diff options
729 files changed, 28539 insertions, 23767 deletions
diff --git a/Documentation/feature-removal-schedule.txt b/Documentation/feature-removal-schedule.txt index bf0e3df8e7a1..164c89394cff 100644 --- a/Documentation/feature-removal-schedule.txt +++ b/Documentation/feature-removal-schedule.txt @@ -212,7 +212,7 @@ Who: Stephen Hemminger <shemminger@linux-foundation.org> --------------------------- What: i386/x86_64 bzImage symlinks -When: April 2008 +When: April 2010 Why: The i386/x86_64 merge provides a symlink to the old bzImage location so not yet updated user space tools, e.g. package diff --git a/Documentation/i386/boot.txt b/Documentation/i386/boot.txt index fc49b79bc1ab..2eb16100bb3f 100644 --- a/Documentation/i386/boot.txt +++ b/Documentation/i386/boot.txt @@ -170,6 +170,8 @@ Offset Proto Name Meaning 0238/4 2.06+ cmdline_size Maximum size of the kernel command line 023C/4 2.07+ hardware_subarch Hardware subarchitecture 0240/8 2.07+ hardware_subarch_data Subarchitecture-specific data +0248/4 2.08+ payload_offset Offset of kernel payload +024C/4 2.08+ payload_length Length of kernel payload (1) For backwards compatibility, if the setup_sects field contains 0, the real value is 4. @@ -512,6 +514,32 @@ Protocol: 2.07+ A pointer to data that is specific to hardware subarch +Field name: payload_offset +Type: read +Offset/size: 0x248/4 +Protocol: 2.08+ + + If non-zero then this field contains the offset from the end of the + real-mode code to the payload. + + The payload may be compressed. The format of both the compressed and + uncompressed data should be determined using the standard magic + numbers. Currently only gzip compressed ELF is used. + +Field name: payload_length +Type: read +Offset/size: 0x24c/4 +Protocol: 2.08+ + + The length of the payload. + +**** THE IMAGE CHECKSUM + +From boot protocol version 2.08 onwards the CRC-32 is calculated over +the entire file using the characteristic polynomial 0x04C11DB7 and an +initial remainder of 0xffffffff. The checksum is appended to the +file; therefore the CRC of the file up to the limit specified in the +syssize field of the header is always 0. **** THE KERNEL COMMAND LINE diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt index c3de1767edd4..c867f506366d 100644 --- a/Documentation/kernel-parameters.txt +++ b/Documentation/kernel-parameters.txt @@ -812,6 +812,19 @@ and is between 256 and 4096 characters. It is defined in the file inttest= [IA64] + iommu= [x86] + off + force + noforce + biomerge + panic + nopanic + merge + nomerge + forcesac + soft + + intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option off Disable intel iommu driver. @@ -1139,6 +1152,11 @@ and is between 256 and 4096 characters. It is defined in the file or memmap=0x10000$0x18690000 + memtest= [KNL,X86_64] Enable memtest + Format: <integer> + range: 0,4 : pattern number + default : 0 <disable> + meye.*= [HW] Set MotionEye Camera parameters See Documentation/video4linux/meye.txt. diff --git a/Documentation/s390/s390dbf.txt b/Documentation/s390/s390dbf.txt index 0eb7c58916de..e05420973698 100644 --- a/Documentation/s390/s390dbf.txt +++ b/Documentation/s390/s390dbf.txt @@ -115,6 +115,27 @@ Return Value: Handle for generated debug area Description: Allocates memory for a debug log Must not be called within an interrupt handler +---------------------------------------------------------------------------- +debug_info_t *debug_register_mode(char *name, int pages, int nr_areas, + int buf_size, mode_t mode, uid_t uid, + gid_t gid); + +Parameter: name: Name of debug log (e.g. used for debugfs entry) + pages: Number of pages, which will be allocated per area + nr_areas: Number of debug areas + buf_size: Size of data area in each debug entry + mode: File mode for debugfs files. E.g. S_IRWXUGO + uid: User ID for debugfs files. Currently only 0 is + supported. + gid: Group ID for debugfs files. Currently only 0 is + supported. + +Return Value: Handle for generated debug area + NULL if register failed + +Description: Allocates memory for a debug log + Must not be called within an interrupt handler + --------------------------------------------------------------------------- void debug_unregister (debug_info_t * id); diff --git a/Documentation/x86/pat.txt b/Documentation/x86/pat.txt new file mode 100644 index 000000000000..17965f927c15 --- /dev/null +++ b/Documentation/x86/pat.txt @@ -0,0 +1,100 @@ + +PAT (Page Attribute Table) + +x86 Page Attribute Table (PAT) allows for setting the memory attribute at the +page level granularity. PAT is complementary to the MTRR settings which allows +for setting of memory types over physical address ranges. However, PAT is +more flexible than MTRR due to its capability to set attributes at page level +and also due to the fact that there are no hardware limitations on number of +such attribute settings allowed. Added flexibility comes with guidelines for +not having memory type aliasing for the same physical memory with multiple +virtual addresses. + +PAT allows for different types of memory attributes. The most commonly used +ones that will be supported at this time are Write-back, Uncached, +Write-combined and Uncached Minus. + +There are many different APIs in the kernel that allows setting of memory +attributes at the page level. In order to avoid aliasing, these interfaces +should be used thoughtfully. Below is a table of interfaces available, +their intended usage and their memory attribute relationships. Internally, +these APIs use a reserve_memtype()/free_memtype() interface on the physical +address range to avoid any aliasing. + + +------------------------------------------------------------------- +API | RAM | ACPI,... | Reserved/Holes | +-----------------------|----------|------------|------------------| + | | | | +ioremap | -- | UC | UC | + | | | | +ioremap_cache | -- | WB | WB | + | | | | +ioremap_nocache | -- | UC | UC | + | | | | +ioremap_wc | -- | -- | WC | + | | | | +set_memory_uc | UC | -- | -- | + set_memory_wb | | | | + | | | | +set_memory_wc | WC | -- | -- | + set_memory_wb | | | | + | | | | +pci sysfs resource | -- | -- | UC | + | | | | +pci sysfs resource_wc | -- | -- | WC | + is IORESOURCE_PREFETCH| | | | + | | | | +pci proc | -- | -- | UC | + !PCIIOC_WRITE_COMBINE | | | | + | | | | +pci proc | -- | -- | WC | + PCIIOC_WRITE_COMBINE | | | | + | | | | +/dev/mem | -- | UC | UC | + read-write | | | | + | | | | +/dev/mem | -- | UC | UC | + mmap SYNC flag | | | | + | | | | +/dev/mem | -- | WB/WC/UC | WB/WC/UC | + mmap !SYNC flag | |(from exist-| (from exist- | + and | | ing alias)| ing alias) | + any alias to this area| | | | + | | | | +/dev/mem | -- | WB | WB | + mmap !SYNC flag | | | | + no alias to this area | | | | + and | | | | + MTRR says WB | | | | + | | | | +/dev/mem | -- | -- | UC_MINUS | + mmap !SYNC flag | | | | + no alias to this area | | | | + and | | | | + MTRR says !WB | | | | + | | | | +------------------------------------------------------------------- + +Notes: + +-- in the above table mean "Not suggested usage for the API". Some of the --'s +are strictly enforced by the kernel. Some others are not really enforced +today, but may be enforced in future. + +For ioremap and pci access through /sys or /proc - The actual type returned +can be more restrictive, in case of any existing aliasing for that address. +For example: If there is an existing uncached mapping, a new ioremap_wc can +return uncached mapping in place of write-combine requested. + +set_memory_[uc|wc] and set_memory_wb should be used in pairs, where driver will +first make a region uc or wc and switch it back to wb after use. + +Over time writes to /proc/mtrr will be deprecated in favor of using PAT based +interfaces. Users writing to /proc/mtrr are suggested to use above interfaces. + +Drivers should use ioremap_[uc|wc] to access PCI BARs with [uc|wc] access +types. + +Drivers should use set_memory_[uc|wc] to set access type for RAM ranges. + diff --git a/Documentation/x86_64/boot-options.txt b/Documentation/x86_64/boot-options.txt index 34abae4e9442..b0c7b6c4abda 100644 --- a/Documentation/x86_64/boot-options.txt +++ b/Documentation/x86_64/boot-options.txt @@ -307,3 +307,8 @@ Debugging stuck (default) Miscellaneous + + nogbpages + Do not use GB pages for kernel direct mappings. + gbpages + Use GB pages for kernel direct mappings. diff --git a/arch/alpha/kernel/Makefile b/arch/alpha/kernel/Makefile index dccf05245d4d..ac706c1d7ada 100644 --- a/arch/alpha/kernel/Makefile +++ b/arch/alpha/kernel/Makefile @@ -7,7 +7,7 @@ EXTRA_AFLAGS := $(KBUILD_CFLAGS) EXTRA_CFLAGS := -Werror -Wno-sign-compare obj-y := entry.o traps.o process.o init_task.o osf_sys.o irq.o \ - irq_alpha.o signal.o setup.o ptrace.o time.o semaphore.o \ + irq_alpha.o signal.o setup.o ptrace.o time.o \ alpha_ksyms.o systbls.o err_common.o io.o obj-$(CONFIG_VGA_HOSE) += console.o diff --git a/arch/alpha/kernel/alpha_ksyms.c b/arch/alpha/kernel/alpha_ksyms.c index e9762a33b043..d96e742d4dc2 100644 --- a/arch/alpha/kernel/alpha_ksyms.c +++ b/arch/alpha/kernel/alpha_ksyms.c @@ -77,15 +77,6 @@ EXPORT_SYMBOL(__do_clear_user); EXPORT_SYMBOL(__strncpy_from_user); EXPORT_SYMBOL(__strnlen_user); -/* Semaphore helper functions. */ -EXPORT_SYMBOL(__down_failed); -EXPORT_SYMBOL(__down_failed_interruptible); -EXPORT_SYMBOL(__up_wakeup); -EXPORT_SYMBOL(down); -EXPORT_SYMBOL(down_interruptible); -EXPORT_SYMBOL(down_trylock); -EXPORT_SYMBOL(up); - /* * SMP-specific symbols. */ diff --git a/arch/alpha/kernel/semaphore.c b/arch/alpha/kernel/semaphore.c deleted file mode 100644 index 8d2982aa1b8d..000000000000 --- a/arch/alpha/kernel/semaphore.c +++ /dev/null @@ -1,224 +0,0 @@ -/* - * Alpha semaphore implementation. - * - * (C) Copyright 1996 Linus Torvalds - * (C) Copyright 1999, 2000 Richard Henderson - */ - -#include <linux/errno.h> -#include <linux/sched.h> -#include <linux/init.h> - -/* - * This is basically the PPC semaphore scheme ported to use - * the Alpha ll/sc sequences, so see the PPC code for - * credits. - */ - -/* - * Atomically update sem->count. - * This does the equivalent of the following: - * - * old_count = sem->count; - * tmp = MAX(old_count, 0) + incr; - * sem->count = tmp; - * return old_count; - */ -static inline int __sem_update_count(struct semaphore *sem, int incr) -{ - long old_count, tmp = 0; - - __asm__ __volatile__( - "1: ldl_l %0,%2\n" - " cmovgt %0,%0,%1\n" - " addl %1,%3,%1\n" - " stl_c %1,%2\n" - " beq %1,2f\n" - " mb\n" - ".subsection 2\n" - "2: br 1b\n" - ".previous" - : "=&r" (old_count), "=&r" (tmp), "=m" (sem->count) - : "Ir" (incr), "1" (tmp), "m" (sem->count)); - - return old_count; -} - -/* - * Perform the "down" function. Return zero for semaphore acquired, - * return negative for signalled out of the function. - * - * If called from down, the return is ignored and the wait loop is - * not interruptible. This means that a task waiting on a semaphore - * using "down()" cannot be killed until someone does an "up()" on - * the semaphore. - * - * If called from down_interruptible, the return value gets checked - * upon return. If the return value is negative then the task continues - * with the negative value in the return register (it can be tested by - * the caller). - * - * Either form may be used in conjunction with "up()". - */ - -void __sched -__down_failed(struct semaphore *sem) -{ - struct task_struct *tsk = current; - DECLARE_WAITQUEUE(wait, tsk); - -#ifdef CONFIG_DEBUG_SEMAPHORE - printk("%s(%d): down failed(%p)\n", - tsk->comm, task_pid_nr(tsk), sem); -#endif - - tsk->state = TASK_UNINTERRUPTIBLE; - wmb(); - add_wait_queue_exclusive(&sem->wait, &wait); - - /* - * Try to get the semaphore. If the count is > 0, then we've - * got the semaphore; we decrement count and exit the loop. - * If the count is 0 or negative, we set it to -1, indicating - * that we are asleep, and then sleep. - */ - while (__sem_update_count(sem, -1) <= 0) { - schedule(); - set_task_state(tsk, TASK_UNINTERRUPTIBLE); - } - remove_wait_queue(&sem->wait, &wait); - tsk->state = TASK_RUNNING; - - /* - * If there are any more sleepers, wake one of them up so - * that it can either get the semaphore, or set count to -1 - * indicating that there are still processes sleeping. - */ - wake_up(&sem->wait); - -#ifdef CONFIG_DEBUG_SEMAPHORE - printk("%s(%d): down acquired(%p)\n", - tsk->comm, task_pid_nr(tsk), sem); -#endif -} - -int __sched -__down_failed_interruptible(struct semaphore *sem) -{ - struct task_struct *tsk = current; - DECLARE_WAITQUEUE(wait, tsk); - long ret = 0; - -#ifdef CONFIG_DEBUG_SEMAPHORE - printk("%s(%d): down failed(%p)\n", - tsk->comm, task_pid_nr(tsk), sem); -#endif - - tsk->state = TASK_INTERRUPTIBLE; - wmb(); - add_wait_queue_exclusive(&sem->wait, &wait); - - while (__sem_update_count(sem, -1) <= 0) { - if (signal_pending(current)) { - /* - * A signal is pending - give up trying. - * Set sem->count to 0 if it is negative, - * since we are no longer sleeping. - */ - __sem_update_count(sem, 0); - ret = -EINTR; - break; - } - schedule(); - set_task_state(tsk, TASK_INTERRUPTIBLE); - } - - remove_wait_queue(&sem->wait, &wait); - tsk->state = TASK_RUNNING; - wake_up(&sem->wait); - -#ifdef CONFIG_DEBUG_SEMAPHORE - printk("%s(%d): down %s(%p)\n", - current->comm, task_pid_nr(current), - (ret < 0 ? "interrupted" : "acquired"), sem); -#endif - return ret; -} - -void -__up_wakeup(struct semaphore *sem) -{ - /* - * Note that we incremented count in up() before we came here, - * but that was ineffective since the result was <= 0, and - * any negative value of count is equivalent to 0. - * This ends up setting count to 1, unless count is now > 0 - * (i.e. because some other cpu has called up() in the meantime), - * in which case we just increment count. - */ - __sem_update_count(sem, 1); - wake_up(&sem->wait); -} - -void __sched -down(struct semaphore *sem) -{ -#ifdef WAITQUEUE_DEBUG - CHECK_MAGIC(sem->__magic); -#endif -#ifdef CONFIG_DEBUG_SEMAPHORE - printk("%s(%d): down(%p) <count=%d> from %p\n", - current->comm, task_pid_nr(current), sem, - atomic_read(&sem->count), __builtin_return_address(0)); -#endif - __down(sem); -} - -int __sched -down_interruptible(struct semaphore *sem) -{ -#ifdef WAITQUEUE_DEBUG - CHECK_MAGIC(sem->__magic); -#endif -#ifdef CONFIG_DEBUG_SEMAPHORE - printk("%s(%d): down(%p) <count=%d> from %p\n", - current->comm, task_pid_nr(current), sem, - atomic_read(&sem->count), __builtin_return_address(0)); -#endif - return __down_interruptible(sem); -} - -int -down_trylock(struct semaphore *sem) -{ - int ret; - -#ifdef WAITQUEUE_DEBUG - CHECK_MAGIC(sem->__magic); -#endif - - ret = __down_trylock(sem); - -#ifdef CONFIG_DEBUG_SEMAPHORE - printk("%s(%d): down_trylock %s from %p\n", - current->comm, task_pid_nr(current), - ret ? "failed" : "acquired", - __builtin_return_address(0)); -#endif - - return ret; -} - -void -up(struct semaphore *sem) -{ -#ifdef WAITQUEUE_DEBUG - CHECK_MAGIC(sem->__magic); -#endif -#ifdef CONFIG_DEBUG_SEMAPHORE - printk("%s(%d): up(%p) <count=%d> from %p\n", - current->comm, task_pid_nr(current), sem, - atomic_read(&sem->count), __builtin_return_address(0)); -#endif - __up(sem); -} diff --git a/arch/arm/kernel/Makefile b/arch/arm/kernel/Makefile index 00d44c6fbfe9..6235f72a14f0 100644 --- a/arch/arm/kernel/Makefile +++ b/arch/arm/kernel/Makefile @@ -7,7 +7,7 @@ AFLAGS_head.o := -DTEXT_OFFSET=$(TEXT_OFFSET) # Object file lists. obj-y := compat.o entry-armv.o entry-common.o irq.o \ - process.o ptrace.o semaphore.o setup.o signal.o \ + process.o ptrace.o setup.o signal.o \ sys_arm.o stacktrace.o time.o traps.o obj-$(CONFIG_ISA_DMA_API) += dma.o diff --git a/arch/arm/kernel/semaphore.c b/arch/arm/kernel/semaphore.c deleted file mode 100644 index 981fe5c6ccbe..000000000000 --- a/arch/arm/kernel/semaphore.c +++ /dev/null @@ -1,221 +0,0 @@ -/* - * ARM semaphore implementation, taken from - * - * i386 semaphore implementation. - * - * (C) Copyright 1999 Linus Torvalds - * - * Modified for ARM by Russell King - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License version 2 as - * published by the Free Software Foundation. - */ -#include <linux/module.h> -#include <linux/sched.h> -#include <linux/errno.h> -#include <linux/init.h> - -#include <asm/semaphore.h> - -/* - * Semaphores are implemented using a two-way counter: - * The "count" variable is decremented for each process - * that tries to acquire the semaphore, while the "sleeping" - * variable is a count of such acquires. - * - * Notably, the inline "up()" and "down()" functions can - * efficiently test if they need to do any extra work (up - * needs to do something only if count was negative before - * the increment operation. - * - * "sleeping" and the contention routine ordering is - * protected by the semaphore spinlock. - * - * Note that these functions are only called when there is - * contention on the lock, and as such all this is the - * "non-critical" part of the whole semaphore business. The - * critical part is the inline stuff in <asm/semaphore.h> - * where we want to avoid any extra jumps and calls. - */ - -/* - * Logic: - * - only on a boundary condition do we need to care. When we go - * from a negative count to a non-negative, we wake people up. - * - when we go from a non-negative count to a negative do we - * (a) synchronize with the "sleeper" count and (b) make sure - * that we're on the wakeup list before we synchronize so that - * we cannot lose wakeup events. - */ - -void __up(struct semaphore *sem) -{ - wake_up(&sem->wait); -} - -static DEFINE_SPINLOCK(semaphore_lock); - -void __sched __down(struct semaphore * sem) -{ - struct task_struct *tsk = current; - DECLARE_WAITQUEUE(wait, tsk); - tsk->state = TASK_UNINTERRUPTIBLE; - add_wait_queue_exclusive(&sem->wait, &wait); - - spin_lock_irq(&semaphore_lock); - sem->sleepers++; - for (;;) { - int sleepers = sem->sleepers; - - /* - * Add "everybody else" into it. They aren't - * playing, because we own the spinlock. - */ - if (!atomic_add_negative(sleepers - 1, &sem->count)) { - sem->sleepers = 0; - break; - } - sem->sleepers = 1; /* us - see -1 above */ - spin_unlock_irq(&semaphore_lock); - - schedule(); - tsk->state = TASK_UNINTERRUPTIBLE; - spin_lock_irq(&semaphore_lock); - } - spin_unlock_irq(&semaphore_lock); - remove_wait_queue(&sem->wait, &wait); - tsk->state = TASK_RUNNING; - wake_up(&sem->wait); -} - -int __sched __down_interruptible(struct semaphore * sem) -{ - int retval = 0; - struct task_struct *tsk = current; - DECLARE_WAITQUEUE(wait, tsk); - tsk->state = TASK_INTERRUPTIBLE; - add_wait_queue_exclusive(&sem->wait, &wait); - - spin_lock_irq(&semaphore_lock); - sem->sleepers ++; - for (;;) { - int sleepers = sem->sleepers; - - /* - * With signals pending, this turns into - * the trylock failure case - we won't be - * sleeping, and we* can't get the lock as - * it has contention. Just correct the count - * and exit. - */ - if (signal_pending(current)) { - retval = -EINTR; - sem->sleepers = 0; - atomic_add(sleepers, &sem->count); - break; - } - - /* - * Add "everybody else" into it. They aren't - * playing, because we own the spinlock. The - * "-1" is because we're still hoping to get - * the lock. - */ - if (!atomic_add_negative(sleepers - 1, &sem->count)) { - sem->sleepers = 0; - break; - } - sem->sleepers = 1; /* us - see -1 above */ - spin_unlock_irq(&semaphore_lock); - - schedule(); - tsk->state = TASK_INTERRUPTIBLE; - spin_lock_irq(&semaphore_lock); - } - spin_unlock_irq(&semaphore_lock); - tsk->state = TASK_RUNNING; - remove_wait_queue(&sem->wait, &wait); - wake_up(&sem->wait); - return retval; -} - -/* - * Trylock failed - make sure we correct for - * having decremented the count. - * - * We could have done the trylock with a - * single "cmpxchg" without failure cases, - * but then it wouldn't work on a 386. - */ -int __down_trylock(struct semaphore * sem) -{ - int sleepers; - unsigned long flags; - - spin_lock_irqsave(&semaphore_lock, flags); - sleepers = sem->sleepers + 1; - sem->sleepers = 0; - - /* - * Add "everybody else" and us into it. They aren't - * playing, because we own the spinlock. - */ - if (!atomic_add_negative(sleepers, &sem->count)) - wake_up(&sem->wait); - - spin_unlock_irqrestore(&semaphore_lock, flags); - return 1; -} - -/* - * The semaphore operations have a special calling sequence that - * allow us to do a simpler in-line version of them. These routines - * need to convert that sequence back into the C sequence when - * there is contention on the semaphore. - * - * ip contains the semaphore pointer on entry. Save the C-clobbered - * registers (r0 to r3 and lr), but not ip, as we use it as a return - * value in some cases.. - * To remain AAPCS compliant (64-bit stack align) we save r4 as well. - */ -asm(" .section .sched.text,\"ax\",%progbits \n\ - .align 5 \n\ - .globl __down_failed \n\ -__down_failed: \n\ - stmfd sp!, {r0 - r4, lr} \n\ - mov r0, ip \n\ - bl __down \n\ - ldmfd sp!, {r0 - r4, pc} \n\ - \n\ - .align 5 \n\ - .globl __down_interruptible_failed \n\ -__down_interruptible_failed: \n\ - stmfd sp!, {r0 - r4, lr} \n\ - mov r0, ip \n\ - bl __down_interruptible \n\ - mov ip, r0 \n\ - ldmfd sp!, {r0 - r4, pc} \n\ - \n\ - .align 5 \n\ - .globl __down_trylock_failed \n\ -__down_trylock_failed: \n\ - stmfd sp!, {r0 - r4, lr} \n\ - mov r0, ip \n\ - bl __down_trylock \n\ - mov ip, r0 \n\ - ldmfd sp!, {r0 - r4, pc} \n\ - \n\ - .align 5 \n\ - .globl __up_wakeup \n\ -__up_wakeup: \n\ - stmfd sp!, {r0 - r4, lr} \n\ - mov r0, ip \n\ - bl __up \n\ - ldmfd sp!, {r0 - r4, pc} \n\ - "); - -EXPORT_SYMBOL(__down_failed); -EXPORT_SYMBOL(__down_interruptible_failed); -EXPORT_SYMBOL(__down_trylock_failed); -EXPORT_SYMBOL(__up_wakeup); diff --git a/arch/avr32/kernel/Makefile b/arch/avr32/kernel/Makefile index e4b6d122b033..18229d0d1861 100644 --- a/arch/avr32/kernel/Makefile +++ b/arch/avr32/kernel/Makefile @@ -6,7 +6,7 @@ extra-y := head.o vmlinux.lds obj-$(CONFIG_SUBARCH_AVR32B) += entry-avr32b.o obj-y += syscall_table.o syscall-stubs.o irq.o -obj-y += setup.o traps.o semaphore.o ocd.o ptrace.o +obj-y += setup.o traps.o ocd.o ptrace.o obj-y += signal.o sys_avr32.o process.o time.o obj-y += init_task.o switch_to.o cpu.o obj-$(CONFIG_MODULES) += module.o avr32_ksyms.o diff --git a/arch/avr32/kernel/semaphore.c b/arch/avr32/kernel/semaphore.c deleted file mode 100644 index 1e2705a05016..000000000000 --- a/arch/avr32/kernel/semaphore.c +++ /dev/null @@ -1,148 +0,0 @@ -/* - * AVR32 sempahore implementation. - * - * Copyright (C) 2004-2006 Atmel Corporation - * - * Based on linux/arch/i386/kernel/semaphore.c - * Copyright (C) 1999 Linus Torvalds - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License version 2 as - * published by the Free Software Foundation. - */ - -#include <linux/sched.h> -#include <linux/errno.h> -#include <linux/module.h> - -#include <asm/semaphore.h> -#include <asm/atomic.h> - -/* - * Semaphores are implemented using a two-way counter: - * The "count" variable is decremented for each process - * that tries to acquire the semaphore, while the "sleeping" - * variable is a count of such acquires. - * - * Notably, the inline "up()" and "down()" functions can - * efficiently test if they need to do any extra work (up - * needs to do something only if count was negative before - * the increment operation. - * - * "sleeping" and the contention routine ordering is protected - * by the spinlock in the semaphore's waitqueue head. - * - * Note that these functions are only called when there is - * contention on the lock, and as such all this is the - * "non-critical" part of the whole semaphore business. The - * critical part is the inline stuff in <asm/semaphore.h> - * where we want to avoid any extra jumps and calls. - */ - -/* - * Logic: - * - only on a boundary condition do we need to care. When we go - * from a negative count to a non-negative, we wake people up. - * - when we go from a non-negative count to a negative do we - * (a) synchronize with the "sleeper" count and (b) make sure - * that we're on the wakeup list before we synchronize so that - * we cannot lose wakeup events. - */ - -void __up(struct semaphore *sem) -{ - wake_up(&sem->wait); -} -EXPORT_SYMBOL(__up); - -void __sched __down(struct semaphore *sem) -{ - struct task_struct *tsk = current; - DECLARE_WAITQUEUE(wait, tsk); - unsigned long flags; - - tsk->state = TASK_UNINTERRUPTIBLE; - spin_lock_irqsave(&sem->wait.lock, flags); - add_wait_queue_exclusive_locked(&sem->wait, &wait); - - sem->sleepers++; - for (;;) { - int sleepers = sem->sleepers; - - /* - * Add "everybody else" into it. They aren't - * playing, because we own the spinlock in - * the wait_queue_head. - */ - if (atomic_add_return(sleepers - 1, &sem->count) >= 0) { - sem->sleepers = 0; - break; - } - sem->sleepers = 1; /* us - see -1 above */ - spin_unlock_irqrestore(&sem->wait.lock, flags); - - schedule(); - - spin_lock_irqsave(&sem->wait.lock, flags); - tsk->state = TASK_UNINTERRUPTIBLE; - } - remove_wait_queue_locked(&sem->wait, &wait); - wake_up_locked(&sem->wait); - spin_unlock_irqrestore(&sem->wait.lock, flags); - tsk->state = TASK_RUNNING; -} -EXPORT_SYMBOL(__down); - -int __sched __down_interruptible(struct semaphore *sem) -{ - int retval = 0; - struct task_struct *tsk = current; - DECLARE_WAITQUEUE(wait, tsk); - unsigned long flags; - - tsk->state = TASK_INTERRUPTIBLE; - spin_lock_irqsave(&sem->wait.lock, flags); - add_wait_queue_exclusive_locked(&sem->wait, &wait); - - sem->sleepers++; - for (;;) { - int sleepers = sem->sleepers; - - /* - * With signals pending, this turns into the trylock - * failure case - we won't be sleeping, and we can't - * get the lock as it has contention. Just correct the - * count and exit. - */ - if (signal_pending(current)) { - retval = -EINTR; - sem->sleepers = 0; - atomic_add(sleepers, &sem->count); - break; - } - - /* - * Add "everybody else" into it. They aren't - * playing, because we own the spinlock in - * the wait_queue_head. - */ - if (atomic_add_return(sleepers - 1, &sem->count) >= 0) { - sem->sleepers = 0; - break; - } - sem->sleepers = 1; /* us - see -1 above */ - spin_unlock_irqrestore(&sem->wait.lock, flags); - - schedule(); - - spin_lock_irqsave(&sem->wait.lock, flags); - tsk->state = TASK_INTERRUPTIBLE; - } - remove_wait_queue_locked(&sem->wait, &wait); - wake_up_locked(&sem->wait); - spin_unlock_irqrestore(&sem->wait.lock, flags); - - tsk->state = TASK_RUNNING; - return retval; -} -EXPORT_SYMBOL(__down_interruptible); diff --git a/arch/blackfin/Kconfig b/arch/blackfin/Kconfig index 589c6aca4803..2dd1f300a5cf 100644 --- a/arch/blackfin/Kconfig +++ b/arch/blackfin/Kconfig @@ -31,10 +31,6 @@ config ZONE_DMA bool default y -config SEMAPHORE_SLEEPERS - bool - default y - config GENERIC_FIND_NEXT_BIT bool default y diff --git a/arch/blackfin/kernel/bfin_ksyms.c b/arch/blackfin/kernel/bfin_ksyms.c index 0bfbb269e350..053edff6c0d8 100644 --- a/arch/blackfin/kernel/bfin_ksyms.c +++ b/arch/blackfin/kernel/bfin_ksyms.c @@ -42,11 +42,6 @@ EXPORT_SYMBOL(ip_fast_csum); EXPORT_SYMBOL(kernel_thread); -EXPORT_SYMBOL(__up); -EXPORT_SYMBOL(__down); -EXPORT_SYMBOL(__down_trylock); -EXPORT_SYMBOL(__down_interruptible); - EXPORT_SYMBOL(is_in_rom); EXPORT_SYMBOL(bfin_return_from_exception); diff --git a/arch/cris/kernel/Makefile b/arch/cris/kernel/Makefile index c8e8ea570989..ee7bcd4d20b2 100644 --- a/arch/cris/kernel/Makefile +++ b/arch/cris/kernel/Makefile @@ -5,8 +5,7 @@ extra-y := vmlinux.lds -obj-y := process.o traps.o irq.o ptrace.o setup.o \ - time.o sys_cris.o semaphore.o +obj-y := process.o traps.o irq.o ptrace.o setup.o time.o sys_cris.o obj-$(CONFIG_MODULES) += crisksyms.o obj-$(CONFIG_MODULES) += module.o diff --git a/arch/cris/kernel/crisksyms.c b/arch/cris/kernel/crisksyms.c index 62f0e752915a..7ac000f6a888 100644 --- a/arch/cris/kernel/crisksyms.c +++ b/arch/cris/kernel/crisksyms.c @@ -9,7 +9,6 @@ #include <linux/string.h> #include <linux/tty.h> -#include <asm/semaphore.h> #include <asm/processor.h> #include <asm/uaccess.h> #include <asm/checksum.h> @@ -49,12 +48,6 @@ EXPORT_SYMBOL(__negdi2); EXPORT_SYMBOL(__ioremap); EXPORT_SYMBOL(iounmap); -/* Semaphore functions */ -EXPORT_SYMBOL(__up); -EXPORT_SYMBOL(__down); -EXPORT_SYMBOL(__down_interruptible); -EXPORT_SYMBOL(__down_trylock); - /* Userspace access functions */ EXPORT_SYMBOL(__copy_user_zeroing); EXPORT_SYMBOL(__copy_user); diff --git a/arch/cris/kernel/semaphore.c b/arch/cris/kernel/semaphore.c deleted file mode 100644 index f137a439041f..000000000000 --- a/arch/cris/kernel/semaphore.c +++ /dev/null @@ -1,129 +0,0 @@ -/* - * Generic semaphore code. Buyer beware. Do your own - * specific changes in <asm/semaphore-helper.h> - */ - -#include <linux/sched.h> -#include <asm/semaphore-helper.h> - -/* - * Semaphores are implemented using a two-way counter: - * The "count" variable is decremented for each process - * that tries to sleep, while the "waking" variable is - * incremented when the "up()" code goes to wake up waiting - * processes. - * - * Notably, the inline "up()" and "down()" functions can - * efficiently test if they need to do any extra work (up - * needs to do something only if count was negative before - * the increment operation. - * - * waking_non_zero() (from asm/semaphore.h) must execute - * atomically. - * - * When __up() is called, the count was negative before - * incrementing it, and we need to wake up somebody. - * - * This routine adds one to the count of processes that need to - * wake up and exit. ALL waiting processes actually wake up but - * only the one that gets to the "waking" field first will gate - * through and acquire the semaphore. The others will go back - * to sleep. - * - * Note that these functions are only called when there is - * contention on the lock, and as such all this is the - * "non-critical" part of the whole semaphore business. The - * critical part is the inline stuff in <asm/semaphore.h> - * where we want to avoid any extra jumps and calls. - */ -void __up(struct semaphore *sem) -{ - wake_one_more(sem); - wake_up(&sem->wait); -} - -/* - * Perform the "down" function. Return zero for semaphore acquired, - * return negative for signalled out of the function. - * - * If called from __down, the return is ignored and the wait loop is - * not interruptible. This means that a task waiting on a semaphore - * using "down()" cannot be killed until someone does an "up()" on - * the semaphore. - * - * If called from __down_interruptible, the return value gets checked - * upon return. If the return value is negative then the task continues - * with the negative value in the return register (it can be tested by - * the caller). - * - * Either form may be used in conjunction with "up()". - * - */ - -#define DOWN_VAR \ - struct task_struct *tsk = current; \ - wait_queue_t wait; \ - init_waitqueue_entry(&wait, tsk); - -#define DOWN_HEAD(task_state) \ - \ - \ - tsk->state = (task_state); \ - add_wait_queue(&sem->wait, &wait); \ - \ - /* \ - * Ok, we're set up. sem->count is known to be less than zero \ - * so we must wait. \ - * \ - * We can let go the lock for purposes of waiting. \ - * We re-acquire it after awaking so as to protect \ - * all semaphore operations. \ - * \ - * If "up()" is called before we call waking_non_zero() then \ - * we will catch it right away. If it is called later then \ - * we will have to go through a wakeup cycle to catch it. \ - * \ - * Multiple waiters contend for the semaphore lock to see \ - * who gets to gate through and who has to wait some more. \ - */ \ - for (;;) { - -#define DOWN_TAIL(task_state) \ - tsk->state = (task_state); \ - } \ - tsk->state = TASK_RUNNING; \ - remove_wait_queue(&sem->wait, &wait); - -void __sched __down(struct semaphore * sem) -{ - DOWN_VAR - DOWN_HEAD(TASK_UNINTERRUPTIBLE) - if (waking_non_zero(sem)) - break; - schedule(); - DOWN_TAIL(TASK_UNINTERRUPTIBLE) -} - -int __sched __down_interruptible(struct semaphore * sem) -{ - int ret = 0; - DOWN_VAR - DOWN_HEAD(TASK_INTERRUPTIBLE) - - ret = waking_non_zero_interruptible(sem, tsk); - if (ret) - { - if (ret == 1) - /* ret != 0 only if we get interrupted -arca */ - ret = 0; - break; - } - schedule(); - DOWN_TAIL(TASK_INTERRUPTIBLE) - return ret; -} - -int __down_trylock(struct semaphore * sem) -{ - return waking_non_zero_trylock(sem); -} diff --git a/arch/frv/kernel/Makefile b/arch/frv/kernel/Makefile index e8f73ed28b52..c36f70b6699a 100644 --- a/arch/frv/kernel/Makefile +++ b/arch/frv/kernel/Makefile @@ -9,7 +9,7 @@ extra-y:= head.o init_task.o vmlinux.lds obj-y := $(heads-y) entry.o entry-table.o break.o switch_to.o kernel_thread.o \ kernel_execve.o process.o traps.o ptrace.o signal.o dma.o \ - sys_frv.o time.o semaphore.o setup.o frv_ksyms.o \ + sys_frv.o time.o setup.o frv_ksyms.o \ debug-stub.o irq.o sleep.o uaccess.o obj-$(CONFIG_GDBSTUB) += gdb-stub.o gdb-io.o diff --git a/arch/frv/kernel/frv_ksyms.c b/arch/frv/kernel/frv_ksyms.c index f772704b3d28..0316b3c50eff 100644 --- a/arch/frv/kernel/frv_ksyms.c +++ b/arch/frv/kernel/frv_ksyms.c @@ -12,7 +12,6 @@ #include <asm/pgalloc.h> #include <asm/irq.h> #include <asm/io.h> -#include <asm/semaphore.h> #include <asm/checksum.h> #include <asm/hardirq.h> #include <asm/cacheflush.h> diff --git a/arch/frv/kernel/semaphore.c b/arch/frv/kernel/semaphore.c deleted file mode 100644 index 7ee3a147b471..000000000000 --- a/arch/frv/kernel/semaphore.c +++ /dev/null @@ -1,155 +0,0 @@ -/* semaphore.c: FR-V semaphores - * - * Copyright (C) 2003 Red Hat, Inc. All Rights Reserved. - * Written by David Howells (dhowells@redhat.com) - * - Derived from lib/rwsem-spinlock.c - * - * This program is free software; you can redistribute it and/or - * modify it under the terms of the GNU General Public License - * as published by the Free Software Foundation; either version - * 2 of the License, or (at your option) any later version. - */ - -#include <linux/sched.h> -#include <linux/module.h> -#include <asm/semaphore.h> - -struct sem_waiter { - struct list_head list; - struct task_struct *task; -}; - -#ifdef CONFIG_DEBUG_SEMAPHORE -void semtrace(struct semaphore *sem, const char *str) -{ - if (sem->debug) - printk("[%d] %s({%d,%d})\n", - current->pid, - str, - sem->counter, - list_empty(&sem->wait_list) ? 0 : 1); -} -#else -#define semtrace(SEM,STR) do { } while(0) -#endif - -/* - * wait for a token to be granted from a semaphore - * - entered with lock held and interrupts disabled - */ -void __down(struct semaphore *sem, unsigned long flags) -{ - struct task_struct *tsk = current; - struct sem_waiter waiter; - - semtrace(sem, "Entering __down"); - - /* set up my own style of waitqueue */ - waiter.task = tsk; - get_task_struct(tsk); - - list_add_tail(&waiter.list, &sem->wait_list); - - /* we don't need to touch the semaphore struct anymore */ - spin_unlock_irqrestore(&sem->wait_lock, flags); - - /* wait to be given the semaphore */ - set_task_state(tsk, TASK_UNINTERRUPTIBLE); - - for (;;) { - if (list_empty(&waiter.list)) - break; - schedule(); - set_task_state(tsk, TASK_UNINTERRUPTIBLE); - } - - tsk->state = TASK_RUNNING; - semtrace(sem, "Leaving __down"); -} - -EXPORT_SYMBOL(__down); - -/* - * interruptibly wait for a token to be granted from a semaphore - * - entered with lock held and interrupts disabled - */ -int __down_interruptible(struct semaphore *sem, unsigned long flags) -{ - struct task_struct *tsk = current; - struct sem_waiter waiter; - int ret; - - semtrace(sem,"Entering __down_interruptible"); - - /* set up my own style of waitqueue */ - waiter.task = tsk; - get_task_struct(tsk); - - list_add_tail(&waiter.list, &sem->wait_list); - - /* we don't need to touch the semaphore struct anymore */ - set_task_state(tsk, TASK_INTERRUPTIBLE); - - spin_unlock_irqrestore(&sem->wait_lock, flags); - - /* wait to be given the semaphore */ - ret = 0; - for (;;) { - if (list_empty(&waiter.list)) - break; - if (unlikely(signal_pending(current))) - goto interrupted; - schedule(); - set_task_state(tsk, TASK_INTERRUPTIBLE); - } - - out: - tsk->state = TASK_RUNNING; - semtrace(sem, "Leaving __down_interruptible"); - return ret; - - interrupted: - spin_lock_irqsave(&sem->wait_lock, flags); - - if (!list_empty(&waiter.list)) { - list_del(&waiter.list); - ret = -EINTR; - } - - spin_unlock_irqrestore(&sem->wait_lock, flags); - if (ret == -EINTR) - put_task_struct(current); - goto out; -} - -EXPORT_SYMBOL(__down_interruptible); - -/* - * release a single token back to a semaphore - * - entered with lock held and interrupts disabled - */ -void __up(struct semaphore *sem) -{ - struct task_struct *tsk; - struct sem_waiter *waiter; - - semtrace(sem,"Entering __up"); - - /* grant the token to the process at the front of the queue */ - waiter = list_entry(sem->wait_list.next, struct sem_waiter, list); - - /* We must be careful not to touch 'waiter' after we set ->task = NULL. - * It is allocated on the waiter's stack and may become invalid at - * any time after that point (due to a wakeup from another source). - */ - list_del_init(&waiter->list); - tsk = waiter->task; - mb(); - waiter->task = NULL; - wake_up_process(tsk); - put_task_struct(tsk); - - semtrace(sem,"Leaving __up"); -} - -EXPORT_SYMBOL(__up); diff --git a/arch/h8300/kernel/Makefile b/arch/h8300/kernel/Makefile index 874f6aefee65..6c248c3c5c3b 100644 --- a/arch/h8300/kernel/Makefile +++ b/arch/h8300/kernel/Makefile @@ -5,7 +5,7 @@ extra-y := vmlinux.lds obj-y := process.o traps.o ptrace.o irq.o \ - sys_h8300.o time.o semaphore.o signal.o \ + sys_h8300.o time.o signal.o \ setup.o gpio.o init_task.o syscalls.o \ entry.o diff --git a/arch/h8300/kernel/h8300_ksyms.c b/arch/h8300/kernel/h8300_ksyms.c index d1b15267ac81..6866bd9c7fb4 100644 --- a/arch/h8300/kernel/h8300_ksyms.c +++ b/arch/h8300/kernel/h8300_ksyms.c @@ -12,7 +12,6 @@ #include <asm/pgalloc.h> #include <asm/irq.h> #include <asm/io.h> -#include <asm/semaphore.h> #include <asm/checksum.h> #include <asm/current.h> #include <asm/gpio.h> diff --git a/arch/h8300/kernel/semaphore.c b/arch/h8300/kernel/semaphore.c deleted file mode 100644 index d12cbbfe6ebd..000000000000 --- a/arch/h8300/kernel/semaphore.c +++ /dev/null @@ -1,132 +0,0 @@ -/* - * Generic semaphore code. Buyer beware. Do your own - * specific changes in <asm/semaphore-helper.h> - */ - -#include <linux/sched.h> -#include <linux/init.h> -#include <asm/semaphore-helper.h> - -#ifndef CONFIG_RMW_INSNS -spinlock_t semaphore_wake_lock; -#endif - -/* - * Semaphores are implemented using a two-way counter: - * The "count" variable is decremented for each process - * that tries to sleep, while the "waking" variable is - * incremented when the "up()" code goes to wake up waiting - * processes. - * - * Notably, the inline "up()" and "down()" functions can - * efficiently test if they need to do any extra work (up - * needs to do something only if count was negative before - * the increment operation. - * - * waking_non_zero() (from asm/semaphore.h) must execute - * atomically. - * - * When __up() is called, the count was negative before - * incrementing it, and we need to wake up somebody. - * - * This routine adds one to the count of processes that need to - * wake up and exit. ALL waiting processes actually wake up but - * only the one that gets to the "waking" field first will gate - * through and acquire the semaphore. The others will go back - * to sleep. - * - * Note that these functions are only called when there is - * contention on the lock, and as such all this is the - * "non-critical" part of the whole semaphore business. The - * critical part is the inline stuff in <asm/semaphore.h> - * where we want to avoid any extra jumps and calls. - */ -void __up(struct semaphore *sem) -{ - wake_one_more(sem); - wake_up(&sem->wait); -} - -/* - * Perform the "down" function. Return zero for semaphore acquired, - * return negative for signalled out of the function. - * - * If called from __down, the return is ignored and the wait loop is - * not interruptible. This means that a task waiting on a semaphore - * using "down()" cannot be killed until someone does an "up()" on - * the semaphore. - * - * If called from __down_interruptible, the return value gets checked - * upon return. If the return value is negative then the task continues - * with the negative value in the return register (it can be tested by - * the caller). - * - * Either form may be used in conjunction with "up()". - * - */ - - -#define DOWN_HEAD(task_state) \ - \ - \ - current->state = (task_state); \ - add_wait_queue(&sem->wait, &wait); \ - \ - /* \ - * Ok, we're set up. sem->count is known to be less than zero \ - * so we must wait. \ - * \ - * We can let go the lock for purposes of waiting. \ - * We re-acquire it after awaking so as to protect \ - * all semaphore operations. \ - * \ - * If "up()" is called before we call waking_non_zero() then \ - * we will catch it right away. If it is called later then \ - * we will have to go through a wakeup cycle to catch it. \ - * \ - * Multiple waiters contend for the semaphore lock to see \ - * who gets to gate through and who has to wait some more. \ - */ \ - for (;;) { - -#define DOWN_TAIL(task_state) \ - current->state = (task_state); \ - } \ - current->state = TASK_RUNNING; \ - remove_wait_queue(&sem->wait, &wait); - -void __sched __down(struct semaphore * sem) -{ - DECLARE_WAITQUEUE(wait, current); - - DOWN_HEAD(TASK_UNINTERRUPTIBLE) - if (waking_non_zero(sem)) - break; - schedule(); - DOWN_TAIL(TASK_UNINTERRUPTIBLE) -} - -int __sched __down_interruptible(struct semaphore * sem) -{ - DECLARE_WAITQUEUE(wait, current); - int ret = 0; - - DOWN_HEAD(TASK_INTERRUPTIBLE) - - ret = waking_non_zero_interruptible(sem, current); - if (ret) - { - if (ret == 1) - /* ret != 0 only if we get interrupted -arca */ - ret = 0; - break; - } - schedule(); - DOWN_TAIL(TASK_INTERRUPTIBLE) - return ret; -} - -int __down_trylock(struct semaphore * sem) -{ - return waking_non_zero_trylock(sem); -} diff --git a/arch/ia64/kernel/Makefile b/arch/ia64/kernel/Makefile index 33e5a598672d..13fd10e8699e 100644 --- a/arch/ia64/kernel/Makefile +++ b/arch/ia64/kernel/Makefile @@ -6,7 +6,7 @@ extra-y := head.o init_task.o vmlinux.lds obj-y := acpi.o entry.o efi.o efi_stub.o gate-data.o fsys.o ia64_ksyms.o irq.o irq_ia64.o \ irq_lsapic.o ivt.o machvec.o pal.o patch.o process.o perfmon.o ptrace.o sal.o \ - salinfo.o semaphore.o setup.o signal.o sys_ia64.o time.o traps.o unaligned.o \ + salinfo.o setup.o signal.o sys_ia64.o time.o traps.o unaligned.o \ unwind.o mca.o mca_asm.o topology.o obj-$(CONFIG_IA64_BRL_EMU) += brl_emu.o diff --git a/arch/ia64/kernel/ia64_ksyms.c b/arch/ia64/kernel/ia64_ksyms.c index 8e7193d55528..6da1f20d7372 100644 --- a/arch/ia64/kernel/ia64_ksyms.c +++ b/arch/ia64/kernel/ia64_ksyms.c @@ -19,12 +19,6 @@ EXPORT_SYMBOL_GPL(empty_zero_page); EXPORT_SYMBOL(ip_fast_csum); /* hand-coded assembly */ EXPORT_SYMBOL(csum_ipv6_magic); -#include <asm/semaphore.h> -EXPORT_SYMBOL(__down); -EXPORT_SYMBOL(__down_interruptible); -EXPORT_SYMBOL(__down_trylock); -EXPORT_SYMBOL(__up); - #include <asm/page.h> EXPORT_SYMBOL(clear_page); diff --git a/arch/ia64/kernel/semaphore.c b/arch/ia64/kernel/semaphore.c deleted file mode 100644 index 2724ef3fbae2..000000000000 --- a/arch/ia64/kernel/semaphore.c +++ /dev/null @@ -1,165 +0,0 @@ -/* - * IA-64 semaphore implementation (derived from x86 version). - * - * Copyright (C) 1999-2000, 2002 Hewlett-Packard Co - * David Mosberger-Tang <davidm@hpl.hp.com> - */ - -/* - * Semaphores are implemented using a two-way counter: The "count" - * variable is decremented for each process that tries to acquire the - * semaphore, while the "sleepers" variable is a count of such - * acquires. - * - * Notably, the inline "up()" and "down()" functions can efficiently - * test if they need to do any extra work (up needs to do something - * only if count was negative before the increment operation. - * - * "sleeping" and the contention routine ordering is protected - * by the spinlock in the semaphore's waitqueue head. - * - * Note that these functions are only called when there is contention - * on the lock, and as such all this is the "non-critical" part of the - * whole semaphore business. The critical part is the inline stuff in - * <asm/semaphore.h> where we want to avoid any extra jumps and calls. - */ -#include <linux/sched.h> -#include <linux/init.h> - -#include <asm/errno.h> -#include <asm/semaphore.h> - -/* - * Logic: - * - Only on a boundary condition do we need to care. When we go - * from a negative count to a non-negative, we wake people up. - * - When we go from a non-negative count to a negative do we - * (a) synchronize with the "sleepers" count and (b) make sure - * that we're on the wakeup list before we synchronize so that - * we cannot lose wakeup events. - */ - -void -__up (struct semaphore *sem) -{ - wake_up(&sem->wait); -} - -void __sched __down (struct semaphore *sem) -{ - struct task_struct *tsk = current; - DECLARE_WAITQUEUE(wait, tsk); - unsigned long flags; - - tsk->state = TASK_UNINTERRUPTIBLE; - spin_lock_irqsave(&sem->wait.lock, flags); - add_wait_queue_exclusive_locked(&sem->wait, &wait); - - sem->sleepers++; - for (;;) { - int sleepers = sem->sleepers; - - /* - * Add "everybody else" into it. They aren't - * playing, because we own the spinlock in - * the wait_queue_head. - */ - if (!atomic_add_negative(sleepers - 1, &sem->count)) { - sem->sleepers = 0; - break; - } - sem->sleepers = 1; /* us - see -1 above */ - spin_unlock_irqrestore(&sem->wait.lock, flags); - - schedule(); - - spin_lock_irqsave(&sem->wait.lock, flags); - tsk->state = TASK_UNINTERRUPTIBLE; - } - remove_wait_queue_locked(&sem->wait, &wait); - wake_up_locked(&sem->wait); - spin_unlock_irqrestore(&sem->wait.lock, flags); - tsk->state = TASK_RUNNING; -} - -int __sched __down_interruptible (struct semaphore * sem) -{ - int retval = 0; - struct task_struct *tsk = current; - DECLARE_WAITQUEUE(wait, tsk); - unsigned long flags; - - tsk->state = TASK_INTERRUPTIBLE; - spin_lock_irqsave(&sem->wait.lock, flags); - add_wait_queue_exclusive_locked(&sem->wait, &wait); - - sem->sleepers ++; - for (;;) { - int sleepers = sem->sleepers; - - /* - * With signals pending, this turns into - * the trylock failure case - we won't be - * sleeping, and we* can't get the lock as - * it has contention. Just correct the count - * and exit. - */ - if (signal_pending(current)) { - retval = -EINTR; - sem->sleepers = 0; - atomic_add(sleepers, &sem->count); - break; - } - - /* - * Add "everybody else" into it. They aren't - * playing, because we own the spinlock in - * wait_queue_head. The "-1" is because we're - * still hoping to get the semaphore. - */ - if (!atomic_add_negative(sleepers - 1, &sem->count)) { - sem->sleepers = 0; - break; - } - sem->sleepers = 1; /* us - see -1 above */ - spin_unlock_irqrestore(&sem->wait.lock, flags); - - schedule(); - - spin_lock_irqsave(&sem->wait.lock, flags); - tsk->state = TASK_INTERRUPTIBLE; - } - remove_wait_queue_locked(&sem->wait, &wait); - wake_up_locked(&sem->wait); - spin_unlock_irqrestore(&sem->wait.lock, flags); - - tsk->state = TASK_RUNNING; - return retval; -} - -/* - * Trylock failed - make sure we correct for having decremented the - * count. - */ -int -__down_trylock (struct semaphore *sem) -{ - unsigned long flags; - int sleepers; - - spin_lock_irqsave(&sem->wait.lock, flags); - sleepers = sem->sleepers + 1; - sem->sleepers = 0; - - /* - * Add "everybody else" and us into it. They aren't - * playing, because we own the spinlock in the - * wait_queue_head. - */ - if (!atomic_add_negative(sleepers, &sem->count)) { - wake_up_locked(&sem->wait); - } - - spin_unlock_irqrestore(&sem->wait.lock, flags); - return 1; -} diff --git a/arch/m32r/kernel/Makefile b/arch/m32r/kernel/Makefile index e97e26e87c9e..09200d4886e3 100644 --- a/arch/m32r/kernel/Makefile +++ b/arch/m32r/kernel/Makefile @@ -5,7 +5,7 @@ extra-y := head.o init_task.o vmlinux.lds obj-y := process.o entry.o traps.o align.o irq.o setup.o time.o \ - m32r_ksyms.o sys_m32r.o semaphore.o signal.o ptrace.o + m32r_ksyms.o sys_m32r.o signal.o ptrace.o obj-$(CONFIG_SMP) += smp.o smpboot.o obj-$(CONFIG_MODULES) += module.o diff --git a/arch/m32r/kernel/m32r_ksyms.c b/arch/m32r/kernel/m32r_ksyms.c index 41a4c95e06d6..e6709fe950ba 100644 --- a/arch/m32r/kernel/m32r_ksyms.c +++ b/arch/m32r/kernel/m32r_ksyms.c @@ -7,7 +7,6 @@ #include <linux/interrupt.h> #include <linux/string.h> -#include <asm/semaphore.h> #include <asm/processor.h> #include <asm/uaccess.h> #include <asm/checksum.h> @@ -22,10 +21,6 @@ EXPORT_SYMBOL(dump_fpu); EXPORT_SYMBOL(__ioremap); EXPORT_SYMBOL(iounmap); EXPORT_SYMBOL(kernel_thread); -EXPORT_SYMBOL(__down); -EXPORT_SYMBOL(__down_interruptible); -EXPORT_SYMBOL(__up); -EXPORT_SYMBOL(__down_trylock); /* Networking helper routines. */ /* Delay loops */ diff --git a/arch/m32r/kernel/semaphore.c b/arch/m32r/kernel/semaphore.c deleted file mode 100644 index 940c2d37cfd1..000000000000 --- a/arch/m32r/kernel/semaphore.c +++ /dev/null @@ -1,185 +0,0 @@ -/* - * linux/arch/m32r/semaphore.c - * orig : i386 2.6.4 - * - * M32R semaphore implementation. - * - * Copyright (c) 2002 - 2004 Hitoshi Yamamoto - */ - -/* - * i386 semaphore implementation. - * - * (C) Copyright 1999 Linus Torvalds - * - * Portions Copyright 1999 Red Hat, Inc. - * - * This program is free software; you can redistribute it and/or - * modify it under the terms of the GNU General Public License - * as published by the Free Software Foundation; either version - * 2 of the License, or (at your option) any later version. - * - * rw semaphores implemented November 1999 by Benjamin LaHaise <bcrl@kvack.org> - */ -#include <linux/sched.h> -#include <linux/err.h> -#include <linux/init.h> -#include <asm/semaphore.h> - -/* - * Semaphores are implemented using a two-way counter: - * The "count" variable is decremented for each process - * that tries to acquire the semaphore, while the "sleeping" - * variable is a count of such acquires. - * - * Notably, the inline "up()" and "down()" functions can - * efficiently test if they need to do any extra work (up - * needs to do something only if count was negative before - * the increment operation. - * - * "sleeping" and the contention routine ordering is protected - * by the spinlock in the semaphore's waitqueue head. - * - * Note that these functions are only called when there is - * contention on the lock, and as such all this is the - * "non-critical" part of the whole semaphore business. The - * critical part is the inline stuff in <asm/semaphore.h> - * where we want to avoid any extra jumps and calls. - */ - -/* - * Logic: - * - only on a boundary condition do we need to care. When we go - * from a negative count to a non-negative, we wake people up. - * - when we go from a non-negative count to a negative do we - * (a) synchronize with the "sleeper" count and (b) make sure - * that we're on the wakeup list before we synchronize so that - * we cannot lose wakeup events. - */ - -asmlinkage void __up(struct semaphore *sem) -{ - wake_up(&sem->wait); -} - -asmlinkage void __sched __down(struct semaphore * sem) -{ - struct task_struct *tsk = current; - DECLARE_WAITQUEUE(wait, tsk); - unsigned long flags; - - tsk->state = TASK_UNINTERRUPTIBLE; - spin_lock_irqsave(&sem->wait.lock, flags); - add_wait_queue_exclusive_locked(&sem->wait, &wait); - - sem->sleepers++; - for (;;) { - int sleepers = sem->sleepers; - - /* - * Add "everybody else" into it. They aren't - * playing, because we own the spinlock in - * the wait_queue_head. - */ - if (!atomic_add_negative(sleepers - 1, &sem->count)) { - sem->sleepers = 0; - break; - } - sem->sleepers = 1; /* us - see -1 above */ - spin_unlock_irqrestore(&sem->wait.lock, flags); - - schedule(); - - spin_lock_irqsave(&sem->wait.lock, flags); - tsk->state = TASK_UNINTERRUPTIBLE; - } - remove_wait_queue_locked(&sem->wait, &wait); - wake_up_locked(&sem->wait); - spin_unlock_irqrestore(&sem->wait.lock, flags); - tsk->state = TASK_RUNNING; -} - -asmlinkage int __sched __down_interruptible(struct semaphore * sem) -{ - int retval = 0; - struct task_struct *tsk = current; - DECLARE_WAITQUEUE(wait, tsk); - unsigned long flags; - - tsk->state = TASK_INTERRUPTIBLE; - spin_lock_irqsave(&sem->wait.lock, flags); - add_wait_queue_exclusive_locked(&sem->wait, &wait); - - sem->sleepers++; - for (;;) { - int sleepers = sem->sleepers; - - /* - * With signals pending, this turns into - * the trylock failure case - we won't be - * sleeping, and we* can't get the lock as - * it has contention. Just correct the count - * and exit. - */ - if (signal_pending(current)) { - retval = -EINTR; - sem->sleepers = 0; - atomic_add(sleepers, &sem->count); - break; - } - - /* - * Add "everybody else" into it. They aren't - * playing, because we own the spinlock in - * wait_queue_head. The "-1" is because we're - * still hoping to get the semaphore. - */ - if (!atomic_add_negative(sleepers - 1, &sem->count)) { - sem->sleepers = 0; - break; - } - sem->sleepers = 1; /* us - see -1 above */ - spin_unlock_irqrestore(&sem->wait.lock, flags); - - schedule(); - - spin_lock_irqsave(&sem->wait.lock, flags); - tsk->state = TASK_INTERRUPTIBLE; - } - remove_wait_queue_locked(&sem->wait, &wait); - wake_up_locked(&sem->wait); - spin_unlock_irqrestore(&sem->wait.lock, flags); - - tsk->state = TASK_RUNNING; - return retval; -} - -/* - * Trylock failed - make sure we correct for - * having decremented the count. - * - * We could have done the trylock with a - * single "cmpxchg" without failure cases, - * but then it wouldn't work on a 386. - */ -asmlinkage int __down_trylock(struct semaphore * sem) -{ - int sleepers; - unsigned long flags; - - spin_lock_irqsave(&sem->wait.lock, flags); - sleepers = sem->sleepers + 1; - sem->sleepers = 0; - - /* - * Add "everybody else" and us into it. They aren't - * playing, because we own the spinlock in the - * wait_queue_head. - */ - if (!atomic_add_negative(sleepers, &sem->count)) { - wake_up_locked(&sem->wait); - } - - spin_unlock_irqrestore(&sem->wait.lock, flags); - return 1; -} diff --git a/arch/m68k/kernel/Makefile b/arch/m68k/kernel/Makefile index a806208c7fb5..7a62a718143b 100644 --- a/arch/m68k/kernel/Makefile +++ b/arch/m68k/kernel/Makefile @@ -10,7 +10,7 @@ endif extra-y += vmlinux.lds obj-y := entry.o process.o traps.o ints.o signal.o ptrace.o module.o \ - sys_m68k.o time.o semaphore.o setup.o m68k_ksyms.o devres.o + sys_m68k.o time.o setup.o m68k_ksyms.o devres.o devres-y = ../../../kernel/irq/devres.o diff --git a/arch/m68k/kernel/m68k_ksyms.c b/arch/m68k/kernel/m68k_ksyms.c index 6fc69c74fe2e..d900e77e5363 100644 --- a/arch/m68k/kernel/m68k_ksyms.c +++ b/arch/m68k/kernel/m68k_ksyms.c @@ -1,5 +1,4 @@ #include <linux/module.h> -#include <asm/semaphore.h> asmlinkage long long __ashldi3 (long long, int); asmlinkage long long __ashrdi3 (long long, int); @@ -15,8 +14,3 @@ EXPORT_SYMBOL(__ashrdi3); EXPORT_SYMBOL(__lshrdi3); EXPORT_SYMBOL(__muldi3); -EXPORT_SYMBOL(__down_failed); -EXPORT_SYMBOL(__down_failed_interruptible); -EXPORT_SYMBOL(__down_failed_trylock); -EXPORT_SYMBOL(__up_wakeup); - diff --git a/arch/m68k/kernel/semaphore.c b/arch/m68k/kernel/semaphore.c deleted file mode 100644 index d12cbbfe6ebd..000000000000 --- a/arch/m68k/kernel/semaphore.c +++ /dev/null @@ -1,132 +0,0 @@ -/* - * Generic semaphore code. Buyer beware. Do your own - * specific changes in <asm/semaphore-helper.h> - */ - -#include <linux/sched.h> -#include <linux/init.h> -#include <asm/semaphore-helper.h> - -#ifndef CONFIG_RMW_INSNS -spinlock_t semaphore_wake_lock; -#endif - -/* - * Semaphores are implemented using a two-way counter: - * The "count" variable is decremented for each process - * that tries to sleep, while the "waking" variable is - * incremented when the "up()" code goes to wake up waiting - * processes. - * - * Notably, the inline "up()" and "down()" functions can - * efficiently test if they need to do any extra work (up - * needs to do something only if count was negative before - * the increment operation. - * - * waking_non_zero() (from asm/semaphore.h) must execute - * atomically. - * - * When __up() is called, the count was negative before - * incrementing it, and we need to wake up somebody. - * - * This routine adds one to the count of processes that need to - * wake up and exit. ALL waiting processes actually wake up but - * only the one that gets to the "waking" field first will gate - * through and acquire the semaphore. The others will go back - * to sleep. - * - * Note that these functions are only called when there is - * contention on the lock, and as such all this is the - * "non-critical" part of the whole semaphore business. The - * critical part is the inline stuff in <asm/semaphore.h> - * where we want to avoid any extra jumps and calls. - */ -void __up(struct semaphore *sem) -{ - wake_one_more(sem); - wake_up(&sem->wait); -} - -/* - * Perform the "down" function. Return zero for semaphore acquired, - * return negative for signalled out of the function. - * - * If called from __down, the return is ignored and the wait loop is - * not interruptible. This means that a task waiting on a semaphore - * using "down()" cannot be killed until someone does an "up()" on - * the semaphore. - * - * If called from __down_interruptible, the return value gets checked - * upon return. If the return value is negative then the task continues - * with the negative value in the return register (it can be tested by - * the caller). - * - * Either form may be used in conjunction with "up()". - * - */ - - -#define DOWN_HEAD(task_state) \ - \ - \ - current->state = (task_state); \ - add_wait_queue(&sem->wait, &wait); \ - \ - /* \ - * Ok, we're set up. sem->count is known to be less than zero \ - * so we must wait. \ - * \ - * We can let go the lock for purposes of waiting. \ - * We re-acquire it after awaking so as to protect \ - * all semaphore operations. \ - * \ - * If "up()" is called before we call waking_non_zero() then \ - * we will catch it right away. If it is called later then \ - * we will have to go through a wakeup cycle to catch it. \ - * \ - * Multiple waiters contend for the semaphore lock to see \ - * who gets to gate through and who has to wait some more. \ - */ \ - for (;;) { - -#define DOWN_TAIL(task_state) \ - current->state = (task_state); \ - } \ - current->state = TASK_RUNNING; \ - remove_wait_queue(&sem->wait, &wait); - -void __sched __down(struct semaphore * sem) -{ - DECLARE_WAITQUEUE(wait, current); - - DOWN_HEAD(TASK_UNINTERRUPTIBLE) - if (waking_non_zero(sem)) - break; - schedule(); - DOWN_TAIL(TASK_UNINTERRUPTIBLE) -} - -int __sched __down_interruptible(struct semaphore * sem) -{ - DECLARE_WAITQUEUE(wait, current); - int ret = 0; - - DOWN_HEAD(TASK_INTERRUPTIBLE) - - ret = waking_non_zero_interruptible(sem, current); - if (ret) - { - if (ret == 1) - /* ret != 0 only if we get interrupted -arca */ - ret = 0; - break; - } - schedule(); - DOWN_TAIL(TASK_INTERRUPTIBLE) - return ret; -} - -int __down_trylock(struct semaphore * sem) -{ - return waking_non_zero_trylock(sem); -} diff --git a/arch/m68k/lib/Makefile b/arch/m68k/lib/Makefile index 6bbf19f96007..a18af095cd7c 100644 --- a/arch/m68k/lib/Makefile +++ b/arch/m68k/lib/Makefile @@ -5,4 +5,4 @@ EXTRA_AFLAGS := -traditional lib-y := ashldi3.o ashrdi3.o lshrdi3.o muldi3.o \ - checksum.o string.o semaphore.o uaccess.o + checksum.o string.o uaccess.o diff --git a/arch/m68k/lib/semaphore.S b/arch/m68k/lib/semaphore.S deleted file mode 100644 index 0215624c1602..000000000000 --- a/arch/m68k/lib/semaphore.S +++ /dev/null @@ -1,53 +0,0 @@ -/* - * linux/arch/m68k/lib/semaphore.S - * - * Copyright (C) 1996 Linus Torvalds - * - * m68k version by Andreas Schwab - */ - -#include <linux/linkage.h> -#include <asm/semaphore.h> - -/* - * The semaphore operations have a special calling sequence that - * allow us to do a simpler in-line version of them. These routines - * need to convert that sequence back into the C sequence when - * there is contention on the semaphore. - */ -ENTRY(__down_failed) - moveml %a0/%d0/%d1,-(%sp) - movel %a1,-(%sp) - jbsr __down - movel (%sp)+,%a1 - moveml (%sp)+,%a0/%d0/%d1 - rts - -ENTRY(__down_failed_interruptible) - movel %a0,-(%sp) - movel %d1,-(%sp) - movel %a1,-(%sp) - jbsr __down_interruptible - movel (%sp)+,%a1 - movel (%sp)+,%d1 - movel (%sp)+,%a0 - rts - -ENTRY(__down_failed_trylock) - movel %a0,-(%sp) - movel %d1,-(%sp) - movel %a1,-(%sp) - jbsr __down_trylock - movel (%sp)+,%a1 - movel (%sp)+,%d1 - movel (%sp)+,%a0 - rts - -ENTRY(__up_wakeup) - moveml %a0/%d0/%d1,-(%sp) - movel %a1,-(%sp) - jbsr __up - movel (%sp)+,%a1 - moveml (%sp)+,%a0/%d0/%d1 - rts - diff --git a/arch/m68knommu/kernel/Makefile b/arch/m68knommu/kernel/Makefile index 1524b39ad63f..f0eab3dedb5a 100644 --- a/arch/m68knommu/kernel/Makefile +++ b/arch/m68knommu/kernel/Makefile @@ -5,7 +5,7 @@ extra-y := vmlinux.lds obj-y += dma.o entry.o init_task.o irq.o m68k_ksyms.o process.o ptrace.o \ - semaphore.o setup.o signal.o syscalltable.o sys_m68k.o time.o traps.o + setup.o signal.o syscalltable.o sys_m68k.o time.o traps.o obj-$(CONFIG_MODULES) += module.o obj-$(CONFIG_COMEMPCI) += comempci.o diff --git a/arch/m68knommu/kernel/m68k_ksyms.c b/arch/m68knommu/kernel/m68k_ksyms.c index 53fad1490282..39fe0a7aec32 100644 --- a/arch/m68knommu/kernel/m68k_ksyms.c +++ b/arch/m68knommu/kernel/m68k_ksyms.c @@ -13,7 +13,6 @@ #include <asm/pgalloc.h> #include <asm/irq.h> #include <asm/io.h> -#include <asm/semaphore.h> #include <asm/checksum.h> #include <asm/current.h> @@ -39,11 +38,6 @@ EXPORT_SYMBOL(csum_partial_copy_nocheck); EXPORT_SYMBOL(memcpy); EXPORT_SYMBOL(memset); -EXPORT_SYMBOL(__down_failed); -EXPORT_SYMBOL(__down_failed_interruptible); -EXPORT_SYMBOL(__down_failed_trylock); -EXPORT_SYMBOL(__up_wakeup); - /* * libgcc functions - functions that are used internally by the * compiler... (prototypes are not correct though, but that diff --git a/arch/m68knommu/kernel/semaphore.c b/arch/m68knommu/kernel/semaphore.c deleted file mode 100644 index bce2bc7d87c6..000000000000 --- a/arch/m68knommu/kernel/semaphore.c +++ /dev/null @@ -1,133 +0,0 @@ -/* - * Generic semaphore code. Buyer beware. Do your own - * specific changes in <asm/semaphore-helper.h> - */ - -#include <linux/sched.h> -#include <linux/err.h> -#include <linux/init.h> -#include <asm/semaphore-helper.h> - -#ifndef CONFIG_RMW_INSNS -spinlock_t semaphore_wake_lock; -#endif - -/* - * Semaphores are implemented using a two-way counter: - * The "count" variable is decremented for each process - * that tries to sleep, while the "waking" variable is - * incremented when the "up()" code goes to wake up waiting - * processes. - * - * Notably, the inline "up()" and "down()" functions can - * efficiently test if they need to do any extra work (up - * needs to do something only if count was negative before - * the increment operation. - * - * waking_non_zero() (from asm/semaphore.h) must execute - * atomically. - * - * When __up() is called, the count was negative before - * incrementing it, and we need to wake up somebody. - * - * This routine adds one to the count of processes that need to - * wake up and exit. ALL waiting processes actually wake up but - * only the one that gets to the "waking" field first will gate - * through and acquire the semaphore. The others will go back - * to sleep. - * - * Note that these functions are only called when there is - * contention on the lock, and as such all this is the - * "non-critical" part of the whole semaphore business. The - * critical part is the inline stuff in <asm/semaphore.h> - * where we want to avoid any extra jumps and calls. - */ -void __up(struct semaphore *sem) -{ - wake_one_more(sem); - wake_up(&sem->wait); -} - -/* - * Perform the "down" function. Return zero for semaphore acquired, - * return negative for signalled out of the function. - * - * If called from __down, the return is ignored and the wait loop is - * not interruptible. This means that a task waiting on a semaphore - * using "down()" cannot be killed until someone does an "up()" on - * the semaphore. - * - * If called from __down_interruptible, the return value gets checked - * upon return. If the return value is negative then the task continues - * with the negative value in the return register (it can be tested by - * the caller). - * - * Either form may be used in conjunction with "up()". - * - */ - - -#define DOWN_HEAD(task_state) \ - \ - \ - current->state = (task_state); \ - add_wait_queue(&sem->wait, &wait); \ - \ - /* \ - * Ok, we're set up. sem->count is known to be less than zero \ - * so we must wait. \ - * \ - * We can let go the lock for purposes of waiting. \ - * We re-acquire it after awaking so as to protect \ - * all semaphore operations. \ - * \ - * If "up()" is called before we call waking_non_zero() then \ - * we will catch it right away. If it is called later then \ - * we will have to go through a wakeup cycle to catch it. \ - * \ - * Multiple waiters contend for the semaphore lock to see \ - * who gets to gate through and who has to wait some more. \ - */ \ - for (;;) { - -#define DOWN_TAIL(task_state) \ - current->state = (task_state); \ - } \ - current->state = TASK_RUNNING; \ - remove_wait_queue(&sem->wait, &wait); - -void __sched __down(struct semaphore * sem) -{ - DECLARE_WAITQUEUE(wait, current); - - DOWN_HEAD(TASK_UNINTERRUPTIBLE) - if (waking_non_zero(sem)) - break; - schedule(); - DOWN_TAIL(TASK_UNINTERRUPTIBLE) -} - -int __sched __down_interruptible(struct semaphore * sem) -{ - DECLARE_WAITQUEUE(wait, current); - int ret = 0; - - DOWN_HEAD(TASK_INTERRUPTIBLE) - - ret = waking_non_zero_interruptible(sem, current); - if (ret) - { - if (ret == 1) - /* ret != 0 only if we get interrupted -arca */ - ret = 0; - break; - } - schedule(); - DOWN_TAIL(TASK_INTERRUPTIBLE) - return ret; -} - -int __down_trylock(struct semaphore * sem) -{ - return waking_non_zero_trylock(sem); -} diff --git a/arch/m68knommu/lib/Makefile b/arch/m68knommu/lib/Makefile index e051a7913987..d94d709665aa 100644 --- a/arch/m68knommu/lib/Makefile +++ b/arch/m68knommu/lib/Makefile @@ -4,4 +4,4 @@ lib-y := ashldi3.o ashrdi3.o lshrdi3.o \ muldi3.o mulsi3.o divsi3.o udivsi3.o modsi3.o umodsi3.o \ - checksum.o semaphore.o memcpy.o memset.o delay.o + checksum.o memcpy.o memset.o delay.o diff --git a/arch/m68knommu/lib/semaphore.S b/arch/m68knommu/lib/semaphore.S deleted file mode 100644 index 87c746034376..000000000000 --- a/arch/m68knommu/lib/semaphore.S +++ /dev/null @@ -1,66 +0,0 @@ -/* - * linux/arch/m68k/lib/semaphore.S - * - * Copyright (C) 1996 Linus Torvalds - * - * m68k version by Andreas Schwab - * - * MAR/1999 -- modified to support ColdFire (gerg@snapgear.com) - */ - -#include <linux/linkage.h> -#include <asm/semaphore.h> - -/* - * "down_failed" is called with the eventual return address - * in %a0, and the address of the semaphore in %a1. We need - * to increment the number of waiters on the semaphore, - * call "__down()", and then eventually return to try again. - */ -ENTRY(__down_failed) -#ifdef CONFIG_COLDFIRE - subl #12,%sp - moveml %a0/%d0/%d1,(%sp) -#else - moveml %a0/%d0/%d1,-(%sp) -#endif - movel %a1,-(%sp) - jbsr __down - movel (%sp)+,%a1 - movel (%sp)+,%d0 - movel (%sp)+,%d1 - rts - -ENTRY(__down_failed_interruptible) - movel %a0,-(%sp) - movel %d1,-(%sp) - movel %a1,-(%sp) - jbsr __down_interruptible - movel (%sp)+,%a1 - movel (%sp)+,%d1 - rts - -ENTRY(__up_wakeup) -#ifdef CONFIG_COLDFIRE - subl #12,%sp - moveml %a0/%d0/%d1,(%sp) -#else - moveml %a0/%d0/%d1,-(%sp) -#endif - movel %a1,-(%sp) - jbsr __up - movel (%sp)+,%a1 - movel (%sp)+,%d0 - movel (%sp)+,%d1 - rts - -ENTRY(__down_failed_trylock) - movel %a0,-(%sp) - movel %d1,-(%sp) - movel %a1,-(%sp) - jbsr __down_trylock - movel (%sp)+,%a1 - movel (%sp)+,%d1 - movel (%sp)+,%a0 - rts - diff --git a/arch/mips/kernel/Makefile b/arch/mips/kernel/Makefile index 9e78e1a4ca17..6fcdb6fda2e2 100644 --- a/arch/mips/kernel/Makefile +++ b/arch/mips/kernel/Makefile @@ -5,7 +5,7 @@ extra-y := head.o init_task.o vmlinux.lds obj-y += cpu-probe.o branch.o entry.o genex.o irq.o process.o \ - ptrace.o reset.o semaphore.o setup.o signal.o syscall.o \ + ptrace.o reset.o setup.o signal.o syscall.o \ time.o topology.o traps.o unaligned.o obj-$(CONFIG_CEVT_BCM1480) += cevt-bcm1480.o diff --git a/arch/mips/kernel/semaphore.c b/arch/mips/kernel/semaphore.c deleted file mode 100644 index 1265358cdca1..000000000000 --- a/arch/mips/kernel/semaphore.c +++ /dev/null @@ -1,168 +0,0 @@ -/* - * MIPS-specific semaphore code. - * - * Copyright (C) 1999 Cort Dougan <cort@cs.nmt.edu> - * Copyright (C) 2004 Ralf Baechle <ralf@linux-mips.org> - * - * This program is free software; you can redistribute it and/or - * modify it under the terms of the GNU General Public License - * as published by the Free Software Foundation; either version - * 2 of the License, or (at your option) any later version. - * - * April 2001 - Reworked by Paul Mackerras <paulus@samba.org> - * to eliminate the SMP races in the old version between the updates - * of `count' and `waking'. Now we use negative `count' values to - * indicate that some process(es) are waiting for the semaphore. - */ - -#include <linux/module.h> -#include <linux/sched.h> -#include <linux/init.h> -#include <asm/atomic.h> -#include <asm/cpu-features.h> -#include <asm/errno.h> -#include <asm/semaphore.h> -#include <asm/war.h> -/* - * Atomically update sem->count. - * This does the equivalent of the following: - * - * old_count = sem->count; - * tmp = MAX(old_count, 0) + incr; - * sem->count = tmp; - * return old_count; - * - * On machines without lld/scd we need a spinlock to make the manipulation of - * sem->count and sem->waking atomic. Scalability isn't an issue because - * this lock is used on UP only so it's just an empty variable. - */ -static inline int __sem_update_count(struct semaphore *sem, int incr) -{ - int old_count, tmp; - - if (cpu_has_llsc && R10000_LLSC_WAR) { - __asm__ __volatile__( - " .set mips3 \n" - "1: ll %0, %2 # __sem_update_count \n" - " sra %1, %0, 31 \n" - " not %1 \n" - " and %1, %0, %1 \n" - " addu %1, %1, %3 \n" - " sc %1, %2 \n" - " beqzl %1, 1b \n" - " .set mips0 \n" - : "=&r" (old_count), "=&r" (tmp), "=m" (sem->count) - : "r" (incr), "m" (sem->count)); - } else if (cpu_has_llsc) { - __asm__ __volatile__( - " .set mips3 \n" - "1: ll %0, %2 # __sem_update_count \n" - " sra %1, %0, 31 \n" - " not %1 \n" - " and %1, %0, %1 \n" - " addu %1, %1, %3 \n" - " sc %1, %2 \n" - " beqz %1, 1b \n" - " .set mips0 \n" - : "=&r" (old_count), "=&r" (tmp), "=m" (sem->count) - : "r" (incr), "m" (sem->count)); - } else { - static DEFINE_SPINLOCK(semaphore_lock); - unsigned long flags; - - spin_lock_irqsave(&semaphore_lock, flags); - old_count = atomic_read(&sem->count); - tmp = max_t(int, old_count, 0) + incr; - atomic_set(&sem->count, tmp); - spin_unlock_irqrestore(&semaphore_lock, flags); - } - - return old_count; -} - -void __up(struct semaphore *sem) -{ - /* - * Note that we incremented count in up() before we came here, - * but that was ineffective since the result was <= 0, and - * any negative value of count is equivalent to 0. - * This ends up setting count to 1, unless count is now > 0 - * (i.e. because some other cpu has called up() in the meantime), - * in which case we just increment count. - */ - __sem_update_count(sem, 1); - wake_up(&sem->wait); -} - -EXPORT_SYMBOL(__up); - -/* - * Note that when we come in to __down or __down_interruptible, - * we have already decremented count, but that decrement was - * ineffective since the result was < 0, and any negative value - * of count is equivalent to 0. - * Thus it is only when we decrement count from some value > 0 - * that we have actually got the semaphore. - */ -void __sched __down(struct semaphore *sem) -{ - struct task_struct *tsk = current; - DECLARE_WAITQUEUE(wait, tsk); - - __set_task_state(tsk, TASK_UNINTERRUPTIBLE); - add_wait_queue_exclusive(&sem->wait, &wait); - - /* - * Try to get the semaphore. If the count is > 0, then we've - * got the semaphore; we decrement count and exit the loop. - * If the count is 0 or negative, we set it to -1, indicating - * that we are asleep, and then sleep. - */ - while (__sem_update_count(sem, -1) <= 0) { - schedule(); - set_task_state(tsk, TASK_UNINTERRUPTIBLE); - } - remove_wait_queue(&sem->wait, &wait); - __set_task_state(tsk, TASK_RUNNING); - - /* - * If there are any more sleepers, wake one of them up so - * that it can either get the semaphore, or set count to -1 - * indicating that there are still processes sleeping. - */ - wake_up(&sem->wait); -} - -EXPORT_SYMBOL(__down); - -int __sched __down_interruptible(struct semaphore * sem) -{ - int retval = 0; - struct task_struct *tsk = current; - DECLARE_WAITQUEUE(wait, tsk); - - __set_task_state(tsk, TASK_INTERRUPTIBLE); - add_wait_queue_exclusive(&sem->wait, &wait); - - while (__sem_update_count(sem, -1) <= 0) { - if (signal_pending(current)) { - /* - * A signal is pending - give up trying. - * Set sem->count to 0 if it is negative, - * since we are no longer sleeping. - */ - __sem_update_count(sem, 0); - retval = -EINTR; - break; - } - schedule(); - set_task_state(tsk, TASK_INTERRUPTIBLE); - } - remove_wait_queue(&sem->wait, &wait); - __set_task_state(tsk, TASK_RUNNING); - - wake_up(&sem->wait); - return retval; -} - -EXPORT_SYMBOL(__down_interruptible); diff --git a/arch/mn10300/kernel/Makefile b/arch/mn10300/kernel/Makefile index ef07c956170a..23f2ab67574c 100644 --- a/arch/mn10300/kernel/Makefile +++ b/arch/mn10300/kernel/Makefile @@ -3,7 +3,7 @@ # extra-y := head.o init_task.o vmlinux.lds -obj-y := process.o semaphore.o signal.o entry.o fpu.o traps.o irq.o \ +obj-y := process.o signal.o entry.o fpu.o traps.o irq.o \ ptrace.o setup.o time.o sys_mn10300.o io.o kthread.o \ switch_to.o mn10300_ksyms.o kernel_execve.o diff --git a/arch/mn10300/kernel/semaphore.c b/arch/mn10300/kernel/semaphore.c deleted file mode 100644 index 9153c4039fd2..000000000000 --- a/arch/mn10300/kernel/semaphore.c +++ /dev/null @@ -1,149 +0,0 @@ -/* MN10300 Semaphore implementation - * - * Copyright (C) 2007 Red Hat, Inc. All Rights Reserved. - * Written by David Howells (dhowells@redhat.com) - * - * This program is free software; you can redistribute it and/or - * modify it under the terms of the GNU General Public Licence - * as published by the Free Software Foundation; either version - * 2 of the Licence, or (at your option) any later version. - */ -#include <linux/sched.h> -#include <linux/module.h> -#include <asm/semaphore.h> - -struct sem_waiter { - struct list_head list; - struct task_struct *task; -}; - -#if SEMAPHORE_DEBUG -void semtrace(struct semaphore *sem, const char *str) -{ - if (sem->debug) - printk(KERN_DEBUG "[%d] %s({%d,%d})\n", - current->pid, - str, - atomic_read(&sem->count), - list_empty(&sem->wait_list) ? 0 : 1); -} -#else -#define semtrace(SEM, STR) do { } while (0) -#endif - -/* - * wait for a token to be granted from a semaphore - * - entered with lock held and interrupts disabled - */ -void __down(struct semaphore *sem, unsigned long flags) -{ - struct task_struct *tsk = current; - struct sem_waiter waiter; - - semtrace(sem, "Entering __down"); - - /* set up my own style of waitqueue */ - waiter.task = tsk; - get_task_struct(tsk); - - list_add_tail(&waiter.list, &sem->wait_list); - - /* we don't need to touch the semaphore struct anymore */ - spin_unlock_irqrestore(&sem->wait_lock, flags); - - /* wait to be given the semaphore */ - set_task_state(tsk, TASK_UNINTERRUPTIBLE); - - for (;;) { - if (!waiter.task) - break; - schedule(); - set_task_state(tsk, TASK_UNINTERRUPTIBLE); - } - - tsk->state = TASK_RUNNING; - semtrace(sem, "Leaving __down"); -} -EXPORT_SYMBOL(__down); - -/* - * interruptibly wait for a token to be granted from a semaphore - * - entered with lock held and interrupts disabled - */ -int __down_interruptible(struct semaphore *sem, unsigned long flags) -{ - struct task_struct *tsk = current; - struct sem_waiter waiter; - int ret; - - semtrace(sem, "Entering __down_interruptible"); - - /* set up my own style of waitqueue */ - waiter.task = tsk; - get_task_struct(tsk); - - list_add_tail(&waiter.list, &sem->wait_list); - - /* we don't need to touch the semaphore struct anymore */ - set_task_state(tsk, TASK_INTERRUPTIBLE); - - spin_unlock_irqrestore(&sem->wait_lock, flags); - - /* wait to be given the semaphore */ - ret = 0; - for (;;) { - if (!waiter.task) - break; - if (unlikely(signal_pending(current))) - goto interrupted; - schedule(); - set_task_state(tsk, TASK_INTERRUPTIBLE); - } - - out: - tsk->state = TASK_RUNNING; - semtrace(sem, "Leaving __down_interruptible"); - return ret; - - interrupted: - spin_lock_irqsave(&sem->wait_lock, flags); - list_del(&waiter.list); - spin_unlock_irqrestore(&sem->wait_lock, flags); - - ret = 0; - if (!waiter.task) { - put_task_struct(current); - ret = -EINTR; - } - goto out; -} -EXPORT_SYMBOL(__down_interruptible); - -/* - * release a single token back to a semaphore - * - entered with lock held and interrupts disabled - */ -void __up(struct semaphore *sem) -{ - struct task_struct *tsk; - struct sem_waiter *waiter; - - semtrace(sem, "Entering __up"); - - /* grant the token to the process at the front of the queue */ - waiter = list_entry(sem->wait_list.next, struct sem_waiter, list); - - /* We must be careful not to touch 'waiter' after we set ->task = NULL. - * It is an allocated on the waiter's stack and may become invalid at - * any time after that point (due to a wakeup from another source). - */ - list_del_init(&waiter->list); - tsk = waiter->task; - smp_mb(); - waiter->task = NULL; - wake_up_process(tsk); - put_task_struct(tsk); - - semtrace(sem, "Leaving __up"); -} -EXPORT_SYMBOL(__up); diff --git a/arch/parisc/kernel/Makefile b/arch/parisc/kernel/Makefile index 27827bc3717e..1f6585a56f97 100644 --- a/arch/parisc/kernel/Makefile +++ b/arch/parisc/kernel/Makefile @@ -9,7 +9,7 @@ AFLAGS_pacache.o := -traditional obj-y := cache.o pacache.o setup.o traps.o time.o irq.o \ pa7300lc.o syscall.o entry.o sys_parisc.o firmware.o \ - ptrace.o hardware.o inventory.o drivers.o semaphore.o \ + ptrace.o hardware.o inventory.o drivers.o \ signal.o hpmc.o real2.o parisc_ksyms.o unaligned.o \ process.o processor.o pdc_cons.o pdc_chassis.o unwind.o \ topology.o diff --git a/arch/parisc/kernel/parisc_ksyms.c b/arch/parisc/kernel/parisc_ksyms.c index 7aca704e96f0..5b7fc4aa044d 100644 --- a/arch/parisc/kernel/parisc_ksyms.c +++ b/arch/parisc/kernel/parisc_ksyms.c @@ -69,11 +69,6 @@ EXPORT_SYMBOL(memcpy_toio); EXPORT_SYMBOL(memcpy_fromio); EXPORT_SYMBOL(memset_io); -#include <asm/semaphore.h> -EXPORT_SYMBOL(__up); -EXPORT_SYMBOL(__down_interruptible); -EXPORT_SYMBOL(__down); - extern void $$divI(void); extern void $$divU(void); extern void $$remI(void); diff --git a/arch/parisc/kernel/semaphore.c b/arch/parisc/kernel/semaphore.c deleted file mode 100644 index ee806bcc3726..000000000000 --- a/arch/parisc/kernel/semaphore.c +++ /dev/null @@ -1,102 +0,0 @@ -/* - * Semaphore implementation Copyright (c) 2001 Matthew Wilcox, Hewlett-Packard - */ - -#include <linux/sched.h> -#include <linux/spinlock.h> -#include <linux/errno.h> -#include <linux/init.h> - -/* - * Semaphores are complex as we wish to avoid using two variables. - * `count' has multiple roles, depending on its value. If it is positive - * or zero, there are no waiters. The functions here will never be - * called; see <asm/semaphore.h> - * - * When count is -1 it indicates there is at least one task waiting - * for the semaphore. - * - * When count is less than that, there are '- count - 1' wakeups - * pending. ie if it has value -3, there are 2 wakeups pending. - * - * Note that these functions are only called when there is contention - * on the lock, and as such all this is the "non-critical" part of the - * whole semaphore business. The critical part is the inline stuff in - * <asm/semaphore.h> where we want to avoid any extra jumps and calls. - */ -void __up(struct semaphore *sem) -{ - sem->count--; - wake_up(&sem->wait); -} - -#define wakers(count) (-1 - count) - -#define DOWN_HEAD \ - int ret = 0; \ - DECLARE_WAITQUEUE(wait, current); \ - \ - /* Note that someone is waiting */ \ - if (sem->count == 0) \ - sem->count = -1; \ - \ - /* protected by the sentry still -- use unlocked version */ \ - wait.flags = WQ_FLAG_EXCLUSIVE; \ - __add_wait_queue_tail(&sem->wait, &wait); \ - lost_race: \ - spin_unlock_irq(&sem->sentry); \ - -#define DOWN_TAIL \ - spin_lock_irq(&sem->sentry); \ - if (wakers(sem->count) == 0 && ret == 0) \ - goto lost_race; /* Someone stole our wakeup */ \ - __remove_wait_queue(&sem->wait, &wait); \ - current->state = TASK_RUNNING; \ - if (!waitqueue_active(&sem->wait) && (sem->count < 0)) \ - sem->count = wakers(sem->count); - -#define UPDATE_COUNT \ - sem->count += (sem->count < 0) ? 1 : - 1; - - -void __sched __down(struct semaphore * sem) -{ - DOWN_HEAD - - for(;;) { - set_task_state(current, TASK_UNINTERRUPTIBLE); - /* we can _read_ this without the sentry */ - if (sem->count != -1) - break; - schedule(); - } - - DOWN_TAIL - UPDATE_COUNT -} - -int __sched __down_interruptible(struct semaphore * sem) -{ - DOWN_HEAD - - for(;;) { - set_task_state(current, TASK_INTERRUPTIBLE); - /* we can _read_ this without the sentry */ - if (sem->count != -1) - break; - - if (signal_pending(current)) { - ret = -EINTR; - break; - } - schedule(); - } - - DOWN_TAIL - - if (!ret) { - UPDATE_COUNT - } - - return ret; -} diff --git a/arch/powerpc/kernel/Makefile b/arch/powerpc/kernel/Makefile index c1baf9d5903f..b9dbfff9afe9 100644 --- a/arch/powerpc/kernel/Makefile +++ b/arch/powerpc/kernel/Makefile @@ -12,7 +12,7 @@ CFLAGS_prom_init.o += -fPIC CFLAGS_btext.o += -fPIC endif -obj-y := semaphore.o cputable.o ptrace.o syscalls.o \ +obj-y := cputable.o ptrace.o syscalls.o \ irq.o align.o signal_32.o pmc.o vdso.o \ init_task.o process.o systbl.o idle.o \ signal.o diff --git a/arch/powerpc/kernel/ppc_ksyms.c b/arch/powerpc/kernel/ppc_ksyms.c index 9c98424277a8..65d14e6ddc3c 100644 --- a/arch/powerpc/kernel/ppc_ksyms.c +++ b/arch/powerpc/kernel/ppc_ksyms.c @@ -15,7 +15,6 @@ #include <linux/bitops.h> #include <asm/page.h> -#include <asm/semaphore.h> #include <asm/processor.h> #include <asm/cacheflush.h> #include <asm/uaccess.h> diff --git a/arch/powerpc/kernel/semaphore.c b/arch/powerpc/kernel/semaphore.c deleted file mode 100644 index 2f8c3c951394..000000000000 --- a/arch/powerpc/kernel/semaphore.c +++ /dev/null @@ -1,135 +0,0 @@ -/* - * PowerPC-specific semaphore code. - * - * Copyright (C) 1999 Cort Dougan <cort@cs.nmt.edu> - * - * This program is free software; you can redistribute it and/or - * modify it under the terms of the GNU General Public License - * as published by the Free Software Foundation; either version - * 2 of the License, or (at your option) any later version. - * - * April 2001 - Reworked by Paul Mackerras <paulus@samba.org> - * to eliminate the SMP races in the old version between the updates - * of `count' and `waking'. Now we use negative `count' values to - * indicate that some process(es) are waiting for the semaphore. - */ - -#include <linux/sched.h> -#include <linux/init.h> -#include <linux/module.h> - -#include <asm/atomic.h> -#include <asm/semaphore.h> -#include <asm/errno.h> - -/* - * Atomically update sem->count. - * This does the equivalent of the following: - * - * old_count = sem->count; - * tmp = MAX(old_count, 0) + incr; - * sem->count = tmp; - * return old_count; - */ -static inline int __sem_update_count(struct semaphore *sem, int incr) -{ - int old_count, tmp; - - __asm__ __volatile__("\n" -"1: lwarx %0,0,%3\n" -" srawi %1,%0,31\n" -" andc %1,%0,%1\n" -" add %1,%1,%4\n" - PPC405_ERR77(0,%3) -" stwcx. %1,0,%3\n" -" bne 1b" - : "=&r" (old_count), "=&r" (tmp), "=m" (sem->count) - : "r" (&sem->count), "r" (incr), "m" (sem->count) - : "cc"); - - return old_count; -} - -void __up(struct semaphore *sem) -{ - /* - * Note that we incremented count in up() before we came here, - * but that was ineffective since the result was <= 0, and - * any negative value of count is equivalent to 0. - * This ends up setting count to 1, unless count is now > 0 - * (i.e. because some other cpu has called up() in the meantime), - * in which case we just increment count. - */ - __sem_update_count(sem, 1); - wake_up(&sem->wait); -} -EXPORT_SYMBOL(__up); - -/* - * Note that when we come in to __down or __down_interruptible, - * we have already decremented count, but that decrement was - * ineffective since the result was < 0, and any negative value - * of count is equivalent to 0. - * Thus it is only when we decrement count from some value > 0 - * that we have actually got the semaphore. - */ -void __sched __down(struct semaphore *sem) -{ - struct task_struct *tsk = current; - DECLARE_WAITQUEUE(wait, tsk); - - __set_task_state(tsk, TASK_UNINTERRUPTIBLE); - add_wait_queue_exclusive(&sem->wait, &wait); - - /* - * Try to get the semaphore. If the count is > 0, then we've - * got the semaphore; we decrement count and exit the loop. - * If the count is 0 or negative, we set it to -1, indicating - * that we are asleep, and then sleep. - */ - while (__sem_update_count(sem, -1) <= 0) { - schedule(); - set_task_state(tsk, TASK_UNINTERRUPTIBLE); - } - remove_wait_queue(&sem->wait, &wait); - __set_task_state(tsk, TASK_RUNNING); - - /* - * If there are any more sleepers, wake one of them up so - * that it can either get the semaphore, or set count to -1 - * indicating that there are still processes sleeping. - */ - wake_up(&sem->wait); -} -EXPORT_SYMBOL(__down); - -int __sched __down_interruptible(struct semaphore * sem) -{ - int retval = 0; - struct task_struct *tsk = current; - DECLARE_WAITQUEUE(wait, tsk); - - __set_task_state(tsk, TASK_INTERRUPTIBLE); - add_wait_queue_exclusive(&sem->wait, &wait); - - while (__sem_update_count(sem, -1) <= 0) { - if (signal_pending(current)) { - /* - * A signal is pending - give up trying. - * Set sem->count to 0 if it is negative, - * since we are no longer sleeping. - */ - __sem_update_count(sem, 0); - retval = -EINTR; - break; - } - schedule(); - set_task_state(tsk, TASK_INTERRUPTIBLE); - } - remove_wait_queue(&sem->wait, &wait); - __set_task_state(tsk, TASK_RUNNING); - - wake_up(&sem->wait); - return retval; -} -EXPORT_SYMBOL(__down_interruptible); diff --git a/arch/ppc/kernel/semaphore.c b/arch/ppc/kernel/semaphore.c deleted file mode 100644 index 2fe429b27c14..000000000000 --- a/arch/ppc/kernel/semaphore.c +++ /dev/null @@ -1,131 +0,0 @@ -/* - * PowerPC-specific semaphore code. - * - * Copyright (C) 1999 Cort Dougan <cort@cs.nmt.edu> - * - * This program is free software; you can redistribute it and/or - * modify it under the terms of the GNU General Public License - * as published by the Free Software Foundation; either version - * 2 of the License, or (at your option) any later version. - * - * April 2001 - Reworked by Paul Mackerras <paulus@samba.org> - * to eliminate the SMP races in the old version between the updates - * of `count' and `waking'. Now we use negative `count' values to - * indicate that some process(es) are waiting for the semaphore. - */ - -#include <linux/sched.h> -#include <linux/init.h> -#include <asm/atomic.h> -#include <asm/semaphore.h> -#include <asm/errno.h> - -/* - * Atomically update sem->count. - * This does the equivalent of the following: - * - * old_count = sem->count; - * tmp = MAX(old_count, 0) + incr; - * sem->count = tmp; - * return old_count; - */ -static inline int __sem_update_count(struct semaphore *sem, int incr) -{ - int old_count, tmp; - - __asm__ __volatile__("\n" -"1: lwarx %0,0,%3\n" -" srawi %1,%0,31\n" -" andc %1,%0,%1\n" -" add %1,%1,%4\n" - PPC405_ERR77(0,%3) -" stwcx. %1,0,%3\n" -" bne 1b" - : "=&r" (old_count), "=&r" (tmp), "=m" (sem->count) - : "r" (&sem->count), "r" (incr), "m" (sem->count) - : "cc"); - - return old_count; -} - -void __up(struct semaphore *sem) -{ - /* - * Note that we incremented count in up() before we came here, - * but that was ineffective since the result was <= 0, and - * any negative value of count is equivalent to 0. - * This ends up setting count to 1, unless count is now > 0 - * (i.e. because some other cpu has called up() in the meantime), - * in which case we just increment count. - */ - __sem_update_count(sem, 1); - wake_up(&sem->wait); -} - -/* - * Note that when we come in to __down or __down_interruptible, - * we have already decremented count, but that decrement was - * ineffective since the result was < 0, and any negative value - * of count is equivalent to 0. - * Thus it is only when we decrement count from some value > 0 - * that we have actually got the semaphore. - */ -void __sched __down(struct semaphore *sem) -{ - struct task_struct *tsk = current; - DECLARE_WAITQUEUE(wait, tsk); - - tsk->state = TASK_UNINTERRUPTIBLE; - add_wait_queue_exclusive(&sem->wait, &wait); - smp_wmb(); - - /* - * Try to get the semaphore. If the count is > 0, then we've - * got the semaphore; we decrement count and exit the loop. - * If the count is 0 or negative, we set it to -1, indicating - * that we are asleep, and then sleep. - */ - while (__sem_update_count(sem, -1) <= 0) { - schedule(); - tsk->state = TASK_UNINTERRUPTIBLE; - } - remove_wait_queue(&sem->wait, &wait); - tsk->state = TASK_RUNNING; - - /* - * If there are any more sleepers, wake one of them up so - * that it can either get the semaphore, or set count to -1 - * indicating that there are still processes sleeping. - */ - wake_up(&sem->wait); -} - -int __sched __down_interruptible(struct semaphore * sem) -{ - int retval = 0; - struct task_struct *tsk = current; - DECLARE_WAITQUEUE(wait, tsk); - - tsk->state = TASK_INTERRUPTIBLE; - add_wait_queue_exclusive(&sem->wait, &wait); - smp_wmb(); - - while (__sem_update_count(sem, -1) <= 0) { - if (signal_pending(current)) { - /* - * A signal is pending - give up trying. - * Set sem->count to 0 if it is negative, - * since we are no longer sleeping. - */ - __sem_update_count(sem, 0); - retval = -EINTR; - break; - } - schedule(); - tsk->state = TASK_INTERRUPTIBLE; - } - tsk->state = TASK_RUNNING; - remove_wait_queue(&sem->wait, &wait); - wake_up(&sem->wait); - return retval; -} diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig index 1831833c430e..f6a68e178fc5 100644 --- a/arch/s390/Kconfig +++ b/arch/s390/Kconfig @@ -3,6 +3,10 @@ # see Documentation/kbuild/kconfig-language.txt. # +config SCHED_MC + def_bool y + depends on SMP + config MMU def_bool y @@ -39,6 +43,9 @@ config GENERIC_HWEIGHT config GENERIC_TIME def_bool y +config GENERIC_CLOCKEVENTS + def_bool y + config GENERIC_BUG bool depends on BUG @@ -69,6 +76,8 @@ menu "Base setup" comment "Processor type and features" +source "kernel/time/Kconfig" + config 64BIT bool "64 bit kernel" help @@ -301,10 +310,7 @@ config QDIO tristate "QDIO support" ---help--- This driver provides the Queued Direct I/O base support for - IBM mainframes. - - For details please refer to the documentation provided by IBM at - <http://www10.software.ibm.com/developerworks/opensource/linux390> + IBM System z. To compile this driver as a module, choose M here: the module will be called qdio. @@ -486,25 +492,6 @@ config APPLDATA_NET_SUM source kernel/Kconfig.hz -config NO_IDLE_HZ - bool "No HZ timer ticks in idle" - help - Switches the regular HZ timer off when the system is going idle. - This helps z/VM to detect that the Linux system is idle. VM can - then "swap-out" this guest which reduces memory usage. It also - reduces the overhead of idle systems. - - The HZ timer can be switched on/off via /proc/sys/kernel/hz_timer. - hz_timer=0 means HZ timer is disabled. hz_timer=1 means HZ - timer is active. - -config NO_IDLE_HZ_INIT - bool "HZ timer in idle off by default" - depends on NO_IDLE_HZ - help - The HZ timer is switched off in idle by default. That means the - HZ timer is already disabled at boot time. - config S390_HYPFS_FS bool "s390 hypervisor file system support" select SYS_HYPERVISOR diff --git a/arch/s390/crypto/aes_s390.c b/arch/s390/crypto/aes_s390.c index a3f67f8b5427..e33f32b54c08 100644 --- a/arch/s390/crypto/aes_s390.c +++ b/arch/s390/crypto/aes_s390.c @@ -499,7 +499,7 @@ static struct crypto_alg cbc_aes_alg = { } }; -static int __init aes_init(void) +static int __init aes_s390_init(void) { int ret; @@ -542,15 +542,15 @@ aes_err: goto out; } -static void __exit aes_fini(void) +static void __exit aes_s390_fini(void) { crypto_unregister_alg(&cbc_aes_alg); crypto_unregister_alg(&ecb_aes_alg); crypto_unregister_alg(&aes_alg); } -module_init(aes_init); -module_exit(aes_fini); +module_init(aes_s390_init); +module_exit(aes_s390_fini); MODULE_ALIAS("aes"); diff --git a/arch/s390/crypto/des_s390.c b/arch/s390/crypto/des_s390.c index ea22707f435f..4aba83b31596 100644 --- a/arch/s390/crypto/des_s390.c +++ b/arch/s390/crypto/des_s390.c @@ -550,7 +550,7 @@ static struct crypto_alg cbc_des3_192_alg = { } }; -static int init(void) +static int des_s390_init(void) { int ret = 0; @@ -612,7 +612,7 @@ des_err: goto out; } -static void __exit fini(void) +static void __exit des_s390_fini(void) { crypto_unregister_alg(&cbc_des3_192_alg); crypto_unregister_alg(&ecb_des3_192_alg); @@ -625,8 +625,8 @@ static void __exit fini(void) crypto_unregister_alg(&des_alg); } -module_init(init); -module_exit(fini); +module_init(des_s390_init); +module_exit(des_s390_fini); MODULE_ALIAS("des"); MODULE_ALIAS("des3_ede"); diff --git a/arch/s390/crypto/sha1_s390.c b/arch/s390/crypto/sha1_s390.c index 5a834f6578ab..9cf9eca22747 100644 --- a/arch/s390/crypto/sha1_s390.c +++ b/arch/s390/crypto/sha1_s390.c @@ -137,7 +137,7 @@ static struct crypto_alg alg = { .dia_final = sha1_final } } }; -static int __init init(void) +static int __init sha1_s390_init(void) { if (!crypt_s390_func_available(KIMD_SHA_1)) return -EOPNOTSUPP; @@ -145,13 +145,13 @@ static int __init init(void) return crypto_register_alg(&alg); } -static void __exit fini(void) +static void __exit sha1_s390_fini(void) { crypto_unregister_alg(&alg); } -module_init(init); -module_exit(fini); +module_init(sha1_s390_init); +module_exit(sha1_s390_fini); MODULE_ALIAS("sha1"); diff --git a/arch/s390/crypto/sha256_s390.c b/arch/s390/crypto/sha256_s390.c index ccf8633c4f65..2a3d756b35d4 100644 --- a/arch/s390/crypto/sha256_s390.c +++ b/arch/s390/crypto/sha256_s390.c @@ -133,7 +133,7 @@ static struct crypto_alg alg = { .dia_final = sha256_final } } }; -static int init(void) +static int sha256_s390_init(void) { if (!crypt_s390_func_available(KIMD_SHA_256)) return -EOPNOTSUPP; @@ -141,13 +141,13 @@ static int init(void) return crypto_register_alg(&alg); } -static void __exit fini(void) +static void __exit sha256_s390_fini(void) { crypto_unregister_alg(&alg); } -module_init(init); -module_exit(fini); +module_init(sha256_s390_init); +module_exit(sha256_s390_fini); MODULE_ALIAS("sha256"); diff --git a/arch/s390/defconfig b/arch/s390/defconfig index 62f6b5a606dd..dcc3ec2ef643 100644 --- a/arch/s390/defconfig +++ b/arch/s390/defconfig @@ -3,6 +3,7 @@ # Linux kernel version: 2.6.25-rc4 # Wed Mar 5 11:22:59 2008 # +CONFIG_SCHED_MC=y CONFIG_MMU=y CONFIG_ZONE_DMA=y CONFIG_LOCKDEP_SUPPORT=y diff --git a/arch/s390/kernel/Makefile b/arch/s390/kernel/Makefile index 4d3e38392cb1..77051cd27925 100644 --- a/arch/s390/kernel/Makefile +++ b/arch/s390/kernel/Makefile @@ -11,7 +11,7 @@ CFLAGS_smp.o := -Wno-nonnull obj-y := bitmap.o traps.o time.o process.o base.o early.o \ setup.o sys_s390.o ptrace.o signal.o cpcmd.o ebcdic.o \ - semaphore.o s390_ext.o debug.o irq.o ipl.o dis.o diag.o + s390_ext.o debug.o irq.o ipl.o dis.o diag.o obj-y += $(if $(CONFIG_64BIT),entry64.o,entry.o) obj-y += $(if $(CONFIG_64BIT),reipl64.o,reipl.o) @@ -19,7 +19,7 @@ obj-y += $(if $(CONFIG_64BIT),reipl64.o,reipl.o) extra-y += head.o init_task.o vmlinux.lds obj-$(CONFIG_MODULES) += s390_ksyms.o module.o -obj-$(CONFIG_SMP) += smp.o +obj-$(CONFIG_SMP) += smp.o topology.o obj-$(CONFIG_AUDIT) += audit.o compat-obj-$(CONFIG_AUDIT) += compat_audit.o diff --git a/arch/s390/kernel/compat_linux.h b/arch/s390/kernel/compat_linux.h index e89f8c0c42a0..20723a062017 100644 --- a/arch/s390/kernel/compat_linux.h +++ b/arch/s390/kernel/compat_linux.h @@ -162,4 +162,77 @@ struct ucontext32 { compat_sigset_t uc_sigmask; /* mask last for extensibility */ }; +struct __sysctl_args32; +struct stat64_emu31; +struct mmap_arg_struct_emu31; +struct fadvise64_64_args; +struct old_sigaction32; +struct old_sigaction32; + +long sys32_chown16(const char __user * filename, u16 user, u16 group); +long sys32_lchown16(const char __user * filename, u16 user, u16 group); +long sys32_fchown16(unsigned int fd, u16 user, u16 group); +long sys32_setregid16(u16 rgid, u16 egid); +long sys32_setgid16(u16 gid); +long sys32_setreuid16(u16 ruid, u16 euid); +long sys32_setuid16(u16 uid); +long sys32_setresuid16(u16 ruid, u16 euid, u16 suid); +long sys32_getresuid16(u16 __user *ruid, u16 __user *euid, u16 __user *suid); +long sys32_setresgid16(u16 rgid, u16 egid, u16 sgid); +long sys32_getresgid16(u16 __user *rgid, u16 __user *egid, u16 __user *sgid); +long sys32_setfsuid16(u16 uid); +long sys32_setfsgid16(u16 gid); +long sys32_getgroups16(int gidsetsize, u16 __user *grouplist); +long sys32_setgroups16(int gidsetsize, u16 __user *grouplist); +long sys32_getuid16(void); +long sys32_geteuid16(void); +long sys32_getgid16(void); +long sys32_getegid16(void); +long sys32_ipc(u32 call, int first, int second, int third, u32 ptr); +long sys32_truncate64(const char __user * path, unsigned long high, + unsigned long low); +long sys32_ftruncate64(unsigned int fd, unsigned long high, unsigned long low); +long sys32_sched_rr_get_interval(compat_pid_t pid, + struct compat_timespec __user *interval); +long sys32_rt_sigprocmask(int how, compat_sigset_t __user *set, + compat_sigset_t __user *oset, size_t sigsetsize); +long sys32_rt_sigpending(compat_sigset_t __user *set, size_t sigsetsize); +long sys32_rt_sigqueueinfo(int pid, int sig, compat_siginfo_t __user *uinfo); +long sys32_execve(void); +long sys32_init_module(void __user *umod, unsigned long len, + const char __user *uargs); +long sys32_delete_module(const char __user *name_user, unsigned int flags); +long sys32_gettimeofday(struct compat_timeval __user *tv, + struct timezone __user *tz); +long sys32_settimeofday(struct compat_timeval __user *tv, + struct timezone __user *tz); +long sys32_pause(void); +long sys32_pread64(unsigned int fd, char __user *ubuf, size_t count, + u32 poshi, u32 poslo); +long sys32_pwrite64(unsigned int fd, const char __user *ubuf, + size_t count, u32 poshi, u32 poslo); +compat_ssize_t sys32_readahead(int fd, u32 offhi, u32 offlo, s32 count); +long sys32_sendfile(int out_fd, int in_fd, compat_off_t __user *offset, + size_t count); +long sys32_sendfile64(int out_fd, int in_fd, compat_loff_t __user *offset, + s32 count); +long sys32_sysctl(struct __sysctl_args32 __user *args); +long sys32_stat64(char __user * filename, struct stat64_emu31 __user * statbuf); +long sys32_lstat64(char __user * filename, + struct stat64_emu31 __user * statbuf); +long sys32_fstat64(unsigned long fd, struct stat64_emu31 __user * statbuf); +long sys32_fstatat64(unsigned int dfd, char __user *filename, + struct stat64_emu31 __user* statbuf, int flag); +unsigned long old32_mmap(struct mmap_arg_struct_emu31 __user *arg); +long sys32_mmap2(struct mmap_arg_struct_emu31 __user *arg); +long sys32_read(unsigned int fd, char __user * buf, size_t count); +long sys32_write(unsigned int fd, char __user * buf, size_t count); +long sys32_clone(void); +long sys32_fadvise64(int fd, loff_t offset, size_t len, int advise); +long sys32_fadvise64_64(struct fadvise64_64_args __user *args); +long sys32_sigaction(int sig, const struct old_sigaction32 __user *act, + struct old_sigaction32 __user *oact); +long sys32_rt_sigaction(int sig, const struct sigaction32 __user *act, + struct sigaction32 __user *oact, size_t sigsetsize); +long sys32_sigaltstack(const stack_t32 __user *uss, stack_t32 __user *uoss); #endif /* _ASM_S390X_S390_H */ diff --git a/arch/s390/kernel/compat_signal.c b/arch/s390/kernel/compat_signal.c index a5692c460bad..c7f02e777af2 100644 --- a/arch/s390/kernel/compat_signal.c +++ b/arch/s390/kernel/compat_signal.c @@ -29,6 +29,7 @@ #include <asm/lowcore.h> #include "compat_linux.h" #include "compat_ptrace.h" +#include "entry.h" #define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP))) @@ -428,6 +429,10 @@ get_sigframe(struct k_sigaction *ka, struct pt_regs * regs, size_t frame_size) /* Default to using normal stack */ sp = (unsigned long) A(regs->gprs[15]); + /* Overflow on alternate signal stack gives SIGSEGV. */ + if (on_sig_stack(sp) && !on_sig_stack((sp - frame_size) & -8UL)) + return (void __user *) -1UL; + /* This is the X/Open sanctioned signal stack switching. */ if (ka->sa.sa_flags & SA_ONSTACK) { if (! sas_ss_flags(sp)) @@ -461,6 +466,9 @@ static int setup_frame32(int sig, struct k_sigaction *ka, if (!access_ok(VERIFY_WRITE, frame, sizeof(sigframe32))) goto give_sigsegv; + if (frame == (void __user *) -1UL) + goto give_sigsegv; + if (__copy_to_user(&frame->sc.oldmask, &set->sig, _SIGMASK_COPY_SIZE32)) goto give_sigsegv; @@ -514,6 +522,9 @@ static int setup_rt_frame32(int sig, struct k_sigaction *ka, siginfo_t *info, if (!access_ok(VERIFY_WRITE, frame, sizeof(rt_sigframe32))) goto give_sigsegv; + if (frame == (void __user *) -1UL) + goto give_sigsegv; + if (copy_siginfo_to_user32(&frame->info, info)) goto give_sigsegv; diff --git a/arch/s390/kernel/debug.c b/arch/s390/kernel/debug.c index 1b2f5ce45320..1e7d4ac7068b 100644 --- a/arch/s390/kernel/debug.c +++ b/arch/s390/kernel/debug.c @@ -73,7 +73,7 @@ static ssize_t debug_input(struct file *file, const char __user *user_buf, static int debug_open(struct inode *inode, struct file *file); static int debug_close(struct inode *inode, struct file *file); static debug_info_t* debug_info_create(char *name, int pages_per_area, - int nr_areas, int buf_size); + int nr_areas, int buf_size, mode_t mode); static void debug_info_get(debug_info_t *); static void debug_info_put(debug_info_t *); static int debug_prolog_level_fn(debug_info_t * id, @@ -157,7 +157,7 @@ struct debug_view debug_sprintf_view = { }; /* used by dump analysis tools to determine version of debug feature */ -unsigned int debug_feature_version = __DEBUG_FEATURE_VERSION; +static unsigned int __used debug_feature_version = __DEBUG_FEATURE_VERSION; /* static globals */ @@ -327,7 +327,8 @@ debug_info_free(debug_info_t* db_info){ */ static debug_info_t* -debug_info_create(char *name, int pages_per_area, int nr_areas, int buf_size) +debug_info_create(char *name, int pages_per_area, int nr_areas, int buf_size, + mode_t mode) { debug_info_t* rc; @@ -336,6 +337,8 @@ debug_info_create(char *name, int pages_per_area, int nr_areas, int buf_size) if(!rc) goto out; + rc->mode = mode & ~S_IFMT; + /* create root directory */ rc->debugfs_root_entry = debugfs_create_dir(rc->name, debug_debugfs_root_entry); @@ -676,23 +679,30 @@ debug_close(struct inode *inode, struct file *file) } /* - * debug_register: - * - creates and initializes debug area for the caller - * - returns handle for debug area + * debug_register_mode: + * - Creates and initializes debug area for the caller + * The mode parameter allows to specify access rights for the s390dbf files + * - Returns handle for debug area */ -debug_info_t* -debug_register (char *name, int pages_per_area, int nr_areas, int buf_size) +debug_info_t *debug_register_mode(char *name, int pages_per_area, int nr_areas, + int buf_size, mode_t mode, uid_t uid, + gid_t gid) { debug_info_t *rc = NULL; + /* Since debugfs currently does not support uid/gid other than root, */ + /* we do not allow gid/uid != 0 until we get support for that. */ + if ((uid != 0) || (gid != 0)) + printk(KERN_WARNING "debug: Warning - Currently only uid/gid " + "= 0 are supported. Using root as owner now!"); if (!initialized) BUG(); mutex_lock(&debug_mutex); /* create new debug_info */ - rc = debug_info_create(name, pages_per_area, nr_areas, buf_size); + rc = debug_info_create(name, pages_per_area, nr_areas, buf_size, mode); if(!rc) goto out; debug_register_view(rc, &debug_level_view); @@ -705,6 +715,20 @@ out: mutex_unlock(&debug_mutex); return rc; } +EXPORT_SYMBOL(debug_register_mode); + +/* + * debug_register: + * - creates and initializes debug area for the caller + * - returns handle for debug area + */ + +debug_info_t *debug_register(char *name, int pages_per_area, int nr_areas, + int buf_size) +{ + return debug_register_mode(name, pages_per_area, nr_areas, buf_size, + S_IRUSR | S_IWUSR, 0, 0); +} /* * debug_unregister: @@ -1073,15 +1097,16 @@ debug_register_view(debug_info_t * id, struct debug_view *view) int rc = 0; int i; unsigned long flags; - mode_t mode = S_IFREG; + mode_t mode; struct dentry *pde; if (!id) goto out; - if (view->prolog_proc || view->format_proc || view->header_proc) - mode |= S_IRUSR; - if (view->input_proc) - mode |= S_IWUSR; + mode = (id->mode | S_IFREG) & ~S_IXUGO; + if (!(view->prolog_proc || view->format_proc || view->header_proc)) + mode &= ~(S_IRUSR | S_IRGRP | S_IROTH); + if (!view->input_proc) + mode &= ~(S_IWUSR | S_IWGRP | S_IWOTH); pde = debugfs_create_file(view->name, mode, id->debugfs_root_entry, id , &debug_file_ops); if (!pde){ diff --git a/arch/s390/kernel/early.c b/arch/s390/kernel/early.c index 01832c440636..540a67f979b6 100644 --- a/arch/s390/kernel/early.c +++ b/arch/s390/kernel/early.c @@ -21,6 +21,7 @@ #include <asm/setup.h> #include <asm/cpcmd.h> #include <asm/sclp.h> +#include "entry.h" /* * Create a Kernel NSS if the SAVESYS= parameter is defined diff --git a/arch/s390/kernel/entry.h b/arch/s390/kernel/entry.h new file mode 100644 index 000000000000..6b1896345eda --- /dev/null +++ b/arch/s390/kernel/entry.h @@ -0,0 +1,60 @@ +#ifndef _ENTRY_H +#define _ENTRY_H + +#include <linux/types.h> +#include <linux/signal.h> +#include <asm/ptrace.h> + +typedef void pgm_check_handler_t(struct pt_regs *, long); +extern pgm_check_handler_t *pgm_check_table[128]; +pgm_check_handler_t do_protection_exception; +pgm_check_handler_t do_dat_exception; + +extern int sysctl_userprocess_debug; + +void do_single_step(struct pt_regs *regs); +void syscall_trace(struct pt_regs *regs, int entryexit); +void kernel_stack_overflow(struct pt_regs * regs); +void do_signal(struct pt_regs *regs); +int handle_signal32(unsigned long sig, struct k_sigaction *ka, + siginfo_t *info, sigset_t *oldset, struct pt_regs *regs); + +void do_extint(struct pt_regs *regs, unsigned short code); +int __cpuinit start_secondary(void *cpuvoid); +void __init startup_init(void); +void die(const char * str, struct pt_regs * regs, long err); + +struct new_utsname; +struct mmap_arg_struct; +struct fadvise64_64_args; +struct old_sigaction; +struct sel_arg_struct; + +long sys_pipe(unsigned long __user *fildes); +long sys_mmap2(struct mmap_arg_struct __user *arg); +long old_mmap(struct mmap_arg_struct __user *arg); +long sys_ipc(uint call, int first, unsigned long second, + unsigned long third, void __user *ptr); +long s390x_newuname(struct new_utsname __user *name); +long s390x_personality(unsigned long personality); +long s390_fadvise64(int fd, u32 offset_high, u32 offset_low, + size_t len, int advice); +long s390_fadvise64_64(struct fadvise64_64_args __user *args); +long s390_fallocate(int fd, int mode, loff_t offset, u32 len_high, u32 len_low); +long sys_fork(void); +long sys_clone(void); +long sys_vfork(void); +void execve_tail(void); +long sys_execve(void); +int sys_sigsuspend(int history0, int history1, old_sigset_t mask); +long sys_sigaction(int sig, const struct old_sigaction __user *act, + struct old_sigaction __user *oact); +long sys_sigaltstack(const stack_t __user *uss, stack_t __user *uoss); +long sys_sigreturn(void); +long sys_rt_sigreturn(void); +long sys32_sigreturn(void); +long sys32_rt_sigreturn(void); +long old_select(struct sel_arg_struct __user *arg); +long sys_ptrace(long request, long pid, long addr, long data); + +#endif /* _ENTRY_H */ diff --git a/arch/s390/kernel/entry64.S b/arch/s390/kernel/entry64.S index efde6e178f6c..cd959c0b2e16 100644 --- a/arch/s390/kernel/entry64.S +++ b/arch/s390/kernel/entry64.S @@ -475,6 +475,7 @@ pgm_check_handler: pgm_no_vtime: #endif lg %r9,__LC_THREAD_INFO # load pointer to thread_info struct + mvc SP_ARGS(8,%r15),__LC_LAST_BREAK TRACE_IRQS_OFF lgf %r3,__LC_PGM_ILC # load program interruption code lghi %r8,0x7f @@ -847,6 +848,7 @@ stack_overflow: je 0f la %r1,__LC_SAVE_AREA+32 0: mvc SP_R12(32,%r15),0(%r1) # move %r12-%r15 to stack + mvc SP_ARGS(8,%r15),__LC_LAST_BREAK xc __SF_BACKCHAIN(8,%r15),__SF_BACKCHAIN(%r15) # clear back chain la %r2,SP_PTREGS(%r15) # load pt_regs jg kernel_stack_overflow diff --git a/arch/s390/kernel/ipl.c b/arch/s390/kernel/ipl.c index 375232c46c7a..532542447d66 100644 --- a/arch/s390/kernel/ipl.c +++ b/arch/s390/kernel/ipl.c @@ -655,7 +655,7 @@ static struct kobj_attribute reipl_type_attr = static struct kset *reipl_kset; -void reipl_run(struct shutdown_trigger *trigger) +static void reipl_run(struct shutdown_trigger *trigger) { struct ccw_dev_id devid; static char buf[100]; diff --git a/arch/s390/kernel/kprobes.c b/arch/s390/kernel/kprobes.c index c5549a206284..ed04d1372d5d 100644 --- a/arch/s390/kernel/kprobes.c +++ b/arch/s390/kernel/kprobes.c @@ -360,7 +360,7 @@ no_kprobe: * - When the probed function returns, this probe * causes the handlers to fire */ -void kretprobe_trampoline_holder(void) +static void __used kretprobe_trampoline_holder(void) { asm volatile(".global kretprobe_trampoline\n" "kretprobe_trampoline: bcr 0,0\n"); diff --git a/arch/s390/kernel/process.c b/arch/s390/kernel/process.c index ce203154d8ce..c1aff194141d 100644 --- a/arch/s390/kernel/process.c +++ b/arch/s390/kernel/process.c @@ -36,6 +36,8 @@ #include <linux/module.h> #include <linux/notifier.h> #include <linux/utsname.h> +#include <linux/tick.h> +#include <linux/elfcore.h> #include <asm/uaccess.h> #include <asm/pgtable.h> #include <asm/system.h> @@ -44,6 +46,7 @@ #include <asm/irq.h> #include <asm/timer.h> #include <asm/cpu.h> +#include "entry.h" asmlinkage void ret_from_fork(void) asm ("ret_from_fork"); @@ -76,6 +79,7 @@ unsigned long thread_saved_pc(struct task_struct *tsk) * Need to know about CPUs going idle? */ static ATOMIC_NOTIFIER_HEAD(idle_chain); +DEFINE_PER_CPU(struct s390_idle_data, s390_idle); int register_idle_notifier(struct notifier_block *nb) { @@ -89,9 +93,33 @@ int unregister_idle_notifier(struct notifier_block *nb) } EXPORT_SYMBOL(unregister_idle_notifier); -void do_monitor_call(struct pt_regs *regs, long interruption_code) +static int s390_idle_enter(void) +{ + struct s390_idle_data *idle; + int nr_calls = 0; + void *hcpu; + int rc; + + hcpu = (void *)(long)smp_processor_id(); + rc = __atomic_notifier_call_chain(&idle_chain, S390_CPU_IDLE, hcpu, -1, + &nr_calls); + if (rc == NOTIFY_BAD) { + nr_calls--; + __atomic_notifier_call_chain(&idle_chain, S390_CPU_NOT_IDLE, + hcpu, nr_calls, NULL); + return rc; + } + idle = &__get_cpu_var(s390_idle); + spin_lock(&idle->lock); + idle->idle_count++; + idle->in_idle = 1; + idle->idle_enter = get_clock(); + spin_unlock(&idle->lock); + return NOTIFY_OK; +} + +void s390_idle_leave(void) { -#ifdef CONFIG_SMP struct s390_idle_data *idle; idle = &__get_cpu_var(s390_idle); @@ -99,10 +127,6 @@ void do_monitor_call(struct pt_regs *regs, long interruption_code) idle->idle_time += get_clock() - idle->idle_enter; idle->in_idle = 0; spin_unlock(&idle->lock); -#endif - /* disable monitor call class 0 */ - __ctl_clear_bit(8, 15); - atomic_notifier_call_chain(&idle_chain, S390_CPU_NOT_IDLE, (void *)(long) smp_processor_id()); } @@ -113,61 +137,30 @@ extern void s390_handle_mcck(void); */ static void default_idle(void) { - int cpu, rc; - int nr_calls = 0; - void *hcpu; -#ifdef CONFIG_SMP - struct s390_idle_data *idle; -#endif - /* CPU is going idle. */ - cpu = smp_processor_id(); - hcpu = (void *)(long)cpu; local_irq_disable(); if (need_resched()) { local_irq_enable(); return; } - - rc = __atomic_notifier_call_chain(&idle_chain, S390_CPU_IDLE, hcpu, -1, - &nr_calls); - if (rc == NOTIFY_BAD) { - nr_calls--; - __atomic_notifier_call_chain(&idle_chain, S390_CPU_NOT_IDLE, - hcpu, nr_calls, NULL); + if (s390_idle_enter() == NOTIFY_BAD) { local_irq_enable(); return; } - - /* enable monitor call class 0 */ - __ctl_set_bit(8, 15); - #ifdef CONFIG_HOTPLUG_CPU - if (cpu_is_offline(cpu)) { + if (cpu_is_offline(smp_processor_id())) { preempt_enable_no_resched(); cpu_die(); } #endif - local_mcck_disable(); if (test_thread_flag(TIF_MCCK_PENDING)) { local_mcck_enable(); - /* disable monitor call class 0 */ - __ctl_clear_bit(8, 15); - atomic_notifier_call_chain(&idle_chain, S390_CPU_NOT_IDLE, - hcpu); + s390_idle_leave(); local_irq_enable(); s390_handle_mcck(); return; } -#ifdef CONFIG_SMP - idle = &__get_cpu_var(s390_idle); - spin_lock(&idle->lock); - idle->idle_count++; - idle->in_idle = 1; - idle->idle_enter = get_clock(); - spin_unlock(&idle->lock); -#endif trace_hardirqs_on(); /* Wait for external, I/O or machine check interrupt. */ __load_psw_mask(psw_kernel_bits | PSW_MASK_WAIT | @@ -177,9 +170,10 @@ static void default_idle(void) void cpu_idle(void) { for (;;) { + tick_nohz_stop_sched_tick(); while (!need_resched()) default_idle(); - + tick_nohz_restart_sched_tick(); preempt_enable_no_resched(); schedule(); preempt_disable(); @@ -201,6 +195,7 @@ void show_regs(struct pt_regs *regs) /* Show stack backtrace if pt_regs is from kernel mode */ if (!(regs->psw.mask & PSW_MASK_PSTATE)) show_trace(NULL, (unsigned long *) regs->gprs[15]); + show_last_breaking_event(regs); } extern void kernel_thread_starter(void); diff --git a/arch/s390/kernel/ptrace.c b/arch/s390/kernel/ptrace.c index 6e036bae9875..58a064296987 100644 --- a/arch/s390/kernel/ptrace.c +++ b/arch/s390/kernel/ptrace.c @@ -41,6 +41,7 @@ #include <asm/system.h> #include <asm/uaccess.h> #include <asm/unistd.h> +#include "entry.h" #ifdef CONFIG_COMPAT #include "compat_ptrace.h" diff --git a/arch/s390/kernel/s390_ext.c b/arch/s390/kernel/s390_ext.c index acf93dba7727..e019b419efc6 100644 --- a/arch/s390/kernel/s390_ext.c +++ b/arch/s390/kernel/s390_ext.c @@ -13,11 +13,12 @@ #include <linux/errno.h> #include <linux/kernel_stat.h> #include <linux/interrupt.h> - +#include <asm/cpu.h> #include <asm/lowcore.h> #include <asm/s390_ext.h> #include <asm/irq_regs.h> #include <asm/irq.h> +#include "entry.h" /* * ext_int_hash[index] is the start of the list for all external interrupts @@ -119,13 +120,10 @@ void do_extint(struct pt_regs *regs, unsigned short code) old_regs = set_irq_regs(regs); irq_enter(); - asm volatile ("mc 0,0"); - if (S390_lowcore.int_clock >= S390_lowcore.jiffy_timer) - /** - * Make sure that the i/o interrupt did not "overtake" - * the last HZ timer interrupt. - */ - account_ticks(S390_lowcore.int_clock); + s390_idle_check(); + if (S390_lowcore.int_clock >= S390_lowcore.clock_comparator) + /* Serve timer interrupts first. */ + clock_comparator_work(); kstat_cpu(smp_processor_id()).irqs[EXTERNAL_INTERRUPT]++; index = ext_hash(code); for (p = ext_int_hash[index]; p; p = p->next) { diff --git a/arch/s390/kernel/s390_ksyms.c b/arch/s390/kernel/s390_ksyms.c index 7234c737f825..48238a114ce9 100644 --- a/arch/s390/kernel/s390_ksyms.c +++ b/arch/s390/kernel/s390_ksyms.c @@ -27,13 +27,6 @@ EXPORT_SYMBOL(_zb_findmap); EXPORT_SYMBOL(_sb_findmap); /* - * semaphore ops - */ -EXPORT_SYMBOL(__up); -EXPORT_SYMBOL(__down); -EXPORT_SYMBOL(__down_interruptible); - -/* * binfmt_elf loader */ extern int dump_fpu (struct pt_regs * regs, s390_fp_regs *fpregs); diff --git a/arch/s390/kernel/semaphore.c b/arch/s390/kernel/semaphore.c deleted file mode 100644 index 191303f6c1d8..000000000000 --- a/arch/s390/kernel/semaphore.c +++ /dev/null @@ -1,108 +0,0 @@ -/* - * linux/arch/s390/kernel/semaphore.c - * - * S390 version - * Copyright (C) 1998-2000 IBM Corporation - * Author(s): Martin Schwidefsky - * - * Derived from "linux/arch/i386/kernel/semaphore.c - * Copyright (C) 1999, Linus Torvalds - * - */ -#include <linux/sched.h> -#include <linux/errno.h> -#include <linux/init.h> - -#include <asm/semaphore.h> - -/* - * Atomically update sem->count. Equivalent to: - * old_val = sem->count.counter; - * new_val = ((old_val >= 0) ? old_val : 0) + incr; - * sem->count.counter = new_val; - * return old_val; - */ -static inline int __sem_update_count(struct semaphore *sem, int incr) -{ - int old_val, new_val; - - asm volatile( - " l %0,0(%3)\n" - "0: ltr %1,%0\n" - " jhe 1f\n" - " lhi %1,0\n" - "1: ar %1,%4\n" - " cs %0,%1,0(%3)\n" - " jl 0b\n" - : "=&d" (old_val), "=&d" (new_val), "=m" (sem->count) - : "a" (&sem->count), "d" (incr), "m" (sem->count) - : "cc"); - return old_val; -} - -/* - * The inline function up() incremented count but the result - * was <= 0. This indicates that some process is waiting on - * the semaphore. The semaphore is free and we'll wake the - * first sleeping process, so we set count to 1 unless some - * other cpu has called up in the meantime in which case - * we just increment count by 1. - */ -void __up(struct semaphore *sem) -{ - __sem_update_count(sem, 1); - wake_up(&sem->wait); -} - -/* - * The inline function down() decremented count and the result - * was < 0. The wait loop will atomically test and update the - * semaphore counter following the rules: - * count > 0: decrement count, wake up queue and exit. - * count <= 0: set count to -1, go to sleep. - */ -void __sched __down(struct semaphore * sem) -{ - struct task_struct *tsk = current; - DECLARE_WAITQUEUE(wait, tsk); - - __set_task_state(tsk, TASK_UNINTERRUPTIBLE); - add_wait_queue_exclusive(&sem->wait, &wait); - while (__sem_update_count(sem, -1) <= 0) { - schedule(); - set_task_state(tsk, TASK_UNINTERRUPTIBLE); - } - remove_wait_queue(&sem->wait, &wait); - __set_task_state(tsk, TASK_RUNNING); - wake_up(&sem->wait); -} - -/* - * Same as __down() with an additional test for signals. - * If a signal is pending the count is updated as follows: - * count > 0: wake up queue and exit. - * count <= 0: set count to 0, wake up queue and exit. - */ -int __sched __down_interruptible(struct semaphore * sem) -{ - int retval = 0; - struct task_struct *tsk = current; - DECLARE_WAITQUEUE(wait, tsk); - - __set_task_state(tsk, TASK_INTERRUPTIBLE); - add_wait_queue_exclusive(&sem->wait, &wait); - while (__sem_update_count(sem, -1) <= 0) { - if (signal_pending(current)) { - __sem_update_count(sem, 0); - retval = -EINTR; - break; - } - schedule(); - set_task_state(tsk, TASK_INTERRUPTIBLE); - } - remove_wait_queue(&sem->wait, &wait); - __set_task_state(tsk, TASK_RUNNING); - wake_up(&sem->wait); - return retval; -} - diff --git a/arch/s390/kernel/setup.c b/arch/s390/kernel/setup.c index 290e504061a3..7141147e6b63 100644 --- a/arch/s390/kernel/setup.c +++ b/arch/s390/kernel/setup.c @@ -39,6 +39,7 @@ #include <linux/pfn.h> #include <linux/ctype.h> #include <linux/reboot.h> +#include <linux/topology.h> #include <asm/ipl.h> #include <asm/uaccess.h> @@ -427,7 +428,7 @@ setup_lowcore(void) lc->io_new_psw.mask = psw_kernel_bits; lc->io_new_psw.addr = PSW_ADDR_AMODE | (unsigned long) io_int_handler; lc->ipl_device = S390_lowcore.ipl_device; - lc->jiffy_timer = -1LL; + lc->clock_comparator = -1ULL; lc->kernel_stack = ((unsigned long) &init_thread_union) + THREAD_SIZE; lc->async_stack = (unsigned long) __alloc_bootmem(ASYNC_SIZE, ASYNC_SIZE, 0) + ASYNC_SIZE; @@ -687,7 +688,7 @@ static __init unsigned int stfl(void) return S390_lowcore.stfl_fac_list; } -static __init int stfle(unsigned long long *list, int doublewords) +static int __init __stfle(unsigned long long *list, int doublewords) { typedef struct { unsigned long long _[doublewords]; } addrtype; register unsigned long __nr asm("0") = doublewords - 1; @@ -697,6 +698,13 @@ static __init int stfle(unsigned long long *list, int doublewords) return __nr + 1; } +int __init stfle(unsigned long long *list, int doublewords) +{ + if (!(stfl() & (1UL << 24))) + return -EOPNOTSUPP; + return __stfle(list, doublewords); +} + /* * Setup hardware capabilities. */ @@ -741,7 +749,7 @@ static void __init setup_hwcaps(void) * HWCAP_S390_DFP bit 6. */ if ((elf_hwcap & (1UL << 2)) && - stfle(&facility_list_extended, 1) > 0) { + __stfle(&facility_list_extended, 1) > 0) { if (facility_list_extended & (1ULL << (64 - 43))) elf_hwcap |= 1UL << 6; } @@ -823,6 +831,7 @@ setup_arch(char **cmdline_p) cpu_init(); __cpu_logical_map[0] = S390_lowcore.cpu_data.cpu_addr; + s390_init_cpu_topology(); /* * Setup capabilities (ELF_HWCAP & ELF_PLATFORM). diff --git a/arch/s390/kernel/signal.c b/arch/s390/kernel/signal.c index 4449bf32cbf1..b97682040215 100644 --- a/arch/s390/kernel/signal.c +++ b/arch/s390/kernel/signal.c @@ -27,6 +27,7 @@ #include <asm/ucontext.h> #include <asm/uaccess.h> #include <asm/lowcore.h> +#include "entry.h" #define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP))) @@ -235,6 +236,10 @@ get_sigframe(struct k_sigaction *ka, struct pt_regs * regs, size_t frame_size) /* Default to using normal stack */ sp = regs->gprs[15]; + /* Overflow on alternate signal stack gives SIGSEGV. */ + if (on_sig_stack(sp) && !on_sig_stack((sp - frame_size) & -8UL)) + return (void __user *) -1UL; + /* This is the X/Open sanctioned signal stack switching. */ if (ka->sa.sa_flags & SA_ONSTACK) { if (! sas_ss_flags(sp)) @@ -270,6 +275,9 @@ static int setup_frame(int sig, struct k_sigaction *ka, if (!access_ok(VERIFY_WRITE, frame, sizeof(sigframe))) goto give_sigsegv; + if (frame == (void __user *) -1UL) + goto give_sigsegv; + if (__copy_to_user(&frame->sc.oldmask, &set->sig, _SIGMASK_COPY_SIZE)) goto give_sigsegv; @@ -327,6 +335,9 @@ static int setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, if (!access_ok(VERIFY_WRITE, frame, sizeof(rt_sigframe))) goto give_sigsegv; + if (frame == (void __user *) -1UL) + goto give_sigsegv; + if (copy_siginfo_to_user(&frame->info, info)) goto give_sigsegv; @@ -474,11 +485,6 @@ void do_signal(struct pt_regs *regs) int ret; #ifdef CONFIG_COMPAT if (test_thread_flag(TIF_31BIT)) { - extern int handle_signal32(unsigned long sig, - struct k_sigaction *ka, - siginfo_t *info, - sigset_t *oldset, - struct pt_regs *regs); ret = handle_signal32(signr, &ka, &info, oldset, regs); } else diff --git a/arch/s390/kernel/smp.c b/arch/s390/kernel/smp.c index 8f894d380a62..0dfa988c1b26 100644 --- a/arch/s390/kernel/smp.c +++ b/arch/s390/kernel/smp.c @@ -44,6 +44,7 @@ #include <asm/lowcore.h> #include <asm/sclp.h> #include <asm/cpu.h> +#include "entry.h" /* * An array with a pointer the lowcore of every CPU. @@ -67,13 +68,12 @@ enum s390_cpu_state { CPU_STATE_CONFIGURED, }; -#ifdef CONFIG_HOTPLUG_CPU -static DEFINE_MUTEX(smp_cpu_state_mutex); -#endif +DEFINE_MUTEX(smp_cpu_state_mutex); +int smp_cpu_polarization[NR_CPUS]; static int smp_cpu_state[NR_CPUS]; +static int cpu_management; static DEFINE_PER_CPU(struct cpu, cpu_devices); -DEFINE_PER_CPU(struct s390_idle_data, s390_idle); static void smp_ext_bitcall(int, ec_bit_sig); @@ -298,7 +298,7 @@ static void smp_ext_bitcall(int cpu, ec_bit_sig sig) /* * this function sends a 'purge tlb' signal to another CPU. */ -void smp_ptlb_callback(void *info) +static void smp_ptlb_callback(void *info) { __tlb_flush_local(); } @@ -456,6 +456,7 @@ static int smp_rescan_cpus_sigp(cpumask_t avail) if (cpu_known(cpu_id)) continue; __cpu_logical_map[logical_cpu] = cpu_id; + smp_cpu_polarization[logical_cpu] = POLARIZATION_UNKNWN; if (!cpu_stopped(logical_cpu)) continue; cpu_set(logical_cpu, cpu_present_map); @@ -489,6 +490,7 @@ static int smp_rescan_cpus_sclp(cpumask_t avail) if (cpu_known(cpu_id)) continue; __cpu_logical_map[logical_cpu] = cpu_id; + smp_cpu_polarization[logical_cpu] = POLARIZATION_UNKNWN; cpu_set(logical_cpu, cpu_present_map); if (cpu >= info->configured) smp_cpu_state[logical_cpu] = CPU_STATE_STANDBY; @@ -846,6 +848,7 @@ void __init smp_prepare_boot_cpu(void) S390_lowcore.percpu_offset = __per_cpu_offset[0]; current_set[0] = current; smp_cpu_state[0] = CPU_STATE_CONFIGURED; + smp_cpu_polarization[0] = POLARIZATION_UNKNWN; spin_lock_init(&(&__get_cpu_var(s390_idle))->lock); } @@ -897,15 +900,19 @@ static ssize_t cpu_configure_store(struct sys_device *dev, const char *buf, case 0: if (smp_cpu_state[cpu] == CPU_STATE_CONFIGURED) { rc = sclp_cpu_deconfigure(__cpu_logical_map[cpu]); - if (!rc) + if (!rc) { smp_cpu_state[cpu] = CPU_STATE_STANDBY; + smp_cpu_polarization[cpu] = POLARIZATION_UNKNWN; + } } break; case 1: if (smp_cpu_state[cpu] == CPU_STATE_STANDBY) { rc = sclp_cpu_configure(__cpu_logical_map[cpu]); - if (!rc) + if (!rc) { smp_cpu_state[cpu] = CPU_STATE_CONFIGURED; + smp_cpu_polarization[cpu] = POLARIZATION_UNKNWN; + } } break; default: @@ -919,6 +926,34 @@ out: static SYSDEV_ATTR(configure, 0644, cpu_configure_show, cpu_configure_store); #endif /* CONFIG_HOTPLUG_CPU */ +static ssize_t cpu_polarization_show(struct sys_device *dev, char *buf) +{ + int cpu = dev->id; + ssize_t count; + + mutex_lock(&smp_cpu_state_mutex); + switch (smp_cpu_polarization[cpu]) { + case POLARIZATION_HRZ: + count = sprintf(buf, "horizontal\n"); + break; + case POLARIZATION_VL: + count = sprintf(buf, "vertical:low\n"); + break; + case POLARIZATION_VM: + count = sprintf(buf, "vertical:medium\n"); + break; + case POLARIZATION_VH: + count = sprintf(buf, "vertical:high\n"); + break; + default: + count = sprintf(buf, "unknown\n"); + break; + } + mutex_unlock(&smp_cpu_state_mutex); + return count; +} +static SYSDEV_ATTR(polarization, 0444, cpu_polarization_show, NULL); + static ssize_t show_cpu_address(struct sys_device *dev, char *buf) { return sprintf(buf, "%d\n", __cpu_logical_map[dev->id]); @@ -931,6 +966,7 @@ static struct attribute *cpu_common_attrs[] = { &attr_configure.attr, #endif &attr_address.attr, + &attr_polarization.attr, NULL, }; @@ -1075,11 +1111,48 @@ static ssize_t __ref rescan_store(struct sys_device *dev, out: put_online_cpus(); mutex_unlock(&smp_cpu_state_mutex); + if (!cpus_empty(newcpus)) + topology_schedule_update(); return rc ? rc : count; } static SYSDEV_ATTR(rescan, 0200, NULL, rescan_store); #endif /* CONFIG_HOTPLUG_CPU */ +static ssize_t dispatching_show(struct sys_device *dev, char *buf) +{ + ssize_t count; + + mutex_lock(&smp_cpu_state_mutex); + count = sprintf(buf, "%d\n", cpu_management); + mutex_unlock(&smp_cpu_state_mutex); + return count; +} + +static ssize_t dispatching_store(struct sys_device *dev, const char *buf, + size_t count) +{ + int val, rc; + char delim; + + if (sscanf(buf, "%d %c", &val, &delim) != 1) + return -EINVAL; + if (val != 0 && val != 1) + return -EINVAL; + rc = 0; + mutex_lock(&smp_cpu_state_mutex); + get_online_cpus(); + if (cpu_management == val) + goto out; + rc = topology_set_cpu_management(val); + if (!rc) + cpu_management = val; +out: + put_online_cpus(); + mutex_unlock(&smp_cpu_state_mutex); + return rc ? rc : count; +} +static SYSDEV_ATTR(dispatching, 0644, dispatching_show, dispatching_store); + static int __init topology_init(void) { int cpu; @@ -1093,6 +1166,10 @@ static int __init topology_init(void) if (rc) return rc; #endif + rc = sysfs_create_file(&cpu_sysdev_class.kset.kobj, + &attr_dispatching.attr); + if (rc) + return rc; for_each_present_cpu(cpu) { rc = smp_add_present_cpu(cpu); if (rc) diff --git a/arch/s390/kernel/sys_s390.c b/arch/s390/kernel/sys_s390.c index fefee99f28aa..988d0d64c2c8 100644 --- a/arch/s390/kernel/sys_s390.c +++ b/arch/s390/kernel/sys_s390.c @@ -29,8 +29,8 @@ #include <linux/personality.h> #include <linux/unistd.h> #include <linux/ipc.h> - #include <asm/uaccess.h> +#include "entry.h" /* * sys_pipe() is the normal C calling standard for creating diff --git a/arch/s390/kernel/time.c b/arch/s390/kernel/time.c index cb232c155360..7aec676fefd5 100644 --- a/arch/s390/kernel/time.c +++ b/arch/s390/kernel/time.c @@ -30,7 +30,7 @@ #include <linux/timex.h> #include <linux/notifier.h> #include <linux/clocksource.h> - +#include <linux/clockchips.h> #include <asm/uaccess.h> #include <asm/delay.h> #include <asm/s390_ext.h> @@ -39,6 +39,7 @@ #include <asm/irq_regs.h> #include <asm/timer.h> #include <asm/etr.h> +#include <asm/cio.h> /* change this if you have some constant time drift */ #define USECS_PER_JIFFY ((unsigned long) 1000000/HZ) @@ -57,16 +58,16 @@ static ext_int_info_t ext_int_info_cc; static ext_int_info_t ext_int_etr_cc; -static u64 init_timer_cc; static u64 jiffies_timer_cc; -static u64 xtime_cc; + +static DEFINE_PER_CPU(struct clock_event_device, comparators); /* * Scheduler clock - returns current time in nanosec units. */ unsigned long long sched_clock(void) { - return ((get_clock() - jiffies_timer_cc) * 125) >> 9; + return ((get_clock_xt() - jiffies_timer_cc) * 125) >> 9; } /* @@ -95,162 +96,40 @@ void tod_to_timeval(__u64 todval, struct timespec *xtime) #define s390_do_profile() do { ; } while(0) #endif /* CONFIG_PROFILING */ -/* - * Advance the per cpu tick counter up to the time given with the - * "time" argument. The per cpu update consists of accounting - * the virtual cpu time, calling update_process_times and calling - * the profiling hook. If xtime is before time it is advanced as well. - */ -void account_ticks(u64 time) +void clock_comparator_work(void) { - __u32 ticks; - __u64 tmp; - - /* Calculate how many ticks have passed. */ - if (time < S390_lowcore.jiffy_timer) - return; - tmp = time - S390_lowcore.jiffy_timer; - if (tmp >= 2*CLK_TICKS_PER_JIFFY) { /* more than two ticks ? */ - ticks = __div(tmp, CLK_TICKS_PER_JIFFY) + 1; - S390_lowcore.jiffy_timer += - CLK_TICKS_PER_JIFFY * (__u64) ticks; - } else if (tmp >= CLK_TICKS_PER_JIFFY) { - ticks = 2; - S390_lowcore.jiffy_timer += 2*CLK_TICKS_PER_JIFFY; - } else { - ticks = 1; - S390_lowcore.jiffy_timer += CLK_TICKS_PER_JIFFY; - } - -#ifdef CONFIG_SMP - /* - * Do not rely on the boot cpu to do the calls to do_timer. - * Spread it over all cpus instead. - */ - write_seqlock(&xtime_lock); - if (S390_lowcore.jiffy_timer > xtime_cc) { - __u32 xticks; - tmp = S390_lowcore.jiffy_timer - xtime_cc; - if (tmp >= 2*CLK_TICKS_PER_JIFFY) { - xticks = __div(tmp, CLK_TICKS_PER_JIFFY); - xtime_cc += (__u64) xticks * CLK_TICKS_PER_JIFFY; - } else { - xticks = 1; - xtime_cc += CLK_TICKS_PER_JIFFY; - } - do_timer(xticks); - } - write_sequnlock(&xtime_lock); -#else - do_timer(ticks); -#endif - - while (ticks--) - update_process_times(user_mode(get_irq_regs())); + struct clock_event_device *cd; + S390_lowcore.clock_comparator = -1ULL; + set_clock_comparator(S390_lowcore.clock_comparator); + cd = &__get_cpu_var(comparators); + cd->event_handler(cd); s390_do_profile(); } -#ifdef CONFIG_NO_IDLE_HZ - -#ifdef CONFIG_NO_IDLE_HZ_INIT -int sysctl_hz_timer = 0; -#else -int sysctl_hz_timer = 1; -#endif - -/* - * Stop the HZ tick on the current CPU. - * Only cpu_idle may call this function. - */ -static void stop_hz_timer(void) -{ - unsigned long flags; - unsigned long seq, next; - __u64 timer, todval; - int cpu = smp_processor_id(); - - if (sysctl_hz_timer != 0) - return; - - cpu_set(cpu, nohz_cpu_mask); - - /* - * Leave the clock comparator set up for the next timer - * tick if either rcu or a softirq is pending. - */ - if (rcu_needs_cpu(cpu) || local_softirq_pending()) { - cpu_clear(cpu, nohz_cpu_mask); - return; - } - - /* - * This cpu is going really idle. Set up the clock comparator - * for the next event. - */ - next = next_timer_interrupt(); - do { - seq = read_seqbegin_irqsave(&xtime_lock, flags); - timer = ((__u64) next) - ((__u64) jiffies) + jiffies_64; - } while (read_seqretry_irqrestore(&xtime_lock, seq, flags)); - todval = -1ULL; - /* Be careful about overflows. */ - if (timer < (-1ULL / CLK_TICKS_PER_JIFFY)) { - timer = jiffies_timer_cc + timer * CLK_TICKS_PER_JIFFY; - if (timer >= jiffies_timer_cc) - todval = timer; - } - set_clock_comparator(todval); -} - /* - * Start the HZ tick on the current CPU. - * Only cpu_idle may call this function. + * Fixup the clock comparator. */ -static void start_hz_timer(void) +static void fixup_clock_comparator(unsigned long long delta) { - if (!cpu_isset(smp_processor_id(), nohz_cpu_mask)) + /* If nobody is waiting there's nothing to fix. */ + if (S390_lowcore.clock_comparator == -1ULL) return; - account_ticks(get_clock()); - set_clock_comparator(S390_lowcore.jiffy_timer + CPU_DEVIATION); - cpu_clear(smp_processor_id(), nohz_cpu_mask); -} - -static int nohz_idle_notify(struct notifier_block *self, - unsigned long action, void *hcpu) -{ - switch (action) { - case S390_CPU_IDLE: - stop_hz_timer(); - break; - case S390_CPU_NOT_IDLE: - start_hz_timer(); - break; - } - return NOTIFY_OK; + S390_lowcore.clock_comparator += delta; + set_clock_comparator(S390_lowcore.clock_comparator); } -static struct notifier_block nohz_idle_nb = { - .notifier_call = nohz_idle_notify, -}; - -static void __init nohz_init(void) +static int s390_next_event(unsigned long delta, + struct clock_event_device *evt) { - if (register_idle_notifier(&nohz_idle_nb)) - panic("Couldn't register idle notifier"); + S390_lowcore.clock_comparator = get_clock() + delta; + set_clock_comparator(S390_lowcore.clock_comparator); + return 0; } -#endif - -/* - * Set up per cpu jiffy timer and set the clock comparator. - */ -static void setup_jiffy_timer(void) +static void s390_set_mode(enum clock_event_mode mode, + struct clock_event_device *evt) { - /* Set up clock comparator to next jiffy. */ - S390_lowcore.jiffy_timer = - jiffies_timer_cc + (jiffies_64 + 1) * CLK_TICKS_PER_JIFFY; - set_clock_comparator(S390_lowcore.jiffy_timer + CPU_DEVIATION); } /* @@ -259,7 +138,26 @@ static void setup_jiffy_timer(void) */ void init_cpu_timer(void) { - setup_jiffy_timer(); + struct clock_event_device *cd; + int cpu; + + S390_lowcore.clock_comparator = -1ULL; + set_clock_comparator(S390_lowcore.clock_comparator); + + cpu = smp_processor_id(); + cd = &per_cpu(comparators, cpu); + cd->name = "comparator"; + cd->features = CLOCK_EVT_FEAT_ONESHOT; + cd->mult = 16777; + cd->shift = 12; + cd->min_delta_ns = 1; + cd->max_delta_ns = LONG_MAX; + cd->rating = 400; + cd->cpumask = cpumask_of_cpu(cpu); + cd->set_next_event = s390_next_event; + cd->set_mode = s390_set_mode; + + clockevents_register_device(cd); /* Enable clock comparator timer interrupt. */ __ctl_set_bit(0,11); @@ -270,8 +168,6 @@ void init_cpu_timer(void) static void clock_comparator_interrupt(__u16 code) { - /* set clock comparator for next tick */ - set_clock_comparator(S390_lowcore.jiffy_timer + CPU_DEVIATION); } static void etr_reset(void); @@ -316,8 +212,9 @@ static struct clocksource clocksource_tod = { */ void __init time_init(void) { + u64 init_timer_cc; + init_timer_cc = reset_tod_clock(); - xtime_cc = init_timer_cc + CLK_TICKS_PER_JIFFY; jiffies_timer_cc = init_timer_cc - jiffies_64 * CLK_TICKS_PER_JIFFY; /* set xtime */ @@ -342,10 +239,6 @@ void __init time_init(void) /* Enable TOD clock interrupts on the boot cpu. */ init_cpu_timer(); -#ifdef CONFIG_NO_IDLE_HZ - nohz_init(); -#endif - #ifdef CONFIG_VIRT_TIMER vtime_init(); #endif @@ -699,53 +592,49 @@ static int etr_aib_follows(struct etr_aib *a1, struct etr_aib *a2, int p) } /* - * The time is "clock". xtime is what we think the time is. + * The time is "clock". old is what we think the time is. * Adjust the value by a multiple of jiffies and add the delta to ntp. * "delay" is an approximation how long the synchronization took. If * the time correction is positive, then "delay" is subtracted from * the time difference and only the remaining part is passed to ntp. */ -static void etr_adjust_time(unsigned long long clock, unsigned long long delay) +static unsigned long long etr_adjust_time(unsigned long long old, + unsigned long long clock, + unsigned long long delay) { unsigned long long delta, ticks; struct timex adjust; - /* - * We don't have to take the xtime lock because the cpu - * executing etr_adjust_time is running disabled in - * tasklet context and all other cpus are looping in - * etr_sync_cpu_start. - */ - if (clock > xtime_cc) { + if (clock > old) { /* It is later than we thought. */ - delta = ticks = clock - xtime_cc; + delta = ticks = clock - old; delta = ticks = (delta < delay) ? 0 : delta - delay; delta -= do_div(ticks, CLK_TICKS_PER_JIFFY); - init_timer_cc = init_timer_cc + delta; - jiffies_timer_cc = jiffies_timer_cc + delta; - xtime_cc = xtime_cc + delta; adjust.offset = ticks * (1000000 / HZ); } else { /* It is earlier than we thought. */ - delta = ticks = xtime_cc - clock; + delta = ticks = old - clock; delta -= do_div(ticks, CLK_TICKS_PER_JIFFY); - init_timer_cc = init_timer_cc - delta; - jiffies_timer_cc = jiffies_timer_cc - delta; - xtime_cc = xtime_cc - delta; + delta = -delta; adjust.offset = -ticks * (1000000 / HZ); } + jiffies_timer_cc += delta; if (adjust.offset != 0) { printk(KERN_NOTICE "etr: time adjusted by %li micro-seconds\n", adjust.offset); adjust.modes = ADJ_OFFSET_SINGLESHOT; do_adjtimex(&adjust); } + return delta; } +static struct { + int in_sync; + unsigned long long fixup_cc; +} etr_sync; + static void etr_sync_cpu_start(void *dummy) { - int *in_sync = dummy; - etr_enable_sync_clock(); /* * This looks like a busy wait loop but it isn't. etr_sync_cpus @@ -753,7 +642,7 @@ static void etr_sync_cpu_start(void *dummy) * __udelay will stop the cpu on an enabled wait psw until the * TOD is running again. */ - while (*in_sync == 0) { + while (etr_sync.in_sync == 0) { __udelay(1); /* * A different cpu changes *in_sync. Therefore use @@ -761,14 +650,14 @@ static void etr_sync_cpu_start(void *dummy) */ barrier(); } - if (*in_sync != 1) + if (etr_sync.in_sync != 1) /* Didn't work. Clear per-cpu in sync bit again. */ etr_disable_sync_clock(NULL); /* * This round of TOD syncing is done. Set the clock comparator * to the next tick and let the processor continue. */ - setup_jiffy_timer(); + fixup_clock_comparator(etr_sync.fixup_cc); } static void etr_sync_cpu_end(void *dummy) @@ -783,8 +672,8 @@ static void etr_sync_cpu_end(void *dummy) static int etr_sync_clock(struct etr_aib *aib, int port) { struct etr_aib *sync_port; - unsigned long long clock, delay; - int in_sync, follows; + unsigned long long clock, old_clock, delay, delta; + int follows; int rc; /* Check if the current aib is adjacent to the sync port aib. */ @@ -799,9 +688,9 @@ static int etr_sync_clock(struct etr_aib *aib, int port) * successfully synced the clock. smp_call_function will * return after all other cpus are in etr_sync_cpu_start. */ - in_sync = 0; + memset(&etr_sync, 0, sizeof(etr_sync)); preempt_disable(); - smp_call_function(etr_sync_cpu_start,&in_sync,0,0); + smp_call_function(etr_sync_cpu_start, NULL, 0, 0); local_irq_disable(); etr_enable_sync_clock(); @@ -809,6 +698,7 @@ static int etr_sync_clock(struct etr_aib *aib, int port) __ctl_set_bit(14, 21); __ctl_set_bit(0, 29); clock = ((unsigned long long) (aib->edf2.etv + 1)) << 32; + old_clock = get_clock(); if (set_clock(clock) == 0) { __udelay(1); /* Wait for the clock to start. */ __ctl_clear_bit(0, 29); @@ -817,16 +707,17 @@ static int etr_sync_clock(struct etr_aib *aib, int port) /* Adjust Linux timing variables. */ delay = (unsigned long long) (aib->edf2.etv - sync_port->edf2.etv) << 32; - etr_adjust_time(clock, delay); - setup_jiffy_timer(); + delta = etr_adjust_time(old_clock, clock, delay); + etr_sync.fixup_cc = delta; + fixup_clock_comparator(delta); /* Verify that the clock is properly set. */ if (!etr_aib_follows(sync_port, aib, port)) { /* Didn't work. */ etr_disable_sync_clock(NULL); - in_sync = -EAGAIN; + etr_sync.in_sync = -EAGAIN; rc = -EAGAIN; } else { - in_sync = 1; + etr_sync.in_sync = 1; rc = 0; } } else { @@ -834,7 +725,7 @@ static int etr_sync_clock(struct etr_aib *aib, int port) __ctl_clear_bit(0, 29); __ctl_clear_bit(14, 21); etr_disable_sync_clock(NULL); - in_sync = -EAGAIN; + etr_sync.in_sync = -EAGAIN; rc = -EAGAIN; } local_irq_enable(); diff --git a/arch/s390/kernel/topology.c b/arch/s390/kernel/topology.c new file mode 100644 index 000000000000..12b39b3d9c38 --- /dev/null +++ b/arch/s390/kernel/topology.c @@ -0,0 +1,314 @@ +/* + * Copyright IBM Corp. 2007 + * Author(s): Heiko Carstens <heiko.carstens@de.ibm.com> + */ + +#include <linux/kernel.h> +#include <linux/mm.h> +#include <linux/init.h> +#include <linux/device.h> +#include <linux/bootmem.h> +#include <linux/sched.h> +#include <linux/workqueue.h> +#include <linux/cpu.h> +#include <linux/smp.h> +#include <asm/delay.h> +#include <asm/s390_ext.h> +#include <asm/sysinfo.h> + +#define CPU_BITS 64 +#define NR_MAG 6 + +#define PTF_HORIZONTAL (0UL) +#define PTF_VERTICAL (1UL) +#define PTF_CHECK (2UL) + +struct tl_cpu { + unsigned char reserved0[4]; + unsigned char :6; + unsigned char pp:2; + unsigned char reserved1; + unsigned short origin; + unsigned long mask[CPU_BITS / BITS_PER_LONG]; +}; + +struct tl_container { + unsigned char reserved[8]; +}; + +union tl_entry { + unsigned char nl; + struct tl_cpu cpu; + struct tl_container container; +}; + +struct tl_info { + unsigned char reserved0[2]; + unsigned short length; + unsigned char mag[NR_MAG]; + unsigned char reserved1; + unsigned char mnest; + unsigned char reserved2[4]; + union tl_entry tle[0]; +}; + +struct core_info { + struct core_info *next; + cpumask_t mask; +}; + +static void topology_work_fn(struct work_struct *work); +static struct tl_info *tl_info; +static struct core_info core_info; +static int machine_has_topology; +static int machine_has_topology_irq; +static struct timer_list topology_timer; +static void set_topology_timer(void); +static DECLARE_WORK(topology_work, topology_work_fn); + +cpumask_t cpu_coregroup_map(unsigned int cpu) +{ + struct core_info *core = &core_info; + cpumask_t mask; + + cpus_clear(mask); + if (!machine_has_topology) + return cpu_present_map; + mutex_lock(&smp_cpu_state_mutex); + while (core) { + if (cpu_isset(cpu, core->mask)) { + mask = core->mask; + break; + } + core = core->next; + } + mutex_unlock(&smp_cpu_state_mutex); + if (cpus_empty(mask)) + mask = cpumask_of_cpu(cpu); + return mask; +} + +static void add_cpus_to_core(struct tl_cpu *tl_cpu, struct core_info *core) +{ + unsigned int cpu; + + for (cpu = find_first_bit(&tl_cpu->mask[0], CPU_BITS); + cpu < CPU_BITS; + cpu = find_next_bit(&tl_cpu->mask[0], CPU_BITS, cpu + 1)) + { + unsigned int rcpu, lcpu; + + rcpu = CPU_BITS - 1 - cpu + tl_cpu->origin; + for_each_present_cpu(lcpu) { + if (__cpu_logical_map[lcpu] == rcpu) { + cpu_set(lcpu, core->mask); + smp_cpu_polarization[lcpu] = tl_cpu->pp; + } + } + } +} + +static void clear_cores(void) +{ + struct core_info *core = &core_info; + + while (core) { + cpus_clear(core->mask); + core = core->next; + } +} + +static union tl_entry *next_tle(union tl_entry *tle) +{ + if (tle->nl) + return (union tl_entry *)((struct tl_container *)tle + 1); + else + return (union tl_entry *)((struct tl_cpu *)tle + 1); +} + +static void tl_to_cores(struct tl_info *info) +{ + union tl_entry *tle, *end; + struct core_info *core = &core_info; + + mutex_lock(&smp_cpu_state_mutex); + clear_cores(); + tle = info->tle; + end = (union tl_entry *)((unsigned long)info + info->length); + while (tle < end) { + switch (tle->nl) { + case 5: + case 4: + case 3: + case 2: + break; + case 1: + core = core->next; + break; + case 0: + add_cpus_to_core(&tle->cpu, core); + break; + default: + clear_cores(); + machine_has_topology = 0; + return; + } + tle = next_tle(tle); + } + mutex_unlock(&smp_cpu_state_mutex); +} + +static void topology_update_polarization_simple(void) +{ + int cpu; + + mutex_lock(&smp_cpu_state_mutex); + for_each_present_cpu(cpu) + smp_cpu_polarization[cpu] = POLARIZATION_HRZ; + mutex_unlock(&smp_cpu_state_mutex); +} + +static int ptf(unsigned long fc) +{ + int rc; + + asm volatile( + " .insn rre,0xb9a20000,%1,%1\n" + " ipm %0\n" + " srl %0,28\n" + : "=d" (rc) + : "d" (fc) : "cc"); + return rc; +} + +int topology_set_cpu_management(int fc) +{ + int cpu; + int rc; + + if (!machine_has_topology) + return -EOPNOTSUPP; + if (fc) + rc = ptf(PTF_VERTICAL); + else + rc = ptf(PTF_HORIZONTAL); + if (rc) + return -EBUSY; + for_each_present_cpu(cpu) + smp_cpu_polarization[cpu] = POLARIZATION_UNKNWN; + return rc; +} + +void arch_update_cpu_topology(void) +{ + struct tl_info *info = tl_info; + struct sys_device *sysdev; + int cpu; + + if (!machine_has_topology) { + topology_update_polarization_simple(); + return; + } + stsi(info, 15, 1, 2); + tl_to_cores(info); + for_each_online_cpu(cpu) { + sysdev = get_cpu_sysdev(cpu); + kobject_uevent(&sysdev->kobj, KOBJ_CHANGE); + } +} + +static void topology_work_fn(struct work_struct *work) +{ + arch_reinit_sched_domains(); +} + +void topology_schedule_update(void) +{ + schedule_work(&topology_work); +} + +static void topology_timer_fn(unsigned long ignored) +{ + if (ptf(PTF_CHECK)) + topology_schedule_update(); + set_topology_timer(); +} + +static void set_topology_timer(void) +{ + topology_timer.function = topology_timer_fn; + topology_timer.data = 0; + topology_timer.expires = jiffies + 60 * HZ; + add_timer(&topology_timer); +} + +static void topology_interrupt(__u16 code) +{ + schedule_work(&topology_work); +} + +static int __init init_topology_update(void) +{ + int rc; + + if (!machine_has_topology) { + topology_update_polarization_simple(); + return 0; + } + init_timer_deferrable(&topology_timer); + if (machine_has_topology_irq) { + rc = register_external_interrupt(0x2005, topology_interrupt); + if (rc) + return rc; + ctl_set_bit(0, 8); + } + else + set_topology_timer(); + return 0; +} +__initcall(init_topology_update); + +void __init s390_init_cpu_topology(void) +{ + unsigned long long facility_bits; + struct tl_info *info; + struct core_info *core; + int nr_cores; + int i; + + if (stfle(&facility_bits, 1) <= 0) + return; + if (!(facility_bits & (1ULL << 52)) || !(facility_bits & (1ULL << 61))) + return; + machine_has_topology = 1; + + if (facility_bits & (1ULL << 51)) + machine_has_topology_irq = 1; + + tl_info = alloc_bootmem_pages(PAGE_SIZE); + if (!tl_info) + goto error; + info = tl_info; + stsi(info, 15, 1, 2); + + nr_cores = info->mag[NR_MAG - 2]; + for (i = 0; i < info->mnest - 2; i++) + nr_cores *= info->mag[NR_MAG - 3 - i]; + + printk(KERN_INFO "CPU topology:"); + for (i = 0; i < NR_MAG; i++) + printk(" %d", info->mag[i]); + printk(" / %d\n", info->mnest); + + core = &core_info; + for (i = 0; i < nr_cores; i++) { + core->next = alloc_bootmem(sizeof(struct core_info)); + core = core->next; + if (!core) + goto error; + } + return; +error: + machine_has_topology = 0; + machine_has_topology_irq = 0; +} diff --git a/arch/s390/kernel/traps.c b/arch/s390/kernel/traps.c index 60f728aeaf12..57b607b61100 100644 --- a/arch/s390/kernel/traps.c +++ b/arch/s390/kernel/traps.c @@ -42,11 +42,8 @@ #include <asm/s390_ext.h> #include <asm/lowcore.h> #include <asm/debug.h> +#include "entry.h" -/* Called from entry.S only */ -extern void handle_per_exception(struct pt_regs *regs); - -typedef void pgm_check_handler_t(struct pt_regs *, long); pgm_check_handler_t *pgm_check_table[128]; #ifdef CONFIG_SYSCTL @@ -59,7 +56,6 @@ int sysctl_userprocess_debug = 0; extern pgm_check_handler_t do_protection_exception; extern pgm_check_handler_t do_dat_exception; -extern pgm_check_handler_t do_monitor_call; extern pgm_check_handler_t do_asce_exception; #define stack_pointer ({ void **sp; asm("la %0,0(15)" : "=&d" (sp)); sp; }) @@ -138,7 +134,6 @@ void show_trace(struct task_struct *task, unsigned long *stack) else __show_trace(sp, S390_lowcore.thread_info, S390_lowcore.thread_info + THREAD_SIZE); - printk("\n"); if (!task) task = current; debug_show_held_locks(task); @@ -166,6 +161,15 @@ void show_stack(struct task_struct *task, unsigned long *sp) show_trace(task, sp); } +#ifdef CONFIG_64BIT +void show_last_breaking_event(struct pt_regs *regs) +{ + printk("Last Breaking-Event-Address:\n"); + printk(" [<%016lx>] ", regs->args[0] & PSW_ADDR_INSN); + print_symbol("%s\n", regs->args[0] & PSW_ADDR_INSN); +} +#endif + /* * The architecture-independent dump_stack generator */ @@ -739,6 +743,5 @@ void __init trap_init(void) pgm_check_table[0x15] = &operand_exception; pgm_check_table[0x1C] = &space_switch_exception; pgm_check_table[0x1D] = &hfp_sqrt_exception; - pgm_check_table[0x40] = &do_monitor_call; pfault_irq_init(); } diff --git a/arch/s390/lib/delay.c b/arch/s390/lib/delay.c index 70f2a862b670..eae21a8ac72d 100644 --- a/arch/s390/lib/delay.c +++ b/arch/s390/lib/delay.c @@ -34,7 +34,7 @@ void __delay(unsigned long loops) */ void __udelay(unsigned long usecs) { - u64 end, time, jiffy_timer = 0; + u64 end, time, old_cc = 0; unsigned long flags, cr0, mask, dummy; int irq_context; @@ -43,8 +43,8 @@ void __udelay(unsigned long usecs) local_bh_disable(); local_irq_save(flags); if (raw_irqs_disabled_flags(flags)) { - jiffy_timer = S390_lowcore.jiffy_timer; - S390_lowcore.jiffy_timer = -1ULL - (4096 << 12); + old_cc = S390_lowcore.clock_comparator; + S390_lowcore.clock_comparator = -1ULL; __ctl_store(cr0, 0, 0); dummy = (cr0 & 0xffff00e0) | 0x00000800; __ctl_load(dummy , 0, 0); @@ -55,8 +55,8 @@ void __udelay(unsigned long usecs) end = get_clock() + ((u64) usecs << 12); do { - time = end < S390_lowcore.jiffy_timer ? - end : S390_lowcore.jiffy_timer; + time = end < S390_lowcore.clock_comparator ? + end : S390_lowcore.clock_comparator; set_clock_comparator(time); trace_hardirqs_on(); __load_psw_mask(mask); @@ -65,10 +65,10 @@ void __udelay(unsigned long usecs) if (raw_irqs_disabled_flags(flags)) { __ctl_load(cr0, 0, 0); - S390_lowcore.jiffy_timer = jiffy_timer; + S390_lowcore.clock_comparator = old_cc; } if (!irq_context) _local_bh_enable(); - set_clock_comparator(S390_lowcore.jiffy_timer); + set_clock_comparator(S390_lowcore.clock_comparator); local_irq_restore(flags); } diff --git a/arch/s390/lib/uaccess_pt.c b/arch/s390/lib/uaccess_pt.c index 5efdfe9f5e76..d66215b0fde9 100644 --- a/arch/s390/lib/uaccess_pt.c +++ b/arch/s390/lib/uaccess_pt.c @@ -302,6 +302,10 @@ static size_t copy_in_user_pt(size_t n, void __user *to, pte_t *pte_from, *pte_to; int write_user; + if (segment_eq(get_fs(), KERNEL_DS)) { + memcpy((void __force *) to, (void __force *) from, n); + return 0; + } done = 0; retry: spin_lock(&mm->page_table_lock); @@ -361,18 +365,10 @@ fault: : "0" (-EFAULT), "d" (oparg), "a" (uaddr), \ "m" (*uaddr) : "cc" ); -int futex_atomic_op_pt(int op, int __user *uaddr, int oparg, int *old) +static int __futex_atomic_op_pt(int op, int __user *uaddr, int oparg, int *old) { int oldval = 0, newval, ret; - spin_lock(¤t->mm->page_table_lock); - uaddr = (int __user *) __dat_user_addr((unsigned long) uaddr); - if (!uaddr) { - spin_unlock(¤t->mm->page_table_lock); - return -EFAULT; - } - get_page(virt_to_page(uaddr)); - spin_unlock(¤t->mm->page_table_lock); switch (op) { case FUTEX_OP_SET: __futex_atomic_op("lr %2,%5\n", @@ -397,17 +393,17 @@ int futex_atomic_op_pt(int op, int __user *uaddr, int oparg, int *old) default: ret = -ENOSYS; } - put_page(virt_to_page(uaddr)); - *old = oldval; + if (ret == 0) + *old = oldval; return ret; } -int futex_atomic_cmpxchg_pt(int __user *uaddr, int oldval, int newval) +int futex_atomic_op_pt(int op, int __user *uaddr, int oparg, int *old) { int ret; - if (!current->mm) - return -EFAULT; + if (segment_eq(get_fs(), KERNEL_DS)) + return __futex_atomic_op_pt(op, uaddr, oparg, old); spin_lock(¤t->mm->page_table_lock); uaddr = (int __user *) __dat_user_addr((unsigned long) uaddr); if (!uaddr) { @@ -416,13 +412,40 @@ int futex_atomic_cmpxchg_pt(int __user *uaddr, int oldval, int newval) } get_page(virt_to_page(uaddr)); spin_unlock(¤t->mm->page_table_lock); - asm volatile(" cs %1,%4,0(%5)\n" - "0: lr %0,%1\n" - "1:\n" - EX_TABLE(0b,1b) + ret = __futex_atomic_op_pt(op, uaddr, oparg, old); + put_page(virt_to_page(uaddr)); + return ret; +} + +static int __futex_atomic_cmpxchg_pt(int __user *uaddr, int oldval, int newval) +{ + int ret; + + asm volatile("0: cs %1,%4,0(%5)\n" + "1: lr %0,%1\n" + "2:\n" + EX_TABLE(0b,2b) EX_TABLE(1b,2b) : "=d" (ret), "+d" (oldval), "=m" (*uaddr) : "0" (-EFAULT), "d" (newval), "a" (uaddr), "m" (*uaddr) : "cc", "memory" ); + return ret; +} + +int futex_atomic_cmpxchg_pt(int __user *uaddr, int oldval, int newval) +{ + int ret; + + if (segment_eq(get_fs(), KERNEL_DS)) + return __futex_atomic_cmpxchg_pt(uaddr, oldval, newval); + spin_lock(¤t->mm->page_table_lock); + uaddr = (int __user *) __dat_user_addr((unsigned long) uaddr); + if (!uaddr) { + spin_unlock(¤t->mm->page_table_lock); + return -EFAULT; + } + get_page(virt_to_page(uaddr)); + spin_unlock(¤t->mm->page_table_lock); + ret = __futex_atomic_cmpxchg_pt(uaddr, oldval, newval); put_page(virt_to_page(uaddr)); return ret; } diff --git a/arch/s390/mm/extmem.c b/arch/s390/mm/extmem.c index 880b0ebf894b..ed2af0a3303b 100644 --- a/arch/s390/mm/extmem.c +++ b/arch/s390/mm/extmem.c @@ -289,22 +289,8 @@ __segment_load (char *name, int do_nonshared, unsigned long *addr, unsigned long rc = add_shared_memory(seg->start_addr, seg->end - seg->start_addr + 1); - switch (rc) { - case 0: - break; - case -ENOSPC: - PRINT_WARN("segment_load: not loading segment %s - overlaps " - "storage/segment\n", name); - goto out_free; - case -ERANGE: - PRINT_WARN("segment_load: not loading segment %s - exceeds " - "kernel mapping range\n", name); - goto out_free; - default: - PRINT_WARN("segment_load: not loading segment %s (rc: %d)\n", - name, rc); + if (rc) goto out_free; - } seg->res = kzalloc(sizeof(struct resource), GFP_KERNEL); if (seg->res == NULL) { @@ -582,8 +568,59 @@ out: mutex_unlock(&dcss_lock); } +/* + * print appropriate error message for segment_load()/segment_type() + * return code + */ +void segment_warning(int rc, char *seg_name) +{ + switch (rc) { + case -ENOENT: + PRINT_WARN("cannot load/query segment %s, " + "does not exist\n", seg_name); + break; + case -ENOSYS: + PRINT_WARN("cannot load/query segment %s, " + "not running on VM\n", seg_name); + break; + case -EIO: + PRINT_WARN("cannot load/query segment %s, " + "hardware error\n", seg_name); + break; + case -ENOTSUPP: + PRINT_WARN("cannot load/query segment %s, " + "is a multi-part segment\n", seg_name); + break; + case -ENOSPC: + PRINT_WARN("cannot load/query segment %s, " + "overlaps with storage\n", seg_name); + break; + case -EBUSY: + PRINT_WARN("cannot load/query segment %s, " + "overlaps with already loaded dcss\n", seg_name); + break; + case -EPERM: + PRINT_WARN("cannot load/query segment %s, " + "already loaded in incompatible mode\n", seg_name); + break; + case -ENOMEM: + PRINT_WARN("cannot load/query segment %s, " + "out of memory\n", seg_name); + break; + case -ERANGE: + PRINT_WARN("cannot load/query segment %s, " + "exceeds kernel mapping range\n", seg_name); + break; + default: + PRINT_WARN("cannot load/query segment %s, " + "return value %i\n", seg_name, rc); + break; + } +} + EXPORT_SYMBOL(segment_load); EXPORT_SYMBOL(segment_unload); EXPORT_SYMBOL(segment_save); EXPORT_SYMBOL(segment_type); EXPORT_SYMBOL(segment_modify_shared); +EXPORT_SYMBOL(segment_warning); diff --git a/arch/s390/mm/fault.c b/arch/s390/mm/fault.c index ed13d429a487..2650f46001d0 100644 --- a/arch/s390/mm/fault.c +++ b/arch/s390/mm/fault.c @@ -28,11 +28,11 @@ #include <linux/hardirq.h> #include <linux/kprobes.h> #include <linux/uaccess.h> - #include <asm/system.h> #include <asm/pgtable.h> #include <asm/s390_ext.h> #include <asm/mmu_context.h> +#include "../kernel/entry.h" #ifndef CONFIG_64BIT #define __FAIL_ADDR_MASK 0x7ffff000 @@ -50,8 +50,6 @@ extern int sysctl_userprocess_debug; #endif -extern void die(const char *,struct pt_regs *,long); - #ifdef CONFIG_KPROBES static inline int notify_page_fault(struct pt_regs *regs, long err) { @@ -245,11 +243,6 @@ static void do_sigbus(struct pt_regs *regs, unsigned long error_code, } #ifdef CONFIG_S390_EXEC_PROTECT -extern long sys_sigreturn(struct pt_regs *regs); -extern long sys_rt_sigreturn(struct pt_regs *regs); -extern long sys32_sigreturn(struct pt_regs *regs); -extern long sys32_rt_sigreturn(struct pt_regs *regs); - static int signal_return(struct mm_struct *mm, struct pt_regs *regs, unsigned long address, unsigned long error_code) { @@ -270,15 +263,15 @@ static int signal_return(struct mm_struct *mm, struct pt_regs *regs, #ifdef CONFIG_COMPAT compat = test_tsk_thread_flag(current, TIF_31BIT); if (compat && instruction == 0x0a77) - sys32_sigreturn(regs); + sys32_sigreturn(); else if (compat && instruction == 0x0aad) - sys32_rt_sigreturn(regs); + sys32_rt_sigreturn(); else #endif if (instruction == 0x0a77) - sys_sigreturn(regs); + sys_sigreturn(); else if (instruction == 0x0aad) - sys_rt_sigreturn(regs); + sys_rt_sigreturn(); else { current->thread.prot_addr = address; current->thread.trap_no = error_code; @@ -424,7 +417,7 @@ no_context: } void __kprobes do_protection_exception(struct pt_regs *regs, - unsigned long error_code) + long error_code) { /* Protection exception is supressing, decrement psw address. */ regs->psw.addr -= (error_code >> 16); @@ -440,7 +433,7 @@ void __kprobes do_protection_exception(struct pt_regs *regs, do_exception(regs, 4, 1); } -void __kprobes do_dat_exception(struct pt_regs *regs, unsigned long error_code) +void __kprobes do_dat_exception(struct pt_regs *regs, long error_code) { do_exception(regs, error_code & 0xff, 0); } diff --git a/arch/s390/mm/init.c b/arch/s390/mm/init.c index 8053245fe259..202c952a29b4 100644 --- a/arch/s390/mm/init.c +++ b/arch/s390/mm/init.c @@ -50,7 +50,6 @@ void show_mem(void) printk("Mem-info:\n"); show_free_areas(); - printk("Free swap: %6ldkB\n", nr_swap_pages << (PAGE_SHIFT - 10)); i = max_mapnr; while (i-- > 0) { if (!pfn_valid(i)) diff --git a/arch/sh/kernel/Makefile_32 b/arch/sh/kernel/Makefile_32 index 62bf373266f7..4bbdce36b92b 100644 --- a/arch/sh/kernel/Makefile_32 +++ b/arch/sh/kernel/Makefile_32 @@ -5,7 +5,7 @@ extra-y := head_32.o init_task.o vmlinux.lds obj-y := debugtraps.o io.o io_generic.o irq.o machvec.o process_32.o \ - ptrace_32.o semaphore.o setup.o signal_32.o sys_sh.o sys_sh32.o \ + ptrace_32.o setup.o signal_32.o sys_sh.o sys_sh32.o \ syscalls_32.o time_32.o topology.o traps.o traps_32.o obj-y += cpu/ timers/ diff --git a/arch/sh/kernel/Makefile_64 b/arch/sh/kernel/Makefile_64 index e01283d49cbf..6edf53b93d94 100644 --- a/arch/sh/kernel/Makefile_64 +++ b/arch/sh/kernel/Makefile_64 @@ -1,7 +1,7 @@ extra-y := head_64.o init_task.o vmlinux.lds obj-y := debugtraps.o io.o io_generic.o irq.o machvec.o process_64.o \ - ptrace_64.o semaphore.o setup.o signal_64.o sys_sh.o sys_sh64.o \ + ptrace_64.o setup.o signal_64.o sys_sh.o sys_sh64.o \ syscalls_64.o time_64.o topology.o traps.o traps_64.o obj-y += cpu/ timers/ diff --git a/arch/sh/kernel/semaphore.c b/arch/sh/kernel/semaphore.c deleted file mode 100644 index 184119eeae56..000000000000 --- a/arch/sh/kernel/semaphore.c +++ /dev/null @@ -1,139 +0,0 @@ -/* - * Just taken from alpha implementation. - * This can't work well, perhaps. - */ -/* - * Generic semaphore code. Buyer beware. Do your own - * specific changes in <asm/semaphore-helper.h> - */ - -#include <linux/errno.h> -#include <linux/sched.h> -#include <linux/wait.h> -#include <linux/init.h> -#include <asm/semaphore.h> -#include <asm/semaphore-helper.h> - -DEFINE_SPINLOCK(semaphore_wake_lock); - -/* - * Semaphores are implemented using a two-way counter: - * The "count" variable is decremented for each process - * that tries to sleep, while the "waking" variable is - * incremented when the "up()" code goes to wake up waiting - * processes. - * - * Notably, the inline "up()" and "down()" functions can - * efficiently test if they need to do any extra work (up - * needs to do something only if count was negative before - * the increment operation. - * - * waking_non_zero() (from asm/semaphore.h) must execute - * atomically. - * - * When __up() is called, the count was negative before - * incrementing it, and we need to wake up somebody. - * - * This routine adds one to the count of processes that need to - * wake up and exit. ALL waiting processes actually wake up but - * only the one that gets to the "waking" field first will gate - * through and acquire the semaphore. The others will go back - * to sleep. - * - * Note that these functions are only called when there is - * contention on the lock, and as such all this is the - * "non-critical" part of the whole semaphore business. The - * critical part is the inline stuff in <asm/semaphore.h> - * where we want to avoid any extra jumps and calls. - */ -void __up(struct semaphore *sem) -{ - wake_one_more(sem); - wake_up(&sem->wait); -} - -/* - * Perform the "down" function. Return zero for semaphore acquired, - * return negative for signalled out of the function. - * - * If called from __down, the return is ignored and the wait loop is - * not interruptible. This means that a task waiting on a semaphore - * using "down()" cannot be killed until someone does an "up()" on - * the semaphore. - * - * If called from __down_interruptible, the return value gets checked - * upon return. If the return value is negative then the task continues - * with the negative value in the return register (it can be tested by - * the caller). - * - * Either form may be used in conjunction with "up()". - * - */ - -#define DOWN_VAR \ - struct task_struct *tsk = current; \ - wait_queue_t wait; \ - init_waitqueue_entry(&wait, tsk); - -#define DOWN_HEAD(task_state) \ - \ - \ - tsk->state = (task_state); \ - add_wait_queue(&sem->wait, &wait); \ - \ - /* \ - * Ok, we're set up. sem->count is known to be less than zero \ - * so we must wait. \ - * \ - * We can let go the lock for purposes of waiting. \ - * We re-acquire it after awaking so as to protect \ - * all semaphore operations. \ - * \ - * If "up()" is called before we call waking_non_zero() then \ - * we will catch it right away. If it is called later then \ - * we will have to go through a wakeup cycle to catch it. \ - * \ - * Multiple waiters contend for the semaphore lock to see \ - * who gets to gate through and who has to wait some more. \ - */ \ - for (;;) { - -#define DOWN_TAIL(task_state) \ - tsk->state = (task_state); \ - } \ - tsk->state = TASK_RUNNING; \ - remove_wait_queue(&sem->wait, &wait); - -void __sched __down(struct semaphore * sem) -{ - DOWN_VAR - DOWN_HEAD(TASK_UNINTERRUPTIBLE) - if (waking_non_zero(sem)) - break; - schedule(); - DOWN_TAIL(TASK_UNINTERRUPTIBLE) -} - -int __sched __down_interruptible(struct semaphore * sem) -{ - int ret = 0; - DOWN_VAR - DOWN_HEAD(TASK_INTERRUPTIBLE) - - ret = waking_non_zero_interruptible(sem, tsk); - if (ret) - { - if (ret == 1) - /* ret != 0 only if we get interrupted -arca */ - ret = 0; - break; - } - schedule(); - DOWN_TAIL(TASK_INTERRUPTIBLE) - return ret; -} - -int __down_trylock(struct semaphore * sem) -{ - return waking_non_zero_trylock(sem); -} diff --git a/arch/sh/kernel/sh_ksyms_32.c b/arch/sh/kernel/sh_ksyms_32.c index 45bb333fd9ec..6d405462cee8 100644 --- a/arch/sh/kernel/sh_ksyms_32.c +++ b/arch/sh/kernel/sh_ksyms_32.c @@ -9,7 +9,6 @@ #include <linux/pci.h> #include <linux/irq.h> #include <asm/sections.h> -#include <asm/semaphore.h> #include <asm/processor.h> #include <asm/uaccess.h> #include <asm/checksum.h> @@ -48,12 +47,6 @@ EXPORT_SYMBOL(__copy_user); EXPORT_SYMBOL(get_vm_area); #endif -/* semaphore exports */ -EXPORT_SYMBOL(__up); -EXPORT_SYMBOL(__down); -EXPORT_SYMBOL(__down_interruptible); -EXPORT_SYMBOL(__down_trylock); - EXPORT_SYMBOL(__udelay); EXPORT_SYMBOL(__ndelay); EXPORT_SYMBOL(__const_udelay); diff --git a/arch/sh/kernel/sh_ksyms_64.c b/arch/sh/kernel/sh_ksyms_64.c index b6410ce4bd1d..a310c9707f03 100644 --- a/arch/sh/kernel/sh_ksyms_64.c +++ b/arch/sh/kernel/sh_ksyms_64.c @@ -16,7 +16,6 @@ #include <linux/in6.h> #include <linux/interrupt.h> #include <linux/screen_info.h> -#include <asm/semaphore.h> #include <asm/processor.h> #include <asm/uaccess.h> #include <asm/checksum.h> @@ -37,9 +36,6 @@ EXPORT_SYMBOL(csum_partial_copy_nocheck); EXPORT_SYMBOL(screen_info); #endif -EXPORT_SYMBOL(__down); -EXPORT_SYMBOL(__down_trylock); -EXPORT_SYMBOL(__up); EXPORT_SYMBOL(__put_user_asm_l); EXPORT_SYMBOL(__get_user_asm_l); EXPORT_SYMBOL(copy_page); diff --git a/arch/sparc/kernel/Makefile b/arch/sparc/kernel/Makefile index bf1b15d3f6f5..2712bb166f6f 100644 --- a/arch/sparc/kernel/Makefile +++ b/arch/sparc/kernel/Makefile @@ -12,7 +12,7 @@ obj-y := entry.o wof.o wuf.o etrap.o rtrap.o traps.o $(IRQ_OBJS) \ sys_sparc.o sunos_asm.o systbls.o \ time.o windows.o cpu.o devices.o sclow.o \ tadpole.o tick14.o ptrace.o sys_solaris.o \ - unaligned.o una_asm.o muldiv.o semaphore.o \ + unaligned.o una_asm.o muldiv.o \ prom.o of_device.o devres.o devres-y = ../../../kernel/irq/devres.o diff --git a/arch/sparc/kernel/semaphore.c b/arch/sparc/kernel/semaphore.c deleted file mode 100644 index 0c37c1a7cd7e..000000000000 --- a/arch/sparc/kernel/semaphore.c +++ /dev/null @@ -1,155 +0,0 @@ -/* $Id: semaphore.c,v 1.7 2001/04/18 21:06:05 davem Exp $ */ - -/* sparc32 semaphore implementation, based on i386 version */ - -#include <linux/sched.h> -#include <linux/errno.h> -#include <linux/init.h> - -#include <asm/semaphore.h> - -/* - * Semaphores are implemented using a two-way counter: - * The "count" variable is decremented for each process - * that tries to acquire the semaphore, while the "sleeping" - * variable is a count of such acquires. - * - * Notably, the inline "up()" and "down()" functions can - * efficiently test if they need to do any extra work (up - * needs to do something only if count was negative before - * the increment operation. - * - * "sleeping" and the contention routine ordering is - * protected by the semaphore spinlock. - * - * Note that these functions are only called when there is - * contention on the lock, and as such all this is the - * "non-critical" part of the whole semaphore business. The - * critical part is the inline stuff in <asm/semaphore.h> - * where we want to avoid any extra jumps and calls. - */ - -/* - * Logic: - * - only on a boundary condition do we need to care. When we go - * from a negative count to a non-negative, we wake people up. - * - when we go from a non-negative count to a negative do we - * (a) synchronize with the "sleeper" count and (b) make sure - * that we're on the wakeup list before we synchronize so that - * we cannot lose wakeup events. - */ - -void __up(struct semaphore *sem) -{ - wake_up(&sem->wait); -} - -static DEFINE_SPINLOCK(semaphore_lock); - -void __sched __down(struct semaphore * sem) -{ - struct task_struct *tsk = current; - DECLARE_WAITQUEUE(wait, tsk); - tsk->state = TASK_UNINTERRUPTIBLE; - add_wait_queue_exclusive(&sem->wait, &wait); - - spin_lock_irq(&semaphore_lock); - sem->sleepers++; - for (;;) { - int sleepers = sem->sleepers; - - /* - * Add "everybody else" into it. They aren't - * playing, because we own the spinlock. - */ - if (!atomic24_add_negative(sleepers - 1, &sem->count)) { - sem->sleepers = 0; - break; - } - sem->sleepers = 1; /* us - see -1 above */ - spin_unlock_irq(&semaphore_lock); - - schedule(); - tsk->state = TASK_UNINTERRUPTIBLE; - spin_lock_irq(&semaphore_lock); - } - spin_unlock_irq(&semaphore_lock); - remove_wait_queue(&sem->wait, &wait); - tsk->state = TASK_RUNNING; - wake_up(&sem->wait); -} - -int __sched __down_interruptible(struct semaphore * sem) -{ - int retval = 0; - struct task_struct *tsk = current; - DECLARE_WAITQUEUE(wait, tsk); - tsk->state = TASK_INTERRUPTIBLE; - add_wait_queue_exclusive(&sem->wait, &wait); - - spin_lock_irq(&semaphore_lock); - sem->sleepers ++; - for (;;) { - int sleepers = sem->sleepers; - - /* - * With signals pending, this turns into - * the trylock failure case - we won't be - * sleeping, and we* can't get the lock as - * it has contention. Just correct the count - * and exit. - */ - if (signal_pending(current)) { - retval = -EINTR; - sem->sleepers = 0; - atomic24_add(sleepers, &sem->count); - break; - } - - /* - * Add "everybody else" into it. They aren't - * playing, because we own the spinlock. The - * "-1" is because we're still hoping to get - * the lock. - */ - if (!atomic24_add_negative(sleepers - 1, &sem->count)) { - sem->sleepers = 0; - break; - } - sem->sleepers = 1; /* us - see -1 above */ - spin_unlock_irq(&semaphore_lock); - - schedule(); - tsk->state = TASK_INTERRUPTIBLE; - spin_lock_irq(&semaphore_lock); - } - spin_unlock_irq(&semaphore_lock); - tsk->state = TASK_RUNNING; - remove_wait_queue(&sem->wait, &wait); - wake_up(&sem->wait); - return retval; -} - -/* - * Trylock failed - make sure we correct for - * having decremented the count. - */ -int __down_trylock(struct semaphore * sem) -{ - int sleepers; - unsigned long flags; - - spin_lock_irqsave(&semaphore_lock, flags); - sleepers = sem->sleepers + 1; - sem->sleepers = 0; - - /* - * Add "everybody else" and us into it. They aren't - * playing, because we own the spinlock. - */ - if (!atomic24_add_negative(sleepers, &sem->count)) - wake_up(&sem->wait); - - spin_unlock_irqrestore(&semaphore_lock, flags); - return 1; -} diff --git a/arch/sparc/kernel/sparc_ksyms.c b/arch/sparc/kernel/sparc_ksyms.c index c1025e551650..97b1de0e9094 100644 --- a/arch/sparc/kernel/sparc_ksyms.c +++ b/arch/sparc/kernel/sparc_ksyms.c @@ -107,11 +107,6 @@ EXPORT_SYMBOL(___rw_read_try); EXPORT_SYMBOL(___rw_read_exit); EXPORT_SYMBOL(___rw_write_enter); #endif -/* semaphores */ -EXPORT_SYMBOL(__up); -EXPORT_SYMBOL(__down); -EXPORT_SYMBOL(__down_trylock); -EXPORT_SYMBOL(__down_interruptible); EXPORT_SYMBOL(sparc_valid_addr_bitmap); EXPORT_SYMBOL(phys_base); diff --git a/arch/sparc64/kernel/Makefile b/arch/sparc64/kernel/Makefile index 1bf5b187de49..459462e80a12 100644 --- a/arch/sparc64/kernel/Makefile +++ b/arch/sparc64/kernel/Makefile @@ -10,7 +10,7 @@ extra-y := head.o init_task.o vmlinux.lds obj-y := process.o setup.o cpu.o idprom.o \ traps.o auxio.o una_asm.o sysfs.o iommu.o \ irq.o ptrace.o time.o sys_sparc.o signal.o \ - unaligned.o central.o pci.o starfire.o semaphore.o \ + unaligned.o central.o pci.o starfire.o \ power.o sbus.o sparc64_ksyms.o chmc.o \ visemul.o prom.o of_device.o hvapi.o sstate.o mdesc.o diff --git a/arch/sparc64/kernel/semaphore.c b/arch/sparc64/kernel/semaphore.c deleted file mode 100644 index 9974a6899551..000000000000 --- a/arch/sparc64/kernel/semaphore.c +++ /dev/null @@ -1,254 +0,0 @@ -/* semaphore.c: Sparc64 semaphore implementation. - * - * This is basically the PPC semaphore scheme ported to use - * the sparc64 atomic instructions, so see the PPC code for - * credits. - */ - -#include <linux/sched.h> -#include <linux/errno.h> -#include <linux/init.h> - -/* - * Atomically update sem->count. - * This does the equivalent of the following: - * - * old_count = sem->count; - * tmp = MAX(old_count, 0) + incr; - * sem->count = tmp; - * return old_count; - */ -static inline int __sem_update_count(struct semaphore *sem, int incr) -{ - int old_count, tmp; - - __asm__ __volatile__("\n" -" ! __sem_update_count old_count(%0) tmp(%1) incr(%4) &sem->count(%3)\n" -"1: ldsw [%3], %0\n" -" mov %0, %1\n" -" cmp %0, 0\n" -" movl %%icc, 0, %1\n" -" add %1, %4, %1\n" -" cas [%3], %0, %1\n" -" cmp %0, %1\n" -" membar #StoreLoad | #StoreStore\n" -" bne,pn %%icc, 1b\n" -" nop\n" - : "=&r" (old_count), "=&r" (tmp), "=m" (sem->count) - : "r" (&sem->count), "r" (incr), "m" (sem->count) - : "cc"); - - return old_count; -} - -static void __up(struct semaphore *sem) -{ - __sem_update_count(sem, 1); - wake_up(&sem->wait); -} - -void up(struct semaphore *sem) -{ - /* This atomically does: - * old_val = sem->count; - * new_val = sem->count + 1; - * sem->count = new_val; - * if (old_val < 0) - * __up(sem); - * - * The (old_val < 0) test is equivalent to - * the more straightforward (new_val <= 0), - * but it is easier to test the former because - * of how the CAS instruction works. - */ - - __asm__ __volatile__("\n" -" ! up sem(%0)\n" -" membar #StoreLoad | #LoadLoad\n" -"1: lduw [%0], %%g1\n" -" add %%g1, 1, %%g7\n" -" cas [%0], %%g1, %%g7\n" -" cmp %%g1, %%g7\n" -" bne,pn %%icc, 1b\n" -" addcc %%g7, 1, %%g0\n" -" membar #StoreLoad | #StoreStore\n" -" ble,pn %%icc, 3f\n" -" nop\n" -"2:\n" -" .subsection 2\n" -"3: mov %0, %%g1\n" -" save %%sp, -160, %%sp\n" -" call %1\n" -" mov %%g1, %%o0\n" -" ba,pt %%xcc, 2b\n" -" restore\n" -" .previous\n" - : : "r" (sem), "i" (__up) - : "g1", "g2", "g3", "g7", "memory", "cc"); -} - -static void __sched __down(struct semaphore * sem) -{ - struct task_struct *tsk = current; - DECLARE_WAITQUEUE(wait, tsk); - - tsk->state = TASK_UNINTERRUPTIBLE; - add_wait_queue_exclusive(&sem->wait, &wait); - - while (__sem_update_count(sem, -1) <= 0) { - schedule(); - tsk->state = TASK_UNINTERRUPTIBLE; - } - remove_wait_queue(&sem->wait, &wait); - tsk->state = TASK_RUNNING; - - wake_up(&sem->wait); -} - -void __sched down(struct semaphore *sem) -{ - might_sleep(); - /* This atomically does: - * old_val = sem->count; - * new_val = sem->count - 1; - * sem->count = new_val; - * if (old_val < 1) - * __down(sem); - * - * The (old_val < 1) test is equivalent to - * the more straightforward (new_val < 0), - * but it is easier to test the former because - * of how the CAS instruction works. - */ - - __asm__ __volatile__("\n" -" ! down sem(%0)\n" -"1: lduw [%0], %%g1\n" -" sub %%g1, 1, %%g7\n" -" cas [%0], %%g1, %%g7\n" -" cmp %%g1, %%g7\n" -" bne,pn %%icc, 1b\n" -" cmp %%g7, 1\n" -" membar #StoreLoad | #StoreStore\n" -" bl,pn %%icc, 3f\n" -" nop\n" -"2:\n" -" .subsection 2\n" -"3: mov %0, %%g1\n" -" save %%sp, -160, %%sp\n" -" call %1\n" -" mov %%g1, %%o0\n" -" ba,pt %%xcc, 2b\n" -" restore\n" -" .previous\n" - : : "r" (sem), "i" (__down) - : "g1", "g2", "g3", "g7", "memory", "cc"); -} - -int down_trylock(struct semaphore *sem) -{ - int ret; - - /* This atomically does: - * old_val = sem->count; - * new_val = sem->count - 1; - * if (old_val < 1) { - * ret = 1; - * } else { - * sem->count = new_val; - * ret = 0; - * } - * - * The (old_val < 1) test is equivalent to - * the more straightforward (new_val < 0), - * but it is easier to test the former because - * of how the CAS instruction works. - */ - - __asm__ __volatile__("\n" -" ! down_trylock sem(%1) ret(%0)\n" -"1: lduw [%1], %%g1\n" -" sub %%g1, 1, %%g7\n" -" cmp %%g1, 1\n" -" bl,pn %%icc, 2f\n" -" mov 1, %0\n" -" cas [%1], %%g1, %%g7\n" -" cmp %%g1, %%g7\n" -" bne,pn %%icc, 1b\n" -" mov 0, %0\n" -" membar #StoreLoad | #StoreStore\n" -"2:\n" - : "=&r" (ret) - : "r" (sem) - : "g1", "g7", "memory", "cc"); - - return ret; -} - -static int __sched __down_interruptible(struct semaphore * sem) -{ - int retval = 0; - struct task_struct *tsk = current; - DECLARE_WAITQUEUE(wait, tsk); - - tsk->state = TASK_INTERRUPTIBLE; - add_wait_queue_exclusive(&sem->wait, &wait); - - while (__sem_update_count(sem, -1) <= 0) { - if (signal_pending(current)) { - __sem_update_count(sem, 0); - retval = -EINTR; - break; - } - schedule(); - tsk->state = TASK_INTERRUPTIBLE; - } - tsk->state = TASK_RUNNING; - remove_wait_queue(&sem->wait, &wait); - wake_up(&sem->wait); - return retval; -} - -int __sched down_interruptible(struct semaphore *sem) -{ - int ret = 0; - - might_sleep(); - /* This atomically does: - * old_val = sem->count; - * new_val = sem->count - 1; - * sem->count = new_val; - * if (old_val < 1) - * ret = __down_interruptible(sem); - * - * The (old_val < 1) test is equivalent to - * the more straightforward (new_val < 0), - * but it is easier to test the former because - * of how the CAS instruction works. - */ - - __asm__ __volatile__("\n" -" ! down_interruptible sem(%2) ret(%0)\n" -"1: lduw [%2], %%g1\n" -" sub %%g1, 1, %%g7\n" -" cas [%2], %%g1, %%g7\n" -" cmp %%g1, %%g7\n" -" bne,pn %%icc, 1b\n" -" cmp %%g7, 1\n" -" membar #StoreLoad | #StoreStore\n" -" bl,pn %%icc, 3f\n" -" nop\n" -"2:\n" -" .subsection 2\n" -"3: mov %2, %%g1\n" -" save %%sp, -160, %%sp\n" -" call %3\n" -" mov %%g1, %%o0\n" -" ba,pt %%xcc, 2b\n" -" restore\n" -" .previous\n" - : "=r" (ret) - : "0" (ret), "r" (sem), "i" (__down_interruptible) - : "g1", "g2", "g3", "g7", "memory", "cc"); - return ret; -} diff --git a/arch/sparc64/kernel/sparc64_ksyms.c b/arch/sparc64/kernel/sparc64_ksyms.c index 51fa773f38c9..051b8d9cb989 100644 --- a/arch/sparc64/kernel/sparc64_ksyms.c +++ b/arch/sparc64/kernel/sparc64_ksyms.c @@ -130,12 +130,6 @@ EXPORT_SYMBOL(_mcount); EXPORT_SYMBOL(sparc64_get_clock_tick); -/* semaphores */ -EXPORT_SYMBOL(down); -EXPORT_SYMBOL(down_trylock); -EXPORT_SYMBOL(down_interruptible); -EXPORT_SYMBOL(up); - /* RW semaphores */ EXPORT_SYMBOL(__down_read); EXPORT_SYMBOL(__down_read_trylock); diff --git a/arch/um/Kconfig.i386 b/arch/um/Kconfig.i386 index 3cd8a04d66d8..e09edfa560da 100644 --- a/arch/um/Kconfig.i386 +++ b/arch/um/Kconfig.i386 @@ -19,10 +19,6 @@ config 64BIT bool default n -config SEMAPHORE_SLEEPERS - bool - default y - config 3_LEVEL_PGTABLES bool "Three-level pagetables (EXPERIMENTAL)" default n diff --git a/arch/um/Kconfig.x86_64 b/arch/um/Kconfig.x86_64 index 6533b349f061..3fbe69e359ed 100644 --- a/arch/um/Kconfig.x86_64 +++ b/arch/um/Kconfig.x86_64 @@ -11,10 +11,6 @@ config RWSEM_GENERIC_SPINLOCK bool default y -config SEMAPHORE_SLEEPERS - bool - default y - config 3_LEVEL_PGTABLES bool default y diff --git a/arch/um/sys-i386/ksyms.c b/arch/um/sys-i386/ksyms.c index 2a1eac1859ce..bfbefd30db8f 100644 --- a/arch/um/sys-i386/ksyms.c +++ b/arch/um/sys-i386/ksyms.c @@ -1,17 +1,5 @@ #include "linux/module.h" -#include "linux/in6.h" -#include "linux/rwsem.h" -#include "asm/byteorder.h" -#include "asm/delay.h" -#include "asm/semaphore.h" -#include "asm/uaccess.h" #include "asm/checksum.h" -#include "asm/errno.h" - -EXPORT_SYMBOL(__down_failed); -EXPORT_SYMBOL(__down_failed_interruptible); -EXPORT_SYMBOL(__down_failed_trylock); -EXPORT_SYMBOL(__up_wakeup); /* Networking helper routines. */ EXPORT_SYMBOL(csum_partial); diff --git a/arch/um/sys-ppc/Makefile b/arch/um/sys-ppc/Makefile index 08901526e893..b8bc844fd2c4 100644 --- a/arch/um/sys-ppc/Makefile +++ b/arch/um/sys-ppc/Makefile @@ -3,7 +3,7 @@ OBJ = built-in.o .S.o: $(CC) $(KBUILD_AFLAGS) -D__ASSEMBLY__ -D__UM_PPC__ -c $< -o $*.o -OBJS = ptrace.o sigcontext.o semaphore.o checksum.o miscthings.o misc.o \ +OBJS = ptrace.o sigcontext.o checksum.o miscthings.o misc.o \ ptrace_user.o sysrq.o EXTRA_AFLAGS := -DCONFIG_PPC32 -I. -I$(srctree)/arch/ppc/kernel @@ -20,10 +20,6 @@ ptrace_user.o: ptrace_user.c sigcontext.o: sigcontext.c $(CC) $(USER_CFLAGS) $(EXTRA_CFLAGS) -c -o $@ $< -semaphore.c: - rm -f $@ - ln -s $(srctree)/arch/ppc/kernel/$@ $@ - checksum.S: rm -f $@ ln -s $(srctree)/arch/ppc/lib/$@ $@ @@ -66,4 +62,4 @@ misc.o: misc.S ppc_defs.h $(CC) $(EXTRA_AFLAGS) $(KBUILD_AFLAGS) -D__ASSEMBLY__ -D__UM_PPC__ -c $< -o $*.o rm -f asm -clean-files := $(OBJS) ppc_defs.h checksum.S semaphore.c mk_defs.c +clean-files := $(OBJS) ppc_defs.h checksum.S mk_defs.c diff --git a/arch/um/sys-x86_64/ksyms.c b/arch/um/sys-x86_64/ksyms.c index 12c593607c59..4d7d1a812d8f 100644 --- a/arch/um/sys-x86_64/ksyms.c +++ b/arch/um/sys-x86_64/ksyms.c @@ -1,16 +1,5 @@ #include "linux/module.h" -#include "linux/in6.h" -#include "linux/rwsem.h" -#include "asm/byteorder.h" -#include "asm/semaphore.h" -#include "asm/uaccess.h" -#include "asm/checksum.h" -#include "asm/errno.h" - -EXPORT_SYMBOL(__down_failed); -EXPORT_SYMBOL(__down_failed_interruptible); -EXPORT_SYMBOL(__down_failed_trylock); -EXPORT_SYMBOL(__up_wakeup); +#include "asm/string.h" /*XXX: we need them because they would be exported by x86_64 */ EXPORT_SYMBOL(__memcpy); diff --git a/arch/v850/kernel/Makefile b/arch/v850/kernel/Makefile index 3930482bddc4..da5889c53576 100644 --- a/arch/v850/kernel/Makefile +++ b/arch/v850/kernel/Makefile @@ -11,7 +11,7 @@ extra-y := head.o init_task.o vmlinux.lds -obj-y += intv.o entry.o process.o syscalls.o time.o semaphore.o setup.o \ +obj-y += intv.o entry.o process.o syscalls.o time.o setup.o \ signal.o irq.o mach.o ptrace.o bug.o obj-$(CONFIG_MODULES) += module.o v850_ksyms.o # chip-specific code diff --git a/arch/v850/kernel/semaphore.c b/arch/v850/kernel/semaphore.c deleted file mode 100644 index fc89fd661c99..000000000000 --- a/arch/v850/kernel/semaphore.c +++ /dev/null @@ -1,166 +0,0 @@ -/* - * arch/v850/kernel/semaphore.c -- Semaphore support - * - * Copyright (C) 1998-2000 IBM Corporation - * Copyright (C) 1999 Linus Torvalds - * - * This file is subject to the terms and conditions of the GNU General - * Public License. See the file COPYING in the main directory of this - * archive for more details. - * - * This file is a copy of the s390 version, arch/s390/kernel/semaphore.c - * Author(s): Martin Schwidefsky - * which was derived from the i386 version, linux/arch/i386/kernel/semaphore.c - */ - -#include <linux/errno.h> -#include <linux/sched.h> -#include <linux/init.h> - -#include <asm/semaphore.h> - -/* - * Semaphores are implemented using a two-way counter: - * The "count" variable is decremented for each process - * that tries to acquire the semaphore, while the "sleeping" - * variable is a count of such acquires. - * - * Notably, the inline "up()" and "down()" functions can - * efficiently test if they need to do any extra work (up - * needs to do something only if count was negative before - * the increment operation. - * - * "sleeping" and the contention routine ordering is - * protected by the semaphore spinlock. - * - * Note that these functions are only called when there is - * contention on the lock, and as such all this is the - * "non-critical" part of the whole semaphore business. The - * critical part is the inline stuff in <asm/semaphore.h> - * where we want to avoid any extra jumps and calls. - */ - -/* - * Logic: - * - only on a boundary condition do we need to care. When we go - * from a negative count to a non-negative, we wake people up. - * - when we go from a non-negative count to a negative do we - * (a) synchronize with the "sleeper" count and (b) make sure - * that we're on the wakeup list before we synchronize so that - * we cannot lose wakeup events. - */ - -void __up(struct semaphore *sem) -{ - wake_up(&sem->wait); -} - -static DEFINE_SPINLOCK(semaphore_lock); - -void __sched __down(struct semaphore * sem) -{ - struct task_struct *tsk = current; - DECLARE_WAITQUEUE(wait, tsk); - tsk->state = TASK_UNINTERRUPTIBLE; - add_wait_queue_exclusive(&sem->wait, &wait); - - spin_lock_irq(&semaphore_lock); - sem->sleepers++; - for (;;) { - int sleepers = sem->sleepers; - - /* - * Add "everybody else" into it. They aren't - * playing, because we own the spinlock. - */ - if (!atomic_add_negative(sleepers - 1, &sem->count)) { - sem->sleepers = 0; - break; - } - sem->sleepers = 1; /* us - see -1 above */ - spin_unlock_irq(&semaphore_lock); - - schedule(); - tsk->state = TASK_UNINTERRUPTIBLE; - spin_lock_irq(&semaphore_lock); - } - spin_unlock_irq(&semaphore_lock); - remove_wait_queue(&sem->wait, &wait); - tsk->state = TASK_RUNNING; - wake_up(&sem->wait); -} - -int __sched __down_interruptible(struct semaphore * sem) -{ - int retval = 0; - struct task_struct *tsk = current; - DECLARE_WAITQUEUE(wait, tsk); - tsk->state = TASK_INTERRUPTIBLE; - add_wait_queue_exclusive(&sem->wait, &wait); - - spin_lock_irq(&semaphore_lock); - sem->sleepers ++; - for (;;) { - int sleepers = sem->sleepers; - - /* - * With signals pending, this turns into - * the trylock failure case - we won't be - * sleeping, and we* can't get the lock as - * it has contention. Just correct the count - * and exit. - */ - if (signal_pending(current)) { - retval = -EINTR; - sem->sleepers = 0; - atomic_add(sleepers, &sem->count); - break; - } - - /* - * Add "everybody else" into it. They aren't - * playing, because we own the spinlock. The - * "-1" is because we're still hoping to get - * the lock. - */ - if (!atomic_add_negative(sleepers - 1, &sem->count)) { - sem->sleepers = 0; - break; - } - sem->sleepers = 1; /* us - see -1 above */ - spin_unlock_irq(&semaphore_lock); - - schedule(); - tsk->state = TASK_INTERRUPTIBLE; - spin_lock_irq(&semaphore_lock); - } - spin_unlock_irq(&semaphore_lock); - tsk->state = TASK_RUNNING; - remove_wait_queue(&sem->wait, &wait); - wake_up(&sem->wait); - return retval; -} - -/* - * Trylock failed - make sure we correct for - * having decremented the count. - */ -int __down_trylock(struct semaphore * sem) -{ - unsigned long flags; - int sleepers; - - spin_lock_irqsave(&semaphore_lock, flags); - sleepers = sem->sleepers + 1; - sem->sleepers = 0; - - /* - * Add "everybody else" and us into it. They aren't - * playing, because we own the spinlock. - */ - if (!atomic_add_negative(sleepers, &sem->count)) - wake_up(&sem->wait); - - spin_unlock_irqrestore(&semaphore_lock, flags); - return 1; -} diff --git a/arch/v850/kernel/v850_ksyms.c b/arch/v850/kernel/v850_ksyms.c index 93575fdc874d..8d386a5dbc4a 100644 --- a/arch/v850/kernel/v850_ksyms.c +++ b/arch/v850/kernel/v850_ksyms.c @@ -11,7 +11,6 @@ #include <asm/pgalloc.h> #include <asm/irq.h> #include <asm/io.h> -#include <asm/semaphore.h> #include <asm/checksum.h> #include <asm/current.h> @@ -34,12 +33,6 @@ EXPORT_SYMBOL (memset); EXPORT_SYMBOL (memcpy); EXPORT_SYMBOL (memmove); -/* semaphores */ -EXPORT_SYMBOL (__down); -EXPORT_SYMBOL (__down_interruptible); -EXPORT_SYMBOL (__down_trylock); -EXPORT_SYMBOL (__up); - /* * libgcc functions - functions that are used internally by the * compiler... (prototypes are not correct though, but that diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index 5c4c8d7a46cc..2a59dbb28248 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -54,9 +54,6 @@ config STACKTRACE_SUPPORT config HAVE_LATENCYTOP_SUPPORT def_bool y -config SEMAPHORE_SLEEPERS - def_bool y - config FAST_CMPXCHG_LOCAL bool default y @@ -118,7 +115,7 @@ config ARCH_HAS_CPU_RELAX def_bool y config HAVE_SETUP_PER_CPU_AREA - def_bool X86_64 + def_bool X86_64 || (X86_SMP && !X86_VOYAGER) config ARCH_HIBERNATION_POSSIBLE def_bool y @@ -172,7 +169,7 @@ config X86_64_SMP config X86_HT bool depends on SMP - depends on (X86_32 && !(X86_VISWS || X86_VOYAGER)) || (X86_64 && !MK8) + depends on (X86_32 && !(X86_VISWS || X86_VOYAGER)) || X86_64 default y config X86_BIOS_REBOOT @@ -182,7 +179,7 @@ config X86_BIOS_REBOOT config X86_TRAMPOLINE bool - depends on X86_SMP || (X86_VOYAGER && SMP) + depends on X86_SMP || (X86_VOYAGER && SMP) || (64BIT && ACPI_SLEEP) default y config KTIME_SCALAR @@ -242,8 +239,7 @@ config X86_ELAN config X86_VOYAGER bool "Voyager (NCR)" - depends on X86_32 - select SMP if !BROKEN + depends on X86_32 && (SMP || BROKEN) help Voyager is an MCA-based 32-way capable SMP architecture proprietary to NCR Corp. Machine classes 345x/35xx/4100/51xx are Voyager-based. @@ -255,9 +251,8 @@ config X86_VOYAGER config X86_NUMAQ bool "NUMAQ (IBM/Sequent)" - select SMP + depends on SMP && X86_32 select NUMA - depends on X86_32 help This option is used for getting Linux to run on a (IBM/Sequent) NUMA multiquad box. This changes the way that processors are bootstrapped, @@ -328,8 +323,9 @@ config X86_RDC321X config X86_VSMP bool "Support for ScaleMP vSMP" - depends on X86_64 && PCI - help + select PARAVIRT + depends on X86_64 + help Support for ScaleMP vSMP systems. Say 'Y' here if this kernel is supposed to run on these EM64T-based machines. Only choose this option if you have one of these machines. @@ -384,6 +380,35 @@ config PARAVIRT endif +config MEMTEST_BOOTPARAM + bool "Memtest boot parameter" + depends on X86_64 + default y + help + This option adds a kernel parameter 'memtest', which allows memtest + to be disabled at boot. If this option is selected, memtest + functionality can be disabled with memtest=0 on the kernel + command line. The purpose of this option is to allow a single + kernel image to be distributed with memtest built in, but not + necessarily enabled. + + If you are unsure how to answer this question, answer Y. + +config MEMTEST_BOOTPARAM_VALUE + int "Memtest boot parameter default value (0-4)" + depends on MEMTEST_BOOTPARAM + range 0 4 + default 0 + help + This option sets the default value for the kernel parameter + 'memtest', which allows memtest to be disabled at boot. If this + option is set to 0 (zero), the memtest kernel parameter will + default to 0, disabling memtest at bootup. If this option is + set to 4, the memtest kernel parameter will default to 4, + enabling memtest at bootup, and use that as pattern number. + + If you are unsure how to answer this question, answer 0. + config ACPI_SRAT def_bool y depends on X86_32 && ACPI && NUMA && (X86_SUMMIT || X86_GENERICARCH) @@ -508,7 +533,7 @@ config NR_CPUS config SCHED_SMT bool "SMT (Hyperthreading) scheduler support" - depends on (X86_64 && SMP) || (X86_32 && X86_HT) + depends on X86_HT help SMT scheduler support improves the CPU scheduler's decision making when dealing with Intel Pentium 4 chips with HyperThreading at a @@ -518,7 +543,7 @@ config SCHED_SMT config SCHED_MC def_bool y prompt "Multi-core scheduler support" - depends on (X86_64 && SMP) || (X86_32 && X86_HT) + depends on X86_HT help Multi-core scheduler support improves the CPU scheduler's decision making when dealing with multi-core CPU chips at a cost of slightly @@ -887,7 +912,7 @@ config NUMA_EMU number of nodes. This is only useful for debugging. config NODES_SHIFT - int + int "Max num nodes shift(1-15)" range 1 15 if X86_64 default "6" if X86_64 default "4" if X86_NUMAQ @@ -1011,6 +1036,21 @@ config MTRR See <file:Documentation/mtrr.txt> for more information. +config X86_PAT + def_bool y + prompt "x86 PAT support" + depends on MTRR && NONPROMISC_DEVMEM + help + Use PAT attributes to setup page level cache control. + + PATs are the modern equivalents of MTRRs and are much more + flexible than MTRRs. + + Say N here if you see bootup problems (boot crash, boot hang, + spontaneous reboots) or a non-working video driver. + + If unsure, say Y. + config EFI def_bool n prompt "EFI runtime service support" @@ -1079,6 +1119,7 @@ source kernel/Kconfig.hz config KEXEC bool "kexec system call" + depends on X86_64 || X86_BIOS_REBOOT help kexec is a system call that implements the ability to shutdown your current kernel, and to start another kernel. It is like a reboot @@ -1380,7 +1421,7 @@ endmenu menu "Bus options (PCI etc.)" config PCI - bool "PCI support" if !X86_VISWS + bool "PCI support" if !X86_VISWS && !X86_VSMP depends on !X86_VOYAGER default y select ARCH_SUPPORTS_MSI if (X86_LOCAL_APIC && X86_IO_APIC) diff --git a/arch/x86/Kconfig.cpu b/arch/x86/Kconfig.cpu index 9304bfba7d45..57072f2716f9 100644 --- a/arch/x86/Kconfig.cpu +++ b/arch/x86/Kconfig.cpu @@ -388,7 +388,7 @@ config X86_OOSTORE # config X86_P6_NOP def_bool y - depends on (X86_64 || !X86_GENERIC) && (M686 || MPENTIUMII || MPENTIUMIII || MPENTIUMM || MCORE2 || MPENTIUM4) + depends on (X86_64 || !X86_GENERIC) && (M686 || MPENTIUMII || MPENTIUMIII || MPENTIUMM || MCORE2 || MPENTIUM4 || MPSC) config X86_TSC def_bool y diff --git a/arch/x86/Kconfig.debug b/arch/x86/Kconfig.debug index 702eb39901ca..610aaecc19f8 100644 --- a/arch/x86/Kconfig.debug +++ b/arch/x86/Kconfig.debug @@ -54,6 +54,18 @@ config DEBUG_PER_CPU_MAPS Say N if unsure. +config X86_PTDUMP + bool "Export kernel pagetable layout to userspace via debugfs" + depends on DEBUG_KERNEL + select DEBUG_FS + help + Say Y here if you want to show the kernel pagetable layout in a + debugfs file. This information is only useful for kernel developers + who are working in architecture specific areas of the kernel. + It is probably not a good idea to enable this feature in a production + kernel. + If in doubt, say "N" + config DEBUG_RODATA bool "Write protect kernel read-only data structures" default y @@ -64,6 +76,18 @@ config DEBUG_RODATA data. This is recommended so that we can catch kernel bugs sooner. If in doubt, say "Y". +config DIRECT_GBPAGES + bool "Enable gbpages-mapped kernel pagetables" + depends on DEBUG_KERNEL && EXPERIMENTAL && X86_64 + help + Enable gigabyte pages support (if the CPU supports it). This can + improve the kernel's performance a tiny bit by reducing TLB + pressure. + + This is experimental code. + + If in doubt, say "N". + config DEBUG_RODATA_TEST bool "Testcase for the DEBUG_RODATA feature" depends on DEBUG_RODATA @@ -82,8 +106,8 @@ config DEBUG_NX_TEST config 4KSTACKS bool "Use 4Kb for kernel stacks instead of 8Kb" - depends on DEBUG_KERNEL depends on X86_32 + default y help If you say Y here the kernel will use a 4Kb stacksize for the kernel stack attached to each process/thread. This facilitates diff --git a/arch/x86/Makefile b/arch/x86/Makefile index f1e739a43d41..3cff3c894cf3 100644 --- a/arch/x86/Makefile +++ b/arch/x86/Makefile @@ -151,7 +151,6 @@ mflags-y += -Iinclude/asm-x86/mach-default # 64 bit does not support subarch support - clear sub arch variables fcore-$(CONFIG_X86_64) := mcore-$(CONFIG_X86_64) := -mflags-$(CONFIG_X86_64) := KBUILD_CFLAGS += $(mflags-y) KBUILD_AFLAGS += $(mflags-y) @@ -159,9 +158,9 @@ KBUILD_AFLAGS += $(mflags-y) ### # Kernel objects -head-y := arch/x86/kernel/head_$(BITS).o -head-$(CONFIG_X86_64) += arch/x86/kernel/head64.o -head-y += arch/x86/kernel/init_task.o +head-y := arch/x86/kernel/head_$(BITS).o +head-y += arch/x86/kernel/head$(BITS).o +head-y += arch/x86/kernel/init_task.o libs-y += arch/x86/lib/ diff --git a/arch/x86/boot/Makefile b/arch/x86/boot/Makefile index f88458e83ef0..7ee102f9c4f8 100644 --- a/arch/x86/boot/Makefile +++ b/arch/x86/boot/Makefile @@ -30,7 +30,7 @@ subdir- := compressed setup-y += a20.o cmdline.o copy.o cpu.o cpucheck.o edd.o setup-y += header.o main.o mca.o memory.o pm.o pmjump.o -setup-y += printf.o string.o tty.o video.o version.o +setup-y += printf.o string.o tty.o video.o video-mode.o version.o setup-$(CONFIG_X86_APM_BOOT) += apm.o setup-$(CONFIG_X86_VOYAGER) += voyager.o @@ -94,6 +94,20 @@ $(obj)/vmlinux.bin: $(obj)/compressed/vmlinux FORCE SETUP_OBJS = $(addprefix $(obj)/,$(setup-y)) +sed-offsets := -e 's/^00*/0/' \ + -e 's/^\([0-9a-fA-F]*\) . \(input_data\|input_data_end\)$$/\#define \2 0x\1/p' + +quiet_cmd_offsets = OFFSETS $@ + cmd_offsets = $(NM) $< | sed -n $(sed-offsets) > $@ + +$(obj)/offsets.h: $(obj)/compressed/vmlinux FORCE + $(call if_changed,offsets) + +targets += offsets.h + +AFLAGS_header.o += -I$(obj) +$(obj)/header.o: $(obj)/offsets.h + LDFLAGS_setup.elf := -T $(obj)/setup.elf: $(src)/setup.ld $(SETUP_OBJS) FORCE $(call if_changed,ld) diff --git a/arch/x86/boot/boot.h b/arch/x86/boot/boot.h index 7822a4983da2..09578070bfba 100644 --- a/arch/x86/boot/boot.h +++ b/arch/x86/boot/boot.h @@ -286,6 +286,11 @@ int getchar_timeout(void); /* video.c */ void set_video(void); +/* video-mode.c */ +int set_mode(u16 mode); +int mode_defined(u16 mode); +void probe_cards(int unsafe); + /* video-vesa.c */ void vesa_store_edid(void); diff --git a/arch/x86/boot/compressed/Makefile b/arch/x86/boot/compressed/Makefile index d2b9f3bb87c0..92fdd35bd93e 100644 --- a/arch/x86/boot/compressed/Makefile +++ b/arch/x86/boot/compressed/Makefile @@ -22,7 +22,7 @@ $(obj)/vmlinux: $(src)/vmlinux_$(BITS).lds $(obj)/head_$(BITS).o $(obj)/misc.o $ $(call if_changed,ld) @: -OBJCOPYFLAGS_vmlinux.bin := -O binary -R .note -R .comment -S +OBJCOPYFLAGS_vmlinux.bin := -R .comment -S $(obj)/vmlinux.bin: vmlinux FORCE $(call if_changed,objcopy) diff --git a/arch/x86/boot/compressed/misc.c b/arch/x86/boot/compressed/misc.c index 8182e32c1b42..dad4e699f5a3 100644 --- a/arch/x86/boot/compressed/misc.c +++ b/arch/x86/boot/compressed/misc.c @@ -15,6 +15,10 @@ * we just keep it from happening */ #undef CONFIG_PARAVIRT +#ifdef CONFIG_X86_32 +#define _ASM_DESC_H_ 1 +#endif + #ifdef CONFIG_X86_64 #define _LINUX_STRING_H_ 1 #define __LINUX_BITMAP_H 1 @@ -22,6 +26,7 @@ #include <linux/linkage.h> #include <linux/screen_info.h> +#include <linux/elf.h> #include <asm/io.h> #include <asm/page.h> #include <asm/boot.h> @@ -53,8 +58,8 @@ * 1 bit (last block flag) * 2 bits (block type) * - * 1 block occurs every 32K -1 bytes or when there 50% compression has been achieved. - * The smallest block type encoding is always used. + * 1 block occurs every 32K -1 bytes or when there 50% compression + * has been achieved. The smallest block type encoding is always used. * * stored: * 32 bits length in bytes. @@ -90,9 +95,9 @@ * * All of which is enough to compute an amount of extra data that is required * to be safe. To avoid problems at the block level allocating 5 extra bytes - * per 32767 bytes of data is sufficient. To avoind problems internal to a block - * adding an extra 32767 bytes (the worst case uncompressed block size) is - * sufficient, to ensure that in the worst case the decompressed data for + * per 32767 bytes of data is sufficient. To avoind problems internal to a + * block adding an extra 32767 bytes (the worst case uncompressed block size) + * is sufficient, to ensure that in the worst case the decompressed data for * block will stop the byte before the compressed data for a block begins. * To avoid problems with the compressed data's meta information an extra 18 * bytes are needed. Leading to the formula: @@ -111,58 +116,66 @@ * gzip declarations */ -#define OF(args) args -#define STATIC static +#define OF(args) args +#define STATIC static #undef memset #undef memcpy -#define memzero(s, n) memset ((s), 0, (n)) +#define memzero(s, n) memset((s), 0, (n)) + +typedef unsigned char uch; +typedef unsigned short ush; +typedef unsigned long ulg; + +/* + * Window size must be at least 32k, and a power of two. + * We don't actually have a window just a huge output buffer, + * so we report a 2G window size, as that should always be + * larger than our output buffer: + */ +#define WSIZE 0x80000000 + +/* Input buffer: */ +static unsigned char *inbuf; -typedef unsigned char uch; -typedef unsigned short ush; -typedef unsigned long ulg; +/* Sliding window buffer (and final output buffer): */ +static unsigned char *window; -#define WSIZE 0x80000000 /* Window size must be at least 32k, - * and a power of two - * We don't actually have a window just - * a huge output buffer so I report - * a 2G windows size, as that should - * always be larger than our output buffer. - */ +/* Valid bytes in inbuf: */ +static unsigned insize; -static uch *inbuf; /* input buffer */ -static uch *window; /* Sliding window buffer, (and final output buffer) */ +/* Index of next byte to be processed in inbuf: */ +static unsigned inptr; -static unsigned insize; /* valid bytes in inbuf */ -static unsigned inptr; /* index of next byte to be processed in inbuf */ -static unsigned outcnt; /* bytes in output buffer */ +/* Bytes in output buffer: */ +static unsigned outcnt; /* gzip flag byte */ -#define ASCII_FLAG 0x01 /* bit 0 set: file probably ASCII text */ -#define CONTINUATION 0x02 /* bit 1 set: continuation of multi-part gzip file */ -#define EXTRA_FIELD 0x04 /* bit 2 set: extra field present */ -#define ORIG_NAME 0x08 /* bit 3 set: original file name present */ -#define COMMENT 0x10 /* bit 4 set: file comment present */ -#define ENCRYPTED 0x20 /* bit 5 set: file is encrypted */ -#define RESERVED 0xC0 /* bit 6,7: reserved */ - -#define get_byte() (inptr < insize ? inbuf[inptr++] : fill_inbuf()) - +#define ASCII_FLAG 0x01 /* bit 0 set: file probably ASCII text */ +#define CONTINUATION 0x02 /* bit 1 set: continuation of multi-part gz file */ +#define EXTRA_FIELD 0x04 /* bit 2 set: extra field present */ +#define ORIG_NAM 0x08 /* bit 3 set: original file name present */ +#define COMMENT 0x10 /* bit 4 set: file comment present */ +#define ENCRYPTED 0x20 /* bit 5 set: file is encrypted */ +#define RESERVED 0xC0 /* bit 6, 7: reserved */ + +#define get_byte() (inptr < insize ? inbuf[inptr++] : fill_inbuf()) + /* Diagnostic functions */ #ifdef DEBUG -# define Assert(cond,msg) {if(!(cond)) error(msg);} -# define Trace(x) fprintf x -# define Tracev(x) {if (verbose) fprintf x ;} -# define Tracevv(x) {if (verbose>1) fprintf x ;} -# define Tracec(c,x) {if (verbose && (c)) fprintf x ;} -# define Tracecv(c,x) {if (verbose>1 && (c)) fprintf x ;} +# define Assert(cond, msg) do { if (!(cond)) error(msg); } while (0) +# define Trace(x) do { fprintf x; } while (0) +# define Tracev(x) do { if (verbose) fprintf x ; } while (0) +# define Tracevv(x) do { if (verbose > 1) fprintf x ; } while (0) +# define Tracec(c, x) do { if (verbose && (c)) fprintf x ; } while (0) +# define Tracecv(c, x) do { if (verbose > 1 && (c)) fprintf x ; } while (0) #else -# define Assert(cond,msg) +# define Assert(cond, msg) # define Trace(x) # define Tracev(x) # define Tracevv(x) -# define Tracec(c,x) -# define Tracecv(c,x) +# define Tracec(c, x) +# define Tracecv(c, x) #endif static int fill_inbuf(void); @@ -170,7 +183,7 @@ static void flush_window(void); static void error(char *m); static void gzip_mark(void **); static void gzip_release(void **); - + /* * This is set up by the setup-routine at boot-time */ @@ -185,7 +198,7 @@ static unsigned char *real_mode; /* Pointer to real-mode data */ extern unsigned char input_data[]; extern int input_len; -static long bytes_out = 0; +static long bytes_out; static void *malloc(int size); static void free(void *where); @@ -210,7 +223,7 @@ static memptr free_mem_end_ptr; #define HEAP_SIZE 0x4000 #endif -static char *vidmem = (char *)0xb8000; +static char *vidmem; static int vidport; static int lines, cols; @@ -224,8 +237,10 @@ static void *malloc(int size) { void *p; - if (size <0) error("Malloc error"); - if (free_mem_ptr <= 0) error("Memory error"); + if (size < 0) + error("Malloc error"); + if (free_mem_ptr <= 0) + error("Memory error"); free_mem_ptr = (free_mem_ptr + 3) & ~3; /* Align */ @@ -251,19 +266,19 @@ static void gzip_release(void **ptr) { free_mem_ptr = (memptr) *ptr; } - + static void scroll(void) { int i; - memcpy ( vidmem, vidmem + cols * 2, ( lines - 1 ) * cols * 2 ); - for ( i = ( lines - 1 ) * cols * 2; i < lines * cols * 2; i += 2 ) + memcpy(vidmem, vidmem + cols * 2, (lines - 1) * cols * 2); + for (i = (lines - 1) * cols * 2; i < lines * cols * 2; i += 2) vidmem[i] = ' '; } static void putstr(const char *s) { - int x,y,pos; + int x, y, pos; char c; #ifdef CONFIG_X86_32 @@ -274,18 +289,18 @@ static void putstr(const char *s) x = RM_SCREEN_INFO.orig_x; y = RM_SCREEN_INFO.orig_y; - while ( ( c = *s++ ) != '\0' ) { - if ( c == '\n' ) { + while ((c = *s++) != '\0') { + if (c == '\n') { x = 0; - if ( ++y >= lines ) { + if (++y >= lines) { scroll(); y--; } } else { vidmem [(x + cols * y) * 2] = c; - if ( ++x >= cols ) { + if (++x >= cols) { x = 0; - if ( ++y >= lines ) { + if (++y >= lines) { scroll(); y--; } @@ -303,22 +318,22 @@ static void putstr(const char *s) outb(0xff & (pos >> 1), vidport+1); } -static void* memset(void* s, int c, unsigned n) +static void *memset(void *s, int c, unsigned n) { int i; char *ss = s; - for (i=0;i<n;i++) ss[i] = c; + for (i = 0; i < n; i++) ss[i] = c; return s; } -static void* memcpy(void* dest, const void* src, unsigned n) +static void *memcpy(void *dest, const void *src, unsigned n) { int i; const char *s = src; char *d = dest; - for (i=0;i<n;i++) d[i] = s[i]; + for (i = 0; i < n; i++) d[i] = s[i]; return dest; } @@ -341,9 +356,9 @@ static void flush_window(void) /* With my window equal to my output buffer * I only need to compute the crc here. */ - ulg c = crc; /* temporary variable */ + unsigned long c = crc; /* temporary variable */ unsigned n; - uch *in, ch; + unsigned char *in, ch; in = window; for (n = 0; n < outcnt; n++) { @@ -351,7 +366,7 @@ static void flush_window(void) c = crc_32_tab[((int)c ^ ch) & 0xff] ^ (c >> 8); } crc = c; - bytes_out += (ulg)outcnt; + bytes_out += (unsigned long)outcnt; outcnt = 0; } @@ -365,9 +380,59 @@ static void error(char *x) asm("hlt"); } +static void parse_elf(void *output) +{ +#ifdef CONFIG_X86_64 + Elf64_Ehdr ehdr; + Elf64_Phdr *phdrs, *phdr; +#else + Elf32_Ehdr ehdr; + Elf32_Phdr *phdrs, *phdr; +#endif + void *dest; + int i; + + memcpy(&ehdr, output, sizeof(ehdr)); + if (ehdr.e_ident[EI_MAG0] != ELFMAG0 || + ehdr.e_ident[EI_MAG1] != ELFMAG1 || + ehdr.e_ident[EI_MAG2] != ELFMAG2 || + ehdr.e_ident[EI_MAG3] != ELFMAG3) { + error("Kernel is not a valid ELF file"); + return; + } + + putstr("Parsing ELF... "); + + phdrs = malloc(sizeof(*phdrs) * ehdr.e_phnum); + if (!phdrs) + error("Failed to allocate space for phdrs"); + + memcpy(phdrs, output + ehdr.e_phoff, sizeof(*phdrs) * ehdr.e_phnum); + + for (i = 0; i < ehdr.e_phnum; i++) { + phdr = &phdrs[i]; + + switch (phdr->p_type) { + case PT_LOAD: +#ifdef CONFIG_RELOCATABLE + dest = output; + dest += (phdr->p_paddr - LOAD_PHYSICAL_ADDR); +#else + dest = (void *)(phdr->p_paddr); +#endif + memcpy(dest, + output + phdr->p_offset, + phdr->p_filesz); + break; + default: /* Ignore other PT_* */ break; + } + } +} + asmlinkage void decompress_kernel(void *rmode, memptr heap, - uch *input_data, unsigned long input_len, - uch *output) + unsigned char *input_data, + unsigned long input_len, + unsigned char *output) { real_mode = rmode; @@ -390,12 +455,12 @@ asmlinkage void decompress_kernel(void *rmode, memptr heap, inptr = 0; #ifdef CONFIG_X86_64 - if ((ulg)output & (__KERNEL_ALIGN - 1)) + if ((unsigned long)output & (__KERNEL_ALIGN - 1)) error("Destination address not 2M aligned"); - if ((ulg)output >= 0xffffffffffUL) + if ((unsigned long)output >= 0xffffffffffUL) error("Destination address too large"); #else - if ((u32)output & (CONFIG_PHYSICAL_ALIGN -1)) + if ((u32)output & (CONFIG_PHYSICAL_ALIGN - 1)) error("Destination address not CONFIG_PHYSICAL_ALIGN aligned"); if (heap > ((-__PAGE_OFFSET-(512<<20)-1) & 0x7fffffff)) error("Destination address too large"); @@ -408,6 +473,7 @@ asmlinkage void decompress_kernel(void *rmode, memptr heap, makecrc(); putstr("\nDecompressing Linux... "); gunzip(); + parse_elf(output); putstr("done.\nBooting the kernel.\n"); return; } diff --git a/arch/x86/boot/cpucheck.c b/arch/x86/boot/cpucheck.c index 769065bd23d7..2462c88689ed 100644 --- a/arch/x86/boot/cpucheck.c +++ b/arch/x86/boot/cpucheck.c @@ -56,27 +56,27 @@ static const u32 req_flags[NCAPINTS] = REQUIRED_MASK7, }; -#define A32(a,b,c,d) (((d) << 24)+((c) << 16)+((b) << 8)+(a)) +#define A32(a, b, c, d) (((d) << 24)+((c) << 16)+((b) << 8)+(a)) static int is_amd(void) { - return cpu_vendor[0] == A32('A','u','t','h') && - cpu_vendor[1] == A32('e','n','t','i') && - cpu_vendor[2] == A32('c','A','M','D'); + return cpu_vendor[0] == A32('A', 'u', 't', 'h') && + cpu_vendor[1] == A32('e', 'n', 't', 'i') && + cpu_vendor[2] == A32('c', 'A', 'M', 'D'); } static int is_centaur(void) { - return cpu_vendor[0] == A32('C','e','n','t') && - cpu_vendor[1] == A32('a','u','r','H') && - cpu_vendor[2] == A32('a','u','l','s'); + return cpu_vendor[0] == A32('C', 'e', 'n', 't') && + cpu_vendor[1] == A32('a', 'u', 'r', 'H') && + cpu_vendor[2] == A32('a', 'u', 'l', 's'); } static int is_transmeta(void) { - return cpu_vendor[0] == A32('G','e','n','u') && - cpu_vendor[1] == A32('i','n','e','T') && - cpu_vendor[2] == A32('M','x','8','6'); + return cpu_vendor[0] == A32('G', 'e', 'n', 'u') && + cpu_vendor[1] == A32('i', 'n', 'e', 'T') && + cpu_vendor[2] == A32('M', 'x', '8', '6'); } static int has_fpu(void) diff --git a/arch/x86/boot/header.S b/arch/x86/boot/header.S index 64ad9016585a..6d2df8d61c54 100644 --- a/arch/x86/boot/header.S +++ b/arch/x86/boot/header.S @@ -22,6 +22,7 @@ #include <asm/page.h> #include <asm/setup.h> #include "boot.h" +#include "offsets.h" SETUPSECTS = 4 /* default nr of setup-sectors */ BOOTSEG = 0x07C0 /* original address of boot-sector */ @@ -119,7 +120,7 @@ _start: # Part 2 of the header, from the old setup.S .ascii "HdrS" # header signature - .word 0x0207 # header version number (>= 0x0105) + .word 0x0208 # header version number (>= 0x0105) # or else old loadlin-1.5 will fail) .globl realmode_swtch realmode_swtch: .word 0, 0 # default_switch, SETUPSEG @@ -223,6 +224,9 @@ hardware_subarch: .long 0 # subarchitecture, added with 2.07 hardware_subarch_data: .quad 0 +payload_offset: .long input_data +payload_length: .long input_data_end-input_data + # End of setup header ##################################################### .section ".inittext", "ax" diff --git a/arch/x86/boot/pm.c b/arch/x86/boot/pm.c index 1a0f936c160b..a93cb8bded4d 100644 --- a/arch/x86/boot/pm.c +++ b/arch/x86/boot/pm.c @@ -100,7 +100,7 @@ static void reset_coprocessor(void) /* * Set up the GDT */ -#define GDT_ENTRY(flags,base,limit) \ +#define GDT_ENTRY(flags, base, limit) \ (((u64)(base & 0xff000000) << 32) | \ ((u64)flags << 40) | \ ((u64)(limit & 0x00ff0000) << 32) | \ diff --git a/arch/x86/boot/tools/build.c b/arch/x86/boot/tools/build.c index b4248740ff0d..44dc1923c0e3 100644 --- a/arch/x86/boot/tools/build.c +++ b/arch/x86/boot/tools/build.c @@ -50,6 +50,75 @@ typedef unsigned long u32; u8 buf[SETUP_SECT_MAX*512]; int is_big_kernel; +/*----------------------------------------------------------------------*/ + +static const u32 crctab32[] = { + 0x00000000, 0x77073096, 0xee0e612c, 0x990951ba, 0x076dc419, + 0x706af48f, 0xe963a535, 0x9e6495a3, 0x0edb8832, 0x79dcb8a4, + 0xe0d5e91e, 0x97d2d988, 0x09b64c2b, 0x7eb17cbd, 0xe7b82d07, + 0x90bf1d91, 0x1db71064, 0x6ab020f2, 0xf3b97148, 0x84be41de, + 0x1adad47d, 0x6ddde4eb, 0xf4d4b551, 0x83d385c7, 0x136c9856, + 0x646ba8c0, 0xfd62f97a, 0x8a65c9ec, 0x14015c4f, 0x63066cd9, + 0xfa0f3d63, 0x8d080df5, 0x3b6e20c8, 0x4c69105e, 0xd56041e4, + 0xa2677172, 0x3c03e4d1, 0x4b04d447, 0xd20d85fd, 0xa50ab56b, + 0x35b5a8fa, 0x42b2986c, 0xdbbbc9d6, 0xacbcf940, 0x32d86ce3, + 0x45df5c75, 0xdcd60dcf, 0xabd13d59, 0x26d930ac, 0x51de003a, + 0xc8d75180, 0xbfd06116, 0x21b4f4b5, 0x56b3c423, 0xcfba9599, + 0xb8bda50f, 0x2802b89e, 0x5f058808, 0xc60cd9b2, 0xb10be924, + 0x2f6f7c87, 0x58684c11, 0xc1611dab, 0xb6662d3d, 0x76dc4190, + 0x01db7106, 0x98d220bc, 0xefd5102a, 0x71b18589, 0x06b6b51f, + 0x9fbfe4a5, 0xe8b8d433, 0x7807c9a2, 0x0f00f934, 0x9609a88e, + 0xe10e9818, 0x7f6a0dbb, 0x086d3d2d, 0x91646c97, 0xe6635c01, + 0x6b6b51f4, 0x1c6c6162, 0x856530d8, 0xf262004e, 0x6c0695ed, + 0x1b01a57b, 0x8208f4c1, 0xf50fc457, 0x65b0d9c6, 0x12b7e950, + 0x8bbeb8ea, 0xfcb9887c, 0x62dd1ddf, 0x15da2d49, 0x8cd37cf3, + 0xfbd44c65, 0x4db26158, 0x3ab551ce, 0xa3bc0074, 0xd4bb30e2, + 0x4adfa541, 0x3dd895d7, 0xa4d1c46d, 0xd3d6f4fb, 0x4369e96a, + 0x346ed9fc, 0xad678846, 0xda60b8d0, 0x44042d73, 0x33031de5, + 0xaa0a4c5f, 0xdd0d7cc9, 0x5005713c, 0x270241aa, 0xbe0b1010, + 0xc90c2086, 0x5768b525, 0x206f85b3, 0xb966d409, 0xce61e49f, + 0x5edef90e, 0x29d9c998, 0xb0d09822, 0xc7d7a8b4, 0x59b33d17, + 0x2eb40d81, 0xb7bd5c3b, 0xc0ba6cad, 0xedb88320, 0x9abfb3b6, + 0x03b6e20c, 0x74b1d29a, 0xead54739, 0x9dd277af, 0x04db2615, + 0x73dc1683, 0xe3630b12, 0x94643b84, 0x0d6d6a3e, 0x7a6a5aa8, + 0xe40ecf0b, 0x9309ff9d, 0x0a00ae27, 0x7d079eb1, 0xf00f9344, + 0x8708a3d2, 0x1e01f268, 0x6906c2fe, 0xf762575d, 0x806567cb, + 0x196c3671, 0x6e6b06e7, 0xfed41b76, 0x89d32be0, 0x10da7a5a, + 0x67dd4acc, 0xf9b9df6f, 0x8ebeeff9, 0x17b7be43, 0x60b08ed5, + 0xd6d6a3e8, 0xa1d1937e, 0x38d8c2c4, 0x4fdff252, 0xd1bb67f1, + 0xa6bc5767, 0x3fb506dd, 0x48b2364b, 0xd80d2bda, 0xaf0a1b4c, + 0x36034af6, 0x41047a60, 0xdf60efc3, 0xa867df55, 0x316e8eef, + 0x4669be79, 0xcb61b38c, 0xbc66831a, 0x256fd2a0, 0x5268e236, + 0xcc0c7795, 0xbb0b4703, 0x220216b9, 0x5505262f, 0xc5ba3bbe, + 0xb2bd0b28, 0x2bb45a92, 0x5cb36a04, 0xc2d7ffa7, 0xb5d0cf31, + 0x2cd99e8b, 0x5bdeae1d, 0x9b64c2b0, 0xec63f226, 0x756aa39c, + 0x026d930a, 0x9c0906a9, 0xeb0e363f, 0x72076785, 0x05005713, + 0x95bf4a82, 0xe2b87a14, 0x7bb12bae, 0x0cb61b38, 0x92d28e9b, + 0xe5d5be0d, 0x7cdcefb7, 0x0bdbdf21, 0x86d3d2d4, 0xf1d4e242, + 0x68ddb3f8, 0x1fda836e, 0x81be16cd, 0xf6b9265b, 0x6fb077e1, + 0x18b74777, 0x88085ae6, 0xff0f6a70, 0x66063bca, 0x11010b5c, + 0x8f659eff, 0xf862ae69, 0x616bffd3, 0x166ccf45, 0xa00ae278, + 0xd70dd2ee, 0x4e048354, 0x3903b3c2, 0xa7672661, 0xd06016f7, + 0x4969474d, 0x3e6e77db, 0xaed16a4a, 0xd9d65adc, 0x40df0b66, + 0x37d83bf0, 0xa9bcae53, 0xdebb9ec5, 0x47b2cf7f, 0x30b5ffe9, + 0xbdbdf21c, 0xcabac28a, 0x53b39330, 0x24b4a3a6, 0xbad03605, + 0xcdd70693, 0x54de5729, 0x23d967bf, 0xb3667a2e, 0xc4614ab8, + 0x5d681b02, 0x2a6f2b94, 0xb40bbe37, 0xc30c8ea1, 0x5a05df1b, + 0x2d02ef8d +}; + +static u32 partial_crc32_one(u8 c, u32 crc) +{ + return crctab32[(crc ^ c) & 0xff] ^ (crc >> 8); +} + +static u32 partial_crc32(const u8 *s, int len, u32 crc) +{ + while (len--) + crc = partial_crc32_one(*s++, crc); + return crc; +} + static void die(const char * str, ...) { va_list args; @@ -74,6 +143,7 @@ int main(int argc, char ** argv) FILE *file; int fd; void *kernel; + u32 crc = 0xffffffffUL; if (argc > 2 && !strcmp(argv[1], "-b")) { @@ -144,7 +214,8 @@ int main(int argc, char ** argv) kernel = mmap(NULL, sz, PROT_READ, MAP_SHARED, fd, 0); if (kernel == MAP_FAILED) die("Unable to mmap '%s': %m", argv[2]); - sys_size = (sz + 15) / 16; + /* Number of 16-byte paragraphs, including space for a 4-byte CRC */ + sys_size = (sz + 15 + 4) / 16; if (!is_big_kernel && sys_size > DEF_SYSSIZE) die("System is too big. Try using bzImage or modules."); @@ -155,12 +226,27 @@ int main(int argc, char ** argv) buf[0x1f6] = sys_size >> 16; buf[0x1f7] = sys_size >> 24; + crc = partial_crc32(buf, i, crc); if (fwrite(buf, 1, i, stdout) != i) die("Writing setup failed"); /* Copy the kernel code */ + crc = partial_crc32(kernel, sz, crc); if (fwrite(kernel, 1, sz, stdout) != sz) die("Writing kernel failed"); + + /* Add padding leaving 4 bytes for the checksum */ + while (sz++ < (sys_size*16) - 4) { + crc = partial_crc32_one('\0', crc); + if (fwrite("\0", 1, 1, stdout) != 1) + die("Writing padding failed"); + } + + /* Write the CRC */ + fprintf(stderr, "CRC %lx\n", crc); + if (fwrite(&crc, 1, 4, stdout) != 4) + die("Writing CRC failed"); + close(fd); /* Everything is OK */ diff --git a/arch/x86/boot/video-bios.c b/arch/x86/boot/video-bios.c index ff664a117096..39e247e96172 100644 --- a/arch/x86/boot/video-bios.c +++ b/arch/x86/boot/video-bios.c @@ -50,6 +50,7 @@ static int set_bios_mode(u8 mode) if (new_mode == mode) return 0; /* Mode change OK */ +#ifndef _WAKEUP if (new_mode != boot_params.screen_info.orig_video_mode) { /* Mode setting failed, but we didn't end up where we started. That's bad. Try to revert to the original @@ -59,13 +60,18 @@ static int set_bios_mode(u8 mode) : "+a" (ax) : : "ebx", "ecx", "edx", "esi", "edi"); } +#endif return -1; } static int bios_probe(void) { u8 mode; +#ifdef _WAKEUP + u8 saved_mode = 0x03; +#else u8 saved_mode = boot_params.screen_info.orig_video_mode; +#endif u16 crtc; struct mode_info *mi; int nmodes = 0; diff --git a/arch/x86/boot/video-mode.c b/arch/x86/boot/video-mode.c new file mode 100644 index 000000000000..748e8d06290a --- /dev/null +++ b/arch/x86/boot/video-mode.c @@ -0,0 +1,173 @@ +/* -*- linux-c -*- ------------------------------------------------------- * + * + * Copyright (C) 1991, 1992 Linus Torvalds + * Copyright 2007-2008 rPath, Inc. - All Rights Reserved + * + * This file is part of the Linux kernel, and is made available under + * the terms of the GNU General Public License version 2. + * + * ----------------------------------------------------------------------- */ + +/* + * arch/i386/boot/video-mode.c + * + * Set the video mode. This is separated out into a different + * file in order to be shared with the ACPI wakeup code. + */ + +#include "boot.h" +#include "video.h" +#include "vesa.h" + +/* + * Common variables + */ +int adapter; /* 0=CGA/MDA/HGC, 1=EGA, 2=VGA+ */ +u16 video_segment; +int force_x, force_y; /* Don't query the BIOS for cols/rows */ + +int do_restore; /* Screen contents changed during mode flip */ +int graphic_mode; /* Graphic mode with linear frame buffer */ + +/* Probe the video drivers and have them generate their mode lists. */ +void probe_cards(int unsafe) +{ + struct card_info *card; + static u8 probed[2]; + + if (probed[unsafe]) + return; + + probed[unsafe] = 1; + + for (card = video_cards; card < video_cards_end; card++) { + if (card->unsafe == unsafe) { + if (card->probe) + card->nmodes = card->probe(); + else + card->nmodes = 0; + } + } +} + +/* Test if a mode is defined */ +int mode_defined(u16 mode) +{ + struct card_info *card; + struct mode_info *mi; + int i; + + for (card = video_cards; card < video_cards_end; card++) { + mi = card->modes; + for (i = 0; i < card->nmodes; i++, mi++) { + if (mi->mode == mode) + return 1; + } + } + + return 0; +} + +/* Set mode (without recalc) */ +static int raw_set_mode(u16 mode, u16 *real_mode) +{ + int nmode, i; + struct card_info *card; + struct mode_info *mi; + + /* Drop the recalc bit if set */ + mode &= ~VIDEO_RECALC; + + /* Scan for mode based on fixed ID, position, or resolution */ + nmode = 0; + for (card = video_cards; card < video_cards_end; card++) { + mi = card->modes; + for (i = 0; i < card->nmodes; i++, mi++) { + int visible = mi->x || mi->y; + + if ((mode == nmode && visible) || + mode == mi->mode || + mode == (mi->y << 8)+mi->x) { + *real_mode = mi->mode; + return card->set_mode(mi); + } + + if (visible) + nmode++; + } + } + + /* Nothing found? Is it an "exceptional" (unprobed) mode? */ + for (card = video_cards; card < video_cards_end; card++) { + if (mode >= card->xmode_first && + mode < card->xmode_first+card->xmode_n) { + struct mode_info mix; + *real_mode = mix.mode = mode; + mix.x = mix.y = 0; + return card->set_mode(&mix); + } + } + + /* Otherwise, failure... */ + return -1; +} + +/* + * Recalculate the vertical video cutoff (hack!) + */ +static void vga_recalc_vertical(void) +{ + unsigned int font_size, rows; + u16 crtc; + u8 pt, ov; + + set_fs(0); + font_size = rdfs8(0x485); /* BIOS: font size (pixels) */ + rows = force_y ? force_y : rdfs8(0x484)+1; /* Text rows */ + + rows *= font_size; /* Visible scan lines */ + rows--; /* ... minus one */ + + crtc = vga_crtc(); + + pt = in_idx(crtc, 0x11); + pt &= ~0x80; /* Unlock CR0-7 */ + out_idx(pt, crtc, 0x11); + + out_idx((u8)rows, crtc, 0x12); /* Lower height register */ + + ov = in_idx(crtc, 0x07); /* Overflow register */ + ov &= 0xbd; + ov |= (rows >> (8-1)) & 0x02; + ov |= (rows >> (9-6)) & 0x40; + out_idx(ov, crtc, 0x07); +} + +/* Set mode (with recalc if specified) */ +int set_mode(u16 mode) +{ + int rv; + u16 real_mode; + + /* Very special mode numbers... */ + if (mode == VIDEO_CURRENT_MODE) + return 0; /* Nothing to do... */ + else if (mode == NORMAL_VGA) + mode = VIDEO_80x25; + else if (mode == EXTENDED_VGA) + mode = VIDEO_8POINT; + + rv = raw_set_mode(mode, &real_mode); + if (rv) + return rv; + + if (mode & VIDEO_RECALC) + vga_recalc_vertical(); + + /* Save the canonical mode number for the kernel, not + an alias, size specification or menu position */ +#ifndef _WAKEUP + boot_params.hdr.vid_mode = real_mode; +#endif + return 0; +} diff --git a/arch/x86/boot/video-vesa.c b/arch/x86/boot/video-vesa.c index 419b5c273374..5d5a3f6e8b5c 100644 --- a/arch/x86/boot/video-vesa.c +++ b/arch/x86/boot/video-vesa.c @@ -24,7 +24,11 @@ static struct vesa_mode_info vminfo; __videocard video_vesa; +#ifndef _WAKEUP static void vesa_store_mode_params_graphics(void); +#else /* _WAKEUP */ +static inline void vesa_store_mode_params_graphics(void) {} +#endif /* _WAKEUP */ static int vesa_probe(void) { @@ -165,6 +169,8 @@ static int vesa_set_mode(struct mode_info *mode) } +#ifndef _WAKEUP + /* Switch DAC to 8-bit mode */ static void vesa_dac_set_8bits(void) { @@ -288,6 +294,8 @@ void vesa_store_edid(void) #endif /* CONFIG_FIRMWARE_EDID */ } +#endif /* not _WAKEUP */ + __videocard video_vesa = { .card_name = "VESA", diff --git a/arch/x86/boot/video-vga.c b/arch/x86/boot/video-vga.c index 7259387b7d19..330d6589a2ad 100644 --- a/arch/x86/boot/video-vga.c +++ b/arch/x86/boot/video-vga.c @@ -210,6 +210,8 @@ static int vga_set_mode(struct mode_info *mode) */ static int vga_probe(void) { + u16 ega_bx; + static const char *card_name[] = { "CGA/MDA/HGC", "EGA", "VGA" }; @@ -226,12 +228,16 @@ static int vga_probe(void) u8 vga_flag; asm(INT10 - : "=b" (boot_params.screen_info.orig_video_ega_bx) + : "=b" (ega_bx) : "a" (0x1200), "b" (0x10) /* Check EGA/VGA */ : "ecx", "edx", "esi", "edi"); +#ifndef _WAKEUP + boot_params.screen_info.orig_video_ega_bx = ega_bx; +#endif + /* If we have MDA/CGA/HGC then BL will be unchanged at 0x10 */ - if ((u8)boot_params.screen_info.orig_video_ega_bx != 0x10) { + if ((u8)ega_bx != 0x10) { /* EGA/VGA */ asm(INT10 : "=a" (vga_flag) @@ -240,7 +246,9 @@ static int vga_probe(void) if (vga_flag == 0x1a) { adapter = ADAPTER_VGA; +#ifndef _WAKEUP boot_params.screen_info.orig_video_isVGA = 1; +#endif } else { adapter = ADAPTER_EGA; } diff --git a/arch/x86/boot/video.c b/arch/x86/boot/video.c index 696d08f3843c..c1c47ba069ef 100644 --- a/arch/x86/boot/video.c +++ b/arch/x86/boot/video.c @@ -18,21 +18,6 @@ #include "video.h" #include "vesa.h" -/* - * Mode list variables - */ -static struct card_info cards[]; /* List of cards to probe for */ - -/* - * Common variables - */ -int adapter; /* 0=CGA/MDA/HGC, 1=EGA, 2=VGA+ */ -u16 video_segment; -int force_x, force_y; /* Don't query the BIOS for cols/rows */ - -int do_restore = 0; /* Screen contents changed during mode flip */ -int graphic_mode; /* Graphic mode with linear frame buffer */ - static void store_cursor_position(void) { u16 curpos; @@ -107,147 +92,6 @@ static void store_mode_params(void) boot_params.screen_info.orig_video_lines = y; } -/* Probe the video drivers and have them generate their mode lists. */ -static void probe_cards(int unsafe) -{ - struct card_info *card; - static u8 probed[2]; - - if (probed[unsafe]) - return; - - probed[unsafe] = 1; - - for (card = video_cards; card < video_cards_end; card++) { - if (card->unsafe == unsafe) { - if (card->probe) - card->nmodes = card->probe(); - else - card->nmodes = 0; - } - } -} - -/* Test if a mode is defined */ -int mode_defined(u16 mode) -{ - struct card_info *card; - struct mode_info *mi; - int i; - - for (card = video_cards; card < video_cards_end; card++) { - mi = card->modes; - for (i = 0; i < card->nmodes; i++, mi++) { - if (mi->mode == mode) - return 1; - } - } - - return 0; -} - -/* Set mode (without recalc) */ -static int raw_set_mode(u16 mode, u16 *real_mode) -{ - int nmode, i; - struct card_info *card; - struct mode_info *mi; - - /* Drop the recalc bit if set */ - mode &= ~VIDEO_RECALC; - - /* Scan for mode based on fixed ID, position, or resolution */ - nmode = 0; - for (card = video_cards; card < video_cards_end; card++) { - mi = card->modes; - for (i = 0; i < card->nmodes; i++, mi++) { - int visible = mi->x || mi->y; - - if ((mode == nmode && visible) || - mode == mi->mode || - mode == (mi->y << 8)+mi->x) { - *real_mode = mi->mode; - return card->set_mode(mi); - } - - if (visible) - nmode++; - } - } - - /* Nothing found? Is it an "exceptional" (unprobed) mode? */ - for (card = video_cards; card < video_cards_end; card++) { - if (mode >= card->xmode_first && - mode < card->xmode_first+card->xmode_n) { - struct mode_info mix; - *real_mode = mix.mode = mode; - mix.x = mix.y = 0; - return card->set_mode(&mix); - } - } - - /* Otherwise, failure... */ - return -1; -} - -/* - * Recalculate the vertical video cutoff (hack!) - */ -static void vga_recalc_vertical(void) -{ - unsigned int font_size, rows; - u16 crtc; - u8 pt, ov; - - set_fs(0); - font_size = rdfs8(0x485); /* BIOS: font size (pixels) */ - rows = force_y ? force_y : rdfs8(0x484)+1; /* Text rows */ - - rows *= font_size; /* Visible scan lines */ - rows--; /* ... minus one */ - - crtc = vga_crtc(); - - pt = in_idx(crtc, 0x11); - pt &= ~0x80; /* Unlock CR0-7 */ - out_idx(pt, crtc, 0x11); - - out_idx((u8)rows, crtc, 0x12); /* Lower height register */ - - ov = in_idx(crtc, 0x07); /* Overflow register */ - ov &= 0xbd; - ov |= (rows >> (8-1)) & 0x02; - ov |= (rows >> (9-6)) & 0x40; - out_idx(ov, crtc, 0x07); -} - -/* Set mode (with recalc if specified) */ -static int set_mode(u16 mode) -{ - int rv; - u16 real_mode; - - /* Very special mode numbers... */ - if (mode == VIDEO_CURRENT_MODE) - return 0; /* Nothing to do... */ - else if (mode == NORMAL_VGA) - mode = VIDEO_80x25; - else if (mode == EXTENDED_VGA) - mode = VIDEO_8POINT; - - rv = raw_set_mode(mode, &real_mode); - if (rv) - return rv; - - if (mode & VIDEO_RECALC) - vga_recalc_vertical(); - - /* Save the canonical mode number for the kernel, not - an alias, size specification or menu position */ - boot_params.hdr.vid_mode = real_mode; - return 0; -} - static unsigned int get_entry(void) { char entry_buf[4]; @@ -486,6 +330,7 @@ void set_video(void) printf("Undefined video mode number: %x\n", mode); mode = ASK_VGA; } + boot_params.hdr.vid_mode = mode; vesa_store_edid(); store_mode_params(); diff --git a/arch/x86/ia32/ia32_signal.c b/arch/x86/ia32/ia32_signal.c index 5e7771a3ba2f..05e155d3fb6c 100644 --- a/arch/x86/ia32/ia32_signal.c +++ b/arch/x86/ia32/ia32_signal.c @@ -468,7 +468,7 @@ int ia32_setup_frame(int sig, struct k_sigaction *ka, restorer = ka->sa.sa_restorer; } else { /* Return stub is in 32bit vsyscall page */ - if (current->binfmt->hasvdso) + if (current->mm->context.vdso) restorer = VDSO32_SYMBOL(current->mm->context.vdso, sigreturn); else diff --git a/arch/x86/ia32/ia32entry.S b/arch/x86/ia32/ia32entry.S index 8022d3c695c0..ae7158bce4d6 100644 --- a/arch/x86/ia32/ia32entry.S +++ b/arch/x86/ia32/ia32entry.S @@ -162,12 +162,14 @@ sysenter_tracesys: SAVE_REST CLEAR_RREGS movq %r9,R9(%rsp) - movq $-ENOSYS,RAX(%rsp) /* really needed? */ + movq $-ENOSYS,RAX(%rsp)/* ptrace can change this for a bad syscall */ movq %rsp,%rdi /* &pt_regs -> arg1 */ call syscall_trace_enter LOAD_ARGS32 ARGOFFSET /* reload args from stack in case ptrace changed it */ RESTORE_REST xchgl %ebp,%r9d + cmpl $(IA32_NR_syscalls-1),%eax + ja int_ret_from_sys_call /* sysenter_tracesys has set RAX(%rsp) */ jmp sysenter_do_call CFI_ENDPROC ENDPROC(ia32_sysenter_target) @@ -261,13 +263,15 @@ cstar_tracesys: SAVE_REST CLEAR_RREGS movq %r9,R9(%rsp) - movq $-ENOSYS,RAX(%rsp) /* really needed? */ + movq $-ENOSYS,RAX(%rsp) /* ptrace can change this for a bad syscall */ movq %rsp,%rdi /* &pt_regs -> arg1 */ call syscall_trace_enter LOAD_ARGS32 ARGOFFSET /* reload args from stack in case ptrace changed it */ RESTORE_REST xchgl %ebp,%r9d movl RSP-ARGOFFSET(%rsp), %r8d + cmpl $(IA32_NR_syscalls-1),%eax + ja int_ret_from_sys_call /* cstar_tracesys has set RAX(%rsp) */ jmp cstar_do_call END(ia32_cstar_target) @@ -325,7 +329,7 @@ ENTRY(ia32_syscall) jnz ia32_tracesys ia32_do_syscall: cmpl $(IA32_NR_syscalls-1),%eax - ja ia32_badsys + ja int_ret_from_sys_call /* ia32_tracesys has set RAX(%rsp) */ IA32_ARG_FIXUP call *ia32_sys_call_table(,%rax,8) # xxx: rip relative ia32_sysret: @@ -335,7 +339,7 @@ ia32_sysret: ia32_tracesys: SAVE_REST CLEAR_RREGS - movq $-ENOSYS,RAX(%rsp) /* really needed? */ + movq $-ENOSYS,RAX(%rsp) /* ptrace can change this for a bad syscall */ movq %rsp,%rdi /* &pt_regs -> arg1 */ call syscall_trace_enter LOAD_ARGS32 ARGOFFSET /* reload args from stack in case ptrace changed it */ diff --git a/arch/x86/ia32/sys_ia32.c b/arch/x86/ia32/sys_ia32.c index abf71d26fc2a..7cede7a9e0dc 100644 --- a/arch/x86/ia32/sys_ia32.c +++ b/arch/x86/ia32/sys_ia32.c @@ -26,51 +26,27 @@ #include <linux/file.h> #include <linux/signal.h> #include <linux/syscalls.h> -#include <linux/resource.h> #include <linux/times.h> #include <linux/utsname.h> -#include <linux/smp.h> #include <linux/smp_lock.h> -#include <linux/sem.h> -#include <linux/msg.h> #include <linux/mm.h> -#include <linux/shm.h> -#include <linux/slab.h> #include <linux/uio.h> -#include <linux/nfs_fs.h> -#include <linux/quota.h> -#include <linux/module.h> -#include <linux/sunrpc/svc.h> -#include <linux/nfsd/nfsd.h> -#include <linux/nfsd/cache.h> -#include <linux/nfsd/xdr.h> -#include <linux/nfsd/syscall.h> #include <linux/poll.h> #include <linux/personality.h> #include <linux/stat.h> -#include <linux/ipc.h> #include <linux/rwsem.h> -#include <linux/binfmts.h> -#include <linux/init.h> -#include <linux/aio_abi.h> -#include <linux/aio.h> #include <linux/compat.h> #include <linux/vfs.h> #include <linux/ptrace.h> #include <linux/highuid.h> -#include <linux/vmalloc.h> -#include <linux/fsnotify.h> #include <linux/sysctl.h> #include <asm/mman.h> #include <asm/types.h> #include <asm/uaccess.h> #include <asm/semaphore.h> #include <asm/atomic.h> -#include <asm/ldt.h> - -#include <net/scm.h> -#include <net/sock.h> #include <asm/ia32.h> +#include <asm/vgtod.h> #define AA(__x) ((unsigned long)(__x)) @@ -804,11 +780,6 @@ asmlinkage long sys32_execve(char __user *name, compat_uptr_t __user *argv, if (IS_ERR(filename)) return error; error = compat_do_execve(filename, argv, envp, regs); - if (error == 0) { - task_lock(current); - current->ptrace &= ~PT_DTRACE; - task_unlock(current); - } putname(filename); return error; } diff --git a/arch/x86/kernel/Makefile b/arch/x86/kernel/Makefile index 4a4260c7f672..c3920ea8ac56 100644 --- a/arch/x86/kernel/Makefile +++ b/arch/x86/kernel/Makefile @@ -2,8 +2,7 @@ # Makefile for the linux kernel. # -extra-y := head_$(BITS).o init_task.o vmlinux.lds -extra-$(CONFIG_X86_64) += head64.o +extra-y := head_$(BITS).o head$(BITS).o init_task.o vmlinux.lds CPPFLAGS_vmlinux.lds += -U$(UTS_MACHINE) @@ -19,7 +18,7 @@ CFLAGS_tsc_64.o := $(nostackp) obj-y := process_$(BITS).o signal_$(BITS).o entry_$(BITS).o obj-y += traps_$(BITS).o irq_$(BITS).o obj-y += time_$(BITS).o ioport.o ldt.o -obj-y += setup_$(BITS).o i8259_$(BITS).o +obj-y += setup_$(BITS).o i8259_$(BITS).o setup.o obj-$(CONFIG_X86_32) += sys_i386_32.o i386_ksyms_32.o obj-$(CONFIG_X86_64) += sys_x86_64.o x8664_ksyms_64.o obj-$(CONFIG_X86_64) += syscall_64.o vsyscall_64.o setup64.o @@ -29,6 +28,7 @@ obj-y += alternative.o i8253.o obj-$(CONFIG_X86_64) += pci-nommu_64.o bugs_64.o obj-y += tsc_$(BITS).o io_delay.o rtc.o +obj-$(CONFIG_X86_TRAMPOLINE) += trampoline.o obj-y += i387.o obj-y += ptrace.o obj-y += ds.o @@ -47,11 +47,12 @@ obj-$(CONFIG_MICROCODE) += microcode.o obj-$(CONFIG_PCI) += early-quirks.o apm-y := apm_32.o obj-$(CONFIG_APM) += apm.o -obj-$(CONFIG_X86_SMP) += smp_$(BITS).o smpboot_$(BITS).o tsc_sync.o -obj-$(CONFIG_X86_32_SMP) += smpcommon_32.o -obj-$(CONFIG_X86_64_SMP) += smp_64.o smpboot_64.o tsc_sync.o +obj-$(CONFIG_X86_SMP) += smp.o +obj-$(CONFIG_X86_SMP) += smpboot.o tsc_sync.o ipi.o tlb_$(BITS).o +obj-$(CONFIG_X86_32_SMP) += smpcommon.o +obj-$(CONFIG_X86_64_SMP) += tsc_sync.o smpcommon.o obj-$(CONFIG_X86_TRAMPOLINE) += trampoline_$(BITS).o -obj-$(CONFIG_X86_MPPARSE) += mpparse_$(BITS).o +obj-$(CONFIG_X86_MPPARSE) += mpparse.o obj-$(CONFIG_X86_LOCAL_APIC) += apic_$(BITS).o nmi_$(BITS).o obj-$(CONFIG_X86_IO_APIC) += io_apic_$(BITS).o obj-$(CONFIG_X86_REBOOTFIXUPS) += reboot_fixups_32.o @@ -60,7 +61,7 @@ obj-$(CONFIG_KEXEC) += relocate_kernel_$(BITS).o crash.o obj-$(CONFIG_CRASH_DUMP) += crash_dump_$(BITS).o obj-$(CONFIG_X86_NUMAQ) += numaq_32.o obj-$(CONFIG_X86_SUMMIT_NUMA) += summit_32.o -obj-$(CONFIG_X86_VSMP) += vsmp_64.o +obj-y += vsmp_64.o obj-$(CONFIG_KPROBES) += kprobes.o obj-$(CONFIG_MODULES) += module_$(BITS).o obj-$(CONFIG_ACPI_SRAT) += srat_32.o @@ -90,7 +91,7 @@ scx200-y += scx200_32.o ### # 64 bit specific files ifeq ($(CONFIG_X86_64),y) - obj-y += genapic_64.o genapic_flat_64.o + obj-y += genapic_64.o genapic_flat_64.o genx2apic_uv_x.o obj-$(CONFIG_X86_PM_TIMER) += pmtimer_64.o obj-$(CONFIG_AUDIT) += audit_64.o diff --git a/arch/x86/kernel/acpi/Makefile b/arch/x86/kernel/acpi/Makefile index 19d3d6e9d09b..7335959b6aff 100644 --- a/arch/x86/kernel/acpi/Makefile +++ b/arch/x86/kernel/acpi/Makefile @@ -1,7 +1,14 @@ +subdir- := realmode + obj-$(CONFIG_ACPI) += boot.o -obj-$(CONFIG_ACPI_SLEEP) += sleep.o wakeup_$(BITS).o +obj-$(CONFIG_ACPI_SLEEP) += sleep.o wakeup_rm.o wakeup_$(BITS).o ifneq ($(CONFIG_ACPI_PROCESSOR),) obj-y += cstate.o processor.o endif +$(obj)/wakeup_rm.o: $(obj)/realmode/wakeup.bin + +$(obj)/realmode/wakeup.bin: FORCE + $(Q)$(MAKE) $(build)=$(obj)/realmode $@ + diff --git a/arch/x86/kernel/acpi/boot.c b/arch/x86/kernel/acpi/boot.c index 2cdc9de9371d..057ccf1d5ad4 100644 --- a/arch/x86/kernel/acpi/boot.c +++ b/arch/x86/kernel/acpi/boot.c @@ -39,6 +39,11 @@ #include <asm/apic.h> #include <asm/io.h> #include <asm/mpspec.h> +#include <asm/smp.h> + +#ifdef CONFIG_X86_LOCAL_APIC +# include <mach_apic.h> +#endif static int __initdata acpi_force = 0; @@ -52,9 +57,7 @@ EXPORT_SYMBOL(acpi_disabled); #ifdef CONFIG_X86_64 #include <asm/proto.h> - -static inline int acpi_madt_oem_check(char *oem_id, char *oem_table_id) { return 0; } - +#include <asm/genapic.h> #else /* X86 */ @@ -111,7 +114,7 @@ char *__init __acpi_map_table(unsigned long phys_addr, unsigned long size) if (!phys_addr || !size) return NULL; - if (phys_addr+size <= (end_pfn_map << PAGE_SHIFT) + PAGE_SIZE) + if (phys_addr+size <= (max_pfn_mapped << PAGE_SHIFT) + PAGE_SIZE) return __va(phys_addr); return NULL; @@ -237,6 +240,16 @@ static int __init acpi_parse_madt(struct acpi_table_header *table) return 0; } +static void __cpuinit acpi_register_lapic(int id, u8 enabled) +{ + if (!enabled) { + ++disabled_cpus; + return; + } + + generic_processor_info(id, 0); +} + static int __init acpi_parse_lapic(struct acpi_subtable_header * header, const unsigned long end) { @@ -256,8 +269,26 @@ acpi_parse_lapic(struct acpi_subtable_header * header, const unsigned long end) * to not preallocating memory for all NR_CPUS * when we use CPU hotplug. */ - mp_register_lapic(processor->id, /* APIC ID */ - processor->lapic_flags & ACPI_MADT_ENABLED); /* Enabled? */ + acpi_register_lapic(processor->id, /* APIC ID */ + processor->lapic_flags & ACPI_MADT_ENABLED); + + return 0; +} + +static int __init +acpi_parse_sapic(struct acpi_subtable_header *header, const unsigned long end) +{ + struct acpi_madt_local_sapic *processor = NULL; + + processor = (struct acpi_madt_local_sapic *)header; + + if (BAD_MADT_ENTRY(processor, end)) + return -EINVAL; + + acpi_table_print_madt_entry(header); + + acpi_register_lapic((processor->id << 8) | processor->eid,/* APIC ID */ + processor->lapic_flags & ACPI_MADT_ENABLED); return 0; } @@ -300,6 +331,8 @@ acpi_parse_lapic_nmi(struct acpi_subtable_header * header, const unsigned long e #ifdef CONFIG_X86_IO_APIC +struct mp_ioapic_routing mp_ioapic_routing[MAX_IO_APICS]; + static int __init acpi_parse_ioapic(struct acpi_subtable_header * header, const unsigned long end) { @@ -532,7 +565,7 @@ static int __cpuinit _acpi_map_lsapic(acpi_handle handle, int *pcpu) buffer.pointer = NULL; tmp_map = cpu_present_map; - mp_register_lapic(physid, lapic->lapic_flags & ACPI_MADT_ENABLED); + acpi_register_lapic(physid, lapic->lapic_flags & ACPI_MADT_ENABLED); /* * If mp_register_lapic successfully generates a new logical cpu @@ -732,6 +765,16 @@ static int __init acpi_parse_fadt(struct acpi_table_header *table) * Parse LAPIC entries in MADT * returns 0 on success, < 0 on error */ + +static void __init acpi_register_lapic_address(unsigned long address) +{ + mp_lapic_addr = address; + + set_fixmap_nocache(FIX_APIC_BASE, address); + if (boot_cpu_physical_apicid == -1U) + boot_cpu_physical_apicid = GET_APIC_ID(read_apic_id()); +} + static int __init acpi_parse_madt_lapic_entries(void) { int count; @@ -753,10 +796,14 @@ static int __init acpi_parse_madt_lapic_entries(void) return count; } - mp_register_lapic_address(acpi_lapic_addr); + acpi_register_lapic_address(acpi_lapic_addr); + + count = acpi_table_parse_madt(ACPI_MADT_TYPE_LOCAL_SAPIC, + acpi_parse_sapic, MAX_APICS); - count = acpi_table_parse_madt(ACPI_MADT_TYPE_LOCAL_APIC, acpi_parse_lapic, - MAX_APICS); + if (!count) + count = acpi_table_parse_madt(ACPI_MADT_TYPE_LOCAL_APIC, + acpi_parse_lapic, MAX_APICS); if (!count) { printk(KERN_ERR PREFIX "No LAPIC entries present\n"); /* TBD: Cleanup to allow fallback to MPS */ diff --git a/arch/x86/kernel/acpi/realmode/Makefile b/arch/x86/kernel/acpi/realmode/Makefile new file mode 100644 index 000000000000..092900854acc --- /dev/null +++ b/arch/x86/kernel/acpi/realmode/Makefile @@ -0,0 +1,57 @@ +# +# arch/x86/kernel/acpi/realmode/Makefile +# +# This file is subject to the terms and conditions of the GNU General Public +# License. See the file "COPYING" in the main directory of this archive +# for more details. +# + +targets := wakeup.bin wakeup.elf + +wakeup-y += wakeup.o wakemain.o video-mode.o copy.o + +# The link order of the video-*.o modules can matter. In particular, +# video-vga.o *must* be listed first, followed by video-vesa.o. +# Hardware-specific drivers should follow in the order they should be +# probed, and video-bios.o should typically be last. +wakeup-y += video-vga.o +wakeup-y += video-vesa.o +wakeup-y += video-bios.o + +targets += $(wakeup-y) + +bootsrc := $(src)/../../../boot + +# --------------------------------------------------------------------------- + +# How to compile the 16-bit code. Note we always compile for -march=i386, +# that way we can complain to the user if the CPU is insufficient. +# Compile with _SETUP since this is similar to the boot-time setup code. +KBUILD_CFLAGS := $(LINUXINCLUDE) -g -Os -D_SETUP -D_WAKEUP -D__KERNEL__ \ + -I$(srctree)/$(bootsrc) \ + $(cflags-y) \ + -Wall -Wstrict-prototypes \ + -march=i386 -mregparm=3 \ + -include $(srctree)/$(bootsrc)/code16gcc.h \ + -fno-strict-aliasing -fomit-frame-pointer \ + $(call cc-option, -ffreestanding) \ + $(call cc-option, -fno-toplevel-reorder,\ + $(call cc-option, -fno-unit-at-a-time)) \ + $(call cc-option, -fno-stack-protector) \ + $(call cc-option, -mpreferred-stack-boundary=2) +KBUILD_CFLAGS += $(call cc-option, -m32) +KBUILD_AFLAGS := $(KBUILD_CFLAGS) -D__ASSEMBLY__ + +WAKEUP_OBJS = $(addprefix $(obj)/,$(wakeup-y)) + +LDFLAGS_wakeup.elf := -T + +CPPFLAGS_wakeup.lds += -P -C + +$(obj)/wakeup.elf: $(src)/wakeup.lds $(WAKEUP_OBJS) FORCE + $(call if_changed,ld) + +OBJCOPYFLAGS_wakeup.bin := -O binary + +$(obj)/wakeup.bin: $(obj)/wakeup.elf FORCE + $(call if_changed,objcopy) diff --git a/arch/x86/kernel/acpi/realmode/copy.S b/arch/x86/kernel/acpi/realmode/copy.S new file mode 100644 index 000000000000..dc59ebee69d8 --- /dev/null +++ b/arch/x86/kernel/acpi/realmode/copy.S @@ -0,0 +1 @@ +#include "../../../boot/copy.S" diff --git a/arch/x86/kernel/acpi/realmode/video-bios.c b/arch/x86/kernel/acpi/realmode/video-bios.c new file mode 100644 index 000000000000..7deabc144a27 --- /dev/null +++ b/arch/x86/kernel/acpi/realmode/video-bios.c @@ -0,0 +1 @@ +#include "../../../boot/video-bios.c" diff --git a/arch/x86/kernel/acpi/realmode/video-mode.c b/arch/x86/kernel/acpi/realmode/video-mode.c new file mode 100644 index 000000000000..328ad209f113 --- /dev/null +++ b/arch/x86/kernel/acpi/realmode/video-mode.c @@ -0,0 +1 @@ +#include "../../../boot/video-mode.c" diff --git a/arch/x86/kernel/acpi/realmode/video-vesa.c b/arch/x86/kernel/acpi/realmode/video-vesa.c new file mode 100644 index 000000000000..9dbb9672226a --- /dev/null +++ b/arch/x86/kernel/acpi/realmode/video-vesa.c @@ -0,0 +1 @@ +#include "../../../boot/video-vesa.c" diff --git a/arch/x86/kernel/acpi/realmode/video-vga.c b/arch/x86/kernel/acpi/realmode/video-vga.c new file mode 100644 index 000000000000..bcc81255f374 --- /dev/null +++ b/arch/x86/kernel/acpi/realmode/video-vga.c @@ -0,0 +1 @@ +#include "../../../boot/video-vga.c" diff --git a/arch/x86/kernel/acpi/realmode/wakemain.c b/arch/x86/kernel/acpi/realmode/wakemain.c new file mode 100644 index 000000000000..883962d9eef2 --- /dev/null +++ b/arch/x86/kernel/acpi/realmode/wakemain.c @@ -0,0 +1,81 @@ +#include "wakeup.h" +#include "boot.h" + +static void udelay(int loops) +{ + while (loops--) + io_delay(); /* Approximately 1 us */ +} + +static void beep(unsigned int hz) +{ + u8 enable; + + if (!hz) { + enable = 0x00; /* Turn off speaker */ + } else { + u16 div = 1193181/hz; + + outb(0xb6, 0x43); /* Ctr 2, squarewave, load, binary */ + io_delay(); + outb(div, 0x42); /* LSB of counter */ + io_delay(); + outb(div >> 8, 0x42); /* MSB of counter */ + io_delay(); + + enable = 0x03; /* Turn on speaker */ + } + inb(0x61); /* Dummy read of System Control Port B */ + io_delay(); + outb(enable, 0x61); /* Enable timer 2 output to speaker */ + io_delay(); +} + +#define DOT_HZ 880 +#define DASH_HZ 587 +#define US_PER_DOT 125000 + +/* Okay, this is totally silly, but it's kind of fun. */ +static void send_morse(const char *pattern) +{ + char s; + + while ((s = *pattern++)) { + switch (s) { + case '.': + beep(DOT_HZ); + udelay(US_PER_DOT); + beep(0); + udelay(US_PER_DOT); + break; + case '-': + beep(DASH_HZ); + udelay(US_PER_DOT * 3); + beep(0); + udelay(US_PER_DOT); + break; + default: /* Assume it's a space */ + udelay(US_PER_DOT * 3); + break; + } + } +} + +void main(void) +{ + /* Kill machine if structures are wrong */ + if (wakeup_header.real_magic != 0x12345678) + while (1); + + if (wakeup_header.realmode_flags & 4) + send_morse("...-"); + + if (wakeup_header.realmode_flags & 1) + asm volatile("lcallw $0xc000,$3"); + + if (wakeup_header.realmode_flags & 2) { + /* Need to call BIOS */ + probe_cards(0); + set_mode(wakeup_header.video_mode); + } +} diff --git a/arch/x86/kernel/acpi/realmode/wakeup.S b/arch/x86/kernel/acpi/realmode/wakeup.S new file mode 100644 index 000000000000..f9b77fb37e5b --- /dev/null +++ b/arch/x86/kernel/acpi/realmode/wakeup.S @@ -0,0 +1,113 @@ +/* + * ACPI wakeup real mode startup stub + */ +#include <asm/segment.h> +#include <asm/msr-index.h> +#include <asm/page.h> +#include <asm/pgtable.h> + + .code16 + .section ".header", "a" + +/* This should match the structure in wakeup.h */ + .globl wakeup_header +wakeup_header: +video_mode: .short 0 /* Video mode number */ +pmode_return: .byte 0x66, 0xea /* ljmpl */ + .long 0 /* offset goes here */ + .short __KERNEL_CS +pmode_cr0: .long 0 /* Saved %cr0 */ +pmode_cr3: .long 0 /* Saved %cr3 */ +pmode_cr4: .long 0 /* Saved %cr4 */ +pmode_efer: .quad 0 /* Saved EFER */ +pmode_gdt: .quad 0 +realmode_flags: .long 0 +real_magic: .long 0 +trampoline_segment: .word 0 +signature: .long 0x51ee1111 + + .text + .globl _start + .code16 +wakeup_code: +_start: + cli + cld + + /* Set up segments */ + movw %cs, %ax + movw %ax, %ds + movw %ax, %es + movw %ax, %ss + + movl $wakeup_stack_end, %esp + + /* Clear the EFLAGS */ + pushl $0 + popfl + + /* Check header signature... */ + movl signature, %eax + cmpl $0x51ee1111, %eax + jne bogus_real_magic + + /* Check we really have everything... */ + movl end_signature, %eax + cmpl $0x65a22c82, %eax + jne bogus_real_magic + + /* Call the C code */ + calll main + + /* Do any other stuff... */ + +#ifndef CONFIG_64BIT + /* This could also be done in C code... */ + movl pmode_cr3, %eax + movl %eax, %cr3 + + movl pmode_cr4, %ecx + jecxz 1f + movl %ecx, %cr4 +1: + movl pmode_efer, %eax + movl pmode_efer + 4, %edx + movl %eax, %ecx + orl %edx, %ecx + jz 1f + movl $0xc0000080, %ecx + wrmsr +1: + + lgdtl pmode_gdt + + /* This really couldn't... */ + movl pmode_cr0, %eax + movl %eax, %cr0 + jmp pmode_return +#else + pushw $0 + pushw trampoline_segment + pushw $0 + lret +#endif + +bogus_real_magic: +1: + hlt + jmp 1b + + .data + .balign 4 + .globl HEAP, heap_end +HEAP: + .long wakeup_heap +heap_end: + .long wakeup_stack + + .bss +wakeup_heap: + .space 2048 +wakeup_stack: + .space 2048 +wakeup_stack_end: diff --git a/arch/x86/kernel/acpi/realmode/wakeup.h b/arch/x86/kernel/acpi/realmode/wakeup.h new file mode 100644 index 000000000000..ef8166fe8020 --- /dev/null +++ b/arch/x86/kernel/acpi/realmode/wakeup.h @@ -0,0 +1,36 @@ +/* + * Definitions for the wakeup data structure at the head of the + * wakeup code. + */ + +#ifndef ARCH_X86_KERNEL_ACPI_RM_WAKEUP_H +#define ARCH_X86_KERNEL_ACPI_RM_WAKEUP_H + +#ifndef __ASSEMBLY__ +#include <linux/types.h> + +/* This must match data at wakeup.S */ +struct wakeup_header { + u16 video_mode; /* Video mode number */ + u16 _jmp1; /* ljmpl opcode, 32-bit only */ + u32 pmode_entry; /* Protected mode resume point, 32-bit only */ + u16 _jmp2; /* CS value, 32-bit only */ + u32 pmode_cr0; /* Protected mode cr0 */ + u32 pmode_cr3; /* Protected mode cr3 */ + u32 pmode_cr4; /* Protected mode cr4 */ + u32 pmode_efer_low; /* Protected mode EFER */ + u32 pmode_efer_high; + u64 pmode_gdt; + u32 realmode_flags; + u32 real_magic; + u16 trampoline_segment; /* segment with trampoline code, 64-bit only */ + u32 signature; /* To check we have correct structure */ +} __attribute__((__packed__)); + +extern struct wakeup_header wakeup_header; +#endif + +#define HEADER_OFFSET 0x3f00 +#define WAKEUP_SIZE 0x4000 + +#endif /* ARCH_X86_KERNEL_ACPI_RM_WAKEUP_H */ diff --git a/arch/x86/kernel/acpi/realmode/wakeup.lds.S b/arch/x86/kernel/acpi/realmode/wakeup.lds.S new file mode 100644 index 000000000000..22fab6c4be15 --- /dev/null +++ b/arch/x86/kernel/acpi/realmode/wakeup.lds.S @@ -0,0 +1,61 @@ +/* + * wakeup.ld + * + * Linker script for the real-mode wakeup code + */ +#undef i386 +#include "wakeup.h" + +OUTPUT_FORMAT("elf32-i386", "elf32-i386", "elf32-i386") +OUTPUT_ARCH(i386) +ENTRY(_start) + +SECTIONS +{ + . = HEADER_OFFSET; + .header : { + *(.header) + } + + . = 0; + .text : { + *(.text*) + } + + . = ALIGN(16); + .rodata : { + *(.rodata*) + } + + .videocards : { + video_cards = .; + *(.videocards) + video_cards_end = .; + } + + . = ALIGN(16); + .data : { + *(.data*) + } + + .signature : { + end_signature = .; + LONG(0x65a22c82) + } + + . = ALIGN(16); + .bss : { + __bss_start = .; + *(.bss) + __bss_end = .; + } + + . = ALIGN(16); + _end = .; + + /DISCARD/ : { + *(.note*) + } + + . = ASSERT(_end <= WAKEUP_SIZE, "Wakeup too big!"); +} diff --git a/arch/x86/kernel/acpi/sleep.c b/arch/x86/kernel/acpi/sleep.c index 6bc815cd8cb3..afc25ee9964b 100644 --- a/arch/x86/kernel/acpi/sleep.c +++ b/arch/x86/kernel/acpi/sleep.c @@ -10,30 +10,72 @@ #include <linux/dmi.h> #include <linux/cpumask.h> -#include <asm/smp.h> +#include "realmode/wakeup.h" +#include "sleep.h" -/* address in low memory of the wakeup routine. */ -unsigned long acpi_wakeup_address = 0; +unsigned long acpi_wakeup_address; unsigned long acpi_realmode_flags; -extern char wakeup_start, wakeup_end; -extern unsigned long acpi_copy_wakeup_routine(unsigned long); +/* address in low memory of the wakeup routine. */ +static unsigned long acpi_realmode; + +#ifdef CONFIG_64BIT +static char temp_stack[10240]; +#endif /** * acpi_save_state_mem - save kernel state * * Create an identity mapped page table and copy the wakeup routine to * low memory. + * + * Note that this is too late to change acpi_wakeup_address. */ int acpi_save_state_mem(void) { - if (!acpi_wakeup_address) { - printk(KERN_ERR "Could not allocate memory during boot, S3 disabled\n"); + struct wakeup_header *header; + + if (!acpi_realmode) { + printk(KERN_ERR "Could not allocate memory during boot, " + "S3 disabled\n"); return -ENOMEM; } - memcpy((void *)acpi_wakeup_address, &wakeup_start, - &wakeup_end - &wakeup_start); - acpi_copy_wakeup_routine(acpi_wakeup_address); + memcpy((void *)acpi_realmode, &wakeup_code_start, WAKEUP_SIZE); + + header = (struct wakeup_header *)(acpi_realmode + HEADER_OFFSET); + if (header->signature != 0x51ee1111) { + printk(KERN_ERR "wakeup header does not match\n"); + return -EINVAL; + } + + header->video_mode = saved_video_mode; + +#ifndef CONFIG_64BIT + store_gdt((struct desc_ptr *)&header->pmode_gdt); + + header->pmode_efer_low = nx_enabled; + if (header->pmode_efer_low & 1) { + /* This is strange, why not save efer, always? */ + rdmsr(MSR_EFER, header->pmode_efer_low, + header->pmode_efer_high); + } +#endif /* !CONFIG_64BIT */ + + header->pmode_cr0 = read_cr0(); + header->pmode_cr4 = read_cr4(); + header->realmode_flags = acpi_realmode_flags; + header->real_magic = 0x12345678; + +#ifndef CONFIG_64BIT + header->pmode_entry = (u32)&wakeup_pmode_return; + header->pmode_cr3 = (u32)(swsusp_pg_dir - __PAGE_OFFSET); + saved_magic = 0x12345678; +#else /* CONFIG_64BIT */ + header->trampoline_segment = setup_trampoline() >> 4; + init_rsp = (unsigned long)temp_stack + 4096; + initial_code = (unsigned long)wakeup_long64; + saved_magic = 0x123456789abcdef0; +#endif /* CONFIG_64BIT */ return 0; } @@ -56,15 +98,20 @@ void acpi_restore_state_mem(void) */ void __init acpi_reserve_bootmem(void) { - if ((&wakeup_end - &wakeup_start) > PAGE_SIZE*2) { + if ((&wakeup_code_end - &wakeup_code_start) > WAKEUP_SIZE) { printk(KERN_ERR "ACPI: Wakeup code way too big, S3 disabled.\n"); return; } - acpi_wakeup_address = (unsigned long)alloc_bootmem_low(PAGE_SIZE*2); - if (!acpi_wakeup_address) + acpi_realmode = (unsigned long)alloc_bootmem_low(WAKEUP_SIZE); + + if (!acpi_realmode) { printk(KERN_ERR "ACPI: Cannot allocate lowmem, S3 disabled.\n"); + return; + } + + acpi_wakeup_address = acpi_realmode; } diff --git a/arch/x86/kernel/acpi/sleep.h b/arch/x86/kernel/acpi/sleep.h new file mode 100644 index 000000000000..adbcbaa6f1df --- /dev/null +++ b/arch/x86/kernel/acpi/sleep.h @@ -0,0 +1,16 @@ +/* + * Variables and functions used by the code in sleep.c + */ + +#include <asm/trampoline.h> + +extern char wakeup_code_start, wakeup_code_end; + +extern unsigned long saved_video_mode; +extern long saved_magic; + +extern int wakeup_pmode_return; +extern char swsusp_pg_dir[PAGE_SIZE]; + +extern unsigned long acpi_copy_wakeup_routine(unsigned long); +extern void wakeup_long64(void); diff --git a/arch/x86/kernel/acpi/sleep_32.c b/arch/x86/kernel/acpi/sleep_32.c deleted file mode 100644 index 63fe5525e026..000000000000 --- a/arch/x86/kernel/acpi/sleep_32.c +++ /dev/null @@ -1,40 +0,0 @@ -/* - * sleep.c - x86-specific ACPI sleep support. - * - * Copyright (C) 2001-2003 Patrick Mochel - * Copyright (C) 2001-2003 Pavel Machek <pavel@suse.cz> - */ - -#include <linux/acpi.h> -#include <linux/bootmem.h> -#include <linux/dmi.h> -#include <linux/cpumask.h> - -#include <asm/smp.h> - -/* Ouch, we want to delete this. We already have better version in userspace, in - s2ram from suspend.sf.net project */ -static __init int reset_videomode_after_s3(const struct dmi_system_id *d) -{ - acpi_realmode_flags |= 2; - return 0; -} - -static __initdata struct dmi_system_id acpisleep_dmi_table[] = { - { /* Reset video mode after returning from ACPI S3 sleep */ - .callback = reset_videomode_after_s3, - .ident = "Toshiba Satellite 4030cdt", - .matches = { - DMI_MATCH(DMI_PRODUCT_NAME, "S4030CDT/4.3"), - }, - }, - {} -}; - -static int __init acpisleep_dmi_init(void) -{ - dmi_check_system(acpisleep_dmi_table); - return 0; -} - -core_initcall(acpisleep_dmi_init); diff --git a/arch/x86/kernel/acpi/wakeup_32.S b/arch/x86/kernel/acpi/wakeup_32.S index f53e3277f8e5..a12e6a9fb659 100644 --- a/arch/x86/kernel/acpi/wakeup_32.S +++ b/arch/x86/kernel/acpi/wakeup_32.S @@ -3,178 +3,12 @@ #include <asm/segment.h> #include <asm/page.h> -# -# wakeup_code runs in real mode, and at unknown address (determined at run-time). -# Therefore it must only use relative jumps/calls. -# -# Do we need to deal with A20? It is okay: ACPI specs says A20 must be enabled -# -# If physical address of wakeup_code is 0x12345, BIOS should call us with -# cs = 0x1234, eip = 0x05 -# - -#define BEEP \ - inb $97, %al; \ - outb %al, $0x80; \ - movb $3, %al; \ - outb %al, $97; \ - outb %al, $0x80; \ - movb $-74, %al; \ - outb %al, $67; \ - outb %al, $0x80; \ - movb $-119, %al; \ - outb %al, $66; \ - outb %al, $0x80; \ - movb $15, %al; \ - outb %al, $66; - -ALIGN - .align 4096 -ENTRY(wakeup_start) -wakeup_code: - wakeup_code_start = . - .code16 - - cli - cld - - # setup data segment - movw %cs, %ax - movw %ax, %ds # Make ds:0 point to wakeup_start - movw %ax, %ss - - testl $4, realmode_flags - wakeup_code - jz 1f - BEEP -1: - mov $(wakeup_stack - wakeup_code), %sp # Private stack is needed for ASUS board - - pushl $0 # Kill any dangerous flags - popfl - - movl real_magic - wakeup_code, %eax - cmpl $0x12345678, %eax - jne bogus_real_magic - - testl $1, realmode_flags - wakeup_code - jz 1f - lcall $0xc000,$3 - movw %cs, %ax - movw %ax, %ds # Bios might have played with that - movw %ax, %ss -1: - - testl $2, realmode_flags - wakeup_code - jz 1f - mov video_mode - wakeup_code, %ax - call mode_set -1: - - # set up page table - movl $swsusp_pg_dir-__PAGE_OFFSET, %eax - movl %eax, %cr3 - - testl $1, real_efer_save_restore - wakeup_code - jz 4f - # restore efer setting - movl real_save_efer_edx - wakeup_code, %edx - movl real_save_efer_eax - wakeup_code, %eax - mov $0xc0000080, %ecx - wrmsr -4: - # make sure %cr4 is set correctly (features, etc) - movl real_save_cr4 - wakeup_code, %eax - movl %eax, %cr4 - - # need a gdt -- use lgdtl to force 32-bit operands, in case - # the GDT is located past 16 megabytes. - lgdtl real_save_gdt - wakeup_code - - movl real_save_cr0 - wakeup_code, %eax - movl %eax, %cr0 - jmp 1f -1: - movl real_magic - wakeup_code, %eax - cmpl $0x12345678, %eax - jne bogus_real_magic - - testl $8, realmode_flags - wakeup_code - jz 1f - BEEP -1: - ljmpl $__KERNEL_CS, $wakeup_pmode_return - -real_save_gdt: .word 0 - .long 0 -real_save_cr0: .long 0 -real_save_cr3: .long 0 -real_save_cr4: .long 0 -real_magic: .long 0 -video_mode: .long 0 -realmode_flags: .long 0 -real_efer_save_restore: .long 0 -real_save_efer_edx: .long 0 -real_save_efer_eax: .long 0 - -bogus_real_magic: - jmp bogus_real_magic - -/* This code uses an extended set of video mode numbers. These include: - * Aliases for standard modes - * NORMAL_VGA (-1) - * EXTENDED_VGA (-2) - * ASK_VGA (-3) - * Video modes numbered by menu position -- NOT RECOMMENDED because of lack - * of compatibility when extending the table. These are between 0x00 and 0xff. - */ -#define VIDEO_FIRST_MENU 0x0000 - -/* Standard BIOS video modes (BIOS number + 0x0100) */ -#define VIDEO_FIRST_BIOS 0x0100 - -/* VESA BIOS video modes (VESA number + 0x0200) */ -#define VIDEO_FIRST_VESA 0x0200 - -/* Video7 special modes (BIOS number + 0x0900) */ -#define VIDEO_FIRST_V7 0x0900 - -# Setting of user mode (AX=mode ID) => CF=success - -# For now, we only handle VESA modes (0x0200..0x03ff). To handle other -# modes, we should probably compile in the video code from the boot -# directory. -mode_set: - movw %ax, %bx - subb $VIDEO_FIRST_VESA>>8, %bh - cmpb $2, %bh - jb check_vesa - -setbad: - clc - ret - -check_vesa: - orw $0x4000, %bx # Use linear frame buffer - movw $0x4f02, %ax # VESA BIOS mode set call - int $0x10 - cmpw $0x004f, %ax # AL=4f if implemented - jnz setbad # AH=0 if OK - - stc - ret +# Copyright 2003, 2008 Pavel Machek <pavel@suse.cz>, distribute under GPLv2 .code32 ALIGN -.org 0x800 -wakeup_stack_begin: # Stack grows down - -.org 0xff0 # Just below end of page -wakeup_stack: -ENTRY(wakeup_end) - -.org 0x1000 - +ENTRY(wakeup_pmode_return) wakeup_pmode_return: movw $__KERNEL_DS, %ax movw %ax, %ss @@ -187,7 +21,7 @@ wakeup_pmode_return: lgdt saved_gdt lidt saved_idt lldt saved_ldt - ljmp $(__KERNEL_CS),$1f + ljmp $(__KERNEL_CS), $1f 1: movl %cr3, %eax movl %eax, %cr3 @@ -201,82 +35,41 @@ wakeup_pmode_return: jne bogus_magic # jump to place where we left off - movl saved_eip,%eax + movl saved_eip, %eax jmp *%eax bogus_magic: jmp bogus_magic -## -# acpi_copy_wakeup_routine -# -# Copy the above routine to low memory. -# -# Parameters: -# %eax: place to copy wakeup routine to -# -# Returned address is location of code in low memory (past data and stack) -# -ENTRY(acpi_copy_wakeup_routine) - pushl %ebx +save_registers: sgdt saved_gdt sidt saved_idt sldt saved_ldt str saved_tss - movl nx_enabled, %edx - movl %edx, real_efer_save_restore - wakeup_start (%eax) - testl $1, real_efer_save_restore - wakeup_start (%eax) - jz 2f - # save efer setting - pushl %eax - movl %eax, %ebx - mov $0xc0000080, %ecx - rdmsr - movl %edx, real_save_efer_edx - wakeup_start (%ebx) - movl %eax, real_save_efer_eax - wakeup_start (%ebx) - popl %eax -2: - - movl %cr3, %edx - movl %edx, real_save_cr3 - wakeup_start (%eax) - movl %cr4, %edx - movl %edx, real_save_cr4 - wakeup_start (%eax) - movl %cr0, %edx - movl %edx, real_save_cr0 - wakeup_start (%eax) - sgdt real_save_gdt - wakeup_start (%eax) - - movl saved_videomode, %edx - movl %edx, video_mode - wakeup_start (%eax) - movl acpi_realmode_flags, %edx - movl %edx, realmode_flags - wakeup_start (%eax) - movl $0x12345678, real_magic - wakeup_start (%eax) - movl $0x12345678, saved_magic - popl %ebx - ret - -save_registers: leal 4(%esp), %eax movl %eax, saved_context_esp - movl %ebx, saved_context_ebx - movl %ebp, saved_context_ebp - movl %esi, saved_context_esi - movl %edi, saved_context_edi - pushfl ; popl saved_context_eflags - - movl $ret_point, saved_eip + movl %ebx, saved_context_ebx + movl %ebp, saved_context_ebp + movl %esi, saved_context_esi + movl %edi, saved_context_edi + pushfl + popl saved_context_eflags + + movl $ret_point, saved_eip ret restore_registers: - movl saved_context_ebp, %ebp - movl saved_context_ebx, %ebx - movl saved_context_esi, %esi - movl saved_context_edi, %edi - pushl saved_context_eflags ; popfl - ret + movl saved_context_ebp, %ebp + movl saved_context_ebx, %ebx + movl saved_context_esi, %esi + movl saved_context_edi, %edi + pushl saved_context_eflags + popfl + ret ENTRY(do_suspend_lowlevel) call save_processor_state diff --git a/arch/x86/kernel/acpi/wakeup_64.S b/arch/x86/kernel/acpi/wakeup_64.S index 2e1b9e0d0767..bcc293423a70 100644 --- a/arch/x86/kernel/acpi/wakeup_64.S +++ b/arch/x86/kernel/acpi/wakeup_64.S @@ -7,191 +7,18 @@ #include <asm/asm-offsets.h> # Copyright 2003 Pavel Machek <pavel@suse.cz>, distribute under GPLv2 -# -# wakeup_code runs in real mode, and at unknown address (determined at run-time). -# Therefore it must only use relative jumps/calls. -# -# Do we need to deal with A20? It is okay: ACPI specs says A20 must be enabled -# -# If physical address of wakeup_code is 0x12345, BIOS should call us with -# cs = 0x1234, eip = 0x05 -# - -#define BEEP \ - inb $97, %al; \ - outb %al, $0x80; \ - movb $3, %al; \ - outb %al, $97; \ - outb %al, $0x80; \ - movb $-74, %al; \ - outb %al, $67; \ - outb %al, $0x80; \ - movb $-119, %al; \ - outb %al, $66; \ - outb %al, $0x80; \ - movb $15, %al; \ - outb %al, $66; - - -ALIGN - .align 16 -ENTRY(wakeup_start) -wakeup_code: - wakeup_code_start = . - .code16 - -# Running in *copy* of this code, somewhere in low 1MB. - - cli - cld - # setup data segment - movw %cs, %ax - movw %ax, %ds # Make ds:0 point to wakeup_start - movw %ax, %ss - - # Data segment must be set up before we can see whether to beep. - testl $4, realmode_flags - wakeup_code - jz 1f - BEEP -1: - - # Private stack is needed for ASUS board - mov $(wakeup_stack - wakeup_code), %sp - - pushl $0 # Kill any dangerous flags - popfl - - movl real_magic - wakeup_code, %eax - cmpl $0x12345678, %eax - jne bogus_real_magic - - testl $1, realmode_flags - wakeup_code - jz 1f - lcall $0xc000,$3 - movw %cs, %ax - movw %ax, %ds # Bios might have played with that - movw %ax, %ss -1: - - testl $2, realmode_flags - wakeup_code - jz 1f - mov video_mode - wakeup_code, %ax - call mode_set -1: - - mov %ds, %ax # Find 32bit wakeup_code addr - movzx %ax, %esi # (Convert %ds:gdt to a liner ptr) - shll $4, %esi - # Fix up the vectors - addl %esi, wakeup_32_vector - wakeup_code - addl %esi, wakeup_long64_vector - wakeup_code - addl %esi, gdt_48a + 2 - wakeup_code # Fixup the gdt pointer - - lidtl %ds:idt_48a - wakeup_code - lgdtl %ds:gdt_48a - wakeup_code # load gdt with whatever is - # appropriate - - movl $1, %eax # protected mode (PE) bit - lmsw %ax # This is it! - jmp 1f -1: - - ljmpl *(wakeup_32_vector - wakeup_code) - - .balign 4 -wakeup_32_vector: - .long wakeup_32 - wakeup_code - .word __KERNEL32_CS, 0 - - .code32 -wakeup_32: -# Running in this code, but at low address; paging is not yet turned on. - - movl $__KERNEL_DS, %eax - movl %eax, %ds - - /* - * Prepare for entering 64bits mode - */ - - /* Enable PAE */ - xorl %eax, %eax - btsl $5, %eax - movl %eax, %cr4 - - /* Setup early boot stage 4 level pagetables */ - leal (wakeup_level4_pgt - wakeup_code)(%esi), %eax - movl %eax, %cr3 - - /* Check if nx is implemented */ - movl $0x80000001, %eax - cpuid - movl %edx,%edi - - /* Enable Long Mode */ - xorl %eax, %eax - btsl $_EFER_LME, %eax - - /* No Execute supported? */ - btl $20,%edi - jnc 1f - btsl $_EFER_NX, %eax - - /* Make changes effective */ -1: movl $MSR_EFER, %ecx - xorl %edx, %edx - wrmsr - - xorl %eax, %eax - btsl $31, %eax /* Enable paging and in turn activate Long Mode */ - btsl $0, %eax /* Enable protected mode */ - - /* Make changes effective */ - movl %eax, %cr0 - - /* At this point: - CR4.PAE must be 1 - CS.L must be 0 - CR3 must point to PML4 - Next instruction must be a branch - This must be on identity-mapped page - */ - /* - * At this point we're in long mode but in 32bit compatibility mode - * with EFER.LME = 1, CS.L = 0, CS.D = 1 (and in turn - * EFER.LMA = 1). Now we want to jump in 64bit mode, to do that we load - * the new gdt/idt that has __KERNEL_CS with CS.L = 1. - */ - - /* Finally jump in 64bit mode */ - ljmp *(wakeup_long64_vector - wakeup_code)(%esi) - - .balign 4 -wakeup_long64_vector: - .long wakeup_long64 - wakeup_code - .word __KERNEL_CS, 0 .code64 - - /* Hooray, we are in Long 64-bit mode (but still running in - * low memory) - */ -wakeup_long64: /* - * We must switch to a new descriptor in kernel space for the GDT - * because soon the kernel won't have access anymore to the userspace - * addresses where we're currently running on. We have to do that here - * because in 32bit we couldn't load a 64bit linear address. + * Hooray, we are in Long 64-bit mode (but still running in low memory) */ - lgdt cpu_gdt_descr - - movq saved_magic, %rax - movq $0x123456789abcdef0, %rdx - cmpq %rdx, %rax - jne bogus_64_magic +ENTRY(wakeup_long64) +wakeup_long64: + movq saved_magic, %rax + movq $0x123456789abcdef0, %rdx + cmpq %rdx, %rax + jne bogus_64_magic - nop - nop movw $__KERNEL_DS, %ax movw %ax, %ss movw %ax, %ds @@ -208,130 +35,8 @@ wakeup_long64: movq saved_rip, %rax jmp *%rax -.code32 - - .align 64 -gdta: - /* Its good to keep gdt in sync with one in trampoline.S */ - .word 0, 0, 0, 0 # dummy - /* ??? Why I need the accessed bit set in order for this to work? */ - .quad 0x00cf9b000000ffff # __KERNEL32_CS - .quad 0x00af9b000000ffff # __KERNEL_CS - .quad 0x00cf93000000ffff # __KERNEL_DS - -idt_48a: - .word 0 # idt limit = 0 - .word 0, 0 # idt base = 0L - -gdt_48a: - .word 0x800 # gdt limit=2048, - # 256 GDT entries - .long gdta - wakeup_code # gdt base (relocated in later) - -real_magic: .quad 0 -video_mode: .quad 0 -realmode_flags: .quad 0 - -.code16 -bogus_real_magic: - jmp bogus_real_magic - -.code64 bogus_64_magic: - jmp bogus_64_magic - -/* This code uses an extended set of video mode numbers. These include: - * Aliases for standard modes - * NORMAL_VGA (-1) - * EXTENDED_VGA (-2) - * ASK_VGA (-3) - * Video modes numbered by menu position -- NOT RECOMMENDED because of lack - * of compatibility when extending the table. These are between 0x00 and 0xff. - */ -#define VIDEO_FIRST_MENU 0x0000 - -/* Standard BIOS video modes (BIOS number + 0x0100) */ -#define VIDEO_FIRST_BIOS 0x0100 - -/* VESA BIOS video modes (VESA number + 0x0200) */ -#define VIDEO_FIRST_VESA 0x0200 - -/* Video7 special modes (BIOS number + 0x0900) */ -#define VIDEO_FIRST_V7 0x0900 - -# Setting of user mode (AX=mode ID) => CF=success - -# For now, we only handle VESA modes (0x0200..0x03ff). To handle other -# modes, we should probably compile in the video code from the boot -# directory. -.code16 -mode_set: - movw %ax, %bx - subb $VIDEO_FIRST_VESA>>8, %bh - cmpb $2, %bh - jb check_vesa - -setbad: - clc - ret - -check_vesa: - orw $0x4000, %bx # Use linear frame buffer - movw $0x4f02, %ax # VESA BIOS mode set call - int $0x10 - cmpw $0x004f, %ax # AL=4f if implemented - jnz setbad # AH=0 if OK - - stc - ret - -wakeup_stack_begin: # Stack grows down - -.org 0xff0 -wakeup_stack: # Just below end of page - -.org 0x1000 -ENTRY(wakeup_level4_pgt) - .quad level3_ident_pgt - __START_KERNEL_map + _KERNPG_TABLE - .fill 510,8,0 - /* (2^48-(2*1024*1024*1024))/(2^39) = 511 */ - .quad level3_kernel_pgt - __START_KERNEL_map + _KERNPG_TABLE - -ENTRY(wakeup_end) - -## -# acpi_copy_wakeup_routine -# -# Copy the above routine to low memory. -# -# Parameters: -# %rdi: place to copy wakeup routine to -# -# Returned address is location of code in low memory (past data and stack) -# - .code64 -ENTRY(acpi_copy_wakeup_routine) - pushq %rax - pushq %rdx - - movl saved_video_mode, %edx - movl %edx, video_mode - wakeup_start (,%rdi) - movl acpi_realmode_flags, %edx - movl %edx, realmode_flags - wakeup_start (,%rdi) - movq $0x12345678, real_magic - wakeup_start (,%rdi) - movq $0x123456789abcdef0, %rdx - movq %rdx, saved_magic - - movq saved_magic, %rax - movq $0x123456789abcdef0, %rdx - cmpq %rdx, %rax - jne bogus_64_magic - - # restore the regs we used - popq %rdx - popq %rax -ENTRY(do_suspend_lowlevel_s4bios) - ret + jmp bogus_64_magic .align 2 .p2align 4,,15 @@ -414,7 +119,7 @@ do_suspend_lowlevel: jmp restore_processor_state .LFE5: .Lfe5: - .size do_suspend_lowlevel,.Lfe5-do_suspend_lowlevel + .size do_suspend_lowlevel, .Lfe5-do_suspend_lowlevel .data ALIGN diff --git a/arch/x86/kernel/acpi/wakeup_rm.S b/arch/x86/kernel/acpi/wakeup_rm.S new file mode 100644 index 000000000000..6ff3b5730575 --- /dev/null +++ b/arch/x86/kernel/acpi/wakeup_rm.S @@ -0,0 +1,10 @@ +/* + * Wrapper script for the realmode binary as a transport object + * before copying to low memory. + */ + .section ".rodata","a" + .globl wakeup_code_start, wakeup_code_end +wakeup_code_start: + .incbin "arch/x86/kernel/acpi/realmode/wakeup.bin" +wakeup_code_end: + .size wakeup_code_start, .-wakeup_code_start diff --git a/arch/x86/kernel/alternative.c b/arch/x86/kernel/alternative.c index 5fed98ca0e1f..df4099dc1c68 100644 --- a/arch/x86/kernel/alternative.c +++ b/arch/x86/kernel/alternative.c @@ -11,6 +11,8 @@ #include <asm/mce.h> #include <asm/nmi.h> #include <asm/vsyscall.h> +#include <asm/cacheflush.h> +#include <asm/io.h> #define MAX_PATCH_LEN (255-1) @@ -177,7 +179,7 @@ static const unsigned char*const * find_nop_table(void) #endif /* CONFIG_X86_64 */ /* Use this to add nops to a buffer, then text_poke the whole buffer. */ -static void add_nops(void *insns, unsigned int len) +void add_nops(void *insns, unsigned int len) { const unsigned char *const *noptable = find_nop_table(); @@ -190,6 +192,7 @@ static void add_nops(void *insns, unsigned int len) len -= noplen; } } +EXPORT_SYMBOL_GPL(add_nops); extern struct alt_instr __alt_instructions[], __alt_instructions_end[]; extern u8 *__smp_locks[], *__smp_locks_end[]; @@ -205,7 +208,7 @@ void apply_alternatives(struct alt_instr *start, struct alt_instr *end) struct alt_instr *a; char insnbuf[MAX_PATCH_LEN]; - DPRINTK("%s: alt table %p -> %p\n", __FUNCTION__, start, end); + DPRINTK("%s: alt table %p -> %p\n", __func__, start, end); for (a = start; a < end; a++) { u8 *instr = a->instr; BUG_ON(a->replacementlen > a->instrlen); @@ -217,13 +220,13 @@ void apply_alternatives(struct alt_instr *start, struct alt_instr *end) if (instr >= (u8 *)VSYSCALL_START && instr < (u8*)VSYSCALL_END) { instr = __va(instr - (u8*)VSYSCALL_START + (u8*)__pa_symbol(&__vsyscall_0)); DPRINTK("%s: vsyscall fixup: %p => %p\n", - __FUNCTION__, a->instr, instr); + __func__, a->instr, instr); } #endif memcpy(insnbuf, a->replacement, a->replacementlen); add_nops(insnbuf + a->replacementlen, a->instrlen - a->replacementlen); - text_poke(instr, insnbuf, a->instrlen); + text_poke_early(instr, insnbuf, a->instrlen); } } @@ -284,7 +287,6 @@ void alternatives_smp_module_add(struct module *mod, char *name, void *text, void *text_end) { struct smp_alt_module *smp; - unsigned long flags; if (noreplace_smp) return; @@ -307,42 +309,40 @@ void alternatives_smp_module_add(struct module *mod, char *name, smp->text = text; smp->text_end = text_end; DPRINTK("%s: locks %p -> %p, text %p -> %p, name %s\n", - __FUNCTION__, smp->locks, smp->locks_end, + __func__, smp->locks, smp->locks_end, smp->text, smp->text_end, smp->name); - spin_lock_irqsave(&smp_alt, flags); + spin_lock(&smp_alt); list_add_tail(&smp->next, &smp_alt_modules); if (boot_cpu_has(X86_FEATURE_UP)) alternatives_smp_unlock(smp->locks, smp->locks_end, smp->text, smp->text_end); - spin_unlock_irqrestore(&smp_alt, flags); + spin_unlock(&smp_alt); } void alternatives_smp_module_del(struct module *mod) { struct smp_alt_module *item; - unsigned long flags; if (smp_alt_once || noreplace_smp) return; - spin_lock_irqsave(&smp_alt, flags); + spin_lock(&smp_alt); list_for_each_entry(item, &smp_alt_modules, next) { if (mod != item->mod) continue; list_del(&item->next); - spin_unlock_irqrestore(&smp_alt, flags); - DPRINTK("%s: %s\n", __FUNCTION__, item->name); + spin_unlock(&smp_alt); + DPRINTK("%s: %s\n", __func__, item->name); kfree(item); return; } - spin_unlock_irqrestore(&smp_alt, flags); + spin_unlock(&smp_alt); } void alternatives_smp_switch(int smp) { struct smp_alt_module *mod; - unsigned long flags; #ifdef CONFIG_LOCKDEP /* @@ -359,7 +359,7 @@ void alternatives_smp_switch(int smp) return; BUG_ON(!smp && (num_online_cpus() > 1)); - spin_lock_irqsave(&smp_alt, flags); + spin_lock(&smp_alt); /* * Avoid unnecessary switches because it forces JIT based VMs to @@ -383,7 +383,7 @@ void alternatives_smp_switch(int smp) mod->text, mod->text_end); } smp_mode = smp; - spin_unlock_irqrestore(&smp_alt, flags); + spin_unlock(&smp_alt); } #endif @@ -411,7 +411,7 @@ void apply_paravirt(struct paravirt_patch_site *start, /* Pad the rest with nops */ add_nops(insnbuf + used, p->len - used); - text_poke(p->instr, insnbuf, p->len); + text_poke_early(p->instr, insnbuf, p->len); } } extern struct paravirt_patch_site __start_parainstructions[], @@ -420,8 +420,6 @@ extern struct paravirt_patch_site __start_parainstructions[], void __init alternative_instructions(void) { - unsigned long flags; - /* The patching is not fully atomic, so try to avoid local interruptions that might execute the to be patched code. Other CPUs are not running. */ @@ -430,7 +428,6 @@ void __init alternative_instructions(void) stop_mce(); #endif - local_irq_save(flags); apply_alternatives(__alt_instructions, __alt_instructions_end); /* switch to patch-once-at-boottime-only mode and free the @@ -462,7 +459,6 @@ void __init alternative_instructions(void) } #endif apply_paravirt(__parainstructions, __parainstructions_end); - local_irq_restore(flags); if (smp_alt_once) free_init_pages("SMP alternatives", @@ -475,18 +471,71 @@ void __init alternative_instructions(void) #endif } -/* - * Warning: +/** + * text_poke_early - Update instructions on a live kernel at boot time + * @addr: address to modify + * @opcode: source of the copy + * @len: length to copy + * * When you use this code to patch more than one byte of an instruction * you need to make sure that other CPUs cannot execute this code in parallel. - * Also no thread must be currently preempted in the middle of these instructions. - * And on the local CPU you need to be protected again NMI or MCE handlers - * seeing an inconsistent instruction while you patch. + * Also no thread must be currently preempted in the middle of these + * instructions. And on the local CPU you need to be protected again NMI or MCE + * handlers seeing an inconsistent instruction while you patch. */ -void __kprobes text_poke(void *addr, unsigned char *opcode, int len) +void *text_poke_early(void *addr, const void *opcode, size_t len) { + unsigned long flags; + local_irq_save(flags); memcpy(addr, opcode, len); + local_irq_restore(flags); + sync_core(); + /* Could also do a CLFLUSH here to speed up CPU recovery; but + that causes hangs on some VIA CPUs. */ + return addr; +} + +/** + * text_poke - Update instructions on a live kernel + * @addr: address to modify + * @opcode: source of the copy + * @len: length to copy + * + * Only atomic text poke/set should be allowed when not doing early patching. + * It means the size must be writable atomically and the address must be aligned + * in a way that permits an atomic write. It also makes sure we fit on a single + * page. + */ +void *__kprobes text_poke(void *addr, const void *opcode, size_t len) +{ + unsigned long flags; + char *vaddr; + int nr_pages = 2; + + BUG_ON(len > sizeof(long)); + BUG_ON((((long)addr + len - 1) & ~(sizeof(long) - 1)) + - ((long)addr & ~(sizeof(long) - 1))); + if (kernel_text_address((unsigned long)addr)) { + struct page *pages[2] = { virt_to_page(addr), + virt_to_page(addr + PAGE_SIZE) }; + if (!pages[1]) + nr_pages = 1; + vaddr = vmap(pages, nr_pages, VM_MAP, PAGE_KERNEL); + BUG_ON(!vaddr); + local_irq_save(flags); + memcpy(&vaddr[(unsigned long)addr & ~PAGE_MASK], opcode, len); + local_irq_restore(flags); + vunmap(vaddr); + } else { + /* + * modules are in vmalloc'ed memory, always writable. + */ + local_irq_save(flags); + memcpy(addr, opcode, len); + local_irq_restore(flags); + } sync_core(); /* Could also do a CLFLUSH here to speed up CPU recovery; but that causes hangs on some VIA CPUs. */ + return addr; } diff --git a/arch/x86/kernel/aperture_64.c b/arch/x86/kernel/aperture_64.c index 00df126169b4..479926d9e004 100644 --- a/arch/x86/kernel/aperture_64.c +++ b/arch/x86/kernel/aperture_64.c @@ -27,11 +27,11 @@ #include <asm/k8.h> int gart_iommu_aperture; -int gart_iommu_aperture_disabled __initdata = 0; -int gart_iommu_aperture_allowed __initdata = 0; +int gart_iommu_aperture_disabled __initdata; +int gart_iommu_aperture_allowed __initdata; int fallback_aper_order __initdata = 1; /* 64MB */ -int fallback_aper_force __initdata = 0; +int fallback_aper_force __initdata; int fix_aperture __initdata = 1; diff --git a/arch/x86/kernel/apic_32.c b/arch/x86/kernel/apic_32.c index 35a568ea8400..687208190b06 100644 --- a/arch/x86/kernel/apic_32.c +++ b/arch/x86/kernel/apic_32.c @@ -50,6 +50,11 @@ # error SPURIOUS_APIC_VECTOR definition error #endif +unsigned long mp_lapic_addr; + +DEFINE_PER_CPU(u16, x86_bios_cpu_apicid) = BAD_APICID; +EXPORT_PER_CPU_SYMBOL(x86_bios_cpu_apicid); + /* * Knob to control our willingness to enable the local APIC. * @@ -621,6 +626,35 @@ int setup_profiling_timer(unsigned int multiplier) } /* + * Setup extended LVT, AMD specific (K8, family 10h) + * + * Vector mappings are hard coded. On K8 only offset 0 (APIC500) and + * MCE interrupts are supported. Thus MCE offset must be set to 0. + */ + +#define APIC_EILVT_LVTOFF_MCE 0 +#define APIC_EILVT_LVTOFF_IBS 1 + +static void setup_APIC_eilvt(u8 lvt_off, u8 vector, u8 msg_type, u8 mask) +{ + unsigned long reg = (lvt_off << 4) + APIC_EILVT0; + unsigned int v = (mask << 16) | (msg_type << 8) | vector; + apic_write(reg, v); +} + +u8 setup_APIC_eilvt_mce(u8 vector, u8 msg_type, u8 mask) +{ + setup_APIC_eilvt(APIC_EILVT_LVTOFF_MCE, vector, msg_type, mask); + return APIC_EILVT_LVTOFF_MCE; +} + +u8 setup_APIC_eilvt_ibs(u8 vector, u8 msg_type, u8 mask) +{ + setup_APIC_eilvt(APIC_EILVT_LVTOFF_IBS, vector, msg_type, mask); + return APIC_EILVT_LVTOFF_IBS; +} + +/* * Local APIC start and shutdown */ @@ -868,12 +902,50 @@ void __init init_bsp_APIC(void) apic_write_around(APIC_LVT1, value); } +void __cpuinit lapic_setup_esr(void) +{ + unsigned long oldvalue, value, maxlvt; + if (lapic_is_integrated() && !esr_disable) { + /* !82489DX */ + maxlvt = lapic_get_maxlvt(); + if (maxlvt > 3) /* Due to the Pentium erratum 3AP. */ + apic_write(APIC_ESR, 0); + oldvalue = apic_read(APIC_ESR); + + /* enables sending errors */ + value = ERROR_APIC_VECTOR; + apic_write_around(APIC_LVTERR, value); + /* + * spec says clear errors after enabling vector. + */ + if (maxlvt > 3) + apic_write(APIC_ESR, 0); + value = apic_read(APIC_ESR); + if (value != oldvalue) + apic_printk(APIC_VERBOSE, "ESR value before enabling " + "vector: 0x%08lx after: 0x%08lx\n", + oldvalue, value); + } else { + if (esr_disable) + /* + * Something untraceable is creating bad interrupts on + * secondary quads ... for the moment, just leave the + * ESR disabled - we can't do anything useful with the + * errors anyway - mbligh + */ + printk(KERN_INFO "Leaving ESR disabled.\n"); + else + printk(KERN_INFO "No ESR for 82489DX.\n"); + } +} + + /** * setup_local_APIC - setup the local APIC */ void __cpuinit setup_local_APIC(void) { - unsigned long oldvalue, value, maxlvt, integrated; + unsigned long value, integrated; int i, j; /* Pound the ESR really hard over the head with a big hammer - mbligh */ @@ -997,40 +1069,13 @@ void __cpuinit setup_local_APIC(void) if (!integrated) /* 82489DX */ value |= APIC_LVT_LEVEL_TRIGGER; apic_write_around(APIC_LVT1, value); +} - if (integrated && !esr_disable) { - /* !82489DX */ - maxlvt = lapic_get_maxlvt(); - if (maxlvt > 3) /* Due to the Pentium erratum 3AP. */ - apic_write(APIC_ESR, 0); - oldvalue = apic_read(APIC_ESR); - - /* enables sending errors */ - value = ERROR_APIC_VECTOR; - apic_write_around(APIC_LVTERR, value); - /* - * spec says clear errors after enabling vector. - */ - if (maxlvt > 3) - apic_write(APIC_ESR, 0); - value = apic_read(APIC_ESR); - if (value != oldvalue) - apic_printk(APIC_VERBOSE, "ESR value before enabling " - "vector: 0x%08lx after: 0x%08lx\n", - oldvalue, value); - } else { - if (esr_disable) - /* - * Something untraceable is creating bad interrupts on - * secondary quads ... for the moment, just leave the - * ESR disabled - we can't do anything useful with the - * errors anyway - mbligh - */ - printk(KERN_INFO "Leaving ESR disabled.\n"); - else - printk(KERN_INFO "No ESR for 82489DX.\n"); - } +void __cpuinit end_local_APIC_setup(void) +{ + unsigned long value; + lapic_setup_esr(); /* Disable the local apic timer */ value = apic_read(APIC_LVTT); value |= (APIC_LVT_MASKED | LOCAL_TIMER_VECTOR); @@ -1147,7 +1192,7 @@ void __init init_apic_mappings(void) * default configuration (or the MP table is broken). */ if (boot_cpu_physical_apicid == -1U) - boot_cpu_physical_apicid = GET_APIC_ID(apic_read(APIC_ID)); + boot_cpu_physical_apicid = GET_APIC_ID(read_apic_id()); #ifdef CONFIG_X86_IO_APIC { @@ -1185,6 +1230,9 @@ fake_ioapic_page: * This initializes the IO-APIC and APIC hardware if this is * a UP kernel. */ + +int apic_version[MAX_APICS]; + int __init APIC_init_uniprocessor(void) { if (enable_local_apic < 0) @@ -1214,12 +1262,13 @@ int __init APIC_init_uniprocessor(void) * might be zero if read from MP tables. Get it from LAPIC. */ #ifdef CONFIG_CRASH_DUMP - boot_cpu_physical_apicid = GET_APIC_ID(apic_read(APIC_ID)); + boot_cpu_physical_apicid = GET_APIC_ID(read_apic_id()); #endif phys_cpu_present_map = physid_mask_of_physid(boot_cpu_physical_apicid); setup_local_APIC(); + end_local_APIC_setup(); #ifdef CONFIG_X86_IO_APIC if (smp_found_config) if (!skip_ioapic_setup && nr_ioapics) @@ -1288,6 +1337,29 @@ void smp_error_interrupt(struct pt_regs *regs) irq_exit(); } +#ifdef CONFIG_SMP +void __init smp_intr_init(void) +{ + /* + * IRQ0 must be given a fixed assignment and initialized, + * because it's used before the IO-APIC is set up. + */ + set_intr_gate(FIRST_DEVICE_VECTOR, interrupt[0]); + + /* + * The reschedule interrupt is a CPU-to-CPU reschedule-helper + * IPI, driven by wakeup. + */ + set_intr_gate(RESCHEDULE_VECTOR, reschedule_interrupt); + + /* IPI for invalidation */ + set_intr_gate(INVALIDATE_TLB_VECTOR, invalidate_interrupt); + + /* IPI for generic function call */ + set_intr_gate(CALL_FUNCTION_VECTOR, call_function_interrupt); +} +#endif + /* * Initialize APIC interrupts */ @@ -1394,6 +1466,88 @@ void disconnect_bsp_APIC(int virt_wire_setup) } } +unsigned int __cpuinitdata maxcpus = NR_CPUS; + +void __cpuinit generic_processor_info(int apicid, int version) +{ + int cpu; + cpumask_t tmp_map; + physid_mask_t phys_cpu; + + /* + * Validate version + */ + if (version == 0x0) { + printk(KERN_WARNING "BIOS bug, APIC version is 0 for CPU#%d! " + "fixing up to 0x10. (tell your hw vendor)\n", + version); + version = 0x10; + } + apic_version[apicid] = version; + + phys_cpu = apicid_to_cpu_present(apicid); + physids_or(phys_cpu_present_map, phys_cpu_present_map, phys_cpu); + + if (num_processors >= NR_CPUS) { + printk(KERN_WARNING "WARNING: NR_CPUS limit of %i reached." + " Processor ignored.\n", NR_CPUS); + return; + } + + if (num_processors >= maxcpus) { + printk(KERN_WARNING "WARNING: maxcpus limit of %i reached." + " Processor ignored.\n", maxcpus); + return; + } + + num_processors++; + cpus_complement(tmp_map, cpu_present_map); + cpu = first_cpu(tmp_map); + + if (apicid == boot_cpu_physical_apicid) + /* + * x86_bios_cpu_apicid is required to have processors listed + * in same order as logical cpu numbers. Hence the first + * entry is BSP, and so on. + */ + cpu = 0; + + /* + * Would be preferable to switch to bigsmp when CONFIG_HOTPLUG_CPU=y + * but we need to work other dependencies like SMP_SUSPEND etc + * before this can be done without some confusion. + * if (CPU_HOTPLUG_ENABLED || num_processors > 8) + * - Ashok Raj <ashok.raj@intel.com> + */ + if (num_processors > 8) { + switch (boot_cpu_data.x86_vendor) { + case X86_VENDOR_INTEL: + if (!APIC_XAPIC(version)) { + def_to_bigsmp = 0; + break; + } + /* If P4 and above fall through */ + case X86_VENDOR_AMD: + def_to_bigsmp = 1; + } + } +#ifdef CONFIG_SMP + /* are we being called early in kernel startup? */ + if (x86_cpu_to_apicid_early_ptr) { + u16 *cpu_to_apicid = x86_cpu_to_apicid_early_ptr; + u16 *bios_cpu_apicid = x86_bios_cpu_apicid_early_ptr; + + cpu_to_apicid[cpu] = apicid; + bios_cpu_apicid[cpu] = apicid; + } else { + per_cpu(x86_cpu_to_apicid, cpu) = apicid; + per_cpu(x86_bios_cpu_apicid, cpu) = apicid; + } +#endif + cpu_set(cpu, cpu_possible_map); + cpu_set(cpu, cpu_present_map); +} + /* * Power management */ diff --git a/arch/x86/kernel/apic_64.c b/arch/x86/kernel/apic_64.c index d8d03e09dea2..9e8e5c050c55 100644 --- a/arch/x86/kernel/apic_64.c +++ b/arch/x86/kernel/apic_64.c @@ -34,13 +34,15 @@ #include <asm/mpspec.h> #include <asm/hpet.h> #include <asm/pgalloc.h> -#include <asm/mach_apic.h> #include <asm/nmi.h> #include <asm/idle.h> #include <asm/proto.h> #include <asm/timex.h> #include <asm/apic.h> +#include <mach_ipi.h> +#include <mach_apic.h> + int disable_apic_timer __cpuinitdata; static int apic_calibrate_pmtmr __initdata; int disable_apic; @@ -83,6 +85,12 @@ static DEFINE_PER_CPU(struct clock_event_device, lapic_events); static unsigned long apic_phys; +unsigned long mp_lapic_addr; + +DEFINE_PER_CPU(u16, x86_bios_cpu_apicid) = BAD_APICID; +EXPORT_PER_CPU_SYMBOL(x86_bios_cpu_apicid); + +unsigned int __cpuinitdata maxcpus = NR_CPUS; /* * Get the LAPIC version */ @@ -431,7 +439,8 @@ void __cpuinit check_boot_apic_timer_broadcast(void) lapic_clockevent.features |= CLOCK_EVT_FEAT_DUMMY; local_irq_enable(); - clockevents_notify(CLOCK_EVT_NOTIFY_BROADCAST_FORCE, &boot_cpu_id); + clockevents_notify(CLOCK_EVT_NOTIFY_BROADCAST_FORCE, + &boot_cpu_physical_apicid); local_irq_disable(); } @@ -640,10 +649,10 @@ int __init verify_local_APIC(void) /* * The ID register is read/write in a real APIC. */ - reg0 = apic_read(APIC_ID); + reg0 = read_apic_id(); apic_printk(APIC_DEBUG, "Getting ID: %x\n", reg0); apic_write(APIC_ID, reg0 ^ APIC_ID_MASK); - reg1 = apic_read(APIC_ID); + reg1 = read_apic_id(); apic_printk(APIC_DEBUG, "Getting ID: %x\n", reg1); apic_write(APIC_ID, reg0); if (reg1 != (reg0 ^ APIC_ID_MASK)) @@ -728,6 +737,7 @@ void __cpuinit setup_local_APIC(void) unsigned int value; int i, j; + preempt_disable(); value = apic_read(APIC_LVR); BUILD_BUG_ON((SPURIOUS_APIC_VECTOR & 0x0f) != 0x0f); @@ -821,6 +831,7 @@ void __cpuinit setup_local_APIC(void) else value = APIC_DM_NMI | APIC_LVT_MASKED; apic_write(APIC_LVT1, value); + preempt_enable(); } void __cpuinit lapic_setup_esr(void) @@ -857,10 +868,34 @@ static int __init detect_init_APIC(void) } mp_lapic_addr = APIC_DEFAULT_PHYS_BASE; - boot_cpu_id = 0; + boot_cpu_physical_apicid = 0; return 0; } +void __init early_init_lapic_mapping(void) +{ + unsigned long apic_phys; + + /* + * If no local APIC can be found then go out + * : it means there is no mpatable and MADT + */ + if (!smp_found_config) + return; + + apic_phys = mp_lapic_addr; + + set_fixmap_nocache(FIX_APIC_BASE, apic_phys); + apic_printk(APIC_VERBOSE, "mapped APIC to %16lx (%16lx)\n", + APIC_BASE, apic_phys); + + /* + * Fetch the APIC ID of the BSP in case we have a + * default configuration (or the MP table is broken). + */ + boot_cpu_physical_apicid = GET_APIC_ID(read_apic_id()); +} + /** * init_apic_mappings - initialize APIC mappings */ @@ -881,16 +916,11 @@ void __init init_apic_mappings(void) apic_printk(APIC_VERBOSE, "mapped APIC to %16lx (%16lx)\n", APIC_BASE, apic_phys); - /* Put local APIC into the resource map. */ - lapic_resource.start = apic_phys; - lapic_resource.end = lapic_resource.start + PAGE_SIZE - 1; - insert_resource(&iomem_resource, &lapic_resource); - /* * Fetch the APIC ID of the BSP in case we have a * default configuration (or the MP table is broken). */ - boot_cpu_id = GET_APIC_ID(apic_read(APIC_ID)); + boot_cpu_physical_apicid = GET_APIC_ID(read_apic_id()); } /* @@ -911,8 +941,8 @@ int __init APIC_init_uniprocessor(void) verify_local_APIC(); - phys_cpu_present_map = physid_mask_of_physid(boot_cpu_id); - apic_write(APIC_ID, SET_APIC_ID(boot_cpu_id)); + phys_cpu_present_map = physid_mask_of_physid(boot_cpu_physical_apicid); + apic_write(APIC_ID, SET_APIC_ID(boot_cpu_physical_apicid)); setup_local_APIC(); @@ -1029,6 +1059,52 @@ void disconnect_bsp_APIC(int virt_wire_setup) apic_write(APIC_LVT1, value); } +void __cpuinit generic_processor_info(int apicid, int version) +{ + int cpu; + cpumask_t tmp_map; + + if (num_processors >= NR_CPUS) { + printk(KERN_WARNING "WARNING: NR_CPUS limit of %i reached." + " Processor ignored.\n", NR_CPUS); + return; + } + + if (num_processors >= maxcpus) { + printk(KERN_WARNING "WARNING: maxcpus limit of %i reached." + " Processor ignored.\n", maxcpus); + return; + } + + num_processors++; + cpus_complement(tmp_map, cpu_present_map); + cpu = first_cpu(tmp_map); + + physid_set(apicid, phys_cpu_present_map); + if (apicid == boot_cpu_physical_apicid) { + /* + * x86_bios_cpu_apicid is required to have processors listed + * in same order as logical cpu numbers. Hence the first + * entry is BSP, and so on. + */ + cpu = 0; + } + /* are we being called early in kernel startup? */ + if (x86_cpu_to_apicid_early_ptr) { + u16 *cpu_to_apicid = x86_cpu_to_apicid_early_ptr; + u16 *bios_cpu_apicid = x86_bios_cpu_apicid_early_ptr; + + cpu_to_apicid[cpu] = apicid; + bios_cpu_apicid[cpu] = apicid; + } else { + per_cpu(x86_cpu_to_apicid, cpu) = apicid; + per_cpu(x86_bios_cpu_apicid, cpu) = apicid; + } + + cpu_set(cpu, cpu_possible_map); + cpu_set(cpu, cpu_present_map); +} + /* * Power management */ @@ -1065,7 +1141,7 @@ static int lapic_suspend(struct sys_device *dev, pm_message_t state) maxlvt = lapic_get_maxlvt(); - apic_pm_state.apic_id = apic_read(APIC_ID); + apic_pm_state.apic_id = read_apic_id(); apic_pm_state.apic_taskpri = apic_read(APIC_TASKPRI); apic_pm_state.apic_ldr = apic_read(APIC_LDR); apic_pm_state.apic_dfr = apic_read(APIC_DFR); @@ -1180,9 +1256,19 @@ __cpuinit int apic_is_clustered_box(void) { int i, clusters, zeros; unsigned id; - u16 *bios_cpu_apicid = x86_bios_cpu_apicid_early_ptr; + u16 *bios_cpu_apicid; DECLARE_BITMAP(clustermap, NUM_APIC_CLUSTERS); + /* + * there is not this kind of box with AMD CPU yet. + * Some AMD box with quadcore cpu and 8 sockets apicid + * will be [4, 0x23] or [8, 0x27] could be thought to + * vsmp box still need checking... + */ + if ((boot_cpu_data.x86_vendor == X86_VENDOR_AMD) && !is_vsmp_box()) + return 0; + + bios_cpu_apicid = x86_bios_cpu_apicid_early_ptr; bitmap_zero(clustermap, NUM_APIC_CLUSTERS); for (i = 0; i < NR_CPUS; i++) { @@ -1219,6 +1305,12 @@ __cpuinit int apic_is_clustered_box(void) ++zeros; } + /* ScaleMP vSMPowered boxes have one cluster per board and TSCs are + * not guaranteed to be synced between boards + */ + if (is_vsmp_box() && clusters > 1) + return 1; + /* * If clusters > 2, then should be multi-chassis. * May have to revisit this when multi-core + hyperthreaded CPUs come @@ -1290,3 +1382,21 @@ static __init int setup_apicpmtimer(char *s) } __setup("apicpmtimer", setup_apicpmtimer); +static int __init lapic_insert_resource(void) +{ + if (!apic_phys) + return -1; + + /* Put local APIC into the resource map. */ + lapic_resource.start = apic_phys; + lapic_resource.end = lapic_resource.start + PAGE_SIZE - 1; + insert_resource(&iomem_resource, &lapic_resource); + + return 0; +} + +/* + * need call insert after e820_reserve_resources() + * that is using request_resource + */ +late_initcall(lapic_insert_resource); diff --git a/arch/x86/kernel/apm_32.c b/arch/x86/kernel/apm_32.c index d4438ef296d8..f0030a0999c7 100644 --- a/arch/x86/kernel/apm_32.c +++ b/arch/x86/kernel/apm_32.c @@ -2217,7 +2217,6 @@ static struct dmi_system_id __initdata apm_dmi_table[] = { */ static int __init apm_init(void) { - struct proc_dir_entry *apm_proc; struct desc_struct *gdt; int err; @@ -2322,9 +2321,7 @@ static int __init apm_init(void) set_base(gdt[APM_DS >> 3], __va((unsigned long)apm_info.bios.dseg << 4)); - apm_proc = create_proc_entry("apm", 0, NULL); - if (apm_proc) - apm_proc->proc_fops = &apm_file_ops; + proc_create("apm", 0, NULL, &apm_file_ops); kapmd_task = kthread_create(apm, NULL, "kapmd"); if (IS_ERR(kapmd_task)) { diff --git a/arch/x86/kernel/asm-offsets_32.c b/arch/x86/kernel/asm-offsets_32.c index 8ea040124f7d..670c3c311289 100644 --- a/arch/x86/kernel/asm-offsets_32.c +++ b/arch/x86/kernel/asm-offsets_32.c @@ -10,7 +10,7 @@ #include <linux/personality.h> #include <linux/suspend.h> #include <asm/ucontext.h> -#include "sigframe_32.h" +#include "sigframe.h" #include <asm/pgtable.h> #include <asm/fixmap.h> #include <asm/processor.h> diff --git a/arch/x86/kernel/bugs_64.c b/arch/x86/kernel/bugs_64.c index 8f520f93ffd4..9a3ed0649d4e 100644 --- a/arch/x86/kernel/bugs_64.c +++ b/arch/x86/kernel/bugs_64.c @@ -9,13 +9,25 @@ #include <asm/bugs.h> #include <asm/processor.h> #include <asm/mtrr.h> +#include <asm/cacheflush.h> void __init check_bugs(void) { - identify_cpu(&boot_cpu_data); + identify_boot_cpu(); #if !defined(CONFIG_SMP) printk("CPU: "); print_cpu_info(&boot_cpu_data); #endif alternative_instructions(); + + /* + * Make sure the first 2MB area is not mapped by huge pages + * There are typically fixed size MTRRs in there and overlapping + * MTRRs into large pages causes slow downs. + * + * Right now we don't do that with gbpages because there seems + * very little benefit for that case. + */ + if (!direct_gbpages) + set_memory_4k((unsigned long)__va(0), 1); } diff --git a/arch/x86/kernel/cpu/Makefile b/arch/x86/kernel/cpu/Makefile index a0c4d7c5dbd7..ee7c45235e54 100644 --- a/arch/x86/kernel/cpu/Makefile +++ b/arch/x86/kernel/cpu/Makefile @@ -3,9 +3,9 @@ # obj-y := intel_cacheinfo.o addon_cpuid_features.o -obj-y += feature_names.o +obj-y += proc.o feature_names.o -obj-$(CONFIG_X86_32) += common.o proc.o bugs.o +obj-$(CONFIG_X86_32) += common.o bugs.o obj-$(CONFIG_X86_32) += amd.o obj-$(CONFIG_X86_32) += cyrix.o obj-$(CONFIG_X86_32) += centaur.o diff --git a/arch/x86/kernel/cpu/amd.c b/arch/x86/kernel/cpu/amd.c index 693e353999cd..0173065dc3b7 100644 --- a/arch/x86/kernel/cpu/amd.c +++ b/arch/x86/kernel/cpu/amd.c @@ -4,8 +4,8 @@ #include <asm/io.h> #include <asm/processor.h> #include <asm/apic.h> -#include <asm/mach_apic.h> +#include <mach_apic.h> #include "cpu.h" /* @@ -20,7 +20,7 @@ * the chip setting when fixing the bug but they also tweaked some * performance at the same time.. */ - + extern void vide(void); __asm__(".align 4\nvide: ret"); @@ -63,12 +63,12 @@ static __cpuinit int amd_apic_timer_broken(void) int force_mwait __cpuinitdata; -void __cpuinit early_init_amd(struct cpuinfo_x86 *c) +static void __cpuinit early_init_amd(struct cpuinfo_x86 *c) { if (cpuid_eax(0x80000000) >= 0x80000007) { c->x86_power = cpuid_edx(0x80000007); if (c->x86_power & (1<<8)) - set_bit(X86_FEATURE_CONSTANT_TSC, c->x86_capability); + set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC); } } @@ -81,7 +81,8 @@ static void __cpuinit init_amd(struct cpuinfo_x86 *c) #ifdef CONFIG_SMP unsigned long long value; - /* Disable TLB flush filter by setting HWCR.FFDIS on K8 + /* + * Disable TLB flush filter by setting HWCR.FFDIS on K8 * bit 6 of msr C001_0015 * * Errata 63 for SH-B3 steppings @@ -102,15 +103,16 @@ static void __cpuinit init_amd(struct cpuinfo_x86 *c) * no bus pipeline) */ - /* Bit 31 in normal CPUID used for nonstandard 3DNow ID; - 3DNow is IDd by bit 31 in extended CPUID (1*32+31) anyway */ - clear_bit(0*32+31, c->x86_capability); - + /* + * Bit 31 in normal CPUID used for nonstandard 3DNow ID; + * 3DNow is IDd by bit 31 in extended CPUID (1*32+31) anyway + */ + clear_cpu_cap(c, 0*32+31); + r = get_model_name(c); - switch(c->x86) - { - case 4: + switch (c->x86) { + case 4: /* * General Systems BIOSen alias the cpu frequency registers * of the Elan at 0x000df000. Unfortuantly, one of the Linux @@ -120,61 +122,60 @@ static void __cpuinit init_amd(struct cpuinfo_x86 *c) #define CBAR (0xfffc) /* Configuration Base Address (32-bit) */ #define CBAR_ENB (0x80000000) #define CBAR_KEY (0X000000CB) - if (c->x86_model==9 || c->x86_model == 10) { + if (c->x86_model == 9 || c->x86_model == 10) { if (inl (CBAR) & CBAR_ENB) outl (0 | CBAR_KEY, CBAR); } break; - case 5: - if( c->x86_model < 6 ) - { + case 5: + if (c->x86_model < 6) { /* Based on AMD doc 20734R - June 2000 */ - if ( c->x86_model == 0 ) { - clear_bit(X86_FEATURE_APIC, c->x86_capability); - set_bit(X86_FEATURE_PGE, c->x86_capability); + if (c->x86_model == 0) { + clear_cpu_cap(c, X86_FEATURE_APIC); + set_cpu_cap(c, X86_FEATURE_PGE); } break; } - - if ( c->x86_model == 6 && c->x86_mask == 1 ) { + + if (c->x86_model == 6 && c->x86_mask == 1) { const int K6_BUG_LOOP = 1000000; int n; void (*f_vide)(void); unsigned long d, d2; - + printk(KERN_INFO "AMD K6 stepping B detected - "); - + /* - * It looks like AMD fixed the 2.6.2 bug and improved indirect + * It looks like AMD fixed the 2.6.2 bug and improved indirect * calls at the same time. */ n = K6_BUG_LOOP; f_vide = vide; rdtscl(d); - while (n--) + while (n--) f_vide(); rdtscl(d2); d = d2-d; - if (d > 20*K6_BUG_LOOP) + if (d > 20*K6_BUG_LOOP) printk("system stability may be impaired when more than 32 MB are used.\n"); - else + else printk("probably OK (after B9730xxxx).\n"); printk(KERN_INFO "Please see http://membres.lycos.fr/poulot/k6bug.html\n"); } /* K6 with old style WHCR */ if (c->x86_model < 8 || - (c->x86_model== 8 && c->x86_mask < 8)) { + (c->x86_model == 8 && c->x86_mask < 8)) { /* We can only write allocate on the low 508Mb */ - if(mbytes>508) - mbytes=508; + if (mbytes > 508) + mbytes = 508; rdmsr(MSR_K6_WHCR, l, h); - if ((l&0x0000FFFF)==0) { + if ((l&0x0000FFFF) == 0) { unsigned long flags; - l=(1<<0)|((mbytes/4)<<1); + l = (1<<0)|((mbytes/4)<<1); local_irq_save(flags); wbinvd(); wrmsr(MSR_K6_WHCR, l, h); @@ -185,17 +186,17 @@ static void __cpuinit init_amd(struct cpuinfo_x86 *c) break; } - if ((c->x86_model == 8 && c->x86_mask >7) || + if ((c->x86_model == 8 && c->x86_mask > 7) || c->x86_model == 9 || c->x86_model == 13) { /* The more serious chips .. */ - if(mbytes>4092) - mbytes=4092; + if (mbytes > 4092) + mbytes = 4092; rdmsr(MSR_K6_WHCR, l, h); - if ((l&0xFFFF0000)==0) { + if ((l&0xFFFF0000) == 0) { unsigned long flags; - l=((mbytes>>2)<<22)|(1<<16); + l = ((mbytes>>2)<<22)|(1<<16); local_irq_save(flags); wbinvd(); wrmsr(MSR_K6_WHCR, l, h); @@ -207,7 +208,7 @@ static void __cpuinit init_amd(struct cpuinfo_x86 *c) /* Set MTRR capability flag if appropriate */ if (c->x86_model == 13 || c->x86_model == 9 || (c->x86_model == 8 && c->x86_mask >= 8)) - set_bit(X86_FEATURE_K6_MTRR, c->x86_capability); + set_cpu_cap(c, X86_FEATURE_K6_MTRR); break; } @@ -217,10 +218,11 @@ static void __cpuinit init_amd(struct cpuinfo_x86 *c) break; } break; - case 6: /* An Athlon/Duron */ - - /* Bit 15 of Athlon specific MSR 15, needs to be 0 - * to enable SSE on Palomino/Morgan/Barton CPU's. + case 6: /* An Athlon/Duron */ + + /* + * Bit 15 of Athlon specific MSR 15, needs to be 0 + * to enable SSE on Palomino/Morgan/Barton CPU's. * If the BIOS didn't enable it already, enable it here. */ if (c->x86_model >= 6 && c->x86_model <= 10) { @@ -229,15 +231,16 @@ static void __cpuinit init_amd(struct cpuinfo_x86 *c) rdmsr(MSR_K7_HWCR, l, h); l &= ~0x00008000; wrmsr(MSR_K7_HWCR, l, h); - set_bit(X86_FEATURE_XMM, c->x86_capability); + set_cpu_cap(c, X86_FEATURE_XMM); } } - /* It's been determined by AMD that Athlons since model 8 stepping 1 + /* + * It's been determined by AMD that Athlons since model 8 stepping 1 * are more robust with CLK_CTL set to 200xxxxx instead of 600xxxxx * As per AMD technical note 27212 0.2 */ - if ((c->x86_model == 8 && c->x86_mask>=1) || (c->x86_model > 8)) { + if ((c->x86_model == 8 && c->x86_mask >= 1) || (c->x86_model > 8)) { rdmsr(MSR_K7_CLK_CTL, l, h); if ((l & 0xfff00000) != 0x20000000) { printk ("CPU: CLK_CTL MSR was %x. Reprogramming to %x\n", l, @@ -253,20 +256,19 @@ static void __cpuinit init_amd(struct cpuinfo_x86 *c) /* Use K8 tuning for Fam10h and Fam11h */ case 0x10: case 0x11: - set_bit(X86_FEATURE_K8, c->x86_capability); + set_cpu_cap(c, X86_FEATURE_K8); break; case 6: - set_bit(X86_FEATURE_K7, c->x86_capability); + set_cpu_cap(c, X86_FEATURE_K7); break; } if (c->x86 >= 6) - set_bit(X86_FEATURE_FXSAVE_LEAK, c->x86_capability); + set_cpu_cap(c, X86_FEATURE_FXSAVE_LEAK); display_cacheinfo(c); - if (cpuid_eax(0x80000000) >= 0x80000008) { + if (cpuid_eax(0x80000000) >= 0x80000008) c->x86_max_cores = (cpuid_ecx(0x80000008) & 0xff) + 1; - } #ifdef CONFIG_X86_HT /* @@ -302,20 +304,20 @@ static void __cpuinit init_amd(struct cpuinfo_x86 *c) /* K6s reports MCEs but don't actually have all the MSRs */ if (c->x86 < 6) - clear_bit(X86_FEATURE_MCE, c->x86_capability); + clear_cpu_cap(c, X86_FEATURE_MCE); if (cpu_has_xmm2) - set_bit(X86_FEATURE_MFENCE_RDTSC, c->x86_capability); + set_cpu_cap(c, X86_FEATURE_MFENCE_RDTSC); } -static unsigned int __cpuinit amd_size_cache(struct cpuinfo_x86 * c, unsigned int size) +static unsigned int __cpuinit amd_size_cache(struct cpuinfo_x86 *c, unsigned int size) { /* AMD errata T13 (order #21922) */ if ((c->x86 == 6)) { if (c->x86_model == 3 && c->x86_mask == 0) /* Duron Rev A0 */ size = 64; if (c->x86_model == 4 && - (c->x86_mask==0 || c->x86_mask==1)) /* Tbird rev A1/A2 */ + (c->x86_mask == 0 || c->x86_mask == 1)) /* Tbird rev A1/A2 */ size = 256; } return size; @@ -323,19 +325,20 @@ static unsigned int __cpuinit amd_size_cache(struct cpuinfo_x86 * c, unsigned in static struct cpu_dev amd_cpu_dev __cpuinitdata = { .c_vendor = "AMD", - .c_ident = { "AuthenticAMD" }, + .c_ident = { "AuthenticAMD" }, .c_models = { { .vendor = X86_VENDOR_AMD, .family = 4, .model_names = { [3] = "486 DX/2", [7] = "486 DX/2-WB", - [8] = "486 DX/4", - [9] = "486 DX/4-WB", + [8] = "486 DX/4", + [9] = "486 DX/4-WB", [14] = "Am5x86-WT", - [15] = "Am5x86-WB" + [15] = "Am5x86-WB" } }, }, + .c_early_init = early_init_amd, .c_init = init_amd, .c_size_cache = amd_size_cache, }; @@ -345,3 +348,5 @@ int __init amd_init_cpu(void) cpu_devs[X86_VENDOR_AMD] = &amd_cpu_dev; return 0; } + +cpu_vendor_dev_register(X86_VENDOR_AMD, &amd_cpu_dev); diff --git a/arch/x86/kernel/cpu/centaur.c b/arch/x86/kernel/cpu/centaur.c index 9681fa15ddf0..e0f45edd6a55 100644 --- a/arch/x86/kernel/cpu/centaur.c +++ b/arch/x86/kernel/cpu/centaur.c @@ -1,31 +1,34 @@ #include <linux/kernel.h> #include <linux/init.h> #include <linux/bitops.h> + #include <asm/processor.h> #include <asm/msr.h> #include <asm/e820.h> #include <asm/mtrr.h> + #include "cpu.h" #ifdef CONFIG_X86_OOSTORE static u32 __cpuinit power2(u32 x) { - u32 s=1; - while(s<=x) - s<<=1; - return s>>=1; + u32 s = 1; + + while (s <= x) + s <<= 1; + + return s >>= 1; } /* - * Set up an actual MCR + * Set up an actual MCR */ - static void __cpuinit centaur_mcr_insert(int reg, u32 base, u32 size, int key) { u32 lo, hi; - + hi = base & ~0xFFF; lo = ~(size-1); /* Size is a power of 2 so this makes a mask */ lo &= ~0xFFF; /* Remove the ctrl value bits */ @@ -35,30 +38,28 @@ static void __cpuinit centaur_mcr_insert(int reg, u32 base, u32 size, int key) } /* - * Figure what we can cover with MCR's + * Figure what we can cover with MCR's * - * Shortcut: We know you can't put 4Gig of RAM on a winchip + * Shortcut: We know you can't put 4Gig of RAM on a winchip */ - -static u32 __cpuinit ramtop(void) /* 16388 */ +static u32 __cpuinit ramtop(void) { - int i; - u32 top = 0; u32 clip = 0xFFFFFFFFUL; - + u32 top = 0; + int i; + for (i = 0; i < e820.nr_map; i++) { unsigned long start, end; if (e820.map[i].addr > 0xFFFFFFFFUL) continue; /* - * Don't MCR over reserved space. Ignore the ISA hole - * we frob around that catastrophe already + * Don't MCR over reserved space. Ignore the ISA hole + * we frob around that catastrophe already */ - - if (e820.map[i].type == E820_RESERVED) - { - if(e820.map[i].addr >= 0x100000UL && e820.map[i].addr < clip) + if (e820.map[i].type == E820_RESERVED) { + if (e820.map[i].addr >= 0x100000UL && + e820.map[i].addr < clip) clip = e820.map[i].addr; continue; } @@ -69,28 +70,27 @@ static u32 __cpuinit ramtop(void) /* 16388 */ if (end > top) top = end; } - /* Everything below 'top' should be RAM except for the ISA hole. - Because of the limited MCR's we want to map NV/ACPI into our - MCR range for gunk in RAM - - Clip might cause us to MCR insufficient RAM but that is an - acceptable failure mode and should only bite obscure boxes with - a VESA hole at 15Mb - - The second case Clip sometimes kicks in is when the EBDA is marked - as reserved. Again we fail safe with reasonable results - */ - - if(top>clip) - top=clip; - + /* + * Everything below 'top' should be RAM except for the ISA hole. + * Because of the limited MCR's we want to map NV/ACPI into our + * MCR range for gunk in RAM + * + * Clip might cause us to MCR insufficient RAM but that is an + * acceptable failure mode and should only bite obscure boxes with + * a VESA hole at 15Mb + * + * The second case Clip sometimes kicks in is when the EBDA is marked + * as reserved. Again we fail safe with reasonable results + */ + if (top > clip) + top = clip; + return top; } /* - * Compute a set of MCR's to give maximum coverage + * Compute a set of MCR's to give maximum coverage */ - static int __cpuinit centaur_mcr_compute(int nr, int key) { u32 mem = ramtop(); @@ -99,141 +99,131 @@ static int __cpuinit centaur_mcr_compute(int nr, int key) u32 top = root; u32 floor = 0; int ct = 0; - - while(ct<nr) - { + + while (ct < nr) { u32 fspace = 0; + u32 high; + u32 low; /* - * Find the largest block we will fill going upwards + * Find the largest block we will fill going upwards */ - - u32 high = power2(mem-top); + high = power2(mem-top); /* - * Find the largest block we will fill going downwards + * Find the largest block we will fill going downwards */ - - u32 low = base/2; + low = base/2; /* - * Don't fill below 1Mb going downwards as there - * is an ISA hole in the way. - */ - - if(base <= 1024*1024) + * Don't fill below 1Mb going downwards as there + * is an ISA hole in the way. + */ + if (base <= 1024*1024) low = 0; - + /* - * See how much space we could cover by filling below - * the ISA hole + * See how much space we could cover by filling below + * the ISA hole */ - - if(floor == 0) + + if (floor == 0) fspace = 512*1024; - else if(floor ==512*1024) + else if (floor == 512*1024) fspace = 128*1024; /* And forget ROM space */ - + /* - * Now install the largest coverage we get + * Now install the largest coverage we get */ - - if(fspace > high && fspace > low) - { + if (fspace > high && fspace > low) { centaur_mcr_insert(ct, floor, fspace, key); floor += fspace; - } - else if(high > low) - { + } else if (high > low) { centaur_mcr_insert(ct, top, high, key); top += high; - } - else if(low > 0) - { + } else if (low > 0) { base -= low; centaur_mcr_insert(ct, base, low, key); - } - else break; + } else + break; ct++; } /* - * We loaded ct values. We now need to set the mask. The caller - * must do this bit. + * We loaded ct values. We now need to set the mask. The caller + * must do this bit. */ - return ct; } static void __cpuinit centaur_create_optimal_mcr(void) { + int used; int i; + /* - * Allocate up to 6 mcrs to mark as much of ram as possible - * as write combining and weak write ordered. + * Allocate up to 6 mcrs to mark as much of ram as possible + * as write combining and weak write ordered. * - * To experiment with: Linux never uses stack operations for - * mmio spaces so we could globally enable stack operation wc + * To experiment with: Linux never uses stack operations for + * mmio spaces so we could globally enable stack operation wc * - * Load the registers with type 31 - full write combining, all - * writes weakly ordered. + * Load the registers with type 31 - full write combining, all + * writes weakly ordered. */ - int used = centaur_mcr_compute(6, 31); + used = centaur_mcr_compute(6, 31); /* - * Wipe unused MCRs + * Wipe unused MCRs */ - - for(i=used;i<8;i++) + for (i = used; i < 8; i++) wrmsr(MSR_IDT_MCR0+i, 0, 0); } static void __cpuinit winchip2_create_optimal_mcr(void) { u32 lo, hi; + int used; int i; /* - * Allocate up to 6 mcrs to mark as much of ram as possible - * as write combining, weak store ordered. + * Allocate up to 6 mcrs to mark as much of ram as possible + * as write combining, weak store ordered. * - * Load the registers with type 25 - * 8 - weak write ordering - * 16 - weak read ordering - * 1 - write combining + * Load the registers with type 25 + * 8 - weak write ordering + * 16 - weak read ordering + * 1 - write combining */ + used = centaur_mcr_compute(6, 25); - int used = centaur_mcr_compute(6, 25); - /* - * Mark the registers we are using. + * Mark the registers we are using. */ - rdmsr(MSR_IDT_MCR_CTRL, lo, hi); - for(i=0;i<used;i++) - lo|=1<<(9+i); + for (i = 0; i < used; i++) + lo |= 1<<(9+i); wrmsr(MSR_IDT_MCR_CTRL, lo, hi); - + /* - * Wipe unused MCRs + * Wipe unused MCRs */ - - for(i=used;i<8;i++) + + for (i = used; i < 8; i++) wrmsr(MSR_IDT_MCR0+i, 0, 0); } /* - * Handle the MCR key on the Winchip 2. + * Handle the MCR key on the Winchip 2. */ - static void __cpuinit winchip2_unprotect_mcr(void) { u32 lo, hi; u32 key; - + rdmsr(MSR_IDT_MCR_CTRL, lo, hi); - lo&=~0x1C0; /* blank bits 8-6 */ + lo &= ~0x1C0; /* blank bits 8-6 */ key = (lo>>17) & 7; lo |= key<<6; /* replace with unlock key */ wrmsr(MSR_IDT_MCR_CTRL, lo, hi); @@ -242,9 +232,9 @@ static void __cpuinit winchip2_unprotect_mcr(void) static void __cpuinit winchip2_protect_mcr(void) { u32 lo, hi; - + rdmsr(MSR_IDT_MCR_CTRL, lo, hi); - lo&=~0x1C0; /* blank bits 8-6 */ + lo &= ~0x1C0; /* blank bits 8-6 */ wrmsr(MSR_IDT_MCR_CTRL, lo, hi); } #endif /* CONFIG_X86_OOSTORE */ @@ -267,17 +257,17 @@ static void __cpuinit init_c3(struct cpuinfo_x86 *c) /* enable ACE unit, if present and disabled */ if ((tmp & (ACE_PRESENT | ACE_ENABLED)) == ACE_PRESENT) { - rdmsr (MSR_VIA_FCR, lo, hi); + rdmsr(MSR_VIA_FCR, lo, hi); lo |= ACE_FCR; /* enable ACE unit */ - wrmsr (MSR_VIA_FCR, lo, hi); + wrmsr(MSR_VIA_FCR, lo, hi); printk(KERN_INFO "CPU: Enabled ACE h/w crypto\n"); } /* enable RNG unit, if present and disabled */ if ((tmp & (RNG_PRESENT | RNG_ENABLED)) == RNG_PRESENT) { - rdmsr (MSR_VIA_RNG, lo, hi); + rdmsr(MSR_VIA_RNG, lo, hi); lo |= RNG_ENABLE; /* enable RNG unit */ - wrmsr (MSR_VIA_RNG, lo, hi); + wrmsr(MSR_VIA_RNG, lo, hi); printk(KERN_INFO "CPU: Enabled h/w RNG\n"); } @@ -288,171 +278,183 @@ static void __cpuinit init_c3(struct cpuinfo_x86 *c) } /* Cyrix III family needs CX8 & PGE explicitly enabled. */ - if (c->x86_model >=6 && c->x86_model <= 9) { - rdmsr (MSR_VIA_FCR, lo, hi); + if (c->x86_model >= 6 && c->x86_model <= 9) { + rdmsr(MSR_VIA_FCR, lo, hi); lo |= (1<<1 | 1<<7); - wrmsr (MSR_VIA_FCR, lo, hi); - set_bit(X86_FEATURE_CX8, c->x86_capability); + wrmsr(MSR_VIA_FCR, lo, hi); + set_cpu_cap(c, X86_FEATURE_CX8); } /* Before Nehemiah, the C3's had 3dNOW! */ - if (c->x86_model >=6 && c->x86_model <9) - set_bit(X86_FEATURE_3DNOW, c->x86_capability); + if (c->x86_model >= 6 && c->x86_model < 9) + set_cpu_cap(c, X86_FEATURE_3DNOW); get_model_name(c); display_cacheinfo(c); } +enum { + ECX8 = 1<<1, + EIERRINT = 1<<2, + DPM = 1<<3, + DMCE = 1<<4, + DSTPCLK = 1<<5, + ELINEAR = 1<<6, + DSMC = 1<<7, + DTLOCK = 1<<8, + EDCTLB = 1<<8, + EMMX = 1<<9, + DPDC = 1<<11, + EBRPRED = 1<<12, + DIC = 1<<13, + DDC = 1<<14, + DNA = 1<<15, + ERETSTK = 1<<16, + E2MMX = 1<<19, + EAMD3D = 1<<20, +}; + static void __cpuinit init_centaur(struct cpuinfo_x86 *c) { - enum { - ECX8=1<<1, - EIERRINT=1<<2, - DPM=1<<3, - DMCE=1<<4, - DSTPCLK=1<<5, - ELINEAR=1<<6, - DSMC=1<<7, - DTLOCK=1<<8, - EDCTLB=1<<8, - EMMX=1<<9, - DPDC=1<<11, - EBRPRED=1<<12, - DIC=1<<13, - DDC=1<<14, - DNA=1<<15, - ERETSTK=1<<16, - E2MMX=1<<19, - EAMD3D=1<<20, - }; char *name; - u32 fcr_set=0; - u32 fcr_clr=0; - u32 lo,hi,newlo; - u32 aa,bb,cc,dd; + u32 fcr_set = 0; + u32 fcr_clr = 0; + u32 lo, hi, newlo; + u32 aa, bb, cc, dd; - /* Bit 31 in normal CPUID used for nonstandard 3DNow ID; - 3DNow is IDd by bit 31 in extended CPUID (1*32+31) anyway */ - clear_bit(0*32+31, c->x86_capability); + /* + * Bit 31 in normal CPUID used for nonstandard 3DNow ID; + * 3DNow is IDd by bit 31 in extended CPUID (1*32+31) anyway + */ + clear_cpu_cap(c, 0*32+31); switch (c->x86) { - - case 5: - switch(c->x86_model) { - case 4: - name="C6"; - fcr_set=ECX8|DSMC|EDCTLB|EMMX|ERETSTK; - fcr_clr=DPDC; - printk(KERN_NOTICE "Disabling bugged TSC.\n"); - clear_bit(X86_FEATURE_TSC, c->x86_capability); + case 5: + switch (c->x86_model) { + case 4: + name = "C6"; + fcr_set = ECX8|DSMC|EDCTLB|EMMX|ERETSTK; + fcr_clr = DPDC; + printk(KERN_NOTICE "Disabling bugged TSC.\n"); + clear_cpu_cap(c, X86_FEATURE_TSC); #ifdef CONFIG_X86_OOSTORE - centaur_create_optimal_mcr(); - /* Enable - write combining on non-stack, non-string - write combining on string, all types - weak write ordering - - The C6 original lacks weak read order - - Note 0x120 is write only on Winchip 1 */ - - wrmsr(MSR_IDT_MCR_CTRL, 0x01F0001F, 0); -#endif + centaur_create_optimal_mcr(); + /* + * Enable: + * write combining on non-stack, non-string + * write combining on string, all types + * weak write ordering + * + * The C6 original lacks weak read order + * + * Note 0x120 is write only on Winchip 1 + */ + wrmsr(MSR_IDT_MCR_CTRL, 0x01F0001F, 0); +#endif + break; + case 8: + switch (c->x86_mask) { + default: + name = "2"; + break; + case 7 ... 9: + name = "2A"; break; - case 8: - switch(c->x86_mask) { - default: - name="2"; - break; - case 7 ... 9: - name="2A"; - break; - case 10 ... 15: - name="2B"; - break; - } - fcr_set=ECX8|DSMC|DTLOCK|EMMX|EBRPRED|ERETSTK|E2MMX|EAMD3D; - fcr_clr=DPDC; + case 10 ... 15: + name = "2B"; + break; + } + fcr_set = ECX8|DSMC|DTLOCK|EMMX|EBRPRED|ERETSTK| + E2MMX|EAMD3D; + fcr_clr = DPDC; #ifdef CONFIG_X86_OOSTORE - winchip2_unprotect_mcr(); - winchip2_create_optimal_mcr(); - rdmsr(MSR_IDT_MCR_CTRL, lo, hi); - /* Enable - write combining on non-stack, non-string - write combining on string, all types - weak write ordering - */ - lo|=31; - wrmsr(MSR_IDT_MCR_CTRL, lo, hi); - winchip2_protect_mcr(); + winchip2_unprotect_mcr(); + winchip2_create_optimal_mcr(); + rdmsr(MSR_IDT_MCR_CTRL, lo, hi); + /* + * Enable: + * write combining on non-stack, non-string + * write combining on string, all types + * weak write ordering + */ + lo |= 31; + wrmsr(MSR_IDT_MCR_CTRL, lo, hi); + winchip2_protect_mcr(); #endif - break; - case 9: - name="3"; - fcr_set=ECX8|DSMC|DTLOCK|EMMX|EBRPRED|ERETSTK|E2MMX|EAMD3D; - fcr_clr=DPDC; + break; + case 9: + name = "3"; + fcr_set = ECX8|DSMC|DTLOCK|EMMX|EBRPRED|ERETSTK| + E2MMX|EAMD3D; + fcr_clr = DPDC; #ifdef CONFIG_X86_OOSTORE - winchip2_unprotect_mcr(); - winchip2_create_optimal_mcr(); - rdmsr(MSR_IDT_MCR_CTRL, lo, hi); - /* Enable - write combining on non-stack, non-string - write combining on string, all types - weak write ordering - */ - lo|=31; - wrmsr(MSR_IDT_MCR_CTRL, lo, hi); - winchip2_protect_mcr(); + winchip2_unprotect_mcr(); + winchip2_create_optimal_mcr(); + rdmsr(MSR_IDT_MCR_CTRL, lo, hi); + /* + * Enable: + * write combining on non-stack, non-string + * write combining on string, all types + * weak write ordering + */ + lo |= 31; + wrmsr(MSR_IDT_MCR_CTRL, lo, hi); + winchip2_protect_mcr(); #endif - break; - default: - name="??"; - } + break; + default: + name = "??"; + } - rdmsr(MSR_IDT_FCR1, lo, hi); - newlo=(lo|fcr_set) & (~fcr_clr); + rdmsr(MSR_IDT_FCR1, lo, hi); + newlo = (lo|fcr_set) & (~fcr_clr); - if (newlo!=lo) { - printk(KERN_INFO "Centaur FCR was 0x%X now 0x%X\n", lo, newlo ); - wrmsr(MSR_IDT_FCR1, newlo, hi ); - } else { - printk(KERN_INFO "Centaur FCR is 0x%X\n",lo); - } - /* Emulate MTRRs using Centaur's MCR. */ - set_bit(X86_FEATURE_CENTAUR_MCR, c->x86_capability); - /* Report CX8 */ - set_bit(X86_FEATURE_CX8, c->x86_capability); - /* Set 3DNow! on Winchip 2 and above. */ - if (c->x86_model >=8) - set_bit(X86_FEATURE_3DNOW, c->x86_capability); - /* See if we can find out some more. */ - if ( cpuid_eax(0x80000000) >= 0x80000005 ) { - /* Yes, we can. */ - cpuid(0x80000005,&aa,&bb,&cc,&dd); - /* Add L1 data and code cache sizes. */ - c->x86_cache_size = (cc>>24)+(dd>>24); - } - sprintf( c->x86_model_id, "WinChip %s", name ); - break; + if (newlo != lo) { + printk(KERN_INFO "Centaur FCR was 0x%X now 0x%X\n", + lo, newlo); + wrmsr(MSR_IDT_FCR1, newlo, hi); + } else { + printk(KERN_INFO "Centaur FCR is 0x%X\n", lo); + } + /* Emulate MTRRs using Centaur's MCR. */ + set_cpu_cap(c, X86_FEATURE_CENTAUR_MCR); + /* Report CX8 */ + set_cpu_cap(c, X86_FEATURE_CX8); + /* Set 3DNow! on Winchip 2 and above. */ + if (c->x86_model >= 8) + set_cpu_cap(c, X86_FEATURE_3DNOW); + /* See if we can find out some more. */ + if (cpuid_eax(0x80000000) >= 0x80000005) { + /* Yes, we can. */ + cpuid(0x80000005, &aa, &bb, &cc, &dd); + /* Add L1 data and code cache sizes. */ + c->x86_cache_size = (cc>>24)+(dd>>24); + } + sprintf(c->x86_model_id, "WinChip %s", name); + break; - case 6: - init_c3(c); - break; + case 6: + init_c3(c); + break; } } -static unsigned int __cpuinit centaur_size_cache(struct cpuinfo_x86 * c, unsigned int size) +static unsigned int __cpuinit +centaur_size_cache(struct cpuinfo_x86 *c, unsigned int size) { /* VIA C3 CPUs (670-68F) need further shifting. */ if ((c->x86 == 6) && ((c->x86_model == 7) || (c->x86_model == 8))) size >>= 8; - /* VIA also screwed up Nehemiah stepping 1, and made - it return '65KB' instead of '64KB' - - Note, it seems this may only be in engineering samples. */ - if ((c->x86==6) && (c->x86_model==9) && (c->x86_mask==1) && (size==65)) - size -=1; + /* + * There's also an erratum in Nehemiah stepping 1, which + * returns '65KB' instead of '64KB' + * - Note, it seems this may only be in engineering samples. + */ + if ((c->x86 == 6) && (c->x86_model == 9) && + (c->x86_mask == 1) && (size == 65)) + size -= 1; return size; } @@ -464,8 +466,4 @@ static struct cpu_dev centaur_cpu_dev __cpuinitdata = { .c_size_cache = centaur_size_cache, }; -int __init centaur_init_cpu(void) -{ - cpu_devs[X86_VENDOR_CENTAUR] = ¢aur_cpu_dev; - return 0; -} +cpu_vendor_dev_register(X86_VENDOR_CENTAUR, ¢aur_cpu_dev); diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c index a38aafaefc23..d999d7833bc2 100644 --- a/arch/x86/kernel/cpu/common.c +++ b/arch/x86/kernel/cpu/common.c @@ -62,9 +62,9 @@ __u32 cleared_cpu_caps[NCAPINTS] __cpuinitdata; static int cachesize_override __cpuinitdata = -1; static int disable_x86_serial_nr __cpuinitdata = 1; -struct cpu_dev * cpu_devs[X86_VENDOR_NUM] = {}; +struct cpu_dev *cpu_devs[X86_VENDOR_NUM] = {}; -static void __cpuinit default_init(struct cpuinfo_x86 * c) +static void __cpuinit default_init(struct cpuinfo_x86 *c) { /* Not much we can do here... */ /* Check if at least it has cpuid */ @@ -81,11 +81,11 @@ static struct cpu_dev __cpuinitdata default_cpu = { .c_init = default_init, .c_vendor = "Unknown", }; -static struct cpu_dev * this_cpu __cpuinitdata = &default_cpu; +static struct cpu_dev *this_cpu __cpuinitdata = &default_cpu; static int __init cachesize_setup(char *str) { - get_option (&str, &cachesize_override); + get_option(&str, &cachesize_override); return 1; } __setup("cachesize=", cachesize_setup); @@ -107,12 +107,12 @@ int __cpuinit get_model_name(struct cpuinfo_x86 *c) /* Intel chips right-justify this string for some dumb reason; undo that brain damage */ p = q = &c->x86_model_id[0]; - while ( *p == ' ' ) + while (*p == ' ') p++; - if ( p != q ) { - while ( *p ) + if (p != q) { + while (*p) *q++ = *p++; - while ( q <= &c->x86_model_id[48] ) + while (q <= &c->x86_model_id[48]) *q++ = '\0'; /* Zero-pad the rest */ } @@ -130,7 +130,7 @@ void __cpuinit display_cacheinfo(struct cpuinfo_x86 *c) cpuid(0x80000005, &dummy, &dummy, &ecx, &edx); printk(KERN_INFO "CPU: L1 I Cache: %dK (%d bytes/line), D cache %dK (%d bytes/line)\n", edx>>24, edx&0xFF, ecx>>24, ecx&0xFF); - c->x86_cache_size=(ecx>>24)+(edx>>24); + c->x86_cache_size = (ecx>>24)+(edx>>24); } if (n < 0x80000006) /* Some chips just has a large L1. */ @@ -138,16 +138,16 @@ void __cpuinit display_cacheinfo(struct cpuinfo_x86 *c) ecx = cpuid_ecx(0x80000006); l2size = ecx >> 16; - + /* do processor-specific cache resizing */ if (this_cpu->c_size_cache) - l2size = this_cpu->c_size_cache(c,l2size); + l2size = this_cpu->c_size_cache(c, l2size); /* Allow user to override all this if necessary. */ if (cachesize_override != -1) l2size = cachesize_override; - if ( l2size == 0 ) + if (l2size == 0) return; /* Again, no L2 cache is possible */ c->x86_cache_size = l2size; @@ -156,16 +156,19 @@ void __cpuinit display_cacheinfo(struct cpuinfo_x86 *c) l2size, ecx & 0xFF); } -/* Naming convention should be: <Name> [(<Codename>)] */ -/* This table only is used unless init_<vendor>() below doesn't set it; */ -/* in particular, if CPUID levels 0x80000002..4 are supported, this isn't used */ +/* + * Naming convention should be: <Name> [(<Codename>)] + * This table only is used unless init_<vendor>() below doesn't set it; + * in particular, if CPUID levels 0x80000002..4 are supported, this isn't used + * + */ /* Look up CPU names by table lookup. */ static char __cpuinit *table_lookup_model(struct cpuinfo_x86 *c) { struct cpu_model_info *info; - if ( c->x86_model >= 16 ) + if (c->x86_model >= 16) return NULL; /* Range check */ if (!this_cpu) @@ -190,9 +193,9 @@ static void __cpuinit get_cpu_vendor(struct cpuinfo_x86 *c, int early) for (i = 0; i < X86_VENDOR_NUM; i++) { if (cpu_devs[i]) { - if (!strcmp(v,cpu_devs[i]->c_ident[0]) || - (cpu_devs[i]->c_ident[1] && - !strcmp(v,cpu_devs[i]->c_ident[1]))) { + if (!strcmp(v, cpu_devs[i]->c_ident[0]) || + (cpu_devs[i]->c_ident[1] && + !strcmp(v, cpu_devs[i]->c_ident[1]))) { c->x86_vendor = i; if (!early) this_cpu = cpu_devs[i]; @@ -210,7 +213,7 @@ static void __cpuinit get_cpu_vendor(struct cpuinfo_x86 *c, int early) } -static int __init x86_fxsr_setup(char * s) +static int __init x86_fxsr_setup(char *s) { setup_clear_cpu_cap(X86_FEATURE_FXSR); setup_clear_cpu_cap(X86_FEATURE_XMM); @@ -219,7 +222,7 @@ static int __init x86_fxsr_setup(char * s) __setup("nofxsr", x86_fxsr_setup); -static int __init x86_sep_setup(char * s) +static int __init x86_sep_setup(char *s) { setup_clear_cpu_cap(X86_FEATURE_SEP); return 1; @@ -306,14 +309,30 @@ static void __cpuinit early_get_cap(struct cpuinfo_x86 *c) } -} + clear_cpu_cap(c, X86_FEATURE_PAT); + + switch (c->x86_vendor) { + case X86_VENDOR_AMD: + if (c->x86 >= 0xf && c->x86 <= 0x11) + set_cpu_cap(c, X86_FEATURE_PAT); + break; + case X86_VENDOR_INTEL: + if (c->x86 == 0xF || (c->x86 == 6 && c->x86_model >= 15)) + set_cpu_cap(c, X86_FEATURE_PAT); + break; + } -/* Do minimum CPU detection early. - Fields really needed: vendor, cpuid_level, family, model, mask, cache alignment. - The others are not touched to avoid unwanted side effects. +} - WARNING: this function is only called on the BP. Don't add code here - that is supposed to run on all CPUs. */ +/* + * Do minimum CPU detection early. + * Fields really needed: vendor, cpuid_level, family, model, mask, + * cache alignment. + * The others are not touched to avoid unwanted side effects. + * + * WARNING: this function is only called on the BP. Don't add code here + * that is supposed to run on all CPUs. + */ static void __init early_cpu_detect(void) { struct cpuinfo_x86 *c = &boot_cpu_data; @@ -328,19 +347,14 @@ static void __init early_cpu_detect(void) get_cpu_vendor(c, 1); - switch (c->x86_vendor) { - case X86_VENDOR_AMD: - early_init_amd(c); - break; - case X86_VENDOR_INTEL: - early_init_intel(c); - break; - } + if (c->x86_vendor != X86_VENDOR_UNKNOWN && + cpu_devs[c->x86_vendor]->c_early_init) + cpu_devs[c->x86_vendor]->c_early_init(c); early_get_cap(c); } -static void __cpuinit generic_identify(struct cpuinfo_x86 * c) +static void __cpuinit generic_identify(struct cpuinfo_x86 *c) { u32 tfms, xlvl; unsigned int ebx; @@ -351,13 +365,12 @@ static void __cpuinit generic_identify(struct cpuinfo_x86 * c) (unsigned int *)&c->x86_vendor_id[0], (unsigned int *)&c->x86_vendor_id[8], (unsigned int *)&c->x86_vendor_id[4]); - + get_cpu_vendor(c, 0); /* Initialize the standard set of capabilities */ /* Note that the vendor-specific code below might override */ - /* Intel-defined flags: level 0x00000001 */ - if ( c->cpuid_level >= 0x00000001 ) { + if (c->cpuid_level >= 0x00000001) { u32 capability, excap; cpuid(0x00000001, &tfms, &ebx, &excap, &capability); c->x86_capability[0] = capability; @@ -369,12 +382,14 @@ static void __cpuinit generic_identify(struct cpuinfo_x86 * c) if (c->x86 >= 0x6) c->x86_model += ((tfms >> 16) & 0xF) << 4; c->x86_mask = tfms & 15; + c->initial_apicid = (ebx >> 24) & 0xFF; #ifdef CONFIG_X86_HT - c->apicid = phys_pkg_id((ebx >> 24) & 0xFF, 0); + c->apicid = phys_pkg_id(c->initial_apicid, 0); + c->phys_proc_id = c->initial_apicid; #else - c->apicid = (ebx >> 24) & 0xFF; + c->apicid = c->initial_apicid; #endif - if (c->x86_capability[0] & (1<<19)) + if (test_cpu_cap(c, X86_FEATURE_CLFLSH)) c->x86_clflush_size = ((ebx >> 8) & 0xff) * 8; } else { /* Have CPUID level 0 only - unheard of */ @@ -383,33 +398,42 @@ static void __cpuinit generic_identify(struct cpuinfo_x86 * c) /* AMD-defined flags: level 0x80000001 */ xlvl = cpuid_eax(0x80000000); - if ( (xlvl & 0xffff0000) == 0x80000000 ) { - if ( xlvl >= 0x80000001 ) { + if ((xlvl & 0xffff0000) == 0x80000000) { + if (xlvl >= 0x80000001) { c->x86_capability[1] = cpuid_edx(0x80000001); c->x86_capability[6] = cpuid_ecx(0x80000001); } - if ( xlvl >= 0x80000004 ) + if (xlvl >= 0x80000004) get_model_name(c); /* Default name */ } init_scattered_cpuid_features(c); } -#ifdef CONFIG_X86_HT - c->phys_proc_id = (cpuid_ebx(1) >> 24) & 0xff; -#endif + clear_cpu_cap(c, X86_FEATURE_PAT); + + switch (c->x86_vendor) { + case X86_VENDOR_AMD: + if (c->x86 >= 0xf && c->x86 <= 0x11) + set_cpu_cap(c, X86_FEATURE_PAT); + break; + case X86_VENDOR_INTEL: + if (c->x86 == 0xF || (c->x86 == 6 && c->x86_model >= 15)) + set_cpu_cap(c, X86_FEATURE_PAT); + break; + } } static void __cpuinit squash_the_stupid_serial_number(struct cpuinfo_x86 *c) { - if (cpu_has(c, X86_FEATURE_PN) && disable_x86_serial_nr ) { + if (cpu_has(c, X86_FEATURE_PN) && disable_x86_serial_nr) { /* Disable processor serial number */ - unsigned long lo,hi; - rdmsr(MSR_IA32_BBL_CR_CTL,lo,hi); + unsigned long lo, hi; + rdmsr(MSR_IA32_BBL_CR_CTL, lo, hi); lo |= 0x200000; - wrmsr(MSR_IA32_BBL_CR_CTL,lo,hi); + wrmsr(MSR_IA32_BBL_CR_CTL, lo, hi); printk(KERN_NOTICE "CPU serial number disabled.\n"); - clear_bit(X86_FEATURE_PN, c->x86_capability); + clear_cpu_cap(c, X86_FEATURE_PN); /* Disabling the serial number may affect the cpuid level */ c->cpuid_level = cpuid_eax(0); @@ -444,9 +468,11 @@ void __cpuinit identify_cpu(struct cpuinfo_x86 *c) memset(&c->x86_capability, 0, sizeof c->x86_capability); if (!have_cpuid_p()) { - /* First of all, decide if this is a 486 or higher */ - /* It's a 486 if we can modify the AC flag */ - if ( flag_is_changeable_p(X86_EFLAGS_AC) ) + /* + * First of all, decide if this is a 486 or higher + * It's a 486 if we can modify the AC flag + */ + if (flag_is_changeable_p(X86_EFLAGS_AC)) c->x86 = 4; else c->x86 = 3; @@ -479,10 +505,10 @@ void __cpuinit identify_cpu(struct cpuinfo_x86 *c) */ /* If the model name is still unset, do table lookup. */ - if ( !c->x86_model_id[0] ) { + if (!c->x86_model_id[0]) { char *p; p = table_lookup_model(c); - if ( p ) + if (p) strcpy(c->x86_model_id, p); else /* Last resort... */ @@ -496,9 +522,9 @@ void __cpuinit identify_cpu(struct cpuinfo_x86 *c) * common between the CPUs. The first time this routine gets * executed, c == &boot_cpu_data. */ - if ( c != &boot_cpu_data ) { + if (c != &boot_cpu_data) { /* AND the already accumulated flags with these */ - for ( i = 0 ; i < NCAPINTS ; i++ ) + for (i = 0 ; i < NCAPINTS ; i++) boot_cpu_data.x86_capability[i] &= c->x86_capability[i]; } @@ -542,7 +568,7 @@ void __cpuinit detect_ht(struct cpuinfo_x86 *c) if (smp_num_siblings == 1) { printk(KERN_INFO "CPU: Hyper-Threading is disabled\n"); - } else if (smp_num_siblings > 1 ) { + } else if (smp_num_siblings > 1) { if (smp_num_siblings > NR_CPUS) { printk(KERN_WARNING "CPU: Unsupported number of the " @@ -552,7 +578,7 @@ void __cpuinit detect_ht(struct cpuinfo_x86 *c) } index_msb = get_count_order(smp_num_siblings); - c->phys_proc_id = phys_pkg_id((ebx >> 24) & 0xFF, index_msb); + c->phys_proc_id = phys_pkg_id(c->initial_apicid, index_msb); printk(KERN_INFO "CPU: Physical Processor ID: %d\n", c->phys_proc_id); @@ -563,7 +589,7 @@ void __cpuinit detect_ht(struct cpuinfo_x86 *c) core_bits = get_count_order(c->x86_max_cores); - c->cpu_core_id = phys_pkg_id((ebx >> 24) & 0xFF, index_msb) & + c->cpu_core_id = phys_pkg_id(c->initial_apicid, index_msb) & ((1 << core_bits) - 1); if (c->x86_max_cores > 1) @@ -597,7 +623,7 @@ void __cpuinit print_cpu_info(struct cpuinfo_x86 *c) else printk("%s", c->x86_model_id); - if (c->x86_mask || c->cpuid_level >= 0) + if (c->x86_mask || c->cpuid_level >= 0) printk(" stepping %02x\n", c->x86_mask); else printk("\n"); @@ -616,23 +642,15 @@ __setup("clearcpuid=", setup_disablecpuid); cpumask_t cpu_initialized __cpuinitdata = CPU_MASK_NONE; -/* This is hacky. :) - * We're emulating future behavior. - * In the future, the cpu-specific init functions will be called implicitly - * via the magic of initcalls. - * They will insert themselves into the cpu_devs structure. - * Then, when cpu_init() is called, we can just iterate over that array. - */ void __init early_cpu_init(void) { - intel_cpu_init(); - cyrix_init_cpu(); - nsc_init_cpu(); - amd_init_cpu(); - centaur_init_cpu(); - transmeta_init_cpu(); - nexgen_init_cpu(); - umc_init_cpu(); + struct cpu_vendor_dev *cvdev; + + for (cvdev = __x86cpuvendor_start ; + cvdev < __x86cpuvendor_end ; + cvdev++) + cpu_devs[cvdev->vendor] = cvdev->cpu_dev; + early_cpu_detect(); } @@ -666,7 +684,7 @@ void __cpuinit cpu_init(void) { int cpu = smp_processor_id(); struct task_struct *curr = current; - struct tss_struct * t = &per_cpu(init_tss, cpu); + struct tss_struct *t = &per_cpu(init_tss, cpu); struct thread_struct *thread = &curr->thread; if (cpu_test_and_set(cpu, cpu_initialized)) { @@ -692,7 +710,7 @@ void __cpuinit cpu_init(void) enter_lazy_tlb(&init_mm, curr); load_sp0(t, thread); - set_tss_desc(cpu,t); + set_tss_desc(cpu, t); load_TR_desc(); load_LDT(&init_mm.context); diff --git a/arch/x86/kernel/cpu/cpu.h b/arch/x86/kernel/cpu/cpu.h index e0b38c33d842..783691b2a738 100644 --- a/arch/x86/kernel/cpu/cpu.h +++ b/arch/x86/kernel/cpu/cpu.h @@ -14,6 +14,7 @@ struct cpu_dev { struct cpu_model_info c_models[4]; + void (*c_early_init)(struct cpuinfo_x86 *c); void (*c_init)(struct cpuinfo_x86 * c); void (*c_identify)(struct cpuinfo_x86 * c); unsigned int (*c_size_cache)(struct cpuinfo_x86 * c, unsigned int size); @@ -21,18 +22,17 @@ struct cpu_dev { extern struct cpu_dev * cpu_devs [X86_VENDOR_NUM]; +struct cpu_vendor_dev { + int vendor; + struct cpu_dev *cpu_dev; +}; + +#define cpu_vendor_dev_register(cpu_vendor_id, cpu_dev) \ + static struct cpu_vendor_dev __cpu_vendor_dev_##cpu_vendor_id __used \ + __attribute__((__section__(".x86cpuvendor.init"))) = \ + { cpu_vendor_id, cpu_dev } + +extern struct cpu_vendor_dev __x86cpuvendor_start[], __x86cpuvendor_end[]; + extern int get_model_name(struct cpuinfo_x86 *c); extern void display_cacheinfo(struct cpuinfo_x86 *c); - -extern void early_init_intel(struct cpuinfo_x86 *c); -extern void early_init_amd(struct cpuinfo_x86 *c); - -/* Specific CPU type init functions */ -int intel_cpu_init(void); -int amd_init_cpu(void); -int cyrix_init_cpu(void); -int nsc_init_cpu(void); -int centaur_init_cpu(void); -int transmeta_init_cpu(void); -int nexgen_init_cpu(void); -int umc_init_cpu(void); diff --git a/arch/x86/kernel/cpu/cyrix.c b/arch/x86/kernel/cpu/cyrix.c index 7139b0262703..3fd7a67bb06a 100644 --- a/arch/x86/kernel/cpu/cyrix.c +++ b/arch/x86/kernel/cpu/cyrix.c @@ -19,7 +19,7 @@ static void __cpuinit do_cyrix_devid(unsigned char *dir0, unsigned char *dir1) { unsigned char ccr2, ccr3; unsigned long flags; - + /* we test for DEVID by checking whether CCR3 is writable */ local_irq_save(flags); ccr3 = getCx86(CX86_CCR3); @@ -37,8 +37,7 @@ static void __cpuinit do_cyrix_devid(unsigned char *dir0, unsigned char *dir1) setCx86(CX86_CCR2, ccr2); *dir0 = 0xfe; } - } - else { + } else { setCx86(CX86_CCR3, ccr3); /* restore CCR3 */ /* read DIR0 and DIR1 CPU registers */ @@ -86,7 +85,7 @@ static char cyrix_model_mult2[] __cpuinitdata = "12233445"; static void __cpuinit check_cx686_slop(struct cpuinfo_x86 *c) { unsigned long flags; - + if (Cx86_dir0_msb == 3) { unsigned char ccr3, ccr5; @@ -132,7 +131,7 @@ static void __cpuinit set_cx86_memwb(void) /* set 'Not Write-through' */ write_cr0(read_cr0() | X86_CR0_NW); /* CCR2 bit 2: lock NW bit and set WT1 */ - setCx86(CX86_CCR2, getCx86(CX86_CCR2) | 0x14 ); + setCx86(CX86_CCR2, getCx86(CX86_CCR2) | 0x14); } static void __cpuinit set_cx86_inc(void) @@ -148,7 +147,7 @@ static void __cpuinit set_cx86_inc(void) setCx86(CX86_PCR1, getCx86(CX86_PCR1) | 0x02); /* PCR0 -- Performance Control */ /* Incrementor Margin 10 */ - setCx86(CX86_PCR0, getCx86(CX86_PCR0) | 0x04); + setCx86(CX86_PCR0, getCx86(CX86_PCR0) | 0x04); setCx86(CX86_CCR3, ccr3); /* disable MAPEN */ } @@ -167,16 +166,16 @@ static void __cpuinit geode_configure(void) ccr3 = getCx86(CX86_CCR3); setCx86(CX86_CCR3, (ccr3 & 0x0f) | 0x10); /* enable MAPEN */ - + /* FPU fast, DTE cache, Mem bypass */ setCx86(CX86_CCR4, getCx86(CX86_CCR4) | 0x38); setCx86(CX86_CCR3, ccr3); /* disable MAPEN */ - + set_cx86_memwb(); - set_cx86_reorder(); + set_cx86_reorder(); set_cx86_inc(); - + local_irq_restore(flags); } @@ -187,14 +186,16 @@ static void __cpuinit init_cyrix(struct cpuinfo_x86 *c) char *buf = c->x86_model_id; const char *p = NULL; - /* Bit 31 in normal CPUID used for nonstandard 3DNow ID; - 3DNow is IDd by bit 31 in extended CPUID (1*32+31) anyway */ - clear_bit(0*32+31, c->x86_capability); + /* + * Bit 31 in normal CPUID used for nonstandard 3DNow ID; + * 3DNow is IDd by bit 31 in extended CPUID (1*32+31) anyway + */ + clear_cpu_cap(c, 0*32+31); /* Cyrix used bit 24 in extended (AMD) CPUID for Cyrix MMX extensions */ - if ( test_bit(1*32+24, c->x86_capability) ) { - clear_bit(1*32+24, c->x86_capability); - set_bit(X86_FEATURE_CXMMX, c->x86_capability); + if (test_cpu_cap(c, 1*32+24)) { + clear_cpu_cap(c, 1*32+24); + set_cpu_cap(c, X86_FEATURE_CXMMX); } do_cyrix_devid(&dir0, &dir1); @@ -213,7 +214,7 @@ static void __cpuinit init_cyrix(struct cpuinfo_x86 *c) * the model, multiplier and stepping. Black magic included, * to make the silicon step/rev numbers match the printed ones. */ - + switch (dir0_msn) { unsigned char tmp; @@ -241,7 +242,7 @@ static void __cpuinit init_cyrix(struct cpuinfo_x86 *c) } else /* 686 */ p = Cx86_cb+1; /* Emulate MTRRs using Cyrix's ARRs. */ - set_bit(X86_FEATURE_CYRIX_ARR, c->x86_capability); + set_cpu_cap(c, X86_FEATURE_CYRIX_ARR); /* 6x86's contain this bug */ c->coma_bug = 1; break; @@ -250,17 +251,18 @@ static void __cpuinit init_cyrix(struct cpuinfo_x86 *c) #ifdef CONFIG_PCI { u32 vendor, device; - /* It isn't really a PCI quirk directly, but the cure is the - same. The MediaGX has deep magic SMM stuff that handles the - SB emulation. It throws away the fifo on disable_dma() which - is wrong and ruins the audio. - - Bug2: VSA1 has a wrap bug so that using maximum sized DMA - causes bad things. According to NatSemi VSA2 has another - bug to do with 'hlt'. I've not seen any boards using VSA2 - and X doesn't seem to support it either so who cares 8). - VSA1 we work around however. - */ + /* + * It isn't really a PCI quirk directly, but the cure is the + * same. The MediaGX has deep magic SMM stuff that handles the + * SB emulation. It throws away the fifo on disable_dma() which + * is wrong and ruins the audio. + * + * Bug2: VSA1 has a wrap bug so that using maximum sized DMA + * causes bad things. According to NatSemi VSA2 has another + * bug to do with 'hlt'. I've not seen any boards using VSA2 + * and X doesn't seem to support it either so who cares 8). + * VSA1 we work around however. + */ printk(KERN_INFO "Working around Cyrix MediaGX virtual DMA bugs.\n"); isa_dma_bridge_buggy = 2; @@ -273,55 +275,51 @@ static void __cpuinit init_cyrix(struct cpuinfo_x86 *c) /* * The 5510/5520 companion chips have a funky PIT. - */ + */ if (vendor == PCI_VENDOR_ID_CYRIX && (device == PCI_DEVICE_ID_CYRIX_5510 || device == PCI_DEVICE_ID_CYRIX_5520)) mark_tsc_unstable("cyrix 5510/5520 detected"); } #endif - c->x86_cache_size=16; /* Yep 16K integrated cache thats it */ + c->x86_cache_size = 16; /* Yep 16K integrated cache thats it */ /* GXm supports extended cpuid levels 'ala' AMD */ if (c->cpuid_level == 2) { /* Enable cxMMX extensions (GX1 Datasheet 54) */ setCx86(CX86_CCR7, getCx86(CX86_CCR7) | 1); - + /* * GXm : 0x30 ... 0x5f GXm datasheet 51 * GXlv: 0x6x GXlv datasheet 54 * ? : 0x7x * GX1 : 0x8x GX1 datasheet 56 */ - if((0x30 <= dir1 && dir1 <= 0x6f) || (0x80 <=dir1 && dir1 <= 0x8f)) + if ((0x30 <= dir1 && dir1 <= 0x6f) || (0x80 <= dir1 && dir1 <= 0x8f)) geode_configure(); get_model_name(c); /* get CPU marketing name */ return; - } - else { /* MediaGX */ + } else { /* MediaGX */ Cx86_cb[2] = (dir0_lsn & 1) ? '3' : '4'; p = Cx86_cb+2; c->x86_model = (dir1 & 0x20) ? 1 : 2; } break; - case 5: /* 6x86MX/M II */ - if (dir1 > 7) - { + case 5: /* 6x86MX/M II */ + if (dir1 > 7) { dir0_msn++; /* M II */ /* Enable MMX extensions (App note 108) */ setCx86(CX86_CCR7, getCx86(CX86_CCR7)|1); - } - else - { + } else { c->coma_bug = 1; /* 6x86MX, it has the bug. */ } tmp = (!(dir0_lsn & 7) || dir0_lsn & 1) ? 2 : 0; Cx86_cb[tmp] = cyrix_model_mult2[dir0_lsn & 7]; p = Cx86_cb+tmp; - if (((dir1 & 0x0f) > 4) || ((dir1 & 0xf0) == 0x20)) + if (((dir1 & 0x0f) > 4) || ((dir1 & 0xf0) == 0x20)) (c->x86_model)++; /* Emulate MTRRs using Cyrix's ARRs. */ - set_bit(X86_FEATURE_CYRIX_ARR, c->x86_capability); + set_cpu_cap(c, X86_FEATURE_CYRIX_ARR); break; case 0xf: /* Cyrix 486 without DEVID registers */ @@ -343,7 +341,8 @@ static void __cpuinit init_cyrix(struct cpuinfo_x86 *c) break; } strcpy(buf, Cx86_model[dir0_msn & 7]); - if (p) strcat(buf, p); + if (p) + strcat(buf, p); return; } @@ -352,7 +351,8 @@ static void __cpuinit init_cyrix(struct cpuinfo_x86 *c) */ static void __cpuinit init_nsc(struct cpuinfo_x86 *c) { - /* There may be GX1 processors in the wild that are branded + /* + * There may be GX1 processors in the wild that are branded * NSC and not Cyrix. * * This function only handles the GX processor, and kicks every @@ -377,7 +377,7 @@ static void __cpuinit init_nsc(struct cpuinfo_x86 *c) * by the fact that they preserve the flags across the division of 5/2. * PII and PPro exhibit this behavior too, but they have cpuid available. */ - + /* * Perform the Cyrix 5/2 test. A Cyrix won't change * the flags, while other 486 chips will. @@ -398,27 +398,26 @@ static inline int test_cyrix_52div(void) return (unsigned char) (test >> 8) == 0x02; } -static void __cpuinit cyrix_identify(struct cpuinfo_x86 * c) +static void __cpuinit cyrix_identify(struct cpuinfo_x86 *c) { /* Detect Cyrix with disabled CPUID */ - if ( c->x86 == 4 && test_cyrix_52div() ) { + if (c->x86 == 4 && test_cyrix_52div()) { unsigned char dir0, dir1; - + strcpy(c->x86_vendor_id, "CyrixInstead"); - c->x86_vendor = X86_VENDOR_CYRIX; - - /* Actually enable cpuid on the older cyrix */ - - /* Retrieve CPU revisions */ - + c->x86_vendor = X86_VENDOR_CYRIX; + + /* Actually enable cpuid on the older cyrix */ + + /* Retrieve CPU revisions */ + do_cyrix_devid(&dir0, &dir1); - dir0>>=4; - + dir0 >>= 4; + /* Check it is an affected model */ - - if (dir0 == 5 || dir0 == 3) - { + + if (dir0 == 5 || dir0 == 3) { unsigned char ccr3; unsigned long flags; printk(KERN_INFO "Enabling CPUID on Cyrix processor.\n"); @@ -434,26 +433,17 @@ static void __cpuinit cyrix_identify(struct cpuinfo_x86 * c) static struct cpu_dev cyrix_cpu_dev __cpuinitdata = { .c_vendor = "Cyrix", - .c_ident = { "CyrixInstead" }, + .c_ident = { "CyrixInstead" }, .c_init = init_cyrix, .c_identify = cyrix_identify, }; -int __init cyrix_init_cpu(void) -{ - cpu_devs[X86_VENDOR_CYRIX] = &cyrix_cpu_dev; - return 0; -} +cpu_vendor_dev_register(X86_VENDOR_CYRIX, &cyrix_cpu_dev); static struct cpu_dev nsc_cpu_dev __cpuinitdata = { .c_vendor = "NSC", - .c_ident = { "Geode by NSC" }, + .c_ident = { "Geode by NSC" }, .c_init = init_nsc, }; -int __init nsc_init_cpu(void) -{ - cpu_devs[X86_VENDOR_NSC] = &nsc_cpu_dev; - return 0; -} - +cpu_vendor_dev_register(X86_VENDOR_NSC, &nsc_cpu_dev); diff --git a/arch/x86/kernel/cpu/feature_names.c b/arch/x86/kernel/cpu/feature_names.c index ee975ac6bbcb..e43ad4ad4cba 100644 --- a/arch/x86/kernel/cpu/feature_names.c +++ b/arch/x86/kernel/cpu/feature_names.c @@ -4,7 +4,7 @@ * This file must not contain any executable code. */ -#include "asm/cpufeature.h" +#include <asm/cpufeature.h> /* * These flag bits must match the definitions in <asm/cpufeature.h>. diff --git a/arch/x86/kernel/cpu/intel.c b/arch/x86/kernel/cpu/intel.c index fae31ce747bd..fe9224c51d37 100644 --- a/arch/x86/kernel/cpu/intel.c +++ b/arch/x86/kernel/cpu/intel.c @@ -30,7 +30,7 @@ struct movsl_mask movsl_mask __read_mostly; #endif -void __cpuinit early_init_intel(struct cpuinfo_x86 *c) +static void __cpuinit early_init_intel(struct cpuinfo_x86 *c) { /* Netburst reports 64 bytes clflush size, but does IO in 128 bytes */ if (c->x86 == 15 && c->x86_cache_alignment == 64) @@ -45,7 +45,7 @@ void __cpuinit early_init_intel(struct cpuinfo_x86 *c) * * This is called before we do cpu ident work */ - + int __cpuinit ppro_with_ram_bug(void) { /* Uses data from early_cpu_detect now */ @@ -58,7 +58,7 @@ int __cpuinit ppro_with_ram_bug(void) } return 0; } - + /* * P4 Xeon errata 037 workaround. @@ -69,7 +69,7 @@ static void __cpuinit Intel_errata_workarounds(struct cpuinfo_x86 *c) unsigned long lo, hi; if ((c->x86 == 15) && (c->x86_model == 1) && (c->x86_mask == 1)) { - rdmsr (MSR_IA32_MISC_ENABLE, lo, hi); + rdmsr(MSR_IA32_MISC_ENABLE, lo, hi); if ((lo & (1<<9)) == 0) { printk (KERN_INFO "CPU: C0 stepping P4 Xeon detected.\n"); printk (KERN_INFO "CPU: Disabling hardware prefetching (Errata 037)\n"); @@ -127,10 +127,10 @@ static void __cpuinit init_intel(struct cpuinfo_x86 *c) */ c->f00f_bug = 0; if (!paravirt_enabled() && c->x86 == 5) { - static int f00f_workaround_enabled = 0; + static int f00f_workaround_enabled; c->f00f_bug = 1; - if ( !f00f_workaround_enabled ) { + if (!f00f_workaround_enabled) { trap_init_f00f_bug(); printk(KERN_NOTICE "Intel Pentium with F0 0F bug - workaround enabled.\n"); f00f_workaround_enabled = 1; @@ -139,20 +139,22 @@ static void __cpuinit init_intel(struct cpuinfo_x86 *c) #endif l2 = init_intel_cacheinfo(c); - if (c->cpuid_level > 9 ) { + if (c->cpuid_level > 9) { unsigned eax = cpuid_eax(10); /* Check for version and the number of counters */ if ((eax & 0xff) && (((eax>>8) & 0xff) > 1)) - set_bit(X86_FEATURE_ARCH_PERFMON, c->x86_capability); + set_cpu_cap(c, X86_FEATURE_ARCH_PERFMON); } /* SEP CPUID bug: Pentium Pro reports SEP but doesn't have it until model 3 mask 3 */ if ((c->x86<<8 | c->x86_model<<4 | c->x86_mask) < 0x633) - clear_bit(X86_FEATURE_SEP, c->x86_capability); + clear_cpu_cap(c, X86_FEATURE_SEP); - /* Names for the Pentium II/Celeron processors - detectable only by also checking the cache size. - Dixon is NOT a Celeron. */ + /* + * Names for the Pentium II/Celeron processors + * detectable only by also checking the cache size. + * Dixon is NOT a Celeron. + */ if (c->x86 == 6) { switch (c->x86_model) { case 5: @@ -163,14 +165,14 @@ static void __cpuinit init_intel(struct cpuinfo_x86 *c) p = "Mobile Pentium II (Dixon)"; } break; - + case 6: if (l2 == 128) p = "Celeron (Mendocino)"; else if (c->x86_mask == 0 || c->x86_mask == 5) p = "Celeron-A"; break; - + case 8: if (l2 == 128) p = "Celeron (Coppermine)"; @@ -178,9 +180,9 @@ static void __cpuinit init_intel(struct cpuinfo_x86 *c) } } - if ( p ) + if (p) strcpy(c->x86_model_id, p); - + c->x86_max_cores = num_cpu_cores(c); detect_ht(c); @@ -207,28 +209,29 @@ static void __cpuinit init_intel(struct cpuinfo_x86 *c) #endif if (cpu_has_xmm2) - set_bit(X86_FEATURE_LFENCE_RDTSC, c->x86_capability); + set_cpu_cap(c, X86_FEATURE_LFENCE_RDTSC); if (c->x86 == 15) { - set_bit(X86_FEATURE_P4, c->x86_capability); + set_cpu_cap(c, X86_FEATURE_P4); } - if (c->x86 == 6) - set_bit(X86_FEATURE_P3, c->x86_capability); + if (c->x86 == 6) + set_cpu_cap(c, X86_FEATURE_P3); if (cpu_has_ds) { unsigned int l1; rdmsr(MSR_IA32_MISC_ENABLE, l1, l2); if (!(l1 & (1<<11))) - set_bit(X86_FEATURE_BTS, c->x86_capability); + set_cpu_cap(c, X86_FEATURE_BTS); if (!(l1 & (1<<12))) - set_bit(X86_FEATURE_PEBS, c->x86_capability); + set_cpu_cap(c, X86_FEATURE_PEBS); } if (cpu_has_bts) ds_init_intel(c); } -static unsigned int __cpuinit intel_size_cache(struct cpuinfo_x86 * c, unsigned int size) +static unsigned int __cpuinit intel_size_cache(struct cpuinfo_x86 *c, unsigned int size) { - /* Intel PIII Tualatin. This comes in two flavours. + /* + * Intel PIII Tualatin. This comes in two flavours. * One has 256kb of cache, the other 512. We have no way * to determine which, so we use a boottime override * for the 512kb model, and assume 256 otherwise. @@ -240,42 +243,42 @@ static unsigned int __cpuinit intel_size_cache(struct cpuinfo_x86 * c, unsigned static struct cpu_dev intel_cpu_dev __cpuinitdata = { .c_vendor = "Intel", - .c_ident = { "GenuineIntel" }, + .c_ident = { "GenuineIntel" }, .c_models = { - { .vendor = X86_VENDOR_INTEL, .family = 4, .model_names = - { - [0] = "486 DX-25/33", - [1] = "486 DX-50", - [2] = "486 SX", - [3] = "486 DX/2", - [4] = "486 SL", - [5] = "486 SX/2", - [7] = "486 DX/2-WB", - [8] = "486 DX/4", + { .vendor = X86_VENDOR_INTEL, .family = 4, .model_names = + { + [0] = "486 DX-25/33", + [1] = "486 DX-50", + [2] = "486 SX", + [3] = "486 DX/2", + [4] = "486 SL", + [5] = "486 SX/2", + [7] = "486 DX/2-WB", + [8] = "486 DX/4", [9] = "486 DX/4-WB" } }, { .vendor = X86_VENDOR_INTEL, .family = 5, .model_names = - { - [0] = "Pentium 60/66 A-step", - [1] = "Pentium 60/66", + { + [0] = "Pentium 60/66 A-step", + [1] = "Pentium 60/66", [2] = "Pentium 75 - 200", - [3] = "OverDrive PODP5V83", + [3] = "OverDrive PODP5V83", [4] = "Pentium MMX", - [7] = "Mobile Pentium 75 - 200", + [7] = "Mobile Pentium 75 - 200", [8] = "Mobile Pentium MMX" } }, { .vendor = X86_VENDOR_INTEL, .family = 6, .model_names = - { + { [0] = "Pentium Pro A-step", - [1] = "Pentium Pro", - [3] = "Pentium II (Klamath)", - [4] = "Pentium II (Deschutes)", - [5] = "Pentium II (Deschutes)", + [1] = "Pentium Pro", + [3] = "Pentium II (Klamath)", + [4] = "Pentium II (Deschutes)", + [5] = "Pentium II (Deschutes)", [6] = "Mobile Pentium II", - [7] = "Pentium III (Katmai)", - [8] = "Pentium III (Coppermine)", + [7] = "Pentium III (Katmai)", + [8] = "Pentium III (Coppermine)", [10] = "Pentium III (Cascades)", [11] = "Pentium III (Tualatin)", } @@ -290,15 +293,12 @@ static struct cpu_dev intel_cpu_dev __cpuinitdata = { } }, }, + .c_early_init = early_init_intel, .c_init = init_intel, .c_size_cache = intel_size_cache, }; -__init int intel_cpu_init(void) -{ - cpu_devs[X86_VENDOR_INTEL] = &intel_cpu_dev; - return 0; -} +cpu_vendor_dev_register(X86_VENDOR_INTEL, &intel_cpu_dev); #ifndef CONFIG_X86_CMPXCHG unsigned long cmpxchg_386_u8(volatile void *ptr, u8 old, u8 new) @@ -364,5 +364,5 @@ unsigned long long cmpxchg_486_u64(volatile void *ptr, u64 old, u64 new) EXPORT_SYMBOL(cmpxchg_486_u64); #endif -// arch_initcall(intel_cpu_init); +/* arch_initcall(intel_cpu_init); */ diff --git a/arch/x86/kernel/cpu/mcheck/mce_32.c b/arch/x86/kernel/cpu/mcheck/mce_32.c index a5182dcd94ae..774d87cfd8cd 100644 --- a/arch/x86/kernel/cpu/mcheck/mce_32.c +++ b/arch/x86/kernel/cpu/mcheck/mce_32.c @@ -10,20 +10,20 @@ #include <linux/smp.h> #include <linux/thread_info.h> -#include <asm/processor.h> +#include <asm/processor.h> #include <asm/system.h> #include <asm/mce.h> #include "mce.h" -int mce_disabled = 0; +int mce_disabled; int nr_mce_banks; EXPORT_SYMBOL_GPL(nr_mce_banks); /* non-fatal.o */ /* Handle unconfigured int18 (should never happen) */ -static void unexpected_machine_check(struct pt_regs * regs, long error_code) -{ +static void unexpected_machine_check(struct pt_regs *regs, long error_code) +{ printk(KERN_ERR "CPU#%d: Unexpected int18 (Machine Check).\n", smp_processor_id()); } @@ -33,30 +33,30 @@ void (*machine_check_vector)(struct pt_regs *, long error_code) = unexpected_mac /* This has to be run for each processor */ void mcheck_init(struct cpuinfo_x86 *c) { - if (mce_disabled==1) + if (mce_disabled == 1) return; switch (c->x86_vendor) { - case X86_VENDOR_AMD: - amd_mcheck_init(c); - break; - - case X86_VENDOR_INTEL: - if (c->x86==5) - intel_p5_mcheck_init(c); - if (c->x86==6) - intel_p6_mcheck_init(c); - if (c->x86==15) - intel_p4_mcheck_init(c); - break; - - case X86_VENDOR_CENTAUR: - if (c->x86==5) - winchip_mcheck_init(c); - break; - - default: - break; + case X86_VENDOR_AMD: + amd_mcheck_init(c); + break; + + case X86_VENDOR_INTEL: + if (c->x86 == 5) + intel_p5_mcheck_init(c); + if (c->x86 == 6) + intel_p6_mcheck_init(c); + if (c->x86 == 15) + intel_p4_mcheck_init(c); + break; + + case X86_VENDOR_CENTAUR: + if (c->x86 == 5) + winchip_mcheck_init(c); + break; + + default: + break; } } diff --git a/arch/x86/kernel/cpu/mcheck/non-fatal.c b/arch/x86/kernel/cpu/mcheck/non-fatal.c index bf39409b3838..00ccb6c14ec2 100644 --- a/arch/x86/kernel/cpu/mcheck/non-fatal.c +++ b/arch/x86/kernel/cpu/mcheck/non-fatal.c @@ -16,7 +16,7 @@ #include <linux/smp.h> #include <linux/module.h> -#include <asm/processor.h> +#include <asm/processor.h> #include <asm/system.h> #include <asm/msr.h> @@ -26,23 +26,26 @@ static int firstbank; #define MCE_RATE 15*HZ /* timer rate is 15s */ -static void mce_checkregs (void *info) +static void mce_checkregs(void *info) { u32 low, high; int i; - for (i=firstbank; i<nr_mce_banks; i++) { - rdmsr (MSR_IA32_MC0_STATUS+i*4, low, high); + for (i = firstbank; i < nr_mce_banks; i++) { + rdmsr(MSR_IA32_MC0_STATUS+i*4, low, high); if (high & (1<<31)) { printk(KERN_INFO "MCE: The hardware reports a non " "fatal, correctable incident occurred on " "CPU %d.\n", smp_processor_id()); - printk (KERN_INFO "Bank %d: %08x%08x\n", i, high, low); + printk(KERN_INFO "Bank %d: %08x%08x\n", i, high, low); - /* Scrub the error so we don't pick it up in MCE_RATE seconds time. */ - wrmsr (MSR_IA32_MC0_STATUS+i*4, 0UL, 0UL); + /* + * Scrub the error so we don't pick it up in MCE_RATE + * seconds time. + */ + wrmsr(MSR_IA32_MC0_STATUS+i*4, 0UL, 0UL); /* Serialize */ wmb(); @@ -55,10 +58,10 @@ static void mce_work_fn(struct work_struct *work); static DECLARE_DELAYED_WORK(mce_work, mce_work_fn); static void mce_work_fn(struct work_struct *work) -{ +{ on_each_cpu(mce_checkregs, NULL, 1, 1); schedule_delayed_work(&mce_work, round_jiffies_relative(MCE_RATE)); -} +} static int __init init_nonfatal_mce_checker(void) { diff --git a/arch/x86/kernel/cpu/mcheck/p5.c b/arch/x86/kernel/cpu/mcheck/p5.c index a18310aaae0c..bfa5817afdda 100644 --- a/arch/x86/kernel/cpu/mcheck/p5.c +++ b/arch/x86/kernel/cpu/mcheck/p5.c @@ -9,20 +9,20 @@ #include <linux/interrupt.h> #include <linux/smp.h> -#include <asm/processor.h> +#include <asm/processor.h> #include <asm/system.h> #include <asm/msr.h> #include "mce.h" /* Machine check handler for Pentium class Intel */ -static void pentium_machine_check(struct pt_regs * regs, long error_code) +static void pentium_machine_check(struct pt_regs *regs, long error_code) { u32 loaddr, hi, lotype; rdmsr(MSR_IA32_P5_MC_ADDR, loaddr, hi); rdmsr(MSR_IA32_P5_MC_TYPE, lotype, hi); printk(KERN_EMERG "CPU#%d: Machine Check Exception: 0x%8X (type 0x%8X).\n", smp_processor_id(), loaddr, lotype); - if(lotype&(1<<5)) + if (lotype&(1<<5)) printk(KERN_EMERG "CPU#%d: Possible thermal failure (CPU on fire ?).\n", smp_processor_id()); add_taint(TAINT_MACHINE_CHECK); } @@ -31,13 +31,13 @@ static void pentium_machine_check(struct pt_regs * regs, long error_code) void intel_p5_mcheck_init(struct cpuinfo_x86 *c) { u32 l, h; - + /*Check for MCE support */ - if( !cpu_has(c, X86_FEATURE_MCE) ) - return; + if (!cpu_has(c, X86_FEATURE_MCE)) + return; /* Default P5 to off as its often misconnected */ - if(mce_disabled != -1) + if (mce_disabled != -1) return; machine_check_vector = pentium_machine_check; wmb(); @@ -47,7 +47,7 @@ void intel_p5_mcheck_init(struct cpuinfo_x86 *c) rdmsr(MSR_IA32_P5_MC_TYPE, l, h); printk(KERN_INFO "Intel old style machine check architecture supported.\n"); - /* Enable MCE */ + /* Enable MCE */ set_in_cr4(X86_CR4_MCE); printk(KERN_INFO "Intel old style machine check reporting enabled on CPU#%d.\n", smp_processor_id()); } diff --git a/arch/x86/kernel/cpu/mcheck/p6.c b/arch/x86/kernel/cpu/mcheck/p6.c index 74342604d30e..62efc9c2b3af 100644 --- a/arch/x86/kernel/cpu/mcheck/p6.c +++ b/arch/x86/kernel/cpu/mcheck/p6.c @@ -9,23 +9,23 @@ #include <linux/interrupt.h> #include <linux/smp.h> -#include <asm/processor.h> +#include <asm/processor.h> #include <asm/system.h> #include <asm/msr.h> #include "mce.h" /* Machine Check Handler For PII/PIII */ -static void intel_machine_check(struct pt_regs * regs, long error_code) +static void intel_machine_check(struct pt_regs *regs, long error_code) { - int recover=1; + int recover = 1; u32 alow, ahigh, high, low; u32 mcgstl, mcgsth; int i; - rdmsr (MSR_IA32_MCG_STATUS, mcgstl, mcgsth); + rdmsr(MSR_IA32_MCG_STATUS, mcgstl, mcgsth); if (mcgstl & (1<<0)) /* Recoverable ? */ - recover=0; + recover = 0; printk(KERN_EMERG "CPU %d: Machine Check Exception: %08x%08x\n", smp_processor_id(), mcgsth, mcgstl); @@ -55,30 +55,30 @@ static void intel_machine_check(struct pt_regs * regs, long error_code) } if (recover & 2) - panic ("CPU context corrupt"); + panic("CPU context corrupt"); if (recover & 1) - panic ("Unable to continue"); + panic("Unable to continue"); - printk (KERN_EMERG "Attempting to continue.\n"); - /* - * Do not clear the MSR_IA32_MCi_STATUS if the error is not + printk(KERN_EMERG "Attempting to continue.\n"); + /* + * Do not clear the MSR_IA32_MCi_STATUS if the error is not * recoverable/continuable.This will allow BIOS to look at the MSRs * for errors if the OS could not log the error. */ - for (i=0; i<nr_mce_banks; i++) { + for (i = 0; i < nr_mce_banks; i++) { unsigned int msr; msr = MSR_IA32_MC0_STATUS+i*4; - rdmsr (msr,low, high); + rdmsr(msr, low, high); if (high & (1<<31)) { /* Clear it */ - wrmsr (msr, 0UL, 0UL); + wrmsr(msr, 0UL, 0UL); /* Serialize */ wmb(); add_taint(TAINT_MACHINE_CHECK); } } mcgstl &= ~(1<<2); - wrmsr (MSR_IA32_MCG_STATUS,mcgstl, mcgsth); + wrmsr(MSR_IA32_MCG_STATUS, mcgstl, mcgsth); } /* Set up machine check reporting for processors with Intel style MCE */ @@ -86,21 +86,21 @@ void intel_p6_mcheck_init(struct cpuinfo_x86 *c) { u32 l, h; int i; - + /* Check for MCE support */ if (!cpu_has(c, X86_FEATURE_MCE)) return; /* Check for PPro style MCA */ - if (!cpu_has(c, X86_FEATURE_MCA)) + if (!cpu_has(c, X86_FEATURE_MCA)) return; /* Ok machine check is available */ machine_check_vector = intel_machine_check; wmb(); - printk (KERN_INFO "Intel machine check architecture supported.\n"); - rdmsr (MSR_IA32_MCG_CAP, l, h); + printk(KERN_INFO "Intel machine check architecture supported.\n"); + rdmsr(MSR_IA32_MCG_CAP, l, h); if (l & (1<<8)) /* Control register present ? */ wrmsr(MSR_IA32_MCG_CTL, 0xffffffff, 0xffffffff); nr_mce_banks = l & 0xff; @@ -110,13 +110,13 @@ void intel_p6_mcheck_init(struct cpuinfo_x86 *c) * - MC0_CTL should not be written * - Status registers on all banks should be cleared on reset */ - for (i=1; i<nr_mce_banks; i++) - wrmsr (MSR_IA32_MC0_CTL+4*i, 0xffffffff, 0xffffffff); + for (i = 1; i < nr_mce_banks; i++) + wrmsr(MSR_IA32_MC0_CTL+4*i, 0xffffffff, 0xffffffff); - for (i=0; i<nr_mce_banks; i++) - wrmsr (MSR_IA32_MC0_STATUS+4*i, 0x0, 0x0); + for (i = 0; i < nr_mce_banks; i++) + wrmsr(MSR_IA32_MC0_STATUS+4*i, 0x0, 0x0); - set_in_cr4 (X86_CR4_MCE); - printk (KERN_INFO "Intel machine check reporting enabled on CPU#%d.\n", + set_in_cr4(X86_CR4_MCE); + printk(KERN_INFO "Intel machine check reporting enabled on CPU#%d.\n", smp_processor_id()); } diff --git a/arch/x86/kernel/cpu/mcheck/winchip.c b/arch/x86/kernel/cpu/mcheck/winchip.c index 3d428d5afc52..f2be3e190c6b 100644 --- a/arch/x86/kernel/cpu/mcheck/winchip.c +++ b/arch/x86/kernel/cpu/mcheck/winchip.c @@ -8,14 +8,14 @@ #include <linux/kernel.h> #include <linux/interrupt.h> -#include <asm/processor.h> +#include <asm/processor.h> #include <asm/system.h> #include <asm/msr.h> #include "mce.h" /* Machine check handler for WinChip C6 */ -static void winchip_machine_check(struct pt_regs * regs, long error_code) +static void winchip_machine_check(struct pt_regs *regs, long error_code) { printk(KERN_EMERG "CPU0: Machine Check Exception.\n"); add_taint(TAINT_MACHINE_CHECK); @@ -28,8 +28,8 @@ void winchip_mcheck_init(struct cpuinfo_x86 *c) machine_check_vector = winchip_machine_check; wmb(); rdmsr(MSR_IDT_FCR1, lo, hi); - lo|= (1<<2); /* Enable EIERRINT (int 18 MCE) */ - lo&= ~(1<<4); /* Enable MCE */ + lo |= (1<<2); /* Enable EIERRINT (int 18 MCE) */ + lo &= ~(1<<4); /* Enable MCE */ wrmsr(MSR_IDT_FCR1, lo, hi); set_in_cr4(X86_CR4_MCE); printk(KERN_INFO "Winchip machine check reporting enabled on CPU#0.\n"); diff --git a/arch/x86/kernel/cpu/mtrr/generic.c b/arch/x86/kernel/cpu/mtrr/generic.c index 3e18db4cefee..353efe4f5017 100644 --- a/arch/x86/kernel/cpu/mtrr/generic.c +++ b/arch/x86/kernel/cpu/mtrr/generic.c @@ -11,6 +11,7 @@ #include <asm/cpufeature.h> #include <asm/processor-flags.h> #include <asm/tlbflush.h> +#include <asm/pat.h> #include "mtrr.h" struct mtrr_state { @@ -35,6 +36,8 @@ static struct fixed_range_block fixed_range_blocks[] = { static unsigned long smp_changes_mask; static struct mtrr_state mtrr_state = {}; +static int mtrr_state_set; +static u64 tom2; #undef MODULE_PARAM_PREFIX #define MODULE_PARAM_PREFIX "mtrr." @@ -42,6 +45,111 @@ static struct mtrr_state mtrr_state = {}; static int mtrr_show; module_param_named(show, mtrr_show, bool, 0); +/* + * Returns the effective MTRR type for the region + * Error returns: + * - 0xFE - when the range is "not entirely covered" by _any_ var range MTRR + * - 0xFF - when MTRR is not enabled + */ +u8 mtrr_type_lookup(u64 start, u64 end) +{ + int i; + u64 base, mask; + u8 prev_match, curr_match; + + if (!mtrr_state_set) + return 0xFF; + + if (!mtrr_state.enabled) + return 0xFF; + + /* Make end inclusive end, instead of exclusive */ + end--; + + /* Look in fixed ranges. Just return the type as per start */ + if (mtrr_state.have_fixed && (start < 0x100000)) { + int idx; + + if (start < 0x80000) { + idx = 0; + idx += (start >> 16); + return mtrr_state.fixed_ranges[idx]; + } else if (start < 0xC0000) { + idx = 1 * 8; + idx += ((start - 0x80000) >> 14); + return mtrr_state.fixed_ranges[idx]; + } else if (start < 0x1000000) { + idx = 3 * 8; + idx += ((start - 0xC0000) >> 12); + return mtrr_state.fixed_ranges[idx]; + } + } + + /* + * Look in variable ranges + * Look of multiple ranges matching this address and pick type + * as per MTRR precedence + */ + if (!mtrr_state.enabled & 2) { + return mtrr_state.def_type; + } + + prev_match = 0xFF; + for (i = 0; i < num_var_ranges; ++i) { + unsigned short start_state, end_state; + + if (!(mtrr_state.var_ranges[i].mask_lo & (1 << 11))) + continue; + + base = (((u64)mtrr_state.var_ranges[i].base_hi) << 32) + + (mtrr_state.var_ranges[i].base_lo & PAGE_MASK); + mask = (((u64)mtrr_state.var_ranges[i].mask_hi) << 32) + + (mtrr_state.var_ranges[i].mask_lo & PAGE_MASK); + + start_state = ((start & mask) == (base & mask)); + end_state = ((end & mask) == (base & mask)); + if (start_state != end_state) + return 0xFE; + + if ((start & mask) != (base & mask)) { + continue; + } + + curr_match = mtrr_state.var_ranges[i].base_lo & 0xff; + if (prev_match == 0xFF) { + prev_match = curr_match; + continue; + } + + if (prev_match == MTRR_TYPE_UNCACHABLE || + curr_match == MTRR_TYPE_UNCACHABLE) { + return MTRR_TYPE_UNCACHABLE; + } + + if ((prev_match == MTRR_TYPE_WRBACK && + curr_match == MTRR_TYPE_WRTHROUGH) || + (prev_match == MTRR_TYPE_WRTHROUGH && + curr_match == MTRR_TYPE_WRBACK)) { + prev_match = MTRR_TYPE_WRTHROUGH; + curr_match = MTRR_TYPE_WRTHROUGH; + } + + if (prev_match != curr_match) { + return MTRR_TYPE_UNCACHABLE; + } + } + + if (tom2) { + if (start >= (1ULL<<32) && (end < tom2)) + return MTRR_TYPE_WRBACK; + } + + if (prev_match != 0xFF) + return prev_match; + + return mtrr_state.def_type; +} + /* Get the MSR pair relating to a var range */ static void get_mtrr_var_range(unsigned int index, struct mtrr_var_range *vr) @@ -79,12 +187,16 @@ static void print_fixed(unsigned base, unsigned step, const mtrr_type*types) base, base + step - 1, mtrr_attrib_to_str(*types)); } +static void prepare_set(void); +static void post_set(void); + /* Grab all of the MTRR state for this CPU into *state */ void __init get_mtrr_state(void) { unsigned int i; struct mtrr_var_range *vrs; unsigned lo, dummy; + unsigned long flags; vrs = mtrr_state.var_ranges; @@ -100,6 +212,15 @@ void __init get_mtrr_state(void) mtrr_state.def_type = (lo & 0xff); mtrr_state.enabled = (lo & 0xc00) >> 10; + if (amd_special_default_mtrr()) { + unsigned lo, hi; + /* TOP_MEM2 */ + rdmsr(MSR_K8_TOP_MEM2, lo, hi); + tom2 = hi; + tom2 <<= 32; + tom2 |= lo; + tom2 &= 0xffffff8000000ULL; + } if (mtrr_show) { int high_width; @@ -130,7 +251,22 @@ void __init get_mtrr_state(void) else printk(KERN_INFO "MTRR %u disabled\n", i); } + if (tom2) { + printk(KERN_INFO "TOM2: %016llx aka %lldM\n", + tom2, tom2>>20); + } } + mtrr_state_set = 1; + + /* PAT setup for BP. We need to go through sync steps here */ + local_irq_save(flags); + prepare_set(); + + pat_init(); + + post_set(); + local_irq_restore(flags); + } /* Some BIOS's are fucked and don't set all MTRRs the same! */ @@ -397,6 +533,9 @@ static void generic_set_all(void) /* Actually set the state */ mask = set_mtrr_state(); + /* also set PAT */ + pat_init(); + post_set(); local_irq_restore(flags); diff --git a/arch/x86/kernel/cpu/mtrr/if.c b/arch/x86/kernel/cpu/mtrr/if.c index 91e150acb46c..1960f1985e5e 100644 --- a/arch/x86/kernel/cpu/mtrr/if.c +++ b/arch/x86/kernel/cpu/mtrr/if.c @@ -424,11 +424,10 @@ static int __init mtrr_if_init(void) return -ENODEV; proc_root_mtrr = - create_proc_entry("mtrr", S_IWUSR | S_IRUGO, &proc_root); - if (proc_root_mtrr) { + proc_create("mtrr", S_IWUSR | S_IRUGO, &proc_root, &mtrr_fops); + + if (proc_root_mtrr) proc_root_mtrr->owner = THIS_MODULE; - proc_root_mtrr->proc_fops = &mtrr_fops; - } return 0; } diff --git a/arch/x86/kernel/cpu/mtrr/main.c b/arch/x86/kernel/cpu/mtrr/main.c index a6450b3ae759..6a1e278d9323 100644 --- a/arch/x86/kernel/cpu/mtrr/main.c +++ b/arch/x86/kernel/cpu/mtrr/main.c @@ -627,7 +627,7 @@ early_param("disable_mtrr_trim", disable_mtrr_trim_setup); #define Tom2Enabled (1U << 21) #define Tom2ForceMemTypeWB (1U << 22) -static __init int amd_special_default_mtrr(void) +int __init amd_special_default_mtrr(void) { u32 l, h; diff --git a/arch/x86/kernel/cpu/mtrr/state.c b/arch/x86/kernel/cpu/mtrr/state.c index 9f8ba923d1c9..7f7e2753685b 100644 --- a/arch/x86/kernel/cpu/mtrr/state.c +++ b/arch/x86/kernel/cpu/mtrr/state.c @@ -19,13 +19,15 @@ void set_mtrr_prepare_save(struct set_mtrr_context *ctxt) if (use_intel() || is_cpu(CYRIX)) { /* Save value of CR4 and clear Page Global Enable (bit 7) */ - if ( cpu_has_pge ) { + if (cpu_has_pge) { ctxt->cr4val = read_cr4(); write_cr4(ctxt->cr4val & ~X86_CR4_PGE); } - /* Disable and flush caches. Note that wbinvd flushes the TLBs as - a side-effect */ + /* + * Disable and flush caches. Note that wbinvd flushes the TLBs + * as a side-effect + */ cr0 = read_cr0() | X86_CR0_CD; wbinvd(); write_cr0(cr0); @@ -42,7 +44,7 @@ void set_mtrr_prepare_save(struct set_mtrr_context *ctxt) void set_mtrr_cache_disable(struct set_mtrr_context *ctxt) { - if (use_intel()) + if (use_intel()) /* Disable MTRRs, and set the default type to uncached */ mtrr_wrmsr(MTRRdefType_MSR, ctxt->deftype_lo & 0xf300UL, ctxt->deftype_hi); @@ -66,12 +68,12 @@ void set_mtrr_done(struct set_mtrr_context *ctxt) else /* Cyrix ARRs - everything else was excluded at the top */ setCx86(CX86_CCR3, ctxt->ccr3); - + /* Enable caches */ write_cr0(read_cr0() & 0xbfffffff); /* Restore value of CR4 */ - if ( cpu_has_pge ) + if (cpu_has_pge) write_cr4(ctxt->cr4val); } /* Re-enable interrupts locally (if enabled previously) */ diff --git a/arch/x86/kernel/cpu/nexgen.c b/arch/x86/kernel/cpu/nexgen.c index 961fbe1a748f..5d5e1c134123 100644 --- a/arch/x86/kernel/cpu/nexgen.c +++ b/arch/x86/kernel/cpu/nexgen.c @@ -9,11 +9,11 @@ * Detect a NexGen CPU running without BIOS hypercode new enough * to have CPUID. (Thanks to Herbert Oppmann) */ - + static int __cpuinit deep_magic_nexgen_probe(void) { int ret; - + __asm__ __volatile__ ( " movw $0x5555, %%ax\n" " xorw %%dx,%%dx\n" @@ -22,22 +22,21 @@ static int __cpuinit deep_magic_nexgen_probe(void) " movl $0, %%eax\n" " jnz 1f\n" " movl $1, %%eax\n" - "1:\n" - : "=a" (ret) : : "cx", "dx" ); + "1:\n" + : "=a" (ret) : : "cx", "dx"); return ret; } -static void __cpuinit init_nexgen(struct cpuinfo_x86 * c) +static void __cpuinit init_nexgen(struct cpuinfo_x86 *c) { c->x86_cache_size = 256; /* A few had 1 MB... */ } -static void __cpuinit nexgen_identify(struct cpuinfo_x86 * c) +static void __cpuinit nexgen_identify(struct cpuinfo_x86 *c) { /* Detect NexGen with old hypercode */ - if ( deep_magic_nexgen_probe() ) { + if (deep_magic_nexgen_probe()) strcpy(c->x86_vendor_id, "NexGenDriven"); - } } static struct cpu_dev nexgen_cpu_dev __cpuinitdata = { diff --git a/arch/x86/kernel/cpu/proc.c b/arch/x86/kernel/cpu/proc.c index af11d31dce0a..0978a4a39418 100644 --- a/arch/x86/kernel/cpu/proc.c +++ b/arch/x86/kernel/cpu/proc.c @@ -8,78 +8,139 @@ /* * Get CPU information for use by the procfs. */ +#ifdef CONFIG_X86_32 +static void show_cpuinfo_core(struct seq_file *m, struct cpuinfo_x86 *c, + unsigned int cpu) +{ +#ifdef CONFIG_X86_HT + if (c->x86_max_cores * smp_num_siblings > 1) { + seq_printf(m, "physical id\t: %d\n", c->phys_proc_id); + seq_printf(m, "siblings\t: %d\n", + cpus_weight(per_cpu(cpu_core_map, cpu))); + seq_printf(m, "core id\t\t: %d\n", c->cpu_core_id); + seq_printf(m, "cpu cores\t: %d\n", c->booted_cores); + seq_printf(m, "apicid\t\t: %d\n", c->apicid); + seq_printf(m, "initial apicid\t: %d\n", c->initial_apicid); + } +#endif +} + +static void show_cpuinfo_misc(struct seq_file *m, struct cpuinfo_x86 *c) +{ + /* + * We use exception 16 if we have hardware math and we've either seen + * it or the CPU claims it is internal + */ + int fpu_exception = c->hard_math && (ignore_fpu_irq || cpu_has_fpu); + seq_printf(m, + "fdiv_bug\t: %s\n" + "hlt_bug\t\t: %s\n" + "f00f_bug\t: %s\n" + "coma_bug\t: %s\n" + "fpu\t\t: %s\n" + "fpu_exception\t: %s\n" + "cpuid level\t: %d\n" + "wp\t\t: %s\n", + c->fdiv_bug ? "yes" : "no", + c->hlt_works_ok ? "no" : "yes", + c->f00f_bug ? "yes" : "no", + c->coma_bug ? "yes" : "no", + c->hard_math ? "yes" : "no", + fpu_exception ? "yes" : "no", + c->cpuid_level, + c->wp_works_ok ? "yes" : "no"); +} +#else +static void show_cpuinfo_core(struct seq_file *m, struct cpuinfo_x86 *c, + unsigned int cpu) +{ +#ifdef CONFIG_SMP + if (c->x86_max_cores * smp_num_siblings > 1) { + seq_printf(m, "physical id\t: %d\n", c->phys_proc_id); + seq_printf(m, "siblings\t: %d\n", + cpus_weight(per_cpu(cpu_core_map, cpu))); + seq_printf(m, "core id\t\t: %d\n", c->cpu_core_id); + seq_printf(m, "cpu cores\t: %d\n", c->booted_cores); + seq_printf(m, "apicid\t\t: %d\n", c->apicid); + seq_printf(m, "initial apicid\t: %d\n", c->initial_apicid); + } +#endif +} + +static void show_cpuinfo_misc(struct seq_file *m, struct cpuinfo_x86 *c) +{ + seq_printf(m, + "fpu\t\t: yes\n" + "fpu_exception\t: yes\n" + "cpuid level\t: %d\n" + "wp\t\t: yes\n", + c->cpuid_level); +} +#endif + static int show_cpuinfo(struct seq_file *m, void *v) { struct cpuinfo_x86 *c = v; - int i, n = 0; - int fpu_exception; + unsigned int cpu = 0; + int i; #ifdef CONFIG_SMP - n = c->cpu_index; + cpu = c->cpu_index; #endif - seq_printf(m, "processor\t: %d\n" - "vendor_id\t: %s\n" - "cpu family\t: %d\n" - "model\t\t: %d\n" - "model name\t: %s\n", - n, - c->x86_vendor_id[0] ? c->x86_vendor_id : "unknown", - c->x86, - c->x86_model, - c->x86_model_id[0] ? c->x86_model_id : "unknown"); + seq_printf(m, "processor\t: %u\n" + "vendor_id\t: %s\n" + "cpu family\t: %d\n" + "model\t\t: %u\n" + "model name\t: %s\n", + cpu, + c->x86_vendor_id[0] ? c->x86_vendor_id : "unknown", + c->x86, + c->x86_model, + c->x86_model_id[0] ? c->x86_model_id : "unknown"); if (c->x86_mask || c->cpuid_level >= 0) seq_printf(m, "stepping\t: %d\n", c->x86_mask); else seq_printf(m, "stepping\t: unknown\n"); - if ( cpu_has(c, X86_FEATURE_TSC) ) { - unsigned int freq = cpufreq_quick_get(n); + if (cpu_has(c, X86_FEATURE_TSC)) { + unsigned int freq = cpufreq_quick_get(cpu); + if (!freq) freq = cpu_khz; seq_printf(m, "cpu MHz\t\t: %u.%03u\n", - freq / 1000, (freq % 1000)); + freq / 1000, (freq % 1000)); } /* Cache size */ if (c->x86_cache_size >= 0) seq_printf(m, "cache size\t: %d KB\n", c->x86_cache_size); -#ifdef CONFIG_X86_HT - if (c->x86_max_cores * smp_num_siblings > 1) { - seq_printf(m, "physical id\t: %d\n", c->phys_proc_id); - seq_printf(m, "siblings\t: %d\n", - cpus_weight(per_cpu(cpu_core_map, n))); - seq_printf(m, "core id\t\t: %d\n", c->cpu_core_id); - seq_printf(m, "cpu cores\t: %d\n", c->booted_cores); - } -#endif - - /* We use exception 16 if we have hardware math and we've either seen it or the CPU claims it is internal */ - fpu_exception = c->hard_math && (ignore_fpu_irq || cpu_has_fpu); - seq_printf(m, "fdiv_bug\t: %s\n" - "hlt_bug\t\t: %s\n" - "f00f_bug\t: %s\n" - "coma_bug\t: %s\n" - "fpu\t\t: %s\n" - "fpu_exception\t: %s\n" - "cpuid level\t: %d\n" - "wp\t\t: %s\n" - "flags\t\t:", - c->fdiv_bug ? "yes" : "no", - c->hlt_works_ok ? "no" : "yes", - c->f00f_bug ? "yes" : "no", - c->coma_bug ? "yes" : "no", - c->hard_math ? "yes" : "no", - fpu_exception ? "yes" : "no", - c->cpuid_level, - c->wp_works_ok ? "yes" : "no"); - - for ( i = 0 ; i < 32*NCAPINTS ; i++ ) - if ( test_bit(i, c->x86_capability) && - x86_cap_flags[i] != NULL ) + + show_cpuinfo_core(m, c, cpu); + show_cpuinfo_misc(m, c); + + seq_printf(m, "flags\t\t:"); + for (i = 0; i < 32*NCAPINTS; i++) + if (cpu_has(c, i) && x86_cap_flags[i] != NULL) seq_printf(m, " %s", x86_cap_flags[i]); - for (i = 0; i < 32; i++) + seq_printf(m, "\nbogomips\t: %lu.%02lu\n", + c->loops_per_jiffy/(500000/HZ), + (c->loops_per_jiffy/(5000/HZ)) % 100); + +#ifdef CONFIG_X86_64 + if (c->x86_tlbsize > 0) + seq_printf(m, "TLB size\t: %d 4K pages\n", c->x86_tlbsize); +#endif + seq_printf(m, "clflush size\t: %u\n", c->x86_clflush_size); +#ifdef CONFIG_X86_64 + seq_printf(m, "cache_alignment\t: %d\n", c->x86_cache_alignment); + seq_printf(m, "address sizes\t: %u bits physical, %u bits virtual\n", + c->x86_phys_bits, c->x86_virt_bits); +#endif + + seq_printf(m, "power management:"); + for (i = 0; i < 32; i++) { if (c->x86_power & (1 << i)) { if (i < ARRAY_SIZE(x86_power_flags) && x86_power_flags[i]) @@ -89,11 +150,9 @@ static int show_cpuinfo(struct seq_file *m, void *v) else seq_printf(m, " [%d]", i); } + } - seq_printf(m, "\nbogomips\t: %lu.%02lu\n", - c->loops_per_jiffy/(500000/HZ), - (c->loops_per_jiffy/(5000/HZ)) % 100); - seq_printf(m, "clflush size\t: %u\n\n", c->x86_clflush_size); + seq_printf(m, "\n\n"); return 0; } @@ -106,14 +165,17 @@ static void *c_start(struct seq_file *m, loff_t *pos) return &cpu_data(*pos); return NULL; } + static void *c_next(struct seq_file *m, void *v, loff_t *pos) { *pos = next_cpu(*pos, cpu_online_map); return c_start(m, pos); } + static void c_stop(struct seq_file *m, void *v) { } + const struct seq_operations cpuinfo_op = { .start = c_start, .next = c_next, diff --git a/arch/x86/kernel/cpu/transmeta.c b/arch/x86/kernel/cpu/transmeta.c index e8b422c1c512..b911a2c61b8f 100644 --- a/arch/x86/kernel/cpu/transmeta.c +++ b/arch/x86/kernel/cpu/transmeta.c @@ -18,8 +18,8 @@ static void __cpuinit init_transmeta(struct cpuinfo_x86 *c) /* Print CMS and CPU revision */ max = cpuid_eax(0x80860000); cpu_rev = 0; - if ( max >= 0x80860001 ) { - cpuid(0x80860001, &dummy, &cpu_rev, &cpu_freq, &cpu_flags); + if (max >= 0x80860001) { + cpuid(0x80860001, &dummy, &cpu_rev, &cpu_freq, &cpu_flags); if (cpu_rev != 0x02000000) { printk(KERN_INFO "CPU: Processor revision %u.%u.%u.%u, %u MHz\n", (cpu_rev >> 24) & 0xff, @@ -29,7 +29,7 @@ static void __cpuinit init_transmeta(struct cpuinfo_x86 *c) cpu_freq); } } - if ( max >= 0x80860002 ) { + if (max >= 0x80860002) { cpuid(0x80860002, &new_cpu_rev, &cms_rev1, &cms_rev2, &dummy); if (cpu_rev == 0x02000000) { printk(KERN_INFO "CPU: Processor revision %08X, %u MHz\n", @@ -42,7 +42,7 @@ static void __cpuinit init_transmeta(struct cpuinfo_x86 *c) cms_rev1 & 0xff, cms_rev2); } - if ( max >= 0x80860006 ) { + if (max >= 0x80860006) { cpuid(0x80860003, (void *)&cpu_info[0], (void *)&cpu_info[4], @@ -74,23 +74,25 @@ static void __cpuinit init_transmeta(struct cpuinfo_x86 *c) wrmsr(0x80860004, cap_mask, uk); /* All Transmeta CPUs have a constant TSC */ - set_bit(X86_FEATURE_CONSTANT_TSC, c->x86_capability); - + set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC); + #ifdef CONFIG_SYSCTL - /* randomize_va_space slows us down enormously; - it probably triggers retranslation of x86->native bytecode */ + /* + * randomize_va_space slows us down enormously; + * it probably triggers retranslation of x86->native bytecode + */ randomize_va_space = 0; #endif } -static void __cpuinit transmeta_identify(struct cpuinfo_x86 * c) +static void __cpuinit transmeta_identify(struct cpuinfo_x86 *c) { u32 xlvl; /* Transmeta-defined flags: level 0x80860001 */ xlvl = cpuid_eax(0x80860000); - if ( (xlvl & 0xffff0000) == 0x80860000 ) { - if ( xlvl >= 0x80860001 ) + if ((xlvl & 0xffff0000) == 0x80860000) { + if (xlvl >= 0x80860001) c->x86_capability[2] = cpuid_edx(0x80860001); } } @@ -102,8 +104,4 @@ static struct cpu_dev transmeta_cpu_dev __cpuinitdata = { .c_identify = transmeta_identify, }; -int __init transmeta_init_cpu(void) -{ - cpu_devs[X86_VENDOR_TRANSMETA] = &transmeta_cpu_dev; - return 0; -} +cpu_vendor_dev_register(X86_VENDOR_TRANSMETA, &transmeta_cpu_dev); diff --git a/arch/x86/kernel/cpu/umc.c b/arch/x86/kernel/cpu/umc.c index a7a4e75bdcd7..b1fc90989d75 100644 --- a/arch/x86/kernel/cpu/umc.c +++ b/arch/x86/kernel/cpu/umc.c @@ -3,24 +3,23 @@ #include <asm/processor.h> #include "cpu.h" -/* UMC chips appear to be only either 386 or 486, so no special init takes place. +/* + * UMC chips appear to be only either 386 or 486, + * so no special init takes place. */ static struct cpu_dev umc_cpu_dev __cpuinitdata = { .c_vendor = "UMC", - .c_ident = { "UMC UMC UMC" }, + .c_ident = { "UMC UMC UMC" }, .c_models = { { .vendor = X86_VENDOR_UMC, .family = 4, .model_names = - { - [1] = "U5D", - [2] = "U5S", + { + [1] = "U5D", + [2] = "U5S", } }, }, }; -int __init umc_init_cpu(void) -{ - cpu_devs[X86_VENDOR_UMC] = &umc_cpu_dev; - return 0; -} +cpu_vendor_dev_register(X86_VENDOR_UMC, &umc_cpu_dev); + diff --git a/arch/x86/kernel/crash.c b/arch/x86/kernel/crash.c index 9a5fa0abfcc7..2251d0ae9570 100644 --- a/arch/x86/kernel/crash.c +++ b/arch/x86/kernel/crash.c @@ -26,11 +26,7 @@ #include <linux/kdebug.h> #include <asm/smp.h> -#ifdef CONFIG_X86_32 #include <mach_ipi.h> -#else -#include <asm/mach_apic.h> -#endif /* This keeps a track of which one is crashing cpu. */ static int crashing_cpu; diff --git a/arch/x86/kernel/ds.c b/arch/x86/kernel/ds.c index dcd918c1580d..11c11b8ec48d 100644 --- a/arch/x86/kernel/ds.c +++ b/arch/x86/kernel/ds.c @@ -220,11 +220,11 @@ int ds_allocate(void **dsp, size_t bts_size_in_bytes) int ds_free(void **dsp) { - if (*dsp) + if (*dsp) { kfree((void *)get_bts_buffer_base(*dsp)); - kfree(*dsp); - *dsp = NULL; - + kfree(*dsp); + *dsp = NULL; + } return 0; } diff --git a/arch/x86/kernel/e820_32.c b/arch/x86/kernel/e820_32.c index 80444c5c9b14..0240cd778365 100644 --- a/arch/x86/kernel/e820_32.c +++ b/arch/x86/kernel/e820_32.c @@ -450,38 +450,25 @@ int __init sanitize_e820_map(struct e820entry * biosmap, char * pnr_map) * thinkpad 560x, for example, does not cooperate with the memory * detection code.) */ -int __init copy_e820_map(struct e820entry * biosmap, int nr_map) +int __init copy_e820_map(struct e820entry *biosmap, int nr_map) { /* Only one memory region (or negative)? Ignore it */ if (nr_map < 2) return -1; do { - unsigned long long start = biosmap->addr; - unsigned long long size = biosmap->size; - unsigned long long end = start + size; - unsigned long type = biosmap->type; + u64 start = biosmap->addr; + u64 size = biosmap->size; + u64 end = start + size; + u32 type = biosmap->type; /* Overflow in 64 bits? Ignore the memory map. */ if (start > end) return -1; - /* - * Some BIOSes claim RAM in the 640k - 1M region. - * Not right. Fix it up. - */ - if (type == E820_RAM) { - if (start < 0x100000ULL && end > 0xA0000ULL) { - if (start < 0xA0000ULL) - add_memory_region(start, 0xA0000ULL-start, type); - if (end <= 0x100000ULL) - continue; - start = 0x100000ULL; - size = end - start; - } - } add_memory_region(start, size, type); - } while (biosmap++,--nr_map); + } while (biosmap++, --nr_map); + return 0; } diff --git a/arch/x86/kernel/e820_64.c b/arch/x86/kernel/e820_64.c index 9be697126013..7f6c0c85c8f6 100644 --- a/arch/x86/kernel/e820_64.c +++ b/arch/x86/kernel/e820_64.c @@ -27,6 +27,7 @@ #include <asm/setup.h> #include <asm/sections.h> #include <asm/kdebug.h> +#include <asm/trampoline.h> struct e820map e820; @@ -36,11 +37,11 @@ struct e820map e820; unsigned long end_pfn; /* - * end_pfn only includes RAM, while end_pfn_map includes all e820 entries. - * The direct mapping extends to end_pfn_map, so that we can directly access + * end_pfn only includes RAM, while max_pfn_mapped includes all e820 entries. + * The direct mapping extends to max_pfn_mapped, so that we can directly access * apertures, ACPI and other tables without having to play with fixmaps. */ -unsigned long end_pfn_map; +unsigned long max_pfn_mapped; /* * Last pfn which the user wants to use. @@ -58,8 +59,8 @@ struct early_res { }; static struct early_res early_res[MAX_EARLY_RES] __initdata = { { 0, PAGE_SIZE, "BIOS data page" }, /* BIOS data page */ -#ifdef CONFIG_SMP - { SMP_TRAMPOLINE_BASE, SMP_TRAMPOLINE_BASE + 2*PAGE_SIZE, "SMP_TRAMPOLINE" }, +#ifdef CONFIG_X86_TRAMPOLINE + { TRAMPOLINE_BASE, TRAMPOLINE_BASE + 2 * PAGE_SIZE, "TRAMPOLINE" }, #endif {} }; @@ -95,7 +96,8 @@ void __init early_res_to_bootmem(void) } /* Check for already reserved areas */ -static inline int bad_addr(unsigned long *addrp, unsigned long size) +static inline int +bad_addr(unsigned long *addrp, unsigned long size, unsigned long align) { int i; unsigned long addr = *addrp, last; @@ -105,7 +107,7 @@ again: for (i = 0; i < MAX_EARLY_RES && early_res[i].end; i++) { struct early_res *r = &early_res[i]; if (last >= r->start && addr < r->end) { - *addrp = addr = r->end; + *addrp = addr = round_up(r->end, align); changed = 1; goto again; } @@ -113,6 +115,40 @@ again: return changed; } +/* Check for already reserved areas */ +static inline int +bad_addr_size(unsigned long *addrp, unsigned long *sizep, unsigned long align) +{ + int i; + unsigned long addr = *addrp, last; + unsigned long size = *sizep; + int changed = 0; +again: + last = addr + size; + for (i = 0; i < MAX_EARLY_RES && early_res[i].end; i++) { + struct early_res *r = &early_res[i]; + if (last > r->start && addr < r->start) { + size = r->start - addr; + changed = 1; + goto again; + } + if (last > r->end && addr < r->end) { + addr = round_up(r->end, align); + size = last - addr; + changed = 1; + goto again; + } + if (last <= r->end && addr >= r->start) { + (*sizep)++; + return 0; + } + } + if (changed) { + *addrp = addr; + *sizep = size; + } + return changed; +} /* * This function checks if any part of the range <start,end> is mapped * with type. @@ -174,26 +210,27 @@ int __init e820_all_mapped(unsigned long start, unsigned long end, * Find a free area with specified alignment in a specific range. */ unsigned long __init find_e820_area(unsigned long start, unsigned long end, - unsigned size, unsigned long align) + unsigned long size, unsigned long align) { int i; - unsigned long mask = ~(align - 1); for (i = 0; i < e820.nr_map; i++) { struct e820entry *ei = &e820.map[i]; - unsigned long addr = ei->addr, last; + unsigned long addr, last; + unsigned long ei_last; if (ei->type != E820_RAM) continue; + addr = round_up(ei->addr, align); + ei_last = ei->addr + ei->size; if (addr < start) - addr = start; - if (addr > ei->addr + ei->size) + addr = round_up(start, align); + if (addr >= ei_last) continue; - while (bad_addr(&addr, size) && addr+size <= ei->addr+ei->size) + while (bad_addr(&addr, size, align) && addr+size <= ei_last) ; - addr = (addr + align - 1) & mask; last = addr + size; - if (last > ei->addr + ei->size) + if (last > ei_last) continue; if (last > end) continue; @@ -203,6 +240,40 @@ unsigned long __init find_e820_area(unsigned long start, unsigned long end, } /* + * Find next free range after *start + */ +unsigned long __init find_e820_area_size(unsigned long start, + unsigned long *sizep, + unsigned long align) +{ + int i; + + for (i = 0; i < e820.nr_map; i++) { + struct e820entry *ei = &e820.map[i]; + unsigned long addr, last; + unsigned long ei_last; + + if (ei->type != E820_RAM) + continue; + addr = round_up(ei->addr, align); + ei_last = ei->addr + ei->size; + if (addr < start) + addr = round_up(start, align); + if (addr >= ei_last) + continue; + *sizep = ei_last - addr; + while (bad_addr_size(&addr, sizep, align) && + addr + *sizep <= ei_last) + ; + last = addr + *sizep; + if (last > ei_last) + continue; + return addr; + } + return -1UL; + +} +/* * Find the highest page frame number we have available */ unsigned long __init e820_end_of_ram(void) @@ -211,29 +282,29 @@ unsigned long __init e820_end_of_ram(void) end_pfn = find_max_pfn_with_active_regions(); - if (end_pfn > end_pfn_map) - end_pfn_map = end_pfn; - if (end_pfn_map > MAXMEM>>PAGE_SHIFT) - end_pfn_map = MAXMEM>>PAGE_SHIFT; + if (end_pfn > max_pfn_mapped) + max_pfn_mapped = end_pfn; + if (max_pfn_mapped > MAXMEM>>PAGE_SHIFT) + max_pfn_mapped = MAXMEM>>PAGE_SHIFT; if (end_pfn > end_user_pfn) end_pfn = end_user_pfn; - if (end_pfn > end_pfn_map) - end_pfn = end_pfn_map; + if (end_pfn > max_pfn_mapped) + end_pfn = max_pfn_mapped; - printk(KERN_INFO "end_pfn_map = %lu\n", end_pfn_map); + printk(KERN_INFO "max_pfn_mapped = %lu\n", max_pfn_mapped); return end_pfn; } /* * Mark e820 reserved areas as busy for the resource manager. */ -void __init e820_reserve_resources(struct resource *code_resource, - struct resource *data_resource, struct resource *bss_resource) +void __init e820_reserve_resources(void) { int i; + struct resource *res; + + res = alloc_bootmem_low(sizeof(struct resource) * e820.nr_map); for (i = 0; i < e820.nr_map; i++) { - struct resource *res; - res = alloc_bootmem_low(sizeof(struct resource)); switch (e820.map[i].type) { case E820_RAM: res->name = "System RAM"; break; case E820_ACPI: res->name = "ACPI Tables"; break; @@ -243,21 +314,8 @@ void __init e820_reserve_resources(struct resource *code_resource, res->start = e820.map[i].addr; res->end = res->start + e820.map[i].size - 1; res->flags = IORESOURCE_MEM | IORESOURCE_BUSY; - request_resource(&iomem_resource, res); - if (e820.map[i].type == E820_RAM) { - /* - * We don't know which RAM region contains kernel data, - * so we try it repeatedly and let the resource manager - * test it. - */ - request_resource(res, code_resource); - request_resource(res, data_resource); - request_resource(res, bss_resource); -#ifdef CONFIG_KEXEC - if (crashk_res.start != crashk_res.end) - request_resource(res, &crashk_res); -#endif - } + insert_resource(&iomem_resource, res); + res++; } } @@ -309,9 +367,9 @@ static int __init e820_find_active_region(const struct e820entry *ei, if (*ei_startpfn >= *ei_endpfn) return 0; - /* Check if end_pfn_map should be updated */ - if (ei->type != E820_RAM && *ei_endpfn > end_pfn_map) - end_pfn_map = *ei_endpfn; + /* Check if max_pfn_mapped should be updated */ + if (ei->type != E820_RAM && *ei_endpfn > max_pfn_mapped) + max_pfn_mapped = *ei_endpfn; /* Skip if map is outside the node */ if (ei->type != E820_RAM || *ei_endpfn <= start_pfn || @@ -634,10 +692,10 @@ static int __init copy_e820_map(struct e820entry *biosmap, int nr_map) return -1; do { - unsigned long start = biosmap->addr; - unsigned long size = biosmap->size; - unsigned long end = start + size; - unsigned long type = biosmap->type; + u64 start = biosmap->addr; + u64 size = biosmap->size; + u64 end = start + size; + u32 type = biosmap->type; /* Overflow in 64 bits? Ignore the memory map. */ if (start > end) @@ -702,7 +760,7 @@ static int __init parse_memmap_opt(char *p) saved_max_pfn = e820_end_of_ram(); remove_all_active_ranges(); #endif - end_pfn_map = 0; + max_pfn_mapped = 0; e820.nr_map = 0; userdef = 1; return 0; diff --git a/arch/x86/kernel/early_printk.c b/arch/x86/kernel/early_printk.c index cff84cd9987f..643fd861b724 100644 --- a/arch/x86/kernel/early_printk.c +++ b/arch/x86/kernel/early_printk.c @@ -13,7 +13,7 @@ #define VGABASE (__ISA_IO_base + 0xb8000) static int max_ypos = 25, max_xpos = 80; -static int current_ypos = 25, current_xpos = 0; +static int current_ypos = 25, current_xpos; static void early_vga_write(struct console *con, const char *str, unsigned n) { @@ -108,12 +108,12 @@ static __init void early_serial_init(char *s) if (*s) { unsigned port; - if (!strncmp(s,"0x",2)) { + if (!strncmp(s, "0x", 2)) { early_serial_base = simple_strtoul(s, &e, 16); } else { static int bases[] = { 0x3f8, 0x2f8 }; - if (!strncmp(s,"ttyS",4)) + if (!strncmp(s, "ttyS", 4)) s += 4; port = simple_strtoul(s, &e, 10); if (port > 1 || s == e) @@ -194,7 +194,7 @@ static struct console simnow_console = { /* Direct interface for emergencies */ static struct console *early_console = &early_vga_console; -static int early_console_initialized = 0; +static int early_console_initialized; void early_printk(const char *fmt, ...) { @@ -202,9 +202,9 @@ void early_printk(const char *fmt, ...) int n; va_list ap; - va_start(ap,fmt); - n = vscnprintf(buf,512,fmt,ap); - early_console->write(early_console,buf,n); + va_start(ap, fmt); + n = vscnprintf(buf, 512, fmt, ap); + early_console->write(early_console, buf, n); va_end(ap); } @@ -229,15 +229,15 @@ static int __init setup_early_printk(char *buf) early_serial_init(buf); early_console = &early_serial_console; } else if (!strncmp(buf, "vga", 3) - && boot_params.screen_info.orig_video_isVGA == 1) { + && boot_params.screen_info.orig_video_isVGA == 1) { max_xpos = boot_params.screen_info.orig_video_cols; max_ypos = boot_params.screen_info.orig_video_lines; current_ypos = boot_params.screen_info.orig_y; early_console = &early_vga_console; - } else if (!strncmp(buf, "simnow", 6)) { - simnow_init(buf + 6); - early_console = &simnow_console; - keep_early = 1; + } else if (!strncmp(buf, "simnow", 6)) { + simnow_init(buf + 6); + early_console = &simnow_console; + keep_early = 1; #ifdef CONFIG_HVC_XEN } else if (!strncmp(buf, "xen", 3)) { early_console = &xenboot_console; diff --git a/arch/x86/kernel/entry_32.S b/arch/x86/kernel/entry_32.S index 4b87c32b639f..9ba49a26dff8 100644 --- a/arch/x86/kernel/entry_32.S +++ b/arch/x86/kernel/entry_32.S @@ -51,6 +51,7 @@ #include <asm/desc.h> #include <asm/percpu.h> #include <asm/dwarf2.h> +#include <asm/processor-flags.h> #include "irq_vectors.h" /* @@ -68,13 +69,6 @@ #define nr_syscalls ((syscall_table_size)/4) -CF_MASK = 0x00000001 -TF_MASK = 0x00000100 -IF_MASK = 0x00000200 -DF_MASK = 0x00000400 -NT_MASK = 0x00004000 -VM_MASK = 0x00020000 - #ifdef CONFIG_PREEMPT #define preempt_stop(clobbers) DISABLE_INTERRUPTS(clobbers); TRACE_IRQS_OFF #else @@ -84,7 +78,7 @@ VM_MASK = 0x00020000 .macro TRACE_IRQS_IRET #ifdef CONFIG_TRACE_IRQFLAGS - testl $IF_MASK,PT_EFLAGS(%esp) # interrupts off? + testl $X86_EFLAGS_IF,PT_EFLAGS(%esp) # interrupts off? jz 1f TRACE_IRQS_ON 1: @@ -246,7 +240,7 @@ ret_from_intr: check_userspace: movl PT_EFLAGS(%esp), %eax # mix EFLAGS and CS movb PT_CS(%esp), %al - andl $(VM_MASK | SEGMENT_RPL_MASK), %eax + andl $(X86_EFLAGS_VM | SEGMENT_RPL_MASK), %eax cmpl $USER_RPL, %eax jb resume_kernel # not returning to v8086 or userspace @@ -271,7 +265,7 @@ need_resched: movl TI_flags(%ebp), %ecx # need_resched set ? testb $_TIF_NEED_RESCHED, %cl jz restore_all - testl $IF_MASK,PT_EFLAGS(%esp) # interrupts off (exception path) ? + testl $X86_EFLAGS_IF,PT_EFLAGS(%esp) # interrupts off (exception path) ? jz restore_all call preempt_schedule_irq jmp need_resched @@ -291,10 +285,10 @@ ENTRY(ia32_sysenter_target) movl TSS_sysenter_sp0(%esp),%esp sysenter_past_esp: /* - * No need to follow this irqs on/off section: the syscall - * disabled irqs and here we enable it straight after entry: + * Interrupts are disabled here, but we can't trace it until + * enough kernel state to call TRACE_IRQS_OFF can be called - but + * we immediately enable interrupts at that point anyway. */ - ENABLE_INTERRUPTS(CLBR_NONE) pushl $(__USER_DS) CFI_ADJUST_CFA_OFFSET 4 /*CFI_REL_OFFSET ss, 0*/ @@ -302,6 +296,7 @@ sysenter_past_esp: CFI_ADJUST_CFA_OFFSET 4 CFI_REL_OFFSET esp, 0 pushfl + orl $X86_EFLAGS_IF, (%esp) CFI_ADJUST_CFA_OFFSET 4 pushl $(__USER_CS) CFI_ADJUST_CFA_OFFSET 4 @@ -315,6 +310,11 @@ sysenter_past_esp: CFI_ADJUST_CFA_OFFSET 4 CFI_REL_OFFSET eip, 0 + pushl %eax + CFI_ADJUST_CFA_OFFSET 4 + SAVE_ALL + ENABLE_INTERRUPTS(CLBR_NONE) + /* * Load the potential sixth argument from user stack. * Careful about security. @@ -322,14 +322,12 @@ sysenter_past_esp: cmpl $__PAGE_OFFSET-3,%ebp jae syscall_fault 1: movl (%ebp),%ebp + movl %ebp,PT_EBP(%esp) .section __ex_table,"a" .align 4 .long 1b,syscall_fault .previous - pushl %eax - CFI_ADJUST_CFA_OFFSET 4 - SAVE_ALL GET_THREAD_INFO(%ebp) /* Note, _TIF_SECCOMP is bit number 8, and so it needs testw and not testb */ @@ -384,7 +382,7 @@ syscall_exit: # setting need_resched or sigpending # between sampling and the iret TRACE_IRQS_OFF - testl $TF_MASK,PT_EFLAGS(%esp) # If tracing set singlestep flag on exit + testl $X86_EFLAGS_TF,PT_EFLAGS(%esp) # If tracing set singlestep flag on exit jz no_singlestep orl $_TIF_SINGLESTEP,TI_flags(%ebp) no_singlestep: @@ -399,7 +397,7 @@ restore_all: # See comments in process.c:copy_thread() for details. movb PT_OLDSS(%esp), %ah movb PT_CS(%esp), %al - andl $(VM_MASK | (SEGMENT_TI_MASK << 8) | SEGMENT_RPL_MASK), %eax + andl $(X86_EFLAGS_VM | (SEGMENT_TI_MASK << 8) | SEGMENT_RPL_MASK), %eax cmpl $((SEGMENT_LDT << 8) | USER_RPL), %eax CFI_REMEMBER_STATE je ldt_ss # returning to user-space with LDT SS @@ -486,7 +484,7 @@ work_resched: work_notifysig: # deal with pending signals and # notify-resume requests #ifdef CONFIG_VM86 - testl $VM_MASK, PT_EFLAGS(%esp) + testl $X86_EFLAGS_VM, PT_EFLAGS(%esp) movl %esp, %eax jne work_notifysig_v86 # returning to kernel-space or # vm86-space @@ -543,9 +541,6 @@ END(syscall_exit_work) RING0_INT_FRAME # can't unwind into user space anyway syscall_fault: - pushl %eax # save orig_eax - CFI_ADJUST_CFA_OFFSET 4 - SAVE_ALL GET_THREAD_INFO(%ebp) movl $-EFAULT,PT_EAX(%esp) jmp resume_userspace diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S index c20c9e7e08dd..556a8df522a7 100644 --- a/arch/x86/kernel/entry_64.S +++ b/arch/x86/kernel/entry_64.S @@ -319,19 +319,17 @@ badsys: /* Do syscall tracing */ tracesys: SAVE_REST - movq $-ENOSYS,RAX(%rsp) + movq $-ENOSYS,RAX(%rsp) /* ptrace can change this for a bad syscall */ FIXUP_TOP_OF_STACK %rdi movq %rsp,%rdi call syscall_trace_enter LOAD_ARGS ARGOFFSET /* reload args from stack in case ptrace changed it */ RESTORE_REST cmpq $__NR_syscall_max,%rax - movq $-ENOSYS,%rcx - cmova %rcx,%rax - ja 1f + ja int_ret_from_sys_call /* RAX(%rsp) set to -ENOSYS above */ movq %r10,%rcx /* fixup for C */ call *sys_call_table(,%rax,8) -1: movq %rax,RAX-ARGOFFSET(%rsp) + movq %rax,RAX-ARGOFFSET(%rsp) /* Use IRET because user could have changed frame */ /* diff --git a/arch/x86/kernel/genapic_64.c b/arch/x86/kernel/genapic_64.c index 4ae7b6440260..9546ef408b92 100644 --- a/arch/x86/kernel/genapic_64.c +++ b/arch/x86/kernel/genapic_64.c @@ -15,6 +15,7 @@ #include <linux/kernel.h> #include <linux/ctype.h> #include <linux/init.h> +#include <linux/hardirq.h> #include <asm/smp.h> #include <asm/ipi.h> @@ -24,20 +25,20 @@ #include <acpi/acpi_bus.h> #endif -/* which logical CPU number maps to which CPU (physical APIC ID) */ -u16 x86_cpu_to_apicid_init[NR_CPUS] __initdata - = { [0 ... NR_CPUS-1] = BAD_APICID }; -void *x86_cpu_to_apicid_early_ptr; -DEFINE_PER_CPU(u16, x86_cpu_to_apicid) = BAD_APICID; -EXPORT_PER_CPU_SYMBOL(x86_cpu_to_apicid); +DEFINE_PER_CPU(int, x2apic_extra_bits); struct genapic __read_mostly *genapic = &apic_flat; +static enum uv_system_type uv_system_type; + /* * Check the APIC IDs in bios_cpu_apicid and choose the APIC mode. */ void __init setup_apic_routing(void) { + if (uv_system_type == UV_NON_UNIQUE_APIC) + genapic = &apic_x2apic_uv_x; + else #ifdef CONFIG_ACPI /* * Quirk: some x86_64 machines can only use physical APIC mode @@ -64,3 +65,37 @@ void send_IPI_self(int vector) { __send_IPI_shortcut(APIC_DEST_SELF, vector, APIC_DEST_PHYSICAL); } + +int __init acpi_madt_oem_check(char *oem_id, char *oem_table_id) +{ + if (!strcmp(oem_id, "SGI")) { + if (!strcmp(oem_table_id, "UVL")) + uv_system_type = UV_LEGACY_APIC; + else if (!strcmp(oem_table_id, "UVX")) + uv_system_type = UV_X2APIC; + else if (!strcmp(oem_table_id, "UVH")) + uv_system_type = UV_NON_UNIQUE_APIC; + } + return 0; +} + +unsigned int read_apic_id(void) +{ + unsigned int id; + + WARN_ON(preemptible()); + id = apic_read(APIC_ID); + if (uv_system_type >= UV_X2APIC) + id |= __get_cpu_var(x2apic_extra_bits); + return id; +} + +enum uv_system_type get_uv_system_type(void) +{ + return uv_system_type; +} + +int is_uv_system(void) +{ + return uv_system_type != UV_NONE; +} diff --git a/arch/x86/kernel/genapic_flat_64.c b/arch/x86/kernel/genapic_flat_64.c index 07352b74bda6..1a9c68845ee8 100644 --- a/arch/x86/kernel/genapic_flat_64.c +++ b/arch/x86/kernel/genapic_flat_64.c @@ -97,7 +97,7 @@ static void flat_send_IPI_all(int vector) static int flat_apic_id_registered(void) { - return physid_isset(GET_APIC_ID(apic_read(APIC_ID)), phys_cpu_present_map); + return physid_isset(GET_APIC_ID(read_apic_id()), phys_cpu_present_map); } static unsigned int flat_cpu_mask_to_apicid(cpumask_t cpumask) @@ -138,12 +138,9 @@ static cpumask_t physflat_target_cpus(void) static cpumask_t physflat_vector_allocation_domain(int cpu) { - cpumask_t domain = CPU_MASK_NONE; - cpu_set(cpu, domain); - return domain; + return cpumask_of_cpu(cpu); } - static void physflat_send_IPI_mask(cpumask_t cpumask, int vector) { send_IPI_mask_sequence(cpumask, vector); diff --git a/arch/x86/kernel/genx2apic_uv_x.c b/arch/x86/kernel/genx2apic_uv_x.c new file mode 100644 index 000000000000..5d77c9cd8e15 --- /dev/null +++ b/arch/x86/kernel/genx2apic_uv_x.c @@ -0,0 +1,245 @@ +/* + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * SGI UV APIC functions (note: not an Intel compatible APIC) + * + * Copyright (C) 2007 Silicon Graphics, Inc. All rights reserved. + */ + +#include <linux/threads.h> +#include <linux/cpumask.h> +#include <linux/string.h> +#include <linux/kernel.h> +#include <linux/ctype.h> +#include <linux/init.h> +#include <linux/sched.h> +#include <linux/bootmem.h> +#include <linux/module.h> +#include <asm/smp.h> +#include <asm/ipi.h> +#include <asm/genapic.h> +#include <asm/uv/uv_mmrs.h> +#include <asm/uv/uv_hub.h> + +DEFINE_PER_CPU(struct uv_hub_info_s, __uv_hub_info); +EXPORT_PER_CPU_SYMBOL_GPL(__uv_hub_info); + +struct uv_blade_info *uv_blade_info; +EXPORT_SYMBOL_GPL(uv_blade_info); + +short *uv_node_to_blade; +EXPORT_SYMBOL_GPL(uv_node_to_blade); + +short *uv_cpu_to_blade; +EXPORT_SYMBOL_GPL(uv_cpu_to_blade); + +short uv_possible_blades; +EXPORT_SYMBOL_GPL(uv_possible_blades); + +/* Start with all IRQs pointing to boot CPU. IRQ balancing will shift them. */ + +static cpumask_t uv_target_cpus(void) +{ + return cpumask_of_cpu(0); +} + +static cpumask_t uv_vector_allocation_domain(int cpu) +{ + cpumask_t domain = CPU_MASK_NONE; + cpu_set(cpu, domain); + return domain; +} + +int uv_wakeup_secondary(int phys_apicid, unsigned int start_rip) +{ + unsigned long val; + int nasid; + + nasid = uv_apicid_to_nasid(phys_apicid); + val = (1UL << UVH_IPI_INT_SEND_SHFT) | + (phys_apicid << UVH_IPI_INT_APIC_ID_SHFT) | + (((long)start_rip << UVH_IPI_INT_VECTOR_SHFT) >> 12) | + (6 << UVH_IPI_INT_DELIVERY_MODE_SHFT); + uv_write_global_mmr64(nasid, UVH_IPI_INT, val); + return 0; +} + +static void uv_send_IPI_one(int cpu, int vector) +{ + unsigned long val, apicid; + int nasid; + + apicid = per_cpu(x86_cpu_to_apicid, cpu); /* ZZZ - cache node-local ? */ + nasid = uv_apicid_to_nasid(apicid); + val = + (1UL << UVH_IPI_INT_SEND_SHFT) | (apicid << + UVH_IPI_INT_APIC_ID_SHFT) | + (vector << UVH_IPI_INT_VECTOR_SHFT); + uv_write_global_mmr64(nasid, UVH_IPI_INT, val); + printk(KERN_DEBUG + "UV: IPI to cpu %d, apicid 0x%lx, vec %d, nasid%d, val 0x%lx\n", + cpu, apicid, vector, nasid, val); +} + +static void uv_send_IPI_mask(cpumask_t mask, int vector) +{ + unsigned int cpu; + + for (cpu = 0; cpu < NR_CPUS; ++cpu) + if (cpu_isset(cpu, mask)) + uv_send_IPI_one(cpu, vector); +} + +static void uv_send_IPI_allbutself(int vector) +{ + cpumask_t mask = cpu_online_map; + + cpu_clear(smp_processor_id(), mask); + + if (!cpus_empty(mask)) + uv_send_IPI_mask(mask, vector); +} + +static void uv_send_IPI_all(int vector) +{ + uv_send_IPI_mask(cpu_online_map, vector); +} + +static int uv_apic_id_registered(void) +{ + return 1; +} + +static unsigned int uv_cpu_mask_to_apicid(cpumask_t cpumask) +{ + int cpu; + + /* + * We're using fixed IRQ delivery, can only return one phys APIC ID. + * May as well be the first. + */ + cpu = first_cpu(cpumask); + if ((unsigned)cpu < NR_CPUS) + return per_cpu(x86_cpu_to_apicid, cpu); + else + return BAD_APICID; +} + +static unsigned int phys_pkg_id(int index_msb) +{ + return GET_APIC_ID(read_apic_id()) >> index_msb; +} + +#ifdef ZZZ /* Needs x2apic patch */ +static void uv_send_IPI_self(int vector) +{ + apic_write(APIC_SELF_IPI, vector); +} +#endif + +struct genapic apic_x2apic_uv_x = { + .name = "UV large system", + .int_delivery_mode = dest_Fixed, + .int_dest_mode = (APIC_DEST_PHYSICAL != 0), + .target_cpus = uv_target_cpus, + .vector_allocation_domain = uv_vector_allocation_domain,/* Fixme ZZZ */ + .apic_id_registered = uv_apic_id_registered, + .send_IPI_all = uv_send_IPI_all, + .send_IPI_allbutself = uv_send_IPI_allbutself, + .send_IPI_mask = uv_send_IPI_mask, + /* ZZZ.send_IPI_self = uv_send_IPI_self, */ + .cpu_mask_to_apicid = uv_cpu_mask_to_apicid, + .phys_pkg_id = phys_pkg_id, /* Fixme ZZZ */ +}; + +static __cpuinit void set_x2apic_extra_bits(int nasid) +{ + __get_cpu_var(x2apic_extra_bits) = ((nasid >> 1) << 6); +} + +/* + * Called on boot cpu. + */ +static __init void uv_system_init(void) +{ + union uvh_si_addr_map_config_u m_n_config; + int bytes, nid, cpu, lcpu, nasid, last_nasid, blade; + unsigned long mmr_base; + + m_n_config.v = uv_read_local_mmr(UVH_SI_ADDR_MAP_CONFIG); + mmr_base = + uv_read_local_mmr(UVH_RH_GAM_MMR_OVERLAY_CONFIG_MMR) & + ~UV_MMR_ENABLE; + printk(KERN_DEBUG "UV: global MMR base 0x%lx\n", mmr_base); + + last_nasid = -1; + for_each_possible_cpu(cpu) { + nid = cpu_to_node(cpu); + nasid = uv_apicid_to_nasid(per_cpu(x86_cpu_to_apicid, cpu)); + if (nasid != last_nasid) + uv_possible_blades++; + last_nasid = nasid; + } + printk(KERN_DEBUG "UV: Found %d blades\n", uv_num_possible_blades()); + + bytes = sizeof(struct uv_blade_info) * uv_num_possible_blades(); + uv_blade_info = alloc_bootmem_pages(bytes); + + bytes = sizeof(uv_node_to_blade[0]) * num_possible_nodes(); + uv_node_to_blade = alloc_bootmem_pages(bytes); + memset(uv_node_to_blade, 255, bytes); + + bytes = sizeof(uv_cpu_to_blade[0]) * num_possible_cpus(); + uv_cpu_to_blade = alloc_bootmem_pages(bytes); + memset(uv_cpu_to_blade, 255, bytes); + + last_nasid = -1; + blade = -1; + lcpu = -1; + for_each_possible_cpu(cpu) { + nid = cpu_to_node(cpu); + nasid = uv_apicid_to_nasid(per_cpu(x86_cpu_to_apicid, cpu)); + if (nasid != last_nasid) { + blade++; + lcpu = -1; + uv_blade_info[blade].nr_posible_cpus = 0; + uv_blade_info[blade].nr_online_cpus = 0; + } + last_nasid = nasid; + lcpu++; + + uv_cpu_hub_info(cpu)->m_val = m_n_config.s.m_skt; + uv_cpu_hub_info(cpu)->n_val = m_n_config.s.n_skt; + uv_cpu_hub_info(cpu)->numa_blade_id = blade; + uv_cpu_hub_info(cpu)->blade_processor_id = lcpu; + uv_cpu_hub_info(cpu)->local_nasid = nasid; + uv_cpu_hub_info(cpu)->gnode_upper = + nasid & ~((1 << uv_hub_info->n_val) - 1); + uv_cpu_hub_info(cpu)->global_mmr_base = mmr_base; + uv_cpu_hub_info(cpu)->coherency_domain_number = 0;/* ZZZ */ + uv_blade_info[blade].nasid = nasid; + uv_blade_info[blade].nr_posible_cpus++; + uv_node_to_blade[nid] = blade; + uv_cpu_to_blade[cpu] = blade; + + printk(KERN_DEBUG "UV cpu %d, apicid 0x%x, nasid %d, nid %d\n", + cpu, per_cpu(x86_cpu_to_apicid, cpu), nasid, nid); + printk(KERN_DEBUG "UV lcpu %d, blade %d\n", lcpu, blade); + } +} + +/* + * Called on each cpu to initialize the per_cpu UV data area. + */ +void __cpuinit uv_cpu_init(void) +{ + if (!uv_node_to_blade) + uv_system_init(); + + uv_blade_info[uv_numa_blade_id()].nr_online_cpus++; + + if (get_uv_system_type() == UV_NON_UNIQUE_APIC) + set_x2apic_extra_bits(uv_hub_info->local_nasid); +} diff --git a/arch/x86/kernel/head32.c b/arch/x86/kernel/head32.c new file mode 100644 index 000000000000..3db059058927 --- /dev/null +++ b/arch/x86/kernel/head32.c @@ -0,0 +1,14 @@ +/* + * linux/arch/i386/kernel/head32.c -- prepare to run common code + * + * Copyright (C) 2000 Andrea Arcangeli <andrea@suse.de> SuSE + * Copyright (C) 2007 Eric Biederman <ebiederm@xmission.com> + */ + +#include <linux/init.h> +#include <linux/start_kernel.h> + +void __init i386_start_kernel(void) +{ + start_kernel(); +} diff --git a/arch/x86/kernel/head64.c b/arch/x86/kernel/head64.c index ad2440832de0..d6d54faa84df 100644 --- a/arch/x86/kernel/head64.c +++ b/arch/x86/kernel/head64.c @@ -49,39 +49,75 @@ static void __init copy_bootdata(char *real_mode_data) } } -#define EBDA_ADDR_POINTER 0x40E +#define BIOS_EBDA_SEGMENT 0x40E +#define BIOS_LOWMEM_KILOBYTES 0x413 -static __init void reserve_ebda(void) +/* + * The BIOS places the EBDA/XBDA at the top of conventional + * memory, and usually decreases the reported amount of + * conventional memory (int 0x12) too. This also contains a + * workaround for Dell systems that neglect to reserve EBDA. + * The same workaround also avoids a problem with the AMD768MPX + * chipset: reserve a page before VGA to prevent PCI prefetch + * into it (errata #56). Usually the page is reserved anyways, + * unless you have no PS/2 mouse plugged in. + */ +static void __init reserve_ebda_region(void) { - unsigned ebda_addr, ebda_size; + unsigned int lowmem, ebda_addr; + + /* To determine the position of the EBDA and the */ + /* end of conventional memory, we need to look at */ + /* the BIOS data area. In a paravirtual environment */ + /* that area is absent. We'll just have to assume */ + /* that the paravirt case can handle memory setup */ + /* correctly, without our help. */ + if (paravirt_enabled()) + return; - /* - * there is a real-mode segmented pointer pointing to the - * 4K EBDA area at 0x40E - */ - ebda_addr = *(unsigned short *)__va(EBDA_ADDR_POINTER); + /* end of low (conventional) memory */ + lowmem = *(unsigned short *)__va(BIOS_LOWMEM_KILOBYTES); + lowmem <<= 10; + + /* start of EBDA area */ + ebda_addr = *(unsigned short *)__va(BIOS_EBDA_SEGMENT); ebda_addr <<= 4; - if (!ebda_addr) - return; + /* Fixup: bios puts an EBDA in the top 64K segment */ + /* of conventional memory, but does not adjust lowmem. */ + if ((lowmem - ebda_addr) <= 0x10000) + lowmem = ebda_addr; - ebda_size = *(unsigned short *)__va(ebda_addr); + /* Fixup: bios does not report an EBDA at all. */ + /* Some old Dells seem to need 4k anyhow (bugzilla 2990) */ + if ((ebda_addr == 0) && (lowmem >= 0x9f000)) + lowmem = 0x9f000; - /* Round EBDA up to pages */ - if (ebda_size == 0) - ebda_size = 1; - ebda_size <<= 10; - ebda_size = round_up(ebda_size + (ebda_addr & ~PAGE_MASK), PAGE_SIZE); - if (ebda_size > 64*1024) - ebda_size = 64*1024; + /* Paranoia: should never happen, but... */ + if ((lowmem == 0) || (lowmem >= 0x100000)) + lowmem = 0x9f000; - reserve_early(ebda_addr, ebda_addr + ebda_size, "EBDA"); + /* reserve all memory between lowmem and the 1MB mark */ + reserve_early(lowmem, 0x100000, "BIOS reserved"); } void __init x86_64_start_kernel(char * real_mode_data) { int i; + /* + * Build-time sanity checks on the kernel image and module + * area mappings. (these are purely build-time and produce no code) + */ + BUILD_BUG_ON(MODULES_VADDR < KERNEL_IMAGE_START); + BUILD_BUG_ON(MODULES_VADDR-KERNEL_IMAGE_START < KERNEL_IMAGE_SIZE); + BUILD_BUG_ON(MODULES_LEN + KERNEL_IMAGE_SIZE > 2*PUD_SIZE); + BUILD_BUG_ON((KERNEL_IMAGE_START & ~PMD_MASK) != 0); + BUILD_BUG_ON((MODULES_VADDR & ~PMD_MASK) != 0); + BUILD_BUG_ON(!(MODULES_VADDR > __START_KERNEL)); + BUILD_BUG_ON(!(((MODULES_END - 1) & PGDIR_MASK) == + (__START_KERNEL & PGDIR_MASK))); + /* clear bss before set_intr_gate with early_idt_handler */ clear_bss(); @@ -91,7 +127,7 @@ void __init x86_64_start_kernel(char * real_mode_data) /* Cleanup the over mapped high alias */ cleanup_highmap(); - for (i = 0; i < IDT_ENTRIES; i++) { + for (i = 0; i < NUM_EXCEPTION_VECTORS; i++) { #ifdef CONFIG_EARLY_PRINTK set_intr_gate(i, &early_idt_handlers[i]); #else @@ -118,7 +154,7 @@ void __init x86_64_start_kernel(char * real_mode_data) reserve_early(ramdisk_image, ramdisk_end, "RAMDISK"); } - reserve_ebda(); + reserve_ebda_region(); /* * At this point everything still needed from the boot loader diff --git a/arch/x86/kernel/head_32.S b/arch/x86/kernel/head_32.S index 74d87ea85b5c..826988a6e964 100644 --- a/arch/x86/kernel/head_32.S +++ b/arch/x86/kernel/head_32.S @@ -450,7 +450,7 @@ is386: movl $2,%ecx # set MP jmp initialize_secondary # all other CPUs call initialize_secondary 1: #endif /* CONFIG_SMP */ - jmp start_kernel + jmp i386_start_kernel /* * We depend on ET to be correct. This checks for 287/387. diff --git a/arch/x86/kernel/head_64.S b/arch/x86/kernel/head_64.S index a007454133a3..10a1955bb1d1 100644 --- a/arch/x86/kernel/head_64.S +++ b/arch/x86/kernel/head_64.S @@ -132,10 +132,6 @@ ident_complete: addq %rbp, trampoline_level4_pgt + 0(%rip) addq %rbp, trampoline_level4_pgt + (511*8)(%rip) #endif -#ifdef CONFIG_ACPI_SLEEP - addq %rbp, wakeup_level4_pgt + 0(%rip) - addq %rbp, wakeup_level4_pgt + (511*8)(%rip) -#endif /* Due to ENTRY(), sometimes the empty space gets filled with * zeros. Better take a jmp than relying on empty space being @@ -267,21 +263,16 @@ ENTRY(secondary_startup_64) bad_address: jmp bad_address + .section ".init.text","ax" #ifdef CONFIG_EARLY_PRINTK -.macro early_idt_tramp first, last - .ifgt \last-\first - early_idt_tramp \first, \last-1 - .endif - movl $\last,%esi - jmp early_idt_handler -.endm - .globl early_idt_handlers early_idt_handlers: - early_idt_tramp 0, 63 - early_idt_tramp 64, 127 - early_idt_tramp 128, 191 - early_idt_tramp 192, 255 + i = 0 + .rept NUM_EXCEPTION_VECTORS + movl $i, %esi + jmp early_idt_handler + i = i + 1 + .endr #endif ENTRY(early_idt_handler) @@ -327,6 +318,7 @@ early_idt_msg: early_idt_ripmsg: .asciz "RIP %s\n" #endif /* CONFIG_EARLY_PRINTK */ + .previous .balign PAGE_SIZE @@ -383,12 +375,12 @@ NEXT_PAGE(level2_ident_pgt) NEXT_PAGE(level2_kernel_pgt) /* - * 128 MB kernel mapping. We spend a full page on this pagetable + * 512 MB kernel mapping. We spend a full page on this pagetable * anyway. * * The kernel code+data+bss must not be bigger than that. * - * (NOTE: at +128MB starts the module area, see MODULES_VADDR. + * (NOTE: at +512MB starts the module area, see MODULES_VADDR. * If you want to increase this then increase MODULES_VADDR * too.) */ diff --git a/arch/x86/kernel/i386_ksyms_32.c b/arch/x86/kernel/i386_ksyms_32.c index 061627806a2d..deb43785e923 100644 --- a/arch/x86/kernel/i386_ksyms_32.c +++ b/arch/x86/kernel/i386_ksyms_32.c @@ -1,13 +1,8 @@ #include <linux/module.h> -#include <asm/semaphore.h> #include <asm/checksum.h> #include <asm/desc.h> #include <asm/pgtable.h> -EXPORT_SYMBOL(__down_failed); -EXPORT_SYMBOL(__down_failed_interruptible); -EXPORT_SYMBOL(__down_failed_trylock); -EXPORT_SYMBOL(__up_wakeup); /* Networking helper routines. */ EXPORT_SYMBOL(csum_partial_copy_generic); diff --git a/arch/x86/kernel/i387.c b/arch/x86/kernel/i387.c index d2e39e69aaf8..8f8102d967b3 100644 --- a/arch/x86/kernel/i387.c +++ b/arch/x86/kernel/i387.c @@ -5,45 +5,41 @@ * General FPU state handling cleanups * Gareth Hughes <gareth@valinux.com>, May 2000 */ - -#include <linux/sched.h> #include <linux/module.h> #include <linux/regset.h> +#include <linux/sched.h> + +#include <asm/sigcontext.h> #include <asm/processor.h> -#include <asm/i387.h> #include <asm/math_emu.h> -#include <asm/sigcontext.h> -#include <asm/user.h> -#include <asm/ptrace.h> #include <asm/uaccess.h> +#include <asm/ptrace.h> +#include <asm/i387.h> +#include <asm/user.h> #ifdef CONFIG_X86_64 - -#include <asm/sigcontext32.h> -#include <asm/user32.h> - +# include <asm/sigcontext32.h> +# include <asm/user32.h> #else - -#define save_i387_ia32 save_i387 -#define restore_i387_ia32 restore_i387 - -#define _fpstate_ia32 _fpstate -#define user_i387_ia32_struct user_i387_struct -#define user32_fxsr_struct user_fxsr_struct - +# define save_i387_ia32 save_i387 +# define restore_i387_ia32 restore_i387 +# define _fpstate_ia32 _fpstate +# define user_i387_ia32_struct user_i387_struct +# define user32_fxsr_struct user_fxsr_struct #endif #ifdef CONFIG_MATH_EMULATION -#define HAVE_HWFP (boot_cpu_data.hard_math) +# define HAVE_HWFP (boot_cpu_data.hard_math) #else -#define HAVE_HWFP 1 +# define HAVE_HWFP 1 #endif -static unsigned int mxcsr_feature_mask __read_mostly = 0xffffffffu; +static unsigned int mxcsr_feature_mask __read_mostly = 0xffffffffu; void mxcsr_feature_mask_init(void) { unsigned long mask = 0; + clts(); if (cpu_has_fxsr) { memset(¤t->thread.i387.fxsave, 0, @@ -69,10 +65,11 @@ void __cpuinit fpu_init(void) if (offsetof(struct task_struct, thread.i387.fxsave) & 15) __bad_fxsave_alignment(); + set_in_cr4(X86_CR4_OSFXSR); set_in_cr4(X86_CR4_OSXMMEXCPT); - write_cr0(oldcr0 & ~((1UL<<3)|(1UL<<2))); /* clear TS and EM */ + write_cr0(oldcr0 & ~(X86_CR0_TS|X86_CR0_EM)); /* clear TS and EM */ mxcsr_feature_mask_init(); /* clean state in init */ @@ -178,6 +175,7 @@ static inline unsigned short twd_i387_to_fxsr(unsigned short twd) tmp = (tmp | (tmp >> 1)) & 0x3333; /* 00VV00VV00VV00VV */ tmp = (tmp | (tmp >> 2)) & 0x0f0f; /* 0000VVVV0000VVVV */ tmp = (tmp | (tmp >> 4)) & 0x00ff; /* 00000000VVVVVVVV */ + return tmp; } @@ -232,8 +230,8 @@ static inline u32 twd_fxsr_to_i387(struct i387_fxsave_struct *fxsave) * FXSR floating point environment conversions. */ -static void convert_from_fxsr(struct user_i387_ia32_struct *env, - struct task_struct *tsk) +static void +convert_from_fxsr(struct user_i387_ia32_struct *env, struct task_struct *tsk) { struct i387_fxsave_struct *fxsave = &tsk->thread.i387.fxsave; struct _fpreg *to = (struct _fpreg *) &env->st_space[0]; @@ -252,10 +250,11 @@ static void convert_from_fxsr(struct user_i387_ia32_struct *env, * should be actually ds/cs at fpu exception time, but * that information is not available in 64bit mode. */ - asm("mov %%ds,%0" : "=r" (env->fos)); - asm("mov %%cs,%0" : "=r" (env->fcs)); + asm("mov %%ds, %[fos]" : [fos] "=r" (env->fos)); + asm("mov %%cs, %[fcs]" : [fcs] "=r" (env->fcs)); } else { struct pt_regs *regs = task_pt_regs(tsk); + env->fos = 0xffff0000 | tsk->thread.ds; env->fcs = regs->cs; } @@ -309,9 +308,10 @@ int fpregs_get(struct task_struct *target, const struct user_regset *regset, init_fpu(target); - if (!cpu_has_fxsr) + if (!cpu_has_fxsr) { return user_regset_copyout(&pos, &count, &kbuf, &ubuf, &target->thread.i387.fsave, 0, -1); + } if (kbuf && pos == 0 && count == sizeof(env)) { convert_from_fxsr(kbuf, target); @@ -319,6 +319,7 @@ int fpregs_get(struct task_struct *target, const struct user_regset *regset, } convert_from_fxsr(&env, target); + return user_regset_copyout(&pos, &count, &kbuf, &ubuf, &env, 0, -1); } @@ -335,9 +336,10 @@ int fpregs_set(struct task_struct *target, const struct user_regset *regset, init_fpu(target); set_stopped_child_used_math(target); - if (!cpu_has_fxsr) + if (!cpu_has_fxsr) { return user_regset_copyin(&pos, &count, &kbuf, &ubuf, &target->thread.i387.fsave, 0, -1); + } if (pos > 0 || count < sizeof(env)) convert_from_fxsr(&env, target); @@ -392,28 +394,28 @@ int save_i387_ia32(struct _fpstate_ia32 __user *buf) { if (!used_math()) return 0; - - /* This will cause a "finit" to be triggered by the next + /* + * This will cause a "finit" to be triggered by the next * attempted FPU operation by the 'current' process. */ clear_used_math(); - if (HAVE_HWFP) { - if (cpu_has_fxsr) { - return save_i387_fxsave(buf); - } else { - return save_i387_fsave(buf); - } - } else { + if (!HAVE_HWFP) { return fpregs_soft_get(current, NULL, 0, sizeof(struct user_i387_ia32_struct), NULL, buf) ? -1 : 1; } + + if (cpu_has_fxsr) + return save_i387_fxsave(buf); + else + return save_i387_fsave(buf); } static inline int restore_i387_fsave(struct _fpstate_ia32 __user *buf) { struct task_struct *tsk = current; + clear_fpu(tsk); return __copy_from_user(&tsk->thread.i387.fsave, buf, sizeof(struct i387_fsave_struct)); @@ -421,9 +423,10 @@ static inline int restore_i387_fsave(struct _fpstate_ia32 __user *buf) static int restore_i387_fxsave(struct _fpstate_ia32 __user *buf) { - int err; struct task_struct *tsk = current; struct user_i387_ia32_struct env; + int err; + clear_fpu(tsk); err = __copy_from_user(&tsk->thread.i387.fxsave, &buf->_fxsr_env[0], sizeof(struct i387_fxsave_struct)); @@ -432,6 +435,7 @@ static int restore_i387_fxsave(struct _fpstate_ia32 __user *buf) if (err || __copy_from_user(&env, buf, sizeof(env))) return 1; convert_to_fxsr(tsk, &env); + return 0; } @@ -440,17 +444,17 @@ int restore_i387_ia32(struct _fpstate_ia32 __user *buf) int err; if (HAVE_HWFP) { - if (cpu_has_fxsr) { + if (cpu_has_fxsr) err = restore_i387_fxsave(buf); - } else { + else err = restore_i387_fsave(buf); - } } else { err = fpregs_soft_set(current, NULL, 0, sizeof(struct user_i387_ia32_struct), NULL, buf) != 0; } set_used_math(); + return err; } @@ -463,8 +467,8 @@ int restore_i387_ia32(struct _fpstate_ia32 __user *buf) */ int dump_fpu(struct pt_regs *regs, struct user_i387_struct *fpu) { - int fpvalid; struct task_struct *tsk = current; + int fpvalid; fpvalid = !!used_math(); if (fpvalid) diff --git a/arch/x86/kernel/io_apic_32.c b/arch/x86/kernel/io_apic_32.c index 4ca548632c8d..2e2f42074e18 100644 --- a/arch/x86/kernel/io_apic_32.c +++ b/arch/x86/kernel/io_apic_32.c @@ -71,6 +71,16 @@ int sis_apic_bug = -1; */ int nr_ioapic_registers[MAX_IO_APICS]; +/* I/O APIC entries */ +struct mpc_config_ioapic mp_ioapics[MAX_IO_APICS]; +int nr_ioapics; + +/* MP IRQ source entries */ +struct mpc_config_intsrc mp_irqs[MAX_IRQ_SOURCES]; + +/* # of MP IRQ source entries */ +int mp_irq_entries; + static int disable_timer_pin_1 __initdata; /* @@ -810,10 +820,7 @@ static int __init find_isa_irq_pin(int irq, int type) for (i = 0; i < mp_irq_entries; i++) { int lbus = mp_irqs[i].mpc_srcbus; - if ((mp_bus_id_to_type[lbus] == MP_BUS_ISA || - mp_bus_id_to_type[lbus] == MP_BUS_EISA || - mp_bus_id_to_type[lbus] == MP_BUS_MCA - ) && + if (test_bit(lbus, mp_bus_not_pci) && (mp_irqs[i].mpc_irqtype == type) && (mp_irqs[i].mpc_srcbusirq == irq)) @@ -829,10 +836,7 @@ static int __init find_isa_irq_apic(int irq, int type) for (i = 0; i < mp_irq_entries; i++) { int lbus = mp_irqs[i].mpc_srcbus; - if ((mp_bus_id_to_type[lbus] == MP_BUS_ISA || - mp_bus_id_to_type[lbus] == MP_BUS_EISA || - mp_bus_id_to_type[lbus] == MP_BUS_MCA - ) && + if (test_bit(lbus, mp_bus_not_pci) && (mp_irqs[i].mpc_irqtype == type) && (mp_irqs[i].mpc_srcbusirq == irq)) break; @@ -872,7 +876,7 @@ int IO_APIC_get_PCI_irq_vector(int bus, int slot, int pin) mp_irqs[i].mpc_dstapic == MP_APIC_ALL) break; - if ((mp_bus_id_to_type[lbus] == MP_BUS_PCI) && + if (!test_bit(lbus, mp_bus_not_pci) && !mp_irqs[i].mpc_irqtype && (bus == lbus) && (slot == ((mp_irqs[i].mpc_srcbusirq >> 2) & 0x1f))) { @@ -921,6 +925,7 @@ void __init setup_ioapic_dest(void) } #endif +#if defined(CONFIG_EISA) || defined(CONFIG_MCA) /* * EISA Edge/Level control register, ELCR */ @@ -934,6 +939,13 @@ static int EISA_ELCR(unsigned int irq) "Broken MPtable reports ISA irq %d\n", irq); return 0; } +#endif + +/* ISA interrupts are always polarity zero edge triggered, + * when listed as conforming in the MP table. */ + +#define default_ISA_trigger(idx) (0) +#define default_ISA_polarity(idx) (0) /* EISA interrupts are always polarity zero and can be edge or level * trigger depending on the ELCR value. If an interrupt is listed as @@ -941,13 +953,7 @@ static int EISA_ELCR(unsigned int irq) * be read in from the ELCR */ #define default_EISA_trigger(idx) (EISA_ELCR(mp_irqs[idx].mpc_srcbusirq)) -#define default_EISA_polarity(idx) (0) - -/* ISA interrupts are always polarity zero edge triggered, - * when listed as conforming in the MP table. */ - -#define default_ISA_trigger(idx) (0) -#define default_ISA_polarity(idx) (0) +#define default_EISA_polarity(idx) default_ISA_polarity(idx) /* PCI interrupts are always polarity one level triggered, * when listed as conforming in the MP table. */ @@ -959,7 +965,7 @@ static int EISA_ELCR(unsigned int irq) * when listed as conforming in the MP table. */ #define default_MCA_trigger(idx) (1) -#define default_MCA_polarity(idx) (0) +#define default_MCA_polarity(idx) default_ISA_polarity(idx) static int MPBIOS_polarity(int idx) { @@ -973,35 +979,9 @@ static int MPBIOS_polarity(int idx) { case 0: /* conforms, ie. bus-type dependent polarity */ { - switch (mp_bus_id_to_type[bus]) - { - case MP_BUS_ISA: /* ISA pin */ - { - polarity = default_ISA_polarity(idx); - break; - } - case MP_BUS_EISA: /* EISA pin */ - { - polarity = default_EISA_polarity(idx); - break; - } - case MP_BUS_PCI: /* PCI pin */ - { - polarity = default_PCI_polarity(idx); - break; - } - case MP_BUS_MCA: /* MCA pin */ - { - polarity = default_MCA_polarity(idx); - break; - } - default: - { - printk(KERN_WARNING "broken BIOS!!\n"); - polarity = 1; - break; - } - } + polarity = test_bit(bus, mp_bus_not_pci)? + default_ISA_polarity(idx): + default_PCI_polarity(idx); break; } case 1: /* high active */ @@ -1042,11 +1022,15 @@ static int MPBIOS_trigger(int idx) { case 0: /* conforms, ie. bus-type dependent */ { + trigger = test_bit(bus, mp_bus_not_pci)? + default_ISA_trigger(idx): + default_PCI_trigger(idx); +#if defined(CONFIG_EISA) || defined(CONFIG_MCA) switch (mp_bus_id_to_type[bus]) { case MP_BUS_ISA: /* ISA pin */ { - trigger = default_ISA_trigger(idx); + /* set before the switch */ break; } case MP_BUS_EISA: /* EISA pin */ @@ -1056,7 +1040,7 @@ static int MPBIOS_trigger(int idx) } case MP_BUS_PCI: /* PCI pin */ { - trigger = default_PCI_trigger(idx); + /* set before the switch */ break; } case MP_BUS_MCA: /* MCA pin */ @@ -1071,6 +1055,7 @@ static int MPBIOS_trigger(int idx) break; } } +#endif break; } case 1: /* edge */ @@ -1120,39 +1105,22 @@ static int pin_2_irq(int idx, int apic, int pin) if (mp_irqs[idx].mpc_dstirq != pin) printk(KERN_ERR "broken BIOS or MPTABLE parser, ayiee!!\n"); - switch (mp_bus_id_to_type[bus]) - { - case MP_BUS_ISA: /* ISA pin */ - case MP_BUS_EISA: - case MP_BUS_MCA: - { - irq = mp_irqs[idx].mpc_srcbusirq; - break; - } - case MP_BUS_PCI: /* PCI pin */ - { - /* - * PCI IRQs are mapped in order - */ - i = irq = 0; - while (i < apic) - irq += nr_ioapic_registers[i++]; - irq += pin; - - /* - * For MPS mode, so far only needed by ES7000 platform - */ - if (ioapic_renumber_irq) - irq = ioapic_renumber_irq(apic, irq); + if (test_bit(bus, mp_bus_not_pci)) + irq = mp_irqs[idx].mpc_srcbusirq; + else { + /* + * PCI IRQs are mapped in order + */ + i = irq = 0; + while (i < apic) + irq += nr_ioapic_registers[i++]; + irq += pin; - break; - } - default: - { - printk(KERN_ERR "unknown bus type %d.\n",bus); - irq = 0; - break; - } + /* + * For MPS mode, so far only needed by ES7000 platform + */ + if (ioapic_renumber_irq) + irq = ioapic_renumber_irq(apic, irq); } /* @@ -1260,7 +1228,6 @@ static void __init setup_IO_APIC_irqs(void) { struct IO_APIC_route_entry entry; int apic, pin, idx, irq, first_notcon = 1, vector; - unsigned long flags; apic_printk(APIC_VERBOSE, KERN_DEBUG "init IO_APIC IRQs\n"); @@ -1326,9 +1293,7 @@ static void __init setup_IO_APIC_irqs(void) if (!apic && (irq < 16)) disable_8259A_irq(irq); } - spin_lock_irqsave(&ioapic_lock, flags); - __ioapic_write_entry(apic, pin, entry); - spin_unlock_irqrestore(&ioapic_lock, flags); + ioapic_write_entry(apic, pin, entry); } } @@ -1524,8 +1489,8 @@ void /*__init*/ print_local_APIC(void * dummy) printk("\n" KERN_DEBUG "printing local APIC contents on CPU#%d/%d:\n", smp_processor_id(), hard_smp_processor_id()); - v = apic_read(APIC_ID); - printk(KERN_INFO "... APIC ID: %08x (%01x)\n", v, GET_APIC_ID(v)); + printk(KERN_INFO "... APIC ID: %08x (%01x)\n", v, + GET_APIC_ID(read_apic_id())); v = apic_read(APIC_LVR); printk(KERN_INFO "... APIC VERSION: %08x\n", v); ver = GET_APIC_VERSION(v); @@ -1734,7 +1699,7 @@ void disable_IO_APIC(void) entry.delivery_mode = dest_ExtINT; /* ExtInt */ entry.vector = 0; entry.dest.physical.physical_dest = - GET_APIC_ID(apic_read(APIC_ID)); + GET_APIC_ID(read_apic_id()); /* * Add it to the IO-APIC irq-routing table: @@ -2031,8 +1996,7 @@ static inline void init_IO_APIC_traps(void) * 0x80, because int 0x80 is hm, kind of importantish. ;) */ for (irq = 0; irq < NR_IRQS ; irq++) { - int tmp = irq; - if (IO_APIC_IRQ(tmp) && !irq_vector[tmp]) { + if (IO_APIC_IRQ(irq) && !irq_vector[irq]) { /* * Hmm.. We don't have an entry for this, * so default to an old-fashioned 8259 @@ -2156,8 +2120,6 @@ static inline void unlock_ExtINT_logic(void) ioapic_write_entry(apic, pin, entry0); } -int timer_uses_ioapic_pin_0; - /* * This code may look a bit paranoid, but it's supposed to cooperate with * a wide range of boards and BIOS bugs. Fortunately only the timer IRQ @@ -2168,10 +2130,14 @@ static inline void __init check_timer(void) { int apic1, pin1, apic2, pin2; int vector; + unsigned int ver; unsigned long flags; local_irq_save(flags); + ver = apic_read(APIC_LVR); + ver = GET_APIC_VERSION(ver); + /* * get/set the timer IRQ vector: */ @@ -2184,11 +2150,15 @@ static inline void __init check_timer(void) * mode for the 8259A whenever interrupts are routed * through I/O APICs. Also IRQ0 has to be enabled in * the 8259A which implies the virtual wire has to be - * disabled in the local APIC. + * disabled in the local APIC. Finally timer interrupts + * need to be acknowledged manually in the 8259A for + * timer_interrupt() and for the i82489DX when using + * the NMI watchdog. */ apic_write_around(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_EXTINT); init_8259A(1); - timer_ack = 1; + timer_ack = !cpu_has_tsc; + timer_ack |= (nmi_watchdog == NMI_IO_APIC && !APIC_INTEGRATED(ver)); if (timer_over_8254 > 0) enable_8259A_irq(0); @@ -2197,9 +2167,6 @@ static inline void __init check_timer(void) pin2 = ioapic_i8259.pin; apic2 = ioapic_i8259.apic; - if (pin1 == 0) - timer_uses_ioapic_pin_0 = 1; - printk(KERN_INFO "..TIMER: vector=0x%02X apic1=%d pin1=%d apic2=%d pin2=%d\n", vector, apic1, pin1, apic2, pin2); @@ -2789,7 +2756,6 @@ int __init io_apic_get_redir_entries (int ioapic) int io_apic_set_pci_routing (int ioapic, int pin, int irq, int edge_level, int active_high_low) { struct IO_APIC_route_entry entry; - unsigned long flags; if (!IO_APIC_IRQ(irq)) { printk(KERN_ERR "IOAPIC[%d]: Invalid reference to IRQ 0\n", @@ -2830,9 +2796,7 @@ int io_apic_set_pci_routing (int ioapic, int pin, int irq, int edge_level, int a if (!ioapic && (irq < 16)) disable_8259A_irq(irq); - spin_lock_irqsave(&ioapic_lock, flags); - __ioapic_write_entry(ioapic, pin, entry); - spin_unlock_irqrestore(&ioapic_lock, flags); + ioapic_write_entry(ioapic, pin, entry); return 0; } diff --git a/arch/x86/kernel/io_apic_64.c b/arch/x86/kernel/io_apic_64.c index 1627c0d53e0b..b54464b26658 100644 --- a/arch/x86/kernel/io_apic_64.c +++ b/arch/x86/kernel/io_apic_64.c @@ -43,13 +43,15 @@ #include <asm/smp.h> #include <asm/desc.h> #include <asm/proto.h> -#include <asm/mach_apic.h> #include <asm/acpi.h> #include <asm/dma.h> #include <asm/nmi.h> #include <asm/msidef.h> #include <asm/hypertransport.h> +#include <mach_ipi.h> +#include <mach_apic.h> + struct irq_cfg { cpumask_t domain; cpumask_t old_domain; @@ -101,6 +103,16 @@ DEFINE_SPINLOCK(vector_lock); */ int nr_ioapic_registers[MAX_IO_APICS]; +/* I/O APIC entries */ +struct mpc_config_ioapic mp_ioapics[MAX_IO_APICS]; +int nr_ioapics; + +/* MP IRQ source entries */ +struct mpc_config_intsrc mp_irqs[MAX_IRQ_SOURCES]; + +/* # of MP IRQ source entries */ +int mp_irq_entries; + /* * Rough estimation of how many shared IRQs there are, can * be changed anytime. @@ -155,11 +167,10 @@ static inline void io_apic_modify(unsigned int apic, unsigned int value) writel(value, &io_apic->data); } -static int io_apic_level_ack_pending(unsigned int irq) +static bool io_apic_level_ack_pending(unsigned int irq) { struct irq_pin_list *entry; unsigned long flags; - int pending = 0; spin_lock_irqsave(&ioapic_lock, flags); entry = irq_2_pin + irq; @@ -172,13 +183,17 @@ static int io_apic_level_ack_pending(unsigned int irq) break; reg = io_apic_read(entry->apic, 0x10 + pin*2); /* Is the remote IRR bit set? */ - pending |= (reg >> 14) & 1; + if ((reg >> 14) & 1) { + spin_unlock_irqrestore(&ioapic_lock, flags); + return true; + } if (!entry->next) break; entry = irq_2_pin + entry->next; } spin_unlock_irqrestore(&ioapic_lock, flags); - return pending; + + return false; } /* @@ -902,9 +917,8 @@ static void __init setup_IO_APIC_irqs(void) static void __init setup_ExtINT_IRQ0_pin(unsigned int apic, unsigned int pin, int vector) { struct IO_APIC_route_entry entry; - unsigned long flags; - memset(&entry,0,sizeof(entry)); + memset(&entry, 0, sizeof(entry)); disable_8259A_irq(0); @@ -932,10 +946,7 @@ static void __init setup_ExtINT_IRQ0_pin(unsigned int apic, unsigned int pin, in /* * Add it to the IO-APIC irq-routing table: */ - spin_lock_irqsave(&ioapic_lock, flags); - io_apic_write(apic, 0x11+2*pin, *(((int *)&entry)+1)); - io_apic_write(apic, 0x10+2*pin, *(((int *)&entry)+0)); - spin_unlock_irqrestore(&ioapic_lock, flags); + ioapic_write_entry(apic, pin, entry); enable_8259A_irq(0); } @@ -1066,8 +1077,7 @@ void __apicdebuginit print_local_APIC(void * dummy) printk("\n" KERN_DEBUG "printing local APIC contents on CPU#%d/%d:\n", smp_processor_id(), hard_smp_processor_id()); - v = apic_read(APIC_ID); - printk(KERN_INFO "... APIC ID: %08x (%01x)\n", v, GET_APIC_ID(v)); + printk(KERN_INFO "... APIC ID: %08x (%01x)\n", v, GET_APIC_ID(read_apic_id())); v = apic_read(APIC_LVR); printk(KERN_INFO "... APIC VERSION: %08x\n", v); ver = GET_APIC_VERSION(v); @@ -1261,7 +1271,7 @@ void disable_IO_APIC(void) entry.dest_mode = 0; /* Physical */ entry.delivery_mode = dest_ExtINT; /* ExtInt */ entry.vector = 0; - entry.dest = GET_APIC_ID(apic_read(APIC_ID)); + entry.dest = GET_APIC_ID(read_apic_id()); /* * Add it to the IO-APIC irq-routing table: @@ -1352,9 +1362,7 @@ static int ioapic_retrigger_irq(unsigned int irq) unsigned long flags; spin_lock_irqsave(&vector_lock, flags); - cpus_clear(mask); - cpu_set(first_cpu(cfg->domain), mask); - + mask = cpumask_of_cpu(first_cpu(cfg->domain)); send_IPI_mask(mask, cfg->vector); spin_unlock_irqrestore(&vector_lock, flags); @@ -1517,8 +1525,7 @@ static inline void init_IO_APIC_traps(void) * 0x80, because int 0x80 is hm, kind of importantish. ;) */ for (irq = 0; irq < NR_IRQS ; irq++) { - int tmp = irq; - if (IO_APIC_IRQ(tmp) && !irq_cfg[tmp].vector) { + if (IO_APIC_IRQ(irq) && !irq_cfg[irq].vector) { /* * Hmm.. We don't have an entry for this, * so default to an old-fashioned 8259 @@ -1597,17 +1604,14 @@ static inline void unlock_ExtINT_logic(void) int apic, pin, i; struct IO_APIC_route_entry entry0, entry1; unsigned char save_control, save_freq_select; - unsigned long flags; pin = find_isa_irq_pin(8, mp_INT); apic = find_isa_irq_apic(8, mp_INT); if (pin == -1) return; - spin_lock_irqsave(&ioapic_lock, flags); - *(((int *)&entry0) + 1) = io_apic_read(apic, 0x11 + 2 * pin); - *(((int *)&entry0) + 0) = io_apic_read(apic, 0x10 + 2 * pin); - spin_unlock_irqrestore(&ioapic_lock, flags); + entry0 = ioapic_read_entry(apic, pin); + clear_IO_APIC_pin(apic, pin); memset(&entry1, 0, sizeof(entry1)); @@ -1620,10 +1624,7 @@ static inline void unlock_ExtINT_logic(void) entry1.trigger = 0; entry1.vector = 0; - spin_lock_irqsave(&ioapic_lock, flags); - io_apic_write(apic, 0x11 + 2 * pin, *(((int *)&entry1) + 1)); - io_apic_write(apic, 0x10 + 2 * pin, *(((int *)&entry1) + 0)); - spin_unlock_irqrestore(&ioapic_lock, flags); + ioapic_write_entry(apic, pin, entry1); save_control = CMOS_READ(RTC_CONTROL); save_freq_select = CMOS_READ(RTC_FREQ_SELECT); @@ -1642,10 +1643,7 @@ static inline void unlock_ExtINT_logic(void) CMOS_WRITE(save_freq_select, RTC_FREQ_SELECT); clear_IO_APIC_pin(apic, pin); - spin_lock_irqsave(&ioapic_lock, flags); - io_apic_write(apic, 0x11 + 2 * pin, *(((int *)&entry0) + 1)); - io_apic_write(apic, 0x10 + 2 * pin, *(((int *)&entry0) + 0)); - spin_unlock_irqrestore(&ioapic_lock, flags); + ioapic_write_entry(apic, pin, entry0); } /* @@ -2314,7 +2312,6 @@ static struct resource * __init ioapic_setup_resources(void) res = (void *)mem; if (mem != NULL) { - memset(mem, 0, n); mem += sizeof(struct resource) * nr_ioapics; for (i = 0; i < nr_ioapics; i++) { diff --git a/arch/x86/kernel/ipi.c b/arch/x86/kernel/ipi.c new file mode 100644 index 000000000000..c0df7b89ca23 --- /dev/null +++ b/arch/x86/kernel/ipi.c @@ -0,0 +1,178 @@ +#include <linux/cpumask.h> +#include <linux/interrupt.h> +#include <linux/init.h> + +#include <linux/mm.h> +#include <linux/delay.h> +#include <linux/spinlock.h> +#include <linux/kernel_stat.h> +#include <linux/mc146818rtc.h> +#include <linux/cache.h> +#include <linux/interrupt.h> +#include <linux/cpu.h> +#include <linux/module.h> + +#include <asm/smp.h> +#include <asm/mtrr.h> +#include <asm/tlbflush.h> +#include <asm/mmu_context.h> +#include <asm/apic.h> +#include <asm/proto.h> + +#ifdef CONFIG_X86_32 +#include <mach_apic.h> +/* + * the following functions deal with sending IPIs between CPUs. + * + * We use 'broadcast', CPU->CPU IPIs and self-IPIs too. + */ + +static inline int __prepare_ICR(unsigned int shortcut, int vector) +{ + unsigned int icr = shortcut | APIC_DEST_LOGICAL; + + switch (vector) { + default: + icr |= APIC_DM_FIXED | vector; + break; + case NMI_VECTOR: + icr |= APIC_DM_NMI; + break; + } + return icr; +} + +static inline int __prepare_ICR2(unsigned int mask) +{ + return SET_APIC_DEST_FIELD(mask); +} + +void __send_IPI_shortcut(unsigned int shortcut, int vector) +{ + /* + * Subtle. In the case of the 'never do double writes' workaround + * we have to lock out interrupts to be safe. As we don't care + * of the value read we use an atomic rmw access to avoid costly + * cli/sti. Otherwise we use an even cheaper single atomic write + * to the APIC. + */ + unsigned int cfg; + + /* + * Wait for idle. + */ + apic_wait_icr_idle(); + + /* + * No need to touch the target chip field + */ + cfg = __prepare_ICR(shortcut, vector); + + /* + * Send the IPI. The write to APIC_ICR fires this off. + */ + apic_write_around(APIC_ICR, cfg); +} + +void send_IPI_self(int vector) +{ + __send_IPI_shortcut(APIC_DEST_SELF, vector); +} + +/* + * This is used to send an IPI with no shorthand notation (the destination is + * specified in bits 56 to 63 of the ICR). + */ +static inline void __send_IPI_dest_field(unsigned long mask, int vector) +{ + unsigned long cfg; + + /* + * Wait for idle. + */ + if (unlikely(vector == NMI_VECTOR)) + safe_apic_wait_icr_idle(); + else + apic_wait_icr_idle(); + + /* + * prepare target chip field + */ + cfg = __prepare_ICR2(mask); + apic_write_around(APIC_ICR2, cfg); + + /* + * program the ICR + */ + cfg = __prepare_ICR(0, vector); + + /* + * Send the IPI. The write to APIC_ICR fires this off. + */ + apic_write_around(APIC_ICR, cfg); +} + +/* + * This is only used on smaller machines. + */ +void send_IPI_mask_bitmask(cpumask_t cpumask, int vector) +{ + unsigned long mask = cpus_addr(cpumask)[0]; + unsigned long flags; + + local_irq_save(flags); + WARN_ON(mask & ~cpus_addr(cpu_online_map)[0]); + __send_IPI_dest_field(mask, vector); + local_irq_restore(flags); +} + +void send_IPI_mask_sequence(cpumask_t mask, int vector) +{ + unsigned long flags; + unsigned int query_cpu; + + /* + * Hack. The clustered APIC addressing mode doesn't allow us to send + * to an arbitrary mask, so I do a unicasts to each CPU instead. This + * should be modified to do 1 message per cluster ID - mbligh + */ + + local_irq_save(flags); + for_each_possible_cpu(query_cpu) { + if (cpu_isset(query_cpu, mask)) { + __send_IPI_dest_field(cpu_to_logical_apicid(query_cpu), + vector); + } + } + local_irq_restore(flags); +} + +/* must come after the send_IPI functions above for inlining */ +#include <mach_ipi.h> +static int convert_apicid_to_cpu(int apic_id) +{ + int i; + + for_each_possible_cpu(i) { + if (per_cpu(x86_cpu_to_apicid, i) == apic_id) + return i; + } + return -1; +} + +int safe_smp_processor_id(void) +{ + int apicid, cpuid; + + if (!boot_cpu_has(X86_FEATURE_APIC)) + return 0; + + apicid = hard_smp_processor_id(); + if (apicid == BAD_APICID) + return 0; + + cpuid = convert_apicid_to_cpu(apicid); + + return cpuid >= 0 ? cpuid : 0; +} +#endif diff --git a/arch/x86/kernel/irq_32.c b/arch/x86/kernel/irq_32.c index cef054b09d27..6ea67b76a214 100644 --- a/arch/x86/kernel/irq_32.c +++ b/arch/x86/kernel/irq_32.c @@ -79,7 +79,7 @@ unsigned int do_IRQ(struct pt_regs *regs) if (unlikely((unsigned)irq >= NR_IRQS)) { printk(KERN_EMERG "%s: cannot handle IRQ %d\n", - __FUNCTION__, irq); + __func__, irq); BUG(); } diff --git a/arch/x86/kernel/kprobes.c b/arch/x86/kernel/kprobes.c index 34a591283f5d..b8c6743a13da 100644 --- a/arch/x86/kernel/kprobes.c +++ b/arch/x86/kernel/kprobes.c @@ -410,13 +410,13 @@ static void __kprobes set_current_kprobe(struct kprobe *p, struct pt_regs *regs, static void __kprobes clear_btf(void) { if (test_thread_flag(TIF_DEBUGCTLMSR)) - wrmsrl(MSR_IA32_DEBUGCTLMSR, 0); + update_debugctlmsr(0); } static void __kprobes restore_btf(void) { if (test_thread_flag(TIF_DEBUGCTLMSR)) - wrmsrl(MSR_IA32_DEBUGCTLMSR, current->thread.debugctlmsr); + update_debugctlmsr(current->thread.debugctlmsr); } static void __kprobes prepare_singlestep(struct kprobe *p, struct pt_regs *regs) @@ -489,7 +489,7 @@ static int __kprobes reenter_kprobe(struct kprobe *p, struct pt_regs *regs, break; case KPROBE_HIT_SS: if (p == kprobe_running()) { - regs->flags &= ~TF_MASK; + regs->flags &= ~X86_EFLAGS_TF; regs->flags |= kcb->kprobe_saved_flags; return 0; } else { @@ -858,15 +858,15 @@ static int __kprobes post_kprobe_handler(struct pt_regs *regs) if (!cur) return 0; + resume_execution(cur, regs, kcb); + regs->flags |= kcb->kprobe_saved_flags; + trace_hardirqs_fixup_flags(regs->flags); + if ((kcb->kprobe_status != KPROBE_REENTER) && cur->post_handler) { kcb->kprobe_status = KPROBE_HIT_SSDONE; cur->post_handler(cur, regs, 0); } - resume_execution(cur, regs, kcb); - regs->flags |= kcb->kprobe_saved_flags; - trace_hardirqs_fixup_flags(regs->flags); - /* Restore back the original saved kprobes variables and continue. */ if (kcb->kprobe_status == KPROBE_REENTER) { restore_previous_kprobe(kcb); diff --git a/arch/x86/kernel/mca_32.c b/arch/x86/kernel/mca_32.c index 9482033ed0fe..2dc183758be3 100644 --- a/arch/x86/kernel/mca_32.c +++ b/arch/x86/kernel/mca_32.c @@ -53,9 +53,9 @@ #include <linux/init.h> #include <asm/arch_hooks.h> -static unsigned char which_scsi = 0; +static unsigned char which_scsi; -int MCA_bus = 0; +int MCA_bus; EXPORT_SYMBOL(MCA_bus); /* @@ -68,15 +68,17 @@ static DEFINE_SPINLOCK(mca_lock); /* Build the status info for the adapter */ -static void mca_configure_adapter_status(struct mca_device *mca_dev) { +static void mca_configure_adapter_status(struct mca_device *mca_dev) +{ mca_dev->status = MCA_ADAPTER_NONE; mca_dev->pos_id = mca_dev->pos[0] + (mca_dev->pos[1] << 8); - if(!mca_dev->pos_id && mca_dev->slot < MCA_MAX_SLOT_NR) { + if (!mca_dev->pos_id && mca_dev->slot < MCA_MAX_SLOT_NR) { - /* id = 0x0000 usually indicates hardware failure, + /* + * id = 0x0000 usually indicates hardware failure, * however, ZP Gu (zpg@castle.net> reports that his 9556 * has 0x0000 as id and everything still works. There * also seem to be an adapter with id = 0x0000; the @@ -87,9 +89,10 @@ static void mca_configure_adapter_status(struct mca_device *mca_dev) { mca_dev->status = MCA_ADAPTER_ERROR; return; - } else if(mca_dev->pos_id != 0xffff) { + } else if (mca_dev->pos_id != 0xffff) { - /* 0xffff usually indicates that there's no adapter, + /* + * 0xffff usually indicates that there's no adapter, * however, some integrated adapters may have 0xffff as * their id and still be valid. Examples are on-board * VGA of the 55sx, the integrated SCSI of the 56 & 57, @@ -99,19 +102,19 @@ static void mca_configure_adapter_status(struct mca_device *mca_dev) { mca_dev->status = MCA_ADAPTER_NORMAL; } - if((mca_dev->pos_id == 0xffff || + if ((mca_dev->pos_id == 0xffff || mca_dev->pos_id == 0x0000) && mca_dev->slot >= MCA_MAX_SLOT_NR) { int j; - for(j = 2; j < 8; j++) { - if(mca_dev->pos[j] != 0xff) { + for (j = 2; j < 8; j++) { + if (mca_dev->pos[j] != 0xff) { mca_dev->status = MCA_ADAPTER_NORMAL; break; } } } - if(!(mca_dev->pos[2] & MCA_ENABLED)) { + if (!(mca_dev->pos[2] & MCA_ENABLED)) { /* enabled bit is in POS 2 */ @@ -133,7 +136,7 @@ static struct resource mca_standard_resources[] = { #define MCA_STANDARD_RESOURCES ARRAY_SIZE(mca_standard_resources) -/** +/* * mca_read_and_store_pos - read the POS registers into a memory buffer * @pos: a char pointer to 8 bytes, contains the POS register value on * successful return @@ -141,12 +144,14 @@ static struct resource mca_standard_resources[] = { * Returns 1 if a card actually exists (i.e. the pos isn't * all 0xff) or 0 otherwise */ -static int mca_read_and_store_pos(unsigned char *pos) { +static int mca_read_and_store_pos(unsigned char *pos) +{ int j; int found = 0; - for(j=0; j<8; j++) { - if((pos[j] = inb_p(MCA_POS_REG(j))) != 0xff) { + for (j = 0; j < 8; j++) { + pos[j] = inb_p(MCA_POS_REG(j)); + if (pos[j] != 0xff) { /* 0xff all across means no device. 0x00 means * something's broken, but a device is * probably there. However, if you get 0x00 @@ -167,11 +172,11 @@ static unsigned char mca_pc_read_pos(struct mca_device *mca_dev, int reg) unsigned char byte; unsigned long flags; - if(reg < 0 || reg >= 8) + if (reg < 0 || reg >= 8) return 0; spin_lock_irqsave(&mca_lock, flags); - if(mca_dev->pos_register) { + if (mca_dev->pos_register) { /* Disable adapter setup, enable motherboard setup */ outb_p(0, MCA_ADAPTER_SETUP_REG); @@ -203,7 +208,7 @@ static void mca_pc_write_pos(struct mca_device *mca_dev, int reg, { unsigned long flags; - if(reg < 0 || reg >= 8) + if (reg < 0 || reg >= 8) return; spin_lock_irqsave(&mca_lock, flags); @@ -227,17 +232,17 @@ static void mca_pc_write_pos(struct mca_device *mca_dev, int reg, } /* for the primary MCA bus, we have identity transforms */ -static int mca_dummy_transform_irq(struct mca_device * mca_dev, int irq) +static int mca_dummy_transform_irq(struct mca_device *mca_dev, int irq) { return irq; } -static int mca_dummy_transform_ioport(struct mca_device * mca_dev, int port) +static int mca_dummy_transform_ioport(struct mca_device *mca_dev, int port) { return port; } -static void *mca_dummy_transform_memory(struct mca_device * mca_dev, void *mem) +static void *mca_dummy_transform_memory(struct mca_device *mca_dev, void *mem) { return mem; } @@ -251,7 +256,8 @@ static int __init mca_init(void) short mca_builtin_scsi_ports[] = {0xf7, 0xfd, 0x00}; struct mca_bus *bus; - /* WARNING: Be careful when making changes here. Putting an adapter + /* + * WARNING: Be careful when making changes here. Putting an adapter * and the motherboard simultaneously into setup mode may result in * damage to chips (according to The Indispensible PC Hardware Book * by Hans-Peter Messmer). Also, we disable system interrupts (so @@ -283,7 +289,7 @@ static int __init mca_init(void) /* get the motherboard device */ mca_dev = kzalloc(sizeof(struct mca_device), GFP_KERNEL); - if(unlikely(!mca_dev)) + if (unlikely(!mca_dev)) goto out_nomem; /* @@ -309,7 +315,7 @@ static int __init mca_init(void) mca_register_device(MCA_PRIMARY_BUS, mca_dev); mca_dev = kzalloc(sizeof(struct mca_device), GFP_ATOMIC); - if(unlikely(!mca_dev)) + if (unlikely(!mca_dev)) goto out_unlock_nomem; /* Put motherboard into video setup mode, read integrated video @@ -326,7 +332,8 @@ static int __init mca_init(void) mca_dev->slot = MCA_INTEGVIDEO; mca_register_device(MCA_PRIMARY_BUS, mca_dev); - /* Put motherboard into scsi setup mode, read integrated scsi + /* + * Put motherboard into scsi setup mode, read integrated scsi * POS registers, and turn motherboard setup off. * * It seems there are two possible SCSI registers. Martin says that @@ -338,18 +345,18 @@ static int __init mca_init(void) * machine. */ - for(i = 0; (which_scsi = mca_builtin_scsi_ports[i]) != 0; i++) { + for (i = 0; (which_scsi = mca_builtin_scsi_ports[i]) != 0; i++) { outb_p(which_scsi, MCA_MOTHERBOARD_SETUP_REG); - if(mca_read_and_store_pos(pos)) + if (mca_read_and_store_pos(pos)) break; } - if(which_scsi) { + if (which_scsi) { /* found a scsi card */ mca_dev = kzalloc(sizeof(struct mca_device), GFP_ATOMIC); - if(unlikely(!mca_dev)) + if (unlikely(!mca_dev)) goto out_unlock_nomem; - for(j = 0; j < 8; j++) + for (j = 0; j < 8; j++) mca_dev->pos[j] = pos[j]; mca_configure_adapter_status(mca_dev); @@ -364,21 +371,22 @@ static int __init mca_init(void) outb_p(0xff, MCA_MOTHERBOARD_SETUP_REG); - /* Now loop over MCA slots: put each adapter into setup mode, and + /* + * Now loop over MCA slots: put each adapter into setup mode, and * read its POS registers. Then put adapter setup off. */ - for(i=0; i<MCA_MAX_SLOT_NR; i++) { + for (i = 0; i < MCA_MAX_SLOT_NR; i++) { outb_p(0x8|(i&0xf), MCA_ADAPTER_SETUP_REG); - if(!mca_read_and_store_pos(pos)) + if (!mca_read_and_store_pos(pos)) continue; mca_dev = kzalloc(sizeof(struct mca_device), GFP_ATOMIC); - if(unlikely(!mca_dev)) + if (unlikely(!mca_dev)) goto out_unlock_nomem; - for(j=0; j<8; j++) - mca_dev->pos[j]=pos[j]; + for (j = 0; j < 8; j++) + mca_dev->pos[j] = pos[j]; mca_dev->driver_loaded = 0; mca_dev->slot = i; @@ -414,20 +422,20 @@ mca_handle_nmi_device(struct mca_device *mca_dev, int check_flag) { int slot = mca_dev->slot; - if(slot == MCA_INTEGSCSI) { + if (slot == MCA_INTEGSCSI) { printk(KERN_CRIT "NMI: caused by MCA integrated SCSI adapter (%s)\n", mca_dev->name); - } else if(slot == MCA_INTEGVIDEO) { + } else if (slot == MCA_INTEGVIDEO) { printk(KERN_CRIT "NMI: caused by MCA integrated video adapter (%s)\n", mca_dev->name); - } else if(slot == MCA_MOTHERBOARD) { + } else if (slot == MCA_MOTHERBOARD) { printk(KERN_CRIT "NMI: caused by motherboard (%s)\n", mca_dev->name); } /* More info available in POS 6 and 7? */ - if(check_flag) { + if (check_flag) { unsigned char pos6, pos7; pos6 = mca_device_read_pos(mca_dev, 6); @@ -447,8 +455,9 @@ static int __kprobes mca_handle_nmi_callback(struct device *dev, void *data) pos5 = mca_device_read_pos(mca_dev, 5); - if(!(pos5 & 0x80)) { - /* Bit 7 of POS 5 is reset when this adapter has a hardware + if (!(pos5 & 0x80)) { + /* + * Bit 7 of POS 5 is reset when this adapter has a hardware * error. Bit 7 it reset if there's error information * available in POS 6 and 7. */ @@ -460,7 +469,8 @@ static int __kprobes mca_handle_nmi_callback(struct device *dev, void *data) void __kprobes mca_handle_nmi(void) { - /* First try - scan the various adapters and see if a specific + /* + * First try - scan the various adapters and see if a specific * adapter was responsible for the error. */ bus_for_each_dev(&mca_bus_type, NULL, NULL, mca_handle_nmi_callback); diff --git a/arch/x86/kernel/microcode.c b/arch/x86/kernel/microcode.c index f2702d01b8a8..25cf6dee4e56 100644 --- a/arch/x86/kernel/microcode.c +++ b/arch/x86/kernel/microcode.c @@ -290,7 +290,7 @@ static int get_maching_microcode(void *mc, int cpu) } return 0; find: - pr_debug("microcode: CPU %d found a matching microcode update with" + pr_debug("microcode: CPU%d found a matching microcode update with" " version 0x%x (current=0x%x)\n", cpu, mc_header->rev,uci->rev); new_mc = vmalloc(total_size); if (!new_mc) { @@ -336,11 +336,11 @@ static void apply_microcode(int cpu) spin_unlock_irqrestore(µcode_update_lock, flags); if (val[1] != uci->mc->hdr.rev) { - printk(KERN_ERR "microcode: CPU%d updated from revision " + printk(KERN_ERR "microcode: CPU%d update from revision " "0x%x to 0x%x failed\n", cpu_num, uci->rev, val[1]); return; } - pr_debug("microcode: CPU%d updated from revision " + printk(KERN_INFO "microcode: CPU%d updated from revision " "0x%x to 0x%x, date = %08x \n", cpu_num, uci->rev, val[1], uci->mc->hdr.date); uci->rev = val[1]; @@ -534,7 +534,7 @@ static int cpu_request_microcode(int cpu) c->x86, c->x86_model, c->x86_mask); error = request_firmware(&firmware, name, µcode_pdev->dev); if (error) { - pr_debug("ucode data file %s load failed\n", name); + pr_debug("microcode: ucode data file %s load failed\n", name); return error; } buf = firmware->data; @@ -709,7 +709,7 @@ static int __mc_sysdev_add(struct sys_device *sys_dev, int resume) if (!cpu_online(cpu)) return 0; - pr_debug("Microcode:CPU %d added\n", cpu); + pr_debug("microcode: CPU%d added\n", cpu); memset(uci, 0, sizeof(*uci)); err = sysfs_create_group(&sys_dev->kobj, &mc_attr_group); @@ -733,7 +733,7 @@ static int mc_sysdev_remove(struct sys_device *sys_dev) if (!cpu_online(cpu)) return 0; - pr_debug("Microcode:CPU %d removed\n", cpu); + pr_debug("microcode: CPU%d removed\n", cpu); microcode_fini_cpu(cpu); sysfs_remove_group(&sys_dev->kobj, &mc_attr_group); return 0; @@ -745,7 +745,7 @@ static int mc_sysdev_resume(struct sys_device *dev) if (!cpu_online(cpu)) return 0; - pr_debug("Microcode:CPU %d resumed\n", cpu); + pr_debug("microcode: CPU%d resumed\n", cpu); /* only CPU 0 will apply ucode here */ apply_microcode(0); return 0; @@ -783,7 +783,7 @@ mc_cpu_callback(struct notifier_block *nb, unsigned long action, void *hcpu) } case CPU_DOWN_FAILED_FROZEN: if (sysfs_create_group(&sys_dev->kobj, &mc_attr_group)) - printk(KERN_ERR "Microcode: Failed to create the sysfs " + printk(KERN_ERR "microcode: Failed to create the sysfs " "group for CPU%d\n", cpu); break; case CPU_DOWN_PREPARE: diff --git a/arch/x86/kernel/mpparse_32.c b/arch/x86/kernel/mpparse.c index f349e68e45a0..70744e344fa1 100644 --- a/arch/x86/kernel/mpparse_32.c +++ b/arch/x86/kernel/mpparse.c @@ -4,82 +4,56 @@ * * (c) 1995 Alan Cox, Building #3 <alan@redhat.com> * (c) 1998, 1999, 2000 Ingo Molnar <mingo@redhat.com> - * - * Fixes - * Erich Boleyn : MP v1.4 and additional changes. - * Alan Cox : Added EBDA scanning - * Ingo Molnar : various cleanups and rewrites - * Maciej W. Rozycki: Bits for default MP configurations - * Paul Diefenbaugh: Added full ACPI support + * (c) 2008 Alexey Starikovskiy <astarikovskiy@suse.de> */ #include <linux/mm.h> #include <linux/init.h> -#include <linux/acpi.h> #include <linux/delay.h> #include <linux/bootmem.h> #include <linux/kernel_stat.h> #include <linux/mc146818rtc.h> #include <linux/bitops.h> +#include <linux/acpi.h> +#include <linux/module.h> #include <asm/smp.h> -#include <asm/acpi.h> #include <asm/mtrr.h> #include <asm/mpspec.h> +#include <asm/pgalloc.h> #include <asm/io_apic.h> +#include <asm/proto.h> +#include <asm/acpi.h> +#include <asm/bios_ebda.h> #include <mach_apic.h> +#ifdef CONFIG_X86_32 #include <mach_apicdef.h> #include <mach_mpparse.h> -#include <bios_ebda.h> +#endif /* Have we found an MP table */ int smp_found_config; -unsigned int __cpuinitdata maxcpus = NR_CPUS; /* * Various Linux-internal data structures created from the * MP-table. */ -int apic_version [MAX_APICS]; -int mp_bus_id_to_type [MAX_MP_BUSSES]; -int mp_bus_id_to_node [MAX_MP_BUSSES]; -int mp_bus_id_to_local [MAX_MP_BUSSES]; -int quad_local_to_mp_bus_id [NR_CPUS/4][4]; -int mp_bus_id_to_pci_bus [MAX_MP_BUSSES] = { [0 ... MAX_MP_BUSSES-1] = -1 }; -static int mp_current_pci_id; - -/* I/O APIC entries */ -struct mpc_config_ioapic mp_ioapics[MAX_IO_APICS]; - -/* # of MP IRQ source entries */ -struct mpc_config_intsrc mp_irqs[MAX_IRQ_SOURCES]; +#if defined (CONFIG_MCA) || defined (CONFIG_EISA) +int mp_bus_id_to_type[MAX_MP_BUSSES]; +#endif -/* MP IRQ source entries */ -int mp_irq_entries; +DECLARE_BITMAP(mp_bus_not_pci, MAX_MP_BUSSES); +int mp_bus_id_to_pci_bus[MAX_MP_BUSSES] = {[0 ... MAX_MP_BUSSES - 1] = -1 }; -int nr_ioapics; +static int mp_current_pci_id; int pic_mode; -unsigned long mp_lapic_addr; - -unsigned int def_to_bigsmp = 0; - -/* Processor that is doing the boot up */ -unsigned int boot_cpu_physical_apicid = -1U; -/* Internal processor count */ -unsigned int num_processors; - -/* Bitmask of physically existing CPUs */ -physid_mask_t phys_cpu_present_map; - -u8 bios_cpu_apicid[NR_CPUS] = { [0 ... NR_CPUS-1] = BAD_APICID }; /* * Intel MP BIOS table parsing routines: */ - /* * Checksum an MP configuration block. */ @@ -94,216 +68,153 @@ static int __init mpf_checksum(unsigned char *mp, int len) return sum & 0xFF; } +#ifdef CONFIG_X86_NUMAQ /* * Have to match translation table entries to main table entries by counter * hence the mpc_record variable .... can't see a less disgusting way of * doing this .... */ -static int mpc_record; -static struct mpc_config_translation *translation_table[MAX_MPC_ENTRY] __cpuinitdata; +static int mpc_record; +static struct mpc_config_translation *translation_table[MAX_MPC_ENTRY] + __cpuinitdata; +#endif -static void __cpuinit MP_processor_info (struct mpc_config_processor *m) +static void __cpuinit MP_processor_info(struct mpc_config_processor *m) { - int ver, apicid; - physid_mask_t phys_cpu; - - if (!(m->mpc_cpuflag & CPU_ENABLED)) - return; + int apicid; + char *bootup_cpu = ""; + if (!(m->mpc_cpuflag & CPU_ENABLED)) { + disabled_cpus++; + return; + } +#ifdef CONFIG_X86_NUMAQ apicid = mpc_apic_id(m, translation_table[mpc_record]); - - if (m->mpc_featureflag&(1<<0)) - Dprintk(" Floating point unit present.\n"); - if (m->mpc_featureflag&(1<<7)) - Dprintk(" Machine Exception supported.\n"); - if (m->mpc_featureflag&(1<<8)) - Dprintk(" 64 bit compare & exchange supported.\n"); - if (m->mpc_featureflag&(1<<9)) - Dprintk(" Internal APIC present.\n"); - if (m->mpc_featureflag&(1<<11)) - Dprintk(" SEP present.\n"); - if (m->mpc_featureflag&(1<<12)) - Dprintk(" MTRR present.\n"); - if (m->mpc_featureflag&(1<<13)) - Dprintk(" PGE present.\n"); - if (m->mpc_featureflag&(1<<14)) - Dprintk(" MCA present.\n"); - if (m->mpc_featureflag&(1<<15)) - Dprintk(" CMOV present.\n"); - if (m->mpc_featureflag&(1<<16)) - Dprintk(" PAT present.\n"); - if (m->mpc_featureflag&(1<<17)) - Dprintk(" PSE present.\n"); - if (m->mpc_featureflag&(1<<18)) - Dprintk(" PSN present.\n"); - if (m->mpc_featureflag&(1<<19)) - Dprintk(" Cache Line Flush Instruction present.\n"); - /* 20 Reserved */ - if (m->mpc_featureflag&(1<<21)) - Dprintk(" Debug Trace and EMON Store present.\n"); - if (m->mpc_featureflag&(1<<22)) - Dprintk(" ACPI Thermal Throttle Registers present.\n"); - if (m->mpc_featureflag&(1<<23)) - Dprintk(" MMX present.\n"); - if (m->mpc_featureflag&(1<<24)) - Dprintk(" FXSR present.\n"); - if (m->mpc_featureflag&(1<<25)) - Dprintk(" XMM present.\n"); - if (m->mpc_featureflag&(1<<26)) - Dprintk(" Willamette New Instructions present.\n"); - if (m->mpc_featureflag&(1<<27)) - Dprintk(" Self Snoop present.\n"); - if (m->mpc_featureflag&(1<<28)) - Dprintk(" HT present.\n"); - if (m->mpc_featureflag&(1<<29)) - Dprintk(" Thermal Monitor present.\n"); - /* 30, 31 Reserved */ - - +#else + apicid = m->mpc_apicid; +#endif if (m->mpc_cpuflag & CPU_BOOTPROCESSOR) { - Dprintk(" Bootup CPU\n"); + bootup_cpu = " (Bootup-CPU)"; boot_cpu_physical_apicid = m->mpc_apicid; } - ver = m->mpc_apicver; - - /* - * Validate version - */ - if (ver == 0x0) { - printk(KERN_WARNING "BIOS bug, APIC version is 0 for CPU#%d! " - "fixing up to 0x10. (tell your hw vendor)\n", - m->mpc_apicid); - ver = 0x10; - } - apic_version[m->mpc_apicid] = ver; - - phys_cpu = apicid_to_cpu_present(apicid); - physids_or(phys_cpu_present_map, phys_cpu_present_map, phys_cpu); - - if (num_processors >= NR_CPUS) { - printk(KERN_WARNING "WARNING: NR_CPUS limit of %i reached." - " Processor ignored.\n", NR_CPUS); - return; - } - - if (num_processors >= maxcpus) { - printk(KERN_WARNING "WARNING: maxcpus limit of %i reached." - " Processor ignored.\n", maxcpus); - return; - } - - cpu_set(num_processors, cpu_possible_map); - num_processors++; - - /* - * Would be preferable to switch to bigsmp when CONFIG_HOTPLUG_CPU=y - * but we need to work other dependencies like SMP_SUSPEND etc - * before this can be done without some confusion. - * if (CPU_HOTPLUG_ENABLED || num_processors > 8) - * - Ashok Raj <ashok.raj@intel.com> - */ - if (num_processors > 8) { - switch (boot_cpu_data.x86_vendor) { - case X86_VENDOR_INTEL: - if (!APIC_XAPIC(ver)) { - def_to_bigsmp = 0; - break; - } - /* If P4 and above fall through */ - case X86_VENDOR_AMD: - def_to_bigsmp = 1; - } - } - bios_cpu_apicid[num_processors - 1] = m->mpc_apicid; + printk(KERN_INFO "Processor #%d%s\n", m->mpc_apicid, bootup_cpu); + generic_processor_info(apicid, m->mpc_apicver); } -static void __init MP_bus_info (struct mpc_config_bus *m) +static void __init MP_bus_info(struct mpc_config_bus *m) { char str[7]; memcpy(str, m->mpc_bustype, 6); str[6] = 0; +#ifdef CONFIG_X86_NUMAQ mpc_oem_bus_info(m, str, translation_table[mpc_record]); +#else + Dprintk("Bus #%d is %s\n", m->mpc_busid, str); +#endif #if MAX_MP_BUSSES < 256 if (m->mpc_busid >= MAX_MP_BUSSES) { printk(KERN_WARNING "MP table busid value (%d) for bustype %s " - " is too large, max. supported is %d\n", - m->mpc_busid, str, MAX_MP_BUSSES - 1); + " is too large, max. supported is %d\n", + m->mpc_busid, str, MAX_MP_BUSSES - 1); return; } #endif - if (strncmp(str, BUSTYPE_ISA, sizeof(BUSTYPE_ISA)-1) == 0) { + if (strncmp(str, BUSTYPE_ISA, sizeof(BUSTYPE_ISA) - 1) == 0) { + set_bit(m->mpc_busid, mp_bus_not_pci); +#if defined(CONFIG_EISA) || defined (CONFIG_MCA) mp_bus_id_to_type[m->mpc_busid] = MP_BUS_ISA; - } else if (strncmp(str, BUSTYPE_EISA, sizeof(BUSTYPE_EISA)-1) == 0) { - mp_bus_id_to_type[m->mpc_busid] = MP_BUS_EISA; - } else if (strncmp(str, BUSTYPE_PCI, sizeof(BUSTYPE_PCI)-1) == 0) { +#endif + } else if (strncmp(str, BUSTYPE_PCI, sizeof(BUSTYPE_PCI) - 1) == 0) { +#ifdef CONFIG_X86_NUMAQ mpc_oem_pci_bus(m, translation_table[mpc_record]); - mp_bus_id_to_type[m->mpc_busid] = MP_BUS_PCI; +#endif + clear_bit(m->mpc_busid, mp_bus_not_pci); mp_bus_id_to_pci_bus[m->mpc_busid] = mp_current_pci_id; mp_current_pci_id++; - } else if (strncmp(str, BUSTYPE_MCA, sizeof(BUSTYPE_MCA)-1) == 0) { +#if defined(CONFIG_EISA) || defined (CONFIG_MCA) + mp_bus_id_to_type[m->mpc_busid] = MP_BUS_PCI; + } else if (strncmp(str, BUSTYPE_EISA, sizeof(BUSTYPE_EISA) - 1) == 0) { + mp_bus_id_to_type[m->mpc_busid] = MP_BUS_EISA; + } else if (strncmp(str, BUSTYPE_MCA, sizeof(BUSTYPE_MCA) - 1) == 0) { mp_bus_id_to_type[m->mpc_busid] = MP_BUS_MCA; - } else { +#endif + } else printk(KERN_WARNING "Unknown bustype %s - ignoring\n", str); +} + +#ifdef CONFIG_X86_IO_APIC + +static int bad_ioapic(unsigned long address) +{ + if (nr_ioapics >= MAX_IO_APICS) { + printk(KERN_ERR "ERROR: Max # of I/O APICs (%d) exceeded " + "(found %d)\n", MAX_IO_APICS, nr_ioapics); + panic("Recompile kernel with bigger MAX_IO_APICS!\n"); + } + if (!address) { + printk(KERN_ERR "WARNING: Bogus (zero) I/O APIC address" + " found in table, skipping!\n"); + return 1; } + return 0; } -static void __init MP_ioapic_info (struct mpc_config_ioapic *m) +static void __init MP_ioapic_info(struct mpc_config_ioapic *m) { if (!(m->mpc_flags & MPC_APIC_USABLE)) return; printk(KERN_INFO "I/O APIC #%d Version %d at 0x%X.\n", - m->mpc_apicid, m->mpc_apicver, m->mpc_apicaddr); - if (nr_ioapics >= MAX_IO_APICS) { - printk(KERN_CRIT "Max # of I/O APICs (%d) exceeded (found %d).\n", - MAX_IO_APICS, nr_ioapics); - panic("Recompile kernel with bigger MAX_IO_APICS!.\n"); - } - if (!m->mpc_apicaddr) { - printk(KERN_ERR "WARNING: bogus zero I/O APIC address" - " found in MP table, skipping!\n"); + m->mpc_apicid, m->mpc_apicver, m->mpc_apicaddr); + + if (bad_ioapic(m->mpc_apicaddr)) return; - } + mp_ioapics[nr_ioapics] = *m; nr_ioapics++; } -static void __init MP_intsrc_info (struct mpc_config_intsrc *m) +static void __init MP_intsrc_info(struct mpc_config_intsrc *m) { - mp_irqs [mp_irq_entries] = *m; + mp_irqs[mp_irq_entries] = *m; Dprintk("Int: type %d, pol %d, trig %d, bus %d," " IRQ %02x, APIC ID %x, APIC INT %02x\n", - m->mpc_irqtype, m->mpc_irqflag & 3, - (m->mpc_irqflag >> 2) & 3, m->mpc_srcbus, - m->mpc_srcbusirq, m->mpc_dstapic, m->mpc_dstirq); + m->mpc_irqtype, m->mpc_irqflag & 3, + (m->mpc_irqflag >> 2) & 3, m->mpc_srcbus, + m->mpc_srcbusirq, m->mpc_dstapic, m->mpc_dstirq); if (++mp_irq_entries == MAX_IRQ_SOURCES) panic("Max # of irq sources exceeded!!\n"); } -static void __init MP_lintsrc_info (struct mpc_config_lintsrc *m) +#endif + +static void __init MP_lintsrc_info(struct mpc_config_lintsrc *m) { Dprintk("Lint: type %d, pol %d, trig %d, bus %d," " IRQ %02x, APIC ID %x, APIC LINT %02x\n", - m->mpc_irqtype, m->mpc_irqflag & 3, - (m->mpc_irqflag >> 2) &3, m->mpc_srcbusid, - m->mpc_srcbusirq, m->mpc_destapic, m->mpc_destapiclint); + m->mpc_irqtype, m->mpc_irqflag & 3, + (m->mpc_irqflag >> 2) & 3, m->mpc_srcbusid, + m->mpc_srcbusirq, m->mpc_destapic, m->mpc_destapiclint); } #ifdef CONFIG_X86_NUMAQ -static void __init MP_translation_info (struct mpc_config_translation *m) +static void __init MP_translation_info(struct mpc_config_translation *m) { - printk(KERN_INFO "Translation: record %d, type %d, quad %d, global %d, local %d\n", mpc_record, m->trans_type, m->trans_quad, m->trans_global, m->trans_local); + printk(KERN_INFO + "Translation: record %d, type %d, quad %d, global %d, local %d\n", + mpc_record, m->trans_type, m->trans_quad, m->trans_global, + m->trans_local); - if (mpc_record >= MAX_MPC_ENTRY) + if (mpc_record >= MAX_MPC_ENTRY) printk(KERN_ERR "MAX_MPC_ENTRY exceeded!\n"); else - translation_table[mpc_record] = m; /* stash this for later */ + translation_table[mpc_record] = m; /* stash this for later */ if (m->trans_quad < MAX_NUMNODES && !node_online(m->trans_quad)) node_set_online(m->trans_quad); } @@ -312,118 +223,124 @@ static void __init MP_translation_info (struct mpc_config_translation *m) * Read/parse the MPC oem tables */ -static void __init smp_read_mpc_oem(struct mp_config_oemtable *oemtable, \ - unsigned short oemsize) +static void __init smp_read_mpc_oem(struct mp_config_oemtable *oemtable, + unsigned short oemsize) { - int count = sizeof (*oemtable); /* the header size */ - unsigned char *oemptr = ((unsigned char *)oemtable)+count; - + int count = sizeof(*oemtable); /* the header size */ + unsigned char *oemptr = ((unsigned char *)oemtable) + count; + mpc_record = 0; - printk(KERN_INFO "Found an OEM MPC table at %8p - parsing it ... \n", oemtable); - if (memcmp(oemtable->oem_signature,MPC_OEM_SIGNATURE,4)) - { - printk(KERN_WARNING "SMP mpc oemtable: bad signature [%c%c%c%c]!\n", - oemtable->oem_signature[0], - oemtable->oem_signature[1], - oemtable->oem_signature[2], - oemtable->oem_signature[3]); + printk(KERN_INFO "Found an OEM MPC table at %8p - parsing it ... \n", + oemtable); + if (memcmp(oemtable->oem_signature, MPC_OEM_SIGNATURE, 4)) { + printk(KERN_WARNING + "SMP mpc oemtable: bad signature [%c%c%c%c]!\n", + oemtable->oem_signature[0], oemtable->oem_signature[1], + oemtable->oem_signature[2], oemtable->oem_signature[3]); return; } - if (mpf_checksum((unsigned char *)oemtable,oemtable->oem_length)) - { + if (mpf_checksum((unsigned char *)oemtable, oemtable->oem_length)) { printk(KERN_WARNING "SMP oem mptable: checksum error!\n"); return; } while (count < oemtable->oem_length) { switch (*oemptr) { - case MP_TRANSLATION: + case MP_TRANSLATION: { - struct mpc_config_translation *m= - (struct mpc_config_translation *)oemptr; + struct mpc_config_translation *m = + (struct mpc_config_translation *)oemptr; MP_translation_info(m); oemptr += sizeof(*m); count += sizeof(*m); ++mpc_record; break; } - default: + default: { - printk(KERN_WARNING "Unrecognised OEM table entry type! - %d\n", (int) *oemptr); + printk(KERN_WARNING + "Unrecognised OEM table entry type! - %d\n", + (int)*oemptr); return; } } - } + } } static inline void mps_oem_check(struct mp_config_table *mpc, char *oem, - char *productid) + char *productid) { if (strncmp(oem, "IBM NUMA", 8)) printk("Warning! May not be a NUMA-Q system!\n"); if (mpc->mpc_oemptr) - smp_read_mpc_oem((struct mp_config_oemtable *) mpc->mpc_oemptr, - mpc->mpc_oemsize); + smp_read_mpc_oem((struct mp_config_oemtable *)mpc->mpc_oemptr, + mpc->mpc_oemsize); } -#endif /* CONFIG_X86_NUMAQ */ +#endif /* CONFIG_X86_NUMAQ */ /* * Read/parse the MPC */ -static int __init smp_read_mpc(struct mp_config_table *mpc) +static int __init smp_read_mpc(struct mp_config_table *mpc, unsigned early) { char str[16]; char oem[10]; - int count=sizeof(*mpc); - unsigned char *mpt=((unsigned char *)mpc)+count; + int count = sizeof(*mpc); + unsigned char *mpt = ((unsigned char *)mpc) + count; - if (memcmp(mpc->mpc_signature,MPC_SIGNATURE,4)) { - printk(KERN_ERR "SMP mptable: bad signature [0x%x]!\n", - *(u32 *)mpc->mpc_signature); + if (memcmp(mpc->mpc_signature, MPC_SIGNATURE, 4)) { + printk(KERN_ERR "MPTABLE: bad signature [%c%c%c%c]!\n", + mpc->mpc_signature[0], mpc->mpc_signature[1], + mpc->mpc_signature[2], mpc->mpc_signature[3]); return 0; } - if (mpf_checksum((unsigned char *)mpc,mpc->mpc_length)) { - printk(KERN_ERR "SMP mptable: checksum error!\n"); + if (mpf_checksum((unsigned char *)mpc, mpc->mpc_length)) { + printk(KERN_ERR "MPTABLE: checksum error!\n"); return 0; } - if (mpc->mpc_spec!=0x01 && mpc->mpc_spec!=0x04) { - printk(KERN_ERR "SMP mptable: bad table version (%d)!!\n", - mpc->mpc_spec); + if (mpc->mpc_spec != 0x01 && mpc->mpc_spec != 0x04) { + printk(KERN_ERR "MPTABLE: bad table version (%d)!!\n", + mpc->mpc_spec); return 0; } if (!mpc->mpc_lapic) { - printk(KERN_ERR "SMP mptable: null local APIC address!\n"); + printk(KERN_ERR "MPTABLE: null local APIC address!\n"); return 0; } - memcpy(oem,mpc->mpc_oem,8); - oem[8]=0; - printk(KERN_INFO "OEM ID: %s ",oem); + memcpy(oem, mpc->mpc_oem, 8); + oem[8] = 0; + printk(KERN_INFO "MPTABLE: OEM ID: %s ", oem); - memcpy(str,mpc->mpc_productid,12); - str[12]=0; - printk("Product ID: %s ",str); + memcpy(str, mpc->mpc_productid, 12); + str[12] = 0; + printk("Product ID: %s ", str); +#ifdef CONFIG_X86_32 mps_oem_check(mpc, oem, str); +#endif + printk(KERN_INFO "MPTABLE: Product ID: %s ", str); - printk("APIC at: 0x%X\n", mpc->mpc_lapic); + printk(KERN_INFO "MPTABLE: APIC at: 0x%X\n", mpc->mpc_lapic); - /* - * Save the local APIC address (it might be non-default) -- but only - * if we're not using ACPI. - */ + /* save the local APIC address, it might be non-default */ if (!acpi_lapic) mp_lapic_addr = mpc->mpc_lapic; + if (early) + return 1; + /* - * Now process the configuration blocks. + * Now process the configuration blocks. */ +#ifdef CONFIG_X86_NUMAQ mpc_record = 0; +#endif while (count < mpc->mpc_length) { - switch(*mpt) { - case MP_PROCESSOR: + switch (*mpt) { + case MP_PROCESSOR: { - struct mpc_config_processor *m= - (struct mpc_config_processor *)mpt; + struct mpc_config_processor *m = + (struct mpc_config_processor *)mpt; /* ACPI may have already provided this data */ if (!acpi_lapic) MP_processor_info(m); @@ -431,57 +348,68 @@ static int __init smp_read_mpc(struct mp_config_table *mpc) count += sizeof(*m); break; } - case MP_BUS: + case MP_BUS: { - struct mpc_config_bus *m= - (struct mpc_config_bus *)mpt; + struct mpc_config_bus *m = + (struct mpc_config_bus *)mpt; MP_bus_info(m); mpt += sizeof(*m); count += sizeof(*m); break; } - case MP_IOAPIC: + case MP_IOAPIC: { - struct mpc_config_ioapic *m= - (struct mpc_config_ioapic *)mpt; +#ifdef CONFIG_X86_IO_APIC + struct mpc_config_ioapic *m = + (struct mpc_config_ioapic *)mpt; MP_ioapic_info(m); - mpt+=sizeof(*m); - count+=sizeof(*m); +#endif + mpt += sizeof(struct mpc_config_ioapic); + count += sizeof(struct mpc_config_ioapic); break; } - case MP_INTSRC: + case MP_INTSRC: { - struct mpc_config_intsrc *m= - (struct mpc_config_intsrc *)mpt; +#ifdef CONFIG_X86_IO_APIC + struct mpc_config_intsrc *m = + (struct mpc_config_intsrc *)mpt; MP_intsrc_info(m); - mpt+=sizeof(*m); - count+=sizeof(*m); +#endif + mpt += sizeof(struct mpc_config_intsrc); + count += sizeof(struct mpc_config_intsrc); break; } - case MP_LINTSRC: + case MP_LINTSRC: { - struct mpc_config_lintsrc *m= - (struct mpc_config_lintsrc *)mpt; + struct mpc_config_lintsrc *m = + (struct mpc_config_lintsrc *)mpt; MP_lintsrc_info(m); - mpt+=sizeof(*m); - count+=sizeof(*m); - break; - } - default: - { - count = mpc->mpc_length; + mpt += sizeof(*m); + count += sizeof(*m); break; } + default: + /* wrong mptable */ + printk(KERN_ERR "Your mptable is wrong, contact your HW vendor!\n"); + printk(KERN_ERR "type %x\n", *mpt); + print_hex_dump(KERN_ERR, " ", DUMP_PREFIX_ADDRESS, 16, + 1, mpc, mpc->mpc_length, 1); + count = mpc->mpc_length; + break; } +#ifdef CONFIG_X86_NUMAQ ++mpc_record; +#endif } setup_apic_routing(); if (!num_processors) - printk(KERN_ERR "SMP mptable: no processors registered!\n"); + printk(KERN_ERR "MPTABLE: no processors registered!\n"); return num_processors; } +#ifdef CONFIG_X86_IO_APIC + static int __init ELCR_trigger(unsigned int irq) { unsigned int port; @@ -497,7 +425,7 @@ static void __init construct_default_ioirq_mptable(int mpc_default_type) int ELCR_fallback = 0; intsrc.mpc_type = MP_INTSRC; - intsrc.mpc_irqflag = 0; /* conforming */ + intsrc.mpc_irqflag = 0; /* conforming */ intsrc.mpc_srcbus = 0; intsrc.mpc_dstapic = mp_ioapics[0].mpc_apicid; @@ -512,12 +440,16 @@ static void __init construct_default_ioirq_mptable(int mpc_default_type) * If it does, we assume it's valid. */ if (mpc_default_type == 5) { - printk(KERN_INFO "ISA/PCI bus type with no IRQ information... falling back to ELCR\n"); + printk(KERN_INFO "ISA/PCI bus type with no IRQ information... " + "falling back to ELCR\n"); - if (ELCR_trigger(0) || ELCR_trigger(1) || ELCR_trigger(2) || ELCR_trigger(13)) - printk(KERN_WARNING "ELCR contains invalid data... not using ELCR\n"); + if (ELCR_trigger(0) || ELCR_trigger(1) || ELCR_trigger(2) || + ELCR_trigger(13)) + printk(KERN_ERR "ELCR contains invalid data... " + "not using ELCR\n"); else { - printk(KERN_INFO "Using ELCR to identify PCI interrupts\n"); + printk(KERN_INFO + "Using ELCR to identify PCI interrupts\n"); ELCR_fallback = 1; } } @@ -546,21 +478,25 @@ static void __init construct_default_ioirq_mptable(int mpc_default_type) } intsrc.mpc_srcbusirq = i; - intsrc.mpc_dstirq = i ? i : 2; /* IRQ0 to INTIN2 */ + intsrc.mpc_dstirq = i ? i : 2; /* IRQ0 to INTIN2 */ MP_intsrc_info(&intsrc); } intsrc.mpc_irqtype = mp_ExtINT; intsrc.mpc_srcbusirq = 0; - intsrc.mpc_dstirq = 0; /* 8259A to INTIN0 */ + intsrc.mpc_dstirq = 0; /* 8259A to INTIN0 */ MP_intsrc_info(&intsrc); } +#endif + static inline void __init construct_default_ISA_mptable(int mpc_default_type) { struct mpc_config_processor processor; struct mpc_config_bus bus; +#ifdef CONFIG_X86_IO_APIC struct mpc_config_ioapic ioapic; +#endif struct mpc_config_lintsrc lintsrc; int linttypes[2] = { mp_ExtINT, mp_NMI }; int i; @@ -578,8 +514,7 @@ static inline void __init construct_default_ISA_mptable(int mpc_default_type) processor.mpc_apicver = mpc_default_type > 4 ? 0x10 : 0x01; processor.mpc_cpuflag = CPU_ENABLED; processor.mpc_cpufeature = (boot_cpu_data.x86 << 8) | - (boot_cpu_data.x86_model << 4) | - boot_cpu_data.x86_mask; + (boot_cpu_data.x86_model << 4) | boot_cpu_data.x86_mask; processor.mpc_featureflag = boot_cpu_data.x86_capability[0]; processor.mpc_reserved[0] = 0; processor.mpc_reserved[1] = 0; @@ -591,23 +526,22 @@ static inline void __init construct_default_ISA_mptable(int mpc_default_type) bus.mpc_type = MP_BUS; bus.mpc_busid = 0; switch (mpc_default_type) { - default: - printk("???\n"); - printk(KERN_ERR "Unknown standard configuration %d\n", - mpc_default_type); - /* fall through */ - case 1: - case 5: - memcpy(bus.mpc_bustype, "ISA ", 6); - break; - case 2: - case 6: - case 3: - memcpy(bus.mpc_bustype, "EISA ", 6); - break; - case 4: - case 7: - memcpy(bus.mpc_bustype, "MCA ", 6); + default: + printk(KERN_ERR "???\nUnknown standard configuration %d\n", + mpc_default_type); + /* fall through */ + case 1: + case 5: + memcpy(bus.mpc_bustype, "ISA ", 6); + break; + case 2: + case 6: + case 3: + memcpy(bus.mpc_bustype, "EISA ", 6); + break; + case 4: + case 7: + memcpy(bus.mpc_bustype, "MCA ", 6); } MP_bus_info(&bus); if (mpc_default_type > 4) { @@ -616,6 +550,7 @@ static inline void __init construct_default_ISA_mptable(int mpc_default_type) MP_bus_info(&bus); } +#ifdef CONFIG_X86_IO_APIC ioapic.mpc_type = MP_IOAPIC; ioapic.mpc_apicid = 2; ioapic.mpc_apicver = mpc_default_type > 4 ? 0x10 : 0x01; @@ -627,9 +562,9 @@ static inline void __init construct_default_ISA_mptable(int mpc_default_type) * We set up most of the low 16 IO-APIC pins according to MPS rules. */ construct_default_ioirq_mptable(mpc_default_type); - +#endif lintsrc.mpc_type = MP_LINTSRC; - lintsrc.mpc_irqflag = 0; /* conforming */ + lintsrc.mpc_irqflag = 0; /* conforming */ lintsrc.mpc_srcbusid = 0; lintsrc.mpc_srcbusirq = 0; lintsrc.mpc_destapic = MP_APIC_ALL; @@ -645,36 +580,49 @@ static struct intel_mp_floating *mpf_found; /* * Scan the memory blocks for an SMP configuration block. */ -void __init get_smp_config (void) +static void __init __get_smp_config(unsigned early) { struct intel_mp_floating *mpf = mpf_found; + if (acpi_lapic && early) + return; /* - * ACPI supports both logical (e.g. Hyper-Threading) and physical + * ACPI supports both logical (e.g. Hyper-Threading) and physical * processors, where MPS only supports physical. */ if (acpi_lapic && acpi_ioapic) { - printk(KERN_INFO "Using ACPI (MADT) for SMP configuration information\n"); + printk(KERN_INFO "Using ACPI (MADT) for SMP configuration " + "information\n"); return; - } - else if (acpi_lapic) - printk(KERN_INFO "Using ACPI for processor (LAPIC) configuration information\n"); - - printk(KERN_INFO "Intel MultiProcessor Specification v1.%d\n", mpf->mpf_specification); - if (mpf->mpf_feature2 & (1<<7)) { + } else if (acpi_lapic) + printk(KERN_INFO "Using ACPI for processor (LAPIC) " + "configuration information\n"); + + printk(KERN_INFO "Intel MultiProcessor Specification v1.%d\n", + mpf->mpf_specification); +#ifdef CONFIG_X86_32 + if (mpf->mpf_feature2 & (1 << 7)) { printk(KERN_INFO " IMCR and PIC compatibility mode.\n"); pic_mode = 1; } else { printk(KERN_INFO " Virtual Wire compatibility mode.\n"); pic_mode = 0; } - +#endif /* * Now see if we need to read further. */ if (mpf->mpf_feature1 != 0) { + if (early) { + /* + * local APIC has default address + */ + mp_lapic_addr = APIC_DEFAULT_PHYS_BASE; + return; + } - printk(KERN_INFO "Default MP configuration #%d\n", mpf->mpf_feature1); + printk(KERN_INFO "Default MP configuration #%d\n", + mpf->mpf_feature1); construct_default_ISA_mptable(mpf->mpf_feature1); } else if (mpf->mpf_physptr) { @@ -683,12 +631,18 @@ void __init get_smp_config (void) * Read the physical hardware table. Anything here will * override the defaults. */ - if (!smp_read_mpc(phys_to_virt(mpf->mpf_physptr))) { + if (!smp_read_mpc(phys_to_virt(mpf->mpf_physptr), early)) { smp_found_config = 0; - printk(KERN_ERR "BIOS bug, MP table errors detected!...\n"); - printk(KERN_ERR "... disabling SMP support. (tell your hw vendor)\n"); + printk(KERN_ERR + "BIOS bug, MP table errors detected!...\n"); + printk(KERN_ERR "... disabling SMP support. " + "(tell your hw vendor)\n"); return; } + + if (early) + return; +#ifdef CONFIG_X86_IO_APIC /* * If there are no explicit MP IRQ entries, then we are * broken. We set up most of the low 16 IO-APIC pins to @@ -697,7 +651,9 @@ void __init get_smp_config (void) if (!mp_irq_entries) { struct mpc_config_bus bus; - printk(KERN_ERR "BIOS bug, no explicit IRQ entries, using default mptable. (tell your hw vendor)\n"); + printk(KERN_ERR "BIOS bug, no explicit IRQ entries, " + "using default mptable. " + "(tell your hw vendor)\n"); bus.mpc_type = MP_BUS; bus.mpc_busid = 0; @@ -706,36 +662,51 @@ void __init get_smp_config (void) construct_default_ioirq_mptable(0); } - +#endif } else BUG(); - printk(KERN_INFO "Processors: %d\n", num_processors); + if (!early) + printk(KERN_INFO "Processors: %d\n", num_processors); /* * Only use the first configuration found. */ } -static int __init smp_scan_config (unsigned long base, unsigned long length) +void __init early_get_smp_config(void) +{ + __get_smp_config(1); +} + +void __init get_smp_config(void) { - unsigned long *bp = phys_to_virt(base); + __get_smp_config(0); +} + +static int __init smp_scan_config(unsigned long base, unsigned long length, + unsigned reserve) +{ + extern void __bad_mpf_size(void); + unsigned int *bp = phys_to_virt(base); struct intel_mp_floating *mpf; - printk(KERN_INFO "Scan SMP from %p for %ld bytes.\n", bp,length); + Dprintk("Scan SMP from %p for %ld bytes.\n", bp, length); if (sizeof(*mpf) != 16) - printk("Error: MPF size\n"); + __bad_mpf_size(); while (length > 0) { mpf = (struct intel_mp_floating *)bp; if ((*bp == SMP_MAGIC_IDENT) && - (mpf->mpf_length == 1) && - !mpf_checksum((unsigned char *)bp, 16) && - ((mpf->mpf_specification == 1) - || (mpf->mpf_specification == 4)) ) { + (mpf->mpf_length == 1) && + !mpf_checksum((unsigned char *)bp, 16) && + ((mpf->mpf_specification == 1) + || (mpf->mpf_specification == 4))) { smp_found_config = 1; + mpf_found = mpf; +#ifdef CONFIG_X86_32 printk(KERN_INFO "found SMP MP-table at [%p] %08lx\n", - mpf, virt_to_phys(mpf)); + mpf, virt_to_phys(mpf)); reserve_bootmem(virt_to_phys(mpf), PAGE_SIZE, BOOTMEM_DEFAULT); if (mpf->mpf_physptr) { @@ -756,8 +727,16 @@ static int __init smp_scan_config (unsigned long base, unsigned long length) BOOTMEM_DEFAULT); } - mpf_found = mpf; - return 1; +#else + if (!reserve) + return 1; + + reserve_bootmem_generic(virt_to_phys(mpf), PAGE_SIZE); + if (mpf->mpf_physptr) + reserve_bootmem_generic(mpf->mpf_physptr, + PAGE_SIZE); +#endif + return 1; } bp += 4; length -= 16; @@ -765,7 +744,7 @@ static int __init smp_scan_config (unsigned long base, unsigned long length) return 0; } -void __init find_smp_config (void) +static void __init __find_smp_config(unsigned reserve) { unsigned int address; @@ -777,9 +756,9 @@ void __init find_smp_config (void) * 2) Scan the top 1K of base RAM * 3) Scan the 64K of bios */ - if (smp_scan_config(0x0,0x400) || - smp_scan_config(639*0x400,0x400) || - smp_scan_config(0xF0000,0x10000)) + if (smp_scan_config(0x0, 0x400, reserve) || + smp_scan_config(639 * 0x400, 0x400, reserve) || + smp_scan_config(0xF0000, 0x10000, reserve)) return; /* * If it is an SMP machine we should know now, unless the @@ -800,144 +779,113 @@ void __init find_smp_config (void) address = get_bios_ebda(); if (address) - smp_scan_config(address, 0x400); + smp_scan_config(address, 0x400, reserve); } -int es7000_plat; - -/* -------------------------------------------------------------------------- - ACPI-based MP Configuration - -------------------------------------------------------------------------- */ - -#ifdef CONFIG_ACPI - -void __init mp_register_lapic_address(u64 address) +void __init early_find_smp_config(void) { - mp_lapic_addr = (unsigned long) address; - - set_fixmap_nocache(FIX_APIC_BASE, mp_lapic_addr); - - if (boot_cpu_physical_apicid == -1U) - boot_cpu_physical_apicid = GET_APIC_ID(apic_read(APIC_ID)); - - Dprintk("Boot CPU = %d\n", boot_cpu_physical_apicid); + __find_smp_config(0); } -void __cpuinit mp_register_lapic (u8 id, u8 enabled) +void __init find_smp_config(void) { - struct mpc_config_processor processor; - int boot_cpu = 0; - - if (MAX_APICS - id <= 0) { - printk(KERN_WARNING "Processor #%d invalid (max %d)\n", - id, MAX_APICS); - return; - } - - if (id == boot_cpu_physical_apicid) - boot_cpu = 1; + __find_smp_config(1); +} - processor.mpc_type = MP_PROCESSOR; - processor.mpc_apicid = id; - processor.mpc_apicver = GET_APIC_VERSION(apic_read(APIC_LVR)); - processor.mpc_cpuflag = (enabled ? CPU_ENABLED : 0); - processor.mpc_cpuflag |= (boot_cpu ? CPU_BOOTPROCESSOR : 0); - processor.mpc_cpufeature = (boot_cpu_data.x86 << 8) | - (boot_cpu_data.x86_model << 4) | boot_cpu_data.x86_mask; - processor.mpc_featureflag = boot_cpu_data.x86_capability[0]; - processor.mpc_reserved[0] = 0; - processor.mpc_reserved[1] = 0; +/* -------------------------------------------------------------------------- + ACPI-based MP Configuration + -------------------------------------------------------------------------- */ - MP_processor_info(&processor); -} +#ifdef CONFIG_ACPI #ifdef CONFIG_X86_IO_APIC #define MP_ISA_BUS 0 #define MP_MAX_IOAPIC_PIN 127 -static struct mp_ioapic_routing { - int apic_id; - int gsi_base; - int gsi_end; - u32 pin_programmed[4]; -} mp_ioapic_routing[MAX_IO_APICS]; +extern struct mp_ioapic_routing mp_ioapic_routing[MAX_IO_APICS]; -static int mp_find_ioapic (int gsi) +static int mp_find_ioapic(int gsi) { int i = 0; /* Find the IOAPIC that manages this GSI. */ for (i = 0; i < nr_ioapics; i++) { if ((gsi >= mp_ioapic_routing[i].gsi_base) - && (gsi <= mp_ioapic_routing[i].gsi_end)) + && (gsi <= mp_ioapic_routing[i].gsi_end)) return i; } printk(KERN_ERR "ERROR: Unable to locate IOAPIC for GSI %d\n", gsi); - return -1; } -void __init mp_register_ioapic(u8 id, u32 address, u32 gsi_base) +static u8 uniq_ioapic_id(u8 id) +{ +#ifdef CONFIG_X86_32 + if ((boot_cpu_data.x86_vendor == X86_VENDOR_INTEL) && + !APIC_XAPIC(apic_version[boot_cpu_physical_apicid])) + return io_apic_get_unique_id(nr_ioapics, id); + else + return id; +#else + int i; + DECLARE_BITMAP(used, 256); + bitmap_zero(used, 256); + for (i = 0; i < nr_ioapics; i++) { + struct mpc_config_ioapic *ia = &mp_ioapics[i]; + __set_bit(ia->mpc_apicid, used); + } + if (!test_bit(id, used)) + return id; + return find_first_zero_bit(used, 256); +#endif +} + +void __init mp_register_ioapic(int id, u32 address, u32 gsi_base) { int idx = 0; - int tmpid; - if (nr_ioapics >= MAX_IO_APICS) { - printk(KERN_ERR "ERROR: Max # of I/O APICs (%d) exceeded " - "(found %d)\n", MAX_IO_APICS, nr_ioapics); - panic("Recompile kernel with bigger MAX_IO_APICS!\n"); - } - if (!address) { - printk(KERN_ERR "WARNING: Bogus (zero) I/O APIC address" - " found in MADT table, skipping!\n"); + if (bad_ioapic(address)) return; - } - idx = nr_ioapics++; + idx = nr_ioapics; mp_ioapics[idx].mpc_type = MP_IOAPIC; mp_ioapics[idx].mpc_flags = MPC_APIC_USABLE; mp_ioapics[idx].mpc_apicaddr = address; set_fixmap_nocache(FIX_IO_APIC_BASE_0 + idx, address); - if ((boot_cpu_data.x86_vendor == X86_VENDOR_INTEL) - && !APIC_XAPIC(apic_version[boot_cpu_physical_apicid])) - tmpid = io_apic_get_unique_id(idx, id); - else - tmpid = id; - if (tmpid == -1) { - nr_ioapics--; - return; - } - mp_ioapics[idx].mpc_apicid = tmpid; + mp_ioapics[idx].mpc_apicid = uniq_ioapic_id(id); +#ifdef CONFIG_X86_32 mp_ioapics[idx].mpc_apicver = io_apic_get_version(idx); - - /* +#else + mp_ioapics[idx].mpc_apicver = 0; +#endif + /* * Build basic GSI lookup table to facilitate gsi->io_apic lookups * and to prevent reprogramming of IOAPIC pins (PCI GSIs). */ mp_ioapic_routing[idx].apic_id = mp_ioapics[idx].mpc_apicid; mp_ioapic_routing[idx].gsi_base = gsi_base; mp_ioapic_routing[idx].gsi_end = gsi_base + - io_apic_get_redir_entries(idx); + io_apic_get_redir_entries(idx); - printk("IOAPIC[%d]: apic_id %d, version %d, address 0x%x, " + printk(KERN_INFO "IOAPIC[%d]: apic_id %d, version %d, address 0x%x, " "GSI %d-%d\n", idx, mp_ioapics[idx].mpc_apicid, mp_ioapics[idx].mpc_apicver, mp_ioapics[idx].mpc_apicaddr, - mp_ioapic_routing[idx].gsi_base, - mp_ioapic_routing[idx].gsi_end); + mp_ioapic_routing[idx].gsi_base, mp_ioapic_routing[idx].gsi_end); + + nr_ioapics++; } -void __init -mp_override_legacy_irq(u8 bus_irq, u8 polarity, u8 trigger, u32 gsi) +void __init mp_override_legacy_irq(u8 bus_irq, u8 polarity, u8 trigger, u32 gsi) { struct mpc_config_intsrc intsrc; - int ioapic = -1; - int pin = -1; + int ioapic = -1; + int pin = -1; - /* + /* * Convert 'gsi' to 'ioapic.pin'. */ ioapic = mp_find_ioapic(gsi); @@ -947,7 +895,7 @@ mp_override_legacy_irq(u8 bus_irq, u8 polarity, u8 trigger, u32 gsi) /* * TBD: This check is for faulty timer entries, where the override - * erroneously sets the trigger to level, resulting in a HUGE + * erroneously sets the trigger to level, resulting in a HUGE * increase of timer interrupts! */ if ((bus_irq == 0) && (trigger == 3)) @@ -957,13 +905,13 @@ mp_override_legacy_irq(u8 bus_irq, u8 polarity, u8 trigger, u32 gsi) intsrc.mpc_irqtype = mp_INT; intsrc.mpc_irqflag = (trigger << 2) | polarity; intsrc.mpc_srcbus = MP_ISA_BUS; - intsrc.mpc_srcbusirq = bus_irq; /* IRQ */ - intsrc.mpc_dstapic = mp_ioapics[ioapic].mpc_apicid; /* APIC ID */ - intsrc.mpc_dstirq = pin; /* INTIN# */ + intsrc.mpc_srcbusirq = bus_irq; /* IRQ */ + intsrc.mpc_dstapic = mp_ioapics[ioapic].mpc_apicid; /* APIC ID */ + intsrc.mpc_dstirq = pin; /* INTIN# */ Dprintk("Int: type %d, pol %d, trig %d, bus %d, irq %d, %d-%d\n", - intsrc.mpc_irqtype, intsrc.mpc_irqflag & 3, - (intsrc.mpc_irqflag >> 2) & 3, intsrc.mpc_srcbus, + intsrc.mpc_irqtype, intsrc.mpc_irqflag & 3, + (intsrc.mpc_irqflag >> 2) & 3, intsrc.mpc_srcbus, intsrc.mpc_srcbusirq, intsrc.mpc_dstapic, intsrc.mpc_dstirq); mp_irqs[mp_irq_entries] = intsrc; @@ -971,16 +919,21 @@ mp_override_legacy_irq(u8 bus_irq, u8 polarity, u8 trigger, u32 gsi) panic("Max # of irq sources exceeded!\n"); } -void __init mp_config_acpi_legacy_irqs (void) +int es7000_plat; + +void __init mp_config_acpi_legacy_irqs(void) { struct mpc_config_intsrc intsrc; int i = 0; int ioapic = -1; - /* +#if defined (CONFIG_MCA) || defined (CONFIG_EISA) + /* * Fabricate the legacy ISA bus (bus #31). */ mp_bus_id_to_type[MP_ISA_BUS] = MP_BUS_ISA; +#endif + set_bit(MP_ISA_BUS, mp_bus_not_pci); Dprintk("Bus #%d is ISA\n", MP_ISA_BUS); /* @@ -989,19 +942,20 @@ void __init mp_config_acpi_legacy_irqs (void) if (es7000_plat == 1) return; - /* - * Locate the IOAPIC that manages the ISA IRQs (0-15). + /* + * Locate the IOAPIC that manages the ISA IRQs (0-15). */ ioapic = mp_find_ioapic(0); if (ioapic < 0) return; intsrc.mpc_type = MP_INTSRC; - intsrc.mpc_irqflag = 0; /* Conforming */ + intsrc.mpc_irqflag = 0; /* Conforming */ intsrc.mpc_srcbus = MP_ISA_BUS; +#ifdef CONFIG_X86_IO_APIC intsrc.mpc_dstapic = mp_ioapics[ioapic].mpc_apicid; - - /* +#endif + /* * Use the default configuration for the IRQs 0-15. Unless * overridden by (MADT) interrupt source override entries. */ @@ -1012,28 +966,29 @@ void __init mp_config_acpi_legacy_irqs (void) struct mpc_config_intsrc *irq = mp_irqs + idx; /* Do we already have a mapping for this ISA IRQ? */ - if (irq->mpc_srcbus == MP_ISA_BUS && irq->mpc_srcbusirq == i) + if (irq->mpc_srcbus == MP_ISA_BUS + && irq->mpc_srcbusirq == i) break; /* Do we already have a mapping for this IOAPIC pin */ if ((irq->mpc_dstapic == intsrc.mpc_dstapic) && - (irq->mpc_dstirq == i)) + (irq->mpc_dstirq == i)) break; } if (idx != mp_irq_entries) { printk(KERN_DEBUG "ACPI: IRQ%d used by override.\n", i); - continue; /* IRQ already used */ + continue; /* IRQ already used */ } intsrc.mpc_irqtype = mp_INT; - intsrc.mpc_srcbusirq = i; /* Identity mapped */ + intsrc.mpc_srcbusirq = i; /* Identity mapped */ intsrc.mpc_dstirq = i; Dprintk("Int: type %d, pol %d, trig %d, bus %d, irq %d, " - "%d-%d\n", intsrc.mpc_irqtype, intsrc.mpc_irqflag & 3, - (intsrc.mpc_irqflag >> 2) & 3, intsrc.mpc_srcbus, - intsrc.mpc_srcbusirq, intsrc.mpc_dstapic, + "%d-%d\n", intsrc.mpc_irqtype, intsrc.mpc_irqflag & 3, + (intsrc.mpc_irqflag >> 2) & 3, intsrc.mpc_srcbus, + intsrc.mpc_srcbusirq, intsrc.mpc_dstapic, intsrc.mpc_dstirq); mp_irqs[mp_irq_entries] = intsrc; @@ -1042,21 +997,27 @@ void __init mp_config_acpi_legacy_irqs (void) } } -#define MAX_GSI_NUM 4096 -#define IRQ_COMPRESSION_START 64 - int mp_register_gsi(u32 gsi, int triggering, int polarity) { int ioapic = -1; int ioapic_pin = 0; int idx, bit = 0; +#ifdef CONFIG_X86_32 +#define MAX_GSI_NUM 4096 +#define IRQ_COMPRESSION_START 64 + static int pci_irq = IRQ_COMPRESSION_START; /* * Mapping between Global System Interrupts, which * represent all possible interrupts, and IRQs * assigned to actual devices. */ - static int gsi_to_irq[MAX_GSI_NUM]; + static int gsi_to_irq[MAX_GSI_NUM]; +#else + + if (acpi_irq_model != ACPI_IRQ_MODEL_IOAPIC) + return gsi; +#endif /* Don't set up the ACPI SCI because it's already set up */ if (acpi_gbl_FADT.sci_interrupt == gsi) @@ -1070,11 +1031,13 @@ int mp_register_gsi(u32 gsi, int triggering, int polarity) ioapic_pin = gsi - mp_ioapic_routing[ioapic].gsi_base; +#ifdef CONFIG_X86_32 if (ioapic_renumber_irq) gsi = ioapic_renumber_irq(ioapic, gsi); +#endif - /* - * Avoid pin reprogramming. PRTs typically include entries + /* + * Avoid pin reprogramming. PRTs typically include entries * with redundant pin->gsi mappings (but unique PCI devices); * we only program the IOAPIC on the first. */ @@ -1082,23 +1045,27 @@ int mp_register_gsi(u32 gsi, int triggering, int polarity) idx = (ioapic_pin < 32) ? 0 : (ioapic_pin / 32); if (idx > 3) { printk(KERN_ERR "Invalid reference to IOAPIC pin " - "%d-%d\n", mp_ioapic_routing[ioapic].apic_id, - ioapic_pin); + "%d-%d\n", mp_ioapic_routing[ioapic].apic_id, + ioapic_pin); return gsi; } - if ((1<<bit) & mp_ioapic_routing[ioapic].pin_programmed[idx]) { + if ((1 << bit) & mp_ioapic_routing[ioapic].pin_programmed[idx]) { Dprintk(KERN_DEBUG "Pin %d-%d already programmed\n", mp_ioapic_routing[ioapic].apic_id, ioapic_pin); +#ifdef CONFIG_X86_32 return (gsi < IRQ_COMPRESSION_START ? gsi : gsi_to_irq[gsi]); +#else + return gsi; +#endif } - mp_ioapic_routing[ioapic].pin_programmed[idx] |= (1<<bit); - + mp_ioapic_routing[ioapic].pin_programmed[idx] |= (1 << bit); +#ifdef CONFIG_X86_32 /* * For GSI >= 64, use IRQ compression */ if ((gsi >= IRQ_COMPRESSION_START) - && (triggering == ACPI_LEVEL_SENSITIVE)) { + && (triggering == ACPI_LEVEL_SENSITIVE)) { /* * For PCI devices assign IRQs in order, avoiding gaps * due to unused I/O APIC pins. @@ -1115,8 +1082,7 @@ int mp_register_gsi(u32 gsi, int triggering, int polarity) * So test for this condition, and if necessary, avoid * the pin collision. */ - if (gsi > 15 || (gsi == 0 && !timer_uses_ioapic_pin_0)) - gsi = pci_irq++; + gsi = pci_irq++; /* * Don't assign IRQ used by ACPI SCI */ @@ -1128,10 +1094,10 @@ int mp_register_gsi(u32 gsi, int triggering, int polarity) return gsi; } } - +#endif io_apic_set_pci_routing(ioapic, ioapic_pin, gsi, - triggering == ACPI_EDGE_SENSITIVE ? 0 : 1, - polarity == ACPI_ACTIVE_HIGH ? 0 : 1); + triggering == ACPI_EDGE_SENSITIVE ? 0 : 1, + polarity == ACPI_ACTIVE_HIGH ? 0 : 1); return gsi; } diff --git a/arch/x86/kernel/mpparse_64.c b/arch/x86/kernel/mpparse_64.c deleted file mode 100644 index 72ab1403fed7..000000000000 --- a/arch/x86/kernel/mpparse_64.c +++ /dev/null @@ -1,867 +0,0 @@ -/* - * Intel Multiprocessor Specification 1.1 and 1.4 - * compliant MP-table parsing routines. - * - * (c) 1995 Alan Cox, Building #3 <alan@redhat.com> - * (c) 1998, 1999, 2000 Ingo Molnar <mingo@redhat.com> - * - * Fixes - * Erich Boleyn : MP v1.4 and additional changes. - * Alan Cox : Added EBDA scanning - * Ingo Molnar : various cleanups and rewrites - * Maciej W. Rozycki: Bits for default MP configurations - * Paul Diefenbaugh: Added full ACPI support - */ - -#include <linux/mm.h> -#include <linux/init.h> -#include <linux/delay.h> -#include <linux/bootmem.h> -#include <linux/kernel_stat.h> -#include <linux/mc146818rtc.h> -#include <linux/acpi.h> -#include <linux/module.h> - -#include <asm/smp.h> -#include <asm/mtrr.h> -#include <asm/mpspec.h> -#include <asm/pgalloc.h> -#include <asm/io_apic.h> -#include <asm/proto.h> -#include <asm/acpi.h> - -/* Have we found an MP table */ -int smp_found_config; - -/* - * Various Linux-internal data structures created from the - * MP-table. - */ -DECLARE_BITMAP(mp_bus_not_pci, MAX_MP_BUSSES); -int mp_bus_id_to_pci_bus [MAX_MP_BUSSES] = { [0 ... MAX_MP_BUSSES-1] = -1 }; - -static int mp_current_pci_id = 0; -/* I/O APIC entries */ -struct mpc_config_ioapic mp_ioapics[MAX_IO_APICS]; - -/* # of MP IRQ source entries */ -struct mpc_config_intsrc mp_irqs[MAX_IRQ_SOURCES]; - -/* MP IRQ source entries */ -int mp_irq_entries; - -int nr_ioapics; -unsigned long mp_lapic_addr = 0; - - - -/* Processor that is doing the boot up */ -unsigned int boot_cpu_id = -1U; -EXPORT_SYMBOL(boot_cpu_id); - -/* Internal processor count */ -unsigned int num_processors; - -unsigned disabled_cpus __cpuinitdata; - -/* Bitmask of physically existing CPUs */ -physid_mask_t phys_cpu_present_map = PHYSID_MASK_NONE; - -u16 x86_bios_cpu_apicid_init[NR_CPUS] __initdata - = { [0 ... NR_CPUS-1] = BAD_APICID }; -void *x86_bios_cpu_apicid_early_ptr; -DEFINE_PER_CPU(u16, x86_bios_cpu_apicid) = BAD_APICID; -EXPORT_PER_CPU_SYMBOL(x86_bios_cpu_apicid); - - -/* - * Intel MP BIOS table parsing routines: - */ - -/* - * Checksum an MP configuration block. - */ - -static int __init mpf_checksum(unsigned char *mp, int len) -{ - int sum = 0; - - while (len--) - sum += *mp++; - - return sum & 0xFF; -} - -static void __cpuinit MP_processor_info(struct mpc_config_processor *m) -{ - int cpu; - cpumask_t tmp_map; - char *bootup_cpu = ""; - - if (!(m->mpc_cpuflag & CPU_ENABLED)) { - disabled_cpus++; - return; - } - if (m->mpc_cpuflag & CPU_BOOTPROCESSOR) { - bootup_cpu = " (Bootup-CPU)"; - boot_cpu_id = m->mpc_apicid; - } - - printk(KERN_INFO "Processor #%d%s\n", m->mpc_apicid, bootup_cpu); - - if (num_processors >= NR_CPUS) { - printk(KERN_WARNING "WARNING: NR_CPUS limit of %i reached." - " Processor ignored.\n", NR_CPUS); - return; - } - - num_processors++; - cpus_complement(tmp_map, cpu_present_map); - cpu = first_cpu(tmp_map); - - physid_set(m->mpc_apicid, phys_cpu_present_map); - if (m->mpc_cpuflag & CPU_BOOTPROCESSOR) { - /* - * x86_bios_cpu_apicid is required to have processors listed - * in same order as logical cpu numbers. Hence the first - * entry is BSP, and so on. - */ - cpu = 0; - } - /* are we being called early in kernel startup? */ - if (x86_cpu_to_apicid_early_ptr) { - u16 *cpu_to_apicid = x86_cpu_to_apicid_early_ptr; - u16 *bios_cpu_apicid = x86_bios_cpu_apicid_early_ptr; - - cpu_to_apicid[cpu] = m->mpc_apicid; - bios_cpu_apicid[cpu] = m->mpc_apicid; - } else { - per_cpu(x86_cpu_to_apicid, cpu) = m->mpc_apicid; - per_cpu(x86_bios_cpu_apicid, cpu) = m->mpc_apicid; - } - - cpu_set(cpu, cpu_possible_map); - cpu_set(cpu, cpu_present_map); -} - -static void __init MP_bus_info (struct mpc_config_bus *m) -{ - char str[7]; - - memcpy(str, m->mpc_bustype, 6); - str[6] = 0; - Dprintk("Bus #%d is %s\n", m->mpc_busid, str); - - if (strncmp(str, "ISA", 3) == 0) { - set_bit(m->mpc_busid, mp_bus_not_pci); - } else if (strncmp(str, "PCI", 3) == 0) { - clear_bit(m->mpc_busid, mp_bus_not_pci); - mp_bus_id_to_pci_bus[m->mpc_busid] = mp_current_pci_id; - mp_current_pci_id++; - } else { - printk(KERN_ERR "Unknown bustype %s\n", str); - } -} - -static int bad_ioapic(unsigned long address) -{ - if (nr_ioapics >= MAX_IO_APICS) { - printk(KERN_ERR "ERROR: Max # of I/O APICs (%d) exceeded " - "(found %d)\n", MAX_IO_APICS, nr_ioapics); - panic("Recompile kernel with bigger MAX_IO_APICS!\n"); - } - if (!address) { - printk(KERN_ERR "WARNING: Bogus (zero) I/O APIC address" - " found in table, skipping!\n"); - return 1; - } - return 0; -} - -static void __init MP_ioapic_info (struct mpc_config_ioapic *m) -{ - if (!(m->mpc_flags & MPC_APIC_USABLE)) - return; - - printk("I/O APIC #%d at 0x%X.\n", - m->mpc_apicid, m->mpc_apicaddr); - - if (bad_ioapic(m->mpc_apicaddr)) - return; - - mp_ioapics[nr_ioapics] = *m; - nr_ioapics++; -} - -static void __init MP_intsrc_info (struct mpc_config_intsrc *m) -{ - mp_irqs [mp_irq_entries] = *m; - Dprintk("Int: type %d, pol %d, trig %d, bus %d," - " IRQ %02x, APIC ID %x, APIC INT %02x\n", - m->mpc_irqtype, m->mpc_irqflag & 3, - (m->mpc_irqflag >> 2) & 3, m->mpc_srcbus, - m->mpc_srcbusirq, m->mpc_dstapic, m->mpc_dstirq); - if (++mp_irq_entries >= MAX_IRQ_SOURCES) - panic("Max # of irq sources exceeded!!\n"); -} - -static void __init MP_lintsrc_info (struct mpc_config_lintsrc *m) -{ - Dprintk("Lint: type %d, pol %d, trig %d, bus %d," - " IRQ %02x, APIC ID %x, APIC LINT %02x\n", - m->mpc_irqtype, m->mpc_irqflag & 3, - (m->mpc_irqflag >> 2) &3, m->mpc_srcbusid, - m->mpc_srcbusirq, m->mpc_destapic, m->mpc_destapiclint); -} - -/* - * Read/parse the MPC - */ - -static int __init smp_read_mpc(struct mp_config_table *mpc) -{ - char str[16]; - int count=sizeof(*mpc); - unsigned char *mpt=((unsigned char *)mpc)+count; - - if (memcmp(mpc->mpc_signature,MPC_SIGNATURE,4)) { - printk("MPTABLE: bad signature [%c%c%c%c]!\n", - mpc->mpc_signature[0], - mpc->mpc_signature[1], - mpc->mpc_signature[2], - mpc->mpc_signature[3]); - return 0; - } - if (mpf_checksum((unsigned char *)mpc,mpc->mpc_length)) { - printk("MPTABLE: checksum error!\n"); - return 0; - } - if (mpc->mpc_spec!=0x01 && mpc->mpc_spec!=0x04) { - printk(KERN_ERR "MPTABLE: bad table version (%d)!!\n", - mpc->mpc_spec); - return 0; - } - if (!mpc->mpc_lapic) { - printk(KERN_ERR "MPTABLE: null local APIC address!\n"); - return 0; - } - memcpy(str,mpc->mpc_oem,8); - str[8] = 0; - printk(KERN_INFO "MPTABLE: OEM ID: %s ",str); - - memcpy(str,mpc->mpc_productid,12); - str[12] = 0; - printk("MPTABLE: Product ID: %s ",str); - - printk("MPTABLE: APIC at: 0x%X\n",mpc->mpc_lapic); - - /* save the local APIC address, it might be non-default */ - if (!acpi_lapic) - mp_lapic_addr = mpc->mpc_lapic; - - /* - * Now process the configuration blocks. - */ - while (count < mpc->mpc_length) { - switch(*mpt) { - case MP_PROCESSOR: - { - struct mpc_config_processor *m= - (struct mpc_config_processor *)mpt; - if (!acpi_lapic) - MP_processor_info(m); - mpt += sizeof(*m); - count += sizeof(*m); - break; - } - case MP_BUS: - { - struct mpc_config_bus *m= - (struct mpc_config_bus *)mpt; - MP_bus_info(m); - mpt += sizeof(*m); - count += sizeof(*m); - break; - } - case MP_IOAPIC: - { - struct mpc_config_ioapic *m= - (struct mpc_config_ioapic *)mpt; - MP_ioapic_info(m); - mpt += sizeof(*m); - count += sizeof(*m); - break; - } - case MP_INTSRC: - { - struct mpc_config_intsrc *m= - (struct mpc_config_intsrc *)mpt; - - MP_intsrc_info(m); - mpt += sizeof(*m); - count += sizeof(*m); - break; - } - case MP_LINTSRC: - { - struct mpc_config_lintsrc *m= - (struct mpc_config_lintsrc *)mpt; - MP_lintsrc_info(m); - mpt += sizeof(*m); - count += sizeof(*m); - break; - } - } - } - setup_apic_routing(); - if (!num_processors) - printk(KERN_ERR "MPTABLE: no processors registered!\n"); - return num_processors; -} - -static int __init ELCR_trigger(unsigned int irq) -{ - unsigned int port; - - port = 0x4d0 + (irq >> 3); - return (inb(port) >> (irq & 7)) & 1; -} - -static void __init construct_default_ioirq_mptable(int mpc_default_type) -{ - struct mpc_config_intsrc intsrc; - int i; - int ELCR_fallback = 0; - - intsrc.mpc_type = MP_INTSRC; - intsrc.mpc_irqflag = 0; /* conforming */ - intsrc.mpc_srcbus = 0; - intsrc.mpc_dstapic = mp_ioapics[0].mpc_apicid; - - intsrc.mpc_irqtype = mp_INT; - - /* - * If true, we have an ISA/PCI system with no IRQ entries - * in the MP table. To prevent the PCI interrupts from being set up - * incorrectly, we try to use the ELCR. The sanity check to see if - * there is good ELCR data is very simple - IRQ0, 1, 2 and 13 can - * never be level sensitive, so we simply see if the ELCR agrees. - * If it does, we assume it's valid. - */ - if (mpc_default_type == 5) { - printk(KERN_INFO "ISA/PCI bus type with no IRQ information... falling back to ELCR\n"); - - if (ELCR_trigger(0) || ELCR_trigger(1) || ELCR_trigger(2) || ELCR_trigger(13)) - printk(KERN_ERR "ELCR contains invalid data... not using ELCR\n"); - else { - printk(KERN_INFO "Using ELCR to identify PCI interrupts\n"); - ELCR_fallback = 1; - } - } - - for (i = 0; i < 16; i++) { - switch (mpc_default_type) { - case 2: - if (i == 0 || i == 13) - continue; /* IRQ0 & IRQ13 not connected */ - /* fall through */ - default: - if (i == 2) - continue; /* IRQ2 is never connected */ - } - - if (ELCR_fallback) { - /* - * If the ELCR indicates a level-sensitive interrupt, we - * copy that information over to the MP table in the - * irqflag field (level sensitive, active high polarity). - */ - if (ELCR_trigger(i)) - intsrc.mpc_irqflag = 13; - else - intsrc.mpc_irqflag = 0; - } - - intsrc.mpc_srcbusirq = i; - intsrc.mpc_dstirq = i ? i : 2; /* IRQ0 to INTIN2 */ - MP_intsrc_info(&intsrc); - } - - intsrc.mpc_irqtype = mp_ExtINT; - intsrc.mpc_srcbusirq = 0; - intsrc.mpc_dstirq = 0; /* 8259A to INTIN0 */ - MP_intsrc_info(&intsrc); -} - -static inline void __init construct_default_ISA_mptable(int mpc_default_type) -{ - struct mpc_config_processor processor; - struct mpc_config_bus bus; - struct mpc_config_ioapic ioapic; - struct mpc_config_lintsrc lintsrc; - int linttypes[2] = { mp_ExtINT, mp_NMI }; - int i; - - /* - * local APIC has default address - */ - mp_lapic_addr = APIC_DEFAULT_PHYS_BASE; - - /* - * 2 CPUs, numbered 0 & 1. - */ - processor.mpc_type = MP_PROCESSOR; - processor.mpc_apicver = 0; - processor.mpc_cpuflag = CPU_ENABLED; - processor.mpc_cpufeature = 0; - processor.mpc_featureflag = 0; - processor.mpc_reserved[0] = 0; - processor.mpc_reserved[1] = 0; - for (i = 0; i < 2; i++) { - processor.mpc_apicid = i; - MP_processor_info(&processor); - } - - bus.mpc_type = MP_BUS; - bus.mpc_busid = 0; - switch (mpc_default_type) { - default: - printk(KERN_ERR "???\nUnknown standard configuration %d\n", - mpc_default_type); - /* fall through */ - case 1: - case 5: - memcpy(bus.mpc_bustype, "ISA ", 6); - break; - } - MP_bus_info(&bus); - if (mpc_default_type > 4) { - bus.mpc_busid = 1; - memcpy(bus.mpc_bustype, "PCI ", 6); - MP_bus_info(&bus); - } - - ioapic.mpc_type = MP_IOAPIC; - ioapic.mpc_apicid = 2; - ioapic.mpc_apicver = 0; - ioapic.mpc_flags = MPC_APIC_USABLE; - ioapic.mpc_apicaddr = 0xFEC00000; - MP_ioapic_info(&ioapic); - - /* - * We set up most of the low 16 IO-APIC pins according to MPS rules. - */ - construct_default_ioirq_mptable(mpc_default_type); - - lintsrc.mpc_type = MP_LINTSRC; - lintsrc.mpc_irqflag = 0; /* conforming */ - lintsrc.mpc_srcbusid = 0; - lintsrc.mpc_srcbusirq = 0; - lintsrc.mpc_destapic = MP_APIC_ALL; - for (i = 0; i < 2; i++) { - lintsrc.mpc_irqtype = linttypes[i]; - lintsrc.mpc_destapiclint = i; - MP_lintsrc_info(&lintsrc); - } -} - -static struct intel_mp_floating *mpf_found; - -/* - * Scan the memory blocks for an SMP configuration block. - */ -void __init get_smp_config (void) -{ - struct intel_mp_floating *mpf = mpf_found; - - /* - * ACPI supports both logical (e.g. Hyper-Threading) and physical - * processors, where MPS only supports physical. - */ - if (acpi_lapic && acpi_ioapic) { - printk(KERN_INFO "Using ACPI (MADT) for SMP configuration information\n"); - return; - } - else if (acpi_lapic) - printk(KERN_INFO "Using ACPI for processor (LAPIC) configuration information\n"); - - printk("Intel MultiProcessor Specification v1.%d\n", mpf->mpf_specification); - - /* - * Now see if we need to read further. - */ - if (mpf->mpf_feature1 != 0) { - - printk(KERN_INFO "Default MP configuration #%d\n", mpf->mpf_feature1); - construct_default_ISA_mptable(mpf->mpf_feature1); - - } else if (mpf->mpf_physptr) { - - /* - * Read the physical hardware table. Anything here will - * override the defaults. - */ - if (!smp_read_mpc(phys_to_virt(mpf->mpf_physptr))) { - smp_found_config = 0; - printk(KERN_ERR "BIOS bug, MP table errors detected!...\n"); - printk(KERN_ERR "... disabling SMP support. (tell your hw vendor)\n"); - return; - } - /* - * If there are no explicit MP IRQ entries, then we are - * broken. We set up most of the low 16 IO-APIC pins to - * ISA defaults and hope it will work. - */ - if (!mp_irq_entries) { - struct mpc_config_bus bus; - - printk(KERN_ERR "BIOS bug, no explicit IRQ entries, using default mptable. (tell your hw vendor)\n"); - - bus.mpc_type = MP_BUS; - bus.mpc_busid = 0; - memcpy(bus.mpc_bustype, "ISA ", 6); - MP_bus_info(&bus); - - construct_default_ioirq_mptable(0); - } - - } else - BUG(); - - printk(KERN_INFO "Processors: %d\n", num_processors); - /* - * Only use the first configuration found. - */ -} - -static int __init smp_scan_config (unsigned long base, unsigned long length) -{ - extern void __bad_mpf_size(void); - unsigned int *bp = phys_to_virt(base); - struct intel_mp_floating *mpf; - - Dprintk("Scan SMP from %p for %ld bytes.\n", bp,length); - if (sizeof(*mpf) != 16) - __bad_mpf_size(); - - while (length > 0) { - mpf = (struct intel_mp_floating *)bp; - if ((*bp == SMP_MAGIC_IDENT) && - (mpf->mpf_length == 1) && - !mpf_checksum((unsigned char *)bp, 16) && - ((mpf->mpf_specification == 1) - || (mpf->mpf_specification == 4)) ) { - - smp_found_config = 1; - reserve_bootmem_generic(virt_to_phys(mpf), PAGE_SIZE); - if (mpf->mpf_physptr) - reserve_bootmem_generic(mpf->mpf_physptr, PAGE_SIZE); - mpf_found = mpf; - return 1; - } - bp += 4; - length -= 16; - } - return 0; -} - -void __init find_smp_config(void) -{ - unsigned int address; - - /* - * FIXME: Linux assumes you have 640K of base ram.. - * this continues the error... - * - * 1) Scan the bottom 1K for a signature - * 2) Scan the top 1K of base RAM - * 3) Scan the 64K of bios - */ - if (smp_scan_config(0x0,0x400) || - smp_scan_config(639*0x400,0x400) || - smp_scan_config(0xF0000,0x10000)) - return; - /* - * If it is an SMP machine we should know now. - * - * there is a real-mode segmented pointer pointing to the - * 4K EBDA area at 0x40E, calculate and scan it here. - * - * NOTE! There are Linux loaders that will corrupt the EBDA - * area, and as such this kind of SMP config may be less - * trustworthy, simply because the SMP table may have been - * stomped on during early boot. These loaders are buggy and - * should be fixed. - */ - - address = *(unsigned short *)phys_to_virt(0x40E); - address <<= 4; - if (smp_scan_config(address, 0x1000)) - return; - - /* If we have come this far, we did not find an MP table */ - printk(KERN_INFO "No mptable found.\n"); -} - -/* -------------------------------------------------------------------------- - ACPI-based MP Configuration - -------------------------------------------------------------------------- */ - -#ifdef CONFIG_ACPI - -void __init mp_register_lapic_address(u64 address) -{ - mp_lapic_addr = (unsigned long) address; - set_fixmap_nocache(FIX_APIC_BASE, mp_lapic_addr); - if (boot_cpu_id == -1U) - boot_cpu_id = GET_APIC_ID(apic_read(APIC_ID)); -} - -void __cpuinit mp_register_lapic (u8 id, u8 enabled) -{ - struct mpc_config_processor processor; - int boot_cpu = 0; - - if (id == boot_cpu_id) - boot_cpu = 1; - - processor.mpc_type = MP_PROCESSOR; - processor.mpc_apicid = id; - processor.mpc_apicver = 0; - processor.mpc_cpuflag = (enabled ? CPU_ENABLED : 0); - processor.mpc_cpuflag |= (boot_cpu ? CPU_BOOTPROCESSOR : 0); - processor.mpc_cpufeature = 0; - processor.mpc_featureflag = 0; - processor.mpc_reserved[0] = 0; - processor.mpc_reserved[1] = 0; - - MP_processor_info(&processor); -} - -#define MP_ISA_BUS 0 -#define MP_MAX_IOAPIC_PIN 127 - -static struct mp_ioapic_routing { - int apic_id; - int gsi_start; - int gsi_end; - u32 pin_programmed[4]; -} mp_ioapic_routing[MAX_IO_APICS]; - -static int mp_find_ioapic(int gsi) -{ - int i = 0; - - /* Find the IOAPIC that manages this GSI. */ - for (i = 0; i < nr_ioapics; i++) { - if ((gsi >= mp_ioapic_routing[i].gsi_start) - && (gsi <= mp_ioapic_routing[i].gsi_end)) - return i; - } - - printk(KERN_ERR "ERROR: Unable to locate IOAPIC for GSI %d\n", gsi); - return -1; -} - -static u8 uniq_ioapic_id(u8 id) -{ - int i; - DECLARE_BITMAP(used, 256); - bitmap_zero(used, 256); - for (i = 0; i < nr_ioapics; i++) { - struct mpc_config_ioapic *ia = &mp_ioapics[i]; - __set_bit(ia->mpc_apicid, used); - } - if (!test_bit(id, used)) - return id; - return find_first_zero_bit(used, 256); -} - -void __init mp_register_ioapic(u8 id, u32 address, u32 gsi_base) -{ - int idx = 0; - - if (bad_ioapic(address)) - return; - - idx = nr_ioapics; - - mp_ioapics[idx].mpc_type = MP_IOAPIC; - mp_ioapics[idx].mpc_flags = MPC_APIC_USABLE; - mp_ioapics[idx].mpc_apicaddr = address; - - set_fixmap_nocache(FIX_IO_APIC_BASE_0 + idx, address); - mp_ioapics[idx].mpc_apicid = uniq_ioapic_id(id); - mp_ioapics[idx].mpc_apicver = 0; - - /* - * Build basic IRQ lookup table to facilitate gsi->io_apic lookups - * and to prevent reprogramming of IOAPIC pins (PCI IRQs). - */ - mp_ioapic_routing[idx].apic_id = mp_ioapics[idx].mpc_apicid; - mp_ioapic_routing[idx].gsi_start = gsi_base; - mp_ioapic_routing[idx].gsi_end = gsi_base + - io_apic_get_redir_entries(idx); - - printk(KERN_INFO "IOAPIC[%d]: apic_id %d, address 0x%x, " - "GSI %d-%d\n", idx, mp_ioapics[idx].mpc_apicid, - mp_ioapics[idx].mpc_apicaddr, - mp_ioapic_routing[idx].gsi_start, - mp_ioapic_routing[idx].gsi_end); - - nr_ioapics++; -} - -void __init -mp_override_legacy_irq(u8 bus_irq, u8 polarity, u8 trigger, u32 gsi) -{ - struct mpc_config_intsrc intsrc; - int ioapic = -1; - int pin = -1; - - /* - * Convert 'gsi' to 'ioapic.pin'. - */ - ioapic = mp_find_ioapic(gsi); - if (ioapic < 0) - return; - pin = gsi - mp_ioapic_routing[ioapic].gsi_start; - - /* - * TBD: This check is for faulty timer entries, where the override - * erroneously sets the trigger to level, resulting in a HUGE - * increase of timer interrupts! - */ - if ((bus_irq == 0) && (trigger == 3)) - trigger = 1; - - intsrc.mpc_type = MP_INTSRC; - intsrc.mpc_irqtype = mp_INT; - intsrc.mpc_irqflag = (trigger << 2) | polarity; - intsrc.mpc_srcbus = MP_ISA_BUS; - intsrc.mpc_srcbusirq = bus_irq; /* IRQ */ - intsrc.mpc_dstapic = mp_ioapics[ioapic].mpc_apicid; /* APIC ID */ - intsrc.mpc_dstirq = pin; /* INTIN# */ - - Dprintk("Int: type %d, pol %d, trig %d, bus %d, irq %d, %d-%d\n", - intsrc.mpc_irqtype, intsrc.mpc_irqflag & 3, - (intsrc.mpc_irqflag >> 2) & 3, intsrc.mpc_srcbus, - intsrc.mpc_srcbusirq, intsrc.mpc_dstapic, intsrc.mpc_dstirq); - - mp_irqs[mp_irq_entries] = intsrc; - if (++mp_irq_entries == MAX_IRQ_SOURCES) - panic("Max # of irq sources exceeded!\n"); -} - -void __init mp_config_acpi_legacy_irqs(void) -{ - struct mpc_config_intsrc intsrc; - int i = 0; - int ioapic = -1; - - /* - * Fabricate the legacy ISA bus (bus #31). - */ - set_bit(MP_ISA_BUS, mp_bus_not_pci); - - /* - * Locate the IOAPIC that manages the ISA IRQs (0-15). - */ - ioapic = mp_find_ioapic(0); - if (ioapic < 0) - return; - - intsrc.mpc_type = MP_INTSRC; - intsrc.mpc_irqflag = 0; /* Conforming */ - intsrc.mpc_srcbus = MP_ISA_BUS; - intsrc.mpc_dstapic = mp_ioapics[ioapic].mpc_apicid; - - /* - * Use the default configuration for the IRQs 0-15. Unless - * overridden by (MADT) interrupt source override entries. - */ - for (i = 0; i < 16; i++) { - int idx; - - for (idx = 0; idx < mp_irq_entries; idx++) { - struct mpc_config_intsrc *irq = mp_irqs + idx; - - /* Do we already have a mapping for this ISA IRQ? */ - if (irq->mpc_srcbus == MP_ISA_BUS && irq->mpc_srcbusirq == i) - break; - - /* Do we already have a mapping for this IOAPIC pin */ - if ((irq->mpc_dstapic == intsrc.mpc_dstapic) && - (irq->mpc_dstirq == i)) - break; - } - - if (idx != mp_irq_entries) { - printk(KERN_DEBUG "ACPI: IRQ%d used by override.\n", i); - continue; /* IRQ already used */ - } - - intsrc.mpc_irqtype = mp_INT; - intsrc.mpc_srcbusirq = i; /* Identity mapped */ - intsrc.mpc_dstirq = i; - - Dprintk("Int: type %d, pol %d, trig %d, bus %d, irq %d, " - "%d-%d\n", intsrc.mpc_irqtype, intsrc.mpc_irqflag & 3, - (intsrc.mpc_irqflag >> 2) & 3, intsrc.mpc_srcbus, - intsrc.mpc_srcbusirq, intsrc.mpc_dstapic, - intsrc.mpc_dstirq); - - mp_irqs[mp_irq_entries] = intsrc; - if (++mp_irq_entries == MAX_IRQ_SOURCES) - panic("Max # of irq sources exceeded!\n"); - } -} - -int mp_register_gsi(u32 gsi, int triggering, int polarity) -{ - int ioapic = -1; - int ioapic_pin = 0; - int idx, bit = 0; - - if (acpi_irq_model != ACPI_IRQ_MODEL_IOAPIC) - return gsi; - - /* Don't set up the ACPI SCI because it's already set up */ - if (acpi_gbl_FADT.sci_interrupt == gsi) - return gsi; - - ioapic = mp_find_ioapic(gsi); - if (ioapic < 0) { - printk(KERN_WARNING "No IOAPIC for GSI %u\n", gsi); - return gsi; - } - - ioapic_pin = gsi - mp_ioapic_routing[ioapic].gsi_start; - - /* - * Avoid pin reprogramming. PRTs typically include entries - * with redundant pin->gsi mappings (but unique PCI devices); - * we only program the IOAPIC on the first. - */ - bit = ioapic_pin % 32; - idx = (ioapic_pin < 32) ? 0 : (ioapic_pin / 32); - if (idx > 3) { - printk(KERN_ERR "Invalid reference to IOAPIC pin " - "%d-%d\n", mp_ioapic_routing[ioapic].apic_id, - ioapic_pin); - return gsi; - } - if ((1<<bit) & mp_ioapic_routing[ioapic].pin_programmed[idx]) { - Dprintk(KERN_DEBUG "Pin %d-%d already programmed\n", - mp_ioapic_routing[ioapic].apic_id, ioapic_pin); - return gsi; - } - - mp_ioapic_routing[ioapic].pin_programmed[idx] |= (1<<bit); - - io_apic_set_pci_routing(ioapic, ioapic_pin, gsi, - triggering == ACPI_EDGE_SENSITIVE ? 0 : 1, - polarity == ACPI_ACTIVE_HIGH ? 0 : 1); - return gsi; -} -#endif /*CONFIG_ACPI*/ diff --git a/arch/x86/kernel/msr.c b/arch/x86/kernel/msr.c index af51ea8400b2..4dfb40530057 100644 --- a/arch/x86/kernel/msr.c +++ b/arch/x86/kernel/msr.c @@ -65,8 +65,8 @@ static loff_t msr_seek(struct file *file, loff_t offset, int orig) return ret; } -static ssize_t msr_read(struct file *file, char __user * buf, - size_t count, loff_t * ppos) +static ssize_t msr_read(struct file *file, char __user *buf, + size_t count, loff_t *ppos) { u32 __user *tmp = (u32 __user *) buf; u32 data[2]; diff --git a/arch/x86/kernel/nmi_32.c b/arch/x86/kernel/nmi_32.c index 6a0aa7038685..8421d0ac6f22 100644 --- a/arch/x86/kernel/nmi_32.c +++ b/arch/x86/kernel/nmi_32.c @@ -22,9 +22,11 @@ #include <linux/cpumask.h> #include <linux/kernel_stat.h> #include <linux/kdebug.h> +#include <linux/slab.h> #include <asm/smp.h> #include <asm/nmi.h> +#include <asm/timer.h> #include "mach_traps.h" @@ -67,7 +69,7 @@ static __init void nmi_cpu_busy(void *data) } #endif -static int __init check_nmi_watchdog(void) +int __init check_nmi_watchdog(void) { unsigned int *prev_nmi_count; int cpu; @@ -80,7 +82,7 @@ static int __init check_nmi_watchdog(void) prev_nmi_count = kmalloc(NR_CPUS * sizeof(int), GFP_KERNEL); if (!prev_nmi_count) - return -1; + goto error; printk(KERN_INFO "Testing NMI watchdog ... "); @@ -117,7 +119,7 @@ static int __init check_nmi_watchdog(void) if (!atomic_read(&nmi_active)) { kfree(prev_nmi_count); atomic_set(&nmi_active, -1); - return -1; + goto error; } printk("OK.\n"); @@ -128,9 +130,11 @@ static int __init check_nmi_watchdog(void) kfree(prev_nmi_count); return 0; +error: + timer_ack = !cpu_has_tsc; + + return -1; } -/* This needs to happen later in boot so counters are working */ -late_initcall(check_nmi_watchdog); static int __init setup_nmi_watchdog(char *str) { diff --git a/arch/x86/kernel/nmi_64.c b/arch/x86/kernel/nmi_64.c index 9a4fde74bee1..11f9130ac513 100644 --- a/arch/x86/kernel/nmi_64.c +++ b/arch/x86/kernel/nmi_64.c @@ -26,6 +26,8 @@ #include <asm/proto.h> #include <asm/mce.h> +#include <mach_traps.h> + int unknown_nmi_panic; int nmi_watchdog_enabled; int panic_on_unrecovered_nmi; diff --git a/arch/x86/kernel/paravirt.c b/arch/x86/kernel/paravirt.c index 075962cc75ab..3733412d1357 100644 --- a/arch/x86/kernel/paravirt.c +++ b/arch/x86/kernel/paravirt.c @@ -206,13 +206,6 @@ static struct resource reserve_ioports = { .flags = IORESOURCE_IO | IORESOURCE_BUSY, }; -static struct resource reserve_iomem = { - .start = 0, - .end = -1, - .name = "paravirt-iomem", - .flags = IORESOURCE_MEM | IORESOURCE_BUSY, -}; - /* * Reserve the whole legacy IO space to prevent any legacy drivers * from wasting time probing for their hardware. This is a fairly @@ -222,16 +215,7 @@ static struct resource reserve_iomem = { */ int paravirt_disable_iospace(void) { - int ret; - - ret = request_resource(&ioport_resource, &reserve_ioports); - if (ret == 0) { - ret = request_resource(&iomem_resource, &reserve_iomem); - if (ret) - release_resource(&reserve_ioports); - } - - return ret; + return request_resource(&ioport_resource, &reserve_ioports); } static DEFINE_PER_CPU(enum paravirt_lazy_mode, paravirt_lazy_mode) = PARAVIRT_LAZY_NONE; diff --git a/arch/x86/kernel/pci-dma_64.c b/arch/x86/kernel/pci-dma_64.c index 375cb2bc45be..ada5a0604992 100644 --- a/arch/x86/kernel/pci-dma_64.c +++ b/arch/x86/kernel/pci-dma_64.c @@ -232,32 +232,32 @@ static __init int iommu_setup(char *p) return -EINVAL; while (*p) { - if (!strncmp(p,"off",3)) + if (!strncmp(p, "off", 3)) no_iommu = 1; /* gart_parse_options has more force support */ - if (!strncmp(p,"force",5)) + if (!strncmp(p, "force", 5)) force_iommu = 1; - if (!strncmp(p,"noforce",7)) { + if (!strncmp(p, "noforce", 7)) { iommu_merge = 0; force_iommu = 0; } - if (!strncmp(p, "biomerge",8)) { + if (!strncmp(p, "biomerge", 8)) { iommu_bio_merge = 4096; iommu_merge = 1; force_iommu = 1; } - if (!strncmp(p, "panic",5)) + if (!strncmp(p, "panic", 5)) panic_on_overflow = 1; - if (!strncmp(p, "nopanic",7)) + if (!strncmp(p, "nopanic", 7)) panic_on_overflow = 0; - if (!strncmp(p, "merge",5)) { + if (!strncmp(p, "merge", 5)) { iommu_merge = 1; force_iommu = 1; } - if (!strncmp(p, "nomerge",7)) + if (!strncmp(p, "nomerge", 7)) iommu_merge = 0; - if (!strncmp(p, "forcesac",8)) + if (!strncmp(p, "forcesac", 8)) iommu_sac_force = 1; if (!strncmp(p, "allowdac", 8)) forbid_dac = 0; @@ -265,7 +265,7 @@ static __init int iommu_setup(char *p) forbid_dac = -1; #ifdef CONFIG_SWIOTLB - if (!strncmp(p, "soft",4)) + if (!strncmp(p, "soft", 4)) swiotlb = 1; #endif diff --git a/arch/x86/kernel/process_32.c b/arch/x86/kernel/process_32.c index 43930e73f657..3903a8f2eb97 100644 --- a/arch/x86/kernel/process_32.c +++ b/arch/x86/kernel/process_32.c @@ -113,20 +113,13 @@ void default_idle(void) local_irq_disable(); if (!need_resched()) { - ktime_t t0, t1; - u64 t0n, t1n; - - t0 = ktime_get(); - t0n = ktime_to_ns(t0); safe_halt(); /* enables interrupts racelessly */ local_irq_disable(); - t1 = ktime_get(); - t1n = ktime_to_ns(t1); - sched_clock_idle_wakeup_event(t1n - t0n); } local_irq_enable(); current_thread_info()->status |= TS_POLLING; } else { + local_irq_enable(); /* loop is done by the caller */ cpu_relax(); } @@ -142,6 +135,7 @@ EXPORT_SYMBOL(default_idle); */ static void poll_idle(void) { + local_irq_enable(); cpu_relax(); } @@ -248,8 +242,11 @@ void mwait_idle_with_hints(unsigned long ax, unsigned long cx) __monitor((void *)¤t_thread_info()->flags, 0, 0); smp_mb(); if (!need_resched()) - __mwait(ax, cx); - } + __sti_mwait(ax, cx); + else + local_irq_enable(); + } else + local_irq_enable(); } /* Default MONITOR/MWAIT with no hints, used for default C1 state */ @@ -332,7 +329,7 @@ void __show_registers(struct pt_regs *regs, int all) init_utsname()->version); printk("EIP: %04x:[<%08lx>] EFLAGS: %08lx CPU: %d\n", - 0xffff & regs->cs, regs->ip, regs->flags, + (u16)regs->cs, regs->ip, regs->flags, smp_processor_id()); print_symbol("EIP is at %s\n", regs->ip); @@ -341,8 +338,7 @@ void __show_registers(struct pt_regs *regs, int all) printk("ESI: %08lx EDI: %08lx EBP: %08lx ESP: %08lx\n", regs->si, regs->di, regs->bp, sp); printk(" DS: %04x ES: %04x FS: %04x GS: %04x SS: %04x\n", - regs->ds & 0xffff, regs->es & 0xffff, - regs->fs & 0xffff, gs, ss); + (u16)regs->ds, (u16)regs->es, (u16)regs->fs, gs, ss); if (!all) return; @@ -513,6 +509,21 @@ int copy_thread(int nr, unsigned long clone_flags, unsigned long sp, return err; } +void +start_thread(struct pt_regs *regs, unsigned long new_ip, unsigned long new_sp) +{ + __asm__("movl %0, %%gs" :: "r"(0)); + regs->fs = 0; + set_fs(USER_DS); + regs->ds = __USER_DS; + regs->es = __USER_DS; + regs->ss = __USER_DS; + regs->cs = __USER_CS; + regs->ip = new_ip; + regs->sp = new_sp; +} +EXPORT_SYMBOL_GPL(start_thread); + #ifdef CONFIG_SECCOMP static void hard_disable_TSC(void) { @@ -550,12 +561,12 @@ __switch_to_xtra(struct task_struct *prev_p, struct task_struct *next_p, /* we clear debugctl to make sure DS * is not in use when we change it */ debugctl = 0; - wrmsrl(MSR_IA32_DEBUGCTLMSR, 0); + update_debugctlmsr(0); wrmsr(MSR_IA32_DS_AREA, next->ds_area_msr, 0); } if (next->debugctlmsr != debugctl) - wrmsr(MSR_IA32_DEBUGCTLMSR, next->debugctlmsr, 0); + update_debugctlmsr(next->debugctlmsr); if (test_tsk_thread_flag(next_p, TIF_DEBUG)) { set_debugreg(next->debugreg0, 0); diff --git a/arch/x86/kernel/process_64.c b/arch/x86/kernel/process_64.c index 46c4c546b499..e75ccc8a2b87 100644 --- a/arch/x86/kernel/process_64.c +++ b/arch/x86/kernel/process_64.c @@ -107,16 +107,8 @@ void default_idle(void) smp_mb(); local_irq_disable(); if (!need_resched()) { - ktime_t t0, t1; - u64 t0n, t1n; - - t0 = ktime_get(); - t0n = ktime_to_ns(t0); safe_halt(); /* enables interrupts racelessly */ local_irq_disable(); - t1 = ktime_get(); - t1n = ktime_to_ns(t1); - sched_clock_idle_wakeup_event(t1n - t0n); } local_irq_enable(); current_thread_info()->status |= TS_POLLING; @@ -528,6 +520,21 @@ out: return err; } +void +start_thread(struct pt_regs *regs, unsigned long new_ip, unsigned long new_sp) +{ + asm volatile("movl %0, %%fs; movl %0, %%es; movl %0, %%ds" :: "r"(0)); + load_gs_index(0); + regs->ip = new_ip; + regs->sp = new_sp; + write_pda(oldrsp, new_sp); + regs->cs = __USER_CS; + regs->ss = __USER_DS; + regs->flags = 0x200; + set_fs(USER_DS); +} +EXPORT_SYMBOL_GPL(start_thread); + /* * This special macro can be used to load a debugging register */ @@ -548,12 +555,12 @@ static inline void __switch_to_xtra(struct task_struct *prev_p, /* we clear debugctl to make sure DS * is not in use when we change it */ debugctl = 0; - wrmsrl(MSR_IA32_DEBUGCTLMSR, 0); + update_debugctlmsr(0); wrmsrl(MSR_IA32_DS_AREA, next->ds_area_msr); } if (next->debugctlmsr != debugctl) - wrmsrl(MSR_IA32_DEBUGCTLMSR, next->debugctlmsr); + update_debugctlmsr(next->debugctlmsr); if (test_tsk_thread_flag(next_p, TIF_DEBUG)) { loaddebug(next, 0); diff --git a/arch/x86/kernel/ptrace.c b/arch/x86/kernel/ptrace.c index eb92ccbb3502..559c1b027417 100644 --- a/arch/x86/kernel/ptrace.c +++ b/arch/x86/kernel/ptrace.c @@ -1456,7 +1456,6 @@ void send_sigtrap(struct task_struct *tsk, struct pt_regs *regs, int error_code) /* notification of system call entry/exit * - triggered by current->work.syscall_trace */ -__attribute__((regparm(3))) int do_syscall_trace(struct pt_regs *regs, int entryexit) { int is_sysemu = test_thread_flag(TIF_SYSCALL_EMU); diff --git a/arch/x86/kernel/reboot.c b/arch/x86/kernel/reboot.c index 484c4a80d38a..9692202d3bfb 100644 --- a/arch/x86/kernel/reboot.c +++ b/arch/x86/kernel/reboot.c @@ -1,5 +1,4 @@ #include <linux/module.h> -#include <linux/init.h> #include <linux/reboot.h> #include <linux/init.h> #include <linux/pm.h> @@ -412,12 +411,12 @@ static void native_machine_shutdown(void) #ifdef CONFIG_X86_32 /* See if there has been given a command line override */ if ((reboot_cpu != -1) && (reboot_cpu < NR_CPUS) && - cpu_isset(reboot_cpu, cpu_online_map)) + cpu_online(reboot_cpu)) reboot_cpu_id = reboot_cpu; #endif /* Make certain the cpu I'm about to reboot on is online */ - if (!cpu_isset(reboot_cpu_id, cpu_online_map)) + if (!cpu_online(reboot_cpu_id)) reboot_cpu_id = smp_processor_id(); /* Make certain I only run on the appropriate processor */ diff --git a/arch/x86/kernel/relocate_kernel_32.S b/arch/x86/kernel/relocate_kernel_32.S index f151d6fae462..c30fe25d470d 100644 --- a/arch/x86/kernel/relocate_kernel_32.S +++ b/arch/x86/kernel/relocate_kernel_32.S @@ -9,18 +9,19 @@ #include <linux/linkage.h> #include <asm/page.h> #include <asm/kexec.h> +#include <asm/processor-flags.h> +#include <asm/pgtable.h> /* * Must be relocatable PIC code callable as a C function */ #define PTR(x) (x << 2) -#define PAGE_ALIGNED (1 << PAGE_SHIFT) -#define PAGE_ATTR 0x63 /* _PAGE_PRESENT|_PAGE_RW|_PAGE_ACCESSED|_PAGE_DIRTY */ -#define PAE_PGD_ATTR 0x01 /* _PAGE_PRESENT */ +#define PAGE_ATTR (_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED | _PAGE_DIRTY) +#define PAE_PGD_ATTR (_PAGE_PRESENT) .text - .align PAGE_ALIGNED + .align PAGE_SIZE .globl relocate_kernel relocate_kernel: movl 8(%esp), %ebp /* list of pages */ @@ -155,7 +156,7 @@ relocate_new_kernel: movl %eax, %cr3 /* setup a new stack at the end of the physical control page */ - lea 4096(%edi), %esp + lea PAGE_SIZE(%edi), %esp /* jump to identity mapped page */ movl %edi, %eax @@ -168,16 +169,16 @@ identity_mapped: pushl %edx /* Set cr0 to a known state: - * 31 0 == Paging disabled - * 18 0 == Alignment check disabled - * 16 0 == Write protect disabled - * 3 0 == No task switch - * 2 0 == Don't do FP software emulation. - * 0 1 == Proctected mode enabled + * - Paging disabled + * - Alignment check disabled + * - Write protect disabled + * - No task switch + * - Don't do FP software emulation. + * - Proctected mode enabled */ movl %cr0, %eax - andl $~((1<<31)|(1<<18)|(1<<16)|(1<<3)|(1<<2)), %eax - orl $(1<<0), %eax + andl $~(X86_CR0_PG | X86_CR0_AM | X86_CR0_WP | X86_CR0_TS | X86_CR0_EM), %eax + orl $(X86_CR0_PE), %eax movl %eax, %cr0 /* clear cr4 if applicable */ @@ -186,8 +187,7 @@ identity_mapped: /* Set cr4 to a known state: * Setting everything to zero seems safe. */ - movl %cr4, %eax - andl $0, %eax + xorl %eax, %eax movl %eax, %cr4 jmp 1f diff --git a/arch/x86/kernel/relocate_kernel_64.S b/arch/x86/kernel/relocate_kernel_64.S index 14e95872c6a3..f5afe665a82b 100644 --- a/arch/x86/kernel/relocate_kernel_64.S +++ b/arch/x86/kernel/relocate_kernel_64.S @@ -9,17 +9,18 @@ #include <linux/linkage.h> #include <asm/page.h> #include <asm/kexec.h> +#include <asm/processor-flags.h> +#include <asm/pgtable.h> /* * Must be relocatable PIC code callable as a C function */ #define PTR(x) (x << 3) -#define PAGE_ALIGNED (1 << PAGE_SHIFT) -#define PAGE_ATTR 0x63 /* _PAGE_PRESENT|_PAGE_RW|_PAGE_ACCESSED|_PAGE_DIRTY */ +#define PAGE_ATTR (_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED | _PAGE_DIRTY) .text - .align PAGE_ALIGNED + .align PAGE_SIZE .code64 .globl relocate_kernel relocate_kernel: @@ -160,7 +161,7 @@ relocate_new_kernel: movq %r9, %cr3 /* setup a new stack at the end of the physical control page */ - lea 4096(%r8), %rsp + lea PAGE_SIZE(%r8), %rsp /* jump to identity mapped page */ addq $(identity_mapped - relocate_kernel), %r8 @@ -172,33 +173,22 @@ identity_mapped: pushq %rdx /* Set cr0 to a known state: - * 31 1 == Paging enabled - * 18 0 == Alignment check disabled - * 16 0 == Write protect disabled - * 3 0 == No task switch - * 2 0 == Don't do FP software emulation. - * 0 1 == Proctected mode enabled + * - Paging enabled + * - Alignment check disabled + * - Write protect disabled + * - No task switch + * - Don't do FP software emulation. + * - Proctected mode enabled */ movq %cr0, %rax - andq $~((1<<18)|(1<<16)|(1<<3)|(1<<2)), %rax - orl $((1<<31)|(1<<0)), %eax + andq $~(X86_CR0_AM | X86_CR0_WP | X86_CR0_TS | X86_CR0_EM), %rax + orl $(X86_CR0_PG | X86_CR0_PE), %eax movq %rax, %cr0 /* Set cr4 to a known state: - * 10 0 == xmm exceptions disabled - * 9 0 == xmm registers instructions disabled - * 8 0 == performance monitoring counter disabled - * 7 0 == page global disabled - * 6 0 == machine check exceptions disabled - * 5 1 == physical address extension enabled - * 4 0 == page size extensions disabled - * 3 0 == Debug extensions disabled - * 2 0 == Time stamp disable (disabled) - * 1 0 == Protected mode virtual interrupts disabled - * 0 0 == VME disabled + * - physical address extension enabled */ - - movq $((1<<5)), %rax + movq $X86_CR4_PAE, %rax movq %rax, %cr4 jmp 1f diff --git a/arch/x86/kernel/rtc.c b/arch/x86/kernel/rtc.c index eb9b1a198f5e..9615eee9b775 100644 --- a/arch/x86/kernel/rtc.c +++ b/arch/x86/kernel/rtc.c @@ -9,7 +9,6 @@ #include <asm/vsyscall.h> #ifdef CONFIG_X86_32 -# define CMOS_YEARS_OFFS 1900 /* * This is a special lock that is owned by the CPU and holds the index * register we are working with. It is required for NMI access to the @@ -17,14 +16,11 @@ */ volatile unsigned long cmos_lock = 0; EXPORT_SYMBOL(cmos_lock); -#else -/* - * x86-64 systems only exists since 2002. - * This will work up to Dec 31, 2100 - */ -# define CMOS_YEARS_OFFS 2000 #endif +/* For two digit years assume time is always after that */ +#define CMOS_YEARS_OFFS 2000 + DEFINE_SPINLOCK(rtc_lock); EXPORT_SYMBOL(rtc_lock); @@ -98,7 +94,7 @@ int mach_set_rtc_mmss(unsigned long nowtime) unsigned long mach_get_cmos_time(void) { - unsigned int year, mon, day, hour, min, sec, century = 0; + unsigned int status, year, mon, day, hour, min, sec, century = 0; /* * If UIP is clear, then we have >= 244 microseconds before @@ -116,14 +112,16 @@ unsigned long mach_get_cmos_time(void) mon = CMOS_READ(RTC_MONTH); year = CMOS_READ(RTC_YEAR); -#if defined(CONFIG_ACPI) && defined(CONFIG_X86_64) - /* CHECKME: Is this really 64bit only ??? */ +#ifdef CONFIG_ACPI if (acpi_gbl_FADT.header.revision >= FADT2_REVISION_ID && acpi_gbl_FADT.century) century = CMOS_READ(acpi_gbl_FADT.century); #endif - if (RTC_ALWAYS_BCD || !(CMOS_READ(RTC_CONTROL) & RTC_DM_BINARY)) { + status = CMOS_READ(RTC_CONTROL); + WARN_ON_ONCE(RTC_ALWAYS_BCD && (status & RTC_DM_BINARY)); + + if (RTC_ALWAYS_BCD || !(status & RTC_DM_BINARY)) { BCD_TO_BIN(sec); BCD_TO_BIN(min); BCD_TO_BIN(hour); @@ -136,11 +134,8 @@ unsigned long mach_get_cmos_time(void) BCD_TO_BIN(century); year += century * 100; printk(KERN_INFO "Extended CMOS year: %d\n", century * 100); - } else { + } else year += CMOS_YEARS_OFFS; - if (year < 1970) - year += 100; - } return mktime(year, mon, day, hour, min, sec); } @@ -151,8 +146,8 @@ unsigned char rtc_cmos_read(unsigned char addr) unsigned char val; lock_cmos_prefix(addr); - outb_p(addr, RTC_PORT(0)); - val = inb_p(RTC_PORT(1)); + outb(addr, RTC_PORT(0)); + val = inb(RTC_PORT(1)); lock_cmos_suffix(addr); return val; } @@ -161,8 +156,8 @@ EXPORT_SYMBOL(rtc_cmos_read); void rtc_cmos_write(unsigned char val, unsigned char addr) { lock_cmos_prefix(addr); - outb_p(addr, RTC_PORT(0)); - outb_p(val, RTC_PORT(1)); + outb(addr, RTC_PORT(0)); + outb(val, RTC_PORT(1)); lock_cmos_suffix(addr); } EXPORT_SYMBOL(rtc_cmos_write); diff --git a/arch/x86/kernel/setup.c b/arch/x86/kernel/setup.c new file mode 100644 index 000000000000..ed157c90412e --- /dev/null +++ b/arch/x86/kernel/setup.c @@ -0,0 +1,113 @@ +#include <linux/kernel.h> +#include <linux/module.h> +#include <linux/init.h> +#include <linux/bootmem.h> +#include <linux/percpu.h> +#include <asm/smp.h> +#include <asm/percpu.h> +#include <asm/sections.h> +#include <asm/processor.h> +#include <asm/setup.h> +#include <asm/topology.h> +#include <asm/mpspec.h> +#include <asm/apicdef.h> + +unsigned int num_processors; +unsigned disabled_cpus __cpuinitdata; +/* Processor that is doing the boot up */ +unsigned int boot_cpu_physical_apicid = -1U; +EXPORT_SYMBOL(boot_cpu_physical_apicid); + +physid_mask_t phys_cpu_present_map; + +DEFINE_PER_CPU(u16, x86_cpu_to_apicid) = BAD_APICID; +EXPORT_PER_CPU_SYMBOL(x86_cpu_to_apicid); + +/* Bitmask of physically existing CPUs */ +physid_mask_t phys_cpu_present_map; + +#if defined(CONFIG_HAVE_SETUP_PER_CPU_AREA) && defined(CONFIG_SMP) +/* + * Copy data used in early init routines from the initial arrays to the + * per cpu data areas. These arrays then become expendable and the + * *_early_ptr's are zeroed indicating that the static arrays are gone. + */ +static void __init setup_per_cpu_maps(void) +{ + int cpu; + + for_each_possible_cpu(cpu) { + per_cpu(x86_cpu_to_apicid, cpu) = x86_cpu_to_apicid_init[cpu]; + per_cpu(x86_bios_cpu_apicid, cpu) = + x86_bios_cpu_apicid_init[cpu]; +#ifdef CONFIG_NUMA + per_cpu(x86_cpu_to_node_map, cpu) = + x86_cpu_to_node_map_init[cpu]; +#endif + } + + /* indicate the early static arrays will soon be gone */ + x86_cpu_to_apicid_early_ptr = NULL; + x86_bios_cpu_apicid_early_ptr = NULL; +#ifdef CONFIG_NUMA + x86_cpu_to_node_map_early_ptr = NULL; +#endif +} + +#ifdef CONFIG_X86_32 +/* + * Great future not-so-futuristic plan: make i386 and x86_64 do it + * the same way + */ +unsigned long __per_cpu_offset[NR_CPUS] __read_mostly; +EXPORT_SYMBOL(__per_cpu_offset); +#endif + +/* + * Great future plan: + * Declare PDA itself and support (irqstack,tss,pgd) as per cpu data. + * Always point %gs to its beginning + */ +void __init setup_per_cpu_areas(void) +{ + int i; + unsigned long size; + +#ifdef CONFIG_HOTPLUG_CPU + prefill_possible_map(); +#endif + + /* Copy section for each CPU (we discard the original) */ + size = PERCPU_ENOUGH_ROOM; + printk(KERN_INFO "PERCPU: Allocating %lu bytes of per cpu data\n", + size); + + for_each_possible_cpu(i) { + char *ptr; +#ifndef CONFIG_NEED_MULTIPLE_NODES + ptr = alloc_bootmem_pages(size); +#else + int node = early_cpu_to_node(i); + if (!node_online(node) || !NODE_DATA(node)) { + ptr = alloc_bootmem_pages(size); + printk(KERN_INFO + "cpu %d has no node or node-local memory\n", i); + } + else + ptr = alloc_bootmem_pages_node(NODE_DATA(node), size); +#endif + if (!ptr) + panic("Cannot allocate cpu data for CPU %d\n", i); +#ifdef CONFIG_X86_64 + cpu_pda(i)->data_offset = ptr - __per_cpu_start; +#else + __per_cpu_offset[i] = ptr - __per_cpu_start; +#endif + memcpy(ptr, __per_cpu_start, __per_cpu_end - __per_cpu_start); + } + + /* Setup percpu data maps */ + setup_per_cpu_maps(); +} + +#endif diff --git a/arch/x86/kernel/setup64.c b/arch/x86/kernel/setup64.c index 143aa78c566b..9042fb0e36f5 100644 --- a/arch/x86/kernel/setup64.c +++ b/arch/x86/kernel/setup64.c @@ -24,6 +24,7 @@ #include <asm/proto.h> #include <asm/sections.h> #include <asm/setup.h> +#include <asm/genapic.h> #ifndef CONFIG_DEBUG_BOOT_PARAMS struct boot_params __initdata boot_params; @@ -86,83 +87,6 @@ static int __init nonx32_setup(char *str) } __setup("noexec32=", nonx32_setup); -/* - * Copy data used in early init routines from the initial arrays to the - * per cpu data areas. These arrays then become expendable and the - * *_early_ptr's are zeroed indicating that the static arrays are gone. - */ -static void __init setup_per_cpu_maps(void) -{ - int cpu; - - for_each_possible_cpu(cpu) { -#ifdef CONFIG_SMP - if (per_cpu_offset(cpu)) { -#endif - per_cpu(x86_cpu_to_apicid, cpu) = - x86_cpu_to_apicid_init[cpu]; - per_cpu(x86_bios_cpu_apicid, cpu) = - x86_bios_cpu_apicid_init[cpu]; -#ifdef CONFIG_NUMA - per_cpu(x86_cpu_to_node_map, cpu) = - x86_cpu_to_node_map_init[cpu]; -#endif -#ifdef CONFIG_SMP - } - else - printk(KERN_NOTICE "per_cpu_offset zero for cpu %d\n", - cpu); -#endif - } - - /* indicate the early static arrays will soon be gone */ - x86_cpu_to_apicid_early_ptr = NULL; - x86_bios_cpu_apicid_early_ptr = NULL; -#ifdef CONFIG_NUMA - x86_cpu_to_node_map_early_ptr = NULL; -#endif -} - -/* - * Great future plan: - * Declare PDA itself and support (irqstack,tss,pgd) as per cpu data. - * Always point %gs to its beginning - */ -void __init setup_per_cpu_areas(void) -{ - int i; - unsigned long size; - -#ifdef CONFIG_HOTPLUG_CPU - prefill_possible_map(); -#endif - - /* Copy section for each CPU (we discard the original) */ - size = PERCPU_ENOUGH_ROOM; - - printk(KERN_INFO "PERCPU: Allocating %lu bytes of per cpu data\n", size); - for_each_cpu_mask (i, cpu_possible_map) { - char *ptr; -#ifndef CONFIG_NEED_MULTIPLE_NODES - ptr = alloc_bootmem_pages(size); -#else - int node = early_cpu_to_node(i); - - if (!node_online(node) || !NODE_DATA(node)) - ptr = alloc_bootmem_pages(size); - else - ptr = alloc_bootmem_pages_node(NODE_DATA(node), size); -#endif - if (!ptr) - panic("Cannot allocate cpu data for CPU %d\n", i); - cpu_pda(i)->data_offset = ptr - __per_cpu_start; - memcpy(ptr, __per_cpu_start, __per_cpu_end - __per_cpu_start); - } - - /* setup percpu data maps early */ - setup_per_cpu_maps(); -} - void pda_init(int cpu) { struct x8664_pda *pda = cpu_pda(cpu); @@ -357,4 +281,7 @@ void __cpuinit cpu_init (void) fpu_init(); raw_local_save_flags(kernel_eflags); + + if (is_uv_system()) + uv_cpu_init(); } diff --git a/arch/x86/kernel/setup_32.c b/arch/x86/kernel/setup_32.c index 2b3e5d45176b..5b0bffb7fcc9 100644 --- a/arch/x86/kernel/setup_32.c +++ b/arch/x86/kernel/setup_32.c @@ -62,8 +62,9 @@ #include <asm/io.h> #include <asm/vmi.h> #include <setup_arch.h> -#include <bios_ebda.h> +#include <asm/bios_ebda.h> #include <asm/cacheflush.h> +#include <asm/processor.h> /* This value is set up by the early boot code to point to the value immediately after the boot time page tables. It contains a *physical* @@ -154,6 +155,8 @@ struct cpuinfo_x86 new_cpu_data __cpuinitdata = { 0, 0, 0, 0, -1, 1, 0, 0, -1 }; struct cpuinfo_x86 boot_cpu_data __read_mostly = { 0, 0, 0, 0, -1, 1, 0, 0, -1 }; EXPORT_SYMBOL(boot_cpu_data); +unsigned int def_to_bigsmp; + #ifndef CONFIG_X86_PAE unsigned long mmu_cr4_features; #else @@ -189,7 +192,7 @@ EXPORT_SYMBOL(ist_info); extern void early_cpu_init(void); extern int root_mountflags; -unsigned long saved_videomode; +unsigned long saved_video_mode; #define RAMDISK_IMAGE_START_MASK 0x07FF #define RAMDISK_PROMPT_FLAG 0x8000 @@ -227,7 +230,7 @@ static inline void copy_edd(void) } #endif -int __initdata user_defined_memmap = 0; +int __initdata user_defined_memmap; /* * "mem=nopentium" disables the 4MB page tables. @@ -385,15 +388,56 @@ unsigned long __init find_max_low_pfn(void) return max_low_pfn; } +#define BIOS_EBDA_SEGMENT 0x40E +#define BIOS_LOWMEM_KILOBYTES 0x413 + /* - * workaround for Dell systems that neglect to reserve EBDA + * The BIOS places the EBDA/XBDA at the top of conventional + * memory, and usually decreases the reported amount of + * conventional memory (int 0x12) too. This also contains a + * workaround for Dell systems that neglect to reserve EBDA. + * The same workaround also avoids a problem with the AMD768MPX + * chipset: reserve a page before VGA to prevent PCI prefetch + * into it (errata #56). Usually the page is reserved anyways, + * unless you have no PS/2 mouse plugged in. */ static void __init reserve_ebda_region(void) { - unsigned int addr; - addr = get_bios_ebda(); - if (addr) - reserve_bootmem(addr, PAGE_SIZE, BOOTMEM_DEFAULT); + unsigned int lowmem, ebda_addr; + + /* To determine the position of the EBDA and the */ + /* end of conventional memory, we need to look at */ + /* the BIOS data area. In a paravirtual environment */ + /* that area is absent. We'll just have to assume */ + /* that the paravirt case can handle memory setup */ + /* correctly, without our help. */ + if (paravirt_enabled()) + return; + + /* end of low (conventional) memory */ + lowmem = *(unsigned short *)__va(BIOS_LOWMEM_KILOBYTES); + lowmem <<= 10; + + /* start of EBDA area */ + ebda_addr = *(unsigned short *)__va(BIOS_EBDA_SEGMENT); + ebda_addr <<= 4; + + /* Fixup: bios puts an EBDA in the top 64K segment */ + /* of conventional memory, but does not adjust lowmem. */ + if ((lowmem - ebda_addr) <= 0x10000) + lowmem = ebda_addr; + + /* Fixup: bios does not report an EBDA at all. */ + /* Some old Dells seem to need 4k anyhow (bugzilla 2990) */ + if ((ebda_addr == 0) && (lowmem >= 0x9f000)) + lowmem = 0x9f000; + + /* Paranoia: should never happen, but... */ + if ((lowmem == 0) || (lowmem >= 0x100000)) + lowmem = 0x9f000; + + /* reserve all memory between lowmem and the 1MB mark */ + reserve_bootmem(lowmem, 0x100000 - lowmem, BOOTMEM_DEFAULT); } #ifndef CONFIG_NEED_MULTIPLE_NODES @@ -617,16 +661,9 @@ void __init setup_bootmem_allocator(void) */ reserve_bootmem(0, PAGE_SIZE, BOOTMEM_DEFAULT); - /* reserve EBDA region, it's a 4K region */ + /* reserve EBDA region */ reserve_ebda_region(); - /* could be an AMD 768MPX chipset. Reserve a page before VGA to prevent - PCI prefetch into it (errata #56). Usually the page is reserved anyways, - unless you have no PS/2 mouse plugged in. */ - if (boot_cpu_data.x86_vendor == X86_VENDOR_AMD && - boot_cpu_data.x86 == 6) - reserve_bootmem(0xa0000 - 4096, 4096, BOOTMEM_DEFAULT); - #ifdef CONFIG_SMP /* * But first pinch a few for the stack/trampoline stuff @@ -687,6 +724,18 @@ char * __init __attribute__((weak)) memory_setup(void) return machine_specific_memory_setup(); } +#ifdef CONFIG_NUMA +/* + * In the golden day, when everything among i386 and x86_64 will be + * integrated, this will not live here + */ +void *x86_cpu_to_node_map_early_ptr; +int x86_cpu_to_node_map_init[NR_CPUS] = { + [0 ... NR_CPUS-1] = NUMA_NO_NODE +}; +DEFINE_PER_CPU(int, x86_cpu_to_node_map) = NUMA_NO_NODE; +#endif + /* * Determine if we were loaded by an EFI loader. If so, then we have also been * passed the efi memmap, systab, etc., so we should use these data structures @@ -714,7 +763,7 @@ void __init setup_arch(char **cmdline_p) edid_info = boot_params.edid_info; apm_info.bios = boot_params.apm_bios_info; ist_info = boot_params.ist_info; - saved_videomode = boot_params.hdr.vid_mode; + saved_video_mode = boot_params.hdr.vid_mode; if( boot_params.sys_desc_table.length != 0 ) { set_mca_bus(boot_params.sys_desc_table.table[3] & 0x2); machine_id = boot_params.sys_desc_table.table[0]; @@ -820,6 +869,18 @@ void __init setup_arch(char **cmdline_p) io_delay_init(); +#ifdef CONFIG_X86_SMP + /* + * setup to use the early static init tables during kernel startup + * X86_SMP will exclude sub-arches that don't deal well with it. + */ + x86_cpu_to_apicid_early_ptr = (void *)x86_cpu_to_apicid_init; + x86_bios_cpu_apicid_early_ptr = (void *)x86_bios_cpu_apicid_init; +#ifdef CONFIG_NUMA + x86_cpu_to_node_map_early_ptr = (void *)x86_cpu_to_node_map_init; +#endif +#endif + #ifdef CONFIG_X86_GENERICARCH generic_apic_probe(); #endif diff --git a/arch/x86/kernel/setup_64.c b/arch/x86/kernel/setup_64.c index f4f7ecfb898c..674ef3510cdf 100644 --- a/arch/x86/kernel/setup_64.c +++ b/arch/x86/kernel/setup_64.c @@ -58,7 +58,6 @@ #include <asm/mmu_context.h> #include <asm/proto.h> #include <asm/setup.h> -#include <asm/mach_apic.h> #include <asm/numa.h> #include <asm/sections.h> #include <asm/dmi.h> @@ -66,7 +65,9 @@ #include <asm/mce.h> #include <asm/ds.h> #include <asm/topology.h> +#include <asm/trampoline.h> +#include <mach_apic.h> #ifdef CONFIG_PARAVIRT #include <asm/paravirt.h> #else @@ -248,6 +249,7 @@ static void __init reserve_crashkernel(void) (unsigned long)(total_mem >> 20)); crashk_res.start = crash_base; crashk_res.end = crash_base + crash_size - 1; + insert_resource(&iomem_resource, &crashk_res); } } #else @@ -322,6 +324,11 @@ void __init setup_arch(char **cmdline_p) finish_e820_parsing(); + /* after parse_early_param, so could debug it */ + insert_resource(&iomem_resource, &code_resource); + insert_resource(&iomem_resource, &data_resource); + insert_resource(&iomem_resource, &bss_resource); + early_gart_iommu_check(); e820_register_active_regions(0, 0, -1UL); @@ -341,10 +348,12 @@ void __init setup_arch(char **cmdline_p) check_efer(); - init_memory_mapping(0, (end_pfn_map << PAGE_SHIFT)); + max_pfn_mapped = init_memory_mapping(0, (max_pfn_mapped << PAGE_SHIFT)); if (efi_enabled) efi_init(); + vsmp_init(); + dmi_scan_machine(); io_delay_init(); @@ -450,7 +459,7 @@ void __init setup_arch(char **cmdline_p) /* * We trust e820 completely. No explicit ROM probing in memory. */ - e820_reserve_resources(&code_resource, &data_resource, &bss_resource); + e820_reserve_resources(); e820_mark_nosave_regions(); /* request I/O space for devices used on all i[345]86 PCs */ @@ -552,9 +561,9 @@ static void __cpuinit amd_detect_cmp(struct cpuinfo_x86 *c) bits = c->x86_coreid_bits; /* Low order bits define the core id (index of core in socket) */ - c->cpu_core_id = c->phys_proc_id & ((1 << bits)-1); - /* Convert the APIC ID into the socket ID */ - c->phys_proc_id = phys_pkg_id(bits); + c->cpu_core_id = c->initial_apicid & ((1 << bits)-1); + /* Convert the initial APIC ID into the socket ID */ + c->phys_proc_id = c->initial_apicid >> bits; #ifdef CONFIG_NUMA node = c->phys_proc_id; @@ -571,7 +580,7 @@ static void __cpuinit amd_detect_cmp(struct cpuinfo_x86 *c) If that doesn't result in a usable node fall back to the path for the previous case. */ - int ht_nodeid = apicid - (cpu_data(0).phys_proc_id << bits); + int ht_nodeid = c->initial_apicid; if (ht_nodeid >= 0 && apicid_to_node[ht_nodeid] != NUMA_NO_NODE) @@ -677,7 +686,7 @@ static void __cpuinit init_amd(struct cpuinfo_x86 *c) /* Bit 31 in normal CPUID used for nonstandard 3DNow ID; 3DNow is IDd by bit 31 in extended CPUID (1*32+31) anyway */ - clear_bit(0*32+31, (unsigned long *)&c->x86_capability); + clear_cpu_cap(c, 0*32+31); /* On C+ stepping K8 rep microcode works well for copy/memset */ level = cpuid_eax(1); @@ -721,6 +730,19 @@ static void __cpuinit init_amd(struct cpuinfo_x86 *c) if (amd_apic_timer_broken()) disable_apic_timer = 1; + + if (c == &boot_cpu_data && c->x86 >= 0xf && c->x86 <= 0x11) { + unsigned long long tseg; + + /* + * Split up direct mapping around the TSEG SMM area. + * Don't do it for gbpages because there seems very little + * benefit in doing so. + */ + if (!rdmsrl_safe(MSR_K8_TSEG_ADDR, &tseg) && + (tseg >> PMD_SHIFT) < (max_pfn_mapped >> (PMD_SHIFT-PAGE_SHIFT))) + set_memory_4k((unsigned long)__va(tseg), 1); + } } void __cpuinit detect_ht(struct cpuinfo_x86 *c) @@ -813,7 +835,7 @@ static void __cpuinit early_init_intel(struct cpuinfo_x86 *c) { if ((c->x86 == 0xf && c->x86_model >= 0x03) || (c->x86 == 0x6 && c->x86_model >= 0x0e)) - set_bit(X86_FEATURE_CONSTANT_TSC, &c->x86_capability); + set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC); } static void __cpuinit init_intel(struct cpuinfo_x86 *c) @@ -856,9 +878,6 @@ static void __cpuinit init_intel(struct cpuinfo_x86 *c) if (c->x86 == 15) c->x86_cache_alignment = c->x86_clflush_size * 2; - if ((c->x86 == 0xf && c->x86_model >= 0x03) || - (c->x86 == 0x6 && c->x86_model >= 0x0e)) - set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC); if (c->x86 == 6) set_cpu_cap(c, X86_FEATURE_REP_GOOD); set_cpu_cap(c, X86_FEATURE_LFENCE_RDTSC); @@ -867,6 +886,32 @@ static void __cpuinit init_intel(struct cpuinfo_x86 *c) srat_detect_node(); } +static void __cpuinit early_init_centaur(struct cpuinfo_x86 *c) +{ + if (c->x86 == 0x6 && c->x86_model >= 0xf) + set_bit(X86_FEATURE_CONSTANT_TSC, &c->x86_capability); +} + +static void __cpuinit init_centaur(struct cpuinfo_x86 *c) +{ + /* Cache sizes */ + unsigned n; + + n = c->extended_cpuid_level; + if (n >= 0x80000008) { + unsigned eax = cpuid_eax(0x80000008); + c->x86_virt_bits = (eax >> 8) & 0xff; + c->x86_phys_bits = eax & 0xff; + } + + if (c->x86 == 0x6 && c->x86_model >= 0xf) { + c->x86_cache_alignment = c->x86_clflush_size * 2; + set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC); + set_cpu_cap(c, X86_FEATURE_REP_GOOD); + } + set_cpu_cap(c, X86_FEATURE_LFENCE_RDTSC); +} + static void __cpuinit get_cpu_vendor(struct cpuinfo_x86 *c) { char *v = c->x86_vendor_id; @@ -875,6 +920,8 @@ static void __cpuinit get_cpu_vendor(struct cpuinfo_x86 *c) c->x86_vendor = X86_VENDOR_AMD; else if (!strcmp(v, "GenuineIntel")) c->x86_vendor = X86_VENDOR_INTEL; + else if (!strcmp(v, "CentaurHauls")) + c->x86_vendor = X86_VENDOR_CENTAUR; else c->x86_vendor = X86_VENDOR_UNKNOWN; } @@ -922,15 +969,16 @@ static void __cpuinit early_identify_cpu(struct cpuinfo_x86 *c) c->x86 += (tfms >> 20) & 0xff; if (c->x86 >= 0x6) c->x86_model += ((tfms >> 16) & 0xF) << 4; - if (c->x86_capability[0] & (1<<19)) + if (test_cpu_cap(c, X86_FEATURE_CLFLSH)) c->x86_clflush_size = ((misc >> 8) & 0xff) * 8; } else { /* Have CPUID level 0 only - unheard of */ c->x86 = 4; } + c->initial_apicid = (cpuid_ebx(1) >> 24) & 0xff; #ifdef CONFIG_SMP - c->phys_proc_id = (cpuid_ebx(1) >> 24) & 0xff; + c->phys_proc_id = c->initial_apicid; #endif /* AMD-defined flags: level 0x80000001 */ xlvl = cpuid_eax(0x80000000); @@ -956,12 +1004,22 @@ static void __cpuinit early_identify_cpu(struct cpuinfo_x86 *c) if (c->extended_cpuid_level >= 0x80000007) c->x86_power = cpuid_edx(0x80000007); + + clear_cpu_cap(c, X86_FEATURE_PAT); + switch (c->x86_vendor) { case X86_VENDOR_AMD: early_init_amd(c); + if (c->x86 >= 0xf && c->x86 <= 0x11) + set_cpu_cap(c, X86_FEATURE_PAT); break; case X86_VENDOR_INTEL: early_init_intel(c); + if (c->x86 == 0xF || (c->x86 == 6 && c->x86_model >= 15)) + set_cpu_cap(c, X86_FEATURE_PAT); + break; + case X86_VENDOR_CENTAUR: + early_init_centaur(c); break; } @@ -999,6 +1057,10 @@ void __cpuinit identify_cpu(struct cpuinfo_x86 *c) init_intel(c); break; + case X86_VENDOR_CENTAUR: + init_centaur(c); + break; + case X86_VENDOR_UNKNOWN: default: display_cacheinfo(c); @@ -1028,14 +1090,24 @@ void __cpuinit identify_cpu(struct cpuinfo_x86 *c) #endif select_idle_routine(c); - if (c != &boot_cpu_data) - mtrr_ap_init(); #ifdef CONFIG_NUMA numa_add_cpu(smp_processor_id()); #endif } +void __cpuinit identify_boot_cpu(void) +{ + identify_cpu(&boot_cpu_data); +} + +void __cpuinit identify_secondary_cpu(struct cpuinfo_x86 *c) +{ + BUG_ON(c == &boot_cpu_data); + identify_cpu(c); + mtrr_ap_init(); +} + static __init int setup_noclflush(char *arg) { setup_clear_cpu_cap(X86_FEATURE_CLFLSH); @@ -1064,123 +1136,3 @@ static __init int setup_disablecpuid(char *arg) return 1; } __setup("clearcpuid=", setup_disablecpuid); - -/* - * Get CPU information for use by the procfs. - */ - -static int show_cpuinfo(struct seq_file *m, void *v) -{ - struct cpuinfo_x86 *c = v; - int cpu = 0, i; - -#ifdef CONFIG_SMP - cpu = c->cpu_index; -#endif - - seq_printf(m, "processor\t: %u\n" - "vendor_id\t: %s\n" - "cpu family\t: %d\n" - "model\t\t: %d\n" - "model name\t: %s\n", - (unsigned)cpu, - c->x86_vendor_id[0] ? c->x86_vendor_id : "unknown", - c->x86, - (int)c->x86_model, - c->x86_model_id[0] ? c->x86_model_id : "unknown"); - - if (c->x86_mask || c->cpuid_level >= 0) - seq_printf(m, "stepping\t: %d\n", c->x86_mask); - else - seq_printf(m, "stepping\t: unknown\n"); - - if (cpu_has(c, X86_FEATURE_TSC)) { - unsigned int freq = cpufreq_quick_get((unsigned)cpu); - - if (!freq) - freq = cpu_khz; - seq_printf(m, "cpu MHz\t\t: %u.%03u\n", - freq / 1000, (freq % 1000)); - } - - /* Cache size */ - if (c->x86_cache_size >= 0) - seq_printf(m, "cache size\t: %d KB\n", c->x86_cache_size); - -#ifdef CONFIG_SMP - if (smp_num_siblings * c->x86_max_cores > 1) { - seq_printf(m, "physical id\t: %d\n", c->phys_proc_id); - seq_printf(m, "siblings\t: %d\n", - cpus_weight(per_cpu(cpu_core_map, cpu))); - seq_printf(m, "core id\t\t: %d\n", c->cpu_core_id); - seq_printf(m, "cpu cores\t: %d\n", c->booted_cores); - } -#endif - - seq_printf(m, - "fpu\t\t: yes\n" - "fpu_exception\t: yes\n" - "cpuid level\t: %d\n" - "wp\t\t: yes\n" - "flags\t\t:", - c->cpuid_level); - - for (i = 0; i < 32*NCAPINTS; i++) - if (cpu_has(c, i) && x86_cap_flags[i] != NULL) - seq_printf(m, " %s", x86_cap_flags[i]); - - seq_printf(m, "\nbogomips\t: %lu.%02lu\n", - c->loops_per_jiffy/(500000/HZ), - (c->loops_per_jiffy/(5000/HZ)) % 100); - - if (c->x86_tlbsize > 0) - seq_printf(m, "TLB size\t: %d 4K pages\n", c->x86_tlbsize); - seq_printf(m, "clflush size\t: %d\n", c->x86_clflush_size); - seq_printf(m, "cache_alignment\t: %d\n", c->x86_cache_alignment); - - seq_printf(m, "address sizes\t: %u bits physical, %u bits virtual\n", - c->x86_phys_bits, c->x86_virt_bits); - - seq_printf(m, "power management:"); - for (i = 0; i < 32; i++) { - if (c->x86_power & (1 << i)) { - if (i < ARRAY_SIZE(x86_power_flags) && - x86_power_flags[i]) - seq_printf(m, "%s%s", - x86_power_flags[i][0]?" ":"", - x86_power_flags[i]); - else - seq_printf(m, " [%d]", i); - } - } - - seq_printf(m, "\n\n"); - - return 0; -} - -static void *c_start(struct seq_file *m, loff_t *pos) -{ - if (*pos == 0) /* just in case, cpu 0 is not the first */ - *pos = first_cpu(cpu_online_map); - if ((*pos) < NR_CPUS && cpu_online(*pos)) - return &cpu_data(*pos); - return NULL; -} - -static void *c_next(struct seq_file *m, void *v, loff_t *pos) -{ - *pos = next_cpu(*pos, cpu_online_map); - return c_start(m, pos); -} - -static void c_stop(struct seq_file *m, void *v) -{ -} - -const struct seq_operations cpuinfo_op = { - .start = c_start, - .next = c_next, - .stop = c_stop, - .show = show_cpuinfo, -}; diff --git a/arch/x86/kernel/sigframe_32.h b/arch/x86/kernel/sigframe.h index 0b2221711dad..72bbb519d2dc 100644 --- a/arch/x86/kernel/sigframe_32.h +++ b/arch/x86/kernel/sigframe.h @@ -1,5 +1,5 @@ -struct sigframe -{ +#ifdef CONFIG_X86_32 +struct sigframe { char __user *pretcode; int sig; struct sigcontext sc; @@ -8,8 +8,7 @@ struct sigframe char retcode[8]; }; -struct rt_sigframe -{ +struct rt_sigframe { char __user *pretcode; int sig; struct siginfo __user *pinfo; @@ -19,3 +18,10 @@ struct rt_sigframe struct _fpstate fpstate; char retcode[8]; }; +#else +struct rt_sigframe { + char __user *pretcode; + struct ucontext uc; + struct siginfo info; +}; +#endif diff --git a/arch/x86/kernel/signal_32.c b/arch/x86/kernel/signal_32.c index 0157a6f0f41f..f1b117930837 100644 --- a/arch/x86/kernel/signal_32.c +++ b/arch/x86/kernel/signal_32.c @@ -4,32 +4,44 @@ * 1997-11-28 Modified for POSIX.1b signals by Richard Henderson * 2000-06-20 Pentium III FXSR, SSE support by Gareth Hughes */ +#include <linux/list.h> -#include <linux/sched.h> -#include <linux/mm.h> -#include <linux/smp.h> +#include <linux/personality.h> +#include <linux/binfmts.h> +#include <linux/suspend.h> #include <linux/kernel.h> +#include <linux/ptrace.h> #include <linux/signal.h> +#include <linux/stddef.h> +#include <linux/unistd.h> #include <linux/errno.h> +#include <linux/sched.h> #include <linux/wait.h> -#include <linux/unistd.h> -#include <linux/stddef.h> -#include <linux/personality.h> -#include <linux/suspend.h> -#include <linux/ptrace.h> #include <linux/elf.h> -#include <linux/binfmts.h> +#include <linux/smp.h> +#include <linux/mm.h> + #include <asm/processor.h> #include <asm/ucontext.h> #include <asm/uaccess.h> #include <asm/i387.h> #include <asm/vdso.h> -#include "sigframe_32.h" -#define DEBUG_SIG 0 +#include "sigframe.h" #define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP))) +#define __FIX_EFLAGS (X86_EFLAGS_AC | X86_EFLAGS_OF | \ + X86_EFLAGS_DF | X86_EFLAGS_TF | X86_EFLAGS_SF | \ + X86_EFLAGS_ZF | X86_EFLAGS_AF | X86_EFLAGS_PF | \ + X86_EFLAGS_CF) + +#ifdef CONFIG_X86_32 +# define FIX_EFLAGS (__FIX_EFLAGS | X86_EFLAGS_RF) +#else +# define FIX_EFLAGS __FIX_EFLAGS +#endif + /* * Atomically swap in the new signal mask, and wait for a signal. */ @@ -46,10 +58,11 @@ sys_sigsuspend(int history0, int history1, old_sigset_t mask) current->state = TASK_INTERRUPTIBLE; schedule(); set_thread_flag(TIF_RESTORE_SIGMASK); + return -ERESTARTNOHAND; } -asmlinkage int +asmlinkage int sys_sigaction(int sig, const struct old_sigaction __user *act, struct old_sigaction __user *oact) { @@ -58,10 +71,12 @@ sys_sigaction(int sig, const struct old_sigaction __user *act, if (act) { old_sigset_t mask; + if (!access_ok(VERIFY_READ, act, sizeof(*act)) || __get_user(new_ka.sa.sa_handler, &act->sa_handler) || __get_user(new_ka.sa.sa_restorer, &act->sa_restorer)) return -EFAULT; + __get_user(new_ka.sa.sa_flags, &act->sa_flags); __get_user(mask, &act->sa_mask); siginitset(&new_ka.sa.sa_mask, mask); @@ -74,6 +89,7 @@ sys_sigaction(int sig, const struct old_sigaction __user *act, __put_user(old_ka.sa.sa_handler, &oact->sa_handler) || __put_user(old_ka.sa.sa_restorer, &oact->sa_restorer)) return -EFAULT; + __put_user(old_ka.sa.sa_flags, &oact->sa_flags); __put_user(old_ka.sa.sa_mask.sig[0], &oact->sa_mask); } @@ -81,10 +97,12 @@ sys_sigaction(int sig, const struct old_sigaction __user *act, return ret; } -asmlinkage int -sys_sigaltstack(unsigned long bx) +asmlinkage int sys_sigaltstack(unsigned long bx) { - /* This is needed to make gcc realize it doesn't own the "struct pt_regs" */ + /* + * This is needed to make gcc realize it doesn't own the + * "struct pt_regs" + */ struct pt_regs *regs = (struct pt_regs *)&bx; const stack_t __user *uss = (const stack_t __user *)bx; stack_t __user *uoss = (stack_t __user *)regs->cx; @@ -96,9 +114,9 @@ sys_sigaltstack(unsigned long bx) /* * Do a signal return; undo the signal stack. */ - static int -restore_sigcontext(struct pt_regs *regs, struct sigcontext __user *sc, int *peax) +restore_sigcontext(struct pt_regs *regs, struct sigcontext __user *sc, + unsigned long *pax) { unsigned int err = 0; @@ -120,37 +138,29 @@ restore_sigcontext(struct pt_regs *regs, struct sigcontext __user *sc, int *peax #define GET_SEG(seg) \ { unsigned short tmp; \ err |= __get_user(tmp, &sc->seg); \ - loadsegment(seg,tmp); } - -#define FIX_EFLAGS (X86_EFLAGS_AC | X86_EFLAGS_RF | \ - X86_EFLAGS_OF | X86_EFLAGS_DF | \ - X86_EFLAGS_TF | X86_EFLAGS_SF | X86_EFLAGS_ZF | \ - X86_EFLAGS_AF | X86_EFLAGS_PF | X86_EFLAGS_CF) + loadsegment(seg, tmp); } GET_SEG(gs); COPY_SEG(fs); COPY_SEG(es); COPY_SEG(ds); - COPY(di); - COPY(si); - COPY(bp); - COPY(sp); - COPY(bx); - COPY(dx); - COPY(cx); - COPY(ip); + COPY(di); COPY(si); COPY(bp); COPY(sp); COPY(bx); + COPY(dx); COPY(cx); COPY(ip); COPY_SEG_STRICT(cs); COPY_SEG_STRICT(ss); - + { unsigned int tmpflags; + err |= __get_user(tmpflags, &sc->flags); - regs->flags = (regs->flags & ~FIX_EFLAGS) | (tmpflags & FIX_EFLAGS); + regs->flags = (regs->flags & ~FIX_EFLAGS) | + (tmpflags & FIX_EFLAGS); regs->orig_ax = -1; /* disable syscall checks */ } { - struct _fpstate __user * buf; + struct _fpstate __user *buf; + err |= __get_user(buf, &sc->fpstate); if (buf) { if (!access_ok(VERIFY_READ, buf, sizeof(*buf))) @@ -158,6 +168,7 @@ restore_sigcontext(struct pt_regs *regs, struct sigcontext __user *sc, int *peax err |= restore_i387(buf); } else { struct task_struct *me = current; + if (used_math()) { clear_fpu(me); clear_used_math(); @@ -165,24 +176,26 @@ restore_sigcontext(struct pt_regs *regs, struct sigcontext __user *sc, int *peax } } - err |= __get_user(*peax, &sc->ax); + err |= __get_user(*pax, &sc->ax); return err; badframe: return 1; } -asmlinkage int sys_sigreturn(unsigned long __unused) +asmlinkage unsigned long sys_sigreturn(unsigned long __unused) { - struct pt_regs *regs = (struct pt_regs *) &__unused; - struct sigframe __user *frame = (struct sigframe __user *)(regs->sp - 8); + struct sigframe __user *frame; + struct pt_regs *regs; + unsigned long ax; sigset_t set; - int ax; + + regs = (struct pt_regs *) &__unused; + frame = (struct sigframe __user *)(regs->sp - 8); if (!access_ok(VERIFY_READ, frame, sizeof(*frame))) goto badframe; - if (__get_user(set.sig[0], &frame->sc.oldmask) - || (_NSIG_WORDS > 1 + if (__get_user(set.sig[0], &frame->sc.oldmask) || (_NSIG_WORDS > 1 && __copy_from_user(&set.sig[1], &frame->extramask, sizeof(frame->extramask)))) goto badframe; @@ -192,33 +205,35 @@ asmlinkage int sys_sigreturn(unsigned long __unused) current->blocked = set; recalc_sigpending(); spin_unlock_irq(¤t->sighand->siglock); - + if (restore_sigcontext(regs, &frame->sc, &ax)) goto badframe; return ax; badframe: if (show_unhandled_signals && printk_ratelimit()) { - printk("%s%s[%d] bad frame in sigreturn frame:%p ip:%lx" - " sp:%lx oeax:%lx", + printk(KERN_INFO "%s%s[%d] bad frame in sigreturn frame:" + "%p ip:%lx sp:%lx oeax:%lx", task_pid_nr(current) > 1 ? KERN_INFO : KERN_EMERG, current->comm, task_pid_nr(current), frame, regs->ip, regs->sp, regs->orig_ax); print_vma_addr(" in ", regs->ip); - printk("\n"); + printk(KERN_CONT "\n"); } force_sig(SIGSEGV, current); + return 0; -} +} asmlinkage int sys_rt_sigreturn(unsigned long __unused) { - struct pt_regs *regs = (struct pt_regs *) &__unused; - struct rt_sigframe __user *frame = (struct rt_sigframe __user *)(regs->sp - 4); + struct pt_regs *regs = (struct pt_regs *)&__unused; + struct rt_sigframe __user *frame; + unsigned long ax; sigset_t set; - int ax; + frame = (struct rt_sigframe __user *)(regs->sp - sizeof(long)); if (!access_ok(VERIFY_READ, frame, sizeof(*frame))) goto badframe; if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set))) @@ -229,7 +244,7 @@ asmlinkage int sys_rt_sigreturn(unsigned long __unused) current->blocked = set; recalc_sigpending(); spin_unlock_irq(¤t->sighand->siglock); - + if (restore_sigcontext(regs, &frame->uc.uc_mcontext, &ax)) goto badframe; @@ -241,12 +256,11 @@ asmlinkage int sys_rt_sigreturn(unsigned long __unused) badframe: force_sig(SIGSEGV, current); return 0; -} +} /* * Set up a signal frame. */ - static int setup_sigcontext(struct sigcontext __user *sc, struct _fpstate __user *fpstate, struct pt_regs *regs, unsigned long mask) @@ -277,9 +291,9 @@ setup_sigcontext(struct sigcontext __user *sc, struct _fpstate __user *fpstate, tmp = save_i387(fpstate); if (tmp < 0) - err = 1; + err = 1; else - err |= __put_user(tmp ? fpstate : NULL, &sc->fpstate); + err |= __put_user(tmp ? fpstate : NULL, &sc->fpstate); /* non-iBCS2 extensions.. */ err |= __put_user(mask, &sc->oldmask); @@ -292,7 +306,7 @@ setup_sigcontext(struct sigcontext __user *sc, struct _fpstate __user *fpstate, * Determine which stack to use.. */ static inline void __user * -get_sigframe(struct k_sigaction *ka, struct pt_regs * regs, size_t frame_size) +get_sigframe(struct k_sigaction *ka, struct pt_regs *regs, size_t frame_size) { unsigned long sp; @@ -310,32 +324,30 @@ get_sigframe(struct k_sigaction *ka, struct pt_regs * regs, size_t frame_size) if (ka->sa.sa_flags & SA_ONSTACK) { if (sas_ss_flags(sp) == 0) sp = current->sas_ss_sp + current->sas_ss_size; - } - - /* This is the legacy signal stack switching. */ - else if ((regs->ss & 0xffff) != __USER_DS && - !(ka->sa.sa_flags & SA_RESTORER) && - ka->sa.sa_restorer) { - sp = (unsigned long) ka->sa.sa_restorer; + } else { + /* This is the legacy signal stack switching. */ + if ((regs->ss & 0xffff) != __USER_DS && + !(ka->sa.sa_flags & SA_RESTORER) && + ka->sa.sa_restorer) + sp = (unsigned long) ka->sa.sa_restorer; } sp -= frame_size; - /* Align the stack pointer according to the i386 ABI, - * i.e. so that on function entry ((sp + 4) & 15) == 0. */ + /* + * Align the stack pointer according to the i386 ABI, + * i.e. so that on function entry ((sp + 4) & 15) == 0. + */ sp = ((sp + 4) & -16ul) - 4; + return (void __user *) sp; } -/* These symbols are defined with the addresses in the vsyscall page. - See vsyscall-sigreturn.S. */ -extern void __user __kernel_sigreturn; -extern void __user __kernel_rt_sigreturn; - -static int setup_frame(int sig, struct k_sigaction *ka, - sigset_t *set, struct pt_regs * regs) +static int +setup_frame(int sig, struct k_sigaction *ka, sigset_t *set, + struct pt_regs *regs) { - void __user *restorer; struct sigframe __user *frame; + void __user *restorer; int err = 0; int usig; @@ -365,7 +377,7 @@ static int setup_frame(int sig, struct k_sigaction *ka, goto give_sigsegv; } - if (current->binfmt->hasvdso) + if (current->mm->context.vdso) restorer = VDSO32_SYMBOL(current->mm->context.vdso, sigreturn); else restorer = &frame->retcode; @@ -374,9 +386,9 @@ static int setup_frame(int sig, struct k_sigaction *ka, /* Set up to return from userspace. */ err |= __put_user(restorer, &frame->pretcode); - + /* - * This is popl %eax ; movl $,%eax ; int $0x80 + * This is popl %eax ; movl $__NR_sigreturn, %eax ; int $0x80 * * WE DO NOT USE IT ANY MORE! It's only left here for historical * reasons and because gdb uses it as a signature to notice @@ -390,11 +402,11 @@ static int setup_frame(int sig, struct k_sigaction *ka, goto give_sigsegv; /* Set up registers for signal handler */ - regs->sp = (unsigned long) frame; - regs->ip = (unsigned long) ka->sa.sa_handler; - regs->ax = (unsigned long) sig; - regs->dx = (unsigned long) 0; - regs->cx = (unsigned long) 0; + regs->sp = (unsigned long)frame; + regs->ip = (unsigned long)ka->sa.sa_handler; + regs->ax = (unsigned long)sig; + regs->dx = 0; + regs->cx = 0; regs->ds = __USER_DS; regs->es = __USER_DS; @@ -407,15 +419,10 @@ static int setup_frame(int sig, struct k_sigaction *ka, * The tracer may want to single-step inside the * handler too. */ - regs->flags &= ~(TF_MASK | X86_EFLAGS_DF); + regs->flags &= ~(X86_EFLAGS_TF | X86_EFLAGS_DF); if (test_thread_flag(TIF_SINGLESTEP)) ptrace_notify(SIGTRAP); -#if DEBUG_SIG - printk("SIG deliver (%s:%d): sp=%p pc=%p ra=%p\n", - current->comm, current->pid, frame, regs->ip, frame->pretcode); -#endif - return 0; give_sigsegv: @@ -424,10 +431,10 @@ give_sigsegv: } static int setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, - sigset_t *set, struct pt_regs * regs) + sigset_t *set, struct pt_regs *regs) { - void __user *restorer; struct rt_sigframe __user *frame; + void __user *restorer; int err = 0; int usig; @@ -457,7 +464,7 @@ static int setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, &frame->uc.uc_stack.ss_flags); err |= __put_user(current->sas_ss_size, &frame->uc.uc_stack.ss_size); err |= setup_sigcontext(&frame->uc.uc_mcontext, &frame->fpstate, - regs, set->sig[0]); + regs, set->sig[0]); err |= __copy_to_user(&frame->uc.uc_sigmask, set, sizeof(*set)); if (err) goto give_sigsegv; @@ -467,9 +474,9 @@ static int setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, if (ka->sa.sa_flags & SA_RESTORER) restorer = ka->sa.sa_restorer; err |= __put_user(restorer, &frame->pretcode); - + /* - * This is movl $,%ax ; int $0x80 + * This is movl $__NR_rt_sigreturn, %ax ; int $0x80 * * WE DO NOT USE IT ANY MORE! It's only left here for historical * reasons and because gdb uses it as a signature to notice @@ -483,11 +490,11 @@ static int setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, goto give_sigsegv; /* Set up registers for signal handler */ - regs->sp = (unsigned long) frame; - regs->ip = (unsigned long) ka->sa.sa_handler; - regs->ax = (unsigned long) usig; - regs->dx = (unsigned long) &frame->info; - regs->cx = (unsigned long) &frame->uc; + regs->sp = (unsigned long)frame; + regs->ip = (unsigned long)ka->sa.sa_handler; + regs->ax = (unsigned long)usig; + regs->dx = (unsigned long)&frame->info; + regs->cx = (unsigned long)&frame->uc; regs->ds = __USER_DS; regs->es = __USER_DS; @@ -500,15 +507,10 @@ static int setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, * The tracer may want to single-step inside the * handler too. */ - regs->flags &= ~(TF_MASK | X86_EFLAGS_DF); + regs->flags &= ~(X86_EFLAGS_TF | X86_EFLAGS_DF); if (test_thread_flag(TIF_SINGLESTEP)) ptrace_notify(SIGTRAP); -#if DEBUG_SIG - printk("SIG deliver (%s:%d): sp=%p pc=%p ra=%p\n", - current->comm, current->pid, frame, regs->ip, frame->pretcode); -#endif - return 0; give_sigsegv: @@ -517,33 +519,33 @@ give_sigsegv: } /* - * OK, we're invoking a handler - */ - + * OK, we're invoking a handler: + */ static int handle_signal(unsigned long sig, siginfo_t *info, struct k_sigaction *ka, - sigset_t *oldset, struct pt_regs * regs) + sigset_t *oldset, struct pt_regs *regs) { int ret; /* Are we from a system call? */ - if (regs->orig_ax >= 0) { + if ((long)regs->orig_ax >= 0) { /* If so, check system call restarting.. */ switch (regs->ax) { - case -ERESTART_RESTARTBLOCK: - case -ERESTARTNOHAND: + case -ERESTART_RESTARTBLOCK: + case -ERESTARTNOHAND: + regs->ax = -EINTR; + break; + + case -ERESTARTSYS: + if (!(ka->sa.sa_flags & SA_RESTART)) { regs->ax = -EINTR; break; - - case -ERESTARTSYS: - if (!(ka->sa.sa_flags & SA_RESTART)) { - regs->ax = -EINTR; - break; - } - /* fallthrough */ - case -ERESTARTNOINTR: - regs->ax = regs->orig_ax; - regs->ip -= 2; + } + /* fallthrough */ + case -ERESTARTNOINTR: + regs->ax = regs->orig_ax; + regs->ip -= 2; + break; } } @@ -561,16 +563,17 @@ handle_signal(unsigned long sig, siginfo_t *info, struct k_sigaction *ka, else ret = setup_frame(sig, ka, oldset, regs); - if (ret == 0) { - spin_lock_irq(¤t->sighand->siglock); - sigorsets(¤t->blocked,¤t->blocked,&ka->sa.sa_mask); - if (!(ka->sa.sa_flags & SA_NODEFER)) - sigaddset(¤t->blocked,sig); - recalc_sigpending(); - spin_unlock_irq(¤t->sighand->siglock); - } + if (ret) + return ret; - return ret; + spin_lock_irq(¤t->sighand->siglock); + sigorsets(¤t->blocked, ¤t->blocked, &ka->sa.sa_mask); + if (!(ka->sa.sa_flags & SA_NODEFER)) + sigaddset(¤t->blocked, sig); + recalc_sigpending(); + spin_unlock_irq(¤t->sighand->siglock); + + return 0; } /* @@ -580,18 +583,17 @@ handle_signal(unsigned long sig, siginfo_t *info, struct k_sigaction *ka, */ static void do_signal(struct pt_regs *regs) { + struct k_sigaction ka; siginfo_t info; int signr; - struct k_sigaction ka; sigset_t *oldset; /* - * We want the common case to go fast, which - * is why we may in certain cases get here from - * kernel mode. Just return without doing anything - * if so. vm86 regs switched out by assembly code - * before reaching here, so testing against kernel - * CS suffices. + * We want the common case to go fast, which is why we may in certain + * cases get here from kernel mode. Just return without doing anything + * if so. + * X86_32: vm86 regs switched out by assembly code before reaching + * here, so testing against kernel CS suffices. */ if (!user_mode(regs)) return; @@ -603,29 +605,31 @@ static void do_signal(struct pt_regs *regs) signr = get_signal_to_deliver(&info, &ka, regs, NULL); if (signr > 0) { - /* Re-enable any watchpoints before delivering the + /* + * Re-enable any watchpoints before delivering the * signal to user space. The processor register will * have been cleared if the watchpoint triggered * inside the kernel. */ - if (unlikely(current->thread.debugreg7)) + if (current->thread.debugreg7) set_debugreg(current->thread.debugreg7, 7); - /* Whee! Actually deliver the signal. */ + /* Whee! Actually deliver the signal. */ if (handle_signal(signr, &info, &ka, oldset, regs) == 0) { - /* a signal was successfully delivered; the saved + /* + * a signal was successfully delivered; the saved * sigmask will have been stored in the signal frame, * and will be restored by sigreturn, so we can simply - * clear the TIF_RESTORE_SIGMASK flag */ + * clear the TIF_RESTORE_SIGMASK flag + */ if (test_thread_flag(TIF_RESTORE_SIGMASK)) clear_thread_flag(TIF_RESTORE_SIGMASK); } - return; } /* Did we come from a system call? */ - if (regs->orig_ax >= 0) { + if ((long)regs->orig_ax >= 0) { /* Restart the system call - no handlers present */ switch (regs->ax) { case -ERESTARTNOHAND: @@ -642,8 +646,10 @@ static void do_signal(struct pt_regs *regs) } } - /* if there's no signal to deliver, we just put the saved sigmask - * back */ + /* + * If there's no signal to deliver, we just put the saved sigmask + * back. + */ if (test_thread_flag(TIF_RESTORE_SIGMASK)) { clear_thread_flag(TIF_RESTORE_SIGMASK); sigprocmask(SIG_SETMASK, ¤t->saved_sigmask, NULL); @@ -654,13 +660,12 @@ static void do_signal(struct pt_regs *regs) * notification of userspace execution resumption * - triggered by the TIF_WORK_MASK flags */ -__attribute__((regparm(3))) -void do_notify_resume(struct pt_regs *regs, void *_unused, - __u32 thread_info_flags) +void +do_notify_resume(struct pt_regs *regs, void *unused, __u32 thread_info_flags) { /* Pending single-step? */ if (thread_info_flags & _TIF_SINGLESTEP) { - regs->flags |= TF_MASK; + regs->flags |= X86_EFLAGS_TF; clear_thread_flag(TIF_SINGLESTEP); } @@ -670,6 +675,6 @@ void do_notify_resume(struct pt_regs *regs, void *_unused, if (thread_info_flags & _TIF_HRTICK_RESCHED) hrtick_resched(); - + clear_thread_flag(TIF_IRET); } diff --git a/arch/x86/kernel/signal_64.c b/arch/x86/kernel/signal_64.c index 1c83e5124c65..827179c5b32a 100644 --- a/arch/x86/kernel/signal_64.c +++ b/arch/x86/kernel/signal_64.c @@ -19,17 +19,28 @@ #include <linux/stddef.h> #include <linux/personality.h> #include <linux/compiler.h> +#include <asm/processor.h> #include <asm/ucontext.h> #include <asm/uaccess.h> #include <asm/i387.h> #include <asm/proto.h> #include <asm/ia32_unistd.h> #include <asm/mce.h> - -/* #define DEBUG_SIG 1 */ +#include "sigframe.h" #define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP))) +#define __FIX_EFLAGS (X86_EFLAGS_AC | X86_EFLAGS_OF | \ + X86_EFLAGS_DF | X86_EFLAGS_TF | X86_EFLAGS_SF | \ + X86_EFLAGS_ZF | X86_EFLAGS_AF | X86_EFLAGS_PF | \ + X86_EFLAGS_CF) + +#ifdef CONFIG_X86_32 +# define FIX_EFLAGS (__FIX_EFLAGS | X86_EFLAGS_RF) +#else +# define FIX_EFLAGS __FIX_EFLAGS +#endif + int ia32_setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, sigset_t *set, struct pt_regs * regs); int ia32_setup_frame(int sig, struct k_sigaction *ka, @@ -46,16 +57,9 @@ sys_sigaltstack(const stack_t __user *uss, stack_t __user *uoss, /* * Do a signal return; undo the signal stack. */ - -struct rt_sigframe -{ - char __user *pretcode; - struct ucontext uc; - struct siginfo info; -}; - static int -restore_sigcontext(struct pt_regs *regs, struct sigcontext __user *sc, unsigned long *prax) +restore_sigcontext(struct pt_regs *regs, struct sigcontext __user *sc, + unsigned long *pax) { unsigned int err = 0; @@ -87,7 +91,7 @@ restore_sigcontext(struct pt_regs *regs, struct sigcontext __user *sc, unsigned { unsigned int tmpflags; err |= __get_user(tmpflags, &sc->flags); - regs->flags = (regs->flags & ~0x40DD5) | (tmpflags & 0x40DD5); + regs->flags = (regs->flags & ~FIX_EFLAGS) | (tmpflags & FIX_EFLAGS); regs->orig_ax = -1; /* disable syscall checks */ } @@ -108,7 +112,7 @@ restore_sigcontext(struct pt_regs *regs, struct sigcontext __user *sc, unsigned } } - err |= __get_user(*prax, &sc->ax); + err |= __get_user(*pax, &sc->ax); return err; badframe: @@ -121,13 +125,11 @@ asmlinkage long sys_rt_sigreturn(struct pt_regs *regs) sigset_t set; unsigned long ax; - frame = (struct rt_sigframe __user *)(regs->sp - 8); - if (!access_ok(VERIFY_READ, frame, sizeof(*frame))) { + frame = (struct rt_sigframe __user *)(regs->sp - sizeof(long)); + if (!access_ok(VERIFY_READ, frame, sizeof(*frame))) goto badframe; - } - if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set))) { + if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set))) goto badframe; - } sigdelsetmask(&set, ~_BLOCKABLE); spin_lock_irq(¤t->sighand->siglock); @@ -138,10 +140,6 @@ asmlinkage long sys_rt_sigreturn(struct pt_regs *regs) if (restore_sigcontext(regs, &frame->uc.uc_mcontext, &ax)) goto badframe; -#ifdef DEBUG_SIG - printk("%d sigreturn ip:%lx sp:%lx frame:%p ax:%lx\n",current->pid,regs->ip,regs->sp,frame,ax); -#endif - if (do_sigaltstack(&frame->uc.uc_stack, NULL, regs->sp) == -EFAULT) goto badframe; @@ -270,10 +268,6 @@ static int setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, if (err) goto give_sigsegv; -#ifdef DEBUG_SIG - printk("%d old ip %lx old sp %lx old ax %lx\n", current->pid,regs->ip,regs->sp,regs->ax); -#endif - /* Set up registers for signal handler */ regs->di = sig; /* In case the signal handler was declared without prototypes */ @@ -298,10 +292,6 @@ static int setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, regs->flags &= ~(X86_EFLAGS_TF | X86_EFLAGS_DF); if (test_thread_flag(TIF_SINGLESTEP)) ptrace_notify(SIGTRAP); -#ifdef DEBUG_SIG - printk("SIG deliver (%s:%d): sp=%p pc=%lx ra=%p\n", - current->comm, current->pid, frame, regs->ip, frame->pretcode); -#endif return 0; @@ -345,35 +335,29 @@ static long current_syscall_ret(struct pt_regs *regs) static int handle_signal(unsigned long sig, siginfo_t *info, struct k_sigaction *ka, - sigset_t *oldset, struct pt_regs *regs) + sigset_t *oldset, struct pt_regs *regs) { int ret; -#ifdef DEBUG_SIG - printk("handle_signal pid:%d sig:%lu ip:%lx sp:%lx regs=%p\n", - current->pid, sig, - regs->ip, regs->sp, regs); -#endif - /* Are we from a system call? */ if (current_syscall(regs) >= 0) { /* If so, check system call restarting.. */ switch (current_syscall_ret(regs)) { - case -ERESTART_RESTARTBLOCK: - case -ERESTARTNOHAND: - regs->ax = -EINTR; - break; + case -ERESTART_RESTARTBLOCK: + case -ERESTARTNOHAND: + regs->ax = -EINTR; + break; - case -ERESTARTSYS: - if (!(ka->sa.sa_flags & SA_RESTART)) { - regs->ax = -EINTR; - break; - } - /* fallthrough */ - case -ERESTARTNOINTR: - regs->ax = regs->orig_ax; - regs->ip -= 2; + case -ERESTARTSYS: + if (!(ka->sa.sa_flags & SA_RESTART)) { + regs->ax = -EINTR; break; + } + /* fallthrough */ + case -ERESTARTNOINTR: + regs->ax = regs->orig_ax; + regs->ip -= 2; + break; } } @@ -420,10 +404,11 @@ static void do_signal(struct pt_regs *regs) sigset_t *oldset; /* - * We want the common case to go fast, which - * is why we may in certain cases get here from - * kernel mode. Just return without doing anything + * We want the common case to go fast, which is why we may in certain + * cases get here from kernel mode. Just return without doing anything * if so. + * X86_32: vm86 regs switched out by assembly code before reaching + * here, so testing against kernel CS suffices. */ if (!user_mode(regs)) return; @@ -473,22 +458,19 @@ static void do_signal(struct pt_regs *regs) } } - /* if there's no signal to deliver, we just put the saved sigmask - back. */ + /* + * If there's no signal to deliver, we just put the saved sigmask + * back. + */ if (test_thread_flag(TIF_RESTORE_SIGMASK)) { clear_thread_flag(TIF_RESTORE_SIGMASK); sigprocmask(SIG_SETMASK, ¤t->saved_sigmask, NULL); } } -void -do_notify_resume(struct pt_regs *regs, void *unused, __u32 thread_info_flags) +void do_notify_resume(struct pt_regs *regs, void *unused, + __u32 thread_info_flags) { -#ifdef DEBUG_SIG - printk("do_notify_resume flags:%x ip:%lx sp:%lx caller:%p pending:%x\n", - thread_info_flags, regs->ip, regs->sp, __builtin_return_address(0),signal_pending(current)); -#endif - /* Pending single-step? */ if (thread_info_flags & _TIF_SINGLESTEP) { regs->flags |= X86_EFLAGS_TF; @@ -502,7 +484,7 @@ do_notify_resume(struct pt_regs *regs, void *unused, __u32 thread_info_flags) #endif /* CONFIG_X86_MCE */ /* deal with pending signal delivery */ - if (thread_info_flags & (_TIF_SIGPENDING|_TIF_RESTORE_SIGMASK)) + if (thread_info_flags & (_TIF_SIGPENDING | _TIF_RESTORE_SIGMASK)) do_signal(regs); if (thread_info_flags & _TIF_HRTICK_RESCHED) diff --git a/arch/x86/kernel/smp.c b/arch/x86/kernel/smp.c new file mode 100644 index 000000000000..8f75893a6467 --- /dev/null +++ b/arch/x86/kernel/smp.c @@ -0,0 +1,343 @@ +/* + * Intel SMP support routines. + * + * (c) 1995 Alan Cox, Building #3 <alan@redhat.com> + * (c) 1998-99, 2000 Ingo Molnar <mingo@redhat.com> + * (c) 2002,2003 Andi Kleen, SuSE Labs. + * + * i386 and x86_64 integration by Glauber Costa <gcosta@redhat.com> + * + * This code is released under the GNU General Public License version 2 or + * later. + */ + +#include <linux/init.h> + +#include <linux/mm.h> +#include <linux/delay.h> +#include <linux/spinlock.h> +#include <linux/kernel_stat.h> +#include <linux/mc146818rtc.h> +#include <linux/cache.h> +#include <linux/interrupt.h> +#include <linux/cpu.h> + +#include <asm/mtrr.h> +#include <asm/tlbflush.h> +#include <asm/mmu_context.h> +#include <asm/proto.h> +#include <mach_ipi.h> +#include <mach_apic.h> +/* + * Some notes on x86 processor bugs affecting SMP operation: + * + * Pentium, Pentium Pro, II, III (and all CPUs) have bugs. + * The Linux implications for SMP are handled as follows: + * + * Pentium III / [Xeon] + * None of the E1AP-E3AP errata are visible to the user. + * + * E1AP. see PII A1AP + * E2AP. see PII A2AP + * E3AP. see PII A3AP + * + * Pentium II / [Xeon] + * None of the A1AP-A3AP errata are visible to the user. + * + * A1AP. see PPro 1AP + * A2AP. see PPro 2AP + * A3AP. see PPro 7AP + * + * Pentium Pro + * None of 1AP-9AP errata are visible to the normal user, + * except occasional delivery of 'spurious interrupt' as trap #15. + * This is very rare and a non-problem. + * + * 1AP. Linux maps APIC as non-cacheable + * 2AP. worked around in hardware + * 3AP. fixed in C0 and above steppings microcode update. + * Linux does not use excessive STARTUP_IPIs. + * 4AP. worked around in hardware + * 5AP. symmetric IO mode (normal Linux operation) not affected. + * 'noapic' mode has vector 0xf filled out properly. + * 6AP. 'noapic' mode might be affected - fixed in later steppings + * 7AP. We do not assume writes to the LVT deassering IRQs + * 8AP. We do not enable low power mode (deep sleep) during MP bootup + * 9AP. We do not use mixed mode + * + * Pentium + * There is a marginal case where REP MOVS on 100MHz SMP + * machines with B stepping processors can fail. XXX should provide + * an L1cache=Writethrough or L1cache=off option. + * + * B stepping CPUs may hang. There are hardware work arounds + * for this. We warn about it in case your board doesn't have the work + * arounds. Basically that's so I can tell anyone with a B stepping + * CPU and SMP problems "tough". + * + * Specific items [From Pentium Processor Specification Update] + * + * 1AP. Linux doesn't use remote read + * 2AP. Linux doesn't trust APIC errors + * 3AP. We work around this + * 4AP. Linux never generated 3 interrupts of the same priority + * to cause a lost local interrupt. + * 5AP. Remote read is never used + * 6AP. not affected - worked around in hardware + * 7AP. not affected - worked around in hardware + * 8AP. worked around in hardware - we get explicit CS errors if not + * 9AP. only 'noapic' mode affected. Might generate spurious + * interrupts, we log only the first one and count the + * rest silently. + * 10AP. not affected - worked around in hardware + * 11AP. Linux reads the APIC between writes to avoid this, as per + * the documentation. Make sure you preserve this as it affects + * the C stepping chips too. + * 12AP. not affected - worked around in hardware + * 13AP. not affected - worked around in hardware + * 14AP. we always deassert INIT during bootup + * 15AP. not affected - worked around in hardware + * 16AP. not affected - worked around in hardware + * 17AP. not affected - worked around in hardware + * 18AP. not affected - worked around in hardware + * 19AP. not affected - worked around in BIOS + * + * If this sounds worrying believe me these bugs are either ___RARE___, + * or are signal timing bugs worked around in hardware and there's + * about nothing of note with C stepping upwards. + */ + +/* + * this function sends a 'reschedule' IPI to another CPU. + * it goes straight through and wastes no time serializing + * anything. Worst case is that we lose a reschedule ... + */ +static void native_smp_send_reschedule(int cpu) +{ + if (unlikely(cpu_is_offline(cpu))) { + WARN_ON(1); + return; + } + send_IPI_mask(cpumask_of_cpu(cpu), RESCHEDULE_VECTOR); +} + +/* + * Structure and data for smp_call_function(). This is designed to minimise + * static memory requirements. It also looks cleaner. + */ +static DEFINE_SPINLOCK(call_lock); + +struct call_data_struct { + void (*func) (void *info); + void *info; + atomic_t started; + atomic_t finished; + int wait; +}; + +void lock_ipi_call_lock(void) +{ + spin_lock_irq(&call_lock); +} + +void unlock_ipi_call_lock(void) +{ + spin_unlock_irq(&call_lock); +} + +static struct call_data_struct *call_data; + +static void __smp_call_function(void (*func) (void *info), void *info, + int nonatomic, int wait) +{ + struct call_data_struct data; + int cpus = num_online_cpus() - 1; + + if (!cpus) + return; + + data.func = func; + data.info = info; + atomic_set(&data.started, 0); + data.wait = wait; + if (wait) + atomic_set(&data.finished, 0); + + call_data = &data; + mb(); + + /* Send a message to all other CPUs and wait for them to respond */ + send_IPI_allbutself(CALL_FUNCTION_VECTOR); + + /* Wait for response */ + while (atomic_read(&data.started) != cpus) + cpu_relax(); + + if (wait) + while (atomic_read(&data.finished) != cpus) + cpu_relax(); +} + + +/** + * smp_call_function_mask(): Run a function on a set of other CPUs. + * @mask: The set of cpus to run on. Must not include the current cpu. + * @func: The function to run. This must be fast and non-blocking. + * @info: An arbitrary pointer to pass to the function. + * @wait: If true, wait (atomically) until function has completed on other CPUs. + * + * Returns 0 on success, else a negative status code. + * + * If @wait is true, then returns once @func has returned; otherwise + * it returns just before the target cpu calls @func. + * + * You must not call this function with disabled interrupts or from a + * hardware interrupt handler or from a bottom half handler. + */ +static int +native_smp_call_function_mask(cpumask_t mask, + void (*func)(void *), void *info, + int wait) +{ + struct call_data_struct data; + cpumask_t allbutself; + int cpus; + + /* Can deadlock when called with interrupts disabled */ + WARN_ON(irqs_disabled()); + + /* Holding any lock stops cpus from going down. */ + spin_lock(&call_lock); + + allbutself = cpu_online_map; + cpu_clear(smp_processor_id(), allbutself); + + cpus_and(mask, mask, allbutself); + cpus = cpus_weight(mask); + + if (!cpus) { + spin_unlock(&call_lock); + return 0; + } + + data.func = func; + data.info = info; + atomic_set(&data.started, 0); + data.wait = wait; + if (wait) + atomic_set(&data.finished, 0); + + call_data = &data; + wmb(); + + /* Send a message to other CPUs */ + if (cpus_equal(mask, allbutself)) + send_IPI_allbutself(CALL_FUNCTION_VECTOR); + else + send_IPI_mask(mask, CALL_FUNCTION_VECTOR); + + /* Wait for response */ + while (atomic_read(&data.started) != cpus) + cpu_relax(); + + if (wait) + while (atomic_read(&data.finished) != cpus) + cpu_relax(); + spin_unlock(&call_lock); + + return 0; +} + +static void stop_this_cpu(void *dummy) +{ + local_irq_disable(); + /* + * Remove this CPU: + */ + cpu_clear(smp_processor_id(), cpu_online_map); + disable_local_APIC(); + if (hlt_works(smp_processor_id())) + for (;;) halt(); + for (;;); +} + +/* + * this function calls the 'stop' function on all other CPUs in the system. + */ + +static void native_smp_send_stop(void) +{ + int nolock; + unsigned long flags; + + if (reboot_force) + return; + + /* Don't deadlock on the call lock in panic */ + nolock = !spin_trylock(&call_lock); + local_irq_save(flags); + __smp_call_function(stop_this_cpu, NULL, 0, 0); + if (!nolock) + spin_unlock(&call_lock); + disable_local_APIC(); + local_irq_restore(flags); +} + +/* + * Reschedule call back. Nothing to do, + * all the work is done automatically when + * we return from the interrupt. + */ +void smp_reschedule_interrupt(struct pt_regs *regs) +{ + ack_APIC_irq(); +#ifdef CONFIG_X86_32 + __get_cpu_var(irq_stat).irq_resched_count++; +#else + add_pda(irq_resched_count, 1); +#endif +} + +void smp_call_function_interrupt(struct pt_regs *regs) +{ + void (*func) (void *info) = call_data->func; + void *info = call_data->info; + int wait = call_data->wait; + + ack_APIC_irq(); + /* + * Notify initiating CPU that I've grabbed the data and am + * about to execute the function + */ + mb(); + atomic_inc(&call_data->started); + /* + * At this point the info structure may be out of scope unless wait==1 + */ + irq_enter(); + (*func)(info); +#ifdef CONFIG_X86_32 + __get_cpu_var(irq_stat).irq_call_count++; +#else + add_pda(irq_call_count, 1); +#endif + irq_exit(); + + if (wait) { + mb(); + atomic_inc(&call_data->finished); + } +} + +struct smp_ops smp_ops = { + .smp_prepare_boot_cpu = native_smp_prepare_boot_cpu, + .smp_prepare_cpus = native_smp_prepare_cpus, + .cpu_up = native_cpu_up, + .smp_cpus_done = native_smp_cpus_done, + + .smp_send_stop = native_smp_send_stop, + .smp_send_reschedule = native_smp_send_reschedule, + .smp_call_function_mask = native_smp_call_function_mask, +}; +EXPORT_SYMBOL_GPL(smp_ops); + diff --git a/arch/x86/kernel/smp_32.c b/arch/x86/kernel/smp_32.c deleted file mode 100644 index dc0cde9d16fb..000000000000 --- a/arch/x86/kernel/smp_32.c +++ /dev/null @@ -1,712 +0,0 @@ -/* - * Intel SMP support routines. - * - * (c) 1995 Alan Cox, Building #3 <alan@redhat.com> - * (c) 1998-99, 2000 Ingo Molnar <mingo@redhat.com> - * - * This code is released under the GNU General Public License version 2 or - * later. - */ - -#include <linux/init.h> - -#include <linux/mm.h> -#include <linux/delay.h> -#include <linux/spinlock.h> -#include <linux/kernel_stat.h> -#include <linux/mc146818rtc.h> -#include <linux/cache.h> -#include <linux/interrupt.h> -#include <linux/cpu.h> -#include <linux/module.h> - -#include <asm/mtrr.h> -#include <asm/tlbflush.h> -#include <asm/mmu_context.h> -#include <mach_apic.h> - -/* - * Some notes on x86 processor bugs affecting SMP operation: - * - * Pentium, Pentium Pro, II, III (and all CPUs) have bugs. - * The Linux implications for SMP are handled as follows: - * - * Pentium III / [Xeon] - * None of the E1AP-E3AP errata are visible to the user. - * - * E1AP. see PII A1AP - * E2AP. see PII A2AP - * E3AP. see PII A3AP - * - * Pentium II / [Xeon] - * None of the A1AP-A3AP errata are visible to the user. - * - * A1AP. see PPro 1AP - * A2AP. see PPro 2AP - * A3AP. see PPro 7AP - * - * Pentium Pro - * None of 1AP-9AP errata are visible to the normal user, - * except occasional delivery of 'spurious interrupt' as trap #15. - * This is very rare and a non-problem. - * - * 1AP. Linux maps APIC as non-cacheable - * 2AP. worked around in hardware - * 3AP. fixed in C0 and above steppings microcode update. - * Linux does not use excessive STARTUP_IPIs. - * 4AP. worked around in hardware - * 5AP. symmetric IO mode (normal Linux operation) not affected. - * 'noapic' mode has vector 0xf filled out properly. - * 6AP. 'noapic' mode might be affected - fixed in later steppings - * 7AP. We do not assume writes to the LVT deassering IRQs - * 8AP. We do not enable low power mode (deep sleep) during MP bootup - * 9AP. We do not use mixed mode - * - * Pentium - * There is a marginal case where REP MOVS on 100MHz SMP - * machines with B stepping processors can fail. XXX should provide - * an L1cache=Writethrough or L1cache=off option. - * - * B stepping CPUs may hang. There are hardware work arounds - * for this. We warn about it in case your board doesn't have the work - * arounds. Basically that's so I can tell anyone with a B stepping - * CPU and SMP problems "tough". - * - * Specific items [From Pentium Processor Specification Update] - * - * 1AP. Linux doesn't use remote read - * 2AP. Linux doesn't trust APIC errors - * 3AP. We work around this - * 4AP. Linux never generated 3 interrupts of the same priority - * to cause a lost local interrupt. - * 5AP. Remote read is never used - * 6AP. not affected - worked around in hardware - * 7AP. not affected - worked around in hardware - * 8AP. worked around in hardware - we get explicit CS errors if not - * 9AP. only 'noapic' mode affected. Might generate spurious - * interrupts, we log only the first one and count the - * rest silently. - * 10AP. not affected - worked around in hardware - * 11AP. Linux reads the APIC between writes to avoid this, as per - * the documentation. Make sure you preserve this as it affects - * the C stepping chips too. - * 12AP. not affected - worked around in hardware - * 13AP. not affected - worked around in hardware - * 14AP. we always deassert INIT during bootup - * 15AP. not affected - worked around in hardware - * 16AP. not affected - worked around in hardware - * 17AP. not affected - worked around in hardware - * 18AP. not affected - worked around in hardware - * 19AP. not affected - worked around in BIOS - * - * If this sounds worrying believe me these bugs are either ___RARE___, - * or are signal timing bugs worked around in hardware and there's - * about nothing of note with C stepping upwards. - */ - -DEFINE_PER_CPU(struct tlb_state, cpu_tlbstate) ____cacheline_aligned = { &init_mm, 0, }; - -/* - * the following functions deal with sending IPIs between CPUs. - * - * We use 'broadcast', CPU->CPU IPIs and self-IPIs too. - */ - -static inline int __prepare_ICR (unsigned int shortcut, int vector) -{ - unsigned int icr = shortcut | APIC_DEST_LOGICAL; - - switch (vector) { - default: - icr |= APIC_DM_FIXED | vector; - break; - case NMI_VECTOR: - icr |= APIC_DM_NMI; - break; - } - return icr; -} - -static inline int __prepare_ICR2 (unsigned int mask) -{ - return SET_APIC_DEST_FIELD(mask); -} - -void __send_IPI_shortcut(unsigned int shortcut, int vector) -{ - /* - * Subtle. In the case of the 'never do double writes' workaround - * we have to lock out interrupts to be safe. As we don't care - * of the value read we use an atomic rmw access to avoid costly - * cli/sti. Otherwise we use an even cheaper single atomic write - * to the APIC. - */ - unsigned int cfg; - - /* - * Wait for idle. - */ - apic_wait_icr_idle(); - - /* - * No need to touch the target chip field - */ - cfg = __prepare_ICR(shortcut, vector); - - /* - * Send the IPI. The write to APIC_ICR fires this off. - */ - apic_write_around(APIC_ICR, cfg); -} - -void send_IPI_self(int vector) -{ - __send_IPI_shortcut(APIC_DEST_SELF, vector); -} - -/* - * This is used to send an IPI with no shorthand notation (the destination is - * specified in bits 56 to 63 of the ICR). - */ -static inline void __send_IPI_dest_field(unsigned long mask, int vector) -{ - unsigned long cfg; - - /* - * Wait for idle. - */ - if (unlikely(vector == NMI_VECTOR)) - safe_apic_wait_icr_idle(); - else - apic_wait_icr_idle(); - - /* - * prepare target chip field - */ - cfg = __prepare_ICR2(mask); - apic_write_around(APIC_ICR2, cfg); - - /* - * program the ICR - */ - cfg = __prepare_ICR(0, vector); - - /* - * Send the IPI. The write to APIC_ICR fires this off. - */ - apic_write_around(APIC_ICR, cfg); -} - -/* - * This is only used on smaller machines. - */ -void send_IPI_mask_bitmask(cpumask_t cpumask, int vector) -{ - unsigned long mask = cpus_addr(cpumask)[0]; - unsigned long flags; - - local_irq_save(flags); - WARN_ON(mask & ~cpus_addr(cpu_online_map)[0]); - __send_IPI_dest_field(mask, vector); - local_irq_restore(flags); -} - -void send_IPI_mask_sequence(cpumask_t mask, int vector) -{ - unsigned long flags; - unsigned int query_cpu; - - /* - * Hack. The clustered APIC addressing mode doesn't allow us to send - * to an arbitrary mask, so I do a unicasts to each CPU instead. This - * should be modified to do 1 message per cluster ID - mbligh - */ - - local_irq_save(flags); - for_each_possible_cpu(query_cpu) { - if (cpu_isset(query_cpu, mask)) { - __send_IPI_dest_field(cpu_to_logical_apicid(query_cpu), - vector); - } - } - local_irq_restore(flags); -} - -#include <mach_ipi.h> /* must come after the send_IPI functions above for inlining */ - -/* - * Smarter SMP flushing macros. - * c/o Linus Torvalds. - * - * These mean you can really definitely utterly forget about - * writing to user space from interrupts. (Its not allowed anyway). - * - * Optimizations Manfred Spraul <manfred@colorfullife.com> - */ - -static cpumask_t flush_cpumask; -static struct mm_struct * flush_mm; -static unsigned long flush_va; -static DEFINE_SPINLOCK(tlbstate_lock); - -/* - * We cannot call mmdrop() because we are in interrupt context, - * instead update mm->cpu_vm_mask. - * - * We need to reload %cr3 since the page tables may be going - * away from under us.. - */ -void leave_mm(int cpu) -{ - if (per_cpu(cpu_tlbstate, cpu).state == TLBSTATE_OK) - BUG(); - cpu_clear(cpu, per_cpu(cpu_tlbstate, cpu).active_mm->cpu_vm_mask); - load_cr3(swapper_pg_dir); -} -EXPORT_SYMBOL_GPL(leave_mm); - -/* - * - * The flush IPI assumes that a thread switch happens in this order: - * [cpu0: the cpu that switches] - * 1) switch_mm() either 1a) or 1b) - * 1a) thread switch to a different mm - * 1a1) cpu_clear(cpu, old_mm->cpu_vm_mask); - * Stop ipi delivery for the old mm. This is not synchronized with - * the other cpus, but smp_invalidate_interrupt ignore flush ipis - * for the wrong mm, and in the worst case we perform a superfluous - * tlb flush. - * 1a2) set cpu_tlbstate to TLBSTATE_OK - * Now the smp_invalidate_interrupt won't call leave_mm if cpu0 - * was in lazy tlb mode. - * 1a3) update cpu_tlbstate[].active_mm - * Now cpu0 accepts tlb flushes for the new mm. - * 1a4) cpu_set(cpu, new_mm->cpu_vm_mask); - * Now the other cpus will send tlb flush ipis. - * 1a4) change cr3. - * 1b) thread switch without mm change - * cpu_tlbstate[].active_mm is correct, cpu0 already handles - * flush ipis. - * 1b1) set cpu_tlbstate to TLBSTATE_OK - * 1b2) test_and_set the cpu bit in cpu_vm_mask. - * Atomically set the bit [other cpus will start sending flush ipis], - * and test the bit. - * 1b3) if the bit was 0: leave_mm was called, flush the tlb. - * 2) switch %%esp, ie current - * - * The interrupt must handle 2 special cases: - * - cr3 is changed before %%esp, ie. it cannot use current->{active_,}mm. - * - the cpu performs speculative tlb reads, i.e. even if the cpu only - * runs in kernel space, the cpu could load tlb entries for user space - * pages. - * - * The good news is that cpu_tlbstate is local to each cpu, no - * write/read ordering problems. - */ - -/* - * TLB flush IPI: - * - * 1) Flush the tlb entries if the cpu uses the mm that's being flushed. - * 2) Leave the mm if we are in the lazy tlb mode. - */ - -void smp_invalidate_interrupt(struct pt_regs *regs) -{ - unsigned long cpu; - - cpu = get_cpu(); - - if (!cpu_isset(cpu, flush_cpumask)) - goto out; - /* - * This was a BUG() but until someone can quote me the - * line from the intel manual that guarantees an IPI to - * multiple CPUs is retried _only_ on the erroring CPUs - * its staying as a return - * - * BUG(); - */ - - if (flush_mm == per_cpu(cpu_tlbstate, cpu).active_mm) { - if (per_cpu(cpu_tlbstate, cpu).state == TLBSTATE_OK) { - if (flush_va == TLB_FLUSH_ALL) - local_flush_tlb(); - else - __flush_tlb_one(flush_va); - } else - leave_mm(cpu); - } - ack_APIC_irq(); - smp_mb__before_clear_bit(); - cpu_clear(cpu, flush_cpumask); - smp_mb__after_clear_bit(); -out: - put_cpu_no_resched(); - __get_cpu_var(irq_stat).irq_tlb_count++; -} - -void native_flush_tlb_others(const cpumask_t *cpumaskp, struct mm_struct *mm, - unsigned long va) -{ - cpumask_t cpumask = *cpumaskp; - - /* - * A couple of (to be removed) sanity checks: - * - * - current CPU must not be in mask - * - mask must exist :) - */ - BUG_ON(cpus_empty(cpumask)); - BUG_ON(cpu_isset(smp_processor_id(), cpumask)); - BUG_ON(!mm); - -#ifdef CONFIG_HOTPLUG_CPU - /* If a CPU which we ran on has gone down, OK. */ - cpus_and(cpumask, cpumask, cpu_online_map); - if (unlikely(cpus_empty(cpumask))) - return; -#endif - - /* - * i'm not happy about this global shared spinlock in the - * MM hot path, but we'll see how contended it is. - * AK: x86-64 has a faster method that could be ported. - */ - spin_lock(&tlbstate_lock); - - flush_mm = mm; - flush_va = va; - cpus_or(flush_cpumask, cpumask, flush_cpumask); - /* - * We have to send the IPI only to - * CPUs affected. - */ - send_IPI_mask(cpumask, INVALIDATE_TLB_VECTOR); - - while (!cpus_empty(flush_cpumask)) - /* nothing. lockup detection does not belong here */ - cpu_relax(); - - flush_mm = NULL; - flush_va = 0; - spin_unlock(&tlbstate_lock); -} - -void flush_tlb_current_task(void) -{ - struct mm_struct *mm = current->mm; - cpumask_t cpu_mask; - - preempt_disable(); - cpu_mask = mm->cpu_vm_mask; - cpu_clear(smp_processor_id(), cpu_mask); - - local_flush_tlb(); - if (!cpus_empty(cpu_mask)) - flush_tlb_others(cpu_mask, mm, TLB_FLUSH_ALL); - preempt_enable(); -} - -void flush_tlb_mm (struct mm_struct * mm) -{ - cpumask_t cpu_mask; - - preempt_disable(); - cpu_mask = mm->cpu_vm_mask; - cpu_clear(smp_processor_id(), cpu_mask); - - if (current->active_mm == mm) { - if (current->mm) - local_flush_tlb(); - else - leave_mm(smp_processor_id()); - } - if (!cpus_empty(cpu_mask)) - flush_tlb_others(cpu_mask, mm, TLB_FLUSH_ALL); - - preempt_enable(); -} - -void flush_tlb_page(struct vm_area_struct * vma, unsigned long va) -{ - struct mm_struct *mm = vma->vm_mm; - cpumask_t cpu_mask; - - preempt_disable(); - cpu_mask = mm->cpu_vm_mask; - cpu_clear(smp_processor_id(), cpu_mask); - - if (current->active_mm == mm) { - if(current->mm) - __flush_tlb_one(va); - else - leave_mm(smp_processor_id()); - } - - if (!cpus_empty(cpu_mask)) - flush_tlb_others(cpu_mask, mm, va); - - preempt_enable(); -} -EXPORT_SYMBOL(flush_tlb_page); - -static void do_flush_tlb_all(void* info) -{ - unsigned long cpu = smp_processor_id(); - - __flush_tlb_all(); - if (per_cpu(cpu_tlbstate, cpu).state == TLBSTATE_LAZY) - leave_mm(cpu); -} - -void flush_tlb_all(void) -{ - on_each_cpu(do_flush_tlb_all, NULL, 1, 1); -} - -/* - * this function sends a 'reschedule' IPI to another CPU. - * it goes straight through and wastes no time serializing - * anything. Worst case is that we lose a reschedule ... - */ -static void native_smp_send_reschedule(int cpu) -{ - WARN_ON(cpu_is_offline(cpu)); - send_IPI_mask(cpumask_of_cpu(cpu), RESCHEDULE_VECTOR); -} - -/* - * Structure and data for smp_call_function(). This is designed to minimise - * static memory requirements. It also looks cleaner. - */ -static DEFINE_SPINLOCK(call_lock); - -struct call_data_struct { - void (*func) (void *info); - void *info; - atomic_t started; - atomic_t finished; - int wait; -}; - -void lock_ipi_call_lock(void) -{ - spin_lock_irq(&call_lock); -} - -void unlock_ipi_call_lock(void) -{ - spin_unlock_irq(&call_lock); -} - -static struct call_data_struct *call_data; - -static void __smp_call_function(void (*func) (void *info), void *info, - int nonatomic, int wait) -{ - struct call_data_struct data; - int cpus = num_online_cpus() - 1; - - if (!cpus) - return; - - data.func = func; - data.info = info; - atomic_set(&data.started, 0); - data.wait = wait; - if (wait) - atomic_set(&data.finished, 0); - - call_data = &data; - mb(); - - /* Send a message to all other CPUs and wait for them to respond */ - send_IPI_allbutself(CALL_FUNCTION_VECTOR); - - /* Wait for response */ - while (atomic_read(&data.started) != cpus) - cpu_relax(); - - if (wait) - while (atomic_read(&data.finished) != cpus) - cpu_relax(); -} - - -/** - * smp_call_function_mask(): Run a function on a set of other CPUs. - * @mask: The set of cpus to run on. Must not include the current cpu. - * @func: The function to run. This must be fast and non-blocking. - * @info: An arbitrary pointer to pass to the function. - * @wait: If true, wait (atomically) until function has completed on other CPUs. - * - * Returns 0 on success, else a negative status code. - * - * If @wait is true, then returns once @func has returned; otherwise - * it returns just before the target cpu calls @func. - * - * You must not call this function with disabled interrupts or from a - * hardware interrupt handler or from a bottom half handler. - */ -static int -native_smp_call_function_mask(cpumask_t mask, - void (*func)(void *), void *info, - int wait) -{ - struct call_data_struct data; - cpumask_t allbutself; - int cpus; - - /* Can deadlock when called with interrupts disabled */ - WARN_ON(irqs_disabled()); - - /* Holding any lock stops cpus from going down. */ - spin_lock(&call_lock); - - allbutself = cpu_online_map; - cpu_clear(smp_processor_id(), allbutself); - - cpus_and(mask, mask, allbutself); - cpus = cpus_weight(mask); - - if (!cpus) { - spin_unlock(&call_lock); - return 0; - } - - data.func = func; - data.info = info; - atomic_set(&data.started, 0); - data.wait = wait; - if (wait) - atomic_set(&data.finished, 0); - - call_data = &data; - mb(); - - /* Send a message to other CPUs */ - if (cpus_equal(mask, allbutself)) - send_IPI_allbutself(CALL_FUNCTION_VECTOR); - else - send_IPI_mask(mask, CALL_FUNCTION_VECTOR); - - /* Wait for response */ - while (atomic_read(&data.started) != cpus) - cpu_relax(); - - if (wait) - while (atomic_read(&data.finished) != cpus) - cpu_relax(); - spin_unlock(&call_lock); - - return 0; -} - -static void stop_this_cpu (void * dummy) -{ - local_irq_disable(); - /* - * Remove this CPU: - */ - cpu_clear(smp_processor_id(), cpu_online_map); - disable_local_APIC(); - if (cpu_data(smp_processor_id()).hlt_works_ok) - for(;;) halt(); - for (;;); -} - -/* - * this function calls the 'stop' function on all other CPUs in the system. - */ - -static void native_smp_send_stop(void) -{ - /* Don't deadlock on the call lock in panic */ - int nolock = !spin_trylock(&call_lock); - unsigned long flags; - - local_irq_save(flags); - __smp_call_function(stop_this_cpu, NULL, 0, 0); - if (!nolock) - spin_unlock(&call_lock); - disable_local_APIC(); - local_irq_restore(flags); -} - -/* - * Reschedule call back. Nothing to do, - * all the work is done automatically when - * we return from the interrupt. - */ -void smp_reschedule_interrupt(struct pt_regs *regs) -{ - ack_APIC_irq(); - __get_cpu_var(irq_stat).irq_resched_count++; -} - -void smp_call_function_interrupt(struct pt_regs *regs) -{ - void (*func) (void *info) = call_data->func; - void *info = call_data->info; - int wait = call_data->wait; - - ack_APIC_irq(); - /* - * Notify initiating CPU that I've grabbed the data and am - * about to execute the function - */ - mb(); - atomic_inc(&call_data->started); - /* - * At this point the info structure may be out of scope unless wait==1 - */ - irq_enter(); - (*func)(info); - __get_cpu_var(irq_stat).irq_call_count++; - irq_exit(); - - if (wait) { - mb(); - atomic_inc(&call_data->finished); - } -} - -static int convert_apicid_to_cpu(int apic_id) -{ - int i; - - for_each_possible_cpu(i) { - if (per_cpu(x86_cpu_to_apicid, i) == apic_id) - return i; - } - return -1; -} - -int safe_smp_processor_id(void) -{ - int apicid, cpuid; - - if (!boot_cpu_has(X86_FEATURE_APIC)) - return 0; - - apicid = hard_smp_processor_id(); - if (apicid == BAD_APICID) - return 0; - - cpuid = convert_apicid_to_cpu(apicid); - - return cpuid >= 0 ? cpuid : 0; -} - -struct smp_ops smp_ops = { - .smp_prepare_boot_cpu = native_smp_prepare_boot_cpu, - .smp_prepare_cpus = native_smp_prepare_cpus, - .cpu_up = native_cpu_up, - .smp_cpus_done = native_smp_cpus_done, - - .smp_send_stop = native_smp_send_stop, - .smp_send_reschedule = native_smp_send_reschedule, - .smp_call_function_mask = native_smp_call_function_mask, -}; -EXPORT_SYMBOL_GPL(smp_ops); diff --git a/arch/x86/kernel/smpboot_32.c b/arch/x86/kernel/smpboot.c index 579b9b740c7c..e6abe8a49b1f 100644 --- a/arch/x86/kernel/smpboot_32.c +++ b/arch/x86/kernel/smpboot.c @@ -3,6 +3,7 @@ * * (c) 1995 Alan Cox, Building #3 <alan@redhat.com> * (c) 1998, 1999, 2000 Ingo Molnar <mingo@redhat.com> + * Copyright 2001 Andi Kleen, SuSE Labs. * * Much of the core SMP work is based on previous work by Thomas Radke, to * whom a great many thanks are extended. @@ -29,53 +30,90 @@ * Ingo Molnar : various cleanups and rewrites * Tigran Aivazian : fixed "0.00 in /proc/uptime on SMP" bug. * Maciej W. Rozycki : Bits for genuine 82489DX APICs + * Andi Kleen : Changed for SMP boot into long mode. * Martin J. Bligh : Added support for multi-quad systems * Dave Jones : Report invalid combinations of Athlon CPUs. -* Rusty Russell : Hacked into shape for new "hotplug" boot process. */ + * Rusty Russell : Hacked into shape for new "hotplug" boot process. + * Andi Kleen : Converted to new state machine. + * Ashok Raj : CPU hotplug support + * Glauber Costa : i386 and x86_64 integration + */ -#include <linux/module.h> #include <linux/init.h> -#include <linux/kernel.h> - -#include <linux/mm.h> +#include <linux/smp.h> +#include <linux/module.h> #include <linux/sched.h> -#include <linux/kernel_stat.h> -#include <linux/bootmem.h> -#include <linux/notifier.h> -#include <linux/cpu.h> #include <linux/percpu.h> +#include <linux/bootmem.h> +#include <linux/err.h> #include <linux/nmi.h> -#include <linux/delay.h> -#include <linux/mc146818rtc.h> -#include <asm/tlbflush.h> +#include <asm/acpi.h> #include <asm/desc.h> -#include <asm/arch_hooks.h> #include <asm/nmi.h> +#include <asm/irq.h> +#include <asm/smp.h> +#include <asm/trampoline.h> +#include <asm/cpu.h> +#include <asm/numa.h> +#include <asm/pgtable.h> +#include <asm/tlbflush.h> +#include <asm/mtrr.h> +#include <asm/nmi.h> +#include <asm/vmi.h> +#include <linux/mc146818rtc.h> #include <mach_apic.h> #include <mach_wakecpu.h> #include <smpboot_hooks.h> -#include <asm/vmi.h> -#include <asm/mtrr.h> -/* Set if we find a B stepping CPU */ -static int __cpuinitdata smp_b_stepping; +/* + * FIXME: For x86_64, those are defined in other files. But moving them here, + * would make the setup areas dependent on smp, which is a loss. When we + * integrate apic between arches, we can probably do a better job, but + * right now, they'll stay here -- glommer + */ + +/* which logical CPU number maps to which CPU (physical APIC ID) */ +u16 x86_cpu_to_apicid_init[NR_CPUS] __initdata = + { [0 ... NR_CPUS-1] = BAD_APICID }; +void *x86_cpu_to_apicid_early_ptr; + +u16 x86_bios_cpu_apicid_init[NR_CPUS] __initdata + = { [0 ... NR_CPUS-1] = BAD_APICID }; +void *x86_bios_cpu_apicid_early_ptr; + +#ifdef CONFIG_X86_32 +u8 apicid_2_node[MAX_APICID]; +#endif + +/* State of each CPU */ +DEFINE_PER_CPU(int, cpu_state) = { 0 }; + +/* Store all idle threads, this can be reused instead of creating +* a new thread. Also avoids complicated thread destroy functionality +* for idle threads. +*/ +#ifdef CONFIG_HOTPLUG_CPU +/* + * Needed only for CONFIG_HOTPLUG_CPU because __cpuinitdata is + * removed after init for !CONFIG_HOTPLUG_CPU. + */ +static DEFINE_PER_CPU(struct task_struct *, idle_thread_array); +#define get_idle_for_cpu(x) (per_cpu(idle_thread_array, x)) +#define set_idle_for_cpu(x, p) (per_cpu(idle_thread_array, x) = (p)) +#else +struct task_struct *idle_thread_array[NR_CPUS] __cpuinitdata ; +#define get_idle_for_cpu(x) (idle_thread_array[(x)]) +#define set_idle_for_cpu(x, p) (idle_thread_array[(x)] = (p)) +#endif /* Number of siblings per CPU package */ int smp_num_siblings = 1; EXPORT_SYMBOL(smp_num_siblings); /* Last level cache ID of each logical CPU */ -DEFINE_PER_CPU(u8, cpu_llc_id) = BAD_APICID; - -/* representing HT siblings of each logical CPU */ -DEFINE_PER_CPU(cpumask_t, cpu_sibling_map); -EXPORT_PER_CPU_SYMBOL(cpu_sibling_map); - -/* representing HT and core siblings of each logical CPU */ -DEFINE_PER_CPU(cpumask_t, cpu_core_map); -EXPORT_PER_CPU_SYMBOL(cpu_core_map); +DEFINE_PER_CPU(u16, cpu_llc_id) = BAD_APICID; /* bitmap of online cpus */ cpumask_t cpu_online_map __read_mostly; @@ -85,126 +123,94 @@ cpumask_t cpu_callin_map; cpumask_t cpu_callout_map; cpumask_t cpu_possible_map; EXPORT_SYMBOL(cpu_possible_map); -static cpumask_t smp_commenced_mask; + +/* representing HT siblings of each logical CPU */ +DEFINE_PER_CPU(cpumask_t, cpu_sibling_map); +EXPORT_PER_CPU_SYMBOL(cpu_sibling_map); + +/* representing HT and core siblings of each logical CPU */ +DEFINE_PER_CPU(cpumask_t, cpu_core_map); +EXPORT_PER_CPU_SYMBOL(cpu_core_map); /* Per CPU bogomips and other parameters */ DEFINE_PER_CPU_SHARED_ALIGNED(struct cpuinfo_x86, cpu_info); EXPORT_PER_CPU_SYMBOL(cpu_info); -/* which logical CPU number maps to which CPU (physical APIC ID) */ -u8 x86_cpu_to_apicid_init[NR_CPUS] __initdata = - { [0 ... NR_CPUS-1] = BAD_APICID }; -void *x86_cpu_to_apicid_early_ptr; -DEFINE_PER_CPU(u8, x86_cpu_to_apicid) = BAD_APICID; -EXPORT_PER_CPU_SYMBOL(x86_cpu_to_apicid); - -u8 apicid_2_node[MAX_APICID]; +static atomic_t init_deasserted; -/* - * Trampoline 80x86 program as an array. - */ +static int boot_cpu_logical_apicid; -extern const unsigned char trampoline_data []; -extern const unsigned char trampoline_end []; -static unsigned char *trampoline_base; +/* representing cpus for which sibling maps can be computed */ +static cpumask_t cpu_sibling_setup_map; -static void map_cpu_to_logical_apicid(void); +/* Set if we find a B stepping CPU */ +int __cpuinitdata smp_b_stepping; -/* State of each CPU. */ -DEFINE_PER_CPU(int, cpu_state) = { 0 }; +#if defined(CONFIG_NUMA) && defined(CONFIG_X86_32) -/* - * Currently trivial. Write the real->protected mode - * bootstrap into the page concerned. The caller - * has made sure it's suitably aligned. - */ +/* which logical CPUs are on which nodes */ +cpumask_t node_to_cpumask_map[MAX_NUMNODES] __read_mostly = + { [0 ... MAX_NUMNODES-1] = CPU_MASK_NONE }; +EXPORT_SYMBOL(node_to_cpumask_map); +/* which node each logical CPU is on */ +int cpu_to_node_map[NR_CPUS] __read_mostly = { [0 ... NR_CPUS-1] = 0 }; +EXPORT_SYMBOL(cpu_to_node_map); -static unsigned long __cpuinit setup_trampoline(void) +/* set up a mapping between cpu and node. */ +static void map_cpu_to_node(int cpu, int node) { - memcpy(trampoline_base, trampoline_data, trampoline_end - trampoline_data); - return virt_to_phys(trampoline_base); + printk(KERN_INFO "Mapping cpu %d to node %d\n", cpu, node); + cpu_set(cpu, node_to_cpumask_map[node]); + cpu_to_node_map[cpu] = node; } -/* - * We are called very early to get the low memory for the - * SMP bootup trampoline page. - */ -void __init smp_alloc_memory(void) +/* undo a mapping between cpu and node. */ +static void unmap_cpu_to_node(int cpu) { - trampoline_base = alloc_bootmem_low_pages(PAGE_SIZE); - /* - * Has to be in very low memory so we can execute - * real-mode AP code. - */ - if (__pa(trampoline_base) >= 0x9F000) - BUG(); + int node; + + printk(KERN_INFO "Unmapping cpu %d from all nodes\n", cpu); + for (node = 0; node < MAX_NUMNODES; node++) + cpu_clear(cpu, node_to_cpumask_map[node]); + cpu_to_node_map[cpu] = 0; } +#else /* !(CONFIG_NUMA && CONFIG_X86_32) */ +#define map_cpu_to_node(cpu, node) ({}) +#define unmap_cpu_to_node(cpu) ({}) +#endif -/* - * The bootstrap kernel entry code has set these up. Save them for - * a given CPU - */ +#ifdef CONFIG_X86_32 +u8 cpu_2_logical_apicid[NR_CPUS] __read_mostly = + { [0 ... NR_CPUS-1] = BAD_APICID }; -void __cpuinit smp_store_cpu_info(int id) +void map_cpu_to_logical_apicid(void) { - struct cpuinfo_x86 *c = &cpu_data(id); - - *c = boot_cpu_data; - c->cpu_index = id; - if (id!=0) - identify_secondary_cpu(c); - /* - * Mask B, Pentium, but not Pentium MMX - */ - if (c->x86_vendor == X86_VENDOR_INTEL && - c->x86 == 5 && - c->x86_mask >= 1 && c->x86_mask <= 4 && - c->x86_model <= 3) - /* - * Remember we have B step Pentia with bugs - */ - smp_b_stepping = 1; - - /* - * Certain Athlons might work (for various values of 'work') in SMP - * but they are not certified as MP capable. - */ - if ((c->x86_vendor == X86_VENDOR_AMD) && (c->x86 == 6)) { - - if (num_possible_cpus() == 1) - goto valid_k7; - - /* Athlon 660/661 is valid. */ - if ((c->x86_model==6) && ((c->x86_mask==0) || (c->x86_mask==1))) - goto valid_k7; - - /* Duron 670 is valid */ - if ((c->x86_model==7) && (c->x86_mask==0)) - goto valid_k7; - - /* - * Athlon 662, Duron 671, and Athlon >model 7 have capability bit. - * It's worth noting that the A5 stepping (662) of some Athlon XP's - * have the MP bit set. - * See http://www.heise.de/newsticker/data/jow-18.10.01-000 for more. - */ - if (((c->x86_model==6) && (c->x86_mask>=2)) || - ((c->x86_model==7) && (c->x86_mask>=1)) || - (c->x86_model> 7)) - if (cpu_has_mp) - goto valid_k7; + int cpu = smp_processor_id(); + int apicid = logical_smp_processor_id(); + int node = apicid_to_node(apicid); - /* If we get here, it's not a certified SMP capable AMD system. */ - add_taint(TAINT_UNSAFE_SMP); - } + if (!node_online(node)) + node = first_online_node; -valid_k7: - ; + cpu_2_logical_apicid[cpu] = apicid; + map_cpu_to_node(cpu, node); } -static atomic_t init_deasserted; +void unmap_cpu_to_logical_apicid(int cpu) +{ + cpu_2_logical_apicid[cpu] = BAD_APICID; + unmap_cpu_to_node(cpu); +} +#else +#define unmap_cpu_to_logical_apicid(cpu) do {} while (0) +#define map_cpu_to_logical_apicid() do {} while (0) +#endif -static void __cpuinit smp_callin(void) +/* + * Report back to the Boot Processor. + * Running on AP. + */ +void __cpuinit smp_callin(void) { int cpuid, phys_id; unsigned long timeout; @@ -220,12 +226,11 @@ static void __cpuinit smp_callin(void) /* * (This works even if the APIC is not enabled.) */ - phys_id = GET_APIC_ID(apic_read(APIC_ID)); + phys_id = GET_APIC_ID(read_apic_id()); cpuid = smp_processor_id(); if (cpu_isset(cpuid, cpu_callin_map)) { - printk("huh, phys CPU#%d, CPU#%d already present??\n", + panic("%s: phys CPU#%d, CPU#%d already present??\n", __func__, phys_id, cpuid); - BUG(); } Dprintk("CPU#%d (phys ID: %d) waiting for CALLOUT\n", cpuid, phys_id); @@ -247,13 +252,12 @@ static void __cpuinit smp_callin(void) */ if (cpu_isset(cpuid, cpu_callout_map)) break; - rep_nop(); + cpu_relax(); } if (!time_before(jiffies, timeout)) { - printk("BUG: CPU%d started up but did not get a callout!\n", - cpuid); - BUG(); + panic("%s: CPU%d started up but did not get a callout!\n", + __func__, cpuid); } /* @@ -266,13 +270,19 @@ static void __cpuinit smp_callin(void) Dprintk("CALLIN, before setup_local_APIC().\n"); smp_callin_clear_local_apic(); setup_local_APIC(); + end_local_APIC_setup(); map_cpu_to_logical_apicid(); /* * Get our bogomips. + * + * Need to enable IRQs because it can take longer and then + * the NMI watchdog might kill us. */ + local_irq_enable(); calibrate_delay(); - Dprintk("Stack at about %p\n",&cpuid); + local_irq_disable(); + Dprintk("Stack at about %p\n", &cpuid); /* * Save our processor parameters @@ -285,91 +295,10 @@ static void __cpuinit smp_callin(void) cpu_set(cpuid, cpu_callin_map); } -static int cpucount; - -/* maps the cpu to the sched domain representing multi-core */ -cpumask_t cpu_coregroup_map(int cpu) -{ - struct cpuinfo_x86 *c = &cpu_data(cpu); - /* - * For perf, we return last level cache shared map. - * And for power savings, we return cpu_core_map - */ - if (sched_mc_power_savings || sched_smt_power_savings) - return per_cpu(cpu_core_map, cpu); - else - return c->llc_shared_map; -} - -/* representing cpus for which sibling maps can be computed */ -static cpumask_t cpu_sibling_setup_map; - -void __cpuinit set_cpu_sibling_map(int cpu) -{ - int i; - struct cpuinfo_x86 *c = &cpu_data(cpu); - - cpu_set(cpu, cpu_sibling_setup_map); - - if (smp_num_siblings > 1) { - for_each_cpu_mask(i, cpu_sibling_setup_map) { - if (c->phys_proc_id == cpu_data(i).phys_proc_id && - c->cpu_core_id == cpu_data(i).cpu_core_id) { - cpu_set(i, per_cpu(cpu_sibling_map, cpu)); - cpu_set(cpu, per_cpu(cpu_sibling_map, i)); - cpu_set(i, per_cpu(cpu_core_map, cpu)); - cpu_set(cpu, per_cpu(cpu_core_map, i)); - cpu_set(i, c->llc_shared_map); - cpu_set(cpu, cpu_data(i).llc_shared_map); - } - } - } else { - cpu_set(cpu, per_cpu(cpu_sibling_map, cpu)); - } - - cpu_set(cpu, c->llc_shared_map); - - if (current_cpu_data.x86_max_cores == 1) { - per_cpu(cpu_core_map, cpu) = per_cpu(cpu_sibling_map, cpu); - c->booted_cores = 1; - return; - } - - for_each_cpu_mask(i, cpu_sibling_setup_map) { - if (per_cpu(cpu_llc_id, cpu) != BAD_APICID && - per_cpu(cpu_llc_id, cpu) == per_cpu(cpu_llc_id, i)) { - cpu_set(i, c->llc_shared_map); - cpu_set(cpu, cpu_data(i).llc_shared_map); - } - if (c->phys_proc_id == cpu_data(i).phys_proc_id) { - cpu_set(i, per_cpu(cpu_core_map, cpu)); - cpu_set(cpu, per_cpu(cpu_core_map, i)); - /* - * Does this new cpu bringup a new core? - */ - if (cpus_weight(per_cpu(cpu_sibling_map, cpu)) == 1) { - /* - * for each core in package, increment - * the booted_cores for this new cpu - */ - if (first_cpu(per_cpu(cpu_sibling_map, i)) == i) - c->booted_cores++; - /* - * increment the core count for all - * the other cpus in this package - */ - if (i != cpu) - cpu_data(i).booted_cores++; - } else if (i != cpu && !c->booted_cores) - c->booted_cores = cpu_data(i).booted_cores; - } - } -} - /* * Activate a secondary processor. */ -static void __cpuinit start_secondary(void *unused) +void __cpuinit start_secondary(void *unused) { /* * Don't put *anything* before cpu_init(), SMP booting is too @@ -382,24 +311,19 @@ static void __cpuinit start_secondary(void *unused) cpu_init(); preempt_disable(); smp_callin(); - while (!cpu_isset(smp_processor_id(), smp_commenced_mask)) - rep_nop(); + + /* otherwise gcc will move up smp_processor_id before the cpu_init */ + barrier(); /* * Check TSC synchronization with the BP: */ check_tsc_sync_target(); - setup_secondary_clock(); if (nmi_watchdog == NMI_IO_APIC) { disable_8259A_irq(0); enable_NMI_through_LVT0(); enable_8259A_irq(0); } - /* - * low-memory mappings have been cleared, flush them from - * the local TLBs too. - */ - local_flush_tlb(); /* This must be done before setting cpu_online_map */ set_cpu_sibling_map(raw_smp_processor_id()); @@ -414,17 +338,27 @@ static void __cpuinit start_secondary(void *unused) * smp_call_function(). */ lock_ipi_call_lock(); +#ifdef CONFIG_X86_64 + spin_lock(&vector_lock); + + /* Setup the per cpu irq handling data structures */ + __setup_vector_irq(smp_processor_id()); + /* + * Allow the master to continue. + */ + spin_unlock(&vector_lock); +#endif cpu_set(smp_processor_id(), cpu_online_map); unlock_ipi_call_lock(); per_cpu(cpu_state, smp_processor_id()) = CPU_ONLINE; - /* We can take interrupts now: we're officially "up". */ - local_irq_enable(); + setup_secondary_clock(); wmb(); cpu_idle(); } +#ifdef CONFIG_X86_32 /* * Everything has been set up for the secondary * CPUs - they just need to reload everything @@ -442,89 +376,233 @@ void __devinit initialize_secondary(void) "movl %0,%%esp\n\t" "jmp *%1" : - :"m" (current->thread.sp),"m" (current->thread.ip)); + :"m" (current->thread.sp), "m" (current->thread.ip)); } +#endif -/* Static state in head.S used to set up a CPU */ -extern struct { - void * sp; - unsigned short ss; -} stack_start; +static void __cpuinit smp_apply_quirks(struct cpuinfo_x86 *c) +{ +#ifdef CONFIG_X86_32 + /* + * Mask B, Pentium, but not Pentium MMX + */ + if (c->x86_vendor == X86_VENDOR_INTEL && + c->x86 == 5 && + c->x86_mask >= 1 && c->x86_mask <= 4 && + c->x86_model <= 3) + /* + * Remember we have B step Pentia with bugs + */ + smp_b_stepping = 1; -#ifdef CONFIG_NUMA + /* + * Certain Athlons might work (for various values of 'work') in SMP + * but they are not certified as MP capable. + */ + if ((c->x86_vendor == X86_VENDOR_AMD) && (c->x86 == 6)) { -/* which logical CPUs are on which nodes */ -cpumask_t node_to_cpumask_map[MAX_NUMNODES] __read_mostly = - { [0 ... MAX_NUMNODES-1] = CPU_MASK_NONE }; -EXPORT_SYMBOL(node_to_cpumask_map); -/* which node each logical CPU is on */ -int cpu_to_node_map[NR_CPUS] __read_mostly = { [0 ... NR_CPUS-1] = 0 }; -EXPORT_SYMBOL(cpu_to_node_map); + if (num_possible_cpus() == 1) + goto valid_k7; -/* set up a mapping between cpu and node. */ -static inline void map_cpu_to_node(int cpu, int node) -{ - printk("Mapping cpu %d to node %d\n", cpu, node); - cpu_set(cpu, node_to_cpumask_map[node]); - cpu_to_node_map[cpu] = node; + /* Athlon 660/661 is valid. */ + if ((c->x86_model == 6) && ((c->x86_mask == 0) || + (c->x86_mask == 1))) + goto valid_k7; + + /* Duron 670 is valid */ + if ((c->x86_model == 7) && (c->x86_mask == 0)) + goto valid_k7; + + /* + * Athlon 662, Duron 671, and Athlon >model 7 have capability + * bit. It's worth noting that the A5 stepping (662) of some + * Athlon XP's have the MP bit set. + * See http://www.heise.de/newsticker/data/jow-18.10.01-000 for + * more. + */ + if (((c->x86_model == 6) && (c->x86_mask >= 2)) || + ((c->x86_model == 7) && (c->x86_mask >= 1)) || + (c->x86_model > 7)) + if (cpu_has_mp) + goto valid_k7; + + /* If we get here, not a certified SMP capable AMD system. */ + add_taint(TAINT_UNSAFE_SMP); + } + +valid_k7: + ; +#endif } -/* undo a mapping between cpu and node. */ -static inline void unmap_cpu_to_node(int cpu) +void __cpuinit smp_checks(void) { - int node; + if (smp_b_stepping) + printk(KERN_WARNING "WARNING: SMP operation may be unreliable" + "with B stepping processors.\n"); - printk("Unmapping cpu %d from all nodes\n", cpu); - for (node = 0; node < MAX_NUMNODES; node ++) - cpu_clear(cpu, node_to_cpumask_map[node]); - cpu_to_node_map[cpu] = 0; + /* + * Don't taint if we are running SMP kernel on a single non-MP + * approved Athlon + */ + if (tainted & TAINT_UNSAFE_SMP) { + if (num_online_cpus()) + printk(KERN_INFO "WARNING: This combination of AMD" + "processors is not suitable for SMP.\n"); + else + tainted &= ~TAINT_UNSAFE_SMP; + } } -#else /* !CONFIG_NUMA */ -#define map_cpu_to_node(cpu, node) ({}) -#define unmap_cpu_to_node(cpu) ({}) +/* + * The bootstrap kernel entry code has set these up. Save them for + * a given CPU + */ -#endif /* CONFIG_NUMA */ +void __cpuinit smp_store_cpu_info(int id) +{ + struct cpuinfo_x86 *c = &cpu_data(id); + + *c = boot_cpu_data; + c->cpu_index = id; + if (id != 0) + identify_secondary_cpu(c); + smp_apply_quirks(c); +} -u8 cpu_2_logical_apicid[NR_CPUS] __read_mostly = { [0 ... NR_CPUS-1] = BAD_APICID }; -static void map_cpu_to_logical_apicid(void) +void __cpuinit set_cpu_sibling_map(int cpu) { - int cpu = smp_processor_id(); - int apicid = logical_smp_processor_id(); - int node = apicid_to_node(apicid); + int i; + struct cpuinfo_x86 *c = &cpu_data(cpu); - if (!node_online(node)) - node = first_online_node; + cpu_set(cpu, cpu_sibling_setup_map); - cpu_2_logical_apicid[cpu] = apicid; - map_cpu_to_node(cpu, node); + if (smp_num_siblings > 1) { + for_each_cpu_mask(i, cpu_sibling_setup_map) { + if (c->phys_proc_id == cpu_data(i).phys_proc_id && + c->cpu_core_id == cpu_data(i).cpu_core_id) { + cpu_set(i, per_cpu(cpu_sibling_map, cpu)); + cpu_set(cpu, per_cpu(cpu_sibling_map, i)); + cpu_set(i, per_cpu(cpu_core_map, cpu)); + cpu_set(cpu, per_cpu(cpu_core_map, i)); + cpu_set(i, c->llc_shared_map); + cpu_set(cpu, cpu_data(i).llc_shared_map); + } + } + } else { + cpu_set(cpu, per_cpu(cpu_sibling_map, cpu)); + } + + cpu_set(cpu, c->llc_shared_map); + + if (current_cpu_data.x86_max_cores == 1) { + per_cpu(cpu_core_map, cpu) = per_cpu(cpu_sibling_map, cpu); + c->booted_cores = 1; + return; + } + + for_each_cpu_mask(i, cpu_sibling_setup_map) { + if (per_cpu(cpu_llc_id, cpu) != BAD_APICID && + per_cpu(cpu_llc_id, cpu) == per_cpu(cpu_llc_id, i)) { + cpu_set(i, c->llc_shared_map); + cpu_set(cpu, cpu_data(i).llc_shared_map); + } + if (c->phys_proc_id == cpu_data(i).phys_proc_id) { + cpu_set(i, per_cpu(cpu_core_map, cpu)); + cpu_set(cpu, per_cpu(cpu_core_map, i)); + /* + * Does this new cpu bringup a new core? + */ + if (cpus_weight(per_cpu(cpu_sibling_map, cpu)) == 1) { + /* + * for each core in package, increment + * the booted_cores for this new cpu + */ + if (first_cpu(per_cpu(cpu_sibling_map, i)) == i) + c->booted_cores++; + /* + * increment the core count for all + * the other cpus in this package + */ + if (i != cpu) + cpu_data(i).booted_cores++; + } else if (i != cpu && !c->booted_cores) + c->booted_cores = cpu_data(i).booted_cores; + } + } } -static void unmap_cpu_to_logical_apicid(int cpu) +/* maps the cpu to the sched domain representing multi-core */ +cpumask_t cpu_coregroup_map(int cpu) { - cpu_2_logical_apicid[cpu] = BAD_APICID; - unmap_cpu_to_node(cpu); + struct cpuinfo_x86 *c = &cpu_data(cpu); + /* + * For perf, we return last level cache shared map. + * And for power savings, we return cpu_core_map + */ + if (sched_mc_power_savings || sched_smt_power_savings) + return per_cpu(cpu_core_map, cpu); + else + return c->llc_shared_map; +} + +#ifdef CONFIG_X86_32 +/* + * We are called very early to get the low memory for the + * SMP bootup trampoline page. + */ +void __init smp_alloc_memory(void) +{ + trampoline_base = alloc_bootmem_low_pages(PAGE_SIZE); + /* + * Has to be in very low memory so we can execute + * real-mode AP code. + */ + if (__pa(trampoline_base) >= 0x9F000) + BUG(); +} +#endif + +void impress_friends(void) +{ + int cpu; + unsigned long bogosum = 0; + /* + * Allow the user to impress friends. + */ + Dprintk("Before bogomips.\n"); + for_each_possible_cpu(cpu) + if (cpu_isset(cpu, cpu_callout_map)) + bogosum += cpu_data(cpu).loops_per_jiffy; + printk(KERN_INFO + "Total of %d processors activated (%lu.%02lu BogoMIPS).\n", + num_online_cpus(), + bogosum/(500000/HZ), + (bogosum/(5000/HZ))%100); + + Dprintk("Before bogocount - setting activated=1.\n"); } static inline void __inquire_remote_apic(int apicid) { - int i, regs[] = { APIC_ID >> 4, APIC_LVR >> 4, APIC_SPIV >> 4 }; + unsigned i, regs[] = { APIC_ID >> 4, APIC_LVR >> 4, APIC_SPIV >> 4 }; char *names[] = { "ID", "VERSION", "SPIV" }; int timeout; - unsigned long status; + u32 status; - printk("Inquiring remote APIC #%d...\n", apicid); + printk(KERN_INFO "Inquiring remote APIC #%d...\n", apicid); for (i = 0; i < ARRAY_SIZE(regs); i++) { - printk("... APIC #%d %s: ", apicid, names[i]); + printk(KERN_INFO "... APIC #%d %s: ", apicid, names[i]); /* * Wait for idle. */ status = safe_apic_wait_icr_idle(); if (status) - printk("a previous APIC delivery may have failed\n"); + printk(KERN_CONT + "a previous APIC delivery may have failed\n"); apic_write_around(APIC_ICR2, SET_APIC_DEST_FIELD(apicid)); apic_write_around(APIC_ICR, APIC_DM_REMRD | regs[i]); @@ -538,16 +616,16 @@ static inline void __inquire_remote_apic(int apicid) switch (status) { case APIC_ICR_RR_VALID: status = apic_read(APIC_RRR); - printk("%lx\n", status); + printk(KERN_CONT "%08x\n", status); break; default: - printk("failed\n"); + printk(KERN_CONT "failed\n"); } } } #ifdef WAKE_SECONDARY_VIA_NMI -/* +/* * Poke the other CPU in the eye via NMI to wake it up. Remember that the normal * INIT, INIT, STARTUP sequence will reset the chip hard for us, and this * won't ... remember to clear down the APIC, etc later. @@ -584,9 +662,9 @@ wakeup_secondary_cpu(int logical_apicid, unsigned long start_eip) Dprintk("NMI sent.\n"); if (send_status) - printk("APIC never delivered???\n"); + printk(KERN_ERR "APIC never delivered???\n"); if (accept_status) - printk("APIC delivery error (%lx).\n", accept_status); + printk(KERN_ERR "APIC delivery error (%lx).\n", accept_status); return (send_status | accept_status); } @@ -637,6 +715,7 @@ wakeup_secondary_cpu(int phys_apicid, unsigned long start_eip) Dprintk("Waiting for send to finish...\n"); send_status = safe_apic_wait_icr_idle(); + mb(); atomic_set(&init_deasserted, 1); /* @@ -655,7 +734,11 @@ wakeup_secondary_cpu(int phys_apicid, unsigned long start_eip) * target processor state. */ startup_ipi_hook(phys_apicid, (unsigned long) start_secondary, - (unsigned long) stack_start.sp); +#ifdef CONFIG_X86_64 + (unsigned long)init_rsp); +#else + (unsigned long)stack_start.sp); +#endif /* * Run STARTUP IPI loop. @@ -665,7 +748,7 @@ wakeup_secondary_cpu(int phys_apicid, unsigned long start_eip) maxlvt = lapic_get_maxlvt(); for (j = 1; j <= num_starts; j++) { - Dprintk("Sending STARTUP #%d.\n",j); + Dprintk("Sending STARTUP #%d.\n", j); apic_read_around(APIC_SPIV); apic_write(APIC_ESR, 0); apic_read(APIC_ESR); @@ -711,49 +794,29 @@ wakeup_secondary_cpu(int phys_apicid, unsigned long start_eip) Dprintk("After Startup.\n"); if (send_status) - printk("APIC never delivered???\n"); + printk(KERN_ERR "APIC never delivered???\n"); if (accept_status) - printk("APIC delivery error (%lx).\n", accept_status); + printk(KERN_ERR "APIC delivery error (%lx).\n", accept_status); return (send_status | accept_status); } #endif /* WAKE_SECONDARY_VIA_INIT */ -extern cpumask_t cpu_initialized; -static inline int alloc_cpu_id(void) -{ - cpumask_t tmp_map; +struct create_idle { + struct work_struct work; + struct task_struct *idle; + struct completion done; int cpu; - cpus_complement(tmp_map, cpu_present_map); - cpu = first_cpu(tmp_map); - if (cpu >= NR_CPUS) - return -ENODEV; - return cpu; -} +}; -#ifdef CONFIG_HOTPLUG_CPU -static struct task_struct * __cpuinitdata cpu_idle_tasks[NR_CPUS]; -static inline struct task_struct * __cpuinit alloc_idle_task(int cpu) +static void __cpuinit do_fork_idle(struct work_struct *work) { - struct task_struct *idle; + struct create_idle *c_idle = + container_of(work, struct create_idle, work); - if ((idle = cpu_idle_tasks[cpu]) != NULL) { - /* initialize thread_struct. we really want to avoid destroy - * idle tread - */ - idle->thread.sp = (unsigned long)task_pt_regs(idle); - init_idle(idle, cpu); - return idle; - } - idle = fork_idle(cpu); - - if (!IS_ERR(idle)) - cpu_idle_tasks[cpu] = idle; - return idle; + c_idle->idle = fork_idle(c_idle->cpu); + complete(&c_idle->done); } -#else -#define alloc_idle_task(cpu) fork_idle(cpu) -#endif static int __cpuinit do_boot_cpu(int apicid, int cpu) /* @@ -762,45 +825,92 @@ static int __cpuinit do_boot_cpu(int apicid, int cpu) * Returns zero if CPU booted OK, else error code from wakeup_secondary_cpu. */ { - struct task_struct *idle; - unsigned long boot_error; + unsigned long boot_error = 0; int timeout; - unsigned long start_eip; + unsigned long start_ip; unsigned short nmi_high = 0, nmi_low = 0; + struct create_idle c_idle = { + .cpu = cpu, + .done = COMPLETION_INITIALIZER_ONSTACK(c_idle.done), + }; + INIT_WORK(&c_idle.work, do_fork_idle); +#ifdef CONFIG_X86_64 + /* allocate memory for gdts of secondary cpus. Hotplug is considered */ + if (!cpu_gdt_descr[cpu].address && + !(cpu_gdt_descr[cpu].address = get_zeroed_page(GFP_KERNEL))) { + printk(KERN_ERR "Failed to allocate GDT for CPU %d\n", cpu); + return -1; + } - /* - * Save current MTRR state in case it was changed since early boot - * (e.g. by the ACPI SMI) to initialize new CPUs with MTRRs in sync: - */ - mtrr_save_state(); + /* Allocate node local memory for AP pdas */ + if (cpu_pda(cpu) == &boot_cpu_pda[cpu]) { + struct x8664_pda *newpda, *pda; + int node = cpu_to_node(cpu); + pda = cpu_pda(cpu); + newpda = kmalloc_node(sizeof(struct x8664_pda), GFP_ATOMIC, + node); + if (newpda) { + memcpy(newpda, pda, sizeof(struct x8664_pda)); + cpu_pda(cpu) = newpda; + } else + printk(KERN_ERR + "Could not allocate node local PDA for CPU %d on node %d\n", + cpu, node); + } +#endif + + alternatives_smp_switch(1); + + c_idle.idle = get_idle_for_cpu(cpu); /* * We can't use kernel_thread since we must avoid to * reschedule the child. */ - idle = alloc_idle_task(cpu); - if (IS_ERR(idle)) - panic("failed fork for CPU %d", cpu); + if (c_idle.idle) { + c_idle.idle->thread.sp = (unsigned long) (((struct pt_regs *) + (THREAD_SIZE + task_stack_page(c_idle.idle))) - 1); + init_idle(c_idle.idle, cpu); + goto do_rest; + } + if (!keventd_up() || current_is_keventd()) + c_idle.work.func(&c_idle.work); + else { + schedule_work(&c_idle.work); + wait_for_completion(&c_idle.done); + } + + if (IS_ERR(c_idle.idle)) { + printk("failed fork for CPU %d\n", cpu); + return PTR_ERR(c_idle.idle); + } + + set_idle_for_cpu(cpu, c_idle.idle); +do_rest: +#ifdef CONFIG_X86_32 + per_cpu(current_task, cpu) = c_idle.idle; init_gdt(cpu); - per_cpu(current_task, cpu) = idle; early_gdt_descr.address = (unsigned long)get_cpu_gdt_table(cpu); + c_idle.idle->thread.ip = (unsigned long) start_secondary; + /* Stack for startup_32 can be just as for start_secondary onwards */ + stack_start.sp = (void *) c_idle.idle->thread.sp; + irq_ctx_init(cpu); +#else + cpu_pda(cpu)->pcurrent = c_idle.idle; + init_rsp = c_idle.idle->thread.sp; + load_sp0(&per_cpu(init_tss, cpu), &c_idle.idle->thread); + initial_code = (unsigned long)start_secondary; + clear_tsk_thread_flag(c_idle.idle, TIF_FORK); +#endif - idle->thread.ip = (unsigned long) start_secondary; - /* start_eip had better be page-aligned! */ - start_eip = setup_trampoline(); - - ++cpucount; - alternatives_smp_switch(1); + /* start_ip had better be page-aligned! */ + start_ip = setup_trampoline(); /* So we see what's up */ - printk("Booting processor %d/%d ip %lx\n", cpu, apicid, start_eip); - /* Stack for startup_32 can be just as for start_secondary onwards */ - stack_start.sp = (void *) idle->thread.sp; + printk(KERN_INFO "Booting processor %d/%d ip %lx\n", + cpu, apicid, start_ip); - irq_ctx_init(cpu); - - per_cpu(x86_cpu_to_apicid, cpu) = apicid; /* * This grunge runs the startup process for * the targeted processor. @@ -812,12 +922,17 @@ static int __cpuinit do_boot_cpu(int apicid, int cpu) store_NMI_vector(&nmi_high, &nmi_low); - smpboot_setup_warm_reset_vector(start_eip); + smpboot_setup_warm_reset_vector(start_ip); + /* + * Be paranoid about clearing APIC errors. + */ + apic_write(APIC_ESR, 0); + apic_read(APIC_ESR); /* * Starting actual IPI sequence... */ - boot_error = wakeup_secondary_cpu(apicid, start_eip); + boot_error = wakeup_secondary_cpu(apicid, start_ip); if (!boot_error) { /* @@ -839,18 +954,18 @@ static int __cpuinit do_boot_cpu(int apicid, int cpu) if (cpu_isset(cpu, cpu_callin_map)) { /* number CPUs logically, starting from 1 (BSP is 0) */ Dprintk("OK.\n"); - printk("CPU%d: ", cpu); + printk(KERN_INFO "CPU%d: ", cpu); print_cpu_info(&cpu_data(cpu)); Dprintk("CPU has booted.\n"); } else { - boot_error= 1; + boot_error = 1; if (*((volatile unsigned char *)trampoline_base) == 0xA5) /* trampoline started but...? */ - printk("Stuck ??\n"); + printk(KERN_ERR "Stuck ??\n"); else /* trampoline code not run */ - printk("Not responding.\n"); + printk(KERN_ERR "Not responding.\n"); inquire_remote_apic(apicid); } } @@ -858,156 +973,159 @@ static int __cpuinit do_boot_cpu(int apicid, int cpu) if (boot_error) { /* Try to put things back the way they were before ... */ unmap_cpu_to_logical_apicid(cpu); - cpu_clear(cpu, cpu_callout_map); /* was set here (do_boot_cpu()) */ +#ifdef CONFIG_X86_64 + clear_node_cpumask(cpu); /* was set by numa_add_cpu */ +#endif + cpu_clear(cpu, cpu_callout_map); /* was set by do_boot_cpu() */ cpu_clear(cpu, cpu_initialized); /* was set by cpu_init() */ - cpucount--; - } else { - per_cpu(x86_cpu_to_apicid, cpu) = apicid; - cpu_set(cpu, cpu_present_map); + cpu_clear(cpu, cpu_possible_map); + cpu_clear(cpu, cpu_present_map); + per_cpu(x86_cpu_to_apicid, cpu) = BAD_APICID; } /* mark "stuck" area as not stuck */ *((volatile unsigned long *)trampoline_base) = 0; + /* + * Cleanup possible dangling ends... + */ + smpboot_restore_warm_reset_vector(); + return boot_error; } -#ifdef CONFIG_HOTPLUG_CPU -void cpu_exit_clear(void) +int __cpuinit native_cpu_up(unsigned int cpu) { - int cpu = raw_smp_processor_id(); - - idle_task_exit(); - - cpucount --; - cpu_uninit(); - irq_ctx_exit(cpu); - - cpu_clear(cpu, cpu_callout_map); - cpu_clear(cpu, cpu_callin_map); + int apicid = cpu_present_to_apicid(cpu); + unsigned long flags; + int err; - cpu_clear(cpu, smp_commenced_mask); - unmap_cpu_to_logical_apicid(cpu); -} + WARN_ON(irqs_disabled()); -struct warm_boot_cpu_info { - struct completion *complete; - struct work_struct task; - int apicid; - int cpu; -}; + Dprintk("++++++++++++++++++++=_---CPU UP %u\n", cpu); -static void __cpuinit do_warm_boot_cpu(struct work_struct *work) -{ - struct warm_boot_cpu_info *info = - container_of(work, struct warm_boot_cpu_info, task); - do_boot_cpu(info->apicid, info->cpu); - complete(info->complete); -} + if (apicid == BAD_APICID || apicid == boot_cpu_physical_apicid || + !physid_isset(apicid, phys_cpu_present_map)) { + printk(KERN_ERR "%s: bad cpu %d\n", __func__, cpu); + return -EINVAL; + } -static int __cpuinit __smp_prepare_cpu(int cpu) -{ - DECLARE_COMPLETION_ONSTACK(done); - struct warm_boot_cpu_info info; - int apicid, ret; - - apicid = per_cpu(x86_cpu_to_apicid, cpu); - if (apicid == BAD_APICID) { - ret = -ENODEV; - goto exit; + /* + * Already booted CPU? + */ + if (cpu_isset(cpu, cpu_callin_map)) { + Dprintk("do_boot_cpu %d Already started\n", cpu); + return -ENOSYS; } - info.complete = &done; - info.apicid = apicid; - info.cpu = cpu; - INIT_WORK(&info.task, do_warm_boot_cpu); + /* + * Save current MTRR state in case it was changed since early boot + * (e.g. by the ACPI SMI) to initialize new CPUs with MTRRs in sync: + */ + mtrr_save_state(); + + per_cpu(cpu_state, cpu) = CPU_UP_PREPARE; +#ifdef CONFIG_X86_32 /* init low mem mapping */ clone_pgd_range(swapper_pg_dir, swapper_pg_dir + USER_PGD_PTRS, min_t(unsigned long, KERNEL_PGD_PTRS, USER_PGD_PTRS)); flush_tlb_all(); - schedule_work(&info.task); - wait_for_completion(&done); - - zap_low_mappings(); - ret = 0; -exit: - return ret; -} -#endif - -/* - * Cycle through the processors sending APIC IPIs to boot each. - */ - -static int boot_cpu_logical_apicid; -/* Where the IO area was mapped on multiquad, always 0 otherwise */ -void *xquad_portio; -#ifdef CONFIG_X86_NUMAQ -EXPORT_SYMBOL(xquad_portio); #endif -static void __init smp_boot_cpus(unsigned int max_cpus) -{ - int apicid, cpu, bit, kicked; - unsigned long bogosum = 0; + err = do_boot_cpu(apicid, cpu); + if (err < 0) { + Dprintk("do_boot_cpu failed %d\n", err); + return err; + } /* - * Setup boot CPU information + * Check TSC synchronization with the AP (keep irqs disabled + * while doing so): */ - smp_store_cpu_info(0); /* Final full version of the data */ - printk("CPU%d: ", 0); - print_cpu_info(&cpu_data(0)); + local_irq_save(flags); + check_tsc_sync_source(cpu); + local_irq_restore(flags); - boot_cpu_physical_apicid = GET_APIC_ID(apic_read(APIC_ID)); - boot_cpu_logical_apicid = logical_smp_processor_id(); - per_cpu(x86_cpu_to_apicid, 0) = boot_cpu_physical_apicid; + while (!cpu_isset(cpu, cpu_online_map)) { + cpu_relax(); + touch_nmi_watchdog(); + } - current_thread_info()->cpu = 0; + return 0; +} - set_cpu_sibling_map(0); +/* + * Fall back to non SMP mode after errors. + * + * RED-PEN audit/test this more. I bet there is more state messed up here. + */ +static __init void disable_smp(void) +{ + cpu_present_map = cpumask_of_cpu(0); + cpu_possible_map = cpumask_of_cpu(0); +#ifdef CONFIG_X86_32 + smpboot_clear_io_apic_irqs(); +#endif + if (smp_found_config) + phys_cpu_present_map = + physid_mask_of_physid(boot_cpu_physical_apicid); + else + phys_cpu_present_map = physid_mask_of_physid(0); + map_cpu_to_logical_apicid(); + cpu_set(0, per_cpu(cpu_sibling_map, 0)); + cpu_set(0, per_cpu(cpu_core_map, 0)); +} + +/* + * Various sanity checks. + */ +static int __init smp_sanity_check(unsigned max_cpus) +{ + preempt_disable(); + if (!physid_isset(hard_smp_processor_id(), phys_cpu_present_map)) { + printk(KERN_WARNING "weird, boot CPU (#%d) not listed" + "by the BIOS.\n", hard_smp_processor_id()); + physid_set(hard_smp_processor_id(), phys_cpu_present_map); + } /* * If we couldn't find an SMP configuration at boot time, * get out of here now! */ if (!smp_found_config && !acpi_lapic) { + preempt_enable(); printk(KERN_NOTICE "SMP motherboard not detected.\n"); - smpboot_clear_io_apic_irqs(); - phys_cpu_present_map = physid_mask_of_physid(0); + disable_smp(); if (APIC_init_uniprocessor()) printk(KERN_NOTICE "Local APIC not detected." " Using dummy APIC emulation.\n"); - map_cpu_to_logical_apicid(); - cpu_set(0, per_cpu(cpu_sibling_map, 0)); - cpu_set(0, per_cpu(cpu_core_map, 0)); - return; + return -1; } /* * Should not be necessary because the MP table should list the boot * CPU too, but we do it for the sake of robustness anyway. - * Makes no sense to do this check in clustered apic mode, so skip it */ if (!check_phys_apicid_present(boot_cpu_physical_apicid)) { - printk("weird, boot CPU (#%d) not listed by the BIOS.\n", - boot_cpu_physical_apicid); + printk(KERN_NOTICE + "weird, boot CPU (#%d) not listed by the BIOS.\n", + boot_cpu_physical_apicid); physid_set(hard_smp_processor_id(), phys_cpu_present_map); } + preempt_enable(); /* * If we couldn't find a local APIC, then get out of here now! */ - if (APIC_INTEGRATED(apic_version[boot_cpu_physical_apicid]) && !cpu_has_apic) { + if (APIC_INTEGRATED(apic_version[boot_cpu_physical_apicid]) && + !cpu_has_apic) { printk(KERN_ERR "BIOS bug, local APIC #%d not detected!...\n", boot_cpu_physical_apicid); - printk(KERN_ERR "... forcing use of dummy APIC emulation. (tell your hw vendor)\n"); - smpboot_clear_io_apic_irqs(); - phys_cpu_present_map = physid_mask_of_physid(0); - map_cpu_to_logical_apicid(); - cpu_set(0, per_cpu(cpu_sibling_map, 0)); - cpu_set(0, per_cpu(cpu_core_map, 0)); - return; + printk(KERN_ERR "... forcing use of dummy APIC emulation." + "(tell your hw vendor)\n"); + smpboot_clear_io_apic(); + return -1; } verify_local_APIC(); @@ -1016,137 +1134,148 @@ static void __init smp_boot_cpus(unsigned int max_cpus) * If SMP should be disabled, then really disable it! */ if (!max_cpus) { - smp_found_config = 0; - printk(KERN_INFO "SMP mode deactivated, forcing use of dummy APIC emulation.\n"); - + printk(KERN_INFO "SMP mode deactivated," + "forcing use of dummy APIC emulation.\n"); + smpboot_clear_io_apic(); +#ifdef CONFIG_X86_32 if (nmi_watchdog == NMI_LOCAL_APIC) { - printk(KERN_INFO "activating minimal APIC for NMI watchdog use.\n"); + printk(KERN_INFO "activating minimal APIC for" + "NMI watchdog use.\n"); connect_bsp_APIC(); setup_local_APIC(); + end_local_APIC_setup(); } - smpboot_clear_io_apic_irqs(); - phys_cpu_present_map = physid_mask_of_physid(0); - map_cpu_to_logical_apicid(); - cpu_set(0, per_cpu(cpu_sibling_map, 0)); - cpu_set(0, per_cpu(cpu_core_map, 0)); - return; +#endif + return -1; } - connect_bsp_APIC(); - setup_local_APIC(); - map_cpu_to_logical_apicid(); + return 0; +} +static void __init smp_cpu_index_default(void) +{ + int i; + struct cpuinfo_x86 *c; - setup_portio_remap(); + for_each_cpu_mask(i, cpu_possible_map) { + c = &cpu_data(i); + /* mark all to hotplug */ + c->cpu_index = NR_CPUS; + } +} +/* + * Prepare for SMP bootup. The MP table or ACPI has been read + * earlier. Just do some sanity checking here and enable APIC mode. + */ +void __init native_smp_prepare_cpus(unsigned int max_cpus) +{ + nmi_watchdog_default(); + smp_cpu_index_default(); + current_cpu_data = boot_cpu_data; + cpu_callin_map = cpumask_of_cpu(0); + mb(); /* - * Scan the CPU present map and fire up the other CPUs via do_boot_cpu - * - * In clustered apic mode, phys_cpu_present_map is a constructed thus: - * bits 0-3 are quad0, 4-7 are quad1, etc. A perverse twist on the - * clustered apic ID. + * Setup boot CPU information */ - Dprintk("CPU present map: %lx\n", physids_coerce(phys_cpu_present_map)); - - kicked = 1; - for (bit = 0; kicked < NR_CPUS && bit < MAX_APICS; bit++) { - apicid = cpu_present_to_apicid(bit); - /* - * Don't even attempt to start the boot CPU! - */ - if ((apicid == boot_cpu_apicid) || (apicid == BAD_APICID)) - continue; + smp_store_cpu_info(0); /* Final full version of the data */ + boot_cpu_logical_apicid = logical_smp_processor_id(); + current_thread_info()->cpu = 0; /* needed? */ + set_cpu_sibling_map(0); - if (!check_apicid_present(bit)) - continue; - if (max_cpus <= cpucount+1) - continue; + if (smp_sanity_check(max_cpus) < 0) { + printk(KERN_INFO "SMP disabled\n"); + disable_smp(); + return; + } - if (((cpu = alloc_cpu_id()) <= 0) || do_boot_cpu(apicid, cpu)) - printk("CPU #%d not responding - cannot use it.\n", - apicid); - else - ++kicked; + preempt_disable(); + if (GET_APIC_ID(read_apic_id()) != boot_cpu_physical_apicid) { + panic("Boot APIC ID in local APIC unexpected (%d vs %d)", + GET_APIC_ID(read_apic_id()), boot_cpu_physical_apicid); + /* Or can we switch back to PIC here? */ } + preempt_enable(); +#ifdef CONFIG_X86_32 + connect_bsp_APIC(); +#endif /* - * Cleanup possible dangling ends... + * Switch from PIC to APIC mode. */ - smpboot_restore_warm_reset_vector(); + setup_local_APIC(); +#ifdef CONFIG_X86_64 /* - * Allow the user to impress friends. + * Enable IO APIC before setting up error vector */ - Dprintk("Before bogomips.\n"); - for_each_possible_cpu(cpu) - if (cpu_isset(cpu, cpu_callout_map)) - bogosum += cpu_data(cpu).loops_per_jiffy; - printk(KERN_INFO - "Total of %d processors activated (%lu.%02lu BogoMIPS).\n", - cpucount+1, - bogosum/(500000/HZ), - (bogosum/(5000/HZ))%100); - - Dprintk("Before bogocount - setting activated=1.\n"); - - if (smp_b_stepping) - printk(KERN_WARNING "WARNING: SMP operation may be unreliable with B stepping processors.\n"); + if (!skip_ioapic_setup && nr_ioapics) + enable_IO_APIC(); +#endif + end_local_APIC_setup(); - /* - * Don't taint if we are running SMP kernel on a single non-MP - * approved Athlon - */ - if (tainted & TAINT_UNSAFE_SMP) { - if (cpucount) - printk (KERN_INFO "WARNING: This combination of AMD processors is not suitable for SMP.\n"); - else - tainted &= ~TAINT_UNSAFE_SMP; - } + map_cpu_to_logical_apicid(); - Dprintk("Boot done.\n"); + setup_portio_remap(); + smpboot_setup_io_apic(); /* - * construct cpu_sibling_map, so that we can tell sibling CPUs - * efficiently. + * Set up local APIC timer on boot CPU. */ - for_each_possible_cpu(cpu) { - cpus_clear(per_cpu(cpu_sibling_map, cpu)); - cpus_clear(per_cpu(cpu_core_map, cpu)); - } - - cpu_set(0, per_cpu(cpu_sibling_map, 0)); - cpu_set(0, per_cpu(cpu_core_map, 0)); - - smpboot_setup_io_apic(); + printk(KERN_INFO "CPU%d: ", 0); + print_cpu_info(&cpu_data(0)); setup_boot_clock(); } +/* + * Early setup to make printk work. + */ +void __init native_smp_prepare_boot_cpu(void) +{ + int me = smp_processor_id(); +#ifdef CONFIG_X86_32 + init_gdt(me); + switch_to_new_gdt(); +#endif + /* already set me in cpu_online_map in boot_cpu_init() */ + cpu_set(me, cpu_callout_map); + per_cpu(cpu_state, me) = CPU_ONLINE; +} -/* These are wrappers to interface to the new boot process. Someone - who understands all this stuff should rewrite it properly. --RR 15/Jul/02 */ -void __init native_smp_prepare_cpus(unsigned int max_cpus) +void __init native_smp_cpus_done(unsigned int max_cpus) { - smp_commenced_mask = cpumask_of_cpu(0); - cpu_callin_map = cpumask_of_cpu(0); - mb(); - smp_boot_cpus(max_cpus); + Dprintk("Boot done.\n"); + + impress_friends(); + smp_checks(); +#ifdef CONFIG_X86_IO_APIC + setup_ioapic_dest(); +#endif + check_nmi_watchdog(); +#ifdef CONFIG_X86_32 + zap_low_mappings(); +#endif } -void __init native_smp_prepare_boot_cpu(void) +#ifdef CONFIG_HOTPLUG_CPU + +# ifdef CONFIG_X86_32 +void cpu_exit_clear(void) { - unsigned int cpu = smp_processor_id(); + int cpu = raw_smp_processor_id(); - init_gdt(cpu); - switch_to_new_gdt(); + idle_task_exit(); + + cpu_uninit(); + irq_ctx_exit(cpu); + + cpu_clear(cpu, cpu_callout_map); + cpu_clear(cpu, cpu_callin_map); - cpu_set(cpu, cpu_online_map); - cpu_set(cpu, cpu_callout_map); - cpu_set(cpu, cpu_present_map); - cpu_set(cpu, cpu_possible_map); - __get_cpu_var(cpu_state) = CPU_ONLINE; + unmap_cpu_to_logical_apicid(cpu); } +# endif /* CONFIG_X86_32 */ -#ifdef CONFIG_HOTPLUG_CPU void remove_siblinginfo(int cpu) { int sibling; @@ -1160,7 +1289,7 @@ void remove_siblinginfo(int cpu) if (cpus_weight(per_cpu(cpu_sibling_map, cpu)) == 1) cpu_data(sibling).booted_cores--; } - + for_each_cpu_mask(sibling, per_cpu(cpu_sibling_map, cpu)) cpu_clear(cpu, per_cpu(cpu_sibling_map, sibling)); cpus_clear(per_cpu(cpu_sibling_map, cpu)); @@ -1170,35 +1299,99 @@ void remove_siblinginfo(int cpu) cpu_clear(cpu, cpu_sibling_setup_map); } +int additional_cpus __initdata = -1; + +static __init int setup_additional_cpus(char *s) +{ + return s && get_option(&s, &additional_cpus) ? 0 : -EINVAL; +} +early_param("additional_cpus", setup_additional_cpus); + +/* + * cpu_possible_map should be static, it cannot change as cpu's + * are onlined, or offlined. The reason is per-cpu data-structures + * are allocated by some modules at init time, and dont expect to + * do this dynamically on cpu arrival/departure. + * cpu_present_map on the other hand can change dynamically. + * In case when cpu_hotplug is not compiled, then we resort to current + * behaviour, which is cpu_possible == cpu_present. + * - Ashok Raj + * + * Three ways to find out the number of additional hotplug CPUs: + * - If the BIOS specified disabled CPUs in ACPI/mptables use that. + * - The user can overwrite it with additional_cpus=NUM + * - Otherwise don't reserve additional CPUs. + * We do this because additional CPUs waste a lot of memory. + * -AK + */ +__init void prefill_possible_map(void) +{ + int i; + int possible; + + if (additional_cpus == -1) { + if (disabled_cpus > 0) + additional_cpus = disabled_cpus; + else + additional_cpus = 0; + } + possible = num_processors + additional_cpus; + if (possible > NR_CPUS) + possible = NR_CPUS; + + printk(KERN_INFO "SMP: Allowing %d CPUs, %d hotplug CPUs\n", + possible, max_t(int, possible - num_processors, 0)); + + for (i = 0; i < possible; i++) + cpu_set(i, cpu_possible_map); +} + +static void __ref remove_cpu_from_maps(int cpu) +{ + cpu_clear(cpu, cpu_online_map); +#ifdef CONFIG_X86_64 + cpu_clear(cpu, cpu_callout_map); + cpu_clear(cpu, cpu_callin_map); + /* was set by cpu_init() */ + clear_bit(cpu, (unsigned long *)&cpu_initialized); + clear_node_cpumask(cpu); +#endif +} + int __cpu_disable(void) { - cpumask_t map = cpu_online_map; int cpu = smp_processor_id(); /* * Perhaps use cpufreq to drop frequency, but that could go * into generic code. - * + * * We won't take down the boot processor on i386 due to some * interrupts only being able to be serviced by the BSP. * Especially so if we're not using an IOAPIC -zwane */ if (cpu == 0) return -EBUSY; + if (nmi_watchdog == NMI_LOCAL_APIC) stop_apic_nmi_watchdog(NULL); clear_local_APIC(); - /* Allow any queued timer interrupts to get serviced */ + + /* + * HACK: + * Allow any queued timer interrupts to get serviced + * This is only a temporary solution until we cleanup + * fixup_irqs as we do for IA64. + */ local_irq_enable(); mdelay(1); - local_irq_disable(); + local_irq_disable(); remove_siblinginfo(cpu); - cpu_clear(cpu, map); - fixup_irqs(map); /* It's now safe to remove this processor from the online map */ - cpu_clear(cpu, cpu_online_map); + remove_cpu_from_maps(cpu); + fixup_irqs(cpu_online_map); return 0; } @@ -1210,14 +1403,14 @@ void __cpu_die(unsigned int cpu) for (i = 0; i < 10; i++) { /* They ack this in play_dead by setting CPU_DEAD */ if (per_cpu(cpu_state, cpu) == CPU_DEAD) { - printk ("CPU %d is now offline\n", cpu); + printk(KERN_INFO "CPU %d is now offline\n", cpu); if (1 == num_online_cpus()) alternatives_smp_switch(0); return; } msleep(100); } - printk(KERN_ERR "CPU %u didn't die...\n", cpu); + printk(KERN_ERR "CPU %u didn't die...\n", cpu); } #else /* ... !CONFIG_HOTPLUG_CPU */ int __cpu_disable(void) @@ -1230,81 +1423,7 @@ void __cpu_die(unsigned int cpu) /* We said "no" in __cpu_disable */ BUG(); } -#endif /* CONFIG_HOTPLUG_CPU */ - -int __cpuinit native_cpu_up(unsigned int cpu) -{ - unsigned long flags; -#ifdef CONFIG_HOTPLUG_CPU - int ret = 0; - - /* - * We do warm boot only on cpus that had booted earlier - * Otherwise cold boot is all handled from smp_boot_cpus(). - * cpu_callin_map is set during AP kickstart process. Its reset - * when a cpu is taken offline from cpu_exit_clear(). - */ - if (!cpu_isset(cpu, cpu_callin_map)) - ret = __smp_prepare_cpu(cpu); - - if (ret) - return -EIO; -#endif - - /* In case one didn't come up */ - if (!cpu_isset(cpu, cpu_callin_map)) { - printk(KERN_DEBUG "skipping cpu%d, didn't come online\n", cpu); - return -EIO; - } - - per_cpu(cpu_state, cpu) = CPU_UP_PREPARE; - /* Unleash the CPU! */ - cpu_set(cpu, smp_commenced_mask); - - /* - * Check TSC synchronization with the AP (keep irqs disabled - * while doing so): - */ - local_irq_save(flags); - check_tsc_sync_source(cpu); - local_irq_restore(flags); - - while (!cpu_isset(cpu, cpu_online_map)) { - cpu_relax(); - touch_nmi_watchdog(); - } - - return 0; -} - -void __init native_smp_cpus_done(unsigned int max_cpus) -{ -#ifdef CONFIG_X86_IO_APIC - setup_ioapic_dest(); #endif - zap_low_mappings(); -} - -void __init smp_intr_init(void) -{ - /* - * IRQ0 must be given a fixed assignment and initialized, - * because it's used before the IO-APIC is set up. - */ - set_intr_gate(FIRST_DEVICE_VECTOR, interrupt[0]); - - /* - * The reschedule interrupt is a CPU-to-CPU reschedule-helper - * IPI, driven by wakeup. - */ - set_intr_gate(RESCHEDULE_VECTOR, reschedule_interrupt); - - /* IPI for invalidation */ - set_intr_gate(INVALIDATE_TLB_VECTOR, invalidate_interrupt); - - /* IPI for generic function call */ - set_intr_gate(CALL_FUNCTION_VECTOR, call_function_interrupt); -} /* * If the BIOS enumerates physical processors before logical, diff --git a/arch/x86/kernel/smpboot_64.c b/arch/x86/kernel/smpboot_64.c deleted file mode 100644 index 0880f2c388a9..000000000000 --- a/arch/x86/kernel/smpboot_64.c +++ /dev/null @@ -1,1108 +0,0 @@ -/* - * x86 SMP booting functions - * - * (c) 1995 Alan Cox, Building #3 <alan@redhat.com> - * (c) 1998, 1999, 2000 Ingo Molnar <mingo@redhat.com> - * Copyright 2001 Andi Kleen, SuSE Labs. - * - * Much of the core SMP work is based on previous work by Thomas Radke, to - * whom a great many thanks are extended. - * - * Thanks to Intel for making available several different Pentium, - * Pentium Pro and Pentium-II/Xeon MP machines. - * Original development of Linux SMP code supported by Caldera. - * - * This code is released under the GNU General Public License version 2 - * - * Fixes - * Felix Koop : NR_CPUS used properly - * Jose Renau : Handle single CPU case. - * Alan Cox : By repeated request 8) - Total BogoMIP report. - * Greg Wright : Fix for kernel stacks panic. - * Erich Boleyn : MP v1.4 and additional changes. - * Matthias Sattler : Changes for 2.1 kernel map. - * Michel Lespinasse : Changes for 2.1 kernel map. - * Michael Chastain : Change trampoline.S to gnu as. - * Alan Cox : Dumb bug: 'B' step PPro's are fine - * Ingo Molnar : Added APIC timers, based on code - * from Jose Renau - * Ingo Molnar : various cleanups and rewrites - * Tigran Aivazian : fixed "0.00 in /proc/uptime on SMP" bug. - * Maciej W. Rozycki : Bits for genuine 82489DX APICs - * Andi Kleen : Changed for SMP boot into long mode. - * Rusty Russell : Hacked into shape for new "hotplug" boot process. - * Andi Kleen : Converted to new state machine. - * Various cleanups. - * Probably mostly hotplug CPU ready now. - * Ashok Raj : CPU hotplug support - */ - - -#include <linux/init.h> - -#include <linux/mm.h> -#include <linux/kernel_stat.h> -#include <linux/bootmem.h> -#include <linux/thread_info.h> -#include <linux/module.h> -#include <linux/delay.h> -#include <linux/mc146818rtc.h> -#include <linux/smp.h> -#include <linux/kdebug.h> - -#include <asm/mtrr.h> -#include <asm/pgalloc.h> -#include <asm/desc.h> -#include <asm/tlbflush.h> -#include <asm/proto.h> -#include <asm/nmi.h> -#include <asm/irq.h> -#include <asm/hw_irq.h> -#include <asm/numa.h> - -/* Number of siblings per CPU package */ -int smp_num_siblings = 1; -EXPORT_SYMBOL(smp_num_siblings); - -/* Last level cache ID of each logical CPU */ -DEFINE_PER_CPU(u16, cpu_llc_id) = BAD_APICID; - -/* Bitmask of currently online CPUs */ -cpumask_t cpu_online_map __read_mostly; - -EXPORT_SYMBOL(cpu_online_map); - -/* - * Private maps to synchronize booting between AP and BP. - * Probably not needed anymore, but it makes for easier debugging. -AK - */ -cpumask_t cpu_callin_map; -cpumask_t cpu_callout_map; -cpumask_t cpu_possible_map; -EXPORT_SYMBOL(cpu_possible_map); - -/* Per CPU bogomips and other parameters */ -DEFINE_PER_CPU_SHARED_ALIGNED(struct cpuinfo_x86, cpu_info); -EXPORT_PER_CPU_SYMBOL(cpu_info); - -/* Set when the idlers are all forked */ -int smp_threads_ready; - -/* representing HT siblings of each logical CPU */ -DEFINE_PER_CPU(cpumask_t, cpu_sibling_map); -EXPORT_PER_CPU_SYMBOL(cpu_sibling_map); - -/* representing HT and core siblings of each logical CPU */ -DEFINE_PER_CPU(cpumask_t, cpu_core_map); -EXPORT_PER_CPU_SYMBOL(cpu_core_map); - -/* - * Trampoline 80x86 program as an array. - */ - -extern const unsigned char trampoline_data[]; -extern const unsigned char trampoline_end[]; - -/* State of each CPU */ -DEFINE_PER_CPU(int, cpu_state) = { 0 }; - -/* - * Store all idle threads, this can be reused instead of creating - * a new thread. Also avoids complicated thread destroy functionality - * for idle threads. - */ -#ifdef CONFIG_HOTPLUG_CPU -/* - * Needed only for CONFIG_HOTPLUG_CPU because __cpuinitdata is - * removed after init for !CONFIG_HOTPLUG_CPU. - */ -static DEFINE_PER_CPU(struct task_struct *, idle_thread_array); -#define get_idle_for_cpu(x) (per_cpu(idle_thread_array, x)) -#define set_idle_for_cpu(x,p) (per_cpu(idle_thread_array, x) = (p)) -#else -struct task_struct *idle_thread_array[NR_CPUS] __cpuinitdata ; -#define get_idle_for_cpu(x) (idle_thread_array[(x)]) -#define set_idle_for_cpu(x,p) (idle_thread_array[(x)] = (p)) -#endif - - -/* - * Currently trivial. Write the real->protected mode - * bootstrap into the page concerned. The caller - * has made sure it's suitably aligned. - */ - -static unsigned long __cpuinit setup_trampoline(void) -{ - void *tramp = __va(SMP_TRAMPOLINE_BASE); - memcpy(tramp, trampoline_data, trampoline_end - trampoline_data); - return virt_to_phys(tramp); -} - -/* - * The bootstrap kernel entry code has set these up. Save them for - * a given CPU - */ - -static void __cpuinit smp_store_cpu_info(int id) -{ - struct cpuinfo_x86 *c = &cpu_data(id); - - *c = boot_cpu_data; - c->cpu_index = id; - identify_cpu(c); - print_cpu_info(c); -} - -static atomic_t init_deasserted __cpuinitdata; - -/* - * Report back to the Boot Processor. - * Running on AP. - */ -void __cpuinit smp_callin(void) -{ - int cpuid, phys_id; - unsigned long timeout; - - /* - * If waken up by an INIT in an 82489DX configuration - * we may get here before an INIT-deassert IPI reaches - * our local APIC. We have to wait for the IPI or we'll - * lock up on an APIC access. - */ - while (!atomic_read(&init_deasserted)) - cpu_relax(); - - /* - * (This works even if the APIC is not enabled.) - */ - phys_id = GET_APIC_ID(apic_read(APIC_ID)); - cpuid = smp_processor_id(); - if (cpu_isset(cpuid, cpu_callin_map)) { - panic("smp_callin: phys CPU#%d, CPU#%d already present??\n", - phys_id, cpuid); - } - Dprintk("CPU#%d (phys ID: %d) waiting for CALLOUT\n", cpuid, phys_id); - - /* - * STARTUP IPIs are fragile beasts as they might sometimes - * trigger some glue motherboard logic. Complete APIC bus - * silence for 1 second, this overestimates the time the - * boot CPU is spending to send the up to 2 STARTUP IPIs - * by a factor of two. This should be enough. - */ - - /* - * Waiting 2s total for startup (udelay is not yet working) - */ - timeout = jiffies + 2*HZ; - while (time_before(jiffies, timeout)) { - /* - * Has the boot CPU finished it's STARTUP sequence? - */ - if (cpu_isset(cpuid, cpu_callout_map)) - break; - cpu_relax(); - } - - if (!time_before(jiffies, timeout)) { - panic("smp_callin: CPU%d started up but did not get a callout!\n", - cpuid); - } - - /* - * the boot CPU has finished the init stage and is spinning - * on callin_map until we finish. We are free to set up this - * CPU, first the APIC. (this is probably redundant on most - * boards) - */ - - Dprintk("CALLIN, before setup_local_APIC().\n"); - setup_local_APIC(); - end_local_APIC_setup(); - - /* - * Get our bogomips. - * - * Need to enable IRQs because it can take longer and then - * the NMI watchdog might kill us. - */ - local_irq_enable(); - calibrate_delay(); - local_irq_disable(); - Dprintk("Stack at about %p\n",&cpuid); - - /* - * Save our processor parameters - */ - smp_store_cpu_info(cpuid); - - /* - * Allow the master to continue. - */ - cpu_set(cpuid, cpu_callin_map); -} - -/* maps the cpu to the sched domain representing multi-core */ -cpumask_t cpu_coregroup_map(int cpu) -{ - struct cpuinfo_x86 *c = &cpu_data(cpu); - /* - * For perf, we return last level cache shared map. - * And for power savings, we return cpu_core_map - */ - if (sched_mc_power_savings || sched_smt_power_savings) - return per_cpu(cpu_core_map, cpu); - else - return c->llc_shared_map; -} - -/* representing cpus for which sibling maps can be computed */ -static cpumask_t cpu_sibling_setup_map; - -static inline void set_cpu_sibling_map(int cpu) -{ - int i; - struct cpuinfo_x86 *c = &cpu_data(cpu); - - cpu_set(cpu, cpu_sibling_setup_map); - - if (smp_num_siblings > 1) { - for_each_cpu_mask(i, cpu_sibling_setup_map) { - if (c->phys_proc_id == cpu_data(i).phys_proc_id && - c->cpu_core_id == cpu_data(i).cpu_core_id) { - cpu_set(i, per_cpu(cpu_sibling_map, cpu)); - cpu_set(cpu, per_cpu(cpu_sibling_map, i)); - cpu_set(i, per_cpu(cpu_core_map, cpu)); - cpu_set(cpu, per_cpu(cpu_core_map, i)); - cpu_set(i, c->llc_shared_map); - cpu_set(cpu, cpu_data(i).llc_shared_map); - } - } - } else { - cpu_set(cpu, per_cpu(cpu_sibling_map, cpu)); - } - - cpu_set(cpu, c->llc_shared_map); - - if (current_cpu_data.x86_max_cores == 1) { - per_cpu(cpu_core_map, cpu) = per_cpu(cpu_sibling_map, cpu); - c->booted_cores = 1; - return; - } - - for_each_cpu_mask(i, cpu_sibling_setup_map) { - if (per_cpu(cpu_llc_id, cpu) != BAD_APICID && - per_cpu(cpu_llc_id, cpu) == per_cpu(cpu_llc_id, i)) { - cpu_set(i, c->llc_shared_map); - cpu_set(cpu, cpu_data(i).llc_shared_map); - } - if (c->phys_proc_id == cpu_data(i).phys_proc_id) { - cpu_set(i, per_cpu(cpu_core_map, cpu)); - cpu_set(cpu, per_cpu(cpu_core_map, i)); - /* - * Does this new cpu bringup a new core? - */ - if (cpus_weight(per_cpu(cpu_sibling_map, cpu)) == 1) { - /* - * for each core in package, increment - * the booted_cores for this new cpu - */ - if (first_cpu(per_cpu(cpu_sibling_map, i)) == i) - c->booted_cores++; - /* - * increment the core count for all - * the other cpus in this package - */ - if (i != cpu) - cpu_data(i).booted_cores++; - } else if (i != cpu && !c->booted_cores) - c->booted_cores = cpu_data(i).booted_cores; - } - } -} - -/* - * Setup code on secondary processor (after comming out of the trampoline) - */ -void __cpuinit start_secondary(void) -{ - /* - * Dont put anything before smp_callin(), SMP - * booting is too fragile that we want to limit the - * things done here to the most necessary things. - */ - cpu_init(); - preempt_disable(); - smp_callin(); - - /* otherwise gcc will move up the smp_processor_id before the cpu_init */ - barrier(); - - /* - * Check TSC sync first: - */ - check_tsc_sync_target(); - - if (nmi_watchdog == NMI_IO_APIC) { - disable_8259A_irq(0); - enable_NMI_through_LVT0(); - enable_8259A_irq(0); - } - - /* - * The sibling maps must be set before turing the online map on for - * this cpu - */ - set_cpu_sibling_map(smp_processor_id()); - - /* - * We need to hold call_lock, so there is no inconsistency - * between the time smp_call_function() determines number of - * IPI recipients, and the time when the determination is made - * for which cpus receive the IPI in genapic_flat.c. Holding this - * lock helps us to not include this cpu in a currently in progress - * smp_call_function(). - */ - lock_ipi_call_lock(); - spin_lock(&vector_lock); - - /* Setup the per cpu irq handling data structures */ - __setup_vector_irq(smp_processor_id()); - /* - * Allow the master to continue. - */ - cpu_set(smp_processor_id(), cpu_online_map); - per_cpu(cpu_state, smp_processor_id()) = CPU_ONLINE; - spin_unlock(&vector_lock); - - unlock_ipi_call_lock(); - - setup_secondary_clock(); - - cpu_idle(); -} - -extern volatile unsigned long init_rsp; -extern void (*initial_code)(void); - -#ifdef APIC_DEBUG -static void inquire_remote_apic(int apicid) -{ - unsigned i, regs[] = { APIC_ID >> 4, APIC_LVR >> 4, APIC_SPIV >> 4 }; - char *names[] = { "ID", "VERSION", "SPIV" }; - int timeout; - u32 status; - - printk(KERN_INFO "Inquiring remote APIC #%d...\n", apicid); - - for (i = 0; i < ARRAY_SIZE(regs); i++) { - printk(KERN_INFO "... APIC #%d %s: ", apicid, names[i]); - - /* - * Wait for idle. - */ - status = safe_apic_wait_icr_idle(); - if (status) - printk(KERN_CONT - "a previous APIC delivery may have failed\n"); - - apic_write(APIC_ICR2, SET_APIC_DEST_FIELD(apicid)); - apic_write(APIC_ICR, APIC_DM_REMRD | regs[i]); - - timeout = 0; - do { - udelay(100); - status = apic_read(APIC_ICR) & APIC_ICR_RR_MASK; - } while (status == APIC_ICR_RR_INPROG && timeout++ < 1000); - - switch (status) { - case APIC_ICR_RR_VALID: - status = apic_read(APIC_RRR); - printk(KERN_CONT "%08x\n", status); - break; - default: - printk(KERN_CONT "failed\n"); - } - } -} -#endif - -/* - * Kick the secondary to wake up. - */ -static int __cpuinit wakeup_secondary_via_INIT(int phys_apicid, unsigned int start_rip) -{ - unsigned long send_status, accept_status = 0; - int maxlvt, num_starts, j; - - Dprintk("Asserting INIT.\n"); - - /* - * Turn INIT on target chip - */ - apic_write(APIC_ICR2, SET_APIC_DEST_FIELD(phys_apicid)); - - /* - * Send IPI - */ - apic_write(APIC_ICR, APIC_INT_LEVELTRIG | APIC_INT_ASSERT - | APIC_DM_INIT); - - Dprintk("Waiting for send to finish...\n"); - send_status = safe_apic_wait_icr_idle(); - - mdelay(10); - - Dprintk("Deasserting INIT.\n"); - - /* Target chip */ - apic_write(APIC_ICR2, SET_APIC_DEST_FIELD(phys_apicid)); - - /* Send IPI */ - apic_write(APIC_ICR, APIC_INT_LEVELTRIG | APIC_DM_INIT); - - Dprintk("Waiting for send to finish...\n"); - send_status = safe_apic_wait_icr_idle(); - - mb(); - atomic_set(&init_deasserted, 1); - - num_starts = 2; - - /* - * Run STARTUP IPI loop. - */ - Dprintk("#startup loops: %d.\n", num_starts); - - maxlvt = lapic_get_maxlvt(); - - for (j = 1; j <= num_starts; j++) { - Dprintk("Sending STARTUP #%d.\n",j); - apic_write(APIC_ESR, 0); - apic_read(APIC_ESR); - Dprintk("After apic_write.\n"); - - /* - * STARTUP IPI - */ - - /* Target chip */ - apic_write(APIC_ICR2, SET_APIC_DEST_FIELD(phys_apicid)); - - /* Boot on the stack */ - /* Kick the second */ - apic_write(APIC_ICR, APIC_DM_STARTUP | (start_rip >> 12)); - - /* - * Give the other CPU some time to accept the IPI. - */ - udelay(300); - - Dprintk("Startup point 1.\n"); - - Dprintk("Waiting for send to finish...\n"); - send_status = safe_apic_wait_icr_idle(); - - /* - * Give the other CPU some time to accept the IPI. - */ - udelay(200); - /* - * Due to the Pentium erratum 3AP. - */ - if (maxlvt > 3) { - apic_write(APIC_ESR, 0); - } - accept_status = (apic_read(APIC_ESR) & 0xEF); - if (send_status || accept_status) - break; - } - Dprintk("After Startup.\n"); - - if (send_status) - printk(KERN_ERR "APIC never delivered???\n"); - if (accept_status) - printk(KERN_ERR "APIC delivery error (%lx).\n", accept_status); - - return (send_status | accept_status); -} - -struct create_idle { - struct work_struct work; - struct task_struct *idle; - struct completion done; - int cpu; -}; - -static void __cpuinit do_fork_idle(struct work_struct *work) -{ - struct create_idle *c_idle = - container_of(work, struct create_idle, work); - - c_idle->idle = fork_idle(c_idle->cpu); - complete(&c_idle->done); -} - -/* - * Boot one CPU. - */ -static int __cpuinit do_boot_cpu(int cpu, int apicid) -{ - unsigned long boot_error; - int timeout; - unsigned long start_rip; - struct create_idle c_idle = { - .cpu = cpu, - .done = COMPLETION_INITIALIZER_ONSTACK(c_idle.done), - }; - INIT_WORK(&c_idle.work, do_fork_idle); - - /* allocate memory for gdts of secondary cpus. Hotplug is considered */ - if (!cpu_gdt_descr[cpu].address && - !(cpu_gdt_descr[cpu].address = get_zeroed_page(GFP_KERNEL))) { - printk(KERN_ERR "Failed to allocate GDT for CPU %d\n", cpu); - return -1; - } - - /* Allocate node local memory for AP pdas */ - if (cpu_pda(cpu) == &boot_cpu_pda[cpu]) { - struct x8664_pda *newpda, *pda; - int node = cpu_to_node(cpu); - pda = cpu_pda(cpu); - newpda = kmalloc_node(sizeof (struct x8664_pda), GFP_ATOMIC, - node); - if (newpda) { - memcpy(newpda, pda, sizeof (struct x8664_pda)); - cpu_pda(cpu) = newpda; - } else - printk(KERN_ERR - "Could not allocate node local PDA for CPU %d on node %d\n", - cpu, node); - } - - alternatives_smp_switch(1); - - c_idle.idle = get_idle_for_cpu(cpu); - - if (c_idle.idle) { - c_idle.idle->thread.sp = (unsigned long) (((struct pt_regs *) - (THREAD_SIZE + task_stack_page(c_idle.idle))) - 1); - init_idle(c_idle.idle, cpu); - goto do_rest; - } - - /* - * During cold boot process, keventd thread is not spun up yet. - * When we do cpu hot-add, we create idle threads on the fly, we should - * not acquire any attributes from the calling context. Hence the clean - * way to create kernel_threads() is to do that from keventd(). - * We do the current_is_keventd() due to the fact that ACPI notifier - * was also queuing to keventd() and when the caller is already running - * in context of keventd(), we would end up with locking up the keventd - * thread. - */ - if (!keventd_up() || current_is_keventd()) - c_idle.work.func(&c_idle.work); - else { - schedule_work(&c_idle.work); - wait_for_completion(&c_idle.done); - } - - if (IS_ERR(c_idle.idle)) { - printk("failed fork for CPU %d\n", cpu); - return PTR_ERR(c_idle.idle); - } - - set_idle_for_cpu(cpu, c_idle.idle); - -do_rest: - - cpu_pda(cpu)->pcurrent = c_idle.idle; - - start_rip = setup_trampoline(); - - init_rsp = c_idle.idle->thread.sp; - load_sp0(&per_cpu(init_tss, cpu), &c_idle.idle->thread); - initial_code = start_secondary; - clear_tsk_thread_flag(c_idle.idle, TIF_FORK); - - printk(KERN_INFO "Booting processor %d/%d APIC 0x%x\n", cpu, - cpus_weight(cpu_present_map), - apicid); - - /* - * This grunge runs the startup process for - * the targeted processor. - */ - - atomic_set(&init_deasserted, 0); - - Dprintk("Setting warm reset code and vector.\n"); - - CMOS_WRITE(0xa, 0xf); - local_flush_tlb(); - Dprintk("1.\n"); - *((volatile unsigned short *) phys_to_virt(0x469)) = start_rip >> 4; - Dprintk("2.\n"); - *((volatile unsigned short *) phys_to_virt(0x467)) = start_rip & 0xf; - Dprintk("3.\n"); - - /* - * Be paranoid about clearing APIC errors. - */ - apic_write(APIC_ESR, 0); - apic_read(APIC_ESR); - - /* - * Status is now clean - */ - boot_error = 0; - - /* - * Starting actual IPI sequence... - */ - boot_error = wakeup_secondary_via_INIT(apicid, start_rip); - - if (!boot_error) { - /* - * allow APs to start initializing. - */ - Dprintk("Before Callout %d.\n", cpu); - cpu_set(cpu, cpu_callout_map); - Dprintk("After Callout %d.\n", cpu); - - /* - * Wait 5s total for a response - */ - for (timeout = 0; timeout < 50000; timeout++) { - if (cpu_isset(cpu, cpu_callin_map)) - break; /* It has booted */ - udelay(100); - } - - if (cpu_isset(cpu, cpu_callin_map)) { - /* number CPUs logically, starting from 1 (BSP is 0) */ - Dprintk("CPU has booted.\n"); - } else { - boot_error = 1; - if (*((volatile unsigned char *)phys_to_virt(SMP_TRAMPOLINE_BASE)) - == 0xA5) - /* trampoline started but...? */ - printk("Stuck ??\n"); - else - /* trampoline code not run */ - printk("Not responding.\n"); -#ifdef APIC_DEBUG - inquire_remote_apic(apicid); -#endif - } - } - if (boot_error) { - cpu_clear(cpu, cpu_callout_map); /* was set here (do_boot_cpu()) */ - clear_bit(cpu, (unsigned long *)&cpu_initialized); /* was set by cpu_init() */ - clear_node_cpumask(cpu); /* was set by numa_add_cpu */ - cpu_clear(cpu, cpu_present_map); - cpu_clear(cpu, cpu_possible_map); - per_cpu(x86_cpu_to_apicid, cpu) = BAD_APICID; - return -EIO; - } - - return 0; -} - -cycles_t cacheflush_time; -unsigned long cache_decay_ticks; - -/* - * Cleanup possible dangling ends... - */ -static __cpuinit void smp_cleanup_boot(void) -{ - /* - * Paranoid: Set warm reset code and vector here back - * to default values. - */ - CMOS_WRITE(0, 0xf); - - /* - * Reset trampoline flag - */ - *((volatile int *) phys_to_virt(0x467)) = 0; -} - -/* - * Fall back to non SMP mode after errors. - * - * RED-PEN audit/test this more. I bet there is more state messed up here. - */ -static __init void disable_smp(void) -{ - cpu_present_map = cpumask_of_cpu(0); - cpu_possible_map = cpumask_of_cpu(0); - if (smp_found_config) - phys_cpu_present_map = physid_mask_of_physid(boot_cpu_id); - else - phys_cpu_present_map = physid_mask_of_physid(0); - cpu_set(0, per_cpu(cpu_sibling_map, 0)); - cpu_set(0, per_cpu(cpu_core_map, 0)); -} - -#ifdef CONFIG_HOTPLUG_CPU - -int additional_cpus __initdata = -1; - -/* - * cpu_possible_map should be static, it cannot change as cpu's - * are onlined, or offlined. The reason is per-cpu data-structures - * are allocated by some modules at init time, and dont expect to - * do this dynamically on cpu arrival/departure. - * cpu_present_map on the other hand can change dynamically. - * In case when cpu_hotplug is not compiled, then we resort to current - * behaviour, which is cpu_possible == cpu_present. - * - Ashok Raj - * - * Three ways to find out the number of additional hotplug CPUs: - * - If the BIOS specified disabled CPUs in ACPI/mptables use that. - * - The user can overwrite it with additional_cpus=NUM - * - Otherwise don't reserve additional CPUs. - * We do this because additional CPUs waste a lot of memory. - * -AK - */ -__init void prefill_possible_map(void) -{ - int i; - int possible; - - if (additional_cpus == -1) { - if (disabled_cpus > 0) - additional_cpus = disabled_cpus; - else - additional_cpus = 0; - } - possible = num_processors + additional_cpus; - if (possible > NR_CPUS) - possible = NR_CPUS; - - printk(KERN_INFO "SMP: Allowing %d CPUs, %d hotplug CPUs\n", - possible, - max_t(int, possible - num_processors, 0)); - - for (i = 0; i < possible; i++) - cpu_set(i, cpu_possible_map); -} -#endif - -/* - * Various sanity checks. - */ -static int __init smp_sanity_check(unsigned max_cpus) -{ - if (!physid_isset(hard_smp_processor_id(), phys_cpu_present_map)) { - printk("weird, boot CPU (#%d) not listed by the BIOS.\n", - hard_smp_processor_id()); - physid_set(hard_smp_processor_id(), phys_cpu_present_map); - } - - /* - * If we couldn't find an SMP configuration at boot time, - * get out of here now! - */ - if (!smp_found_config) { - printk(KERN_NOTICE "SMP motherboard not detected.\n"); - disable_smp(); - if (APIC_init_uniprocessor()) - printk(KERN_NOTICE "Local APIC not detected." - " Using dummy APIC emulation.\n"); - return -1; - } - - /* - * Should not be necessary because the MP table should list the boot - * CPU too, but we do it for the sake of robustness anyway. - */ - if (!physid_isset(boot_cpu_id, phys_cpu_present_map)) { - printk(KERN_NOTICE "weird, boot CPU (#%d) not listed by the BIOS.\n", - boot_cpu_id); - physid_set(hard_smp_processor_id(), phys_cpu_present_map); - } - - /* - * If we couldn't find a local APIC, then get out of here now! - */ - if (!cpu_has_apic) { - printk(KERN_ERR "BIOS bug, local APIC #%d not detected!...\n", - boot_cpu_id); - printk(KERN_ERR "... forcing use of dummy APIC emulation. (tell your hw vendor)\n"); - nr_ioapics = 0; - return -1; - } - - /* - * If SMP should be disabled, then really disable it! - */ - if (!max_cpus) { - printk(KERN_INFO "SMP mode deactivated, forcing use of dummy APIC emulation.\n"); - nr_ioapics = 0; - return -1; - } - - return 0; -} - -static void __init smp_cpu_index_default(void) -{ - int i; - struct cpuinfo_x86 *c; - - for_each_cpu_mask(i, cpu_possible_map) { - c = &cpu_data(i); - /* mark all to hotplug */ - c->cpu_index = NR_CPUS; - } -} - -/* - * Prepare for SMP bootup. The MP table or ACPI has been read - * earlier. Just do some sanity checking here and enable APIC mode. - */ -void __init smp_prepare_cpus(unsigned int max_cpus) -{ - nmi_watchdog_default(); - smp_cpu_index_default(); - current_cpu_data = boot_cpu_data; - current_thread_info()->cpu = 0; /* needed? */ - set_cpu_sibling_map(0); - - if (smp_sanity_check(max_cpus) < 0) { - printk(KERN_INFO "SMP disabled\n"); - disable_smp(); - return; - } - - - /* - * Switch from PIC to APIC mode. - */ - setup_local_APIC(); - - /* - * Enable IO APIC before setting up error vector - */ - if (!skip_ioapic_setup && nr_ioapics) - enable_IO_APIC(); - end_local_APIC_setup(); - - if (GET_APIC_ID(apic_read(APIC_ID)) != boot_cpu_id) { - panic("Boot APIC ID in local APIC unexpected (%d vs %d)", - GET_APIC_ID(apic_read(APIC_ID)), boot_cpu_id); - /* Or can we switch back to PIC here? */ - } - - /* - * Now start the IO-APICs - */ - if (!skip_ioapic_setup && nr_ioapics) - setup_IO_APIC(); - else - nr_ioapics = 0; - - /* - * Set up local APIC timer on boot CPU. - */ - - setup_boot_clock(); -} - -/* - * Early setup to make printk work. - */ -void __init smp_prepare_boot_cpu(void) -{ - int me = smp_processor_id(); - /* already set me in cpu_online_map in boot_cpu_init() */ - cpu_set(me, cpu_callout_map); - per_cpu(cpu_state, me) = CPU_ONLINE; -} - -/* - * Entry point to boot a CPU. - */ -int __cpuinit __cpu_up(unsigned int cpu) -{ - int apicid = cpu_present_to_apicid(cpu); - unsigned long flags; - int err; - - WARN_ON(irqs_disabled()); - - Dprintk("++++++++++++++++++++=_---CPU UP %u\n", cpu); - - if (apicid == BAD_APICID || apicid == boot_cpu_id || - !physid_isset(apicid, phys_cpu_present_map)) { - printk("__cpu_up: bad cpu %d\n", cpu); - return -EINVAL; - } - - /* - * Already booted CPU? - */ - if (cpu_isset(cpu, cpu_callin_map)) { - Dprintk("do_boot_cpu %d Already started\n", cpu); - return -ENOSYS; - } - - /* - * Save current MTRR state in case it was changed since early boot - * (e.g. by the ACPI SMI) to initialize new CPUs with MTRRs in sync: - */ - mtrr_save_state(); - - per_cpu(cpu_state, cpu) = CPU_UP_PREPARE; - /* Boot it! */ - err = do_boot_cpu(cpu, apicid); - if (err < 0) { - Dprintk("do_boot_cpu failed %d\n", err); - return err; - } - - /* Unleash the CPU! */ - Dprintk("waiting for cpu %d\n", cpu); - - /* - * Make sure and check TSC sync: - */ - local_irq_save(flags); - check_tsc_sync_source(cpu); - local_irq_restore(flags); - - while (!cpu_isset(cpu, cpu_online_map)) - cpu_relax(); - err = 0; - - return err; -} - -/* - * Finish the SMP boot. - */ -void __init smp_cpus_done(unsigned int max_cpus) -{ - smp_cleanup_boot(); - setup_ioapic_dest(); - check_nmi_watchdog(); -} - -#ifdef CONFIG_HOTPLUG_CPU - -static void remove_siblinginfo(int cpu) -{ - int sibling; - struct cpuinfo_x86 *c = &cpu_data(cpu); - - for_each_cpu_mask(sibling, per_cpu(cpu_core_map, cpu)) { - cpu_clear(cpu, per_cpu(cpu_core_map, sibling)); - /* - * last thread sibling in this cpu core going down - */ - if (cpus_weight(per_cpu(cpu_sibling_map, cpu)) == 1) - cpu_data(sibling).booted_cores--; - } - - for_each_cpu_mask(sibling, per_cpu(cpu_sibling_map, cpu)) - cpu_clear(cpu, per_cpu(cpu_sibling_map, sibling)); - cpus_clear(per_cpu(cpu_sibling_map, cpu)); - cpus_clear(per_cpu(cpu_core_map, cpu)); - c->phys_proc_id = 0; - c->cpu_core_id = 0; - cpu_clear(cpu, cpu_sibling_setup_map); -} - -static void __ref remove_cpu_from_maps(void) -{ - int cpu = smp_processor_id(); - - cpu_clear(cpu, cpu_callout_map); - cpu_clear(cpu, cpu_callin_map); - clear_bit(cpu, (unsigned long *)&cpu_initialized); /* was set by cpu_init() */ - clear_node_cpumask(cpu); -} - -int __cpu_disable(void) -{ - int cpu = smp_processor_id(); - - /* - * Perhaps use cpufreq to drop frequency, but that could go - * into generic code. - * - * We won't take down the boot processor on i386 due to some - * interrupts only being able to be serviced by the BSP. - * Especially so if we're not using an IOAPIC -zwane - */ - if (cpu == 0) - return -EBUSY; - - if (nmi_watchdog == NMI_LOCAL_APIC) - stop_apic_nmi_watchdog(NULL); - clear_local_APIC(); - - /* - * HACK: - * Allow any queued timer interrupts to get serviced - * This is only a temporary solution until we cleanup - * fixup_irqs as we do for IA64. - */ - local_irq_enable(); - mdelay(1); - - local_irq_disable(); - remove_siblinginfo(cpu); - - spin_lock(&vector_lock); - /* It's now safe to remove this processor from the online map */ - cpu_clear(cpu, cpu_online_map); - spin_unlock(&vector_lock); - remove_cpu_from_maps(); - fixup_irqs(cpu_online_map); - return 0; -} - -void __cpu_die(unsigned int cpu) -{ - /* We don't do anything here: idle task is faking death itself. */ - unsigned int i; - - for (i = 0; i < 10; i++) { - /* They ack this in play_dead by setting CPU_DEAD */ - if (per_cpu(cpu_state, cpu) == CPU_DEAD) { - printk ("CPU %d is now offline\n", cpu); - if (1 == num_online_cpus()) - alternatives_smp_switch(0); - return; - } - msleep(100); - } - printk(KERN_ERR "CPU %u didn't die...\n", cpu); -} - -static __init int setup_additional_cpus(char *s) -{ - return s && get_option(&s, &additional_cpus) ? 0 : -EINVAL; -} -early_param("additional_cpus", setup_additional_cpus); - -#else /* ... !CONFIG_HOTPLUG_CPU */ - -int __cpu_disable(void) -{ - return -ENOSYS; -} - -void __cpu_die(unsigned int cpu) -{ - /* We said "no" in __cpu_disable */ - BUG(); -} -#endif /* CONFIG_HOTPLUG_CPU */ diff --git a/arch/x86/kernel/smpcommon.c b/arch/x86/kernel/smpcommon.c new file mode 100644 index 000000000000..3449064d141a --- /dev/null +++ b/arch/x86/kernel/smpcommon.c @@ -0,0 +1,83 @@ +/* + * SMP stuff which is common to all sub-architectures. + */ +#include <linux/module.h> +#include <asm/smp.h> + +#ifdef CONFIG_X86_32 +DEFINE_PER_CPU(unsigned long, this_cpu_off); +EXPORT_PER_CPU_SYMBOL(this_cpu_off); + +/* Initialize the CPU's GDT. This is either the boot CPU doing itself + (still using the master per-cpu area), or a CPU doing it for a + secondary which will soon come up. */ +__cpuinit void init_gdt(int cpu) +{ + struct desc_struct *gdt = get_cpu_gdt_table(cpu); + + pack_descriptor(&gdt[GDT_ENTRY_PERCPU], + __per_cpu_offset[cpu], 0xFFFFF, + 0x2 | DESCTYPE_S, 0x8); + + gdt[GDT_ENTRY_PERCPU].s = 1; + + per_cpu(this_cpu_off, cpu) = __per_cpu_offset[cpu]; + per_cpu(cpu_number, cpu) = cpu; +} +#endif + +/** + * smp_call_function(): Run a function on all other CPUs. + * @func: The function to run. This must be fast and non-blocking. + * @info: An arbitrary pointer to pass to the function. + * @nonatomic: Unused. + * @wait: If true, wait (atomically) until function has completed on other CPUs. + * + * Returns 0 on success, else a negative status code. + * + * If @wait is true, then returns once @func has returned; otherwise + * it returns just before the target cpu calls @func. + * + * You must not call this function with disabled interrupts or from a + * hardware interrupt handler or from a bottom half handler. + */ +int smp_call_function(void (*func) (void *info), void *info, int nonatomic, + int wait) +{ + return smp_call_function_mask(cpu_online_map, func, info, wait); +} +EXPORT_SYMBOL(smp_call_function); + +/** + * smp_call_function_single - Run a function on a specific CPU + * @cpu: The target CPU. Cannot be the calling CPU. + * @func: The function to run. This must be fast and non-blocking. + * @info: An arbitrary pointer to pass to the function. + * @nonatomic: Unused. + * @wait: If true, wait until function has completed on other CPUs. + * + * Returns 0 on success, else a negative status code. + * + * If @wait is true, then returns once @func has returned; otherwise + * it returns just before the target cpu calls @func. + */ +int smp_call_function_single(int cpu, void (*func) (void *info), void *info, + int nonatomic, int wait) +{ + /* prevent preemption and reschedule on another processor */ + int ret; + int me = get_cpu(); + if (cpu == me) { + local_irq_disable(); + func(info); + local_irq_enable(); + put_cpu(); + return 0; + } + + ret = smp_call_function_mask(cpumask_of_cpu(cpu), func, info, wait); + + put_cpu(); + return ret; +} +EXPORT_SYMBOL(smp_call_function_single); diff --git a/arch/x86/kernel/smpcommon_32.c b/arch/x86/kernel/smpcommon_32.c index 8bc38af29aef..8b137891791f 100644 --- a/arch/x86/kernel/smpcommon_32.c +++ b/arch/x86/kernel/smpcommon_32.c @@ -1,82 +1 @@ -/* - * SMP stuff which is common to all sub-architectures. - */ -#include <linux/module.h> -#include <asm/smp.h> -DEFINE_PER_CPU(unsigned long, this_cpu_off); -EXPORT_PER_CPU_SYMBOL(this_cpu_off); - -/* Initialize the CPU's GDT. This is either the boot CPU doing itself - (still using the master per-cpu area), or a CPU doing it for a - secondary which will soon come up. */ -__cpuinit void init_gdt(int cpu) -{ - struct desc_struct *gdt = get_cpu_gdt_table(cpu); - - pack_descriptor(&gdt[GDT_ENTRY_PERCPU], - __per_cpu_offset[cpu], 0xFFFFF, - 0x2 | DESCTYPE_S, 0x8); - - gdt[GDT_ENTRY_PERCPU].s = 1; - - per_cpu(this_cpu_off, cpu) = __per_cpu_offset[cpu]; - per_cpu(cpu_number, cpu) = cpu; -} - - -/** - * smp_call_function(): Run a function on all other CPUs. - * @func: The function to run. This must be fast and non-blocking. - * @info: An arbitrary pointer to pass to the function. - * @nonatomic: Unused. - * @wait: If true, wait (atomically) until function has completed on other CPUs. - * - * Returns 0 on success, else a negative status code. - * - * If @wait is true, then returns once @func has returned; otherwise - * it returns just before the target cpu calls @func. - * - * You must not call this function with disabled interrupts or from a - * hardware interrupt handler or from a bottom half handler. - */ -int smp_call_function(void (*func) (void *info), void *info, int nonatomic, - int wait) -{ - return smp_call_function_mask(cpu_online_map, func, info, wait); -} -EXPORT_SYMBOL(smp_call_function); - -/** - * smp_call_function_single - Run a function on a specific CPU - * @cpu: The target CPU. Cannot be the calling CPU. - * @func: The function to run. This must be fast and non-blocking. - * @info: An arbitrary pointer to pass to the function. - * @nonatomic: Unused. - * @wait: If true, wait until function has completed on other CPUs. - * - * Returns 0 on success, else a negative status code. - * - * If @wait is true, then returns once @func has returned; otherwise - * it returns just before the target cpu calls @func. - */ -int smp_call_function_single(int cpu, void (*func) (void *info), void *info, - int nonatomic, int wait) -{ - /* prevent preemption and reschedule on another processor */ - int ret; - int me = get_cpu(); - if (cpu == me) { - local_irq_disable(); - func(info); - local_irq_enable(); - put_cpu(); - return 0; - } - - ret = smp_call_function_mask(cpumask_of_cpu(cpu), func, info, wait); - - put_cpu(); - return ret; -} -EXPORT_SYMBOL(smp_call_function_single); diff --git a/arch/x86/kernel/srat_32.c b/arch/x86/kernel/srat_32.c index b72e61359c36..70e4a374b4e8 100644 --- a/arch/x86/kernel/srat_32.c +++ b/arch/x86/kernel/srat_32.c @@ -277,14 +277,14 @@ int __init get_memcfg_from_srat(void) rsdp_address = acpi_os_get_root_pointer(); if (!rsdp_address) { printk("%s: System description tables not found\n", - __FUNCTION__); + __func__); goto out_err; } - printk("%s: assigning address to rsdp\n", __FUNCTION__); + printk("%s: assigning address to rsdp\n", __func__); rsdp = (struct acpi_table_rsdp *)(u32)rsdp_address; if (!rsdp) { - printk("%s: Didn't find ACPI root!\n", __FUNCTION__); + printk("%s: Didn't find ACPI root!\n", __func__); goto out_err; } @@ -292,7 +292,7 @@ int __init get_memcfg_from_srat(void) rsdp->oem_id); if (strncmp(rsdp->signature, ACPI_SIG_RSDP,strlen(ACPI_SIG_RSDP))) { - printk(KERN_WARNING "%s: RSDP table signature incorrect\n", __FUNCTION__); + printk(KERN_WARNING "%s: RSDP table signature incorrect\n", __func__); goto out_err; } @@ -302,7 +302,7 @@ int __init get_memcfg_from_srat(void) if (!rsdt) { printk(KERN_WARNING "%s: ACPI: Invalid root system description tables (RSDT)\n", - __FUNCTION__); + __func__); goto out_err; } diff --git a/arch/x86/kernel/step.c b/arch/x86/kernel/step.c index 071ff4798236..92c20fee6781 100644 --- a/arch/x86/kernel/step.c +++ b/arch/x86/kernel/step.c @@ -148,7 +148,7 @@ static void write_debugctlmsr(struct task_struct *child, unsigned long val) if (child != current) return; - wrmsrl(MSR_IA32_DEBUGCTLMSR, val); + update_debugctlmsr(val); } /* diff --git a/arch/x86/kernel/summit_32.c b/arch/x86/kernel/summit_32.c index 72f463401592..6878a9c2df5d 100644 --- a/arch/x86/kernel/summit_32.c +++ b/arch/x86/kernel/summit_32.c @@ -35,43 +35,47 @@ static struct rio_table_hdr *rio_table_hdr __initdata; static struct scal_detail *scal_devs[MAX_NUMNODES] __initdata; static struct rio_detail *rio_devs[MAX_NUMNODES*4] __initdata; +static int mp_bus_id_to_node[MAX_MP_BUSSES] __initdata; + static int __init setup_pci_node_map_for_wpeg(int wpeg_num, int last_bus) { int twister = 0, node = 0; int i, bus, num_buses; - for(i = 0; i < rio_table_hdr->num_rio_dev; i++){ - if (rio_devs[i]->node_id == rio_devs[wpeg_num]->owner_id){ + for (i = 0; i < rio_table_hdr->num_rio_dev; i++) { + if (rio_devs[i]->node_id == rio_devs[wpeg_num]->owner_id) { twister = rio_devs[i]->owner_id; break; } } - if (i == rio_table_hdr->num_rio_dev){ - printk(KERN_ERR "%s: Couldn't find owner Cyclone for Winnipeg!\n", __FUNCTION__); + if (i == rio_table_hdr->num_rio_dev) { + printk(KERN_ERR "%s: Couldn't find owner Cyclone for Winnipeg!\n", __func__); return last_bus; } - for(i = 0; i < rio_table_hdr->num_scal_dev; i++){ - if (scal_devs[i]->node_id == twister){ + for (i = 0; i < rio_table_hdr->num_scal_dev; i++) { + if (scal_devs[i]->node_id == twister) { node = scal_devs[i]->node_id; break; } } - if (i == rio_table_hdr->num_scal_dev){ - printk(KERN_ERR "%s: Couldn't find owner Twister for Cyclone!\n", __FUNCTION__); + if (i == rio_table_hdr->num_scal_dev) { + printk(KERN_ERR "%s: Couldn't find owner Twister for Cyclone!\n", __func__); return last_bus; } - switch (rio_devs[wpeg_num]->type){ + switch (rio_devs[wpeg_num]->type) { case CompatWPEG: - /* The Compatibility Winnipeg controls the 2 legacy buses, + /* + * The Compatibility Winnipeg controls the 2 legacy buses, * the 66MHz PCI bus [2 slots] and the 2 "extra" buses in case * a PCI-PCI bridge card is used in either slot: total 5 buses. */ num_buses = 5; break; case AltWPEG: - /* The Alternate Winnipeg controls the 2 133MHz buses [1 slot + /* + * The Alternate Winnipeg controls the 2 133MHz buses [1 slot * each], their 2 "extra" buses, the 100MHz bus [2 slots] and * the "extra" buses for each of those slots: total 7 buses. */ @@ -79,17 +83,18 @@ static int __init setup_pci_node_map_for_wpeg(int wpeg_num, int last_bus) break; case LookOutAWPEG: case LookOutBWPEG: - /* A Lookout Winnipeg controls 3 100MHz buses [2 slots each] + /* + * A Lookout Winnipeg controls 3 100MHz buses [2 slots each] * & the "extra" buses for each of those slots: total 9 buses. */ num_buses = 9; break; default: - printk(KERN_INFO "%s: Unsupported Winnipeg type!\n", __FUNCTION__); + printk(KERN_INFO "%s: Unsupported Winnipeg type!\n", __func__); return last_bus; } - for(bus = last_bus; bus < last_bus + num_buses; bus++) + for (bus = last_bus; bus < last_bus + num_buses; bus++) mp_bus_id_to_node[bus] = node; return bus; } @@ -99,14 +104,14 @@ static int __init build_detail_arrays(void) unsigned long ptr; int i, scal_detail_size, rio_detail_size; - if (rio_table_hdr->num_scal_dev > MAX_NUMNODES){ - printk(KERN_WARNING "%s: MAX_NUMNODES too low! Defined as %d, but system has %d nodes.\n", __FUNCTION__, MAX_NUMNODES, rio_table_hdr->num_scal_dev); + if (rio_table_hdr->num_scal_dev > MAX_NUMNODES) { + printk(KERN_WARNING "%s: MAX_NUMNODES too low! Defined as %d, but system has %d nodes.\n", __func__, MAX_NUMNODES, rio_table_hdr->num_scal_dev); return 0; } - switch (rio_table_hdr->version){ + switch (rio_table_hdr->version) { default: - printk(KERN_WARNING "%s: Invalid Rio Grande Table Version: %d\n", __FUNCTION__, rio_table_hdr->version); + printk(KERN_WARNING "%s: Invalid Rio Grande Table Version: %d\n", __func__, rio_table_hdr->version); return 0; case 2: scal_detail_size = 11; @@ -119,10 +124,10 @@ static int __init build_detail_arrays(void) } ptr = (unsigned long)rio_table_hdr + 3; - for(i = 0; i < rio_table_hdr->num_scal_dev; i++, ptr += scal_detail_size) + for (i = 0; i < rio_table_hdr->num_scal_dev; i++, ptr += scal_detail_size) scal_devs[i] = (struct scal_detail *)ptr; - for(i = 0; i < rio_table_hdr->num_rio_dev; i++, ptr += rio_detail_size) + for (i = 0; i < rio_table_hdr->num_rio_dev; i++, ptr += rio_detail_size) rio_devs[i] = (struct rio_detail *)ptr; return 1; @@ -140,9 +145,9 @@ void __init setup_summit(void) rio_table_hdr = NULL; offset = 0x180; - while (offset){ + while (offset) { /* The block id is stored in the 2nd word */ - if (*((unsigned short *)(ptr + offset + 2)) == 0x4752){ + if (*((unsigned short *)(ptr + offset + 2)) == 0x4752) { /* set the pointer past the offset & block id */ rio_table_hdr = (struct rio_table_hdr *)(ptr + offset + 4); break; @@ -150,8 +155,8 @@ void __init setup_summit(void) /* The next offset is stored in the 1st word. 0 means no more */ offset = *((unsigned short *)(ptr + offset)); } - if (!rio_table_hdr){ - printk(KERN_ERR "%s: Unable to locate Rio Grande Table in EBDA - bailing!\n", __FUNCTION__); + if (!rio_table_hdr) { + printk(KERN_ERR "%s: Unable to locate Rio Grande Table in EBDA - bailing!\n", __func__); return; } @@ -161,8 +166,8 @@ void __init setup_summit(void) /* The first Winnipeg we're looking for has an index of 0 */ next_wpeg = 0; do { - for(i = 0; i < rio_table_hdr->num_rio_dev; i++){ - if (is_WPEG(rio_devs[i]) && rio_devs[i]->WP_index == next_wpeg){ + for (i = 0; i < rio_table_hdr->num_rio_dev; i++) { + if (is_WPEG(rio_devs[i]) && rio_devs[i]->WP_index == next_wpeg) { /* It's the Winnipeg we're looking for! */ next_bus = setup_pci_node_map_for_wpeg(i, next_bus); next_wpeg++; diff --git a/arch/x86/kernel/syscall_64.c b/arch/x86/kernel/syscall_64.c index 9d498c2f8eea..170d43c17487 100644 --- a/arch/x86/kernel/syscall_64.c +++ b/arch/x86/kernel/syscall_64.c @@ -1,4 +1,4 @@ -/* System call table for x86-64. */ +/* System call table for x86-64. */ #include <linux/linkage.h> #include <linux/sys.h> @@ -7,20 +7,23 @@ #define __NO_STUBS -#define __SYSCALL(nr, sym) extern asmlinkage void sym(void) ; +#define __SYSCALL(nr, sym) extern asmlinkage void sym(void) ; #undef _ASM_X86_64_UNISTD_H_ #include <asm/unistd_64.h> #undef __SYSCALL -#define __SYSCALL(nr, sym) [ nr ] = sym, +#define __SYSCALL(nr, sym) [nr] = sym, #undef _ASM_X86_64_UNISTD_H_ -typedef void (*sys_call_ptr_t)(void); +typedef void (*sys_call_ptr_t)(void); extern void sys_ni_syscall(void); const sys_call_ptr_t sys_call_table[__NR_syscall_max+1] = { - /* Smells like a like a compiler bug -- it doesn't work when the & below is removed. */ + /* + *Smells like a like a compiler bug -- it doesn't work + *when the & below is removed. + */ [0 ... __NR_syscall_max] = &sys_ni_syscall, #include <asm/unistd_64.h> }; diff --git a/arch/x86/kernel/test_nx.c b/arch/x86/kernel/test_nx.c index 10b8a6f69f84..787a5e499dd1 100644 --- a/arch/x86/kernel/test_nx.c +++ b/arch/x86/kernel/test_nx.c @@ -11,6 +11,8 @@ */ #include <linux/module.h> #include <linux/sort.h> +#include <linux/slab.h> + #include <asm/uaccess.h> #include <asm/asm.h> diff --git a/arch/x86/kernel/tlb_32.c b/arch/x86/kernel/tlb_32.c new file mode 100644 index 000000000000..9bb2363851af --- /dev/null +++ b/arch/x86/kernel/tlb_32.c @@ -0,0 +1,243 @@ +#include <linux/spinlock.h> +#include <linux/cpu.h> +#include <linux/interrupt.h> + +#include <asm/tlbflush.h> + +DEFINE_PER_CPU(struct tlb_state, cpu_tlbstate) + ____cacheline_aligned = { &init_mm, 0, }; + +/* must come after the send_IPI functions above for inlining */ +#include <mach_ipi.h> + +/* + * Smarter SMP flushing macros. + * c/o Linus Torvalds. + * + * These mean you can really definitely utterly forget about + * writing to user space from interrupts. (Its not allowed anyway). + * + * Optimizations Manfred Spraul <manfred@colorfullife.com> + */ + +static cpumask_t flush_cpumask; +static struct mm_struct *flush_mm; +static unsigned long flush_va; +static DEFINE_SPINLOCK(tlbstate_lock); + +/* + * We cannot call mmdrop() because we are in interrupt context, + * instead update mm->cpu_vm_mask. + * + * We need to reload %cr3 since the page tables may be going + * away from under us.. + */ +void leave_mm(int cpu) +{ + if (per_cpu(cpu_tlbstate, cpu).state == TLBSTATE_OK) + BUG(); + cpu_clear(cpu, per_cpu(cpu_tlbstate, cpu).active_mm->cpu_vm_mask); + load_cr3(swapper_pg_dir); +} +EXPORT_SYMBOL_GPL(leave_mm); + +/* + * + * The flush IPI assumes that a thread switch happens in this order: + * [cpu0: the cpu that switches] + * 1) switch_mm() either 1a) or 1b) + * 1a) thread switch to a different mm + * 1a1) cpu_clear(cpu, old_mm->cpu_vm_mask); + * Stop ipi delivery for the old mm. This is not synchronized with + * the other cpus, but smp_invalidate_interrupt ignore flush ipis + * for the wrong mm, and in the worst case we perform a superfluous + * tlb flush. + * 1a2) set cpu_tlbstate to TLBSTATE_OK + * Now the smp_invalidate_interrupt won't call leave_mm if cpu0 + * was in lazy tlb mode. + * 1a3) update cpu_tlbstate[].active_mm + * Now cpu0 accepts tlb flushes for the new mm. + * 1a4) cpu_set(cpu, new_mm->cpu_vm_mask); + * Now the other cpus will send tlb flush ipis. + * 1a4) change cr3. + * 1b) thread switch without mm change + * cpu_tlbstate[].active_mm is correct, cpu0 already handles + * flush ipis. + * 1b1) set cpu_tlbstate to TLBSTATE_OK + * 1b2) test_and_set the cpu bit in cpu_vm_mask. + * Atomically set the bit [other cpus will start sending flush ipis], + * and test the bit. + * 1b3) if the bit was 0: leave_mm was called, flush the tlb. + * 2) switch %%esp, ie current + * + * The interrupt must handle 2 special cases: + * - cr3 is changed before %%esp, ie. it cannot use current->{active_,}mm. + * - the cpu performs speculative tlb reads, i.e. even if the cpu only + * runs in kernel space, the cpu could load tlb entries for user space + * pages. + * + * The good news is that cpu_tlbstate is local to each cpu, no + * write/read ordering problems. + */ + +/* + * TLB flush IPI: + * + * 1) Flush the tlb entries if the cpu uses the mm that's being flushed. + * 2) Leave the mm if we are in the lazy tlb mode. + */ + +void smp_invalidate_interrupt(struct pt_regs *regs) +{ + unsigned long cpu; + + cpu = get_cpu(); + + if (!cpu_isset(cpu, flush_cpumask)) + goto out; + /* + * This was a BUG() but until someone can quote me the + * line from the intel manual that guarantees an IPI to + * multiple CPUs is retried _only_ on the erroring CPUs + * its staying as a return + * + * BUG(); + */ + + if (flush_mm == per_cpu(cpu_tlbstate, cpu).active_mm) { + if (per_cpu(cpu_tlbstate, cpu).state == TLBSTATE_OK) { + if (flush_va == TLB_FLUSH_ALL) + local_flush_tlb(); + else + __flush_tlb_one(flush_va); + } else + leave_mm(cpu); + } + ack_APIC_irq(); + smp_mb__before_clear_bit(); + cpu_clear(cpu, flush_cpumask); + smp_mb__after_clear_bit(); +out: + put_cpu_no_resched(); + __get_cpu_var(irq_stat).irq_tlb_count++; +} + +void native_flush_tlb_others(const cpumask_t *cpumaskp, struct mm_struct *mm, + unsigned long va) +{ + cpumask_t cpumask = *cpumaskp; + + /* + * A couple of (to be removed) sanity checks: + * + * - current CPU must not be in mask + * - mask must exist :) + */ + BUG_ON(cpus_empty(cpumask)); + BUG_ON(cpu_isset(smp_processor_id(), cpumask)); + BUG_ON(!mm); + +#ifdef CONFIG_HOTPLUG_CPU + /* If a CPU which we ran on has gone down, OK. */ + cpus_and(cpumask, cpumask, cpu_online_map); + if (unlikely(cpus_empty(cpumask))) + return; +#endif + + /* + * i'm not happy about this global shared spinlock in the + * MM hot path, but we'll see how contended it is. + * AK: x86-64 has a faster method that could be ported. + */ + spin_lock(&tlbstate_lock); + + flush_mm = mm; + flush_va = va; + cpus_or(flush_cpumask, cpumask, flush_cpumask); + /* + * We have to send the IPI only to + * CPUs affected. + */ + send_IPI_mask(cpumask, INVALIDATE_TLB_VECTOR); + + while (!cpus_empty(flush_cpumask)) + /* nothing. lockup detection does not belong here */ + cpu_relax(); + + flush_mm = NULL; + flush_va = 0; + spin_unlock(&tlbstate_lock); +} + +void flush_tlb_current_task(void) +{ + struct mm_struct *mm = current->mm; + cpumask_t cpu_mask; + + preempt_disable(); + cpu_mask = mm->cpu_vm_mask; + cpu_clear(smp_processor_id(), cpu_mask); + + local_flush_tlb(); + if (!cpus_empty(cpu_mask)) + flush_tlb_others(cpu_mask, mm, TLB_FLUSH_ALL); + preempt_enable(); +} + +void flush_tlb_mm(struct mm_struct *mm) +{ + cpumask_t cpu_mask; + + preempt_disable(); + cpu_mask = mm->cpu_vm_mask; + cpu_clear(smp_processor_id(), cpu_mask); + + if (current->active_mm == mm) { + if (current->mm) + local_flush_tlb(); + else + leave_mm(smp_processor_id()); + } + if (!cpus_empty(cpu_mask)) + flush_tlb_others(cpu_mask, mm, TLB_FLUSH_ALL); + + preempt_enable(); +} + +void flush_tlb_page(struct vm_area_struct *vma, unsigned long va) +{ + struct mm_struct *mm = vma->vm_mm; + cpumask_t cpu_mask; + + preempt_disable(); + cpu_mask = mm->cpu_vm_mask; + cpu_clear(smp_processor_id(), cpu_mask); + + if (current->active_mm == mm) { + if (current->mm) + __flush_tlb_one(va); + else + leave_mm(smp_processor_id()); + } + + if (!cpus_empty(cpu_mask)) + flush_tlb_others(cpu_mask, mm, va); + + preempt_enable(); +} +EXPORT_SYMBOL(flush_tlb_page); + +static void do_flush_tlb_all(void *info) +{ + unsigned long cpu = smp_processor_id(); + + __flush_tlb_all(); + if (per_cpu(cpu_tlbstate, cpu).state == TLBSTATE_LAZY) + leave_mm(cpu); +} + +void flush_tlb_all(void) +{ + on_each_cpu(do_flush_tlb_all, NULL, 1, 1); +} + diff --git a/arch/x86/kernel/smp_64.c b/arch/x86/kernel/tlb_64.c index 2fd74b06db67..1558e513757e 100644 --- a/arch/x86/kernel/smp_64.c +++ b/arch/x86/kernel/tlb_64.c @@ -1,14 +1,3 @@ -/* - * Intel SMP support routines. - * - * (c) 1995 Alan Cox, Building #3 <alan@redhat.com> - * (c) 1998-99, 2000 Ingo Molnar <mingo@redhat.com> - * (c) 2002,2003 Andi Kleen, SuSE Labs. - * - * This code is released under the GNU General Public License version 2 or - * later. - */ - #include <linux/init.h> #include <linux/mm.h> @@ -22,12 +11,12 @@ #include <asm/mtrr.h> #include <asm/pgalloc.h> #include <asm/tlbflush.h> -#include <asm/mach_apic.h> #include <asm/mmu_context.h> #include <asm/proto.h> #include <asm/apicdef.h> #include <asm/idle.h> +#include <mach_ipi.h> /* * Smarter SMP flushing macros. * c/o Linus Torvalds. @@ -228,7 +217,7 @@ void flush_tlb_current_task(void) preempt_enable(); } -void flush_tlb_mm (struct mm_struct * mm) +void flush_tlb_mm(struct mm_struct *mm) { cpumask_t cpu_mask; @@ -248,7 +237,7 @@ void flush_tlb_mm (struct mm_struct * mm) preempt_enable(); } -void flush_tlb_page(struct vm_area_struct * vma, unsigned long va) +void flush_tlb_page(struct vm_area_struct *vma, unsigned long va) { struct mm_struct *mm = vma->vm_mm; cpumask_t cpu_mask; @@ -258,7 +247,7 @@ void flush_tlb_page(struct vm_area_struct * vma, unsigned long va) cpu_clear(smp_processor_id(), cpu_mask); if (current->active_mm == mm) { - if(current->mm) + if (current->mm) __flush_tlb_one(va); else leave_mm(smp_processor_id()); @@ -270,7 +259,7 @@ void flush_tlb_page(struct vm_area_struct * vma, unsigned long va) preempt_enable(); } -static void do_flush_tlb_all(void* info) +static void do_flush_tlb_all(void *info) { unsigned long cpu = smp_processor_id(); @@ -283,248 +272,3 @@ void flush_tlb_all(void) { on_each_cpu(do_flush_tlb_all, NULL, 1, 1); } - -/* - * this function sends a 'reschedule' IPI to another CPU. - * it goes straight through and wastes no time serializing - * anything. Worst case is that we lose a reschedule ... - */ - -void smp_send_reschedule(int cpu) -{ - send_IPI_mask(cpumask_of_cpu(cpu), RESCHEDULE_VECTOR); -} - -/* - * Structure and data for smp_call_function(). This is designed to minimise - * static memory requirements. It also looks cleaner. - */ -static DEFINE_SPINLOCK(call_lock); - -struct call_data_struct { - void (*func) (void *info); - void *info; - atomic_t started; - atomic_t finished; - int wait; -}; - -static struct call_data_struct * call_data; - -void lock_ipi_call_lock(void) -{ - spin_lock_irq(&call_lock); -} - -void unlock_ipi_call_lock(void) -{ - spin_unlock_irq(&call_lock); -} - -/* - * this function sends a 'generic call function' IPI to all other CPU - * of the system defined in the mask. - */ -static int __smp_call_function_mask(cpumask_t mask, - void (*func)(void *), void *info, - int wait) -{ - struct call_data_struct data; - cpumask_t allbutself; - int cpus; - - allbutself = cpu_online_map; - cpu_clear(smp_processor_id(), allbutself); - - cpus_and(mask, mask, allbutself); - cpus = cpus_weight(mask); - - if (!cpus) - return 0; - - data.func = func; - data.info = info; - atomic_set(&data.started, 0); - data.wait = wait; - if (wait) - atomic_set(&data.finished, 0); - - call_data = &data; - wmb(); - - /* Send a message to other CPUs */ - if (cpus_equal(mask, allbutself)) - send_IPI_allbutself(CALL_FUNCTION_VECTOR); - else - send_IPI_mask(mask, CALL_FUNCTION_VECTOR); - - /* Wait for response */ - while (atomic_read(&data.started) != cpus) - cpu_relax(); - - if (!wait) - return 0; - - while (atomic_read(&data.finished) != cpus) - cpu_relax(); - - return 0; -} -/** - * smp_call_function_mask(): Run a function on a set of other CPUs. - * @mask: The set of cpus to run on. Must not include the current cpu. - * @func: The function to run. This must be fast and non-blocking. - * @info: An arbitrary pointer to pass to the function. - * @wait: If true, wait (atomically) until function has completed on other CPUs. - * - * Returns 0 on success, else a negative status code. - * - * If @wait is true, then returns once @func has returned; otherwise - * it returns just before the target cpu calls @func. - * - * You must not call this function with disabled interrupts or from a - * hardware interrupt handler or from a bottom half handler. - */ -int smp_call_function_mask(cpumask_t mask, - void (*func)(void *), void *info, - int wait) -{ - int ret; - - /* Can deadlock when called with interrupts disabled */ - WARN_ON(irqs_disabled()); - - spin_lock(&call_lock); - ret = __smp_call_function_mask(mask, func, info, wait); - spin_unlock(&call_lock); - return ret; -} -EXPORT_SYMBOL(smp_call_function_mask); - -/* - * smp_call_function_single - Run a function on a specific CPU - * @func: The function to run. This must be fast and non-blocking. - * @info: An arbitrary pointer to pass to the function. - * @nonatomic: Currently unused. - * @wait: If true, wait until function has completed on other CPUs. - * - * Retrurns 0 on success, else a negative status code. - * - * Does not return until the remote CPU is nearly ready to execute <func> - * or is or has executed. - */ - -int smp_call_function_single (int cpu, void (*func) (void *info), void *info, - int nonatomic, int wait) -{ - /* prevent preemption and reschedule on another processor */ - int ret, me = get_cpu(); - - /* Can deadlock when called with interrupts disabled */ - WARN_ON(irqs_disabled()); - - if (cpu == me) { - local_irq_disable(); - func(info); - local_irq_enable(); - put_cpu(); - return 0; - } - - ret = smp_call_function_mask(cpumask_of_cpu(cpu), func, info, wait); - - put_cpu(); - return ret; -} -EXPORT_SYMBOL(smp_call_function_single); - -/* - * smp_call_function - run a function on all other CPUs. - * @func: The function to run. This must be fast and non-blocking. - * @info: An arbitrary pointer to pass to the function. - * @nonatomic: currently unused. - * @wait: If true, wait (atomically) until function has completed on other - * CPUs. - * - * Returns 0 on success, else a negative status code. Does not return until - * remote CPUs are nearly ready to execute func or are or have executed. - * - * You must not call this function with disabled interrupts or from a - * hardware interrupt handler or from a bottom half handler. - * Actually there are a few legal cases, like panic. - */ -int smp_call_function (void (*func) (void *info), void *info, int nonatomic, - int wait) -{ - return smp_call_function_mask(cpu_online_map, func, info, wait); -} -EXPORT_SYMBOL(smp_call_function); - -static void stop_this_cpu(void *dummy) -{ - local_irq_disable(); - /* - * Remove this CPU: - */ - cpu_clear(smp_processor_id(), cpu_online_map); - disable_local_APIC(); - for (;;) - halt(); -} - -void smp_send_stop(void) -{ - int nolock; - unsigned long flags; - - if (reboot_force) - return; - - /* Don't deadlock on the call lock in panic */ - nolock = !spin_trylock(&call_lock); - local_irq_save(flags); - __smp_call_function_mask(cpu_online_map, stop_this_cpu, NULL, 0); - if (!nolock) - spin_unlock(&call_lock); - disable_local_APIC(); - local_irq_restore(flags); -} - -/* - * Reschedule call back. Nothing to do, - * all the work is done automatically when - * we return from the interrupt. - */ -asmlinkage void smp_reschedule_interrupt(void) -{ - ack_APIC_irq(); - add_pda(irq_resched_count, 1); -} - -asmlinkage void smp_call_function_interrupt(void) -{ - void (*func) (void *info) = call_data->func; - void *info = call_data->info; - int wait = call_data->wait; - - ack_APIC_irq(); - /* - * Notify initiating CPU that I've grabbed the data and am - * about to execute the function - */ - mb(); - atomic_inc(&call_data->started); - /* - * At this point the info structure may be out of scope unless wait==1 - */ - exit_idle(); - irq_enter(); - (*func)(info); - add_pda(irq_call_count, 1); - irq_exit(); - if (wait) { - mb(); - atomic_inc(&call_data->finished); - } -} - diff --git a/arch/x86/kernel/trampoline.c b/arch/x86/kernel/trampoline.c new file mode 100644 index 000000000000..abbf199adebb --- /dev/null +++ b/arch/x86/kernel/trampoline.c @@ -0,0 +1,18 @@ +#include <linux/io.h> + +#include <asm/trampoline.h> + +/* ready for x86_64, no harm for x86, since it will overwrite after alloc */ +unsigned char *trampoline_base = __va(TRAMPOLINE_BASE); + +/* + * Currently trivial. Write the real->protected mode + * bootstrap into the page concerned. The caller + * has made sure it's suitably aligned. + */ +unsigned long setup_trampoline(void) +{ + memcpy(trampoline_base, trampoline_data, + trampoline_end - trampoline_data); + return virt_to_phys(trampoline_base); +} diff --git a/arch/x86/kernel/trampoline_64.S b/arch/x86/kernel/trampoline_64.S index 4aedd0bcee4c..894293c598db 100644 --- a/arch/x86/kernel/trampoline_64.S +++ b/arch/x86/kernel/trampoline_64.S @@ -30,12 +30,7 @@ #include <asm/msr.h> #include <asm/segment.h> -/* We can free up trampoline after bootup if cpu hotplug is not supported. */ -#ifndef CONFIG_HOTPLUG_CPU -.section .init.data, "aw", @progbits -#else .section .rodata, "a", @progbits -#endif .code16 diff --git a/arch/x86/kernel/traps_32.c b/arch/x86/kernel/traps_32.c index c5421f30d678..65791ca2824a 100644 --- a/arch/x86/kernel/traps_32.c +++ b/arch/x86/kernel/traps_32.c @@ -9,26 +9,28 @@ * 'Traps.c' handles hardware traps and faults after we have saved some * state in 'asm.s'. */ -#include <linux/sched.h> +#include <linux/interrupt.h> +#include <linux/kallsyms.h> +#include <linux/spinlock.h> +#include <linux/highmem.h> +#include <linux/kprobes.h> +#include <linux/uaccess.h> +#include <linux/utsname.h> +#include <linux/kdebug.h> #include <linux/kernel.h> +#include <linux/module.h> +#include <linux/ptrace.h> #include <linux/string.h> +#include <linux/unwind.h> +#include <linux/delay.h> #include <linux/errno.h> +#include <linux/kexec.h> +#include <linux/sched.h> #include <linux/timer.h> -#include <linux/mm.h> #include <linux/init.h> -#include <linux/delay.h> -#include <linux/spinlock.h> -#include <linux/interrupt.h> -#include <linux/highmem.h> -#include <linux/kallsyms.h> -#include <linux/ptrace.h> -#include <linux/utsname.h> -#include <linux/kprobes.h> -#include <linux/kexec.h> -#include <linux/unwind.h> -#include <linux/uaccess.h> -#include <linux/nmi.h> #include <linux/bug.h> +#include <linux/nmi.h> +#include <linux/mm.h> #ifdef CONFIG_EISA #include <linux/ioport.h> @@ -43,21 +45,18 @@ #include <linux/edac.h> #endif +#include <asm/arch_hooks.h> +#include <asm/stacktrace.h> #include <asm/processor.h> -#include <asm/system.h> -#include <asm/io.h> -#include <asm/atomic.h> #include <asm/debugreg.h> +#include <asm/atomic.h> +#include <asm/system.h> +#include <asm/unwind.h> #include <asm/desc.h> #include <asm/i387.h> #include <asm/nmi.h> -#include <asm/unwind.h> #include <asm/smp.h> -#include <asm/arch_hooks.h> -#include <linux/kdebug.h> -#include <asm/stacktrace.h> - -#include <linux/module.h> +#include <asm/io.h> #include "mach_traps.h" @@ -69,7 +68,7 @@ EXPORT_SYMBOL_GPL(used_vectors); asmlinkage int system_call(void); /* Do we ignore FPU interrupts ? */ -char ignore_fpu_irq = 0; +char ignore_fpu_irq; /* * The IDT has to be page-aligned to simplify the Pentium @@ -105,12 +104,13 @@ static unsigned int code_bytes = 64; void printk_address(unsigned long address, int reliable) { #ifdef CONFIG_KALLSYMS - unsigned long offset = 0, symsize; + char namebuf[KSYM_NAME_LEN]; + unsigned long offset = 0; + unsigned long symsize; const char *symname; - char *modname; - char *delim = ":"; - char namebuf[128]; char reliab[4] = ""; + char *delim = ":"; + char *modname; symname = kallsyms_lookup(address, &symsize, &offset, &modname, namebuf); @@ -138,13 +138,14 @@ static inline int valid_stack_ptr(struct thread_info *tinfo, void *p, unsigned s /* The form of the top of the frame on the stack */ struct stack_frame { - struct stack_frame *next_frame; - unsigned long return_address; + struct stack_frame *next_frame; + unsigned long return_address; }; -static inline unsigned long print_context_stack(struct thread_info *tinfo, - unsigned long *stack, unsigned long bp, - const struct stacktrace_ops *ops, void *data) +static inline unsigned long +print_context_stack(struct thread_info *tinfo, + unsigned long *stack, unsigned long bp, + const struct stacktrace_ops *ops, void *data) { struct stack_frame *frame = (struct stack_frame *)bp; @@ -166,7 +167,7 @@ static inline unsigned long print_context_stack(struct thread_info *tinfo, return bp; } -#define MSG(msg) ops->warning(data, msg) +#define MSG(msg) ops->warning(data, msg) void dump_trace(struct task_struct *task, struct pt_regs *regs, unsigned long *stack, unsigned long bp, @@ -177,6 +178,7 @@ void dump_trace(struct task_struct *task, struct pt_regs *regs, if (!stack) { unsigned long dummy; + stack = &dummy; if (task != current) stack = (unsigned long *)task->thread.sp; @@ -186,7 +188,7 @@ void dump_trace(struct task_struct *task, struct pt_regs *regs, if (!bp) { if (task == current) { /* Grab bp right from our regs */ - asm ("movl %%ebp, %0" : "=r" (bp) : ); + asm("movl %%ebp, %0" : "=r" (bp) :); } else { /* bp is the last reg pushed by switch_to */ bp = *(unsigned long *) task->thread.sp; @@ -196,15 +198,18 @@ void dump_trace(struct task_struct *task, struct pt_regs *regs, while (1) { struct thread_info *context; + context = (struct thread_info *) ((unsigned long)stack & (~(THREAD_SIZE - 1))); bp = print_context_stack(context, stack, bp, ops, data); - /* Should be after the line below, but somewhere - in early boot context comes out corrupted and we - can't reference it -AK */ + /* + * Should be after the line below, but somewhere + * in early boot context comes out corrupted and we + * can't reference it: + */ if (ops->stack(data, "IRQ") < 0) break; - stack = (unsigned long*)context->previous_esp; + stack = (unsigned long *)context->previous_esp; if (!stack) break; touch_nmi_watchdog(); @@ -243,15 +248,15 @@ static void print_trace_address(void *data, unsigned long addr, int reliable) } static const struct stacktrace_ops print_trace_ops = { - .warning = print_trace_warning, - .warning_symbol = print_trace_warning_symbol, - .stack = print_trace_stack, - .address = print_trace_address, + .warning = print_trace_warning, + .warning_symbol = print_trace_warning_symbol, + .stack = print_trace_stack, + .address = print_trace_address, }; static void show_trace_log_lvl(struct task_struct *task, struct pt_regs *regs, - unsigned long *stack, unsigned long bp, char *log_lvl) + unsigned long *stack, unsigned long bp, char *log_lvl) { dump_trace(task, regs, stack, bp, &print_trace_ops, log_lvl); printk("%s =======================\n", log_lvl); @@ -263,21 +268,22 @@ void show_trace(struct task_struct *task, struct pt_regs *regs, show_trace_log_lvl(task, regs, stack, bp, ""); } -static void show_stack_log_lvl(struct task_struct *task, struct pt_regs *regs, - unsigned long *sp, unsigned long bp, char *log_lvl) +static void +show_stack_log_lvl(struct task_struct *task, struct pt_regs *regs, + unsigned long *sp, unsigned long bp, char *log_lvl) { unsigned long *stack; int i; if (sp == NULL) { if (task) - sp = (unsigned long*)task->thread.sp; + sp = (unsigned long *)task->thread.sp; else sp = (unsigned long *)&sp; } stack = sp; - for(i = 0; i < kstack_depth_to_print; i++) { + for (i = 0; i < kstack_depth_to_print; i++) { if (kstack_end(stack)) break; if (i && ((i % 8) == 0)) @@ -285,6 +291,7 @@ static void show_stack_log_lvl(struct task_struct *task, struct pt_regs *regs, printk("%08lx ", *stack++); } printk("\n%sCall Trace:\n", log_lvl); + show_trace_log_lvl(task, regs, sp, bp, log_lvl); } @@ -299,8 +306,8 @@ void show_stack(struct task_struct *task, unsigned long *sp) */ void dump_stack(void) { - unsigned long stack; unsigned long bp = 0; + unsigned long stack; #ifdef CONFIG_FRAME_POINTER if (!bp) @@ -312,6 +319,7 @@ void dump_stack(void) init_utsname()->release, (int)strcspn(init_utsname()->version, " "), init_utsname()->version); + show_trace(current, NULL, &stack, bp); } @@ -323,6 +331,7 @@ void show_registers(struct pt_regs *regs) print_modules(); __show_registers(regs, 0); + printk(KERN_EMERG "Process %.*s (pid: %d, ti=%p task=%p task.ti=%p)", TASK_COMM_LEN, current->comm, task_pid_nr(current), current_thread_info(), current, task_thread_info(current)); @@ -331,10 +340,10 @@ void show_registers(struct pt_regs *regs) * time of the fault.. */ if (!user_mode_vm(regs)) { - u8 *ip; unsigned int code_prologue = code_bytes * 43 / 64; unsigned int code_len = code_bytes; unsigned char c; + u8 *ip; printk("\n" KERN_EMERG "Stack: "); show_stack_log_lvl(NULL, regs, ®s->sp, 0, KERN_EMERG); @@ -361,7 +370,7 @@ void show_registers(struct pt_regs *regs) } } printk("\n"); -} +} int is_valid_bugaddr(unsigned long ip) { @@ -377,10 +386,10 @@ int is_valid_bugaddr(unsigned long ip) static int die_counter; -int __kprobes __die(const char * str, struct pt_regs * regs, long err) +int __kprobes __die(const char *str, struct pt_regs *regs, long err) { - unsigned long sp; unsigned short ss; + unsigned long sp; printk(KERN_EMERG "%s: %04lx [#%d] ", str, err & 0xffff, ++die_counter); #ifdef CONFIG_PREEMPT @@ -395,8 +404,8 @@ int __kprobes __die(const char * str, struct pt_regs * regs, long err) printk("\n"); if (notify_die(DIE_OOPS, str, regs, err, - current->thread.trap_no, SIGSEGV) != - NOTIFY_STOP) { + current->thread.trap_no, SIGSEGV) != NOTIFY_STOP) { + show_registers(regs); /* Executive summary in case the oops scrolled away */ sp = (unsigned long) (®s->sp); @@ -408,17 +417,18 @@ int __kprobes __die(const char * str, struct pt_regs * regs, long err) printk(KERN_EMERG "EIP: [<%08lx>] ", regs->ip); print_symbol("%s", regs->ip); printk(" SS:ESP %04x:%08lx\n", ss, sp); + return 0; - } else { - return 1; } + + return 1; } /* - * This is gone through when something in the kernel has done something bad and - * is about to be terminated. + * This is gone through when something in the kernel has done something bad + * and is about to be terminated: */ -void die(const char * str, struct pt_regs * regs, long err) +void die(const char *str, struct pt_regs *regs, long err) { static struct { raw_spinlock_t lock; @@ -440,8 +450,9 @@ void die(const char * str, struct pt_regs * regs, long err) die.lock_owner = smp_processor_id(); die.lock_owner_depth = 0; bust_spinlocks(1); - } else + } else { raw_local_irq_save(flags); + } if (++die.lock_owner_depth < 3) { report_bug(regs->ip, regs); @@ -474,19 +485,20 @@ void die(const char * str, struct pt_regs * regs, long err) do_exit(SIGSEGV); } -static inline void die_if_kernel(const char * str, struct pt_regs * regs, long err) +static inline void +die_if_kernel(const char *str, struct pt_regs *regs, long err) { if (!user_mode_vm(regs)) die(str, regs, err); } -static void __kprobes do_trap(int trapnr, int signr, char *str, int vm86, - struct pt_regs * regs, long error_code, - siginfo_t *info) +static void __kprobes +do_trap(int trapnr, int signr, char *str, int vm86, struct pt_regs *regs, + long error_code, siginfo_t *info) { struct task_struct *tsk = current; - if (regs->flags & VM_MASK) { + if (regs->flags & X86_VM_MASK) { if (vm86) goto vm86_trap; goto trap_signal; @@ -495,111 +507,112 @@ static void __kprobes do_trap(int trapnr, int signr, char *str, int vm86, if (!user_mode(regs)) goto kernel_trap; - trap_signal: { - /* - * We want error_code and trap_no set for userspace faults and - * kernelspace faults which result in die(), but not - * kernelspace faults which are fixed up. die() gives the - * process no chance to handle the signal and notice the - * kernel fault information, so that won't result in polluting - * the information about previously queued, but not yet - * delivered, faults. See also do_general_protection below. - */ - tsk->thread.error_code = error_code; - tsk->thread.trap_no = trapnr; +trap_signal: + /* + * We want error_code and trap_no set for userspace faults and + * kernelspace faults which result in die(), but not + * kernelspace faults which are fixed up. die() gives the + * process no chance to handle the signal and notice the + * kernel fault information, so that won't result in polluting + * the information about previously queued, but not yet + * delivered, faults. See also do_general_protection below. + */ + tsk->thread.error_code = error_code; + tsk->thread.trap_no = trapnr; - if (info) - force_sig_info(signr, info, tsk); - else - force_sig(signr, tsk); - return; - } + if (info) + force_sig_info(signr, info, tsk); + else + force_sig(signr, tsk); + return; - kernel_trap: { - if (!fixup_exception(regs)) { - tsk->thread.error_code = error_code; - tsk->thread.trap_no = trapnr; - die(str, regs, error_code); - } - return; +kernel_trap: + if (!fixup_exception(regs)) { + tsk->thread.error_code = error_code; + tsk->thread.trap_no = trapnr; + die(str, regs, error_code); } + return; - vm86_trap: { - int ret = handle_vm86_trap((struct kernel_vm86_regs *) regs, error_code, trapnr); - if (ret) goto trap_signal; - return; - } +vm86_trap: + if (handle_vm86_trap((struct kernel_vm86_regs *) regs, + error_code, trapnr)) + goto trap_signal; + return; } -#define DO_ERROR(trapnr, signr, str, name) \ -void do_##name(struct pt_regs * regs, long error_code) \ -{ \ - if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \ - == NOTIFY_STOP) \ - return; \ - do_trap(trapnr, signr, str, 0, regs, error_code, NULL); \ +#define DO_ERROR(trapnr, signr, str, name) \ +void do_##name(struct pt_regs *regs, long error_code) \ +{ \ + if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \ + == NOTIFY_STOP) \ + return; \ + do_trap(trapnr, signr, str, 0, regs, error_code, NULL); \ } -#define DO_ERROR_INFO(trapnr, signr, str, name, sicode, siaddr, irq) \ -void do_##name(struct pt_regs * regs, long error_code) \ -{ \ - siginfo_t info; \ - if (irq) \ - local_irq_enable(); \ - info.si_signo = signr; \ - info.si_errno = 0; \ - info.si_code = sicode; \ - info.si_addr = (void __user *)siaddr; \ - if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \ - == NOTIFY_STOP) \ - return; \ - do_trap(trapnr, signr, str, 0, regs, error_code, &info); \ +#define DO_ERROR_INFO(trapnr, signr, str, name, sicode, siaddr, irq) \ +void do_##name(struct pt_regs *regs, long error_code) \ +{ \ + siginfo_t info; \ + if (irq) \ + local_irq_enable(); \ + info.si_signo = signr; \ + info.si_errno = 0; \ + info.si_code = sicode; \ + info.si_addr = (void __user *)siaddr; \ + if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \ + == NOTIFY_STOP) \ + return; \ + do_trap(trapnr, signr, str, 0, regs, error_code, &info); \ } -#define DO_VM86_ERROR(trapnr, signr, str, name) \ -void do_##name(struct pt_regs * regs, long error_code) \ -{ \ - if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \ - == NOTIFY_STOP) \ - return; \ - do_trap(trapnr, signr, str, 1, regs, error_code, NULL); \ +#define DO_VM86_ERROR(trapnr, signr, str, name) \ +void do_##name(struct pt_regs *regs, long error_code) \ +{ \ + if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \ + == NOTIFY_STOP) \ + return; \ + do_trap(trapnr, signr, str, 1, regs, error_code, NULL); \ } -#define DO_VM86_ERROR_INFO(trapnr, signr, str, name, sicode, siaddr) \ -void do_##name(struct pt_regs * regs, long error_code) \ -{ \ - siginfo_t info; \ - info.si_signo = signr; \ - info.si_errno = 0; \ - info.si_code = sicode; \ - info.si_addr = (void __user *)siaddr; \ - trace_hardirqs_fixup(); \ - if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \ - == NOTIFY_STOP) \ - return; \ - do_trap(trapnr, signr, str, 1, regs, error_code, &info); \ +#define DO_VM86_ERROR_INFO(trapnr, signr, str, name, sicode, siaddr) \ +void do_##name(struct pt_regs *regs, long error_code) \ +{ \ + siginfo_t info; \ + info.si_signo = signr; \ + info.si_errno = 0; \ + info.si_code = sicode; \ + info.si_addr = (void __user *)siaddr; \ + trace_hardirqs_fixup(); \ + if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \ + == NOTIFY_STOP) \ + return; \ + do_trap(trapnr, signr, str, 1, regs, error_code, &info); \ } -DO_VM86_ERROR_INFO( 0, SIGFPE, "divide error", divide_error, FPE_INTDIV, regs->ip) +DO_VM86_ERROR_INFO(0, SIGFPE, "divide error", divide_error, FPE_INTDIV, regs->ip) #ifndef CONFIG_KPROBES -DO_VM86_ERROR( 3, SIGTRAP, "int3", int3) +DO_VM86_ERROR(3, SIGTRAP, "int3", int3) #endif -DO_VM86_ERROR( 4, SIGSEGV, "overflow", overflow) -DO_VM86_ERROR( 5, SIGSEGV, "bounds", bounds) -DO_ERROR_INFO( 6, SIGILL, "invalid opcode", invalid_op, ILL_ILLOPN, regs->ip, 0) -DO_ERROR( 9, SIGFPE, "coprocessor segment overrun", coprocessor_segment_overrun) +DO_VM86_ERROR(4, SIGSEGV, "overflow", overflow) +DO_VM86_ERROR(5, SIGSEGV, "bounds", bounds) +DO_ERROR_INFO(6, SIGILL, "invalid opcode", invalid_op, ILL_ILLOPN, regs->ip, 0) +DO_ERROR(9, SIGFPE, "coprocessor segment overrun", coprocessor_segment_overrun) DO_ERROR(10, SIGSEGV, "invalid TSS", invalid_TSS) DO_ERROR(11, SIGBUS, "segment not present", segment_not_present) DO_ERROR(12, SIGBUS, "stack segment", stack_segment) DO_ERROR_INFO(17, SIGBUS, "alignment check", alignment_check, BUS_ADRALN, 0, 0) DO_ERROR_INFO(32, SIGSEGV, "iret exception", iret_error, ILL_BADSTK, 0, 1) -void __kprobes do_general_protection(struct pt_regs * regs, - long error_code) +void __kprobes do_general_protection(struct pt_regs *regs, long error_code) { - int cpu = get_cpu(); - struct tss_struct *tss = &per_cpu(init_tss, cpu); - struct thread_struct *thread = ¤t->thread; + struct thread_struct *thread; + struct tss_struct *tss; + int cpu; + + cpu = get_cpu(); + tss = &per_cpu(init_tss, cpu); + thread = ¤t->thread; /* * Perform the lazy TSS's I/O bitmap copy. If the TSS has an @@ -616,19 +629,21 @@ void __kprobes do_general_protection(struct pt_regs * regs, * If the previously set map was extending to higher ports * than the current one, pad extra space with 0xff (no access). */ - if (thread->io_bitmap_max < tss->io_bitmap_max) + if (thread->io_bitmap_max < tss->io_bitmap_max) { memset((char *) tss->io_bitmap + thread->io_bitmap_max, 0xff, tss->io_bitmap_max - thread->io_bitmap_max); + } tss->io_bitmap_max = thread->io_bitmap_max; tss->x86_tss.io_bitmap_base = IO_BITMAP_OFFSET; tss->io_bitmap_owner = thread; put_cpu(); + return; } put_cpu(); - if (regs->flags & VM_MASK) + if (regs->flags & X86_VM_MASK) goto gp_in_vm86; if (!user_mode(regs)) @@ -636,6 +651,7 @@ void __kprobes do_general_protection(struct pt_regs * regs, current->thread.error_code = error_code; current->thread.trap_no = 13; + if (show_unhandled_signals && unhandled_signal(current, SIGSEGV) && printk_ratelimit()) { printk(KERN_INFO @@ -666,21 +682,24 @@ gp_in_kernel: } static __kprobes void -mem_parity_error(unsigned char reason, struct pt_regs * regs) +mem_parity_error(unsigned char reason, struct pt_regs *regs) { - printk(KERN_EMERG "Uhhuh. NMI received for unknown reason %02x on " - "CPU %d.\n", reason, smp_processor_id()); - printk(KERN_EMERG "You have some hardware problem, likely on the PCI bus.\n"); + printk(KERN_EMERG + "Uhhuh. NMI received for unknown reason %02x on CPU %d.\n", + reason, smp_processor_id()); + + printk(KERN_EMERG + "You have some hardware problem, likely on the PCI bus.\n"); #if defined(CONFIG_EDAC) - if(edac_handler_set()) { + if (edac_handler_set()) { edac_atomic_assert_error(); return; } #endif if (panic_on_unrecovered_nmi) - panic("NMI: Not continuing"); + panic("NMI: Not continuing"); printk(KERN_EMERG "Dazed and confused, but trying to continue\n"); @@ -689,7 +708,7 @@ mem_parity_error(unsigned char reason, struct pt_regs * regs) } static __kprobes void -io_check_error(unsigned char reason, struct pt_regs * regs) +io_check_error(unsigned char reason, struct pt_regs *regs) { unsigned long i; @@ -699,30 +718,37 @@ io_check_error(unsigned char reason, struct pt_regs * regs) /* Re-enable the IOCK line, wait for a few seconds */ reason = (reason & 0xf) | 8; outb(reason, 0x61); + i = 2000; - while (--i) udelay(1000); + while (--i) + udelay(1000); + reason &= ~8; outb(reason, 0x61); } static __kprobes void -unknown_nmi_error(unsigned char reason, struct pt_regs * regs) +unknown_nmi_error(unsigned char reason, struct pt_regs *regs) { if (notify_die(DIE_NMIUNKNOWN, "nmi", regs, reason, 2, SIGINT) == NOTIFY_STOP) return; #ifdef CONFIG_MCA - /* Might actually be able to figure out what the guilty party - * is. */ - if( MCA_bus ) { + /* + * Might actually be able to figure out what the guilty party + * is: + */ + if (MCA_bus) { mca_handle_nmi(); return; } #endif - printk(KERN_EMERG "Uhhuh. NMI received for unknown reason %02x on " - "CPU %d.\n", reason, smp_processor_id()); + printk(KERN_EMERG + "Uhhuh. NMI received for unknown reason %02x on CPU %d.\n", + reason, smp_processor_id()); + printk(KERN_EMERG "Do you have a strange power saving mode enabled?\n"); if (panic_on_unrecovered_nmi) - panic("NMI: Not continuing"); + panic("NMI: Not continuing"); printk(KERN_EMERG "Dazed and confused, but trying to continue\n"); } @@ -731,14 +757,13 @@ static DEFINE_SPINLOCK(nmi_print_lock); void __kprobes die_nmi(struct pt_regs *regs, const char *msg) { - if (notify_die(DIE_NMIWATCHDOG, msg, regs, 0, 2, SIGINT) == - NOTIFY_STOP) + if (notify_die(DIE_NMIWATCHDOG, msg, regs, 0, 2, SIGINT) == NOTIFY_STOP) return; spin_lock(&nmi_print_lock); /* * We are in trouble anyway, lets at least try - * to get a message out. + * to get a message out: */ bust_spinlocks(1); printk(KERN_EMERG "%s", msg); @@ -749,9 +774,10 @@ void __kprobes die_nmi(struct pt_regs *regs, const char *msg) spin_unlock(&nmi_print_lock); bust_spinlocks(0); - /* If we are in kernel we are probably nested up pretty bad - * and might aswell get out now while we still can. - */ + /* + * If we are in kernel we are probably nested up pretty bad + * and might aswell get out now while we still can: + */ if (!user_mode_vm(regs)) { current->thread.trap_no = 2; crash_kexec(regs); @@ -760,14 +786,14 @@ void __kprobes die_nmi(struct pt_regs *regs, const char *msg) do_exit(SIGSEGV); } -static __kprobes void default_do_nmi(struct pt_regs * regs) +static __kprobes void default_do_nmi(struct pt_regs *regs) { unsigned char reason = 0; - /* Only the BSP gets external NMIs from the system. */ + /* Only the BSP gets external NMIs from the system: */ if (!smp_processor_id()) reason = get_nmi_reason(); - + if (!(reason & 0xc0)) { if (notify_die(DIE_NMI_IPI, "nmi_ipi", regs, reason, 2, SIGINT) == NOTIFY_STOP) @@ -780,8 +806,10 @@ static __kprobes void default_do_nmi(struct pt_regs * regs) if (nmi_watchdog_tick(regs, reason)) return; if (!do_nmi_callback(regs, smp_processor_id())) -#endif unknown_nmi_error(reason, regs); +#else + unknown_nmi_error(reason, regs); +#endif return; } @@ -793,14 +821,14 @@ static __kprobes void default_do_nmi(struct pt_regs * regs) io_check_error(reason, regs); /* * Reassert NMI in case it became active meanwhile - * as it's edge-triggered. + * as it's edge-triggered: */ reassert_nmi(); } static int ignore_nmis; -__kprobes void do_nmi(struct pt_regs * regs, long error_code) +__kprobes void do_nmi(struct pt_regs *regs, long error_code) { int cpu; @@ -836,9 +864,12 @@ void __kprobes do_int3(struct pt_regs *regs, long error_code) if (notify_die(DIE_INT3, "int3", regs, error_code, 3, SIGTRAP) == NOTIFY_STOP) return; - /* This is an interrupt gate, because kprobes wants interrupts - disabled. Normal trap handlers don't. */ + /* + * This is an interrupt gate, because kprobes wants interrupts + * disabled. Normal trap handlers don't. + */ restore_interrupts(regs); + do_trap(3, SIGTRAP, "int3", 1, regs, error_code, NULL); } #endif @@ -853,7 +884,7 @@ void __kprobes do_int3(struct pt_regs *regs, long error_code) * from user space. Such code must not hold kernel locks (since it * can equally take a page fault), therefore it is safe to call * force_sig_info even though that claims and releases locks. - * + * * Code in ./signal.c ensures that the debug control register * is restored before we deliver any signal, and therefore that * user code runs with the correct debug control register even though @@ -865,10 +896,10 @@ void __kprobes do_int3(struct pt_regs *regs, long error_code) * find every occurrence of the TF bit that could be saved away even * by user code) */ -void __kprobes do_debug(struct pt_regs * regs, long error_code) +void __kprobes do_debug(struct pt_regs *regs, long error_code) { - unsigned int condition; struct task_struct *tsk = current; + unsigned int condition; trace_hardirqs_fixup(); @@ -893,7 +924,7 @@ void __kprobes do_debug(struct pt_regs * regs, long error_code) goto clear_dr7; } - if (regs->flags & VM_MASK) + if (regs->flags & X86_VM_MASK) goto debug_vm86; /* Save debug status register where ptrace can see it */ @@ -916,7 +947,8 @@ void __kprobes do_debug(struct pt_regs * regs, long error_code) /* Ok, finally something we can handle */ send_sigtrap(tsk, regs, error_code); - /* Disable additional traps. They'll be re-enabled when + /* + * Disable additional traps. They'll be re-enabled when * the signal is delivered. */ clear_dr7: @@ -929,7 +961,7 @@ debug_vm86: clear_TF_reenable: set_tsk_thread_flag(tsk, TIF_SINGLESTEP); - regs->flags &= ~TF_MASK; + regs->flags &= ~X86_EFLAGS_TF; return; } @@ -940,9 +972,10 @@ clear_TF_reenable: */ void math_error(void __user *ip) { - struct task_struct * task; + struct task_struct *task; + unsigned short cwd; + unsigned short swd; siginfo_t info; - unsigned short cwd, swd; /* * Save the info for the exception handler and clear the error. @@ -968,36 +1001,36 @@ void math_error(void __user *ip) cwd = get_fpu_cwd(task); swd = get_fpu_swd(task); switch (swd & ~cwd & 0x3f) { - case 0x000: /* No unmasked exception */ - return; - default: /* Multiple exceptions */ - break; - case 0x001: /* Invalid Op */ - /* - * swd & 0x240 == 0x040: Stack Underflow - * swd & 0x240 == 0x240: Stack Overflow - * User must clear the SF bit (0x40) if set - */ - info.si_code = FPE_FLTINV; - break; - case 0x002: /* Denormalize */ - case 0x010: /* Underflow */ - info.si_code = FPE_FLTUND; - break; - case 0x004: /* Zero Divide */ - info.si_code = FPE_FLTDIV; - break; - case 0x008: /* Overflow */ - info.si_code = FPE_FLTOVF; - break; - case 0x020: /* Precision */ - info.si_code = FPE_FLTRES; - break; + case 0x000: /* No unmasked exception */ + return; + default: /* Multiple exceptions */ + break; + case 0x001: /* Invalid Op */ + /* + * swd & 0x240 == 0x040: Stack Underflow + * swd & 0x240 == 0x240: Stack Overflow + * User must clear the SF bit (0x40) if set + */ + info.si_code = FPE_FLTINV; + break; + case 0x002: /* Denormalize */ + case 0x010: /* Underflow */ + info.si_code = FPE_FLTUND; + break; + case 0x004: /* Zero Divide */ + info.si_code = FPE_FLTDIV; + break; + case 0x008: /* Overflow */ + info.si_code = FPE_FLTOVF; + break; + case 0x020: /* Precision */ + info.si_code = FPE_FLTRES; + break; } force_sig_info(SIGFPE, &info, task); } -void do_coprocessor_error(struct pt_regs * regs, long error_code) +void do_coprocessor_error(struct pt_regs *regs, long error_code) { ignore_fpu_irq = 1; math_error((void __user *)regs->ip); @@ -1005,9 +1038,9 @@ void do_coprocessor_error(struct pt_regs * regs, long error_code) static void simd_math_error(void __user *ip) { - struct task_struct * task; - siginfo_t info; + struct task_struct *task; unsigned short mxcsr; + siginfo_t info; /* * Save the info for the exception handler and clear the error. @@ -1028,82 +1061,80 @@ static void simd_math_error(void __user *ip) */ mxcsr = get_fpu_mxcsr(task); switch (~((mxcsr & 0x1f80) >> 7) & (mxcsr & 0x3f)) { - case 0x000: - default: - break; - case 0x001: /* Invalid Op */ - info.si_code = FPE_FLTINV; - break; - case 0x002: /* Denormalize */ - case 0x010: /* Underflow */ - info.si_code = FPE_FLTUND; - break; - case 0x004: /* Zero Divide */ - info.si_code = FPE_FLTDIV; - break; - case 0x008: /* Overflow */ - info.si_code = FPE_FLTOVF; - break; - case 0x020: /* Precision */ - info.si_code = FPE_FLTRES; - break; + case 0x000: + default: + break; + case 0x001: /* Invalid Op */ + info.si_code = FPE_FLTINV; + break; + case 0x002: /* Denormalize */ + case 0x010: /* Underflow */ + info.si_code = FPE_FLTUND; + break; + case 0x004: /* Zero Divide */ + info.si_code = FPE_FLTDIV; + break; + case 0x008: /* Overflow */ + info.si_code = FPE_FLTOVF; + break; + case 0x020: /* Precision */ + info.si_code = FPE_FLTRES; + break; } force_sig_info(SIGFPE, &info, task); } -void do_simd_coprocessor_error(struct pt_regs * regs, - long error_code) +void do_simd_coprocessor_error(struct pt_regs *regs, long error_code) { if (cpu_has_xmm) { /* Handle SIMD FPU exceptions on PIII+ processors. */ ignore_fpu_irq = 1; simd_math_error((void __user *)regs->ip); - } else { - /* - * Handle strange cache flush from user space exception - * in all other cases. This is undocumented behaviour. - */ - if (regs->flags & VM_MASK) { - handle_vm86_fault((struct kernel_vm86_regs *)regs, - error_code); - return; - } - current->thread.trap_no = 19; - current->thread.error_code = error_code; - die_if_kernel("cache flush denied", regs, error_code); - force_sig(SIGSEGV, current); + return; + } + /* + * Handle strange cache flush from user space exception + * in all other cases. This is undocumented behaviour. + */ + if (regs->flags & X86_VM_MASK) { + handle_vm86_fault((struct kernel_vm86_regs *)regs, error_code); + return; } + current->thread.trap_no = 19; + current->thread.error_code = error_code; + die_if_kernel("cache flush denied", regs, error_code); + force_sig(SIGSEGV, current); } -void do_spurious_interrupt_bug(struct pt_regs * regs, - long error_code) +void do_spurious_interrupt_bug(struct pt_regs *regs, long error_code) { #if 0 /* No need to warn about this any longer. */ - printk("Ignoring P6 Local APIC Spurious Interrupt Bug...\n"); + printk(KERN_INFO "Ignoring P6 Local APIC Spurious Interrupt Bug...\n"); #endif } -unsigned long patch_espfix_desc(unsigned long uesp, - unsigned long kesp) +unsigned long patch_espfix_desc(unsigned long uesp, unsigned long kesp) { struct desc_struct *gdt = __get_cpu_var(gdt_page).gdt; unsigned long base = (kesp - uesp) & -THREAD_SIZE; unsigned long new_kesp = kesp - base; unsigned long lim_pages = (new_kesp | (THREAD_SIZE - 1)) >> PAGE_SHIFT; __u64 desc = *(__u64 *)&gdt[GDT_ENTRY_ESPFIX_SS]; + /* Set up base for espfix segment */ - desc &= 0x00f0ff0000000000ULL; - desc |= ((((__u64)base) << 16) & 0x000000ffffff0000ULL) | + desc &= 0x00f0ff0000000000ULL; + desc |= ((((__u64)base) << 16) & 0x000000ffffff0000ULL) | ((((__u64)base) << 32) & 0xff00000000000000ULL) | ((((__u64)lim_pages) << 32) & 0x000f000000000000ULL) | (lim_pages & 0xffff); *(__u64 *)&gdt[GDT_ENTRY_ESPFIX_SS] = desc; + return new_kesp; } /* - * 'math_state_restore()' saves the current math information in the + * 'math_state_restore()' saves the current math information in the * old math state array, and gets the new ones from the current task * * Careful.. There are problems with IBM-designed IRQ13 behaviour. @@ -1117,7 +1148,7 @@ asmlinkage void math_state_restore(void) struct thread_info *thread = current_thread_info(); struct task_struct *tsk = thread->task; - clts(); /* Allow maths ops (or we recurse) */ + clts(); /* Allow maths ops (or we recurse) */ if (!tsk_used_math(tsk)) init_fpu(tsk); restore_fpu(tsk); @@ -1130,53 +1161,52 @@ EXPORT_SYMBOL_GPL(math_state_restore); asmlinkage void math_emulate(long arg) { - printk(KERN_EMERG "math-emulation not enabled and no coprocessor found.\n"); - printk(KERN_EMERG "killing %s.\n",current->comm); - force_sig(SIGFPE,current); + printk(KERN_EMERG + "math-emulation not enabled and no coprocessor found.\n"); + printk(KERN_EMERG "killing %s.\n", current->comm); + force_sig(SIGFPE, current); schedule(); } #endif /* CONFIG_MATH_EMULATION */ - void __init trap_init(void) { int i; #ifdef CONFIG_EISA void __iomem *p = early_ioremap(0x0FFFD9, 4); - if (readl(p) == 'E'+('I'<<8)+('S'<<16)+('A'<<24)) { + + if (readl(p) == 'E' + ('I'<<8) + ('S'<<16) + ('A'<<24)) EISA_bus = 1; - } early_iounmap(p, 4); #endif #ifdef CONFIG_X86_LOCAL_APIC init_apic_mappings(); #endif - - set_trap_gate(0,÷_error); - set_intr_gate(1,&debug); - set_intr_gate(2,&nmi); + set_trap_gate(0, ÷_error); + set_intr_gate(1, &debug); + set_intr_gate(2, &nmi); set_system_intr_gate(3, &int3); /* int3/4 can be called from all */ - set_system_gate(4,&overflow); - set_trap_gate(5,&bounds); - set_trap_gate(6,&invalid_op); - set_trap_gate(7,&device_not_available); - set_task_gate(8,GDT_ENTRY_DOUBLEFAULT_TSS); - set_trap_gate(9,&coprocessor_segment_overrun); - set_trap_gate(10,&invalid_TSS); - set_trap_gate(11,&segment_not_present); - set_trap_gate(12,&stack_segment); - set_trap_gate(13,&general_protection); - set_intr_gate(14,&page_fault); - set_trap_gate(15,&spurious_interrupt_bug); - set_trap_gate(16,&coprocessor_error); - set_trap_gate(17,&alignment_check); + set_system_gate(4, &overflow); + set_trap_gate(5, &bounds); + set_trap_gate(6, &invalid_op); + set_trap_gate(7, &device_not_available); + set_task_gate(8, GDT_ENTRY_DOUBLEFAULT_TSS); + set_trap_gate(9, &coprocessor_segment_overrun); + set_trap_gate(10, &invalid_TSS); + set_trap_gate(11, &segment_not_present); + set_trap_gate(12, &stack_segment); + set_trap_gate(13, &general_protection); + set_intr_gate(14, &page_fault); + set_trap_gate(15, &spurious_interrupt_bug); + set_trap_gate(16, &coprocessor_error); + set_trap_gate(17, &alignment_check); #ifdef CONFIG_X86_MCE - set_trap_gate(18,&machine_check); + set_trap_gate(18, &machine_check); #endif - set_trap_gate(19,&simd_coprocessor_error); + set_trap_gate(19, &simd_coprocessor_error); /* * Verify that the FXSAVE/FXRSTOR data will be 16-byte aligned. @@ -1189,21 +1219,22 @@ void __init trap_init(void) printk("done.\n"); } if (cpu_has_xmm) { - printk(KERN_INFO "Enabling unmasked SIMD FPU exception " - "support... "); + printk(KERN_INFO + "Enabling unmasked SIMD FPU exception support... "); set_in_cr4(X86_CR4_OSXMMEXCPT); printk("done.\n"); } - set_system_gate(SYSCALL_VECTOR,&system_call); + set_system_gate(SYSCALL_VECTOR, &system_call); - /* Reserve all the builtin and the syscall vector. */ + /* Reserve all the builtin and the syscall vector: */ for (i = 0; i < FIRST_EXTERNAL_VECTOR; i++) set_bit(i, used_vectors); + set_bit(SYSCALL_VECTOR, used_vectors); /* - * Should be a barrier for any external CPU state. + * Should be a barrier for any external CPU state: */ cpu_init(); @@ -1213,6 +1244,7 @@ void __init trap_init(void) static int __init kstack_setup(char *s) { kstack_depth_to_print = simple_strtoul(s, NULL, 0); + return 1; } __setup("kstack=", kstack_setup); diff --git a/arch/x86/kernel/traps_64.c b/arch/x86/kernel/traps_64.c index 4e073320e70a..79aa6fc0815c 100644 --- a/arch/x86/kernel/traps_64.c +++ b/arch/x86/kernel/traps_64.c @@ -33,6 +33,8 @@ #include <linux/kdebug.h> #include <linux/utsname.h> +#include <mach_traps.h> + #if defined(CONFIG_EDAC) #include <linux/edac.h> #endif diff --git a/arch/x86/kernel/tsc_32.c b/arch/x86/kernel/tsc_32.c index c2241e04ea5f..3d7e6e9fa6c2 100644 --- a/arch/x86/kernel/tsc_32.c +++ b/arch/x86/kernel/tsc_32.c @@ -84,8 +84,8 @@ DEFINE_PER_CPU(unsigned long, cyc2ns); static void set_cyc2ns_scale(unsigned long cpu_khz, int cpu) { - unsigned long flags, prev_scale, *scale; unsigned long long tsc_now, ns_now; + unsigned long flags, *scale; local_irq_save(flags); sched_clock_idle_sleep_event(); @@ -95,7 +95,6 @@ static void set_cyc2ns_scale(unsigned long cpu_khz, int cpu) rdtscll(tsc_now); ns_now = __cycles_2_ns(tsc_now); - prev_scale = *scale; if (cpu_khz) *scale = (NSEC_PER_MSEC << CYC2NS_SCALE_FACTOR)/cpu_khz; @@ -392,13 +391,15 @@ void __init tsc_init(void) int cpu; if (!cpu_has_tsc) - goto out_no_tsc; + return; cpu_khz = calculate_cpu_khz(); tsc_khz = cpu_khz; - if (!cpu_khz) - goto out_no_tsc; + if (!cpu_khz) { + mark_tsc_unstable("could not calculate TSC khz"); + return; + } printk("Detected %lu.%03lu MHz processor.\n", (unsigned long)cpu_khz / 1000, @@ -431,9 +432,4 @@ void __init tsc_init(void) tsc_enabled = 1; clocksource_register(&clocksource_tsc); - - return; - -out_no_tsc: - setup_clear_cpu_cap(X86_FEATURE_TSC); } diff --git a/arch/x86/kernel/tsc_64.c b/arch/x86/kernel/tsc_64.c index d3bebaaad842..ceeba01e7f47 100644 --- a/arch/x86/kernel/tsc_64.c +++ b/arch/x86/kernel/tsc_64.c @@ -44,8 +44,8 @@ DEFINE_PER_CPU(unsigned long, cyc2ns); static void set_cyc2ns_scale(unsigned long cpu_khz, int cpu) { - unsigned long flags, prev_scale, *scale; unsigned long long tsc_now, ns_now; + unsigned long flags, *scale; local_irq_save(flags); sched_clock_idle_sleep_event(); @@ -55,7 +55,6 @@ static void set_cyc2ns_scale(unsigned long cpu_khz, int cpu) rdtscll(tsc_now); ns_now = __cycles_2_ns(tsc_now); - prev_scale = *scale; if (cpu_khz) *scale = (NSEC_PER_MSEC << CYC2NS_SCALE_FACTOR)/cpu_khz; diff --git a/arch/x86/kernel/vm86_32.c b/arch/x86/kernel/vm86_32.c index 738c2104df30..38f566fa27d2 100644 --- a/arch/x86/kernel/vm86_32.c +++ b/arch/x86/kernel/vm86_32.c @@ -64,7 +64,7 @@ #define KVM86 ((struct kernel_vm86_struct *)regs) -#define VMPI KVM86->vm86plus +#define VMPI KVM86->vm86plus /* @@ -81,7 +81,7 @@ #define VFLAGS (*(unsigned short *)&(current->thread.v86flags)) #define VEFLAGS (current->thread.v86flags) -#define set_flags(X,new,mask) \ +#define set_flags(X, new, mask) \ ((X) = ((X) & ~(mask)) | ((new) & (mask))) #define SAFE_MASK (0xDD5) @@ -93,8 +93,10 @@ static int copy_vm86_regs_to_user(struct vm86_regs __user *user, { int ret = 0; - /* kernel_vm86_regs is missing gs, so copy everything up to - (but not including) orig_eax, and then rest including orig_eax. */ + /* + * kernel_vm86_regs is missing gs, so copy everything up to + * (but not including) orig_eax, and then rest including orig_eax. + */ ret += copy_to_user(user, regs, offsetof(struct kernel_vm86_regs, pt.orig_ax)); ret += copy_to_user(&user->orig_eax, ®s->pt.orig_ax, sizeof(struct kernel_vm86_regs) - @@ -120,7 +122,7 @@ static int copy_vm86_regs_from_user(struct kernel_vm86_regs *regs, return ret; } -struct pt_regs * save_v86_state(struct kernel_vm86_regs * regs) +struct pt_regs *save_v86_state(struct kernel_vm86_regs *regs) { struct tss_struct *tss; struct pt_regs *ret; @@ -137,9 +139,9 @@ struct pt_regs * save_v86_state(struct kernel_vm86_regs * regs) printk("no vm86_info: BAD\n"); do_exit(SIGSEGV); } - set_flags(regs->pt.flags, VEFLAGS, VIF_MASK | current->thread.v86mask); - tmp = copy_vm86_regs_to_user(¤t->thread.vm86_info->regs,regs); - tmp += put_user(current->thread.screen_bitmap,¤t->thread.vm86_info->screen_bitmap); + set_flags(regs->pt.flags, VEFLAGS, X86_EFLAGS_VIF | current->thread.v86mask); + tmp = copy_vm86_regs_to_user(¤t->thread.vm86_info->regs, regs); + tmp += put_user(current->thread.screen_bitmap, ¤t->thread.vm86_info->screen_bitmap); if (tmp) { printk("vm86: could not access userspace vm86_info\n"); do_exit(SIGSEGV); @@ -237,20 +239,21 @@ asmlinkage int sys_vm86(struct pt_regs regs) tsk = current; switch (regs.bx) { - case VM86_REQUEST_IRQ: - case VM86_FREE_IRQ: - case VM86_GET_IRQ_BITS: - case VM86_GET_AND_RESET_IRQ: - ret = do_vm86_irq_handling(regs.bx, (int)regs.cx); - goto out; - case VM86_PLUS_INSTALL_CHECK: - /* NOTE: on old vm86 stuff this will return the error - from access_ok(), because the subfunction is - interpreted as (invalid) address to vm86_struct. - So the installation check works. - */ - ret = 0; - goto out; + case VM86_REQUEST_IRQ: + case VM86_FREE_IRQ: + case VM86_GET_IRQ_BITS: + case VM86_GET_AND_RESET_IRQ: + ret = do_vm86_irq_handling(regs.bx, (int)regs.cx); + goto out; + case VM86_PLUS_INSTALL_CHECK: + /* + * NOTE: on old vm86 stuff this will return the error + * from access_ok(), because the subfunction is + * interpreted as (invalid) address to vm86_struct. + * So the installation check works. + */ + ret = 0; + goto out; } /* we come here only for functions VM86_ENTER, VM86_ENTER_NO_BYPASS */ @@ -296,21 +299,21 @@ static void do_sys_vm86(struct kernel_vm86_struct *info, struct task_struct *tsk VEFLAGS = info->regs.pt.flags; info->regs.pt.flags &= SAFE_MASK; info->regs.pt.flags |= info->regs32->flags & ~SAFE_MASK; - info->regs.pt.flags |= VM_MASK; + info->regs.pt.flags |= X86_VM_MASK; switch (info->cpu_type) { - case CPU_286: - tsk->thread.v86mask = 0; - break; - case CPU_386: - tsk->thread.v86mask = NT_MASK | IOPL_MASK; - break; - case CPU_486: - tsk->thread.v86mask = AC_MASK | NT_MASK | IOPL_MASK; - break; - default: - tsk->thread.v86mask = ID_MASK | AC_MASK | NT_MASK | IOPL_MASK; - break; + case CPU_286: + tsk->thread.v86mask = 0; + break; + case CPU_386: + tsk->thread.v86mask = X86_EFLAGS_NT | X86_EFLAGS_IOPL; + break; + case CPU_486: + tsk->thread.v86mask = X86_EFLAGS_AC | X86_EFLAGS_NT | X86_EFLAGS_IOPL; + break; + default: + tsk->thread.v86mask = X86_EFLAGS_ID | X86_EFLAGS_AC | X86_EFLAGS_NT | X86_EFLAGS_IOPL; + break; } /* @@ -346,9 +349,9 @@ static void do_sys_vm86(struct kernel_vm86_struct *info, struct task_struct *tsk /* we never return here */ } -static inline void return_to_32bit(struct kernel_vm86_regs * regs16, int retval) +static inline void return_to_32bit(struct kernel_vm86_regs *regs16, int retval) { - struct pt_regs * regs32; + struct pt_regs *regs32; regs32 = save_v86_state(regs16); regs32->ax = retval; @@ -358,29 +361,30 @@ static inline void return_to_32bit(struct kernel_vm86_regs * regs16, int retval) : : "r" (regs32), "r" (current_thread_info())); } -static inline void set_IF(struct kernel_vm86_regs * regs) +static inline void set_IF(struct kernel_vm86_regs *regs) { - VEFLAGS |= VIF_MASK; - if (VEFLAGS & VIP_MASK) + VEFLAGS |= X86_EFLAGS_VIF; + if (VEFLAGS & X86_EFLAGS_VIP) return_to_32bit(regs, VM86_STI); } -static inline void clear_IF(struct kernel_vm86_regs * regs) +static inline void clear_IF(struct kernel_vm86_regs *regs) { - VEFLAGS &= ~VIF_MASK; + VEFLAGS &= ~X86_EFLAGS_VIF; } -static inline void clear_TF(struct kernel_vm86_regs * regs) +static inline void clear_TF(struct kernel_vm86_regs *regs) { - regs->pt.flags &= ~TF_MASK; + regs->pt.flags &= ~X86_EFLAGS_TF; } -static inline void clear_AC(struct kernel_vm86_regs * regs) +static inline void clear_AC(struct kernel_vm86_regs *regs) { - regs->pt.flags &= ~AC_MASK; + regs->pt.flags &= ~X86_EFLAGS_AC; } -/* It is correct to call set_IF(regs) from the set_vflags_* +/* + * It is correct to call set_IF(regs) from the set_vflags_* * functions. However someone forgot to call clear_IF(regs) * in the opposite case. * After the command sequence CLI PUSHF STI POPF you should @@ -391,41 +395,41 @@ static inline void clear_AC(struct kernel_vm86_regs * regs) * [KD] */ -static inline void set_vflags_long(unsigned long flags, struct kernel_vm86_regs * regs) +static inline void set_vflags_long(unsigned long flags, struct kernel_vm86_regs *regs) { set_flags(VEFLAGS, flags, current->thread.v86mask); set_flags(regs->pt.flags, flags, SAFE_MASK); - if (flags & IF_MASK) + if (flags & X86_EFLAGS_IF) set_IF(regs); else clear_IF(regs); } -static inline void set_vflags_short(unsigned short flags, struct kernel_vm86_regs * regs) +static inline void set_vflags_short(unsigned short flags, struct kernel_vm86_regs *regs) { set_flags(VFLAGS, flags, current->thread.v86mask); set_flags(regs->pt.flags, flags, SAFE_MASK); - if (flags & IF_MASK) + if (flags & X86_EFLAGS_IF) set_IF(regs); else clear_IF(regs); } -static inline unsigned long get_vflags(struct kernel_vm86_regs * regs) +static inline unsigned long get_vflags(struct kernel_vm86_regs *regs) { unsigned long flags = regs->pt.flags & RETURN_MASK; - if (VEFLAGS & VIF_MASK) - flags |= IF_MASK; - flags |= IOPL_MASK; + if (VEFLAGS & X86_EFLAGS_VIF) + flags |= X86_EFLAGS_IF; + flags |= X86_EFLAGS_IOPL; return flags | (VEFLAGS & current->thread.v86mask); } -static inline int is_revectored(int nr, struct revectored_struct * bitmap) +static inline int is_revectored(int nr, struct revectored_struct *bitmap) { __asm__ __volatile__("btl %2,%1\n\tsbbl %0,%0" :"=r" (nr) - :"m" (*bitmap),"r" (nr)); + :"m" (*bitmap), "r" (nr)); return nr; } @@ -437,7 +441,7 @@ static inline int is_revectored(int nr, struct revectored_struct * bitmap) ptr--; \ if (put_user(__val, base + ptr) < 0) \ goto err_label; \ - } while(0) + } while (0) #define pushw(base, ptr, val, err_label) \ do { \ @@ -448,7 +452,7 @@ static inline int is_revectored(int nr, struct revectored_struct * bitmap) ptr--; \ if (put_user(val_byte(__val, 0), base + ptr) < 0) \ goto err_label; \ - } while(0) + } while (0) #define pushl(base, ptr, val, err_label) \ do { \ @@ -465,7 +469,7 @@ static inline int is_revectored(int nr, struct revectored_struct * bitmap) ptr--; \ if (put_user(val_byte(__val, 0), base + ptr) < 0) \ goto err_label; \ - } while(0) + } while (0) #define popb(base, ptr, err_label) \ ({ \ @@ -512,7 +516,7 @@ static inline int is_revectored(int nr, struct revectored_struct * bitmap) * in userspace is always better than an Oops anyway.) [KD] */ static void do_int(struct kernel_vm86_regs *regs, int i, - unsigned char __user * ssp, unsigned short sp) + unsigned char __user *ssp, unsigned short sp) { unsigned long __user *intr_ptr; unsigned long segoffs; @@ -521,7 +525,7 @@ static void do_int(struct kernel_vm86_regs *regs, int i, goto cannot_handle; if (is_revectored(i, &KVM86->int_revectored)) goto cannot_handle; - if (i==0x21 && is_revectored(AH(regs),&KVM86->int21_revectored)) + if (i == 0x21 && is_revectored(AH(regs), &KVM86->int21_revectored)) goto cannot_handle; intr_ptr = (unsigned long __user *) (i << 2); if (get_user(segoffs, intr_ptr)) @@ -543,30 +547,23 @@ cannot_handle: return_to_32bit(regs, VM86_INTx + (i << 8)); } -int handle_vm86_trap(struct kernel_vm86_regs * regs, long error_code, int trapno) +int handle_vm86_trap(struct kernel_vm86_regs *regs, long error_code, int trapno) { if (VMPI.is_vm86pus) { - if ( (trapno==3) || (trapno==1) ) + if ((trapno == 3) || (trapno == 1)) return_to_32bit(regs, VM86_TRAP + (trapno << 8)); do_int(regs, trapno, (unsigned char __user *) (regs->pt.ss << 4), SP(regs)); return 0; } - if (trapno !=1) + if (trapno != 1) return 1; /* we let this handle by the calling routine */ - if (current->ptrace & PT_PTRACED) { - unsigned long flags; - spin_lock_irqsave(¤t->sighand->siglock, flags); - sigdelset(¤t->blocked, SIGTRAP); - recalc_sigpending(); - spin_unlock_irqrestore(¤t->sighand->siglock, flags); - } - send_sig(SIGTRAP, current, 1); current->thread.trap_no = trapno; current->thread.error_code = error_code; + force_sig(SIGTRAP, current); return 0; } -void handle_vm86_fault(struct kernel_vm86_regs * regs, long error_code) +void handle_vm86_fault(struct kernel_vm86_regs *regs, long error_code) { unsigned char opcode; unsigned char __user *csp; @@ -576,11 +573,11 @@ void handle_vm86_fault(struct kernel_vm86_regs * regs, long error_code) #define CHECK_IF_IN_TRAP \ if (VMPI.vm86dbg_active && VMPI.vm86dbg_TFpendig) \ - newflags |= TF_MASK + newflags |= X86_EFLAGS_TF #define VM86_FAULT_RETURN do { \ - if (VMPI.force_return_for_pic && (VEFLAGS & (IF_MASK | VIF_MASK))) \ + if (VMPI.force_return_for_pic && (VEFLAGS & (X86_EFLAGS_IF | X86_EFLAGS_VIF))) \ return_to_32bit(regs, VM86_PICRETURN); \ - if (orig_flags & TF_MASK) \ + if (orig_flags & X86_EFLAGS_TF) \ handle_vm86_trap(regs, 0, 1); \ return; } while (0) @@ -595,17 +592,17 @@ void handle_vm86_fault(struct kernel_vm86_regs * regs, long error_code) pref_done = 0; do { switch (opcode = popb(csp, ip, simulate_sigsegv)) { - case 0x66: /* 32-bit data */ data32=1; break; - case 0x67: /* 32-bit address */ break; - case 0x2e: /* CS */ break; - case 0x3e: /* DS */ break; - case 0x26: /* ES */ break; - case 0x36: /* SS */ break; - case 0x65: /* GS */ break; - case 0x64: /* FS */ break; - case 0xf2: /* repnz */ break; - case 0xf3: /* rep */ break; - default: pref_done = 1; + case 0x66: /* 32-bit data */ data32 = 1; break; + case 0x67: /* 32-bit address */ break; + case 0x2e: /* CS */ break; + case 0x3e: /* DS */ break; + case 0x26: /* ES */ break; + case 0x36: /* SS */ break; + case 0x65: /* GS */ break; + case 0x64: /* FS */ break; + case 0xf2: /* repnz */ break; + case 0xf3: /* rep */ break; + default: pref_done = 1; } } while (!pref_done); @@ -628,7 +625,7 @@ void handle_vm86_fault(struct kernel_vm86_regs * regs, long error_code) { unsigned long newflags; if (data32) { - newflags=popl(ssp, sp, simulate_sigsegv); + newflags = popl(ssp, sp, simulate_sigsegv); SP(regs) += 4; } else { newflags = popw(ssp, sp, simulate_sigsegv); @@ -636,20 +633,20 @@ void handle_vm86_fault(struct kernel_vm86_regs * regs, long error_code) } IP(regs) = ip; CHECK_IF_IN_TRAP; - if (data32) { + if (data32) set_vflags_long(newflags, regs); - } else { + else set_vflags_short(newflags, regs); - } + VM86_FAULT_RETURN; } /* int xx */ case 0xcd: { - int intno=popb(csp, ip, simulate_sigsegv); + int intno = popb(csp, ip, simulate_sigsegv); IP(regs) = ip; if (VMPI.vm86dbg_active) { - if ( (1 << (intno &7)) & VMPI.vm86dbg_intxxtab[intno >> 3] ) + if ((1 << (intno & 7)) & VMPI.vm86dbg_intxxtab[intno >> 3]) return_to_32bit(regs, VM86_INTx + (intno << 8)); } do_int(regs, intno, ssp, sp); @@ -663,9 +660,9 @@ void handle_vm86_fault(struct kernel_vm86_regs * regs, long error_code) unsigned long newcs; unsigned long newflags; if (data32) { - newip=popl(ssp, sp, simulate_sigsegv); - newcs=popl(ssp, sp, simulate_sigsegv); - newflags=popl(ssp, sp, simulate_sigsegv); + newip = popl(ssp, sp, simulate_sigsegv); + newcs = popl(ssp, sp, simulate_sigsegv); + newflags = popl(ssp, sp, simulate_sigsegv); SP(regs) += 12; } else { newip = popw(ssp, sp, simulate_sigsegv); @@ -734,18 +731,18 @@ static struct vm86_irqs { static DEFINE_SPINLOCK(irqbits_lock); static int irqbits; -#define ALLOWED_SIGS ( 1 /* 0 = don't send a signal */ \ +#define ALLOWED_SIGS (1 /* 0 = don't send a signal */ \ | (1 << SIGUSR1) | (1 << SIGUSR2) | (1 << SIGIO) | (1 << SIGURG) \ - | (1 << SIGUNUSED) ) - + | (1 << SIGUNUSED)) + static irqreturn_t irq_handler(int intno, void *dev_id) { int irq_bit; unsigned long flags; - spin_lock_irqsave(&irqbits_lock, flags); + spin_lock_irqsave(&irqbits_lock, flags); irq_bit = 1 << intno; - if ((irqbits & irq_bit) || ! vm86_irqs[intno].tsk) + if ((irqbits & irq_bit) || !vm86_irqs[intno].tsk) goto out; irqbits |= irq_bit; if (vm86_irqs[intno].sig) @@ -759,7 +756,7 @@ static irqreturn_t irq_handler(int intno, void *dev_id) return IRQ_HANDLED; out: - spin_unlock_irqrestore(&irqbits_lock, flags); + spin_unlock_irqrestore(&irqbits_lock, flags); return IRQ_NONE; } @@ -770,9 +767,9 @@ static inline void free_vm86_irq(int irqnumber) free_irq(irqnumber, NULL); vm86_irqs[irqnumber].tsk = NULL; - spin_lock_irqsave(&irqbits_lock, flags); + spin_lock_irqsave(&irqbits_lock, flags); irqbits &= ~(1 << irqnumber); - spin_unlock_irqrestore(&irqbits_lock, flags); + spin_unlock_irqrestore(&irqbits_lock, flags); } void release_vm86_irqs(struct task_struct *task) @@ -788,10 +785,10 @@ static inline int get_and_reset_irq(int irqnumber) int bit; unsigned long flags; int ret = 0; - + if (invalid_vm86_irq(irqnumber)) return 0; if (vm86_irqs[irqnumber].tsk != current) return 0; - spin_lock_irqsave(&irqbits_lock, flags); + spin_lock_irqsave(&irqbits_lock, flags); bit = irqbits & (1 << irqnumber); irqbits &= ~bit; if (bit) { @@ -799,7 +796,7 @@ static inline int get_and_reset_irq(int irqnumber) ret = 1; } - spin_unlock_irqrestore(&irqbits_lock, flags); + spin_unlock_irqrestore(&irqbits_lock, flags); return ret; } diff --git a/arch/x86/kernel/vmlinux_32.lds.S b/arch/x86/kernel/vmlinux_32.lds.S index 2ffa9656fe7a..ce5ed083a1e9 100644 --- a/arch/x86/kernel/vmlinux_32.lds.S +++ b/arch/x86/kernel/vmlinux_32.lds.S @@ -149,6 +149,11 @@ SECTIONS *(.con_initcall.init) __con_initcall_end = .; } + .x86cpuvendor.init : AT(ADDR(.x86cpuvendor.init) - LOAD_OFFSET) { + __x86cpuvendor_start = .; + *(.x86cpuvendor.init) + __x86cpuvendor_end = .; + } SECURITY_INIT . = ALIGN(4); .altinstructions : AT(ADDR(.altinstructions) - LOAD_OFFSET) { diff --git a/arch/x86/kernel/vmlinux_64.lds.S b/arch/x86/kernel/vmlinux_64.lds.S index fab132299735..b7ab3c335fae 100644 --- a/arch/x86/kernel/vmlinux_64.lds.S +++ b/arch/x86/kernel/vmlinux_64.lds.S @@ -177,6 +177,11 @@ SECTIONS *(.con_initcall.init) } __con_initcall_end = .; + __x86cpuvendor_start = .; + .x86cpuvendor.init : AT(ADDR(.x86cpuvendor.init) - LOAD_OFFSET) { + *(.x86cpuvendor.init) + } + __x86cpuvendor_end = .; SECURITY_INIT . = ALIGN(8); @@ -247,3 +252,9 @@ SECTIONS DWARF_DEBUG } + +/* + * Build-time check on the image size: + */ +ASSERT((_end - _text <= KERNEL_IMAGE_SIZE), + "kernel image bigger than KERNEL_IMAGE_SIZE") diff --git a/arch/x86/kernel/vsmp_64.c b/arch/x86/kernel/vsmp_64.c index d971210a6d36..caf2a26f5cfd 100644 --- a/arch/x86/kernel/vsmp_64.c +++ b/arch/x86/kernel/vsmp_64.c @@ -8,6 +8,8 @@ * * Ravikiran Thirumalai <kiran@scalemp.com>, * Shai Fultheim <shai@scalemp.com> + * Paravirt ops integration: Glauber de Oliveira Costa <gcosta@redhat.com>, + * Ravikiran Thirumalai <kiran@scalemp.com> */ #include <linux/init.h> @@ -15,38 +17,137 @@ #include <linux/pci_regs.h> #include <asm/pci-direct.h> #include <asm/io.h> +#include <asm/paravirt.h> -static int __init vsmp_init(void) +#if defined CONFIG_PCI && defined CONFIG_PARAVIRT +/* + * Interrupt control on vSMPowered systems: + * ~AC is a shadow of IF. If IF is 'on' AC should be 'off' + * and vice versa. + */ + +static unsigned long vsmp_save_fl(void) { - void *address; - unsigned int cap, ctl; + unsigned long flags = native_save_fl(); - if (!early_pci_allowed()) - return 0; + if (!(flags & X86_EFLAGS_IF) || (flags & X86_EFLAGS_AC)) + flags &= ~X86_EFLAGS_IF; + return flags; +} - /* Check if we are running on a ScaleMP vSMP box */ - if ((read_pci_config_16(0, 0x1f, 0, PCI_VENDOR_ID) != - PCI_VENDOR_ID_SCALEMP) || - (read_pci_config_16(0, 0x1f, 0, PCI_DEVICE_ID) != - PCI_DEVICE_ID_SCALEMP_VSMP_CTL)) - return 0; +static void vsmp_restore_fl(unsigned long flags) +{ + if (flags & X86_EFLAGS_IF) + flags &= ~X86_EFLAGS_AC; + else + flags |= X86_EFLAGS_AC; + native_restore_fl(flags); +} + +static void vsmp_irq_disable(void) +{ + unsigned long flags = native_save_fl(); + + native_restore_fl((flags & ~X86_EFLAGS_IF) | X86_EFLAGS_AC); +} + +static void vsmp_irq_enable(void) +{ + unsigned long flags = native_save_fl(); + + native_restore_fl((flags | X86_EFLAGS_IF) & (~X86_EFLAGS_AC)); +} + +static unsigned __init vsmp_patch(u8 type, u16 clobbers, void *ibuf, + unsigned long addr, unsigned len) +{ + switch (type) { + case PARAVIRT_PATCH(pv_irq_ops.irq_enable): + case PARAVIRT_PATCH(pv_irq_ops.irq_disable): + case PARAVIRT_PATCH(pv_irq_ops.save_fl): + case PARAVIRT_PATCH(pv_irq_ops.restore_fl): + return paravirt_patch_default(type, clobbers, ibuf, addr, len); + default: + return native_patch(type, clobbers, ibuf, addr, len); + } + +} + +static void __init set_vsmp_pv_ops(void) +{ + void *address; + unsigned int cap, ctl, cfg; /* set vSMP magic bits to indicate vSMP capable kernel */ - address = ioremap(read_pci_config(0, 0x1f, 0, PCI_BASE_ADDRESS_0), 8); + cfg = read_pci_config(0, 0x1f, 0, PCI_BASE_ADDRESS_0); + address = early_ioremap(cfg, 8); cap = readl(address); ctl = readl(address + 4); printk(KERN_INFO "vSMP CTL: capabilities:0x%08x control:0x%08x\n", cap, ctl); if (cap & ctl & (1 << 4)) { - /* Turn on vSMP IRQ fastpath handling (see system.h) */ + /* Setup irq ops and turn on vSMP IRQ fastpath handling */ + pv_irq_ops.irq_disable = vsmp_irq_disable; + pv_irq_ops.irq_enable = vsmp_irq_enable; + pv_irq_ops.save_fl = vsmp_save_fl; + pv_irq_ops.restore_fl = vsmp_restore_fl; + pv_init_ops.patch = vsmp_patch; + ctl &= ~(1 << 4); writel(ctl, address + 4); ctl = readl(address + 4); printk(KERN_INFO "vSMP CTL: control set to:0x%08x\n", ctl); } - iounmap(address); + early_iounmap(address, 8); +} +#else +static void __init set_vsmp_pv_ops(void) +{ +} +#endif + +#ifdef CONFIG_PCI +static int is_vsmp = -1; + +static void __init detect_vsmp_box(void) +{ + is_vsmp = 0; + + if (!early_pci_allowed()) + return; + + /* Check if we are running on a ScaleMP vSMPowered box */ + if (read_pci_config(0, 0x1f, 0, PCI_VENDOR_ID) == + (PCI_VENDOR_ID_SCALEMP | (PCI_DEVICE_ID_SCALEMP_VSMP_CTL << 16))) + is_vsmp = 1; +} + +int is_vsmp_box(void) +{ + if (is_vsmp != -1) + return is_vsmp; + else { + WARN_ON_ONCE(1); + return 0; + } +} +#else +static int __init detect_vsmp_box(void) +{ +} +int is_vsmp_box(void) +{ return 0; } +#endif -core_initcall(vsmp_init); +void __init vsmp_init(void) +{ + detect_vsmp_box(); + if (!is_vsmp_box()) + return; + + set_vsmp_pv_ops(); + return; +} diff --git a/arch/x86/kernel/x8664_ksyms_64.c b/arch/x86/kernel/x8664_ksyms_64.c index a66e9c1a0537..58882f9f2637 100644 --- a/arch/x86/kernel/x8664_ksyms_64.c +++ b/arch/x86/kernel/x8664_ksyms_64.c @@ -4,7 +4,6 @@ #include <linux/module.h> #include <linux/smp.h> -#include <asm/semaphore.h> #include <asm/processor.h> #include <asm/uaccess.h> #include <asm/pgtable.h> @@ -12,11 +11,6 @@ EXPORT_SYMBOL(kernel_thread); -EXPORT_SYMBOL(__down_failed); -EXPORT_SYMBOL(__down_failed_interruptible); -EXPORT_SYMBOL(__down_failed_trylock); -EXPORT_SYMBOL(__up_wakeup); - EXPORT_SYMBOL(__get_user_1); EXPORT_SYMBOL(__get_user_2); EXPORT_SYMBOL(__get_user_4); @@ -35,15 +29,17 @@ EXPORT_SYMBOL(__copy_from_user_inatomic); EXPORT_SYMBOL(copy_page); EXPORT_SYMBOL(clear_page); -/* Export string functions. We normally rely on gcc builtin for most of these, - but gcc sometimes decides not to inline them. */ +/* + * Export string functions. We normally rely on gcc builtin for most of these, + * but gcc sometimes decides not to inline them. + */ #undef memcpy #undef memset #undef memmove -extern void * memset(void *,int,__kernel_size_t); -extern void * memcpy(void *,const void *,__kernel_size_t); -extern void * __memcpy(void *,const void *,__kernel_size_t); +extern void *memset(void *, int, __kernel_size_t); +extern void *memcpy(void *, const void *, __kernel_size_t); +extern void *__memcpy(void *, const void *, __kernel_size_t); EXPORT_SYMBOL(memset); EXPORT_SYMBOL(memcpy); diff --git a/arch/x86/lguest/boot.c b/arch/x86/lguest/boot.c index 3335b4595efd..af65b2da3ba0 100644 --- a/arch/x86/lguest/boot.c +++ b/arch/x86/lguest/boot.c @@ -661,7 +661,7 @@ static int lguest_clockevent_set_next_event(unsigned long delta, if (delta < LG_CLOCK_MIN_DELTA) { if (printk_ratelimit()) printk(KERN_DEBUG "%s: small delta %lu ns\n", - __FUNCTION__, delta); + __func__, delta); return -ETIME; } diff --git a/arch/x86/lib/memcpy_32.c b/arch/x86/lib/memcpy_32.c index 37756b6fb329..5415a9d06f53 100644 --- a/arch/x86/lib/memcpy_32.c +++ b/arch/x86/lib/memcpy_32.c @@ -25,7 +25,7 @@ void *memmove(void *dest, const void *src, size_t n) int d0, d1, d2; if (dest < src) { - memcpy(dest,src,n); + memcpy(dest, src, n); } else { __asm__ __volatile__( "std\n\t" diff --git a/arch/x86/lib/memmove_64.c b/arch/x86/lib/memmove_64.c index 80175e47b190..0a33909bf122 100644 --- a/arch/x86/lib/memmove_64.c +++ b/arch/x86/lib/memmove_64.c @@ -6,10 +6,10 @@ #include <linux/module.h> #undef memmove -void *memmove(void * dest,const void *src,size_t count) +void *memmove(void *dest, const void *src, size_t count) { - if (dest < src) { - return memcpy(dest,src,count); + if (dest < src) { + return memcpy(dest, src, count); } else { char *p = dest + count; const char *s = src + count; @@ -17,5 +17,5 @@ void *memmove(void * dest,const void *src,size_t count) *--p = *--s; } return dest; -} +} EXPORT_SYMBOL(memmove); diff --git a/arch/x86/lib/mmx_32.c b/arch/x86/lib/mmx_32.c index cc9b4a4450f3..c9f2d9ba8dd8 100644 --- a/arch/x86/lib/mmx_32.c +++ b/arch/x86/lib/mmx_32.c @@ -1,32 +1,30 @@ -#include <linux/types.h> -#include <linux/string.h> -#include <linux/sched.h> -#include <linux/hardirq.h> -#include <linux/module.h> - -#include <asm/asm.h> -#include <asm/i387.h> - - /* * MMX 3DNow! library helper functions * * To do: - * We can use MMX just for prefetch in IRQ's. This may be a win. + * We can use MMX just for prefetch in IRQ's. This may be a win. * (reported so on K6-III) * We should use a better code neutral filler for the short jump * leal ebx. [ebx] is apparently best for K6-2, but Cyrix ?? * We also want to clobber the filler register so we don't get any - * register forwarding stalls on the filler. + * register forwarding stalls on the filler. * * Add *user handling. Checksums are not a win with MMX on any CPU * tested so far for any MMX solution figured. * - * 22/09/2000 - Arjan van de Ven - * Improved for non-egineering-sample Athlons + * 22/09/2000 - Arjan van de Ven + * Improved for non-egineering-sample Athlons * */ - +#include <linux/hardirq.h> +#include <linux/string.h> +#include <linux/module.h> +#include <linux/sched.h> +#include <linux/types.h> + +#include <asm/i387.h> +#include <asm/asm.h> + void *_mmx_memcpy(void *to, const void *from, size_t len) { void *p; @@ -51,12 +49,10 @@ void *_mmx_memcpy(void *to, const void *from, size_t len) "3: movw $0x1AEB, 1b\n" /* jmp on 26 bytes */ " jmp 2b\n" ".previous\n" - _ASM_EXTABLE(1b,3b) - : : "r" (from) ); - - - for(; i>5; i--) - { + _ASM_EXTABLE(1b, 3b) + : : "r" (from)); + + for ( ; i > 5; i--) { __asm__ __volatile__ ( "1: prefetch 320(%0)\n" "2: movq (%0), %%mm0\n" @@ -79,14 +75,14 @@ void *_mmx_memcpy(void *to, const void *from, size_t len) "3: movw $0x05EB, 1b\n" /* jmp on 5 bytes */ " jmp 2b\n" ".previous\n" - _ASM_EXTABLE(1b,3b) - : : "r" (from), "r" (to) : "memory"); - from+=64; - to+=64; + _ASM_EXTABLE(1b, 3b) + : : "r" (from), "r" (to) : "memory"); + + from += 64; + to += 64; } - for(; i>0; i--) - { + for ( ; i > 0; i--) { __asm__ __volatile__ ( " movq (%0), %%mm0\n" " movq 8(%0), %%mm1\n" @@ -104,17 +100,20 @@ void *_mmx_memcpy(void *to, const void *from, size_t len) " movq %%mm1, 40(%1)\n" " movq %%mm2, 48(%1)\n" " movq %%mm3, 56(%1)\n" - : : "r" (from), "r" (to) : "memory"); - from+=64; - to+=64; + : : "r" (from), "r" (to) : "memory"); + + from += 64; + to += 64; } /* - * Now do the tail of the block + * Now do the tail of the block: */ - __memcpy(to, from, len&63); + __memcpy(to, from, len & 63); kernel_fpu_end(); + return p; } +EXPORT_SYMBOL(_mmx_memcpy); #ifdef CONFIG_MK7 @@ -128,13 +127,12 @@ static void fast_clear_page(void *page) int i; kernel_fpu_begin(); - + __asm__ __volatile__ ( " pxor %%mm0, %%mm0\n" : : ); - for(i=0;i<4096/64;i++) - { + for (i = 0; i < 4096/64; i++) { __asm__ __volatile__ ( " movntq %%mm0, (%0)\n" " movntq %%mm0, 8(%0)\n" @@ -145,14 +143,15 @@ static void fast_clear_page(void *page) " movntq %%mm0, 48(%0)\n" " movntq %%mm0, 56(%0)\n" : : "r" (page) : "memory"); - page+=64; + page += 64; } - /* since movntq is weakly-ordered, a "sfence" is needed to become - * ordered again. + + /* + * Since movntq is weakly-ordered, a "sfence" is needed to become + * ordered again: */ - __asm__ __volatile__ ( - " sfence \n" : : - ); + __asm__ __volatile__("sfence\n"::); + kernel_fpu_end(); } @@ -162,10 +161,11 @@ static void fast_copy_page(void *to, void *from) kernel_fpu_begin(); - /* maybe the prefetch stuff can go before the expensive fnsave... + /* + * maybe the prefetch stuff can go before the expensive fnsave... * but that is for later. -AV */ - __asm__ __volatile__ ( + __asm__ __volatile__( "1: prefetch (%0)\n" " prefetch 64(%0)\n" " prefetch 128(%0)\n" @@ -176,11 +176,9 @@ static void fast_copy_page(void *to, void *from) "3: movw $0x1AEB, 1b\n" /* jmp on 26 bytes */ " jmp 2b\n" ".previous\n" - _ASM_EXTABLE(1b,3b) - : : "r" (from) ); + _ASM_EXTABLE(1b, 3b) : : "r" (from)); - for(i=0; i<(4096-320)/64; i++) - { + for (i = 0; i < (4096-320)/64; i++) { __asm__ __volatile__ ( "1: prefetch 320(%0)\n" "2: movq (%0), %%mm0\n" @@ -203,13 +201,13 @@ static void fast_copy_page(void *to, void *from) "3: movw $0x05EB, 1b\n" /* jmp on 5 bytes */ " jmp 2b\n" ".previous\n" - _ASM_EXTABLE(1b,3b) - : : "r" (from), "r" (to) : "memory"); - from+=64; - to+=64; + _ASM_EXTABLE(1b, 3b) : : "r" (from), "r" (to) : "memory"); + + from += 64; + to += 64; } - for(i=(4096-320)/64; i<4096/64; i++) - { + + for (i = (4096-320)/64; i < 4096/64; i++) { __asm__ __volatile__ ( "2: movq (%0), %%mm0\n" " movntq %%mm0, (%1)\n" @@ -227,37 +225,34 @@ static void fast_copy_page(void *to, void *from) " movntq %%mm6, 48(%1)\n" " movq 56(%0), %%mm7\n" " movntq %%mm7, 56(%1)\n" - : : "r" (from), "r" (to) : "memory"); - from+=64; - to+=64; + : : "r" (from), "r" (to) : "memory"); + from += 64; + to += 64; } - /* since movntq is weakly-ordered, a "sfence" is needed to become - * ordered again. + /* + * Since movntq is weakly-ordered, a "sfence" is needed to become + * ordered again: */ - __asm__ __volatile__ ( - " sfence \n" : : - ); + __asm__ __volatile__("sfence \n"::); kernel_fpu_end(); } -#else +#else /* CONFIG_MK7 */ /* * Generic MMX implementation without K7 specific streaming */ - static void fast_clear_page(void *page) { int i; - + kernel_fpu_begin(); - + __asm__ __volatile__ ( " pxor %%mm0, %%mm0\n" : : ); - for(i=0;i<4096/128;i++) - { + for (i = 0; i < 4096/128; i++) { __asm__ __volatile__ ( " movq %%mm0, (%0)\n" " movq %%mm0, 8(%0)\n" @@ -275,8 +270,8 @@ static void fast_clear_page(void *page) " movq %%mm0, 104(%0)\n" " movq %%mm0, 112(%0)\n" " movq %%mm0, 120(%0)\n" - : : "r" (page) : "memory"); - page+=128; + : : "r" (page) : "memory"); + page += 128; } kernel_fpu_end(); @@ -285,8 +280,7 @@ static void fast_clear_page(void *page) static void fast_copy_page(void *to, void *from) { int i; - - + kernel_fpu_begin(); __asm__ __volatile__ ( @@ -300,11 +294,9 @@ static void fast_copy_page(void *to, void *from) "3: movw $0x1AEB, 1b\n" /* jmp on 26 bytes */ " jmp 2b\n" ".previous\n" - _ASM_EXTABLE(1b,3b) - : : "r" (from) ); + _ASM_EXTABLE(1b, 3b) : : "r" (from)); - for(i=0; i<4096/64; i++) - { + for (i = 0; i < 4096/64; i++) { __asm__ __volatile__ ( "1: prefetch 320(%0)\n" "2: movq (%0), %%mm0\n" @@ -327,60 +319,59 @@ static void fast_copy_page(void *to, void *from) "3: movw $0x05EB, 1b\n" /* jmp on 5 bytes */ " jmp 2b\n" ".previous\n" - _ASM_EXTABLE(1b,3b) - : : "r" (from), "r" (to) : "memory"); - from+=64; - to+=64; + _ASM_EXTABLE(1b, 3b) + : : "r" (from), "r" (to) : "memory"); + + from += 64; + to += 64; } kernel_fpu_end(); } - -#endif +#endif /* !CONFIG_MK7 */ /* - * Favour MMX for page clear and copy. + * Favour MMX for page clear and copy: */ - -static void slow_zero_page(void * page) +static void slow_zero_page(void *page) { int d0, d1; - __asm__ __volatile__( \ - "cld\n\t" \ - "rep ; stosl" \ - : "=&c" (d0), "=&D" (d1) - :"a" (0),"1" (page),"0" (1024) - :"memory"); + + __asm__ __volatile__( + "cld\n\t" + "rep ; stosl" + + : "=&c" (d0), "=&D" (d1) + :"a" (0), "1" (page), "0" (1024) + :"memory"); } - -void mmx_clear_page(void * page) + +void mmx_clear_page(void *page) { - if(unlikely(in_interrupt())) + if (unlikely(in_interrupt())) slow_zero_page(page); else fast_clear_page(page); } +EXPORT_SYMBOL(mmx_clear_page); static void slow_copy_page(void *to, void *from) { int d0, d1, d2; - __asm__ __volatile__( \ - "cld\n\t" \ - "rep ; movsl" \ - : "=&c" (d0), "=&D" (d1), "=&S" (d2) \ - : "0" (1024),"1" ((long) to),"2" ((long) from) \ + + __asm__ __volatile__( + "cld\n\t" + "rep ; movsl" + : "=&c" (d0), "=&D" (d1), "=&S" (d2) + : "0" (1024), "1" ((long) to), "2" ((long) from) : "memory"); } - void mmx_copy_page(void *to, void *from) { - if(unlikely(in_interrupt())) + if (unlikely(in_interrupt())) slow_copy_page(to, from); else fast_copy_page(to, from); } - -EXPORT_SYMBOL(_mmx_memcpy); -EXPORT_SYMBOL(mmx_clear_page); EXPORT_SYMBOL(mmx_copy_page); diff --git a/arch/x86/lib/semaphore_32.S b/arch/x86/lib/semaphore_32.S index 3899bd37fdf0..648fe4741782 100644 --- a/arch/x86/lib/semaphore_32.S +++ b/arch/x86/lib/semaphore_32.S @@ -30,89 +30,6 @@ * value or just clobbered.. */ .section .sched.text, "ax" -ENTRY(__down_failed) - CFI_STARTPROC - FRAME - pushl %edx - CFI_ADJUST_CFA_OFFSET 4 - CFI_REL_OFFSET edx,0 - pushl %ecx - CFI_ADJUST_CFA_OFFSET 4 - CFI_REL_OFFSET ecx,0 - call __down - popl %ecx - CFI_ADJUST_CFA_OFFSET -4 - CFI_RESTORE ecx - popl %edx - CFI_ADJUST_CFA_OFFSET -4 - CFI_RESTORE edx - ENDFRAME - ret - CFI_ENDPROC - ENDPROC(__down_failed) - -ENTRY(__down_failed_interruptible) - CFI_STARTPROC - FRAME - pushl %edx - CFI_ADJUST_CFA_OFFSET 4 - CFI_REL_OFFSET edx,0 - pushl %ecx - CFI_ADJUST_CFA_OFFSET 4 - CFI_REL_OFFSET ecx,0 - call __down_interruptible - popl %ecx - CFI_ADJUST_CFA_OFFSET -4 - CFI_RESTORE ecx - popl %edx - CFI_ADJUST_CFA_OFFSET -4 - CFI_RESTORE edx - ENDFRAME - ret - CFI_ENDPROC - ENDPROC(__down_failed_interruptible) - -ENTRY(__down_failed_trylock) - CFI_STARTPROC - FRAME - pushl %edx - CFI_ADJUST_CFA_OFFSET 4 - CFI_REL_OFFSET edx,0 - pushl %ecx - CFI_ADJUST_CFA_OFFSET 4 - CFI_REL_OFFSET ecx,0 - call __down_trylock - popl %ecx - CFI_ADJUST_CFA_OFFSET -4 - CFI_RESTORE ecx - popl %edx - CFI_ADJUST_CFA_OFFSET -4 - CFI_RESTORE edx - ENDFRAME - ret - CFI_ENDPROC - ENDPROC(__down_failed_trylock) - -ENTRY(__up_wakeup) - CFI_STARTPROC - FRAME - pushl %edx - CFI_ADJUST_CFA_OFFSET 4 - CFI_REL_OFFSET edx,0 - pushl %ecx - CFI_ADJUST_CFA_OFFSET 4 - CFI_REL_OFFSET ecx,0 - call __up - popl %ecx - CFI_ADJUST_CFA_OFFSET -4 - CFI_RESTORE ecx - popl %edx - CFI_ADJUST_CFA_OFFSET -4 - CFI_RESTORE edx - ENDFRAME - ret - CFI_ENDPROC - ENDPROC(__up_wakeup) /* * rw spinlock fallbacks diff --git a/arch/x86/lib/string_32.c b/arch/x86/lib/string_32.c index c2c0504a3071..94972e7c094d 100644 --- a/arch/x86/lib/string_32.c +++ b/arch/x86/lib/string_32.c @@ -14,25 +14,25 @@ #include <linux/module.h> #ifdef __HAVE_ARCH_STRCPY -char *strcpy(char * dest,const char *src) +char *strcpy(char *dest, const char *src) { int d0, d1, d2; - asm volatile( "1:\tlodsb\n\t" + asm volatile("1:\tlodsb\n\t" "stosb\n\t" "testb %%al,%%al\n\t" "jne 1b" : "=&S" (d0), "=&D" (d1), "=&a" (d2) - :"0" (src),"1" (dest) : "memory"); + :"0" (src), "1" (dest) : "memory"); return dest; } EXPORT_SYMBOL(strcpy); #endif #ifdef __HAVE_ARCH_STRNCPY -char *strncpy(char * dest,const char *src,size_t count) +char *strncpy(char *dest, const char *src, size_t count) { int d0, d1, d2, d3; - asm volatile( "1:\tdecl %2\n\t" + asm volatile("1:\tdecl %2\n\t" "js 2f\n\t" "lodsb\n\t" "stosb\n\t" @@ -42,17 +42,17 @@ char *strncpy(char * dest,const char *src,size_t count) "stosb\n" "2:" : "=&S" (d0), "=&D" (d1), "=&c" (d2), "=&a" (d3) - :"0" (src),"1" (dest),"2" (count) : "memory"); + :"0" (src), "1" (dest), "2" (count) : "memory"); return dest; } EXPORT_SYMBOL(strncpy); #endif #ifdef __HAVE_ARCH_STRCAT -char *strcat(char * dest,const char * src) +char *strcat(char *dest, const char *src) { int d0, d1, d2, d3; - asm volatile( "repne\n\t" + asm volatile("repne\n\t" "scasb\n\t" "decl %1\n" "1:\tlodsb\n\t" @@ -67,10 +67,10 @@ EXPORT_SYMBOL(strcat); #endif #ifdef __HAVE_ARCH_STRNCAT -char *strncat(char * dest,const char * src,size_t count) +char *strncat(char *dest, const char *src, size_t count) { int d0, d1, d2, d3; - asm volatile( "repne\n\t" + asm volatile("repne\n\t" "scasb\n\t" "decl %1\n\t" "movl %8,%3\n" @@ -83,7 +83,7 @@ char *strncat(char * dest,const char * src,size_t count) "2:\txorl %2,%2\n\t" "stosb" : "=&S" (d0), "=&D" (d1), "=&a" (d2), "=&c" (d3) - : "0" (src),"1" (dest),"2" (0),"3" (0xffffffffu), "g" (count) + : "0" (src), "1" (dest), "2" (0), "3" (0xffffffffu), "g" (count) : "memory"); return dest; } @@ -91,11 +91,11 @@ EXPORT_SYMBOL(strncat); #endif #ifdef __HAVE_ARCH_STRCMP -int strcmp(const char * cs,const char * ct) +int strcmp(const char *cs, const char *ct) { int d0, d1; int res; - asm volatile( "1:\tlodsb\n\t" + asm volatile("1:\tlodsb\n\t" "scasb\n\t" "jne 2f\n\t" "testb %%al,%%al\n\t" @@ -106,7 +106,7 @@ int strcmp(const char * cs,const char * ct) "orb $1,%%al\n" "3:" :"=a" (res), "=&S" (d0), "=&D" (d1) - :"1" (cs),"2" (ct) + :"1" (cs), "2" (ct) :"memory"); return res; } @@ -114,11 +114,11 @@ EXPORT_SYMBOL(strcmp); #endif #ifdef __HAVE_ARCH_STRNCMP -int strncmp(const char * cs,const char * ct,size_t count) +int strncmp(const char *cs, const char *ct, size_t count) { int res; int d0, d1, d2; - asm volatile( "1:\tdecl %3\n\t" + asm volatile("1:\tdecl %3\n\t" "js 2f\n\t" "lodsb\n\t" "scasb\n\t" @@ -131,7 +131,7 @@ int strncmp(const char * cs,const char * ct,size_t count) "orb $1,%%al\n" "4:" :"=a" (res), "=&S" (d0), "=&D" (d1), "=&c" (d2) - :"1" (cs),"2" (ct),"3" (count) + :"1" (cs), "2" (ct), "3" (count) :"memory"); return res; } @@ -139,11 +139,11 @@ EXPORT_SYMBOL(strncmp); #endif #ifdef __HAVE_ARCH_STRCHR -char *strchr(const char * s, int c) +char *strchr(const char *s, int c) { int d0; - char * res; - asm volatile( "movb %%al,%%ah\n" + char *res; + asm volatile("movb %%al,%%ah\n" "1:\tlodsb\n\t" "cmpb %%ah,%%al\n\t" "je 2f\n\t" @@ -153,7 +153,7 @@ char *strchr(const char * s, int c) "2:\tmovl %1,%0\n\t" "decl %0" :"=a" (res), "=&S" (d0) - :"1" (s),"0" (c) + :"1" (s), "0" (c) :"memory"); return res; } @@ -161,16 +161,16 @@ EXPORT_SYMBOL(strchr); #endif #ifdef __HAVE_ARCH_STRLEN -size_t strlen(const char * s) +size_t strlen(const char *s) { int d0; int res; - asm volatile( "repne\n\t" + asm volatile("repne\n\t" "scasb\n\t" "notl %0\n\t" "decl %0" :"=c" (res), "=&D" (d0) - :"1" (s),"a" (0), "0" (0xffffffffu) + :"1" (s), "a" (0), "0" (0xffffffffu) :"memory"); return res; } @@ -178,19 +178,19 @@ EXPORT_SYMBOL(strlen); #endif #ifdef __HAVE_ARCH_MEMCHR -void *memchr(const void *cs,int c,size_t count) +void *memchr(const void *cs, int c, size_t count) { int d0; void *res; if (!count) return NULL; - asm volatile( "repne\n\t" + asm volatile("repne\n\t" "scasb\n\t" "je 1f\n\t" "movl $1,%0\n" "1:\tdecl %0" :"=D" (res), "=&c" (d0) - :"a" (c),"0" (cs),"1" (count) + :"a" (c), "0" (cs), "1" (count) :"memory"); return res; } @@ -198,7 +198,7 @@ EXPORT_SYMBOL(memchr); #endif #ifdef __HAVE_ARCH_MEMSCAN -void *memscan(void * addr, int c, size_t size) +void *memscan(void *addr, int c, size_t size) { if (!size) return addr; @@ -219,7 +219,7 @@ size_t strnlen(const char *s, size_t count) { int d0; int res; - asm volatile( "movl %2,%0\n\t" + asm volatile("movl %2,%0\n\t" "jmp 2f\n" "1:\tcmpb $0,(%0)\n\t" "je 3f\n\t" @@ -229,7 +229,7 @@ size_t strnlen(const char *s, size_t count) "jne 1b\n" "3:\tsubl %2,%0" :"=a" (res), "=&d" (d0) - :"c" (s),"1" (count) + :"c" (s), "1" (count) :"memory"); return res; } diff --git a/arch/x86/lib/strstr_32.c b/arch/x86/lib/strstr_32.c index a3dafbf59dae..42e8a50303f3 100644 --- a/arch/x86/lib/strstr_32.c +++ b/arch/x86/lib/strstr_32.c @@ -1,9 +1,9 @@ #include <linux/string.h> -char * strstr(const char * cs,const char * ct) +char *strstr(const char *cs, const char *ct) { int d0, d1; -register char * __res; +register char *__res; __asm__ __volatile__( "movl %6,%%edi\n\t" "repne\n\t" diff --git a/arch/x86/lib/thunk_64.S b/arch/x86/lib/thunk_64.S index 8b92d428ab02..e009251d4e9f 100644 --- a/arch/x86/lib/thunk_64.S +++ b/arch/x86/lib/thunk_64.S @@ -41,11 +41,6 @@ thunk rwsem_downgrade_thunk,rwsem_downgrade_wake #endif - thunk __down_failed,__down - thunk_retrax __down_failed_interruptible,__down_interruptible - thunk_retrax __down_failed_trylock,__down_trylock - thunk __up_wakeup,__up - #ifdef CONFIG_TRACE_IRQFLAGS thunk trace_hardirqs_on_thunk,trace_hardirqs_on thunk trace_hardirqs_off_thunk,trace_hardirqs_off diff --git a/arch/x86/lib/usercopy_32.c b/arch/x86/lib/usercopy_32.c index e849b9998b0e..24e60944971a 100644 --- a/arch/x86/lib/usercopy_32.c +++ b/arch/x86/lib/usercopy_32.c @@ -1,4 +1,4 @@ -/* +/* * User address space access functions. * The non inlined parts of asm-i386/uaccess.h are here. * @@ -22,14 +22,14 @@ static inline int __movsl_is_ok(unsigned long a1, unsigned long a2, unsigned lon #endif return 1; } -#define movsl_is_ok(a1,a2,n) \ - __movsl_is_ok((unsigned long)(a1),(unsigned long)(a2),(n)) +#define movsl_is_ok(a1, a2, n) \ + __movsl_is_ok((unsigned long)(a1), (unsigned long)(a2), (n)) /* * Copy a null terminated string from userspace. */ -#define __do_strncpy_from_user(dst,src,count,res) \ +#define __do_strncpy_from_user(dst, src, count, res) \ do { \ int __d0, __d1, __d2; \ might_sleep(); \ @@ -61,7 +61,7 @@ do { \ * least @count bytes long. * @src: Source address, in user space. * @count: Maximum number of bytes to copy, including the trailing NUL. - * + * * Copies a NUL-terminated string from userspace to kernel space. * Caller must check the specified block with access_ok() before calling * this function. @@ -90,7 +90,7 @@ EXPORT_SYMBOL(__strncpy_from_user); * least @count bytes long. * @src: Source address, in user space. * @count: Maximum number of bytes to copy, including the trailing NUL. - * + * * Copies a NUL-terminated string from userspace to kernel space. * * On success, returns the length of the string (not including the trailing @@ -120,7 +120,7 @@ EXPORT_SYMBOL(strncpy_from_user); do { \ int __d0; \ might_sleep(); \ - __asm__ __volatile__( \ + __asm__ __volatile__( \ "0: rep; stosl\n" \ " movl %2,%0\n" \ "1: rep; stosb\n" \ @@ -333,17 +333,17 @@ __copy_user_zeroing_intel(void *to, const void __user *from, unsigned long size) __asm__ __volatile__( " .align 2,0x90\n" "0: movl 32(%4), %%eax\n" - " cmpl $67, %0\n" - " jbe 2f\n" + " cmpl $67, %0\n" + " jbe 2f\n" "1: movl 64(%4), %%eax\n" - " .align 2,0x90\n" - "2: movl 0(%4), %%eax\n" - "21: movl 4(%4), %%edx\n" - " movl %%eax, 0(%3)\n" - " movl %%edx, 4(%3)\n" - "3: movl 8(%4), %%eax\n" - "31: movl 12(%4),%%edx\n" - " movl %%eax, 8(%3)\n" + " .align 2,0x90\n" + "2: movl 0(%4), %%eax\n" + "21: movl 4(%4), %%edx\n" + " movl %%eax, 0(%3)\n" + " movl %%edx, 4(%3)\n" + "3: movl 8(%4), %%eax\n" + "31: movl 12(%4),%%edx\n" + " movl %%eax, 8(%3)\n" " movl %%edx, 12(%3)\n" "4: movl 16(%4), %%eax\n" "41: movl 20(%4), %%edx\n" @@ -369,38 +369,38 @@ __copy_user_zeroing_intel(void *to, const void __user *from, unsigned long size) "91: movl 60(%4), %%edx\n" " movl %%eax, 56(%3)\n" " movl %%edx, 60(%3)\n" - " addl $-64, %0\n" - " addl $64, %4\n" - " addl $64, %3\n" - " cmpl $63, %0\n" - " ja 0b\n" - "5: movl %0, %%eax\n" - " shrl $2, %0\n" - " andl $3, %%eax\n" - " cld\n" - "6: rep; movsl\n" + " addl $-64, %0\n" + " addl $64, %4\n" + " addl $64, %3\n" + " cmpl $63, %0\n" + " ja 0b\n" + "5: movl %0, %%eax\n" + " shrl $2, %0\n" + " andl $3, %%eax\n" + " cld\n" + "6: rep; movsl\n" " movl %%eax,%0\n" - "7: rep; movsb\n" - "8:\n" + "7: rep; movsb\n" + "8:\n" ".section .fixup,\"ax\"\n" - "9: lea 0(%%eax,%0,4),%0\n" - "16: pushl %0\n" - " pushl %%eax\n" + "9: lea 0(%%eax,%0,4),%0\n" + "16: pushl %0\n" + " pushl %%eax\n" " xorl %%eax,%%eax\n" - " rep; stosb\n" - " popl %%eax\n" - " popl %0\n" - " jmp 8b\n" - ".previous\n" + " rep; stosb\n" + " popl %%eax\n" + " popl %0\n" + " jmp 8b\n" + ".previous\n" ".section __ex_table,\"a\"\n" - " .align 4\n" - " .long 0b,16b\n" + " .align 4\n" + " .long 0b,16b\n" " .long 1b,16b\n" " .long 2b,16b\n" " .long 21b,16b\n" - " .long 3b,16b\n" + " .long 3b,16b\n" " .long 31b,16b\n" - " .long 4b,16b\n" + " .long 4b,16b\n" " .long 41b,16b\n" " .long 10b,16b\n" " .long 51b,16b\n" @@ -412,9 +412,9 @@ __copy_user_zeroing_intel(void *to, const void __user *from, unsigned long size) " .long 81b,16b\n" " .long 14b,16b\n" " .long 91b,16b\n" - " .long 6b,9b\n" - " .long 7b,16b\n" - ".previous" + " .long 6b,9b\n" + " .long 7b,16b\n" + ".previous" : "=&c"(size), "=&D" (d0), "=&S" (d1) : "1"(to), "2"(from), "0"(size) : "eax", "edx", "memory"); @@ -429,7 +429,7 @@ __copy_user_zeroing_intel(void *to, const void __user *from, unsigned long size) static unsigned long __copy_user_zeroing_intel_nocache(void *to, const void __user *from, unsigned long size) { - int d0, d1; + int d0, d1; __asm__ __volatile__( " .align 2,0x90\n" @@ -526,7 +526,7 @@ static unsigned long __copy_user_zeroing_intel_nocache(void *to, static unsigned long __copy_user_intel_nocache(void *to, const void __user *from, unsigned long size) { - int d0, d1; + int d0, d1; __asm__ __volatile__( " .align 2,0x90\n" @@ -629,7 +629,7 @@ unsigned long __copy_user_zeroing_intel_nocache(void *to, #endif /* CONFIG_X86_INTEL_USERCOPY */ /* Generic arbitrary sized copy. */ -#define __copy_user(to,from,size) \ +#define __copy_user(to, from, size) \ do { \ int __d0, __d1, __d2; \ __asm__ __volatile__( \ @@ -665,7 +665,7 @@ do { \ : "memory"); \ } while (0) -#define __copy_user_zeroing(to,from,size) \ +#define __copy_user_zeroing(to, from, size) \ do { \ int __d0, __d1, __d2; \ __asm__ __volatile__( \ @@ -712,7 +712,7 @@ unsigned long __copy_to_user_ll(void __user *to, const void *from, { #ifndef CONFIG_X86_WP_WORKS_OK if (unlikely(boot_cpu_data.wp_works_ok == 0) && - ((unsigned long )to) < TASK_SIZE) { + ((unsigned long)to) < TASK_SIZE) { /* * When we are in an atomic section (see * mm/filemap.c:file_read_actor), return the full @@ -721,26 +721,26 @@ unsigned long __copy_to_user_ll(void __user *to, const void *from, if (in_atomic()) return n; - /* + /* * CPU does not honor the WP bit when writing * from supervisory mode, and due to preemption or SMP, * the page tables can change at any time. * Do it manually. Manfred <manfred@colorfullife.com> */ while (n) { - unsigned long offset = ((unsigned long)to)%PAGE_SIZE; + unsigned long offset = ((unsigned long)to)%PAGE_SIZE; unsigned long len = PAGE_SIZE - offset; int retval; struct page *pg; void *maddr; - + if (len > n) len = n; survive: down_read(¤t->mm->mmap_sem); retval = get_user_pages(current, current->mm, - (unsigned long )to, 1, 1, 0, &pg, NULL); + (unsigned long)to, 1, 1, 0, &pg, NULL); if (retval == -ENOMEM && is_global_init(current)) { up_read(¤t->mm->mmap_sem); @@ -750,8 +750,8 @@ survive: if (retval != 1) { up_read(¤t->mm->mmap_sem); - break; - } + break; + } maddr = kmap_atomic(pg, KM_USER0); memcpy(maddr + offset, from, len); @@ -802,12 +802,12 @@ unsigned long __copy_from_user_ll_nocache(void *to, const void __user *from, unsigned long n) { #ifdef CONFIG_X86_INTEL_USERCOPY - if ( n > 64 && cpu_has_xmm2) - n = __copy_user_zeroing_intel_nocache(to, from, n); + if (n > 64 && cpu_has_xmm2) + n = __copy_user_zeroing_intel_nocache(to, from, n); else __copy_user_zeroing(to, from, n); #else - __copy_user_zeroing(to, from, n); + __copy_user_zeroing(to, from, n); #endif return n; } @@ -817,12 +817,12 @@ unsigned long __copy_from_user_ll_nocache_nozero(void *to, const void __user *fr unsigned long n) { #ifdef CONFIG_X86_INTEL_USERCOPY - if ( n > 64 && cpu_has_xmm2) - n = __copy_user_intel_nocache(to, from, n); + if (n > 64 && cpu_has_xmm2) + n = __copy_user_intel_nocache(to, from, n); else __copy_user(to, from, n); #else - __copy_user(to, from, n); + __copy_user(to, from, n); #endif return n; } diff --git a/arch/x86/mach-generic/bigsmp.c b/arch/x86/mach-generic/bigsmp.c index 292a225edabe..95fc463056d0 100644 --- a/arch/x86/mach-generic/bigsmp.c +++ b/arch/x86/mach-generic/bigsmp.c @@ -1,4 +1,4 @@ -/* +/* * APIC driver for "bigsmp" XAPIC machines with more than 8 virtual CPUs. * Drives the local APIC in "clustered mode". */ @@ -32,26 +32,26 @@ static int hp_ht_bigsmp(const struct dmi_system_id *d) static const struct dmi_system_id bigsmp_dmi_table[] = { - { hp_ht_bigsmp, "HP ProLiant DL760 G2", { - DMI_MATCH(DMI_BIOS_VENDOR, "HP"), - DMI_MATCH(DMI_BIOS_VERSION, "P44-"), - }}, - - { hp_ht_bigsmp, "HP ProLiant DL740", { - DMI_MATCH(DMI_BIOS_VENDOR, "HP"), - DMI_MATCH(DMI_BIOS_VERSION, "P47-"), - }}, + { hp_ht_bigsmp, "HP ProLiant DL760 G2", + { DMI_MATCH(DMI_BIOS_VENDOR, "HP"), + DMI_MATCH(DMI_BIOS_VERSION, "P44-"),} + }, + + { hp_ht_bigsmp, "HP ProLiant DL740", + { DMI_MATCH(DMI_BIOS_VENDOR, "HP"), + DMI_MATCH(DMI_BIOS_VERSION, "P47-"),} + }, { } }; static int probe_bigsmp(void) -{ +{ if (def_to_bigsmp) - dmi_bigsmp = 1; + dmi_bigsmp = 1; else dmi_check_system(bigsmp_dmi_table); - return dmi_bigsmp; -} + return dmi_bigsmp; +} -struct genapic apic_bigsmp = APIC_INIT("bigsmp", probe_bigsmp); +struct genapic apic_bigsmp = APIC_INIT("bigsmp", probe_bigsmp); diff --git a/arch/x86/mach-generic/default.c b/arch/x86/mach-generic/default.c index 1af0cc7648f0..9e835a11a13a 100644 --- a/arch/x86/mach-generic/default.c +++ b/arch/x86/mach-generic/default.c @@ -1,4 +1,4 @@ -/* +/* * Default generic APIC driver. This handles up to 8 CPUs. */ #define APIC_DEFINITION 1 @@ -19,8 +19,8 @@ /* should be called last. */ static int probe_default(void) -{ +{ return 1; -} +} -struct genapic apic_default = APIC_INIT("default", probe_default); +struct genapic apic_default = APIC_INIT("default", probe_default); diff --git a/arch/x86/mach-generic/probe.c b/arch/x86/mach-generic/probe.c index f410d3cb5659..c5ae751b994a 100644 --- a/arch/x86/mach-generic/probe.c +++ b/arch/x86/mach-generic/probe.c @@ -1,8 +1,9 @@ -/* Copyright 2003 Andi Kleen, SuSE Labs. - * Subject to the GNU Public License, v.2 - * +/* + * Copyright 2003 Andi Kleen, SuSE Labs. + * Subject to the GNU Public License, v.2 + * * Generic x86 APIC driver probe layer. - */ + */ #include <linux/threads.h> #include <linux/cpumask.h> #include <linux/string.h> @@ -24,7 +25,7 @@ struct genapic *genapic = &apic_default; static struct genapic *apic_probe[] __initdata = { &apic_summit, - &apic_bigsmp, + &apic_bigsmp, &apic_es7000, &apic_default, /* must be last */ NULL, @@ -69,7 +70,7 @@ void __init generic_bigsmp_probe(void) } void __init generic_apic_probe(void) -{ +{ if (!cmdline_apic) { int i; for (i = 0; apic_probe[i]; i++) { @@ -83,40 +84,40 @@ void __init generic_apic_probe(void) panic("Didn't find an APIC driver"); } printk(KERN_INFO "Using APIC driver %s\n", genapic->name); -} +} /* These functions can switch the APIC even after the initial ->probe() */ int __init mps_oem_check(struct mp_config_table *mpc, char *oem, char *productid) -{ +{ int i; - for (i = 0; apic_probe[i]; ++i) { - if (apic_probe[i]->mps_oem_check(mpc,oem,productid)) { + for (i = 0; apic_probe[i]; ++i) { + if (apic_probe[i]->mps_oem_check(mpc, oem, productid)) { if (!cmdline_apic) { genapic = apic_probe[i]; printk(KERN_INFO "Switched to APIC driver `%s'.\n", genapic->name); } return 1; - } - } + } + } return 0; -} +} int __init acpi_madt_oem_check(char *oem_id, char *oem_table_id) { int i; - for (i = 0; apic_probe[i]; ++i) { - if (apic_probe[i]->acpi_madt_oem_check(oem_id, oem_table_id)) { + for (i = 0; apic_probe[i]; ++i) { + if (apic_probe[i]->acpi_madt_oem_check(oem_id, oem_table_id)) { if (!cmdline_apic) { genapic = apic_probe[i]; printk(KERN_INFO "Switched to APIC driver `%s'.\n", genapic->name); } return 1; - } - } - return 0; + } + } + return 0; } int hard_smp_processor_id(void) diff --git a/arch/x86/mach-generic/summit.c b/arch/x86/mach-generic/summit.c index 74883ccb8f73..a97ea0f35b1e 100644 --- a/arch/x86/mach-generic/summit.c +++ b/arch/x86/mach-generic/summit.c @@ -1,4 +1,4 @@ -/* +/* * APIC driver for the IBM "Summit" chipset. */ #define APIC_DEFINITION 1 @@ -19,9 +19,9 @@ #include <asm/mach-summit/mach_mpparse.h> static int probe_summit(void) -{ +{ /* probed later in mptable/ACPI hooks */ return 0; -} +} -struct genapic apic_summit = APIC_INIT("summit", probe_summit); +struct genapic apic_summit = APIC_INIT("summit", probe_summit); diff --git a/arch/x86/mach-rdc321x/Makefile b/arch/x86/mach-rdc321x/Makefile index 1faac8125e3d..8325b4ca431c 100644 --- a/arch/x86/mach-rdc321x/Makefile +++ b/arch/x86/mach-rdc321x/Makefile @@ -1,5 +1,5 @@ # # Makefile for the RDC321x specific parts of the kernel # -obj-$(CONFIG_X86_RDC321X) := gpio.o platform.o wdt.o +obj-$(CONFIG_X86_RDC321X) := gpio.o platform.o diff --git a/arch/x86/mach-rdc321x/wdt.c b/arch/x86/mach-rdc321x/wdt.c deleted file mode 100644 index ec5625ae7061..000000000000 --- a/arch/x86/mach-rdc321x/wdt.c +++ /dev/null @@ -1,275 +0,0 @@ -/* - * RDC321x watchdog driver - * - * Copyright (C) 2007 Florian Fainelli <florian@openwrt.org> - * - * This driver is highly inspired from the cpu5_wdt driver - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. - * - */ - -#include <linux/module.h> -#include <linux/moduleparam.h> -#include <linux/types.h> -#include <linux/errno.h> -#include <linux/miscdevice.h> -#include <linux/fs.h> -#include <linux/init.h> -#include <linux/ioport.h> -#include <linux/timer.h> -#include <linux/completion.h> -#include <linux/jiffies.h> -#include <linux/platform_device.h> -#include <linux/watchdog.h> -#include <linux/io.h> -#include <linux/uaccess.h> - -#include <asm/mach-rdc321x/rdc321x_defs.h> - -#define RDC_WDT_MASK 0x80000000 /* Mask */ -#define RDC_WDT_EN 0x00800000 /* Enable bit */ -#define RDC_WDT_WTI 0x00200000 /* Generate CPU reset/NMI/WDT on timeout */ -#define RDC_WDT_RST 0x00100000 /* Reset bit */ -#define RDC_WDT_WIF 0x00040000 /* WDT IRQ Flag */ -#define RDC_WDT_IRT 0x00000100 /* IRQ Routing table */ -#define RDC_WDT_CNT 0x00000001 /* WDT count */ - -#define RDC_CLS_TMR 0x80003844 /* Clear timer */ - -#define RDC_WDT_INTERVAL (HZ/10+1) - -int nowayout = WATCHDOG_NOWAYOUT; -module_param(nowayout, int, 0); -MODULE_PARM_DESC(nowayout, "Watchdog cannot be stopped once started (default=" __MODULE_STRING(WATCHDOG_NOWAYOUT) ")"); - -static int ticks = 1000; - -/* some device data */ - -static struct { - struct completion stop; - volatile int running; - struct timer_list timer; - volatile int queue; - int default_ticks; - unsigned long inuse; -} rdc321x_wdt_device; - -/* generic helper functions */ - -static void rdc321x_wdt_trigger(unsigned long unused) -{ - if (rdc321x_wdt_device.running) - ticks--; - - /* keep watchdog alive */ - outl(RDC_WDT_EN|inl(RDC3210_CFGREG_DATA), RDC3210_CFGREG_DATA); - - /* requeue?? */ - if (rdc321x_wdt_device.queue && ticks) - mod_timer(&rdc321x_wdt_device.timer, - jiffies + RDC_WDT_INTERVAL); - else { - /* ticks doesn't matter anyway */ - complete(&rdc321x_wdt_device.stop); - } - -} - -static void rdc321x_wdt_reset(void) -{ - ticks = rdc321x_wdt_device.default_ticks; -} - -static void rdc321x_wdt_start(void) -{ - if (!rdc321x_wdt_device.queue) { - rdc321x_wdt_device.queue = 1; - - /* Clear the timer */ - outl(RDC_CLS_TMR, RDC3210_CFGREG_ADDR); - - /* Enable watchdog and set the timeout to 81.92 us */ - outl(RDC_WDT_EN|RDC_WDT_CNT, RDC3210_CFGREG_DATA); - - mod_timer(&rdc321x_wdt_device.timer, - jiffies + RDC_WDT_INTERVAL); - } - - /* if process dies, counter is not decremented */ - rdc321x_wdt_device.running++; -} - -static int rdc321x_wdt_stop(void) -{ - if (rdc321x_wdt_device.running) - rdc321x_wdt_device.running = 0; - - ticks = rdc321x_wdt_device.default_ticks; - - return -EIO; -} - -/* filesystem operations */ - -static int rdc321x_wdt_open(struct inode *inode, struct file *file) -{ - if (test_and_set_bit(0, &rdc321x_wdt_device.inuse)) - return -EBUSY; - - return nonseekable_open(inode, file); -} - -static int rdc321x_wdt_release(struct inode *inode, struct file *file) -{ - clear_bit(0, &rdc321x_wdt_device.inuse); - return 0; -} - -static int rdc321x_wdt_ioctl(struct inode *inode, struct file *file, - unsigned int cmd, unsigned long arg) -{ - void __user *argp = (void __user *)arg; - unsigned int value; - static struct watchdog_info ident = { - .options = WDIOF_CARDRESET, - .identity = "RDC321x WDT", - }; - - switch (cmd) { - case WDIOC_KEEPALIVE: - rdc321x_wdt_reset(); - break; - case WDIOC_GETSTATUS: - /* Read the value from the DATA register */ - value = inl(RDC3210_CFGREG_DATA); - if (copy_to_user(argp, &value, sizeof(int))) - return -EFAULT; - break; - case WDIOC_GETSUPPORT: - if (copy_to_user(argp, &ident, sizeof(ident))) - return -EFAULT; - break; - case WDIOC_SETOPTIONS: - if (copy_from_user(&value, argp, sizeof(int))) - return -EFAULT; - switch (value) { - case WDIOS_ENABLECARD: - rdc321x_wdt_start(); - break; - case WDIOS_DISABLECARD: - return rdc321x_wdt_stop(); - default: - return -EINVAL; - } - break; - default: - return -ENOTTY; - } - return 0; -} - -static ssize_t rdc321x_wdt_write(struct file *file, const char __user *buf, - size_t count, loff_t *ppos) -{ - if (!count) - return -EIO; - - rdc321x_wdt_reset(); - - return count; -} - -static const struct file_operations rdc321x_wdt_fops = { - .owner = THIS_MODULE, - .llseek = no_llseek, - .ioctl = rdc321x_wdt_ioctl, - .open = rdc321x_wdt_open, - .write = rdc321x_wdt_write, - .release = rdc321x_wdt_release, -}; - -static struct miscdevice rdc321x_wdt_misc = { - .minor = WATCHDOG_MINOR, - .name = "watchdog", - .fops = &rdc321x_wdt_fops, -}; - -static int __devinit rdc321x_wdt_probe(struct platform_device *pdev) -{ - int err; - - err = misc_register(&rdc321x_wdt_misc); - if (err < 0) { - printk(KERN_ERR PFX "watchdog misc_register failed\n"); - return err; - } - - /* Reset the watchdog */ - outl(RDC_WDT_RST, RDC3210_CFGREG_DATA); - - init_completion(&rdc321x_wdt_device.stop); - rdc321x_wdt_device.queue = 0; - - clear_bit(0, &rdc321x_wdt_device.inuse); - - setup_timer(&rdc321x_wdt_device.timer, rdc321x_wdt_trigger, 0); - - rdc321x_wdt_device.default_ticks = ticks; - - printk(KERN_INFO PFX "watchdog init success\n"); - - return 0; -} - -static int rdc321x_wdt_remove(struct platform_device *pdev) -{ - if (rdc321x_wdt_device.queue) { - rdc321x_wdt_device.queue = 0; - wait_for_completion(&rdc321x_wdt_device.stop); - } - - misc_deregister(&rdc321x_wdt_misc); - - return 0; -} - -static struct platform_driver rdc321x_wdt_driver = { - .probe = rdc321x_wdt_probe, - .remove = rdc321x_wdt_remove, - .driver = { - .owner = THIS_MODULE, - .name = "rdc321x-wdt", - }, -}; - -static int __init rdc321x_wdt_init(void) -{ - return platform_driver_register(&rdc321x_wdt_driver); -} - -static void __exit rdc321x_wdt_exit(void) -{ - platform_driver_unregister(&rdc321x_wdt_driver); -} - -module_init(rdc321x_wdt_init); -module_exit(rdc321x_wdt_exit); - -MODULE_AUTHOR("Florian Fainelli <florian@openwrt.org>"); -MODULE_DESCRIPTION("RDC321x watchdog driver"); -MODULE_LICENSE("GPL"); -MODULE_ALIAS_MISCDEV(WATCHDOG_MINOR); diff --git a/arch/x86/mach-voyager/voyager_smp.c b/arch/x86/mach-voyager/voyager_smp.c index 3cc8eb2f36a9..be7235bf105d 100644 --- a/arch/x86/mach-voyager/voyager_smp.c +++ b/arch/x86/mach-voyager/voyager_smp.c @@ -27,6 +27,7 @@ #include <asm/pgalloc.h> #include <asm/tlbflush.h> #include <asm/arch_hooks.h> +#include <asm/trampoline.h> /* TLB state -- visible externally, indexed physically */ DEFINE_PER_CPU_SHARED_ALIGNED(struct tlb_state, cpu_tlbstate) = { &init_mm, 0 }; @@ -210,7 +211,7 @@ static int cpucount = 0; /* steal a page from the bottom of memory for the trampoline and * squirrel its address away here. This will be in kernel virtual * space */ -static __u32 trampoline_base; +unsigned char *trampoline_base; /* The per cpu profile stuff - used in smp_local_timer_interrupt */ static DEFINE_PER_CPU(int, prof_multiplier) = 1; @@ -429,15 +430,15 @@ void __init smp_store_cpu_info(int id) } /* set up the trampoline and return the physical address of the code */ -static __u32 __init setup_trampoline(void) +unsigned long __init setup_trampoline(void) { /* these two are global symbols in trampoline.S */ extern const __u8 trampoline_end[]; extern const __u8 trampoline_data[]; - memcpy((__u8 *) trampoline_base, trampoline_data, + memcpy(trampoline_base, trampoline_data, trampoline_end - trampoline_data); - return virt_to_phys((__u8 *) trampoline_base); + return virt_to_phys(trampoline_base); } /* Routine initially called when a non-boot CPU is brought online */ @@ -520,13 +521,6 @@ static void __init do_boot_cpu(__u8 cpu) & ~(voyager_extended_vic_processors & voyager_allowed_boot_processors); - /* This is an area in head.S which was used to set up the - * initial kernel stack. We need to alter this to give the - * booting CPU a new stack (taken from its idle process) */ - extern struct { - __u8 *sp; - unsigned short ss; - } stack_start; /* This is the format of the CPI IDT gate (in real mode) which * we're hijacking to boot the CPU */ union IDTFormat { @@ -1166,7 +1160,7 @@ void flush_tlb_all(void) * is sorted out */ void __init smp_alloc_memory(void) { - trampoline_base = (__u32) alloc_bootmem_low_pages(PAGE_SIZE); + trampoline_base = alloc_bootmem_low_pages(PAGE_SIZE); if (__pa(trampoline_base) >= 0x93000) BUG(); } diff --git a/arch/x86/math-emu/fpu_entry.c b/arch/x86/math-emu/fpu_entry.c index 760baeea5f07..4bab3b145392 100644 --- a/arch/x86/math-emu/fpu_entry.c +++ b/arch/x86/math-emu/fpu_entry.c @@ -276,6 +276,7 @@ asmlinkage void math_emulate(long arg) entry_sel_off.offset = FPU_ORIG_EIP; entry_sel_off.selector = FPU_CS; entry_sel_off.opcode = (byte1 << 8) | FPU_modrm; + entry_sel_off.empty = 0; FPU_rm = FPU_modrm & 7; diff --git a/arch/x86/math-emu/reg_ld_str.c b/arch/x86/math-emu/reg_ld_str.c index 799d4af5be66..02af772a24db 100644 --- a/arch/x86/math-emu/reg_ld_str.c +++ b/arch/x86/math-emu/reg_ld_str.c @@ -383,15 +383,15 @@ int FPU_store_double(FPU_REG *st0_ptr, u_char st0_tag, double __user *dfloat) int exp; FPU_REG tmp; + l[0] = 0; + l[1] = 0; if (st0_tag == TAG_Valid) { reg_copy(st0_ptr, &tmp); exp = exponent(&tmp); if (exp < DOUBLE_Emin) { /* It may be a denormal */ addexponent(&tmp, -DOUBLE_Emin + 52); /* largest exp to be 51 */ - - denormal_arg: - +denormal_arg: if ((precision_loss = FPU_round_to_int(&tmp, st0_tag))) { #ifdef PECULIAR_486 /* Did it round to a non-denormal ? */ @@ -477,8 +477,7 @@ int FPU_store_double(FPU_REG *st0_ptr, u_char st0_tag, double __user *dfloat) /* This is a special case: see sec 16.2.5.1 of the 80486 book */ /* Overflow to infinity */ - l[0] = 0x00000000; /* Set to */ - l[1] = 0x7ff00000; /* + INF */ + l[1] = 0x7ff00000; /* Set to + INF */ } else { if (precision_loss) { if (increment) @@ -492,8 +491,6 @@ int FPU_store_double(FPU_REG *st0_ptr, u_char st0_tag, double __user *dfloat) } } else if (st0_tag == TAG_Zero) { /* Number is zero */ - l[0] = 0; - l[1] = 0; } else if (st0_tag == TAG_Special) { st0_tag = FPU_Special(st0_ptr); if (st0_tag == TW_Denormal) { @@ -508,7 +505,6 @@ int FPU_store_double(FPU_REG *st0_ptr, u_char st0_tag, double __user *dfloat) reg_copy(st0_ptr, &tmp); goto denormal_arg; } else if (st0_tag == TW_Infinity) { - l[0] = 0; l[1] = 0x7ff00000; } else if (st0_tag == TW_NaN) { /* Is it really a NaN ? */ @@ -532,7 +528,6 @@ int FPU_store_double(FPU_REG *st0_ptr, u_char st0_tag, double __user *dfloat) EXCEPTION(EX_Invalid); if (!(control_word & CW_Invalid)) return 0; - l[0] = 0; l[1] = 0xfff80000; } } diff --git a/arch/x86/mm/Makefile b/arch/x86/mm/Makefile index 983291096848..20941d2954e2 100644 --- a/arch/x86/mm/Makefile +++ b/arch/x86/mm/Makefile @@ -1,5 +1,17 @@ +obj-y := init_$(BITS).o fault.o ioremap.o extable.o pageattr.o mmap.o \ + pat.o + +obj-$(CONFIG_X86_32) += pgtable_32.o + +obj-$(CONFIG_HUGETLB_PAGE) += hugetlbpage.o +obj-$(CONFIG_X86_PTDUMP) += dump_pagetables.o + +obj-$(CONFIG_HIGHMEM) += highmem_32.o + ifeq ($(CONFIG_X86_32),y) -include ${srctree}/arch/x86/mm/Makefile_32 +obj-$(CONFIG_NUMA) += discontig_32.o else -include ${srctree}/arch/x86/mm/Makefile_64 +obj-$(CONFIG_NUMA) += numa_64.o +obj-$(CONFIG_K8_NUMA) += k8topology_64.o +obj-$(CONFIG_ACPI_NUMA) += srat_64.o endif diff --git a/arch/x86/mm/Makefile_32 b/arch/x86/mm/Makefile_32 deleted file mode 100644 index c36ae88bb543..000000000000 --- a/arch/x86/mm/Makefile_32 +++ /dev/null @@ -1,9 +0,0 @@ -# -# Makefile for the linux i386-specific parts of the memory manager. -# - -obj-y := init_32.o pgtable_32.o fault.o ioremap.o extable.o pageattr.o mmap.o - -obj-$(CONFIG_NUMA) += discontig_32.o -obj-$(CONFIG_HUGETLB_PAGE) += hugetlbpage.o -obj-$(CONFIG_HIGHMEM) += highmem_32.o diff --git a/arch/x86/mm/Makefile_64 b/arch/x86/mm/Makefile_64 deleted file mode 100644 index 688c8c28ac8f..000000000000 --- a/arch/x86/mm/Makefile_64 +++ /dev/null @@ -1,9 +0,0 @@ -# -# Makefile for the linux x86_64-specific parts of the memory manager. -# - -obj-y := init_64.o fault.o ioremap.o extable.o pageattr.o mmap.o -obj-$(CONFIG_HUGETLB_PAGE) += hugetlbpage.o -obj-$(CONFIG_NUMA) += numa_64.o -obj-$(CONFIG_K8_NUMA) += k8topology_64.o -obj-$(CONFIG_ACPI_NUMA) += srat_64.o diff --git a/arch/x86/mm/discontig_32.c b/arch/x86/mm/discontig_32.c index 8e25e06ff730..eba0bbede7a6 100644 --- a/arch/x86/mm/discontig_32.c +++ b/arch/x86/mm/discontig_32.c @@ -37,7 +37,7 @@ #include <asm/e820.h> #include <asm/setup.h> #include <asm/mmzone.h> -#include <bios_ebda.h> +#include <asm/bios_ebda.h> struct pglist_data *node_data[MAX_NUMNODES] __read_mostly; EXPORT_SYMBOL(node_data); diff --git a/arch/x86/mm/dump_pagetables.c b/arch/x86/mm/dump_pagetables.c new file mode 100644 index 000000000000..6791b8334bc6 --- /dev/null +++ b/arch/x86/mm/dump_pagetables.c @@ -0,0 +1,354 @@ +/* + * Debug helper to dump the current kernel pagetables of the system + * so that we can see what the various memory ranges are set to. + * + * (C) Copyright 2008 Intel Corporation + * + * Author: Arjan van de Ven <arjan@linux.intel.com> + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; version 2 + * of the License. + */ + +#include <linux/debugfs.h> +#include <linux/mm.h> +#include <linux/module.h> +#include <linux/seq_file.h> + +#include <asm/pgtable.h> + +/* + * The dumper groups pagetable entries of the same type into one, and for + * that it needs to keep some state when walking, and flush this state + * when a "break" in the continuity is found. + */ +struct pg_state { + int level; + pgprot_t current_prot; + unsigned long start_address; + unsigned long current_address; + const struct addr_marker *marker; +}; + +struct addr_marker { + unsigned long start_address; + const char *name; +}; + +/* Address space markers hints */ +static struct addr_marker address_markers[] = { + { 0, "User Space" }, +#ifdef CONFIG_X86_64 + { 0x8000000000000000UL, "Kernel Space" }, + { 0xffff810000000000UL, "Low Kernel Mapping" }, + { VMALLOC_START, "vmalloc() Area" }, + { VMEMMAP_START, "Vmemmap" }, + { __START_KERNEL_map, "High Kernel Mapping" }, + { MODULES_VADDR, "Modules" }, + { MODULES_END, "End Modules" }, +#else + { PAGE_OFFSET, "Kernel Mapping" }, + { 0/* VMALLOC_START */, "vmalloc() Area" }, + { 0/*VMALLOC_END*/, "vmalloc() End" }, +# ifdef CONFIG_HIGHMEM + { 0/*PKMAP_BASE*/, "Persisent kmap() Area" }, +# endif + { 0/*FIXADDR_START*/, "Fixmap Area" }, +#endif + { -1, NULL } /* End of list */ +}; + +/* Multipliers for offsets within the PTEs */ +#define PTE_LEVEL_MULT (PAGE_SIZE) +#define PMD_LEVEL_MULT (PTRS_PER_PTE * PTE_LEVEL_MULT) +#define PUD_LEVEL_MULT (PTRS_PER_PMD * PMD_LEVEL_MULT) +#define PGD_LEVEL_MULT (PTRS_PER_PUD * PUD_LEVEL_MULT) + +/* + * Print a readable form of a pgprot_t to the seq_file + */ +static void printk_prot(struct seq_file *m, pgprot_t prot, int level) +{ + pgprotval_t pr = pgprot_val(prot); + static const char * const level_name[] = + { "cr3", "pgd", "pud", "pmd", "pte" }; + + if (!pgprot_val(prot)) { + /* Not present */ + seq_printf(m, " "); + } else { + if (pr & _PAGE_USER) + seq_printf(m, "USR "); + else + seq_printf(m, " "); + if (pr & _PAGE_RW) + seq_printf(m, "RW "); + else + seq_printf(m, "ro "); + if (pr & _PAGE_PWT) + seq_printf(m, "PWT "); + else + seq_printf(m, " "); + if (pr & _PAGE_PCD) + seq_printf(m, "PCD "); + else + seq_printf(m, " "); + + /* Bit 9 has a different meaning on level 3 vs 4 */ + if (level <= 3) { + if (pr & _PAGE_PSE) + seq_printf(m, "PSE "); + else + seq_printf(m, " "); + } else { + if (pr & _PAGE_PAT) + seq_printf(m, "pat "); + else + seq_printf(m, " "); + } + if (pr & _PAGE_GLOBAL) + seq_printf(m, "GLB "); + else + seq_printf(m, " "); + if (pr & _PAGE_NX) + seq_printf(m, "NX "); + else + seq_printf(m, "x "); + } + seq_printf(m, "%s\n", level_name[level]); +} + +/* + * On 64 bits, sign-extend the 48 bit address to 64 bit + */ +static unsigned long normalize_addr(unsigned long u) +{ +#ifdef CONFIG_X86_64 + return (signed long)(u << 16) >> 16; +#else + return u; +#endif +} + +/* + * This function gets called on a break in a continuous series + * of PTE entries; the next one is different so we need to + * print what we collected so far. + */ +static void note_page(struct seq_file *m, struct pg_state *st, + pgprot_t new_prot, int level) +{ + pgprotval_t prot, cur; + static const char units[] = "KMGTPE"; + + /* + * If we have a "break" in the series, we need to flush the state that + * we have now. "break" is either changing perms, levels or + * address space marker. + */ + prot = pgprot_val(new_prot) & ~(PTE_MASK); + cur = pgprot_val(st->current_prot) & ~(PTE_MASK); + + if (!st->level) { + /* First entry */ + st->current_prot = new_prot; + st->level = level; + st->marker = address_markers; + seq_printf(m, "---[ %s ]---\n", st->marker->name); + } else if (prot != cur || level != st->level || + st->current_address >= st->marker[1].start_address) { + const char *unit = units; + unsigned long delta; + + /* + * Now print the actual finished series + */ + seq_printf(m, "0x%p-0x%p ", + (void *)st->start_address, + (void *)st->current_address); + + delta = (st->current_address - st->start_address) >> 10; + while (!(delta & 1023) && unit[1]) { + delta >>= 10; + unit++; + } + seq_printf(m, "%9lu%c ", delta, *unit); + printk_prot(m, st->current_prot, st->level); + + /* + * We print markers for special areas of address space, + * such as the start of vmalloc space etc. + * This helps in the interpretation. + */ + if (st->current_address >= st->marker[1].start_address) { + st->marker++; + seq_printf(m, "---[ %s ]---\n", st->marker->name); + } + + st->start_address = st->current_address; + st->current_prot = new_prot; + st->level = level; + } +} + +static void walk_pte_level(struct seq_file *m, struct pg_state *st, pmd_t addr, + unsigned long P) +{ + int i; + pte_t *start; + + start = (pte_t *) pmd_page_vaddr(addr); + for (i = 0; i < PTRS_PER_PTE; i++) { + pgprot_t prot = pte_pgprot(*start); + + st->current_address = normalize_addr(P + i * PTE_LEVEL_MULT); + note_page(m, st, prot, 4); + start++; + } +} + +#if PTRS_PER_PMD > 1 + +static void walk_pmd_level(struct seq_file *m, struct pg_state *st, pud_t addr, + unsigned long P) +{ + int i; + pmd_t *start; + + start = (pmd_t *) pud_page_vaddr(addr); + for (i = 0; i < PTRS_PER_PMD; i++) { + st->current_address = normalize_addr(P + i * PMD_LEVEL_MULT); + if (!pmd_none(*start)) { + pgprotval_t prot = pmd_val(*start) & ~PTE_MASK; + + if (pmd_large(*start) || !pmd_present(*start)) + note_page(m, st, __pgprot(prot), 3); + else + walk_pte_level(m, st, *start, + P + i * PMD_LEVEL_MULT); + } else + note_page(m, st, __pgprot(0), 3); + start++; + } +} + +#else +#define walk_pmd_level(m,s,a,p) walk_pte_level(m,s,__pmd(pud_val(a)),p) +#define pud_large(a) pmd_large(__pmd(pud_val(a))) +#define pud_none(a) pmd_none(__pmd(pud_val(a))) +#endif + +#if PTRS_PER_PUD > 1 + +static void walk_pud_level(struct seq_file *m, struct pg_state *st, pgd_t addr, + unsigned long P) +{ + int i; + pud_t *start; + + start = (pud_t *) pgd_page_vaddr(addr); + + for (i = 0; i < PTRS_PER_PUD; i++) { + st->current_address = normalize_addr(P + i * PUD_LEVEL_MULT); + if (!pud_none(*start)) { + pgprotval_t prot = pud_val(*start) & ~PTE_MASK; + + if (pud_large(*start) || !pud_present(*start)) + note_page(m, st, __pgprot(prot), 2); + else + walk_pmd_level(m, st, *start, + P + i * PUD_LEVEL_MULT); + } else + note_page(m, st, __pgprot(0), 2); + + start++; + } +} + +#else +#define walk_pud_level(m,s,a,p) walk_pmd_level(m,s,__pud(pgd_val(a)),p) +#define pgd_large(a) pud_large(__pud(pgd_val(a))) +#define pgd_none(a) pud_none(__pud(pgd_val(a))) +#endif + +static void walk_pgd_level(struct seq_file *m) +{ +#ifdef CONFIG_X86_64 + pgd_t *start = (pgd_t *) &init_level4_pgt; +#else + pgd_t *start = swapper_pg_dir; +#endif + int i; + struct pg_state st; + + memset(&st, 0, sizeof(st)); + + for (i = 0; i < PTRS_PER_PGD; i++) { + st.current_address = normalize_addr(i * PGD_LEVEL_MULT); + if (!pgd_none(*start)) { + pgprotval_t prot = pgd_val(*start) & ~PTE_MASK; + + if (pgd_large(*start) || !pgd_present(*start)) + note_page(m, &st, __pgprot(prot), 1); + else + walk_pud_level(m, &st, *start, + i * PGD_LEVEL_MULT); + } else + note_page(m, &st, __pgprot(0), 1); + + start++; + } + + /* Flush out the last page */ + st.current_address = normalize_addr(PTRS_PER_PGD*PGD_LEVEL_MULT); + note_page(m, &st, __pgprot(0), 0); +} + +static int ptdump_show(struct seq_file *m, void *v) +{ + walk_pgd_level(m); + return 0; +} + +static int ptdump_open(struct inode *inode, struct file *filp) +{ + return single_open(filp, ptdump_show, NULL); +} + +static const struct file_operations ptdump_fops = { + .open = ptdump_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; + +int pt_dump_init(void) +{ + struct dentry *pe; + +#ifdef CONFIG_X86_32 + /* Not a compile-time constant on x86-32 */ + address_markers[2].start_address = VMALLOC_START; + address_markers[3].start_address = VMALLOC_END; +# ifdef CONFIG_HIGHMEM + address_markers[4].start_address = PKMAP_BASE; + address_markers[5].start_address = FIXADDR_START; +# else + address_markers[4].start_address = FIXADDR_START; +# endif +#endif + + pe = debugfs_create_file("kernel_page_tables", 0600, NULL, NULL, + &ptdump_fops); + if (!pe) + return -ENOMEM; + + return 0; +} + +__initcall(pt_dump_init); +MODULE_LICENSE("GPL"); +MODULE_AUTHOR("Arjan van de Ven <arjan@linux.intel.com>"); +MODULE_DESCRIPTION("Kernel debugging helper that dumps pagetables"); diff --git a/arch/x86/mm/fault.c b/arch/x86/mm/fault.c index ec08d8389850..fd7e1798c75a 100644 --- a/arch/x86/mm/fault.c +++ b/arch/x86/mm/fault.c @@ -639,7 +639,7 @@ void __kprobes do_page_fault(struct pt_regs *regs, unsigned long error_code) #ifdef CONFIG_X86_32 /* It's safe to allow irq's after cr2 has been saved and the vmalloc fault has been handled. */ - if (regs->flags & (X86_EFLAGS_IF|VM_MASK)) + if (regs->flags & (X86_EFLAGS_IF | X86_VM_MASK)) local_irq_enable(); /* @@ -976,9 +976,5 @@ void vmalloc_sync_all(void) if (address == start) start = address + PGDIR_SIZE; } - /* Check that there is no need to do the same for the modules area. */ - BUILD_BUG_ON(!(MODULES_VADDR > __START_KERNEL)); - BUILD_BUG_ON(!(((MODULES_END - 1) & PGDIR_MASK) == - (__START_KERNEL & PGDIR_MASK))); #endif } diff --git a/arch/x86/mm/init_32.c b/arch/x86/mm/init_32.c index ee1091a46964..1500dc8d63e4 100644 --- a/arch/x86/mm/init_32.c +++ b/arch/x86/mm/init_32.c @@ -51,6 +51,8 @@ unsigned int __VMALLOC_RESERVE = 128 << 20; +unsigned long max_pfn_mapped; + DEFINE_PER_CPU(struct mmu_gather, mmu_gathers); unsigned long highstart_pfn, highend_pfn; @@ -179,8 +181,13 @@ static void __init kernel_physical_mapping_init(pgd_t *pgd_base) /* * Map with big pages if possible, otherwise * create normal page tables: + * + * Don't use a large page for the first 2/4MB of memory + * because there are often fixed size MTRRs in there + * and overlapping MTRRs into large pages can cause + * slowdowns. */ - if (cpu_has_pse) { + if (cpu_has_pse && !(pgd_idx == 0 && pmd_idx == 0)) { unsigned int addr2; pgprot_t prot = PAGE_KERNEL_LARGE; @@ -194,6 +201,7 @@ static void __init kernel_physical_mapping_init(pgd_t *pgd_base) set_pmd(pmd, pfn_pmd(pfn, prot)); pfn += PTRS_PER_PTE; + max_pfn_mapped = pfn; continue; } pte = one_page_table_init(pmd); @@ -208,6 +216,7 @@ static void __init kernel_physical_mapping_init(pgd_t *pgd_base) set_pte(pte, pfn_pte(pfn, prot)); } + max_pfn_mapped = pfn; } } } @@ -723,25 +732,17 @@ void mark_rodata_ro(void) unsigned long start = PFN_ALIGN(_text); unsigned long size = PFN_ALIGN(_etext) - start; -#ifndef CONFIG_KPROBES -#ifdef CONFIG_HOTPLUG_CPU - /* It must still be possible to apply SMP alternatives. */ - if (num_possible_cpus() <= 1) -#endif - { - set_pages_ro(virt_to_page(start), size >> PAGE_SHIFT); - printk(KERN_INFO "Write protecting the kernel text: %luk\n", - size >> 10); + set_pages_ro(virt_to_page(start), size >> PAGE_SHIFT); + printk(KERN_INFO "Write protecting the kernel text: %luk\n", + size >> 10); #ifdef CONFIG_CPA_DEBUG - printk(KERN_INFO "Testing CPA: Reverting %lx-%lx\n", - start, start+size); - set_pages_rw(virt_to_page(start), size>>PAGE_SHIFT); + printk(KERN_INFO "Testing CPA: Reverting %lx-%lx\n", + start, start+size); + set_pages_rw(virt_to_page(start), size>>PAGE_SHIFT); - printk(KERN_INFO "Testing CPA: write protecting again\n"); - set_pages_ro(virt_to_page(start), size>>PAGE_SHIFT); -#endif - } + printk(KERN_INFO "Testing CPA: write protecting again\n"); + set_pages_ro(virt_to_page(start), size>>PAGE_SHIFT); #endif start += size; size = (unsigned long)__end_rodata - start; diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c index a02a14f0f324..1076097dcab2 100644 --- a/arch/x86/mm/init_64.c +++ b/arch/x86/mm/init_64.c @@ -54,6 +54,26 @@ static unsigned long dma_reserve __initdata; DEFINE_PER_CPU(struct mmu_gather, mmu_gathers); +int direct_gbpages __meminitdata +#ifdef CONFIG_DIRECT_GBPAGES + = 1 +#endif +; + +static int __init parse_direct_gbpages_off(char *arg) +{ + direct_gbpages = 0; + return 0; +} +early_param("nogbpages", parse_direct_gbpages_off); + +static int __init parse_direct_gbpages_on(char *arg) +{ + direct_gbpages = 1; + return 0; +} +early_param("gbpages", parse_direct_gbpages_on); + /* * NOTE: pagetable_init alloc all the fixmap pagetables contiguous on the * physical space so we can cache the place of the first one and move @@ -69,9 +89,6 @@ void show_mem(void) printk(KERN_INFO "Mem-info:\n"); show_free_areas(); - printk(KERN_INFO "Free swap: %6ldkB\n", - nr_swap_pages << (PAGE_SHIFT-10)); - for_each_online_pgdat(pgdat) { for (i = 0; i < pgdat->node_spanned_pages; ++i) { /* @@ -296,7 +313,7 @@ __meminit void early_iounmap(void *addr, unsigned long size) __flush_tlb_all(); } -static void __meminit +static unsigned long __meminit phys_pmd_init(pmd_t *pmd_page, unsigned long address, unsigned long end) { int i = pmd_index(address); @@ -318,21 +335,26 @@ phys_pmd_init(pmd_t *pmd_page, unsigned long address, unsigned long end) set_pte((pte_t *)pmd, pfn_pte(address >> PAGE_SHIFT, PAGE_KERNEL_LARGE)); } + return address; } -static void __meminit +static unsigned long __meminit phys_pmd_update(pud_t *pud, unsigned long address, unsigned long end) { pmd_t *pmd = pmd_offset(pud, 0); + unsigned long last_map_addr; + spin_lock(&init_mm.page_table_lock); - phys_pmd_init(pmd, address, end); + last_map_addr = phys_pmd_init(pmd, address, end); spin_unlock(&init_mm.page_table_lock); __flush_tlb_all(); + return last_map_addr; } -static void __meminit +static unsigned long __meminit phys_pud_init(pud_t *pud_page, unsigned long addr, unsigned long end) { + unsigned long last_map_addr = end; int i = pud_index(addr); for (; i < PTRS_PER_PUD; i++, addr = (addr & PUD_MASK) + PUD_SIZE) { @@ -350,7 +372,15 @@ phys_pud_init(pud_t *pud_page, unsigned long addr, unsigned long end) } if (pud_val(*pud)) { - phys_pmd_update(pud, addr, end); + if (!pud_large(*pud)) + last_map_addr = phys_pmd_update(pud, addr, end); + continue; + } + + if (direct_gbpages) { + set_pte((pte_t *)pud, + pfn_pte(addr >> PAGE_SHIFT, PAGE_KERNEL_LARGE)); + last_map_addr = (addr & PUD_MASK) + PUD_SIZE; continue; } @@ -358,12 +388,14 @@ phys_pud_init(pud_t *pud_page, unsigned long addr, unsigned long end) spin_lock(&init_mm.page_table_lock); set_pud(pud, __pud(pmd_phys | _KERNPG_TABLE)); - phys_pmd_init(pmd, addr, end); + last_map_addr = phys_pmd_init(pmd, addr, end); spin_unlock(&init_mm.page_table_lock); unmap_low_page(pmd); } __flush_tlb_all(); + + return last_map_addr >> PAGE_SHIFT; } static void __init find_early_table_space(unsigned long end) @@ -371,9 +403,11 @@ static void __init find_early_table_space(unsigned long end) unsigned long puds, pmds, tables, start; puds = (end + PUD_SIZE - 1) >> PUD_SHIFT; - pmds = (end + PMD_SIZE - 1) >> PMD_SHIFT; - tables = round_up(puds * sizeof(pud_t), PAGE_SIZE) + - round_up(pmds * sizeof(pmd_t), PAGE_SIZE); + tables = round_up(puds * sizeof(pud_t), PAGE_SIZE); + if (!direct_gbpages) { + pmds = (end + PMD_SIZE - 1) >> PMD_SHIFT; + tables += round_up(pmds * sizeof(pmd_t), PAGE_SIZE); + } /* * RED-PEN putting page tables only on node 0 could @@ -393,16 +427,135 @@ static void __init find_early_table_space(unsigned long end) (table_start << PAGE_SHIFT) + tables); } +static void __init init_gbpages(void) +{ + if (direct_gbpages && cpu_has_gbpages) + printk(KERN_INFO "Using GB pages for direct mapping\n"); + else + direct_gbpages = 0; +} + +#ifdef CONFIG_MEMTEST_BOOTPARAM + +static void __init memtest(unsigned long start_phys, unsigned long size, + unsigned pattern) +{ + unsigned long i; + unsigned long *start; + unsigned long start_bad; + unsigned long last_bad; + unsigned long val; + unsigned long start_phys_aligned; + unsigned long count; + unsigned long incr; + + switch (pattern) { + case 0: + val = 0UL; + break; + case 1: + val = -1UL; + break; + case 2: + val = 0x5555555555555555UL; + break; + case 3: + val = 0xaaaaaaaaaaaaaaaaUL; + break; + default: + return; + } + + incr = sizeof(unsigned long); + start_phys_aligned = ALIGN(start_phys, incr); + count = (size - (start_phys_aligned - start_phys))/incr; + start = __va(start_phys_aligned); + start_bad = 0; + last_bad = 0; + + for (i = 0; i < count; i++) + start[i] = val; + for (i = 0; i < count; i++, start++, start_phys_aligned += incr) { + if (*start != val) { + if (start_phys_aligned == last_bad + incr) { + last_bad += incr; + } else { + if (start_bad) { + printk(KERN_CONT "\n %016lx bad mem addr %016lx - %016lx reserved", + val, start_bad, last_bad + incr); + reserve_early(start_bad, last_bad - start_bad, "BAD RAM"); + } + start_bad = last_bad = start_phys_aligned; + } + } + } + if (start_bad) { + printk(KERN_CONT "\n %016lx bad mem addr %016lx - %016lx reserved", + val, start_bad, last_bad + incr); + reserve_early(start_bad, last_bad - start_bad, "BAD RAM"); + } + +} + +static int memtest_pattern __initdata = CONFIG_MEMTEST_BOOTPARAM_VALUE; + +static int __init parse_memtest(char *arg) +{ + if (arg) + memtest_pattern = simple_strtoul(arg, NULL, 0); + return 0; +} + +early_param("memtest", parse_memtest); + +static void __init early_memtest(unsigned long start, unsigned long end) +{ + unsigned long t_start, t_size; + unsigned pattern; + + if (!memtest_pattern) + return; + + printk(KERN_INFO "early_memtest: pattern num %d", memtest_pattern); + for (pattern = 0; pattern < memtest_pattern; pattern++) { + t_start = start; + t_size = 0; + while (t_start < end) { + t_start = find_e820_area_size(t_start, &t_size, 1); + + /* done ? */ + if (t_start >= end) + break; + if (t_start + t_size > end) + t_size = end - t_start; + + printk(KERN_CONT "\n %016lx - %016lx pattern %d", + t_start, t_start + t_size, pattern); + + memtest(t_start, t_size, pattern); + + t_start += t_size; + } + } + printk(KERN_CONT "\n"); +} +#else +static void __init early_memtest(unsigned long start, unsigned long end) +{ +} +#endif + /* * Setup the direct mapping of the physical memory at PAGE_OFFSET. * This runs before bootmem is initialized and gets pages directly from * the physical memory. To access them they are temporarily mapped. */ -void __init_refok init_memory_mapping(unsigned long start, unsigned long end) +unsigned long __init_refok init_memory_mapping(unsigned long start, unsigned long end) { - unsigned long next; + unsigned long next, last_map_addr = end; + unsigned long start_phys = start, end_phys = end; - pr_debug("init_memory_mapping\n"); + printk(KERN_INFO "init_memory_mapping\n"); /* * Find space for the kernel direct mapping tables. @@ -411,8 +564,10 @@ void __init_refok init_memory_mapping(unsigned long start, unsigned long end) * memory mapped. Unfortunately this is done currently before the * nodes are discovered. */ - if (!after_bootmem) + if (!after_bootmem) { + init_gbpages(); find_early_table_space(end); + } start = (unsigned long)__va(start); end = (unsigned long)__va(end); @@ -430,7 +585,7 @@ void __init_refok init_memory_mapping(unsigned long start, unsigned long end) next = start + PGDIR_SIZE; if (next > end) next = end; - phys_pud_init(pud, __pa(start), __pa(next)); + last_map_addr = phys_pud_init(pud, __pa(start), __pa(next)); if (!after_bootmem) set_pgd(pgd_offset_k(start), mk_kernel_pgd(pud_phys)); unmap_low_page(pud); @@ -443,6 +598,11 @@ void __init_refok init_memory_mapping(unsigned long start, unsigned long end) if (!after_bootmem) reserve_early(table_start << PAGE_SHIFT, table_end << PAGE_SHIFT, "PGTABLE"); + + if (!after_bootmem) + early_memtest(start_phys, end_phys); + + return last_map_addr; } #ifndef CONFIG_NUMA @@ -482,11 +642,13 @@ int arch_add_memory(int nid, u64 start, u64 size) { struct pglist_data *pgdat = NODE_DATA(nid); struct zone *zone = pgdat->node_zones + ZONE_NORMAL; - unsigned long start_pfn = start >> PAGE_SHIFT; + unsigned long last_mapped_pfn, start_pfn = start >> PAGE_SHIFT; unsigned long nr_pages = size >> PAGE_SHIFT; int ret; - init_memory_mapping(start, start + size-1); + last_mapped_pfn = init_memory_mapping(start, start + size-1); + if (last_mapped_pfn > max_pfn_mapped) + max_pfn_mapped = last_mapped_pfn; ret = __add_pages(zone, start_pfn, nr_pages); WARN_ON(1); @@ -596,24 +758,7 @@ EXPORT_SYMBOL_GPL(rodata_test_data); void mark_rodata_ro(void) { - unsigned long start = (unsigned long)_stext, end; - -#ifdef CONFIG_HOTPLUG_CPU - /* It must still be possible to apply SMP alternatives. */ - if (num_possible_cpus() > 1) - start = (unsigned long)_etext; -#endif - -#ifdef CONFIG_KPROBES - start = (unsigned long)__start_rodata; -#endif - - end = (unsigned long)__end_rodata; - start = (start + PAGE_SIZE - 1) & PAGE_MASK; - end &= PAGE_MASK; - if (end <= start) - return; - + unsigned long start = PFN_ALIGN(_stext), end = PFN_ALIGN(__end_rodata); printk(KERN_INFO "Write protecting the kernel read-only data: %luk\n", (end - start) >> 10); @@ -636,6 +781,7 @@ void mark_rodata_ro(void) set_memory_ro(start, (end-start) >> PAGE_SHIFT); #endif } + #endif #ifdef CONFIG_BLK_DEV_INITRD @@ -657,7 +803,7 @@ void __init reserve_bootmem_generic(unsigned long phys, unsigned len) * This can happen with kdump kernels when accessing * firmware tables: */ - if (pfn < end_pfn_map) + if (pfn < max_pfn_mapped) return; printk(KERN_ERR "reserve_bootmem: illegal reserve %lx %u\n", diff --git a/arch/x86/mm/ioremap.c b/arch/x86/mm/ioremap.c index 794895c6dcc9..c590fd200e29 100644 --- a/arch/x86/mm/ioremap.c +++ b/arch/x86/mm/ioremap.c @@ -19,11 +19,7 @@ #include <asm/pgtable.h> #include <asm/tlbflush.h> #include <asm/pgalloc.h> - -enum ioremap_mode { - IOR_MODE_UNCACHED, - IOR_MODE_CACHED, -}; +#include <asm/pat.h> #ifdef CONFIG_X86_64 @@ -35,11 +31,23 @@ unsigned long __phys_addr(unsigned long x) } EXPORT_SYMBOL(__phys_addr); +static inline int phys_addr_valid(unsigned long addr) +{ + return addr < (1UL << boot_cpu_data.x86_phys_bits); +} + +#else + +static inline int phys_addr_valid(unsigned long addr) +{ + return 1; +} + #endif int page_is_ram(unsigned long pagenr) { - unsigned long addr, end; + resource_size_t addr, end; int i; /* @@ -78,19 +86,22 @@ int page_is_ram(unsigned long pagenr) * Fix up the linear direct mapping of the kernel to avoid cache attribute * conflicts. */ -static int ioremap_change_attr(unsigned long vaddr, unsigned long size, - enum ioremap_mode mode) +int ioremap_change_attr(unsigned long vaddr, unsigned long size, + unsigned long prot_val) { unsigned long nrpages = size >> PAGE_SHIFT; int err; - switch (mode) { - case IOR_MODE_UNCACHED: + switch (prot_val) { + case _PAGE_CACHE_UC: default: - err = set_memory_uc(vaddr, nrpages); + err = _set_memory_uc(vaddr, nrpages); + break; + case _PAGE_CACHE_WC: + err = _set_memory_wc(vaddr, nrpages); break; - case IOR_MODE_CACHED: - err = set_memory_wb(vaddr, nrpages); + case _PAGE_CACHE_WB: + err = _set_memory_wb(vaddr, nrpages); break; } @@ -107,17 +118,27 @@ static int ioremap_change_attr(unsigned long vaddr, unsigned long size, * caller shouldn't need to know that small detail. */ static void __iomem *__ioremap(resource_size_t phys_addr, unsigned long size, - enum ioremap_mode mode) + unsigned long prot_val) { - unsigned long pfn, offset, last_addr, vaddr; + unsigned long pfn, offset, vaddr; + resource_size_t last_addr; struct vm_struct *area; + unsigned long new_prot_val; pgprot_t prot; + int retval; /* Don't allow wraparound or zero size */ last_addr = phys_addr + size - 1; if (!size || last_addr < phys_addr) return NULL; + if (!phys_addr_valid(phys_addr)) { + printk(KERN_WARNING "ioremap: invalid physical address %llx\n", + phys_addr); + WARN_ON_ONCE(1); + return NULL; + } + /* * Don't remap the low PCI/ISA area, it's always mapped.. */ @@ -127,25 +148,14 @@ static void __iomem *__ioremap(resource_size_t phys_addr, unsigned long size, /* * Don't allow anybody to remap normal RAM that we're using.. */ - for (pfn = phys_addr >> PAGE_SHIFT; pfn < max_pfn_mapped && - (pfn << PAGE_SHIFT) < last_addr; pfn++) { - if (page_is_ram(pfn) && pfn_valid(pfn) && - !PageReserved(pfn_to_page(pfn))) - return NULL; - } + for (pfn = phys_addr >> PAGE_SHIFT; + (pfn << PAGE_SHIFT) < last_addr; pfn++) { - switch (mode) { - case IOR_MODE_UNCACHED: - default: - /* - * FIXME: we will use UC MINUS for now, as video fb drivers - * depend on it. Upcoming ioremap_wc() will fix this behavior. - */ - prot = PAGE_KERNEL_UC_MINUS; - break; - case IOR_MODE_CACHED: - prot = PAGE_KERNEL; - break; + int is_ram = page_is_ram(pfn); + + if (is_ram && pfn_valid(pfn) && !PageReserved(pfn_to_page(pfn))) + return NULL; + WARN_ON_ONCE(is_ram); } /* @@ -155,6 +165,49 @@ static void __iomem *__ioremap(resource_size_t phys_addr, unsigned long size, phys_addr &= PAGE_MASK; size = PAGE_ALIGN(last_addr+1) - phys_addr; + retval = reserve_memtype(phys_addr, phys_addr + size, + prot_val, &new_prot_val); + if (retval) { + pr_debug("Warning: reserve_memtype returned %d\n", retval); + return NULL; + } + + if (prot_val != new_prot_val) { + /* + * Do not fallback to certain memory types with certain + * requested type: + * - request is uncached, return cannot be write-back + * - request is uncached, return cannot be write-combine + * - request is write-combine, return cannot be write-back + */ + if ((prot_val == _PAGE_CACHE_UC && + (new_prot_val == _PAGE_CACHE_WB || + new_prot_val == _PAGE_CACHE_WC)) || + (prot_val == _PAGE_CACHE_WC && + new_prot_val == _PAGE_CACHE_WB)) { + pr_debug( + "ioremap error for 0x%llx-0x%llx, requested 0x%lx, got 0x%lx\n", + phys_addr, phys_addr + size, + prot_val, new_prot_val); + free_memtype(phys_addr, phys_addr + size); + return NULL; + } + prot_val = new_prot_val; + } + + switch (prot_val) { + case _PAGE_CACHE_UC: + default: + prot = PAGE_KERNEL_NOCACHE; + break; + case _PAGE_CACHE_WC: + prot = PAGE_KERNEL_WC; + break; + case _PAGE_CACHE_WB: + prot = PAGE_KERNEL; + break; + } + /* * Ok, go for it.. */ @@ -164,11 +217,13 @@ static void __iomem *__ioremap(resource_size_t phys_addr, unsigned long size, area->phys_addr = phys_addr; vaddr = (unsigned long) area->addr; if (ioremap_page_range(vaddr, vaddr + size, phys_addr, prot)) { + free_memtype(phys_addr, phys_addr + size); free_vm_area(area); return NULL; } - if (ioremap_change_attr(vaddr, size, mode) < 0) { + if (ioremap_change_attr(vaddr, size, prot_val) < 0) { + free_memtype(phys_addr, phys_addr + size); vunmap(area->addr); return NULL; } @@ -199,13 +254,32 @@ static void __iomem *__ioremap(resource_size_t phys_addr, unsigned long size, */ void __iomem *ioremap_nocache(resource_size_t phys_addr, unsigned long size) { - return __ioremap(phys_addr, size, IOR_MODE_UNCACHED); + return __ioremap(phys_addr, size, _PAGE_CACHE_UC); } EXPORT_SYMBOL(ioremap_nocache); +/** + * ioremap_wc - map memory into CPU space write combined + * @offset: bus address of the memory + * @size: size of the resource to map + * + * This version of ioremap ensures that the memory is marked write combining. + * Write combining allows faster writes to some hardware devices. + * + * Must be freed with iounmap. + */ +void __iomem *ioremap_wc(unsigned long phys_addr, unsigned long size) +{ + if (pat_wc_enabled) + return __ioremap(phys_addr, size, _PAGE_CACHE_WC); + else + return ioremap_nocache(phys_addr, size); +} +EXPORT_SYMBOL(ioremap_wc); + void __iomem *ioremap_cache(resource_size_t phys_addr, unsigned long size) { - return __ioremap(phys_addr, size, IOR_MODE_CACHED); + return __ioremap(phys_addr, size, _PAGE_CACHE_WB); } EXPORT_SYMBOL(ioremap_cache); @@ -252,6 +326,8 @@ void iounmap(volatile void __iomem *addr) return; } + free_memtype(p->phys_addr, p->phys_addr + get_vm_area_size(p)); + /* Finally remove it */ o = remove_vm_area((void *)addr); BUG_ON(p != o || o == NULL); @@ -272,8 +348,8 @@ static int __init early_ioremap_debug_setup(char *str) early_param("early_ioremap_debug", early_ioremap_debug_setup); static __initdata int after_paging_init; -static __initdata pte_t bm_pte[PAGE_SIZE/sizeof(pte_t)] - __attribute__((aligned(PAGE_SIZE))); +static pte_t bm_pte[PAGE_SIZE/sizeof(pte_t)] + __section(.bss.page_aligned); static inline pmd_t * __init early_ioremap_pmd(unsigned long addr) { diff --git a/arch/x86/mm/numa_64.c b/arch/x86/mm/numa_64.c index 16b82ad34b96..2ea56f48f29b 100644 --- a/arch/x86/mm/numa_64.c +++ b/arch/x86/mm/numa_64.c @@ -31,13 +31,15 @@ bootmem_data_t plat_node_bdata[MAX_NUMNODES]; struct memnode memnode; +#ifdef CONFIG_SMP int x86_cpu_to_node_map_init[NR_CPUS] = { [0 ... NR_CPUS-1] = NUMA_NO_NODE }; void *x86_cpu_to_node_map_early_ptr; +EXPORT_SYMBOL(x86_cpu_to_node_map_early_ptr); +#endif DEFINE_PER_CPU(int, x86_cpu_to_node_map) = NUMA_NO_NODE; EXPORT_PER_CPU_SYMBOL(x86_cpu_to_node_map); -EXPORT_SYMBOL(x86_cpu_to_node_map_early_ptr); s16 apicid_to_node[MAX_LOCAL_APIC] __cpuinitdata = { [0 ... MAX_LOCAL_APIC-1] = NUMA_NO_NODE @@ -548,8 +550,6 @@ void __cpuinit numa_set_node(int cpu, int node) { int *cpu_to_node_map = x86_cpu_to_node_map_early_ptr; - cpu_pda(cpu)->nodenumber = node; - if(cpu_to_node_map) cpu_to_node_map[cpu] = node; else if(per_cpu_offset(cpu)) diff --git a/arch/x86/mm/pageattr.c b/arch/x86/mm/pageattr.c index 7b79f6be4e7d..7d9517abc9af 100644 --- a/arch/x86/mm/pageattr.c +++ b/arch/x86/mm/pageattr.c @@ -9,6 +9,8 @@ #include <linux/slab.h> #include <linux/mm.h> #include <linux/interrupt.h> +#include <linux/seq_file.h> +#include <linux/debugfs.h> #include <asm/e820.h> #include <asm/processor.h> @@ -17,6 +19,7 @@ #include <asm/uaccess.h> #include <asm/pgalloc.h> #include <asm/proto.h> +#include <asm/pat.h> /* * The current flushing context - we pass it instead of 5 arguments: @@ -28,6 +31,7 @@ struct cpa_data { int numpages; int flushtlb; unsigned long pfn; + unsigned force_split : 1; }; #ifdef CONFIG_X86_64 @@ -259,6 +263,9 @@ try_preserve_large_page(pte_t *kpte, unsigned long address, int i, do_split = 1; unsigned int level; + if (cpa->force_split) + return 1; + spin_lock_irqsave(&pgd_lock, flags); /* * Check for races, another CPU might have split this page @@ -693,7 +700,8 @@ static inline int cache_attr(pgprot_t attr) } static int change_page_attr_set_clr(unsigned long addr, int numpages, - pgprot_t mask_set, pgprot_t mask_clr) + pgprot_t mask_set, pgprot_t mask_clr, + int force_split) { struct cpa_data cpa; int ret, cache, checkalias; @@ -704,7 +712,7 @@ static int change_page_attr_set_clr(unsigned long addr, int numpages, */ mask_set = canon_pgprot(mask_set); mask_clr = canon_pgprot(mask_clr); - if (!pgprot_val(mask_set) && !pgprot_val(mask_clr)) + if (!pgprot_val(mask_set) && !pgprot_val(mask_clr) && !force_split) return 0; /* Ensure we are PAGE_SIZE aligned */ @@ -721,6 +729,7 @@ static int change_page_attr_set_clr(unsigned long addr, int numpages, cpa.mask_set = mask_set; cpa.mask_clr = mask_clr; cpa.flushtlb = 0; + cpa.force_split = force_split; /* No alias checking for _NX bit modifications */ checkalias = (pgprot_val(mask_set) | pgprot_val(mask_clr)) != _PAGE_NX; @@ -759,26 +768,61 @@ out: static inline int change_page_attr_set(unsigned long addr, int numpages, pgprot_t mask) { - return change_page_attr_set_clr(addr, numpages, mask, __pgprot(0)); + return change_page_attr_set_clr(addr, numpages, mask, __pgprot(0), 0); } static inline int change_page_attr_clear(unsigned long addr, int numpages, pgprot_t mask) { - return change_page_attr_set_clr(addr, numpages, __pgprot(0), mask); + return change_page_attr_set_clr(addr, numpages, __pgprot(0), mask, 0); } -int set_memory_uc(unsigned long addr, int numpages) +int _set_memory_uc(unsigned long addr, int numpages) { return change_page_attr_set(addr, numpages, - __pgprot(_PAGE_PCD)); + __pgprot(_PAGE_CACHE_UC)); +} + +int set_memory_uc(unsigned long addr, int numpages) +{ + if (reserve_memtype(addr, addr + numpages * PAGE_SIZE, + _PAGE_CACHE_UC, NULL)) + return -EINVAL; + + return _set_memory_uc(addr, numpages); } EXPORT_SYMBOL(set_memory_uc); -int set_memory_wb(unsigned long addr, int numpages) +int _set_memory_wc(unsigned long addr, int numpages) +{ + return change_page_attr_set(addr, numpages, + __pgprot(_PAGE_CACHE_WC)); +} + +int set_memory_wc(unsigned long addr, int numpages) +{ + if (!pat_wc_enabled) + return set_memory_uc(addr, numpages); + + if (reserve_memtype(addr, addr + numpages * PAGE_SIZE, + _PAGE_CACHE_WC, NULL)) + return -EINVAL; + + return _set_memory_wc(addr, numpages); +} +EXPORT_SYMBOL(set_memory_wc); + +int _set_memory_wb(unsigned long addr, int numpages) { return change_page_attr_clear(addr, numpages, - __pgprot(_PAGE_PCD | _PAGE_PWT)); + __pgprot(_PAGE_CACHE_MASK)); +} + +int set_memory_wb(unsigned long addr, int numpages) +{ + free_memtype(addr, addr + numpages * PAGE_SIZE); + + return _set_memory_wb(addr, numpages); } EXPORT_SYMBOL(set_memory_wb); @@ -809,6 +853,12 @@ int set_memory_np(unsigned long addr, int numpages) return change_page_attr_clear(addr, numpages, __pgprot(_PAGE_PRESENT)); } +int set_memory_4k(unsigned long addr, int numpages) +{ + return change_page_attr_set_clr(addr, numpages, __pgprot(0), + __pgprot(0), 1); +} + int set_pages_uc(struct page *page, int numpages) { unsigned long addr = (unsigned long)page_address(page); @@ -918,6 +968,45 @@ void kernel_map_pages(struct page *page, int numpages, int enable) cpa_fill_pool(NULL); } +#ifdef CONFIG_DEBUG_FS +static int dpa_show(struct seq_file *m, void *v) +{ + seq_puts(m, "DEBUG_PAGEALLOC\n"); + seq_printf(m, "pool_size : %lu\n", pool_size); + seq_printf(m, "pool_pages : %lu\n", pool_pages); + seq_printf(m, "pool_low : %lu\n", pool_low); + seq_printf(m, "pool_used : %lu\n", pool_used); + seq_printf(m, "pool_failed : %lu\n", pool_failed); + + return 0; +} + +static int dpa_open(struct inode *inode, struct file *filp) +{ + return single_open(filp, dpa_show, NULL); +} + +static const struct file_operations dpa_fops = { + .open = dpa_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; + +int __init debug_pagealloc_proc_init(void) +{ + struct dentry *de; + + de = debugfs_create_file("debug_pagealloc", 0600, NULL, NULL, + &dpa_fops); + if (!de) + return -ENOMEM; + + return 0; +} +__initcall(debug_pagealloc_proc_init); +#endif + #ifdef CONFIG_HIBERNATION bool kernel_page_present(struct page *page) diff --git a/arch/x86/mm/pat.c b/arch/x86/mm/pat.c new file mode 100644 index 000000000000..72c0f6097402 --- /dev/null +++ b/arch/x86/mm/pat.c @@ -0,0 +1,421 @@ +/* + * Handle caching attributes in page tables (PAT) + * + * Authors: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com> + * Suresh B Siddha <suresh.b.siddha@intel.com> + * + * Loosely based on earlier PAT patchset from Eric Biederman and Andi Kleen. + */ + +#include <linux/mm.h> +#include <linux/kernel.h> +#include <linux/gfp.h> +#include <linux/fs.h> + +#include <asm/msr.h> +#include <asm/tlbflush.h> +#include <asm/processor.h> +#include <asm/pgtable.h> +#include <asm/pat.h> +#include <asm/e820.h> +#include <asm/cacheflush.h> +#include <asm/fcntl.h> +#include <asm/mtrr.h> + +int pat_wc_enabled = 1; + +static u64 __read_mostly boot_pat_state; + +static int nopat(char *str) +{ + pat_wc_enabled = 0; + printk(KERN_INFO "x86: PAT support disabled.\n"); + + return 0; +} +early_param("nopat", nopat); + +static int pat_known_cpu(void) +{ + if (!pat_wc_enabled) + return 0; + + if (cpu_has_pat) + return 1; + + pat_wc_enabled = 0; + printk(KERN_INFO "CPU and/or kernel does not support PAT.\n"); + return 0; +} + +enum { + PAT_UC = 0, /* uncached */ + PAT_WC = 1, /* Write combining */ + PAT_WT = 4, /* Write Through */ + PAT_WP = 5, /* Write Protected */ + PAT_WB = 6, /* Write Back (default) */ + PAT_UC_MINUS = 7, /* UC, but can be overriden by MTRR */ +}; + +#define PAT(x,y) ((u64)PAT_ ## y << ((x)*8)) + +void pat_init(void) +{ + u64 pat; + +#ifndef CONFIG_X86_PAT + nopat(NULL); +#endif + + /* Boot CPU enables PAT based on CPU feature */ + if (!smp_processor_id() && !pat_known_cpu()) + return; + + /* APs enable PAT iff boot CPU has enabled it before */ + if (smp_processor_id() && !pat_wc_enabled) + return; + + /* Set PWT to Write-Combining. All other bits stay the same */ + /* + * PTE encoding used in Linux: + * PAT + * |PCD + * ||PWT + * ||| + * 000 WB _PAGE_CACHE_WB + * 001 WC _PAGE_CACHE_WC + * 010 UC- _PAGE_CACHE_UC_MINUS + * 011 UC _PAGE_CACHE_UC + * PAT bit unused + */ + pat = PAT(0,WB) | PAT(1,WC) | PAT(2,UC_MINUS) | PAT(3,UC) | + PAT(4,WB) | PAT(5,WC) | PAT(6,UC_MINUS) | PAT(7,UC); + + /* Boot CPU check */ + if (!smp_processor_id()) { + rdmsrl(MSR_IA32_CR_PAT, boot_pat_state); + } + + wrmsrl(MSR_IA32_CR_PAT, pat); + printk(KERN_INFO "x86 PAT enabled: cpu %d, old 0x%Lx, new 0x%Lx\n", + smp_processor_id(), boot_pat_state, pat); +} + +#undef PAT + +static char *cattr_name(unsigned long flags) +{ + switch (flags & _PAGE_CACHE_MASK) { + case _PAGE_CACHE_UC: return "uncached"; + case _PAGE_CACHE_UC_MINUS: return "uncached-minus"; + case _PAGE_CACHE_WB: return "write-back"; + case _PAGE_CACHE_WC: return "write-combining"; + default: return "broken"; + } +} + +/* + * The global memtype list keeps track of memory type for specific + * physical memory areas. Conflicting memory types in different + * mappings can cause CPU cache corruption. To avoid this we keep track. + * + * The list is sorted based on starting address and can contain multiple + * entries for each address (this allows reference counting for overlapping + * areas). All the aliases have the same cache attributes of course. + * Zero attributes are represented as holes. + * + * Currently the data structure is a list because the number of mappings + * are expected to be relatively small. If this should be a problem + * it could be changed to a rbtree or similar. + * + * memtype_lock protects the whole list. + */ + +struct memtype { + u64 start; + u64 end; + unsigned long type; + struct list_head nd; +}; + +static LIST_HEAD(memtype_list); +static DEFINE_SPINLOCK(memtype_lock); /* protects memtype list */ + +/* + * Does intersection of PAT memory type and MTRR memory type and returns + * the resulting memory type as PAT understands it. + * (Type in pat and mtrr will not have same value) + * The intersection is based on "Effective Memory Type" tables in IA-32 + * SDM vol 3a + */ +static int pat_x_mtrr_type(u64 start, u64 end, unsigned long prot, + unsigned long *ret_prot) +{ + unsigned long pat_type; + u8 mtrr_type; + + mtrr_type = mtrr_type_lookup(start, end); + if (mtrr_type == 0xFF) { /* MTRR not enabled */ + *ret_prot = prot; + return 0; + } + if (mtrr_type == 0xFE) { /* MTRR match error */ + *ret_prot = _PAGE_CACHE_UC; + return -1; + } + if (mtrr_type != MTRR_TYPE_UNCACHABLE && + mtrr_type != MTRR_TYPE_WRBACK && + mtrr_type != MTRR_TYPE_WRCOMB) { /* MTRR type unhandled */ + *ret_prot = _PAGE_CACHE_UC; + return -1; + } + + pat_type = prot & _PAGE_CACHE_MASK; + prot &= (~_PAGE_CACHE_MASK); + + /* Currently doing intersection by hand. Optimize it later. */ + if (pat_type == _PAGE_CACHE_WC) { + *ret_prot = prot | _PAGE_CACHE_WC; + } else if (pat_type == _PAGE_CACHE_UC_MINUS) { + *ret_prot = prot | _PAGE_CACHE_UC_MINUS; + } else if (pat_type == _PAGE_CACHE_UC || + mtrr_type == MTRR_TYPE_UNCACHABLE) { + *ret_prot = prot | _PAGE_CACHE_UC; + } else if (mtrr_type == MTRR_TYPE_WRCOMB) { + *ret_prot = prot | _PAGE_CACHE_WC; + } else { + *ret_prot = prot | _PAGE_CACHE_WB; + } + + return 0; +} + +int reserve_memtype(u64 start, u64 end, unsigned long req_type, + unsigned long *ret_type) +{ + struct memtype *new_entry = NULL; + struct memtype *parse; + unsigned long actual_type; + int err = 0; + + /* Only track when pat_wc_enabled */ + if (!pat_wc_enabled) { + if (ret_type) + *ret_type = req_type; + + return 0; + } + + /* Low ISA region is always mapped WB in page table. No need to track */ + if (start >= ISA_START_ADDRESS && (end - 1) <= ISA_END_ADDRESS) { + if (ret_type) + *ret_type = _PAGE_CACHE_WB; + + return 0; + } + + req_type &= _PAGE_CACHE_MASK; + err = pat_x_mtrr_type(start, end, req_type, &actual_type); + if (err) { + if (ret_type) + *ret_type = actual_type; + + return -EINVAL; + } + + new_entry = kmalloc(sizeof(struct memtype), GFP_KERNEL); + if (!new_entry) + return -ENOMEM; + + new_entry->start = start; + new_entry->end = end; + new_entry->type = actual_type; + + if (ret_type) + *ret_type = actual_type; + + spin_lock(&memtype_lock); + + /* Search for existing mapping that overlaps the current range */ + list_for_each_entry(parse, &memtype_list, nd) { + struct memtype *saved_ptr; + + if (parse->start >= end) { + printk("New Entry\n"); + list_add(&new_entry->nd, parse->nd.prev); + new_entry = NULL; + break; + } + + if (start <= parse->start && end >= parse->start) { + if (actual_type != parse->type && ret_type) { + actual_type = parse->type; + *ret_type = actual_type; + new_entry->type = actual_type; + } + + if (actual_type != parse->type) { + printk( + KERN_INFO "%s:%d conflicting memory types %Lx-%Lx %s<->%s\n", + current->comm, current->pid, + start, end, + cattr_name(actual_type), + cattr_name(parse->type)); + err = -EBUSY; + break; + } + + saved_ptr = parse; + /* + * Check to see whether the request overlaps more + * than one entry in the list + */ + list_for_each_entry_continue(parse, &memtype_list, nd) { + if (end <= parse->start) { + break; + } + + if (actual_type != parse->type) { + printk( + KERN_INFO "%s:%d conflicting memory types %Lx-%Lx %s<->%s\n", + current->comm, current->pid, + start, end, + cattr_name(actual_type), + cattr_name(parse->type)); + err = -EBUSY; + break; + } + } + + if (err) { + break; + } + + printk("Overlap at 0x%Lx-0x%Lx\n", + saved_ptr->start, saved_ptr->end); + /* No conflict. Go ahead and add this new entry */ + list_add(&new_entry->nd, saved_ptr->nd.prev); + new_entry = NULL; + break; + } + + if (start < parse->end) { + if (actual_type != parse->type && ret_type) { + actual_type = parse->type; + *ret_type = actual_type; + new_entry->type = actual_type; + } + + if (actual_type != parse->type) { + printk( + KERN_INFO "%s:%d conflicting memory types %Lx-%Lx %s<->%s\n", + current->comm, current->pid, + start, end, + cattr_name(actual_type), + cattr_name(parse->type)); + err = -EBUSY; + break; + } + + saved_ptr = parse; + /* + * Check to see whether the request overlaps more + * than one entry in the list + */ + list_for_each_entry_continue(parse, &memtype_list, nd) { + if (end <= parse->start) { + break; + } + + if (actual_type != parse->type) { + printk( + KERN_INFO "%s:%d conflicting memory types %Lx-%Lx %s<->%s\n", + current->comm, current->pid, + start, end, + cattr_name(actual_type), + cattr_name(parse->type)); + err = -EBUSY; + break; + } + } + + if (err) { + break; + } + + printk("Overlap at 0x%Lx-0x%Lx\n", + saved_ptr->start, saved_ptr->end); + /* No conflict. Go ahead and add this new entry */ + list_add(&new_entry->nd, &saved_ptr->nd); + new_entry = NULL; + break; + } + } + + if (err) { + printk( + "reserve_memtype failed 0x%Lx-0x%Lx, track %s, req %s\n", + start, end, cattr_name(new_entry->type), + cattr_name(req_type)); + kfree(new_entry); + spin_unlock(&memtype_lock); + return err; + } + + if (new_entry) { + /* No conflict. Not yet added to the list. Add to the tail */ + list_add_tail(&new_entry->nd, &memtype_list); + printk("New Entry\n"); + } + + if (ret_type) { + printk( + "reserve_memtype added 0x%Lx-0x%Lx, track %s, req %s, ret %s\n", + start, end, cattr_name(actual_type), + cattr_name(req_type), cattr_name(*ret_type)); + } else { + printk( + "reserve_memtype added 0x%Lx-0x%Lx, track %s, req %s\n", + start, end, cattr_name(actual_type), + cattr_name(req_type)); + } + + spin_unlock(&memtype_lock); + return err; +} + +int free_memtype(u64 start, u64 end) +{ + struct memtype *ml; + int err = -EINVAL; + + /* Only track when pat_wc_enabled */ + if (!pat_wc_enabled) { + return 0; + } + + /* Low ISA region is always mapped WB. No need to track */ + if (start >= ISA_START_ADDRESS && end <= ISA_END_ADDRESS) { + return 0; + } + + spin_lock(&memtype_lock); + list_for_each_entry(ml, &memtype_list, nd) { + if (ml->start == start && ml->end == end) { + list_del(&ml->nd); + kfree(ml); + err = 0; + break; + } + } + spin_unlock(&memtype_lock); + + if (err) { + printk(KERN_DEBUG "%s:%d freeing invalid memtype %Lx-%Lx\n", + current->comm, current->pid, start, end); + } + + printk( "free_memtype request 0x%Lx-0x%Lx\n", start, end); + return err; +} + diff --git a/arch/x86/mm/pgtable_32.c b/arch/x86/mm/pgtable_32.c index 2f9e9afcb9f4..3165ec0672bd 100644 --- a/arch/x86/mm/pgtable_32.c +++ b/arch/x86/mm/pgtable_32.c @@ -36,7 +36,6 @@ void show_mem(void) printk(KERN_INFO "Mem-info:\n"); show_free_areas(); - printk(KERN_INFO "Free swap: %6ldkB\n", nr_swap_pages<<(PAGE_SHIFT-10)); for_each_online_pgdat(pgdat) { pgdat_resize_lock(pgdat, &flags); for (i = 0; i < pgdat->node_spanned_pages; ++i) { @@ -381,3 +380,10 @@ void __pmd_free_tlb(struct mmu_gather *tlb, pmd_t *pmd) } #endif + +int pmd_bad(pmd_t pmd) +{ + WARN_ON_ONCE(pmd_bad_v1(pmd) != pmd_bad_v2(pmd)); + + return pmd_bad_v1(pmd); +} diff --git a/arch/x86/mm/srat_64.c b/arch/x86/mm/srat_64.c index 845001c617cc..1bae9c855ceb 100644 --- a/arch/x86/mm/srat_64.c +++ b/arch/x86/mm/srat_64.c @@ -20,6 +20,7 @@ #include <asm/proto.h> #include <asm/numa.h> #include <asm/e820.h> +#include <asm/genapic.h> int acpi_numa __initdata; @@ -132,7 +133,6 @@ acpi_numa_processor_affinity_init(struct acpi_srat_cpu_affinity *pa) int pxm, node; int apic_id; - apic_id = pa->apic_id; if (srat_disabled()) return; if (pa->header.length != sizeof(struct acpi_srat_cpu_affinity)) { @@ -148,6 +148,11 @@ acpi_numa_processor_affinity_init(struct acpi_srat_cpu_affinity *pa) bad_srat(); return; } + + if (is_uv_system()) + apic_id = (pa->apic_id << 8) | pa->local_sapic_eid; + else + apic_id = pa->apic_id; apicid_to_node[apic_id] = node; acpi_numa = 1; printk(KERN_INFO "SRAT: PXM %u -> APIC %u -> Node %u\n", diff --git a/arch/x86/oprofile/init.c b/arch/x86/oprofile/init.c index 5341d481d92f..cdfe4c54deca 100644 --- a/arch/x86/oprofile/init.c +++ b/arch/x86/oprofile/init.c @@ -10,18 +10,19 @@ #include <linux/oprofile.h> #include <linux/init.h> #include <linux/errno.h> - -/* We support CPUs that have performance counters like the Pentium Pro + +/* + * We support CPUs that have performance counters like the Pentium Pro * with the NMI mode driver. */ - -extern int op_nmi_init(struct oprofile_operations * ops); -extern int op_nmi_timer_init(struct oprofile_operations * ops); + +extern int op_nmi_init(struct oprofile_operations *ops); +extern int op_nmi_timer_init(struct oprofile_operations *ops); extern void op_nmi_exit(void); extern void x86_backtrace(struct pt_regs * const regs, unsigned int depth); -int __init oprofile_arch_init(struct oprofile_operations * ops) +int __init oprofile_arch_init(struct oprofile_operations *ops) { int ret; diff --git a/arch/x86/oprofile/nmi_timer_int.c b/arch/x86/oprofile/nmi_timer_int.c index 1418e36ae7ab..e3ecb71b5790 100644 --- a/arch/x86/oprofile/nmi_timer_int.c +++ b/arch/x86/oprofile/nmi_timer_int.c @@ -17,14 +17,14 @@ #include <asm/nmi.h> #include <asm/apic.h> #include <asm/ptrace.h> - + static int profile_timer_exceptions_notify(struct notifier_block *self, unsigned long val, void *data) { struct die_args *args = (struct die_args *)data; int ret = NOTIFY_DONE; - switch(val) { + switch (val) { case DIE_NMI: oprofile_add_sample(args->regs, 0); ret = NOTIFY_STOP; @@ -56,7 +56,7 @@ static void timer_stop(void) } -int __init op_nmi_timer_init(struct oprofile_operations * ops) +int __init op_nmi_timer_init(struct oprofile_operations *ops) { if ((nmi_watchdog != NMI_IO_APIC) || (atomic_read(&nmi_active) <= 0)) return -ENODEV; diff --git a/arch/x86/oprofile/op_model_athlon.c b/arch/x86/oprofile/op_model_athlon.c index c3ee43333f26..3d534879a9dc 100644 --- a/arch/x86/oprofile/op_model_athlon.c +++ b/arch/x86/oprofile/op_model_athlon.c @@ -1,4 +1,4 @@ -/** +/* * @file op_model_athlon.h * athlon / K7 / K8 / Family 10h model-specific MSR operations * @@ -14,28 +14,28 @@ #include <asm/ptrace.h> #include <asm/msr.h> #include <asm/nmi.h> - + #include "op_x86_model.h" #include "op_counter.h" #define NUM_COUNTERS 4 #define NUM_CONTROLS 4 -#define CTR_IS_RESERVED(msrs,c) (msrs->counters[(c)].addr ? 1 : 0) -#define CTR_READ(l,h,msrs,c) do {rdmsr(msrs->counters[(c)].addr, (l), (h));} while (0) -#define CTR_WRITE(l,msrs,c) do {wrmsr(msrs->counters[(c)].addr, -(unsigned int)(l), -1);} while (0) +#define CTR_IS_RESERVED(msrs, c) (msrs->counters[(c)].addr ? 1 : 0) +#define CTR_READ(l, h, msrs, c) do {rdmsr(msrs->counters[(c)].addr, (l), (h)); } while (0) +#define CTR_WRITE(l, msrs, c) do {wrmsr(msrs->counters[(c)].addr, -(unsigned int)(l), -1); } while (0) #define CTR_OVERFLOWED(n) (!((n) & (1U<<31))) -#define CTRL_IS_RESERVED(msrs,c) (msrs->controls[(c)].addr ? 1 : 0) -#define CTRL_READ(l,h,msrs,c) do {rdmsr(msrs->controls[(c)].addr, (l), (h));} while (0) -#define CTRL_WRITE(l,h,msrs,c) do {wrmsr(msrs->controls[(c)].addr, (l), (h));} while (0) +#define CTRL_IS_RESERVED(msrs, c) (msrs->controls[(c)].addr ? 1 : 0) +#define CTRL_READ(l, h, msrs, c) do {rdmsr(msrs->controls[(c)].addr, (l), (h)); } while (0) +#define CTRL_WRITE(l, h, msrs, c) do {wrmsr(msrs->controls[(c)].addr, (l), (h)); } while (0) #define CTRL_SET_ACTIVE(n) (n |= (1<<22)) #define CTRL_SET_INACTIVE(n) (n &= ~(1<<22)) #define CTRL_CLEAR_LO(x) (x &= (1<<21)) #define CTRL_CLEAR_HI(x) (x &= 0xfffffcf0) #define CTRL_SET_ENABLE(val) (val |= 1<<20) -#define CTRL_SET_USR(val,u) (val |= ((u & 1) << 16)) -#define CTRL_SET_KERN(val,k) (val |= ((k & 1) << 17)) +#define CTRL_SET_USR(val, u) (val |= ((u & 1) << 16)) +#define CTRL_SET_KERN(val, k) (val |= ((k & 1) << 17)) #define CTRL_SET_UM(val, m) (val |= (m << 8)) #define CTRL_SET_EVENT_LOW(val, e) (val |= (e & 0xff)) #define CTRL_SET_EVENT_HIGH(val, e) (val |= ((e >> 8) & 0xf)) @@ -43,19 +43,19 @@ #define CTRL_SET_GUEST_ONLY(val, h) (val |= ((h & 1) << 8)) static unsigned long reset_value[NUM_COUNTERS]; - + static void athlon_fill_in_addresses(struct op_msrs * const msrs) { int i; - for (i=0; i < NUM_COUNTERS; i++) { + for (i = 0; i < NUM_COUNTERS; i++) { if (reserve_perfctr_nmi(MSR_K7_PERFCTR0 + i)) msrs->counters[i].addr = MSR_K7_PERFCTR0 + i; else msrs->counters[i].addr = 0; } - for (i=0; i < NUM_CONTROLS; i++) { + for (i = 0; i < NUM_CONTROLS; i++) { if (reserve_evntsel_nmi(MSR_K7_EVNTSEL0 + i)) msrs->controls[i].addr = MSR_K7_EVNTSEL0 + i; else @@ -63,15 +63,15 @@ static void athlon_fill_in_addresses(struct op_msrs * const msrs) } } - + static void athlon_setup_ctrs(struct op_msrs const * const msrs) { unsigned int low, high; int i; - + /* clear all counters */ for (i = 0 ; i < NUM_CONTROLS; ++i) { - if (unlikely(!CTRL_IS_RESERVED(msrs,i))) + if (unlikely(!CTRL_IS_RESERVED(msrs, i))) continue; CTRL_READ(low, high, msrs, i); CTRL_CLEAR_LO(low); @@ -81,14 +81,14 @@ static void athlon_setup_ctrs(struct op_msrs const * const msrs) /* avoid a false detection of ctr overflows in NMI handler */ for (i = 0; i < NUM_COUNTERS; ++i) { - if (unlikely(!CTR_IS_RESERVED(msrs,i))) + if (unlikely(!CTR_IS_RESERVED(msrs, i))) continue; CTR_WRITE(1, msrs, i); } /* enable active counters */ for (i = 0; i < NUM_COUNTERS; ++i) { - if ((counter_config[i].enabled) && (CTR_IS_RESERVED(msrs,i))) { + if ((counter_config[i].enabled) && (CTR_IS_RESERVED(msrs, i))) { reset_value[i] = counter_config[i].count; CTR_WRITE(counter_config[i].count, msrs, i); @@ -112,7 +112,7 @@ static void athlon_setup_ctrs(struct op_msrs const * const msrs) } } - + static int athlon_check_ctrs(struct pt_regs * const regs, struct op_msrs const * const msrs) { @@ -133,7 +133,7 @@ static int athlon_check_ctrs(struct pt_regs * const regs, return 1; } - + static void athlon_start(struct op_msrs const * const msrs) { unsigned int low, high; @@ -150,7 +150,7 @@ static void athlon_start(struct op_msrs const * const msrs) static void athlon_stop(struct op_msrs const * const msrs) { - unsigned int low,high; + unsigned int low, high; int i; /* Subtle: stop on all counters to avoid race with @@ -169,11 +169,11 @@ static void athlon_shutdown(struct op_msrs const * const msrs) int i; for (i = 0 ; i < NUM_COUNTERS ; ++i) { - if (CTR_IS_RESERVED(msrs,i)) + if (CTR_IS_RESERVED(msrs, i)) release_perfctr_nmi(MSR_K7_PERFCTR0 + i); } for (i = 0 ; i < NUM_CONTROLS ; ++i) { - if (CTRL_IS_RESERVED(msrs,i)) + if (CTRL_IS_RESERVED(msrs, i)) release_evntsel_nmi(MSR_K7_EVNTSEL0 + i); } } diff --git a/arch/x86/oprofile/op_model_ppro.c b/arch/x86/oprofile/op_model_ppro.c index c554f52cb808..eff431f6c57b 100644 --- a/arch/x86/oprofile/op_model_ppro.c +++ b/arch/x86/oprofile/op_model_ppro.c @@ -1,4 +1,4 @@ -/** +/* * @file op_model_ppro.h * pentium pro / P6 model-specific MSR operations * @@ -15,45 +15,45 @@ #include <asm/msr.h> #include <asm/apic.h> #include <asm/nmi.h> - + #include "op_x86_model.h" #include "op_counter.h" #define NUM_COUNTERS 2 #define NUM_CONTROLS 2 -#define CTR_IS_RESERVED(msrs,c) (msrs->counters[(c)].addr ? 1 : 0) -#define CTR_READ(l,h,msrs,c) do {rdmsr(msrs->counters[(c)].addr, (l), (h));} while (0) -#define CTR_32BIT_WRITE(l,msrs,c) \ - do {wrmsr(msrs->counters[(c)].addr, -(u32)(l), 0);} while (0) +#define CTR_IS_RESERVED(msrs, c) (msrs->counters[(c)].addr ? 1 : 0) +#define CTR_READ(l, h, msrs, c) do {rdmsr(msrs->counters[(c)].addr, (l), (h)); } while (0) +#define CTR_32BIT_WRITE(l, msrs, c) \ + do {wrmsr(msrs->counters[(c)].addr, -(u32)(l), 0); } while (0) #define CTR_OVERFLOWED(n) (!((n) & (1U<<31))) -#define CTRL_IS_RESERVED(msrs,c) (msrs->controls[(c)].addr ? 1 : 0) -#define CTRL_READ(l,h,msrs,c) do {rdmsr((msrs->controls[(c)].addr), (l), (h));} while (0) -#define CTRL_WRITE(l,h,msrs,c) do {wrmsr((msrs->controls[(c)].addr), (l), (h));} while (0) +#define CTRL_IS_RESERVED(msrs, c) (msrs->controls[(c)].addr ? 1 : 0) +#define CTRL_READ(l, h, msrs, c) do {rdmsr((msrs->controls[(c)].addr), (l), (h)); } while (0) +#define CTRL_WRITE(l, h, msrs, c) do {wrmsr((msrs->controls[(c)].addr), (l), (h)); } while (0) #define CTRL_SET_ACTIVE(n) (n |= (1<<22)) #define CTRL_SET_INACTIVE(n) (n &= ~(1<<22)) #define CTRL_CLEAR(x) (x &= (1<<21)) #define CTRL_SET_ENABLE(val) (val |= 1<<20) -#define CTRL_SET_USR(val,u) (val |= ((u & 1) << 16)) -#define CTRL_SET_KERN(val,k) (val |= ((k & 1) << 17)) +#define CTRL_SET_USR(val, u) (val |= ((u & 1) << 16)) +#define CTRL_SET_KERN(val, k) (val |= ((k & 1) << 17)) #define CTRL_SET_UM(val, m) (val |= (m << 8)) #define CTRL_SET_EVENT(val, e) (val |= e) static unsigned long reset_value[NUM_COUNTERS]; - + static void ppro_fill_in_addresses(struct op_msrs * const msrs) { int i; - for (i=0; i < NUM_COUNTERS; i++) { + for (i = 0; i < NUM_COUNTERS; i++) { if (reserve_perfctr_nmi(MSR_P6_PERFCTR0 + i)) msrs->counters[i].addr = MSR_P6_PERFCTR0 + i; else msrs->counters[i].addr = 0; } - - for (i=0; i < NUM_CONTROLS; i++) { + + for (i = 0; i < NUM_CONTROLS; i++) { if (reserve_evntsel_nmi(MSR_P6_EVNTSEL0 + i)) msrs->controls[i].addr = MSR_P6_EVNTSEL0 + i; else @@ -69,23 +69,23 @@ static void ppro_setup_ctrs(struct op_msrs const * const msrs) /* clear all counters */ for (i = 0 ; i < NUM_CONTROLS; ++i) { - if (unlikely(!CTRL_IS_RESERVED(msrs,i))) + if (unlikely(!CTRL_IS_RESERVED(msrs, i))) continue; CTRL_READ(low, high, msrs, i); CTRL_CLEAR(low); CTRL_WRITE(low, high, msrs, i); } - + /* avoid a false detection of ctr overflows in NMI handler */ for (i = 0; i < NUM_COUNTERS; ++i) { - if (unlikely(!CTR_IS_RESERVED(msrs,i))) + if (unlikely(!CTR_IS_RESERVED(msrs, i))) continue; CTR_32BIT_WRITE(1, msrs, i); } /* enable active counters */ for (i = 0; i < NUM_COUNTERS; ++i) { - if ((counter_config[i].enabled) && (CTR_IS_RESERVED(msrs,i))) { + if ((counter_config[i].enabled) && (CTR_IS_RESERVED(msrs, i))) { reset_value[i] = counter_config[i].count; CTR_32BIT_WRITE(counter_config[i].count, msrs, i); @@ -104,13 +104,13 @@ static void ppro_setup_ctrs(struct op_msrs const * const msrs) } } - + static int ppro_check_ctrs(struct pt_regs * const regs, struct op_msrs const * const msrs) { unsigned int low, high; int i; - + for (i = 0 ; i < NUM_COUNTERS; ++i) { if (!reset_value[i]) continue; @@ -135,10 +135,10 @@ static int ppro_check_ctrs(struct pt_regs * const regs, return 1; } - + static void ppro_start(struct op_msrs const * const msrs) { - unsigned int low,high; + unsigned int low, high; int i; for (i = 0; i < NUM_COUNTERS; ++i) { @@ -153,7 +153,7 @@ static void ppro_start(struct op_msrs const * const msrs) static void ppro_stop(struct op_msrs const * const msrs) { - unsigned int low,high; + unsigned int low, high; int i; for (i = 0; i < NUM_COUNTERS; ++i) { @@ -170,11 +170,11 @@ static void ppro_shutdown(struct op_msrs const * const msrs) int i; for (i = 0 ; i < NUM_COUNTERS ; ++i) { - if (CTR_IS_RESERVED(msrs,i)) + if (CTR_IS_RESERVED(msrs, i)) release_perfctr_nmi(MSR_P6_PERFCTR0 + i); } for (i = 0 ; i < NUM_CONTROLS ; ++i) { - if (CTRL_IS_RESERVED(msrs,i)) + if (CTRL_IS_RESERVED(msrs, i)) release_evntsel_nmi(MSR_P6_EVNTSEL0 + i); } } diff --git a/arch/x86/pci/i386.c b/arch/x86/pci/i386.c index 103b9dff1213..2ead72363077 100644 --- a/arch/x86/pci/i386.c +++ b/arch/x86/pci/i386.c @@ -30,6 +30,9 @@ #include <linux/init.h> #include <linux/ioport.h> #include <linux/errno.h> +#include <linux/bootmem.h> + +#include <asm/pat.h> #include "pci.h" @@ -297,10 +300,35 @@ void pcibios_set_master(struct pci_dev *dev) pci_write_config_byte(dev, PCI_LATENCY_TIMER, lat); } +static void pci_unmap_page_range(struct vm_area_struct *vma) +{ + u64 addr = (u64)vma->vm_pgoff << PAGE_SHIFT; + free_memtype(addr, addr + vma->vm_end - vma->vm_start); +} + +static void pci_track_mmap_page_range(struct vm_area_struct *vma) +{ + u64 addr = (u64)vma->vm_pgoff << PAGE_SHIFT; + unsigned long flags = pgprot_val(vma->vm_page_prot) + & _PAGE_CACHE_MASK; + + reserve_memtype(addr, addr + vma->vm_end - vma->vm_start, flags, NULL); +} + +static struct vm_operations_struct pci_mmap_ops = { + .open = pci_track_mmap_page_range, + .close = pci_unmap_page_range, +}; + int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma, enum pci_mmap_state mmap_state, int write_combine) { unsigned long prot; + u64 addr = vma->vm_pgoff << PAGE_SHIFT; + unsigned long len = vma->vm_end - vma->vm_start; + unsigned long flags; + unsigned long new_flags; + int retval; /* I/O space cannot be accessed via normal processor loads and * stores on this platform. @@ -308,21 +336,50 @@ int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma, if (mmap_state == pci_mmap_io) return -EINVAL; - /* Leave vm_pgoff as-is, the PCI space address is the physical - * address on this platform. - */ prot = pgprot_val(vma->vm_page_prot); - if (boot_cpu_data.x86 > 3) - prot |= _PAGE_PCD | _PAGE_PWT; + if (pat_wc_enabled && write_combine) + prot |= _PAGE_CACHE_WC; + else if (boot_cpu_data.x86 > 3) + prot |= _PAGE_CACHE_UC; + vma->vm_page_prot = __pgprot(prot); - /* Write-combine setting is ignored, it is changed via the mtrr - * interfaces on this platform. - */ + flags = pgprot_val(vma->vm_page_prot) & _PAGE_CACHE_MASK; + retval = reserve_memtype(addr, addr + len, flags, &new_flags); + if (retval) + return retval; + + if (flags != new_flags) { + /* + * Do not fallback to certain memory types with certain + * requested type: + * - request is uncached, return cannot be write-back + * - request is uncached, return cannot be write-combine + * - request is write-combine, return cannot be write-back + */ + if ((flags == _PAGE_CACHE_UC && + (new_flags == _PAGE_CACHE_WB || + new_flags == _PAGE_CACHE_WC)) || + (flags == _PAGE_CACHE_WC && + new_flags == _PAGE_CACHE_WB)) { + free_memtype(addr, addr+len); + return -EINVAL; + } + flags = new_flags; + } + + if (vma->vm_pgoff <= max_pfn_mapped && + ioremap_change_attr((unsigned long)__va(addr), len, flags)) { + free_memtype(addr, addr + len); + return -EINVAL; + } + if (io_remap_pfn_range(vma, vma->vm_start, vma->vm_pgoff, vma->vm_end - vma->vm_start, vma->vm_page_prot)) return -EAGAIN; + vma->vm_ops = &pci_mmap_ops; + return 0; } diff --git a/arch/x86/pci/irq.c b/arch/x86/pci/irq.c index a8715861877e..579745ca6b66 100644 --- a/arch/x86/pci/irq.c +++ b/arch/x86/pci/irq.c @@ -200,7 +200,7 @@ static int pirq_ali_get(struct pci_dev *router, struct pci_dev *dev, int pirq) { static const unsigned char irqmap[16] = { 0, 9, 3, 10, 4, 5, 7, 6, 1, 11, 0, 12, 0, 14, 0, 15 }; - WARN_ON_ONCE(pirq >= 16); + WARN_ON_ONCE(pirq > 16); return irqmap[read_config_nybble(router, 0x48, pirq-1)]; } @@ -209,7 +209,7 @@ static int pirq_ali_set(struct pci_dev *router, struct pci_dev *dev, int pirq, i static const unsigned char irqmap[16] = { 0, 8, 0, 2, 4, 5, 7, 6, 0, 1, 3, 9, 11, 0, 13, 15 }; unsigned int val = irqmap[irq]; - WARN_ON_ONCE(pirq >= 16); + WARN_ON_ONCE(pirq > 16); if (val) { write_config_nybble(router, 0x48, pirq-1, val); return 1; @@ -260,7 +260,7 @@ static int pirq_via586_get(struct pci_dev *router, struct pci_dev *dev, int pirq { static const unsigned int pirqmap[5] = { 3, 2, 5, 1, 1 }; - WARN_ON_ONCE(pirq >= 5); + WARN_ON_ONCE(pirq > 5); return read_config_nybble(router, 0x55, pirqmap[pirq-1]); } @@ -268,7 +268,7 @@ static int pirq_via586_set(struct pci_dev *router, struct pci_dev *dev, int pirq { static const unsigned int pirqmap[5] = { 3, 2, 5, 1, 1 }; - WARN_ON_ONCE(pirq >= 5); + WARN_ON_ONCE(pirq > 5); write_config_nybble(router, 0x55, pirqmap[pirq-1], irq); return 1; } @@ -282,7 +282,7 @@ static int pirq_ite_get(struct pci_dev *router, struct pci_dev *dev, int pirq) { static const unsigned char pirqmap[4] = { 1, 0, 2, 3 }; - WARN_ON_ONCE(pirq >= 4); + WARN_ON_ONCE(pirq > 4); return read_config_nybble(router,0x43, pirqmap[pirq-1]); } @@ -290,7 +290,7 @@ static int pirq_ite_set(struct pci_dev *router, struct pci_dev *dev, int pirq, i { static const unsigned char pirqmap[4] = { 1, 0, 2, 3 }; - WARN_ON_ONCE(pirq >= 4); + WARN_ON_ONCE(pirq > 4); write_config_nybble(router, 0x43, pirqmap[pirq-1], irq); return 1; } diff --git a/arch/x86/pci/numa.c b/arch/x86/pci/numa.c index 55270c26237c..d9afbae5092b 100644 --- a/arch/x86/pci/numa.c +++ b/arch/x86/pci/numa.c @@ -11,11 +11,41 @@ #define XQUAD_PORTIO_BASE 0xfe400000 #define XQUAD_PORTIO_QUAD 0x40000 /* 256k per quad. */ +int mp_bus_id_to_node[MAX_MP_BUSSES]; #define BUS2QUAD(global) (mp_bus_id_to_node[global]) + +int mp_bus_id_to_local[MAX_MP_BUSSES]; #define BUS2LOCAL(global) (mp_bus_id_to_local[global]) + +void mpc_oem_bus_info(struct mpc_config_bus *m, char *name, + struct mpc_config_translation *translation) +{ + int quad = translation->trans_quad; + int local = translation->trans_local; + + mp_bus_id_to_node[m->mpc_busid] = quad; + mp_bus_id_to_local[m->mpc_busid] = local; + printk(KERN_INFO "Bus #%d is %s (node %d)\n", + m->mpc_busid, name, quad); +} + +int quad_local_to_mp_bus_id [NR_CPUS/4][4]; #define QUADLOCAL2BUS(quad,local) (quad_local_to_mp_bus_id[quad][local]) +void mpc_oem_pci_bus(struct mpc_config_bus *m, + struct mpc_config_translation *translation) +{ + int quad = translation->trans_quad; + int local = translation->trans_local; + + quad_local_to_mp_bus_id[quad][local] = m->mpc_busid; +} + +/* Where the IO area was mapped on multiquad, always 0 otherwise */ +void *xquad_portio; +#ifdef CONFIG_X86_NUMAQ +EXPORT_SYMBOL(xquad_portio); +#endif -extern void *xquad_portio; /* Where the IO area was mapped */ #define XQUAD_PORT_ADDR(port, quad) (xquad_portio + (XQUAD_PORTIO_QUAD*quad) + port) #define PCI_CONF1_MQ_ADDRESS(bus, devfn, reg) \ diff --git a/arch/x86/power/cpu_32.c b/arch/x86/power/cpu_32.c index 7f9c6da04a4c..7dc5d5cf50a2 100644 --- a/arch/x86/power/cpu_32.c +++ b/arch/x86/power/cpu_32.c @@ -27,17 +27,17 @@ static void __save_processor_state(struct saved_context *ctxt) /* * descriptor tables */ - store_gdt(&ctxt->gdt); - store_idt(&ctxt->idt); - store_tr(ctxt->tr); + store_gdt(&ctxt->gdt); + store_idt(&ctxt->idt); + store_tr(ctxt->tr); /* * segment registers */ - savesegment(es, ctxt->es); - savesegment(fs, ctxt->fs); - savesegment(gs, ctxt->gs); - savesegment(ss, ctxt->ss); + savesegment(es, ctxt->es); + savesegment(fs, ctxt->fs); + savesegment(gs, ctxt->gs); + savesegment(ss, ctxt->ss); /* * control registers @@ -48,10 +48,12 @@ static void __save_processor_state(struct saved_context *ctxt) ctxt->cr4 = read_cr4(); } +/* Needed by apm.c */ void save_processor_state(void) { __save_processor_state(&saved_context); } +EXPORT_SYMBOL(save_processor_state); static void do_fpu_end(void) { @@ -64,9 +66,14 @@ static void do_fpu_end(void) static void fix_processor_context(void) { int cpu = smp_processor_id(); - struct tss_struct * t = &per_cpu(init_tss, cpu); + struct tss_struct *t = &per_cpu(init_tss, cpu); - set_tss_desc(cpu,t); /* This just modifies memory; should not be necessary. But... This is necessary, because 386 hardware has concept of busy TSS or some similar stupidity. */ + set_tss_desc(cpu, t); /* + * This just modifies memory; should not be + * necessary. But... This is necessary, because + * 386 hardware has concept of busy TSS or some + * similar stupidity. + */ load_TR_desc(); /* This does ltr */ load_LDT(¤t->active_mm->context); /* This does lldt */ @@ -100,16 +107,16 @@ static void __restore_processor_state(struct saved_context *ctxt) * now restore the descriptor tables to their proper values * ltr is done i fix_processor_context(). */ - load_gdt(&ctxt->gdt); - load_idt(&ctxt->idt); + load_gdt(&ctxt->gdt); + load_idt(&ctxt->idt); /* * segment registers */ - loadsegment(es, ctxt->es); - loadsegment(fs, ctxt->fs); - loadsegment(gs, ctxt->gs); - loadsegment(ss, ctxt->ss); + loadsegment(es, ctxt->es); + loadsegment(fs, ctxt->fs); + loadsegment(gs, ctxt->gs); + loadsegment(ss, ctxt->ss); /* * sysenter MSRs @@ -123,11 +130,9 @@ static void __restore_processor_state(struct saved_context *ctxt) mcheck_init(&boot_cpu_data); } +/* Needed by apm.c */ void restore_processor_state(void) { __restore_processor_state(&saved_context); } - -/* Needed by apm.c */ -EXPORT_SYMBOL(save_processor_state); EXPORT_SYMBOL(restore_processor_state); diff --git a/arch/x86/vdso/Makefile b/arch/x86/vdso/Makefile index 0a8f4742ef51..17a6b057856b 100644 --- a/arch/x86/vdso/Makefile +++ b/arch/x86/vdso/Makefile @@ -39,7 +39,7 @@ $(obj)/%.so: $(obj)/%.so.dbg FORCE CFL := $(PROFILING) -mcmodel=small -fPIC -g0 -O2 -fasynchronous-unwind-tables -m64 -$(vobjs): KBUILD_CFLAGS = $(CFL) +$(vobjs): KBUILD_CFLAGS += $(CFL) targets += vdso-syms.lds obj-$(VDSO64-y) += vdso-syms.lds diff --git a/arch/x86/vdso/vdso32-setup.c b/arch/x86/vdso/vdso32-setup.c index 348f1341e1c8..e2af8eee80e3 100644 --- a/arch/x86/vdso/vdso32-setup.c +++ b/arch/x86/vdso/vdso32-setup.c @@ -210,8 +210,12 @@ static int use_sysenter __read_mostly = -1; /* May not be __init: called during resume */ void syscall32_cpu_init(void) { - if (use_sysenter < 0) - use_sysenter = (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL); + if (use_sysenter < 0) { + if (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL) + use_sysenter = 1; + if (boot_cpu_data.x86_vendor == X86_VENDOR_CENTAUR) + use_sysenter = 1; + } /* Load these always in case some future AMD CPU supports SYSENTER from compat mode too. */ @@ -325,6 +329,9 @@ int arch_setup_additional_pages(struct linux_binprm *bprm, int exstack) int ret = 0; bool compat; + if (vdso_enabled == VDSO_DISABLED) + return 0; + down_write(&mm->mmap_sem); /* Test compat mode once here, in case someone diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c index 27ee26aedf94..c0388220cf97 100644 --- a/arch/x86/xen/enlighten.c +++ b/arch/x86/xen/enlighten.c @@ -25,6 +25,7 @@ #include <linux/mm.h> #include <linux/page-flags.h> #include <linux/highmem.h> +#include <linux/console.h> #include <xen/interface/xen.h> #include <xen/interface/physdev.h> @@ -889,7 +890,6 @@ void __init xen_setup_vcpu_info_placement(void) pv_irq_ops.irq_disable = xen_irq_disable_direct; pv_irq_ops.irq_enable = xen_irq_enable_direct; pv_mmu_ops.read_cr2 = xen_read_cr2_direct; - pv_cpu_ops.iret = xen_iret_direct; } } @@ -993,7 +993,7 @@ static const struct pv_cpu_ops xen_cpu_ops __initdata = { .read_tsc = native_read_tsc, .read_pmc = native_read_pmc, - .iret = (void *)&hypercall_page[__HYPERVISOR_iret], + .iret = xen_iret, .irq_enable_syscall_ret = NULL, /* never called */ .load_tr_desc = paravirt_nop, @@ -1228,6 +1228,9 @@ asmlinkage void __init xen_start_kernel(void) ? __pa(xen_start_info->mod_start) : 0; boot_params.hdr.ramdisk_size = xen_start_info->mod_len; + if (!is_initial_xendomain()) + add_preferred_console("hvc", 0, NULL); + /* Start the world */ start_kernel(); } diff --git a/arch/x86/xen/multicalls.c b/arch/x86/xen/multicalls.c index 5e6f36f6d876..5791eb2e3750 100644 --- a/arch/x86/xen/multicalls.c +++ b/arch/x86/xen/multicalls.c @@ -76,7 +76,7 @@ void xen_mc_flush(void) if (ret) { printk(KERN_ERR "%d multicall(s) failed: cpu %d\n", ret, smp_processor_id()); - for(i = 0; i < b->mcidx; i++) { + for (i = 0; i < b->mcidx; i++) { printk(" call %2d/%d: op=%lu arg=[%lx] result=%ld\n", i+1, b->mcidx, b->debug[i].op, @@ -93,7 +93,7 @@ void xen_mc_flush(void) local_irq_restore(flags); - for(i = 0; i < b->cbidx; i++) { + for (i = 0; i < b->cbidx; i++) { struct callback *cb = &b->callbacks[i]; (*cb->fn)(cb->data); diff --git a/arch/x86/xen/smp.c b/arch/x86/xen/smp.c index aafc54437403..e340ff92f6b6 100644 --- a/arch/x86/xen/smp.c +++ b/arch/x86/xen/smp.c @@ -35,7 +35,7 @@ #include "xen-ops.h" #include "mmu.h" -static cpumask_t cpu_initialized_map; +static cpumask_t xen_cpu_initialized_map; static DEFINE_PER_CPU(int, resched_irq); static DEFINE_PER_CPU(int, callfunc_irq); @@ -179,7 +179,7 @@ void __init xen_smp_prepare_cpus(unsigned int max_cpus) if (xen_smp_intr_init(0)) BUG(); - cpu_initialized_map = cpumask_of_cpu(0); + xen_cpu_initialized_map = cpumask_of_cpu(0); /* Restrict the possible_map according to max_cpus. */ while ((num_possible_cpus() > 1) && (num_possible_cpus() > max_cpus)) { @@ -210,7 +210,7 @@ cpu_initialize_context(unsigned int cpu, struct task_struct *idle) struct vcpu_guest_context *ctxt; struct gdt_page *gdt = &per_cpu(gdt_page, cpu); - if (cpu_test_and_set(cpu, cpu_initialized_map)) + if (cpu_test_and_set(cpu, xen_cpu_initialized_map)) return 0; ctxt = kzalloc(sizeof(*ctxt), GFP_KERNEL); diff --git a/arch/x86/xen/xen-asm.S b/arch/x86/xen/xen-asm.S index 6b7190449d07..fe161ed4b01e 100644 --- a/arch/x86/xen/xen-asm.S +++ b/arch/x86/xen/xen-asm.S @@ -135,13 +135,8 @@ ENDPATCH(xen_restore_fl_direct) current stack state in whatever form its in, we keep things simple by only using a single register which is pushed/popped on the stack. - - Non-direct iret could be done in the same way, but it would - require an annoying amount of code duplication. We'll assume - that direct mode will be the common case once the hypervisor - support becomes commonplace. */ -ENTRY(xen_iret_direct) +ENTRY(xen_iret) /* test eflags for special cases */ testl $(X86_EFLAGS_VM | XEN_EFLAGS_NMI), 8(%esp) jnz hyper_iret @@ -155,9 +150,9 @@ ENTRY(xen_iret_direct) GET_THREAD_INFO(%eax) movl TI_cpu(%eax),%eax movl __per_cpu_offset(,%eax,4),%eax - lea per_cpu__xen_vcpu_info(%eax),%eax + mov per_cpu__xen_vcpu(%eax),%eax #else - movl $per_cpu__xen_vcpu_info, %eax + movl per_cpu__xen_vcpu, %eax #endif /* check IF state we're restoring */ diff --git a/arch/x86/xen/xen-ops.h b/arch/x86/xen/xen-ops.h index b02a909bfd4c..956a491ea998 100644 --- a/arch/x86/xen/xen-ops.h +++ b/arch/x86/xen/xen-ops.h @@ -63,5 +63,5 @@ DECL_ASM(void, xen_irq_disable_direct, void); DECL_ASM(unsigned long, xen_save_fl_direct, void); DECL_ASM(void, xen_restore_fl_direct, unsigned long); -void xen_iret_direct(void); +void xen_iret(void); #endif /* XEN_OPS_H */ diff --git a/arch/xtensa/kernel/Makefile b/arch/xtensa/kernel/Makefile index f582d6a24ec2..7419dbccf027 100644 --- a/arch/xtensa/kernel/Makefile +++ b/arch/xtensa/kernel/Makefile @@ -5,7 +5,7 @@ extra-y := head.o vmlinux.lds -obj-y := align.o entry.o irq.o coprocessor.o process.o ptrace.o semaphore.o \ +obj-y := align.o entry.o irq.o coprocessor.o process.o ptrace.o \ setup.o signal.o syscall.o time.o traps.o vectors.o platform.o \ pci-dma.o init_task.o io.o diff --git a/arch/xtensa/kernel/semaphore.c b/arch/xtensa/kernel/semaphore.c deleted file mode 100644 index 995c6410ae10..000000000000 --- a/arch/xtensa/kernel/semaphore.c +++ /dev/null @@ -1,226 +0,0 @@ -/* - * arch/xtensa/kernel/semaphore.c - * - * Generic semaphore code. Buyer beware. Do your own specific changes - * in <asm/semaphore-helper.h> - * - * This file is subject to the terms and conditions of the GNU General Public - * License. See the file "COPYING" in the main directory of this archive - * for more details. - * - * Copyright (C) 2001 - 2005 Tensilica Inc. - * - * Joe Taylor <joe@tensilica.com, joetylr@yahoo.com> - * Chris Zankel <chris@zankel.net> - * Marc Gauthier<marc@tensilica.com, marc@alumni.uwaterloo.ca> - * Kevin Chea - */ - -#include <linux/sched.h> -#include <linux/wait.h> -#include <linux/init.h> -#include <asm/semaphore.h> -#include <asm/errno.h> - -/* - * These two _must_ execute atomically wrt each other. - */ - -static __inline__ void wake_one_more(struct semaphore * sem) -{ - atomic_inc((atomic_t *)&sem->sleepers); -} - -static __inline__ int waking_non_zero(struct semaphore *sem) -{ - unsigned long flags; - int ret = 0; - - spin_lock_irqsave(&semaphore_wake_lock, flags); - if (sem->sleepers > 0) { - sem->sleepers--; - ret = 1; - } - spin_unlock_irqrestore(&semaphore_wake_lock, flags); - return ret; -} - -/* - * waking_non_zero_interruptible: - * 1 got the lock - * 0 go to sleep - * -EINTR interrupted - * - * We must undo the sem->count down_interruptible() increment while we are - * protected by the spinlock in order to make atomic this atomic_inc() with the - * atomic_read() in wake_one_more(), otherwise we can race. -arca - */ - -static __inline__ int waking_non_zero_interruptible(struct semaphore *sem, - struct task_struct *tsk) -{ - unsigned long flags; - int ret = 0; - - spin_lock_irqsave(&semaphore_wake_lock, flags); - if (sem->sleepers > 0) { - sem->sleepers--; - ret = 1; - } else if (signal_pending(tsk)) { - atomic_inc(&sem->count); - ret = -EINTR; - } - spin_unlock_irqrestore(&semaphore_wake_lock, flags); - return ret; -} - -/* - * waking_non_zero_trylock: - * 1 failed to lock - * 0 got the lock - * - * We must undo the sem->count down_trylock() increment while we are - * protected by the spinlock in order to make atomic this atomic_inc() with the - * atomic_read() in wake_one_more(), otherwise we can race. -arca - */ - -static __inline__ int waking_non_zero_trylock(struct semaphore *sem) -{ - unsigned long flags; - int ret = 1; - - spin_lock_irqsave(&semaphore_wake_lock, flags); - if (sem->sleepers <= 0) - atomic_inc(&sem->count); - else { - sem->sleepers--; - ret = 0; - } - spin_unlock_irqrestore(&semaphore_wake_lock, flags); - return ret; -} - -DEFINE_SPINLOCK(semaphore_wake_lock); - -/* - * Semaphores are implemented using a two-way counter: - * The "count" variable is decremented for each process - * that tries to sleep, while the "waking" variable is - * incremented when the "up()" code goes to wake up waiting - * processes. - * - * Notably, the inline "up()" and "down()" functions can - * efficiently test if they need to do any extra work (up - * needs to do something only if count was negative before - * the increment operation. - * - * waking_non_zero() (from asm/semaphore.h) must execute - * atomically. - * - * When __up() is called, the count was negative before - * incrementing it, and we need to wake up somebody. - * - * This routine adds one to the count of processes that need to - * wake up and exit. ALL waiting processes actually wake up but - * only the one that gets to the "waking" field first will gate - * through and acquire the semaphore. The others will go back - * to sleep. - * - * Note that these functions are only called when there is - * contention on the lock, and as such all this is the - * "non-critical" part of the whole semaphore business. The - * critical part is the inline stuff in <asm/semaphore.h> - * where we want to avoid any extra jumps and calls. - */ - -void __up(struct semaphore *sem) -{ - wake_one_more(sem); - wake_up(&sem->wait); -} - -/* - * Perform the "down" function. Return zero for semaphore acquired, - * return negative for signalled out of the function. - * - * If called from __down, the return is ignored and the wait loop is - * not interruptible. This means that a task waiting on a semaphore - * using "down()" cannot be killed until someone does an "up()" on - * the semaphore. - * - * If called from __down_interruptible, the return value gets checked - * upon return. If the return value is negative then the task continues - * with the negative value in the return register (it can be tested by - * the caller). - * - * Either form may be used in conjunction with "up()". - * - */ - -#define DOWN_VAR \ - struct task_struct *tsk = current; \ - wait_queue_t wait; \ - init_waitqueue_entry(&wait, tsk); - -#define DOWN_HEAD(task_state) \ - \ - \ - tsk->state = (task_state); \ - add_wait_queue(&sem->wait, &wait); \ - \ - /* \ - * Ok, we're set up. sem->count is known to be less than zero \ - * so we must wait. \ - * \ - * We can let go the lock for purposes of waiting. \ - * We re-acquire it after awaking so as to protect \ - * all semaphore operations. \ - * \ - * If "up()" is called before we call waking_non_zero() then \ - * we will catch it right away. If it is called later then \ - * we will have to go through a wakeup cycle to catch it. \ - * \ - * Multiple waiters contend for the semaphore lock to see \ - * who gets to gate through and who has to wait some more. \ - */ \ - for (;;) { - -#define DOWN_TAIL(task_state) \ - tsk->state = (task_state); \ - } \ - tsk->state = TASK_RUNNING; \ - remove_wait_queue(&sem->wait, &wait); - -void __sched __down(struct semaphore * sem) -{ - DOWN_VAR - DOWN_HEAD(TASK_UNINTERRUPTIBLE) - if (waking_non_zero(sem)) - break; - schedule(); - DOWN_TAIL(TASK_UNINTERRUPTIBLE) -} - -int __sched __down_interruptible(struct semaphore * sem) -{ - int ret = 0; - DOWN_VAR - DOWN_HEAD(TASK_INTERRUPTIBLE) - - ret = waking_non_zero_interruptible(sem, tsk); - if (ret) - { - if (ret == 1) - /* ret != 0 only if we get interrupted -arca */ - ret = 0; - break; - } - schedule(); - DOWN_TAIL(TASK_INTERRUPTIBLE) - return ret; -} - -int __down_trylock(struct semaphore * sem) -{ - return waking_non_zero_trylock(sem); -} diff --git a/arch/xtensa/kernel/xtensa_ksyms.c b/arch/xtensa/kernel/xtensa_ksyms.c index 60dbdb43fb4c..6e52cdd6166f 100644 --- a/arch/xtensa/kernel/xtensa_ksyms.c +++ b/arch/xtensa/kernel/xtensa_ksyms.c @@ -26,7 +26,6 @@ #include <asm/io.h> #include <asm/page.h> #include <asm/pgalloc.h> -#include <asm/semaphore.h> #ifdef CONFIG_BLK_DEV_FD #include <asm/floppy.h> #endif @@ -71,14 +70,6 @@ EXPORT_SYMBOL(__umodsi3); EXPORT_SYMBOL(__udivdi3); EXPORT_SYMBOL(__umoddi3); -/* - * Semaphore operations - */ -EXPORT_SYMBOL(__down); -EXPORT_SYMBOL(__down_interruptible); -EXPORT_SYMBOL(__down_trylock); -EXPORT_SYMBOL(__up); - #ifdef CONFIG_NET /* * Networking support diff --git a/drivers/acpi/osl.c b/drivers/acpi/osl.c index a697fb6cf050..a498a6cc68fe 100644 --- a/drivers/acpi/osl.c +++ b/drivers/acpi/osl.c @@ -4,6 +4,8 @@ * Copyright (C) 2000 Andrew Henroid * Copyright (C) 2001, 2002 Andy Grover <andrew.grover@intel.com> * Copyright (C) 2001, 2002 Paul Diefenbaugh <paul.s.diefenbaugh@intel.com> + * Copyright (c) 2008 Intel Corporation + * Author: Matthew Wilcox <willy@linux.intel.com> * * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ * @@ -37,15 +39,18 @@ #include <linux/workqueue.h> #include <linux/nmi.h> #include <linux/acpi.h> -#include <acpi/acpi.h> -#include <asm/io.h> -#include <acpi/acpi_bus.h> -#include <acpi/processor.h> -#include <asm/uaccess.h> - #include <linux/efi.h> #include <linux/ioport.h> #include <linux/list.h> +#include <linux/jiffies.h> +#include <linux/semaphore.h> + +#include <asm/io.h> +#include <asm/uaccess.h> + +#include <acpi/acpi.h> +#include <acpi/acpi_bus.h> +#include <acpi/processor.h> #define _COMPONENT ACPI_OS_SERVICES ACPI_MODULE_NAME("osl"); @@ -764,7 +769,6 @@ acpi_os_create_semaphore(u32 max_units, u32 initial_units, acpi_handle * handle) { struct semaphore *sem = NULL; - sem = acpi_os_allocate(sizeof(struct semaphore)); if (!sem) return AE_NO_MEMORY; @@ -791,12 +795,12 @@ acpi_status acpi_os_delete_semaphore(acpi_handle handle) { struct semaphore *sem = (struct semaphore *)handle; - if (!sem) return AE_BAD_PARAMETER; ACPI_DEBUG_PRINT((ACPI_DB_MUTEX, "Deleting semaphore[%p].\n", handle)); + BUG_ON(!list_empty(&sem->wait_list)); kfree(sem); sem = NULL; @@ -804,21 +808,15 @@ acpi_status acpi_os_delete_semaphore(acpi_handle handle) } /* - * TODO: The kernel doesn't have a 'down_timeout' function -- had to - * improvise. The process is to sleep for one scheduler quantum - * until the semaphore becomes available. Downside is that this - * may result in starvation for timeout-based waits when there's - * lots of semaphore activity. - * * TODO: Support for units > 1? */ acpi_status acpi_os_wait_semaphore(acpi_handle handle, u32 units, u16 timeout) { acpi_status status = AE_OK; struct semaphore *sem = (struct semaphore *)handle; + long jiffies; int ret = 0; - if (!sem || (units < 1)) return AE_BAD_PARAMETER; @@ -828,58 +826,14 @@ acpi_status acpi_os_wait_semaphore(acpi_handle handle, u32 units, u16 timeout) ACPI_DEBUG_PRINT((ACPI_DB_MUTEX, "Waiting for semaphore[%p|%d|%d]\n", handle, units, timeout)); - /* - * This can be called during resume with interrupts off. - * Like boot-time, we should be single threaded and will - * always get the lock if we try -- timeout or not. - * If this doesn't succeed, then we will oops courtesy of - * might_sleep() in down(). - */ - if (!down_trylock(sem)) - return AE_OK; - - switch (timeout) { - /* - * No Wait: - * -------- - * A zero timeout value indicates that we shouldn't wait - just - * acquire the semaphore if available otherwise return AE_TIME - * (a.k.a. 'would block'). - */ - case 0: - if (down_trylock(sem)) - status = AE_TIME; - break; - - /* - * Wait Indefinitely: - * ------------------ - */ - case ACPI_WAIT_FOREVER: - down(sem); - break; - - /* - * Wait w/ Timeout: - * ---------------- - */ - default: - // TODO: A better timeout algorithm? - { - int i = 0; - static const int quantum_ms = 1000 / HZ; - - ret = down_trylock(sem); - for (i = timeout; (i > 0 && ret != 0); i -= quantum_ms) { - schedule_timeout_interruptible(1); - ret = down_trylock(sem); - } - - if (ret != 0) - status = AE_TIME; - } - break; - } + if (timeout == ACPI_WAIT_FOREVER) + jiffies = MAX_SCHEDULE_TIMEOUT; + else + jiffies = msecs_to_jiffies(timeout); + + ret = down_timeout(sem, jiffies); + if (ret) + status = AE_TIME; if (ACPI_FAILURE(status)) { ACPI_DEBUG_PRINT((ACPI_DB_MUTEX, @@ -902,7 +856,6 @@ acpi_status acpi_os_signal_semaphore(acpi_handle handle, u32 units) { struct semaphore *sem = (struct semaphore *)handle; - if (!sem || (units < 1)) return AE_BAD_PARAMETER; diff --git a/drivers/char/hw_random/core.c b/drivers/char/hw_random/core.c index 84cdf9025737..349b6edc5794 100644 --- a/drivers/char/hw_random/core.c +++ b/drivers/char/hw_random/core.c @@ -116,6 +116,10 @@ static ssize_t rng_dev_read(struct file *filp, char __user *buf, err = -EAGAIN; if (!bytes_read && (filp->f_flags & O_NONBLOCK)) goto out; + if (bytes_read < 0) { + err = bytes_read; + goto out; + } err = -EFAULT; while (bytes_read && size) { diff --git a/drivers/crypto/Kconfig b/drivers/crypto/Kconfig index 6b658d84d521..6d2f0c8d419a 100644 --- a/drivers/crypto/Kconfig +++ b/drivers/crypto/Kconfig @@ -64,6 +64,7 @@ config ZCRYPT tristate "Support for PCI-attached cryptographic adapters" depends on S390 select ZCRYPT_MONOLITHIC if ZCRYPT="y" + select HW_RANDOM help Select this option if you want to use a PCI-attached cryptographic adapter like: diff --git a/drivers/infiniband/core/cm.c b/drivers/infiniband/core/cm.c index faa7ce318a6d..a47fe64e5c39 100644 --- a/drivers/infiniband/core/cm.c +++ b/drivers/infiniband/core/cm.c @@ -467,6 +467,31 @@ static int cm_compare_private_data(u8 *private_data, return memcmp(src, dst_data->data, IB_CM_COMPARE_SIZE); } +/* + * Trivial helpers to strip endian annotation and compare; the + * endianness doesn't actually matter since we just need a stable + * order for the RB tree. + */ +static int be32_lt(__be32 a, __be32 b) +{ + return (__force u32) a < (__force u32) b; +} + +static int be32_gt(__be32 a, __be32 b) +{ + return (__force u32) a > (__force u32) b; +} + +static int be64_lt(__be64 a, __be64 b) +{ + return (__force u64) a < (__force u64) b; +} + +static int be64_gt(__be64 a, __be64 b) +{ + return (__force u64) a > (__force u64) b; +} + static struct cm_id_private * cm_insert_listen(struct cm_id_private *cm_id_priv) { struct rb_node **link = &cm.listen_service_table.rb_node; @@ -492,9 +517,9 @@ static struct cm_id_private * cm_insert_listen(struct cm_id_private *cm_id_priv) link = &(*link)->rb_left; else if (cm_id_priv->id.device > cur_cm_id_priv->id.device) link = &(*link)->rb_right; - else if (service_id < cur_cm_id_priv->id.service_id) + else if (be64_lt(service_id, cur_cm_id_priv->id.service_id)) link = &(*link)->rb_left; - else if (service_id > cur_cm_id_priv->id.service_id) + else if (be64_gt(service_id, cur_cm_id_priv->id.service_id)) link = &(*link)->rb_right; else if (data_cmp < 0) link = &(*link)->rb_left; @@ -527,9 +552,9 @@ static struct cm_id_private * cm_find_listen(struct ib_device *device, node = node->rb_left; else if (device > cm_id_priv->id.device) node = node->rb_right; - else if (service_id < cm_id_priv->id.service_id) + else if (be64_lt(service_id, cm_id_priv->id.service_id)) node = node->rb_left; - else if (service_id > cm_id_priv->id.service_id) + else if (be64_gt(service_id, cm_id_priv->id.service_id)) node = node->rb_right; else if (data_cmp < 0) node = node->rb_left; @@ -552,13 +577,13 @@ static struct cm_timewait_info * cm_insert_remote_id(struct cm_timewait_info parent = *link; cur_timewait_info = rb_entry(parent, struct cm_timewait_info, remote_id_node); - if (remote_id < cur_timewait_info->work.remote_id) + if (be32_lt(remote_id, cur_timewait_info->work.remote_id)) link = &(*link)->rb_left; - else if (remote_id > cur_timewait_info->work.remote_id) + else if (be32_gt(remote_id, cur_timewait_info->work.remote_id)) link = &(*link)->rb_right; - else if (remote_ca_guid < cur_timewait_info->remote_ca_guid) + else if (be64_lt(remote_ca_guid, cur_timewait_info->remote_ca_guid)) link = &(*link)->rb_left; - else if (remote_ca_guid > cur_timewait_info->remote_ca_guid) + else if (be64_gt(remote_ca_guid, cur_timewait_info->remote_ca_guid)) link = &(*link)->rb_right; else return cur_timewait_info; @@ -578,13 +603,13 @@ static struct cm_timewait_info * cm_find_remote_id(__be64 remote_ca_guid, while (node) { timewait_info = rb_entry(node, struct cm_timewait_info, remote_id_node); - if (remote_id < timewait_info->work.remote_id) + if (be32_lt(remote_id, timewait_info->work.remote_id)) node = node->rb_left; - else if (remote_id > timewait_info->work.remote_id) + else if (be32_gt(remote_id, timewait_info->work.remote_id)) node = node->rb_right; - else if (remote_ca_guid < timewait_info->remote_ca_guid) + else if (be64_lt(remote_ca_guid, timewait_info->remote_ca_guid)) node = node->rb_left; - else if (remote_ca_guid > timewait_info->remote_ca_guid) + else if (be64_gt(remote_ca_guid, timewait_info->remote_ca_guid)) node = node->rb_right; else return timewait_info; @@ -605,13 +630,13 @@ static struct cm_timewait_info * cm_insert_remote_qpn(struct cm_timewait_info parent = *link; cur_timewait_info = rb_entry(parent, struct cm_timewait_info, remote_qp_node); - if (remote_qpn < cur_timewait_info->remote_qpn) + if (be32_lt(remote_qpn, cur_timewait_info->remote_qpn)) link = &(*link)->rb_left; - else if (remote_qpn > cur_timewait_info->remote_qpn) + else if (be32_gt(remote_qpn, cur_timewait_info->remote_qpn)) link = &(*link)->rb_right; - else if (remote_ca_guid < cur_timewait_info->remote_ca_guid) + else if (be64_lt(remote_ca_guid, cur_timewait_info->remote_ca_guid)) link = &(*link)->rb_left; - else if (remote_ca_guid > cur_timewait_info->remote_ca_guid) + else if (be64_gt(remote_ca_guid, cur_timewait_info->remote_ca_guid)) link = &(*link)->rb_right; else return cur_timewait_info; @@ -635,9 +660,9 @@ static struct cm_id_private * cm_insert_remote_sidr(struct cm_id_private parent = *link; cur_cm_id_priv = rb_entry(parent, struct cm_id_private, sidr_id_node); - if (remote_id < cur_cm_id_priv->id.remote_id) + if (be32_lt(remote_id, cur_cm_id_priv->id.remote_id)) link = &(*link)->rb_left; - else if (remote_id > cur_cm_id_priv->id.remote_id) + else if (be32_gt(remote_id, cur_cm_id_priv->id.remote_id)) link = &(*link)->rb_right; else { int cmp; @@ -2848,7 +2873,7 @@ static void cm_format_sidr_req(struct cm_sidr_req_msg *sidr_req_msg, cm_format_mad_hdr(&sidr_req_msg->hdr, CM_SIDR_REQ_ATTR_ID, cm_form_tid(cm_id_priv, CM_MSG_SEQUENCE_SIDR)); sidr_req_msg->request_id = cm_id_priv->id.local_id; - sidr_req_msg->pkey = cpu_to_be16(param->path->pkey); + sidr_req_msg->pkey = param->path->pkey; sidr_req_msg->service_id = param->service_id; if (param->private_data && param->private_data_len) diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c index d81c156a22b4..671f13738054 100644 --- a/drivers/infiniband/core/cma.c +++ b/drivers/infiniband/core/cma.c @@ -1289,7 +1289,7 @@ static int iw_conn_req_handler(struct iw_cm_id *cm_id, new_cm_id = rdma_create_id(listen_id->id.event_handler, listen_id->id.context, RDMA_PS_TCP); - if (!new_cm_id) { + if (IS_ERR(new_cm_id)) { ret = -ENOMEM; goto out; } diff --git a/drivers/infiniband/core/fmr_pool.c b/drivers/infiniband/core/fmr_pool.c index 06d502c06a4d..1286dc1b98b2 100644 --- a/drivers/infiniband/core/fmr_pool.c +++ b/drivers/infiniband/core/fmr_pool.c @@ -158,8 +158,7 @@ static void ib_fmr_batch_release(struct ib_fmr_pool *pool) #endif } - list_splice(&pool->dirty_list, &unmap_list); - INIT_LIST_HEAD(&pool->dirty_list); + list_splice_init(&pool->dirty_list, &unmap_list); pool->dirty_len = 0; spin_unlock_irq(&pool->pool_lock); diff --git a/drivers/infiniband/core/ucma.c b/drivers/infiniband/core/ucma.c index 15937eb38aae..ca4cf3a511ab 100644 --- a/drivers/infiniband/core/ucma.c +++ b/drivers/infiniband/core/ucma.c @@ -614,7 +614,7 @@ static ssize_t ucma_query_route(struct ucma_file *file, if (!ctx->cm_id->device) goto out; - resp.node_guid = ctx->cm_id->device->node_guid; + resp.node_guid = (__force __u64) ctx->cm_id->device->node_guid; resp.port_num = ctx->cm_id->port_num; switch (rdma_node_get_transport(ctx->cm_id->device->node_type)) { case RDMA_TRANSPORT_IB: diff --git a/drivers/infiniband/core/uverbs.h b/drivers/infiniband/core/uverbs.h index c75eb6c9bd49..2cad8b4b5292 100644 --- a/drivers/infiniband/core/uverbs.h +++ b/drivers/infiniband/core/uverbs.h @@ -81,13 +81,13 @@ struct ib_uverbs_device { struct ib_uverbs_event_file { struct kref ref; - struct file *file; struct ib_uverbs_file *uverbs_file; spinlock_t lock; - int is_async; wait_queue_head_t poll_wait; struct fasync_struct *async_queue; struct list_head event_list; + int is_async; + int is_closed; }; struct ib_uverbs_file { diff --git a/drivers/infiniband/core/uverbs_cmd.c b/drivers/infiniband/core/uverbs_cmd.c index 495c803fb11d..2c3bff5fe867 100644 --- a/drivers/infiniband/core/uverbs_cmd.c +++ b/drivers/infiniband/core/uverbs_cmd.c @@ -1065,6 +1065,7 @@ ssize_t ib_uverbs_create_qp(struct ib_uverbs_file *file, attr.srq = srq; attr.sq_sig_type = cmd.sq_sig_all ? IB_SIGNAL_ALL_WR : IB_SIGNAL_REQ_WR; attr.qp_type = cmd.qp_type; + attr.create_flags = 0; attr.cap.max_send_wr = cmd.max_send_wr; attr.cap.max_recv_wr = cmd.max_recv_wr; @@ -1462,7 +1463,6 @@ ssize_t ib_uverbs_post_send(struct ib_uverbs_file *file, next->num_sge = user_wr->num_sge; next->opcode = user_wr->opcode; next->send_flags = user_wr->send_flags; - next->imm_data = (__be32 __force) user_wr->imm_data; if (is_ud) { next->wr.ud.ah = idr_read_ah(user_wr->wr.ud.ah, @@ -1475,14 +1475,24 @@ ssize_t ib_uverbs_post_send(struct ib_uverbs_file *file, next->wr.ud.remote_qkey = user_wr->wr.ud.remote_qkey; } else { switch (next->opcode) { - case IB_WR_RDMA_WRITE: case IB_WR_RDMA_WRITE_WITH_IMM: + next->ex.imm_data = + (__be32 __force) user_wr->ex.imm_data; + case IB_WR_RDMA_WRITE: case IB_WR_RDMA_READ: next->wr.rdma.remote_addr = user_wr->wr.rdma.remote_addr; next->wr.rdma.rkey = user_wr->wr.rdma.rkey; break; + case IB_WR_SEND_WITH_IMM: + next->ex.imm_data = + (__be32 __force) user_wr->ex.imm_data; + break; + case IB_WR_SEND_WITH_INV: + next->ex.invalidate_rkey = + user_wr->ex.invalidate_rkey; + break; case IB_WR_ATOMIC_CMP_AND_SWP: case IB_WR_ATOMIC_FETCH_AND_ADD: next->wr.atomic.remote_addr = diff --git a/drivers/infiniband/core/uverbs_main.c b/drivers/infiniband/core/uverbs_main.c index 7c2ac3905582..f49f94653a96 100644 --- a/drivers/infiniband/core/uverbs_main.c +++ b/drivers/infiniband/core/uverbs_main.c @@ -352,7 +352,7 @@ static int ib_uverbs_event_close(struct inode *inode, struct file *filp) struct ib_uverbs_event *entry, *tmp; spin_lock_irq(&file->lock); - file->file = NULL; + file->is_closed = 1; list_for_each_entry_safe(entry, tmp, &file->event_list, list) { if (entry->counter) list_del(&entry->obj_list); @@ -390,7 +390,7 @@ void ib_uverbs_comp_handler(struct ib_cq *cq, void *cq_context) return; spin_lock_irqsave(&file->lock, flags); - if (!file->file) { + if (file->is_closed) { spin_unlock_irqrestore(&file->lock, flags); return; } @@ -423,7 +423,7 @@ static void ib_uverbs_async_handler(struct ib_uverbs_file *file, unsigned long flags; spin_lock_irqsave(&file->async_file->lock, flags); - if (!file->async_file->file) { + if (!file->async_file->is_closed) { spin_unlock_irqrestore(&file->async_file->lock, flags); return; } @@ -509,6 +509,7 @@ struct file *ib_uverbs_alloc_event_file(struct ib_uverbs_file *uverbs_file, ev_file->uverbs_file = uverbs_file; ev_file->async_queue = NULL; ev_file->is_async = is_async; + ev_file->is_closed = 0; *fd = get_unused_fd(); if (*fd < 0) { @@ -516,25 +517,18 @@ struct file *ib_uverbs_alloc_event_file(struct ib_uverbs_file *uverbs_file, goto err; } - filp = get_empty_filp(); - if (!filp) { - ret = -ENFILE; - goto err_fd; - } - - ev_file->file = filp; - /* * fops_get() can't fail here, because we're coming from a * system call on a uverbs file, which will already have a * module reference. */ - filp->f_op = fops_get(&uverbs_event_fops); - filp->f_path.mnt = mntget(uverbs_event_mnt); - filp->f_path.dentry = dget(uverbs_event_mnt->mnt_root); - filp->f_mapping = filp->f_path.dentry->d_inode->i_mapping; - filp->f_flags = O_RDONLY; - filp->f_mode = FMODE_READ; + filp = alloc_file(uverbs_event_mnt, dget(uverbs_event_mnt->mnt_root), + FMODE_READ, fops_get(&uverbs_event_fops)); + if (!filp) { + ret = -ENFILE; + goto err_fd; + } + filp->private_data = ev_file; return filp; diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c index 86ed8af9c7e6..05042089de6e 100644 --- a/drivers/infiniband/core/verbs.c +++ b/drivers/infiniband/core/verbs.c @@ -248,7 +248,9 @@ int ib_modify_srq(struct ib_srq *srq, struct ib_srq_attr *srq_attr, enum ib_srq_attr_mask srq_attr_mask) { - return srq->device->modify_srq(srq, srq_attr, srq_attr_mask, NULL); + return srq->device->modify_srq ? + srq->device->modify_srq(srq, srq_attr, srq_attr_mask, NULL) : + -ENOSYS; } EXPORT_SYMBOL(ib_modify_srq); @@ -628,6 +630,13 @@ struct ib_cq *ib_create_cq(struct ib_device *device, } EXPORT_SYMBOL(ib_create_cq); +int ib_modify_cq(struct ib_cq *cq, u16 cq_count, u16 cq_period) +{ + return cq->device->modify_cq ? + cq->device->modify_cq(cq, cq_count, cq_period) : -ENOSYS; +} +EXPORT_SYMBOL(ib_modify_cq); + int ib_destroy_cq(struct ib_cq *cq) { if (atomic_read(&cq->usecnt)) @@ -672,6 +681,9 @@ struct ib_mr *ib_reg_phys_mr(struct ib_pd *pd, { struct ib_mr *mr; + if (!pd->device->reg_phys_mr) + return ERR_PTR(-ENOSYS); + mr = pd->device->reg_phys_mr(pd, phys_buf_array, num_phys_buf, mr_access_flags, iova_start); diff --git a/drivers/infiniband/hw/amso1100/c2.c b/drivers/infiniband/hw/amso1100/c2.c index f283a9f0c23b..113f3c03c5b5 100644 --- a/drivers/infiniband/hw/amso1100/c2.c +++ b/drivers/infiniband/hw/amso1100/c2.c @@ -130,10 +130,10 @@ static int c2_tx_ring_alloc(struct c2_ring *tx_ring, void *vaddr, tx_desc->status = 0; /* Set TXP_HTXD_UNINIT */ - __raw_writeq(cpu_to_be64(0x1122334455667788ULL), + __raw_writeq((__force u64) cpu_to_be64(0x1122334455667788ULL), (void __iomem *) txp_desc + C2_TXP_ADDR); __raw_writew(0, (void __iomem *) txp_desc + C2_TXP_LEN); - __raw_writew(cpu_to_be16(TXP_HTXD_UNINIT), + __raw_writew((__force u16) cpu_to_be16(TXP_HTXD_UNINIT), (void __iomem *) txp_desc + C2_TXP_FLAGS); elem->skb = NULL; @@ -179,13 +179,13 @@ static int c2_rx_ring_alloc(struct c2_ring *rx_ring, void *vaddr, rx_desc->status = 0; /* Set RXP_HRXD_UNINIT */ - __raw_writew(cpu_to_be16(RXP_HRXD_OK), + __raw_writew((__force u16) cpu_to_be16(RXP_HRXD_OK), (void __iomem *) rxp_desc + C2_RXP_STATUS); __raw_writew(0, (void __iomem *) rxp_desc + C2_RXP_COUNT); __raw_writew(0, (void __iomem *) rxp_desc + C2_RXP_LEN); - __raw_writeq(cpu_to_be64(0x99aabbccddeeffULL), + __raw_writeq((__force u64) cpu_to_be64(0x99aabbccddeeffULL), (void __iomem *) rxp_desc + C2_RXP_ADDR); - __raw_writew(cpu_to_be16(RXP_HRXD_UNINIT), + __raw_writew((__force u16) cpu_to_be16(RXP_HRXD_UNINIT), (void __iomem *) rxp_desc + C2_RXP_FLAGS); elem->skb = NULL; @@ -239,10 +239,11 @@ static inline int c2_rx_alloc(struct c2_port *c2_port, struct c2_element *elem) rxp_hdr->flags = RXP_HRXD_READY; __raw_writew(0, elem->hw_desc + C2_RXP_STATUS); - __raw_writew(cpu_to_be16((u16) maplen - sizeof(*rxp_hdr)), + __raw_writew((__force u16) cpu_to_be16((u16) maplen - sizeof(*rxp_hdr)), elem->hw_desc + C2_RXP_LEN); - __raw_writeq(cpu_to_be64(mapaddr), elem->hw_desc + C2_RXP_ADDR); - __raw_writew(cpu_to_be16(RXP_HRXD_READY), elem->hw_desc + C2_RXP_FLAGS); + __raw_writeq((__force u64) cpu_to_be64(mapaddr), elem->hw_desc + C2_RXP_ADDR); + __raw_writew((__force u16) cpu_to_be16(RXP_HRXD_READY), + elem->hw_desc + C2_RXP_FLAGS); elem->skb = skb; elem->mapaddr = mapaddr; @@ -290,9 +291,9 @@ static void c2_rx_clean(struct c2_port *c2_port) __raw_writew(0, elem->hw_desc + C2_RXP_STATUS); __raw_writew(0, elem->hw_desc + C2_RXP_COUNT); __raw_writew(0, elem->hw_desc + C2_RXP_LEN); - __raw_writeq(cpu_to_be64(0x99aabbccddeeffULL), + __raw_writeq((__force u64) cpu_to_be64(0x99aabbccddeeffULL), elem->hw_desc + C2_RXP_ADDR); - __raw_writew(cpu_to_be16(RXP_HRXD_UNINIT), + __raw_writew((__force u16) cpu_to_be16(RXP_HRXD_UNINIT), elem->hw_desc + C2_RXP_FLAGS); if (elem->skb) { @@ -346,16 +347,16 @@ static void c2_tx_clean(struct c2_port *c2_port) elem->hw_desc + C2_TXP_LEN); __raw_writeq(0, elem->hw_desc + C2_TXP_ADDR); - __raw_writew(cpu_to_be16(TXP_HTXD_DONE), + __raw_writew((__force u16) cpu_to_be16(TXP_HTXD_DONE), elem->hw_desc + C2_TXP_FLAGS); c2_port->netstats.tx_dropped++; break; } else { __raw_writew(0, elem->hw_desc + C2_TXP_LEN); - __raw_writeq(cpu_to_be64(0x1122334455667788ULL), + __raw_writeq((__force u64) cpu_to_be64(0x1122334455667788ULL), elem->hw_desc + C2_TXP_ADDR); - __raw_writew(cpu_to_be16(TXP_HTXD_UNINIT), + __raw_writew((__force u16) cpu_to_be16(TXP_HTXD_UNINIT), elem->hw_desc + C2_TXP_FLAGS); } @@ -390,7 +391,7 @@ static void c2_tx_interrupt(struct net_device *netdev) for (elem = tx_ring->to_clean; elem != tx_ring->to_use; elem = elem->next) { txp_htxd.flags = - be16_to_cpu(readw(elem->hw_desc + C2_TXP_FLAGS)); + be16_to_cpu((__force __be16) readw(elem->hw_desc + C2_TXP_FLAGS)); if (txp_htxd.flags != TXP_HTXD_DONE) break; @@ -398,7 +399,7 @@ static void c2_tx_interrupt(struct net_device *netdev) if (netif_msg_tx_done(c2_port)) { /* PCI reads are expensive in fast path */ txp_htxd.len = - be16_to_cpu(readw(elem->hw_desc + C2_TXP_LEN)); + be16_to_cpu((__force __be16) readw(elem->hw_desc + C2_TXP_LEN)); pr_debug("%s: tx done slot %3Zu status 0x%x len " "%5u bytes\n", netdev->name, elem - tx_ring->start, @@ -448,10 +449,12 @@ static void c2_rx_error(struct c2_port *c2_port, struct c2_element *elem) /* Write the descriptor to the adapter's rx ring */ __raw_writew(0, elem->hw_desc + C2_RXP_STATUS); __raw_writew(0, elem->hw_desc + C2_RXP_COUNT); - __raw_writew(cpu_to_be16((u16) elem->maplen - sizeof(*rxp_hdr)), + __raw_writew((__force u16) cpu_to_be16((u16) elem->maplen - sizeof(*rxp_hdr)), elem->hw_desc + C2_RXP_LEN); - __raw_writeq(cpu_to_be64(elem->mapaddr), elem->hw_desc + C2_RXP_ADDR); - __raw_writew(cpu_to_be16(RXP_HRXD_READY), elem->hw_desc + C2_RXP_FLAGS); + __raw_writeq((__force u64) cpu_to_be64(elem->mapaddr), + elem->hw_desc + C2_RXP_ADDR); + __raw_writew((__force u16) cpu_to_be16(RXP_HRXD_READY), + elem->hw_desc + C2_RXP_FLAGS); pr_debug("packet dropped\n"); c2_port->netstats.rx_dropped++; @@ -653,7 +656,7 @@ static int c2_up(struct net_device *netdev) i++, elem++) { rxp_hdr = (struct c2_rxp_hdr *) elem->skb->data; rxp_hdr->flags = 0; - __raw_writew(cpu_to_be16(RXP_HRXD_READY), + __raw_writew((__force u16) cpu_to_be16(RXP_HRXD_READY), elem->hw_desc + C2_RXP_FLAGS); } @@ -787,9 +790,12 @@ static int c2_xmit_frame(struct sk_buff *skb, struct net_device *netdev) elem->maplen = maplen; /* Tell HW to xmit */ - __raw_writeq(cpu_to_be64(mapaddr), elem->hw_desc + C2_TXP_ADDR); - __raw_writew(cpu_to_be16(maplen), elem->hw_desc + C2_TXP_LEN); - __raw_writew(cpu_to_be16(TXP_HTXD_READY), elem->hw_desc + C2_TXP_FLAGS); + __raw_writeq((__force u64) cpu_to_be64(mapaddr), + elem->hw_desc + C2_TXP_ADDR); + __raw_writew((__force u16) cpu_to_be16(maplen), + elem->hw_desc + C2_TXP_LEN); + __raw_writew((__force u16) cpu_to_be16(TXP_HTXD_READY), + elem->hw_desc + C2_TXP_FLAGS); c2_port->netstats.tx_packets++; c2_port->netstats.tx_bytes += maplen; @@ -810,11 +816,11 @@ static int c2_xmit_frame(struct sk_buff *skb, struct net_device *netdev) elem->maplen = maplen; /* Tell HW to xmit */ - __raw_writeq(cpu_to_be64(mapaddr), + __raw_writeq((__force u64) cpu_to_be64(mapaddr), elem->hw_desc + C2_TXP_ADDR); - __raw_writew(cpu_to_be16(maplen), + __raw_writew((__force u16) cpu_to_be16(maplen), elem->hw_desc + C2_TXP_LEN); - __raw_writew(cpu_to_be16(TXP_HTXD_READY), + __raw_writew((__force u16) cpu_to_be16(TXP_HTXD_READY), elem->hw_desc + C2_TXP_FLAGS); c2_port->netstats.tx_packets++; @@ -1005,7 +1011,7 @@ static int __devinit c2_probe(struct pci_dev *pcidev, /* Remap the adapter PCI registers in BAR4 */ mmio_regs = ioremap_nocache(reg4_start + C2_PCI_REGS_OFFSET, sizeof(struct c2_adapter_pci_regs)); - if (mmio_regs == 0UL) { + if (!mmio_regs) { printk(KERN_ERR PFX "Unable to remap adapter PCI registers in BAR4\n"); ret = -EIO; @@ -1029,10 +1035,10 @@ static int __devinit c2_probe(struct pci_dev *pcidev, } /* Validate the adapter version */ - if (be32_to_cpu(readl(mmio_regs + C2_REGS_VERS)) != C2_VERSION) { + if (be32_to_cpu((__force __be32) readl(mmio_regs + C2_REGS_VERS)) != C2_VERSION) { printk(KERN_ERR PFX "Version mismatch " "[fw=%u, c2=%u], Adapter not claimed\n", - be32_to_cpu(readl(mmio_regs + C2_REGS_VERS)), + be32_to_cpu((__force __be32) readl(mmio_regs + C2_REGS_VERS)), C2_VERSION); ret = -EINVAL; iounmap(mmio_regs); @@ -1040,12 +1046,12 @@ static int __devinit c2_probe(struct pci_dev *pcidev, } /* Validate the adapter IVN */ - if (be32_to_cpu(readl(mmio_regs + C2_REGS_IVN)) != C2_IVN) { + if (be32_to_cpu((__force __be32) readl(mmio_regs + C2_REGS_IVN)) != C2_IVN) { printk(KERN_ERR PFX "Downlevel FIrmware level. You should be using " "the OpenIB device support kit. " "[fw=0x%x, c2=0x%x], Adapter not claimed\n", - be32_to_cpu(readl(mmio_regs + C2_REGS_IVN)), - C2_IVN); + be32_to_cpu((__force __be32) readl(mmio_regs + C2_REGS_IVN)), + C2_IVN); ret = -EINVAL; iounmap(mmio_regs); goto bail2; @@ -1068,7 +1074,7 @@ static int __devinit c2_probe(struct pci_dev *pcidev, /* Get the last RX index */ c2dev->cur_rx = - (be32_to_cpu(readl(mmio_regs + C2_REGS_HRX_CUR)) - + (be32_to_cpu((__force __be32) readl(mmio_regs + C2_REGS_HRX_CUR)) - 0xffffc000) / sizeof(struct c2_rxp_desc); /* Request an interrupt line for the driver */ @@ -1090,7 +1096,7 @@ static int __devinit c2_probe(struct pci_dev *pcidev, } /* Save off the actual size prior to unmapping mmio_regs */ - kva_map_size = be32_to_cpu(readl(mmio_regs + C2_REGS_PCI_WINSIZE)); + kva_map_size = be32_to_cpu((__force __be32) readl(mmio_regs + C2_REGS_PCI_WINSIZE)); /* Unmap the adapter PCI registers in BAR4 */ iounmap(mmio_regs); @@ -1109,7 +1115,7 @@ static int __devinit c2_probe(struct pci_dev *pcidev, /* Remap the adapter HRXDQ PA space to kernel VA space */ c2dev->mmio_rxp_ring = ioremap_nocache(reg4_start + C2_RXP_HRXDQ_OFFSET, C2_RXP_HRXDQ_SIZE); - if (c2dev->mmio_rxp_ring == 0UL) { + if (!c2dev->mmio_rxp_ring) { printk(KERN_ERR PFX "Unable to remap MMIO HRXDQ region\n"); ret = -EIO; goto bail6; @@ -1118,7 +1124,7 @@ static int __devinit c2_probe(struct pci_dev *pcidev, /* Remap the adapter HTXDQ PA space to kernel VA space */ c2dev->mmio_txp_ring = ioremap_nocache(reg4_start + C2_TXP_HTXDQ_OFFSET, C2_TXP_HTXDQ_SIZE); - if (c2dev->mmio_txp_ring == 0UL) { + if (!c2dev->mmio_txp_ring) { printk(KERN_ERR PFX "Unable to remap MMIO HTXDQ region\n"); ret = -EIO; goto bail7; @@ -1129,7 +1135,7 @@ static int __devinit c2_probe(struct pci_dev *pcidev, /* Remap the PCI registers in adapter BAR0 to kernel VA space */ c2dev->regs = ioremap_nocache(reg0_start, reg0_len); - if (c2dev->regs == 0UL) { + if (!c2dev->regs) { printk(KERN_ERR PFX "Unable to remap BAR0\n"); ret = -EIO; goto bail8; @@ -1139,7 +1145,7 @@ static int __devinit c2_probe(struct pci_dev *pcidev, c2dev->pa = reg4_start + C2_PCI_REGS_OFFSET; c2dev->kva = ioremap_nocache(reg4_start + C2_PCI_REGS_OFFSET, kva_map_size); - if (c2dev->kva == 0UL) { + if (!c2dev->kva) { printk(KERN_ERR PFX "Unable to remap BAR4\n"); ret = -EIO; goto bail9; diff --git a/drivers/infiniband/hw/amso1100/c2.h b/drivers/infiniband/hw/amso1100/c2.h index fa58200217a1..ed38ab8d9c0c 100644 --- a/drivers/infiniband/hw/amso1100/c2.h +++ b/drivers/infiniband/hw/amso1100/c2.h @@ -346,7 +346,7 @@ struct c2_dev { // spinlock_t aeq_lock; // spinlock_t rnic_lock; - u16 *hint_count; + __be16 *hint_count; dma_addr_t hint_count_dma; u16 hints_read; @@ -425,10 +425,10 @@ static inline void __raw_writeq(u64 val, void __iomem * addr) #endif #define C2_SET_CUR_RX(c2dev, cur_rx) \ - __raw_writel(cpu_to_be32(cur_rx), c2dev->mmio_txp_ring + 4092) + __raw_writel((__force u32) cpu_to_be32(cur_rx), c2dev->mmio_txp_ring + 4092) #define C2_GET_CUR_RX(c2dev) \ - be32_to_cpu(readl(c2dev->mmio_txp_ring + 4092)) + be32_to_cpu((__force __be32) readl(c2dev->mmio_txp_ring + 4092)) static inline struct c2_dev *to_c2dev(struct ib_device *ibdev) { @@ -485,8 +485,8 @@ extern void c2_unregister_device(struct c2_dev *c2dev); extern int c2_rnic_init(struct c2_dev *c2dev); extern void c2_rnic_term(struct c2_dev *c2dev); extern void c2_rnic_interrupt(struct c2_dev *c2dev); -extern int c2_del_addr(struct c2_dev *c2dev, u32 inaddr, u32 inmask); -extern int c2_add_addr(struct c2_dev *c2dev, u32 inaddr, u32 inmask); +extern int c2_del_addr(struct c2_dev *c2dev, __be32 inaddr, __be32 inmask); +extern int c2_add_addr(struct c2_dev *c2dev, __be32 inaddr, __be32 inmask); /* QPs */ extern int c2_alloc_qp(struct c2_dev *c2dev, struct c2_pd *pd, @@ -545,7 +545,7 @@ extern void c2_ae_event(struct c2_dev *c2dev, u32 mq_index); extern int c2_init_mqsp_pool(struct c2_dev *c2dev, gfp_t gfp_mask, struct sp_chunk **root); extern void c2_free_mqsp_pool(struct c2_dev *c2dev, struct sp_chunk *root); -extern u16 *c2_alloc_mqsp(struct c2_dev *c2dev, struct sp_chunk *head, - dma_addr_t *dma_addr, gfp_t gfp_mask); -extern void c2_free_mqsp(u16 * mqsp); +extern __be16 *c2_alloc_mqsp(struct c2_dev *c2dev, struct sp_chunk *head, + dma_addr_t *dma_addr, gfp_t gfp_mask); +extern void c2_free_mqsp(__be16* mqsp); #endif diff --git a/drivers/infiniband/hw/amso1100/c2_ae.c b/drivers/infiniband/hw/amso1100/c2_ae.c index a31439bd3b67..62af74295dbe 100644 --- a/drivers/infiniband/hw/amso1100/c2_ae.c +++ b/drivers/infiniband/hw/amso1100/c2_ae.c @@ -61,7 +61,7 @@ static int c2_convert_cm_status(u32 c2_status) default: printk(KERN_ERR PFX "%s - Unable to convert CM status: %d\n", - __FUNCTION__, c2_status); + __func__, c2_status); return -EIO; } } @@ -193,9 +193,9 @@ void c2_ae_event(struct c2_dev *c2dev, u32 mq_index) pr_debug("%s: event = %s, user_context=%llx, " "resource_type=%x, " "resource=%x, qp_state=%s\n", - __FUNCTION__, + __func__, to_event_str(event_id), - (unsigned long long) be64_to_cpu(wr->ae.ae_generic.user_context), + (unsigned long long) wr->ae.ae_generic.user_context, be32_to_cpu(wr->ae.ae_generic.resource_type), be32_to_cpu(wr->ae.ae_generic.resource), to_qp_state_str(be32_to_cpu(wr->ae.ae_generic.qp_state))); @@ -259,7 +259,7 @@ void c2_ae_event(struct c2_dev *c2dev, u32 mq_index) BUG_ON(1); pr_debug("%s:%d Unexpected event_id=%d on QP=%p, " "CM_ID=%p\n", - __FUNCTION__, __LINE__, + __func__, __LINE__, event_id, qp, cm_id); break; } @@ -276,7 +276,7 @@ void c2_ae_event(struct c2_dev *c2dev, u32 mq_index) pr_debug("C2_RES_IND_EP event_id=%d\n", event_id); if (event_id != CCAE_CONNECTION_REQUEST) { pr_debug("%s: Invalid event_id: %d\n", - __FUNCTION__, event_id); + __func__, event_id); break; } cm_event.event = IW_CM_EVENT_CONNECT_REQUEST; diff --git a/drivers/infiniband/hw/amso1100/c2_alloc.c b/drivers/infiniband/hw/amso1100/c2_alloc.c index 0315f99e4191..e9110163aeff 100644 --- a/drivers/infiniband/hw/amso1100/c2_alloc.c +++ b/drivers/infiniband/hw/amso1100/c2_alloc.c @@ -87,8 +87,8 @@ void c2_free_mqsp_pool(struct c2_dev *c2dev, struct sp_chunk *root) } } -u16 *c2_alloc_mqsp(struct c2_dev *c2dev, struct sp_chunk *head, - dma_addr_t *dma_addr, gfp_t gfp_mask) +__be16 *c2_alloc_mqsp(struct c2_dev *c2dev, struct sp_chunk *head, + dma_addr_t *dma_addr, gfp_t gfp_mask) { u16 mqsp; @@ -113,14 +113,14 @@ u16 *c2_alloc_mqsp(struct c2_dev *c2dev, struct sp_chunk *head, *dma_addr = head->dma_addr + ((unsigned long) &(head->shared_ptr[mqsp]) - (unsigned long) head); - pr_debug("%s addr %p dma_addr %llx\n", __FUNCTION__, + pr_debug("%s addr %p dma_addr %llx\n", __func__, &(head->shared_ptr[mqsp]), (unsigned long long) *dma_addr); - return &(head->shared_ptr[mqsp]); + return (__force __be16 *) &(head->shared_ptr[mqsp]); } return NULL; } -void c2_free_mqsp(u16 * mqsp) +void c2_free_mqsp(__be16 *mqsp) { struct sp_chunk *head; u16 idx; @@ -129,7 +129,7 @@ void c2_free_mqsp(u16 * mqsp) head = (struct sp_chunk *) ((unsigned long) mqsp & PAGE_MASK); /* Link head to new mqsp */ - *mqsp = head->head; + *mqsp = (__force __be16) head->head; /* Compute the shared_ptr index */ idx = ((unsigned long) mqsp & ~PAGE_MASK) >> 1; diff --git a/drivers/infiniband/hw/amso1100/c2_cq.c b/drivers/infiniband/hw/amso1100/c2_cq.c index d2b3366786d6..bb17cce3cb59 100644 --- a/drivers/infiniband/hw/amso1100/c2_cq.c +++ b/drivers/infiniband/hw/amso1100/c2_cq.c @@ -422,8 +422,8 @@ void c2_free_cq(struct c2_dev *c2dev, struct c2_cq *cq) goto bail1; reply = (struct c2wr_cq_destroy_rep *) (unsigned long) (vq_req->reply_msg); - - vq_repbuf_free(c2dev, reply); + if (reply) + vq_repbuf_free(c2dev, reply); bail1: vq_req_free(c2dev, vq_req); bail0: diff --git a/drivers/infiniband/hw/amso1100/c2_intr.c b/drivers/infiniband/hw/amso1100/c2_intr.c index 0d0bc33ca30a..3b5095470cb3 100644 --- a/drivers/infiniband/hw/amso1100/c2_intr.c +++ b/drivers/infiniband/hw/amso1100/c2_intr.c @@ -174,7 +174,11 @@ static void handle_vq(struct c2_dev *c2dev, u32 mq_index) return; } - err = c2_errno(reply_msg); + if (reply_msg) + err = c2_errno(reply_msg); + else + err = -ENOMEM; + if (!err) switch (req->event) { case IW_CM_EVENT_ESTABLISHED: c2_set_qp_state(req->qp, diff --git a/drivers/infiniband/hw/amso1100/c2_mm.c b/drivers/infiniband/hw/amso1100/c2_mm.c index 1e4f46493fcb..b506fe22b4d4 100644 --- a/drivers/infiniband/hw/amso1100/c2_mm.c +++ b/drivers/infiniband/hw/amso1100/c2_mm.c @@ -45,7 +45,7 @@ * Reply buffer _is_ freed by this function. */ static int -send_pbl_messages(struct c2_dev *c2dev, u32 stag_index, +send_pbl_messages(struct c2_dev *c2dev, __be32 stag_index, unsigned long va, u32 pbl_depth, struct c2_vq_req *vq_req, int pbl_type) { diff --git a/drivers/infiniband/hw/amso1100/c2_mq.c b/drivers/infiniband/hw/amso1100/c2_mq.c index b88a75592102..0cddc49beae1 100644 --- a/drivers/infiniband/hw/amso1100/c2_mq.c +++ b/drivers/infiniband/hw/amso1100/c2_mq.c @@ -64,7 +64,7 @@ void c2_mq_produce(struct c2_mq *q) q->priv = (q->priv + 1) % q->q_size; q->hint_count++; /* Update peer's offset. */ - __raw_writew(cpu_to_be16(q->priv), &q->peer->shared); + __raw_writew((__force u16) cpu_to_be16(q->priv), &q->peer->shared); } } @@ -105,7 +105,7 @@ void c2_mq_free(struct c2_mq *q) #endif q->priv = (q->priv + 1) % q->q_size; /* Update peer's offset. */ - __raw_writew(cpu_to_be16(q->priv), &q->peer->shared); + __raw_writew((__force u16) cpu_to_be16(q->priv), &q->peer->shared); } } diff --git a/drivers/infiniband/hw/amso1100/c2_mq.h b/drivers/infiniband/hw/amso1100/c2_mq.h index 9185bbb21658..acede007b94a 100644 --- a/drivers/infiniband/hw/amso1100/c2_mq.h +++ b/drivers/infiniband/hw/amso1100/c2_mq.h @@ -75,7 +75,7 @@ struct c2_mq { u16 hint_count; u16 priv; struct c2_mq_shared __iomem *peer; - u16 *shared; + __be16 *shared; dma_addr_t shared_dma; u32 q_size; u32 msg_size; diff --git a/drivers/infiniband/hw/amso1100/c2_provider.c b/drivers/infiniband/hw/amso1100/c2_provider.c index 7a6cece6ea9d..e10d27a6e145 100644 --- a/drivers/infiniband/hw/amso1100/c2_provider.c +++ b/drivers/infiniband/hw/amso1100/c2_provider.c @@ -67,7 +67,7 @@ static int c2_query_device(struct ib_device *ibdev, { struct c2_dev *c2dev = to_c2dev(ibdev); - pr_debug("%s:%u\n", __FUNCTION__, __LINE__); + pr_debug("%s:%u\n", __func__, __LINE__); *props = c2dev->props; return 0; @@ -76,7 +76,7 @@ static int c2_query_device(struct ib_device *ibdev, static int c2_query_port(struct ib_device *ibdev, u8 port, struct ib_port_attr *props) { - pr_debug("%s:%u\n", __FUNCTION__, __LINE__); + pr_debug("%s:%u\n", __func__, __LINE__); props->max_mtu = IB_MTU_4096; props->lid = 0; @@ -102,14 +102,14 @@ static int c2_modify_port(struct ib_device *ibdev, u8 port, int port_modify_mask, struct ib_port_modify *props) { - pr_debug("%s:%u\n", __FUNCTION__, __LINE__); + pr_debug("%s:%u\n", __func__, __LINE__); return 0; } static int c2_query_pkey(struct ib_device *ibdev, u8 port, u16 index, u16 * pkey) { - pr_debug("%s:%u\n", __FUNCTION__, __LINE__); + pr_debug("%s:%u\n", __func__, __LINE__); *pkey = 0; return 0; } @@ -119,7 +119,7 @@ static int c2_query_gid(struct ib_device *ibdev, u8 port, { struct c2_dev *c2dev = to_c2dev(ibdev); - pr_debug("%s:%u\n", __FUNCTION__, __LINE__); + pr_debug("%s:%u\n", __func__, __LINE__); memset(&(gid->raw[0]), 0, sizeof(gid->raw)); memcpy(&(gid->raw[0]), c2dev->pseudo_netdev->dev_addr, 6); @@ -134,7 +134,7 @@ static struct ib_ucontext *c2_alloc_ucontext(struct ib_device *ibdev, { struct c2_ucontext *context; - pr_debug("%s:%u\n", __FUNCTION__, __LINE__); + pr_debug("%s:%u\n", __func__, __LINE__); context = kmalloc(sizeof(*context), GFP_KERNEL); if (!context) return ERR_PTR(-ENOMEM); @@ -144,14 +144,14 @@ static struct ib_ucontext *c2_alloc_ucontext(struct ib_device *ibdev, static int c2_dealloc_ucontext(struct ib_ucontext *context) { - pr_debug("%s:%u\n", __FUNCTION__, __LINE__); + pr_debug("%s:%u\n", __func__, __LINE__); kfree(context); return 0; } static int c2_mmap_uar(struct ib_ucontext *context, struct vm_area_struct *vma) { - pr_debug("%s:%u\n", __FUNCTION__, __LINE__); + pr_debug("%s:%u\n", __func__, __LINE__); return -ENOSYS; } @@ -162,7 +162,7 @@ static struct ib_pd *c2_alloc_pd(struct ib_device *ibdev, struct c2_pd *pd; int err; - pr_debug("%s:%u\n", __FUNCTION__, __LINE__); + pr_debug("%s:%u\n", __func__, __LINE__); pd = kmalloc(sizeof(*pd), GFP_KERNEL); if (!pd) @@ -187,7 +187,7 @@ static struct ib_pd *c2_alloc_pd(struct ib_device *ibdev, static int c2_dealloc_pd(struct ib_pd *pd) { - pr_debug("%s:%u\n", __FUNCTION__, __LINE__); + pr_debug("%s:%u\n", __func__, __LINE__); c2_pd_free(to_c2dev(pd->device), to_c2pd(pd)); kfree(pd); @@ -196,13 +196,13 @@ static int c2_dealloc_pd(struct ib_pd *pd) static struct ib_ah *c2_ah_create(struct ib_pd *pd, struct ib_ah_attr *ah_attr) { - pr_debug("%s:%u\n", __FUNCTION__, __LINE__); + pr_debug("%s:%u\n", __func__, __LINE__); return ERR_PTR(-ENOSYS); } static int c2_ah_destroy(struct ib_ah *ah) { - pr_debug("%s:%u\n", __FUNCTION__, __LINE__); + pr_debug("%s:%u\n", __func__, __LINE__); return -ENOSYS; } @@ -230,7 +230,7 @@ struct ib_qp *c2_get_qp(struct ib_device *device, int qpn) qp = c2_find_qpn(c2dev, qpn); pr_debug("%s Returning QP=%p for QPN=%d, device=%p, refcount=%d\n", - __FUNCTION__, qp, qpn, device, + __func__, qp, qpn, device, (qp?atomic_read(&qp->refcount):0)); return (qp?&qp->ibqp:NULL); @@ -243,13 +243,16 @@ static struct ib_qp *c2_create_qp(struct ib_pd *pd, struct c2_qp *qp; int err; - pr_debug("%s:%u\n", __FUNCTION__, __LINE__); + pr_debug("%s:%u\n", __func__, __LINE__); + + if (init_attr->create_flags) + return ERR_PTR(-EINVAL); switch (init_attr->qp_type) { case IB_QPT_RC: qp = kzalloc(sizeof(*qp), GFP_KERNEL); if (!qp) { - pr_debug("%s: Unable to allocate QP\n", __FUNCTION__); + pr_debug("%s: Unable to allocate QP\n", __func__); return ERR_PTR(-ENOMEM); } spin_lock_init(&qp->lock); @@ -266,7 +269,7 @@ static struct ib_qp *c2_create_qp(struct ib_pd *pd, break; default: - pr_debug("%s: Invalid QP type: %d\n", __FUNCTION__, + pr_debug("%s: Invalid QP type: %d\n", __func__, init_attr->qp_type); return ERR_PTR(-EINVAL); break; @@ -285,7 +288,7 @@ static int c2_destroy_qp(struct ib_qp *ib_qp) struct c2_qp *qp = to_c2qp(ib_qp); pr_debug("%s:%u qp=%p,qp->state=%d\n", - __FUNCTION__, __LINE__,ib_qp,qp->state); + __func__, __LINE__, ib_qp, qp->state); c2_free_qp(to_c2dev(ib_qp->device), qp); kfree(qp); return 0; @@ -300,13 +303,13 @@ static struct ib_cq *c2_create_cq(struct ib_device *ibdev, int entries, int vect cq = kmalloc(sizeof(*cq), GFP_KERNEL); if (!cq) { - pr_debug("%s: Unable to allocate CQ\n", __FUNCTION__); + pr_debug("%s: Unable to allocate CQ\n", __func__); return ERR_PTR(-ENOMEM); } err = c2_init_cq(to_c2dev(ibdev), entries, NULL, cq); if (err) { - pr_debug("%s: error initializing CQ\n", __FUNCTION__); + pr_debug("%s: error initializing CQ\n", __func__); kfree(cq); return ERR_PTR(err); } @@ -318,7 +321,7 @@ static int c2_destroy_cq(struct ib_cq *ib_cq) { struct c2_cq *cq = to_c2cq(ib_cq); - pr_debug("%s:%u\n", __FUNCTION__, __LINE__); + pr_debug("%s:%u\n", __func__, __LINE__); c2_free_cq(to_c2dev(ib_cq->device), cq); kfree(cq); @@ -400,7 +403,7 @@ static struct ib_mr *c2_reg_phys_mr(struct ib_pd *ib_pd, mr->umem = NULL; pr_debug("%s - page shift %d, pbl_depth %d, total_len %u, " "*iova_start %llx, first pa %llx, last pa %llx\n", - __FUNCTION__, page_shift, pbl_depth, total_len, + __func__, page_shift, pbl_depth, total_len, (unsigned long long) *iova_start, (unsigned long long) page_list[0], (unsigned long long) page_list[pbl_depth-1]); @@ -422,7 +425,7 @@ static struct ib_mr *c2_get_dma_mr(struct ib_pd *pd, int acc) struct ib_phys_buf bl; u64 kva = 0; - pr_debug("%s:%u\n", __FUNCTION__, __LINE__); + pr_debug("%s:%u\n", __func__, __LINE__); /* AMSO1100 limit */ bl.size = 0xffffffff; @@ -442,7 +445,7 @@ static struct ib_mr *c2_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, struct c2_pd *c2pd = to_c2pd(pd); struct c2_mr *c2mr; - pr_debug("%s:%u\n", __FUNCTION__, __LINE__); + pr_debug("%s:%u\n", __func__, __LINE__); c2mr = kmalloc(sizeof(*c2mr), GFP_KERNEL); if (!c2mr) @@ -506,7 +509,7 @@ static int c2_dereg_mr(struct ib_mr *ib_mr) struct c2_mr *mr = to_c2mr(ib_mr); int err; - pr_debug("%s:%u\n", __FUNCTION__, __LINE__); + pr_debug("%s:%u\n", __func__, __LINE__); err = c2_stag_dealloc(to_c2dev(ib_mr->device), ib_mr->lkey); if (err) @@ -523,14 +526,14 @@ static int c2_dereg_mr(struct ib_mr *ib_mr) static ssize_t show_rev(struct class_device *cdev, char *buf) { struct c2_dev *dev = container_of(cdev, struct c2_dev, ibdev.class_dev); - pr_debug("%s:%u\n", __FUNCTION__, __LINE__); + pr_debug("%s:%u\n", __func__, __LINE__); return sprintf(buf, "%x\n", dev->props.hw_ver); } static ssize_t show_fw_ver(struct class_device *cdev, char *buf) { struct c2_dev *dev = container_of(cdev, struct c2_dev, ibdev.class_dev); - pr_debug("%s:%u\n", __FUNCTION__, __LINE__); + pr_debug("%s:%u\n", __func__, __LINE__); return sprintf(buf, "%x.%x.%x\n", (int) (dev->props.fw_ver >> 32), (int) (dev->props.fw_ver >> 16) & 0xffff, @@ -539,13 +542,13 @@ static ssize_t show_fw_ver(struct class_device *cdev, char *buf) static ssize_t show_hca(struct class_device *cdev, char *buf) { - pr_debug("%s:%u\n", __FUNCTION__, __LINE__); + pr_debug("%s:%u\n", __func__, __LINE__); return sprintf(buf, "AMSO1100\n"); } static ssize_t show_board(struct class_device *cdev, char *buf) { - pr_debug("%s:%u\n", __FUNCTION__, __LINE__); + pr_debug("%s:%u\n", __func__, __LINE__); return sprintf(buf, "%.*s\n", 32, "AMSO1100 Board ID"); } @@ -575,13 +578,13 @@ static int c2_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, static int c2_multicast_attach(struct ib_qp *ibqp, union ib_gid *gid, u16 lid) { - pr_debug("%s:%u\n", __FUNCTION__, __LINE__); + pr_debug("%s:%u\n", __func__, __LINE__); return -ENOSYS; } static int c2_multicast_detach(struct ib_qp *ibqp, union ib_gid *gid, u16 lid) { - pr_debug("%s:%u\n", __FUNCTION__, __LINE__); + pr_debug("%s:%u\n", __func__, __LINE__); return -ENOSYS; } @@ -592,13 +595,13 @@ static int c2_process_mad(struct ib_device *ibdev, struct ib_grh *in_grh, struct ib_mad *in_mad, struct ib_mad *out_mad) { - pr_debug("%s:%u\n", __FUNCTION__, __LINE__); + pr_debug("%s:%u\n", __func__, __LINE__); return -ENOSYS; } static int c2_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *iw_param) { - pr_debug("%s:%u\n", __FUNCTION__, __LINE__); + pr_debug("%s:%u\n", __func__, __LINE__); /* Request a connection */ return c2_llp_connect(cm_id, iw_param); @@ -606,7 +609,7 @@ static int c2_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *iw_param) static int c2_accept(struct iw_cm_id *cm_id, struct iw_cm_conn_param *iw_param) { - pr_debug("%s:%u\n", __FUNCTION__, __LINE__); + pr_debug("%s:%u\n", __func__, __LINE__); /* Accept the new connection */ return c2_llp_accept(cm_id, iw_param); @@ -616,7 +619,7 @@ static int c2_reject(struct iw_cm_id *cm_id, const void *pdata, u8 pdata_len) { int err; - pr_debug("%s:%u\n", __FUNCTION__, __LINE__); + pr_debug("%s:%u\n", __func__, __LINE__); err = c2_llp_reject(cm_id, pdata, pdata_len); return err; @@ -626,10 +629,10 @@ static int c2_service_create(struct iw_cm_id *cm_id, int backlog) { int err; - pr_debug("%s:%u\n", __FUNCTION__, __LINE__); + pr_debug("%s:%u\n", __func__, __LINE__); err = c2_llp_service_create(cm_id, backlog); pr_debug("%s:%u err=%d\n", - __FUNCTION__, __LINE__, + __func__, __LINE__, err); return err; } @@ -637,7 +640,7 @@ static int c2_service_create(struct iw_cm_id *cm_id, int backlog) static int c2_service_destroy(struct iw_cm_id *cm_id) { int err; - pr_debug("%s:%u\n", __FUNCTION__, __LINE__); + pr_debug("%s:%u\n", __func__, __LINE__); err = c2_llp_service_destroy(cm_id); @@ -743,7 +746,7 @@ static struct net_device *c2_pseudo_netdev_init(struct c2_dev *c2dev) netdev = alloc_netdev(sizeof(*netdev), name, setup); if (!netdev) { printk(KERN_ERR PFX "%s - etherdev alloc failed", - __FUNCTION__); + __func__); return NULL; } @@ -780,7 +783,7 @@ int c2_register_device(struct c2_dev *dev) if (ret) goto out2; - pr_debug("%s:%u\n", __FUNCTION__, __LINE__); + pr_debug("%s:%u\n", __func__, __LINE__); strlcpy(dev->ibdev.name, "amso%d", IB_DEVICE_NAME_MAX); dev->ibdev.owner = THIS_MODULE; dev->ibdev.uverbs_cmd_mask = @@ -873,13 +876,13 @@ out1: out2: free_netdev(dev->pseudo_netdev); out3: - pr_debug("%s:%u ret=%d\n", __FUNCTION__, __LINE__, ret); + pr_debug("%s:%u ret=%d\n", __func__, __LINE__, ret); return ret; } void c2_unregister_device(struct c2_dev *dev) { - pr_debug("%s:%u\n", __FUNCTION__, __LINE__); + pr_debug("%s:%u\n", __func__, __LINE__); unregister_netdev(dev->pseudo_netdev); free_netdev(dev->pseudo_netdev); ib_unregister_device(&dev->ibdev); diff --git a/drivers/infiniband/hw/amso1100/c2_qp.c b/drivers/infiniband/hw/amso1100/c2_qp.c index 01d07862ea86..a6d89440ad2c 100644 --- a/drivers/infiniband/hw/amso1100/c2_qp.c +++ b/drivers/infiniband/hw/amso1100/c2_qp.c @@ -121,7 +121,7 @@ void c2_set_qp_state(struct c2_qp *qp, int c2_state) int new_state = to_ib_state(c2_state); pr_debug("%s: qp[%p] state modify %s --> %s\n", - __FUNCTION__, + __func__, qp, to_ib_state_str(qp->state), to_ib_state_str(new_state)); @@ -141,7 +141,7 @@ int c2_qp_modify(struct c2_dev *c2dev, struct c2_qp *qp, int err; pr_debug("%s:%d qp=%p, %s --> %s\n", - __FUNCTION__, __LINE__, + __func__, __LINE__, qp, to_ib_state_str(qp->state), to_ib_state_str(attr->qp_state)); @@ -224,7 +224,7 @@ int c2_qp_modify(struct c2_dev *c2dev, struct c2_qp *qp, qp->state = next_state; #ifdef DEBUG else - pr_debug("%s: c2_errno=%d\n", __FUNCTION__, err); + pr_debug("%s: c2_errno=%d\n", __func__, err); #endif /* * If we're going to error and generating the event here, then @@ -243,7 +243,7 @@ int c2_qp_modify(struct c2_dev *c2dev, struct c2_qp *qp, vq_req_free(c2dev, vq_req); pr_debug("%s:%d qp=%p, cur_state=%s\n", - __FUNCTION__, __LINE__, + __func__, __LINE__, qp, to_ib_state_str(qp->state)); return err; @@ -811,16 +811,24 @@ int c2_post_send(struct ib_qp *ibqp, struct ib_send_wr *ib_wr, switch (ib_wr->opcode) { case IB_WR_SEND: - if (ib_wr->send_flags & IB_SEND_SOLICITED) { - c2_wr_set_id(&wr, C2_WR_TYPE_SEND_SE); - msg_size = sizeof(struct c2wr_send_req); + case IB_WR_SEND_WITH_INV: + if (ib_wr->opcode == IB_WR_SEND) { + if (ib_wr->send_flags & IB_SEND_SOLICITED) + c2_wr_set_id(&wr, C2_WR_TYPE_SEND_SE); + else + c2_wr_set_id(&wr, C2_WR_TYPE_SEND); + wr.sqwr.send.remote_stag = 0; } else { - c2_wr_set_id(&wr, C2_WR_TYPE_SEND); - msg_size = sizeof(struct c2wr_send_req); + if (ib_wr->send_flags & IB_SEND_SOLICITED) + c2_wr_set_id(&wr, C2_WR_TYPE_SEND_SE_INV); + else + c2_wr_set_id(&wr, C2_WR_TYPE_SEND_INV); + wr.sqwr.send.remote_stag = + cpu_to_be32(ib_wr->ex.invalidate_rkey); } - wr.sqwr.send.remote_stag = 0; - msg_size += sizeof(struct c2_data_addr) * ib_wr->num_sge; + msg_size = sizeof(struct c2wr_send_req) + + sizeof(struct c2_data_addr) * ib_wr->num_sge; if (ib_wr->num_sge > qp->send_sgl_depth) { err = -EINVAL; break; diff --git a/drivers/infiniband/hw/amso1100/c2_rnic.c b/drivers/infiniband/hw/amso1100/c2_rnic.c index 1687c511cb2f..9a054c6941a4 100644 --- a/drivers/infiniband/hw/amso1100/c2_rnic.c +++ b/drivers/infiniband/hw/amso1100/c2_rnic.c @@ -208,7 +208,7 @@ static int c2_rnic_query(struct c2_dev *c2dev, struct ib_device_attr *props) /* * Add an IP address to the RNIC interface */ -int c2_add_addr(struct c2_dev *c2dev, u32 inaddr, u32 inmask) +int c2_add_addr(struct c2_dev *c2dev, __be32 inaddr, __be32 inmask) { struct c2_vq_req *vq_req; struct c2wr_rnic_setconfig_req *wr; @@ -270,7 +270,7 @@ int c2_add_addr(struct c2_dev *c2dev, u32 inaddr, u32 inmask) /* * Delete an IP address from the RNIC interface */ -int c2_del_addr(struct c2_dev *c2dev, u32 inaddr, u32 inmask) +int c2_del_addr(struct c2_dev *c2dev, __be32 inaddr, __be32 inmask) { struct c2_vq_req *vq_req; struct c2wr_rnic_setconfig_req *wr; @@ -455,7 +455,8 @@ int __devinit c2_rnic_init(struct c2_dev *c2dev) IB_DEVICE_CURR_QP_STATE_MOD | IB_DEVICE_SYS_IMAGE_GUID | IB_DEVICE_ZERO_STAG | - IB_DEVICE_SEND_W_INV | IB_DEVICE_MEM_WINDOW); + IB_DEVICE_MEM_WINDOW | + IB_DEVICE_SEND_W_INV); /* Allocate the qptr_array */ c2dev->qptr_array = vmalloc(C2_MAX_CQS * sizeof(void *)); @@ -506,17 +507,17 @@ int __devinit c2_rnic_init(struct c2_dev *c2dev) mmio_regs = c2dev->kva; /* Initialize the Verbs Request Queue */ c2_mq_req_init(&c2dev->req_vq, 0, - be32_to_cpu(readl(mmio_regs + C2_REGS_Q0_QSIZE)), - be32_to_cpu(readl(mmio_regs + C2_REGS_Q0_MSGSIZE)), + be32_to_cpu((__force __be32) readl(mmio_regs + C2_REGS_Q0_QSIZE)), + be32_to_cpu((__force __be32) readl(mmio_regs + C2_REGS_Q0_MSGSIZE)), mmio_regs + - be32_to_cpu(readl(mmio_regs + C2_REGS_Q0_POOLSTART)), + be32_to_cpu((__force __be32) readl(mmio_regs + C2_REGS_Q0_POOLSTART)), mmio_regs + - be32_to_cpu(readl(mmio_regs + C2_REGS_Q0_SHARED)), + be32_to_cpu((__force __be32) readl(mmio_regs + C2_REGS_Q0_SHARED)), C2_MQ_ADAPTER_TARGET); /* Initialize the Verbs Reply Queue */ - qsize = be32_to_cpu(readl(mmio_regs + C2_REGS_Q1_QSIZE)); - msgsize = be32_to_cpu(readl(mmio_regs + C2_REGS_Q1_MSGSIZE)); + qsize = be32_to_cpu((__force __be32) readl(mmio_regs + C2_REGS_Q1_QSIZE)); + msgsize = be32_to_cpu((__force __be32) readl(mmio_regs + C2_REGS_Q1_MSGSIZE)); q1_pages = dma_alloc_coherent(&c2dev->pcidev->dev, qsize * msgsize, &c2dev->rep_vq.host_dma, GFP_KERNEL); if (!q1_pages) { @@ -524,7 +525,7 @@ int __devinit c2_rnic_init(struct c2_dev *c2dev) goto bail1; } pci_unmap_addr_set(&c2dev->rep_vq, mapping, c2dev->rep_vq.host_dma); - pr_debug("%s rep_vq va %p dma %llx\n", __FUNCTION__, q1_pages, + pr_debug("%s rep_vq va %p dma %llx\n", __func__, q1_pages, (unsigned long long) c2dev->rep_vq.host_dma); c2_mq_rep_init(&c2dev->rep_vq, 1, @@ -532,12 +533,12 @@ int __devinit c2_rnic_init(struct c2_dev *c2dev) msgsize, q1_pages, mmio_regs + - be32_to_cpu(readl(mmio_regs + C2_REGS_Q1_SHARED)), + be32_to_cpu((__force __be32) readl(mmio_regs + C2_REGS_Q1_SHARED)), C2_MQ_HOST_TARGET); /* Initialize the Asynchronus Event Queue */ - qsize = be32_to_cpu(readl(mmio_regs + C2_REGS_Q2_QSIZE)); - msgsize = be32_to_cpu(readl(mmio_regs + C2_REGS_Q2_MSGSIZE)); + qsize = be32_to_cpu((__force __be32) readl(mmio_regs + C2_REGS_Q2_QSIZE)); + msgsize = be32_to_cpu((__force __be32) readl(mmio_regs + C2_REGS_Q2_MSGSIZE)); q2_pages = dma_alloc_coherent(&c2dev->pcidev->dev, qsize * msgsize, &c2dev->aeq.host_dma, GFP_KERNEL); if (!q2_pages) { @@ -545,7 +546,7 @@ int __devinit c2_rnic_init(struct c2_dev *c2dev) goto bail2; } pci_unmap_addr_set(&c2dev->aeq, mapping, c2dev->aeq.host_dma); - pr_debug("%s aeq va %p dma %llx\n", __FUNCTION__, q2_pages, + pr_debug("%s aeq va %p dma %llx\n", __func__, q2_pages, (unsigned long long) c2dev->aeq.host_dma); c2_mq_rep_init(&c2dev->aeq, 2, @@ -553,7 +554,7 @@ int __devinit c2_rnic_init(struct c2_dev *c2dev) msgsize, q2_pages, mmio_regs + - be32_to_cpu(readl(mmio_regs + C2_REGS_Q2_SHARED)), + be32_to_cpu((__force __be32) readl(mmio_regs + C2_REGS_Q2_SHARED)), C2_MQ_HOST_TARGET); /* Initialize the verbs request allocator */ diff --git a/drivers/infiniband/hw/amso1100/c2_vq.c b/drivers/infiniband/hw/amso1100/c2_vq.c index cfdacb1ec279..9ce7819b7b2e 100644 --- a/drivers/infiniband/hw/amso1100/c2_vq.c +++ b/drivers/infiniband/hw/amso1100/c2_vq.c @@ -197,7 +197,7 @@ int vq_send_wr(struct c2_dev *c2dev, union c2wr *wr) */ while (msg == NULL) { pr_debug("%s:%d no available msg in VQ, waiting...\n", - __FUNCTION__, __LINE__); + __func__, __LINE__); init_waitqueue_entry(&__wait, current); add_wait_queue(&c2dev->req_vq_wo, &__wait); spin_unlock(&c2dev->vqlock); diff --git a/drivers/infiniband/hw/amso1100/c2_wr.h b/drivers/infiniband/hw/amso1100/c2_wr.h index 3ec6c43bb0ef..c65fbdd6e469 100644 --- a/drivers/infiniband/hw/amso1100/c2_wr.h +++ b/drivers/infiniband/hw/amso1100/c2_wr.h @@ -180,8 +180,8 @@ enum c2_wr_type { }; struct c2_netaddr { - u32 ip_addr; - u32 netmask; + __be32 ip_addr; + __be32 netmask; u32 mtu; }; @@ -199,9 +199,9 @@ struct c2_route { * A Scatter Gather Entry. */ struct c2_data_addr { - u32 stag; - u32 length; - u64 to; + __be32 stag; + __be32 length; + __be64 to; }; /* @@ -274,7 +274,7 @@ struct c2wr_hdr { * from the host to adapter by libccil, but we copy it anyway * to make the memcpy to the adapter better aligned. */ - u32 wqe_count; + __be32 wqe_count; /* Put these fields next so that later 32- and 64-bit * quantities are naturally aligned. @@ -316,8 +316,8 @@ enum c2_rnic_flags { struct c2wr_rnic_open_req { struct c2wr_hdr hdr; u64 user_context; - u16 flags; /* See enum c2_rnic_flags */ - u16 port_num; + __be16 flags; /* See enum c2_rnic_flags */ + __be16 port_num; } __attribute__((packed)); struct c2wr_rnic_open_rep { @@ -341,30 +341,30 @@ struct c2wr_rnic_query_req { struct c2wr_rnic_query_rep { struct c2wr_hdr hdr; u64 user_context; - u32 vendor_id; - u32 part_number; - u32 hw_version; - u32 fw_ver_major; - u32 fw_ver_minor; - u32 fw_ver_patch; + __be32 vendor_id; + __be32 part_number; + __be32 hw_version; + __be32 fw_ver_major; + __be32 fw_ver_minor; + __be32 fw_ver_patch; char fw_ver_build_str[WR_BUILD_STR_LEN]; - u32 max_qps; - u32 max_qp_depth; + __be32 max_qps; + __be32 max_qp_depth; u32 max_srq_depth; u32 max_send_sgl_depth; u32 max_rdma_sgl_depth; - u32 max_cqs; - u32 max_cq_depth; + __be32 max_cqs; + __be32 max_cq_depth; u32 max_cq_event_handlers; - u32 max_mrs; + __be32 max_mrs; u32 max_pbl_depth; - u32 max_pds; - u32 max_global_ird; + __be32 max_pds; + __be32 max_global_ird; u32 max_global_ord; - u32 max_qp_ird; - u32 max_qp_ord; + __be32 max_qp_ird; + __be32 max_qp_ord; u32 flags; - u32 max_mws; + __be32 max_mws; u32 pbe_range_low; u32 pbe_range_high; u32 max_srqs; @@ -405,7 +405,7 @@ union c2wr_rnic_getconfig { struct c2wr_rnic_setconfig_req { struct c2wr_hdr hdr; u32 rnic_handle; - u32 option; /* See c2_setconfig_cmd_t */ + __be32 option; /* See c2_setconfig_cmd_t */ /* variable data and pad. See c2_netaddr and c2_route */ u8 data[0]; } __attribute__((packed)) ; @@ -441,18 +441,18 @@ union c2wr_rnic_close { */ struct c2wr_cq_create_req { struct c2wr_hdr hdr; - u64 shared_ht; + __be64 shared_ht; u64 user_context; - u64 msg_pool; + __be64 msg_pool; u32 rnic_handle; - u32 msg_size; - u32 depth; + __be32 msg_size; + __be32 depth; } __attribute__((packed)) ; struct c2wr_cq_create_rep { struct c2wr_hdr hdr; - u32 mq_index; - u32 adapter_shared; + __be32 mq_index; + __be32 adapter_shared; u32 cq_handle; } __attribute__((packed)) ; @@ -585,40 +585,40 @@ enum c2wr_qp_flags { struct c2wr_qp_create_req { struct c2wr_hdr hdr; - u64 shared_sq_ht; - u64 shared_rq_ht; + __be64 shared_sq_ht; + __be64 shared_rq_ht; u64 user_context; u32 rnic_handle; u32 sq_cq_handle; u32 rq_cq_handle; - u32 sq_depth; - u32 rq_depth; + __be32 sq_depth; + __be32 rq_depth; u32 srq_handle; u32 srq_limit; - u32 flags; /* see enum c2wr_qp_flags */ - u32 send_sgl_depth; - u32 recv_sgl_depth; - u32 rdma_write_sgl_depth; - u32 ord; - u32 ird; + __be32 flags; /* see enum c2wr_qp_flags */ + __be32 send_sgl_depth; + __be32 recv_sgl_depth; + __be32 rdma_write_sgl_depth; + __be32 ord; + __be32 ird; u32 pd_id; } __attribute__((packed)) ; struct c2wr_qp_create_rep { struct c2wr_hdr hdr; - u32 sq_depth; - u32 rq_depth; + __be32 sq_depth; + __be32 rq_depth; u32 send_sgl_depth; u32 recv_sgl_depth; u32 rdma_write_sgl_depth; u32 ord; u32 ird; - u32 sq_msg_size; - u32 sq_mq_index; - u32 sq_mq_start; - u32 rq_msg_size; - u32 rq_mq_index; - u32 rq_mq_start; + __be32 sq_msg_size; + __be32 sq_mq_index; + __be32 sq_mq_start; + __be32 rq_msg_size; + __be32 rq_mq_index; + __be32 rq_mq_start; u32 qp_handle; } __attribute__((packed)) ; @@ -667,11 +667,11 @@ struct c2wr_qp_modify_req { u32 stream_msg_length; u32 rnic_handle; u32 qp_handle; - u32 next_qp_state; - u32 ord; - u32 ird; - u32 sq_depth; - u32 rq_depth; + __be32 next_qp_state; + __be32 ord; + __be32 ird; + __be32 sq_depth; + __be32 rq_depth; u32 llp_ep_handle; } __attribute__((packed)) ; @@ -721,10 +721,10 @@ struct c2wr_qp_connect_req { struct c2wr_hdr hdr; u32 rnic_handle; u32 qp_handle; - u32 remote_addr; - u16 remote_port; + __be32 remote_addr; + __be16 remote_port; u16 pad; - u32 private_data_length; + __be32 private_data_length; u8 private_data[0]; /* Private data in-line. */ } __attribute__((packed)) ; @@ -759,25 +759,25 @@ union c2wr_nsmr_stag_alloc { struct c2wr_nsmr_register_req { struct c2wr_hdr hdr; - u64 va; + __be64 va; u32 rnic_handle; - u16 flags; + __be16 flags; u8 stag_key; u8 pad; u32 pd_id; - u32 pbl_depth; - u32 pbe_size; - u32 fbo; - u32 length; - u32 addrs_length; + __be32 pbl_depth; + __be32 pbe_size; + __be32 fbo; + __be32 length; + __be32 addrs_length; /* array of paddrs (must be aligned on a 64bit boundary) */ - u64 paddrs[0]; + __be64 paddrs[0]; } __attribute__((packed)) ; struct c2wr_nsmr_register_rep { struct c2wr_hdr hdr; u32 pbl_depth; - u32 stag_index; + __be32 stag_index; } __attribute__((packed)) ; union c2wr_nsmr_register { @@ -788,11 +788,11 @@ union c2wr_nsmr_register { struct c2wr_nsmr_pbl_req { struct c2wr_hdr hdr; u32 rnic_handle; - u32 flags; - u32 stag_index; - u32 addrs_length; + __be32 flags; + __be32 stag_index; + __be32 addrs_length; /* array of paddrs (must be aligned on a 64bit boundary) */ - u64 paddrs[0]; + __be64 paddrs[0]; } __attribute__((packed)) ; struct c2wr_nsmr_pbl_rep { @@ -847,7 +847,7 @@ union c2wr_mw_query { struct c2wr_stag_dealloc_req { struct c2wr_hdr hdr; u32 rnic_handle; - u32 stag_index; + __be32 stag_index; } __attribute__((packed)) ; struct c2wr_stag_dealloc_rep { @@ -949,7 +949,7 @@ struct c2wr_ce { u64 qp_user_context; /* c2_user_qp_t * */ u32 qp_state; /* Current QP State */ u32 handle; /* QPID or EP Handle */ - u32 bytes_rcvd; /* valid for RECV WCs */ + __be32 bytes_rcvd; /* valid for RECV WCs */ u32 stag; } __attribute__((packed)) ; @@ -984,8 +984,8 @@ struct c2_rq_hdr { */ struct c2wr_send_req { struct c2_sq_hdr sq_hdr; - u32 sge_len; - u32 remote_stag; + __be32 sge_len; + __be32 remote_stag; u8 data[0]; /* SGE array */ } __attribute__((packed)); @@ -996,9 +996,9 @@ union c2wr_send { struct c2wr_rdma_write_req { struct c2_sq_hdr sq_hdr; - u64 remote_to; - u32 remote_stag; - u32 sge_len; + __be64 remote_to; + __be32 remote_stag; + __be32 sge_len; u8 data[0]; /* SGE array */ } __attribute__((packed)); @@ -1009,11 +1009,11 @@ union c2wr_rdma_write { struct c2wr_rdma_read_req { struct c2_sq_hdr sq_hdr; - u64 local_to; - u64 remote_to; - u32 local_stag; - u32 remote_stag; - u32 length; + __be64 local_to; + __be64 remote_to; + __be32 local_stag; + __be32 remote_stag; + __be32 length; } __attribute__((packed)); union c2wr_rdma_read { @@ -1113,9 +1113,9 @@ union c2wr_recv { struct c2wr_ae_hdr { struct c2wr_hdr hdr; u64 user_context; /* user context for this res. */ - u32 resource_type; /* see enum c2_resource_indicator */ - u32 resource; /* handle for resource */ - u32 qp_state; /* current QP State */ + __be32 resource_type; /* see enum c2_resource_indicator */ + __be32 resource; /* handle for resource */ + __be32 qp_state; /* current QP State */ } __attribute__((packed)); /* @@ -1124,11 +1124,11 @@ struct c2wr_ae_hdr { */ struct c2wr_ae_active_connect_results { struct c2wr_ae_hdr ae_hdr; - u32 laddr; - u32 raddr; - u16 lport; - u16 rport; - u32 private_data_length; + __be32 laddr; + __be32 raddr; + __be16 lport; + __be16 rport; + __be32 private_data_length; u8 private_data[0]; /* data is in-line in the msg. */ } __attribute__((packed)); @@ -1142,11 +1142,11 @@ struct c2wr_ae_active_connect_results { struct c2wr_ae_connection_request { struct c2wr_ae_hdr ae_hdr; u32 cr_handle; /* connreq handle (sock ptr) */ - u32 laddr; - u32 raddr; - u16 lport; - u16 rport; - u32 private_data_length; + __be32 laddr; + __be32 raddr; + __be16 lport; + __be16 rport; + __be32 private_data_length; u8 private_data[0]; /* data is in-line in the msg. */ } __attribute__((packed)); @@ -1158,12 +1158,12 @@ union c2wr_ae { struct c2wr_init_req { struct c2wr_hdr hdr; - u64 hint_count; - u64 q0_host_shared; - u64 q1_host_shared; - u64 q1_host_msg_pool; - u64 q2_host_shared; - u64 q2_host_msg_pool; + __be64 hint_count; + __be64 q0_host_shared; + __be64 q1_host_shared; + __be64 q1_host_msg_pool; + __be64 q2_host_shared; + __be64 q2_host_msg_pool; } __attribute__((packed)); struct c2wr_init_rep { @@ -1276,10 +1276,10 @@ struct c2wr_ep_listen_create_req { struct c2wr_hdr hdr; u64 user_context; /* returned in AEs. */ u32 rnic_handle; - u32 local_addr; /* local addr, or 0 */ - u16 local_port; /* 0 means "pick one" */ + __be32 local_addr; /* local addr, or 0 */ + __be16 local_port; /* 0 means "pick one" */ u16 pad; - u32 backlog; /* tradional tcp listen bl */ + __be32 backlog; /* tradional tcp listen bl */ } __attribute__((packed)); struct c2wr_ep_listen_create_rep { @@ -1340,7 +1340,7 @@ struct c2wr_cr_accept_req { u32 rnic_handle; u32 qp_handle; /* QP to bind to this LLP conn */ u32 ep_handle; /* LLP handle to accept */ - u32 private_data_length; + __be32 private_data_length; u8 private_data[0]; /* data in-line in msg. */ } __attribute__((packed)); @@ -1508,7 +1508,7 @@ static __inline__ void c2_wr_set_sge_count(void *wr, u8 sge_count) { ((struct c2wr_hdr *) wr)->sge_count = sge_count; } -static __inline__ u32 c2_wr_get_wqe_count(void *wr) +static __inline__ __be32 c2_wr_get_wqe_count(void *wr) { return ((struct c2wr_hdr *) wr)->wqe_count; } diff --git a/drivers/infiniband/hw/cxgb3/cxio_dbg.c b/drivers/infiniband/hw/cxgb3/cxio_dbg.c index 75f7b16a271d..a8d24d53f307 100644 --- a/drivers/infiniband/hw/cxgb3/cxio_dbg.c +++ b/drivers/infiniband/hw/cxgb3/cxio_dbg.c @@ -45,16 +45,16 @@ void cxio_dump_tpt(struct cxio_rdev *rdev, u32 stag) m = kmalloc(sizeof(*m) + size, GFP_ATOMIC); if (!m) { - PDBG("%s couldn't allocate memory.\n", __FUNCTION__); + PDBG("%s couldn't allocate memory.\n", __func__); return; } m->mem_id = MEM_PMRX; m->addr = (stag>>8) * 32 + rdev->rnic_info.tpt_base; m->len = size; - PDBG("%s TPT addr 0x%x len %d\n", __FUNCTION__, m->addr, m->len); + PDBG("%s TPT addr 0x%x len %d\n", __func__, m->addr, m->len); rc = rdev->t3cdev_p->ctl(rdev->t3cdev_p, RDMA_GET_MEM, m); if (rc) { - PDBG("%s toectl returned error %d\n", __FUNCTION__, rc); + PDBG("%s toectl returned error %d\n", __func__, rc); kfree(m); return; } @@ -82,17 +82,17 @@ void cxio_dump_pbl(struct cxio_rdev *rdev, u32 pbl_addr, uint len, u8 shift) m = kmalloc(sizeof(*m) + size, GFP_ATOMIC); if (!m) { - PDBG("%s couldn't allocate memory.\n", __FUNCTION__); + PDBG("%s couldn't allocate memory.\n", __func__); return; } m->mem_id = MEM_PMRX; m->addr = pbl_addr; m->len = size; PDBG("%s PBL addr 0x%x len %d depth %d\n", - __FUNCTION__, m->addr, m->len, npages); + __func__, m->addr, m->len, npages); rc = rdev->t3cdev_p->ctl(rdev->t3cdev_p, RDMA_GET_MEM, m); if (rc) { - PDBG("%s toectl returned error %d\n", __FUNCTION__, rc); + PDBG("%s toectl returned error %d\n", __func__, rc); kfree(m); return; } @@ -144,16 +144,16 @@ void cxio_dump_rqt(struct cxio_rdev *rdev, u32 hwtid, int nents) m = kmalloc(sizeof(*m) + size, GFP_ATOMIC); if (!m) { - PDBG("%s couldn't allocate memory.\n", __FUNCTION__); + PDBG("%s couldn't allocate memory.\n", __func__); return; } m->mem_id = MEM_PMRX; m->addr = ((hwtid)<<10) + rdev->rnic_info.rqt_base; m->len = size; - PDBG("%s RQT addr 0x%x len %d\n", __FUNCTION__, m->addr, m->len); + PDBG("%s RQT addr 0x%x len %d\n", __func__, m->addr, m->len); rc = rdev->t3cdev_p->ctl(rdev->t3cdev_p, RDMA_GET_MEM, m); if (rc) { - PDBG("%s toectl returned error %d\n", __FUNCTION__, rc); + PDBG("%s toectl returned error %d\n", __func__, rc); kfree(m); return; } @@ -177,16 +177,16 @@ void cxio_dump_tcb(struct cxio_rdev *rdev, u32 hwtid) m = kmalloc(sizeof(*m) + size, GFP_ATOMIC); if (!m) { - PDBG("%s couldn't allocate memory.\n", __FUNCTION__); + PDBG("%s couldn't allocate memory.\n", __func__); return; } m->mem_id = MEM_CM; m->addr = hwtid * size; m->len = size; - PDBG("%s TCB %d len %d\n", __FUNCTION__, m->addr, m->len); + PDBG("%s TCB %d len %d\n", __func__, m->addr, m->len); rc = rdev->t3cdev_p->ctl(rdev->t3cdev_p, RDMA_GET_MEM, m); if (rc) { - PDBG("%s toectl returned error %d\n", __FUNCTION__, rc); + PDBG("%s toectl returned error %d\n", __func__, rc); kfree(m); return; } diff --git a/drivers/infiniband/hw/cxgb3/cxio_hal.c b/drivers/infiniband/hw/cxgb3/cxio_hal.c index 03c5ff62889a..66eb7030aea8 100644 --- a/drivers/infiniband/hw/cxgb3/cxio_hal.c +++ b/drivers/infiniband/hw/cxgb3/cxio_hal.c @@ -140,7 +140,7 @@ static int cxio_hal_clear_qp_ctx(struct cxio_rdev *rdev_p, u32 qpid) struct t3_modify_qp_wr *wqe; struct sk_buff *skb = alloc_skb(sizeof(*wqe), GFP_KERNEL); if (!skb) { - PDBG("%s alloc_skb failed\n", __FUNCTION__); + PDBG("%s alloc_skb failed\n", __func__); return -ENOMEM; } wqe = (struct t3_modify_qp_wr *) skb_put(skb, sizeof(*wqe)); @@ -225,7 +225,7 @@ static u32 get_qpid(struct cxio_rdev *rdev_p, struct cxio_ucontext *uctx) } out: mutex_unlock(&uctx->lock); - PDBG("%s qpid 0x%x\n", __FUNCTION__, qpid); + PDBG("%s qpid 0x%x\n", __func__, qpid); return qpid; } @@ -237,7 +237,7 @@ static void put_qpid(struct cxio_rdev *rdev_p, u32 qpid, entry = kmalloc(sizeof *entry, GFP_KERNEL); if (!entry) return; - PDBG("%s qpid 0x%x\n", __FUNCTION__, qpid); + PDBG("%s qpid 0x%x\n", __func__, qpid); entry->qpid = qpid; mutex_lock(&uctx->lock); list_add_tail(&entry->entry, &uctx->qpids); @@ -300,7 +300,7 @@ int cxio_create_qp(struct cxio_rdev *rdev_p, u32 kernel_domain, if (!kernel_domain) wq->udb = (u64)rdev_p->rnic_info.udbell_physbase + (wq->qpid << rdev_p->qpshift); - PDBG("%s qpid 0x%x doorbell 0x%p udb 0x%llx\n", __FUNCTION__, + PDBG("%s qpid 0x%x doorbell 0x%p udb 0x%llx\n", __func__, wq->qpid, wq->doorbell, (unsigned long long) wq->udb); return 0; err4: @@ -345,7 +345,7 @@ static void insert_recv_cqe(struct t3_wq *wq, struct t3_cq *cq) { struct t3_cqe cqe; - PDBG("%s wq %p cq %p sw_rptr 0x%x sw_wptr 0x%x\n", __FUNCTION__, + PDBG("%s wq %p cq %p sw_rptr 0x%x sw_wptr 0x%x\n", __func__, wq, cq, cq->sw_rptr, cq->sw_wptr); memset(&cqe, 0, sizeof(cqe)); cqe.header = cpu_to_be32(V_CQE_STATUS(TPT_ERR_SWFLUSH) | @@ -363,10 +363,10 @@ void cxio_flush_rq(struct t3_wq *wq, struct t3_cq *cq, int count) { u32 ptr; - PDBG("%s wq %p cq %p\n", __FUNCTION__, wq, cq); + PDBG("%s wq %p cq %p\n", __func__, wq, cq); /* flush RQ */ - PDBG("%s rq_rptr %u rq_wptr %u skip count %u\n", __FUNCTION__, + PDBG("%s rq_rptr %u rq_wptr %u skip count %u\n", __func__, wq->rq_rptr, wq->rq_wptr, count); ptr = wq->rq_rptr + count; while (ptr++ != wq->rq_wptr) @@ -378,7 +378,7 @@ static void insert_sq_cqe(struct t3_wq *wq, struct t3_cq *cq, { struct t3_cqe cqe; - PDBG("%s wq %p cq %p sw_rptr 0x%x sw_wptr 0x%x\n", __FUNCTION__, + PDBG("%s wq %p cq %p sw_rptr 0x%x sw_wptr 0x%x\n", __func__, wq, cq, cq->sw_rptr, cq->sw_wptr); memset(&cqe, 0, sizeof(cqe)); cqe.header = cpu_to_be32(V_CQE_STATUS(TPT_ERR_SWFLUSH) | @@ -415,11 +415,11 @@ void cxio_flush_hw_cq(struct t3_cq *cq) { struct t3_cqe *cqe, *swcqe; - PDBG("%s cq %p cqid 0x%x\n", __FUNCTION__, cq, cq->cqid); + PDBG("%s cq %p cqid 0x%x\n", __func__, cq, cq->cqid); cqe = cxio_next_hw_cqe(cq); while (cqe) { PDBG("%s flushing hwcq rptr 0x%x to swcq wptr 0x%x\n", - __FUNCTION__, cq->rptr, cq->sw_wptr); + __func__, cq->rptr, cq->sw_wptr); swcqe = cq->sw_queue + Q_PTR2IDX(cq->sw_wptr, cq->size_log2); *swcqe = *cqe; swcqe->header |= cpu_to_be32(V_CQE_SWCQE(1)); @@ -461,7 +461,7 @@ void cxio_count_scqes(struct t3_cq *cq, struct t3_wq *wq, int *count) (*count)++; ptr++; } - PDBG("%s cq %p count %d\n", __FUNCTION__, cq, *count); + PDBG("%s cq %p count %d\n", __func__, cq, *count); } void cxio_count_rcqes(struct t3_cq *cq, struct t3_wq *wq, int *count) @@ -470,7 +470,7 @@ void cxio_count_rcqes(struct t3_cq *cq, struct t3_wq *wq, int *count) u32 ptr; *count = 0; - PDBG("%s count zero %d\n", __FUNCTION__, *count); + PDBG("%s count zero %d\n", __func__, *count); ptr = cq->sw_rptr; while (!Q_EMPTY(ptr, cq->sw_wptr)) { cqe = cq->sw_queue + (Q_PTR2IDX(ptr, cq->size_log2)); @@ -479,7 +479,7 @@ void cxio_count_rcqes(struct t3_cq *cq, struct t3_wq *wq, int *count) (*count)++; ptr++; } - PDBG("%s cq %p count %d\n", __FUNCTION__, cq, *count); + PDBG("%s cq %p count %d\n", __func__, cq, *count); } static int cxio_hal_init_ctrl_cq(struct cxio_rdev *rdev_p) @@ -506,12 +506,12 @@ static int cxio_hal_init_ctrl_qp(struct cxio_rdev *rdev_p) skb = alloc_skb(sizeof(*wqe), GFP_KERNEL); if (!skb) { - PDBG("%s alloc_skb failed\n", __FUNCTION__); + PDBG("%s alloc_skb failed\n", __func__); return -ENOMEM; } err = cxio_hal_init_ctrl_cq(rdev_p); if (err) { - PDBG("%s err %d initializing ctrl_cq\n", __FUNCTION__, err); + PDBG("%s err %d initializing ctrl_cq\n", __func__, err); goto err; } rdev_p->ctrl_qp.workq = dma_alloc_coherent( @@ -521,7 +521,7 @@ static int cxio_hal_init_ctrl_qp(struct cxio_rdev *rdev_p) &(rdev_p->ctrl_qp.dma_addr), GFP_KERNEL); if (!rdev_p->ctrl_qp.workq) { - PDBG("%s dma_alloc_coherent failed\n", __FUNCTION__); + PDBG("%s dma_alloc_coherent failed\n", __func__); err = -ENOMEM; goto err; } @@ -591,25 +591,25 @@ static int cxio_hal_ctrl_qp_write_mem(struct cxio_rdev *rdev_p, u32 addr, addr &= 0x7FFFFFF; nr_wqe = len % 96 ? len / 96 + 1 : len / 96; /* 96B max per WQE */ PDBG("%s wptr 0x%x rptr 0x%x len %d, nr_wqe %d data %p addr 0x%0x\n", - __FUNCTION__, rdev_p->ctrl_qp.wptr, rdev_p->ctrl_qp.rptr, len, + __func__, rdev_p->ctrl_qp.wptr, rdev_p->ctrl_qp.rptr, len, nr_wqe, data, addr); utx_len = 3; /* in 32B unit */ for (i = 0; i < nr_wqe; i++) { if (Q_FULL(rdev_p->ctrl_qp.rptr, rdev_p->ctrl_qp.wptr, T3_CTRL_QP_SIZE_LOG2)) { PDBG("%s ctrl_qp full wtpr 0x%0x rptr 0x%0x, " - "wait for more space i %d\n", __FUNCTION__, + "wait for more space i %d\n", __func__, rdev_p->ctrl_qp.wptr, rdev_p->ctrl_qp.rptr, i); if (wait_event_interruptible(rdev_p->ctrl_qp.waitq, !Q_FULL(rdev_p->ctrl_qp.rptr, rdev_p->ctrl_qp.wptr, T3_CTRL_QP_SIZE_LOG2))) { PDBG("%s ctrl_qp workq interrupted\n", - __FUNCTION__); + __func__); return -ERESTARTSYS; } PDBG("%s ctrl_qp wakeup, continue posting work request " - "i %d\n", __FUNCTION__, i); + "i %d\n", __func__, i); } wqe = (__be64 *)(rdev_p->ctrl_qp.workq + (rdev_p->ctrl_qp.wptr % (1 << T3_CTRL_QP_SIZE_LOG2))); @@ -630,7 +630,7 @@ static int cxio_hal_ctrl_qp_write_mem(struct cxio_rdev *rdev_p, u32 addr, if ((i != 0) && (i % (((1 << T3_CTRL_QP_SIZE_LOG2)) >> 1) == 0)) { flag = T3_COMPLETION_FLAG; - PDBG("%s force completion at i %d\n", __FUNCTION__, i); + PDBG("%s force completion at i %d\n", __func__, i); } /* build the utx mem command */ @@ -701,7 +701,7 @@ static int __cxio_tpt_op(struct cxio_rdev *rdev_p, u32 reset_tpt_entry, *stag = (stag_idx << 8) | ((*stag) & 0xFF); } PDBG("%s stag_state 0x%0x type 0x%0x pdid 0x%0x, stag_idx 0x%x\n", - __FUNCTION__, stag_state, type, pdid, stag_idx); + __func__, stag_state, type, pdid, stag_idx); if (reset_tpt_entry) cxio_hal_pblpool_free(rdev_p, *pbl_addr, *pbl_size << 3); @@ -718,7 +718,7 @@ static int __cxio_tpt_op(struct cxio_rdev *rdev_p, u32 reset_tpt_entry, if (pbl) { PDBG("%s *pdb_addr 0x%x, pbl_base 0x%x, pbl_size %d\n", - __FUNCTION__, *pbl_addr, rdev_p->rnic_info.pbl_base, + __func__, *pbl_addr, rdev_p->rnic_info.pbl_base, *pbl_size); err = cxio_hal_ctrl_qp_write_mem(rdev_p, (*pbl_addr >> 5), @@ -814,7 +814,7 @@ int cxio_rdma_init(struct cxio_rdev *rdev_p, struct t3_rdma_init_attr *attr) struct sk_buff *skb = alloc_skb(sizeof(*wqe), GFP_ATOMIC); if (!skb) return -ENOMEM; - PDBG("%s rdev_p %p\n", __FUNCTION__, rdev_p); + PDBG("%s rdev_p %p\n", __func__, rdev_p); wqe = (struct t3_rdma_init_wr *) __skb_put(skb, sizeof(*wqe)); wqe->wrh.op_seop_flags = cpu_to_be32(V_FW_RIWR_OP(T3_WR_INIT)); wqe->wrh.gen_tid_len = cpu_to_be32(V_FW_RIWR_TID(attr->tid) | @@ -856,7 +856,7 @@ static int cxio_hal_ev_handler(struct t3cdev *t3cdev_p, struct sk_buff *skb) struct respQ_msg_t *rsp_msg = (struct respQ_msg_t *) skb->data; PDBG("%d: %s cq_id 0x%x cq_ptr 0x%x genbit %0x overflow %0x an %0x" " se %0x notify %0x cqbranch %0x creditth %0x\n", - cnt, __FUNCTION__, RSPQ_CQID(rsp_msg), RSPQ_CQPTR(rsp_msg), + cnt, __func__, RSPQ_CQID(rsp_msg), RSPQ_CQPTR(rsp_msg), RSPQ_GENBIT(rsp_msg), RSPQ_OVERFLOW(rsp_msg), RSPQ_AN(rsp_msg), RSPQ_SE(rsp_msg), RSPQ_NOTIFY(rsp_msg), RSPQ_CQBRANCH(rsp_msg), RSPQ_CREDIT_THRESH(rsp_msg)); @@ -868,7 +868,7 @@ static int cxio_hal_ev_handler(struct t3cdev *t3cdev_p, struct sk_buff *skb) CQE_WRID_HI(rsp_msg->cqe), CQE_WRID_LOW(rsp_msg->cqe)); rdev_p = (struct cxio_rdev *)t3cdev_p->ulp; if (!rdev_p) { - PDBG("%s called by t3cdev %p with null ulp\n", __FUNCTION__, + PDBG("%s called by t3cdev %p with null ulp\n", __func__, t3cdev_p); return 0; } @@ -908,13 +908,13 @@ int cxio_rdev_open(struct cxio_rdev *rdev_p) strncpy(rdev_p->dev_name, rdev_p->t3cdev_p->name, T3_MAX_DEV_NAME_LEN); } else { - PDBG("%s t3cdev_p or dev_name must be set\n", __FUNCTION__); + PDBG("%s t3cdev_p or dev_name must be set\n", __func__); return -EINVAL; } list_add_tail(&rdev_p->entry, &rdev_list); - PDBG("%s opening rnic dev %s\n", __FUNCTION__, rdev_p->dev_name); + PDBG("%s opening rnic dev %s\n", __func__, rdev_p->dev_name); memset(&rdev_p->ctrl_qp, 0, sizeof(rdev_p->ctrl_qp)); if (!rdev_p->t3cdev_p) rdev_p->t3cdev_p = dev2t3cdev(netdev_p); @@ -923,14 +923,14 @@ int cxio_rdev_open(struct cxio_rdev *rdev_p) &(rdev_p->rnic_info)); if (err) { printk(KERN_ERR "%s t3cdev_p(%p)->ctl returned error %d.\n", - __FUNCTION__, rdev_p->t3cdev_p, err); + __func__, rdev_p->t3cdev_p, err); goto err1; } err = rdev_p->t3cdev_p->ctl(rdev_p->t3cdev_p, GET_PORTS, &(rdev_p->port_info)); if (err) { printk(KERN_ERR "%s t3cdev_p(%p)->ctl returned error %d.\n", - __FUNCTION__, rdev_p->t3cdev_p, err); + __func__, rdev_p->t3cdev_p, err); goto err1; } @@ -947,7 +947,7 @@ int cxio_rdev_open(struct cxio_rdev *rdev_p) rdev_p->qpmask = (65536 >> ilog2(rdev_p->qpnr)) - 1; PDBG("%s rnic %s info: tpt_base 0x%0x tpt_top 0x%0x num stags %d " "pbl_base 0x%0x pbl_top 0x%0x rqt_base 0x%0x, rqt_top 0x%0x\n", - __FUNCTION__, rdev_p->dev_name, rdev_p->rnic_info.tpt_base, + __func__, rdev_p->dev_name, rdev_p->rnic_info.tpt_base, rdev_p->rnic_info.tpt_top, cxio_num_stags(rdev_p), rdev_p->rnic_info.pbl_base, rdev_p->rnic_info.pbl_top, rdev_p->rnic_info.rqt_base, @@ -961,7 +961,7 @@ int cxio_rdev_open(struct cxio_rdev *rdev_p) err = cxio_hal_init_ctrl_qp(rdev_p); if (err) { printk(KERN_ERR "%s error %d initializing ctrl_qp.\n", - __FUNCTION__, err); + __func__, err); goto err1; } err = cxio_hal_init_resource(rdev_p, cxio_num_stags(rdev_p), 0, @@ -969,19 +969,19 @@ int cxio_rdev_open(struct cxio_rdev *rdev_p) T3_MAX_NUM_PD); if (err) { printk(KERN_ERR "%s error %d initializing hal resources.\n", - __FUNCTION__, err); + __func__, err); goto err2; } err = cxio_hal_pblpool_create(rdev_p); if (err) { printk(KERN_ERR "%s error %d initializing pbl mem pool.\n", - __FUNCTION__, err); + __func__, err); goto err3; } err = cxio_hal_rqtpool_create(rdev_p); if (err) { printk(KERN_ERR "%s error %d initializing rqt mem pool.\n", - __FUNCTION__, err); + __func__, err); goto err4; } return 0; @@ -1043,7 +1043,7 @@ static void flush_completed_wrs(struct t3_wq *wq, struct t3_cq *cq) * Insert this completed cqe into the swcq. */ PDBG("%s moving cqe into swcq sq idx %ld cq idx %ld\n", - __FUNCTION__, Q_PTR2IDX(ptr, wq->sq_size_log2), + __func__, Q_PTR2IDX(ptr, wq->sq_size_log2), Q_PTR2IDX(cq->sw_wptr, cq->size_log2)); sqp->cqe.header |= htonl(V_CQE_SWCQE(1)); *(cq->sw_queue + Q_PTR2IDX(cq->sw_wptr, cq->size_log2)) @@ -1112,7 +1112,7 @@ int cxio_poll_cq(struct t3_wq *wq, struct t3_cq *cq, struct t3_cqe *cqe, PDBG("%s CQE OOO %d qpid 0x%0x genbit %d type %d status 0x%0x" " opcode 0x%0x len 0x%0x wrid_hi_stag 0x%x wrid_low_msn 0x%x\n", - __FUNCTION__, CQE_OOO(*hw_cqe), CQE_QPID(*hw_cqe), + __func__, CQE_OOO(*hw_cqe), CQE_QPID(*hw_cqe), CQE_GENBIT(*hw_cqe), CQE_TYPE(*hw_cqe), CQE_STATUS(*hw_cqe), CQE_OPCODE(*hw_cqe), CQE_LEN(*hw_cqe), CQE_WRID_HI(*hw_cqe), CQE_WRID_LOW(*hw_cqe)); @@ -1215,7 +1215,7 @@ int cxio_poll_cq(struct t3_wq *wq, struct t3_cq *cq, struct t3_cqe *cqe, struct t3_swsq *sqp; PDBG("%s out of order completion going in swsq at idx %ld\n", - __FUNCTION__, + __func__, Q_PTR2IDX(CQE_WRID_SQ_WPTR(*hw_cqe), wq->sq_size_log2)); sqp = wq->sq + Q_PTR2IDX(CQE_WRID_SQ_WPTR(*hw_cqe), wq->sq_size_log2); @@ -1234,13 +1234,13 @@ proc_cqe: */ if (SQ_TYPE(*hw_cqe)) { wq->sq_rptr = CQE_WRID_SQ_WPTR(*hw_cqe); - PDBG("%s completing sq idx %ld\n", __FUNCTION__, + PDBG("%s completing sq idx %ld\n", __func__, Q_PTR2IDX(wq->sq_rptr, wq->sq_size_log2)); *cookie = (wq->sq + Q_PTR2IDX(wq->sq_rptr, wq->sq_size_log2))->wr_id; wq->sq_rptr++; } else { - PDBG("%s completing rq idx %ld\n", __FUNCTION__, + PDBG("%s completing rq idx %ld\n", __func__, Q_PTR2IDX(wq->rq_rptr, wq->rq_size_log2)); *cookie = *(wq->rq + Q_PTR2IDX(wq->rq_rptr, wq->rq_size_log2)); wq->rq_rptr++; @@ -1255,11 +1255,11 @@ flush_wq: skip_cqe: if (SW_CQE(*hw_cqe)) { PDBG("%s cq %p cqid 0x%x skip sw cqe sw_rptr 0x%x\n", - __FUNCTION__, cq, cq->cqid, cq->sw_rptr); + __func__, cq, cq->cqid, cq->sw_rptr); ++cq->sw_rptr; } else { PDBG("%s cq %p cqid 0x%x skip hw cqe rptr 0x%x\n", - __FUNCTION__, cq, cq->cqid, cq->rptr); + __func__, cq, cq->cqid, cq->rptr); ++cq->rptr; /* diff --git a/drivers/infiniband/hw/cxgb3/cxio_resource.c b/drivers/infiniband/hw/cxgb3/cxio_resource.c index d3095ae5bc2e..45ed4f25ef78 100644 --- a/drivers/infiniband/hw/cxgb3/cxio_resource.c +++ b/drivers/infiniband/hw/cxgb3/cxio_resource.c @@ -206,13 +206,13 @@ void cxio_hal_put_stag(struct cxio_hal_resource *rscp, u32 stag) u32 cxio_hal_get_qpid(struct cxio_hal_resource *rscp) { u32 qpid = cxio_hal_get_resource(rscp->qpid_fifo); - PDBG("%s qpid 0x%x\n", __FUNCTION__, qpid); + PDBG("%s qpid 0x%x\n", __func__, qpid); return qpid; } void cxio_hal_put_qpid(struct cxio_hal_resource *rscp, u32 qpid) { - PDBG("%s qpid 0x%x\n", __FUNCTION__, qpid); + PDBG("%s qpid 0x%x\n", __func__, qpid); cxio_hal_put_resource(rscp->qpid_fifo, qpid); } @@ -255,13 +255,13 @@ void cxio_hal_destroy_resource(struct cxio_hal_resource *rscp) u32 cxio_hal_pblpool_alloc(struct cxio_rdev *rdev_p, int size) { unsigned long addr = gen_pool_alloc(rdev_p->pbl_pool, size); - PDBG("%s addr 0x%x size %d\n", __FUNCTION__, (u32)addr, size); + PDBG("%s addr 0x%x size %d\n", __func__, (u32)addr, size); return (u32)addr; } void cxio_hal_pblpool_free(struct cxio_rdev *rdev_p, u32 addr, int size) { - PDBG("%s addr 0x%x size %d\n", __FUNCTION__, addr, size); + PDBG("%s addr 0x%x size %d\n", __func__, addr, size); gen_pool_free(rdev_p->pbl_pool, (unsigned long)addr, size); } @@ -292,13 +292,13 @@ void cxio_hal_pblpool_destroy(struct cxio_rdev *rdev_p) u32 cxio_hal_rqtpool_alloc(struct cxio_rdev *rdev_p, int size) { unsigned long addr = gen_pool_alloc(rdev_p->rqt_pool, size << 6); - PDBG("%s addr 0x%x size %d\n", __FUNCTION__, (u32)addr, size << 6); + PDBG("%s addr 0x%x size %d\n", __func__, (u32)addr, size << 6); return (u32)addr; } void cxio_hal_rqtpool_free(struct cxio_rdev *rdev_p, u32 addr, int size) { - PDBG("%s addr 0x%x size %d\n", __FUNCTION__, addr, size << 6); + PDBG("%s addr 0x%x size %d\n", __func__, addr, size << 6); gen_pool_free(rdev_p->rqt_pool, (unsigned long)addr, size << 6); } diff --git a/drivers/infiniband/hw/cxgb3/iwch.c b/drivers/infiniband/hw/cxgb3/iwch.c index 0315c9d9fce9..6ba4138c8ec3 100644 --- a/drivers/infiniband/hw/cxgb3/iwch.c +++ b/drivers/infiniband/hw/cxgb3/iwch.c @@ -65,7 +65,7 @@ static DEFINE_MUTEX(dev_mutex); static void rnic_init(struct iwch_dev *rnicp) { - PDBG("%s iwch_dev %p\n", __FUNCTION__, rnicp); + PDBG("%s iwch_dev %p\n", __func__, rnicp); idr_init(&rnicp->cqidr); idr_init(&rnicp->qpidr); idr_init(&rnicp->mmidr); @@ -106,7 +106,7 @@ static void open_rnic_dev(struct t3cdev *tdev) struct iwch_dev *rnicp; static int vers_printed; - PDBG("%s t3cdev %p\n", __FUNCTION__, tdev); + PDBG("%s t3cdev %p\n", __func__, tdev); if (!vers_printed++) printk(KERN_INFO MOD "Chelsio T3 RDMA Driver - version %s\n", DRV_VERSION); @@ -144,7 +144,7 @@ static void open_rnic_dev(struct t3cdev *tdev) static void close_rnic_dev(struct t3cdev *tdev) { struct iwch_dev *dev, *tmp; - PDBG("%s t3cdev %p\n", __FUNCTION__, tdev); + PDBG("%s t3cdev %p\n", __func__, tdev); mutex_lock(&dev_mutex); list_for_each_entry_safe(dev, tmp, &dev_list, entry) { if (dev->rdev.t3cdev_p == tdev) { diff --git a/drivers/infiniband/hw/cxgb3/iwch.h b/drivers/infiniband/hw/cxgb3/iwch.h index caf4e6007a44..9ad9b1e7c8c1 100644 --- a/drivers/infiniband/hw/cxgb3/iwch.h +++ b/drivers/infiniband/hw/cxgb3/iwch.h @@ -147,7 +147,7 @@ static inline int insert_handle(struct iwch_dev *rhp, struct idr *idr, void *handle, u32 id) { int ret; - u32 newid; + int newid; do { if (!idr_pre_get(idr, GFP_KERNEL)) { diff --git a/drivers/infiniband/hw/cxgb3/iwch_cm.c b/drivers/infiniband/hw/cxgb3/iwch_cm.c index 99f2f2a46bf7..72ca360c3dbc 100644 --- a/drivers/infiniband/hw/cxgb3/iwch_cm.c +++ b/drivers/infiniband/hw/cxgb3/iwch_cm.c @@ -110,9 +110,9 @@ static void connect_reply_upcall(struct iwch_ep *ep, int status); static void start_ep_timer(struct iwch_ep *ep) { - PDBG("%s ep %p\n", __FUNCTION__, ep); + PDBG("%s ep %p\n", __func__, ep); if (timer_pending(&ep->timer)) { - PDBG("%s stopped / restarted timer ep %p\n", __FUNCTION__, ep); + PDBG("%s stopped / restarted timer ep %p\n", __func__, ep); del_timer_sync(&ep->timer); } else get_ep(&ep->com); @@ -124,7 +124,7 @@ static void start_ep_timer(struct iwch_ep *ep) static void stop_ep_timer(struct iwch_ep *ep) { - PDBG("%s ep %p\n", __FUNCTION__, ep); + PDBG("%s ep %p\n", __func__, ep); del_timer_sync(&ep->timer); put_ep(&ep->com); } @@ -190,7 +190,7 @@ int iwch_resume_tid(struct iwch_ep *ep) static void set_emss(struct iwch_ep *ep, u16 opt) { - PDBG("%s ep %p opt %u\n", __FUNCTION__, ep, opt); + PDBG("%s ep %p opt %u\n", __func__, ep, opt); ep->emss = T3C_DATA(ep->com.tdev)->mtus[G_TCPOPT_MSS(opt)] - 40; if (G_TCPOPT_TSTAMP(opt)) ep->emss -= 12; @@ -220,7 +220,7 @@ static void state_set(struct iwch_ep_common *epc, enum iwch_ep_state new) unsigned long flags; spin_lock_irqsave(&epc->lock, flags); - PDBG("%s - %s -> %s\n", __FUNCTION__, states[epc->state], states[new]); + PDBG("%s - %s -> %s\n", __func__, states[epc->state], states[new]); __state_set(epc, new); spin_unlock_irqrestore(&epc->lock, flags); return; @@ -236,7 +236,7 @@ static void *alloc_ep(int size, gfp_t gfp) spin_lock_init(&epc->lock); init_waitqueue_head(&epc->waitq); } - PDBG("%s alloc ep %p\n", __FUNCTION__, epc); + PDBG("%s alloc ep %p\n", __func__, epc); return epc; } @@ -244,13 +244,13 @@ void __free_ep(struct kref *kref) { struct iwch_ep_common *epc; epc = container_of(kref, struct iwch_ep_common, kref); - PDBG("%s ep %p state %s\n", __FUNCTION__, epc, states[state_read(epc)]); + PDBG("%s ep %p state %s\n", __func__, epc, states[state_read(epc)]); kfree(epc); } static void release_ep_resources(struct iwch_ep *ep) { - PDBG("%s ep %p tid %d\n", __FUNCTION__, ep, ep->hwtid); + PDBG("%s ep %p tid %d\n", __func__, ep, ep->hwtid); cxgb3_remove_tid(ep->com.tdev, (void *)ep, ep->hwtid); dst_release(ep->dst); l2t_release(L2DATA(ep->com.tdev), ep->l2t); @@ -349,7 +349,7 @@ static unsigned int find_best_mtu(const struct t3c_data *d, unsigned short mtu) static void arp_failure_discard(struct t3cdev *dev, struct sk_buff *skb) { - PDBG("%s t3cdev %p\n", __FUNCTION__, dev); + PDBG("%s t3cdev %p\n", __func__, dev); kfree_skb(skb); } @@ -370,7 +370,7 @@ static void abort_arp_failure(struct t3cdev *dev, struct sk_buff *skb) { struct cpl_abort_req *req = cplhdr(skb); - PDBG("%s t3cdev %p\n", __FUNCTION__, dev); + PDBG("%s t3cdev %p\n", __func__, dev); req->cmd = CPL_ABORT_NO_RST; cxgb3_ofld_send(dev, skb); } @@ -380,10 +380,10 @@ static int send_halfclose(struct iwch_ep *ep, gfp_t gfp) struct cpl_close_con_req *req; struct sk_buff *skb; - PDBG("%s ep %p\n", __FUNCTION__, ep); + PDBG("%s ep %p\n", __func__, ep); skb = get_skb(NULL, sizeof(*req), gfp); if (!skb) { - printk(KERN_ERR MOD "%s - failed to alloc skb\n", __FUNCTION__); + printk(KERN_ERR MOD "%s - failed to alloc skb\n", __func__); return -ENOMEM; } skb->priority = CPL_PRIORITY_DATA; @@ -400,11 +400,11 @@ static int send_abort(struct iwch_ep *ep, struct sk_buff *skb, gfp_t gfp) { struct cpl_abort_req *req; - PDBG("%s ep %p\n", __FUNCTION__, ep); + PDBG("%s ep %p\n", __func__, ep); skb = get_skb(skb, sizeof(*req), gfp); if (!skb) { printk(KERN_ERR MOD "%s - failed to alloc skb.\n", - __FUNCTION__); + __func__); return -ENOMEM; } skb->priority = CPL_PRIORITY_DATA; @@ -426,12 +426,12 @@ static int send_connect(struct iwch_ep *ep) unsigned int mtu_idx; int wscale; - PDBG("%s ep %p\n", __FUNCTION__, ep); + PDBG("%s ep %p\n", __func__, ep); skb = get_skb(NULL, sizeof(*req), GFP_KERNEL); if (!skb) { printk(KERN_ERR MOD "%s - failed to alloc skb.\n", - __FUNCTION__); + __func__); return -ENOMEM; } mtu_idx = find_best_mtu(T3C_DATA(ep->com.tdev), dst_mtu(ep->dst)); @@ -470,7 +470,7 @@ static void send_mpa_req(struct iwch_ep *ep, struct sk_buff *skb) struct mpa_message *mpa; int len; - PDBG("%s ep %p pd_len %d\n", __FUNCTION__, ep, ep->plen); + PDBG("%s ep %p pd_len %d\n", __func__, ep, ep->plen); BUG_ON(skb_cloned(skb)); @@ -530,13 +530,13 @@ static int send_mpa_reject(struct iwch_ep *ep, const void *pdata, u8 plen) struct mpa_message *mpa; struct sk_buff *skb; - PDBG("%s ep %p plen %d\n", __FUNCTION__, ep, plen); + PDBG("%s ep %p plen %d\n", __func__, ep, plen); mpalen = sizeof(*mpa) + plen; skb = get_skb(NULL, mpalen + sizeof(*req), GFP_KERNEL); if (!skb) { - printk(KERN_ERR MOD "%s - cannot alloc skb!\n", __FUNCTION__); + printk(KERN_ERR MOD "%s - cannot alloc skb!\n", __func__); return -ENOMEM; } skb_reserve(skb, sizeof(*req)); @@ -580,13 +580,13 @@ static int send_mpa_reply(struct iwch_ep *ep, const void *pdata, u8 plen) int len; struct sk_buff *skb; - PDBG("%s ep %p plen %d\n", __FUNCTION__, ep, plen); + PDBG("%s ep %p plen %d\n", __func__, ep, plen); mpalen = sizeof(*mpa) + plen; skb = get_skb(NULL, mpalen + sizeof(*req), GFP_KERNEL); if (!skb) { - printk(KERN_ERR MOD "%s - cannot alloc skb!\n", __FUNCTION__); + printk(KERN_ERR MOD "%s - cannot alloc skb!\n", __func__); return -ENOMEM; } skb->priority = CPL_PRIORITY_DATA; @@ -630,7 +630,7 @@ static int act_establish(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) struct cpl_act_establish *req = cplhdr(skb); unsigned int tid = GET_TID(req); - PDBG("%s ep %p tid %d\n", __FUNCTION__, ep, tid); + PDBG("%s ep %p tid %d\n", __func__, ep, tid); dst_confirm(ep->dst); @@ -663,7 +663,7 @@ static void close_complete_upcall(struct iwch_ep *ep) { struct iw_cm_event event; - PDBG("%s ep %p\n", __FUNCTION__, ep); + PDBG("%s ep %p\n", __func__, ep); memset(&event, 0, sizeof(event)); event.event = IW_CM_EVENT_CLOSE; if (ep->com.cm_id) { @@ -680,7 +680,7 @@ static void peer_close_upcall(struct iwch_ep *ep) { struct iw_cm_event event; - PDBG("%s ep %p\n", __FUNCTION__, ep); + PDBG("%s ep %p\n", __func__, ep); memset(&event, 0, sizeof(event)); event.event = IW_CM_EVENT_DISCONNECT; if (ep->com.cm_id) { @@ -694,7 +694,7 @@ static void peer_abort_upcall(struct iwch_ep *ep) { struct iw_cm_event event; - PDBG("%s ep %p\n", __FUNCTION__, ep); + PDBG("%s ep %p\n", __func__, ep); memset(&event, 0, sizeof(event)); event.event = IW_CM_EVENT_CLOSE; event.status = -ECONNRESET; @@ -712,7 +712,7 @@ static void connect_reply_upcall(struct iwch_ep *ep, int status) { struct iw_cm_event event; - PDBG("%s ep %p status %d\n", __FUNCTION__, ep, status); + PDBG("%s ep %p status %d\n", __func__, ep, status); memset(&event, 0, sizeof(event)); event.event = IW_CM_EVENT_CONNECT_REPLY; event.status = status; @@ -724,7 +724,7 @@ static void connect_reply_upcall(struct iwch_ep *ep, int status) event.private_data = ep->mpa_pkt + sizeof(struct mpa_message); } if (ep->com.cm_id) { - PDBG("%s ep %p tid %d status %d\n", __FUNCTION__, ep, + PDBG("%s ep %p tid %d status %d\n", __func__, ep, ep->hwtid, status); ep->com.cm_id->event_handler(ep->com.cm_id, &event); } @@ -739,7 +739,7 @@ static void connect_request_upcall(struct iwch_ep *ep) { struct iw_cm_event event; - PDBG("%s ep %p tid %d\n", __FUNCTION__, ep, ep->hwtid); + PDBG("%s ep %p tid %d\n", __func__, ep, ep->hwtid); memset(&event, 0, sizeof(event)); event.event = IW_CM_EVENT_CONNECT_REQUEST; event.local_addr = ep->com.local_addr; @@ -759,11 +759,11 @@ static void established_upcall(struct iwch_ep *ep) { struct iw_cm_event event; - PDBG("%s ep %p\n", __FUNCTION__, ep); + PDBG("%s ep %p\n", __func__, ep); memset(&event, 0, sizeof(event)); event.event = IW_CM_EVENT_ESTABLISHED; if (ep->com.cm_id) { - PDBG("%s ep %p tid %d\n", __FUNCTION__, ep, ep->hwtid); + PDBG("%s ep %p tid %d\n", __func__, ep, ep->hwtid); ep->com.cm_id->event_handler(ep->com.cm_id, &event); } } @@ -773,7 +773,7 @@ static int update_rx_credits(struct iwch_ep *ep, u32 credits) struct cpl_rx_data_ack *req; struct sk_buff *skb; - PDBG("%s ep %p credits %u\n", __FUNCTION__, ep, credits); + PDBG("%s ep %p credits %u\n", __func__, ep, credits); skb = get_skb(NULL, sizeof(*req), GFP_KERNEL); if (!skb) { printk(KERN_ERR MOD "update_rx_credits - cannot alloc skb!\n"); @@ -797,7 +797,7 @@ static void process_mpa_reply(struct iwch_ep *ep, struct sk_buff *skb) enum iwch_qp_attr_mask mask; int err; - PDBG("%s ep %p\n", __FUNCTION__, ep); + PDBG("%s ep %p\n", __func__, ep); /* * Stop mpa timer. If it expired, then the state has @@ -884,7 +884,7 @@ static void process_mpa_reply(struct iwch_ep *ep, struct sk_buff *skb) ep->mpa_attr.xmit_marker_enabled = mpa->flags & MPA_MARKERS ? 1 : 0; ep->mpa_attr.version = mpa_rev; PDBG("%s - crc_enabled=%d, recv_marker_enabled=%d, " - "xmit_marker_enabled=%d, version=%d\n", __FUNCTION__, + "xmit_marker_enabled=%d, version=%d\n", __func__, ep->mpa_attr.crc_enabled, ep->mpa_attr.recv_marker_enabled, ep->mpa_attr.xmit_marker_enabled, ep->mpa_attr.version); @@ -915,7 +915,7 @@ static void process_mpa_request(struct iwch_ep *ep, struct sk_buff *skb) struct mpa_message *mpa; u16 plen; - PDBG("%s ep %p\n", __FUNCTION__, ep); + PDBG("%s ep %p\n", __func__, ep); /* * Stop mpa timer. If it expired, then the state has @@ -935,7 +935,7 @@ static void process_mpa_request(struct iwch_ep *ep, struct sk_buff *skb) return; } - PDBG("%s enter (%s line %u)\n", __FUNCTION__, __FILE__, __LINE__); + PDBG("%s enter (%s line %u)\n", __func__, __FILE__, __LINE__); /* * Copy the new data into our accumulation buffer. @@ -950,7 +950,7 @@ static void process_mpa_request(struct iwch_ep *ep, struct sk_buff *skb) */ if (ep->mpa_pkt_len < sizeof(*mpa)) return; - PDBG("%s enter (%s line %u)\n", __FUNCTION__, __FILE__, __LINE__); + PDBG("%s enter (%s line %u)\n", __func__, __FILE__, __LINE__); mpa = (struct mpa_message *) ep->mpa_pkt; /* @@ -1000,7 +1000,7 @@ static void process_mpa_request(struct iwch_ep *ep, struct sk_buff *skb) ep->mpa_attr.xmit_marker_enabled = mpa->flags & MPA_MARKERS ? 1 : 0; ep->mpa_attr.version = mpa_rev; PDBG("%s - crc_enabled=%d, recv_marker_enabled=%d, " - "xmit_marker_enabled=%d, version=%d\n", __FUNCTION__, + "xmit_marker_enabled=%d, version=%d\n", __func__, ep->mpa_attr.crc_enabled, ep->mpa_attr.recv_marker_enabled, ep->mpa_attr.xmit_marker_enabled, ep->mpa_attr.version); @@ -1017,7 +1017,7 @@ static int rx_data(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) struct cpl_rx_data *hdr = cplhdr(skb); unsigned int dlen = ntohs(hdr->len); - PDBG("%s ep %p dlen %u\n", __FUNCTION__, ep, dlen); + PDBG("%s ep %p dlen %u\n", __func__, ep, dlen); skb_pull(skb, sizeof(*hdr)); skb_trim(skb, dlen); @@ -1037,7 +1037,7 @@ static int rx_data(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) default: printk(KERN_ERR MOD "%s Unexpected streaming data." " ep %p state %d tid %d\n", - __FUNCTION__, ep, state_read(&ep->com), ep->hwtid); + __func__, ep, state_read(&ep->com), ep->hwtid); /* * The ep will timeout and inform the ULP of the failure. @@ -1063,7 +1063,7 @@ static int tx_ack(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) struct cpl_wr_ack *hdr = cplhdr(skb); unsigned int credits = ntohs(hdr->credits); - PDBG("%s ep %p credits %u\n", __FUNCTION__, ep, credits); + PDBG("%s ep %p credits %u\n", __func__, ep, credits); if (credits == 0) return CPL_RET_BUF_DONE; @@ -1084,7 +1084,7 @@ static int abort_rpl(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) { struct iwch_ep *ep = ctx; - PDBG("%s ep %p\n", __FUNCTION__, ep); + PDBG("%s ep %p\n", __func__, ep); /* * We get 2 abort replies from the HW. The first one must @@ -1115,7 +1115,7 @@ static int act_open_rpl(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) struct iwch_ep *ep = ctx; struct cpl_act_open_rpl *rpl = cplhdr(skb); - PDBG("%s ep %p status %u errno %d\n", __FUNCTION__, ep, rpl->status, + PDBG("%s ep %p status %u errno %d\n", __func__, ep, rpl->status, status2errno(rpl->status)); connect_reply_upcall(ep, status2errno(rpl->status)); state_set(&ep->com, DEAD); @@ -1133,7 +1133,7 @@ static int listen_start(struct iwch_listen_ep *ep) struct sk_buff *skb; struct cpl_pass_open_req *req; - PDBG("%s ep %p\n", __FUNCTION__, ep); + PDBG("%s ep %p\n", __func__, ep); skb = get_skb(NULL, sizeof(*req), GFP_KERNEL); if (!skb) { printk(KERN_ERR MOD "t3c_listen_start failed to alloc skb!\n"); @@ -1162,7 +1162,7 @@ static int pass_open_rpl(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) struct iwch_listen_ep *ep = ctx; struct cpl_pass_open_rpl *rpl = cplhdr(skb); - PDBG("%s ep %p status %d error %d\n", __FUNCTION__, ep, + PDBG("%s ep %p status %d error %d\n", __func__, ep, rpl->status, status2errno(rpl->status)); ep->com.rpl_err = status2errno(rpl->status); ep->com.rpl_done = 1; @@ -1176,10 +1176,10 @@ static int listen_stop(struct iwch_listen_ep *ep) struct sk_buff *skb; struct cpl_close_listserv_req *req; - PDBG("%s ep %p\n", __FUNCTION__, ep); + PDBG("%s ep %p\n", __func__, ep); skb = get_skb(NULL, sizeof(*req), GFP_KERNEL); if (!skb) { - printk(KERN_ERR MOD "%s - failed to alloc skb\n", __FUNCTION__); + printk(KERN_ERR MOD "%s - failed to alloc skb\n", __func__); return -ENOMEM; } req = (struct cpl_close_listserv_req *) skb_put(skb, sizeof(*req)); @@ -1197,7 +1197,7 @@ static int close_listsrv_rpl(struct t3cdev *tdev, struct sk_buff *skb, struct iwch_listen_ep *ep = ctx; struct cpl_close_listserv_rpl *rpl = cplhdr(skb); - PDBG("%s ep %p\n", __FUNCTION__, ep); + PDBG("%s ep %p\n", __func__, ep); ep->com.rpl_err = status2errno(rpl->status); ep->com.rpl_done = 1; wake_up(&ep->com.waitq); @@ -1211,7 +1211,7 @@ static void accept_cr(struct iwch_ep *ep, __be32 peer_ip, struct sk_buff *skb) u32 opt0h, opt0l, opt2; int wscale; - PDBG("%s ep %p\n", __FUNCTION__, ep); + PDBG("%s ep %p\n", __func__, ep); BUG_ON(skb_cloned(skb)); skb_trim(skb, sizeof(*rpl)); skb_get(skb); @@ -1244,7 +1244,7 @@ static void accept_cr(struct iwch_ep *ep, __be32 peer_ip, struct sk_buff *skb) static void reject_cr(struct t3cdev *tdev, u32 hwtid, __be32 peer_ip, struct sk_buff *skb) { - PDBG("%s t3cdev %p tid %u peer_ip %x\n", __FUNCTION__, tdev, hwtid, + PDBG("%s t3cdev %p tid %u peer_ip %x\n", __func__, tdev, hwtid, peer_ip); BUG_ON(skb_cloned(skb)); skb_trim(skb, sizeof(struct cpl_tid_release)); @@ -1279,11 +1279,11 @@ static int pass_accept_req(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) struct rtable *rt; struct iff_mac tim; - PDBG("%s parent ep %p tid %u\n", __FUNCTION__, parent_ep, hwtid); + PDBG("%s parent ep %p tid %u\n", __func__, parent_ep, hwtid); if (state_read(&parent_ep->com) != LISTEN) { printk(KERN_ERR "%s - listening ep not in LISTEN\n", - __FUNCTION__); + __func__); goto reject; } @@ -1295,7 +1295,7 @@ static int pass_accept_req(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) if (tdev->ctl(tdev, GET_IFF_FROM_MAC, &tim) < 0 || !tim.dev) { printk(KERN_ERR "%s bad dst mac %02x %02x %02x %02x %02x %02x\n", - __FUNCTION__, + __func__, req->dst_mac[0], req->dst_mac[1], req->dst_mac[2], @@ -1313,21 +1313,21 @@ static int pass_accept_req(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) req->peer_port, G_PASS_OPEN_TOS(ntohl(req->tos_tid))); if (!rt) { printk(KERN_ERR MOD "%s - failed to find dst entry!\n", - __FUNCTION__); + __func__); goto reject; } dst = &rt->u.dst; l2t = t3_l2t_get(tdev, dst->neighbour, dst->neighbour->dev); if (!l2t) { printk(KERN_ERR MOD "%s - failed to allocate l2t entry!\n", - __FUNCTION__); + __func__); dst_release(dst); goto reject; } child_ep = alloc_ep(sizeof(*child_ep), GFP_KERNEL); if (!child_ep) { printk(KERN_ERR MOD "%s - failed to allocate ep entry!\n", - __FUNCTION__); + __func__); l2t_release(L2DATA(tdev), l2t); dst_release(dst); goto reject; @@ -1362,7 +1362,7 @@ static int pass_establish(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) struct iwch_ep *ep = ctx; struct cpl_pass_establish *req = cplhdr(skb); - PDBG("%s ep %p\n", __FUNCTION__, ep); + PDBG("%s ep %p\n", __func__, ep); ep->snd_seq = ntohl(req->snd_isn); ep->rcv_seq = ntohl(req->rcv_isn); @@ -1383,7 +1383,7 @@ static int peer_close(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) int disconnect = 1; int release = 0; - PDBG("%s ep %p\n", __FUNCTION__, ep); + PDBG("%s ep %p\n", __func__, ep); dst_confirm(ep->dst); spin_lock_irqsave(&ep->com.lock, flags); @@ -1473,7 +1473,7 @@ static int peer_abort(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) int state; if (is_neg_adv_abort(req->status)) { - PDBG("%s neg_adv_abort ep %p tid %d\n", __FUNCTION__, ep, + PDBG("%s neg_adv_abort ep %p tid %d\n", __func__, ep, ep->hwtid); t3_l2t_send_event(ep->com.tdev, ep->l2t); return CPL_RET_BUF_DONE; @@ -1489,7 +1489,7 @@ static int peer_abort(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) } state = state_read(&ep->com); - PDBG("%s ep %p state %u\n", __FUNCTION__, ep, state); + PDBG("%s ep %p state %u\n", __func__, ep, state); switch (state) { case CONNECTING: break; @@ -1528,14 +1528,14 @@ static int peer_abort(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) if (ret) printk(KERN_ERR MOD "%s - qp <- error failed!\n", - __FUNCTION__); + __func__); } peer_abort_upcall(ep); break; case ABORTING: break; case DEAD: - PDBG("%s PEER_ABORT IN DEAD STATE!!!!\n", __FUNCTION__); + PDBG("%s PEER_ABORT IN DEAD STATE!!!!\n", __func__); return CPL_RET_BUF_DONE; default: BUG_ON(1); @@ -1546,7 +1546,7 @@ static int peer_abort(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) rpl_skb = get_skb(skb, sizeof(*rpl), GFP_KERNEL); if (!rpl_skb) { printk(KERN_ERR MOD "%s - cannot allocate skb!\n", - __FUNCTION__); + __func__); dst_release(ep->dst); l2t_release(L2DATA(ep->com.tdev), ep->l2t); put_ep(&ep->com); @@ -1573,7 +1573,7 @@ static int close_con_rpl(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) unsigned long flags; int release = 0; - PDBG("%s ep %p\n", __FUNCTION__, ep); + PDBG("%s ep %p\n", __func__, ep); BUG_ON(!ep); /* The cm_id may be null if we failed to connect */ @@ -1624,9 +1624,9 @@ static int terminate(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) { struct iwch_ep *ep = ctx; - PDBG("%s ep %p\n", __FUNCTION__, ep); + PDBG("%s ep %p\n", __func__, ep); skb_pull(skb, sizeof(struct cpl_rdma_terminate)); - PDBG("%s saving %d bytes of term msg\n", __FUNCTION__, skb->len); + PDBG("%s saving %d bytes of term msg\n", __func__, skb->len); skb_copy_from_linear_data(skb, ep->com.qp->attr.terminate_buffer, skb->len); ep->com.qp->attr.terminate_msg_len = skb->len; @@ -1639,13 +1639,13 @@ static int ec_status(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) struct cpl_rdma_ec_status *rep = cplhdr(skb); struct iwch_ep *ep = ctx; - PDBG("%s ep %p tid %u status %d\n", __FUNCTION__, ep, ep->hwtid, + PDBG("%s ep %p tid %u status %d\n", __func__, ep, ep->hwtid, rep->status); if (rep->status) { struct iwch_qp_attributes attrs; printk(KERN_ERR MOD "%s BAD CLOSE - Aborting tid %u\n", - __FUNCTION__, ep->hwtid); + __func__, ep->hwtid); stop_ep_timer(ep); attrs.next_state = IWCH_QP_STATE_ERROR; iwch_modify_qp(ep->com.qp->rhp, @@ -1663,7 +1663,7 @@ static void ep_timeout(unsigned long arg) unsigned long flags; spin_lock_irqsave(&ep->com.lock, flags); - PDBG("%s ep %p tid %u state %d\n", __FUNCTION__, ep, ep->hwtid, + PDBG("%s ep %p tid %u state %d\n", __func__, ep, ep->hwtid, ep->com.state); switch (ep->com.state) { case MPA_REQ_SENT: @@ -1693,7 +1693,7 @@ int iwch_reject_cr(struct iw_cm_id *cm_id, const void *pdata, u8 pdata_len) { int err; struct iwch_ep *ep = to_ep(cm_id); - PDBG("%s ep %p tid %u\n", __FUNCTION__, ep, ep->hwtid); + PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); if (state_read(&ep->com) == DEAD) { put_ep(&ep->com); @@ -1718,7 +1718,7 @@ int iwch_accept_cr(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) struct iwch_dev *h = to_iwch_dev(cm_id->device); struct iwch_qp *qp = get_qhp(h, conn_param->qpn); - PDBG("%s ep %p tid %u\n", __FUNCTION__, ep, ep->hwtid); + PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); if (state_read(&ep->com) == DEAD) return -ECONNRESET; @@ -1739,7 +1739,7 @@ int iwch_accept_cr(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) ep->com.rpl_err = 0; ep->ird = conn_param->ird; ep->ord = conn_param->ord; - PDBG("%s %d ird %d ord %d\n", __FUNCTION__, __LINE__, ep->ird, ep->ord); + PDBG("%s %d ird %d ord %d\n", __func__, __LINE__, ep->ird, ep->ord); get_ep(&ep->com); @@ -1810,7 +1810,7 @@ int iwch_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) ep = alloc_ep(sizeof(*ep), GFP_KERNEL); if (!ep) { - printk(KERN_ERR MOD "%s - cannot alloc ep.\n", __FUNCTION__); + printk(KERN_ERR MOD "%s - cannot alloc ep.\n", __func__); err = -ENOMEM; goto out; } @@ -1827,7 +1827,7 @@ int iwch_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) ep->com.cm_id = cm_id; ep->com.qp = get_qhp(h, conn_param->qpn); BUG_ON(!ep->com.qp); - PDBG("%s qpn 0x%x qp %p cm_id %p\n", __FUNCTION__, conn_param->qpn, + PDBG("%s qpn 0x%x qp %p cm_id %p\n", __func__, conn_param->qpn, ep->com.qp, cm_id); /* @@ -1835,7 +1835,7 @@ int iwch_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) */ ep->atid = cxgb3_alloc_atid(h->rdev.t3cdev_p, &t3c_client, ep); if (ep->atid == -1) { - printk(KERN_ERR MOD "%s - cannot alloc atid.\n", __FUNCTION__); + printk(KERN_ERR MOD "%s - cannot alloc atid.\n", __func__); err = -ENOMEM; goto fail2; } @@ -1847,7 +1847,7 @@ int iwch_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) cm_id->local_addr.sin_port, cm_id->remote_addr.sin_port, IPTOS_LOWDELAY); if (!rt) { - printk(KERN_ERR MOD "%s - cannot find route.\n", __FUNCTION__); + printk(KERN_ERR MOD "%s - cannot find route.\n", __func__); err = -EHOSTUNREACH; goto fail3; } @@ -1857,7 +1857,7 @@ int iwch_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) ep->l2t = t3_l2t_get(ep->com.tdev, ep->dst->neighbour, ep->dst->neighbour->dev); if (!ep->l2t) { - printk(KERN_ERR MOD "%s - cannot alloc l2e.\n", __FUNCTION__); + printk(KERN_ERR MOD "%s - cannot alloc l2e.\n", __func__); err = -ENOMEM; goto fail4; } @@ -1894,11 +1894,11 @@ int iwch_create_listen(struct iw_cm_id *cm_id, int backlog) ep = alloc_ep(sizeof(*ep), GFP_KERNEL); if (!ep) { - printk(KERN_ERR MOD "%s - cannot alloc ep.\n", __FUNCTION__); + printk(KERN_ERR MOD "%s - cannot alloc ep.\n", __func__); err = -ENOMEM; goto fail1; } - PDBG("%s ep %p\n", __FUNCTION__, ep); + PDBG("%s ep %p\n", __func__, ep); ep->com.tdev = h->rdev.t3cdev_p; cm_id->add_ref(cm_id); ep->com.cm_id = cm_id; @@ -1910,7 +1910,7 @@ int iwch_create_listen(struct iw_cm_id *cm_id, int backlog) */ ep->stid = cxgb3_alloc_stid(h->rdev.t3cdev_p, &t3c_client, ep); if (ep->stid == -1) { - printk(KERN_ERR MOD "%s - cannot alloc atid.\n", __FUNCTION__); + printk(KERN_ERR MOD "%s - cannot alloc atid.\n", __func__); err = -ENOMEM; goto fail2; } @@ -1942,7 +1942,7 @@ int iwch_destroy_listen(struct iw_cm_id *cm_id) int err; struct iwch_listen_ep *ep = to_listen_ep(cm_id); - PDBG("%s ep %p\n", __FUNCTION__, ep); + PDBG("%s ep %p\n", __func__, ep); might_sleep(); state_set(&ep->com, DEAD); @@ -1965,11 +1965,11 @@ int iwch_ep_disconnect(struct iwch_ep *ep, int abrupt, gfp_t gfp) spin_lock_irqsave(&ep->com.lock, flags); - PDBG("%s ep %p state %s, abrupt %d\n", __FUNCTION__, ep, + PDBG("%s ep %p state %s, abrupt %d\n", __func__, ep, states[ep->com.state], abrupt); if (ep->com.state == DEAD) { - PDBG("%s already dead ep %p\n", __FUNCTION__, ep); + PDBG("%s already dead ep %p\n", __func__, ep); goto out; } @@ -2020,7 +2020,7 @@ int iwch_ep_redirect(void *ctx, struct dst_entry *old, struct dst_entry *new, if (ep->dst != old) return 0; - PDBG("%s ep %p redirect to dst %p l2t %p\n", __FUNCTION__, ep, new, + PDBG("%s ep %p redirect to dst %p l2t %p\n", __func__, ep, new, l2t); dst_hold(new); l2t_release(L2DATA(ep->com.tdev), ep->l2t); diff --git a/drivers/infiniband/hw/cxgb3/iwch_cm.h b/drivers/infiniband/hw/cxgb3/iwch_cm.h index 6107e7cd9b57..2bb7fbdb3ff4 100644 --- a/drivers/infiniband/hw/cxgb3/iwch_cm.h +++ b/drivers/infiniband/hw/cxgb3/iwch_cm.h @@ -54,13 +54,13 @@ #define MPA_FLAGS_MASK 0xE0 #define put_ep(ep) { \ - PDBG("put_ep (via %s:%u) ep %p refcnt %d\n", __FUNCTION__, __LINE__, \ + PDBG("put_ep (via %s:%u) ep %p refcnt %d\n", __func__, __LINE__, \ ep, atomic_read(&((ep)->kref.refcount))); \ kref_put(&((ep)->kref), __free_ep); \ } #define get_ep(ep) { \ - PDBG("get_ep (via %s:%u) ep %p, refcnt %d\n", __FUNCTION__, __LINE__, \ + PDBG("get_ep (via %s:%u) ep %p, refcnt %d\n", __func__, __LINE__, \ ep, atomic_read(&((ep)->kref.refcount))); \ kref_get(&((ep)->kref)); \ } diff --git a/drivers/infiniband/hw/cxgb3/iwch_cq.c b/drivers/infiniband/hw/cxgb3/iwch_cq.c index d7624c170ee7..4ee8ccd0a9e5 100644 --- a/drivers/infiniband/hw/cxgb3/iwch_cq.c +++ b/drivers/infiniband/hw/cxgb3/iwch_cq.c @@ -67,7 +67,7 @@ static int iwch_poll_cq_one(struct iwch_dev *rhp, struct iwch_cq *chp, ret = cxio_poll_cq(wq, &(chp->cq), &cqe, &cqe_flushed, &cookie, &credit); if (t3a_device(chp->rhp) && credit) { - PDBG("%s updating %d cq credits on id %d\n", __FUNCTION__, + PDBG("%s updating %d cq credits on id %d\n", __func__, credit, chp->cq.cqid); cxio_hal_cq_op(&rhp->rdev, &chp->cq, CQ_CREDIT_UPDATE, credit); } @@ -83,7 +83,7 @@ static int iwch_poll_cq_one(struct iwch_dev *rhp, struct iwch_cq *chp, wc->vendor_err = CQE_STATUS(cqe); PDBG("%s qpid 0x%x type %d opcode %d status 0x%x wrid hi 0x%x " - "lo 0x%x cookie 0x%llx\n", __FUNCTION__, + "lo 0x%x cookie 0x%llx\n", __func__, CQE_QPID(cqe), CQE_TYPE(cqe), CQE_OPCODE(cqe), CQE_STATUS(cqe), CQE_WRID_HI(cqe), CQE_WRID_LOW(cqe), (unsigned long long) cookie); diff --git a/drivers/infiniband/hw/cxgb3/iwch_ev.c b/drivers/infiniband/hw/cxgb3/iwch_ev.c index b40676662a8a..7b67a6771720 100644 --- a/drivers/infiniband/hw/cxgb3/iwch_ev.c +++ b/drivers/infiniband/hw/cxgb3/iwch_ev.c @@ -52,7 +52,7 @@ static void post_qp_event(struct iwch_dev *rnicp, struct iwch_cq *chp, if (!qhp) { printk(KERN_ERR "%s unaffiliated error 0x%x qpid 0x%x\n", - __FUNCTION__, CQE_STATUS(rsp_msg->cqe), + __func__, CQE_STATUS(rsp_msg->cqe), CQE_QPID(rsp_msg->cqe)); spin_unlock(&rnicp->lock); return; @@ -61,14 +61,14 @@ static void post_qp_event(struct iwch_dev *rnicp, struct iwch_cq *chp, if ((qhp->attr.state == IWCH_QP_STATE_ERROR) || (qhp->attr.state == IWCH_QP_STATE_TERMINATE)) { PDBG("%s AE received after RTS - " - "qp state %d qpid 0x%x status 0x%x\n", __FUNCTION__, + "qp state %d qpid 0x%x status 0x%x\n", __func__, qhp->attr.state, qhp->wq.qpid, CQE_STATUS(rsp_msg->cqe)); spin_unlock(&rnicp->lock); return; } printk(KERN_ERR "%s - AE qpid 0x%x opcode %d status 0x%x " - "type %d wrid.hi 0x%x wrid.lo 0x%x \n", __FUNCTION__, + "type %d wrid.hi 0x%x wrid.lo 0x%x \n", __func__, CQE_QPID(rsp_msg->cqe), CQE_OPCODE(rsp_msg->cqe), CQE_STATUS(rsp_msg->cqe), CQE_TYPE(rsp_msg->cqe), CQE_WRID_HI(rsp_msg->cqe), CQE_WRID_LOW(rsp_msg->cqe)); @@ -132,10 +132,10 @@ void iwch_ev_dispatch(struct cxio_rdev *rdev_p, struct sk_buff *skb) (CQE_STATUS(rsp_msg->cqe) == 0)) { if (SQ_TYPE(rsp_msg->cqe)) { PDBG("%s QPID 0x%x ep %p disconnecting\n", - __FUNCTION__, qhp->wq.qpid, qhp->ep); + __func__, qhp->wq.qpid, qhp->ep); iwch_ep_disconnect(qhp->ep, 0, GFP_ATOMIC); } else { - PDBG("%s post REQ_ERR AE QPID 0x%x\n", __FUNCTION__, + PDBG("%s post REQ_ERR AE QPID 0x%x\n", __func__, qhp->wq.qpid); post_qp_event(rnicp, chp, rsp_msg, IB_EVENT_QP_REQ_ERR, 0); @@ -180,7 +180,7 @@ void iwch_ev_dispatch(struct cxio_rdev *rdev_p, struct sk_buff *skb) case TPT_ERR_INVALIDATE_SHARED_MR: case TPT_ERR_INVALIDATE_MR_WITH_MW_BOUND: printk(KERN_ERR "%s - CQE Err qpid 0x%x opcode %d status 0x%x " - "type %d wrid.hi 0x%x wrid.lo 0x%x \n", __FUNCTION__, + "type %d wrid.hi 0x%x wrid.lo 0x%x \n", __func__, CQE_QPID(rsp_msg->cqe), CQE_OPCODE(rsp_msg->cqe), CQE_STATUS(rsp_msg->cqe), CQE_TYPE(rsp_msg->cqe), CQE_WRID_HI(rsp_msg->cqe), CQE_WRID_LOW(rsp_msg->cqe)); diff --git a/drivers/infiniband/hw/cxgb3/iwch_mem.c b/drivers/infiniband/hw/cxgb3/iwch_mem.c index b8797c66676d..58c3d61bcd14 100644 --- a/drivers/infiniband/hw/cxgb3/iwch_mem.c +++ b/drivers/infiniband/hw/cxgb3/iwch_mem.c @@ -62,7 +62,7 @@ int iwch_register_mem(struct iwch_dev *rhp, struct iwch_pd *php, mmid = stag >> 8; mhp->ibmr.rkey = mhp->ibmr.lkey = stag; insert_handle(rhp, &rhp->mmidr, mhp, mmid); - PDBG("%s mmid 0x%x mhp %p\n", __FUNCTION__, mmid, mhp); + PDBG("%s mmid 0x%x mhp %p\n", __func__, mmid, mhp); return 0; } @@ -96,7 +96,7 @@ int iwch_reregister_mem(struct iwch_dev *rhp, struct iwch_pd *php, mmid = stag >> 8; mhp->ibmr.rkey = mhp->ibmr.lkey = stag; insert_handle(rhp, &rhp->mmidr, mhp, mmid); - PDBG("%s mmid 0x%x mhp %p\n", __FUNCTION__, mmid, mhp); + PDBG("%s mmid 0x%x mhp %p\n", __func__, mmid, mhp); return 0; } @@ -163,7 +163,7 @@ int build_phys_page_list(struct ib_phys_buf *buffer_list, ((u64) j << *shift)); PDBG("%s va 0x%llx mask 0x%llx shift %d len %lld pbl_size %d\n", - __FUNCTION__, (unsigned long long) *iova_start, + __func__, (unsigned long long) *iova_start, (unsigned long long) mask, *shift, (unsigned long long) *total_size, *npages); diff --git a/drivers/infiniband/hw/cxgb3/iwch_provider.c b/drivers/infiniband/hw/cxgb3/iwch_provider.c index b2ea9210467f..ca7265443c05 100644 --- a/drivers/infiniband/hw/cxgb3/iwch_provider.c +++ b/drivers/infiniband/hw/cxgb3/iwch_provider.c @@ -101,7 +101,7 @@ static int iwch_dealloc_ucontext(struct ib_ucontext *context) struct iwch_ucontext *ucontext = to_iwch_ucontext(context); struct iwch_mm_entry *mm, *tmp; - PDBG("%s context %p\n", __FUNCTION__, context); + PDBG("%s context %p\n", __func__, context); list_for_each_entry_safe(mm, tmp, &ucontext->mmaps, entry) kfree(mm); cxio_release_ucontext(&rhp->rdev, &ucontext->uctx); @@ -115,7 +115,7 @@ static struct ib_ucontext *iwch_alloc_ucontext(struct ib_device *ibdev, struct iwch_ucontext *context; struct iwch_dev *rhp = to_iwch_dev(ibdev); - PDBG("%s ibdev %p\n", __FUNCTION__, ibdev); + PDBG("%s ibdev %p\n", __func__, ibdev); context = kzalloc(sizeof(*context), GFP_KERNEL); if (!context) return ERR_PTR(-ENOMEM); @@ -129,7 +129,7 @@ static int iwch_destroy_cq(struct ib_cq *ib_cq) { struct iwch_cq *chp; - PDBG("%s ib_cq %p\n", __FUNCTION__, ib_cq); + PDBG("%s ib_cq %p\n", __func__, ib_cq); chp = to_iwch_cq(ib_cq); remove_handle(chp->rhp, &chp->rhp->cqidr, chp->cq.cqid); @@ -151,7 +151,7 @@ static struct ib_cq *iwch_create_cq(struct ib_device *ibdev, int entries, int ve struct iwch_create_cq_req ureq; struct iwch_ucontext *ucontext = NULL; - PDBG("%s ib_dev %p entries %d\n", __FUNCTION__, ibdev, entries); + PDBG("%s ib_dev %p entries %d\n", __func__, ibdev, entries); rhp = to_iwch_dev(ibdev); chp = kzalloc(sizeof(*chp), GFP_KERNEL); if (!chp) @@ -233,7 +233,7 @@ static int iwch_resize_cq(struct ib_cq *cq, int cqe, struct ib_udata *udata) struct t3_cq oldcq, newcq; int ret; - PDBG("%s ib_cq %p cqe %d\n", __FUNCTION__, cq, cqe); + PDBG("%s ib_cq %p cqe %d\n", __func__, cq, cqe); /* We don't downsize... */ if (cqe <= cq->cqe) @@ -281,7 +281,7 @@ static int iwch_resize_cq(struct ib_cq *cq, int cqe, struct ib_udata *udata) ret = cxio_destroy_cq(&chp->rhp->rdev, &oldcq); if (ret) { printk(KERN_ERR MOD "%s - cxio_destroy_cq failed %d\n", - __FUNCTION__, ret); + __func__, ret); } /* add user hooks here */ @@ -316,7 +316,7 @@ static int iwch_arm_cq(struct ib_cq *ibcq, enum ib_cq_notify_flags flags) chp->cq.rptr = rptr; } else spin_lock_irqsave(&chp->lock, flag); - PDBG("%s rptr 0x%x\n", __FUNCTION__, chp->cq.rptr); + PDBG("%s rptr 0x%x\n", __func__, chp->cq.rptr); err = cxio_hal_cq_op(&rhp->rdev, &chp->cq, cq_op, 0); spin_unlock_irqrestore(&chp->lock, flag); if (err < 0) @@ -337,7 +337,7 @@ static int iwch_mmap(struct ib_ucontext *context, struct vm_area_struct *vma) struct iwch_ucontext *ucontext; u64 addr; - PDBG("%s pgoff 0x%lx key 0x%x len %d\n", __FUNCTION__, vma->vm_pgoff, + PDBG("%s pgoff 0x%lx key 0x%x len %d\n", __func__, vma->vm_pgoff, key, len); if (vma->vm_start & (PAGE_SIZE-1)) { @@ -390,7 +390,7 @@ static int iwch_deallocate_pd(struct ib_pd *pd) php = to_iwch_pd(pd); rhp = php->rhp; - PDBG("%s ibpd %p pdid 0x%x\n", __FUNCTION__, pd, php->pdid); + PDBG("%s ibpd %p pdid 0x%x\n", __func__, pd, php->pdid); cxio_hal_put_pdid(rhp->rdev.rscp, php->pdid); kfree(php); return 0; @@ -404,7 +404,7 @@ static struct ib_pd *iwch_allocate_pd(struct ib_device *ibdev, u32 pdid; struct iwch_dev *rhp; - PDBG("%s ibdev %p\n", __FUNCTION__, ibdev); + PDBG("%s ibdev %p\n", __func__, ibdev); rhp = (struct iwch_dev *) ibdev; pdid = cxio_hal_get_pdid(rhp->rdev.rscp); if (!pdid) @@ -422,7 +422,7 @@ static struct ib_pd *iwch_allocate_pd(struct ib_device *ibdev, return ERR_PTR(-EFAULT); } } - PDBG("%s pdid 0x%0x ptr 0x%p\n", __FUNCTION__, pdid, php); + PDBG("%s pdid 0x%0x ptr 0x%p\n", __func__, pdid, php); return &php->ibpd; } @@ -432,7 +432,7 @@ static int iwch_dereg_mr(struct ib_mr *ib_mr) struct iwch_mr *mhp; u32 mmid; - PDBG("%s ib_mr %p\n", __FUNCTION__, ib_mr); + PDBG("%s ib_mr %p\n", __func__, ib_mr); /* There can be no memory windows */ if (atomic_read(&ib_mr->usecnt)) return -EINVAL; @@ -447,7 +447,7 @@ static int iwch_dereg_mr(struct ib_mr *ib_mr) kfree((void *) (unsigned long) mhp->kva); if (mhp->umem) ib_umem_release(mhp->umem); - PDBG("%s mmid 0x%x ptr %p\n", __FUNCTION__, mmid, mhp); + PDBG("%s mmid 0x%x ptr %p\n", __func__, mmid, mhp); kfree(mhp); return 0; } @@ -467,7 +467,7 @@ static struct ib_mr *iwch_register_phys_mem(struct ib_pd *pd, struct iwch_mr *mhp; int ret; - PDBG("%s ib_pd %p\n", __FUNCTION__, pd); + PDBG("%s ib_pd %p\n", __func__, pd); php = to_iwch_pd(pd); rhp = php->rhp; @@ -531,7 +531,7 @@ static int iwch_reregister_phys_mem(struct ib_mr *mr, int npages; int ret; - PDBG("%s ib_mr %p ib_pd %p\n", __FUNCTION__, mr, pd); + PDBG("%s ib_mr %p ib_pd %p\n", __func__, mr, pd); /* There can be no memory windows */ if (atomic_read(&mr->usecnt)) @@ -594,7 +594,7 @@ static struct ib_mr *iwch_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, struct iwch_mr *mhp; struct iwch_reg_user_mr_resp uresp; - PDBG("%s ib_pd %p\n", __FUNCTION__, pd); + PDBG("%s ib_pd %p\n", __func__, pd); php = to_iwch_pd(pd); rhp = php->rhp; @@ -649,7 +649,7 @@ static struct ib_mr *iwch_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, if (udata && !t3a_device(rhp)) { uresp.pbl_addr = (mhp->attr.pbl_addr - rhp->rdev.rnic_info.pbl_base) >> 3; - PDBG("%s user resp pbl_addr 0x%x\n", __FUNCTION__, + PDBG("%s user resp pbl_addr 0x%x\n", __func__, uresp.pbl_addr); if (ib_copy_to_udata(udata, &uresp, sizeof (uresp))) { @@ -673,7 +673,7 @@ static struct ib_mr *iwch_get_dma_mr(struct ib_pd *pd, int acc) u64 kva; struct ib_mr *ibmr; - PDBG("%s ib_pd %p\n", __FUNCTION__, pd); + PDBG("%s ib_pd %p\n", __func__, pd); /* * T3 only supports 32 bits of size. @@ -710,7 +710,7 @@ static struct ib_mw *iwch_alloc_mw(struct ib_pd *pd) mhp->attr.stag = stag; mmid = (stag) >> 8; insert_handle(rhp, &rhp->mmidr, mhp, mmid); - PDBG("%s mmid 0x%x mhp %p stag 0x%x\n", __FUNCTION__, mmid, mhp, stag); + PDBG("%s mmid 0x%x mhp %p stag 0x%x\n", __func__, mmid, mhp, stag); return &(mhp->ibmw); } @@ -726,7 +726,7 @@ static int iwch_dealloc_mw(struct ib_mw *mw) cxio_deallocate_window(&rhp->rdev, mhp->attr.stag); remove_handle(rhp, &rhp->mmidr, mmid); kfree(mhp); - PDBG("%s ib_mw %p mmid 0x%x ptr %p\n", __FUNCTION__, mw, mmid, mhp); + PDBG("%s ib_mw %p mmid 0x%x ptr %p\n", __func__, mw, mmid, mhp); return 0; } @@ -754,7 +754,7 @@ static int iwch_destroy_qp(struct ib_qp *ib_qp) cxio_destroy_qp(&rhp->rdev, &qhp->wq, ucontext ? &ucontext->uctx : &rhp->rdev.uctx); - PDBG("%s ib_qp %p qpid 0x%0x qhp %p\n", __FUNCTION__, + PDBG("%s ib_qp %p qpid 0x%0x qhp %p\n", __func__, ib_qp, qhp->wq.qpid, qhp); kfree(qhp); return 0; @@ -773,7 +773,7 @@ static struct ib_qp *iwch_create_qp(struct ib_pd *pd, int wqsize, sqsize, rqsize; struct iwch_ucontext *ucontext; - PDBG("%s ib_pd %p\n", __FUNCTION__, pd); + PDBG("%s ib_pd %p\n", __func__, pd); if (attrs->qp_type != IB_QPT_RC) return ERR_PTR(-EINVAL); php = to_iwch_pd(pd); @@ -805,7 +805,7 @@ static struct ib_qp *iwch_create_qp(struct ib_pd *pd, */ sqsize = roundup_pow_of_two(attrs->cap.max_send_wr); wqsize = roundup_pow_of_two(rqsize + sqsize); - PDBG("%s wqsize %d sqsize %d rqsize %d\n", __FUNCTION__, + PDBG("%s wqsize %d sqsize %d rqsize %d\n", __func__, wqsize, sqsize, rqsize); qhp = kzalloc(sizeof(*qhp), GFP_KERNEL); if (!qhp) @@ -898,7 +898,7 @@ static struct ib_qp *iwch_create_qp(struct ib_pd *pd, init_timer(&(qhp->timer)); PDBG("%s sq_num_entries %d, rq_num_entries %d " "qpid 0x%0x qhp %p dma_addr 0x%llx size %d\n", - __FUNCTION__, qhp->attr.sq_num_entries, qhp->attr.rq_num_entries, + __func__, qhp->attr.sq_num_entries, qhp->attr.rq_num_entries, qhp->wq.qpid, qhp, (unsigned long long) qhp->wq.dma_addr, 1 << qhp->wq.size_log2); return &qhp->ibqp; @@ -912,7 +912,7 @@ static int iwch_ib_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, enum iwch_qp_attr_mask mask = 0; struct iwch_qp_attributes attrs; - PDBG("%s ib_qp %p\n", __FUNCTION__, ibqp); + PDBG("%s ib_qp %p\n", __func__, ibqp); /* iwarp does not support the RTR state */ if ((attr_mask & IB_QP_STATE) && (attr->qp_state == IB_QPS_RTR)) @@ -945,20 +945,20 @@ static int iwch_ib_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, void iwch_qp_add_ref(struct ib_qp *qp) { - PDBG("%s ib_qp %p\n", __FUNCTION__, qp); + PDBG("%s ib_qp %p\n", __func__, qp); atomic_inc(&(to_iwch_qp(qp)->refcnt)); } void iwch_qp_rem_ref(struct ib_qp *qp) { - PDBG("%s ib_qp %p\n", __FUNCTION__, qp); + PDBG("%s ib_qp %p\n", __func__, qp); if (atomic_dec_and_test(&(to_iwch_qp(qp)->refcnt))) wake_up(&(to_iwch_qp(qp)->wait)); } static struct ib_qp *iwch_get_qp(struct ib_device *dev, int qpn) { - PDBG("%s ib_dev %p qpn 0x%x\n", __FUNCTION__, dev, qpn); + PDBG("%s ib_dev %p qpn 0x%x\n", __func__, dev, qpn); return (struct ib_qp *)get_qhp(to_iwch_dev(dev), qpn); } @@ -966,7 +966,7 @@ static struct ib_qp *iwch_get_qp(struct ib_device *dev, int qpn) static int iwch_query_pkey(struct ib_device *ibdev, u8 port, u16 index, u16 * pkey) { - PDBG("%s ibdev %p\n", __FUNCTION__, ibdev); + PDBG("%s ibdev %p\n", __func__, ibdev); *pkey = 0; return 0; } @@ -977,7 +977,7 @@ static int iwch_query_gid(struct ib_device *ibdev, u8 port, struct iwch_dev *dev; PDBG("%s ibdev %p, port %d, index %d, gid %p\n", - __FUNCTION__, ibdev, port, index, gid); + __func__, ibdev, port, index, gid); dev = to_iwch_dev(ibdev); BUG_ON(port == 0 || port > 2); memset(&(gid->raw[0]), 0, sizeof(gid->raw)); @@ -990,7 +990,7 @@ static int iwch_query_device(struct ib_device *ibdev, { struct iwch_dev *dev; - PDBG("%s ibdev %p\n", __FUNCTION__, ibdev); + PDBG("%s ibdev %p\n", __func__, ibdev); dev = to_iwch_dev(ibdev); memset(props, 0, sizeof *props); @@ -1017,7 +1017,7 @@ static int iwch_query_device(struct ib_device *ibdev, static int iwch_query_port(struct ib_device *ibdev, u8 port, struct ib_port_attr *props) { - PDBG("%s ibdev %p\n", __FUNCTION__, ibdev); + PDBG("%s ibdev %p\n", __func__, ibdev); props->max_mtu = IB_MTU_4096; props->lid = 0; props->lmc = 0; @@ -1045,7 +1045,7 @@ static ssize_t show_rev(struct class_device *cdev, char *buf) { struct iwch_dev *dev = container_of(cdev, struct iwch_dev, ibdev.class_dev); - PDBG("%s class dev 0x%p\n", __FUNCTION__, cdev); + PDBG("%s class dev 0x%p\n", __func__, cdev); return sprintf(buf, "%d\n", dev->rdev.t3cdev_p->type); } @@ -1056,7 +1056,7 @@ static ssize_t show_fw_ver(struct class_device *cdev, char *buf) struct ethtool_drvinfo info; struct net_device *lldev = dev->rdev.t3cdev_p->lldev; - PDBG("%s class dev 0x%p\n", __FUNCTION__, cdev); + PDBG("%s class dev 0x%p\n", __func__, cdev); rtnl_lock(); lldev->ethtool_ops->get_drvinfo(lldev, &info); rtnl_unlock(); @@ -1070,7 +1070,7 @@ static ssize_t show_hca(struct class_device *cdev, char *buf) struct ethtool_drvinfo info; struct net_device *lldev = dev->rdev.t3cdev_p->lldev; - PDBG("%s class dev 0x%p\n", __FUNCTION__, cdev); + PDBG("%s class dev 0x%p\n", __func__, cdev); rtnl_lock(); lldev->ethtool_ops->get_drvinfo(lldev, &info); rtnl_unlock(); @@ -1081,7 +1081,7 @@ static ssize_t show_board(struct class_device *cdev, char *buf) { struct iwch_dev *dev = container_of(cdev, struct iwch_dev, ibdev.class_dev); - PDBG("%s class dev 0x%p\n", __FUNCTION__, dev); + PDBG("%s class dev 0x%p\n", __func__, dev); return sprintf(buf, "%x.%x\n", dev->rdev.rnic_info.pdev->vendor, dev->rdev.rnic_info.pdev->device); } @@ -1103,14 +1103,13 @@ int iwch_register_device(struct iwch_dev *dev) int ret; int i; - PDBG("%s iwch_dev %p\n", __FUNCTION__, dev); + PDBG("%s iwch_dev %p\n", __func__, dev); strlcpy(dev->ibdev.name, "cxgb3_%d", IB_DEVICE_NAME_MAX); memset(&dev->ibdev.node_guid, 0, sizeof(dev->ibdev.node_guid)); memcpy(&dev->ibdev.node_guid, dev->rdev.t3cdev_p->lldev->dev_addr, 6); dev->ibdev.owner = THIS_MODULE; dev->device_cap_flags = - (IB_DEVICE_ZERO_STAG | - IB_DEVICE_SEND_W_INV | IB_DEVICE_MEM_WINDOW); + (IB_DEVICE_ZERO_STAG | IB_DEVICE_MEM_WINDOW); dev->ibdev.uverbs_cmd_mask = (1ull << IB_USER_VERBS_CMD_GET_CONTEXT) | @@ -1207,7 +1206,7 @@ void iwch_unregister_device(struct iwch_dev *dev) { int i; - PDBG("%s iwch_dev %p\n", __FUNCTION__, dev); + PDBG("%s iwch_dev %p\n", __func__, dev); for (i = 0; i < ARRAY_SIZE(iwch_class_attributes); ++i) class_device_remove_file(&dev->ibdev.class_dev, iwch_class_attributes[i]); diff --git a/drivers/infiniband/hw/cxgb3/iwch_provider.h b/drivers/infiniband/hw/cxgb3/iwch_provider.h index 48833f3f3bd0..61356f91109d 100644 --- a/drivers/infiniband/hw/cxgb3/iwch_provider.h +++ b/drivers/infiniband/hw/cxgb3/iwch_provider.h @@ -213,7 +213,7 @@ static inline struct iwch_mm_entry *remove_mmap(struct iwch_ucontext *ucontext, if (mm->key == key && mm->len == len) { list_del_init(&mm->entry); spin_unlock(&ucontext->mmap_lock); - PDBG("%s key 0x%x addr 0x%llx len %d\n", __FUNCTION__, + PDBG("%s key 0x%x addr 0x%llx len %d\n", __func__, key, (unsigned long long) mm->addr, mm->len); return mm; } @@ -226,7 +226,7 @@ static inline void insert_mmap(struct iwch_ucontext *ucontext, struct iwch_mm_entry *mm) { spin_lock(&ucontext->mmap_lock); - PDBG("%s key 0x%x addr 0x%llx len %d\n", __FUNCTION__, + PDBG("%s key 0x%x addr 0x%llx len %d\n", __func__, mm->key, (unsigned long long) mm->addr, mm->len); list_add_tail(&mm->entry, &ucontext->mmaps); spin_unlock(&ucontext->mmap_lock); diff --git a/drivers/infiniband/hw/cxgb3/iwch_qp.c b/drivers/infiniband/hw/cxgb3/iwch_qp.c index ea2cdd73dd85..8891c3b0a3d5 100644 --- a/drivers/infiniband/hw/cxgb3/iwch_qp.c +++ b/drivers/infiniband/hw/cxgb3/iwch_qp.c @@ -72,7 +72,7 @@ static int iwch_build_rdma_send(union t3_wr *wqe, struct ib_send_wr *wr, wqe->send.reserved[2] = 0; if (wr->opcode == IB_WR_SEND_WITH_IMM) { plen = 4; - wqe->send.sgl[0].stag = wr->imm_data; + wqe->send.sgl[0].stag = wr->ex.imm_data; wqe->send.sgl[0].len = __constant_cpu_to_be32(0); wqe->send.num_sgle = __constant_cpu_to_be32(0); *flit_cnt = 5; @@ -112,7 +112,7 @@ static int iwch_build_rdma_write(union t3_wr *wqe, struct ib_send_wr *wr, if (wr->opcode == IB_WR_RDMA_WRITE_WITH_IMM) { plen = 4; - wqe->write.sgl[0].stag = wr->imm_data; + wqe->write.sgl[0].stag = wr->ex.imm_data; wqe->write.sgl[0].len = __constant_cpu_to_be32(0); wqe->write.num_sgle = __constant_cpu_to_be32(0); *flit_cnt = 6; @@ -168,30 +168,30 @@ static int iwch_sgl2pbl_map(struct iwch_dev *rhp, struct ib_sge *sg_list, mhp = get_mhp(rhp, (sg_list[i].lkey) >> 8); if (!mhp) { - PDBG("%s %d\n", __FUNCTION__, __LINE__); + PDBG("%s %d\n", __func__, __LINE__); return -EIO; } if (!mhp->attr.state) { - PDBG("%s %d\n", __FUNCTION__, __LINE__); + PDBG("%s %d\n", __func__, __LINE__); return -EIO; } if (mhp->attr.zbva) { - PDBG("%s %d\n", __FUNCTION__, __LINE__); + PDBG("%s %d\n", __func__, __LINE__); return -EIO; } if (sg_list[i].addr < mhp->attr.va_fbo) { - PDBG("%s %d\n", __FUNCTION__, __LINE__); + PDBG("%s %d\n", __func__, __LINE__); return -EINVAL; } if (sg_list[i].addr + ((u64) sg_list[i].length) < sg_list[i].addr) { - PDBG("%s %d\n", __FUNCTION__, __LINE__); + PDBG("%s %d\n", __func__, __LINE__); return -EINVAL; } if (sg_list[i].addr + ((u64) sg_list[i].length) > mhp->attr.va_fbo + ((u64) mhp->attr.len)) { - PDBG("%s %d\n", __FUNCTION__, __LINE__); + PDBG("%s %d\n", __func__, __LINE__); return -EINVAL; } offset = sg_list[i].addr - mhp->attr.va_fbo; @@ -290,7 +290,7 @@ int iwch_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, qhp->wq.oldest_read = sqp; break; default: - PDBG("%s post of type=%d TBD!\n", __FUNCTION__, + PDBG("%s post of type=%d TBD!\n", __func__, wr->opcode); err = -EINVAL; } @@ -309,7 +309,7 @@ int iwch_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, Q_GENBIT(qhp->wq.wptr, qhp->wq.size_log2), 0, t3_wr_flit_cnt); PDBG("%s cookie 0x%llx wq idx 0x%x swsq idx %ld opcode %d\n", - __FUNCTION__, (unsigned long long) wr->wr_id, idx, + __func__, (unsigned long long) wr->wr_id, idx, Q_PTR2IDX(qhp->wq.sq_wptr, qhp->wq.sq_size_log2), sqp->opcode); wr = wr->next; @@ -361,7 +361,7 @@ int iwch_post_receive(struct ib_qp *ibqp, struct ib_recv_wr *wr, Q_GENBIT(qhp->wq.wptr, qhp->wq.size_log2), 0, sizeof(struct t3_receive_wr) >> 3); PDBG("%s cookie 0x%llx idx 0x%x rq_wptr 0x%x rw_rptr 0x%x " - "wqe %p \n", __FUNCTION__, (unsigned long long) wr->wr_id, + "wqe %p \n", __func__, (unsigned long long) wr->wr_id, idx, qhp->wq.rq_wptr, qhp->wq.rq_rptr, wqe); ++(qhp->wq.rq_wptr); ++(qhp->wq.wptr); @@ -407,7 +407,7 @@ int iwch_bind_mw(struct ib_qp *qp, return -ENOMEM; } idx = Q_PTR2IDX(qhp->wq.wptr, qhp->wq.size_log2); - PDBG("%s: idx 0x%0x, mw 0x%p, mw_bind 0x%p\n", __FUNCTION__, idx, + PDBG("%s: idx 0x%0x, mw 0x%p, mw_bind 0x%p\n", __func__, idx, mw, mw_bind); wqe = (union t3_wr *) (qhp->wq.queue + idx); @@ -595,10 +595,10 @@ int iwch_post_terminate(struct iwch_qp *qhp, struct respQ_msg_t *rsp_msg) struct terminate_message *term; struct sk_buff *skb; - PDBG("%s %d\n", __FUNCTION__, __LINE__); + PDBG("%s %d\n", __func__, __LINE__); skb = alloc_skb(40, GFP_ATOMIC); if (!skb) { - printk(KERN_ERR "%s cannot send TERMINATE!\n", __FUNCTION__); + printk(KERN_ERR "%s cannot send TERMINATE!\n", __func__); return -ENOMEM; } wqe = (union t3_wr *)skb_put(skb, 40); @@ -629,7 +629,7 @@ static void __flush_qp(struct iwch_qp *qhp, unsigned long *flag) rchp = get_chp(qhp->rhp, qhp->attr.rcq); schp = get_chp(qhp->rhp, qhp->attr.scq); - PDBG("%s qhp %p rchp %p schp %p\n", __FUNCTION__, qhp, rchp, schp); + PDBG("%s qhp %p rchp %p schp %p\n", __func__, qhp, rchp, schp); /* take a ref on the qhp since we must release the lock */ atomic_inc(&qhp->refcnt); spin_unlock_irqrestore(&qhp->lock, *flag); @@ -720,11 +720,11 @@ static int rdma_init(struct iwch_dev *rhp, struct iwch_qp *qhp, init_attr.flags |= capable(CAP_NET_BIND_SERVICE) ? PRIV_QP : 0; init_attr.irs = qhp->ep->rcv_seq; PDBG("%s init_attr.rq_addr 0x%x init_attr.rq_size = %d " - "flags 0x%x qpcaps 0x%x\n", __FUNCTION__, + "flags 0x%x qpcaps 0x%x\n", __func__, init_attr.rq_addr, init_attr.rq_size, init_attr.flags, init_attr.qpcaps); ret = cxio_rdma_init(&rhp->rdev, &init_attr); - PDBG("%s ret %d\n", __FUNCTION__, ret); + PDBG("%s ret %d\n", __func__, ret); return ret; } @@ -742,7 +742,7 @@ int iwch_modify_qp(struct iwch_dev *rhp, struct iwch_qp *qhp, int free = 0; struct iwch_ep *ep = NULL; - PDBG("%s qhp %p qpid 0x%x ep %p state %d -> %d\n", __FUNCTION__, + PDBG("%s qhp %p qpid 0x%x ep %p state %d -> %d\n", __func__, qhp, qhp->wq.qpid, qhp->ep, qhp->attr.state, (mask & IWCH_QP_ATTR_NEXT_STATE) ? attrs->next_state : -1); @@ -899,14 +899,14 @@ int iwch_modify_qp(struct iwch_dev *rhp, struct iwch_qp *qhp, break; default: printk(KERN_ERR "%s in a bad state %d\n", - __FUNCTION__, qhp->attr.state); + __func__, qhp->attr.state); ret = -EINVAL; goto err; break; } goto out; err: - PDBG("%s disassociating ep %p qpid 0x%x\n", __FUNCTION__, qhp->ep, + PDBG("%s disassociating ep %p qpid 0x%x\n", __func__, qhp->ep, qhp->wq.qpid); /* disassociate the LLP connection */ @@ -939,7 +939,7 @@ out: if (free) put_ep(&ep->com); - PDBG("%s exit state %d\n", __FUNCTION__, qhp->attr.state); + PDBG("%s exit state %d\n", __func__, qhp->attr.state); return ret; } diff --git a/drivers/infiniband/hw/ehca/ehca_av.c b/drivers/infiniband/hw/ehca/ehca_av.c index 194c1c30cf63..56735ea2fc57 100644 --- a/drivers/infiniband/hw/ehca/ehca_av.c +++ b/drivers/infiniband/hw/ehca/ehca_av.c @@ -41,9 +41,6 @@ * POSSIBILITY OF SUCH DAMAGE. */ - -#include <asm/current.h> - #include "ehca_tools.h" #include "ehca_iverbs.h" #include "hcp_if.h" @@ -170,17 +167,8 @@ int ehca_modify_ah(struct ib_ah *ah, struct ib_ah_attr *ah_attr) { struct ehca_av *av; struct ehca_ud_av new_ehca_av; - struct ehca_pd *my_pd = container_of(ah->pd, struct ehca_pd, ib_pd); struct ehca_shca *shca = container_of(ah->pd->device, struct ehca_shca, ib_device); - u32 cur_pid = current->tgid; - - if (my_pd->ib_pd.uobject && my_pd->ib_pd.uobject->context && - my_pd->ownpid != cur_pid) { - ehca_err(ah->device, "Invalid caller pid=%x ownpid=%x", - cur_pid, my_pd->ownpid); - return -EINVAL; - } memset(&new_ehca_av, 0, sizeof(new_ehca_av)); new_ehca_av.sl = ah_attr->sl; @@ -242,15 +230,6 @@ int ehca_modify_ah(struct ib_ah *ah, struct ib_ah_attr *ah_attr) int ehca_query_ah(struct ib_ah *ah, struct ib_ah_attr *ah_attr) { struct ehca_av *av = container_of(ah, struct ehca_av, ib_ah); - struct ehca_pd *my_pd = container_of(ah->pd, struct ehca_pd, ib_pd); - u32 cur_pid = current->tgid; - - if (my_pd->ib_pd.uobject && my_pd->ib_pd.uobject->context && - my_pd->ownpid != cur_pid) { - ehca_err(ah->device, "Invalid caller pid=%x ownpid=%x", - cur_pid, my_pd->ownpid); - return -EINVAL; - } memcpy(&ah_attr->grh.dgid, &av->av.grh.word_3, sizeof(ah_attr->grh.dgid)); @@ -273,16 +252,6 @@ int ehca_query_ah(struct ib_ah *ah, struct ib_ah_attr *ah_attr) int ehca_destroy_ah(struct ib_ah *ah) { - struct ehca_pd *my_pd = container_of(ah->pd, struct ehca_pd, ib_pd); - u32 cur_pid = current->tgid; - - if (my_pd->ib_pd.uobject && my_pd->ib_pd.uobject->context && - my_pd->ownpid != cur_pid) { - ehca_err(ah->device, "Invalid caller pid=%x ownpid=%x", - cur_pid, my_pd->ownpid); - return -EINVAL; - } - kmem_cache_free(av_cache, container_of(ah, struct ehca_av, ib_ah)); return 0; diff --git a/drivers/infiniband/hw/ehca/ehca_classes.h b/drivers/infiniband/hw/ehca/ehca_classes.h index 92cce8aacbb7..0d13fe0a260b 100644 --- a/drivers/infiniband/hw/ehca/ehca_classes.h +++ b/drivers/infiniband/hw/ehca/ehca_classes.h @@ -132,7 +132,6 @@ struct ehca_shca { struct ehca_pd { struct ib_pd ib_pd; struct ipz_pd fw_pd; - u32 ownpid; /* small queue mgmt */ struct mutex lock; struct list_head free[2]; @@ -215,7 +214,6 @@ struct ehca_cq { atomic_t nr_events; /* #events seen */ wait_queue_head_t wait_completion; spinlock_t task_lock; - u32 ownpid; /* mmap counter for resources mapped into user space */ u32 mm_count_queue; u32 mm_count_galpa; diff --git a/drivers/infiniband/hw/ehca/ehca_cq.c b/drivers/infiniband/hw/ehca/ehca_cq.c index 0467c158d4a9..ec0cfcf3073f 100644 --- a/drivers/infiniband/hw/ehca/ehca_cq.c +++ b/drivers/infiniband/hw/ehca/ehca_cq.c @@ -43,8 +43,6 @@ * POSSIBILITY OF SUCH DAMAGE. */ -#include <asm/current.h> - #include "ehca_iverbs.h" #include "ehca_classes.h" #include "ehca_irq.h" @@ -148,7 +146,6 @@ struct ib_cq *ehca_create_cq(struct ib_device *device, int cqe, int comp_vector, spin_lock_init(&my_cq->task_lock); atomic_set(&my_cq->nr_events, 0); init_waitqueue_head(&my_cq->wait_completion); - my_cq->ownpid = current->tgid; cq = &my_cq->ib_cq; @@ -320,7 +317,6 @@ int ehca_destroy_cq(struct ib_cq *cq) struct ehca_shca *shca = container_of(device, struct ehca_shca, ib_device); struct ipz_adapter_handle adapter_handle = shca->ipz_hca_handle; - u32 cur_pid = current->tgid; unsigned long flags; if (cq->uobject) { @@ -329,12 +325,6 @@ int ehca_destroy_cq(struct ib_cq *cq) "user space cq_num=%x", my_cq->cq_number); return -EINVAL; } - if (my_cq->ownpid != cur_pid) { - ehca_err(device, "Invalid caller pid=%x ownpid=%x " - "cq_num=%x", - cur_pid, my_cq->ownpid, my_cq->cq_number); - return -EINVAL; - } } /* @@ -374,15 +364,6 @@ int ehca_destroy_cq(struct ib_cq *cq) int ehca_resize_cq(struct ib_cq *cq, int cqe, struct ib_udata *udata) { - struct ehca_cq *my_cq = container_of(cq, struct ehca_cq, ib_cq); - u32 cur_pid = current->tgid; - - if (cq->uobject && my_cq->ownpid != cur_pid) { - ehca_err(cq->device, "Invalid caller pid=%x ownpid=%x", - cur_pid, my_cq->ownpid); - return -EINVAL; - } - /* TODO: proper resize needs to be done */ ehca_err(cq->device, "not implemented yet"); diff --git a/drivers/infiniband/hw/ehca/ehca_hca.c b/drivers/infiniband/hw/ehca/ehca_hca.c index 5bd7b591987e..2515cbde7e65 100644 --- a/drivers/infiniband/hw/ehca/ehca_hca.c +++ b/drivers/infiniband/hw/ehca/ehca_hca.c @@ -43,6 +43,11 @@ #include "ehca_iverbs.h" #include "hcp_if.h" +static unsigned int limit_uint(unsigned int value) +{ + return min_t(unsigned int, value, INT_MAX); +} + int ehca_query_device(struct ib_device *ibdev, struct ib_device_attr *props) { int i, ret = 0; @@ -83,37 +88,40 @@ int ehca_query_device(struct ib_device *ibdev, struct ib_device_attr *props) props->vendor_id = rblock->vendor_id >> 8; props->vendor_part_id = rblock->vendor_part_id >> 16; props->hw_ver = rblock->hw_ver; - props->max_qp = min_t(unsigned, rblock->max_qp, INT_MAX); - props->max_qp_wr = min_t(unsigned, rblock->max_wqes_wq, INT_MAX); - props->max_sge = min_t(unsigned, rblock->max_sge, INT_MAX); - props->max_sge_rd = min_t(unsigned, rblock->max_sge_rd, INT_MAX); - props->max_cq = min_t(unsigned, rblock->max_cq, INT_MAX); - props->max_cqe = min_t(unsigned, rblock->max_cqe, INT_MAX); - props->max_mr = min_t(unsigned, rblock->max_mr, INT_MAX); - props->max_mw = min_t(unsigned, rblock->max_mw, INT_MAX); - props->max_pd = min_t(unsigned, rblock->max_pd, INT_MAX); - props->max_ah = min_t(unsigned, rblock->max_ah, INT_MAX); - props->max_fmr = min_t(unsigned, rblock->max_mr, INT_MAX); + props->max_qp = limit_uint(rblock->max_qp); + props->max_qp_wr = limit_uint(rblock->max_wqes_wq); + props->max_sge = limit_uint(rblock->max_sge); + props->max_sge_rd = limit_uint(rblock->max_sge_rd); + props->max_cq = limit_uint(rblock->max_cq); + props->max_cqe = limit_uint(rblock->max_cqe); + props->max_mr = limit_uint(rblock->max_mr); + props->max_mw = limit_uint(rblock->max_mw); + props->max_pd = limit_uint(rblock->max_pd); + props->max_ah = limit_uint(rblock->max_ah); + props->max_ee = limit_uint(rblock->max_rd_ee_context); + props->max_rdd = limit_uint(rblock->max_rd_domain); + props->max_fmr = limit_uint(rblock->max_mr); + props->local_ca_ack_delay = limit_uint(rblock->local_ca_ack_delay); + props->max_qp_rd_atom = limit_uint(rblock->max_rr_qp); + props->max_ee_rd_atom = limit_uint(rblock->max_rr_ee_context); + props->max_res_rd_atom = limit_uint(rblock->max_rr_hca); + props->max_qp_init_rd_atom = limit_uint(rblock->max_act_wqs_qp); + props->max_ee_init_rd_atom = limit_uint(rblock->max_act_wqs_ee_context); if (EHCA_BMASK_GET(HCA_CAP_SRQ, shca->hca_cap)) { - props->max_srq = props->max_qp; - props->max_srq_wr = props->max_qp_wr; + props->max_srq = limit_uint(props->max_qp); + props->max_srq_wr = limit_uint(props->max_qp_wr); props->max_srq_sge = 3; } - props->max_pkeys = 16; - props->local_ca_ack_delay - = rblock->local_ca_ack_delay; - props->max_raw_ipv6_qp - = min_t(unsigned, rblock->max_raw_ipv6_qp, INT_MAX); - props->max_raw_ethy_qp - = min_t(unsigned, rblock->max_raw_ethy_qp, INT_MAX); - props->max_mcast_grp - = min_t(unsigned, rblock->max_mcast_grp, INT_MAX); - props->max_mcast_qp_attach - = min_t(unsigned, rblock->max_mcast_qp_attach, INT_MAX); + props->max_pkeys = 16; + props->local_ca_ack_delay = limit_uint(rblock->local_ca_ack_delay); + props->max_raw_ipv6_qp = limit_uint(rblock->max_raw_ipv6_qp); + props->max_raw_ethy_qp = limit_uint(rblock->max_raw_ethy_qp); + props->max_mcast_grp = limit_uint(rblock->max_mcast_grp); + props->max_mcast_qp_attach = limit_uint(rblock->max_mcast_qp_attach); props->max_total_mcast_qp_attach - = min_t(unsigned, rblock->max_total_mcast_qp_attach, INT_MAX); + = limit_uint(rblock->max_total_mcast_qp_attach); /* translate device capabilities */ props->device_cap_flags = IB_DEVICE_SYS_IMAGE_GUID | @@ -128,6 +136,46 @@ query_device1: return ret; } +static int map_mtu(struct ehca_shca *shca, u32 fw_mtu) +{ + switch (fw_mtu) { + case 0x1: + return IB_MTU_256; + case 0x2: + return IB_MTU_512; + case 0x3: + return IB_MTU_1024; + case 0x4: + return IB_MTU_2048; + case 0x5: + return IB_MTU_4096; + default: + ehca_err(&shca->ib_device, "Unknown MTU size: %x.", + fw_mtu); + return 0; + } +} + +static int map_number_of_vls(struct ehca_shca *shca, u32 vl_cap) +{ + switch (vl_cap) { + case 0x1: + return 1; + case 0x2: + return 2; + case 0x3: + return 4; + case 0x4: + return 8; + case 0x5: + return 15; + default: + ehca_err(&shca->ib_device, "invalid Vl Capability: %x.", + vl_cap); + return 0; + } +} + int ehca_query_port(struct ib_device *ibdev, u8 port, struct ib_port_attr *props) { @@ -152,31 +200,13 @@ int ehca_query_port(struct ib_device *ibdev, memset(props, 0, sizeof(struct ib_port_attr)); - switch (rblock->max_mtu) { - case 0x1: - props->active_mtu = props->max_mtu = IB_MTU_256; - break; - case 0x2: - props->active_mtu = props->max_mtu = IB_MTU_512; - break; - case 0x3: - props->active_mtu = props->max_mtu = IB_MTU_1024; - break; - case 0x4: - props->active_mtu = props->max_mtu = IB_MTU_2048; - break; - case 0x5: - props->active_mtu = props->max_mtu = IB_MTU_4096; - break; - default: - ehca_err(&shca->ib_device, "Unknown MTU size: %x.", - rblock->max_mtu); - break; - } - + props->active_mtu = props->max_mtu = map_mtu(shca, rblock->max_mtu); props->port_cap_flags = rblock->capability_mask; props->gid_tbl_len = rblock->gid_tbl_len; - props->max_msg_sz = rblock->max_msg_sz; + if (rblock->max_msg_sz) + props->max_msg_sz = rblock->max_msg_sz; + else + props->max_msg_sz = 0x1 << 31; props->bad_pkey_cntr = rblock->bad_pkey_cntr; props->qkey_viol_cntr = rblock->qkey_viol_cntr; props->pkey_tbl_len = rblock->pkey_tbl_len; @@ -186,6 +216,7 @@ int ehca_query_port(struct ib_device *ibdev, props->sm_sl = rblock->sm_sl; props->subnet_timeout = rblock->subnet_timeout; props->init_type_reply = rblock->init_type_reply; + props->max_vl_num = map_number_of_vls(shca, rblock->vl_cap); if (rblock->state && rblock->phys_width) { props->phys_state = rblock->phys_pstate; @@ -314,7 +345,7 @@ query_gid1: return ret; } -const u32 allowed_port_caps = ( +static const u32 allowed_port_caps = ( IB_PORT_SM | IB_PORT_LED_INFO_SUP | IB_PORT_CM_SUP | IB_PORT_SNMP_TUNNEL_SUP | IB_PORT_DEVICE_MGMT_SUP | IB_PORT_VENDOR_CLASS_SUP); diff --git a/drivers/infiniband/hw/ehca/ehca_main.c b/drivers/infiniband/hw/ehca/ehca_main.c index a86ebcc79a95..65b3362cdb9b 100644 --- a/drivers/infiniband/hw/ehca/ehca_main.c +++ b/drivers/infiniband/hw/ehca/ehca_main.c @@ -57,16 +57,17 @@ MODULE_AUTHOR("Christoph Raisch <raisch@de.ibm.com>"); MODULE_DESCRIPTION("IBM eServer HCA InfiniBand Device Driver"); MODULE_VERSION(HCAD_VERSION); -int ehca_open_aqp1 = 0; +static int ehca_open_aqp1 = 0; +static int ehca_hw_level = 0; +static int ehca_poll_all_eqs = 1; +static int ehca_mr_largepage = 1; + int ehca_debug_level = 0; -int ehca_hw_level = 0; int ehca_nr_ports = 2; int ehca_use_hp_mr = 0; int ehca_port_act_time = 30; -int ehca_poll_all_eqs = 1; int ehca_static_rate = -1; int ehca_scaling_code = 0; -int ehca_mr_largepage = 1; int ehca_lock_hcalls = -1; module_param_named(open_aqp1, ehca_open_aqp1, int, S_IRUGO); @@ -396,7 +397,7 @@ init_node_guid1: return ret; } -int ehca_init_device(struct ehca_shca *shca) +static int ehca_init_device(struct ehca_shca *shca) { int ret; @@ -579,8 +580,8 @@ static ssize_t ehca_store_debug_level(struct device_driver *ddp, return 1; } -DRIVER_ATTR(debug_level, S_IRUSR | S_IWUSR, - ehca_show_debug_level, ehca_store_debug_level); +static DRIVER_ATTR(debug_level, S_IRUSR | S_IWUSR, + ehca_show_debug_level, ehca_store_debug_level); static struct attribute *ehca_drv_attrs[] = { &driver_attr_debug_level.attr, @@ -941,7 +942,7 @@ void ehca_poll_eqs(unsigned long data) spin_unlock(&shca_list_lock); } -int __init ehca_module_init(void) +static int __init ehca_module_init(void) { int ret; @@ -988,7 +989,7 @@ module_init1: return ret; }; -void __exit ehca_module_exit(void) +static void __exit ehca_module_exit(void) { if (ehca_poll_all_eqs == 1) del_timer_sync(&poll_eqs_timer); diff --git a/drivers/infiniband/hw/ehca/ehca_mrmw.c b/drivers/infiniband/hw/ehca/ehca_mrmw.c index e239bbf54da1..f26997fc00f8 100644 --- a/drivers/infiniband/hw/ehca/ehca_mrmw.c +++ b/drivers/infiniband/hw/ehca/ehca_mrmw.c @@ -40,8 +40,6 @@ * POSSIBILITY OF SUCH DAMAGE. */ -#include <asm/current.h> - #include <rdma/ib_umem.h> #include "ehca_iverbs.h" @@ -419,7 +417,6 @@ int ehca_rereg_phys_mr(struct ib_mr *mr, struct ehca_shca *shca = container_of(mr->device, struct ehca_shca, ib_device); struct ehca_mr *e_mr = container_of(mr, struct ehca_mr, ib.ib_mr); - struct ehca_pd *my_pd = container_of(mr->pd, struct ehca_pd, ib_pd); u64 new_size; u64 *new_start; u32 new_acl; @@ -429,15 +426,6 @@ int ehca_rereg_phys_mr(struct ib_mr *mr, u32 num_kpages = 0; u32 num_hwpages = 0; struct ehca_mr_pginfo pginfo; - u32 cur_pid = current->tgid; - - if (my_pd->ib_pd.uobject && my_pd->ib_pd.uobject->context && - (my_pd->ownpid != cur_pid)) { - ehca_err(mr->device, "Invalid caller pid=%x ownpid=%x", - cur_pid, my_pd->ownpid); - ret = -EINVAL; - goto rereg_phys_mr_exit0; - } if (!(mr_rereg_mask & IB_MR_REREG_TRANS)) { /* TODO not supported, because PHYP rereg hCall needs pages */ @@ -577,19 +565,9 @@ int ehca_query_mr(struct ib_mr *mr, struct ib_mr_attr *mr_attr) struct ehca_shca *shca = container_of(mr->device, struct ehca_shca, ib_device); struct ehca_mr *e_mr = container_of(mr, struct ehca_mr, ib.ib_mr); - struct ehca_pd *my_pd = container_of(mr->pd, struct ehca_pd, ib_pd); - u32 cur_pid = current->tgid; unsigned long sl_flags; struct ehca_mr_hipzout_parms hipzout; - if (my_pd->ib_pd.uobject && my_pd->ib_pd.uobject->context && - (my_pd->ownpid != cur_pid)) { - ehca_err(mr->device, "Invalid caller pid=%x ownpid=%x", - cur_pid, my_pd->ownpid); - ret = -EINVAL; - goto query_mr_exit0; - } - if ((e_mr->flags & EHCA_MR_FLAG_FMR)) { ehca_err(mr->device, "not supported for FMR, mr=%p e_mr=%p " "e_mr->flags=%x", mr, e_mr, e_mr->flags); @@ -634,16 +612,6 @@ int ehca_dereg_mr(struct ib_mr *mr) struct ehca_shca *shca = container_of(mr->device, struct ehca_shca, ib_device); struct ehca_mr *e_mr = container_of(mr, struct ehca_mr, ib.ib_mr); - struct ehca_pd *my_pd = container_of(mr->pd, struct ehca_pd, ib_pd); - u32 cur_pid = current->tgid; - - if (my_pd->ib_pd.uobject && my_pd->ib_pd.uobject->context && - (my_pd->ownpid != cur_pid)) { - ehca_err(mr->device, "Invalid caller pid=%x ownpid=%x", - cur_pid, my_pd->ownpid); - ret = -EINVAL; - goto dereg_mr_exit0; - } if ((e_mr->flags & EHCA_MR_FLAG_FMR)) { ehca_err(mr->device, "not supported for FMR, mr=%p e_mr=%p " @@ -1952,9 +1920,8 @@ next_kpage: return ret; } -int ehca_set_pagebuf_phys(struct ehca_mr_pginfo *pginfo, - u32 number, - u64 *kpage) +static int ehca_set_pagebuf_phys(struct ehca_mr_pginfo *pginfo, + u32 number, u64 *kpage) { int ret = 0; struct ib_phys_buf *pbuf; @@ -2012,9 +1979,8 @@ int ehca_set_pagebuf_phys(struct ehca_mr_pginfo *pginfo, return ret; } -int ehca_set_pagebuf_fmr(struct ehca_mr_pginfo *pginfo, - u32 number, - u64 *kpage) +static int ehca_set_pagebuf_fmr(struct ehca_mr_pginfo *pginfo, + u32 number, u64 *kpage) { int ret = 0; u64 *fmrlist; diff --git a/drivers/infiniband/hw/ehca/ehca_pd.c b/drivers/infiniband/hw/ehca/ehca_pd.c index 43bcf085fcf2..2fe554855fa5 100644 --- a/drivers/infiniband/hw/ehca/ehca_pd.c +++ b/drivers/infiniband/hw/ehca/ehca_pd.c @@ -38,8 +38,6 @@ * POSSIBILITY OF SUCH DAMAGE. */ -#include <asm/current.h> - #include "ehca_tools.h" #include "ehca_iverbs.h" @@ -58,7 +56,6 @@ struct ib_pd *ehca_alloc_pd(struct ib_device *device, return ERR_PTR(-ENOMEM); } - pd->ownpid = current->tgid; for (i = 0; i < 2; i++) { INIT_LIST_HEAD(&pd->free[i]); INIT_LIST_HEAD(&pd->full[i]); @@ -85,18 +82,10 @@ struct ib_pd *ehca_alloc_pd(struct ib_device *device, int ehca_dealloc_pd(struct ib_pd *pd) { - u32 cur_pid = current->tgid; struct ehca_pd *my_pd = container_of(pd, struct ehca_pd, ib_pd); int i, leftovers = 0; struct ipz_small_queue_page *page, *tmp; - if (my_pd->ib_pd.uobject && my_pd->ib_pd.uobject->context && - my_pd->ownpid != cur_pid) { - ehca_err(pd->device, "Invalid caller pid=%x ownpid=%x", - cur_pid, my_pd->ownpid); - return -EINVAL; - } - for (i = 0; i < 2; i++) { list_splice(&my_pd->full[i], &my_pd->free[i]); list_for_each_entry_safe(page, tmp, &my_pd->free[i], list) { diff --git a/drivers/infiniband/hw/ehca/ehca_qp.c b/drivers/infiniband/hw/ehca/ehca_qp.c index 1012f15a7140..3eb14a52cbf2 100644 --- a/drivers/infiniband/hw/ehca/ehca_qp.c +++ b/drivers/infiniband/hw/ehca/ehca_qp.c @@ -43,9 +43,6 @@ * POSSIBILITY OF SUCH DAMAGE. */ - -#include <asm/current.h> - #include "ehca_classes.h" #include "ehca_tools.h" #include "ehca_qes.h" @@ -424,6 +421,9 @@ static struct ehca_qp *internal_create_qp( u32 swqe_size = 0, rwqe_size = 0, ib_qp_num; unsigned long flags; + if (init_attr->create_flags) + return ERR_PTR(-EINVAL); + memset(&parms, 0, sizeof(parms)); qp_type = init_attr->qp_type; @@ -1526,16 +1526,6 @@ int ehca_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, int attr_mask, struct ehca_shca *shca = container_of(ibqp->device, struct ehca_shca, ib_device); struct ehca_qp *my_qp = container_of(ibqp, struct ehca_qp, ib_qp); - struct ehca_pd *my_pd = container_of(my_qp->ib_qp.pd, struct ehca_pd, - ib_pd); - u32 cur_pid = current->tgid; - - if (my_pd->ib_pd.uobject && my_pd->ib_pd.uobject->context && - my_pd->ownpid != cur_pid) { - ehca_err(ibqp->pd->device, "Invalid caller pid=%x ownpid=%x", - cur_pid, my_pd->ownpid); - return -EINVAL; - } /* The if-block below caches qp_attr to be modified for GSI and SMI * qps during the initialization by ib_mad. When the respective port @@ -1636,23 +1626,13 @@ int ehca_query_qp(struct ib_qp *qp, int qp_attr_mask, struct ib_qp_init_attr *qp_init_attr) { struct ehca_qp *my_qp = container_of(qp, struct ehca_qp, ib_qp); - struct ehca_pd *my_pd = container_of(my_qp->ib_qp.pd, struct ehca_pd, - ib_pd); struct ehca_shca *shca = container_of(qp->device, struct ehca_shca, ib_device); struct ipz_adapter_handle adapter_handle = shca->ipz_hca_handle; struct hcp_modify_qp_control_block *qpcb; - u32 cur_pid = current->tgid; int cnt, ret = 0; u64 h_ret; - if (my_pd->ib_pd.uobject && my_pd->ib_pd.uobject->context && - my_pd->ownpid != cur_pid) { - ehca_err(qp->device, "Invalid caller pid=%x ownpid=%x", - cur_pid, my_pd->ownpid); - return -EINVAL; - } - if (qp_attr_mask & QP_ATTR_QUERY_NOT_SUPPORTED) { ehca_err(qp->device, "Invalid attribute mask " "ehca_qp=%p qp_num=%x qp_attr_mask=%x ", @@ -1797,8 +1777,6 @@ int ehca_modify_srq(struct ib_srq *ibsrq, struct ib_srq_attr *attr, { struct ehca_qp *my_qp = container_of(ibsrq, struct ehca_qp, ib_srq); - struct ehca_pd *my_pd = - container_of(ibsrq->pd, struct ehca_pd, ib_pd); struct ehca_shca *shca = container_of(ibsrq->pd->device, struct ehca_shca, ib_device); struct hcp_modify_qp_control_block *mqpcb; @@ -1806,14 +1784,6 @@ int ehca_modify_srq(struct ib_srq *ibsrq, struct ib_srq_attr *attr, u64 h_ret; int ret = 0; - u32 cur_pid = current->tgid; - if (my_pd->ib_pd.uobject && my_pd->ib_pd.uobject->context && - my_pd->ownpid != cur_pid) { - ehca_err(ibsrq->pd->device, "Invalid caller pid=%x ownpid=%x", - cur_pid, my_pd->ownpid); - return -EINVAL; - } - mqpcb = ehca_alloc_fw_ctrlblock(GFP_KERNEL); if (!mqpcb) { ehca_err(ibsrq->device, "Could not get zeroed page for mqpcb " @@ -1864,22 +1834,13 @@ modify_srq_exit0: int ehca_query_srq(struct ib_srq *srq, struct ib_srq_attr *srq_attr) { struct ehca_qp *my_qp = container_of(srq, struct ehca_qp, ib_srq); - struct ehca_pd *my_pd = container_of(srq->pd, struct ehca_pd, ib_pd); struct ehca_shca *shca = container_of(srq->device, struct ehca_shca, ib_device); struct ipz_adapter_handle adapter_handle = shca->ipz_hca_handle; struct hcp_modify_qp_control_block *qpcb; - u32 cur_pid = current->tgid; int ret = 0; u64 h_ret; - if (my_pd->ib_pd.uobject && my_pd->ib_pd.uobject->context && - my_pd->ownpid != cur_pid) { - ehca_err(srq->device, "Invalid caller pid=%x ownpid=%x", - cur_pid, my_pd->ownpid); - return -EINVAL; - } - qpcb = ehca_alloc_fw_ctrlblock(GFP_KERNEL); if (!qpcb) { ehca_err(srq->device, "Out of memory for qpcb " @@ -1919,7 +1880,6 @@ static int internal_destroy_qp(struct ib_device *dev, struct ehca_qp *my_qp, struct ehca_pd *my_pd = container_of(my_qp->ib_qp.pd, struct ehca_pd, ib_pd); struct ehca_sport *sport = &shca->sport[my_qp->init_attr.port_num - 1]; - u32 cur_pid = current->tgid; u32 qp_num = my_qp->real_qp_num; int ret; u64 h_ret; @@ -1934,11 +1894,6 @@ static int internal_destroy_qp(struct ib_device *dev, struct ehca_qp *my_qp, "user space qp_num=%x", qp_num); return -EINVAL; } - if (my_pd->ownpid != cur_pid) { - ehca_err(dev, "Invalid caller pid=%x ownpid=%x", - cur_pid, my_pd->ownpid); - return -EINVAL; - } } if (my_qp->send_cq) { diff --git a/drivers/infiniband/hw/ehca/ehca_reqs.c b/drivers/infiniband/hw/ehca/ehca_reqs.c index 2ce8cffb8664..a20bbf466188 100644 --- a/drivers/infiniband/hw/ehca/ehca_reqs.c +++ b/drivers/infiniband/hw/ehca/ehca_reqs.c @@ -188,7 +188,7 @@ static inline int ehca_write_swqe(struct ehca_qp *qp, if (send_wr->opcode == IB_WR_SEND_WITH_IMM || send_wr->opcode == IB_WR_RDMA_WRITE_WITH_IMM) { /* this might not work as long as HW does not support it */ - wqe_p->immediate_data = be32_to_cpu(send_wr->imm_data); + wqe_p->immediate_data = be32_to_cpu(send_wr->ex.imm_data); wqe_p->wr_flag |= WQE_WRFLAG_IMM_DATA_PRESENT; } diff --git a/drivers/infiniband/hw/ehca/ehca_tools.h b/drivers/infiniband/hw/ehca/ehca_tools.h index 4a8346a2bc9e..ec950bf8c479 100644 --- a/drivers/infiniband/hw/ehca/ehca_tools.h +++ b/drivers/infiniband/hw/ehca/ehca_tools.h @@ -73,37 +73,37 @@ extern int ehca_debug_level; if (unlikely(ehca_debug_level)) \ dev_printk(KERN_DEBUG, (ib_dev)->dma_device, \ "PU%04x EHCA_DBG:%s " format "\n", \ - raw_smp_processor_id(), __FUNCTION__, \ + raw_smp_processor_id(), __func__, \ ## arg); \ } while (0) #define ehca_info(ib_dev, format, arg...) \ dev_info((ib_dev)->dma_device, "PU%04x EHCA_INFO:%s " format "\n", \ - raw_smp_processor_id(), __FUNCTION__, ## arg) + raw_smp_processor_id(), __func__, ## arg) #define ehca_warn(ib_dev, format, arg...) \ dev_warn((ib_dev)->dma_device, "PU%04x EHCA_WARN:%s " format "\n", \ - raw_smp_processor_id(), __FUNCTION__, ## arg) + raw_smp_processor_id(), __func__, ## arg) #define ehca_err(ib_dev, format, arg...) \ dev_err((ib_dev)->dma_device, "PU%04x EHCA_ERR:%s " format "\n", \ - raw_smp_processor_id(), __FUNCTION__, ## arg) + raw_smp_processor_id(), __func__, ## arg) /* use this one only if no ib_dev available */ #define ehca_gen_dbg(format, arg...) \ do { \ if (unlikely(ehca_debug_level)) \ printk(KERN_DEBUG "PU%04x EHCA_DBG:%s " format "\n", \ - raw_smp_processor_id(), __FUNCTION__, ## arg); \ + raw_smp_processor_id(), __func__, ## arg); \ } while (0) #define ehca_gen_warn(format, arg...) \ printk(KERN_INFO "PU%04x EHCA_WARN:%s " format "\n", \ - raw_smp_processor_id(), __FUNCTION__, ## arg) + raw_smp_processor_id(), __func__, ## arg) #define ehca_gen_err(format, arg...) \ printk(KERN_ERR "PU%04x EHCA_ERR:%s " format "\n", \ - raw_smp_processor_id(), __FUNCTION__, ## arg) + raw_smp_processor_id(), __func__, ## arg) /** * ehca_dmp - printk a memory block, whose length is n*8 bytes. @@ -118,7 +118,7 @@ extern int ehca_debug_level; for (x = 0; x < l; x += 16) { \ printk(KERN_INFO "EHCA_DMP:%s " format \ " adr=%p ofs=%04x %016lx %016lx\n", \ - __FUNCTION__, ##args, deb, x, \ + __func__, ##args, deb, x, \ *((u64 *)&deb[0]), *((u64 *)&deb[8])); \ deb += 16; \ } \ diff --git a/drivers/infiniband/hw/ehca/ehca_uverbs.c b/drivers/infiniband/hw/ehca/ehca_uverbs.c index 5234d6c15c49..1b07f2beafaf 100644 --- a/drivers/infiniband/hw/ehca/ehca_uverbs.c +++ b/drivers/infiniband/hw/ehca/ehca_uverbs.c @@ -40,8 +40,6 @@ * POSSIBILITY OF SUCH DAMAGE. */ -#include <asm/current.h> - #include "ehca_classes.h" #include "ehca_iverbs.h" #include "ehca_mrmw.h" @@ -253,11 +251,9 @@ int ehca_mmap(struct ib_ucontext *context, struct vm_area_struct *vma) u32 idr_handle = fileoffset & 0x1FFFFFF; u32 q_type = (fileoffset >> 27) & 0x1; /* CQ, QP,... */ u32 rsrc_type = (fileoffset >> 25) & 0x3; /* sq,rq,cmnd_window */ - u32 cur_pid = current->tgid; u32 ret; struct ehca_cq *cq; struct ehca_qp *qp; - struct ehca_pd *pd; struct ib_uobject *uobject; switch (q_type) { @@ -270,13 +266,6 @@ int ehca_mmap(struct ib_ucontext *context, struct vm_area_struct *vma) if (!cq) return -EINVAL; - if (cq->ownpid != cur_pid) { - ehca_err(cq->ib_cq.device, - "Invalid caller pid=%x ownpid=%x", - cur_pid, cq->ownpid); - return -ENOMEM; - } - if (!cq->ib_cq.uobject || cq->ib_cq.uobject->context != context) return -EINVAL; @@ -298,14 +287,6 @@ int ehca_mmap(struct ib_ucontext *context, struct vm_area_struct *vma) if (!qp) return -EINVAL; - pd = container_of(qp->ib_qp.pd, struct ehca_pd, ib_pd); - if (pd->ownpid != cur_pid) { - ehca_err(qp->ib_qp.device, - "Invalid caller pid=%x ownpid=%x", - cur_pid, pd->ownpid); - return -ENOMEM; - } - uobject = IS_SRQ(qp) ? qp->ib_srq.uobject : qp->ib_qp.uobject; if (!uobject || uobject->context != context) return -EINVAL; diff --git a/drivers/infiniband/hw/ipath/Makefile b/drivers/infiniband/hw/ipath/Makefile index fe6738826865..75a6c91944c4 100644 --- a/drivers/infiniband/hw/ipath/Makefile +++ b/drivers/infiniband/hw/ipath/Makefile @@ -20,17 +20,20 @@ ib_ipath-y := \ ipath_qp.o \ ipath_rc.o \ ipath_ruc.o \ + ipath_sdma.o \ ipath_srq.o \ ipath_stats.o \ ipath_sysfs.o \ ipath_uc.o \ ipath_ud.o \ ipath_user_pages.o \ + ipath_user_sdma.o \ ipath_verbs_mcast.o \ ipath_verbs.o ib_ipath-$(CONFIG_HT_IRQ) += ipath_iba6110.o ib_ipath-$(CONFIG_PCI_MSI) += ipath_iba6120.o +ib_ipath-$(CONFIG_PCI_MSI) += ipath_iba7220.o ipath_sd7220.o ipath_sd7220_img.o ib_ipath-$(CONFIG_X86_64) += ipath_wc_x86_64.o ib_ipath-$(CONFIG_PPC64) += ipath_wc_ppc64.o diff --git a/drivers/infiniband/hw/ipath/ipath_7220.h b/drivers/infiniband/hw/ipath/ipath_7220.h new file mode 100644 index 000000000000..74fa5cc5131d --- /dev/null +++ b/drivers/infiniband/hw/ipath/ipath_7220.h @@ -0,0 +1,57 @@ +#ifndef _IPATH_7220_H +#define _IPATH_7220_H +/* + * Copyright (c) 2007 QLogic Corporation. All rights reserved. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * OpenIB.org BSD license below: + * + * Redistribution and use in source and binary forms, with or + * without modification, are permitted provided that the following + * conditions are met: + * + * - Redistributions of source code must retain the above + * copyright notice, this list of conditions and the following + * disclaimer. + * + * - Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials + * provided with the distribution. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + */ + +/* + * This header file provides the declarations and common definitions + * for (mostly) manipulation of the SerDes blocks within the IBA7220. + * the functions declared should only be called from within other + * 7220-related files such as ipath_iba7220.c or ipath_sd7220.c. + */ +int ipath_sd7220_presets(struct ipath_devdata *dd); +int ipath_sd7220_init(struct ipath_devdata *dd, int was_reset); +int ipath_sd7220_prog_ld(struct ipath_devdata *dd, int sdnum, u8 *img, + int len, int offset); +int ipath_sd7220_prog_vfy(struct ipath_devdata *dd, int sdnum, const u8 *img, + int len, int offset); +/* + * Below used for sdnum parameter, selecting one of the two sections + * used for PCIe, or the single SerDes used for IB, which is the + * only one currently used + */ +#define IB_7220_SERDES 2 + +int ipath_sd7220_ib_load(struct ipath_devdata *dd); +int ipath_sd7220_ib_vfy(struct ipath_devdata *dd); + +#endif /* _IPATH_7220_H */ diff --git a/drivers/infiniband/hw/ipath/ipath_common.h b/drivers/infiniband/hw/ipath/ipath_common.h index 591901aab6b7..28cfe97cf1e9 100644 --- a/drivers/infiniband/hw/ipath/ipath_common.h +++ b/drivers/infiniband/hw/ipath/ipath_common.h @@ -1,5 +1,5 @@ /* - * Copyright (c) 2006, 2007 QLogic Corporation. All rights reserved. + * Copyright (c) 2006, 2007, 2008 QLogic Corporation. All rights reserved. * Copyright (c) 2003, 2004, 2005, 2006 PathScale, Inc. All rights reserved. * * This software is available to you under a choice of one of two @@ -80,6 +80,8 @@ #define IPATH_IB_LINKDOWN_DISABLE 5 #define IPATH_IB_LINK_LOOPBACK 6 /* enable local loopback */ #define IPATH_IB_LINK_EXTERNAL 7 /* normal, disable local loopback */ +#define IPATH_IB_LINK_NO_HRTBT 8 /* disable Heartbeat, e.g. for loopback */ +#define IPATH_IB_LINK_HRTBT 9 /* enable heartbeat, normal, non-loopback */ /* * These 3 values (SDR and DDR may be ORed for auto-speed @@ -198,7 +200,8 @@ typedef enum _ipath_ureg { #define IPATH_RUNTIME_FORCE_WC_ORDER 0x4 #define IPATH_RUNTIME_RCVHDR_COPY 0x8 #define IPATH_RUNTIME_MASTER 0x10 -/* 0x20 and 0x40 are no longer used, but are reserved for ABI compatibility */ +#define IPATH_RUNTIME_NODMA_RTAIL 0x80 +#define IPATH_RUNTIME_SDMA 0x200 #define IPATH_RUNTIME_FORCE_PIOAVAIL 0x400 #define IPATH_RUNTIME_PIO_REGSWAPPED 0x800 @@ -444,8 +447,9 @@ struct ipath_user_info { #define IPATH_CMD_PIOAVAILUPD 27 /* force an update of PIOAvail reg */ #define IPATH_CMD_POLL_TYPE 28 /* set the kind of polling we want */ #define IPATH_CMD_ARMLAUNCH_CTRL 29 /* armlaunch detection control */ - -#define IPATH_CMD_MAX 29 +/* 30 is unused */ +#define IPATH_CMD_SDMA_INFLIGHT 31 /* sdma inflight counter request */ +#define IPATH_CMD_SDMA_COMPLETE 32 /* sdma completion counter request */ /* * Poll types @@ -483,6 +487,17 @@ struct ipath_cmd { union { struct ipath_tid_info tid_info; struct ipath_user_info user_info; + + /* + * address in userspace where we should put the sdma + * inflight counter + */ + __u64 sdma_inflight; + /* + * address in userspace where we should put the sdma + * completion counter + */ + __u64 sdma_complete; /* address in userspace of struct ipath_port_info to write result to */ __u64 port_info; @@ -537,7 +552,7 @@ struct ipath_diag_pkt { /* The second diag_pkt struct is the expanded version that allows * more control over the packet, specifically, by allowing a custom - * pbc (+ extra) qword, so that special modes and deliberate + * pbc (+ static rate) qword, so that special modes and deliberate * changes to CRCs can be used. The elements were also re-ordered * for better alignment and to avoid padding issues. */ @@ -662,8 +677,12 @@ struct infinipath_counters { #define INFINIPATH_RHF_LENGTH_SHIFT 0 #define INFINIPATH_RHF_RCVTYPE_MASK 0x7 #define INFINIPATH_RHF_RCVTYPE_SHIFT 11 -#define INFINIPATH_RHF_EGRINDEX_MASK 0x7FF +#define INFINIPATH_RHF_EGRINDEX_MASK 0xFFF #define INFINIPATH_RHF_EGRINDEX_SHIFT 16 +#define INFINIPATH_RHF_SEQ_MASK 0xF +#define INFINIPATH_RHF_SEQ_SHIFT 0 +#define INFINIPATH_RHF_HDRQ_OFFSET_MASK 0x7FF +#define INFINIPATH_RHF_HDRQ_OFFSET_SHIFT 4 #define INFINIPATH_RHF_H_ICRCERR 0x80000000 #define INFINIPATH_RHF_H_VCRCERR 0x40000000 #define INFINIPATH_RHF_H_PARITYERR 0x20000000 @@ -673,6 +692,8 @@ struct infinipath_counters { #define INFINIPATH_RHF_H_TIDERR 0x02000000 #define INFINIPATH_RHF_H_MKERR 0x01000000 #define INFINIPATH_RHF_H_IBERR 0x00800000 +#define INFINIPATH_RHF_H_ERR_MASK 0xFF800000 +#define INFINIPATH_RHF_L_USE_EGR 0x80000000 #define INFINIPATH_RHF_L_SWA 0x00008000 #define INFINIPATH_RHF_L_SWB 0x00004000 @@ -696,6 +717,7 @@ struct infinipath_counters { /* SendPIO per-buffer control */ #define INFINIPATH_SP_TEST 0x40 #define INFINIPATH_SP_TESTEBP 0x20 +#define INFINIPATH_SP_TRIGGER_SHIFT 15 /* SendPIOAvail bits */ #define INFINIPATH_SENDPIOAVAIL_BUSY_SHIFT 1 @@ -762,6 +784,7 @@ struct ether_header { #define IPATH_MSN_MASK 0xFFFFFF #define IPATH_QPN_MASK 0xFFFFFF #define IPATH_MULTICAST_LID_BASE 0xC000 +#define IPATH_EAGER_TID_ID INFINIPATH_I_TID_MASK #define IPATH_MULTICAST_QPN 0xFFFFFF /* Receive Header Queue: receive type (from infinipath) */ @@ -781,7 +804,7 @@ struct ether_header { */ static inline __u32 ipath_hdrget_err_flags(const __le32 * rbuf) { - return __le32_to_cpu(rbuf[1]); + return __le32_to_cpu(rbuf[1]) & INFINIPATH_RHF_H_ERR_MASK; } static inline __u32 ipath_hdrget_rcv_type(const __le32 * rbuf) @@ -802,6 +825,23 @@ static inline __u32 ipath_hdrget_index(const __le32 * rbuf) & INFINIPATH_RHF_EGRINDEX_MASK; } +static inline __u32 ipath_hdrget_seq(const __le32 *rbuf) +{ + return (__le32_to_cpu(rbuf[1]) >> INFINIPATH_RHF_SEQ_SHIFT) + & INFINIPATH_RHF_SEQ_MASK; +} + +static inline __u32 ipath_hdrget_offset(const __le32 *rbuf) +{ + return (__le32_to_cpu(rbuf[1]) >> INFINIPATH_RHF_HDRQ_OFFSET_SHIFT) + & INFINIPATH_RHF_HDRQ_OFFSET_MASK; +} + +static inline __u32 ipath_hdrget_use_egr_buf(const __le32 *rbuf) +{ + return __le32_to_cpu(rbuf[0]) & INFINIPATH_RHF_L_USE_EGR; +} + static inline __u32 ipath_hdrget_ipath_ver(__le32 hdrword) { return (__le32_to_cpu(hdrword) >> INFINIPATH_I_VERS_SHIFT) diff --git a/drivers/infiniband/hw/ipath/ipath_debug.h b/drivers/infiniband/hw/ipath/ipath_debug.h index d6f69532d83f..65926cd35759 100644 --- a/drivers/infiniband/hw/ipath/ipath_debug.h +++ b/drivers/infiniband/hw/ipath/ipath_debug.h @@ -66,6 +66,7 @@ #define __IPATH_IPATHERR 0x40000 /* Ethernet (IPATH) errors */ #define __IPATH_IPATHPD 0x80000 /* Ethernet (IPATH) packet dump */ #define __IPATH_IPATHTABLE 0x100000 /* Ethernet (IPATH) table dump */ +#define __IPATH_LINKVERBDBG 0x200000 /* very verbose linkchange debug */ #else /* _IPATH_DEBUGGING */ @@ -89,6 +90,7 @@ #define __IPATH_IPATHERR 0x0 /* Ethernet (IPATH) errors on */ #define __IPATH_IPATHPD 0x0 /* Ethernet (IPATH) packet dump on */ #define __IPATH_IPATHTABLE 0x0 /* Ethernet (IPATH) packet dump on */ +#define __IPATH_LINKVERBDBG 0x0 /* very verbose linkchange debug */ #endif /* _IPATH_DEBUGGING */ diff --git a/drivers/infiniband/hw/ipath/ipath_diag.c b/drivers/infiniband/hw/ipath/ipath_diag.c index 4137c7770f1b..6d49d2f18a88 100644 --- a/drivers/infiniband/hw/ipath/ipath_diag.c +++ b/drivers/infiniband/hw/ipath/ipath_diag.c @@ -1,5 +1,5 @@ /* - * Copyright (c) 2006, 2007 QLogic Corporation. All rights reserved. + * Copyright (c) 2006, 2007, 2008 QLogic Corporation. All rights reserved. * Copyright (c) 2003, 2004, 2005, 2006 PathScale, Inc. All rights reserved. * * This software is available to you under a choice of one of two @@ -330,13 +330,19 @@ static ssize_t ipath_diagpkt_write(struct file *fp, struct ipath_devdata *dd; ssize_t ret = 0; u64 val; + u32 l_state, lt_state; /* LinkState, LinkTrainingState */ - if (count != sizeof(dp)) { + if (count < sizeof(odp)) { ret = -EINVAL; goto bail; } - if (copy_from_user(&dp, data, sizeof(dp))) { + if (count == sizeof(dp)) { + if (copy_from_user(&dp, data, sizeof(dp))) { + ret = -EFAULT; + goto bail; + } + } else if (copy_from_user(&odp, data, sizeof(odp))) { ret = -EFAULT; goto bail; } @@ -396,10 +402,17 @@ static ssize_t ipath_diagpkt_write(struct file *fp, ret = -ENODEV; goto bail; } - /* Check link state, but not if we have custom PBC */ - val = dd->ipath_lastibcstat & IPATH_IBSTATE_MASK; - if (!dp.pbc_wd && val != IPATH_IBSTATE_INIT && - val != IPATH_IBSTATE_ARM && val != IPATH_IBSTATE_ACTIVE) { + /* + * Want to skip check for l_state if using custom PBC, + * because we might be trying to force an SM packet out. + * first-cut, skip _all_ state checking in that case. + */ + val = ipath_ib_state(dd, dd->ipath_lastibcstat); + lt_state = ipath_ib_linktrstate(dd, dd->ipath_lastibcstat); + l_state = ipath_ib_linkstate(dd, dd->ipath_lastibcstat); + if (!dp.pbc_wd && (lt_state != INFINIPATH_IBCS_LT_STATE_LINKUP || + (val != dd->ib_init && val != dd->ib_arm && + val != dd->ib_active))) { ipath_cdbg(VERBOSE, "unit %u not ready (state %llx)\n", dd->ipath_unit, (unsigned long long) val); ret = -EINVAL; @@ -431,15 +444,17 @@ static ssize_t ipath_diagpkt_write(struct file *fp, goto bail; } - piobuf = ipath_getpiobuf(dd, &pbufn); + plen >>= 2; /* in dwords */ + + piobuf = ipath_getpiobuf(dd, plen, &pbufn); if (!piobuf) { ipath_cdbg(VERBOSE, "No PIO buffers avail unit for %u\n", dd->ipath_unit); ret = -EBUSY; goto bail; } - - plen >>= 2; /* in dwords */ + /* disarm it just to be extra sure */ + ipath_disarm_piobufs(dd, pbufn, 1); if (ipath_debug & __IPATH_PKTDBG) ipath_cdbg(VERBOSE, "unit %u 0x%x+1w pio%d\n", diff --git a/drivers/infiniband/hw/ipath/ipath_driver.c b/drivers/infiniband/hw/ipath/ipath_driver.c index ca4d0acc6786..e0a64f070b97 100644 --- a/drivers/infiniband/hw/ipath/ipath_driver.c +++ b/drivers/infiniband/hw/ipath/ipath_driver.c @@ -1,5 +1,5 @@ /* - * Copyright (c) 2006, 2007 QLogic Corporation. All rights reserved. + * Copyright (c) 2006, 2007, 2008 QLogic Corporation. All rights reserved. * Copyright (c) 2003, 2004, 2005, 2006 PathScale, Inc. All rights reserved. * * This software is available to you under a choice of one of two @@ -41,7 +41,6 @@ #include "ipath_kernel.h" #include "ipath_verbs.h" -#include "ipath_common.h" static void ipath_update_pio_bufs(struct ipath_devdata *); @@ -73,10 +72,27 @@ module_param_named(debug, ipath_debug, uint, S_IWUSR | S_IRUGO); MODULE_PARM_DESC(debug, "mask for debug prints"); EXPORT_SYMBOL_GPL(ipath_debug); +unsigned ipath_mtu4096 = 1; /* max 4KB IB mtu by default, if supported */ +module_param_named(mtu4096, ipath_mtu4096, uint, S_IRUGO); +MODULE_PARM_DESC(mtu4096, "enable MTU of 4096 bytes, if supported"); + +static unsigned ipath_hol_timeout_ms = 13000; +module_param_named(hol_timeout_ms, ipath_hol_timeout_ms, uint, S_IRUGO); +MODULE_PARM_DESC(hol_timeout_ms, + "duration of user app suspension after link failure"); + +unsigned ipath_linkrecovery = 1; +module_param_named(linkrecovery, ipath_linkrecovery, uint, S_IWUSR | S_IRUGO); +MODULE_PARM_DESC(linkrecovery, "enable workaround for link recovery issue"); + MODULE_LICENSE("GPL"); -MODULE_AUTHOR("QLogic <support@pathscale.com>"); +MODULE_AUTHOR("QLogic <support@qlogic.com>"); MODULE_DESCRIPTION("QLogic InfiniPath driver"); +/* + * Table to translate the LINKTRAININGSTATE portion of + * IBCStatus to a human-readable form. + */ const char *ipath_ibcstatus_str[] = { "Disabled", "LinkUp", @@ -91,9 +107,20 @@ const char *ipath_ibcstatus_str[] = { "CfgWaitRmt", "CfgIdle", "RecovRetrain", - "LState0xD", /* unused */ + "CfgTxRevLane", /* unused before IBA7220 */ "RecovWaitRmt", "RecovIdle", + /* below were added for IBA7220 */ + "CfgEnhanced", + "CfgTest", + "CfgWaitRmtTest", + "CfgWaitCfgEnhanced", + "SendTS_T", + "SendTstIdles", + "RcvTS_T", + "SendTst_TS1s", + "LTState18", "LTState19", "LTState1A", "LTState1B", + "LTState1C", "LTState1D", "LTState1E", "LTState1F" }; static void __devexit ipath_remove_one(struct pci_dev *); @@ -102,8 +129,10 @@ static int __devinit ipath_init_one(struct pci_dev *, /* Only needed for registration, nothing else needs this info */ #define PCI_VENDOR_ID_PATHSCALE 0x1fc1 +#define PCI_VENDOR_ID_QLOGIC 0x1077 #define PCI_DEVICE_ID_INFINIPATH_HT 0xd #define PCI_DEVICE_ID_INFINIPATH_PE800 0x10 +#define PCI_DEVICE_ID_INFINIPATH_7220 0x7220 /* Number of seconds before our card status check... */ #define STATUS_TIMEOUT 60 @@ -111,6 +140,7 @@ static int __devinit ipath_init_one(struct pci_dev *, static const struct pci_device_id ipath_pci_tbl[] = { { PCI_DEVICE(PCI_VENDOR_ID_PATHSCALE, PCI_DEVICE_ID_INFINIPATH_HT) }, { PCI_DEVICE(PCI_VENDOR_ID_PATHSCALE, PCI_DEVICE_ID_INFINIPATH_PE800) }, + { PCI_DEVICE(PCI_VENDOR_ID_QLOGIC, PCI_DEVICE_ID_INFINIPATH_7220) }, { 0, } }; @@ -126,19 +156,6 @@ static struct pci_driver ipath_driver = { }, }; -static void ipath_check_status(struct work_struct *work) -{ - struct ipath_devdata *dd = container_of(work, struct ipath_devdata, - status_work.work); - - /* - * If we don't have any interrupts, let the user know and - * don't bother checking again. - */ - if (dd->ipath_int_counter == 0) - dev_err(&dd->pcidev->dev, "No interrupts detected.\n"); -} - static inline void read_bars(struct ipath_devdata *dd, struct pci_dev *dev, u32 *bar0, u32 *bar1) { @@ -206,8 +223,6 @@ static struct ipath_devdata *ipath_alloc_devdata(struct pci_dev *pdev) dd->pcidev = pdev; pci_set_drvdata(pdev, dd); - INIT_DELAYED_WORK(&dd->status_work, ipath_check_status); - list_add(&dd->ipath_list, &ipath_dev_list); bail_unlock: @@ -234,12 +249,12 @@ struct ipath_devdata *ipath_lookup(int unit) return dd; } -int ipath_count_units(int *npresentp, int *nupp, u32 *maxportsp) +int ipath_count_units(int *npresentp, int *nupp, int *maxportsp) { int nunits, npresent, nup; struct ipath_devdata *dd; unsigned long flags; - u32 maxports; + int maxports; nunits = npresent = nup = maxports = 0; @@ -304,7 +319,7 @@ static void ipath_verify_pioperf(struct ipath_devdata *dd) u32 *addr; u64 msecs, emsecs; - piobuf = ipath_getpiobuf(dd, &pbnum); + piobuf = ipath_getpiobuf(dd, 0, &pbnum); if (!piobuf) { dev_info(&dd->pcidev->dev, "No PIObufs for checking perf, skipping\n"); @@ -336,7 +351,14 @@ static void ipath_verify_pioperf(struct ipath_devdata *dd) ipath_disable_armlaunch(dd); - writeq(0, piobuf); /* length 0, no dwords actually sent */ + /* + * length 0, no dwords actually sent, and mark as VL15 + * on chips where that may matter (due to IB flowcontrol) + */ + if ((dd->ipath_flags & IPATH_HAS_PBC_CNT)) + writeq(1UL << 63, piobuf); + else + writeq(0, piobuf); ipath_flush_wc(); /* @@ -377,6 +399,7 @@ static int __devinit ipath_init_one(struct pci_dev *pdev, struct ipath_devdata *dd; unsigned long long addr; u32 bar0 = 0, bar1 = 0; + u8 rev; dd = ipath_alloc_devdata(pdev); if (IS_ERR(dd)) { @@ -408,7 +431,7 @@ static int __devinit ipath_init_one(struct pci_dev *pdev, } addr = pci_resource_start(pdev, 0); len = pci_resource_len(pdev, 0); - ipath_cdbg(VERBOSE, "regbase (0) %llx len %d pdev->irq %d, vend %x/%x " + ipath_cdbg(VERBOSE, "regbase (0) %llx len %d irq %d, vend %x/%x " "driver_data %lx\n", addr, len, pdev->irq, ent->vendor, ent->device, ent->driver_data); @@ -512,6 +535,13 @@ static int __devinit ipath_init_one(struct pci_dev *pdev, "CONFIG_PCI_MSI is not enabled\n", ent->device); return -ENODEV; #endif + case PCI_DEVICE_ID_INFINIPATH_7220: +#ifndef CONFIG_PCI_MSI + ipath_dbg("CONFIG_PCI_MSI is not enabled, " + "using IntX for unit %u\n", dd->ipath_unit); +#endif + ipath_init_iba7220_funcs(dd); + break; default: ipath_dev_err(dd, "Found unknown QLogic deviceid 0x%x, " "failing\n", ent->device); @@ -533,7 +563,13 @@ static int __devinit ipath_init_one(struct pci_dev *pdev, goto bail_regions; } - dd->ipath_pcirev = pdev->revision; + ret = pci_read_config_byte(pdev, PCI_REVISION_ID, &rev); + if (ret) { + ipath_dev_err(dd, "Failed to read PCI revision ID unit " + "%u: err %d\n", dd->ipath_unit, -ret); + goto bail_regions; /* shouldn't ever happen */ + } + dd->ipath_pcirev = rev; #if defined(__powerpc__) /* There isn't a generic way to specify writethrough mappings */ @@ -556,14 +592,6 @@ static int __devinit ipath_init_one(struct pci_dev *pdev, ipath_cdbg(VERBOSE, "mapped io addr %llx to kregbase %p\n", addr, dd->ipath_kregbase); - /* - * clear ipath_flags here instead of in ipath_init_chip as it is set - * by ipath_setup_htconfig. - */ - dd->ipath_flags = 0; - dd->ipath_lli_counter = 0; - dd->ipath_lli_errors = 0; - if (dd->ipath_f_bus(dd, pdev)) ipath_dev_err(dd, "Failed to setup config space; " "continuing anyway\n"); @@ -608,13 +636,11 @@ static int __devinit ipath_init_one(struct pci_dev *pdev, ipath_diag_add(dd); ipath_register_ib_device(dd); - /* Check that card status in STATUS_TIMEOUT seconds. */ - schedule_delayed_work(&dd->status_work, HZ * STATUS_TIMEOUT); - goto bail; bail_irqsetup: - if (pdev->irq) free_irq(pdev->irq, dd); + if (pdev->irq) + free_irq(pdev->irq, dd); bail_iounmap: iounmap((volatile void __iomem *) dd->ipath_kregbase); @@ -654,6 +680,10 @@ static void __devexit cleanup_device(struct ipath_devdata *dd) ipath_disable_wc(dd); } + if (dd->ipath_spectriggerhit) + dev_info(&dd->pcidev->dev, "%lu special trigger hits\n", + dd->ipath_spectriggerhit); + if (dd->ipath_pioavailregs_dma) { dma_free_coherent(&dd->pcidev->dev, PAGE_SIZE, (void *) dd->ipath_pioavailregs_dma, @@ -706,6 +736,8 @@ static void __devexit cleanup_device(struct ipath_devdata *dd) tmpp = dd->ipath_pageshadow; dd->ipath_pageshadow = NULL; vfree(tmpp); + + dd->ipath_egrtidbase = NULL; } /* @@ -738,7 +770,6 @@ static void __devexit ipath_remove_one(struct pci_dev *pdev) */ ipath_shutdown_device(dd); - cancel_delayed_work(&dd->status_work); flush_scheduled_work(); if (dd->verbs_dev) @@ -823,20 +854,8 @@ void ipath_disarm_piobufs(struct ipath_devdata *dd, unsigned first, ipath_read_kreg64(dd, dd->ipath_kregs->kr_scratch); spin_unlock_irqrestore(&dd->ipath_sendctrl_lock, flags); } - - /* - * Disable PIOAVAILUPD, then re-enable, reading scratch in - * between. This seems to avoid a chip timing race that causes - * pioavail updates to memory to stop. We xor as we don't - * know the state of the bit when we're called. - */ - spin_lock_irqsave(&dd->ipath_sendctrl_lock, flags); - ipath_write_kreg(dd, dd->ipath_kregs->kr_sendctrl, - dd->ipath_sendctrl ^ INFINIPATH_S_PIOBUFAVAILUPD); - ipath_read_kreg64(dd, dd->ipath_kregs->kr_scratch); - ipath_write_kreg(dd, dd->ipath_kregs->kr_sendctrl, - dd->ipath_sendctrl); - spin_unlock_irqrestore(&dd->ipath_sendctrl_lock, flags); + /* on some older chips, update may not happen after cancel */ + ipath_force_pio_avail_update(dd); } /** @@ -873,18 +892,52 @@ int ipath_wait_linkstate(struct ipath_devdata *dd, u32 state, int msecs) (unsigned long long) ipath_read_kreg64( dd, dd->ipath_kregs->kr_ibcctrl), (unsigned long long) val, - ipath_ibcstatus_str[val & 0xf]); + ipath_ibcstatus_str[val & dd->ibcs_lts_mask]); } return (dd->ipath_flags & state) ? 0 : -ETIMEDOUT; } +static void decode_sdma_errs(struct ipath_devdata *dd, ipath_err_t err, + char *buf, size_t blen) +{ + static const struct { + ipath_err_t err; + const char *msg; + } errs[] = { + { INFINIPATH_E_SDMAGENMISMATCH, "SDmaGenMismatch" }, + { INFINIPATH_E_SDMAOUTOFBOUND, "SDmaOutOfBound" }, + { INFINIPATH_E_SDMATAILOUTOFBOUND, "SDmaTailOutOfBound" }, + { INFINIPATH_E_SDMABASE, "SDmaBase" }, + { INFINIPATH_E_SDMA1STDESC, "SDma1stDesc" }, + { INFINIPATH_E_SDMARPYTAG, "SDmaRpyTag" }, + { INFINIPATH_E_SDMADWEN, "SDmaDwEn" }, + { INFINIPATH_E_SDMAMISSINGDW, "SDmaMissingDw" }, + { INFINIPATH_E_SDMAUNEXPDATA, "SDmaUnexpData" }, + { INFINIPATH_E_SDMADESCADDRMISALIGN, "SDmaDescAddrMisalign" }, + { INFINIPATH_E_SENDBUFMISUSE, "SendBufMisuse" }, + { INFINIPATH_E_SDMADISABLED, "SDmaDisabled" }, + }; + int i; + int expected; + size_t bidx = 0; + + for (i = 0; i < ARRAY_SIZE(errs); i++) { + expected = (errs[i].err != INFINIPATH_E_SDMADISABLED) ? 0 : + test_bit(IPATH_SDMA_ABORTING, &dd->ipath_sdma_status); + if ((err & errs[i].err) && !expected) + bidx += snprintf(buf + bidx, blen - bidx, + "%s ", errs[i].msg); + } +} + /* * Decode the error status into strings, deciding whether to always * print * it or not depending on "normal packet errors" vs everything * else. Return 1 if "real" errors, otherwise 0 if only packet * errors, so caller can decide what to print with the string. */ -int ipath_decode_err(char *buf, size_t blen, ipath_err_t err) +int ipath_decode_err(struct ipath_devdata *dd, char *buf, size_t blen, + ipath_err_t err) { int iserr = 1; *buf = '\0'; @@ -922,6 +975,8 @@ int ipath_decode_err(char *buf, size_t blen, ipath_err_t err) strlcat(buf, "rbadversion ", blen); if (err & INFINIPATH_E_RHDR) strlcat(buf, "rhdr ", blen); + if (err & INFINIPATH_E_SENDSPECIALTRIGGER) + strlcat(buf, "sendspecialtrigger ", blen); if (err & INFINIPATH_E_RLONGPKTLEN) strlcat(buf, "rlongpktlen ", blen); if (err & INFINIPATH_E_RMAXPKTLEN) @@ -964,6 +1019,10 @@ int ipath_decode_err(char *buf, size_t blen, ipath_err_t err) strlcat(buf, "hardware ", blen); if (err & INFINIPATH_E_RESET) strlcat(buf, "reset ", blen); + if (err & INFINIPATH_E_SDMAERRS) + decode_sdma_errs(dd, err, buf, blen); + if (err & INFINIPATH_E_INVALIDEEPCMD) + strlcat(buf, "invalideepromcmd ", blen); done: return iserr; } @@ -1076,18 +1135,17 @@ static void ipath_rcv_hdrerr(struct ipath_devdata *dd, u32 eflags, u32 l, u32 etail, - u64 *rc) + __le32 *rhf_addr, + struct ipath_message_header *hdr) { char emsg[128]; - struct ipath_message_header *hdr; get_rhf_errstring(eflags, emsg, sizeof emsg); - hdr = (struct ipath_message_header *)&rc[1]; ipath_cdbg(PKT, "RHFerrs %x hdrqtail=%x typ=%u " "tlen=%x opcode=%x egridx=%x: %s\n", eflags, l, - ipath_hdrget_rcv_type((__le32 *) rc), - ipath_hdrget_length_in_bytes((__le32 *) rc), + ipath_hdrget_rcv_type(rhf_addr), + ipath_hdrget_length_in_bytes(rhf_addr), be32_to_cpu(hdr->bth[0]) >> 24, etail, emsg); @@ -1112,55 +1170,52 @@ static void ipath_rcv_hdrerr(struct ipath_devdata *dd, */ void ipath_kreceive(struct ipath_portdata *pd) { - u64 *rc; struct ipath_devdata *dd = pd->port_dd; + __le32 *rhf_addr; void *ebuf; const u32 rsize = dd->ipath_rcvhdrentsize; /* words */ const u32 maxcnt = dd->ipath_rcvhdrcnt * rsize; /* words */ u32 etail = -1, l, hdrqtail; struct ipath_message_header *hdr; - u32 eflags, i, etype, tlen, pkttot = 0, updegr=0, reloop=0; + u32 eflags, i, etype, tlen, pkttot = 0, updegr = 0, reloop = 0; static u64 totcalls; /* stats, may eventually remove */ - - if (!dd->ipath_hdrqtailptr) { - ipath_dev_err(dd, - "hdrqtailptr not set, can't do receives\n"); - goto bail; - } + int last; l = pd->port_head; - hdrqtail = ipath_get_rcvhdrtail(pd); - if (l == hdrqtail) - goto bail; - -reloop: - for (i = 0; l != hdrqtail; i++) { - u32 qp; - u8 *bthbytes; + rhf_addr = (__le32 *) pd->port_rcvhdrq + l + dd->ipath_rhf_offset; + if (dd->ipath_flags & IPATH_NODMA_RTAIL) { + u32 seq = ipath_hdrget_seq(rhf_addr); - rc = (u64 *) (pd->port_rcvhdrq + (l << 2)); - hdr = (struct ipath_message_header *)&rc[1]; - /* - * could make a network order version of IPATH_KD_QP, and - * do the obvious shift before masking to speed this up. - */ - qp = ntohl(hdr->bth[1]) & 0xffffff; - bthbytes = (u8 *) hdr->bth; + if (seq != pd->port_seq_cnt) + goto bail; + hdrqtail = 0; + } else { + hdrqtail = ipath_get_rcvhdrtail(pd); + if (l == hdrqtail) + goto bail; + smp_rmb(); + } - eflags = ipath_hdrget_err_flags((__le32 *) rc); - etype = ipath_hdrget_rcv_type((__le32 *) rc); +reloop: + for (last = 0, i = 1; !last; i++) { + hdr = dd->ipath_f_get_msgheader(dd, rhf_addr); + eflags = ipath_hdrget_err_flags(rhf_addr); + etype = ipath_hdrget_rcv_type(rhf_addr); /* total length */ - tlen = ipath_hdrget_length_in_bytes((__le32 *) rc); + tlen = ipath_hdrget_length_in_bytes(rhf_addr); ebuf = NULL; - if (etype != RCVHQ_RCV_TYPE_EXPECTED) { + if ((dd->ipath_flags & IPATH_NODMA_RTAIL) ? + ipath_hdrget_use_egr_buf(rhf_addr) : + (etype != RCVHQ_RCV_TYPE_EXPECTED)) { /* - * it turns out that the chips uses an eager buffer + * It turns out that the chip uses an eager buffer * for all non-expected packets, whether it "needs" * one or not. So always get the index, but don't * set ebuf (so we try to copy data) unless the * length requires it. */ - etail = ipath_hdrget_index((__le32 *) rc); + etail = ipath_hdrget_index(rhf_addr); + updegr = 1; if (tlen > sizeof(*hdr) || etype == RCVHQ_RCV_TYPE_NON_KD) ebuf = ipath_get_egrbuf(dd, etail); @@ -1171,75 +1226,91 @@ reloop: * packets; only ipathhdrerr should be set. */ - if (etype != RCVHQ_RCV_TYPE_NON_KD && etype != - RCVHQ_RCV_TYPE_ERROR && ipath_hdrget_ipath_ver( - hdr->iph.ver_port_tid_offset) != - IPS_PROTO_VERSION) { + if (etype != RCVHQ_RCV_TYPE_NON_KD && + etype != RCVHQ_RCV_TYPE_ERROR && + ipath_hdrget_ipath_ver(hdr->iph.ver_port_tid_offset) != + IPS_PROTO_VERSION) ipath_cdbg(PKT, "Bad InfiniPath protocol version " "%x\n", etype); - } if (unlikely(eflags)) - ipath_rcv_hdrerr(dd, eflags, l, etail, rc); + ipath_rcv_hdrerr(dd, eflags, l, etail, rhf_addr, hdr); else if (etype == RCVHQ_RCV_TYPE_NON_KD) { - ipath_ib_rcv(dd->verbs_dev, rc + 1, ebuf, tlen); + ipath_ib_rcv(dd->verbs_dev, (u32 *)hdr, ebuf, tlen); if (dd->ipath_lli_counter) dd->ipath_lli_counter--; + } else if (etype == RCVHQ_RCV_TYPE_EAGER) { + u8 opcode = be32_to_cpu(hdr->bth[0]) >> 24; + u32 qp = be32_to_cpu(hdr->bth[1]) & 0xffffff; ipath_cdbg(PKT, "typ %x, opcode %x (eager, " "qp=%x), len %x; ignored\n", - etype, bthbytes[0], qp, tlen); + etype, opcode, qp, tlen); } - else if (etype == RCVHQ_RCV_TYPE_EAGER) - ipath_cdbg(PKT, "typ %x, opcode %x (eager, " - "qp=%x), len %x; ignored\n", - etype, bthbytes[0], qp, tlen); else if (etype == RCVHQ_RCV_TYPE_EXPECTED) ipath_dbg("Bug: Expected TID, opcode %x; ignored\n", - be32_to_cpu(hdr->bth[0]) & 0xff); + be32_to_cpu(hdr->bth[0]) >> 24); else { /* * error packet, type of error unknown. * Probably type 3, but we don't know, so don't * even try to print the opcode, etc. + * Usually caused by a "bad packet", that has no + * BTH, when the LRH says it should. */ - ipath_dbg("Error Pkt, but no eflags! egrbuf %x, " - "len %x\nhdrq@%lx;hdrq+%x rhf: %llx; " - "hdr %llx %llx %llx %llx %llx\n", - etail, tlen, (unsigned long) rc, l, - (unsigned long long) rc[0], - (unsigned long long) rc[1], - (unsigned long long) rc[2], - (unsigned long long) rc[3], - (unsigned long long) rc[4], - (unsigned long long) rc[5]); + ipath_cdbg(ERRPKT, "Error Pkt, but no eflags! egrbuf" + " %x, len %x hdrq+%x rhf: %Lx\n", + etail, tlen, l, + le64_to_cpu(*(__le64 *) rhf_addr)); + if (ipath_debug & __IPATH_ERRPKTDBG) { + u32 j, *d, dw = rsize-2; + if (rsize > (tlen>>2)) + dw = tlen>>2; + d = (u32 *)hdr; + printk(KERN_DEBUG "EPkt rcvhdr(%x dw):\n", + dw); + for (j = 0; j < dw; j++) + printk(KERN_DEBUG "%8x%s", d[j], + (j%8) == 7 ? "\n" : " "); + printk(KERN_DEBUG ".\n"); + } } l += rsize; if (l >= maxcnt) l = 0; - if (etype != RCVHQ_RCV_TYPE_EXPECTED) - updegr = 1; + rhf_addr = (__le32 *) pd->port_rcvhdrq + + l + dd->ipath_rhf_offset; + if (dd->ipath_flags & IPATH_NODMA_RTAIL) { + u32 seq = ipath_hdrget_seq(rhf_addr); + + if (++pd->port_seq_cnt > 13) + pd->port_seq_cnt = 1; + if (seq != pd->port_seq_cnt) + last = 1; + } else if (l == hdrqtail) + last = 1; /* * update head regs on last packet, and every 16 packets. * Reduce bus traffic, while still trying to prevent * rcvhdrq overflows, for when the queue is nearly full */ - if (l == hdrqtail || (i && !(i&0xf))) { - u64 lval; - if (l == hdrqtail) - /* request IBA6120 interrupt only on last */ - lval = dd->ipath_rhdrhead_intr_off | l; - else - lval = l; - (void)ipath_write_ureg(dd, ur_rcvhdrhead, lval, 0); + if (last || !(i & 0xf)) { + u64 lval = l; + + /* request IBA6120 and 7220 interrupt only on last */ + if (last) + lval |= dd->ipath_rhdrhead_intr_off; + ipath_write_ureg(dd, ur_rcvhdrhead, lval, + pd->port_port); if (updegr) { - (void)ipath_write_ureg(dd, ur_rcvegrindexhead, - etail, 0); + ipath_write_ureg(dd, ur_rcvegrindexhead, + etail, pd->port_port); updegr = 0; } } } - if (!dd->ipath_rhdrhead_intr_off && !reloop) { + if (!dd->ipath_rhdrhead_intr_off && !reloop && + !(dd->ipath_flags & IPATH_NODMA_RTAIL)) { /* IBA6110 workaround; we can have a race clearing chip * interrupt with another interrupt about to be delivered, * and can clear it before it is delivered on the GPIO @@ -1301,7 +1372,6 @@ static void ipath_update_pio_bufs(struct ipath_devdata *dd) * happens when all buffers are in use, so only cpu overhead, not * latency or bandwidth is affected. */ -#define _IPATH_ALL_CHECKBITS 0x5555555555555555ULL if (!dd->ipath_pioavailregs_dma) { ipath_dbg("Update shadow pioavail, but regs_dma NULL!\n"); return; @@ -1346,7 +1416,7 @@ static void ipath_update_pio_bufs(struct ipath_devdata *dd) piov = le64_to_cpu(dd->ipath_pioavailregs_dma[i ^ 1]); else piov = le64_to_cpu(dd->ipath_pioavailregs_dma[i]); - pchg = _IPATH_ALL_CHECKBITS & + pchg = dd->ipath_pioavailkernel[i] & ~(dd->ipath_pioavailshadow[i] ^ piov); pchbusy = pchg << INFINIPATH_SENDPIOAVAIL_BUSY_SHIFT; if (pchg && (pchbusy & dd->ipath_pioavailshadow[i])) { @@ -1397,27 +1467,63 @@ int ipath_setrcvhdrsize(struct ipath_devdata *dd, unsigned rhdrsize) return ret; } -/** - * ipath_getpiobuf - find an available pio buffer - * @dd: the infinipath device - * @pbufnum: the buffer number is placed here +/* + * debugging code and stats updates if no pio buffers available. + */ +static noinline void no_pio_bufs(struct ipath_devdata *dd) +{ + unsigned long *shadow = dd->ipath_pioavailshadow; + __le64 *dma = (__le64 *)dd->ipath_pioavailregs_dma; + + dd->ipath_upd_pio_shadow = 1; + + /* + * not atomic, but if we lose a stat count in a while, that's OK + */ + ipath_stats.sps_nopiobufs++; + if (!(++dd->ipath_consec_nopiobuf % 100000)) { + ipath_dbg("%u pio sends with no bufavail; dmacopy: " + "%llx %llx %llx %llx; shadow: %lx %lx %lx %lx\n", + dd->ipath_consec_nopiobuf, + (unsigned long long) le64_to_cpu(dma[0]), + (unsigned long long) le64_to_cpu(dma[1]), + (unsigned long long) le64_to_cpu(dma[2]), + (unsigned long long) le64_to_cpu(dma[3]), + shadow[0], shadow[1], shadow[2], shadow[3]); + /* + * 4 buffers per byte, 4 registers above, cover rest + * below + */ + if ((dd->ipath_piobcnt2k + dd->ipath_piobcnt4k) > + (sizeof(shadow[0]) * 4 * 4)) + ipath_dbg("2nd group: dmacopy: %llx %llx " + "%llx %llx; shadow: %lx %lx %lx %lx\n", + (unsigned long long)le64_to_cpu(dma[4]), + (unsigned long long)le64_to_cpu(dma[5]), + (unsigned long long)le64_to_cpu(dma[6]), + (unsigned long long)le64_to_cpu(dma[7]), + shadow[4], shadow[5], shadow[6], + shadow[7]); + } +} + +/* + * common code for normal driver pio buffer allocation, and reserved + * allocation. * * do appropriate marking as busy, etc. * returns buffer number if one found (>=0), negative number is error. - * Used by ipath_layer_send */ -u32 __iomem *ipath_getpiobuf(struct ipath_devdata *dd, u32 * pbufnum) +static u32 __iomem *ipath_getpiobuf_range(struct ipath_devdata *dd, + u32 *pbufnum, u32 first, u32 last, u32 firsti) { - int i, j, starti, updated = 0; - unsigned piobcnt, iter; + int i, j, updated = 0; + unsigned piobcnt; unsigned long flags; unsigned long *shadow = dd->ipath_pioavailshadow; u32 __iomem *buf; - piobcnt = (unsigned)(dd->ipath_piobcnt2k - + dd->ipath_piobcnt4k); - starti = dd->ipath_lastport_piobuf; - iter = piobcnt - starti; + piobcnt = last - first; if (dd->ipath_upd_pio_shadow) { /* * Minor optimization. If we had no buffers on last call, @@ -1425,12 +1531,10 @@ u32 __iomem *ipath_getpiobuf(struct ipath_devdata *dd, u32 * pbufnum) * if no buffers were updated, to be paranoid */ ipath_update_pio_bufs(dd); - /* we scanned here, don't do it at end of scan */ - updated = 1; - i = starti; + updated++; + i = first; } else - i = dd->ipath_lastpioindex; - + i = firsti; rescan: /* * while test_and_set_bit() is atomic, we do that and then the @@ -1438,104 +1542,141 @@ rescan: * of the remaining armlaunch errors. */ spin_lock_irqsave(&ipath_pioavail_lock, flags); - for (j = 0; j < iter; j++, i++) { - if (i >= piobcnt) - i = starti; - /* - * To avoid bus lock overhead, we first find a candidate - * buffer, then do the test and set, and continue if that - * fails. - */ - if (test_bit((2 * i) + 1, shadow) || - test_and_set_bit((2 * i) + 1, shadow)) + for (j = 0; j < piobcnt; j++, i++) { + if (i >= last) + i = first; + if (__test_and_set_bit((2 * i) + 1, shadow)) continue; /* flip generation bit */ - change_bit(2 * i, shadow); + __change_bit(2 * i, shadow); break; } spin_unlock_irqrestore(&ipath_pioavail_lock, flags); - if (j == iter) { - volatile __le64 *dma = dd->ipath_pioavailregs_dma; - - /* - * first time through; shadow exhausted, but may be real - * buffers available, so go see; if any updated, rescan - * (once) - */ + if (j == piobcnt) { if (!updated) { + /* + * first time through; shadow exhausted, but may be + * buffers available, try an update and then rescan. + */ ipath_update_pio_bufs(dd); - updated = 1; - i = starti; + updated++; + i = first; goto rescan; - } - dd->ipath_upd_pio_shadow = 1; - /* - * not atomic, but if we lose one once in a while, that's OK - */ - ipath_stats.sps_nopiobufs++; - if (!(++dd->ipath_consec_nopiobuf % 100000)) { - ipath_dbg( - "%u pio sends with no bufavail; dmacopy: " - "%llx %llx %llx %llx; shadow: " - "%lx %lx %lx %lx\n", - dd->ipath_consec_nopiobuf, - (unsigned long long) le64_to_cpu(dma[0]), - (unsigned long long) le64_to_cpu(dma[1]), - (unsigned long long) le64_to_cpu(dma[2]), - (unsigned long long) le64_to_cpu(dma[3]), - shadow[0], shadow[1], shadow[2], - shadow[3]); + } else if (updated == 1 && piobcnt <= + ((dd->ipath_sendctrl + >> INFINIPATH_S_UPDTHRESH_SHIFT) & + INFINIPATH_S_UPDTHRESH_MASK)) { /* - * 4 buffers per byte, 4 registers above, cover rest - * below + * for chips supporting and using the update + * threshold we need to force an update of the + * in-memory copy if the count is less than the + * thershold, then check one more time. */ - if ((dd->ipath_piobcnt2k + dd->ipath_piobcnt4k) > - (sizeof(shadow[0]) * 4 * 4)) - ipath_dbg("2nd group: dmacopy: %llx %llx " - "%llx %llx; shadow: %lx %lx " - "%lx %lx\n", - (unsigned long long) - le64_to_cpu(dma[4]), - (unsigned long long) - le64_to_cpu(dma[5]), - (unsigned long long) - le64_to_cpu(dma[6]), - (unsigned long long) - le64_to_cpu(dma[7]), - shadow[4], shadow[5], - shadow[6], shadow[7]); + ipath_force_pio_avail_update(dd); + ipath_update_pio_bufs(dd); + updated++; + i = first; + goto rescan; } + + no_pio_bufs(dd); buf = NULL; - goto bail; + } else { + if (i < dd->ipath_piobcnt2k) + buf = (u32 __iomem *) (dd->ipath_pio2kbase + + i * dd->ipath_palign); + else + buf = (u32 __iomem *) + (dd->ipath_pio4kbase + + (i - dd->ipath_piobcnt2k) * dd->ipath_4kalign); + if (pbufnum) + *pbufnum = i; } - /* - * set next starting place. Since it's just an optimization, - * it doesn't matter who wins on this, so no locking - */ - dd->ipath_lastpioindex = i + 1; - if (dd->ipath_upd_pio_shadow) - dd->ipath_upd_pio_shadow = 0; - if (dd->ipath_consec_nopiobuf) - dd->ipath_consec_nopiobuf = 0; - if (i < dd->ipath_piobcnt2k) - buf = (u32 __iomem *) (dd->ipath_pio2kbase + - i * dd->ipath_palign); - else - buf = (u32 __iomem *) - (dd->ipath_pio4kbase + - (i - dd->ipath_piobcnt2k) * dd->ipath_4kalign); - ipath_cdbg(VERBOSE, "Return piobuf%u %uk @ %p\n", - i, (i < dd->ipath_piobcnt2k) ? 2 : 4, buf); - if (pbufnum) - *pbufnum = i; + return buf; +} -bail: +/** + * ipath_getpiobuf - find an available pio buffer + * @dd: the infinipath device + * @plen: the size of the PIO buffer needed in 32-bit words + * @pbufnum: the buffer number is placed here + */ +u32 __iomem *ipath_getpiobuf(struct ipath_devdata *dd, u32 plen, u32 *pbufnum) +{ + u32 __iomem *buf; + u32 pnum, nbufs; + u32 first, lasti; + + if (plen + 1 >= IPATH_SMALLBUF_DWORDS) { + first = dd->ipath_piobcnt2k; + lasti = dd->ipath_lastpioindexl; + } else { + first = 0; + lasti = dd->ipath_lastpioindex; + } + nbufs = dd->ipath_piobcnt2k + dd->ipath_piobcnt4k; + buf = ipath_getpiobuf_range(dd, &pnum, first, nbufs, lasti); + + if (buf) { + /* + * Set next starting place. It's just an optimization, + * it doesn't matter who wins on this, so no locking + */ + if (plen + 1 >= IPATH_SMALLBUF_DWORDS) + dd->ipath_lastpioindexl = pnum + 1; + else + dd->ipath_lastpioindex = pnum + 1; + if (dd->ipath_upd_pio_shadow) + dd->ipath_upd_pio_shadow = 0; + if (dd->ipath_consec_nopiobuf) + dd->ipath_consec_nopiobuf = 0; + ipath_cdbg(VERBOSE, "Return piobuf%u %uk @ %p\n", + pnum, (pnum < dd->ipath_piobcnt2k) ? 2 : 4, buf); + if (pbufnum) + *pbufnum = pnum; + + } return buf; } /** + * ipath_chg_pioavailkernel - change which send buffers are available for kernel + * @dd: the infinipath device + * @start: the starting send buffer number + * @len: the number of send buffers + * @avail: true if the buffers are available for kernel use, false otherwise + */ +void ipath_chg_pioavailkernel(struct ipath_devdata *dd, unsigned start, + unsigned len, int avail) +{ + unsigned long flags; + unsigned end; + + /* There are two bits per send buffer (busy and generation) */ + start *= 2; + len *= 2; + end = start + len; + + /* Set or clear the generation bits. */ + spin_lock_irqsave(&ipath_pioavail_lock, flags); + while (start < end) { + if (avail) { + __clear_bit(start + INFINIPATH_SENDPIOAVAIL_BUSY_SHIFT, + dd->ipath_pioavailshadow); + __set_bit(start, dd->ipath_pioavailkernel); + } else { + __set_bit(start + INFINIPATH_SENDPIOAVAIL_BUSY_SHIFT, + dd->ipath_pioavailshadow); + __clear_bit(start, dd->ipath_pioavailkernel); + } + start += 2; + } + spin_unlock_irqrestore(&ipath_pioavail_lock, flags); +} + +/** * ipath_create_rcvhdrq - create a receive header queue * @dd: the infinipath device * @pd: the port data @@ -1566,19 +1707,27 @@ int ipath_create_rcvhdrq(struct ipath_devdata *dd, ret = -ENOMEM; goto bail; } - pd->port_rcvhdrtail_kvaddr = dma_alloc_coherent( - &dd->pcidev->dev, PAGE_SIZE, &phys_hdrqtail, GFP_KERNEL); - if (!pd->port_rcvhdrtail_kvaddr) { - ipath_dev_err(dd, "attempt to allocate 1 page " - "for port %u rcvhdrqtailaddr failed\n", - pd->port_port); - ret = -ENOMEM; - dma_free_coherent(&dd->pcidev->dev, amt, - pd->port_rcvhdrq, pd->port_rcvhdrq_phys); - pd->port_rcvhdrq = NULL; - goto bail; + + if (!(dd->ipath_flags & IPATH_NODMA_RTAIL)) { + pd->port_rcvhdrtail_kvaddr = dma_alloc_coherent( + &dd->pcidev->dev, PAGE_SIZE, &phys_hdrqtail, + GFP_KERNEL); + if (!pd->port_rcvhdrtail_kvaddr) { + ipath_dev_err(dd, "attempt to allocate 1 page " + "for port %u rcvhdrqtailaddr " + "failed\n", pd->port_port); + ret = -ENOMEM; + dma_free_coherent(&dd->pcidev->dev, amt, + pd->port_rcvhdrq, + pd->port_rcvhdrq_phys); + pd->port_rcvhdrq = NULL; + goto bail; + } + pd->port_rcvhdrqtailaddr_phys = phys_hdrqtail; + ipath_cdbg(VERBOSE, "port %d hdrtailaddr, %llx " + "physical\n", pd->port_port, + (unsigned long long) phys_hdrqtail); } - pd->port_rcvhdrqtailaddr_phys = phys_hdrqtail; pd->port_rcvhdrq_size = amt; @@ -1588,10 +1737,6 @@ int ipath_create_rcvhdrq(struct ipath_devdata *dd, (unsigned long) pd->port_rcvhdrq_phys, (unsigned long) pd->port_rcvhdrq_size, pd->port_port); - - ipath_cdbg(VERBOSE, "port %d hdrtailaddr, %llx physical\n", - pd->port_port, - (unsigned long long) phys_hdrqtail); } else ipath_cdbg(VERBOSE, "reuse port %d rcvhdrq @%p %llx phys; " @@ -1615,7 +1760,6 @@ int ipath_create_rcvhdrq(struct ipath_devdata *dd, ipath_write_kreg_port(dd, dd->ipath_kregs->kr_rcvhdraddr, pd->port_port, pd->port_rcvhdrq_phys); - ret = 0; bail: return ret; } @@ -1632,52 +1776,149 @@ bail: */ void ipath_cancel_sends(struct ipath_devdata *dd, int restore_sendctrl) { + unsigned long flags; + + if (dd->ipath_flags & IPATH_IB_AUTONEG_INPROG) { + ipath_cdbg(VERBOSE, "Ignore while in autonegotiation\n"); + goto bail; + } + /* + * If we have SDMA, and it's not disabled, we have to kick off the + * abort state machine, provided we aren't already aborting. + * If we are in the process of aborting SDMA (!DISABLED, but ABORTING), + * we skip the rest of this routine. It is already "in progress" + */ + if (dd->ipath_flags & IPATH_HAS_SEND_DMA) { + int skip_cancel; + u64 *statp = &dd->ipath_sdma_status; + + spin_lock_irqsave(&dd->ipath_sdma_lock, flags); + skip_cancel = + !test_bit(IPATH_SDMA_DISABLED, statp) && + test_and_set_bit(IPATH_SDMA_ABORTING, statp); + spin_unlock_irqrestore(&dd->ipath_sdma_lock, flags); + if (skip_cancel) + goto bail; + } + ipath_dbg("Cancelling all in-progress send buffers\n"); - dd->ipath_lastcancel = jiffies+HZ/2; /* skip armlaunch errs a bit */ + + /* skip armlaunch errs for a while */ + dd->ipath_lastcancel = jiffies + HZ / 2; + /* - * the abort bit is auto-clearing. We read scratch to be sure - * that cancels and the abort have taken effect in the chip. + * The abort bit is auto-clearing. We also don't want pioavail + * update happening during this, and we don't want any other + * sends going out, so turn those off for the duration. We read + * the scratch register to be sure that cancels and the abort + * have taken effect in the chip. Otherwise two parts are same + * as ipath_force_pio_avail_update() */ + spin_lock_irqsave(&dd->ipath_sendctrl_lock, flags); + dd->ipath_sendctrl &= ~(INFINIPATH_S_PIOBUFAVAILUPD + | INFINIPATH_S_PIOENABLE); ipath_write_kreg(dd, dd->ipath_kregs->kr_sendctrl, - INFINIPATH_S_ABORT); + dd->ipath_sendctrl | INFINIPATH_S_ABORT); ipath_read_kreg64(dd, dd->ipath_kregs->kr_scratch); + spin_unlock_irqrestore(&dd->ipath_sendctrl_lock, flags); + + /* disarm all send buffers */ ipath_disarm_piobufs(dd, 0, - (unsigned)(dd->ipath_piobcnt2k + dd->ipath_piobcnt4k)); - if (restore_sendctrl) /* else done by caller later */ + dd->ipath_piobcnt2k + dd->ipath_piobcnt4k); + + if (restore_sendctrl) { + /* else done by caller later if needed */ + spin_lock_irqsave(&dd->ipath_sendctrl_lock, flags); + dd->ipath_sendctrl |= INFINIPATH_S_PIOBUFAVAILUPD | + INFINIPATH_S_PIOENABLE; ipath_write_kreg(dd, dd->ipath_kregs->kr_sendctrl, - dd->ipath_sendctrl); + dd->ipath_sendctrl); + /* and again, be sure all have hit the chip */ + ipath_read_kreg64(dd, dd->ipath_kregs->kr_scratch); + spin_unlock_irqrestore(&dd->ipath_sendctrl_lock, flags); + } - /* and again, be sure all have hit the chip */ - ipath_read_kreg64(dd, dd->ipath_kregs->kr_scratch); + if ((dd->ipath_flags & IPATH_HAS_SEND_DMA) && + !test_bit(IPATH_SDMA_DISABLED, &dd->ipath_sdma_status) && + test_bit(IPATH_SDMA_RUNNING, &dd->ipath_sdma_status)) { + spin_lock_irqsave(&dd->ipath_sdma_lock, flags); + /* only wait so long for intr */ + dd->ipath_sdma_abort_intr_timeout = jiffies + HZ; + dd->ipath_sdma_reset_wait = 200; + __set_bit(IPATH_SDMA_DISARMED, &dd->ipath_sdma_status); + if (!test_bit(IPATH_SDMA_SHUTDOWN, &dd->ipath_sdma_status)) + tasklet_hi_schedule(&dd->ipath_sdma_abort_task); + spin_unlock_irqrestore(&dd->ipath_sdma_lock, flags); + } +bail:; } +/* + * Force an update of in-memory copy of the pioavail registers, when + * needed for any of a variety of reasons. We read the scratch register + * to make it highly likely that the update will have happened by the + * time we return. If already off (as in cancel_sends above), this + * routine is a nop, on the assumption that the caller will "do the + * right thing". + */ +void ipath_force_pio_avail_update(struct ipath_devdata *dd) +{ + unsigned long flags; + + spin_lock_irqsave(&dd->ipath_sendctrl_lock, flags); + if (dd->ipath_sendctrl & INFINIPATH_S_PIOBUFAVAILUPD) { + ipath_write_kreg(dd, dd->ipath_kregs->kr_sendctrl, + dd->ipath_sendctrl & ~INFINIPATH_S_PIOBUFAVAILUPD); + ipath_read_kreg64(dd, dd->ipath_kregs->kr_scratch); + ipath_write_kreg(dd, dd->ipath_kregs->kr_sendctrl, + dd->ipath_sendctrl); + ipath_read_kreg64(dd, dd->ipath_kregs->kr_scratch); + } + spin_unlock_irqrestore(&dd->ipath_sendctrl_lock, flags); +} -static void ipath_set_ib_lstate(struct ipath_devdata *dd, int which) +static void ipath_set_ib_lstate(struct ipath_devdata *dd, int linkcmd, + int linitcmd) { + u64 mod_wd; static const char *what[4] = { [0] = "NOP", [INFINIPATH_IBCC_LINKCMD_DOWN] = "DOWN", [INFINIPATH_IBCC_LINKCMD_ARMED] = "ARMED", [INFINIPATH_IBCC_LINKCMD_ACTIVE] = "ACTIVE" }; - int linkcmd = (which >> INFINIPATH_IBCC_LINKCMD_SHIFT) & - INFINIPATH_IBCC_LINKCMD_MASK; - - ipath_cdbg(VERBOSE, "Trying to move unit %u to %s, current ltstate " - "is %s\n", dd->ipath_unit, - what[linkcmd], - ipath_ibcstatus_str[ - (ipath_read_kreg64 - (dd, dd->ipath_kregs->kr_ibcstatus) >> - INFINIPATH_IBCS_LINKTRAININGSTATE_SHIFT) & - INFINIPATH_IBCS_LINKTRAININGSTATE_MASK]); - /* flush all queued sends when going to DOWN to be sure that - * they don't block MAD packets */ - if (linkcmd == INFINIPATH_IBCC_LINKCMD_DOWN) - ipath_cancel_sends(dd, 1); + + if (linitcmd == INFINIPATH_IBCC_LINKINITCMD_DISABLE) { + /* + * If we are told to disable, note that so link-recovery + * code does not attempt to bring us back up. + */ + preempt_disable(); + dd->ipath_flags |= IPATH_IB_LINK_DISABLED; + preempt_enable(); + } else if (linitcmd) { + /* + * Any other linkinitcmd will lead to LINKDOWN and then + * to INIT (if all is well), so clear flag to let + * link-recovery code attempt to bring us back up. + */ + preempt_disable(); + dd->ipath_flags &= ~IPATH_IB_LINK_DISABLED; + preempt_enable(); + } + + mod_wd = (linkcmd << dd->ibcc_lc_shift) | + (linitcmd << INFINIPATH_IBCC_LINKINITCMD_SHIFT); + ipath_cdbg(VERBOSE, + "Moving unit %u to %s (initcmd=0x%x), current ltstate is %s\n", + dd->ipath_unit, what[linkcmd], linitcmd, + ipath_ibcstatus_str[ipath_ib_linktrstate(dd, + ipath_read_kreg64(dd, dd->ipath_kregs->kr_ibcstatus))]); ipath_write_kreg(dd, dd->ipath_kregs->kr_ibcctrl, - dd->ipath_ibcctrl | which); + dd->ipath_ibcctrl | mod_wd); + /* read from chip so write is flushed */ + (void) ipath_read_kreg64(dd, dd->ipath_kregs->kr_ibcstatus); } int ipath_set_linkstate(struct ipath_devdata *dd, u8 newstate) @@ -1687,30 +1928,28 @@ int ipath_set_linkstate(struct ipath_devdata *dd, u8 newstate) switch (newstate) { case IPATH_IB_LINKDOWN_ONLY: - ipath_set_ib_lstate(dd, INFINIPATH_IBCC_LINKCMD_DOWN << - INFINIPATH_IBCC_LINKCMD_SHIFT); + ipath_set_ib_lstate(dd, INFINIPATH_IBCC_LINKCMD_DOWN, 0); /* don't wait */ ret = 0; goto bail; case IPATH_IB_LINKDOWN: - ipath_set_ib_lstate(dd, INFINIPATH_IBCC_LINKINITCMD_POLL << - INFINIPATH_IBCC_LINKINITCMD_SHIFT); + ipath_set_ib_lstate(dd, INFINIPATH_IBCC_LINKCMD_DOWN, + INFINIPATH_IBCC_LINKINITCMD_POLL); /* don't wait */ ret = 0; goto bail; case IPATH_IB_LINKDOWN_SLEEP: - ipath_set_ib_lstate(dd, INFINIPATH_IBCC_LINKINITCMD_SLEEP << - INFINIPATH_IBCC_LINKINITCMD_SHIFT); + ipath_set_ib_lstate(dd, INFINIPATH_IBCC_LINKCMD_DOWN, + INFINIPATH_IBCC_LINKINITCMD_SLEEP); /* don't wait */ ret = 0; goto bail; case IPATH_IB_LINKDOWN_DISABLE: - ipath_set_ib_lstate(dd, - INFINIPATH_IBCC_LINKINITCMD_DISABLE << - INFINIPATH_IBCC_LINKINITCMD_SHIFT); + ipath_set_ib_lstate(dd, INFINIPATH_IBCC_LINKCMD_DOWN, + INFINIPATH_IBCC_LINKINITCMD_DISABLE); /* don't wait */ ret = 0; goto bail; @@ -1725,8 +1964,8 @@ int ipath_set_linkstate(struct ipath_devdata *dd, u8 newstate) ret = -EINVAL; goto bail; } - ipath_set_ib_lstate(dd, INFINIPATH_IBCC_LINKCMD_ARMED << - INFINIPATH_IBCC_LINKCMD_SHIFT); + ipath_set_ib_lstate(dd, INFINIPATH_IBCC_LINKCMD_ARMED, 0); + /* * Since the port can transition to ACTIVE by receiving * a non VL 15 packet, wait for either state. @@ -1743,8 +1982,7 @@ int ipath_set_linkstate(struct ipath_devdata *dd, u8 newstate) ret = -EINVAL; goto bail; } - ipath_set_ib_lstate(dd, INFINIPATH_IBCC_LINKCMD_ACTIVE << - INFINIPATH_IBCC_LINKCMD_SHIFT); + ipath_set_ib_lstate(dd, INFINIPATH_IBCC_LINKCMD_ACTIVE, 0); lstate = IPATH_LINKACTIVE; break; @@ -1753,16 +1991,41 @@ int ipath_set_linkstate(struct ipath_devdata *dd, u8 newstate) dd->ipath_ibcctrl |= INFINIPATH_IBCC_LOOPBACK; ipath_write_kreg(dd, dd->ipath_kregs->kr_ibcctrl, dd->ipath_ibcctrl); + + /* turn heartbeat off, as it causes loopback to fail */ + dd->ipath_f_set_ib_cfg(dd, IPATH_IB_CFG_HRTBT, + IPATH_IB_HRTBT_OFF); + /* don't wait */ ret = 0; - goto bail; // no state change to wait for + goto bail; case IPATH_IB_LINK_EXTERNAL: - dev_info(&dd->pcidev->dev, "Disabling IB local loopback (normal)\n"); + dev_info(&dd->pcidev->dev, + "Disabling IB local loopback (normal)\n"); + dd->ipath_f_set_ib_cfg(dd, IPATH_IB_CFG_HRTBT, + IPATH_IB_HRTBT_ON); dd->ipath_ibcctrl &= ~INFINIPATH_IBCC_LOOPBACK; ipath_write_kreg(dd, dd->ipath_kregs->kr_ibcctrl, dd->ipath_ibcctrl); + /* don't wait */ ret = 0; - goto bail; // no state change to wait for + goto bail; + + /* + * Heartbeat can be explicitly enabled by the user via + * "hrtbt_enable" "file", and if disabled, trying to enable here + * will have no effect. Implicit changes (heartbeat off when + * loopback on, and vice versa) are included to ease testing. + */ + case IPATH_IB_LINK_HRTBT: + ret = dd->ipath_f_set_ib_cfg(dd, IPATH_IB_CFG_HRTBT, + IPATH_IB_HRTBT_ON); + goto bail; + + case IPATH_IB_LINK_NO_HRTBT: + ret = dd->ipath_f_set_ib_cfg(dd, IPATH_IB_CFG_HRTBT, + IPATH_IB_HRTBT_OFF); + goto bail; default: ipath_dbg("Invalid linkstate 0x%x requested\n", newstate); @@ -1785,7 +2048,7 @@ bail: * sanity checking on this, and we don't deal with what happens to * programs that are already running when the size changes. * NOTE: changing the MTU will usually cause the IBC to go back to - * link initialize (IPATH_IBSTATE_INIT) state... + * link INIT state... */ int ipath_set_mtu(struct ipath_devdata *dd, u16 arg) { @@ -1800,7 +2063,7 @@ int ipath_set_mtu(struct ipath_devdata *dd, u16 arg) * piosize). We check that it's one of the valid IB sizes. */ if (arg != 256 && arg != 512 && arg != 1024 && arg != 2048 && - arg != 4096) { + (arg != 4096 || !ipath_mtu4096)) { ipath_dbg("Trying to set invalid mtu %u, failing\n", arg); ret = -EINVAL; goto bail; @@ -1816,6 +2079,8 @@ int ipath_set_mtu(struct ipath_devdata *dd, u16 arg) if (arg >= (piosize - IPATH_PIO_MAXIBHDR)) { /* Only if it's not the initial value (or reset to it) */ if (piosize != dd->ipath_init_ibmaxlen) { + if (arg > piosize && arg <= dd->ipath_init_ibmaxlen) + piosize = dd->ipath_init_ibmaxlen; dd->ipath_ibmaxlen = piosize; changed = 1; } @@ -1829,24 +2094,17 @@ int ipath_set_mtu(struct ipath_devdata *dd, u16 arg) } if (changed) { + u64 ibc = dd->ipath_ibcctrl, ibdw; /* - * set the IBC maxpktlength to the size of our pio - * buffers in words + * update our housekeeping variables, and set IBC max + * size, same as init code; max IBC is max we allow in + * buffer, less the qword pbc, plus 1 for ICRC, in dwords */ - u64 ibc = dd->ipath_ibcctrl; + dd->ipath_ibmaxlen = piosize - 2 * sizeof(u32); + ibdw = (dd->ipath_ibmaxlen >> 2) + 1; ibc &= ~(INFINIPATH_IBCC_MAXPKTLEN_MASK << - INFINIPATH_IBCC_MAXPKTLEN_SHIFT); - - piosize = piosize - 2 * sizeof(u32); /* ignore pbc */ - dd->ipath_ibmaxlen = piosize; - piosize /= sizeof(u32); /* in words */ - /* - * for ICRC, which we only send in diag test pkt mode, and - * we don't need to worry about that for mtu - */ - piosize += 1; - - ibc |= piosize << INFINIPATH_IBCC_MAXPKTLEN_SHIFT; + dd->ibcc_mpl_shift); + ibc |= ibdw << dd->ibcc_mpl_shift; dd->ipath_ibcctrl = ibc; ipath_write_kreg(dd, dd->ipath_kregs->kr_ibcctrl, dd->ipath_ibcctrl); @@ -1859,11 +2117,16 @@ bail: return ret; } -int ipath_set_lid(struct ipath_devdata *dd, u32 arg, u8 lmc) +int ipath_set_lid(struct ipath_devdata *dd, u32 lid, u8 lmc) { - dd->ipath_lid = arg; + dd->ipath_lid = lid; dd->ipath_lmc = lmc; + dd->ipath_f_set_ib_cfg(dd, IPATH_IB_CFG_LIDLMC, lid | + (~((1U << lmc) - 1)) << 16); + + dev_info(&dd->pcidev->dev, "We got a lid: 0x%x\n", lid); + return 0; } @@ -1925,10 +2188,8 @@ static void ipath_run_led_override(unsigned long opaque) * but leave that to per-chip functions. */ val = ipath_read_kreg64(dd, dd->ipath_kregs->kr_ibcstatus); - ltstate = (val >> INFINIPATH_IBCS_LINKTRAININGSTATE_SHIFT) & - INFINIPATH_IBCS_LINKTRAININGSTATE_MASK; - lstate = (val >> INFINIPATH_IBCS_LINKSTATE_SHIFT) & - INFINIPATH_IBCS_LINKSTATE_MASK; + ltstate = ipath_ib_linktrstate(dd, val); + lstate = ipath_ib_linkstate(dd, val); dd->ipath_f_setextled(dd, lstate, ltstate); mod_timer(&dd->ipath_led_override_timer, jiffies + timeoff); @@ -1969,9 +2230,8 @@ void ipath_set_led_override(struct ipath_devdata *dd, unsigned int val) dd->ipath_led_override_timer.data = (unsigned long) dd; dd->ipath_led_override_timer.expires = jiffies + 1; add_timer(&dd->ipath_led_override_timer); - } else { + } else atomic_dec(&dd->ipath_led_override_timer_active); - } } /** @@ -1989,6 +2249,8 @@ void ipath_shutdown_device(struct ipath_devdata *dd) ipath_dbg("Shutting down the device\n"); + ipath_hol_up(dd); /* make sure user processes aren't suspended */ + dd->ipath_flags |= IPATH_LINKUNK; dd->ipath_flags &= ~(IPATH_INITTED | IPATH_LINKDOWN | IPATH_LINKINIT | IPATH_LINKARMED | @@ -2003,6 +2265,9 @@ void ipath_shutdown_device(struct ipath_devdata *dd) ipath_write_kreg(dd, dd->ipath_kregs->kr_rcvctrl, dd->ipath_rcvctrl); + if (dd->ipath_flags & IPATH_HAS_SEND_DMA) + teardown_sdma(dd); + /* * gracefully stop all sends allowing any in progress to trickle out * first. @@ -2020,10 +2285,16 @@ void ipath_shutdown_device(struct ipath_devdata *dd) */ udelay(5); - ipath_set_ib_lstate(dd, INFINIPATH_IBCC_LINKINITCMD_DISABLE << - INFINIPATH_IBCC_LINKINITCMD_SHIFT); + dd->ipath_f_setextled(dd, 0, 0); /* make sure LEDs are off */ + + ipath_set_ib_lstate(dd, 0, INFINIPATH_IBCC_LINKINITCMD_DISABLE); ipath_cancel_sends(dd, 0); + /* + * we are shutting down, so tell components that care. We don't do + * this on just a link state change, much like ethernet, a cable + * unplug, etc. doesn't change driver state + */ signal_ib_event(dd, IB_EVENT_PORT_ERR); /* disable IBC */ @@ -2038,10 +2309,20 @@ void ipath_shutdown_device(struct ipath_devdata *dd) */ dd->ipath_f_quiet_serdes(dd); + /* stop all the timers that might still be running */ + del_timer_sync(&dd->ipath_hol_timer); if (dd->ipath_stats_timer_active) { del_timer_sync(&dd->ipath_stats_timer); dd->ipath_stats_timer_active = 0; } + if (dd->ipath_intrchk_timer.data) { + del_timer_sync(&dd->ipath_intrchk_timer); + dd->ipath_intrchk_timer.data = 0; + } + if (atomic_read(&dd->ipath_led_override_timer_active)) { + del_timer_sync(&dd->ipath_led_override_timer); + atomic_set(&dd->ipath_led_override_timer_active, 0); + } /* * clear all interrupts and errors, so that the next time the driver @@ -2115,13 +2396,13 @@ void ipath_free_pddata(struct ipath_devdata *dd, struct ipath_portdata *pd) ipath_cdbg(VERBOSE, "free closed port %d " "ipath_port0_skbinfo @ %p\n", pd->port_port, skbinfo); - for (e = 0; e < dd->ipath_rcvegrcnt; e++) - if (skbinfo[e].skb) { - pci_unmap_single(dd->pcidev, skbinfo[e].phys, - dd->ipath_ibmaxlen, - PCI_DMA_FROMDEVICE); - dev_kfree_skb(skbinfo[e].skb); - } + for (e = 0; e < dd->ipath_p0_rcvegrcnt; e++) + if (skbinfo[e].skb) { + pci_unmap_single(dd->pcidev, skbinfo[e].phys, + dd->ipath_ibmaxlen, + PCI_DMA_FROMDEVICE); + dev_kfree_skb(skbinfo[e].skb); + } vfree(skbinfo); } kfree(pd->port_tid_pg_list); @@ -2144,6 +2425,7 @@ static int __init infinipath_init(void) */ idr_init(&unit_table); if (!idr_pre_get(&unit_table, GFP_KERNEL)) { + printk(KERN_ERR IPATH_DRV_NAME ": idr_pre_get() failed\n"); ret = -ENOMEM; goto bail; } @@ -2235,13 +2517,18 @@ int ipath_reset_device(int unit) } } + if (dd->ipath_flags & IPATH_HAS_SEND_DMA) + teardown_sdma(dd); + dd->ipath_flags &= ~IPATH_INITTED; + ipath_write_kreg(dd, dd->ipath_kregs->kr_intmask, 0ULL); ret = dd->ipath_f_reset(dd); - if (ret != 1) - ipath_dbg("reset was not successful\n"); - ipath_dbg("Trying to reinitialize unit %u after reset attempt\n", - unit); - ret = ipath_init_chip(dd, 1); + if (ret == 1) { + ipath_dbg("Reinitializing unit %u after reset attempt\n", + unit); + ret = ipath_init_chip(dd, 1); + } else + ret = -EAGAIN; if (ret) ipath_dev_err(dd, "Reinitialize unit %u after " "reset failed with %d\n", unit, ret); @@ -2253,13 +2540,121 @@ bail: return ret; } +/* + * send a signal to all the processes that have the driver open + * through the normal interfaces (i.e., everything other than diags + * interface). Returns number of signalled processes. + */ +static int ipath_signal_procs(struct ipath_devdata *dd, int sig) +{ + int i, sub, any = 0; + pid_t pid; + + if (!dd->ipath_pd) + return 0; + for (i = 1; i < dd->ipath_cfgports; i++) { + if (!dd->ipath_pd[i] || !dd->ipath_pd[i]->port_cnt || + !dd->ipath_pd[i]->port_pid) + continue; + pid = dd->ipath_pd[i]->port_pid; + dev_info(&dd->pcidev->dev, "context %d in use " + "(PID %u), sending signal %d\n", + i, pid, sig); + kill_proc(pid, sig, 1); + any++; + for (sub = 0; sub < INFINIPATH_MAX_SUBPORT; sub++) { + pid = dd->ipath_pd[i]->port_subpid[sub]; + if (!pid) + continue; + dev_info(&dd->pcidev->dev, "sub-context " + "%d:%d in use (PID %u), sending " + "signal %d\n", i, sub, pid, sig); + kill_proc(pid, sig, 1); + any++; + } + } + return any; +} + +static void ipath_hol_signal_down(struct ipath_devdata *dd) +{ + if (ipath_signal_procs(dd, SIGSTOP)) + ipath_dbg("Stopped some processes\n"); + ipath_cancel_sends(dd, 1); +} + + +static void ipath_hol_signal_up(struct ipath_devdata *dd) +{ + if (ipath_signal_procs(dd, SIGCONT)) + ipath_dbg("Continued some processes\n"); +} + +/* + * link is down, stop any users processes, and flush pending sends + * to prevent HoL blocking, then start the HoL timer that + * periodically continues, then stop procs, so they can detect + * link down if they want, and do something about it. + * Timer may already be running, so use __mod_timer, not add_timer. + */ +void ipath_hol_down(struct ipath_devdata *dd) +{ + dd->ipath_hol_state = IPATH_HOL_DOWN; + ipath_hol_signal_down(dd); + dd->ipath_hol_next = IPATH_HOL_DOWNCONT; + dd->ipath_hol_timer.expires = jiffies + + msecs_to_jiffies(ipath_hol_timeout_ms); + __mod_timer(&dd->ipath_hol_timer, dd->ipath_hol_timer.expires); +} + +/* + * link is up, continue any user processes, and ensure timer + * is a nop, if running. Let timer keep running, if set; it + * will nop when it sees the link is up + */ +void ipath_hol_up(struct ipath_devdata *dd) +{ + ipath_hol_signal_up(dd); + dd->ipath_hol_state = IPATH_HOL_UP; +} + +/* + * toggle the running/not running state of user proceses + * to prevent HoL blocking on chip resources, but still allow + * user processes to do link down special case handling. + * Should only be called via the timer + */ +void ipath_hol_event(unsigned long opaque) +{ + struct ipath_devdata *dd = (struct ipath_devdata *)opaque; + + if (dd->ipath_hol_next == IPATH_HOL_DOWNSTOP + && dd->ipath_hol_state != IPATH_HOL_UP) { + dd->ipath_hol_next = IPATH_HOL_DOWNCONT; + ipath_dbg("Stopping processes\n"); + ipath_hol_signal_down(dd); + } else { /* may do "extra" if also in ipath_hol_up() */ + dd->ipath_hol_next = IPATH_HOL_DOWNSTOP; + ipath_dbg("Continuing processes\n"); + ipath_hol_signal_up(dd); + } + if (dd->ipath_hol_state == IPATH_HOL_UP) + ipath_dbg("link's up, don't resched timer\n"); + else { + dd->ipath_hol_timer.expires = jiffies + + msecs_to_jiffies(ipath_hol_timeout_ms); + __mod_timer(&dd->ipath_hol_timer, + dd->ipath_hol_timer.expires); + } +} + int ipath_set_rx_pol_inv(struct ipath_devdata *dd, u8 new_pol_inv) { u64 val; - if ( new_pol_inv > INFINIPATH_XGXS_RX_POL_MASK ) { + + if (new_pol_inv > INFINIPATH_XGXS_RX_POL_MASK) return -1; - } - if ( dd->ipath_rx_pol_inv != new_pol_inv ) { + if (dd->ipath_rx_pol_inv != new_pol_inv) { dd->ipath_rx_pol_inv = new_pol_inv; val = ipath_read_kreg64(dd, dd->ipath_kregs->kr_xgxsconfig); val &= ~(INFINIPATH_XGXS_RX_POL_MASK << diff --git a/drivers/infiniband/hw/ipath/ipath_eeprom.c b/drivers/infiniband/hw/ipath/ipath_eeprom.c index e28a42f53769..dc37277f1c80 100644 --- a/drivers/infiniband/hw/ipath/ipath_eeprom.c +++ b/drivers/infiniband/hw/ipath/ipath_eeprom.c @@ -1,5 +1,5 @@ /* - * Copyright (c) 2006, 2007 QLogic Corporation. All rights reserved. + * Copyright (c) 2006, 2007, 2008 QLogic Corporation. All rights reserved. * Copyright (c) 2003, 2004, 2005, 2006 PathScale, Inc. All rights reserved. * * This software is available to you under a choice of one of two @@ -62,6 +62,33 @@ * accessing eeprom contents from within the kernel, only via sysfs. */ +/* Added functionality for IBA7220-based cards */ +#define IPATH_EEPROM_DEV_V1 0xA0 +#define IPATH_EEPROM_DEV_V2 0xA2 +#define IPATH_TEMP_DEV 0x98 +#define IPATH_BAD_DEV (IPATH_EEPROM_DEV_V2+2) +#define IPATH_NO_DEV (0xFF) + +/* + * The number of I2C chains is proliferating. Table below brings + * some order to the madness. The basic principle is that the + * table is scanned from the top, and a "probe" is made to the + * device probe_dev. If that succeeds, the chain is considered + * to be of that type, and dd->i2c_chain_type is set to the index+1 + * of the entry. + * The +1 is so static initialization can mean "unknown, do probe." + */ +static struct i2c_chain_desc { + u8 probe_dev; /* If seen at probe, chain is this type */ + u8 eeprom_dev; /* Dev addr (if any) for EEPROM */ + u8 temp_dev; /* Dev Addr (if any) for Temp-sense */ +} i2c_chains[] = { + { IPATH_BAD_DEV, IPATH_NO_DEV, IPATH_NO_DEV }, /* pre-iba7220 bds */ + { IPATH_EEPROM_DEV_V1, IPATH_EEPROM_DEV_V1, IPATH_TEMP_DEV}, /* V1 */ + { IPATH_EEPROM_DEV_V2, IPATH_EEPROM_DEV_V2, IPATH_TEMP_DEV}, /* V2 */ + { IPATH_NO_DEV } +}; + enum i2c_type { i2c_line_scl = 0, i2c_line_sda @@ -75,13 +102,6 @@ enum i2c_state { #define READ_CMD 1 #define WRITE_CMD 0 -static int eeprom_init; - -/* - * The gpioval manipulation really should be protected by spinlocks - * or be converted to use atomic operations. - */ - /** * i2c_gpio_set - set a GPIO line * @dd: the infinipath device @@ -241,6 +261,27 @@ static int i2c_ackrcv(struct ipath_devdata *dd) } /** + * rd_byte - read a byte, leaving ACK, STOP, etc up to caller + * @dd: the infinipath device + * + * Returns byte shifted out of device + */ +static int rd_byte(struct ipath_devdata *dd) +{ + int bit_cntr, data; + + data = 0; + + for (bit_cntr = 7; bit_cntr >= 0; --bit_cntr) { + data <<= 1; + scl_out(dd, i2c_line_high); + data |= sda_in(dd, 0); + scl_out(dd, i2c_line_low); + } + return data; +} + +/** * wr_byte - write a byte, one bit at a time * @dd: the infinipath device * @data: the byte to write @@ -331,7 +372,6 @@ static int eeprom_reset(struct ipath_devdata *dd) ipath_cdbg(VERBOSE, "Resetting i2c eeprom; initial gpioout reg " "is %llx\n", (unsigned long long) *gpioval); - eeprom_init = 1; /* * This is to get the i2c into a known state, by first going low, * then tristate sda (and then tristate scl as first thing @@ -340,12 +380,17 @@ static int eeprom_reset(struct ipath_devdata *dd) scl_out(dd, i2c_line_low); sda_out(dd, i2c_line_high); + /* Clock up to 9 cycles looking for SDA hi, then issue START and STOP */ while (clock_cycles_left--) { scl_out(dd, i2c_line_high); + /* SDA seen high, issue START by dropping it while SCL high */ if (sda_in(dd, 0)) { sda_out(dd, i2c_line_low); scl_out(dd, i2c_line_low); + /* ATMEL spec says must be followed by STOP. */ + scl_out(dd, i2c_line_high); + sda_out(dd, i2c_line_high); ret = 0; goto bail; } @@ -359,29 +404,121 @@ bail: return ret; } -/** - * ipath_eeprom_read - receives bytes from the eeprom via I2C - * @dd: the infinipath device - * @eeprom_offset: address to read from - * @buffer: where to store result - * @len: number of bytes to receive +/* + * Probe for I2C device at specified address. Returns 0 for "success" + * to match rest of this file. + * Leave bus in "reasonable" state for further commands. */ +static int i2c_probe(struct ipath_devdata *dd, int devaddr) +{ + int ret = 0; + + ret = eeprom_reset(dd); + if (ret) { + ipath_dev_err(dd, "Failed reset probing device 0x%02X\n", + devaddr); + return ret; + } + /* + * Reset no longer leaves bus in start condition, so normal + * i2c_startcmd() will do. + */ + ret = i2c_startcmd(dd, devaddr | READ_CMD); + if (ret) + ipath_cdbg(VERBOSE, "Failed startcmd for device 0x%02X\n", + devaddr); + else { + /* + * Device did respond. Complete a single-byte read, because some + * devices apparently cannot handle STOP immediately after they + * ACK the start-cmd. + */ + int data; + data = rd_byte(dd); + stop_cmd(dd); + ipath_cdbg(VERBOSE, "Response from device 0x%02X\n", devaddr); + } + return ret; +} + +/* + * Returns the "i2c type". This is a pointer to a struct that describes + * the I2C chain on this board. To minimize impact on struct ipath_devdata, + * the (small integer) index into the table is actually memoized, rather + * then the pointer. + * Memoization is because the type is determined on the first call per chip. + * An alternative would be to move type determination to early + * init code. + */ +static struct i2c_chain_desc *ipath_i2c_type(struct ipath_devdata *dd) +{ + int idx; + + /* Get memoized index, from previous successful probes */ + idx = dd->ipath_i2c_chain_type - 1; + if (idx >= 0 && idx < (ARRAY_SIZE(i2c_chains) - 1)) + goto done; + + idx = 0; + while (i2c_chains[idx].probe_dev != IPATH_NO_DEV) { + /* if probe succeeds, this is type */ + if (!i2c_probe(dd, i2c_chains[idx].probe_dev)) + break; + ++idx; + } + + /* + * Old EEPROM (first entry) may require a reset after probe, + * rather than being able to "start" after "stop" + */ + if (idx == 0) + eeprom_reset(dd); + + if (i2c_chains[idx].probe_dev == IPATH_NO_DEV) + idx = -1; + else + dd->ipath_i2c_chain_type = idx + 1; +done: + return (idx >= 0) ? i2c_chains + idx : NULL; +} static int ipath_eeprom_internal_read(struct ipath_devdata *dd, u8 eeprom_offset, void *buffer, int len) { - /* compiler complains unless initialized */ - u8 single_byte = 0; - int bit_cntr; int ret; + struct i2c_chain_desc *icd; + u8 *bp = buffer; - if (!eeprom_init) - eeprom_reset(dd); - - eeprom_offset = (eeprom_offset << 1) | READ_CMD; + ret = 1; + icd = ipath_i2c_type(dd); + if (!icd) + goto bail; - if (i2c_startcmd(dd, eeprom_offset)) { - ipath_dbg("Failed startcmd\n"); + if (icd->eeprom_dev == IPATH_NO_DEV) { + /* legacy not-really-I2C */ + ipath_cdbg(VERBOSE, "Start command only address\n"); + eeprom_offset = (eeprom_offset << 1) | READ_CMD; + ret = i2c_startcmd(dd, eeprom_offset); + } else { + /* Actual I2C */ + ipath_cdbg(VERBOSE, "Start command uses devaddr\n"); + if (i2c_startcmd(dd, icd->eeprom_dev | WRITE_CMD)) { + ipath_dbg("Failed EEPROM startcmd\n"); + stop_cmd(dd); + ret = 1; + goto bail; + } + ret = wr_byte(dd, eeprom_offset); + stop_cmd(dd); + if (ret) { + ipath_dev_err(dd, "Failed to write EEPROM address\n"); + ret = 1; + goto bail; + } + ret = i2c_startcmd(dd, icd->eeprom_dev | READ_CMD); + } + if (ret) { + ipath_dbg("Failed startcmd for dev %02X\n", icd->eeprom_dev); stop_cmd(dd); ret = 1; goto bail; @@ -392,22 +529,11 @@ static int ipath_eeprom_internal_read(struct ipath_devdata *dd, * incrementing the address. */ while (len-- > 0) { - /* get data */ - single_byte = 0; - for (bit_cntr = 8; bit_cntr; bit_cntr--) { - u8 bit; - scl_out(dd, i2c_line_high); - bit = sda_in(dd, 0); - single_byte |= bit << (bit_cntr - 1); - scl_out(dd, i2c_line_low); - } - + /* get and store data */ + *bp++ = rd_byte(dd); /* send ack if not the last byte */ if (len) send_ack(dd); - - *((u8 *) buffer) = single_byte; - buffer++; } stop_cmd(dd); @@ -418,31 +544,40 @@ bail: return ret; } - -/** - * ipath_eeprom_write - writes data to the eeprom via I2C - * @dd: the infinipath device - * @eeprom_offset: where to place data - * @buffer: data to write - * @len: number of bytes to write - */ static int ipath_eeprom_internal_write(struct ipath_devdata *dd, u8 eeprom_offset, const void *buffer, int len) { - u8 single_byte; int sub_len; const u8 *bp = buffer; int max_wait_time, i; int ret; + struct i2c_chain_desc *icd; - if (!eeprom_init) - eeprom_reset(dd); + ret = 1; + icd = ipath_i2c_type(dd); + if (!icd) + goto bail; while (len > 0) { - if (i2c_startcmd(dd, (eeprom_offset << 1) | WRITE_CMD)) { - ipath_dbg("Failed to start cmd offset %u\n", - eeprom_offset); - goto failed_write; + if (icd->eeprom_dev == IPATH_NO_DEV) { + if (i2c_startcmd(dd, + (eeprom_offset << 1) | WRITE_CMD)) { + ipath_dbg("Failed to start cmd offset %u\n", + eeprom_offset); + goto failed_write; + } + } else { + /* Real I2C */ + if (i2c_startcmd(dd, icd->eeprom_dev | WRITE_CMD)) { + ipath_dbg("Failed EEPROM startcmd\n"); + goto failed_write; + } + ret = wr_byte(dd, eeprom_offset); + if (ret) { + ipath_dev_err(dd, "Failed to write EEPROM " + "address\n"); + goto failed_write; + } } sub_len = min(len, 4); @@ -468,9 +603,11 @@ static int ipath_eeprom_internal_write(struct ipath_devdata *dd, u8 eeprom_offse * the writes have completed. We do this inline to avoid * the debug prints that are in the real read routine * if the startcmd fails. + * We also use the proper device address, so it doesn't matter + * whether we have real eeprom_dev. legacy likes any address. */ max_wait_time = 100; - while (i2c_startcmd(dd, READ_CMD)) { + while (i2c_startcmd(dd, icd->eeprom_dev | READ_CMD)) { stop_cmd(dd); if (!--max_wait_time) { ipath_dbg("Did not get successful read to " @@ -478,15 +615,8 @@ static int ipath_eeprom_internal_write(struct ipath_devdata *dd, u8 eeprom_offse goto failed_write; } } - /* now read the zero byte */ - for (i = single_byte = 0; i < 8; i++) { - u8 bit; - scl_out(dd, i2c_line_high); - bit = sda_in(dd, 0); - scl_out(dd, i2c_line_low); - single_byte <<= 1; - single_byte |= bit; - } + /* now read (and ignore) the resulting byte */ + rd_byte(dd); stop_cmd(dd); } @@ -501,9 +631,12 @@ bail: return ret; } -/* - * The public entry-points ipath_eeprom_read() and ipath_eeprom_write() - * are now just wrappers around the internal functions. +/** + * ipath_eeprom_read - receives bytes from the eeprom via I2C + * @dd: the infinipath device + * @eeprom_offset: address to read from + * @buffer: where to store result + * @len: number of bytes to receive */ int ipath_eeprom_read(struct ipath_devdata *dd, u8 eeprom_offset, void *buff, int len) @@ -519,6 +652,13 @@ int ipath_eeprom_read(struct ipath_devdata *dd, u8 eeprom_offset, return ret; } +/** + * ipath_eeprom_write - writes data to the eeprom via I2C + * @dd: the infinipath device + * @eeprom_offset: where to place data + * @buffer: data to write + * @len: number of bytes to write + */ int ipath_eeprom_write(struct ipath_devdata *dd, u8 eeprom_offset, const void *buff, int len) { @@ -820,7 +960,7 @@ int ipath_update_eeprom_log(struct ipath_devdata *dd) * if we log an hour at 31 minutes, then we would need to set * active_time to -29 to accurately count the _next_ hour. */ - if (new_time > 3600) { + if (new_time >= 3600) { new_hrs = new_time / 3600; atomic_sub((new_hrs * 3600), &dd->ipath_active_time); new_hrs += dd->ipath_eep_hrs; @@ -885,3 +1025,159 @@ void ipath_inc_eeprom_err(struct ipath_devdata *dd, u32 eidx, u32 incr) spin_unlock_irqrestore(&dd->ipath_eep_st_lock, flags); return; } + +static int ipath_tempsense_internal_read(struct ipath_devdata *dd, u8 regnum) +{ + int ret; + struct i2c_chain_desc *icd; + + ret = -ENOENT; + + icd = ipath_i2c_type(dd); + if (!icd) + goto bail; + + if (icd->temp_dev == IPATH_NO_DEV) { + /* tempsense only exists on new, real-I2C boards */ + ret = -ENXIO; + goto bail; + } + + if (i2c_startcmd(dd, icd->temp_dev | WRITE_CMD)) { + ipath_dbg("Failed tempsense startcmd\n"); + stop_cmd(dd); + ret = -ENXIO; + goto bail; + } + ret = wr_byte(dd, regnum); + stop_cmd(dd); + if (ret) { + ipath_dev_err(dd, "Failed tempsense WR command %02X\n", + regnum); + ret = -ENXIO; + goto bail; + } + if (i2c_startcmd(dd, icd->temp_dev | READ_CMD)) { + ipath_dbg("Failed tempsense RD startcmd\n"); + stop_cmd(dd); + ret = -ENXIO; + goto bail; + } + /* + * We can only clock out one byte per command, sensibly + */ + ret = rd_byte(dd); + stop_cmd(dd); + +bail: + return ret; +} + +#define VALID_TS_RD_REG_MASK 0xBF + +/** + * ipath_tempsense_read - read register of temp sensor via I2C + * @dd: the infinipath device + * @regnum: register to read from + * + * returns reg contents (0..255) or < 0 for error + */ +int ipath_tempsense_read(struct ipath_devdata *dd, u8 regnum) +{ + int ret; + + if (regnum > 7) + return -EINVAL; + + /* return a bogus value for (the one) register we do not have */ + if (!((1 << regnum) & VALID_TS_RD_REG_MASK)) + return 0; + + ret = mutex_lock_interruptible(&dd->ipath_eep_lock); + if (!ret) { + ret = ipath_tempsense_internal_read(dd, regnum); + mutex_unlock(&dd->ipath_eep_lock); + } + + /* + * There are three possibilities here: + * ret is actual value (0..255) + * ret is -ENXIO or -EINVAL from code in this file + * ret is -EINTR from mutex_lock_interruptible. + */ + return ret; +} + +static int ipath_tempsense_internal_write(struct ipath_devdata *dd, + u8 regnum, u8 data) +{ + int ret = -ENOENT; + struct i2c_chain_desc *icd; + + icd = ipath_i2c_type(dd); + if (!icd) + goto bail; + + if (icd->temp_dev == IPATH_NO_DEV) { + /* tempsense only exists on new, real-I2C boards */ + ret = -ENXIO; + goto bail; + } + if (i2c_startcmd(dd, icd->temp_dev | WRITE_CMD)) { + ipath_dbg("Failed tempsense startcmd\n"); + stop_cmd(dd); + ret = -ENXIO; + goto bail; + } + ret = wr_byte(dd, regnum); + if (ret) { + stop_cmd(dd); + ipath_dev_err(dd, "Failed to write tempsense command %02X\n", + regnum); + ret = -ENXIO; + goto bail; + } + ret = wr_byte(dd, data); + stop_cmd(dd); + ret = i2c_startcmd(dd, icd->temp_dev | READ_CMD); + if (ret) { + ipath_dev_err(dd, "Failed tempsense data wrt to %02X\n", + regnum); + ret = -ENXIO; + } + +bail: + return ret; +} + +#define VALID_TS_WR_REG_MASK ((1 << 9) | (1 << 0xB) | (1 << 0xD)) + +/** + * ipath_tempsense_write - write register of temp sensor via I2C + * @dd: the infinipath device + * @regnum: register to write + * @data: data to write + * + * returns 0 for success or < 0 for error + */ +int ipath_tempsense_write(struct ipath_devdata *dd, u8 regnum, u8 data) +{ + int ret; + + if (regnum > 15 || !((1 << regnum) & VALID_TS_WR_REG_MASK)) + return -EINVAL; + + ret = mutex_lock_interruptible(&dd->ipath_eep_lock); + if (!ret) { + ret = ipath_tempsense_internal_write(dd, regnum, data); + mutex_unlock(&dd->ipath_eep_lock); + } + + /* + * There are three possibilities here: + * ret is 0 for success + * ret is -ENXIO or -EINVAL from code in this file + * ret is -EINTR from mutex_lock_interruptible. + */ + return ret; +} diff --git a/drivers/infiniband/hw/ipath/ipath_file_ops.c b/drivers/infiniband/hw/ipath/ipath_file_ops.c index 7e025c8e01b6..1e627aab18bf 100644 --- a/drivers/infiniband/hw/ipath/ipath_file_ops.c +++ b/drivers/infiniband/hw/ipath/ipath_file_ops.c @@ -1,5 +1,5 @@ /* - * Copyright (c) 2006, 2007 QLogic Corporation. All rights reserved. + * Copyright (c) 2006, 2007, 2008 QLogic Corporation. All rights reserved. * Copyright (c) 2003, 2004, 2005, 2006 PathScale, Inc. All rights reserved. * * This software is available to you under a choice of one of two @@ -36,21 +36,28 @@ #include <linux/cdev.h> #include <linux/swap.h> #include <linux/vmalloc.h> +#include <linux/highmem.h> +#include <linux/io.h> +#include <linux/jiffies.h> #include <asm/pgtable.h> #include "ipath_kernel.h" #include "ipath_common.h" +#include "ipath_user_sdma.h" static int ipath_open(struct inode *, struct file *); static int ipath_close(struct inode *, struct file *); static ssize_t ipath_write(struct file *, const char __user *, size_t, loff_t *); +static ssize_t ipath_writev(struct kiocb *, const struct iovec *, + unsigned long , loff_t); static unsigned int ipath_poll(struct file *, struct poll_table_struct *); static int ipath_mmap(struct file *, struct vm_area_struct *); static const struct file_operations ipath_file_ops = { .owner = THIS_MODULE, .write = ipath_write, + .aio_write = ipath_writev, .open = ipath_open, .release = ipath_close, .poll = ipath_poll, @@ -184,6 +191,29 @@ static int ipath_get_base_info(struct file *fp, kinfo->spi_piobufbase = (u64) pd->port_piobufs + dd->ipath_palign * kinfo->spi_piocnt * slave; } + + /* + * Set the PIO avail update threshold to no larger + * than the number of buffers per process. Note that + * we decrease it here, but won't ever increase it. + */ + if (dd->ipath_pioupd_thresh && + kinfo->spi_piocnt < dd->ipath_pioupd_thresh) { + unsigned long flags; + + dd->ipath_pioupd_thresh = kinfo->spi_piocnt; + ipath_dbg("Decreased pio update threshold to %u\n", + dd->ipath_pioupd_thresh); + spin_lock_irqsave(&dd->ipath_sendctrl_lock, flags); + dd->ipath_sendctrl &= ~(INFINIPATH_S_UPDTHRESH_MASK + << INFINIPATH_S_UPDTHRESH_SHIFT); + dd->ipath_sendctrl |= dd->ipath_pioupd_thresh + << INFINIPATH_S_UPDTHRESH_SHIFT; + ipath_write_kreg(dd, dd->ipath_kregs->kr_sendctrl, + dd->ipath_sendctrl); + spin_unlock_irqrestore(&dd->ipath_sendctrl_lock, flags); + } + if (shared) { kinfo->spi_port_uregbase = (u64) dd->ipath_uregbase + dd->ipath_ureg_align * pd->port_port; @@ -219,7 +249,12 @@ static int ipath_get_base_info(struct file *fp, kinfo->spi_pioalign = dd->ipath_palign; kinfo->spi_qpair = IPATH_KD_QP; - kinfo->spi_piosize = dd->ipath_ibmaxlen; + /* + * user mode PIO buffers are always 2KB, even when 4KB can + * be received, and sent via the kernel; this is ibmaxlen + * for 2K MTU. + */ + kinfo->spi_piosize = dd->ipath_piosize2k - 2 * sizeof(u32); kinfo->spi_mtu = dd->ipath_ibmaxlen; /* maxlen, not ibmtu */ kinfo->spi_port = pd->port_port; kinfo->spi_subport = subport_fp(fp); @@ -1598,6 +1633,9 @@ static int try_alloc_port(struct ipath_devdata *dd, int port, port_fp(fp) = pd; pd->port_pid = current->pid; strncpy(pd->port_comm, current->comm, sizeof(pd->port_comm)); + ipath_chg_pioavailkernel(dd, + dd->ipath_pbufsport * (pd->port_port - 1), + dd->ipath_pbufsport, 0); ipath_stats.sps_ports++; ret = 0; } else @@ -1760,7 +1798,7 @@ static int find_shared_port(struct file *fp, for (ndev = 0; ndev < devmax; ndev++) { struct ipath_devdata *dd = ipath_lookup(ndev); - if (!dd) + if (!usable(dd)) continue; for (i = 1; i < dd->ipath_cfgports; i++) { struct ipath_portdata *pd = dd->ipath_pd[i]; @@ -1839,10 +1877,9 @@ static int ipath_assign_port(struct file *fp, if (ipath_compatible_subports(swmajor, swminor) && uinfo->spu_subport_cnt && (ret = find_shared_port(fp, uinfo))) { - mutex_unlock(&ipath_mutex); if (ret > 0) ret = 0; - goto done; + goto done_chk_sdma; } i_minor = iminor(fp->f_path.dentry->d_inode) - IPATH_USER_MINOR_BASE; @@ -1854,6 +1891,21 @@ static int ipath_assign_port(struct file *fp, else ret = find_best_unit(fp, uinfo); +done_chk_sdma: + if (!ret) { + struct ipath_filedata *fd = fp->private_data; + const struct ipath_portdata *pd = fd->pd; + const struct ipath_devdata *dd = pd->port_dd; + + fd->pq = ipath_user_sdma_queue_create(&dd->pcidev->dev, + dd->ipath_unit, + pd->port_port, + fd->subport); + + if (!fd->pq) + ret = -ENOMEM; + } + mutex_unlock(&ipath_mutex); done: @@ -1922,22 +1974,25 @@ static int ipath_do_user_init(struct file *fp, pd->port_hdrqfull_poll = pd->port_hdrqfull; /* - * now enable the port; the tail registers will be written to memory - * by the chip as soon as it sees the write to - * dd->ipath_kregs->kr_rcvctrl. The update only happens on - * transition from 0 to 1, so clear it first, then set it as part of - * enabling the port. This will (very briefly) affect any other - * open ports, but it shouldn't be long enough to be an issue. - * We explictly set the in-memory copy to 0 beforehand, so we don't - * have to wait to be sure the DMA update has happened. + * Now enable the port for receive. + * For chips that are set to DMA the tail register to memory + * when they change (and when the update bit transitions from + * 0 to 1. So for those chips, we turn it off and then back on. + * This will (very briefly) affect any other open ports, but the + * duration is very short, and therefore isn't an issue. We + * explictly set the in-memory tail copy to 0 beforehand, so we + * don't have to wait to be sure the DMA update has happened + * (chip resets head/tail to 0 on transition to enable). */ - if (pd->port_rcvhdrtail_kvaddr) - ipath_clear_rcvhdrtail(pd); set_bit(dd->ipath_r_portenable_shift + pd->port_port, &dd->ipath_rcvctrl); - ipath_write_kreg(dd, dd->ipath_kregs->kr_rcvctrl, + if (!(dd->ipath_flags & IPATH_NODMA_RTAIL)) { + if (pd->port_rcvhdrtail_kvaddr) + ipath_clear_rcvhdrtail(pd); + ipath_write_kreg(dd, dd->ipath_kregs->kr_rcvctrl, dd->ipath_rcvctrl & ~(1ULL << dd->ipath_r_tailupd_shift)); + } ipath_write_kreg(dd, dd->ipath_kregs->kr_rcvctrl, dd->ipath_rcvctrl); /* Notify any waiting slaves */ @@ -1965,14 +2020,15 @@ static void unlock_expected_tids(struct ipath_portdata *pd) ipath_cdbg(VERBOSE, "Port %u unlocking any locked expTID pages\n", pd->port_port); for (i = port_tidbase; i < maxtid; i++) { - if (!dd->ipath_pageshadow[i]) + struct page *ps = dd->ipath_pageshadow[i]; + + if (!ps) continue; + dd->ipath_pageshadow[i] = NULL; pci_unmap_page(dd->pcidev, dd->ipath_physshadow[i], PAGE_SIZE, PCI_DMA_FROMDEVICE); - ipath_release_user_pages_on_close(&dd->ipath_pageshadow[i], - 1); - dd->ipath_pageshadow[i] = NULL; + ipath_release_user_pages_on_close(&ps, 1); cnt++; ipath_stats.sps_pageunlocks++; } @@ -2007,6 +2063,13 @@ static int ipath_close(struct inode *in, struct file *fp) mutex_unlock(&ipath_mutex); goto bail; } + + dd = pd->port_dd; + + /* drain user sdma queue */ + ipath_user_sdma_queue_drain(dd, fd->pq); + ipath_user_sdma_queue_destroy(fd->pq); + if (--pd->port_cnt) { /* * XXX If the master closes the port before the slave(s), @@ -2019,7 +2082,6 @@ static int ipath_close(struct inode *in, struct file *fp) goto bail; } port = pd->port_port; - dd = pd->port_dd; if (pd->port_hdrqfull) { ipath_cdbg(PROC, "%s[%u] had %u rcvhdrqfull errors " @@ -2039,7 +2101,7 @@ static int ipath_close(struct inode *in, struct file *fp) pd->port_rcvnowait = pd->port_pionowait = 0; } if (pd->port_flag) { - ipath_dbg("port %u port_flag still set to 0x%lx\n", + ipath_cdbg(PROC, "port %u port_flag set: 0x%lx\n", pd->port_port, pd->port_flag); pd->port_flag = 0; } @@ -2076,6 +2138,7 @@ static int ipath_close(struct inode *in, struct file *fp) i = dd->ipath_pbufsport * (port - 1); ipath_disarm_piobufs(dd, i, dd->ipath_pbufsport); + ipath_chg_pioavailkernel(dd, i, dd->ipath_pbufsport, 1); dd->ipath_f_clear_tids(dd, pd->port_port); @@ -2140,17 +2203,31 @@ static int ipath_get_slave_info(struct ipath_portdata *pd, return ret; } -static int ipath_force_pio_avail_update(struct ipath_devdata *dd) +static int ipath_sdma_get_inflight(struct ipath_user_sdma_queue *pq, + u32 __user *inflightp) { - unsigned long flags; + const u32 val = ipath_user_sdma_inflight_counter(pq); - spin_lock_irqsave(&dd->ipath_sendctrl_lock, flags); - ipath_write_kreg(dd, dd->ipath_kregs->kr_sendctrl, - dd->ipath_sendctrl & ~INFINIPATH_S_PIOBUFAVAILUPD); - ipath_read_kreg64(dd, dd->ipath_kregs->kr_scratch); - ipath_write_kreg(dd, dd->ipath_kregs->kr_sendctrl, dd->ipath_sendctrl); - ipath_read_kreg64(dd, dd->ipath_kregs->kr_scratch); - spin_unlock_irqrestore(&dd->ipath_sendctrl_lock, flags); + if (put_user(val, inflightp)) + return -EFAULT; + + return 0; +} + +static int ipath_sdma_get_complete(struct ipath_devdata *dd, + struct ipath_user_sdma_queue *pq, + u32 __user *completep) +{ + u32 val; + int err; + + err = ipath_user_sdma_make_progress(dd, pq); + if (err < 0) + return err; + + val = ipath_user_sdma_complete_counter(pq); + if (put_user(val, completep)) + return -EFAULT; return 0; } @@ -2229,6 +2306,16 @@ static ssize_t ipath_write(struct file *fp, const char __user *data, dest = &cmd.cmd.armlaunch_ctrl; src = &ucmd->cmd.armlaunch_ctrl; break; + case IPATH_CMD_SDMA_INFLIGHT: + copy = sizeof(cmd.cmd.sdma_inflight); + dest = &cmd.cmd.sdma_inflight; + src = &ucmd->cmd.sdma_inflight; + break; + case IPATH_CMD_SDMA_COMPLETE: + copy = sizeof(cmd.cmd.sdma_complete); + dest = &cmd.cmd.sdma_complete; + src = &ucmd->cmd.sdma_complete; + break; default: ret = -EINVAL; goto bail; @@ -2299,7 +2386,7 @@ static ssize_t ipath_write(struct file *fp, const char __user *data, cmd.cmd.slave_mask_addr); break; case IPATH_CMD_PIOAVAILUPD: - ret = ipath_force_pio_avail_update(pd->port_dd); + ipath_force_pio_avail_update(pd->port_dd); break; case IPATH_CMD_POLL_TYPE: pd->poll_type = cmd.cmd.poll_type; @@ -2310,6 +2397,17 @@ static ssize_t ipath_write(struct file *fp, const char __user *data, else ipath_disable_armlaunch(pd->port_dd); break; + case IPATH_CMD_SDMA_INFLIGHT: + ret = ipath_sdma_get_inflight(user_sdma_queue_fp(fp), + (u32 __user *) (unsigned long) + cmd.cmd.sdma_inflight); + break; + case IPATH_CMD_SDMA_COMPLETE: + ret = ipath_sdma_get_complete(pd->port_dd, + user_sdma_queue_fp(fp), + (u32 __user *) (unsigned long) + cmd.cmd.sdma_complete); + break; } if (ret >= 0) @@ -2319,6 +2417,20 @@ bail: return ret; } +static ssize_t ipath_writev(struct kiocb *iocb, const struct iovec *iov, + unsigned long dim, loff_t off) +{ + struct file *filp = iocb->ki_filp; + struct ipath_filedata *fp = filp->private_data; + struct ipath_portdata *pd = port_fp(filp); + struct ipath_user_sdma_queue *pq = fp->pq; + + if (!dim) + return -EINVAL; + + return ipath_user_sdma_writev(pd->port_dd, pq, iov, dim); +} + static struct class *ipath_class; static int init_cdev(int minor, char *name, const struct file_operations *fops, diff --git a/drivers/infiniband/hw/ipath/ipath_iba6110.c b/drivers/infiniband/hw/ipath/ipath_iba6110.c index 9e2ced3cdc5e..02831ad070b8 100644 --- a/drivers/infiniband/hw/ipath/ipath_iba6110.c +++ b/drivers/infiniband/hw/ipath/ipath_iba6110.c @@ -40,6 +40,7 @@ #include <linux/pci.h> #include <linux/delay.h> #include <linux/htirq.h> +#include <rdma/ib_verbs.h> #include "ipath_kernel.h" #include "ipath_registers.h" @@ -305,7 +306,9 @@ static const struct ipath_cregs ipath_ht_cregs = { /* kr_intstatus, kr_intclear, kr_intmask bits */ #define INFINIPATH_I_RCVURG_MASK ((1U<<9)-1) +#define INFINIPATH_I_RCVURG_SHIFT 0 #define INFINIPATH_I_RCVAVAIL_MASK ((1U<<9)-1) +#define INFINIPATH_I_RCVAVAIL_SHIFT 12 /* kr_hwerrclear, kr_hwerrmask, kr_hwerrstatus, bits */ #define INFINIPATH_HWE_HTCMEMPARITYERR_SHIFT 0 @@ -476,7 +479,13 @@ static const struct ipath_hwerror_msgs ipath_6110_hwerror_msgs[] = { #define RXE_EAGER_PARITY (INFINIPATH_HWE_RXEMEMPARITYERR_EAGERTID \ << INFINIPATH_HWE_RXEMEMPARITYERR_SHIFT) -static int ipath_ht_txe_recover(struct ipath_devdata *); +static void ipath_ht_txe_recover(struct ipath_devdata *dd) +{ + ++ipath_stats.sps_txeparity; + dev_info(&dd->pcidev->dev, + "Recovering from TXE PIO parity error\n"); +} + /** * ipath_ht_handle_hwerrors - display hardware errors. @@ -557,11 +566,11 @@ static void ipath_ht_handle_hwerrors(struct ipath_devdata *dd, char *msg, * occur if a processor speculative read is done to the PIO * buffer while we are sending a packet, for example. */ - if ((hwerrs & TXE_PIO_PARITY) && ipath_ht_txe_recover(dd)) + if (hwerrs & TXE_PIO_PARITY) { + ipath_ht_txe_recover(dd); hwerrs &= ~TXE_PIO_PARITY; - if (hwerrs & RXE_EAGER_PARITY) - ipath_dev_err(dd, "RXE parity, Eager TID error is not " - "recoverable\n"); + } + if (!hwerrs) { ipath_dbg("Clearing freezemode on ignored or " "recovered hardware error\n"); @@ -735,11 +744,10 @@ static int ipath_ht_boardname(struct ipath_devdata *dd, char *name, */ dd->ipath_flags |= IPATH_32BITCOUNTERS; dd->ipath_flags |= IPATH_GPIO_INTR; - if (dd->ipath_htspeed != 800) + if (dd->ipath_lbus_speed != 800) ipath_dev_err(dd, "Incorrectly configured for HT @ %uMHz\n", - dd->ipath_htspeed); - ret = 0; + dd->ipath_lbus_speed); /* * set here, not in ipath_init_*_funcs because we have to do @@ -839,7 +847,7 @@ static void slave_or_pri_blk(struct ipath_devdata *dd, struct pci_dev *pdev, /* * now write them back to clear the error. */ - pci_write_config_byte(pdev, link_off, + pci_write_config_word(pdev, link_off, linkctrl & (0xf << 8)); } } @@ -904,7 +912,7 @@ static void slave_or_pri_blk(struct ipath_devdata *dd, struct pci_dev *pdev, break; } - dd->ipath_htwidth = width; + dd->ipath_lbus_width = width; if (linkwidth != 0x11) { ipath_dev_err(dd, "Not configured for 16 bit HT " @@ -952,8 +960,13 @@ static void slave_or_pri_blk(struct ipath_devdata *dd, struct pci_dev *pdev, speed = 200; break; } - dd->ipath_htspeed = speed; + dd->ipath_lbus_speed = speed; } + + snprintf(dd->ipath_lbus_info, sizeof(dd->ipath_lbus_info), + "HyperTransport,%uMHz,x%u\n", + dd->ipath_lbus_speed, + dd->ipath_lbus_width); } static int ipath_ht_intconfig(struct ipath_devdata *dd) @@ -1653,22 +1666,6 @@ static int ipath_ht_early_init(struct ipath_devdata *dd) } -static int ipath_ht_txe_recover(struct ipath_devdata *dd) -{ - int cnt = ++ipath_stats.sps_txeparity; - if (cnt >= IPATH_MAX_PARITY_ATTEMPTS) { - if (cnt == IPATH_MAX_PARITY_ATTEMPTS) - ipath_dev_err(dd, - "Too many attempts to recover from " - "TXE parity, giving up\n"); - return 0; - } - dev_info(&dd->pcidev->dev, - "Recovering from TXE PIO parity error\n"); - return 1; -} - - /** * ipath_init_ht_get_base_info - set chip-specific flags for user code * @dd: the infinipath device diff --git a/drivers/infiniband/hw/ipath/ipath_iba6120.c b/drivers/infiniband/hw/ipath/ipath_iba6120.c index c7a2f50824c0..421cc2af891f 100644 --- a/drivers/infiniband/hw/ipath/ipath_iba6120.c +++ b/drivers/infiniband/hw/ipath/ipath_iba6120.c @@ -1,5 +1,5 @@ /* - * Copyright (c) 2006, 2007 QLogic Corporation. All rights reserved. + * Copyright (c) 2006, 2007, 2008 QLogic Corporation. All rights reserved. * Copyright (c) 2003, 2004, 2005, 2006 PathScale, Inc. All rights reserved. * * This software is available to you under a choice of one of two @@ -38,7 +38,7 @@ #include <linux/interrupt.h> #include <linux/pci.h> #include <linux/delay.h> - +#include <rdma/ib_verbs.h> #include "ipath_kernel.h" #include "ipath_registers.h" @@ -311,9 +311,14 @@ static const struct ipath_cregs ipath_pe_cregs = { .cr_ibsymbolerrcnt = IPATH_CREG_OFFSET(IBSymbolErrCnt) }; +/* kr_control bits */ +#define INFINIPATH_C_RESET 1U + /* kr_intstatus, kr_intclear, kr_intmask bits */ #define INFINIPATH_I_RCVURG_MASK ((1U<<5)-1) +#define INFINIPATH_I_RCVURG_SHIFT 0 #define INFINIPATH_I_RCVAVAIL_MASK ((1U<<5)-1) +#define INFINIPATH_I_RCVAVAIL_SHIFT 12 /* kr_hwerrclear, kr_hwerrmask, kr_hwerrstatus, bits */ #define INFINIPATH_HWE_PCIEMEMPARITYERR_MASK 0x000000000000003fULL @@ -338,6 +343,9 @@ static const struct ipath_cregs ipath_pe_cregs = { #define INFINIPATH_EXTS_MEMBIST_ENDTEST 0x0000000000004000 #define INFINIPATH_EXTS_MEMBIST_FOUND 0x0000000000008000 +/* kr_xgxsconfig bits */ +#define INFINIPATH_XGXS_RESET 0x5ULL + #define _IPATH_GPIO_SDA_NUM 1 #define _IPATH_GPIO_SCL_NUM 0 @@ -346,6 +354,16 @@ static const struct ipath_cregs ipath_pe_cregs = { #define IPATH_GPIO_SCL (1ULL << \ (_IPATH_GPIO_SCL_NUM+INFINIPATH_EXTC_GPIOOE_SHIFT)) +#define INFINIPATH_RT_BUFSIZE_MASK 0xe0000000ULL +#define INFINIPATH_RT_BUFSIZE_SHIFTVAL(tid) \ + ((((tid) & INFINIPATH_RT_BUFSIZE_MASK) >> 29) + 11 - 1) +#define INFINIPATH_RT_BUFSIZE(tid) (1 << INFINIPATH_RT_BUFSIZE_SHIFTVAL(tid)) +#define INFINIPATH_RT_IS_VALID(tid) \ + (((tid) & INFINIPATH_RT_BUFSIZE_MASK) && \ + ((((tid) & INFINIPATH_RT_BUFSIZE_MASK) != INFINIPATH_RT_BUFSIZE_MASK))) +#define INFINIPATH_RT_ADDR_MASK 0x1FFFFFFFULL /* 29 bits valid */ +#define INFINIPATH_RT_ADDR_SHIFT 10 + #define INFINIPATH_R_INTRAVAIL_SHIFT 16 #define INFINIPATH_R_TAILUPD_SHIFT 31 @@ -372,6 +390,8 @@ static const struct ipath_hwerror_msgs ipath_6120_hwerror_msgs[] = { #define TXE_PIO_PARITY ((INFINIPATH_HWE_TXEMEMPARITYERR_PIOBUF | \ INFINIPATH_HWE_TXEMEMPARITYERR_PIOPBC) \ << INFINIPATH_HWE_TXEMEMPARITYERR_SHIFT) +#define RXE_EAGER_PARITY (INFINIPATH_HWE_RXEMEMPARITYERR_EAGERTID \ + << INFINIPATH_HWE_RXEMEMPARITYERR_SHIFT) static void ipath_pe_put_tid_2(struct ipath_devdata *, u64 __iomem *, u32, unsigned long); @@ -450,10 +470,8 @@ static void ipath_pe_handle_hwerrors(struct ipath_devdata *dd, char *msg, * make sure we get this much out, unless told to be quiet, * or it's occurred within the last 5 seconds */ - if ((hwerrs & ~(dd->ipath_lasthwerror | - ((INFINIPATH_HWE_TXEMEMPARITYERR_PIOBUF | - INFINIPATH_HWE_TXEMEMPARITYERR_PIOPBC) - << INFINIPATH_HWE_TXEMEMPARITYERR_SHIFT))) || + if ((hwerrs & ~(dd->ipath_lasthwerror | TXE_PIO_PARITY | + RXE_EAGER_PARITY)) || (ipath_debug & __IPATH_VERBDBG)) dev_info(&dd->pcidev->dev, "Hardware error: hwerr=0x%llx " "(cleared)\n", (unsigned long long) hwerrs); @@ -465,7 +483,7 @@ static void ipath_pe_handle_hwerrors(struct ipath_devdata *dd, char *msg, (hwerrs & ~dd->ipath_hwe_bitsextant)); ctrl = ipath_read_kreg32(dd, dd->ipath_kregs->kr_control); - if (ctrl & INFINIPATH_C_FREEZEMODE) { + if ((ctrl & INFINIPATH_C_FREEZEMODE) && !ipath_diag_inuse) { /* * parity errors in send memory are recoverable, * just cancel the send (if indicated in * sendbuffererror), @@ -540,12 +558,40 @@ static void ipath_pe_handle_hwerrors(struct ipath_devdata *dd, char *msg, dd->ipath_hwerrmask); } - if (*msg) + if (hwerrs) { + /* + * if any set that we aren't ignoring; only + * make the complaint once, in case it's stuck + * or recurring, and we get here multiple + * times. + */ ipath_dev_err(dd, "%s hardware error\n", msg); - if (isfatal && !ipath_diag_inuse && dd->ipath_freezemsg) { + if (dd->ipath_flags & IPATH_INITTED) { + ipath_set_linkstate(dd, IPATH_IB_LINKDOWN); + ipath_setup_pe_setextled(dd, + INFINIPATH_IBCS_L_STATE_DOWN, + INFINIPATH_IBCS_LT_STATE_DISABLED); + ipath_dev_err(dd, "Fatal Hardware Error (freeze " + "mode), no longer usable, SN %.16s\n", + dd->ipath_serial); + isfatal = 1; + } + *dd->ipath_statusp &= ~IPATH_STATUS_IB_READY; + /* mark as having had error */ + *dd->ipath_statusp |= IPATH_STATUS_HWERROR; /* - * for /sys status file ; if no trailing } is copied, we'll - * know it was truncated. + * mark as not usable, at a minimum until driver + * is reloaded, probably until reboot, since no + * other reset is possible. + */ + dd->ipath_flags &= ~IPATH_INITTED; + } else + *msg = 0; /* recovered from all of them */ + + if (isfatal && !ipath_diag_inuse && dd->ipath_freezemsg && msg) { + /* + * for /sys status file ; if no trailing brace is copied, + * we'll know it was truncated. */ snprintf(dd->ipath_freezemsg, dd->ipath_freezelen, "{%s}", msg); @@ -610,7 +656,6 @@ static int ipath_pe_boardname(struct ipath_devdata *dd, char *name, dd->ipath_f_put_tid = ipath_pe_put_tid_2; } - /* * set here, not in ipath_init_*_funcs because we have to do * it after we can read chip registers. @@ -838,7 +883,7 @@ static void ipath_setup_pe_setextled(struct ipath_devdata *dd, u64 lst, extctl = dd->ipath_extctrl & ~(INFINIPATH_EXTC_LED1PRIPORT_ON | INFINIPATH_EXTC_LED2PRIPORT_ON); - if (ltst & INFINIPATH_IBCS_LT_STATE_LINKUP) + if (ltst == INFINIPATH_IBCS_LT_STATE_LINKUP) extctl |= INFINIPATH_EXTC_LED2PRIPORT_ON; if (lst == INFINIPATH_IBCS_L_STATE_ACTIVE) extctl |= INFINIPATH_EXTC_LED1PRIPORT_ON; @@ -863,6 +908,62 @@ static void ipath_setup_pe_cleanup(struct ipath_devdata *dd) pci_disable_msi(dd->pcidev); } +static void ipath_6120_pcie_params(struct ipath_devdata *dd) +{ + u16 linkstat, speed; + int pos; + + pos = pci_find_capability(dd->pcidev, PCI_CAP_ID_EXP); + if (!pos) { + ipath_dev_err(dd, "Can't find PCI Express capability!\n"); + goto bail; + } + + pci_read_config_word(dd->pcidev, pos + PCI_EXP_LNKSTA, + &linkstat); + /* + * speed is bits 0-4, linkwidth is bits 4-8 + * no defines for them in headers + */ + speed = linkstat & 0xf; + linkstat >>= 4; + linkstat &= 0x1f; + dd->ipath_lbus_width = linkstat; + + switch (speed) { + case 1: + dd->ipath_lbus_speed = 2500; /* Gen1, 2.5GHz */ + break; + case 2: + dd->ipath_lbus_speed = 5000; /* Gen1, 5GHz */ + break; + default: /* not defined, assume gen1 */ + dd->ipath_lbus_speed = 2500; + break; + } + + if (linkstat < 8) + ipath_dev_err(dd, + "PCIe width %u (x8 HCA), performance reduced\n", + linkstat); + else + ipath_cdbg(VERBOSE, "PCIe speed %u width %u (x8 HCA)\n", + dd->ipath_lbus_speed, linkstat); + + if (speed != 1) + ipath_dev_err(dd, + "PCIe linkspeed %u is incorrect; " + "should be 1 (2500)!\n", speed); +bail: + /* fill in string, even on errors */ + snprintf(dd->ipath_lbus_info, sizeof(dd->ipath_lbus_info), + "PCIe,%uMHz,x%u\n", + dd->ipath_lbus_speed, + dd->ipath_lbus_width); + + return; +} + /** * ipath_setup_pe_config - setup PCIe config related stuff * @dd: the infinipath device @@ -920,19 +1021,8 @@ static int ipath_setup_pe_config(struct ipath_devdata *dd, } else ipath_dev_err(dd, "Can't find MSI capability, " "can't save MSI settings for reset\n"); - if ((pos = pci_find_capability(dd->pcidev, PCI_CAP_ID_EXP))) { - u16 linkstat; - pci_read_config_word(dd->pcidev, pos + PCI_EXP_LNKSTA, - &linkstat); - linkstat >>= 4; - linkstat &= 0x1f; - if (linkstat != 8) - ipath_dev_err(dd, "PCIe width %u, " - "performance reduced\n", linkstat); - } - else - ipath_dev_err(dd, "Can't find PCI Express " - "capability!\n"); + + ipath_6120_pcie_params(dd); dd->ipath_link_width_supported = IB_WIDTH_1X | IB_WIDTH_4X; dd->ipath_link_speed_supported = IPATH_IB_SDR; @@ -1065,10 +1155,7 @@ static void ipath_init_pe_variables(struct ipath_devdata *dd) INFINIPATH_HWE_RXEMEMPARITYERR_MASK << INFINIPATH_HWE_RXEMEMPARITYERR_SHIFT; - dd->ipath_eep_st_masks[2].errs_to_log = - INFINIPATH_E_INVALIDADDR | INFINIPATH_E_RESET; - - + dd->ipath_eep_st_masks[2].errs_to_log = INFINIPATH_E_RESET; dd->delay_mult = 2; /* SDR, 4X, can't change */ } @@ -1142,6 +1229,9 @@ static int ipath_setup_pe_reset(struct ipath_devdata *dd) u64 val; int i; int ret; + u16 cmdval; + + pci_read_config_word(dd->pcidev, PCI_COMMAND, &cmdval); /* Use ERROR so it shows up in logs, etc. */ ipath_dev_err(dd, "Resetting InfiniPath unit %u\n", dd->ipath_unit); @@ -1169,10 +1259,14 @@ static int ipath_setup_pe_reset(struct ipath_devdata *dd) ipath_dev_err(dd, "rewrite of BAR1 failed: %d\n", r); /* now re-enable memory access */ + pci_write_config_word(dd->pcidev, PCI_COMMAND, cmdval); if ((r = pci_enable_device(dd->pcidev))) ipath_dev_err(dd, "pci_enable_device failed after " "reset: %d\n", r); - /* whether it worked or not, mark as present, again */ + /* + * whether it fully enabled or not, mark as present, + * again (but not INITTED) + */ dd->ipath_flags |= IPATH_PRESENT; val = ipath_read_kreg64(dd, dd->ipath_kregs->kr_revision); if (val == dd->ipath_revision) { @@ -1190,6 +1284,8 @@ static int ipath_setup_pe_reset(struct ipath_devdata *dd) ret = 0; /* failed */ bail: + if (ret) + ipath_6120_pcie_params(dd); return ret; } @@ -1209,16 +1305,21 @@ static void ipath_pe_put_tid(struct ipath_devdata *dd, u64 __iomem *tidptr, { u32 __iomem *tidp32 = (u32 __iomem *)tidptr; unsigned long flags = 0; /* keep gcc quiet */ + int tidx; + spinlock_t *tidlockp; + + if (!dd->ipath_kregbase) + return; if (pa != dd->ipath_tidinvalid) { if (pa & ((1U << 11) - 1)) { dev_info(&dd->pcidev->dev, "BUG: physaddr %lx " - "not 4KB aligned!\n", pa); + "not 2KB aligned!\n", pa); return; } pa >>= 11; /* paranoia check */ - if (pa & (7<<29)) + if (pa & ~INFINIPATH_RT_ADDR_MASK) ipath_dev_err(dd, "BUG: Physical page address 0x%lx " "has bits set in 31-29\n", pa); @@ -1238,14 +1339,22 @@ static void ipath_pe_put_tid(struct ipath_devdata *dd, u64 __iomem *tidptr, * call can be done from interrupt level for the port 0 eager TIDs, * so we have to use irqsave locks. */ - spin_lock_irqsave(&dd->ipath_tid_lock, flags); + /* + * Assumes tidptr always > ipath_egrtidbase + * if type == RCVHQ_RCV_TYPE_EAGER. + */ + tidx = tidptr - dd->ipath_egrtidbase; + + tidlockp = (type == RCVHQ_RCV_TYPE_EAGER && tidx < dd->ipath_rcvegrcnt) + ? &dd->ipath_kernel_tid_lock : &dd->ipath_user_tid_lock; + spin_lock_irqsave(tidlockp, flags); ipath_write_kreg(dd, dd->ipath_kregs->kr_scratch, 0xfeeddeaf); - if (dd->ipath_kregbase) - writel(pa, tidp32); + writel(pa, tidp32); ipath_write_kreg(dd, dd->ipath_kregs->kr_scratch, 0xdeadbeef); mmiowb(); - spin_unlock_irqrestore(&dd->ipath_tid_lock, flags); + spin_unlock_irqrestore(tidlockp, flags); } + /** * ipath_pe_put_tid_2 - write a TID in chip, Revision 2 or higher * @dd: the infinipath device @@ -1261,6 +1370,10 @@ static void ipath_pe_put_tid_2(struct ipath_devdata *dd, u64 __iomem *tidptr, u32 type, unsigned long pa) { u32 __iomem *tidp32 = (u32 __iomem *)tidptr; + u32 tidx; + + if (!dd->ipath_kregbase) + return; if (pa != dd->ipath_tidinvalid) { if (pa & ((1U << 11) - 1)) { @@ -1270,7 +1383,7 @@ static void ipath_pe_put_tid_2(struct ipath_devdata *dd, u64 __iomem *tidptr, } pa >>= 11; /* paranoia check */ - if (pa & (7<<29)) + if (pa & ~INFINIPATH_RT_ADDR_MASK) ipath_dev_err(dd, "BUG: Physical page address 0x%lx " "has bits set in 31-29\n", pa); @@ -1280,8 +1393,8 @@ static void ipath_pe_put_tid_2(struct ipath_devdata *dd, u64 __iomem *tidptr, else /* for now, always full 4KB page */ pa |= 2 << 29; } - if (dd->ipath_kregbase) - writel(pa, tidp32); + tidx = tidptr - dd->ipath_egrtidbase; + writel(pa, tidp32); mmiowb(); } @@ -1379,17 +1492,13 @@ static int ipath_pe_early_init(struct ipath_devdata *dd) dd->ipath_egrtidbase = (u64 __iomem *) ((char __iomem *) dd->ipath_kregbase + dd->ipath_rcvegrbase); - /* - * To truly support a 4KB MTU (for usermode), we need to - * bump this to a larger value. For now, we use them for - * the kernel only. - */ - dd->ipath_rcvegrbufsize = 2048; + dd->ipath_rcvegrbufsize = ipath_mtu4096 ? 4096 : 2048; /* * the min() check here is currently a nop, but it may not always * be, depending on just how we do ipath_rcvegrbufsize */ - dd->ipath_ibmaxlen = min(dd->ipath_piosize2k, + dd->ipath_ibmaxlen = min(ipath_mtu4096 ? dd->ipath_piosize4k : + dd->ipath_piosize2k, dd->ipath_rcvegrbufsize + (dd->ipath_rcvhdrentsize << 2)); dd->ipath_init_ibmaxlen = dd->ipath_ibmaxlen; diff --git a/drivers/infiniband/hw/ipath/ipath_iba7220.c b/drivers/infiniband/hw/ipath/ipath_iba7220.c new file mode 100644 index 000000000000..1b2de2cfb69b --- /dev/null +++ b/drivers/infiniband/hw/ipath/ipath_iba7220.c @@ -0,0 +1,2571 @@ +/* + * Copyright (c) 2006, 2007, 2008 QLogic Corporation. All rights reserved. + * Copyright (c) 2003, 2004, 2005, 2006 PathScale, Inc. All rights reserved. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * OpenIB.org BSD license below: + * + * Redistribution and use in source and binary forms, with or + * without modification, are permitted provided that the following + * conditions are met: + * + * - Redistributions of source code must retain the above + * copyright notice, this list of conditions and the following + * disclaimer. + * + * - Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials + * provided with the distribution. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + */ +/* + * This file contains all of the code that is specific to the + * InfiniPath 7220 chip (except that specific to the SerDes) + */ + +#include <linux/interrupt.h> +#include <linux/pci.h> +#include <linux/delay.h> +#include <linux/io.h> +#include <rdma/ib_verbs.h> + +#include "ipath_kernel.h" +#include "ipath_registers.h" +#include "ipath_7220.h" + +static void ipath_setup_7220_setextled(struct ipath_devdata *, u64, u64); + +static unsigned ipath_compat_ddr_negotiate = 1; + +module_param_named(compat_ddr_negotiate, ipath_compat_ddr_negotiate, uint, + S_IWUSR | S_IRUGO); +MODULE_PARM_DESC(compat_ddr_negotiate, + "Attempt pre-IBTA 1.2 DDR speed negotiation"); + +static unsigned ipath_sdma_fetch_arb = 1; +module_param_named(fetch_arb, ipath_sdma_fetch_arb, uint, S_IRUGO); +MODULE_PARM_DESC(fetch_arb, "IBA7220: change SDMA descriptor arbitration"); + +/* + * This file contains almost all the chip-specific register information and + * access functions for the QLogic InfiniPath 7220 PCI-Express chip, with the + * exception of SerDes support, which in in ipath_sd7220.c. + * + * This lists the InfiniPath registers, in the actual chip layout. + * This structure should never be directly accessed. + */ +struct _infinipath_do_not_use_kernel_regs { + unsigned long long Revision; + unsigned long long Control; + unsigned long long PageAlign; + unsigned long long PortCnt; + unsigned long long DebugPortSelect; + unsigned long long DebugSigsIntSel; /* was Reserved0;*/ + unsigned long long SendRegBase; + unsigned long long UserRegBase; + unsigned long long CounterRegBase; + unsigned long long Scratch; + unsigned long long EEPROMAddrCmd; /* was Reserved1; */ + unsigned long long EEPROMData; /* was Reserved2; */ + unsigned long long IntBlocked; + unsigned long long IntMask; + unsigned long long IntStatus; + unsigned long long IntClear; + unsigned long long ErrorMask; + unsigned long long ErrorStatus; + unsigned long long ErrorClear; + unsigned long long HwErrMask; + unsigned long long HwErrStatus; + unsigned long long HwErrClear; + unsigned long long HwDiagCtrl; + unsigned long long MDIO; + unsigned long long IBCStatus; + unsigned long long IBCCtrl; + unsigned long long ExtStatus; + unsigned long long ExtCtrl; + unsigned long long GPIOOut; + unsigned long long GPIOMask; + unsigned long long GPIOStatus; + unsigned long long GPIOClear; + unsigned long long RcvCtrl; + unsigned long long RcvBTHQP; + unsigned long long RcvHdrSize; + unsigned long long RcvHdrCnt; + unsigned long long RcvHdrEntSize; + unsigned long long RcvTIDBase; + unsigned long long RcvTIDCnt; + unsigned long long RcvEgrBase; + unsigned long long RcvEgrCnt; + unsigned long long RcvBufBase; + unsigned long long RcvBufSize; + unsigned long long RxIntMemBase; + unsigned long long RxIntMemSize; + unsigned long long RcvPartitionKey; + unsigned long long RcvQPMulticastPort; + unsigned long long RcvPktLEDCnt; + unsigned long long IBCDDRCtrl; + unsigned long long HRTBT_GUID; + unsigned long long IB_SDTEST_IF_TX; + unsigned long long IB_SDTEST_IF_RX; + unsigned long long IBCDDRCtrl2; + unsigned long long IBCDDRStatus; + unsigned long long JIntReload; + unsigned long long IBNCModeCtrl; + unsigned long long SendCtrl; + unsigned long long SendBufBase; + unsigned long long SendBufSize; + unsigned long long SendBufCnt; + unsigned long long SendAvailAddr; + unsigned long long TxIntMemBase; + unsigned long long TxIntMemSize; + unsigned long long SendDmaBase; + unsigned long long SendDmaLenGen; + unsigned long long SendDmaTail; + unsigned long long SendDmaHead; + unsigned long long SendDmaHeadAddr; + unsigned long long SendDmaBufMask0; + unsigned long long SendDmaBufMask1; + unsigned long long SendDmaBufMask2; + unsigned long long SendDmaStatus; + unsigned long long SendBufferError; + unsigned long long SendBufferErrorCONT1; + unsigned long long SendBufErr2; /* was Reserved6SBE[0/6] */ + unsigned long long Reserved6L[2]; + unsigned long long AvailUpdCount; + unsigned long long RcvHdrAddr0; + unsigned long long RcvHdrAddrs[16]; /* Why enumerate? */ + unsigned long long Reserved7hdtl; /* Align next to 300 */ + unsigned long long RcvHdrTailAddr0; /* 300, like others */ + unsigned long long RcvHdrTailAddrs[16]; + unsigned long long Reserved9SW[7]; /* was [8]; we have 17 ports */ + unsigned long long IbsdEpbAccCtl; /* IB Serdes EPB access control */ + unsigned long long IbsdEpbTransReg; /* IB Serdes EPB Transaction */ + unsigned long long Reserved10sds; /* was SerdesStatus on */ + unsigned long long XGXSConfig; + unsigned long long IBSerDesCtrl; /* Was IBPLLCfg on Monty */ + unsigned long long EEPCtlStat; /* for "boot" EEPROM/FLASH */ + unsigned long long EEPAddrCmd; + unsigned long long EEPData; + unsigned long long PcieEpbAccCtl; + unsigned long long PcieEpbTransCtl; + unsigned long long EfuseCtl; /* E-Fuse control */ + unsigned long long EfuseData[4]; + unsigned long long ProcMon; + /* this chip moves following two from previous 200, 208 */ + unsigned long long PCIeRBufTestReg0; + unsigned long long PCIeRBufTestReg1; + /* added for this chip */ + unsigned long long PCIeRBufTestReg2; + unsigned long long PCIeRBufTestReg3; + /* added for this chip, debug only */ + unsigned long long SPC_JTAG_ACCESS_REG; + unsigned long long LAControlReg; + unsigned long long GPIODebugSelReg; + unsigned long long DebugPortValueReg; + /* added for this chip, DMA */ + unsigned long long SendDmaBufUsed[3]; + unsigned long long SendDmaReqTagUsed; + /* + * added for this chip, EFUSE: note that these program 64-bit + * words 2 and 3 */ + unsigned long long efuse_pgm_data[2]; + unsigned long long Reserved11LAalign[10]; /* Skip 4B0..4F8 */ + /* we have 30 regs for DDS and RXEQ in IB SERDES */ + unsigned long long SerDesDDSRXEQ[30]; + unsigned long long Reserved12LAalign[2]; /* Skip 5F0, 5F8 */ + /* added for LA debug support */ + unsigned long long LAMemory[32]; +}; + +struct _infinipath_do_not_use_counters { + __u64 LBIntCnt; + __u64 LBFlowStallCnt; + __u64 TxSDmaDescCnt; /* was Reserved1 */ + __u64 TxUnsupVLErrCnt; + __u64 TxDataPktCnt; + __u64 TxFlowPktCnt; + __u64 TxDwordCnt; + __u64 TxLenErrCnt; + __u64 TxMaxMinLenErrCnt; + __u64 TxUnderrunCnt; + __u64 TxFlowStallCnt; + __u64 TxDroppedPktCnt; + __u64 RxDroppedPktCnt; + __u64 RxDataPktCnt; + __u64 RxFlowPktCnt; + __u64 RxDwordCnt; + __u64 RxLenErrCnt; + __u64 RxMaxMinLenErrCnt; + __u64 RxICRCErrCnt; + __u64 RxVCRCErrCnt; + __u64 RxFlowCtrlErrCnt; + __u64 RxBadFormatCnt; + __u64 RxLinkProblemCnt; + __u64 RxEBPCnt; + __u64 RxLPCRCErrCnt; + __u64 RxBufOvflCnt; + __u64 RxTIDFullErrCnt; + __u64 RxTIDValidErrCnt; + __u64 RxPKeyMismatchCnt; + __u64 RxP0HdrEgrOvflCnt; + __u64 RxP1HdrEgrOvflCnt; + __u64 RxP2HdrEgrOvflCnt; + __u64 RxP3HdrEgrOvflCnt; + __u64 RxP4HdrEgrOvflCnt; + __u64 RxP5HdrEgrOvflCnt; + __u64 RxP6HdrEgrOvflCnt; + __u64 RxP7HdrEgrOvflCnt; + __u64 RxP8HdrEgrOvflCnt; + __u64 RxP9HdrEgrOvflCnt; /* was Reserved6 */ + __u64 RxP10HdrEgrOvflCnt; /* was Reserved7 */ + __u64 RxP11HdrEgrOvflCnt; /* new for IBA7220 */ + __u64 RxP12HdrEgrOvflCnt; /* new for IBA7220 */ + __u64 RxP13HdrEgrOvflCnt; /* new for IBA7220 */ + __u64 RxP14HdrEgrOvflCnt; /* new for IBA7220 */ + __u64 RxP15HdrEgrOvflCnt; /* new for IBA7220 */ + __u64 RxP16HdrEgrOvflCnt; /* new for IBA7220 */ + __u64 IBStatusChangeCnt; + __u64 IBLinkErrRecoveryCnt; + __u64 IBLinkDownedCnt; + __u64 IBSymbolErrCnt; + /* The following are new for IBA7220 */ + __u64 RxVL15DroppedPktCnt; + __u64 RxOtherLocalPhyErrCnt; + __u64 PcieRetryBufDiagQwordCnt; + __u64 ExcessBufferOvflCnt; + __u64 LocalLinkIntegrityErrCnt; + __u64 RxVlErrCnt; + __u64 RxDlidFltrCnt; + __u64 Reserved8[7]; + __u64 PSStat; + __u64 PSStart; + __u64 PSInterval; + __u64 PSRcvDataCount; + __u64 PSRcvPktsCount; + __u64 PSXmitDataCount; + __u64 PSXmitPktsCount; + __u64 PSXmitWaitCount; +}; + +#define IPATH_KREG_OFFSET(field) (offsetof( \ + struct _infinipath_do_not_use_kernel_regs, field) / sizeof(u64)) +#define IPATH_CREG_OFFSET(field) (offsetof( \ + struct _infinipath_do_not_use_counters, field) / sizeof(u64)) + +static const struct ipath_kregs ipath_7220_kregs = { + .kr_control = IPATH_KREG_OFFSET(Control), + .kr_counterregbase = IPATH_KREG_OFFSET(CounterRegBase), + .kr_debugportselect = IPATH_KREG_OFFSET(DebugPortSelect), + .kr_errorclear = IPATH_KREG_OFFSET(ErrorClear), + .kr_errormask = IPATH_KREG_OFFSET(ErrorMask), + .kr_errorstatus = IPATH_KREG_OFFSET(ErrorStatus), + .kr_extctrl = IPATH_KREG_OFFSET(ExtCtrl), + .kr_extstatus = IPATH_KREG_OFFSET(ExtStatus), + .kr_gpio_clear = IPATH_KREG_OFFSET(GPIOClear), + .kr_gpio_mask = IPATH_KREG_OFFSET(GPIOMask), + .kr_gpio_out = IPATH_KREG_OFFSET(GPIOOut), + .kr_gpio_status = IPATH_KREG_OFFSET(GPIOStatus), + .kr_hwdiagctrl = IPATH_KREG_OFFSET(HwDiagCtrl), + .kr_hwerrclear = IPATH_KREG_OFFSET(HwErrClear), + .kr_hwerrmask = IPATH_KREG_OFFSET(HwErrMask), + .kr_hwerrstatus = IPATH_KREG_OFFSET(HwErrStatus), + .kr_ibcctrl = IPATH_KREG_OFFSET(IBCCtrl), + .kr_ibcstatus = IPATH_KREG_OFFSET(IBCStatus), + .kr_intblocked = IPATH_KREG_OFFSET(IntBlocked), + .kr_intclear = IPATH_KREG_OFFSET(IntClear), + .kr_intmask = IPATH_KREG_OFFSET(IntMask), + .kr_intstatus = IPATH_KREG_OFFSET(IntStatus), + .kr_mdio = IPATH_KREG_OFFSET(MDIO), + .kr_pagealign = IPATH_KREG_OFFSET(PageAlign), + .kr_partitionkey = IPATH_KREG_OFFSET(RcvPartitionKey), + .kr_portcnt = IPATH_KREG_OFFSET(PortCnt), + .kr_rcvbthqp = IPATH_KREG_OFFSET(RcvBTHQP), + .kr_rcvbufbase = IPATH_KREG_OFFSET(RcvBufBase), + .kr_rcvbufsize = IPATH_KREG_OFFSET(RcvBufSize), + .kr_rcvctrl = IPATH_KREG_OFFSET(RcvCtrl), + .kr_rcvegrbase = IPATH_KREG_OFFSET(RcvEgrBase), + .kr_rcvegrcnt = IPATH_KREG_OFFSET(RcvEgrCnt), + .kr_rcvhdrcnt = IPATH_KREG_OFFSET(RcvHdrCnt), + .kr_rcvhdrentsize = IPATH_KREG_OFFSET(RcvHdrEntSize), + .kr_rcvhdrsize = IPATH_KREG_OFFSET(RcvHdrSize), + .kr_rcvintmembase = IPATH_KREG_OFFSET(RxIntMemBase), + .kr_rcvintmemsize = IPATH_KREG_OFFSET(RxIntMemSize), + .kr_rcvtidbase = IPATH_KREG_OFFSET(RcvTIDBase), + .kr_rcvtidcnt = IPATH_KREG_OFFSET(RcvTIDCnt), + .kr_revision = IPATH_KREG_OFFSET(Revision), + .kr_scratch = IPATH_KREG_OFFSET(Scratch), + .kr_sendbuffererror = IPATH_KREG_OFFSET(SendBufferError), + .kr_sendctrl = IPATH_KREG_OFFSET(SendCtrl), + .kr_sendpioavailaddr = IPATH_KREG_OFFSET(SendAvailAddr), + .kr_sendpiobufbase = IPATH_KREG_OFFSET(SendBufBase), + .kr_sendpiobufcnt = IPATH_KREG_OFFSET(SendBufCnt), + .kr_sendpiosize = IPATH_KREG_OFFSET(SendBufSize), + .kr_sendregbase = IPATH_KREG_OFFSET(SendRegBase), + .kr_txintmembase = IPATH_KREG_OFFSET(TxIntMemBase), + .kr_txintmemsize = IPATH_KREG_OFFSET(TxIntMemSize), + .kr_userregbase = IPATH_KREG_OFFSET(UserRegBase), + + .kr_xgxsconfig = IPATH_KREG_OFFSET(XGXSConfig), + + /* send dma related regs */ + .kr_senddmabase = IPATH_KREG_OFFSET(SendDmaBase), + .kr_senddmalengen = IPATH_KREG_OFFSET(SendDmaLenGen), + .kr_senddmatail = IPATH_KREG_OFFSET(SendDmaTail), + .kr_senddmahead = IPATH_KREG_OFFSET(SendDmaHead), + .kr_senddmaheadaddr = IPATH_KREG_OFFSET(SendDmaHeadAddr), + .kr_senddmabufmask0 = IPATH_KREG_OFFSET(SendDmaBufMask0), + .kr_senddmabufmask1 = IPATH_KREG_OFFSET(SendDmaBufMask1), + .kr_senddmabufmask2 = IPATH_KREG_OFFSET(SendDmaBufMask2), + .kr_senddmastatus = IPATH_KREG_OFFSET(SendDmaStatus), + + /* SerDes related regs */ + .kr_ibserdesctrl = IPATH_KREG_OFFSET(IBSerDesCtrl), + .kr_ib_epbacc = IPATH_KREG_OFFSET(IbsdEpbAccCtl), + .kr_ib_epbtrans = IPATH_KREG_OFFSET(IbsdEpbTransReg), + .kr_pcie_epbacc = IPATH_KREG_OFFSET(PcieEpbAccCtl), + .kr_pcie_epbtrans = IPATH_KREG_OFFSET(PcieEpbTransCtl), + .kr_ib_ddsrxeq = IPATH_KREG_OFFSET(SerDesDDSRXEQ), + + /* + * These should not be used directly via ipath_read_kreg64(), + * use them with ipath_read_kreg64_port() + */ + .kr_rcvhdraddr = IPATH_KREG_OFFSET(RcvHdrAddr0), + .kr_rcvhdrtailaddr = IPATH_KREG_OFFSET(RcvHdrTailAddr0), + + /* + * The rcvpktled register controls one of the debug port signals, so + * a packet activity LED can be connected to it. + */ + .kr_rcvpktledcnt = IPATH_KREG_OFFSET(RcvPktLEDCnt), + .kr_pcierbuftestreg0 = IPATH_KREG_OFFSET(PCIeRBufTestReg0), + .kr_pcierbuftestreg1 = IPATH_KREG_OFFSET(PCIeRBufTestReg1), + + .kr_hrtbt_guid = IPATH_KREG_OFFSET(HRTBT_GUID), + .kr_ibcddrctrl = IPATH_KREG_OFFSET(IBCDDRCtrl), + .kr_ibcddrstatus = IPATH_KREG_OFFSET(IBCDDRStatus), + .kr_jintreload = IPATH_KREG_OFFSET(JIntReload) +}; + +static const struct ipath_cregs ipath_7220_cregs = { + .cr_badformatcnt = IPATH_CREG_OFFSET(RxBadFormatCnt), + .cr_erricrccnt = IPATH_CREG_OFFSET(RxICRCErrCnt), + .cr_errlinkcnt = IPATH_CREG_OFFSET(RxLinkProblemCnt), + .cr_errlpcrccnt = IPATH_CREG_OFFSET(RxLPCRCErrCnt), + .cr_errpkey = IPATH_CREG_OFFSET(RxPKeyMismatchCnt), + .cr_errrcvflowctrlcnt = IPATH_CREG_OFFSET(RxFlowCtrlErrCnt), + .cr_err_rlencnt = IPATH_CREG_OFFSET(RxLenErrCnt), + .cr_errslencnt = IPATH_CREG_OFFSET(TxLenErrCnt), + .cr_errtidfull = IPATH_CREG_OFFSET(RxTIDFullErrCnt), + .cr_errtidvalid = IPATH_CREG_OFFSET(RxTIDValidErrCnt), + .cr_errvcrccnt = IPATH_CREG_OFFSET(RxVCRCErrCnt), + .cr_ibstatuschange = IPATH_CREG_OFFSET(IBStatusChangeCnt), + .cr_intcnt = IPATH_CREG_OFFSET(LBIntCnt), + .cr_invalidrlencnt = IPATH_CREG_OFFSET(RxMaxMinLenErrCnt), + .cr_invalidslencnt = IPATH_CREG_OFFSET(TxMaxMinLenErrCnt), + .cr_lbflowstallcnt = IPATH_CREG_OFFSET(LBFlowStallCnt), + .cr_pktrcvcnt = IPATH_CREG_OFFSET(RxDataPktCnt), + .cr_pktrcvflowctrlcnt = IPATH_CREG_OFFSET(RxFlowPktCnt), + .cr_pktsendcnt = IPATH_CREG_OFFSET(TxDataPktCnt), + .cr_pktsendflowcnt = IPATH_CREG_OFFSET(TxFlowPktCnt), + .cr_portovflcnt = IPATH_CREG_OFFSET(RxP0HdrEgrOvflCnt), + .cr_rcvebpcnt = IPATH_CREG_OFFSET(RxEBPCnt), + .cr_rcvovflcnt = IPATH_CREG_OFFSET(RxBufOvflCnt), + .cr_senddropped = IPATH_CREG_OFFSET(TxDroppedPktCnt), + .cr_sendstallcnt = IPATH_CREG_OFFSET(TxFlowStallCnt), + .cr_sendunderruncnt = IPATH_CREG_OFFSET(TxUnderrunCnt), + .cr_wordrcvcnt = IPATH_CREG_OFFSET(RxDwordCnt), + .cr_wordsendcnt = IPATH_CREG_OFFSET(TxDwordCnt), + .cr_unsupvlcnt = IPATH_CREG_OFFSET(TxUnsupVLErrCnt), + .cr_rxdroppktcnt = IPATH_CREG_OFFSET(RxDroppedPktCnt), + .cr_iblinkerrrecovcnt = IPATH_CREG_OFFSET(IBLinkErrRecoveryCnt), + .cr_iblinkdowncnt = IPATH_CREG_OFFSET(IBLinkDownedCnt), + .cr_ibsymbolerrcnt = IPATH_CREG_OFFSET(IBSymbolErrCnt), + .cr_vl15droppedpktcnt = IPATH_CREG_OFFSET(RxVL15DroppedPktCnt), + .cr_rxotherlocalphyerrcnt = + IPATH_CREG_OFFSET(RxOtherLocalPhyErrCnt), + .cr_excessbufferovflcnt = IPATH_CREG_OFFSET(ExcessBufferOvflCnt), + .cr_locallinkintegrityerrcnt = + IPATH_CREG_OFFSET(LocalLinkIntegrityErrCnt), + .cr_rxvlerrcnt = IPATH_CREG_OFFSET(RxVlErrCnt), + .cr_rxdlidfltrcnt = IPATH_CREG_OFFSET(RxDlidFltrCnt), + .cr_psstat = IPATH_CREG_OFFSET(PSStat), + .cr_psstart = IPATH_CREG_OFFSET(PSStart), + .cr_psinterval = IPATH_CREG_OFFSET(PSInterval), + .cr_psrcvdatacount = IPATH_CREG_OFFSET(PSRcvDataCount), + .cr_psrcvpktscount = IPATH_CREG_OFFSET(PSRcvPktsCount), + .cr_psxmitdatacount = IPATH_CREG_OFFSET(PSXmitDataCount), + .cr_psxmitpktscount = IPATH_CREG_OFFSET(PSXmitPktsCount), + .cr_psxmitwaitcount = IPATH_CREG_OFFSET(PSXmitWaitCount), +}; + +/* kr_control bits */ +#define INFINIPATH_C_RESET (1U<<7) + +/* kr_intstatus, kr_intclear, kr_intmask bits */ +#define INFINIPATH_I_RCVURG_MASK ((1ULL<<17)-1) +#define INFINIPATH_I_RCVURG_SHIFT 32 +#define INFINIPATH_I_RCVAVAIL_MASK ((1ULL<<17)-1) +#define INFINIPATH_I_RCVAVAIL_SHIFT 0 +#define INFINIPATH_I_SERDESTRIMDONE (1ULL<<27) + +/* kr_hwerrclear, kr_hwerrmask, kr_hwerrstatus, bits */ +#define INFINIPATH_HWE_PCIEMEMPARITYERR_MASK 0x00000000000000ffULL +#define INFINIPATH_HWE_PCIEMEMPARITYERR_SHIFT 0 +#define INFINIPATH_HWE_PCIEPOISONEDTLP 0x0000000010000000ULL +#define INFINIPATH_HWE_PCIECPLTIMEOUT 0x0000000020000000ULL +#define INFINIPATH_HWE_PCIEBUSPARITYXTLH 0x0000000040000000ULL +#define INFINIPATH_HWE_PCIEBUSPARITYXADM 0x0000000080000000ULL +#define INFINIPATH_HWE_PCIEBUSPARITYRADM 0x0000000100000000ULL +#define INFINIPATH_HWE_COREPLL_FBSLIP 0x0080000000000000ULL +#define INFINIPATH_HWE_COREPLL_RFSLIP 0x0100000000000000ULL +#define INFINIPATH_HWE_PCIE1PLLFAILED 0x0400000000000000ULL +#define INFINIPATH_HWE_PCIE0PLLFAILED 0x0800000000000000ULL +#define INFINIPATH_HWE_SERDESPLLFAILED 0x1000000000000000ULL +/* specific to this chip */ +#define INFINIPATH_HWE_PCIECPLDATAQUEUEERR 0x0000000000000040ULL +#define INFINIPATH_HWE_PCIECPLHDRQUEUEERR 0x0000000000000080ULL +#define INFINIPATH_HWE_SDMAMEMREADERR 0x0000000010000000ULL +#define INFINIPATH_HWE_CLK_UC_PLLNOTLOCKED 0x2000000000000000ULL +#define INFINIPATH_HWE_PCIESERDESQ0PCLKNOTDETECT 0x0100000000000000ULL +#define INFINIPATH_HWE_PCIESERDESQ1PCLKNOTDETECT 0x0200000000000000ULL +#define INFINIPATH_HWE_PCIESERDESQ2PCLKNOTDETECT 0x0400000000000000ULL +#define INFINIPATH_HWE_PCIESERDESQ3PCLKNOTDETECT 0x0800000000000000ULL +#define INFINIPATH_HWE_DDSRXEQMEMORYPARITYERR 0x0000008000000000ULL +#define INFINIPATH_HWE_IB_UC_MEMORYPARITYERR 0x0000004000000000ULL +#define INFINIPATH_HWE_PCIE_UC_OCT0MEMORYPARITYERR 0x0000001000000000ULL +#define INFINIPATH_HWE_PCIE_UC_OCT1MEMORYPARITYERR 0x0000002000000000ULL + +#define IBA7220_IBCS_LINKTRAININGSTATE_MASK 0x1F +#define IBA7220_IBCS_LINKSTATE_SHIFT 5 +#define IBA7220_IBCS_LINKSPEED_SHIFT 8 +#define IBA7220_IBCS_LINKWIDTH_SHIFT 9 + +#define IBA7220_IBCC_LINKINITCMD_MASK 0x7ULL +#define IBA7220_IBCC_LINKCMD_SHIFT 19 +#define IBA7220_IBCC_MAXPKTLEN_SHIFT 21 + +/* kr_ibcddrctrl bits */ +#define IBA7220_IBC_DLIDLMC_MASK 0xFFFFFFFFUL +#define IBA7220_IBC_DLIDLMC_SHIFT 32 +#define IBA7220_IBC_HRTBT_MASK 3 +#define IBA7220_IBC_HRTBT_SHIFT 16 +#define IBA7220_IBC_HRTBT_ENB 0x10000UL +#define IBA7220_IBC_LANE_REV_SUPPORTED (1<<8) +#define IBA7220_IBC_LREV_MASK 1 +#define IBA7220_IBC_LREV_SHIFT 8 +#define IBA7220_IBC_RXPOL_MASK 1 +#define IBA7220_IBC_RXPOL_SHIFT 7 +#define IBA7220_IBC_WIDTH_SHIFT 5 +#define IBA7220_IBC_WIDTH_MASK 0x3 +#define IBA7220_IBC_WIDTH_1X_ONLY (0<<IBA7220_IBC_WIDTH_SHIFT) +#define IBA7220_IBC_WIDTH_4X_ONLY (1<<IBA7220_IBC_WIDTH_SHIFT) +#define IBA7220_IBC_WIDTH_AUTONEG (2<<IBA7220_IBC_WIDTH_SHIFT) +#define IBA7220_IBC_SPEED_AUTONEG (1<<1) +#define IBA7220_IBC_SPEED_SDR (1<<2) +#define IBA7220_IBC_SPEED_DDR (1<<3) +#define IBA7220_IBC_SPEED_AUTONEG_MASK (0x7<<1) +#define IBA7220_IBC_IBTA_1_2_MASK (1) + +/* kr_ibcddrstatus */ +/* link latency shift is 0, don't bother defining */ +#define IBA7220_DDRSTAT_LINKLAT_MASK 0x3ffffff + +/* kr_extstatus bits */ +#define INFINIPATH_EXTS_FREQSEL 0x2 +#define INFINIPATH_EXTS_SERDESSEL 0x4 +#define INFINIPATH_EXTS_MEMBIST_ENDTEST 0x0000000000004000 +#define INFINIPATH_EXTS_MEMBIST_DISABLED 0x0000000000008000 + +/* kr_xgxsconfig bits */ +#define INFINIPATH_XGXS_RESET 0x5ULL +#define INFINIPATH_XGXS_FC_SAFE (1ULL<<63) + +/* kr_rcvpktledcnt */ +#define IBA7220_LEDBLINK_ON_SHIFT 32 /* 4ns period on after packet */ +#define IBA7220_LEDBLINK_OFF_SHIFT 0 /* 4ns period off before next on */ + +#define _IPATH_GPIO_SDA_NUM 1 +#define _IPATH_GPIO_SCL_NUM 0 + +#define IPATH_GPIO_SDA (1ULL << \ + (_IPATH_GPIO_SDA_NUM+INFINIPATH_EXTC_GPIOOE_SHIFT)) +#define IPATH_GPIO_SCL (1ULL << \ + (_IPATH_GPIO_SCL_NUM+INFINIPATH_EXTC_GPIOOE_SHIFT)) + +#define IBA7220_R_INTRAVAIL_SHIFT 17 +#define IBA7220_R_TAILUPD_SHIFT 35 +#define IBA7220_R_PORTCFG_SHIFT 36 + +#define INFINIPATH_JINT_PACKETSHIFT 16 +#define INFINIPATH_JINT_DEFAULT_IDLE_TICKS 0 +#define INFINIPATH_JINT_DEFAULT_MAX_PACKETS 0 + +#define IBA7220_HDRHEAD_PKTINT_SHIFT 32 /* interrupt cnt in upper 32 bits */ + +/* + * the size bits give us 2^N, in KB units. 0 marks as invalid, + * and 7 is reserved. We currently use only 2KB and 4KB + */ +#define IBA7220_TID_SZ_SHIFT 37 /* shift to 3bit size selector */ +#define IBA7220_TID_SZ_2K (1UL<<IBA7220_TID_SZ_SHIFT) /* 2KB */ +#define IBA7220_TID_SZ_4K (2UL<<IBA7220_TID_SZ_SHIFT) /* 4KB */ +#define IBA7220_TID_PA_SHIFT 11U /* TID addr in chip stored w/o low bits */ + +#define IPATH_AUTONEG_TRIES 5 /* sequential retries to negotiate DDR */ + +static char int_type[16] = "auto"; +module_param_string(interrupt_type, int_type, sizeof(int_type), 0444); +MODULE_PARM_DESC(int_type, " interrupt_type=auto|force_msi|force_intx\n"); + +/* packet rate matching delay; chip has support */ +static u8 rate_to_delay[2][2] = { + /* 1x, 4x */ + { 8, 2 }, /* SDR */ + { 4, 1 } /* DDR */ +}; + +/* 7220 specific hardware errors... */ +static const struct ipath_hwerror_msgs ipath_7220_hwerror_msgs[] = { + INFINIPATH_HWE_MSG(PCIEPOISONEDTLP, "PCIe Poisoned TLP"), + INFINIPATH_HWE_MSG(PCIECPLTIMEOUT, "PCIe completion timeout"), + /* + * In practice, it's unlikely wthat we'll see PCIe PLL, or bus + * parity or memory parity error failures, because most likely we + * won't be able to talk to the core of the chip. Nonetheless, we + * might see them, if they are in parts of the PCIe core that aren't + * essential. + */ + INFINIPATH_HWE_MSG(PCIE1PLLFAILED, "PCIePLL1"), + INFINIPATH_HWE_MSG(PCIE0PLLFAILED, "PCIePLL0"), + INFINIPATH_HWE_MSG(PCIEBUSPARITYXTLH, "PCIe XTLH core parity"), + INFINIPATH_HWE_MSG(PCIEBUSPARITYXADM, "PCIe ADM TX core parity"), + INFINIPATH_HWE_MSG(PCIEBUSPARITYRADM, "PCIe ADM RX core parity"), + INFINIPATH_HWE_MSG(RXDSYNCMEMPARITYERR, "Rx Dsync"), + INFINIPATH_HWE_MSG(SERDESPLLFAILED, "SerDes PLL"), + INFINIPATH_HWE_MSG(PCIECPLDATAQUEUEERR, "PCIe cpl header queue"), + INFINIPATH_HWE_MSG(PCIECPLHDRQUEUEERR, "PCIe cpl data queue"), + INFINIPATH_HWE_MSG(SDMAMEMREADERR, "Send DMA memory read"), + INFINIPATH_HWE_MSG(CLK_UC_PLLNOTLOCKED, "uC PLL clock not locked"), + INFINIPATH_HWE_MSG(PCIESERDESQ0PCLKNOTDETECT, + "PCIe serdes Q0 no clock"), + INFINIPATH_HWE_MSG(PCIESERDESQ1PCLKNOTDETECT, + "PCIe serdes Q1 no clock"), + INFINIPATH_HWE_MSG(PCIESERDESQ2PCLKNOTDETECT, + "PCIe serdes Q2 no clock"), + INFINIPATH_HWE_MSG(PCIESERDESQ3PCLKNOTDETECT, + "PCIe serdes Q3 no clock"), + INFINIPATH_HWE_MSG(DDSRXEQMEMORYPARITYERR, + "DDS RXEQ memory parity"), + INFINIPATH_HWE_MSG(IB_UC_MEMORYPARITYERR, "IB uC memory parity"), + INFINIPATH_HWE_MSG(PCIE_UC_OCT0MEMORYPARITYERR, + "PCIe uC oct0 memory parity"), + INFINIPATH_HWE_MSG(PCIE_UC_OCT1MEMORYPARITYERR, + "PCIe uC oct1 memory parity"), +}; + +static void autoneg_work(struct work_struct *); + +/* + * the offset is different for different configured port numbers, since + * port0 is fixed in size, but others can vary. Make it a function to + * make the issue more obvious. +*/ +static inline u32 port_egrtid_idx(struct ipath_devdata *dd, unsigned port) +{ + return port ? dd->ipath_p0_rcvegrcnt + + (port-1) * dd->ipath_rcvegrcnt : 0; +} + +static void ipath_7220_txe_recover(struct ipath_devdata *dd) +{ + ++ipath_stats.sps_txeparity; + + dev_info(&dd->pcidev->dev, + "Recovering from TXE PIO parity error\n"); + ipath_disarm_senderrbufs(dd, 1); +} + + +/** + * ipath_7220_handle_hwerrors - display hardware errors. + * @dd: the infinipath device + * @msg: the output buffer + * @msgl: the size of the output buffer + * + * Use same msg buffer as regular errors to avoid excessive stack + * use. Most hardware errors are catastrophic, but for right now, + * we'll print them and continue. We reuse the same message buffer as + * ipath_handle_errors() to avoid excessive stack usage. + */ +static void ipath_7220_handle_hwerrors(struct ipath_devdata *dd, char *msg, + size_t msgl) +{ + ipath_err_t hwerrs; + u32 bits, ctrl; + int isfatal = 0; + char bitsmsg[64]; + int log_idx; + + hwerrs = ipath_read_kreg64(dd, dd->ipath_kregs->kr_hwerrstatus); + if (!hwerrs) { + /* + * better than printing cofusing messages + * This seems to be related to clearing the crc error, or + * the pll error during init. + */ + ipath_cdbg(VERBOSE, "Called but no hardware errors set\n"); + goto bail; + } else if (hwerrs == ~0ULL) { + ipath_dev_err(dd, "Read of hardware error status failed " + "(all bits set); ignoring\n"); + goto bail; + } + ipath_stats.sps_hwerrs++; + + /* + * Always clear the error status register, except MEMBISTFAIL, + * regardless of whether we continue or stop using the chip. + * We want that set so we know it failed, even across driver reload. + * We'll still ignore it in the hwerrmask. We do this partly for + * diagnostics, but also for support. + */ + ipath_write_kreg(dd, dd->ipath_kregs->kr_hwerrclear, + hwerrs&~INFINIPATH_HWE_MEMBISTFAILED); + + hwerrs &= dd->ipath_hwerrmask; + + /* We log some errors to EEPROM, check if we have any of those. */ + for (log_idx = 0; log_idx < IPATH_EEP_LOG_CNT; ++log_idx) + if (hwerrs & dd->ipath_eep_st_masks[log_idx].hwerrs_to_log) + ipath_inc_eeprom_err(dd, log_idx, 1); + /* + * Make sure we get this much out, unless told to be quiet, + * or it's occurred within the last 5 seconds. + */ + if ((hwerrs & ~(dd->ipath_lasthwerror | + ((INFINIPATH_HWE_TXEMEMPARITYERR_PIOBUF | + INFINIPATH_HWE_TXEMEMPARITYERR_PIOPBC) + << INFINIPATH_HWE_TXEMEMPARITYERR_SHIFT))) || + (ipath_debug & __IPATH_VERBDBG)) + dev_info(&dd->pcidev->dev, "Hardware error: hwerr=0x%llx " + "(cleared)\n", (unsigned long long) hwerrs); + dd->ipath_lasthwerror |= hwerrs; + + if (hwerrs & ~dd->ipath_hwe_bitsextant) + ipath_dev_err(dd, "hwerror interrupt with unknown errors " + "%llx set\n", (unsigned long long) + (hwerrs & ~dd->ipath_hwe_bitsextant)); + + if (hwerrs & INFINIPATH_HWE_IB_UC_MEMORYPARITYERR) + ipath_sd7220_clr_ibpar(dd); + + ctrl = ipath_read_kreg32(dd, dd->ipath_kregs->kr_control); + if ((ctrl & INFINIPATH_C_FREEZEMODE) && !ipath_diag_inuse) { + /* + * Parity errors in send memory are recoverable, + * just cancel the send (if indicated in * sendbuffererror), + * count the occurrence, unfreeze (if no other handled + * hardware error bits are set), and continue. + */ + if (hwerrs & ((INFINIPATH_HWE_TXEMEMPARITYERR_PIOBUF | + INFINIPATH_HWE_TXEMEMPARITYERR_PIOPBC) + << INFINIPATH_HWE_TXEMEMPARITYERR_SHIFT)) { + ipath_7220_txe_recover(dd); + hwerrs &= ~((INFINIPATH_HWE_TXEMEMPARITYERR_PIOBUF | + INFINIPATH_HWE_TXEMEMPARITYERR_PIOPBC) + << INFINIPATH_HWE_TXEMEMPARITYERR_SHIFT); + if (!hwerrs) { + /* else leave in freeze mode */ + ipath_write_kreg(dd, + dd->ipath_kregs->kr_control, + dd->ipath_control); + goto bail; + } + } + if (hwerrs) { + /* + * If any set that we aren't ignoring only make the + * complaint once, in case it's stuck or recurring, + * and we get here multiple times + * Force link down, so switch knows, and + * LEDs are turned off. + */ + if (dd->ipath_flags & IPATH_INITTED) { + ipath_set_linkstate(dd, IPATH_IB_LINKDOWN); + ipath_setup_7220_setextled(dd, + INFINIPATH_IBCS_L_STATE_DOWN, + INFINIPATH_IBCS_LT_STATE_DISABLED); + ipath_dev_err(dd, "Fatal Hardware Error " + "(freeze mode), no longer" + " usable, SN %.16s\n", + dd->ipath_serial); + isfatal = 1; + } + /* + * Mark as having had an error for driver, and also + * for /sys and status word mapped to user programs. + * This marks unit as not usable, until reset. + */ + *dd->ipath_statusp &= ~IPATH_STATUS_IB_READY; + *dd->ipath_statusp |= IPATH_STATUS_HWERROR; + dd->ipath_flags &= ~IPATH_INITTED; + } else { + ipath_dbg("Clearing freezemode on ignored hardware " + "error\n"); + ipath_clear_freeze(dd); + } + } + + *msg = '\0'; + + if (hwerrs & INFINIPATH_HWE_MEMBISTFAILED) { + strlcat(msg, "[Memory BIST test failed, " + "InfiniPath hardware unusable]", msgl); + /* ignore from now on, so disable until driver reloaded */ + *dd->ipath_statusp |= IPATH_STATUS_HWERROR; + dd->ipath_hwerrmask &= ~INFINIPATH_HWE_MEMBISTFAILED; + ipath_write_kreg(dd, dd->ipath_kregs->kr_hwerrmask, + dd->ipath_hwerrmask); + } + + ipath_format_hwerrors(hwerrs, + ipath_7220_hwerror_msgs, + ARRAY_SIZE(ipath_7220_hwerror_msgs), + msg, msgl); + + if (hwerrs & (INFINIPATH_HWE_PCIEMEMPARITYERR_MASK + << INFINIPATH_HWE_PCIEMEMPARITYERR_SHIFT)) { + bits = (u32) ((hwerrs >> + INFINIPATH_HWE_PCIEMEMPARITYERR_SHIFT) & + INFINIPATH_HWE_PCIEMEMPARITYERR_MASK); + snprintf(bitsmsg, sizeof bitsmsg, + "[PCIe Mem Parity Errs %x] ", bits); + strlcat(msg, bitsmsg, msgl); + } + +#define _IPATH_PLL_FAIL (INFINIPATH_HWE_COREPLL_FBSLIP | \ + INFINIPATH_HWE_COREPLL_RFSLIP) + + if (hwerrs & _IPATH_PLL_FAIL) { + snprintf(bitsmsg, sizeof bitsmsg, + "[PLL failed (%llx), InfiniPath hardware unusable]", + (unsigned long long) hwerrs & _IPATH_PLL_FAIL); + strlcat(msg, bitsmsg, msgl); + /* ignore from now on, so disable until driver reloaded */ + dd->ipath_hwerrmask &= ~(hwerrs & _IPATH_PLL_FAIL); + ipath_write_kreg(dd, dd->ipath_kregs->kr_hwerrmask, + dd->ipath_hwerrmask); + } + + if (hwerrs & INFINIPATH_HWE_SERDESPLLFAILED) { + /* + * If it occurs, it is left masked since the eternal + * interface is unused. + */ + dd->ipath_hwerrmask &= ~INFINIPATH_HWE_SERDESPLLFAILED; + ipath_write_kreg(dd, dd->ipath_kregs->kr_hwerrmask, + dd->ipath_hwerrmask); + } + + ipath_dev_err(dd, "%s hardware error\n", msg); + /* + * For /sys status file. if no trailing } is copied, we'll + * know it was truncated. + */ + if (isfatal && !ipath_diag_inuse && dd->ipath_freezemsg) + snprintf(dd->ipath_freezemsg, dd->ipath_freezelen, + "{%s}", msg); +bail:; +} + +/** + * ipath_7220_boardname - fill in the board name + * @dd: the infinipath device + * @name: the output buffer + * @namelen: the size of the output buffer + * + * info is based on the board revision register + */ +static int ipath_7220_boardname(struct ipath_devdata *dd, char *name, + size_t namelen) +{ + char *n = NULL; + u8 boardrev = dd->ipath_boardrev; + int ret; + + if (boardrev == 15) { + /* + * Emulator sometimes comes up all-ones, rather than zero. + */ + boardrev = 0; + dd->ipath_boardrev = boardrev; + } + switch (boardrev) { + case 0: + n = "InfiniPath_7220_Emulation"; + break; + case 1: + n = "InfiniPath_QLE7240"; + break; + case 2: + n = "InfiniPath_QLE7280"; + break; + case 3: + n = "InfiniPath_QLE7242"; + break; + case 4: + n = "InfiniPath_QEM7240"; + break; + case 5: + n = "InfiniPath_QMI7240"; + break; + case 6: + n = "InfiniPath_QMI7264"; + break; + case 7: + n = "InfiniPath_QMH7240"; + break; + case 8: + n = "InfiniPath_QME7240"; + break; + case 9: + n = "InfiniPath_QLE7250"; + break; + case 10: + n = "InfiniPath_QLE7290"; + break; + case 11: + n = "InfiniPath_QEM7250"; + break; + case 12: + n = "InfiniPath_QLE-Bringup"; + break; + default: + ipath_dev_err(dd, + "Don't yet know about board with ID %u\n", + boardrev); + snprintf(name, namelen, "Unknown_InfiniPath_PCIe_%u", + boardrev); + break; + } + if (n) + snprintf(name, namelen, "%s", n); + + if (dd->ipath_majrev != 5 || !dd->ipath_minrev || + dd->ipath_minrev > 2) { + ipath_dev_err(dd, "Unsupported InfiniPath hardware " + "revision %u.%u!\n", + dd->ipath_majrev, dd->ipath_minrev); + ret = 1; + } else if (dd->ipath_minrev == 1) { + /* Rev1 chips are prototype. Complain, but allow use */ + ipath_dev_err(dd, "Unsupported hardware " + "revision %u.%u, Contact support@qlogic.com\n", + dd->ipath_majrev, dd->ipath_minrev); + ret = 0; + } else + ret = 0; + + /* + * Set here not in ipath_init_*_funcs because we have to do + * it after we can read chip registers. + */ + dd->ipath_ureg_align = 0x10000; /* 64KB alignment */ + + return ret; +} + +/** + * ipath_7220_init_hwerrors - enable hardware errors + * @dd: the infinipath device + * + * now that we have finished initializing everything that might reasonably + * cause a hardware error, and cleared those errors bits as they occur, + * we can enable hardware errors in the mask (potentially enabling + * freeze mode), and enable hardware errors as errors (along with + * everything else) in errormask + */ +static void ipath_7220_init_hwerrors(struct ipath_devdata *dd) +{ + ipath_err_t val; + u64 extsval; + + extsval = ipath_read_kreg64(dd, dd->ipath_kregs->kr_extstatus); + + if (!(extsval & (INFINIPATH_EXTS_MEMBIST_ENDTEST | + INFINIPATH_EXTS_MEMBIST_DISABLED))) + ipath_dev_err(dd, "MemBIST did not complete!\n"); + if (extsval & INFINIPATH_EXTS_MEMBIST_DISABLED) + dev_info(&dd->pcidev->dev, "MemBIST is disabled.\n"); + + val = ~0ULL; /* barring bugs, all hwerrors become interrupts, */ + + if (!dd->ipath_boardrev) /* no PLL for Emulator */ + val &= ~INFINIPATH_HWE_SERDESPLLFAILED; + + if (dd->ipath_minrev == 1) + val &= ~(1ULL << 42); /* TXE LaunchFIFO Parity rev1 issue */ + + val &= ~INFINIPATH_HWE_IB_UC_MEMORYPARITYERR; + dd->ipath_hwerrmask = val; + + /* + * special trigger "error" is for debugging purposes. It + * works around a processor/chipset problem. The error + * interrupt allows us to count occurrences, but we don't + * want to pay the overhead for normal use. Emulation only + */ + if (!dd->ipath_boardrev) + dd->ipath_maskederrs = INFINIPATH_E_SENDSPECIALTRIGGER; +} + +/* + * All detailed interaction with the SerDes has been moved to ipath_sd7220.c + * + * The portion of IBA7220-specific bringup_serdes() that actually deals with + * registers and memory within the SerDes itself is ipath_sd7220_init(). + */ + +/** + * ipath_7220_bringup_serdes - bring up the serdes + * @dd: the infinipath device + */ +static int ipath_7220_bringup_serdes(struct ipath_devdata *dd) +{ + int ret = 0; + u64 val, prev_val, guid; + int was_reset; /* Note whether uC was reset */ + + ipath_dbg("Trying to bringup serdes\n"); + + if (ipath_read_kreg64(dd, dd->ipath_kregs->kr_hwerrstatus) & + INFINIPATH_HWE_SERDESPLLFAILED) { + ipath_dbg("At start, serdes PLL failed bit set " + "in hwerrstatus, clearing and continuing\n"); + ipath_write_kreg(dd, dd->ipath_kregs->kr_hwerrclear, + INFINIPATH_HWE_SERDESPLLFAILED); + } + + if (!dd->ipath_ibcddrctrl) { + /* not on re-init after reset */ + dd->ipath_ibcddrctrl = + ipath_read_kreg64(dd, dd->ipath_kregs->kr_ibcddrctrl); + + if (dd->ipath_link_speed_enabled == + (IPATH_IB_SDR | IPATH_IB_DDR)) + dd->ipath_ibcddrctrl |= + IBA7220_IBC_SPEED_AUTONEG_MASK | + IBA7220_IBC_IBTA_1_2_MASK; + else + dd->ipath_ibcddrctrl |= + dd->ipath_link_speed_enabled == IPATH_IB_DDR + ? IBA7220_IBC_SPEED_DDR : + IBA7220_IBC_SPEED_SDR; + if ((dd->ipath_link_width_enabled & (IB_WIDTH_1X | + IB_WIDTH_4X)) == (IB_WIDTH_1X | IB_WIDTH_4X)) + dd->ipath_ibcddrctrl |= IBA7220_IBC_WIDTH_AUTONEG; + else + dd->ipath_ibcddrctrl |= + dd->ipath_link_width_enabled == IB_WIDTH_4X + ? IBA7220_IBC_WIDTH_4X_ONLY : + IBA7220_IBC_WIDTH_1X_ONLY; + + /* always enable these on driver reload, not sticky */ + dd->ipath_ibcddrctrl |= + IBA7220_IBC_RXPOL_MASK << IBA7220_IBC_RXPOL_SHIFT; + dd->ipath_ibcddrctrl |= + IBA7220_IBC_HRTBT_MASK << IBA7220_IBC_HRTBT_SHIFT; + /* + * automatic lane reversal detection for receive + * doesn't work correctly in rev 1, so disable it + * on that rev, otherwise enable (disabling not + * sticky across reload for >rev1) + */ + if (dd->ipath_minrev == 1) + dd->ipath_ibcddrctrl &= + ~IBA7220_IBC_LANE_REV_SUPPORTED; + else + dd->ipath_ibcddrctrl |= + IBA7220_IBC_LANE_REV_SUPPORTED; + } + + ipath_write_kreg(dd, dd->ipath_kregs->kr_ibcddrctrl, + dd->ipath_ibcddrctrl); + + ipath_write_kreg(dd, IPATH_KREG_OFFSET(IBNCModeCtrl), 0Ull); + + /* IBA7220 has SERDES MPU reset in D0 of what _was_ IBPLLCfg */ + val = ipath_read_kreg64(dd, dd->ipath_kregs->kr_ibserdesctrl); + /* remember if uC was in Reset or not, for dactrim */ + was_reset = (val & 1); + ipath_cdbg(VERBOSE, "IBReset %s xgxsconfig %llx\n", + was_reset ? "Asserted" : "Negated", (unsigned long long) + ipath_read_kreg64(dd, dd->ipath_kregs->kr_xgxsconfig)); + + if (dd->ipath_boardrev) { + /* + * Hardware is not emulator, and may have been reset. Init it. + * Below will release reset, but needs to know if chip was + * originally in reset, to only trim DACs on first time + * after chip reset or powercycle (not driver reload) + */ + ret = ipath_sd7220_init(dd, was_reset); + } + + val = ipath_read_kreg64(dd, dd->ipath_kregs->kr_xgxsconfig); + prev_val = val; + val |= INFINIPATH_XGXS_FC_SAFE; + if (val != prev_val) { + ipath_write_kreg(dd, dd->ipath_kregs->kr_xgxsconfig, val); + ipath_read_kreg32(dd, dd->ipath_kregs->kr_scratch); + } + if (val & INFINIPATH_XGXS_RESET) + val &= ~INFINIPATH_XGXS_RESET; + if (val != prev_val) + ipath_write_kreg(dd, dd->ipath_kregs->kr_xgxsconfig, val); + + ipath_cdbg(VERBOSE, "done: xgxs=%llx from %llx\n", + (unsigned long long) + ipath_read_kreg64(dd, dd->ipath_kregs->kr_xgxsconfig), + prev_val); + + guid = be64_to_cpu(dd->ipath_guid); + + if (!guid) { + /* have to have something, so use likely unique tsc */ + guid = get_cycles(); + ipath_dbg("No GUID for heartbeat, faking %llx\n", + (unsigned long long)guid); + } else + ipath_cdbg(VERBOSE, "Wrote %llX to HRTBT_GUID\n", guid); + ipath_write_kreg(dd, dd->ipath_kregs->kr_hrtbt_guid, guid); + return ret; +} + +static void ipath_7220_config_jint(struct ipath_devdata *dd, + u16 idle_ticks, u16 max_packets) +{ + + /* + * We can request a receive interrupt for 1 or more packets + * from current offset. + */ + if (idle_ticks == 0 || max_packets == 0) + /* interrupt after one packet if no mitigation */ + dd->ipath_rhdrhead_intr_off = + 1ULL << IBA7220_HDRHEAD_PKTINT_SHIFT; + else + /* Turn off RcvHdrHead interrupts if using mitigation */ + dd->ipath_rhdrhead_intr_off = 0ULL; + + /* refresh kernel RcvHdrHead registers... */ + ipath_write_ureg(dd, ur_rcvhdrhead, + dd->ipath_rhdrhead_intr_off | + dd->ipath_pd[0]->port_head, 0); + + dd->ipath_jint_max_packets = max_packets; + dd->ipath_jint_idle_ticks = idle_ticks; + ipath_write_kreg(dd, dd->ipath_kregs->kr_jintreload, + ((u64) max_packets << INFINIPATH_JINT_PACKETSHIFT) | + idle_ticks); +} + +/** + * ipath_7220_quiet_serdes - set serdes to txidle + * @dd: the infinipath device + * Called when driver is being unloaded + */ +static void ipath_7220_quiet_serdes(struct ipath_devdata *dd) +{ + u64 val; + dd->ipath_flags &= ~IPATH_IB_AUTONEG_INPROG; + wake_up(&dd->ipath_autoneg_wait); + cancel_delayed_work(&dd->ipath_autoneg_work); + flush_scheduled_work(); + ipath_shutdown_relock_poll(dd); + val = ipath_read_kreg64(dd, dd->ipath_kregs->kr_xgxsconfig); + val |= INFINIPATH_XGXS_RESET; + ipath_write_kreg(dd, dd->ipath_kregs->kr_xgxsconfig, val); +} + +static int ipath_7220_intconfig(struct ipath_devdata *dd) +{ + ipath_7220_config_jint(dd, dd->ipath_jint_idle_ticks, + dd->ipath_jint_max_packets); + return 0; +} + +/** + * ipath_setup_7220_setextled - set the state of the two external LEDs + * @dd: the infinipath device + * @lst: the L state + * @ltst: the LT state + * + * These LEDs indicate the physical and logical state of IB link. + * For this chip (at least with recommended board pinouts), LED1 + * is Yellow (logical state) and LED2 is Green (physical state), + * + * Note: We try to match the Mellanox HCA LED behavior as best + * we can. Green indicates physical link state is OK (something is + * plugged in, and we can train). + * Amber indicates the link is logically up (ACTIVE). + * Mellanox further blinks the amber LED to indicate data packet + * activity, but we have no hardware support for that, so it would + * require waking up every 10-20 msecs and checking the counters + * on the chip, and then turning the LED off if appropriate. That's + * visible overhead, so not something we will do. + * + */ +static void ipath_setup_7220_setextled(struct ipath_devdata *dd, u64 lst, + u64 ltst) +{ + u64 extctl, ledblink = 0; + unsigned long flags = 0; + + /* the diags use the LED to indicate diag info, so we leave + * the external LED alone when the diags are running */ + if (ipath_diag_inuse) + return; + + /* Allow override of LED display for, e.g. Locating system in rack */ + if (dd->ipath_led_override) { + ltst = (dd->ipath_led_override & IPATH_LED_PHYS) + ? INFINIPATH_IBCS_LT_STATE_LINKUP + : INFINIPATH_IBCS_LT_STATE_DISABLED; + lst = (dd->ipath_led_override & IPATH_LED_LOG) + ? INFINIPATH_IBCS_L_STATE_ACTIVE + : INFINIPATH_IBCS_L_STATE_DOWN; + } + + spin_lock_irqsave(&dd->ipath_gpio_lock, flags); + extctl = dd->ipath_extctrl & ~(INFINIPATH_EXTC_LED1PRIPORT_ON | + INFINIPATH_EXTC_LED2PRIPORT_ON); + if (ltst == INFINIPATH_IBCS_LT_STATE_LINKUP) { + extctl |= INFINIPATH_EXTC_LED1PRIPORT_ON; + /* + * counts are in chip clock (4ns) periods. + * This is 1/16 sec (66.6ms) on, + * 3/16 sec (187.5 ms) off, with packets rcvd + */ + ledblink = ((66600*1000UL/4) << IBA7220_LEDBLINK_ON_SHIFT) + | ((187500*1000UL/4) << IBA7220_LEDBLINK_OFF_SHIFT); + } + if (lst == INFINIPATH_IBCS_L_STATE_ACTIVE) + extctl |= INFINIPATH_EXTC_LED2PRIPORT_ON; + dd->ipath_extctrl = extctl; + ipath_write_kreg(dd, dd->ipath_kregs->kr_extctrl, extctl); + spin_unlock_irqrestore(&dd->ipath_gpio_lock, flags); + + if (ledblink) /* blink the LED on packet receive */ + ipath_write_kreg(dd, dd->ipath_kregs->kr_rcvpktledcnt, + ledblink); +} + +/* + * Similar to pci_intx(pdev, 1), except that we make sure + * msi is off... + */ +static void ipath_enable_intx(struct pci_dev *pdev) +{ + u16 cw, new; + int pos; + + /* first, turn on INTx */ + pci_read_config_word(pdev, PCI_COMMAND, &cw); + new = cw & ~PCI_COMMAND_INTX_DISABLE; + if (new != cw) + pci_write_config_word(pdev, PCI_COMMAND, new); + + /* then turn off MSI */ + pos = pci_find_capability(pdev, PCI_CAP_ID_MSI); + if (pos) { + pci_read_config_word(pdev, pos + PCI_MSI_FLAGS, &cw); + new = cw & ~PCI_MSI_FLAGS_ENABLE; + if (new != cw) + pci_write_config_word(pdev, pos + PCI_MSI_FLAGS, new); + } +} + +static int ipath_msi_enabled(struct pci_dev *pdev) +{ + int pos, ret = 0; + + pos = pci_find_capability(pdev, PCI_CAP_ID_MSI); + if (pos) { + u16 cw; + + pci_read_config_word(pdev, pos + PCI_MSI_FLAGS, &cw); + ret = !!(cw & PCI_MSI_FLAGS_ENABLE); + } + return ret; +} + +/* + * disable msi interrupt if enabled, and clear the flag. + * flag is used primarily for the fallback to IntX, but + * is also used in reinit after reset as a flag. + */ +static void ipath_7220_nomsi(struct ipath_devdata *dd) +{ + dd->ipath_msi_lo = 0; +#ifdef CONFIG_PCI_MSI + if (ipath_msi_enabled(dd->pcidev)) { + /* + * free, but don't zero; later kernels require + * it be freed before disable_msi, so the intx + * setup has to request it again. + */ + if (dd->ipath_irq) + free_irq(dd->ipath_irq, dd); + pci_disable_msi(dd->pcidev); + } +#endif +} + +/* + * ipath_setup_7220_cleanup - clean up any per-chip chip-specific stuff + * @dd: the infinipath device + * + * Nothing but msi interrupt cleanup for now. + * + * This is called during driver unload. + */ +static void ipath_setup_7220_cleanup(struct ipath_devdata *dd) +{ + ipath_7220_nomsi(dd); +} + + +static void ipath_7220_pcie_params(struct ipath_devdata *dd, u32 boardrev) +{ + u16 linkstat, minwidth, speed; + int pos; + + pos = pci_find_capability(dd->pcidev, PCI_CAP_ID_EXP); + if (!pos) { + ipath_dev_err(dd, "Can't find PCI Express capability!\n"); + goto bail; + } + + pci_read_config_word(dd->pcidev, pos + PCI_EXP_LNKSTA, + &linkstat); + /* + * speed is bits 0-4, linkwidth is bits 4-8 + * no defines for them in headers + */ + speed = linkstat & 0xf; + linkstat >>= 4; + linkstat &= 0x1f; + dd->ipath_lbus_width = linkstat; + switch (boardrev) { + case 0: + case 2: + case 10: + case 12: + minwidth = 16; /* x16 capable boards */ + break; + default: + minwidth = 8; /* x8 capable boards */ + break; + } + + switch (speed) { + case 1: + dd->ipath_lbus_speed = 2500; /* Gen1, 2.5GHz */ + break; + case 2: + dd->ipath_lbus_speed = 5000; /* Gen1, 5GHz */ + break; + default: /* not defined, assume gen1 */ + dd->ipath_lbus_speed = 2500; + break; + } + + if (linkstat < minwidth) + ipath_dev_err(dd, + "PCIe width %u (x%u HCA), performance " + "reduced\n", linkstat, minwidth); + else + ipath_cdbg(VERBOSE, "PCIe speed %u width %u (x%u HCA)\n", + dd->ipath_lbus_speed, linkstat, minwidth); + + if (speed != 1) + ipath_dev_err(dd, + "PCIe linkspeed %u is incorrect; " + "should be 1 (2500)!\n", speed); + +bail: + /* fill in string, even on errors */ + snprintf(dd->ipath_lbus_info, sizeof(dd->ipath_lbus_info), + "PCIe,%uMHz,x%u\n", + dd->ipath_lbus_speed, + dd->ipath_lbus_width); + return; +} + + +/** + * ipath_setup_7220_config - setup PCIe config related stuff + * @dd: the infinipath device + * @pdev: the PCI device + * + * The pci_enable_msi() call will fail on systems with MSI quirks + * such as those with AMD8131, even if the device of interest is not + * attached to that device, (in the 2.6.13 - 2.6.15 kernels, at least, fixed + * late in 2.6.16). + * All that can be done is to edit the kernel source to remove the quirk + * check until that is fixed. + * We do not need to call enable_msi() for our HyperTransport chip, + * even though it uses MSI, and we want to avoid the quirk warning, so + * So we call enable_msi only for PCIe. If we do end up needing + * pci_enable_msi at some point in the future for HT, we'll move the + * call back into the main init_one code. + * We save the msi lo and hi values, so we can restore them after + * chip reset (the kernel PCI infrastructure doesn't yet handle that + * correctly). + */ +static int ipath_setup_7220_config(struct ipath_devdata *dd, + struct pci_dev *pdev) +{ + int pos, ret = -1; + u32 boardrev; + + dd->ipath_msi_lo = 0; /* used as a flag during reset processing */ +#ifdef CONFIG_PCI_MSI + pos = pci_find_capability(pdev, PCI_CAP_ID_MSI); + if (!strcmp(int_type, "force_msi") || !strcmp(int_type, "auto")) + ret = pci_enable_msi(pdev); + if (ret) { + if (!strcmp(int_type, "force_msi")) { + ipath_dev_err(dd, "pci_enable_msi failed: %d, " + "force_msi is on, so not continuing.\n", + ret); + return ret; + } + + ipath_enable_intx(pdev); + if (!strcmp(int_type, "auto")) + ipath_dev_err(dd, "pci_enable_msi failed: %d, " + "falling back to INTx\n", ret); + } else if (pos) { + u16 control; + pci_read_config_dword(pdev, pos + PCI_MSI_ADDRESS_LO, + &dd->ipath_msi_lo); + pci_read_config_dword(pdev, pos + PCI_MSI_ADDRESS_HI, + &dd->ipath_msi_hi); + pci_read_config_word(pdev, pos + PCI_MSI_FLAGS, + &control); + /* now save the data (vector) info */ + pci_read_config_word(pdev, + pos + ((control & PCI_MSI_FLAGS_64BIT) + ? PCI_MSI_DATA_64 : + PCI_MSI_DATA_32), + &dd->ipath_msi_data); + } else + ipath_dev_err(dd, "Can't find MSI capability, " + "can't save MSI settings for reset\n"); +#else + ipath_dbg("PCI_MSI not configured, using IntX interrupts\n"); + ipath_enable_intx(pdev); +#endif + + dd->ipath_irq = pdev->irq; + + /* + * We save the cachelinesize also, although it doesn't + * really matter. + */ + pci_read_config_byte(pdev, PCI_CACHE_LINE_SIZE, + &dd->ipath_pci_cacheline); + + /* + * this function called early, ipath_boardrev not set yet. Can't + * use ipath_read_kreg64() yet, too early in init, so use readq() + */ + boardrev = (readq(&dd->ipath_kregbase[dd->ipath_kregs->kr_revision]) + >> INFINIPATH_R_BOARDID_SHIFT) & INFINIPATH_R_BOARDID_MASK; + + ipath_7220_pcie_params(dd, boardrev); + + dd->ipath_flags |= IPATH_NODMA_RTAIL | IPATH_HAS_SEND_DMA | + IPATH_HAS_PBC_CNT | IPATH_HAS_THRESH_UPDATE; + dd->ipath_pioupd_thresh = 4U; /* set default update threshold */ + return 0; +} + +static void ipath_init_7220_variables(struct ipath_devdata *dd) +{ + /* + * setup the register offsets, since they are different for each + * chip + */ + dd->ipath_kregs = &ipath_7220_kregs; + dd->ipath_cregs = &ipath_7220_cregs; + + /* + * bits for selecting i2c direction and values, + * used for I2C serial flash + */ + dd->ipath_gpio_sda_num = _IPATH_GPIO_SDA_NUM; + dd->ipath_gpio_scl_num = _IPATH_GPIO_SCL_NUM; + dd->ipath_gpio_sda = IPATH_GPIO_SDA; + dd->ipath_gpio_scl = IPATH_GPIO_SCL; + + /* + * Fill in data for field-values that change in IBA7220. + * We dynamically specify only the mask for LINKTRAININGSTATE + * and only the shift for LINKSTATE, as they are the only ones + * that change. Also precalculate the 3 link states of interest + * and the combined mask. + */ + dd->ibcs_ls_shift = IBA7220_IBCS_LINKSTATE_SHIFT; + dd->ibcs_lts_mask = IBA7220_IBCS_LINKTRAININGSTATE_MASK; + dd->ibcs_mask = (INFINIPATH_IBCS_LINKSTATE_MASK << + dd->ibcs_ls_shift) | dd->ibcs_lts_mask; + dd->ib_init = (INFINIPATH_IBCS_LT_STATE_LINKUP << + INFINIPATH_IBCS_LINKTRAININGSTATE_SHIFT) | + (INFINIPATH_IBCS_L_STATE_INIT << dd->ibcs_ls_shift); + dd->ib_arm = (INFINIPATH_IBCS_LT_STATE_LINKUP << + INFINIPATH_IBCS_LINKTRAININGSTATE_SHIFT) | + (INFINIPATH_IBCS_L_STATE_ARM << dd->ibcs_ls_shift); + dd->ib_active = (INFINIPATH_IBCS_LT_STATE_LINKUP << + INFINIPATH_IBCS_LINKTRAININGSTATE_SHIFT) | + (INFINIPATH_IBCS_L_STATE_ACTIVE << dd->ibcs_ls_shift); + + /* + * Fill in data for ibcc field-values that change in IBA7220. + * We dynamically specify only the mask for LINKINITCMD + * and only the shift for LINKCMD and MAXPKTLEN, as they are + * the only ones that change. + */ + dd->ibcc_lic_mask = IBA7220_IBCC_LINKINITCMD_MASK; + dd->ibcc_lc_shift = IBA7220_IBCC_LINKCMD_SHIFT; + dd->ibcc_mpl_shift = IBA7220_IBCC_MAXPKTLEN_SHIFT; + + /* Fill in shifts for RcvCtrl. */ + dd->ipath_r_portenable_shift = INFINIPATH_R_PORTENABLE_SHIFT; + dd->ipath_r_intravail_shift = IBA7220_R_INTRAVAIL_SHIFT; + dd->ipath_r_tailupd_shift = IBA7220_R_TAILUPD_SHIFT; + dd->ipath_r_portcfg_shift = IBA7220_R_PORTCFG_SHIFT; + + /* variables for sanity checking interrupt and errors */ + dd->ipath_hwe_bitsextant = + (INFINIPATH_HWE_RXEMEMPARITYERR_MASK << + INFINIPATH_HWE_RXEMEMPARITYERR_SHIFT) | + (INFINIPATH_HWE_TXEMEMPARITYERR_MASK << + INFINIPATH_HWE_TXEMEMPARITYERR_SHIFT) | + (INFINIPATH_HWE_PCIEMEMPARITYERR_MASK << + INFINIPATH_HWE_PCIEMEMPARITYERR_SHIFT) | + INFINIPATH_HWE_PCIE1PLLFAILED | + INFINIPATH_HWE_PCIE0PLLFAILED | + INFINIPATH_HWE_PCIEPOISONEDTLP | + INFINIPATH_HWE_PCIECPLTIMEOUT | + INFINIPATH_HWE_PCIEBUSPARITYXTLH | + INFINIPATH_HWE_PCIEBUSPARITYXADM | + INFINIPATH_HWE_PCIEBUSPARITYRADM | + INFINIPATH_HWE_MEMBISTFAILED | + INFINIPATH_HWE_COREPLL_FBSLIP | + INFINIPATH_HWE_COREPLL_RFSLIP | + INFINIPATH_HWE_SERDESPLLFAILED | + INFINIPATH_HWE_IBCBUSTOSPCPARITYERR | + INFINIPATH_HWE_IBCBUSFRSPCPARITYERR | + INFINIPATH_HWE_PCIECPLDATAQUEUEERR | + INFINIPATH_HWE_PCIECPLHDRQUEUEERR | + INFINIPATH_HWE_SDMAMEMREADERR | + INFINIPATH_HWE_CLK_UC_PLLNOTLOCKED | + INFINIPATH_HWE_PCIESERDESQ0PCLKNOTDETECT | + INFINIPATH_HWE_PCIESERDESQ1PCLKNOTDETECT | + INFINIPATH_HWE_PCIESERDESQ2PCLKNOTDETECT | + INFINIPATH_HWE_PCIESERDESQ3PCLKNOTDETECT | + INFINIPATH_HWE_DDSRXEQMEMORYPARITYERR | + INFINIPATH_HWE_IB_UC_MEMORYPARITYERR | + INFINIPATH_HWE_PCIE_UC_OCT0MEMORYPARITYERR | + INFINIPATH_HWE_PCIE_UC_OCT1MEMORYPARITYERR; + dd->ipath_i_bitsextant = + INFINIPATH_I_SDMAINT | INFINIPATH_I_SDMADISABLED | + (INFINIPATH_I_RCVURG_MASK << INFINIPATH_I_RCVURG_SHIFT) | + (INFINIPATH_I_RCVAVAIL_MASK << + INFINIPATH_I_RCVAVAIL_SHIFT) | + INFINIPATH_I_ERROR | INFINIPATH_I_SPIOSENT | + INFINIPATH_I_SPIOBUFAVAIL | INFINIPATH_I_GPIO | + INFINIPATH_I_JINT | INFINIPATH_I_SERDESTRIMDONE; + dd->ipath_e_bitsextant = + INFINIPATH_E_RFORMATERR | INFINIPATH_E_RVCRC | + INFINIPATH_E_RICRC | INFINIPATH_E_RMINPKTLEN | + INFINIPATH_E_RMAXPKTLEN | INFINIPATH_E_RLONGPKTLEN | + INFINIPATH_E_RSHORTPKTLEN | INFINIPATH_E_RUNEXPCHAR | + INFINIPATH_E_RUNSUPVL | INFINIPATH_E_REBP | + INFINIPATH_E_RIBFLOW | INFINIPATH_E_RBADVERSION | + INFINIPATH_E_RRCVEGRFULL | INFINIPATH_E_RRCVHDRFULL | + INFINIPATH_E_RBADTID | INFINIPATH_E_RHDRLEN | + INFINIPATH_E_RHDR | INFINIPATH_E_RIBLOSTLINK | + INFINIPATH_E_SENDSPECIALTRIGGER | + INFINIPATH_E_SDMADISABLED | INFINIPATH_E_SMINPKTLEN | + INFINIPATH_E_SMAXPKTLEN | INFINIPATH_E_SUNDERRUN | + INFINIPATH_E_SPKTLEN | INFINIPATH_E_SDROPPEDSMPPKT | + INFINIPATH_E_SDROPPEDDATAPKT | + INFINIPATH_E_SPIOARMLAUNCH | INFINIPATH_E_SUNEXPERRPKTNUM | + INFINIPATH_E_SUNSUPVL | INFINIPATH_E_SENDBUFMISUSE | + INFINIPATH_E_SDMAGENMISMATCH | INFINIPATH_E_SDMAOUTOFBOUND | + INFINIPATH_E_SDMATAILOUTOFBOUND | INFINIPATH_E_SDMABASE | + INFINIPATH_E_SDMA1STDESC | INFINIPATH_E_SDMARPYTAG | + INFINIPATH_E_SDMADWEN | INFINIPATH_E_SDMAMISSINGDW | + INFINIPATH_E_SDMAUNEXPDATA | + INFINIPATH_E_IBSTATUSCHANGED | INFINIPATH_E_INVALIDADDR | + INFINIPATH_E_RESET | INFINIPATH_E_HARDWARE | + INFINIPATH_E_SDMADESCADDRMISALIGN | + INFINIPATH_E_INVALIDEEPCMD; + + dd->ipath_i_rcvavail_mask = INFINIPATH_I_RCVAVAIL_MASK; + dd->ipath_i_rcvurg_mask = INFINIPATH_I_RCVURG_MASK; + dd->ipath_i_rcvavail_shift = INFINIPATH_I_RCVAVAIL_SHIFT; + dd->ipath_i_rcvurg_shift = INFINIPATH_I_RCVURG_SHIFT; + dd->ipath_flags |= IPATH_INTREG_64 | IPATH_HAS_MULT_IB_SPEED + | IPATH_HAS_LINK_LATENCY; + + /* + * EEPROM error log 0 is TXE Parity errors. 1 is RXE Parity. + * 2 is Some Misc, 3 is reserved for future. + */ + dd->ipath_eep_st_masks[0].hwerrs_to_log = + INFINIPATH_HWE_TXEMEMPARITYERR_MASK << + INFINIPATH_HWE_TXEMEMPARITYERR_SHIFT; + + dd->ipath_eep_st_masks[1].hwerrs_to_log = + INFINIPATH_HWE_RXEMEMPARITYERR_MASK << + INFINIPATH_HWE_RXEMEMPARITYERR_SHIFT; + + dd->ipath_eep_st_masks[2].errs_to_log = INFINIPATH_E_RESET; + + ipath_linkrecovery = 0; + + init_waitqueue_head(&dd->ipath_autoneg_wait); + INIT_DELAYED_WORK(&dd->ipath_autoneg_work, autoneg_work); + + dd->ipath_link_width_supported = IB_WIDTH_1X | IB_WIDTH_4X; + dd->ipath_link_speed_supported = IPATH_IB_SDR | IPATH_IB_DDR; + + dd->ipath_link_width_enabled = dd->ipath_link_width_supported; + dd->ipath_link_speed_enabled = dd->ipath_link_speed_supported; + /* + * set the initial values to reasonable default, will be set + * for real when link is up. + */ + dd->ipath_link_width_active = IB_WIDTH_4X; + dd->ipath_link_speed_active = IPATH_IB_SDR; + dd->delay_mult = rate_to_delay[0][1]; +} + + +/* + * Setup the MSI stuff again after a reset. I'd like to just call + * pci_enable_msi() and request_irq() again, but when I do that, + * the MSI enable bit doesn't get set in the command word, and + * we switch to to a different interrupt vector, which is confusing, + * so I instead just do it all inline. Perhaps somehow can tie this + * into the PCIe hotplug support at some point + * Note, because I'm doing it all here, I don't call pci_disable_msi() + * or free_irq() at the start of ipath_setup_7220_reset(). + */ +static int ipath_reinit_msi(struct ipath_devdata *dd) +{ + int ret = 0; +#ifdef CONFIG_PCI_MSI + int pos; + u16 control; + if (!dd->ipath_msi_lo) /* Using intX, or init problem */ + goto bail; + + pos = pci_find_capability(dd->pcidev, PCI_CAP_ID_MSI); + if (!pos) { + ipath_dev_err(dd, "Can't find MSI capability, " + "can't restore MSI settings\n"); + goto bail; + } + ipath_cdbg(VERBOSE, "Writing msi_lo 0x%x to config offset 0x%x\n", + dd->ipath_msi_lo, pos + PCI_MSI_ADDRESS_LO); + pci_write_config_dword(dd->pcidev, pos + PCI_MSI_ADDRESS_LO, + dd->ipath_msi_lo); + ipath_cdbg(VERBOSE, "Writing msi_lo 0x%x to config offset 0x%x\n", + dd->ipath_msi_hi, pos + PCI_MSI_ADDRESS_HI); + pci_write_config_dword(dd->pcidev, pos + PCI_MSI_ADDRESS_HI, + dd->ipath_msi_hi); + pci_read_config_word(dd->pcidev, pos + PCI_MSI_FLAGS, &control); + if (!(control & PCI_MSI_FLAGS_ENABLE)) { + ipath_cdbg(VERBOSE, "MSI control at off %x was %x, " + "setting MSI enable (%x)\n", pos + PCI_MSI_FLAGS, + control, control | PCI_MSI_FLAGS_ENABLE); + control |= PCI_MSI_FLAGS_ENABLE; + pci_write_config_word(dd->pcidev, pos + PCI_MSI_FLAGS, + control); + } + /* now rewrite the data (vector) info */ + pci_write_config_word(dd->pcidev, pos + + ((control & PCI_MSI_FLAGS_64BIT) ? 12 : 8), + dd->ipath_msi_data); + ret = 1; +bail: +#endif + if (!ret) { + ipath_dbg("Using IntX, MSI disabled or not configured\n"); + ipath_enable_intx(dd->pcidev); + ret = 1; + } + /* + * We restore the cachelinesize also, although it doesn't really + * matter. + */ + pci_write_config_byte(dd->pcidev, PCI_CACHE_LINE_SIZE, + dd->ipath_pci_cacheline); + /* and now set the pci master bit again */ + pci_set_master(dd->pcidev); + + return ret; +} + +/* + * This routine sleeps, so it can only be called from user context, not + * from interrupt context. If we need interrupt context, we can split + * it into two routines. + */ +static int ipath_setup_7220_reset(struct ipath_devdata *dd) +{ + u64 val; + int i; + int ret; + u16 cmdval; + + pci_read_config_word(dd->pcidev, PCI_COMMAND, &cmdval); + + /* Use dev_err so it shows up in logs, etc. */ + ipath_dev_err(dd, "Resetting InfiniPath unit %u\n", dd->ipath_unit); + + /* keep chip from being accessed in a few places */ + dd->ipath_flags &= ~(IPATH_INITTED | IPATH_PRESENT); + val = dd->ipath_control | INFINIPATH_C_RESET; + ipath_write_kreg(dd, dd->ipath_kregs->kr_control, val); + mb(); + + for (i = 1; i <= 5; i++) { + int r; + + /* + * Allow MBIST, etc. to complete; longer on each retry. + * We sometimes get machine checks from bus timeout if no + * response, so for now, make it *really* long. + */ + msleep(1000 + (1 + i) * 2000); + r = pci_write_config_dword(dd->pcidev, PCI_BASE_ADDRESS_0, + dd->ipath_pcibar0); + if (r) + ipath_dev_err(dd, "rewrite of BAR0 failed: %d\n", r); + r = pci_write_config_dword(dd->pcidev, PCI_BASE_ADDRESS_1, + dd->ipath_pcibar1); + if (r) + ipath_dev_err(dd, "rewrite of BAR1 failed: %d\n", r); + /* now re-enable memory access */ + pci_write_config_word(dd->pcidev, PCI_COMMAND, cmdval); + r = pci_enable_device(dd->pcidev); + if (r) + ipath_dev_err(dd, "pci_enable_device failed after " + "reset: %d\n", r); + /* + * whether it fully enabled or not, mark as present, + * again (but not INITTED) + */ + dd->ipath_flags |= IPATH_PRESENT; + val = ipath_read_kreg64(dd, dd->ipath_kregs->kr_revision); + if (val == dd->ipath_revision) { + ipath_cdbg(VERBOSE, "Got matching revision " + "register %llx on try %d\n", + (unsigned long long) val, i); + ret = ipath_reinit_msi(dd); + goto bail; + } + /* Probably getting -1 back */ + ipath_dbg("Didn't get expected revision register, " + "got %llx, try %d\n", (unsigned long long) val, + i + 1); + } + ret = 0; /* failed */ + +bail: + if (ret) + ipath_7220_pcie_params(dd, dd->ipath_boardrev); + + return ret; +} + +/** + * ipath_7220_put_tid - write a TID to the chip + * @dd: the infinipath device + * @tidptr: pointer to the expected TID (in chip) to udpate + * @tidtype: 0 for eager, 1 for expected + * @pa: physical address of in memory buffer; ipath_tidinvalid if freeing + * + * This exists as a separate routine to allow for selection of the + * appropriate "flavor". The static calls in cleanup just use the + * revision-agnostic form, as they are not performance critical. + */ +static void ipath_7220_put_tid(struct ipath_devdata *dd, u64 __iomem *tidptr, + u32 type, unsigned long pa) +{ + if (pa != dd->ipath_tidinvalid) { + u64 chippa = pa >> IBA7220_TID_PA_SHIFT; + + /* paranoia checks */ + if (pa != (chippa << IBA7220_TID_PA_SHIFT)) { + dev_info(&dd->pcidev->dev, "BUG: physaddr %lx " + "not 2KB aligned!\n", pa); + return; + } + if (pa >= (1UL << IBA7220_TID_SZ_SHIFT)) { + ipath_dev_err(dd, + "BUG: Physical page address 0x%lx " + "larger than supported\n", pa); + return; + } + + if (type == RCVHQ_RCV_TYPE_EAGER) + chippa |= dd->ipath_tidtemplate; + else /* for now, always full 4KB page */ + chippa |= IBA7220_TID_SZ_4K; + writeq(chippa, tidptr); + } else + writeq(pa, tidptr); + mmiowb(); +} + +/** + * ipath_7220_clear_tid - clear all TID entries for a port, expected and eager + * @dd: the infinipath device + * @port: the port + * + * clear all TID entries for a port, expected and eager. + * Used from ipath_close(). On this chip, TIDs are only 32 bits, + * not 64, but they are still on 64 bit boundaries, so tidbase + * is declared as u64 * for the pointer math, even though we write 32 bits + */ +static void ipath_7220_clear_tids(struct ipath_devdata *dd, unsigned port) +{ + u64 __iomem *tidbase; + unsigned long tidinv; + int i; + + if (!dd->ipath_kregbase) + return; + + ipath_cdbg(VERBOSE, "Invalidate TIDs for port %u\n", port); + + tidinv = dd->ipath_tidinvalid; + tidbase = (u64 __iomem *) + ((char __iomem *)(dd->ipath_kregbase) + + dd->ipath_rcvtidbase + + port * dd->ipath_rcvtidcnt * sizeof(*tidbase)); + + for (i = 0; i < dd->ipath_rcvtidcnt; i++) + ipath_7220_put_tid(dd, &tidbase[i], RCVHQ_RCV_TYPE_EXPECTED, + tidinv); + + tidbase = (u64 __iomem *) + ((char __iomem *)(dd->ipath_kregbase) + + dd->ipath_rcvegrbase + port_egrtid_idx(dd, port) + * sizeof(*tidbase)); + + for (i = port ? dd->ipath_rcvegrcnt : dd->ipath_p0_rcvegrcnt; i; i--) + ipath_7220_put_tid(dd, &tidbase[i-1], RCVHQ_RCV_TYPE_EAGER, + tidinv); +} + +/** + * ipath_7220_tidtemplate - setup constants for TID updates + * @dd: the infinipath device + * + * We setup stuff that we use a lot, to avoid calculating each time + */ +static void ipath_7220_tidtemplate(struct ipath_devdata *dd) +{ + /* For now, we always allocate 4KB buffers (at init) so we can + * receive max size packets. We may want a module parameter to + * specify 2KB or 4KB and/or make be per port instead of per device + * for those who want to reduce memory footprint. Note that the + * ipath_rcvhdrentsize size must be large enough to hold the largest + * IB header (currently 96 bytes) that we expect to handle (plus of + * course the 2 dwords of RHF). + */ + if (dd->ipath_rcvegrbufsize == 2048) + dd->ipath_tidtemplate = IBA7220_TID_SZ_2K; + else if (dd->ipath_rcvegrbufsize == 4096) + dd->ipath_tidtemplate = IBA7220_TID_SZ_4K; + else { + dev_info(&dd->pcidev->dev, "BUG: unsupported egrbufsize " + "%u, using %u\n", dd->ipath_rcvegrbufsize, + 4096); + dd->ipath_tidtemplate = IBA7220_TID_SZ_4K; + } + dd->ipath_tidinvalid = 0; +} + +static int ipath_7220_early_init(struct ipath_devdata *dd) +{ + u32 i, s; + + if (strcmp(int_type, "auto") && + strcmp(int_type, "force_msi") && + strcmp(int_type, "force_intx")) { + ipath_dev_err(dd, "Invalid interrupt_type: '%s', expecting " + "auto, force_msi or force_intx\n", int_type); + return -EINVAL; + } + + /* + * Control[4] has been added to change the arbitration within + * the SDMA engine between favoring data fetches over descriptor + * fetches. ipath_sdma_fetch_arb==0 gives data fetches priority. + */ + if (ipath_sdma_fetch_arb && (dd->ipath_minrev > 1)) + dd->ipath_control |= 1<<4; + + dd->ipath_flags |= IPATH_4BYTE_TID; + + /* + * For openfabrics, we need to be able to handle an IB header of + * 24 dwords. HT chip has arbitrary sized receive buffers, so we + * made them the same size as the PIO buffers. This chip does not + * handle arbitrary size buffers, so we need the header large enough + * to handle largest IB header, but still have room for a 2KB MTU + * standard IB packet. + */ + dd->ipath_rcvhdrentsize = 24; + dd->ipath_rcvhdrsize = IPATH_DFLT_RCVHDRSIZE; + dd->ipath_rhf_offset = + dd->ipath_rcvhdrentsize - sizeof(u64) / sizeof(u32); + + dd->ipath_rcvegrbufsize = ipath_mtu4096 ? 4096 : 2048; + /* + * the min() check here is currently a nop, but it may not always + * be, depending on just how we do ipath_rcvegrbufsize + */ + dd->ipath_ibmaxlen = min(ipath_mtu4096 ? dd->ipath_piosize4k : + dd->ipath_piosize2k, + dd->ipath_rcvegrbufsize + + (dd->ipath_rcvhdrentsize << 2)); + dd->ipath_init_ibmaxlen = dd->ipath_ibmaxlen; + + ipath_7220_config_jint(dd, INFINIPATH_JINT_DEFAULT_IDLE_TICKS, + INFINIPATH_JINT_DEFAULT_MAX_PACKETS); + + if (dd->ipath_boardrev) /* no eeprom on emulator */ + ipath_get_eeprom_info(dd); + + /* start of code to check and print procmon */ + s = ipath_read_kreg32(dd, IPATH_KREG_OFFSET(ProcMon)); + s &= ~(1U<<31); /* clear done bit */ + s |= 1U<<14; /* clear counter (write 1 to clear) */ + ipath_write_kreg(dd, IPATH_KREG_OFFSET(ProcMon), s); + /* make sure clear_counter low long enough before start */ + ipath_read_kreg32(dd, dd->ipath_kregs->kr_scratch); + ipath_read_kreg32(dd, dd->ipath_kregs->kr_scratch); + + s &= ~(1U<<14); /* allow counter to count (before starting) */ + ipath_write_kreg(dd, IPATH_KREG_OFFSET(ProcMon), s); + ipath_read_kreg32(dd, dd->ipath_kregs->kr_scratch); + ipath_read_kreg32(dd, dd->ipath_kregs->kr_scratch); + s = ipath_read_kreg32(dd, IPATH_KREG_OFFSET(ProcMon)); + + s |= 1U<<15; /* start the counter */ + s &= ~(1U<<31); /* clear done bit */ + s &= ~0x7ffU; /* clear frequency bits */ + s |= 0xe29; /* set frequency bits, in case cleared */ + ipath_write_kreg(dd, IPATH_KREG_OFFSET(ProcMon), s); + + s = 0; + for (i = 500; i > 0 && !(s&(1ULL<<31)); i--) { + ipath_read_kreg32(dd, dd->ipath_kregs->kr_scratch); + s = ipath_read_kreg32(dd, IPATH_KREG_OFFSET(ProcMon)); + } + if (!(s&(1U<<31))) + ipath_dev_err(dd, "ProcMon register not valid: 0x%x\n", s); + else + ipath_dbg("ProcMon=0x%x, count=0x%x\n", s, (s>>16)&0x1ff); + + return 0; +} + +/** + * ipath_init_7220_get_base_info - set chip-specific flags for user code + * @pd: the infinipath port + * @kbase: ipath_base_info pointer + * + * We set the PCIE flag because the lower bandwidth on PCIe vs + * HyperTransport can affect some user packet algorithims. + */ +static int ipath_7220_get_base_info(struct ipath_portdata *pd, void *kbase) +{ + struct ipath_base_info *kinfo = kbase; + + kinfo->spi_runtime_flags |= + IPATH_RUNTIME_PCIE | IPATH_RUNTIME_NODMA_RTAIL | + IPATH_RUNTIME_SDMA; + + return 0; +} + +static void ipath_7220_free_irq(struct ipath_devdata *dd) +{ + free_irq(dd->ipath_irq, dd); + dd->ipath_irq = 0; +} + +static struct ipath_message_header * +ipath_7220_get_msgheader(struct ipath_devdata *dd, __le32 *rhf_addr) +{ + u32 offset = ipath_hdrget_offset(rhf_addr); + + return (struct ipath_message_header *) + (rhf_addr - dd->ipath_rhf_offset + offset); +} + +static void ipath_7220_config_ports(struct ipath_devdata *dd, ushort cfgports) +{ + u32 nchipports; + + nchipports = ipath_read_kreg32(dd, dd->ipath_kregs->kr_portcnt); + if (!cfgports) { + int ncpus = num_online_cpus(); + + if (ncpus <= 4) + dd->ipath_portcnt = 5; + else if (ncpus <= 8) + dd->ipath_portcnt = 9; + if (dd->ipath_portcnt) + ipath_dbg("Auto-configured for %u ports, %d cpus " + "online\n", dd->ipath_portcnt, ncpus); + } else if (cfgports <= nchipports) + dd->ipath_portcnt = cfgports; + if (!dd->ipath_portcnt) /* none of the above, set to max */ + dd->ipath_portcnt = nchipports; + /* + * chip can be configured for 5, 9, or 17 ports, and choice + * affects number of eager TIDs per port (1K, 2K, 4K). + */ + if (dd->ipath_portcnt > 9) + dd->ipath_rcvctrl |= 2ULL << IBA7220_R_PORTCFG_SHIFT; + else if (dd->ipath_portcnt > 5) + dd->ipath_rcvctrl |= 1ULL << IBA7220_R_PORTCFG_SHIFT; + /* else configure for default 5 receive ports */ + ipath_write_kreg(dd, dd->ipath_kregs->kr_rcvctrl, + dd->ipath_rcvctrl); + dd->ipath_p0_rcvegrcnt = 2048; /* always */ + if (dd->ipath_flags & IPATH_HAS_SEND_DMA) + dd->ipath_pioreserved = 1; /* reserve a buffer */ +} + + +static int ipath_7220_get_ib_cfg(struct ipath_devdata *dd, int which) +{ + int lsb, ret = 0; + u64 maskr; /* right-justified mask */ + + switch (which) { + case IPATH_IB_CFG_HRTBT: /* Get Heartbeat off/enable/auto */ + lsb = IBA7220_IBC_HRTBT_SHIFT; + maskr = IBA7220_IBC_HRTBT_MASK; + break; + + case IPATH_IB_CFG_LWID_ENB: /* Get allowed Link-width */ + ret = dd->ipath_link_width_enabled; + goto done; + + case IPATH_IB_CFG_LWID: /* Get currently active Link-width */ + ret = dd->ipath_link_width_active; + goto done; + + case IPATH_IB_CFG_SPD_ENB: /* Get allowed Link speeds */ + ret = dd->ipath_link_speed_enabled; + goto done; + + case IPATH_IB_CFG_SPD: /* Get current Link spd */ + ret = dd->ipath_link_speed_active; + goto done; + + case IPATH_IB_CFG_RXPOL_ENB: /* Get Auto-RX-polarity enable */ + lsb = IBA7220_IBC_RXPOL_SHIFT; + maskr = IBA7220_IBC_RXPOL_MASK; + break; + + case IPATH_IB_CFG_LREV_ENB: /* Get Auto-Lane-reversal enable */ + lsb = IBA7220_IBC_LREV_SHIFT; + maskr = IBA7220_IBC_LREV_MASK; + break; + + case IPATH_IB_CFG_LINKLATENCY: + ret = ipath_read_kreg64(dd, dd->ipath_kregs->kr_ibcddrstatus) + & IBA7220_DDRSTAT_LINKLAT_MASK; + goto done; + + default: + ret = -ENOTSUPP; + goto done; + } + ret = (int)((dd->ipath_ibcddrctrl >> lsb) & maskr); +done: + return ret; +} + +static int ipath_7220_set_ib_cfg(struct ipath_devdata *dd, int which, u32 val) +{ + int lsb, ret = 0, setforce = 0; + u64 maskr; /* right-justified mask */ + + switch (which) { + case IPATH_IB_CFG_LIDLMC: + /* + * Set LID and LMC. Combined to avoid possible hazard + * caller puts LMC in 16MSbits, DLID in 16LSbits of val + */ + lsb = IBA7220_IBC_DLIDLMC_SHIFT; + maskr = IBA7220_IBC_DLIDLMC_MASK; + break; + + case IPATH_IB_CFG_HRTBT: /* set Heartbeat off/enable/auto */ + if (val & IPATH_IB_HRTBT_ON && + (dd->ipath_flags & IPATH_NO_HRTBT)) + goto bail; + lsb = IBA7220_IBC_HRTBT_SHIFT; + maskr = IBA7220_IBC_HRTBT_MASK; + break; + + case IPATH_IB_CFG_LWID_ENB: /* set allowed Link-width */ + /* + * As with speed, only write the actual register if + * the link is currently down, otherwise takes effect + * on next link change. + */ + dd->ipath_link_width_enabled = val; + if ((dd->ipath_flags & (IPATH_LINKDOWN|IPATH_LINKINIT)) != + IPATH_LINKDOWN) + goto bail; + /* + * We set the IPATH_IB_FORCE_NOTIFY bit so updown + * will get called because we want update + * link_width_active, and the change may not take + * effect for some time (if we are in POLL), so this + * flag will force the updown routine to be called + * on the next ibstatuschange down interrupt, even + * if it's not an down->up transition. + */ + val--; /* convert from IB to chip */ + maskr = IBA7220_IBC_WIDTH_MASK; + lsb = IBA7220_IBC_WIDTH_SHIFT; + setforce = 1; + dd->ipath_flags |= IPATH_IB_FORCE_NOTIFY; + break; + + case IPATH_IB_CFG_SPD_ENB: /* set allowed Link speeds */ + /* + * If we turn off IB1.2, need to preset SerDes defaults, + * but not right now. Set a flag for the next time + * we command the link down. As with width, only write the + * actual register if the link is currently down, otherwise + * takes effect on next link change. Since setting is being + * explictly requested (via MAD or sysfs), clear autoneg + * failure status if speed autoneg is enabled. + */ + dd->ipath_link_speed_enabled = val; + if (dd->ipath_ibcddrctrl & IBA7220_IBC_IBTA_1_2_MASK && + !(val & (val - 1))) + dd->ipath_presets_needed = 1; + if ((dd->ipath_flags & (IPATH_LINKDOWN|IPATH_LINKINIT)) != + IPATH_LINKDOWN) + goto bail; + /* + * We set the IPATH_IB_FORCE_NOTIFY bit so updown + * will get called because we want update + * link_speed_active, and the change may not take + * effect for some time (if we are in POLL), so this + * flag will force the updown routine to be called + * on the next ibstatuschange down interrupt, even + * if it's not an down->up transition. When setting + * speed autoneg, clear AUTONEG_FAILED. + */ + if (val == (IPATH_IB_SDR | IPATH_IB_DDR)) { + val = IBA7220_IBC_SPEED_AUTONEG_MASK | + IBA7220_IBC_IBTA_1_2_MASK; + dd->ipath_flags &= ~IPATH_IB_AUTONEG_FAILED; + } else + val = val == IPATH_IB_DDR ? IBA7220_IBC_SPEED_DDR + : IBA7220_IBC_SPEED_SDR; + maskr = IBA7220_IBC_SPEED_AUTONEG_MASK | + IBA7220_IBC_IBTA_1_2_MASK; + lsb = 0; /* speed bits are low bits */ + setforce = 1; + break; + + case IPATH_IB_CFG_RXPOL_ENB: /* set Auto-RX-polarity enable */ + lsb = IBA7220_IBC_RXPOL_SHIFT; + maskr = IBA7220_IBC_RXPOL_MASK; + break; + + case IPATH_IB_CFG_LREV_ENB: /* set Auto-Lane-reversal enable */ + lsb = IBA7220_IBC_LREV_SHIFT; + maskr = IBA7220_IBC_LREV_MASK; + break; + + default: + ret = -ENOTSUPP; + goto bail; + } + dd->ipath_ibcddrctrl &= ~(maskr << lsb); + dd->ipath_ibcddrctrl |= (((u64) val & maskr) << lsb); + ipath_write_kreg(dd, dd->ipath_kregs->kr_ibcddrctrl, + dd->ipath_ibcddrctrl); + if (setforce) + dd->ipath_flags |= IPATH_IB_FORCE_NOTIFY; +bail: + return ret; +} + +static void ipath_7220_read_counters(struct ipath_devdata *dd, + struct infinipath_counters *cntrs) +{ + u64 *counters = (u64 *) cntrs; + int i; + + for (i = 0; i < sizeof(*cntrs) / sizeof(u64); i++) + counters[i] = ipath_snap_cntr(dd, i); +} + +/* if we are using MSI, try to fallback to IntX */ +static int ipath_7220_intr_fallback(struct ipath_devdata *dd) +{ + if (dd->ipath_msi_lo) { + dev_info(&dd->pcidev->dev, "MSI interrupt not detected," + " trying IntX interrupts\n"); + ipath_7220_nomsi(dd); + ipath_enable_intx(dd->pcidev); + /* + * some newer kernels require free_irq before disable_msi, + * and irq can be changed during disable and intx enable + * and we need to therefore use the pcidev->irq value, + * not our saved MSI value. + */ + dd->ipath_irq = dd->pcidev->irq; + if (request_irq(dd->ipath_irq, ipath_intr, IRQF_SHARED, + IPATH_DRV_NAME, dd)) + ipath_dev_err(dd, + "Could not re-request_irq for IntX\n"); + return 1; + } + return 0; +} + +/* + * reset the XGXS (between serdes and IBC). Slightly less intrusive + * than resetting the IBC or external link state, and useful in some + * cases to cause some retraining. To do this right, we reset IBC + * as well. + */ +static void ipath_7220_xgxs_reset(struct ipath_devdata *dd) +{ + u64 val, prev_val; + + prev_val = ipath_read_kreg64(dd, dd->ipath_kregs->kr_xgxsconfig); + val = prev_val | INFINIPATH_XGXS_RESET; + prev_val &= ~INFINIPATH_XGXS_RESET; /* be sure */ + ipath_write_kreg(dd, dd->ipath_kregs->kr_control, + dd->ipath_control & ~INFINIPATH_C_LINKENABLE); + ipath_write_kreg(dd, dd->ipath_kregs->kr_xgxsconfig, val); + ipath_read_kreg32(dd, dd->ipath_kregs->kr_scratch); + ipath_write_kreg(dd, dd->ipath_kregs->kr_xgxsconfig, prev_val); + ipath_write_kreg(dd, dd->ipath_kregs->kr_control, + dd->ipath_control); +} + + +/* Still needs cleanup, too much hardwired stuff */ +static void autoneg_send(struct ipath_devdata *dd, + u32 *hdr, u32 dcnt, u32 *data) +{ + int i; + u64 cnt; + u32 __iomem *piobuf; + u32 pnum; + + i = 0; + cnt = 7 + dcnt + 1; /* 7 dword header, dword data, icrc */ + while (!(piobuf = ipath_getpiobuf(dd, cnt, &pnum))) { + if (i++ > 15) { + ipath_dbg("Couldn't get pio buffer for send\n"); + return; + } + udelay(2); + } + if (dd->ipath_flags&IPATH_HAS_PBC_CNT) + cnt |= 0x80000000UL<<32; /* mark as VL15 */ + writeq(cnt, piobuf); + ipath_flush_wc(); + __iowrite32_copy(piobuf + 2, hdr, 7); + __iowrite32_copy(piobuf + 9, data, dcnt); + ipath_flush_wc(); +} + +/* + * _start packet gets sent twice at start, _done gets sent twice at end + */ +static void ipath_autoneg_send(struct ipath_devdata *dd, int which) +{ + static u32 swapped; + u32 dw, i, hcnt, dcnt, *data; + static u32 hdr[7] = { 0xf002ffff, 0x48ffff, 0x6400abba }; + static u32 madpayload_start[0x40] = { + 0x1810103, 0x1, 0x0, 0x0, 0x2c90000, 0x2c9, 0x0, 0x0, + 0xffffffff, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, + 0x1, 0x1388, 0x15e, 0x1, /* rest 0's */ + }; + static u32 madpayload_done[0x40] = { + 0x1810103, 0x1, 0x0, 0x0, 0x2c90000, 0x2c9, 0x0, 0x0, + 0xffffffff, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, + 0x40000001, 0x1388, 0x15e, /* rest 0's */ + }; + dcnt = sizeof(madpayload_start)/sizeof(madpayload_start[0]); + hcnt = sizeof(hdr)/sizeof(hdr[0]); + if (!swapped) { + /* for maintainability, do it at runtime */ + for (i = 0; i < hcnt; i++) { + dw = (__force u32) cpu_to_be32(hdr[i]); + hdr[i] = dw; + } + for (i = 0; i < dcnt; i++) { + dw = (__force u32) cpu_to_be32(madpayload_start[i]); + madpayload_start[i] = dw; + dw = (__force u32) cpu_to_be32(madpayload_done[i]); + madpayload_done[i] = dw; + } + swapped = 1; + } + + data = which ? madpayload_done : madpayload_start; + ipath_cdbg(PKT, "Sending %s special MADs\n", which?"done":"start"); + + autoneg_send(dd, hdr, dcnt, data); + ipath_read_kreg64(dd, dd->ipath_kregs->kr_scratch); + udelay(2); + autoneg_send(dd, hdr, dcnt, data); + ipath_read_kreg64(dd, dd->ipath_kregs->kr_scratch); + udelay(2); +} + + + +/* + * Do the absolute minimum to cause an IB speed change, and make it + * ready, but don't actually trigger the change. The caller will + * do that when ready (if link is in Polling training state, it will + * happen immediately, otherwise when link next goes down) + * + * This routine should only be used as part of the DDR autonegotation + * code for devices that are not compliant with IB 1.2 (or code that + * fixes things up for same). + * + * When link has gone down, and autoneg enabled, or autoneg has + * failed and we give up until next time we set both speeds, and + * then we want IBTA enabled as well as "use max enabled speed. + */ +static void set_speed_fast(struct ipath_devdata *dd, u32 speed) +{ + dd->ipath_ibcddrctrl &= ~(IBA7220_IBC_SPEED_AUTONEG_MASK | + IBA7220_IBC_IBTA_1_2_MASK | + (IBA7220_IBC_WIDTH_MASK << IBA7220_IBC_WIDTH_SHIFT)); + + if (speed == (IPATH_IB_SDR | IPATH_IB_DDR)) + dd->ipath_ibcddrctrl |= IBA7220_IBC_SPEED_AUTONEG_MASK | + IBA7220_IBC_IBTA_1_2_MASK; + else + dd->ipath_ibcddrctrl |= speed == IPATH_IB_DDR ? + IBA7220_IBC_SPEED_DDR : IBA7220_IBC_SPEED_SDR; + + /* + * Convert from IB-style 1 = 1x, 2 = 4x, 3 = auto + * to chip-centric 0 = 1x, 1 = 4x, 2 = auto + */ + dd->ipath_ibcddrctrl |= (u64)(dd->ipath_link_width_enabled - 1) << + IBA7220_IBC_WIDTH_SHIFT; + ipath_write_kreg(dd, dd->ipath_kregs->kr_ibcddrctrl, + dd->ipath_ibcddrctrl); + ipath_cdbg(VERBOSE, "setup for IB speed (%x) done\n", speed); +} + + +/* + * this routine is only used when we are not talking to another + * IB 1.2-compliant device that we think can do DDR. + * (This includes all existing switch chips as of Oct 2007.) + * 1.2-compliant devices go directly to DDR prior to reaching INIT + */ +static void try_auto_neg(struct ipath_devdata *dd) +{ + /* + * required for older non-IB1.2 DDR switches. Newer + * non-IB-compliant switches don't need it, but so far, + * aren't bothered by it either. "Magic constant" + */ + ipath_write_kreg(dd, IPATH_KREG_OFFSET(IBNCModeCtrl), + 0x3b9dc07); + dd->ipath_flags |= IPATH_IB_AUTONEG_INPROG; + ipath_autoneg_send(dd, 0); + set_speed_fast(dd, IPATH_IB_DDR); + ipath_toggle_rclkrls(dd); + /* 2 msec is minimum length of a poll cycle */ + schedule_delayed_work(&dd->ipath_autoneg_work, + msecs_to_jiffies(2)); +} + + +static int ipath_7220_ib_updown(struct ipath_devdata *dd, int ibup, u64 ibcs) +{ + int ret = 0; + u32 ltstate = ipath_ib_linkstate(dd, ibcs); + + dd->ipath_link_width_active = + ((ibcs >> IBA7220_IBCS_LINKWIDTH_SHIFT) & 1) ? + IB_WIDTH_4X : IB_WIDTH_1X; + dd->ipath_link_speed_active = + ((ibcs >> IBA7220_IBCS_LINKSPEED_SHIFT) & 1) ? + IPATH_IB_DDR : IPATH_IB_SDR; + + if (!ibup) { + /* + * when link goes down we don't want aeq running, so it + * won't't interfere with IBC training, etc., and we need + * to go back to the static SerDes preset values + */ + if (dd->ipath_x1_fix_tries && + ltstate <= INFINIPATH_IBCS_LT_STATE_SLEEPQUIET && + ltstate != INFINIPATH_IBCS_LT_STATE_LINKUP) + dd->ipath_x1_fix_tries = 0; + if (!(dd->ipath_flags & (IPATH_IB_AUTONEG_FAILED | + IPATH_IB_AUTONEG_INPROG))) + set_speed_fast(dd, dd->ipath_link_speed_enabled); + if (!(dd->ipath_flags & IPATH_IB_AUTONEG_INPROG)) { + ipath_cdbg(VERBOSE, "Setting RXEQ defaults\n"); + ipath_sd7220_presets(dd); + } + /* this might better in ipath_sd7220_presets() */ + ipath_set_relock_poll(dd, ibup); + } else { + if (ipath_compat_ddr_negotiate && + !(dd->ipath_flags & (IPATH_IB_AUTONEG_FAILED | + IPATH_IB_AUTONEG_INPROG)) && + dd->ipath_link_speed_active == IPATH_IB_SDR && + (dd->ipath_link_speed_enabled & + (IPATH_IB_DDR | IPATH_IB_SDR)) == + (IPATH_IB_DDR | IPATH_IB_SDR) && + dd->ipath_autoneg_tries < IPATH_AUTONEG_TRIES) { + /* we are SDR, and DDR auto-negotiation enabled */ + ++dd->ipath_autoneg_tries; + ipath_dbg("DDR negotiation try, %u/%u\n", + dd->ipath_autoneg_tries, + IPATH_AUTONEG_TRIES); + try_auto_neg(dd); + ret = 1; /* no other IB status change processing */ + } else if ((dd->ipath_flags & IPATH_IB_AUTONEG_INPROG) + && dd->ipath_link_speed_active == IPATH_IB_SDR) { + ipath_autoneg_send(dd, 1); + set_speed_fast(dd, IPATH_IB_DDR); + udelay(2); + ipath_toggle_rclkrls(dd); + ret = 1; /* no other IB status change processing */ + } else { + if ((dd->ipath_flags & IPATH_IB_AUTONEG_INPROG) && + (dd->ipath_link_speed_active & IPATH_IB_DDR)) { + ipath_dbg("Got to INIT with DDR autoneg\n"); + dd->ipath_flags &= ~(IPATH_IB_AUTONEG_INPROG + | IPATH_IB_AUTONEG_FAILED); + dd->ipath_autoneg_tries = 0; + /* re-enable SDR, for next link down */ + set_speed_fast(dd, + dd->ipath_link_speed_enabled); + wake_up(&dd->ipath_autoneg_wait); + } else if (dd->ipath_flags & IPATH_IB_AUTONEG_FAILED) { + /* + * clear autoneg failure flag, and do setup + * so we'll try next time link goes down and + * back to INIT (possibly connected to different + * device). + */ + ipath_dbg("INIT %sDR after autoneg failure\n", + (dd->ipath_link_speed_active & + IPATH_IB_DDR) ? "D" : "S"); + dd->ipath_flags &= ~IPATH_IB_AUTONEG_FAILED; + dd->ipath_ibcddrctrl |= + IBA7220_IBC_IBTA_1_2_MASK; + ipath_write_kreg(dd, + IPATH_KREG_OFFSET(IBNCModeCtrl), 0); + } + } + /* + * if we are in 1X, and are in autoneg width, it + * could be due to an xgxs problem, so if we haven't + * already tried, try twice to get to 4X; if we + * tried, and couldn't, report it, since it will + * probably not be what is desired. + */ + if ((dd->ipath_link_width_enabled & (IB_WIDTH_1X | + IB_WIDTH_4X)) == (IB_WIDTH_1X | IB_WIDTH_4X) + && dd->ipath_link_width_active == IB_WIDTH_1X + && dd->ipath_x1_fix_tries < 3) { + if (++dd->ipath_x1_fix_tries == 3) + dev_info(&dd->pcidev->dev, + "IB link is in 1X mode\n"); + else { + ipath_cdbg(VERBOSE, "IB 1X in " + "auto-width, try %u to be " + "sure it's really 1X; " + "ltstate %u\n", + dd->ipath_x1_fix_tries, + ltstate); + dd->ipath_f_xgxs_reset(dd); + ret = 1; /* skip other processing */ + } + } + + if (!ret) { + dd->delay_mult = rate_to_delay + [(ibcs >> IBA7220_IBCS_LINKSPEED_SHIFT) & 1] + [(ibcs >> IBA7220_IBCS_LINKWIDTH_SHIFT) & 1]; + + ipath_set_relock_poll(dd, ibup); + } + } + + if (!ret) + ipath_setup_7220_setextled(dd, ipath_ib_linkstate(dd, ibcs), + ltstate); + return ret; +} + + +/* + * Handle the empirically determined mechanism for auto-negotiation + * of DDR speed with switches. + */ +static void autoneg_work(struct work_struct *work) +{ + struct ipath_devdata *dd; + u64 startms; + u32 lastlts, i; + + dd = container_of(work, struct ipath_devdata, + ipath_autoneg_work.work); + + startms = jiffies_to_msecs(jiffies); + + /* + * busy wait for this first part, it should be at most a + * few hundred usec, since we scheduled ourselves for 2msec. + */ + for (i = 0; i < 25; i++) { + lastlts = ipath_ib_linktrstate(dd, dd->ipath_lastibcstat); + if (lastlts == INFINIPATH_IBCS_LT_STATE_POLLQUIET) { + ipath_set_linkstate(dd, IPATH_IB_LINKDOWN_DISABLE); + break; + } + udelay(100); + } + + if (!(dd->ipath_flags & IPATH_IB_AUTONEG_INPROG)) + goto done; /* we got there early or told to stop */ + + /* we expect this to timeout */ + if (wait_event_timeout(dd->ipath_autoneg_wait, + !(dd->ipath_flags & IPATH_IB_AUTONEG_INPROG), + msecs_to_jiffies(90))) + goto done; + + ipath_toggle_rclkrls(dd); + + /* we expect this to timeout */ + if (wait_event_timeout(dd->ipath_autoneg_wait, + !(dd->ipath_flags & IPATH_IB_AUTONEG_INPROG), + msecs_to_jiffies(1700))) + goto done; + + set_speed_fast(dd, IPATH_IB_SDR); + ipath_toggle_rclkrls(dd); + + /* + * wait up to 250 msec for link to train and get to INIT at DDR; + * this should terminate early + */ + wait_event_timeout(dd->ipath_autoneg_wait, + !(dd->ipath_flags & IPATH_IB_AUTONEG_INPROG), + msecs_to_jiffies(250)); +done: + if (dd->ipath_flags & IPATH_IB_AUTONEG_INPROG) { + ipath_dbg("Did not get to DDR INIT (%x) after %Lu msecs\n", + ipath_ib_state(dd, dd->ipath_lastibcstat), + jiffies_to_msecs(jiffies)-startms); + dd->ipath_flags &= ~IPATH_IB_AUTONEG_INPROG; + if (dd->ipath_autoneg_tries == IPATH_AUTONEG_TRIES) { + dd->ipath_flags |= IPATH_IB_AUTONEG_FAILED; + ipath_dbg("Giving up on DDR until next IB " + "link Down\n"); + dd->ipath_autoneg_tries = 0; + } + set_speed_fast(dd, dd->ipath_link_speed_enabled); + } +} + + +/** + * ipath_init_iba7220_funcs - set up the chip-specific function pointers + * @dd: the infinipath device + * + * This is global, and is called directly at init to set up the + * chip-specific function pointers for later use. + */ +void ipath_init_iba7220_funcs(struct ipath_devdata *dd) +{ + dd->ipath_f_intrsetup = ipath_7220_intconfig; + dd->ipath_f_bus = ipath_setup_7220_config; + dd->ipath_f_reset = ipath_setup_7220_reset; + dd->ipath_f_get_boardname = ipath_7220_boardname; + dd->ipath_f_init_hwerrors = ipath_7220_init_hwerrors; + dd->ipath_f_early_init = ipath_7220_early_init; + dd->ipath_f_handle_hwerrors = ipath_7220_handle_hwerrors; + dd->ipath_f_quiet_serdes = ipath_7220_quiet_serdes; + dd->ipath_f_bringup_serdes = ipath_7220_bringup_serdes; + dd->ipath_f_clear_tids = ipath_7220_clear_tids; + dd->ipath_f_put_tid = ipath_7220_put_tid; + dd->ipath_f_cleanup = ipath_setup_7220_cleanup; + dd->ipath_f_setextled = ipath_setup_7220_setextled; + dd->ipath_f_get_base_info = ipath_7220_get_base_info; + dd->ipath_f_free_irq = ipath_7220_free_irq; + dd->ipath_f_tidtemplate = ipath_7220_tidtemplate; + dd->ipath_f_intr_fallback = ipath_7220_intr_fallback; + dd->ipath_f_xgxs_reset = ipath_7220_xgxs_reset; + dd->ipath_f_get_ib_cfg = ipath_7220_get_ib_cfg; + dd->ipath_f_set_ib_cfg = ipath_7220_set_ib_cfg; + dd->ipath_f_config_jint = ipath_7220_config_jint; + dd->ipath_f_config_ports = ipath_7220_config_ports; + dd->ipath_f_read_counters = ipath_7220_read_counters; + dd->ipath_f_get_msgheader = ipath_7220_get_msgheader; + dd->ipath_f_ib_updown = ipath_7220_ib_updown; + + /* initialize chip-specific variables */ + ipath_init_7220_variables(dd); +} diff --git a/drivers/infiniband/hw/ipath/ipath_init_chip.c b/drivers/infiniband/hw/ipath/ipath_init_chip.c index 4471674975cd..27dd89476660 100644 --- a/drivers/infiniband/hw/ipath/ipath_init_chip.c +++ b/drivers/infiniband/hw/ipath/ipath_init_chip.c @@ -1,5 +1,5 @@ /* - * Copyright (c) 2006, 2007 QLogic Corporation. All rights reserved. + * Copyright (c) 2006, 2007, 2008 QLogic Corporation. All rights reserved. * Copyright (c) 2003, 2004, 2005, 2006 PathScale, Inc. All rights reserved. * * This software is available to you under a choice of one of two @@ -155,24 +155,13 @@ static int bringup_link(struct ipath_devdata *dd) dd->ipath_control); /* - * Note that prior to try 14 or 15 of IB, the credit scaling - * wasn't working, because it was swapped for writes with the - * 1 bit default linkstate field + * set initial max size pkt IBC will send, including ICRC; it's the + * PIO buffer size in dwords, less 1; also see ipath_set_mtu() */ + val = (dd->ipath_ibmaxlen >> 2) + 1; + ibc = val << dd->ibcc_mpl_shift; - /* ignore pbc and align word */ - val = dd->ipath_piosize2k - 2 * sizeof(u32); - /* - * for ICRC, which we only send in diag test pkt mode, and we - * don't need to worry about that for mtu - */ - val += 1; - /* - * Set the IBC maxpktlength to the size of our pio buffers the - * maxpktlength is in words. This is *not* the IB data MTU. - */ - ibc = (val / sizeof(u32)) << INFINIPATH_IBCC_MAXPKTLEN_SHIFT; - /* in KB */ + /* flowcontrolwatermark is in units of KBytes */ ibc |= 0x5ULL << INFINIPATH_IBCC_FLOWCTRLWATERMARK_SHIFT; /* * How often flowctrl sent. More or less in usecs; balance against @@ -191,10 +180,13 @@ static int bringup_link(struct ipath_devdata *dd) /* * Want to start out with both LINKCMD and LINKINITCMD in NOP * (0 and 0). Don't put linkinitcmd in ipath_ibcctrl, want that - * to stay a NOP + * to stay a NOP. Flag that we are disabled, for the (unlikely) + * case that some recovery path is trying to bring the link up + * before we are ready. */ ibc |= INFINIPATH_IBCC_LINKINITCMD_DISABLE << INFINIPATH_IBCC_LINKINITCMD_SHIFT; + dd->ipath_flags |= IPATH_IB_LINK_DISABLED; ipath_cdbg(VERBOSE, "Writing 0x%llx to ibcctrl\n", (unsigned long long) ibc); ipath_write_kreg(dd, dd->ipath_kregs->kr_ibcctrl, ibc); @@ -227,17 +219,26 @@ static struct ipath_portdata *create_portdata0(struct ipath_devdata *dd) pd->port_cnt = 1; /* The port 0 pkey table is used by the layer interface. */ pd->port_pkeys[0] = IPATH_DEFAULT_P_KEY; + pd->port_seq_cnt = 1; } return pd; } -static int init_chip_first(struct ipath_devdata *dd, - struct ipath_portdata **pdp) +static int init_chip_first(struct ipath_devdata *dd) { - struct ipath_portdata *pd = NULL; + struct ipath_portdata *pd; int ret = 0; u64 val; + spin_lock_init(&dd->ipath_kernel_tid_lock); + spin_lock_init(&dd->ipath_user_tid_lock); + spin_lock_init(&dd->ipath_sendctrl_lock); + spin_lock_init(&dd->ipath_sdma_lock); + spin_lock_init(&dd->ipath_gpio_lock); + spin_lock_init(&dd->ipath_eep_st_lock); + spin_lock_init(&dd->ipath_sdepb_lock); + mutex_init(&dd->ipath_eep_lock); + /* * skip cfgports stuff because we are not allocating memory, * and we don't want problems if the portcnt changed due to @@ -250,12 +251,14 @@ static int init_chip_first(struct ipath_devdata *dd, else if (ipath_cfgports <= dd->ipath_portcnt) { dd->ipath_cfgports = ipath_cfgports; ipath_dbg("Configured to use %u ports out of %u in chip\n", - dd->ipath_cfgports, dd->ipath_portcnt); + dd->ipath_cfgports, ipath_read_kreg32(dd, + dd->ipath_kregs->kr_portcnt)); } else { dd->ipath_cfgports = dd->ipath_portcnt; ipath_dbg("Tried to configured to use %u ports; chip " "only supports %u\n", ipath_cfgports, - dd->ipath_portcnt); + ipath_read_kreg32(dd, + dd->ipath_kregs->kr_portcnt)); } /* * Allocate full portcnt array, rather than just cfgports, because @@ -295,12 +298,9 @@ static int init_chip_first(struct ipath_devdata *dd, val = ipath_read_kreg64(dd, dd->ipath_kregs->kr_sendpiosize); dd->ipath_piosize2k = val & ~0U; dd->ipath_piosize4k = val >> 32; - /* - * Note: the chips support a maximum MTU of 4096, but the driver - * hasn't implemented this feature yet, so set the initial value - * to 2048. - */ - dd->ipath_ibmtu = 2048; + if (dd->ipath_piosize4k == 0 && ipath_mtu4096) + ipath_mtu4096 = 0; /* 4KB not supported by this chip */ + dd->ipath_ibmtu = ipath_mtu4096 ? 4096 : 2048; val = ipath_read_kreg64(dd, dd->ipath_kregs->kr_sendpiobufcnt); dd->ipath_piobcnt2k = val & ~0U; dd->ipath_piobcnt4k = val >> 32; @@ -328,43 +328,46 @@ static int init_chip_first(struct ipath_devdata *dd, else ipath_dbg("%u 2k piobufs @ %p\n", dd->ipath_piobcnt2k, dd->ipath_pio2kbase); - spin_lock_init(&dd->ipath_tid_lock); - spin_lock_init(&dd->ipath_sendctrl_lock); - spin_lock_init(&dd->ipath_gpio_lock); - spin_lock_init(&dd->ipath_eep_st_lock); - mutex_init(&dd->ipath_eep_lock); - done: - *pdp = pd; return ret; } /** * init_chip_reset - re-initialize after a reset, or enable * @dd: the infinipath device - * @pdp: output for port data * * sanity check at least some of the values after reset, and * ensure no receive or transmit (explictly, in case reset * failed */ -static int init_chip_reset(struct ipath_devdata *dd, - struct ipath_portdata **pdp) +static int init_chip_reset(struct ipath_devdata *dd) { u32 rtmp; + int i; + unsigned long flags; + + /* + * ensure chip does no sends or receives, tail updates, or + * pioavail updates while we re-initialize + */ + dd->ipath_rcvctrl &= ~(1ULL << dd->ipath_r_tailupd_shift); + for (i = 0; i < dd->ipath_portcnt; i++) { + clear_bit(dd->ipath_r_portenable_shift + i, + &dd->ipath_rcvctrl); + clear_bit(dd->ipath_r_intravail_shift + i, + &dd->ipath_rcvctrl); + } + ipath_write_kreg(dd, dd->ipath_kregs->kr_rcvctrl, + dd->ipath_rcvctrl); - *pdp = dd->ipath_pd[0]; - /* ensure chip does no sends or receives while we re-initialize */ - dd->ipath_control = dd->ipath_sendctrl = dd->ipath_rcvctrl = 0U; - ipath_write_kreg(dd, dd->ipath_kregs->kr_rcvctrl, dd->ipath_rcvctrl); + spin_lock_irqsave(&dd->ipath_sendctrl_lock, flags); + dd->ipath_sendctrl = 0U; /* no sdma, etc */ ipath_write_kreg(dd, dd->ipath_kregs->kr_sendctrl, dd->ipath_sendctrl); - ipath_write_kreg(dd, dd->ipath_kregs->kr_control, dd->ipath_control); + ipath_read_kreg64(dd, dd->ipath_kregs->kr_scratch); + spin_unlock_irqrestore(&dd->ipath_sendctrl_lock, flags); + + ipath_write_kreg(dd, dd->ipath_kregs->kr_control, 0ULL); - rtmp = ipath_read_kreg32(dd, dd->ipath_kregs->kr_portcnt); - if (dd->ipath_portcnt != rtmp) - dev_info(&dd->pcidev->dev, "portcnt was %u before " - "reset, now %u, using original\n", - dd->ipath_portcnt, rtmp); rtmp = ipath_read_kreg32(dd, dd->ipath_kregs->kr_rcvtidcnt); if (rtmp != dd->ipath_rcvtidcnt) dev_info(&dd->pcidev->dev, "tidcnt was %u before " @@ -467,10 +470,10 @@ static void init_shadow_tids(struct ipath_devdata *dd) dd->ipath_physshadow = addrs; } -static void enable_chip(struct ipath_devdata *dd, - struct ipath_portdata *pd, int reinit) +static void enable_chip(struct ipath_devdata *dd, int reinit) { u32 val; + u64 rcvmask; unsigned long flags; int i; @@ -484,17 +487,28 @@ static void enable_chip(struct ipath_devdata *dd, /* Enable PIO send, and update of PIOavail regs to memory. */ dd->ipath_sendctrl = INFINIPATH_S_PIOENABLE | INFINIPATH_S_PIOBUFAVAILUPD; + + /* + * Set the PIO avail update threshold to host memory + * on chips that support it. + */ + if (dd->ipath_pioupd_thresh) + dd->ipath_sendctrl |= dd->ipath_pioupd_thresh + << INFINIPATH_S_UPDTHRESH_SHIFT; ipath_write_kreg(dd, dd->ipath_kregs->kr_sendctrl, dd->ipath_sendctrl); ipath_read_kreg64(dd, dd->ipath_kregs->kr_scratch); spin_unlock_irqrestore(&dd->ipath_sendctrl_lock, flags); /* - * enable port 0 receive, and receive interrupt. other ports - * done as user opens and inits them. + * Enable kernel ports' receive and receive interrupt. + * Other ports done as user opens and inits them. */ - dd->ipath_rcvctrl = (1ULL << dd->ipath_r_tailupd_shift) | - (1ULL << dd->ipath_r_portenable_shift) | - (1ULL << dd->ipath_r_intravail_shift); + rcvmask = 1ULL; + dd->ipath_rcvctrl |= (rcvmask << dd->ipath_r_portenable_shift) | + (rcvmask << dd->ipath_r_intravail_shift); + if (!(dd->ipath_flags & IPATH_NODMA_RTAIL)) + dd->ipath_rcvctrl |= (1ULL << dd->ipath_r_tailupd_shift); + ipath_write_kreg(dd, dd->ipath_kregs->kr_rcvctrl, dd->ipath_rcvctrl); @@ -505,16 +519,16 @@ static void enable_chip(struct ipath_devdata *dd, dd->ipath_flags |= IPATH_INITTED; /* - * init our shadow copies of head from tail values, and write - * head values to match. + * Init our shadow copies of head from tail values, + * and write head values to match. */ val = ipath_read_ureg32(dd, ur_rcvegrindextail, 0); - (void)ipath_write_ureg(dd, ur_rcvegrindexhead, val, 0); + ipath_write_ureg(dd, ur_rcvegrindexhead, val, 0); /* Initialize so we interrupt on next packet received */ - (void)ipath_write_ureg(dd, ur_rcvhdrhead, - dd->ipath_rhdrhead_intr_off | - dd->ipath_pd[0]->port_head, 0); + ipath_write_ureg(dd, ur_rcvhdrhead, + dd->ipath_rhdrhead_intr_off | + dd->ipath_pd[0]->port_head, 0); /* * by now pioavail updates to memory should have occurred, so @@ -523,25 +537,26 @@ static void enable_chip(struct ipath_devdata *dd, * initial values of the generation bit correct. */ for (i = 0; i < dd->ipath_pioavregs; i++) { - __le64 val; + __le64 pioavail; /* * Chip Errata bug 6641; even and odd qwords>3 are swapped. */ if (i > 3 && (dd->ipath_flags & IPATH_SWAP_PIOBUFS)) - val = dd->ipath_pioavailregs_dma[i ^ 1]; + pioavail = dd->ipath_pioavailregs_dma[i ^ 1]; else - val = dd->ipath_pioavailregs_dma[i]; - dd->ipath_pioavailshadow[i] = le64_to_cpu(val); + pioavail = dd->ipath_pioavailregs_dma[i]; + dd->ipath_pioavailshadow[i] = le64_to_cpu(pioavail) | + (~dd->ipath_pioavailkernel[i] << + INFINIPATH_SENDPIOAVAIL_BUSY_SHIFT); } /* can get counters, stats, etc. */ dd->ipath_flags |= IPATH_PRESENT; } -static int init_housekeeping(struct ipath_devdata *dd, - struct ipath_portdata **pdp, int reinit) +static int init_housekeeping(struct ipath_devdata *dd, int reinit) { - char boardn[32]; + char boardn[40]; int ret = 0; /* @@ -600,18 +615,9 @@ static int init_housekeeping(struct ipath_devdata *dd, ipath_write_kreg(dd, dd->ipath_kregs->kr_errorclear, INFINIPATH_E_RESET); - if (reinit) - ret = init_chip_reset(dd, pdp); - else - ret = init_chip_first(dd, pdp); - - if (ret) - goto done; - - ipath_cdbg(VERBOSE, "Revision %llx (PCI %x), %u ports, %u tids, " - "%u egrtids\n", (unsigned long long) dd->ipath_revision, - dd->ipath_pcirev, dd->ipath_portcnt, dd->ipath_rcvtidcnt, - dd->ipath_rcvegrcnt); + ipath_cdbg(VERBOSE, "Revision %llx (PCI %x)\n", + (unsigned long long) dd->ipath_revision, + dd->ipath_pcirev); if (((dd->ipath_revision >> INFINIPATH_R_SOFTWARE_SHIFT) & INFINIPATH_R_SOFTWARE_MASK) != IPATH_CHIP_SWVERSION) { @@ -650,10 +656,39 @@ static int init_housekeeping(struct ipath_devdata *dd, ipath_dbg("%s", dd->ipath_boardversion); + if (ret) + goto done; + + if (reinit) + ret = init_chip_reset(dd); + else + ret = init_chip_first(dd); + done: return ret; } +static void verify_interrupt(unsigned long opaque) +{ + struct ipath_devdata *dd = (struct ipath_devdata *) opaque; + + if (!dd) + return; /* being torn down */ + + /* + * If we don't have any interrupts, let the user know and + * don't bother checking again. + */ + if (dd->ipath_int_counter == 0) { + if (!dd->ipath_f_intr_fallback(dd)) + dev_err(&dd->pcidev->dev, "No interrupts detected, " + "not usable.\n"); + else /* re-arm the timer to see if fallback works */ + mod_timer(&dd->ipath_intrchk_timer, jiffies + HZ/2); + } else + ipath_cdbg(VERBOSE, "%u interrupts at timer check\n", + dd->ipath_int_counter); +} /** * ipath_init_chip - do the actual initialization sequence on the chip @@ -676,11 +711,11 @@ int ipath_init_chip(struct ipath_devdata *dd, int reinit) u32 val32, kpiobufs; u32 piobufs, uports; u64 val; - struct ipath_portdata *pd = NULL; /* keep gcc4 happy */ + struct ipath_portdata *pd; gfp_t gfp_flags = GFP_USER | __GFP_COMP; unsigned long flags; - ret = init_housekeeping(dd, &pd, reinit); + ret = init_housekeeping(dd, reinit); if (ret) goto done; @@ -700,7 +735,7 @@ int ipath_init_chip(struct ipath_devdata *dd, int reinit) * we now use routines that backend onto __get_free_pages, the * rest would be wasted. */ - dd->ipath_rcvhdrcnt = dd->ipath_rcvegrcnt; + dd->ipath_rcvhdrcnt = max(dd->ipath_p0_rcvegrcnt, dd->ipath_rcvegrcnt); ipath_write_kreg(dd, dd->ipath_kregs->kr_rcvhdrcnt, dd->ipath_rcvhdrcnt); @@ -731,8 +766,8 @@ int ipath_init_chip(struct ipath_devdata *dd, int reinit) if (kpiobufs + (uports * IPATH_MIN_USER_PORT_BUFCNT) > piobufs) { int i = (int) piobufs - (int) (uports * IPATH_MIN_USER_PORT_BUFCNT); - if (i < 0) - i = 0; + if (i < 1) + i = 1; dev_info(&dd->pcidev->dev, "Allocating %d PIO bufs of " "%d for kernel leaves too few for %d user ports " "(%d each); using %u\n", kpiobufs, @@ -751,24 +786,40 @@ int ipath_init_chip(struct ipath_devdata *dd, int reinit) ipath_dbg("allocating %u pbufs/port leaves %u unused, " "add to kernel\n", dd->ipath_pbufsport, val32); dd->ipath_lastport_piobuf -= val32; + kpiobufs += val32; ipath_dbg("%u pbufs/port leaves %u unused, add to kernel\n", dd->ipath_pbufsport, val32); } - dd->ipath_lastpioindex = dd->ipath_lastport_piobuf; + dd->ipath_lastpioindex = 0; + dd->ipath_lastpioindexl = dd->ipath_piobcnt2k; + ipath_chg_pioavailkernel(dd, 0, piobufs, 1); ipath_cdbg(VERBOSE, "%d PIO bufs for kernel out of %d total %u " "each for %u user ports\n", kpiobufs, piobufs, dd->ipath_pbufsport, uports); + if (dd->ipath_pioupd_thresh) { + if (dd->ipath_pbufsport < dd->ipath_pioupd_thresh) + dd->ipath_pioupd_thresh = dd->ipath_pbufsport; + if (kpiobufs < dd->ipath_pioupd_thresh) + dd->ipath_pioupd_thresh = kpiobufs; + } + + ret = dd->ipath_f_early_init(dd); + if (ret) { + ipath_dev_err(dd, "Early initialization failure\n"); + goto done; + } - dd->ipath_f_early_init(dd); /* - * cancel any possible active sends from early driver load. + * Cancel any possible active sends from early driver load. * Follows early_init because some chips have to initialize * PIO buffers in early_init to avoid false parity errors. */ ipath_cancel_sends(dd, 0); - /* early_init sets rcvhdrentsize and rcvhdrsize, so this must be - * done after early_init */ + /* + * Early_init sets rcvhdrentsize and rcvhdrsize, so this must be + * done after early_init. + */ dd->ipath_hdrqlast = dd->ipath_rcvhdrentsize * (dd->ipath_rcvhdrcnt - 1); ipath_write_kreg(dd, dd->ipath_kregs->kr_rcvhdrentsize, @@ -783,8 +834,8 @@ int ipath_init_chip(struct ipath_devdata *dd, int reinit) goto done; } - (void)ipath_write_kreg(dd, dd->ipath_kregs->kr_sendpioavailaddr, - dd->ipath_pioavailregs_phys); + ipath_write_kreg(dd, dd->ipath_kregs->kr_sendpioavailaddr, + dd->ipath_pioavailregs_phys); /* * this is to detect s/w errors, which the h/w works around by * ignoring the low 6 bits of address, if it wasn't aligned. @@ -843,58 +894,65 @@ int ipath_init_chip(struct ipath_devdata *dd, int reinit) /* enable errors that are masked, at least this first time. */ ipath_write_kreg(dd, dd->ipath_kregs->kr_errormask, ~dd->ipath_maskederrs); - dd->ipath_errormask = ipath_read_kreg64(dd, - dd->ipath_kregs->kr_errormask); + dd->ipath_maskederrs = 0; /* don't re-enable ignored in timer */ + dd->ipath_errormask = + ipath_read_kreg64(dd, dd->ipath_kregs->kr_errormask); /* clear any interrupts up to this point (ints still not enabled) */ ipath_write_kreg(dd, dd->ipath_kregs->kr_intclear, -1LL); + dd->ipath_f_tidtemplate(dd); + /* * Set up the port 0 (kernel) rcvhdr q and egr TIDs. If doing * re-init, the simplest way to handle this is to free * existing, and re-allocate. * Need to re-create rest of port 0 portdata as well. */ + pd = dd->ipath_pd[0]; if (reinit) { - /* Alloc and init new ipath_portdata for port0, + struct ipath_portdata *npd; + + /* + * Alloc and init new ipath_portdata for port0, * Then free old pd. Could lead to fragmentation, but also * makes later support for hot-swap easier. */ - struct ipath_portdata *npd; npd = create_portdata0(dd); if (npd) { ipath_free_pddata(dd, pd); - dd->ipath_pd[0] = pd = npd; + dd->ipath_pd[0] = npd; + pd = npd; } else { - ipath_dev_err(dd, "Unable to allocate portdata for" - " port 0, failing\n"); + ipath_dev_err(dd, "Unable to allocate portdata" + " for port 0, failing\n"); ret = -ENOMEM; goto done; } } - dd->ipath_f_tidtemplate(dd); ret = ipath_create_rcvhdrq(dd, pd); - if (!ret) { - dd->ipath_hdrqtailptr = - (volatile __le64 *)pd->port_rcvhdrtail_kvaddr; + if (!ret) ret = create_port0_egr(dd); - } - if (ret) - ipath_dev_err(dd, "failed to allocate port 0 (kernel) " + if (ret) { + ipath_dev_err(dd, "failed to allocate kernel port's " "rcvhdrq and/or egr bufs\n"); + goto done; + } else - enable_chip(dd, pd, reinit); - + enable_chip(dd, reinit); - if (!ret && !reinit) { - /* used when we close a port, for DMA already in flight at close */ + if (!reinit) { + /* + * Used when we close a port, for DMA already in flight + * at close. + */ dd->ipath_dummy_hdrq = dma_alloc_coherent( - &dd->pcidev->dev, pd->port_rcvhdrq_size, + &dd->pcidev->dev, dd->ipath_pd[0]->port_rcvhdrq_size, &dd->ipath_dummy_hdrq_phys, gfp_flags); - if (!dd->ipath_dummy_hdrq ) { + if (!dd->ipath_dummy_hdrq) { dev_info(&dd->pcidev->dev, "Couldn't allocate 0x%lx bytes for dummy hdrq\n", - pd->port_rcvhdrq_size); + dd->ipath_pd[0]->port_rcvhdrq_size); /* fallback to just 0'ing */ dd->ipath_dummy_hdrq_phys = 0UL; } @@ -906,7 +964,7 @@ int ipath_init_chip(struct ipath_devdata *dd, int reinit) */ ipath_write_kreg(dd, dd->ipath_kregs->kr_intclear, 0ULL); - if(!dd->ipath_stats_timer_active) { + if (!dd->ipath_stats_timer_active) { /* * first init, or after an admin disable/enable * set up stats retrieval timer, even if we had errors @@ -922,6 +980,16 @@ int ipath_init_chip(struct ipath_devdata *dd, int reinit) dd->ipath_stats_timer_active = 1; } + /* Set up SendDMA if chip supports it */ + if (dd->ipath_flags & IPATH_HAS_SEND_DMA) + ret = setup_sdma(dd); + + /* Set up HoL state */ + init_timer(&dd->ipath_hol_timer); + dd->ipath_hol_timer.function = ipath_hol_event; + dd->ipath_hol_timer.data = (unsigned long)dd; + dd->ipath_hol_state = IPATH_HOL_UP; + done: if (!ret) { *dd->ipath_statusp |= IPATH_STATUS_CHIP_PRESENT; @@ -934,6 +1002,20 @@ done: 0ULL); /* chip is usable; mark it as initialized */ *dd->ipath_statusp |= IPATH_STATUS_INITTED; + + /* + * setup to verify we get an interrupt, and fallback + * to an alternate if necessary and possible + */ + if (!reinit) { + init_timer(&dd->ipath_intrchk_timer); + dd->ipath_intrchk_timer.function = + verify_interrupt; + dd->ipath_intrchk_timer.data = + (unsigned long) dd; + } + dd->ipath_intrchk_timer.expires = jiffies + HZ/2; + add_timer(&dd->ipath_intrchk_timer); } else ipath_dev_err(dd, "No interrupts enabled, couldn't " "setup interrupt address\n"); diff --git a/drivers/infiniband/hw/ipath/ipath_intr.c b/drivers/infiniband/hw/ipath/ipath_intr.c index 92e58c921522..1b58f4737c71 100644 --- a/drivers/infiniband/hw/ipath/ipath_intr.c +++ b/drivers/infiniband/hw/ipath/ipath_intr.c @@ -1,5 +1,5 @@ /* - * Copyright (c) 2006, 2007 QLogic Corporation. All rights reserved. + * Copyright (c) 2006, 2007, 2008 QLogic Corporation. All rights reserved. * Copyright (c) 2003, 2004, 2005, 2006 PathScale, Inc. All rights reserved. * * This software is available to you under a choice of one of two @@ -32,6 +32,7 @@ */ #include <linux/pci.h> +#include <linux/delay.h> #include "ipath_kernel.h" #include "ipath_verbs.h" @@ -59,9 +60,11 @@ static void ipath_clrpiobuf(struct ipath_devdata *dd, u32 pnum) dev_info(&dd->pcidev->dev, "Rewrite PIO buffer %u, to recover from parity error\n", pnum); - *pbuf = dwcnt+1; /* no flush required, since already in freeze */ - while(--dwcnt) - *pbuf++ = 0; + + /* no flush required, since already in freeze */ + writel(dwcnt + 1, pbuf); + while (--dwcnt) + writel(0, pbuf++); } /* @@ -70,7 +73,7 @@ static void ipath_clrpiobuf(struct ipath_devdata *dd, u32 pnum) * If rewrite is true, and bits are set in the sendbufferror registers, * we'll write to the buffer, for error recovery on parity errors. */ -static void ipath_disarm_senderrbufs(struct ipath_devdata *dd, int rewrite) +void ipath_disarm_senderrbufs(struct ipath_devdata *dd, int rewrite) { u32 piobcnt; unsigned long sbuf[4]; @@ -84,12 +87,14 @@ static void ipath_disarm_senderrbufs(struct ipath_devdata *dd, int rewrite) dd, dd->ipath_kregs->kr_sendbuffererror); sbuf[1] = ipath_read_kreg64( dd, dd->ipath_kregs->kr_sendbuffererror + 1); - if (piobcnt > 128) { + if (piobcnt > 128) sbuf[2] = ipath_read_kreg64( dd, dd->ipath_kregs->kr_sendbuffererror + 2); + if (piobcnt > 192) sbuf[3] = ipath_read_kreg64( dd, dd->ipath_kregs->kr_sendbuffererror + 3); - } + else + sbuf[3] = 0; if (sbuf[0] || sbuf[1] || (piobcnt > 128 && (sbuf[2] || sbuf[3]))) { int i; @@ -254,24 +259,20 @@ void ipath_format_hwerrors(u64 hwerrs, } /* return the strings for the most common link states */ -static char *ib_linkstate(u32 linkstate) +static char *ib_linkstate(struct ipath_devdata *dd, u64 ibcs) { char *ret; + u32 state; - switch (linkstate) { - case IPATH_IBSTATE_INIT: + state = ipath_ib_state(dd, ibcs); + if (state == dd->ib_init) ret = "Init"; - break; - case IPATH_IBSTATE_ARM: + else if (state == dd->ib_arm) ret = "Arm"; - break; - case IPATH_IBSTATE_ACTIVE: + else if (state == dd->ib_active) ret = "Active"; - break; - default: + else ret = "Down"; - } - return ret; } @@ -286,103 +287,172 @@ void signal_ib_event(struct ipath_devdata *dd, enum ib_event_type ev) } static void handle_e_ibstatuschanged(struct ipath_devdata *dd, - ipath_err_t errs, int noprint) + ipath_err_t errs) { - u64 val; - u32 ltstate, lstate; + u32 ltstate, lstate, ibstate, lastlstate; + u32 init = dd->ib_init; + u32 arm = dd->ib_arm; + u32 active = dd->ib_active; + const u64 ibcs = ipath_read_kreg64(dd, dd->ipath_kregs->kr_ibcstatus); + + lstate = ipath_ib_linkstate(dd, ibcs); /* linkstate */ + ibstate = ipath_ib_state(dd, ibcs); + /* linkstate at last interrupt */ + lastlstate = ipath_ib_linkstate(dd, dd->ipath_lastibcstat); + ltstate = ipath_ib_linktrstate(dd, ibcs); /* linktrainingtate */ /* - * even if diags are enabled, we want to notice LINKINIT, etc. - * We just don't want to change the LED state, or - * dd->ipath_kregs->kr_ibcctrl + * Since going into a recovery state causes the link state to go + * down and since recovery is transitory, it is better if we "miss" + * ever seeing the link training state go into recovery (i.e., + * ignore this transition for link state special handling purposes) + * without even updating ipath_lastibcstat. */ - val = ipath_read_kreg64(dd, dd->ipath_kregs->kr_ibcstatus); - lstate = val & IPATH_IBSTATE_MASK; + if ((ltstate == INFINIPATH_IBCS_LT_STATE_RECOVERRETRAIN) || + (ltstate == INFINIPATH_IBCS_LT_STATE_RECOVERWAITRMT) || + (ltstate == INFINIPATH_IBCS_LT_STATE_RECOVERIDLE)) + goto done; /* - * this is confusing enough when it happens that I want to always put it - * on the console and in the logs. If it was a requested state change, - * we'll have already cleared the flags, so we won't print this warning + * if linkstate transitions into INIT from any of the various down + * states, or if it transitions from any of the up (INIT or better) + * states into any of the down states (except link recovery), then + * call the chip-specific code to take appropriate actions. */ - if ((lstate != IPATH_IBSTATE_ARM && lstate != IPATH_IBSTATE_ACTIVE) - && (dd->ipath_flags & (IPATH_LINKARMED | IPATH_LINKACTIVE))) { - dev_info(&dd->pcidev->dev, "Link state changed from %s to %s\n", - (dd->ipath_flags & IPATH_LINKARMED) ? "ARM" : "ACTIVE", - ib_linkstate(lstate)); - /* - * Flush all queued sends when link went to DOWN or INIT, - * to be sure that they don't block SMA and other MAD packets - */ - ipath_cancel_sends(dd, 1); - } - else if (lstate == IPATH_IBSTATE_INIT || lstate == IPATH_IBSTATE_ARM || - lstate == IPATH_IBSTATE_ACTIVE) { - /* - * only print at SMA if there is a change, debug if not - * (sometimes we want to know that, usually not). - */ - if (lstate == ((unsigned) dd->ipath_lastibcstat - & IPATH_IBSTATE_MASK)) { - ipath_dbg("Status change intr but no change (%s)\n", - ib_linkstate(lstate)); + if (lstate >= INFINIPATH_IBCS_L_STATE_INIT && + lastlstate == INFINIPATH_IBCS_L_STATE_DOWN) { + /* transitioned to UP */ + if (dd->ipath_f_ib_updown(dd, 1, ibcs)) { + /* link came up, so we must no longer be disabled */ + dd->ipath_flags &= ~IPATH_IB_LINK_DISABLED; + ipath_cdbg(LINKVERB, "LinkUp handled, skipped\n"); + goto skip_ibchange; /* chip-code handled */ + } + } else if ((lastlstate >= INFINIPATH_IBCS_L_STATE_INIT || + (dd->ipath_flags & IPATH_IB_FORCE_NOTIFY)) && + ltstate <= INFINIPATH_IBCS_LT_STATE_CFGWAITRMT && + ltstate != INFINIPATH_IBCS_LT_STATE_LINKUP) { + int handled; + handled = dd->ipath_f_ib_updown(dd, 0, ibcs); + dd->ipath_flags &= ~IPATH_IB_FORCE_NOTIFY; + if (handled) { + ipath_cdbg(LINKVERB, "LinkDown handled, skipped\n"); + goto skip_ibchange; /* chip-code handled */ } - else - ipath_cdbg(VERBOSE, "Unit %u link state %s, last " - "was %s\n", dd->ipath_unit, - ib_linkstate(lstate), - ib_linkstate((unsigned) - dd->ipath_lastibcstat - & IPATH_IBSTATE_MASK)); } - else { - lstate = dd->ipath_lastibcstat & IPATH_IBSTATE_MASK; - if (lstate == IPATH_IBSTATE_INIT || - lstate == IPATH_IBSTATE_ARM || - lstate == IPATH_IBSTATE_ACTIVE) - ipath_cdbg(VERBOSE, "Unit %u link state down" - " (state 0x%x), from %s\n", - dd->ipath_unit, - (u32)val & IPATH_IBSTATE_MASK, - ib_linkstate(lstate)); - else - ipath_cdbg(VERBOSE, "Unit %u link state changed " - "to 0x%x from down (%x)\n", - dd->ipath_unit, (u32) val, lstate); + + /* + * Significant enough to always print and get into logs, if it was + * unexpected. If it was a requested state change, we'll have + * already cleared the flags, so we won't print this warning + */ + if ((ibstate != arm && ibstate != active) && + (dd->ipath_flags & (IPATH_LINKARMED | IPATH_LINKACTIVE))) { + dev_info(&dd->pcidev->dev, "Link state changed from %s " + "to %s\n", (dd->ipath_flags & IPATH_LINKARMED) ? + "ARM" : "ACTIVE", ib_linkstate(dd, ibcs)); } - ltstate = (val >> INFINIPATH_IBCS_LINKTRAININGSTATE_SHIFT) & - INFINIPATH_IBCS_LINKTRAININGSTATE_MASK; - lstate = (val >> INFINIPATH_IBCS_LINKSTATE_SHIFT) & - INFINIPATH_IBCS_LINKSTATE_MASK; if (ltstate == INFINIPATH_IBCS_LT_STATE_POLLACTIVE || ltstate == INFINIPATH_IBCS_LT_STATE_POLLQUIET) { - u32 last_ltstate; - + u32 lastlts; + lastlts = ipath_ib_linktrstate(dd, dd->ipath_lastibcstat); /* - * Ignore cycling back and forth from Polling.Active - * to Polling.Quiet while waiting for the other end of - * the link to come up. We will cycle back and forth - * between them if no cable is plugged in, - * the other device is powered off or disabled, etc. + * Ignore cycling back and forth from Polling.Active to + * Polling.Quiet while waiting for the other end of the link + * to come up, except to try and decide if we are connected + * to a live IB device or not. We will cycle back and + * forth between them if no cable is plugged in, the other + * device is powered off or disabled, etc. */ - last_ltstate = (dd->ipath_lastibcstat >> - INFINIPATH_IBCS_LINKTRAININGSTATE_SHIFT) - & INFINIPATH_IBCS_LINKTRAININGSTATE_MASK; - if (last_ltstate == INFINIPATH_IBCS_LT_STATE_POLLACTIVE - || last_ltstate == - INFINIPATH_IBCS_LT_STATE_POLLQUIET) { - if (dd->ipath_ibpollcnt > 40) { + if (lastlts == INFINIPATH_IBCS_LT_STATE_POLLACTIVE || + lastlts == INFINIPATH_IBCS_LT_STATE_POLLQUIET) { + if (!(dd->ipath_flags & IPATH_IB_AUTONEG_INPROG) && + (++dd->ipath_ibpollcnt == 40)) { dd->ipath_flags |= IPATH_NOCABLE; *dd->ipath_statusp |= IPATH_STATUS_IB_NOCABLE; - } else - dd->ipath_ibpollcnt++; + ipath_cdbg(LINKVERB, "Set NOCABLE\n"); + } + ipath_cdbg(LINKVERB, "POLL change to %s (%x)\n", + ipath_ibcstatus_str[ltstate], ibstate); goto skip_ibchange; } } - dd->ipath_ibpollcnt = 0; /* some state other than 2 or 3 */ + + dd->ipath_ibpollcnt = 0; /* not poll*, now */ ipath_stats.sps_iblink++; - if (ltstate != INFINIPATH_IBCS_LT_STATE_LINKUP) { + + if (ibstate != init && dd->ipath_lastlinkrecov && ipath_linkrecovery) { + u64 linkrecov; + linkrecov = ipath_snap_cntr(dd, + dd->ipath_cregs->cr_iblinkerrrecovcnt); + if (linkrecov != dd->ipath_lastlinkrecov) { + ipath_dbg("IB linkrecov up %Lx (%s %s) recov %Lu\n", + ibcs, ib_linkstate(dd, ibcs), + ipath_ibcstatus_str[ltstate], + linkrecov); + /* and no more until active again */ + dd->ipath_lastlinkrecov = 0; + ipath_set_linkstate(dd, IPATH_IB_LINKDOWN); + goto skip_ibchange; + } + } + + if (ibstate == init || ibstate == arm || ibstate == active) { + *dd->ipath_statusp &= ~IPATH_STATUS_IB_NOCABLE; + if (ibstate == init || ibstate == arm) { + *dd->ipath_statusp &= ~IPATH_STATUS_IB_READY; + if (dd->ipath_flags & IPATH_LINKACTIVE) + signal_ib_event(dd, IB_EVENT_PORT_ERR); + } + if (ibstate == arm) { + dd->ipath_flags |= IPATH_LINKARMED; + dd->ipath_flags &= ~(IPATH_LINKUNK | + IPATH_LINKINIT | IPATH_LINKDOWN | + IPATH_LINKACTIVE | IPATH_NOCABLE); + ipath_hol_down(dd); + } else if (ibstate == init) { + /* + * set INIT and DOWN. Down is checked by + * most of the other code, but INIT is + * useful to know in a few places. + */ + dd->ipath_flags |= IPATH_LINKINIT | + IPATH_LINKDOWN; + dd->ipath_flags &= ~(IPATH_LINKUNK | + IPATH_LINKARMED | IPATH_LINKACTIVE | + IPATH_NOCABLE); + ipath_hol_down(dd); + } else { /* active */ + dd->ipath_lastlinkrecov = ipath_snap_cntr(dd, + dd->ipath_cregs->cr_iblinkerrrecovcnt); + *dd->ipath_statusp |= + IPATH_STATUS_IB_READY | IPATH_STATUS_IB_CONF; + dd->ipath_flags |= IPATH_LINKACTIVE; + dd->ipath_flags &= ~(IPATH_LINKUNK | IPATH_LINKINIT + | IPATH_LINKDOWN | IPATH_LINKARMED | + IPATH_NOCABLE); + if (dd->ipath_flags & IPATH_HAS_SEND_DMA) + ipath_restart_sdma(dd); + signal_ib_event(dd, IB_EVENT_PORT_ACTIVE); + /* LED active not handled in chip _f_updown */ + dd->ipath_f_setextled(dd, lstate, ltstate); + ipath_hol_up(dd); + } + + /* + * print after we've already done the work, so as not to + * delay the state changes and notifications, for debugging + */ + if (lstate == lastlstate) + ipath_cdbg(LINKVERB, "Unchanged from last: %s " + "(%x)\n", ib_linkstate(dd, ibcs), ibstate); + else + ipath_cdbg(VERBOSE, "Unit %u: link up to %s %s (%x)\n", + dd->ipath_unit, ib_linkstate(dd, ibcs), + ipath_ibcstatus_str[ltstate], ibstate); + } else { /* down */ if (dd->ipath_flags & IPATH_LINKACTIVE) signal_ib_event(dd, IB_EVENT_PORT_ERR); dd->ipath_flags |= IPATH_LINKDOWN; @@ -391,69 +461,28 @@ static void handle_e_ibstatuschanged(struct ipath_devdata *dd, IPATH_LINKARMED); *dd->ipath_statusp &= ~IPATH_STATUS_IB_READY; dd->ipath_lli_counter = 0; - if (!noprint) { - if (((dd->ipath_lastibcstat >> - INFINIPATH_IBCS_LINKSTATE_SHIFT) & - INFINIPATH_IBCS_LINKSTATE_MASK) - == INFINIPATH_IBCS_L_STATE_ACTIVE) - /* if from up to down be more vocal */ - ipath_cdbg(VERBOSE, - "Unit %u link now down (%s)\n", - dd->ipath_unit, - ipath_ibcstatus_str[ltstate]); - else - ipath_cdbg(VERBOSE, "Unit %u link is " - "down (%s)\n", dd->ipath_unit, - ipath_ibcstatus_str[ltstate]); - } - dd->ipath_f_setextled(dd, lstate, ltstate); - } else if ((val & IPATH_IBSTATE_MASK) == IPATH_IBSTATE_ACTIVE) { - dd->ipath_flags |= IPATH_LINKACTIVE; - dd->ipath_flags &= - ~(IPATH_LINKUNK | IPATH_LINKINIT | IPATH_LINKDOWN | - IPATH_LINKARMED | IPATH_NOCABLE); - *dd->ipath_statusp &= ~IPATH_STATUS_IB_NOCABLE; - *dd->ipath_statusp |= - IPATH_STATUS_IB_READY | IPATH_STATUS_IB_CONF; - dd->ipath_f_setextled(dd, lstate, ltstate); - signal_ib_event(dd, IB_EVENT_PORT_ACTIVE); - } else if ((val & IPATH_IBSTATE_MASK) == IPATH_IBSTATE_INIT) { - if (dd->ipath_flags & IPATH_LINKACTIVE) - signal_ib_event(dd, IB_EVENT_PORT_ERR); - /* - * set INIT and DOWN. Down is checked by most of the other - * code, but INIT is useful to know in a few places. - */ - dd->ipath_flags |= IPATH_LINKINIT | IPATH_LINKDOWN; - dd->ipath_flags &= - ~(IPATH_LINKUNK | IPATH_LINKACTIVE | IPATH_LINKARMED - | IPATH_NOCABLE); - *dd->ipath_statusp &= ~(IPATH_STATUS_IB_NOCABLE - | IPATH_STATUS_IB_READY); - dd->ipath_f_setextled(dd, lstate, ltstate); - } else if ((val & IPATH_IBSTATE_MASK) == IPATH_IBSTATE_ARM) { - if (dd->ipath_flags & IPATH_LINKACTIVE) - signal_ib_event(dd, IB_EVENT_PORT_ERR); - dd->ipath_flags |= IPATH_LINKARMED; - dd->ipath_flags &= - ~(IPATH_LINKUNK | IPATH_LINKDOWN | IPATH_LINKINIT | - IPATH_LINKACTIVE | IPATH_NOCABLE); - *dd->ipath_statusp &= ~(IPATH_STATUS_IB_NOCABLE - | IPATH_STATUS_IB_READY); - dd->ipath_f_setextled(dd, lstate, ltstate); - } else { - if (!noprint) - ipath_dbg("IBstatuschange unit %u: %s (%x)\n", - dd->ipath_unit, - ipath_ibcstatus_str[ltstate], ltstate); + if (lastlstate != INFINIPATH_IBCS_L_STATE_DOWN) + ipath_cdbg(VERBOSE, "Unit %u link state down " + "(state 0x%x), from %s\n", + dd->ipath_unit, lstate, + ib_linkstate(dd, dd->ipath_lastibcstat)); + else + ipath_cdbg(LINKVERB, "Unit %u link state changed " + "to %s (0x%x) from down (%x)\n", + dd->ipath_unit, + ipath_ibcstatus_str[ltstate], + ibstate, lastlstate); } + skip_ibchange: - dd->ipath_lastibcstat = val; + dd->ipath_lastibcstat = ibcs; +done: + return; } static void handle_supp_msgs(struct ipath_devdata *dd, - unsigned supp_msgs, char *msg, int msgsz) + unsigned supp_msgs, char *msg, u32 msgsz) { /* * Print the message unless it's ibc status change only, which @@ -461,12 +490,19 @@ static void handle_supp_msgs(struct ipath_devdata *dd, */ if (dd->ipath_lasterror & ~INFINIPATH_E_IBSTATUSCHANGED) { int iserr; - iserr = ipath_decode_err(msg, msgsz, + ipath_err_t mask; + iserr = ipath_decode_err(dd, msg, msgsz, dd->ipath_lasterror & ~INFINIPATH_E_IBSTATUSCHANGED); - if (dd->ipath_lasterror & - ~(INFINIPATH_E_RRCVEGRFULL | - INFINIPATH_E_RRCVHDRFULL | INFINIPATH_E_PKTERRS)) + + mask = INFINIPATH_E_RRCVEGRFULL | INFINIPATH_E_RRCVHDRFULL | + INFINIPATH_E_PKTERRS | INFINIPATH_E_SDMADISABLED; + + /* if we're in debug, then don't mask SDMADISABLED msgs */ + if (ipath_debug & __IPATH_DBG) + mask &= ~INFINIPATH_E_SDMADISABLED; + + if (dd->ipath_lasterror & ~mask) ipath_dev_err(dd, "Suppressed %u messages for " "fast-repeating errors (%s) (%llx)\n", supp_msgs, msg, @@ -493,7 +529,7 @@ static void handle_supp_msgs(struct ipath_devdata *dd, static unsigned handle_frequent_errors(struct ipath_devdata *dd, ipath_err_t errs, char *msg, - int msgsz, int *noprint) + u32 msgsz, int *noprint) { unsigned long nc; static unsigned long nextmsg_time; @@ -523,19 +559,125 @@ static unsigned handle_frequent_errors(struct ipath_devdata *dd, return supp_msgs; } +static void handle_sdma_errors(struct ipath_devdata *dd, ipath_err_t errs) +{ + unsigned long flags; + int expected; + + if (ipath_debug & __IPATH_DBG) { + char msg[128]; + ipath_decode_err(dd, msg, sizeof msg, errs & + INFINIPATH_E_SDMAERRS); + ipath_dbg("errors %lx (%s)\n", (unsigned long)errs, msg); + } + if (ipath_debug & __IPATH_VERBDBG) { + unsigned long tl, hd, status, lengen; + tl = ipath_read_kreg64(dd, dd->ipath_kregs->kr_senddmatail); + hd = ipath_read_kreg64(dd, dd->ipath_kregs->kr_senddmahead); + status = ipath_read_kreg64(dd + , dd->ipath_kregs->kr_senddmastatus); + lengen = ipath_read_kreg64(dd, + dd->ipath_kregs->kr_senddmalengen); + ipath_cdbg(VERBOSE, "sdma tl 0x%lx hd 0x%lx status 0x%lx " + "lengen 0x%lx\n", tl, hd, status, lengen); + } + + spin_lock_irqsave(&dd->ipath_sdma_lock, flags); + __set_bit(IPATH_SDMA_DISABLED, &dd->ipath_sdma_status); + expected = test_bit(IPATH_SDMA_ABORTING, &dd->ipath_sdma_status); + spin_unlock_irqrestore(&dd->ipath_sdma_lock, flags); + if (!expected) + ipath_cancel_sends(dd, 1); +} + +static void handle_sdma_intr(struct ipath_devdata *dd, u64 istat) +{ + unsigned long flags; + int expected; + + if ((istat & INFINIPATH_I_SDMAINT) && + !test_bit(IPATH_SDMA_SHUTDOWN, &dd->ipath_sdma_status)) + ipath_sdma_intr(dd); + + if (istat & INFINIPATH_I_SDMADISABLED) { + expected = test_bit(IPATH_SDMA_ABORTING, + &dd->ipath_sdma_status); + ipath_dbg("%s SDmaDisabled intr\n", + expected ? "expected" : "unexpected"); + spin_lock_irqsave(&dd->ipath_sdma_lock, flags); + __set_bit(IPATH_SDMA_DISABLED, &dd->ipath_sdma_status); + spin_unlock_irqrestore(&dd->ipath_sdma_lock, flags); + if (!expected) + ipath_cancel_sends(dd, 1); + if (!test_bit(IPATH_SDMA_SHUTDOWN, &dd->ipath_sdma_status)) + tasklet_hi_schedule(&dd->ipath_sdma_abort_task); + } +} + +static int handle_hdrq_full(struct ipath_devdata *dd) +{ + int chkerrpkts = 0; + u32 hd, tl; + u32 i; + + ipath_stats.sps_hdrqfull++; + for (i = 0; i < dd->ipath_cfgports; i++) { + struct ipath_portdata *pd = dd->ipath_pd[i]; + + if (i == 0) { + /* + * For kernel receive queues, we just want to know + * if there are packets in the queue that we can + * process. + */ + if (pd->port_head != ipath_get_hdrqtail(pd)) + chkerrpkts |= 1 << i; + continue; + } + + /* Skip if user context is not open */ + if (!pd || !pd->port_cnt) + continue; + + /* Don't report the same point multiple times. */ + if (dd->ipath_flags & IPATH_NODMA_RTAIL) + tl = ipath_read_ureg32(dd, ur_rcvhdrtail, i); + else + tl = ipath_get_rcvhdrtail(pd); + if (tl == pd->port_lastrcvhdrqtail) + continue; + + hd = ipath_read_ureg32(dd, ur_rcvhdrhead, i); + if (hd == (tl + 1) || (!hd && tl == dd->ipath_hdrqlast)) { + pd->port_lastrcvhdrqtail = tl; + pd->port_hdrqfull++; + /* flush hdrqfull so that poll() sees it */ + wmb(); + wake_up_interruptible(&pd->port_wait); + } + } + + return chkerrpkts; +} + static int handle_errors(struct ipath_devdata *dd, ipath_err_t errs) { char msg[128]; u64 ignore_this_time = 0; - int i, iserr = 0; + u64 iserr = 0; int chkerrpkts = 0, noprint = 0; unsigned supp_msgs; int log_idx; - supp_msgs = handle_frequent_errors(dd, errs, msg, sizeof msg, &noprint); + /* + * don't report errors that are masked, either at init + * (not set in ipath_errormask), or temporarily (set in + * ipath_maskederrs) + */ + errs &= dd->ipath_errormask & ~dd->ipath_maskederrs; - /* don't report errors that are masked */ - errs &= ~dd->ipath_maskederrs; + supp_msgs = handle_frequent_errors(dd, errs, msg, (u32)sizeof msg, + &noprint); /* do these first, they are most important */ if (errs & INFINIPATH_E_HARDWARE) { @@ -550,6 +692,9 @@ static int handle_errors(struct ipath_devdata *dd, ipath_err_t errs) } } + if (errs & INFINIPATH_E_SDMAERRS) + handle_sdma_errors(dd, errs); + if (!noprint && (errs & ~dd->ipath_e_bitsextant)) ipath_dev_err(dd, "error interrupt with unknown errors " "%llx set\n", (unsigned long long) @@ -580,18 +725,19 @@ static int handle_errors(struct ipath_devdata *dd, ipath_err_t errs) * ones on this particular interrupt, which also isn't great */ dd->ipath_maskederrs |= dd->ipath_lasterror | errs; + dd->ipath_errormask &= ~dd->ipath_maskederrs; ipath_write_kreg(dd, dd->ipath_kregs->kr_errormask, - dd->ipath_errormask); - s_iserr = ipath_decode_err(msg, sizeof msg, - dd->ipath_maskederrs); + dd->ipath_errormask); + s_iserr = ipath_decode_err(dd, msg, sizeof msg, + dd->ipath_maskederrs); if (dd->ipath_maskederrs & - ~(INFINIPATH_E_RRCVEGRFULL | - INFINIPATH_E_RRCVHDRFULL | INFINIPATH_E_PKTERRS)) + ~(INFINIPATH_E_RRCVEGRFULL | + INFINIPATH_E_RRCVHDRFULL | INFINIPATH_E_PKTERRS)) ipath_dev_err(dd, "Temporarily disabling " "error(s) %llx reporting; too frequent (%s)\n", - (unsigned long long)dd->ipath_maskederrs, + (unsigned long long) dd->ipath_maskederrs, msg); else { /* @@ -633,26 +779,43 @@ static int handle_errors(struct ipath_devdata *dd, ipath_err_t errs) INFINIPATH_E_IBSTATUSCHANGED); } - /* likely due to cancel, so suppress */ + if (errs & INFINIPATH_E_SENDSPECIALTRIGGER) { + dd->ipath_spectriggerhit++; + ipath_dbg("%lu special trigger hits\n", + dd->ipath_spectriggerhit); + } + + /* likely due to cancel; so suppress message unless verbose */ if ((errs & (INFINIPATH_E_SPKTLEN | INFINIPATH_E_SPIOARMLAUNCH)) && dd->ipath_lastcancel > jiffies) { - ipath_dbg("Suppressed armlaunch/spktlen after error send cancel\n"); + /* armlaunch takes precedence; it often causes both. */ + ipath_cdbg(VERBOSE, + "Suppressed %s error (%llx) after sendbuf cancel\n", + (errs & INFINIPATH_E_SPIOARMLAUNCH) ? + "armlaunch" : "sendpktlen", (unsigned long long)errs); errs &= ~(INFINIPATH_E_SPIOARMLAUNCH | INFINIPATH_E_SPKTLEN); } if (!errs) return 0; - if (!noprint) + if (!noprint) { + ipath_err_t mask; /* - * the ones we mask off are handled specially below or above + * The ones we mask off are handled specially below + * or above. Also mask SDMADISABLED by default as it + * is too chatty. */ - ipath_decode_err(msg, sizeof msg, - errs & ~(INFINIPATH_E_IBSTATUSCHANGED | - INFINIPATH_E_RRCVEGRFULL | - INFINIPATH_E_RRCVHDRFULL | - INFINIPATH_E_HARDWARE)); - else + mask = INFINIPATH_E_IBSTATUSCHANGED | + INFINIPATH_E_RRCVEGRFULL | INFINIPATH_E_RRCVHDRFULL | + INFINIPATH_E_HARDWARE | INFINIPATH_E_SDMADISABLED; + + /* if we're in debug, then don't mask SDMADISABLED msgs */ + if (ipath_debug & __IPATH_DBG) + mask &= ~INFINIPATH_E_SDMADISABLED; + + ipath_decode_err(dd, msg, sizeof msg, errs & ~mask); + } else /* so we don't need if (!noprint) at strlcat's below */ *msg = 0; @@ -677,40 +840,8 @@ static int handle_errors(struct ipath_devdata *dd, ipath_err_t errs) * fast_stats, no more than every 5 seconds, user ports get printed * on close */ - if (errs & INFINIPATH_E_RRCVHDRFULL) { - u32 hd, tl; - ipath_stats.sps_hdrqfull++; - for (i = 0; i < dd->ipath_cfgports; i++) { - struct ipath_portdata *pd = dd->ipath_pd[i]; - if (i == 0) { - hd = pd->port_head; - tl = (u32) le64_to_cpu( - *dd->ipath_hdrqtailptr); - } else if (pd && pd->port_cnt && - pd->port_rcvhdrtail_kvaddr) { - /* - * don't report same point multiple times, - * except kernel - */ - tl = *(u64 *) pd->port_rcvhdrtail_kvaddr; - if (tl == pd->port_lastrcvhdrqtail) - continue; - hd = ipath_read_ureg32(dd, ur_rcvhdrhead, - i); - } else - continue; - if (hd == (tl + 1) || - (!hd && tl == dd->ipath_hdrqlast)) { - if (i == 0) - chkerrpkts = 1; - pd->port_lastrcvhdrqtail = tl; - pd->port_hdrqfull++; - /* flush hdrqfull so that poll() sees it */ - wmb(); - wake_up_interruptible(&pd->port_wait); - } - } - } + if (errs & INFINIPATH_E_RRCVHDRFULL) + chkerrpkts |= handle_hdrq_full(dd); if (errs & INFINIPATH_E_RRCVEGRFULL) { struct ipath_portdata *pd = dd->ipath_pd[0]; @@ -721,9 +852,8 @@ static int handle_errors(struct ipath_devdata *dd, ipath_err_t errs) * vs user) */ ipath_stats.sps_etidfull++; - if (pd->port_head != - (u32) le64_to_cpu(*dd->ipath_hdrqtailptr)) - chkerrpkts = 1; + if (pd->port_head != ipath_get_hdrqtail(pd)) + chkerrpkts |= 1; } /* @@ -741,16 +871,13 @@ static int handle_errors(struct ipath_devdata *dd, ipath_err_t errs) dd->ipath_flags &= ~(IPATH_LINKUNK | IPATH_LINKINIT | IPATH_LINKARMED | IPATH_LINKACTIVE); *dd->ipath_statusp &= ~IPATH_STATUS_IB_READY; - if (!noprint) { - u64 st = ipath_read_kreg64( - dd, dd->ipath_kregs->kr_ibcstatus); - ipath_dbg("Lost link, link now down (%s)\n", - ipath_ibcstatus_str[st & 0xf]); - } + ipath_dbg("Lost link, link now down (%s)\n", + ipath_ibcstatus_str[ipath_read_kreg64(dd, + dd->ipath_kregs->kr_ibcstatus) & 0xf]); } if (errs & INFINIPATH_E_IBSTATUSCHANGED) - handle_e_ibstatuschanged(dd, errs, noprint); + handle_e_ibstatuschanged(dd, errs); if (errs & INFINIPATH_E_RESET) { if (!noprint) @@ -765,9 +892,6 @@ static int handle_errors(struct ipath_devdata *dd, ipath_err_t errs) if (!noprint && *msg) { if (iserr) ipath_dev_err(dd, "%s error\n", msg); - else - dev_info(&dd->pcidev->dev, "%s packet problems\n", - msg); } if (dd->ipath_state_wanted & dd->ipath_flags) { ipath_cdbg(VERBOSE, "driver wanted state %x, iflags now %x, " @@ -779,7 +903,6 @@ static int handle_errors(struct ipath_devdata *dd, ipath_err_t errs) return chkerrpkts; } - /* * try to cleanup as much as possible for anything that might have gone * wrong while in freeze mode, such as pio buffers being written by user @@ -796,8 +919,7 @@ static int handle_errors(struct ipath_devdata *dd, ipath_err_t errs) void ipath_clear_freeze(struct ipath_devdata *dd) { int i, im; - __le64 val; - unsigned long flags; + u64 val; /* disable error interrupts, to avoid confusion */ ipath_write_kreg(dd, dd->ipath_kregs->kr_errormask, 0ULL); @@ -816,14 +938,7 @@ void ipath_clear_freeze(struct ipath_devdata *dd) dd->ipath_control); /* ensure pio avail updates continue */ - spin_lock_irqsave(&dd->ipath_sendctrl_lock, flags); - ipath_write_kreg(dd, dd->ipath_kregs->kr_sendctrl, - dd->ipath_sendctrl & ~INFINIPATH_S_PIOBUFAVAILUPD); - ipath_read_kreg64(dd, dd->ipath_kregs->kr_scratch); - ipath_write_kreg(dd, dd->ipath_kregs->kr_sendctrl, - dd->ipath_sendctrl); - ipath_read_kreg64(dd, dd->ipath_kregs->kr_scratch); - spin_unlock_irqrestore(&dd->ipath_sendctrl_lock, flags); + ipath_force_pio_avail_update(dd); /* * We just enabled pioavailupdate, so dma copy is almost certainly @@ -831,10 +946,13 @@ void ipath_clear_freeze(struct ipath_devdata *dd) */ for (i = 0; i < dd->ipath_pioavregs; i++) { /* deal with 6110 chip bug */ - im = i > 3 ? i ^ 1 : i; + im = (i > 3 && (dd->ipath_flags & IPATH_SWAP_PIOBUFS)) ? + i ^ 1 : i; val = ipath_read_kreg64(dd, (0x1000 / sizeof(u64)) + im); - dd->ipath_pioavailregs_dma[i] = dd->ipath_pioavailshadow[i] - = le64_to_cpu(val); + dd->ipath_pioavailregs_dma[i] = cpu_to_le64(val); + dd->ipath_pioavailshadow[i] = val | + (~dd->ipath_pioavailkernel[i] << + INFINIPATH_SENDPIOAVAIL_BUSY_SHIFT); } /* @@ -950,7 +1068,7 @@ set: * process was waiting for a packet to arrive, and didn't want * to poll */ -static void handle_urcv(struct ipath_devdata *dd, u32 istat) +static void handle_urcv(struct ipath_devdata *dd, u64 istat) { u64 portr; int i; @@ -966,12 +1084,13 @@ static void handle_urcv(struct ipath_devdata *dd, u32 istat) * and ipath_poll_next()... */ rmb(); - portr = ((istat >> INFINIPATH_I_RCVAVAIL_SHIFT) & - dd->ipath_i_rcvavail_mask) - | ((istat >> INFINIPATH_I_RCVURG_SHIFT) & - dd->ipath_i_rcvurg_mask); + portr = ((istat >> dd->ipath_i_rcvavail_shift) & + dd->ipath_i_rcvavail_mask) | + ((istat >> dd->ipath_i_rcvurg_shift) & + dd->ipath_i_rcvurg_mask); for (i = 1; i < dd->ipath_cfgports; i++) { struct ipath_portdata *pd = dd->ipath_pd[i]; + if (portr & (1 << i) && pd && pd->port_cnt) { if (test_and_clear_bit(IPATH_PORT_WAITING_RCV, &pd->port_flag)) { @@ -988,7 +1107,7 @@ static void handle_urcv(struct ipath_devdata *dd, u32 istat) } if (rcvdint) { /* only want to take one interrupt, so turn off the rcv - * interrupt for all the ports that we did the wakeup on + * interrupt for all the ports that we set the rcv_waiting * (but never for kernel port) */ ipath_write_kreg(dd, dd->ipath_kregs->kr_rcvctrl, @@ -999,12 +1118,11 @@ static void handle_urcv(struct ipath_devdata *dd, u32 istat) irqreturn_t ipath_intr(int irq, void *data) { struct ipath_devdata *dd = data; - u32 istat, chk0rcv = 0; + u64 istat, chk0rcv = 0; ipath_err_t estat = 0; irqreturn_t ret; static unsigned unexpected = 0; - static const u32 port0rbits = (1U<<INFINIPATH_I_RCVAVAIL_SHIFT) | - (1U<<INFINIPATH_I_RCVURG_SHIFT); + u64 kportrbits; ipath_stats.sps_ints++; @@ -1053,17 +1171,17 @@ irqreturn_t ipath_intr(int irq, void *data) if (unlikely(istat & ~dd->ipath_i_bitsextant)) ipath_dev_err(dd, - "interrupt with unknown interrupts %x set\n", - istat & (u32) ~ dd->ipath_i_bitsextant); - else - ipath_cdbg(VERBOSE, "intr stat=0x%x\n", istat); + "interrupt with unknown interrupts %Lx set\n", + istat & ~dd->ipath_i_bitsextant); + else if (istat & ~INFINIPATH_I_ERROR) /* errors do own printing */ + ipath_cdbg(VERBOSE, "intr stat=0x%Lx\n", istat); - if (unlikely(istat & INFINIPATH_I_ERROR)) { + if (istat & INFINIPATH_I_ERROR) { ipath_stats.sps_errints++; estat = ipath_read_kreg64(dd, dd->ipath_kregs->kr_errorstatus); if (!estat) - dev_info(&dd->pcidev->dev, "error interrupt (%x), " + dev_info(&dd->pcidev->dev, "error interrupt (%Lx), " "but no error bits set!\n", istat); else if (estat == -1LL) /* @@ -1073,9 +1191,7 @@ irqreturn_t ipath_intr(int irq, void *data) ipath_dev_err(dd, "Read of error status failed " "(all bits set); ignoring\n"); else - if (handle_errors(dd, estat)) - /* force calling ipath_kreceive() */ - chk0rcv = 1; + chk0rcv |= handle_errors(dd, estat); } if (istat & INFINIPATH_I_GPIO) { @@ -1093,8 +1209,7 @@ irqreturn_t ipath_intr(int irq, void *data) gpiostatus = ipath_read_kreg32( dd, dd->ipath_kregs->kr_gpio_status); - /* First the error-counter case. - */ + /* First the error-counter case. */ if ((gpiostatus & IPATH_GPIO_ERRINTR_MASK) && (dd->ipath_flags & IPATH_GPIO_ERRINTRS)) { /* want to clear the bits we see asserted. */ @@ -1156,7 +1271,6 @@ irqreturn_t ipath_intr(int irq, void *data) (u64) to_clear); } } - chk0rcv |= istat & port0rbits; /* * Clear the interrupt bits we found set, unless they are receive @@ -1169,22 +1283,25 @@ irqreturn_t ipath_intr(int irq, void *data) ipath_write_kreg(dd, dd->ipath_kregs->kr_intclear, istat); /* - * handle port0 receive before checking for pio buffers available, - * since receives can overflow; piobuf waiters can afford a few - * extra cycles, since they were waiting anyway, and user's waiting - * for receive are at the bottom. + * Handle kernel receive queues before checking for pio buffers + * available since receives can overflow; piobuf waiters can afford + * a few extra cycles, since they were waiting anyway, and user's + * waiting for receive are at the bottom. */ - if (chk0rcv) { + kportrbits = (1ULL << dd->ipath_i_rcvavail_shift) | + (1ULL << dd->ipath_i_rcvurg_shift); + if (chk0rcv || (istat & kportrbits)) { + istat &= ~kportrbits; ipath_kreceive(dd->ipath_pd[0]); - istat &= ~port0rbits; } - if (istat & ((dd->ipath_i_rcvavail_mask << - INFINIPATH_I_RCVAVAIL_SHIFT) - | (dd->ipath_i_rcvurg_mask << - INFINIPATH_I_RCVURG_SHIFT))) + if (istat & ((dd->ipath_i_rcvavail_mask << dd->ipath_i_rcvavail_shift) | + (dd->ipath_i_rcvurg_mask << dd->ipath_i_rcvurg_shift))) handle_urcv(dd, istat); + if (istat & (INFINIPATH_I_SDMAINT | INFINIPATH_I_SDMADISABLED)) + handle_sdma_intr(dd, istat); + if (istat & INFINIPATH_I_SPIOBUFAVAIL) { unsigned long flags; @@ -1195,7 +1312,10 @@ irqreturn_t ipath_intr(int irq, void *data) ipath_read_kreg64(dd, dd->ipath_kregs->kr_scratch); spin_unlock_irqrestore(&dd->ipath_sendctrl_lock, flags); - handle_layer_pioavail(dd); + if (!(dd->ipath_flags & IPATH_HAS_SEND_DMA)) + handle_layer_pioavail(dd); + else + ipath_dbg("unexpected BUFAVAIL intr\n"); } ret = IRQ_HANDLED; diff --git a/drivers/infiniband/hw/ipath/ipath_kernel.h b/drivers/infiniband/hw/ipath/ipath_kernel.h index ecf3f7ff7717..5863cbe99303 100644 --- a/drivers/infiniband/hw/ipath/ipath_kernel.h +++ b/drivers/infiniband/hw/ipath/ipath_kernel.h @@ -1,7 +1,7 @@ #ifndef _IPATH_KERNEL_H #define _IPATH_KERNEL_H /* - * Copyright (c) 2006, 2007 QLogic Corporation. All rights reserved. + * Copyright (c) 2006, 2007, 2008 QLogic Corporation. All rights reserved. * Copyright (c) 2003, 2004, 2005, 2006 PathScale, Inc. All rights reserved. * * This software is available to you under a choice of one of two @@ -42,6 +42,8 @@ #include <linux/pci.h> #include <linux/dma-mapping.h> #include <linux/mutex.h> +#include <linux/list.h> +#include <linux/scatterlist.h> #include <asm/io.h> #include <rdma/ib_verbs.h> @@ -175,9 +177,13 @@ struct ipath_portdata { u16 poll_type; /* port rcvhdrq head offset */ u32 port_head; + /* receive packet sequence counter */ + u32 port_seq_cnt; }; struct sk_buff; +struct ipath_sge_state; +struct ipath_verbs_txreq; /* * control information for layered drivers @@ -191,6 +197,40 @@ struct ipath_skbinfo { dma_addr_t phys; }; +struct ipath_sdma_txreq { + int flags; + int sg_count; + union { + struct scatterlist *sg; + void *map_addr; + }; + void (*callback)(void *, int); + void *callback_cookie; + int callback_status; + u16 start_idx; /* sdma private */ + u16 next_descq_idx; /* sdma private */ + struct list_head list; /* sdma private */ +}; + +struct ipath_sdma_desc { + __le64 qw[2]; +}; + +#define IPATH_SDMA_TXREQ_F_USELARGEBUF 0x1 +#define IPATH_SDMA_TXREQ_F_HEADTOHOST 0x2 +#define IPATH_SDMA_TXREQ_F_INTREQ 0x4 +#define IPATH_SDMA_TXREQ_F_FREEBUF 0x8 +#define IPATH_SDMA_TXREQ_F_FREEDESC 0x10 +#define IPATH_SDMA_TXREQ_F_VL15 0x20 + +#define IPATH_SDMA_TXREQ_S_OK 0 +#define IPATH_SDMA_TXREQ_S_SENDERROR 1 +#define IPATH_SDMA_TXREQ_S_ABORTED 2 +#define IPATH_SDMA_TXREQ_S_SHUTDOWN 3 + +/* max dwords in small buffer packet */ +#define IPATH_SMALLBUF_DWORDS (dd->ipath_piosize2k >> 2) + /* * Possible IB config parameters for ipath_f_get/set_ib_cfg() */ @@ -221,11 +261,6 @@ struct ipath_devdata { unsigned long ipath_physaddr; /* base of memory alloced for ipath_kregbase, for free */ u64 *ipath_kregalloc; - /* - * virtual address where port0 rcvhdrqtail updated for this unit. - * only written to by the chip, not the driver. - */ - volatile __le64 *ipath_hdrqtailptr; /* ipath_cfgports pointers */ struct ipath_portdata **ipath_pd; /* sk_buffs used by port 0 eager receive queue */ @@ -283,6 +318,7 @@ struct ipath_devdata { /* per chip actions needed for IB Link up/down changes */ int (*ipath_f_ib_updown)(struct ipath_devdata *, int, u64); + unsigned ipath_lastegr_idx; struct ipath_ibdev *verbs_dev; struct timer_list verbs_timer; /* total dwords sent (summed from counter) */ @@ -309,6 +345,7 @@ struct ipath_devdata { ipath_err_t ipath_lasthwerror; /* errors masked because they occur too fast */ ipath_err_t ipath_maskederrs; + u64 ipath_lastlinkrecov; /* link recoveries at last ACTIVE */ /* time in jiffies at which to re-enable maskederrs */ unsigned long ipath_unmasktime; /* count of egrfull errors, combined for all ports */ @@ -347,6 +384,7 @@ struct ipath_devdata { u32 ipath_lastrpkts; /* pio bufs allocated per port */ u32 ipath_pbufsport; + u32 ipath_pioupd_thresh; /* update threshold, some chips */ /* * number of ports configured as max; zero is set to number chip * supports, less gives more pio bufs/port, etc. @@ -365,6 +403,7 @@ struct ipath_devdata { * get to multiple devices */ u32 ipath_lastpioindex; + u32 ipath_lastpioindexl; /* max length of freezemsg */ u32 ipath_freezelen; /* @@ -381,6 +420,15 @@ struct ipath_devdata { u32 ipath_pcibar0; /* so we can rewrite it after a chip reset */ u32 ipath_pcibar1; + u32 ipath_x1_fix_tries; + u32 ipath_autoneg_tries; + u32 serdes_first_init_done; + + struct ipath_relock { + atomic_t ipath_relock_timer_active; + struct timer_list ipath_relock_timer; + unsigned int ipath_relock_interval; /* in jiffies */ + } ipath_relock_singleton; /* interrupt number */ int ipath_irq; @@ -403,7 +451,7 @@ struct ipath_devdata { u64 __iomem *ipath_egrtidbase; /* lock to workaround chip bug 9437 and others */ spinlock_t ipath_kernel_tid_lock; - spinlock_t ipath_tid_lock; + spinlock_t ipath_user_tid_lock; spinlock_t ipath_sendctrl_lock; /* @@ -422,11 +470,48 @@ struct ipath_devdata { struct class_device *diag_class_dev; /* timer used to prevent stats overflow, error throttling, etc. */ struct timer_list ipath_stats_timer; + /* timer to verify interrupts work, and fallback if possible */ + struct timer_list ipath_intrchk_timer; void *ipath_dummy_hdrq; /* used after port close */ dma_addr_t ipath_dummy_hdrq_phys; + /* SendDMA related entries */ + spinlock_t ipath_sdma_lock; + u64 ipath_sdma_status; + unsigned long ipath_sdma_abort_jiffies; + unsigned long ipath_sdma_abort_intr_timeout; + unsigned long ipath_sdma_buf_jiffies; + struct ipath_sdma_desc *ipath_sdma_descq; + u64 ipath_sdma_descq_added; + u64 ipath_sdma_descq_removed; + int ipath_sdma_desc_nreserved; + u16 ipath_sdma_descq_cnt; + u16 ipath_sdma_descq_tail; + u16 ipath_sdma_descq_head; + u16 ipath_sdma_next_intr; + u16 ipath_sdma_reset_wait; + u8 ipath_sdma_generation; + struct tasklet_struct ipath_sdma_abort_task; + struct tasklet_struct ipath_sdma_notify_task; + struct list_head ipath_sdma_activelist; + struct list_head ipath_sdma_notifylist; + atomic_t ipath_sdma_vl15_count; + struct timer_list ipath_sdma_vl15_timer; + + dma_addr_t ipath_sdma_descq_phys; + volatile __le64 *ipath_sdma_head_dma; + dma_addr_t ipath_sdma_head_phys; + unsigned long ipath_ureg_align; /* user register alignment */ + struct delayed_work ipath_autoneg_work; + wait_queue_head_t ipath_autoneg_wait; + + /* HoL blocking / user app forward-progress state */ + unsigned ipath_hol_state; + unsigned ipath_hol_next; + struct timer_list ipath_hol_timer; + /* * Shadow copies of registers; size indicates read access size. * Most of them are readonly, but some are write-only register, @@ -447,6 +532,8 @@ struct ipath_devdata { * init time. */ unsigned long ipath_pioavailshadow[8]; + /* bitmap of send buffers available for the kernel to use with PIO. */ + unsigned long ipath_pioavailkernel[8]; /* shadow of kr_gpio_out, for rmw ops */ u64 ipath_gpio_out; /* shadow the gpio mask register */ @@ -472,6 +559,8 @@ struct ipath_devdata { u64 ipath_intconfig; /* kr_sendpiobufbase value */ u64 ipath_piobufbase; + /* kr_ibcddrctrl shadow */ + u64 ipath_ibcddrctrl; /* these are the "32 bit" regs */ @@ -488,7 +577,10 @@ struct ipath_devdata { unsigned long ipath_rcvctrl; /* shadow kr_sendctrl */ unsigned long ipath_sendctrl; - unsigned long ipath_lastcancel; /* to not count armlaunch after cancel */ + /* to not count armlaunch after cancel */ + unsigned long ipath_lastcancel; + /* count cases where special trigger was needed (double write) */ + unsigned long ipath_spectriggerhit; /* value we put in kr_rcvhdrcnt */ u32 ipath_rcvhdrcnt; @@ -510,6 +602,7 @@ struct ipath_devdata { u32 ipath_piobcnt4k; /* size in bytes of "4KB" PIO buffers */ u32 ipath_piosize4k; + u32 ipath_pioreserved; /* reserved special-inkernel; */ /* kr_rcvegrbase value */ u32 ipath_rcvegrbase; /* kr_rcvegrcnt value */ @@ -546,10 +639,10 @@ struct ipath_devdata { u32 ipath_init_ibmaxlen; /* size of each rcvegrbuffer */ u32 ipath_rcvegrbufsize; - /* width (2,4,8,16,32) from HT config reg */ - u32 ipath_htwidth; - /* HT speed (200,400,800,1000) from HT config */ - u32 ipath_htspeed; + /* localbus width (1, 2,4,8,16,32) from config space */ + u32 ipath_lbus_width; + /* localbus speed (HT: 200,400,800,1000; PCIe 2500) */ + u32 ipath_lbus_speed; /* * number of sequential ibcstatus change for polling active/quiet * (i.e., link not coming up). @@ -573,21 +666,14 @@ struct ipath_devdata { */ u8 ipath_serial[16]; /* human readable board version */ - u8 ipath_boardversion[80]; + u8 ipath_boardversion[96]; + u8 ipath_lbus_info[32]; /* human readable localbus info */ /* chip major rev, from ipath_revision */ u8 ipath_majrev; /* chip minor rev, from ipath_revision */ u8 ipath_minrev; /* board rev, from ipath_revision */ u8 ipath_boardrev; - - u8 ipath_r_portenable_shift; - u8 ipath_r_intravail_shift; - u8 ipath_r_tailupd_shift; - u8 ipath_r_portcfg_shift; - - /* unit # of this chip, if present */ - int ipath_unit; /* saved for restore after reset */ u8 ipath_pci_cacheline; /* LID mask control */ @@ -603,6 +689,14 @@ struct ipath_devdata { /* Rx Polarity inversion (compensate for ~tx on partner) */ u8 ipath_rx_pol_inv; + u8 ipath_r_portenable_shift; + u8 ipath_r_intravail_shift; + u8 ipath_r_tailupd_shift; + u8 ipath_r_portcfg_shift; + + /* unit # of this chip, if present */ + int ipath_unit; + /* local link integrity counter */ u32 ipath_lli_counter; /* local link integrity errors */ @@ -617,9 +711,6 @@ struct ipath_devdata { u32 ipath_overrun_thresh_errs; u32 ipath_lli_errs; - /* status check work */ - struct delayed_work status_work; - /* * Not all devices managed by a driver instance are the same * type, so these fields must be per-device. @@ -632,8 +723,8 @@ struct ipath_devdata { * Below should be computable from number of ports, * since they are never modified. */ - u32 ipath_i_rcvavail_mask; - u32 ipath_i_rcvurg_mask; + u64 ipath_i_rcvavail_mask; + u64 ipath_i_rcvurg_mask; u16 ipath_i_rcvurg_shift; u16 ipath_i_rcvavail_shift; @@ -641,8 +732,9 @@ struct ipath_devdata { * Register bits for selecting i2c direction and values, used for * I2C serial flash. */ - u16 ipath_gpio_sda_num; - u16 ipath_gpio_scl_num; + u8 ipath_gpio_sda_num; + u8 ipath_gpio_scl_num; + u8 ipath_i2c_chain_type; u64 ipath_gpio_sda; u64 ipath_gpio_scl; @@ -703,13 +795,51 @@ struct ipath_devdata { /* interrupt mitigation reload register info */ u16 ipath_jint_idle_ticks; /* idle clock ticks */ u16 ipath_jint_max_packets; /* max packets across all ports */ + + /* + * lock for access to SerDes, and flags to sequence preset + * versus steady-state. 7220-only at the moment. + */ + spinlock_t ipath_sdepb_lock; + u8 ipath_presets_needed; /* Set if presets to be restored next DOWN */ }; +/* ipath_hol_state values (stopping/starting user proc, send flushing) */ +#define IPATH_HOL_UP 0 +#define IPATH_HOL_DOWN 1 +/* ipath_hol_next toggle values, used when hol_state IPATH_HOL_DOWN */ +#define IPATH_HOL_DOWNSTOP 0 +#define IPATH_HOL_DOWNCONT 1 + +/* bit positions for sdma_status */ +#define IPATH_SDMA_ABORTING 0 +#define IPATH_SDMA_DISARMED 1 +#define IPATH_SDMA_DISABLED 2 +#define IPATH_SDMA_LAYERBUF 3 +#define IPATH_SDMA_RUNNING 62 +#define IPATH_SDMA_SHUTDOWN 63 + +/* bit combinations that correspond to abort states */ +#define IPATH_SDMA_ABORT_NONE 0 +#define IPATH_SDMA_ABORT_ABORTING (1UL << IPATH_SDMA_ABORTING) +#define IPATH_SDMA_ABORT_DISARMED ((1UL << IPATH_SDMA_ABORTING) | \ + (1UL << IPATH_SDMA_DISARMED)) +#define IPATH_SDMA_ABORT_DISABLED ((1UL << IPATH_SDMA_ABORTING) | \ + (1UL << IPATH_SDMA_DISABLED)) +#define IPATH_SDMA_ABORT_ABORTED ((1UL << IPATH_SDMA_ABORTING) | \ + (1UL << IPATH_SDMA_DISARMED) | (1UL << IPATH_SDMA_DISABLED)) +#define IPATH_SDMA_ABORT_MASK ((1UL<<IPATH_SDMA_ABORTING) | \ + (1UL << IPATH_SDMA_DISARMED) | (1UL << IPATH_SDMA_DISABLED)) + +#define IPATH_SDMA_BUF_NONE 0 +#define IPATH_SDMA_BUF_MASK (1UL<<IPATH_SDMA_LAYERBUF) + /* Private data for file operations */ struct ipath_filedata { struct ipath_portdata *pd; unsigned subport; unsigned tidcursor; + struct ipath_user_sdma_queue *pq; }; extern struct list_head ipath_dev_list; extern spinlock_t ipath_devs_lock; @@ -718,7 +848,7 @@ extern struct ipath_devdata *ipath_lookup(int unit); int ipath_init_chip(struct ipath_devdata *, int); int ipath_enable_wc(struct ipath_devdata *dd); void ipath_disable_wc(struct ipath_devdata *dd); -int ipath_count_units(int *npresentp, int *nupp, u32 *maxportsp); +int ipath_count_units(int *npresentp, int *nupp, int *maxportsp); void ipath_shutdown_device(struct ipath_devdata *); void ipath_clear_freeze(struct ipath_devdata *); @@ -741,7 +871,8 @@ struct sk_buff *ipath_alloc_skb(struct ipath_devdata *dd, gfp_t); extern int ipath_diag_inuse; irqreturn_t ipath_intr(int irq, void *devid); -int ipath_decode_err(char *buf, size_t blen, ipath_err_t err); +int ipath_decode_err(struct ipath_devdata *dd, char *buf, size_t blen, + ipath_err_t err); #if __IPATH_INFO || __IPATH_DBG extern const char *ipath_ibcstatus_str[]; #endif @@ -774,6 +905,13 @@ int ipath_set_lid(struct ipath_devdata *, u32, u8); int ipath_set_rx_pol_inv(struct ipath_devdata *dd, u8 new_pol_inv); void ipath_enable_armlaunch(struct ipath_devdata *); void ipath_disable_armlaunch(struct ipath_devdata *); +void ipath_hol_down(struct ipath_devdata *); +void ipath_hol_up(struct ipath_devdata *); +void ipath_hol_event(unsigned long); +void ipath_toggle_rclkrls(struct ipath_devdata *); +void ipath_sd7220_clr_ibpar(struct ipath_devdata *); +void ipath_set_relock_poll(struct ipath_devdata *, int); +void ipath_shutdown_relock_poll(struct ipath_devdata *); /* for use in system calls, where we want to know device type, etc. */ #define port_fp(fp) ((struct ipath_filedata *)(fp)->private_data)->pd @@ -781,11 +919,15 @@ void ipath_disable_armlaunch(struct ipath_devdata *); ((struct ipath_filedata *)(fp)->private_data)->subport #define tidcursor_fp(fp) \ ((struct ipath_filedata *)(fp)->private_data)->tidcursor +#define user_sdma_queue_fp(fp) \ + ((struct ipath_filedata *)(fp)->private_data)->pq /* * values for ipath_flags */ -/* The chip is up and initted */ + /* chip can report link latency (IB 1.2) */ +#define IPATH_HAS_LINK_LATENCY 0x1 + /* The chip is up and initted */ #define IPATH_INITTED 0x2 /* set if any user code has set kr_rcvhdrsize */ #define IPATH_RCVHDRSZ_SET 0x4 @@ -809,6 +951,8 @@ void ipath_disable_armlaunch(struct ipath_devdata *); #define IPATH_LINKUNK 0x400 /* Write combining flush needed for PIO */ #define IPATH_PIO_FLUSH_WC 0x1000 + /* DMA Receive tail pointer */ +#define IPATH_NODMA_RTAIL 0x2000 /* no IB cable, or no device on IB cable */ #define IPATH_NOCABLE 0x4000 /* Supports port zero per packet receive interrupts via @@ -819,16 +963,26 @@ void ipath_disable_armlaunch(struct ipath_devdata *); /* packet/word counters are 32 bit, else those 4 counters * are 64bit */ #define IPATH_32BITCOUNTERS 0x20000 - /* can miss port0 rx interrupts */ /* Interrupt register is 64 bits */ #define IPATH_INTREG_64 0x40000 + /* can miss port0 rx interrupts */ #define IPATH_DISABLED 0x80000 /* administratively disabled */ /* Use GPIO interrupts for new counters */ #define IPATH_GPIO_ERRINTRS 0x100000 #define IPATH_SWAP_PIOBUFS 0x200000 + /* Supports Send DMA */ +#define IPATH_HAS_SEND_DMA 0x400000 + /* Supports Send Count (not just word count) in PBC */ +#define IPATH_HAS_PBC_CNT 0x800000 /* Suppress heartbeat, even if turning off loopback */ #define IPATH_NO_HRTBT 0x1000000 +#define IPATH_HAS_THRESH_UPDATE 0x4000000 #define IPATH_HAS_MULT_IB_SPEED 0x8000000 +#define IPATH_IB_AUTONEG_INPROG 0x10000000 +#define IPATH_IB_AUTONEG_FAILED 0x20000000 + /* Linkdown-disable intentionally, Do not attempt to bring up */ +#define IPATH_IB_LINK_DISABLED 0x40000000 +#define IPATH_IB_FORCE_NOTIFY 0x80000000 /* force notify on next ib change */ /* Bits in GPIO for the added interrupts */ #define IPATH_GPIO_PORT0_BIT 2 @@ -847,13 +1001,18 @@ void ipath_disable_armlaunch(struct ipath_devdata *); /* free up any allocated data at closes */ void ipath_free_data(struct ipath_portdata *dd); -u32 __iomem *ipath_getpiobuf(struct ipath_devdata *, u32 *); +u32 __iomem *ipath_getpiobuf(struct ipath_devdata *, u32, u32 *); +void ipath_chg_pioavailkernel(struct ipath_devdata *dd, unsigned start, + unsigned len, int avail); +void ipath_init_iba7220_funcs(struct ipath_devdata *); void ipath_init_iba6120_funcs(struct ipath_devdata *); void ipath_init_iba6110_funcs(struct ipath_devdata *); void ipath_get_eeprom_info(struct ipath_devdata *); int ipath_update_eeprom_log(struct ipath_devdata *dd); void ipath_inc_eeprom_err(struct ipath_devdata *dd, u32 eidx, u32 incr); u64 ipath_snap_cntr(struct ipath_devdata *, ipath_creg); +void ipath_disarm_senderrbufs(struct ipath_devdata *, int); +void ipath_force_pio_avail_update(struct ipath_devdata *); void signal_ib_event(struct ipath_devdata *dd, enum ib_event_type ev); /* @@ -865,6 +1024,34 @@ void signal_ib_event(struct ipath_devdata *dd, enum ib_event_type ev); #define IPATH_LED_LOG 2 /* Logical (link) YELLOW LED */ void ipath_set_led_override(struct ipath_devdata *dd, unsigned int val); +/* send dma routines */ +int setup_sdma(struct ipath_devdata *); +void teardown_sdma(struct ipath_devdata *); +void ipath_restart_sdma(struct ipath_devdata *); +void ipath_sdma_intr(struct ipath_devdata *); +int ipath_sdma_verbs_send(struct ipath_devdata *, struct ipath_sge_state *, + u32, struct ipath_verbs_txreq *); +/* ipath_sdma_lock should be locked before calling this. */ +int ipath_sdma_make_progress(struct ipath_devdata *dd); + +/* must be called under ipath_sdma_lock */ +static inline u16 ipath_sdma_descq_freecnt(const struct ipath_devdata *dd) +{ + return dd->ipath_sdma_descq_cnt - + (dd->ipath_sdma_descq_added - dd->ipath_sdma_descq_removed) - + 1 - dd->ipath_sdma_desc_nreserved; +} + +static inline void ipath_sdma_desc_reserve(struct ipath_devdata *dd, u16 cnt) +{ + dd->ipath_sdma_desc_nreserved += cnt; +} + +static inline void ipath_sdma_desc_unreserve(struct ipath_devdata *dd, u16 cnt) +{ + dd->ipath_sdma_desc_nreserved -= cnt; +} + /* * number of words used for protocol header if not set by ipath_userinit(); */ @@ -875,6 +1062,8 @@ void ipath_release_user_pages(struct page **, size_t); void ipath_release_user_pages_on_close(struct page **, size_t); int ipath_eeprom_read(struct ipath_devdata *, u8, void *, int); int ipath_eeprom_write(struct ipath_devdata *, u8, const void *, int); +int ipath_tempsense_read(struct ipath_devdata *, u8 regnum); +int ipath_tempsense_write(struct ipath_devdata *, u8 regnum, u8 data); /* these are used for the registers that vary with port */ void ipath_write_kreg_port(const struct ipath_devdata *, ipath_kreg, @@ -891,8 +1080,7 @@ void ipath_write_kreg_port(const struct ipath_devdata *, ipath_kreg, /* * At the moment, none of the s-registers are writable, so no - * ipath_write_sreg(), and none of the c-registers are writable, so no - * ipath_write_creg(). + * ipath_write_sreg(). */ /** @@ -1001,6 +1189,27 @@ static inline u32 ipath_get_rcvhdrtail(const struct ipath_portdata *pd) pd->port_rcvhdrtail_kvaddr)); } +static inline u32 ipath_get_hdrqtail(const struct ipath_portdata *pd) +{ + const struct ipath_devdata *dd = pd->port_dd; + u32 hdrqtail; + + if (dd->ipath_flags & IPATH_NODMA_RTAIL) { + __le32 *rhf_addr; + u32 seq; + + rhf_addr = (__le32 *) pd->port_rcvhdrq + + pd->port_head + dd->ipath_rhf_offset; + seq = ipath_hdrget_seq(rhf_addr); + hdrqtail = pd->port_head; + if (seq == pd->port_seq_cnt) + hdrqtail++; + } else + hdrqtail = ipath_get_rcvhdrtail(pd); + + return hdrqtail; +} + static inline u64 ipath_read_ireg(const struct ipath_devdata *dd, ipath_kreg r) { return (dd->ipath_flags & IPATH_INTREG_64) ? @@ -1029,6 +1238,21 @@ static inline u32 ipath_ib_linktrstate(struct ipath_devdata *dd, u64 ibcs) } /* + * from contents of IBCStatus (or a saved copy), return logical link state + * combination of link state and linktraining state (down, active, init, + * arm, etc. + */ +static inline u32 ipath_ib_state(struct ipath_devdata *dd, u64 ibcs) +{ + u32 ibs; + ibs = (u32)(ibcs >> INFINIPATH_IBCS_LINKTRAININGSTATE_SHIFT) & + dd->ibcs_lts_mask; + ibs |= (u32)(ibcs & + (INFINIPATH_IBCS_LINKSTATE_MASK << dd->ibcs_ls_shift)); + return ibs; +} + +/* * sysfs interface. */ @@ -1053,6 +1277,7 @@ int ipathfs_remove_device(struct ipath_devdata *); dma_addr_t ipath_map_page(struct pci_dev *, struct page *, unsigned long, size_t, int); dma_addr_t ipath_map_single(struct pci_dev *, void *, size_t, int); +const char *ipath_get_unit_name(int unit); /* * Flush write combining store buffers (if present) and perform a write @@ -1065,11 +1290,8 @@ dma_addr_t ipath_map_single(struct pci_dev *, void *, size_t, int); #endif extern unsigned ipath_debug; /* debugging bit mask */ - -#define IPATH_MAX_PARITY_ATTEMPTS 10000 /* max times to try recovery */ - -const char *ipath_get_unit_name(int unit); - +extern unsigned ipath_linkrecovery; +extern unsigned ipath_mtu4096; extern struct mutex ipath_mutex; #define IPATH_DRV_NAME "ib_ipath" @@ -1096,7 +1318,7 @@ extern struct mutex ipath_mutex; # define __IPATH_DBG_WHICH(which,fmt,...) \ do { \ - if(unlikely(ipath_debug&(which))) \ + if (unlikely(ipath_debug & (which))) \ printk(KERN_DEBUG IPATH_DRV_NAME ": %s: " fmt, \ __func__,##__VA_ARGS__); \ } while(0) diff --git a/drivers/infiniband/hw/ipath/ipath_mad.c b/drivers/infiniband/hw/ipath/ipath_mad.c index b34b91d3723a..1ff46ae7dd99 100644 --- a/drivers/infiniband/hw/ipath/ipath_mad.c +++ b/drivers/infiniband/hw/ipath/ipath_mad.c @@ -1,5 +1,5 @@ /* - * Copyright (c) 2006, 2007 QLogic Corporation. All rights reserved. + * Copyright (c) 2006, 2007, 2008 QLogic Corporation. All rights reserved. * Copyright (c) 2005, 2006 PathScale, Inc. All rights reserved. * * This software is available to you under a choice of one of two @@ -146,6 +146,15 @@ static int recv_subn_get_guidinfo(struct ib_smp *smp, return reply(smp); } +static void set_link_width_enabled(struct ipath_devdata *dd, u32 w) +{ + (void) dd->ipath_f_set_ib_cfg(dd, IPATH_IB_CFG_LWID_ENB, w); +} + +static void set_link_speed_enabled(struct ipath_devdata *dd, u32 s) +{ + (void) dd->ipath_f_set_ib_cfg(dd, IPATH_IB_CFG_SPD_ENB, s); +} static int get_overrunthreshold(struct ipath_devdata *dd) { @@ -226,6 +235,7 @@ static int recv_subn_get_portinfo(struct ib_smp *smp, struct ib_device *ibdev, u8 port) { struct ipath_ibdev *dev; + struct ipath_devdata *dd; struct ib_port_info *pip = (struct ib_port_info *)smp->data; u16 lid; u8 ibcstat; @@ -239,6 +249,7 @@ static int recv_subn_get_portinfo(struct ib_smp *smp, } dev = to_idev(ibdev); + dd = dev->dd; /* Clear all fields. Only set the non-zero fields. */ memset(smp->data, 0, sizeof(smp->data)); @@ -248,25 +259,28 @@ static int recv_subn_get_portinfo(struct ib_smp *smp, dev->mkeyprot == 0) pip->mkey = dev->mkey; pip->gid_prefix = dev->gid_prefix; - lid = dev->dd->ipath_lid; + lid = dd->ipath_lid; pip->lid = lid ? cpu_to_be16(lid) : IB_LID_PERMISSIVE; pip->sm_lid = cpu_to_be16(dev->sm_lid); pip->cap_mask = cpu_to_be32(dev->port_cap_flags); /* pip->diag_code; */ pip->mkey_lease_period = cpu_to_be16(dev->mkey_lease_period); pip->local_port_num = port; - pip->link_width_enabled = dev->link_width_enabled; - pip->link_width_supported = 3; /* 1x or 4x */ - pip->link_width_active = 2; /* 4x */ - pip->linkspeed_portstate = 0x10; /* 2.5Gbps */ - ibcstat = dev->dd->ipath_lastibcstat; - pip->linkspeed_portstate |= ((ibcstat >> 4) & 0x3) + 1; + pip->link_width_enabled = dd->ipath_link_width_enabled; + pip->link_width_supported = dd->ipath_link_width_supported; + pip->link_width_active = dd->ipath_link_width_active; + pip->linkspeed_portstate = dd->ipath_link_speed_supported << 4; + ibcstat = dd->ipath_lastibcstat; + /* map LinkState to IB portinfo values. */ + pip->linkspeed_portstate |= ipath_ib_linkstate(dd, ibcstat) + 1; + pip->portphysstate_linkdown = - (ipath_cvt_physportstate[ibcstat & 0xf] << 4) | - (get_linkdowndefaultstate(dev->dd) ? 1 : 2); - pip->mkeyprot_resv_lmc = (dev->mkeyprot << 6) | dev->dd->ipath_lmc; - pip->linkspeedactive_enabled = 0x11; /* 2.5Gbps, 2.5Gbps */ - switch (dev->dd->ipath_ibmtu) { + (ipath_cvt_physportstate[ibcstat & dd->ibcs_lts_mask] << 4) | + (get_linkdowndefaultstate(dd) ? 1 : 2); + pip->mkeyprot_resv_lmc = (dev->mkeyprot << 6) | dd->ipath_lmc; + pip->linkspeedactive_enabled = (dd->ipath_link_speed_active << 4) | + dd->ipath_link_speed_enabled; + switch (dd->ipath_ibmtu) { case 4096: mtu = IB_MTU_4096; break; @@ -292,19 +306,15 @@ static int recv_subn_get_portinfo(struct ib_smp *smp, /* pip->vl_arb_high_cap; // only one VL */ /* pip->vl_arb_low_cap; // only one VL */ /* InitTypeReply = 0 */ - /* - * Note: the chips support a maximum MTU of 4096, but the driver - * hasn't implemented this feature yet, so set the maximum value - * to 2048. - */ - pip->inittypereply_mtucap = IB_MTU_2048; - // HCAs ignore VLStallCount and HOQLife + /* our mtu cap depends on whether 4K MTU enabled or not */ + pip->inittypereply_mtucap = ipath_mtu4096 ? IB_MTU_4096 : IB_MTU_2048; + /* HCAs ignore VLStallCount and HOQLife */ /* pip->vlstallcnt_hoqlife; */ pip->operationalvl_pei_peo_fpi_fpo = 0x10; /* OVLs = 1 */ pip->mkey_violations = cpu_to_be16(dev->mkey_violations); /* P_KeyViolations are counted by hardware. */ pip->pkey_violations = - cpu_to_be16((ipath_get_cr_errpkey(dev->dd) - + cpu_to_be16((ipath_get_cr_errpkey(dd) - dev->z_pkey_violations) & 0xFFFF); pip->qkey_violations = cpu_to_be16(dev->qkey_violations); /* Only the hardware GUID is supported for now */ @@ -313,10 +323,17 @@ static int recv_subn_get_portinfo(struct ib_smp *smp, /* 32.768 usec. response time (guessing) */ pip->resv_resptimevalue = 3; pip->localphyerrors_overrunerrors = - (get_phyerrthreshold(dev->dd) << 4) | - get_overrunthreshold(dev->dd); + (get_phyerrthreshold(dd) << 4) | + get_overrunthreshold(dd); /* pip->max_credit_hint; */ - /* pip->link_roundtrip_latency[3]; */ + if (dev->port_cap_flags & IB_PORT_LINK_LATENCY_SUP) { + u32 v; + + v = dd->ipath_f_get_ib_cfg(dd, IPATH_IB_CFG_LINKLATENCY); + pip->link_roundtrip_latency[0] = v >> 16; + pip->link_roundtrip_latency[1] = v >> 8; + pip->link_roundtrip_latency[2] = v; + } ret = reply(smp); @@ -444,19 +461,25 @@ static int recv_subn_set_portinfo(struct ib_smp *smp, ib_dispatch_event(&event); } - /* Only 4x supported but allow 1x or 4x to be set (see 14.2.6.6). */ + /* Allow 1x or 4x to be set (see 14.2.6.6). */ lwe = pip->link_width_enabled; - if ((lwe >= 4 && lwe <= 8) || (lwe >= 0xC && lwe <= 0xFE)) - goto err; - if (lwe == 0xFF) - dev->link_width_enabled = 3; /* 1x or 4x */ - else if (lwe) - dev->link_width_enabled = lwe; + if (lwe) { + if (lwe == 0xFF) + lwe = dd->ipath_link_width_supported; + else if (lwe >= 16 || (lwe & ~dd->ipath_link_width_supported)) + goto err; + set_link_width_enabled(dd, lwe); + } - /* Only 2.5 Gbs supported. */ + /* Allow 2.5 or 5.0 Gbs. */ lse = pip->linkspeedactive_enabled & 0xF; - if (lse >= 2 && lse <= 0xE) - goto err; + if (lse) { + if (lse == 15) + lse = dd->ipath_link_speed_supported; + else if (lse >= 8 || (lse & ~dd->ipath_link_speed_supported)) + goto err; + set_link_speed_enabled(dd, lse); + } /* Set link down default state. */ switch (pip->portphysstate_linkdown & 0xF) { @@ -491,6 +514,8 @@ static int recv_subn_set_portinfo(struct ib_smp *smp, mtu = 2048; break; case IB_MTU_4096: + if (!ipath_mtu4096) + goto err; mtu = 4096; break; default: @@ -565,6 +590,10 @@ static int recv_subn_set_portinfo(struct ib_smp *smp, else goto err; ipath_set_linkstate(dd, lstate); + if (lstate == IPATH_IB_LINKDOWN_DISABLE) { + ret = IB_MAD_RESULT_SUCCESS | IB_MAD_RESULT_CONSUMED; + goto done; + } ipath_wait_linkstate(dd, IPATH_LINKINIT | IPATH_LINKARMED | IPATH_LINKACTIVE, 1000); break; @@ -948,10 +977,14 @@ static int recv_pma_get_portsamplescontrol(struct ib_perf *pmp, * nsec. 0 == 4 nsec., 1 == 8 nsec., ..., 255 == 1020 nsec. Sample * intervals are counted in ticks. Since we use Linux timers, that * count in jiffies, we can't sample for less than 1000 ticks if HZ - * == 1000 (4000 ticks if HZ is 250). + * == 1000 (4000 ticks if HZ is 250). link_speed_active returns 2 for + * DDR, 1 for SDR, set the tick to 1 for DDR, 0 for SDR on chips that + * have hardware support for delaying packets. */ - /* XXX This is WRONG. */ - p->tick = 250; /* 1 usec. */ + if (crp->cr_psstat) + p->tick = dev->dd->ipath_link_speed_active - 1; + else + p->tick = 250; /* 1 usec. */ p->counter_width = 4; /* 32 bit counters */ p->counter_mask0_9 = COUNTER_MASK0_9; spin_lock_irqsave(&dev->pending_lock, flags); @@ -1364,7 +1397,8 @@ static int process_subn(struct ib_device *ibdev, int mad_flags, } /* Is the mkey in the process of expiring? */ - if (dev->mkey_lease_timeout && jiffies >= dev->mkey_lease_timeout) { + if (dev->mkey_lease_timeout && + time_after_eq(jiffies, dev->mkey_lease_timeout)) { /* Clear timeout and mkey protection field. */ dev->mkey_lease_timeout = 0; dev->mkeyprot = 0; diff --git a/drivers/infiniband/hw/ipath/ipath_qp.c b/drivers/infiniband/hw/ipath/ipath_qp.c index 087ed3166479..dd5b6e9d57c2 100644 --- a/drivers/infiniband/hw/ipath/ipath_qp.c +++ b/drivers/infiniband/hw/ipath/ipath_qp.c @@ -1,5 +1,5 @@ /* - * Copyright (c) 2006, 2007 QLogic Corporation. All rights reserved. + * Copyright (c) 2006, 2007, 2008 QLogic Corporation. All rights reserved. * Copyright (c) 2005, 2006 PathScale, Inc. All rights reserved. * * This software is available to you under a choice of one of two @@ -340,6 +340,7 @@ static void ipath_reset_qp(struct ipath_qp *qp, enum ib_qp_type type) qp->s_flags &= IPATH_S_SIGNAL_REQ_WR; qp->s_hdrwords = 0; qp->s_wqe = NULL; + qp->s_pkt_delay = 0; qp->s_psn = 0; qp->r_psn = 0; qp->r_msn = 0; @@ -392,7 +393,6 @@ int ipath_error_qp(struct ipath_qp *qp, enum ib_wc_status err) qp->ibqp.qp_num, qp->remote_qpn, err); spin_lock(&dev->pending_lock); - /* XXX What if its already removed by the timeout code? */ if (!list_empty(&qp->timerwait)) list_del_init(&qp->timerwait); if (!list_empty(&qp->piowait)) @@ -516,13 +516,13 @@ int ipath_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, goto inval; /* - * Note: the chips support a maximum MTU of 4096, but the driver - * hasn't implemented this feature yet, so don't allow Path MTU - * values greater than 2048. + * don't allow invalid Path MTU values or greater than 2048 + * unless we are configured for a 4KB MTU */ - if (attr_mask & IB_QP_PATH_MTU) - if (attr->path_mtu > IB_MTU_2048) - goto inval; + if ((attr_mask & IB_QP_PATH_MTU) && + (ib_mtu_enum_to_int(attr->path_mtu) == -1 || + (attr->path_mtu > IB_MTU_2048 && !ipath_mtu4096))) + goto inval; if (attr_mask & IB_QP_PATH_MIG_STATE) if (attr->path_mig_state != IB_MIG_MIGRATED && @@ -564,8 +564,10 @@ int ipath_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, if (attr_mask & IB_QP_ACCESS_FLAGS) qp->qp_access_flags = attr->qp_access_flags; - if (attr_mask & IB_QP_AV) + if (attr_mask & IB_QP_AV) { qp->remote_ah_attr = attr->ah_attr; + qp->s_dmult = ipath_ib_rate_to_mult(attr->ah_attr.static_rate); + } if (attr_mask & IB_QP_PATH_MTU) qp->path_mtu = attr->path_mtu; @@ -748,22 +750,33 @@ struct ib_qp *ipath_create_qp(struct ib_pd *ibpd, size_t sz; struct ib_qp *ret; - if (init_attr->cap.max_send_sge > ib_ipath_max_sges || - init_attr->cap.max_recv_sge > ib_ipath_max_sges || - init_attr->cap.max_send_wr > ib_ipath_max_qp_wrs || - init_attr->cap.max_recv_wr > ib_ipath_max_qp_wrs) { - ret = ERR_PTR(-ENOMEM); + if (init_attr->create_flags) { + ret = ERR_PTR(-EINVAL); goto bail; } - if (init_attr->cap.max_send_sge + - init_attr->cap.max_recv_sge + - init_attr->cap.max_send_wr + - init_attr->cap.max_recv_wr == 0) { + if (init_attr->cap.max_send_sge > ib_ipath_max_sges || + init_attr->cap.max_send_wr > ib_ipath_max_qp_wrs) { ret = ERR_PTR(-EINVAL); goto bail; } + /* Check receive queue parameters if no SRQ is specified. */ + if (!init_attr->srq) { + if (init_attr->cap.max_recv_sge > ib_ipath_max_sges || + init_attr->cap.max_recv_wr > ib_ipath_max_qp_wrs) { + ret = ERR_PTR(-EINVAL); + goto bail; + } + if (init_attr->cap.max_send_sge + + init_attr->cap.max_send_wr + + init_attr->cap.max_recv_sge + + init_attr->cap.max_recv_wr == 0) { + ret = ERR_PTR(-EINVAL); + goto bail; + } + } + switch (init_attr->qp_type) { case IB_QPT_UC: case IB_QPT_RC: @@ -840,6 +853,7 @@ struct ib_qp *ipath_create_qp(struct ib_pd *ibpd, goto bail_qp; } qp->ip = NULL; + qp->s_tx = NULL; ipath_reset_qp(qp, init_attr->qp_type); break; @@ -945,12 +959,20 @@ int ipath_destroy_qp(struct ib_qp *ibqp) /* Stop the sending tasklet. */ tasklet_kill(&qp->s_task); + if (qp->s_tx) { + atomic_dec(&qp->refcount); + if (qp->s_tx->txreq.flags & IPATH_SDMA_TXREQ_F_FREEBUF) + kfree(qp->s_tx->txreq.map_addr); + } + /* Make sure the QP isn't on the timeout list. */ spin_lock_irqsave(&dev->pending_lock, flags); if (!list_empty(&qp->timerwait)) list_del_init(&qp->timerwait); if (!list_empty(&qp->piowait)) list_del_init(&qp->piowait); + if (qp->s_tx) + list_add(&qp->s_tx->txreq.list, &dev->txreq_free); spin_unlock_irqrestore(&dev->pending_lock, flags); /* @@ -1021,7 +1043,6 @@ void ipath_sqerror_qp(struct ipath_qp *qp, struct ib_wc *wc) qp->ibqp.qp_num, qp->remote_qpn, wc->status); spin_lock(&dev->pending_lock); - /* XXX What if its already removed by the timeout code? */ if (!list_empty(&qp->timerwait)) list_del_init(&qp->timerwait); if (!list_empty(&qp->piowait)) diff --git a/drivers/infiniband/hw/ipath/ipath_rc.c b/drivers/infiniband/hw/ipath/ipath_rc.c index 40f3e37d7adc..c405dfba5531 100644 --- a/drivers/infiniband/hw/ipath/ipath_rc.c +++ b/drivers/infiniband/hw/ipath/ipath_rc.c @@ -1,5 +1,5 @@ /* - * Copyright (c) 2006, 2007 QLogic Corporation. All rights reserved. + * Copyright (c) 2006, 2007, 2008 QLogic Corporation. All rights reserved. * Copyright (c) 2005, 2006 PathScale, Inc. All rights reserved. * * This software is available to you under a choice of one of two @@ -31,6 +31,8 @@ * SOFTWARE. */ +#include <linux/io.h> + #include "ipath_verbs.h" #include "ipath_kernel.h" @@ -306,7 +308,7 @@ int ipath_make_rc_req(struct ipath_qp *qp) else { qp->s_state = OP(SEND_ONLY_WITH_IMMEDIATE); /* Immediate data comes after the BTH */ - ohdr->u.imm_data = wqe->wr.imm_data; + ohdr->u.imm_data = wqe->wr.ex.imm_data; hwords += 1; } if (wqe->wr.send_flags & IB_SEND_SOLICITED) @@ -344,7 +346,7 @@ int ipath_make_rc_req(struct ipath_qp *qp) qp->s_state = OP(RDMA_WRITE_ONLY_WITH_IMMEDIATE); /* Immediate data comes after RETH */ - ohdr->u.rc.imm_data = wqe->wr.imm_data; + ohdr->u.rc.imm_data = wqe->wr.ex.imm_data; hwords += 1; if (wqe->wr.send_flags & IB_SEND_SOLICITED) bth0 |= 1 << 23; @@ -488,7 +490,7 @@ int ipath_make_rc_req(struct ipath_qp *qp) else { qp->s_state = OP(SEND_LAST_WITH_IMMEDIATE); /* Immediate data comes after the BTH */ - ohdr->u.imm_data = wqe->wr.imm_data; + ohdr->u.imm_data = wqe->wr.ex.imm_data; hwords += 1; } if (wqe->wr.send_flags & IB_SEND_SOLICITED) @@ -524,7 +526,7 @@ int ipath_make_rc_req(struct ipath_qp *qp) else { qp->s_state = OP(RDMA_WRITE_LAST_WITH_IMMEDIATE); /* Immediate data comes after the BTH */ - ohdr->u.imm_data = wqe->wr.imm_data; + ohdr->u.imm_data = wqe->wr.ex.imm_data; hwords += 1; if (wqe->wr.send_flags & IB_SEND_SOLICITED) bth0 |= 1 << 23; @@ -585,19 +587,39 @@ bail: static void send_rc_ack(struct ipath_qp *qp) { struct ipath_ibdev *dev = to_idev(qp->ibqp.device); + struct ipath_devdata *dd; u16 lrh0; u32 bth0; u32 hwords; + u32 __iomem *piobuf; struct ipath_ib_header hdr; struct ipath_other_headers *ohdr; unsigned long flags; + spin_lock_irqsave(&qp->s_lock, flags); + /* Don't send ACK or NAK if a RDMA read or atomic is pending. */ if (qp->r_head_ack_queue != qp->s_tail_ack_queue || (qp->s_flags & IPATH_S_ACK_PENDING) || qp->s_ack_state != OP(ACKNOWLEDGE)) goto queue_ack; + spin_unlock_irqrestore(&qp->s_lock, flags); + + dd = dev->dd; + piobuf = ipath_getpiobuf(dd, 0, NULL); + if (!piobuf) { + /* + * We are out of PIO buffers at the moment. + * Pass responsibility for sending the ACK to the + * send tasklet so that when a PIO buffer becomes + * available, the ACK is sent ahead of other outgoing + * packets. + */ + spin_lock_irqsave(&qp->s_lock, flags); + goto queue_ack; + } + /* Construct the header. */ ohdr = &hdr.u.oth; lrh0 = IPATH_LRH_BTH; @@ -611,7 +633,7 @@ static void send_rc_ack(struct ipath_qp *qp) lrh0 = IPATH_LRH_GRH; } /* read pkey_index w/o lock (its atomic) */ - bth0 = ipath_get_pkey(dev->dd, qp->s_pkey_index) | + bth0 = ipath_get_pkey(dd, qp->s_pkey_index) | (OP(ACKNOWLEDGE) << 24) | (1 << 22); if (qp->r_nak_state) ohdr->u.aeth = cpu_to_be32((qp->r_msn & IPATH_MSN_MASK) | @@ -623,30 +645,29 @@ static void send_rc_ack(struct ipath_qp *qp) hdr.lrh[0] = cpu_to_be16(lrh0); hdr.lrh[1] = cpu_to_be16(qp->remote_ah_attr.dlid); hdr.lrh[2] = cpu_to_be16(hwords + SIZE_OF_CRC); - hdr.lrh[3] = cpu_to_be16(dev->dd->ipath_lid); + hdr.lrh[3] = cpu_to_be16(dd->ipath_lid); ohdr->bth[0] = cpu_to_be32(bth0); ohdr->bth[1] = cpu_to_be32(qp->remote_qpn); ohdr->bth[2] = cpu_to_be32(qp->r_ack_psn & IPATH_PSN_MASK); - /* - * If we can send the ACK, clear the ACK state. - */ - if (ipath_verbs_send(qp, &hdr, hwords, NULL, 0) == 0) { - dev->n_unicast_xmit++; - goto done; - } + writeq(hwords + 1, piobuf); - /* - * We are out of PIO buffers at the moment. - * Pass responsibility for sending the ACK to the - * send tasklet so that when a PIO buffer becomes - * available, the ACK is sent ahead of other outgoing - * packets. - */ - dev->n_rc_qacks++; + if (dd->ipath_flags & IPATH_PIO_FLUSH_WC) { + u32 *hdrp = (u32 *) &hdr; + + ipath_flush_wc(); + __iowrite32_copy(piobuf + 2, hdrp, hwords - 1); + ipath_flush_wc(); + __raw_writel(hdrp[hwords - 1], piobuf + hwords + 1); + } else + __iowrite32_copy(piobuf + 2, (u32 *) &hdr, hwords); + + ipath_flush_wc(); + + dev->n_unicast_xmit++; + goto done; queue_ack: - spin_lock_irqsave(&qp->s_lock, flags); dev->n_rc_qacks++; qp->s_flags |= IPATH_S_ACK_PENDING; qp->s_nak_state = qp->r_nak_state; diff --git a/drivers/infiniband/hw/ipath/ipath_registers.h b/drivers/infiniband/hw/ipath/ipath_registers.h index 92ad73a7fff0..8f44d0cf3833 100644 --- a/drivers/infiniband/hw/ipath/ipath_registers.h +++ b/drivers/infiniband/hw/ipath/ipath_registers.h @@ -63,67 +63,92 @@ /* kr_control bits */ #define INFINIPATH_C_FREEZEMODE 0x00000002 #define INFINIPATH_C_LINKENABLE 0x00000004 -#define INFINIPATH_C_RESET 0x00000001 /* kr_sendctrl bits */ #define INFINIPATH_S_DISARMPIOBUF_SHIFT 16 +#define INFINIPATH_S_UPDTHRESH_SHIFT 24 +#define INFINIPATH_S_UPDTHRESH_MASK 0x1f #define IPATH_S_ABORT 0 #define IPATH_S_PIOINTBUFAVAIL 1 #define IPATH_S_PIOBUFAVAILUPD 2 #define IPATH_S_PIOENABLE 3 +#define IPATH_S_SDMAINTENABLE 9 +#define IPATH_S_SDMASINGLEDESCRIPTOR 10 +#define IPATH_S_SDMAENABLE 11 +#define IPATH_S_SDMAHALT 12 #define IPATH_S_DISARM 31 #define INFINIPATH_S_ABORT (1U << IPATH_S_ABORT) #define INFINIPATH_S_PIOINTBUFAVAIL (1U << IPATH_S_PIOINTBUFAVAIL) #define INFINIPATH_S_PIOBUFAVAILUPD (1U << IPATH_S_PIOBUFAVAILUPD) #define INFINIPATH_S_PIOENABLE (1U << IPATH_S_PIOENABLE) +#define INFINIPATH_S_SDMAINTENABLE (1U << IPATH_S_SDMAINTENABLE) +#define INFINIPATH_S_SDMASINGLEDESCRIPTOR \ + (1U << IPATH_S_SDMASINGLEDESCRIPTOR) +#define INFINIPATH_S_SDMAENABLE (1U << IPATH_S_SDMAENABLE) +#define INFINIPATH_S_SDMAHALT (1U << IPATH_S_SDMAHALT) #define INFINIPATH_S_DISARM (1U << IPATH_S_DISARM) -/* kr_rcvctrl bits */ +/* kr_rcvctrl bits that are the same on multiple chips */ #define INFINIPATH_R_PORTENABLE_SHIFT 0 #define INFINIPATH_R_QPMAP_ENABLE (1ULL << 38) /* kr_intstatus, kr_intclear, kr_intmask bits */ -#define INFINIPATH_I_RCVURG_SHIFT 0 -#define INFINIPATH_I_RCVAVAIL_SHIFT 12 -#define INFINIPATH_I_ERROR 0x80000000 -#define INFINIPATH_I_SPIOSENT 0x40000000 -#define INFINIPATH_I_SPIOBUFAVAIL 0x20000000 -#define INFINIPATH_I_GPIO 0x10000000 +#define INFINIPATH_I_SDMAINT 0x8000000000000000ULL +#define INFINIPATH_I_SDMADISABLED 0x4000000000000000ULL +#define INFINIPATH_I_ERROR 0x0000000080000000ULL +#define INFINIPATH_I_SPIOSENT 0x0000000040000000ULL +#define INFINIPATH_I_SPIOBUFAVAIL 0x0000000020000000ULL +#define INFINIPATH_I_GPIO 0x0000000010000000ULL +#define INFINIPATH_I_JINT 0x0000000004000000ULL /* kr_errorstatus, kr_errorclear, kr_errormask bits */ -#define INFINIPATH_E_RFORMATERR 0x0000000000000001ULL -#define INFINIPATH_E_RVCRC 0x0000000000000002ULL -#define INFINIPATH_E_RICRC 0x0000000000000004ULL -#define INFINIPATH_E_RMINPKTLEN 0x0000000000000008ULL -#define INFINIPATH_E_RMAXPKTLEN 0x0000000000000010ULL -#define INFINIPATH_E_RLONGPKTLEN 0x0000000000000020ULL -#define INFINIPATH_E_RSHORTPKTLEN 0x0000000000000040ULL -#define INFINIPATH_E_RUNEXPCHAR 0x0000000000000080ULL -#define INFINIPATH_E_RUNSUPVL 0x0000000000000100ULL -#define INFINIPATH_E_REBP 0x0000000000000200ULL -#define INFINIPATH_E_RIBFLOW 0x0000000000000400ULL -#define INFINIPATH_E_RBADVERSION 0x0000000000000800ULL -#define INFINIPATH_E_RRCVEGRFULL 0x0000000000001000ULL -#define INFINIPATH_E_RRCVHDRFULL 0x0000000000002000ULL -#define INFINIPATH_E_RBADTID 0x0000000000004000ULL -#define INFINIPATH_E_RHDRLEN 0x0000000000008000ULL -#define INFINIPATH_E_RHDR 0x0000000000010000ULL -#define INFINIPATH_E_RIBLOSTLINK 0x0000000000020000ULL -#define INFINIPATH_E_SMINPKTLEN 0x0000000020000000ULL -#define INFINIPATH_E_SMAXPKTLEN 0x0000000040000000ULL -#define INFINIPATH_E_SUNDERRUN 0x0000000080000000ULL -#define INFINIPATH_E_SPKTLEN 0x0000000100000000ULL -#define INFINIPATH_E_SDROPPEDSMPPKT 0x0000000200000000ULL -#define INFINIPATH_E_SDROPPEDDATAPKT 0x0000000400000000ULL -#define INFINIPATH_E_SPIOARMLAUNCH 0x0000000800000000ULL -#define INFINIPATH_E_SUNEXPERRPKTNUM 0x0000001000000000ULL -#define INFINIPATH_E_SUNSUPVL 0x0000002000000000ULL -#define INFINIPATH_E_IBSTATUSCHANGED 0x0001000000000000ULL -#define INFINIPATH_E_INVALIDADDR 0x0002000000000000ULL -#define INFINIPATH_E_RESET 0x0004000000000000ULL -#define INFINIPATH_E_HARDWARE 0x0008000000000000ULL +#define INFINIPATH_E_RFORMATERR 0x0000000000000001ULL +#define INFINIPATH_E_RVCRC 0x0000000000000002ULL +#define INFINIPATH_E_RICRC 0x0000000000000004ULL +#define INFINIPATH_E_RMINPKTLEN 0x0000000000000008ULL +#define INFINIPATH_E_RMAXPKTLEN 0x0000000000000010ULL +#define INFINIPATH_E_RLONGPKTLEN 0x0000000000000020ULL +#define INFINIPATH_E_RSHORTPKTLEN 0x0000000000000040ULL +#define INFINIPATH_E_RUNEXPCHAR 0x0000000000000080ULL +#define INFINIPATH_E_RUNSUPVL 0x0000000000000100ULL +#define INFINIPATH_E_REBP 0x0000000000000200ULL +#define INFINIPATH_E_RIBFLOW 0x0000000000000400ULL +#define INFINIPATH_E_RBADVERSION 0x0000000000000800ULL +#define INFINIPATH_E_RRCVEGRFULL 0x0000000000001000ULL +#define INFINIPATH_E_RRCVHDRFULL 0x0000000000002000ULL +#define INFINIPATH_E_RBADTID 0x0000000000004000ULL +#define INFINIPATH_E_RHDRLEN 0x0000000000008000ULL +#define INFINIPATH_E_RHDR 0x0000000000010000ULL +#define INFINIPATH_E_RIBLOSTLINK 0x0000000000020000ULL +#define INFINIPATH_E_SENDSPECIALTRIGGER 0x0000000008000000ULL +#define INFINIPATH_E_SDMADISABLED 0x0000000010000000ULL +#define INFINIPATH_E_SMINPKTLEN 0x0000000020000000ULL +#define INFINIPATH_E_SMAXPKTLEN 0x0000000040000000ULL +#define INFINIPATH_E_SUNDERRUN 0x0000000080000000ULL +#define INFINIPATH_E_SPKTLEN 0x0000000100000000ULL +#define INFINIPATH_E_SDROPPEDSMPPKT 0x0000000200000000ULL +#define INFINIPATH_E_SDROPPEDDATAPKT 0x0000000400000000ULL +#define INFINIPATH_E_SPIOARMLAUNCH 0x0000000800000000ULL +#define INFINIPATH_E_SUNEXPERRPKTNUM 0x0000001000000000ULL +#define INFINIPATH_E_SUNSUPVL 0x0000002000000000ULL +#define INFINIPATH_E_SENDBUFMISUSE 0x0000004000000000ULL +#define INFINIPATH_E_SDMAGENMISMATCH 0x0000008000000000ULL +#define INFINIPATH_E_SDMAOUTOFBOUND 0x0000010000000000ULL +#define INFINIPATH_E_SDMATAILOUTOFBOUND 0x0000020000000000ULL +#define INFINIPATH_E_SDMABASE 0x0000040000000000ULL +#define INFINIPATH_E_SDMA1STDESC 0x0000080000000000ULL +#define INFINIPATH_E_SDMARPYTAG 0x0000100000000000ULL +#define INFINIPATH_E_SDMADWEN 0x0000200000000000ULL +#define INFINIPATH_E_SDMAMISSINGDW 0x0000400000000000ULL +#define INFINIPATH_E_SDMAUNEXPDATA 0x0000800000000000ULL +#define INFINIPATH_E_IBSTATUSCHANGED 0x0001000000000000ULL +#define INFINIPATH_E_INVALIDADDR 0x0002000000000000ULL +#define INFINIPATH_E_RESET 0x0004000000000000ULL +#define INFINIPATH_E_HARDWARE 0x0008000000000000ULL +#define INFINIPATH_E_SDMADESCADDRMISALIGN 0x0010000000000000ULL +#define INFINIPATH_E_INVALIDEEPCMD 0x0020000000000000ULL /* * this is used to print "common" packet errors only when the @@ -134,6 +159,17 @@ | INFINIPATH_E_RICRC | INFINIPATH_E_RSHORTPKTLEN \ | INFINIPATH_E_REBP ) +/* Convenience for decoding Send DMA errors */ +#define INFINIPATH_E_SDMAERRS ( \ + INFINIPATH_E_SDMAGENMISMATCH | INFINIPATH_E_SDMAOUTOFBOUND | \ + INFINIPATH_E_SDMATAILOUTOFBOUND | INFINIPATH_E_SDMABASE | \ + INFINIPATH_E_SDMA1STDESC | INFINIPATH_E_SDMARPYTAG | \ + INFINIPATH_E_SDMADWEN | INFINIPATH_E_SDMAMISSINGDW | \ + INFINIPATH_E_SDMAUNEXPDATA | \ + INFINIPATH_E_SDMADESCADDRMISALIGN | \ + INFINIPATH_E_SDMADISABLED | \ + INFINIPATH_E_SENDBUFMISUSE) + /* kr_hwerrclear, kr_hwerrmask, kr_hwerrstatus, bits */ /* TXEMEMPARITYERR bit 0: PIObuf, 1: PIOpbc, 2: launchfifo * RXEMEMPARITYERR bit 0: rcvbuf, 1: lookupq, 2: expTID, 3: eagerTID @@ -158,7 +194,7 @@ #define INFINIPATH_HWE_RXEMEMPARITYERR_HDRINFO 0x40ULL /* waldo specific -- find the rest in ipath_6110.c */ #define INFINIPATH_HWE_RXDSYNCMEMPARITYERR 0x0000000400000000ULL -/* monty specific -- find the rest in ipath_6120.c */ +/* 6120/7220 specific -- find the rest in ipath_6120.c and ipath_7220.c */ #define INFINIPATH_HWE_MEMBISTFAILED 0x0040000000000000ULL /* kr_hwdiagctrl bits */ @@ -185,8 +221,8 @@ #define INFINIPATH_IBCC_LINKINITCMD_SLEEP 3 #define INFINIPATH_IBCC_LINKINITCMD_SHIFT 16 #define INFINIPATH_IBCC_LINKCMD_MASK 0x3ULL -#define INFINIPATH_IBCC_LINKCMD_DOWN 1 /* move to 0x11 */ -#define INFINIPATH_IBCC_LINKCMD_ARMED 2 /* move to 0x21 */ +#define INFINIPATH_IBCC_LINKCMD_DOWN 1 /* move to 0x11 */ +#define INFINIPATH_IBCC_LINKCMD_ARMED 2 /* move to 0x21 */ #define INFINIPATH_IBCC_LINKCMD_ACTIVE 3 /* move to 0x31 */ #define INFINIPATH_IBCC_LINKCMD_SHIFT 18 #define INFINIPATH_IBCC_MAXPKTLEN_MASK 0x7FFULL @@ -201,10 +237,9 @@ #define INFINIPATH_IBCC_LINKDOWNDEFAULTSTATE 0x4000000000000000ULL /* kr_ibcstatus bits */ -#define INFINIPATH_IBCS_LINKTRAININGSTATE_MASK 0xF #define INFINIPATH_IBCS_LINKTRAININGSTATE_SHIFT 0 #define INFINIPATH_IBCS_LINKSTATE_MASK 0x7 -#define INFINIPATH_IBCS_LINKSTATE_SHIFT 4 + #define INFINIPATH_IBCS_TXREADY 0x40000000 #define INFINIPATH_IBCS_TXCREDITOK 0x80000000 /* link training states (shift by @@ -222,30 +257,13 @@ #define INFINIPATH_IBCS_LT_STATE_RECOVERRETRAIN 0x0c #define INFINIPATH_IBCS_LT_STATE_RECOVERWAITRMT 0x0e #define INFINIPATH_IBCS_LT_STATE_RECOVERIDLE 0x0f -/* link state machine states (shift by INFINIPATH_IBCS_LINKSTATE_SHIFT) */ +/* link state machine states (shift by ibcs_ls_shift) */ #define INFINIPATH_IBCS_L_STATE_DOWN 0x0 #define INFINIPATH_IBCS_L_STATE_INIT 0x1 #define INFINIPATH_IBCS_L_STATE_ARM 0x2 #define INFINIPATH_IBCS_L_STATE_ACTIVE 0x3 #define INFINIPATH_IBCS_L_STATE_ACT_DEFER 0x4 -/* combination link status states that we use with some frequency */ -#define IPATH_IBSTATE_MASK ((INFINIPATH_IBCS_LINKTRAININGSTATE_MASK \ - << INFINIPATH_IBCS_LINKTRAININGSTATE_SHIFT) | \ - (INFINIPATH_IBCS_LINKSTATE_MASK \ - <<INFINIPATH_IBCS_LINKSTATE_SHIFT)) -#define IPATH_IBSTATE_INIT ((INFINIPATH_IBCS_L_STATE_INIT \ - << INFINIPATH_IBCS_LINKSTATE_SHIFT) | \ - (INFINIPATH_IBCS_LT_STATE_LINKUP \ - <<INFINIPATH_IBCS_LINKTRAININGSTATE_SHIFT)) -#define IPATH_IBSTATE_ARM ((INFINIPATH_IBCS_L_STATE_ARM \ - << INFINIPATH_IBCS_LINKSTATE_SHIFT) | \ - (INFINIPATH_IBCS_LT_STATE_LINKUP \ - <<INFINIPATH_IBCS_LINKTRAININGSTATE_SHIFT)) -#define IPATH_IBSTATE_ACTIVE ((INFINIPATH_IBCS_L_STATE_ACTIVE \ - << INFINIPATH_IBCS_LINKSTATE_SHIFT) | \ - (INFINIPATH_IBCS_LT_STATE_LINKUP \ - <<INFINIPATH_IBCS_LINKTRAININGSTATE_SHIFT)) /* kr_extstatus bits */ #define INFINIPATH_EXTS_SERDESPLLLOCK 0x1 @@ -286,8 +304,7 @@ /* L1 Power down; use with RXDETECT, Otherwise not used on IB side */ #define INFINIPATH_SERDC0_L1PWR_DN 0xF0ULL -/* kr_xgxsconfig bits */ -#define INFINIPATH_XGXS_RESET 0x7ULL +/* common kr_xgxsconfig bits (or safe in all, even if not implemented) */ #define INFINIPATH_XGXS_RX_POL_SHIFT 19 #define INFINIPATH_XGXS_RX_POL_MASK 0xfULL @@ -417,6 +434,29 @@ struct ipath_kregs { ipath_kreg kr_pcieq1serdesconfig0; ipath_kreg kr_pcieq1serdesconfig1; ipath_kreg kr_pcieq1serdesstatus; + ipath_kreg kr_hrtbt_guid; + ipath_kreg kr_ibcddrctrl; + ipath_kreg kr_ibcddrstatus; + ipath_kreg kr_jintreload; + + /* send dma related regs */ + ipath_kreg kr_senddmabase; + ipath_kreg kr_senddmalengen; + ipath_kreg kr_senddmatail; + ipath_kreg kr_senddmahead; + ipath_kreg kr_senddmaheadaddr; + ipath_kreg kr_senddmabufmask0; + ipath_kreg kr_senddmabufmask1; + ipath_kreg kr_senddmabufmask2; + ipath_kreg kr_senddmastatus; + + /* SerDes related regs (IBA7220-only) */ + ipath_kreg kr_ibserdesctrl; + ipath_kreg kr_ib_epbacc; + ipath_kreg kr_ib_epbtrans; + ipath_kreg kr_pcie_epbacc; + ipath_kreg kr_pcie_epbtrans; + ipath_kreg kr_ib_ddsrxeq; }; struct ipath_cregs { diff --git a/drivers/infiniband/hw/ipath/ipath_ruc.c b/drivers/infiniband/hw/ipath/ipath_ruc.c index a59bdbd0ed87..8ac5c1d82ccd 100644 --- a/drivers/infiniband/hw/ipath/ipath_ruc.c +++ b/drivers/infiniband/hw/ipath/ipath_ruc.c @@ -310,7 +310,7 @@ again: switch (wqe->wr.opcode) { case IB_WR_SEND_WITH_IMM: wc.wc_flags = IB_WC_WITH_IMM; - wc.imm_data = wqe->wr.imm_data; + wc.imm_data = wqe->wr.ex.imm_data; /* FALLTHROUGH */ case IB_WR_SEND: if (!ipath_get_rwqe(qp, 0)) { @@ -339,7 +339,7 @@ again: goto err; } wc.wc_flags = IB_WC_WITH_IMM; - wc.imm_data = wqe->wr.imm_data; + wc.imm_data = wqe->wr.ex.imm_data; if (!ipath_get_rwqe(qp, 1)) goto rnr_nak; /* FALLTHROUGH */ @@ -483,14 +483,16 @@ done: static void want_buffer(struct ipath_devdata *dd) { - unsigned long flags; - - spin_lock_irqsave(&dd->ipath_sendctrl_lock, flags); - dd->ipath_sendctrl |= INFINIPATH_S_PIOINTBUFAVAIL; - ipath_write_kreg(dd, dd->ipath_kregs->kr_sendctrl, - dd->ipath_sendctrl); - ipath_read_kreg64(dd, dd->ipath_kregs->kr_scratch); - spin_unlock_irqrestore(&dd->ipath_sendctrl_lock, flags); + if (!(dd->ipath_flags & IPATH_HAS_SEND_DMA)) { + unsigned long flags; + + spin_lock_irqsave(&dd->ipath_sendctrl_lock, flags); + dd->ipath_sendctrl |= INFINIPATH_S_PIOINTBUFAVAIL; + ipath_write_kreg(dd, dd->ipath_kregs->kr_sendctrl, + dd->ipath_sendctrl); + ipath_read_kreg64(dd, dd->ipath_kregs->kr_scratch); + spin_unlock_irqrestore(&dd->ipath_sendctrl_lock, flags); + } } /** diff --git a/drivers/infiniband/hw/ipath/ipath_sd7220.c b/drivers/infiniband/hw/ipath/ipath_sd7220.c new file mode 100644 index 000000000000..aa47eb549520 --- /dev/null +++ b/drivers/infiniband/hw/ipath/ipath_sd7220.c @@ -0,0 +1,1462 @@ +/* + * Copyright (c) 2006, 2007, 2008 QLogic Corporation. All rights reserved. + * Copyright (c) 2003, 2004, 2005, 2006 PathScale, Inc. All rights reserved. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * OpenIB.org BSD license below: + * + * Redistribution and use in source and binary forms, with or + * without modification, are permitted provided that the following + * conditions are met: + * + * - Redistributions of source code must retain the above + * copyright notice, this list of conditions and the following + * disclaimer. + * + * - Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials + * provided with the distribution. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + */ +/* + * This file contains all of the code that is specific to the SerDes + * on the InfiniPath 7220 chip. + */ + +#include <linux/pci.h> +#include <linux/delay.h> + +#include "ipath_kernel.h" +#include "ipath_registers.h" +#include "ipath_7220.h" + +/* + * The IBSerDesMappTable is a memory that holds values to be stored in + * various SerDes registers by IBC. It is not part of the normal kregs + * map and is used in exactly one place, hence the #define below. + */ +#define KR_IBSerDesMappTable (0x94000 / (sizeof(uint64_t))) + +/* + * Below used for sdnum parameter, selecting one of the two sections + * used for PCIe, or the single SerDes used for IB. + */ +#define PCIE_SERDES0 0 +#define PCIE_SERDES1 1 + +/* + * The EPB requires addressing in a particular form. EPB_LOC() is intended + * to make #definitions a little more readable. + */ +#define EPB_ADDR_SHF 8 +#define EPB_LOC(chn, elt, reg) \ + (((elt & 0xf) | ((chn & 7) << 4) | ((reg & 0x3f) << 9)) << \ + EPB_ADDR_SHF) +#define EPB_IB_QUAD0_CS_SHF (25) +#define EPB_IB_QUAD0_CS (1U << EPB_IB_QUAD0_CS_SHF) +#define EPB_IB_UC_CS_SHF (26) +#define EPB_PCIE_UC_CS_SHF (27) +#define EPB_GLOBAL_WR (1U << (EPB_ADDR_SHF + 8)) + +/* Forward declarations. */ +static int ipath_sd7220_reg_mod(struct ipath_devdata *dd, int sdnum, u32 loc, + u32 data, u32 mask); +static int ibsd_mod_allchnls(struct ipath_devdata *dd, int loc, int val, + int mask); +static int ipath_sd_trimdone_poll(struct ipath_devdata *dd); +static void ipath_sd_trimdone_monitor(struct ipath_devdata *dd, + const char *where); +static int ipath_sd_setvals(struct ipath_devdata *dd); +static int ipath_sd_early(struct ipath_devdata *dd); +static int ipath_sd_dactrim(struct ipath_devdata *dd); +/* Set the registers that IBC may muck with to their default "preset" values */ +int ipath_sd7220_presets(struct ipath_devdata *dd); +static int ipath_internal_presets(struct ipath_devdata *dd); +/* Tweak the register (CMUCTRL5) that contains the TRIMSELF controls */ +static int ipath_sd_trimself(struct ipath_devdata *dd, int val); +static int epb_access(struct ipath_devdata *dd, int sdnum, int claim); + +void ipath_set_relock_poll(struct ipath_devdata *dd, int ibup); + +/* + * Below keeps track of whether the "once per power-on" initialization has + * been done, because uC code Version 1.32.17 or higher allows the uC to + * be reset at will, and Automatic Equalization may require it. So the + * state of the reset "pin", as reflected in was_reset parameter to + * ipath_sd7220_init() is no longer valid. Instead, we check for the + * actual uC code having been loaded. + */ +static int ipath_ibsd_ucode_loaded(struct ipath_devdata *dd) +{ + if (!dd->serdes_first_init_done && (ipath_sd7220_ib_vfy(dd) > 0)) + dd->serdes_first_init_done = 1; + return dd->serdes_first_init_done; +} + +/* repeat #define for local use. "Real" #define is in ipath_iba7220.c */ +#define INFINIPATH_HWE_IB_UC_MEMORYPARITYERR 0x0000004000000000ULL +#define IB_MPREG5 (EPB_LOC(6, 0, 0xE) | (1L << EPB_IB_UC_CS_SHF)) +#define IB_MPREG6 (EPB_LOC(6, 0, 0xF) | (1U << EPB_IB_UC_CS_SHF)) +#define UC_PAR_CLR_D 8 +#define UC_PAR_CLR_M 0xC +#define IB_CTRL2(chn) (EPB_LOC(chn, 7, 3) | EPB_IB_QUAD0_CS) +#define START_EQ1(chan) EPB_LOC(chan, 7, 0x27) + +void ipath_sd7220_clr_ibpar(struct ipath_devdata *dd) +{ + int ret; + + /* clear, then re-enable parity errs */ + ret = ipath_sd7220_reg_mod(dd, IB_7220_SERDES, IB_MPREG6, + UC_PAR_CLR_D, UC_PAR_CLR_M); + if (ret < 0) { + ipath_dev_err(dd, "Failed clearing IBSerDes Parity err\n"); + goto bail; + } + ret = ipath_sd7220_reg_mod(dd, IB_7220_SERDES, IB_MPREG6, 0, + UC_PAR_CLR_M); + + ipath_read_kreg32(dd, dd->ipath_kregs->kr_scratch); + udelay(4); + ipath_write_kreg(dd, dd->ipath_kregs->kr_hwerrclear, + INFINIPATH_HWE_IB_UC_MEMORYPARITYERR); + ipath_read_kreg32(dd, dd->ipath_kregs->kr_scratch); +bail: + return; +} + +/* + * After a reset or other unusual event, the epb interface may need + * to be re-synchronized, between the host and the uC. + * returns <0 for failure to resync within IBSD_RESYNC_TRIES (not expected) + */ +#define IBSD_RESYNC_TRIES 3 +#define IB_PGUDP(chn) (EPB_LOC((chn), 2, 1) | EPB_IB_QUAD0_CS) +#define IB_CMUDONE(chn) (EPB_LOC((chn), 7, 0xF) | EPB_IB_QUAD0_CS) + +static int ipath_resync_ibepb(struct ipath_devdata *dd) +{ + int ret, pat, tries, chn; + u32 loc; + + ret = -1; + chn = 0; + for (tries = 0; tries < (4 * IBSD_RESYNC_TRIES); ++tries) { + loc = IB_PGUDP(chn); + ret = ipath_sd7220_reg_mod(dd, IB_7220_SERDES, loc, 0, 0); + if (ret < 0) { + ipath_dev_err(dd, "Failed read in resync\n"); + continue; + } + if (ret != 0xF0 && ret != 0x55 && tries == 0) + ipath_dev_err(dd, "unexpected pattern in resync\n"); + pat = ret ^ 0xA5; /* alternate F0 and 55 */ + ret = ipath_sd7220_reg_mod(dd, IB_7220_SERDES, loc, pat, 0xFF); + if (ret < 0) { + ipath_dev_err(dd, "Failed write in resync\n"); + continue; + } + ret = ipath_sd7220_reg_mod(dd, IB_7220_SERDES, loc, 0, 0); + if (ret < 0) { + ipath_dev_err(dd, "Failed re-read in resync\n"); + continue; + } + if (ret != pat) { + ipath_dev_err(dd, "Failed compare1 in resync\n"); + continue; + } + loc = IB_CMUDONE(chn); + ret = ipath_sd7220_reg_mod(dd, IB_7220_SERDES, loc, 0, 0); + if (ret < 0) { + ipath_dev_err(dd, "Failed CMUDONE rd in resync\n"); + continue; + } + if ((ret & 0x70) != ((chn << 4) | 0x40)) { + ipath_dev_err(dd, "Bad CMUDONE value %02X, chn %d\n", + ret, chn); + continue; + } + if (++chn == 4) + break; /* Success */ + } + ipath_cdbg(VERBOSE, "Resync in %d tries\n", tries); + return (ret > 0) ? 0 : ret; +} + +/* + * Localize the stuff that should be done to change IB uC reset + * returns <0 for errors. + */ +static int ipath_ibsd_reset(struct ipath_devdata *dd, int assert_rst) +{ + u64 rst_val; + int ret = 0; + unsigned long flags; + + rst_val = ipath_read_kreg64(dd, dd->ipath_kregs->kr_ibserdesctrl); + if (assert_rst) { + /* + * Vendor recommends "interrupting" uC before reset, to + * minimize possible glitches. + */ + spin_lock_irqsave(&dd->ipath_sdepb_lock, flags); + epb_access(dd, IB_7220_SERDES, 1); + rst_val |= 1ULL; + /* Squelch possible parity error from _asserting_ reset */ + ipath_write_kreg(dd, dd->ipath_kregs->kr_hwerrmask, + dd->ipath_hwerrmask & + ~INFINIPATH_HWE_IB_UC_MEMORYPARITYERR); + ipath_write_kreg(dd, dd->ipath_kregs->kr_ibserdesctrl, rst_val); + /* flush write, delay to ensure it took effect */ + ipath_read_kreg32(dd, dd->ipath_kregs->kr_scratch); + udelay(2); + /* once it's reset, can remove interrupt */ + epb_access(dd, IB_7220_SERDES, -1); + spin_unlock_irqrestore(&dd->ipath_sdepb_lock, flags); + } else { + /* + * Before we de-assert reset, we need to deal with + * possible glitch on the Parity-error line. + * Suppress it around the reset, both in chip-level + * hwerrmask and in IB uC control reg. uC will allow + * it again during startup. + */ + u64 val; + rst_val &= ~(1ULL); + ipath_write_kreg(dd, dd->ipath_kregs->kr_hwerrmask, + dd->ipath_hwerrmask & + ~INFINIPATH_HWE_IB_UC_MEMORYPARITYERR); + + ret = ipath_resync_ibepb(dd); + if (ret < 0) + ipath_dev_err(dd, "unable to re-sync IB EPB\n"); + + /* set uC control regs to suppress parity errs */ + ret = ipath_sd7220_reg_mod(dd, IB_7220_SERDES, IB_MPREG5, 1, 1); + if (ret < 0) + goto bail; + /* IB uC code past Version 1.32.17 allow suppression of wdog */ + ret = ipath_sd7220_reg_mod(dd, IB_7220_SERDES, IB_MPREG6, 0x80, + 0x80); + if (ret < 0) { + ipath_dev_err(dd, "Failed to set WDOG disable\n"); + goto bail; + } + ipath_write_kreg(dd, dd->ipath_kregs->kr_ibserdesctrl, rst_val); + /* flush write, delay for startup */ + ipath_read_kreg32(dd, dd->ipath_kregs->kr_scratch); + udelay(1); + /* clear, then re-enable parity errs */ + ipath_sd7220_clr_ibpar(dd); + val = ipath_read_kreg64(dd, dd->ipath_kregs->kr_hwerrstatus); + if (val & INFINIPATH_HWE_IB_UC_MEMORYPARITYERR) { + ipath_dev_err(dd, "IBUC Parity still set after RST\n"); + dd->ipath_hwerrmask &= + ~INFINIPATH_HWE_IB_UC_MEMORYPARITYERR; + } + ipath_write_kreg(dd, dd->ipath_kregs->kr_hwerrmask, + dd->ipath_hwerrmask); + } + +bail: + return ret; +} + +static void ipath_sd_trimdone_monitor(struct ipath_devdata *dd, + const char *where) +{ + int ret, chn, baduns; + u64 val; + + if (!where) + where = "?"; + + /* give time for reset to settle out in EPB */ + udelay(2); + + ret = ipath_resync_ibepb(dd); + if (ret < 0) + ipath_dev_err(dd, "not able to re-sync IB EPB (%s)\n", where); + + /* Do "sacrificial read" to get EPB in sane state after reset */ + ret = ipath_sd7220_reg_mod(dd, IB_7220_SERDES, IB_CTRL2(0), 0, 0); + if (ret < 0) + ipath_dev_err(dd, "Failed TRIMDONE 1st read, (%s)\n", where); + + /* Check/show "summary" Trim-done bit in IBCStatus */ + val = ipath_read_kreg64(dd, dd->ipath_kregs->kr_ibcstatus); + if (val & (1ULL << 11)) + ipath_cdbg(VERBOSE, "IBCS TRIMDONE set (%s)\n", where); + else + ipath_dev_err(dd, "IBCS TRIMDONE clear (%s)\n", where); + + udelay(2); + + ret = ipath_sd7220_reg_mod(dd, IB_7220_SERDES, IB_MPREG6, 0x80, 0x80); + if (ret < 0) + ipath_dev_err(dd, "Failed Dummy RMW, (%s)\n", where); + udelay(10); + + baduns = 0; + + for (chn = 3; chn >= 0; --chn) { + /* Read CTRL reg for each channel to check TRIMDONE */ + ret = ipath_sd7220_reg_mod(dd, IB_7220_SERDES, + IB_CTRL2(chn), 0, 0); + if (ret < 0) + ipath_dev_err(dd, "Failed checking TRIMDONE, chn %d" + " (%s)\n", chn, where); + + if (!(ret & 0x10)) { + int probe; + baduns |= (1 << chn); + ipath_dev_err(dd, "TRIMDONE cleared on chn %d (%02X)." + " (%s)\n", chn, ret, where); + probe = ipath_sd7220_reg_mod(dd, IB_7220_SERDES, + IB_PGUDP(0), 0, 0); + ipath_dev_err(dd, "probe is %d (%02X)\n", + probe, probe); + probe = ipath_sd7220_reg_mod(dd, IB_7220_SERDES, + IB_CTRL2(chn), 0, 0); + ipath_dev_err(dd, "re-read: %d (%02X)\n", + probe, probe); + ret = ipath_sd7220_reg_mod(dd, IB_7220_SERDES, + IB_CTRL2(chn), 0x10, 0x10); + if (ret < 0) + ipath_dev_err(dd, + "Err on TRIMDONE rewrite1\n"); + } + } + for (chn = 3; chn >= 0; --chn) { + /* Read CTRL reg for each channel to check TRIMDONE */ + if (baduns & (1 << chn)) { + ipath_dev_err(dd, + "Reseting TRIMDONE on chn %d (%s)\n", + chn, where); + ret = ipath_sd7220_reg_mod(dd, IB_7220_SERDES, + IB_CTRL2(chn), 0x10, 0x10); + if (ret < 0) + ipath_dev_err(dd, "Failed re-setting " + "TRIMDONE, chn %d (%s)\n", + chn, where); + } + } +} + +/* + * Below is portion of IBA7220-specific bringup_serdes() that actually + * deals with registers and memory within the SerDes itself. + * Post IB uC code version 1.32.17, was_reset being 1 is not really + * informative, so we double-check. + */ +int ipath_sd7220_init(struct ipath_devdata *dd, int was_reset) +{ + int ret = 1; /* default to failure */ + int first_reset; + int val_stat; + + if (!was_reset) { + /* entered with reset not asserted, we need to do it */ + ipath_ibsd_reset(dd, 1); + ipath_sd_trimdone_monitor(dd, "Driver-reload"); + } + + /* Substitute our deduced value for was_reset */ + ret = ipath_ibsd_ucode_loaded(dd); + if (ret < 0) { + ret = 1; + goto done; + } + first_reset = !ret; /* First reset if IBSD uCode not yet loaded */ + + /* + * Alter some regs per vendor latest doc, reset-defaults + * are not right for IB. + */ + ret = ipath_sd_early(dd); + if (ret < 0) { + ipath_dev_err(dd, "Failed to set IB SERDES early defaults\n"); + ret = 1; + goto done; + } + + /* + * Set DAC manual trim IB. + * We only do this once after chip has been reset (usually + * same as once per system boot). + */ + if (first_reset) { + ret = ipath_sd_dactrim(dd); + if (ret < 0) { + ipath_dev_err(dd, "Failed IB SERDES DAC trim\n"); + ret = 1; + goto done; + } + } + + /* + * Set various registers (DDS and RXEQ) that will be + * controlled by IBC (in 1.2 mode) to reasonable preset values + * Calling the "internal" version avoids the "check for needed" + * and "trimdone monitor" that might be counter-productive. + */ + ret = ipath_internal_presets(dd); + if (ret < 0) { + ipath_dev_err(dd, "Failed to set IB SERDES presets\n"); + ret = 1; + goto done; + } + ret = ipath_sd_trimself(dd, 0x80); + if (ret < 0) { + ipath_dev_err(dd, "Failed to set IB SERDES TRIMSELF\n"); + ret = 1; + goto done; + } + + /* Load image, then try to verify */ + ret = 0; /* Assume success */ + if (first_reset) { + int vfy; + int trim_done; + ipath_dbg("SerDes uC was reset, reloading PRAM\n"); + ret = ipath_sd7220_ib_load(dd); + if (ret < 0) { + ipath_dev_err(dd, "Failed to load IB SERDES image\n"); + ret = 1; + goto done; + } + + /* Loaded image, try to verify */ + vfy = ipath_sd7220_ib_vfy(dd); + if (vfy != ret) { + ipath_dev_err(dd, "SERDES PRAM VFY failed\n"); + ret = 1; + goto done; + } + /* + * Loaded and verified. Almost good... + * hold "success" in ret + */ + ret = 0; + + /* + * Prev steps all worked, continue bringup + * De-assert RESET to uC, only in first reset, to allow + * trimming. + * + * Since our default setup sets START_EQ1 to + * PRESET, we need to clear that for this very first run. + */ + ret = ibsd_mod_allchnls(dd, START_EQ1(0), 0, 0x38); + if (ret < 0) { + ipath_dev_err(dd, "Failed clearing START_EQ1\n"); + ret = 1; + goto done; + } + + ipath_ibsd_reset(dd, 0); + /* + * If this is not the first reset, trimdone should be set + * already. + */ + trim_done = ipath_sd_trimdone_poll(dd); + /* + * Whether or not trimdone succeeded, we need to put the + * uC back into reset to avoid a possible fight with the + * IBC state-machine. + */ + ipath_ibsd_reset(dd, 1); + + if (!trim_done) { + ipath_dev_err(dd, "No TRIMDONE seen\n"); + ret = 1; + goto done; + } + + ipath_sd_trimdone_monitor(dd, "First-reset"); + /* Remember so we do not re-do the load, dactrim, etc. */ + dd->serdes_first_init_done = 1; + } + /* + * Setup for channel training and load values for + * RxEq and DDS in tables used by IBC in IB1.2 mode + */ + + val_stat = ipath_sd_setvals(dd); + if (val_stat < 0) + ret = 1; +done: + /* start relock timer regardless, but start at 1 second */ + ipath_set_relock_poll(dd, -1); + return ret; +} + +#define EPB_ACC_REQ 1 +#define EPB_ACC_GNT 0x100 +#define EPB_DATA_MASK 0xFF +#define EPB_RD (1ULL << 24) +#define EPB_TRANS_RDY (1ULL << 31) +#define EPB_TRANS_ERR (1ULL << 30) +#define EPB_TRANS_TRIES 5 + +/* + * query, claim, release ownership of the EPB (External Parallel Bus) + * for a specified SERDES. + * the "claim" parameter is >0 to claim, <0 to release, 0 to query. + * Returns <0 for errors, >0 if we had ownership, else 0. + */ +static int epb_access(struct ipath_devdata *dd, int sdnum, int claim) +{ + u16 acc; + u64 accval; + int owned = 0; + u64 oct_sel = 0; + + switch (sdnum) { + case IB_7220_SERDES : + /* + * The IB SERDES "ownership" is fairly simple. A single each + * request/grant. + */ + acc = dd->ipath_kregs->kr_ib_epbacc; + break; + case PCIE_SERDES0 : + case PCIE_SERDES1 : + /* PCIe SERDES has two "octants", need to select which */ + acc = dd->ipath_kregs->kr_pcie_epbacc; + oct_sel = (2 << (sdnum - PCIE_SERDES0)); + break; + default : + return 0; + } + + /* Make sure any outstanding transaction was seen */ + ipath_read_kreg32(dd, dd->ipath_kregs->kr_scratch); + udelay(15); + + accval = ipath_read_kreg32(dd, acc); + + owned = !!(accval & EPB_ACC_GNT); + if (claim < 0) { + /* Need to release */ + u64 pollval; + /* + * The only writeable bits are the request and CS. + * Both should be clear + */ + u64 newval = 0; + ipath_write_kreg(dd, acc, newval); + /* First read after write is not trustworthy */ + pollval = ipath_read_kreg32(dd, acc); + udelay(5); + pollval = ipath_read_kreg32(dd, acc); + if (pollval & EPB_ACC_GNT) + owned = -1; + } else if (claim > 0) { + /* Need to claim */ + u64 pollval; + u64 newval = EPB_ACC_REQ | oct_sel; + ipath_write_kreg(dd, acc, newval); + /* First read after write is not trustworthy */ + pollval = ipath_read_kreg32(dd, acc); + udelay(5); + pollval = ipath_read_kreg32(dd, acc); + if (!(pollval & EPB_ACC_GNT)) + owned = -1; + } + return owned; +} + +/* + * Lemma to deal with race condition of write..read to epb regs + */ +static int epb_trans(struct ipath_devdata *dd, u16 reg, u64 i_val, u64 *o_vp) +{ + int tries; + u64 transval; + + + ipath_write_kreg(dd, reg, i_val); + /* Throw away first read, as RDY bit may be stale */ + transval = ipath_read_kreg64(dd, reg); + + for (tries = EPB_TRANS_TRIES; tries; --tries) { + transval = ipath_read_kreg32(dd, reg); + if (transval & EPB_TRANS_RDY) + break; + udelay(5); + } + if (transval & EPB_TRANS_ERR) + return -1; + if (tries > 0 && o_vp) + *o_vp = transval; + return tries; +} + +/** + * + * ipath_sd7220_reg_mod - modify SERDES register + * @dd: the infinipath device + * @sdnum: which SERDES to access + * @loc: location - channel, element, register, as packed by EPB_LOC() macro. + * @wd: Write Data - value to set in register + * @mask: ones where data should be spliced into reg. + * + * Basic register read/modify/write, with un-needed acesses elided. That is, + * a mask of zero will prevent write, while a mask of 0xFF will prevent read. + * returns current (presumed, if a write was done) contents of selected + * register, or <0 if errors. + */ +static int ipath_sd7220_reg_mod(struct ipath_devdata *dd, int sdnum, u32 loc, + u32 wd, u32 mask) +{ + u16 trans; + u64 transval; + int owned; + int tries, ret; + unsigned long flags; + + switch (sdnum) { + case IB_7220_SERDES : + trans = dd->ipath_kregs->kr_ib_epbtrans; + break; + case PCIE_SERDES0 : + case PCIE_SERDES1 : + trans = dd->ipath_kregs->kr_pcie_epbtrans; + break; + default : + return -1; + } + + /* + * All access is locked in software (vs other host threads) and + * hardware (vs uC access). + */ + spin_lock_irqsave(&dd->ipath_sdepb_lock, flags); + + owned = epb_access(dd, sdnum, 1); + if (owned < 0) { + spin_unlock_irqrestore(&dd->ipath_sdepb_lock, flags); + return -1; + } + ret = 0; + for (tries = EPB_TRANS_TRIES; tries; --tries) { + transval = ipath_read_kreg32(dd, trans); + if (transval & EPB_TRANS_RDY) + break; + udelay(5); + } + + if (tries > 0) { + tries = 1; /* to make read-skip work */ + if (mask != 0xFF) { + /* + * Not a pure write, so need to read. + * loc encodes chip-select as well as address + */ + transval = loc | EPB_RD; + tries = epb_trans(dd, trans, transval, &transval); + } + if (tries > 0 && mask != 0) { + /* + * Not a pure read, so need to write. + */ + wd = (wd & mask) | (transval & ~mask); + transval = loc | (wd & EPB_DATA_MASK); + tries = epb_trans(dd, trans, transval, &transval); + } + } + /* else, failed to see ready, what error-handling? */ + + /* + * Release bus. Failure is an error. + */ + if (epb_access(dd, sdnum, -1) < 0) + ret = -1; + else + ret = transval & EPB_DATA_MASK; + + spin_unlock_irqrestore(&dd->ipath_sdepb_lock, flags); + if (tries <= 0) + ret = -1; + return ret; +} + +#define EPB_ROM_R (2) +#define EPB_ROM_W (1) +/* + * Below, all uC-related, use appropriate UC_CS, depending + * on which SerDes is used. + */ +#define EPB_UC_CTL EPB_LOC(6, 0, 0) +#define EPB_MADDRL EPB_LOC(6, 0, 2) +#define EPB_MADDRH EPB_LOC(6, 0, 3) +#define EPB_ROMDATA EPB_LOC(6, 0, 4) +#define EPB_RAMDATA EPB_LOC(6, 0, 5) + +/* Transfer date to/from uC Program RAM of IB or PCIe SerDes */ +static int ipath_sd7220_ram_xfer(struct ipath_devdata *dd, int sdnum, u32 loc, + u8 *buf, int cnt, int rd_notwr) +{ + u16 trans; + u64 transval; + u64 csbit; + int owned; + int tries; + int sofar; + int addr; + int ret; + unsigned long flags; + const char *op; + + /* Pick appropriate transaction reg and "Chip select" for this serdes */ + switch (sdnum) { + case IB_7220_SERDES : + csbit = 1ULL << EPB_IB_UC_CS_SHF; + trans = dd->ipath_kregs->kr_ib_epbtrans; + break; + case PCIE_SERDES0 : + case PCIE_SERDES1 : + /* PCIe SERDES has uC "chip select" in different bit, too */ + csbit = 1ULL << EPB_PCIE_UC_CS_SHF; + trans = dd->ipath_kregs->kr_pcie_epbtrans; + break; + default : + return -1; + } + + op = rd_notwr ? "Rd" : "Wr"; + spin_lock_irqsave(&dd->ipath_sdepb_lock, flags); + + owned = epb_access(dd, sdnum, 1); + if (owned < 0) { + spin_unlock_irqrestore(&dd->ipath_sdepb_lock, flags); + ipath_dbg("Could not get %s access to %s EPB: %X, loc %X\n", + op, (sdnum == IB_7220_SERDES) ? "IB" : "PCIe", + owned, loc); + return -1; + } + + /* + * In future code, we may need to distinguish several address ranges, + * and select various memories based on this. For now, just trim + * "loc" (location including address and memory select) to + * "addr" (address within memory). we will only support PRAM + * The memory is 8KB. + */ + addr = loc & 0x1FFF; + for (tries = EPB_TRANS_TRIES; tries; --tries) { + transval = ipath_read_kreg32(dd, trans); + if (transval & EPB_TRANS_RDY) + break; + udelay(5); + } + + sofar = 0; + if (tries <= 0) + ipath_dbg("No initial RDY on EPB access request\n"); + else { + /* + * Every "memory" access is doubly-indirect. + * We set two bytes of address, then read/write + * one or mores bytes of data. + */ + + /* First, we set control to "Read" or "Write" */ + transval = csbit | EPB_UC_CTL | + (rd_notwr ? EPB_ROM_R : EPB_ROM_W); + tries = epb_trans(dd, trans, transval, &transval); + if (tries <= 0) + ipath_dbg("No EPB response to uC %s cmd\n", op); + while (tries > 0 && sofar < cnt) { + if (!sofar) { + /* Only set address at start of chunk */ + int addrbyte = (addr + sofar) >> 8; + transval = csbit | EPB_MADDRH | addrbyte; + tries = epb_trans(dd, trans, transval, + &transval); + if (tries <= 0) { + ipath_dbg("No EPB response ADDRH\n"); + break; + } + addrbyte = (addr + sofar) & 0xFF; + transval = csbit | EPB_MADDRL | addrbyte; + tries = epb_trans(dd, trans, transval, + &transval); + if (tries <= 0) { + ipath_dbg("No EPB response ADDRL\n"); + break; + } + } + + if (rd_notwr) + transval = csbit | EPB_ROMDATA | EPB_RD; + else + transval = csbit | EPB_ROMDATA | buf[sofar]; + tries = epb_trans(dd, trans, transval, &transval); + if (tries <= 0) { + ipath_dbg("No EPB response DATA\n"); + break; + } + if (rd_notwr) + buf[sofar] = transval & EPB_DATA_MASK; + ++sofar; + } + /* Finally, clear control-bit for Read or Write */ + transval = csbit | EPB_UC_CTL; + tries = epb_trans(dd, trans, transval, &transval); + if (tries <= 0) + ipath_dbg("No EPB response to drop of uC %s cmd\n", op); + } + + ret = sofar; + /* Release bus. Failure is an error */ + if (epb_access(dd, sdnum, -1) < 0) + ret = -1; + + spin_unlock_irqrestore(&dd->ipath_sdepb_lock, flags); + if (tries <= 0) { + ipath_dbg("SERDES PRAM %s failed after %d bytes\n", op, sofar); + ret = -1; + } + return ret; +} + +#define PROG_CHUNK 64 + +int ipath_sd7220_prog_ld(struct ipath_devdata *dd, int sdnum, + u8 *img, int len, int offset) +{ + int cnt, sofar, req; + + sofar = 0; + while (sofar < len) { + req = len - sofar; + if (req > PROG_CHUNK) + req = PROG_CHUNK; + cnt = ipath_sd7220_ram_xfer(dd, sdnum, offset + sofar, + img + sofar, req, 0); + if (cnt < req) { + sofar = -1; + break; + } + sofar += req; + } + return sofar; +} + +#define VFY_CHUNK 64 +#define SD_PRAM_ERROR_LIMIT 42 + +int ipath_sd7220_prog_vfy(struct ipath_devdata *dd, int sdnum, + const u8 *img, int len, int offset) +{ + int cnt, sofar, req, idx, errors; + unsigned char readback[VFY_CHUNK]; + + errors = 0; + sofar = 0; + while (sofar < len) { + req = len - sofar; + if (req > VFY_CHUNK) + req = VFY_CHUNK; + cnt = ipath_sd7220_ram_xfer(dd, sdnum, sofar + offset, + readback, req, 1); + if (cnt < req) { + /* failed in read itself */ + sofar = -1; + break; + } + for (idx = 0; idx < cnt; ++idx) { + if (readback[idx] != img[idx+sofar]) + ++errors; + } + sofar += cnt; + } + return errors ? -errors : sofar; +} + +/* IRQ not set up at this point in init, so we poll. */ +#define IB_SERDES_TRIM_DONE (1ULL << 11) +#define TRIM_TMO (30) + +static int ipath_sd_trimdone_poll(struct ipath_devdata *dd) +{ + int trim_tmo, ret; + uint64_t val; + + /* + * Default to failure, so IBC will not start + * without IB_SERDES_TRIM_DONE. + */ + ret = 0; + for (trim_tmo = 0; trim_tmo < TRIM_TMO; ++trim_tmo) { + val = ipath_read_kreg64(dd, dd->ipath_kregs->kr_ibcstatus); + if (val & IB_SERDES_TRIM_DONE) { + ipath_cdbg(VERBOSE, "TRIMDONE after %d\n", trim_tmo); + ret = 1; + break; + } + msleep(10); + } + if (trim_tmo >= TRIM_TMO) { + ipath_dev_err(dd, "No TRIMDONE in %d tries\n", trim_tmo); + ret = 0; + } + return ret; +} + +#define TX_FAST_ELT (9) + +/* + * Set the "negotiation" values for SERDES. These are used by the IB1.2 + * link negotiation. Macros below are attempt to keep the values a + * little more human-editable. + * First, values related to Drive De-emphasis Settings. + */ + +#define NUM_DDS_REGS 6 +#define DDS_REG_MAP 0x76A910 /* LSB-first list of regs (in elt 9) to mod */ + +#define DDS_VAL(amp_d, main_d, ipst_d, ipre_d, amp_s, main_s, ipst_s, ipre_s) \ + { { ((amp_d & 0x1F) << 1) | 1, ((amp_s & 0x1F) << 1) | 1, \ + (main_d << 3) | 4 | (ipre_d >> 2), \ + (main_s << 3) | 4 | (ipre_s >> 2), \ + ((ipst_d & 0xF) << 1) | ((ipre_d & 3) << 6) | 0x21, \ + ((ipst_s & 0xF) << 1) | ((ipre_s & 3) << 6) | 0x21 } } + +static struct dds_init { + uint8_t reg_vals[NUM_DDS_REGS]; +} dds_init_vals[] = { + /* DDR(FDR) SDR(HDR) */ + /* Vendor recommends below for 3m cable */ +#define DDS_3M 0 + DDS_VAL(31, 19, 12, 0, 29, 22, 9, 0), + DDS_VAL(31, 12, 15, 4, 31, 15, 15, 1), + DDS_VAL(31, 13, 15, 3, 31, 16, 15, 0), + DDS_VAL(31, 14, 15, 2, 31, 17, 14, 0), + DDS_VAL(31, 15, 15, 1, 31, 18, 13, 0), + DDS_VAL(31, 16, 15, 0, 31, 19, 12, 0), + DDS_VAL(31, 17, 14, 0, 31, 20, 11, 0), + DDS_VAL(31, 18, 13, 0, 30, 21, 10, 0), + DDS_VAL(31, 20, 11, 0, 28, 23, 8, 0), + DDS_VAL(31, 21, 10, 0, 27, 24, 7, 0), + DDS_VAL(31, 22, 9, 0, 26, 25, 6, 0), + DDS_VAL(30, 23, 8, 0, 25, 26, 5, 0), + DDS_VAL(29, 24, 7, 0, 23, 27, 4, 0), + /* Vendor recommends below for 1m cable */ +#define DDS_1M 13 + DDS_VAL(28, 25, 6, 0, 21, 28, 3, 0), + DDS_VAL(27, 26, 5, 0, 19, 29, 2, 0), + DDS_VAL(25, 27, 4, 0, 17, 30, 1, 0) +}; + +/* + * Next, values related to Receive Equalization. + * In comments, FDR (Full) is IB DDR, HDR (Half) is IB SDR + */ +/* Hardware packs an element number and register address thus: */ +#define RXEQ_INIT_RDESC(elt, addr) (((elt) & 0xF) | ((addr) << 4)) +#define RXEQ_VAL(elt, adr, val0, val1, val2, val3) \ + {RXEQ_INIT_RDESC((elt), (adr)), {(val0), (val1), (val2), (val3)} } + +#define RXEQ_VAL_ALL(elt, adr, val) \ + {RXEQ_INIT_RDESC((elt), (adr)), {(val), (val), (val), (val)} } + +#define RXEQ_SDR_DFELTH 0 +#define RXEQ_SDR_TLTH 0 +#define RXEQ_SDR_G1CNT_Z1CNT 0x11 +#define RXEQ_SDR_ZCNT 23 + +static struct rxeq_init { + u16 rdesc; /* in form used in SerDesDDSRXEQ */ + u8 rdata[4]; +} rxeq_init_vals[] = { + /* Set Rcv Eq. to Preset node */ + RXEQ_VAL_ALL(7, 0x27, 0x10), + /* Set DFELTHFDR/HDR thresholds */ + RXEQ_VAL(7, 8, 0, 0, 0, 0), /* FDR */ + RXEQ_VAL(7, 0x21, 0, 0, 0, 0), /* HDR */ + /* Set TLTHFDR/HDR theshold */ + RXEQ_VAL(7, 9, 2, 2, 2, 2), /* FDR */ + RXEQ_VAL(7, 0x23, 2, 2, 2, 2), /* HDR */ + /* Set Preamp setting 2 (ZFR/ZCNT) */ + RXEQ_VAL(7, 0x1B, 12, 12, 12, 12), /* FDR */ + RXEQ_VAL(7, 0x1C, 12, 12, 12, 12), /* HDR */ + /* Set Preamp DC gain and Setting 1 (GFR/GHR) */ + RXEQ_VAL(7, 0x1E, 0x10, 0x10, 0x10, 0x10), /* FDR */ + RXEQ_VAL(7, 0x1F, 0x10, 0x10, 0x10, 0x10), /* HDR */ + /* Toggle RELOCK (in VCDL_CTRL0) to lock to data */ + RXEQ_VAL_ALL(6, 6, 0x20), /* Set D5 High */ + RXEQ_VAL_ALL(6, 6, 0), /* Set D5 Low */ +}; + +/* There are 17 values from vendor, but IBC only accesses the first 16 */ +#define DDS_ROWS (16) +#define RXEQ_ROWS ARRAY_SIZE(rxeq_init_vals) + +static int ipath_sd_setvals(struct ipath_devdata *dd) +{ + int idx, midx; + int min_idx; /* Minimum index for this portion of table */ + uint32_t dds_reg_map; + u64 __iomem *taddr, *iaddr; + uint64_t data; + uint64_t sdctl; + + taddr = dd->ipath_kregbase + KR_IBSerDesMappTable; + iaddr = dd->ipath_kregbase + dd->ipath_kregs->kr_ib_ddsrxeq; + + /* + * Init the DDS section of the table. + * Each "row" of the table provokes NUM_DDS_REG writes, to the + * registers indicated in DDS_REG_MAP. + */ + sdctl = ipath_read_kreg64(dd, dd->ipath_kregs->kr_ibserdesctrl); + sdctl = (sdctl & ~(0x1f << 8)) | (NUM_DDS_REGS << 8); + sdctl = (sdctl & ~(0x1f << 13)) | (RXEQ_ROWS << 13); + ipath_write_kreg(dd, dd->ipath_kregs->kr_ibserdesctrl, sdctl); + + /* + * Iterate down table within loop for each register to store. + */ + dds_reg_map = DDS_REG_MAP; + for (idx = 0; idx < NUM_DDS_REGS; ++idx) { + data = ((dds_reg_map & 0xF) << 4) | TX_FAST_ELT; + writeq(data, iaddr + idx); + mmiowb(); + ipath_read_kreg32(dd, dd->ipath_kregs->kr_scratch); + dds_reg_map >>= 4; + for (midx = 0; midx < DDS_ROWS; ++midx) { + u64 __iomem *daddr = taddr + ((midx << 4) + idx); + data = dds_init_vals[midx].reg_vals[idx]; + writeq(data, daddr); + mmiowb(); + ipath_read_kreg32(dd, dd->ipath_kregs->kr_scratch); + } /* End inner for (vals for this reg, each row) */ + } /* end outer for (regs to be stored) */ + + /* + * Init the RXEQ section of the table. As explained above the table + * rxeq_init_vals[], this runs in a different order, as the pattern + * of register references is more complex, but there are only + * four "data" values per register. + */ + min_idx = idx; /* RXEQ indices pick up where DDS left off */ + taddr += 0x100; /* RXEQ data is in second half of table */ + /* Iterate through RXEQ register addresses */ + for (idx = 0; idx < RXEQ_ROWS; ++idx) { + int didx; /* "destination" */ + int vidx; + + /* didx is offset by min_idx to address RXEQ range of regs */ + didx = idx + min_idx; + /* Store the next RXEQ register address */ + writeq(rxeq_init_vals[idx].rdesc, iaddr + didx); + mmiowb(); + ipath_read_kreg32(dd, dd->ipath_kregs->kr_scratch); + /* Iterate through RXEQ values */ + for (vidx = 0; vidx < 4; vidx++) { + data = rxeq_init_vals[idx].rdata[vidx]; + writeq(data, taddr + (vidx << 6) + idx); + mmiowb(); + ipath_read_kreg32(dd, dd->ipath_kregs->kr_scratch); + } + } /* end outer for (Reg-writes for RXEQ) */ + return 0; +} + +#define CMUCTRL5 EPB_LOC(7, 0, 0x15) +#define RXHSCTRL0(chan) EPB_LOC(chan, 6, 0) +#define VCDL_DAC2(chan) EPB_LOC(chan, 6, 5) +#define VCDL_CTRL0(chan) EPB_LOC(chan, 6, 6) +#define VCDL_CTRL2(chan) EPB_LOC(chan, 6, 8) +#define START_EQ2(chan) EPB_LOC(chan, 7, 0x28) + +static int ibsd_sto_noisy(struct ipath_devdata *dd, int loc, int val, int mask) +{ + int ret = -1; + int sloc; /* shifted loc, for messages */ + + loc |= (1U << EPB_IB_QUAD0_CS_SHF); + sloc = loc >> EPB_ADDR_SHF; + + ret = ipath_sd7220_reg_mod(dd, IB_7220_SERDES, loc, val, mask); + if (ret < 0) + ipath_dev_err(dd, "Write failed: elt %d," + " addr 0x%X, chnl %d, val 0x%02X, mask 0x%02X\n", + (sloc & 0xF), (sloc >> 9) & 0x3f, (sloc >> 4) & 7, + val & 0xFF, mask & 0xFF); + return ret; +} + +/* + * Repeat a "store" across all channels of the IB SerDes. + * Although nominally it inherits the "read value" of the last + * channel it modified, the only really useful return is <0 for + * failure, >= 0 for success. The parameter 'loc' is assumed to + * be the location for the channel-0 copy of the register to + * be modified. + */ +static int ibsd_mod_allchnls(struct ipath_devdata *dd, int loc, int val, + int mask) +{ + int ret = -1; + int chnl; + + if (loc & EPB_GLOBAL_WR) { + /* + * Our caller has assured us that we can set all four + * channels at once. Trust that. If mask is not 0xFF, + * we will read the _specified_ channel for our starting + * value. + */ + loc |= (1U << EPB_IB_QUAD0_CS_SHF); + chnl = (loc >> (4 + EPB_ADDR_SHF)) & 7; + if (mask != 0xFF) { + ret = ipath_sd7220_reg_mod(dd, IB_7220_SERDES, + loc & ~EPB_GLOBAL_WR, 0, 0); + if (ret < 0) { + int sloc = loc >> EPB_ADDR_SHF; + ipath_dev_err(dd, "pre-read failed: elt %d," + " addr 0x%X, chnl %d\n", (sloc & 0xF), + (sloc >> 9) & 0x3f, chnl); + return ret; + } + val = (ret & ~mask) | (val & mask); + } + loc &= ~(7 << (4+EPB_ADDR_SHF)); + ret = ipath_sd7220_reg_mod(dd, IB_7220_SERDES, loc, val, 0xFF); + if (ret < 0) { + int sloc = loc >> EPB_ADDR_SHF; + ipath_dev_err(dd, "Global WR failed: elt %d," + " addr 0x%X, val %02X\n", + (sloc & 0xF), (sloc >> 9) & 0x3f, val); + } + return ret; + } + /* Clear "channel" and set CS so we can simply iterate */ + loc &= ~(7 << (4+EPB_ADDR_SHF)); + loc |= (1U << EPB_IB_QUAD0_CS_SHF); + for (chnl = 0; chnl < 4; ++chnl) { + int cloc; + cloc = loc | (chnl << (4+EPB_ADDR_SHF)); + ret = ipath_sd7220_reg_mod(dd, IB_7220_SERDES, cloc, val, mask); + if (ret < 0) { + int sloc = loc >> EPB_ADDR_SHF; + ipath_dev_err(dd, "Write failed: elt %d," + " addr 0x%X, chnl %d, val 0x%02X," + " mask 0x%02X\n", + (sloc & 0xF), (sloc >> 9) & 0x3f, chnl, + val & 0xFF, mask & 0xFF); + break; + } + } + return ret; +} + +/* + * Set the Tx values normally modified by IBC in IB1.2 mode to default + * values, as gotten from first row of init table. + */ +static int set_dds_vals(struct ipath_devdata *dd, struct dds_init *ddi) +{ + int ret; + int idx, reg, data; + uint32_t regmap; + + regmap = DDS_REG_MAP; + for (idx = 0; idx < NUM_DDS_REGS; ++idx) { + reg = (regmap & 0xF); + regmap >>= 4; + data = ddi->reg_vals[idx]; + /* Vendor says RMW not needed for these regs, use 0xFF mask */ + ret = ibsd_mod_allchnls(dd, EPB_LOC(0, 9, reg), data, 0xFF); + if (ret < 0) + break; + } + return ret; +} + +/* + * Set the Rx values normally modified by IBC in IB1.2 mode to default + * values, as gotten from selected column of init table. + */ +static int set_rxeq_vals(struct ipath_devdata *dd, int vsel) +{ + int ret; + int ridx; + int cnt = ARRAY_SIZE(rxeq_init_vals); + + for (ridx = 0; ridx < cnt; ++ridx) { + int elt, reg, val, loc; + elt = rxeq_init_vals[ridx].rdesc & 0xF; + reg = rxeq_init_vals[ridx].rdesc >> 4; + loc = EPB_LOC(0, elt, reg); + val = rxeq_init_vals[ridx].rdata[vsel]; + /* mask of 0xFF, because hardware does full-byte store. */ + ret = ibsd_mod_allchnls(dd, loc, val, 0xFF); + if (ret < 0) + break; + } + return ret; +} + +/* + * Set the default values (row 0) for DDR Driver Demphasis. + * we do this initially and whenever we turn off IB-1.2 + * The "default" values for Rx equalization are also stored to + * SerDes registers. Formerly (and still default), we used set 2. + * For experimenting with cables and link-partners, we allow changing + * that via a module parameter. + */ +static unsigned ipath_rxeq_set = 2; +module_param_named(rxeq_default_set, ipath_rxeq_set, uint, + S_IWUSR | S_IRUGO); +MODULE_PARM_DESC(rxeq_default_set, + "Which set [0..3] of Rx Equalization values is default"); + +static int ipath_internal_presets(struct ipath_devdata *dd) +{ + int ret = 0; + + ret = set_dds_vals(dd, dds_init_vals + DDS_3M); + + if (ret < 0) + ipath_dev_err(dd, "Failed to set default DDS values\n"); + ret = set_rxeq_vals(dd, ipath_rxeq_set & 3); + if (ret < 0) + ipath_dev_err(dd, "Failed to set default RXEQ values\n"); + return ret; +} + +int ipath_sd7220_presets(struct ipath_devdata *dd) +{ + int ret = 0; + + if (!dd->ipath_presets_needed) + return ret; + dd->ipath_presets_needed = 0; + /* Assert uC reset, so we don't clash with it. */ + ipath_ibsd_reset(dd, 1); + udelay(2); + ipath_sd_trimdone_monitor(dd, "link-down"); + + ret = ipath_internal_presets(dd); +return ret; +} + +static int ipath_sd_trimself(struct ipath_devdata *dd, int val) +{ + return ibsd_sto_noisy(dd, CMUCTRL5, val, 0xFF); +} + +static int ipath_sd_early(struct ipath_devdata *dd) +{ + int ret = -1; /* Default failed */ + int chnl; + + for (chnl = 0; chnl < 4; ++chnl) { + ret = ibsd_sto_noisy(dd, RXHSCTRL0(chnl), 0xD4, 0xFF); + if (ret < 0) + goto bail; + } + for (chnl = 0; chnl < 4; ++chnl) { + ret = ibsd_sto_noisy(dd, VCDL_DAC2(chnl), 0x2D, 0xFF); + if (ret < 0) + goto bail; + } + /* more fine-tuning of what will be default */ + for (chnl = 0; chnl < 4; ++chnl) { + ret = ibsd_sto_noisy(dd, VCDL_CTRL2(chnl), 3, 0xF); + if (ret < 0) + goto bail; + } + for (chnl = 0; chnl < 4; ++chnl) { + ret = ibsd_sto_noisy(dd, START_EQ1(chnl), 0x10, 0xFF); + if (ret < 0) + goto bail; + } + for (chnl = 0; chnl < 4; ++chnl) { + ret = ibsd_sto_noisy(dd, START_EQ2(chnl), 0x30, 0xFF); + if (ret < 0) + goto bail; + } +bail: + return ret; +} + +#define BACTRL(chnl) EPB_LOC(chnl, 6, 0x0E) +#define LDOUTCTRL1(chnl) EPB_LOC(chnl, 7, 6) +#define RXHSSTATUS(chnl) EPB_LOC(chnl, 6, 0xF) + +static int ipath_sd_dactrim(struct ipath_devdata *dd) +{ + int ret = -1; /* Default failed */ + int chnl; + + for (chnl = 0; chnl < 4; ++chnl) { + ret = ibsd_sto_noisy(dd, BACTRL(chnl), 0x40, 0xFF); + if (ret < 0) + goto bail; + } + for (chnl = 0; chnl < 4; ++chnl) { + ret = ibsd_sto_noisy(dd, LDOUTCTRL1(chnl), 0x04, 0xFF); + if (ret < 0) + goto bail; + } + for (chnl = 0; chnl < 4; ++chnl) { + ret = ibsd_sto_noisy(dd, RXHSSTATUS(chnl), 0x04, 0xFF); + if (ret < 0) + goto bail; + } + /* + * delay for max possible number of steps, with slop. + * Each step is about 4usec. + */ + udelay(415); + for (chnl = 0; chnl < 4; ++chnl) { + ret = ibsd_sto_noisy(dd, LDOUTCTRL1(chnl), 0x00, 0xFF); + if (ret < 0) + goto bail; + } +bail: + return ret; +} + +#define RELOCK_FIRST_MS 3 +#define RXLSPPM(chan) EPB_LOC(chan, 0, 2) +void ipath_toggle_rclkrls(struct ipath_devdata *dd) +{ + int loc = RXLSPPM(0) | EPB_GLOBAL_WR; + int ret; + + ret = ibsd_mod_allchnls(dd, loc, 0, 0x80); + if (ret < 0) + ipath_dev_err(dd, "RCLKRLS failed to clear D7\n"); + else { + udelay(1); + ibsd_mod_allchnls(dd, loc, 0x80, 0x80); + } + /* And again for good measure */ + udelay(1); + ret = ibsd_mod_allchnls(dd, loc, 0, 0x80); + if (ret < 0) + ipath_dev_err(dd, "RCLKRLS failed to clear D7\n"); + else { + udelay(1); + ibsd_mod_allchnls(dd, loc, 0x80, 0x80); + } + /* Now reset xgxs and IBC to complete the recovery */ + dd->ipath_f_xgxs_reset(dd); +} + +/* + * Shut down the timer that polls for relock occasions, if needed + * this is "hooked" from ipath_7220_quiet_serdes(), which is called + * just before ipath_shutdown_device() in ipath_driver.c shuts down all + * the other timers + */ +void ipath_shutdown_relock_poll(struct ipath_devdata *dd) +{ + struct ipath_relock *irp = &dd->ipath_relock_singleton; + if (atomic_read(&irp->ipath_relock_timer_active)) { + del_timer_sync(&irp->ipath_relock_timer); + atomic_set(&irp->ipath_relock_timer_active, 0); + } +} + +static unsigned ipath_relock_by_timer = 1; +module_param_named(relock_by_timer, ipath_relock_by_timer, uint, + S_IWUSR | S_IRUGO); +MODULE_PARM_DESC(relock_by_timer, "Allow relock attempt if link not up"); + +static void ipath_run_relock(unsigned long opaque) +{ + struct ipath_devdata *dd = (struct ipath_devdata *)opaque; + struct ipath_relock *irp = &dd->ipath_relock_singleton; + u64 val, ltstate; + + if (!(dd->ipath_flags & IPATH_INITTED)) { + /* Not yet up, just reenable the timer for later */ + irp->ipath_relock_interval = HZ; + mod_timer(&irp->ipath_relock_timer, jiffies + HZ); + return; + } + + /* + * Check link-training state for "stuck" state. + * if found, try relock and schedule another try at + * exponentially growing delay, maxed at one second. + * if not stuck, our work is done. + */ + val = ipath_read_kreg64(dd, dd->ipath_kregs->kr_ibcstatus); + ltstate = ipath_ib_linktrstate(dd, val); + + if (ltstate <= INFINIPATH_IBCS_LT_STATE_CFGWAITRMT + && ltstate != INFINIPATH_IBCS_LT_STATE_LINKUP) { + int timeoff; + /* Not up yet. Try again, if allowed by module-param */ + if (ipath_relock_by_timer) { + if (dd->ipath_flags & IPATH_IB_AUTONEG_INPROG) + ipath_cdbg(VERBOSE, "Skip RELOCK in AUTONEG\n"); + else if (!(dd->ipath_flags & IPATH_IB_LINK_DISABLED)) { + ipath_cdbg(VERBOSE, "RELOCK\n"); + ipath_toggle_rclkrls(dd); + } + } + /* re-set timer for next check */ + timeoff = irp->ipath_relock_interval << 1; + if (timeoff > HZ) + timeoff = HZ; + irp->ipath_relock_interval = timeoff; + + mod_timer(&irp->ipath_relock_timer, jiffies + timeoff); + } else { + /* Up, so no more need to check so often */ + mod_timer(&irp->ipath_relock_timer, jiffies + HZ); + } +} + +void ipath_set_relock_poll(struct ipath_devdata *dd, int ibup) +{ + struct ipath_relock *irp = &dd->ipath_relock_singleton; + + if (ibup > 0) { + /* we are now up, so relax timer to 1 second interval */ + if (atomic_read(&irp->ipath_relock_timer_active)) + mod_timer(&irp->ipath_relock_timer, jiffies + HZ); + } else { + /* Transition to down, (re-)set timer to short interval. */ + int timeout; + timeout = (HZ * ((ibup == -1) ? 1000 : RELOCK_FIRST_MS))/1000; + if (timeout == 0) + timeout = 1; + /* If timer has not yet been started, do so. */ + if (atomic_inc_return(&irp->ipath_relock_timer_active) == 1) { + init_timer(&irp->ipath_relock_timer); + irp->ipath_relock_timer.function = ipath_run_relock; + irp->ipath_relock_timer.data = (unsigned long) dd; + irp->ipath_relock_interval = timeout; + irp->ipath_relock_timer.expires = jiffies + timeout; + add_timer(&irp->ipath_relock_timer); + } else { + irp->ipath_relock_interval = timeout; + mod_timer(&irp->ipath_relock_timer, jiffies + timeout); + atomic_dec(&irp->ipath_relock_timer_active); + } + } +} + diff --git a/drivers/infiniband/hw/ipath/ipath_sd7220_img.c b/drivers/infiniband/hw/ipath/ipath_sd7220_img.c new file mode 100644 index 000000000000..5ef59da9270a --- /dev/null +++ b/drivers/infiniband/hw/ipath/ipath_sd7220_img.c @@ -0,0 +1,1082 @@ +/* + * Copyright (c) 2007, 2008 QLogic Corporation. All rights reserved. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * OpenIB.org BSD license below: + * + * Redistribution and use in source and binary forms, with or + * without modification, are permitted provided that the following + * conditions are met: + * + * - Redistributions of source code must retain the above + * copyright notice, this list of conditions and the following + * disclaimer. + * + * - Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials + * provided with the distribution. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + */ + +/* + * This file contains the memory image from the vendor, to be copied into + * the IB SERDES of the IBA7220 during initialization. + * The file also includes the two functions which use this image. + */ +#include <linux/pci.h> +#include <linux/delay.h> + +#include "ipath_kernel.h" +#include "ipath_registers.h" +#include "ipath_7220.h" + +static unsigned char ipath_sd7220_ib_img[] = { +/*0000*/0x02, 0x0A, 0x29, 0x02, 0x0A, 0x87, 0xE5, 0xE6, + 0x30, 0xE6, 0x04, 0x7F, 0x01, 0x80, 0x02, 0x7F, +/*0010*/0x00, 0xE5, 0xE2, 0x30, 0xE4, 0x04, 0x7E, 0x01, + 0x80, 0x02, 0x7E, 0x00, 0xEE, 0x5F, 0x60, 0x08, +/*0020*/0x53, 0xF9, 0xF7, 0xE4, 0xF5, 0xFE, 0x80, 0x08, + 0x7F, 0x0A, 0x12, 0x17, 0x31, 0x12, 0x0E, 0xA2, +/*0030*/0x75, 0xFC, 0x08, 0xE4, 0xF5, 0xFD, 0xE5, 0xE7, + 0x20, 0xE7, 0x03, 0x43, 0xF9, 0x08, 0x22, 0x00, +/*0040*/0x01, 0x20, 0x11, 0x00, 0x04, 0x20, 0x00, 0x75, + 0x51, 0x01, 0xE4, 0xF5, 0x52, 0xF5, 0x53, 0xF5, +/*0050*/0x52, 0xF5, 0x7E, 0x7F, 0x04, 0x02, 0x04, 0x38, + 0xC2, 0x36, 0x05, 0x52, 0xE5, 0x52, 0xD3, 0x94, +/*0060*/0x0C, 0x40, 0x05, 0x75, 0x52, 0x01, 0xD2, 0x36, + 0x90, 0x07, 0x0C, 0x74, 0x07, 0xF0, 0xA3, 0x74, +/*0070*/0xFF, 0xF0, 0xE4, 0xF5, 0x0C, 0xA3, 0xF0, 0x90, + 0x07, 0x14, 0xF0, 0xA3, 0xF0, 0x75, 0x0B, 0x20, +/*0080*/0xF5, 0x09, 0xE4, 0xF5, 0x08, 0xE5, 0x08, 0xD3, + 0x94, 0x30, 0x40, 0x03, 0x02, 0x04, 0x04, 0x12, +/*0090*/0x00, 0x06, 0x15, 0x0B, 0xE5, 0x08, 0x70, 0x04, + 0x7F, 0x01, 0x80, 0x02, 0x7F, 0x00, 0xE5, 0x09, +/*00A0*/0x70, 0x04, 0x7E, 0x01, 0x80, 0x02, 0x7E, 0x00, + 0xEE, 0x5F, 0x60, 0x05, 0x12, 0x18, 0x71, 0xD2, +/*00B0*/0x35, 0x53, 0xE1, 0xF7, 0xE5, 0x08, 0x45, 0x09, + 0xFF, 0xE5, 0x0B, 0x25, 0xE0, 0x25, 0xE0, 0x24, +/*00C0*/0x83, 0xF5, 0x82, 0xE4, 0x34, 0x07, 0xF5, 0x83, + 0xEF, 0xF0, 0x85, 0xE2, 0x20, 0xE5, 0x52, 0xD3, +/*00D0*/0x94, 0x01, 0x40, 0x0D, 0x12, 0x19, 0xF3, 0xE0, + 0x54, 0xA0, 0x64, 0x40, 0x70, 0x03, 0x02, 0x03, +/*00E0*/0xFB, 0x53, 0xF9, 0xF8, 0x90, 0x94, 0x70, 0xE4, + 0xF0, 0xE0, 0xF5, 0x10, 0xAF, 0x09, 0x12, 0x1E, +/*00F0*/0xB3, 0xAF, 0x08, 0xEF, 0x44, 0x08, 0xF5, 0x82, + 0x75, 0x83, 0x80, 0xE0, 0xF5, 0x29, 0xEF, 0x44, +/*0100*/0x07, 0x12, 0x1A, 0x3C, 0xF5, 0x22, 0x54, 0x40, + 0xD3, 0x94, 0x00, 0x40, 0x1E, 0xE5, 0x29, 0x54, +/*0110*/0xF0, 0x70, 0x21, 0x12, 0x19, 0xF3, 0xE0, 0x44, + 0x80, 0xF0, 0xE5, 0x22, 0x54, 0x30, 0x65, 0x08, +/*0120*/0x70, 0x09, 0x12, 0x19, 0xF3, 0xE0, 0x54, 0xBF, + 0xF0, 0x80, 0x09, 0x12, 0x19, 0xF3, 0x74, 0x40, +/*0130*/0xF0, 0x02, 0x03, 0xFB, 0x12, 0x1A, 0x12, 0x75, + 0x83, 0xAE, 0x74, 0xFF, 0xF0, 0xAF, 0x08, 0x7E, +/*0140*/0x00, 0xEF, 0x44, 0x07, 0xF5, 0x82, 0xE0, 0xFD, + 0xE5, 0x0B, 0x25, 0xE0, 0x25, 0xE0, 0x24, 0x81, +/*0150*/0xF5, 0x82, 0xE4, 0x34, 0x07, 0xF5, 0x83, 0xED, + 0xF0, 0x90, 0x07, 0x0E, 0xE0, 0x04, 0xF0, 0xEF, +/*0160*/0x44, 0x07, 0xF5, 0x82, 0x75, 0x83, 0x98, 0xE0, + 0xF5, 0x28, 0x12, 0x1A, 0x23, 0x40, 0x0C, 0x12, +/*0170*/0x19, 0xF3, 0xE0, 0x44, 0x01, 0x12, 0x1A, 0x32, + 0x02, 0x03, 0xF6, 0xAF, 0x08, 0x7E, 0x00, 0x74, +/*0180*/0x80, 0xCD, 0xEF, 0xCD, 0x8D, 0x82, 0xF5, 0x83, + 0xE0, 0x30, 0xE0, 0x0A, 0x12, 0x19, 0xF3, 0xE0, +/*0190*/0x44, 0x20, 0xF0, 0x02, 0x03, 0xFB, 0x12, 0x19, + 0xF3, 0xE0, 0x54, 0xDF, 0xF0, 0xEE, 0x44, 0xAE, +/*01A0*/0x12, 0x1A, 0x43, 0x30, 0xE4, 0x03, 0x02, 0x03, + 0xFB, 0x74, 0x9E, 0x12, 0x1A, 0x05, 0x20, 0xE0, +/*01B0*/0x03, 0x02, 0x03, 0xFB, 0x8F, 0x82, 0x8E, 0x83, + 0xE0, 0x20, 0xE0, 0x03, 0x02, 0x03, 0xFB, 0x12, +/*01C0*/0x19, 0xF3, 0xE0, 0x44, 0x10, 0xF0, 0xE5, 0xE3, + 0x20, 0xE7, 0x08, 0xE5, 0x08, 0x12, 0x1A, 0x3A, +/*01D0*/0x44, 0x04, 0xF0, 0xAF, 0x08, 0x7E, 0x00, 0xEF, + 0x12, 0x1A, 0x3A, 0x20, 0xE2, 0x34, 0x12, 0x19, +/*01E0*/0xF3, 0xE0, 0x44, 0x08, 0xF0, 0xE5, 0xE4, 0x30, + 0xE6, 0x04, 0x7D, 0x01, 0x80, 0x02, 0x7D, 0x00, +/*01F0*/0xE5, 0x7E, 0xC3, 0x94, 0x04, 0x50, 0x04, 0x7C, + 0x01, 0x80, 0x02, 0x7C, 0x00, 0xEC, 0x4D, 0x60, +/*0200*/0x05, 0xC2, 0x35, 0x02, 0x03, 0xFB, 0xEE, 0x44, + 0xD2, 0x12, 0x1A, 0x43, 0x44, 0x40, 0xF0, 0x02, +/*0210*/0x03, 0xFB, 0x12, 0x19, 0xF3, 0xE0, 0x54, 0xF7, + 0xF0, 0x12, 0x1A, 0x12, 0x75, 0x83, 0xD2, 0xE0, +/*0220*/0x54, 0xBF, 0xF0, 0x90, 0x07, 0x14, 0xE0, 0x04, + 0xF0, 0xE5, 0x7E, 0x70, 0x03, 0x75, 0x7E, 0x01, +/*0230*/0xAF, 0x08, 0x7E, 0x00, 0x12, 0x1A, 0x23, 0x40, + 0x12, 0x12, 0x19, 0xF3, 0xE0, 0x44, 0x01, 0x12, +/*0240*/0x19, 0xF2, 0xE0, 0x54, 0x02, 0x12, 0x1A, 0x32, + 0x02, 0x03, 0xFB, 0x12, 0x19, 0xF3, 0xE0, 0x44, +/*0250*/0x02, 0x12, 0x19, 0xF2, 0xE0, 0x54, 0xFE, 0xF0, + 0xC2, 0x35, 0xEE, 0x44, 0x8A, 0x8F, 0x82, 0xF5, +/*0260*/0x83, 0xE0, 0xF5, 0x17, 0x54, 0x8F, 0x44, 0x40, + 0xF0, 0x74, 0x90, 0xFC, 0xE5, 0x08, 0x44, 0x07, +/*0270*/0xFD, 0xF5, 0x82, 0x8C, 0x83, 0xE0, 0x54, 0x3F, + 0x90, 0x07, 0x02, 0xF0, 0xE0, 0x54, 0xC0, 0x8D, +/*0280*/0x82, 0x8C, 0x83, 0xF0, 0x74, 0x92, 0x12, 0x1A, + 0x05, 0x90, 0x07, 0x03, 0x12, 0x1A, 0x19, 0x74, +/*0290*/0x82, 0x12, 0x1A, 0x05, 0x90, 0x07, 0x04, 0x12, + 0x1A, 0x19, 0x74, 0xB4, 0x12, 0x1A, 0x05, 0x90, +/*02A0*/0x07, 0x05, 0x12, 0x1A, 0x19, 0x74, 0x94, 0xFE, + 0xE5, 0x08, 0x44, 0x06, 0x12, 0x1A, 0x0A, 0xF5, +/*02B0*/0x10, 0x30, 0xE0, 0x04, 0xD2, 0x37, 0x80, 0x02, + 0xC2, 0x37, 0xE5, 0x10, 0x54, 0x7F, 0x8F, 0x82, +/*02C0*/0x8E, 0x83, 0xF0, 0x30, 0x44, 0x30, 0x12, 0x1A, + 0x03, 0x54, 0x80, 0xD3, 0x94, 0x00, 0x40, 0x04, +/*02D0*/0xD2, 0x39, 0x80, 0x02, 0xC2, 0x39, 0x8F, 0x82, + 0x8E, 0x83, 0xE0, 0x44, 0x80, 0xF0, 0x12, 0x1A, +/*02E0*/0x03, 0x54, 0x40, 0xD3, 0x94, 0x00, 0x40, 0x04, + 0xD2, 0x3A, 0x80, 0x02, 0xC2, 0x3A, 0x8F, 0x82, +/*02F0*/0x8E, 0x83, 0xE0, 0x44, 0x40, 0xF0, 0x74, 0x92, + 0xFE, 0xE5, 0x08, 0x44, 0x06, 0x12, 0x1A, 0x0A, +/*0300*/0x30, 0xE7, 0x04, 0xD2, 0x38, 0x80, 0x02, 0xC2, + 0x38, 0x8F, 0x82, 0x8E, 0x83, 0xE0, 0x54, 0x7F, +/*0310*/0xF0, 0x12, 0x1E, 0x46, 0xE4, 0xF5, 0x0A, 0x20, + 0x03, 0x02, 0x80, 0x03, 0x30, 0x43, 0x03, 0x12, +/*0320*/0x19, 0x95, 0x20, 0x02, 0x02, 0x80, 0x03, 0x30, + 0x42, 0x03, 0x12, 0x0C, 0x8F, 0x30, 0x30, 0x06, +/*0330*/0x12, 0x19, 0x95, 0x12, 0x0C, 0x8F, 0x12, 0x0D, + 0x47, 0x12, 0x19, 0xF3, 0xE0, 0x54, 0xFB, 0xF0, +/*0340*/0xE5, 0x0A, 0xC3, 0x94, 0x01, 0x40, 0x46, 0x43, + 0xE1, 0x08, 0x12, 0x19, 0xF3, 0xE0, 0x44, 0x04, +/*0350*/0xF0, 0xE5, 0xE4, 0x20, 0xE7, 0x2A, 0x12, 0x1A, + 0x12, 0x75, 0x83, 0xD2, 0xE0, 0x54, 0x08, 0xD3, +/*0360*/0x94, 0x00, 0x40, 0x04, 0x7F, 0x01, 0x80, 0x02, + 0x7F, 0x00, 0xE5, 0x0A, 0xC3, 0x94, 0x01, 0x40, +/*0370*/0x04, 0x7E, 0x01, 0x80, 0x02, 0x7E, 0x00, 0xEF, + 0x5E, 0x60, 0x05, 0x12, 0x1D, 0xD7, 0x80, 0x17, +/*0380*/0x12, 0x1A, 0x12, 0x75, 0x83, 0xD2, 0xE0, 0x44, + 0x08, 0xF0, 0x02, 0x03, 0xFB, 0x12, 0x1A, 0x12, +/*0390*/0x75, 0x83, 0xD2, 0xE0, 0x54, 0xF7, 0xF0, 0x12, + 0x1E, 0x46, 0x7F, 0x08, 0x12, 0x17, 0x31, 0x74, +/*03A0*/0x8E, 0xFE, 0x12, 0x1A, 0x12, 0x8E, 0x83, 0xE0, + 0xF5, 0x10, 0x54, 0xFE, 0xF0, 0xE5, 0x10, 0x44, +/*03B0*/0x01, 0xFF, 0xE5, 0x08, 0xFD, 0xED, 0x44, 0x07, + 0xF5, 0x82, 0xEF, 0xF0, 0xE5, 0x10, 0x54, 0xFE, +/*03C0*/0xFF, 0xED, 0x44, 0x07, 0xF5, 0x82, 0xEF, 0x12, + 0x1A, 0x11, 0x75, 0x83, 0x86, 0xE0, 0x44, 0x10, +/*03D0*/0x12, 0x1A, 0x11, 0xE0, 0x44, 0x10, 0xF0, 0x12, + 0x19, 0xF3, 0xE0, 0x54, 0xFD, 0x44, 0x01, 0xFF, +/*03E0*/0x12, 0x19, 0xF3, 0xEF, 0x12, 0x1A, 0x32, 0x30, + 0x32, 0x0C, 0xE5, 0x08, 0x44, 0x08, 0xF5, 0x82, +/*03F0*/0x75, 0x83, 0x82, 0x74, 0x05, 0xF0, 0xAF, 0x0B, + 0x12, 0x18, 0xD7, 0x74, 0x10, 0x25, 0x08, 0xF5, +/*0400*/0x08, 0x02, 0x00, 0x85, 0x05, 0x09, 0xE5, 0x09, + 0xD3, 0x94, 0x07, 0x50, 0x03, 0x02, 0x00, 0x82, +/*0410*/0xE5, 0x7E, 0xD3, 0x94, 0x00, 0x40, 0x04, 0x7F, + 0x01, 0x80, 0x02, 0x7F, 0x00, 0xE5, 0x7E, 0xC3, +/*0420*/0x94, 0xFA, 0x50, 0x04, 0x7E, 0x01, 0x80, 0x02, + 0x7E, 0x00, 0xEE, 0x5F, 0x60, 0x02, 0x05, 0x7E, +/*0430*/0x30, 0x35, 0x0B, 0x43, 0xE1, 0x01, 0x7F, 0x09, + 0x12, 0x17, 0x31, 0x02, 0x00, 0x58, 0x53, 0xE1, +/*0440*/0xFE, 0x02, 0x00, 0x58, 0x8E, 0x6A, 0x8F, 0x6B, + 0x8C, 0x6C, 0x8D, 0x6D, 0x75, 0x6E, 0x01, 0x75, +/*0450*/0x6F, 0x01, 0x75, 0x70, 0x01, 0xE4, 0xF5, 0x73, + 0xF5, 0x74, 0xF5, 0x75, 0x90, 0x07, 0x2F, 0xF0, +/*0460*/0xF5, 0x3C, 0xF5, 0x3E, 0xF5, 0x46, 0xF5, 0x47, + 0xF5, 0x3D, 0xF5, 0x3F, 0xF5, 0x6F, 0xE5, 0x6F, +/*0470*/0x70, 0x0F, 0xE5, 0x6B, 0x45, 0x6A, 0x12, 0x07, + 0x2A, 0x75, 0x83, 0x80, 0x74, 0x3A, 0xF0, 0x80, +/*0480*/0x09, 0x12, 0x07, 0x2A, 0x75, 0x83, 0x80, 0x74, + 0x1A, 0xF0, 0xE4, 0xF5, 0x6E, 0xC3, 0x74, 0x3F, +/*0490*/0x95, 0x6E, 0xFF, 0x12, 0x08, 0x65, 0x75, 0x83, + 0x82, 0xEF, 0xF0, 0x12, 0x1A, 0x4D, 0x12, 0x08, +/*04A0*/0xC6, 0xE5, 0x33, 0xF0, 0x12, 0x08, 0xFA, 0x12, + 0x08, 0xB1, 0x40, 0xE1, 0xE5, 0x6F, 0x70, 0x0B, +/*04B0*/0x12, 0x07, 0x2A, 0x75, 0x83, 0x80, 0x74, 0x36, + 0xF0, 0x80, 0x09, 0x12, 0x07, 0x2A, 0x75, 0x83, +/*04C0*/0x80, 0x74, 0x16, 0xF0, 0x75, 0x6E, 0x01, 0x12, + 0x07, 0x2A, 0x75, 0x83, 0xB4, 0xE5, 0x6E, 0xF0, +/*04D0*/0x12, 0x1A, 0x4D, 0x74, 0x3F, 0x25, 0x6E, 0xF5, + 0x82, 0xE4, 0x34, 0x00, 0xF5, 0x83, 0xE5, 0x33, +/*04E0*/0xF0, 0x74, 0xBF, 0x25, 0x6E, 0xF5, 0x82, 0xE4, + 0x34, 0x00, 0x12, 0x08, 0xB1, 0x40, 0xD8, 0xE4, +/*04F0*/0xF5, 0x70, 0xF5, 0x46, 0xF5, 0x47, 0xF5, 0x6E, + 0x12, 0x08, 0xFA, 0xF5, 0x83, 0xE0, 0xFE, 0x12, +/*0500*/0x08, 0xC6, 0xE0, 0x7C, 0x00, 0x24, 0x00, 0xFF, + 0xEC, 0x3E, 0xFE, 0xAD, 0x3B, 0xD3, 0xEF, 0x9D, +/*0510*/0xEE, 0x9C, 0x50, 0x04, 0x7B, 0x01, 0x80, 0x02, + 0x7B, 0x00, 0xE5, 0x70, 0x70, 0x04, 0x7A, 0x01, +/*0520*/0x80, 0x02, 0x7A, 0x00, 0xEB, 0x5A, 0x60, 0x06, + 0x85, 0x6E, 0x46, 0x75, 0x70, 0x01, 0xD3, 0xEF, +/*0530*/0x9D, 0xEE, 0x9C, 0x50, 0x04, 0x7F, 0x01, 0x80, + 0x02, 0x7F, 0x00, 0xE5, 0x70, 0xB4, 0x01, 0x04, +/*0540*/0x7E, 0x01, 0x80, 0x02, 0x7E, 0x00, 0xEF, 0x5E, + 0x60, 0x03, 0x85, 0x6E, 0x47, 0x05, 0x6E, 0xE5, +/*0550*/0x6E, 0x64, 0x7F, 0x70, 0xA3, 0xE5, 0x46, 0x60, + 0x05, 0xE5, 0x47, 0xB4, 0x7E, 0x03, 0x85, 0x46, +/*0560*/0x47, 0xE5, 0x6F, 0x70, 0x08, 0x85, 0x46, 0x76, + 0x85, 0x47, 0x77, 0x80, 0x0E, 0xC3, 0x74, 0x7F, +/*0570*/0x95, 0x46, 0xF5, 0x78, 0xC3, 0x74, 0x7F, 0x95, + 0x47, 0xF5, 0x79, 0xE5, 0x6F, 0x70, 0x37, 0xE5, +/*0580*/0x46, 0x65, 0x47, 0x70, 0x0C, 0x75, 0x73, 0x01, + 0x75, 0x74, 0x01, 0xF5, 0x3C, 0xF5, 0x3D, 0x80, +/*0590*/0x35, 0xE4, 0xF5, 0x4E, 0xC3, 0xE5, 0x47, 0x95, + 0x46, 0xF5, 0x3C, 0xC3, 0x13, 0xF5, 0x71, 0x25, +/*05A0*/0x46, 0xF5, 0x72, 0xC3, 0x94, 0x3F, 0x40, 0x05, + 0xE4, 0xF5, 0x3D, 0x80, 0x40, 0xC3, 0x74, 0x3F, +/*05B0*/0x95, 0x72, 0xF5, 0x3D, 0x80, 0x37, 0xE5, 0x46, + 0x65, 0x47, 0x70, 0x0F, 0x75, 0x73, 0x01, 0x75, +/*05C0*/0x75, 0x01, 0xF5, 0x3E, 0xF5, 0x3F, 0x75, 0x4E, + 0x01, 0x80, 0x22, 0xE4, 0xF5, 0x4E, 0xC3, 0xE5, +/*05D0*/0x47, 0x95, 0x46, 0xF5, 0x3E, 0xC3, 0x13, 0xF5, + 0x71, 0x25, 0x46, 0xF5, 0x72, 0xD3, 0x94, 0x3F, +/*05E0*/0x50, 0x05, 0xE4, 0xF5, 0x3F, 0x80, 0x06, 0xE5, + 0x72, 0x24, 0xC1, 0xF5, 0x3F, 0x05, 0x6F, 0xE5, +/*05F0*/0x6F, 0xC3, 0x94, 0x02, 0x50, 0x03, 0x02, 0x04, + 0x6E, 0xE5, 0x6D, 0x45, 0x6C, 0x70, 0x02, 0x80, +/*0600*/0x04, 0xE5, 0x74, 0x45, 0x75, 0x90, 0x07, 0x2F, + 0xF0, 0x7F, 0x01, 0xE5, 0x3E, 0x60, 0x04, 0xE5, +/*0610*/0x3C, 0x70, 0x14, 0xE4, 0xF5, 0x3C, 0xF5, 0x3D, + 0xF5, 0x3E, 0xF5, 0x3F, 0x12, 0x08, 0xD2, 0x70, +/*0620*/0x04, 0xF0, 0x02, 0x06, 0xA4, 0x80, 0x7A, 0xE5, + 0x3C, 0xC3, 0x95, 0x3E, 0x40, 0x07, 0xE5, 0x3C, +/*0630*/0x95, 0x3E, 0xFF, 0x80, 0x06, 0xC3, 0xE5, 0x3E, + 0x95, 0x3C, 0xFF, 0xE5, 0x76, 0xD3, 0x95, 0x79, +/*0640*/0x40, 0x05, 0x85, 0x76, 0x7A, 0x80, 0x03, 0x85, + 0x79, 0x7A, 0xE5, 0x77, 0xC3, 0x95, 0x78, 0x50, +/*0650*/0x05, 0x85, 0x77, 0x7B, 0x80, 0x03, 0x85, 0x78, + 0x7B, 0xE5, 0x7B, 0xD3, 0x95, 0x7A, 0x40, 0x30, +/*0660*/0xE5, 0x7B, 0x95, 0x7A, 0xF5, 0x3C, 0xF5, 0x3E, + 0xC3, 0xE5, 0x7B, 0x95, 0x7A, 0x90, 0x07, 0x19, +/*0670*/0xF0, 0xE5, 0x3C, 0xC3, 0x13, 0xF5, 0x71, 0x25, + 0x7A, 0xF5, 0x72, 0xC3, 0x94, 0x3F, 0x40, 0x05, +/*0680*/0xE4, 0xF5, 0x3D, 0x80, 0x1F, 0xC3, 0x74, 0x3F, + 0x95, 0x72, 0xF5, 0x3D, 0xF5, 0x3F, 0x80, 0x14, +/*0690*/0xE4, 0xF5, 0x3C, 0xF5, 0x3E, 0x90, 0x07, 0x19, + 0xF0, 0x12, 0x08, 0xD2, 0x70, 0x03, 0xF0, 0x80, +/*06A0*/0x03, 0x74, 0x01, 0xF0, 0x12, 0x08, 0x65, 0x75, + 0x83, 0xD0, 0xE0, 0x54, 0x0F, 0xFE, 0xAD, 0x3C, +/*06B0*/0x70, 0x02, 0x7E, 0x07, 0xBE, 0x0F, 0x02, 0x7E, + 0x80, 0xEE, 0xFB, 0xEF, 0xD3, 0x9B, 0x74, 0x80, +/*06C0*/0xF8, 0x98, 0x40, 0x1F, 0xE4, 0xF5, 0x3C, 0xF5, + 0x3E, 0x12, 0x08, 0xD2, 0x70, 0x03, 0xF0, 0x80, +/*06D0*/0x12, 0x74, 0x01, 0xF0, 0xE5, 0x08, 0xFB, 0xEB, + 0x44, 0x07, 0xF5, 0x82, 0x75, 0x83, 0xD2, 0xE0, +/*06E0*/0x44, 0x10, 0xF0, 0xE5, 0x08, 0xFB, 0xEB, 0x44, + 0x09, 0xF5, 0x82, 0x75, 0x83, 0x9E, 0xED, 0xF0, +/*06F0*/0xEB, 0x44, 0x07, 0xF5, 0x82, 0x75, 0x83, 0xCA, + 0xED, 0xF0, 0x12, 0x08, 0x65, 0x75, 0x83, 0xCC, +/*0700*/0xEF, 0xF0, 0x22, 0xE5, 0x08, 0x44, 0x07, 0xF5, + 0x82, 0x75, 0x83, 0xBC, 0xE0, 0x54, 0xF0, 0xF0, +/*0710*/0xE5, 0x08, 0x44, 0x07, 0xF5, 0x82, 0x75, 0x83, + 0xBE, 0xE0, 0x54, 0xF0, 0xF0, 0xE5, 0x08, 0x44, +/*0720*/0x07, 0xF5, 0x82, 0x75, 0x83, 0xC0, 0xE0, 0x54, + 0xF0, 0xF0, 0xE5, 0x08, 0x44, 0x07, 0xF5, 0x82, +/*0730*/0x22, 0xF0, 0x90, 0x07, 0x28, 0xE0, 0xFE, 0xA3, + 0xE0, 0xF5, 0x82, 0x8E, 0x83, 0x22, 0x85, 0x42, +/*0740*/0x42, 0x85, 0x41, 0x41, 0x85, 0x40, 0x40, 0x74, + 0xC0, 0x2F, 0xF5, 0x82, 0x74, 0x02, 0x3E, 0xF5, +/*0750*/0x83, 0xE5, 0x42, 0xF0, 0x74, 0xE0, 0x2F, 0xF5, + 0x82, 0x74, 0x02, 0x3E, 0xF5, 0x83, 0x22, 0xE5, +/*0760*/0x42, 0x29, 0xFD, 0xE4, 0x33, 0xFC, 0xE5, 0x3C, + 0xC3, 0x9D, 0xEC, 0x64, 0x80, 0xF8, 0x74, 0x80, +/*0770*/0x98, 0x22, 0xF5, 0x83, 0xE0, 0x90, 0x07, 0x22, + 0x54, 0x1F, 0xFD, 0xE0, 0xFA, 0xA3, 0xE0, 0xF5, +/*0780*/0x82, 0x8A, 0x83, 0xED, 0xF0, 0x22, 0x90, 0x07, + 0x22, 0xE0, 0xFC, 0xA3, 0xE0, 0xF5, 0x82, 0x8C, +/*0790*/0x83, 0x22, 0x90, 0x07, 0x24, 0xFF, 0xED, 0x44, + 0x07, 0xCF, 0xF0, 0xA3, 0xEF, 0xF0, 0x22, 0x85, +/*07A0*/0x38, 0x38, 0x85, 0x39, 0x39, 0x85, 0x3A, 0x3A, + 0x74, 0xC0, 0x2F, 0xF5, 0x82, 0x74, 0x02, 0x3E, +/*07B0*/0xF5, 0x83, 0x22, 0x90, 0x07, 0x26, 0xFF, 0xED, + 0x44, 0x07, 0xCF, 0xF0, 0xA3, 0xEF, 0xF0, 0x22, +/*07C0*/0xF0, 0x74, 0xA0, 0x2F, 0xF5, 0x82, 0x74, 0x02, + 0x3E, 0xF5, 0x83, 0x22, 0x74, 0xC0, 0x25, 0x11, +/*07D0*/0xF5, 0x82, 0xE4, 0x34, 0x01, 0xF5, 0x83, 0x22, + 0x74, 0x00, 0x25, 0x11, 0xF5, 0x82, 0xE4, 0x34, +/*07E0*/0x02, 0xF5, 0x83, 0x22, 0x74, 0x60, 0x25, 0x11, + 0xF5, 0x82, 0xE4, 0x34, 0x03, 0xF5, 0x83, 0x22, +/*07F0*/0x74, 0x80, 0x25, 0x11, 0xF5, 0x82, 0xE4, 0x34, + 0x03, 0xF5, 0x83, 0x22, 0x74, 0xE0, 0x25, 0x11, +/*0800*/0xF5, 0x82, 0xE4, 0x34, 0x03, 0xF5, 0x83, 0x22, + 0x74, 0x40, 0x25, 0x11, 0xF5, 0x82, 0xE4, 0x34, +/*0810*/0x06, 0xF5, 0x83, 0x22, 0x74, 0x80, 0x2F, 0xF5, + 0x82, 0x74, 0x02, 0x3E, 0xF5, 0x83, 0x22, 0xAF, +/*0820*/0x08, 0x7E, 0x00, 0xEF, 0x44, 0x07, 0xF5, 0x82, + 0x22, 0xF5, 0x83, 0xE5, 0x82, 0x44, 0x07, 0xF5, +/*0830*/0x82, 0xE5, 0x40, 0xF0, 0x22, 0x74, 0x40, 0x25, + 0x11, 0xF5, 0x82, 0xE4, 0x34, 0x02, 0xF5, 0x83, +/*0840*/0x22, 0x74, 0xC0, 0x25, 0x11, 0xF5, 0x82, 0xE4, + 0x34, 0x03, 0xF5, 0x83, 0x22, 0x74, 0x00, 0x25, +/*0850*/0x11, 0xF5, 0x82, 0xE4, 0x34, 0x06, 0xF5, 0x83, + 0x22, 0x74, 0x20, 0x25, 0x11, 0xF5, 0x82, 0xE4, +/*0860*/0x34, 0x06, 0xF5, 0x83, 0x22, 0xE5, 0x08, 0xFD, + 0xED, 0x44, 0x07, 0xF5, 0x82, 0x22, 0xE5, 0x41, +/*0870*/0xF0, 0xE5, 0x65, 0x64, 0x01, 0x45, 0x64, 0x22, + 0x7E, 0x00, 0xFB, 0x7A, 0x00, 0xFD, 0x7C, 0x00, +/*0880*/0x22, 0x74, 0x20, 0x25, 0x11, 0xF5, 0x82, 0xE4, + 0x34, 0x02, 0x22, 0x74, 0xA0, 0x25, 0x11, 0xF5, +/*0890*/0x82, 0xE4, 0x34, 0x03, 0x22, 0x85, 0x3E, 0x42, + 0x85, 0x3F, 0x41, 0x8F, 0x40, 0x22, 0x85, 0x3C, +/*08A0*/0x42, 0x85, 0x3D, 0x41, 0x8F, 0x40, 0x22, 0x75, + 0x45, 0x3F, 0x90, 0x07, 0x20, 0xE4, 0xF0, 0xA3, +/*08B0*/0x22, 0xF5, 0x83, 0xE5, 0x32, 0xF0, 0x05, 0x6E, + 0xE5, 0x6E, 0xC3, 0x94, 0x40, 0x22, 0xF0, 0xE5, +/*08C0*/0x08, 0x44, 0x06, 0xF5, 0x82, 0x22, 0x74, 0x00, + 0x25, 0x6E, 0xF5, 0x82, 0xE4, 0x34, 0x00, 0xF5, +/*08D0*/0x83, 0x22, 0xE5, 0x6D, 0x45, 0x6C, 0x90, 0x07, + 0x2F, 0x22, 0xE4, 0xF9, 0xE5, 0x3C, 0xD3, 0x95, +/*08E0*/0x3E, 0x22, 0x74, 0x80, 0x2E, 0xF5, 0x82, 0xE4, + 0x34, 0x02, 0xF5, 0x83, 0xE0, 0x22, 0x74, 0xA0, +/*08F0*/0x2E, 0xF5, 0x82, 0xE4, 0x34, 0x02, 0xF5, 0x83, + 0xE0, 0x22, 0x74, 0x80, 0x25, 0x6E, 0xF5, 0x82, +/*0900*/0xE4, 0x34, 0x00, 0x22, 0x25, 0x42, 0xFD, 0xE4, + 0x33, 0xFC, 0x22, 0x85, 0x42, 0x42, 0x85, 0x41, +/*0910*/0x41, 0x85, 0x40, 0x40, 0x22, 0xED, 0x4C, 0x60, + 0x03, 0x02, 0x09, 0xE5, 0xEF, 0x4E, 0x70, 0x37, +/*0920*/0x90, 0x07, 0x26, 0x12, 0x07, 0x89, 0xE0, 0xFD, + 0x12, 0x07, 0xCC, 0xED, 0xF0, 0x90, 0x07, 0x28, +/*0930*/0x12, 0x07, 0x89, 0xE0, 0xFD, 0x12, 0x07, 0xD8, + 0xED, 0xF0, 0x12, 0x07, 0x86, 0xE0, 0x54, 0x1F, +/*0940*/0xFD, 0x12, 0x08, 0x81, 0xF5, 0x83, 0xED, 0xF0, + 0x90, 0x07, 0x24, 0x12, 0x07, 0x89, 0xE0, 0x54, +/*0950*/0x1F, 0xFD, 0x12, 0x08, 0x35, 0xED, 0xF0, 0xEF, + 0x64, 0x04, 0x4E, 0x70, 0x37, 0x90, 0x07, 0x26, +/*0960*/0x12, 0x07, 0x89, 0xE0, 0xFD, 0x12, 0x07, 0xE4, + 0xED, 0xF0, 0x90, 0x07, 0x28, 0x12, 0x07, 0x89, +/*0970*/0xE0, 0xFD, 0x12, 0x07, 0xF0, 0xED, 0xF0, 0x12, + 0x07, 0x86, 0xE0, 0x54, 0x1F, 0xFD, 0x12, 0x08, +/*0980*/0x8B, 0xF5, 0x83, 0xED, 0xF0, 0x90, 0x07, 0x24, + 0x12, 0x07, 0x89, 0xE0, 0x54, 0x1F, 0xFD, 0x12, +/*0990*/0x08, 0x41, 0xED, 0xF0, 0xEF, 0x64, 0x01, 0x4E, + 0x70, 0x04, 0x7D, 0x01, 0x80, 0x02, 0x7D, 0x00, +/*09A0*/0xEF, 0x64, 0x02, 0x4E, 0x70, 0x04, 0x7F, 0x01, + 0x80, 0x02, 0x7F, 0x00, 0xEF, 0x4D, 0x60, 0x78, +/*09B0*/0x90, 0x07, 0x26, 0x12, 0x07, 0x35, 0xE0, 0xFF, + 0x12, 0x07, 0xFC, 0xEF, 0x12, 0x07, 0x31, 0xE0, +/*09C0*/0xFF, 0x12, 0x08, 0x08, 0xEF, 0xF0, 0x90, 0x07, + 0x22, 0x12, 0x07, 0x35, 0xE0, 0x54, 0x1F, 0xFF, +/*09D0*/0x12, 0x08, 0x4D, 0xEF, 0xF0, 0x90, 0x07, 0x24, + 0x12, 0x07, 0x35, 0xE0, 0x54, 0x1F, 0xFF, 0x12, +/*09E0*/0x08, 0x59, 0xEF, 0xF0, 0x22, 0x12, 0x07, 0xCC, + 0xE4, 0xF0, 0x12, 0x07, 0xD8, 0xE4, 0xF0, 0x12, +/*09F0*/0x08, 0x81, 0xF5, 0x83, 0xE4, 0xF0, 0x12, 0x08, + 0x35, 0x74, 0x14, 0xF0, 0x12, 0x07, 0xE4, 0xE4, +/*0A00*/0xF0, 0x12, 0x07, 0xF0, 0xE4, 0xF0, 0x12, 0x08, + 0x8B, 0xF5, 0x83, 0xE4, 0xF0, 0x12, 0x08, 0x41, +/*0A10*/0x74, 0x14, 0xF0, 0x12, 0x07, 0xFC, 0xE4, 0xF0, + 0x12, 0x08, 0x08, 0xE4, 0xF0, 0x12, 0x08, 0x4D, +/*0A20*/0xE4, 0xF0, 0x12, 0x08, 0x59, 0x74, 0x14, 0xF0, + 0x22, 0x53, 0xF9, 0xF7, 0x75, 0xFC, 0x10, 0xE4, +/*0A30*/0xF5, 0xFD, 0x75, 0xFE, 0x30, 0xF5, 0xFF, 0xE5, + 0xE7, 0x20, 0xE7, 0x03, 0x43, 0xF9, 0x08, 0xE5, +/*0A40*/0xE6, 0x20, 0xE7, 0x0B, 0x78, 0xFF, 0xE4, 0xF6, + 0xD8, 0xFD, 0x53, 0xE6, 0xFE, 0x80, 0x09, 0x78, +/*0A50*/0x08, 0xE4, 0xF6, 0xD8, 0xFD, 0x53, 0xE6, 0xFE, + 0x75, 0x81, 0x80, 0xE4, 0xF5, 0xA8, 0xD2, 0xA8, +/*0A60*/0xC2, 0xA9, 0xD2, 0xAF, 0xE5, 0xE2, 0x20, 0xE5, + 0x05, 0x20, 0xE6, 0x02, 0x80, 0x03, 0x43, 0xE1, +/*0A70*/0x02, 0xE5, 0xE2, 0x20, 0xE0, 0x0E, 0x90, 0x00, + 0x00, 0x7F, 0x00, 0x7E, 0x08, 0xE4, 0xF0, 0xA3, +/*0A80*/0xDF, 0xFC, 0xDE, 0xFA, 0x02, 0x0A, 0xDB, 0x43, + 0xFA, 0x01, 0xC0, 0xE0, 0xC0, 0xF0, 0xC0, 0x83, +/*0A90*/0xC0, 0x82, 0xC0, 0xD0, 0x12, 0x1C, 0xE7, 0xD0, + 0xD0, 0xD0, 0x82, 0xD0, 0x83, 0xD0, 0xF0, 0xD0, +/*0AA0*/0xE0, 0x53, 0xFA, 0xFE, 0x32, 0x02, 0x1B, 0x55, + 0xE4, 0x93, 0xA3, 0xF8, 0xE4, 0x93, 0xA3, 0xF6, +/*0AB0*/0x08, 0xDF, 0xF9, 0x80, 0x29, 0xE4, 0x93, 0xA3, + 0xF8, 0x54, 0x07, 0x24, 0x0C, 0xC8, 0xC3, 0x33, +/*0AC0*/0xC4, 0x54, 0x0F, 0x44, 0x20, 0xC8, 0x83, 0x40, + 0x04, 0xF4, 0x56, 0x80, 0x01, 0x46, 0xF6, 0xDF, +/*0AD0*/0xE4, 0x80, 0x0B, 0x01, 0x02, 0x04, 0x08, 0x10, + 0x20, 0x40, 0x80, 0x90, 0x00, 0x3F, 0xE4, 0x7E, +/*0AE0*/0x01, 0x93, 0x60, 0xC1, 0xA3, 0xFF, 0x54, 0x3F, + 0x30, 0xE5, 0x09, 0x54, 0x1F, 0xFE, 0xE4, 0x93, +/*0AF0*/0xA3, 0x60, 0x01, 0x0E, 0xCF, 0x54, 0xC0, 0x25, + 0xE0, 0x60, 0xAD, 0x40, 0xB8, 0x80, 0xFE, 0x8C, +/*0B00*/0x64, 0x8D, 0x65, 0x8A, 0x66, 0x8B, 0x67, 0xE4, + 0xF5, 0x69, 0xEF, 0x4E, 0x70, 0x03, 0x02, 0x1D, +/*0B10*/0x55, 0xE4, 0xF5, 0x68, 0xE5, 0x67, 0x45, 0x66, + 0x70, 0x32, 0x12, 0x07, 0x2A, 0x75, 0x83, 0x90, +/*0B20*/0xE4, 0x12, 0x07, 0x29, 0x75, 0x83, 0xC2, 0xE4, + 0x12, 0x07, 0x29, 0x75, 0x83, 0xC4, 0xE4, 0x12, +/*0B30*/0x08, 0x70, 0x70, 0x29, 0x12, 0x07, 0x2A, 0x75, + 0x83, 0x92, 0xE4, 0x12, 0x07, 0x29, 0x75, 0x83, +/*0B40*/0xC6, 0xE4, 0x12, 0x07, 0x29, 0x75, 0x83, 0xC8, + 0xE4, 0xF0, 0x80, 0x11, 0x90, 0x07, 0x26, 0x12, +/*0B50*/0x07, 0x35, 0xE4, 0x12, 0x08, 0x70, 0x70, 0x05, + 0x12, 0x07, 0x32, 0xE4, 0xF0, 0x12, 0x1D, 0x55, +/*0B60*/0x12, 0x1E, 0xBF, 0xE5, 0x67, 0x45, 0x66, 0x70, + 0x33, 0x12, 0x07, 0x2A, 0x75, 0x83, 0x90, 0xE5, +/*0B70*/0x41, 0x12, 0x07, 0x29, 0x75, 0x83, 0xC2, 0xE5, + 0x41, 0x12, 0x07, 0x29, 0x75, 0x83, 0xC4, 0x12, +/*0B80*/0x08, 0x6E, 0x70, 0x29, 0x12, 0x07, 0x2A, 0x75, + 0x83, 0x92, 0xE5, 0x40, 0x12, 0x07, 0x29, 0x75, +/*0B90*/0x83, 0xC6, 0xE5, 0x40, 0x12, 0x07, 0x29, 0x75, + 0x83, 0xC8, 0x80, 0x0E, 0x90, 0x07, 0x26, 0x12, +/*0BA0*/0x07, 0x35, 0x12, 0x08, 0x6E, 0x70, 0x06, 0x12, + 0x07, 0x32, 0xE5, 0x40, 0xF0, 0xAF, 0x69, 0x7E, +/*0BB0*/0x00, 0xAD, 0x67, 0xAC, 0x66, 0x12, 0x04, 0x44, + 0x12, 0x07, 0x2A, 0x75, 0x83, 0xCA, 0xE0, 0xD3, +/*0BC0*/0x94, 0x00, 0x50, 0x0C, 0x05, 0x68, 0xE5, 0x68, + 0xC3, 0x94, 0x05, 0x50, 0x03, 0x02, 0x0B, 0x14, +/*0BD0*/0x22, 0x8C, 0x60, 0x8D, 0x61, 0x12, 0x08, 0xDA, + 0x74, 0x20, 0x40, 0x0D, 0x2F, 0xF5, 0x82, 0x74, +/*0BE0*/0x03, 0x3E, 0xF5, 0x83, 0xE5, 0x3E, 0xF0, 0x80, + 0x0B, 0x2F, 0xF5, 0x82, 0x74, 0x03, 0x3E, 0xF5, +/*0BF0*/0x83, 0xE5, 0x3C, 0xF0, 0xE5, 0x3C, 0xD3, 0x95, + 0x3E, 0x40, 0x3C, 0xE5, 0x61, 0x45, 0x60, 0x70, +/*0C00*/0x10, 0xE9, 0x12, 0x09, 0x04, 0xE5, 0x3E, 0x12, + 0x07, 0x68, 0x40, 0x3B, 0x12, 0x08, 0x95, 0x80, +/*0C10*/0x18, 0xE5, 0x3E, 0xC3, 0x95, 0x38, 0x40, 0x1D, + 0x85, 0x3E, 0x38, 0xE5, 0x3E, 0x60, 0x05, 0x85, +/*0C20*/0x3F, 0x39, 0x80, 0x03, 0x85, 0x39, 0x39, 0x8F, + 0x3A, 0x12, 0x08, 0x14, 0xE5, 0x3E, 0x12, 0x07, +/*0C30*/0xC0, 0xE5, 0x3F, 0xF0, 0x22, 0x80, 0x43, 0xE5, + 0x61, 0x45, 0x60, 0x70, 0x19, 0x12, 0x07, 0x5F, +/*0C40*/0x40, 0x05, 0x12, 0x08, 0x9E, 0x80, 0x27, 0x12, + 0x09, 0x0B, 0x12, 0x08, 0x14, 0xE5, 0x42, 0x12, +/*0C50*/0x07, 0xC0, 0xE5, 0x41, 0xF0, 0x22, 0xE5, 0x3C, + 0xC3, 0x95, 0x38, 0x40, 0x1D, 0x85, 0x3C, 0x38, +/*0C60*/0xE5, 0x3C, 0x60, 0x05, 0x85, 0x3D, 0x39, 0x80, + 0x03, 0x85, 0x39, 0x39, 0x8F, 0x3A, 0x12, 0x08, +/*0C70*/0x14, 0xE5, 0x3C, 0x12, 0x07, 0xC0, 0xE5, 0x3D, + 0xF0, 0x22, 0x85, 0x38, 0x38, 0x85, 0x39, 0x39, +/*0C80*/0x85, 0x3A, 0x3A, 0x12, 0x08, 0x14, 0xE5, 0x38, + 0x12, 0x07, 0xC0, 0xE5, 0x39, 0xF0, 0x22, 0x7F, +/*0C90*/0x06, 0x12, 0x17, 0x31, 0x12, 0x1D, 0x23, 0x12, + 0x0E, 0x04, 0x12, 0x0E, 0x33, 0xE0, 0x44, 0x0A, +/*0CA0*/0xF0, 0x74, 0x8E, 0xFE, 0x12, 0x0E, 0x04, 0x12, + 0x0E, 0x0B, 0xEF, 0xF0, 0xE5, 0x28, 0x30, 0xE5, +/*0CB0*/0x03, 0xD3, 0x80, 0x01, 0xC3, 0x40, 0x05, 0x75, + 0x14, 0x20, 0x80, 0x03, 0x75, 0x14, 0x08, 0x12, +/*0CC0*/0x0E, 0x04, 0x75, 0x83, 0x8A, 0xE5, 0x14, 0xF0, + 0xB4, 0xFF, 0x05, 0x75, 0x12, 0x80, 0x80, 0x06, +/*0CD0*/0xE5, 0x14, 0xC3, 0x13, 0xF5, 0x12, 0xE4, 0xF5, + 0x16, 0xF5, 0x7F, 0x12, 0x19, 0x36, 0x12, 0x13, +/*0CE0*/0xA3, 0xE5, 0x0A, 0xC3, 0x94, 0x01, 0x50, 0x09, + 0x05, 0x16, 0xE5, 0x16, 0xC3, 0x94, 0x14, 0x40, +/*0CF0*/0xEA, 0xE5, 0xE4, 0x20, 0xE7, 0x28, 0x12, 0x0E, + 0x04, 0x75, 0x83, 0xD2, 0xE0, 0x54, 0x08, 0xD3, +/*0D00*/0x94, 0x00, 0x40, 0x04, 0x7F, 0x01, 0x80, 0x02, + 0x7F, 0x00, 0xE5, 0x0A, 0xC3, 0x94, 0x01, 0x40, +/*0D10*/0x04, 0x7E, 0x01, 0x80, 0x02, 0x7E, 0x00, 0xEF, + 0x5E, 0x60, 0x03, 0x12, 0x1D, 0xD7, 0xE5, 0x7F, +/*0D20*/0xC3, 0x94, 0x11, 0x40, 0x14, 0x12, 0x0E, 0x04, + 0x75, 0x83, 0xD2, 0xE0, 0x44, 0x80, 0xF0, 0xE5, +/*0D30*/0xE4, 0x20, 0xE7, 0x0F, 0x12, 0x1D, 0xD7, 0x80, + 0x0A, 0x12, 0x0E, 0x04, 0x75, 0x83, 0xD2, 0xE0, +/*0D40*/0x54, 0x7F, 0xF0, 0x12, 0x1D, 0x23, 0x22, 0x74, + 0x8A, 0x85, 0x08, 0x82, 0xF5, 0x83, 0xE5, 0x17, +/*0D50*/0xF0, 0x12, 0x0E, 0x3A, 0xE4, 0xF0, 0x90, 0x07, + 0x02, 0xE0, 0x12, 0x0E, 0x17, 0x75, 0x83, 0x90, +/*0D60*/0xEF, 0xF0, 0x74, 0x92, 0xFE, 0xE5, 0x08, 0x44, + 0x07, 0xFF, 0xF5, 0x82, 0x8E, 0x83, 0xE0, 0x54, +/*0D70*/0xC0, 0xFD, 0x90, 0x07, 0x03, 0xE0, 0x54, 0x3F, + 0x4D, 0x8F, 0x82, 0x8E, 0x83, 0xF0, 0x90, 0x07, +/*0D80*/0x04, 0xE0, 0x12, 0x0E, 0x17, 0x75, 0x83, 0x82, + 0xEF, 0xF0, 0x90, 0x07, 0x05, 0xE0, 0xFF, 0xED, +/*0D90*/0x44, 0x07, 0xF5, 0x82, 0x75, 0x83, 0xB4, 0xEF, + 0x12, 0x0E, 0x03, 0x75, 0x83, 0x80, 0xE0, 0x54, +/*0DA0*/0xBF, 0xF0, 0x30, 0x37, 0x0A, 0x12, 0x0E, 0x91, + 0x75, 0x83, 0x94, 0xE0, 0x44, 0x80, 0xF0, 0x30, +/*0DB0*/0x38, 0x0A, 0x12, 0x0E, 0x91, 0x75, 0x83, 0x92, + 0xE0, 0x44, 0x80, 0xF0, 0xE5, 0x28, 0x30, 0xE4, +/*0DC0*/0x1A, 0x20, 0x39, 0x0A, 0x12, 0x0E, 0x04, 0x75, + 0x83, 0x88, 0xE0, 0x54, 0x7F, 0xF0, 0x20, 0x3A, +/*0DD0*/0x0A, 0x12, 0x0E, 0x04, 0x75, 0x83, 0x88, 0xE0, + 0x54, 0xBF, 0xF0, 0x74, 0x8C, 0xFE, 0x12, 0x0E, +/*0DE0*/0x04, 0x8E, 0x83, 0xE0, 0x54, 0x0F, 0x12, 0x0E, + 0x03, 0x75, 0x83, 0x86, 0xE0, 0x54, 0xBF, 0xF0, +/*0DF0*/0xE5, 0x08, 0x44, 0x06, 0x12, 0x0D, 0xFD, 0x75, + 0x83, 0x8A, 0xE4, 0xF0, 0x22, 0xF5, 0x82, 0x75, +/*0E00*/0x83, 0x82, 0xE4, 0xF0, 0xE5, 0x08, 0x44, 0x07, + 0xF5, 0x82, 0x22, 0x8E, 0x83, 0xE0, 0xF5, 0x10, +/*0E10*/0x54, 0xFE, 0xF0, 0xE5, 0x10, 0x44, 0x01, 0xFF, + 0xE5, 0x08, 0xFD, 0xED, 0x44, 0x07, 0xF5, 0x82, +/*0E20*/0x22, 0xE5, 0x15, 0xC4, 0x54, 0x07, 0xFF, 0xE5, + 0x08, 0xFD, 0xED, 0x44, 0x08, 0xF5, 0x82, 0x75, +/*0E30*/0x83, 0x82, 0x22, 0x75, 0x83, 0x80, 0xE0, 0x44, + 0x40, 0xF0, 0xE5, 0x08, 0x44, 0x08, 0xF5, 0x82, +/*0E40*/0x75, 0x83, 0x8A, 0x22, 0xE5, 0x16, 0x25, 0xE0, + 0x25, 0xE0, 0x24, 0xAF, 0xF5, 0x82, 0xE4, 0x34, +/*0E50*/0x1A, 0xF5, 0x83, 0xE4, 0x93, 0xF5, 0x0D, 0x22, + 0x43, 0xE1, 0x10, 0x43, 0xE1, 0x80, 0x53, 0xE1, +/*0E60*/0xFD, 0x85, 0xE1, 0x10, 0x22, 0xE5, 0x16, 0x25, + 0xE0, 0x25, 0xE0, 0x24, 0xB2, 0xF5, 0x82, 0xE4, +/*0E70*/0x34, 0x1A, 0xF5, 0x83, 0xE4, 0x93, 0x22, 0x85, + 0x55, 0x82, 0x85, 0x54, 0x83, 0xE5, 0x15, 0xF0, +/*0E80*/0x22, 0xE5, 0xE2, 0x54, 0x20, 0xD3, 0x94, 0x00, + 0x22, 0xE5, 0xE2, 0x54, 0x40, 0xD3, 0x94, 0x00, +/*0E90*/0x22, 0xE5, 0x08, 0x44, 0x06, 0xF5, 0x82, 0x22, + 0xFD, 0xE5, 0x08, 0xFB, 0xEB, 0x44, 0x07, 0xF5, +/*0EA0*/0x82, 0x22, 0x53, 0xF9, 0xF7, 0x75, 0xFE, 0x30, + 0x22, 0xEF, 0x4E, 0x70, 0x26, 0x12, 0x07, 0xCC, +/*0EB0*/0xE0, 0xFD, 0x90, 0x07, 0x26, 0x12, 0x07, 0x7B, + 0x12, 0x07, 0xD8, 0xE0, 0xFD, 0x90, 0x07, 0x28, +/*0EC0*/0x12, 0x07, 0x7B, 0x12, 0x08, 0x81, 0x12, 0x07, + 0x72, 0x12, 0x08, 0x35, 0xE0, 0x90, 0x07, 0x24, +/*0ED0*/0x12, 0x07, 0x78, 0xEF, 0x64, 0x04, 0x4E, 0x70, + 0x29, 0x12, 0x07, 0xE4, 0xE0, 0xFD, 0x90, 0x07, +/*0EE0*/0x26, 0x12, 0x07, 0x7B, 0x12, 0x07, 0xF0, 0xE0, + 0xFD, 0x90, 0x07, 0x28, 0x12, 0x07, 0x7B, 0x12, +/*0EF0*/0x08, 0x8B, 0x12, 0x07, 0x72, 0x12, 0x08, 0x41, + 0xE0, 0x54, 0x1F, 0xFD, 0x90, 0x07, 0x24, 0x12, +/*0F00*/0x07, 0x7B, 0xEF, 0x64, 0x01, 0x4E, 0x70, 0x04, + 0x7D, 0x01, 0x80, 0x02, 0x7D, 0x00, 0xEF, 0x64, +/*0F10*/0x02, 0x4E, 0x70, 0x04, 0x7F, 0x01, 0x80, 0x02, + 0x7F, 0x00, 0xEF, 0x4D, 0x60, 0x35, 0x12, 0x07, +/*0F20*/0xFC, 0xE0, 0xFF, 0x90, 0x07, 0x26, 0x12, 0x07, + 0x89, 0xEF, 0xF0, 0x12, 0x08, 0x08, 0xE0, 0xFF, +/*0F30*/0x90, 0x07, 0x28, 0x12, 0x07, 0x89, 0xEF, 0xF0, + 0x12, 0x08, 0x4D, 0xE0, 0x54, 0x1F, 0xFF, 0x12, +/*0F40*/0x07, 0x86, 0xEF, 0xF0, 0x12, 0x08, 0x59, 0xE0, + 0x54, 0x1F, 0xFF, 0x90, 0x07, 0x24, 0x12, 0x07, +/*0F50*/0x89, 0xEF, 0xF0, 0x22, 0xE4, 0xF5, 0x53, 0x12, + 0x0E, 0x81, 0x40, 0x04, 0x7F, 0x01, 0x80, 0x02, +/*0F60*/0x7F, 0x00, 0x12, 0x0E, 0x89, 0x40, 0x04, 0x7E, + 0x01, 0x80, 0x02, 0x7E, 0x00, 0xEE, 0x4F, 0x70, +/*0F70*/0x03, 0x02, 0x0F, 0xF6, 0x85, 0xE1, 0x10, 0x43, + 0xE1, 0x02, 0x53, 0xE1, 0x0F, 0x85, 0xE1, 0x10, +/*0F80*/0xE4, 0xF5, 0x51, 0xE5, 0xE3, 0x54, 0x3F, 0xF5, + 0x52, 0x12, 0x0E, 0x89, 0x40, 0x1D, 0xAD, 0x52, +/*0F90*/0xAF, 0x51, 0x12, 0x11, 0x18, 0xEF, 0x60, 0x08, + 0x85, 0xE1, 0x10, 0x43, 0xE1, 0x40, 0x80, 0x0B, +/*0FA0*/0x53, 0xE1, 0xBF, 0x12, 0x0E, 0x58, 0x12, 0x00, + 0x06, 0x80, 0xFB, 0xE5, 0xE3, 0x54, 0x3F, 0xF5, +/*0FB0*/0x51, 0xE5, 0xE4, 0x54, 0x3F, 0xF5, 0x52, 0x12, + 0x0E, 0x81, 0x40, 0x1D, 0xAD, 0x52, 0xAF, 0x51, +/*0FC0*/0x12, 0x11, 0x18, 0xEF, 0x60, 0x08, 0x85, 0xE1, + 0x10, 0x43, 0xE1, 0x20, 0x80, 0x0B, 0x53, 0xE1, +/*0FD0*/0xDF, 0x12, 0x0E, 0x58, 0x12, 0x00, 0x06, 0x80, + 0xFB, 0x12, 0x0E, 0x81, 0x40, 0x04, 0x7F, 0x01, +/*0FE0*/0x80, 0x02, 0x7F, 0x00, 0x12, 0x0E, 0x89, 0x40, + 0x04, 0x7E, 0x01, 0x80, 0x02, 0x7E, 0x00, 0xEE, +/*0FF0*/0x4F, 0x60, 0x03, 0x12, 0x0E, 0x5B, 0x22, 0x12, + 0x0E, 0x21, 0xEF, 0xF0, 0x12, 0x10, 0x91, 0x22, +/*1000*/0x02, 0x11, 0x00, 0x02, 0x10, 0x40, 0x02, 0x10, + 0x90, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, +/*1010*/0x01, 0x20, 0x01, 0x20, 0xE4, 0xF5, 0x57, 0x12, + 0x16, 0xBD, 0x12, 0x16, 0x44, 0xE4, 0x12, 0x10, +/*1020*/0x56, 0x12, 0x14, 0xB7, 0x90, 0x07, 0x26, 0x12, + 0x07, 0x35, 0xE4, 0x12, 0x07, 0x31, 0xE4, 0xF0, +/*1030*/0x12, 0x10, 0x56, 0x12, 0x14, 0xB7, 0x90, 0x07, + 0x26, 0x12, 0x07, 0x35, 0xE5, 0x41, 0x12, 0x07, +/*1040*/0x31, 0xE5, 0x40, 0xF0, 0xAF, 0x57, 0x7E, 0x00, + 0xAD, 0x56, 0x7C, 0x00, 0x12, 0x04, 0x44, 0xAF, +/*1050*/0x56, 0x7E, 0x00, 0x02, 0x11, 0xEE, 0xFF, 0x90, + 0x07, 0x20, 0xA3, 0xE0, 0xFD, 0xE4, 0xF5, 0x56, +/*1060*/0xF5, 0x40, 0xFE, 0xFC, 0xAB, 0x56, 0xFA, 0x12, + 0x11, 0x51, 0x7F, 0x0F, 0x7D, 0x18, 0xE4, 0xF5, +/*1070*/0x56, 0xF5, 0x40, 0xFE, 0xFC, 0xAB, 0x56, 0xFA, + 0x12, 0x15, 0x41, 0xAF, 0x56, 0x7E, 0x00, 0x12, +/*1080*/0x1A, 0xFF, 0xE4, 0xFF, 0xF5, 0x56, 0x7D, 0x1F, + 0xF5, 0x40, 0xFE, 0xFC, 0xAB, 0x56, 0xFA, 0x22, +/*1090*/0x22, 0xE4, 0xF5, 0x55, 0xE5, 0x08, 0xFD, 0x74, + 0xA0, 0xF5, 0x56, 0xED, 0x44, 0x07, 0xF5, 0x57, +/*10A0*/0xE5, 0x28, 0x30, 0xE5, 0x03, 0xD3, 0x80, 0x01, + 0xC3, 0x40, 0x05, 0x7F, 0x28, 0xEF, 0x80, 0x04, +/*10B0*/0x7F, 0x14, 0xEF, 0xC3, 0x13, 0xF5, 0x54, 0xE4, + 0xF9, 0x12, 0x0E, 0x18, 0x75, 0x83, 0x8E, 0xE0, +/*10C0*/0xF5, 0x10, 0xCE, 0xEF, 0xCE, 0xEE, 0xD3, 0x94, + 0x00, 0x40, 0x26, 0xE5, 0x10, 0x54, 0xFE, 0x12, +/*10D0*/0x0E, 0x98, 0x75, 0x83, 0x8E, 0xED, 0xF0, 0xE5, + 0x10, 0x44, 0x01, 0xFD, 0xEB, 0x44, 0x07, 0xF5, +/*10E0*/0x82, 0xED, 0xF0, 0x85, 0x57, 0x82, 0x85, 0x56, + 0x83, 0xE0, 0x30, 0xE3, 0x01, 0x09, 0x1E, 0x80, +/*10F0*/0xD4, 0xC2, 0x34, 0xE9, 0xC3, 0x95, 0x54, 0x40, + 0x02, 0xD2, 0x34, 0x22, 0x02, 0x00, 0x06, 0x22, +/*1100*/0x30, 0x30, 0x11, 0x90, 0x10, 0x00, 0xE4, 0x93, + 0xF5, 0x10, 0x90, 0x10, 0x10, 0xE4, 0x93, 0xF5, +/*1110*/0x10, 0x12, 0x10, 0x90, 0x12, 0x11, 0x50, 0x22, + 0xE4, 0xFC, 0xC3, 0xED, 0x9F, 0xFA, 0xEF, 0xF5, +/*1120*/0x83, 0x75, 0x82, 0x00, 0x79, 0xFF, 0xE4, 0x93, + 0xCC, 0x6C, 0xCC, 0xA3, 0xD9, 0xF8, 0xDA, 0xF6, +/*1130*/0xE5, 0xE2, 0x30, 0xE4, 0x02, 0x8C, 0xE5, 0xED, + 0x24, 0xFF, 0xFF, 0xEF, 0x75, 0x82, 0xFF, 0xF5, +/*1140*/0x83, 0xE4, 0x93, 0x6C, 0x70, 0x03, 0x7F, 0x01, + 0x22, 0x7F, 0x00, 0x22, 0x22, 0x11, 0x00, 0x00, +/*1150*/0x22, 0x8E, 0x58, 0x8F, 0x59, 0x8C, 0x5A, 0x8D, + 0x5B, 0x8A, 0x5C, 0x8B, 0x5D, 0x75, 0x5E, 0x01, +/*1160*/0xE4, 0xF5, 0x5F, 0xF5, 0x60, 0xF5, 0x62, 0x12, + 0x07, 0x2A, 0x75, 0x83, 0xD0, 0xE0, 0xFF, 0xC4, +/*1170*/0x54, 0x0F, 0xF5, 0x61, 0x12, 0x1E, 0xA5, 0x85, + 0x59, 0x5E, 0xD3, 0xE5, 0x5E, 0x95, 0x5B, 0xE5, +/*1180*/0x5A, 0x12, 0x07, 0x6B, 0x50, 0x4B, 0x12, 0x07, + 0x03, 0x75, 0x83, 0xBC, 0xE0, 0x45, 0x5E, 0x12, +/*1190*/0x07, 0x29, 0x75, 0x83, 0xBE, 0xE0, 0x45, 0x5E, + 0x12, 0x07, 0x29, 0x75, 0x83, 0xC0, 0xE0, 0x45, +/*11A0*/0x5E, 0xF0, 0xAF, 0x5F, 0xE5, 0x60, 0x12, 0x08, + 0x78, 0x12, 0x0A, 0xFF, 0xAF, 0x62, 0x7E, 0x00, +/*11B0*/0xAD, 0x5D, 0xAC, 0x5C, 0x12, 0x04, 0x44, 0xE5, + 0x61, 0xAF, 0x5E, 0x7E, 0x00, 0xB4, 0x03, 0x05, +/*11C0*/0x12, 0x1E, 0x21, 0x80, 0x07, 0xAD, 0x5D, 0xAC, + 0x5C, 0x12, 0x13, 0x17, 0x05, 0x5E, 0x02, 0x11, +/*11D0*/0x7A, 0x12, 0x07, 0x03, 0x75, 0x83, 0xBC, 0xE0, + 0x45, 0x40, 0x12, 0x07, 0x29, 0x75, 0x83, 0xBE, +/*11E0*/0xE0, 0x45, 0x40, 0x12, 0x07, 0x29, 0x75, 0x83, + 0xC0, 0xE0, 0x45, 0x40, 0xF0, 0x22, 0x8E, 0x58, +/*11F0*/0x8F, 0x59, 0x75, 0x5A, 0x01, 0x79, 0x01, 0x75, + 0x5B, 0x01, 0xE4, 0xFB, 0x12, 0x07, 0x2A, 0x75, +/*1200*/0x83, 0xAE, 0xE0, 0x54, 0x1A, 0xFF, 0x12, 0x08, + 0x65, 0xE0, 0xC4, 0x13, 0x54, 0x07, 0xFE, 0xEF, +/*1210*/0x70, 0x0C, 0xEE, 0x65, 0x35, 0x70, 0x07, 0x90, + 0x07, 0x2F, 0xE0, 0xB4, 0x01, 0x0D, 0xAF, 0x35, +/*1220*/0x7E, 0x00, 0x12, 0x0E, 0xA9, 0xCF, 0xEB, 0xCF, + 0x02, 0x1E, 0x60, 0xE5, 0x59, 0x64, 0x02, 0x45, +/*1230*/0x58, 0x70, 0x04, 0x7F, 0x01, 0x80, 0x02, 0x7F, + 0x00, 0xE5, 0x59, 0x45, 0x58, 0x70, 0x04, 0x7E, +/*1240*/0x01, 0x80, 0x02, 0x7E, 0x00, 0xEE, 0x4F, 0x60, + 0x23, 0x85, 0x41, 0x49, 0x85, 0x40, 0x4B, 0xE5, +/*1250*/0x59, 0x45, 0x58, 0x70, 0x2C, 0xAF, 0x5A, 0xFE, + 0xCD, 0xE9, 0xCD, 0xFC, 0xAB, 0x59, 0xAA, 0x58, +/*1260*/0x12, 0x0A, 0xFF, 0xAF, 0x5B, 0x7E, 0x00, 0x12, + 0x1E, 0x60, 0x80, 0x15, 0xAF, 0x5B, 0x7E, 0x00, +/*1270*/0x12, 0x1E, 0x60, 0x90, 0x07, 0x26, 0x12, 0x07, + 0x35, 0xE5, 0x49, 0x12, 0x07, 0x31, 0xE5, 0x4B, +/*1280*/0xF0, 0xE4, 0xFD, 0xAF, 0x35, 0xFE, 0xFC, 0x12, + 0x09, 0x15, 0x22, 0x8C, 0x64, 0x8D, 0x65, 0x12, +/*1290*/0x08, 0xDA, 0x40, 0x3C, 0xE5, 0x65, 0x45, 0x64, + 0x70, 0x10, 0x12, 0x09, 0x04, 0xC3, 0xE5, 0x3E, +/*12A0*/0x12, 0x07, 0x69, 0x40, 0x3B, 0x12, 0x08, 0x95, + 0x80, 0x18, 0xE5, 0x3E, 0xC3, 0x95, 0x38, 0x40, +/*12B0*/0x1D, 0x85, 0x3E, 0x38, 0xE5, 0x3E, 0x60, 0x05, + 0x85, 0x3F, 0x39, 0x80, 0x03, 0x85, 0x39, 0x39, +/*12C0*/0x8F, 0x3A, 0x12, 0x07, 0xA8, 0xE5, 0x3E, 0x12, + 0x07, 0x53, 0xE5, 0x3F, 0xF0, 0x22, 0x80, 0x3B, +/*12D0*/0xE5, 0x65, 0x45, 0x64, 0x70, 0x11, 0x12, 0x07, + 0x5F, 0x40, 0x05, 0x12, 0x08, 0x9E, 0x80, 0x1F, +/*12E0*/0x12, 0x07, 0x3E, 0xE5, 0x41, 0xF0, 0x22, 0xE5, + 0x3C, 0xC3, 0x95, 0x38, 0x40, 0x1D, 0x85, 0x3C, +/*12F0*/0x38, 0xE5, 0x3C, 0x60, 0x05, 0x85, 0x3D, 0x39, + 0x80, 0x03, 0x85, 0x39, 0x39, 0x8F, 0x3A, 0x12, +/*1300*/0x07, 0xA8, 0xE5, 0x3C, 0x12, 0x07, 0x53, 0xE5, + 0x3D, 0xF0, 0x22, 0x12, 0x07, 0x9F, 0xE5, 0x38, +/*1310*/0x12, 0x07, 0x53, 0xE5, 0x39, 0xF0, 0x22, 0x8C, + 0x63, 0x8D, 0x64, 0x12, 0x08, 0xDA, 0x40, 0x3C, +/*1320*/0xE5, 0x64, 0x45, 0x63, 0x70, 0x10, 0x12, 0x09, + 0x04, 0xC3, 0xE5, 0x3E, 0x12, 0x07, 0x69, 0x40, +/*1330*/0x3B, 0x12, 0x08, 0x95, 0x80, 0x18, 0xE5, 0x3E, + 0xC3, 0x95, 0x38, 0x40, 0x1D, 0x85, 0x3E, 0x38, +/*1340*/0xE5, 0x3E, 0x60, 0x05, 0x85, 0x3F, 0x39, 0x80, + 0x03, 0x85, 0x39, 0x39, 0x8F, 0x3A, 0x12, 0x07, +/*1350*/0xA8, 0xE5, 0x3E, 0x12, 0x07, 0x53, 0xE5, 0x3F, + 0xF0, 0x22, 0x80, 0x3B, 0xE5, 0x64, 0x45, 0x63, +/*1360*/0x70, 0x11, 0x12, 0x07, 0x5F, 0x40, 0x05, 0x12, + 0x08, 0x9E, 0x80, 0x1F, 0x12, 0x07, 0x3E, 0xE5, +/*1370*/0x41, 0xF0, 0x22, 0xE5, 0x3C, 0xC3, 0x95, 0x38, + 0x40, 0x1D, 0x85, 0x3C, 0x38, 0xE5, 0x3C, 0x60, +/*1380*/0x05, 0x85, 0x3D, 0x39, 0x80, 0x03, 0x85, 0x39, + 0x39, 0x8F, 0x3A, 0x12, 0x07, 0xA8, 0xE5, 0x3C, +/*1390*/0x12, 0x07, 0x53, 0xE5, 0x3D, 0xF0, 0x22, 0x12, + 0x07, 0x9F, 0xE5, 0x38, 0x12, 0x07, 0x53, 0xE5, +/*13A0*/0x39, 0xF0, 0x22, 0xE5, 0x0D, 0xFE, 0xE5, 0x08, + 0x8E, 0x54, 0x44, 0x05, 0xF5, 0x55, 0x75, 0x15, +/*13B0*/0x0F, 0xF5, 0x82, 0x12, 0x0E, 0x7A, 0x12, 0x17, + 0xA3, 0x20, 0x31, 0x05, 0x75, 0x15, 0x03, 0x80, +/*13C0*/0x03, 0x75, 0x15, 0x0B, 0xE5, 0x0A, 0xC3, 0x94, + 0x01, 0x50, 0x38, 0x12, 0x14, 0x20, 0x20, 0x31, +/*13D0*/0x06, 0x05, 0x15, 0x05, 0x15, 0x80, 0x04, 0x15, + 0x15, 0x15, 0x15, 0xE5, 0x0A, 0xC3, 0x94, 0x01, +/*13E0*/0x50, 0x21, 0x12, 0x14, 0x20, 0x20, 0x31, 0x04, + 0x05, 0x15, 0x80, 0x02, 0x15, 0x15, 0xE5, 0x0A, +/*13F0*/0xC3, 0x94, 0x01, 0x50, 0x0E, 0x12, 0x0E, 0x77, + 0x12, 0x17, 0xA3, 0x20, 0x31, 0x05, 0x05, 0x15, +/*1400*/0x12, 0x0E, 0x77, 0xE5, 0x15, 0xB4, 0x08, 0x04, + 0x7F, 0x01, 0x80, 0x02, 0x7F, 0x00, 0xE5, 0x15, +/*1410*/0xB4, 0x07, 0x04, 0x7E, 0x01, 0x80, 0x02, 0x7E, + 0x00, 0xEE, 0x4F, 0x60, 0x02, 0x05, 0x7F, 0x22, +/*1420*/0x85, 0x55, 0x82, 0x85, 0x54, 0x83, 0xE5, 0x15, + 0xF0, 0x12, 0x17, 0xA3, 0x22, 0x12, 0x07, 0x2A, +/*1430*/0x75, 0x83, 0xAE, 0x74, 0xFF, 0x12, 0x07, 0x29, + 0xE0, 0x54, 0x1A, 0xF5, 0x34, 0xE0, 0xC4, 0x13, +/*1440*/0x54, 0x07, 0xF5, 0x35, 0x24, 0xFE, 0x60, 0x24, + 0x24, 0xFE, 0x60, 0x3C, 0x24, 0x04, 0x70, 0x63, +/*1450*/0x75, 0x31, 0x2D, 0xE5, 0x08, 0xFD, 0x74, 0xB6, + 0x12, 0x07, 0x92, 0x74, 0xBC, 0x90, 0x07, 0x22, +/*1460*/0x12, 0x07, 0x95, 0x74, 0x90, 0x12, 0x07, 0xB3, + 0x74, 0x92, 0x80, 0x3C, 0x75, 0x31, 0x3A, 0xE5, +/*1470*/0x08, 0xFD, 0x74, 0xBA, 0x12, 0x07, 0x92, 0x74, + 0xC0, 0x90, 0x07, 0x22, 0x12, 0x07, 0xB6, 0x74, +/*1480*/0xC4, 0x12, 0x07, 0xB3, 0x74, 0xC8, 0x80, 0x20, + 0x75, 0x31, 0x35, 0xE5, 0x08, 0xFD, 0x74, 0xB8, +/*1490*/0x12, 0x07, 0x92, 0x74, 0xBE, 0xFF, 0xED, 0x44, + 0x07, 0x90, 0x07, 0x22, 0xCF, 0xF0, 0xA3, 0xEF, +/*14A0*/0xF0, 0x74, 0xC2, 0x12, 0x07, 0xB3, 0x74, 0xC6, + 0xFF, 0xED, 0x44, 0x07, 0xA3, 0xCF, 0xF0, 0xA3, +/*14B0*/0xEF, 0xF0, 0x22, 0x75, 0x34, 0x01, 0x22, 0x8E, + 0x58, 0x8F, 0x59, 0x8C, 0x5A, 0x8D, 0x5B, 0x8A, +/*14C0*/0x5C, 0x8B, 0x5D, 0x75, 0x5E, 0x01, 0xE4, 0xF5, + 0x5F, 0x12, 0x1E, 0xA5, 0x85, 0x59, 0x5E, 0xD3, +/*14D0*/0xE5, 0x5E, 0x95, 0x5B, 0xE5, 0x5A, 0x12, 0x07, + 0x6B, 0x50, 0x57, 0xE5, 0x5D, 0x45, 0x5C, 0x70, +/*14E0*/0x30, 0x12, 0x07, 0x2A, 0x75, 0x83, 0x92, 0xE5, + 0x5E, 0x12, 0x07, 0x29, 0x75, 0x83, 0xC6, 0xE5, +/*14F0*/0x5E, 0x12, 0x07, 0x29, 0x75, 0x83, 0xC8, 0xE5, + 0x5E, 0x12, 0x07, 0x29, 0x75, 0x83, 0x90, 0xE5, +/*1500*/0x5E, 0x12, 0x07, 0x29, 0x75, 0x83, 0xC2, 0xE5, + 0x5E, 0x12, 0x07, 0x29, 0x75, 0x83, 0xC4, 0x80, +/*1510*/0x03, 0x12, 0x07, 0x32, 0xE5, 0x5E, 0xF0, 0xAF, + 0x5F, 0x7E, 0x00, 0xAD, 0x5D, 0xAC, 0x5C, 0x12, +/*1520*/0x04, 0x44, 0xAF, 0x5E, 0x7E, 0x00, 0xAD, 0x5D, + 0xAC, 0x5C, 0x12, 0x0B, 0xD1, 0x05, 0x5E, 0x02, +/*1530*/0x14, 0xCF, 0xAB, 0x5D, 0xAA, 0x5C, 0xAD, 0x5B, + 0xAC, 0x5A, 0xAF, 0x59, 0xAE, 0x58, 0x02, 0x1B, +/*1540*/0xFB, 0x8C, 0x5C, 0x8D, 0x5D, 0x8A, 0x5E, 0x8B, + 0x5F, 0x75, 0x60, 0x01, 0xE4, 0xF5, 0x61, 0xF5, +/*1550*/0x62, 0xF5, 0x63, 0x12, 0x1E, 0xA5, 0x8F, 0x60, + 0xD3, 0xE5, 0x60, 0x95, 0x5D, 0xE5, 0x5C, 0x12, +/*1560*/0x07, 0x6B, 0x50, 0x61, 0xE5, 0x5F, 0x45, 0x5E, + 0x70, 0x27, 0x12, 0x07, 0x2A, 0x75, 0x83, 0xB6, +/*1570*/0xE5, 0x60, 0x12, 0x07, 0x29, 0x75, 0x83, 0xB8, + 0xE5, 0x60, 0x12, 0x07, 0x29, 0x75, 0x83, 0xBA, +/*1580*/0xE5, 0x60, 0xF0, 0xAF, 0x61, 0x7E, 0x00, 0xE5, + 0x62, 0x12, 0x08, 0x7A, 0x12, 0x0A, 0xFF, 0x80, +/*1590*/0x19, 0x90, 0x07, 0x24, 0x12, 0x07, 0x35, 0xE5, + 0x60, 0x12, 0x07, 0x29, 0x75, 0x83, 0x8E, 0xE4, +/*15A0*/0x12, 0x07, 0x29, 0x74, 0x01, 0x12, 0x07, 0x29, + 0xE4, 0xF0, 0xAF, 0x63, 0x7E, 0x00, 0xAD, 0x5F, +/*15B0*/0xAC, 0x5E, 0x12, 0x04, 0x44, 0xAF, 0x60, 0x7E, + 0x00, 0xAD, 0x5F, 0xAC, 0x5E, 0x12, 0x12, 0x8B, +/*15C0*/0x05, 0x60, 0x02, 0x15, 0x58, 0x22, 0x90, 0x11, + 0x4D, 0xE4, 0x93, 0x90, 0x07, 0x2E, 0xF0, 0x12, +/*15D0*/0x08, 0x1F, 0x75, 0x83, 0xAE, 0xE0, 0x54, 0x1A, + 0xF5, 0x34, 0x70, 0x67, 0xEF, 0x44, 0x07, 0xF5, +/*15E0*/0x82, 0x75, 0x83, 0xCE, 0xE0, 0xFF, 0x13, 0x13, + 0x13, 0x54, 0x07, 0xF5, 0x36, 0x54, 0x0F, 0xD3, +/*15F0*/0x94, 0x00, 0x40, 0x06, 0x12, 0x14, 0x2D, 0x12, + 0x1B, 0xA9, 0xE5, 0x36, 0x54, 0x0F, 0x24, 0xFE, +/*1600*/0x60, 0x0C, 0x14, 0x60, 0x0C, 0x14, 0x60, 0x19, + 0x24, 0x03, 0x70, 0x37, 0x80, 0x10, 0x02, 0x1E, +/*1610*/0x91, 0x12, 0x1E, 0x91, 0x12, 0x07, 0x2A, 0x75, + 0x83, 0xCE, 0xE0, 0x54, 0xEF, 0xF0, 0x02, 0x1D, +/*1620*/0xAE, 0x12, 0x10, 0x14, 0xE4, 0xF5, 0x55, 0x12, + 0x1D, 0x85, 0x05, 0x55, 0xE5, 0x55, 0xC3, 0x94, +/*1630*/0x05, 0x40, 0xF4, 0x12, 0x07, 0x2A, 0x75, 0x83, + 0xCE, 0xE0, 0x54, 0xC7, 0x12, 0x07, 0x29, 0xE0, +/*1640*/0x44, 0x08, 0xF0, 0x22, 0xE4, 0xF5, 0x58, 0xF5, + 0x59, 0xAF, 0x08, 0xEF, 0x44, 0x07, 0xF5, 0x82, +/*1650*/0x75, 0x83, 0xD0, 0xE0, 0xFD, 0xC4, 0x54, 0x0F, + 0xF5, 0x5A, 0xEF, 0x44, 0x07, 0xF5, 0x82, 0x75, +/*1660*/0x83, 0x80, 0x74, 0x01, 0xF0, 0x12, 0x08, 0x21, + 0x75, 0x83, 0x82, 0xE5, 0x45, 0xF0, 0xEF, 0x44, +/*1670*/0x07, 0xF5, 0x82, 0x75, 0x83, 0x8A, 0x74, 0xFF, + 0xF0, 0x12, 0x1A, 0x4D, 0x12, 0x07, 0x2A, 0x75, +/*1680*/0x83, 0xBC, 0xE0, 0x54, 0xEF, 0x12, 0x07, 0x29, + 0x75, 0x83, 0xBE, 0xE0, 0x54, 0xEF, 0x12, 0x07, +/*1690*/0x29, 0x75, 0x83, 0xC0, 0xE0, 0x54, 0xEF, 0x12, + 0x07, 0x29, 0x75, 0x83, 0xBC, 0xE0, 0x44, 0x10, +/*16A0*/0x12, 0x07, 0x29, 0x75, 0x83, 0xBE, 0xE0, 0x44, + 0x10, 0x12, 0x07, 0x29, 0x75, 0x83, 0xC0, 0xE0, +/*16B0*/0x44, 0x10, 0xF0, 0xAF, 0x58, 0xE5, 0x59, 0x12, + 0x08, 0x78, 0x02, 0x0A, 0xFF, 0xE4, 0xF5, 0x58, +/*16C0*/0x7D, 0x01, 0xF5, 0x59, 0xAF, 0x35, 0xFE, 0xFC, + 0x12, 0x09, 0x15, 0x12, 0x07, 0x2A, 0x75, 0x83, +/*16D0*/0xB6, 0x74, 0x10, 0x12, 0x07, 0x29, 0x75, 0x83, + 0xB8, 0x74, 0x10, 0x12, 0x07, 0x29, 0x75, 0x83, +/*16E0*/0xBA, 0x74, 0x10, 0x12, 0x07, 0x29, 0x75, 0x83, + 0xBC, 0x74, 0x10, 0x12, 0x07, 0x29, 0x75, 0x83, +/*16F0*/0xBE, 0x74, 0x10, 0x12, 0x07, 0x29, 0x75, 0x83, + 0xC0, 0x74, 0x10, 0x12, 0x07, 0x29, 0x75, 0x83, +/*1700*/0x90, 0xE4, 0x12, 0x07, 0x29, 0x75, 0x83, 0xC2, + 0xE4, 0x12, 0x07, 0x29, 0x75, 0x83, 0xC4, 0xE4, +/*1710*/0x12, 0x07, 0x29, 0x75, 0x83, 0x92, 0xE4, 0x12, + 0x07, 0x29, 0x75, 0x83, 0xC6, 0xE4, 0x12, 0x07, +/*1720*/0x29, 0x75, 0x83, 0xC8, 0xE4, 0xF0, 0xAF, 0x58, + 0xFE, 0xE5, 0x59, 0x12, 0x08, 0x7A, 0x02, 0x0A, +/*1730*/0xFF, 0xE5, 0xE2, 0x30, 0xE4, 0x6C, 0xE5, 0xE7, + 0x54, 0xC0, 0x64, 0x40, 0x70, 0x64, 0xE5, 0x09, +/*1740*/0xC4, 0x54, 0x30, 0xFE, 0xE5, 0x08, 0x25, 0xE0, + 0x25, 0xE0, 0x54, 0xC0, 0x4E, 0xFE, 0xEF, 0x54, +/*1750*/0x3F, 0x4E, 0xFD, 0xE5, 0x2B, 0xAE, 0x2A, 0x78, + 0x02, 0xC3, 0x33, 0xCE, 0x33, 0xCE, 0xD8, 0xF9, +/*1760*/0xF5, 0x82, 0x8E, 0x83, 0xED, 0xF0, 0xE5, 0x2B, + 0xAE, 0x2A, 0x78, 0x02, 0xC3, 0x33, 0xCE, 0x33, +/*1770*/0xCE, 0xD8, 0xF9, 0xFF, 0xF5, 0x82, 0x8E, 0x83, + 0xA3, 0xE5, 0xFE, 0xF0, 0x8F, 0x82, 0x8E, 0x83, +/*1780*/0xA3, 0xA3, 0xE5, 0xFD, 0xF0, 0x8F, 0x82, 0x8E, + 0x83, 0xA3, 0xA3, 0xA3, 0xE5, 0xFC, 0xF0, 0xC3, +/*1790*/0xE5, 0x2B, 0x94, 0xFA, 0xE5, 0x2A, 0x94, 0x00, + 0x50, 0x08, 0x05, 0x2B, 0xE5, 0x2B, 0x70, 0x02, +/*17A0*/0x05, 0x2A, 0x22, 0xE4, 0xFF, 0xE4, 0xF5, 0x58, + 0xF5, 0x56, 0xF5, 0x57, 0x74, 0x82, 0xFC, 0x12, +/*17B0*/0x0E, 0x04, 0x8C, 0x83, 0xE0, 0xF5, 0x10, 0x54, + 0x7F, 0xF0, 0xE5, 0x10, 0x44, 0x80, 0x12, 0x0E, +/*17C0*/0x98, 0xED, 0xF0, 0x7E, 0x0A, 0x12, 0x0E, 0x04, + 0x75, 0x83, 0xA0, 0xE0, 0x20, 0xE0, 0x26, 0xDE, +/*17D0*/0xF4, 0x05, 0x57, 0xE5, 0x57, 0x70, 0x02, 0x05, + 0x56, 0xE5, 0x14, 0x24, 0x01, 0xFD, 0xE4, 0x33, +/*17E0*/0xFC, 0xD3, 0xE5, 0x57, 0x9D, 0xE5, 0x56, 0x9C, + 0x40, 0xD9, 0xE5, 0x0A, 0x94, 0x20, 0x50, 0x02, +/*17F0*/0x05, 0x0A, 0x43, 0xE1, 0x08, 0xC2, 0x31, 0x12, + 0x0E, 0x04, 0x75, 0x83, 0xA6, 0xE0, 0x55, 0x12, +/*1800*/0x65, 0x12, 0x70, 0x03, 0xD2, 0x31, 0x22, 0xC2, + 0x31, 0x22, 0x90, 0x07, 0x26, 0xE0, 0xFA, 0xA3, +/*1810*/0xE0, 0xF5, 0x82, 0x8A, 0x83, 0xE0, 0xF5, 0x41, + 0xE5, 0x39, 0xC3, 0x95, 0x41, 0x40, 0x26, 0xE5, +/*1820*/0x39, 0x95, 0x41, 0xC3, 0x9F, 0xEE, 0x12, 0x07, + 0x6B, 0x40, 0x04, 0x7C, 0x01, 0x80, 0x02, 0x7C, +/*1830*/0x00, 0xE5, 0x41, 0x64, 0x3F, 0x60, 0x04, 0x7B, + 0x01, 0x80, 0x02, 0x7B, 0x00, 0xEC, 0x5B, 0x60, +/*1840*/0x29, 0x05, 0x41, 0x80, 0x28, 0xC3, 0xE5, 0x41, + 0x95, 0x39, 0xC3, 0x9F, 0xEE, 0x12, 0x07, 0x6B, +/*1850*/0x40, 0x04, 0x7F, 0x01, 0x80, 0x02, 0x7F, 0x00, + 0xE5, 0x41, 0x60, 0x04, 0x7E, 0x01, 0x80, 0x02, +/*1860*/0x7E, 0x00, 0xEF, 0x5E, 0x60, 0x04, 0x15, 0x41, + 0x80, 0x03, 0x85, 0x39, 0x41, 0x85, 0x3A, 0x40, +/*1870*/0x22, 0xE5, 0xE2, 0x30, 0xE4, 0x60, 0xE5, 0xE1, + 0x30, 0xE2, 0x5B, 0xE5, 0x09, 0x70, 0x04, 0x7F, +/*1880*/0x01, 0x80, 0x02, 0x7F, 0x00, 0xE5, 0x08, 0x70, + 0x04, 0x7E, 0x01, 0x80, 0x02, 0x7E, 0x00, 0xEE, +/*1890*/0x5F, 0x60, 0x43, 0x53, 0xF9, 0xF8, 0xE5, 0xE2, + 0x30, 0xE4, 0x3B, 0xE5, 0xE1, 0x30, 0xE2, 0x2E, +/*18A0*/0x43, 0xFA, 0x02, 0x53, 0xFA, 0xFB, 0xE4, 0xF5, + 0x10, 0x90, 0x94, 0x70, 0xE5, 0x10, 0xF0, 0xE5, +/*18B0*/0xE1, 0x30, 0xE2, 0xE7, 0x90, 0x94, 0x70, 0xE0, + 0x65, 0x10, 0x60, 0x03, 0x43, 0xFA, 0x04, 0x05, +/*18C0*/0x10, 0x90, 0x94, 0x70, 0xE5, 0x10, 0xF0, 0x70, + 0xE6, 0x12, 0x00, 0x06, 0x80, 0xE1, 0x53, 0xFA, +/*18D0*/0xFD, 0x53, 0xFA, 0xFB, 0x80, 0xC0, 0x22, 0x8F, + 0x54, 0x12, 0x00, 0x06, 0xE5, 0xE1, 0x30, 0xE0, +/*18E0*/0x04, 0x7F, 0x01, 0x80, 0x02, 0x7F, 0x00, 0xE5, + 0x7E, 0xD3, 0x94, 0x05, 0x40, 0x04, 0x7E, 0x01, +/*18F0*/0x80, 0x02, 0x7E, 0x00, 0xEE, 0x4F, 0x60, 0x3D, + 0x85, 0x54, 0x11, 0xE5, 0xE2, 0x20, 0xE1, 0x32, +/*1900*/0x74, 0xCE, 0x12, 0x1A, 0x05, 0x30, 0xE7, 0x04, + 0x7D, 0x01, 0x80, 0x02, 0x7D, 0x00, 0x8F, 0x82, +/*1910*/0x8E, 0x83, 0xE0, 0x30, 0xE6, 0x04, 0x7F, 0x01, + 0x80, 0x02, 0x7F, 0x00, 0xEF, 0x5D, 0x70, 0x15, +/*1920*/0x12, 0x15, 0xC6, 0x74, 0xCE, 0x12, 0x1A, 0x05, + 0x30, 0xE6, 0x07, 0xE0, 0x44, 0x80, 0xF0, 0x43, +/*1930*/0xF9, 0x80, 0x12, 0x18, 0x71, 0x22, 0x12, 0x0E, + 0x44, 0xE5, 0x16, 0x25, 0xE0, 0x25, 0xE0, 0x24, +/*1940*/0xB0, 0xF5, 0x82, 0xE4, 0x34, 0x1A, 0xF5, 0x83, + 0xE4, 0x93, 0xF5, 0x0F, 0xE5, 0x16, 0x25, 0xE0, +/*1950*/0x25, 0xE0, 0x24, 0xB1, 0xF5, 0x82, 0xE4, 0x34, + 0x1A, 0xF5, 0x83, 0xE4, 0x93, 0xF5, 0x0E, 0x12, +/*1960*/0x0E, 0x65, 0xF5, 0x10, 0xE5, 0x0F, 0x54, 0xF0, + 0x12, 0x0E, 0x17, 0x75, 0x83, 0x8C, 0xEF, 0xF0, +/*1970*/0xE5, 0x0F, 0x30, 0xE0, 0x0C, 0x12, 0x0E, 0x04, + 0x75, 0x83, 0x86, 0xE0, 0x44, 0x40, 0xF0, 0x80, +/*1980*/0x0A, 0x12, 0x0E, 0x04, 0x75, 0x83, 0x86, 0xE0, + 0x54, 0xBF, 0xF0, 0x12, 0x0E, 0x91, 0x75, 0x83, +/*1990*/0x82, 0xE5, 0x0E, 0xF0, 0x22, 0x7F, 0x05, 0x12, + 0x17, 0x31, 0x12, 0x0E, 0x04, 0x12, 0x0E, 0x33, +/*19A0*/0x74, 0x02, 0xF0, 0x74, 0x8E, 0xFE, 0x12, 0x0E, + 0x04, 0x12, 0x0E, 0x0B, 0xEF, 0xF0, 0x75, 0x15, +/*19B0*/0x70, 0x12, 0x0F, 0xF7, 0x20, 0x34, 0x05, 0x75, + 0x15, 0x10, 0x80, 0x03, 0x75, 0x15, 0x50, 0x12, +/*19C0*/0x0F, 0xF7, 0x20, 0x34, 0x04, 0x74, 0x10, 0x80, + 0x02, 0x74, 0xF0, 0x25, 0x15, 0xF5, 0x15, 0x12, +/*19D0*/0x0E, 0x21, 0xEF, 0xF0, 0x12, 0x10, 0x91, 0x20, + 0x34, 0x17, 0xE5, 0x15, 0x64, 0x30, 0x60, 0x0C, +/*19E0*/0x74, 0x10, 0x25, 0x15, 0xF5, 0x15, 0xB4, 0x80, + 0x03, 0xE4, 0xF5, 0x15, 0x12, 0x0E, 0x21, 0xEF, +/*19F0*/0xF0, 0x22, 0xF0, 0xE5, 0x0B, 0x25, 0xE0, 0x25, + 0xE0, 0x24, 0x82, 0xF5, 0x82, 0xE4, 0x34, 0x07, +/*1A00*/0xF5, 0x83, 0x22, 0x74, 0x88, 0xFE, 0xE5, 0x08, + 0x44, 0x07, 0xFF, 0xF5, 0x82, 0x8E, 0x83, 0xE0, +/*1A10*/0x22, 0xF0, 0xE5, 0x08, 0x44, 0x07, 0xF5, 0x82, + 0x22, 0xF0, 0xE0, 0x54, 0xC0, 0x8F, 0x82, 0x8E, +/*1A20*/0x83, 0xF0, 0x22, 0xEF, 0x44, 0x07, 0xF5, 0x82, + 0x75, 0x83, 0x86, 0xE0, 0x54, 0x10, 0xD3, 0x94, +/*1A30*/0x00, 0x22, 0xF0, 0x90, 0x07, 0x15, 0xE0, 0x04, + 0xF0, 0x22, 0x44, 0x06, 0xF5, 0x82, 0x75, 0x83, +/*1A40*/0x9E, 0xE0, 0x22, 0xFE, 0xEF, 0x44, 0x07, 0xF5, + 0x82, 0x8E, 0x83, 0xE0, 0x22, 0xE4, 0x90, 0x07, +/*1A50*/0x2A, 0xF0, 0xA3, 0xF0, 0x12, 0x07, 0x2A, 0x75, + 0x83, 0x82, 0xE0, 0x54, 0x7F, 0x12, 0x07, 0x29, +/*1A60*/0xE0, 0x44, 0x80, 0xF0, 0x12, 0x10, 0xFC, 0x12, + 0x08, 0x1F, 0x75, 0x83, 0xA0, 0xE0, 0x20, 0xE0, +/*1A70*/0x1A, 0x90, 0x07, 0x2B, 0xE0, 0x04, 0xF0, 0x70, + 0x06, 0x90, 0x07, 0x2A, 0xE0, 0x04, 0xF0, 0x90, +/*1A80*/0x07, 0x2A, 0xE0, 0xB4, 0x10, 0xE1, 0xA3, 0xE0, + 0xB4, 0x00, 0xDC, 0xEE, 0x44, 0xA6, 0xFC, 0xEF, +/*1A90*/0x44, 0x07, 0xF5, 0x82, 0x8C, 0x83, 0xE0, 0xF5, + 0x32, 0xEE, 0x44, 0xA8, 0xFE, 0xEF, 0x44, 0x07, +/*1AA0*/0xF5, 0x82, 0x8E, 0x83, 0xE0, 0xF5, 0x33, 0x22, + 0x01, 0x20, 0x11, 0x00, 0x04, 0x20, 0x00, 0x90, +/*1AB0*/0x00, 0x20, 0x0F, 0x92, 0x00, 0x21, 0x0F, 0x94, + 0x00, 0x22, 0x0F, 0x96, 0x00, 0x23, 0x0F, 0x98, +/*1AC0*/0x00, 0x24, 0x0F, 0x9A, 0x00, 0x25, 0x0F, 0x9C, + 0x00, 0x26, 0x0F, 0x9E, 0x00, 0x27, 0x0F, 0xA0, +/*1AD0*/0x01, 0x20, 0x01, 0xA2, 0x01, 0x21, 0x01, 0xA4, + 0x01, 0x22, 0x01, 0xA6, 0x01, 0x23, 0x01, 0xA8, +/*1AE0*/0x01, 0x24, 0x01, 0xAA, 0x01, 0x25, 0x01, 0xAC, + 0x01, 0x26, 0x01, 0xAE, 0x01, 0x27, 0x01, 0xB0, +/*1AF0*/0x01, 0x28, 0x01, 0xB4, 0x00, 0x28, 0x0F, 0xB6, + 0x40, 0x28, 0x0F, 0xB8, 0x61, 0x28, 0x01, 0xCB, +/*1B00*/0xEF, 0xCB, 0xCA, 0xEE, 0xCA, 0x7F, 0x01, 0xE4, + 0xFD, 0xEB, 0x4A, 0x70, 0x24, 0xE5, 0x08, 0xF5, +/*1B10*/0x82, 0x74, 0xB6, 0x12, 0x08, 0x29, 0xE5, 0x08, + 0xF5, 0x82, 0x74, 0xB8, 0x12, 0x08, 0x29, 0xE5, +/*1B20*/0x08, 0xF5, 0x82, 0x74, 0xBA, 0x12, 0x08, 0x29, + 0x7E, 0x00, 0x7C, 0x00, 0x12, 0x0A, 0xFF, 0x80, +/*1B30*/0x12, 0x90, 0x07, 0x26, 0x12, 0x07, 0x35, 0xE5, + 0x41, 0xF0, 0x90, 0x07, 0x24, 0x12, 0x07, 0x35, +/*1B40*/0xE5, 0x40, 0xF0, 0x12, 0x07, 0x2A, 0x75, 0x83, + 0x8E, 0xE4, 0x12, 0x07, 0x29, 0x74, 0x01, 0x12, +/*1B50*/0x07, 0x29, 0xE4, 0xF0, 0x22, 0xE4, 0xF5, 0x26, + 0xF5, 0x27, 0x53, 0xE1, 0xFE, 0xF5, 0x2A, 0x75, +/*1B60*/0x2B, 0x01, 0xF5, 0x08, 0x7F, 0x01, 0x12, 0x17, + 0x31, 0x30, 0x30, 0x1C, 0x90, 0x1A, 0xA9, 0xE4, +/*1B70*/0x93, 0xF5, 0x10, 0x90, 0x1F, 0xF9, 0xE4, 0x93, + 0xF5, 0x10, 0x90, 0x00, 0x41, 0xE4, 0x93, 0xF5, +/*1B80*/0x10, 0x90, 0x1E, 0xCA, 0xE4, 0x93, 0xF5, 0x10, + 0x7F, 0x02, 0x12, 0x17, 0x31, 0x12, 0x0F, 0x54, +/*1B90*/0x7F, 0x03, 0x12, 0x17, 0x31, 0x12, 0x00, 0x06, + 0xE5, 0xE2, 0x30, 0xE7, 0x09, 0x12, 0x10, 0x00, +/*1BA0*/0x30, 0x30, 0x03, 0x12, 0x11, 0x00, 0x02, 0x00, + 0x47, 0x12, 0x08, 0x1F, 0x75, 0x83, 0xD0, 0xE0, +/*1BB0*/0xC4, 0x54, 0x0F, 0xFD, 0x75, 0x43, 0x01, 0x75, + 0x44, 0xFF, 0x12, 0x08, 0xAA, 0x74, 0x04, 0xF0, +/*1BC0*/0x75, 0x3B, 0x01, 0xED, 0x14, 0x60, 0x0C, 0x14, + 0x60, 0x0B, 0x14, 0x60, 0x0F, 0x24, 0x03, 0x70, +/*1BD0*/0x0B, 0x80, 0x09, 0x80, 0x00, 0x12, 0x08, 0xA7, + 0x04, 0xF0, 0x80, 0x06, 0x12, 0x08, 0xA7, 0x74, +/*1BE0*/0x04, 0xF0, 0xEE, 0x44, 0x82, 0xFE, 0xEF, 0x44, + 0x07, 0xF5, 0x82, 0x8E, 0x83, 0xE5, 0x45, 0x12, +/*1BF0*/0x08, 0xBE, 0x75, 0x83, 0x82, 0xE5, 0x31, 0xF0, + 0x02, 0x11, 0x4C, 0x8E, 0x60, 0x8F, 0x61, 0x12, +/*1C00*/0x1E, 0xA5, 0xE4, 0xFF, 0xCE, 0xED, 0xCE, 0xEE, + 0xD3, 0x95, 0x61, 0xE5, 0x60, 0x12, 0x07, 0x6B, +/*1C10*/0x40, 0x39, 0x74, 0x20, 0x2E, 0xF5, 0x82, 0xE4, + 0x34, 0x03, 0xF5, 0x83, 0xE0, 0x70, 0x03, 0xFF, +/*1C20*/0x80, 0x26, 0x12, 0x08, 0xE2, 0xFD, 0xC3, 0x9F, + 0x40, 0x1E, 0xCF, 0xED, 0xCF, 0xEB, 0x4A, 0x70, +/*1C30*/0x0B, 0x8D, 0x42, 0x12, 0x08, 0xEE, 0xF5, 0x41, + 0x8E, 0x40, 0x80, 0x0C, 0x12, 0x08, 0xE2, 0xF5, +/*1C40*/0x38, 0x12, 0x08, 0xEE, 0xF5, 0x39, 0x8E, 0x3A, + 0x1E, 0x80, 0xBC, 0x22, 0x75, 0x58, 0x01, 0xE5, +/*1C50*/0x35, 0x70, 0x0C, 0x12, 0x07, 0xCC, 0xE0, 0xF5, + 0x4A, 0x12, 0x07, 0xD8, 0xE0, 0xF5, 0x4C, 0xE5, +/*1C60*/0x35, 0xB4, 0x04, 0x0C, 0x12, 0x07, 0xE4, 0xE0, + 0xF5, 0x4A, 0x12, 0x07, 0xF0, 0xE0, 0xF5, 0x4C, +/*1C70*/0xE5, 0x35, 0xB4, 0x01, 0x04, 0x7F, 0x01, 0x80, + 0x02, 0x7F, 0x00, 0xE5, 0x35, 0xB4, 0x02, 0x04, +/*1C80*/0x7E, 0x01, 0x80, 0x02, 0x7E, 0x00, 0xEE, 0x4F, + 0x60, 0x0C, 0x12, 0x07, 0xFC, 0xE0, 0xF5, 0x4A, +/*1C90*/0x12, 0x08, 0x08, 0xE0, 0xF5, 0x4C, 0x85, 0x41, + 0x49, 0x85, 0x40, 0x4B, 0x22, 0x75, 0x5B, 0x01, +/*1CA0*/0x90, 0x07, 0x24, 0x12, 0x07, 0x35, 0xE0, 0x54, + 0x1F, 0xFF, 0xD3, 0x94, 0x02, 0x50, 0x04, 0x8F, +/*1CB0*/0x58, 0x80, 0x05, 0xEF, 0x24, 0xFE, 0xF5, 0x58, + 0xEF, 0xC3, 0x94, 0x18, 0x40, 0x05, 0x75, 0x59, +/*1CC0*/0x18, 0x80, 0x04, 0xEF, 0x04, 0xF5, 0x59, 0x85, + 0x43, 0x5A, 0xAF, 0x58, 0x7E, 0x00, 0xAD, 0x59, +/*1CD0*/0x7C, 0x00, 0xAB, 0x5B, 0x7A, 0x00, 0x12, 0x15, + 0x41, 0xAF, 0x5A, 0x7E, 0x00, 0x12, 0x18, 0x0A, +/*1CE0*/0xAF, 0x5B, 0x7E, 0x00, 0x02, 0x1A, 0xFF, 0xE5, + 0xE2, 0x30, 0xE7, 0x0E, 0x12, 0x10, 0x03, 0xC2, +/*1CF0*/0x30, 0x30, 0x30, 0x03, 0x12, 0x10, 0xFF, 0x20, + 0x33, 0x28, 0xE5, 0xE7, 0x30, 0xE7, 0x05, 0x12, +/*1D00*/0x0E, 0xA2, 0x80, 0x0D, 0xE5, 0xFE, 0xC3, 0x94, + 0x20, 0x50, 0x06, 0x12, 0x0E, 0xA2, 0x43, 0xF9, +/*1D10*/0x08, 0xE5, 0xF2, 0x30, 0xE7, 0x03, 0x53, 0xF9, + 0x7F, 0xE5, 0xF1, 0x54, 0x70, 0xD3, 0x94, 0x00, +/*1D20*/0x50, 0xD8, 0x22, 0x12, 0x0E, 0x04, 0x75, 0x83, + 0x80, 0xE4, 0xF0, 0xE5, 0x08, 0x44, 0x07, 0x12, +/*1D30*/0x0D, 0xFD, 0x75, 0x83, 0x84, 0x12, 0x0E, 0x02, + 0x75, 0x83, 0x86, 0x12, 0x0E, 0x02, 0x75, 0x83, +/*1D40*/0x8C, 0xE0, 0x54, 0xF3, 0x12, 0x0E, 0x03, 0x75, + 0x83, 0x8E, 0x12, 0x0E, 0x02, 0x75, 0x83, 0x94, +/*1D50*/0xE0, 0x54, 0xFB, 0xF0, 0x22, 0x12, 0x07, 0x2A, + 0x75, 0x83, 0x8E, 0xE4, 0x12, 0x07, 0x29, 0x74, +/*1D60*/0x01, 0x12, 0x07, 0x29, 0xE4, 0x12, 0x08, 0xBE, + 0x75, 0x83, 0x8C, 0xE0, 0x44, 0x20, 0x12, 0x08, +/*1D70*/0xBE, 0xE0, 0x54, 0xDF, 0xF0, 0x74, 0x84, 0x85, + 0x08, 0x82, 0xF5, 0x83, 0xE0, 0x54, 0x7F, 0xF0, +/*1D80*/0xE0, 0x44, 0x80, 0xF0, 0x22, 0x75, 0x56, 0x01, + 0xE4, 0xFD, 0xF5, 0x57, 0xAF, 0x35, 0xFE, 0xFC, +/*1D90*/0x12, 0x09, 0x15, 0x12, 0x1C, 0x9D, 0x12, 0x1E, + 0x7A, 0x12, 0x1C, 0x4C, 0xAF, 0x57, 0x7E, 0x00, +/*1DA0*/0xAD, 0x56, 0x7C, 0x00, 0x12, 0x04, 0x44, 0xAF, + 0x56, 0x7E, 0x00, 0x02, 0x11, 0xEE, 0x75, 0x56, +/*1DB0*/0x01, 0xE4, 0xFD, 0xF5, 0x57, 0xAF, 0x35, 0xFE, + 0xFC, 0x12, 0x09, 0x15, 0x12, 0x1C, 0x9D, 0x12, +/*1DC0*/0x1E, 0x7A, 0x12, 0x1C, 0x4C, 0xAF, 0x57, 0x7E, + 0x00, 0xAD, 0x56, 0x7C, 0x00, 0x12, 0x04, 0x44, +/*1DD0*/0xAF, 0x56, 0x7E, 0x00, 0x02, 0x11, 0xEE, 0xE4, + 0xF5, 0x16, 0x12, 0x0E, 0x44, 0xFE, 0xE5, 0x08, +/*1DE0*/0x44, 0x05, 0xFF, 0x12, 0x0E, 0x65, 0x8F, 0x82, + 0x8E, 0x83, 0xF0, 0x05, 0x16, 0xE5, 0x16, 0xC3, +/*1DF0*/0x94, 0x14, 0x40, 0xE6, 0xE5, 0x08, 0x12, 0x0E, + 0x2B, 0xE4, 0xF0, 0x22, 0xE4, 0xF5, 0x58, 0xF5, +/*1E00*/0x59, 0xF5, 0x5A, 0xFF, 0xFE, 0xAD, 0x58, 0xFC, + 0x12, 0x09, 0x15, 0x7F, 0x04, 0x7E, 0x00, 0xAD, +/*1E10*/0x58, 0x7C, 0x00, 0x12, 0x09, 0x15, 0x7F, 0x02, + 0x7E, 0x00, 0xAD, 0x58, 0x7C, 0x00, 0x02, 0x09, +/*1E20*/0x15, 0xE5, 0x3C, 0x25, 0x3E, 0xFC, 0xE5, 0x42, + 0x24, 0x00, 0xFB, 0xE4, 0x33, 0xFA, 0xEC, 0xC3, +/*1E30*/0x9B, 0xEA, 0x12, 0x07, 0x6B, 0x40, 0x0B, 0x8C, + 0x42, 0xE5, 0x3D, 0x25, 0x3F, 0xF5, 0x41, 0x8F, +/*1E40*/0x40, 0x22, 0x12, 0x09, 0x0B, 0x22, 0x74, 0x84, + 0xF5, 0x18, 0x85, 0x08, 0x19, 0x85, 0x19, 0x82, +/*1E50*/0x85, 0x18, 0x83, 0xE0, 0x54, 0x7F, 0xF0, 0xE0, + 0x44, 0x80, 0xF0, 0xE0, 0x44, 0x80, 0xF0, 0x22, +/*1E60*/0xEF, 0x4E, 0x70, 0x0B, 0x12, 0x07, 0x2A, 0x75, + 0x83, 0xD2, 0xE0, 0x54, 0xDF, 0xF0, 0x22, 0x12, +/*1E70*/0x07, 0x2A, 0x75, 0x83, 0xD2, 0xE0, 0x44, 0x20, + 0xF0, 0x22, 0x75, 0x58, 0x01, 0x90, 0x07, 0x26, +/*1E80*/0x12, 0x07, 0x35, 0xE0, 0x54, 0x3F, 0xF5, 0x41, + 0x12, 0x07, 0x32, 0xE0, 0x54, 0x3F, 0xF5, 0x40, +/*1E90*/0x22, 0x75, 0x56, 0x02, 0xE4, 0xF5, 0x57, 0x12, + 0x1D, 0xFC, 0xAF, 0x57, 0x7E, 0x00, 0xAD, 0x56, +/*1EA0*/0x7C, 0x00, 0x02, 0x04, 0x44, 0xE4, 0xF5, 0x42, + 0xF5, 0x41, 0xF5, 0x40, 0xF5, 0x38, 0xF5, 0x39, +/*1EB0*/0xF5, 0x3A, 0x22, 0xEF, 0x54, 0x07, 0xFF, 0xE5, + 0xF9, 0x54, 0xF8, 0x4F, 0xF5, 0xF9, 0x22, 0x7F, +/*1EC0*/0x01, 0xE4, 0xFE, 0x0F, 0x0E, 0xBE, 0xFF, 0xFB, + 0x22, 0x01, 0x20, 0x00, 0x01, 0x04, 0x20, 0x00, +/*1ED0*/0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, +/*1EE0*/0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, +/*1EF0*/0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, +/*1F00*/0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, +/*1F10*/0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, +/*1F20*/0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, +/*1F30*/0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, +/*1F40*/0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, +/*1F50*/0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, +/*1F60*/0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, +/*1F70*/0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, +/*1F80*/0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, +/*1F90*/0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, +/*1FA0*/0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, +/*1FB0*/0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, +/*1FC0*/0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, +/*1FD0*/0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, +/*1FE0*/0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, +/*1FF0*/0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x01, 0x20, 0x11, 0x00, 0x04, 0x20, 0x00, 0x81 +}; + +int ipath_sd7220_ib_load(struct ipath_devdata *dd) +{ + return ipath_sd7220_prog_ld(dd, IB_7220_SERDES, ipath_sd7220_ib_img, + sizeof(ipath_sd7220_ib_img), 0); +} + +int ipath_sd7220_ib_vfy(struct ipath_devdata *dd) +{ + return ipath_sd7220_prog_vfy(dd, IB_7220_SERDES, ipath_sd7220_ib_img, + sizeof(ipath_sd7220_ib_img), 0); +} diff --git a/drivers/infiniband/hw/ipath/ipath_sdma.c b/drivers/infiniband/hw/ipath/ipath_sdma.c new file mode 100644 index 000000000000..1974df7a9f78 --- /dev/null +++ b/drivers/infiniband/hw/ipath/ipath_sdma.c @@ -0,0 +1,790 @@ +/* + * Copyright (c) 2007, 2008 QLogic Corporation. All rights reserved. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * OpenIB.org BSD license below: + * + * Redistribution and use in source and binary forms, with or + * without modification, are permitted provided that the following + * conditions are met: + * + * - Redistributions of source code must retain the above + * copyright notice, this list of conditions and the following + * disclaimer. + * + * - Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials + * provided with the distribution. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + */ + +#include <linux/spinlock.h> + +#include "ipath_kernel.h" +#include "ipath_verbs.h" +#include "ipath_common.h" + +#define SDMA_DESCQ_SZ PAGE_SIZE /* 256 entries per 4KB page */ + +static void vl15_watchdog_enq(struct ipath_devdata *dd) +{ + /* ipath_sdma_lock must already be held */ + if (atomic_inc_return(&dd->ipath_sdma_vl15_count) == 1) { + unsigned long interval = (HZ + 19) / 20; + dd->ipath_sdma_vl15_timer.expires = jiffies + interval; + add_timer(&dd->ipath_sdma_vl15_timer); + } +} + +static void vl15_watchdog_deq(struct ipath_devdata *dd) +{ + /* ipath_sdma_lock must already be held */ + if (atomic_dec_return(&dd->ipath_sdma_vl15_count) != 0) { + unsigned long interval = (HZ + 19) / 20; + mod_timer(&dd->ipath_sdma_vl15_timer, jiffies + interval); + } else { + del_timer(&dd->ipath_sdma_vl15_timer); + } +} + +static void vl15_watchdog_timeout(unsigned long opaque) +{ + struct ipath_devdata *dd = (struct ipath_devdata *)opaque; + + if (atomic_read(&dd->ipath_sdma_vl15_count) != 0) { + ipath_dbg("vl15 watchdog timeout - clearing\n"); + ipath_cancel_sends(dd, 1); + ipath_hol_down(dd); + } else { + ipath_dbg("vl15 watchdog timeout - " + "condition already cleared\n"); + } +} + +static void unmap_desc(struct ipath_devdata *dd, unsigned head) +{ + __le64 *descqp = &dd->ipath_sdma_descq[head].qw[0]; + u64 desc[2]; + dma_addr_t addr; + size_t len; + + desc[0] = le64_to_cpu(descqp[0]); + desc[1] = le64_to_cpu(descqp[1]); + + addr = (desc[1] << 32) | (desc[0] >> 32); + len = (desc[0] >> 14) & (0x7ffULL << 2); + dma_unmap_single(&dd->pcidev->dev, addr, len, DMA_TO_DEVICE); +} + +/* + * ipath_sdma_lock should be locked before calling this. + */ +int ipath_sdma_make_progress(struct ipath_devdata *dd) +{ + struct list_head *lp = NULL; + struct ipath_sdma_txreq *txp = NULL; + u16 dmahead; + u16 start_idx = 0; + int progress = 0; + + if (!list_empty(&dd->ipath_sdma_activelist)) { + lp = dd->ipath_sdma_activelist.next; + txp = list_entry(lp, struct ipath_sdma_txreq, list); + start_idx = txp->start_idx; + } + + /* + * Read the SDMA head register in order to know that the + * interrupt clear has been written to the chip. + * Otherwise, we may not get an interrupt for the last + * descriptor in the queue. + */ + dmahead = (u16)ipath_read_kreg32(dd, dd->ipath_kregs->kr_senddmahead); + /* sanity check return value for error handling (chip reset, etc.) */ + if (dmahead >= dd->ipath_sdma_descq_cnt) + goto done; + + while (dd->ipath_sdma_descq_head != dmahead) { + if (txp && txp->flags & IPATH_SDMA_TXREQ_F_FREEDESC && + dd->ipath_sdma_descq_head == start_idx) { + unmap_desc(dd, dd->ipath_sdma_descq_head); + start_idx++; + if (start_idx == dd->ipath_sdma_descq_cnt) + start_idx = 0; + } + + /* increment free count and head */ + dd->ipath_sdma_descq_removed++; + if (++dd->ipath_sdma_descq_head == dd->ipath_sdma_descq_cnt) + dd->ipath_sdma_descq_head = 0; + + if (txp && txp->next_descq_idx == dd->ipath_sdma_descq_head) { + /* move to notify list */ + if (txp->flags & IPATH_SDMA_TXREQ_F_VL15) + vl15_watchdog_deq(dd); + list_move_tail(lp, &dd->ipath_sdma_notifylist); + if (!list_empty(&dd->ipath_sdma_activelist)) { + lp = dd->ipath_sdma_activelist.next; + txp = list_entry(lp, struct ipath_sdma_txreq, + list); + start_idx = txp->start_idx; + } else { + lp = NULL; + txp = NULL; + } + } + progress = 1; + } + + if (progress) + tasklet_hi_schedule(&dd->ipath_sdma_notify_task); + +done: + return progress; +} + +static void ipath_sdma_notify(struct ipath_devdata *dd, struct list_head *list) +{ + struct ipath_sdma_txreq *txp, *txp_next; + + list_for_each_entry_safe(txp, txp_next, list, list) { + list_del_init(&txp->list); + + if (txp->callback) + (*txp->callback)(txp->callback_cookie, + txp->callback_status); + } +} + +static void sdma_notify_taskbody(struct ipath_devdata *dd) +{ + unsigned long flags; + struct list_head list; + + INIT_LIST_HEAD(&list); + + spin_lock_irqsave(&dd->ipath_sdma_lock, flags); + + list_splice_init(&dd->ipath_sdma_notifylist, &list); + + spin_unlock_irqrestore(&dd->ipath_sdma_lock, flags); + + ipath_sdma_notify(dd, &list); + + /* + * The IB verbs layer needs to see the callback before getting + * the call to ipath_ib_piobufavail() because the callback + * handles releasing resources the next send will need. + * Otherwise, we could do these calls in + * ipath_sdma_make_progress(). + */ + ipath_ib_piobufavail(dd->verbs_dev); +} + +static void sdma_notify_task(unsigned long opaque) +{ + struct ipath_devdata *dd = (struct ipath_devdata *)opaque; + + if (!test_bit(IPATH_SDMA_SHUTDOWN, &dd->ipath_sdma_status)) + sdma_notify_taskbody(dd); +} + +static void dump_sdma_state(struct ipath_devdata *dd) +{ + unsigned long reg; + + reg = ipath_read_kreg64(dd, dd->ipath_kregs->kr_senddmastatus); + ipath_cdbg(VERBOSE, "kr_senddmastatus: 0x%016lx\n", reg); + + reg = ipath_read_kreg64(dd, dd->ipath_kregs->kr_sendctrl); + ipath_cdbg(VERBOSE, "kr_sendctrl: 0x%016lx\n", reg); + + reg = ipath_read_kreg64(dd, dd->ipath_kregs->kr_senddmabufmask0); + ipath_cdbg(VERBOSE, "kr_senddmabufmask0: 0x%016lx\n", reg); + + reg = ipath_read_kreg64(dd, dd->ipath_kregs->kr_senddmabufmask1); + ipath_cdbg(VERBOSE, "kr_senddmabufmask1: 0x%016lx\n", reg); + + reg = ipath_read_kreg64(dd, dd->ipath_kregs->kr_senddmabufmask2); + ipath_cdbg(VERBOSE, "kr_senddmabufmask2: 0x%016lx\n", reg); + + reg = ipath_read_kreg64(dd, dd->ipath_kregs->kr_senddmatail); + ipath_cdbg(VERBOSE, "kr_senddmatail: 0x%016lx\n", reg); + + reg = ipath_read_kreg64(dd, dd->ipath_kregs->kr_senddmahead); + ipath_cdbg(VERBOSE, "kr_senddmahead: 0x%016lx\n", reg); +} + +static void sdma_abort_task(unsigned long opaque) +{ + struct ipath_devdata *dd = (struct ipath_devdata *) opaque; + u64 status; + unsigned long flags; + + if (test_bit(IPATH_SDMA_SHUTDOWN, &dd->ipath_sdma_status)) + return; + + spin_lock_irqsave(&dd->ipath_sdma_lock, flags); + + status = dd->ipath_sdma_status & IPATH_SDMA_ABORT_MASK; + + /* nothing to do */ + if (status == IPATH_SDMA_ABORT_NONE) + goto unlock; + + /* ipath_sdma_abort() is done, waiting for interrupt */ + if (status == IPATH_SDMA_ABORT_DISARMED) { + if (jiffies < dd->ipath_sdma_abort_intr_timeout) + goto resched_noprint; + /* give up, intr got lost somewhere */ + ipath_dbg("give up waiting for SDMADISABLED intr\n"); + __set_bit(IPATH_SDMA_DISABLED, &dd->ipath_sdma_status); + status = IPATH_SDMA_ABORT_ABORTED; + } + + /* everything is stopped, time to clean up and restart */ + if (status == IPATH_SDMA_ABORT_ABORTED) { + struct ipath_sdma_txreq *txp, *txpnext; + u64 hwstatus; + int notify = 0; + + hwstatus = ipath_read_kreg64(dd, + dd->ipath_kregs->kr_senddmastatus); + + if (/* ScoreBoardDrainInProg */ + test_bit(63, &hwstatus) || + /* AbortInProg */ + test_bit(62, &hwstatus) || + /* InternalSDmaEnable */ + test_bit(61, &hwstatus) || + /* ScbEmpty */ + !test_bit(30, &hwstatus)) { + if (dd->ipath_sdma_reset_wait > 0) { + /* not done shutting down sdma */ + --dd->ipath_sdma_reset_wait; + goto resched; + } + ipath_cdbg(VERBOSE, "gave up waiting for quiescent " + "status after SDMA reset, continuing\n"); + dump_sdma_state(dd); + } + + /* dequeue all "sent" requests */ + list_for_each_entry_safe(txp, txpnext, + &dd->ipath_sdma_activelist, list) { + txp->callback_status = IPATH_SDMA_TXREQ_S_ABORTED; + if (txp->flags & IPATH_SDMA_TXREQ_F_VL15) + vl15_watchdog_deq(dd); + list_move_tail(&txp->list, &dd->ipath_sdma_notifylist); + notify = 1; + } + if (notify) + tasklet_hi_schedule(&dd->ipath_sdma_notify_task); + + /* reset our notion of head and tail */ + dd->ipath_sdma_descq_tail = 0; + dd->ipath_sdma_descq_head = 0; + dd->ipath_sdma_head_dma[0] = 0; + dd->ipath_sdma_generation = 0; + dd->ipath_sdma_descq_removed = dd->ipath_sdma_descq_added; + + /* Reset SendDmaLenGen */ + ipath_write_kreg(dd, dd->ipath_kregs->kr_senddmalengen, + (u64) dd->ipath_sdma_descq_cnt | (1ULL << 18)); + + /* done with sdma state for a bit */ + spin_unlock_irqrestore(&dd->ipath_sdma_lock, flags); + + /* + * Don't restart sdma here. Wait until link is up to ACTIVE. + * VL15 MADs used to bring the link up use PIO, and multiple + * link transitions otherwise cause the sdma engine to be + * stopped and started multiple times. + * The disable is done here, including the shadow, so the + * state is kept consistent. + * See ipath_restart_sdma() for the actual starting of sdma. + */ + spin_lock_irqsave(&dd->ipath_sendctrl_lock, flags); + dd->ipath_sendctrl &= ~INFINIPATH_S_SDMAENABLE; + ipath_write_kreg(dd, dd->ipath_kregs->kr_sendctrl, + dd->ipath_sendctrl); + ipath_read_kreg64(dd, dd->ipath_kregs->kr_scratch); + spin_unlock_irqrestore(&dd->ipath_sendctrl_lock, flags); + + /* make sure I see next message */ + dd->ipath_sdma_abort_jiffies = 0; + + goto done; + } + +resched: + /* + * for now, keep spinning + * JAG - this is bad to just have default be a loop without + * state change + */ + if (jiffies > dd->ipath_sdma_abort_jiffies) { + ipath_dbg("looping with status 0x%016llx\n", + dd->ipath_sdma_status); + dd->ipath_sdma_abort_jiffies = jiffies + 5 * HZ; + } +resched_noprint: + spin_unlock_irqrestore(&dd->ipath_sdma_lock, flags); + if (!test_bit(IPATH_SDMA_SHUTDOWN, &dd->ipath_sdma_status)) + tasklet_hi_schedule(&dd->ipath_sdma_abort_task); + return; + +unlock: + spin_unlock_irqrestore(&dd->ipath_sdma_lock, flags); +done: + return; +} + +/* + * This is called from interrupt context. + */ +void ipath_sdma_intr(struct ipath_devdata *dd) +{ + unsigned long flags; + + spin_lock_irqsave(&dd->ipath_sdma_lock, flags); + + (void) ipath_sdma_make_progress(dd); + + spin_unlock_irqrestore(&dd->ipath_sdma_lock, flags); +} + +static int alloc_sdma(struct ipath_devdata *dd) +{ + int ret = 0; + + /* Allocate memory for SendDMA descriptor FIFO */ + dd->ipath_sdma_descq = dma_alloc_coherent(&dd->pcidev->dev, + SDMA_DESCQ_SZ, &dd->ipath_sdma_descq_phys, GFP_KERNEL); + + if (!dd->ipath_sdma_descq) { + ipath_dev_err(dd, "failed to allocate SendDMA descriptor " + "FIFO memory\n"); + ret = -ENOMEM; + goto done; + } + + dd->ipath_sdma_descq_cnt = + SDMA_DESCQ_SZ / sizeof(struct ipath_sdma_desc); + + /* Allocate memory for DMA of head register to memory */ + dd->ipath_sdma_head_dma = dma_alloc_coherent(&dd->pcidev->dev, + PAGE_SIZE, &dd->ipath_sdma_head_phys, GFP_KERNEL); + if (!dd->ipath_sdma_head_dma) { + ipath_dev_err(dd, "failed to allocate SendDMA head memory\n"); + ret = -ENOMEM; + goto cleanup_descq; + } + dd->ipath_sdma_head_dma[0] = 0; + + init_timer(&dd->ipath_sdma_vl15_timer); + dd->ipath_sdma_vl15_timer.function = vl15_watchdog_timeout; + dd->ipath_sdma_vl15_timer.data = (unsigned long)dd; + atomic_set(&dd->ipath_sdma_vl15_count, 0); + + goto done; + +cleanup_descq: + dma_free_coherent(&dd->pcidev->dev, SDMA_DESCQ_SZ, + (void *)dd->ipath_sdma_descq, dd->ipath_sdma_descq_phys); + dd->ipath_sdma_descq = NULL; + dd->ipath_sdma_descq_phys = 0; +done: + return ret; +} + +int setup_sdma(struct ipath_devdata *dd) +{ + int ret = 0; + unsigned i, n; + u64 tmp64; + u64 senddmabufmask[3] = { 0 }; + unsigned long flags; + + ret = alloc_sdma(dd); + if (ret) + goto done; + + if (!dd->ipath_sdma_descq) { + ipath_dev_err(dd, "SendDMA memory not allocated\n"); + goto done; + } + + dd->ipath_sdma_status = 0; + dd->ipath_sdma_abort_jiffies = 0; + dd->ipath_sdma_generation = 0; + dd->ipath_sdma_descq_tail = 0; + dd->ipath_sdma_descq_head = 0; + dd->ipath_sdma_descq_removed = 0; + dd->ipath_sdma_descq_added = 0; + + /* Set SendDmaBase */ + ipath_write_kreg(dd, dd->ipath_kregs->kr_senddmabase, + dd->ipath_sdma_descq_phys); + /* Set SendDmaLenGen */ + tmp64 = dd->ipath_sdma_descq_cnt; + tmp64 |= 1<<18; /* enable generation checking */ + ipath_write_kreg(dd, dd->ipath_kregs->kr_senddmalengen, tmp64); + /* Set SendDmaTail */ + ipath_write_kreg(dd, dd->ipath_kregs->kr_senddmatail, + dd->ipath_sdma_descq_tail); + /* Set SendDmaHeadAddr */ + ipath_write_kreg(dd, dd->ipath_kregs->kr_senddmaheadaddr, + dd->ipath_sdma_head_phys); + + /* Reserve all the former "kernel" piobufs */ + n = dd->ipath_piobcnt2k + dd->ipath_piobcnt4k - dd->ipath_pioreserved; + for (i = dd->ipath_lastport_piobuf; i < n; ++i) { + unsigned word = i / 64; + unsigned bit = i & 63; + BUG_ON(word >= 3); + senddmabufmask[word] |= 1ULL << bit; + } + ipath_chg_pioavailkernel(dd, dd->ipath_lastport_piobuf, + n - dd->ipath_lastport_piobuf, 0); + ipath_write_kreg(dd, dd->ipath_kregs->kr_senddmabufmask0, + senddmabufmask[0]); + ipath_write_kreg(dd, dd->ipath_kregs->kr_senddmabufmask1, + senddmabufmask[1]); + ipath_write_kreg(dd, dd->ipath_kregs->kr_senddmabufmask2, + senddmabufmask[2]); + + INIT_LIST_HEAD(&dd->ipath_sdma_activelist); + INIT_LIST_HEAD(&dd->ipath_sdma_notifylist); + + tasklet_init(&dd->ipath_sdma_notify_task, sdma_notify_task, + (unsigned long) dd); + tasklet_init(&dd->ipath_sdma_abort_task, sdma_abort_task, + (unsigned long) dd); + + /* + * No use to turn on SDMA here, as link is probably not ACTIVE + * Just mark it RUNNING and enable the interrupt, and let the + * ipath_restart_sdma() on link transition to ACTIVE actually + * enable it. + */ + spin_lock_irqsave(&dd->ipath_sendctrl_lock, flags); + dd->ipath_sendctrl |= INFINIPATH_S_SDMAINTENABLE; + ipath_write_kreg(dd, dd->ipath_kregs->kr_sendctrl, dd->ipath_sendctrl); + ipath_read_kreg64(dd, dd->ipath_kregs->kr_scratch); + __set_bit(IPATH_SDMA_RUNNING, &dd->ipath_sdma_status); + spin_unlock_irqrestore(&dd->ipath_sendctrl_lock, flags); + +done: + return ret; +} + +void teardown_sdma(struct ipath_devdata *dd) +{ + struct ipath_sdma_txreq *txp, *txpnext; + unsigned long flags; + dma_addr_t sdma_head_phys = 0; + dma_addr_t sdma_descq_phys = 0; + void *sdma_descq = NULL; + void *sdma_head_dma = NULL; + + spin_lock_irqsave(&dd->ipath_sdma_lock, flags); + __clear_bit(IPATH_SDMA_RUNNING, &dd->ipath_sdma_status); + __set_bit(IPATH_SDMA_ABORTING, &dd->ipath_sdma_status); + __set_bit(IPATH_SDMA_SHUTDOWN, &dd->ipath_sdma_status); + spin_unlock_irqrestore(&dd->ipath_sdma_lock, flags); + + tasklet_kill(&dd->ipath_sdma_abort_task); + tasklet_kill(&dd->ipath_sdma_notify_task); + + /* turn off sdma */ + spin_lock_irqsave(&dd->ipath_sendctrl_lock, flags); + dd->ipath_sendctrl &= ~INFINIPATH_S_SDMAENABLE; + ipath_write_kreg(dd, dd->ipath_kregs->kr_sendctrl, + dd->ipath_sendctrl); + ipath_read_kreg64(dd, dd->ipath_kregs->kr_scratch); + spin_unlock_irqrestore(&dd->ipath_sendctrl_lock, flags); + + spin_lock_irqsave(&dd->ipath_sdma_lock, flags); + /* dequeue all "sent" requests */ + list_for_each_entry_safe(txp, txpnext, &dd->ipath_sdma_activelist, + list) { + txp->callback_status = IPATH_SDMA_TXREQ_S_SHUTDOWN; + if (txp->flags & IPATH_SDMA_TXREQ_F_VL15) + vl15_watchdog_deq(dd); + list_move_tail(&txp->list, &dd->ipath_sdma_notifylist); + } + spin_unlock_irqrestore(&dd->ipath_sdma_lock, flags); + + sdma_notify_taskbody(dd); + + del_timer_sync(&dd->ipath_sdma_vl15_timer); + + spin_lock_irqsave(&dd->ipath_sdma_lock, flags); + + dd->ipath_sdma_abort_jiffies = 0; + + ipath_write_kreg(dd, dd->ipath_kregs->kr_senddmabase, 0); + ipath_write_kreg(dd, dd->ipath_kregs->kr_senddmalengen, 0); + ipath_write_kreg(dd, dd->ipath_kregs->kr_senddmatail, 0); + ipath_write_kreg(dd, dd->ipath_kregs->kr_senddmaheadaddr, 0); + ipath_write_kreg(dd, dd->ipath_kregs->kr_senddmabufmask0, 0); + ipath_write_kreg(dd, dd->ipath_kregs->kr_senddmabufmask1, 0); + ipath_write_kreg(dd, dd->ipath_kregs->kr_senddmabufmask2, 0); + + if (dd->ipath_sdma_head_dma) { + sdma_head_dma = (void *) dd->ipath_sdma_head_dma; + sdma_head_phys = dd->ipath_sdma_head_phys; + dd->ipath_sdma_head_dma = NULL; + dd->ipath_sdma_head_phys = 0; + } + + if (dd->ipath_sdma_descq) { + sdma_descq = dd->ipath_sdma_descq; + sdma_descq_phys = dd->ipath_sdma_descq_phys; + dd->ipath_sdma_descq = NULL; + dd->ipath_sdma_descq_phys = 0; + } + + spin_unlock_irqrestore(&dd->ipath_sdma_lock, flags); + + if (sdma_head_dma) + dma_free_coherent(&dd->pcidev->dev, PAGE_SIZE, + sdma_head_dma, sdma_head_phys); + + if (sdma_descq) + dma_free_coherent(&dd->pcidev->dev, SDMA_DESCQ_SZ, + sdma_descq, sdma_descq_phys); +} + +/* + * [Re]start SDMA, if we use it, and it's not already OK. + * This is called on transition to link ACTIVE, either the first or + * subsequent times. + */ +void ipath_restart_sdma(struct ipath_devdata *dd) +{ + unsigned long flags; + int needed = 1; + + if (!(dd->ipath_flags & IPATH_HAS_SEND_DMA)) + goto bail; + + /* + * First, make sure we should, which is to say, + * check that we are "RUNNING" (not in teardown) + * and not "SHUTDOWN" + */ + spin_lock_irqsave(&dd->ipath_sdma_lock, flags); + if (!test_bit(IPATH_SDMA_RUNNING, &dd->ipath_sdma_status) + || test_bit(IPATH_SDMA_SHUTDOWN, &dd->ipath_sdma_status)) + needed = 0; + else { + __clear_bit(IPATH_SDMA_DISABLED, &dd->ipath_sdma_status); + __clear_bit(IPATH_SDMA_DISARMED, &dd->ipath_sdma_status); + __clear_bit(IPATH_SDMA_ABORTING, &dd->ipath_sdma_status); + } + spin_unlock_irqrestore(&dd->ipath_sdma_lock, flags); + if (!needed) { + ipath_dbg("invalid attempt to restart SDMA, status 0x%016llx\n", + dd->ipath_sdma_status); + goto bail; + } + spin_lock_irqsave(&dd->ipath_sendctrl_lock, flags); + /* + * First clear, just to be safe. Enable is only done + * in chip on 0->1 transition + */ + dd->ipath_sendctrl &= ~INFINIPATH_S_SDMAENABLE; + ipath_write_kreg(dd, dd->ipath_kregs->kr_sendctrl, dd->ipath_sendctrl); + ipath_read_kreg64(dd, dd->ipath_kregs->kr_scratch); + dd->ipath_sendctrl |= INFINIPATH_S_SDMAENABLE; + ipath_write_kreg(dd, dd->ipath_kregs->kr_sendctrl, dd->ipath_sendctrl); + ipath_read_kreg64(dd, dd->ipath_kregs->kr_scratch); + spin_unlock_irqrestore(&dd->ipath_sendctrl_lock, flags); + +bail: + return; +} + +static inline void make_sdma_desc(struct ipath_devdata *dd, + u64 *sdmadesc, u64 addr, u64 dwlen, u64 dwoffset) +{ + WARN_ON(addr & 3); + /* SDmaPhyAddr[47:32] */ + sdmadesc[1] = addr >> 32; + /* SDmaPhyAddr[31:0] */ + sdmadesc[0] = (addr & 0xfffffffcULL) << 32; + /* SDmaGeneration[1:0] */ + sdmadesc[0] |= (dd->ipath_sdma_generation & 3ULL) << 30; + /* SDmaDwordCount[10:0] */ + sdmadesc[0] |= (dwlen & 0x7ffULL) << 16; + /* SDmaBufOffset[12:2] */ + sdmadesc[0] |= dwoffset & 0x7ffULL; +} + +/* + * This function queues one IB packet onto the send DMA queue per call. + * The caller is responsible for checking: + * 1) The number of send DMA descriptor entries is less than the size of + * the descriptor queue. + * 2) The IB SGE addresses and lengths are 32-bit aligned + * (except possibly the last SGE's length) + * 3) The SGE addresses are suitable for passing to dma_map_single(). + */ +int ipath_sdma_verbs_send(struct ipath_devdata *dd, + struct ipath_sge_state *ss, u32 dwords, + struct ipath_verbs_txreq *tx) +{ + + unsigned long flags; + struct ipath_sge *sge; + int ret = 0; + u16 tail; + __le64 *descqp; + u64 sdmadesc[2]; + u32 dwoffset; + dma_addr_t addr; + + if ((tx->map_len + (dwords<<2)) > dd->ipath_ibmaxlen) { + ipath_dbg("packet size %X > ibmax %X, fail\n", + tx->map_len + (dwords<<2), dd->ipath_ibmaxlen); + ret = -EMSGSIZE; + goto fail; + } + + spin_lock_irqsave(&dd->ipath_sdma_lock, flags); + +retry: + if (unlikely(test_bit(IPATH_SDMA_ABORTING, &dd->ipath_sdma_status))) { + ret = -EBUSY; + goto unlock; + } + + if (tx->txreq.sg_count > ipath_sdma_descq_freecnt(dd)) { + if (ipath_sdma_make_progress(dd)) + goto retry; + ret = -ENOBUFS; + goto unlock; + } + + addr = dma_map_single(&dd->pcidev->dev, tx->txreq.map_addr, + tx->map_len, DMA_TO_DEVICE); + if (dma_mapping_error(addr)) { + ret = -EIO; + goto unlock; + } + + dwoffset = tx->map_len >> 2; + make_sdma_desc(dd, sdmadesc, (u64) addr, dwoffset, 0); + + /* SDmaFirstDesc */ + sdmadesc[0] |= 1ULL << 12; + if (tx->txreq.flags & IPATH_SDMA_TXREQ_F_USELARGEBUF) + sdmadesc[0] |= 1ULL << 14; /* SDmaUseLargeBuf */ + + /* write to the descq */ + tail = dd->ipath_sdma_descq_tail; + descqp = &dd->ipath_sdma_descq[tail].qw[0]; + *descqp++ = cpu_to_le64(sdmadesc[0]); + *descqp++ = cpu_to_le64(sdmadesc[1]); + + if (tx->txreq.flags & IPATH_SDMA_TXREQ_F_FREEDESC) + tx->txreq.start_idx = tail; + + /* increment the tail */ + if (++tail == dd->ipath_sdma_descq_cnt) { + tail = 0; + descqp = &dd->ipath_sdma_descq[0].qw[0]; + ++dd->ipath_sdma_generation; + } + + sge = &ss->sge; + while (dwords) { + u32 dw; + u32 len; + + len = dwords << 2; + if (len > sge->length) + len = sge->length; + if (len > sge->sge_length) + len = sge->sge_length; + BUG_ON(len == 0); + dw = (len + 3) >> 2; + addr = dma_map_single(&dd->pcidev->dev, sge->vaddr, dw << 2, + DMA_TO_DEVICE); + make_sdma_desc(dd, sdmadesc, (u64) addr, dw, dwoffset); + /* SDmaUseLargeBuf has to be set in every descriptor */ + if (tx->txreq.flags & IPATH_SDMA_TXREQ_F_USELARGEBUF) + sdmadesc[0] |= 1ULL << 14; + /* write to the descq */ + *descqp++ = cpu_to_le64(sdmadesc[0]); + *descqp++ = cpu_to_le64(sdmadesc[1]); + + /* increment the tail */ + if (++tail == dd->ipath_sdma_descq_cnt) { + tail = 0; + descqp = &dd->ipath_sdma_descq[0].qw[0]; + ++dd->ipath_sdma_generation; + } + sge->vaddr += len; + sge->length -= len; + sge->sge_length -= len; + if (sge->sge_length == 0) { + if (--ss->num_sge) + *sge = *ss->sg_list++; + } else if (sge->length == 0 && sge->mr != NULL) { + if (++sge->n >= IPATH_SEGSZ) { + if (++sge->m >= sge->mr->mapsz) + break; + sge->n = 0; + } + sge->vaddr = + sge->mr->map[sge->m]->segs[sge->n].vaddr; + sge->length = + sge->mr->map[sge->m]->segs[sge->n].length; + } + + dwoffset += dw; + dwords -= dw; + } + + if (!tail) + descqp = &dd->ipath_sdma_descq[dd->ipath_sdma_descq_cnt].qw[0]; + descqp -= 2; + /* SDmaLastDesc */ + descqp[0] |= __constant_cpu_to_le64(1ULL << 11); + if (tx->txreq.flags & IPATH_SDMA_TXREQ_F_INTREQ) { + /* SDmaIntReq */ + descqp[0] |= __constant_cpu_to_le64(1ULL << 15); + } + + /* Commit writes to memory and advance the tail on the chip */ + wmb(); + ipath_write_kreg(dd, dd->ipath_kregs->kr_senddmatail, tail); + + tx->txreq.next_descq_idx = tail; + tx->txreq.callback_status = IPATH_SDMA_TXREQ_S_OK; + dd->ipath_sdma_descq_tail = tail; + dd->ipath_sdma_descq_added += tx->txreq.sg_count; + list_add_tail(&tx->txreq.list, &dd->ipath_sdma_activelist); + if (tx->txreq.flags & IPATH_SDMA_TXREQ_F_VL15) + vl15_watchdog_enq(dd); + +unlock: + spin_unlock_irqrestore(&dd->ipath_sdma_lock, flags); +fail: + return ret; +} diff --git a/drivers/infiniband/hw/ipath/ipath_srq.c b/drivers/infiniband/hw/ipath/ipath_srq.c index f772102e4713..e3d80ca84c1a 100644 --- a/drivers/infiniband/hw/ipath/ipath_srq.c +++ b/drivers/infiniband/hw/ipath/ipath_srq.c @@ -1,5 +1,5 @@ /* - * Copyright (c) 2006, 2007 QLogic Corporation. All rights reserved. + * Copyright (c) 2006, 2007, 2008 QLogic Corporation. All rights reserved. * Copyright (c) 2005, 2006 PathScale, Inc. All rights reserved. * * This software is available to you under a choice of one of two @@ -245,7 +245,8 @@ int ipath_modify_srq(struct ib_srq *ibsrq, struct ib_srq_attr *attr, sizeof(offset_addr)); if (ret) goto bail_free; - udata->outbuf = (void __user *) offset_addr; + udata->outbuf = + (void __user *) (unsigned long) offset_addr; ret = ib_copy_to_udata(udata, &offset, sizeof(offset)); if (ret) diff --git a/drivers/infiniband/hw/ipath/ipath_stats.c b/drivers/infiniband/hw/ipath/ipath_stats.c index d2725cd11bdc..c8e3d65f0de8 100644 --- a/drivers/infiniband/hw/ipath/ipath_stats.c +++ b/drivers/infiniband/hw/ipath/ipath_stats.c @@ -1,5 +1,5 @@ /* - * Copyright (c) 2006, 2007 QLogic Corporation. All rights reserved. + * Copyright (c) 2006, 2007, 2008 QLogic Corporation. All rights reserved. * Copyright (c) 2003, 2004, 2005, 2006 PathScale, Inc. All rights reserved. * * This software is available to you under a choice of one of two @@ -136,6 +136,7 @@ static void ipath_qcheck(struct ipath_devdata *dd) struct ipath_portdata *pd = dd->ipath_pd[0]; size_t blen = 0; char buf[128]; + u32 hdrqtail; *buf = 0; if (pd->port_hdrqfull != dd->ipath_p0_hdrqfull) { @@ -174,17 +175,18 @@ static void ipath_qcheck(struct ipath_devdata *dd) if (blen) ipath_dbg("%s\n", buf); - if (pd->port_head != (u32) - le64_to_cpu(*dd->ipath_hdrqtailptr)) { + hdrqtail = ipath_get_hdrqtail(pd); + if (pd->port_head != hdrqtail) { if (dd->ipath_lastport0rcv_cnt == ipath_stats.sps_port0pkts) { ipath_cdbg(PKT, "missing rcv interrupts? " - "port0 hd=%llx tl=%x; port0pkts %llx\n", - (unsigned long long) - le64_to_cpu(*dd->ipath_hdrqtailptr), - pd->port_head, + "port0 hd=%x tl=%x; port0pkts %llx; write" + " hd (w/intr)\n", + pd->port_head, hdrqtail, (unsigned long long) ipath_stats.sps_port0pkts); + ipath_write_ureg(dd, ur_rcvhdrhead, hdrqtail | + dd->ipath_rhdrhead_intr_off, pd->port_port); } dd->ipath_lastport0rcv_cnt = ipath_stats.sps_port0pkts; } @@ -290,11 +292,11 @@ void ipath_get_faststats(unsigned long opaque) && time_after(jiffies, dd->ipath_unmasktime)) { char ebuf[256]; int iserr; - iserr = ipath_decode_err(ebuf, sizeof ebuf, - dd->ipath_maskederrs); + iserr = ipath_decode_err(dd, ebuf, sizeof ebuf, + dd->ipath_maskederrs); if (dd->ipath_maskederrs & - ~(INFINIPATH_E_RRCVEGRFULL | INFINIPATH_E_RRCVHDRFULL | - INFINIPATH_E_PKTERRS )) + ~(INFINIPATH_E_RRCVEGRFULL | INFINIPATH_E_RRCVHDRFULL | + INFINIPATH_E_PKTERRS)) ipath_dev_err(dd, "Re-enabling masked errors " "(%s)\n", ebuf); else { @@ -306,17 +308,18 @@ void ipath_get_faststats(unsigned long opaque) * level. */ if (iserr) - ipath_dbg("Re-enabling queue full errors (%s)\n", - ebuf); + ipath_dbg( + "Re-enabling queue full errors (%s)\n", + ebuf); else ipath_cdbg(ERRPKT, "Re-enabling packet" - " problem interrupt (%s)\n", ebuf); + " problem interrupt (%s)\n", ebuf); } /* re-enable masked errors */ dd->ipath_errormask |= dd->ipath_maskederrs; ipath_write_kreg(dd, dd->ipath_kregs->kr_errormask, - dd->ipath_errormask); + dd->ipath_errormask); dd->ipath_maskederrs = 0; } diff --git a/drivers/infiniband/hw/ipath/ipath_sysfs.c b/drivers/infiniband/hw/ipath/ipath_sysfs.c index 56dfc8a2344c..a6c8efbdc0c9 100644 --- a/drivers/infiniband/hw/ipath/ipath_sysfs.c +++ b/drivers/infiniband/hw/ipath/ipath_sysfs.c @@ -1,5 +1,5 @@ /* - * Copyright (c) 2006, 2007 QLogic Corporation. All rights reserved. + * Copyright (c) 2006, 2007, 2008 QLogic Corporation. All rights reserved. * Copyright (c) 2006 PathScale, Inc. All rights reserved. * * This software is available to you under a choice of one of two @@ -34,6 +34,7 @@ #include <linux/ctype.h> #include "ipath_kernel.h" +#include "ipath_verbs.h" #include "ipath_common.h" /** @@ -163,6 +164,15 @@ static ssize_t show_boardversion(struct device *dev, return scnprintf(buf, PAGE_SIZE, "%s", dd->ipath_boardversion); } +static ssize_t show_localbus_info(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct ipath_devdata *dd = dev_get_drvdata(dev); + /* The string printed here is already newline-terminated. */ + return scnprintf(buf, PAGE_SIZE, "%s", dd->ipath_lbus_info); +} + static ssize_t show_lmc(struct device *dev, struct device_attribute *attr, char *buf) @@ -311,6 +321,8 @@ static ssize_t store_guid(struct device *dev, dd->ipath_guid = new_guid; dd->ipath_nguid = 1; + if (dd->verbs_dev) + dd->verbs_dev->ibdev.node_guid = new_guid; ret = strlen(buf); goto bail; @@ -919,21 +931,21 @@ static ssize_t store_rx_polinv_enb(struct device *dev, u16 val; ret = ipath_parse_ushort(buf, &val); - if (ret < 0 || val > 1) - goto invalid; + if (ret >= 0 && val > 1) { + ipath_dev_err(dd, + "attempt to set invalid Rx Polarity (enable)\n"); + ret = -EINVAL; + goto bail; + } r = dd->ipath_f_set_ib_cfg(dd, IPATH_IB_CFG_RXPOL_ENB, val); - if (r < 0) { + if (r < 0) ret = r; - goto bail; - } - goto bail; -invalid: - ipath_dev_err(dd, "attempt to set invalid Rx Polarity (enable)\n"); bail: return ret; } + /* * Get/Set RX lane-reversal enable. 0=no, 1=yes. */ @@ -988,6 +1000,75 @@ static struct attribute_group driver_attr_group = { .attrs = driver_attributes }; +static ssize_t store_tempsense(struct device *dev, + struct device_attribute *attr, + const char *buf, + size_t count) +{ + struct ipath_devdata *dd = dev_get_drvdata(dev); + int ret, stat; + u16 val; + + ret = ipath_parse_ushort(buf, &val); + if (ret <= 0) { + ipath_dev_err(dd, "attempt to set invalid tempsense config\n"); + goto bail; + } + /* If anything but the highest limit, enable T_CRIT_A "interrupt" */ + stat = ipath_tempsense_write(dd, 9, (val == 0x7f7f) ? 0x80 : 0); + if (stat) { + ipath_dev_err(dd, "Unable to set tempsense config\n"); + ret = -1; + goto bail; + } + stat = ipath_tempsense_write(dd, 0xB, (u8) (val & 0xFF)); + if (stat) { + ipath_dev_err(dd, "Unable to set local Tcrit\n"); + ret = -1; + goto bail; + } + stat = ipath_tempsense_write(dd, 0xD, (u8) (val >> 8)); + if (stat) { + ipath_dev_err(dd, "Unable to set remote Tcrit\n"); + ret = -1; + goto bail; + } + +bail: + return ret; +} + +/* + * dump tempsense regs. in decimal, to ease shell-scripts. + */ +static ssize_t show_tempsense(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct ipath_devdata *dd = dev_get_drvdata(dev); + int ret; + int idx; + u8 regvals[8]; + + ret = -ENXIO; + for (idx = 0; idx < 8; ++idx) { + if (idx == 6) + continue; + ret = ipath_tempsense_read(dd, idx); + if (ret < 0) + break; + regvals[idx] = ret; + } + if (idx == 8) + ret = scnprintf(buf, PAGE_SIZE, "%d %d %02X %02X %d %d\n", + *(signed char *)(regvals), + *(signed char *)(regvals + 1), + regvals[2], regvals[3], + *(signed char *)(regvals + 5), + *(signed char *)(regvals + 7)); + return ret; +} + struct attribute_group *ipath_driver_attr_groups[] = { &driver_attr_group, NULL, @@ -1011,10 +1092,13 @@ static DEVICE_ATTR(unit, S_IRUGO, show_unit, NULL); static DEVICE_ATTR(rx_pol_inv, S_IWUSR, NULL, store_rx_pol_inv); static DEVICE_ATTR(led_override, S_IWUSR, NULL, store_led_override); static DEVICE_ATTR(logged_errors, S_IRUGO, show_logged_errs, NULL); +static DEVICE_ATTR(localbus_info, S_IRUGO, show_localbus_info, NULL); static DEVICE_ATTR(jint_max_packets, S_IWUSR | S_IRUGO, show_jint_max_packets, store_jint_max_packets); static DEVICE_ATTR(jint_idle_ticks, S_IWUSR | S_IRUGO, show_jint_idle_ticks, store_jint_idle_ticks); +static DEVICE_ATTR(tempsense, S_IWUSR | S_IRUGO, + show_tempsense, store_tempsense); static struct attribute *dev_attributes[] = { &dev_attr_guid.attr, @@ -1034,6 +1118,8 @@ static struct attribute *dev_attributes[] = { &dev_attr_rx_pol_inv.attr, &dev_attr_led_override.attr, &dev_attr_logged_errors.attr, + &dev_attr_tempsense.attr, + &dev_attr_localbus_info.attr, NULL }; diff --git a/drivers/infiniband/hw/ipath/ipath_uc.c b/drivers/infiniband/hw/ipath/ipath_uc.c index 2dd8de20d221..bfe8926b5514 100644 --- a/drivers/infiniband/hw/ipath/ipath_uc.c +++ b/drivers/infiniband/hw/ipath/ipath_uc.c @@ -94,7 +94,7 @@ int ipath_make_uc_req(struct ipath_qp *qp) qp->s_state = OP(SEND_ONLY_WITH_IMMEDIATE); /* Immediate data comes after the BTH */ - ohdr->u.imm_data = wqe->wr.imm_data; + ohdr->u.imm_data = wqe->wr.ex.imm_data; hwords += 1; } if (wqe->wr.send_flags & IB_SEND_SOLICITED) @@ -123,7 +123,7 @@ int ipath_make_uc_req(struct ipath_qp *qp) qp->s_state = OP(RDMA_WRITE_ONLY_WITH_IMMEDIATE); /* Immediate data comes after the RETH */ - ohdr->u.rc.imm_data = wqe->wr.imm_data; + ohdr->u.rc.imm_data = wqe->wr.ex.imm_data; hwords += 1; if (wqe->wr.send_flags & IB_SEND_SOLICITED) bth0 |= 1 << 23; @@ -152,7 +152,7 @@ int ipath_make_uc_req(struct ipath_qp *qp) else { qp->s_state = OP(SEND_LAST_WITH_IMMEDIATE); /* Immediate data comes after the BTH */ - ohdr->u.imm_data = wqe->wr.imm_data; + ohdr->u.imm_data = wqe->wr.ex.imm_data; hwords += 1; } if (wqe->wr.send_flags & IB_SEND_SOLICITED) @@ -177,7 +177,7 @@ int ipath_make_uc_req(struct ipath_qp *qp) qp->s_state = OP(RDMA_WRITE_LAST_WITH_IMMEDIATE); /* Immediate data comes after the BTH */ - ohdr->u.imm_data = wqe->wr.imm_data; + ohdr->u.imm_data = wqe->wr.ex.imm_data; hwords += 1; if (wqe->wr.send_flags & IB_SEND_SOLICITED) bth0 |= 1 << 23; diff --git a/drivers/infiniband/hw/ipath/ipath_ud.c b/drivers/infiniband/hw/ipath/ipath_ud.c index de67eed08ed0..8b6a261c89e3 100644 --- a/drivers/infiniband/hw/ipath/ipath_ud.c +++ b/drivers/infiniband/hw/ipath/ipath_ud.c @@ -1,5 +1,5 @@ /* - * Copyright (c) 2006, 2007 QLogic Corporation. All rights reserved. + * Copyright (c) 2006, 2007, 2008 QLogic Corporation. All rights reserved. * Copyright (c) 2005, 2006 PathScale, Inc. All rights reserved. * * This software is available to you under a choice of one of two @@ -95,7 +95,7 @@ static void ipath_ud_loopback(struct ipath_qp *sqp, struct ipath_swqe *swqe) if (swqe->wr.opcode == IB_WR_SEND_WITH_IMM) { wc.wc_flags = IB_WC_WITH_IMM; - wc.imm_data = swqe->wr.imm_data; + wc.imm_data = swqe->wr.ex.imm_data; } else { wc.wc_flags = 0; wc.imm_data = 0; @@ -303,6 +303,7 @@ int ipath_make_ud_req(struct ipath_qp *qp) qp->s_hdrwords = 7; qp->s_cur_size = wqe->length; qp->s_cur_sge = &qp->s_sge; + qp->s_dmult = ah_attr->static_rate; qp->s_wqe = wqe; qp->s_sge.sge = wqe->sg_list[0]; qp->s_sge.sg_list = wqe->sg_list + 1; @@ -326,7 +327,7 @@ int ipath_make_ud_req(struct ipath_qp *qp) } if (wqe->wr.opcode == IB_WR_SEND_WITH_IMM) { qp->s_hdrwords++; - ohdr->u.ud.imm_data = wqe->wr.imm_data; + ohdr->u.ud.imm_data = wqe->wr.ex.imm_data; bth0 = IB_OPCODE_UD_SEND_ONLY_WITH_IMMEDIATE << 24; } else bth0 = IB_OPCODE_UD_SEND_ONLY << 24; diff --git a/drivers/infiniband/hw/ipath/ipath_user_sdma.c b/drivers/infiniband/hw/ipath/ipath_user_sdma.c new file mode 100644 index 000000000000..86e016916cd1 --- /dev/null +++ b/drivers/infiniband/hw/ipath/ipath_user_sdma.c @@ -0,0 +1,879 @@ +/* + * Copyright (c) 2007, 2008 QLogic Corporation. All rights reserved. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * OpenIB.org BSD license below: + * + * Redistribution and use in source and binary forms, with or + * without modification, are permitted provided that the following + * conditions are met: + * + * - Redistributions of source code must retain the above + * copyright notice, this list of conditions and the following + * disclaimer. + * + * - Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials + * provided with the distribution. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + */ +#include <linux/mm.h> +#include <linux/types.h> +#include <linux/device.h> +#include <linux/dmapool.h> +#include <linux/slab.h> +#include <linux/list.h> +#include <linux/highmem.h> +#include <linux/io.h> +#include <linux/uio.h> +#include <linux/rbtree.h> +#include <linux/spinlock.h> +#include <linux/delay.h> + +#include "ipath_kernel.h" +#include "ipath_user_sdma.h" + +/* minimum size of header */ +#define IPATH_USER_SDMA_MIN_HEADER_LENGTH 64 +/* expected size of headers (for dma_pool) */ +#define IPATH_USER_SDMA_EXP_HEADER_LENGTH 64 +/* length mask in PBC (lower 11 bits) */ +#define IPATH_PBC_LENGTH_MASK ((1 << 11) - 1) + +struct ipath_user_sdma_pkt { + u8 naddr; /* dimension of addr (1..3) ... */ + u32 counter; /* sdma pkts queued counter for this entry */ + u64 added; /* global descq number of entries */ + + struct { + u32 offset; /* offset for kvaddr, addr */ + u32 length; /* length in page */ + u8 put_page; /* should we put_page? */ + u8 dma_mapped; /* is page dma_mapped? */ + struct page *page; /* may be NULL (coherent mem) */ + void *kvaddr; /* FIXME: only for pio hack */ + dma_addr_t addr; + } addr[4]; /* max pages, any more and we coalesce */ + struct list_head list; /* list element */ +}; + +struct ipath_user_sdma_queue { + /* + * pkts sent to dma engine are queued on this + * list head. the type of the elements of this + * list are struct ipath_user_sdma_pkt... + */ + struct list_head sent; + + /* headers with expected length are allocated from here... */ + char header_cache_name[64]; + struct dma_pool *header_cache; + + /* packets are allocated from the slab cache... */ + char pkt_slab_name[64]; + struct kmem_cache *pkt_slab; + + /* as packets go on the queued queue, they are counted... */ + u32 counter; + u32 sent_counter; + + /* dma page table */ + struct rb_root dma_pages_root; + + /* protect everything above... */ + struct mutex lock; +}; + +struct ipath_user_sdma_queue * +ipath_user_sdma_queue_create(struct device *dev, int unit, int port, int sport) +{ + struct ipath_user_sdma_queue *pq = + kmalloc(sizeof(struct ipath_user_sdma_queue), GFP_KERNEL); + + if (!pq) + goto done; + + pq->counter = 0; + pq->sent_counter = 0; + INIT_LIST_HEAD(&pq->sent); + + mutex_init(&pq->lock); + + snprintf(pq->pkt_slab_name, sizeof(pq->pkt_slab_name), + "ipath-user-sdma-pkts-%u-%02u.%02u", unit, port, sport); + pq->pkt_slab = kmem_cache_create(pq->pkt_slab_name, + sizeof(struct ipath_user_sdma_pkt), + 0, 0, NULL); + + if (!pq->pkt_slab) + goto err_kfree; + + snprintf(pq->header_cache_name, sizeof(pq->header_cache_name), + "ipath-user-sdma-headers-%u-%02u.%02u", unit, port, sport); + pq->header_cache = dma_pool_create(pq->header_cache_name, + dev, + IPATH_USER_SDMA_EXP_HEADER_LENGTH, + 4, 0); + if (!pq->header_cache) + goto err_slab; + + pq->dma_pages_root = RB_ROOT; + + goto done; + +err_slab: + kmem_cache_destroy(pq->pkt_slab); +err_kfree: + kfree(pq); + pq = NULL; + +done: + return pq; +} + +static void ipath_user_sdma_init_frag(struct ipath_user_sdma_pkt *pkt, + int i, size_t offset, size_t len, + int put_page, int dma_mapped, + struct page *page, + void *kvaddr, dma_addr_t dma_addr) +{ + pkt->addr[i].offset = offset; + pkt->addr[i].length = len; + pkt->addr[i].put_page = put_page; + pkt->addr[i].dma_mapped = dma_mapped; + pkt->addr[i].page = page; + pkt->addr[i].kvaddr = kvaddr; + pkt->addr[i].addr = dma_addr; +} + +static void ipath_user_sdma_init_header(struct ipath_user_sdma_pkt *pkt, + u32 counter, size_t offset, + size_t len, int dma_mapped, + struct page *page, + void *kvaddr, dma_addr_t dma_addr) +{ + pkt->naddr = 1; + pkt->counter = counter; + ipath_user_sdma_init_frag(pkt, 0, offset, len, 0, dma_mapped, page, + kvaddr, dma_addr); +} + +/* we've too many pages in the iovec, coalesce to a single page */ +static int ipath_user_sdma_coalesce(const struct ipath_devdata *dd, + struct ipath_user_sdma_pkt *pkt, + const struct iovec *iov, + unsigned long niov) { + int ret = 0; + struct page *page = alloc_page(GFP_KERNEL); + void *mpage_save; + char *mpage; + int i; + int len = 0; + dma_addr_t dma_addr; + + if (!page) { + ret = -ENOMEM; + goto done; + } + + mpage = kmap(page); + mpage_save = mpage; + for (i = 0; i < niov; i++) { + int cfur; + + cfur = copy_from_user(mpage, + iov[i].iov_base, iov[i].iov_len); + if (cfur) { + ret = -EFAULT; + goto free_unmap; + } + + mpage += iov[i].iov_len; + len += iov[i].iov_len; + } + + dma_addr = dma_map_page(&dd->pcidev->dev, page, 0, len, + DMA_TO_DEVICE); + if (dma_mapping_error(dma_addr)) { + ret = -ENOMEM; + goto free_unmap; + } + + ipath_user_sdma_init_frag(pkt, 1, 0, len, 0, 1, page, mpage_save, + dma_addr); + pkt->naddr = 2; + + goto done; + +free_unmap: + kunmap(page); + __free_page(page); +done: + return ret; +} + +/* how many pages in this iovec element? */ +static int ipath_user_sdma_num_pages(const struct iovec *iov) +{ + const unsigned long addr = (unsigned long) iov->iov_base; + const unsigned long len = iov->iov_len; + const unsigned long spage = addr & PAGE_MASK; + const unsigned long epage = (addr + len - 1) & PAGE_MASK; + + return 1 + ((epage - spage) >> PAGE_SHIFT); +} + +/* truncate length to page boundry */ +static int ipath_user_sdma_page_length(unsigned long addr, unsigned long len) +{ + const unsigned long offset = addr & ~PAGE_MASK; + + return ((offset + len) > PAGE_SIZE) ? (PAGE_SIZE - offset) : len; +} + +static void ipath_user_sdma_free_pkt_frag(struct device *dev, + struct ipath_user_sdma_queue *pq, + struct ipath_user_sdma_pkt *pkt, + int frag) +{ + const int i = frag; + + if (pkt->addr[i].page) { + if (pkt->addr[i].dma_mapped) + dma_unmap_page(dev, + pkt->addr[i].addr, + pkt->addr[i].length, + DMA_TO_DEVICE); + + if (pkt->addr[i].kvaddr) + kunmap(pkt->addr[i].page); + + if (pkt->addr[i].put_page) + put_page(pkt->addr[i].page); + else + __free_page(pkt->addr[i].page); + } else if (pkt->addr[i].kvaddr) + /* free coherent mem from cache... */ + dma_pool_free(pq->header_cache, + pkt->addr[i].kvaddr, pkt->addr[i].addr); +} + +/* return number of pages pinned... */ +static int ipath_user_sdma_pin_pages(const struct ipath_devdata *dd, + struct ipath_user_sdma_pkt *pkt, + unsigned long addr, int tlen, int npages) +{ + struct page *pages[2]; + int j; + int ret; + + ret = get_user_pages(current, current->mm, addr, + npages, 0, 1, pages, NULL); + + if (ret != npages) { + int i; + + for (i = 0; i < ret; i++) + put_page(pages[i]); + + ret = -ENOMEM; + goto done; + } + + for (j = 0; j < npages; j++) { + /* map the pages... */ + const int flen = + ipath_user_sdma_page_length(addr, tlen); + dma_addr_t dma_addr = + dma_map_page(&dd->pcidev->dev, + pages[j], 0, flen, DMA_TO_DEVICE); + unsigned long fofs = addr & ~PAGE_MASK; + + if (dma_mapping_error(dma_addr)) { + ret = -ENOMEM; + goto done; + } + + ipath_user_sdma_init_frag(pkt, pkt->naddr, fofs, flen, 1, 1, + pages[j], kmap(pages[j]), + dma_addr); + + pkt->naddr++; + addr += flen; + tlen -= flen; + } + +done: + return ret; +} + +static int ipath_user_sdma_pin_pkt(const struct ipath_devdata *dd, + struct ipath_user_sdma_queue *pq, + struct ipath_user_sdma_pkt *pkt, + const struct iovec *iov, + unsigned long niov) +{ + int ret = 0; + unsigned long idx; + + for (idx = 0; idx < niov; idx++) { + const int npages = ipath_user_sdma_num_pages(iov + idx); + const unsigned long addr = (unsigned long) iov[idx].iov_base; + + ret = ipath_user_sdma_pin_pages(dd, pkt, + addr, iov[idx].iov_len, + npages); + if (ret < 0) + goto free_pkt; + } + + goto done; + +free_pkt: + for (idx = 0; idx < pkt->naddr; idx++) + ipath_user_sdma_free_pkt_frag(&dd->pcidev->dev, pq, pkt, idx); + +done: + return ret; +} + +static int ipath_user_sdma_init_payload(const struct ipath_devdata *dd, + struct ipath_user_sdma_queue *pq, + struct ipath_user_sdma_pkt *pkt, + const struct iovec *iov, + unsigned long niov, int npages) +{ + int ret = 0; + + if (npages >= ARRAY_SIZE(pkt->addr)) + ret = ipath_user_sdma_coalesce(dd, pkt, iov, niov); + else + ret = ipath_user_sdma_pin_pkt(dd, pq, pkt, iov, niov); + + return ret; +} + +/* free a packet list -- return counter value of last packet */ +static void ipath_user_sdma_free_pkt_list(struct device *dev, + struct ipath_user_sdma_queue *pq, + struct list_head *list) +{ + struct ipath_user_sdma_pkt *pkt, *pkt_next; + + list_for_each_entry_safe(pkt, pkt_next, list, list) { + int i; + + for (i = 0; i < pkt->naddr; i++) + ipath_user_sdma_free_pkt_frag(dev, pq, pkt, i); + + kmem_cache_free(pq->pkt_slab, pkt); + } +} + +/* + * copy headers, coalesce etc -- pq->lock must be held + * + * we queue all the packets to list, returning the + * number of bytes total. list must be empty initially, + * as, if there is an error we clean it... + */ +static int ipath_user_sdma_queue_pkts(const struct ipath_devdata *dd, + struct ipath_user_sdma_queue *pq, + struct list_head *list, + const struct iovec *iov, + unsigned long niov, + int maxpkts) +{ + unsigned long idx = 0; + int ret = 0; + int npkts = 0; + struct page *page = NULL; + __le32 *pbc; + dma_addr_t dma_addr; + struct ipath_user_sdma_pkt *pkt = NULL; + size_t len; + size_t nw; + u32 counter = pq->counter; + int dma_mapped = 0; + + while (idx < niov && npkts < maxpkts) { + const unsigned long addr = (unsigned long) iov[idx].iov_base; + const unsigned long idx_save = idx; + unsigned pktnw; + unsigned pktnwc; + int nfrags = 0; + int npages = 0; + int cfur; + + dma_mapped = 0; + len = iov[idx].iov_len; + nw = len >> 2; + page = NULL; + + pkt = kmem_cache_alloc(pq->pkt_slab, GFP_KERNEL); + if (!pkt) { + ret = -ENOMEM; + goto free_list; + } + + if (len < IPATH_USER_SDMA_MIN_HEADER_LENGTH || + len > PAGE_SIZE || len & 3 || addr & 3) { + ret = -EINVAL; + goto free_pkt; + } + + if (len == IPATH_USER_SDMA_EXP_HEADER_LENGTH) + pbc = dma_pool_alloc(pq->header_cache, GFP_KERNEL, + &dma_addr); + else + pbc = NULL; + + if (!pbc) { + page = alloc_page(GFP_KERNEL); + if (!page) { + ret = -ENOMEM; + goto free_pkt; + } + pbc = kmap(page); + } + + cfur = copy_from_user(pbc, iov[idx].iov_base, len); + if (cfur) { + ret = -EFAULT; + goto free_pbc; + } + + /* + * this assignment is a bit strange. it's because the + * the pbc counts the number of 32 bit words in the full + * packet _except_ the first word of the pbc itself... + */ + pktnwc = nw - 1; + + /* + * pktnw computation yields the number of 32 bit words + * that the caller has indicated in the PBC. note that + * this is one less than the total number of words that + * goes to the send DMA engine as the first 32 bit word + * of the PBC itself is not counted. Armed with this count, + * we can verify that the packet is consistent with the + * iovec lengths. + */ + pktnw = le32_to_cpu(*pbc) & IPATH_PBC_LENGTH_MASK; + if (pktnw < pktnwc || pktnw > pktnwc + (PAGE_SIZE >> 2)) { + ret = -EINVAL; + goto free_pbc; + } + + + idx++; + while (pktnwc < pktnw && idx < niov) { + const size_t slen = iov[idx].iov_len; + const unsigned long faddr = + (unsigned long) iov[idx].iov_base; + + if (slen & 3 || faddr & 3 || !slen || + slen > PAGE_SIZE) { + ret = -EINVAL; + goto free_pbc; + } + + npages++; + if ((faddr & PAGE_MASK) != + ((faddr + slen - 1) & PAGE_MASK)) + npages++; + + pktnwc += slen >> 2; + idx++; + nfrags++; + } + + if (pktnwc != pktnw) { + ret = -EINVAL; + goto free_pbc; + } + + if (page) { + dma_addr = dma_map_page(&dd->pcidev->dev, + page, 0, len, DMA_TO_DEVICE); + if (dma_mapping_error(dma_addr)) { + ret = -ENOMEM; + goto free_pbc; + } + + dma_mapped = 1; + } + + ipath_user_sdma_init_header(pkt, counter, 0, len, dma_mapped, + page, pbc, dma_addr); + + if (nfrags) { + ret = ipath_user_sdma_init_payload(dd, pq, pkt, + iov + idx_save + 1, + nfrags, npages); + if (ret < 0) + goto free_pbc_dma; + } + + counter++; + npkts++; + + list_add_tail(&pkt->list, list); + } + + ret = idx; + goto done; + +free_pbc_dma: + if (dma_mapped) + dma_unmap_page(&dd->pcidev->dev, dma_addr, len, DMA_TO_DEVICE); +free_pbc: + if (page) { + kunmap(page); + __free_page(page); + } else + dma_pool_free(pq->header_cache, pbc, dma_addr); +free_pkt: + kmem_cache_free(pq->pkt_slab, pkt); +free_list: + ipath_user_sdma_free_pkt_list(&dd->pcidev->dev, pq, list); +done: + return ret; +} + +static void ipath_user_sdma_set_complete_counter(struct ipath_user_sdma_queue *pq, + u32 c) +{ + pq->sent_counter = c; +} + +/* try to clean out queue -- needs pq->lock */ +static int ipath_user_sdma_queue_clean(const struct ipath_devdata *dd, + struct ipath_user_sdma_queue *pq) +{ + struct list_head free_list; + struct ipath_user_sdma_pkt *pkt; + struct ipath_user_sdma_pkt *pkt_prev; + int ret = 0; + + INIT_LIST_HEAD(&free_list); + + list_for_each_entry_safe(pkt, pkt_prev, &pq->sent, list) { + s64 descd = dd->ipath_sdma_descq_removed - pkt->added; + + if (descd < 0) + break; + + list_move_tail(&pkt->list, &free_list); + + /* one more packet cleaned */ + ret++; + } + + if (!list_empty(&free_list)) { + u32 counter; + + pkt = list_entry(free_list.prev, + struct ipath_user_sdma_pkt, list); + counter = pkt->counter; + + ipath_user_sdma_free_pkt_list(&dd->pcidev->dev, pq, &free_list); + ipath_user_sdma_set_complete_counter(pq, counter); + } + + return ret; +} + +void ipath_user_sdma_queue_destroy(struct ipath_user_sdma_queue *pq) +{ + if (!pq) + return; + + kmem_cache_destroy(pq->pkt_slab); + dma_pool_destroy(pq->header_cache); + kfree(pq); +} + +/* clean descriptor queue, returns > 0 if some elements cleaned */ +static int ipath_user_sdma_hwqueue_clean(struct ipath_devdata *dd) +{ + int ret; + unsigned long flags; + + spin_lock_irqsave(&dd->ipath_sdma_lock, flags); + ret = ipath_sdma_make_progress(dd); + spin_unlock_irqrestore(&dd->ipath_sdma_lock, flags); + + return ret; +} + +/* we're in close, drain packets so that we can cleanup successfully... */ +void ipath_user_sdma_queue_drain(struct ipath_devdata *dd, + struct ipath_user_sdma_queue *pq) +{ + int i; + + if (!pq) + return; + + for (i = 0; i < 100; i++) { + mutex_lock(&pq->lock); + if (list_empty(&pq->sent)) { + mutex_unlock(&pq->lock); + break; + } + ipath_user_sdma_hwqueue_clean(dd); + ipath_user_sdma_queue_clean(dd, pq); + mutex_unlock(&pq->lock); + msleep(10); + } + + if (!list_empty(&pq->sent)) { + struct list_head free_list; + + printk(KERN_INFO "drain: lists not empty: forcing!\n"); + INIT_LIST_HEAD(&free_list); + mutex_lock(&pq->lock); + list_splice_init(&pq->sent, &free_list); + ipath_user_sdma_free_pkt_list(&dd->pcidev->dev, pq, &free_list); + mutex_unlock(&pq->lock); + } +} + +static inline __le64 ipath_sdma_make_desc0(struct ipath_devdata *dd, + u64 addr, u64 dwlen, u64 dwoffset) +{ + return cpu_to_le64(/* SDmaPhyAddr[31:0] */ + ((addr & 0xfffffffcULL) << 32) | + /* SDmaGeneration[1:0] */ + ((dd->ipath_sdma_generation & 3ULL) << 30) | + /* SDmaDwordCount[10:0] */ + ((dwlen & 0x7ffULL) << 16) | + /* SDmaBufOffset[12:2] */ + (dwoffset & 0x7ffULL)); +} + +static inline __le64 ipath_sdma_make_first_desc0(__le64 descq) +{ + return descq | __constant_cpu_to_le64(1ULL << 12); +} + +static inline __le64 ipath_sdma_make_last_desc0(__le64 descq) +{ + /* last */ /* dma head */ + return descq | __constant_cpu_to_le64(1ULL << 11 | 1ULL << 13); +} + +static inline __le64 ipath_sdma_make_desc1(u64 addr) +{ + /* SDmaPhyAddr[47:32] */ + return cpu_to_le64(addr >> 32); +} + +static void ipath_user_sdma_send_frag(struct ipath_devdata *dd, + struct ipath_user_sdma_pkt *pkt, int idx, + unsigned ofs, u16 tail) +{ + const u64 addr = (u64) pkt->addr[idx].addr + + (u64) pkt->addr[idx].offset; + const u64 dwlen = (u64) pkt->addr[idx].length / 4; + __le64 *descqp; + __le64 descq0; + + descqp = &dd->ipath_sdma_descq[tail].qw[0]; + + descq0 = ipath_sdma_make_desc0(dd, addr, dwlen, ofs); + if (idx == 0) + descq0 = ipath_sdma_make_first_desc0(descq0); + if (idx == pkt->naddr - 1) + descq0 = ipath_sdma_make_last_desc0(descq0); + + descqp[0] = descq0; + descqp[1] = ipath_sdma_make_desc1(addr); +} + +/* pq->lock must be held, get packets on the wire... */ +static int ipath_user_sdma_push_pkts(struct ipath_devdata *dd, + struct ipath_user_sdma_queue *pq, + struct list_head *pktlist) +{ + int ret = 0; + unsigned long flags; + u16 tail; + + if (list_empty(pktlist)) + return 0; + + if (unlikely(!(dd->ipath_flags & IPATH_LINKACTIVE))) + return -ECOMM; + + spin_lock_irqsave(&dd->ipath_sdma_lock, flags); + + if (unlikely(dd->ipath_sdma_status & IPATH_SDMA_ABORT_MASK)) { + ret = -ECOMM; + goto unlock; + } + + tail = dd->ipath_sdma_descq_tail; + while (!list_empty(pktlist)) { + struct ipath_user_sdma_pkt *pkt = + list_entry(pktlist->next, struct ipath_user_sdma_pkt, + list); + int i; + unsigned ofs = 0; + u16 dtail = tail; + + if (pkt->naddr > ipath_sdma_descq_freecnt(dd)) + goto unlock_check_tail; + + for (i = 0; i < pkt->naddr; i++) { + ipath_user_sdma_send_frag(dd, pkt, i, ofs, tail); + ofs += pkt->addr[i].length >> 2; + + if (++tail == dd->ipath_sdma_descq_cnt) { + tail = 0; + ++dd->ipath_sdma_generation; + } + } + + if ((ofs<<2) > dd->ipath_ibmaxlen) { + ipath_dbg("packet size %X > ibmax %X, fail\n", + ofs<<2, dd->ipath_ibmaxlen); + ret = -EMSGSIZE; + goto unlock; + } + + /* + * if the packet is >= 2KB mtu equivalent, we have to use + * the large buffers, and have to mark each descriptor as + * part of a large buffer packet. + */ + if (ofs >= IPATH_SMALLBUF_DWORDS) { + for (i = 0; i < pkt->naddr; i++) { + dd->ipath_sdma_descq[dtail].qw[0] |= + __constant_cpu_to_le64(1ULL << 14); + if (++dtail == dd->ipath_sdma_descq_cnt) + dtail = 0; + } + } + + dd->ipath_sdma_descq_added += pkt->naddr; + pkt->added = dd->ipath_sdma_descq_added; + list_move_tail(&pkt->list, &pq->sent); + ret++; + } + +unlock_check_tail: + /* advance the tail on the chip if necessary */ + if (dd->ipath_sdma_descq_tail != tail) { + wmb(); + ipath_write_kreg(dd, dd->ipath_kregs->kr_senddmatail, tail); + dd->ipath_sdma_descq_tail = tail; + } + +unlock: + spin_unlock_irqrestore(&dd->ipath_sdma_lock, flags); + + return ret; +} + +int ipath_user_sdma_writev(struct ipath_devdata *dd, + struct ipath_user_sdma_queue *pq, + const struct iovec *iov, + unsigned long dim) +{ + int ret = 0; + struct list_head list; + int npkts = 0; + + INIT_LIST_HEAD(&list); + + mutex_lock(&pq->lock); + + if (dd->ipath_sdma_descq_added != dd->ipath_sdma_descq_removed) { + ipath_user_sdma_hwqueue_clean(dd); + ipath_user_sdma_queue_clean(dd, pq); + } + + while (dim) { + const int mxp = 8; + + down_write(¤t->mm->mmap_sem); + ret = ipath_user_sdma_queue_pkts(dd, pq, &list, iov, dim, mxp); + up_write(¤t->mm->mmap_sem); + + if (ret <= 0) + goto done_unlock; + else { + dim -= ret; + iov += ret; + } + + /* force packets onto the sdma hw queue... */ + if (!list_empty(&list)) { + /* + * lazily clean hw queue. the 4 is a guess of about + * how many sdma descriptors a packet will take (it + * doesn't have to be perfect). + */ + if (ipath_sdma_descq_freecnt(dd) < ret * 4) { + ipath_user_sdma_hwqueue_clean(dd); + ipath_user_sdma_queue_clean(dd, pq); + } + + ret = ipath_user_sdma_push_pkts(dd, pq, &list); + if (ret < 0) + goto done_unlock; + else { + npkts += ret; + pq->counter += ret; + + if (!list_empty(&list)) + goto done_unlock; + } + } + } + +done_unlock: + if (!list_empty(&list)) + ipath_user_sdma_free_pkt_list(&dd->pcidev->dev, pq, &list); + mutex_unlock(&pq->lock); + + return (ret < 0) ? ret : npkts; +} + +int ipath_user_sdma_make_progress(struct ipath_devdata *dd, + struct ipath_user_sdma_queue *pq) +{ + int ret = 0; + + mutex_lock(&pq->lock); + ipath_user_sdma_hwqueue_clean(dd); + ret = ipath_user_sdma_queue_clean(dd, pq); + mutex_unlock(&pq->lock); + + return ret; +} + +u32 ipath_user_sdma_complete_counter(const struct ipath_user_sdma_queue *pq) +{ + return pq->sent_counter; +} + +u32 ipath_user_sdma_inflight_counter(struct ipath_user_sdma_queue *pq) +{ + return pq->counter; +} + diff --git a/drivers/infiniband/hw/ipath/ipath_user_sdma.h b/drivers/infiniband/hw/ipath/ipath_user_sdma.h new file mode 100644 index 000000000000..e70946c1428c --- /dev/null +++ b/drivers/infiniband/hw/ipath/ipath_user_sdma.h @@ -0,0 +1,54 @@ +/* + * Copyright (c) 2007, 2008 QLogic Corporation. All rights reserved. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * OpenIB.org BSD license below: + * + * Redistribution and use in source and binary forms, with or + * without modification, are permitted provided that the following + * conditions are met: + * + * - Redistributions of source code must retain the above + * copyright notice, this list of conditions and the following + * disclaimer. + * + * - Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials + * provided with the distribution. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + */ +#include <linux/device.h> + +struct ipath_user_sdma_queue; + +struct ipath_user_sdma_queue * +ipath_user_sdma_queue_create(struct device *dev, int unit, int port, int sport); +void ipath_user_sdma_queue_destroy(struct ipath_user_sdma_queue *pq); + +int ipath_user_sdma_writev(struct ipath_devdata *dd, + struct ipath_user_sdma_queue *pq, + const struct iovec *iov, + unsigned long dim); + +int ipath_user_sdma_make_progress(struct ipath_devdata *dd, + struct ipath_user_sdma_queue *pq); + +int ipath_user_sdma_pkt_sent(const struct ipath_user_sdma_queue *pq, + u32 counter); +void ipath_user_sdma_queue_drain(struct ipath_devdata *dd, + struct ipath_user_sdma_queue *pq); + +u32 ipath_user_sdma_complete_counter(const struct ipath_user_sdma_queue *pq); +u32 ipath_user_sdma_inflight_counter(struct ipath_user_sdma_queue *pq); diff --git a/drivers/infiniband/hw/ipath/ipath_verbs.c b/drivers/infiniband/hw/ipath/ipath_verbs.c index 32d8f882e56c..320a6d018de7 100644 --- a/drivers/infiniband/hw/ipath/ipath_verbs.c +++ b/drivers/infiniband/hw/ipath/ipath_verbs.c @@ -1,5 +1,5 @@ /* - * Copyright (c) 2006, 2007 QLogic Corporation. All rights reserved. + * Copyright (c) 2006, 2007, 2008 QLogic Corporation. All rights reserved. * Copyright (c) 2005, 2006 PathScale, Inc. All rights reserved. * * This software is available to you under a choice of one of two @@ -242,6 +242,93 @@ static void ipath_flush_wqe(struct ipath_qp *qp, struct ib_send_wr *wr) ipath_cq_enter(to_icq(qp->ibqp.send_cq), &wc, 1); } +/* + * Count the number of DMA descriptors needed to send length bytes of data. + * Don't modify the ipath_sge_state to get the count. + * Return zero if any of the segments is not aligned. + */ +static u32 ipath_count_sge(struct ipath_sge_state *ss, u32 length) +{ + struct ipath_sge *sg_list = ss->sg_list; + struct ipath_sge sge = ss->sge; + u8 num_sge = ss->num_sge; + u32 ndesc = 1; /* count the header */ + + while (length) { + u32 len = sge.length; + + if (len > length) + len = length; + if (len > sge.sge_length) + len = sge.sge_length; + BUG_ON(len == 0); + if (((long) sge.vaddr & (sizeof(u32) - 1)) || + (len != length && (len & (sizeof(u32) - 1)))) { + ndesc = 0; + break; + } + ndesc++; + sge.vaddr += len; + sge.length -= len; + sge.sge_length -= len; + if (sge.sge_length == 0) { + if (--num_sge) + sge = *sg_list++; + } else if (sge.length == 0 && sge.mr != NULL) { + if (++sge.n >= IPATH_SEGSZ) { + if (++sge.m >= sge.mr->mapsz) + break; + sge.n = 0; + } + sge.vaddr = + sge.mr->map[sge.m]->segs[sge.n].vaddr; + sge.length = + sge.mr->map[sge.m]->segs[sge.n].length; + } + length -= len; + } + return ndesc; +} + +/* + * Copy from the SGEs to the data buffer. + */ +static void ipath_copy_from_sge(void *data, struct ipath_sge_state *ss, + u32 length) +{ + struct ipath_sge *sge = &ss->sge; + + while (length) { + u32 len = sge->length; + + if (len > length) + len = length; + if (len > sge->sge_length) + len = sge->sge_length; + BUG_ON(len == 0); + memcpy(data, sge->vaddr, len); + sge->vaddr += len; + sge->length -= len; + sge->sge_length -= len; + if (sge->sge_length == 0) { + if (--ss->num_sge) + *sge = *ss->sg_list++; + } else if (sge->length == 0 && sge->mr != NULL) { + if (++sge->n >= IPATH_SEGSZ) { + if (++sge->m >= sge->mr->mapsz) + break; + sge->n = 0; + } + sge->vaddr = + sge->mr->map[sge->m]->segs[sge->n].vaddr; + sge->length = + sge->mr->map[sge->m]->segs[sge->n].length; + } + data += len; + length -= len; + } +} + /** * ipath_post_one_send - post one RC, UC, or UD send work request * @qp: the QP to post on @@ -866,27 +953,257 @@ static void copy_io(u32 __iomem *piobuf, struct ipath_sge_state *ss, __raw_writel(last, piobuf); } -static int ipath_verbs_send_pio(struct ipath_qp *qp, u32 *hdr, u32 hdrwords, +/* + * Convert IB rate to delay multiplier. + */ +unsigned ipath_ib_rate_to_mult(enum ib_rate rate) +{ + switch (rate) { + case IB_RATE_2_5_GBPS: return 8; + case IB_RATE_5_GBPS: return 4; + case IB_RATE_10_GBPS: return 2; + case IB_RATE_20_GBPS: return 1; + default: return 0; + } +} + +/* + * Convert delay multiplier to IB rate + */ +static enum ib_rate ipath_mult_to_ib_rate(unsigned mult) +{ + switch (mult) { + case 8: return IB_RATE_2_5_GBPS; + case 4: return IB_RATE_5_GBPS; + case 2: return IB_RATE_10_GBPS; + case 1: return IB_RATE_20_GBPS; + default: return IB_RATE_PORT_CURRENT; + } +} + +static inline struct ipath_verbs_txreq *get_txreq(struct ipath_ibdev *dev) +{ + struct ipath_verbs_txreq *tx = NULL; + unsigned long flags; + + spin_lock_irqsave(&dev->pending_lock, flags); + if (!list_empty(&dev->txreq_free)) { + struct list_head *l = dev->txreq_free.next; + + list_del(l); + tx = list_entry(l, struct ipath_verbs_txreq, txreq.list); + } + spin_unlock_irqrestore(&dev->pending_lock, flags); + return tx; +} + +static inline void put_txreq(struct ipath_ibdev *dev, + struct ipath_verbs_txreq *tx) +{ + unsigned long flags; + + spin_lock_irqsave(&dev->pending_lock, flags); + list_add(&tx->txreq.list, &dev->txreq_free); + spin_unlock_irqrestore(&dev->pending_lock, flags); +} + +static void sdma_complete(void *cookie, int status) +{ + struct ipath_verbs_txreq *tx = cookie; + struct ipath_qp *qp = tx->qp; + struct ipath_ibdev *dev = to_idev(qp->ibqp.device); + + /* Generate a completion queue entry if needed */ + if (qp->ibqp.qp_type != IB_QPT_RC && tx->wqe) { + enum ib_wc_status ibs = status == IPATH_SDMA_TXREQ_S_OK ? + IB_WC_SUCCESS : IB_WC_WR_FLUSH_ERR; + + ipath_send_complete(qp, tx->wqe, ibs); + } + + if (tx->txreq.flags & IPATH_SDMA_TXREQ_F_FREEBUF) + kfree(tx->txreq.map_addr); + put_txreq(dev, tx); + + if (atomic_dec_and_test(&qp->refcount)) + wake_up(&qp->wait); +} + +/* + * Compute the number of clock cycles of delay before sending the next packet. + * The multipliers reflect the number of clocks for the fastest rate so + * one tick at 4xDDR is 8 ticks at 1xSDR. + * If the destination port will take longer to receive a packet than + * the outgoing link can send it, we need to delay sending the next packet + * by the difference in time it takes the receiver to receive and the sender + * to send this packet. + * Note that this delay is always correct for UC and RC but not always + * optimal for UD. For UD, the destination HCA can be different for each + * packet, in which case, we could send packets to a different destination + * while "waiting" for the delay. The overhead for doing this without + * HW support is more than just paying the cost of delaying some packets + * unnecessarily. + */ +static inline unsigned ipath_pkt_delay(u32 plen, u8 snd_mult, u8 rcv_mult) +{ + return (rcv_mult > snd_mult) ? + (plen * (rcv_mult - snd_mult) + 1) >> 1 : 0; +} + +static int ipath_verbs_send_dma(struct ipath_qp *qp, + struct ipath_ib_header *hdr, u32 hdrwords, + struct ipath_sge_state *ss, u32 len, + u32 plen, u32 dwords) +{ + struct ipath_ibdev *dev = to_idev(qp->ibqp.device); + struct ipath_devdata *dd = dev->dd; + struct ipath_verbs_txreq *tx; + u32 *piobuf; + u32 control; + u32 ndesc; + int ret; + + tx = qp->s_tx; + if (tx) { + qp->s_tx = NULL; + /* resend previously constructed packet */ + ret = ipath_sdma_verbs_send(dd, tx->ss, tx->len, tx); + if (ret) + qp->s_tx = tx; + goto bail; + } + + tx = get_txreq(dev); + if (!tx) { + ret = -EBUSY; + goto bail; + } + + /* + * Get the saved delay count we computed for the previous packet + * and save the delay count for this packet to be used next time + * we get here. + */ + control = qp->s_pkt_delay; + qp->s_pkt_delay = ipath_pkt_delay(plen, dd->delay_mult, qp->s_dmult); + + tx->qp = qp; + atomic_inc(&qp->refcount); + tx->wqe = qp->s_wqe; + tx->txreq.callback = sdma_complete; + tx->txreq.callback_cookie = tx; + tx->txreq.flags = IPATH_SDMA_TXREQ_F_HEADTOHOST | + IPATH_SDMA_TXREQ_F_INTREQ | IPATH_SDMA_TXREQ_F_FREEDESC; + if (plen + 1 >= IPATH_SMALLBUF_DWORDS) + tx->txreq.flags |= IPATH_SDMA_TXREQ_F_USELARGEBUF; + + /* VL15 packets bypass credit check */ + if ((be16_to_cpu(hdr->lrh[0]) >> 12) == 15) { + control |= 1ULL << 31; + tx->txreq.flags |= IPATH_SDMA_TXREQ_F_VL15; + } + + if (len) { + /* + * Don't try to DMA if it takes more descriptors than + * the queue holds. + */ + ndesc = ipath_count_sge(ss, len); + if (ndesc >= dd->ipath_sdma_descq_cnt) + ndesc = 0; + } else + ndesc = 1; + if (ndesc) { + tx->hdr.pbc[0] = cpu_to_le32(plen); + tx->hdr.pbc[1] = cpu_to_le32(control); + memcpy(&tx->hdr.hdr, hdr, hdrwords << 2); + tx->txreq.sg_count = ndesc; + tx->map_len = (hdrwords + 2) << 2; + tx->txreq.map_addr = &tx->hdr; + ret = ipath_sdma_verbs_send(dd, ss, dwords, tx); + if (ret) { + /* save ss and length in dwords */ + tx->ss = ss; + tx->len = dwords; + qp->s_tx = tx; + } + goto bail; + } + + /* Allocate a buffer and copy the header and payload to it. */ + tx->map_len = (plen + 1) << 2; + piobuf = kmalloc(tx->map_len, GFP_ATOMIC); + if (unlikely(piobuf == NULL)) { + ret = -EBUSY; + goto err_tx; + } + tx->txreq.map_addr = piobuf; + tx->txreq.flags |= IPATH_SDMA_TXREQ_F_FREEBUF; + tx->txreq.sg_count = 1; + + *piobuf++ = (__force u32) cpu_to_le32(plen); + *piobuf++ = (__force u32) cpu_to_le32(control); + memcpy(piobuf, hdr, hdrwords << 2); + ipath_copy_from_sge(piobuf + hdrwords, ss, len); + + ret = ipath_sdma_verbs_send(dd, NULL, 0, tx); + /* + * If we couldn't queue the DMA request, save the info + * and try again later rather than destroying the + * buffer and undoing the side effects of the copy. + */ + if (ret) { + tx->ss = NULL; + tx->len = 0; + qp->s_tx = tx; + } + dev->n_unaligned++; + goto bail; + +err_tx: + if (atomic_dec_and_test(&qp->refcount)) + wake_up(&qp->wait); + put_txreq(dev, tx); +bail: + return ret; +} + +static int ipath_verbs_send_pio(struct ipath_qp *qp, + struct ipath_ib_header *ibhdr, u32 hdrwords, struct ipath_sge_state *ss, u32 len, u32 plen, u32 dwords) { struct ipath_devdata *dd = to_idev(qp->ibqp.device)->dd; + u32 *hdr = (u32 *) ibhdr; u32 __iomem *piobuf; unsigned flush_wc; + u32 control; int ret; - piobuf = ipath_getpiobuf(dd, NULL); + piobuf = ipath_getpiobuf(dd, plen, NULL); if (unlikely(piobuf == NULL)) { ret = -EBUSY; goto bail; } /* - * Write len to control qword, no flags. + * Get the saved delay count we computed for the previous packet + * and save the delay count for this packet to be used next time + * we get here. + */ + control = qp->s_pkt_delay; + qp->s_pkt_delay = ipath_pkt_delay(plen, dd->delay_mult, qp->s_dmult); + + /* VL15 packets bypass credit check */ + if ((be16_to_cpu(ibhdr->lrh[0]) >> 12) == 15) + control |= 1ULL << 31; + + /* + * Write the length to the control qword plus any needed flags. * We have to flush after the PBC for correctness on some cpus * or WC buffer can be written out of order. */ - writeq(plen, piobuf); + writeq(((u64) control << 32) | plen, piobuf); piobuf += 2; flush_wc = dd->ipath_flags & IPATH_PIO_FLUSH_WC; @@ -961,15 +1278,25 @@ int ipath_verbs_send(struct ipath_qp *qp, struct ipath_ib_header *hdr, */ plen = hdrwords + dwords + 1; - /* Drop non-VL15 packets if we are not in the active state */ - if (!(dd->ipath_flags & IPATH_LINKACTIVE) && - qp->ibqp.qp_type != IB_QPT_SMI) { + /* + * VL15 packets (IB_QPT_SMI) will always use PIO, so we + * can defer SDMA restart until link goes ACTIVE without + * worrying about just how we got there. + */ + if (qp->ibqp.qp_type == IB_QPT_SMI) + ret = ipath_verbs_send_pio(qp, hdr, hdrwords, ss, len, + plen, dwords); + /* All non-VL15 packets are dropped if link is not ACTIVE */ + else if (!(dd->ipath_flags & IPATH_LINKACTIVE)) { if (qp->s_wqe) ipath_send_complete(qp, qp->s_wqe, IB_WC_SUCCESS); ret = 0; - } else - ret = ipath_verbs_send_pio(qp, (u32 *) hdr, hdrwords, - ss, len, plen, dwords); + } else if (dd->ipath_flags & IPATH_HAS_SEND_DMA) + ret = ipath_verbs_send_dma(qp, hdr, hdrwords, ss, len, + plen, dwords); + else + ret = ipath_verbs_send_pio(qp, hdr, hdrwords, ss, len, + plen, dwords); return ret; } @@ -1038,6 +1365,12 @@ int ipath_get_counters(struct ipath_devdata *dd, ipath_snap_cntr(dd, crp->cr_errlpcrccnt) + ipath_snap_cntr(dd, crp->cr_badformatcnt) + dd->ipath_rxfc_unsupvl_errs; + if (crp->cr_rxotherlocalphyerrcnt) + cntrs->port_rcv_errors += + ipath_snap_cntr(dd, crp->cr_rxotherlocalphyerrcnt); + if (crp->cr_rxvlerrcnt) + cntrs->port_rcv_errors += + ipath_snap_cntr(dd, crp->cr_rxvlerrcnt); cntrs->port_rcv_remphys_errors = ipath_snap_cntr(dd, crp->cr_rcvebpcnt); cntrs->port_xmit_discards = ipath_snap_cntr(dd, crp->cr_unsupvlcnt); @@ -1046,9 +1379,16 @@ int ipath_get_counters(struct ipath_devdata *dd, cntrs->port_xmit_packets = ipath_snap_cntr(dd, crp->cr_pktsendcnt); cntrs->port_rcv_packets = ipath_snap_cntr(dd, crp->cr_pktrcvcnt); cntrs->local_link_integrity_errors = - (dd->ipath_flags & IPATH_GPIO_ERRINTRS) ? - dd->ipath_lli_errs : dd->ipath_lli_errors; - cntrs->excessive_buffer_overrun_errors = dd->ipath_overrun_thresh_errs; + crp->cr_locallinkintegrityerrcnt ? + ipath_snap_cntr(dd, crp->cr_locallinkintegrityerrcnt) : + ((dd->ipath_flags & IPATH_GPIO_ERRINTRS) ? + dd->ipath_lli_errs : dd->ipath_lli_errors); + cntrs->excessive_buffer_overrun_errors = + crp->cr_excessbufferovflcnt ? + ipath_snap_cntr(dd, crp->cr_excessbufferovflcnt) : + dd->ipath_overrun_thresh_errs; + cntrs->vl15_dropped = crp->cr_vl15droppedpktcnt ? + ipath_snap_cntr(dd, crp->cr_vl15droppedpktcnt) : 0; ret = 0; @@ -1183,7 +1523,9 @@ static int ipath_query_port(struct ib_device *ibdev, props->sm_lid = dev->sm_lid; props->sm_sl = dev->sm_sl; ibcstat = dd->ipath_lastibcstat; - props->state = ((ibcstat >> 4) & 0x3) + 1; + /* map LinkState to IB portinfo values. */ + props->state = ipath_ib_linkstate(dd, ibcstat) + 1; + /* See phys_state_show() */ props->phys_state = /* MEA: assumes shift == 0 */ ipath_cvt_physportstate[dd->ipath_lastibcstat & @@ -1195,18 +1537,13 @@ static int ipath_query_port(struct ib_device *ibdev, props->bad_pkey_cntr = ipath_get_cr_errpkey(dd) - dev->z_pkey_violations; props->qkey_viol_cntr = dev->qkey_violations; - props->active_width = IB_WIDTH_4X; + props->active_width = dd->ipath_link_width_active; /* See rate_show() */ - props->active_speed = 1; /* Regular 10Mbs speed. */ + props->active_speed = dd->ipath_link_speed_active; props->max_vl_num = 1; /* VLCap = VL0 */ props->init_type_reply = 0; - /* - * Note: the chip supports a maximum MTU of 4096, but the driver - * hasn't implemented this feature yet, so set the maximum value - * to 2048. - */ - props->max_mtu = IB_MTU_2048; + props->max_mtu = ipath_mtu4096 ? IB_MTU_4096 : IB_MTU_2048; switch (dd->ipath_ibmtu) { case 4096: mtu = IB_MTU_4096; @@ -1399,6 +1736,7 @@ static struct ib_ah *ipath_create_ah(struct ib_pd *pd, /* ib_create_ah() will initialize ah->ibah. */ ah->attr = *ah_attr; + ah->attr.static_rate = ipath_ib_rate_to_mult(ah_attr->static_rate); ret = &ah->ibah; @@ -1432,6 +1770,7 @@ static int ipath_query_ah(struct ib_ah *ibah, struct ib_ah_attr *ah_attr) struct ipath_ah *ah = to_iah(ibah); *ah_attr = ah->attr; + ah_attr->static_rate = ipath_mult_to_ib_rate(ah->attr.static_rate); return 0; } @@ -1581,6 +1920,8 @@ int ipath_register_ib_device(struct ipath_devdata *dd) struct ipath_verbs_counters cntrs; struct ipath_ibdev *idev; struct ib_device *dev; + struct ipath_verbs_txreq *tx; + unsigned i; int ret; idev = (struct ipath_ibdev *)ib_alloc_device(sizeof *idev); @@ -1591,6 +1932,17 @@ int ipath_register_ib_device(struct ipath_devdata *dd) dev = &idev->ibdev; + if (dd->ipath_sdma_descq_cnt) { + tx = kmalloc(dd->ipath_sdma_descq_cnt * sizeof *tx, + GFP_KERNEL); + if (tx == NULL) { + ret = -ENOMEM; + goto err_tx; + } + } else + tx = NULL; + idev->txreq_bufs = tx; + /* Only need to initialize non-zero fields. */ spin_lock_init(&idev->n_pds_lock); spin_lock_init(&idev->n_ahs_lock); @@ -1631,15 +1983,17 @@ int ipath_register_ib_device(struct ipath_devdata *dd) INIT_LIST_HEAD(&idev->pending[2]); INIT_LIST_HEAD(&idev->piowait); INIT_LIST_HEAD(&idev->rnrwait); + INIT_LIST_HEAD(&idev->txreq_free); idev->pending_index = 0; idev->port_cap_flags = IB_PORT_SYS_IMAGE_GUID_SUP | IB_PORT_CLIENT_REG_SUP; + if (dd->ipath_flags & IPATH_HAS_LINK_LATENCY) + idev->port_cap_flags |= IB_PORT_LINK_LATENCY_SUP; idev->pma_counter_select[0] = IB_PMA_PORT_XMIT_DATA; idev->pma_counter_select[1] = IB_PMA_PORT_RCV_DATA; idev->pma_counter_select[2] = IB_PMA_PORT_XMIT_PKTS; idev->pma_counter_select[3] = IB_PMA_PORT_RCV_PKTS; idev->pma_counter_select[4] = IB_PMA_PORT_XMIT_WAIT; - idev->link_width_enabled = 3; /* 1x or 4x */ /* Snapshot current HW counters to "clear" them. */ ipath_get_counters(dd, &cntrs); @@ -1661,6 +2015,9 @@ int ipath_register_ib_device(struct ipath_devdata *dd) cntrs.excessive_buffer_overrun_errors; idev->z_vl15_dropped = cntrs.vl15_dropped; + for (i = 0; i < dd->ipath_sdma_descq_cnt; i++, tx++) + list_add(&tx->txreq.list, &idev->txreq_free); + /* * The system image GUID is supposed to be the same for all * IB HCAs in a single system but since there can be other @@ -1710,6 +2067,7 @@ int ipath_register_ib_device(struct ipath_devdata *dd) dev->phys_port_cnt = 1; dev->num_comp_vectors = 1; dev->dma_device = &dd->pcidev->dev; + dev->class_dev.dev = dev->dma_device; dev->query_device = ipath_query_device; dev->modify_device = ipath_modify_device; dev->query_port = ipath_query_port; @@ -1774,6 +2132,8 @@ err_reg: err_lk: kfree(idev->qp_table.table); err_qp: + kfree(idev->txreq_bufs); +err_tx: ib_dealloc_device(dev); ipath_dev_err(dd, "cannot register verbs: %d!\n", -ret); idev = NULL; @@ -1808,6 +2168,7 @@ void ipath_unregister_ib_device(struct ipath_ibdev *dev) ipath_free_all_qps(&dev->qp_table); kfree(dev->qp_table.table); kfree(dev->lk_table.table); + kfree(dev->txreq_bufs); ib_dealloc_device(ibdev); } @@ -1855,13 +2216,15 @@ static ssize_t show_stats(struct class_device *cdev, char *buf) "RC stalls %d\n" "piobuf wait %d\n" "no piobuf %d\n" + "unaligned %d\n" "PKT drops %d\n" "WQE errs %d\n", dev->n_rc_resends, dev->n_rc_qacks, dev->n_rc_acks, dev->n_seq_naks, dev->n_rdma_seq, dev->n_rnr_naks, dev->n_other_naks, dev->n_timeouts, dev->n_rdma_dup_busy, dev->n_rc_stalls, dev->n_piowait, - dev->n_no_piobuf, dev->n_pkt_drops, dev->n_wqe_errs); + dev->n_no_piobuf, dev->n_unaligned, + dev->n_pkt_drops, dev->n_wqe_errs); for (i = 0; i < ARRAY_SIZE(dev->opstats); i++) { const struct ipath_opcode_stats *si = &dev->opstats[i]; diff --git a/drivers/infiniband/hw/ipath/ipath_verbs.h b/drivers/infiniband/hw/ipath/ipath_verbs.h index 3d59736b49b2..6514aa8306cd 100644 --- a/drivers/infiniband/hw/ipath/ipath_verbs.h +++ b/drivers/infiniband/hw/ipath/ipath_verbs.h @@ -1,5 +1,5 @@ /* - * Copyright (c) 2006, 2007 QLogic Corporation. All rights reserved. + * Copyright (c) 2006, 2007, 2008 QLogic Corporation. All rights reserved. * Copyright (c) 2005, 2006 PathScale, Inc. All rights reserved. * * This software is available to you under a choice of one of two @@ -138,6 +138,11 @@ struct ipath_ib_header { } u; } __attribute__ ((packed)); +struct ipath_pio_header { + __le32 pbc[2]; + struct ipath_ib_header hdr; +} __attribute__ ((packed)); + /* * There is one struct ipath_mcast for each multicast GID. * All attached QPs are then stored as a list of @@ -319,6 +324,7 @@ struct ipath_sge_state { struct ipath_sge *sg_list; /* next SGE to be used if any */ struct ipath_sge sge; /* progress state for the current SGE */ u8 num_sge; + u8 static_rate; }; /* @@ -356,6 +362,7 @@ struct ipath_qp { struct tasklet_struct s_task; struct ipath_mmap_info *ip; struct ipath_sge_state *s_cur_sge; + struct ipath_verbs_txreq *s_tx; struct ipath_sge_state s_sge; /* current send request data */ struct ipath_ack_entry s_ack_queue[IPATH_MAX_RDMA_ATOMIC + 1]; struct ipath_sge_state s_ack_rdma_sge; @@ -363,7 +370,8 @@ struct ipath_qp { struct ipath_sge_state r_sge; /* current receive data */ spinlock_t s_lock; unsigned long s_busy; - u32 s_hdrwords; /* size of s_hdr in 32 bit words */ + u16 s_pkt_delay; + u16 s_hdrwords; /* size of s_hdr in 32 bit words */ u32 s_cur_size; /* size of send packet in bytes */ u32 s_len; /* total length of s_sge */ u32 s_rdma_read_len; /* total length of s_rdma_read_sge */ @@ -387,7 +395,6 @@ struct ipath_qp { u8 r_nak_state; /* non-zero if NAK is pending */ u8 r_min_rnr_timer; /* retry timeout value for RNR NAKs */ u8 r_reuse_sge; /* for UC receive errors */ - u8 r_sge_inx; /* current index into sg_list */ u8 r_wrid_valid; /* r_wrid set but CQ entry not yet made */ u8 r_max_rd_atomic; /* max number of RDMA read/atomic to receive */ u8 r_head_ack_queue; /* index into s_ack_queue[] */ @@ -403,6 +410,7 @@ struct ipath_qp { u8 s_num_rd_atomic; /* number of RDMA read/atomic pending */ u8 s_tail_ack_queue; /* index into s_ack_queue[] */ u8 s_flags; + u8 s_dmult; u8 timeout; /* Timeout for this QP */ enum ib_mtu path_mtu; u32 remote_qpn; @@ -510,6 +518,8 @@ struct ipath_ibdev { struct ipath_lkey_table lk_table; struct list_head pending[3]; /* FIFO of QPs waiting for ACKs */ struct list_head piowait; /* list for wait PIO buf */ + struct list_head txreq_free; + void *txreq_bufs; /* list of QPs waiting for RNR timer */ struct list_head rnrwait; spinlock_t pending_lock; @@ -570,6 +580,7 @@ struct ipath_ibdev { u32 n_rdma_dup_busy; u32 n_piowait; u32 n_no_piobuf; + u32 n_unaligned; u32 port_cap_flags; u32 pma_sample_start; u32 pma_sample_interval; @@ -581,7 +592,6 @@ struct ipath_ibdev { u16 pending_index; /* which pending queue is active */ u8 pma_sample_status; u8 subnet_timeout; - u8 link_width_enabled; u8 vl_high_limit; struct ipath_opcode_stats opstats[128]; }; @@ -602,6 +612,16 @@ struct ipath_verbs_counters { u32 vl15_dropped; }; +struct ipath_verbs_txreq { + struct ipath_qp *qp; + struct ipath_swqe *wqe; + u32 map_len; + u32 len; + struct ipath_sge_state *ss; + struct ipath_pio_header hdr; + struct ipath_sdma_txreq txreq; +}; + static inline struct ipath_mr *to_imr(struct ib_mr *ibmr) { return container_of(ibmr, struct ipath_mr, ibmr); @@ -694,11 +714,11 @@ void ipath_sqerror_qp(struct ipath_qp *qp, struct ib_wc *wc); void ipath_get_credit(struct ipath_qp *qp, u32 aeth); +unsigned ipath_ib_rate_to_mult(enum ib_rate rate); + int ipath_verbs_send(struct ipath_qp *qp, struct ipath_ib_header *hdr, u32 hdrwords, struct ipath_sge_state *ss, u32 len); -void ipath_cq_enter(struct ipath_cq *cq, struct ib_wc *entry, int sig); - void ipath_copy_sge(struct ipath_sge_state *ss, void *data, u32 length); void ipath_skip_sge(struct ipath_sge_state *ss, u32 length); diff --git a/drivers/infiniband/hw/mlx4/cq.c b/drivers/infiniband/hw/mlx4/cq.c index 7360bbafbe84..3557e7edc9b6 100644 --- a/drivers/infiniband/hw/mlx4/cq.c +++ b/drivers/infiniband/hw/mlx4/cq.c @@ -85,6 +85,82 @@ static struct mlx4_cqe *next_cqe_sw(struct mlx4_ib_cq *cq) return get_sw_cqe(cq, cq->mcq.cons_index); } +int mlx4_ib_modify_cq(struct ib_cq *cq, u16 cq_count, u16 cq_period) +{ + struct mlx4_ib_cq *mcq = to_mcq(cq); + struct mlx4_ib_dev *dev = to_mdev(cq->device); + + return mlx4_cq_modify(dev->dev, &mcq->mcq, cq_count, cq_period); +} + +static int mlx4_ib_alloc_cq_buf(struct mlx4_ib_dev *dev, struct mlx4_ib_cq_buf *buf, int nent) +{ + int err; + + err = mlx4_buf_alloc(dev->dev, nent * sizeof(struct mlx4_cqe), + PAGE_SIZE * 2, &buf->buf); + + if (err) + goto out; + + err = mlx4_mtt_init(dev->dev, buf->buf.npages, buf->buf.page_shift, + &buf->mtt); + if (err) + goto err_buf; + + err = mlx4_buf_write_mtt(dev->dev, &buf->mtt, &buf->buf); + if (err) + goto err_mtt; + + return 0; + +err_mtt: + mlx4_mtt_cleanup(dev->dev, &buf->mtt); + +err_buf: + mlx4_buf_free(dev->dev, nent * sizeof(struct mlx4_cqe), + &buf->buf); + +out: + return err; +} + +static void mlx4_ib_free_cq_buf(struct mlx4_ib_dev *dev, struct mlx4_ib_cq_buf *buf, int cqe) +{ + mlx4_buf_free(dev->dev, (cqe + 1) * sizeof(struct mlx4_cqe), &buf->buf); +} + +static int mlx4_ib_get_cq_umem(struct mlx4_ib_dev *dev, struct ib_ucontext *context, + struct mlx4_ib_cq_buf *buf, struct ib_umem **umem, + u64 buf_addr, int cqe) +{ + int err; + + *umem = ib_umem_get(context, buf_addr, cqe * sizeof (struct mlx4_cqe), + IB_ACCESS_LOCAL_WRITE); + if (IS_ERR(*umem)) + return PTR_ERR(*umem); + + err = mlx4_mtt_init(dev->dev, ib_umem_page_count(*umem), + ilog2((*umem)->page_size), &buf->mtt); + if (err) + goto err_buf; + + err = mlx4_ib_umem_write_mtt(dev, &buf->mtt, *umem); + if (err) + goto err_mtt; + + return 0; + +err_mtt: + mlx4_mtt_cleanup(dev->dev, &buf->mtt); + +err_buf: + ib_umem_release(*umem); + + return err; +} + struct ib_cq *mlx4_ib_create_cq(struct ib_device *ibdev, int entries, int vector, struct ib_ucontext *context, struct ib_udata *udata) @@ -92,7 +168,6 @@ struct ib_cq *mlx4_ib_create_cq(struct ib_device *ibdev, int entries, int vector struct mlx4_ib_dev *dev = to_mdev(ibdev); struct mlx4_ib_cq *cq; struct mlx4_uar *uar; - int buf_size; int err; if (entries < 1 || entries > dev->dev->caps.max_cqes) @@ -104,8 +179,10 @@ struct ib_cq *mlx4_ib_create_cq(struct ib_device *ibdev, int entries, int vector entries = roundup_pow_of_two(entries + 1); cq->ibcq.cqe = entries - 1; - buf_size = entries * sizeof (struct mlx4_cqe); + mutex_init(&cq->resize_mutex); spin_lock_init(&cq->lock); + cq->resize_buf = NULL; + cq->resize_umem = NULL; if (context) { struct mlx4_ib_create_cq ucmd; @@ -115,21 +192,10 @@ struct ib_cq *mlx4_ib_create_cq(struct ib_device *ibdev, int entries, int vector goto err_cq; } - cq->umem = ib_umem_get(context, ucmd.buf_addr, buf_size, - IB_ACCESS_LOCAL_WRITE); - if (IS_ERR(cq->umem)) { - err = PTR_ERR(cq->umem); - goto err_cq; - } - - err = mlx4_mtt_init(dev->dev, ib_umem_page_count(cq->umem), - ilog2(cq->umem->page_size), &cq->buf.mtt); - if (err) - goto err_buf; - - err = mlx4_ib_umem_write_mtt(dev, &cq->buf.mtt, cq->umem); + err = mlx4_ib_get_cq_umem(dev, context, &cq->buf, &cq->umem, + ucmd.buf_addr, entries); if (err) - goto err_mtt; + goto err_cq; err = mlx4_ib_db_map_user(to_mucontext(context), ucmd.db_addr, &cq->db); @@ -147,19 +213,9 @@ struct ib_cq *mlx4_ib_create_cq(struct ib_device *ibdev, int entries, int vector *cq->mcq.set_ci_db = 0; *cq->mcq.arm_db = 0; - if (mlx4_buf_alloc(dev->dev, buf_size, PAGE_SIZE * 2, &cq->buf.buf)) { - err = -ENOMEM; - goto err_db; - } - - err = mlx4_mtt_init(dev->dev, cq->buf.buf.npages, cq->buf.buf.page_shift, - &cq->buf.mtt); + err = mlx4_ib_alloc_cq_buf(dev, &cq->buf, entries); if (err) - goto err_buf; - - err = mlx4_buf_write_mtt(dev->dev, &cq->buf.mtt, &cq->buf.buf); - if (err) - goto err_mtt; + goto err_db; uar = &dev->priv_uar; } @@ -187,12 +243,10 @@ err_dbmap: err_mtt: mlx4_mtt_cleanup(dev->dev, &cq->buf.mtt); -err_buf: if (context) ib_umem_release(cq->umem); else - mlx4_buf_free(dev->dev, entries * sizeof (struct mlx4_cqe), - &cq->buf.buf); + mlx4_ib_free_cq_buf(dev, &cq->buf, entries); err_db: if (!context) @@ -204,6 +258,170 @@ err_cq: return ERR_PTR(err); } +static int mlx4_alloc_resize_buf(struct mlx4_ib_dev *dev, struct mlx4_ib_cq *cq, + int entries) +{ + int err; + + if (cq->resize_buf) + return -EBUSY; + + cq->resize_buf = kmalloc(sizeof *cq->resize_buf, GFP_ATOMIC); + if (!cq->resize_buf) + return -ENOMEM; + + err = mlx4_ib_alloc_cq_buf(dev, &cq->resize_buf->buf, entries); + if (err) { + kfree(cq->resize_buf); + cq->resize_buf = NULL; + return err; + } + + cq->resize_buf->cqe = entries - 1; + + return 0; +} + +static int mlx4_alloc_resize_umem(struct mlx4_ib_dev *dev, struct mlx4_ib_cq *cq, + int entries, struct ib_udata *udata) +{ + struct mlx4_ib_resize_cq ucmd; + int err; + + if (cq->resize_umem) + return -EBUSY; + + if (ib_copy_from_udata(&ucmd, udata, sizeof ucmd)) + return -EFAULT; + + cq->resize_buf = kmalloc(sizeof *cq->resize_buf, GFP_ATOMIC); + if (!cq->resize_buf) + return -ENOMEM; + + err = mlx4_ib_get_cq_umem(dev, cq->umem->context, &cq->resize_buf->buf, + &cq->resize_umem, ucmd.buf_addr, entries); + if (err) { + kfree(cq->resize_buf); + cq->resize_buf = NULL; + return err; + } + + cq->resize_buf->cqe = entries - 1; + + return 0; +} + +static int mlx4_ib_get_outstanding_cqes(struct mlx4_ib_cq *cq) +{ + u32 i; + + i = cq->mcq.cons_index; + while (get_sw_cqe(cq, i & cq->ibcq.cqe)) + ++i; + + return i - cq->mcq.cons_index; +} + +static void mlx4_ib_cq_resize_copy_cqes(struct mlx4_ib_cq *cq) +{ + struct mlx4_cqe *cqe; + int i; + + i = cq->mcq.cons_index; + cqe = get_cqe(cq, i & cq->ibcq.cqe); + while ((cqe->owner_sr_opcode & MLX4_CQE_OPCODE_MASK) != MLX4_CQE_OPCODE_RESIZE) { + memcpy(get_cqe_from_buf(&cq->resize_buf->buf, + (i + 1) & cq->resize_buf->cqe), + get_cqe(cq, i & cq->ibcq.cqe), sizeof(struct mlx4_cqe)); + cqe = get_cqe(cq, ++i & cq->ibcq.cqe); + } + ++cq->mcq.cons_index; +} + +int mlx4_ib_resize_cq(struct ib_cq *ibcq, int entries, struct ib_udata *udata) +{ + struct mlx4_ib_dev *dev = to_mdev(ibcq->device); + struct mlx4_ib_cq *cq = to_mcq(ibcq); + int outst_cqe; + int err; + + mutex_lock(&cq->resize_mutex); + + if (entries < 1 || entries > dev->dev->caps.max_cqes) { + err = -EINVAL; + goto out; + } + + entries = roundup_pow_of_two(entries + 1); + if (entries == ibcq->cqe + 1) { + err = 0; + goto out; + } + + if (ibcq->uobject) { + err = mlx4_alloc_resize_umem(dev, cq, entries, udata); + if (err) + goto out; + } else { + /* Can't be smaller then the number of outstanding CQEs */ + outst_cqe = mlx4_ib_get_outstanding_cqes(cq); + if (entries < outst_cqe + 1) { + err = 0; + goto out; + } + + err = mlx4_alloc_resize_buf(dev, cq, entries); + if (err) + goto out; + } + + err = mlx4_cq_resize(dev->dev, &cq->mcq, entries, &cq->resize_buf->buf.mtt); + if (err) + goto err_buf; + + if (ibcq->uobject) { + cq->buf = cq->resize_buf->buf; + cq->ibcq.cqe = cq->resize_buf->cqe; + ib_umem_release(cq->umem); + cq->umem = cq->resize_umem; + + kfree(cq->resize_buf); + cq->resize_buf = NULL; + cq->resize_umem = NULL; + } else { + spin_lock_irq(&cq->lock); + if (cq->resize_buf) { + mlx4_ib_cq_resize_copy_cqes(cq); + mlx4_ib_free_cq_buf(dev, &cq->buf, cq->ibcq.cqe); + cq->buf = cq->resize_buf->buf; + cq->ibcq.cqe = cq->resize_buf->cqe; + + kfree(cq->resize_buf); + cq->resize_buf = NULL; + } + spin_unlock_irq(&cq->lock); + } + + goto out; + +err_buf: + if (!ibcq->uobject) + mlx4_ib_free_cq_buf(dev, &cq->resize_buf->buf, + cq->resize_buf->cqe); + + kfree(cq->resize_buf); + cq->resize_buf = NULL; + + if (cq->resize_umem) { + ib_umem_release(cq->resize_umem); + cq->resize_umem = NULL; + } + +out: + mutex_unlock(&cq->resize_mutex); + return err; +} + int mlx4_ib_destroy_cq(struct ib_cq *cq) { struct mlx4_ib_dev *dev = to_mdev(cq->device); @@ -216,8 +434,7 @@ int mlx4_ib_destroy_cq(struct ib_cq *cq) mlx4_ib_db_unmap_user(to_mucontext(cq->uobject->context), &mcq->db); ib_umem_release(mcq->umem); } else { - mlx4_buf_free(dev->dev, (cq->cqe + 1) * sizeof (struct mlx4_cqe), - &mcq->buf.buf); + mlx4_ib_free_cq_buf(dev, &mcq->buf, cq->cqe + 1); mlx4_ib_db_free(dev, &mcq->db); } @@ -297,6 +514,20 @@ static void mlx4_ib_handle_error_cqe(struct mlx4_err_cqe *cqe, wc->vendor_err = cqe->vendor_err_syndrome; } +static int mlx4_ib_ipoib_csum_ok(__be32 status, __be16 checksum) +{ + return ((status & cpu_to_be32(MLX4_CQE_IPOIB_STATUS_IPV4 | + MLX4_CQE_IPOIB_STATUS_IPV4F | + MLX4_CQE_IPOIB_STATUS_IPV4OPT | + MLX4_CQE_IPOIB_STATUS_IPV6 | + MLX4_CQE_IPOIB_STATUS_IPOK)) == + cpu_to_be32(MLX4_CQE_IPOIB_STATUS_IPV4 | + MLX4_CQE_IPOIB_STATUS_IPOK)) && + (status & cpu_to_be32(MLX4_CQE_IPOIB_STATUS_UDP | + MLX4_CQE_IPOIB_STATUS_TCP)) && + checksum == cpu_to_be16(0xffff); +} + static int mlx4_ib_poll_one(struct mlx4_ib_cq *cq, struct mlx4_ib_qp **cur_qp, struct ib_wc *wc) @@ -310,6 +541,7 @@ static int mlx4_ib_poll_one(struct mlx4_ib_cq *cq, u32 g_mlpath_rqpn; u16 wqe_ctr; +repoll: cqe = next_cqe_sw(cq); if (!cqe) return -EAGAIN; @@ -332,6 +564,22 @@ static int mlx4_ib_poll_one(struct mlx4_ib_cq *cq, return -EINVAL; } + /* Resize CQ in progress */ + if (unlikely((cqe->owner_sr_opcode & MLX4_CQE_OPCODE_MASK) == MLX4_CQE_OPCODE_RESIZE)) { + if (cq->resize_buf) { + struct mlx4_ib_dev *dev = to_mdev(cq->ibcq.device); + + mlx4_ib_free_cq_buf(dev, &cq->buf, cq->ibcq.cqe); + cq->buf = cq->resize_buf->buf; + cq->ibcq.cqe = cq->resize_buf->cqe; + + kfree(cq->resize_buf); + cq->resize_buf = NULL; + } + + goto repoll; + } + if (!*cur_qp || (be32_to_cpu(cqe->my_qpn) & 0xffffff) != (*cur_qp)->mqp.qpn) { /* @@ -406,6 +654,9 @@ static int mlx4_ib_poll_one(struct mlx4_ib_cq *cq, case MLX4_OPCODE_BIND_MW: wc->opcode = IB_WC_BIND_MW; break; + case MLX4_OPCODE_LSO: + wc->opcode = IB_WC_LSO; + break; } } else { wc->byte_len = be32_to_cpu(cqe->byte_cnt); @@ -434,6 +685,8 @@ static int mlx4_ib_poll_one(struct mlx4_ib_cq *cq, wc->dlid_path_bits = (g_mlpath_rqpn >> 24) & 0x7f; wc->wc_flags |= g_mlpath_rqpn & 0x80000000 ? IB_WC_GRH : 0; wc->pkey_index = be32_to_cpu(cqe->immed_rss_invalid) & 0x7f; + wc->csum_ok = mlx4_ib_ipoib_csum_ok(cqe->ipoib_status, + cqe->checksum); } return 0; diff --git a/drivers/infiniband/hw/mlx4/mad.c b/drivers/infiniband/hw/mlx4/mad.c index 0ed02b7834da..4c1e72fc8f57 100644 --- a/drivers/infiniband/hw/mlx4/mad.c +++ b/drivers/infiniband/hw/mlx4/mad.c @@ -165,7 +165,7 @@ static void smp_snoop(struct ib_device *ibdev, u8 port_num, struct ib_mad *mad) event.device = ibdev; event.element.port_num = port_num; - if(pinfo->clientrereg_resv_subnetto & 0x80) + if (pinfo->clientrereg_resv_subnetto & 0x80) event.event = IB_EVENT_CLIENT_REREGISTER; else event.event = IB_EVENT_LID_CHANGE; diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c index 96a39b5c9254..136c76c7b4e7 100644 --- a/drivers/infiniband/hw/mlx4/main.c +++ b/drivers/infiniband/hw/mlx4/main.c @@ -44,8 +44,8 @@ #include "user.h" #define DRV_NAME "mlx4_ib" -#define DRV_VERSION "0.01" -#define DRV_RELDATE "May 1, 2006" +#define DRV_VERSION "1.0" +#define DRV_RELDATE "April 4, 2008" MODULE_AUTHOR("Roland Dreier"); MODULE_DESCRIPTION("Mellanox ConnectX HCA InfiniBand driver"); @@ -99,6 +99,10 @@ static int mlx4_ib_query_device(struct ib_device *ibdev, props->device_cap_flags |= IB_DEVICE_AUTO_PATH_MIG; if (dev->dev->caps.flags & MLX4_DEV_CAP_FLAG_UD_AV_PORT) props->device_cap_flags |= IB_DEVICE_UD_AV_PORT_ENFORCE; + if (dev->dev->caps.flags & MLX4_DEV_CAP_FLAG_IPOIB_CSUM) + props->device_cap_flags |= IB_DEVICE_UD_IP_CSUM; + if (dev->dev->caps.max_gso_sz) + props->device_cap_flags |= IB_DEVICE_UD_TSO; props->vendor_id = be32_to_cpup((__be32 *) (out_mad->data + 36)) & 0xffffff; @@ -567,6 +571,7 @@ static void *mlx4_ib_add(struct mlx4_dev *dev) (1ull << IB_USER_VERBS_CMD_DEREG_MR) | (1ull << IB_USER_VERBS_CMD_CREATE_COMP_CHANNEL) | (1ull << IB_USER_VERBS_CMD_CREATE_CQ) | + (1ull << IB_USER_VERBS_CMD_RESIZE_CQ) | (1ull << IB_USER_VERBS_CMD_DESTROY_CQ) | (1ull << IB_USER_VERBS_CMD_CREATE_QP) | (1ull << IB_USER_VERBS_CMD_MODIFY_QP) | @@ -605,6 +610,8 @@ static void *mlx4_ib_add(struct mlx4_dev *dev) ibdev->ib_dev.post_send = mlx4_ib_post_send; ibdev->ib_dev.post_recv = mlx4_ib_post_recv; ibdev->ib_dev.create_cq = mlx4_ib_create_cq; + ibdev->ib_dev.modify_cq = mlx4_ib_modify_cq; + ibdev->ib_dev.resize_cq = mlx4_ib_resize_cq; ibdev->ib_dev.destroy_cq = mlx4_ib_destroy_cq; ibdev->ib_dev.poll_cq = mlx4_ib_poll_cq; ibdev->ib_dev.req_notify_cq = mlx4_ib_arm_cq; @@ -675,18 +682,20 @@ static void mlx4_ib_remove(struct mlx4_dev *dev, void *ibdev_ptr) } static void mlx4_ib_event(struct mlx4_dev *dev, void *ibdev_ptr, - enum mlx4_dev_event event, int subtype, - int port) + enum mlx4_dev_event event, int port) { struct ib_event ibev; switch (event) { - case MLX4_EVENT_TYPE_PORT_CHANGE: - ibev.event = subtype == MLX4_PORT_CHANGE_SUBTYPE_ACTIVE ? - IB_EVENT_PORT_ACTIVE : IB_EVENT_PORT_ERR; + case MLX4_DEV_EVENT_PORT_UP: + ibev.event = IB_EVENT_PORT_ACTIVE; break; - case MLX4_EVENT_TYPE_LOCAL_CATAS_ERROR: + case MLX4_DEV_EVENT_PORT_DOWN: + ibev.event = IB_EVENT_PORT_ERR; + break; + + case MLX4_DEV_EVENT_CATASTROPHIC_ERROR: ibev.event = IB_EVENT_DEVICE_FATAL; break; diff --git a/drivers/infiniband/hw/mlx4/mlx4_ib.h b/drivers/infiniband/hw/mlx4/mlx4_ib.h index 3726e451a327..9e637323c155 100644 --- a/drivers/infiniband/hw/mlx4/mlx4_ib.h +++ b/drivers/infiniband/hw/mlx4/mlx4_ib.h @@ -78,13 +78,21 @@ struct mlx4_ib_cq_buf { struct mlx4_mtt mtt; }; +struct mlx4_ib_cq_resize { + struct mlx4_ib_cq_buf buf; + int cqe; +}; + struct mlx4_ib_cq { struct ib_cq ibcq; struct mlx4_cq mcq; struct mlx4_ib_cq_buf buf; + struct mlx4_ib_cq_resize *resize_buf; struct mlx4_ib_db db; spinlock_t lock; + struct mutex resize_mutex; struct ib_umem *umem; + struct ib_umem *resize_umem; }; struct mlx4_ib_mr { @@ -110,6 +118,10 @@ struct mlx4_ib_wq { unsigned tail; }; +enum mlx4_ib_qp_flags { + MLX4_IB_QP_LSO = 1 << 0 +}; + struct mlx4_ib_qp { struct ib_qp ibqp; struct mlx4_qp mqp; @@ -129,6 +141,7 @@ struct mlx4_ib_qp { struct mlx4_mtt mtt; int buf_size; struct mutex mutex; + u32 flags; u8 port; u8 alt_port; u8 atomic_rd_en; @@ -249,6 +262,8 @@ struct ib_mr *mlx4_ib_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, struct ib_udata *udata); int mlx4_ib_dereg_mr(struct ib_mr *mr); +int mlx4_ib_modify_cq(struct ib_cq *cq, u16 cq_count, u16 cq_period); +int mlx4_ib_resize_cq(struct ib_cq *ibcq, int entries, struct ib_udata *udata); struct ib_cq *mlx4_ib_create_cq(struct ib_device *ibdev, int entries, int vector, struct ib_ucontext *context, struct ib_udata *udata); diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c index 958e205b6d7c..b75efae7e449 100644 --- a/drivers/infiniband/hw/mlx4/qp.c +++ b/drivers/infiniband/hw/mlx4/qp.c @@ -71,6 +71,7 @@ enum { static const __be32 mlx4_ib_opcode[] = { [IB_WR_SEND] = __constant_cpu_to_be32(MLX4_OPCODE_SEND), + [IB_WR_LSO] = __constant_cpu_to_be32(MLX4_OPCODE_LSO), [IB_WR_SEND_WITH_IMM] = __constant_cpu_to_be32(MLX4_OPCODE_SEND_IMM), [IB_WR_RDMA_WRITE] = __constant_cpu_to_be32(MLX4_OPCODE_RDMA_WRITE), [IB_WR_RDMA_WRITE_WITH_IMM] = __constant_cpu_to_be32(MLX4_OPCODE_RDMA_WRITE_IMM), @@ -122,7 +123,7 @@ static void *get_send_wqe(struct mlx4_ib_qp *qp, int n) */ static void stamp_send_wqe(struct mlx4_ib_qp *qp, int n, int size) { - u32 *wqe; + __be32 *wqe; int i; int s; int ind; @@ -143,7 +144,7 @@ static void stamp_send_wqe(struct mlx4_ib_qp *qp, int n, int size) buf = get_send_wqe(qp, n & (qp->sq.wqe_cnt - 1)); for (i = 64; i < s; i += 64) { wqe = buf + i; - *wqe = 0xffffffff; + *wqe = cpu_to_be32(0xffffffff); } } } @@ -242,7 +243,7 @@ static void mlx4_ib_qp_event(struct mlx4_qp *qp, enum mlx4_event type) } } -static int send_wqe_overhead(enum ib_qp_type type) +static int send_wqe_overhead(enum ib_qp_type type, u32 flags) { /* * UD WQEs must have a datagram segment. @@ -253,7 +254,8 @@ static int send_wqe_overhead(enum ib_qp_type type) switch (type) { case IB_QPT_UD: return sizeof (struct mlx4_wqe_ctrl_seg) + - sizeof (struct mlx4_wqe_datagram_seg); + sizeof (struct mlx4_wqe_datagram_seg) + + ((flags & MLX4_IB_QP_LSO) ? 64 : 0); case IB_QPT_UC: return sizeof (struct mlx4_wqe_ctrl_seg) + sizeof (struct mlx4_wqe_raddr_seg); @@ -315,7 +317,7 @@ static int set_kernel_sq_size(struct mlx4_ib_dev *dev, struct ib_qp_cap *cap, /* Sanity check SQ size before proceeding */ if (cap->max_send_wr > dev->dev->caps.max_wqes || cap->max_send_sge > dev->dev->caps.max_sq_sg || - cap->max_inline_data + send_wqe_overhead(type) + + cap->max_inline_data + send_wqe_overhead(type, qp->flags) + sizeof (struct mlx4_wqe_inline_seg) > dev->dev->caps.max_sq_desc_sz) return -EINVAL; @@ -329,7 +331,7 @@ static int set_kernel_sq_size(struct mlx4_ib_dev *dev, struct ib_qp_cap *cap, s = max(cap->max_send_sge * sizeof (struct mlx4_wqe_data_seg), cap->max_inline_data + sizeof (struct mlx4_wqe_inline_seg)) + - send_wqe_overhead(type); + send_wqe_overhead(type, qp->flags); /* * Hermon supports shrinking WQEs, such that a single work @@ -394,7 +396,8 @@ static int set_kernel_sq_size(struct mlx4_ib_dev *dev, struct ib_qp_cap *cap, } qp->sq.max_gs = ((qp->sq_max_wqes_per_wr << qp->sq.wqe_shift) - - send_wqe_overhead(type)) / sizeof (struct mlx4_wqe_data_seg); + send_wqe_overhead(type, qp->flags)) / + sizeof (struct mlx4_wqe_data_seg); qp->buf_size = (qp->rq.wqe_cnt << qp->rq.wqe_shift) + (qp->sq.wqe_cnt << qp->sq.wqe_shift); @@ -503,6 +506,9 @@ static int create_qp_common(struct mlx4_ib_dev *dev, struct ib_pd *pd, } else { qp->sq_no_prefetch = 0; + if (init_attr->create_flags & IB_QP_CREATE_IPOIB_UD_LSO) + qp->flags |= MLX4_IB_QP_LSO; + err = set_kernel_sq_size(dev, &init_attr->cap, init_attr->qp_type, qp); if (err) goto err; @@ -673,6 +679,13 @@ struct ib_qp *mlx4_ib_create_qp(struct ib_pd *pd, struct mlx4_ib_qp *qp; int err; + /* We only support LSO, and only for kernel UD QPs. */ + if (init_attr->create_flags & ~IB_QP_CREATE_IPOIB_UD_LSO) + return ERR_PTR(-EINVAL); + if (init_attr->create_flags & IB_QP_CREATE_IPOIB_UD_LSO && + (pd->uobject || init_attr->qp_type != IB_QPT_UD)) + return ERR_PTR(-EINVAL); + switch (init_attr->qp_type) { case IB_QPT_RC: case IB_QPT_UC: @@ -876,10 +889,15 @@ static int __mlx4_ib_modify_qp(struct ib_qp *ibqp, } } - if (ibqp->qp_type == IB_QPT_GSI || ibqp->qp_type == IB_QPT_SMI || - ibqp->qp_type == IB_QPT_UD) + if (ibqp->qp_type == IB_QPT_GSI || ibqp->qp_type == IB_QPT_SMI) context->mtu_msgmax = (IB_MTU_4096 << 5) | 11; - else if (attr_mask & IB_QP_PATH_MTU) { + else if (ibqp->qp_type == IB_QPT_UD) { + if (qp->flags & MLX4_IB_QP_LSO) + context->mtu_msgmax = (IB_MTU_4096 << 5) | + ilog2(dev->dev->caps.max_gso_sz); + else + context->mtu_msgmax = (IB_MTU_4096 << 5) | 11; + } else if (attr_mask & IB_QP_PATH_MTU) { if (attr->path_mtu < IB_MTU_256 || attr->path_mtu > IB_MTU_4096) { printk(KERN_ERR "path MTU (%u) is invalid\n", attr->path_mtu); @@ -1182,7 +1200,7 @@ out: } static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_send_wr *wr, - void *wqe) + void *wqe, unsigned *mlx_seg_len) { struct ib_device *ib_dev = &to_mdev(sqp->qp.ibqp.device)->ib_dev; struct mlx4_wqe_mlx_seg *mlx = wqe; @@ -1231,7 +1249,7 @@ static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_send_wr *wr, case IB_WR_SEND_WITH_IMM: sqp->ud_header.bth.opcode = IB_OPCODE_UD_SEND_ONLY_WITH_IMMEDIATE; sqp->ud_header.immediate_present = 1; - sqp->ud_header.immediate_data = wr->imm_data; + sqp->ud_header.immediate_data = wr->ex.imm_data; break; default: return -EINVAL; @@ -1303,7 +1321,9 @@ static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_send_wr *wr, i = 2; } - return ALIGN(i * sizeof (struct mlx4_wqe_inline_seg) + header_size, 16); + *mlx_seg_len = + ALIGN(i * sizeof (struct mlx4_wqe_inline_seg) + header_size, 16); + return 0; } static int mlx4_wq_overflow(struct mlx4_ib_wq *wq, int nreq, struct ib_cq *ib_cq) @@ -1396,6 +1416,34 @@ static void __set_data_seg(struct mlx4_wqe_data_seg *dseg, struct ib_sge *sg) dseg->addr = cpu_to_be64(sg->addr); } +static int build_lso_seg(struct mlx4_lso_seg *wqe, struct ib_send_wr *wr, + struct mlx4_ib_qp *qp, unsigned *lso_seg_len) +{ + unsigned halign = ALIGN(sizeof *wqe + wr->wr.ud.hlen, 16); + + /* + * This is a temporary limitation and will be removed in + * a forthcoming FW release: + */ + if (unlikely(halign > 64)) + return -EINVAL; + + if (unlikely(!(qp->flags & MLX4_IB_QP_LSO) && + wr->num_sge > qp->sq.max_gs - (halign >> 4))) + return -EINVAL; + + memcpy(wqe->header, wr->wr.ud.header, wr->wr.ud.hlen); + + /* make sure LSO header is written before overwriting stamping */ + wmb(); + + wqe->mss_hdr_size = cpu_to_be32((wr->wr.ud.mss - wr->wr.ud.hlen) << 16 | + wr->wr.ud.hlen); + + *lso_seg_len = halign; + return 0; +} + int mlx4_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, struct ib_send_wr **bad_wr) { @@ -1409,6 +1457,7 @@ int mlx4_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, unsigned ind; int uninitialized_var(stamp); int uninitialized_var(size); + unsigned seglen; int i; spin_lock_irqsave(&qp->sq.lock, flags); @@ -1436,11 +1485,14 @@ int mlx4_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, cpu_to_be32(MLX4_WQE_CTRL_CQ_UPDATE) : 0) | (wr->send_flags & IB_SEND_SOLICITED ? cpu_to_be32(MLX4_WQE_CTRL_SOLICITED) : 0) | + ((wr->send_flags & IB_SEND_IP_CSUM) ? + cpu_to_be32(MLX4_WQE_CTRL_IP_CSUM | + MLX4_WQE_CTRL_TCP_UDP_CSUM) : 0) | qp->sq_signal_bits; if (wr->opcode == IB_WR_SEND_WITH_IMM || wr->opcode == IB_WR_RDMA_WRITE_WITH_IMM) - ctrl->imm = wr->imm_data; + ctrl->imm = wr->ex.imm_data; else ctrl->imm = 0; @@ -1484,19 +1536,27 @@ int mlx4_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, set_datagram_seg(wqe, wr); wqe += sizeof (struct mlx4_wqe_datagram_seg); size += sizeof (struct mlx4_wqe_datagram_seg) / 16; + + if (wr->opcode == IB_WR_LSO) { + err = build_lso_seg(wqe, wr, qp, &seglen); + if (unlikely(err)) { + *bad_wr = wr; + goto out; + } + wqe += seglen; + size += seglen / 16; + } break; case IB_QPT_SMI: case IB_QPT_GSI: - err = build_mlx_header(to_msqp(qp), wr, ctrl); - if (err < 0) { + err = build_mlx_header(to_msqp(qp), wr, ctrl, &seglen); + if (unlikely(err)) { *bad_wr = wr; goto out; } - wqe += err; - size += err / 16; - - err = 0; + wqe += seglen; + size += seglen / 16; break; default: @@ -1725,7 +1785,9 @@ int mlx4_ib_query_qp(struct ib_qp *ibqp, struct ib_qp_attr *qp_attr, int qp_attr struct mlx4_ib_qp *qp = to_mqp(ibqp); struct mlx4_qp_context context; int mlx4_state; - int err; + int err = 0; + + mutex_lock(&qp->mutex); if (qp->state == IB_QPS_RESET) { qp_attr->qp_state = IB_QPS_RESET; @@ -1733,12 +1795,15 @@ int mlx4_ib_query_qp(struct ib_qp *ibqp, struct ib_qp_attr *qp_attr, int qp_attr } err = mlx4_qp_query(dev->dev, &qp->mqp, &context); - if (err) - return -EINVAL; + if (err) { + err = -EINVAL; + goto out; + } mlx4_state = be32_to_cpu(context.flags) >> 28; - qp_attr->qp_state = to_ib_qp_state(mlx4_state); + qp->state = to_ib_qp_state(mlx4_state); + qp_attr->qp_state = qp->state; qp_attr->path_mtu = context.mtu_msgmax >> 5; qp_attr->path_mig_state = to_ib_mig_state((be32_to_cpu(context.flags) >> 11) & 0x3); @@ -1797,6 +1862,8 @@ done: qp_init_attr->cap = qp_attr->cap; - return 0; +out: + mutex_unlock(&qp->mutex); + return err; } diff --git a/drivers/infiniband/hw/mthca/mthca_cmd.c b/drivers/infiniband/hw/mthca/mthca_cmd.c index 09a30dd12b14..54d230ee7d63 100644 --- a/drivers/infiniband/hw/mthca/mthca_cmd.c +++ b/drivers/infiniband/hw/mthca/mthca_cmd.c @@ -219,7 +219,7 @@ static void mthca_cmd_post_dbell(struct mthca_dev *dev, __raw_writel((__force u32) cpu_to_be32((1 << HCR_GO_BIT) | (1 << HCA_E_BIT) | (op_modifier << HCR_OPMOD_SHIFT) | - op), ptr + offs[6]); + op), ptr + offs[6]); wmb(); __raw_writel((__force u32) 0, ptr + offs[7]); wmb(); @@ -1339,6 +1339,10 @@ int mthca_INIT_HCA(struct mthca_dev *dev, /* Check port for UD address vector: */ *(inbox + INIT_HCA_FLAGS2_OFFSET / 4) |= cpu_to_be32(1); + /* Enable IPoIB checksumming if we can: */ + if (dev->device_cap_flags & IB_DEVICE_UD_IP_CSUM) + *(inbox + INIT_HCA_FLAGS2_OFFSET / 4) |= cpu_to_be32(7 << 3); + /* We leave wqe_quota, responder_exu, etc as 0 (default) */ /* QPC/EEC/CQC/EQC/RDB attributes */ diff --git a/drivers/infiniband/hw/mthca/mthca_cmd.h b/drivers/infiniband/hw/mthca/mthca_cmd.h index 2f976f2051d6..8928ca4a9325 100644 --- a/drivers/infiniband/hw/mthca/mthca_cmd.h +++ b/drivers/infiniband/hw/mthca/mthca_cmd.h @@ -103,6 +103,7 @@ enum { DEV_LIM_FLAG_RAW_IPV6 = 1 << 4, DEV_LIM_FLAG_RAW_ETHER = 1 << 5, DEV_LIM_FLAG_SRQ = 1 << 6, + DEV_LIM_FLAG_IPOIB_CSUM = 1 << 7, DEV_LIM_FLAG_BAD_PKEY_CNTR = 1 << 8, DEV_LIM_FLAG_BAD_QKEY_CNTR = 1 << 9, DEV_LIM_FLAG_MW = 1 << 16, diff --git a/drivers/infiniband/hw/mthca/mthca_cq.c b/drivers/infiniband/hw/mthca/mthca_cq.c index 1e1e336d3ef9..20401d2ba6b2 100644 --- a/drivers/infiniband/hw/mthca/mthca_cq.c +++ b/drivers/infiniband/hw/mthca/mthca_cq.c @@ -119,7 +119,8 @@ struct mthca_cqe { __be32 my_qpn; __be32 my_ee; __be32 rqpn; - __be16 sl_g_mlpath; + u8 sl_ipok; + u8 g_mlpath; __be16 rlid; __be32 imm_etype_pkey_eec; __be32 byte_cnt; @@ -493,6 +494,7 @@ static inline int mthca_poll_one(struct mthca_dev *dev, int is_send; int free_cqe = 1; int err = 0; + u16 checksum; cqe = next_cqe_sw(cq); if (!cqe) @@ -635,12 +637,14 @@ static inline int mthca_poll_one(struct mthca_dev *dev, break; } entry->slid = be16_to_cpu(cqe->rlid); - entry->sl = be16_to_cpu(cqe->sl_g_mlpath) >> 12; + entry->sl = cqe->sl_ipok >> 4; entry->src_qp = be32_to_cpu(cqe->rqpn) & 0xffffff; - entry->dlid_path_bits = be16_to_cpu(cqe->sl_g_mlpath) & 0x7f; + entry->dlid_path_bits = cqe->g_mlpath & 0x7f; entry->pkey_index = be32_to_cpu(cqe->imm_etype_pkey_eec) >> 16; - entry->wc_flags |= be16_to_cpu(cqe->sl_g_mlpath) & 0x80 ? - IB_WC_GRH : 0; + entry->wc_flags |= cqe->g_mlpath & 0x80 ? IB_WC_GRH : 0; + checksum = (be32_to_cpu(cqe->rqpn) >> 24) | + ((be32_to_cpu(cqe->my_ee) >> 16) & 0xff00); + entry->csum_ok = (cqe->sl_ipok & 1 && checksum == 0xffff); } entry->status = IB_WC_SUCCESS; diff --git a/drivers/infiniband/hw/mthca/mthca_dev.h b/drivers/infiniband/hw/mthca/mthca_dev.h index 7bbdd1f4e6c7..0e842e023400 100644 --- a/drivers/infiniband/hw/mthca/mthca_dev.h +++ b/drivers/infiniband/hw/mthca/mthca_dev.h @@ -54,8 +54,8 @@ #define DRV_NAME "ib_mthca" #define PFX DRV_NAME ": " -#define DRV_VERSION "0.08" -#define DRV_RELDATE "February 14, 2006" +#define DRV_VERSION "1.0" +#define DRV_RELDATE "April 4, 2008" enum { MTHCA_FLAG_DDR_HIDDEN = 1 << 1, @@ -390,11 +390,11 @@ extern void __buggy_use_of_MTHCA_PUT(void); do { \ void *__p = (char *) (source) + (offset); \ switch (sizeof (dest)) { \ - case 1: (dest) = *(u8 *) __p; break; \ - case 2: (dest) = be16_to_cpup(__p); break; \ - case 4: (dest) = be32_to_cpup(__p); break; \ - case 8: (dest) = be64_to_cpup(__p); break; \ - default: __buggy_use_of_MTHCA_GET(); \ + case 1: (dest) = *(u8 *) __p; break; \ + case 2: (dest) = be16_to_cpup(__p); break; \ + case 4: (dest) = be32_to_cpup(__p); break; \ + case 8: (dest) = be64_to_cpup(__p); break; \ + default: __buggy_use_of_MTHCA_GET(); \ } \ } while (0) diff --git a/drivers/infiniband/hw/mthca/mthca_eq.c b/drivers/infiniband/hw/mthca/mthca_eq.c index b60eb5df96e8..8bde7f98e58a 100644 --- a/drivers/infiniband/hw/mthca/mthca_eq.c +++ b/drivers/infiniband/hw/mthca/mthca_eq.c @@ -232,9 +232,9 @@ static inline struct mthca_eqe *get_eqe(struct mthca_eq *eq, u32 entry) return eq->page_list[off / PAGE_SIZE].buf + off % PAGE_SIZE; } -static inline struct mthca_eqe* next_eqe_sw(struct mthca_eq *eq) +static inline struct mthca_eqe *next_eqe_sw(struct mthca_eq *eq) { - struct mthca_eqe* eqe; + struct mthca_eqe *eqe; eqe = get_eqe(eq, eq->cons_index); return (MTHCA_EQ_ENTRY_OWNER_HW & eqe->owner) ? NULL : eqe; } diff --git a/drivers/infiniband/hw/mthca/mthca_mad.c b/drivers/infiniband/hw/mthca/mthca_mad.c index acfa41d968ee..8b7e83e6e88f 100644 --- a/drivers/infiniband/hw/mthca/mthca_mad.c +++ b/drivers/infiniband/hw/mthca/mthca_mad.c @@ -125,7 +125,7 @@ static void smp_snoop(struct ib_device *ibdev, event.device = ibdev; event.element.port_num = port_num; - if(pinfo->clientrereg_resv_subnetto & 0x80) + if (pinfo->clientrereg_resv_subnetto & 0x80) event.event = IB_EVENT_CLIENT_REREGISTER; else event.event = IB_EVENT_LID_CHANGE; diff --git a/drivers/infiniband/hw/mthca/mthca_main.c b/drivers/infiniband/hw/mthca/mthca_main.c index cd3d8adbef9f..9ebadd6e0cfb 100644 --- a/drivers/infiniband/hw/mthca/mthca_main.c +++ b/drivers/infiniband/hw/mthca/mthca_main.c @@ -267,11 +267,16 @@ static int mthca_dev_lim(struct mthca_dev *mdev, struct mthca_dev_lim *dev_lim) if (dev_lim->flags & DEV_LIM_FLAG_SRQ) mdev->mthca_flags |= MTHCA_FLAG_SRQ; + if (mthca_is_memfree(mdev)) + if (dev_lim->flags & DEV_LIM_FLAG_IPOIB_CSUM) + mdev->device_cap_flags |= IB_DEVICE_UD_IP_CSUM; + return 0; } static int mthca_init_tavor(struct mthca_dev *mdev) { + s64 size; u8 status; int err; struct mthca_dev_lim dev_lim; @@ -324,9 +329,11 @@ static int mthca_init_tavor(struct mthca_dev *mdev) if (mdev->mthca_flags & MTHCA_FLAG_SRQ) profile.num_srq = dev_lim.max_srqs; - err = mthca_make_profile(mdev, &profile, &dev_lim, &init_hca); - if (err < 0) + size = mthca_make_profile(mdev, &profile, &dev_lim, &init_hca); + if (size < 0) { + err = size; goto err_disable; + } err = mthca_INIT_HCA(mdev, &init_hca, &status); if (err) { @@ -605,7 +612,7 @@ static int mthca_init_arbel(struct mthca_dev *mdev) struct mthca_dev_lim dev_lim; struct mthca_profile profile; struct mthca_init_hca_param init_hca; - u64 icm_size; + s64 icm_size; u8 status; int err; @@ -653,7 +660,7 @@ static int mthca_init_arbel(struct mthca_dev *mdev) profile.num_srq = dev_lim.max_srqs; icm_size = mthca_make_profile(mdev, &profile, &dev_lim, &init_hca); - if ((int) icm_size < 0) { + if (icm_size < 0) { err = icm_size; goto err_stop_fw; } diff --git a/drivers/infiniband/hw/mthca/mthca_memfree.c b/drivers/infiniband/hw/mthca/mthca_memfree.c index 252db0822f6c..b224079d4e1f 100644 --- a/drivers/infiniband/hw/mthca/mthca_memfree.c +++ b/drivers/infiniband/hw/mthca/mthca_memfree.c @@ -359,12 +359,14 @@ struct mthca_icm_table *mthca_alloc_icm_table(struct mthca_dev *dev, int use_lowmem, int use_coherent) { struct mthca_icm_table *table; + int obj_per_chunk; int num_icm; unsigned chunk_size; int i; u8 status; - num_icm = (obj_size * nobj + MTHCA_TABLE_CHUNK_SIZE - 1) / MTHCA_TABLE_CHUNK_SIZE; + obj_per_chunk = MTHCA_TABLE_CHUNK_SIZE / obj_size; + num_icm = DIV_ROUND_UP(nobj, obj_per_chunk); table = kmalloc(sizeof *table + num_icm * sizeof *table->icm, GFP_KERNEL); if (!table) @@ -412,7 +414,7 @@ err: if (table->icm[i]) { mthca_UNMAP_ICM(dev, virt + i * MTHCA_TABLE_CHUNK_SIZE, MTHCA_TABLE_CHUNK_SIZE / MTHCA_ICM_PAGE_SIZE, - &status); + &status); mthca_free_icm(dev, table->icm[i], table->coherent); } diff --git a/drivers/infiniband/hw/mthca/mthca_profile.c b/drivers/infiniband/hw/mthca/mthca_profile.c index 26bf86d1cfcd..605a8d57fac6 100644 --- a/drivers/infiniband/hw/mthca/mthca_profile.c +++ b/drivers/infiniband/hw/mthca/mthca_profile.c @@ -63,7 +63,7 @@ enum { MTHCA_NUM_PDS = 1 << 15 }; -u64 mthca_make_profile(struct mthca_dev *dev, +s64 mthca_make_profile(struct mthca_dev *dev, struct mthca_profile *request, struct mthca_dev_lim *dev_lim, struct mthca_init_hca_param *init_hca) @@ -77,7 +77,7 @@ u64 mthca_make_profile(struct mthca_dev *dev, }; u64 mem_base, mem_avail; - u64 total_size = 0; + s64 total_size = 0; struct mthca_resource *profile; struct mthca_resource tmp; int i, j; diff --git a/drivers/infiniband/hw/mthca/mthca_profile.h b/drivers/infiniband/hw/mthca/mthca_profile.h index 94641808f97f..e76cb62d8e32 100644 --- a/drivers/infiniband/hw/mthca/mthca_profile.h +++ b/drivers/infiniband/hw/mthca/mthca_profile.h @@ -53,7 +53,7 @@ struct mthca_profile { int fmr_reserved_mtts; }; -u64 mthca_make_profile(struct mthca_dev *mdev, +s64 mthca_make_profile(struct mthca_dev *mdev, struct mthca_profile *request, struct mthca_dev_lim *dev_lim, struct mthca_init_hca_param *init_hca); diff --git a/drivers/infiniband/hw/mthca/mthca_provider.c b/drivers/infiniband/hw/mthca/mthca_provider.c index 9e491df6419c..81b257e18bb6 100644 --- a/drivers/infiniband/hw/mthca/mthca_provider.c +++ b/drivers/infiniband/hw/mthca/mthca_provider.c @@ -60,7 +60,7 @@ static int mthca_query_device(struct ib_device *ibdev, struct ib_smp *in_mad = NULL; struct ib_smp *out_mad = NULL; int err = -ENOMEM; - struct mthca_dev* mdev = to_mdev(ibdev); + struct mthca_dev *mdev = to_mdev(ibdev); u8 status; @@ -540,6 +540,9 @@ static struct ib_qp *mthca_create_qp(struct ib_pd *pd, struct mthca_qp *qp; int err; + if (init_attr->create_flags) + return ERR_PTR(-EINVAL); + switch (init_attr->qp_type) { case IB_QPT_RC: case IB_QPT_UC: diff --git a/drivers/infiniband/hw/mthca/mthca_qp.c b/drivers/infiniband/hw/mthca/mthca_qp.c index db5595bbf7f0..09dc3614cf2c 100644 --- a/drivers/infiniband/hw/mthca/mthca_qp.c +++ b/drivers/infiniband/hw/mthca/mthca_qp.c @@ -437,29 +437,34 @@ int mthca_query_qp(struct ib_qp *ibqp, struct ib_qp_attr *qp_attr, int qp_attr_m int mthca_state; u8 status; + mutex_lock(&qp->mutex); + if (qp->state == IB_QPS_RESET) { qp_attr->qp_state = IB_QPS_RESET; goto done; } mailbox = mthca_alloc_mailbox(dev, GFP_KERNEL); - if (IS_ERR(mailbox)) - return PTR_ERR(mailbox); + if (IS_ERR(mailbox)) { + err = PTR_ERR(mailbox); + goto out; + } err = mthca_QUERY_QP(dev, qp->qpn, 0, mailbox, &status); if (err) - goto out; + goto out_mailbox; if (status) { mthca_warn(dev, "QUERY_QP returned status %02x\n", status); err = -EINVAL; - goto out; + goto out_mailbox; } qp_param = mailbox->buf; context = &qp_param->context; mthca_state = be32_to_cpu(context->flags) >> 28; - qp_attr->qp_state = to_ib_qp_state(mthca_state); + qp->state = to_ib_qp_state(mthca_state); + qp_attr->qp_state = qp->state; qp_attr->path_mtu = context->mtu_msgmax >> 5; qp_attr->path_mig_state = to_ib_mig_state((be32_to_cpu(context->flags) >> 11) & 0x3); @@ -506,8 +511,11 @@ done: qp_init_attr->cap = qp_attr->cap; -out: +out_mailbox: mthca_free_mailbox(dev, mailbox); + +out: + mutex_unlock(&qp->mutex); return err; } @@ -1532,7 +1540,7 @@ static int build_mlx_header(struct mthca_dev *dev, struct mthca_sqp *sqp, case IB_WR_SEND_WITH_IMM: sqp->ud_header.bth.opcode = IB_OPCODE_UD_SEND_ONLY_WITH_IMMEDIATE; sqp->ud_header.immediate_present = 1; - sqp->ud_header.immediate_data = wr->imm_data; + sqp->ud_header.immediate_data = wr->ex.imm_data; break; default: return -EINVAL; @@ -1679,7 +1687,7 @@ int mthca_tavor_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, cpu_to_be32(1); if (wr->opcode == IB_WR_SEND_WITH_IMM || wr->opcode == IB_WR_RDMA_WRITE_WITH_IMM) - ((struct mthca_next_seg *) wqe)->imm = wr->imm_data; + ((struct mthca_next_seg *) wqe)->imm = wr->ex.imm_data; wqe += sizeof (struct mthca_next_seg); size = sizeof (struct mthca_next_seg) / 16; @@ -2015,10 +2023,12 @@ int mthca_arbel_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, cpu_to_be32(MTHCA_NEXT_CQ_UPDATE) : 0) | ((wr->send_flags & IB_SEND_SOLICITED) ? cpu_to_be32(MTHCA_NEXT_SOLICIT) : 0) | + ((wr->send_flags & IB_SEND_IP_CSUM) ? + cpu_to_be32(MTHCA_NEXT_IP_CSUM | MTHCA_NEXT_TCP_UDP_CSUM) : 0) | cpu_to_be32(1); if (wr->opcode == IB_WR_SEND_WITH_IMM || wr->opcode == IB_WR_RDMA_WRITE_WITH_IMM) - ((struct mthca_next_seg *) wqe)->imm = wr->imm_data; + ((struct mthca_next_seg *) wqe)->imm = wr->ex.imm_data; wqe += sizeof (struct mthca_next_seg); size = sizeof (struct mthca_next_seg) / 16; diff --git a/drivers/infiniband/hw/mthca/mthca_wqe.h b/drivers/infiniband/hw/mthca/mthca_wqe.h index f6a66fe78e48..b3551a8dea1d 100644 --- a/drivers/infiniband/hw/mthca/mthca_wqe.h +++ b/drivers/infiniband/hw/mthca/mthca_wqe.h @@ -38,14 +38,16 @@ #include <linux/types.h> enum { - MTHCA_NEXT_DBD = 1 << 7, - MTHCA_NEXT_FENCE = 1 << 6, - MTHCA_NEXT_CQ_UPDATE = 1 << 3, - MTHCA_NEXT_EVENT_GEN = 1 << 2, - MTHCA_NEXT_SOLICIT = 1 << 1, - - MTHCA_MLX_VL15 = 1 << 17, - MTHCA_MLX_SLR = 1 << 16 + MTHCA_NEXT_DBD = 1 << 7, + MTHCA_NEXT_FENCE = 1 << 6, + MTHCA_NEXT_CQ_UPDATE = 1 << 3, + MTHCA_NEXT_EVENT_GEN = 1 << 2, + MTHCA_NEXT_SOLICIT = 1 << 1, + MTHCA_NEXT_IP_CSUM = 1 << 4, + MTHCA_NEXT_TCP_UDP_CSUM = 1 << 5, + + MTHCA_MLX_VL15 = 1 << 17, + MTHCA_MLX_SLR = 1 << 16 }; enum { diff --git a/drivers/infiniband/hw/nes/nes.c b/drivers/infiniband/hw/nes/nes.c index b2112f5a422f..b00b0e3a91dc 100644 --- a/drivers/infiniband/hw/nes/nes.c +++ b/drivers/infiniband/hw/nes/nes.c @@ -65,7 +65,6 @@ MODULE_LICENSE("Dual BSD/GPL"); MODULE_VERSION(DRV_VERSION); int max_mtu = 9000; -int nics_per_function = 1; int interrupt_mod_interval = 0; @@ -93,15 +92,9 @@ module_param_named(debug_level, nes_debug_level, uint, 0644); MODULE_PARM_DESC(debug_level, "Enable debug output level"); LIST_HEAD(nes_adapter_list); -LIST_HEAD(nes_dev_list); +static LIST_HEAD(nes_dev_list); atomic_t qps_destroyed; -atomic_t cqp_reqs_allocated; -atomic_t cqp_reqs_freed; -atomic_t cqp_reqs_dynallocated; -atomic_t cqp_reqs_dynfreed; -atomic_t cqp_reqs_queued; -atomic_t cqp_reqs_redriven; static void nes_print_macaddr(struct net_device *netdev); static irqreturn_t nes_interrupt(int, void *); @@ -310,7 +303,7 @@ void nes_rem_ref(struct ib_qp *ibqp) if (atomic_read(&nesqp->refcount) == 0) { printk(KERN_INFO PFX "%s: Reference count already 0 for QP%d, last aeq = 0x%04X.\n", - __FUNCTION__, ibqp->qp_num, nesqp->last_aeq); + __func__, ibqp->qp_num, nesqp->last_aeq); BUG(); } @@ -751,13 +744,13 @@ static void __devexit nes_remove(struct pci_dev *pcidev) list_del(&nesdev->list); nes_destroy_cqp(nesdev); + + free_irq(pcidev->irq, nesdev); tasklet_kill(&nesdev->dpc_tasklet); /* Deallocate the Adapter Structure */ nes_destroy_adapter(nesdev->nesadapter); - free_irq(pcidev->irq, nesdev); - if (nesdev->msi_enabled) { pci_disable_msi(pcidev); } diff --git a/drivers/infiniband/hw/nes/nes.h b/drivers/infiniband/hw/nes/nes.h index a48b288618ec..1626124a156d 100644 --- a/drivers/infiniband/hw/nes/nes.h +++ b/drivers/infiniband/hw/nes/nes.h @@ -143,12 +143,12 @@ #ifdef CONFIG_INFINIBAND_NES_DEBUG #define nes_debug(level, fmt, args...) \ if (level & nes_debug_level) \ - printk(KERN_ERR PFX "%s[%u]: " fmt, __FUNCTION__, __LINE__, ##args) + printk(KERN_ERR PFX "%s[%u]: " fmt, __func__, __LINE__, ##args) #define assert(expr) \ if (!(expr)) { \ printk(KERN_ERR PFX "Assertion failed! %s, %s, %s, line %d\n", \ - #expr, __FILE__, __FUNCTION__, __LINE__); \ + #expr, __FILE__, __func__, __LINE__); \ } #define NES_EVENT_TIMEOUT 1200000 @@ -166,7 +166,6 @@ if (!(expr)) { \ #include "nes_cm.h" extern int max_mtu; -extern int nics_per_function; #define max_frame_len (max_mtu+ETH_HLEN) extern int interrupt_mod_interval; extern int nes_if_count; @@ -177,9 +176,6 @@ extern unsigned int nes_drv_opt; extern unsigned int nes_debug_level; extern struct list_head nes_adapter_list; -extern struct list_head nes_dev_list; - -extern struct nes_cm_core *g_cm_core; extern atomic_t cm_connects; extern atomic_t cm_accepts; @@ -209,7 +205,6 @@ extern atomic_t cm_nodes_destroyed; extern atomic_t cm_accel_dropped_pkts; extern atomic_t cm_resets_recvd; -extern u32 crit_err_count; extern u32 int_mod_timer_init; extern u32 int_mod_cq_depth_256; extern u32 int_mod_cq_depth_128; @@ -219,14 +214,6 @@ extern u32 int_mod_cq_depth_16; extern u32 int_mod_cq_depth_4; extern u32 int_mod_cq_depth_1; -extern atomic_t cqp_reqs_allocated; -extern atomic_t cqp_reqs_freed; -extern atomic_t cqp_reqs_dynallocated; -extern atomic_t cqp_reqs_dynfreed; -extern atomic_t cqp_reqs_queued; -extern atomic_t cqp_reqs_redriven; - - struct nes_device { struct nes_adapter *nesadapter; void __iomem *regs; @@ -412,7 +399,7 @@ static inline int nes_alloc_resource(struct nes_adapter *nesadapter, if (resource_num >= max_resources) { resource_num = find_first_zero_bit(resource_array, max_resources); if (resource_num >= max_resources) { - printk(KERN_ERR PFX "%s: No available resourcess.\n", __FUNCTION__); + printk(KERN_ERR PFX "%s: No available resourcess.\n", __func__); spin_unlock_irqrestore(&nesadapter->resource_lock, flags); return -EMFILE; } @@ -510,9 +497,6 @@ struct ib_qp *nes_get_qp(struct ib_device *, int); /* nes_hw.c */ struct nes_adapter *nes_init_adapter(struct nes_device *, u8); void nes_nic_init_timer_defaults(struct nes_device *, u8); -unsigned int nes_reset_adapter_ne020(struct nes_device *, u8 *); -int nes_init_serdes(struct nes_device *, u8, u8, u8); -void nes_init_csr_ne020(struct nes_device *, u8, u8); void nes_destroy_adapter(struct nes_adapter *); int nes_init_cqp(struct nes_device *); int nes_init_phy(struct nes_device *); @@ -520,20 +504,12 @@ int nes_init_nic_qp(struct nes_device *, struct net_device *); void nes_destroy_nic_qp(struct nes_vnic *); int nes_napi_isr(struct nes_device *); void nes_dpc(unsigned long); -void nes_process_ceq(struct nes_device *, struct nes_hw_ceq *); -void nes_process_aeq(struct nes_device *, struct nes_hw_aeq *); -void nes_process_mac_intr(struct nes_device *, u32); -void nes_nic_napi_ce_handler(struct nes_device *, struct nes_hw_nic_cq *); void nes_nic_ce_handler(struct nes_device *, struct nes_hw_nic_cq *); -void nes_cqp_ce_handler(struct nes_device *, struct nes_hw_cq *); -void nes_process_iwarp_aeqe(struct nes_device *, struct nes_hw_aeqe *); void nes_iwarp_ce_handler(struct nes_device *, struct nes_hw_cq *); int nes_destroy_cqp(struct nes_device *); int nes_nic_cm_xmit(struct sk_buff *, struct net_device *); /* nes_nic.c */ -void nes_netdev_set_multicast_list(struct net_device *); -void nes_netdev_exit(struct nes_vnic *); struct net_device *nes_netdev_init(struct nes_device *, void __iomem *); void nes_netdev_destroy(struct net_device *); int nes_nic_cm_xmit(struct sk_buff *, struct net_device *); @@ -544,7 +520,6 @@ int nes_cm_recv(struct sk_buff *, struct net_device *); void nes_update_arp(unsigned char *, u32, u32, u16, u16); void nes_manage_arp_cache(struct net_device *, unsigned char *, u32, u32); void nes_sock_release(struct nes_qp *, unsigned long *); -struct nes_cm_core *nes_cm_alloc_core(void); void flush_wqes(struct nes_device *nesdev, struct nes_qp *, u32, u32); int nes_manage_apbvt(struct nes_vnic *, u32, u32, u32); int nes_cm_disconn(struct nes_qp *); @@ -556,7 +531,6 @@ int nes_modify_qp(struct ib_qp *, struct ib_qp_attr *, int, struct ib_udata *); struct nes_ib_device *nes_init_ofa_device(struct net_device *); void nes_destroy_ofa_device(struct nes_ib_device *); int nes_register_ofa_device(struct nes_ib_device *); -void nes_unregister_ofa_device(struct nes_ib_device *); /* nes_util.c */ int nes_read_eeprom_values(struct nes_device *, struct nes_adapter *); diff --git a/drivers/infiniband/hw/nes/nes_cm.c b/drivers/infiniband/hw/nes/nes_cm.c index 0bef878e0f65..d0738623bcf3 100644 --- a/drivers/infiniband/hw/nes/nes_cm.c +++ b/drivers/infiniband/hw/nes/nes_cm.c @@ -80,7 +80,30 @@ static struct nes_cm_listener *mini_cm_listen(struct nes_cm_core *, static int add_ref_cm_node(struct nes_cm_node *); static int rem_ref_cm_node(struct nes_cm_core *, struct nes_cm_node *); static int mini_cm_del_listen(struct nes_cm_core *, struct nes_cm_listener *); - +static struct sk_buff *form_cm_frame(struct sk_buff *, struct nes_cm_node *, + void *, u32, void *, u32, u8); +static struct sk_buff *get_free_pkt(struct nes_cm_node *cm_node); + +static struct nes_cm_node *mini_cm_connect(struct nes_cm_core *, + struct nes_vnic *, + struct ietf_mpa_frame *, + struct nes_cm_info *); +static int mini_cm_accept(struct nes_cm_core *, struct ietf_mpa_frame *, + struct nes_cm_node *); +static int mini_cm_reject(struct nes_cm_core *, struct ietf_mpa_frame *, + struct nes_cm_node *); +static int mini_cm_close(struct nes_cm_core *, struct nes_cm_node *); +static int mini_cm_recv_pkt(struct nes_cm_core *, struct nes_vnic *, + struct sk_buff *); +static int mini_cm_dealloc_core(struct nes_cm_core *); +static int mini_cm_get(struct nes_cm_core *); +static int mini_cm_set(struct nes_cm_core *, u32, u32); +static int nes_cm_disconn_true(struct nes_qp *); +static int nes_cm_post_event(struct nes_cm_event *event); +static int nes_disconnect(struct nes_qp *nesqp, int abrupt); +static void nes_disconnect_worker(struct work_struct *work); +static int send_ack(struct nes_cm_node *cm_node); +static int send_fin(struct nes_cm_node *cm_node, struct sk_buff *skb); /* External CM API Interface */ /* instance of function pointers for client API */ @@ -99,7 +122,7 @@ static struct nes_cm_ops nes_cm_api = { mini_cm_set }; -struct nes_cm_core *g_cm_core; +static struct nes_cm_core *g_cm_core; atomic_t cm_connects; atomic_t cm_accepts; @@ -149,7 +172,7 @@ static struct nes_cm_event *create_event(struct nes_cm_node *cm_node, /** * send_mpa_request */ -int send_mpa_request(struct nes_cm_node *cm_node) +static int send_mpa_request(struct nes_cm_node *cm_node) { struct sk_buff *skb; int ret; @@ -243,8 +266,9 @@ static int handle_exception_pkt(struct nes_cm_node *cm_node, struct sk_buff *skb * form_cm_frame - get a free packet and build empty frame Use * node info to build. */ -struct sk_buff *form_cm_frame(struct sk_buff *skb, struct nes_cm_node *cm_node, - void *options, u32 optionsize, void *data, u32 datasize, u8 flags) +static struct sk_buff *form_cm_frame(struct sk_buff *skb, struct nes_cm_node *cm_node, + void *options, u32 optionsize, void *data, + u32 datasize, u8 flags) { struct tcphdr *tcph; struct iphdr *iph; @@ -342,7 +366,6 @@ static void print_core(struct nes_cm_core *core) if (!core) return; nes_debug(NES_DBG_CM, "---------------------------------------------\n"); - nes_debug(NES_DBG_CM, "Session ID : %u \n", atomic_read(&core->session_id)); nes_debug(NES_DBG_CM, "State : %u \n", core->state); @@ -395,7 +418,7 @@ int schedule_nes_timer(struct nes_cm_node *cm_node, struct sk_buff *skb, } if (type == NES_TIMER_TYPE_SEND) { - new_send->seq_num = htonl(tcp_hdr(skb)->seq); + new_send->seq_num = ntohl(tcp_hdr(skb)->seq); atomic_inc(&new_send->skb->users); ret = nes_nic_cm_xmit(new_send->skb, cm_node->netdev); @@ -420,7 +443,7 @@ int schedule_nes_timer(struct nes_cm_node *cm_node, struct sk_buff *skb, spin_unlock_irqrestore(&cm_node->retrans_list_lock, flags); } if (type == NES_TIMER_TYPE_RECV) { - new_send->seq_num = htonl(tcp_hdr(skb)->seq); + new_send->seq_num = ntohl(tcp_hdr(skb)->seq); new_send->timetosend = jiffies; spin_lock_irqsave(&cm_node->recv_list_lock, flags); list_add_tail(&new_send->list, &cm_node->recv_list); @@ -442,7 +465,7 @@ int schedule_nes_timer(struct nes_cm_node *cm_node, struct sk_buff *skb, /** * nes_cm_timer_tick */ -void nes_cm_timer_tick(unsigned long pass) +static void nes_cm_timer_tick(unsigned long pass) { unsigned long flags, qplockflags; unsigned long nexttimeout = jiffies + NES_LONG_TIME; @@ -644,7 +667,7 @@ void nes_cm_timer_tick(unsigned long pass) /** * send_syn */ -int send_syn(struct nes_cm_node *cm_node, u32 sendack) +static int send_syn(struct nes_cm_node *cm_node, u32 sendack) { int ret; int flags = SET_SYN; @@ -710,7 +733,7 @@ int send_syn(struct nes_cm_node *cm_node, u32 sendack) /** * send_reset */ -int send_reset(struct nes_cm_node *cm_node) +static int send_reset(struct nes_cm_node *cm_node) { int ret; struct sk_buff *skb = get_free_pkt(cm_node); @@ -732,7 +755,7 @@ int send_reset(struct nes_cm_node *cm_node) /** * send_ack */ -int send_ack(struct nes_cm_node *cm_node) +static int send_ack(struct nes_cm_node *cm_node) { int ret; struct sk_buff *skb = get_free_pkt(cm_node); @@ -752,7 +775,7 @@ int send_ack(struct nes_cm_node *cm_node) /** * send_fin */ -int send_fin(struct nes_cm_node *cm_node, struct sk_buff *skb) +static int send_fin(struct nes_cm_node *cm_node, struct sk_buff *skb) { int ret; @@ -775,7 +798,7 @@ int send_fin(struct nes_cm_node *cm_node, struct sk_buff *skb) /** * get_free_pkt */ -struct sk_buff *get_free_pkt(struct nes_cm_node *cm_node) +static struct sk_buff *get_free_pkt(struct nes_cm_node *cm_node) { struct sk_buff *skb, *new_skb; @@ -820,7 +843,6 @@ static struct nes_cm_node *find_node(struct nes_cm_core *cm_core, { unsigned long flags; u32 hashkey; - struct list_head *list_pos; struct list_head *hte; struct nes_cm_node *cm_node; @@ -835,8 +857,7 @@ static struct nes_cm_node *find_node(struct nes_cm_core *cm_core, /* walk list and find cm_node associated with this session ID */ spin_lock_irqsave(&cm_core->ht_lock, flags); - list_for_each(list_pos, hte) { - cm_node = container_of(list_pos, struct nes_cm_node, list); + list_for_each_entry(cm_node, hte, list) { /* compare quad, return node handle if a match */ nes_debug(NES_DBG_CM, "finding node %x:%x =? %x:%x ^ %x:%x =? %x:%x\n", cm_node->loc_addr, cm_node->loc_port, @@ -864,13 +885,11 @@ static struct nes_cm_listener *find_listener(struct nes_cm_core *cm_core, nes_addr_t dst_addr, u16 dst_port, enum nes_cm_listener_state listener_state) { unsigned long flags; - struct list_head *listen_list; struct nes_cm_listener *listen_node; /* walk list and find cm_node associated with this session ID */ spin_lock_irqsave(&cm_core->listen_list_lock, flags); - list_for_each(listen_list, &cm_core->listen_list.list) { - listen_node = container_of(listen_list, struct nes_cm_listener, list); + list_for_each_entry(listen_node, &cm_core->listen_list.list, list) { /* compare node pair, return node handle if a match */ if (((listen_node->loc_addr == dst_addr) || listen_node->loc_addr == 0x00000000) && @@ -1014,7 +1033,7 @@ static void nes_addr_send_arp(u32 dst_ip) fl.nl_u.ip4_u.daddr = htonl(dst_ip); if (ip_route_output_key(&init_net, &rt, &fl)) { printk("%s: ip_route_output_key failed for 0x%08X\n", - __FUNCTION__, dst_ip); + __func__, dst_ip); return; } @@ -1077,8 +1096,6 @@ static struct nes_cm_node *make_cm_node(struct nes_cm_core *cm_core, cm_node->tcp_cntxt.rcv_nxt = 0; /* get a unique session ID , add thread_id to an upcounter to handle race */ atomic_inc(&cm_core->node_cnt); - atomic_inc(&cm_core->session_id); - cm_node->session_id = (u32)(atomic_read(&cm_core->session_id) + current->tgid); cm_node->conn_type = cm_info->conn_type; cm_node->apbvt_set = 0; cm_node->accept_pend = 0; @@ -1239,7 +1256,7 @@ static int process_options(struct nes_cm_node *cm_node, u8 *optionsloc, u32 opti continue; case OPTION_NUMBER_MSS: nes_debug(NES_DBG_CM, "%s: MSS Length: %d Offset: %d Size: %d\n", - __FUNCTION__, + __func__, all_options->as_mss.length, offset, optionsize); got_mss_option = 1; if (all_options->as_mss.length != 4) { @@ -1272,8 +1289,8 @@ static int process_options(struct nes_cm_node *cm_node, u8 *optionsloc, u32 opti /** * process_packet */ -int process_packet(struct nes_cm_node *cm_node, struct sk_buff *skb, - struct nes_cm_core *cm_core) +static int process_packet(struct nes_cm_node *cm_node, struct sk_buff *skb, + struct nes_cm_core *cm_core) { int optionsize; int datasize; @@ -1360,7 +1377,7 @@ int process_packet(struct nes_cm_node *cm_node, struct sk_buff *skb, if (optionsize) { u8 *optionsloc = (u8 *)&tcph[1]; if (process_options(cm_node, optionsloc, optionsize, (u32)tcph->syn)) { - nes_debug(NES_DBG_CM, "%s: Node %p, Sending RESET\n", __FUNCTION__, cm_node); + nes_debug(NES_DBG_CM, "%s: Node %p, Sending RESET\n", __func__, cm_node); send_reset(cm_node); if (cm_node->state != NES_CM_STATE_SYN_SENT) rem_ref_cm_node(cm_core, cm_node); @@ -1605,9 +1622,7 @@ static struct nes_cm_listener *mini_cm_listen(struct nes_cm_core *cm_core, listener->cm_core = cm_core; listener->nesvnic = nesvnic; atomic_inc(&cm_core->node_cnt); - atomic_inc(&cm_core->session_id); - listener->session_id = (u32)(atomic_read(&cm_core->session_id) + current->tgid); listener->conn_type = cm_info->conn_type; listener->backlog = cm_info->backlog; listener->listener_state = NES_CM_LISTENER_ACTIVE_STATE; @@ -1631,9 +1646,10 @@ static struct nes_cm_listener *mini_cm_listen(struct nes_cm_core *cm_core, /** * mini_cm_connect - make a connection node with params */ -struct nes_cm_node *mini_cm_connect(struct nes_cm_core *cm_core, - struct nes_vnic *nesvnic, struct ietf_mpa_frame *mpa_frame, - struct nes_cm_info *cm_info) +static struct nes_cm_node *mini_cm_connect(struct nes_cm_core *cm_core, + struct nes_vnic *nesvnic, + struct ietf_mpa_frame *mpa_frame, + struct nes_cm_info *cm_info) { int ret = 0; struct nes_cm_node *cm_node; @@ -1717,8 +1733,8 @@ struct nes_cm_node *mini_cm_connect(struct nes_cm_core *cm_core, * mini_cm_accept - accept a connection * This function is never called */ -int mini_cm_accept(struct nes_cm_core *cm_core, struct ietf_mpa_frame *mpa_frame, - struct nes_cm_node *cm_node) +static int mini_cm_accept(struct nes_cm_core *cm_core, struct ietf_mpa_frame *mpa_frame, + struct nes_cm_node *cm_node) { return 0; } @@ -1727,9 +1743,9 @@ int mini_cm_accept(struct nes_cm_core *cm_core, struct ietf_mpa_frame *mpa_frame /** * mini_cm_reject - reject and teardown a connection */ -int mini_cm_reject(struct nes_cm_core *cm_core, - struct ietf_mpa_frame *mpa_frame, - struct nes_cm_node *cm_node) +static int mini_cm_reject(struct nes_cm_core *cm_core, + struct ietf_mpa_frame *mpa_frame, + struct nes_cm_node *cm_node) { int ret = 0; struct sk_buff *skb; @@ -1761,7 +1777,7 @@ int mini_cm_reject(struct nes_cm_core *cm_core, /** * mini_cm_close */ -int mini_cm_close(struct nes_cm_core *cm_core, struct nes_cm_node *cm_node) +static int mini_cm_close(struct nes_cm_core *cm_core, struct nes_cm_node *cm_node) { int ret = 0; @@ -1808,8 +1824,8 @@ int mini_cm_close(struct nes_cm_core *cm_core, struct nes_cm_node *cm_node) * recv_pkt - recv an ETHERNET packet, and process it through CM * node state machine */ -int mini_cm_recv_pkt(struct nes_cm_core *cm_core, struct nes_vnic *nesvnic, - struct sk_buff *skb) +static int mini_cm_recv_pkt(struct nes_cm_core *cm_core, struct nes_vnic *nesvnic, + struct sk_buff *skb) { struct nes_cm_node *cm_node = NULL; struct nes_cm_listener *listener = NULL; @@ -1898,7 +1914,7 @@ int mini_cm_recv_pkt(struct nes_cm_core *cm_core, struct nes_vnic *nesvnic, /** * nes_cm_alloc_core - allocate a top level instance of a cm core */ -struct nes_cm_core *nes_cm_alloc_core(void) +static struct nes_cm_core *nes_cm_alloc_core(void) { int i; @@ -1919,7 +1935,6 @@ struct nes_cm_core *nes_cm_alloc_core(void) cm_core->state = NES_CM_STATE_INITED; cm_core->free_tx_pkt_max = NES_CM_DEFAULT_FREE_PKTS; - atomic_set(&cm_core->session_id, 0); atomic_set(&cm_core->events_posted, 0); /* init the packet lists */ @@ -1958,7 +1973,7 @@ struct nes_cm_core *nes_cm_alloc_core(void) /** * mini_cm_dealloc_core - deallocate a top level instance of a cm core */ -int mini_cm_dealloc_core(struct nes_cm_core *cm_core) +static int mini_cm_dealloc_core(struct nes_cm_core *cm_core) { nes_debug(NES_DBG_CM, "De-Alloc CM Core (%p)\n", cm_core); @@ -1983,7 +1998,7 @@ int mini_cm_dealloc_core(struct nes_cm_core *cm_core) /** * mini_cm_get */ -int mini_cm_get(struct nes_cm_core *cm_core) +static int mini_cm_get(struct nes_cm_core *cm_core) { return cm_core->state; } @@ -1992,7 +2007,7 @@ int mini_cm_get(struct nes_cm_core *cm_core) /** * mini_cm_set */ -int mini_cm_set(struct nes_cm_core *cm_core, u32 type, u32 value) +static int mini_cm_set(struct nes_cm_core *cm_core, u32 type, u32 value) { int ret = 0; @@ -2109,7 +2124,7 @@ int nes_cm_disconn(struct nes_qp *nesqp) /** * nes_disconnect_worker */ -void nes_disconnect_worker(struct work_struct *work) +static void nes_disconnect_worker(struct work_struct *work) { struct nes_qp *nesqp = container_of(work, struct nes_qp, disconn_work); @@ -2122,7 +2137,7 @@ void nes_disconnect_worker(struct work_struct *work) /** * nes_cm_disconn_true */ -int nes_cm_disconn_true(struct nes_qp *nesqp) +static int nes_cm_disconn_true(struct nes_qp *nesqp) { unsigned long flags; int ret = 0; @@ -2265,7 +2280,7 @@ int nes_cm_disconn_true(struct nes_qp *nesqp) /** * nes_disconnect */ -int nes_disconnect(struct nes_qp *nesqp, int abrupt) +static int nes_disconnect(struct nes_qp *nesqp, int abrupt) { int ret = 0; struct nes_vnic *nesvnic; @@ -2482,7 +2497,7 @@ int nes_accept(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) } if (ret) printk("%s[%u] OFA CM event_handler returned, ret=%d\n", - __FUNCTION__, __LINE__, ret); + __func__, __LINE__, ret); return 0; } @@ -2650,7 +2665,7 @@ int nes_create_listen(struct iw_cm_id *cm_id, int backlog) cm_node = g_cm_core->api->listen(g_cm_core, nesvnic, &cm_info); if (!cm_node) { printk("%s[%u] Error returned from listen API call\n", - __FUNCTION__, __LINE__); + __func__, __LINE__); return -ENOMEM; } @@ -2740,7 +2755,7 @@ int nes_cm_stop(void) * cm_event_connected * handle a connected event, setup QPs and HW */ -void cm_event_connected(struct nes_cm_event *event) +static void cm_event_connected(struct nes_cm_event *event) { u64 u64temp; struct nes_qp *nesqp; @@ -2864,7 +2879,7 @@ void cm_event_connected(struct nes_cm_event *event) if (ret) printk("%s[%u] OFA CM event_handler returned, ret=%d\n", - __FUNCTION__, __LINE__, ret); + __func__, __LINE__, ret); nes_debug(NES_DBG_CM, "Exiting connect thread for QP%u. jiffies = %lu\n", nesqp->hwqp.qp_id, jiffies ); @@ -2877,7 +2892,7 @@ void cm_event_connected(struct nes_cm_event *event) /** * cm_event_connect_error */ -void cm_event_connect_error(struct nes_cm_event *event) +static void cm_event_connect_error(struct nes_cm_event *event) { struct nes_qp *nesqp; struct iw_cm_id *cm_id; @@ -2919,7 +2934,7 @@ void cm_event_connect_error(struct nes_cm_event *event) nes_debug(NES_DBG_CM, "OFA CM event_handler returned, ret=%d\n", ret); if (ret) printk("%s[%u] OFA CM event_handler returned, ret=%d\n", - __FUNCTION__, __LINE__, ret); + __func__, __LINE__, ret); nes_rem_ref(&nesqp->ibqp); cm_id->rem_ref(cm_id); @@ -2930,7 +2945,7 @@ void cm_event_connect_error(struct nes_cm_event *event) /** * cm_event_reset */ -void cm_event_reset(struct nes_cm_event *event) +static void cm_event_reset(struct nes_cm_event *event) { struct nes_qp *nesqp; struct iw_cm_id *cm_id; @@ -2973,7 +2988,7 @@ void cm_event_reset(struct nes_cm_event *event) /** * cm_event_mpa_req */ -void cm_event_mpa_req(struct nes_cm_event *event) +static void cm_event_mpa_req(struct nes_cm_event *event) { struct iw_cm_id *cm_id; struct iw_cm_event cm_event; @@ -3007,7 +3022,7 @@ void cm_event_mpa_req(struct nes_cm_event *event) ret = cm_id->event_handler(cm_id, &cm_event); if (ret) printk("%s[%u] OFA CM event_handler returned, ret=%d\n", - __FUNCTION__, __LINE__, ret); + __func__, __LINE__, ret); return; } @@ -3019,7 +3034,7 @@ static void nes_cm_event_handler(struct work_struct *); * nes_cm_post_event * post an event to the cm event handler */ -int nes_cm_post_event(struct nes_cm_event *event) +static int nes_cm_post_event(struct nes_cm_event *event) { atomic_inc(&event->cm_node->cm_core->events_posted); add_ref_cm_node(event->cm_node); diff --git a/drivers/infiniband/hw/nes/nes_cm.h b/drivers/infiniband/hw/nes/nes_cm.h index a59f0a7fb278..7717cb2ab500 100644 --- a/drivers/infiniband/hw/nes/nes_cm.h +++ b/drivers/infiniband/hw/nes/nes_cm.h @@ -225,7 +225,6 @@ enum nes_cm_listener_state { struct nes_cm_listener { struct list_head list; - u64 session_id; struct nes_cm_core *cm_core; u8 loc_mac[ETH_ALEN]; nes_addr_t loc_addr; @@ -242,7 +241,6 @@ struct nes_cm_listener { /* per connection node and node state information */ struct nes_cm_node { - u64 session_id; u32 hashkey; nes_addr_t loc_addr, rem_addr; @@ -327,7 +325,6 @@ struct nes_cm_event { struct nes_cm_core { enum nes_cm_node_state state; - atomic_t session_id; atomic_t listen_node_cnt; struct nes_cm_node listen_list; @@ -383,35 +380,10 @@ struct nes_cm_ops { int (*set)(struct nes_cm_core *, u32, u32); }; - -int send_mpa_request(struct nes_cm_node *); -struct sk_buff *form_cm_frame(struct sk_buff *, struct nes_cm_node *, - void *, u32, void *, u32, u8); int schedule_nes_timer(struct nes_cm_node *, struct sk_buff *, enum nes_timer_type, int, int); -void nes_cm_timer_tick(unsigned long); -int send_syn(struct nes_cm_node *, u32); -int send_reset(struct nes_cm_node *); -int send_ack(struct nes_cm_node *); -int send_fin(struct nes_cm_node *, struct sk_buff *); -struct sk_buff *get_free_pkt(struct nes_cm_node *); -int process_packet(struct nes_cm_node *, struct sk_buff *, struct nes_cm_core *); - -struct nes_cm_node * mini_cm_connect(struct nes_cm_core *, - struct nes_vnic *, struct ietf_mpa_frame *, struct nes_cm_info *); -int mini_cm_accept(struct nes_cm_core *, struct ietf_mpa_frame *, struct nes_cm_node *); -int mini_cm_reject(struct nes_cm_core *, struct ietf_mpa_frame *, struct nes_cm_node *); -int mini_cm_close(struct nes_cm_core *, struct nes_cm_node *); -int mini_cm_recv_pkt(struct nes_cm_core *, struct nes_vnic *, struct sk_buff *); -struct nes_cm_core *mini_cm_alloc_core(struct nes_cm_info *); -int mini_cm_dealloc_core(struct nes_cm_core *); -int mini_cm_get(struct nes_cm_core *); -int mini_cm_set(struct nes_cm_core *, u32, u32); int nes_cm_disconn(struct nes_qp *); -void nes_disconnect_worker(struct work_struct *); -int nes_cm_disconn_true(struct nes_qp *); -int nes_disconnect(struct nes_qp *, int); int nes_accept(struct iw_cm_id *, struct iw_cm_conn_param *); int nes_reject(struct iw_cm_id *, const void *, u8); @@ -423,11 +395,4 @@ int nes_cm_recv(struct sk_buff *, struct net_device *); int nes_cm_start(void); int nes_cm_stop(void); -/* CM event handler functions */ -void cm_event_connected(struct nes_cm_event *); -void cm_event_connect_error(struct nes_cm_event *); -void cm_event_reset(struct nes_cm_event *); -void cm_event_mpa_req(struct nes_cm_event *); -int nes_cm_post_event(struct nes_cm_event *); - #endif /* NES_CM_H */ diff --git a/drivers/infiniband/hw/nes/nes_hw.c b/drivers/infiniband/hw/nes/nes_hw.c index 49e53e4c1ebe..aa53aab91bf8 100644 --- a/drivers/infiniband/hw/nes/nes_hw.c +++ b/drivers/infiniband/hw/nes/nes_hw.c @@ -41,7 +41,7 @@ #include "nes.h" -u32 crit_err_count = 0; +static u32 crit_err_count; u32 int_mod_timer_init; u32 int_mod_cq_depth_256; u32 int_mod_cq_depth_128; @@ -53,6 +53,17 @@ u32 int_mod_cq_depth_1; #include "nes_cm.h" +static void nes_cqp_ce_handler(struct nes_device *nesdev, struct nes_hw_cq *cq); +static void nes_init_csr_ne020(struct nes_device *nesdev, u8 hw_rev, u8 port_count); +static int nes_init_serdes(struct nes_device *nesdev, u8 hw_rev, u8 port_count, + u8 OneG_Mode); +static void nes_nic_napi_ce_handler(struct nes_device *nesdev, struct nes_hw_nic_cq *cq); +static void nes_process_aeq(struct nes_device *nesdev, struct nes_hw_aeq *aeq); +static void nes_process_ceq(struct nes_device *nesdev, struct nes_hw_ceq *ceq); +static void nes_process_iwarp_aeqe(struct nes_device *nesdev, + struct nes_hw_aeqe *aeqe); +static void nes_process_mac_intr(struct nes_device *nesdev, u32 mac_number); +static unsigned int nes_reset_adapter_ne020(struct nes_device *nesdev, u8 *OneG_Mode); #ifdef CONFIG_INFINIBAND_NES_DEBUG static unsigned char *nes_iwarp_state_str[] = { @@ -370,7 +381,7 @@ struct nes_adapter *nes_init_adapter(struct nes_device *nesdev, u8 hw_rev) { nesadapter->et_use_adaptive_rx_coalesce = 1; nesadapter->timer_int_limit = NES_TIMER_INT_LIMIT_DYNAMIC; nesadapter->et_rx_coalesce_usecs_irq = 0; - printk(PFX "%s: Using Adaptive Interrupt Moderation\n", __FUNCTION__); + printk(PFX "%s: Using Adaptive Interrupt Moderation\n", __func__); } /* Setup and enable the periodic timer */ if (nesadapter->et_rx_coalesce_usecs_irq) @@ -382,7 +393,7 @@ struct nes_adapter *nes_init_adapter(struct nes_device *nesdev, u8 hw_rev) { nesadapter->base_pd = 1; nesadapter->device_cap_flags = - IB_DEVICE_ZERO_STAG | IB_DEVICE_SEND_W_INV | IB_DEVICE_MEM_WINDOW; + IB_DEVICE_ZERO_STAG | IB_DEVICE_MEM_WINDOW; nesadapter->allocated_qps = (unsigned long *)&(((unsigned char *)nesadapter) [(sizeof(struct nes_adapter)+(sizeof(unsigned long)-1))&(~(sizeof(unsigned long)-1))]); @@ -572,7 +583,7 @@ struct nes_adapter *nes_init_adapter(struct nes_device *nesdev, u8 hw_rev) { if (vendor_id == 0xffff) break; } - nes_debug(NES_DBG_INIT, "%s %d functions found for %s.\n", __FUNCTION__, + nes_debug(NES_DBG_INIT, "%s %d functions found for %s.\n", __func__, func_index, pci_name(nesdev->pcidev)); nesadapter->adapter_fcn_count = func_index; @@ -583,7 +594,7 @@ struct nes_adapter *nes_init_adapter(struct nes_device *nesdev, u8 hw_rev) { /** * nes_reset_adapter_ne020 */ -unsigned int nes_reset_adapter_ne020(struct nes_device *nesdev, u8 *OneG_Mode) +static unsigned int nes_reset_adapter_ne020(struct nes_device *nesdev, u8 *OneG_Mode) { u32 port_count; u32 u32temp; @@ -691,7 +702,8 @@ unsigned int nes_reset_adapter_ne020(struct nes_device *nesdev, u8 *OneG_Mode) /** * nes_init_serdes */ -int nes_init_serdes(struct nes_device *nesdev, u8 hw_rev, u8 port_count, u8 OneG_Mode) +static int nes_init_serdes(struct nes_device *nesdev, u8 hw_rev, u8 port_count, + u8 OneG_Mode) { int i; u32 u32temp; @@ -739,7 +751,7 @@ int nes_init_serdes(struct nes_device *nesdev, u8 hw_rev, u8 port_count, u8 One & 0x0000000f)) != 0x0000000f) && (i++ < 5000)) mdelay(1); if (i >= 5000) { - printk("%s: Init: serdes 1 not ready, status=%x\n", __FUNCTION__, u32temp); + printk("%s: Init: serdes 1 not ready, status=%x\n", __func__, u32temp); /* return 1; */ } nes_write_indexed(nesdev, NES_IDX_ETH_SERDES_TX_EMP1, 0x000bdef7); @@ -760,7 +772,7 @@ int nes_init_serdes(struct nes_device *nesdev, u8 hw_rev, u8 port_count, u8 One * nes_init_csr_ne020 * Initialize registers for ne020 hardware */ -void nes_init_csr_ne020(struct nes_device *nesdev, u8 hw_rev, u8 port_count) +static void nes_init_csr_ne020(struct nes_device *nesdev, u8 hw_rev, u8 port_count) { u32 u32temp; @@ -1204,7 +1216,7 @@ int nes_init_phy(struct nes_device *nesdev) if (nesadapter->OneG_Mode) { nes_debug(NES_DBG_PHY, "1G PHY, mac_index = %d.\n", mac_index); if (nesadapter->phy_type[mac_index] == NES_PHY_TYPE_1G) { - printk(PFX "%s: Programming mdc config for 1G\n", __FUNCTION__); + printk(PFX "%s: Programming mdc config for 1G\n", __func__); tx_config = nes_read_indexed(nesdev, NES_IDX_MAC_TX_CONFIG); tx_config |= 0x04; nes_write_indexed(nesdev, NES_IDX_MAC_TX_CONFIG, tx_config); @@ -1358,7 +1370,7 @@ static void nes_replenish_nic_rq(struct nes_vnic *nesvnic) static void nes_rq_wqes_timeout(unsigned long parm) { struct nes_vnic *nesvnic = (struct nes_vnic *)parm; - printk("%s: Timer fired.\n", __FUNCTION__); + printk("%s: Timer fired.\n", __func__); atomic_set(&nesvnic->rx_skb_timer_running, 0); if (atomic_read(&nesvnic->rx_skbs_needed)) nes_replenish_nic_rq(nesvnic); @@ -1909,7 +1921,7 @@ void nes_dpc(unsigned long param) /** * nes_process_ceq */ -void nes_process_ceq(struct nes_device *nesdev, struct nes_hw_ceq *ceq) +static void nes_process_ceq(struct nes_device *nesdev, struct nes_hw_ceq *ceq) { u64 u64temp; struct nes_hw_cq *cq; @@ -1949,7 +1961,7 @@ void nes_process_ceq(struct nes_device *nesdev, struct nes_hw_ceq *ceq) /** * nes_process_aeq */ -void nes_process_aeq(struct nes_device *nesdev, struct nes_hw_aeq *aeq) +static void nes_process_aeq(struct nes_device *nesdev, struct nes_hw_aeq *aeq) { // u64 u64temp; u32 head; @@ -2060,7 +2072,7 @@ static void nes_reset_link(struct nes_device *nesdev, u32 mac_index) /** * nes_process_mac_intr */ -void nes_process_mac_intr(struct nes_device *nesdev, u32 mac_number) +static void nes_process_mac_intr(struct nes_device *nesdev, u32 mac_number) { unsigned long flags; u32 pcs_control_status; @@ -2163,7 +2175,7 @@ void nes_process_mac_intr(struct nes_device *nesdev, u32 mac_number) temp_phy_data = phy_data; } while (1); nes_debug(NES_DBG_PHY, "%s: Phy data = 0x%04X, link was %s.\n", - __FUNCTION__, phy_data, nesadapter->mac_link_down ? "DOWN" : "UP"); + __func__, phy_data, nesadapter->mac_link_down ? "DOWN" : "UP"); } else { phy_data = (0x0f0f0000 == (pcs_control_status & 0x0f1f0000)) ? 4 : 0; @@ -2205,7 +2217,7 @@ void nes_process_mac_intr(struct nes_device *nesdev, u32 mac_number) -void nes_nic_napi_ce_handler(struct nes_device *nesdev, struct nes_hw_nic_cq *cq) +static void nes_nic_napi_ce_handler(struct nes_device *nesdev, struct nes_hw_nic_cq *cq) { struct nes_vnic *nesvnic = container_of(cq, struct nes_vnic, nic_cq); @@ -2428,7 +2440,7 @@ void nes_nic_ce_handler(struct nes_device *nesdev, struct nes_hw_nic_cq *cq) /** * nes_cqp_ce_handler */ -void nes_cqp_ce_handler(struct nes_device *nesdev, struct nes_hw_cq *cq) +static void nes_cqp_ce_handler(struct nes_device *nesdev, struct nes_hw_cq *cq) { u64 u64temp; unsigned long flags; @@ -2567,7 +2579,8 @@ void nes_cqp_ce_handler(struct nes_device *nesdev, struct nes_hw_cq *cq) /** * nes_process_iwarp_aeqe */ -void nes_process_iwarp_aeqe(struct nes_device *nesdev, struct nes_hw_aeqe *aeqe) +static void nes_process_iwarp_aeqe(struct nes_device *nesdev, + struct nes_hw_aeqe *aeqe) { u64 context; u64 aeqe_context = 0; @@ -2819,7 +2832,7 @@ void nes_process_iwarp_aeqe(struct nes_device *nesdev, struct nes_hw_aeqe *aeqe) le32_to_cpu(aeqe->aeqe_words[NES_AEQE_COMP_QP_CQ_ID_IDX])); if (resource_allocated) { printk(KERN_ERR PFX "%s: Processing an NES_AEQE_AEID_CQ_OPERATION_ERROR event on CQ%u\n", - __FUNCTION__, le32_to_cpu(aeqe->aeqe_words[NES_AEQE_COMP_QP_CQ_ID_IDX])); + __func__, le32_to_cpu(aeqe->aeqe_words[NES_AEQE_COMP_QP_CQ_ID_IDX])); } break; case NES_AEQE_AEID_DDP_UBE_DDP_MESSAGE_TOO_LONG_FOR_AVAILABLE_BUFFER: diff --git a/drivers/infiniband/hw/nes/nes_nic.c b/drivers/infiniband/hw/nes/nes_nic.c index eee77da61935..34166641f207 100644 --- a/drivers/infiniband/hw/nes/nes_nic.c +++ b/drivers/infiniband/hw/nes/nes_nic.c @@ -802,7 +802,7 @@ static int nes_netdev_set_mac_address(struct net_device *netdev, void *p) memcpy(netdev->dev_addr, mac_addr->sa_data, netdev->addr_len); printk(PFX "%s: Address length = %d, Address = %02X%02X%02X%02X%02X%02X..\n", - __FUNCTION__, netdev->addr_len, + __func__, netdev->addr_len, mac_addr->sa_data[0], mac_addr->sa_data[1], mac_addr->sa_data[2], mac_addr->sa_data[3], mac_addr->sa_data[4], mac_addr->sa_data[5]); @@ -832,7 +832,7 @@ static int nes_netdev_set_mac_address(struct net_device *netdev, void *p) /** * nes_netdev_set_multicast_list */ -void nes_netdev_set_multicast_list(struct net_device *netdev) +static void nes_netdev_set_multicast_list(struct net_device *netdev) { struct nes_vnic *nesvnic = netdev_priv(netdev); struct nes_device *nesdev = nesvnic->nesdev; @@ -947,28 +947,6 @@ static int nes_netdev_change_mtu(struct net_device *netdev, int new_mtu) return ret; } - -/** - * nes_netdev_exit - destroy network device - */ -void nes_netdev_exit(struct nes_vnic *nesvnic) -{ - struct net_device *netdev = nesvnic->netdev; - struct nes_ib_device *nesibdev = nesvnic->nesibdev; - - nes_debug(NES_DBG_SHUTDOWN, "\n"); - - // destroy the ibdevice if RDMA enabled - if ((nesvnic->rdma_enabled)&&(nesvnic->of_device_registered)) { - nes_destroy_ofa_device( nesibdev ); - nesvnic->of_device_registered = 0; - nesvnic->nesibdev = NULL; - } - unregister_netdev(netdev); - nes_debug(NES_DBG_SHUTDOWN, "\n"); -} - - #define NES_ETHTOOL_STAT_COUNT 55 static const char nes_ethtool_stringset[NES_ETHTOOL_STAT_COUNT][ETH_GSTRING_LEN] = { "Link Change Interrupts", diff --git a/drivers/infiniband/hw/nes/nes_utils.c b/drivers/infiniband/hw/nes/nes_utils.c index c4ec6ac63461..f9db07c2717d 100644 --- a/drivers/infiniband/hw/nes/nes_utils.c +++ b/drivers/infiniband/hw/nes/nes_utils.c @@ -566,7 +566,7 @@ struct nes_cqp_request *nes_get_cqp_request(struct nes_device *nesdev) cqp_request); } else printk(KERN_ERR PFX "%s: Could not allocated a CQP request.\n", - __FUNCTION__); + __func__); return cqp_request; } diff --git a/drivers/infiniband/hw/nes/nes_verbs.c b/drivers/infiniband/hw/nes/nes_verbs.c index a651e9d9f0ef..7c27420c2240 100644 --- a/drivers/infiniband/hw/nes/nes_verbs.c +++ b/drivers/infiniband/hw/nes/nes_verbs.c @@ -49,6 +49,7 @@ atomic_t mod_qp_timouts; atomic_t qps_created; atomic_t sw_qps_destroyed; +static void nes_unregister_ofa_device(struct nes_ib_device *nesibdev); /** * nes_alloc_mw @@ -1043,10 +1044,10 @@ static int nes_setup_virt_qp(struct nes_qp *nesqp, struct nes_pbl *nespbl, u8 sq_pbl_entries; pbl_entries = nespbl->pbl_size >> 3; - nes_debug(NES_DBG_QP, "Userspace PBL, pbl_size=%u, pbl_entries = %d pbl_vbase=%p, pbl_pbase=%p\n", + nes_debug(NES_DBG_QP, "Userspace PBL, pbl_size=%u, pbl_entries = %d pbl_vbase=%p, pbl_pbase=%lx\n", nespbl->pbl_size, pbl_entries, (void *)nespbl->pbl_vbase, - (void *)nespbl->pbl_pbase); + (unsigned long) nespbl->pbl_pbase); pbl = (__le64 *) nespbl->pbl_vbase; /* points to first pbl entry */ /* now lets set the sq_vbase as well as rq_vbase addrs we will assign */ /* the first pbl to be fro the rq_vbase... */ @@ -1074,9 +1075,9 @@ static int nes_setup_virt_qp(struct nes_qp *nesqp, struct nes_pbl *nespbl, /* nesqp->hwqp.rq_vbase = bus_to_virt(*pbl); */ /*nesqp->hwqp.rq_vbase = phys_to_virt(*pbl); */ - nes_debug(NES_DBG_QP, "QP sq_vbase= %p sq_pbase=%p rq_vbase=%p rq_pbase=%p\n", - nesqp->hwqp.sq_vbase, (void *)nesqp->hwqp.sq_pbase, - nesqp->hwqp.rq_vbase, (void *)nesqp->hwqp.rq_pbase); + nes_debug(NES_DBG_QP, "QP sq_vbase= %p sq_pbase=%lx rq_vbase=%p rq_pbase=%lx\n", + nesqp->hwqp.sq_vbase, (unsigned long) nesqp->hwqp.sq_pbase, + nesqp->hwqp.rq_vbase, (unsigned long) nesqp->hwqp.rq_pbase); spin_lock_irqsave(&nesadapter->pbl_lock, flags); if (!nesadapter->free_256pbl) { pci_free_consistent(nesdev->pcidev, nespbl->pbl_size, nespbl->pbl_vbase, @@ -1251,6 +1252,9 @@ static struct ib_qp *nes_create_qp(struct ib_pd *ibpd, u8 rq_encoded_size; /* int counter; */ + if (init_attr->create_flags) + return ERR_PTR(-EINVAL); + atomic_inc(&qps_created); switch (init_attr->qp_type) { case IB_QPT_RC: @@ -1908,13 +1912,13 @@ static int nes_destroy_cq(struct ib_cq *ib_cq) nesadapter->free_256pbl++; if (nesadapter->free_256pbl > nesadapter->max_256pbl) { printk(KERN_ERR PFX "%s: free 256B PBLs(%u) has exceeded the max(%u)\n", - __FUNCTION__, nesadapter->free_256pbl, nesadapter->max_256pbl); + __func__, nesadapter->free_256pbl, nesadapter->max_256pbl); } } else if (nescq->virtual_cq == 2) { nesadapter->free_4kpbl++; if (nesadapter->free_4kpbl > nesadapter->max_4kpbl) { printk(KERN_ERR PFX "%s: free 4K PBLs(%u) has exceeded the max(%u)\n", - __FUNCTION__, nesadapter->free_4kpbl, nesadapter->max_4kpbl); + __func__, nesadapter->free_4kpbl, nesadapter->max_4kpbl); } opcode |= NES_CQP_CQ_4KB_CHUNK; } @@ -2653,10 +2657,10 @@ static struct ib_mr *nes_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, nespbl->pbl_vbase = (u64 *)pbl; nespbl->user_base = start; - nes_debug(NES_DBG_MR, "Allocated PBL memory, %u bytes, pbl_pbase=%p," + nes_debug(NES_DBG_MR, "Allocated PBL memory, %u bytes, pbl_pbase=%lx," " pbl_vbase=%p user_base=0x%lx\n", - nespbl->pbl_size, (void *)nespbl->pbl_pbase, - (void*)nespbl->pbl_vbase, nespbl->user_base); + nespbl->pbl_size, (unsigned long) nespbl->pbl_pbase, + (void *) nespbl->pbl_vbase, nespbl->user_base); list_for_each_entry(chunk, ®ion->chunk_list, list) { for (nmap_index = 0; nmap_index < chunk->nmap; ++nmap_index) { @@ -3895,14 +3899,11 @@ int nes_register_ofa_device(struct nes_ib_device *nesibdev) /** * nes_unregister_ofa_device */ -void nes_unregister_ofa_device(struct nes_ib_device *nesibdev) +static void nes_unregister_ofa_device(struct nes_ib_device *nesibdev) { struct nes_vnic *nesvnic = nesibdev->nesvnic; int i; - if (nesibdev == NULL) - return; - for (i = 0; i < ARRAY_SIZE(nes_class_attributes); ++i) { class_device_remove_file(&nesibdev->ibdev.class_dev, nes_class_attributes[i]); } diff --git a/drivers/infiniband/ulp/ipoib/Makefile b/drivers/infiniband/ulp/ipoib/Makefile index 98ee38e8c2c4..3090100f0de7 100644 --- a/drivers/infiniband/ulp/ipoib/Makefile +++ b/drivers/infiniband/ulp/ipoib/Makefile @@ -4,7 +4,8 @@ ib_ipoib-y := ipoib_main.o \ ipoib_ib.o \ ipoib_multicast.o \ ipoib_verbs.o \ - ipoib_vlan.o + ipoib_vlan.o \ + ipoib_ethtool.o ib_ipoib-$(CONFIG_INFINIBAND_IPOIB_CM) += ipoib_cm.o ib_ipoib-$(CONFIG_INFINIBAND_IPOIB_DEBUG) += ipoib_fs.o diff --git a/drivers/infiniband/ulp/ipoib/ipoib.h b/drivers/infiniband/ulp/ipoib/ipoib.h index 054fab8e27a0..73b2b176ad0e 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib.h +++ b/drivers/infiniband/ulp/ipoib/ipoib.h @@ -87,6 +87,7 @@ enum { IPOIB_MCAST_STARTED = 8, IPOIB_FLAG_ADMIN_CM = 9, IPOIB_FLAG_UMCAST = 10, + IPOIB_FLAG_CSUM = 11, IPOIB_MAX_BACKOFF_SECONDS = 16, @@ -241,6 +242,11 @@ struct ipoib_cm_dev_priv { int num_frags; }; +struct ipoib_ethtool_st { + u16 coalesce_usecs; + u16 max_coalesced_frames; +}; + /* * Device private locking: tx_lock protects members used in TX fast * path (and we use LLTX so upper layers don't do extra locking). @@ -318,6 +324,8 @@ struct ipoib_dev_priv { struct dentry *mcg_dentry; struct dentry *path_dentry; #endif + int hca_caps; + struct ipoib_ethtool_st ethtool; }; struct ipoib_ah { @@ -458,6 +466,8 @@ void ipoib_pkey_poll(struct work_struct *work); int ipoib_pkey_dev_delay_open(struct net_device *dev); void ipoib_drain_cq(struct net_device *dev); +void ipoib_set_ethtool_ops(struct net_device *dev); + #ifdef CONFIG_INFINIBAND_IPOIB_CM #define IPOIB_FLAGS_RC 0x80 diff --git a/drivers/infiniband/ulp/ipoib/ipoib_cm.c b/drivers/infiniband/ulp/ipoib/ipoib_cm.c index 2490b2d79dbb..9db7b0bd9134 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_cm.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_cm.c @@ -1007,9 +1007,9 @@ static int ipoib_cm_modify_tx_init(struct net_device *dev, struct ipoib_dev_priv *priv = netdev_priv(dev); struct ib_qp_attr qp_attr; int qp_attr_mask, ret; - ret = ib_find_cached_pkey(priv->ca, priv->port, priv->pkey, &qp_attr.pkey_index); + ret = ib_find_pkey(priv->ca, priv->port, priv->pkey, &qp_attr.pkey_index); if (ret) { - ipoib_warn(priv, "pkey 0x%x not in cache: %d\n", priv->pkey, ret); + ipoib_warn(priv, "pkey 0x%x not found: %d\n", priv->pkey, ret); return ret; } @@ -1383,6 +1383,10 @@ static ssize_t set_mode(struct device *d, struct device_attribute *attr, set_bit(IPOIB_FLAG_ADMIN_CM, &priv->flags); ipoib_warn(priv, "enabling connected mode " "will cause multicast packet drops\n"); + + dev->features &= ~(NETIF_F_IP_CSUM | NETIF_F_SG | NETIF_F_TSO); + priv->tx_wr.send_flags &= ~IB_SEND_IP_CSUM; + ipoib_flush_paths(dev); return count; } @@ -1391,6 +1395,13 @@ static ssize_t set_mode(struct device *d, struct device_attribute *attr, clear_bit(IPOIB_FLAG_ADMIN_CM, &priv->flags); dev->mtu = min(priv->mcast_mtu, dev->mtu); ipoib_flush_paths(dev); + + if (test_bit(IPOIB_FLAG_CSUM, &priv->flags)) { + dev->features |= NETIF_F_IP_CSUM | NETIF_F_SG; + if (priv->hca_caps & IB_DEVICE_UD_TSO) + dev->features |= NETIF_F_TSO; + } + return count; } diff --git a/drivers/infiniband/ulp/ipoib/ipoib_ethtool.c b/drivers/infiniband/ulp/ipoib/ipoib_ethtool.c new file mode 100644 index 000000000000..9a47428366c9 --- /dev/null +++ b/drivers/infiniband/ulp/ipoib/ipoib_ethtool.c @@ -0,0 +1,99 @@ +/* + * Copyright (c) 2007 Mellanox Technologies. All rights reserved. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * OpenIB.org BSD license below: + * + * Redistribution and use in source and binary forms, with or + * without modification, are permitted provided that the following + * conditions are met: + * + * - Redistributions of source code must retain the above + * copyright notice, this list of conditions and the following + * disclaimer. + * + * - Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials + * provided with the distribution. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + */ + +#include <linux/kernel.h> +#include <linux/ethtool.h> +#include <linux/netdevice.h> + +#include "ipoib.h" + +static void ipoib_get_drvinfo(struct net_device *netdev, + struct ethtool_drvinfo *drvinfo) +{ + strncpy(drvinfo->driver, "ipoib", sizeof(drvinfo->driver) - 1); +} + +static int ipoib_get_coalesce(struct net_device *dev, + struct ethtool_coalesce *coal) +{ + struct ipoib_dev_priv *priv = netdev_priv(dev); + + coal->rx_coalesce_usecs = priv->ethtool.coalesce_usecs; + coal->tx_coalesce_usecs = priv->ethtool.coalesce_usecs; + coal->rx_max_coalesced_frames = priv->ethtool.max_coalesced_frames; + coal->tx_max_coalesced_frames = priv->ethtool.max_coalesced_frames; + + return 0; +} + +static int ipoib_set_coalesce(struct net_device *dev, + struct ethtool_coalesce *coal) +{ + struct ipoib_dev_priv *priv = netdev_priv(dev); + int ret; + + /* + * Since IPoIB uses a single CQ for both rx and tx, we assume + * that rx params dictate the configuration. These values are + * saved in the private data and returned when ipoib_get_coalesce() + * is called. + */ + if (coal->rx_coalesce_usecs > 0xffff || + coal->rx_max_coalesced_frames > 0xffff) + return -EINVAL; + + ret = ib_modify_cq(priv->cq, coal->rx_max_coalesced_frames, + coal->rx_coalesce_usecs); + if (ret && ret != -ENOSYS) { + ipoib_warn(priv, "failed modifying CQ (%d)\n", ret); + return ret; + } + + coal->tx_coalesce_usecs = coal->rx_coalesce_usecs; + coal->tx_max_coalesced_frames = coal->rx_max_coalesced_frames; + priv->ethtool.coalesce_usecs = coal->rx_coalesce_usecs; + priv->ethtool.max_coalesced_frames = coal->rx_max_coalesced_frames; + + return 0; +} + +static const struct ethtool_ops ipoib_ethtool_ops = { + .get_drvinfo = ipoib_get_drvinfo, + .get_tso = ethtool_op_get_tso, + .get_coalesce = ipoib_get_coalesce, + .set_coalesce = ipoib_set_coalesce, +}; + +void ipoib_set_ethtool_ops(struct net_device *dev) +{ + SET_ETHTOOL_OPS(dev, &ipoib_ethtool_ops); +} diff --git a/drivers/infiniband/ulp/ipoib/ipoib_ib.c b/drivers/infiniband/ulp/ipoib/ipoib_ib.c index 08c4396cf418..0205eb7c1bd3 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_ib.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_ib.c @@ -39,6 +39,8 @@ #include <linux/dma-mapping.h> #include <rdma/ib_cache.h> +#include <linux/ip.h> +#include <linux/tcp.h> #include "ipoib.h" @@ -231,6 +233,10 @@ static void ipoib_ib_handle_rx_wc(struct net_device *dev, struct ib_wc *wc) skb->dev = dev; /* XXX get correct PACKET_ type here */ skb->pkt_type = PACKET_HOST; + + if (test_bit(IPOIB_FLAG_CSUM, &priv->flags) && likely(wc->csum_ok)) + skb->ip_summed = CHECKSUM_UNNECESSARY; + netif_receive_skb(skb); repost: @@ -245,29 +251,37 @@ static int ipoib_dma_map_tx(struct ib_device *ca, struct sk_buff *skb = tx_req->skb; u64 *mapping = tx_req->mapping; int i; + int off; - mapping[0] = ib_dma_map_single(ca, skb->data, skb_headlen(skb), - DMA_TO_DEVICE); - if (unlikely(ib_dma_mapping_error(ca, mapping[0]))) - return -EIO; + if (skb_headlen(skb)) { + mapping[0] = ib_dma_map_single(ca, skb->data, skb_headlen(skb), + DMA_TO_DEVICE); + if (unlikely(ib_dma_mapping_error(ca, mapping[0]))) + return -EIO; + + off = 1; + } else + off = 0; for (i = 0; i < skb_shinfo(skb)->nr_frags; ++i) { skb_frag_t *frag = &skb_shinfo(skb)->frags[i]; - mapping[i + 1] = ib_dma_map_page(ca, frag->page, + mapping[i + off] = ib_dma_map_page(ca, frag->page, frag->page_offset, frag->size, DMA_TO_DEVICE); - if (unlikely(ib_dma_mapping_error(ca, mapping[i + 1]))) + if (unlikely(ib_dma_mapping_error(ca, mapping[i + off]))) goto partial_error; } return 0; partial_error: - ib_dma_unmap_single(ca, mapping[0], skb_headlen(skb), DMA_TO_DEVICE); - for (; i > 0; --i) { skb_frag_t *frag = &skb_shinfo(skb)->frags[i - 1]; - ib_dma_unmap_page(ca, mapping[i], frag->size, DMA_TO_DEVICE); + ib_dma_unmap_page(ca, mapping[i - !off], frag->size, DMA_TO_DEVICE); } + + if (off) + ib_dma_unmap_single(ca, mapping[0], skb_headlen(skb), DMA_TO_DEVICE); + return -EIO; } @@ -277,12 +291,17 @@ static void ipoib_dma_unmap_tx(struct ib_device *ca, struct sk_buff *skb = tx_req->skb; u64 *mapping = tx_req->mapping; int i; + int off; - ib_dma_unmap_single(ca, mapping[0], skb_headlen(skb), DMA_TO_DEVICE); + if (skb_headlen(skb)) { + ib_dma_unmap_single(ca, mapping[0], skb_headlen(skb), DMA_TO_DEVICE); + off = 1; + } else + off = 0; for (i = 0; i < skb_shinfo(skb)->nr_frags; ++i) { skb_frag_t *frag = &skb_shinfo(skb)->frags[i]; - ib_dma_unmap_page(ca, mapping[i + 1], frag->size, + ib_dma_unmap_page(ca, mapping[i + off], frag->size, DMA_TO_DEVICE); } } @@ -388,24 +407,40 @@ void ipoib_ib_completion(struct ib_cq *cq, void *dev_ptr) static inline int post_send(struct ipoib_dev_priv *priv, unsigned int wr_id, struct ib_ah *address, u32 qpn, - u64 *mapping, int headlen, - skb_frag_t *frags, - int nr_frags) + struct ipoib_tx_buf *tx_req, + void *head, int hlen) { struct ib_send_wr *bad_wr; - int i; + int i, off; + struct sk_buff *skb = tx_req->skb; + skb_frag_t *frags = skb_shinfo(skb)->frags; + int nr_frags = skb_shinfo(skb)->nr_frags; + u64 *mapping = tx_req->mapping; + + if (skb_headlen(skb)) { + priv->tx_sge[0].addr = mapping[0]; + priv->tx_sge[0].length = skb_headlen(skb); + off = 1; + } else + off = 0; - priv->tx_sge[0].addr = mapping[0]; - priv->tx_sge[0].length = headlen; for (i = 0; i < nr_frags; ++i) { - priv->tx_sge[i + 1].addr = mapping[i + 1]; - priv->tx_sge[i + 1].length = frags[i].size; + priv->tx_sge[i + off].addr = mapping[i + off]; + priv->tx_sge[i + off].length = frags[i].size; } - priv->tx_wr.num_sge = nr_frags + 1; + priv->tx_wr.num_sge = nr_frags + off; priv->tx_wr.wr_id = wr_id; priv->tx_wr.wr.ud.remote_qpn = qpn; priv->tx_wr.wr.ud.ah = address; + if (head) { + priv->tx_wr.wr.ud.mss = skb_shinfo(skb)->gso_size; + priv->tx_wr.wr.ud.header = head; + priv->tx_wr.wr.ud.hlen = hlen; + priv->tx_wr.opcode = IB_WR_LSO; + } else + priv->tx_wr.opcode = IB_WR_SEND; + return ib_post_send(priv->qp, &priv->tx_wr, &bad_wr); } @@ -414,14 +449,30 @@ void ipoib_send(struct net_device *dev, struct sk_buff *skb, { struct ipoib_dev_priv *priv = netdev_priv(dev); struct ipoib_tx_buf *tx_req; - - if (unlikely(skb->len > priv->mcast_mtu + IPOIB_ENCAP_LEN)) { - ipoib_warn(priv, "packet len %d (> %d) too long to send, dropping\n", - skb->len, priv->mcast_mtu + IPOIB_ENCAP_LEN); - ++dev->stats.tx_dropped; - ++dev->stats.tx_errors; - ipoib_cm_skb_too_long(dev, skb, priv->mcast_mtu); - return; + int hlen; + void *phead; + + if (skb_is_gso(skb)) { + hlen = skb_transport_offset(skb) + tcp_hdrlen(skb); + phead = skb->data; + if (unlikely(!skb_pull(skb, hlen))) { + ipoib_warn(priv, "linear data too small\n"); + ++dev->stats.tx_dropped; + ++dev->stats.tx_errors; + dev_kfree_skb_any(skb); + return; + } + } else { + if (unlikely(skb->len > priv->mcast_mtu + IPOIB_ENCAP_LEN)) { + ipoib_warn(priv, "packet len %d (> %d) too long to send, dropping\n", + skb->len, priv->mcast_mtu + IPOIB_ENCAP_LEN); + ++dev->stats.tx_dropped; + ++dev->stats.tx_errors; + ipoib_cm_skb_too_long(dev, skb, priv->mcast_mtu); + return; + } + phead = NULL; + hlen = 0; } ipoib_dbg_data(priv, "sending packet, length=%d address=%p qpn=0x%06x\n", @@ -442,10 +493,13 @@ void ipoib_send(struct net_device *dev, struct sk_buff *skb, return; } + if (skb->ip_summed == CHECKSUM_PARTIAL) + priv->tx_wr.send_flags |= IB_SEND_IP_CSUM; + else + priv->tx_wr.send_flags &= ~IB_SEND_IP_CSUM; + if (unlikely(post_send(priv, priv->tx_head & (ipoib_sendq_size - 1), - address->ah, qpn, - tx_req->mapping, skb_headlen(skb), - skb_shinfo(skb)->frags, skb_shinfo(skb)->nr_frags))) { + address->ah, qpn, tx_req, phead, hlen))) { ipoib_warn(priv, "post_send failed\n"); ++dev->stats.tx_errors; ipoib_dma_unmap_tx(priv->ca, tx_req); @@ -540,7 +594,7 @@ static void ipoib_pkey_dev_check_presence(struct net_device *dev) struct ipoib_dev_priv *priv = netdev_priv(dev); u16 pkey_index = 0; - if (ib_find_cached_pkey(priv->ca, priv->port, priv->pkey, &pkey_index)) + if (ib_find_pkey(priv->ca, priv->port, priv->pkey, &pkey_index)) clear_bit(IPOIB_PKEY_ASSIGNED, &priv->flags); else set_bit(IPOIB_PKEY_ASSIGNED, &priv->flags); @@ -781,13 +835,13 @@ static void __ipoib_ib_dev_flush(struct ipoib_dev_priv *priv, int pkey_event) clear_bit(IPOIB_PKEY_ASSIGNED, &priv->flags); ipoib_ib_dev_down(dev, 0); ipoib_ib_dev_stop(dev, 0); - ipoib_pkey_dev_delay_open(dev); - return; + if (ipoib_pkey_dev_delay_open(dev)) + return; } - set_bit(IPOIB_PKEY_ASSIGNED, &priv->flags); /* restart QP only if P_Key index is changed */ - if (new_index == priv->pkey_index) { + if (test_and_set_bit(IPOIB_PKEY_ASSIGNED, &priv->flags) && + new_index == priv->pkey_index) { ipoib_dbg(priv, "Not flushing - P_Key index not changed.\n"); return; } diff --git a/drivers/infiniband/ulp/ipoib/ipoib_main.c b/drivers/infiniband/ulp/ipoib/ipoib_main.c index 57282048865c..bd07f02cf02b 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_main.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c @@ -359,8 +359,7 @@ void ipoib_flush_paths(struct net_device *dev) spin_lock_irq(&priv->tx_lock); spin_lock(&priv->lock); - list_splice(&priv->path_list, &remove_list); - INIT_LIST_HEAD(&priv->path_list); + list_splice_init(&priv->path_list, &remove_list); list_for_each_entry(path, &remove_list, list) rb_erase(&path->rb_node, &priv->path_tree); @@ -952,6 +951,8 @@ static void ipoib_setup(struct net_device *dev) dev->set_multicast_list = ipoib_set_mcast_list; dev->neigh_setup = ipoib_neigh_setup_dev; + ipoib_set_ethtool_ops(dev); + netif_napi_add(dev, &priv->napi, ipoib_poll, 100); dev->watchdog_timeo = HZ; @@ -1105,6 +1106,7 @@ static struct net_device *ipoib_add_port(const char *format, struct ib_device *hca, u8 port) { struct ipoib_dev_priv *priv; + struct ib_device_attr *device_attr; int result = -ENOMEM; priv = ipoib_intf_alloc(format); @@ -1120,6 +1122,29 @@ static struct net_device *ipoib_add_port(const char *format, goto device_init_failed; } + device_attr = kmalloc(sizeof *device_attr, GFP_KERNEL); + if (!device_attr) { + printk(KERN_WARNING "%s: allocation of %zu bytes failed\n", + hca->name, sizeof *device_attr); + goto device_init_failed; + } + + result = ib_query_device(hca, device_attr); + if (result) { + printk(KERN_WARNING "%s: ib_query_device failed (ret = %d)\n", + hca->name, result); + kfree(device_attr); + goto device_init_failed; + } + priv->hca_caps = device_attr->device_cap_flags; + + kfree(device_attr); + + if (priv->hca_caps & IB_DEVICE_UD_IP_CSUM) { + set_bit(IPOIB_FLAG_CSUM, &priv->flags); + priv->dev->features |= NETIF_F_SG | NETIF_F_IP_CSUM; + } + /* * Set the full membership bit, so that we join the right * broadcast group, etc. @@ -1137,7 +1162,6 @@ static struct net_device *ipoib_add_port(const char *format, } else memcpy(priv->dev->dev_addr + 4, priv->local_gid.raw, sizeof (union ib_gid)); - result = ipoib_dev_init(priv->dev, hca, port); if (result < 0) { printk(KERN_WARNING "%s: failed to initialize port %d (ret = %d)\n", @@ -1155,6 +1179,9 @@ static struct net_device *ipoib_add_port(const char *format, goto event_failed; } + if (priv->dev->features & NETIF_F_SG && priv->hca_caps & IB_DEVICE_UD_TSO) + priv->dev->features |= NETIF_F_TSO; + result = register_netdev(priv->dev); if (result) { printk(KERN_WARNING "%s: couldn't register ipoib port %d; error %d\n", diff --git a/drivers/infiniband/ulp/ipoib/ipoib_verbs.c b/drivers/infiniband/ulp/ipoib/ipoib_verbs.c index a3aeb911f024..8a20e3742c43 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_verbs.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_verbs.c @@ -192,6 +192,9 @@ int ipoib_transport_dev_init(struct net_device *dev, struct ib_device *ca) init_attr.send_cq = priv->cq; init_attr.recv_cq = priv->cq; + if (priv->hca_caps & IB_DEVICE_UD_TSO) + init_attr.create_flags = IB_QP_CREATE_IPOIB_UD_LSO; + if (dev->features & NETIF_F_SG) init_attr.cap.max_send_sge = MAX_SKB_FRAGS + 1; diff --git a/drivers/infiniband/ulp/iser/iser_initiator.c b/drivers/infiniband/ulp/iser/iser_initiator.c index 83247f1fdf72..08dc81c46f41 100644 --- a/drivers/infiniband/ulp/iser/iser_initiator.c +++ b/drivers/infiniband/ulp/iser/iser_initiator.c @@ -405,7 +405,7 @@ int iser_send_data_out(struct iscsi_conn *conn, struct iser_dto *send_dto = NULL; unsigned long buf_offset; unsigned long data_seg_len; - unsigned int itt; + uint32_t itt; int err = 0; if (!iser_conn_state_comp(iser_conn->ib_conn, ISER_CONN_UP)) { @@ -416,7 +416,7 @@ int iser_send_data_out(struct iscsi_conn *conn, if (iser_check_xmit(conn, ctask)) return -ENOBUFS; - itt = ntohl(hdr->itt); + itt = (__force uint32_t)hdr->itt; data_seg_len = ntoh24(hdr->dlength); buf_offset = ntohl(hdr->offset); diff --git a/drivers/infiniband/ulp/iser/iser_verbs.c b/drivers/infiniband/ulp/iser/iser_verbs.c index 993f0a8ff28f..d19cfe605ebb 100644 --- a/drivers/infiniband/ulp/iser/iser_verbs.c +++ b/drivers/infiniband/ulp/iser/iser_verbs.c @@ -473,11 +473,8 @@ static int iser_cma_handler(struct rdma_cm_id *cma_id, struct rdma_cm_event *eve iser_connect_error(cma_id); break; case RDMA_CM_EVENT_DISCONNECTED: - iser_disconnected_handler(cma_id); - break; case RDMA_CM_EVENT_DEVICE_REMOVAL: - iser_err("Device removal is currently unsupported\n"); - BUG(); + iser_disconnected_handler(cma_id); break; default: iser_err("Unexpected RDMA CM event (%d)\n", event->event); diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c index fd4a49fc4773..125765aa9d59 100644 --- a/drivers/infiniband/ulp/srp/ib_srp.c +++ b/drivers/infiniband/ulp/srp/ib_srp.c @@ -68,7 +68,7 @@ static int srp_max_iu_len; module_param(srp_sg_tablesize, int, 0444); MODULE_PARM_DESC(srp_sg_tablesize, - "Max number of gather/scatter entries per I/O (default is 12)"); + "Max number of gather/scatter entries per I/O (default is 12, max 255)"); static int topspin_workarounds = 1; @@ -2138,6 +2138,11 @@ static int __init srp_init_module(void) { int ret; + if (srp_sg_tablesize > 255) { + printk(KERN_WARNING PFX "Clamping srp_sg_tablesize to 255\n"); + srp_sg_tablesize = 255; + } + ib_srp_transport_template = srp_attach_transport(&ib_srp_transport_functions); if (!ib_srp_transport_template) diff --git a/drivers/net/mlx4/catas.c b/drivers/net/mlx4/catas.c index 6b32ec94b3a8..aa9528779044 100644 --- a/drivers/net/mlx4/catas.c +++ b/drivers/net/mlx4/catas.c @@ -69,7 +69,7 @@ static void poll_catas(unsigned long dev_ptr) if (readl(priv->catas_err.map)) { dump_err_buf(dev); - mlx4_dispatch_event(dev, MLX4_EVENT_TYPE_LOCAL_CATAS_ERROR, 0, 0); + mlx4_dispatch_event(dev, MLX4_DEV_EVENT_CATASTROPHIC_ERROR, 0); if (internal_err_reset) { spin_lock(&catas_lock); diff --git a/drivers/net/mlx4/cmd.c b/drivers/net/mlx4/cmd.c index db49051b97b1..70dff94a8bc6 100644 --- a/drivers/net/mlx4/cmd.c +++ b/drivers/net/mlx4/cmd.c @@ -106,7 +106,8 @@ struct mlx4_cmd_context { u16 token; }; -static int mlx4_status_to_errno(u8 status) { +static int mlx4_status_to_errno(u8 status) +{ static const int trans_table[] = { [CMD_STAT_INTERNAL_ERR] = -EIO, [CMD_STAT_BAD_OP] = -EPERM, diff --git a/drivers/net/mlx4/cq.c b/drivers/net/mlx4/cq.c index d4441fee3d80..caa5bcf54e35 100644 --- a/drivers/net/mlx4/cq.c +++ b/drivers/net/mlx4/cq.c @@ -38,6 +38,7 @@ #include <linux/hardirq.h> #include <linux/mlx4/cmd.h> +#include <linux/mlx4/cq.h> #include "mlx4.h" #include "icm.h" @@ -47,21 +48,19 @@ struct mlx4_cq_context { u16 reserved1[3]; __be16 page_offset; __be32 logsize_usrpage; - u8 reserved2; - u8 cq_period; - u8 reserved3; - u8 cq_max_count; - u8 reserved4[3]; + __be16 cq_period; + __be16 cq_max_count; + u8 reserved2[3]; u8 comp_eqn; u8 log_page_size; - u8 reserved5[2]; + u8 reserved3[2]; u8 mtt_base_addr_h; __be32 mtt_base_addr_l; __be32 last_notified_index; __be32 solicit_producer_index; __be32 consumer_index; __be32 producer_index; - u32 reserved6[2]; + u32 reserved4[2]; __be64 db_rec_addr; }; @@ -121,6 +120,13 @@ static int mlx4_SW2HW_CQ(struct mlx4_dev *dev, struct mlx4_cmd_mailbox *mailbox, MLX4_CMD_TIME_CLASS_A); } +static int mlx4_MODIFY_CQ(struct mlx4_dev *dev, struct mlx4_cmd_mailbox *mailbox, + int cq_num, u32 opmod) +{ + return mlx4_cmd(dev, mailbox->dma, cq_num, opmod, MLX4_CMD_MODIFY_CQ, + MLX4_CMD_TIME_CLASS_A); +} + static int mlx4_HW2SW_CQ(struct mlx4_dev *dev, struct mlx4_cmd_mailbox *mailbox, int cq_num) { @@ -129,6 +135,58 @@ static int mlx4_HW2SW_CQ(struct mlx4_dev *dev, struct mlx4_cmd_mailbox *mailbox, MLX4_CMD_TIME_CLASS_A); } +int mlx4_cq_modify(struct mlx4_dev *dev, struct mlx4_cq *cq, + u16 count, u16 period) +{ + struct mlx4_cmd_mailbox *mailbox; + struct mlx4_cq_context *cq_context; + int err; + + mailbox = mlx4_alloc_cmd_mailbox(dev); + if (IS_ERR(mailbox)) + return PTR_ERR(mailbox); + + cq_context = mailbox->buf; + memset(cq_context, 0, sizeof *cq_context); + + cq_context->cq_max_count = cpu_to_be16(count); + cq_context->cq_period = cpu_to_be16(period); + + err = mlx4_MODIFY_CQ(dev, mailbox, cq->cqn, 1); + + mlx4_free_cmd_mailbox(dev, mailbox); + return err; +} +EXPORT_SYMBOL_GPL(mlx4_cq_modify); + +int mlx4_cq_resize(struct mlx4_dev *dev, struct mlx4_cq *cq, + int entries, struct mlx4_mtt *mtt) +{ + struct mlx4_cmd_mailbox *mailbox; + struct mlx4_cq_context *cq_context; + u64 mtt_addr; + int err; + + mailbox = mlx4_alloc_cmd_mailbox(dev); + if (IS_ERR(mailbox)) + return PTR_ERR(mailbox); + + cq_context = mailbox->buf; + memset(cq_context, 0, sizeof *cq_context); + + cq_context->logsize_usrpage = cpu_to_be32(ilog2(entries) << 24); + cq_context->log_page_size = mtt->page_shift - 12; + mtt_addr = mlx4_mtt_addr(dev, mtt); + cq_context->mtt_base_addr_h = mtt_addr >> 32; + cq_context->mtt_base_addr_l = cpu_to_be32(mtt_addr & 0xffffffff); + + err = mlx4_MODIFY_CQ(dev, mailbox, cq->cqn, 1); + + mlx4_free_cmd_mailbox(dev, mailbox); + return err; +} +EXPORT_SYMBOL_GPL(mlx4_cq_resize); + int mlx4_cq_alloc(struct mlx4_dev *dev, int nent, struct mlx4_mtt *mtt, struct mlx4_uar *uar, u64 db_rec, struct mlx4_cq *cq) { diff --git a/drivers/net/mlx4/eq.c b/drivers/net/mlx4/eq.c index 9c36c2034030..e141a1513f07 100644 --- a/drivers/net/mlx4/eq.c +++ b/drivers/net/mlx4/eq.c @@ -202,7 +202,10 @@ static int mlx4_eq_int(struct mlx4_dev *dev, struct mlx4_eq *eq) break; case MLX4_EVENT_TYPE_PORT_CHANGE: - mlx4_dispatch_event(dev, eqe->type, eqe->subtype, + mlx4_dispatch_event(dev, + eqe->subtype == MLX4_PORT_CHANGE_SUBTYPE_ACTIVE ? + MLX4_DEV_EVENT_PORT_UP : + MLX4_DEV_EVENT_PORT_DOWN, be32_to_cpu(eqe->event.port_change.port) >> 28); break; diff --git a/drivers/net/mlx4/fw.c b/drivers/net/mlx4/fw.c index 61dc4951d6b0..d82f2751d2c7 100644 --- a/drivers/net/mlx4/fw.c +++ b/drivers/net/mlx4/fw.c @@ -133,6 +133,7 @@ int mlx4_QUERY_DEV_CAP(struct mlx4_dev *dev, struct mlx4_dev_cap *dev_cap) #define QUERY_DEV_CAP_MAX_AV_OFFSET 0x27 #define QUERY_DEV_CAP_MAX_REQ_QP_OFFSET 0x29 #define QUERY_DEV_CAP_MAX_RES_QP_OFFSET 0x2b +#define QUERY_DEV_CAP_MAX_GSO_OFFSET 0x2d #define QUERY_DEV_CAP_MAX_RDMA_OFFSET 0x2f #define QUERY_DEV_CAP_RSZ_SRQ_OFFSET 0x33 #define QUERY_DEV_CAP_ACK_DELAY_OFFSET 0x35 @@ -215,6 +216,13 @@ int mlx4_QUERY_DEV_CAP(struct mlx4_dev *dev, struct mlx4_dev_cap *dev_cap) dev_cap->max_requester_per_qp = 1 << (field & 0x3f); MLX4_GET(field, outbox, QUERY_DEV_CAP_MAX_RES_QP_OFFSET); dev_cap->max_responder_per_qp = 1 << (field & 0x3f); + MLX4_GET(field, outbox, QUERY_DEV_CAP_MAX_GSO_OFFSET); + field &= 0x1f; + if (!field) + dev_cap->max_gso_sz = 0; + else + dev_cap->max_gso_sz = 1 << field; + MLX4_GET(field, outbox, QUERY_DEV_CAP_MAX_RDMA_OFFSET); dev_cap->max_rdma_global = 1 << (field & 0x3f); MLX4_GET(field, outbox, QUERY_DEV_CAP_ACK_DELAY_OFFSET); @@ -377,6 +385,7 @@ int mlx4_QUERY_DEV_CAP(struct mlx4_dev *dev, struct mlx4_dev_cap *dev_cap) dev_cap->max_sq_desc_sz, dev_cap->max_sq_sg); mlx4_dbg(dev, "Max RQ desc size: %d, max RQ S/G: %d\n", dev_cap->max_rq_desc_sz, dev_cap->max_rq_sg); + mlx4_dbg(dev, "Max GSO size: %d\n", dev_cap->max_gso_sz); dump_dev_cap_flags(dev, dev_cap->flags); @@ -696,6 +705,10 @@ int mlx4_INIT_HCA(struct mlx4_dev *dev, struct mlx4_init_hca_param *param) /* Check port for UD address vector: */ *(inbox + INIT_HCA_FLAGS_OFFSET / 4) |= cpu_to_be32(1); + /* Enable IPoIB checksumming if we can: */ + if (dev->caps.flags & MLX4_DEV_CAP_FLAG_IPOIB_CSUM) + *(inbox + INIT_HCA_FLAGS_OFFSET / 4) |= cpu_to_be32(1 << 3); + /* QPC/EEC/CQC/EQC/RDMARC attributes */ MLX4_PUT(inbox, param->qpc_base, INIT_HCA_QPC_BASE_OFFSET); diff --git a/drivers/net/mlx4/fw.h b/drivers/net/mlx4/fw.h index e16dec890413..306cb9b0242d 100644 --- a/drivers/net/mlx4/fw.h +++ b/drivers/net/mlx4/fw.h @@ -96,6 +96,7 @@ struct mlx4_dev_cap { u8 bmme_flags; u32 reserved_lkey; u64 max_icm_sz; + int max_gso_sz; }; struct mlx4_adapter { diff --git a/drivers/net/mlx4/intf.c b/drivers/net/mlx4/intf.c index be5d9e90ccf2..4a6c4d526f1b 100644 --- a/drivers/net/mlx4/intf.c +++ b/drivers/net/mlx4/intf.c @@ -30,8 +30,6 @@ * SOFTWARE. */ -#include <linux/mlx4/driver.h> - #include "mlx4.h" struct mlx4_device_context { @@ -113,8 +111,7 @@ void mlx4_unregister_interface(struct mlx4_interface *intf) } EXPORT_SYMBOL_GPL(mlx4_unregister_interface); -void mlx4_dispatch_event(struct mlx4_dev *dev, enum mlx4_event type, - int subtype, int port) +void mlx4_dispatch_event(struct mlx4_dev *dev, enum mlx4_dev_event type, int port) { struct mlx4_priv *priv = mlx4_priv(dev); struct mlx4_device_context *dev_ctx; @@ -124,8 +121,7 @@ void mlx4_dispatch_event(struct mlx4_dev *dev, enum mlx4_event type, list_for_each_entry(dev_ctx, &priv->ctx_list, list) if (dev_ctx->intf->event) - dev_ctx->intf->event(dev, dev_ctx->context, type, - subtype, port); + dev_ctx->intf->event(dev, dev_ctx->context, type, port); spin_unlock_irqrestore(&priv->ctx_lock, flags); } diff --git a/drivers/net/mlx4/main.c b/drivers/net/mlx4/main.c index 08bfc130a33e..49a4acab5e82 100644 --- a/drivers/net/mlx4/main.c +++ b/drivers/net/mlx4/main.c @@ -76,7 +76,7 @@ static char mlx4_version[] __devinitdata = DRV_VERSION " (" DRV_RELDATE ")\n"; static struct mlx4_profile default_profile = { - .num_qp = 1 << 16, + .num_qp = 1 << 17, .num_srq = 1 << 16, .rdmarc_per_qp = 1 << 4, .num_cq = 1 << 16, @@ -159,6 +159,7 @@ static int mlx4_dev_cap(struct mlx4_dev *dev, struct mlx4_dev_cap *dev_cap) dev->caps.page_size_cap = ~(u32) (dev_cap->min_page_sz - 1); dev->caps.flags = dev_cap->flags; dev->caps.stat_rate_support = dev_cap->stat_rate_support; + dev->caps.max_gso_sz = dev_cap->max_gso_sz; return 0; } @@ -735,8 +736,7 @@ static int __mlx4_init_one(struct pci_dev *pdev, const struct pci_device_id *id) } /* - * Check for BARs. We expect 0: 1MB, 2: 8MB, 4: DDR (may not - * be present) + * Check for BARs. We expect 0: 1MB */ if (!(pci_resource_flags(pdev, 0) & IORESOURCE_MEM) || pci_resource_len(pdev, 0) != 1 << 20) { diff --git a/drivers/net/mlx4/mcg.c b/drivers/net/mlx4/mcg.c index a99e7729d333..57f7f1f0d4ec 100644 --- a/drivers/net/mlx4/mcg.c +++ b/drivers/net/mlx4/mcg.c @@ -190,10 +190,6 @@ int mlx4_multicast_attach(struct mlx4_dev *dev, struct mlx4_qp *qp, u8 gid[16]) } index += dev->caps.num_mgms; - err = mlx4_READ_MCG(dev, index, mailbox); - if (err) - goto out; - memset(mgm, 0, sizeof *mgm); memcpy(mgm->gid, gid, 16); } @@ -301,12 +297,10 @@ int mlx4_multicast_detach(struct mlx4_dev *dev, struct mlx4_qp *qp, u8 gid[16]) mgm->qp[loc] = mgm->qp[i - 1]; mgm->qp[i - 1] = 0; - err = mlx4_WRITE_MCG(dev, index, mailbox); - if (err) - goto out; - - if (i != 1) + if (i != 1) { + err = mlx4_WRITE_MCG(dev, index, mailbox); goto out; + } if (prev == -1) { /* Remove entry from MGM */ diff --git a/drivers/net/mlx4/mlx4.h b/drivers/net/mlx4/mlx4.h index 53a1cdddfc13..73336810e652 100644 --- a/drivers/net/mlx4/mlx4.h +++ b/drivers/net/mlx4/mlx4.h @@ -42,6 +42,7 @@ #include <linux/timer.h> #include <linux/mlx4/device.h> +#include <linux/mlx4/driver.h> #include <linux/mlx4/doorbell.h> #define DRV_NAME "mlx4_core" @@ -313,8 +314,7 @@ void mlx4_catas_cleanup(void); int mlx4_restart_one(struct pci_dev *pdev); int mlx4_register_device(struct mlx4_dev *dev); void mlx4_unregister_device(struct mlx4_dev *dev); -void mlx4_dispatch_event(struct mlx4_dev *dev, enum mlx4_event type, - int subtype, int port); +void mlx4_dispatch_event(struct mlx4_dev *dev, enum mlx4_dev_event type, int port); struct mlx4_dev_cap; struct mlx4_init_hca_param; diff --git a/drivers/s390/block/Kconfig b/drivers/s390/block/Kconfig index e879b212cf43..07883197f474 100644 --- a/drivers/s390/block/Kconfig +++ b/drivers/s390/block/Kconfig @@ -20,6 +20,7 @@ config DCSSBLK config DASD tristate "Support for DASD devices" depends on CCW && BLOCK + select IOSCHED_DEADLINE help Enable this option if you want to access DASDs directly utilizing S/390s channel subsystem commands. This is necessary for running diff --git a/drivers/s390/block/dasd.c b/drivers/s390/block/dasd.c index ccf46c96adb4..ac6d4d3218b3 100644 --- a/drivers/s390/block/dasd.c +++ b/drivers/s390/block/dasd.c @@ -980,12 +980,12 @@ void dasd_int_handler(struct ccw_device *cdev, unsigned long intparm, break; case -ETIMEDOUT: printk(KERN_WARNING"%s(%s): request timed out\n", - __FUNCTION__, cdev->dev.bus_id); + __func__, cdev->dev.bus_id); //FIXME - dasd uses own timeout interface... break; default: printk(KERN_WARNING"%s(%s): unknown error %ld\n", - __FUNCTION__, cdev->dev.bus_id, PTR_ERR(irb)); + __func__, cdev->dev.bus_id, PTR_ERR(irb)); } return; } @@ -1956,6 +1956,7 @@ static int dasd_alloc_queue(struct dasd_block *block) block->request_queue->queuedata = block; elevator_exit(block->request_queue->elevator); + block->request_queue->elevator = NULL; rc = elevator_init(block->request_queue, "deadline"); if (rc) { blk_cleanup_queue(block->request_queue); @@ -2298,9 +2299,8 @@ int dasd_generic_set_offline(struct ccw_device *cdev) * in the other openers. */ if (device->block) { - struct dasd_block *block = device->block; - max_count = block->bdev ? 0 : -1; - open_count = (int) atomic_read(&block->open_count); + max_count = device->block->bdev ? 0 : -1; + open_count = atomic_read(&device->block->open_count); if (open_count > max_count) { if (open_count > 0) printk(KERN_WARNING "Can't offline dasd " diff --git a/drivers/s390/block/dasd_3990_erp.c b/drivers/s390/block/dasd_3990_erp.c index b19db20a0bef..e6700df52df4 100644 --- a/drivers/s390/block/dasd_3990_erp.c +++ b/drivers/s390/block/dasd_3990_erp.c @@ -1996,6 +1996,36 @@ dasd_3990_erp_compound(struct dasd_ccw_req * erp, char *sense) } /* end dasd_3990_erp_compound */ /* + *DASD_3990_ERP_HANDLE_SIM + * + *DESCRIPTION + * inspects the SIM SENSE data and starts an appropriate action + * + * PARAMETER + * sense sense data of the actual error + * + * RETURN VALUES + * none + */ +void +dasd_3990_erp_handle_sim(struct dasd_device *device, char *sense) +{ + /* print message according to log or message to operator mode */ + if ((sense[24] & DASD_SIM_MSG_TO_OP) || (sense[1] & 0x10)) { + + /* print SIM SRC from RefCode */ + DEV_MESSAGE(KERN_ERR, device, "SIM - SRC: " + "%02x%02x%02x%02x", sense[22], + sense[23], sense[11], sense[12]); + } else if (sense[24] & DASD_SIM_LOG) { + /* print SIM SRC Refcode */ + DEV_MESSAGE(KERN_WARNING, device, "SIM - SRC: " + "%02x%02x%02x%02x", sense[22], + sense[23], sense[11], sense[12]); + } +} + +/* * DASD_3990_ERP_INSPECT_32 * * DESCRIPTION @@ -2018,6 +2048,10 @@ dasd_3990_erp_inspect_32(struct dasd_ccw_req * erp, char *sense) erp->function = dasd_3990_erp_inspect_32; + /* check for SIM sense data */ + if ((sense[6] & DASD_SIM_SENSE) == DASD_SIM_SENSE) + dasd_3990_erp_handle_sim(device, sense); + if (sense[25] & DASD_SENSE_BIT_0) { /* compound program action codes (byte25 bit 0 == '1') */ diff --git a/drivers/s390/block/dasd_alias.c b/drivers/s390/block/dasd_alias.c index 3a40bee9d358..2d8df0b30538 100644 --- a/drivers/s390/block/dasd_alias.c +++ b/drivers/s390/block/dasd_alias.c @@ -745,6 +745,19 @@ static void flush_all_alias_devices_on_lcu(struct alias_lcu *lcu) spin_unlock_irqrestore(&lcu->lock, flags); } +static void __stop_device_on_lcu(struct dasd_device *device, + struct dasd_device *pos) +{ + /* If pos == device then device is already locked! */ + if (pos == device) { + pos->stopped |= DASD_STOPPED_SU; + return; + } + spin_lock(get_ccwdev_lock(pos->cdev)); + pos->stopped |= DASD_STOPPED_SU; + spin_unlock(get_ccwdev_lock(pos->cdev)); +} + /* * This function is called in interrupt context, so the * cdev lock for device is already locked! @@ -755,35 +768,15 @@ static void _stop_all_devices_on_lcu(struct alias_lcu *lcu, struct alias_pav_group *pavgroup; struct dasd_device *pos; - list_for_each_entry(pos, &lcu->active_devices, alias_list) { - if (pos != device) - spin_lock(get_ccwdev_lock(pos->cdev)); - pos->stopped |= DASD_STOPPED_SU; - if (pos != device) - spin_unlock(get_ccwdev_lock(pos->cdev)); - } - list_for_each_entry(pos, &lcu->inactive_devices, alias_list) { - if (pos != device) - spin_lock(get_ccwdev_lock(pos->cdev)); - pos->stopped |= DASD_STOPPED_SU; - if (pos != device) - spin_unlock(get_ccwdev_lock(pos->cdev)); - } + list_for_each_entry(pos, &lcu->active_devices, alias_list) + __stop_device_on_lcu(device, pos); + list_for_each_entry(pos, &lcu->inactive_devices, alias_list) + __stop_device_on_lcu(device, pos); list_for_each_entry(pavgroup, &lcu->grouplist, group) { - list_for_each_entry(pos, &pavgroup->baselist, alias_list) { - if (pos != device) - spin_lock(get_ccwdev_lock(pos->cdev)); - pos->stopped |= DASD_STOPPED_SU; - if (pos != device) - spin_unlock(get_ccwdev_lock(pos->cdev)); - } - list_for_each_entry(pos, &pavgroup->aliaslist, alias_list) { - if (pos != device) - spin_lock(get_ccwdev_lock(pos->cdev)); - pos->stopped |= DASD_STOPPED_SU; - if (pos != device) - spin_unlock(get_ccwdev_lock(pos->cdev)); - } + list_for_each_entry(pos, &pavgroup->baselist, alias_list) + __stop_device_on_lcu(device, pos); + list_for_each_entry(pos, &pavgroup->aliaslist, alias_list) + __stop_device_on_lcu(device, pos); } } diff --git a/drivers/s390/block/dasd_eckd.c b/drivers/s390/block/dasd_eckd.c index 61f16937c1e0..a0edae091b5e 100644 --- a/drivers/s390/block/dasd_eckd.c +++ b/drivers/s390/block/dasd_eckd.c @@ -1415,6 +1415,13 @@ static void dasd_eckd_handle_unsolicited_interrupt(struct dasd_device *device, return; } + + /* service information message SIM */ + if ((irb->ecw[6] & DASD_SIM_SENSE) == DASD_SIM_SENSE) { + dasd_3990_erp_handle_sim(device, irb->ecw); + return; + } + /* just report other unsolicited interrupts */ DEV_MESSAGE(KERN_DEBUG, device, "%s", "unsolicited interrupt received"); diff --git a/drivers/s390/block/dasd_fba.c b/drivers/s390/block/dasd_fba.c index d13ea05089a7..116611583df8 100644 --- a/drivers/s390/block/dasd_fba.c +++ b/drivers/s390/block/dasd_fba.c @@ -125,7 +125,8 @@ dasd_fba_check_characteristics(struct dasd_device *device) private = (struct dasd_fba_private *) device->private; if (private == NULL) { - private = kzalloc(sizeof(struct dasd_fba_private), GFP_KERNEL); + private = kzalloc(sizeof(struct dasd_fba_private), + GFP_KERNEL | GFP_DMA); if (private == NULL) { DEV_MESSAGE(KERN_WARNING, device, "%s", "memory allocation failed for private " diff --git a/drivers/s390/block/dasd_int.h b/drivers/s390/block/dasd_int.h index 44b2984dfbee..6c624bf44617 100644 --- a/drivers/s390/block/dasd_int.h +++ b/drivers/s390/block/dasd_int.h @@ -72,6 +72,11 @@ struct dasd_block; #define DASD_SENSE_BIT_2 0x20 #define DASD_SENSE_BIT_3 0x10 +/* BIT DEFINITIONS FOR SIM SENSE */ +#define DASD_SIM_SENSE 0x0F +#define DASD_SIM_MSG_TO_OP 0x03 +#define DASD_SIM_LOG 0x0C + /* * SECTION: MACROs for klogd and s390 debug feature (dbf) */ @@ -621,6 +626,7 @@ void dasd_log_sense(struct dasd_ccw_req *, struct irb *); /* externals in dasd_3990_erp.c */ struct dasd_ccw_req *dasd_3990_erp_action(struct dasd_ccw_req *); +void dasd_3990_erp_handle_sim(struct dasd_device *, char *); /* externals in dasd_eer.c */ #ifdef CONFIG_DASD_EER diff --git a/drivers/s390/block/dcssblk.c b/drivers/s390/block/dcssblk.c index e6c94dbfdeaa..04787eab1016 100644 --- a/drivers/s390/block/dcssblk.c +++ b/drivers/s390/block/dcssblk.c @@ -142,57 +142,6 @@ dcssblk_get_device_by_name(char *name) return NULL; } -/* - * print appropriate error message for segment_load()/segment_type() - * return code - */ -static void -dcssblk_segment_warn(int rc, char* seg_name) -{ - switch (rc) { - case -ENOENT: - PRINT_WARN("cannot load/query segment %s, does not exist\n", - seg_name); - break; - case -ENOSYS: - PRINT_WARN("cannot load/query segment %s, not running on VM\n", - seg_name); - break; - case -EIO: - PRINT_WARN("cannot load/query segment %s, hardware error\n", - seg_name); - break; - case -ENOTSUPP: - PRINT_WARN("cannot load/query segment %s, is a multi-part " - "segment\n", seg_name); - break; - case -ENOSPC: - PRINT_WARN("cannot load/query segment %s, overlaps with " - "storage\n", seg_name); - break; - case -EBUSY: - PRINT_WARN("cannot load/query segment %s, overlaps with " - "already loaded dcss\n", seg_name); - break; - case -EPERM: - PRINT_WARN("cannot load/query segment %s, already loaded in " - "incompatible mode\n", seg_name); - break; - case -ENOMEM: - PRINT_WARN("cannot load/query segment %s, out of memory\n", - seg_name); - break; - case -ERANGE: - PRINT_WARN("cannot load/query segment %s, exceeds kernel " - "mapping range\n", seg_name); - break; - default: - PRINT_WARN("cannot load/query segment %s, return value %i\n", - seg_name, rc); - break; - } -} - static void dcssblk_unregister_callback(struct device *dev) { device_unregister(dev); @@ -423,7 +372,7 @@ dcssblk_add_store(struct device *dev, struct device_attribute *attr, const char rc = segment_load(local_buf, SEGMENT_SHARED, &dev_info->start, &dev_info->end); if (rc < 0) { - dcssblk_segment_warn(rc, dev_info->segment_name); + segment_warning(rc, dev_info->segment_name); goto dealloc_gendisk; } seg_byte_size = (dev_info->end - dev_info->start + 1); diff --git a/drivers/s390/char/monreader.c b/drivers/s390/char/monreader.c index 67009bfa093e..1e1f50655bbf 100644 --- a/drivers/s390/char/monreader.c +++ b/drivers/s390/char/monreader.c @@ -111,56 +111,6 @@ static void dcss_mkname(char *ascii_name, char *ebcdic_name) ASCEBC(ebcdic_name, 8); } -/* - * print appropriate error message for segment_load()/segment_type() - * return code - */ -static void mon_segment_warn(int rc, char* seg_name) -{ - switch (rc) { - case -ENOENT: - P_WARNING("cannot load/query segment %s, does not exist\n", - seg_name); - break; - case -ENOSYS: - P_WARNING("cannot load/query segment %s, not running on VM\n", - seg_name); - break; - case -EIO: - P_WARNING("cannot load/query segment %s, hardware error\n", - seg_name); - break; - case -ENOTSUPP: - P_WARNING("cannot load/query segment %s, is a multi-part " - "segment\n", seg_name); - break; - case -ENOSPC: - P_WARNING("cannot load/query segment %s, overlaps with " - "storage\n", seg_name); - break; - case -EBUSY: - P_WARNING("cannot load/query segment %s, overlaps with " - "already loaded dcss\n", seg_name); - break; - case -EPERM: - P_WARNING("cannot load/query segment %s, already loaded in " - "incompatible mode\n", seg_name); - break; - case -ENOMEM: - P_WARNING("cannot load/query segment %s, out of memory\n", - seg_name); - break; - case -ERANGE: - P_WARNING("cannot load/query segment %s, exceeds kernel " - "mapping range\n", seg_name); - break; - default: - P_WARNING("cannot load/query segment %s, return value %i\n", - seg_name, rc); - break; - } -} - static inline unsigned long mon_mca_start(struct mon_msg *monmsg) { return *(u32 *) &monmsg->msg.rmmsg; @@ -585,7 +535,7 @@ static int __init mon_init(void) rc = segment_type(mon_dcss_name); if (rc < 0) { - mon_segment_warn(rc, mon_dcss_name); + segment_warning(rc, mon_dcss_name); goto out_iucv; } if (rc != SEG_TYPE_SC) { @@ -598,7 +548,7 @@ static int __init mon_init(void) rc = segment_load(mon_dcss_name, SEGMENT_SHARED, &mon_dcss_start, &mon_dcss_end); if (rc < 0) { - mon_segment_warn(rc, mon_dcss_name); + segment_warning(rc, mon_dcss_name); rc = -EINVAL; goto out_iucv; } diff --git a/drivers/s390/char/sclp_tty.c b/drivers/s390/char/sclp_tty.c index 2e616e33891d..e3b3d390b4a3 100644 --- a/drivers/s390/char/sclp_tty.c +++ b/drivers/s390/char/sclp_tty.c @@ -332,7 +332,7 @@ sclp_tty_write_string(const unsigned char *str, int count) if (sclp_ttybuf == NULL) { while (list_empty(&sclp_tty_pages)) { spin_unlock_irqrestore(&sclp_tty_lock, flags); - if (in_atomic()) + if (in_interrupt()) sclp_sync_wait(); else wait_event(sclp_tty_waitq, diff --git a/drivers/s390/char/sclp_vt220.c b/drivers/s390/char/sclp_vt220.c index f7b258dfd52c..ed507594e62b 100644 --- a/drivers/s390/char/sclp_vt220.c +++ b/drivers/s390/char/sclp_vt220.c @@ -383,7 +383,7 @@ sclp_vt220_timeout(unsigned long data) */ static int __sclp_vt220_write(const unsigned char *buf, int count, int do_schedule, - int convertlf) + int convertlf, int may_schedule) { unsigned long flags; void *page; @@ -398,9 +398,8 @@ __sclp_vt220_write(const unsigned char *buf, int count, int do_schedule, /* Create a sclp output buffer if none exists yet */ if (sclp_vt220_current_request == NULL) { while (list_empty(&sclp_vt220_empty)) { - spin_unlock_irqrestore(&sclp_vt220_lock, - flags); - if (in_atomic()) + spin_unlock_irqrestore(&sclp_vt220_lock, flags); + if (in_interrupt() || !may_schedule) sclp_sync_wait(); else wait_event(sclp_vt220_waitq, @@ -450,7 +449,7 @@ __sclp_vt220_write(const unsigned char *buf, int count, int do_schedule, static int sclp_vt220_write(struct tty_struct *tty, const unsigned char *buf, int count) { - return __sclp_vt220_write(buf, count, 1, 0); + return __sclp_vt220_write(buf, count, 1, 0, 1); } #define SCLP_VT220_SESSION_ENDED 0x01 @@ -529,7 +528,7 @@ sclp_vt220_close(struct tty_struct *tty, struct file *filp) static void sclp_vt220_put_char(struct tty_struct *tty, unsigned char ch) { - __sclp_vt220_write(&ch, 1, 0, 0); + __sclp_vt220_write(&ch, 1, 0, 0, 1); } /* @@ -746,7 +745,7 @@ __initcall(sclp_vt220_tty_init); static void sclp_vt220_con_write(struct console *con, const char *buf, unsigned int count) { - __sclp_vt220_write((const unsigned char *) buf, count, 1, 1); + __sclp_vt220_write((const unsigned char *) buf, count, 1, 1, 0); } static struct tty_driver * diff --git a/drivers/s390/char/tape_34xx.c b/drivers/s390/char/tape_34xx.c index 5b47e9cce75f..874adf365e46 100644 --- a/drivers/s390/char/tape_34xx.c +++ b/drivers/s390/char/tape_34xx.c @@ -394,7 +394,7 @@ tape_34xx_unit_check(struct tape_device *device, struct tape_request *request, return tape_34xx_erp_failed(request, -ENOSPC); default: PRINT_ERR("Invalid op in %s:%i\n", - __FUNCTION__, __LINE__); + __func__, __LINE__); return tape_34xx_erp_failed(request, 0); } } diff --git a/drivers/s390/char/tape_char.c b/drivers/s390/char/tape_char.c index b830a8cbef78..ebe84067bae9 100644 --- a/drivers/s390/char/tape_char.c +++ b/drivers/s390/char/tape_char.c @@ -83,9 +83,9 @@ tapechar_setup_device(struct tape_device * device) void tapechar_cleanup_device(struct tape_device *device) { - unregister_tape_dev(device->rt); + unregister_tape_dev(&device->cdev->dev, device->rt); device->rt = NULL; - unregister_tape_dev(device->nt); + unregister_tape_dev(&device->cdev->dev, device->nt); device->nt = NULL; } diff --git a/drivers/s390/char/tape_class.c b/drivers/s390/char/tape_class.c index aa7f166f4034..6dfdb7c17981 100644 --- a/drivers/s390/char/tape_class.c +++ b/drivers/s390/char/tape_class.c @@ -99,11 +99,10 @@ fail_with_tcd: } EXPORT_SYMBOL(register_tape_dev); -void unregister_tape_dev(struct tape_class_device *tcd) +void unregister_tape_dev(struct device *device, struct tape_class_device *tcd) { if (tcd != NULL && !IS_ERR(tcd)) { - sysfs_remove_link(&tcd->class_device->kobj, - tcd->mode_name); + sysfs_remove_link(&device->kobj, tcd->mode_name); device_destroy(tape_class, tcd->char_device->dev); cdev_del(tcd->char_device); kfree(tcd); diff --git a/drivers/s390/char/tape_class.h b/drivers/s390/char/tape_class.h index e2b5ac918acf..707b7f48c232 100644 --- a/drivers/s390/char/tape_class.h +++ b/drivers/s390/char/tape_class.h @@ -56,6 +56,6 @@ struct tape_class_device *register_tape_dev( char * device_name, char * node_name ); -void unregister_tape_dev(struct tape_class_device *tcd); +void unregister_tape_dev(struct device *device, struct tape_class_device *tcd); #endif /* __TAPE_CLASS_H__ */ diff --git a/drivers/s390/char/vmur.c b/drivers/s390/char/vmur.c index 7689b500a104..83ae9a852f00 100644 --- a/drivers/s390/char/vmur.c +++ b/drivers/s390/char/vmur.c @@ -100,7 +100,8 @@ static struct urdev *urdev_alloc(struct ccw_device *cdev) urd->reclen = cdev->id.driver_info; ccw_device_get_id(cdev, &urd->dev_id); mutex_init(&urd->io_mutex); - mutex_init(&urd->open_mutex); + init_waitqueue_head(&urd->wait); + spin_lock_init(&urd->open_lock); atomic_set(&urd->ref_count, 1); urd->cdev = cdev; get_device(&cdev->dev); @@ -678,17 +679,21 @@ static int ur_open(struct inode *inode, struct file *file) if (!urd) return -ENXIO; - if (file->f_flags & O_NONBLOCK) { - if (!mutex_trylock(&urd->open_mutex)) { + spin_lock(&urd->open_lock); + while (urd->open_flag) { + spin_unlock(&urd->open_lock); + if (file->f_flags & O_NONBLOCK) { rc = -EBUSY; goto fail_put; } - } else { - if (mutex_lock_interruptible(&urd->open_mutex)) { + if (wait_event_interruptible(urd->wait, urd->open_flag == 0)) { rc = -ERESTARTSYS; goto fail_put; } + spin_lock(&urd->open_lock); } + urd->open_flag++; + spin_unlock(&urd->open_lock); TRACE("ur_open\n"); @@ -720,7 +725,9 @@ static int ur_open(struct inode *inode, struct file *file) fail_urfile_free: urfile_free(urf); fail_unlock: - mutex_unlock(&urd->open_mutex); + spin_lock(&urd->open_lock); + urd->open_flag--; + spin_unlock(&urd->open_lock); fail_put: urdev_put(urd); return rc; @@ -731,7 +738,10 @@ static int ur_release(struct inode *inode, struct file *file) struct urfile *urf = file->private_data; TRACE("ur_release\n"); - mutex_unlock(&urf->urd->open_mutex); + spin_lock(&urf->urd->open_lock); + urf->urd->open_flag--; + spin_unlock(&urf->urd->open_lock); + wake_up_interruptible(&urf->urd->wait); urdev_put(urf->urd); urfile_free(urf); return 0; diff --git a/drivers/s390/char/vmur.h b/drivers/s390/char/vmur.h index fa959644735a..fa320ad4593d 100644 --- a/drivers/s390/char/vmur.h +++ b/drivers/s390/char/vmur.h @@ -62,7 +62,6 @@ struct file_control_block { struct urdev { struct ccw_device *cdev; /* Backpointer to ccw device */ struct mutex io_mutex; /* Serialises device IO */ - struct mutex open_mutex; /* Serialises access to device */ struct completion *io_done; /* do_ur_io waits; irq completes */ struct device *device; struct cdev *char_device; @@ -71,6 +70,9 @@ struct urdev { int class; /* VM device class */ int io_request_rc; /* return code from I/O request */ atomic_t ref_count; /* reference counter */ + wait_queue_head_t wait; /* wait queue to serialize open */ + int open_flag; /* "urdev is open" flag */ + spinlock_t open_lock; /* serialize critical sections */ }; /* diff --git a/drivers/s390/char/vmwatchdog.c b/drivers/s390/char/vmwatchdog.c index 6f40facb1c4d..19f8389291b6 100644 --- a/drivers/s390/char/vmwatchdog.c +++ b/drivers/s390/char/vmwatchdog.c @@ -96,7 +96,7 @@ static int vmwdt_keepalive(void) if (ret) { printk(KERN_WARNING "%s: problem setting interval %d, " - "cmd %s\n", __FUNCTION__, vmwdt_interval, + "cmd %s\n", __func__, vmwdt_interval, vmwdt_cmd); } return ret; @@ -107,7 +107,7 @@ static int vmwdt_disable(void) int ret = __diag288(wdt_cancel, 0, "", 0); if (ret) { printk(KERN_WARNING "%s: problem disabling watchdog\n", - __FUNCTION__); + __func__); } return ret; } diff --git a/drivers/s390/char/zcore.c b/drivers/s390/char/zcore.c index f523501e6e6c..bbbd14e9d48f 100644 --- a/drivers/s390/char/zcore.c +++ b/drivers/s390/char/zcore.c @@ -224,7 +224,7 @@ static int __init init_cpu_info(enum arch_id arch) sa = kmalloc(sizeof(*sa), GFP_KERNEL); if (!sa) { - ERROR_MSG("kmalloc failed: %s: %i\n",__FUNCTION__, __LINE__); + ERROR_MSG("kmalloc failed: %s: %i\n",__func__, __LINE__); return -ENOMEM; } if (memcpy_hsa_kernel(sa, sys_info.sa_base, sys_info.sa_size) < 0) { diff --git a/drivers/s390/cio/chsc.c b/drivers/s390/cio/chsc.c index 007aaeb4f532..5de86908b0d0 100644 --- a/drivers/s390/cio/chsc.c +++ b/drivers/s390/cio/chsc.c @@ -217,6 +217,8 @@ void chsc_chp_offline(struct chp_id chpid) if (chp_get_status(chpid) <= 0) return; + /* Wait until previous actions have settled. */ + css_wait_for_slow_path(); for_each_subchannel_staged(s390_subchannel_remove_chpid, NULL, &chpid); } @@ -303,7 +305,8 @@ static void s390_process_res_acc (struct res_acc_data *res_data) sprintf(dbf_txt, "fla%x", res_data->fla); CIO_TRACE_EVENT( 2, dbf_txt); } - + /* Wait until previous actions have settled. */ + css_wait_for_slow_path(); /* * I/O resources may have become accessible. * Scan through all subchannels that may be concerned and @@ -561,9 +564,12 @@ void chsc_chp_online(struct chp_id chpid) sprintf(dbf_txt, "cadd%x.%02x", chpid.cssid, chpid.id); CIO_TRACE_EVENT(2, dbf_txt); - if (chp_get_status(chpid) != 0) + if (chp_get_status(chpid) != 0) { + /* Wait until previous actions have settled. */ + css_wait_for_slow_path(); for_each_subchannel_staged(__chp_add, __chp_add_new_sch, &chpid); + } } static void __s390_subchannel_vary_chpid(struct subchannel *sch, @@ -650,6 +656,8 @@ __s390_vary_chpid_on(struct subchannel_id schid, void *data) */ int chsc_chp_vary(struct chp_id chpid, int on) { + /* Wait until previous actions have settled. */ + css_wait_for_slow_path(); /* * Redo PathVerification on the devices the chpid connects to */ @@ -758,7 +766,6 @@ chsc_secm(struct channel_subsystem *css, int enable) if (!secm_area) return -ENOMEM; - mutex_lock(&css->mutex); if (enable && !css->cm_enabled) { css->cub_addr1 = (void *)get_zeroed_page(GFP_KERNEL | GFP_DMA); css->cub_addr2 = (void *)get_zeroed_page(GFP_KERNEL | GFP_DMA); @@ -766,7 +773,6 @@ chsc_secm(struct channel_subsystem *css, int enable) free_page((unsigned long)css->cub_addr1); free_page((unsigned long)css->cub_addr2); free_page((unsigned long)secm_area); - mutex_unlock(&css->mutex); return -ENOMEM; } } @@ -787,7 +793,6 @@ chsc_secm(struct channel_subsystem *css, int enable) free_page((unsigned long)css->cub_addr1); free_page((unsigned long)css->cub_addr2); } - mutex_unlock(&css->mutex); free_page((unsigned long)secm_area); return ret; } diff --git a/drivers/s390/cio/cio.c b/drivers/s390/cio/cio.c index 60590a12d529..23ffcc4768a7 100644 --- a/drivers/s390/cio/cio.c +++ b/drivers/s390/cio/cio.c @@ -24,6 +24,7 @@ #include <asm/ipl.h> #include <asm/chpid.h> #include <asm/airq.h> +#include <asm/cpu.h> #include "cio.h" #include "css.h" #include "chsc.h" @@ -649,13 +650,10 @@ do_IRQ (struct pt_regs *regs) old_regs = set_irq_regs(regs); irq_enter(); - asm volatile ("mc 0,0"); - if (S390_lowcore.int_clock >= S390_lowcore.jiffy_timer) - /** - * Make sure that the i/o interrupt did not "overtake" - * the last HZ timer interrupt. - */ - account_ticks(S390_lowcore.int_clock); + s390_idle_check(); + if (S390_lowcore.int_clock >= S390_lowcore.clock_comparator) + /* Serve timer interrupts first. */ + clock_comparator_work(); /* * Get interrupt information from lowcore */ @@ -672,10 +670,14 @@ do_IRQ (struct pt_regs *regs) continue; } sch = (struct subchannel *)(unsigned long)tpi_info->intparm; - if (sch) - spin_lock(sch->lock); + if (!sch) { + /* Clear pending interrupt condition. */ + tsch(tpi_info->schid, irb); + continue; + } + spin_lock(sch->lock); /* Store interrupt response block to lowcore. */ - if (tsch (tpi_info->schid, irb) == 0 && sch) { + if (tsch(tpi_info->schid, irb) == 0) { /* Keep subchannel information word up to date. */ memcpy (&sch->schib.scsw, &irb->scsw, sizeof (irb->scsw)); @@ -683,8 +685,7 @@ do_IRQ (struct pt_regs *regs) if (sch->driver && sch->driver->irq) sch->driver->irq(sch); } - if (sch) - spin_unlock(sch->lock); + spin_unlock(sch->lock); /* * Are more interrupts pending? * If so, the tpi instruction will update the lowcore @@ -710,8 +711,9 @@ void *cio_get_console_priv(void) /* * busy wait for the next interrupt on the console */ -void -wait_cons_dev (void) +void wait_cons_dev(void) + __releases(console_subchannel.lock) + __acquires(console_subchannel.lock) { unsigned long cr6 __attribute__ ((aligned (8))); unsigned long save_cr6 __attribute__ ((aligned (8))); diff --git a/drivers/s390/cio/cio.h b/drivers/s390/cio/cio.h index 52afa4c784de..08f2235c5a6f 100644 --- a/drivers/s390/cio/cio.h +++ b/drivers/s390/cio/cio.h @@ -100,6 +100,7 @@ extern int cio_modify (struct subchannel *); int cio_create_sch_lock(struct subchannel *); void do_adapter_IO(void); +void do_IRQ(struct pt_regs *); /* Use with care. */ #ifdef CONFIG_CCW_CONSOLE diff --git a/drivers/s390/cio/css.c b/drivers/s390/cio/css.c index 3b45bbe6cce0..c1afab5f72d6 100644 --- a/drivers/s390/cio/css.c +++ b/drivers/s390/cio/css.c @@ -533,6 +533,12 @@ void css_schedule_eval_all(void) spin_unlock_irqrestore(&slow_subchannel_lock, flags); } +void css_wait_for_slow_path(void) +{ + flush_workqueue(ccw_device_notify_work); + flush_workqueue(slow_path_wq); +} + /* Reprobe subchannel if unregistered. */ static int reprobe_subchannel(struct subchannel_id schid, void *data) { @@ -683,10 +689,14 @@ css_cm_enable_show(struct device *dev, struct device_attribute *attr, char *buf) { struct channel_subsystem *css = to_css(dev); + int ret; if (!css) return 0; - return sprintf(buf, "%x\n", css->cm_enabled); + mutex_lock(&css->mutex); + ret = sprintf(buf, "%x\n", css->cm_enabled); + mutex_unlock(&css->mutex); + return ret; } static ssize_t @@ -696,6 +706,7 @@ css_cm_enable_store(struct device *dev, struct device_attribute *attr, struct channel_subsystem *css = to_css(dev); int ret; + mutex_lock(&css->mutex); switch (buf[0]) { case '0': ret = css->cm_enabled ? chsc_secm(css, 0) : 0; @@ -706,6 +717,7 @@ css_cm_enable_store(struct device *dev, struct device_attribute *attr, default: ret = -EINVAL; } + mutex_unlock(&css->mutex); return ret < 0 ? ret : count; } @@ -752,9 +764,11 @@ static int css_reboot_event(struct notifier_block *this, struct channel_subsystem *css; css = channel_subsystems[i]; + mutex_lock(&css->mutex); if (css->cm_enabled) if (chsc_secm(css, 0)) ret = NOTIFY_BAD; + mutex_unlock(&css->mutex); } return ret; diff --git a/drivers/s390/cio/css.h b/drivers/s390/cio/css.h index b70554523552..e1913518f354 100644 --- a/drivers/s390/cio/css.h +++ b/drivers/s390/cio/css.h @@ -144,6 +144,7 @@ struct schib; int css_sch_is_valid(struct schib *); extern struct workqueue_struct *slow_path_wq; +void css_wait_for_slow_path(void); extern struct attribute_group *subch_attr_groups[]; #endif diff --git a/drivers/s390/cio/device.c b/drivers/s390/cio/device.c index fec004f62bcf..e0c7adb8958e 100644 --- a/drivers/s390/cio/device.c +++ b/drivers/s390/cio/device.c @@ -577,7 +577,6 @@ static DEVICE_ATTR(devtype, 0444, devtype_show, NULL); static DEVICE_ATTR(cutype, 0444, cutype_show, NULL); static DEVICE_ATTR(modalias, 0444, modalias_show, NULL); static DEVICE_ATTR(online, 0644, online_show, online_store); -extern struct device_attribute dev_attr_cmb_enable; static DEVICE_ATTR(availability, 0444, available_show, NULL); static struct attribute * subch_attrs[] = { diff --git a/drivers/s390/cio/device.h b/drivers/s390/cio/device.h index d40a2ffaa000..cb08092be39f 100644 --- a/drivers/s390/cio/device.h +++ b/drivers/s390/cio/device.h @@ -127,4 +127,5 @@ extern struct bus_type ccw_bus_type; void retry_set_schib(struct ccw_device *cdev); void cmf_retry_copy_block(struct ccw_device *); int cmf_reenable(struct ccw_device *); +extern struct device_attribute dev_attr_cmb_enable; #endif diff --git a/drivers/s390/cio/device_ops.c b/drivers/s390/cio/device_ops.c index 49b58eb0fab8..a1718a0aa539 100644 --- a/drivers/s390/cio/device_ops.c +++ b/drivers/s390/cio/device_ops.c @@ -193,8 +193,15 @@ int ccw_device_start_key(struct ccw_device *cdev, struct ccw1 *cpa, return -EACCES; } ret = cio_start_key (sch, cpa, lpm, key); - if (ret == 0) + switch (ret) { + case 0: cdev->private->intparm = intparm; + break; + case -EACCES: + case -ENODEV: + dev_fsm_event(cdev, DEV_EVENT_VERIFY); + break; + } return ret; } diff --git a/drivers/s390/cio/device_status.c b/drivers/s390/cio/device_status.c index ebe0848cfe33..4a38993000f2 100644 --- a/drivers/s390/cio/device_status.c +++ b/drivers/s390/cio/device_status.c @@ -62,7 +62,7 @@ ccw_device_path_notoper(struct ccw_device *cdev) stsch (sch->schid, &sch->schib); CIO_MSG_EVENT(0, "%s(0.%x.%04x) - path(s) %02x are " - "not operational \n", __FUNCTION__, + "not operational \n", __func__, sch->schid.ssid, sch->schid.sch_no, sch->schib.pmcw.pnom); @@ -312,6 +312,7 @@ ccw_device_do_sense(struct ccw_device *cdev, struct irb *irb) { struct subchannel *sch; struct ccw1 *sense_ccw; + int rc; sch = to_subchannel(cdev->dev.parent); @@ -337,7 +338,10 @@ ccw_device_do_sense(struct ccw_device *cdev, struct irb *irb) /* Reset internal retry indication. */ cdev->private->flags.intretry = 0; - return cio_start(sch, sense_ccw, 0xff); + rc = cio_start(sch, sense_ccw, 0xff); + if (rc == -ENODEV || rc == -EACCES) + dev_fsm_event(cdev, DEV_EVENT_VERIFY); + return rc; } /* diff --git a/drivers/s390/cio/qdio.c b/drivers/s390/cio/qdio.c index 2b5bfb7c69e5..c359386708e9 100644 --- a/drivers/s390/cio/qdio.c +++ b/drivers/s390/cio/qdio.c @@ -1399,7 +1399,7 @@ __tiqdio_inbound_processing(struct qdio_q *q, int spare_ind_was_set) * q->dev_st_chg_ind is the indicator, be it shared or not. * only clear it, if indicator is non-shared */ - if (!spare_ind_was_set) + if (q->dev_st_chg_ind != &spare_indicator) tiqdio_clear_summary_bit((__u32*)q->dev_st_chg_ind); if (q->hydra_gives_outbound_pcis) { @@ -2217,9 +2217,78 @@ qdio_synchronize(struct ccw_device *cdev, unsigned int flags, return cc; } +static int +qdio_get_ssqd_information(struct subchannel_id *schid, + struct qdio_chsc_ssqd **ssqd_area) +{ + int result; + + QDIO_DBF_TEXT0(0, setup, "getssqd"); + *ssqd_area = mempool_alloc(qdio_mempool_scssc, GFP_ATOMIC); + if (!ssqd_area) { + QDIO_PRINT_WARN("Could not get memory for chsc on sch x%x.\n", + schid->sch_no); + return -ENOMEM; + } + + (*ssqd_area)->request = (struct chsc_header) { + .length = 0x0010, + .code = 0x0024, + }; + (*ssqd_area)->first_sch = schid->sch_no; + (*ssqd_area)->last_sch = schid->sch_no; + (*ssqd_area)->ssid = schid->ssid; + result = chsc(*ssqd_area); + + if (result) { + QDIO_PRINT_WARN("CHSC returned cc %i on sch 0.%x.%x.\n", + result, schid->ssid, schid->sch_no); + goto out; + } + + if ((*ssqd_area)->response.code != QDIO_CHSC_RESPONSE_CODE_OK) { + QDIO_PRINT_WARN("CHSC response is 0x%x on sch 0.%x.%x.\n", + (*ssqd_area)->response.code, + schid->ssid, schid->sch_no); + goto out; + } + if (!((*ssqd_area)->flags & CHSC_FLAG_QDIO_CAPABILITY) || + !((*ssqd_area)->flags & CHSC_FLAG_VALIDITY) || + ((*ssqd_area)->sch != schid->sch_no)) { + QDIO_PRINT_WARN("huh? problems checking out sch 0.%x.%x... " \ + "using all SIGAs.\n", + schid->ssid, schid->sch_no); + goto out; + } + return 0; +out: + return -EINVAL; +} + +int +qdio_get_ssqd_pct(struct ccw_device *cdev) +{ + struct qdio_chsc_ssqd *ssqd_area; + struct subchannel_id schid; + char dbf_text[15]; + int rc; + int pct = 0; + + QDIO_DBF_TEXT0(0, setup, "getpct"); + schid = ccw_device_get_subchannel_id(cdev); + rc = qdio_get_ssqd_information(&schid, &ssqd_area); + if (!rc) + pct = (int)ssqd_area->pct; + if (rc != -ENOMEM) + mempool_free(ssqd_area, qdio_mempool_scssc); + sprintf(dbf_text, "pct: %d", pct); + QDIO_DBF_TEXT2(0, setup, dbf_text); + return pct; +} +EXPORT_SYMBOL(qdio_get_ssqd_pct); + static void -qdio_check_subchannel_qebsm(struct qdio_irq *irq_ptr, unsigned char qdioac, - unsigned long token) +qdio_check_subchannel_qebsm(struct qdio_irq *irq_ptr, unsigned long token) { struct qdio_q *q; int i; @@ -2227,7 +2296,7 @@ qdio_check_subchannel_qebsm(struct qdio_irq *irq_ptr, unsigned char qdioac, char dbf_text[15]; /*check if QEBSM is disabled */ - if (!(irq_ptr->is_qebsm) || !(qdioac & 0x01)) { + if (!(irq_ptr->is_qebsm) || !(irq_ptr->qdioac & 0x01)) { irq_ptr->is_qebsm = 0; irq_ptr->sch_token = 0; irq_ptr->qib.rflags &= ~QIB_RFLAGS_ENABLE_QEBSM; @@ -2256,102 +2325,27 @@ qdio_check_subchannel_qebsm(struct qdio_irq *irq_ptr, unsigned char qdioac, } static void -qdio_get_ssqd_information(struct qdio_irq *irq_ptr) +qdio_get_ssqd_siga(struct qdio_irq *irq_ptr) { - int result; - unsigned char qdioac; - struct { - struct chsc_header request; - u16 reserved1:10; - u16 ssid:2; - u16 fmt:4; - u16 first_sch; - u16 reserved2; - u16 last_sch; - u32 reserved3; - struct chsc_header response; - u32 reserved4; - u8 flags; - u8 reserved5; - u16 sch; - u8 qfmt; - u8 parm; - u8 qdioac1; - u8 sch_class; - u8 reserved7; - u8 icnt; - u8 reserved8; - u8 ocnt; - u8 reserved9; - u8 mbccnt; - u16 qdioac2; - u64 sch_token; - } *ssqd_area; + int rc; + struct qdio_chsc_ssqd *ssqd_area; QDIO_DBF_TEXT0(0,setup,"getssqd"); - qdioac = 0; - ssqd_area = mempool_alloc(qdio_mempool_scssc, GFP_ATOMIC); - if (!ssqd_area) { - QDIO_PRINT_WARN("Could not get memory for chsc. Using all " \ - "SIGAs for sch x%x.\n", irq_ptr->schid.sch_no); + irq_ptr->qdioac = 0; + rc = qdio_get_ssqd_information(&irq_ptr->schid, &ssqd_area); + if (rc) { + QDIO_PRINT_WARN("using all SIGAs for sch x%x.n", + irq_ptr->schid.sch_no); irq_ptr->qdioac = CHSC_FLAG_SIGA_INPUT_NECESSARY | CHSC_FLAG_SIGA_OUTPUT_NECESSARY | CHSC_FLAG_SIGA_SYNC_NECESSARY; /* all flags set */ irq_ptr->is_qebsm = 0; - irq_ptr->sch_token = 0; - irq_ptr->qib.rflags &= ~QIB_RFLAGS_ENABLE_QEBSM; - return; - } - - ssqd_area->request = (struct chsc_header) { - .length = 0x0010, - .code = 0x0024, - }; - ssqd_area->first_sch = irq_ptr->schid.sch_no; - ssqd_area->last_sch = irq_ptr->schid.sch_no; - ssqd_area->ssid = irq_ptr->schid.ssid; - result = chsc(ssqd_area); - - if (result) { - QDIO_PRINT_WARN("CHSC returned cc %i. Using all " \ - "SIGAs for sch 0.%x.%x.\n", result, - irq_ptr->schid.ssid, irq_ptr->schid.sch_no); - qdioac = CHSC_FLAG_SIGA_INPUT_NECESSARY | - CHSC_FLAG_SIGA_OUTPUT_NECESSARY | - CHSC_FLAG_SIGA_SYNC_NECESSARY; /* all flags set */ - irq_ptr->is_qebsm = 0; - goto out; - } + } else + irq_ptr->qdioac = ssqd_area->qdioac1; - if (ssqd_area->response.code != QDIO_CHSC_RESPONSE_CODE_OK) { - QDIO_PRINT_WARN("response upon checking SIGA needs " \ - "is 0x%x. Using all SIGAs for sch 0.%x.%x.\n", - ssqd_area->response.code, - irq_ptr->schid.ssid, irq_ptr->schid.sch_no); - qdioac = CHSC_FLAG_SIGA_INPUT_NECESSARY | - CHSC_FLAG_SIGA_OUTPUT_NECESSARY | - CHSC_FLAG_SIGA_SYNC_NECESSARY; /* all flags set */ - irq_ptr->is_qebsm = 0; - goto out; - } - if (!(ssqd_area->flags & CHSC_FLAG_QDIO_CAPABILITY) || - !(ssqd_area->flags & CHSC_FLAG_VALIDITY) || - (ssqd_area->sch != irq_ptr->schid.sch_no)) { - QDIO_PRINT_WARN("huh? problems checking out sch 0.%x.%x... " \ - "using all SIGAs.\n", - irq_ptr->schid.ssid, irq_ptr->schid.sch_no); - qdioac = CHSC_FLAG_SIGA_INPUT_NECESSARY | - CHSC_FLAG_SIGA_OUTPUT_NECESSARY | - CHSC_FLAG_SIGA_SYNC_NECESSARY; /* worst case */ - irq_ptr->is_qebsm = 0; - goto out; - } - qdioac = ssqd_area->qdioac1; -out: - qdio_check_subchannel_qebsm(irq_ptr, qdioac, - ssqd_area->sch_token); - mempool_free(ssqd_area, qdio_mempool_scssc); - irq_ptr->qdioac = qdioac; + qdio_check_subchannel_qebsm(irq_ptr, ssqd_area->sch_token); + if (rc != -ENOMEM) + mempool_free(ssqd_area, qdio_mempool_scssc); } static unsigned int @@ -3227,7 +3221,7 @@ qdio_establish(struct qdio_initialize *init_data) return -EIO; } - qdio_get_ssqd_information(irq_ptr); + qdio_get_ssqd_siga(irq_ptr); /* if this gets set once, we're running under VM and can omit SVSes */ if (irq_ptr->qdioac&CHSC_FLAG_SIGA_SYNC_NECESSARY) omit_svs=1; diff --git a/drivers/s390/cio/qdio.h b/drivers/s390/cio/qdio.h index da8a272fd75b..c3df6b2c38b7 100644 --- a/drivers/s390/cio/qdio.h +++ b/drivers/s390/cio/qdio.h @@ -406,6 +406,34 @@ do_clear_global_summary(void) #define CHSC_FLAG_SIGA_SYNC_DONE_ON_THININTS 0x08 #define CHSC_FLAG_SIGA_SYNC_DONE_ON_OUTB_PCIS 0x04 +struct qdio_chsc_ssqd { + struct chsc_header request; + u16 reserved1:10; + u16 ssid:2; + u16 fmt:4; + u16 first_sch; + u16 reserved2; + u16 last_sch; + u32 reserved3; + struct chsc_header response; + u32 reserved4; + u8 flags; + u8 reserved5; + u16 sch; + u8 qfmt; + u8 parm; + u8 qdioac1; + u8 sch_class; + u8 pct; + u8 icnt; + u8 reserved7; + u8 ocnt; + u8 reserved8; + u8 mbccnt; + u16 qdioac2; + u64 sch_token; +}; + struct qdio_perf_stats { #ifdef CONFIG_64BIT atomic64_t tl_runs; diff --git a/drivers/s390/crypto/ap_bus.c b/drivers/s390/crypto/ap_bus.c index 7b0b81901297..a1ab3e3efd11 100644 --- a/drivers/s390/crypto/ap_bus.c +++ b/drivers/s390/crypto/ap_bus.c @@ -45,7 +45,7 @@ static int ap_poll_thread_start(void); static void ap_poll_thread_stop(void); static void ap_request_timeout(unsigned long); -/** +/* * Module description. */ MODULE_AUTHOR("IBM Corporation"); @@ -53,7 +53,7 @@ MODULE_DESCRIPTION("Adjunct Processor Bus driver, " "Copyright 2006 IBM Corporation"); MODULE_LICENSE("GPL"); -/** +/* * Module parameter */ int ap_domain_index = -1; /* Adjunct Processor Domain Index */ @@ -69,7 +69,7 @@ static struct device *ap_root_device = NULL; static DEFINE_SPINLOCK(ap_device_lock); static LIST_HEAD(ap_device_list); -/** +/* * Workqueue & timer for bus rescan. */ static struct workqueue_struct *ap_work_queue; @@ -77,7 +77,7 @@ static struct timer_list ap_config_timer; static int ap_config_time = AP_CONFIG_TIME; static DECLARE_WORK(ap_config_work, ap_scan_bus); -/** +/* * Tasklet & timer for AP request polling. */ static struct timer_list ap_poll_timer = TIMER_INITIALIZER(ap_poll_timeout,0,0); @@ -88,9 +88,9 @@ static struct task_struct *ap_poll_kthread = NULL; static DEFINE_MUTEX(ap_poll_thread_mutex); /** - * Test if ap instructions are available. + * ap_intructions_available() - Test if AP instructions are available. * - * Returns 0 if the ap instructions are installed. + * Returns 0 if the AP instructions are installed. */ static inline int ap_instructions_available(void) { @@ -108,12 +108,12 @@ static inline int ap_instructions_available(void) } /** - * Test adjunct processor queue. - * @qid: the ap queue number - * @queue_depth: pointer to queue depth value - * @device_type: pointer to device type value + * ap_test_queue(): Test adjunct processor queue. + * @qid: The AP queue number + * @queue_depth: Pointer to queue depth value + * @device_type: Pointer to device type value * - * Returns ap queue status structure. + * Returns AP queue status structure. */ static inline struct ap_queue_status ap_test_queue(ap_qid_t qid, int *queue_depth, int *device_type) @@ -130,10 +130,10 @@ ap_test_queue(ap_qid_t qid, int *queue_depth, int *device_type) } /** - * Reset adjunct processor queue. - * @qid: the ap queue number + * ap_reset_queue(): Reset adjunct processor queue. + * @qid: The AP queue number * - * Returns ap queue status structure. + * Returns AP queue status structure. */ static inline struct ap_queue_status ap_reset_queue(ap_qid_t qid) { @@ -148,16 +148,14 @@ static inline struct ap_queue_status ap_reset_queue(ap_qid_t qid) } /** - * Send message to adjunct processor queue. - * @qid: the ap queue number - * @psmid: the program supplied message identifier - * @msg: the message text - * @length: the message length - * - * Returns ap queue status structure. + * __ap_send(): Send message to adjunct processor queue. + * @qid: The AP queue number + * @psmid: The program supplied message identifier + * @msg: The message text + * @length: The message length * + * Returns AP queue status structure. * Condition code 1 on NQAP can't happen because the L bit is 1. - * * Condition code 2 on NQAP also means the send is incomplete, * because a segment boundary was reached. The NQAP is repeated. */ @@ -198,23 +196,20 @@ int ap_send(ap_qid_t qid, unsigned long long psmid, void *msg, size_t length) } EXPORT_SYMBOL(ap_send); -/* - * Receive message from adjunct processor queue. - * @qid: the ap queue number - * @psmid: pointer to program supplied message identifier - * @msg: the message text - * @length: the message length - * - * Returns ap queue status structure. +/** + * __ap_recv(): Receive message from adjunct processor queue. + * @qid: The AP queue number + * @psmid: Pointer to program supplied message identifier + * @msg: The message text + * @length: The message length * + * Returns AP queue status structure. * Condition code 1 on DQAP means the receive has taken place * but only partially. The response is incomplete, hence the * DQAP is repeated. - * * Condition code 2 on DQAP also means the receive is incomplete, * this time because a segment boundary was reached. Again, the * DQAP is repeated. - * * Note that gpr2 is used by the DQAP instruction to keep track of * any 'residual' length, in case the instruction gets interrupted. * Hence it gets zeroed before the instruction. @@ -263,11 +258,12 @@ int ap_recv(ap_qid_t qid, unsigned long long *psmid, void *msg, size_t length) EXPORT_SYMBOL(ap_recv); /** - * Check if an AP queue is available. The test is repeated for - * AP_MAX_RESET times. - * @qid: the ap queue number - * @queue_depth: pointer to queue depth value - * @device_type: pointer to device type value + * ap_query_queue(): Check if an AP queue is available. + * @qid: The AP queue number + * @queue_depth: Pointer to queue depth value + * @device_type: Pointer to device type value + * + * The test is repeated for AP_MAX_RESET times. */ static int ap_query_queue(ap_qid_t qid, int *queue_depth, int *device_type) { @@ -308,8 +304,10 @@ static int ap_query_queue(ap_qid_t qid, int *queue_depth, int *device_type) } /** + * ap_init_queue(): Reset an AP queue. + * @qid: The AP queue number + * * Reset an AP queue and wait for it to become available again. - * @qid: the ap queue number */ static int ap_init_queue(ap_qid_t qid) { @@ -346,7 +344,10 @@ static int ap_init_queue(ap_qid_t qid) } /** - * Arm request timeout if a AP device was idle and a new request is submitted. + * ap_increase_queue_count(): Arm request timeout. + * @ap_dev: Pointer to an AP device. + * + * Arm request timeout if an AP device was idle and a new request is submitted. */ static void ap_increase_queue_count(struct ap_device *ap_dev) { @@ -360,7 +361,10 @@ static void ap_increase_queue_count(struct ap_device *ap_dev) } /** - * AP device is still alive, re-schedule request timeout if there are still + * ap_decrease_queue_count(): Decrease queue count. + * @ap_dev: Pointer to an AP device. + * + * If AP device is still alive, re-schedule request timeout if there are still * pending requests. */ static void ap_decrease_queue_count(struct ap_device *ap_dev) @@ -371,7 +375,7 @@ static void ap_decrease_queue_count(struct ap_device *ap_dev) if (ap_dev->queue_count > 0) mod_timer(&ap_dev->timeout, jiffies + timeout); else - /** + /* * The timeout timer should to be disabled now - since * del_timer_sync() is very expensive, we just tell via the * reset flag to ignore the pending timeout timer. @@ -379,7 +383,7 @@ static void ap_decrease_queue_count(struct ap_device *ap_dev) ap_dev->reset = AP_RESET_IGNORE; } -/** +/* * AP device related attributes. */ static ssize_t ap_hwtype_show(struct device *dev, @@ -433,6 +437,10 @@ static struct attribute_group ap_dev_attr_group = { }; /** + * ap_bus_match() + * @dev: Pointer to device + * @drv: Pointer to device_driver + * * AP bus driver registration/unregistration. */ static int ap_bus_match(struct device *dev, struct device_driver *drv) @@ -441,7 +449,7 @@ static int ap_bus_match(struct device *dev, struct device_driver *drv) struct ap_driver *ap_drv = to_ap_drv(drv); struct ap_device_id *id; - /** + /* * Compare device type of the device with the list of * supported types of the device_driver. */ @@ -455,8 +463,12 @@ static int ap_bus_match(struct device *dev, struct device_driver *drv) } /** - * uevent function for AP devices. It sets up a single environment - * variable DEV_TYPE which contains the hardware device type. + * ap_uevent(): Uevent function for AP devices. + * @dev: Pointer to device + * @env: Pointer to kobj_uevent_env + * + * It sets up a single environment variable DEV_TYPE which contains the + * hardware device type. */ static int ap_uevent (struct device *dev, struct kobj_uevent_env *env) { @@ -500,8 +512,10 @@ static int ap_device_probe(struct device *dev) } /** + * __ap_flush_queue(): Flush requests. + * @ap_dev: Pointer to the AP device + * * Flush all requests from the request/pending queue of an AP device. - * @ap_dev: pointer to the AP device. */ static void __ap_flush_queue(struct ap_device *ap_dev) { @@ -565,7 +579,7 @@ void ap_driver_unregister(struct ap_driver *ap_drv) } EXPORT_SYMBOL(ap_driver_unregister); -/** +/* * AP bus attributes. */ static ssize_t ap_domain_show(struct bus_type *bus, char *buf) @@ -630,14 +644,16 @@ static struct bus_attribute *const ap_bus_attrs[] = { }; /** - * Pick one of the 16 ap domains. + * ap_select_domain(): Select an AP domain. + * + * Pick one of the 16 AP domains. */ static int ap_select_domain(void) { int queue_depth, device_type, count, max_count, best_domain; int rc, i, j; - /** + /* * We want to use a single domain. Either the one specified with * the "domain=" parameter or the domain with the maximum number * of devices. @@ -669,8 +685,10 @@ static int ap_select_domain(void) } /** - * Find the device type if query queue returned a device type of 0. + * ap_probe_device_type(): Find the device type of an AP. * @ap_dev: pointer to the AP device. + * + * Find the device type if query queue returned a device type of 0. */ static int ap_probe_device_type(struct ap_device *ap_dev) { @@ -764,7 +782,11 @@ out: } /** - * Scan the ap bus for new devices. + * __ap_scan_bus(): Scan the AP bus. + * @dev: Pointer to device + * @data: Pointer to data + * + * Scan the AP bus for new devices. */ static int __ap_scan_bus(struct device *dev, void *data) { @@ -867,6 +889,8 @@ ap_config_timeout(unsigned long ptr) } /** + * ap_schedule_poll_timer(): Schedule poll timer. + * * Set up the timer to run the poll tasklet */ static inline void ap_schedule_poll_timer(void) @@ -877,10 +901,11 @@ static inline void ap_schedule_poll_timer(void) } /** - * Receive pending reply messages from an AP device. + * ap_poll_read(): Receive pending reply messages from an AP device. * @ap_dev: pointer to the AP device * @flags: pointer to control flags, bit 2^0 is set if another poll is * required, bit 2^1 is set if the poll timer needs to get armed + * * Returns 0 if the device is still present, -ENODEV if not. */ static int ap_poll_read(struct ap_device *ap_dev, unsigned long *flags) @@ -925,10 +950,11 @@ static int ap_poll_read(struct ap_device *ap_dev, unsigned long *flags) } /** - * Send messages from the request queue to an AP device. + * ap_poll_write(): Send messages from the request queue to an AP device. * @ap_dev: pointer to the AP device * @flags: pointer to control flags, bit 2^0 is set if another poll is * required, bit 2^1 is set if the poll timer needs to get armed + * * Returns 0 if the device is still present, -ENODEV if not. */ static int ap_poll_write(struct ap_device *ap_dev, unsigned long *flags) @@ -968,11 +994,13 @@ static int ap_poll_write(struct ap_device *ap_dev, unsigned long *flags) } /** - * Poll AP device for pending replies and send new messages. If either - * ap_poll_read or ap_poll_write returns -ENODEV unregister the device. + * ap_poll_queue(): Poll AP device for pending replies and send new messages. * @ap_dev: pointer to the bus device * @flags: pointer to control flags, bit 2^0 is set if another poll is * required, bit 2^1 is set if the poll timer needs to get armed + * + * Poll AP device for pending replies and send new messages. If either + * ap_poll_read or ap_poll_write returns -ENODEV unregister the device. * Returns 0. */ static inline int ap_poll_queue(struct ap_device *ap_dev, unsigned long *flags) @@ -986,9 +1014,11 @@ static inline int ap_poll_queue(struct ap_device *ap_dev, unsigned long *flags) } /** - * Queue a message to a device. + * __ap_queue_message(): Queue a message to a device. * @ap_dev: pointer to the AP device * @ap_msg: the message to be queued + * + * Queue a message to a device. Returns 0 if successful. */ static int __ap_queue_message(struct ap_device *ap_dev, struct ap_message *ap_msg) { @@ -1055,12 +1085,14 @@ void ap_queue_message(struct ap_device *ap_dev, struct ap_message *ap_msg) EXPORT_SYMBOL(ap_queue_message); /** + * ap_cancel_message(): Cancel a crypto request. + * @ap_dev: The AP device that has the message queued + * @ap_msg: The message that is to be removed + * * Cancel a crypto request. This is done by removing the request - * from the devive pendingq or requestq queue. Note that the + * from the device pending or request queue. Note that the * request stays on the AP queue. When it finishes the message * reply will be discarded because the psmid can't be found. - * @ap_dev: AP device that has the message queued - * @ap_msg: the message that is to be removed */ void ap_cancel_message(struct ap_device *ap_dev, struct ap_message *ap_msg) { @@ -1082,7 +1114,10 @@ void ap_cancel_message(struct ap_device *ap_dev, struct ap_message *ap_msg) EXPORT_SYMBOL(ap_cancel_message); /** - * AP receive polling for finished AP requests + * ap_poll_timeout(): AP receive polling for finished AP requests. + * @unused: Unused variable. + * + * Schedules the AP tasklet. */ static void ap_poll_timeout(unsigned long unused) { @@ -1090,6 +1125,9 @@ static void ap_poll_timeout(unsigned long unused) } /** + * ap_reset(): Reset a not responding AP device. + * @ap_dev: Pointer to the AP device + * * Reset a not responding AP device and move all requests from the * pending queue to the request queue. */ @@ -1108,11 +1146,6 @@ static void ap_reset(struct ap_device *ap_dev) ap_dev->unregistered = 1; } -/** - * Poll all AP devices on the bus in a round robin fashion. Continue - * polling until bit 2^0 of the control flags is not set. If bit 2^1 - * of the control flags has been set arm the poll timer. - */ static int __ap_poll_all(struct ap_device *ap_dev, unsigned long *flags) { spin_lock(&ap_dev->lock); @@ -1126,6 +1159,14 @@ static int __ap_poll_all(struct ap_device *ap_dev, unsigned long *flags) return 0; } +/** + * ap_poll_all(): Poll all AP devices. + * @dummy: Unused variable + * + * Poll all AP devices on the bus in a round robin fashion. Continue + * polling until bit 2^0 of the control flags is not set. If bit 2^1 + * of the control flags has been set arm the poll timer. + */ static void ap_poll_all(unsigned long dummy) { unsigned long flags; @@ -1144,6 +1185,9 @@ static void ap_poll_all(unsigned long dummy) } /** + * ap_poll_thread(): Thread that polls for finished requests. + * @data: Unused pointer + * * AP bus poll thread. The purpose of this thread is to poll for * finished requests in a loop if there is a "free" cpu - that is * a cpu that doesn't have anything better to do. The polling stops @@ -1213,7 +1257,10 @@ static void ap_poll_thread_stop(void) } /** - * Handling of request timeouts + * ap_request_timeout(): Handling of request timeouts + * @data: Holds the AP device. + * + * Handles request timeouts. */ static void ap_request_timeout(unsigned long data) { @@ -1246,7 +1293,9 @@ static struct reset_call ap_reset_call = { }; /** - * The module initialization code. + * ap_module_init(): The module initialization code. + * + * Initializes the module. */ int __init ap_module_init(void) { @@ -1288,7 +1337,7 @@ int __init ap_module_init(void) if (ap_select_domain() == 0) ap_scan_bus(NULL); - /* Setup the ap bus rescan timer. */ + /* Setup the AP bus rescan timer. */ init_timer(&ap_config_timer); ap_config_timer.function = ap_config_timeout; ap_config_timer.data = 0; @@ -1325,7 +1374,9 @@ static int __ap_match_all(struct device *dev, void *data) } /** - * The module termination code + * ap_modules_exit(): The module termination code + * + * Terminates the module. */ void ap_module_exit(void) { diff --git a/drivers/s390/crypto/ap_bus.h b/drivers/s390/crypto/ap_bus.h index 87c2d6442875..c1e1200c43fc 100644 --- a/drivers/s390/crypto/ap_bus.h +++ b/drivers/s390/crypto/ap_bus.h @@ -50,6 +50,15 @@ typedef unsigned int ap_qid_t; #define AP_QID_QUEUE(_qid) ((_qid) & 15) /** + * structy ap_queue_status - Holds the AP queue status. + * @queue_empty: Shows if queue is empty + * @replies_waiting: Waiting replies + * @queue_full: Is 1 if the queue is full + * @pad: A 4 bit pad + * @int_enabled: Shows if interrupts are enabled for the AP + * @response_conde: Holds the 8 bit response code + * @pad2: A 16 bit pad + * * The ap queue status word is returned by all three AP functions * (PQAP, NQAP and DQAP). There's a set of flags in the first * byte, followed by a 1 byte response code. @@ -75,7 +84,7 @@ struct ap_queue_status { #define AP_RESPONSE_NO_FIRST_PART 0x13 #define AP_RESPONSE_MESSAGE_TOO_BIG 0x15 -/** +/* * Known device types */ #define AP_DEVICE_TYPE_PCICC 3 @@ -84,7 +93,7 @@ struct ap_queue_status { #define AP_DEVICE_TYPE_CEX2A 6 #define AP_DEVICE_TYPE_CEX2C 7 -/** +/* * AP reset flag states */ #define AP_RESET_IGNORE 0 /* request timeout will be ignored */ @@ -152,7 +161,7 @@ struct ap_message { .dev_type=(dt), \ .match_flags=AP_DEVICE_ID_MATCH_DEVICE_TYPE, -/** +/* * Note: don't use ap_send/ap_recv after using ap_queue_message * for the first time. Otherwise the ap message queue will get * confused. diff --git a/drivers/s390/crypto/zcrypt_api.c b/drivers/s390/crypto/zcrypt_api.c index e3625a47a596..4d36e805a234 100644 --- a/drivers/s390/crypto/zcrypt_api.c +++ b/drivers/s390/crypto/zcrypt_api.c @@ -36,10 +36,11 @@ #include <linux/compat.h> #include <asm/atomic.h> #include <asm/uaccess.h> +#include <linux/hw_random.h> #include "zcrypt_api.h" -/** +/* * Module description. */ MODULE_AUTHOR("IBM Corporation"); @@ -52,7 +53,10 @@ static LIST_HEAD(zcrypt_device_list); static int zcrypt_device_count = 0; static atomic_t zcrypt_open_count = ATOMIC_INIT(0); -/** +static int zcrypt_rng_device_add(void); +static void zcrypt_rng_device_remove(void); + +/* * Device attributes common for all crypto devices. */ static ssize_t zcrypt_type_show(struct device *dev, @@ -99,6 +103,9 @@ static struct attribute_group zcrypt_device_attr_group = { }; /** + * __zcrypt_increase_preference(): Increase preference of a crypto device. + * @zdev: Pointer the crypto device + * * Move the device towards the head of the device list. * Need to be called while holding the zcrypt device list lock. * Note: cards with speed_rating of 0 are kept at the end of the list. @@ -125,6 +132,9 @@ static void __zcrypt_increase_preference(struct zcrypt_device *zdev) } /** + * __zcrypt_decrease_preference(): Decrease preference of a crypto device. + * @zdev: Pointer to a crypto device. + * * Move the device towards the tail of the device list. * Need to be called while holding the zcrypt device list lock. * Note: cards with speed_rating of 0 are kept at the end of the list. @@ -198,7 +208,10 @@ void zcrypt_device_free(struct zcrypt_device *zdev) EXPORT_SYMBOL(zcrypt_device_free); /** - * Register a crypto device. + * zcrypt_device_register() - Register a crypto device. + * @zdev: Pointer to a crypto device + * + * Register a crypto device. Returns 0 if successful. */ int zcrypt_device_register(struct zcrypt_device *zdev) { @@ -216,16 +229,37 @@ int zcrypt_device_register(struct zcrypt_device *zdev) __zcrypt_increase_preference(zdev); zcrypt_device_count++; spin_unlock_bh(&zcrypt_device_lock); + if (zdev->ops->rng) { + rc = zcrypt_rng_device_add(); + if (rc) + goto out_unregister; + } + return 0; + +out_unregister: + spin_lock_bh(&zcrypt_device_lock); + zcrypt_device_count--; + list_del_init(&zdev->list); + spin_unlock_bh(&zcrypt_device_lock); + sysfs_remove_group(&zdev->ap_dev->device.kobj, + &zcrypt_device_attr_group); + put_device(&zdev->ap_dev->device); + zcrypt_device_put(zdev); out: return rc; } EXPORT_SYMBOL(zcrypt_device_register); /** + * zcrypt_device_unregister(): Unregister a crypto device. + * @zdev: Pointer to crypto device + * * Unregister a crypto device. */ void zcrypt_device_unregister(struct zcrypt_device *zdev) { + if (zdev->ops->rng) + zcrypt_rng_device_remove(); spin_lock_bh(&zcrypt_device_lock); zcrypt_device_count--; list_del_init(&zdev->list); @@ -238,7 +272,9 @@ void zcrypt_device_unregister(struct zcrypt_device *zdev) EXPORT_SYMBOL(zcrypt_device_unregister); /** - * zcrypt_read is not be supported beyond zcrypt 1.3.1 + * zcrypt_read (): Not supported beyond zcrypt 1.3.1. + * + * This function is not supported beyond zcrypt 1.3.1. */ static ssize_t zcrypt_read(struct file *filp, char __user *buf, size_t count, loff_t *f_pos) @@ -247,6 +283,8 @@ static ssize_t zcrypt_read(struct file *filp, char __user *buf, } /** + * zcrypt_write(): Not allowed. + * * Write is is not allowed */ static ssize_t zcrypt_write(struct file *filp, const char __user *buf, @@ -256,7 +294,9 @@ static ssize_t zcrypt_write(struct file *filp, const char __user *buf, } /** - * Device open/close functions to count number of users. + * zcrypt_open(): Count number of users. + * + * Device open function to count number of users. */ static int zcrypt_open(struct inode *inode, struct file *filp) { @@ -264,13 +304,18 @@ static int zcrypt_open(struct inode *inode, struct file *filp) return 0; } +/** + * zcrypt_release(): Count number of users. + * + * Device close function to count number of users. + */ static int zcrypt_release(struct inode *inode, struct file *filp) { atomic_dec(&zcrypt_open_count); return 0; } -/** +/* * zcrypt ioctls. */ static long zcrypt_rsa_modexpo(struct ica_rsa_modexpo *mex) @@ -280,7 +325,7 @@ static long zcrypt_rsa_modexpo(struct ica_rsa_modexpo *mex) if (mex->outputdatalength < mex->inputdatalength) return -EINVAL; - /** + /* * As long as outputdatalength is big enough, we can set the * outputdatalength equal to the inputdatalength, since that is the * number of bytes we will copy in any case @@ -326,7 +371,7 @@ static long zcrypt_rsa_crt(struct ica_rsa_modexpo_crt *crt) if (crt->outputdatalength < crt->inputdatalength || (crt->inputdatalength & 1)) return -EINVAL; - /** + /* * As long as outputdatalength is big enough, we can set the * outputdatalength equal to the inputdatalength, since that is the * number of bytes we will copy in any case @@ -343,7 +388,7 @@ static long zcrypt_rsa_crt(struct ica_rsa_modexpo_crt *crt) zdev->max_mod_size < crt->inputdatalength) continue; if (zdev->short_crt && crt->inputdatalength > 240) { - /** + /* * Check inputdata for leading zeros for cards * that can't handle np_prime, bp_key, or * u_mult_inv > 128 bytes. @@ -359,7 +404,7 @@ static long zcrypt_rsa_crt(struct ica_rsa_modexpo_crt *crt) copy_from_user(&z3, crt->u_mult_inv, len)) return -EFAULT; copied = 1; - /** + /* * We have to restart device lookup - * the device list may have changed by now. */ @@ -427,6 +472,37 @@ static long zcrypt_send_cprb(struct ica_xcRB *xcRB) return -ENODEV; } +static long zcrypt_rng(char *buffer) +{ + struct zcrypt_device *zdev; + int rc; + + spin_lock_bh(&zcrypt_device_lock); + list_for_each_entry(zdev, &zcrypt_device_list, list) { + if (!zdev->online || !zdev->ops->rng) + continue; + zcrypt_device_get(zdev); + get_device(&zdev->ap_dev->device); + zdev->request_count++; + __zcrypt_decrease_preference(zdev); + if (try_module_get(zdev->ap_dev->drv->driver.owner)) { + spin_unlock_bh(&zcrypt_device_lock); + rc = zdev->ops->rng(zdev, buffer); + spin_lock_bh(&zcrypt_device_lock); + module_put(zdev->ap_dev->drv->driver.owner); + } else + rc = -EAGAIN; + zdev->request_count--; + __zcrypt_increase_preference(zdev); + put_device(&zdev->ap_dev->device); + zcrypt_device_put(zdev); + spin_unlock_bh(&zcrypt_device_lock); + return rc; + } + spin_unlock_bh(&zcrypt_device_lock); + return -ENODEV; +} + static void zcrypt_status_mask(char status[AP_DEVICES]) { struct zcrypt_device *zdev; @@ -514,6 +590,8 @@ static int zcrypt_count_type(int type) } /** + * zcrypt_ica_status(): Old, depracted combi status call. + * * Old, deprecated combi status call. */ static long zcrypt_ica_status(struct file *filp, unsigned long arg) @@ -615,7 +693,7 @@ static long zcrypt_unlocked_ioctl(struct file *filp, unsigned int cmd, (int __user *) arg); case Z90STAT_DOMAIN_INDEX: return put_user(ap_domain_index, (int __user *) arg); - /** + /* * Deprecated ioctls. Don't add another device count ioctl, * you can count them yourself in the user space with the * output of the Z90STAT_STATUS_MASK ioctl. @@ -653,7 +731,7 @@ static long zcrypt_unlocked_ioctl(struct file *filp, unsigned int cmd, } #ifdef CONFIG_COMPAT -/** +/* * ioctl32 conversion routines */ struct compat_ica_rsa_modexpo { @@ -804,7 +882,7 @@ static long zcrypt_compat_ioctl(struct file *filp, unsigned int cmd, } #endif -/** +/* * Misc device file operations. */ static const struct file_operations zcrypt_fops = { @@ -819,7 +897,7 @@ static const struct file_operations zcrypt_fops = { .release = zcrypt_release }; -/** +/* * Misc device. */ static struct miscdevice zcrypt_misc_device = { @@ -828,7 +906,7 @@ static struct miscdevice zcrypt_misc_device = { .fops = &zcrypt_fops, }; -/** +/* * Deprecated /proc entry support. */ static struct proc_dir_entry *zcrypt_entry; @@ -1022,7 +1100,7 @@ static int zcrypt_status_write(struct file *file, const char __user *buffer, } for (j = 0; j < 64 && *ptr; ptr++) { - /** + /* * '0' for no device, '1' for PCICA, '2' for PCICC, * '3' for PCIXCC_MCL2, '4' for PCIXCC_MCL3, * '5' for CEX2C and '6' for CEX2A' @@ -1041,7 +1119,76 @@ out: return count; } +static int zcrypt_rng_device_count; +static u32 *zcrypt_rng_buffer; +static int zcrypt_rng_buffer_index; +static DEFINE_MUTEX(zcrypt_rng_mutex); + +static int zcrypt_rng_data_read(struct hwrng *rng, u32 *data) +{ + int rc; + + /* + * We don't need locking here because the RNG API guarantees serialized + * read method calls. + */ + if (zcrypt_rng_buffer_index == 0) { + rc = zcrypt_rng((char *) zcrypt_rng_buffer); + if (rc < 0) + return -EIO; + zcrypt_rng_buffer_index = rc / sizeof *data; + } + *data = zcrypt_rng_buffer[--zcrypt_rng_buffer_index]; + return sizeof *data; +} + +static struct hwrng zcrypt_rng_dev = { + .name = "zcrypt", + .data_read = zcrypt_rng_data_read, +}; + +static int zcrypt_rng_device_add(void) +{ + int rc = 0; + + mutex_lock(&zcrypt_rng_mutex); + if (zcrypt_rng_device_count == 0) { + zcrypt_rng_buffer = (u32 *) get_zeroed_page(GFP_KERNEL); + if (!zcrypt_rng_buffer) { + rc = -ENOMEM; + goto out; + } + zcrypt_rng_buffer_index = 0; + rc = hwrng_register(&zcrypt_rng_dev); + if (rc) + goto out_free; + zcrypt_rng_device_count = 1; + } else + zcrypt_rng_device_count++; + mutex_unlock(&zcrypt_rng_mutex); + return 0; + +out_free: + free_page((unsigned long) zcrypt_rng_buffer); +out: + mutex_unlock(&zcrypt_rng_mutex); + return rc; +} + +static void zcrypt_rng_device_remove(void) +{ + mutex_lock(&zcrypt_rng_mutex); + zcrypt_rng_device_count--; + if (zcrypt_rng_device_count == 0) { + hwrng_unregister(&zcrypt_rng_dev); + free_page((unsigned long) zcrypt_rng_buffer); + } + mutex_unlock(&zcrypt_rng_mutex); +} + /** + * zcrypt_api_init(): Module initialization. + * * The module initialization code. */ int __init zcrypt_api_init(void) @@ -1076,6 +1223,8 @@ out: } /** + * zcrypt_api_exit(): Module termination. + * * The module termination code. */ void zcrypt_api_exit(void) diff --git a/drivers/s390/crypto/zcrypt_api.h b/drivers/s390/crypto/zcrypt_api.h index de4877ee618f..5c6e222b2ac4 100644 --- a/drivers/s390/crypto/zcrypt_api.h +++ b/drivers/s390/crypto/zcrypt_api.h @@ -43,17 +43,17 @@ #define DEV_NAME "zcrypt" #define PRINTK(fmt, args...) \ - printk(KERN_DEBUG DEV_NAME ": %s -> " fmt, __FUNCTION__ , ## args) + printk(KERN_DEBUG DEV_NAME ": %s -> " fmt, __func__ , ## args) #define PRINTKN(fmt, args...) \ printk(KERN_DEBUG DEV_NAME ": " fmt, ## args) #define PRINTKW(fmt, args...) \ - printk(KERN_WARNING DEV_NAME ": %s -> " fmt, __FUNCTION__ , ## args) + printk(KERN_WARNING DEV_NAME ": %s -> " fmt, __func__ , ## args) #define PRINTKC(fmt, args...) \ - printk(KERN_CRIT DEV_NAME ": %s -> " fmt, __FUNCTION__ , ## args) + printk(KERN_CRIT DEV_NAME ": %s -> " fmt, __func__ , ## args) #ifdef ZCRYPT_DEBUG #define PDEBUG(fmt, args...) \ - printk(KERN_DEBUG DEV_NAME ": %s -> " fmt, __FUNCTION__ , ## args) + printk(KERN_DEBUG DEV_NAME ": %s -> " fmt, __func__ , ## args) #else #define PDEBUG(fmt, args...) do {} while (0) #endif @@ -100,6 +100,13 @@ struct ica_z90_status { #define ZCRYPT_CEX2C 5 #define ZCRYPT_CEX2A 6 +/** + * Large random numbers are pulled in 4096 byte chunks from the crypto cards + * and stored in a page. Be carefull when increasing this buffer due to size + * limitations for AP requests. + */ +#define ZCRYPT_RNG_BUFFER_SIZE 4096 + struct zcrypt_device; struct zcrypt_ops { @@ -107,6 +114,7 @@ struct zcrypt_ops { long (*rsa_modexpo_crt)(struct zcrypt_device *, struct ica_rsa_modexpo_crt *); long (*send_cprb)(struct zcrypt_device *, struct ica_xcRB *); + long (*rng)(struct zcrypt_device *, char *); }; struct zcrypt_device { diff --git a/drivers/s390/crypto/zcrypt_cca_key.h b/drivers/s390/crypto/zcrypt_cca_key.h index 8dbcf0eef3e5..ed82f2f59b17 100644 --- a/drivers/s390/crypto/zcrypt_cca_key.h +++ b/drivers/s390/crypto/zcrypt_cca_key.h @@ -174,7 +174,7 @@ static inline int zcrypt_type6_mex_key_de(struct ica_rsa_modexpo *mex, key->pvtMeHdr = static_pvt_me_hdr; key->pvtMeSec = static_pvt_me_sec; key->pubMeSec = static_pub_me_sec; - /** + /* * In a private key, the modulus doesn't appear in the public * section. So, an arbitrary public exponent of 0x010001 will be * used. @@ -338,7 +338,7 @@ static inline int zcrypt_type6_crt_key(struct ica_rsa_modexpo_crt *crt, pub = (struct cca_public_sec *)(key->key_parts + key_len); *pub = static_cca_pub_sec; pub->modulus_bit_len = 8 * crt->inputdatalength; - /** + /* * In a private key, the modulus doesn't appear in the public * section. So, an arbitrary public exponent of 0x010001 will be * used. diff --git a/drivers/s390/crypto/zcrypt_error.h b/drivers/s390/crypto/zcrypt_error.h index 2cb616ba8bec..3e27fe77d207 100644 --- a/drivers/s390/crypto/zcrypt_error.h +++ b/drivers/s390/crypto/zcrypt_error.h @@ -108,7 +108,7 @@ static inline int convert_error(struct zcrypt_device *zdev, return -EINVAL; case REP82_ERROR_MESSAGE_TYPE: // REP88_ERROR_MESSAGE_TYPE // '20' CEX2A - /** + /* * To sent a message of the wrong type is a bug in the * device driver. Warn about it, disable the device * and then repeat the request. diff --git a/drivers/s390/crypto/zcrypt_pcicc.c b/drivers/s390/crypto/zcrypt_pcicc.c index d6d59bf9ac38..17ea56ce1c11 100644 --- a/drivers/s390/crypto/zcrypt_pcicc.c +++ b/drivers/s390/crypto/zcrypt_pcicc.c @@ -42,7 +42,7 @@ #define PCICC_MAX_MOD_SIZE_OLD 128 /* 1024 bits */ #define PCICC_MAX_MOD_SIZE 256 /* 2048 bits */ -/** +/* * PCICC cards need a speed rating of 0. This keeps them at the end of * the zcrypt device list (see zcrypt_api.c). PCICC cards are only * used if no other cards are present because they are slow and can only @@ -388,7 +388,7 @@ static int convert_type86(struct zcrypt_device *zdev, reply_len = le16_to_cpu(msg->length) - 2; if (reply_len > outputdatalength) return -EINVAL; - /** + /* * For all encipher requests, the length of the ciphertext (reply_len) * will always equal the modulus length. For MEX decipher requests * the output needs to get padded. Minimum pad size is 10. diff --git a/drivers/s390/crypto/zcrypt_pcixcc.c b/drivers/s390/crypto/zcrypt_pcixcc.c index 70b9ddc8cf9d..0bc9b3188e64 100644 --- a/drivers/s390/crypto/zcrypt_pcixcc.c +++ b/drivers/s390/crypto/zcrypt_pcixcc.c @@ -356,6 +356,55 @@ static int XCRB_msg_to_type6CPRB_msgX(struct zcrypt_device *zdev, } /** + * Prepare a type6 CPRB message for random number generation + * + * @ap_dev: AP device pointer + * @ap_msg: pointer to AP message + */ +static void rng_type6CPRB_msgX(struct ap_device *ap_dev, + struct ap_message *ap_msg, + unsigned random_number_length) +{ + struct { + struct type6_hdr hdr; + struct CPRBX cprbx; + char function_code[2]; + short int rule_length; + char rule[8]; + short int verb_length; + short int key_length; + } __attribute__((packed)) *msg = ap_msg->message; + static struct type6_hdr static_type6_hdrX = { + .type = 0x06, + .offset1 = 0x00000058, + .agent_id = {'C', 'A'}, + .function_code = {'R', 'L'}, + .ToCardLen1 = sizeof *msg - sizeof(msg->hdr), + .FromCardLen1 = sizeof *msg - sizeof(msg->hdr), + }; + static struct CPRBX static_cprbx = { + .cprb_len = 0x00dc, + .cprb_ver_id = 0x02, + .func_id = {0x54, 0x32}, + .req_parml = sizeof *msg - sizeof(msg->hdr) - + sizeof(msg->cprbx), + .rpl_msgbl = sizeof *msg - sizeof(msg->hdr), + }; + + msg->hdr = static_type6_hdrX; + msg->hdr.FromCardLen2 = random_number_length, + msg->cprbx = static_cprbx; + msg->cprbx.rpl_datal = random_number_length, + msg->cprbx.domain = AP_QID_QUEUE(ap_dev->qid); + memcpy(msg->function_code, msg->hdr.function_code, 0x02); + msg->rule_length = 0x0a; + memcpy(msg->rule, "RANDOM ", 8); + msg->verb_length = 0x02; + msg->key_length = 0x02; + ap_msg->length = sizeof *msg; +} + +/** * Copy results from a type 86 ICA reply message back to user space. * * @zdev: crypto device pointer @@ -452,7 +501,7 @@ static int convert_type86_ica(struct zcrypt_device *zdev, reply_len = msg->length - 2; if (reply_len > outputdatalength) return -EINVAL; - /** + /* * For all encipher requests, the length of the ciphertext (reply_len) * will always equal the modulus length. For MEX decipher requests * the output needs to get padded. Minimum pad size is 10. @@ -509,6 +558,26 @@ static int convert_type86_xcrb(struct zcrypt_device *zdev, return 0; } +static int convert_type86_rng(struct zcrypt_device *zdev, + struct ap_message *reply, + char *buffer) +{ + struct { + struct type86_hdr hdr; + struct type86_fmt2_ext fmt2; + struct CPRBX cprbx; + } __attribute__((packed)) *msg = reply->message; + char *data = reply->message; + + if (msg->cprbx.ccp_rtcode != 0 || msg->cprbx.ccp_rscode != 0) { + PDEBUG("RNG response error on PCIXCC/CEX2C rc=%hu/rs=%hu\n", + rc, rs); + return -EINVAL; + } + memcpy(buffer, data + msg->fmt2.offset2, msg->fmt2.count2); + return msg->fmt2.count2; +} + static int convert_response_ica(struct zcrypt_device *zdev, struct ap_message *reply, char __user *outputdata, @@ -567,6 +636,31 @@ static int convert_response_xcrb(struct zcrypt_device *zdev, } } +static int convert_response_rng(struct zcrypt_device *zdev, + struct ap_message *reply, + char *data) +{ + struct type86x_reply *msg = reply->message; + + switch (msg->hdr.type) { + case TYPE82_RSP_CODE: + case TYPE88_RSP_CODE: + return -EINVAL; + case TYPE86_RSP_CODE: + if (msg->hdr.reply_code) + return -EINVAL; + if (msg->cprbx.cprb_ver_id == 0x02) + return convert_type86_rng(zdev, reply, data); + /* no break, incorrect cprb version is an unknown response */ + default: /* Unknown response type, this should NEVER EVER happen */ + PRINTK("Unrecognized Message Header: %08x%08x\n", + *(unsigned int *) reply->message, + *(unsigned int *) (reply->message+4)); + zdev->online = 0; + return -EAGAIN; /* repeat the request on a different device. */ + } +} + /** * This function is called from the AP bus code after a crypto request * "msg" has finished with the reply message "reply". @@ -736,6 +830,42 @@ out_free: } /** + * The request distributor calls this function if it picked the PCIXCC/CEX2C + * device to generate random data. + * @zdev: pointer to zcrypt_device structure that identifies the + * PCIXCC/CEX2C device to the request distributor + * @buffer: pointer to a memory page to return random data + */ + +static long zcrypt_pcixcc_rng(struct zcrypt_device *zdev, + char *buffer) +{ + struct ap_message ap_msg; + struct response_type resp_type = { + .type = PCIXCC_RESPONSE_TYPE_XCRB, + }; + int rc; + + ap_msg.message = kmalloc(PCIXCC_MAX_XCRB_MESSAGE_SIZE, GFP_KERNEL); + if (!ap_msg.message) + return -ENOMEM; + ap_msg.psmid = (((unsigned long long) current->pid) << 32) + + atomic_inc_return(&zcrypt_step); + ap_msg.private = &resp_type; + rng_type6CPRB_msgX(zdev->ap_dev, &ap_msg, ZCRYPT_RNG_BUFFER_SIZE); + init_completion(&resp_type.work); + ap_queue_message(zdev->ap_dev, &ap_msg); + rc = wait_for_completion_interruptible(&resp_type.work); + if (rc == 0) + rc = convert_response_rng(zdev, &ap_msg, buffer); + else + /* Signal pending. */ + ap_cancel_message(zdev->ap_dev, &ap_msg); + kfree(ap_msg.message); + return rc; +} + +/** * The crypto operations for a PCIXCC/CEX2C card. */ static struct zcrypt_ops zcrypt_pcixcc_ops = { @@ -744,6 +874,13 @@ static struct zcrypt_ops zcrypt_pcixcc_ops = { .send_cprb = zcrypt_pcixcc_send_cprb, }; +static struct zcrypt_ops zcrypt_pcixcc_with_rng_ops = { + .rsa_modexpo = zcrypt_pcixcc_modexpo, + .rsa_modexpo_crt = zcrypt_pcixcc_modexpo_crt, + .send_cprb = zcrypt_pcixcc_send_cprb, + .rng = zcrypt_pcixcc_rng, +}; + /** * Micro-code detection function. Its sends a message to a pcixcc card * to find out the microcode level. @@ -859,6 +996,58 @@ out_free: } /** + * Large random number detection function. Its sends a message to a pcixcc + * card to find out if large random numbers are supported. + * @ap_dev: pointer to the AP device. + * + * Returns 1 if large random numbers are supported, 0 if not and < 0 on error. + */ +static int zcrypt_pcixcc_rng_supported(struct ap_device *ap_dev) +{ + struct ap_message ap_msg; + unsigned long long psmid; + struct { + struct type86_hdr hdr; + struct type86_fmt2_ext fmt2; + struct CPRBX cprbx; + } __attribute__((packed)) *reply; + int rc, i; + + ap_msg.message = (void *) get_zeroed_page(GFP_KERNEL); + if (!ap_msg.message) + return -ENOMEM; + + rng_type6CPRB_msgX(ap_dev, &ap_msg, 4); + rc = ap_send(ap_dev->qid, 0x0102030405060708ULL, ap_msg.message, + ap_msg.length); + if (rc) + goto out_free; + + /* Wait for the test message to complete. */ + for (i = 0; i < 2 * HZ; i++) { + msleep(1000 / HZ); + rc = ap_recv(ap_dev->qid, &psmid, ap_msg.message, 4096); + if (rc == 0 && psmid == 0x0102030405060708ULL) + break; + } + + if (i >= 2 * HZ) { + /* Got no answer. */ + rc = -ENODEV; + goto out_free; + } + + reply = ap_msg.message; + if (reply->cprbx.ccp_rtcode == 0 && reply->cprbx.ccp_rscode == 0) + rc = 1; + else + rc = 0; +out_free: + free_page((unsigned long) ap_msg.message); + return rc; +} + +/** * Probe function for PCIXCC/CEX2C cards. It always accepts the AP device * since the bus_match already checked the hardware type. The PCIXCC * cards come in two flavours: micro code level 2 and micro code level 3. @@ -874,7 +1063,6 @@ static int zcrypt_pcixcc_probe(struct ap_device *ap_dev) if (!zdev) return -ENOMEM; zdev->ap_dev = ap_dev; - zdev->ops = &zcrypt_pcixcc_ops; zdev->online = 1; if (ap_dev->device_type == AP_DEVICE_TYPE_PCIXCC) { rc = zcrypt_pcixcc_mcl(ap_dev); @@ -901,6 +1089,15 @@ static int zcrypt_pcixcc_probe(struct ap_device *ap_dev) zdev->min_mod_size = PCIXCC_MIN_MOD_SIZE; zdev->max_mod_size = PCIXCC_MAX_MOD_SIZE; } + rc = zcrypt_pcixcc_rng_supported(ap_dev); + if (rc < 0) { + zcrypt_device_free(zdev); + return rc; + } + if (rc) + zdev->ops = &zcrypt_pcixcc_with_rng_ops; + else + zdev->ops = &zcrypt_pcixcc_ops; ap_dev->reply = &zdev->reply; ap_dev->private = zdev; rc = zcrypt_device_register(zdev); diff --git a/drivers/s390/net/claw.c b/drivers/s390/net/claw.c index d8a5c229c5a7..04a1d7bf678c 100644 --- a/drivers/s390/net/claw.c +++ b/drivers/s390/net/claw.c @@ -299,7 +299,7 @@ claw_probe(struct ccwgroup_device *cgdev) struct claw_privbk *privptr=NULL; #ifdef FUNCTRACE - printk(KERN_INFO "%s Enter\n",__FUNCTION__); + printk(KERN_INFO "%s Enter\n",__func__); #endif CLAW_DBF_TEXT(2,setup,"probe"); if (!get_device(&cgdev->dev)) @@ -313,7 +313,7 @@ claw_probe(struct ccwgroup_device *cgdev) probe_error(cgdev); put_device(&cgdev->dev); printk(KERN_WARNING "Out of memory %s %s Exit Line %d \n", - cgdev->cdev[0]->dev.bus_id,__FUNCTION__,__LINE__); + cgdev->cdev[0]->dev.bus_id,__func__,__LINE__); CLAW_DBF_TEXT_(2,setup,"probex%d",-ENOMEM); return -ENOMEM; } @@ -323,7 +323,7 @@ claw_probe(struct ccwgroup_device *cgdev) probe_error(cgdev); put_device(&cgdev->dev); printk(KERN_WARNING "Out of memory %s %s Exit Line %d \n", - cgdev->cdev[0]->dev.bus_id,__FUNCTION__,__LINE__); + cgdev->cdev[0]->dev.bus_id,__func__,__LINE__); CLAW_DBF_TEXT_(2,setup,"probex%d",-ENOMEM); return -ENOMEM; } @@ -340,7 +340,7 @@ claw_probe(struct ccwgroup_device *cgdev) probe_error(cgdev); put_device(&cgdev->dev); printk(KERN_WARNING "add_files failed %s %s Exit Line %d \n", - cgdev->cdev[0]->dev.bus_id,__FUNCTION__,__LINE__); + cgdev->cdev[0]->dev.bus_id,__func__,__LINE__); CLAW_DBF_TEXT_(2,setup,"probex%d",rc); return rc; } @@ -351,7 +351,7 @@ claw_probe(struct ccwgroup_device *cgdev) cgdev->dev.driver_data = privptr; #ifdef FUNCTRACE printk(KERN_INFO "claw:%s exit on line %d, " - "rc = 0\n",__FUNCTION__,__LINE__); + "rc = 0\n",__func__,__LINE__); #endif CLAW_DBF_TEXT(2,setup,"prbext 0"); @@ -371,7 +371,7 @@ claw_tx(struct sk_buff *skb, struct net_device *dev) struct chbk *p_ch; #ifdef FUNCTRACE - printk(KERN_INFO "%s:%s enter\n",dev->name,__FUNCTION__); + printk(KERN_INFO "%s:%s enter\n",dev->name,__func__); #endif CLAW_DBF_TEXT(4,trace,"claw_tx"); p_ch=&privptr->channel[WRITE]; @@ -381,7 +381,7 @@ claw_tx(struct sk_buff *skb, struct net_device *dev) privptr->stats.tx_dropped++; #ifdef FUNCTRACE printk(KERN_INFO "%s: %s() exit on line %d, rc = EIO\n", - dev->name,__FUNCTION__, __LINE__); + dev->name,__func__, __LINE__); #endif CLAW_DBF_TEXT_(2,trace,"clawtx%d",-EIO); return -EIO; @@ -398,7 +398,7 @@ claw_tx(struct sk_buff *skb, struct net_device *dev) spin_unlock_irqrestore(get_ccwdev_lock(p_ch->cdev), saveflags); #ifdef FUNCTRACE printk(KERN_INFO "%s:%s exit on line %d, rc = %d\n", - dev->name, __FUNCTION__, __LINE__, rc); + dev->name, __func__, __LINE__, rc); #endif CLAW_DBF_TEXT_(4,trace,"clawtx%d",rc); return rc; @@ -460,7 +460,7 @@ claw_pack_skb(struct claw_privbk *privptr) #ifdef IOTRACE printk(KERN_INFO "%s: %s() Packed %d len %d\n", p_env->ndev->name, - __FUNCTION__,pkt_cnt,new_skb->len); + __func__,pkt_cnt,new_skb->len); #endif } CLAW_DBF_TEXT(4,trace,"PackSKBx"); @@ -478,7 +478,7 @@ claw_change_mtu(struct net_device *dev, int new_mtu) struct claw_privbk *privptr=dev->priv; int buff_size; #ifdef FUNCTRACE - printk(KERN_INFO "%s:%s Enter \n",dev->name,__FUNCTION__); + printk(KERN_INFO "%s:%s Enter \n",dev->name,__func__); #endif #ifdef DEBUGMSG printk(KERN_INFO "variable dev =\n"); @@ -491,14 +491,14 @@ claw_change_mtu(struct net_device *dev, int new_mtu) #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d, rc=EINVAL\n", dev->name, - __FUNCTION__, __LINE__); + __func__, __LINE__); #endif return -EINVAL; } dev->mtu = new_mtu; #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d\n",dev->name, - __FUNCTION__, __LINE__); + __func__, __LINE__); #endif return 0; } /* end of claw_change_mtu */ @@ -522,7 +522,7 @@ claw_open(struct net_device *dev) struct ccwbk *p_buf; #ifdef FUNCTRACE - printk(KERN_INFO "%s:%s Enter \n",dev->name,__FUNCTION__); + printk(KERN_INFO "%s:%s Enter \n",dev->name,__func__); #endif CLAW_DBF_TEXT(4,trace,"open"); if (!dev || (dev->name[0] == 0x00)) { @@ -537,7 +537,7 @@ claw_open(struct net_device *dev) if (rc) { printk(KERN_INFO "%s:%s Exit on line %d, rc=ENOMEM\n", dev->name, - __FUNCTION__, __LINE__); + __func__, __LINE__); CLAW_DBF_TEXT(2,trace,"openmem"); return -ENOMEM; } @@ -661,7 +661,7 @@ claw_open(struct net_device *dev) claw_clear_busy(dev); #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d, rc=EIO\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif CLAW_DBF_TEXT(2,trace,"open EIO"); return -EIO; @@ -673,7 +673,7 @@ claw_open(struct net_device *dev) #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d, rc=0\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif CLAW_DBF_TEXT(4,trace,"openok"); return 0; @@ -696,7 +696,7 @@ claw_irq_handler(struct ccw_device *cdev, #ifdef FUNCTRACE - printk(KERN_INFO "%s enter \n",__FUNCTION__); + printk(KERN_INFO "%s enter \n",__func__); #endif CLAW_DBF_TEXT(4,trace,"clawirq"); /* Bypass all 'unsolicited interrupts' */ @@ -706,7 +706,7 @@ claw_irq_handler(struct ccw_device *cdev, cdev->dev.bus_id,irb->scsw.cstat, irb->scsw.dstat); #ifdef FUNCTRACE printk(KERN_INFO "claw: %s() " - "exit on line %d\n",__FUNCTION__,__LINE__); + "exit on line %d\n",__func__,__LINE__); #endif CLAW_DBF_TEXT(2,trace,"badirq"); return; @@ -752,7 +752,7 @@ claw_irq_handler(struct ccw_device *cdev, #endif #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif CLAW_DBF_TEXT(2,trace,"chanchk"); /* return; */ @@ -777,7 +777,7 @@ claw_irq_handler(struct ccw_device *cdev, (SCSW_STCTL_ALERT_STATUS | SCSW_STCTL_STATUS_PEND)))) { #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif return; } @@ -788,7 +788,7 @@ claw_irq_handler(struct ccw_device *cdev, #endif #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif CLAW_DBF_TEXT(4,trace,"stop"); return; @@ -804,7 +804,7 @@ claw_irq_handler(struct ccw_device *cdev, (SCSW_STCTL_ALERT_STATUS | SCSW_STCTL_STATUS_PEND)))) { #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif CLAW_DBF_TEXT(4,trace,"haltio"); return; @@ -838,7 +838,7 @@ claw_irq_handler(struct ccw_device *cdev, #endif #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif CLAW_DBF_TEXT(4,trace,"haltio"); return; @@ -858,7 +858,7 @@ claw_irq_handler(struct ccw_device *cdev, } #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif CLAW_DBF_TEXT(4,trace,"notrdy"); return; @@ -874,7 +874,7 @@ claw_irq_handler(struct ccw_device *cdev, } #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif CLAW_DBF_TEXT(4,trace,"PCI_read"); return; @@ -885,7 +885,7 @@ claw_irq_handler(struct ccw_device *cdev, (SCSW_STCTL_ALERT_STATUS | SCSW_STCTL_STATUS_PEND)))) { #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif CLAW_DBF_TEXT(4,trace,"SPend_rd"); return; @@ -906,7 +906,7 @@ claw_irq_handler(struct ccw_device *cdev, #endif #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif CLAW_DBF_TEXT(4,trace,"RdIRQXit"); return; @@ -929,7 +929,7 @@ claw_irq_handler(struct ccw_device *cdev, } #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif CLAW_DBF_TEXT(4,trace,"rstrtwrt"); return; @@ -946,7 +946,7 @@ claw_irq_handler(struct ccw_device *cdev, (SCSW_STCTL_ALERT_STATUS | SCSW_STCTL_STATUS_PEND)))) { #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif CLAW_DBF_TEXT(4,trace,"writeUE"); return; @@ -969,7 +969,7 @@ claw_irq_handler(struct ccw_device *cdev, #endif #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif CLAW_DBF_TEXT(4,trace,"StWtExit"); return; @@ -978,7 +978,7 @@ claw_irq_handler(struct ccw_device *cdev, "state=%d\n",dev->name,p_ch->claw_state); #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif CLAW_DBF_TEXT(2,trace,"badIRQ"); return; @@ -1001,7 +1001,7 @@ claw_irq_tasklet ( unsigned long data ) p_ch = (struct chbk *) data; dev = (struct net_device *)p_ch->ndev; #ifdef FUNCTRACE - printk(KERN_INFO "%s:%s Enter \n",dev->name,__FUNCTION__); + printk(KERN_INFO "%s:%s Enter \n",dev->name,__func__); #endif #ifdef DEBUGMSG printk(KERN_INFO "%s: variable p_ch =\n",dev->name); @@ -1021,7 +1021,7 @@ claw_irq_tasklet ( unsigned long data ) CLAW_DBF_TEXT(4,trace,"TskletXt"); #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif return; } /* end of claw_irq_bh */ @@ -1048,7 +1048,7 @@ claw_release(struct net_device *dev) if (!privptr) return 0; #ifdef FUNCTRACE - printk(KERN_INFO "%s:%s Enter \n",dev->name,__FUNCTION__); + printk(KERN_INFO "%s:%s Enter \n",dev->name,__func__); #endif CLAW_DBF_TEXT(4,trace,"release"); #ifdef DEBUGMSG @@ -1090,7 +1090,7 @@ claw_release(struct net_device *dev) if(privptr->buffs_alloc != 1) { #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif CLAW_DBF_TEXT(4,trace,"none2fre"); return 0; @@ -1171,7 +1171,7 @@ claw_release(struct net_device *dev) } #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif CLAW_DBF_TEXT(4,trace,"rlsexit"); return 0; @@ -1192,7 +1192,7 @@ claw_write_retry ( struct chbk *p_ch ) #ifdef FUNCTRACE - printk(KERN_INFO "%s:%s Enter\n",dev->name,__FUNCTION__); + printk(KERN_INFO "%s:%s Enter\n",dev->name,__func__); printk(KERN_INFO "claw: variable p_ch =\n"); dumpit((char *) p_ch, sizeof(struct chbk)); #endif @@ -1200,20 +1200,20 @@ claw_write_retry ( struct chbk *p_ch ) if (p_ch->claw_state == CLAW_STOP) { #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif return; } #ifdef DEBUGMSG printk( KERN_INFO "%s:%s state-%02x\n" , dev->name, - __FUNCTION__, + __func__, p_ch->claw_state); #endif claw_strt_out_IO( dev ); #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif CLAW_DBF_TEXT(4,trace,"rtry_xit"); return; @@ -1235,7 +1235,7 @@ claw_write_next ( struct chbk * p_ch ) int rc; #ifdef FUNCTRACE - printk(KERN_INFO "%s:%s Enter \n",p_ch->ndev->name,__FUNCTION__); + printk(KERN_INFO "%s:%s Enter \n",p_ch->ndev->name,__func__); printk(KERN_INFO "%s: variable p_ch =\n",p_ch->ndev->name); dumpit((char *) p_ch, sizeof(struct chbk)); #endif @@ -1262,7 +1262,7 @@ claw_write_next ( struct chbk * p_ch ) #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif return; } /* end of claw_write_next */ @@ -1276,7 +1276,7 @@ static void claw_timer ( struct chbk * p_ch ) { #ifdef FUNCTRACE - printk(KERN_INFO "%s:%s Entry\n",p_ch->ndev->name,__FUNCTION__); + printk(KERN_INFO "%s:%s Entry\n",p_ch->ndev->name,__func__); printk(KERN_INFO "%s: variable p_ch =\n",p_ch->ndev->name); dumpit((char *) p_ch, sizeof(struct chbk)); #endif @@ -1285,7 +1285,7 @@ claw_timer ( struct chbk * p_ch ) wake_up(&p_ch->wait); #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d\n", - p_ch->ndev->name,__FUNCTION__,__LINE__); + p_ch->ndev->name,__func__,__LINE__); #endif return; } /* end of claw_timer */ @@ -1312,7 +1312,7 @@ pages_to_order_of_mag(int num_of_pages) int order_of_mag=1; /* assume 2 pages */ int nump=2; #ifdef FUNCTRACE - printk(KERN_INFO "%s Enter pages = %d \n",__FUNCTION__,num_of_pages); + printk(KERN_INFO "%s Enter pages = %d \n",__func__,num_of_pages); #endif CLAW_DBF_TEXT_(5,trace,"pages%d",num_of_pages); if (num_of_pages == 1) {return 0; } /* magnitude of 0 = 1 page */ @@ -1327,7 +1327,7 @@ pages_to_order_of_mag(int num_of_pages) if (order_of_mag > 9) { order_of_mag = 9; } /* I know it's paranoid */ #ifdef FUNCTRACE printk(KERN_INFO "%s Exit on line %d, order = %d\n", - __FUNCTION__,__LINE__, order_of_mag); + __func__,__LINE__, order_of_mag); #endif CLAW_DBF_TEXT_(5,trace,"mag%d",order_of_mag); return order_of_mag; @@ -1349,7 +1349,7 @@ add_claw_reads(struct net_device *dev, struct ccwbk* p_first, struct ccwbk* p_buf; #endif #ifdef FUNCTRACE - printk(KERN_INFO "%s:%s Enter \n",dev->name,__FUNCTION__); + printk(KERN_INFO "%s:%s Enter \n",dev->name,__func__); #endif #ifdef DEBUGMSG printk(KERN_INFO "dev\n"); @@ -1369,7 +1369,7 @@ add_claw_reads(struct net_device *dev, struct ccwbk* p_first, if ( p_first==NULL) { #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif CLAW_DBF_TEXT(4,trace,"addexit"); return 0; @@ -1400,9 +1400,9 @@ add_claw_reads(struct net_device *dev, struct ccwbk* p_first, if ( privptr-> p_read_active_first ==NULL ) { #ifdef DEBUGMSG printk(KERN_INFO "%s:%s p_read_active_first == NULL \n", - dev->name,__FUNCTION__); + dev->name,__func__); printk(KERN_INFO "%s:%s Read active first/last changed \n", - dev->name,__FUNCTION__); + dev->name,__func__); #endif privptr-> p_read_active_first= p_first; /* set new first */ privptr-> p_read_active_last = p_last; /* set new last */ @@ -1411,7 +1411,7 @@ add_claw_reads(struct net_device *dev, struct ccwbk* p_first, #ifdef DEBUGMSG printk(KERN_INFO "%s:%s Read in progress \n", - dev->name,__FUNCTION__); + dev->name,__func__); #endif /* set up TIC ccw */ temp_ccw.cda= (__u32)__pa(&p_first->read); @@ -1450,15 +1450,15 @@ add_claw_reads(struct net_device *dev, struct ccwbk* p_first, privptr->p_read_active_last=p_last; } /* end of if ( privptr-> p_read_active_first ==NULL) */ #ifdef IOTRACE - printk(KERN_INFO "%s:%s dump p_last CCW BK \n",dev->name,__FUNCTION__); + printk(KERN_INFO "%s:%s dump p_last CCW BK \n",dev->name,__func__); dumpit((char *)p_last, sizeof(struct ccwbk)); - printk(KERN_INFO "%s:%s dump p_end CCW BK \n",dev->name,__FUNCTION__); + printk(KERN_INFO "%s:%s dump p_end CCW BK \n",dev->name,__func__); dumpit((char *)p_end, sizeof(struct endccw)); - printk(KERN_INFO "%s:%s dump p_first CCW BK \n",dev->name,__FUNCTION__); + printk(KERN_INFO "%s:%s dump p_first CCW BK \n",dev->name,__func__); dumpit((char *)p_first, sizeof(struct ccwbk)); printk(KERN_INFO "%s:%s Dump Active CCW chain \n", - dev->name,__FUNCTION__); + dev->name,__func__); p_buf=privptr->p_read_active_first; while (p_buf!=NULL) { dumpit((char *)p_buf, sizeof(struct ccwbk)); @@ -1467,7 +1467,7 @@ add_claw_reads(struct net_device *dev, struct ccwbk* p_first, #endif #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif CLAW_DBF_TEXT(4,trace,"addexit"); return 0; @@ -1483,7 +1483,7 @@ ccw_check_return_code(struct ccw_device *cdev, int return_code) { #ifdef FUNCTRACE printk(KERN_INFO "%s: %s() > enter \n", - cdev->dev.bus_id,__FUNCTION__); + cdev->dev.bus_id,__func__); #endif CLAW_DBF_TEXT(4,trace,"ccwret"); #ifdef DEBUGMSG @@ -1516,7 +1516,7 @@ ccw_check_return_code(struct ccw_device *cdev, int return_code) } #ifdef FUNCTRACE printk(KERN_INFO "%s: %s() > exit on line %d\n", - cdev->dev.bus_id,__FUNCTION__,__LINE__); + cdev->dev.bus_id,__func__,__LINE__); #endif CLAW_DBF_TEXT(4,trace,"ccwret"); } /* end of ccw_check_return_code */ @@ -1531,7 +1531,7 @@ ccw_check_unit_check(struct chbk * p_ch, unsigned char sense ) struct net_device *dev = p_ch->ndev; #ifdef FUNCTRACE - printk(KERN_INFO "%s: %s() > enter\n",dev->name,__FUNCTION__); + printk(KERN_INFO "%s: %s() > enter\n",dev->name,__func__); #endif #ifdef DEBUGMSG printk(KERN_INFO "%s: variable dev =\n",dev->name); @@ -1578,7 +1578,7 @@ ccw_check_unit_check(struct chbk * p_ch, unsigned char sense ) #ifdef FUNCTRACE printk(KERN_INFO "%s: %s() exit on line %d\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif } /* end of ccw_check_unit_check */ @@ -1706,7 +1706,7 @@ find_link(struct net_device *dev, char *host_name, char *ws_name ) int rc=0; #ifdef FUNCTRACE - printk(KERN_INFO "%s:%s > enter \n",dev->name,__FUNCTION__); + printk(KERN_INFO "%s:%s > enter \n",dev->name,__func__); #endif CLAW_DBF_TEXT(2,setup,"findlink"); #ifdef DEBUGMSG @@ -1739,7 +1739,7 @@ find_link(struct net_device *dev, char *host_name, char *ws_name ) #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif return 0; } /* end of find_link */ @@ -1773,7 +1773,7 @@ claw_hw_tx(struct sk_buff *skb, struct net_device *dev, long linkid) struct ccwbk *p_buf; #endif #ifdef FUNCTRACE - printk(KERN_INFO "%s: %s() > enter\n",dev->name,__FUNCTION__); + printk(KERN_INFO "%s: %s() > enter\n",dev->name,__func__); #endif CLAW_DBF_TEXT(4,trace,"hw_tx"); #ifdef DEBUGMSG @@ -1787,7 +1787,7 @@ claw_hw_tx(struct sk_buff *skb, struct net_device *dev, long linkid) p_ch=(struct chbk *)&privptr->channel[WRITE]; p_env =privptr->p_env; #ifdef IOTRACE - printk(KERN_INFO "%s: %s() dump sk_buff \n",dev->name,__FUNCTION__); + printk(KERN_INFO "%s: %s() dump sk_buff \n",dev->name,__func__); dumpit((char *)skb ,sizeof(struct sk_buff)); #endif claw_free_wrt_buf(dev); /* Clean up free chain if posible */ @@ -1877,7 +1877,7 @@ claw_hw_tx(struct sk_buff *skb, struct net_device *dev, long linkid) while (len_of_data > 0) { #ifdef DEBUGMSG printk(KERN_INFO "%s: %s() length-of-data is %ld \n", - dev->name ,__FUNCTION__,len_of_data); + dev->name ,__func__,len_of_data); dumpit((char *)pDataAddress ,64); #endif p_this_ccw=privptr->p_write_free_chain; /* get a block */ @@ -1913,7 +1913,7 @@ claw_hw_tx(struct sk_buff *skb, struct net_device *dev, long linkid) p_last_ccw=p_this_ccw; /* save new last block */ #ifdef IOTRACE printk(KERN_INFO "%s: %s() > CCW and Buffer %ld bytes long \n", - dev->name,__FUNCTION__,bytesInThisBuffer); + dev->name,__func__,bytesInThisBuffer); dumpit((char *)p_this_ccw, sizeof(struct ccwbk)); dumpit((char *)p_this_ccw->p_buffer, 64); #endif @@ -1998,7 +1998,7 @@ claw_hw_tx(struct sk_buff *skb, struct net_device *dev, long linkid) #ifdef IOTRACE printk(KERN_INFO "%s: %s() > Dump Active CCW chain \n", - dev->name,__FUNCTION__); + dev->name,__func__); p_buf=privptr->p_write_active_first; while (p_buf!=NULL) { dumpit((char *)p_buf, sizeof(struct ccwbk)); @@ -2018,7 +2018,7 @@ claw_hw_tx(struct sk_buff *skb, struct net_device *dev, long linkid) /* if write free count is zero , set NOBUFFER */ #ifdef DEBUGMSG printk(KERN_INFO "%s: %s() > free_count is %d\n", - dev->name,__FUNCTION__, + dev->name,__func__, (int) privptr->write_free_count ); #endif if (privptr->write_free_count==0) { @@ -2029,7 +2029,7 @@ Done2: Done: #ifdef FUNCTRACE printk(KERN_INFO "%s: %s() > exit on line %d, rc = %d \n", - dev->name,__FUNCTION__,__LINE__, rc); + dev->name,__func__,__LINE__, rc); #endif return(rc); } /* end of claw_hw_tx */ @@ -2063,7 +2063,7 @@ init_ccw_bk(struct net_device *dev) addr_t real_TIC_address; int i,j; #ifdef FUNCTRACE - printk(KERN_INFO "%s: %s() enter \n",dev->name,__FUNCTION__); + printk(KERN_INFO "%s: %s() enter \n",dev->name,__func__); #endif CLAW_DBF_TEXT(4,trace,"init_ccw"); #ifdef DEBUGMSG @@ -2097,15 +2097,15 @@ init_ccw_bk(struct net_device *dev) #ifdef DEBUGMSG printk(KERN_INFO "%s: %s() " "ccw_blocks_required=%d\n", - dev->name,__FUNCTION__, + dev->name,__func__, ccw_blocks_required); printk(KERN_INFO "%s: %s() " "PAGE_SIZE=0x%x\n", - dev->name,__FUNCTION__, + dev->name,__func__, (unsigned int)PAGE_SIZE); printk(KERN_INFO "%s: %s() > " "PAGE_MASK=0x%x\n", - dev->name,__FUNCTION__, + dev->name,__func__, (unsigned int)PAGE_MASK); #endif /* @@ -2117,10 +2117,10 @@ init_ccw_bk(struct net_device *dev) #ifdef DEBUGMSG printk(KERN_INFO "%s: %s() > ccw_blocks_perpage=%d\n", - dev->name,__FUNCTION__, + dev->name,__func__, ccw_blocks_perpage); printk(KERN_INFO "%s: %s() > ccw_pages_required=%d\n", - dev->name,__FUNCTION__, + dev->name,__func__, ccw_pages_required); #endif /* @@ -2156,29 +2156,29 @@ init_ccw_bk(struct net_device *dev) #ifdef DEBUGMSG if (privptr->p_env->read_size < PAGE_SIZE) { printk(KERN_INFO "%s: %s() reads_perpage=%d\n", - dev->name,__FUNCTION__, + dev->name,__func__, claw_reads_perpage); } else { printk(KERN_INFO "%s: %s() pages_perread=%d\n", - dev->name,__FUNCTION__, + dev->name,__func__, privptr->p_buff_pages_perread); } printk(KERN_INFO "%s: %s() read_pages=%d\n", - dev->name,__FUNCTION__, + dev->name,__func__, claw_read_pages); if (privptr->p_env->write_size < PAGE_SIZE) { printk(KERN_INFO "%s: %s() writes_perpage=%d\n", - dev->name,__FUNCTION__, + dev->name,__func__, claw_writes_perpage); } else { printk(KERN_INFO "%s: %s() pages_perwrite=%d\n", - dev->name,__FUNCTION__, + dev->name,__func__, privptr->p_buff_pages_perwrite); } printk(KERN_INFO "%s: %s() write_pages=%d\n", - dev->name,__FUNCTION__, + dev->name,__func__, claw_write_pages); #endif @@ -2194,12 +2194,12 @@ init_ccw_bk(struct net_device *dev) printk(KERN_INFO "%s: %s() " "__get_free_pages for CCWs failed : " "pages is %d\n", - dev->name,__FUNCTION__, + dev->name,__func__, ccw_pages_required ); #ifdef FUNCTRACE printk(KERN_INFO "%s: %s() > " "exit on line %d, rc = ENOMEM\n", - dev->name,__FUNCTION__, + dev->name,__func__, __LINE__); #endif return -ENOMEM; @@ -2218,7 +2218,7 @@ init_ccw_bk(struct net_device *dev) /* Initialize ending CCW block */ #ifdef DEBUGMSG printk(KERN_INFO "%s: %s() begin initialize ending CCW blocks\n", - dev->name,__FUNCTION__); + dev->name,__func__); #endif p_endccw=privptr->p_end_ccw; @@ -2276,7 +2276,7 @@ init_ccw_bk(struct net_device *dev) #ifdef IOTRACE printk(KERN_INFO "%s: %s() dump claw ending CCW BK \n", - dev->name,__FUNCTION__); + dev->name,__func__); dumpit((char *)p_endccw, sizeof(struct endccw)); #endif @@ -2287,7 +2287,7 @@ init_ccw_bk(struct net_device *dev) #ifdef DEBUGMSG printk(KERN_INFO "%s: %s() Begin build a chain of CCW buffer \n", - dev->name,__FUNCTION__); + dev->name,__func__); #endif p_buff=privptr->p_buff_ccw; @@ -2306,7 +2306,7 @@ init_ccw_bk(struct net_device *dev) #ifdef DEBUGMSG printk(KERN_INFO "%s: %s() " "End build a chain of CCW buffer \n", - dev->name,__FUNCTION__); + dev->name,__func__); p_buf=p_free_chain; while (p_buf!=NULL) { dumpit((char *)p_buf, sizeof(struct ccwbk)); @@ -2321,7 +2321,7 @@ init_ccw_bk(struct net_device *dev) #ifdef DEBUGMSG printk(KERN_INFO "%s: %s() " "Begin initialize ClawSignalBlock \n", - dev->name,__FUNCTION__); + dev->name,__func__); #endif if (privptr->p_claw_signal_blk==NULL) { privptr->p_claw_signal_blk=p_free_chain; @@ -2334,7 +2334,7 @@ init_ccw_bk(struct net_device *dev) #ifdef DEBUGMSG printk(KERN_INFO "%s: %s() > End initialize " "ClawSignalBlock\n", - dev->name,__FUNCTION__); + dev->name,__func__); dumpit((char *)privptr->p_claw_signal_blk, sizeof(struct ccwbk)); #endif @@ -2349,14 +2349,14 @@ init_ccw_bk(struct net_device *dev) if (privptr->p_buff_write==NULL) { printk(KERN_INFO "%s: %s() __get_free_pages for write" " bufs failed : get is for %d pages\n", - dev->name,__FUNCTION__,claw_write_pages ); + dev->name,__func__,claw_write_pages ); free_pages((unsigned long)privptr->p_buff_ccw, (int)pages_to_order_of_mag(privptr->p_buff_ccw_num)); privptr->p_buff_ccw=NULL; #ifdef FUNCTRACE printk(KERN_INFO "%s: %s() > exit on line %d," "rc = ENOMEM\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif return -ENOMEM; } @@ -2369,7 +2369,7 @@ init_ccw_bk(struct net_device *dev) ccw_pages_required * PAGE_SIZE); #ifdef DEBUGMSG printk(KERN_INFO "%s: %s() Begin build claw write free " - "chain \n",dev->name,__FUNCTION__); + "chain \n",dev->name,__func__); #endif privptr->p_write_free_chain=NULL; @@ -2409,14 +2409,14 @@ init_ccw_bk(struct net_device *dev) #ifdef IOTRACE printk(KERN_INFO "%s:%s __get_free_pages " "for writes buf: get for %d pages\n", - dev->name,__FUNCTION__, + dev->name,__func__, privptr->p_buff_pages_perwrite); #endif if (p_buff==NULL) { printk(KERN_INFO "%s:%s __get_free_pages " "for writes buf failed : get is for %d pages\n", dev->name, - __FUNCTION__, + __func__, privptr->p_buff_pages_perwrite ); free_pages((unsigned long)privptr->p_buff_ccw, (int)pages_to_order_of_mag( @@ -2433,7 +2433,7 @@ init_ccw_bk(struct net_device *dev) #ifdef FUNCTRACE printk(KERN_INFO "%s: %s exit on line %d, rc = ENOMEM\n", dev->name, - __FUNCTION__, + __func__, __LINE__); #endif return -ENOMEM; @@ -2466,7 +2466,7 @@ init_ccw_bk(struct net_device *dev) #ifdef DEBUGMSG printk(KERN_INFO "%s:%s End build claw write free chain \n", - dev->name,__FUNCTION__); + dev->name,__func__); p_buf=privptr->p_write_free_chain; while (p_buf!=NULL) { dumpit((char *)p_buf, sizeof(struct ccwbk)); @@ -2485,7 +2485,7 @@ init_ccw_bk(struct net_device *dev) printk(KERN_INFO "%s: %s() " "__get_free_pages for read buf failed : " "get is for %d pages\n", - dev->name,__FUNCTION__,claw_read_pages ); + dev->name,__func__,claw_read_pages ); free_pages((unsigned long)privptr->p_buff_ccw, (int)pages_to_order_of_mag( privptr->p_buff_ccw_num)); @@ -2497,7 +2497,7 @@ init_ccw_bk(struct net_device *dev) privptr->p_buff_write=NULL; #ifdef FUNCTRACE printk(KERN_INFO "%s: %s() > exit on line %d, rc =" - " ENOMEM\n",dev->name,__FUNCTION__,__LINE__); + " ENOMEM\n",dev->name,__func__,__LINE__); #endif return -ENOMEM; } @@ -2509,7 +2509,7 @@ init_ccw_bk(struct net_device *dev) */ #ifdef DEBUGMSG printk(KERN_INFO "%s: %s() Begin build claw read free chain \n", - dev->name,__FUNCTION__); + dev->name,__func__); #endif p_buff=privptr->p_buff_read; for (i=0 ; i< privptr->p_env->read_buffers ; i++) { @@ -2590,7 +2590,7 @@ init_ccw_bk(struct net_device *dev) #ifdef DEBUGMSG printk(KERN_INFO "%s: %s() Begin build claw read free chain \n", - dev->name,__FUNCTION__); + dev->name,__func__); #endif for (i=0 ; i< privptr->p_env->read_buffers ; i++) { p_buff = (void *)__get_free_pages(__GFP_DMA, @@ -2598,7 +2598,7 @@ init_ccw_bk(struct net_device *dev) if (p_buff==NULL) { printk(KERN_INFO "%s: %s() __get_free_pages for read " "buf failed : get is for %d pages\n", - dev->name,__FUNCTION__, + dev->name,__func__, privptr->p_buff_pages_perread ); free_pages((unsigned long)privptr->p_buff_ccw, (int)pages_to_order_of_mag(privptr->p_buff_ccw_num)); @@ -2622,7 +2622,7 @@ init_ccw_bk(struct net_device *dev) privptr->p_buff_write=NULL; #ifdef FUNCTRACE printk(KERN_INFO "%s: %s() exit on line %d, rc = ENOMEM\n", - dev->name,__FUNCTION__, + dev->name,__func__, __LINE__); #endif return -ENOMEM; @@ -2695,7 +2695,7 @@ init_ccw_bk(struct net_device *dev) } /* pBuffread = NULL */ #ifdef DEBUGMSG printk(KERN_INFO "%s: %s() > End build claw read free chain \n", - dev->name,__FUNCTION__); + dev->name,__func__); p_buf=p_first_CCWB; while (p_buf!=NULL) { dumpit((char *)p_buf, sizeof(struct ccwbk)); @@ -2707,7 +2707,7 @@ init_ccw_bk(struct net_device *dev) privptr->buffs_alloc = 1; #ifdef FUNCTRACE printk(KERN_INFO "%s: %s() exit on line %d\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif return 0; } /* end of init_ccw_bk */ @@ -2723,11 +2723,11 @@ probe_error( struct ccwgroup_device *cgdev) { struct claw_privbk *privptr; #ifdef FUNCTRACE - printk(KERN_INFO "%s enter \n",__FUNCTION__); + printk(KERN_INFO "%s enter \n",__func__); #endif CLAW_DBF_TEXT(4,trace,"proberr"); #ifdef DEBUGMSG - printk(KERN_INFO "%s variable cgdev =\n",__FUNCTION__); + printk(KERN_INFO "%s variable cgdev =\n",__func__); dumpit((char *) cgdev, sizeof(struct ccwgroup_device)); #endif privptr=(struct claw_privbk *)cgdev->dev.driver_data; @@ -2741,7 +2741,7 @@ probe_error( struct ccwgroup_device *cgdev) } #ifdef FUNCTRACE printk(KERN_INFO "%s > exit on line %d\n", - __FUNCTION__,__LINE__); + __func__,__LINE__); #endif return; @@ -2772,7 +2772,7 @@ claw_process_control( struct net_device *dev, struct ccwbk * p_ccw) struct chbk *p_ch = NULL; #ifdef FUNCTRACE printk(KERN_INFO "%s: %s() > enter \n", - dev->name,__FUNCTION__); + dev->name,__func__); #endif CLAW_DBF_TEXT(2,setup,"clw_cntl"); #ifdef DEBUGMSG @@ -2794,7 +2794,7 @@ claw_process_control( struct net_device *dev, struct ccwbk * p_ccw) #ifdef FUNCTRACE printk(KERN_INFO "%s: %s() > " "exit on line %d, rc=0\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif return 0; } @@ -3057,7 +3057,7 @@ claw_process_control( struct net_device *dev, struct ccwbk * p_ccw) #ifdef FUNCTRACE printk(KERN_INFO "%s: %s() exit on line %d, rc = 0\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif return 0; @@ -3080,7 +3080,7 @@ claw_send_control(struct net_device *dev, __u8 type, __u8 link, struct sk_buff *skb; #ifdef FUNCTRACE - printk(KERN_INFO "%s:%s > enter \n",dev->name,__FUNCTION__); + printk(KERN_INFO "%s:%s > enter \n",dev->name,__func__); #endif CLAW_DBF_TEXT(2,setup,"sndcntl"); #ifdef DEBUGMSG @@ -3143,10 +3143,10 @@ claw_send_control(struct net_device *dev, __u8 type, __u8 link, skb = dev_alloc_skb(sizeof(struct clawctl)); if (!skb) { printk( "%s:%s low on mem, returning...\n", - dev->name,__FUNCTION__); + dev->name,__func__); #ifdef DEBUG printk(KERN_INFO "%s:%s Exit, rc = ENOMEM\n", - dev->name,__FUNCTION__); + dev->name,__func__); #endif return -ENOMEM; } @@ -3162,7 +3162,7 @@ claw_send_control(struct net_device *dev, __u8 type, __u8 link, claw_hw_tx(skb, dev, 0); #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif return 0; @@ -3180,7 +3180,7 @@ claw_snd_conn_req(struct net_device *dev, __u8 link) struct clawctl *p_ctl; #ifdef FUNCTRACE - printk(KERN_INFO "%s:%s Enter \n",dev->name,__FUNCTION__); + printk(KERN_INFO "%s:%s Enter \n",dev->name,__func__); #endif CLAW_DBF_TEXT(2,setup,"snd_conn"); #ifdef DEBUGMSG @@ -3193,7 +3193,7 @@ claw_snd_conn_req(struct net_device *dev, __u8 link) if ( privptr->system_validate_comp==0x00 ) { #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d, rc = 1\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif return rc; } @@ -3209,7 +3209,7 @@ claw_snd_conn_req(struct net_device *dev, __u8 link) HOST_APPL_NAME, privptr->p_env->api_type); #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d, rc = %d\n", - dev->name,__FUNCTION__,__LINE__, rc); + dev->name,__func__,__LINE__, rc); #endif return rc; @@ -3228,7 +3228,7 @@ claw_snd_disc(struct net_device *dev, struct clawctl * p_ctl) struct conncmd * p_connect; #ifdef FUNCTRACE - printk(KERN_INFO "%s:%s Enter\n",dev->name,__FUNCTION__); + printk(KERN_INFO "%s:%s Enter\n",dev->name,__func__); #endif CLAW_DBF_TEXT(2,setup,"snd_dsc"); #ifdef DEBUGMSG @@ -3244,7 +3244,7 @@ claw_snd_disc(struct net_device *dev, struct clawctl * p_ctl) p_connect->host_name, p_connect->WS_name); #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d, rc = %d\n", - dev->name,__FUNCTION__, __LINE__, rc); + dev->name,__func__, __LINE__, rc); #endif return rc; } /* end of claw_snd_disc */ @@ -3265,7 +3265,7 @@ claw_snd_sys_validate_rsp(struct net_device *dev, #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Enter\n", - dev->name,__FUNCTION__); + dev->name,__func__); #endif CLAW_DBF_TEXT(2,setup,"chkresp"); #ifdef DEBUGMSG @@ -3285,7 +3285,7 @@ claw_snd_sys_validate_rsp(struct net_device *dev, p_env->adapter_name ); #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d, rc = %d\n", - dev->name,__FUNCTION__,__LINE__, rc); + dev->name,__func__,__LINE__, rc); #endif return rc; } /* end of claw_snd_sys_validate_rsp */ @@ -3301,7 +3301,7 @@ claw_strt_conn_req(struct net_device *dev ) int rc; #ifdef FUNCTRACE - printk(KERN_INFO "%s:%s Enter\n",dev->name,__FUNCTION__); + printk(KERN_INFO "%s:%s Enter\n",dev->name,__func__); #endif CLAW_DBF_TEXT(2,setup,"conn_req"); #ifdef DEBUGMSG @@ -3311,7 +3311,7 @@ claw_strt_conn_req(struct net_device *dev ) rc=claw_snd_conn_req(dev, 1); #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d, rc = %d\n", - dev->name,__FUNCTION__,__LINE__, rc); + dev->name,__func__,__LINE__, rc); #endif return rc; } /* end of claw_strt_conn_req */ @@ -3327,13 +3327,13 @@ net_device_stats *claw_stats(struct net_device *dev) { struct claw_privbk *privptr; #ifdef FUNCTRACE - printk(KERN_INFO "%s:%s Enter\n",dev->name,__FUNCTION__); + printk(KERN_INFO "%s:%s Enter\n",dev->name,__func__); #endif CLAW_DBF_TEXT(4,trace,"stats"); privptr = dev->priv; #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif return &privptr->stats; } /* end of claw_stats */ @@ -3366,7 +3366,7 @@ unpack_read(struct net_device *dev ) int p=0; #ifdef FUNCTRACE - printk(KERN_INFO "%s:%s enter \n",dev->name,__FUNCTION__); + printk(KERN_INFO "%s:%s enter \n",dev->name,__func__); #endif CLAW_DBF_TEXT(4,trace,"unpkread"); p_first_ccw=NULL; @@ -3408,7 +3408,7 @@ unpack_read(struct net_device *dev ) if ((p_this_ccw->header.opcode & MORE_to_COME_FLAG)!=0) { #ifdef DEBUGMSG printk(KERN_INFO "%s: %s > More_to_come is ON\n", - dev->name,__FUNCTION__); + dev->name,__func__); #endif mtc_this_frm=1; if (p_this_ccw->header.length!= @@ -3435,7 +3435,7 @@ unpack_read(struct net_device *dev ) #ifdef DEBUGMSG printk(KERN_INFO "%s:%s goto next " "frame from MoretoComeSkip \n", - dev->name,__FUNCTION__); + dev->name,__func__); #endif goto NextFrame; } @@ -3445,7 +3445,7 @@ unpack_read(struct net_device *dev ) #ifdef DEBUGMSG printk(KERN_INFO "%s:%s goto next " "frame from claw_process_control \n", - dev->name,__FUNCTION__); + dev->name,__func__); #endif CLAW_DBF_TEXT(4,trace,"UnpkCntl"); goto NextFrame; @@ -3468,7 +3468,7 @@ unpack_next: if (privptr->mtc_logical_link<0) { #ifdef DEBUGMSG printk(KERN_INFO "%s: %s mtc_logical_link < 0 \n", - dev->name,__FUNCTION__); + dev->name,__func__); #endif /* @@ -3487,7 +3487,7 @@ unpack_next: printk(KERN_INFO "%s: %s > goto next " "frame from MoretoComeSkip \n", dev->name, - __FUNCTION__); + __func__); printk(KERN_INFO " bytes_to_mov %d > (MAX_ENVELOPE_" "SIZE-privptr->mtc_offset %d)\n", bytes_to_mov,(MAX_ENVELOPE_SIZE- privptr->mtc_offset)); @@ -3505,13 +3505,13 @@ unpack_next: } #ifdef DEBUGMSG printk(KERN_INFO "%s: %s() received data \n", - dev->name,__FUNCTION__); + dev->name,__func__); if (p_env->packing == DO_PACKED) dumpit((char *)p_packd+sizeof(struct clawph),32); else dumpit((char *)p_this_ccw->p_buffer, 32); printk(KERN_INFO "%s: %s() bytelength %d \n", - dev->name,__FUNCTION__,bytes_to_mov); + dev->name,__func__,bytes_to_mov); #endif if (mtc_this_frm==0) { len_of_data=privptr->mtc_offset+bytes_to_mov; @@ -3530,13 +3530,13 @@ unpack_next: #ifdef DEBUGMSG printk(KERN_INFO "%s: %s() netif_" "rx(skb) completed \n", - dev->name,__FUNCTION__); + dev->name,__func__); #endif } else { privptr->stats.rx_dropped++; printk(KERN_WARNING "%s: %s() low on memory\n", - dev->name,__FUNCTION__); + dev->name,__func__); } privptr->mtc_offset=0; privptr->mtc_logical_link=-1; @@ -3575,10 +3575,10 @@ NextFrame: #ifdef IOTRACE printk(KERN_INFO "%s:%s processed frame is %d \n", - dev->name,__FUNCTION__,i); + dev->name,__func__,i); printk(KERN_INFO "%s:%s F:%lx L:%lx\n", dev->name, - __FUNCTION__, + __func__, (unsigned long)p_first_ccw, (unsigned long)p_last_ccw); #endif @@ -3588,7 +3588,7 @@ NextFrame: claw_strt_read(dev, LOCK_YES); #ifdef FUNCTRACE printk(KERN_INFO "%s: %s exit on line %d\n", - dev->name, __FUNCTION__, __LINE__); + dev->name, __func__, __LINE__); #endif return; } /* end of unpack_read */ @@ -3610,7 +3610,7 @@ claw_strt_read (struct net_device *dev, int lock ) p_ch=&privptr->channel[READ]; #ifdef FUNCTRACE - printk(KERN_INFO "%s:%s Enter \n",dev->name,__FUNCTION__); + printk(KERN_INFO "%s:%s Enter \n",dev->name,__func__); printk(KERN_INFO "%s: variable lock = %d, dev =\n",dev->name, lock); dumpit((char *) dev, sizeof(struct net_device)); #endif @@ -3626,7 +3626,7 @@ claw_strt_read (struct net_device *dev, int lock ) } #ifdef DEBUGMSG printk(KERN_INFO "%s:%s state-%02x\n" , - dev->name,__FUNCTION__, p_ch->claw_state); + dev->name,__func__, p_ch->claw_state); #endif if (lock==LOCK_YES) { spin_lock_irqsave(get_ccwdev_lock(p_ch->cdev), saveflags); @@ -3634,7 +3634,7 @@ claw_strt_read (struct net_device *dev, int lock ) if (test_and_set_bit(0, (void *)&p_ch->IO_active) == 0) { #ifdef DEBUGMSG printk(KERN_INFO "%s: HOT READ started in %s\n" , - dev->name,__FUNCTION__); + dev->name,__func__); p_clawh=(struct clawh *)privptr->p_claw_signal_blk; dumpit((char *)&p_clawh->flag , 1); #endif @@ -3650,7 +3650,7 @@ claw_strt_read (struct net_device *dev, int lock ) else { #ifdef DEBUGMSG printk(KERN_INFO "%s: No READ started by %s() In progress\n" , - dev->name,__FUNCTION__); + dev->name,__func__); #endif CLAW_DBF_TEXT(2,trace,"ReadAct"); } @@ -3660,7 +3660,7 @@ claw_strt_read (struct net_device *dev, int lock ) } #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif CLAW_DBF_TEXT(4,trace,"StRdExit"); return; @@ -3681,7 +3681,7 @@ claw_strt_out_IO( struct net_device *dev ) struct ccwbk *p_first_ccw; #ifdef FUNCTRACE - printk(KERN_INFO "%s:%s Enter\n",dev->name,__FUNCTION__); + printk(KERN_INFO "%s:%s Enter\n",dev->name,__func__); #endif if (!dev) { return; @@ -3691,7 +3691,7 @@ claw_strt_out_IO( struct net_device *dev ) #ifdef DEBUGMSG printk(KERN_INFO "%s:%s state-%02x\n" , - dev->name,__FUNCTION__,p_ch->claw_state); + dev->name,__func__,p_ch->claw_state); #endif CLAW_DBF_TEXT(4,trace,"strt_io"); p_first_ccw=privptr->p_write_active_first; @@ -3701,14 +3701,14 @@ claw_strt_out_IO( struct net_device *dev ) if (p_first_ccw == NULL) { #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif return; } if (test_and_set_bit(0, (void *)&p_ch->IO_active) == 0) { parm = (unsigned long) p_ch; #ifdef DEBUGMSG - printk(KERN_INFO "%s:%s do_io \n" ,dev->name,__FUNCTION__); + printk(KERN_INFO "%s:%s do_io \n" ,dev->name,__func__); dumpit((char *)p_first_ccw, sizeof(struct ccwbk)); #endif CLAW_DBF_TEXT(2,trace,"StWrtIO"); @@ -3721,7 +3721,7 @@ claw_strt_out_IO( struct net_device *dev ) dev->trans_start = jiffies; #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d\n", - dev->name,__FUNCTION__,__LINE__); + dev->name,__func__,__LINE__); #endif return; @@ -3745,7 +3745,7 @@ claw_free_wrt_buf( struct net_device *dev ) struct ccwbk*p_buf; #endif #ifdef FUNCTRACE - printk(KERN_INFO "%s:%s Enter\n",dev->name,__FUNCTION__); + printk(KERN_INFO "%s:%s Enter\n",dev->name,__func__); printk(KERN_INFO "%s: free count = %d variable dev =\n", dev->name,privptr->write_free_count); #endif @@ -3798,7 +3798,7 @@ claw_free_wrt_buf( struct net_device *dev ) privptr->p_write_active_last=NULL; #ifdef DEBUGMSG printk(KERN_INFO "%s:%s p_write_" - "active_first==NULL\n",dev->name,__FUNCTION__); + "active_first==NULL\n",dev->name,__func__); #endif } #ifdef IOTRACE @@ -3819,7 +3819,7 @@ claw_free_wrt_buf( struct net_device *dev ) CLAW_DBF_TEXT_(4,trace,"FWC=%d",privptr->write_free_count); #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d free_count =%d\n", - dev->name,__FUNCTION__, __LINE__,privptr->write_free_count); + dev->name,__func__, __LINE__,privptr->write_free_count); #endif return; } @@ -3833,7 +3833,7 @@ claw_free_netdevice(struct net_device * dev, int free_dev) { struct claw_privbk *privptr; #ifdef FUNCTRACE - printk(KERN_INFO "%s:%s Enter\n",dev->name,__FUNCTION__); + printk(KERN_INFO "%s:%s Enter\n",dev->name,__func__); #endif CLAW_DBF_TEXT(2,setup,"free_dev"); @@ -3854,7 +3854,7 @@ claw_free_netdevice(struct net_device * dev, int free_dev) #endif CLAW_DBF_TEXT(2,setup,"feee_ok"); #ifdef FUNCTRACE - printk(KERN_INFO "%s:%s Exit\n",dev->name,__FUNCTION__); + printk(KERN_INFO "%s:%s Exit\n",dev->name,__func__); #endif } @@ -3867,13 +3867,13 @@ static void claw_init_netdevice(struct net_device * dev) { #ifdef FUNCTRACE - printk(KERN_INFO "%s:%s Enter\n",dev->name,__FUNCTION__); + printk(KERN_INFO "%s:%s Enter\n",dev->name,__func__); #endif CLAW_DBF_TEXT(2,setup,"init_dev"); CLAW_DBF_TEXT_(2,setup,"%s",dev->name); if (!dev) { printk(KERN_WARNING "claw:%s BAD Device exit line %d\n", - __FUNCTION__,__LINE__); + __func__,__LINE__); CLAW_DBF_TEXT(2,setup,"baddev"); return; } @@ -3889,7 +3889,7 @@ claw_init_netdevice(struct net_device * dev) dev->tx_queue_len = 1300; dev->flags = IFF_POINTOPOINT | IFF_NOARP; #ifdef FUNCTRACE - printk(KERN_INFO "%s:%s Exit\n",dev->name,__FUNCTION__); + printk(KERN_INFO "%s:%s Exit\n",dev->name,__func__); #endif CLAW_DBF_TEXT(2,setup,"initok"); return; @@ -3909,7 +3909,7 @@ add_channel(struct ccw_device *cdev,int i,struct claw_privbk *privptr) struct ccw_dev_id dev_id; #ifdef FUNCTRACE - printk(KERN_INFO "%s:%s Enter\n",cdev->dev.bus_id,__FUNCTION__); + printk(KERN_INFO "%s:%s Enter\n",cdev->dev.bus_id,__func__); #endif CLAW_DBF_TEXT_(2,setup,"%s",cdev->dev.bus_id); privptr->channel[i].flag = i+1; /* Read is 1 Write is 2 */ @@ -3920,16 +3920,16 @@ add_channel(struct ccw_device *cdev,int i,struct claw_privbk *privptr) p_ch->devno = dev_id.devno; if ((p_ch->irb = kzalloc(sizeof (struct irb),GFP_KERNEL)) == NULL) { printk(KERN_WARNING "%s Out of memory in %s for irb\n", - p_ch->id,__FUNCTION__); + p_ch->id,__func__); #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d\n", - p_ch->id,__FUNCTION__,__LINE__); + p_ch->id,__func__,__LINE__); #endif return -ENOMEM; } #ifdef FUNCTRACE printk(KERN_INFO "%s:%s Exit on line %d\n", - cdev->dev.bus_id,__FUNCTION__,__LINE__); + cdev->dev.bus_id,__func__,__LINE__); #endif return 0; } @@ -3952,7 +3952,7 @@ claw_new_device(struct ccwgroup_device *cgdev) int ret; struct ccw_dev_id dev_id; - pr_debug("%s() called\n", __FUNCTION__); + pr_debug("%s() called\n", __func__); printk(KERN_INFO "claw: add for %s\n",cgdev->cdev[READ]->dev.bus_id); CLAW_DBF_TEXT(2,setup,"new_dev"); privptr = cgdev->dev.driver_data; @@ -3990,7 +3990,7 @@ claw_new_device(struct ccwgroup_device *cgdev) } dev = alloc_netdev(0,"claw%d",claw_init_netdevice); if (!dev) { - printk(KERN_WARNING "%s:alloc_netdev failed\n",__FUNCTION__); + printk(KERN_WARNING "%s:alloc_netdev failed\n",__func__); goto out; } dev->priv = privptr; @@ -4065,7 +4065,7 @@ claw_shutdown_device(struct ccwgroup_device *cgdev) struct net_device *ndev; int ret; - pr_debug("%s() called\n", __FUNCTION__); + pr_debug("%s() called\n", __func__); CLAW_DBF_TEXT_(2,setup,"%s",cgdev->dev.bus_id); priv = cgdev->dev.driver_data; if (!priv) @@ -4095,15 +4095,15 @@ claw_remove_device(struct ccwgroup_device *cgdev) { struct claw_privbk *priv; - pr_debug("%s() called\n", __FUNCTION__); + pr_debug("%s() called\n", __func__); CLAW_DBF_TEXT_(2,setup,"%s",cgdev->dev.bus_id); priv = cgdev->dev.driver_data; if (!priv) { - printk(KERN_WARNING "claw: %s() no Priv exiting\n",__FUNCTION__); + printk(KERN_WARNING "claw: %s() no Priv exiting\n",__func__); return; } printk(KERN_INFO "claw: %s() called %s will be removed.\n", - __FUNCTION__,cgdev->cdev[0]->dev.bus_id); + __func__,cgdev->cdev[0]->dev.bus_id); if (cgdev->state == CCWGROUP_ONLINE) claw_shutdown_device(cgdev); claw_remove_files(&cgdev->dev); @@ -4346,7 +4346,7 @@ static struct attribute_group claw_attr_group = { static int claw_add_files(struct device *dev) { - pr_debug("%s() called\n", __FUNCTION__); + pr_debug("%s() called\n", __func__); CLAW_DBF_TEXT(2,setup,"add_file"); return sysfs_create_group(&dev->kobj, &claw_attr_group); } @@ -4354,7 +4354,7 @@ claw_add_files(struct device *dev) static void claw_remove_files(struct device *dev) { - pr_debug("%s() called\n", __FUNCTION__); + pr_debug("%s() called\n", __func__); CLAW_DBF_TEXT(2,setup,"rem_file"); sysfs_remove_group(&dev->kobj, &claw_attr_group); } @@ -4385,12 +4385,12 @@ claw_init(void) printk(KERN_INFO "claw: starting driver\n"); #ifdef FUNCTRACE - printk(KERN_INFO "claw: %s() enter \n",__FUNCTION__); + printk(KERN_INFO "claw: %s() enter \n",__func__); #endif ret = claw_register_debug_facility(); if (ret) { printk(KERN_WARNING "claw: %s() debug_register failed %d\n", - __FUNCTION__,ret); + __func__,ret); return ret; } CLAW_DBF_TEXT(2,setup,"init_mod"); @@ -4398,10 +4398,10 @@ claw_init(void) if (ret) { claw_unregister_debug_facility(); printk(KERN_WARNING "claw; %s() cu3088 register failed %d\n", - __FUNCTION__,ret); + __func__,ret); } #ifdef FUNCTRACE - printk(KERN_INFO "claw: %s() exit \n",__FUNCTION__); + printk(KERN_INFO "claw: %s() exit \n",__func__); #endif return ret; } diff --git a/drivers/s390/net/netiucv.c b/drivers/s390/net/netiucv.c index 874a19994489..8f876f6ab367 100644 --- a/drivers/s390/net/netiucv.c +++ b/drivers/s390/net/netiucv.c @@ -670,7 +670,7 @@ static void conn_action_rx(fsm_instance *fi, int event, void *arg) struct netiucv_priv *privptr = netdev_priv(conn->netdev); int rc; - IUCV_DBF_TEXT(trace, 4, __FUNCTION__); + IUCV_DBF_TEXT(trace, 4, __func__); if (!conn->netdev) { iucv_message_reject(conn->path, msg); @@ -718,7 +718,7 @@ static void conn_action_txdone(fsm_instance *fi, int event, void *arg) struct ll_header header; int rc; - IUCV_DBF_TEXT(trace, 4, __FUNCTION__); + IUCV_DBF_TEXT(trace, 4, __func__); if (conn && conn->netdev) privptr = netdev_priv(conn->netdev); @@ -799,7 +799,7 @@ static void conn_action_connaccept(fsm_instance *fi, int event, void *arg) struct netiucv_priv *privptr = netdev_priv(netdev); int rc; - IUCV_DBF_TEXT(trace, 3, __FUNCTION__); + IUCV_DBF_TEXT(trace, 3, __func__); conn->path = path; path->msglim = NETIUCV_QUEUELEN_DEFAULT; @@ -821,7 +821,7 @@ static void conn_action_connreject(fsm_instance *fi, int event, void *arg) struct iucv_event *ev = arg; struct iucv_path *path = ev->data; - IUCV_DBF_TEXT(trace, 3, __FUNCTION__); + IUCV_DBF_TEXT(trace, 3, __func__); iucv_path_sever(path, NULL); } @@ -831,7 +831,7 @@ static void conn_action_connack(fsm_instance *fi, int event, void *arg) struct net_device *netdev = conn->netdev; struct netiucv_priv *privptr = netdev_priv(netdev); - IUCV_DBF_TEXT(trace, 3, __FUNCTION__); + IUCV_DBF_TEXT(trace, 3, __func__); fsm_deltimer(&conn->timer); fsm_newstate(fi, CONN_STATE_IDLE); netdev->tx_queue_len = conn->path->msglim; @@ -842,7 +842,7 @@ static void conn_action_conntimsev(fsm_instance *fi, int event, void *arg) { struct iucv_connection *conn = arg; - IUCV_DBF_TEXT(trace, 3, __FUNCTION__); + IUCV_DBF_TEXT(trace, 3, __func__); fsm_deltimer(&conn->timer); iucv_path_sever(conn->path, NULL); fsm_newstate(fi, CONN_STATE_STARTWAIT); @@ -854,7 +854,7 @@ static void conn_action_connsever(fsm_instance *fi, int event, void *arg) struct net_device *netdev = conn->netdev; struct netiucv_priv *privptr = netdev_priv(netdev); - IUCV_DBF_TEXT(trace, 3, __FUNCTION__); + IUCV_DBF_TEXT(trace, 3, __func__); fsm_deltimer(&conn->timer); iucv_path_sever(conn->path, NULL); @@ -870,7 +870,7 @@ static void conn_action_start(fsm_instance *fi, int event, void *arg) struct iucv_connection *conn = arg; int rc; - IUCV_DBF_TEXT(trace, 3, __FUNCTION__); + IUCV_DBF_TEXT(trace, 3, __func__); fsm_newstate(fi, CONN_STATE_STARTWAIT); PRINT_DEBUG("%s('%s'): connecting ...\n", @@ -948,7 +948,7 @@ static void conn_action_stop(fsm_instance *fi, int event, void *arg) struct net_device *netdev = conn->netdev; struct netiucv_priv *privptr = netdev_priv(netdev); - IUCV_DBF_TEXT(trace, 3, __FUNCTION__); + IUCV_DBF_TEXT(trace, 3, __func__); fsm_deltimer(&conn->timer); fsm_newstate(fi, CONN_STATE_STOPPED); @@ -1024,7 +1024,7 @@ static void dev_action_start(fsm_instance *fi, int event, void *arg) struct net_device *dev = arg; struct netiucv_priv *privptr = netdev_priv(dev); - IUCV_DBF_TEXT(trace, 3, __FUNCTION__); + IUCV_DBF_TEXT(trace, 3, __func__); fsm_newstate(fi, DEV_STATE_STARTWAIT); fsm_event(privptr->conn->fsm, CONN_EVENT_START, privptr->conn); @@ -1044,7 +1044,7 @@ dev_action_stop(fsm_instance *fi, int event, void *arg) struct netiucv_priv *privptr = netdev_priv(dev); struct iucv_event ev; - IUCV_DBF_TEXT(trace, 3, __FUNCTION__); + IUCV_DBF_TEXT(trace, 3, __func__); ev.conn = privptr->conn; @@ -1066,7 +1066,7 @@ dev_action_connup(fsm_instance *fi, int event, void *arg) struct net_device *dev = arg; struct netiucv_priv *privptr = netdev_priv(dev); - IUCV_DBF_TEXT(trace, 3, __FUNCTION__); + IUCV_DBF_TEXT(trace, 3, __func__); switch (fsm_getstate(fi)) { case DEV_STATE_STARTWAIT: @@ -1097,7 +1097,7 @@ dev_action_connup(fsm_instance *fi, int event, void *arg) static void dev_action_conndown(fsm_instance *fi, int event, void *arg) { - IUCV_DBF_TEXT(trace, 3, __FUNCTION__); + IUCV_DBF_TEXT(trace, 3, __func__); switch (fsm_getstate(fi)) { case DEV_STATE_RUNNING: @@ -1288,7 +1288,7 @@ static int netiucv_tx(struct sk_buff *skb, struct net_device *dev) struct netiucv_priv *privptr = netdev_priv(dev); int rc; - IUCV_DBF_TEXT(trace, 4, __FUNCTION__); + IUCV_DBF_TEXT(trace, 4, __func__); /** * Some sanity checks ... */ @@ -1344,7 +1344,7 @@ static struct net_device_stats *netiucv_stats (struct net_device * dev) { struct netiucv_priv *priv = netdev_priv(dev); - IUCV_DBF_TEXT(trace, 5, __FUNCTION__); + IUCV_DBF_TEXT(trace, 5, __func__); return &priv->stats; } @@ -1360,7 +1360,7 @@ static struct net_device_stats *netiucv_stats (struct net_device * dev) */ static int netiucv_change_mtu(struct net_device * dev, int new_mtu) { - IUCV_DBF_TEXT(trace, 3, __FUNCTION__); + IUCV_DBF_TEXT(trace, 3, __func__); if (new_mtu < 576 || new_mtu > NETIUCV_MTU_MAX) { IUCV_DBF_TEXT(setup, 2, "given MTU out of valid range\n"); return -EINVAL; @@ -1378,7 +1378,7 @@ static ssize_t user_show(struct device *dev, struct device_attribute *attr, { struct netiucv_priv *priv = dev->driver_data; - IUCV_DBF_TEXT(trace, 5, __FUNCTION__); + IUCV_DBF_TEXT(trace, 5, __func__); return sprintf(buf, "%s\n", netiucv_printname(priv->conn->userid)); } @@ -1393,7 +1393,7 @@ static ssize_t user_write(struct device *dev, struct device_attribute *attr, int i; struct iucv_connection *cp; - IUCV_DBF_TEXT(trace, 3, __FUNCTION__); + IUCV_DBF_TEXT(trace, 3, __func__); if (count > 9) { PRINT_WARN("netiucv: username too long (%d)!\n", (int) count); IUCV_DBF_TEXT_(setup, 2, @@ -1449,7 +1449,7 @@ static ssize_t buffer_show (struct device *dev, struct device_attribute *attr, char *buf) { struct netiucv_priv *priv = dev->driver_data; - IUCV_DBF_TEXT(trace, 5, __FUNCTION__); + IUCV_DBF_TEXT(trace, 5, __func__); return sprintf(buf, "%d\n", priv->conn->max_buffsize); } @@ -1461,7 +1461,7 @@ static ssize_t buffer_write (struct device *dev, struct device_attribute *attr, char *e; int bs1; - IUCV_DBF_TEXT(trace, 3, __FUNCTION__); + IUCV_DBF_TEXT(trace, 3, __func__); if (count >= 39) return -EINVAL; @@ -1513,7 +1513,7 @@ static ssize_t dev_fsm_show (struct device *dev, struct device_attribute *attr, { struct netiucv_priv *priv = dev->driver_data; - IUCV_DBF_TEXT(trace, 5, __FUNCTION__); + IUCV_DBF_TEXT(trace, 5, __func__); return sprintf(buf, "%s\n", fsm_getstate_str(priv->fsm)); } @@ -1524,7 +1524,7 @@ static ssize_t conn_fsm_show (struct device *dev, { struct netiucv_priv *priv = dev->driver_data; - IUCV_DBF_TEXT(trace, 5, __FUNCTION__); + IUCV_DBF_TEXT(trace, 5, __func__); return sprintf(buf, "%s\n", fsm_getstate_str(priv->conn->fsm)); } @@ -1535,7 +1535,7 @@ static ssize_t maxmulti_show (struct device *dev, { struct netiucv_priv *priv = dev->driver_data; - IUCV_DBF_TEXT(trace, 5, __FUNCTION__); + IUCV_DBF_TEXT(trace, 5, __func__); return sprintf(buf, "%ld\n", priv->conn->prof.maxmulti); } @@ -1545,7 +1545,7 @@ static ssize_t maxmulti_write (struct device *dev, { struct netiucv_priv *priv = dev->driver_data; - IUCV_DBF_TEXT(trace, 4, __FUNCTION__); + IUCV_DBF_TEXT(trace, 4, __func__); priv->conn->prof.maxmulti = 0; return count; } @@ -1557,7 +1557,7 @@ static ssize_t maxcq_show (struct device *dev, struct device_attribute *attr, { struct netiucv_priv *priv = dev->driver_data; - IUCV_DBF_TEXT(trace, 5, __FUNCTION__); + IUCV_DBF_TEXT(trace, 5, __func__); return sprintf(buf, "%ld\n", priv->conn->prof.maxcqueue); } @@ -1566,7 +1566,7 @@ static ssize_t maxcq_write (struct device *dev, struct device_attribute *attr, { struct netiucv_priv *priv = dev->driver_data; - IUCV_DBF_TEXT(trace, 4, __FUNCTION__); + IUCV_DBF_TEXT(trace, 4, __func__); priv->conn->prof.maxcqueue = 0; return count; } @@ -1578,7 +1578,7 @@ static ssize_t sdoio_show (struct device *dev, struct device_attribute *attr, { struct netiucv_priv *priv = dev->driver_data; - IUCV_DBF_TEXT(trace, 5, __FUNCTION__); + IUCV_DBF_TEXT(trace, 5, __func__); return sprintf(buf, "%ld\n", priv->conn->prof.doios_single); } @@ -1587,7 +1587,7 @@ static ssize_t sdoio_write (struct device *dev, struct device_attribute *attr, { struct netiucv_priv *priv = dev->driver_data; - IUCV_DBF_TEXT(trace, 4, __FUNCTION__); + IUCV_DBF_TEXT(trace, 4, __func__); priv->conn->prof.doios_single = 0; return count; } @@ -1599,7 +1599,7 @@ static ssize_t mdoio_show (struct device *dev, struct device_attribute *attr, { struct netiucv_priv *priv = dev->driver_data; - IUCV_DBF_TEXT(trace, 5, __FUNCTION__); + IUCV_DBF_TEXT(trace, 5, __func__); return sprintf(buf, "%ld\n", priv->conn->prof.doios_multi); } @@ -1608,7 +1608,7 @@ static ssize_t mdoio_write (struct device *dev, struct device_attribute *attr, { struct netiucv_priv *priv = dev->driver_data; - IUCV_DBF_TEXT(trace, 5, __FUNCTION__); + IUCV_DBF_TEXT(trace, 5, __func__); priv->conn->prof.doios_multi = 0; return count; } @@ -1620,7 +1620,7 @@ static ssize_t txlen_show (struct device *dev, struct device_attribute *attr, { struct netiucv_priv *priv = dev->driver_data; - IUCV_DBF_TEXT(trace, 5, __FUNCTION__); + IUCV_DBF_TEXT(trace, 5, __func__); return sprintf(buf, "%ld\n", priv->conn->prof.txlen); } @@ -1629,7 +1629,7 @@ static ssize_t txlen_write (struct device *dev, struct device_attribute *attr, { struct netiucv_priv *priv = dev->driver_data; - IUCV_DBF_TEXT(trace, 4, __FUNCTION__); + IUCV_DBF_TEXT(trace, 4, __func__); priv->conn->prof.txlen = 0; return count; } @@ -1641,7 +1641,7 @@ static ssize_t txtime_show (struct device *dev, struct device_attribute *attr, { struct netiucv_priv *priv = dev->driver_data; - IUCV_DBF_TEXT(trace, 5, __FUNCTION__); + IUCV_DBF_TEXT(trace, 5, __func__); return sprintf(buf, "%ld\n", priv->conn->prof.tx_time); } @@ -1650,7 +1650,7 @@ static ssize_t txtime_write (struct device *dev, struct device_attribute *attr, { struct netiucv_priv *priv = dev->driver_data; - IUCV_DBF_TEXT(trace, 4, __FUNCTION__); + IUCV_DBF_TEXT(trace, 4, __func__); priv->conn->prof.tx_time = 0; return count; } @@ -1662,7 +1662,7 @@ static ssize_t txpend_show (struct device *dev, struct device_attribute *attr, { struct netiucv_priv *priv = dev->driver_data; - IUCV_DBF_TEXT(trace, 5, __FUNCTION__); + IUCV_DBF_TEXT(trace, 5, __func__); return sprintf(buf, "%ld\n", priv->conn->prof.tx_pending); } @@ -1671,7 +1671,7 @@ static ssize_t txpend_write (struct device *dev, struct device_attribute *attr, { struct netiucv_priv *priv = dev->driver_data; - IUCV_DBF_TEXT(trace, 4, __FUNCTION__); + IUCV_DBF_TEXT(trace, 4, __func__); priv->conn->prof.tx_pending = 0; return count; } @@ -1683,7 +1683,7 @@ static ssize_t txmpnd_show (struct device *dev, struct device_attribute *attr, { struct netiucv_priv *priv = dev->driver_data; - IUCV_DBF_TEXT(trace, 5, __FUNCTION__); + IUCV_DBF_TEXT(trace, 5, __func__); return sprintf(buf, "%ld\n", priv->conn->prof.tx_max_pending); } @@ -1692,7 +1692,7 @@ static ssize_t txmpnd_write (struct device *dev, struct device_attribute *attr, { struct netiucv_priv *priv = dev->driver_data; - IUCV_DBF_TEXT(trace, 4, __FUNCTION__); + IUCV_DBF_TEXT(trace, 4, __func__); priv->conn->prof.tx_max_pending = 0; return count; } @@ -1732,7 +1732,7 @@ static int netiucv_add_files(struct device *dev) { int ret; - IUCV_DBF_TEXT(trace, 3, __FUNCTION__); + IUCV_DBF_TEXT(trace, 3, __func__); ret = sysfs_create_group(&dev->kobj, &netiucv_attr_group); if (ret) return ret; @@ -1744,7 +1744,7 @@ static int netiucv_add_files(struct device *dev) static void netiucv_remove_files(struct device *dev) { - IUCV_DBF_TEXT(trace, 3, __FUNCTION__); + IUCV_DBF_TEXT(trace, 3, __func__); sysfs_remove_group(&dev->kobj, &netiucv_stat_attr_group); sysfs_remove_group(&dev->kobj, &netiucv_attr_group); } @@ -1756,7 +1756,7 @@ static int netiucv_register_device(struct net_device *ndev) int ret; - IUCV_DBF_TEXT(trace, 3, __FUNCTION__); + IUCV_DBF_TEXT(trace, 3, __func__); if (dev) { snprintf(dev->bus_id, BUS_ID_SIZE, "net%s", ndev->name); @@ -1792,7 +1792,7 @@ out_unreg: static void netiucv_unregister_device(struct device *dev) { - IUCV_DBF_TEXT(trace, 3, __FUNCTION__); + IUCV_DBF_TEXT(trace, 3, __func__); netiucv_remove_files(dev); device_unregister(dev); } @@ -1857,7 +1857,7 @@ out: */ static void netiucv_remove_connection(struct iucv_connection *conn) { - IUCV_DBF_TEXT(trace, 3, __FUNCTION__); + IUCV_DBF_TEXT(trace, 3, __func__); write_lock_bh(&iucv_connection_rwlock); list_del_init(&conn->list); write_unlock_bh(&iucv_connection_rwlock); @@ -1881,7 +1881,7 @@ static void netiucv_free_netdevice(struct net_device *dev) { struct netiucv_priv *privptr = netdev_priv(dev); - IUCV_DBF_TEXT(trace, 3, __FUNCTION__); + IUCV_DBF_TEXT(trace, 3, __func__); if (!dev) return; @@ -1963,7 +1963,7 @@ static ssize_t conn_write(struct device_driver *drv, struct netiucv_priv *priv; struct iucv_connection *cp; - IUCV_DBF_TEXT(trace, 3, __FUNCTION__); + IUCV_DBF_TEXT(trace, 3, __func__); if (count>9) { PRINT_WARN("netiucv: username too long (%d)!\n", (int)count); IUCV_DBF_TEXT(setup, 2, "conn_write: too long\n"); @@ -2048,7 +2048,7 @@ static ssize_t remove_write (struct device_driver *drv, const char *p; int i; - IUCV_DBF_TEXT(trace, 3, __FUNCTION__); + IUCV_DBF_TEXT(trace, 3, __func__); if (count >= IFNAMSIZ) count = IFNAMSIZ - 1;; @@ -2116,7 +2116,7 @@ static void __exit netiucv_exit(void) struct netiucv_priv *priv; struct device *dev; - IUCV_DBF_TEXT(trace, 3, __FUNCTION__); + IUCV_DBF_TEXT(trace, 3, __func__); while (!list_empty(&iucv_connection_list)) { cp = list_entry(iucv_connection_list.next, struct iucv_connection, list); @@ -2146,8 +2146,7 @@ static int __init netiucv_init(void) rc = iucv_register(&netiucv_handler, 1); if (rc) goto out_dbf; - IUCV_DBF_TEXT(trace, 3, __FUNCTION__); - netiucv_driver.groups = netiucv_drv_attr_groups; + IUCV_DBF_TEXT(trace, 3, __func__); rc = driver_register(&netiucv_driver); if (rc) { PRINT_ERR("NETIUCV: failed to register driver.\n"); diff --git a/drivers/s390/s390mach.c b/drivers/s390/s390mach.c index 644a06eba828..4d4b54277c43 100644 --- a/drivers/s390/s390mach.c +++ b/drivers/s390/s390mach.c @@ -59,15 +59,15 @@ repeat: printk(KERN_WARNING"%s: Code does not support more " "than two chained crws; please report to " - "linux390@de.ibm.com!\n", __FUNCTION__); + "linux390@de.ibm.com!\n", __func__); ccode = stcrw(&tmp_crw); printk(KERN_WARNING"%s: crw reports slct=%d, oflw=%d, " "chn=%d, rsc=%X, anc=%d, erc=%X, rsid=%X\n", - __FUNCTION__, tmp_crw.slct, tmp_crw.oflw, + __func__, tmp_crw.slct, tmp_crw.oflw, tmp_crw.chn, tmp_crw.rsc, tmp_crw.anc, tmp_crw.erc, tmp_crw.rsid); printk(KERN_WARNING"%s: This was crw number %x in the " - "chain\n", __FUNCTION__, chain); + "chain\n", __func__, chain); if (ccode != 0) break; chain = tmp_crw.chn ? chain + 1 : 0; @@ -83,7 +83,7 @@ repeat: crw[chain].rsid); /* Check for overflows. */ if (crw[chain].oflw) { - pr_debug("%s: crw overflow detected!\n", __FUNCTION__); + pr_debug("%s: crw overflow detected!\n", __func__); css_schedule_eval_all(); chain = 0; continue; diff --git a/drivers/s390/s390mach.h b/drivers/s390/s390mach.h index d3ca4281a494..ca681f9b67fc 100644 --- a/drivers/s390/s390mach.h +++ b/drivers/s390/s390mach.h @@ -105,4 +105,8 @@ static inline int stcrw(struct crw *pcrw ) #define ED_ETR_SYNC 12 /* External damage ETR sync check */ #define ED_ETR_SWITCH 13 /* External damage ETR switch to local */ +struct pt_regs; + +void s390_handle_mcck(void); +void s390_do_machine_check(struct pt_regs *regs); #endif /* __s390mach */ diff --git a/drivers/s390/scsi/zfcp_def.h b/drivers/s390/scsi/zfcp_def.h index 9e9f6c1e4e5d..45a7cd98c140 100644 --- a/drivers/s390/scsi/zfcp_def.h +++ b/drivers/s390/scsi/zfcp_def.h @@ -539,7 +539,7 @@ struct zfcp_rc_entry { /* logging routine for zfcp */ #define _ZFCP_LOG(fmt, args...) \ - printk(KERN_ERR ZFCP_NAME": %s(%d): " fmt, __FUNCTION__, \ + printk(KERN_ERR ZFCP_NAME": %s(%d): " fmt, __func__, \ __LINE__ , ##args) #define ZFCP_LOG(level, fmt, args...) \ diff --git a/drivers/s390/sysinfo.c b/drivers/s390/sysinfo.c index 291ff6235fe2..c3e4ab07b9cc 100644 --- a/drivers/s390/sysinfo.c +++ b/drivers/s390/sysinfo.c @@ -11,111 +11,13 @@ #include <linux/init.h> #include <linux/delay.h> #include <asm/ebcdic.h> +#include <asm/sysinfo.h> /* Sigh, math-emu. Don't ask. */ #include <asm/sfp-util.h> #include <math-emu/soft-fp.h> #include <math-emu/single.h> -struct sysinfo_1_1_1 { - char reserved_0[32]; - char manufacturer[16]; - char type[4]; - char reserved_1[12]; - char model_capacity[16]; - char sequence[16]; - char plant[4]; - char model[16]; -}; - -struct sysinfo_1_2_1 { - char reserved_0[80]; - char sequence[16]; - char plant[4]; - char reserved_1[2]; - unsigned short cpu_address; -}; - -struct sysinfo_1_2_2 { - char format; - char reserved_0[1]; - unsigned short acc_offset; - char reserved_1[24]; - unsigned int secondary_capability; - unsigned int capability; - unsigned short cpus_total; - unsigned short cpus_configured; - unsigned short cpus_standby; - unsigned short cpus_reserved; - unsigned short adjustment[0]; -}; - -struct sysinfo_1_2_2_extension { - unsigned int alt_capability; - unsigned short alt_adjustment[0]; -}; - -struct sysinfo_2_2_1 { - char reserved_0[80]; - char sequence[16]; - char plant[4]; - unsigned short cpu_id; - unsigned short cpu_address; -}; - -struct sysinfo_2_2_2 { - char reserved_0[32]; - unsigned short lpar_number; - char reserved_1; - unsigned char characteristics; - unsigned short cpus_total; - unsigned short cpus_configured; - unsigned short cpus_standby; - unsigned short cpus_reserved; - char name[8]; - unsigned int caf; - char reserved_2[16]; - unsigned short cpus_dedicated; - unsigned short cpus_shared; -}; - -#define LPAR_CHAR_DEDICATED (1 << 7) -#define LPAR_CHAR_SHARED (1 << 6) -#define LPAR_CHAR_LIMITED (1 << 5) - -struct sysinfo_3_2_2 { - char reserved_0[31]; - unsigned char count; - struct { - char reserved_0[4]; - unsigned short cpus_total; - unsigned short cpus_configured; - unsigned short cpus_standby; - unsigned short cpus_reserved; - char name[8]; - unsigned int caf; - char cpi[16]; - char reserved_1[24]; - - } vm[8]; -}; - -static inline int stsi(void *sysinfo, int fc, int sel1, int sel2) -{ - register int r0 asm("0") = (fc << 28) | sel1; - register int r1 asm("1") = sel2; - - asm volatile( - " stsi 0(%2)\n" - "0: jz 2f\n" - "1: lhi %0,%3\n" - "2:\n" - EX_TABLE(0b,1b) - : "+d" (r0) : "d" (r1), "a" (sysinfo), "K" (-ENOSYS) - : "cc", "memory" ); - return r0; -} - static inline int stsi_0(void) { int rc = stsi (NULL, 0, 0, 0); @@ -133,6 +35,8 @@ static int stsi_1_1_1(struct sysinfo_1_1_1 *info, char *page, int len) EBCASC(info->sequence, sizeof(info->sequence)); EBCASC(info->plant, sizeof(info->plant)); EBCASC(info->model_capacity, sizeof(info->model_capacity)); + EBCASC(info->model_perm_cap, sizeof(info->model_perm_cap)); + EBCASC(info->model_temp_cap, sizeof(info->model_temp_cap)); len += sprintf(page + len, "Manufacturer: %-16.16s\n", info->manufacturer); len += sprintf(page + len, "Type: %-4.4s\n", @@ -155,8 +59,18 @@ static int stsi_1_1_1(struct sysinfo_1_1_1 *info, char *page, int len) info->sequence); len += sprintf(page + len, "Plant: %-4.4s\n", info->plant); - len += sprintf(page + len, "Model Capacity: %-16.16s\n", - info->model_capacity); + len += sprintf(page + len, "Model Capacity: %-16.16s %08u\n", + info->model_capacity, *(u32 *) info->model_cap_rating); + if (info->model_perm_cap[0] != '\0') + len += sprintf(page + len, + "Model Perm. Capacity: %-16.16s %08u\n", + info->model_perm_cap, + *(u32 *) info->model_perm_cap_rating); + if (info->model_temp_cap[0] != '\0') + len += sprintf(page + len, + "Model Temp. Capacity: %-16.16s %08u\n", + info->model_temp_cap, + *(u32 *) info->model_temp_cap_rating); return len; } diff --git a/include/asm-alpha/semaphore.h b/include/asm-alpha/semaphore.h index f1e9278a9fe2..d9b2034ed1d2 100644 --- a/include/asm-alpha/semaphore.h +++ b/include/asm-alpha/semaphore.h @@ -1,149 +1 @@ -#ifndef _ALPHA_SEMAPHORE_H -#define _ALPHA_SEMAPHORE_H - -/* - * SMP- and interrupt-safe semaphores.. - * - * (C) Copyright 1996 Linus Torvalds - * (C) Copyright 1996, 2000 Richard Henderson - */ - -#include <asm/current.h> -#include <asm/system.h> -#include <asm/atomic.h> -#include <linux/compiler.h> -#include <linux/wait.h> -#include <linux/rwsem.h> - -struct semaphore { - atomic_t count; - wait_queue_head_t wait; -}; - -#define __SEMAPHORE_INITIALIZER(name, n) \ -{ \ - .count = ATOMIC_INIT(n), \ - .wait = __WAIT_QUEUE_HEAD_INITIALIZER((name).wait), \ -} - -#define __DECLARE_SEMAPHORE_GENERIC(name,count) \ - struct semaphore name = __SEMAPHORE_INITIALIZER(name,count) - -#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name,1) - -static inline void sema_init(struct semaphore *sem, int val) -{ - /* - * Logically, - * *sem = (struct semaphore)__SEMAPHORE_INITIALIZER((*sem),val); - * except that gcc produces better initializing by parts yet. - */ - - atomic_set(&sem->count, val); - init_waitqueue_head(&sem->wait); -} - -static inline void init_MUTEX (struct semaphore *sem) -{ - sema_init(sem, 1); -} - -static inline void init_MUTEX_LOCKED (struct semaphore *sem) -{ - sema_init(sem, 0); -} - -extern void down(struct semaphore *); -extern void __down_failed(struct semaphore *); -extern int down_interruptible(struct semaphore *); -extern int __down_failed_interruptible(struct semaphore *); -extern int down_trylock(struct semaphore *); -extern void up(struct semaphore *); -extern void __up_wakeup(struct semaphore *); - -/* - * Hidden out of line code is fun, but extremely messy. Rely on newer - * compilers to do a respectable job with this. The contention cases - * are handled out of line in arch/alpha/kernel/semaphore.c. - */ - -static inline void __down(struct semaphore *sem) -{ - long count; - might_sleep(); - count = atomic_dec_return(&sem->count); - if (unlikely(count < 0)) - __down_failed(sem); -} - -static inline int __down_interruptible(struct semaphore *sem) -{ - long count; - might_sleep(); - count = atomic_dec_return(&sem->count); - if (unlikely(count < 0)) - return __down_failed_interruptible(sem); - return 0; -} - -/* - * down_trylock returns 0 on success, 1 if we failed to get the lock. - */ - -static inline int __down_trylock(struct semaphore *sem) -{ - long ret; - - /* "Equivalent" C: - - do { - ret = ldl_l; - --ret; - if (ret < 0) - break; - ret = stl_c = ret; - } while (ret == 0); - */ - __asm__ __volatile__( - "1: ldl_l %0,%1\n" - " subl %0,1,%0\n" - " blt %0,2f\n" - " stl_c %0,%1\n" - " beq %0,3f\n" - " mb\n" - "2:\n" - ".subsection 2\n" - "3: br 1b\n" - ".previous" - : "=&r" (ret), "=m" (sem->count) - : "m" (sem->count)); - - return ret < 0; -} - -static inline void __up(struct semaphore *sem) -{ - if (unlikely(atomic_inc_return(&sem->count) <= 0)) - __up_wakeup(sem); -} - -#if !defined(CONFIG_DEBUG_SEMAPHORE) -extern inline void down(struct semaphore *sem) -{ - __down(sem); -} -extern inline int down_interruptible(struct semaphore *sem) -{ - return __down_interruptible(sem); -} -extern inline int down_trylock(struct semaphore *sem) -{ - return __down_trylock(sem); -} -extern inline void up(struct semaphore *sem) -{ - __up(sem); -} -#endif - -#endif +#include <linux/semaphore.h> diff --git a/include/asm-arm/semaphore-helper.h b/include/asm-arm/semaphore-helper.h deleted file mode 100644 index 1d7f1987edb9..000000000000 --- a/include/asm-arm/semaphore-helper.h +++ /dev/null @@ -1,84 +0,0 @@ -#ifndef ASMARM_SEMAPHORE_HELPER_H -#define ASMARM_SEMAPHORE_HELPER_H - -/* - * These two _must_ execute atomically wrt each other. - */ -static inline void wake_one_more(struct semaphore * sem) -{ - unsigned long flags; - - spin_lock_irqsave(&semaphore_wake_lock, flags); - if (atomic_read(&sem->count) <= 0) - sem->waking++; - spin_unlock_irqrestore(&semaphore_wake_lock, flags); -} - -static inline int waking_non_zero(struct semaphore *sem) -{ - unsigned long flags; - int ret = 0; - - spin_lock_irqsave(&semaphore_wake_lock, flags); - if (sem->waking > 0) { - sem->waking--; - ret = 1; - } - spin_unlock_irqrestore(&semaphore_wake_lock, flags); - return ret; -} - -/* - * waking non zero interruptible - * 1 got the lock - * 0 go to sleep - * -EINTR interrupted - * - * We must undo the sem->count down_interruptible() increment while we are - * protected by the spinlock in order to make this atomic_inc() with the - * atomic_read() in wake_one_more(), otherwise we can race. -arca - */ -static inline int waking_non_zero_interruptible(struct semaphore *sem, - struct task_struct *tsk) -{ - unsigned long flags; - int ret = 0; - - spin_lock_irqsave(&semaphore_wake_lock, flags); - if (sem->waking > 0) { - sem->waking--; - ret = 1; - } else if (signal_pending(tsk)) { - atomic_inc(&sem->count); - ret = -EINTR; - } - spin_unlock_irqrestore(&semaphore_wake_lock, flags); - return ret; -} - -/* - * waking_non_zero_try_lock: - * 1 failed to lock - * 0 got the lock - * - * We must undo the sem->count down_interruptible() increment while we are - * protected by the spinlock in order to make this atomic_inc() with the - * atomic_read() in wake_one_more(), otherwise we can race. -arca - */ -static inline int waking_non_zero_trylock(struct semaphore *sem) -{ - unsigned long flags; - int ret = 1; - - spin_lock_irqsave(&semaphore_wake_lock, flags); - if (sem->waking <= 0) - atomic_inc(&sem->count); - else { - sem->waking--; - ret = 0; - } - spin_unlock_irqrestore(&semaphore_wake_lock, flags); - return ret; -} - -#endif diff --git a/include/asm-arm/semaphore.h b/include/asm-arm/semaphore.h index 1c8b441f89e3..d9b2034ed1d2 100644 --- a/include/asm-arm/semaphore.h +++ b/include/asm-arm/semaphore.h @@ -1,98 +1 @@ -/* - * linux/include/asm-arm/semaphore.h - */ -#ifndef __ASM_ARM_SEMAPHORE_H -#define __ASM_ARM_SEMAPHORE_H - -#include <linux/linkage.h> -#include <linux/spinlock.h> -#include <linux/wait.h> -#include <linux/rwsem.h> - -#include <asm/atomic.h> -#include <asm/locks.h> - -struct semaphore { - atomic_t count; - int sleepers; - wait_queue_head_t wait; -}; - -#define __SEMAPHORE_INIT(name, cnt) \ -{ \ - .count = ATOMIC_INIT(cnt), \ - .wait = __WAIT_QUEUE_HEAD_INITIALIZER((name).wait), \ -} - -#define __DECLARE_SEMAPHORE_GENERIC(name,count) \ - struct semaphore name = __SEMAPHORE_INIT(name,count) - -#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name,1) - -static inline void sema_init(struct semaphore *sem, int val) -{ - atomic_set(&sem->count, val); - sem->sleepers = 0; - init_waitqueue_head(&sem->wait); -} - -static inline void init_MUTEX(struct semaphore *sem) -{ - sema_init(sem, 1); -} - -static inline void init_MUTEX_LOCKED(struct semaphore *sem) -{ - sema_init(sem, 0); -} - -/* - * special register calling convention - */ -asmlinkage void __down_failed(void); -asmlinkage int __down_interruptible_failed(void); -asmlinkage int __down_trylock_failed(void); -asmlinkage void __up_wakeup(void); - -extern void __down(struct semaphore * sem); -extern int __down_interruptible(struct semaphore * sem); -extern int __down_trylock(struct semaphore * sem); -extern void __up(struct semaphore * sem); - -/* - * This is ugly, but we want the default case to fall through. - * "__down" is the actual routine that waits... - */ -static inline void down(struct semaphore * sem) -{ - might_sleep(); - __down_op(sem, __down_failed); -} - -/* - * This is ugly, but we want the default case to fall through. - * "__down_interruptible" is the actual routine that waits... - */ -static inline int down_interruptible (struct semaphore * sem) -{ - might_sleep(); - return __down_op_ret(sem, __down_interruptible_failed); -} - -static inline int down_trylock(struct semaphore *sem) -{ - return __down_op_ret(sem, __down_trylock_failed); -} - -/* - * Note! This is subtle. We jump to wake people up only if - * the semaphore was negative (== somebody was waiting on it). - * The default case (no contention) will result in NO - * jumps for both down() and up(). - */ -static inline void up(struct semaphore * sem) -{ - __up_op(sem, __up_wakeup); -} - -#endif +#include <linux/semaphore.h> diff --git a/include/asm-avr32/semaphore.h b/include/asm-avr32/semaphore.h index feaf1d453386..d9b2034ed1d2 100644 --- a/include/asm-avr32/semaphore.h +++ b/include/asm-avr32/semaphore.h @@ -1,108 +1 @@ -/* - * SMP- and interrupt-safe semaphores. - * - * Copyright (C) 2006 Atmel Corporation - * - * Based on include/asm-i386/semaphore.h - * Copyright (C) 1996 Linus Torvalds - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License version 2 as - * published by the Free Software Foundation. - */ -#ifndef __ASM_AVR32_SEMAPHORE_H -#define __ASM_AVR32_SEMAPHORE_H - -#include <linux/linkage.h> - -#include <asm/system.h> -#include <asm/atomic.h> -#include <linux/wait.h> -#include <linux/rwsem.h> - -struct semaphore { - atomic_t count; - int sleepers; - wait_queue_head_t wait; -}; - -#define __SEMAPHORE_INITIALIZER(name, n) \ -{ \ - .count = ATOMIC_INIT(n), \ - .wait = __WAIT_QUEUE_HEAD_INITIALIZER((name).wait) \ -} - -#define __DECLARE_SEMAPHORE_GENERIC(name,count) \ - struct semaphore name = __SEMAPHORE_INITIALIZER(name,count) - -#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name,1) - -static inline void sema_init (struct semaphore *sem, int val) -{ - atomic_set(&sem->count, val); - sem->sleepers = 0; - init_waitqueue_head(&sem->wait); -} - -static inline void init_MUTEX (struct semaphore *sem) -{ - sema_init(sem, 1); -} - -static inline void init_MUTEX_LOCKED (struct semaphore *sem) -{ - sema_init(sem, 0); -} - -void __down(struct semaphore * sem); -int __down_interruptible(struct semaphore * sem); -void __up(struct semaphore * sem); - -/* - * This is ugly, but we want the default case to fall through. - * "__down_failed" is a special asm handler that calls the C - * routine that actually waits. See arch/i386/kernel/semaphore.c - */ -static inline void down(struct semaphore * sem) -{ - might_sleep(); - if (unlikely(atomic_dec_return (&sem->count) < 0)) - __down (sem); -} - -/* - * Interruptible try to acquire a semaphore. If we obtained - * it, return zero. If we were interrupted, returns -EINTR - */ -static inline int down_interruptible(struct semaphore * sem) -{ - int ret = 0; - - might_sleep(); - if (unlikely(atomic_dec_return (&sem->count) < 0)) - ret = __down_interruptible (sem); - return ret; -} - -/* - * Non-blockingly attempt to down() a semaphore. - * Returns zero if we acquired it - */ -static inline int down_trylock(struct semaphore * sem) -{ - return atomic_dec_if_positive(&sem->count) < 0; -} - -/* - * Note! This is subtle. We jump to wake people up only if - * the semaphore was negative (== somebody was waiting on it). - * The default case (no contention) will result in NO - * jumps for both down() and up(). - */ -static inline void up(struct semaphore * sem) -{ - if (unlikely(atomic_inc_return (&sem->count) <= 0)) - __up (sem); -} - -#endif /*__ASM_AVR32_SEMAPHORE_H */ +#include <linux/semaphore.h> diff --git a/include/asm-blackfin/semaphore-helper.h b/include/asm-blackfin/semaphore-helper.h deleted file mode 100644 index 9082b0dc3eb5..000000000000 --- a/include/asm-blackfin/semaphore-helper.h +++ /dev/null @@ -1,82 +0,0 @@ -/* Based on M68K version, Lineo Inc. May 2001 */ - -#ifndef _BFIN_SEMAPHORE_HELPER_H -#define _BFIN_SEMAPHORE_HELPER_H - -/* - * SMP- and interrupt-safe semaphores helper functions. - * - * (C) Copyright 1996 Linus Torvalds - * - */ - -#include <asm/errno.h> - -/* - * These two _must_ execute atomically wrt each other. - */ -static inline void wake_one_more(struct semaphore *sem) -{ - atomic_inc(&sem->waking); -} - -static inline int waking_non_zero(struct semaphore *sem) -{ - int ret; - unsigned long flags = 0; - - spin_lock_irqsave(&semaphore_wake_lock, flags); - ret = 0; - if (atomic_read(&sem->waking) > 0) { - atomic_dec(&sem->waking); - ret = 1; - } - spin_unlock_irqrestore(&semaphore_wake_lock, flags); - return ret; -} - -/* - * waking_non_zero_interruptible: - * 1 got the lock - * 0 go to sleep - * -EINTR interrupted - */ -static inline int waking_non_zero_interruptible(struct semaphore *sem, - struct task_struct *tsk) -{ - int ret = 0; - unsigned long flags = 0; - - spin_lock_irqsave(&semaphore_wake_lock, flags); - if (atomic_read(&sem->waking) > 0) { - atomic_dec(&sem->waking); - ret = 1; - } else if (signal_pending(tsk)) { - atomic_inc(&sem->count); - ret = -EINTR; - } - spin_unlock_irqrestore(&semaphore_wake_lock, flags); - return ret; -} - -/* - * waking_non_zero_trylock: - * 1 failed to lock - * 0 got the lock - */ -static inline int waking_non_zero_trylock(struct semaphore *sem) -{ - int ret = 1; - unsigned long flags = 0; - - spin_lock_irqsave(&semaphore_wake_lock, flags); - if (atomic_read(&sem->waking) > 0) { - atomic_dec(&sem->waking); - ret = 0; - } else - atomic_inc(&sem->count); - spin_unlock_irqrestore(&semaphore_wake_lock, flags); - return ret; -} - -#endif /* _BFIN_SEMAPHORE_HELPER_H */ diff --git a/include/asm-blackfin/semaphore.h b/include/asm-blackfin/semaphore.h index 533f90fb2e4e..d9b2034ed1d2 100644 --- a/include/asm-blackfin/semaphore.h +++ b/include/asm-blackfin/semaphore.h @@ -1,105 +1 @@ -#ifndef _BFIN_SEMAPHORE_H -#define _BFIN_SEMAPHORE_H - -#ifndef __ASSEMBLY__ - -#include <linux/linkage.h> -#include <linux/wait.h> -#include <linux/spinlock.h> -#include <linux/rwsem.h> -#include <asm/atomic.h> - -/* - * Interrupt-safe semaphores.. - * - * (C) Copyright 1996 Linus Torvalds - * - * BFIN version by akbar hussain Lineo Inc April 2001 - * - */ - -struct semaphore { - atomic_t count; - int sleepers; - wait_queue_head_t wait; -}; - -#define __SEMAPHORE_INITIALIZER(name, n) \ -{ \ - .count = ATOMIC_INIT(n), \ - .sleepers = 0, \ - .wait = __WAIT_QUEUE_HEAD_INITIALIZER((name).wait) \ -} - -#define __DECLARE_SEMAPHORE_GENERIC(name,count) \ - struct semaphore name = __SEMAPHORE_INITIALIZER(name,count) - -#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name,1) - -static inline void sema_init(struct semaphore *sem, int val) -{ - *sem = (struct semaphore)__SEMAPHORE_INITIALIZER(*sem, val); -} - -static inline void init_MUTEX(struct semaphore *sem) -{ - sema_init(sem, 1); -} - -static inline void init_MUTEX_LOCKED(struct semaphore *sem) -{ - sema_init(sem, 0); -} - -asmlinkage void __down(struct semaphore *sem); -asmlinkage int __down_interruptible(struct semaphore *sem); -asmlinkage int __down_trylock(struct semaphore *sem); -asmlinkage void __up(struct semaphore *sem); - -extern spinlock_t semaphore_wake_lock; - -/* - * This is ugly, but we want the default case to fall through. - * "down_failed" is a special asm handler that calls the C - * routine that actually waits. - */ -static inline void down(struct semaphore *sem) -{ - might_sleep(); - if (atomic_dec_return(&sem->count) < 0) - __down(sem); -} - -static inline int down_interruptible(struct semaphore *sem) -{ - int ret = 0; - - might_sleep(); - if (atomic_dec_return(&sem->count) < 0) - ret = __down_interruptible(sem); - return (ret); -} - -static inline int down_trylock(struct semaphore *sem) -{ - int ret = 0; - - if (atomic_dec_return(&sem->count) < 0) - ret = __down_trylock(sem); - return ret; -} - -/* - * Note! This is subtle. We jump to wake people up only if - * the semaphore was negative (== somebody was waiting on it). - * The default case (no contention) will result in NO - * jumps for both down() and up(). - */ -static inline void up(struct semaphore *sem) -{ - if (atomic_inc_return(&sem->count) <= 0) - __up(sem); -} - -#endif /* __ASSEMBLY__ */ -#endif /* _BFIN_SEMAPHORE_H */ +#include <linux/semaphore.h> diff --git a/include/asm-cris/semaphore-helper.h b/include/asm-cris/semaphore-helper.h deleted file mode 100644 index 27bfeca1b981..000000000000 --- a/include/asm-cris/semaphore-helper.h +++ /dev/null @@ -1,78 +0,0 @@ -/* $Id: semaphore-helper.h,v 1.3 2001/03/26 15:00:33 orjanf Exp $ - * - * SMP- and interrupt-safe semaphores helper functions. Generic versions, no - * optimizations whatsoever... - * - */ - -#ifndef _ASM_SEMAPHORE_HELPER_H -#define _ASM_SEMAPHORE_HELPER_H - -#include <asm/atomic.h> -#include <linux/errno.h> - -#define read(a) ((a)->counter) -#define inc(a) (((a)->counter)++) -#define dec(a) (((a)->counter)--) - -#define count_inc(a) ((*(a))++) - -/* - * These two _must_ execute atomically wrt each other. - */ -static inline void wake_one_more(struct semaphore * sem) -{ - atomic_inc(&sem->waking); -} - -static inline int waking_non_zero(struct semaphore *sem) -{ - unsigned long flags; - int ret = 0; - - local_irq_save(flags); - if (read(&sem->waking) > 0) { - dec(&sem->waking); - ret = 1; - } - local_irq_restore(flags); - return ret; -} - -static inline int waking_non_zero_interruptible(struct semaphore *sem, - struct task_struct *tsk) -{ - int ret = 0; - unsigned long flags; - - local_irq_save(flags); - if (read(&sem->waking) > 0) { - dec(&sem->waking); - ret = 1; - } else if (signal_pending(tsk)) { - inc(&sem->count); - ret = -EINTR; - } - local_irq_restore(flags); - return ret; -} - -static inline int waking_non_zero_trylock(struct semaphore *sem) -{ - int ret = 1; - unsigned long flags; - - local_irq_save(flags); - if (read(&sem->waking) <= 0) - inc(&sem->count); - else { - dec(&sem->waking); - ret = 0; - } - local_irq_restore(flags); - return ret; -} - -#endif /* _ASM_SEMAPHORE_HELPER_H */ - - diff --git a/include/asm-cris/semaphore.h b/include/asm-cris/semaphore.h index 31a4ac448195..d9b2034ed1d2 100644 --- a/include/asm-cris/semaphore.h +++ b/include/asm-cris/semaphore.h @@ -1,133 +1 @@ -/* $Id: semaphore.h,v 1.3 2001/05/08 13:54:09 bjornw Exp $ */ - -/* On the i386 these are coded in asm, perhaps we should as well. Later.. */ - -#ifndef _CRIS_SEMAPHORE_H -#define _CRIS_SEMAPHORE_H - -#define RW_LOCK_BIAS 0x01000000 - -#include <linux/wait.h> -#include <linux/spinlock.h> -#include <linux/rwsem.h> - -#include <asm/system.h> -#include <asm/atomic.h> - -/* - * CRIS semaphores, implemented in C-only so far. - */ - -struct semaphore { - atomic_t count; - atomic_t waking; - wait_queue_head_t wait; -}; - -#define __SEMAPHORE_INITIALIZER(name, n) \ -{ \ - .count = ATOMIC_INIT(n), \ - .waking = ATOMIC_INIT(0), \ - .wait = __WAIT_QUEUE_HEAD_INITIALIZER((name).wait) \ -} - -#define __DECLARE_SEMAPHORE_GENERIC(name,count) \ - struct semaphore name = __SEMAPHORE_INITIALIZER(name,count) - -#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name,1) - -static inline void sema_init(struct semaphore *sem, int val) -{ - *sem = (struct semaphore)__SEMAPHORE_INITIALIZER((*sem),val); -} - -static inline void init_MUTEX (struct semaphore *sem) -{ - sema_init(sem, 1); -} - -static inline void init_MUTEX_LOCKED (struct semaphore *sem) -{ - sema_init(sem, 0); -} - -extern void __down(struct semaphore * sem); -extern int __down_interruptible(struct semaphore * sem); -extern int __down_trylock(struct semaphore * sem); -extern void __up(struct semaphore * sem); - -/* notice - we probably can do cli/sti here instead of saving */ - -static inline void down(struct semaphore * sem) -{ - unsigned long flags; - int failed; - - might_sleep(); - - /* atomically decrement the semaphores count, and if its negative, we wait */ - cris_atomic_save(sem, flags); - failed = --(sem->count.counter) < 0; - cris_atomic_restore(sem, flags); - if(failed) { - __down(sem); - } -} - -/* - * This version waits in interruptible state so that the waiting - * process can be killed. The down_interruptible routine - * returns negative for signalled and zero for semaphore acquired. - */ - -static inline int down_interruptible(struct semaphore * sem) -{ - unsigned long flags; - int failed; - - might_sleep(); - - /* atomically decrement the semaphores count, and if its negative, we wait */ - cris_atomic_save(sem, flags); - failed = --(sem->count.counter) < 0; - cris_atomic_restore(sem, flags); - if(failed) - failed = __down_interruptible(sem); - return(failed); -} - -static inline int down_trylock(struct semaphore * sem) -{ - unsigned long flags; - int failed; - - cris_atomic_save(sem, flags); - failed = --(sem->count.counter) < 0; - cris_atomic_restore(sem, flags); - if(failed) - failed = __down_trylock(sem); - return(failed); - -} - -/* - * Note! This is subtle. We jump to wake people up only if - * the semaphore was negative (== somebody was waiting on it). - * The default case (no contention) will result in NO - * jumps for both down() and up(). - */ -static inline void up(struct semaphore * sem) -{ - unsigned long flags; - int wakeup; - - /* atomically increment the semaphores count, and if it was negative, we wake people */ - cris_atomic_save(sem, flags); - wakeup = ++(sem->count.counter) <= 0; - cris_atomic_restore(sem, flags); - if(wakeup) { - __up(sem); - } -} - -#endif +#include <linux/semaphore.h> diff --git a/include/asm-frv/semaphore.h b/include/asm-frv/semaphore.h index d7aaa1911a1a..d9b2034ed1d2 100644 --- a/include/asm-frv/semaphore.h +++ b/include/asm-frv/semaphore.h @@ -1,155 +1 @@ -/* semaphore.h: semaphores for the FR-V - * - * Copyright (C) 2003 Red Hat, Inc. All Rights Reserved. - * Written by David Howells (dhowells@redhat.com) - * - * This program is free software; you can redistribute it and/or - * modify it under the terms of the GNU General Public License - * as published by the Free Software Foundation; either version - * 2 of the License, or (at your option) any later version. - */ -#ifndef _ASM_SEMAPHORE_H -#define _ASM_SEMAPHORE_H - -#define RW_LOCK_BIAS 0x01000000 - -#ifndef __ASSEMBLY__ - -#include <linux/linkage.h> -#include <linux/wait.h> -#include <linux/spinlock.h> -#include <linux/rwsem.h> - -/* - * the semaphore definition - * - if counter is >0 then there are tokens available on the semaphore for down to collect - * - if counter is <=0 then there are no spare tokens, and anyone that wants one must wait - * - if wait_list is not empty, then there are processes waiting for the semaphore - */ -struct semaphore { - unsigned counter; - spinlock_t wait_lock; - struct list_head wait_list; -#ifdef CONFIG_DEBUG_SEMAPHORE - unsigned __magic; -#endif -}; - -#ifdef CONFIG_DEBUG_SEMAPHORE -# define __SEM_DEBUG_INIT(name) , (long)&(name).__magic -#else -# define __SEM_DEBUG_INIT(name) -#endif - - -#define __SEMAPHORE_INITIALIZER(name,count) \ -{ count, SPIN_LOCK_UNLOCKED, LIST_HEAD_INIT((name).wait_list) __SEM_DEBUG_INIT(name) } - -#define __DECLARE_SEMAPHORE_GENERIC(name,count) \ - struct semaphore name = __SEMAPHORE_INITIALIZER(name,count) - -#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name,1) - -static inline void sema_init (struct semaphore *sem, int val) -{ - *sem = (struct semaphore) __SEMAPHORE_INITIALIZER(*sem, val); -} - -static inline void init_MUTEX (struct semaphore *sem) -{ - sema_init(sem, 1); -} - -static inline void init_MUTEX_LOCKED (struct semaphore *sem) -{ - sema_init(sem, 0); -} - -extern void __down(struct semaphore *sem, unsigned long flags); -extern int __down_interruptible(struct semaphore *sem, unsigned long flags); -extern void __up(struct semaphore *sem); - -static inline void down(struct semaphore *sem) -{ - unsigned long flags; - -#ifdef CONFIG_DEBUG_SEMAPHORE - CHECK_MAGIC(sem->__magic); -#endif - - spin_lock_irqsave(&sem->wait_lock, flags); - if (likely(sem->counter > 0)) { - sem->counter--; - spin_unlock_irqrestore(&sem->wait_lock, flags); - } - else { - __down(sem, flags); - } -} - -static inline int down_interruptible(struct semaphore *sem) -{ - unsigned long flags; - int ret = 0; - -#ifdef CONFIG_DEBUG_SEMAPHORE - CHECK_MAGIC(sem->__magic); -#endif - - spin_lock_irqsave(&sem->wait_lock, flags); - if (likely(sem->counter > 0)) { - sem->counter--; - spin_unlock_irqrestore(&sem->wait_lock, flags); - } - else { - ret = __down_interruptible(sem, flags); - } - return ret; -} - -/* - * non-blockingly attempt to down() a semaphore. - * - returns zero if we acquired it - */ -static inline int down_trylock(struct semaphore *sem) -{ - unsigned long flags; - int success = 0; - -#ifdef CONFIG_DEBUG_SEMAPHORE - CHECK_MAGIC(sem->__magic); -#endif - - spin_lock_irqsave(&sem->wait_lock, flags); - if (sem->counter > 0) { - sem->counter--; - success = 1; - } - spin_unlock_irqrestore(&sem->wait_lock, flags); - return !success; -} - -static inline void up(struct semaphore *sem) -{ - unsigned long flags; - -#ifdef CONFIG_DEBUG_SEMAPHORE - CHECK_MAGIC(sem->__magic); -#endif - - spin_lock_irqsave(&sem->wait_lock, flags); - if (!list_empty(&sem->wait_list)) - __up(sem); - else - sem->counter++; - spin_unlock_irqrestore(&sem->wait_lock, flags); -} - -static inline int sem_getcount(struct semaphore *sem) -{ - return sem->counter; -} - -#endif /* __ASSEMBLY__ */ - -#endif +#include <linux/semaphore.h> diff --git a/include/asm-h8300/semaphore-helper.h b/include/asm-h8300/semaphore-helper.h deleted file mode 100644 index 4fea36be5fd8..000000000000 --- a/include/asm-h8300/semaphore-helper.h +++ /dev/null @@ -1,85 +0,0 @@ -#ifndef _H8300_SEMAPHORE_HELPER_H -#define _H8300_SEMAPHORE_HELPER_H - -/* - * SMP- and interrupt-safe semaphores helper functions. - * - * (C) Copyright 1996 Linus Torvalds - * - * based on - * m68k version by Andreas Schwab - */ - -#include <linux/errno.h> - -/* - * These two _must_ execute atomically wrt each other. - */ -static inline void wake_one_more(struct semaphore * sem) -{ - atomic_inc((atomic_t *)&sem->sleepers); -} - -static inline int waking_non_zero(struct semaphore *sem) -{ - int ret; - unsigned long flags; - - spin_lock_irqsave(&semaphore_wake_lock, flags); - ret = 0; - if (sem->sleepers > 0) { - sem->sleepers--; - ret = 1; - } - spin_unlock_irqrestore(&semaphore_wake_lock, flags); - return ret; -} - -/* - * waking_non_zero_interruptible: - * 1 got the lock - * 0 go to sleep - * -EINTR interrupted - */ -static inline int waking_non_zero_interruptible(struct semaphore *sem, - struct task_struct *tsk) -{ - int ret; - unsigned long flags; - - spin_lock_irqsave(&semaphore_wake_lock, flags); - ret = 0; - if (sem->sleepers > 0) { - sem->sleepers--; - ret = 1; - } else if (signal_pending(tsk)) { - atomic_inc(&sem->count); - ret = -EINTR; - } - spin_unlock_irqrestore(&semaphore_wake_lock, flags); - return ret; -} - -/* - * waking_non_zero_trylock: - * 1 failed to lock - * 0 got the lock - */ -static inline int waking_non_zero_trylock(struct semaphore *sem) -{ - int ret; - unsigned long flags; - - spin_lock_irqsave(&semaphore_wake_lock, flags); - ret = 1; - if (sem->sleepers <= 0) - atomic_inc(&sem->count); - else { - sem->sleepers--; - ret = 0; - } - spin_unlock_irqrestore(&semaphore_wake_lock, flags); - return ret; -} - -#endif diff --git a/include/asm-h8300/semaphore.h b/include/asm-h8300/semaphore.h index f3ffff83ff09..d9b2034ed1d2 100644 --- a/include/asm-h8300/semaphore.h +++ b/include/asm-h8300/semaphore.h @@ -1,190 +1 @@ -#ifndef _H8300_SEMAPHORE_H -#define _H8300_SEMAPHORE_H - -#define RW_LOCK_BIAS 0x01000000 - -#ifndef __ASSEMBLY__ - -#include <linux/linkage.h> -#include <linux/wait.h> -#include <linux/spinlock.h> -#include <linux/rwsem.h> - -#include <asm/system.h> -#include <asm/atomic.h> - -/* - * Interrupt-safe semaphores.. - * - * (C) Copyright 1996 Linus Torvalds - * - * H8/300 version by Yoshinori Sato - */ - - -struct semaphore { - atomic_t count; - int sleepers; - wait_queue_head_t wait; -}; - -#define __SEMAPHORE_INITIALIZER(name, n) \ -{ \ - .count = ATOMIC_INIT(n), \ - .sleepers = 0, \ - .wait = __WAIT_QUEUE_HEAD_INITIALIZER((name).wait) \ -} - -#define __DECLARE_SEMAPHORE_GENERIC(name,count) \ - struct semaphore name = __SEMAPHORE_INITIALIZER(name,count) - -#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name,1) - -static inline void sema_init (struct semaphore *sem, int val) -{ - *sem = (struct semaphore)__SEMAPHORE_INITIALIZER(*sem, val); -} - -static inline void init_MUTEX (struct semaphore *sem) -{ - sema_init(sem, 1); -} - -static inline void init_MUTEX_LOCKED (struct semaphore *sem) -{ - sema_init(sem, 0); -} - -asmlinkage void __down_failed(void /* special register calling convention */); -asmlinkage int __down_failed_interruptible(void /* params in registers */); -asmlinkage int __down_failed_trylock(void /* params in registers */); -asmlinkage void __up_wakeup(void /* special register calling convention */); - -asmlinkage void __down(struct semaphore * sem); -asmlinkage int __down_interruptible(struct semaphore * sem); -asmlinkage int __down_trylock(struct semaphore * sem); -asmlinkage void __up(struct semaphore * sem); - -extern spinlock_t semaphore_wake_lock; - -/* - * This is ugly, but we want the default case to fall through. - * "down_failed" is a special asm handler that calls the C - * routine that actually waits. See arch/m68k/lib/semaphore.S - */ -static inline void down(struct semaphore * sem) -{ - register atomic_t *count asm("er0"); - - might_sleep(); - - count = &(sem->count); - __asm__ __volatile__( - "stc ccr,r3l\n\t" - "orc #0x80,ccr\n\t" - "mov.l %2, er1\n\t" - "dec.l #1,er1\n\t" - "mov.l er1,%0\n\t" - "bpl 1f\n\t" - "ldc r3l,ccr\n\t" - "mov.l %1,er0\n\t" - "jsr @___down\n\t" - "bra 2f\n" - "1:\n\t" - "ldc r3l,ccr\n" - "2:" - : "=m"(*count) - : "g"(sem),"m"(*count) - : "cc", "er1", "er2", "er3"); -} - -static inline int down_interruptible(struct semaphore * sem) -{ - register atomic_t *count asm("er0"); - - might_sleep(); - - count = &(sem->count); - __asm__ __volatile__( - "stc ccr,r1l\n\t" - "orc #0x80,ccr\n\t" - "mov.l %3, er2\n\t" - "dec.l #1,er2\n\t" - "mov.l er2,%1\n\t" - "bpl 1f\n\t" - "ldc r1l,ccr\n\t" - "mov.l %2,er0\n\t" - "jsr @___down_interruptible\n\t" - "bra 2f\n" - "1:\n\t" - "ldc r1l,ccr\n\t" - "sub.l %0,%0\n\t" - "2:\n\t" - : "=r" (count),"=m" (*count) - : "g"(sem),"m"(*count) - : "cc", "er1", "er2", "er3"); - return (int)count; -} - -static inline int down_trylock(struct semaphore * sem) -{ - register atomic_t *count asm("er0"); - - count = &(sem->count); - __asm__ __volatile__( - "stc ccr,r3l\n\t" - "orc #0x80,ccr\n\t" - "mov.l %3,er2\n\t" - "dec.l #1,er2\n\t" - "mov.l er2,%0\n\t" - "bpl 1f\n\t" - "ldc r3l,ccr\n\t" - "jmp @3f\n\t" - LOCK_SECTION_START(".align 2\n\t") - "3:\n\t" - "mov.l %2,er0\n\t" - "jsr @___down_trylock\n\t" - "jmp @2f\n\t" - LOCK_SECTION_END - "1:\n\t" - "ldc r3l,ccr\n\t" - "sub.l %1,%1\n" - "2:" - : "=m" (*count),"=r"(count) - : "g"(sem),"m"(*count) - : "cc", "er1","er2", "er3"); - return (int)count; -} - -/* - * Note! This is subtle. We jump to wake people up only if - * the semaphore was negative (== somebody was waiting on it). - * The default case (no contention) will result in NO - * jumps for both down() and up(). - */ -static inline void up(struct semaphore * sem) -{ - register atomic_t *count asm("er0"); - - count = &(sem->count); - __asm__ __volatile__( - "stc ccr,r3l\n\t" - "orc #0x80,ccr\n\t" - "mov.l %2,er1\n\t" - "inc.l #1,er1\n\t" - "mov.l er1,%0\n\t" - "ldc r3l,ccr\n\t" - "sub.l er2,er2\n\t" - "cmp.l er2,er1\n\t" - "bgt 1f\n\t" - "mov.l %1,er0\n\t" - "jsr @___up\n" - "1:" - : "=m"(*count) - : "g"(sem),"m"(*count) - : "cc", "er1", "er2", "er3"); -} - -#endif /* __ASSEMBLY__ */ - -#endif +#include <linux/semaphore.h> diff --git a/include/asm-ia64/semaphore.h b/include/asm-ia64/semaphore.h index d8393d11288d..d9b2034ed1d2 100644 --- a/include/asm-ia64/semaphore.h +++ b/include/asm-ia64/semaphore.h @@ -1,99 +1 @@ -#ifndef _ASM_IA64_SEMAPHORE_H -#define _ASM_IA64_SEMAPHORE_H - -/* - * Copyright (C) 1998-2000 Hewlett-Packard Co - * Copyright (C) 1998-2000 David Mosberger-Tang <davidm@hpl.hp.com> - */ - -#include <linux/wait.h> -#include <linux/rwsem.h> - -#include <asm/atomic.h> - -struct semaphore { - atomic_t count; - int sleepers; - wait_queue_head_t wait; -}; - -#define __SEMAPHORE_INITIALIZER(name, n) \ -{ \ - .count = ATOMIC_INIT(n), \ - .sleepers = 0, \ - .wait = __WAIT_QUEUE_HEAD_INITIALIZER((name).wait) \ -} - -#define __DECLARE_SEMAPHORE_GENERIC(name,count) \ - struct semaphore name = __SEMAPHORE_INITIALIZER(name, count) - -#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name, 1) - -static inline void -sema_init (struct semaphore *sem, int val) -{ - *sem = (struct semaphore) __SEMAPHORE_INITIALIZER(*sem, val); -} - -static inline void -init_MUTEX (struct semaphore *sem) -{ - sema_init(sem, 1); -} - -static inline void -init_MUTEX_LOCKED (struct semaphore *sem) -{ - sema_init(sem, 0); -} - -extern void __down (struct semaphore * sem); -extern int __down_interruptible (struct semaphore * sem); -extern int __down_trylock (struct semaphore * sem); -extern void __up (struct semaphore * sem); - -/* - * Atomically decrement the semaphore's count. If it goes negative, - * block the calling thread in the TASK_UNINTERRUPTIBLE state. - */ -static inline void -down (struct semaphore *sem) -{ - might_sleep(); - if (ia64_fetchadd(-1, &sem->count.counter, acq) < 1) - __down(sem); -} - -/* - * Atomically decrement the semaphore's count. If it goes negative, - * block the calling thread in the TASK_INTERRUPTIBLE state. - */ -static inline int -down_interruptible (struct semaphore * sem) -{ - int ret = 0; - - might_sleep(); - if (ia64_fetchadd(-1, &sem->count.counter, acq) < 1) - ret = __down_interruptible(sem); - return ret; -} - -static inline int -down_trylock (struct semaphore *sem) -{ - int ret = 0; - - if (ia64_fetchadd(-1, &sem->count.counter, acq) < 1) - ret = __down_trylock(sem); - return ret; -} - -static inline void -up (struct semaphore * sem) -{ - if (ia64_fetchadd(1, &sem->count.counter, rel) <= -1) - __up(sem); -} - -#endif /* _ASM_IA64_SEMAPHORE_H */ +#include <linux/semaphore.h> diff --git a/include/asm-m32r/semaphore.h b/include/asm-m32r/semaphore.h index b5bf95a6f2b4..d9b2034ed1d2 100644 --- a/include/asm-m32r/semaphore.h +++ b/include/asm-m32r/semaphore.h @@ -1,144 +1 @@ -#ifndef _ASM_M32R_SEMAPHORE_H -#define _ASM_M32R_SEMAPHORE_H - -#include <linux/linkage.h> - -#ifdef __KERNEL__ - -/* - * SMP- and interrupt-safe semaphores.. - * - * Copyright (C) 1996 Linus Torvalds - * Copyright (C) 2004, 2006 Hirokazu Takata <takata at linux-m32r.org> - */ - -#include <linux/wait.h> -#include <linux/rwsem.h> -#include <asm/assembler.h> -#include <asm/system.h> -#include <asm/atomic.h> - -struct semaphore { - atomic_t count; - int sleepers; - wait_queue_head_t wait; -}; - -#define __SEMAPHORE_INITIALIZER(name, n) \ -{ \ - .count = ATOMIC_INIT(n), \ - .sleepers = 0, \ - .wait = __WAIT_QUEUE_HEAD_INITIALIZER((name).wait) \ -} - -#define __DECLARE_SEMAPHORE_GENERIC(name,count) \ - struct semaphore name = __SEMAPHORE_INITIALIZER(name,count) - -#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name,1) - -static inline void sema_init (struct semaphore *sem, int val) -{ -/* - * *sem = (struct semaphore)__SEMAPHORE_INITIALIZER((*sem),val); - * - * i'd rather use the more flexible initialization above, but sadly - * GCC 2.7.2.3 emits a bogus warning. EGCS doesnt. Oh well. - */ - atomic_set(&sem->count, val); - sem->sleepers = 0; - init_waitqueue_head(&sem->wait); -} - -static inline void init_MUTEX (struct semaphore *sem) -{ - sema_init(sem, 1); -} - -static inline void init_MUTEX_LOCKED (struct semaphore *sem) -{ - sema_init(sem, 0); -} - -asmlinkage void __down_failed(void /* special register calling convention */); -asmlinkage int __down_failed_interruptible(void /* params in registers */); -asmlinkage int __down_failed_trylock(void /* params in registers */); -asmlinkage void __up_wakeup(void /* special register calling convention */); - -asmlinkage void __down(struct semaphore * sem); -asmlinkage int __down_interruptible(struct semaphore * sem); -asmlinkage int __down_trylock(struct semaphore * sem); -asmlinkage void __up(struct semaphore * sem); - -/* - * Atomically decrement the semaphore's count. If it goes negative, - * block the calling thread in the TASK_UNINTERRUPTIBLE state. - */ -static inline void down(struct semaphore * sem) -{ - might_sleep(); - if (unlikely(atomic_dec_return(&sem->count) < 0)) - __down(sem); -} - -/* - * Interruptible try to acquire a semaphore. If we obtained - * it, return zero. If we were interrupted, returns -EINTR - */ -static inline int down_interruptible(struct semaphore * sem) -{ - int result = 0; - - might_sleep(); - if (unlikely(atomic_dec_return(&sem->count) < 0)) - result = __down_interruptible(sem); - - return result; -} - -/* - * Non-blockingly attempt to down() a semaphore. - * Returns zero if we acquired it - */ -static inline int down_trylock(struct semaphore * sem) -{ - unsigned long flags; - long count; - int result = 0; - - local_irq_save(flags); - __asm__ __volatile__ ( - "# down_trylock \n\t" - DCACHE_CLEAR("%0", "r4", "%1") - M32R_LOCK" %0, @%1; \n\t" - "addi %0, #-1; \n\t" - M32R_UNLOCK" %0, @%1; \n\t" - : "=&r" (count) - : "r" (&sem->count) - : "memory" -#ifdef CONFIG_CHIP_M32700_TS1 - , "r4" -#endif /* CONFIG_CHIP_M32700_TS1 */ - ); - local_irq_restore(flags); - - if (unlikely(count < 0)) - result = __down_trylock(sem); - - return result; -} - -/* - * Note! This is subtle. We jump to wake people up only if - * the semaphore was negative (== somebody was waiting on it). - * The default case (no contention) will result in NO - * jumps for both down() and up(). - */ -static inline void up(struct semaphore * sem) -{ - if (unlikely(atomic_inc_return(&sem->count) <= 0)) - __up(sem); -} - -#endif /* __KERNEL__ */ - -#endif /* _ASM_M32R_SEMAPHORE_H */ +#include <linux/semaphore.h> diff --git a/include/asm-m68k/semaphore-helper.h b/include/asm-m68k/semaphore-helper.h deleted file mode 100644 index eef30ba0b499..000000000000 --- a/include/asm-m68k/semaphore-helper.h +++ /dev/null @@ -1,142 +0,0 @@ -#ifndef _M68K_SEMAPHORE_HELPER_H -#define _M68K_SEMAPHORE_HELPER_H - -/* - * SMP- and interrupt-safe semaphores helper functions. - * - * (C) Copyright 1996 Linus Torvalds - * - * m68k version by Andreas Schwab - */ - -#include <linux/errno.h> - -/* - * These two _must_ execute atomically wrt each other. - */ -static inline void wake_one_more(struct semaphore * sem) -{ - atomic_inc(&sem->waking); -} - -#ifndef CONFIG_RMW_INSNS -extern spinlock_t semaphore_wake_lock; -#endif - -static inline int waking_non_zero(struct semaphore *sem) -{ - int ret; -#ifndef CONFIG_RMW_INSNS - unsigned long flags; - - spin_lock_irqsave(&semaphore_wake_lock, flags); - ret = 0; - if (atomic_read(&sem->waking) > 0) { - atomic_dec(&sem->waking); - ret = 1; - } - spin_unlock_irqrestore(&semaphore_wake_lock, flags); -#else - int tmp1, tmp2; - - __asm__ __volatile__ - ("1: movel %1,%2\n" - " jle 2f\n" - " subql #1,%2\n" - " casl %1,%2,%3\n" - " jne 1b\n" - " moveq #1,%0\n" - "2:" - : "=d" (ret), "=d" (tmp1), "=d" (tmp2) - : "m" (sem->waking), "0" (0), "1" (sem->waking)); -#endif - - return ret; -} - -/* - * waking_non_zero_interruptible: - * 1 got the lock - * 0 go to sleep - * -EINTR interrupted - */ -static inline int waking_non_zero_interruptible(struct semaphore *sem, - struct task_struct *tsk) -{ - int ret; -#ifndef CONFIG_RMW_INSNS - unsigned long flags; - - spin_lock_irqsave(&semaphore_wake_lock, flags); - ret = 0; - if (atomic_read(&sem->waking) > 0) { - atomic_dec(&sem->waking); - ret = 1; - } else if (signal_pending(tsk)) { - atomic_inc(&sem->count); - ret = -EINTR; - } - spin_unlock_irqrestore(&semaphore_wake_lock, flags); -#else - int tmp1, tmp2; - - __asm__ __volatile__ - ("1: movel %1,%2\n" - " jle 2f\n" - " subql #1,%2\n" - " casl %1,%2,%3\n" - " jne 1b\n" - " moveq #1,%0\n" - " jra %a4\n" - "2:" - : "=d" (ret), "=d" (tmp1), "=d" (tmp2) - : "m" (sem->waking), "i" (&&next), "0" (0), "1" (sem->waking)); - if (signal_pending(tsk)) { - atomic_inc(&sem->count); - ret = -EINTR; - } -next: -#endif - - return ret; -} - -/* - * waking_non_zero_trylock: - * 1 failed to lock - * 0 got the lock - */ -static inline int waking_non_zero_trylock(struct semaphore *sem) -{ - int ret; -#ifndef CONFIG_RMW_INSNS - unsigned long flags; - - spin_lock_irqsave(&semaphore_wake_lock, flags); - ret = 1; - if (atomic_read(&sem->waking) > 0) { - atomic_dec(&sem->waking); - ret = 0; - } else - atomic_inc(&sem->count); - spin_unlock_irqrestore(&semaphore_wake_lock, flags); -#else - int tmp1, tmp2; - - __asm__ __volatile__ - ("1: movel %1,%2\n" - " jle 2f\n" - " subql #1,%2\n" - " casl %1,%2,%3\n" - " jne 1b\n" - " moveq #0,%0\n" - "2:" - : "=d" (ret), "=d" (tmp1), "=d" (tmp2) - : "m" (sem->waking), "0" (1), "1" (sem->waking)); - if (ret) - atomic_inc(&sem->count); -#endif - return ret; -} - -#endif diff --git a/include/asm-m68k/semaphore.h b/include/asm-m68k/semaphore.h index 64d6b119bb0a..d9b2034ed1d2 100644 --- a/include/asm-m68k/semaphore.h +++ b/include/asm-m68k/semaphore.h @@ -1,163 +1 @@ -#ifndef _M68K_SEMAPHORE_H -#define _M68K_SEMAPHORE_H - -#define RW_LOCK_BIAS 0x01000000 - -#ifndef __ASSEMBLY__ - -#include <linux/linkage.h> -#include <linux/wait.h> -#include <linux/spinlock.h> -#include <linux/rwsem.h> -#include <linux/stringify.h> - -#include <asm/system.h> -#include <asm/atomic.h> - -/* - * Interrupt-safe semaphores.. - * - * (C) Copyright 1996 Linus Torvalds - * - * m68k version by Andreas Schwab - */ - - -struct semaphore { - atomic_t count; - atomic_t waking; - wait_queue_head_t wait; -}; - -#define __SEMAPHORE_INITIALIZER(name, n) \ -{ \ - .count = ATOMIC_INIT(n), \ - .waking = ATOMIC_INIT(0), \ - .wait = __WAIT_QUEUE_HEAD_INITIALIZER((name).wait) \ -} - -#define __DECLARE_SEMAPHORE_GENERIC(name,count) \ - struct semaphore name = __SEMAPHORE_INITIALIZER(name,count) - -#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name,1) - -static inline void sema_init(struct semaphore *sem, int val) -{ - *sem = (struct semaphore)__SEMAPHORE_INITIALIZER(*sem, val); -} - -static inline void init_MUTEX (struct semaphore *sem) -{ - sema_init(sem, 1); -} - -static inline void init_MUTEX_LOCKED (struct semaphore *sem) -{ - sema_init(sem, 0); -} - -asmlinkage void __down_failed(void /* special register calling convention */); -asmlinkage int __down_failed_interruptible(void /* params in registers */); -asmlinkage int __down_failed_trylock(void /* params in registers */); -asmlinkage void __up_wakeup(void /* special register calling convention */); - -asmlinkage void __down(struct semaphore * sem); -asmlinkage int __down_interruptible(struct semaphore * sem); -asmlinkage int __down_trylock(struct semaphore * sem); -asmlinkage void __up(struct semaphore * sem); - -/* - * This is ugly, but we want the default case to fall through. - * "down_failed" is a special asm handler that calls the C - * routine that actually waits. See arch/m68k/lib/semaphore.S - */ -static inline void down(struct semaphore *sem) -{ - register struct semaphore *sem1 __asm__ ("%a1") = sem; - - might_sleep(); - __asm__ __volatile__( - "| atomic down operation\n\t" - "subql #1,%0@\n\t" - "jmi 2f\n\t" - "1:\n" - LOCK_SECTION_START(".even\n\t") - "2:\tpea 1b\n\t" - "jbra __down_failed\n" - LOCK_SECTION_END - : /* no outputs */ - : "a" (sem1) - : "memory"); -} - -static inline int down_interruptible(struct semaphore *sem) -{ - register struct semaphore *sem1 __asm__ ("%a1") = sem; - register int result __asm__ ("%d0"); - - might_sleep(); - __asm__ __volatile__( - "| atomic interruptible down operation\n\t" - "subql #1,%1@\n\t" - "jmi 2f\n\t" - "clrl %0\n" - "1:\n" - LOCK_SECTION_START(".even\n\t") - "2:\tpea 1b\n\t" - "jbra __down_failed_interruptible\n" - LOCK_SECTION_END - : "=d" (result) - : "a" (sem1) - : "memory"); - return result; -} - -static inline int down_trylock(struct semaphore *sem) -{ - register struct semaphore *sem1 __asm__ ("%a1") = sem; - register int result __asm__ ("%d0"); - - __asm__ __volatile__( - "| atomic down trylock operation\n\t" - "subql #1,%1@\n\t" - "jmi 2f\n\t" - "clrl %0\n" - "1:\n" - LOCK_SECTION_START(".even\n\t") - "2:\tpea 1b\n\t" - "jbra __down_failed_trylock\n" - LOCK_SECTION_END - : "=d" (result) - : "a" (sem1) - : "memory"); - return result; -} - -/* - * Note! This is subtle. We jump to wake people up only if - * the semaphore was negative (== somebody was waiting on it). - * The default case (no contention) will result in NO - * jumps for both down() and up(). - */ -static inline void up(struct semaphore *sem) -{ - register struct semaphore *sem1 __asm__ ("%a1") = sem; - - __asm__ __volatile__( - "| atomic up operation\n\t" - "addql #1,%0@\n\t" - "jle 2f\n" - "1:\n" - LOCK_SECTION_START(".even\n\t") - "2:\t" - "pea 1b\n\t" - "jbra __up_wakeup\n" - LOCK_SECTION_END - : /* no outputs */ - : "a" (sem1) - : "memory"); -} - -#endif /* __ASSEMBLY__ */ - -#endif +#include <linux/semaphore.h> diff --git a/include/asm-m68knommu/semaphore-helper.h b/include/asm-m68knommu/semaphore-helper.h deleted file mode 100644 index 43da7bc483c7..000000000000 --- a/include/asm-m68knommu/semaphore-helper.h +++ /dev/null @@ -1,82 +0,0 @@ -#ifndef _M68K_SEMAPHORE_HELPER_H -#define _M68K_SEMAPHORE_HELPER_H - -/* - * SMP- and interrupt-safe semaphores helper functions. - * - * (C) Copyright 1996 Linus Torvalds - * - * m68k version by Andreas Schwab - */ - - -/* - * These two _must_ execute atomically wrt each other. - */ -static inline void wake_one_more(struct semaphore * sem) -{ - atomic_inc(&sem->waking); -} - -static inline int waking_non_zero(struct semaphore *sem) -{ - int ret; - unsigned long flags; - - spin_lock_irqsave(&semaphore_wake_lock, flags); - ret = 0; - if (atomic_read(&sem->waking) > 0) { - atomic_dec(&sem->waking); - ret = 1; - } - spin_unlock_irqrestore(&semaphore_wake_lock, flags); - return ret; -} - -/* - * waking_non_zero_interruptible: - * 1 got the lock - * 0 go to sleep - * -EINTR interrupted - */ -static inline int waking_non_zero_interruptible(struct semaphore *sem, - struct task_struct *tsk) -{ - int ret; - unsigned long flags; - - spin_lock_irqsave(&semaphore_wake_lock, flags); - ret = 0; - if (atomic_read(&sem->waking) > 0) { - atomic_dec(&sem->waking); - ret = 1; - } else if (signal_pending(tsk)) { - atomic_inc(&sem->count); - ret = -EINTR; - } - spin_unlock_irqrestore(&semaphore_wake_lock, flags); - return ret; -} - -/* - * waking_non_zero_trylock: - * 1 failed to lock - * 0 got the lock - */ -static inline int waking_non_zero_trylock(struct semaphore *sem) -{ - int ret; - unsigned long flags; - - spin_lock_irqsave(&semaphore_wake_lock, flags); - ret = 1; - if (atomic_read(&sem->waking) > 0) { - atomic_dec(&sem->waking); - ret = 0; - } else - atomic_inc(&sem->count); - spin_unlock_irqrestore(&semaphore_wake_lock, flags); - return ret; -} - -#endif diff --git a/include/asm-m68knommu/semaphore.h b/include/asm-m68knommu/semaphore.h index 5779eb6c0689..d9b2034ed1d2 100644 --- a/include/asm-m68knommu/semaphore.h +++ b/include/asm-m68knommu/semaphore.h @@ -1,153 +1 @@ -#ifndef _M68K_SEMAPHORE_H -#define _M68K_SEMAPHORE_H - -#define RW_LOCK_BIAS 0x01000000 - -#ifndef __ASSEMBLY__ - -#include <linux/linkage.h> -#include <linux/wait.h> -#include <linux/spinlock.h> -#include <linux/rwsem.h> - -#include <asm/system.h> -#include <asm/atomic.h> - -/* - * Interrupt-safe semaphores.. - * - * (C) Copyright 1996 Linus Torvalds - * - * m68k version by Andreas Schwab - */ - - -struct semaphore { - atomic_t count; - atomic_t waking; - wait_queue_head_t wait; -}; - -#define __SEMAPHORE_INITIALIZER(name, n) \ -{ \ - .count = ATOMIC_INIT(n), \ - .waking = ATOMIC_INIT(0), \ - .wait = __WAIT_QUEUE_HEAD_INITIALIZER((name).wait) \ -} - -#define __DECLARE_SEMAPHORE_GENERIC(name,count) \ - struct semaphore name = __SEMAPHORE_INITIALIZER(name,count) - -#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name,1) - -static inline void sema_init (struct semaphore *sem, int val) -{ - *sem = (struct semaphore)__SEMAPHORE_INITIALIZER(*sem, val); -} - -static inline void init_MUTEX (struct semaphore *sem) -{ - sema_init(sem, 1); -} - -static inline void init_MUTEX_LOCKED (struct semaphore *sem) -{ - sema_init(sem, 0); -} - -asmlinkage void __down_failed(void /* special register calling convention */); -asmlinkage int __down_failed_interruptible(void /* params in registers */); -asmlinkage int __down_failed_trylock(void /* params in registers */); -asmlinkage void __up_wakeup(void /* special register calling convention */); - -asmlinkage void __down(struct semaphore * sem); -asmlinkage int __down_interruptible(struct semaphore * sem); -asmlinkage int __down_trylock(struct semaphore * sem); -asmlinkage void __up(struct semaphore * sem); - -extern spinlock_t semaphore_wake_lock; - -/* - * This is ugly, but we want the default case to fall through. - * "down_failed" is a special asm handler that calls the C - * routine that actually waits. See arch/m68k/lib/semaphore.S - */ -static inline void down(struct semaphore * sem) -{ - might_sleep(); - __asm__ __volatile__( - "| atomic down operation\n\t" - "movel %0, %%a1\n\t" - "lea %%pc@(1f), %%a0\n\t" - "subql #1, %%a1@\n\t" - "jmi __down_failed\n" - "1:" - : /* no outputs */ - : "g" (sem) - : "cc", "%a0", "%a1", "memory"); -} - -static inline int down_interruptible(struct semaphore * sem) -{ - int ret; - - might_sleep(); - __asm__ __volatile__( - "| atomic down operation\n\t" - "movel %1, %%a1\n\t" - "lea %%pc@(1f), %%a0\n\t" - "subql #1, %%a1@\n\t" - "jmi __down_failed_interruptible\n\t" - "clrl %%d0\n" - "1: movel %%d0, %0\n" - : "=d" (ret) - : "g" (sem) - : "cc", "%d0", "%a0", "%a1", "memory"); - return(ret); -} - -static inline int down_trylock(struct semaphore * sem) -{ - register struct semaphore *sem1 __asm__ ("%a1") = sem; - register int result __asm__ ("%d0"); - - __asm__ __volatile__( - "| atomic down trylock operation\n\t" - "subql #1,%1@\n\t" - "jmi 2f\n\t" - "clrl %0\n" - "1:\n" - ".section .text.lock,\"ax\"\n" - ".even\n" - "2:\tpea 1b\n\t" - "jbra __down_failed_trylock\n" - ".previous" - : "=d" (result) - : "a" (sem1) - : "memory"); - return result; -} - -/* - * Note! This is subtle. We jump to wake people up only if - * the semaphore was negative (== somebody was waiting on it). - * The default case (no contention) will result in NO - * jumps for both down() and up(). - */ -static inline void up(struct semaphore * sem) -{ - __asm__ __volatile__( - "| atomic up operation\n\t" - "movel %0, %%a1\n\t" - "lea %%pc@(1f), %%a0\n\t" - "addql #1, %%a1@\n\t" - "jle __up_wakeup\n" - "1:" - : /* no outputs */ - : "g" (sem) - : "cc", "%a0", "%a1", "memory"); -} - -#endif /* __ASSEMBLY__ */ - -#endif +#include <linux/semaphore.h> diff --git a/include/asm-mips/semaphore.h b/include/asm-mips/semaphore.h index fdf8042b784b..d9b2034ed1d2 100644 --- a/include/asm-mips/semaphore.h +++ b/include/asm-mips/semaphore.h @@ -1,108 +1 @@ -/* - * This file is subject to the terms and conditions of the GNU General Public - * License. See the file "COPYING" in the main directory of this archive - * for more details. - * - * Copyright (C) 1996 Linus Torvalds - * Copyright (C) 1998, 99, 2000, 01, 04 Ralf Baechle - * Copyright (C) 1999, 2000, 01 Silicon Graphics, Inc. - * Copyright (C) 2000, 01 MIPS Technologies, Inc. - * - * In all honesty, little of the old MIPS code left - the PPC64 variant was - * just looking nice and portable so I ripped it. Credits to whoever wrote - * it. - */ -#ifndef __ASM_SEMAPHORE_H -#define __ASM_SEMAPHORE_H - -/* - * Remove spinlock-based RW semaphores; RW semaphore definitions are - * now in rwsem.h and we use the generic lib/rwsem.c implementation. - * Rework semaphores to use atomic_dec_if_positive. - * -- Paul Mackerras (paulus@samba.org) - */ - -#ifdef __KERNEL__ - -#include <asm/atomic.h> -#include <asm/system.h> -#include <linux/wait.h> -#include <linux/rwsem.h> - -struct semaphore { - /* - * Note that any negative value of count is equivalent to 0, - * but additionally indicates that some process(es) might be - * sleeping on `wait'. - */ - atomic_t count; - wait_queue_head_t wait; -}; - -#define __SEMAPHORE_INITIALIZER(name, n) \ -{ \ - .count = ATOMIC_INIT(n), \ - .wait = __WAIT_QUEUE_HEAD_INITIALIZER((name).wait) \ -} - -#define __DECLARE_SEMAPHORE_GENERIC(name, count) \ - struct semaphore name = __SEMAPHORE_INITIALIZER(name, count) - -#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name, 1) - -static inline void sema_init(struct semaphore *sem, int val) -{ - atomic_set(&sem->count, val); - init_waitqueue_head(&sem->wait); -} - -static inline void init_MUTEX(struct semaphore *sem) -{ - sema_init(sem, 1); -} - -static inline void init_MUTEX_LOCKED(struct semaphore *sem) -{ - sema_init(sem, 0); -} - -extern void __down(struct semaphore * sem); -extern int __down_interruptible(struct semaphore * sem); -extern void __up(struct semaphore * sem); - -static inline void down(struct semaphore * sem) -{ - might_sleep(); - - /* - * Try to get the semaphore, take the slow path if we fail. - */ - if (unlikely(atomic_dec_return(&sem->count) < 0)) - __down(sem); -} - -static inline int down_interruptible(struct semaphore * sem) -{ - int ret = 0; - - might_sleep(); - - if (unlikely(atomic_dec_return(&sem->count) < 0)) - ret = __down_interruptible(sem); - return ret; -} - -static inline int down_trylock(struct semaphore * sem) -{ - return atomic_dec_if_positive(&sem->count) < 0; -} - -static inline void up(struct semaphore * sem) -{ - if (unlikely(atomic_inc_return(&sem->count) <= 0)) - __up(sem); -} - -#endif /* __KERNEL__ */ - -#endif /* __ASM_SEMAPHORE_H */ +#include <linux/semaphore.h> diff --git a/include/asm-mn10300/semaphore.h b/include/asm-mn10300/semaphore.h index 5a9e1ad0b253..d9b2034ed1d2 100644 --- a/include/asm-mn10300/semaphore.h +++ b/include/asm-mn10300/semaphore.h @@ -1,169 +1 @@ -/* MN10300 Semaphores - * - * Copyright (C) 2007 Red Hat, Inc. All Rights Reserved. - * Written by David Howells (dhowells@redhat.com) - * - * This program is free software; you can redistribute it and/or - * modify it under the terms of the GNU General Public Licence - * as published by the Free Software Foundation; either version - * 2 of the Licence, or (at your option) any later version. - */ -#ifndef _ASM_SEMAPHORE_H -#define _ASM_SEMAPHORE_H - -#ifndef __ASSEMBLY__ - -#include <linux/linkage.h> -#include <linux/wait.h> -#include <linux/spinlock.h> -#include <linux/rwsem.h> - -#define SEMAPHORE_DEBUG 0 - -/* - * the semaphore definition - * - if count is >0 then there are tokens available on the semaphore for down - * to collect - * - if count is <=0 then there are no spare tokens, and anyone that wants one - * must wait - * - if wait_list is not empty, then there are processes waiting for the - * semaphore - */ -struct semaphore { - atomic_t count; /* it's not really atomic, it's - * just that certain modules - * expect to be able to access - * it directly */ - spinlock_t wait_lock; - struct list_head wait_list; -#if SEMAPHORE_DEBUG - unsigned __magic; -#endif -}; - -#if SEMAPHORE_DEBUG -# define __SEM_DEBUG_INIT(name) , (long)&(name).__magic -#else -# define __SEM_DEBUG_INIT(name) -#endif - - -#define __SEMAPHORE_INITIALIZER(name, init_count) \ -{ \ - .count = ATOMIC_INIT(init_count), \ - .wait_lock = __SPIN_LOCK_UNLOCKED((name).wait_lock), \ - .wait_list = LIST_HEAD_INIT((name).wait_list) \ - __SEM_DEBUG_INIT(name) \ -} - -#define __DECLARE_SEMAPHORE_GENERIC(name,count) \ - struct semaphore name = __SEMAPHORE_INITIALIZER(name, count) - -#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name, 1) -#define DECLARE_MUTEX_LOCKED(name) __DECLARE_SEMAPHORE_GENERIC(name, 0) - -static inline void sema_init(struct semaphore *sem, int val) -{ - *sem = (struct semaphore) __SEMAPHORE_INITIALIZER(*sem, val); -} - -static inline void init_MUTEX(struct semaphore *sem) -{ - sema_init(sem, 1); -} - -static inline void init_MUTEX_LOCKED(struct semaphore *sem) -{ - sema_init(sem, 0); -} - -extern void __down(struct semaphore *sem, unsigned long flags); -extern int __down_interruptible(struct semaphore *sem, unsigned long flags); -extern void __up(struct semaphore *sem); - -static inline void down(struct semaphore *sem) -{ - unsigned long flags; - int count; - -#if SEMAPHORE_DEBUG - CHECK_MAGIC(sem->__magic); -#endif - - spin_lock_irqsave(&sem->wait_lock, flags); - count = atomic_read(&sem->count); - if (likely(count > 0)) { - atomic_set(&sem->count, count - 1); - spin_unlock_irqrestore(&sem->wait_lock, flags); - } else { - __down(sem, flags); - } -} - -static inline int down_interruptible(struct semaphore *sem) -{ - unsigned long flags; - int count, ret = 0; - -#if SEMAPHORE_DEBUG - CHECK_MAGIC(sem->__magic); -#endif - - spin_lock_irqsave(&sem->wait_lock, flags); - count = atomic_read(&sem->count); - if (likely(count > 0)) { - atomic_set(&sem->count, count - 1); - spin_unlock_irqrestore(&sem->wait_lock, flags); - } else { - ret = __down_interruptible(sem, flags); - } - return ret; -} - -/* - * non-blockingly attempt to down() a semaphore. - * - returns zero if we acquired it - */ -static inline int down_trylock(struct semaphore *sem) -{ - unsigned long flags; - int count, success = 0; - -#if SEMAPHORE_DEBUG - CHECK_MAGIC(sem->__magic); -#endif - - spin_lock_irqsave(&sem->wait_lock, flags); - count = atomic_read(&sem->count); - if (likely(count > 0)) { - atomic_set(&sem->count, count - 1); - success = 1; - } - spin_unlock_irqrestore(&sem->wait_lock, flags); - return !success; -} - -static inline void up(struct semaphore *sem) -{ - unsigned long flags; - -#if SEMAPHORE_DEBUG - CHECK_MAGIC(sem->__magic); -#endif - - spin_lock_irqsave(&sem->wait_lock, flags); - if (!list_empty(&sem->wait_list)) - __up(sem); - else - atomic_set(&sem->count, atomic_read(&sem->count) + 1); - spin_unlock_irqrestore(&sem->wait_lock, flags); -} - -static inline int sem_getcount(struct semaphore *sem) -{ - return atomic_read(&sem->count); -} - -#endif /* __ASSEMBLY__ */ - -#endif +#include <linux/semaphore.h> diff --git a/include/asm-parisc/semaphore-helper.h b/include/asm-parisc/semaphore-helper.h deleted file mode 100644 index 387f7c1277a2..000000000000 --- a/include/asm-parisc/semaphore-helper.h +++ /dev/null @@ -1,89 +0,0 @@ -#ifndef _ASM_PARISC_SEMAPHORE_HELPER_H -#define _ASM_PARISC_SEMAPHORE_HELPER_H - -/* - * SMP- and interrupt-safe semaphores helper functions. - * - * (C) Copyright 1996 Linus Torvalds - * (C) Copyright 1999 Andrea Arcangeli - */ - -/* - * These two _must_ execute atomically wrt each other. - * - * This is trivially done with load_locked/store_cond, - * which we have. Let the rest of the losers suck eggs. - */ -static __inline__ void wake_one_more(struct semaphore * sem) -{ - atomic_inc((atomic_t *)&sem->waking); -} - -static __inline__ int waking_non_zero(struct semaphore *sem) -{ - unsigned long flags; - int ret = 0; - - spin_lock_irqsave(&semaphore_wake_lock, flags); - if (sem->waking > 0) { - sem->waking--; - ret = 1; - } - spin_unlock_irqrestore(&semaphore_wake_lock, flags); - return ret; -} - -/* - * waking_non_zero_interruptible: - * 1 got the lock - * 0 go to sleep - * -EINTR interrupted - * - * We must undo the sem->count down_interruptible() increment while we are - * protected by the spinlock in order to make atomic this atomic_inc() with the - * atomic_read() in wake_one_more(), otherwise we can race. -arca - */ -static __inline__ int waking_non_zero_interruptible(struct semaphore *sem, - struct task_struct *tsk) -{ - unsigned long flags; - int ret = 0; - - spin_lock_irqsave(&semaphore_wake_lock, flags); - if (sem->waking > 0) { - sem->waking--; - ret = 1; - } else if (signal_pending(tsk)) { - atomic_inc(&sem->count); - ret = -EINTR; - } - spin_unlock_irqrestore(&semaphore_wake_lock, flags); - return ret; -} - -/* - * waking_non_zero_trylock: - * 1 failed to lock - * 0 got the lock - * - * We must undo the sem->count down_trylock() increment while we are - * protected by the spinlock in order to make atomic this atomic_inc() with the - * atomic_read() in wake_one_more(), otherwise we can race. -arca - */ -static __inline__ int waking_non_zero_trylock(struct semaphore *sem) -{ - unsigned long flags; - int ret = 1; - - spin_lock_irqsave(&semaphore_wake_lock, flags); - if (sem->waking <= 0) - atomic_inc(&sem->count); - else { - sem->waking--; - ret = 0; - } - spin_unlock_irqrestore(&semaphore_wake_lock, flags); - return ret; -} - -#endif /* _ASM_PARISC_SEMAPHORE_HELPER_H */ diff --git a/include/asm-parisc/semaphore.h b/include/asm-parisc/semaphore.h index a16271cdc748..d9b2034ed1d2 100644 --- a/include/asm-parisc/semaphore.h +++ b/include/asm-parisc/semaphore.h @@ -1,145 +1 @@ -/* SMP- and interrupt-safe semaphores. - * PA-RISC version by Matthew Wilcox - * - * Linux/PA-RISC Project (http://www.parisc-linux.org/) - * Copyright (C) 1996 Linus Torvalds - * Copyright (C) 1999-2001 Matthew Wilcox < willy at debian d0T org > - * Copyright (C) 2000 Grant Grundler < grundler a debian org > - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA - */ - -#ifndef _ASM_PARISC_SEMAPHORE_H -#define _ASM_PARISC_SEMAPHORE_H - -#include <linux/spinlock.h> -#include <linux/wait.h> -#include <linux/rwsem.h> - -#include <asm/system.h> - -/* - * The `count' is initialised to the number of people who are allowed to - * take the lock. (Normally we want a mutex, so this is `1'). if - * `count' is positive, the lock can be taken. if it's 0, no-one is - * waiting on it. if it's -1, at least one task is waiting. - */ -struct semaphore { - spinlock_t sentry; - int count; - wait_queue_head_t wait; -}; - -#define __SEMAPHORE_INITIALIZER(name, n) \ -{ \ - .sentry = SPIN_LOCK_UNLOCKED, \ - .count = n, \ - .wait = __WAIT_QUEUE_HEAD_INITIALIZER((name).wait) \ -} - -#define __DECLARE_SEMAPHORE_GENERIC(name,count) \ - struct semaphore name = __SEMAPHORE_INITIALIZER(name,count) - -#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name,1) - -static inline void sema_init (struct semaphore *sem, int val) -{ - *sem = (struct semaphore)__SEMAPHORE_INITIALIZER((*sem),val); -} - -static inline void init_MUTEX (struct semaphore *sem) -{ - sema_init(sem, 1); -} - -static inline void init_MUTEX_LOCKED (struct semaphore *sem) -{ - sema_init(sem, 0); -} - -static inline int sem_getcount(struct semaphore *sem) -{ - return sem->count; -} - -asmlinkage void __down(struct semaphore * sem); -asmlinkage int __down_interruptible(struct semaphore * sem); -asmlinkage void __up(struct semaphore * sem); - -/* Semaphores can be `tried' from irq context. So we have to disable - * interrupts while we're messing with the semaphore. Sorry. - */ - -static inline void down(struct semaphore * sem) -{ - might_sleep(); - spin_lock_irq(&sem->sentry); - if (sem->count > 0) { - sem->count--; - } else { - __down(sem); - } - spin_unlock_irq(&sem->sentry); -} - -static inline int down_interruptible(struct semaphore * sem) -{ - int ret = 0; - might_sleep(); - spin_lock_irq(&sem->sentry); - if (sem->count > 0) { - sem->count--; - } else { - ret = __down_interruptible(sem); - } - spin_unlock_irq(&sem->sentry); - return ret; -} - -/* - * down_trylock returns 0 on success, 1 if we failed to get the lock. - * May not sleep, but must preserve irq state - */ -static inline int down_trylock(struct semaphore * sem) -{ - unsigned long flags; - int count; - - spin_lock_irqsave(&sem->sentry, flags); - count = sem->count - 1; - if (count >= 0) - sem->count = count; - spin_unlock_irqrestore(&sem->sentry, flags); - return (count < 0); -} - -/* - * Note! This is subtle. We jump to wake people up only if - * the semaphore was negative (== somebody was waiting on it). - */ -static inline void up(struct semaphore * sem) -{ - unsigned long flags; - - spin_lock_irqsave(&sem->sentry, flags); - if (sem->count < 0) { - __up(sem); - } else { - sem->count++; - } - spin_unlock_irqrestore(&sem->sentry, flags); -} - -#endif /* _ASM_PARISC_SEMAPHORE_H */ +#include <linux/semaphore.h> diff --git a/include/asm-powerpc/semaphore.h b/include/asm-powerpc/semaphore.h index 48dd32e07749..d9b2034ed1d2 100644 --- a/include/asm-powerpc/semaphore.h +++ b/include/asm-powerpc/semaphore.h @@ -1,94 +1 @@ -#ifndef _ASM_POWERPC_SEMAPHORE_H -#define _ASM_POWERPC_SEMAPHORE_H - -/* - * Remove spinlock-based RW semaphores; RW semaphore definitions are - * now in rwsem.h and we use the generic lib/rwsem.c implementation. - * Rework semaphores to use atomic_dec_if_positive. - * -- Paul Mackerras (paulus@samba.org) - */ - -#ifdef __KERNEL__ - -#include <asm/atomic.h> -#include <asm/system.h> -#include <linux/wait.h> -#include <linux/rwsem.h> - -struct semaphore { - /* - * Note that any negative value of count is equivalent to 0, - * but additionally indicates that some process(es) might be - * sleeping on `wait'. - */ - atomic_t count; - wait_queue_head_t wait; -}; - -#define __SEMAPHORE_INITIALIZER(name, n) \ -{ \ - .count = ATOMIC_INIT(n), \ - .wait = __WAIT_QUEUE_HEAD_INITIALIZER((name).wait) \ -} - -#define __DECLARE_SEMAPHORE_GENERIC(name, count) \ - struct semaphore name = __SEMAPHORE_INITIALIZER(name,count) - -#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name, 1) - -static inline void sema_init (struct semaphore *sem, int val) -{ - atomic_set(&sem->count, val); - init_waitqueue_head(&sem->wait); -} - -static inline void init_MUTEX (struct semaphore *sem) -{ - sema_init(sem, 1); -} - -static inline void init_MUTEX_LOCKED (struct semaphore *sem) -{ - sema_init(sem, 0); -} - -extern void __down(struct semaphore * sem); -extern int __down_interruptible(struct semaphore * sem); -extern void __up(struct semaphore * sem); - -static inline void down(struct semaphore * sem) -{ - might_sleep(); - - /* - * Try to get the semaphore, take the slow path if we fail. - */ - if (unlikely(atomic_dec_return(&sem->count) < 0)) - __down(sem); -} - -static inline int down_interruptible(struct semaphore * sem) -{ - int ret = 0; - - might_sleep(); - - if (unlikely(atomic_dec_return(&sem->count) < 0)) - ret = __down_interruptible(sem); - return ret; -} - -static inline int down_trylock(struct semaphore * sem) -{ - return atomic_dec_if_positive(&sem->count) < 0; -} - -static inline void up(struct semaphore * sem) -{ - if (unlikely(atomic_inc_return(&sem->count) <= 0)) - __up(sem); -} - -#endif /* __KERNEL__ */ - -#endif /* _ASM_POWERPC_SEMAPHORE_H */ +#include <linux/semaphore.h> diff --git a/include/asm-s390/cio.h b/include/asm-s390/cio.h index 123b557c3ff4..0818ecd30ca6 100644 --- a/include/asm-s390/cio.h +++ b/include/asm-s390/cio.h @@ -397,6 +397,10 @@ struct cio_iplinfo { extern int cio_get_iplinfo(struct cio_iplinfo *iplinfo); +/* Function from drivers/s390/cio/chsc.c */ +int chsc_sstpc(void *page, unsigned int op, u16 ctrl); +int chsc_sstpi(void *page, void *result, size_t size); + #endif #endif diff --git a/include/asm-s390/cpu.h b/include/asm-s390/cpu.h index 352dde194f3c..e5a6a9ba3adf 100644 --- a/include/asm-s390/cpu.h +++ b/include/asm-s390/cpu.h @@ -22,4 +22,12 @@ struct s390_idle_data { DECLARE_PER_CPU(struct s390_idle_data, s390_idle); +void s390_idle_leave(void); + +static inline void s390_idle_check(void) +{ + if ((&__get_cpu_var(s390_idle))->in_idle) + s390_idle_leave(); +} + #endif /* _ASM_S390_CPU_H_ */ diff --git a/include/asm-s390/debug.h b/include/asm-s390/debug.h index c00dd2b3dc50..335baf4fc64f 100644 --- a/include/asm-s390/debug.h +++ b/include/asm-s390/debug.h @@ -73,6 +73,7 @@ typedef struct debug_info { struct dentry* debugfs_entries[DEBUG_MAX_VIEWS]; struct debug_view* views[DEBUG_MAX_VIEWS]; char name[DEBUG_MAX_NAME_LEN]; + mode_t mode; } debug_info_t; typedef int (debug_header_proc_t) (debug_info_t* id, @@ -122,6 +123,10 @@ debug_entry_t* debug_exception_common(debug_info_t* id, int level, debug_info_t* debug_register(char* name, int pages, int nr_areas, int buf_size); +debug_info_t *debug_register_mode(char *name, int pages, int nr_areas, + int buf_size, mode_t mode, uid_t uid, + gid_t gid); + void debug_unregister(debug_info_t* id); void debug_set_level(debug_info_t* id, int new_level); diff --git a/include/asm-s390/extmem.h b/include/asm-s390/extmem.h index c8802c934b74..33837d756184 100644 --- a/include/asm-s390/extmem.h +++ b/include/asm-s390/extmem.h @@ -22,11 +22,12 @@ #define SEGMENT_SHARED 0 #define SEGMENT_EXCLUSIVE 1 -extern int segment_load (char *name,int segtype,unsigned long *addr,unsigned long *length); -extern void segment_unload(char *name); -extern void segment_save(char *name); -extern int segment_type (char* name); -extern int segment_modify_shared (char *name, int do_nonshared); +int segment_load (char *name, int segtype, unsigned long *addr, unsigned long *length); +void segment_unload(char *name); +void segment_save(char *name); +int segment_type (char* name); +int segment_modify_shared (char *name, int do_nonshared); +void segment_warning(int rc, char *seg_name); #endif #endif diff --git a/include/asm-s390/hardirq.h b/include/asm-s390/hardirq.h index 31beb18cb3d1..4b7cb964ff35 100644 --- a/include/asm-s390/hardirq.h +++ b/include/asm-s390/hardirq.h @@ -32,6 +32,6 @@ typedef struct { #define HARDIRQ_BITS 8 -extern void account_ticks(u64 time); +void clock_comparator_work(void); #endif /* __ASM_HARDIRQ_H */ diff --git a/include/asm-s390/lowcore.h b/include/asm-s390/lowcore.h index 801a6fd35b5b..5de3efb31445 100644 --- a/include/asm-s390/lowcore.h +++ b/include/asm-s390/lowcore.h @@ -56,6 +56,8 @@ #define __LC_IO_INT_WORD 0x0C0 #define __LC_MCCK_CODE 0x0E8 +#define __LC_LAST_BREAK 0x110 + #define __LC_RETURN_PSW 0x200 #define __LC_SAVE_AREA 0xC00 @@ -80,7 +82,6 @@ #define __LC_CPUID 0xC60 #define __LC_CPUADDR 0xC68 #define __LC_IPLDEV 0xC7C -#define __LC_JIFFY_TIMER 0xC80 #define __LC_CURRENT 0xC90 #define __LC_INT_CLOCK 0xC98 #else /* __s390x__ */ @@ -103,7 +104,6 @@ #define __LC_CPUID 0xD80 #define __LC_CPUADDR 0xD88 #define __LC_IPLDEV 0xDB8 -#define __LC_JIFFY_TIMER 0xDC0 #define __LC_CURRENT 0xDD8 #define __LC_INT_CLOCK 0xDE8 #endif /* __s390x__ */ @@ -276,7 +276,7 @@ struct _lowcore /* entry.S sensitive area end */ /* SMP info area: defined by DJB */ - __u64 jiffy_timer; /* 0xc80 */ + __u64 clock_comparator; /* 0xc80 */ __u32 ext_call_fast; /* 0xc88 */ __u32 percpu_offset; /* 0xc8c */ __u32 current_task; /* 0xc90 */ @@ -368,11 +368,12 @@ struct _lowcore /* entry.S sensitive area end */ /* SMP info area: defined by DJB */ - __u64 jiffy_timer; /* 0xdc0 */ + __u64 clock_comparator; /* 0xdc0 */ __u64 ext_call_fast; /* 0xdc8 */ __u64 percpu_offset; /* 0xdd0 */ __u64 current_task; /* 0xdd8 */ - __u64 softirq_pending; /* 0xde0 */ + __u32 softirq_pending; /* 0xde0 */ + __u32 pad_0x0de4; /* 0xde4 */ __u64 int_clock; /* 0xde8 */ __u8 pad12[0xe00-0xdf0]; /* 0xdf0 */ diff --git a/include/asm-s390/processor.h b/include/asm-s390/processor.h index 51d88912aa20..8eaf343a12a8 100644 --- a/include/asm-s390/processor.h +++ b/include/asm-s390/processor.h @@ -175,6 +175,13 @@ extern void task_show_regs(struct seq_file *m, struct task_struct *task); extern void show_registers(struct pt_regs *regs); extern void show_code(struct pt_regs *regs); extern void show_trace(struct task_struct *task, unsigned long *sp); +#ifdef CONFIG_64BIT +extern void show_last_breaking_event(struct pt_regs *regs); +#else +static inline void show_last_breaking_event(struct pt_regs *regs) +{ +} +#endif unsigned long get_wchan(struct task_struct *p); #define task_pt_regs(tsk) ((struct pt_regs *) \ diff --git a/include/asm-s390/semaphore.h b/include/asm-s390/semaphore.h index 0e7001ad8392..d9b2034ed1d2 100644 --- a/include/asm-s390/semaphore.h +++ b/include/asm-s390/semaphore.h @@ -1,107 +1 @@ -/* - * include/asm-s390/semaphore.h - * - * S390 version - * Copyright (C) 1999 IBM Deutschland Entwicklung GmbH, IBM Corporation - * - * Derived from "include/asm-i386/semaphore.h" - * (C) Copyright 1996 Linus Torvalds - */ - -#ifndef _S390_SEMAPHORE_H -#define _S390_SEMAPHORE_H - -#include <asm/system.h> -#include <asm/atomic.h> -#include <linux/wait.h> -#include <linux/rwsem.h> - -struct semaphore { - /* - * Note that any negative value of count is equivalent to 0, - * but additionally indicates that some process(es) might be - * sleeping on `wait'. - */ - atomic_t count; - wait_queue_head_t wait; -}; - -#define __SEMAPHORE_INITIALIZER(name,count) \ - { ATOMIC_INIT(count), __WAIT_QUEUE_HEAD_INITIALIZER((name).wait) } - -#define __DECLARE_SEMAPHORE_GENERIC(name,count) \ - struct semaphore name = __SEMAPHORE_INITIALIZER(name,count) - -#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name,1) - -static inline void sema_init (struct semaphore *sem, int val) -{ - atomic_set(&sem->count, val); - init_waitqueue_head(&sem->wait); -} - -static inline void init_MUTEX (struct semaphore *sem) -{ - sema_init(sem, 1); -} - -static inline void init_MUTEX_LOCKED (struct semaphore *sem) -{ - sema_init(sem, 0); -} - -asmlinkage void __down(struct semaphore * sem); -asmlinkage int __down_interruptible(struct semaphore * sem); -asmlinkage int __down_trylock(struct semaphore * sem); -asmlinkage void __up(struct semaphore * sem); - -static inline void down(struct semaphore * sem) -{ - might_sleep(); - if (atomic_dec_return(&sem->count) < 0) - __down(sem); -} - -static inline int down_interruptible(struct semaphore * sem) -{ - int ret = 0; - - might_sleep(); - if (atomic_dec_return(&sem->count) < 0) - ret = __down_interruptible(sem); - return ret; -} - -static inline int down_trylock(struct semaphore * sem) -{ - int old_val, new_val; - - /* - * This inline assembly atomically implements the equivalent - * to the following C code: - * old_val = sem->count.counter; - * if ((new_val = old_val) > 0) - * sem->count.counter = --new_val; - * In the ppc code this is called atomic_dec_if_positive. - */ - asm volatile( - " l %0,0(%3)\n" - "0: ltr %1,%0\n" - " jle 1f\n" - " ahi %1,-1\n" - " cs %0,%1,0(%3)\n" - " jl 0b\n" - "1:" - : "=&d" (old_val), "=&d" (new_val), "=m" (sem->count.counter) - : "a" (&sem->count.counter), "m" (sem->count.counter) - : "cc", "memory"); - return old_val <= 0; -} - -static inline void up(struct semaphore * sem) -{ - if (atomic_inc_return(&sem->count) <= 0) - __up(sem); -} - -#endif +#include <linux/semaphore.h> diff --git a/include/asm-s390/smp.h b/include/asm-s390/smp.h index c7b74326a527..6f3821a6a902 100644 --- a/include/asm-s390/smp.h +++ b/include/asm-s390/smp.h @@ -90,6 +90,9 @@ extern void __cpu_die (unsigned int cpu); extern void cpu_die (void) __attribute__ ((noreturn)); extern int __cpu_up (unsigned int cpu); +extern struct mutex smp_cpu_state_mutex; +extern int smp_cpu_polarization[]; + extern int smp_call_function_mask(cpumask_t mask, void (*func)(void *), void *info, int wait); #endif diff --git a/include/asm-s390/sysinfo.h b/include/asm-s390/sysinfo.h new file mode 100644 index 000000000000..abe10ae15e46 --- /dev/null +++ b/include/asm-s390/sysinfo.h @@ -0,0 +1,116 @@ +/* + * definition for store system information stsi + * + * Copyright IBM Corp. 2001,2008 + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License (version 2 only) + * as published by the Free Software Foundation. + * + * Author(s): Ulrich Weigand <weigand@de.ibm.com> + * Christian Borntraeger <borntraeger@de.ibm.com> + */ + +struct sysinfo_1_1_1 { + char reserved_0[32]; + char manufacturer[16]; + char type[4]; + char reserved_1[12]; + char model_capacity[16]; + char sequence[16]; + char plant[4]; + char model[16]; + char model_perm_cap[16]; + char model_temp_cap[16]; + char model_cap_rating[4]; + char model_perm_cap_rating[4]; + char model_temp_cap_rating[4]; +}; + +struct sysinfo_1_2_1 { + char reserved_0[80]; + char sequence[16]; + char plant[4]; + char reserved_1[2]; + unsigned short cpu_address; +}; + +struct sysinfo_1_2_2 { + char format; + char reserved_0[1]; + unsigned short acc_offset; + char reserved_1[24]; + unsigned int secondary_capability; + unsigned int capability; + unsigned short cpus_total; + unsigned short cpus_configured; + unsigned short cpus_standby; + unsigned short cpus_reserved; + unsigned short adjustment[0]; +}; + +struct sysinfo_1_2_2_extension { + unsigned int alt_capability; + unsigned short alt_adjustment[0]; +}; + +struct sysinfo_2_2_1 { + char reserved_0[80]; + char sequence[16]; + char plant[4]; + unsigned short cpu_id; + unsigned short cpu_address; +}; + +struct sysinfo_2_2_2 { + char reserved_0[32]; + unsigned short lpar_number; + char reserved_1; + unsigned char characteristics; + unsigned short cpus_total; + unsigned short cpus_configured; + unsigned short cpus_standby; + unsigned short cpus_reserved; + char name[8]; + unsigned int caf; + char reserved_2[16]; + unsigned short cpus_dedicated; + unsigned short cpus_shared; +}; + +#define LPAR_CHAR_DEDICATED (1 << 7) +#define LPAR_CHAR_SHARED (1 << 6) +#define LPAR_CHAR_LIMITED (1 << 5) + +struct sysinfo_3_2_2 { + char reserved_0[31]; + unsigned char count; + struct { + char reserved_0[4]; + unsigned short cpus_total; + unsigned short cpus_configured; + unsigned short cpus_standby; + unsigned short cpus_reserved; + char name[8]; + unsigned int caf; + char cpi[16]; + char reserved_1[24]; + + } vm[8]; +}; + +static inline int stsi(void *sysinfo, int fc, int sel1, int sel2) +{ + register int r0 asm("0") = (fc << 28) | sel1; + register int r1 asm("1") = sel2; + + asm volatile( + " stsi 0(%2)\n" + "0: jz 2f\n" + "1: lhi %0,%3\n" + "2:\n" + EX_TABLE(0b, 1b) + : "+d" (r0) : "d" (r1), "a" (sysinfo), "K" (-ENOSYS) + : "cc", "memory"); + return r0; +} diff --git a/include/asm-s390/system.h b/include/asm-s390/system.h index 15aba30601a3..92098df4d6e3 100644 --- a/include/asm-s390/system.h +++ b/include/asm-s390/system.h @@ -406,6 +406,8 @@ __set_psw_mask(unsigned long mask) #define local_mcck_enable() __set_psw_mask(psw_kernel_bits) #define local_mcck_disable() __set_psw_mask(psw_kernel_bits & ~PSW_MASK_MCHECK) +int stfle(unsigned long long *list, int doublewords); + #ifdef CONFIG_SMP extern void smp_ctl_set_bit(int cr, int bit); diff --git a/include/asm-s390/timex.h b/include/asm-s390/timex.h index 98229db24314..d744c3d62de5 100644 --- a/include/asm-s390/timex.h +++ b/include/asm-s390/timex.h @@ -62,16 +62,18 @@ static inline unsigned long long get_clock (void) return clk; } -static inline void get_clock_extended(void *dest) +static inline unsigned long long get_clock_xt(void) { - typedef struct { unsigned long long clk[2]; } __clock_t; + unsigned char clk[16]; #if __GNUC__ > 3 || (__GNUC__ == 3 && __GNUC_MINOR__ > 2) - asm volatile("stcke %0" : "=Q" (*((__clock_t *)dest)) : : "cc"); + asm volatile("stcke %0" : "=Q" (clk) : : "cc"); #else /* __GNUC__ */ - asm volatile("stcke 0(%1)" : "=m" (*((__clock_t *)dest)) - : "a" ((__clock_t *)dest) : "cc"); + asm volatile("stcke 0(%1)" : "=m" (clk) + : "a" (clk) : "cc"); #endif /* __GNUC__ */ + + return *((unsigned long long *)&clk[1]); } static inline cycles_t get_cycles(void) @@ -81,5 +83,6 @@ static inline cycles_t get_cycles(void) int get_sync_clock(unsigned long long *clock); void init_cpu_timer(void); +unsigned long long monotonic_clock(void); #endif diff --git a/include/asm-s390/tlbflush.h b/include/asm-s390/tlbflush.h index 35fb4f9127b2..9e57a93d7de1 100644 --- a/include/asm-s390/tlbflush.h +++ b/include/asm-s390/tlbflush.h @@ -13,12 +13,14 @@ static inline void __tlb_flush_local(void) asm volatile("ptlb" : : : "memory"); } +#ifdef CONFIG_SMP /* * Flush all tlb entries on all cpus. */ +void smp_ptlb_all(void); + static inline void __tlb_flush_global(void) { - extern void smp_ptlb_all(void); register unsigned long reg2 asm("2"); register unsigned long reg3 asm("3"); register unsigned long reg4 asm("4"); @@ -39,6 +41,25 @@ static inline void __tlb_flush_global(void) : : "d" (reg2), "d" (reg3), "d" (reg4), "m" (dummy) : "cc" ); } +static inline void __tlb_flush_full(struct mm_struct *mm) +{ + cpumask_t local_cpumask; + + preempt_disable(); + /* + * If the process only ran on the local cpu, do a local flush. + */ + local_cpumask = cpumask_of_cpu(smp_processor_id()); + if (cpus_equal(mm->cpu_vm_mask, local_cpumask)) + __tlb_flush_local(); + else + __tlb_flush_global(); + preempt_enable(); +} +#else +#define __tlb_flush_full(mm) __tlb_flush_local() +#endif + /* * Flush all tlb entries of a page table on all cpus. */ @@ -51,8 +72,6 @@ static inline void __tlb_flush_idte(unsigned long asce) static inline void __tlb_flush_mm(struct mm_struct * mm) { - cpumask_t local_cpumask; - if (unlikely(cpus_empty(mm->cpu_vm_mask))) return; /* @@ -69,16 +88,7 @@ static inline void __tlb_flush_mm(struct mm_struct * mm) mm->context.asce_bits); return; } - preempt_disable(); - /* - * If the process only ran on the local cpu, do a local flush. - */ - local_cpumask = cpumask_of_cpu(smp_processor_id()); - if (cpus_equal(mm->cpu_vm_mask, local_cpumask)) - __tlb_flush_local(); - else - __tlb_flush_global(); - preempt_enable(); + __tlb_flush_full(mm); } static inline void __tlb_flush_mm_cond(struct mm_struct * mm) diff --git a/include/asm-s390/topology.h b/include/asm-s390/topology.h index 613aa64019da..8e97b06f298a 100644 --- a/include/asm-s390/topology.h +++ b/include/asm-s390/topology.h @@ -1,6 +1,29 @@ #ifndef _ASM_S390_TOPOLOGY_H #define _ASM_S390_TOPOLOGY_H +#include <linux/cpumask.h> + +#define mc_capable() (1) + +cpumask_t cpu_coregroup_map(unsigned int cpu); + +int topology_set_cpu_management(int fc); +void topology_schedule_update(void); + +#define POLARIZATION_UNKNWN (-1) +#define POLARIZATION_HRZ (0) +#define POLARIZATION_VL (1) +#define POLARIZATION_VM (2) +#define POLARIZATION_VH (3) + +#ifdef CONFIG_SMP +void s390_init_cpu_topology(void); +#else +static inline void s390_init_cpu_topology(void) +{ +}; +#endif + #include <asm-generic/topology.h> #endif /* _ASM_S390_TOPOLOGY_H */ diff --git a/include/asm-sh/semaphore-helper.h b/include/asm-sh/semaphore-helper.h deleted file mode 100644 index bd8230c369ca..000000000000 --- a/include/asm-sh/semaphore-helper.h +++ /dev/null @@ -1,89 +0,0 @@ -#ifndef __ASM_SH_SEMAPHORE_HELPER_H -#define __ASM_SH_SEMAPHORE_HELPER_H - -/* - * SMP- and interrupt-safe semaphores helper functions. - * - * (C) Copyright 1996 Linus Torvalds - * (C) Copyright 1999 Andrea Arcangeli - */ - -/* - * These two _must_ execute atomically wrt each other. - * - * This is trivially done with load_locked/store_cond, - * which we have. Let the rest of the losers suck eggs. - */ -static __inline__ void wake_one_more(struct semaphore * sem) -{ - atomic_inc((atomic_t *)&sem->sleepers); -} - -static __inline__ int waking_non_zero(struct semaphore *sem) -{ - unsigned long flags; - int ret = 0; - - spin_lock_irqsave(&semaphore_wake_lock, flags); - if (sem->sleepers > 0) { - sem->sleepers--; - ret = 1; - } - spin_unlock_irqrestore(&semaphore_wake_lock, flags); - return ret; -} - -/* - * waking_non_zero_interruptible: - * 1 got the lock - * 0 go to sleep - * -EINTR interrupted - * - * We must undo the sem->count down_interruptible() increment while we are - * protected by the spinlock in order to make atomic this atomic_inc() with the - * atomic_read() in wake_one_more(), otherwise we can race. -arca - */ -static __inline__ int waking_non_zero_interruptible(struct semaphore *sem, - struct task_struct *tsk) -{ - unsigned long flags; - int ret = 0; - - spin_lock_irqsave(&semaphore_wake_lock, flags); - if (sem->sleepers > 0) { - sem->sleepers--; - ret = 1; - } else if (signal_pending(tsk)) { - atomic_inc(&sem->count); - ret = -EINTR; - } - spin_unlock_irqrestore(&semaphore_wake_lock, flags); - return ret; -} - -/* - * waking_non_zero_trylock: - * 1 failed to lock - * 0 got the lock - * - * We must undo the sem->count down_trylock() increment while we are - * protected by the spinlock in order to make atomic this atomic_inc() with the - * atomic_read() in wake_one_more(), otherwise we can race. -arca - */ -static __inline__ int waking_non_zero_trylock(struct semaphore *sem) -{ - unsigned long flags; - int ret = 1; - - spin_lock_irqsave(&semaphore_wake_lock, flags); - if (sem->sleepers <= 0) - atomic_inc(&sem->count); - else { - sem->sleepers--; - ret = 0; - } - spin_unlock_irqrestore(&semaphore_wake_lock, flags); - return ret; -} - -#endif /* __ASM_SH_SEMAPHORE_HELPER_H */ diff --git a/include/asm-sh/semaphore.h b/include/asm-sh/semaphore.h index 9e5a37c4dce2..d9b2034ed1d2 100644 --- a/include/asm-sh/semaphore.h +++ b/include/asm-sh/semaphore.h @@ -1,115 +1 @@ -#ifndef __ASM_SH_SEMAPHORE_H -#define __ASM_SH_SEMAPHORE_H - -#include <linux/linkage.h> - -#ifdef __KERNEL__ -/* - * SMP- and interrupt-safe semaphores. - * - * (C) Copyright 1996 Linus Torvalds - * - * SuperH verison by Niibe Yutaka - * (Currently no asm implementation but generic C code...) - */ - -#include <linux/spinlock.h> -#include <linux/rwsem.h> -#include <linux/wait.h> - -#include <asm/system.h> -#include <asm/atomic.h> - -struct semaphore { - atomic_t count; - int sleepers; - wait_queue_head_t wait; -}; - -#define __SEMAPHORE_INITIALIZER(name, n) \ -{ \ - .count = ATOMIC_INIT(n), \ - .sleepers = 0, \ - .wait = __WAIT_QUEUE_HEAD_INITIALIZER((name).wait) \ -} - -#define __DECLARE_SEMAPHORE_GENERIC(name,count) \ - struct semaphore name = __SEMAPHORE_INITIALIZER(name,count) - -#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name,1) - -static inline void sema_init (struct semaphore *sem, int val) -{ -/* - * *sem = (struct semaphore)__SEMAPHORE_INITIALIZER((*sem),val); - * - * i'd rather use the more flexible initialization above, but sadly - * GCC 2.7.2.3 emits a bogus warning. EGCS doesn't. Oh well. - */ - atomic_set(&sem->count, val); - sem->sleepers = 0; - init_waitqueue_head(&sem->wait); -} - -static inline void init_MUTEX (struct semaphore *sem) -{ - sema_init(sem, 1); -} - -static inline void init_MUTEX_LOCKED (struct semaphore *sem) -{ - sema_init(sem, 0); -} - -#if 0 -asmlinkage void __down_failed(void /* special register calling convention */); -asmlinkage int __down_failed_interruptible(void /* params in registers */); -asmlinkage int __down_failed_trylock(void /* params in registers */); -asmlinkage void __up_wakeup(void /* special register calling convention */); -#endif - -asmlinkage void __down(struct semaphore * sem); -asmlinkage int __down_interruptible(struct semaphore * sem); -asmlinkage int __down_trylock(struct semaphore * sem); -asmlinkage void __up(struct semaphore * sem); - -extern spinlock_t semaphore_wake_lock; - -static inline void down(struct semaphore * sem) -{ - might_sleep(); - if (atomic_dec_return(&sem->count) < 0) - __down(sem); -} - -static inline int down_interruptible(struct semaphore * sem) -{ - int ret = 0; - - might_sleep(); - if (atomic_dec_return(&sem->count) < 0) - ret = __down_interruptible(sem); - return ret; -} - -static inline int down_trylock(struct semaphore * sem) -{ - int ret = 0; - - if (atomic_dec_return(&sem->count) < 0) - ret = __down_trylock(sem); - return ret; -} - -/* - * Note! This is subtle. We jump to wake people up only if - * the semaphore was negative (== somebody was waiting on it). - */ -static inline void up(struct semaphore * sem) -{ - if (atomic_inc_return(&sem->count) <= 0) - __up(sem); -} - -#endif -#endif /* __ASM_SH_SEMAPHORE_H */ +#include <linux/semaphore.h> diff --git a/include/asm-sparc/semaphore.h b/include/asm-sparc/semaphore.h index 8018f9f4d497..d9b2034ed1d2 100644 --- a/include/asm-sparc/semaphore.h +++ b/include/asm-sparc/semaphore.h @@ -1,192 +1 @@ -#ifndef _SPARC_SEMAPHORE_H -#define _SPARC_SEMAPHORE_H - -/* Dinky, good for nothing, just barely irq safe, Sparc semaphores. */ - -#ifdef __KERNEL__ - -#include <asm/atomic.h> -#include <linux/wait.h> -#include <linux/rwsem.h> - -struct semaphore { - atomic24_t count; - int sleepers; - wait_queue_head_t wait; -}; - -#define __SEMAPHORE_INITIALIZER(name, n) \ -{ \ - .count = ATOMIC24_INIT(n), \ - .sleepers = 0, \ - .wait = __WAIT_QUEUE_HEAD_INITIALIZER((name).wait) \ -} - -#define __DECLARE_SEMAPHORE_GENERIC(name,count) \ - struct semaphore name = __SEMAPHORE_INITIALIZER(name,count) - -#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name,1) - -static inline void sema_init (struct semaphore *sem, int val) -{ - atomic24_set(&sem->count, val); - sem->sleepers = 0; - init_waitqueue_head(&sem->wait); -} - -static inline void init_MUTEX (struct semaphore *sem) -{ - sema_init(sem, 1); -} - -static inline void init_MUTEX_LOCKED (struct semaphore *sem) -{ - sema_init(sem, 0); -} - -extern void __down(struct semaphore * sem); -extern int __down_interruptible(struct semaphore * sem); -extern int __down_trylock(struct semaphore * sem); -extern void __up(struct semaphore * sem); - -static inline void down(struct semaphore * sem) -{ - register volatile int *ptr asm("g1"); - register int increment asm("g2"); - - might_sleep(); - - ptr = &(sem->count.counter); - increment = 1; - - __asm__ __volatile__( - "mov %%o7, %%g4\n\t" - "call ___atomic24_sub\n\t" - " add %%o7, 8, %%o7\n\t" - "tst %%g2\n\t" - "bl 2f\n\t" - " nop\n" - "1:\n\t" - ".subsection 2\n" - "2:\n\t" - "save %%sp, -64, %%sp\n\t" - "mov %%g1, %%l1\n\t" - "mov %%g5, %%l5\n\t" - "call %3\n\t" - " mov %%g1, %%o0\n\t" - "mov %%l1, %%g1\n\t" - "ba 1b\n\t" - " restore %%l5, %%g0, %%g5\n\t" - ".previous\n" - : "=&r" (increment) - : "0" (increment), "r" (ptr), "i" (__down) - : "g3", "g4", "g7", "memory", "cc"); -} - -static inline int down_interruptible(struct semaphore * sem) -{ - register volatile int *ptr asm("g1"); - register int increment asm("g2"); - - might_sleep(); - - ptr = &(sem->count.counter); - increment = 1; - - __asm__ __volatile__( - "mov %%o7, %%g4\n\t" - "call ___atomic24_sub\n\t" - " add %%o7, 8, %%o7\n\t" - "tst %%g2\n\t" - "bl 2f\n\t" - " clr %%g2\n" - "1:\n\t" - ".subsection 2\n" - "2:\n\t" - "save %%sp, -64, %%sp\n\t" - "mov %%g1, %%l1\n\t" - "mov %%g5, %%l5\n\t" - "call %3\n\t" - " mov %%g1, %%o0\n\t" - "mov %%l1, %%g1\n\t" - "mov %%l5, %%g5\n\t" - "ba 1b\n\t" - " restore %%o0, %%g0, %%g2\n\t" - ".previous\n" - : "=&r" (increment) - : "0" (increment), "r" (ptr), "i" (__down_interruptible) - : "g3", "g4", "g7", "memory", "cc"); - - return increment; -} - -static inline int down_trylock(struct semaphore * sem) -{ - register volatile int *ptr asm("g1"); - register int increment asm("g2"); - - ptr = &(sem->count.counter); - increment = 1; - - __asm__ __volatile__( - "mov %%o7, %%g4\n\t" - "call ___atomic24_sub\n\t" - " add %%o7, 8, %%o7\n\t" - "tst %%g2\n\t" - "bl 2f\n\t" - " clr %%g2\n" - "1:\n\t" - ".subsection 2\n" - "2:\n\t" - "save %%sp, -64, %%sp\n\t" - "mov %%g1, %%l1\n\t" - "mov %%g5, %%l5\n\t" - "call %3\n\t" - " mov %%g1, %%o0\n\t" - "mov %%l1, %%g1\n\t" - "mov %%l5, %%g5\n\t" - "ba 1b\n\t" - " restore %%o0, %%g0, %%g2\n\t" - ".previous\n" - : "=&r" (increment) - : "0" (increment), "r" (ptr), "i" (__down_trylock) - : "g3", "g4", "g7", "memory", "cc"); - - return increment; -} - -static inline void up(struct semaphore * sem) -{ - register volatile int *ptr asm("g1"); - register int increment asm("g2"); - - ptr = &(sem->count.counter); - increment = 1; - - __asm__ __volatile__( - "mov %%o7, %%g4\n\t" - "call ___atomic24_add\n\t" - " add %%o7, 8, %%o7\n\t" - "tst %%g2\n\t" - "ble 2f\n\t" - " nop\n" - "1:\n\t" - ".subsection 2\n" - "2:\n\t" - "save %%sp, -64, %%sp\n\t" - "mov %%g1, %%l1\n\t" - "mov %%g5, %%l5\n\t" - "call %3\n\t" - " mov %%g1, %%o0\n\t" - "mov %%l1, %%g1\n\t" - "ba 1b\n\t" - " restore %%l5, %%g0, %%g5\n\t" - ".previous\n" - : "=&r" (increment) - : "0" (increment), "r" (ptr), "i" (__up) - : "g3", "g4", "g7", "memory", "cc"); -} - -#endif /* __KERNEL__ */ - -#endif /* !(_SPARC_SEMAPHORE_H) */ +#include <linux/semaphore.h> diff --git a/include/asm-sparc64/semaphore.h b/include/asm-sparc64/semaphore.h index 7f7c0c4e024f..d9b2034ed1d2 100644 --- a/include/asm-sparc64/semaphore.h +++ b/include/asm-sparc64/semaphore.h @@ -1,53 +1 @@ -#ifndef _SPARC64_SEMAPHORE_H -#define _SPARC64_SEMAPHORE_H - -/* These are actually reasonable on the V9. - * - * See asm-ppc/semaphore.h for implementation commentary, - * only sparc64 specific issues are commented here. - */ -#ifdef __KERNEL__ - -#include <asm/atomic.h> -#include <asm/system.h> -#include <linux/wait.h> -#include <linux/rwsem.h> - -struct semaphore { - atomic_t count; - wait_queue_head_t wait; -}; - -#define __SEMAPHORE_INITIALIZER(name, count) \ - { ATOMIC_INIT(count), \ - __WAIT_QUEUE_HEAD_INITIALIZER((name).wait) } - -#define __DECLARE_SEMAPHORE_GENERIC(name, count) \ - struct semaphore name = __SEMAPHORE_INITIALIZER(name,count) - -#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name, 1) - -static inline void sema_init (struct semaphore *sem, int val) -{ - atomic_set(&sem->count, val); - init_waitqueue_head(&sem->wait); -} - -static inline void init_MUTEX (struct semaphore *sem) -{ - sema_init(sem, 1); -} - -static inline void init_MUTEX_LOCKED (struct semaphore *sem) -{ - sema_init(sem, 0); -} - -extern void up(struct semaphore *sem); -extern void down(struct semaphore *sem); -extern int down_trylock(struct semaphore *sem); -extern int down_interruptible(struct semaphore *sem); - -#endif /* __KERNEL__ */ - -#endif /* !(_SPARC64_SEMAPHORE_H) */ +#include <linux/semaphore.h> diff --git a/include/asm-um/semaphore.h b/include/asm-um/semaphore.h index ff13c34de421..d9b2034ed1d2 100644 --- a/include/asm-um/semaphore.h +++ b/include/asm-um/semaphore.h @@ -1,6 +1 @@ -#ifndef __UM_SEMAPHORE_H -#define __UM_SEMAPHORE_H - -#include "asm/arch/semaphore.h" - -#endif +#include <linux/semaphore.h> diff --git a/include/asm-v850/semaphore.h b/include/asm-v850/semaphore.h index 10ed0ccf37df..d9b2034ed1d2 100644 --- a/include/asm-v850/semaphore.h +++ b/include/asm-v850/semaphore.h @@ -1,84 +1 @@ -#ifndef __V850_SEMAPHORE_H__ -#define __V850_SEMAPHORE_H__ - -#include <linux/linkage.h> -#include <linux/spinlock.h> -#include <linux/wait.h> -#include <linux/rwsem.h> - -#include <asm/atomic.h> - -struct semaphore { - atomic_t count; - int sleepers; - wait_queue_head_t wait; -}; - -#define __SEMAPHORE_INITIALIZER(name,count) \ - { ATOMIC_INIT (count), 0, \ - __WAIT_QUEUE_HEAD_INITIALIZER ((name).wait) } - -#define __DECLARE_SEMAPHORE_GENERIC(name,count) \ - struct semaphore name = __SEMAPHORE_INITIALIZER (name,count) - -#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC (name,1) - -static inline void sema_init (struct semaphore *sem, int val) -{ - *sem = (struct semaphore)__SEMAPHORE_INITIALIZER((*sem),val); -} - -static inline void init_MUTEX (struct semaphore *sem) -{ - sema_init (sem, 1); -} - -static inline void init_MUTEX_LOCKED (struct semaphore *sem) -{ - sema_init (sem, 0); -} - -/* - * special register calling convention - */ -asmlinkage void __down_failed (void); -asmlinkage int __down_interruptible_failed (void); -asmlinkage int __down_trylock_failed (void); -asmlinkage void __up_wakeup (void); - -extern void __down (struct semaphore * sem); -extern int __down_interruptible (struct semaphore * sem); -extern int __down_trylock (struct semaphore * sem); -extern void __up (struct semaphore * sem); - -static inline void down (struct semaphore * sem) -{ - might_sleep(); - if (atomic_dec_return (&sem->count) < 0) - __down (sem); -} - -static inline int down_interruptible (struct semaphore * sem) -{ - int ret = 0; - might_sleep(); - if (atomic_dec_return (&sem->count) < 0) - ret = __down_interruptible (sem); - return ret; -} - -static inline int down_trylock (struct semaphore *sem) -{ - int ret = 0; - if (atomic_dec_return (&sem->count) < 0) - ret = __down_trylock (sem); - return ret; -} - -static inline void up (struct semaphore * sem) -{ - if (atomic_inc_return (&sem->count) <= 0) - __up (sem); -} - -#endif /* __V850_SEMAPHORE_H__ */ +#include <linux/semaphore.h> diff --git a/include/asm-x86/Kbuild b/include/asm-x86/Kbuild index 3b8160a2b47e..1e3554596f72 100644 --- a/include/asm-x86/Kbuild +++ b/include/asm-x86/Kbuild @@ -10,6 +10,7 @@ header-y += prctl.h header-y += ptrace-abi.h header-y += sigcontext32.h header-y += ucontext.h +header-y += processor-flags.h unifdef-y += e820.h unifdef-y += ist.h diff --git a/include/asm-x86/a.out-core.h b/include/asm-x86/a.out-core.h index d2b6e11d3e97..714207a1c387 100644 --- a/include/asm-x86/a.out-core.h +++ b/include/asm-x86/a.out-core.h @@ -29,8 +29,9 @@ static inline void aout_dump_thread(struct pt_regs *regs, struct user *dump) dump->magic = CMAGIC; dump->start_code = 0; dump->start_stack = regs->sp & ~(PAGE_SIZE - 1); - dump->u_tsize = ((unsigned long) current->mm->end_code) >> PAGE_SHIFT; - dump->u_dsize = ((unsigned long) (current->mm->brk + (PAGE_SIZE-1))) >> PAGE_SHIFT; + dump->u_tsize = ((unsigned long)current->mm->end_code) >> PAGE_SHIFT; + dump->u_dsize = ((unsigned long)(current->mm->brk + (PAGE_SIZE - 1))) + >> PAGE_SHIFT; dump->u_dsize -= dump->u_tsize; dump->u_ssize = 0; dump->u_debugreg[0] = current->thread.debugreg0; @@ -43,7 +44,8 @@ static inline void aout_dump_thread(struct pt_regs *regs, struct user *dump) dump->u_debugreg[7] = current->thread.debugreg7; if (dump->start_stack < TASK_SIZE) - dump->u_ssize = ((unsigned long) (TASK_SIZE - dump->start_stack)) >> PAGE_SHIFT; + dump->u_ssize = ((unsigned long)(TASK_SIZE - dump->start_stack)) + >> PAGE_SHIFT; dump->regs.bx = regs->bx; dump->regs.cx = regs->cx; @@ -55,7 +57,7 @@ static inline void aout_dump_thread(struct pt_regs *regs, struct user *dump) dump->regs.ds = (u16)regs->ds; dump->regs.es = (u16)regs->es; dump->regs.fs = (u16)regs->fs; - savesegment(gs,gs); + savesegment(gs, gs); dump->regs.orig_ax = regs->orig_ax; dump->regs.ip = regs->ip; dump->regs.cs = (u16)regs->cs; @@ -63,7 +65,7 @@ static inline void aout_dump_thread(struct pt_regs *regs, struct user *dump) dump->regs.sp = regs->sp; dump->regs.ss = (u16)regs->ss; - dump->u_fpvalid = dump_fpu (regs, &dump->i387); + dump->u_fpvalid = dump_fpu(regs, &dump->i387); } #endif /* CONFIG_X86_32 */ diff --git a/include/asm-x86/acpi.h b/include/asm-x86/acpi.h index 7a72d6aa50be..14411c9de46f 100644 --- a/include/asm-x86/acpi.h +++ b/include/asm-x86/acpi.h @@ -67,16 +67,16 @@ int __acpi_release_global_lock(unsigned int *lock); */ #define ACPI_DIV_64_BY_32(n_hi, n_lo, d32, q32, r32) \ asm("divl %2;" \ - :"=a"(q32), "=d"(r32) \ - :"r"(d32), \ + : "=a"(q32), "=d"(r32) \ + : "r"(d32), \ "0"(n_lo), "1"(n_hi)) #define ACPI_SHIFT_RIGHT_64(n_hi, n_lo) \ asm("shrl $1,%2 ;" \ "rcrl $1,%3;" \ - :"=r"(n_hi), "=r"(n_lo) \ - :"0"(n_hi), "1"(n_lo)) + : "=r"(n_hi), "=r"(n_lo) \ + : "0"(n_hi), "1"(n_lo)) #ifdef CONFIG_ACPI extern int acpi_lapic; diff --git a/include/asm-x86/alternative.h b/include/asm-x86/alternative.h index d8bacf3c4b08..1f6a9ca10126 100644 --- a/include/asm-x86/alternative.h +++ b/include/asm-x86/alternative.h @@ -66,8 +66,8 @@ extern void alternatives_smp_module_del(struct module *mod); extern void alternatives_smp_switch(int smp); #else static inline void alternatives_smp_module_add(struct module *mod, char *name, - void *locks, void *locks_end, - void *text, void *text_end) {} + void *locks, void *locks_end, + void *text, void *text_end) {} static inline void alternatives_smp_module_del(struct module *mod) {} static inline void alternatives_smp_switch(int smp) {} #endif /* CONFIG_SMP */ @@ -148,14 +148,34 @@ struct paravirt_patch_site; void apply_paravirt(struct paravirt_patch_site *start, struct paravirt_patch_site *end); #else -static inline void -apply_paravirt(struct paravirt_patch_site *start, - struct paravirt_patch_site *end) +static inline void apply_paravirt(struct paravirt_patch_site *start, + struct paravirt_patch_site *end) {} #define __parainstructions NULL #define __parainstructions_end NULL #endif -extern void text_poke(void *addr, unsigned char *opcode, int len); +extern void add_nops(void *insns, unsigned int len); + +/* + * Clear and restore the kernel write-protection flag on the local CPU. + * Allows the kernel to edit read-only pages. + * Side-effect: any interrupt handler running between save and restore will have + * the ability to write to read-only pages. + * + * Warning: + * Code patching in the UP case is safe if NMIs and MCE handlers are stopped and + * no thread can be preempted in the instructions being modified (no iret to an + * invalid instruction possible) or if the instructions are changed from a + * consistent state to another consistent state atomically. + * More care must be taken when modifying code in the SMP case because of + * Intel's errata. + * On the local CPU you need to be protected again NMI or MCE handlers seeing an + * inconsistent instruction while you patch. + * The _early version expects the memory to already be RW. + */ + +extern void *text_poke(void *addr, const void *opcode, size_t len); +extern void *text_poke_early(void *addr, const void *opcode, size_t len); #endif /* _ASM_X86_ALTERNATIVE_H */ diff --git a/include/asm-x86/apic.h b/include/asm-x86/apic.h index bcfc07fd3661..be9639a9a186 100644 --- a/include/asm-x86/apic.h +++ b/include/asm-x86/apic.h @@ -44,7 +44,6 @@ extern int apic_runs_main_timer; extern int ioapic_force; extern int disable_apic; extern int disable_apic_timer; -extern unsigned boot_cpu_id; /* * Basic functions accessing APICs. @@ -59,6 +58,8 @@ extern unsigned boot_cpu_id; #define setup_secondary_clock setup_secondary_APIC_clock #endif +extern int is_vsmp_box(void); + static inline void native_apic_write(unsigned long reg, u32 v) { *((volatile u32 *)(APIC_BASE + reg)) = v; @@ -66,7 +67,7 @@ static inline void native_apic_write(unsigned long reg, u32 v) static inline void native_apic_write_atomic(unsigned long reg, u32 v) { - (void) xchg((u32*)(APIC_BASE + reg), v); + (void)xchg((u32 *)(APIC_BASE + reg), v); } static inline u32 native_apic_read(unsigned long reg) @@ -123,7 +124,7 @@ extern void enable_NMI_through_LVT0(void); * On 32bit this is mach-xxx local */ #ifdef CONFIG_X86_64 -extern void setup_apic_routing(void); +extern void early_init_lapic_mapping(void); #endif extern u8 setup_APIC_eilvt_mce(u8 vector, u8 msg_type, u8 mask); diff --git a/include/asm-x86/apicdef.h b/include/asm-x86/apicdef.h index 550af7a6f88e..6b9008c78731 100644 --- a/include/asm-x86/apicdef.h +++ b/include/asm-x86/apicdef.h @@ -12,17 +12,15 @@ #define APIC_ID 0x20 -#ifdef CONFIG_X86_64 -# define APIC_ID_MASK (0xFFu<<24) -# define GET_APIC_ID(x) (((x)>>24)&0xFFu) -# define SET_APIC_ID(x) (((x)<<24)) -#endif - #define APIC_LVR 0x30 #define APIC_LVR_MASK 0xFF00FF -#define GET_APIC_VERSION(x) ((x)&0xFFu) -#define GET_APIC_MAXLVT(x) (((x)>>16)&0xFFu) -#define APIC_INTEGRATED(x) ((x)&0xF0u) +#define GET_APIC_VERSION(x) ((x) & 0xFFu) +#define GET_APIC_MAXLVT(x) (((x) >> 16) & 0xFFu) +#ifdef CONFIG_X86_32 +# define APIC_INTEGRATED(x) ((x) & 0xF0u) +#else +# define APIC_INTEGRATED(x) (1) +#endif #define APIC_XAPIC(x) ((x) >= 0x14) #define APIC_TASKPRI 0x80 #define APIC_TPRI_MASK 0xFFu @@ -33,16 +31,16 @@ #define APIC_EIO_ACK 0x0 #define APIC_RRR 0xC0 #define APIC_LDR 0xD0 -#define APIC_LDR_MASK (0xFFu<<24) -#define GET_APIC_LOGICAL_ID(x) (((x)>>24)&0xFFu) -#define SET_APIC_LOGICAL_ID(x) (((x)<<24)) +#define APIC_LDR_MASK (0xFFu << 24) +#define GET_APIC_LOGICAL_ID(x) (((x) >> 24) & 0xFFu) +#define SET_APIC_LOGICAL_ID(x) (((x) << 24)) #define APIC_ALL_CPUS 0xFFu #define APIC_DFR 0xE0 #define APIC_DFR_CLUSTER 0x0FFFFFFFul #define APIC_DFR_FLAT 0xFFFFFFFFul #define APIC_SPIV 0xF0 -#define APIC_SPIV_FOCUS_DISABLED (1<<9) -#define APIC_SPIV_APIC_ENABLED (1<<8) +#define APIC_SPIV_FOCUS_DISABLED (1 << 9) +#define APIC_SPIV_APIC_ENABLED (1 << 8) #define APIC_ISR 0x100 #define APIC_ISR_NR 0x8 /* Number of 32 bit ISR registers. */ #define APIC_TMR 0x180 @@ -78,27 +76,27 @@ #define APIC_DM_EXTINT 0x00700 #define APIC_VECTOR_MASK 0x000FF #define APIC_ICR2 0x310 -#define GET_APIC_DEST_FIELD(x) (((x)>>24)&0xFF) -#define SET_APIC_DEST_FIELD(x) ((x)<<24) +#define GET_APIC_DEST_FIELD(x) (((x) >> 24) & 0xFF) +#define SET_APIC_DEST_FIELD(x) ((x) << 24) #define APIC_LVTT 0x320 #define APIC_LVTTHMR 0x330 #define APIC_LVTPC 0x340 #define APIC_LVT0 0x350 -#define APIC_LVT_TIMER_BASE_MASK (0x3<<18) -#define GET_APIC_TIMER_BASE(x) (((x)>>18)&0x3) -#define SET_APIC_TIMER_BASE(x) (((x)<<18)) +#define APIC_LVT_TIMER_BASE_MASK (0x3 << 18) +#define GET_APIC_TIMER_BASE(x) (((x) >> 18) & 0x3) +#define SET_APIC_TIMER_BASE(x) (((x) << 18)) #define APIC_TIMER_BASE_CLKIN 0x0 #define APIC_TIMER_BASE_TMBASE 0x1 #define APIC_TIMER_BASE_DIV 0x2 -#define APIC_LVT_TIMER_PERIODIC (1<<17) -#define APIC_LVT_MASKED (1<<16) -#define APIC_LVT_LEVEL_TRIGGER (1<<15) -#define APIC_LVT_REMOTE_IRR (1<<14) -#define APIC_INPUT_POLARITY (1<<13) -#define APIC_SEND_PENDING (1<<12) +#define APIC_LVT_TIMER_PERIODIC (1 << 17) +#define APIC_LVT_MASKED (1 << 16) +#define APIC_LVT_LEVEL_TRIGGER (1 << 15) +#define APIC_LVT_REMOTE_IRR (1 << 14) +#define APIC_INPUT_POLARITY (1 << 13) +#define APIC_SEND_PENDING (1 << 12) #define APIC_MODE_MASK 0x700 -#define GET_APIC_DELIVERY_MODE(x) (((x)>>8)&0x7) -#define SET_APIC_DELIVERY_MODE(x, y) (((x)&~0x700)|((y)<<8)) +#define GET_APIC_DELIVERY_MODE(x) (((x) >> 8) & 0x7) +#define SET_APIC_DELIVERY_MODE(x, y) (((x) & ~0x700) | ((y) << 8)) #define APIC_MODE_FIXED 0x0 #define APIC_MODE_NMI 0x4 #define APIC_MODE_EXTINT 0x7 @@ -107,7 +105,7 @@ #define APIC_TMICT 0x380 #define APIC_TMCCT 0x390 #define APIC_TDCR 0x3E0 -#define APIC_TDR_DIV_TMBASE (1<<2) +#define APIC_TDR_DIV_TMBASE (1 << 2) #define APIC_TDR_DIV_1 0xB #define APIC_TDR_DIV_2 0x0 #define APIC_TDR_DIV_4 0x1 @@ -117,14 +115,14 @@ #define APIC_TDR_DIV_64 0x9 #define APIC_TDR_DIV_128 0xA #define APIC_EILVT0 0x500 -#define APIC_EILVT_NR_AMD_K8 1 /* Number of extended interrupts */ +#define APIC_EILVT_NR_AMD_K8 1 /* # of extended interrupts */ #define APIC_EILVT_NR_AMD_10H 4 -#define APIC_EILVT_LVTOFF(x) (((x)>>4)&0xF) +#define APIC_EILVT_LVTOFF(x) (((x) >> 4) & 0xF) #define APIC_EILVT_MSG_FIX 0x0 #define APIC_EILVT_MSG_SMI 0x2 #define APIC_EILVT_MSG_NMI 0x4 #define APIC_EILVT_MSG_EXT 0x7 -#define APIC_EILVT_MASKED (1<<16) +#define APIC_EILVT_MASKED (1 << 16) #define APIC_EILVT1 0x510 #define APIC_EILVT2 0x520 #define APIC_EILVT3 0x530 @@ -135,7 +133,7 @@ # define MAX_IO_APICS 64 #else # define MAX_IO_APICS 128 -# define MAX_LOCAL_APIC 256 +# define MAX_LOCAL_APIC 32768 #endif /* @@ -408,6 +406,9 @@ struct local_apic { #undef u32 -#define BAD_APICID 0xFFu - +#ifdef CONFIG_X86_32 + #define BAD_APICID 0xFFu +#else + #define BAD_APICID 0xFFFFu +#endif #endif diff --git a/include/asm-x86/atomic_32.h b/include/asm-x86/atomic_32.h index 437aac801711..21a4825148c0 100644 --- a/include/asm-x86/atomic_32.h +++ b/include/asm-x86/atomic_32.h @@ -15,138 +15,133 @@ * on us. We need to use _exactly_ the address the user gave us, * not some alias that contains the same information. */ -typedef struct { int counter; } atomic_t; +typedef struct { + int counter; +} atomic_t; #define ATOMIC_INIT(i) { (i) } /** * atomic_read - read atomic variable * @v: pointer of type atomic_t - * + * * Atomically reads the value of @v. - */ + */ #define atomic_read(v) ((v)->counter) /** * atomic_set - set atomic variable * @v: pointer of type atomic_t * @i: required value - * + * * Atomically sets the value of @v to @i. - */ -#define atomic_set(v,i) (((v)->counter) = (i)) + */ +#define atomic_set(v, i) (((v)->counter) = (i)) /** * atomic_add - add integer to atomic variable * @i: integer value to add * @v: pointer of type atomic_t - * + * * Atomically adds @i to @v. */ -static __inline__ void atomic_add(int i, atomic_t *v) +static inline void atomic_add(int i, atomic_t *v) { - __asm__ __volatile__( - LOCK_PREFIX "addl %1,%0" - :"+m" (v->counter) - :"ir" (i)); + asm volatile(LOCK_PREFIX "addl %1,%0" + : "+m" (v->counter) + : "ir" (i)); } /** * atomic_sub - subtract integer from atomic variable * @i: integer value to subtract * @v: pointer of type atomic_t - * + * * Atomically subtracts @i from @v. */ -static __inline__ void atomic_sub(int i, atomic_t *v) +static inline void atomic_sub(int i, atomic_t *v) { - __asm__ __volatile__( - LOCK_PREFIX "subl %1,%0" - :"+m" (v->counter) - :"ir" (i)); + asm volatile(LOCK_PREFIX "subl %1,%0" + : "+m" (v->counter) + : "ir" (i)); } /** * atomic_sub_and_test - subtract value from variable and test result * @i: integer value to subtract * @v: pointer of type atomic_t - * + * * Atomically subtracts @i from @v and returns * true if the result is zero, or false for all * other cases. */ -static __inline__ int atomic_sub_and_test(int i, atomic_t *v) +static inline int atomic_sub_and_test(int i, atomic_t *v) { unsigned char c; - __asm__ __volatile__( - LOCK_PREFIX "subl %2,%0; sete %1" - :"+m" (v->counter), "=qm" (c) - :"ir" (i) : "memory"); + asm volatile(LOCK_PREFIX "subl %2,%0; sete %1" + : "+m" (v->counter), "=qm" (c) + : "ir" (i) : "memory"); return c; } /** * atomic_inc - increment atomic variable * @v: pointer of type atomic_t - * + * * Atomically increments @v by 1. - */ -static __inline__ void atomic_inc(atomic_t *v) + */ +static inline void atomic_inc(atomic_t *v) { - __asm__ __volatile__( - LOCK_PREFIX "incl %0" - :"+m" (v->counter)); + asm volatile(LOCK_PREFIX "incl %0" + : "+m" (v->counter)); } /** * atomic_dec - decrement atomic variable * @v: pointer of type atomic_t - * + * * Atomically decrements @v by 1. - */ -static __inline__ void atomic_dec(atomic_t *v) + */ +static inline void atomic_dec(atomic_t *v) { - __asm__ __volatile__( - LOCK_PREFIX "decl %0" - :"+m" (v->counter)); + asm volatile(LOCK_PREFIX "decl %0" + : "+m" (v->counter)); } /** * atomic_dec_and_test - decrement and test * @v: pointer of type atomic_t - * + * * Atomically decrements @v by 1 and * returns true if the result is 0, or false for all other * cases. - */ -static __inline__ int atomic_dec_and_test(atomic_t *v) + */ +static inline int atomic_dec_and_test(atomic_t *v) { unsigned char c; - __asm__ __volatile__( - LOCK_PREFIX "decl %0; sete %1" - :"+m" (v->counter), "=qm" (c) - : : "memory"); + asm volatile(LOCK_PREFIX "decl %0; sete %1" + : "+m" (v->counter), "=qm" (c) + : : "memory"); return c != 0; } /** - * atomic_inc_and_test - increment and test + * atomic_inc_and_test - increment and test * @v: pointer of type atomic_t - * + * * Atomically increments @v by 1 * and returns true if the result is zero, or false for all * other cases. - */ -static __inline__ int atomic_inc_and_test(atomic_t *v) + */ +static inline int atomic_inc_and_test(atomic_t *v) { unsigned char c; - __asm__ __volatile__( - LOCK_PREFIX "incl %0; sete %1" - :"+m" (v->counter), "=qm" (c) - : : "memory"); + asm volatile(LOCK_PREFIX "incl %0; sete %1" + : "+m" (v->counter), "=qm" (c) + : : "memory"); return c != 0; } @@ -154,19 +149,18 @@ static __inline__ int atomic_inc_and_test(atomic_t *v) * atomic_add_negative - add and test if negative * @v: pointer of type atomic_t * @i: integer value to add - * + * * Atomically adds @i to @v and returns true * if the result is negative, or false when * result is greater than or equal to zero. - */ -static __inline__ int atomic_add_negative(int i, atomic_t *v) + */ +static inline int atomic_add_negative(int i, atomic_t *v) { unsigned char c; - __asm__ __volatile__( - LOCK_PREFIX "addl %2,%0; sets %1" - :"+m" (v->counter), "=qm" (c) - :"ir" (i) : "memory"); + asm volatile(LOCK_PREFIX "addl %2,%0; sets %1" + : "+m" (v->counter), "=qm" (c) + : "ir" (i) : "memory"); return c; } @@ -177,20 +171,19 @@ static __inline__ int atomic_add_negative(int i, atomic_t *v) * * Atomically adds @i to @v and returns @i + @v */ -static __inline__ int atomic_add_return(int i, atomic_t *v) +static inline int atomic_add_return(int i, atomic_t *v) { int __i; #ifdef CONFIG_M386 unsigned long flags; - if(unlikely(boot_cpu_data.x86 <= 3)) + if (unlikely(boot_cpu_data.x86 <= 3)) goto no_xadd; #endif /* Modern 486+ processor */ __i = i; - __asm__ __volatile__( - LOCK_PREFIX "xaddl %0, %1" - :"+r" (i), "+m" (v->counter) - : : "memory"); + asm volatile(LOCK_PREFIX "xaddl %0, %1" + : "+r" (i), "+m" (v->counter) + : : "memory"); return i + __i; #ifdef CONFIG_M386 @@ -210,9 +203,9 @@ no_xadd: /* Legacy 386 processor */ * * Atomically subtracts @i from @v and returns @v - @i */ -static __inline__ int atomic_sub_return(int i, atomic_t *v) +static inline int atomic_sub_return(int i, atomic_t *v) { - return atomic_add_return(-i,v); + return atomic_add_return(-i, v); } #define atomic_cmpxchg(v, old, new) (cmpxchg(&((v)->counter), (old), (new))) @@ -227,7 +220,7 @@ static __inline__ int atomic_sub_return(int i, atomic_t *v) * Atomically adds @a to @v, so long as @v was not already @u. * Returns non-zero if @v was not @u, and zero otherwise. */ -static __inline__ int atomic_add_unless(atomic_t *v, int a, int u) +static inline int atomic_add_unless(atomic_t *v, int a, int u) { int c, old; c = atomic_read(v); @@ -244,17 +237,17 @@ static __inline__ int atomic_add_unless(atomic_t *v, int a, int u) #define atomic_inc_not_zero(v) atomic_add_unless((v), 1, 0) -#define atomic_inc_return(v) (atomic_add_return(1,v)) -#define atomic_dec_return(v) (atomic_sub_return(1,v)) +#define atomic_inc_return(v) (atomic_add_return(1, v)) +#define atomic_dec_return(v) (atomic_sub_return(1, v)) /* These are x86-specific, used by some header files */ -#define atomic_clear_mask(mask, addr) \ -__asm__ __volatile__(LOCK_PREFIX "andl %0,%1" \ -: : "r" (~(mask)),"m" (*addr) : "memory") +#define atomic_clear_mask(mask, addr) \ + asm volatile(LOCK_PREFIX "andl %0,%1" \ + : : "r" (~(mask)), "m" (*(addr)) : "memory") -#define atomic_set_mask(mask, addr) \ -__asm__ __volatile__(LOCK_PREFIX "orl %0,%1" \ -: : "r" (mask),"m" (*(addr)) : "memory") +#define atomic_set_mask(mask, addr) \ + asm volatile(LOCK_PREFIX "orl %0,%1" \ + : : "r" (mask), "m" (*(addr)) : "memory") /* Atomic operations are already serializing on x86 */ #define smp_mb__before_atomic_dec() barrier() diff --git a/include/asm-x86/atomic_64.h b/include/asm-x86/atomic_64.h index 2d20a7a19f62..3e0cd7d38335 100644 --- a/include/asm-x86/atomic_64.h +++ b/include/asm-x86/atomic_64.h @@ -22,140 +22,135 @@ * on us. We need to use _exactly_ the address the user gave us, * not some alias that contains the same information. */ -typedef struct { int counter; } atomic_t; +typedef struct { + int counter; +} atomic_t; #define ATOMIC_INIT(i) { (i) } /** * atomic_read - read atomic variable * @v: pointer of type atomic_t - * + * * Atomically reads the value of @v. - */ + */ #define atomic_read(v) ((v)->counter) /** * atomic_set - set atomic variable * @v: pointer of type atomic_t * @i: required value - * + * * Atomically sets the value of @v to @i. - */ -#define atomic_set(v,i) (((v)->counter) = (i)) + */ +#define atomic_set(v, i) (((v)->counter) = (i)) /** * atomic_add - add integer to atomic variable * @i: integer value to add * @v: pointer of type atomic_t - * + * * Atomically adds @i to @v. */ -static __inline__ void atomic_add(int i, atomic_t *v) +static inline void atomic_add(int i, atomic_t *v) { - __asm__ __volatile__( - LOCK_PREFIX "addl %1,%0" - :"=m" (v->counter) - :"ir" (i), "m" (v->counter)); + asm volatile(LOCK_PREFIX "addl %1,%0" + : "=m" (v->counter) + : "ir" (i), "m" (v->counter)); } /** * atomic_sub - subtract the atomic variable * @i: integer value to subtract * @v: pointer of type atomic_t - * + * * Atomically subtracts @i from @v. */ -static __inline__ void atomic_sub(int i, atomic_t *v) +static inline void atomic_sub(int i, atomic_t *v) { - __asm__ __volatile__( - LOCK_PREFIX "subl %1,%0" - :"=m" (v->counter) - :"ir" (i), "m" (v->counter)); + asm volatile(LOCK_PREFIX "subl %1,%0" + : "=m" (v->counter) + : "ir" (i), "m" (v->counter)); } /** * atomic_sub_and_test - subtract value from variable and test result * @i: integer value to subtract * @v: pointer of type atomic_t - * + * * Atomically subtracts @i from @v and returns * true if the result is zero, or false for all * other cases. */ -static __inline__ int atomic_sub_and_test(int i, atomic_t *v) +static inline int atomic_sub_and_test(int i, atomic_t *v) { unsigned char c; - __asm__ __volatile__( - LOCK_PREFIX "subl %2,%0; sete %1" - :"=m" (v->counter), "=qm" (c) - :"ir" (i), "m" (v->counter) : "memory"); + asm volatile(LOCK_PREFIX "subl %2,%0; sete %1" + : "=m" (v->counter), "=qm" (c) + : "ir" (i), "m" (v->counter) : "memory"); return c; } /** * atomic_inc - increment atomic variable * @v: pointer of type atomic_t - * + * * Atomically increments @v by 1. - */ -static __inline__ void atomic_inc(atomic_t *v) + */ +static inline void atomic_inc(atomic_t *v) { - __asm__ __volatile__( - LOCK_PREFIX "incl %0" - :"=m" (v->counter) - :"m" (v->counter)); + asm volatile(LOCK_PREFIX "incl %0" + : "=m" (v->counter) + : "m" (v->counter)); } /** * atomic_dec - decrement atomic variable * @v: pointer of type atomic_t - * + * * Atomically decrements @v by 1. - */ -static __inline__ void atomic_dec(atomic_t *v) + */ +static inline void atomic_dec(atomic_t *v) { - __asm__ __volatile__( - LOCK_PREFIX "decl %0" - :"=m" (v->counter) - :"m" (v->counter)); + asm volatile(LOCK_PREFIX "decl %0" + : "=m" (v->counter) + : "m" (v->counter)); } /** * atomic_dec_and_test - decrement and test * @v: pointer of type atomic_t - * + * * Atomically decrements @v by 1 and * returns true if the result is 0, or false for all other * cases. - */ -static __inline__ int atomic_dec_and_test(atomic_t *v) + */ +static inline int atomic_dec_and_test(atomic_t *v) { unsigned char c; - __asm__ __volatile__( - LOCK_PREFIX "decl %0; sete %1" - :"=m" (v->counter), "=qm" (c) - :"m" (v->counter) : "memory"); + asm volatile(LOCK_PREFIX "decl %0; sete %1" + : "=m" (v->counter), "=qm" (c) + : "m" (v->counter) : "memory"); return c != 0; } /** - * atomic_inc_and_test - increment and test + * atomic_inc_and_test - increment and test * @v: pointer of type atomic_t - * + * * Atomically increments @v by 1 * and returns true if the result is zero, or false for all * other cases. - */ -static __inline__ int atomic_inc_and_test(atomic_t *v) + */ +static inline int atomic_inc_and_test(atomic_t *v) { unsigned char c; - __asm__ __volatile__( - LOCK_PREFIX "incl %0; sete %1" - :"=m" (v->counter), "=qm" (c) - :"m" (v->counter) : "memory"); + asm volatile(LOCK_PREFIX "incl %0; sete %1" + : "=m" (v->counter), "=qm" (c) + : "m" (v->counter) : "memory"); return c != 0; } @@ -163,19 +158,18 @@ static __inline__ int atomic_inc_and_test(atomic_t *v) * atomic_add_negative - add and test if negative * @i: integer value to add * @v: pointer of type atomic_t - * + * * Atomically adds @i to @v and returns true * if the result is negative, or false when * result is greater than or equal to zero. - */ -static __inline__ int atomic_add_negative(int i, atomic_t *v) + */ +static inline int atomic_add_negative(int i, atomic_t *v) { unsigned char c; - __asm__ __volatile__( - LOCK_PREFIX "addl %2,%0; sets %1" - :"=m" (v->counter), "=qm" (c) - :"ir" (i), "m" (v->counter) : "memory"); + asm volatile(LOCK_PREFIX "addl %2,%0; sets %1" + : "=m" (v->counter), "=qm" (c) + : "ir" (i), "m" (v->counter) : "memory"); return c; } @@ -186,27 +180,28 @@ static __inline__ int atomic_add_negative(int i, atomic_t *v) * * Atomically adds @i to @v and returns @i + @v */ -static __inline__ int atomic_add_return(int i, atomic_t *v) +static inline int atomic_add_return(int i, atomic_t *v) { int __i = i; - __asm__ __volatile__( - LOCK_PREFIX "xaddl %0, %1" - :"+r" (i), "+m" (v->counter) - : : "memory"); + asm volatile(LOCK_PREFIX "xaddl %0, %1" + : "+r" (i), "+m" (v->counter) + : : "memory"); return i + __i; } -static __inline__ int atomic_sub_return(int i, atomic_t *v) +static inline int atomic_sub_return(int i, atomic_t *v) { - return atomic_add_return(-i,v); + return atomic_add_return(-i, v); } -#define atomic_inc_return(v) (atomic_add_return(1,v)) -#define atomic_dec_return(v) (atomic_sub_return(1,v)) +#define atomic_inc_return(v) (atomic_add_return(1, v)) +#define atomic_dec_return(v) (atomic_sub_return(1, v)) /* An 64bit atomic type */ -typedef struct { long counter; } atomic64_t; +typedef struct { + long counter; +} atomic64_t; #define ATOMIC64_INIT(i) { (i) } @@ -226,7 +221,7 @@ typedef struct { long counter; } atomic64_t; * * Atomically sets the value of @v to @i. */ -#define atomic64_set(v,i) (((v)->counter) = (i)) +#define atomic64_set(v, i) (((v)->counter) = (i)) /** * atomic64_add - add integer to atomic64 variable @@ -235,12 +230,11 @@ typedef struct { long counter; } atomic64_t; * * Atomically adds @i to @v. */ -static __inline__ void atomic64_add(long i, atomic64_t *v) +static inline void atomic64_add(long i, atomic64_t *v) { - __asm__ __volatile__( - LOCK_PREFIX "addq %1,%0" - :"=m" (v->counter) - :"ir" (i), "m" (v->counter)); + asm volatile(LOCK_PREFIX "addq %1,%0" + : "=m" (v->counter) + : "ir" (i), "m" (v->counter)); } /** @@ -250,12 +244,11 @@ static __inline__ void atomic64_add(long i, atomic64_t *v) * * Atomically subtracts @i from @v. */ -static __inline__ void atomic64_sub(long i, atomic64_t *v) +static inline void atomic64_sub(long i, atomic64_t *v) { - __asm__ __volatile__( - LOCK_PREFIX "subq %1,%0" - :"=m" (v->counter) - :"ir" (i), "m" (v->counter)); + asm volatile(LOCK_PREFIX "subq %1,%0" + : "=m" (v->counter) + : "ir" (i), "m" (v->counter)); } /** @@ -267,14 +260,13 @@ static __inline__ void atomic64_sub(long i, atomic64_t *v) * true if the result is zero, or false for all * other cases. */ -static __inline__ int atomic64_sub_and_test(long i, atomic64_t *v) +static inline int atomic64_sub_and_test(long i, atomic64_t *v) { unsigned char c; - __asm__ __volatile__( - LOCK_PREFIX "subq %2,%0; sete %1" - :"=m" (v->counter), "=qm" (c) - :"ir" (i), "m" (v->counter) : "memory"); + asm volatile(LOCK_PREFIX "subq %2,%0; sete %1" + : "=m" (v->counter), "=qm" (c) + : "ir" (i), "m" (v->counter) : "memory"); return c; } @@ -284,12 +276,11 @@ static __inline__ int atomic64_sub_and_test(long i, atomic64_t *v) * * Atomically increments @v by 1. */ -static __inline__ void atomic64_inc(atomic64_t *v) +static inline void atomic64_inc(atomic64_t *v) { - __asm__ __volatile__( - LOCK_PREFIX "incq %0" - :"=m" (v->counter) - :"m" (v->counter)); + asm volatile(LOCK_PREFIX "incq %0" + : "=m" (v->counter) + : "m" (v->counter)); } /** @@ -298,12 +289,11 @@ static __inline__ void atomic64_inc(atomic64_t *v) * * Atomically decrements @v by 1. */ -static __inline__ void atomic64_dec(atomic64_t *v) +static inline void atomic64_dec(atomic64_t *v) { - __asm__ __volatile__( - LOCK_PREFIX "decq %0" - :"=m" (v->counter) - :"m" (v->counter)); + asm volatile(LOCK_PREFIX "decq %0" + : "=m" (v->counter) + : "m" (v->counter)); } /** @@ -314,14 +304,13 @@ static __inline__ void atomic64_dec(atomic64_t *v) * returns true if the result is 0, or false for all other * cases. */ -static __inline__ int atomic64_dec_and_test(atomic64_t *v) +static inline int atomic64_dec_and_test(atomic64_t *v) { unsigned char c; - __asm__ __volatile__( - LOCK_PREFIX "decq %0; sete %1" - :"=m" (v->counter), "=qm" (c) - :"m" (v->counter) : "memory"); + asm volatile(LOCK_PREFIX "decq %0; sete %1" + : "=m" (v->counter), "=qm" (c) + : "m" (v->counter) : "memory"); return c != 0; } @@ -333,14 +322,13 @@ static __inline__ int atomic64_dec_and_test(atomic64_t *v) * and returns true if the result is zero, or false for all * other cases. */ -static __inline__ int atomic64_inc_and_test(atomic64_t *v) +static inline int atomic64_inc_and_test(atomic64_t *v) { unsigned char c; - __asm__ __volatile__( - LOCK_PREFIX "incq %0; sete %1" - :"=m" (v->counter), "=qm" (c) - :"m" (v->counter) : "memory"); + asm volatile(LOCK_PREFIX "incq %0; sete %1" + : "=m" (v->counter), "=qm" (c) + : "m" (v->counter) : "memory"); return c != 0; } @@ -353,14 +341,13 @@ static __inline__ int atomic64_inc_and_test(atomic64_t *v) * if the result is negative, or false when * result is greater than or equal to zero. */ -static __inline__ int atomic64_add_negative(long i, atomic64_t *v) +static inline int atomic64_add_negative(long i, atomic64_t *v) { unsigned char c; - __asm__ __volatile__( - LOCK_PREFIX "addq %2,%0; sets %1" - :"=m" (v->counter), "=qm" (c) - :"ir" (i), "m" (v->counter) : "memory"); + asm volatile(LOCK_PREFIX "addq %2,%0; sets %1" + : "=m" (v->counter), "=qm" (c) + : "ir" (i), "m" (v->counter) : "memory"); return c; } @@ -371,29 +358,28 @@ static __inline__ int atomic64_add_negative(long i, atomic64_t *v) * * Atomically adds @i to @v and returns @i + @v */ -static __inline__ long atomic64_add_return(long i, atomic64_t *v) +static inline long atomic64_add_return(long i, atomic64_t *v) { long __i = i; - __asm__ __volatile__( - LOCK_PREFIX "xaddq %0, %1;" - :"+r" (i), "+m" (v->counter) - : : "memory"); + asm volatile(LOCK_PREFIX "xaddq %0, %1;" + : "+r" (i), "+m" (v->counter) + : : "memory"); return i + __i; } -static __inline__ long atomic64_sub_return(long i, atomic64_t *v) +static inline long atomic64_sub_return(long i, atomic64_t *v) { - return atomic64_add_return(-i,v); + return atomic64_add_return(-i, v); } -#define atomic64_inc_return(v) (atomic64_add_return(1,v)) -#define atomic64_dec_return(v) (atomic64_sub_return(1,v)) +#define atomic64_inc_return(v) (atomic64_add_return(1, (v))) +#define atomic64_dec_return(v) (atomic64_sub_return(1, (v))) -#define atomic64_cmpxchg(v, old, new) (cmpxchg(&((v)->counter), old, new)) +#define atomic64_cmpxchg(v, old, new) (cmpxchg(&((v)->counter), (old), (new))) #define atomic64_xchg(v, new) (xchg(&((v)->counter), new)) -#define atomic_cmpxchg(v, old, new) (cmpxchg(&((v)->counter), old, new)) -#define atomic_xchg(v, new) (xchg(&((v)->counter), new)) +#define atomic_cmpxchg(v, old, new) (cmpxchg(&((v)->counter), (old), (new))) +#define atomic_xchg(v, new) (xchg(&((v)->counter), (new))) /** * atomic_add_unless - add unless the number is a given value @@ -404,7 +390,7 @@ static __inline__ long atomic64_sub_return(long i, atomic64_t *v) * Atomically adds @a to @v, so long as it was not @u. * Returns non-zero if @v was not @u, and zero otherwise. */ -static __inline__ int atomic_add_unless(atomic_t *v, int a, int u) +static inline int atomic_add_unless(atomic_t *v, int a, int u) { int c, old; c = atomic_read(v); @@ -430,7 +416,7 @@ static __inline__ int atomic_add_unless(atomic_t *v, int a, int u) * Atomically adds @a to @v, so long as it was not @u. * Returns non-zero if @v was not @u, and zero otherwise. */ -static __inline__ int atomic64_add_unless(atomic64_t *v, long a, long u) +static inline int atomic64_add_unless(atomic64_t *v, long a, long u) { long c, old; c = atomic64_read(v); @@ -448,13 +434,14 @@ static __inline__ int atomic64_add_unless(atomic64_t *v, long a, long u) #define atomic64_inc_not_zero(v) atomic64_add_unless((v), 1, 0) /* These are x86-specific, used by some header files */ -#define atomic_clear_mask(mask, addr) \ -__asm__ __volatile__(LOCK_PREFIX "andl %0,%1" \ -: : "r" (~(mask)),"m" (*addr) : "memory") - -#define atomic_set_mask(mask, addr) \ -__asm__ __volatile__(LOCK_PREFIX "orl %0,%1" \ -: : "r" ((unsigned)mask),"m" (*(addr)) : "memory") +#define atomic_clear_mask(mask, addr) \ + asm volatile(LOCK_PREFIX "andl %0,%1" \ + : : "r" (~(mask)), "m" (*(addr)) : "memory") + +#define atomic_set_mask(mask, addr) \ + asm volatile(LOCK_PREFIX "orl %0,%1" \ + : : "r" ((unsigned)(mask)), "m" (*(addr)) \ + : "memory") /* Atomic operations are already serializing on x86 */ #define smp_mb__before_atomic_dec() barrier() diff --git a/include/asm-x86/mach-default/bios_ebda.h b/include/asm-x86/bios_ebda.h index 9cbd9a668af8..9cbd9a668af8 100644 --- a/include/asm-x86/mach-default/bios_ebda.h +++ b/include/asm-x86/bios_ebda.h diff --git a/include/asm-x86/bitops.h b/include/asm-x86/bitops.h index 1a23ce1a5697..1ae7b270a1ef 100644 --- a/include/asm-x86/bitops.h +++ b/include/asm-x86/bitops.h @@ -23,10 +23,13 @@ #if __GNUC__ < 4 || (__GNUC__ == 4 && __GNUC_MINOR__ < 1) /* Technically wrong, but this avoids compilation errors on some gcc versions. */ -#define ADDR "=m" (*(volatile long *) addr) +#define ADDR "=m" (*(volatile long *)addr) +#define BIT_ADDR "=m" (((volatile int *)addr)[nr >> 5]) #else #define ADDR "+m" (*(volatile long *) addr) +#define BIT_ADDR "+m" (((volatile int *)addr)[nr >> 5]) #endif +#define BASE_ADDR "m" (*(volatile int *)addr) /** * set_bit - Atomically set a bit in memory @@ -45,9 +48,7 @@ */ static inline void set_bit(int nr, volatile void *addr) { - asm volatile(LOCK_PREFIX "bts %1,%0" - : ADDR - : "Ir" (nr) : "memory"); + asm volatile(LOCK_PREFIX "bts %1,%0" : ADDR : "Ir" (nr) : "memory"); } /** @@ -79,9 +80,7 @@ static inline void __set_bit(int nr, volatile void *addr) */ static inline void clear_bit(int nr, volatile void *addr) { - asm volatile(LOCK_PREFIX "btr %1,%0" - : ADDR - : "Ir" (nr)); + asm volatile(LOCK_PREFIX "btr %1,%2" : BIT_ADDR : "Ir" (nr), BASE_ADDR); } /* @@ -100,7 +99,7 @@ static inline void clear_bit_unlock(unsigned nr, volatile void *addr) static inline void __clear_bit(int nr, volatile void *addr) { - asm volatile("btr %1,%0" : ADDR : "Ir" (nr)); + asm volatile("btr %1,%2" : BIT_ADDR : "Ir" (nr), BASE_ADDR); } /* @@ -135,7 +134,7 @@ static inline void __clear_bit_unlock(unsigned nr, volatile void *addr) */ static inline void __change_bit(int nr, volatile void *addr) { - asm volatile("btc %1,%0" : ADDR : "Ir" (nr)); + asm volatile("btc %1,%2" : BIT_ADDR : "Ir" (nr), BASE_ADDR); } /** @@ -149,8 +148,7 @@ static inline void __change_bit(int nr, volatile void *addr) */ static inline void change_bit(int nr, volatile void *addr) { - asm volatile(LOCK_PREFIX "btc %1,%0" - : ADDR : "Ir" (nr)); + asm volatile(LOCK_PREFIX "btc %1,%2" : BIT_ADDR : "Ir" (nr), BASE_ADDR); } /** @@ -166,9 +164,7 @@ static inline int test_and_set_bit(int nr, volatile void *addr) int oldbit; asm volatile(LOCK_PREFIX "bts %2,%1\n\t" - "sbb %0,%0" - : "=r" (oldbit), ADDR - : "Ir" (nr) : "memory"); + "sbb %0,%0" : "=r" (oldbit), ADDR : "Ir" (nr) : "memory"); return oldbit; } @@ -198,10 +194,9 @@ static inline int __test_and_set_bit(int nr, volatile void *addr) { int oldbit; - asm("bts %2,%1\n\t" - "sbb %0,%0" - : "=r" (oldbit), ADDR - : "Ir" (nr)); + asm volatile("bts %2,%3\n\t" + "sbb %0,%0" + : "=r" (oldbit), BIT_ADDR : "Ir" (nr), BASE_ADDR); return oldbit; } @@ -219,8 +214,7 @@ static inline int test_and_clear_bit(int nr, volatile void *addr) asm volatile(LOCK_PREFIX "btr %2,%1\n\t" "sbb %0,%0" - : "=r" (oldbit), ADDR - : "Ir" (nr) : "memory"); + : "=r" (oldbit), ADDR : "Ir" (nr) : "memory"); return oldbit; } @@ -238,10 +232,9 @@ static inline int __test_and_clear_bit(int nr, volatile void *addr) { int oldbit; - asm volatile("btr %2,%1\n\t" + asm volatile("btr %2,%3\n\t" "sbb %0,%0" - : "=r" (oldbit), ADDR - : "Ir" (nr)); + : "=r" (oldbit), BIT_ADDR : "Ir" (nr), BASE_ADDR); return oldbit; } @@ -250,10 +243,9 @@ static inline int __test_and_change_bit(int nr, volatile void *addr) { int oldbit; - asm volatile("btc %2,%1\n\t" + asm volatile("btc %2,%3\n\t" "sbb %0,%0" - : "=r" (oldbit), ADDR - : "Ir" (nr) : "memory"); + : "=r" (oldbit), BIT_ADDR : "Ir" (nr), BASE_ADDR); return oldbit; } @@ -272,8 +264,7 @@ static inline int test_and_change_bit(int nr, volatile void *addr) asm volatile(LOCK_PREFIX "btc %2,%1\n\t" "sbb %0,%0" - : "=r" (oldbit), ADDR - : "Ir" (nr) : "memory"); + : "=r" (oldbit), ADDR : "Ir" (nr) : "memory"); return oldbit; } @@ -288,10 +279,11 @@ static inline int variable_test_bit(int nr, volatile const void *addr) { int oldbit; - asm volatile("bt %2,%1\n\t" + asm volatile("bt %2,%3\n\t" "sbb %0,%0" : "=r" (oldbit) - : "m" (*(unsigned long *)addr), "Ir" (nr)); + : "m" (((volatile const int *)addr)[nr >> 5]), + "Ir" (nr), BASE_ADDR); return oldbit; } @@ -310,6 +302,8 @@ static int test_bit(int nr, const volatile unsigned long *addr); constant_test_bit((nr),(addr)) : \ variable_test_bit((nr),(addr))) +#undef BASE_ADDR +#undef BIT_ADDR #undef ADDR #ifdef CONFIG_X86_32 diff --git a/include/asm-x86/bitops_32.h b/include/asm-x86/bitops_32.h index e4d75fcf9c03..2513a81f82aa 100644 --- a/include/asm-x86/bitops_32.h +++ b/include/asm-x86/bitops_32.h @@ -20,20 +20,22 @@ static inline int find_first_zero_bit(const unsigned long *addr, unsigned size) if (!size) return 0; - /* This looks at memory. Mark it volatile to tell gcc not to move it around */ - __asm__ __volatile__( - "movl $-1,%%eax\n\t" - "xorl %%edx,%%edx\n\t" - "repe; scasl\n\t" - "je 1f\n\t" - "xorl -4(%%edi),%%eax\n\t" - "subl $4,%%edi\n\t" - "bsfl %%eax,%%edx\n" - "1:\tsubl %%ebx,%%edi\n\t" - "shll $3,%%edi\n\t" - "addl %%edi,%%edx" - :"=d" (res), "=&c" (d0), "=&D" (d1), "=&a" (d2) - :"1" ((size + 31) >> 5), "2" (addr), "b" (addr) : "memory"); + /* This looks at memory. + * Mark it volatile to tell gcc not to move it around + */ + asm volatile("movl $-1,%%eax\n\t" + "xorl %%edx,%%edx\n\t" + "repe; scasl\n\t" + "je 1f\n\t" + "xorl -4(%%edi),%%eax\n\t" + "subl $4,%%edi\n\t" + "bsfl %%eax,%%edx\n" + "1:\tsubl %%ebx,%%edi\n\t" + "shll $3,%%edi\n\t" + "addl %%edi,%%edx" + : "=d" (res), "=&c" (d0), "=&D" (d1), "=&a" (d2) + : "1" ((size + 31) >> 5), "2" (addr), + "b" (addr) : "memory"); return res; } @@ -75,7 +77,7 @@ static inline unsigned find_first_bit(const unsigned long *addr, unsigned size) unsigned long val = *addr++; if (val) return __ffs(val) + x; - x += (sizeof(*addr)<<3); + x += sizeof(*addr) << 3; } return x; } @@ -152,10 +154,10 @@ static inline int fls(int x) #include <asm-generic/bitops/ext2-non-atomic.h> -#define ext2_set_bit_atomic(lock, nr, addr) \ - test_and_set_bit((nr), (unsigned long *)addr) -#define ext2_clear_bit_atomic(lock, nr, addr) \ - test_and_clear_bit((nr), (unsigned long *)addr) +#define ext2_set_bit_atomic(lock, nr, addr) \ + test_and_set_bit((nr), (unsigned long *)(addr)) +#define ext2_clear_bit_atomic(lock, nr, addr) \ + test_and_clear_bit((nr), (unsigned long *)(addr)) #include <asm-generic/bitops/minix.h> diff --git a/include/asm-x86/bitops_64.h b/include/asm-x86/bitops_64.h index aaf15194d536..365f8207ea59 100644 --- a/include/asm-x86/bitops_64.h +++ b/include/asm-x86/bitops_64.h @@ -17,35 +17,35 @@ static inline long __scanbit(unsigned long val, unsigned long max) return val; } -#define find_first_bit(addr,size) \ -((__builtin_constant_p(size) && (size) <= BITS_PER_LONG ? \ - (__scanbit(*(unsigned long *)addr,(size))) : \ - find_first_bit(addr,size))) - #define find_next_bit(addr,size,off) \ ((__builtin_constant_p(size) && (size) <= BITS_PER_LONG ? \ ((off) + (__scanbit((*(unsigned long *)addr) >> (off),(size)-(off)))) : \ find_next_bit(addr,size,off))) -#define find_first_zero_bit(addr,size) \ -((__builtin_constant_p(size) && (size) <= BITS_PER_LONG ? \ - (__scanbit(~*(unsigned long *)addr,(size))) : \ - find_first_zero_bit(addr,size))) - #define find_next_zero_bit(addr,size,off) \ ((__builtin_constant_p(size) && (size) <= BITS_PER_LONG ? \ ((off)+(__scanbit(~(((*(unsigned long *)addr)) >> (off)),(size)-(off)))) : \ find_next_zero_bit(addr,size,off))) -static inline void set_bit_string(unsigned long *bitmap, unsigned long i, - int len) -{ - unsigned long end = i + len; +#define find_first_bit(addr, size) \ + ((__builtin_constant_p((size)) && (size) <= BITS_PER_LONG \ + ? (__scanbit(*(unsigned long *)(addr), (size))) \ + : find_first_bit((addr), (size)))) + +#define find_first_zero_bit(addr, size) \ + ((__builtin_constant_p((size)) && (size) <= BITS_PER_LONG \ + ? (__scanbit(~*(unsigned long *)(addr), (size))) \ + : find_first_zero_bit((addr), (size)))) + +static inline void set_bit_string(unsigned long *bitmap, unsigned long i, + int len) +{ + unsigned long end = i + len; while (i < end) { - __set_bit(i, bitmap); + __set_bit(i, bitmap); i++; } -} +} /** * ffz - find first zero in word. @@ -150,10 +150,10 @@ static inline int fls(int x) #include <asm-generic/bitops/ext2-non-atomic.h> -#define ext2_set_bit_atomic(lock,nr,addr) \ - test_and_set_bit((nr),(unsigned long*)addr) -#define ext2_clear_bit_atomic(lock,nr,addr) \ - test_and_clear_bit((nr),(unsigned long*)addr) +#define ext2_set_bit_atomic(lock, nr, addr) \ + test_and_set_bit((nr), (unsigned long *)(addr)) +#define ext2_clear_bit_atomic(lock, nr, addr) \ + test_and_clear_bit((nr), (unsigned long *)(addr)) #include <asm-generic/bitops/minix.h> diff --git a/include/asm-x86/bug.h b/include/asm-x86/bug.h index 8d477a201392..b69aa64b82a4 100644 --- a/include/asm-x86/bug.h +++ b/include/asm-x86/bug.h @@ -12,25 +12,25 @@ # define __BUG_C0 "2:\t.quad 1b, %c0\n" #endif -#define BUG() \ - do { \ - asm volatile("1:\tud2\n" \ - ".pushsection __bug_table,\"a\"\n" \ - __BUG_C0 \ - "\t.word %c1, 0\n" \ - "\t.org 2b+%c2\n" \ - ".popsection" \ - : : "i" (__FILE__), "i" (__LINE__), \ - "i" (sizeof(struct bug_entry))); \ - for(;;) ; \ - } while(0) +#define BUG() \ +do { \ + asm volatile("1:\tud2\n" \ + ".pushsection __bug_table,\"a\"\n" \ + __BUG_C0 \ + "\t.word %c1, 0\n" \ + "\t.org 2b+%c2\n" \ + ".popsection" \ + : : "i" (__FILE__), "i" (__LINE__), \ + "i" (sizeof(struct bug_entry))); \ + for (;;) ; \ +} while (0) #else -#define BUG() \ - do { \ - asm volatile("ud2"); \ - for(;;) ; \ - } while(0) +#define BUG() \ +do { \ + asm volatile("ud2"); \ + for (;;) ; \ +} while (0) #endif #endif /* !CONFIG_BUG */ diff --git a/include/asm-x86/byteorder.h b/include/asm-x86/byteorder.h index fe2f2e5d51ba..e02ae2d89acf 100644 --- a/include/asm-x86/byteorder.h +++ b/include/asm-x86/byteorder.h @@ -8,50 +8,59 @@ #ifdef __i386__ -static __inline__ __attribute_const__ __u32 ___arch__swab32(__u32 x) +static inline __attribute_const__ __u32 ___arch__swab32(__u32 x) { #ifdef CONFIG_X86_BSWAP - __asm__("bswap %0" : "=r" (x) : "0" (x)); + asm("bswap %0" : "=r" (x) : "0" (x)); #else - __asm__("xchgb %b0,%h0\n\t" /* swap lower bytes */ - "rorl $16,%0\n\t" /* swap words */ - "xchgb %b0,%h0" /* swap higher bytes */ - :"=q" (x) - : "0" (x)); + asm("xchgb %b0,%h0\n\t" /* swap lower bytes */ + "rorl $16,%0\n\t" /* swap words */ + "xchgb %b0,%h0" /* swap higher bytes */ + : "=q" (x) + : "0" (x)); #endif return x; } -static __inline__ __attribute_const__ __u64 ___arch__swab64(__u64 val) +static inline __attribute_const__ __u64 ___arch__swab64(__u64 val) { union { - struct { __u32 a,b; } s; + struct { + __u32 a; + __u32 b; + } s; __u64 u; } v; v.u = val; #ifdef CONFIG_X86_BSWAP - __asm__("bswapl %0 ; bswapl %1 ; xchgl %0,%1" + asm("bswapl %0 ; bswapl %1 ; xchgl %0,%1" : "=r" (v.s.a), "=r" (v.s.b) : "0" (v.s.a), "1" (v.s.b)); #else v.s.a = ___arch__swab32(v.s.a); v.s.b = ___arch__swab32(v.s.b); - __asm__("xchgl %0,%1" : "=r" (v.s.a), "=r" (v.s.b) : "0" (v.s.a), "1" (v.s.b)); + asm("xchgl %0,%1" + : "=r" (v.s.a), "=r" (v.s.b) + : "0" (v.s.a), "1" (v.s.b)); #endif return v.u; } #else /* __i386__ */ -static __inline__ __attribute_const__ __u64 ___arch__swab64(__u64 x) +static inline __attribute_const__ __u64 ___arch__swab64(__u64 x) { - __asm__("bswapq %0" : "=r" (x) : "0" (x)); + asm("bswapq %0" + : "=r" (x) + : "0" (x)); return x; } -static __inline__ __attribute_const__ __u32 ___arch__swab32(__u32 x) +static inline __attribute_const__ __u32 ___arch__swab32(__u32 x) { - __asm__("bswapl %0" : "=r" (x) : "0" (x)); + asm("bswapl %0" + : "=r" (x) + : "0" (x)); return x; } diff --git a/include/asm-x86/cacheflush.h b/include/asm-x86/cacheflush.h index 5396c212d8c0..f4c0ab50d2c2 100644 --- a/include/asm-x86/cacheflush.h +++ b/include/asm-x86/cacheflush.h @@ -14,33 +14,85 @@ #define flush_dcache_mmap_lock(mapping) do { } while (0) #define flush_dcache_mmap_unlock(mapping) do { } while (0) #define flush_icache_range(start, end) do { } while (0) -#define flush_icache_page(vma,pg) do { } while (0) -#define flush_icache_user_range(vma,pg,adr,len) do { } while (0) +#define flush_icache_page(vma, pg) do { } while (0) +#define flush_icache_user_range(vma, pg, adr, len) do { } while (0) #define flush_cache_vmap(start, end) do { } while (0) #define flush_cache_vunmap(start, end) do { } while (0) -#define copy_to_user_page(vma, page, vaddr, dst, src, len) \ - memcpy(dst, src, len) -#define copy_from_user_page(vma, page, vaddr, dst, src, len) \ - memcpy(dst, src, len) +#define copy_to_user_page(vma, page, vaddr, dst, src, len) \ + memcpy((dst), (src), (len)) +#define copy_from_user_page(vma, page, vaddr, dst, src, len) \ + memcpy((dst), (src), (len)) -int __deprecated_for_modules change_page_attr(struct page *page, int numpages, - pgprot_t prot); -int set_pages_uc(struct page *page, int numpages); -int set_pages_wb(struct page *page, int numpages); -int set_pages_x(struct page *page, int numpages); -int set_pages_nx(struct page *page, int numpages); -int set_pages_ro(struct page *page, int numpages); -int set_pages_rw(struct page *page, int numpages); +/* + * The set_memory_* API can be used to change various attributes of a virtual + * address range. The attributes include: + * Cachability : UnCached, WriteCombining, WriteBack + * Executability : eXeutable, NoteXecutable + * Read/Write : ReadOnly, ReadWrite + * Presence : NotPresent + * + * Within a catagory, the attributes are mutually exclusive. + * + * The implementation of this API will take care of various aspects that + * are associated with changing such attributes, such as: + * - Flushing TLBs + * - Flushing CPU caches + * - Making sure aliases of the memory behind the mapping don't violate + * coherency rules as defined by the CPU in the system. + * + * What this API does not do: + * - Provide exclusion between various callers - including callers that + * operation on other mappings of the same physical page + * - Restore default attributes when a page is freed + * - Guarantee that mappings other than the requested one are + * in any state, other than that these do not violate rules for + * the CPU you have. Do not depend on any effects on other mappings, + * CPUs other than the one you have may have more relaxed rules. + * The caller is required to take care of these. + */ +int _set_memory_uc(unsigned long addr, int numpages); +int _set_memory_wc(unsigned long addr, int numpages); +int _set_memory_wb(unsigned long addr, int numpages); int set_memory_uc(unsigned long addr, int numpages); +int set_memory_wc(unsigned long addr, int numpages); int set_memory_wb(unsigned long addr, int numpages); int set_memory_x(unsigned long addr, int numpages); int set_memory_nx(unsigned long addr, int numpages); int set_memory_ro(unsigned long addr, int numpages); int set_memory_rw(unsigned long addr, int numpages); int set_memory_np(unsigned long addr, int numpages); +int set_memory_4k(unsigned long addr, int numpages); + +/* + * For legacy compatibility with the old APIs, a few functions + * are provided that work on a "struct page". + * These functions operate ONLY on the 1:1 kernel mapping of the + * memory that the struct page represents, and internally just + * call the set_memory_* function. See the description of the + * set_memory_* function for more details on conventions. + * + * These APIs should be considered *deprecated* and are likely going to + * be removed in the future. + * The reason for this is the implicit operation on the 1:1 mapping only, + * making this not a generally useful API. + * + * Specifically, many users of the old APIs had a virtual address, + * called virt_to_page() or vmalloc_to_page() on that address to + * get a struct page* that the old API required. + * To convert these cases, use set_memory_*() on the original + * virtual address, do not use these functions. + */ + +int set_pages_uc(struct page *page, int numpages); +int set_pages_wb(struct page *page, int numpages); +int set_pages_x(struct page *page, int numpages); +int set_pages_nx(struct page *page, int numpages); +int set_pages_ro(struct page *page, int numpages); +int set_pages_rw(struct page *page, int numpages); + void clflush_cache_range(void *addr, unsigned int size); diff --git a/include/asm-x86/checksum_32.h b/include/asm-x86/checksum_32.h index 75194abbe8ee..52bbb0d8c4c1 100644 --- a/include/asm-x86/checksum_32.h +++ b/include/asm-x86/checksum_32.h @@ -28,7 +28,8 @@ asmlinkage __wsum csum_partial(const void *buff, int len, __wsum sum); */ asmlinkage __wsum csum_partial_copy_generic(const void *src, void *dst, - int len, __wsum sum, int *src_err_ptr, int *dst_err_ptr); + int len, __wsum sum, + int *src_err_ptr, int *dst_err_ptr); /* * Note: when you get a NULL pointer exception here this means someone @@ -37,20 +38,20 @@ asmlinkage __wsum csum_partial_copy_generic(const void *src, void *dst, * If you use these functions directly please don't forget the * access_ok(). */ -static __inline__ -__wsum csum_partial_copy_nocheck (const void *src, void *dst, - int len, __wsum sum) +static inline __wsum csum_partial_copy_nocheck(const void *src, void *dst, + int len, __wsum sum) { - return csum_partial_copy_generic ( src, dst, len, sum, NULL, NULL); + return csum_partial_copy_generic(src, dst, len, sum, NULL, NULL); } -static __inline__ -__wsum csum_partial_copy_from_user(const void __user *src, void *dst, - int len, __wsum sum, int *err_ptr) +static inline __wsum csum_partial_copy_from_user(const void __user *src, + void *dst, + int len, __wsum sum, + int *err_ptr) { might_sleep(); return csum_partial_copy_generic((__force void *)src, dst, - len, sum, err_ptr, NULL); + len, sum, err_ptr, NULL); } /* @@ -64,30 +65,29 @@ static inline __sum16 ip_fast_csum(const void *iph, unsigned int ihl) { unsigned int sum; - __asm__ __volatile__( - "movl (%1), %0 ;\n" - "subl $4, %2 ;\n" - "jbe 2f ;\n" - "addl 4(%1), %0 ;\n" - "adcl 8(%1), %0 ;\n" - "adcl 12(%1), %0 ;\n" -"1: adcl 16(%1), %0 ;\n" - "lea 4(%1), %1 ;\n" - "decl %2 ;\n" - "jne 1b ;\n" - "adcl $0, %0 ;\n" - "movl %0, %2 ;\n" - "shrl $16, %0 ;\n" - "addw %w2, %w0 ;\n" - "adcl $0, %0 ;\n" - "notl %0 ;\n" -"2: ;\n" + asm volatile("movl (%1), %0 ;\n" + "subl $4, %2 ;\n" + "jbe 2f ;\n" + "addl 4(%1), %0 ;\n" + "adcl 8(%1), %0 ;\n" + "adcl 12(%1), %0;\n" + "1: adcl 16(%1), %0 ;\n" + "lea 4(%1), %1 ;\n" + "decl %2 ;\n" + "jne 1b ;\n" + "adcl $0, %0 ;\n" + "movl %0, %2 ;\n" + "shrl $16, %0 ;\n" + "addw %w2, %w0 ;\n" + "adcl $0, %0 ;\n" + "notl %0 ;\n" + "2: ;\n" /* Since the input registers which are loaded with iph and ihl are modified, we must also specify them as outputs, or gcc will assume they contain their original values. */ - : "=r" (sum), "=r" (iph), "=r" (ihl) - : "1" (iph), "2" (ihl) - : "memory"); + : "=r" (sum), "=r" (iph), "=r" (ihl) + : "1" (iph), "2" (ihl) + : "memory"); return (__force __sum16)sum; } @@ -97,29 +97,27 @@ static inline __sum16 ip_fast_csum(const void *iph, unsigned int ihl) static inline __sum16 csum_fold(__wsum sum) { - __asm__( - "addl %1, %0 ;\n" - "adcl $0xffff, %0 ;\n" - : "=r" (sum) - : "r" ((__force u32)sum << 16), - "0" ((__force u32)sum & 0xffff0000) - ); + asm("addl %1, %0 ;\n" + "adcl $0xffff, %0 ;\n" + : "=r" (sum) + : "r" ((__force u32)sum << 16), + "0" ((__force u32)sum & 0xffff0000)); return (__force __sum16)(~(__force u32)sum >> 16); } static inline __wsum csum_tcpudp_nofold(__be32 saddr, __be32 daddr, - unsigned short len, - unsigned short proto, - __wsum sum) + unsigned short len, + unsigned short proto, + __wsum sum) { - __asm__( - "addl %1, %0 ;\n" - "adcl %2, %0 ;\n" - "adcl %3, %0 ;\n" - "adcl $0, %0 ;\n" - : "=r" (sum) - : "g" (daddr), "g"(saddr), "g"((len + proto) << 8), "0"(sum)); - return sum; + asm("addl %1, %0 ;\n" + "adcl %2, %0 ;\n" + "adcl %3, %0 ;\n" + "adcl $0, %0 ;\n" + : "=r" (sum) + : "g" (daddr), "g"(saddr), + "g" ((len + proto) << 8), "0" (sum)); + return sum; } /* @@ -127,11 +125,11 @@ static inline __wsum csum_tcpudp_nofold(__be32 saddr, __be32 daddr, * returns a 16-bit checksum, already complemented */ static inline __sum16 csum_tcpudp_magic(__be32 saddr, __be32 daddr, - unsigned short len, - unsigned short proto, - __wsum sum) + unsigned short len, + unsigned short proto, + __wsum sum) { - return csum_fold(csum_tcpudp_nofold(saddr,daddr,len,proto,sum)); + return csum_fold(csum_tcpudp_nofold(saddr, daddr, len, proto, sum)); } /* @@ -141,30 +139,29 @@ static inline __sum16 csum_tcpudp_magic(__be32 saddr, __be32 daddr, static inline __sum16 ip_compute_csum(const void *buff, int len) { - return csum_fold (csum_partial(buff, len, 0)); + return csum_fold(csum_partial(buff, len, 0)); } #define _HAVE_ARCH_IPV6_CSUM -static __inline__ __sum16 csum_ipv6_magic(const struct in6_addr *saddr, - const struct in6_addr *daddr, - __u32 len, unsigned short proto, - __wsum sum) +static inline __sum16 csum_ipv6_magic(const struct in6_addr *saddr, + const struct in6_addr *daddr, + __u32 len, unsigned short proto, + __wsum sum) { - __asm__( - "addl 0(%1), %0 ;\n" - "adcl 4(%1), %0 ;\n" - "adcl 8(%1), %0 ;\n" - "adcl 12(%1), %0 ;\n" - "adcl 0(%2), %0 ;\n" - "adcl 4(%2), %0 ;\n" - "adcl 8(%2), %0 ;\n" - "adcl 12(%2), %0 ;\n" - "adcl %3, %0 ;\n" - "adcl %4, %0 ;\n" - "adcl $0, %0 ;\n" - : "=&r" (sum) - : "r" (saddr), "r" (daddr), - "r"(htonl(len)), "r"(htonl(proto)), "0"(sum)); + asm("addl 0(%1), %0 ;\n" + "adcl 4(%1), %0 ;\n" + "adcl 8(%1), %0 ;\n" + "adcl 12(%1), %0 ;\n" + "adcl 0(%2), %0 ;\n" + "adcl 4(%2), %0 ;\n" + "adcl 8(%2), %0 ;\n" + "adcl 12(%2), %0 ;\n" + "adcl %3, %0 ;\n" + "adcl %4, %0 ;\n" + "adcl $0, %0 ;\n" + : "=&r" (sum) + : "r" (saddr), "r" (daddr), + "r" (htonl(len)), "r" (htonl(proto)), "0" (sum)); return csum_fold(sum); } @@ -173,14 +170,15 @@ static __inline__ __sum16 csum_ipv6_magic(const struct in6_addr *saddr, * Copy and checksum to user */ #define HAVE_CSUM_COPY_USER -static __inline__ __wsum csum_and_copy_to_user(const void *src, - void __user *dst, - int len, __wsum sum, - int *err_ptr) +static inline __wsum csum_and_copy_to_user(const void *src, + void __user *dst, + int len, __wsum sum, + int *err_ptr) { might_sleep(); if (access_ok(VERIFY_WRITE, dst, len)) - return csum_partial_copy_generic(src, (__force void *)dst, len, sum, NULL, err_ptr); + return csum_partial_copy_generic(src, (__force void *)dst, + len, sum, NULL, err_ptr); if (len) *err_ptr = -EFAULT; diff --git a/include/asm-x86/checksum_64.h b/include/asm-x86/checksum_64.h index e5f79997decc..8bd861cc5267 100644 --- a/include/asm-x86/checksum_64.h +++ b/include/asm-x86/checksum_64.h @@ -1,33 +1,31 @@ #ifndef _X86_64_CHECKSUM_H #define _X86_64_CHECKSUM_H -/* - * Checksums for x86-64 - * Copyright 2002 by Andi Kleen, SuSE Labs +/* + * Checksums for x86-64 + * Copyright 2002 by Andi Kleen, SuSE Labs * with some code from asm-x86/checksum.h - */ + */ #include <linux/compiler.h> #include <asm/uaccess.h> #include <asm/byteorder.h> -/** +/** * csum_fold - Fold and invert a 32bit checksum. * sum: 32bit unfolded sum - * + * * Fold a 32bit running checksum to 16bit and invert it. This is usually * the last step before putting a checksum into a packet. * Make sure not to mix with 64bit checksums. */ static inline __sum16 csum_fold(__wsum sum) { - __asm__( - " addl %1,%0\n" - " adcl $0xffff,%0" - : "=r" (sum) - : "r" ((__force u32)sum << 16), - "0" ((__force u32)sum & 0xffff0000) - ); + asm(" addl %1,%0\n" + " adcl $0xffff,%0" + : "=r" (sum) + : "r" ((__force u32)sum << 16), + "0" ((__force u32)sum & 0xffff0000)); return (__force __sum16)(~(__force u32)sum >> 16); } @@ -43,46 +41,46 @@ static inline __sum16 csum_fold(__wsum sum) * ip_fast_csum - Compute the IPv4 header checksum efficiently. * iph: ipv4 header * ihl: length of header / 4 - */ + */ static inline __sum16 ip_fast_csum(const void *iph, unsigned int ihl) { unsigned int sum; - asm( " movl (%1), %0\n" - " subl $4, %2\n" - " jbe 2f\n" - " addl 4(%1), %0\n" - " adcl 8(%1), %0\n" - " adcl 12(%1), %0\n" - "1: adcl 16(%1), %0\n" - " lea 4(%1), %1\n" - " decl %2\n" - " jne 1b\n" - " adcl $0, %0\n" - " movl %0, %2\n" - " shrl $16, %0\n" - " addw %w2, %w0\n" - " adcl $0, %0\n" - " notl %0\n" - "2:" + asm(" movl (%1), %0\n" + " subl $4, %2\n" + " jbe 2f\n" + " addl 4(%1), %0\n" + " adcl 8(%1), %0\n" + " adcl 12(%1), %0\n" + "1: adcl 16(%1), %0\n" + " lea 4(%1), %1\n" + " decl %2\n" + " jne 1b\n" + " adcl $0, %0\n" + " movl %0, %2\n" + " shrl $16, %0\n" + " addw %w2, %w0\n" + " adcl $0, %0\n" + " notl %0\n" + "2:" /* Since the input registers which are loaded with iph and ihl are modified, we must also specify them as outputs, or gcc will assume they contain their original values. */ - : "=r" (sum), "=r" (iph), "=r" (ihl) - : "1" (iph), "2" (ihl) - : "memory"); + : "=r" (sum), "=r" (iph), "=r" (ihl) + : "1" (iph), "2" (ihl) + : "memory"); return (__force __sum16)sum; } -/** +/** * csum_tcpup_nofold - Compute an IPv4 pseudo header checksum. * @saddr: source address * @daddr: destination address * @len: length of packet * @proto: ip protocol of packet - * @sum: initial sum to be added in (32bit unfolded) - * - * Returns the pseudo header checksum the input data. Result is + * @sum: initial sum to be added in (32bit unfolded) + * + * Returns the pseudo header checksum the input data. Result is * 32bit unfolded. */ static inline __wsum @@ -93,32 +91,32 @@ csum_tcpudp_nofold(__be32 saddr, __be32 daddr, unsigned short len, " adcl %2, %0\n" " adcl %3, %0\n" " adcl $0, %0\n" - : "=r" (sum) + : "=r" (sum) : "g" (daddr), "g" (saddr), "g" ((len + proto)<<8), "0" (sum)); - return sum; + return sum; } -/** +/** * csum_tcpup_magic - Compute an IPv4 pseudo header checksum. * @saddr: source address * @daddr: destination address * @len: length of packet * @proto: ip protocol of packet - * @sum: initial sum to be added in (32bit unfolded) - * + * @sum: initial sum to be added in (32bit unfolded) + * * Returns the 16bit pseudo header checksum the input data already * complemented and ready to be filled in. */ -static inline __sum16 -csum_tcpudp_magic(__be32 saddr, __be32 daddr, - unsigned short len, unsigned short proto, __wsum sum) +static inline __sum16 csum_tcpudp_magic(__be32 saddr, __be32 daddr, + unsigned short len, + unsigned short proto, __wsum sum) { - return csum_fold(csum_tcpudp_nofold(saddr,daddr,len,proto,sum)); + return csum_fold(csum_tcpudp_nofold(saddr, daddr, len, proto, sum)); } -/** +/** * csum_partial - Compute an internet checksum. * @buff: buffer to be checksummed * @len: length of buffer. @@ -127,7 +125,7 @@ csum_tcpudp_magic(__be32 saddr, __be32 daddr, * Returns the 32bit unfolded internet checksum of the buffer. * Before filling it in it needs to be csum_fold()'ed. * buff should be aligned to a 64bit boundary if possible. - */ + */ extern __wsum csum_partial(const void *buff, int len, __wsum sum); #define _HAVE_ARCH_COPY_AND_CSUM_FROM_USER 1 @@ -136,23 +134,22 @@ extern __wsum csum_partial(const void *buff, int len, __wsum sum); /* Do not call this directly. Use the wrappers below */ extern __wsum csum_partial_copy_generic(const void *src, const void *dst, - int len, - __wsum sum, - int *src_err_ptr, int *dst_err_ptr); + int len, __wsum sum, + int *src_err_ptr, int *dst_err_ptr); extern __wsum csum_partial_copy_from_user(const void __user *src, void *dst, - int len, __wsum isum, int *errp); + int len, __wsum isum, int *errp); extern __wsum csum_partial_copy_to_user(const void *src, void __user *dst, - int len, __wsum isum, int *errp); -extern __wsum csum_partial_copy_nocheck(const void *src, void *dst, int len, - __wsum sum); + int len, __wsum isum, int *errp); +extern __wsum csum_partial_copy_nocheck(const void *src, void *dst, + int len, __wsum sum); /* Old names. To be removed. */ #define csum_and_copy_to_user csum_partial_copy_to_user #define csum_and_copy_from_user csum_partial_copy_from_user -/** +/** * ip_compute_csum - Compute an 16bit IP checksum. * @buff: buffer address. * @len: length of buffer. @@ -170,7 +167,7 @@ extern __sum16 ip_compute_csum(const void *buff, int len); * @proto: protocol of packet * @sum: initial sum (32bit unfolded) to be added in * - * Computes an IPv6 pseudo header checksum. This sum is added the checksum + * Computes an IPv6 pseudo header checksum. This sum is added the checksum * into UDP/TCP packets and contains some link layer information. * Returns the unfolded 32bit checksum. */ @@ -185,11 +182,10 @@ csum_ipv6_magic(const struct in6_addr *saddr, const struct in6_addr *daddr, static inline unsigned add32_with_carry(unsigned a, unsigned b) { asm("addl %2,%0\n\t" - "adcl $0,%0" - : "=r" (a) + "adcl $0,%0" + : "=r" (a) : "0" (a), "r" (b)); return a; } #endif - diff --git a/include/asm-x86/cmpxchg_32.h b/include/asm-x86/cmpxchg_32.h index 959fad00dff5..bf5a69d1329e 100644 --- a/include/asm-x86/cmpxchg_32.h +++ b/include/asm-x86/cmpxchg_32.h @@ -8,9 +8,12 @@ * you need to test for the feature in boot_cpu_data. */ -#define xchg(ptr,v) ((__typeof__(*(ptr)))__xchg((unsigned long)(v),(ptr),sizeof(*(ptr)))) +#define xchg(ptr, v) \ + ((__typeof__(*(ptr)))__xchg((unsigned long)(v), (ptr), sizeof(*(ptr)))) -struct __xchg_dummy { unsigned long a[100]; }; +struct __xchg_dummy { + unsigned long a[100]; +}; #define __xg(x) ((struct __xchg_dummy *)(x)) /* @@ -27,72 +30,74 @@ struct __xchg_dummy { unsigned long a[100]; }; * of the instruction set reference 24319102.pdf. We need * the reader side to see the coherent 64bit value. */ -static inline void __set_64bit (unsigned long long * ptr, - unsigned int low, unsigned int high) +static inline void __set_64bit(unsigned long long *ptr, + unsigned int low, unsigned int high) { - __asm__ __volatile__ ( - "\n1:\t" - "movl (%0), %%eax\n\t" - "movl 4(%0), %%edx\n\t" - LOCK_PREFIX "cmpxchg8b (%0)\n\t" - "jnz 1b" - : /* no outputs */ - : "D"(ptr), - "b"(low), - "c"(high) - : "ax","dx","memory"); + asm volatile("\n1:\t" + "movl (%0), %%eax\n\t" + "movl 4(%0), %%edx\n\t" + LOCK_PREFIX "cmpxchg8b (%0)\n\t" + "jnz 1b" + : /* no outputs */ + : "D"(ptr), + "b"(low), + "c"(high) + : "ax", "dx", "memory"); } -static inline void __set_64bit_constant (unsigned long long *ptr, - unsigned long long value) +static inline void __set_64bit_constant(unsigned long long *ptr, + unsigned long long value) { - __set_64bit(ptr,(unsigned int)(value), (unsigned int)((value)>>32ULL)); + __set_64bit(ptr, (unsigned int)value, (unsigned int)(value >> 32)); } -#define ll_low(x) *(((unsigned int*)&(x))+0) -#define ll_high(x) *(((unsigned int*)&(x))+1) -static inline void __set_64bit_var (unsigned long long *ptr, - unsigned long long value) +#define ll_low(x) *(((unsigned int *)&(x)) + 0) +#define ll_high(x) *(((unsigned int *)&(x)) + 1) + +static inline void __set_64bit_var(unsigned long long *ptr, + unsigned long long value) { - __set_64bit(ptr,ll_low(value), ll_high(value)); + __set_64bit(ptr, ll_low(value), ll_high(value)); } -#define set_64bit(ptr,value) \ -(__builtin_constant_p(value) ? \ - __set_64bit_constant(ptr, value) : \ - __set_64bit_var(ptr, value) ) +#define set_64bit(ptr, value) \ + (__builtin_constant_p((value)) \ + ? __set_64bit_constant((ptr), (value)) \ + : __set_64bit_var((ptr), (value))) -#define _set_64bit(ptr,value) \ -(__builtin_constant_p(value) ? \ - __set_64bit(ptr, (unsigned int)(value), (unsigned int)((value)>>32ULL) ) : \ - __set_64bit(ptr, ll_low(value), ll_high(value)) ) +#define _set_64bit(ptr, value) \ + (__builtin_constant_p(value) \ + ? __set_64bit(ptr, (unsigned int)(value), \ + (unsigned int)((value) >> 32)) \ + : __set_64bit(ptr, ll_low((value)), ll_high((value)))) /* * Note: no "lock" prefix even on SMP: xchg always implies lock anyway * Note 2: xchg has side effect, so that attribute volatile is necessary, * but generally the primitive is invalid, *ptr is output argument. --ANK */ -static inline unsigned long __xchg(unsigned long x, volatile void * ptr, int size) +static inline unsigned long __xchg(unsigned long x, volatile void *ptr, + int size) { switch (size) { - case 1: - __asm__ __volatile__("xchgb %b0,%1" - :"=q" (x) - :"m" (*__xg(ptr)), "0" (x) - :"memory"); - break; - case 2: - __asm__ __volatile__("xchgw %w0,%1" - :"=r" (x) - :"m" (*__xg(ptr)), "0" (x) - :"memory"); - break; - case 4: - __asm__ __volatile__("xchgl %0,%1" - :"=r" (x) - :"m" (*__xg(ptr)), "0" (x) - :"memory"); - break; + case 1: + asm volatile("xchgb %b0,%1" + : "=q" (x) + : "m" (*__xg(ptr)), "0" (x) + : "memory"); + break; + case 2: + asm volatile("xchgw %w0,%1" + : "=r" (x) + : "m" (*__xg(ptr)), "0" (x) + : "memory"); + break; + case 4: + asm volatile("xchgl %0,%1" + : "=r" (x) + : "m" (*__xg(ptr)), "0" (x) + : "memory"); + break; } return x; } @@ -105,24 +110,27 @@ static inline unsigned long __xchg(unsigned long x, volatile void * ptr, int siz #ifdef CONFIG_X86_CMPXCHG #define __HAVE_ARCH_CMPXCHG 1 -#define cmpxchg(ptr, o, n) \ - ((__typeof__(*(ptr)))__cmpxchg((ptr), (unsigned long)(o), \ - (unsigned long)(n), sizeof(*(ptr)))) -#define sync_cmpxchg(ptr, o, n) \ - ((__typeof__(*(ptr)))__sync_cmpxchg((ptr), (unsigned long)(o), \ - (unsigned long)(n), sizeof(*(ptr)))) -#define cmpxchg_local(ptr, o, n) \ - ((__typeof__(*(ptr)))__cmpxchg_local((ptr), (unsigned long)(o), \ - (unsigned long)(n), sizeof(*(ptr)))) +#define cmpxchg(ptr, o, n) \ + ((__typeof__(*(ptr)))__cmpxchg((ptr), (unsigned long)(o), \ + (unsigned long)(n), \ + sizeof(*(ptr)))) +#define sync_cmpxchg(ptr, o, n) \ + ((__typeof__(*(ptr)))__sync_cmpxchg((ptr), (unsigned long)(o), \ + (unsigned long)(n), \ + sizeof(*(ptr)))) +#define cmpxchg_local(ptr, o, n) \ + ((__typeof__(*(ptr)))__cmpxchg_local((ptr), (unsigned long)(o), \ + (unsigned long)(n), \ + sizeof(*(ptr)))) #endif #ifdef CONFIG_X86_CMPXCHG64 -#define cmpxchg64(ptr, o, n) \ - ((__typeof__(*(ptr)))__cmpxchg64((ptr), (unsigned long long)(o), \ - (unsigned long long)(n))) -#define cmpxchg64_local(ptr, o, n) \ - ((__typeof__(*(ptr)))__cmpxchg64_local((ptr), (unsigned long long)(o),\ - (unsigned long long)(n))) +#define cmpxchg64(ptr, o, n) \ + ((__typeof__(*(ptr)))__cmpxchg64((ptr), (unsigned long long)(o), \ + (unsigned long long)(n))) +#define cmpxchg64_local(ptr, o, n) \ + ((__typeof__(*(ptr)))__cmpxchg64_local((ptr), (unsigned long long)(o), \ + (unsigned long long)(n))) #endif static inline unsigned long __cmpxchg(volatile void *ptr, unsigned long old, @@ -131,22 +139,22 @@ static inline unsigned long __cmpxchg(volatile void *ptr, unsigned long old, unsigned long prev; switch (size) { case 1: - __asm__ __volatile__(LOCK_PREFIX "cmpxchgb %b1,%2" - : "=a"(prev) - : "q"(new), "m"(*__xg(ptr)), "0"(old) - : "memory"); + asm volatile(LOCK_PREFIX "cmpxchgb %b1,%2" + : "=a"(prev) + : "q"(new), "m"(*__xg(ptr)), "0"(old) + : "memory"); return prev; case 2: - __asm__ __volatile__(LOCK_PREFIX "cmpxchgw %w1,%2" - : "=a"(prev) - : "r"(new), "m"(*__xg(ptr)), "0"(old) - : "memory"); + asm volatile(LOCK_PREFIX "cmpxchgw %w1,%2" + : "=a"(prev) + : "r"(new), "m"(*__xg(ptr)), "0"(old) + : "memory"); return prev; case 4: - __asm__ __volatile__(LOCK_PREFIX "cmpxchgl %1,%2" - : "=a"(prev) - : "r"(new), "m"(*__xg(ptr)), "0"(old) - : "memory"); + asm volatile(LOCK_PREFIX "cmpxchgl %1,%2" + : "=a"(prev) + : "r"(new), "m"(*__xg(ptr)), "0"(old) + : "memory"); return prev; } return old; @@ -158,85 +166,88 @@ static inline unsigned long __cmpxchg(volatile void *ptr, unsigned long old, * isn't. */ static inline unsigned long __sync_cmpxchg(volatile void *ptr, - unsigned long old, - unsigned long new, int size) + unsigned long old, + unsigned long new, int size) { unsigned long prev; switch (size) { case 1: - __asm__ __volatile__("lock; cmpxchgb %b1,%2" - : "=a"(prev) - : "q"(new), "m"(*__xg(ptr)), "0"(old) - : "memory"); + asm volatile("lock; cmpxchgb %b1,%2" + : "=a"(prev) + : "q"(new), "m"(*__xg(ptr)), "0"(old) + : "memory"); return prev; case 2: - __asm__ __volatile__("lock; cmpxchgw %w1,%2" - : "=a"(prev) - : "r"(new), "m"(*__xg(ptr)), "0"(old) - : "memory"); + asm volatile("lock; cmpxchgw %w1,%2" + : "=a"(prev) + : "r"(new), "m"(*__xg(ptr)), "0"(old) + : "memory"); return prev; case 4: - __asm__ __volatile__("lock; cmpxchgl %1,%2" - : "=a"(prev) - : "r"(new), "m"(*__xg(ptr)), "0"(old) - : "memory"); + asm volatile("lock; cmpxchgl %1,%2" + : "=a"(prev) + : "r"(new), "m"(*__xg(ptr)), "0"(old) + : "memory"); return prev; } return old; } static inline unsigned long __cmpxchg_local(volatile void *ptr, - unsigned long old, unsigned long new, int size) + unsigned long old, + unsigned long new, int size) { unsigned long prev; switch (size) { case 1: - __asm__ __volatile__("cmpxchgb %b1,%2" - : "=a"(prev) - : "q"(new), "m"(*__xg(ptr)), "0"(old) - : "memory"); + asm volatile("cmpxchgb %b1,%2" + : "=a"(prev) + : "q"(new), "m"(*__xg(ptr)), "0"(old) + : "memory"); return prev; case 2: - __asm__ __volatile__("cmpxchgw %w1,%2" - : "=a"(prev) - : "r"(new), "m"(*__xg(ptr)), "0"(old) - : "memory"); + asm volatile("cmpxchgw %w1,%2" + : "=a"(prev) + : "r"(new), "m"(*__xg(ptr)), "0"(old) + : "memory"); return prev; case 4: - __asm__ __volatile__("cmpxchgl %1,%2" - : "=a"(prev) - : "r"(new), "m"(*__xg(ptr)), "0"(old) - : "memory"); + asm volatile("cmpxchgl %1,%2" + : "=a"(prev) + : "r"(new), "m"(*__xg(ptr)), "0"(old) + : "memory"); return prev; } return old; } static inline unsigned long long __cmpxchg64(volatile void *ptr, - unsigned long long old, unsigned long long new) + unsigned long long old, + unsigned long long new) { unsigned long long prev; - __asm__ __volatile__(LOCK_PREFIX "cmpxchg8b %3" - : "=A"(prev) - : "b"((unsigned long)new), - "c"((unsigned long)(new >> 32)), - "m"(*__xg(ptr)), - "0"(old) - : "memory"); + asm volatile(LOCK_PREFIX "cmpxchg8b %3" + : "=A"(prev) + : "b"((unsigned long)new), + "c"((unsigned long)(new >> 32)), + "m"(*__xg(ptr)), + "0"(old) + : "memory"); return prev; } static inline unsigned long long __cmpxchg64_local(volatile void *ptr, - unsigned long long old, unsigned long long new) + unsigned long long old, + unsigned long long new) { unsigned long long prev; - __asm__ __volatile__("cmpxchg8b %3" - : "=A"(prev) - : "b"((unsigned long)new), - "c"((unsigned long)(new >> 32)), - "m"(*__xg(ptr)), - "0"(old) - : "memory"); + asm volatile("cmpxchg8b %3" + : "=A"(prev) + : "b"((unsigned long)new), + "c"((unsigned long)(new >> 32)), + "m"(*__xg(ptr)), + "0"(old) + : "memory"); return prev; } @@ -252,7 +263,7 @@ extern unsigned long cmpxchg_386_u16(volatile void *, u16, u16); extern unsigned long cmpxchg_386_u32(volatile void *, u32, u32); static inline unsigned long cmpxchg_386(volatile void *ptr, unsigned long old, - unsigned long new, int size) + unsigned long new, int size) { switch (size) { case 1: diff --git a/include/asm-x86/cmpxchg_64.h b/include/asm-x86/cmpxchg_64.h index 56f5b41e071c..d9b26b9a28cf 100644 --- a/include/asm-x86/cmpxchg_64.h +++ b/include/asm-x86/cmpxchg_64.h @@ -3,7 +3,8 @@ #include <asm/alternative.h> /* Provides LOCK_PREFIX */ -#define xchg(ptr,v) ((__typeof__(*(ptr)))__xchg((unsigned long)(v),(ptr),sizeof(*(ptr)))) +#define xchg(ptr, v) ((__typeof__(*(ptr)))__xchg((unsigned long)(v), \ + (ptr), sizeof(*(ptr)))) #define __xg(x) ((volatile long *)(x)) @@ -19,33 +20,34 @@ static inline void set_64bit(volatile unsigned long *ptr, unsigned long val) * Note 2: xchg has side effect, so that attribute volatile is necessary, * but generally the primitive is invalid, *ptr is output argument. --ANK */ -static inline unsigned long __xchg(unsigned long x, volatile void * ptr, int size) +static inline unsigned long __xchg(unsigned long x, volatile void *ptr, + int size) { switch (size) { - case 1: - __asm__ __volatile__("xchgb %b0,%1" - :"=q" (x) - :"m" (*__xg(ptr)), "0" (x) - :"memory"); - break; - case 2: - __asm__ __volatile__("xchgw %w0,%1" - :"=r" (x) - :"m" (*__xg(ptr)), "0" (x) - :"memory"); - break; - case 4: - __asm__ __volatile__("xchgl %k0,%1" - :"=r" (x) - :"m" (*__xg(ptr)), "0" (x) - :"memory"); - break; - case 8: - __asm__ __volatile__("xchgq %0,%1" - :"=r" (x) - :"m" (*__xg(ptr)), "0" (x) - :"memory"); - break; + case 1: + asm volatile("xchgb %b0,%1" + : "=q" (x) + : "m" (*__xg(ptr)), "0" (x) + : "memory"); + break; + case 2: + asm volatile("xchgw %w0,%1" + : "=r" (x) + : "m" (*__xg(ptr)), "0" (x) + : "memory"); + break; + case 4: + asm volatile("xchgl %k0,%1" + : "=r" (x) + : "m" (*__xg(ptr)), "0" (x) + : "memory"); + break; + case 8: + asm volatile("xchgq %0,%1" + : "=r" (x) + : "m" (*__xg(ptr)), "0" (x) + : "memory"); + break; } return x; } @@ -64,61 +66,62 @@ static inline unsigned long __cmpxchg(volatile void *ptr, unsigned long old, unsigned long prev; switch (size) { case 1: - __asm__ __volatile__(LOCK_PREFIX "cmpxchgb %b1,%2" - : "=a"(prev) - : "q"(new), "m"(*__xg(ptr)), "0"(old) - : "memory"); + asm volatile(LOCK_PREFIX "cmpxchgb %b1,%2" + : "=a"(prev) + : "q"(new), "m"(*__xg(ptr)), "0"(old) + : "memory"); return prev; case 2: - __asm__ __volatile__(LOCK_PREFIX "cmpxchgw %w1,%2" - : "=a"(prev) - : "r"(new), "m"(*__xg(ptr)), "0"(old) - : "memory"); + asm volatile(LOCK_PREFIX "cmpxchgw %w1,%2" + : "=a"(prev) + : "r"(new), "m"(*__xg(ptr)), "0"(old) + : "memory"); return prev; case 4: - __asm__ __volatile__(LOCK_PREFIX "cmpxchgl %k1,%2" - : "=a"(prev) - : "r"(new), "m"(*__xg(ptr)), "0"(old) - : "memory"); + asm volatile(LOCK_PREFIX "cmpxchgl %k1,%2" + : "=a"(prev) + : "r"(new), "m"(*__xg(ptr)), "0"(old) + : "memory"); return prev; case 8: - __asm__ __volatile__(LOCK_PREFIX "cmpxchgq %1,%2" - : "=a"(prev) - : "r"(new), "m"(*__xg(ptr)), "0"(old) - : "memory"); + asm volatile(LOCK_PREFIX "cmpxchgq %1,%2" + : "=a"(prev) + : "r"(new), "m"(*__xg(ptr)), "0"(old) + : "memory"); return prev; } return old; } static inline unsigned long __cmpxchg_local(volatile void *ptr, - unsigned long old, unsigned long new, int size) + unsigned long old, + unsigned long new, int size) { unsigned long prev; switch (size) { case 1: - __asm__ __volatile__("cmpxchgb %b1,%2" - : "=a"(prev) - : "q"(new), "m"(*__xg(ptr)), "0"(old) - : "memory"); + asm volatile("cmpxchgb %b1,%2" + : "=a"(prev) + : "q"(new), "m"(*__xg(ptr)), "0"(old) + : "memory"); return prev; case 2: - __asm__ __volatile__("cmpxchgw %w1,%2" - : "=a"(prev) - : "r"(new), "m"(*__xg(ptr)), "0"(old) - : "memory"); + asm volatile("cmpxchgw %w1,%2" + : "=a"(prev) + : "r"(new), "m"(*__xg(ptr)), "0"(old) + : "memory"); return prev; case 4: - __asm__ __volatile__("cmpxchgl %k1,%2" - : "=a"(prev) - : "r"(new), "m"(*__xg(ptr)), "0"(old) - : "memory"); + asm volatile("cmpxchgl %k1,%2" + : "=a"(prev) + : "r"(new), "m"(*__xg(ptr)), "0"(old) + : "memory"); return prev; case 8: - __asm__ __volatile__("cmpxchgq %1,%2" - : "=a"(prev) - : "r"(new), "m"(*__xg(ptr)), "0"(old) - : "memory"); + asm volatile("cmpxchgq %1,%2" + : "=a"(prev) + : "r"(new), "m"(*__xg(ptr)), "0"(old) + : "memory"); return prev; } return old; @@ -126,19 +129,20 @@ static inline unsigned long __cmpxchg_local(volatile void *ptr, #define cmpxchg(ptr, o, n) \ ((__typeof__(*(ptr)))__cmpxchg((ptr), (unsigned long)(o), \ - (unsigned long)(n), sizeof(*(ptr)))) + (unsigned long)(n), sizeof(*(ptr)))) #define cmpxchg64(ptr, o, n) \ - ({ \ +({ \ BUILD_BUG_ON(sizeof(*(ptr)) != 8); \ cmpxchg((ptr), (o), (n)); \ - }) +}) #define cmpxchg_local(ptr, o, n) \ ((__typeof__(*(ptr)))__cmpxchg_local((ptr), (unsigned long)(o), \ - (unsigned long)(n), sizeof(*(ptr)))) + (unsigned long)(n), \ + sizeof(*(ptr)))) #define cmpxchg64_local(ptr, o, n) \ - ({ \ +({ \ BUILD_BUG_ON(sizeof(*(ptr)) != 8); \ cmpxchg_local((ptr), (o), (n)); \ - }) +}) #endif diff --git a/include/asm-x86/compat.h b/include/asm-x86/compat.h index d3e8f3e87ee8..1793ac317a30 100644 --- a/include/asm-x86/compat.h +++ b/include/asm-x86/compat.h @@ -204,7 +204,7 @@ static inline compat_uptr_t ptr_to_compat(void __user *uptr) return (u32)(unsigned long)uptr; } -static __inline__ void __user *compat_alloc_user_space(long len) +static inline void __user *compat_alloc_user_space(long len) { struct pt_regs *regs = task_pt_regs(current); return (void __user *)regs->sp - len; diff --git a/include/asm-x86/cpufeature.h b/include/asm-x86/cpufeature.h index 065e92966c7c..0d609c837a41 100644 --- a/include/asm-x86/cpufeature.h +++ b/include/asm-x86/cpufeature.h @@ -120,6 +120,9 @@ extern const char * const x86_cap_flags[NCAPINTS*32]; extern const char * const x86_power_flags[32]; +#define test_cpu_cap(c, bit) \ + test_bit(bit, (unsigned long *)((c)->x86_capability)) + #define cpu_has(c, bit) \ (__builtin_constant_p(bit) && \ ( (((bit)>>5)==0 && (1UL<<((bit)&31) & REQUIRED_MASK0)) || \ @@ -131,7 +134,8 @@ extern const char * const x86_power_flags[32]; (((bit)>>5)==6 && (1UL<<((bit)&31) & REQUIRED_MASK6)) || \ (((bit)>>5)==7 && (1UL<<((bit)&31) & REQUIRED_MASK7)) ) \ ? 1 : \ - test_bit(bit, (unsigned long *)((c)->x86_capability))) + test_cpu_cap(c, bit)) + #define boot_cpu_has(bit) cpu_has(&boot_cpu_data, bit) #define set_cpu_cap(c, bit) set_bit(bit, (unsigned long *)((c)->x86_capability)) @@ -181,6 +185,8 @@ extern const char * const x86_power_flags[32]; #define cpu_has_clflush boot_cpu_has(X86_FEATURE_CLFLSH) #define cpu_has_bts boot_cpu_has(X86_FEATURE_BTS) #define cpu_has_gbpages boot_cpu_has(X86_FEATURE_GBPAGES) +#define cpu_has_arch_perfmon boot_cpu_has(X86_FEATURE_ARCH_PERFMON) +#define cpu_has_pat boot_cpu_has(X86_FEATURE_PAT) #if defined(CONFIG_X86_INVLPG) || defined(CONFIG_X86_64) # define cpu_has_invlpg 1 diff --git a/include/asm-x86/current_32.h b/include/asm-x86/current_32.h index d35248539912..5af9bdb97a16 100644 --- a/include/asm-x86/current_32.h +++ b/include/asm-x86/current_32.h @@ -11,7 +11,7 @@ static __always_inline struct task_struct *get_current(void) { return x86_read_percpu(current_task); } - + #define current get_current() #endif /* !(_I386_CURRENT_H) */ diff --git a/include/asm-x86/current_64.h b/include/asm-x86/current_64.h index bc8adecee66d..2d368ede2fc1 100644 --- a/include/asm-x86/current_64.h +++ b/include/asm-x86/current_64.h @@ -1,23 +1,23 @@ #ifndef _X86_64_CURRENT_H #define _X86_64_CURRENT_H -#if !defined(__ASSEMBLY__) +#if !defined(__ASSEMBLY__) struct task_struct; #include <asm/pda.h> -static inline struct task_struct *get_current(void) -{ - struct task_struct *t = read_pda(pcurrent); +static inline struct task_struct *get_current(void) +{ + struct task_struct *t = read_pda(pcurrent); return t; -} +} #define current get_current() #else #ifndef ASM_OFFSET_H -#include <asm/asm-offsets.h> +#include <asm/asm-offsets.h> #endif #define GET_CURRENT(reg) movq %gs:(pda_pcurrent),reg diff --git a/include/asm-x86/desc.h b/include/asm-x86/desc.h index 5b6a05d3a771..268a012bcd79 100644 --- a/include/asm-x86/desc.h +++ b/include/asm-x86/desc.h @@ -62,8 +62,8 @@ static inline struct desc_struct *get_cpu_gdt_table(unsigned int cpu) } static inline void pack_gate(gate_desc *gate, unsigned char type, - unsigned long base, unsigned dpl, unsigned flags, unsigned short seg) - + unsigned long base, unsigned dpl, unsigned flags, + unsigned short seg) { gate->a = (seg << 16) | (base & 0xffff); gate->b = (base & 0xffff0000) | @@ -84,22 +84,23 @@ static inline int desc_empty(const void *ptr) #define load_TR_desc() native_load_tr_desc() #define load_gdt(dtr) native_load_gdt(dtr) #define load_idt(dtr) native_load_idt(dtr) -#define load_tr(tr) __asm__ __volatile("ltr %0"::"m" (tr)) -#define load_ldt(ldt) __asm__ __volatile("lldt %0"::"m" (ldt)) +#define load_tr(tr) asm volatile("ltr %0"::"m" (tr)) +#define load_ldt(ldt) asm volatile("lldt %0"::"m" (ldt)) #define store_gdt(dtr) native_store_gdt(dtr) #define store_idt(dtr) native_store_idt(dtr) #define store_tr(tr) (tr = native_store_tr()) -#define store_ldt(ldt) __asm__ ("sldt %0":"=m" (ldt)) +#define store_ldt(ldt) asm("sldt %0":"=m" (ldt)) #define load_TLS(t, cpu) native_load_tls(t, cpu) #define set_ldt native_set_ldt -#define write_ldt_entry(dt, entry, desc) \ - native_write_ldt_entry(dt, entry, desc) -#define write_gdt_entry(dt, entry, desc, type) \ - native_write_gdt_entry(dt, entry, desc, type) -#define write_idt_entry(dt, entry, g) native_write_idt_entry(dt, entry, g) +#define write_ldt_entry(dt, entry, desc) \ + native_write_ldt_entry(dt, entry, desc) +#define write_gdt_entry(dt, entry, desc, type) \ + native_write_gdt_entry(dt, entry, desc, type) +#define write_idt_entry(dt, entry, g) \ + native_write_idt_entry(dt, entry, g) #endif static inline void native_write_idt_entry(gate_desc *idt, int entry, @@ -138,8 +139,8 @@ static inline void pack_descriptor(struct desc_struct *desc, unsigned long base, { desc->a = ((base & 0xffff) << 16) | (limit & 0xffff); desc->b = (base & 0xff000000) | ((base & 0xff0000) >> 16) | - (limit & 0x000f0000) | ((type & 0xff) << 8) | - ((flags & 0xf) << 20); + (limit & 0x000f0000) | ((type & 0xff) << 8) | + ((flags & 0xf) << 20); desc->p = 1; } @@ -159,7 +160,6 @@ static inline void set_tssldt_descriptor(void *d, unsigned long addr, desc->base2 = (PTR_MIDDLE(addr) >> 8) & 0xFF; desc->base3 = PTR_HIGH(addr); #else - pack_descriptor((struct desc_struct *)d, addr, size, 0x80 | type, 0); #endif } @@ -177,7 +177,8 @@ static inline void __set_tss_desc(unsigned cpu, unsigned int entry, void *addr) * last valid byte */ set_tssldt_descriptor(&tss, (unsigned long)addr, DESC_TSS, - IO_BITMAP_OFFSET + IO_BITMAP_BYTES + sizeof(unsigned long) - 1); + IO_BITMAP_OFFSET + IO_BITMAP_BYTES + + sizeof(unsigned long) - 1); write_gdt_entry(d, entry, &tss, DESC_TSS); } @@ -186,7 +187,7 @@ static inline void __set_tss_desc(unsigned cpu, unsigned int entry, void *addr) static inline void native_set_ldt(const void *addr, unsigned int entries) { if (likely(entries == 0)) - __asm__ __volatile__("lldt %w0"::"q" (0)); + asm volatile("lldt %w0"::"q" (0)); else { unsigned cpu = smp_processor_id(); ldt_desc ldt; @@ -195,7 +196,7 @@ static inline void native_set_ldt(const void *addr, unsigned int entries) DESC_LDT, entries * sizeof(ldt) - 1); write_gdt_entry(get_cpu_gdt_table(cpu), GDT_ENTRY_LDT, &ldt, DESC_LDT); - __asm__ __volatile__("lldt %w0"::"q" (GDT_ENTRY_LDT*8)); + asm volatile("lldt %w0"::"q" (GDT_ENTRY_LDT*8)); } } @@ -240,15 +241,15 @@ static inline void native_load_tls(struct thread_struct *t, unsigned int cpu) gdt[GDT_ENTRY_TLS_MIN + i] = t->tls_array[i]; } -#define _LDT_empty(info) (\ - (info)->base_addr == 0 && \ - (info)->limit == 0 && \ - (info)->contents == 0 && \ - (info)->read_exec_only == 1 && \ - (info)->seg_32bit == 0 && \ - (info)->limit_in_pages == 0 && \ - (info)->seg_not_present == 1 && \ - (info)->useable == 0) +#define _LDT_empty(info) \ + ((info)->base_addr == 0 && \ + (info)->limit == 0 && \ + (info)->contents == 0 && \ + (info)->read_exec_only == 1 && \ + (info)->seg_32bit == 0 && \ + (info)->limit_in_pages == 0 && \ + (info)->seg_not_present == 1 && \ + (info)->useable == 0) #ifdef CONFIG_X86_64 #define LDT_empty(info) (_LDT_empty(info) && ((info)->lm == 0)) @@ -287,7 +288,7 @@ static inline unsigned long get_desc_limit(const struct desc_struct *desc) } static inline void _set_gate(int gate, unsigned type, void *addr, - unsigned dpl, unsigned ist, unsigned seg) + unsigned dpl, unsigned ist, unsigned seg) { gate_desc s; pack_gate(&s, type, (unsigned long)addr, dpl, ist, seg); @@ -370,10 +371,10 @@ static inline void set_system_gate_ist(int n, void *addr, unsigned ist) * Will read the base address of GDT_ENTRY_ESPFIX_SS and put it into %eax. */ #define GET_DESC_BASE(idx, gdt, base, lo_w, lo_b, hi_b) \ - movb idx*8+4(gdt), lo_b; \ - movb idx*8+7(gdt), hi_b; \ - shll $16, base; \ - movw idx*8+2(gdt), lo_w; + movb idx * 8 + 4(gdt), lo_b; \ + movb idx * 8 + 7(gdt), hi_b; \ + shll $16, base; \ + movw idx * 8 + 2(gdt), lo_w; #endif /* __ASSEMBLY__ */ diff --git a/include/asm-x86/desc_defs.h b/include/asm-x86/desc_defs.h index e33f078b3e54..eccb4ea1f918 100644 --- a/include/asm-x86/desc_defs.h +++ b/include/asm-x86/desc_defs.h @@ -18,17 +18,19 @@ * incrementally. We keep the signature as a struct, rather than an union, * so we can get rid of it transparently in the future -- glommer */ -// 8 byte segment descriptor +/* 8 byte segment descriptor */ struct desc_struct { union { - struct { unsigned int a, b; }; + struct { + unsigned int a; + unsigned int b; + }; struct { u16 limit0; u16 base0; unsigned base1: 8, type: 4, s: 1, dpl: 2, p: 1; unsigned limit: 4, avl: 1, l: 1, d: 1, g: 1, base2: 8; }; - }; } __attribute__((packed)); @@ -39,7 +41,7 @@ enum { GATE_TASK = 0x5, }; -// 16byte gate +/* 16byte gate */ struct gate_struct64 { u16 offset_low; u16 segment; @@ -56,10 +58,10 @@ struct gate_struct64 { enum { DESC_TSS = 0x9, DESC_LDT = 0x2, - DESCTYPE_S = 0x10, /* !system */ + DESCTYPE_S = 0x10, /* !system */ }; -// LDT or TSS descriptor in the GDT. 16 bytes. +/* LDT or TSS descriptor in the GDT. 16 bytes. */ struct ldttss_desc64 { u16 limit0; u16 base0; @@ -84,7 +86,6 @@ struct desc_ptr { unsigned long address; } __attribute__((packed)) ; - #endif /* !__ASSEMBLY__ */ #endif diff --git a/include/asm-x86/div64.h b/include/asm-x86/div64.h index e98d16e7a37a..0dbf8bf3ef0a 100644 --- a/include/asm-x86/div64.h +++ b/include/asm-x86/div64.h @@ -17,18 +17,20 @@ * This ends up being the most efficient "calling * convention" on x86. */ -#define do_div(n,base) ({ \ - unsigned long __upper, __low, __high, __mod, __base; \ - __base = (base); \ - asm("":"=a" (__low), "=d" (__high):"A" (n)); \ - __upper = __high; \ - if (__high) { \ - __upper = __high % (__base); \ - __high = __high / (__base); \ - } \ - asm("divl %2":"=a" (__low), "=d" (__mod):"rm" (__base), "0" (__low), "1" (__upper)); \ - asm("":"=A" (n):"a" (__low),"d" (__high)); \ - __mod; \ +#define do_div(n, base) \ +({ \ + unsigned long __upper, __low, __high, __mod, __base; \ + __base = (base); \ + asm("":"=a" (__low), "=d" (__high) : "A" (n)); \ + __upper = __high; \ + if (__high) { \ + __upper = __high % (__base); \ + __high = __high / (__base); \ + } \ + asm("divl %2":"=a" (__low), "=d" (__mod) \ + : "rm" (__base), "0" (__low), "1" (__upper)); \ + asm("":"=A" (n) : "a" (__low), "d" (__high)); \ + __mod; \ }) /* @@ -37,14 +39,13 @@ * * Warning, this will do an exception if X overflows. */ -#define div_long_long_rem(a,b,c) div_ll_X_l_rem(a,b,c) +#define div_long_long_rem(a, b, c) div_ll_X_l_rem(a, b, c) -static inline long -div_ll_X_l_rem(long long divs, long div, long *rem) +static inline long div_ll_X_l_rem(long long divs, long div, long *rem) { long dum2; - __asm__("divl %2":"=a"(dum2), "=d"(*rem) - : "rm"(div), "A"(divs)); + asm("divl %2":"=a"(dum2), "=d"(*rem) + : "rm"(div), "A"(divs)); return dum2; diff --git a/include/asm-x86/dma.h b/include/asm-x86/dma.h index e9733ce89880..ca1098a7e580 100644 --- a/include/asm-x86/dma.h +++ b/include/asm-x86/dma.h @@ -12,7 +12,6 @@ #include <asm/io.h> /* need byte IO */ #include <linux/delay.h> - #ifdef HAVE_REALLY_SLOW_DMA_CONTROLLER #define dma_outb outb_p #else @@ -74,15 +73,15 @@ #ifdef CONFIG_X86_32 /* The maximum address that we can perform a DMA transfer to on this platform */ -#define MAX_DMA_ADDRESS (PAGE_OFFSET+0x1000000) +#define MAX_DMA_ADDRESS (PAGE_OFFSET + 0x1000000) #else /* 16MB ISA DMA zone */ -#define MAX_DMA_PFN ((16*1024*1024) >> PAGE_SHIFT) +#define MAX_DMA_PFN ((16 * 1024 * 1024) >> PAGE_SHIFT) /* 4GB broken PCI/AGP hardware bus master zone */ -#define MAX_DMA32_PFN ((4UL*1024*1024*1024) >> PAGE_SHIFT) +#define MAX_DMA32_PFN ((4UL * 1024 * 1024 * 1024) >> PAGE_SHIFT) /* Compat define for old dma zone */ #define MAX_DMA_ADDRESS ((unsigned long)__va(MAX_DMA_PFN << PAGE_SHIFT)) @@ -154,20 +153,20 @@ extern spinlock_t dma_spin_lock; -static __inline__ unsigned long claim_dma_lock(void) +static inline unsigned long claim_dma_lock(void) { unsigned long flags; spin_lock_irqsave(&dma_spin_lock, flags); return flags; } -static __inline__ void release_dma_lock(unsigned long flags) +static inline void release_dma_lock(unsigned long flags) { spin_unlock_irqrestore(&dma_spin_lock, flags); } /* enable/disable a specific DMA channel */ -static __inline__ void enable_dma(unsigned int dmanr) +static inline void enable_dma(unsigned int dmanr) { if (dmanr <= 3) dma_outb(dmanr, DMA1_MASK_REG); @@ -175,7 +174,7 @@ static __inline__ void enable_dma(unsigned int dmanr) dma_outb(dmanr & 3, DMA2_MASK_REG); } -static __inline__ void disable_dma(unsigned int dmanr) +static inline void disable_dma(unsigned int dmanr) { if (dmanr <= 3) dma_outb(dmanr | 4, DMA1_MASK_REG); @@ -190,7 +189,7 @@ static __inline__ void disable_dma(unsigned int dmanr) * --- In order to do that, the DMA routines below should --- * --- only be used while holding the DMA lock ! --- */ -static __inline__ void clear_dma_ff(unsigned int dmanr) +static inline void clear_dma_ff(unsigned int dmanr) { if (dmanr <= 3) dma_outb(0, DMA1_CLEAR_FF_REG); @@ -199,7 +198,7 @@ static __inline__ void clear_dma_ff(unsigned int dmanr) } /* set mode (above) for a specific DMA channel */ -static __inline__ void set_dma_mode(unsigned int dmanr, char mode) +static inline void set_dma_mode(unsigned int dmanr, char mode) { if (dmanr <= 3) dma_outb(mode | dmanr, DMA1_MODE_REG); @@ -212,7 +211,7 @@ static __inline__ void set_dma_mode(unsigned int dmanr, char mode) * the lower 16 bits of the DMA current address register, but a 64k boundary * may have been crossed. */ -static __inline__ void set_dma_page(unsigned int dmanr, char pagenr) +static inline void set_dma_page(unsigned int dmanr, char pagenr) { switch (dmanr) { case 0: @@ -243,15 +242,15 @@ static __inline__ void set_dma_page(unsigned int dmanr, char pagenr) /* Set transfer address & page bits for specific DMA channel. * Assumes dma flipflop is clear. */ -static __inline__ void set_dma_addr(unsigned int dmanr, unsigned int a) +static inline void set_dma_addr(unsigned int dmanr, unsigned int a) { set_dma_page(dmanr, a>>16); if (dmanr <= 3) { dma_outb(a & 0xff, ((dmanr & 3) << 1) + IO_DMA1_BASE); dma_outb((a >> 8) & 0xff, ((dmanr & 3) << 1) + IO_DMA1_BASE); } else { - dma_outb((a >> 1) & 0xff, ((dmanr & 3) << 2) + IO_DMA2_BASE); - dma_outb((a >> 9) & 0xff, ((dmanr & 3) << 2) + IO_DMA2_BASE); + dma_outb((a >> 1) & 0xff, ((dmanr & 3) << 2) + IO_DMA2_BASE); + dma_outb((a >> 9) & 0xff, ((dmanr & 3) << 2) + IO_DMA2_BASE); } } @@ -264,18 +263,18 @@ static __inline__ void set_dma_addr(unsigned int dmanr, unsigned int a) * Assumes dma flip-flop is clear. * NOTE 2: "count" represents _bytes_ and must be even for channels 5-7. */ -static __inline__ void set_dma_count(unsigned int dmanr, unsigned int count) +static inline void set_dma_count(unsigned int dmanr, unsigned int count) { count--; if (dmanr <= 3) { - dma_outb(count & 0xff, ((dmanr & 3) << 1) + 1 + IO_DMA1_BASE); - dma_outb((count >> 8) & 0xff, - ((dmanr & 3) << 1) + 1 + IO_DMA1_BASE); + dma_outb(count & 0xff, ((dmanr & 3) << 1) + 1 + IO_DMA1_BASE); + dma_outb((count >> 8) & 0xff, + ((dmanr & 3) << 1) + 1 + IO_DMA1_BASE); } else { - dma_outb((count >> 1) & 0xff, - ((dmanr & 3) << 2) + 2 + IO_DMA2_BASE); - dma_outb((count >> 9) & 0xff, - ((dmanr & 3) << 2) + 2 + IO_DMA2_BASE); + dma_outb((count >> 1) & 0xff, + ((dmanr & 3) << 2) + 2 + IO_DMA2_BASE); + dma_outb((count >> 9) & 0xff, + ((dmanr & 3) << 2) + 2 + IO_DMA2_BASE); } } @@ -288,7 +287,7 @@ static __inline__ void set_dma_count(unsigned int dmanr, unsigned int count) * * Assumes DMA flip-flop is clear. */ -static __inline__ int get_dma_residue(unsigned int dmanr) +static inline int get_dma_residue(unsigned int dmanr) { unsigned int io_port; /* using short to get 16-bit wrap around */ diff --git a/include/asm-x86/dwarf2_64.h b/include/asm-x86/dwarf2_64.h index eedc08526b0b..c950519a264d 100644 --- a/include/asm-x86/dwarf2_64.h +++ b/include/asm-x86/dwarf2_64.h @@ -1,16 +1,15 @@ #ifndef _DWARF2_H #define _DWARF2_H 1 - #ifndef __ASSEMBLY__ #warning "asm/dwarf2.h should be only included in pure assembly files" #endif -/* +/* Macros for dwarf2 CFI unwind table entries. - See "as.info" for details on these pseudo ops. Unfortunately - they are only supported in very new binutils, so define them - away for older version. + See "as.info" for details on these pseudo ops. Unfortunately + they are only supported in very new binutils, so define them + away for older version. */ #ifdef CONFIG_AS_CFI diff --git a/include/asm-x86/e820_32.h b/include/asm-x86/e820_32.h index e7207a6de3e0..43b1a8bd4b34 100644 --- a/include/asm-x86/e820_32.h +++ b/include/asm-x86/e820_32.h @@ -34,8 +34,8 @@ extern void e820_register_memory(void); extern void limit_regions(unsigned long long size); extern void print_memory_map(char *who); extern void init_iomem_resources(struct resource *code_resource, - struct resource *data_resource, - struct resource *bss_resource); + struct resource *data_resource, + struct resource *bss_resource); #if defined(CONFIG_PM) && defined(CONFIG_HIBERNATION) extern void e820_mark_nosave_regions(void); diff --git a/include/asm-x86/e820_64.h b/include/asm-x86/e820_64.h index 22ede73ae724..f478c57eb060 100644 --- a/include/asm-x86/e820_64.h +++ b/include/asm-x86/e820_64.h @@ -14,20 +14,24 @@ #include <linux/ioport.h> #ifndef __ASSEMBLY__ -extern unsigned long find_e820_area(unsigned long start, unsigned long end, - unsigned size, unsigned long align); -extern void add_memory_region(unsigned long start, unsigned long size, +extern unsigned long find_e820_area(unsigned long start, unsigned long end, + unsigned long size, unsigned long align); +extern unsigned long find_e820_area_size(unsigned long start, + unsigned long *sizep, + unsigned long align); +extern void add_memory_region(unsigned long start, unsigned long size, int type); extern void update_memory_range(u64 start, u64 size, unsigned old_type, unsigned new_type); extern void setup_memory_region(void); -extern void contig_e820_setup(void); +extern void contig_e820_setup(void); extern unsigned long e820_end_of_ram(void); -extern void e820_reserve_resources(struct resource *code_resource, - struct resource *data_resource, struct resource *bss_resource); +extern void e820_reserve_resources(void); extern void e820_mark_nosave_regions(void); -extern int e820_any_mapped(unsigned long start, unsigned long end, unsigned type); -extern int e820_all_mapped(unsigned long start, unsigned long end, unsigned type); +extern int e820_any_mapped(unsigned long start, unsigned long end, + unsigned type); +extern int e820_all_mapped(unsigned long start, unsigned long end, + unsigned type); extern int e820_any_non_reserved(unsigned long start, unsigned long end); extern int is_memory_any_valid(unsigned long start, unsigned long end); extern int e820_all_non_reserved(unsigned long start, unsigned long end); @@ -35,8 +39,8 @@ extern int is_memory_all_valid(unsigned long start, unsigned long end); extern unsigned long e820_hole_size(unsigned long start, unsigned long end); extern void e820_setup_gap(void); -extern void e820_register_active_regions(int nid, - unsigned long start_pfn, unsigned long end_pfn); +extern void e820_register_active_regions(int nid, unsigned long start_pfn, + unsigned long end_pfn); extern void finish_e820_parsing(void); diff --git a/include/asm-x86/edac.h b/include/asm-x86/edac.h index cf3200a745ad..a8088f63a30e 100644 --- a/include/asm-x86/edac.h +++ b/include/asm-x86/edac.h @@ -3,7 +3,7 @@ /* ECC atomic, DMA, SMP and interrupt safe scrub function */ -static __inline__ void atomic_scrub(void *va, u32 size) +static inline void atomic_scrub(void *va, u32 size) { u32 i, *virt_addr = va; @@ -12,7 +12,7 @@ static __inline__ void atomic_scrub(void *va, u32 size) * are interrupt, DMA and SMP safe. */ for (i = 0; i < size / 4; i++, virt_addr++) - __asm__ __volatile__("lock; addl $0, %0"::"m"(*virt_addr)); + asm volatile("lock; addl $0, %0"::"m" (*virt_addr)); } #endif diff --git a/include/asm-x86/efi.h b/include/asm-x86/efi.h index ea9734b74aca..d53004b855cc 100644 --- a/include/asm-x86/efi.h +++ b/include/asm-x86/efi.h @@ -20,7 +20,7 @@ extern unsigned long asmlinkage efi_call_phys(void *, ...); */ #define efi_call_virt(f, args...) \ - ((efi_##f##_t __attribute__((regparm(0)))*)efi.systab->runtime->f)(args) + ((efi_##f##_t __attribute__((regparm(0)))*)efi.systab->runtime->f)(args) #define efi_call_virt0(f) efi_call_virt(f) #define efi_call_virt1(f, a1) efi_call_virt(f, a1) diff --git a/include/asm-x86/elf.h b/include/asm-x86/elf.h index fb62f9941e38..8f232dc5b5fe 100644 --- a/include/asm-x86/elf.h +++ b/include/asm-x86/elf.h @@ -11,7 +11,7 @@ typedef unsigned long elf_greg_t; -#define ELF_NGREG (sizeof (struct user_regs_struct) / sizeof(elf_greg_t)) +#define ELF_NGREG (sizeof(struct user_regs_struct) / sizeof(elf_greg_t)) typedef elf_greg_t elf_gregset_t[ELF_NGREG]; typedef struct user_i387_struct elf_fpregset_t; @@ -82,8 +82,9 @@ extern unsigned int vdso_enabled; #define elf_check_arch_ia32(x) \ (((x)->e_machine == EM_386) || ((x)->e_machine == EM_486)) -#ifdef CONFIG_X86_32 #include <asm/processor.h> + +#ifdef CONFIG_X86_32 #include <asm/system.h> /* for savesegment */ #include <asm/desc.h> @@ -99,10 +100,11 @@ extern unsigned int vdso_enabled; We might as well make sure everything else is cleared too (except for %esp), just to make things more deterministic. */ -#define ELF_PLAT_INIT(_r, load_addr) do { \ - _r->bx = 0; _r->cx = 0; _r->dx = 0; \ - _r->si = 0; _r->di = 0; _r->bp = 0; \ - _r->ax = 0; \ +#define ELF_PLAT_INIT(_r, load_addr) \ + do { \ + _r->bx = 0; _r->cx = 0; _r->dx = 0; \ + _r->si = 0; _r->di = 0; _r->bp = 0; \ + _r->ax = 0; \ } while (0) /* @@ -110,24 +112,25 @@ extern unsigned int vdso_enabled; * now struct_user_regs, they are different) */ -#define ELF_CORE_COPY_REGS(pr_reg, regs) do { \ - pr_reg[0] = regs->bx; \ - pr_reg[1] = regs->cx; \ - pr_reg[2] = regs->dx; \ - pr_reg[3] = regs->si; \ - pr_reg[4] = regs->di; \ - pr_reg[5] = regs->bp; \ - pr_reg[6] = regs->ax; \ - pr_reg[7] = regs->ds & 0xffff; \ - pr_reg[8] = regs->es & 0xffff; \ - pr_reg[9] = regs->fs & 0xffff; \ - savesegment(gs, pr_reg[10]); \ - pr_reg[11] = regs->orig_ax; \ - pr_reg[12] = regs->ip; \ - pr_reg[13] = regs->cs & 0xffff; \ - pr_reg[14] = regs->flags; \ - pr_reg[15] = regs->sp; \ - pr_reg[16] = regs->ss & 0xffff; \ +#define ELF_CORE_COPY_REGS(pr_reg, regs) \ +do { \ + pr_reg[0] = regs->bx; \ + pr_reg[1] = regs->cx; \ + pr_reg[2] = regs->dx; \ + pr_reg[3] = regs->si; \ + pr_reg[4] = regs->di; \ + pr_reg[5] = regs->bp; \ + pr_reg[6] = regs->ax; \ + pr_reg[7] = regs->ds & 0xffff; \ + pr_reg[8] = regs->es & 0xffff; \ + pr_reg[9] = regs->fs & 0xffff; \ + savesegment(gs, pr_reg[10]); \ + pr_reg[11] = regs->orig_ax; \ + pr_reg[12] = regs->ip; \ + pr_reg[13] = regs->cs & 0xffff; \ + pr_reg[14] = regs->flags; \ + pr_reg[15] = regs->sp; \ + pr_reg[16] = regs->ss & 0xffff; \ } while (0); #define ELF_PLATFORM (utsname()->machine) @@ -135,12 +138,10 @@ extern unsigned int vdso_enabled; #else /* CONFIG_X86_32 */ -#include <asm/processor.h> - /* * This is used to ensure we don't load something for the wrong architecture. */ -#define elf_check_arch(x) \ +#define elf_check_arch(x) \ ((x)->e_machine == EM_X86_64) #define compat_elf_check_arch(x) elf_check_arch_ia32(x) @@ -169,24 +170,30 @@ static inline void elf_common_init(struct thread_struct *t, t->ds = t->es = ds; } -#define ELF_PLAT_INIT(_r, load_addr) do { \ - elf_common_init(¤t->thread, _r, 0); \ - clear_thread_flag(TIF_IA32); \ +#define ELF_PLAT_INIT(_r, load_addr) \ +do { \ + elf_common_init(¤t->thread, _r, 0); \ + clear_thread_flag(TIF_IA32); \ } while (0) -#define COMPAT_ELF_PLAT_INIT(regs, load_addr) \ +#define COMPAT_ELF_PLAT_INIT(regs, load_addr) \ elf_common_init(¤t->thread, regs, __USER_DS) -#define compat_start_thread(regs, ip, sp) do { \ - start_ia32_thread(regs, ip, sp); \ - set_fs(USER_DS); \ - } while (0) -#define COMPAT_SET_PERSONALITY(ex, ibcs2) do { \ - if (test_thread_flag(TIF_IA32)) \ - clear_thread_flag(TIF_ABI_PENDING); \ - else \ - set_thread_flag(TIF_ABI_PENDING); \ - current->personality |= force_personality32; \ - } while (0) + +#define compat_start_thread(regs, ip, sp) \ +do { \ + start_ia32_thread(regs, ip, sp); \ + set_fs(USER_DS); \ +} while (0) + +#define COMPAT_SET_PERSONALITY(ex, ibcs2) \ +do { \ + if (test_thread_flag(TIF_IA32)) \ + clear_thread_flag(TIF_ABI_PENDING); \ + else \ + set_thread_flag(TIF_ABI_PENDING); \ + current->personality |= force_personality32; \ +} while (0) + #define COMPAT_ELF_PLATFORM ("i686") /* @@ -195,7 +202,8 @@ static inline void elf_common_init(struct thread_struct *t, * getting dumped. */ -#define ELF_CORE_COPY_REGS(pr_reg, regs) do { \ +#define ELF_CORE_COPY_REGS(pr_reg, regs) \ +do { \ unsigned v; \ (pr_reg)[0] = (regs)->r15; \ (pr_reg)[1] = (regs)->r14; \ @@ -269,10 +277,12 @@ extern int force_personality32; struct task_struct; -#define ARCH_DLINFO_IA32(vdso_enabled) \ -do if (vdso_enabled) { \ +#define ARCH_DLINFO_IA32(vdso_enabled) \ +do { \ + if (vdso_enabled) { \ NEW_AUX_ENT(AT_SYSINFO, VDSO_ENTRY); \ NEW_AUX_ENT(AT_SYSINFO_EHDR, VDSO_CURRENT_BASE); \ + } \ } while (0) #ifdef CONFIG_X86_32 @@ -290,9 +300,11 @@ do if (vdso_enabled) { \ /* 1GB for 64bit, 8MB for 32bit */ #define STACK_RND_MASK (test_thread_flag(TIF_IA32) ? 0x7ff : 0x3fffff) -#define ARCH_DLINFO \ -do if (vdso_enabled) { \ - NEW_AUX_ENT(AT_SYSINFO_EHDR,(unsigned long)current->mm->context.vdso);\ +#define ARCH_DLINFO \ +do { \ + if (vdso_enabled) \ + NEW_AUX_ENT(AT_SYSINFO_EHDR, \ + (unsigned long)current->mm->context.vdso); \ } while (0) #define AT_SYSINFO 32 @@ -305,8 +317,8 @@ do if (vdso_enabled) { \ #define VDSO_CURRENT_BASE ((unsigned long)current->mm->context.vdso) -#define VDSO_ENTRY \ - ((unsigned long) VDSO32_SYMBOL(VDSO_CURRENT_BASE, vsyscall)) +#define VDSO_ENTRY \ + ((unsigned long)VDSO32_SYMBOL(VDSO_CURRENT_BASE, vsyscall)) struct linux_binprm; diff --git a/include/asm-x86/fixmap_32.h b/include/asm-x86/fixmap_32.h index a7404d50686b..eb1665125c44 100644 --- a/include/asm-x86/fixmap_32.h +++ b/include/asm-x86/fixmap_32.h @@ -99,8 +99,7 @@ enum fixed_addresses { */ #define NR_FIX_BTMAPS 64 #define FIX_BTMAPS_NESTING 4 - FIX_BTMAP_END = - __end_of_permanent_fixed_addresses + 512 - + FIX_BTMAP_END = __end_of_permanent_fixed_addresses + 512 - (__end_of_permanent_fixed_addresses & 511), FIX_BTMAP_BEGIN = FIX_BTMAP_END + NR_FIX_BTMAPS*FIX_BTMAPS_NESTING - 1, FIX_WP_TEST, @@ -110,20 +109,20 @@ enum fixed_addresses { __end_of_fixed_addresses }; -extern void __set_fixmap (enum fixed_addresses idx, - unsigned long phys, pgprot_t flags); +extern void __set_fixmap(enum fixed_addresses idx, + unsigned long phys, pgprot_t flags); extern void reserve_top_address(unsigned long reserve); -#define set_fixmap(idx, phys) \ - __set_fixmap(idx, phys, PAGE_KERNEL) +#define set_fixmap(idx, phys) \ + __set_fixmap(idx, phys, PAGE_KERNEL) /* * Some hardware wants to get fixmapped without caching. */ -#define set_fixmap_nocache(idx, phys) \ - __set_fixmap(idx, phys, PAGE_KERNEL_NOCACHE) +#define set_fixmap_nocache(idx, phys) \ + __set_fixmap(idx, phys, PAGE_KERNEL_NOCACHE) -#define clear_fixmap(idx) \ - __set_fixmap(idx, 0, __pgprot(0)) +#define clear_fixmap(idx) \ + __set_fixmap(idx, 0, __pgprot(0)) #define FIXADDR_TOP ((unsigned long)__FIXADDR_TOP) @@ -156,7 +155,7 @@ static __always_inline unsigned long fix_to_virt(const unsigned int idx) if (idx >= __end_of_fixed_addresses) __this_fixmap_does_not_exist(); - return __fix_to_virt(idx); + return __fix_to_virt(idx); } static inline unsigned long virt_to_fix(const unsigned long vaddr) diff --git a/include/asm-x86/fixmap_64.h b/include/asm-x86/fixmap_64.h index 70ddb21e6458..f3d76858c0e6 100644 --- a/include/asm-x86/fixmap_64.h +++ b/include/asm-x86/fixmap_64.h @@ -34,32 +34,34 @@ enum fixed_addresses { VSYSCALL_LAST_PAGE, - VSYSCALL_FIRST_PAGE = VSYSCALL_LAST_PAGE + ((VSYSCALL_END-VSYSCALL_START) >> PAGE_SHIFT) - 1, + VSYSCALL_FIRST_PAGE = VSYSCALL_LAST_PAGE + + ((VSYSCALL_END-VSYSCALL_START) >> PAGE_SHIFT) - 1, VSYSCALL_HPET, FIX_DBGP_BASE, FIX_EARLYCON_MEM_BASE, FIX_HPET_BASE, FIX_APIC_BASE, /* local (CPU) APIC) -- required for SMP or not */ FIX_IO_APIC_BASE_0, - FIX_IO_APIC_BASE_END = FIX_IO_APIC_BASE_0 + MAX_IO_APICS-1, + FIX_IO_APIC_BASE_END = FIX_IO_APIC_BASE_0 + MAX_IO_APICS - 1, FIX_EFI_IO_MAP_LAST_PAGE, - FIX_EFI_IO_MAP_FIRST_PAGE = FIX_EFI_IO_MAP_LAST_PAGE+MAX_EFI_IO_PAGES-1, + FIX_EFI_IO_MAP_FIRST_PAGE = FIX_EFI_IO_MAP_LAST_PAGE + + MAX_EFI_IO_PAGES - 1, #ifdef CONFIG_PROVIDE_OHCI1394_DMA_INIT FIX_OHCI1394_BASE, #endif __end_of_fixed_addresses }; -extern void __set_fixmap (enum fixed_addresses idx, - unsigned long phys, pgprot_t flags); +extern void __set_fixmap(enum fixed_addresses idx, + unsigned long phys, pgprot_t flags); -#define set_fixmap(idx, phys) \ - __set_fixmap(idx, phys, PAGE_KERNEL) +#define set_fixmap(idx, phys) \ + __set_fixmap(idx, phys, PAGE_KERNEL) /* * Some hardware wants to get fixmapped without caching. */ -#define set_fixmap_nocache(idx, phys) \ - __set_fixmap(idx, phys, PAGE_KERNEL_NOCACHE) +#define set_fixmap_nocache(idx, phys) \ + __set_fixmap(idx, phys, PAGE_KERNEL_NOCACHE) #define FIXADDR_TOP (VSYSCALL_END-PAGE_SIZE) #define FIXADDR_SIZE (__end_of_fixed_addresses << PAGE_SHIFT) diff --git a/include/asm-x86/floppy.h b/include/asm-x86/floppy.h index a48d7153c097..dbe82a5c5eac 100644 --- a/include/asm-x86/floppy.h +++ b/include/asm-x86/floppy.h @@ -20,20 +20,21 @@ * driver otherwise. It doesn't matter much for performance anyway, as most * floppy accesses go through the track buffer. */ -#define _CROSS_64KB(a,s,vdma) \ -(!(vdma) && ((unsigned long)(a)/K_64 != ((unsigned long)(a) + (s) - 1) / K_64)) +#define _CROSS_64KB(a, s, vdma) \ + (!(vdma) && \ + ((unsigned long)(a)/K_64 != ((unsigned long)(a) + (s) - 1) / K_64)) -#define CROSS_64KB(a,s) _CROSS_64KB(a,s,use_virtual_dma & 1) +#define CROSS_64KB(a, s) _CROSS_64KB(a, s, use_virtual_dma & 1) -#define SW fd_routine[use_virtual_dma&1] +#define SW fd_routine[use_virtual_dma & 1] #define CSW fd_routine[can_use_virtual_dma & 1] #define fd_inb(port) inb_p(port) -#define fd_outb(value,port) outb_p(value,port) +#define fd_outb(value, port) outb_p(value, port) -#define fd_request_dma() CSW._request_dma(FLOPPY_DMA,"floppy") +#define fd_request_dma() CSW._request_dma(FLOPPY_DMA, "floppy") #define fd_free_dma() CSW._free_dma(FLOPPY_DMA) #define fd_enable_irq() enable_irq(FLOPPY_IRQ) #define fd_disable_irq() disable_irq(FLOPPY_IRQ) @@ -52,64 +53,64 @@ static int doing_pdma; static irqreturn_t floppy_hardint(int irq, void *dev_id) { - register unsigned char st; + unsigned char st; #undef TRACE_FLPY_INT #ifdef TRACE_FLPY_INT - static int calls=0; - static int bytes=0; - static int dma_wait=0; + static int calls; + static int bytes; + static int dma_wait; #endif if (!doing_pdma) return floppy_interrupt(irq, dev_id); #ifdef TRACE_FLPY_INT - if(!calls) + if (!calls) bytes = virtual_dma_count; #endif { - register int lcount; - register char *lptr; + int lcount; + char *lptr; st = 1; - for(lcount=virtual_dma_count, lptr=virtual_dma_addr; - lcount; lcount--, lptr++) { - st=inb(virtual_dma_port+4) & 0xa0 ; - if(st != 0xa0) + for (lcount = virtual_dma_count, lptr = virtual_dma_addr; + lcount; lcount--, lptr++) { + st = inb(virtual_dma_port + 4) & 0xa0; + if (st != 0xa0) break; - if(virtual_dma_mode) - outb_p(*lptr, virtual_dma_port+5); + if (virtual_dma_mode) + outb_p(*lptr, virtual_dma_port + 5); else - *lptr = inb_p(virtual_dma_port+5); + *lptr = inb_p(virtual_dma_port + 5); } virtual_dma_count = lcount; virtual_dma_addr = lptr; - st = inb(virtual_dma_port+4); + st = inb(virtual_dma_port + 4); } #ifdef TRACE_FLPY_INT calls++; #endif - if(st == 0x20) + if (st == 0x20) return IRQ_HANDLED; - if(!(st & 0x20)) { + if (!(st & 0x20)) { virtual_dma_residue += virtual_dma_count; - virtual_dma_count=0; + virtual_dma_count = 0; #ifdef TRACE_FLPY_INT printk("count=%x, residue=%x calls=%d bytes=%d dma_wait=%d\n", virtual_dma_count, virtual_dma_residue, calls, bytes, dma_wait); calls = 0; - dma_wait=0; + dma_wait = 0; #endif doing_pdma = 0; floppy_interrupt(irq, dev_id); return IRQ_HANDLED; } #ifdef TRACE_FLPY_INT - if(!virtual_dma_count) + if (!virtual_dma_count) dma_wait++; #endif return IRQ_HANDLED; @@ -117,14 +118,14 @@ static irqreturn_t floppy_hardint(int irq, void *dev_id) static void fd_disable_dma(void) { - if(! (can_use_virtual_dma & 1)) + if (!(can_use_virtual_dma & 1)) disable_dma(FLOPPY_DMA); doing_pdma = 0; virtual_dma_residue += virtual_dma_count; - virtual_dma_count=0; + virtual_dma_count = 0; } -static int vdma_request_dma(unsigned int dmanr, const char * device_id) +static int vdma_request_dma(unsigned int dmanr, const char *device_id) { return 0; } @@ -142,7 +143,7 @@ static int vdma_get_dma_residue(unsigned int dummy) static int fd_request_irq(void) { - if(can_use_virtual_dma) + if (can_use_virtual_dma) return request_irq(FLOPPY_IRQ, floppy_hardint, IRQF_DISABLED, "floppy", NULL); else @@ -152,13 +153,13 @@ static int fd_request_irq(void) static unsigned long dma_mem_alloc(unsigned long size) { - return __get_dma_pages(GFP_KERNEL|__GFP_NORETRY,get_order(size)); + return __get_dma_pages(GFP_KERNEL|__GFP_NORETRY, get_order(size)); } static unsigned long vdma_mem_alloc(unsigned long size) { - return (unsigned long) vmalloc(size); + return (unsigned long)vmalloc(size); } @@ -166,7 +167,7 @@ static unsigned long vdma_mem_alloc(unsigned long size) static void _fd_dma_mem_free(unsigned long addr, unsigned long size) { - if((unsigned long) addr >= (unsigned long) high_memory) + if ((unsigned long)addr >= (unsigned long)high_memory) vfree((void *)addr); else free_pages(addr, get_order(size)); @@ -176,10 +177,10 @@ static void _fd_dma_mem_free(unsigned long addr, unsigned long size) static void _fd_chose_dma_mode(char *addr, unsigned long size) { - if(can_use_virtual_dma == 2) { - if((unsigned long) addr >= (unsigned long) high_memory || - isa_virt_to_bus(addr) >= 0x1000000 || - _CROSS_64KB(addr, size, 0)) + if (can_use_virtual_dma == 2) { + if ((unsigned long)addr >= (unsigned long)high_memory || + isa_virt_to_bus(addr) >= 0x1000000 || + _CROSS_64KB(addr, size, 0)) use_virtual_dma = 1; else use_virtual_dma = 0; @@ -195,7 +196,7 @@ static int vdma_dma_setup(char *addr, unsigned long size, int mode, int io) { doing_pdma = 1; virtual_dma_port = io; - virtual_dma_mode = (mode == DMA_MODE_WRITE); + virtual_dma_mode = (mode == DMA_MODE_WRITE); virtual_dma_addr = addr; virtual_dma_count = size; virtual_dma_residue = 0; @@ -213,18 +214,18 @@ static int hard_dma_setup(char *addr, unsigned long size, int mode, int io) /* actual, physical DMA */ doing_pdma = 0; clear_dma_ff(FLOPPY_DMA); - set_dma_mode(FLOPPY_DMA,mode); - set_dma_addr(FLOPPY_DMA,isa_virt_to_bus(addr)); - set_dma_count(FLOPPY_DMA,size); + set_dma_mode(FLOPPY_DMA, mode); + set_dma_addr(FLOPPY_DMA, isa_virt_to_bus(addr)); + set_dma_count(FLOPPY_DMA, size); enable_dma(FLOPPY_DMA); return 0; } static struct fd_routine_l { - int (*_request_dma)(unsigned int dmanr, const char * device_id); + int (*_request_dma)(unsigned int dmanr, const char *device_id); void (*_free_dma)(unsigned int dmanr); int (*_get_dma_residue)(unsigned int dummy); - unsigned long (*_dma_mem_alloc) (unsigned long size); + unsigned long (*_dma_mem_alloc)(unsigned long size); int (*_dma_setup)(char *addr, unsigned long size, int mode, int io); } fd_routine[] = { { @@ -252,7 +253,8 @@ static int FDC2 = -1; * is needed to prevent corrupted CMOS RAM in case "insmod floppy" * coincides with another rtc CMOS user. Paul G. */ -#define FLOPPY0_TYPE ({ \ +#define FLOPPY0_TYPE \ +({ \ unsigned long flags; \ unsigned char val; \ spin_lock_irqsave(&rtc_lock, flags); \ @@ -261,7 +263,8 @@ static int FDC2 = -1; val; \ }) -#define FLOPPY1_TYPE ({ \ +#define FLOPPY1_TYPE \ +({ \ unsigned long flags; \ unsigned char val; \ spin_lock_irqsave(&rtc_lock, flags); \ diff --git a/include/asm-x86/futex.h b/include/asm-x86/futex.h index c9952ea9f698..ac0fbf24d722 100644 --- a/include/asm-x86/futex.h +++ b/include/asm-x86/futex.h @@ -12,35 +12,32 @@ #include <asm/uaccess.h> #define __futex_atomic_op1(insn, ret, oldval, uaddr, oparg) \ - __asm__ __volatile( \ -"1: " insn "\n" \ -"2: .section .fixup,\"ax\"\n \ -3: mov %3, %1\n \ - jmp 2b\n \ - .previous\n" \ - _ASM_EXTABLE(1b,3b) \ - : "=r" (oldval), "=r" (ret), "+m" (*uaddr) \ - : "i" (-EFAULT), "0" (oparg), "1" (0)) + asm volatile("1:\t" insn "\n" \ + "2:\t.section .fixup,\"ax\"\n" \ + "3:\tmov\t%3, %1\n" \ + "\tjmp\t2b\n" \ + "\t.previous\n" \ + _ASM_EXTABLE(1b, 3b) \ + : "=r" (oldval), "=r" (ret), "+m" (*uaddr) \ + : "i" (-EFAULT), "0" (oparg), "1" (0)) #define __futex_atomic_op2(insn, ret, oldval, uaddr, oparg) \ - __asm__ __volatile( \ -"1: movl %2, %0\n \ - movl %0, %3\n" \ - insn "\n" \ -"2: lock; cmpxchgl %3, %2\n \ - jnz 1b\n \ -3: .section .fixup,\"ax\"\n \ -4: mov %5, %1\n \ - jmp 3b\n \ - .previous\n" \ - _ASM_EXTABLE(1b,4b) \ - _ASM_EXTABLE(2b,4b) \ - : "=&a" (oldval), "=&r" (ret), "+m" (*uaddr), \ - "=&r" (tem) \ - : "r" (oparg), "i" (-EFAULT), "1" (0)) - -static inline int -futex_atomic_op_inuser(int encoded_op, int __user *uaddr) + asm volatile("1:\tmovl %2, %0\n" \ + "\tmovl\t%0, %3\n" \ + "\t" insn "\n" \ + "2:\tlock; cmpxchgl %3, %2\n" \ + "\tjnz\t1b\n" \ + "3:\t.section .fixup,\"ax\"\n" \ + "4:\tmov\t%5, %1\n" \ + "\tjmp\t3b\n" \ + "\t.previous\n" \ + _ASM_EXTABLE(1b, 4b) \ + _ASM_EXTABLE(2b, 4b) \ + : "=&a" (oldval), "=&r" (ret), \ + "+m" (*uaddr), "=&r" (tem) \ + : "r" (oparg), "i" (-EFAULT), "1" (0)) + +static inline int futex_atomic_op_inuser(int encoded_op, int __user *uaddr) { int op = (encoded_op >> 28) & 7; int cmp = (encoded_op >> 24) & 15; @@ -87,20 +84,33 @@ futex_atomic_op_inuser(int encoded_op, int __user *uaddr) if (!ret) { switch (cmp) { - case FUTEX_OP_CMP_EQ: ret = (oldval == cmparg); break; - case FUTEX_OP_CMP_NE: ret = (oldval != cmparg); break; - case FUTEX_OP_CMP_LT: ret = (oldval < cmparg); break; - case FUTEX_OP_CMP_GE: ret = (oldval >= cmparg); break; - case FUTEX_OP_CMP_LE: ret = (oldval <= cmparg); break; - case FUTEX_OP_CMP_GT: ret = (oldval > cmparg); break; - default: ret = -ENOSYS; + case FUTEX_OP_CMP_EQ: + ret = (oldval == cmparg); + break; + case FUTEX_OP_CMP_NE: + ret = (oldval != cmparg); + break; + case FUTEX_OP_CMP_LT: + ret = (oldval < cmparg); + break; + case FUTEX_OP_CMP_GE: + ret = (oldval >= cmparg); + break; + case FUTEX_OP_CMP_LE: + ret = (oldval <= cmparg); + break; + case FUTEX_OP_CMP_GT: + ret = (oldval > cmparg); + break; + default: + ret = -ENOSYS; } } return ret; } -static inline int -futex_atomic_cmpxchg_inatomic(int __user *uaddr, int oldval, int newval) +static inline int futex_atomic_cmpxchg_inatomic(int __user *uaddr, int oldval, + int newval) { #if defined(CONFIG_X86_32) && !defined(CONFIG_X86_BSWAP) @@ -112,16 +122,15 @@ futex_atomic_cmpxchg_inatomic(int __user *uaddr, int oldval, int newval) if (!access_ok(VERIFY_WRITE, uaddr, sizeof(int))) return -EFAULT; - __asm__ __volatile__( - "1: lock; cmpxchgl %3, %1 \n" - "2: .section .fixup, \"ax\" \n" - "3: mov %2, %0 \n" - " jmp 2b \n" - " .previous \n" - _ASM_EXTABLE(1b,3b) - : "=a" (oldval), "+m" (*uaddr) - : "i" (-EFAULT), "r" (newval), "0" (oldval) - : "memory" + asm volatile("1:\tlock; cmpxchgl %3, %1\n" + "2:\t.section .fixup, \"ax\"\n" + "3:\tmov %2, %0\n" + "\tjmp 2b\n" + "\t.previous\n" + _ASM_EXTABLE(1b, 3b) + : "=a" (oldval), "+m" (*uaddr) + : "i" (-EFAULT), "r" (newval), "0" (oldval) + : "memory" ); return oldval; diff --git a/include/asm-x86/genapic_32.h b/include/asm-x86/genapic_32.h index 33e3ffe1766c..f1b96932746b 100644 --- a/include/asm-x86/genapic_32.h +++ b/include/asm-x86/genapic_32.h @@ -14,23 +14,22 @@ * Copyright 2003 Andi Kleen, SuSE Labs. */ -struct mpc_config_translation; struct mpc_config_bus; struct mp_config_table; struct mpc_config_processor; -struct genapic { - char *name; - int (*probe)(void); +struct genapic { + char *name; + int (*probe)(void); int (*apic_id_registered)(void); cpumask_t (*target_cpus)(void); int int_delivery_mode; - int int_dest_mode; + int int_dest_mode; int ESR_DISABLE; int apic_destination_logical; unsigned long (*check_apicid_used)(physid_mask_t bitmap, int apicid); - unsigned long (*check_apicid_present)(int apicid); + unsigned long (*check_apicid_present)(int apicid); int no_balance_irq; int no_ioapic_check; void (*init_apic_ldr)(void); @@ -38,28 +37,21 @@ struct genapic { void (*setup_apic_routing)(void); int (*multi_timer_check)(int apic, int irq); - int (*apicid_to_node)(int logical_apicid); + int (*apicid_to_node)(int logical_apicid); int (*cpu_to_logical_apicid)(int cpu); int (*cpu_present_to_apicid)(int mps_cpu); physid_mask_t (*apicid_to_cpu_present)(int phys_apicid); - int (*mpc_apic_id)(struct mpc_config_processor *m, - struct mpc_config_translation *t); - void (*setup_portio_remap)(void); + void (*setup_portio_remap)(void); int (*check_phys_apicid_present)(int boot_cpu_physical_apicid); void (*enable_apic_mode)(void); u32 (*phys_pkg_id)(u32 cpuid_apic, int index_msb); /* mpparse */ - void (*mpc_oem_bus_info)(struct mpc_config_bus *, char *, - struct mpc_config_translation *); - void (*mpc_oem_pci_bus)(struct mpc_config_bus *, - struct mpc_config_translation *); - /* When one of the next two hooks returns 1 the genapic - is switched to this. Essentially they are additional probe + is switched to this. Essentially they are additional probe functions. */ - int (*mps_oem_check)(struct mp_config_table *mpc, char *oem, - char *productid); + int (*mps_oem_check)(struct mp_config_table *mpc, char *oem, + char *productid); int (*acpi_madt_oem_check)(char *oem_id, char *oem_table_id); unsigned (*get_apic_id)(unsigned long x); @@ -72,7 +64,7 @@ struct genapic { void (*send_IPI_allbutself)(int vector); void (*send_IPI_all)(int vector); #endif -}; +}; #define APICFUNC(x) .x = x, @@ -85,43 +77,46 @@ struct genapic { #define IPIFUNC(x) #endif -#define APIC_INIT(aname, aprobe) { \ - .name = aname, \ - .probe = aprobe, \ - .int_delivery_mode = INT_DELIVERY_MODE, \ - .int_dest_mode = INT_DEST_MODE, \ - .no_balance_irq = NO_BALANCE_IRQ, \ - .ESR_DISABLE = esr_disable, \ - .apic_destination_logical = APIC_DEST_LOGICAL, \ - APICFUNC(apic_id_registered) \ - APICFUNC(target_cpus) \ - APICFUNC(check_apicid_used) \ - APICFUNC(check_apicid_present) \ - APICFUNC(init_apic_ldr) \ - APICFUNC(ioapic_phys_id_map) \ - APICFUNC(setup_apic_routing) \ - APICFUNC(multi_timer_check) \ - APICFUNC(apicid_to_node) \ - APICFUNC(cpu_to_logical_apicid) \ - APICFUNC(cpu_present_to_apicid) \ - APICFUNC(apicid_to_cpu_present) \ - APICFUNC(mpc_apic_id) \ - APICFUNC(setup_portio_remap) \ - APICFUNC(check_phys_apicid_present) \ - APICFUNC(mpc_oem_bus_info) \ - APICFUNC(mpc_oem_pci_bus) \ - APICFUNC(mps_oem_check) \ - APICFUNC(get_apic_id) \ - .apic_id_mask = APIC_ID_MASK, \ - APICFUNC(cpu_mask_to_apicid) \ - APICFUNC(acpi_madt_oem_check) \ - IPIFUNC(send_IPI_mask) \ - IPIFUNC(send_IPI_allbutself) \ - IPIFUNC(send_IPI_all) \ - APICFUNC(enable_apic_mode) \ - APICFUNC(phys_pkg_id) \ - } +#define APIC_INIT(aname, aprobe) \ +{ \ + .name = aname, \ + .probe = aprobe, \ + .int_delivery_mode = INT_DELIVERY_MODE, \ + .int_dest_mode = INT_DEST_MODE, \ + .no_balance_irq = NO_BALANCE_IRQ, \ + .ESR_DISABLE = esr_disable, \ + .apic_destination_logical = APIC_DEST_LOGICAL, \ + APICFUNC(apic_id_registered) \ + APICFUNC(target_cpus) \ + APICFUNC(check_apicid_used) \ + APICFUNC(check_apicid_present) \ + APICFUNC(init_apic_ldr) \ + APICFUNC(ioapic_phys_id_map) \ + APICFUNC(setup_apic_routing) \ + APICFUNC(multi_timer_check) \ + APICFUNC(apicid_to_node) \ + APICFUNC(cpu_to_logical_apicid) \ + APICFUNC(cpu_present_to_apicid) \ + APICFUNC(apicid_to_cpu_present) \ + APICFUNC(setup_portio_remap) \ + APICFUNC(check_phys_apicid_present) \ + APICFUNC(mps_oem_check) \ + APICFUNC(get_apic_id) \ + .apic_id_mask = APIC_ID_MASK, \ + APICFUNC(cpu_mask_to_apicid) \ + APICFUNC(acpi_madt_oem_check) \ + IPIFUNC(send_IPI_mask) \ + IPIFUNC(send_IPI_allbutself) \ + IPIFUNC(send_IPI_all) \ + APICFUNC(enable_apic_mode) \ + APICFUNC(phys_pkg_id) \ +} extern struct genapic *genapic; +enum uv_system_type {UV_NONE, UV_LEGACY_APIC, UV_X2APIC, UV_NON_UNIQUE_APIC}; +#define get_uv_system_type() UV_NONE +#define is_uv_system() 0 + + #endif diff --git a/include/asm-x86/genapic_64.h b/include/asm-x86/genapic_64.h index d7e516ccbaa4..1de931b263ce 100644 --- a/include/asm-x86/genapic_64.h +++ b/include/asm-x86/genapic_64.h @@ -33,5 +33,15 @@ extern struct genapic *genapic; extern struct genapic apic_flat; extern struct genapic apic_physflat; +extern int acpi_madt_oem_check(char *, char *); + +enum uv_system_type {UV_NONE, UV_LEGACY_APIC, UV_X2APIC, UV_NON_UNIQUE_APIC}; +extern enum uv_system_type get_uv_system_type(void); +extern int is_uv_system(void); + +extern struct genapic apic_x2apic_uv_x; +DECLARE_PER_CPU(int, x2apic_extra_bits); +extern void uv_cpu_init(void); +extern int uv_wakeup_secondary(int phys_apicid, unsigned int start_rip); #endif diff --git a/include/asm-x86/geode.h b/include/asm-x86/geode.h index 9e7280092a48..9870cc1f2f8f 100644 --- a/include/asm-x86/geode.h +++ b/include/asm-x86/geode.h @@ -167,7 +167,7 @@ static inline int is_geode(void) /* MFGPTs */ #define MFGPT_MAX_TIMERS 8 -#define MFGPT_TIMER_ANY -1 +#define MFGPT_TIMER_ANY (-1) #define MFGPT_DOMAIN_WORKING 1 #define MFGPT_DOMAIN_STANDBY 2 diff --git a/include/asm-x86/highmem.h b/include/asm-x86/highmem.h index 479767c9195f..e153f3b44774 100644 --- a/include/asm-x86/highmem.h +++ b/include/asm-x86/highmem.h @@ -8,7 +8,7 @@ * Gerhard.Wichert@pdb.siemens.de * * - * Redesigned the x86 32-bit VM architecture to deal with + * Redesigned the x86 32-bit VM architecture to deal with * up to 16 Terabyte physical memory. With current x86 CPUs * we now support up to 64 Gigabytes physical RAM. * diff --git a/include/asm-x86/hw_irq_64.h b/include/asm-x86/hw_irq_64.h index 312a58d6dac6..0062ef390f67 100644 --- a/include/asm-x86/hw_irq_64.h +++ b/include/asm-x86/hw_irq_64.h @@ -36,7 +36,7 @@ * cleanup after irq migration. */ #define IRQ_MOVE_CLEANUP_VECTOR FIRST_EXTERNAL_VECTOR - + /* * Vectors 0x30-0x3f are used for ISA interrupts. */ @@ -159,13 +159,12 @@ extern atomic_t irq_mis_count; * SMP has a few special interrupts for IPI messages */ -#define BUILD_IRQ(nr) \ -asmlinkage void IRQ_NAME(nr); \ -__asm__( \ -"\n.p2align\n" \ -"IRQ" #nr "_interrupt:\n\t" \ - "push $~(" #nr ") ; " \ - "jmp common_interrupt"); +#define BUILD_IRQ(nr) \ + asmlinkage void IRQ_NAME(nr); \ + asm("\n.p2align\n" \ + "IRQ" #nr "_interrupt:\n\t" \ + "push $~(" #nr ") ; " \ + "jmp common_interrupt"); #define platform_legacy_irq(irq) ((irq) < 16) diff --git a/include/asm-x86/hypertransport.h b/include/asm-x86/hypertransport.h index c16c6ff4bdd7..d2bbd238b3e1 100644 --- a/include/asm-x86/hypertransport.h +++ b/include/asm-x86/hypertransport.h @@ -8,12 +8,14 @@ #define HT_IRQ_LOW_BASE 0xf8000000 #define HT_IRQ_LOW_VECTOR_SHIFT 16 -#define HT_IRQ_LOW_VECTOR_MASK 0x00ff0000 -#define HT_IRQ_LOW_VECTOR(v) (((v) << HT_IRQ_LOW_VECTOR_SHIFT) & HT_IRQ_LOW_VECTOR_MASK) +#define HT_IRQ_LOW_VECTOR_MASK 0x00ff0000 +#define HT_IRQ_LOW_VECTOR(v) \ + (((v) << HT_IRQ_LOW_VECTOR_SHIFT) & HT_IRQ_LOW_VECTOR_MASK) #define HT_IRQ_LOW_DEST_ID_SHIFT 8 -#define HT_IRQ_LOW_DEST_ID_MASK 0x0000ff00 -#define HT_IRQ_LOW_DEST_ID(v) (((v) << HT_IRQ_LOW_DEST_ID_SHIFT) & HT_IRQ_LOW_DEST_ID_MASK) +#define HT_IRQ_LOW_DEST_ID_MASK 0x0000ff00 +#define HT_IRQ_LOW_DEST_ID(v) \ + (((v) << HT_IRQ_LOW_DEST_ID_SHIFT) & HT_IRQ_LOW_DEST_ID_MASK) #define HT_IRQ_LOW_DM_PHYSICAL 0x0000000 #define HT_IRQ_LOW_DM_LOGICAL 0x0000040 @@ -36,7 +38,8 @@ #define HT_IRQ_HIGH_DEST_ID_SHIFT 0 -#define HT_IRQ_HIGH_DEST_ID_MASK 0x00ffffff -#define HT_IRQ_HIGH_DEST_ID(v) ((((v) >> 8) << HT_IRQ_HIGH_DEST_ID_SHIFT) & HT_IRQ_HIGH_DEST_ID_MASK) +#define HT_IRQ_HIGH_DEST_ID_MASK 0x00ffffff +#define HT_IRQ_HIGH_DEST_ID(v) \ + ((((v) >> 8) << HT_IRQ_HIGH_DEST_ID_SHIFT) & HT_IRQ_HIGH_DEST_ID_MASK) #endif /* ASM_HYPERTRANSPORT_H */ diff --git a/include/asm-x86/i387.h b/include/asm-x86/i387.h index f377b76b2f34..54522b814f1c 100644 --- a/include/asm-x86/i387.h +++ b/include/asm-x86/i387.h @@ -41,7 +41,7 @@ static inline void tolerant_fwait(void) { asm volatile("1: fwait\n" "2:\n" - _ASM_EXTABLE(1b,2b)); + _ASM_EXTABLE(1b, 2b)); } static inline int restore_fpu_checking(struct i387_fxsave_struct *fx) @@ -54,7 +54,7 @@ static inline int restore_fpu_checking(struct i387_fxsave_struct *fx) "3: movl $-1,%[err]\n" " jmp 2b\n" ".previous\n" - _ASM_EXTABLE(1b,3b) + _ASM_EXTABLE(1b, 3b) : [err] "=r" (err) #if 0 /* See comment in __save_init_fpu() below. */ : [fx] "r" (fx), "m" (*fx), "0" (0)); @@ -76,11 +76,11 @@ static inline int restore_fpu_checking(struct i387_fxsave_struct *fx) static inline void clear_fpu_state(struct i387_fxsave_struct *fx) { if (unlikely(fx->swd & X87_FSW_ES)) - asm volatile("fnclex"); + asm volatile("fnclex"); alternative_input(ASM_NOP8 ASM_NOP2, - " emms\n" /* clear stack tags */ - " fildl %%gs:0", /* load to clear state */ - X86_FEATURE_FXSAVE_LEAK); + " emms\n" /* clear stack tags */ + " fildl %%gs:0", /* load to clear state */ + X86_FEATURE_FXSAVE_LEAK); } static inline int save_i387_checking(struct i387_fxsave_struct __user *fx) @@ -93,14 +93,15 @@ static inline int save_i387_checking(struct i387_fxsave_struct __user *fx) "3: movl $-1,%[err]\n" " jmp 2b\n" ".previous\n" - _ASM_EXTABLE(1b,3b) + _ASM_EXTABLE(1b, 3b) : [err] "=r" (err), "=m" (*fx) #if 0 /* See comment in __fxsave_clear() below. */ : [fx] "r" (fx), "0" (0)); #else : [fx] "cdaSDb" (fx), "0" (0)); #endif - if (unlikely(err) && __clear_user(fx, sizeof(struct i387_fxsave_struct))) + if (unlikely(err) && + __clear_user(fx, sizeof(struct i387_fxsave_struct))) err = -EFAULT; /* No need to clear here because the caller clears USED_MATH */ return err; @@ -156,8 +157,10 @@ static inline int save_i387(struct _fpstate __user *buf) return 0; clear_used_math(); /* trigger finit */ if (task_thread_info(tsk)->status & TS_USEDFPU) { - err = save_i387_checking((struct i387_fxsave_struct __user *)buf); - if (err) return err; + err = save_i387_checking((struct i387_fxsave_struct __user *) + buf); + if (err) + return err; task_thread_info(tsk)->status &= ~TS_USEDFPU; stts(); } else { diff --git a/include/asm-x86/i8259.h b/include/asm-x86/i8259.h index 67c319e0efc7..45d4df3e51e6 100644 --- a/include/asm-x86/i8259.h +++ b/include/asm-x86/i8259.h @@ -1,9 +1,11 @@ #ifndef __ASM_I8259_H__ #define __ASM_I8259_H__ +#include <linux/delay.h> + extern unsigned int cached_irq_mask; -#define __byte(x,y) (((unsigned char *) &(y))[x]) +#define __byte(x, y) (((unsigned char *)&(y))[x]) #define cached_master_mask (__byte(0, cached_irq_mask)) #define cached_slave_mask (__byte(1, cached_irq_mask)) @@ -29,7 +31,28 @@ extern void enable_8259A_irq(unsigned int irq); extern void disable_8259A_irq(unsigned int irq); extern unsigned int startup_8259A_irq(unsigned int irq); -#define inb_pic inb_p -#define outb_pic outb_p +/* the PIC may need a careful delay on some platforms, hence specific calls */ +static inline unsigned char inb_pic(unsigned int port) +{ + unsigned char value = inb(port); + + /* + * delay for some accesses to PIC on motherboard or in chipset + * must be at least one microsecond, so be safe here: + */ + udelay(2); + + return value; +} + +static inline void outb_pic(unsigned char value, unsigned int port) +{ + outb(value, port); + /* + * delay for some accesses to PIC on motherboard or in chipset + * must be at least one microsecond, so be safe here: + */ + udelay(2); +} #endif /* __ASM_I8259_H__ */ diff --git a/include/asm-x86/ia32.h b/include/asm-x86/ia32.h index aa9733206e29..55d3abe5276f 100644 --- a/include/asm-x86/ia32.h +++ b/include/asm-x86/ia32.h @@ -14,19 +14,19 @@ /* signal.h */ struct sigaction32 { - unsigned int sa_handler; /* Really a pointer, but need to deal - with 32 bits */ - unsigned int sa_flags; - unsigned int sa_restorer; /* Another 32 bit pointer */ - compat_sigset_t sa_mask; /* A 32 bit mask */ + unsigned int sa_handler; /* Really a pointer, but need to deal + with 32 bits */ + unsigned int sa_flags; + unsigned int sa_restorer; /* Another 32 bit pointer */ + compat_sigset_t sa_mask; /* A 32 bit mask */ }; struct old_sigaction32 { - unsigned int sa_handler; /* Really a pointer, but need to deal - with 32 bits */ - compat_old_sigset_t sa_mask; /* A 32 bit mask */ - unsigned int sa_flags; - unsigned int sa_restorer; /* Another 32 bit pointer */ + unsigned int sa_handler; /* Really a pointer, but need to deal + with 32 bits */ + compat_old_sigset_t sa_mask; /* A 32 bit mask */ + unsigned int sa_flags; + unsigned int sa_restorer; /* Another 32 bit pointer */ }; typedef struct sigaltstack_ia32 { @@ -65,7 +65,7 @@ struct stat64 { long long st_size; unsigned int st_blksize; - long long st_blocks;/* Number 512-byte blocks allocated. */ + long long st_blocks;/* Number 512-byte blocks allocated */ unsigned st_atime; unsigned st_atime_nsec; @@ -77,13 +77,13 @@ struct stat64 { unsigned long long st_ino; } __attribute__((packed)); -typedef struct compat_siginfo{ +typedef struct compat_siginfo { int si_signo; int si_errno; int si_code; union { - int _pad[((128/sizeof(int)) - 3)]; + int _pad[((128 / sizeof(int)) - 3)]; /* kill() */ struct { @@ -129,28 +129,26 @@ typedef struct compat_siginfo{ } _sifields; } compat_siginfo_t; -struct sigframe32 -{ - u32 pretcode; - int sig; - struct sigcontext_ia32 sc; - struct _fpstate_ia32 fpstate; - unsigned int extramask[_COMPAT_NSIG_WORDS-1]; +struct sigframe32 { + u32 pretcode; + int sig; + struct sigcontext_ia32 sc; + struct _fpstate_ia32 fpstate; + unsigned int extramask[_COMPAT_NSIG_WORDS-1]; }; -struct rt_sigframe32 -{ - u32 pretcode; - int sig; - u32 pinfo; - u32 puc; - compat_siginfo_t info; - struct ucontext_ia32 uc; - struct _fpstate_ia32 fpstate; +struct rt_sigframe32 { + u32 pretcode; + int sig; + u32 pinfo; + u32 puc; + compat_siginfo_t info; + struct ucontext_ia32 uc; + struct _fpstate_ia32 fpstate; }; struct ustat32 { - __u32 f_tfree; + __u32 f_tfree; compat_ino_t f_tinode; char f_fname[6]; char f_fpack[6]; @@ -168,5 +166,5 @@ extern void ia32_pick_mmap_layout(struct mm_struct *mm); #endif #endif /* !CONFIG_IA32_SUPPORT */ - -#endif + +#endif diff --git a/include/asm-x86/io.h b/include/asm-x86/io.h index 5a58b176dd61..7b292d386713 100644 --- a/include/asm-x86/io.h +++ b/include/asm-x86/io.h @@ -1,5 +1,11 @@ +#define ARCH_HAS_IOREMAP_WC + #ifdef CONFIG_X86_32 # include "io_32.h" #else # include "io_64.h" #endif +extern int ioremap_change_attr(unsigned long vaddr, unsigned long size, + unsigned long prot_val); +extern void __iomem *ioremap_wc(unsigned long offset, unsigned long size); + diff --git a/include/asm-x86/io_32.h b/include/asm-x86/io_32.h index d4d8fbd9378c..509045f5fda2 100644 --- a/include/asm-x86/io_32.h +++ b/include/asm-x86/io_32.h @@ -65,14 +65,14 @@ * * The returned physical address is the physical (CPU) mapping for * the memory address given. It is only valid to use this function on - * addresses directly mapped or allocated via kmalloc. + * addresses directly mapped or allocated via kmalloc. * * This function does not give bus mappings for DMA transfers. In * almost all conceivable cases a device driver should not be using * this function */ - -static inline unsigned long virt_to_phys(volatile void * address) + +static inline unsigned long virt_to_phys(volatile void *address) { return __pa(address); } @@ -90,7 +90,7 @@ static inline unsigned long virt_to_phys(volatile void * address) * this function */ -static inline void * phys_to_virt(unsigned long address) +static inline void *phys_to_virt(unsigned long address) { return __va(address); } @@ -169,16 +169,19 @@ extern void __iomem *fix_ioremap(unsigned idx, unsigned long phys); static inline unsigned char readb(const volatile void __iomem *addr) { - return *(volatile unsigned char __force *) addr; + return *(volatile unsigned char __force *)addr; } + static inline unsigned short readw(const volatile void __iomem *addr) { - return *(volatile unsigned short __force *) addr; + return *(volatile unsigned short __force *)addr; } + static inline unsigned int readl(const volatile void __iomem *addr) { return *(volatile unsigned int __force *) addr; } + #define readb_relaxed(addr) readb(addr) #define readw_relaxed(addr) readw(addr) #define readl_relaxed(addr) readl(addr) @@ -188,15 +191,17 @@ static inline unsigned int readl(const volatile void __iomem *addr) static inline void writeb(unsigned char b, volatile void __iomem *addr) { - *(volatile unsigned char __force *) addr = b; + *(volatile unsigned char __force *)addr = b; } + static inline void writew(unsigned short b, volatile void __iomem *addr) { - *(volatile unsigned short __force *) addr = b; + *(volatile unsigned short __force *)addr = b; } + static inline void writel(unsigned int b, volatile void __iomem *addr) { - *(volatile unsigned int __force *) addr = b; + *(volatile unsigned int __force *)addr = b; } #define __raw_writeb writeb #define __raw_writew writew @@ -239,12 +244,12 @@ memcpy_toio(volatile void __iomem *dst, const void *src, int count) * 1. Out of order aware processors * 2. Accidentally out of order processors (PPro errata #51) */ - + #if defined(CONFIG_X86_OOSTORE) || defined(CONFIG_X86_PPRO_FENCE) static inline void flush_write_buffers(void) { - __asm__ __volatile__ ("lock; addl $0,0(%%esp)": : :"memory"); + asm volatile("lock; addl $0,0(%%esp)": : :"memory"); } #else @@ -264,7 +269,8 @@ extern void io_delay_init(void); #include <asm/paravirt.h> #else -static inline void slow_down_io(void) { +static inline void slow_down_io(void) +{ native_io_delay(); #ifdef REALLY_SLOW_IO native_io_delay(); @@ -275,51 +281,74 @@ static inline void slow_down_io(void) { #endif -#define __BUILDIO(bwl,bw,type) \ -static inline void out##bwl(unsigned type value, int port) { \ - out##bwl##_local(value, port); \ -} \ -static inline unsigned type in##bwl(int port) { \ - return in##bwl##_local(port); \ +#define __BUILDIO(bwl, bw, type) \ +static inline void out##bwl(unsigned type value, int port) \ +{ \ + out##bwl##_local(value, port); \ +} \ + \ +static inline unsigned type in##bwl(int port) \ +{ \ + return in##bwl##_local(port); \ } -#define BUILDIO(bwl,bw,type) \ -static inline void out##bwl##_local(unsigned type value, int port) { \ - __asm__ __volatile__("out" #bwl " %" #bw "0, %w1" : : "a"(value), "Nd"(port)); \ -} \ -static inline unsigned type in##bwl##_local(int port) { \ - unsigned type value; \ - __asm__ __volatile__("in" #bwl " %w1, %" #bw "0" : "=a"(value) : "Nd"(port)); \ - return value; \ -} \ -static inline void out##bwl##_local_p(unsigned type value, int port) { \ - out##bwl##_local(value, port); \ - slow_down_io(); \ -} \ -static inline unsigned type in##bwl##_local_p(int port) { \ - unsigned type value = in##bwl##_local(port); \ - slow_down_io(); \ - return value; \ -} \ -__BUILDIO(bwl,bw,type) \ -static inline void out##bwl##_p(unsigned type value, int port) { \ - out##bwl(value, port); \ - slow_down_io(); \ -} \ -static inline unsigned type in##bwl##_p(int port) { \ - unsigned type value = in##bwl(port); \ - slow_down_io(); \ - return value; \ -} \ -static inline void outs##bwl(int port, const void *addr, unsigned long count) { \ - __asm__ __volatile__("rep; outs" #bwl : "+S"(addr), "+c"(count) : "d"(port)); \ -} \ -static inline void ins##bwl(int port, void *addr, unsigned long count) { \ - __asm__ __volatile__("rep; ins" #bwl : "+D"(addr), "+c"(count) : "d"(port)); \ +#define BUILDIO(bwl, bw, type) \ +static inline void out##bwl##_local(unsigned type value, int port) \ +{ \ + asm volatile("out" #bwl " %" #bw "0, %w1" \ + : : "a"(value), "Nd"(port)); \ +} \ + \ +static inline unsigned type in##bwl##_local(int port) \ +{ \ + unsigned type value; \ + asm volatile("in" #bwl " %w1, %" #bw "0" \ + : "=a"(value) : "Nd"(port)); \ + return value; \ +} \ + \ +static inline void out##bwl##_local_p(unsigned type value, int port) \ +{ \ + out##bwl##_local(value, port); \ + slow_down_io(); \ +} \ + \ +static inline unsigned type in##bwl##_local_p(int port) \ +{ \ + unsigned type value = in##bwl##_local(port); \ + slow_down_io(); \ + return value; \ +} \ + \ +__BUILDIO(bwl, bw, type) \ + \ +static inline void out##bwl##_p(unsigned type value, int port) \ +{ \ + out##bwl(value, port); \ + slow_down_io(); \ +} \ + \ +static inline unsigned type in##bwl##_p(int port) \ +{ \ + unsigned type value = in##bwl(port); \ + slow_down_io(); \ + return value; \ +} \ + \ +static inline void outs##bwl(int port, const void *addr, unsigned long count) \ +{ \ + asm volatile("rep; outs" #bwl \ + : "+S"(addr), "+c"(count) : "d"(port)); \ +} \ + \ +static inline void ins##bwl(int port, void *addr, unsigned long count) \ +{ \ + asm volatile("rep; ins" #bwl \ + : "+D"(addr), "+c"(count) : "d"(port)); \ } -BUILDIO(b,b,char) -BUILDIO(w,w,short) -BUILDIO(l,,int) +BUILDIO(b, b, char) +BUILDIO(w, w, short) +BUILDIO(l, , int) #endif diff --git a/include/asm-x86/io_64.h b/include/asm-x86/io_64.h index db0be2011a3c..c2f5eef47b88 100644 --- a/include/asm-x86/io_64.h +++ b/include/asm-x86/io_64.h @@ -58,60 +58,75 @@ static inline void slow_down_io(void) /* * Talk about misusing macros.. */ -#define __OUT1(s,x) \ +#define __OUT1(s, x) \ static inline void out##s(unsigned x value, unsigned short port) { -#define __OUT2(s,s1,s2) \ -__asm__ __volatile__ ("out" #s " %" s1 "0,%" s2 "1" +#define __OUT2(s, s1, s2) \ +asm volatile ("out" #s " %" s1 "0,%" s2 "1" #ifndef REALLY_SLOW_IO #define REALLY_SLOW_IO #define UNSET_REALLY_SLOW_IO #endif -#define __OUT(s,s1,x) \ -__OUT1(s,x) __OUT2(s,s1,"w") : : "a" (value), "Nd" (port)); } \ -__OUT1(s##_p, x) __OUT2(s, s1, "w") : : "a" (value), "Nd" (port)); \ - slow_down_io(); } +#define __OUT(s, s1, x) \ + __OUT1(s, x) __OUT2(s, s1, "w") : : "a" (value), "Nd" (port)); \ + } \ + __OUT1(s##_p, x) __OUT2(s, s1, "w") : : "a" (value), "Nd" (port)); \ + slow_down_io(); \ +} -#define __IN1(s) \ -static inline RETURN_TYPE in##s(unsigned short port) { RETURN_TYPE _v; +#define __IN1(s) \ +static inline RETURN_TYPE in##s(unsigned short port) \ +{ \ + RETURN_TYPE _v; -#define __IN2(s,s1,s2) \ -__asm__ __volatile__ ("in" #s " %" s2 "1,%" s1 "0" +#define __IN2(s, s1, s2) \ + asm volatile ("in" #s " %" s2 "1,%" s1 "0" -#define __IN(s,s1,i...) \ -__IN1(s) __IN2(s, s1, "w") : "=a" (_v) : "Nd" (port), ##i); return _v; } \ -__IN1(s##_p) __IN2(s, s1, "w") : "=a" (_v) : "Nd" (port), ##i); \ - slow_down_io(); return _v; } +#define __IN(s, s1, i...) \ + __IN1(s) __IN2(s, s1, "w") : "=a" (_v) : "Nd" (port), ##i); \ + return _v; \ + } \ + __IN1(s##_p) __IN2(s, s1, "w") : "=a" (_v) : "Nd" (port), ##i); \ + slow_down_io(); \ + return _v; } #ifdef UNSET_REALLY_SLOW_IO #undef REALLY_SLOW_IO #endif -#define __INS(s) \ -static inline void ins##s(unsigned short port, void * addr, unsigned long count) \ -{ __asm__ __volatile__ ("rep ; ins" #s \ -: "=D" (addr), "=c" (count) : "d" (port),"0" (addr),"1" (count)); } +#define __INS(s) \ +static inline void ins##s(unsigned short port, void *addr, \ + unsigned long count) \ +{ \ + asm volatile ("rep ; ins" #s \ + : "=D" (addr), "=c" (count) \ + : "d" (port), "0" (addr), "1" (count)); \ +} -#define __OUTS(s) \ -static inline void outs##s(unsigned short port, const void * addr, unsigned long count) \ -{ __asm__ __volatile__ ("rep ; outs" #s \ -: "=S" (addr), "=c" (count) : "d" (port),"0" (addr),"1" (count)); } +#define __OUTS(s) \ +static inline void outs##s(unsigned short port, const void *addr, \ + unsigned long count) \ +{ \ + asm volatile ("rep ; outs" #s \ + : "=S" (addr), "=c" (count) \ + : "d" (port), "0" (addr), "1" (count)); \ +} #define RETURN_TYPE unsigned char -__IN(b,"") +__IN(b, "") #undef RETURN_TYPE #define RETURN_TYPE unsigned short -__IN(w,"") +__IN(w, "") #undef RETURN_TYPE #define RETURN_TYPE unsigned int -__IN(l,"") +__IN(l, "") #undef RETURN_TYPE -__OUT(b,"b",char) -__OUT(w,"w",short) -__OUT(l,,int) +__OUT(b, "b", char) +__OUT(w, "w", short) +__OUT(l, , int) __INS(b) __INS(w) @@ -132,12 +147,12 @@ __OUTS(l) * Change virtual addresses to physical addresses and vv. * These are pretty trivial */ -static inline unsigned long virt_to_phys(volatile void * address) +static inline unsigned long virt_to_phys(volatile void *address) { return __pa(address); } -static inline void * phys_to_virt(unsigned long address) +static inline void *phys_to_virt(unsigned long address) { return __va(address); } @@ -200,18 +215,22 @@ static inline __u8 __readb(const volatile void __iomem *addr) { return *(__force volatile __u8 *)addr; } + static inline __u16 __readw(const volatile void __iomem *addr) { return *(__force volatile __u16 *)addr; } + static __always_inline __u32 __readl(const volatile void __iomem *addr) { return *(__force volatile __u32 *)addr; } + static inline __u64 __readq(const volatile void __iomem *addr) { return *(__force volatile __u64 *)addr; } + #define readb(x) __readb(x) #define readw(x) __readw(x) #define readl(x) __readl(x) @@ -231,37 +250,44 @@ static inline void __writel(__u32 b, volatile void __iomem *addr) { *(__force volatile __u32 *)addr = b; } + static inline void __writeq(__u64 b, volatile void __iomem *addr) { *(__force volatile __u64 *)addr = b; } + static inline void __writeb(__u8 b, volatile void __iomem *addr) { *(__force volatile __u8 *)addr = b; } + static inline void __writew(__u16 b, volatile void __iomem *addr) { *(__force volatile __u16 *)addr = b; } -#define writeq(val,addr) __writeq((val),(addr)) -#define writel(val,addr) __writel((val),(addr)) -#define writew(val,addr) __writew((val),(addr)) -#define writeb(val,addr) __writeb((val),(addr)) + +#define writeq(val, addr) __writeq((val), (addr)) +#define writel(val, addr) __writel((val), (addr)) +#define writew(val, addr) __writew((val), (addr)) +#define writeb(val, addr) __writeb((val), (addr)) #define __raw_writeb writeb #define __raw_writew writew #define __raw_writel writel #define __raw_writeq writeq -void __memcpy_fromio(void*,unsigned long,unsigned); -void __memcpy_toio(unsigned long,const void*,unsigned); +void __memcpy_fromio(void *, unsigned long, unsigned); +void __memcpy_toio(unsigned long, const void *, unsigned); -static inline void memcpy_fromio(void *to, const volatile void __iomem *from, unsigned len) +static inline void memcpy_fromio(void *to, const volatile void __iomem *from, + unsigned len) { - __memcpy_fromio(to,(unsigned long)from,len); + __memcpy_fromio(to, (unsigned long)from, len); } -static inline void memcpy_toio(volatile void __iomem *to, const void *from, unsigned len) + +static inline void memcpy_toio(volatile void __iomem *to, const void *from, + unsigned len) { - __memcpy_toio((unsigned long)to,from,len); + __memcpy_toio((unsigned long)to, from, len); } void memset_io(volatile void __iomem *a, int b, size_t c); @@ -276,7 +302,7 @@ void memset_io(volatile void __iomem *a, int b, size_t c); */ #define __ISA_IO_base ((char __iomem *)(PAGE_OFFSET)) -#define flush_write_buffers() +#define flush_write_buffers() extern int iommu_bio_merge; #define BIO_VMERGE_BOUNDARY iommu_bio_merge diff --git a/include/asm-x86/io_apic.h b/include/asm-x86/io_apic.h index 0f5b3fef0b08..0c9e17c73e05 100644 --- a/include/asm-x86/io_apic.h +++ b/include/asm-x86/io_apic.h @@ -110,6 +110,13 @@ extern int nr_ioapic_registers[MAX_IO_APICS]; * MP-BIOS irq configuration table structures: */ +struct mp_ioapic_routing { + int apic_id; + int gsi_base; + int gsi_end; + u32 pin_programmed[4]; +}; + /* I/O APIC entries */ extern struct mpc_config_ioapic mp_ioapics[MAX_IO_APICS]; @@ -146,7 +153,6 @@ extern int io_apic_get_version(int ioapic); extern int io_apic_get_redir_entries(int ioapic); extern int io_apic_set_pci_routing(int ioapic, int pin, int irq, int edge_level, int active_high_low); -extern int timer_uses_ioapic_pin_0; #endif /* CONFIG_ACPI */ extern int (*ioapic_renumber_irq)(int ioapic, int irq); diff --git a/include/asm-x86/ioctls.h b/include/asm-x86/ioctls.h index 93c894dc5154..c0c338bd4068 100644 --- a/include/asm-x86/ioctls.h +++ b/include/asm-x86/ioctls.h @@ -47,12 +47,13 @@ #define TIOCSBRK 0x5427 /* BSD compatibility */ #define TIOCCBRK 0x5428 /* BSD compatibility */ #define TIOCGSID 0x5429 /* Return the session ID of FD */ -#define TCGETS2 _IOR('T',0x2A, struct termios2) -#define TCSETS2 _IOW('T',0x2B, struct termios2) -#define TCSETSW2 _IOW('T',0x2C, struct termios2) -#define TCSETSF2 _IOW('T',0x2D, struct termios2) -#define TIOCGPTN _IOR('T',0x30, unsigned int) /* Get Pty Number (of pty-mux device) */ -#define TIOCSPTLCK _IOW('T',0x31, int) /* Lock/unlock Pty */ +#define TCGETS2 _IOR('T', 0x2A, struct termios2) +#define TCSETS2 _IOW('T', 0x2B, struct termios2) +#define TCSETSW2 _IOW('T', 0x2C, struct termios2) +#define TCSETSF2 _IOW('T', 0x2D, struct termios2) +#define TIOCGPTN _IOR('T', 0x30, unsigned int) + /* Get Pty Number (of pty-mux device) */ +#define TIOCSPTLCK _IOW('T', 0x31, int) /* Lock/unlock Pty */ #define FIONCLEX 0x5450 #define FIOCLEX 0x5451 diff --git a/include/asm-x86/ipcbuf.h b/include/asm-x86/ipcbuf.h index 2adf8b39a40b..ee678fd51594 100644 --- a/include/asm-x86/ipcbuf.h +++ b/include/asm-x86/ipcbuf.h @@ -11,8 +11,7 @@ * - 2 miscellaneous 32-bit values */ -struct ipc64_perm -{ +struct ipc64_perm { __kernel_key_t key; __kernel_uid32_t uid; __kernel_gid32_t gid; diff --git a/include/asm-x86/ipi.h b/include/asm-x86/ipi.h index 6d011bd6067d..ecc80f341f37 100644 --- a/include/asm-x86/ipi.h +++ b/include/asm-x86/ipi.h @@ -27,7 +27,8 @@ * We use 'broadcast', CPU->CPU IPIs and self-IPIs too. */ -static inline unsigned int __prepare_ICR (unsigned int shortcut, int vector, unsigned int dest) +static inline unsigned int __prepare_ICR(unsigned int shortcut, int vector, + unsigned int dest) { unsigned int icr = shortcut | dest; @@ -42,12 +43,13 @@ static inline unsigned int __prepare_ICR (unsigned int shortcut, int vector, uns return icr; } -static inline int __prepare_ICR2 (unsigned int mask) +static inline int __prepare_ICR2(unsigned int mask) { return SET_APIC_DEST_FIELD(mask); } -static inline void __send_IPI_shortcut(unsigned int shortcut, int vector, unsigned int dest) +static inline void __send_IPI_shortcut(unsigned int shortcut, int vector, + unsigned int dest) { /* * Subtle. In the case of the 'never do double writes' workaround @@ -78,7 +80,8 @@ static inline void __send_IPI_shortcut(unsigned int shortcut, int vector, unsign * This is used to send an IPI with no shorthand notation (the destination is * specified in bits 56 to 63 of the ICR). */ -static inline void __send_IPI_dest_field(unsigned int mask, int vector, unsigned int dest) +static inline void __send_IPI_dest_field(unsigned int mask, int vector, + unsigned int dest) { unsigned long cfg; diff --git a/include/asm-x86/irq_32.h b/include/asm-x86/irq_32.h index aca9c96e8e6b..0b79f3185243 100644 --- a/include/asm-x86/irq_32.h +++ b/include/asm-x86/irq_32.h @@ -15,7 +15,7 @@ #include "irq_vectors.h" #include <asm/thread_info.h> -static __inline__ int irq_canonicalize(int irq) +static inline int irq_canonicalize(int irq) { return ((irq == 2) ? 9 : irq); } diff --git a/include/asm-x86/irq_64.h b/include/asm-x86/irq_64.h index 5006c6e75656..083d35a62c94 100644 --- a/include/asm-x86/irq_64.h +++ b/include/asm-x86/irq_64.h @@ -31,10 +31,10 @@ #define FIRST_SYSTEM_VECTOR 0xef /* duplicated in hw_irq.h */ -#define NR_IRQS (NR_VECTORS + (32 *NR_CPUS)) +#define NR_IRQS (NR_VECTORS + (32 * NR_CPUS)) #define NR_IRQ_VECTORS NR_IRQS -static __inline__ int irq_canonicalize(int irq) +static inline int irq_canonicalize(int irq) { return ((irq == 2) ? 9 : irq); } diff --git a/include/asm-x86/irqflags.h b/include/asm-x86/irqflags.h index 0e2292483b35..c242527f970e 100644 --- a/include/asm-x86/irqflags.h +++ b/include/asm-x86/irqflags.h @@ -12,25 +12,21 @@ static inline unsigned long native_save_fl(void) { unsigned long flags; - __asm__ __volatile__( - "# __raw_save_flags\n\t" - "pushf ; pop %0" - : "=g" (flags) - : /* no input */ - : "memory" - ); + asm volatile("# __raw_save_flags\n\t" + "pushf ; pop %0" + : "=g" (flags) + : /* no input */ + : "memory"); return flags; } static inline void native_restore_fl(unsigned long flags) { - __asm__ __volatile__( - "push %0 ; popf" - : /* no output */ - :"g" (flags) - :"memory", "cc" - ); + asm volatile("push %0 ; popf" + : /* no output */ + :"g" (flags) + :"memory", "cc"); } static inline void native_irq_disable(void) @@ -70,26 +66,6 @@ static inline void raw_local_irq_restore(unsigned long flags) native_restore_fl(flags); } -#ifdef CONFIG_X86_VSMP - -/* - * Interrupt control for the VSMP architecture: - */ - -static inline void raw_local_irq_disable(void) -{ - unsigned long flags = __raw_local_save_flags(); - raw_local_irq_restore((flags & ~X86_EFLAGS_IF) | X86_EFLAGS_AC); -} - -static inline void raw_local_irq_enable(void) -{ - unsigned long flags = __raw_local_save_flags(); - raw_local_irq_restore((flags | X86_EFLAGS_IF) & (~X86_EFLAGS_AC)); -} - -#else - static inline void raw_local_irq_disable(void) { native_irq_disable(); @@ -100,8 +76,6 @@ static inline void raw_local_irq_enable(void) native_irq_enable(); } -#endif - /* * Used in the idle loop; sti takes one instruction cycle * to complete: @@ -153,23 +127,16 @@ static inline unsigned long __raw_local_irq_save(void) #endif /* CONFIG_PARAVIRT */ #ifndef __ASSEMBLY__ -#define raw_local_save_flags(flags) \ - do { (flags) = __raw_local_save_flags(); } while (0) +#define raw_local_save_flags(flags) \ + do { (flags) = __raw_local_save_flags(); } while (0) -#define raw_local_irq_save(flags) \ - do { (flags) = __raw_local_irq_save(); } while (0) +#define raw_local_irq_save(flags) \ + do { (flags) = __raw_local_irq_save(); } while (0) -#ifdef CONFIG_X86_VSMP -static inline int raw_irqs_disabled_flags(unsigned long flags) -{ - return !(flags & X86_EFLAGS_IF) || (flags & X86_EFLAGS_AC); -} -#else static inline int raw_irqs_disabled_flags(unsigned long flags) { return !(flags & X86_EFLAGS_IF); } -#endif static inline int raw_irqs_disabled(void) { diff --git a/include/asm-x86/kdebug.h b/include/asm-x86/kdebug.h index 35231240ebdc..96651bb59ba1 100644 --- a/include/asm-x86/kdebug.h +++ b/include/asm-x86/kdebug.h @@ -24,12 +24,12 @@ enum die_val { }; extern void printk_address(unsigned long address, int reliable); -extern void die(const char *,struct pt_regs *,long); +extern void die(const char *, struct pt_regs *,long); extern int __must_check __die(const char *, struct pt_regs *, long); extern void show_registers(struct pt_regs *regs); extern void __show_registers(struct pt_regs *, int all); extern void show_trace(struct task_struct *t, struct pt_regs *regs, - unsigned long *sp, unsigned long bp); + unsigned long *sp, unsigned long bp); extern void __show_regs(struct pt_regs *regs); extern void show_regs(struct pt_regs *regs); extern unsigned long oops_begin(void); diff --git a/include/asm-x86/kexec.h b/include/asm-x86/kexec.h index c90d3c77afc2..8f855a15f64d 100644 --- a/include/asm-x86/kexec.h +++ b/include/asm-x86/kexec.h @@ -94,10 +94,9 @@ static inline void crash_fixup_ss_esp(struct pt_regs *newregs, { #ifdef CONFIG_X86_32 newregs->sp = (unsigned long)&(oldregs->sp); - __asm__ __volatile__( - "xorl %%eax, %%eax\n\t" - "movw %%ss, %%ax\n\t" - :"=a"(newregs->ss)); + asm volatile("xorl %%eax, %%eax\n\t" + "movw %%ss, %%ax\n\t" + :"=a"(newregs->ss)); #endif } @@ -114,39 +113,39 @@ static inline void crash_setup_regs(struct pt_regs *newregs, crash_fixup_ss_esp(newregs, oldregs); } else { #ifdef CONFIG_X86_32 - __asm__ __volatile__("movl %%ebx,%0" : "=m"(newregs->bx)); - __asm__ __volatile__("movl %%ecx,%0" : "=m"(newregs->cx)); - __asm__ __volatile__("movl %%edx,%0" : "=m"(newregs->dx)); - __asm__ __volatile__("movl %%esi,%0" : "=m"(newregs->si)); - __asm__ __volatile__("movl %%edi,%0" : "=m"(newregs->di)); - __asm__ __volatile__("movl %%ebp,%0" : "=m"(newregs->bp)); - __asm__ __volatile__("movl %%eax,%0" : "=m"(newregs->ax)); - __asm__ __volatile__("movl %%esp,%0" : "=m"(newregs->sp)); - __asm__ __volatile__("movl %%ss, %%eax;" :"=a"(newregs->ss)); - __asm__ __volatile__("movl %%cs, %%eax;" :"=a"(newregs->cs)); - __asm__ __volatile__("movl %%ds, %%eax;" :"=a"(newregs->ds)); - __asm__ __volatile__("movl %%es, %%eax;" :"=a"(newregs->es)); - __asm__ __volatile__("pushfl; popl %0" :"=m"(newregs->flags)); + asm volatile("movl %%ebx,%0" : "=m"(newregs->bx)); + asm volatile("movl %%ecx,%0" : "=m"(newregs->cx)); + asm volatile("movl %%edx,%0" : "=m"(newregs->dx)); + asm volatile("movl %%esi,%0" : "=m"(newregs->si)); + asm volatile("movl %%edi,%0" : "=m"(newregs->di)); + asm volatile("movl %%ebp,%0" : "=m"(newregs->bp)); + asm volatile("movl %%eax,%0" : "=m"(newregs->ax)); + asm volatile("movl %%esp,%0" : "=m"(newregs->sp)); + asm volatile("movl %%ss, %%eax;" :"=a"(newregs->ss)); + asm volatile("movl %%cs, %%eax;" :"=a"(newregs->cs)); + asm volatile("movl %%ds, %%eax;" :"=a"(newregs->ds)); + asm volatile("movl %%es, %%eax;" :"=a"(newregs->es)); + asm volatile("pushfl; popl %0" :"=m"(newregs->flags)); #else - __asm__ __volatile__("movq %%rbx,%0" : "=m"(newregs->bx)); - __asm__ __volatile__("movq %%rcx,%0" : "=m"(newregs->cx)); - __asm__ __volatile__("movq %%rdx,%0" : "=m"(newregs->dx)); - __asm__ __volatile__("movq %%rsi,%0" : "=m"(newregs->si)); - __asm__ __volatile__("movq %%rdi,%0" : "=m"(newregs->di)); - __asm__ __volatile__("movq %%rbp,%0" : "=m"(newregs->bp)); - __asm__ __volatile__("movq %%rax,%0" : "=m"(newregs->ax)); - __asm__ __volatile__("movq %%rsp,%0" : "=m"(newregs->sp)); - __asm__ __volatile__("movq %%r8,%0" : "=m"(newregs->r8)); - __asm__ __volatile__("movq %%r9,%0" : "=m"(newregs->r9)); - __asm__ __volatile__("movq %%r10,%0" : "=m"(newregs->r10)); - __asm__ __volatile__("movq %%r11,%0" : "=m"(newregs->r11)); - __asm__ __volatile__("movq %%r12,%0" : "=m"(newregs->r12)); - __asm__ __volatile__("movq %%r13,%0" : "=m"(newregs->r13)); - __asm__ __volatile__("movq %%r14,%0" : "=m"(newregs->r14)); - __asm__ __volatile__("movq %%r15,%0" : "=m"(newregs->r15)); - __asm__ __volatile__("movl %%ss, %%eax;" :"=a"(newregs->ss)); - __asm__ __volatile__("movl %%cs, %%eax;" :"=a"(newregs->cs)); - __asm__ __volatile__("pushfq; popq %0" :"=m"(newregs->flags)); + asm volatile("movq %%rbx,%0" : "=m"(newregs->bx)); + asm volatile("movq %%rcx,%0" : "=m"(newregs->cx)); + asm volatile("movq %%rdx,%0" : "=m"(newregs->dx)); + asm volatile("movq %%rsi,%0" : "=m"(newregs->si)); + asm volatile("movq %%rdi,%0" : "=m"(newregs->di)); + asm volatile("movq %%rbp,%0" : "=m"(newregs->bp)); + asm volatile("movq %%rax,%0" : "=m"(newregs->ax)); + asm volatile("movq %%rsp,%0" : "=m"(newregs->sp)); + asm volatile("movq %%r8,%0" : "=m"(newregs->r8)); + asm volatile("movq %%r9,%0" : "=m"(newregs->r9)); + asm volatile("movq %%r10,%0" : "=m"(newregs->r10)); + asm volatile("movq %%r11,%0" : "=m"(newregs->r11)); + asm volatile("movq %%r12,%0" : "=m"(newregs->r12)); + asm volatile("movq %%r13,%0" : "=m"(newregs->r13)); + asm volatile("movq %%r14,%0" : "=m"(newregs->r14)); + asm volatile("movq %%r15,%0" : "=m"(newregs->r15)); + asm volatile("movl %%ss, %%eax;" :"=a"(newregs->ss)); + asm volatile("movl %%cs, %%eax;" :"=a"(newregs->cs)); + asm volatile("pushfq; popq %0" :"=m"(newregs->flags)); #endif newregs->ip = (unsigned long)current_text_addr(); } diff --git a/include/asm-x86/kprobes.h b/include/asm-x86/kprobes.h index 61ad7b5d142e..54980b0b3892 100644 --- a/include/asm-x86/kprobes.h +++ b/include/asm-x86/kprobes.h @@ -35,12 +35,12 @@ typedef u8 kprobe_opcode_t; #define RELATIVEJUMP_INSTRUCTION 0xe9 #define MAX_INSN_SIZE 16 #define MAX_STACK_SIZE 64 -#define MIN_STACK_SIZE(ADDR) (((MAX_STACK_SIZE) < \ - (((unsigned long)current_thread_info()) + THREAD_SIZE \ - - (unsigned long)(ADDR))) \ - ? (MAX_STACK_SIZE) \ - : (((unsigned long)current_thread_info()) + THREAD_SIZE \ - - (unsigned long)(ADDR))) +#define MIN_STACK_SIZE(ADDR) \ + (((MAX_STACK_SIZE) < (((unsigned long)current_thread_info()) + \ + THREAD_SIZE - (unsigned long)(ADDR))) \ + ? (MAX_STACK_SIZE) \ + : (((unsigned long)current_thread_info()) + \ + THREAD_SIZE - (unsigned long)(ADDR))) #define flush_insn_slot(p) do { } while (0) diff --git a/include/asm-x86/kvm_host.h b/include/asm-x86/kvm_host.h index 4702b04b979a..68ee390b2844 100644 --- a/include/asm-x86/kvm_host.h +++ b/include/asm-x86/kvm_host.h @@ -22,15 +22,16 @@ #define CR3_PAE_RESERVED_BITS ((X86_CR3_PWT | X86_CR3_PCD) - 1) #define CR3_NONPAE_RESERVED_BITS ((PAGE_SIZE-1) & ~(X86_CR3_PWT | X86_CR3_PCD)) -#define CR3_L_MODE_RESERVED_BITS (CR3_NONPAE_RESERVED_BITS|0xFFFFFF0000000000ULL) +#define CR3_L_MODE_RESERVED_BITS (CR3_NONPAE_RESERVED_BITS | \ + 0xFFFFFF0000000000ULL) -#define KVM_GUEST_CR0_MASK \ +#define KVM_GUEST_CR0_MASK \ (X86_CR0_PG | X86_CR0_PE | X86_CR0_WP | X86_CR0_NE \ | X86_CR0_NW | X86_CR0_CD) -#define KVM_VM_CR0_ALWAYS_ON \ +#define KVM_VM_CR0_ALWAYS_ON \ (X86_CR0_PG | X86_CR0_PE | X86_CR0_WP | X86_CR0_NE | X86_CR0_TS \ | X86_CR0_MP) -#define KVM_GUEST_CR4_MASK \ +#define KVM_GUEST_CR4_MASK \ (X86_CR4_VME | X86_CR4_PSE | X86_CR4_PAE | X86_CR4_PGE | X86_CR4_VMXE) #define KVM_PMODE_VM_CR4_ALWAYS_ON (X86_CR4_PAE | X86_CR4_VMXE) #define KVM_RMODE_VM_CR4_ALWAYS_ON (X86_CR4_VME | X86_CR4_PAE | X86_CR4_VMXE) @@ -133,12 +134,12 @@ struct kvm_pte_chain { union kvm_mmu_page_role { unsigned word; struct { - unsigned glevels : 4; - unsigned level : 4; - unsigned quadrant : 2; - unsigned pad_for_nice_hex_output : 6; - unsigned metaphysical : 1; - unsigned access : 3; + unsigned glevels:4; + unsigned level:4; + unsigned quadrant:2; + unsigned pad_for_nice_hex_output:6; + unsigned metaphysical:1; + unsigned access:3; }; }; @@ -606,6 +607,7 @@ static inline void kvm_inject_gp(struct kvm_vcpu *vcpu, u32 error_code) #define TSS_BASE_SIZE 0x68 #define TSS_IOPB_SIZE (65536 / 8) #define TSS_REDIRECTION_SIZE (256 / 8) -#define RMODE_TSS_SIZE (TSS_BASE_SIZE + TSS_REDIRECTION_SIZE + TSS_IOPB_SIZE + 1) +#define RMODE_TSS_SIZE \ + (TSS_BASE_SIZE + TSS_REDIRECTION_SIZE + TSS_IOPB_SIZE + 1) #endif diff --git a/include/asm-x86/kvm_x86_emulate.h b/include/asm-x86/kvm_x86_emulate.h index 7db91b9bdcd4..d6337f941c98 100644 --- a/include/asm-x86/kvm_x86_emulate.h +++ b/include/asm-x86/kvm_x86_emulate.h @@ -68,10 +68,10 @@ struct x86_emulate_ops { * @val: [OUT] Value read from memory, zero-extended to 'u_long'. * @bytes: [IN ] Number of bytes to read from memory. */ - int (*read_emulated) (unsigned long addr, - void *val, - unsigned int bytes, - struct kvm_vcpu *vcpu); + int (*read_emulated)(unsigned long addr, + void *val, + unsigned int bytes, + struct kvm_vcpu *vcpu); /* * write_emulated: Read bytes from emulated/special memory area. @@ -80,10 +80,10 @@ struct x86_emulate_ops { * required). * @bytes: [IN ] Number of bytes to write to memory. */ - int (*write_emulated) (unsigned long addr, - const void *val, - unsigned int bytes, - struct kvm_vcpu *vcpu); + int (*write_emulated)(unsigned long addr, + const void *val, + unsigned int bytes, + struct kvm_vcpu *vcpu); /* * cmpxchg_emulated: Emulate an atomic (LOCKed) CMPXCHG operation on an @@ -93,11 +93,11 @@ struct x86_emulate_ops { * @new: [IN ] Value to write to @addr. * @bytes: [IN ] Number of bytes to access using CMPXCHG. */ - int (*cmpxchg_emulated) (unsigned long addr, - const void *old, - const void *new, - unsigned int bytes, - struct kvm_vcpu *vcpu); + int (*cmpxchg_emulated)(unsigned long addr, + const void *old, + const void *new, + unsigned int bytes, + struct kvm_vcpu *vcpu); }; @@ -143,7 +143,7 @@ struct x86_emulate_ctxt { /* Register state before/after emulation. */ struct kvm_vcpu *vcpu; - /* Linear faulting address (if emulating a page-faulting instruction). */ + /* Linear faulting address (if emulating a page-faulting instruction) */ unsigned long eflags; /* Emulated execution mode, represented by an X86EMUL_MODE value. */ diff --git a/include/asm-x86/lguest.h b/include/asm-x86/lguest.h index 9b17571e9bc3..be4a7247fa2b 100644 --- a/include/asm-x86/lguest.h +++ b/include/asm-x86/lguest.h @@ -34,8 +34,7 @@ extern const char lgstart_iret[], lgend_iret[]; extern void lguest_iret(void); extern void lguest_init(void); -struct lguest_regs -{ +struct lguest_regs { /* Manually saved part. */ unsigned long eax, ebx, ecx, edx; unsigned long esi, edi, ebp; @@ -51,8 +50,7 @@ struct lguest_regs }; /* This is a guest-specific page (mapped ro) into the guest. */ -struct lguest_ro_state -{ +struct lguest_ro_state { /* Host information we need to restore when we switch back. */ u32 host_cr3; struct desc_ptr host_idt_desc; @@ -67,8 +65,7 @@ struct lguest_ro_state struct desc_struct guest_gdt[GDT_ENTRIES]; }; -struct lg_cpu_arch -{ +struct lg_cpu_arch { /* The GDT entries copied into lguest_ro_state when running. */ struct desc_struct gdt[GDT_ENTRIES]; @@ -85,7 +82,7 @@ static inline void lguest_set_ts(void) cr0 = read_cr0(); if (!(cr0 & 8)) - write_cr0(cr0|8); + write_cr0(cr0 | 8); } /* Full 4G segment descriptors, suitable for CS and DS. */ diff --git a/include/asm-x86/lguest_hcall.h b/include/asm-x86/lguest_hcall.h index f239e7069cab..a3241f28e34a 100644 --- a/include/asm-x86/lguest_hcall.h +++ b/include/asm-x86/lguest_hcall.h @@ -46,7 +46,7 @@ hcall(unsigned long call, { /* "int" is the Intel instruction to trigger a trap. */ asm volatile("int $" __stringify(LGUEST_TRAP_ENTRY) - /* The call in %eax (aka "a") might be overwritten */ + /* The call in %eax (aka "a") might be overwritten */ : "=a"(call) /* The arguments are in %eax, %edx, %ebx & %ecx */ : "a"(call), "d"(arg1), "b"(arg2), "c"(arg3) @@ -62,8 +62,7 @@ hcall(unsigned long call, #define LGUEST_IRQS (NR_IRQS < 32 ? NR_IRQS: 32) #define LHCALL_RING_SIZE 64 -struct hcall_args -{ +struct hcall_args { /* These map directly onto eax, ebx, ecx, edx in struct lguest_regs */ unsigned long arg0, arg2, arg3, arg1; }; diff --git a/include/asm-x86/linkage.h b/include/asm-x86/linkage.h index c048353f4b85..64e444f8e85b 100644 --- a/include/asm-x86/linkage.h +++ b/include/asm-x86/linkage.h @@ -1,6 +1,9 @@ #ifndef __ASM_LINKAGE_H #define __ASM_LINKAGE_H +#undef notrace +#define notrace __attribute__((no_instrument_function)) + #ifdef CONFIG_X86_64 #define __ALIGN .p2align 4,,15 #define __ALIGN_STR ".p2align 4,,15" diff --git a/include/asm-x86/local.h b/include/asm-x86/local.h index f852c62b3319..330a72496abd 100644 --- a/include/asm-x86/local.h +++ b/include/asm-x86/local.h @@ -18,32 +18,28 @@ typedef struct { static inline void local_inc(local_t *l) { - __asm__ __volatile__( - _ASM_INC "%0" - :"+m" (l->a.counter)); + asm volatile(_ASM_INC "%0" + : "+m" (l->a.counter)); } static inline void local_dec(local_t *l) { - __asm__ __volatile__( - _ASM_DEC "%0" - :"+m" (l->a.counter)); + asm volatile(_ASM_DEC "%0" + : "+m" (l->a.counter)); } static inline void local_add(long i, local_t *l) { - __asm__ __volatile__( - _ASM_ADD "%1,%0" - :"+m" (l->a.counter) - :"ir" (i)); + asm volatile(_ASM_ADD "%1,%0" + : "+m" (l->a.counter) + : "ir" (i)); } static inline void local_sub(long i, local_t *l) { - __asm__ __volatile__( - _ASM_SUB "%1,%0" - :"+m" (l->a.counter) - :"ir" (i)); + asm volatile(_ASM_SUB "%1,%0" + : "+m" (l->a.counter) + : "ir" (i)); } /** @@ -59,10 +55,9 @@ static inline int local_sub_and_test(long i, local_t *l) { unsigned char c; - __asm__ __volatile__( - _ASM_SUB "%2,%0; sete %1" - :"+m" (l->a.counter), "=qm" (c) - :"ir" (i) : "memory"); + asm volatile(_ASM_SUB "%2,%0; sete %1" + : "+m" (l->a.counter), "=qm" (c) + : "ir" (i) : "memory"); return c; } @@ -78,10 +73,9 @@ static inline int local_dec_and_test(local_t *l) { unsigned char c; - __asm__ __volatile__( - _ASM_DEC "%0; sete %1" - :"+m" (l->a.counter), "=qm" (c) - : : "memory"); + asm volatile(_ASM_DEC "%0; sete %1" + : "+m" (l->a.counter), "=qm" (c) + : : "memory"); return c != 0; } @@ -97,10 +91,9 @@ static inline int local_inc_and_test(local_t *l) { unsigned char c; - __asm__ __volatile__( - _ASM_INC "%0; sete %1" - :"+m" (l->a.counter), "=qm" (c) - : : "memory"); + asm volatile(_ASM_INC "%0; sete %1" + : "+m" (l->a.counter), "=qm" (c) + : : "memory"); return c != 0; } @@ -117,10 +110,9 @@ static inline int local_add_negative(long i, local_t *l) { unsigned char c; - __asm__ __volatile__( - _ASM_ADD "%2,%0; sets %1" - :"+m" (l->a.counter), "=qm" (c) - :"ir" (i) : "memory"); + asm volatile(_ASM_ADD "%2,%0; sets %1" + : "+m" (l->a.counter), "=qm" (c) + : "ir" (i) : "memory"); return c; } @@ -141,10 +133,9 @@ static inline long local_add_return(long i, local_t *l) #endif /* Modern 486+ processor */ __i = i; - __asm__ __volatile__( - _ASM_XADD "%0, %1;" - :"+r" (i), "+m" (l->a.counter) - : : "memory"); + asm volatile(_ASM_XADD "%0, %1;" + : "+r" (i), "+m" (l->a.counter) + : : "memory"); return i + __i; #ifdef CONFIG_M386 @@ -182,11 +173,11 @@ static inline long local_sub_return(long i, local_t *l) #define local_add_unless(l, a, u) \ ({ \ long c, old; \ - c = local_read(l); \ + c = local_read((l)); \ for (;;) { \ if (unlikely(c == (u))) \ break; \ - old = local_cmpxchg((l), c, c + (a)); \ + old = local_cmpxchg((l), c, c + (a)); \ if (likely(old == c)) \ break; \ c = old; \ @@ -214,26 +205,30 @@ static inline long local_sub_return(long i, local_t *l) /* Need to disable preemption for the cpu local counters otherwise we could still access a variable of a previous CPU in a non atomic way. */ -#define cpu_local_wrap_v(l) \ - ({ local_t res__; \ - preempt_disable(); \ - res__ = (l); \ - preempt_enable(); \ - res__; }) +#define cpu_local_wrap_v(l) \ +({ \ + local_t res__; \ + preempt_disable(); \ + res__ = (l); \ + preempt_enable(); \ + res__; \ +}) #define cpu_local_wrap(l) \ - ({ preempt_disable(); \ - l; \ - preempt_enable(); }) \ - -#define cpu_local_read(l) cpu_local_wrap_v(local_read(&__get_cpu_var(l))) -#define cpu_local_set(l, i) cpu_local_wrap(local_set(&__get_cpu_var(l), (i))) -#define cpu_local_inc(l) cpu_local_wrap(local_inc(&__get_cpu_var(l))) -#define cpu_local_dec(l) cpu_local_wrap(local_dec(&__get_cpu_var(l))) -#define cpu_local_add(i, l) cpu_local_wrap(local_add((i), &__get_cpu_var(l))) -#define cpu_local_sub(i, l) cpu_local_wrap(local_sub((i), &__get_cpu_var(l))) - -#define __cpu_local_inc(l) cpu_local_inc(l) -#define __cpu_local_dec(l) cpu_local_dec(l) +({ \ + preempt_disable(); \ + (l); \ + preempt_enable(); \ +}) \ + +#define cpu_local_read(l) cpu_local_wrap_v(local_read(&__get_cpu_var((l)))) +#define cpu_local_set(l, i) cpu_local_wrap(local_set(&__get_cpu_var((l)), (i))) +#define cpu_local_inc(l) cpu_local_wrap(local_inc(&__get_cpu_var((l)))) +#define cpu_local_dec(l) cpu_local_wrap(local_dec(&__get_cpu_var((l)))) +#define cpu_local_add(i, l) cpu_local_wrap(local_add((i), &__get_cpu_var((l)))) +#define cpu_local_sub(i, l) cpu_local_wrap(local_sub((i), &__get_cpu_var((l)))) + +#define __cpu_local_inc(l) cpu_local_inc((l)) +#define __cpu_local_dec(l) cpu_local_dec((l)) #define __cpu_local_add(i, l) cpu_local_add((i), (l)) #define __cpu_local_sub(i, l) cpu_local_sub((i), (l)) diff --git a/include/asm-x86/mach-bigsmp/mach_apic.h b/include/asm-x86/mach-bigsmp/mach_apic.h index 6df235e8ea91..8327907c79bf 100644 --- a/include/asm-x86/mach-bigsmp/mach_apic.h +++ b/include/asm-x86/mach-bigsmp/mach_apic.h @@ -1,10 +1,7 @@ #ifndef __ASM_MACH_APIC_H #define __ASM_MACH_APIC_H - -extern u8 bios_cpu_apicid[]; - -#define xapic_phys_to_log_apicid(cpu) (bios_cpu_apicid[cpu]) +#define xapic_phys_to_log_apicid(cpu) (per_cpu(x86_bios_cpu_apicid, cpu)) #define esr_disable (1) static inline int apic_id_registered(void) @@ -90,7 +87,7 @@ static inline int apicid_to_node(int logical_apicid) static inline int cpu_present_to_apicid(int mps_cpu) { if (mps_cpu < NR_CPUS) - return (int) bios_cpu_apicid[mps_cpu]; + return (int) per_cpu(x86_bios_cpu_apicid, mps_cpu); return BAD_APICID; } @@ -109,17 +106,6 @@ static inline int cpu_to_logical_apicid(int cpu) return cpu_physical_id(cpu); } -static inline int mpc_apic_id(struct mpc_config_processor *m, - struct mpc_config_translation *translation_record) -{ - printk("Processor #%d %u:%u APIC version %d\n", - m->mpc_apicid, - (m->mpc_cpufeature & CPU_FAMILY_MASK) >> 8, - (m->mpc_cpufeature & CPU_MODEL_MASK) >> 4, - m->mpc_apicver); - return m->mpc_apicid; -} - static inline physid_mask_t ioapic_phys_id_map(physid_mask_t phys_map) { /* For clustered we don't have a good way to do this yet - hack */ diff --git a/include/asm-x86/mach-default/mach_apic.h b/include/asm-x86/mach-default/mach_apic.h index e3c2c1012c1c..0a6634f62abe 100644 --- a/include/asm-x86/mach-default/mach_apic.h +++ b/include/asm-x86/mach-default/mach_apic.h @@ -1,6 +1,8 @@ #ifndef __ASM_MACH_APIC_H #define __ASM_MACH_APIC_H +#ifdef CONFIG_X86_LOCAL_APIC + #include <mach_apicdef.h> #include <asm/smp.h> @@ -14,24 +16,25 @@ static inline cpumask_t target_cpus(void) return cpumask_of_cpu(0); #endif } -#define TARGET_CPUS (target_cpus()) #define NO_BALANCE_IRQ (0) #define esr_disable (0) +#ifdef CONFIG_X86_64 +#include <asm/genapic.h> +#define INT_DELIVERY_MODE (genapic->int_delivery_mode) +#define INT_DEST_MODE (genapic->int_dest_mode) +#define TARGET_CPUS (genapic->target_cpus()) +#define apic_id_registered (genapic->apic_id_registered) +#define init_apic_ldr (genapic->init_apic_ldr) +#define cpu_mask_to_apicid (genapic->cpu_mask_to_apicid) +#define phys_pkg_id (genapic->phys_pkg_id) +#define vector_allocation_domain (genapic->vector_allocation_domain) +extern void setup_apic_routing(void); +#else #define INT_DELIVERY_MODE dest_LowestPrio #define INT_DEST_MODE 1 /* logical delivery broadcast to all procs */ - -static inline unsigned long check_apicid_used(physid_mask_t bitmap, int apicid) -{ - return physid_isset(apicid, bitmap); -} - -static inline unsigned long check_apicid_present(int bit) -{ - return physid_isset(bit, phys_cpu_present_map); -} - +#define TARGET_CPUS (target_cpus()) /* * Set up the logical destination ID. * @@ -49,23 +52,51 @@ static inline void init_apic_ldr(void) apic_write_around(APIC_LDR, val); } -static inline physid_mask_t ioapic_phys_id_map(physid_mask_t phys_map) +static inline int apic_id_registered(void) { - return phys_map; + return physid_isset(GET_APIC_ID(read_apic_id()), phys_cpu_present_map); +} + +static inline unsigned int cpu_mask_to_apicid(cpumask_t cpumask) +{ + return cpus_addr(cpumask)[0]; +} + +static inline u32 phys_pkg_id(u32 cpuid_apic, int index_msb) +{ + return cpuid_apic >> index_msb; } static inline void setup_apic_routing(void) { +#ifdef CONFIG_X86_IO_APIC printk("Enabling APIC mode: %s. Using %d I/O APICs\n", "Flat", nr_ioapics); +#endif } -static inline int multi_timer_check(int apic, int irq) +static inline int apicid_to_node(int logical_apicid) { return 0; } +#endif -static inline int apicid_to_node(int logical_apicid) +static inline unsigned long check_apicid_used(physid_mask_t bitmap, int apicid) +{ + return physid_isset(apicid, bitmap); +} + +static inline unsigned long check_apicid_present(int bit) +{ + return physid_isset(bit, phys_cpu_present_map); +} + +static inline physid_mask_t ioapic_phys_id_map(physid_mask_t phys_map) +{ + return phys_map; +} + +static inline int multi_timer_check(int apic, int irq) { return 0; } @@ -78,8 +109,13 @@ static inline int cpu_to_logical_apicid(int cpu) static inline int cpu_present_to_apicid(int mps_cpu) { +#ifdef CONFIG_X86_64 + if (cpu_present(mps_cpu)) + return (int)per_cpu(x86_bios_cpu_apicid, mps_cpu); +#else if (mps_cpu < get_physical_broadcast()) return mps_cpu; +#endif else return BAD_APICID; } @@ -89,17 +125,6 @@ static inline physid_mask_t apicid_to_cpu_present(int phys_apicid) return physid_mask_of_physid(phys_apicid); } -static inline int mpc_apic_id(struct mpc_config_processor *m, - struct mpc_config_translation *translation_record) -{ - printk("Processor #%d %u:%u APIC version %d\n", - m->mpc_apicid, - (m->mpc_cpufeature & CPU_FAMILY_MASK) >> 8, - (m->mpc_cpufeature & CPU_MODEL_MASK) >> 4, - m->mpc_apicver); - return m->mpc_apicid; -} - static inline void setup_portio_remap(void) { } @@ -109,23 +134,9 @@ static inline int check_phys_apicid_present(int boot_cpu_physical_apicid) return physid_isset(boot_cpu_physical_apicid, phys_cpu_present_map); } -static inline int apic_id_registered(void) -{ - return physid_isset(GET_APIC_ID(apic_read(APIC_ID)), phys_cpu_present_map); -} - -static inline unsigned int cpu_mask_to_apicid(cpumask_t cpumask) -{ - return cpus_addr(cpumask)[0]; -} - static inline void enable_apic_mode(void) { } -static inline u32 phys_pkg_id(u32 cpuid_apic, int index_msb) -{ - return cpuid_apic >> index_msb; -} - +#endif /* CONFIG_X86_LOCAL_APIC */ #endif /* __ASM_MACH_APIC_H */ diff --git a/include/asm-x86/mach-default/mach_apicdef.h b/include/asm-x86/mach-default/mach_apicdef.h index ae9841319094..e4b29ba37de6 100644 --- a/include/asm-x86/mach-default/mach_apicdef.h +++ b/include/asm-x86/mach-default/mach_apicdef.h @@ -3,10 +3,14 @@ #include <asm/apic.h> +#ifdef CONFIG_X86_64 +#define APIC_ID_MASK (0xFFu<<24) +#define GET_APIC_ID(x) (((x)>>24)&0xFFu) +#define SET_APIC_ID(x) (((x)<<24)) +#else #define APIC_ID_MASK (0xF<<24) - static inline unsigned get_apic_id(unsigned long x) -{ +{ unsigned int ver = GET_APIC_VERSION(apic_read(APIC_LVR)); if (APIC_XAPIC(ver)) return (((x)>>24)&0xFF); @@ -15,5 +19,6 @@ static inline unsigned get_apic_id(unsigned long x) } #define GET_APIC_ID(x) get_apic_id(x) +#endif #endif diff --git a/include/asm-x86/mach-default/mach_ipi.h b/include/asm-x86/mach-default/mach_ipi.h index 0dba244c86db..be323364e68f 100644 --- a/include/asm-x86/mach-default/mach_ipi.h +++ b/include/asm-x86/mach-default/mach_ipi.h @@ -9,10 +9,15 @@ void __send_IPI_shortcut(unsigned int shortcut, int vector); extern int no_broadcast; +#ifdef CONFIG_X86_64 +#include <asm/genapic.h> +#define send_IPI_mask (genapic->send_IPI_mask) +#else static inline void send_IPI_mask(cpumask_t mask, int vector) { send_IPI_mask_bitmask(mask, vector); } +#endif static inline void __local_send_IPI_allbutself(int vector) { @@ -33,6 +38,10 @@ static inline void __local_send_IPI_all(int vector) __send_IPI_shortcut(APIC_DEST_ALLINC, vector); } +#ifdef CONFIG_X86_64 +#define send_IPI_allbutself (genapic->send_IPI_allbutself) +#define send_IPI_all (genapic->send_IPI_all) +#else static inline void send_IPI_allbutself(int vector) { /* @@ -50,5 +59,6 @@ static inline void send_IPI_all(int vector) { __local_send_IPI_all(vector); } +#endif #endif /* __ASM_MACH_IPI_H */ diff --git a/include/asm-x86/mach-default/mach_mpparse.h b/include/asm-x86/mach-default/mach_mpparse.h index 1d3832482580..d14108505bb8 100644 --- a/include/asm-x86/mach-default/mach_mpparse.h +++ b/include/asm-x86/mach-default/mach_mpparse.h @@ -1,17 +1,6 @@ #ifndef __ASM_MACH_MPPARSE_H #define __ASM_MACH_MPPARSE_H -static inline void mpc_oem_bus_info(struct mpc_config_bus *m, char *name, - struct mpc_config_translation *translation) -{ -// Dprintk("Bus #%d is %s\n", m->mpc_busid, name); -} - -static inline void mpc_oem_pci_bus(struct mpc_config_bus *m, - struct mpc_config_translation *translation) -{ -} - static inline int mps_oem_check(struct mp_config_table *mpc, char *oem, char *productid) { diff --git a/include/asm-x86/mach-default/mach_reboot.h b/include/asm-x86/mach-default/mach_reboot.h deleted file mode 100644 index 6adee6a97dec..000000000000 --- a/include/asm-x86/mach-default/mach_reboot.h +++ /dev/null @@ -1,61 +0,0 @@ -/* - * arch/i386/mach-generic/mach_reboot.h - * - * Machine specific reboot functions for generic. - * Split out from reboot.c by Osamu Tomita <tomita@cinet.co.jp> - */ -#ifndef _MACH_REBOOT_H -#define _MACH_REBOOT_H - -static inline void kb_wait(void) -{ - int i; - - for (i = 0; i < 0x10000; i++) - if ((inb_p(0x64) & 0x02) == 0) - break; -} - -static inline void mach_reboot(void) -{ - int i; - - /* old method, works on most machines */ - for (i = 0; i < 10; i++) { - kb_wait(); - udelay(50); - outb(0xfe, 0x64); /* pulse reset low */ - udelay(50); - } - - /* New method: sets the "System flag" which, when set, indicates - * successful completion of the keyboard controller self-test (Basic - * Assurance Test, BAT). This is needed for some machines with no - * keyboard plugged in. This read-modify-write sequence sets only the - * system flag - */ - for (i = 0; i < 10; i++) { - int cmd; - - outb(0x20, 0x64); /* read Controller Command Byte */ - udelay(50); - kb_wait(); - udelay(50); - cmd = inb(0x60); - udelay(50); - kb_wait(); - udelay(50); - outb(0x60, 0x64); /* write Controller Command Byte */ - udelay(50); - kb_wait(); - udelay(50); - outb(cmd | 0x14, 0x60); /* set "System flag" and "Keyboard Disabled" */ - udelay(50); - kb_wait(); - udelay(50); - outb(0xfe, 0x64); /* pulse reset low */ - udelay(50); - } -} - -#endif /* !_MACH_REBOOT_H */ diff --git a/include/asm-x86/mach-default/smpboot_hooks.h b/include/asm-x86/mach-default/smpboot_hooks.h index 7f45f6311059..3ff2c5bff93a 100644 --- a/include/asm-x86/mach-default/smpboot_hooks.h +++ b/include/asm-x86/mach-default/smpboot_hooks.h @@ -41,4 +41,11 @@ static inline void smpboot_setup_io_apic(void) */ if (!skip_ioapic_setup && nr_ioapics) setup_IO_APIC(); + else + nr_ioapics = 0; +} + +static inline void smpboot_clear_io_apic(void) +{ + nr_ioapics = 0; } diff --git a/include/asm-x86/mach-es7000/mach_apic.h b/include/asm-x86/mach-es7000/mach_apic.h index d23011fdf454..fbc8ad256f5a 100644 --- a/include/asm-x86/mach-es7000/mach_apic.h +++ b/include/asm-x86/mach-es7000/mach_apic.h @@ -1,9 +1,7 @@ #ifndef __ASM_MACH_APIC_H #define __ASM_MACH_APIC_H -extern u8 bios_cpu_apicid[]; - -#define xapic_phys_to_log_apicid(cpu) (bios_cpu_apicid[cpu]) +#define xapic_phys_to_log_apicid(cpu) per_cpu(x86_bios_cpu_apicid, cpu) #define esr_disable (1) static inline int apic_id_registered(void) @@ -80,7 +78,7 @@ extern void enable_apic_mode(void); extern int apic_version [MAX_APICS]; static inline void setup_apic_routing(void) { - int apic = bios_cpu_apicid[smp_processor_id()]; + int apic = per_cpu(x86_bios_cpu_apicid, smp_processor_id()); printk("Enabling APIC mode: %s. Using %d I/O APICs, target cpus %lx\n", (apic_version[apic] == 0x14) ? "Physical Cluster" : "Logical Cluster", nr_ioapics, cpus_addr(TARGET_CPUS)[0]); @@ -102,7 +100,7 @@ static inline int cpu_present_to_apicid(int mps_cpu) if (!mps_cpu) return boot_cpu_physical_apicid; else if (mps_cpu < NR_CPUS) - return (int) bios_cpu_apicid[mps_cpu]; + return (int) per_cpu(x86_bios_cpu_apicid, mps_cpu); else return BAD_APICID; } @@ -129,16 +127,6 @@ static inline int cpu_to_logical_apicid(int cpu) #endif } -static inline int mpc_apic_id(struct mpc_config_processor *m, struct mpc_config_translation *unused) -{ - printk("Processor #%d %u:%u APIC version %d\n", - m->mpc_apicid, - (m->mpc_cpufeature & CPU_FAMILY_MASK) >> 8, - (m->mpc_cpufeature & CPU_MODEL_MASK) >> 4, - m->mpc_apicver); - return (m->mpc_apicid); -} - static inline physid_mask_t ioapic_phys_id_map(physid_mask_t phys_map) { /* For clustered we don't have a good way to do this yet - hack */ @@ -153,7 +141,7 @@ static inline void setup_portio_remap(void) extern unsigned int boot_cpu_physical_apicid; static inline int check_phys_apicid_present(int cpu_physical_apicid) { - boot_cpu_physical_apicid = GET_APIC_ID(apic_read(APIC_ID)); + boot_cpu_physical_apicid = GET_APIC_ID(read_apic_id()); return (1); } diff --git a/include/asm-x86/mach-es7000/mach_mpparse.h b/include/asm-x86/mach-es7000/mach_mpparse.h index 52ee75cd0fe1..ef26d3523625 100644 --- a/include/asm-x86/mach-es7000/mach_mpparse.h +++ b/include/asm-x86/mach-es7000/mach_mpparse.h @@ -3,17 +3,6 @@ #include <linux/acpi.h> -static inline void mpc_oem_bus_info(struct mpc_config_bus *m, char *name, - struct mpc_config_translation *translation) -{ - Dprintk("Bus #%d is %s\n", m->mpc_busid, name); -} - -static inline void mpc_oem_pci_bus(struct mpc_config_bus *m, - struct mpc_config_translation *translation) -{ -} - extern int parse_unisys_oem (char *oemptr); extern int find_unisys_acpi_oem_table(unsigned long *oem_addr); extern void setup_unisys(void); diff --git a/include/asm-x86/mach-generic/mach_apic.h b/include/asm-x86/mach-generic/mach_apic.h index a236e7021528..6eff343e1233 100644 --- a/include/asm-x86/mach-generic/mach_apic.h +++ b/include/asm-x86/mach-generic/mach_apic.h @@ -19,7 +19,6 @@ #define cpu_to_logical_apicid (genapic->cpu_to_logical_apicid) #define cpu_present_to_apicid (genapic->cpu_present_to_apicid) #define apicid_to_cpu_present (genapic->apicid_to_cpu_present) -#define mpc_apic_id (genapic->mpc_apic_id) #define setup_portio_remap (genapic->setup_portio_remap) #define check_apicid_present (genapic->check_apicid_present) #define check_phys_apicid_present (genapic->check_phys_apicid_present) diff --git a/include/asm-x86/mach-generic/mach_mpparse.h b/include/asm-x86/mach-generic/mach_mpparse.h index dbd9fce54f4d..0d0b5ba2e9d1 100644 --- a/include/asm-x86/mach-generic/mach_mpparse.h +++ b/include/asm-x86/mach-generic/mach_mpparse.h @@ -1,11 +1,6 @@ #ifndef _MACH_MPPARSE_H #define _MACH_MPPARSE_H 1 -#include <asm/genapic.h> - -#define mpc_oem_bus_info (genapic->mpc_oem_bus_info) -#define mpc_oem_pci_bus (genapic->mpc_oem_pci_bus) - int mps_oem_check(struct mp_config_table *mpc, char *oem, char *productid); int acpi_madt_oem_check(char *oem_id, char *oem_table_id); diff --git a/include/asm-x86/mach-numaq/mach_apic.h b/include/asm-x86/mach-numaq/mach_apic.h index 3b637fac890b..75a56e5afbe7 100644 --- a/include/asm-x86/mach-numaq/mach_apic.h +++ b/include/asm-x86/mach-numaq/mach_apic.h @@ -95,6 +95,16 @@ static inline physid_mask_t apicid_to_cpu_present(int logical_apicid) return physid_mask_of_physid(cpu + 4*node); } +struct mpc_config_translation { + unsigned char mpc_type; + unsigned char trans_len; + unsigned char trans_type; + unsigned char trans_quad; + unsigned char trans_global; + unsigned char trans_local; + unsigned short trans_reserved; +}; + static inline int mpc_apic_id(struct mpc_config_processor *m, struct mpc_config_translation *translation_record) { diff --git a/include/asm-x86/mach-numaq/mach_mpparse.h b/include/asm-x86/mach-numaq/mach_mpparse.h index 51bbac8fc0c2..459b12401187 100644 --- a/include/asm-x86/mach-numaq/mach_mpparse.h +++ b/include/asm-x86/mach-numaq/mach_mpparse.h @@ -1,25 +1,10 @@ #ifndef __ASM_MACH_MPPARSE_H #define __ASM_MACH_MPPARSE_H -static inline void mpc_oem_bus_info(struct mpc_config_bus *m, char *name, - struct mpc_config_translation *translation) -{ - int quad = translation->trans_quad; - int local = translation->trans_local; - - mp_bus_id_to_node[m->mpc_busid] = quad; - mp_bus_id_to_local[m->mpc_busid] = local; - printk("Bus #%d is %s (node %d)\n", m->mpc_busid, name, quad); -} - -static inline void mpc_oem_pci_bus(struct mpc_config_bus *m, - struct mpc_config_translation *translation) -{ - int quad = translation->trans_quad; - int local = translation->trans_local; - - quad_local_to_mp_bus_id[quad][local] = m->mpc_busid; -} +extern void mpc_oem_bus_info(struct mpc_config_bus *m, char *name, + struct mpc_config_translation *translation); +extern void mpc_oem_pci_bus(struct mpc_config_bus *m, + struct mpc_config_translation *translation); /* Hook from generic ACPI tables.c */ static inline void acpi_madt_oem_check(char *oem_id, char *oem_table_id) diff --git a/include/asm-x86/mach-summit/mach_apic.h b/include/asm-x86/mach-summit/mach_apic.h index 062c97f6100b..1f76c2e70232 100644 --- a/include/asm-x86/mach-summit/mach_apic.h +++ b/include/asm-x86/mach-summit/mach_apic.h @@ -40,7 +40,6 @@ static inline unsigned long check_apicid_present(int bit) #define apicid_cluster(apicid) ((apicid) & XAPIC_DEST_CLUSTER_MASK) -extern u8 bios_cpu_apicid[]; extern u8 cpu_2_logical_apicid[]; static inline void init_apic_ldr(void) @@ -110,7 +109,7 @@ static inline int cpu_to_logical_apicid(int cpu) static inline int cpu_present_to_apicid(int mps_cpu) { if (mps_cpu < NR_CPUS) - return (int)bios_cpu_apicid[mps_cpu]; + return (int)per_cpu(x86_bios_cpu_apicid, mps_cpu); else return BAD_APICID; } @@ -126,17 +125,6 @@ static inline physid_mask_t apicid_to_cpu_present(int apicid) return physid_mask_of_physid(0); } -static inline int mpc_apic_id(struct mpc_config_processor *m, - struct mpc_config_translation *translation_record) -{ - printk("Processor #%d %u:%u APIC version %d\n", - m->mpc_apicid, - (m->mpc_cpufeature & CPU_FAMILY_MASK) >> 8, - (m->mpc_cpufeature & CPU_MODEL_MASK) >> 4, - m->mpc_apicver); - return m->mpc_apicid; -} - static inline void setup_portio_remap(void) { } diff --git a/include/asm-x86/mach-summit/mach_mpparse.h b/include/asm-x86/mach-summit/mach_mpparse.h index c2520539d934..fdf591701339 100644 --- a/include/asm-x86/mach-summit/mach_mpparse.h +++ b/include/asm-x86/mach-summit/mach_mpparse.h @@ -12,17 +12,6 @@ extern void setup_summit(void); #define setup_summit() {} #endif -static inline void mpc_oem_bus_info(struct mpc_config_bus *m, char *name, - struct mpc_config_translation *translation) -{ - Dprintk("Bus #%d is %s\n", m->mpc_busid, name); -} - -static inline void mpc_oem_pci_bus(struct mpc_config_bus *m, - struct mpc_config_translation *translation) -{ -} - static inline int mps_oem_check(struct mp_config_table *mpc, char *oem, char *productid) { diff --git a/include/asm-x86/mach-visws/mach_apic.h b/include/asm-x86/mach-visws/mach_apic.h index efac6f0d139f..a9ef33a8a995 100644 --- a/include/asm-x86/mach-visws/mach_apic.h +++ b/include/asm-x86/mach-visws/mach_apic.h @@ -23,7 +23,7 @@ static inline int apic_id_registered(void) { - return physid_isset(GET_APIC_ID(apic_read(APIC_ID)), phys_cpu_present_map); + return physid_isset(GET_APIC_ID(read_apic_id()), phys_cpu_present_map); } /* diff --git a/include/asm-x86/mach-visws/smpboot_hooks.h b/include/asm-x86/mach-visws/smpboot_hooks.h index d926471fa359..c9b83e395a2e 100644 --- a/include/asm-x86/mach-visws/smpboot_hooks.h +++ b/include/asm-x86/mach-visws/smpboot_hooks.h @@ -22,3 +22,7 @@ static inline void smpboot_restore_warm_reset_vector(void) static inline void smpboot_setup_io_apic(void) { } + +static inline void smpboot_clear_io_apic(void) +{ +} diff --git a/include/asm-x86/mach_apic.h b/include/asm-x86/mach_apic.h deleted file mode 100644 index 7b7115a0c1c9..000000000000 --- a/include/asm-x86/mach_apic.h +++ /dev/null @@ -1,29 +0,0 @@ -#ifndef __ASM_MACH_APIC_H -#define __ASM_MACH_APIC_H - -/* - * Copyright 2004 James Cleverdon, IBM. - * Subject to the GNU Public License, v.2 - * - * Generic APIC sub-arch defines. - * - * Hacked for x86-64 by James Cleverdon from i386 architecture code by - * Martin Bligh, Andi Kleen, James Bottomley, John Stultz, and - * James Cleverdon. - */ - -#include <asm/genapic.h> - -#define INT_DELIVERY_MODE (genapic->int_delivery_mode) -#define INT_DEST_MODE (genapic->int_dest_mode) -#define TARGET_CPUS (genapic->target_cpus()) -#define vector_allocation_domain (genapic->vector_allocation_domain) -#define apic_id_registered (genapic->apic_id_registered) -#define init_apic_ldr (genapic->init_apic_ldr) -#define send_IPI_mask (genapic->send_IPI_mask) -#define send_IPI_allbutself (genapic->send_IPI_allbutself) -#define send_IPI_all (genapic->send_IPI_all) -#define cpu_mask_to_apicid (genapic->cpu_mask_to_apicid) -#define phys_pkg_id (genapic->phys_pkg_id) - -#endif /* __ASM_MACH_APIC_H */ diff --git a/include/asm-x86/mc146818rtc.h b/include/asm-x86/mc146818rtc.h index cdd9f965835a..daf1ccde77af 100644 --- a/include/asm-x86/mc146818rtc.h +++ b/include/asm-x86/mc146818rtc.h @@ -42,7 +42,7 @@ extern volatile unsigned long cmos_lock; static inline void lock_cmos(unsigned char reg) { unsigned long new; - new = ((smp_processor_id()+1) << 8) | reg; + new = ((smp_processor_id() + 1) << 8) | reg; for (;;) { if (cmos_lock) { cpu_relax(); @@ -57,22 +57,26 @@ static inline void unlock_cmos(void) { cmos_lock = 0; } + static inline int do_i_have_lock_cmos(void) { - return (cmos_lock >> 8) == (smp_processor_id()+1); + return (cmos_lock >> 8) == (smp_processor_id() + 1); } + static inline unsigned char current_lock_cmos_reg(void) { return cmos_lock & 0xff; } -#define lock_cmos_prefix(reg) \ + +#define lock_cmos_prefix(reg) \ do { \ unsigned long cmos_flags; \ local_irq_save(cmos_flags); \ lock_cmos(reg) -#define lock_cmos_suffix(reg) \ - unlock_cmos(); \ - local_irq_restore(cmos_flags); \ + +#define lock_cmos_suffix(reg) \ + unlock_cmos(); \ + local_irq_restore(cmos_flags); \ } while (0) #else #define lock_cmos_prefix(reg) do {} while (0) diff --git a/include/asm-x86/mca_dma.h b/include/asm-x86/mca_dma.h index fbb1f3b71279..c3dca6edc6b1 100644 --- a/include/asm-x86/mca_dma.h +++ b/include/asm-x86/mca_dma.h @@ -12,18 +12,18 @@ * count by 2 when using 16-bit dma; that is not handled by these functions. * * Ramen Noodles are yummy. - * - * 1998 Tymm Twillman <tymm@computer.org> + * + * 1998 Tymm Twillman <tymm@computer.org> */ /* - * Registers that are used by the DMA controller; FN is the function register + * Registers that are used by the DMA controller; FN is the function register * (tell the controller what to do) and EXE is the execution register (how * to do it) */ #define MCA_DMA_REG_FN 0x18 -#define MCA_DMA_REG_EXE 0x1A +#define MCA_DMA_REG_EXE 0x1A /* * Functions that the DMA controller can do @@ -43,9 +43,9 @@ /* * Modes (used by setting MCA_DMA_FN_MODE in the function register) - * + * * Note that the MODE_READ is read from memory (write to device), and - * MODE_WRITE is vice-versa. + * MODE_WRITE is vice-versa. */ #define MCA_DMA_MODE_XFER 0x04 /* read by default */ @@ -63,7 +63,7 @@ * IRQ context. */ -static __inline__ void mca_enable_dma(unsigned int dmanr) +static inline void mca_enable_dma(unsigned int dmanr) { outb(MCA_DMA_FN_RESET_MASK | dmanr, MCA_DMA_REG_FN); } @@ -76,7 +76,7 @@ static __inline__ void mca_enable_dma(unsigned int dmanr) * IRQ context. */ -static __inline__ void mca_disable_dma(unsigned int dmanr) +static inline void mca_disable_dma(unsigned int dmanr) { outb(MCA_DMA_FN_MASK | dmanr, MCA_DMA_REG_FN); } @@ -87,10 +87,10 @@ static __inline__ void mca_disable_dma(unsigned int dmanr) * @a: 24bit bus address * * Load the address register in the DMA controller. This has a 24bit - * limitation (16Mb). + * limitation (16Mb). */ -static __inline__ void mca_set_dma_addr(unsigned int dmanr, unsigned int a) +static inline void mca_set_dma_addr(unsigned int dmanr, unsigned int a) { outb(MCA_DMA_FN_SET_ADDR | dmanr, MCA_DMA_REG_FN); outb(a & 0xff, MCA_DMA_REG_EXE); @@ -106,14 +106,14 @@ static __inline__ void mca_set_dma_addr(unsigned int dmanr, unsigned int a) * limitation (16Mb). The return is a bus address. */ -static __inline__ unsigned int mca_get_dma_addr(unsigned int dmanr) +static inline unsigned int mca_get_dma_addr(unsigned int dmanr) { unsigned int addr; outb(MCA_DMA_FN_GET_ADDR | dmanr, MCA_DMA_REG_FN); addr = inb(MCA_DMA_REG_EXE); addr |= inb(MCA_DMA_REG_EXE) << 8; - addr |= inb(MCA_DMA_REG_EXE) << 16; + addr |= inb(MCA_DMA_REG_EXE) << 16; return addr; } @@ -127,7 +127,7 @@ static __inline__ unsigned int mca_get_dma_addr(unsigned int dmanr) * Setting a count of zero will not do what you expect. */ -static __inline__ void mca_set_dma_count(unsigned int dmanr, unsigned int count) +static inline void mca_set_dma_count(unsigned int dmanr, unsigned int count) { count--; /* transfers one more than count -- correct for this */ @@ -144,7 +144,7 @@ static __inline__ void mca_set_dma_count(unsigned int dmanr, unsigned int count) * on this DMA channel. */ -static __inline__ unsigned int mca_get_dma_residue(unsigned int dmanr) +static inline unsigned int mca_get_dma_residue(unsigned int dmanr) { unsigned short count; @@ -164,12 +164,12 @@ static __inline__ unsigned int mca_get_dma_residue(unsigned int dmanr) * with an I/O port target. */ -static __inline__ void mca_set_dma_io(unsigned int dmanr, unsigned int io_addr) +static inline void mca_set_dma_io(unsigned int dmanr, unsigned int io_addr) { /* * DMA from a port address -- set the io address */ - + outb(MCA_DMA_FN_SET_IO | dmanr, MCA_DMA_REG_FN); outb(io_addr & 0xff, MCA_DMA_REG_EXE); outb((io_addr >> 8) & 0xff, MCA_DMA_REG_EXE); @@ -192,7 +192,7 @@ static __inline__ void mca_set_dma_io(unsigned int dmanr, unsigned int io_addr) * %MCA_DMA_MODE_16 to do 16bit transfers. */ -static __inline__ void mca_set_dma_mode(unsigned int dmanr, unsigned int mode) +static inline void mca_set_dma_mode(unsigned int dmanr, unsigned int mode) { outb(MCA_DMA_FN_SET_MODE | dmanr, MCA_DMA_REG_FN); outb(mode, MCA_DMA_REG_EXE); diff --git a/include/asm-x86/mmu.h b/include/asm-x86/mmu.h index efa962c38897..00e88679e11f 100644 --- a/include/asm-x86/mmu.h +++ b/include/asm-x86/mmu.h @@ -10,10 +10,10 @@ * * cpu_vm_mask is used to optimize ldt flushing. */ -typedef struct { +typedef struct { void *ldt; #ifdef CONFIG_X86_64 - rwlock_t ldtlock; + rwlock_t ldtlock; #endif int size; struct mutex lock; diff --git a/include/asm-x86/mmu_context_32.h b/include/asm-x86/mmu_context_32.h index 8198d1cca1f3..9756ae0f1dd3 100644 --- a/include/asm-x86/mmu_context_32.h +++ b/include/asm-x86/mmu_context_32.h @@ -62,7 +62,7 @@ static inline void switch_mm(struct mm_struct *prev, BUG_ON(per_cpu(cpu_tlbstate, cpu).active_mm != next); if (!cpu_test_and_set(cpu, next->cpu_vm_mask)) { - /* We were in lazy tlb mode and leave_mm disabled + /* We were in lazy tlb mode and leave_mm disabled * tlb flush IPI delivery. We must reload %cr3. */ load_cr3(next->pgd); @@ -75,10 +75,10 @@ static inline void switch_mm(struct mm_struct *prev, #define deactivate_mm(tsk, mm) \ asm("movl %0,%%gs": :"r" (0)); -#define activate_mm(prev, next) \ - do { \ - paravirt_activate_mm(prev, next); \ - switch_mm((prev),(next),NULL); \ - } while(0); +#define activate_mm(prev, next) \ +do { \ + paravirt_activate_mm((prev), (next)); \ + switch_mm((prev), (next), NULL); \ +} while (0); #endif diff --git a/include/asm-x86/mmu_context_64.h b/include/asm-x86/mmu_context_64.h index ad6dc821ef9e..ca44c71e7fb3 100644 --- a/include/asm-x86/mmu_context_64.h +++ b/include/asm-x86/mmu_context_64.h @@ -20,12 +20,12 @@ void destroy_context(struct mm_struct *mm); static inline void enter_lazy_tlb(struct mm_struct *mm, struct task_struct *tsk) { #ifdef CONFIG_SMP - if (read_pda(mmu_state) == TLBSTATE_OK) + if (read_pda(mmu_state) == TLBSTATE_OK) write_pda(mmu_state, TLBSTATE_LAZY); #endif } -static inline void switch_mm(struct mm_struct *prev, struct mm_struct *next, +static inline void switch_mm(struct mm_struct *prev, struct mm_struct *next, struct task_struct *tsk) { unsigned cpu = smp_processor_id(); @@ -39,7 +39,7 @@ static inline void switch_mm(struct mm_struct *prev, struct mm_struct *next, cpu_set(cpu, next->cpu_vm_mask); load_cr3(next->pgd); - if (unlikely(next->context.ldt != prev->context.ldt)) + if (unlikely(next->context.ldt != prev->context.ldt)) load_LDT_nolock(&next->context); } #ifdef CONFIG_SMP @@ -48,7 +48,7 @@ static inline void switch_mm(struct mm_struct *prev, struct mm_struct *next, if (read_pda(active_mm) != next) BUG(); if (!cpu_test_and_set(cpu, next->cpu_vm_mask)) { - /* We were in lazy tlb mode and leave_mm disabled + /* We were in lazy tlb mode and leave_mm disabled * tlb flush IPI delivery. We must reload CR3 * to make sure to use no freed page tables. */ @@ -59,13 +59,14 @@ static inline void switch_mm(struct mm_struct *prev, struct mm_struct *next, #endif } -#define deactivate_mm(tsk,mm) do { \ - load_gs_index(0); \ - asm volatile("movl %0,%%fs"::"r"(0)); \ -} while(0) +#define deactivate_mm(tsk, mm) \ +do { \ + load_gs_index(0); \ + asm volatile("movl %0,%%fs"::"r"(0)); \ +} while (0) -#define activate_mm(prev, next) \ - switch_mm((prev),(next),NULL) +#define activate_mm(prev, next) \ + switch_mm((prev), (next), NULL) #endif diff --git a/include/asm-x86/mmx.h b/include/asm-x86/mmx.h index 46b71da99869..940881218ff8 100644 --- a/include/asm-x86/mmx.h +++ b/include/asm-x86/mmx.h @@ -6,7 +6,7 @@ */ #include <linux/types.h> - + extern void *_mmx_memcpy(void *to, const void *from, size_t size); extern void mmx_clear_page(void *page); extern void mmx_copy_page(void *to, void *from); diff --git a/include/asm-x86/mmzone_32.h b/include/asm-x86/mmzone_32.h index 274a59566c45..cb2cad0b65a7 100644 --- a/include/asm-x86/mmzone_32.h +++ b/include/asm-x86/mmzone_32.h @@ -18,7 +18,7 @@ extern struct pglist_data *node_data[]; #include <asm/srat.h> #endif -extern int get_memcfg_numa_flat(void ); +extern int get_memcfg_numa_flat(void); /* * This allows any one NUMA architecture to be compiled * for, and still fall back to the flat function if it @@ -129,7 +129,7 @@ static inline int pfn_valid(int pfn) struct pglist_data __maybe_unused \ *__alloc_bootmem_node__pgdat = (pgdat); \ __alloc_bootmem_node(NODE_DATA(0), (x), PAGE_SIZE, \ - __pa(MAX_DMA_ADDRESS)) \ + __pa(MAX_DMA_ADDRESS)); \ }) #define alloc_bootmem_low_pages_node(pgdat, x) \ ({ \ diff --git a/include/asm-x86/mmzone_64.h b/include/asm-x86/mmzone_64.h index ebaf9663aa8a..594bd0dc1d08 100644 --- a/include/asm-x86/mmzone_64.h +++ b/include/asm-x86/mmzone_64.h @@ -7,7 +7,7 @@ #ifdef CONFIG_NUMA -#define VIRTUAL_BUG_ON(x) +#define VIRTUAL_BUG_ON(x) #include <asm/smp.h> @@ -16,7 +16,7 @@ struct memnode { int shift; unsigned int mapsize; s16 *map; - s16 embedded_map[64-8]; + s16 embedded_map[64 - 8]; } ____cacheline_aligned; /* total size = 128 bytes */ extern struct memnode memnode; #define memnode_shift memnode.shift @@ -25,27 +25,27 @@ extern struct memnode memnode; extern struct pglist_data *node_data[]; -static inline __attribute__((pure)) int phys_to_nid(unsigned long addr) -{ - unsigned nid; +static inline __attribute__((pure)) int phys_to_nid(unsigned long addr) +{ + unsigned nid; VIRTUAL_BUG_ON(!memnodemap); VIRTUAL_BUG_ON((addr >> memnode_shift) >= memnodemapsize); - nid = memnodemap[addr >> memnode_shift]; - VIRTUAL_BUG_ON(nid >= MAX_NUMNODES || !node_data[nid]); - return nid; -} + nid = memnodemap[addr >> memnode_shift]; + VIRTUAL_BUG_ON(nid >= MAX_NUMNODES || !node_data[nid]); + return nid; +} #define NODE_DATA(nid) (node_data[nid]) #define node_start_pfn(nid) (NODE_DATA(nid)->node_start_pfn) -#define node_end_pfn(nid) (NODE_DATA(nid)->node_start_pfn + \ +#define node_end_pfn(nid) (NODE_DATA(nid)->node_start_pfn + \ NODE_DATA(nid)->node_spanned_pages) extern int early_pfn_to_nid(unsigned long pfn); #ifdef CONFIG_NUMA_EMU -#define FAKE_NODE_MIN_SIZE (64*1024*1024) -#define FAKE_NODE_MIN_HASH_MASK (~(FAKE_NODE_MIN_SIZE - 1uL)) +#define FAKE_NODE_MIN_SIZE (64 * 1024 * 1024) +#define FAKE_NODE_MIN_HASH_MASK (~(FAKE_NODE_MIN_SIZE - 1UL)) #endif #endif diff --git a/include/asm-x86/mpspec.h b/include/asm-x86/mpspec.h index 781ad74ab9e9..57a991b9c053 100644 --- a/include/asm-x86/mpspec.h +++ b/include/asm-x86/mpspec.h @@ -1,16 +1,13 @@ #ifndef _AM_X86_MPSPEC_H #define _AM_X86_MPSPEC_H +#include <linux/init.h> + #include <asm/mpspec_def.h> #ifdef CONFIG_X86_32 #include <mach_mpspec.h> -extern int mp_bus_id_to_type[MAX_MP_BUSSES]; -extern int mp_bus_id_to_node[MAX_MP_BUSSES]; -extern int mp_bus_id_to_local[MAX_MP_BUSSES]; -extern int quad_local_to_mp_bus_id[NR_CPUS/4][4]; - extern unsigned int def_to_bigsmp; extern int apic_version[MAX_APICS]; extern u8 apicid_2_node[]; @@ -24,27 +21,30 @@ extern int pic_mode; /* Each PCI slot may be a combo card with its own bus. 4 IRQ pins per slot. */ #define MAX_IRQ_SOURCES (MAX_MP_BUSSES * 4) -extern DECLARE_BITMAP(mp_bus_not_pci, MAX_MP_BUSSES); +extern void early_find_smp_config(void); +extern void early_get_smp_config(void); #endif +#if defined(CONFIG_MCA) || defined(CONFIG_EISA) +extern int mp_bus_id_to_type[MAX_MP_BUSSES]; +#endif + +extern DECLARE_BITMAP(mp_bus_not_pci, MAX_MP_BUSSES); + extern int mp_bus_id_to_pci_bus[MAX_MP_BUSSES]; extern unsigned int boot_cpu_physical_apicid; extern int smp_found_config; -extern int nr_ioapics; -extern int mp_irq_entries; -extern struct mpc_config_intsrc mp_irqs[MAX_IRQ_SOURCES]; extern int mpc_default_type; extern unsigned long mp_lapic_addr; extern void find_smp_config(void); extern void get_smp_config(void); +void __cpuinit generic_processor_info(int apicid, int version); #ifdef CONFIG_ACPI -extern void mp_register_lapic(u8 id, u8 enabled); -extern void mp_register_lapic_address(u64 address); -extern void mp_register_ioapic(u8 id, u32 address, u32 gsi_base); +extern void mp_register_ioapic(int id, u32 address, u32 gsi_base); extern void mp_override_legacy_irq(u8 bus_irq, u8 polarity, u8 trigger, u32 gsi); extern void mp_config_acpi_legacy_irqs(void); @@ -53,8 +53,7 @@ extern int mp_register_gsi(u32 gsi, int edge_level, int active_high_low); #define PHYSID_ARRAY_SIZE BITS_TO_LONGS(MAX_APICS) -struct physid_mask -{ +struct physid_mask { unsigned long mask[PHYSID_ARRAY_SIZE]; }; @@ -63,34 +62,34 @@ typedef struct physid_mask physid_mask_t; #define physid_set(physid, map) set_bit(physid, (map).mask) #define physid_clear(physid, map) clear_bit(physid, (map).mask) #define physid_isset(physid, map) test_bit(physid, (map).mask) -#define physid_test_and_set(physid, map) \ +#define physid_test_and_set(physid, map) \ test_and_set_bit(physid, (map).mask) -#define physids_and(dst, src1, src2) \ +#define physids_and(dst, src1, src2) \ bitmap_and((dst).mask, (src1).mask, (src2).mask, MAX_APICS) -#define physids_or(dst, src1, src2) \ +#define physids_or(dst, src1, src2) \ bitmap_or((dst).mask, (src1).mask, (src2).mask, MAX_APICS) -#define physids_clear(map) \ +#define physids_clear(map) \ bitmap_zero((map).mask, MAX_APICS) -#define physids_complement(dst, src) \ +#define physids_complement(dst, src) \ bitmap_complement((dst).mask, (src).mask, MAX_APICS) -#define physids_empty(map) \ +#define physids_empty(map) \ bitmap_empty((map).mask, MAX_APICS) -#define physids_equal(map1, map2) \ +#define physids_equal(map1, map2) \ bitmap_equal((map1).mask, (map2).mask, MAX_APICS) -#define physids_weight(map) \ +#define physids_weight(map) \ bitmap_weight((map).mask, MAX_APICS) -#define physids_shift_right(d, s, n) \ +#define physids_shift_right(d, s, n) \ bitmap_shift_right((d).mask, (s).mask, n, MAX_APICS) -#define physids_shift_left(d, s, n) \ +#define physids_shift_left(d, s, n) \ bitmap_shift_left((d).mask, (s).mask, n, MAX_APICS) #define physids_coerce(map) ((map).mask[0]) diff --git a/include/asm-x86/mpspec_def.h b/include/asm-x86/mpspec_def.h index 3504617fe648..dc6ef85e3624 100644 --- a/include/asm-x86/mpspec_def.h +++ b/include/asm-x86/mpspec_def.h @@ -11,7 +11,7 @@ * information is. */ -#define SMP_MAGIC_IDENT (('_'<<24)|('P'<<16)|('M'<<8)|'_') +#define SMP_MAGIC_IDENT (('_'<<24) | ('P'<<16) | ('M'<<8) | '_') #ifdef CONFIG_X86_32 # define MAX_MPC_ENTRY 1024 @@ -23,8 +23,7 @@ # define MAX_APICS 255 #endif -struct intel_mp_floating -{ +struct intel_mp_floating { char mpf_signature[4]; /* "_MP_" */ unsigned int mpf_physptr; /* Configuration table address */ unsigned char mpf_length; /* Our length (paragraphs) */ @@ -39,14 +38,13 @@ struct intel_mp_floating #define MPC_SIGNATURE "PCMP" -struct mp_config_table -{ +struct mp_config_table { char mpc_signature[4]; unsigned short mpc_length; /* Size of table */ - char mpc_spec; /* 0x01 */ - char mpc_checksum; - char mpc_oem[8]; - char mpc_productid[12]; + char mpc_spec; /* 0x01 */ + char mpc_checksum; + char mpc_oem[8]; + char mpc_productid[12]; unsigned int mpc_oemptr; /* 0 if not present */ unsigned short mpc_oemsize; /* 0 if not present */ unsigned short mpc_oemcount; @@ -71,8 +69,7 @@ struct mp_config_table #define CPU_MODEL_MASK 0x00F0 #define CPU_FAMILY_MASK 0x0F00 -struct mpc_config_processor -{ +struct mpc_config_processor { unsigned char mpc_type; unsigned char mpc_apicid; /* Local APIC number */ unsigned char mpc_apicver; /* Its versions */ @@ -82,8 +79,7 @@ struct mpc_config_processor unsigned int mpc_reserved[2]; }; -struct mpc_config_bus -{ +struct mpc_config_bus { unsigned char mpc_type; unsigned char mpc_busid; unsigned char mpc_bustype[6]; @@ -111,8 +107,7 @@ struct mpc_config_bus #define MPC_APIC_USABLE 0x01 -struct mpc_config_ioapic -{ +struct mpc_config_ioapic { unsigned char mpc_type; unsigned char mpc_apicid; unsigned char mpc_apicver; @@ -120,8 +115,7 @@ struct mpc_config_ioapic unsigned int mpc_apicaddr; }; -struct mpc_config_intsrc -{ +struct mpc_config_intsrc { unsigned char mpc_type; unsigned char mpc_irqtype; unsigned short mpc_irqflag; @@ -144,8 +138,7 @@ enum mp_irq_source_types { #define MP_APIC_ALL 0xFF -struct mpc_config_lintsrc -{ +struct mpc_config_lintsrc { unsigned char mpc_type; unsigned char mpc_irqtype; unsigned short mpc_irqflag; @@ -157,8 +150,7 @@ struct mpc_config_lintsrc #define MPC_OEM_SIGNATURE "_OEM" -struct mp_config_oemtable -{ +struct mp_config_oemtable { char oem_signature[4]; unsigned short oem_length; /* Size of table */ char oem_rev; /* 0x01 */ @@ -166,17 +158,6 @@ struct mp_config_oemtable char mpc_oem[8]; }; -struct mpc_config_translation -{ - unsigned char mpc_type; - unsigned char trans_len; - unsigned char trans_type; - unsigned char trans_quad; - unsigned char trans_global; - unsigned char trans_local; - unsigned short trans_reserved; -}; - /* * Default configurations * @@ -196,4 +177,3 @@ enum mp_bustype { MP_BUS_MCA, }; #endif - diff --git a/include/asm-x86/msidef.h b/include/asm-x86/msidef.h index 5b8acddb70fb..296f29ce426d 100644 --- a/include/asm-x86/msidef.h +++ b/include/asm-x86/msidef.h @@ -11,7 +11,8 @@ #define MSI_DATA_VECTOR_SHIFT 0 #define MSI_DATA_VECTOR_MASK 0x000000ff -#define MSI_DATA_VECTOR(v) (((v) << MSI_DATA_VECTOR_SHIFT) & MSI_DATA_VECTOR_MASK) +#define MSI_DATA_VECTOR(v) (((v) << MSI_DATA_VECTOR_SHIFT) & \ + MSI_DATA_VECTOR_MASK) #define MSI_DATA_DELIVERY_MODE_SHIFT 8 #define MSI_DATA_DELIVERY_FIXED (0 << MSI_DATA_DELIVERY_MODE_SHIFT) @@ -37,11 +38,14 @@ #define MSI_ADDR_DEST_MODE_LOGICAL (1 << MSI_ADDR_DEST_MODE_SHIFT) #define MSI_ADDR_REDIRECTION_SHIFT 3 -#define MSI_ADDR_REDIRECTION_CPU (0 << MSI_ADDR_REDIRECTION_SHIFT) /* dedicated cpu */ -#define MSI_ADDR_REDIRECTION_LOWPRI (1 << MSI_ADDR_REDIRECTION_SHIFT) /* lowest priority */ +#define MSI_ADDR_REDIRECTION_CPU (0 << MSI_ADDR_REDIRECTION_SHIFT) + /* dedicated cpu */ +#define MSI_ADDR_REDIRECTION_LOWPRI (1 << MSI_ADDR_REDIRECTION_SHIFT) + /* lowest priority */ #define MSI_ADDR_DEST_ID_SHIFT 12 #define MSI_ADDR_DEST_ID_MASK 0x00ffff0 -#define MSI_ADDR_DEST_ID(dest) (((dest) << MSI_ADDR_DEST_ID_SHIFT) & MSI_ADDR_DEST_ID_MASK) +#define MSI_ADDR_DEST_ID(dest) (((dest) << MSI_ADDR_DEST_ID_SHIFT) & \ + MSI_ADDR_DEST_ID_MASK) #endif /* ASM_MSIDEF_H */ diff --git a/include/asm-x86/msr-index.h b/include/asm-x86/msr-index.h index fae118a25278..09413ad39d3c 100644 --- a/include/asm-x86/msr-index.h +++ b/include/asm-x86/msr-index.h @@ -57,6 +57,8 @@ #define MSR_MTRRfix4K_F8000 0x0000026f #define MSR_MTRRdefType 0x000002ff +#define MSR_IA32_CR_PAT 0x00000277 + #define MSR_IA32_DEBUGCTLMSR 0x000001d9 #define MSR_IA32_LASTBRANCHFROMIP 0x000001db #define MSR_IA32_LASTBRANCHTOIP 0x000001dc @@ -83,6 +85,7 @@ /* AMD64 MSRs. Not complete. See the architecture manual for a more complete list. */ +#define MSR_AMD64_NB_CFG 0xc001001f #define MSR_AMD64_IBSFETCHCTL 0xc0011030 #define MSR_AMD64_IBSFETCHLINAD 0xc0011031 #define MSR_AMD64_IBSFETCHPHYSAD 0xc0011032 @@ -109,6 +112,7 @@ #define MSR_K8_SYSCFG 0xc0010010 #define MSR_K8_HWCR 0xc0010015 #define MSR_K8_ENABLE_C1E 0xc0010055 +#define MSR_K8_TSEG_ADDR 0xc0010112 #define K8_MTRRFIXRANGE_DRAM_ENABLE 0x00040000 /* MtrrFixDramEn bit */ #define K8_MTRRFIXRANGE_DRAM_MODIFY 0x00080000 /* MtrrFixDramModEn bit */ #define K8_MTRR_RDMEM_WRMEM_MASK 0x18181818 /* Mask: RdMem|WrMem */ diff --git a/include/asm-x86/msr.h b/include/asm-x86/msr.h index 3ca29ebebbb1..3707650a169b 100644 --- a/include/asm-x86/msr.h +++ b/include/asm-x86/msr.h @@ -16,8 +16,8 @@ static inline unsigned long long native_read_tscp(unsigned int *aux) { unsigned long low, high; - asm volatile (".byte 0x0f,0x01,0xf9" - : "=a" (low), "=d" (high), "=c" (*aux)); + asm volatile(".byte 0x0f,0x01,0xf9" + : "=a" (low), "=d" (high), "=c" (*aux)); return low | ((u64)high >> 32); } @@ -29,7 +29,7 @@ static inline unsigned long long native_read_tscp(unsigned int *aux) */ #ifdef CONFIG_X86_64 #define DECLARE_ARGS(val, low, high) unsigned low, high -#define EAX_EDX_VAL(val, low, high) (low | ((u64)(high) << 32)) +#define EAX_EDX_VAL(val, low, high) ((low) | ((u64)(high) << 32)) #define EAX_EDX_ARGS(val, low, high) "a" (low), "d" (high) #define EAX_EDX_RET(val, low, high) "=a" (low), "=d" (high) #else @@ -57,7 +57,7 @@ static inline unsigned long long native_read_msr_safe(unsigned int msr, ".section .fixup,\"ax\"\n\t" "3: mov %3,%0 ; jmp 1b\n\t" ".previous\n\t" - _ASM_EXTABLE(2b,3b) + _ASM_EXTABLE(2b, 3b) : "=r" (*err), EAX_EDX_RET(val, low, high) : "c" (msr), "i" (-EFAULT)); return EAX_EDX_VAL(val, low, high); @@ -78,10 +78,10 @@ static inline int native_write_msr_safe(unsigned int msr, ".section .fixup,\"ax\"\n\t" "3: mov %4,%0 ; jmp 1b\n\t" ".previous\n\t" - _ASM_EXTABLE(2b,3b) + _ASM_EXTABLE(2b, 3b) : "=a" (err) : "c" (msr), "0" (low), "d" (high), - "i" (-EFAULT)); + "i" (-EFAULT)); return err; } @@ -116,23 +116,23 @@ static inline unsigned long long native_read_pmc(int counter) * pointer indirection), this allows gcc to optimize better */ -#define rdmsr(msr,val1,val2) \ - do { \ - u64 __val = native_read_msr(msr); \ - (val1) = (u32)__val; \ - (val2) = (u32)(__val >> 32); \ - } while(0) +#define rdmsr(msr, val1, val2) \ +do { \ + u64 __val = native_read_msr((msr)); \ + (val1) = (u32)__val; \ + (val2) = (u32)(__val >> 32); \ +} while (0) static inline void wrmsr(unsigned msr, unsigned low, unsigned high) { native_write_msr(msr, low, high); } -#define rdmsrl(msr,val) \ - ((val) = native_read_msr(msr)) +#define rdmsrl(msr, val) \ + ((val) = native_read_msr((msr))) #define wrmsrl(msr, val) \ - native_write_msr(msr, (u32)((u64)(val)), (u32)((u64)(val) >> 32)) + native_write_msr((msr), (u32)((u64)(val)), (u32)((u64)(val) >> 32)) /* wrmsr with exception handling */ static inline int wrmsr_safe(unsigned msr, unsigned low, unsigned high) @@ -141,14 +141,22 @@ static inline int wrmsr_safe(unsigned msr, unsigned low, unsigned high) } /* rdmsr with exception handling */ -#define rdmsr_safe(msr,p1,p2) \ - ({ \ - int __err; \ - u64 __val = native_read_msr_safe(msr, &__err); \ - (*p1) = (u32)__val; \ - (*p2) = (u32)(__val >> 32); \ - __err; \ - }) +#define rdmsr_safe(msr, p1, p2) \ +({ \ + int __err; \ + u64 __val = native_read_msr_safe((msr), &__err); \ + (*p1) = (u32)__val; \ + (*p2) = (u32)(__val >> 32); \ + __err; \ +}) + +static inline int rdmsrl_safe(unsigned msr, unsigned long long *p) +{ + int err; + + *p = native_read_msr_safe(msr, &err); + return err; +} #define rdtscl(low) \ ((low) = (u32)native_read_tsc()) @@ -156,35 +164,37 @@ static inline int wrmsr_safe(unsigned msr, unsigned low, unsigned high) #define rdtscll(val) \ ((val) = native_read_tsc()) -#define rdpmc(counter,low,high) \ - do { \ - u64 _l = native_read_pmc(counter); \ - (low) = (u32)_l; \ - (high) = (u32)(_l >> 32); \ - } while(0) +#define rdpmc(counter, low, high) \ +do { \ + u64 _l = native_read_pmc((counter)); \ + (low) = (u32)_l; \ + (high) = (u32)(_l >> 32); \ +} while (0) -#define rdtscp(low, high, aux) \ - do { \ - unsigned long long _val = native_read_tscp(&(aux)); \ - (low) = (u32)_val; \ - (high) = (u32)(_val >> 32); \ - } while (0) +#define rdtscp(low, high, aux) \ +do { \ + unsigned long long _val = native_read_tscp(&(aux)); \ + (low) = (u32)_val; \ + (high) = (u32)(_val >> 32); \ +} while (0) #define rdtscpll(val, aux) (val) = native_read_tscp(&(aux)) #endif /* !CONFIG_PARAVIRT */ -#define checking_wrmsrl(msr,val) wrmsr_safe(msr,(u32)(val),(u32)((val)>>32)) +#define checking_wrmsrl(msr, val) wrmsr_safe((msr), (u32)(val), \ + (u32)((val) >> 32)) -#define write_tsc(val1,val2) wrmsr(0x10, val1, val2) +#define write_tsc(val1, val2) wrmsr(0x10, (val1), (val2)) -#define write_rdtscp_aux(val) wrmsr(0xc0000103, val, 0) +#define write_rdtscp_aux(val) wrmsr(0xc0000103, (val), 0) #ifdef CONFIG_SMP void rdmsr_on_cpu(unsigned int cpu, u32 msr_no, u32 *l, u32 *h); void wrmsr_on_cpu(unsigned int cpu, u32 msr_no, u32 l, u32 h); int rdmsr_safe_on_cpu(unsigned int cpu, u32 msr_no, u32 *l, u32 *h); + int wrmsr_safe_on_cpu(unsigned int cpu, u32 msr_no, u32 l, u32 h); #else /* CONFIG_SMP */ static inline void rdmsr_on_cpu(unsigned int cpu, u32 msr_no, u32 *l, u32 *h) @@ -195,7 +205,8 @@ static inline void wrmsr_on_cpu(unsigned int cpu, u32 msr_no, u32 l, u32 h) { wrmsr(msr_no, l, h); } -static inline int rdmsr_safe_on_cpu(unsigned int cpu, u32 msr_no, u32 *l, u32 *h) +static inline int rdmsr_safe_on_cpu(unsigned int cpu, u32 msr_no, + u32 *l, u32 *h) { return rdmsr_safe(msr_no, l, h); } diff --git a/include/asm-x86/mtrr.h b/include/asm-x86/mtrr.h index 319d065800be..a69a01a51729 100644 --- a/include/asm-x86/mtrr.h +++ b/include/asm-x86/mtrr.h @@ -28,8 +28,7 @@ #define MTRR_IOCTL_BASE 'M' -struct mtrr_sentry -{ +struct mtrr_sentry { unsigned long base; /* Base address */ unsigned int size; /* Size of region */ unsigned int type; /* Type of region */ @@ -41,8 +40,7 @@ struct mtrr_sentry will break. */ #ifdef __i386__ -struct mtrr_gentry -{ +struct mtrr_gentry { unsigned int regnum; /* Register number */ unsigned long base; /* Base address */ unsigned int size; /* Size of region */ @@ -51,8 +49,7 @@ struct mtrr_gentry #else /* __i386__ */ -struct mtrr_gentry -{ +struct mtrr_gentry { unsigned long base; /* Base address */ unsigned int size; /* Size of region */ unsigned int regnum; /* Register number */ @@ -86,38 +83,45 @@ struct mtrr_gentry /* The following functions are for use by other drivers */ # ifdef CONFIG_MTRR +extern u8 mtrr_type_lookup(u64 addr, u64 end); extern void mtrr_save_fixed_ranges(void *); extern void mtrr_save_state(void); -extern int mtrr_add (unsigned long base, unsigned long size, - unsigned int type, bool increment); -extern int mtrr_add_page (unsigned long base, unsigned long size, - unsigned int type, bool increment); -extern int mtrr_del (int reg, unsigned long base, unsigned long size); -extern int mtrr_del_page (int reg, unsigned long base, unsigned long size); +extern int mtrr_add(unsigned long base, unsigned long size, + unsigned int type, bool increment); +extern int mtrr_add_page(unsigned long base, unsigned long size, + unsigned int type, bool increment); +extern int mtrr_del(int reg, unsigned long base, unsigned long size); +extern int mtrr_del_page(int reg, unsigned long base, unsigned long size); extern void mtrr_centaur_report_mcr(int mcr, u32 lo, u32 hi); extern void mtrr_ap_init(void); extern void mtrr_bp_init(void); extern int mtrr_trim_uncached_memory(unsigned long end_pfn); +extern int amd_special_default_mtrr(void); # else +static inline u8 mtrr_type_lookup(u64 addr, u64 end) +{ + /* + * Return no-MTRRs: + */ + return 0xff; +} #define mtrr_save_fixed_ranges(arg) do {} while (0) #define mtrr_save_state() do {} while (0) -static __inline__ int mtrr_add (unsigned long base, unsigned long size, - unsigned int type, bool increment) +static inline int mtrr_add(unsigned long base, unsigned long size, + unsigned int type, bool increment) { return -ENODEV; } -static __inline__ int mtrr_add_page (unsigned long base, unsigned long size, +static inline int mtrr_add_page(unsigned long base, unsigned long size, unsigned int type, bool increment) { return -ENODEV; } -static __inline__ int mtrr_del (int reg, unsigned long base, - unsigned long size) +static inline int mtrr_del(int reg, unsigned long base, unsigned long size) { return -ENODEV; } -static __inline__ int mtrr_del_page (int reg, unsigned long base, - unsigned long size) +static inline int mtrr_del_page(int reg, unsigned long base, unsigned long size) { return -ENODEV; } @@ -125,7 +129,9 @@ static inline int mtrr_trim_uncached_memory(unsigned long end_pfn) { return 0; } -static __inline__ void mtrr_centaur_report_mcr(int mcr, u32 lo, u32 hi) {;} +static inline void mtrr_centaur_report_mcr(int mcr, u32 lo, u32 hi) +{ +} #define mtrr_ap_init() do {} while (0) #define mtrr_bp_init() do {} while (0) @@ -134,15 +140,13 @@ static __inline__ void mtrr_centaur_report_mcr(int mcr, u32 lo, u32 hi) {;} #ifdef CONFIG_COMPAT #include <linux/compat.h> -struct mtrr_sentry32 -{ +struct mtrr_sentry32 { compat_ulong_t base; /* Base address */ compat_uint_t size; /* Size of region */ compat_uint_t type; /* Type of region */ }; -struct mtrr_gentry32 -{ +struct mtrr_gentry32 { compat_ulong_t regnum; /* Register number */ compat_uint_t base; /* Base address */ compat_uint_t size; /* Size of region */ @@ -151,16 +155,17 @@ struct mtrr_gentry32 #define MTRR_IOCTL_BASE 'M' -#define MTRRIOC32_ADD_ENTRY _IOW(MTRR_IOCTL_BASE, 0, struct mtrr_sentry32) -#define MTRRIOC32_SET_ENTRY _IOW(MTRR_IOCTL_BASE, 1, struct mtrr_sentry32) -#define MTRRIOC32_DEL_ENTRY _IOW(MTRR_IOCTL_BASE, 2, struct mtrr_sentry32) -#define MTRRIOC32_GET_ENTRY _IOWR(MTRR_IOCTL_BASE, 3, struct mtrr_gentry32) -#define MTRRIOC32_KILL_ENTRY _IOW(MTRR_IOCTL_BASE, 4, struct mtrr_sentry32) -#define MTRRIOC32_ADD_PAGE_ENTRY _IOW(MTRR_IOCTL_BASE, 5, struct mtrr_sentry32) -#define MTRRIOC32_SET_PAGE_ENTRY _IOW(MTRR_IOCTL_BASE, 6, struct mtrr_sentry32) -#define MTRRIOC32_DEL_PAGE_ENTRY _IOW(MTRR_IOCTL_BASE, 7, struct mtrr_sentry32) -#define MTRRIOC32_GET_PAGE_ENTRY _IOWR(MTRR_IOCTL_BASE, 8, struct mtrr_gentry32) -#define MTRRIOC32_KILL_PAGE_ENTRY _IOW(MTRR_IOCTL_BASE, 9, struct mtrr_sentry32) +#define MTRRIOC32_ADD_ENTRY _IOW(MTRR_IOCTL_BASE, 0, struct mtrr_sentry32) +#define MTRRIOC32_SET_ENTRY _IOW(MTRR_IOCTL_BASE, 1, struct mtrr_sentry32) +#define MTRRIOC32_DEL_ENTRY _IOW(MTRR_IOCTL_BASE, 2, struct mtrr_sentry32) +#define MTRRIOC32_GET_ENTRY _IOWR(MTRR_IOCTL_BASE, 3, struct mtrr_gentry32) +#define MTRRIOC32_KILL_ENTRY _IOW(MTRR_IOCTL_BASE, 4, struct mtrr_sentry32) +#define MTRRIOC32_ADD_PAGE_ENTRY _IOW(MTRR_IOCTL_BASE, 5, struct mtrr_sentry32) +#define MTRRIOC32_SET_PAGE_ENTRY _IOW(MTRR_IOCTL_BASE, 6, struct mtrr_sentry32) +#define MTRRIOC32_DEL_PAGE_ENTRY _IOW(MTRR_IOCTL_BASE, 7, struct mtrr_sentry32) +#define MTRRIOC32_GET_PAGE_ENTRY _IOWR(MTRR_IOCTL_BASE, 8, struct mtrr_gentry32) +#define MTRRIOC32_KILL_PAGE_ENTRY \ + _IOW(MTRR_IOCTL_BASE, 9, struct mtrr_sentry32) #endif /* CONFIG_COMPAT */ #endif /* __KERNEL__ */ diff --git a/include/asm-x86/mutex_32.h b/include/asm-x86/mutex_32.h index bbeefb96ddfd..73e928ef5f03 100644 --- a/include/asm-x86/mutex_32.h +++ b/include/asm-x86/mutex_32.h @@ -9,7 +9,7 @@ #ifndef _ASM_MUTEX_H #define _ASM_MUTEX_H -#include "asm/alternative.h" +#include <asm/alternative.h> /** * __mutex_fastpath_lock - try to take the lock by moving the count @@ -21,22 +21,20 @@ * wasn't 1 originally. This function MUST leave the value lower than 1 * even when the "1" assertion wasn't true. */ -#define __mutex_fastpath_lock(count, fail_fn) \ -do { \ - unsigned int dummy; \ - \ - typecheck(atomic_t *, count); \ +#define __mutex_fastpath_lock(count, fail_fn) \ +do { \ + unsigned int dummy; \ + \ + typecheck(atomic_t *, count); \ typecheck_fn(void (*)(atomic_t *), fail_fn); \ - \ - __asm__ __volatile__( \ - LOCK_PREFIX " decl (%%eax) \n" \ - " jns 1f \n" \ - " call "#fail_fn" \n" \ - "1: \n" \ - \ - :"=a" (dummy) \ - : "a" (count) \ - : "memory", "ecx", "edx"); \ + \ + asm volatile(LOCK_PREFIX " decl (%%eax)\n" \ + " jns 1f \n" \ + " call " #fail_fn "\n" \ + "1:\n" \ + : "=a" (dummy) \ + : "a" (count) \ + : "memory", "ecx", "edx"); \ } while (0) @@ -50,8 +48,8 @@ do { \ * wasn't 1 originally. This function returns 0 if the fastpath succeeds, * or anything the slow path function returns */ -static inline int -__mutex_fastpath_lock_retval(atomic_t *count, int (*fail_fn)(atomic_t *)) +static inline int __mutex_fastpath_lock_retval(atomic_t *count, + int (*fail_fn)(atomic_t *)) { if (unlikely(atomic_dec_return(count) < 0)) return fail_fn(count); @@ -72,22 +70,20 @@ __mutex_fastpath_lock_retval(atomic_t *count, int (*fail_fn)(atomic_t *)) * __mutex_slowpath_needs_to_unlock() macro needs to return 1, it needs * to return 0 otherwise. */ -#define __mutex_fastpath_unlock(count, fail_fn) \ -do { \ - unsigned int dummy; \ - \ - typecheck(atomic_t *, count); \ +#define __mutex_fastpath_unlock(count, fail_fn) \ +do { \ + unsigned int dummy; \ + \ + typecheck(atomic_t *, count); \ typecheck_fn(void (*)(atomic_t *), fail_fn); \ - \ - __asm__ __volatile__( \ - LOCK_PREFIX " incl (%%eax) \n" \ - " jg 1f \n" \ - " call "#fail_fn" \n" \ - "1: \n" \ - \ - :"=a" (dummy) \ - : "a" (count) \ - : "memory", "ecx", "edx"); \ + \ + asm volatile(LOCK_PREFIX " incl (%%eax)\n" \ + " jg 1f\n" \ + " call " #fail_fn "\n" \ + "1:\n" \ + : "=a" (dummy) \ + : "a" (count) \ + : "memory", "ecx", "edx"); \ } while (0) #define __mutex_slowpath_needs_to_unlock() 1 @@ -104,8 +100,8 @@ do { \ * Additionally, if the value was < 0 originally, this function must not leave * it to 0 on failure. */ -static inline int -__mutex_fastpath_trylock(atomic_t *count, int (*fail_fn)(atomic_t *)) +static inline int __mutex_fastpath_trylock(atomic_t *count, + int (*fail_fn)(atomic_t *)) { /* * We have two variants here. The cmpxchg based one is the best one diff --git a/include/asm-x86/mutex_64.h b/include/asm-x86/mutex_64.h index 6c2949a3c677..f3fae9becb38 100644 --- a/include/asm-x86/mutex_64.h +++ b/include/asm-x86/mutex_64.h @@ -16,23 +16,21 @@ * * Atomically decrements @v and calls <fail_fn> if the result is negative. */ -#define __mutex_fastpath_lock(v, fail_fn) \ -do { \ - unsigned long dummy; \ - \ - typecheck(atomic_t *, v); \ - typecheck_fn(void (*)(atomic_t *), fail_fn); \ - \ - __asm__ __volatile__( \ - LOCK_PREFIX " decl (%%rdi) \n" \ - " jns 1f \n" \ - " call "#fail_fn" \n" \ - "1:" \ - \ - :"=D" (dummy) \ - : "D" (v) \ - : "rax", "rsi", "rdx", "rcx", \ - "r8", "r9", "r10", "r11", "memory"); \ +#define __mutex_fastpath_lock(v, fail_fn) \ +do { \ + unsigned long dummy; \ + \ + typecheck(atomic_t *, v); \ + typecheck_fn(void (*)(atomic_t *), fail_fn); \ + \ + asm volatile(LOCK_PREFIX " decl (%%rdi)\n" \ + " jns 1f \n" \ + " call " #fail_fn "\n" \ + "1:" \ + : "=D" (dummy) \ + : "D" (v) \ + : "rax", "rsi", "rdx", "rcx", \ + "r8", "r9", "r10", "r11", "memory"); \ } while (0) /** @@ -45,9 +43,8 @@ do { \ * it wasn't 1 originally. This function returns 0 if the fastpath succeeds, * or anything the slow path function returns */ -static inline int -__mutex_fastpath_lock_retval(atomic_t *count, - int (*fail_fn)(atomic_t *)) +static inline int __mutex_fastpath_lock_retval(atomic_t *count, + int (*fail_fn)(atomic_t *)) { if (unlikely(atomic_dec_return(count) < 0)) return fail_fn(count); @@ -62,23 +59,21 @@ __mutex_fastpath_lock_retval(atomic_t *count, * * Atomically increments @v and calls <fail_fn> if the result is nonpositive. */ -#define __mutex_fastpath_unlock(v, fail_fn) \ -do { \ - unsigned long dummy; \ - \ - typecheck(atomic_t *, v); \ - typecheck_fn(void (*)(atomic_t *), fail_fn); \ - \ - __asm__ __volatile__( \ - LOCK_PREFIX " incl (%%rdi) \n" \ - " jg 1f \n" \ - " call "#fail_fn" \n" \ - "1: " \ - \ - :"=D" (dummy) \ - : "D" (v) \ - : "rax", "rsi", "rdx", "rcx", \ - "r8", "r9", "r10", "r11", "memory"); \ +#define __mutex_fastpath_unlock(v, fail_fn) \ +do { \ + unsigned long dummy; \ + \ + typecheck(atomic_t *, v); \ + typecheck_fn(void (*)(atomic_t *), fail_fn); \ + \ + asm volatile(LOCK_PREFIX " incl (%%rdi)\n" \ + " jg 1f\n" \ + " call " #fail_fn "\n" \ + "1:" \ + : "=D" (dummy) \ + : "D" (v) \ + : "rax", "rsi", "rdx", "rcx", \ + "r8", "r9", "r10", "r11", "memory"); \ } while (0) #define __mutex_slowpath_needs_to_unlock() 1 @@ -93,8 +88,8 @@ do { \ * if it wasn't 1 originally. [the fallback function is never used on * x86_64, because all x86_64 CPUs have a CMPXCHG instruction.] */ -static inline int -__mutex_fastpath_trylock(atomic_t *count, int (*fail_fn)(atomic_t *)) +static inline int __mutex_fastpath_trylock(atomic_t *count, + int (*fail_fn)(atomic_t *)) { if (likely(atomic_cmpxchg(count, 1, 0) == 1)) return 1; diff --git a/include/asm-x86/nmi.h b/include/asm-x86/nmi.h index 53ccac14cead..1e363021e72f 100644 --- a/include/asm-x86/nmi.h +++ b/include/asm-x86/nmi.h @@ -1,5 +1,93 @@ -#ifdef CONFIG_X86_32 -# include "nmi_32.h" +#ifndef _ASM_X86_NMI_H_ +#define _ASM_X86_NMI_H_ + +#include <linux/pm.h> +#include <asm/irq.h> +#include <asm/io.h> + +#ifdef ARCH_HAS_NMI_WATCHDOG + +/** + * do_nmi_callback + * + * Check to see if a callback exists and execute it. Return 1 + * if the handler exists and was handled successfully. + */ +int do_nmi_callback(struct pt_regs *regs, int cpu); + +#ifdef CONFIG_PM + +/** Replace the PM callback routine for NMI. */ +struct pm_dev *set_nmi_pm_callback(pm_callback callback); + +/** Unset the PM callback routine back to the default. */ +void unset_nmi_pm_callback(struct pm_dev *dev); + #else -# include "nmi_64.h" + +static inline struct pm_dev *set_nmi_pm_callback(pm_callback callback) +{ + return 0; +} + +static inline void unset_nmi_pm_callback(struct pm_dev *dev) +{ +} + +#endif /* CONFIG_PM */ + +#ifdef CONFIG_X86_64 +extern void default_do_nmi(struct pt_regs *); +extern void die_nmi(char *str, struct pt_regs *regs, int do_panic); +extern void nmi_watchdog_default(void); +#else +#define nmi_watchdog_default() do {} while (0) +#endif + +extern int check_nmi_watchdog(void); +extern int nmi_watchdog_enabled; +extern int unknown_nmi_panic; +extern int avail_to_resrv_perfctr_nmi_bit(unsigned int); +extern int avail_to_resrv_perfctr_nmi(unsigned int); +extern int reserve_perfctr_nmi(unsigned int); +extern void release_perfctr_nmi(unsigned int); +extern int reserve_evntsel_nmi(unsigned int); +extern void release_evntsel_nmi(unsigned int); + +extern void setup_apic_nmi_watchdog(void *); +extern void stop_apic_nmi_watchdog(void *); +extern void disable_timer_nmi_watchdog(void); +extern void enable_timer_nmi_watchdog(void); +extern int nmi_watchdog_tick(struct pt_regs *regs, unsigned reason); + +extern atomic_t nmi_active; +extern unsigned int nmi_watchdog; +#define NMI_DISABLED -1 +#define NMI_NONE 0 +#define NMI_IO_APIC 1 +#define NMI_LOCAL_APIC 2 +#define NMI_INVALID 3 +#define NMI_DEFAULT NMI_DISABLED + +struct ctl_table; +struct file; +extern int proc_nmi_enabled(struct ctl_table *, int , struct file *, + void __user *, size_t *, loff_t *); +extern int unknown_nmi_panic; + +void __trigger_all_cpu_backtrace(void); +#define trigger_all_cpu_backtrace() __trigger_all_cpu_backtrace() + +#endif + +void lapic_watchdog_stop(void); +int lapic_watchdog_init(unsigned nmi_hz); +int lapic_wd_event(unsigned nmi_hz); +unsigned lapic_adjust_nmi_hz(unsigned hz); +int lapic_watchdog_ok(void); +void disable_lapic_nmi_watchdog(void); +void enable_lapic_nmi_watchdog(void); +void stop_nmi(void); +void restart_nmi(void); + #endif diff --git a/include/asm-x86/nmi_32.h b/include/asm-x86/nmi_32.h deleted file mode 100644 index 7206c7e8a388..000000000000 --- a/include/asm-x86/nmi_32.h +++ /dev/null @@ -1,61 +0,0 @@ -#ifndef ASM_NMI_H -#define ASM_NMI_H - -#include <linux/pm.h> -#include <asm/irq.h> - -#ifdef ARCH_HAS_NMI_WATCHDOG - -/** - * do_nmi_callback - * - * Check to see if a callback exists and execute it. Return 1 - * if the handler exists and was handled successfully. - */ -int do_nmi_callback(struct pt_regs *regs, int cpu); - -extern int nmi_watchdog_enabled; -extern int avail_to_resrv_perfctr_nmi_bit(unsigned int); -extern int avail_to_resrv_perfctr_nmi(unsigned int); -extern int reserve_perfctr_nmi(unsigned int); -extern void release_perfctr_nmi(unsigned int); -extern int reserve_evntsel_nmi(unsigned int); -extern void release_evntsel_nmi(unsigned int); - -extern void setup_apic_nmi_watchdog (void *); -extern void stop_apic_nmi_watchdog (void *); -extern void disable_timer_nmi_watchdog(void); -extern void enable_timer_nmi_watchdog(void); -extern int nmi_watchdog_tick (struct pt_regs * regs, unsigned reason); - -extern atomic_t nmi_active; -extern unsigned int nmi_watchdog; -#define NMI_DISABLED -1 -#define NMI_NONE 0 -#define NMI_IO_APIC 1 -#define NMI_LOCAL_APIC 2 -#define NMI_INVALID 3 -#define NMI_DEFAULT NMI_DISABLED - -struct ctl_table; -struct file; -extern int proc_nmi_enabled(struct ctl_table *, int , struct file *, - void __user *, size_t *, loff_t *); -extern int unknown_nmi_panic; - -void __trigger_all_cpu_backtrace(void); -#define trigger_all_cpu_backtrace() __trigger_all_cpu_backtrace() - -#endif - -void lapic_watchdog_stop(void); -int lapic_watchdog_init(unsigned nmi_hz); -int lapic_wd_event(unsigned nmi_hz); -unsigned lapic_adjust_nmi_hz(unsigned hz); -int lapic_watchdog_ok(void); -void disable_lapic_nmi_watchdog(void); -void enable_lapic_nmi_watchdog(void); -void stop_nmi(void); -void restart_nmi(void); - -#endif /* ASM_NMI_H */ diff --git a/include/asm-x86/nmi_64.h b/include/asm-x86/nmi_64.h deleted file mode 100644 index 2eeb74e5f3ff..000000000000 --- a/include/asm-x86/nmi_64.h +++ /dev/null @@ -1,90 +0,0 @@ -#ifndef ASM_NMI_H -#define ASM_NMI_H - -#include <linux/pm.h> -#include <asm/io.h> - -/** - * do_nmi_callback - * - * Check to see if a callback exists and execute it. Return 1 - * if the handler exists and was handled successfully. - */ -int do_nmi_callback(struct pt_regs *regs, int cpu); - -#ifdef CONFIG_PM - -/** Replace the PM callback routine for NMI. */ -struct pm_dev * set_nmi_pm_callback(pm_callback callback); - -/** Unset the PM callback routine back to the default. */ -void unset_nmi_pm_callback(struct pm_dev * dev); - -#else - -static inline struct pm_dev * set_nmi_pm_callback(pm_callback callback) -{ - return 0; -} - -static inline void unset_nmi_pm_callback(struct pm_dev * dev) -{ -} - -#endif /* CONFIG_PM */ - -extern void default_do_nmi(struct pt_regs *); -extern void die_nmi(char *str, struct pt_regs *regs, int do_panic); - -#define get_nmi_reason() inb(0x61) - -extern int unknown_nmi_panic; -extern int nmi_watchdog_enabled; - -extern int check_nmi_watchdog(void); -extern int avail_to_resrv_perfctr_nmi_bit(unsigned int); -extern int avail_to_resrv_perfctr_nmi(unsigned int); -extern int reserve_perfctr_nmi(unsigned int); -extern void release_perfctr_nmi(unsigned int); -extern int reserve_evntsel_nmi(unsigned int); -extern void release_evntsel_nmi(unsigned int); - -extern void setup_apic_nmi_watchdog (void *); -extern void stop_apic_nmi_watchdog (void *); -extern void disable_timer_nmi_watchdog(void); -extern void enable_timer_nmi_watchdog(void); -extern int nmi_watchdog_tick (struct pt_regs * regs, unsigned reason); - -extern void nmi_watchdog_default(void); - -extern atomic_t nmi_active; -extern unsigned int nmi_watchdog; -#define NMI_DISABLED -1 -#define NMI_NONE 0 -#define NMI_IO_APIC 1 -#define NMI_LOCAL_APIC 2 -#define NMI_INVALID 3 -#define NMI_DEFAULT NMI_DISABLED - -struct ctl_table; -struct file; -extern int proc_nmi_enabled(struct ctl_table *, int , struct file *, - void __user *, size_t *, loff_t *); - -extern int unknown_nmi_panic; - -void __trigger_all_cpu_backtrace(void); -#define trigger_all_cpu_backtrace() __trigger_all_cpu_backtrace() - - -void lapic_watchdog_stop(void); -int lapic_watchdog_init(unsigned nmi_hz); -int lapic_wd_event(unsigned nmi_hz); -unsigned lapic_adjust_nmi_hz(unsigned hz); -int lapic_watchdog_ok(void); -void disable_lapic_nmi_watchdog(void); -void enable_lapic_nmi_watchdog(void); -void stop_nmi(void); -void restart_nmi(void); - -#endif /* ASM_NMI_H */ diff --git a/include/asm-x86/nops.h b/include/asm-x86/nops.h index b3930ae539b3..ad0bedd10b89 100644 --- a/include/asm-x86/nops.h +++ b/include/asm-x86/nops.h @@ -5,6 +5,8 @@ /* generic versions from gas 1: nop + the following instructions are NOT nops in 64-bit mode, + for 64-bit mode use K8 or P6 nops instead 2: movl %esi,%esi 3: leal 0x00(%esi),%esi 4: leal 0x00(,%esi,1),%esi diff --git a/include/asm-x86/numa_64.h b/include/asm-x86/numa_64.h index 15fe07cde586..32c22ae0709f 100644 --- a/include/asm-x86/numa_64.h +++ b/include/asm-x86/numa_64.h @@ -1,11 +1,12 @@ -#ifndef _ASM_X8664_NUMA_H +#ifndef _ASM_X8664_NUMA_H #define _ASM_X8664_NUMA_H 1 #include <linux/nodemask.h> #include <asm/apicdef.h> struct bootnode { - u64 start,end; + u64 start; + u64 end; }; extern int compute_hash_shift(struct bootnode *nodes, int numnodes); diff --git a/include/asm-x86/numaq.h b/include/asm-x86/numaq.h index 38f710dc37f2..94b86c31239a 100644 --- a/include/asm-x86/numaq.h +++ b/include/asm-x86/numaq.h @@ -3,7 +3,7 @@ * * Copyright (C) 2002, IBM Corp. * - * All rights reserved. + * All rights reserved. * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by @@ -33,7 +33,8 @@ extern int get_memcfg_numaq(void); /* * SYS_CFG_DATA_PRIV_ADDR, struct eachquadmem, and struct sys_cfg_data are the */ -#define SYS_CFG_DATA_PRIV_ADDR 0x0009d000 /* place for scd in private quad space */ +#define SYS_CFG_DATA_PRIV_ADDR 0x0009d000 /* place for scd in private + quad space */ /* * Communication area for each processor on lynxer-processor tests. @@ -139,7 +140,7 @@ struct sys_cfg_data { unsigned int low_shrd_mem_base; /* 0 or 512MB or 1GB */ unsigned int low_shrd_mem_quad_offset; /* 0,128M,256M,512M,1G */ /* may not be totally populated */ - unsigned int split_mem_enbl; /* 0 for no low shared memory */ + unsigned int split_mem_enbl; /* 0 for no low shared memory */ unsigned int mmio_sz; /* Size of total system memory mapped I/O */ /* (in MB). */ unsigned int quad_spin_lock; /* Spare location used for quad */ @@ -152,7 +153,7 @@ struct sys_cfg_data { /* * memory configuration area for each quad */ - struct eachquadmem eq[MAX_NUMNODES]; /* indexed by quad id */ + struct eachquadmem eq[MAX_NUMNODES]; /* indexed by quad id */ }; static inline unsigned long *get_zholes_size(int nid) diff --git a/include/asm-x86/page.h b/include/asm-x86/page.h index a05b2896492f..6724a4bc6b7a 100644 --- a/include/asm-x86/page.h +++ b/include/asm-x86/page.h @@ -33,10 +33,8 @@ #ifdef CONFIG_X86_64 #include <asm/page_64.h> -#define max_pfn_mapped end_pfn_map #else #include <asm/page_32.h> -#define max_pfn_mapped max_low_pfn #endif /* CONFIG_X86_64 */ #define PAGE_OFFSET ((unsigned long)__PAGE_OFFSET) @@ -50,6 +48,8 @@ extern int page_is_ram(unsigned long pagenr); +extern unsigned long max_pfn_mapped; + struct page; static inline void clear_user_page(void *page, unsigned long vaddr, diff --git a/include/asm-x86/page_32.h b/include/asm-x86/page_32.h index 5f7257fd589b..424e82f8ae27 100644 --- a/include/asm-x86/page_32.h +++ b/include/asm-x86/page_32.h @@ -47,7 +47,10 @@ typedef unsigned long pgdval_t; typedef unsigned long pgprotval_t; typedef unsigned long phys_addr_t; -typedef union { pteval_t pte, pte_low; } pte_t; +typedef union { + pteval_t pte; + pteval_t pte_low; +} pte_t; #endif /* __ASSEMBLY__ */ #endif /* CONFIG_X86_PAE */ @@ -61,7 +64,7 @@ typedef struct page *pgtable_t; #endif #ifndef __ASSEMBLY__ -#define __phys_addr(x) ((x)-PAGE_OFFSET) +#define __phys_addr(x) ((x) - PAGE_OFFSET) #define __phys_reloc_hide(x) RELOC_HIDE((x), 0) #ifdef CONFIG_FLATMEM @@ -78,7 +81,7 @@ extern unsigned int __VMALLOC_RESERVE; extern int sysctl_legacy_va_layout; #define VMALLOC_RESERVE ((unsigned long)__VMALLOC_RESERVE) -#define MAXMEM (-__PAGE_OFFSET-__VMALLOC_RESERVE) +#define MAXMEM (-__PAGE_OFFSET - __VMALLOC_RESERVE) #ifdef CONFIG_X86_USE_3DNOW #include <asm/mmx.h> diff --git a/include/asm-x86/page_64.h b/include/asm-x86/page_64.h index 143546073b95..6ea72859c491 100644 --- a/include/asm-x86/page_64.h +++ b/include/asm-x86/page_64.h @@ -5,7 +5,7 @@ #define THREAD_ORDER 1 #define THREAD_SIZE (PAGE_SIZE << THREAD_ORDER) -#define CURRENT_MASK (~(THREAD_SIZE-1)) +#define CURRENT_MASK (~(THREAD_SIZE - 1)) #define EXCEPTION_STACK_ORDER 0 #define EXCEPTION_STKSZ (PAGE_SIZE << EXCEPTION_STACK_ORDER) @@ -48,10 +48,10 @@ #define __VIRTUAL_MASK_SHIFT 48 /* - * Kernel image size is limited to 128 MB (see level2_kernel_pgt in + * Kernel image size is limited to 512 MB (see level2_kernel_pgt in * arch/x86/kernel/head_64.S), and it is mapped here: */ -#define KERNEL_IMAGE_SIZE (128*1024*1024) +#define KERNEL_IMAGE_SIZE (512 * 1024 * 1024) #define KERNEL_IMAGE_START _AC(0xffffffff80000000, UL) #ifndef __ASSEMBLY__ @@ -59,7 +59,6 @@ void clear_page(void *page); void copy_page(void *to, void *from); extern unsigned long end_pfn; -extern unsigned long end_pfn_map; extern unsigned long phys_base; extern unsigned long __phys_addr(unsigned long); @@ -81,6 +80,9 @@ typedef struct { pteval_t pte; } pte_t; #define vmemmap ((struct page *)VMEMMAP_START) +extern unsigned long init_memory_mapping(unsigned long start, + unsigned long end); + #endif /* !__ASSEMBLY__ */ #ifdef CONFIG_FLATMEM diff --git a/include/asm-x86/param.h b/include/asm-x86/param.h index c996ec4da0c8..6f0d0422f4ca 100644 --- a/include/asm-x86/param.h +++ b/include/asm-x86/param.h @@ -3,8 +3,8 @@ #ifdef __KERNEL__ # define HZ CONFIG_HZ /* Internal kernel timer frequency */ -# define USER_HZ 100 /* .. some user interfaces are in "ticks" */ -# define CLOCKS_PER_SEC (USER_HZ) /* like times() */ +# define USER_HZ 100 /* some user interfaces are */ +# define CLOCKS_PER_SEC (USER_HZ) /* in "ticks" like times() */ #endif #ifndef HZ diff --git a/include/asm-x86/paravirt.h b/include/asm-x86/paravirt.h index d6236eb46466..3d419398499b 100644 --- a/include/asm-x86/paravirt.h +++ b/include/asm-x86/paravirt.h @@ -231,7 +231,8 @@ struct pv_mmu_ops { void (*set_pte_at)(struct mm_struct *mm, unsigned long addr, pte_t *ptep, pte_t pteval); void (*set_pmd)(pmd_t *pmdp, pmd_t pmdval); - void (*pte_update)(struct mm_struct *mm, unsigned long addr, pte_t *ptep); + void (*pte_update)(struct mm_struct *mm, unsigned long addr, + pte_t *ptep); void (*pte_update_defer)(struct mm_struct *mm, unsigned long addr, pte_t *ptep); @@ -246,7 +247,8 @@ struct pv_mmu_ops { void (*set_pte_atomic)(pte_t *ptep, pte_t pteval); void (*set_pte_present)(struct mm_struct *mm, unsigned long addr, pte_t *ptep, pte_t pte); - void (*pte_clear)(struct mm_struct *mm, unsigned long addr, pte_t *ptep); + void (*pte_clear)(struct mm_struct *mm, unsigned long addr, + pte_t *ptep); void (*pmd_clear)(pmd_t *pmdp); #endif /* CONFIG_X86_PAE */ @@ -274,8 +276,7 @@ struct pv_mmu_ops { /* This contains all the paravirt structures: we get a convenient * number for each function using the offset which we use to indicate * what to patch. */ -struct paravirt_patch_template -{ +struct paravirt_patch_template { struct pv_init_ops pv_init_ops; struct pv_time_ops pv_time_ops; struct pv_cpu_ops pv_cpu_ops; @@ -660,43 +661,56 @@ static inline int paravirt_write_msr(unsigned msr, unsigned low, unsigned high) } /* These should all do BUG_ON(_err), but our headers are too tangled. */ -#define rdmsr(msr,val1,val2) do { \ +#define rdmsr(msr, val1, val2) \ +do { \ int _err; \ u64 _l = paravirt_read_msr(msr, &_err); \ val1 = (u32)_l; \ val2 = _l >> 32; \ -} while(0) +} while (0) -#define wrmsr(msr,val1,val2) do { \ +#define wrmsr(msr, val1, val2) \ +do { \ paravirt_write_msr(msr, val1, val2); \ -} while(0) +} while (0) -#define rdmsrl(msr,val) do { \ +#define rdmsrl(msr, val) \ +do { \ int _err; \ val = paravirt_read_msr(msr, &_err); \ -} while(0) +} while (0) -#define wrmsrl(msr,val) wrmsr(msr, (u32)((u64)(val)), ((u64)(val))>>32) -#define wrmsr_safe(msr,a,b) paravirt_write_msr(msr, a, b) +#define wrmsrl(msr, val) wrmsr(msr, (u32)((u64)(val)), ((u64)(val))>>32) +#define wrmsr_safe(msr, a, b) paravirt_write_msr(msr, a, b) /* rdmsr with exception handling */ -#define rdmsr_safe(msr,a,b) ({ \ +#define rdmsr_safe(msr, a, b) \ +({ \ int _err; \ u64 _l = paravirt_read_msr(msr, &_err); \ (*a) = (u32)_l; \ (*b) = _l >> 32; \ - _err; }) + _err; \ +}) + +static inline int rdmsrl_safe(unsigned msr, unsigned long long *p) +{ + int err; + *p = paravirt_read_msr(msr, &err); + return err; +} static inline u64 paravirt_read_tsc(void) { return PVOP_CALL0(u64, pv_cpu_ops.read_tsc); } -#define rdtscl(low) do { \ +#define rdtscl(low) \ +do { \ u64 _l = paravirt_read_tsc(); \ low = (int)_l; \ -} while(0) +} while (0) #define rdtscll(val) (val = paravirt_read_tsc()) @@ -711,11 +725,12 @@ static inline unsigned long long paravirt_read_pmc(int counter) return PVOP_CALL1(u64, pv_cpu_ops.read_pmc, counter); } -#define rdpmc(counter,low,high) do { \ +#define rdpmc(counter, low, high) \ +do { \ u64 _l = paravirt_read_pmc(counter); \ low = (u32)_l; \ high = _l >> 32; \ -} while(0) +} while (0) static inline unsigned long long paravirt_rdtscp(unsigned int *aux) { @@ -794,7 +809,8 @@ static inline void set_iopl_mask(unsigned mask) } /* The paravirtualized I/O functions */ -static inline void slow_down_io(void) { +static inline void slow_down_io(void) +{ pv_cpu_ops.io_delay(); #ifdef REALLY_SLOW_IO pv_cpu_ops.io_delay(); diff --git a/include/asm-x86/parport.h b/include/asm-x86/parport.h index 019cbca24a38..3c4ffeb467e9 100644 --- a/include/asm-x86/parport.h +++ b/include/asm-x86/parport.h @@ -1,10 +1,10 @@ #ifndef _ASM_X86_PARPORT_H #define _ASM_X86_PARPORT_H -static int __devinit parport_pc_find_isa_ports (int autoirq, int autodma); -static int __devinit parport_pc_find_nonpci_ports (int autoirq, int autodma) +static int __devinit parport_pc_find_isa_ports(int autoirq, int autodma); +static int __devinit parport_pc_find_nonpci_ports(int autoirq, int autodma) { - return parport_pc_find_isa_ports (autoirq, autodma); + return parport_pc_find_isa_ports(autoirq, autodma); } #endif /* _ASM_X86_PARPORT_H */ diff --git a/include/asm-x86/pat.h b/include/asm-x86/pat.h new file mode 100644 index 000000000000..8b822b5a1786 --- /dev/null +++ b/include/asm-x86/pat.h @@ -0,0 +1,16 @@ + +#ifndef _ASM_PAT_H +#define _ASM_PAT_H 1 + +#include <linux/types.h> + +extern int pat_wc_enabled; + +extern void pat_init(void); + +extern int reserve_memtype(u64 start, u64 end, + unsigned long req_type, unsigned long *ret_type); +extern int free_memtype(u64 start, u64 end); + +#endif + diff --git a/include/asm-x86/pci-direct.h b/include/asm-x86/pci-direct.h index 6823fa4f1afa..5b21485be573 100644 --- a/include/asm-x86/pci-direct.h +++ b/include/asm-x86/pci-direct.h @@ -4,7 +4,7 @@ #include <linux/types.h> /* Direct PCI access. This is used for PCI accesses in early boot before - the PCI subsystem works. */ + the PCI subsystem works. */ extern u32 read_pci_config(u8 bus, u8 slot, u8 func, u8 offset); extern u8 read_pci_config_byte(u8 bus, u8 slot, u8 func, u8 offset); diff --git a/include/asm-x86/pci.h b/include/asm-x86/pci.h index c61190cb9e12..ddd8e248fc0a 100644 --- a/include/asm-x86/pci.h +++ b/include/asm-x86/pci.h @@ -8,14 +8,13 @@ #include <asm/scatterlist.h> #include <asm/io.h> - #ifdef __KERNEL__ struct pci_sysdata { int domain; /* PCI domain */ int node; /* NUMA node */ #ifdef CONFIG_X86_64 - void* iommu; /* IOMMU private data */ + void *iommu; /* IOMMU private data */ #endif }; @@ -52,7 +51,7 @@ extern unsigned long pci_mem_start; #define PCIBIOS_MIN_CARDBUS_IO 0x4000 void pcibios_config_init(void); -struct pci_bus * pcibios_scan_root(int bus); +struct pci_bus *pcibios_scan_root(int bus); void pcibios_set_master(struct pci_dev *dev); void pcibios_penalize_isa_irq(int irq, int active); @@ -62,7 +61,8 @@ int pcibios_set_irq_routing(struct pci_dev *dev, int pin, int irq); #define HAVE_PCI_MMAP extern int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma, - enum pci_mmap_state mmap_state, int write_combine); + enum pci_mmap_state mmap_state, + int write_combine); #ifdef CONFIG_PCI diff --git a/include/asm-x86/pci_64.h b/include/asm-x86/pci_64.h index 374690314539..df867e5d80b1 100644 --- a/include/asm-x86/pci_64.h +++ b/include/asm-x86/pci_64.h @@ -1,12 +1,10 @@ #ifndef __x8664_PCI_H #define __x8664_PCI_H - #ifdef __KERNEL__ - #ifdef CONFIG_CALGARY_IOMMU -static inline void* pci_iommu(struct pci_bus *bus) +static inline void *pci_iommu(struct pci_bus *bus) { struct pci_sysdata *sd = bus->sysdata; return sd->iommu; @@ -19,11 +17,10 @@ static inline void set_pci_iommu(struct pci_bus *bus, void *val) } #endif /* CONFIG_CALGARY_IOMMU */ - -extern int (*pci_config_read)(int seg, int bus, int dev, int fn, int reg, int len, u32 *value); -extern int (*pci_config_write)(int seg, int bus, int dev, int fn, int reg, int len, u32 value); - - +extern int (*pci_config_read)(int seg, int bus, int dev, int fn, + int reg, int len, u32 *value); +extern int (*pci_config_write)(int seg, int bus, int dev, int fn, + int reg, int len, u32 value); extern void pci_iommu_alloc(void); @@ -65,5 +62,4 @@ extern void pci_iommu_alloc(void); #endif /* __KERNEL__ */ - #endif /* __x8664_PCI_H */ diff --git a/include/asm-x86/pda.h b/include/asm-x86/pda.h index c0305bff0f19..101fb9e11954 100644 --- a/include/asm-x86/pda.h +++ b/include/asm-x86/pda.h @@ -22,7 +22,6 @@ struct x8664_pda { offset 40!!! */ #endif char *irqstackptr; - unsigned int nodenumber; /* number of current node */ unsigned int __softirq_pending; unsigned int __nmi_count; /* number of NMI on this CPUs */ short mmu_state; @@ -58,34 +57,36 @@ extern struct x8664_pda _proxy_pda; #define pda_offset(field) offsetof(struct x8664_pda, field) -#define pda_to_op(op, field, val) do { \ - typedef typeof(_proxy_pda.field) T__; \ - if (0) { T__ tmp__; tmp__ = (val); } /* type checking */ \ - switch (sizeof(_proxy_pda.field)) { \ - case 2: \ - asm(op "w %1,%%gs:%c2" : \ - "+m" (_proxy_pda.field) : \ - "ri" ((T__)val), \ - "i"(pda_offset(field))); \ - break; \ - case 4: \ - asm(op "l %1,%%gs:%c2" : \ - "+m" (_proxy_pda.field) : \ - "ri" ((T__)val), \ - "i" (pda_offset(field))); \ - break; \ - case 8: \ - asm(op "q %1,%%gs:%c2": \ - "+m" (_proxy_pda.field) : \ - "ri" ((T__)val), \ - "i"(pda_offset(field))); \ - break; \ - default: \ - __bad_pda_field(); \ - } \ - } while (0) +#define pda_to_op(op, field, val) \ +do { \ + typedef typeof(_proxy_pda.field) T__; \ + if (0) { T__ tmp__; tmp__ = (val); } /* type checking */ \ + switch (sizeof(_proxy_pda.field)) { \ + case 2: \ + asm(op "w %1,%%gs:%c2" : \ + "+m" (_proxy_pda.field) : \ + "ri" ((T__)val), \ + "i"(pda_offset(field))); \ + break; \ + case 4: \ + asm(op "l %1,%%gs:%c2" : \ + "+m" (_proxy_pda.field) : \ + "ri" ((T__)val), \ + "i" (pda_offset(field))); \ + break; \ + case 8: \ + asm(op "q %1,%%gs:%c2": \ + "+m" (_proxy_pda.field) : \ + "ri" ((T__)val), \ + "i"(pda_offset(field))); \ + break; \ + default: \ + __bad_pda_field(); \ + } \ +} while (0) -#define pda_from_op(op,field) ({ \ +#define pda_from_op(op, field) \ +({ \ typeof(_proxy_pda.field) ret__; \ switch (sizeof(_proxy_pda.field)) { \ case 2: \ @@ -93,23 +94,24 @@ extern struct x8664_pda _proxy_pda; "=r" (ret__) : \ "i" (pda_offset(field)), \ "m" (_proxy_pda.field)); \ - break; \ + break; \ case 4: \ asm(op "l %%gs:%c1,%0": \ "=r" (ret__): \ "i" (pda_offset(field)), \ "m" (_proxy_pda.field)); \ - break; \ + break; \ case 8: \ asm(op "q %%gs:%c1,%0": \ "=r" (ret__) : \ "i" (pda_offset(field)), \ "m" (_proxy_pda.field)); \ - break; \ + break; \ default: \ __bad_pda_field(); \ - } \ - ret__; }) + } \ + ret__; \ +}) #define read_pda(field) pda_from_op("mov", field) #define write_pda(field, val) pda_to_op("mov", field, val) @@ -118,12 +120,13 @@ extern struct x8664_pda _proxy_pda; #define or_pda(field, val) pda_to_op("or", field, val) /* This is not atomic against other CPUs -- CPU preemption needs to be off */ -#define test_and_clear_bit_pda(bit, field) ({ \ - int old__; \ - asm volatile("btr %2,%%gs:%c3\n\tsbbl %0,%0" \ - : "=r" (old__), "+m" (_proxy_pda.field) \ - : "dIr" (bit), "i" (pda_offset(field)) : "memory"); \ - old__; \ +#define test_and_clear_bit_pda(bit, field) \ +({ \ + int old__; \ + asm volatile("btr %2,%%gs:%c3\n\tsbbl %0,%0" \ + : "=r" (old__), "+m" (_proxy_pda.field) \ + : "dIr" (bit), "i" (pda_offset(field)) : "memory");\ + old__; \ }) #endif diff --git a/include/asm-x86/percpu.h b/include/asm-x86/percpu.h index 0dec00f27eb4..736fc3bb8e1e 100644 --- a/include/asm-x86/percpu.h +++ b/include/asm-x86/percpu.h @@ -85,58 +85,62 @@ DECLARE_PER_CPU(unsigned long, this_cpu_off); * don't give an lvalue though). */ extern void __bad_percpu_size(void); -#define percpu_to_op(op,var,val) \ - do { \ - typedef typeof(var) T__; \ - if (0) { T__ tmp__; tmp__ = (val); } \ - switch (sizeof(var)) { \ - case 1: \ - asm(op "b %1,"__percpu_seg"%0" \ - : "+m" (var) \ - :"ri" ((T__)val)); \ - break; \ - case 2: \ - asm(op "w %1,"__percpu_seg"%0" \ - : "+m" (var) \ - :"ri" ((T__)val)); \ - break; \ - case 4: \ - asm(op "l %1,"__percpu_seg"%0" \ - : "+m" (var) \ - :"ri" ((T__)val)); \ - break; \ - default: __bad_percpu_size(); \ - } \ - } while (0) - -#define percpu_from_op(op,var) \ - ({ \ - typeof(var) ret__; \ - switch (sizeof(var)) { \ - case 1: \ - asm(op "b "__percpu_seg"%1,%0" \ - : "=r" (ret__) \ - : "m" (var)); \ - break; \ - case 2: \ - asm(op "w "__percpu_seg"%1,%0" \ - : "=r" (ret__) \ - : "m" (var)); \ - break; \ - case 4: \ - asm(op "l "__percpu_seg"%1,%0" \ - : "=r" (ret__) \ - : "m" (var)); \ - break; \ - default: __bad_percpu_size(); \ - } \ - ret__; }) +#define percpu_to_op(op, var, val) \ +do { \ + typedef typeof(var) T__; \ + if (0) { \ + T__ tmp__; \ + tmp__ = (val); \ + } \ + switch (sizeof(var)) { \ + case 1: \ + asm(op "b %1,"__percpu_seg"%0" \ + : "+m" (var) \ + : "ri" ((T__)val)); \ + break; \ + case 2: \ + asm(op "w %1,"__percpu_seg"%0" \ + : "+m" (var) \ + : "ri" ((T__)val)); \ + break; \ + case 4: \ + asm(op "l %1,"__percpu_seg"%0" \ + : "+m" (var) \ + : "ri" ((T__)val)); \ + break; \ + default: __bad_percpu_size(); \ + } \ +} while (0) + +#define percpu_from_op(op, var) \ +({ \ + typeof(var) ret__; \ + switch (sizeof(var)) { \ + case 1: \ + asm(op "b "__percpu_seg"%1,%0" \ + : "=r" (ret__) \ + : "m" (var)); \ + break; \ + case 2: \ + asm(op "w "__percpu_seg"%1,%0" \ + : "=r" (ret__) \ + : "m" (var)); \ + break; \ + case 4: \ + asm(op "l "__percpu_seg"%1,%0" \ + : "=r" (ret__) \ + : "m" (var)); \ + break; \ + default: __bad_percpu_size(); \ + } \ + ret__; \ +}) #define x86_read_percpu(var) percpu_from_op("mov", per_cpu__##var) -#define x86_write_percpu(var,val) percpu_to_op("mov", per_cpu__##var, val) -#define x86_add_percpu(var,val) percpu_to_op("add", per_cpu__##var, val) -#define x86_sub_percpu(var,val) percpu_to_op("sub", per_cpu__##var, val) -#define x86_or_percpu(var,val) percpu_to_op("or", per_cpu__##var, val) +#define x86_write_percpu(var, val) percpu_to_op("mov", per_cpu__##var, val) +#define x86_add_percpu(var, val) percpu_to_op("add", per_cpu__##var, val) +#define x86_sub_percpu(var, val) percpu_to_op("sub", per_cpu__##var, val) +#define x86_or_percpu(var, val) percpu_to_op("or", per_cpu__##var, val) #endif /* !__ASSEMBLY__ */ #endif /* !CONFIG_X86_64 */ #endif /* _ASM_X86_PERCPU_H_ */ diff --git a/include/asm-x86/pgtable-2level.h b/include/asm-x86/pgtable-2level.h index 701404fab308..46bc52c0eae1 100644 --- a/include/asm-x86/pgtable-2level.h +++ b/include/asm-x86/pgtable-2level.h @@ -26,7 +26,8 @@ static inline void native_set_pte_atomic(pte_t *ptep, pte_t pte) native_set_pte(ptep, pte); } -static inline void native_set_pte_present(struct mm_struct *mm, unsigned long addr, +static inline void native_set_pte_present(struct mm_struct *mm, + unsigned long addr, pte_t *ptep, pte_t pte) { native_set_pte(ptep, pte); @@ -37,7 +38,8 @@ static inline void native_pmd_clear(pmd_t *pmdp) native_set_pmd(pmdp, __pmd(0)); } -static inline void native_pte_clear(struct mm_struct *mm, unsigned long addr, pte_t *xp) +static inline void native_pte_clear(struct mm_struct *mm, + unsigned long addr, pte_t *xp) { *xp = native_make_pte(0); } @@ -61,16 +63,18 @@ static inline pte_t native_ptep_get_and_clear(pte_t *xp) */ #define PTE_FILE_MAX_BITS 29 -#define pte_to_pgoff(pte) \ - ((((pte).pte_low >> 1) & 0x1f ) + (((pte).pte_low >> 8) << 5 )) +#define pte_to_pgoff(pte) \ + ((((pte).pte_low >> 1) & 0x1f) + (((pte).pte_low >> 8) << 5)) -#define pgoff_to_pte(off) \ - ((pte_t) { .pte_low = (((off) & 0x1f) << 1) + (((off) >> 5) << 8) + _PAGE_FILE }) +#define pgoff_to_pte(off) \ + ((pte_t) { .pte_low = (((off) & 0x1f) << 1) + \ + (((off) >> 5) << 8) + _PAGE_FILE }) /* Encode and de-code a swap entry */ #define __swp_type(x) (((x).val >> 1) & 0x1f) #define __swp_offset(x) ((x).val >> 8) -#define __swp_entry(type, offset) ((swp_entry_t) { ((type) << 1) | ((offset) << 8) }) +#define __swp_entry(type, offset) \ + ((swp_entry_t) { ((type) << 1) | ((offset) << 8) }) #define __pte_to_swp_entry(pte) ((swp_entry_t) { (pte).pte_low }) #define __swp_entry_to_pte(x) ((pte_t) { .pte = (x).val }) diff --git a/include/asm-x86/pgtable-3level.h b/include/asm-x86/pgtable-3level.h index 1d763eec740f..8b4a9d44b7f4 100644 --- a/include/asm-x86/pgtable-3level.h +++ b/include/asm-x86/pgtable-3level.h @@ -8,22 +8,26 @@ * Copyright (C) 1999 Ingo Molnar <mingo@redhat.com> */ -#define pte_ERROR(e) \ - printk("%s:%d: bad pte %p(%08lx%08lx).\n", __FILE__, __LINE__, &(e), (e).pte_high, (e).pte_low) -#define pmd_ERROR(e) \ - printk("%s:%d: bad pmd %p(%016Lx).\n", __FILE__, __LINE__, &(e), pmd_val(e)) -#define pgd_ERROR(e) \ - printk("%s:%d: bad pgd %p(%016Lx).\n", __FILE__, __LINE__, &(e), pgd_val(e)) - +#define pte_ERROR(e) \ + printk("%s:%d: bad pte %p(%08lx%08lx).\n", \ + __FILE__, __LINE__, &(e), (e).pte_high, (e).pte_low) +#define pmd_ERROR(e) \ + printk("%s:%d: bad pmd %p(%016Lx).\n", \ + __FILE__, __LINE__, &(e), pmd_val(e)) +#define pgd_ERROR(e) \ + printk("%s:%d: bad pgd %p(%016Lx).\n", \ + __FILE__, __LINE__, &(e), pgd_val(e)) static inline int pud_none(pud_t pud) { return pud_val(pud) == 0; } + static inline int pud_bad(pud_t pud) { return (pud_val(pud) & ~(PTE_MASK | _KERNPG_TABLE | _PAGE_USER)) != 0; } + static inline int pud_present(pud_t pud) { return pud_val(pud) & _PAGE_PRESENT; @@ -48,7 +52,8 @@ static inline void native_set_pte(pte_t *ptep, pte_t pte) * we are justified in merely clearing the PTE present bit, followed * by a set. The ordering here is important. */ -static inline void native_set_pte_present(struct mm_struct *mm, unsigned long addr, +static inline void native_set_pte_present(struct mm_struct *mm, + unsigned long addr, pte_t *ptep, pte_t pte) { ptep->pte_low = 0; @@ -60,15 +65,17 @@ static inline void native_set_pte_present(struct mm_struct *mm, unsigned long ad static inline void native_set_pte_atomic(pte_t *ptep, pte_t pte) { - set_64bit((unsigned long long *)(ptep),native_pte_val(pte)); + set_64bit((unsigned long long *)(ptep), native_pte_val(pte)); } + static inline void native_set_pmd(pmd_t *pmdp, pmd_t pmd) { - set_64bit((unsigned long long *)(pmdp),native_pmd_val(pmd)); + set_64bit((unsigned long long *)(pmdp), native_pmd_val(pmd)); } + static inline void native_set_pud(pud_t *pudp, pud_t pud) { - set_64bit((unsigned long long *)(pudp),native_pud_val(pud)); + set_64bit((unsigned long long *)(pudp), native_pud_val(pud)); } /* @@ -76,7 +83,8 @@ static inline void native_set_pud(pud_t *pudp, pud_t pud) * entry, so clear the bottom half first and enforce ordering with a compiler * barrier. */ -static inline void native_pte_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep) +static inline void native_pte_clear(struct mm_struct *mm, unsigned long addr, + pte_t *ptep) { ptep->pte_low = 0; smp_wmb(); @@ -107,20 +115,19 @@ static inline void pud_clear(pud_t *pudp) * current pgd to avoid unnecessary TLB flushes. */ pgd = read_cr3(); - if (__pa(pudp) >= pgd && __pa(pudp) < (pgd + sizeof(pgd_t)*PTRS_PER_PGD)) + if (__pa(pudp) >= pgd && __pa(pudp) < + (pgd + sizeof(pgd_t)*PTRS_PER_PGD)) write_cr3(pgd); } -#define pud_page(pud) \ -((struct page *) __va(pud_val(pud) & PAGE_MASK)) +#define pud_page(pud) ((struct page *) __va(pud_val(pud) & PAGE_MASK)) -#define pud_page_vaddr(pud) \ -((unsigned long) __va(pud_val(pud) & PAGE_MASK)) +#define pud_page_vaddr(pud) ((unsigned long) __va(pud_val(pud) & PAGE_MASK)) /* Find an entry in the second-level page table.. */ -#define pmd_offset(pud, address) ((pmd_t *) pud_page(*(pud)) + \ - pmd_index(address)) +#define pmd_offset(pud, address) ((pmd_t *)pud_page(*(pud)) + \ + pmd_index(address)) #ifdef CONFIG_SMP static inline pte_t native_ptep_get_and_clear(pte_t *ptep) @@ -161,7 +168,8 @@ static inline unsigned long pte_pfn(pte_t pte) * put the 32 bits of offset into the high part. */ #define pte_to_pgoff(pte) ((pte).pte_high) -#define pgoff_to_pte(off) ((pte_t) { { .pte_low = _PAGE_FILE, .pte_high = (off) } }) +#define pgoff_to_pte(off) \ + ((pte_t) { { .pte_low = _PAGE_FILE, .pte_high = (off) } }) #define PTE_FILE_MAX_BITS 32 /* Encode and de-code a swap entry */ diff --git a/include/asm-x86/pgtable.h b/include/asm-x86/pgtable.h index 9cf472aeb9ce..f1d9f4a03f6f 100644 --- a/include/asm-x86/pgtable.h +++ b/include/asm-x86/pgtable.h @@ -4,13 +4,13 @@ #define USER_PTRS_PER_PGD ((TASK_SIZE-1)/PGDIR_SIZE+1) #define FIRST_USER_ADDRESS 0 -#define _PAGE_BIT_PRESENT 0 -#define _PAGE_BIT_RW 1 -#define _PAGE_BIT_USER 2 -#define _PAGE_BIT_PWT 3 -#define _PAGE_BIT_PCD 4 -#define _PAGE_BIT_ACCESSED 5 -#define _PAGE_BIT_DIRTY 6 +#define _PAGE_BIT_PRESENT 0 /* is present */ +#define _PAGE_BIT_RW 1 /* writeable */ +#define _PAGE_BIT_USER 2 /* userspace addressable */ +#define _PAGE_BIT_PWT 3 /* page write through */ +#define _PAGE_BIT_PCD 4 /* page cache disabled */ +#define _PAGE_BIT_ACCESSED 5 /* was accessed (raised by CPU) */ +#define _PAGE_BIT_DIRTY 6 /* was written to (raised by CPU) */ #define _PAGE_BIT_FILE 6 #define _PAGE_BIT_PSE 7 /* 4 MB (or 2MB) page */ #define _PAGE_BIT_PAT 7 /* on 4KB pages */ @@ -48,24 +48,39 @@ #endif /* If _PAGE_PRESENT is clear, we use these: */ -#define _PAGE_FILE _PAGE_DIRTY /* nonlinear file mapping, saved PTE; unset:swap */ +#define _PAGE_FILE _PAGE_DIRTY /* nonlinear file mapping, + * saved PTE; unset:swap */ #define _PAGE_PROTNONE _PAGE_PSE /* if the user mapped it with PROT_NONE; pte_present gives true */ -#define _PAGE_TABLE (_PAGE_PRESENT | _PAGE_RW | _PAGE_USER | _PAGE_ACCESSED | _PAGE_DIRTY) -#define _KERNPG_TABLE (_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED | _PAGE_DIRTY) +#define _PAGE_TABLE (_PAGE_PRESENT | _PAGE_RW | _PAGE_USER | \ + _PAGE_ACCESSED | _PAGE_DIRTY) +#define _KERNPG_TABLE (_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED | \ + _PAGE_DIRTY) #define _PAGE_CHG_MASK (PTE_MASK | _PAGE_ACCESSED | _PAGE_DIRTY) -#define PAGE_NONE __pgprot(_PAGE_PROTNONE | _PAGE_ACCESSED) -#define PAGE_SHARED __pgprot(_PAGE_PRESENT | _PAGE_RW | _PAGE_USER | _PAGE_ACCESSED | _PAGE_NX) +#define _PAGE_CACHE_MASK (_PAGE_PCD | _PAGE_PWT) +#define _PAGE_CACHE_WB (0) +#define _PAGE_CACHE_WC (_PAGE_PWT) +#define _PAGE_CACHE_UC_MINUS (_PAGE_PCD) +#define _PAGE_CACHE_UC (_PAGE_PCD | _PAGE_PWT) -#define PAGE_SHARED_EXEC __pgprot(_PAGE_PRESENT | _PAGE_RW | _PAGE_USER | _PAGE_ACCESSED) -#define PAGE_COPY_NOEXEC __pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_ACCESSED | _PAGE_NX) -#define PAGE_COPY_EXEC __pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_ACCESSED) +#define PAGE_NONE __pgprot(_PAGE_PROTNONE | _PAGE_ACCESSED) +#define PAGE_SHARED __pgprot(_PAGE_PRESENT | _PAGE_RW | _PAGE_USER | \ + _PAGE_ACCESSED | _PAGE_NX) + +#define PAGE_SHARED_EXEC __pgprot(_PAGE_PRESENT | _PAGE_RW | \ + _PAGE_USER | _PAGE_ACCESSED) +#define PAGE_COPY_NOEXEC __pgprot(_PAGE_PRESENT | _PAGE_USER | \ + _PAGE_ACCESSED | _PAGE_NX) +#define PAGE_COPY_EXEC __pgprot(_PAGE_PRESENT | _PAGE_USER | \ + _PAGE_ACCESSED) #define PAGE_COPY PAGE_COPY_NOEXEC -#define PAGE_READONLY __pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_ACCESSED | _PAGE_NX) -#define PAGE_READONLY_EXEC __pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_ACCESSED) +#define PAGE_READONLY __pgprot(_PAGE_PRESENT | _PAGE_USER | \ + _PAGE_ACCESSED | _PAGE_NX) +#define PAGE_READONLY_EXEC __pgprot(_PAGE_PRESENT | _PAGE_USER | \ + _PAGE_ACCESSED) #ifdef CONFIG_X86_32 #define _PAGE_KERNEL_EXEC \ @@ -84,6 +99,7 @@ extern pteval_t __PAGE_KERNEL, __PAGE_KERNEL_EXEC; #define __PAGE_KERNEL_RO (__PAGE_KERNEL & ~_PAGE_RW) #define __PAGE_KERNEL_RX (__PAGE_KERNEL_EXEC & ~_PAGE_RW) #define __PAGE_KERNEL_EXEC_NOCACHE (__PAGE_KERNEL_EXEC | _PAGE_PCD | _PAGE_PWT) +#define __PAGE_KERNEL_WC (__PAGE_KERNEL | _PAGE_CACHE_WC) #define __PAGE_KERNEL_NOCACHE (__PAGE_KERNEL | _PAGE_PCD | _PAGE_PWT) #define __PAGE_KERNEL_UC_MINUS (__PAGE_KERNEL | _PAGE_PCD) #define __PAGE_KERNEL_VSYSCALL (__PAGE_KERNEL_RX | _PAGE_USER) @@ -101,6 +117,7 @@ extern pteval_t __PAGE_KERNEL, __PAGE_KERNEL_EXEC; #define PAGE_KERNEL_RO MAKE_GLOBAL(__PAGE_KERNEL_RO) #define PAGE_KERNEL_EXEC MAKE_GLOBAL(__PAGE_KERNEL_EXEC) #define PAGE_KERNEL_RX MAKE_GLOBAL(__PAGE_KERNEL_RX) +#define PAGE_KERNEL_WC MAKE_GLOBAL(__PAGE_KERNEL_WC) #define PAGE_KERNEL_NOCACHE MAKE_GLOBAL(__PAGE_KERNEL_NOCACHE) #define PAGE_KERNEL_UC_MINUS MAKE_GLOBAL(__PAGE_KERNEL_UC_MINUS) #define PAGE_KERNEL_EXEC_NOCACHE MAKE_GLOBAL(__PAGE_KERNEL_EXEC_NOCACHE) @@ -134,7 +151,7 @@ extern pteval_t __PAGE_KERNEL, __PAGE_KERNEL_EXEC; * ZERO_PAGE is a global shared page that is always zero: used * for zero-mapped memory areas etc.. */ -extern unsigned long empty_zero_page[PAGE_SIZE/sizeof(unsigned long)]; +extern unsigned long empty_zero_page[PAGE_SIZE / sizeof(unsigned long)]; #define ZERO_PAGE(vaddr) (virt_to_page(empty_zero_page)) extern spinlock_t pgd_lock; @@ -144,30 +161,101 @@ extern struct list_head pgd_list; * The following only work if pte_present() is true. * Undefined behaviour if not.. */ -static inline int pte_dirty(pte_t pte) { return pte_val(pte) & _PAGE_DIRTY; } -static inline int pte_young(pte_t pte) { return pte_val(pte) & _PAGE_ACCESSED; } -static inline int pte_write(pte_t pte) { return pte_val(pte) & _PAGE_RW; } -static inline int pte_file(pte_t pte) { return pte_val(pte) & _PAGE_FILE; } -static inline int pte_huge(pte_t pte) { return pte_val(pte) & _PAGE_PSE; } -static inline int pte_global(pte_t pte) { return pte_val(pte) & _PAGE_GLOBAL; } -static inline int pte_exec(pte_t pte) { return !(pte_val(pte) & _PAGE_NX); } - -static inline int pmd_large(pmd_t pte) { - return (pmd_val(pte) & (_PAGE_PSE|_PAGE_PRESENT)) == - (_PAGE_PSE|_PAGE_PRESENT); +static inline int pte_dirty(pte_t pte) +{ + return pte_val(pte) & _PAGE_DIRTY; +} + +static inline int pte_young(pte_t pte) +{ + return pte_val(pte) & _PAGE_ACCESSED; +} + +static inline int pte_write(pte_t pte) +{ + return pte_val(pte) & _PAGE_RW; +} + +static inline int pte_file(pte_t pte) +{ + return pte_val(pte) & _PAGE_FILE; +} + +static inline int pte_huge(pte_t pte) +{ + return pte_val(pte) & _PAGE_PSE; +} + +static inline int pte_global(pte_t pte) +{ + return pte_val(pte) & _PAGE_GLOBAL; +} + +static inline int pte_exec(pte_t pte) +{ + return !(pte_val(pte) & _PAGE_NX); +} + +static inline int pmd_large(pmd_t pte) +{ + return (pmd_val(pte) & (_PAGE_PSE | _PAGE_PRESENT)) == + (_PAGE_PSE | _PAGE_PRESENT); +} + +static inline pte_t pte_mkclean(pte_t pte) +{ + return __pte(pte_val(pte) & ~(pteval_t)_PAGE_DIRTY); +} + +static inline pte_t pte_mkold(pte_t pte) +{ + return __pte(pte_val(pte) & ~(pteval_t)_PAGE_ACCESSED); +} + +static inline pte_t pte_wrprotect(pte_t pte) +{ + return __pte(pte_val(pte) & ~(pteval_t)_PAGE_RW); +} + +static inline pte_t pte_mkexec(pte_t pte) +{ + return __pte(pte_val(pte) & ~(pteval_t)_PAGE_NX); +} + +static inline pte_t pte_mkdirty(pte_t pte) +{ + return __pte(pte_val(pte) | _PAGE_DIRTY); +} + +static inline pte_t pte_mkyoung(pte_t pte) +{ + return __pte(pte_val(pte) | _PAGE_ACCESSED); } -static inline pte_t pte_mkclean(pte_t pte) { return __pte(pte_val(pte) & ~(pteval_t)_PAGE_DIRTY); } -static inline pte_t pte_mkold(pte_t pte) { return __pte(pte_val(pte) & ~(pteval_t)_PAGE_ACCESSED); } -static inline pte_t pte_wrprotect(pte_t pte) { return __pte(pte_val(pte) & ~(pteval_t)_PAGE_RW); } -static inline pte_t pte_mkexec(pte_t pte) { return __pte(pte_val(pte) & ~(pteval_t)_PAGE_NX); } -static inline pte_t pte_mkdirty(pte_t pte) { return __pte(pte_val(pte) | _PAGE_DIRTY); } -static inline pte_t pte_mkyoung(pte_t pte) { return __pte(pte_val(pte) | _PAGE_ACCESSED); } -static inline pte_t pte_mkwrite(pte_t pte) { return __pte(pte_val(pte) | _PAGE_RW); } -static inline pte_t pte_mkhuge(pte_t pte) { return __pte(pte_val(pte) | _PAGE_PSE); } -static inline pte_t pte_clrhuge(pte_t pte) { return __pte(pte_val(pte) & ~(pteval_t)_PAGE_PSE); } -static inline pte_t pte_mkglobal(pte_t pte) { return __pte(pte_val(pte) | _PAGE_GLOBAL); } -static inline pte_t pte_clrglobal(pte_t pte) { return __pte(pte_val(pte) & ~(pteval_t)_PAGE_GLOBAL); } +static inline pte_t pte_mkwrite(pte_t pte) +{ + return __pte(pte_val(pte) | _PAGE_RW); +} + +static inline pte_t pte_mkhuge(pte_t pte) +{ + return __pte(pte_val(pte) | _PAGE_PSE); +} + +static inline pte_t pte_clrhuge(pte_t pte) +{ + return __pte(pte_val(pte) & ~(pteval_t)_PAGE_PSE); +} + +static inline pte_t pte_mkglobal(pte_t pte) +{ + return __pte(pte_val(pte) | _PAGE_GLOBAL); +} + +static inline pte_t pte_clrglobal(pte_t pte) +{ + return __pte(pte_val(pte) & ~(pteval_t)_PAGE_GLOBAL); +} extern pteval_t __supported_pte_mask; @@ -334,7 +422,8 @@ static inline void native_set_pte_at(struct mm_struct *mm, unsigned long addr, }) #define __HAVE_ARCH_PTEP_GET_AND_CLEAR -static inline pte_t ptep_get_and_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep) +static inline pte_t ptep_get_and_clear(struct mm_struct *mm, unsigned long addr, + pte_t *ptep) { pte_t pte = native_ptep_get_and_clear(ptep); pte_update(mm, addr, ptep); @@ -342,7 +431,9 @@ static inline pte_t ptep_get_and_clear(struct mm_struct *mm, unsigned long addr, } #define __HAVE_ARCH_PTEP_GET_AND_CLEAR_FULL -static inline pte_t ptep_get_and_clear_full(struct mm_struct *mm, unsigned long addr, pte_t *ptep, int full) +static inline pte_t ptep_get_and_clear_full(struct mm_struct *mm, + unsigned long addr, pte_t *ptep, + int full) { pte_t pte; if (full) { @@ -358,7 +449,8 @@ static inline pte_t ptep_get_and_clear_full(struct mm_struct *mm, unsigned long } #define __HAVE_ARCH_PTEP_SET_WRPROTECT -static inline void ptep_set_wrprotect(struct mm_struct *mm, unsigned long addr, pte_t *ptep) +static inline void ptep_set_wrprotect(struct mm_struct *mm, + unsigned long addr, pte_t *ptep) { clear_bit(_PAGE_BIT_RW, (unsigned long *)&ptep->pte); pte_update(mm, addr, ptep); diff --git a/include/asm-x86/pgtable_32.h b/include/asm-x86/pgtable_32.h index 4e6a0fca0b47..c4a643674458 100644 --- a/include/asm-x86/pgtable_32.h +++ b/include/asm-x86/pgtable_32.h @@ -40,13 +40,13 @@ void paging_init(void); #ifdef CONFIG_X86_PAE # include <asm/pgtable-3level-defs.h> # define PMD_SIZE (1UL << PMD_SHIFT) -# define PMD_MASK (~(PMD_SIZE-1)) +# define PMD_MASK (~(PMD_SIZE - 1)) #else # include <asm/pgtable-2level-defs.h> #endif #define PGDIR_SIZE (1UL << PGDIR_SHIFT) -#define PGDIR_MASK (~(PGDIR_SIZE-1)) +#define PGDIR_MASK (~(PGDIR_SIZE - 1)) #define USER_PGD_PTRS (PAGE_OFFSET >> PGDIR_SHIFT) #define KERNEL_PGD_PTRS (PTRS_PER_PGD-USER_PGD_PTRS) @@ -58,21 +58,22 @@ void paging_init(void); * The vmalloc() routines leaves a hole of 4kB between each vmalloced * area for the same reason. ;) */ -#define VMALLOC_OFFSET (8*1024*1024) -#define VMALLOC_START (((unsigned long) high_memory + \ - 2*VMALLOC_OFFSET-1) & ~(VMALLOC_OFFSET-1)) +#define VMALLOC_OFFSET (8 * 1024 * 1024) +#define VMALLOC_START (((unsigned long)high_memory + 2 * VMALLOC_OFFSET - 1) \ + & ~(VMALLOC_OFFSET - 1)) #ifdef CONFIG_X86_PAE #define LAST_PKMAP 512 #else #define LAST_PKMAP 1024 #endif -#define PKMAP_BASE ((FIXADDR_BOOT_START - PAGE_SIZE*(LAST_PKMAP + 1)) & PMD_MASK) +#define PKMAP_BASE ((FIXADDR_BOOT_START - PAGE_SIZE * (LAST_PKMAP + 1)) \ + & PMD_MASK) #ifdef CONFIG_HIGHMEM -# define VMALLOC_END (PKMAP_BASE-2*PAGE_SIZE) +# define VMALLOC_END (PKMAP_BASE - 2 * PAGE_SIZE) #else -# define VMALLOC_END (FIXADDR_START-2*PAGE_SIZE) +# define VMALLOC_END (FIXADDR_START - 2 * PAGE_SIZE) #endif /* @@ -88,10 +89,16 @@ extern unsigned long pg0[]; #define pte_present(x) ((x).pte_low & (_PAGE_PRESENT | _PAGE_PROTNONE)) /* To avoid harmful races, pmd_none(x) should check only the lower when PAE */ -#define pmd_none(x) (!(unsigned long)pmd_val(x)) -#define pmd_present(x) (pmd_val(x) & _PAGE_PRESENT) -#define pmd_bad(x) ((pmd_val(x) & (~PAGE_MASK & ~_PAGE_USER)) != _KERNPG_TABLE) +#define pmd_none(x) (!(unsigned long)pmd_val((x))) +#define pmd_present(x) (pmd_val((x)) & _PAGE_PRESENT) +extern int pmd_bad(pmd_t pmd); + +#define pmd_bad_v1(x) \ + (_KERNPG_TABLE != (pmd_val((x)) & ~(PAGE_MASK | _PAGE_USER))) +#define pmd_bad_v2(x) \ + (_KERNPG_TABLE != (pmd_val((x)) & ~(PAGE_MASK | _PAGE_USER | \ + _PAGE_PSE | _PAGE_NX))) #define pages_to_mb(x) ((x) >> (20-PAGE_SHIFT)) @@ -117,17 +124,18 @@ static inline void clone_pgd_range(pgd_t *dst, pgd_t *src, int count) } /* - * Macro to mark a page protection value as "uncacheable". On processors which do not support - * it, this is a no-op. + * Macro to mark a page protection value as "uncacheable". + * On processors which do not support it, this is a no-op. */ -#define pgprot_noncached(prot) ((boot_cpu_data.x86 > 3) \ - ? (__pgprot(pgprot_val(prot) | _PAGE_PCD | _PAGE_PWT)) : (prot)) +#define pgprot_noncached(prot) \ + ((boot_cpu_data.x86 > 3) \ + ? (__pgprot(pgprot_val(prot) | _PAGE_PCD | _PAGE_PWT)) \ + : (prot)) /* * Conversion functions: convert a page and protection to a page entry, * and a page entry and page directory to the page they refer to. */ - #define mk_pte(page, pgprot) pfn_pte(page_to_pfn(page), (pgprot)) /* @@ -136,20 +144,20 @@ static inline void clone_pgd_range(pgd_t *dst, pgd_t *src, int count) * this macro returns the index of the entry in the pgd page which would * control the given virtual address */ -#define pgd_index(address) (((address) >> PGDIR_SHIFT) & (PTRS_PER_PGD-1)) -#define pgd_index_k(addr) pgd_index(addr) +#define pgd_index(address) (((address) >> PGDIR_SHIFT) & (PTRS_PER_PGD - 1)) +#define pgd_index_k(addr) pgd_index((addr)) /* * pgd_offset() returns a (pgd_t *) * pgd_index() is used get the offset into the pgd page's array of pgd_t's; */ -#define pgd_offset(mm, address) ((mm)->pgd+pgd_index(address)) +#define pgd_offset(mm, address) ((mm)->pgd + pgd_index((address))) /* * a shortcut which implies the use of the kernel's pgd, instead * of a process's */ -#define pgd_offset_k(address) pgd_offset(&init_mm, address) +#define pgd_offset_k(address) pgd_offset(&init_mm, (address)) static inline int pud_large(pud_t pud) { return 0; } @@ -159,8 +167,8 @@ static inline int pud_large(pud_t pud) { return 0; } * this macro returns the index of the entry in the pmd page which would * control the given virtual address */ -#define pmd_index(address) \ - (((address) >> PMD_SHIFT) & (PTRS_PER_PMD-1)) +#define pmd_index(address) \ + (((address) >> PMD_SHIFT) & (PTRS_PER_PMD - 1)) /* * the pte page can be thought of an array like this: pte_t[PTRS_PER_PTE] @@ -168,43 +176,45 @@ static inline int pud_large(pud_t pud) { return 0; } * this macro returns the index of the entry in the pte page which would * control the given virtual address */ -#define pte_index(address) \ - (((address) >> PAGE_SHIFT) & (PTRS_PER_PTE - 1)) -#define pte_offset_kernel(dir, address) \ - ((pte_t *) pmd_page_vaddr(*(dir)) + pte_index(address)) +#define pte_index(address) \ + (((address) >> PAGE_SHIFT) & (PTRS_PER_PTE - 1)) +#define pte_offset_kernel(dir, address) \ + ((pte_t *)pmd_page_vaddr(*(dir)) + pte_index((address))) -#define pmd_page(pmd) (pfn_to_page(pmd_val(pmd) >> PAGE_SHIFT)) +#define pmd_page(pmd) (pfn_to_page(pmd_val((pmd)) >> PAGE_SHIFT)) -#define pmd_page_vaddr(pmd) \ - ((unsigned long) __va(pmd_val(pmd) & PAGE_MASK)) +#define pmd_page_vaddr(pmd) \ + ((unsigned long)__va(pmd_val((pmd)) & PAGE_MASK)) #if defined(CONFIG_HIGHPTE) -#define pte_offset_map(dir, address) \ - ((pte_t *)kmap_atomic_pte(pmd_page(*(dir)),KM_PTE0) + pte_index(address)) -#define pte_offset_map_nested(dir, address) \ - ((pte_t *)kmap_atomic_pte(pmd_page(*(dir)),KM_PTE1) + pte_index(address)) -#define pte_unmap(pte) kunmap_atomic(pte, KM_PTE0) -#define pte_unmap_nested(pte) kunmap_atomic(pte, KM_PTE1) +#define pte_offset_map(dir, address) \ + ((pte_t *)kmap_atomic_pte(pmd_page(*(dir)), KM_PTE0) + \ + pte_index((address))) +#define pte_offset_map_nested(dir, address) \ + ((pte_t *)kmap_atomic_pte(pmd_page(*(dir)), KM_PTE1) + \ + pte_index((address))) +#define pte_unmap(pte) kunmap_atomic((pte), KM_PTE0) +#define pte_unmap_nested(pte) kunmap_atomic((pte), KM_PTE1) #else -#define pte_offset_map(dir, address) \ - ((pte_t *)page_address(pmd_page(*(dir))) + pte_index(address)) -#define pte_offset_map_nested(dir, address) pte_offset_map(dir, address) +#define pte_offset_map(dir, address) \ + ((pte_t *)page_address(pmd_page(*(dir))) + pte_index((address))) +#define pte_offset_map_nested(dir, address) pte_offset_map((dir), (address)) #define pte_unmap(pte) do { } while (0) #define pte_unmap_nested(pte) do { } while (0) #endif /* Clear a kernel PTE and flush it from the TLB */ -#define kpte_clear_flush(ptep, vaddr) \ -do { \ - pte_clear(&init_mm, vaddr, ptep); \ - __flush_tlb_one(vaddr); \ +#define kpte_clear_flush(ptep, vaddr) \ +do { \ + pte_clear(&init_mm, (vaddr), (ptep)); \ + __flush_tlb_one((vaddr)); \ } while (0) /* * The i386 doesn't have any external MMU info: the kernel page * tables contain all the necessary information. */ -#define update_mmu_cache(vma,address,pte) do { } while (0) +#define update_mmu_cache(vma, address, pte) do { } while (0) void native_pagetable_setup_start(pgd_t *base); void native_pagetable_setup_done(pgd_t *base); @@ -233,7 +243,7 @@ static inline void paravirt_pagetable_setup_done(pgd_t *base) #define kern_addr_valid(kaddr) (0) #endif -#define io_remap_pfn_range(vma, vaddr, pfn, size, prot) \ - remap_pfn_range(vma, vaddr, pfn, size, prot) +#define io_remap_pfn_range(vma, vaddr, pfn, size, prot) \ + remap_pfn_range(vma, vaddr, pfn, size, prot) #endif /* _I386_PGTABLE_H */ diff --git a/include/asm-x86/pgtable_64.h b/include/asm-x86/pgtable_64.h index 0a0b77bc736a..9fd87d0b6477 100644 --- a/include/asm-x86/pgtable_64.h +++ b/include/asm-x86/pgtable_64.h @@ -52,14 +52,18 @@ extern void paging_init(void); #ifndef __ASSEMBLY__ -#define pte_ERROR(e) \ - printk("%s:%d: bad pte %p(%016lx).\n", __FILE__, __LINE__, &(e), pte_val(e)) -#define pmd_ERROR(e) \ - printk("%s:%d: bad pmd %p(%016lx).\n", __FILE__, __LINE__, &(e), pmd_val(e)) -#define pud_ERROR(e) \ - printk("%s:%d: bad pud %p(%016lx).\n", __FILE__, __LINE__, &(e), pud_val(e)) -#define pgd_ERROR(e) \ - printk("%s:%d: bad pgd %p(%016lx).\n", __FILE__, __LINE__, &(e), pgd_val(e)) +#define pte_ERROR(e) \ + printk("%s:%d: bad pte %p(%016lx).\n", \ + __FILE__, __LINE__, &(e), pte_val(e)) +#define pmd_ERROR(e) \ + printk("%s:%d: bad pmd %p(%016lx).\n", \ + __FILE__, __LINE__, &(e), pmd_val(e)) +#define pud_ERROR(e) \ + printk("%s:%d: bad pud %p(%016lx).\n", \ + __FILE__, __LINE__, &(e), pud_val(e)) +#define pgd_ERROR(e) \ + printk("%s:%d: bad pgd %p(%016lx).\n", \ + __FILE__, __LINE__, &(e), pgd_val(e)) #define pgd_none(x) (!pgd_val(x)) #define pud_none(x) (!pud_val(x)) @@ -87,7 +91,8 @@ static inline pte_t native_ptep_get_and_clear(pte_t *xp) #ifdef CONFIG_SMP return native_make_pte(xchg(&xp->pte, 0)); #else - /* native_local_ptep_get_and_clear, but duplicated because of cyclic dependency */ + /* native_local_ptep_get_and_clear, + but duplicated because of cyclic dependency */ pte_t ret = *xp; native_pte_clear(NULL, 0, xp); return ret; @@ -119,7 +124,7 @@ static inline void native_set_pgd(pgd_t *pgdp, pgd_t pgd) *pgdp = pgd; } -static inline void native_pgd_clear(pgd_t * pgd) +static inline void native_pgd_clear(pgd_t *pgd) { native_set_pgd(pgd, native_make_pgd(0)); } @@ -128,19 +133,19 @@ static inline void native_pgd_clear(pgd_t * pgd) #endif /* !__ASSEMBLY__ */ -#define PMD_SIZE (_AC(1,UL) << PMD_SHIFT) -#define PMD_MASK (~(PMD_SIZE-1)) -#define PUD_SIZE (_AC(1,UL) << PUD_SHIFT) -#define PUD_MASK (~(PUD_SIZE-1)) -#define PGDIR_SIZE (_AC(1,UL) << PGDIR_SHIFT) -#define PGDIR_MASK (~(PGDIR_SIZE-1)) +#define PMD_SIZE (_AC(1, UL) << PMD_SHIFT) +#define PMD_MASK (~(PMD_SIZE - 1)) +#define PUD_SIZE (_AC(1, UL) << PUD_SHIFT) +#define PUD_MASK (~(PUD_SIZE - 1)) +#define PGDIR_SIZE (_AC(1, UL) << PGDIR_SHIFT) +#define PGDIR_MASK (~(PGDIR_SIZE - 1)) -#define MAXMEM _AC(0x3fffffffffff, UL) +#define MAXMEM _AC(0x00003fffffffffff, UL) #define VMALLOC_START _AC(0xffffc20000000000, UL) #define VMALLOC_END _AC(0xffffe1ffffffffff, UL) #define VMEMMAP_START _AC(0xffffe20000000000, UL) -#define MODULES_VADDR _AC(0xffffffff88000000, UL) +#define MODULES_VADDR _AC(0xffffffffa0000000, UL) #define MODULES_END _AC(0xfffffffffff00000, UL) #define MODULES_LEN (MODULES_END - MODULES_VADDR) @@ -153,26 +158,28 @@ static inline unsigned long pgd_bad(pgd_t pgd) static inline unsigned long pud_bad(pud_t pud) { - return pud_val(pud) & ~(PTE_MASK | _KERNPG_TABLE | _PAGE_USER); + return pud_val(pud) & + ~(PTE_MASK | _KERNPG_TABLE | _PAGE_USER | _PAGE_PSE | _PAGE_NX); } static inline unsigned long pmd_bad(pmd_t pmd) { - return pmd_val(pmd) & ~(PTE_MASK | _KERNPG_TABLE | _PAGE_USER); + return pmd_val(pmd) & + ~(PTE_MASK | _KERNPG_TABLE | _PAGE_USER | _PAGE_PSE | _PAGE_NX); } -#define pte_none(x) (!pte_val(x)) -#define pte_present(x) (pte_val(x) & (_PAGE_PRESENT | _PAGE_PROTNONE)) +#define pte_none(x) (!pte_val((x))) +#define pte_present(x) (pte_val((x)) & (_PAGE_PRESENT | _PAGE_PROTNONE)) -#define pages_to_mb(x) ((x) >> (20-PAGE_SHIFT)) /* FIXME: is this right? */ -#define pte_page(x) pfn_to_page(pte_pfn(x)) -#define pte_pfn(x) ((pte_val(x) & __PHYSICAL_MASK) >> PAGE_SHIFT) +#define pages_to_mb(x) ((x) >> (20 - PAGE_SHIFT)) /* FIXME: is this right? */ +#define pte_page(x) pfn_to_page(pte_pfn((x))) +#define pte_pfn(x) ((pte_val((x)) & __PHYSICAL_MASK) >> PAGE_SHIFT) /* * Macro to mark a page protection value as "uncacheable". */ -#define pgprot_noncached(prot) (__pgprot(pgprot_val(prot) | _PAGE_PCD | _PAGE_PWT)) - +#define pgprot_noncached(prot) \ + (__pgprot(pgprot_val((prot)) | _PAGE_PCD | _PAGE_PWT)) /* * Conversion functions: convert a page and protection to a page entry, @@ -182,75 +189,81 @@ static inline unsigned long pmd_bad(pmd_t pmd) /* * Level 4 access. */ -#define pgd_page_vaddr(pgd) ((unsigned long) __va((unsigned long)pgd_val(pgd) & PTE_MASK)) -#define pgd_page(pgd) (pfn_to_page(pgd_val(pgd) >> PAGE_SHIFT)) -#define pgd_index(address) (((address) >> PGDIR_SHIFT) & (PTRS_PER_PGD-1)) -#define pgd_offset(mm, addr) ((mm)->pgd + pgd_index(addr)) -#define pgd_offset_k(address) (init_level4_pgt + pgd_index(address)) +#define pgd_page_vaddr(pgd) \ + ((unsigned long)__va((unsigned long)pgd_val((pgd)) & PTE_MASK)) +#define pgd_page(pgd) (pfn_to_page(pgd_val((pgd)) >> PAGE_SHIFT)) +#define pgd_index(address) (((address) >> PGDIR_SHIFT) & (PTRS_PER_PGD - 1)) +#define pgd_offset(mm, address) ((mm)->pgd + pgd_index((address))) +#define pgd_offset_k(address) (init_level4_pgt + pgd_index((address))) #define pgd_present(pgd) (pgd_val(pgd) & _PAGE_PRESENT) static inline int pgd_large(pgd_t pgd) { return 0; } #define mk_kernel_pgd(address) ((pgd_t){ (address) | _KERNPG_TABLE }) /* PUD - Level3 access */ /* to find an entry in a page-table-directory. */ -#define pud_page_vaddr(pud) ((unsigned long) __va(pud_val(pud) & PHYSICAL_PAGE_MASK)) -#define pud_page(pud) (pfn_to_page(pud_val(pud) >> PAGE_SHIFT)) -#define pud_index(address) (((address) >> PUD_SHIFT) & (PTRS_PER_PUD-1)) -#define pud_offset(pgd, address) ((pud_t *) pgd_page_vaddr(*(pgd)) + pud_index(address)) -#define pud_present(pud) (pud_val(pud) & _PAGE_PRESENT) +#define pud_page_vaddr(pud) \ + ((unsigned long)__va(pud_val((pud)) & PHYSICAL_PAGE_MASK)) +#define pud_page(pud) (pfn_to_page(pud_val((pud)) >> PAGE_SHIFT)) +#define pud_index(address) (((address) >> PUD_SHIFT) & (PTRS_PER_PUD - 1)) +#define pud_offset(pgd, address) \ + ((pud_t *)pgd_page_vaddr(*(pgd)) + pud_index((address))) +#define pud_present(pud) (pud_val((pud)) & _PAGE_PRESENT) static inline int pud_large(pud_t pte) { - return (pud_val(pte) & (_PAGE_PSE|_PAGE_PRESENT)) == - (_PAGE_PSE|_PAGE_PRESENT); + return (pud_val(pte) & (_PAGE_PSE | _PAGE_PRESENT)) == + (_PAGE_PSE | _PAGE_PRESENT); } /* PMD - Level 2 access */ -#define pmd_page_vaddr(pmd) ((unsigned long) __va(pmd_val(pmd) & PTE_MASK)) -#define pmd_page(pmd) (pfn_to_page(pmd_val(pmd) >> PAGE_SHIFT)) - -#define pmd_index(address) (((address) >> PMD_SHIFT) & (PTRS_PER_PMD-1)) -#define pmd_offset(dir, address) ((pmd_t *) pud_page_vaddr(*(dir)) + \ - pmd_index(address)) -#define pmd_none(x) (!pmd_val(x)) -#define pmd_present(x) (pmd_val(x) & _PAGE_PRESENT) -#define pfn_pmd(nr,prot) (__pmd(((nr) << PAGE_SHIFT) | pgprot_val(prot))) -#define pmd_pfn(x) ((pmd_val(x) & __PHYSICAL_MASK) >> PAGE_SHIFT) - -#define pte_to_pgoff(pte) ((pte_val(pte) & PHYSICAL_PAGE_MASK) >> PAGE_SHIFT) -#define pgoff_to_pte(off) ((pte_t) { .pte = ((off) << PAGE_SHIFT) | _PAGE_FILE }) +#define pmd_page_vaddr(pmd) ((unsigned long) __va(pmd_val((pmd)) & PTE_MASK)) +#define pmd_page(pmd) (pfn_to_page(pmd_val((pmd)) >> PAGE_SHIFT)) + +#define pmd_index(address) (((address) >> PMD_SHIFT) & (PTRS_PER_PMD - 1)) +#define pmd_offset(dir, address) ((pmd_t *)pud_page_vaddr(*(dir)) + \ + pmd_index(address)) +#define pmd_none(x) (!pmd_val((x))) +#define pmd_present(x) (pmd_val((x)) & _PAGE_PRESENT) +#define pfn_pmd(nr, prot) (__pmd(((nr) << PAGE_SHIFT) | pgprot_val((prot)))) +#define pmd_pfn(x) ((pmd_val((x)) & __PHYSICAL_MASK) >> PAGE_SHIFT) + +#define pte_to_pgoff(pte) ((pte_val((pte)) & PHYSICAL_PAGE_MASK) >> PAGE_SHIFT) +#define pgoff_to_pte(off) ((pte_t) { .pte = ((off) << PAGE_SHIFT) | \ + _PAGE_FILE }) #define PTE_FILE_MAX_BITS __PHYSICAL_MASK_SHIFT /* PTE - Level 1 access. */ /* page, protection -> pte */ -#define mk_pte(page, pgprot) pfn_pte(page_to_pfn(page), (pgprot)) - -#define pte_index(address) \ - (((address) >> PAGE_SHIFT) & (PTRS_PER_PTE - 1)) +#define mk_pte(page, pgprot) pfn_pte(page_to_pfn((page)), (pgprot)) + +#define pte_index(address) (((address) >> PAGE_SHIFT) & (PTRS_PER_PTE - 1)) #define pte_offset_kernel(dir, address) ((pte_t *) pmd_page_vaddr(*(dir)) + \ - pte_index(address)) + pte_index((address))) /* x86-64 always has all page tables mapped. */ -#define pte_offset_map(dir,address) pte_offset_kernel(dir,address) -#define pte_offset_map_nested(dir,address) pte_offset_kernel(dir,address) +#define pte_offset_map(dir, address) pte_offset_kernel((dir), (address)) +#define pte_offset_map_nested(dir, address) pte_offset_kernel((dir), (address)) #define pte_unmap(pte) /* NOP */ -#define pte_unmap_nested(pte) /* NOP */ +#define pte_unmap_nested(pte) /* NOP */ + +#define update_mmu_cache(vma, address, pte) do { } while (0) -#define update_mmu_cache(vma,address,pte) do { } while (0) +extern int direct_gbpages; /* Encode and de-code a swap entry */ #define __swp_type(x) (((x).val >> 1) & 0x3f) #define __swp_offset(x) ((x).val >> 8) -#define __swp_entry(type, offset) ((swp_entry_t) { ((type) << 1) | ((offset) << 8) }) -#define __pte_to_swp_entry(pte) ((swp_entry_t) { pte_val(pte) }) +#define __swp_entry(type, offset) ((swp_entry_t) { ((type) << 1) | \ + ((offset) << 8) }) +#define __pte_to_swp_entry(pte) ((swp_entry_t) { pte_val((pte)) }) #define __swp_entry_to_pte(x) ((pte_t) { .pte = (x).val }) -extern int kern_addr_valid(unsigned long addr); +extern int kern_addr_valid(unsigned long addr); extern void cleanup_highmap(void); -#define io_remap_pfn_range(vma, vaddr, pfn, size, prot) \ - remap_pfn_range(vma, vaddr, pfn, size, prot) +#define io_remap_pfn_range(vma, vaddr, pfn, size, prot) \ + remap_pfn_range(vma, vaddr, pfn, size, prot) #define HAVE_ARCH_UNMAPPED_AREA #define HAVE_ARCH_UNMAPPED_AREA_TOPDOWN @@ -263,8 +276,10 @@ extern void cleanup_highmap(void); /* fs/proc/kcore.c */ #define kc_vaddr_to_offset(v) ((v) & __VIRTUAL_MASK) -#define kc_offset_to_vaddr(o) \ - (((o) & (1UL << (__VIRTUAL_MASK_SHIFT-1))) ? ((o) | (~__VIRTUAL_MASK)) : (o)) +#define kc_offset_to_vaddr(o) \ + (((o) & (1UL << (__VIRTUAL_MASK_SHIFT - 1))) \ + ? ((o) | ~__VIRTUAL_MASK) \ + : (o)) #define __HAVE_ARCH_PTE_SAME #endif /* !__ASSEMBLY__ */ diff --git a/include/asm-x86/posix_types.h b/include/asm-x86/posix_types.h index bb7133dc155d..fe312a5ba204 100644 --- a/include/asm-x86/posix_types.h +++ b/include/asm-x86/posix_types.h @@ -1,11 +1,5 @@ #ifdef __KERNEL__ -# ifdef CONFIG_X86_32 -# include "posix_types_32.h" -# else -# include "posix_types_64.h" -# endif -#else -# ifdef __i386__ +# if defined(CONFIG_X86_32) || defined(__i386__) # include "posix_types_32.h" # else # include "posix_types_64.h" diff --git a/include/asm-x86/posix_types_32.h b/include/asm-x86/posix_types_32.h index 015e539cdef5..b031efda37ec 100644 --- a/include/asm-x86/posix_types_32.h +++ b/include/asm-x86/posix_types_32.h @@ -45,32 +45,39 @@ typedef struct { #if defined(__KERNEL__) #undef __FD_SET -#define __FD_SET(fd,fdsetp) \ - __asm__ __volatile__("btsl %1,%0": \ - "+m" (*(__kernel_fd_set *) (fdsetp)):"r" ((int) (fd))) +#define __FD_SET(fd,fdsetp) \ + asm volatile("btsl %1,%0": \ + "+m" (*(__kernel_fd_set *)(fdsetp)) \ + : "r" ((int)(fd))) #undef __FD_CLR -#define __FD_CLR(fd,fdsetp) \ - __asm__ __volatile__("btrl %1,%0": \ - "+m" (*(__kernel_fd_set *) (fdsetp)):"r" ((int) (fd))) +#define __FD_CLR(fd,fdsetp) \ + asm volatile("btrl %1,%0": \ + "+m" (*(__kernel_fd_set *)(fdsetp)) \ + : "r" ((int) (fd))) #undef __FD_ISSET -#define __FD_ISSET(fd,fdsetp) (__extension__ ({ \ - unsigned char __result; \ - __asm__ __volatile__("btl %1,%2 ; setb %0" \ - :"=q" (__result) :"r" ((int) (fd)), \ - "m" (*(__kernel_fd_set *) (fdsetp))); \ - __result; })) +#define __FD_ISSET(fd,fdsetp) \ + (__extension__ \ + ({ \ + unsigned char __result; \ + asm volatile("btl %1,%2 ; setb %0" \ + : "=q" (__result) \ + : "r" ((int)(fd)), \ + "m" (*(__kernel_fd_set *)(fdsetp))); \ + __result; \ +})) #undef __FD_ZERO -#define __FD_ZERO(fdsetp) \ -do { \ - int __d0, __d1; \ - __asm__ __volatile__("cld ; rep ; stosl" \ - :"=m" (*(__kernel_fd_set *) (fdsetp)), \ - "=&c" (__d0), "=&D" (__d1) \ - :"a" (0), "1" (__FDSET_LONGS), \ - "2" ((__kernel_fd_set *) (fdsetp)) : "memory"); \ +#define __FD_ZERO(fdsetp) \ +do { \ + int __d0, __d1; \ + asm volatile("cld ; rep ; stosl" \ + : "=m" (*(__kernel_fd_set *)(fdsetp)), \ + "=&c" (__d0), "=&D" (__d1) \ + : "a" (0), "1" (__FDSET_LONGS), \ + "2" ((__kernel_fd_set *)(fdsetp)) \ + : "memory"); \ } while (0) #endif /* defined(__KERNEL__) */ diff --git a/include/asm-x86/posix_types_64.h b/include/asm-x86/posix_types_64.h index 9926aa43775b..d6624c95854a 100644 --- a/include/asm-x86/posix_types_64.h +++ b/include/asm-x86/posix_types_64.h @@ -46,7 +46,7 @@ typedef unsigned long __kernel_old_dev_t; #ifdef __KERNEL__ #undef __FD_SET -static __inline__ void __FD_SET(unsigned long fd, __kernel_fd_set *fdsetp) +static inline void __FD_SET(unsigned long fd, __kernel_fd_set *fdsetp) { unsigned long _tmp = fd / __NFDBITS; unsigned long _rem = fd % __NFDBITS; @@ -54,7 +54,7 @@ static __inline__ void __FD_SET(unsigned long fd, __kernel_fd_set *fdsetp) } #undef __FD_CLR -static __inline__ void __FD_CLR(unsigned long fd, __kernel_fd_set *fdsetp) +static inline void __FD_CLR(unsigned long fd, __kernel_fd_set *fdsetp) { unsigned long _tmp = fd / __NFDBITS; unsigned long _rem = fd % __NFDBITS; @@ -62,7 +62,7 @@ static __inline__ void __FD_CLR(unsigned long fd, __kernel_fd_set *fdsetp) } #undef __FD_ISSET -static __inline__ int __FD_ISSET(unsigned long fd, __const__ __kernel_fd_set *p) +static inline int __FD_ISSET(unsigned long fd, __const__ __kernel_fd_set *p) { unsigned long _tmp = fd / __NFDBITS; unsigned long _rem = fd % __NFDBITS; @@ -74,36 +74,36 @@ static __inline__ int __FD_ISSET(unsigned long fd, __const__ __kernel_fd_set *p) * for 256 and 1024-bit fd_sets respectively) */ #undef __FD_ZERO -static __inline__ void __FD_ZERO(__kernel_fd_set *p) +static inline void __FD_ZERO(__kernel_fd_set *p) { unsigned long *tmp = p->fds_bits; int i; if (__builtin_constant_p(__FDSET_LONGS)) { switch (__FDSET_LONGS) { - case 32: - tmp[ 0] = 0; tmp[ 1] = 0; tmp[ 2] = 0; tmp[ 3] = 0; - tmp[ 4] = 0; tmp[ 5] = 0; tmp[ 6] = 0; tmp[ 7] = 0; - tmp[ 8] = 0; tmp[ 9] = 0; tmp[10] = 0; tmp[11] = 0; - tmp[12] = 0; tmp[13] = 0; tmp[14] = 0; tmp[15] = 0; - tmp[16] = 0; tmp[17] = 0; tmp[18] = 0; tmp[19] = 0; - tmp[20] = 0; tmp[21] = 0; tmp[22] = 0; tmp[23] = 0; - tmp[24] = 0; tmp[25] = 0; tmp[26] = 0; tmp[27] = 0; - tmp[28] = 0; tmp[29] = 0; tmp[30] = 0; tmp[31] = 0; - return; - case 16: - tmp[ 0] = 0; tmp[ 1] = 0; tmp[ 2] = 0; tmp[ 3] = 0; - tmp[ 4] = 0; tmp[ 5] = 0; tmp[ 6] = 0; tmp[ 7] = 0; - tmp[ 8] = 0; tmp[ 9] = 0; tmp[10] = 0; tmp[11] = 0; - tmp[12] = 0; tmp[13] = 0; tmp[14] = 0; tmp[15] = 0; - return; - case 8: - tmp[ 0] = 0; tmp[ 1] = 0; tmp[ 2] = 0; tmp[ 3] = 0; - tmp[ 4] = 0; tmp[ 5] = 0; tmp[ 6] = 0; tmp[ 7] = 0; - return; - case 4: - tmp[ 0] = 0; tmp[ 1] = 0; tmp[ 2] = 0; tmp[ 3] = 0; - return; + case 32: + tmp[ 0] = 0; tmp[ 1] = 0; tmp[ 2] = 0; tmp[ 3] = 0; + tmp[ 4] = 0; tmp[ 5] = 0; tmp[ 6] = 0; tmp[ 7] = 0; + tmp[ 8] = 0; tmp[ 9] = 0; tmp[10] = 0; tmp[11] = 0; + tmp[12] = 0; tmp[13] = 0; tmp[14] = 0; tmp[15] = 0; + tmp[16] = 0; tmp[17] = 0; tmp[18] = 0; tmp[19] = 0; + tmp[20] = 0; tmp[21] = 0; tmp[22] = 0; tmp[23] = 0; + tmp[24] = 0; tmp[25] = 0; tmp[26] = 0; tmp[27] = 0; + tmp[28] = 0; tmp[29] = 0; tmp[30] = 0; tmp[31] = 0; + return; + case 16: + tmp[ 0] = 0; tmp[ 1] = 0; tmp[ 2] = 0; tmp[ 3] = 0; + tmp[ 4] = 0; tmp[ 5] = 0; tmp[ 6] = 0; tmp[ 7] = 0; + tmp[ 8] = 0; tmp[ 9] = 0; tmp[10] = 0; tmp[11] = 0; + tmp[12] = 0; tmp[13] = 0; tmp[14] = 0; tmp[15] = 0; + return; + case 8: + tmp[ 0] = 0; tmp[ 1] = 0; tmp[ 2] = 0; tmp[ 3] = 0; + tmp[ 4] = 0; tmp[ 5] = 0; tmp[ 6] = 0; tmp[ 7] = 0; + return; + case 4: + tmp[ 0] = 0; tmp[ 1] = 0; tmp[ 2] = 0; tmp[ 3] = 0; + return; } } i = __FDSET_LONGS; diff --git a/include/asm-x86/processor.h b/include/asm-x86/processor.h index 45a2f0ab33d0..6e26c7c717a2 100644 --- a/include/asm-x86/processor.h +++ b/include/asm-x86/processor.h @@ -3,8 +3,7 @@ #include <asm/processor-flags.h> -/* migration helpers, for KVM - will be removed in 2.6.25: */ -#include <asm/vm86.h> +/* migration helper, for KVM - will be removed in 2.6.25: */ #define Xgt_desc_struct desc_ptr /* Forward declaration, a strange C thing */ @@ -24,6 +23,7 @@ struct mm_struct; #include <asm/msr.h> #include <asm/desc_defs.h> #include <asm/nops.h> + #include <linux/personality.h> #include <linux/cpumask.h> #include <linux/cache.h> @@ -37,16 +37,18 @@ struct mm_struct; static inline void *current_text_addr(void) { void *pc; - asm volatile("mov $1f,%0\n1:":"=r" (pc)); + + asm volatile("mov $1f, %0; 1:":"=r" (pc)); + return pc; } #ifdef CONFIG_X86_VSMP -#define ARCH_MIN_TASKALIGN (1 << INTERNODE_CACHE_SHIFT) -#define ARCH_MIN_MMSTRUCT_ALIGN (1 << INTERNODE_CACHE_SHIFT) +# define ARCH_MIN_TASKALIGN (1 << INTERNODE_CACHE_SHIFT) +# define ARCH_MIN_MMSTRUCT_ALIGN (1 << INTERNODE_CACHE_SHIFT) #else -#define ARCH_MIN_TASKALIGN 16 -#define ARCH_MIN_MMSTRUCT_ALIGN 0 +# define ARCH_MIN_TASKALIGN 16 +# define ARCH_MIN_MMSTRUCT_ALIGN 0 #endif /* @@ -56,69 +58,82 @@ static inline void *current_text_addr(void) */ struct cpuinfo_x86 { - __u8 x86; /* CPU family */ - __u8 x86_vendor; /* CPU vendor */ - __u8 x86_model; - __u8 x86_mask; + __u8 x86; /* CPU family */ + __u8 x86_vendor; /* CPU vendor */ + __u8 x86_model; + __u8 x86_mask; #ifdef CONFIG_X86_32 - char wp_works_ok; /* It doesn't on 386's */ - char hlt_works_ok; /* Problems on some 486Dx4's and old 386's */ - char hard_math; - char rfu; - char fdiv_bug; - char f00f_bug; - char coma_bug; - char pad0; + char wp_works_ok; /* It doesn't on 386's */ + + /* Problems on some 486Dx4's and old 386's: */ + char hlt_works_ok; + char hard_math; + char rfu; + char fdiv_bug; + char f00f_bug; + char coma_bug; + char pad0; #else - /* number of 4K pages in DTLB/ITLB combined(in pages)*/ - int x86_tlbsize; - __u8 x86_virt_bits, x86_phys_bits; - /* cpuid returned core id bits */ - __u8 x86_coreid_bits; - /* Max extended CPUID function supported */ - __u32 extended_cpuid_level; + /* Number of 4K pages in DTLB/ITLB combined(in pages): */ + int x86_tlbsize; + __u8 x86_virt_bits; + __u8 x86_phys_bits; + /* CPUID returned core id bits: */ + __u8 x86_coreid_bits; + /* Max extended CPUID function supported: */ + __u32 extended_cpuid_level; #endif - int cpuid_level; /* Maximum supported CPUID level, -1=no CPUID */ - __u32 x86_capability[NCAPINTS]; - char x86_vendor_id[16]; - char x86_model_id[64]; - int x86_cache_size; /* in KB - valid for CPUS which support this - call */ - int x86_cache_alignment; /* In bytes */ - int x86_power; - unsigned long loops_per_jiffy; + /* Maximum supported CPUID level, -1=no CPUID: */ + int cpuid_level; + __u32 x86_capability[NCAPINTS]; + char x86_vendor_id[16]; + char x86_model_id[64]; + /* in KB - valid for CPUS which support this call: */ + int x86_cache_size; + int x86_cache_alignment; /* In bytes */ + int x86_power; + unsigned long loops_per_jiffy; #ifdef CONFIG_SMP - cpumask_t llc_shared_map; /* cpus sharing the last level cache */ + /* cpus sharing the last level cache: */ + cpumask_t llc_shared_map; #endif - u16 x86_max_cores; /* cpuid returned max cores value */ - u16 apicid; - u16 x86_clflush_size; + /* cpuid returned max cores value: */ + u16 x86_max_cores; + u16 apicid; + u16 initial_apicid; + u16 x86_clflush_size; #ifdef CONFIG_SMP - u16 booted_cores; /* number of cores as seen by OS */ - u16 phys_proc_id; /* Physical processor id. */ - u16 cpu_core_id; /* Core id */ - u16 cpu_index; /* index into per_cpu list */ + /* number of cores as seen by the OS: */ + u16 booted_cores; + /* Physical processor id: */ + u16 phys_proc_id; + /* Core id: */ + u16 cpu_core_id; + /* Index into per_cpu list: */ + u16 cpu_index; #endif } __attribute__((__aligned__(SMP_CACHE_BYTES))); -#define X86_VENDOR_INTEL 0 -#define X86_VENDOR_CYRIX 1 -#define X86_VENDOR_AMD 2 -#define X86_VENDOR_UMC 3 -#define X86_VENDOR_NEXGEN 4 -#define X86_VENDOR_CENTAUR 5 -#define X86_VENDOR_TRANSMETA 7 -#define X86_VENDOR_NSC 8 -#define X86_VENDOR_NUM 9 -#define X86_VENDOR_UNKNOWN 0xff +#define X86_VENDOR_INTEL 0 +#define X86_VENDOR_CYRIX 1 +#define X86_VENDOR_AMD 2 +#define X86_VENDOR_UMC 3 +#define X86_VENDOR_NEXGEN 4 +#define X86_VENDOR_CENTAUR 5 +#define X86_VENDOR_TRANSMETA 7 +#define X86_VENDOR_NSC 8 +#define X86_VENDOR_NUM 9 + +#define X86_VENDOR_UNKNOWN 0xff /* * capabilities of CPUs */ -extern struct cpuinfo_x86 boot_cpu_data; -extern struct cpuinfo_x86 new_cpu_data; -extern struct tss_struct doublefault_tss; -extern __u32 cleared_cpu_caps[NCAPINTS]; +extern struct cpuinfo_x86 boot_cpu_data; +extern struct cpuinfo_x86 new_cpu_data; + +extern struct tss_struct doublefault_tss; +extern __u32 cleared_cpu_caps[NCAPINTS]; #ifdef CONFIG_SMP DECLARE_PER_CPU(struct cpuinfo_x86, cpu_info); @@ -129,7 +144,18 @@ DECLARE_PER_CPU(struct cpuinfo_x86, cpu_info); #define current_cpu_data boot_cpu_data #endif -void cpu_detect(struct cpuinfo_x86 *c); +static inline int hlt_works(int cpu) +{ +#ifdef CONFIG_X86_32 + return cpu_data(cpu).hlt_works_ok; +#else + return 1; +#endif +} + +#define cache_line_size() (boot_cpu_data.x86_cache_alignment) + +extern void cpu_detect(struct cpuinfo_x86 *c); extern void identify_cpu(struct cpuinfo_x86 *); extern void identify_boot_cpu(void); @@ -146,15 +172,15 @@ static inline void detect_ht(struct cpuinfo_x86 *c) {} #endif static inline void native_cpuid(unsigned int *eax, unsigned int *ebx, - unsigned int *ecx, unsigned int *edx) + unsigned int *ecx, unsigned int *edx) { /* ecx is often an input as well as an output. */ - __asm__("cpuid" - : "=a" (*eax), - "=b" (*ebx), - "=c" (*ecx), - "=d" (*edx) - : "0" (*eax), "2" (*ecx)); + asm("cpuid" + : "=a" (*eax), + "=b" (*ebx), + "=c" (*ecx), + "=d" (*edx) + : "0" (*eax), "2" (*ecx)); } static inline void load_cr3(pgd_t *pgdir) @@ -165,54 +191,67 @@ static inline void load_cr3(pgd_t *pgdir) #ifdef CONFIG_X86_32 /* This is the TSS defined by the hardware. */ struct x86_hw_tss { - unsigned short back_link, __blh; - unsigned long sp0; - unsigned short ss0, __ss0h; - unsigned long sp1; - unsigned short ss1, __ss1h; /* ss1 caches MSR_IA32_SYSENTER_CS */ - unsigned long sp2; - unsigned short ss2, __ss2h; - unsigned long __cr3; - unsigned long ip; - unsigned long flags; - unsigned long ax, cx, dx, bx; - unsigned long sp, bp, si, di; - unsigned short es, __esh; - unsigned short cs, __csh; - unsigned short ss, __ssh; - unsigned short ds, __dsh; - unsigned short fs, __fsh; - unsigned short gs, __gsh; - unsigned short ldt, __ldth; - unsigned short trace, io_bitmap_base; + unsigned short back_link, __blh; + unsigned long sp0; + unsigned short ss0, __ss0h; + unsigned long sp1; + /* ss1 caches MSR_IA32_SYSENTER_CS: */ + unsigned short ss1, __ss1h; + unsigned long sp2; + unsigned short ss2, __ss2h; + unsigned long __cr3; + unsigned long ip; + unsigned long flags; + unsigned long ax; + unsigned long cx; + unsigned long dx; + unsigned long bx; + unsigned long sp; + unsigned long bp; + unsigned long si; + unsigned long di; + unsigned short es, __esh; + unsigned short cs, __csh; + unsigned short ss, __ssh; + unsigned short ds, __dsh; + unsigned short fs, __fsh; + unsigned short gs, __gsh; + unsigned short ldt, __ldth; + unsigned short trace; + unsigned short io_bitmap_base; + } __attribute__((packed)); #else struct x86_hw_tss { - u32 reserved1; - u64 sp0; - u64 sp1; - u64 sp2; - u64 reserved2; - u64 ist[7]; - u32 reserved3; - u32 reserved4; - u16 reserved5; - u16 io_bitmap_base; + u32 reserved1; + u64 sp0; + u64 sp1; + u64 sp2; + u64 reserved2; + u64 ist[7]; + u32 reserved3; + u32 reserved4; + u16 reserved5; + u16 io_bitmap_base; + } __attribute__((packed)) ____cacheline_aligned; #endif /* - * Size of io_bitmap. + * IO-bitmap sizes: */ -#define IO_BITMAP_BITS 65536 -#define IO_BITMAP_BYTES (IO_BITMAP_BITS/8) -#define IO_BITMAP_LONGS (IO_BITMAP_BYTES/sizeof(long)) -#define IO_BITMAP_OFFSET offsetof(struct tss_struct, io_bitmap) -#define INVALID_IO_BITMAP_OFFSET 0x8000 -#define INVALID_IO_BITMAP_OFFSET_LAZY 0x9000 +#define IO_BITMAP_BITS 65536 +#define IO_BITMAP_BYTES (IO_BITMAP_BITS/8) +#define IO_BITMAP_LONGS (IO_BITMAP_BYTES/sizeof(long)) +#define IO_BITMAP_OFFSET offsetof(struct tss_struct, io_bitmap) +#define INVALID_IO_BITMAP_OFFSET 0x8000 +#define INVALID_IO_BITMAP_OFFSET_LAZY 0x9000 struct tss_struct { - struct x86_hw_tss x86_tss; + /* + * The hardware state: + */ + struct x86_hw_tss x86_tss; /* * The extra 1 is there because the CPU will access an @@ -220,90 +259,108 @@ struct tss_struct { * bitmap. The extra byte must be all 1 bits, and must * be within the limit. */ - unsigned long io_bitmap[IO_BITMAP_LONGS + 1]; + unsigned long io_bitmap[IO_BITMAP_LONGS + 1]; /* * Cache the current maximum and the last task that used the bitmap: */ - unsigned long io_bitmap_max; - struct thread_struct *io_bitmap_owner; + unsigned long io_bitmap_max; + struct thread_struct *io_bitmap_owner; + /* - * pads the TSS to be cacheline-aligned (size is 0x100) + * Pad the TSS to be cacheline-aligned (size is 0x100): */ - unsigned long __cacheline_filler[35]; + unsigned long __cacheline_filler[35]; /* - * .. and then another 0x100 bytes for emergency kernel stack + * .. and then another 0x100 bytes for the emergency kernel stack: */ - unsigned long stack[64]; + unsigned long stack[64]; + } __attribute__((packed)); DECLARE_PER_CPU(struct tss_struct, init_tss); -/* Save the original ist values for checking stack pointers during debugging */ +/* + * Save the original ist values for checking stack pointers during debugging + */ struct orig_ist { - unsigned long ist[7]; + unsigned long ist[7]; }; #define MXCSR_DEFAULT 0x1f80 struct i387_fsave_struct { - u32 cwd; - u32 swd; - u32 twd; - u32 fip; - u32 fcs; - u32 foo; - u32 fos; - u32 st_space[20]; /* 8*10 bytes for each FP-reg = 80 bytes */ - u32 status; /* software status information */ + u32 cwd; /* FPU Control Word */ + u32 swd; /* FPU Status Word */ + u32 twd; /* FPU Tag Word */ + u32 fip; /* FPU IP Offset */ + u32 fcs; /* FPU IP Selector */ + u32 foo; /* FPU Operand Pointer Offset */ + u32 fos; /* FPU Operand Pointer Selector */ + + /* 8*10 bytes for each FP-reg = 80 bytes: */ + u32 st_space[20]; + + /* Software status information [not touched by FSAVE ]: */ + u32 status; }; struct i387_fxsave_struct { - u16 cwd; - u16 swd; - u16 twd; - u16 fop; + u16 cwd; /* Control Word */ + u16 swd; /* Status Word */ + u16 twd; /* Tag Word */ + u16 fop; /* Last Instruction Opcode */ union { struct { - u64 rip; - u64 rdp; + u64 rip; /* Instruction Pointer */ + u64 rdp; /* Data Pointer */ }; struct { - u32 fip; - u32 fcs; - u32 foo; - u32 fos; + u32 fip; /* FPU IP Offset */ + u32 fcs; /* FPU IP Selector */ + u32 foo; /* FPU Operand Offset */ + u32 fos; /* FPU Operand Selector */ }; }; - u32 mxcsr; - u32 mxcsr_mask; - u32 st_space[32]; /* 8*16 bytes for each FP-reg = 128 bytes */ - u32 xmm_space[64]; /* 16*16 bytes for each XMM-reg = 256 bytes */ - u32 padding[24]; + u32 mxcsr; /* MXCSR Register State */ + u32 mxcsr_mask; /* MXCSR Mask */ + + /* 8*16 bytes for each FP-reg = 128 bytes: */ + u32 st_space[32]; + + /* 16*16 bytes for each XMM-reg = 256 bytes: */ + u32 xmm_space[64]; + + u32 padding[24]; + } __attribute__((aligned(16))); struct i387_soft_struct { - u32 cwd; - u32 swd; - u32 twd; - u32 fip; - u32 fcs; - u32 foo; - u32 fos; - u32 st_space[20]; /* 8*10 bytes for each FP-reg = 80 bytes */ - u8 ftop, changed, lookahead, no_update, rm, alimit; - struct info *info; - u32 entry_eip; + u32 cwd; + u32 swd; + u32 twd; + u32 fip; + u32 fcs; + u32 foo; + u32 fos; + /* 8*10 bytes for each FP-reg = 80 bytes: */ + u32 st_space[20]; + u8 ftop; + u8 changed; + u8 lookahead; + u8 no_update; + u8 rm; + u8 alimit; + struct info *info; + u32 entry_eip; }; union i387_union { struct i387_fsave_struct fsave; struct i387_fxsave_struct fxsave; - struct i387_soft_struct soft; + struct i387_soft_struct soft; }; -#ifdef CONFIG_X86_32 -DECLARE_PER_CPU(u8, cpu_llc_id); -#else +#ifdef CONFIG_X86_64 DECLARE_PER_CPU(struct orig_ist, orig_ist); #endif @@ -313,42 +370,50 @@ extern unsigned int init_intel_cacheinfo(struct cpuinfo_x86 *c); extern unsigned short num_cache_leaves; struct thread_struct { -/* cached TLS descriptors. */ - struct desc_struct tls_array[GDT_ENTRY_TLS_ENTRIES]; - unsigned long sp0; - unsigned long sp; + /* Cached TLS descriptors: */ + struct desc_struct tls_array[GDT_ENTRY_TLS_ENTRIES]; + unsigned long sp0; + unsigned long sp; #ifdef CONFIG_X86_32 - unsigned long sysenter_cs; + unsigned long sysenter_cs; #else - unsigned long usersp; /* Copy from PDA */ - unsigned short es, ds, fsindex, gsindex; + unsigned long usersp; /* Copy from PDA */ + unsigned short es; + unsigned short ds; + unsigned short fsindex; + unsigned short gsindex; #endif - unsigned long ip; - unsigned long fs; - unsigned long gs; -/* Hardware debugging registers */ - unsigned long debugreg0; - unsigned long debugreg1; - unsigned long debugreg2; - unsigned long debugreg3; - unsigned long debugreg6; - unsigned long debugreg7; -/* fault info */ - unsigned long cr2, trap_no, error_code; -/* floating point info */ + unsigned long ip; + unsigned long fs; + unsigned long gs; + /* Hardware debugging registers: */ + unsigned long debugreg0; + unsigned long debugreg1; + unsigned long debugreg2; + unsigned long debugreg3; + unsigned long debugreg6; + unsigned long debugreg7; + /* Fault info: */ + unsigned long cr2; + unsigned long trap_no; + unsigned long error_code; + /* Floating point info: */ union i387_union i387 __attribute__((aligned(16)));; #ifdef CONFIG_X86_32 -/* virtual 86 mode info */ + /* Virtual 86 mode info */ struct vm86_struct __user *vm86_info; unsigned long screen_bitmap; - unsigned long v86flags, v86mask, saved_sp0; - unsigned int saved_fs, saved_gs; + unsigned long v86flags; + unsigned long v86mask; + unsigned long saved_sp0; + unsigned int saved_fs; + unsigned int saved_gs; #endif -/* IO permissions */ - unsigned long *io_bitmap_ptr; - unsigned long iopl; -/* max allowed port in the bitmap, in bytes: */ - unsigned io_bitmap_max; + /* IO permissions: */ + unsigned long *io_bitmap_ptr; + unsigned long iopl; + /* Max allowed port in the bitmap, in bytes: */ + unsigned io_bitmap_max; /* MSR_IA32_DEBUGCTLMSR value to switch in if TIF_DEBUGCTLMSR is set. */ unsigned long debugctlmsr; /* Debug Store - if not 0 points to a DS Save Area configuration; @@ -358,21 +423,27 @@ struct thread_struct { static inline unsigned long native_get_debugreg(int regno) { - unsigned long val = 0; /* Damn you, gcc! */ + unsigned long val = 0; /* Damn you, gcc! */ switch (regno) { case 0: - asm("mov %%db0, %0" :"=r" (val)); break; + asm("mov %%db0, %0" :"=r" (val)); + break; case 1: - asm("mov %%db1, %0" :"=r" (val)); break; + asm("mov %%db1, %0" :"=r" (val)); + break; case 2: - asm("mov %%db2, %0" :"=r" (val)); break; + asm("mov %%db2, %0" :"=r" (val)); + break; case 3: - asm("mov %%db3, %0" :"=r" (val)); break; + asm("mov %%db3, %0" :"=r" (val)); + break; case 6: - asm("mov %%db6, %0" :"=r" (val)); break; + asm("mov %%db6, %0" :"=r" (val)); + break; case 7: - asm("mov %%db7, %0" :"=r" (val)); break; + asm("mov %%db7, %0" :"=r" (val)); + break; default: BUG(); } @@ -383,22 +454,22 @@ static inline void native_set_debugreg(int regno, unsigned long value) { switch (regno) { case 0: - asm("mov %0,%%db0" : /* no output */ :"r" (value)); + asm("mov %0, %%db0" ::"r" (value)); break; case 1: - asm("mov %0,%%db1" : /* no output */ :"r" (value)); + asm("mov %0, %%db1" ::"r" (value)); break; case 2: - asm("mov %0,%%db2" : /* no output */ :"r" (value)); + asm("mov %0, %%db2" ::"r" (value)); break; case 3: - asm("mov %0,%%db3" : /* no output */ :"r" (value)); + asm("mov %0, %%db3" ::"r" (value)); break; case 6: - asm("mov %0,%%db6" : /* no output */ :"r" (value)); + asm("mov %0, %%db6" ::"r" (value)); break; case 7: - asm("mov %0,%%db7" : /* no output */ :"r" (value)); + asm("mov %0, %%db7" ::"r" (value)); break; default: BUG(); @@ -412,23 +483,24 @@ static inline void native_set_iopl_mask(unsigned mask) { #ifdef CONFIG_X86_32 unsigned int reg; - __asm__ __volatile__ ("pushfl;" - "popl %0;" - "andl %1, %0;" - "orl %2, %0;" - "pushl %0;" - "popfl" - : "=&r" (reg) - : "i" (~X86_EFLAGS_IOPL), "r" (mask)); + + asm volatile ("pushfl;" + "popl %0;" + "andl %1, %0;" + "orl %2, %0;" + "pushl %0;" + "popfl" + : "=&r" (reg) + : "i" (~X86_EFLAGS_IOPL), "r" (mask)); #endif } -static inline void native_load_sp0(struct tss_struct *tss, - struct thread_struct *thread) +static inline void +native_load_sp0(struct tss_struct *tss, struct thread_struct *thread) { tss->x86_tss.sp0 = thread->sp0; #ifdef CONFIG_X86_32 - /* Only happens when SEP is enabled, no need to test "SEP"arately */ + /* Only happens when SEP is enabled, no need to test "SEP"arately: */ if (unlikely(tss->x86_tss.ss1 != thread->sysenter_cs)) { tss->x86_tss.ss1 = thread->sysenter_cs; wrmsr(MSR_IA32_SYSENTER_CS, thread->sysenter_cs, 0); @@ -446,8 +518,8 @@ static inline void native_swapgs(void) #ifdef CONFIG_PARAVIRT #include <asm/paravirt.h> #else -#define __cpuid native_cpuid -#define paravirt_enabled() 0 +#define __cpuid native_cpuid +#define paravirt_enabled() 0 /* * These special macros can be used to get or set a debugging register @@ -473,11 +545,12 @@ static inline void load_sp0(struct tss_struct *tss, * enable), so that any CPU's that boot up * after us can get the correct flags. */ -extern unsigned long mmu_cr4_features; +extern unsigned long mmu_cr4_features; static inline void set_in_cr4(unsigned long mask) { unsigned cr4; + mmu_cr4_features |= mask; cr4 = read_cr4(); cr4 |= mask; @@ -487,6 +560,7 @@ static inline void set_in_cr4(unsigned long mask) static inline void clear_in_cr4(unsigned long mask) { unsigned cr4; + mmu_cr4_features &= ~mask; cr4 = read_cr4(); cr4 &= ~mask; @@ -494,42 +568,42 @@ static inline void clear_in_cr4(unsigned long mask) } struct microcode_header { - unsigned int hdrver; - unsigned int rev; - unsigned int date; - unsigned int sig; - unsigned int cksum; - unsigned int ldrver; - unsigned int pf; - unsigned int datasize; - unsigned int totalsize; - unsigned int reserved[3]; + unsigned int hdrver; + unsigned int rev; + unsigned int date; + unsigned int sig; + unsigned int cksum; + unsigned int ldrver; + unsigned int pf; + unsigned int datasize; + unsigned int totalsize; + unsigned int reserved[3]; }; struct microcode { - struct microcode_header hdr; - unsigned int bits[0]; + struct microcode_header hdr; + unsigned int bits[0]; }; -typedef struct microcode microcode_t; -typedef struct microcode_header microcode_header_t; +typedef struct microcode microcode_t; +typedef struct microcode_header microcode_header_t; /* microcode format is extended from prescott processors */ struct extended_signature { - unsigned int sig; - unsigned int pf; - unsigned int cksum; + unsigned int sig; + unsigned int pf; + unsigned int cksum; }; struct extended_sigtable { - unsigned int count; - unsigned int cksum; - unsigned int reserved[3]; + unsigned int count; + unsigned int cksum; + unsigned int reserved[3]; struct extended_signature sigs[0]; }; typedef struct { - unsigned long seg; + unsigned long seg; } mm_segment_t; @@ -541,7 +615,7 @@ extern int kernel_thread(int (*fn)(void *), void *arg, unsigned long flags); /* Free all resources held by a thread. */ extern void release_thread(struct task_struct *); -/* Prepare to copy thread state - unlazy all lazy status */ +/* Prepare to copy thread state - unlazy all lazy state */ extern void prepare_to_copy(struct task_struct *tsk); unsigned long get_wchan(struct task_struct *p); @@ -578,118 +652,137 @@ static inline unsigned int cpuid_eax(unsigned int op) unsigned int eax, ebx, ecx, edx; cpuid(op, &eax, &ebx, &ecx, &edx); + return eax; } + static inline unsigned int cpuid_ebx(unsigned int op) { unsigned int eax, ebx, ecx, edx; cpuid(op, &eax, &ebx, &ecx, &edx); + return ebx; } + static inline unsigned int cpuid_ecx(unsigned int op) { unsigned int eax, ebx, ecx, edx; cpuid(op, &eax, &ebx, &ecx, &edx); + return ecx; } + static inline unsigned int cpuid_edx(unsigned int op) { unsigned int eax, ebx, ecx, edx; cpuid(op, &eax, &ebx, &ecx, &edx); + return edx; } /* REP NOP (PAUSE) is a good thing to insert into busy-wait loops. */ static inline void rep_nop(void) { - __asm__ __volatile__("rep;nop": : :"memory"); + asm volatile("rep; nop" ::: "memory"); } -/* Stop speculative execution */ +static inline void cpu_relax(void) +{ + rep_nop(); +} + +/* Stop speculative execution: */ static inline void sync_core(void) { int tmp; + asm volatile("cpuid" : "=a" (tmp) : "0" (1) - : "ebx", "ecx", "edx", "memory"); + : "ebx", "ecx", "edx", "memory"); } -#define cpu_relax() rep_nop() - static inline void __monitor(const void *eax, unsigned long ecx, - unsigned long edx) + unsigned long edx) { - /* "monitor %eax,%ecx,%edx;" */ - asm volatile( - ".byte 0x0f,0x01,0xc8;" - : :"a" (eax), "c" (ecx), "d"(edx)); + /* "monitor %eax, %ecx, %edx;" */ + asm volatile(".byte 0x0f, 0x01, 0xc8;" + :: "a" (eax), "c" (ecx), "d"(edx)); } static inline void __mwait(unsigned long eax, unsigned long ecx) { - /* "mwait %eax,%ecx;" */ - asm volatile( - ".byte 0x0f,0x01,0xc9;" - : :"a" (eax), "c" (ecx)); + /* "mwait %eax, %ecx;" */ + asm volatile(".byte 0x0f, 0x01, 0xc9;" + :: "a" (eax), "c" (ecx)); } static inline void __sti_mwait(unsigned long eax, unsigned long ecx) { - /* "mwait %eax,%ecx;" */ - asm volatile( - "sti; .byte 0x0f,0x01,0xc9;" - : :"a" (eax), "c" (ecx)); + /* "mwait %eax, %ecx;" */ + asm volatile("sti; .byte 0x0f, 0x01, 0xc9;" + :: "a" (eax), "c" (ecx)); } extern void mwait_idle_with_hints(unsigned long eax, unsigned long ecx); -extern int force_mwait; +extern int force_mwait; extern void select_idle_routine(const struct cpuinfo_x86 *c); -extern unsigned long boot_option_idle_override; +extern unsigned long boot_option_idle_override; extern void enable_sep_cpu(void); extern int sysenter_setup(void); /* Defined in head.S */ -extern struct desc_ptr early_gdt_descr; +extern struct desc_ptr early_gdt_descr; extern void cpu_set_gdt(int); extern void switch_to_new_gdt(void); extern void cpu_init(void); extern void init_gdt(int cpu); -/* from system description table in BIOS. Mostly for MCA use, but - * others may find it useful. */ -extern unsigned int machine_id; -extern unsigned int machine_submodel_id; -extern unsigned int BIOS_revision; +static inline void update_debugctlmsr(unsigned long debugctlmsr) +{ +#ifndef CONFIG_X86_DEBUGCTLMSR + if (boot_cpu_data.x86 < 6) + return; +#endif + wrmsrl(MSR_IA32_DEBUGCTLMSR, debugctlmsr); +} -/* Boot loader type from the setup header */ -extern int bootloader_type; +/* + * from system description table in BIOS. Mostly for MCA use, but + * others may find it useful: + */ +extern unsigned int machine_id; +extern unsigned int machine_submodel_id; +extern unsigned int BIOS_revision; + +/* Boot loader type from the setup header: */ +extern int bootloader_type; -extern char ignore_fpu_irq; -#define cache_line_size() (boot_cpu_data.x86_cache_alignment) +extern char ignore_fpu_irq; #define HAVE_ARCH_PICK_MMAP_LAYOUT 1 #define ARCH_HAS_PREFETCHW #define ARCH_HAS_SPINLOCK_PREFETCH #ifdef CONFIG_X86_32 -#define BASE_PREFETCH ASM_NOP4 -#define ARCH_HAS_PREFETCH +# define BASE_PREFETCH ASM_NOP4 +# define ARCH_HAS_PREFETCH #else -#define BASE_PREFETCH "prefetcht0 (%1)" +# define BASE_PREFETCH "prefetcht0 (%1)" #endif -/* Prefetch instructions for Pentium III and AMD Athlon */ -/* It's not worth to care about 3dnow! prefetches for the K6 - because they are microcoded there and very slow. - However we don't do prefetches for pre XP Athlons currently - That should be fixed. */ +/* + * Prefetch instructions for Pentium III (+) and AMD Athlon (+) + * + * It's not worth to care about 3dnow prefetches for the K6 + * because they are microcoded there and very slow. + */ static inline void prefetch(const void *x) { alternative_input(BASE_PREFETCH, @@ -698,8 +791,11 @@ static inline void prefetch(const void *x) "r" (x)); } -/* 3dnow! prefetch to get an exclusive cache line. Useful for - spinlocks to avoid one state transition in the cache coherency protocol. */ +/* + * 3dnow prefetch to get an exclusive cache line. + * Useful for spinlocks to avoid one state transition in the + * cache coherency protocol: + */ static inline void prefetchw(const void *x) { alternative_input(BASE_PREFETCH, @@ -708,21 +804,25 @@ static inline void prefetchw(const void *x) "r" (x)); } -#define spin_lock_prefetch(x) prefetchw(x) +static inline void spin_lock_prefetch(const void *x) +{ + prefetchw(x); +} + #ifdef CONFIG_X86_32 /* * User space process size: 3GB (default). */ -#define TASK_SIZE (PAGE_OFFSET) -#define STACK_TOP TASK_SIZE -#define STACK_TOP_MAX STACK_TOP - -#define INIT_THREAD { \ - .sp0 = sizeof(init_stack) + (long)&init_stack, \ - .vm86_info = NULL, \ - .sysenter_cs = __KERNEL_CS, \ - .io_bitmap_ptr = NULL, \ - .fs = __KERNEL_PERCPU, \ +#define TASK_SIZE PAGE_OFFSET +#define STACK_TOP TASK_SIZE +#define STACK_TOP_MAX STACK_TOP + +#define INIT_THREAD { \ + .sp0 = sizeof(init_stack) + (long)&init_stack, \ + .vm86_info = NULL, \ + .sysenter_cs = __KERNEL_CS, \ + .io_bitmap_ptr = NULL, \ + .fs = __KERNEL_PERCPU, \ } /* @@ -731,28 +831,15 @@ static inline void prefetchw(const void *x) * permission bitmap. The extra byte must be all 1 bits, and must * be within the limit. */ -#define INIT_TSS { \ - .x86_tss = { \ +#define INIT_TSS { \ + .x86_tss = { \ .sp0 = sizeof(init_stack) + (long)&init_stack, \ - .ss0 = __KERNEL_DS, \ - .ss1 = __KERNEL_CS, \ - .io_bitmap_base = INVALID_IO_BITMAP_OFFSET, \ - }, \ - .io_bitmap = { [0 ... IO_BITMAP_LONGS] = ~0 }, \ -} - -#define start_thread(regs, new_eip, new_esp) do { \ - __asm__("movl %0,%%gs": :"r" (0)); \ - regs->fs = 0; \ - set_fs(USER_DS); \ - regs->ds = __USER_DS; \ - regs->es = __USER_DS; \ - regs->ss = __USER_DS; \ - regs->cs = __USER_CS; \ - regs->ip = new_eip; \ - regs->sp = new_esp; \ -} while (0) - + .ss0 = __KERNEL_DS, \ + .ss1 = __KERNEL_CS, \ + .io_bitmap_base = INVALID_IO_BITMAP_OFFSET, \ + }, \ + .io_bitmap = { [0 ... IO_BITMAP_LONGS] = ~0 }, \ +} extern unsigned long thread_saved_pc(struct task_struct *tsk); @@ -780,24 +867,24 @@ extern unsigned long thread_saved_pc(struct task_struct *tsk); __regs__ - 1; \ }) -#define KSTK_ESP(task) (task_pt_regs(task)->sp) +#define KSTK_ESP(task) (task_pt_regs(task)->sp) #else /* * User space process size. 47bits minus one guard page. */ -#define TASK_SIZE64 (0x800000000000UL - 4096) +#define TASK_SIZE64 ((1UL << 47) - PAGE_SIZE) /* This decides where the kernel will search for a free chunk of vm * space during mmap's. */ -#define IA32_PAGE_OFFSET ((current->personality & ADDR_LIMIT_3GB) ? \ - 0xc0000000 : 0xFFFFe000) +#define IA32_PAGE_OFFSET ((current->personality & ADDR_LIMIT_3GB) ? \ + 0xc0000000 : 0xFFFFe000) -#define TASK_SIZE (test_thread_flag(TIF_IA32) ? \ - IA32_PAGE_OFFSET : TASK_SIZE64) -#define TASK_SIZE_OF(child) ((test_tsk_thread_flag(child, TIF_IA32)) ? \ - IA32_PAGE_OFFSET : TASK_SIZE64) +#define TASK_SIZE (test_thread_flag(TIF_IA32) ? \ + IA32_PAGE_OFFSET : TASK_SIZE64) +#define TASK_SIZE_OF(child) ((test_tsk_thread_flag(child, TIF_IA32)) ? \ + IA32_PAGE_OFFSET : TASK_SIZE64) #define STACK_TOP TASK_SIZE #define STACK_TOP_MAX TASK_SIZE64 @@ -810,33 +897,25 @@ extern unsigned long thread_saved_pc(struct task_struct *tsk); .x86_tss.sp0 = (unsigned long)&init_stack + sizeof(init_stack) \ } -#define start_thread(regs, new_rip, new_rsp) do { \ - asm volatile("movl %0,%%fs; movl %0,%%es; movl %0,%%ds": :"r" (0)); \ - load_gs_index(0); \ - (regs)->ip = (new_rip); \ - (regs)->sp = (new_rsp); \ - write_pda(oldrsp, (new_rsp)); \ - (regs)->cs = __USER_CS; \ - (regs)->ss = __USER_DS; \ - (regs)->flags = 0x200; \ - set_fs(USER_DS); \ -} while (0) - /* * Return saved PC of a blocked thread. * What is this good for? it will be always the scheduler or ret_from_fork. */ -#define thread_saved_pc(t) (*(unsigned long *)((t)->thread.sp - 8)) +#define thread_saved_pc(t) (*(unsigned long *)((t)->thread.sp - 8)) -#define task_pt_regs(tsk) ((struct pt_regs *)(tsk)->thread.sp0 - 1) -#define KSTK_ESP(tsk) -1 /* sorry. doesn't work for syscall. */ +#define task_pt_regs(tsk) ((struct pt_regs *)(tsk)->thread.sp0 - 1) +#define KSTK_ESP(tsk) -1 /* sorry. doesn't work for syscall. */ #endif /* CONFIG_X86_64 */ -/* This decides where the kernel will search for a free chunk of vm +extern void start_thread(struct pt_regs *regs, unsigned long new_ip, + unsigned long new_sp); + +/* + * This decides where the kernel will search for a free chunk of vm * space during mmap's. */ #define TASK_UNMAPPED_BASE (PAGE_ALIGN(TASK_SIZE / 3)) -#define KSTK_EIP(task) (task_pt_regs(task)->ip) +#define KSTK_EIP(task) (task_pt_regs(task)->ip) #endif diff --git a/include/asm-x86/proto.h b/include/asm-x86/proto.h index 68563c0709ac..1e17bcce450e 100644 --- a/include/asm-x86/proto.h +++ b/include/asm-x86/proto.h @@ -7,8 +7,6 @@ extern void early_idt_handler(void); -extern void init_memory_mapping(unsigned long start, unsigned long end); - extern void system_call(void); extern void syscall_init(void); @@ -26,7 +24,7 @@ extern int reboot_force; long do_arch_prctl(struct task_struct *task, int code, unsigned long addr); -#define round_up(x,y) (((x) + (y) - 1) & ~((y)-1)) -#define round_down(x,y) ((x) & ~((y)-1)) +#define round_up(x, y) (((x) + (y) - 1) & ~((y) - 1)) +#define round_down(x, y) ((x) & ~((y) - 1)) #endif diff --git a/include/asm-x86/ptrace.h b/include/asm-x86/ptrace.h index d9e04b46a440..24ec061566c5 100644 --- a/include/asm-x86/ptrace.h +++ b/include/asm-x86/ptrace.h @@ -36,23 +36,23 @@ struct pt_regs { #else /* __KERNEL__ */ struct pt_regs { - long bx; - long cx; - long dx; - long si; - long di; - long bp; - long ax; - int ds; - int es; - int fs; + unsigned long bx; + unsigned long cx; + unsigned long dx; + unsigned long si; + unsigned long di; + unsigned long bp; + unsigned long ax; + unsigned long ds; + unsigned long es; + unsigned long fs; /* int gs; */ - long orig_ax; - long ip; - int cs; - long flags; - long sp; - int ss; + unsigned long orig_ax; + unsigned long ip; + unsigned long cs; + unsigned long flags; + unsigned long sp; + unsigned long ss; }; #include <asm/vm86.h> @@ -140,12 +140,16 @@ extern unsigned long convert_ip_to_linear(struct task_struct *child, struct pt_regs *regs); #ifdef CONFIG_X86_32 -extern void send_sigtrap(struct task_struct *tsk, struct pt_regs *regs, int error_code); +extern void send_sigtrap(struct task_struct *tsk, struct pt_regs *regs, + int error_code); #else void signal_fault(struct pt_regs *regs, void __user *frame, char *where); #endif -#define regs_return_value(regs) ((regs)->ax) +static inline unsigned long regs_return_value(struct pt_regs *regs) +{ + return regs->ax; +} /* * user_mode_vm(regs) determines whether a register set came from user mode. @@ -166,8 +170,8 @@ static inline int user_mode(struct pt_regs *regs) static inline int user_mode_vm(struct pt_regs *regs) { #ifdef CONFIG_X86_32 - return ((regs->cs & SEGMENT_RPL_MASK) | - (regs->flags & VM_MASK)) >= USER_RPL; + return ((regs->cs & SEGMENT_RPL_MASK) | (regs->flags & X86_VM_MASK)) >= + USER_RPL; #else return user_mode(regs); #endif @@ -176,7 +180,7 @@ static inline int user_mode_vm(struct pt_regs *regs) static inline int v8086_mode(struct pt_regs *regs) { #ifdef CONFIG_X86_32 - return (regs->flags & VM_MASK); + return (regs->flags & X86_VM_MASK); #else return 0; /* No V86 mode support in long mode */ #endif diff --git a/include/asm-x86/reboot.h b/include/asm-x86/reboot.h index e9e3ffc22c07..6b5233b4f84b 100644 --- a/include/asm-x86/reboot.h +++ b/include/asm-x86/reboot.h @@ -3,8 +3,7 @@ struct pt_regs; -struct machine_ops -{ +struct machine_ops { void (*restart)(char *cmd); void (*halt)(void); void (*power_off)(void); diff --git a/include/asm-x86/resume-trace.h b/include/asm-x86/resume-trace.h index 46f725b0bc82..2557514d7ef6 100644 --- a/include/asm-x86/resume-trace.h +++ b/include/asm-x86/resume-trace.h @@ -3,16 +3,17 @@ #include <asm/asm.h> -#define TRACE_RESUME(user) do { \ +#define TRACE_RESUME(user) \ +do { \ if (pm_trace_enabled) { \ void *tracedata; \ asm volatile(_ASM_MOV_UL " $1f,%0\n" \ - ".section .tracedata,\"a\"\n" \ - "1:\t.word %c1\n\t" \ - _ASM_PTR " %c2\n" \ - ".previous" \ - :"=r" (tracedata) \ - : "i" (__LINE__), "i" (__FILE__)); \ + ".section .tracedata,\"a\"\n" \ + "1:\t.word %c1\n\t" \ + _ASM_PTR " %c2\n" \ + ".previous" \ + :"=r" (tracedata) \ + : "i" (__LINE__), "i" (__FILE__)); \ generate_resume_trace(tracedata, user); \ } \ } while (0) diff --git a/include/asm-x86/rio.h b/include/asm-x86/rio.h index 97cdcc9887ba..3451c576e6af 100644 --- a/include/asm-x86/rio.h +++ b/include/asm-x86/rio.h @@ -11,53 +11,53 @@ #define RIO_TABLE_VERSION 3 struct rio_table_hdr { - u8 version; /* Version number of this data structure */ - u8 num_scal_dev; /* # of Scalability devices */ - u8 num_rio_dev; /* # of RIO I/O devices */ + u8 version; /* Version number of this data structure */ + u8 num_scal_dev; /* # of Scalability devices */ + u8 num_rio_dev; /* # of RIO I/O devices */ } __attribute__((packed)); struct scal_detail { - u8 node_id; /* Scalability Node ID */ - u32 CBAR; /* Address of 1MB register space */ - u8 port0node; /* Node ID port connected to: 0xFF=None */ - u8 port0port; /* Port num port connected to: 0,1,2, or */ - /* 0xFF=None */ - u8 port1node; /* Node ID port connected to: 0xFF = None */ - u8 port1port; /* Port num port connected to: 0,1,2, or */ - /* 0xFF=None */ - u8 port2node; /* Node ID port connected to: 0xFF = None */ - u8 port2port; /* Port num port connected to: 0,1,2, or */ - /* 0xFF=None */ - u8 chassis_num; /* 1 based Chassis number (1 = boot node) */ + u8 node_id; /* Scalability Node ID */ + u32 CBAR; /* Address of 1MB register space */ + u8 port0node; /* Node ID port connected to: 0xFF=None */ + u8 port0port; /* Port num port connected to: 0,1,2, or */ + /* 0xFF=None */ + u8 port1node; /* Node ID port connected to: 0xFF = None */ + u8 port1port; /* Port num port connected to: 0,1,2, or */ + /* 0xFF=None */ + u8 port2node; /* Node ID port connected to: 0xFF = None */ + u8 port2port; /* Port num port connected to: 0,1,2, or */ + /* 0xFF=None */ + u8 chassis_num; /* 1 based Chassis number (1 = boot node) */ } __attribute__((packed)); struct rio_detail { - u8 node_id; /* RIO Node ID */ - u32 BBAR; /* Address of 1MB register space */ - u8 type; /* Type of device */ - u8 owner_id; /* Node ID of Hurricane that owns this */ - /* node */ - u8 port0node; /* Node ID port connected to: 0xFF=None */ - u8 port0port; /* Port num port connected to: 0,1,2, or */ - /* 0xFF=None */ - u8 port1node; /* Node ID port connected to: 0xFF=None */ - u8 port1port; /* Port num port connected to: 0,1,2, or */ - /* 0xFF=None */ - u8 first_slot; /* Lowest slot number below this Calgary */ - u8 status; /* Bit 0 = 1 : the XAPIC is used */ - /* = 0 : the XAPIC is not used, ie: */ - /* ints fwded to another XAPIC */ - /* Bits1:7 Reserved */ - u8 WP_index; /* instance index - lower ones have */ - /* lower slot numbers/PCI bus numbers */ - u8 chassis_num; /* 1 based Chassis number */ + u8 node_id; /* RIO Node ID */ + u32 BBAR; /* Address of 1MB register space */ + u8 type; /* Type of device */ + u8 owner_id; /* Node ID of Hurricane that owns this */ + /* node */ + u8 port0node; /* Node ID port connected to: 0xFF=None */ + u8 port0port; /* Port num port connected to: 0,1,2, or */ + /* 0xFF=None */ + u8 port1node; /* Node ID port connected to: 0xFF=None */ + u8 port1port; /* Port num port connected to: 0,1,2, or */ + /* 0xFF=None */ + u8 first_slot; /* Lowest slot number below this Calgary */ + u8 status; /* Bit 0 = 1 : the XAPIC is used */ + /* = 0 : the XAPIC is not used, ie: */ + /* ints fwded to another XAPIC */ + /* Bits1:7 Reserved */ + u8 WP_index; /* instance index - lower ones have */ + /* lower slot numbers/PCI bus numbers */ + u8 chassis_num; /* 1 based Chassis number */ } __attribute__((packed)); enum { - HURR_SCALABILTY = 0, /* Hurricane Scalability info */ - HURR_RIOIB = 2, /* Hurricane RIOIB info */ - COMPAT_CALGARY = 4, /* Compatibility Calgary */ - ALT_CALGARY = 5, /* Second Planar Calgary */ + HURR_SCALABILTY = 0, /* Hurricane Scalability info */ + HURR_RIOIB = 2, /* Hurricane RIOIB info */ + COMPAT_CALGARY = 4, /* Compatibility Calgary */ + ALT_CALGARY = 5, /* Second Planar Calgary */ }; /* diff --git a/include/asm-x86/rwsem.h b/include/asm-x86/rwsem.h index 520a379f4b80..750f2a3542b3 100644 --- a/include/asm-x86/rwsem.h +++ b/include/asm-x86/rwsem.h @@ -56,14 +56,16 @@ extern asmregparm struct rw_semaphore * /* * the semaphore definition */ -struct rw_semaphore { - signed long count; + #define RWSEM_UNLOCKED_VALUE 0x00000000 #define RWSEM_ACTIVE_BIAS 0x00000001 #define RWSEM_ACTIVE_MASK 0x0000ffff #define RWSEM_WAITING_BIAS (-0x00010000) #define RWSEM_ACTIVE_READ_BIAS RWSEM_ACTIVE_BIAS #define RWSEM_ACTIVE_WRITE_BIAS (RWSEM_WAITING_BIAS + RWSEM_ACTIVE_BIAS) + +struct rw_semaphore { + signed long count; spinlock_t wait_lock; struct list_head wait_list; #ifdef CONFIG_DEBUG_LOCK_ALLOC @@ -78,11 +80,13 @@ struct rw_semaphore { #endif -#define __RWSEM_INITIALIZER(name) \ -{ RWSEM_UNLOCKED_VALUE, __SPIN_LOCK_UNLOCKED((name).wait_lock), \ - LIST_HEAD_INIT((name).wait_list) __RWSEM_DEP_MAP_INIT(name) } +#define __RWSEM_INITIALIZER(name) \ +{ \ + RWSEM_UNLOCKED_VALUE, __SPIN_LOCK_UNLOCKED((name).wait_lock), \ + LIST_HEAD_INIT((name).wait_list) __RWSEM_DEP_MAP_INIT(name) \ +} -#define DECLARE_RWSEM(name) \ +#define DECLARE_RWSEM(name) \ struct rw_semaphore name = __RWSEM_INITIALIZER(name) extern void __init_rwsem(struct rw_semaphore *sem, const char *name, @@ -100,16 +104,16 @@ do { \ */ static inline void __down_read(struct rw_semaphore *sem) { - __asm__ __volatile__( - "# beginning down_read\n\t" -LOCK_PREFIX " incl (%%eax)\n\t" /* adds 0x00000001, returns the old value */ - " jns 1f\n" - " call call_rwsem_down_read_failed\n" - "1:\n\t" - "# ending down_read\n\t" - : "+m" (sem->count) - : "a" (sem) - : "memory", "cc"); + asm volatile("# beginning down_read\n\t" + LOCK_PREFIX " incl (%%eax)\n\t" + /* adds 0x00000001, returns the old value */ + " jns 1f\n" + " call call_rwsem_down_read_failed\n" + "1:\n\t" + "# ending down_read\n\t" + : "+m" (sem->count) + : "a" (sem) + : "memory", "cc"); } /* @@ -118,21 +122,20 @@ LOCK_PREFIX " incl (%%eax)\n\t" /* adds 0x00000001, returns the old value static inline int __down_read_trylock(struct rw_semaphore *sem) { __s32 result, tmp; - __asm__ __volatile__( - "# beginning __down_read_trylock\n\t" - " movl %0,%1\n\t" - "1:\n\t" - " movl %1,%2\n\t" - " addl %3,%2\n\t" - " jle 2f\n\t" -LOCK_PREFIX " cmpxchgl %2,%0\n\t" - " jnz 1b\n\t" - "2:\n\t" - "# ending __down_read_trylock\n\t" - : "+m" (sem->count), "=&a" (result), "=&r" (tmp) - : "i" (RWSEM_ACTIVE_READ_BIAS) - : "memory", "cc"); - return result>=0 ? 1 : 0; + asm volatile("# beginning __down_read_trylock\n\t" + " movl %0,%1\n\t" + "1:\n\t" + " movl %1,%2\n\t" + " addl %3,%2\n\t" + " jle 2f\n\t" + LOCK_PREFIX " cmpxchgl %2,%0\n\t" + " jnz 1b\n\t" + "2:\n\t" + "# ending __down_read_trylock\n\t" + : "+m" (sem->count), "=&a" (result), "=&r" (tmp) + : "i" (RWSEM_ACTIVE_READ_BIAS) + : "memory", "cc"); + return result >= 0 ? 1 : 0; } /* @@ -143,17 +146,18 @@ static inline void __down_write_nested(struct rw_semaphore *sem, int subclass) int tmp; tmp = RWSEM_ACTIVE_WRITE_BIAS; - __asm__ __volatile__( - "# beginning down_write\n\t" -LOCK_PREFIX " xadd %%edx,(%%eax)\n\t" /* subtract 0x0000ffff, returns the old value */ - " testl %%edx,%%edx\n\t" /* was the count 0 before? */ - " jz 1f\n" - " call call_rwsem_down_write_failed\n" - "1:\n" - "# ending down_write" - : "+m" (sem->count), "=d" (tmp) - : "a" (sem), "1" (tmp) - : "memory", "cc"); + asm volatile("# beginning down_write\n\t" + LOCK_PREFIX " xadd %%edx,(%%eax)\n\t" + /* subtract 0x0000ffff, returns the old value */ + " testl %%edx,%%edx\n\t" + /* was the count 0 before? */ + " jz 1f\n" + " call call_rwsem_down_write_failed\n" + "1:\n" + "# ending down_write" + : "+m" (sem->count), "=d" (tmp) + : "a" (sem), "1" (tmp) + : "memory", "cc"); } static inline void __down_write(struct rw_semaphore *sem) @@ -167,7 +171,7 @@ static inline void __down_write(struct rw_semaphore *sem) static inline int __down_write_trylock(struct rw_semaphore *sem) { signed long ret = cmpxchg(&sem->count, - RWSEM_UNLOCKED_VALUE, + RWSEM_UNLOCKED_VALUE, RWSEM_ACTIVE_WRITE_BIAS); if (ret == RWSEM_UNLOCKED_VALUE) return 1; @@ -180,16 +184,16 @@ static inline int __down_write_trylock(struct rw_semaphore *sem) static inline void __up_read(struct rw_semaphore *sem) { __s32 tmp = -RWSEM_ACTIVE_READ_BIAS; - __asm__ __volatile__( - "# beginning __up_read\n\t" -LOCK_PREFIX " xadd %%edx,(%%eax)\n\t" /* subtracts 1, returns the old value */ - " jns 1f\n\t" - " call call_rwsem_wake\n" - "1:\n" - "# ending __up_read\n" - : "+m" (sem->count), "=d" (tmp) - : "a" (sem), "1" (tmp) - : "memory", "cc"); + asm volatile("# beginning __up_read\n\t" + LOCK_PREFIX " xadd %%edx,(%%eax)\n\t" + /* subtracts 1, returns the old value */ + " jns 1f\n\t" + " call call_rwsem_wake\n" + "1:\n" + "# ending __up_read\n" + : "+m" (sem->count), "=d" (tmp) + : "a" (sem), "1" (tmp) + : "memory", "cc"); } /* @@ -197,17 +201,18 @@ LOCK_PREFIX " xadd %%edx,(%%eax)\n\t" /* subtracts 1, returns the old valu */ static inline void __up_write(struct rw_semaphore *sem) { - __asm__ __volatile__( - "# beginning __up_write\n\t" - " movl %2,%%edx\n\t" -LOCK_PREFIX " xaddl %%edx,(%%eax)\n\t" /* tries to transition 0xffff0001 -> 0x00000000 */ - " jz 1f\n" - " call call_rwsem_wake\n" - "1:\n\t" - "# ending __up_write\n" - : "+m" (sem->count) - : "a" (sem), "i" (-RWSEM_ACTIVE_WRITE_BIAS) - : "memory", "cc", "edx"); + asm volatile("# beginning __up_write\n\t" + " movl %2,%%edx\n\t" + LOCK_PREFIX " xaddl %%edx,(%%eax)\n\t" + /* tries to transition + 0xffff0001 -> 0x00000000 */ + " jz 1f\n" + " call call_rwsem_wake\n" + "1:\n\t" + "# ending __up_write\n" + : "+m" (sem->count) + : "a" (sem), "i" (-RWSEM_ACTIVE_WRITE_BIAS) + : "memory", "cc", "edx"); } /* @@ -215,16 +220,16 @@ LOCK_PREFIX " xaddl %%edx,(%%eax)\n\t" /* tries to transition 0xffff0001 -> */ static inline void __downgrade_write(struct rw_semaphore *sem) { - __asm__ __volatile__( - "# beginning __downgrade_write\n\t" -LOCK_PREFIX " addl %2,(%%eax)\n\t" /* transitions 0xZZZZ0001 -> 0xYYYY0001 */ - " jns 1f\n\t" - " call call_rwsem_downgrade_wake\n" - "1:\n\t" - "# ending __downgrade_write\n" - : "+m" (sem->count) - : "a" (sem), "i" (-RWSEM_WAITING_BIAS) - : "memory", "cc"); + asm volatile("# beginning __downgrade_write\n\t" + LOCK_PREFIX " addl %2,(%%eax)\n\t" + /* transitions 0xZZZZ0001 -> 0xYYYY0001 */ + " jns 1f\n\t" + " call call_rwsem_downgrade_wake\n" + "1:\n\t" + "# ending __downgrade_write\n" + : "+m" (sem->count) + : "a" (sem), "i" (-RWSEM_WAITING_BIAS) + : "memory", "cc"); } /* @@ -232,10 +237,9 @@ LOCK_PREFIX " addl %2,(%%eax)\n\t" /* transitions 0xZZZZ0001 -> 0xYYYY0001 */ static inline void rwsem_atomic_add(int delta, struct rw_semaphore *sem) { - __asm__ __volatile__( -LOCK_PREFIX "addl %1,%0" - : "+m" (sem->count) - : "ir" (delta)); + asm volatile(LOCK_PREFIX "addl %1,%0" + : "+m" (sem->count) + : "ir" (delta)); } /* @@ -245,12 +249,11 @@ static inline int rwsem_atomic_update(int delta, struct rw_semaphore *sem) { int tmp = delta; - __asm__ __volatile__( -LOCK_PREFIX "xadd %0,%1" - : "+r" (tmp), "+m" (sem->count) - : : "memory"); + asm volatile(LOCK_PREFIX "xadd %0,%1" + : "+r" (tmp), "+m" (sem->count) + : : "memory"); - return tmp+delta; + return tmp + delta; } static inline int rwsem_is_locked(struct rw_semaphore *sem) diff --git a/include/asm-x86/segment.h b/include/asm-x86/segment.h index 23f0535fec61..ed5131dd7d92 100644 --- a/include/asm-x86/segment.h +++ b/include/asm-x86/segment.h @@ -191,13 +191,14 @@ #define SEGMENT_TI_MASK 0x4 #define IDT_ENTRIES 256 +#define NUM_EXCEPTION_VECTORS 32 #define GDT_SIZE (GDT_ENTRIES * 8) #define GDT_ENTRY_TLS_ENTRIES 3 #define TLS_SIZE (GDT_ENTRY_TLS_ENTRIES * 8) #ifdef __KERNEL__ #ifndef __ASSEMBLY__ -extern const char early_idt_handlers[IDT_ENTRIES][10]; +extern const char early_idt_handlers[NUM_EXCEPTION_VECTORS][10]; #endif #endif diff --git a/include/asm-x86/semaphore.h b/include/asm-x86/semaphore.h index 572c0b67a6b0..d9b2034ed1d2 100644 --- a/include/asm-x86/semaphore.h +++ b/include/asm-x86/semaphore.h @@ -1,5 +1 @@ -#ifdef CONFIG_X86_32 -# include "semaphore_32.h" -#else -# include "semaphore_64.h" -#endif +#include <linux/semaphore.h> diff --git a/include/asm-x86/semaphore_32.h b/include/asm-x86/semaphore_32.h deleted file mode 100644 index ac96d3804d0c..000000000000 --- a/include/asm-x86/semaphore_32.h +++ /dev/null @@ -1,175 +0,0 @@ -#ifndef _I386_SEMAPHORE_H -#define _I386_SEMAPHORE_H - -#include <linux/linkage.h> - -#ifdef __KERNEL__ - -/* - * SMP- and interrupt-safe semaphores.. - * - * (C) Copyright 1996 Linus Torvalds - * - * Modified 1996-12-23 by Dave Grothe <dave@gcom.com> to fix bugs in - * the original code and to make semaphore waits - * interruptible so that processes waiting on - * semaphores can be killed. - * Modified 1999-02-14 by Andrea Arcangeli, split the sched.c helper - * functions in asm/sempahore-helper.h while fixing a - * potential and subtle race discovered by Ulrich Schmid - * in down_interruptible(). Since I started to play here I - * also implemented the `trylock' semaphore operation. - * 1999-07-02 Artur Skawina <skawina@geocities.com> - * Optimized "0(ecx)" -> "(ecx)" (the assembler does not - * do this). Changed calling sequences from push/jmp to - * traditional call/ret. - * Modified 2001-01-01 Andreas Franck <afranck@gmx.de> - * Some hacks to ensure compatibility with recent - * GCC snapshots, to avoid stack corruption when compiling - * with -fomit-frame-pointer. It's not sure if this will - * be fixed in GCC, as our previous implementation was a - * bit dubious. - * - * If you would like to see an analysis of this implementation, please - * ftp to gcom.com and download the file - * /pub/linux/src/semaphore/semaphore-2.0.24.tar.gz. - * - */ - -#include <asm/system.h> -#include <asm/atomic.h> -#include <linux/wait.h> -#include <linux/rwsem.h> - -struct semaphore { - atomic_t count; - int sleepers; - wait_queue_head_t wait; -}; - - -#define __SEMAPHORE_INITIALIZER(name, n) \ -{ \ - .count = ATOMIC_INIT(n), \ - .sleepers = 0, \ - .wait = __WAIT_QUEUE_HEAD_INITIALIZER((name).wait) \ -} - -#define __DECLARE_SEMAPHORE_GENERIC(name,count) \ - struct semaphore name = __SEMAPHORE_INITIALIZER(name,count) - -#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name,1) - -static inline void sema_init (struct semaphore *sem, int val) -{ -/* - * *sem = (struct semaphore)__SEMAPHORE_INITIALIZER((*sem),val); - * - * i'd rather use the more flexible initialization above, but sadly - * GCC 2.7.2.3 emits a bogus warning. EGCS doesn't. Oh well. - */ - atomic_set(&sem->count, val); - sem->sleepers = 0; - init_waitqueue_head(&sem->wait); -} - -static inline void init_MUTEX (struct semaphore *sem) -{ - sema_init(sem, 1); -} - -static inline void init_MUTEX_LOCKED (struct semaphore *sem) -{ - sema_init(sem, 0); -} - -extern asmregparm void __down_failed(atomic_t *count_ptr); -extern asmregparm int __down_failed_interruptible(atomic_t *count_ptr); -extern asmregparm int __down_failed_trylock(atomic_t *count_ptr); -extern asmregparm void __up_wakeup(atomic_t *count_ptr); - -/* - * This is ugly, but we want the default case to fall through. - * "__down_failed" is a special asm handler that calls the C - * routine that actually waits. See arch/i386/kernel/semaphore.c - */ -static inline void down(struct semaphore * sem) -{ - might_sleep(); - __asm__ __volatile__( - "# atomic down operation\n\t" - LOCK_PREFIX "decl %0\n\t" /* --sem->count */ - "jns 2f\n" - "\tlea %0,%%eax\n\t" - "call __down_failed\n" - "2:" - :"+m" (sem->count) - : - :"memory","ax"); -} - -/* - * Interruptible try to acquire a semaphore. If we obtained - * it, return zero. If we were interrupted, returns -EINTR - */ -static inline int down_interruptible(struct semaphore * sem) -{ - int result; - - might_sleep(); - __asm__ __volatile__( - "# atomic interruptible down operation\n\t" - "xorl %0,%0\n\t" - LOCK_PREFIX "decl %1\n\t" /* --sem->count */ - "jns 2f\n\t" - "lea %1,%%eax\n\t" - "call __down_failed_interruptible\n" - "2:" - :"=&a" (result), "+m" (sem->count) - : - :"memory"); - return result; -} - -/* - * Non-blockingly attempt to down() a semaphore. - * Returns zero if we acquired it - */ -static inline int down_trylock(struct semaphore * sem) -{ - int result; - - __asm__ __volatile__( - "# atomic interruptible down operation\n\t" - "xorl %0,%0\n\t" - LOCK_PREFIX "decl %1\n\t" /* --sem->count */ - "jns 2f\n\t" - "lea %1,%%eax\n\t" - "call __down_failed_trylock\n\t" - "2:\n" - :"=&a" (result), "+m" (sem->count) - : - :"memory"); - return result; -} - -/* - * Note! This is subtle. We jump to wake people up only if - * the semaphore was negative (== somebody was waiting on it). - */ -static inline void up(struct semaphore * sem) -{ - __asm__ __volatile__( - "# atomic up operation\n\t" - LOCK_PREFIX "incl %0\n\t" /* ++sem->count */ - "jg 1f\n\t" - "lea %0,%%eax\n\t" - "call __up_wakeup\n" - "1:" - :"+m" (sem->count) - : - :"memory","ax"); -} - -#endif -#endif diff --git a/include/asm-x86/semaphore_64.h b/include/asm-x86/semaphore_64.h deleted file mode 100644 index 79694306bf7d..000000000000 --- a/include/asm-x86/semaphore_64.h +++ /dev/null @@ -1,180 +0,0 @@ -#ifndef _X86_64_SEMAPHORE_H -#define _X86_64_SEMAPHORE_H - -#include <linux/linkage.h> - -#ifdef __KERNEL__ - -/* - * SMP- and interrupt-safe semaphores.. - * - * (C) Copyright 1996 Linus Torvalds - * - * Modified 1996-12-23 by Dave Grothe <dave@gcom.com> to fix bugs in - * the original code and to make semaphore waits - * interruptible so that processes waiting on - * semaphores can be killed. - * Modified 1999-02-14 by Andrea Arcangeli, split the sched.c helper - * functions in asm/sempahore-helper.h while fixing a - * potential and subtle race discovered by Ulrich Schmid - * in down_interruptible(). Since I started to play here I - * also implemented the `trylock' semaphore operation. - * 1999-07-02 Artur Skawina <skawina@geocities.com> - * Optimized "0(ecx)" -> "(ecx)" (the assembler does not - * do this). Changed calling sequences from push/jmp to - * traditional call/ret. - * Modified 2001-01-01 Andreas Franck <afranck@gmx.de> - * Some hacks to ensure compatibility with recent - * GCC snapshots, to avoid stack corruption when compiling - * with -fomit-frame-pointer. It's not sure if this will - * be fixed in GCC, as our previous implementation was a - * bit dubious. - * - * If you would like to see an analysis of this implementation, please - * ftp to gcom.com and download the file - * /pub/linux/src/semaphore/semaphore-2.0.24.tar.gz. - * - */ - -#include <asm/system.h> -#include <asm/atomic.h> -#include <asm/rwlock.h> -#include <linux/wait.h> -#include <linux/rwsem.h> -#include <linux/stringify.h> - -struct semaphore { - atomic_t count; - int sleepers; - wait_queue_head_t wait; -}; - -#define __SEMAPHORE_INITIALIZER(name, n) \ -{ \ - .count = ATOMIC_INIT(n), \ - .sleepers = 0, \ - .wait = __WAIT_QUEUE_HEAD_INITIALIZER((name).wait) \ -} - -#define __DECLARE_SEMAPHORE_GENERIC(name,count) \ - struct semaphore name = __SEMAPHORE_INITIALIZER(name,count) - -#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name,1) - -static inline void sema_init (struct semaphore *sem, int val) -{ -/* - * *sem = (struct semaphore)__SEMAPHORE_INITIALIZER((*sem),val); - * - * i'd rather use the more flexible initialization above, but sadly - * GCC 2.7.2.3 emits a bogus warning. EGCS doesn't. Oh well. - */ - atomic_set(&sem->count, val); - sem->sleepers = 0; - init_waitqueue_head(&sem->wait); -} - -static inline void init_MUTEX (struct semaphore *sem) -{ - sema_init(sem, 1); -} - -static inline void init_MUTEX_LOCKED (struct semaphore *sem) -{ - sema_init(sem, 0); -} - -asmlinkage void __down_failed(void /* special register calling convention */); -asmlinkage int __down_failed_interruptible(void /* params in registers */); -asmlinkage int __down_failed_trylock(void /* params in registers */); -asmlinkage void __up_wakeup(void /* special register calling convention */); - -asmlinkage void __down(struct semaphore * sem); -asmlinkage int __down_interruptible(struct semaphore * sem); -asmlinkage int __down_trylock(struct semaphore * sem); -asmlinkage void __up(struct semaphore * sem); - -/* - * This is ugly, but we want the default case to fall through. - * "__down_failed" is a special asm handler that calls the C - * routine that actually waits. See arch/x86_64/kernel/semaphore.c - */ -static inline void down(struct semaphore * sem) -{ - might_sleep(); - - __asm__ __volatile__( - "# atomic down operation\n\t" - LOCK_PREFIX "decl %0\n\t" /* --sem->count */ - "jns 1f\n\t" - "call __down_failed\n" - "1:" - :"=m" (sem->count) - :"D" (sem) - :"memory"); -} - -/* - * Interruptible try to acquire a semaphore. If we obtained - * it, return zero. If we were interrupted, returns -EINTR - */ -static inline int down_interruptible(struct semaphore * sem) -{ - int result; - - might_sleep(); - - __asm__ __volatile__( - "# atomic interruptible down operation\n\t" - "xorl %0,%0\n\t" - LOCK_PREFIX "decl %1\n\t" /* --sem->count */ - "jns 2f\n\t" - "call __down_failed_interruptible\n" - "2:\n" - :"=&a" (result), "=m" (sem->count) - :"D" (sem) - :"memory"); - return result; -} - -/* - * Non-blockingly attempt to down() a semaphore. - * Returns zero if we acquired it - */ -static inline int down_trylock(struct semaphore * sem) -{ - int result; - - __asm__ __volatile__( - "# atomic interruptible down operation\n\t" - "xorl %0,%0\n\t" - LOCK_PREFIX "decl %1\n\t" /* --sem->count */ - "jns 2f\n\t" - "call __down_failed_trylock\n\t" - "2:\n" - :"=&a" (result), "=m" (sem->count) - :"D" (sem) - :"memory","cc"); - return result; -} - -/* - * Note! This is subtle. We jump to wake people up only if - * the semaphore was negative (== somebody was waiting on it). - * The default case (no contention) will result in NO - * jumps for both down() and up(). - */ -static inline void up(struct semaphore * sem) -{ - __asm__ __volatile__( - "# atomic up operation\n\t" - LOCK_PREFIX "incl %0\n\t" /* ++sem->count */ - "jg 1f\n\t" - "call __up_wakeup\n" - "1:" - :"=m" (sem->count) - :"D" (sem) - :"memory"); -} -#endif /* __KERNEL__ */ -#endif diff --git a/include/asm-x86/setup.h b/include/asm-x86/setup.h index 071e054abd82..fa6763af8d26 100644 --- a/include/asm-x86/setup.h +++ b/include/asm-x86/setup.h @@ -4,6 +4,10 @@ #define COMMAND_LINE_SIZE 2048 #ifndef __ASSEMBLY__ + +/* Interrupt control for vSMPowered x86_64 systems */ +void vsmp_init(void); + char *machine_specific_memory_setup(void); #ifndef CONFIG_PARAVIRT #define paravirt_post_allocator_init() do {} while (0) @@ -51,8 +55,8 @@ struct e820entry; char * __init machine_specific_memory_setup(void); char *memory_setup(void); -int __init copy_e820_map(struct e820entry * biosmap, int nr_map); -int __init sanitize_e820_map(struct e820entry * biosmap, char * pnr_map); +int __init copy_e820_map(struct e820entry *biosmap, int nr_map); +int __init sanitize_e820_map(struct e820entry *biosmap, char *pnr_map); void __init add_memory_region(unsigned long long start, unsigned long long size, int type); diff --git a/include/asm-x86/sigcontext.h b/include/asm-x86/sigcontext.h index d743947f4c77..2f9c884d2c0f 100644 --- a/include/asm-x86/sigcontext.h +++ b/include/asm-x86/sigcontext.h @@ -79,7 +79,7 @@ struct sigcontext { unsigned long flags; unsigned long sp_at_signal; unsigned short ss, __ssh; - struct _fpstate __user * fpstate; + struct _fpstate __user *fpstate; unsigned long oldmask; unsigned long cr2; }; @@ -107,7 +107,7 @@ struct sigcontext { unsigned long eflags; unsigned long esp_at_signal; unsigned short ss, __ssh; - struct _fpstate __user * fpstate; + struct _fpstate __user *fpstate; unsigned long oldmask; unsigned long cr2; }; @@ -121,7 +121,8 @@ struct sigcontext { struct _fpstate { __u16 cwd; __u16 swd; - __u16 twd; /* Note this is not the same as the 32bit/x87/FSAVE twd */ + __u16 twd; /* Note this is not the same as the + 32bit/x87/FSAVE twd */ __u16 fop; __u64 rip; __u64 rdp; diff --git a/include/asm-x86/sigcontext32.h b/include/asm-x86/sigcontext32.h index 6ffab4fd593a..57a9686fb491 100644 --- a/include/asm-x86/sigcontext32.h +++ b/include/asm-x86/sigcontext32.h @@ -26,7 +26,7 @@ struct _fpstate_ia32 { __u32 cw; __u32 sw; __u32 tag; /* not compatible to 64bit twd */ - __u32 ipoff; + __u32 ipoff; __u32 cssel; __u32 dataoff; __u32 datasel; @@ -39,7 +39,7 @@ struct _fpstate_ia32 { __u32 mxcsr; __u32 reserved; struct _fpxreg _fxsr_st[8]; - struct _xmmreg _xmm[8]; /* It's actually 16 */ + struct _xmmreg _xmm[8]; /* It's actually 16 */ __u32 padding[56]; }; diff --git a/include/asm-x86/signal.h b/include/asm-x86/signal.h index aee7eca585ab..f15186d39c69 100644 --- a/include/asm-x86/signal.h +++ b/include/asm-x86/signal.h @@ -185,61 +185,61 @@ typedef struct sigaltstack { #define __HAVE_ARCH_SIG_BITOPS -#define sigaddset(set,sig) \ - (__builtin_constantp(sig) ? \ - __const_sigaddset((set),(sig)) : \ - __gen_sigaddset((set),(sig))) +#define sigaddset(set,sig) \ + (__builtin_constantp(sig) \ + ? __const_sigaddset((set), (sig)) \ + : __gen_sigaddset((set), (sig))) -static __inline__ void __gen_sigaddset(sigset_t *set, int _sig) +static inline void __gen_sigaddset(sigset_t *set, int _sig) { - __asm__("btsl %1,%0" : "+m"(*set) : "Ir"(_sig - 1) : "cc"); + asm("btsl %1,%0" : "+m"(*set) : "Ir"(_sig - 1) : "cc"); } -static __inline__ void __const_sigaddset(sigset_t *set, int _sig) +static inline void __const_sigaddset(sigset_t *set, int _sig) { unsigned long sig = _sig - 1; set->sig[sig / _NSIG_BPW] |= 1 << (sig % _NSIG_BPW); } -#define sigdelset(set,sig) \ - (__builtin_constant_p(sig) ? \ - __const_sigdelset((set),(sig)) : \ - __gen_sigdelset((set),(sig))) +#define sigdelset(set, sig) \ + (__builtin_constant_p(sig) \ + ? __const_sigdelset((set), (sig)) \ + : __gen_sigdelset((set), (sig))) -static __inline__ void __gen_sigdelset(sigset_t *set, int _sig) +static inline void __gen_sigdelset(sigset_t *set, int _sig) { - __asm__("btrl %1,%0" : "+m"(*set) : "Ir"(_sig - 1) : "cc"); + asm("btrl %1,%0" : "+m"(*set) : "Ir"(_sig - 1) : "cc"); } -static __inline__ void __const_sigdelset(sigset_t *set, int _sig) +static inline void __const_sigdelset(sigset_t *set, int _sig) { unsigned long sig = _sig - 1; set->sig[sig / _NSIG_BPW] &= ~(1 << (sig % _NSIG_BPW)); } -static __inline__ int __const_sigismember(sigset_t *set, int _sig) +static inline int __const_sigismember(sigset_t *set, int _sig) { unsigned long sig = _sig - 1; return 1 & (set->sig[sig / _NSIG_BPW] >> (sig % _NSIG_BPW)); } -static __inline__ int __gen_sigismember(sigset_t *set, int _sig) +static inline int __gen_sigismember(sigset_t *set, int _sig) { int ret; - __asm__("btl %2,%1\n\tsbbl %0,%0" - : "=r"(ret) : "m"(*set), "Ir"(_sig-1) : "cc"); + asm("btl %2,%1\n\tsbbl %0,%0" + : "=r"(ret) : "m"(*set), "Ir"(_sig-1) : "cc"); return ret; } -#define sigismember(set,sig) \ - (__builtin_constant_p(sig) ? \ - __const_sigismember((set),(sig)) : \ - __gen_sigismember((set),(sig))) +#define sigismember(set, sig) \ + (__builtin_constant_p(sig) \ + ? __const_sigismember((set), (sig)) \ + : __gen_sigismember((set), (sig))) -static __inline__ int sigfindinword(unsigned long word) +static inline int sigfindinword(unsigned long word) { - __asm__("bsfl %1,%0" : "=r"(word) : "rm"(word) : "cc"); + asm("bsfl %1,%0" : "=r"(word) : "rm"(word) : "cc"); return word; } diff --git a/include/asm-x86/smp.h b/include/asm-x86/smp.h index f2e8319a6b0b..62ebdec394b9 100644 --- a/include/asm-x86/smp.h +++ b/include/asm-x86/smp.h @@ -1,5 +1,209 @@ -#ifdef CONFIG_X86_32 -# include "smp_32.h" +#ifndef _ASM_X86_SMP_H_ +#define _ASM_X86_SMP_H_ +#ifndef __ASSEMBLY__ +#include <linux/cpumask.h> +#include <linux/init.h> +#include <asm/percpu.h> + +/* + * We need the APIC definitions automatically as part of 'smp.h' + */ +#ifdef CONFIG_X86_LOCAL_APIC +# include <asm/mpspec.h> +# include <asm/apic.h> +# ifdef CONFIG_X86_IO_APIC +# include <asm/io_apic.h> +# endif +#endif +#include <asm/pda.h> +#include <asm/thread_info.h> + +extern cpumask_t cpu_callout_map; +extern cpumask_t cpu_initialized; +extern cpumask_t cpu_callin_map; + +extern void (*mtrr_hook)(void); +extern void zap_low_mappings(void); + +extern int smp_num_siblings; +extern unsigned int num_processors; +extern cpumask_t cpu_initialized; + +#ifdef CONFIG_SMP +extern u16 x86_cpu_to_apicid_init[]; +extern u16 x86_bios_cpu_apicid_init[]; +extern void *x86_cpu_to_apicid_early_ptr; +extern void *x86_bios_cpu_apicid_early_ptr; #else -# include "smp_64.h" +#define x86_cpu_to_apicid_early_ptr NULL +#define x86_bios_cpu_apicid_early_ptr NULL +#endif + +DECLARE_PER_CPU(cpumask_t, cpu_sibling_map); +DECLARE_PER_CPU(cpumask_t, cpu_core_map); +DECLARE_PER_CPU(u16, cpu_llc_id); +DECLARE_PER_CPU(u16, x86_cpu_to_apicid); +DECLARE_PER_CPU(u16, x86_bios_cpu_apicid); + +/* Static state in head.S used to set up a CPU */ +extern struct { + void *sp; + unsigned short ss; +} stack_start; + +struct smp_ops { + void (*smp_prepare_boot_cpu)(void); + void (*smp_prepare_cpus)(unsigned max_cpus); + int (*cpu_up)(unsigned cpu); + void (*smp_cpus_done)(unsigned max_cpus); + + void (*smp_send_stop)(void); + void (*smp_send_reschedule)(int cpu); + int (*smp_call_function_mask)(cpumask_t mask, + void (*func)(void *info), void *info, + int wait); +}; + +/* Globals due to paravirt */ +extern void set_cpu_sibling_map(int cpu); + +#ifdef CONFIG_SMP +#ifndef CONFIG_PARAVIRT +#define startup_ipi_hook(phys_apicid, start_eip, start_esp) do { } while (0) +#endif +extern struct smp_ops smp_ops; + +static inline void smp_send_stop(void) +{ + smp_ops.smp_send_stop(); +} + +static inline void smp_prepare_boot_cpu(void) +{ + smp_ops.smp_prepare_boot_cpu(); +} + +static inline void smp_prepare_cpus(unsigned int max_cpus) +{ + smp_ops.smp_prepare_cpus(max_cpus); +} + +static inline void smp_cpus_done(unsigned int max_cpus) +{ + smp_ops.smp_cpus_done(max_cpus); +} + +static inline int __cpu_up(unsigned int cpu) +{ + return smp_ops.cpu_up(cpu); +} + +static inline void smp_send_reschedule(int cpu) +{ + smp_ops.smp_send_reschedule(cpu); +} + +static inline int smp_call_function_mask(cpumask_t mask, + void (*func) (void *info), void *info, + int wait) +{ + return smp_ops.smp_call_function_mask(mask, func, info, wait); +} + +void native_smp_prepare_boot_cpu(void); +void native_smp_prepare_cpus(unsigned int max_cpus); +void native_smp_cpus_done(unsigned int max_cpus); +int native_cpu_up(unsigned int cpunum); + +extern int __cpu_disable(void); +extern void __cpu_die(unsigned int cpu); + +extern void prefill_possible_map(void); + +void smp_store_cpu_info(int id); +#define cpu_physical_id(cpu) per_cpu(x86_cpu_to_apicid, cpu) + +/* We don't mark CPUs online until __cpu_up(), so we need another measure */ +static inline int num_booting_cpus(void) +{ + return cpus_weight(cpu_callout_map); +} +#endif /* CONFIG_SMP */ + +extern unsigned disabled_cpus __cpuinitdata; + +#ifdef CONFIG_X86_32_SMP +/* + * This function is needed by all SMP systems. It must _always_ be valid + * from the initial startup. We map APIC_BASE very early in page_setup(), + * so this is correct in the x86 case. + */ +DECLARE_PER_CPU(int, cpu_number); +#define raw_smp_processor_id() (x86_read_percpu(cpu_number)) +extern int safe_smp_processor_id(void); + +#elif defined(CONFIG_X86_64_SMP) +#define raw_smp_processor_id() read_pda(cpunumber) + +#define stack_smp_processor_id() \ +({ \ + struct thread_info *ti; \ + __asm__("andq %%rsp,%0; ":"=r" (ti) : "0" (CURRENT_MASK)); \ + ti->cpu; \ +}) +#define safe_smp_processor_id() smp_processor_id() + +#else /* !CONFIG_X86_32_SMP && !CONFIG_X86_64_SMP */ +#define cpu_physical_id(cpu) boot_cpu_physical_apicid +#define safe_smp_processor_id() 0 +#define stack_smp_processor_id() 0 +#endif + +#ifdef CONFIG_X86_LOCAL_APIC + +static inline int logical_smp_processor_id(void) +{ + /* we don't want to mark this access volatile - bad code generation */ + return GET_APIC_LOGICAL_ID(*(u32 *)(APIC_BASE + APIC_LDR)); +} + +#ifndef CONFIG_X86_64 +static inline unsigned int read_apic_id(void) +{ + return *(u32 *)(APIC_BASE + APIC_ID); +} +#else +extern unsigned int read_apic_id(void); +#endif + + +# ifdef APIC_DEFINITION +extern int hard_smp_processor_id(void); +# else +# include <mach_apicdef.h> +static inline int hard_smp_processor_id(void) +{ + /* we don't want to mark this access volatile - bad code generation */ + return GET_APIC_ID(read_apic_id()); +} +# endif /* APIC_DEFINITION */ + +#else /* CONFIG_X86_LOCAL_APIC */ + +# ifndef CONFIG_SMP +# define hard_smp_processor_id() 0 +# endif + +#endif /* CONFIG_X86_LOCAL_APIC */ + +#ifdef CONFIG_HOTPLUG_CPU +extern void cpu_exit_clear(void); +extern void cpu_uninit(void); +extern void remove_siblinginfo(int cpu); +#endif + +extern void smp_alloc_memory(void); +extern void lock_ipi_call_lock(void); +extern void unlock_ipi_call_lock(void); +#endif /* __ASSEMBLY__ */ #endif diff --git a/include/asm-x86/smp_32.h b/include/asm-x86/smp_32.h deleted file mode 100644 index 56152e312287..000000000000 --- a/include/asm-x86/smp_32.h +++ /dev/null @@ -1,165 +0,0 @@ -#ifndef __ASM_SMP_H -#define __ASM_SMP_H - -#ifndef __ASSEMBLY__ -#include <linux/cpumask.h> -#include <linux/init.h> - -/* - * We need the APIC definitions automatically as part of 'smp.h' - */ -#ifdef CONFIG_X86_LOCAL_APIC -# include <asm/mpspec.h> -# include <asm/apic.h> -# ifdef CONFIG_X86_IO_APIC -# include <asm/io_apic.h> -# endif -#endif - -extern cpumask_t cpu_callout_map; -extern cpumask_t cpu_callin_map; - -extern int smp_num_siblings; -extern unsigned int num_processors; - -extern void smp_alloc_memory(void); -extern void lock_ipi_call_lock(void); -extern void unlock_ipi_call_lock(void); - -extern void (*mtrr_hook) (void); -extern void zap_low_mappings (void); - -extern u8 __initdata x86_cpu_to_apicid_init[]; -extern void *x86_cpu_to_apicid_early_ptr; - -DECLARE_PER_CPU(cpumask_t, cpu_sibling_map); -DECLARE_PER_CPU(cpumask_t, cpu_core_map); -DECLARE_PER_CPU(u8, cpu_llc_id); -DECLARE_PER_CPU(u8, x86_cpu_to_apicid); - -#ifdef CONFIG_HOTPLUG_CPU -extern void cpu_exit_clear(void); -extern void cpu_uninit(void); -extern void remove_siblinginfo(int cpu); -#endif - -/* Globals due to paravirt */ -extern void set_cpu_sibling_map(int cpu); - -struct smp_ops -{ - void (*smp_prepare_boot_cpu)(void); - void (*smp_prepare_cpus)(unsigned max_cpus); - int (*cpu_up)(unsigned cpu); - void (*smp_cpus_done)(unsigned max_cpus); - - void (*smp_send_stop)(void); - void (*smp_send_reschedule)(int cpu); - int (*smp_call_function_mask)(cpumask_t mask, - void (*func)(void *info), void *info, - int wait); -}; - -#ifdef CONFIG_SMP -extern struct smp_ops smp_ops; - -static inline void smp_prepare_boot_cpu(void) -{ - smp_ops.smp_prepare_boot_cpu(); -} -static inline void smp_prepare_cpus(unsigned int max_cpus) -{ - smp_ops.smp_prepare_cpus(max_cpus); -} -static inline int __cpu_up(unsigned int cpu) -{ - return smp_ops.cpu_up(cpu); -} -static inline void smp_cpus_done(unsigned int max_cpus) -{ - smp_ops.smp_cpus_done(max_cpus); -} - -static inline void smp_send_stop(void) -{ - smp_ops.smp_send_stop(); -} -static inline void smp_send_reschedule(int cpu) -{ - smp_ops.smp_send_reschedule(cpu); -} -static inline int smp_call_function_mask(cpumask_t mask, - void (*func) (void *info), void *info, - int wait) -{ - return smp_ops.smp_call_function_mask(mask, func, info, wait); -} - -void native_smp_prepare_boot_cpu(void); -void native_smp_prepare_cpus(unsigned int max_cpus); -int native_cpu_up(unsigned int cpunum); -void native_smp_cpus_done(unsigned int max_cpus); - -#ifndef CONFIG_PARAVIRT -#define startup_ipi_hook(phys_apicid, start_eip, start_esp) do { } while (0) -#endif - -extern int __cpu_disable(void); -extern void __cpu_die(unsigned int cpu); - -/* - * This function is needed by all SMP systems. It must _always_ be valid - * from the initial startup. We map APIC_BASE very early in page_setup(), - * so this is correct in the x86 case. - */ -DECLARE_PER_CPU(int, cpu_number); -#define raw_smp_processor_id() (x86_read_percpu(cpu_number)) - -#define cpu_physical_id(cpu) per_cpu(x86_cpu_to_apicid, cpu) - -extern int safe_smp_processor_id(void); - -void __cpuinit smp_store_cpu_info(int id); - -/* We don't mark CPUs online until __cpu_up(), so we need another measure */ -static inline int num_booting_cpus(void) -{ - return cpus_weight(cpu_callout_map); -} - -#else /* CONFIG_SMP */ - -#define safe_smp_processor_id() 0 -#define cpu_physical_id(cpu) boot_cpu_physical_apicid - -#endif /* !CONFIG_SMP */ - -#ifdef CONFIG_X86_LOCAL_APIC - -static __inline int logical_smp_processor_id(void) -{ - /* we don't want to mark this access volatile - bad code generation */ - return GET_APIC_LOGICAL_ID(*(u32 *)(APIC_BASE + APIC_LDR)); -} - -# ifdef APIC_DEFINITION -extern int hard_smp_processor_id(void); -# else -# include <mach_apicdef.h> -static inline int hard_smp_processor_id(void) -{ - /* we don't want to mark this access volatile - bad code generation */ - return GET_APIC_ID(*(u32 *)(APIC_BASE + APIC_ID)); -} -# endif /* APIC_DEFINITION */ - -#else /* CONFIG_X86_LOCAL_APIC */ - -# ifndef CONFIG_SMP -# define hard_smp_processor_id() 0 -# endif - -#endif /* CONFIG_X86_LOCAL_APIC */ - -#endif /* !ASSEMBLY */ -#endif diff --git a/include/asm-x86/smp_64.h b/include/asm-x86/smp_64.h deleted file mode 100644 index e0a75519ad21..000000000000 --- a/include/asm-x86/smp_64.h +++ /dev/null @@ -1,101 +0,0 @@ -#ifndef __ASM_SMP_H -#define __ASM_SMP_H - -#include <linux/cpumask.h> -#include <linux/init.h> - -/* - * We need the APIC definitions automatically as part of 'smp.h' - */ -#include <asm/apic.h> -#include <asm/io_apic.h> -#include <asm/mpspec.h> -#include <asm/pda.h> -#include <asm/thread_info.h> - -extern cpumask_t cpu_callout_map; -extern cpumask_t cpu_initialized; - -extern int smp_num_siblings; -extern unsigned int num_processors; - -extern void smp_alloc_memory(void); -extern void lock_ipi_call_lock(void); -extern void unlock_ipi_call_lock(void); - -extern int smp_call_function_mask(cpumask_t mask, void (*func)(void *), - void *info, int wait); - -extern u16 __initdata x86_cpu_to_apicid_init[]; -extern u16 __initdata x86_bios_cpu_apicid_init[]; -extern void *x86_cpu_to_apicid_early_ptr; -extern void *x86_bios_cpu_apicid_early_ptr; - -DECLARE_PER_CPU(cpumask_t, cpu_sibling_map); -DECLARE_PER_CPU(cpumask_t, cpu_core_map); -DECLARE_PER_CPU(u16, cpu_llc_id); -DECLARE_PER_CPU(u16, x86_cpu_to_apicid); -DECLARE_PER_CPU(u16, x86_bios_cpu_apicid); - -static inline int cpu_present_to_apicid(int mps_cpu) -{ - if (cpu_present(mps_cpu)) - return (int)per_cpu(x86_bios_cpu_apicid, mps_cpu); - else - return BAD_APICID; -} - -#ifdef CONFIG_SMP - -#define SMP_TRAMPOLINE_BASE 0x6000 - -extern int __cpu_disable(void); -extern void __cpu_die(unsigned int cpu); -extern void prefill_possible_map(void); -extern unsigned __cpuinitdata disabled_cpus; - -#define raw_smp_processor_id() read_pda(cpunumber) -#define cpu_physical_id(cpu) per_cpu(x86_cpu_to_apicid, cpu) - -#define stack_smp_processor_id() \ - ({ \ - struct thread_info *ti; \ - __asm__("andq %%rsp,%0; ":"=r" (ti) : "0" (CURRENT_MASK)); \ - ti->cpu; \ -}) - -/* - * On x86 all CPUs are mapped 1:1 to the APIC space. This simplifies - * scheduling and IPI sending and compresses data structures. - */ -static inline int num_booting_cpus(void) -{ - return cpus_weight(cpu_callout_map); -} - -extern void smp_send_reschedule(int cpu); - -#else /* CONFIG_SMP */ - -extern unsigned int boot_cpu_id; -#define cpu_physical_id(cpu) boot_cpu_id -#define stack_smp_processor_id() 0 - -#endif /* !CONFIG_SMP */ - -#define safe_smp_processor_id() smp_processor_id() - -static __inline int logical_smp_processor_id(void) -{ - /* we don't want to mark this access volatile - bad code generation */ - return GET_APIC_LOGICAL_ID(*(u32 *)(APIC_BASE + APIC_LDR)); -} - -static inline int hard_smp_processor_id(void) -{ - /* we don't want to mark this access volatile - bad code generation */ - return GET_APIC_ID(*(u32 *)(APIC_BASE + APIC_ID)); -} - -#endif - diff --git a/include/asm-x86/sparsemem.h b/include/asm-x86/sparsemem.h index fa58cd55411a..9bd48b0a534b 100644 --- a/include/asm-x86/sparsemem.h +++ b/include/asm-x86/sparsemem.h @@ -16,7 +16,7 @@ #ifdef CONFIG_X86_32 # ifdef CONFIG_X86_PAE -# define SECTION_SIZE_BITS 30 +# define SECTION_SIZE_BITS 29 # define MAX_PHYSADDR_BITS 36 # define MAX_PHYSMEM_BITS 36 # else @@ -26,8 +26,8 @@ # endif #else /* CONFIG_X86_32 */ # define SECTION_SIZE_BITS 27 /* matt - 128 is convenient right now */ -# define MAX_PHYSADDR_BITS 40 -# define MAX_PHYSMEM_BITS 40 +# define MAX_PHYSADDR_BITS 44 +# define MAX_PHYSMEM_BITS 44 #endif #endif /* CONFIG_SPARSEMEM */ diff --git a/include/asm-x86/spinlock.h b/include/asm-x86/spinlock.h index 23804c1890ff..bc6376f1bc5a 100644 --- a/include/asm-x86/spinlock.h +++ b/include/asm-x86/spinlock.h @@ -78,11 +78,11 @@ static inline int __raw_spin_is_contended(raw_spinlock_t *lock) return (((tmp >> 8) & 0xff) - (tmp & 0xff)) > 1; } -static inline void __raw_spin_lock(raw_spinlock_t *lock) +static __always_inline void __raw_spin_lock(raw_spinlock_t *lock) { short inc = 0x0100; - __asm__ __volatile__ ( + asm volatile ( LOCK_PREFIX "xaddw %w0, %1\n" "1:\t" "cmpb %h0, %b0\n\t" @@ -92,42 +92,40 @@ static inline void __raw_spin_lock(raw_spinlock_t *lock) /* don't need lfence here, because loads are in-order */ "jmp 1b\n" "2:" - :"+Q" (inc), "+m" (lock->slock) + : "+Q" (inc), "+m" (lock->slock) : - :"memory", "cc"); + : "memory", "cc"); } #define __raw_spin_lock_flags(lock, flags) __raw_spin_lock(lock) -static inline int __raw_spin_trylock(raw_spinlock_t *lock) +static __always_inline int __raw_spin_trylock(raw_spinlock_t *lock) { int tmp; short new; - asm volatile( - "movw %2,%w0\n\t" - "cmpb %h0,%b0\n\t" - "jne 1f\n\t" - "movw %w0,%w1\n\t" - "incb %h1\n\t" - "lock ; cmpxchgw %w1,%2\n\t" - "1:" - "sete %b1\n\t" - "movzbl %b1,%0\n\t" - :"=&a" (tmp), "=Q" (new), "+m" (lock->slock) - : - : "memory", "cc"); + asm volatile("movw %2,%w0\n\t" + "cmpb %h0,%b0\n\t" + "jne 1f\n\t" + "movw %w0,%w1\n\t" + "incb %h1\n\t" + "lock ; cmpxchgw %w1,%2\n\t" + "1:" + "sete %b1\n\t" + "movzbl %b1,%0\n\t" + : "=&a" (tmp), "=Q" (new), "+m" (lock->slock) + : + : "memory", "cc"); return tmp; } -static inline void __raw_spin_unlock(raw_spinlock_t *lock) +static __always_inline void __raw_spin_unlock(raw_spinlock_t *lock) { - __asm__ __volatile__( - UNLOCK_LOCK_PREFIX "incb %0" - :"+m" (lock->slock) - : - :"memory", "cc"); + asm volatile(UNLOCK_LOCK_PREFIX "incb %0" + : "+m" (lock->slock) + : + : "memory", "cc"); } #else static inline int __raw_spin_is_locked(raw_spinlock_t *lock) @@ -144,60 +142,57 @@ static inline int __raw_spin_is_contended(raw_spinlock_t *lock) return (((tmp >> 16) & 0xffff) - (tmp & 0xffff)) > 1; } -static inline void __raw_spin_lock(raw_spinlock_t *lock) +static __always_inline void __raw_spin_lock(raw_spinlock_t *lock) { int inc = 0x00010000; int tmp; - __asm__ __volatile__ ( - "lock ; xaddl %0, %1\n" - "movzwl %w0, %2\n\t" - "shrl $16, %0\n\t" - "1:\t" - "cmpl %0, %2\n\t" - "je 2f\n\t" - "rep ; nop\n\t" - "movzwl %1, %2\n\t" - /* don't need lfence here, because loads are in-order */ - "jmp 1b\n" - "2:" - :"+Q" (inc), "+m" (lock->slock), "=r" (tmp) - : - :"memory", "cc"); + asm volatile("lock ; xaddl %0, %1\n" + "movzwl %w0, %2\n\t" + "shrl $16, %0\n\t" + "1:\t" + "cmpl %0, %2\n\t" + "je 2f\n\t" + "rep ; nop\n\t" + "movzwl %1, %2\n\t" + /* don't need lfence here, because loads are in-order */ + "jmp 1b\n" + "2:" + : "+Q" (inc), "+m" (lock->slock), "=r" (tmp) + : + : "memory", "cc"); } #define __raw_spin_lock_flags(lock, flags) __raw_spin_lock(lock) -static inline int __raw_spin_trylock(raw_spinlock_t *lock) +static __always_inline int __raw_spin_trylock(raw_spinlock_t *lock) { int tmp; int new; - asm volatile( - "movl %2,%0\n\t" - "movl %0,%1\n\t" - "roll $16, %0\n\t" - "cmpl %0,%1\n\t" - "jne 1f\n\t" - "addl $0x00010000, %1\n\t" - "lock ; cmpxchgl %1,%2\n\t" - "1:" - "sete %b1\n\t" - "movzbl %b1,%0\n\t" - :"=&a" (tmp), "=r" (new), "+m" (lock->slock) - : - : "memory", "cc"); + asm volatile("movl %2,%0\n\t" + "movl %0,%1\n\t" + "roll $16, %0\n\t" + "cmpl %0,%1\n\t" + "jne 1f\n\t" + "addl $0x00010000, %1\n\t" + "lock ; cmpxchgl %1,%2\n\t" + "1:" + "sete %b1\n\t" + "movzbl %b1,%0\n\t" + : "=&a" (tmp), "=r" (new), "+m" (lock->slock) + : + : "memory", "cc"); return tmp; } -static inline void __raw_spin_unlock(raw_spinlock_t *lock) +static __always_inline void __raw_spin_unlock(raw_spinlock_t *lock) { - __asm__ __volatile__( - UNLOCK_LOCK_PREFIX "incw %0" - :"+m" (lock->slock) - : - :"memory", "cc"); + asm volatile(UNLOCK_LOCK_PREFIX "incw %0" + : "+m" (lock->slock) + : + : "memory", "cc"); } #endif diff --git a/include/asm-x86/srat.h b/include/asm-x86/srat.h index 165ab4bdc02b..f4bba131d068 100644 --- a/include/asm-x86/srat.h +++ b/include/asm-x86/srat.h @@ -1,5 +1,5 @@ /* - * Some of the code in this file has been gleaned from the 64 bit + * Some of the code in this file has been gleaned from the 64 bit * discontigmem support code base. * * Copyright (C) 2002, IBM Corp. diff --git a/include/asm-x86/string_32.h b/include/asm-x86/string_32.h index c5d13a86dea7..b49369ad9a61 100644 --- a/include/asm-x86/string_32.h +++ b/include/asm-x86/string_32.h @@ -3,7 +3,7 @@ #ifdef __KERNEL__ -/* Let gcc decide wether to inline or use the out of line functions */ +/* Let gcc decide whether to inline or use the out of line functions */ #define __HAVE_ARCH_STRCPY extern char *strcpy(char *dest, const char *src); diff --git a/include/asm-x86/string_64.h b/include/asm-x86/string_64.h index e583da7918fb..52b5ab383395 100644 --- a/include/asm-x86/string_64.h +++ b/include/asm-x86/string_64.h @@ -3,26 +3,24 @@ #ifdef __KERNEL__ -/* Written 2002 by Andi Kleen */ +/* Written 2002 by Andi Kleen */ -/* Only used for special circumstances. Stolen from i386/string.h */ -static __always_inline void * -__inline_memcpy(void * to, const void * from, size_t n) +/* Only used for special circumstances. Stolen from i386/string.h */ +static __always_inline void *__inline_memcpy(void *to, const void *from, size_t n) { -unsigned long d0, d1, d2; -__asm__ __volatile__( - "rep ; movsl\n\t" - "testb $2,%b4\n\t" - "je 1f\n\t" - "movsw\n" - "1:\ttestb $1,%b4\n\t" - "je 2f\n\t" - "movsb\n" - "2:" - : "=&c" (d0), "=&D" (d1), "=&S" (d2) - :"0" (n/4), "q" (n),"1" ((long) to),"2" ((long) from) - : "memory"); -return (to); + unsigned long d0, d1, d2; + asm volatile("rep ; movsl\n\t" + "testb $2,%b4\n\t" + "je 1f\n\t" + "movsw\n" + "1:\ttestb $1,%b4\n\t" + "je 2f\n\t" + "movsb\n" + "2:" + : "=&c" (d0), "=&D" (d1), "=&S" (d2) + : "0" (n / 4), "q" (n), "1" ((long)to), "2" ((long)from) + : "memory"); + return to; } /* Even with __builtin_ the compiler may decide to use the out of line @@ -32,28 +30,30 @@ return (to); #if (__GNUC__ == 4 && __GNUC_MINOR__ >= 3) || __GNUC__ > 4 extern void *memcpy(void *to, const void *from, size_t len); #else -extern void *__memcpy(void *to, const void *from, size_t len); -#define memcpy(dst,src,len) \ - ({ size_t __len = (len); \ - void *__ret; \ - if (__builtin_constant_p(len) && __len >= 64) \ - __ret = __memcpy((dst),(src),__len); \ - else \ - __ret = __builtin_memcpy((dst),(src),__len); \ - __ret; }) +extern void *__memcpy(void *to, const void *from, size_t len); +#define memcpy(dst, src, len) \ +({ \ + size_t __len = (len); \ + void *__ret; \ + if (__builtin_constant_p(len) && __len >= 64) \ + __ret = __memcpy((dst), (src), __len); \ + else \ + __ret = __builtin_memcpy((dst), (src), __len); \ + __ret; \ +}) #endif #define __HAVE_ARCH_MEMSET void *memset(void *s, int c, size_t n); #define __HAVE_ARCH_MEMMOVE -void * memmove(void * dest,const void *src,size_t count); +void *memmove(void *dest, const void *src, size_t count); -int memcmp(const void * cs,const void * ct,size_t count); -size_t strlen(const char * s); -char *strcpy(char * dest,const char *src); -char *strcat(char * dest, const char * src); -int strcmp(const char * cs,const char * ct); +int memcmp(const void *cs, const void *ct, size_t count); +size_t strlen(const char *s); +char *strcpy(char *dest, const char *src); +char *strcat(char *dest, const char *src); +int strcmp(const char *cs, const char *ct); #endif /* __KERNEL__ */ diff --git a/include/asm-x86/suspend_32.h b/include/asm-x86/suspend_32.h index 1bbda3ad7796..24e1c080aa8a 100644 --- a/include/asm-x86/suspend_32.h +++ b/include/asm-x86/suspend_32.h @@ -10,7 +10,7 @@ static inline int arch_prepare_suspend(void) { return 0; } /* image of the saved processor state */ struct saved_context { - u16 es, fs, gs, ss; + u16 es, fs, gs, ss; unsigned long cr0, cr2, cr3, cr4; struct desc_ptr gdt; struct desc_ptr idt; @@ -32,11 +32,11 @@ extern unsigned long saved_edi; static inline void acpi_save_register_state(unsigned long return_point) { saved_eip = return_point; - asm volatile ("movl %%esp,%0" : "=m" (saved_esp)); - asm volatile ("movl %%ebp,%0" : "=m" (saved_ebp)); - asm volatile ("movl %%ebx,%0" : "=m" (saved_ebx)); - asm volatile ("movl %%edi,%0" : "=m" (saved_edi)); - asm volatile ("movl %%esi,%0" : "=m" (saved_esi)); + asm volatile("movl %%esp,%0" : "=m" (saved_esp)); + asm volatile("movl %%ebp,%0" : "=m" (saved_ebp)); + asm volatile("movl %%ebx,%0" : "=m" (saved_ebx)); + asm volatile("movl %%edi,%0" : "=m" (saved_edi)); + asm volatile("movl %%esi,%0" : "=m" (saved_esi)); } #define acpi_restore_register_state() do {} while (0) diff --git a/include/asm-x86/suspend_64.h b/include/asm-x86/suspend_64.h index 2eb92cb81a0d..dc3262b43072 100644 --- a/include/asm-x86/suspend_64.h +++ b/include/asm-x86/suspend_64.h @@ -9,8 +9,7 @@ #include <asm/desc.h> #include <asm/i387.h> -static inline int -arch_prepare_suspend(void) +static inline int arch_prepare_suspend(void) { return 0; } @@ -25,7 +24,7 @@ arch_prepare_suspend(void) */ struct saved_context { struct pt_regs regs; - u16 ds, es, fs, gs, ss; + u16 ds, es, fs, gs, ss; unsigned long gs_base, gs_kernel_base, fs_base; unsigned long cr0, cr2, cr3, cr4, cr8; unsigned long efer; diff --git a/include/asm-x86/swiotlb.h b/include/asm-x86/swiotlb.h index f9c589539a82..f5d9e74b1e4a 100644 --- a/include/asm-x86/swiotlb.h +++ b/include/asm-x86/swiotlb.h @@ -8,15 +8,15 @@ extern dma_addr_t swiotlb_map_single(struct device *hwdev, void *ptr, size_t size, int dir); extern void *swiotlb_alloc_coherent(struct device *hwdev, size_t size, - dma_addr_t *dma_handle, gfp_t flags); + dma_addr_t *dma_handle, gfp_t flags); extern void swiotlb_unmap_single(struct device *hwdev, dma_addr_t dev_addr, - size_t size, int dir); + size_t size, int dir); extern void swiotlb_sync_single_for_cpu(struct device *hwdev, - dma_addr_t dev_addr, - size_t size, int dir); + dma_addr_t dev_addr, + size_t size, int dir); extern void swiotlb_sync_single_for_device(struct device *hwdev, - dma_addr_t dev_addr, - size_t size, int dir); + dma_addr_t dev_addr, + size_t size, int dir); extern void swiotlb_sync_single_range_for_cpu(struct device *hwdev, dma_addr_t dev_addr, unsigned long offset, @@ -26,18 +26,18 @@ extern void swiotlb_sync_single_range_for_device(struct device *hwdev, unsigned long offset, size_t size, int dir); extern void swiotlb_sync_sg_for_cpu(struct device *hwdev, - struct scatterlist *sg, int nelems, - int dir); + struct scatterlist *sg, int nelems, + int dir); extern void swiotlb_sync_sg_for_device(struct device *hwdev, - struct scatterlist *sg, int nelems, - int dir); + struct scatterlist *sg, int nelems, + int dir); extern int swiotlb_map_sg(struct device *hwdev, struct scatterlist *sg, - int nents, int direction); + int nents, int direction); extern void swiotlb_unmap_sg(struct device *hwdev, struct scatterlist *sg, - int nents, int direction); + int nents, int direction); extern int swiotlb_dma_mapping_error(dma_addr_t dma_addr); -extern void swiotlb_free_coherent (struct device *hwdev, size_t size, - void *vaddr, dma_addr_t dma_handle); +extern void swiotlb_free_coherent(struct device *hwdev, size_t size, + void *vaddr, dma_addr_t dma_handle); extern int swiotlb_dma_supported(struct device *hwdev, u64 mask); extern void swiotlb_init(void); diff --git a/include/asm-x86/sync_bitops.h b/include/asm-x86/sync_bitops.h index 6b775c905666..b47a1d0b8a83 100644 --- a/include/asm-x86/sync_bitops.h +++ b/include/asm-x86/sync_bitops.h @@ -13,7 +13,7 @@ * bit 0 is the LSB of addr; bit 32 is the LSB of (addr+1). */ -#define ADDR (*(volatile long *) addr) +#define ADDR (*(volatile long *)addr) /** * sync_set_bit - Atomically set a bit in memory @@ -26,12 +26,12 @@ * Note that @nr may be almost arbitrarily large; this function is not * restricted to acting on a single-word quantity. */ -static inline void sync_set_bit(int nr, volatile unsigned long * addr) +static inline void sync_set_bit(int nr, volatile unsigned long *addr) { - __asm__ __volatile__("lock; btsl %1,%0" - :"+m" (ADDR) - :"Ir" (nr) - : "memory"); + asm volatile("lock; btsl %1,%0" + : "+m" (ADDR) + : "Ir" (nr) + : "memory"); } /** @@ -44,12 +44,12 @@ static inline void sync_set_bit(int nr, volatile unsigned long * addr) * you should call smp_mb__before_clear_bit() and/or smp_mb__after_clear_bit() * in order to ensure changes are visible on other processors. */ -static inline void sync_clear_bit(int nr, volatile unsigned long * addr) +static inline void sync_clear_bit(int nr, volatile unsigned long *addr) { - __asm__ __volatile__("lock; btrl %1,%0" - :"+m" (ADDR) - :"Ir" (nr) - : "memory"); + asm volatile("lock; btrl %1,%0" + : "+m" (ADDR) + : "Ir" (nr) + : "memory"); } /** @@ -61,12 +61,12 @@ static inline void sync_clear_bit(int nr, volatile unsigned long * addr) * Note that @nr may be almost arbitrarily large; this function is not * restricted to acting on a single-word quantity. */ -static inline void sync_change_bit(int nr, volatile unsigned long * addr) +static inline void sync_change_bit(int nr, volatile unsigned long *addr) { - __asm__ __volatile__("lock; btcl %1,%0" - :"+m" (ADDR) - :"Ir" (nr) - : "memory"); + asm volatile("lock; btcl %1,%0" + : "+m" (ADDR) + : "Ir" (nr) + : "memory"); } /** @@ -77,13 +77,13 @@ static inline void sync_change_bit(int nr, volatile unsigned long * addr) * This operation is atomic and cannot be reordered. * It also implies a memory barrier. */ -static inline int sync_test_and_set_bit(int nr, volatile unsigned long * addr) +static inline int sync_test_and_set_bit(int nr, volatile unsigned long *addr) { int oldbit; - __asm__ __volatile__("lock; btsl %2,%1\n\tsbbl %0,%0" - :"=r" (oldbit),"+m" (ADDR) - :"Ir" (nr) : "memory"); + asm volatile("lock; btsl %2,%1\n\tsbbl %0,%0" + : "=r" (oldbit), "+m" (ADDR) + : "Ir" (nr) : "memory"); return oldbit; } @@ -95,13 +95,13 @@ static inline int sync_test_and_set_bit(int nr, volatile unsigned long * addr) * This operation is atomic and cannot be reordered. * It also implies a memory barrier. */ -static inline int sync_test_and_clear_bit(int nr, volatile unsigned long * addr) +static inline int sync_test_and_clear_bit(int nr, volatile unsigned long *addr) { int oldbit; - __asm__ __volatile__("lock; btrl %2,%1\n\tsbbl %0,%0" - :"=r" (oldbit),"+m" (ADDR) - :"Ir" (nr) : "memory"); + asm volatile("lock; btrl %2,%1\n\tsbbl %0,%0" + : "=r" (oldbit), "+m" (ADDR) + : "Ir" (nr) : "memory"); return oldbit; } @@ -113,36 +113,17 @@ static inline int sync_test_and_clear_bit(int nr, volatile unsigned long * addr) * This operation is atomic and cannot be reordered. * It also implies a memory barrier. */ -static inline int sync_test_and_change_bit(int nr, volatile unsigned long* addr) +static inline int sync_test_and_change_bit(int nr, volatile unsigned long *addr) { int oldbit; - __asm__ __volatile__("lock; btcl %2,%1\n\tsbbl %0,%0" - :"=r" (oldbit),"+m" (ADDR) - :"Ir" (nr) : "memory"); + asm volatile("lock; btcl %2,%1\n\tsbbl %0,%0" + : "=r" (oldbit), "+m" (ADDR) + : "Ir" (nr) : "memory"); return oldbit; } -static __always_inline int sync_constant_test_bit(int nr, const volatile unsigned long *addr) -{ - return ((1UL << (nr & 31)) & - (((const volatile unsigned int *)addr)[nr >> 5])) != 0; -} - -static inline int sync_var_test_bit(int nr, const volatile unsigned long * addr) -{ - int oldbit; - - __asm__ __volatile__("btl %2,%1\n\tsbbl %0,%0" - :"=r" (oldbit) - :"m" (ADDR),"Ir" (nr)); - return oldbit; -} - -#define sync_test_bit(nr,addr) \ - (__builtin_constant_p(nr) ? \ - sync_constant_test_bit((nr),(addr)) : \ - sync_var_test_bit((nr),(addr))) +#define sync_test_bit(nr, addr) test_bit(nr, addr) #undef ADDR diff --git a/include/asm-x86/system.h b/include/asm-x86/system.h index 9cff02ffe6c2..a2f04cd79b29 100644 --- a/include/asm-x86/system.h +++ b/include/asm-x86/system.h @@ -27,22 +27,44 @@ struct task_struct *__switch_to(struct task_struct *prev, * Saving eflags is important. It switches not only IOPL between tasks, * it also protects other tasks from NT leaking through sysenter etc. */ -#define switch_to(prev, next, last) do { \ - unsigned long esi, edi; \ - asm volatile("pushfl\n\t" /* Save flags */ \ - "pushl %%ebp\n\t" \ - "movl %%esp,%0\n\t" /* save ESP */ \ - "movl %5,%%esp\n\t" /* restore ESP */ \ - "movl $1f,%1\n\t" /* save EIP */ \ - "pushl %6\n\t" /* restore EIP */ \ - "jmp __switch_to\n" \ +#define switch_to(prev, next, last) \ +do { \ + /* \ + * Context-switching clobbers all registers, so we clobber \ + * them explicitly, via unused output variables. \ + * (EAX and EBP is not listed because EBP is saved/restored \ + * explicitly for wchan access and EAX is the return value of \ + * __switch_to()) \ + */ \ + unsigned long ebx, ecx, edx, esi, edi; \ + \ + asm volatile("pushfl\n\t" /* save flags */ \ + "pushl %%ebp\n\t" /* save EBP */ \ + "movl %%esp,%[prev_sp]\n\t" /* save ESP */ \ + "movl %[next_sp],%%esp\n\t" /* restore ESP */ \ + "movl $1f,%[prev_ip]\n\t" /* save EIP */ \ + "pushl %[next_ip]\n\t" /* restore EIP */ \ + "jmp __switch_to\n" /* regparm call */ \ "1:\t" \ - "popl %%ebp\n\t" \ - "popfl" \ - :"=m" (prev->thread.sp), "=m" (prev->thread.ip), \ - "=a" (last), "=S" (esi), "=D" (edi) \ - :"m" (next->thread.sp), "m" (next->thread.ip), \ - "2" (prev), "d" (next)); \ + "popl %%ebp\n\t" /* restore EBP */ \ + "popfl\n" /* restore flags */ \ + \ + /* output parameters */ \ + : [prev_sp] "=m" (prev->thread.sp), \ + [prev_ip] "=m" (prev->thread.ip), \ + "=a" (last), \ + \ + /* clobbered output registers: */ \ + "=b" (ebx), "=c" (ecx), "=d" (edx), \ + "=S" (esi), "=D" (edi) \ + \ + /* input parameters: */ \ + : [next_sp] "m" (next->thread.sp), \ + [next_ip] "m" (next->thread.ip), \ + \ + /* regparm parameters for __switch_to(): */ \ + [prev] "a" (prev), \ + [next] "d" (next)); \ } while (0) /* @@ -122,35 +144,34 @@ extern void load_gs_index(unsigned); */ #define loadsegment(seg, value) \ asm volatile("\n" \ - "1:\t" \ - "movl %k0,%%" #seg "\n" \ - "2:\n" \ - ".section .fixup,\"ax\"\n" \ - "3:\t" \ - "movl %k1, %%" #seg "\n\t" \ - "jmp 2b\n" \ - ".previous\n" \ - _ASM_EXTABLE(1b,3b) \ - : :"r" (value), "r" (0)) + "1:\t" \ + "movl %k0,%%" #seg "\n" \ + "2:\n" \ + ".section .fixup,\"ax\"\n" \ + "3:\t" \ + "movl %k1, %%" #seg "\n\t" \ + "jmp 2b\n" \ + ".previous\n" \ + _ASM_EXTABLE(1b,3b) \ + : :"r" (value), "r" (0)) /* * Save a segment register away */ -#define savesegment(seg, value) \ +#define savesegment(seg, value) \ asm volatile("mov %%" #seg ",%0":"=rm" (value)) static inline unsigned long get_limit(unsigned long segment) { unsigned long __limit; - __asm__("lsll %1,%0" - :"=r" (__limit):"r" (segment)); - return __limit+1; + asm("lsll %1,%0" : "=r" (__limit) : "r" (segment)); + return __limit + 1; } static inline void native_clts(void) { - asm volatile ("clts"); + asm volatile("clts"); } /* @@ -165,43 +186,43 @@ static unsigned long __force_order; static inline unsigned long native_read_cr0(void) { unsigned long val; - asm volatile("mov %%cr0,%0\n\t" :"=r" (val), "=m" (__force_order)); + asm volatile("mov %%cr0,%0\n\t" : "=r" (val), "=m" (__force_order)); return val; } static inline void native_write_cr0(unsigned long val) { - asm volatile("mov %0,%%cr0": :"r" (val), "m" (__force_order)); + asm volatile("mov %0,%%cr0": : "r" (val), "m" (__force_order)); } static inline unsigned long native_read_cr2(void) { unsigned long val; - asm volatile("mov %%cr2,%0\n\t" :"=r" (val), "=m" (__force_order)); + asm volatile("mov %%cr2,%0\n\t" : "=r" (val), "=m" (__force_order)); return val; } static inline void native_write_cr2(unsigned long val) { - asm volatile("mov %0,%%cr2": :"r" (val), "m" (__force_order)); + asm volatile("mov %0,%%cr2": : "r" (val), "m" (__force_order)); } static inline unsigned long native_read_cr3(void) { unsigned long val; - asm volatile("mov %%cr3,%0\n\t" :"=r" (val), "=m" (__force_order)); + asm volatile("mov %%cr3,%0\n\t" : "=r" (val), "=m" (__force_order)); return val; } static inline void native_write_cr3(unsigned long val) { - asm volatile("mov %0,%%cr3": :"r" (val), "m" (__force_order)); + asm volatile("mov %0,%%cr3": : "r" (val), "m" (__force_order)); } static inline unsigned long native_read_cr4(void) { unsigned long val; - asm volatile("mov %%cr4,%0\n\t" :"=r" (val), "=m" (__force_order)); + asm volatile("mov %%cr4,%0\n\t" : "=r" (val), "=m" (__force_order)); return val; } @@ -213,7 +234,7 @@ static inline unsigned long native_read_cr4_safe(void) #ifdef CONFIG_X86_32 asm volatile("1: mov %%cr4, %0\n" "2:\n" - _ASM_EXTABLE(1b,2b) + _ASM_EXTABLE(1b, 2b) : "=r" (val), "=m" (__force_order) : "0" (0)); #else val = native_read_cr4(); @@ -223,7 +244,7 @@ static inline unsigned long native_read_cr4_safe(void) static inline void native_write_cr4(unsigned long val) { - asm volatile("mov %0,%%cr4": :"r" (val), "m" (__force_order)); + asm volatile("mov %0,%%cr4": : "r" (val), "m" (__force_order)); } #ifdef CONFIG_X86_64 @@ -244,6 +265,7 @@ static inline void native_wbinvd(void) { asm volatile("wbinvd": : :"memory"); } + #ifdef CONFIG_PARAVIRT #include <asm/paravirt.h> #else @@ -276,7 +298,7 @@ static inline void clflush(volatile void *__p) asm volatile("clflush %0" : "+m" (*(volatile char __force *)__p)); } -#define nop() __asm__ __volatile__ ("nop") +#define nop() asm volatile ("nop") void disable_hlt(void); void enable_hlt(void); @@ -296,16 +318,7 @@ void default_idle(void); */ #ifdef CONFIG_X86_32 /* - * For now, "wmb()" doesn't actually do anything, as all - * Intel CPU's follow what Intel calls a *Processor Order*, - * in which all writes are seen in the program order even - * outside the CPU. - * - * I expect future Intel CPU's to have a weaker ordering, - * but I'd also expect them to finally get their act together - * and add some real memory barriers if so. - * - * Some non intel clones support out of order store. wmb() ceases to be a + * Some non-Intel clones support out of order store. wmb() ceases to be a * nop for these. */ #define mb() alternative("lock; addl $0,0(%%esp)", "mfence", X86_FEATURE_XMM2) @@ -384,7 +397,7 @@ void default_idle(void); # define smp_wmb() barrier() #endif #define smp_read_barrier_depends() read_barrier_depends() -#define set_mb(var, value) do { (void) xchg(&var, value); } while (0) +#define set_mb(var, value) do { (void)xchg(&var, value); } while (0) #else #define smp_mb() barrier() #define smp_rmb() barrier() diff --git a/include/asm-x86/tce.h b/include/asm-x86/tce.h index cd955d3d112f..b1a4ea00df78 100644 --- a/include/asm-x86/tce.h +++ b/include/asm-x86/tce.h @@ -39,7 +39,7 @@ struct iommu_table; #define TCE_RPN_MASK 0x0000fffffffff000ULL extern void tce_build(struct iommu_table *tbl, unsigned long index, - unsigned int npages, unsigned long uaddr, int direction); + unsigned int npages, unsigned long uaddr, int direction); extern void tce_free(struct iommu_table *tbl, long index, unsigned int npages); extern void * __init alloc_tce_table(void); extern void __init free_tce_table(void *tbl); diff --git a/include/asm-x86/thread_info_32.h b/include/asm-x86/thread_info_32.h index 5bd508260ffb..4e053fa561a9 100644 --- a/include/asm-x86/thread_info_32.h +++ b/include/asm-x86/thread_info_32.h @@ -20,7 +20,8 @@ * low level task data that entry.S needs immediate access to * - this struct should fit entirely inside of one cache line * - this struct shares the supervisor stack pages - * - if the contents of this structure are changed, the assembly constants must also be changed + * - if the contents of this structure are changed, + * the assembly constants must also be changed */ #ifndef __ASSEMBLY__ @@ -30,18 +31,16 @@ struct thread_info { unsigned long flags; /* low level flags */ unsigned long status; /* thread-synchronous flags */ __u32 cpu; /* current CPU */ - int preempt_count; /* 0 => preemptable, <0 => BUG */ - - + int preempt_count; /* 0 => preemptable, + <0 => BUG */ mm_segment_t addr_limit; /* thread address space: - 0-0xBFFFFFFF for user-thead - 0-0xFFFFFFFF for kernel-thread + 0-0xBFFFFFFF user-thread + 0-0xFFFFFFFF kernel-thread */ void *sysenter_return; struct restart_block restart_block; - - unsigned long previous_esp; /* ESP of the previous stack in case - of nested (IRQ) stacks + unsigned long previous_esp; /* ESP of the previous stack in + case of nested (IRQ) stacks */ __u8 supervisor_stack[0]; }; @@ -90,15 +89,16 @@ register unsigned long current_stack_pointer asm("esp") __used; /* how to get the thread information struct from C */ static inline struct thread_info *current_thread_info(void) { - return (struct thread_info *)(current_stack_pointer & ~(THREAD_SIZE - 1)); + return (struct thread_info *) + (current_stack_pointer & ~(THREAD_SIZE - 1)); } /* thread information allocation */ #ifdef CONFIG_DEBUG_STACK_USAGE -#define alloc_thread_info(tsk) ((struct thread_info *) \ - __get_free_pages(GFP_KERNEL| __GFP_ZERO, get_order(THREAD_SIZE))) +#define alloc_thread_info(tsk) ((struct thread_info *) \ + __get_free_pages(GFP_KERNEL | __GFP_ZERO, get_order(THREAD_SIZE))) #else -#define alloc_thread_info(tsk) ((struct thread_info *) \ +#define alloc_thread_info(tsk) ((struct thread_info *) \ __get_free_pages(GFP_KERNEL, get_order(THREAD_SIZE))) #endif @@ -107,7 +107,7 @@ static inline struct thread_info *current_thread_info(void) #else /* !__ASSEMBLY__ */ /* how to get the thread information struct from ASM */ -#define GET_THREAD_INFO(reg) \ +#define GET_THREAD_INFO(reg) \ movl $-THREAD_SIZE, reg; \ andl %esp, reg @@ -119,14 +119,16 @@ static inline struct thread_info *current_thread_info(void) /* * thread information flags - * - these are process state flags that various assembly files may need to access + * - these are process state flags that various + * assembly files may need to access * - pending work-to-be-done flags are in LSW * - other flags in MSW */ #define TIF_SYSCALL_TRACE 0 /* syscall trace active */ #define TIF_SIGPENDING 1 /* signal pending */ #define TIF_NEED_RESCHED 2 /* rescheduling necessary */ -#define TIF_SINGLESTEP 3 /* restore singlestep on return to user mode */ +#define TIF_SINGLESTEP 3 /* restore singlestep on return to + user mode */ #define TIF_IRET 4 /* return with iret */ #define TIF_SYSCALL_EMU 5 /* syscall emulation active */ #define TIF_SYSCALL_AUDIT 6 /* syscall auditing active */ @@ -143,36 +145,36 @@ static inline struct thread_info *current_thread_info(void) #define TIF_DS_AREA_MSR 23 /* uses thread_struct.ds_area_msr */ #define TIF_BTS_TRACE_TS 24 /* record scheduling event timestamps */ -#define _TIF_SYSCALL_TRACE (1<<TIF_SYSCALL_TRACE) -#define _TIF_SIGPENDING (1<<TIF_SIGPENDING) -#define _TIF_NEED_RESCHED (1<<TIF_NEED_RESCHED) -#define _TIF_SINGLESTEP (1<<TIF_SINGLESTEP) -#define _TIF_IRET (1<<TIF_IRET) -#define _TIF_SYSCALL_EMU (1<<TIF_SYSCALL_EMU) -#define _TIF_SYSCALL_AUDIT (1<<TIF_SYSCALL_AUDIT) -#define _TIF_SECCOMP (1<<TIF_SECCOMP) -#define _TIF_RESTORE_SIGMASK (1<<TIF_RESTORE_SIGMASK) -#define _TIF_HRTICK_RESCHED (1<<TIF_HRTICK_RESCHED) -#define _TIF_DEBUG (1<<TIF_DEBUG) -#define _TIF_IO_BITMAP (1<<TIF_IO_BITMAP) -#define _TIF_FREEZE (1<<TIF_FREEZE) -#define _TIF_NOTSC (1<<TIF_NOTSC) -#define _TIF_FORCED_TF (1<<TIF_FORCED_TF) -#define _TIF_DEBUGCTLMSR (1<<TIF_DEBUGCTLMSR) -#define _TIF_DS_AREA_MSR (1<<TIF_DS_AREA_MSR) -#define _TIF_BTS_TRACE_TS (1<<TIF_BTS_TRACE_TS) +#define _TIF_SYSCALL_TRACE (1 << TIF_SYSCALL_TRACE) +#define _TIF_SIGPENDING (1 << TIF_SIGPENDING) +#define _TIF_NEED_RESCHED (1 << TIF_NEED_RESCHED) +#define _TIF_SINGLESTEP (1 << TIF_SINGLESTEP) +#define _TIF_IRET (1 << TIF_IRET) +#define _TIF_SYSCALL_EMU (1 << TIF_SYSCALL_EMU) +#define _TIF_SYSCALL_AUDIT (1 << TIF_SYSCALL_AUDIT) +#define _TIF_SECCOMP (1 << TIF_SECCOMP) +#define _TIF_RESTORE_SIGMASK (1 << TIF_RESTORE_SIGMASK) +#define _TIF_HRTICK_RESCHED (1 << TIF_HRTICK_RESCHED) +#define _TIF_DEBUG (1 << TIF_DEBUG) +#define _TIF_IO_BITMAP (1 << TIF_IO_BITMAP) +#define _TIF_FREEZE (1 << TIF_FREEZE) +#define _TIF_NOTSC (1 << TIF_NOTSC) +#define _TIF_FORCED_TF (1 << TIF_FORCED_TF) +#define _TIF_DEBUGCTLMSR (1 << TIF_DEBUGCTLMSR) +#define _TIF_DS_AREA_MSR (1 << TIF_DS_AREA_MSR) +#define _TIF_BTS_TRACE_TS (1 << TIF_BTS_TRACE_TS) /* work to do on interrupt/exception return */ -#define _TIF_WORK_MASK \ - (0x0000FFFF & ~(_TIF_SYSCALL_TRACE | _TIF_SYSCALL_AUDIT | \ - _TIF_SECCOMP | _TIF_SYSCALL_EMU)) +#define _TIF_WORK_MASK \ + (0x0000FFFF & ~(_TIF_SYSCALL_TRACE | _TIF_SYSCALL_AUDIT | \ + _TIF_SECCOMP | _TIF_SYSCALL_EMU)) /* work to do on any return to u-space */ #define _TIF_ALLWORK_MASK (0x0000FFFF & ~_TIF_SECCOMP) /* flags to check in __switch_to() */ -#define _TIF_WORK_CTXSW \ - (_TIF_IO_BITMAP | _TIF_NOTSC | _TIF_DEBUGCTLMSR | \ - _TIF_DS_AREA_MSR | _TIF_BTS_TRACE_TS) +#define _TIF_WORK_CTXSW \ + (_TIF_IO_BITMAP | _TIF_NOTSC | _TIF_DEBUGCTLMSR | \ + _TIF_DS_AREA_MSR | _TIF_BTS_TRACE_TS) #define _TIF_WORK_CTXSW_PREV _TIF_WORK_CTXSW #define _TIF_WORK_CTXSW_NEXT (_TIF_WORK_CTXSW | _TIF_DEBUG) @@ -184,8 +186,10 @@ static inline struct thread_info *current_thread_info(void) * ever touches our thread-synchronous status, so we don't * have to worry about atomic accesses. */ -#define TS_USEDFPU 0x0001 /* FPU was used by this task this quantum (SMP) */ -#define TS_POLLING 0x0002 /* True if in idle loop and not sleeping */ +#define TS_USEDFPU 0x0001 /* FPU was used by this task + this quantum (SMP) */ +#define TS_POLLING 0x0002 /* True if in idle loop + and not sleeping */ #define tsk_is_polling(t) (task_thread_info(t)->status & TS_POLLING) diff --git a/include/asm-x86/thread_info_64.h b/include/asm-x86/thread_info_64.h index 6c9b214b8fc3..1e5c6f6152cd 100644 --- a/include/asm-x86/thread_info_64.h +++ b/include/asm-x86/thread_info_64.h @@ -29,9 +29,9 @@ struct thread_info { __u32 flags; /* low level flags */ __u32 status; /* thread synchronous flags */ __u32 cpu; /* current CPU */ - int preempt_count; /* 0 => preemptable, <0 => BUG */ - - mm_segment_t addr_limit; + int preempt_count; /* 0 => preemptable, + <0 => BUG */ + mm_segment_t addr_limit; struct restart_block restart_block; #ifdef CONFIG_IA32_EMULATION void __user *sysenter_return; @@ -61,17 +61,17 @@ struct thread_info { #define init_stack (init_thread_union.stack) static inline struct thread_info *current_thread_info(void) -{ +{ struct thread_info *ti; ti = (void *)(read_pda(kernelstack) + PDA_STACKOFFSET - THREAD_SIZE); - return ti; + return ti; } /* do not use in interrupt context */ static inline struct thread_info *stack_thread_info(void) { struct thread_info *ti; - __asm__("andq %%rsp,%0; ":"=r" (ti) : "0" (~(THREAD_SIZE - 1))); + asm("andq %%rsp,%0; " : "=r" (ti) : "0" (~(THREAD_SIZE - 1))); return ti; } @@ -82,8 +82,8 @@ static inline struct thread_info *stack_thread_info(void) #define THREAD_FLAGS GFP_KERNEL #endif -#define alloc_thread_info(tsk) \ - ((struct thread_info *) __get_free_pages(THREAD_FLAGS, THREAD_ORDER)) +#define alloc_thread_info(tsk) \ + ((struct thread_info *)__get_free_pages(THREAD_FLAGS, THREAD_ORDER)) #define free_thread_info(ti) free_pages((unsigned long) (ti), THREAD_ORDER) @@ -98,7 +98,8 @@ static inline struct thread_info *stack_thread_info(void) /* * thread information flags - * - these are process state flags that various assembly files may need to access + * - these are process state flags that various assembly files + * may need to access * - pending work-to-be-done flags are in LSW * - other flags in MSW * Warning: layout of LSW is hardcoded in entry.S @@ -114,7 +115,7 @@ static inline struct thread_info *stack_thread_info(void) #define TIF_MCE_NOTIFY 10 /* notify userspace of an MCE */ #define TIF_HRTICK_RESCHED 11 /* reprogram hrtick timer */ /* 16 free */ -#define TIF_IA32 17 /* 32bit process */ +#define TIF_IA32 17 /* 32bit process */ #define TIF_FORK 18 /* ret_from_fork */ #define TIF_ABI_PENDING 19 #define TIF_MEMDIE 20 @@ -126,39 +127,40 @@ static inline struct thread_info *stack_thread_info(void) #define TIF_DS_AREA_MSR 26 /* uses thread_struct.ds_area_msr */ #define TIF_BTS_TRACE_TS 27 /* record scheduling event timestamps */ -#define _TIF_SYSCALL_TRACE (1<<TIF_SYSCALL_TRACE) -#define _TIF_SIGPENDING (1<<TIF_SIGPENDING) -#define _TIF_SINGLESTEP (1<<TIF_SINGLESTEP) -#define _TIF_NEED_RESCHED (1<<TIF_NEED_RESCHED) -#define _TIF_IRET (1<<TIF_IRET) -#define _TIF_SYSCALL_AUDIT (1<<TIF_SYSCALL_AUDIT) -#define _TIF_SECCOMP (1<<TIF_SECCOMP) -#define _TIF_RESTORE_SIGMASK (1<<TIF_RESTORE_SIGMASK) -#define _TIF_MCE_NOTIFY (1<<TIF_MCE_NOTIFY) -#define _TIF_HRTICK_RESCHED (1<<TIF_HRTICK_RESCHED) -#define _TIF_IA32 (1<<TIF_IA32) -#define _TIF_FORK (1<<TIF_FORK) -#define _TIF_ABI_PENDING (1<<TIF_ABI_PENDING) -#define _TIF_DEBUG (1<<TIF_DEBUG) -#define _TIF_IO_BITMAP (1<<TIF_IO_BITMAP) -#define _TIF_FREEZE (1<<TIF_FREEZE) -#define _TIF_FORCED_TF (1<<TIF_FORCED_TF) -#define _TIF_DEBUGCTLMSR (1<<TIF_DEBUGCTLMSR) -#define _TIF_DS_AREA_MSR (1<<TIF_DS_AREA_MSR) -#define _TIF_BTS_TRACE_TS (1<<TIF_BTS_TRACE_TS) +#define _TIF_SYSCALL_TRACE (1 << TIF_SYSCALL_TRACE) +#define _TIF_SIGPENDING (1 << TIF_SIGPENDING) +#define _TIF_SINGLESTEP (1 << TIF_SINGLESTEP) +#define _TIF_NEED_RESCHED (1 << TIF_NEED_RESCHED) +#define _TIF_IRET (1 << TIF_IRET) +#define _TIF_SYSCALL_AUDIT (1 << TIF_SYSCALL_AUDIT) +#define _TIF_SECCOMP (1 << TIF_SECCOMP) +#define _TIF_RESTORE_SIGMASK (1 << TIF_RESTORE_SIGMASK) +#define _TIF_MCE_NOTIFY (1 << TIF_MCE_NOTIFY) +#define _TIF_HRTICK_RESCHED (1 << TIF_HRTICK_RESCHED) +#define _TIF_IA32 (1 << TIF_IA32) +#define _TIF_FORK (1 << TIF_FORK) +#define _TIF_ABI_PENDING (1 << TIF_ABI_PENDING) +#define _TIF_DEBUG (1 << TIF_DEBUG) +#define _TIF_IO_BITMAP (1 << TIF_IO_BITMAP) +#define _TIF_FREEZE (1 << TIF_FREEZE) +#define _TIF_FORCED_TF (1 << TIF_FORCED_TF) +#define _TIF_DEBUGCTLMSR (1 << TIF_DEBUGCTLMSR) +#define _TIF_DS_AREA_MSR (1 << TIF_DS_AREA_MSR) +#define _TIF_BTS_TRACE_TS (1 << TIF_BTS_TRACE_TS) /* work to do on interrupt/exception return */ -#define _TIF_WORK_MASK \ - (0x0000FFFF & ~(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SINGLESTEP|_TIF_SECCOMP)) +#define _TIF_WORK_MASK \ + (0x0000FFFF & \ + ~(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SINGLESTEP|_TIF_SECCOMP)) /* work to do on any return to user space */ #define _TIF_ALLWORK_MASK (0x0000FFFF & ~_TIF_SECCOMP) -#define _TIF_DO_NOTIFY_MASK \ +#define _TIF_DO_NOTIFY_MASK \ (_TIF_SIGPENDING|_TIF_SINGLESTEP|_TIF_MCE_NOTIFY|_TIF_HRTICK_RESCHED) /* flags to check in __switch_to() */ -#define _TIF_WORK_CTXSW \ - (_TIF_IO_BITMAP|_TIF_DEBUGCTLMSR|_TIF_DS_AREA_MSR|_TIF_BTS_TRACE_TS) +#define _TIF_WORK_CTXSW \ + (_TIF_IO_BITMAP|_TIF_DEBUGCTLMSR|_TIF_DS_AREA_MSR|_TIF_BTS_TRACE_TS) #define _TIF_WORK_CTXSW_PREV _TIF_WORK_CTXSW #define _TIF_WORK_CTXSW_NEXT (_TIF_WORK_CTXSW|_TIF_DEBUG) @@ -171,9 +173,11 @@ static inline struct thread_info *stack_thread_info(void) * ever touches our thread-synchronous status, so we don't * have to worry about atomic accesses. */ -#define TS_USEDFPU 0x0001 /* FPU was used by this task this quantum (SMP) */ +#define TS_USEDFPU 0x0001 /* FPU was used by this task + this quantum (SMP) */ #define TS_COMPAT 0x0002 /* 32bit syscall active */ -#define TS_POLLING 0x0004 /* true if in idle loop and not sleeping */ +#define TS_POLLING 0x0004 /* true if in idle loop + and not sleeping */ #define tsk_is_polling(t) (task_thread_info(t)->status & TS_POLLING) diff --git a/include/asm-x86/tlbflush.h b/include/asm-x86/tlbflush.h index 3998709ed637..0c0674d94255 100644 --- a/include/asm-x86/tlbflush.h +++ b/include/asm-x86/tlbflush.h @@ -32,7 +32,7 @@ static inline void __native_flush_tlb_global(void) static inline void __native_flush_tlb_single(unsigned long addr) { - __asm__ __volatile__("invlpg (%0)" ::"r" (addr) : "memory"); + asm volatile("invlpg (%0)" ::"r" (addr) : "memory"); } static inline void __flush_tlb_all(void) @@ -134,8 +134,7 @@ void native_flush_tlb_others(const cpumask_t *cpumask, struct mm_struct *mm, #define TLBSTATE_LAZY 2 #ifdef CONFIG_X86_32 -struct tlb_state -{ +struct tlb_state { struct mm_struct *active_mm; int state; char __cacheline_padding[L1_CACHE_BYTES-8]; diff --git a/include/asm-x86/topology.h b/include/asm-x86/topology.h index 8af05a93f097..81a29eb08ac4 100644 --- a/include/asm-x86/topology.h +++ b/include/asm-x86/topology.h @@ -32,13 +32,18 @@ /* Mappings between logical cpu number and node number */ #ifdef CONFIG_X86_32 extern int cpu_to_node_map[]; - #else +/* Returns the number of the current Node. */ +#define numa_node_id() (early_cpu_to_node(raw_smp_processor_id())) +#endif + DECLARE_PER_CPU(int, x86_cpu_to_node_map); + +#ifdef CONFIG_SMP extern int x86_cpu_to_node_map_init[]; extern void *x86_cpu_to_node_map_early_ptr; -/* Returns the number of the current Node. */ -#define numa_node_id() (early_cpu_to_node(raw_smp_processor_id())) +#else +#define x86_cpu_to_node_map_early_ptr NULL #endif extern cpumask_t node_to_cpumask_map[]; @@ -54,6 +59,8 @@ static inline int cpu_to_node(int cpu) } #else /* CONFIG_X86_64 */ + +#ifdef CONFIG_SMP static inline int early_cpu_to_node(int cpu) { int *cpu_to_node_map = x86_cpu_to_node_map_early_ptr; @@ -65,21 +72,21 @@ static inline int early_cpu_to_node(int cpu) else return NUMA_NO_NODE; } +#else +#define early_cpu_to_node(cpu) cpu_to_node(cpu) +#endif static inline int cpu_to_node(int cpu) { #ifdef CONFIG_DEBUG_PER_CPU_MAPS if (x86_cpu_to_node_map_early_ptr) { printk("KERN_NOTICE cpu_to_node(%d): usage too early!\n", - (int)cpu); + (int)cpu); dump_stack(); return ((int *)x86_cpu_to_node_map_early_ptr)[cpu]; } #endif - if (per_cpu_offset(cpu)) - return per_cpu(x86_cpu_to_node_map, cpu); - else - return NUMA_NO_NODE; + return per_cpu(x86_cpu_to_node_map, cpu); } #endif /* CONFIG_X86_64 */ diff --git a/include/asm-x86/trampoline.h b/include/asm-x86/trampoline.h new file mode 100644 index 000000000000..b156b08d0131 --- /dev/null +++ b/include/asm-x86/trampoline.h @@ -0,0 +1,21 @@ +#ifndef __TRAMPOLINE_HEADER +#define __TRAMPOLINE_HEADER + +#ifndef __ASSEMBLY__ + +/* + * Trampoline 80x86 program as an array. + */ +extern const unsigned char trampoline_data []; +extern const unsigned char trampoline_end []; +extern unsigned char *trampoline_base; + +extern unsigned long init_rsp; +extern unsigned long initial_code; + +#define TRAMPOLINE_BASE 0x6000 +extern unsigned long setup_trampoline(void); + +#endif /* __ASSEMBLY__ */ + +#endif /* __TRAMPOLINE_HEADER */ diff --git a/include/asm-x86/tsc.h b/include/asm-x86/tsc.h index 7d3e27f7d484..d2d8eb5b55f5 100644 --- a/include/asm-x86/tsc.h +++ b/include/asm-x86/tsc.h @@ -42,7 +42,7 @@ static inline cycles_t vget_cycles(void) if (!cpu_has_tsc) return 0; #endif - return (cycles_t) __native_read_tsc(); + return (cycles_t)__native_read_tsc(); } extern void tsc_init(void); diff --git a/include/asm-x86/uaccess_32.h b/include/asm-x86/uaccess_32.h index fcc570ec4fee..8e7595c1f34e 100644 --- a/include/asm-x86/uaccess_32.h +++ b/include/asm-x86/uaccess_32.h @@ -32,7 +32,7 @@ #define get_fs() (current_thread_info()->addr_limit) #define set_fs(x) (current_thread_info()->addr_limit = (x)) -#define segment_eq(a,b) ((a).seg == (b).seg) +#define segment_eq(a, b) ((a).seg == (b).seg) /* * movsl can be slow when source and dest are not both 8-byte aligned @@ -43,7 +43,9 @@ extern struct movsl_mask { } ____cacheline_aligned_in_smp movsl_mask; #endif -#define __addr_ok(addr) ((unsigned long __force)(addr) < (current_thread_info()->addr_limit.seg)) +#define __addr_ok(addr) \ + ((unsigned long __force)(addr) < \ + (current_thread_info()->addr_limit.seg)) /* * Test whether a block of memory is a valid user space address. @@ -54,13 +56,16 @@ extern struct movsl_mask { * * This needs 33-bit arithmetic. We have a carry... */ -#define __range_ok(addr,size) ({ \ - unsigned long flag,roksum; \ - __chk_user_ptr(addr); \ - asm("addl %3,%1 ; sbbl %0,%0; cmpl %1,%4; sbbl $0,%0" \ - :"=&r" (flag), "=r" (roksum) \ - :"1" (addr),"g" ((int)(size)),"rm" (current_thread_info()->addr_limit.seg)); \ - flag; }) +#define __range_ok(addr, size) \ +({ \ + unsigned long flag, roksum; \ + __chk_user_ptr(addr); \ + asm("addl %3,%1 ; sbbl %0,%0; cmpl %1,%4; sbbl $0,%0" \ + :"=&r" (flag), "=r" (roksum) \ + :"1" (addr), "g" ((int)(size)), \ + "rm" (current_thread_info()->addr_limit.seg)); \ + flag; \ +}) /** * access_ok: - Checks if a user space pointer is valid @@ -81,7 +86,7 @@ extern struct movsl_mask { * checks that the pointer is in the user space range - after calling * this function, memory access functions may still return -EFAULT. */ -#define access_ok(type,addr,size) (likely(__range_ok(addr,size) == 0)) +#define access_ok(type, addr, size) (likely(__range_ok(addr, size) == 0)) /* * The exception table consists of pairs of addresses: the first is the @@ -96,8 +101,7 @@ extern struct movsl_mask { * on our cache or tlb entries. */ -struct exception_table_entry -{ +struct exception_table_entry { unsigned long insn, fixup; }; @@ -122,13 +126,15 @@ extern void __get_user_1(void); extern void __get_user_2(void); extern void __get_user_4(void); -#define __get_user_x(size,ret,x,ptr) \ - __asm__ __volatile__("call __get_user_" #size \ - :"=a" (ret),"=d" (x) \ - :"0" (ptr)) +#define __get_user_x(size, ret, x, ptr) \ + asm volatile("call __get_user_" #size \ + :"=a" (ret),"=d" (x) \ + :"0" (ptr)) + +/* Careful: we have to cast the result to the type of the pointer + * for sign reasons */ -/* Careful: we have to cast the result to the type of the pointer for sign reasons */ /** * get_user: - Get a simple variable from user space. * @x: Variable to store result. @@ -146,15 +152,24 @@ extern void __get_user_4(void); * Returns zero on success, or -EFAULT on error. * On error, the variable @x is set to zero. */ -#define get_user(x,ptr) \ -({ int __ret_gu; \ +#define get_user(x, ptr) \ +({ \ + int __ret_gu; \ unsigned long __val_gu; \ __chk_user_ptr(ptr); \ - switch(sizeof (*(ptr))) { \ - case 1: __get_user_x(1,__ret_gu,__val_gu,ptr); break; \ - case 2: __get_user_x(2,__ret_gu,__val_gu,ptr); break; \ - case 4: __get_user_x(4,__ret_gu,__val_gu,ptr); break; \ - default: __get_user_x(X,__ret_gu,__val_gu,ptr); break; \ + switch (sizeof(*(ptr))) { \ + case 1: \ + __get_user_x(1, __ret_gu, __val_gu, ptr); \ + break; \ + case 2: \ + __get_user_x(2, __ret_gu, __val_gu, ptr); \ + break; \ + case 4: \ + __get_user_x(4, __ret_gu, __val_gu, ptr); \ + break; \ + default: \ + __get_user_x(X, __ret_gu, __val_gu, ptr); \ + break; \ } \ (x) = (__typeof__(*(ptr)))__val_gu; \ __ret_gu; \ @@ -171,11 +186,25 @@ extern void __put_user_2(void); extern void __put_user_4(void); extern void __put_user_8(void); -#define __put_user_1(x, ptr) __asm__ __volatile__("call __put_user_1":"=a" (__ret_pu):"0" ((typeof(*(ptr)))(x)), "c" (ptr)) -#define __put_user_2(x, ptr) __asm__ __volatile__("call __put_user_2":"=a" (__ret_pu):"0" ((typeof(*(ptr)))(x)), "c" (ptr)) -#define __put_user_4(x, ptr) __asm__ __volatile__("call __put_user_4":"=a" (__ret_pu):"0" ((typeof(*(ptr)))(x)), "c" (ptr)) -#define __put_user_8(x, ptr) __asm__ __volatile__("call __put_user_8":"=a" (__ret_pu):"A" ((typeof(*(ptr)))(x)), "c" (ptr)) -#define __put_user_X(x, ptr) __asm__ __volatile__("call __put_user_X":"=a" (__ret_pu):"c" (ptr)) +#define __put_user_1(x, ptr) \ + asm volatile("call __put_user_1" : "=a" (__ret_pu) \ + : "0" ((typeof(*(ptr)))(x)), "c" (ptr)) + +#define __put_user_2(x, ptr) \ + asm volatile("call __put_user_2" : "=a" (__ret_pu) \ + : "0" ((typeof(*(ptr)))(x)), "c" (ptr)) + +#define __put_user_4(x, ptr) \ + asm volatile("call __put_user_4" : "=a" (__ret_pu) \ + : "0" ((typeof(*(ptr)))(x)), "c" (ptr)) + +#define __put_user_8(x, ptr) \ + asm volatile("call __put_user_8" : "=a" (__ret_pu) \ + : "A" ((typeof(*(ptr)))(x)), "c" (ptr)) + +#define __put_user_X(x, ptr) \ + asm volatile("call __put_user_X" : "=a" (__ret_pu) \ + : "c" (ptr)) /** * put_user: - Write a simple value into user space. @@ -195,32 +224,43 @@ extern void __put_user_8(void); */ #ifdef CONFIG_X86_WP_WORKS_OK -#define put_user(x,ptr) \ -({ int __ret_pu; \ +#define put_user(x, ptr) \ +({ \ + int __ret_pu; \ __typeof__(*(ptr)) __pu_val; \ __chk_user_ptr(ptr); \ __pu_val = x; \ - switch(sizeof(*(ptr))) { \ - case 1: __put_user_1(__pu_val, ptr); break; \ - case 2: __put_user_2(__pu_val, ptr); break; \ - case 4: __put_user_4(__pu_val, ptr); break; \ - case 8: __put_user_8(__pu_val, ptr); break; \ - default:__put_user_X(__pu_val, ptr); break; \ + switch (sizeof(*(ptr))) { \ + case 1: \ + __put_user_1(__pu_val, ptr); \ + break; \ + case 2: \ + __put_user_2(__pu_val, ptr); \ + break; \ + case 4: \ + __put_user_4(__pu_val, ptr); \ + break; \ + case 8: \ + __put_user_8(__pu_val, ptr); \ + break; \ + default: \ + __put_user_X(__pu_val, ptr); \ + break; \ } \ __ret_pu; \ }) #else -#define put_user(x,ptr) \ +#define put_user(x, ptr) \ ({ \ - int __ret_pu; \ - __typeof__(*(ptr)) __pus_tmp = x; \ - __ret_pu=0; \ - if(unlikely(__copy_to_user_ll(ptr, &__pus_tmp, \ - sizeof(*(ptr))) != 0)) \ - __ret_pu=-EFAULT; \ - __ret_pu; \ - }) + int __ret_pu; \ + __typeof__(*(ptr))__pus_tmp = x; \ + __ret_pu = 0; \ + if (unlikely(__copy_to_user_ll(ptr, &__pus_tmp, \ + sizeof(*(ptr))) != 0)) \ + __ret_pu = -EFAULT; \ + __ret_pu; \ +}) #endif @@ -245,8 +285,8 @@ extern void __put_user_8(void); * Returns zero on success, or -EFAULT on error. * On error, the variable @x is set to zero. */ -#define __get_user(x,ptr) \ - __get_user_nocheck((x),(ptr),sizeof(*(ptr))) +#define __get_user(x, ptr) \ + __get_user_nocheck((x), (ptr), sizeof(*(ptr))) /** @@ -268,54 +308,62 @@ extern void __put_user_8(void); * * Returns zero on success, or -EFAULT on error. */ -#define __put_user(x,ptr) \ - __put_user_nocheck((__typeof__(*(ptr)))(x),(ptr),sizeof(*(ptr))) +#define __put_user(x, ptr) \ + __put_user_nocheck((__typeof__(*(ptr)))(x), (ptr), sizeof(*(ptr))) -#define __put_user_nocheck(x,ptr,size) \ +#define __put_user_nocheck(x, ptr, size) \ ({ \ long __pu_err; \ - __put_user_size((x),(ptr),(size),__pu_err,-EFAULT); \ + __put_user_size((x), (ptr), (size), __pu_err, -EFAULT); \ __pu_err; \ }) -#define __put_user_u64(x, addr, err) \ - __asm__ __volatile__( \ - "1: movl %%eax,0(%2)\n" \ - "2: movl %%edx,4(%2)\n" \ - "3:\n" \ - ".section .fixup,\"ax\"\n" \ - "4: movl %3,%0\n" \ - " jmp 3b\n" \ - ".previous\n" \ - _ASM_EXTABLE(1b,4b) \ - _ASM_EXTABLE(2b,4b) \ - : "=r"(err) \ - : "A" (x), "r" (addr), "i"(-EFAULT), "0"(err)) +#define __put_user_u64(x, addr, err) \ + asm volatile("1: movl %%eax,0(%2)\n" \ + "2: movl %%edx,4(%2)\n" \ + "3:\n" \ + ".section .fixup,\"ax\"\n" \ + "4: movl %3,%0\n" \ + " jmp 3b\n" \ + ".previous\n" \ + _ASM_EXTABLE(1b, 4b) \ + _ASM_EXTABLE(2b, 4b) \ + : "=r" (err) \ + : "A" (x), "r" (addr), "i" (-EFAULT), "0" (err)) #ifdef CONFIG_X86_WP_WORKS_OK -#define __put_user_size(x,ptr,size,retval,errret) \ +#define __put_user_size(x, ptr, size, retval, errret) \ do { \ retval = 0; \ __chk_user_ptr(ptr); \ switch (size) { \ - case 1: __put_user_asm(x,ptr,retval,"b","b","iq",errret);break; \ - case 2: __put_user_asm(x,ptr,retval,"w","w","ir",errret);break; \ - case 4: __put_user_asm(x,ptr,retval,"l","","ir",errret); break; \ - case 8: __put_user_u64((__typeof__(*ptr))(x),ptr,retval); break;\ - default: __put_user_bad(); \ + case 1: \ + __put_user_asm(x, ptr, retval, "b", "b", "iq", errret); \ + break; \ + case 2: \ + __put_user_asm(x, ptr, retval, "w", "w", "ir", errret); \ + break; \ + case 4: \ + __put_user_asm(x, ptr, retval, "l", "", "ir", errret); \ + break; \ + case 8: \ + __put_user_u64((__typeof__(*ptr))(x), ptr, retval); \ + break; \ + default: \ + __put_user_bad(); \ } \ } while (0) #else -#define __put_user_size(x,ptr,size,retval,errret) \ +#define __put_user_size(x, ptr, size, retval, errret) \ do { \ - __typeof__(*(ptr)) __pus_tmp = x; \ + __typeof__(*(ptr))__pus_tmp = x; \ retval = 0; \ \ - if(unlikely(__copy_to_user_ll(ptr, &__pus_tmp, size) != 0)) \ + if (unlikely(__copy_to_user_ll(ptr, &__pus_tmp, size) != 0)) \ retval = errret; \ } while (0) @@ -329,65 +377,70 @@ struct __large_struct { unsigned long buf[100]; }; * aliasing issues. */ #define __put_user_asm(x, addr, err, itype, rtype, ltype, errret) \ - __asm__ __volatile__( \ - "1: mov"itype" %"rtype"1,%2\n" \ - "2:\n" \ - ".section .fixup,\"ax\"\n" \ - "3: movl %3,%0\n" \ - " jmp 2b\n" \ - ".previous\n" \ - _ASM_EXTABLE(1b,3b) \ - : "=r"(err) \ - : ltype (x), "m"(__m(addr)), "i"(errret), "0"(err)) - - -#define __get_user_nocheck(x,ptr,size) \ -({ \ - long __gu_err; \ - unsigned long __gu_val; \ - __get_user_size(__gu_val,(ptr),(size),__gu_err,-EFAULT);\ - (x) = (__typeof__(*(ptr)))__gu_val; \ - __gu_err; \ + asm volatile("1: mov"itype" %"rtype"1,%2\n" \ + "2:\n" \ + ".section .fixup,\"ax\"\n" \ + "3: movl %3,%0\n" \ + " jmp 2b\n" \ + ".previous\n" \ + _ASM_EXTABLE(1b, 3b) \ + : "=r"(err) \ + : ltype (x), "m" (__m(addr)), "i" (errret), "0" (err)) + + +#define __get_user_nocheck(x, ptr, size) \ +({ \ + long __gu_err; \ + unsigned long __gu_val; \ + __get_user_size(__gu_val, (ptr), (size), __gu_err, -EFAULT); \ + (x) = (__typeof__(*(ptr)))__gu_val; \ + __gu_err; \ }) extern long __get_user_bad(void); -#define __get_user_size(x,ptr,size,retval,errret) \ +#define __get_user_size(x, ptr, size, retval, errret) \ do { \ retval = 0; \ __chk_user_ptr(ptr); \ switch (size) { \ - case 1: __get_user_asm(x,ptr,retval,"b","b","=q",errret);break; \ - case 2: __get_user_asm(x,ptr,retval,"w","w","=r",errret);break; \ - case 4: __get_user_asm(x,ptr,retval,"l","","=r",errret);break; \ - default: (x) = __get_user_bad(); \ + case 1: \ + __get_user_asm(x, ptr, retval, "b", "b", "=q", errret); \ + break; \ + case 2: \ + __get_user_asm(x, ptr, retval, "w", "w", "=r", errret); \ + break; \ + case 4: \ + __get_user_asm(x, ptr, retval, "l", "", "=r", errret); \ + break; \ + default: \ + (x) = __get_user_bad(); \ } \ } while (0) #define __get_user_asm(x, addr, err, itype, rtype, ltype, errret) \ - __asm__ __volatile__( \ - "1: mov"itype" %2,%"rtype"1\n" \ - "2:\n" \ - ".section .fixup,\"ax\"\n" \ - "3: movl %3,%0\n" \ - " xor"itype" %"rtype"1,%"rtype"1\n" \ - " jmp 2b\n" \ - ".previous\n" \ - _ASM_EXTABLE(1b,3b) \ - : "=r"(err), ltype (x) \ - : "m"(__m(addr)), "i"(errret), "0"(err)) - - -unsigned long __must_check __copy_to_user_ll(void __user *to, - const void *from, unsigned long n); -unsigned long __must_check __copy_from_user_ll(void *to, - const void __user *from, unsigned long n); -unsigned long __must_check __copy_from_user_ll_nozero(void *to, - const void __user *from, unsigned long n); -unsigned long __must_check __copy_from_user_ll_nocache(void *to, - const void __user *from, unsigned long n); -unsigned long __must_check __copy_from_user_ll_nocache_nozero(void *to, - const void __user *from, unsigned long n); + asm volatile("1: mov"itype" %2,%"rtype"1\n" \ + "2:\n" \ + ".section .fixup,\"ax\"\n" \ + "3: movl %3,%0\n" \ + " xor"itype" %"rtype"1,%"rtype"1\n" \ + " jmp 2b\n" \ + ".previous\n" \ + _ASM_EXTABLE(1b, 3b) \ + : "=r" (err), ltype (x) \ + : "m" (__m(addr)), "i" (errret), "0" (err)) + + +unsigned long __must_check __copy_to_user_ll + (void __user *to, const void *from, unsigned long n); +unsigned long __must_check __copy_from_user_ll + (void *to, const void __user *from, unsigned long n); +unsigned long __must_check __copy_from_user_ll_nozero + (void *to, const void __user *from, unsigned long n); +unsigned long __must_check __copy_from_user_ll_nocache + (void *to, const void __user *from, unsigned long n); +unsigned long __must_check __copy_from_user_ll_nocache_nozero + (void *to, const void __user *from, unsigned long n); /** * __copy_to_user_inatomic: - Copy a block of data into user space, with less checking. @@ -416,13 +469,16 @@ __copy_to_user_inatomic(void __user *to, const void *from, unsigned long n) switch (n) { case 1: - __put_user_size(*(u8 *)from, (u8 __user *)to, 1, ret, 1); + __put_user_size(*(u8 *)from, (u8 __user *)to, + 1, ret, 1); return ret; case 2: - __put_user_size(*(u16 *)from, (u16 __user *)to, 2, ret, 2); + __put_user_size(*(u16 *)from, (u16 __user *)to, + 2, ret, 2); return ret; case 4: - __put_user_size(*(u32 *)from, (u32 __user *)to, 4, ret, 4); + __put_user_size(*(u32 *)from, (u32 __user *)to, + 4, ret, 4); return ret; } } @@ -545,19 +601,21 @@ static __always_inline unsigned long __copy_from_user_nocache(void *to, } static __always_inline unsigned long -__copy_from_user_inatomic_nocache(void *to, const void __user *from, unsigned long n) +__copy_from_user_inatomic_nocache(void *to, const void __user *from, + unsigned long n) { return __copy_from_user_ll_nocache_nozero(to, from, n); } unsigned long __must_check copy_to_user(void __user *to, - const void *from, unsigned long n); + const void *from, unsigned long n); unsigned long __must_check copy_from_user(void *to, - const void __user *from, unsigned long n); + const void __user *from, + unsigned long n); long __must_check strncpy_from_user(char *dst, const char __user *src, - long count); + long count); long __must_check __strncpy_from_user(char *dst, - const char __user *src, long count); + const char __user *src, long count); /** * strlen_user: - Get the size of a string in user space. diff --git a/include/asm-x86/uaccess_64.h b/include/asm-x86/uaccess_64.h index b87eb4ba8f9d..b8a2f4339903 100644 --- a/include/asm-x86/uaccess_64.h +++ b/include/asm-x86/uaccess_64.h @@ -29,23 +29,27 @@ #define get_fs() (current_thread_info()->addr_limit) #define set_fs(x) (current_thread_info()->addr_limit = (x)) -#define segment_eq(a,b) ((a).seg == (b).seg) +#define segment_eq(a, b) ((a).seg == (b).seg) -#define __addr_ok(addr) (!((unsigned long)(addr) & (current_thread_info()->addr_limit.seg))) +#define __addr_ok(addr) (!((unsigned long)(addr) & \ + (current_thread_info()->addr_limit.seg))) /* * Uhhuh, this needs 65-bit arithmetic. We have a carry.. */ -#define __range_not_ok(addr,size) ({ \ - unsigned long flag,roksum; \ - __chk_user_ptr(addr); \ - asm("# range_ok\n\r" \ - "addq %3,%1 ; sbbq %0,%0 ; cmpq %1,%4 ; sbbq $0,%0" \ - :"=&r" (flag), "=r" (roksum) \ - :"1" (addr),"g" ((long)(size)),"g" (current_thread_info()->addr_limit.seg)); \ - flag; }) +#define __range_not_ok(addr, size) \ +({ \ + unsigned long flag, roksum; \ + __chk_user_ptr(addr); \ + asm("# range_ok\n\r" \ + "addq %3,%1 ; sbbq %0,%0 ; cmpq %1,%4 ; sbbq $0,%0" \ + : "=&r" (flag), "=r" (roksum) \ + : "1" (addr), "g" ((long)(size)), \ + "g" (current_thread_info()->addr_limit.seg)); \ + flag; \ +}) -#define access_ok(type, addr, size) (__range_not_ok(addr,size) == 0) +#define access_ok(type, addr, size) (__range_not_ok(addr, size) == 0) /* * The exception table consists of pairs of addresses: the first is the @@ -60,8 +64,7 @@ * on our cache or tlb entries. */ -struct exception_table_entry -{ +struct exception_table_entry { unsigned long insn, fixup; }; @@ -84,23 +87,36 @@ extern int fixup_exception(struct pt_regs *regs); * accesses to the same area of user memory). */ -#define __get_user_x(size,ret,x,ptr) \ - asm volatile("call __get_user_" #size \ - :"=a" (ret),"=d" (x) \ - :"c" (ptr) \ - :"r8") +#define __get_user_x(size, ret, x, ptr) \ + asm volatile("call __get_user_" #size \ + : "=a" (ret),"=d" (x) \ + : "c" (ptr) \ + : "r8") + +/* Careful: we have to cast the result to the type of the pointer + * for sign reasons */ -/* Careful: we have to cast the result to the type of the pointer for sign reasons */ -#define get_user(x,ptr) \ -({ unsigned long __val_gu; \ - int __ret_gu; \ +#define get_user(x, ptr) \ +({ \ + unsigned long __val_gu; \ + int __ret_gu; \ __chk_user_ptr(ptr); \ - switch(sizeof (*(ptr))) { \ - case 1: __get_user_x(1,__ret_gu,__val_gu,ptr); break; \ - case 2: __get_user_x(2,__ret_gu,__val_gu,ptr); break; \ - case 4: __get_user_x(4,__ret_gu,__val_gu,ptr); break; \ - case 8: __get_user_x(8,__ret_gu,__val_gu,ptr); break; \ - default: __get_user_bad(); break; \ + switch (sizeof(*(ptr))) { \ + case 1: \ + __get_user_x(1, __ret_gu, __val_gu, ptr); \ + break; \ + case 2: \ + __get_user_x(2, __ret_gu, __val_gu, ptr); \ + break; \ + case 4: \ + __get_user_x(4, __ret_gu, __val_gu, ptr); \ + break; \ + case 8: \ + __get_user_x(8, __ret_gu, __val_gu, ptr); \ + break; \ + default: \ + __get_user_bad(); \ + break; \ } \ (x) = (__force typeof(*(ptr)))__val_gu; \ __ret_gu; \ @@ -112,55 +128,73 @@ extern void __put_user_4(void); extern void __put_user_8(void); extern void __put_user_bad(void); -#define __put_user_x(size,ret,x,ptr) \ - asm volatile("call __put_user_" #size \ - :"=a" (ret) \ - :"c" (ptr),"d" (x) \ - :"r8") +#define __put_user_x(size, ret, x, ptr) \ + asm volatile("call __put_user_" #size \ + :"=a" (ret) \ + :"c" (ptr),"d" (x) \ + :"r8") -#define put_user(x,ptr) \ - __put_user_check((__typeof__(*(ptr)))(x),(ptr),sizeof(*(ptr))) +#define put_user(x, ptr) \ + __put_user_check((__typeof__(*(ptr)))(x), (ptr), sizeof(*(ptr))) -#define __get_user(x,ptr) \ - __get_user_nocheck((x),(ptr),sizeof(*(ptr))) -#define __put_user(x,ptr) \ - __put_user_nocheck((__typeof__(*(ptr)))(x),(ptr),sizeof(*(ptr))) +#define __get_user(x, ptr) \ + __get_user_nocheck((x), (ptr), sizeof(*(ptr))) +#define __put_user(x, ptr) \ + __put_user_nocheck((__typeof__(*(ptr)))(x), (ptr), sizeof(*(ptr))) #define __get_user_unaligned __get_user #define __put_user_unaligned __put_user -#define __put_user_nocheck(x,ptr,size) \ +#define __put_user_nocheck(x, ptr, size) \ ({ \ int __pu_err; \ - __put_user_size((x),(ptr),(size),__pu_err); \ + __put_user_size((x), (ptr), (size), __pu_err); \ __pu_err; \ }) -#define __put_user_check(x,ptr,size) \ -({ \ - int __pu_err; \ - typeof(*(ptr)) __user *__pu_addr = (ptr); \ - switch (size) { \ - case 1: __put_user_x(1,__pu_err,x,__pu_addr); break; \ - case 2: __put_user_x(2,__pu_err,x,__pu_addr); break; \ - case 4: __put_user_x(4,__pu_err,x,__pu_addr); break; \ - case 8: __put_user_x(8,__pu_err,x,__pu_addr); break; \ - default: __put_user_bad(); \ - } \ - __pu_err; \ +#define __put_user_check(x, ptr, size) \ +({ \ + int __pu_err; \ + typeof(*(ptr)) __user *__pu_addr = (ptr); \ + switch (size) { \ + case 1: \ + __put_user_x(1, __pu_err, x, __pu_addr); \ + break; \ + case 2: \ + __put_user_x(2, __pu_err, x, __pu_addr); \ + break; \ + case 4: \ + __put_user_x(4, __pu_err, x, __pu_addr); \ + break; \ + case 8: \ + __put_user_x(8, __pu_err, x, __pu_addr); \ + break; \ + default: \ + __put_user_bad(); \ + } \ + __pu_err; \ }) -#define __put_user_size(x,ptr,size,retval) \ +#define __put_user_size(x, ptr, size, retval) \ do { \ retval = 0; \ __chk_user_ptr(ptr); \ switch (size) { \ - case 1: __put_user_asm(x,ptr,retval,"b","b","iq",-EFAULT); break;\ - case 2: __put_user_asm(x,ptr,retval,"w","w","ir",-EFAULT); break;\ - case 4: __put_user_asm(x,ptr,retval,"l","k","ir",-EFAULT); break;\ - case 8: __put_user_asm(x,ptr,retval,"q","","Zr",-EFAULT); break;\ - default: __put_user_bad(); \ + case 1: \ + __put_user_asm(x, ptr, retval, "b", "b", "iq", -EFAULT);\ + break; \ + case 2: \ + __put_user_asm(x, ptr, retval, "w", "w", "ir", -EFAULT);\ + break; \ + case 4: \ + __put_user_asm(x, ptr, retval, "l", "k", "ir", -EFAULT);\ + break; \ + case 8: \ + __put_user_asm(x, ptr, retval, "q", "", "Zr", -EFAULT); \ + break; \ + default: \ + __put_user_bad(); \ } \ } while (0) @@ -174,23 +208,22 @@ struct __large_struct { unsigned long buf[100]; }; * aliasing issues. */ #define __put_user_asm(x, addr, err, itype, rtype, ltype, errno) \ - asm volatile( \ - "1: mov"itype" %"rtype"1,%2\n" \ - "2:\n" \ - ".section .fixup,\"ax\"\n" \ - "3: mov %3,%0\n" \ - " jmp 2b\n" \ - ".previous\n" \ - _ASM_EXTABLE(1b,3b) \ - : "=r"(err) \ - : ltype (x), "m"(__m(addr)), "i"(errno), "0"(err)) - - -#define __get_user_nocheck(x,ptr,size) \ + asm volatile("1: mov"itype" %"rtype"1,%2\n" \ + "2:\n" \ + ".section .fixup, \"ax\"\n" \ + "3: mov %3,%0\n" \ + " jmp 2b\n" \ + ".previous\n" \ + _ASM_EXTABLE(1b, 3b) \ + : "=r"(err) \ + : ltype (x), "m" (__m(addr)), "i" (errno), "0" (err)) + + +#define __get_user_nocheck(x, ptr, size) \ ({ \ int __gu_err; \ unsigned long __gu_val; \ - __get_user_size(__gu_val,(ptr),(size),__gu_err); \ + __get_user_size(__gu_val, (ptr), (size), __gu_err); \ (x) = (__force typeof(*(ptr)))__gu_val; \ __gu_err; \ }) @@ -201,31 +234,39 @@ extern int __get_user_4(void); extern int __get_user_8(void); extern int __get_user_bad(void); -#define __get_user_size(x,ptr,size,retval) \ +#define __get_user_size(x, ptr, size, retval) \ do { \ retval = 0; \ __chk_user_ptr(ptr); \ switch (size) { \ - case 1: __get_user_asm(x,ptr,retval,"b","b","=q",-EFAULT); break;\ - case 2: __get_user_asm(x,ptr,retval,"w","w","=r",-EFAULT); break;\ - case 4: __get_user_asm(x,ptr,retval,"l","k","=r",-EFAULT); break;\ - case 8: __get_user_asm(x,ptr,retval,"q","","=r",-EFAULT); break;\ - default: (x) = __get_user_bad(); \ + case 1: \ + __get_user_asm(x, ptr, retval, "b", "b", "=q", -EFAULT);\ + break; \ + case 2: \ + __get_user_asm(x, ptr, retval, "w", "w", "=r", -EFAULT);\ + break; \ + case 4: \ + __get_user_asm(x, ptr, retval, "l", "k", "=r", -EFAULT);\ + break; \ + case 8: \ + __get_user_asm(x, ptr, retval, "q", "", "=r", -EFAULT); \ + break; \ + default: \ + (x) = __get_user_bad(); \ } \ } while (0) #define __get_user_asm(x, addr, err, itype, rtype, ltype, errno) \ - asm volatile( \ - "1: mov"itype" %2,%"rtype"1\n" \ - "2:\n" \ - ".section .fixup,\"ax\"\n" \ - "3: mov %3,%0\n" \ - " xor"itype" %"rtype"1,%"rtype"1\n" \ - " jmp 2b\n" \ - ".previous\n" \ - _ASM_EXTABLE(1b,3b) \ - : "=r"(err), ltype (x) \ - : "m"(__m(addr)), "i"(errno), "0"(err)) + asm volatile("1: mov"itype" %2,%"rtype"1\n" \ + "2:\n" \ + ".section .fixup, \"ax\"\n" \ + "3: mov %3,%0\n" \ + " xor"itype" %"rtype"1,%"rtype"1\n" \ + " jmp 2b\n" \ + ".previous\n" \ + _ASM_EXTABLE(1b, 3b) \ + : "=r" (err), ltype (x) \ + : "m" (__m(addr)), "i"(errno), "0"(err)) /* * Copy To/From Userspace @@ -244,110 +285,142 @@ copy_in_user(void __user *to, const void __user *from, unsigned len); static __always_inline __must_check int __copy_from_user(void *dst, const void __user *src, unsigned size) -{ +{ int ret = 0; if (!__builtin_constant_p(size)) - return copy_user_generic(dst,(__force void *)src,size); - switch (size) { - case 1:__get_user_asm(*(u8*)dst,(u8 __user *)src,ret,"b","b","=q",1); + return copy_user_generic(dst, (__force void *)src, size); + switch (size) { + case 1:__get_user_asm(*(u8 *)dst, (u8 __user *)src, + ret, "b", "b", "=q", 1); return ret; - case 2:__get_user_asm(*(u16*)dst,(u16 __user *)src,ret,"w","w","=r",2); + case 2:__get_user_asm(*(u16 *)dst, (u16 __user *)src, + ret, "w", "w", "=r", 2); return ret; - case 4:__get_user_asm(*(u32*)dst,(u32 __user *)src,ret,"l","k","=r",4); + case 4:__get_user_asm(*(u32 *)dst, (u32 __user *)src, + ret, "l", "k", "=r", 4); + return ret; + case 8:__get_user_asm(*(u64 *)dst, (u64 __user *)src, + ret, "q", "", "=r", 8); return ret; - case 8:__get_user_asm(*(u64*)dst,(u64 __user *)src,ret,"q","","=r",8); - return ret; case 10: - __get_user_asm(*(u64*)dst,(u64 __user *)src,ret,"q","","=r",16); - if (unlikely(ret)) return ret; - __get_user_asm(*(u16*)(8+(char*)dst),(u16 __user *)(8+(char __user *)src),ret,"w","w","=r",2); - return ret; + __get_user_asm(*(u64 *)dst, (u64 __user *)src, + ret, "q", "", "=r", 16); + if (unlikely(ret)) + return ret; + __get_user_asm(*(u16 *)(8 + (char *)dst), + (u16 __user *)(8 + (char __user *)src), + ret, "w", "w", "=r", 2); + return ret; case 16: - __get_user_asm(*(u64*)dst,(u64 __user *)src,ret,"q","","=r",16); - if (unlikely(ret)) return ret; - __get_user_asm(*(u64*)(8+(char*)dst),(u64 __user *)(8+(char __user *)src),ret,"q","","=r",8); - return ret; + __get_user_asm(*(u64 *)dst, (u64 __user *)src, + ret, "q", "", "=r", 16); + if (unlikely(ret)) + return ret; + __get_user_asm(*(u64 *)(8 + (char *)dst), + (u64 __user *)(8 + (char __user *)src), + ret, "q", "", "=r", 8); + return ret; default: - return copy_user_generic(dst,(__force void *)src,size); + return copy_user_generic(dst, (__force void *)src, size); } -} +} static __always_inline __must_check int __copy_to_user(void __user *dst, const void *src, unsigned size) -{ +{ int ret = 0; if (!__builtin_constant_p(size)) - return copy_user_generic((__force void *)dst,src,size); - switch (size) { - case 1:__put_user_asm(*(u8*)src,(u8 __user *)dst,ret,"b","b","iq",1); + return copy_user_generic((__force void *)dst, src, size); + switch (size) { + case 1:__put_user_asm(*(u8 *)src, (u8 __user *)dst, + ret, "b", "b", "iq", 1); return ret; - case 2:__put_user_asm(*(u16*)src,(u16 __user *)dst,ret,"w","w","ir",2); + case 2:__put_user_asm(*(u16 *)src, (u16 __user *)dst, + ret, "w", "w", "ir", 2); return ret; - case 4:__put_user_asm(*(u32*)src,(u32 __user *)dst,ret,"l","k","ir",4); + case 4:__put_user_asm(*(u32 *)src, (u32 __user *)dst, + ret, "l", "k", "ir", 4); + return ret; + case 8:__put_user_asm(*(u64 *)src, (u64 __user *)dst, + ret, "q", "", "ir", 8); return ret; - case 8:__put_user_asm(*(u64*)src,(u64 __user *)dst,ret,"q","","ir",8); - return ret; case 10: - __put_user_asm(*(u64*)src,(u64 __user *)dst,ret,"q","","ir",10); - if (unlikely(ret)) return ret; + __put_user_asm(*(u64 *)src, (u64 __user *)dst, + ret, "q", "", "ir", 10); + if (unlikely(ret)) + return ret; asm("":::"memory"); - __put_user_asm(4[(u16*)src],4+(u16 __user *)dst,ret,"w","w","ir",2); - return ret; + __put_user_asm(4[(u16 *)src], 4 + (u16 __user *)dst, + ret, "w", "w", "ir", 2); + return ret; case 16: - __put_user_asm(*(u64*)src,(u64 __user *)dst,ret,"q","","ir",16); - if (unlikely(ret)) return ret; + __put_user_asm(*(u64 *)src, (u64 __user *)dst, + ret, "q", "", "ir", 16); + if (unlikely(ret)) + return ret; asm("":::"memory"); - __put_user_asm(1[(u64*)src],1+(u64 __user *)dst,ret,"q","","ir",8); - return ret; + __put_user_asm(1[(u64 *)src], 1 + (u64 __user *)dst, + ret, "q", "", "ir", 8); + return ret; default: - return copy_user_generic((__force void *)dst,src,size); + return copy_user_generic((__force void *)dst, src, size); } -} +} static __always_inline __must_check int __copy_in_user(void __user *dst, const void __user *src, unsigned size) -{ +{ int ret = 0; if (!__builtin_constant_p(size)) - return copy_user_generic((__force void *)dst,(__force void *)src,size); - switch (size) { - case 1: { + return copy_user_generic((__force void *)dst, + (__force void *)src, size); + switch (size) { + case 1: { u8 tmp; - __get_user_asm(tmp,(u8 __user *)src,ret,"b","b","=q",1); + __get_user_asm(tmp, (u8 __user *)src, + ret, "b", "b", "=q", 1); if (likely(!ret)) - __put_user_asm(tmp,(u8 __user *)dst,ret,"b","b","iq",1); + __put_user_asm(tmp, (u8 __user *)dst, + ret, "b", "b", "iq", 1); return ret; } - case 2: { + case 2: { u16 tmp; - __get_user_asm(tmp,(u16 __user *)src,ret,"w","w","=r",2); + __get_user_asm(tmp, (u16 __user *)src, + ret, "w", "w", "=r", 2); if (likely(!ret)) - __put_user_asm(tmp,(u16 __user *)dst,ret,"w","w","ir",2); + __put_user_asm(tmp, (u16 __user *)dst, + ret, "w", "w", "ir", 2); return ret; } - case 4: { + case 4: { u32 tmp; - __get_user_asm(tmp,(u32 __user *)src,ret,"l","k","=r",4); + __get_user_asm(tmp, (u32 __user *)src, + ret, "l", "k", "=r", 4); if (likely(!ret)) - __put_user_asm(tmp,(u32 __user *)dst,ret,"l","k","ir",4); + __put_user_asm(tmp, (u32 __user *)dst, + ret, "l", "k", "ir", 4); return ret; } - case 8: { + case 8: { u64 tmp; - __get_user_asm(tmp,(u64 __user *)src,ret,"q","","=r",8); + __get_user_asm(tmp, (u64 __user *)src, + ret, "q", "", "=r", 8); if (likely(!ret)) - __put_user_asm(tmp,(u64 __user *)dst,ret,"q","","ir",8); + __put_user_asm(tmp, (u64 __user *)dst, + ret, "q", "", "ir", 8); return ret; } default: - return copy_user_generic((__force void *)dst,(__force void *)src,size); + return copy_user_generic((__force void *)dst, + (__force void *)src, size); } -} +} -__must_check long +__must_check long strncpy_from_user(char *dst, const char __user *src, long count); -__must_check long +__must_check long __strncpy_from_user(char *dst, const char __user *src, long count); __must_check long strnlen_user(const char __user *str, long n); __must_check long __strnlen_user(const char __user *str, long n); @@ -355,7 +428,8 @@ __must_check long strlen_user(const char __user *str); __must_check unsigned long clear_user(void __user *mem, unsigned long len); __must_check unsigned long __clear_user(void __user *mem, unsigned long len); -__must_check long __copy_from_user_inatomic(void *dst, const void __user *src, unsigned size); +__must_check long __copy_from_user_inatomic(void *dst, const void __user *src, + unsigned size); static __must_check __always_inline int __copy_to_user_inatomic(void __user *dst, const void *src, unsigned size) @@ -364,15 +438,19 @@ __copy_to_user_inatomic(void __user *dst, const void *src, unsigned size) } #define ARCH_HAS_NOCACHE_UACCESS 1 -extern long __copy_user_nocache(void *dst, const void __user *src, unsigned size, int zerorest); +extern long __copy_user_nocache(void *dst, const void __user *src, + unsigned size, int zerorest); -static inline int __copy_from_user_nocache(void *dst, const void __user *src, unsigned size) +static inline int __copy_from_user_nocache(void *dst, const void __user *src, + unsigned size) { might_sleep(); return __copy_user_nocache(dst, src, size, 1); } -static inline int __copy_from_user_inatomic_nocache(void *dst, const void __user *src, unsigned size) +static inline int __copy_from_user_inatomic_nocache(void *dst, + const void __user *src, + unsigned size) { return __copy_user_nocache(dst, src, size, 0); } diff --git a/include/asm-x86/unaligned.h b/include/asm-x86/unaligned.h index 913598d4f761..d270ffe72759 100644 --- a/include/asm-x86/unaligned.h +++ b/include/asm-x86/unaligned.h @@ -32,6 +32,6 @@ * * Note that unaligned accesses can be very expensive on some architectures. */ -#define put_unaligned(val, ptr) ((void)( *(ptr) = (val) )) +#define put_unaligned(val, ptr) ((void)(*(ptr) = (val))) #endif /* _ASM_X86_UNALIGNED_H */ diff --git a/include/asm-x86/unistd.h b/include/asm-x86/unistd.h index 2a58ed3e51d8..effc7ad8e12f 100644 --- a/include/asm-x86/unistd.h +++ b/include/asm-x86/unistd.h @@ -1,11 +1,5 @@ #ifdef __KERNEL__ -# ifdef CONFIG_X86_32 -# include "unistd_32.h" -# else -# include "unistd_64.h" -# endif -#else -# ifdef __i386__ +# if defined(CONFIG_X86_32) || defined(__i386__) # include "unistd_32.h" # else # include "unistd_64.h" diff --git a/include/asm-x86/unistd_32.h b/include/asm-x86/unistd_32.h index 984123a68f7c..8317d94771d3 100644 --- a/include/asm-x86/unistd_32.h +++ b/include/asm-x86/unistd_32.h @@ -81,7 +81,7 @@ #define __NR_sigpending 73 #define __NR_sethostname 74 #define __NR_setrlimit 75 -#define __NR_getrlimit 76 /* Back compatible 2Gig limited rlimit */ +#define __NR_getrlimit 76 /* Back compatible 2Gig limited rlimit */ #define __NR_getrusage 77 #define __NR_gettimeofday 78 #define __NR_settimeofday 79 diff --git a/include/asm-x86/unistd_64.h b/include/asm-x86/unistd_64.h index 3883ceb54ef5..fe26e36d0f51 100644 --- a/include/asm-x86/unistd_64.h +++ b/include/asm-x86/unistd_64.h @@ -2,7 +2,7 @@ #define _ASM_X86_64_UNISTD_H_ #ifndef __SYSCALL -#define __SYSCALL(a,b) +#define __SYSCALL(a, b) #endif /* diff --git a/include/asm-x86/user32.h b/include/asm-x86/user32.h index f769872debea..a3d910047879 100644 --- a/include/asm-x86/user32.h +++ b/include/asm-x86/user32.h @@ -1,7 +1,8 @@ #ifndef USER32_H #define USER32_H 1 -/* IA32 compatible user structures for ptrace. These should be used for 32bit coredumps too. */ +/* IA32 compatible user structures for ptrace. + * These should be used for 32bit coredumps too. */ struct user_i387_ia32_struct { u32 cwd; @@ -42,9 +43,9 @@ struct user_regs_struct32 { }; struct user32 { - struct user_regs_struct32 regs; /* Where the registers are actually stored */ + struct user_regs_struct32 regs; /* Where the registers are actually stored */ int u_fpvalid; /* True if math co-processor being used. */ - /* for this mess. Not yet used. */ + /* for this mess. Not yet used. */ struct user_i387_ia32_struct i387; /* Math Co-processor registers. */ /* The rest of this junk is to help gdb figure out what goes where */ __u32 u_tsize; /* Text segment size (pages). */ diff --git a/include/asm-x86/user_32.h b/include/asm-x86/user_32.h index 6157da6f882c..d6e51edc259d 100644 --- a/include/asm-x86/user_32.h +++ b/include/asm-x86/user_32.h @@ -100,10 +100,10 @@ struct user_regs_struct { struct user{ /* We start with the registers, to mimic the way that "memory" is returned from the ptrace(3,...) function. */ - struct user_regs_struct regs; /* Where the registers are actually stored */ + struct user_regs_struct regs; /* Where the registers are actually stored */ /* ptrace does not yet supply these. Someday.... */ int u_fpvalid; /* True if math co-processor being used. */ - /* for this mess. Not yet used. */ + /* for this mess. Not yet used. */ struct user_i387_struct i387; /* Math Co-processor registers. */ /* The rest of this junk is to help gdb figure out what goes where */ unsigned long int u_tsize; /* Text segment size (pages). */ @@ -118,7 +118,7 @@ struct user{ int reserved; /* No longer used */ unsigned long u_ar0; /* Used by gdb to help find the values for */ /* the registers. */ - struct user_i387_struct* u_fpstate; /* Math Co-processor pointer. */ + struct user_i387_struct *u_fpstate; /* Math Co-processor pointer. */ unsigned long magic; /* To uniquely identify a core file */ char u_comm[32]; /* User command that was responsible */ int u_debugreg[8]; diff --git a/include/asm-x86/user_64.h b/include/asm-x86/user_64.h index 963616455609..6037b634c77f 100644 --- a/include/asm-x86/user_64.h +++ b/include/asm-x86/user_64.h @@ -45,12 +45,13 @@ */ /* This matches the 64bit FXSAVE format as defined by AMD. It is the same - as the 32bit format defined by Intel, except that the selector:offset pairs for - data and eip are replaced with flat 64bit pointers. */ + as the 32bit format defined by Intel, except that the selector:offset pairs + for data and eip are replaced with flat 64bit pointers. */ struct user_i387_struct { unsigned short cwd; unsigned short swd; - unsigned short twd; /* Note this is not the same as the 32bit/x87/FSAVE twd */ + unsigned short twd; /* Note this is not the same as + the 32bit/x87/FSAVE twd */ unsigned short fop; __u64 rip; __u64 rdp; @@ -97,13 +98,14 @@ struct user_regs_struct { /* When the kernel dumps core, it starts by dumping the user struct - this will be used by gdb to figure out where the data and stack segments are within the file, and what virtual addresses to use. */ -struct user{ + +struct user { /* We start with the registers, to mimic the way that "memory" is returned from the ptrace(3,...) function. */ - struct user_regs_struct regs; /* Where the registers are actually stored */ + struct user_regs_struct regs; /* Where the registers are actually stored */ /* ptrace does not yet supply these. Someday.... */ int u_fpvalid; /* True if math co-processor being used. */ - /* for this mess. Not yet used. */ + /* for this mess. Not yet used. */ int pad0; struct user_i387_struct i387; /* Math Co-processor registers. */ /* The rest of this junk is to help gdb figure out what goes where */ @@ -120,7 +122,7 @@ struct user{ int pad1; unsigned long u_ar0; /* Used by gdb to help find the values for */ /* the registers. */ - struct user_i387_struct* u_fpstate; /* Math Co-processor pointer. */ + struct user_i387_struct *u_fpstate; /* Math Co-processor pointer. */ unsigned long magic; /* To uniquely identify a core file */ char u_comm[32]; /* User command that was responsible */ unsigned long u_debugreg[8]; diff --git a/include/asm-x86/uv/uv_hub.h b/include/asm-x86/uv/uv_hub.h new file mode 100644 index 000000000000..26b9240d1e23 --- /dev/null +++ b/include/asm-x86/uv/uv_hub.h @@ -0,0 +1,284 @@ +/* + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * SGI UV architectural definitions + * + * Copyright (C) 2007 Silicon Graphics, Inc. All rights reserved. + */ + +#ifndef __ASM_X86_UV_HUB_H__ +#define __ASM_X86_UV_HUB_H__ + +#include <linux/numa.h> +#include <linux/percpu.h> +#include <asm/types.h> +#include <asm/percpu.h> + + +/* + * Addressing Terminology + * + * NASID - network ID of a router, Mbrick or Cbrick. Nasid values of + * routers always have low bit of 1, C/MBricks have low bit + * equal to 0. Most addressing macros that target UV hub chips + * right shift the NASID by 1 to exclude the always-zero bit. + * + * SNASID - NASID right shifted by 1 bit. + * + * + * Memory/UV-HUB Processor Socket Address Format: + * +--------+---------------+---------------------+ + * |00..0000| SNASID | NodeOffset | + * +--------+---------------+---------------------+ + * <--- N bits --->|<--------M bits -----> + * + * M number of node offset bits (35 .. 40) + * N number of SNASID bits (0 .. 10) + * + * Note: M + N cannot currently exceed 44 (x86_64) or 46 (IA64). + * The actual values are configuration dependent and are set at + * boot time + * + * APICID format + * NOTE!!!!!! This is the current format of the APICID. However, code + * should assume that this will change in the future. Use functions + * in this file for all APICID bit manipulations and conversion. + * + * 1111110000000000 + * 5432109876543210 + * nnnnnnnnnnlc0cch + * sssssssssss + * + * n = snasid bits + * l = socket number on board + * c = core + * h = hyperthread + * s = bits that are in the socket CSR + * + * Note: Processor only supports 12 bits in the APICID register. The ACPI + * tables hold all 16 bits. Software needs to be aware of this. + * + * Unless otherwise specified, all references to APICID refer to + * the FULL value contained in ACPI tables, not the subset in the + * processor APICID register. + */ + + +/* + * Maximum number of bricks in all partitions and in all coherency domains. + * This is the total number of bricks accessible in the numalink fabric. It + * includes all C & M bricks. Routers are NOT included. + * + * This value is also the value of the maximum number of non-router NASIDs + * in the numalink fabric. + * + * NOTE: a brick may be 1 or 2 OS nodes. Don't get these confused. + */ +#define UV_MAX_NUMALINK_BLADES 16384 + +/* + * Maximum number of C/Mbricks within a software SSI (hardware may support + * more). + */ +#define UV_MAX_SSI_BLADES 256 + +/* + * The largest possible NASID of a C or M brick (+ 2) + */ +#define UV_MAX_NASID_VALUE (UV_MAX_NUMALINK_NODES * 2) + +/* + * The following defines attributes of the HUB chip. These attributes are + * frequently referenced and are kept in the per-cpu data areas of each cpu. + * They are kept together in a struct to minimize cache misses. + */ +struct uv_hub_info_s { + unsigned long global_mmr_base; + unsigned short local_nasid; + unsigned short gnode_upper; + unsigned short coherency_domain_number; + unsigned short numa_blade_id; + unsigned char blade_processor_id; + unsigned char m_val; + unsigned char n_val; +}; +DECLARE_PER_CPU(struct uv_hub_info_s, __uv_hub_info); +#define uv_hub_info (&__get_cpu_var(__uv_hub_info)) +#define uv_cpu_hub_info(cpu) (&per_cpu(__uv_hub_info, cpu)) + +/* + * Local & Global MMR space macros. + * Note: macros are intended to be used ONLY by inline functions + * in this file - not by other kernel code. + */ +#define UV_SNASID(n) ((n) >> 1) +#define UV_NASID(n) ((n) << 1) + +#define UV_LOCAL_MMR_BASE 0xf4000000UL +#define UV_GLOBAL_MMR32_BASE 0xf8000000UL +#define UV_GLOBAL_MMR64_BASE (uv_hub_info->global_mmr_base) + +#define UV_GLOBAL_MMR32_SNASID_MASK 0x3ff +#define UV_GLOBAL_MMR32_SNASID_SHIFT 15 +#define UV_GLOBAL_MMR64_SNASID_SHIFT 26 + +#define UV_GLOBAL_MMR32_NASID_BITS(n) \ + (((UV_SNASID(n) & UV_GLOBAL_MMR32_SNASID_MASK)) << \ + (UV_GLOBAL_MMR32_SNASID_SHIFT)) + +#define UV_GLOBAL_MMR64_NASID_BITS(n) \ + ((unsigned long)UV_SNASID(n) << UV_GLOBAL_MMR64_SNASID_SHIFT) + +#define UV_APIC_NASID_SHIFT 6 + +/* + * Extract a NASID from an APICID (full apicid, not processor subset) + */ +static inline int uv_apicid_to_nasid(int apicid) +{ + return (UV_NASID(apicid >> UV_APIC_NASID_SHIFT)); +} + +/* + * Access global MMRs using the low memory MMR32 space. This region supports + * faster MMR access but not all MMRs are accessible in this space. + */ +static inline unsigned long *uv_global_mmr32_address(int nasid, + unsigned long offset) +{ + return __va(UV_GLOBAL_MMR32_BASE | + UV_GLOBAL_MMR32_NASID_BITS(nasid) | offset); +} + +static inline void uv_write_global_mmr32(int nasid, unsigned long offset, + unsigned long val) +{ + *uv_global_mmr32_address(nasid, offset) = val; +} + +static inline unsigned long uv_read_global_mmr32(int nasid, + unsigned long offset) +{ + return *uv_global_mmr32_address(nasid, offset); +} + +/* + * Access Global MMR space using the MMR space located at the top of physical + * memory. + */ +static inline unsigned long *uv_global_mmr64_address(int nasid, + unsigned long offset) +{ + return __va(UV_GLOBAL_MMR64_BASE | + UV_GLOBAL_MMR64_NASID_BITS(nasid) | offset); +} + +static inline void uv_write_global_mmr64(int nasid, unsigned long offset, + unsigned long val) +{ + *uv_global_mmr64_address(nasid, offset) = val; +} + +static inline unsigned long uv_read_global_mmr64(int nasid, + unsigned long offset) +{ + return *uv_global_mmr64_address(nasid, offset); +} + +/* + * Access node local MMRs. Faster than using global space but only local MMRs + * are accessible. + */ +static inline unsigned long *uv_local_mmr_address(unsigned long offset) +{ + return __va(UV_LOCAL_MMR_BASE | offset); +} + +static inline unsigned long uv_read_local_mmr(unsigned long offset) +{ + return *uv_local_mmr_address(offset); +} + +static inline void uv_write_local_mmr(unsigned long offset, unsigned long val) +{ + *uv_local_mmr_address(offset) = val; +} + +/* + * Structures and definitions for converting between cpu, node, and blade + * numbers. + */ +struct uv_blade_info { + unsigned short nr_posible_cpus; + unsigned short nr_online_cpus; + unsigned short nasid; +}; +struct uv_blade_info *uv_blade_info; +extern short *uv_node_to_blade; +extern short *uv_cpu_to_blade; +extern short uv_possible_blades; + +/* Blade-local cpu number of current cpu. Numbered 0 .. <# cpus on the blade> */ +static inline int uv_blade_processor_id(void) +{ + return uv_hub_info->blade_processor_id; +} + +/* Blade number of current cpu. Numnbered 0 .. <#blades -1> */ +static inline int uv_numa_blade_id(void) +{ + return uv_hub_info->numa_blade_id; +} + +/* Convert a cpu number to the the UV blade number */ +static inline int uv_cpu_to_blade_id(int cpu) +{ + return uv_cpu_to_blade[cpu]; +} + +/* Convert linux node number to the UV blade number */ +static inline int uv_node_to_blade_id(int nid) +{ + return uv_node_to_blade[nid]; +} + +/* Convert a blade id to the NASID of the blade */ +static inline int uv_blade_to_nasid(int bid) +{ + return uv_blade_info[bid].nasid; +} + +/* Determine the number of possible cpus on a blade */ +static inline int uv_blade_nr_possible_cpus(int bid) +{ + return uv_blade_info[bid].nr_posible_cpus; +} + +/* Determine the number of online cpus on a blade */ +static inline int uv_blade_nr_online_cpus(int bid) +{ + return uv_blade_info[bid].nr_online_cpus; +} + +/* Convert a cpu id to the NASID of the blade containing the cpu */ +static inline int uv_cpu_to_nasid(int cpu) +{ + return uv_blade_info[uv_cpu_to_blade_id(cpu)].nasid; +} + +/* Convert a node number to the NASID of the blade */ +static inline int uv_node_to_nasid(int nid) +{ + return uv_blade_info[uv_node_to_blade_id(nid)].nasid; +} + +/* Maximum possible number of blades */ +static inline int uv_num_possible_blades(void) +{ + return uv_possible_blades; +} + +#endif /* __ASM_X86_UV_HUB__ */ + diff --git a/include/asm-x86/uv/uv_mmrs.h b/include/asm-x86/uv/uv_mmrs.h new file mode 100644 index 000000000000..3b69fe6b6376 --- /dev/null +++ b/include/asm-x86/uv/uv_mmrs.h @@ -0,0 +1,373 @@ +/* + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * SGI UV MMR definitions + * + * Copyright (C) 2007-2008 Silicon Graphics, Inc. All rights reserved. + */ + +#ifndef __ASM_X86_UV_MMRS__ +#define __ASM_X86_UV_MMRS__ + +/* + * AUTO GENERATED - Do not edit + */ + + #define UV_MMR_ENABLE (1UL << 63) + +/* ========================================================================= */ +/* UVH_IPI_INT */ +/* ========================================================================= */ +#define UVH_IPI_INT 0x60500UL +#define UVH_IPI_INT_32 0x0360 + +#define UVH_IPI_INT_VECTOR_SHFT 0 +#define UVH_IPI_INT_VECTOR_MASK 0x00000000000000ffUL +#define UVH_IPI_INT_DELIVERY_MODE_SHFT 8 +#define UVH_IPI_INT_DELIVERY_MODE_MASK 0x0000000000000700UL +#define UVH_IPI_INT_DESTMODE_SHFT 11 +#define UVH_IPI_INT_DESTMODE_MASK 0x0000000000000800UL +#define UVH_IPI_INT_APIC_ID_SHFT 16 +#define UVH_IPI_INT_APIC_ID_MASK 0x0000ffffffff0000UL +#define UVH_IPI_INT_SEND_SHFT 63 +#define UVH_IPI_INT_SEND_MASK 0x8000000000000000UL + +union uvh_ipi_int_u { + unsigned long v; + struct uvh_ipi_int_s { + unsigned long vector_ : 8; /* RW */ + unsigned long delivery_mode : 3; /* RW */ + unsigned long destmode : 1; /* RW */ + unsigned long rsvd_12_15 : 4; /* */ + unsigned long apic_id : 32; /* RW */ + unsigned long rsvd_48_62 : 15; /* */ + unsigned long send : 1; /* WP */ + } s; +}; + +/* ========================================================================= */ +/* UVH_LB_BAU_INTD_PAYLOAD_QUEUE_FIRST */ +/* ========================================================================= */ +#define UVH_LB_BAU_INTD_PAYLOAD_QUEUE_FIRST 0x320050UL +#define UVH_LB_BAU_INTD_PAYLOAD_QUEUE_FIRST_32 0x009f0 + +#define UVH_LB_BAU_INTD_PAYLOAD_QUEUE_FIRST_ADDRESS_SHFT 4 +#define UVH_LB_BAU_INTD_PAYLOAD_QUEUE_FIRST_ADDRESS_MASK 0x000007fffffffff0UL +#define UVH_LB_BAU_INTD_PAYLOAD_QUEUE_FIRST_NODE_ID_SHFT 49 +#define UVH_LB_BAU_INTD_PAYLOAD_QUEUE_FIRST_NODE_ID_MASK 0x7ffe000000000000UL + +union uvh_lb_bau_intd_payload_queue_first_u { + unsigned long v; + struct uvh_lb_bau_intd_payload_queue_first_s { + unsigned long rsvd_0_3: 4; /* */ + unsigned long address : 39; /* RW */ + unsigned long rsvd_43_48: 6; /* */ + unsigned long node_id : 14; /* RW */ + unsigned long rsvd_63 : 1; /* */ + } s; +}; + +/* ========================================================================= */ +/* UVH_LB_BAU_INTD_PAYLOAD_QUEUE_LAST */ +/* ========================================================================= */ +#define UVH_LB_BAU_INTD_PAYLOAD_QUEUE_LAST 0x320060UL +#define UVH_LB_BAU_INTD_PAYLOAD_QUEUE_LAST_32 0x009f8 + +#define UVH_LB_BAU_INTD_PAYLOAD_QUEUE_LAST_ADDRESS_SHFT 4 +#define UVH_LB_BAU_INTD_PAYLOAD_QUEUE_LAST_ADDRESS_MASK 0x000007fffffffff0UL + +union uvh_lb_bau_intd_payload_queue_last_u { + unsigned long v; + struct uvh_lb_bau_intd_payload_queue_last_s { + unsigned long rsvd_0_3: 4; /* */ + unsigned long address : 39; /* RW */ + unsigned long rsvd_43_63: 21; /* */ + } s; +}; + +/* ========================================================================= */ +/* UVH_LB_BAU_INTD_PAYLOAD_QUEUE_TAIL */ +/* ========================================================================= */ +#define UVH_LB_BAU_INTD_PAYLOAD_QUEUE_TAIL 0x320070UL +#define UVH_LB_BAU_INTD_PAYLOAD_QUEUE_TAIL_32 0x00a00 + +#define UVH_LB_BAU_INTD_PAYLOAD_QUEUE_TAIL_ADDRESS_SHFT 4 +#define UVH_LB_BAU_INTD_PAYLOAD_QUEUE_TAIL_ADDRESS_MASK 0x000007fffffffff0UL + +union uvh_lb_bau_intd_payload_queue_tail_u { + unsigned long v; + struct uvh_lb_bau_intd_payload_queue_tail_s { + unsigned long rsvd_0_3: 4; /* */ + unsigned long address : 39; /* RW */ + unsigned long rsvd_43_63: 21; /* */ + } s; +}; + +/* ========================================================================= */ +/* UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE */ +/* ========================================================================= */ +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE 0x320080UL + +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_PENDING_0_SHFT 0 +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_PENDING_0_MASK 0x0000000000000001UL +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_PENDING_1_SHFT 1 +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_PENDING_1_MASK 0x0000000000000002UL +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_PENDING_2_SHFT 2 +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_PENDING_2_MASK 0x0000000000000004UL +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_PENDING_3_SHFT 3 +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_PENDING_3_MASK 0x0000000000000008UL +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_PENDING_4_SHFT 4 +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_PENDING_4_MASK 0x0000000000000010UL +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_PENDING_5_SHFT 5 +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_PENDING_5_MASK 0x0000000000000020UL +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_PENDING_6_SHFT 6 +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_PENDING_6_MASK 0x0000000000000040UL +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_PENDING_7_SHFT 7 +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_PENDING_7_MASK 0x0000000000000080UL +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_TIMEOUT_0_SHFT 8 +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_TIMEOUT_0_MASK 0x0000000000000100UL +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_TIMEOUT_1_SHFT 9 +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_TIMEOUT_1_MASK 0x0000000000000200UL +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_TIMEOUT_2_SHFT 10 +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_TIMEOUT_2_MASK 0x0000000000000400UL +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_TIMEOUT_3_SHFT 11 +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_TIMEOUT_3_MASK 0x0000000000000800UL +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_TIMEOUT_4_SHFT 12 +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_TIMEOUT_4_MASK 0x0000000000001000UL +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_TIMEOUT_5_SHFT 13 +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_TIMEOUT_5_MASK 0x0000000000002000UL +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_TIMEOUT_6_SHFT 14 +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_TIMEOUT_6_MASK 0x0000000000004000UL +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_TIMEOUT_7_SHFT 15 +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_TIMEOUT_7_MASK 0x0000000000008000UL +union uvh_lb_bau_intd_software_acknowledge_u { + unsigned long v; + struct uvh_lb_bau_intd_software_acknowledge_s { + unsigned long pending_0 : 1; /* RW, W1C */ + unsigned long pending_1 : 1; /* RW, W1C */ + unsigned long pending_2 : 1; /* RW, W1C */ + unsigned long pending_3 : 1; /* RW, W1C */ + unsigned long pending_4 : 1; /* RW, W1C */ + unsigned long pending_5 : 1; /* RW, W1C */ + unsigned long pending_6 : 1; /* RW, W1C */ + unsigned long pending_7 : 1; /* RW, W1C */ + unsigned long timeout_0 : 1; /* RW, W1C */ + unsigned long timeout_1 : 1; /* RW, W1C */ + unsigned long timeout_2 : 1; /* RW, W1C */ + unsigned long timeout_3 : 1; /* RW, W1C */ + unsigned long timeout_4 : 1; /* RW, W1C */ + unsigned long timeout_5 : 1; /* RW, W1C */ + unsigned long timeout_6 : 1; /* RW, W1C */ + unsigned long timeout_7 : 1; /* RW, W1C */ + unsigned long rsvd_16_63: 48; /* */ + } s; +}; + +/* ========================================================================= */ +/* UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_ALIAS */ +/* ========================================================================= */ +#define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_ALIAS 0x0000000000320088UL + +/* ========================================================================= */ +/* UVH_LB_BAU_SB_ACTIVATION_CONTROL */ +/* ========================================================================= */ +#define UVH_LB_BAU_SB_ACTIVATION_CONTROL 0x320020UL +#define UVH_LB_BAU_SB_ACTIVATION_CONTROL_32 0x009d8 + +#define UVH_LB_BAU_SB_ACTIVATION_CONTROL_INDEX_SHFT 0 +#define UVH_LB_BAU_SB_ACTIVATION_CONTROL_INDEX_MASK 0x000000000000003fUL +#define UVH_LB_BAU_SB_ACTIVATION_CONTROL_PUSH_SHFT 62 +#define UVH_LB_BAU_SB_ACTIVATION_CONTROL_PUSH_MASK 0x4000000000000000UL +#define UVH_LB_BAU_SB_ACTIVATION_CONTROL_INIT_SHFT 63 +#define UVH_LB_BAU_SB_ACTIVATION_CONTROL_INIT_MASK 0x8000000000000000UL + +union uvh_lb_bau_sb_activation_control_u { + unsigned long v; + struct uvh_lb_bau_sb_activation_control_s { + unsigned long index : 6; /* RW */ + unsigned long rsvd_6_61: 56; /* */ + unsigned long push : 1; /* WP */ + unsigned long init : 1; /* WP */ + } s; +}; + +/* ========================================================================= */ +/* UVH_LB_BAU_SB_ACTIVATION_STATUS_0 */ +/* ========================================================================= */ +#define UVH_LB_BAU_SB_ACTIVATION_STATUS_0 0x320030UL +#define UVH_LB_BAU_SB_ACTIVATION_STATUS_0_32 0x009e0 + +#define UVH_LB_BAU_SB_ACTIVATION_STATUS_0_STATUS_SHFT 0 +#define UVH_LB_BAU_SB_ACTIVATION_STATUS_0_STATUS_MASK 0xffffffffffffffffUL + +union uvh_lb_bau_sb_activation_status_0_u { + unsigned long v; + struct uvh_lb_bau_sb_activation_status_0_s { + unsigned long status : 64; /* RW */ + } s; +}; + +/* ========================================================================= */ +/* UVH_LB_BAU_SB_ACTIVATION_STATUS_1 */ +/* ========================================================================= */ +#define UVH_LB_BAU_SB_ACTIVATION_STATUS_1 0x320040UL +#define UVH_LB_BAU_SB_ACTIVATION_STATUS_1_32 0x009e8 + +#define UVH_LB_BAU_SB_ACTIVATION_STATUS_1_STATUS_SHFT 0 +#define UVH_LB_BAU_SB_ACTIVATION_STATUS_1_STATUS_MASK 0xffffffffffffffffUL + +union uvh_lb_bau_sb_activation_status_1_u { + unsigned long v; + struct uvh_lb_bau_sb_activation_status_1_s { + unsigned long status : 64; /* RW */ + } s; +}; + +/* ========================================================================= */ +/* UVH_LB_BAU_SB_DESCRIPTOR_BASE */ +/* ========================================================================= */ +#define UVH_LB_BAU_SB_DESCRIPTOR_BASE 0x320010UL +#define UVH_LB_BAU_SB_DESCRIPTOR_BASE_32 0x009d0 + +#define UVH_LB_BAU_SB_DESCRIPTOR_BASE_PAGE_ADDRESS_SHFT 12 +#define UVH_LB_BAU_SB_DESCRIPTOR_BASE_PAGE_ADDRESS_MASK 0x000007fffffff000UL +#define UVH_LB_BAU_SB_DESCRIPTOR_BASE_NODE_ID_SHFT 49 +#define UVH_LB_BAU_SB_DESCRIPTOR_BASE_NODE_ID_MASK 0x7ffe000000000000UL + +union uvh_lb_bau_sb_descriptor_base_u { + unsigned long v; + struct uvh_lb_bau_sb_descriptor_base_s { + unsigned long rsvd_0_11 : 12; /* */ + unsigned long page_address : 31; /* RW */ + unsigned long rsvd_43_48 : 6; /* */ + unsigned long node_id : 14; /* RW */ + unsigned long rsvd_63 : 1; /* */ + } s; +}; + +/* ========================================================================= */ +/* UVH_NODE_ID */ +/* ========================================================================= */ +#define UVH_NODE_ID 0x0UL + +#define UVH_NODE_ID_FORCE1_SHFT 0 +#define UVH_NODE_ID_FORCE1_MASK 0x0000000000000001UL +#define UVH_NODE_ID_MANUFACTURER_SHFT 1 +#define UVH_NODE_ID_MANUFACTURER_MASK 0x0000000000000ffeUL +#define UVH_NODE_ID_PART_NUMBER_SHFT 12 +#define UVH_NODE_ID_PART_NUMBER_MASK 0x000000000ffff000UL +#define UVH_NODE_ID_REVISION_SHFT 28 +#define UVH_NODE_ID_REVISION_MASK 0x00000000f0000000UL +#define UVH_NODE_ID_NODE_ID_SHFT 32 +#define UVH_NODE_ID_NODE_ID_MASK 0x00007fff00000000UL +#define UVH_NODE_ID_NODES_PER_BIT_SHFT 48 +#define UVH_NODE_ID_NODES_PER_BIT_MASK 0x007f000000000000UL +#define UVH_NODE_ID_NI_PORT_SHFT 56 +#define UVH_NODE_ID_NI_PORT_MASK 0x0f00000000000000UL + +union uvh_node_id_u { + unsigned long v; + struct uvh_node_id_s { + unsigned long force1 : 1; /* RO */ + unsigned long manufacturer : 11; /* RO */ + unsigned long part_number : 16; /* RO */ + unsigned long revision : 4; /* RO */ + unsigned long node_id : 15; /* RW */ + unsigned long rsvd_47 : 1; /* */ + unsigned long nodes_per_bit : 7; /* RW */ + unsigned long rsvd_55 : 1; /* */ + unsigned long ni_port : 4; /* RO */ + unsigned long rsvd_60_63 : 4; /* */ + } s; +}; + +/* ========================================================================= */ +/* UVH_RH_GAM_GRU_OVERLAY_CONFIG_MMR */ +/* ========================================================================= */ +#define UVH_RH_GAM_GRU_OVERLAY_CONFIG_MMR 0x1600010UL + +#define UVH_RH_GAM_GRU_OVERLAY_CONFIG_MMR_BASE_SHFT 28 +#define UVH_RH_GAM_GRU_OVERLAY_CONFIG_MMR_BASE_MASK 0x00003ffff0000000UL +#define UVH_RH_GAM_GRU_OVERLAY_CONFIG_MMR_GR4_SHFT 46 +#define UVH_RH_GAM_GRU_OVERLAY_CONFIG_MMR_GR4_MASK 0x0000400000000000UL +#define UVH_RH_GAM_GRU_OVERLAY_CONFIG_MMR_N_GRU_SHFT 52 +#define UVH_RH_GAM_GRU_OVERLAY_CONFIG_MMR_N_GRU_MASK 0x00f0000000000000UL +#define UVH_RH_GAM_GRU_OVERLAY_CONFIG_MMR_ENABLE_SHFT 63 +#define UVH_RH_GAM_GRU_OVERLAY_CONFIG_MMR_ENABLE_MASK 0x8000000000000000UL + +union uvh_rh_gam_gru_overlay_config_mmr_u { + unsigned long v; + struct uvh_rh_gam_gru_overlay_config_mmr_s { + unsigned long rsvd_0_27: 28; /* */ + unsigned long base : 18; /* RW */ + unsigned long gr4 : 1; /* RW */ + unsigned long rsvd_47_51: 5; /* */ + unsigned long n_gru : 4; /* RW */ + unsigned long rsvd_56_62: 7; /* */ + unsigned long enable : 1; /* RW */ + } s; +}; + +/* ========================================================================= */ +/* UVH_RH_GAM_MMR_OVERLAY_CONFIG_MMR */ +/* ========================================================================= */ +#define UVH_RH_GAM_MMR_OVERLAY_CONFIG_MMR 0x1600028UL + +#define UVH_RH_GAM_MMR_OVERLAY_CONFIG_MMR_BASE_SHFT 26 +#define UVH_RH_GAM_MMR_OVERLAY_CONFIG_MMR_BASE_MASK 0x00003ffffc000000UL +#define UVH_RH_GAM_MMR_OVERLAY_CONFIG_MMR_DUAL_HUB_SHFT 46 +#define UVH_RH_GAM_MMR_OVERLAY_CONFIG_MMR_DUAL_HUB_MASK 0x0000400000000000UL +#define UVH_RH_GAM_MMR_OVERLAY_CONFIG_MMR_ENABLE_SHFT 63 +#define UVH_RH_GAM_MMR_OVERLAY_CONFIG_MMR_ENABLE_MASK 0x8000000000000000UL + +union uvh_rh_gam_mmr_overlay_config_mmr_u { + unsigned long v; + struct uvh_rh_gam_mmr_overlay_config_mmr_s { + unsigned long rsvd_0_25: 26; /* */ + unsigned long base : 20; /* RW */ + unsigned long dual_hub : 1; /* RW */ + unsigned long rsvd_47_62: 16; /* */ + unsigned long enable : 1; /* RW */ + } s; +}; + +/* ========================================================================= */ +/* UVH_RTC */ +/* ========================================================================= */ +#define UVH_RTC 0x28000UL + +#define UVH_RTC_REAL_TIME_CLOCK_SHFT 0 +#define UVH_RTC_REAL_TIME_CLOCK_MASK 0x00ffffffffffffffUL + +union uvh_rtc_u { + unsigned long v; + struct uvh_rtc_s { + unsigned long real_time_clock : 56; /* RW */ + unsigned long rsvd_56_63 : 8; /* */ + } s; +}; + +/* ========================================================================= */ +/* UVH_SI_ADDR_MAP_CONFIG */ +/* ========================================================================= */ +#define UVH_SI_ADDR_MAP_CONFIG 0xc80000UL + +#define UVH_SI_ADDR_MAP_CONFIG_M_SKT_SHFT 0 +#define UVH_SI_ADDR_MAP_CONFIG_M_SKT_MASK 0x000000000000003fUL +#define UVH_SI_ADDR_MAP_CONFIG_N_SKT_SHFT 8 +#define UVH_SI_ADDR_MAP_CONFIG_N_SKT_MASK 0x0000000000000f00UL + +union uvh_si_addr_map_config_u { + unsigned long v; + struct uvh_si_addr_map_config_s { + unsigned long m_skt : 6; /* RW */ + unsigned long rsvd_6_7: 2; /* */ + unsigned long n_skt : 4; /* RW */ + unsigned long rsvd_12_63: 52; /* */ + } s; +}; + + +#endif /* __ASM_X86_UV_MMRS__ */ diff --git a/include/asm-x86/vdso.h b/include/asm-x86/vdso.h index 629bcb6e8e45..86e085e003d2 100644 --- a/include/asm-x86/vdso.h +++ b/include/asm-x86/vdso.h @@ -8,9 +8,11 @@ extern const char VDSO64_PRELINK[]; * Given a pointer to the vDSO image, find the pointer to VDSO64_name * as that symbol is defined in the vDSO sources or linker script. */ -#define VDSO64_SYMBOL(base, name) ({ \ - extern const char VDSO64_##name[]; \ - (void *) (VDSO64_##name - VDSO64_PRELINK + (unsigned long) (base)); }) +#define VDSO64_SYMBOL(base, name) \ +({ \ + extern const char VDSO64_##name[]; \ + (void *)(VDSO64_##name - VDSO64_PRELINK + (unsigned long)(base)); \ +}) #endif #if defined CONFIG_X86_32 || defined CONFIG_COMPAT @@ -20,9 +22,18 @@ extern const char VDSO32_PRELINK[]; * Given a pointer to the vDSO image, find the pointer to VDSO32_name * as that symbol is defined in the vDSO sources or linker script. */ -#define VDSO32_SYMBOL(base, name) ({ \ - extern const char VDSO32_##name[]; \ - (void *) (VDSO32_##name - VDSO32_PRELINK + (unsigned long) (base)); }) +#define VDSO32_SYMBOL(base, name) \ +({ \ + extern const char VDSO32_##name[]; \ + (void *)(VDSO32_##name - VDSO32_PRELINK + (unsigned long)(base)); \ +}) #endif +/* + * These symbols are defined with the addresses in the vsyscall page. + * See vsyscall-sigreturn.S. + */ +extern void __user __kernel_sigreturn; +extern void __user __kernel_rt_sigreturn; + #endif /* asm-x86/vdso.h */ diff --git a/include/asm-x86/vga.h b/include/asm-x86/vga.h index 0ecf68ac03aa..0ccf804377e6 100644 --- a/include/asm-x86/vga.h +++ b/include/asm-x86/vga.h @@ -12,9 +12,9 @@ * access the videoram directly without any black magic. */ -#define VGA_MAP_MEM(x,s) (unsigned long)phys_to_virt(x) +#define VGA_MAP_MEM(x, s) (unsigned long)phys_to_virt(x) #define vga_readb(x) (*(x)) -#define vga_writeb(x,y) (*(y) = (x)) +#define vga_writeb(x, y) (*(y) = (x)) #endif diff --git a/include/asm-x86/vm86.h b/include/asm-x86/vm86.h index c92fe4af52e8..074b357146df 100644 --- a/include/asm-x86/vm86.h +++ b/include/asm-x86/vm86.h @@ -12,19 +12,13 @@ * Linus */ -#define TF_MASK 0x00000100 -#define IF_MASK 0x00000200 -#define IOPL_MASK 0x00003000 -#define NT_MASK 0x00004000 +#include <asm/processor-flags.h> + #ifdef CONFIG_VM86 -#define VM_MASK 0x00020000 +#define X86_VM_MASK X86_EFLAGS_VM #else -#define VM_MASK 0 /* ignored */ +#define X86_VM_MASK 0 /* No VM86 support */ #endif -#define AC_MASK 0x00040000 -#define VIF_MASK 0x00080000 /* virtual interrupt flag */ -#define VIP_MASK 0x00100000 /* virtual interrupt pending */ -#define ID_MASK 0x00200000 #define BIOSSEG 0x0f000 @@ -42,9 +36,11 @@ #define VM86_ARG(retval) ((retval) >> 8) #define VM86_SIGNAL 0 /* return due to signal */ -#define VM86_UNKNOWN 1 /* unhandled GP fault - IO-instruction or similar */ +#define VM86_UNKNOWN 1 /* unhandled GP fault + - IO-instruction or similar */ #define VM86_INTx 2 /* int3/int x instruction (ARG = x) */ -#define VM86_STI 3 /* sti/popf/iret instruction enabled virtual interrupts */ +#define VM86_STI 3 /* sti/popf/iret instruction enabled + virtual interrupts */ /* * Additional return values when invoking new vm86() @@ -205,7 +201,8 @@ void release_vm86_irqs(struct task_struct *); #define handle_vm86_fault(a, b) #define release_vm86_irqs(a) -static inline int handle_vm86_trap(struct kernel_vm86_regs *a, long b, int c) { +static inline int handle_vm86_trap(struct kernel_vm86_regs *a, long b, int c) +{ return 0; } diff --git a/include/asm-x86/vmi.h b/include/asm-x86/vmi.h index eb8bd892c01e..b7c0dea119fe 100644 --- a/include/asm-x86/vmi.h +++ b/include/asm-x86/vmi.h @@ -155,9 +155,9 @@ #ifndef __ASSEMBLY__ struct vmi_relocation_info { - unsigned char *eip; - unsigned char type; - unsigned char reserved[3]; + unsigned char *eip; + unsigned char type; + unsigned char reserved[3]; }; #endif @@ -173,53 +173,53 @@ struct vmi_relocation_info { #ifndef __ASSEMBLY__ struct vrom_header { - u16 rom_signature; // option ROM signature - u8 rom_length; // ROM length in 512 byte chunks - u8 rom_entry[4]; // 16-bit code entry point - u8 rom_pad0; // 4-byte align pad - u32 vrom_signature; // VROM identification signature - u8 api_version_min;// Minor version of API - u8 api_version_maj;// Major version of API - u8 jump_slots; // Number of jump slots - u8 reserved1; // Reserved for expansion - u32 virtual_top; // Hypervisor virtual address start - u16 reserved2; // Reserved for expansion - u16 license_offs; // Offset to License string - u16 pci_header_offs;// Offset to PCI OPROM header - u16 pnp_header_offs;// Offset to PnP OPROM header - u32 rom_pad3; // PnP reserverd / VMI reserved - u8 reserved[96]; // Reserved for headers - char vmi_init[8]; // VMI_Init jump point - char get_reloc[8]; // VMI_GetRelocationInfo jump point + u16 rom_signature; /* option ROM signature */ + u8 rom_length; /* ROM length in 512 byte chunks */ + u8 rom_entry[4]; /* 16-bit code entry point */ + u8 rom_pad0; /* 4-byte align pad */ + u32 vrom_signature; /* VROM identification signature */ + u8 api_version_min;/* Minor version of API */ + u8 api_version_maj;/* Major version of API */ + u8 jump_slots; /* Number of jump slots */ + u8 reserved1; /* Reserved for expansion */ + u32 virtual_top; /* Hypervisor virtual address start */ + u16 reserved2; /* Reserved for expansion */ + u16 license_offs; /* Offset to License string */ + u16 pci_header_offs;/* Offset to PCI OPROM header */ + u16 pnp_header_offs;/* Offset to PnP OPROM header */ + u32 rom_pad3; /* PnP reserverd / VMI reserved */ + u8 reserved[96]; /* Reserved for headers */ + char vmi_init[8]; /* VMI_Init jump point */ + char get_reloc[8]; /* VMI_GetRelocationInfo jump point */ } __attribute__((packed)); struct pnp_header { - char sig[4]; - char rev; - char size; - short next; - short res; - long devID; - unsigned short manufacturer_offset; - unsigned short product_offset; + char sig[4]; + char rev; + char size; + short next; + short res; + long devID; + unsigned short manufacturer_offset; + unsigned short product_offset; } __attribute__((packed)); struct pci_header { - char sig[4]; - short vendorID; - short deviceID; - short vpdData; - short size; - char rev; - char class; - char subclass; - char interface; - short chunks; - char rom_version_min; - char rom_version_maj; - char codetype; - char lastRom; - short reserved; + char sig[4]; + short vendorID; + short deviceID; + short vpdData; + short size; + char rev; + char class; + char subclass; + char interface; + short chunks; + char rom_version_min; + char rom_version_maj; + char codetype; + char lastRom; + short reserved; } __attribute__((packed)); /* Function prototypes for bootstrapping */ diff --git a/include/asm-x86/voyager.h b/include/asm-x86/voyager.h index 91a9932937ab..9c811d2e6f91 100644 --- a/include/asm-x86/voyager.h +++ b/include/asm-x86/voyager.h @@ -91,8 +91,7 @@ #define VOYAGER_WRITE_CONFIG 0x2 #define VOYAGER_BYPASS 0xff -typedef struct voyager_asic -{ +typedef struct voyager_asic { __u8 asic_addr; /* ASIC address; Level 4 */ __u8 asic_type; /* ASIC type */ __u8 asic_id; /* ASIC id */ @@ -113,7 +112,7 @@ typedef struct voyager_module { __u16 largest_reg; /* Largest register in the scan path */ __u16 smallest_reg; /* Smallest register in the scan path */ voyager_asic_t *asic; /* First ASIC in scan path (CAT_I) */ - struct voyager_module *submodule; /* Submodule pointer */ + struct voyager_module *submodule; /* Submodule pointer */ struct voyager_module *next; /* Next module in linked list */ } voyager_module_t; @@ -135,7 +134,7 @@ typedef struct voyager_eeprom_hdr { __u16 cct_offset; __u16 log_length; /* length of err log */ __u16 xsum_end; /* offset to end of - checksum */ + checksum */ __u8 reserved[4]; __u8 sflag; /* starting sentinal */ __u8 part_number[13]; /* prom part number */ @@ -148,7 +147,8 @@ typedef struct voyager_eeprom_hdr { -#define VOYAGER_EPROM_SIZE_OFFSET ((__u16)(&(((voyager_eprom_hdr_t *)0)->ee_size))) +#define VOYAGER_EPROM_SIZE_OFFSET \ + ((__u16)(&(((voyager_eprom_hdr_t *)0)->ee_size))) #define VOYAGER_XSUM_END_OFFSET 0x2a /* the following three definitions are for internal table layouts @@ -199,7 +199,7 @@ typedef struct voyager_asic_data_table { #define VOYAGER_WCBIC_TOM_L 0x4 #define VOYAGER_WCBIC_TOM_H 0x5 -/* register defines for Voyager Memory Contol (VMC) +/* register defines for Voyager Memory Contol (VMC) * these are present on L4 machines only */ #define VOYAGER_VMC1 0x81 #define VOYAGER_VMC2 0x91 @@ -334,7 +334,7 @@ typedef struct { struct QuadDescription { __u8 Type; /* for type 0 (DYADIC or MONADIC) all fields - * will be zero except for slot */ + * will be zero except for slot */ __u8 StructureVersion; __u32 CPI_BaseAddress; __u32 LARC_BankSize; @@ -342,7 +342,7 @@ struct QuadDescription { __u8 Slot; /* Processor slots 1 - 4 */ } __attribute__((packed)); -struct ProcBoardInfo { +struct ProcBoardInfo { __u8 Type; __u8 StructureVersion; __u8 NumberOfBoards; @@ -382,19 +382,30 @@ struct CPU_Info { * packed in it by our friend the compiler. */ typedef struct { - __u8 Mailbox_SUS; /* Written to by SUS to give commands/response to the OS */ - __u8 Mailbox_OS; /* Written to by the OS to give commands/response to SUS */ - __u8 SUS_MailboxVersion; /* Tells the OS which iteration of the interface SUS supports */ - __u8 OS_MailboxVersion; /* Tells SUS which iteration of the interface the OS supports */ - __u32 OS_Flags; /* Flags set by the OS as info for SUS */ - __u32 SUS_Flags; /* Flags set by SUS as info for the OS */ - __u32 WatchDogPeriod; /* Watchdog period (in seconds) which the DP uses to see if the OS is dead */ + __u8 Mailbox_SUS; /* Written to by SUS to give + commands/response to the OS */ + __u8 Mailbox_OS; /* Written to by the OS to give + commands/response to SUS */ + __u8 SUS_MailboxVersion; /* Tells the OS which iteration of the + interface SUS supports */ + __u8 OS_MailboxVersion; /* Tells SUS which iteration of the + interface the OS supports */ + __u32 OS_Flags; /* Flags set by the OS as info for + SUS */ + __u32 SUS_Flags; /* Flags set by SUS as info + for the OS */ + __u32 WatchDogPeriod; /* Watchdog period (in seconds) which + the DP uses to see if the OS + is dead */ __u32 WatchDogCount; /* Updated by the OS on every tic. */ - __u32 MemoryFor_SUS_ErrorLog; /* Flat 32 bit address which tells SUS where to stuff the SUS error log on a dump */ - MC_SlotInformation_t MC_SlotInfo[NUMBER_OF_MC_BUSSES*SLOTS_PER_MC_BUS]; /* Storage for MCA POS data */ + __u32 MemoryFor_SUS_ErrorLog; /* Flat 32 bit address which tells SUS + where to stuff the SUS error log + on a dump */ + MC_SlotInformation_t MC_SlotInfo[NUMBER_OF_MC_BUSSES*SLOTS_PER_MC_BUS]; + /* Storage for MCA POS data */ /* All new SECOND_PASS_INTERFACE fields added from this point */ - struct ProcBoardInfo *BoardData; - struct CPU_Info *CPU_Data; + struct ProcBoardInfo *BoardData; + struct CPU_Info *CPU_Data; /* All new fields must be added from this point */ } Voyager_KernelSUS_Mbox_t; @@ -478,7 +489,7 @@ struct voyager_SUS { __u32 SUS_errorlog; /* lots of system configuration stuff under here */ }; - + /* Variables exported by voyager_smp */ extern __u32 voyager_extended_vic_processors; extern __u32 voyager_allowed_boot_processors; diff --git a/include/asm-x86/xor_32.h b/include/asm-x86/xor_32.h index a41ef1bdd424..067b5c1835a3 100644 --- a/include/asm-x86/xor_32.h +++ b/include/asm-x86/xor_32.h @@ -16,12 +16,12 @@ * Copyright (C) 1998 Ingo Molnar. */ -#define LD(x,y) " movq 8*("#x")(%1), %%mm"#y" ;\n" -#define ST(x,y) " movq %%mm"#y", 8*("#x")(%1) ;\n" -#define XO1(x,y) " pxor 8*("#x")(%2), %%mm"#y" ;\n" -#define XO2(x,y) " pxor 8*("#x")(%3), %%mm"#y" ;\n" -#define XO3(x,y) " pxor 8*("#x")(%4), %%mm"#y" ;\n" -#define XO4(x,y) " pxor 8*("#x")(%5), %%mm"#y" ;\n" +#define LD(x, y) " movq 8*("#x")(%1), %%mm"#y" ;\n" +#define ST(x, y) " movq %%mm"#y", 8*("#x")(%1) ;\n" +#define XO1(x, y) " pxor 8*("#x")(%2), %%mm"#y" ;\n" +#define XO2(x, y) " pxor 8*("#x")(%3), %%mm"#y" ;\n" +#define XO3(x, y) " pxor 8*("#x")(%4), %%mm"#y" ;\n" +#define XO4(x, y) " pxor 8*("#x")(%5), %%mm"#y" ;\n" #include <asm/i387.h> @@ -32,24 +32,24 @@ xor_pII_mmx_2(unsigned long bytes, unsigned long *p1, unsigned long *p2) kernel_fpu_begin(); - __asm__ __volatile__ ( + asm volatile( #undef BLOCK -#define BLOCK(i) \ - LD(i,0) \ - LD(i+1,1) \ - LD(i+2,2) \ - LD(i+3,3) \ - XO1(i,0) \ - ST(i,0) \ - XO1(i+1,1) \ - ST(i+1,1) \ - XO1(i+2,2) \ - ST(i+2,2) \ - XO1(i+3,3) \ - ST(i+3,3) +#define BLOCK(i) \ + LD(i, 0) \ + LD(i + 1, 1) \ + LD(i + 2, 2) \ + LD(i + 3, 3) \ + XO1(i, 0) \ + ST(i, 0) \ + XO1(i+1, 1) \ + ST(i+1, 1) \ + XO1(i + 2, 2) \ + ST(i + 2, 2) \ + XO1(i + 3, 3) \ + ST(i + 3, 3) " .align 32 ;\n" - " 1: ;\n" + " 1: ;\n" BLOCK(0) BLOCK(4) @@ -76,25 +76,25 @@ xor_pII_mmx_3(unsigned long bytes, unsigned long *p1, unsigned long *p2, kernel_fpu_begin(); - __asm__ __volatile__ ( + asm volatile( #undef BLOCK -#define BLOCK(i) \ - LD(i,0) \ - LD(i+1,1) \ - LD(i+2,2) \ - LD(i+3,3) \ - XO1(i,0) \ - XO1(i+1,1) \ - XO1(i+2,2) \ - XO1(i+3,3) \ - XO2(i,0) \ - ST(i,0) \ - XO2(i+1,1) \ - ST(i+1,1) \ - XO2(i+2,2) \ - ST(i+2,2) \ - XO2(i+3,3) \ - ST(i+3,3) +#define BLOCK(i) \ + LD(i, 0) \ + LD(i + 1, 1) \ + LD(i + 2, 2) \ + LD(i + 3, 3) \ + XO1(i, 0) \ + XO1(i + 1, 1) \ + XO1(i + 2, 2) \ + XO1(i + 3, 3) \ + XO2(i, 0) \ + ST(i, 0) \ + XO2(i + 1, 1) \ + ST(i + 1, 1) \ + XO2(i + 2, 2) \ + ST(i + 2, 2) \ + XO2(i + 3, 3) \ + ST(i + 3, 3) " .align 32 ;\n" " 1: ;\n" @@ -125,29 +125,29 @@ xor_pII_mmx_4(unsigned long bytes, unsigned long *p1, unsigned long *p2, kernel_fpu_begin(); - __asm__ __volatile__ ( + asm volatile( #undef BLOCK -#define BLOCK(i) \ - LD(i,0) \ - LD(i+1,1) \ - LD(i+2,2) \ - LD(i+3,3) \ - XO1(i,0) \ - XO1(i+1,1) \ - XO1(i+2,2) \ - XO1(i+3,3) \ - XO2(i,0) \ - XO2(i+1,1) \ - XO2(i+2,2) \ - XO2(i+3,3) \ - XO3(i,0) \ - ST(i,0) \ - XO3(i+1,1) \ - ST(i+1,1) \ - XO3(i+2,2) \ - ST(i+2,2) \ - XO3(i+3,3) \ - ST(i+3,3) +#define BLOCK(i) \ + LD(i, 0) \ + LD(i + 1, 1) \ + LD(i + 2, 2) \ + LD(i + 3, 3) \ + XO1(i, 0) \ + XO1(i + 1, 1) \ + XO1(i + 2, 2) \ + XO1(i + 3, 3) \ + XO2(i, 0) \ + XO2(i + 1, 1) \ + XO2(i + 2, 2) \ + XO2(i + 3, 3) \ + XO3(i, 0) \ + ST(i, 0) \ + XO3(i + 1, 1) \ + ST(i + 1, 1) \ + XO3(i + 2, 2) \ + ST(i + 2, 2) \ + XO3(i + 3, 3) \ + ST(i + 3, 3) " .align 32 ;\n" " 1: ;\n" @@ -186,35 +186,35 @@ xor_pII_mmx_5(unsigned long bytes, unsigned long *p1, unsigned long *p2, because we modify p4 and p5 there, but we can't mark them as read/write, otherwise we'd overflow the 10-asm-operands limit of GCC < 3.1. */ - __asm__ ("" : "+r" (p4), "+r" (p5)); + asm("" : "+r" (p4), "+r" (p5)); - __asm__ __volatile__ ( + asm volatile( #undef BLOCK -#define BLOCK(i) \ - LD(i,0) \ - LD(i+1,1) \ - LD(i+2,2) \ - LD(i+3,3) \ - XO1(i,0) \ - XO1(i+1,1) \ - XO1(i+2,2) \ - XO1(i+3,3) \ - XO2(i,0) \ - XO2(i+1,1) \ - XO2(i+2,2) \ - XO2(i+3,3) \ - XO3(i,0) \ - XO3(i+1,1) \ - XO3(i+2,2) \ - XO3(i+3,3) \ - XO4(i,0) \ - ST(i,0) \ - XO4(i+1,1) \ - ST(i+1,1) \ - XO4(i+2,2) \ - ST(i+2,2) \ - XO4(i+3,3) \ - ST(i+3,3) +#define BLOCK(i) \ + LD(i, 0) \ + LD(i + 1, 1) \ + LD(i + 2, 2) \ + LD(i + 3, 3) \ + XO1(i, 0) \ + XO1(i + 1, 1) \ + XO1(i + 2, 2) \ + XO1(i + 3, 3) \ + XO2(i, 0) \ + XO2(i + 1, 1) \ + XO2(i + 2, 2) \ + XO2(i + 3, 3) \ + XO3(i, 0) \ + XO3(i + 1, 1) \ + XO3(i + 2, 2) \ + XO3(i + 3, 3) \ + XO4(i, 0) \ + ST(i, 0) \ + XO4(i + 1, 1) \ + ST(i + 1, 1) \ + XO4(i + 2, 2) \ + ST(i + 2, 2) \ + XO4(i + 3, 3) \ + ST(i + 3, 3) " .align 32 ;\n" " 1: ;\n" @@ -233,13 +233,13 @@ xor_pII_mmx_5(unsigned long bytes, unsigned long *p1, unsigned long *p2, " jnz 1b ;\n" : "+r" (lines), "+r" (p1), "+r" (p2), "+r" (p3) - : "r" (p4), "r" (p5) + : "r" (p4), "r" (p5) : "memory"); /* p4 and p5 were modified, and now the variables are dead. Clobber them just to be sure nobody does something stupid like assuming they have some legal value. */ - __asm__ ("" : "=r" (p4), "=r" (p5)); + asm("" : "=r" (p4), "=r" (p5)); kernel_fpu_end(); } @@ -259,7 +259,7 @@ xor_p5_mmx_2(unsigned long bytes, unsigned long *p1, unsigned long *p2) kernel_fpu_begin(); - __asm__ __volatile__ ( + asm volatile( " .align 32 ;\n" " 1: ;\n" " movq (%1), %%mm0 ;\n" @@ -286,7 +286,7 @@ xor_p5_mmx_2(unsigned long bytes, unsigned long *p1, unsigned long *p2) " pxor 56(%2), %%mm7 ;\n" " movq %%mm6, 48(%1) ;\n" " movq %%mm7, 56(%1) ;\n" - + " addl $64, %1 ;\n" " addl $64, %2 ;\n" " decl %0 ;\n" @@ -307,7 +307,7 @@ xor_p5_mmx_3(unsigned long bytes, unsigned long *p1, unsigned long *p2, kernel_fpu_begin(); - __asm__ __volatile__ ( + asm volatile( " .align 32,0x90 ;\n" " 1: ;\n" " movq (%1), %%mm0 ;\n" @@ -342,7 +342,7 @@ xor_p5_mmx_3(unsigned long bytes, unsigned long *p1, unsigned long *p2, " pxor 56(%3), %%mm7 ;\n" " movq %%mm6, 48(%1) ;\n" " movq %%mm7, 56(%1) ;\n" - + " addl $64, %1 ;\n" " addl $64, %2 ;\n" " addl $64, %3 ;\n" @@ -364,7 +364,7 @@ xor_p5_mmx_4(unsigned long bytes, unsigned long *p1, unsigned long *p2, kernel_fpu_begin(); - __asm__ __volatile__ ( + asm volatile( " .align 32,0x90 ;\n" " 1: ;\n" " movq (%1), %%mm0 ;\n" @@ -407,7 +407,7 @@ xor_p5_mmx_4(unsigned long bytes, unsigned long *p1, unsigned long *p2, " pxor 56(%4), %%mm7 ;\n" " movq %%mm6, 48(%1) ;\n" " movq %%mm7, 56(%1) ;\n" - + " addl $64, %1 ;\n" " addl $64, %2 ;\n" " addl $64, %3 ;\n" @@ -436,9 +436,9 @@ xor_p5_mmx_5(unsigned long bytes, unsigned long *p1, unsigned long *p2, because we modify p4 and p5 there, but we can't mark them as read/write, otherwise we'd overflow the 10-asm-operands limit of GCC < 3.1. */ - __asm__ ("" : "+r" (p4), "+r" (p5)); + asm("" : "+r" (p4), "+r" (p5)); - __asm__ __volatile__ ( + asm volatile( " .align 32,0x90 ;\n" " 1: ;\n" " movq (%1), %%mm0 ;\n" @@ -489,7 +489,7 @@ xor_p5_mmx_5(unsigned long bytes, unsigned long *p1, unsigned long *p2, " pxor 56(%5), %%mm7 ;\n" " movq %%mm6, 48(%1) ;\n" " movq %%mm7, 56(%1) ;\n" - + " addl $64, %1 ;\n" " addl $64, %2 ;\n" " addl $64, %3 ;\n" @@ -505,7 +505,7 @@ xor_p5_mmx_5(unsigned long bytes, unsigned long *p1, unsigned long *p2, /* p4 and p5 were modified, and now the variables are dead. Clobber them just to be sure nobody does something stupid like assuming they have some legal value. */ - __asm__ ("" : "=r" (p4), "=r" (p5)); + asm("" : "=r" (p4), "=r" (p5)); kernel_fpu_end(); } @@ -531,11 +531,12 @@ static struct xor_block_template xor_block_p5_mmx = { * Copyright (C) 1999 Zach Brown (with obvious credit due Ingo) */ -#define XMMS_SAVE do { \ +#define XMMS_SAVE \ +do { \ preempt_disable(); \ cr0 = read_cr0(); \ clts(); \ - __asm__ __volatile__ ( \ + asm volatile( \ "movups %%xmm0,(%0) ;\n\t" \ "movups %%xmm1,0x10(%0) ;\n\t" \ "movups %%xmm2,0x20(%0) ;\n\t" \ @@ -543,10 +544,11 @@ static struct xor_block_template xor_block_p5_mmx = { : \ : "r" (xmm_save) \ : "memory"); \ -} while(0) +} while (0) -#define XMMS_RESTORE do { \ - __asm__ __volatile__ ( \ +#define XMMS_RESTORE \ +do { \ + asm volatile( \ "sfence ;\n\t" \ "movups (%0),%%xmm0 ;\n\t" \ "movups 0x10(%0),%%xmm1 ;\n\t" \ @@ -557,76 +559,76 @@ static struct xor_block_template xor_block_p5_mmx = { : "memory"); \ write_cr0(cr0); \ preempt_enable(); \ -} while(0) +} while (0) #define ALIGN16 __attribute__((aligned(16))) #define OFFS(x) "16*("#x")" #define PF_OFFS(x) "256+16*("#x")" #define PF0(x) " prefetchnta "PF_OFFS(x)"(%1) ;\n" -#define LD(x,y) " movaps "OFFS(x)"(%1), %%xmm"#y" ;\n" -#define ST(x,y) " movaps %%xmm"#y", "OFFS(x)"(%1) ;\n" +#define LD(x, y) " movaps "OFFS(x)"(%1), %%xmm"#y" ;\n" +#define ST(x, y) " movaps %%xmm"#y", "OFFS(x)"(%1) ;\n" #define PF1(x) " prefetchnta "PF_OFFS(x)"(%2) ;\n" #define PF2(x) " prefetchnta "PF_OFFS(x)"(%3) ;\n" #define PF3(x) " prefetchnta "PF_OFFS(x)"(%4) ;\n" #define PF4(x) " prefetchnta "PF_OFFS(x)"(%5) ;\n" #define PF5(x) " prefetchnta "PF_OFFS(x)"(%6) ;\n" -#define XO1(x,y) " xorps "OFFS(x)"(%2), %%xmm"#y" ;\n" -#define XO2(x,y) " xorps "OFFS(x)"(%3), %%xmm"#y" ;\n" -#define XO3(x,y) " xorps "OFFS(x)"(%4), %%xmm"#y" ;\n" -#define XO4(x,y) " xorps "OFFS(x)"(%5), %%xmm"#y" ;\n" -#define XO5(x,y) " xorps "OFFS(x)"(%6), %%xmm"#y" ;\n" +#define XO1(x, y) " xorps "OFFS(x)"(%2), %%xmm"#y" ;\n" +#define XO2(x, y) " xorps "OFFS(x)"(%3), %%xmm"#y" ;\n" +#define XO3(x, y) " xorps "OFFS(x)"(%4), %%xmm"#y" ;\n" +#define XO4(x, y) " xorps "OFFS(x)"(%5), %%xmm"#y" ;\n" +#define XO5(x, y) " xorps "OFFS(x)"(%6), %%xmm"#y" ;\n" static void xor_sse_2(unsigned long bytes, unsigned long *p1, unsigned long *p2) { - unsigned long lines = bytes >> 8; + unsigned long lines = bytes >> 8; char xmm_save[16*4] ALIGN16; int cr0; XMMS_SAVE; - __asm__ __volatile__ ( + asm volatile( #undef BLOCK -#define BLOCK(i) \ - LD(i,0) \ - LD(i+1,1) \ +#define BLOCK(i) \ + LD(i, 0) \ + LD(i + 1, 1) \ PF1(i) \ - PF1(i+2) \ - LD(i+2,2) \ - LD(i+3,3) \ - PF0(i+4) \ - PF0(i+6) \ - XO1(i,0) \ - XO1(i+1,1) \ - XO1(i+2,2) \ - XO1(i+3,3) \ - ST(i,0) \ - ST(i+1,1) \ - ST(i+2,2) \ - ST(i+3,3) \ + PF1(i + 2) \ + LD(i + 2, 2) \ + LD(i + 3, 3) \ + PF0(i + 4) \ + PF0(i + 6) \ + XO1(i, 0) \ + XO1(i + 1, 1) \ + XO1(i + 2, 2) \ + XO1(i + 3, 3) \ + ST(i, 0) \ + ST(i + 1, 1) \ + ST(i + 2, 2) \ + ST(i + 3, 3) \ PF0(0) PF0(2) " .align 32 ;\n" - " 1: ;\n" + " 1: ;\n" BLOCK(0) BLOCK(4) BLOCK(8) BLOCK(12) - " addl $256, %1 ;\n" - " addl $256, %2 ;\n" - " decl %0 ;\n" - " jnz 1b ;\n" + " addl $256, %1 ;\n" + " addl $256, %2 ;\n" + " decl %0 ;\n" + " jnz 1b ;\n" : "+r" (lines), "+r" (p1), "+r" (p2) : - : "memory"); + : "memory"); XMMS_RESTORE; } @@ -635,59 +637,59 @@ static void xor_sse_3(unsigned long bytes, unsigned long *p1, unsigned long *p2, unsigned long *p3) { - unsigned long lines = bytes >> 8; + unsigned long lines = bytes >> 8; char xmm_save[16*4] ALIGN16; int cr0; XMMS_SAVE; - __asm__ __volatile__ ( + asm volatile( #undef BLOCK #define BLOCK(i) \ PF1(i) \ - PF1(i+2) \ + PF1(i + 2) \ LD(i,0) \ - LD(i+1,1) \ - LD(i+2,2) \ - LD(i+3,3) \ + LD(i + 1, 1) \ + LD(i + 2, 2) \ + LD(i + 3, 3) \ PF2(i) \ - PF2(i+2) \ - PF0(i+4) \ - PF0(i+6) \ + PF2(i + 2) \ + PF0(i + 4) \ + PF0(i + 6) \ XO1(i,0) \ - XO1(i+1,1) \ - XO1(i+2,2) \ - XO1(i+3,3) \ + XO1(i + 1, 1) \ + XO1(i + 2, 2) \ + XO1(i + 3, 3) \ XO2(i,0) \ - XO2(i+1,1) \ - XO2(i+2,2) \ - XO2(i+3,3) \ + XO2(i + 1, 1) \ + XO2(i + 2, 2) \ + XO2(i + 3, 3) \ ST(i,0) \ - ST(i+1,1) \ - ST(i+2,2) \ - ST(i+3,3) \ + ST(i + 1, 1) \ + ST(i + 2, 2) \ + ST(i + 3, 3) \ PF0(0) PF0(2) " .align 32 ;\n" - " 1: ;\n" + " 1: ;\n" BLOCK(0) BLOCK(4) BLOCK(8) BLOCK(12) - " addl $256, %1 ;\n" - " addl $256, %2 ;\n" - " addl $256, %3 ;\n" - " decl %0 ;\n" - " jnz 1b ;\n" + " addl $256, %1 ;\n" + " addl $256, %2 ;\n" + " addl $256, %3 ;\n" + " decl %0 ;\n" + " jnz 1b ;\n" : "+r" (lines), "+r" (p1), "+r"(p2), "+r"(p3) : - : "memory" ); + : "memory" ); XMMS_RESTORE; } @@ -696,66 +698,66 @@ static void xor_sse_4(unsigned long bytes, unsigned long *p1, unsigned long *p2, unsigned long *p3, unsigned long *p4) { - unsigned long lines = bytes >> 8; + unsigned long lines = bytes >> 8; char xmm_save[16*4] ALIGN16; int cr0; XMMS_SAVE; - __asm__ __volatile__ ( + asm volatile( #undef BLOCK #define BLOCK(i) \ PF1(i) \ - PF1(i+2) \ + PF1(i + 2) \ LD(i,0) \ - LD(i+1,1) \ - LD(i+2,2) \ - LD(i+3,3) \ + LD(i + 1, 1) \ + LD(i + 2, 2) \ + LD(i + 3, 3) \ PF2(i) \ - PF2(i+2) \ + PF2(i + 2) \ XO1(i,0) \ - XO1(i+1,1) \ - XO1(i+2,2) \ - XO1(i+3,3) \ + XO1(i + 1, 1) \ + XO1(i + 2, 2) \ + XO1(i + 3, 3) \ PF3(i) \ - PF3(i+2) \ - PF0(i+4) \ - PF0(i+6) \ + PF3(i + 2) \ + PF0(i + 4) \ + PF0(i + 6) \ XO2(i,0) \ - XO2(i+1,1) \ - XO2(i+2,2) \ - XO2(i+3,3) \ + XO2(i + 1, 1) \ + XO2(i + 2, 2) \ + XO2(i + 3, 3) \ XO3(i,0) \ - XO3(i+1,1) \ - XO3(i+2,2) \ - XO3(i+3,3) \ + XO3(i + 1, 1) \ + XO3(i + 2, 2) \ + XO3(i + 3, 3) \ ST(i,0) \ - ST(i+1,1) \ - ST(i+2,2) \ - ST(i+3,3) \ + ST(i + 1, 1) \ + ST(i + 2, 2) \ + ST(i + 3, 3) \ PF0(0) PF0(2) " .align 32 ;\n" - " 1: ;\n" + " 1: ;\n" BLOCK(0) BLOCK(4) BLOCK(8) BLOCK(12) - " addl $256, %1 ;\n" - " addl $256, %2 ;\n" - " addl $256, %3 ;\n" - " addl $256, %4 ;\n" - " decl %0 ;\n" - " jnz 1b ;\n" + " addl $256, %1 ;\n" + " addl $256, %2 ;\n" + " addl $256, %3 ;\n" + " addl $256, %4 ;\n" + " decl %0 ;\n" + " jnz 1b ;\n" : "+r" (lines), "+r" (p1), "+r" (p2), "+r" (p3), "+r" (p4) : - : "memory" ); + : "memory" ); XMMS_RESTORE; } @@ -764,7 +766,7 @@ static void xor_sse_5(unsigned long bytes, unsigned long *p1, unsigned long *p2, unsigned long *p3, unsigned long *p4, unsigned long *p5) { - unsigned long lines = bytes >> 8; + unsigned long lines = bytes >> 8; char xmm_save[16*4] ALIGN16; int cr0; @@ -776,65 +778,65 @@ xor_sse_5(unsigned long bytes, unsigned long *p1, unsigned long *p2, because we modify p4 and p5 there, but we can't mark them as read/write, otherwise we'd overflow the 10-asm-operands limit of GCC < 3.1. */ - __asm__ ("" : "+r" (p4), "+r" (p5)); + asm("" : "+r" (p4), "+r" (p5)); - __asm__ __volatile__ ( + asm volatile( #undef BLOCK #define BLOCK(i) \ PF1(i) \ - PF1(i+2) \ + PF1(i + 2) \ LD(i,0) \ - LD(i+1,1) \ - LD(i+2,2) \ - LD(i+3,3) \ + LD(i + 1, 1) \ + LD(i + 2, 2) \ + LD(i + 3, 3) \ PF2(i) \ - PF2(i+2) \ + PF2(i + 2) \ XO1(i,0) \ - XO1(i+1,1) \ - XO1(i+2,2) \ - XO1(i+3,3) \ + XO1(i + 1, 1) \ + XO1(i + 2, 2) \ + XO1(i + 3, 3) \ PF3(i) \ - PF3(i+2) \ + PF3(i + 2) \ XO2(i,0) \ - XO2(i+1,1) \ - XO2(i+2,2) \ - XO2(i+3,3) \ + XO2(i + 1, 1) \ + XO2(i + 2, 2) \ + XO2(i + 3, 3) \ PF4(i) \ - PF4(i+2) \ - PF0(i+4) \ - PF0(i+6) \ + PF4(i + 2) \ + PF0(i + 4) \ + PF0(i + 6) \ XO3(i,0) \ - XO3(i+1,1) \ - XO3(i+2,2) \ - XO3(i+3,3) \ + XO3(i + 1, 1) \ + XO3(i + 2, 2) \ + XO3(i + 3, 3) \ XO4(i,0) \ - XO4(i+1,1) \ - XO4(i+2,2) \ - XO4(i+3,3) \ + XO4(i + 1, 1) \ + XO4(i + 2, 2) \ + XO4(i + 3, 3) \ ST(i,0) \ - ST(i+1,1) \ - ST(i+2,2) \ - ST(i+3,3) \ + ST(i + 1, 1) \ + ST(i + 2, 2) \ + ST(i + 3, 3) \ PF0(0) PF0(2) " .align 32 ;\n" - " 1: ;\n" + " 1: ;\n" BLOCK(0) BLOCK(4) BLOCK(8) BLOCK(12) - " addl $256, %1 ;\n" - " addl $256, %2 ;\n" - " addl $256, %3 ;\n" - " addl $256, %4 ;\n" - " addl $256, %5 ;\n" - " decl %0 ;\n" - " jnz 1b ;\n" + " addl $256, %1 ;\n" + " addl $256, %2 ;\n" + " addl $256, %3 ;\n" + " addl $256, %4 ;\n" + " addl $256, %5 ;\n" + " decl %0 ;\n" + " jnz 1b ;\n" : "+r" (lines), "+r" (p1), "+r" (p2), "+r" (p3) : "r" (p4), "r" (p5) @@ -843,17 +845,17 @@ xor_sse_5(unsigned long bytes, unsigned long *p1, unsigned long *p2, /* p4 and p5 were modified, and now the variables are dead. Clobber them just to be sure nobody does something stupid like assuming they have some legal value. */ - __asm__ ("" : "=r" (p4), "=r" (p5)); + asm("" : "=r" (p4), "=r" (p5)); XMMS_RESTORE; } static struct xor_block_template xor_block_pIII_sse = { - .name = "pIII_sse", - .do_2 = xor_sse_2, - .do_3 = xor_sse_3, - .do_4 = xor_sse_4, - .do_5 = xor_sse_5, + .name = "pIII_sse", + .do_2 = xor_sse_2, + .do_3 = xor_sse_3, + .do_4 = xor_sse_4, + .do_5 = xor_sse_5, }; /* Also try the generic routines. */ @@ -861,21 +863,21 @@ static struct xor_block_template xor_block_pIII_sse = { #undef XOR_TRY_TEMPLATES #define XOR_TRY_TEMPLATES \ - do { \ - xor_speed(&xor_block_8regs); \ - xor_speed(&xor_block_8regs_p); \ - xor_speed(&xor_block_32regs); \ - xor_speed(&xor_block_32regs_p); \ - if (cpu_has_xmm) \ - xor_speed(&xor_block_pIII_sse); \ - if (cpu_has_mmx) { \ - xor_speed(&xor_block_pII_mmx); \ - xor_speed(&xor_block_p5_mmx); \ - } \ - } while (0) +do { \ + xor_speed(&xor_block_8regs); \ + xor_speed(&xor_block_8regs_p); \ + xor_speed(&xor_block_32regs); \ + xor_speed(&xor_block_32regs_p); \ + if (cpu_has_xmm) \ + xor_speed(&xor_block_pIII_sse); \ + if (cpu_has_mmx) { \ + xor_speed(&xor_block_pII_mmx); \ + xor_speed(&xor_block_p5_mmx); \ + } \ +} while (0) /* We force the use of the SSE xor block because it can write around L2. We may also be able to load into the L1 only depending on how the cpu deals with a load to a line that is being prefetched. */ -#define XOR_SELECT_TEMPLATE(FASTEST) \ +#define XOR_SELECT_TEMPLATE(FASTEST) \ (cpu_has_xmm ? &xor_block_pIII_sse : FASTEST) diff --git a/include/asm-x86/xor_64.h b/include/asm-x86/xor_64.h index 1eee7fcb2420..24957e39ac8a 100644 --- a/include/asm-x86/xor_64.h +++ b/include/asm-x86/xor_64.h @@ -24,20 +24,23 @@ */ /* - * x86-64 changes / gcc fixes from Andi Kleen. + * x86-64 changes / gcc fixes from Andi Kleen. * Copyright 2002 Andi Kleen, SuSE Labs. * * This hasn't been optimized for the hammer yet, but there are likely * no advantages to be gotten from x86-64 here anyways. */ -typedef struct { unsigned long a,b; } __attribute__((aligned(16))) xmm_store_t; +typedef struct { + unsigned long a, b; +} __attribute__((aligned(16))) xmm_store_t; -/* Doesn't use gcc to save the XMM registers, because there is no easy way to +/* Doesn't use gcc to save the XMM registers, because there is no easy way to tell it to do a clts before the register saving. */ -#define XMMS_SAVE do { \ +#define XMMS_SAVE \ +do { \ preempt_disable(); \ - asm volatile ( \ + asm volatile( \ "movq %%cr0,%0 ;\n\t" \ "clts ;\n\t" \ "movups %%xmm0,(%1) ;\n\t" \ @@ -47,10 +50,11 @@ typedef struct { unsigned long a,b; } __attribute__((aligned(16))) xmm_store_t; : "=&r" (cr0) \ : "r" (xmm_save) \ : "memory"); \ -} while(0) +} while (0) -#define XMMS_RESTORE do { \ - asm volatile ( \ +#define XMMS_RESTORE \ +do { \ + asm volatile( \ "sfence ;\n\t" \ "movups (%1),%%xmm0 ;\n\t" \ "movups 0x10(%1),%%xmm1 ;\n\t" \ @@ -61,72 +65,72 @@ typedef struct { unsigned long a,b; } __attribute__((aligned(16))) xmm_store_t; : "r" (cr0), "r" (xmm_save) \ : "memory"); \ preempt_enable(); \ -} while(0) +} while (0) #define OFFS(x) "16*("#x")" #define PF_OFFS(x) "256+16*("#x")" #define PF0(x) " prefetchnta "PF_OFFS(x)"(%[p1]) ;\n" -#define LD(x,y) " movaps "OFFS(x)"(%[p1]), %%xmm"#y" ;\n" -#define ST(x,y) " movaps %%xmm"#y", "OFFS(x)"(%[p1]) ;\n" +#define LD(x, y) " movaps "OFFS(x)"(%[p1]), %%xmm"#y" ;\n" +#define ST(x, y) " movaps %%xmm"#y", "OFFS(x)"(%[p1]) ;\n" #define PF1(x) " prefetchnta "PF_OFFS(x)"(%[p2]) ;\n" #define PF2(x) " prefetchnta "PF_OFFS(x)"(%[p3]) ;\n" #define PF3(x) " prefetchnta "PF_OFFS(x)"(%[p4]) ;\n" #define PF4(x) " prefetchnta "PF_OFFS(x)"(%[p5]) ;\n" #define PF5(x) " prefetchnta "PF_OFFS(x)"(%[p6]) ;\n" -#define XO1(x,y) " xorps "OFFS(x)"(%[p2]), %%xmm"#y" ;\n" -#define XO2(x,y) " xorps "OFFS(x)"(%[p3]), %%xmm"#y" ;\n" -#define XO3(x,y) " xorps "OFFS(x)"(%[p4]), %%xmm"#y" ;\n" -#define XO4(x,y) " xorps "OFFS(x)"(%[p5]), %%xmm"#y" ;\n" -#define XO5(x,y) " xorps "OFFS(x)"(%[p6]), %%xmm"#y" ;\n" +#define XO1(x, y) " xorps "OFFS(x)"(%[p2]), %%xmm"#y" ;\n" +#define XO2(x, y) " xorps "OFFS(x)"(%[p3]), %%xmm"#y" ;\n" +#define XO3(x, y) " xorps "OFFS(x)"(%[p4]), %%xmm"#y" ;\n" +#define XO4(x, y) " xorps "OFFS(x)"(%[p5]), %%xmm"#y" ;\n" +#define XO5(x, y) " xorps "OFFS(x)"(%[p6]), %%xmm"#y" ;\n" static void xor_sse_2(unsigned long bytes, unsigned long *p1, unsigned long *p2) { - unsigned int lines = bytes >> 8; + unsigned int lines = bytes >> 8; unsigned long cr0; xmm_store_t xmm_save[4]; XMMS_SAVE; - asm volatile ( + asm volatile( #undef BLOCK #define BLOCK(i) \ - LD(i,0) \ - LD(i+1,1) \ + LD(i, 0) \ + LD(i + 1, 1) \ PF1(i) \ - PF1(i+2) \ - LD(i+2,2) \ - LD(i+3,3) \ - PF0(i+4) \ - PF0(i+6) \ - XO1(i,0) \ - XO1(i+1,1) \ - XO1(i+2,2) \ - XO1(i+3,3) \ - ST(i,0) \ - ST(i+1,1) \ - ST(i+2,2) \ - ST(i+3,3) \ + PF1(i + 2) \ + LD(i + 2, 2) \ + LD(i + 3, 3) \ + PF0(i + 4) \ + PF0(i + 6) \ + XO1(i, 0) \ + XO1(i + 1, 1) \ + XO1(i + 2, 2) \ + XO1(i + 3, 3) \ + ST(i, 0) \ + ST(i + 1, 1) \ + ST(i + 2, 2) \ + ST(i + 3, 3) \ PF0(0) PF0(2) " .align 32 ;\n" - " 1: ;\n" + " 1: ;\n" BLOCK(0) BLOCK(4) BLOCK(8) BLOCK(12) - " addq %[inc], %[p1] ;\n" - " addq %[inc], %[p2] ;\n" + " addq %[inc], %[p1] ;\n" + " addq %[inc], %[p2] ;\n" " decl %[cnt] ; jnz 1b" : [p1] "+r" (p1), [p2] "+r" (p2), [cnt] "+r" (lines) - : [inc] "r" (256UL) - : "memory"); + : [inc] "r" (256UL) + : "memory"); XMMS_RESTORE; } @@ -141,52 +145,52 @@ xor_sse_3(unsigned long bytes, unsigned long *p1, unsigned long *p2, XMMS_SAVE; - __asm__ __volatile__ ( + asm volatile( #undef BLOCK #define BLOCK(i) \ PF1(i) \ - PF1(i+2) \ - LD(i,0) \ - LD(i+1,1) \ - LD(i+2,2) \ - LD(i+3,3) \ + PF1(i + 2) \ + LD(i, 0) \ + LD(i + 1, 1) \ + LD(i + 2, 2) \ + LD(i + 3, 3) \ PF2(i) \ - PF2(i+2) \ - PF0(i+4) \ - PF0(i+6) \ - XO1(i,0) \ - XO1(i+1,1) \ - XO1(i+2,2) \ - XO1(i+3,3) \ - XO2(i,0) \ - XO2(i+1,1) \ - XO2(i+2,2) \ - XO2(i+3,3) \ - ST(i,0) \ - ST(i+1,1) \ - ST(i+2,2) \ - ST(i+3,3) \ + PF2(i + 2) \ + PF0(i + 4) \ + PF0(i + 6) \ + XO1(i, 0) \ + XO1(i + 1, 1) \ + XO1(i + 2, 2) \ + XO1(i + 3, 3) \ + XO2(i, 0) \ + XO2(i + 1, 1) \ + XO2(i + 2, 2) \ + XO2(i + 3, 3) \ + ST(i, 0) \ + ST(i + 1, 1) \ + ST(i + 2, 2) \ + ST(i + 3, 3) \ PF0(0) PF0(2) " .align 32 ;\n" - " 1: ;\n" + " 1: ;\n" BLOCK(0) BLOCK(4) BLOCK(8) BLOCK(12) - " addq %[inc], %[p1] ;\n" - " addq %[inc], %[p2] ;\n" - " addq %[inc], %[p3] ;\n" + " addq %[inc], %[p1] ;\n" + " addq %[inc], %[p2] ;\n" + " addq %[inc], %[p3] ;\n" " decl %[cnt] ; jnz 1b" : [cnt] "+r" (lines), [p1] "+r" (p1), [p2] "+r" (p2), [p3] "+r" (p3) : [inc] "r" (256UL) - : "memory"); + : "memory"); XMMS_RESTORE; } @@ -195,64 +199,64 @@ xor_sse_4(unsigned long bytes, unsigned long *p1, unsigned long *p2, unsigned long *p3, unsigned long *p4) { unsigned int lines = bytes >> 8; - xmm_store_t xmm_save[4]; + xmm_store_t xmm_save[4]; unsigned long cr0; XMMS_SAVE; - __asm__ __volatile__ ( + asm volatile( #undef BLOCK #define BLOCK(i) \ PF1(i) \ - PF1(i+2) \ - LD(i,0) \ - LD(i+1,1) \ - LD(i+2,2) \ - LD(i+3,3) \ + PF1(i + 2) \ + LD(i, 0) \ + LD(i + 1, 1) \ + LD(i + 2, 2) \ + LD(i + 3, 3) \ PF2(i) \ - PF2(i+2) \ - XO1(i,0) \ - XO1(i+1,1) \ - XO1(i+2,2) \ - XO1(i+3,3) \ + PF2(i + 2) \ + XO1(i, 0) \ + XO1(i + 1, 1) \ + XO1(i + 2, 2) \ + XO1(i + 3, 3) \ PF3(i) \ - PF3(i+2) \ - PF0(i+4) \ - PF0(i+6) \ - XO2(i,0) \ - XO2(i+1,1) \ - XO2(i+2,2) \ - XO2(i+3,3) \ - XO3(i,0) \ - XO3(i+1,1) \ - XO3(i+2,2) \ - XO3(i+3,3) \ - ST(i,0) \ - ST(i+1,1) \ - ST(i+2,2) \ - ST(i+3,3) \ + PF3(i + 2) \ + PF0(i + 4) \ + PF0(i + 6) \ + XO2(i, 0) \ + XO2(i + 1, 1) \ + XO2(i + 2, 2) \ + XO2(i + 3, 3) \ + XO3(i, 0) \ + XO3(i + 1, 1) \ + XO3(i + 2, 2) \ + XO3(i + 3, 3) \ + ST(i, 0) \ + ST(i + 1, 1) \ + ST(i + 2, 2) \ + ST(i + 3, 3) \ PF0(0) PF0(2) " .align 32 ;\n" - " 1: ;\n" + " 1: ;\n" BLOCK(0) BLOCK(4) BLOCK(8) BLOCK(12) - " addq %[inc], %[p1] ;\n" - " addq %[inc], %[p2] ;\n" - " addq %[inc], %[p3] ;\n" - " addq %[inc], %[p4] ;\n" + " addq %[inc], %[p1] ;\n" + " addq %[inc], %[p2] ;\n" + " addq %[inc], %[p3] ;\n" + " addq %[inc], %[p4] ;\n" " decl %[cnt] ; jnz 1b" : [cnt] "+c" (lines), [p1] "+r" (p1), [p2] "+r" (p2), [p3] "+r" (p3), [p4] "+r" (p4) : [inc] "r" (256UL) - : "memory" ); + : "memory" ); XMMS_RESTORE; } @@ -261,70 +265,70 @@ static void xor_sse_5(unsigned long bytes, unsigned long *p1, unsigned long *p2, unsigned long *p3, unsigned long *p4, unsigned long *p5) { - unsigned int lines = bytes >> 8; + unsigned int lines = bytes >> 8; xmm_store_t xmm_save[4]; unsigned long cr0; XMMS_SAVE; - __asm__ __volatile__ ( + asm volatile( #undef BLOCK #define BLOCK(i) \ PF1(i) \ - PF1(i+2) \ - LD(i,0) \ - LD(i+1,1) \ - LD(i+2,2) \ - LD(i+3,3) \ + PF1(i + 2) \ + LD(i, 0) \ + LD(i + 1, 1) \ + LD(i + 2, 2) \ + LD(i + 3, 3) \ PF2(i) \ - PF2(i+2) \ - XO1(i,0) \ - XO1(i+1,1) \ - XO1(i+2,2) \ - XO1(i+3,3) \ + PF2(i + 2) \ + XO1(i, 0) \ + XO1(i + 1, 1) \ + XO1(i + 2, 2) \ + XO1(i + 3, 3) \ PF3(i) \ - PF3(i+2) \ - XO2(i,0) \ - XO2(i+1,1) \ - XO2(i+2,2) \ - XO2(i+3,3) \ + PF3(i + 2) \ + XO2(i, 0) \ + XO2(i + 1, 1) \ + XO2(i + 2, 2) \ + XO2(i + 3, 3) \ PF4(i) \ - PF4(i+2) \ - PF0(i+4) \ - PF0(i+6) \ - XO3(i,0) \ - XO3(i+1,1) \ - XO3(i+2,2) \ - XO3(i+3,3) \ - XO4(i,0) \ - XO4(i+1,1) \ - XO4(i+2,2) \ - XO4(i+3,3) \ - ST(i,0) \ - ST(i+1,1) \ - ST(i+2,2) \ - ST(i+3,3) \ + PF4(i + 2) \ + PF0(i + 4) \ + PF0(i + 6) \ + XO3(i, 0) \ + XO3(i + 1, 1) \ + XO3(i + 2, 2) \ + XO3(i + 3, 3) \ + XO4(i, 0) \ + XO4(i + 1, 1) \ + XO4(i + 2, 2) \ + XO4(i + 3, 3) \ + ST(i, 0) \ + ST(i + 1, 1) \ + ST(i + 2, 2) \ + ST(i + 3, 3) \ PF0(0) PF0(2) " .align 32 ;\n" - " 1: ;\n" + " 1: ;\n" BLOCK(0) BLOCK(4) BLOCK(8) BLOCK(12) - " addq %[inc], %[p1] ;\n" - " addq %[inc], %[p2] ;\n" - " addq %[inc], %[p3] ;\n" - " addq %[inc], %[p4] ;\n" - " addq %[inc], %[p5] ;\n" + " addq %[inc], %[p1] ;\n" + " addq %[inc], %[p2] ;\n" + " addq %[inc], %[p3] ;\n" + " addq %[inc], %[p4] ;\n" + " addq %[inc], %[p5] ;\n" " decl %[cnt] ; jnz 1b" : [cnt] "+c" (lines), - [p1] "+r" (p1), [p2] "+r" (p2), [p3] "+r" (p3), [p4] "+r" (p4), + [p1] "+r" (p1), [p2] "+r" (p2), [p3] "+r" (p3), [p4] "+r" (p4), [p5] "+r" (p5) : [inc] "r" (256UL) : "memory"); @@ -333,18 +337,18 @@ xor_sse_5(unsigned long bytes, unsigned long *p1, unsigned long *p2, } static struct xor_block_template xor_block_sse = { - .name = "generic_sse", - .do_2 = xor_sse_2, - .do_3 = xor_sse_3, - .do_4 = xor_sse_4, - .do_5 = xor_sse_5, + .name = "generic_sse", + .do_2 = xor_sse_2, + .do_3 = xor_sse_3, + .do_4 = xor_sse_4, + .do_5 = xor_sse_5, }; #undef XOR_TRY_TEMPLATES -#define XOR_TRY_TEMPLATES \ - do { \ - xor_speed(&xor_block_sse); \ - } while (0) +#define XOR_TRY_TEMPLATES \ +do { \ + xor_speed(&xor_block_sse); \ +} while (0) /* We force the use of the SSE xor block because it can write around L2. We may also be able to load into the L1 only depending on how the cpu diff --git a/include/asm-xtensa/semaphore.h b/include/asm-xtensa/semaphore.h index 3e04167cd9dc..d9b2034ed1d2 100644 --- a/include/asm-xtensa/semaphore.h +++ b/include/asm-xtensa/semaphore.h @@ -1,99 +1 @@ -/* - * linux/include/asm-xtensa/semaphore.h - * - * This file is subject to the terms and conditions of the GNU General Public - * License. See the file "COPYING" in the main directory of this archive - * for more details. - * - * Copyright (C) 2001 - 2005 Tensilica Inc. - */ - -#ifndef _XTENSA_SEMAPHORE_H -#define _XTENSA_SEMAPHORE_H - -#include <asm/atomic.h> -#include <asm/system.h> -#include <linux/wait.h> -#include <linux/rwsem.h> - -struct semaphore { - atomic_t count; - int sleepers; - wait_queue_head_t wait; -}; - -#define __SEMAPHORE_INITIALIZER(name,n) \ -{ \ - .count = ATOMIC_INIT(n), \ - .sleepers = 0, \ - .wait = __WAIT_QUEUE_HEAD_INITIALIZER((name).wait) \ -} - -#define __DECLARE_SEMAPHORE_GENERIC(name,count) \ - struct semaphore name = __SEMAPHORE_INITIALIZER(name,count) - -#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name,1) - -static inline void sema_init (struct semaphore *sem, int val) -{ - atomic_set(&sem->count, val); - sem->sleepers = 0; - init_waitqueue_head(&sem->wait); -} - -static inline void init_MUTEX (struct semaphore *sem) -{ - sema_init(sem, 1); -} - -static inline void init_MUTEX_LOCKED (struct semaphore *sem) -{ - sema_init(sem, 0); -} - -asmlinkage void __down(struct semaphore * sem); -asmlinkage int __down_interruptible(struct semaphore * sem); -asmlinkage int __down_trylock(struct semaphore * sem); -asmlinkage void __up(struct semaphore * sem); - -extern spinlock_t semaphore_wake_lock; - -static inline void down(struct semaphore * sem) -{ - might_sleep(); - - if (atomic_sub_return(1, &sem->count) < 0) - __down(sem); -} - -static inline int down_interruptible(struct semaphore * sem) -{ - int ret = 0; - - might_sleep(); - - if (atomic_sub_return(1, &sem->count) < 0) - ret = __down_interruptible(sem); - return ret; -} - -static inline int down_trylock(struct semaphore * sem) -{ - int ret = 0; - - if (atomic_sub_return(1, &sem->count) < 0) - ret = __down_trylock(sem); - return ret; -} - -/* - * Note! This is subtle. We jump to wake people up only if - * the semaphore was negative (== somebody was waiting on it). - */ -static inline void up(struct semaphore * sem) -{ - if (atomic_add_return(1, &sem->count) <= 0) - __up(sem); -} - -#endif /* _XTENSA_SEMAPHORE_H */ +#include <linux/semaphore.h> diff --git a/include/linux/interrupt.h b/include/linux/interrupt.h index f8ab4ce70564..b5fef13148bd 100644 --- a/include/linux/interrupt.h +++ b/include/linux/interrupt.h @@ -102,6 +102,25 @@ extern void disable_irq_nosync(unsigned int irq); extern void disable_irq(unsigned int irq); extern void enable_irq(unsigned int irq); +#if defined(CONFIG_SMP) && defined(CONFIG_GENERIC_HARDIRQS) + +extern int irq_set_affinity(unsigned int irq, cpumask_t cpumask); +extern int irq_can_set_affinity(unsigned int irq); + +#else /* CONFIG_SMP */ + +static inline int irq_set_affinity(unsigned int irq, cpumask_t cpumask) +{ + return -EINVAL; +} + +static inline int irq_can_set_affinity(unsigned int irq) +{ + return 0; +} + +#endif /* CONFIG_SMP && CONFIG_GENERIC_HARDIRQS */ + #ifdef CONFIG_GENERIC_HARDIRQS /* * Special lockdep variants of irq disabling/enabling. diff --git a/include/linux/irq.h b/include/linux/irq.h index 176e5e790a44..1883a85625dd 100644 --- a/include/linux/irq.h +++ b/include/linux/irq.h @@ -228,21 +228,11 @@ static inline void set_pending_irq(unsigned int irq, cpumask_t mask) #endif /* CONFIG_GENERIC_PENDING_IRQ */ -extern int irq_set_affinity(unsigned int irq, cpumask_t cpumask); -extern int irq_can_set_affinity(unsigned int irq); - #else /* CONFIG_SMP */ #define move_native_irq(x) #define move_masked_irq(x) -static inline int irq_set_affinity(unsigned int irq, cpumask_t cpumask) -{ - return -EINVAL; -} - -static inline int irq_can_set_affinity(unsigned int irq) { return 0; } - #endif /* CONFIG_SMP */ #ifdef CONFIG_IRQBALANCE diff --git a/include/linux/mlx4/cmd.h b/include/linux/mlx4/cmd.h index 7d1eaa97de13..77323a72dd3c 100644 --- a/include/linux/mlx4/cmd.h +++ b/include/linux/mlx4/cmd.h @@ -81,7 +81,7 @@ enum { MLX4_CMD_SW2HW_CQ = 0x16, MLX4_CMD_HW2SW_CQ = 0x17, MLX4_CMD_QUERY_CQ = 0x18, - MLX4_CMD_RESIZE_CQ = 0x2c, + MLX4_CMD_MODIFY_CQ = 0x2c, /* SRQ commands */ MLX4_CMD_SW2HW_SRQ = 0x35, diff --git a/include/linux/mlx4/cq.h b/include/linux/mlx4/cq.h index 0181e0a57cbf..071cf96cf01f 100644 --- a/include/linux/mlx4/cq.h +++ b/include/linux/mlx4/cq.h @@ -45,11 +45,11 @@ struct mlx4_cqe { u8 sl; u8 reserved1; __be16 rlid; - u32 reserved2; + __be32 ipoib_status; __be32 byte_cnt; __be16 wqe_index; __be16 checksum; - u8 reserved3[3]; + u8 reserved2[3]; u8 owner_sr_opcode; }; @@ -85,6 +85,16 @@ enum { MLX4_CQE_SYNDROME_REMOTE_ABORTED_ERR = 0x22, }; +enum { + MLX4_CQE_IPOIB_STATUS_IPV4 = 1 << 22, + MLX4_CQE_IPOIB_STATUS_IPV4F = 1 << 23, + MLX4_CQE_IPOIB_STATUS_IPV6 = 1 << 24, + MLX4_CQE_IPOIB_STATUS_IPV4OPT = 1 << 25, + MLX4_CQE_IPOIB_STATUS_TCP = 1 << 26, + MLX4_CQE_IPOIB_STATUS_UDP = 1 << 27, + MLX4_CQE_IPOIB_STATUS_IPOK = 1 << 28, +}; + static inline void mlx4_cq_arm(struct mlx4_cq *cq, u32 cmd, void __iomem *uar_page, spinlock_t *doorbell_lock) @@ -120,4 +130,9 @@ enum { MLX4_CQ_DB_REQ_NOT = 2 << 24 }; +int mlx4_cq_modify(struct mlx4_dev *dev, struct mlx4_cq *cq, + u16 count, u16 period); +int mlx4_cq_resize(struct mlx4_dev *dev, struct mlx4_cq *cq, + int entries, struct mlx4_mtt *mtt); + #endif /* MLX4_CQ_H */ diff --git a/include/linux/mlx4/device.h b/include/linux/mlx4/device.h index 6cdf813cd478..ff7df1a2222f 100644 --- a/include/linux/mlx4/device.h +++ b/include/linux/mlx4/device.h @@ -186,6 +186,7 @@ struct mlx4_caps { u32 flags; u16 stat_rate_support; u8 port_width_cap[MLX4_MAX_PORTS + 1]; + int max_gso_sz; }; struct mlx4_buf_list { diff --git a/include/linux/mlx4/driver.h b/include/linux/mlx4/driver.h index 1b835ca49df1..53c5fdb6eac4 100644 --- a/include/linux/mlx4/driver.h +++ b/include/linux/mlx4/driver.h @@ -48,8 +48,7 @@ struct mlx4_interface { void * (*add) (struct mlx4_dev *dev); void (*remove)(struct mlx4_dev *dev, void *context); void (*event) (struct mlx4_dev *dev, void *context, - enum mlx4_dev_event event, int subtype, - int port); + enum mlx4_dev_event event, int port); struct list_head list; }; diff --git a/include/linux/mlx4/qp.h b/include/linux/mlx4/qp.h index 09a2230923f2..a5e43febee4f 100644 --- a/include/linux/mlx4/qp.h +++ b/include/linux/mlx4/qp.h @@ -158,10 +158,12 @@ struct mlx4_qp_context { #define MLX4_FW_VER_WQE_CTRL_NEC mlx4_fw_ver(2, 2, 232) enum { - MLX4_WQE_CTRL_NEC = 1 << 29, - MLX4_WQE_CTRL_FENCE = 1 << 6, - MLX4_WQE_CTRL_CQ_UPDATE = 3 << 2, - MLX4_WQE_CTRL_SOLICITED = 1 << 1, + MLX4_WQE_CTRL_NEC = 1 << 29, + MLX4_WQE_CTRL_FENCE = 1 << 6, + MLX4_WQE_CTRL_CQ_UPDATE = 3 << 2, + MLX4_WQE_CTRL_SOLICITED = 1 << 1, + MLX4_WQE_CTRL_IP_CSUM = 1 << 4, + MLX4_WQE_CTRL_TCP_UDP_CSUM = 1 << 5, }; struct mlx4_wqe_ctrl_seg { @@ -217,6 +219,11 @@ struct mlx4_wqe_datagram_seg { __be32 reservd[2]; }; +struct mlx4_lso_seg { + __be32 mss_hdr_size; + __be32 header[0]; +}; + struct mlx4_wqe_bind_seg { __be32 flags1; __be32 flags2; diff --git a/include/linux/quota.h b/include/linux/quota.h index 6e0393a5b2ea..eb560d031acd 100644 --- a/include/linux/quota.h +++ b/include/linux/quota.h @@ -160,14 +160,18 @@ enum { #ifdef __KERNEL__ -#include <linux/spinlock.h> -#include <linux/rwsem.h> +#include <linux/list.h> #include <linux/mutex.h> +#include <linux/rwsem.h> +#include <linux/spinlock.h> +#include <linux/wait.h> #include <linux/dqblk_xfs.h> #include <linux/dqblk_v1.h> #include <linux/dqblk_v2.h> +#include <asm/atomic.h> + extern spinlock_t dq_data_lock; /* Maximal numbers of writes for quota operation (insert/delete/update) diff --git a/include/linux/semaphore.h b/include/linux/semaphore.h new file mode 100644 index 000000000000..9cae64b00d6b --- /dev/null +++ b/include/linux/semaphore.h @@ -0,0 +1,51 @@ +/* + * Copyright (c) 2008 Intel Corporation + * Author: Matthew Wilcox <willy@linux.intel.com> + * + * Distributed under the terms of the GNU GPL, version 2 + * + * Please see kernel/semaphore.c for documentation of these functions + */ +#ifndef __LINUX_SEMAPHORE_H +#define __LINUX_SEMAPHORE_H + +#include <linux/list.h> +#include <linux/spinlock.h> + +/* Please don't access any members of this structure directly */ +struct semaphore { + spinlock_t lock; + unsigned int count; + struct list_head wait_list; +}; + +#define __SEMAPHORE_INITIALIZER(name, n) \ +{ \ + .lock = __SPIN_LOCK_UNLOCKED((name).lock), \ + .count = n, \ + .wait_list = LIST_HEAD_INIT((name).wait_list), \ +} + +#define __DECLARE_SEMAPHORE_GENERIC(name, count) \ + struct semaphore name = __SEMAPHORE_INITIALIZER(name, count) + +#define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name, 1) + +static inline void sema_init(struct semaphore *sem, int val) +{ + static struct lock_class_key __key; + *sem = (struct semaphore) __SEMAPHORE_INITIALIZER(*sem, val); + lockdep_init_map(&sem->lock.dep_map, "semaphore->lock", &__key, 0); +} + +#define init_MUTEX(sem) sema_init(sem, 1) +#define init_MUTEX_LOCKED(sem) sema_init(sem, 0) + +extern void down(struct semaphore *sem); +extern int __must_check down_interruptible(struct semaphore *sem); +extern int __must_check down_killable(struct semaphore *sem); +extern int __must_check down_trylock(struct semaphore *sem); +extern int __must_check down_timeout(struct semaphore *sem, long jiffies); +extern void up(struct semaphore *sem); + +#endif /* __LINUX_SEMAPHORE_H */ diff --git a/include/linux/slub_def.h b/include/linux/slub_def.h index b00c1c73eb0a..79d59c937fac 100644 --- a/include/linux/slub_def.h +++ b/include/linux/slub_def.h @@ -45,9 +45,9 @@ struct kmem_cache_cpu { struct kmem_cache_node { spinlock_t list_lock; /* Protect partial list and nr_partial */ unsigned long nr_partial; - atomic_long_t nr_slabs; struct list_head partial; #ifdef CONFIG_SLUB_DEBUG + atomic_long_t nr_slabs; struct list_head full; #endif }; diff --git a/include/rdma/ib_user_verbs.h b/include/rdma/ib_user_verbs.h index 64a721fcbc1c..8d65bf0a625b 100644 --- a/include/rdma/ib_user_verbs.h +++ b/include/rdma/ib_user_verbs.h @@ -533,7 +533,10 @@ struct ib_uverbs_send_wr { __u32 num_sge; __u32 opcode; __u32 send_flags; - __u32 imm_data; + union { + __u32 imm_data; + __u32 invalidate_rkey; + } ex; union { struct { __u64 remote_addr; diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h index 701e7b40560a..95bf4bac44cb 100644 --- a/include/rdma/ib_verbs.h +++ b/include/rdma/ib_verbs.h @@ -94,7 +94,7 @@ enum ib_device_cap_flags { IB_DEVICE_SRQ_RESIZE = (1<<13), IB_DEVICE_N_NOTIFY_CQ = (1<<14), IB_DEVICE_ZERO_STAG = (1<<15), - IB_DEVICE_SEND_W_INV = (1<<16), + IB_DEVICE_RESERVED = (1<<16), /* old SEND_W_INV */ IB_DEVICE_MEM_WINDOW = (1<<17), /* * Devices should set IB_DEVICE_UD_IP_SUM if they support @@ -104,6 +104,8 @@ enum ib_device_cap_flags { * IPoIB driver may set NETIF_F_IP_CSUM for datagram mode. */ IB_DEVICE_UD_IP_CSUM = (1<<18), + IB_DEVICE_UD_TSO = (1<<19), + IB_DEVICE_SEND_W_INV = (1<<21), }; enum ib_atomic_cap { @@ -411,6 +413,7 @@ enum ib_wc_opcode { IB_WC_COMP_SWAP, IB_WC_FETCH_ADD, IB_WC_BIND_MW, + IB_WC_LSO, /* * Set value of IB_WC_RECV so consumers can test if a completion is a * receive by testing (opcode & IB_WC_RECV). @@ -495,6 +498,10 @@ enum ib_qp_type { IB_QPT_RAW_ETY }; +enum ib_qp_create_flags { + IB_QP_CREATE_IPOIB_UD_LSO = 1 << 0, +}; + struct ib_qp_init_attr { void (*event_handler)(struct ib_event *, void *); void *qp_context; @@ -504,6 +511,7 @@ struct ib_qp_init_attr { struct ib_qp_cap cap; enum ib_sig_type sq_sig_type; enum ib_qp_type qp_type; + enum ib_qp_create_flags create_flags; u8 port_num; /* special QP types only */ }; @@ -617,7 +625,9 @@ enum ib_wr_opcode { IB_WR_SEND_WITH_IMM, IB_WR_RDMA_READ, IB_WR_ATOMIC_CMP_AND_SWP, - IB_WR_ATOMIC_FETCH_AND_ADD + IB_WR_ATOMIC_FETCH_AND_ADD, + IB_WR_LSO, + IB_WR_SEND_WITH_INV, }; enum ib_send_flags { @@ -641,7 +651,10 @@ struct ib_send_wr { int num_sge; enum ib_wr_opcode opcode; int send_flags; - __be32 imm_data; + union { + __be32 imm_data; + u32 invalidate_rkey; + } ex; union { struct { u64 remote_addr; @@ -655,6 +668,9 @@ struct ib_send_wr { } atomic; struct { struct ib_ah *ah; + void *header; + int hlen; + int mss; u32 remote_qpn; u32 remote_qkey; u16 pkey_index; /* valid for GSI only */ @@ -730,7 +746,7 @@ struct ib_uobject { struct ib_ucontext *context; /* associated user context */ void *object; /* containing object */ struct list_head list; /* link to context's list */ - u32 id; /* index into kernel idr */ + int id; /* index into kernel idr */ struct kref ref; struct rw_semaphore mutex; /* protects .live */ int live; @@ -971,6 +987,8 @@ struct ib_device { int comp_vector, struct ib_ucontext *context, struct ib_udata *udata); + int (*modify_cq)(struct ib_cq *cq, u16 cq_count, + u16 cq_period); int (*destroy_cq)(struct ib_cq *cq); int (*resize_cq)(struct ib_cq *cq, int cqe, struct ib_udata *udata); @@ -1376,6 +1394,15 @@ struct ib_cq *ib_create_cq(struct ib_device *device, int ib_resize_cq(struct ib_cq *cq, int cqe); /** + * ib_modify_cq - Modifies moderation params of the CQ + * @cq: The CQ to modify. + * @cq_count: number of CQEs that will trigger an event + * @cq_period: max period of time in usec before triggering an event + * + */ +int ib_modify_cq(struct ib_cq *cq, u16 cq_count, u16 cq_period); + +/** * ib_destroy_cq - Destroys the specified CQ. * @cq: The CQ to destroy. */ diff --git a/init/Kconfig b/init/Kconfig index a97924bc5b8d..7fccf09bb95a 100644 --- a/init/Kconfig +++ b/init/Kconfig @@ -763,7 +763,7 @@ endmenu # General setup config SLABINFO bool depends on PROC_FS - depends on SLAB || SLUB + depends on SLAB || SLUB_DEBUG default y config RT_MUTEXES diff --git a/kernel/Makefile b/kernel/Makefile index 05c8003718ee..6c5f081132a4 100644 --- a/kernel/Makefile +++ b/kernel/Makefile @@ -8,7 +8,7 @@ obj-y = sched.o fork.o exec_domain.o panic.o printk.o profile.o \ signal.o sys.o kmod.o workqueue.o pid.o \ rcupdate.o extable.o params.o posix-timers.o \ kthread.o wait.o kfifo.o sys_ni.o posix-cpu-timers.o mutex.o \ - hrtimer.o rwsem.o nsproxy.o srcu.o \ + hrtimer.o rwsem.o nsproxy.o srcu.o semaphore.o \ notifier.o ksysfs.o pm_qos_params.o obj-$(CONFIG_SYSCTL) += sysctl_check.o diff --git a/kernel/cgroup.c b/kernel/cgroup.c index 2727f9238359..6d8de051382b 100644 --- a/kernel/cgroup.c +++ b/kernel/cgroup.c @@ -1722,7 +1722,12 @@ void cgroup_enable_task_cg_lists(void) use_task_css_set_links = 1; do_each_thread(g, p) { task_lock(p); - if (list_empty(&p->cg_list)) + /* + * We should check if the process is exiting, otherwise + * it will race with cgroup_exit() in that the list + * entry won't be deleted though the process has exited. + */ + if (!(p->flags & PF_EXITING) && list_empty(&p->cg_list)) list_add(&p->cg_list, &p->cgroups->tasks); task_unlock(p); } while_each_thread(g, p); diff --git a/kernel/semaphore.c b/kernel/semaphore.c new file mode 100644 index 000000000000..5c2942e768cd --- /dev/null +++ b/kernel/semaphore.c @@ -0,0 +1,264 @@ +/* + * Copyright (c) 2008 Intel Corporation + * Author: Matthew Wilcox <willy@linux.intel.com> + * + * Distributed under the terms of the GNU GPL, version 2 + * + * This file implements counting semaphores. + * A counting semaphore may be acquired 'n' times before sleeping. + * See mutex.c for single-acquisition sleeping locks which enforce + * rules which allow code to be debugged more easily. + */ + +/* + * Some notes on the implementation: + * + * The spinlock controls access to the other members of the semaphore. + * down_trylock() and up() can be called from interrupt context, so we + * have to disable interrupts when taking the lock. It turns out various + * parts of the kernel expect to be able to use down() on a semaphore in + * interrupt context when they know it will succeed, so we have to use + * irqsave variants for down(), down_interruptible() and down_killable() + * too. + * + * The ->count variable represents how many more tasks can acquire this + * semaphore. If it's zero, there may be tasks waiting on the wait_list. + */ + +#include <linux/compiler.h> +#include <linux/kernel.h> +#include <linux/module.h> +#include <linux/sched.h> +#include <linux/semaphore.h> +#include <linux/spinlock.h> + +static noinline void __down(struct semaphore *sem); +static noinline int __down_interruptible(struct semaphore *sem); +static noinline int __down_killable(struct semaphore *sem); +static noinline int __down_timeout(struct semaphore *sem, long jiffies); +static noinline void __up(struct semaphore *sem); + +/** + * down - acquire the semaphore + * @sem: the semaphore to be acquired + * + * Acquires the semaphore. If no more tasks are allowed to acquire the + * semaphore, calling this function will put the task to sleep until the + * semaphore is released. + * + * Use of this function is deprecated, please use down_interruptible() or + * down_killable() instead. + */ +void down(struct semaphore *sem) +{ + unsigned long flags; + + spin_lock_irqsave(&sem->lock, flags); + if (likely(sem->count > 0)) + sem->count--; + else + __down(sem); + spin_unlock_irqrestore(&sem->lock, flags); +} +EXPORT_SYMBOL(down); + +/** + * down_interruptible - acquire the semaphore unless interrupted + * @sem: the semaphore to be acquired + * + * Attempts to acquire the semaphore. If no more tasks are allowed to + * acquire the semaphore, calling this function will put the task to sleep. + * If the sleep is interrupted by a signal, this function will return -EINTR. + * If the semaphore is successfully acquired, this function returns 0. + */ +int down_interruptible(struct semaphore *sem) +{ + unsigned long flags; + int result = 0; + + spin_lock_irqsave(&sem->lock, flags); + if (likely(sem->count > 0)) + sem->count--; + else + result = __down_interruptible(sem); + spin_unlock_irqrestore(&sem->lock, flags); + + return result; +} +EXPORT_SYMBOL(down_interruptible); + +/** + * down_killable - acquire the semaphore unless killed + * @sem: the semaphore to be acquired + * + * Attempts to acquire the semaphore. If no more tasks are allowed to + * acquire the semaphore, calling this function will put the task to sleep. + * If the sleep is interrupted by a fatal signal, this function will return + * -EINTR. If the semaphore is successfully acquired, this function returns + * 0. + */ +int down_killable(struct semaphore *sem) +{ + unsigned long flags; + int result = 0; + + spin_lock_irqsave(&sem->lock, flags); + if (likely(sem->count > 0)) + sem->count--; + else + result = __down_killable(sem); + spin_unlock_irqrestore(&sem->lock, flags); + + return result; +} +EXPORT_SYMBOL(down_killable); + +/** + * down_trylock - try to acquire the semaphore, without waiting + * @sem: the semaphore to be acquired + * + * Try to acquire the semaphore atomically. Returns 0 if the mutex has + * been acquired successfully or 1 if it it cannot be acquired. + * + * NOTE: This return value is inverted from both spin_trylock and + * mutex_trylock! Be careful about this when converting code. + * + * Unlike mutex_trylock, this function can be used from interrupt context, + * and the semaphore can be released by any task or interrupt. + */ +int down_trylock(struct semaphore *sem) +{ + unsigned long flags; + int count; + + spin_lock_irqsave(&sem->lock, flags); + count = sem->count - 1; + if (likely(count >= 0)) + sem->count = count; + spin_unlock_irqrestore(&sem->lock, flags); + + return (count < 0); +} +EXPORT_SYMBOL(down_trylock); + +/** + * down_timeout - acquire the semaphore within a specified time + * @sem: the semaphore to be acquired + * @jiffies: how long to wait before failing + * + * Attempts to acquire the semaphore. If no more tasks are allowed to + * acquire the semaphore, calling this function will put the task to sleep. + * If the semaphore is not released within the specified number of jiffies, + * this function returns -ETIME. It returns 0 if the semaphore was acquired. + */ +int down_timeout(struct semaphore *sem, long jiffies) +{ + unsigned long flags; + int result = 0; + + spin_lock_irqsave(&sem->lock, flags); + if (likely(sem->count > 0)) + sem->count--; + else + result = __down_timeout(sem, jiffies); + spin_unlock_irqrestore(&sem->lock, flags); + + return result; +} +EXPORT_SYMBOL(down_timeout); + +/** + * up - release the semaphore + * @sem: the semaphore to release + * + * Release the semaphore. Unlike mutexes, up() may be called from any + * context and even by tasks which have never called down(). + */ +void up(struct semaphore *sem) +{ + unsigned long flags; + + spin_lock_irqsave(&sem->lock, flags); + if (likely(list_empty(&sem->wait_list))) + sem->count++; + else + __up(sem); + spin_unlock_irqrestore(&sem->lock, flags); +} +EXPORT_SYMBOL(up); + +/* Functions for the contended case */ + +struct semaphore_waiter { + struct list_head list; + struct task_struct *task; + int up; +}; + +/* + * Because this function is inlined, the 'state' parameter will be + * constant, and thus optimised away by the compiler. Likewise the + * 'timeout' parameter for the cases without timeouts. + */ +static inline int __sched __down_common(struct semaphore *sem, long state, + long timeout) +{ + struct task_struct *task = current; + struct semaphore_waiter waiter; + + list_add_tail(&waiter.list, &sem->wait_list); + waiter.task = task; + waiter.up = 0; + + for (;;) { + if (state == TASK_INTERRUPTIBLE && signal_pending(task)) + goto interrupted; + if (state == TASK_KILLABLE && fatal_signal_pending(task)) + goto interrupted; + if (timeout <= 0) + goto timed_out; + __set_task_state(task, state); + spin_unlock_irq(&sem->lock); + timeout = schedule_timeout(timeout); + spin_lock_irq(&sem->lock); + if (waiter.up) + return 0; + } + + timed_out: + list_del(&waiter.list); + return -ETIME; + + interrupted: + list_del(&waiter.list); + return -EINTR; +} + +static noinline void __sched __down(struct semaphore *sem) +{ + __down_common(sem, TASK_UNINTERRUPTIBLE, MAX_SCHEDULE_TIMEOUT); +} + +static noinline int __sched __down_interruptible(struct semaphore *sem) +{ + return __down_common(sem, TASK_INTERRUPTIBLE, MAX_SCHEDULE_TIMEOUT); +} + +static noinline int __sched __down_killable(struct semaphore *sem) +{ + return __down_common(sem, TASK_KILLABLE, MAX_SCHEDULE_TIMEOUT); +} + +static noinline int __sched __down_timeout(struct semaphore *sem, long jiffies) +{ + return __down_common(sem, TASK_UNINTERRUPTIBLE, jiffies); +} + +static noinline void __sched __up(struct semaphore *sem) +{ + struct semaphore_waiter *waiter = list_first_entry(&sem->wait_list, + struct semaphore_waiter, list); + list_del(&waiter->list); + waiter->up = 1; + wake_up_process(waiter->task); +} diff --git a/kernel/signal.c b/kernel/signal.c index 6af1210092c3..cc8303cd093d 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -1757,6 +1757,45 @@ static int do_signal_stop(int signr) return 1; } +static int ptrace_signal(int signr, siginfo_t *info, + struct pt_regs *regs, void *cookie) +{ + if (!(current->ptrace & PT_PTRACED)) + return signr; + + ptrace_signal_deliver(regs, cookie); + + /* Let the debugger run. */ + ptrace_stop(signr, 0, info); + + /* We're back. Did the debugger cancel the sig? */ + signr = current->exit_code; + if (signr == 0) + return signr; + + current->exit_code = 0; + + /* Update the siginfo structure if the signal has + changed. If the debugger wanted something + specific in the siginfo structure then it should + have updated *info via PTRACE_SETSIGINFO. */ + if (signr != info->si_signo) { + info->si_signo = signr; + info->si_errno = 0; + info->si_code = SI_USER; + info->si_pid = task_pid_vnr(current->parent); + info->si_uid = current->parent->uid; + } + + /* If the (new) signal is now blocked, requeue it. */ + if (sigismember(¤t->blocked, signr)) { + specific_send_sig_info(signr, info, current); + signr = 0; + } + + return signr; +} + int get_signal_to_deliver(siginfo_t *info, struct k_sigaction *return_ka, struct pt_regs *regs, void *cookie) { @@ -1785,36 +1824,10 @@ relock: if (!signr) break; /* will return 0 */ - if ((current->ptrace & PT_PTRACED) && signr != SIGKILL) { - ptrace_signal_deliver(regs, cookie); - - /* Let the debugger run. */ - ptrace_stop(signr, 0, info); - - /* We're back. Did the debugger cancel the sig? */ - signr = current->exit_code; - if (signr == 0) - continue; - - current->exit_code = 0; - - /* Update the siginfo structure if the signal has - changed. If the debugger wanted something - specific in the siginfo structure then it should - have updated *info via PTRACE_SETSIGINFO. */ - if (signr != info->si_signo) { - info->si_signo = signr; - info->si_errno = 0; - info->si_code = SI_USER; - info->si_pid = task_pid_vnr(current->parent); - info->si_uid = current->parent->uid; - } - - /* If the (new) signal is now blocked, requeue it. */ - if (sigismember(¤t->blocked, signr)) { - specific_send_sig_info(signr, info, current); + if (signr != SIGKILL) { + signr = ptrace_signal(signr, info, regs, cookie); + if (!signr) continue; - } } ka = ¤t->sighand->action[signr-1]; diff --git a/kernel/time/tick-broadcast.c b/kernel/time/tick-broadcast.c index e1bd50cbbf5d..fdfa0c745bb6 100644 --- a/kernel/time/tick-broadcast.c +++ b/kernel/time/tick-broadcast.c @@ -14,7 +14,7 @@ #include <linux/cpu.h> #include <linux/err.h> #include <linux/hrtimer.h> -#include <linux/irq.h> +#include <linux/interrupt.h> #include <linux/percpu.h> #include <linux/profile.h> #include <linux/sched.h> diff --git a/kernel/time/tick-common.c b/kernel/time/tick-common.c index 1bea399a9ef0..4f3886562b8c 100644 --- a/kernel/time/tick-common.c +++ b/kernel/time/tick-common.c @@ -14,12 +14,14 @@ #include <linux/cpu.h> #include <linux/err.h> #include <linux/hrtimer.h> -#include <linux/irq.h> +#include <linux/interrupt.h> #include <linux/percpu.h> #include <linux/profile.h> #include <linux/sched.h> #include <linux/tick.h> +#include <asm/irq_regs.h> + #include "tick-internal.h" /* diff --git a/kernel/time/tick-oneshot.c b/kernel/time/tick-oneshot.c index 0258d3115d54..450c04935b66 100644 --- a/kernel/time/tick-oneshot.c +++ b/kernel/time/tick-oneshot.c @@ -14,7 +14,7 @@ #include <linux/cpu.h> #include <linux/err.h> #include <linux/hrtimer.h> -#include <linux/irq.h> +#include <linux/interrupt.h> #include <linux/percpu.h> #include <linux/profile.h> #include <linux/sched.h> diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index e601d0e7ac5d..aaba784332e0 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -211,7 +211,7 @@ config SLUB_DEBUG_ON config SLUB_STATS default n bool "Enable SLUB performance statistics" - depends on SLUB + depends on SLUB && SLUB_DEBUG && SYSFS help SLUB statistics are useful to debug SLUBs allocation behavior in order find ways to optimize the allocator. This should never be @@ -265,16 +265,6 @@ config DEBUG_MUTEXES This feature allows mutex semantics violations to be detected and reported. -config DEBUG_SEMAPHORE - bool "Semaphore debugging" - depends on DEBUG_KERNEL - depends on ALPHA || FRV - default n - help - If you say Y here then semaphore processing will issue lots of - verbose debugging messages. If you suspect a semaphore problem or a - kernel hacker asks for this option then say Y. Otherwise say N. - config DEBUG_LOCK_ALLOC bool "Lock debugging: detect incorrect freeing of live locks" depends on DEBUG_KERNEL && TRACE_IRQFLAGS_SUPPORT && STACKTRACE_SUPPORT && LOCKDEP_SUPPORT diff --git a/lib/Makefile b/lib/Makefile index 23de261a4c83..28dba90d5020 100644 --- a/lib/Makefile +++ b/lib/Makefile @@ -29,7 +29,6 @@ obj-$(CONFIG_DEBUG_LOCKING_API_SELFTESTS) += locking-selftest.o obj-$(CONFIG_DEBUG_SPINLOCK) += spinlock_debug.o lib-$(CONFIG_RWSEM_GENERIC_SPINLOCK) += rwsem-spinlock.o lib-$(CONFIG_RWSEM_XCHGADD_ALGORITHM) += rwsem.o -lib-$(CONFIG_SEMAPHORE_SLEEPERS) += semaphore-sleepers.o lib-$(CONFIG_GENERIC_FIND_NEXT_BIT) += find_next_bit.o obj-$(CONFIG_GENERIC_HWEIGHT) += hweight.o obj-$(CONFIG_LOCK_KERNEL) += kernel_lock.o diff --git a/lib/kernel_lock.c b/lib/kernel_lock.c index 812dbf00844b..fbc11a336bc5 100644 --- a/lib/kernel_lock.c +++ b/lib/kernel_lock.c @@ -8,6 +8,7 @@ #include <linux/smp_lock.h> #include <linux/module.h> #include <linux/kallsyms.h> +#include <asm/semaphore.h> /* * The 'big kernel semaphore' diff --git a/lib/semaphore-sleepers.c b/lib/semaphore-sleepers.c deleted file mode 100644 index 0198782cdacb..000000000000 --- a/lib/semaphore-sleepers.c +++ /dev/null @@ -1,176 +0,0 @@ -/* - * i386 and x86-64 semaphore implementation. - * - * (C) Copyright 1999 Linus Torvalds - * - * Portions Copyright 1999 Red Hat, Inc. - * - * This program is free software; you can redistribute it and/or - * modify it under the terms of the GNU General Public License - * as published by the Free Software Foundation; either version - * 2 of the License, or (at your option) any later version. - * - * rw semaphores implemented November 1999 by Benjamin LaHaise <bcrl@kvack.org> - */ -#include <linux/sched.h> -#include <linux/err.h> -#include <linux/init.h> -#include <asm/semaphore.h> - -/* - * Semaphores are implemented using a two-way counter: - * The "count" variable is decremented for each process - * that tries to acquire the semaphore, while the "sleeping" - * variable is a count of such acquires. - * - * Notably, the inline "up()" and "down()" functions can - * efficiently test if they need to do any extra work (up - * needs to do something only if count was negative before - * the increment operation. - * - * "sleeping" and the contention routine ordering is protected - * by the spinlock in the semaphore's waitqueue head. - * - * Note that these functions are only called when there is - * contention on the lock, and as such all this is the - * "non-critical" part of the whole semaphore business. The - * critical part is the inline stuff in <asm/semaphore.h> - * where we want to avoid any extra jumps and calls. - */ - -/* - * Logic: - * - only on a boundary condition do we need to care. When we go - * from a negative count to a non-negative, we wake people up. - * - when we go from a non-negative count to a negative do we - * (a) synchronize with the "sleeper" count and (b) make sure - * that we're on the wakeup list before we synchronize so that - * we cannot lose wakeup events. - */ - -void __up(struct semaphore *sem) -{ - wake_up(&sem->wait); -} - -void __sched __down(struct semaphore *sem) -{ - struct task_struct *tsk = current; - DECLARE_WAITQUEUE(wait, tsk); - unsigned long flags; - - tsk->state = TASK_UNINTERRUPTIBLE; - spin_lock_irqsave(&sem->wait.lock, flags); - add_wait_queue_exclusive_locked(&sem->wait, &wait); - - sem->sleepers++; - for (;;) { - int sleepers = sem->sleepers; - - /* - * Add "everybody else" into it. They aren't - * playing, because we own the spinlock in - * the wait_queue_head. - */ - if (!atomic_add_negative(sleepers - 1, &sem->count)) { - sem->sleepers = 0; - break; - } - sem->sleepers = 1; /* us - see -1 above */ - spin_unlock_irqrestore(&sem->wait.lock, flags); - - schedule(); - - spin_lock_irqsave(&sem->wait.lock, flags); - tsk->state = TASK_UNINTERRUPTIBLE; - } - remove_wait_queue_locked(&sem->wait, &wait); - wake_up_locked(&sem->wait); - spin_unlock_irqrestore(&sem->wait.lock, flags); - tsk->state = TASK_RUNNING; -} - -int __sched __down_interruptible(struct semaphore *sem) -{ - int retval = 0; - struct task_struct *tsk = current; - DECLARE_WAITQUEUE(wait, tsk); - unsigned long flags; - - tsk->state = TASK_INTERRUPTIBLE; - spin_lock_irqsave(&sem->wait.lock, flags); - add_wait_queue_exclusive_locked(&sem->wait, &wait); - - sem->sleepers++; - for (;;) { - int sleepers = sem->sleepers; - - /* - * With signals pending, this turns into - * the trylock failure case - we won't be - * sleeping, and we* can't get the lock as - * it has contention. Just correct the count - * and exit. - */ - if (signal_pending(current)) { - retval = -EINTR; - sem->sleepers = 0; - atomic_add(sleepers, &sem->count); - break; - } - - /* - * Add "everybody else" into it. They aren't - * playing, because we own the spinlock in - * wait_queue_head. The "-1" is because we're - * still hoping to get the semaphore. - */ - if (!atomic_add_negative(sleepers - 1, &sem->count)) { - sem->sleepers = 0; - break; - } - sem->sleepers = 1; /* us - see -1 above */ - spin_unlock_irqrestore(&sem->wait.lock, flags); - - schedule(); - - spin_lock_irqsave(&sem->wait.lock, flags); - tsk->state = TASK_INTERRUPTIBLE; - } - remove_wait_queue_locked(&sem->wait, &wait); - wake_up_locked(&sem->wait); - spin_unlock_irqrestore(&sem->wait.lock, flags); - - tsk->state = TASK_RUNNING; - return retval; -} - -/* - * Trylock failed - make sure we correct for - * having decremented the count. - * - * We could have done the trylock with a - * single "cmpxchg" without failure cases, - * but then it wouldn't work on a 386. - */ -int __down_trylock(struct semaphore *sem) -{ - int sleepers; - unsigned long flags; - - spin_lock_irqsave(&sem->wait.lock, flags); - sleepers = sem->sleepers + 1; - sem->sleepers = 0; - - /* - * Add "everybody else" and us into it. They aren't - * playing, because we own the spinlock in the - * wait_queue_head. - */ - if (!atomic_add_negative(sleepers, &sem->count)) { - wake_up_locked(&sem->wait); - } - - spin_unlock_irqrestore(&sem->wait.lock, flags); - return 1; -} diff --git a/mm/slub.c b/mm/slub.c index acc975fcc8cc..7f8aaa291a4e 100644 --- a/mm/slub.c +++ b/mm/slub.c @@ -837,6 +837,35 @@ static void remove_full(struct kmem_cache *s, struct page *page) spin_unlock(&n->list_lock); } +/* Tracking of the number of slabs for debugging purposes */ +static inline unsigned long slabs_node(struct kmem_cache *s, int node) +{ + struct kmem_cache_node *n = get_node(s, node); + + return atomic_long_read(&n->nr_slabs); +} + +static inline void inc_slabs_node(struct kmem_cache *s, int node) +{ + struct kmem_cache_node *n = get_node(s, node); + + /* + * May be called early in order to allocate a slab for the + * kmem_cache_node structure. Solve the chicken-egg + * dilemma by deferring the increment of the count during + * bootstrap (see early_kmem_cache_node_alloc). + */ + if (!NUMA_BUILD || n) + atomic_long_inc(&n->nr_slabs); +} +static inline void dec_slabs_node(struct kmem_cache *s, int node) +{ + struct kmem_cache_node *n = get_node(s, node); + + atomic_long_dec(&n->nr_slabs); +} + +/* Object debug checks for alloc/free paths */ static void setup_object_debug(struct kmem_cache *s, struct page *page, void *object) { @@ -1028,6 +1057,11 @@ static inline unsigned long kmem_cache_flags(unsigned long objsize, return flags; } #define slub_debug 0 + +static inline unsigned long slabs_node(struct kmem_cache *s, int node) + { return 0; } +static inline void inc_slabs_node(struct kmem_cache *s, int node) {} +static inline void dec_slabs_node(struct kmem_cache *s, int node) {} #endif /* * Slab allocation and freeing @@ -1066,7 +1100,6 @@ static void setup_object(struct kmem_cache *s, struct page *page, static struct page *new_slab(struct kmem_cache *s, gfp_t flags, int node) { struct page *page; - struct kmem_cache_node *n; void *start; void *last; void *p; @@ -1078,9 +1111,7 @@ static struct page *new_slab(struct kmem_cache *s, gfp_t flags, int node) if (!page) goto out; - n = get_node(s, page_to_nid(page)); - if (n) - atomic_long_inc(&n->nr_slabs); + inc_slabs_node(s, page_to_nid(page)); page->slab = s; page->flags |= 1 << PG_slab; if (s->flags & (SLAB_DEBUG_FREE | SLAB_RED_ZONE | SLAB_POISON | @@ -1125,6 +1156,8 @@ static void __free_slab(struct kmem_cache *s, struct page *page) NR_SLAB_RECLAIMABLE : NR_SLAB_UNRECLAIMABLE, -pages); + __ClearPageSlab(page); + reset_page_mapcount(page); __free_pages(page, s->order); } @@ -1151,11 +1184,7 @@ static void free_slab(struct kmem_cache *s, struct page *page) static void discard_slab(struct kmem_cache *s, struct page *page) { - struct kmem_cache_node *n = get_node(s, page_to_nid(page)); - - atomic_long_dec(&n->nr_slabs); - reset_page_mapcount(page); - __ClearPageSlab(page); + dec_slabs_node(s, page_to_nid(page)); free_slab(s, page); } @@ -1886,15 +1915,18 @@ static void init_kmem_cache_cpu(struct kmem_cache *s, c->node = 0; c->offset = s->offset / sizeof(void *); c->objsize = s->objsize; +#ifdef CONFIG_SLUB_STATS + memset(c->stat, 0, NR_SLUB_STAT_ITEMS * sizeof(unsigned)); +#endif } static void init_kmem_cache_node(struct kmem_cache_node *n) { n->nr_partial = 0; - atomic_long_set(&n->nr_slabs, 0); spin_lock_init(&n->list_lock); INIT_LIST_HEAD(&n->partial); #ifdef CONFIG_SLUB_DEBUG + atomic_long_set(&n->nr_slabs, 0); INIT_LIST_HEAD(&n->full); #endif } @@ -2063,7 +2095,7 @@ static struct kmem_cache_node *early_kmem_cache_node_alloc(gfp_t gfpflags, init_tracking(kmalloc_caches, n); #endif init_kmem_cache_node(n); - atomic_long_inc(&n->nr_slabs); + inc_slabs_node(kmalloc_caches, node); /* * lockdep requires consistent irq usage for each lock @@ -2376,7 +2408,7 @@ static inline int kmem_cache_close(struct kmem_cache *s) struct kmem_cache_node *n = get_node(s, node); n->nr_partial -= free_list(s, n, &n->partial); - if (atomic_long_read(&n->nr_slabs)) + if (slabs_node(s, node)) return 1; } free_kmem_cache_nodes(s); @@ -2409,10 +2441,6 @@ EXPORT_SYMBOL(kmem_cache_destroy); struct kmem_cache kmalloc_caches[PAGE_SHIFT + 1] __cacheline_aligned; EXPORT_SYMBOL(kmalloc_caches); -#ifdef CONFIG_ZONE_DMA -static struct kmem_cache *kmalloc_caches_dma[PAGE_SHIFT + 1]; -#endif - static int __init setup_slub_min_order(char *str) { get_option(&str, &slub_min_order); @@ -2472,6 +2500,7 @@ panic: } #ifdef CONFIG_ZONE_DMA +static struct kmem_cache *kmalloc_caches_dma[PAGE_SHIFT + 1]; static void sysfs_add_func(struct work_struct *w) { @@ -2688,21 +2717,6 @@ void kfree(const void *x) } EXPORT_SYMBOL(kfree); -#if defined(CONFIG_SLUB_DEBUG) || defined(CONFIG_SLABINFO) -static unsigned long count_partial(struct kmem_cache_node *n) -{ - unsigned long flags; - unsigned long x = 0; - struct page *page; - - spin_lock_irqsave(&n->list_lock, flags); - list_for_each_entry(page, &n->partial, lru) - x += page->inuse; - spin_unlock_irqrestore(&n->list_lock, flags); - return x; -} -#endif - /* * kmem_cache_shrink removes empty slabs from the partial lists and sorts * the remaining slabs by the number of items in use. The slabs with the @@ -2816,7 +2830,7 @@ static void slab_mem_offline_callback(void *arg) * and offline_pages() function shoudn't call this * callback. So, we must fail. */ - BUG_ON(atomic_long_read(&n->nr_slabs)); + BUG_ON(slabs_node(s, offline_node)); s->node[offline_node] = NULL; kmem_cache_free(kmalloc_caches, n); @@ -3181,6 +3195,21 @@ void *__kmalloc_node_track_caller(size_t size, gfp_t gfpflags, return slab_alloc(s, gfpflags, node, caller); } +#if (defined(CONFIG_SYSFS) && defined(CONFIG_SLUB_DEBUG)) || defined(CONFIG_SLABINFO) +static unsigned long count_partial(struct kmem_cache_node *n) +{ + unsigned long flags; + unsigned long x = 0; + struct page *page; + + spin_lock_irqsave(&n->list_lock, flags); + list_for_each_entry(page, &n->partial, lru) + x += page->inuse; + spin_unlock_irqrestore(&n->list_lock, flags); + return x; +} +#endif + #if defined(CONFIG_SYSFS) && defined(CONFIG_SLUB_DEBUG) static int validate_slab(struct kmem_cache *s, struct page *page, unsigned long *map) @@ -3979,10 +4008,12 @@ static int show_stat(struct kmem_cache *s, char *buf, enum stat_item si) len = sprintf(buf, "%lu", sum); +#ifdef CONFIG_SMP for_each_online_cpu(cpu) { if (data[cpu] && len < PAGE_SIZE - 20) - len += sprintf(buf + len, " c%d=%u", cpu, data[cpu]); + len += sprintf(buf + len, " C%d=%u", cpu, data[cpu]); } +#endif kfree(data); return len + sprintf(buf + len, "\n"); } diff --git a/net/sunrpc/xprtrdma/verbs.c b/net/sunrpc/xprtrdma/verbs.c index ffbf22a1d2ca..8ea283ecc522 100644 --- a/net/sunrpc/xprtrdma/verbs.c +++ b/net/sunrpc/xprtrdma/verbs.c @@ -1573,7 +1573,6 @@ rpcrdma_ep_post(struct rpcrdma_ia *ia, send_wr.sg_list = req->rl_send_iov; send_wr.num_sge = req->rl_niovs; send_wr.opcode = IB_WR_SEND; - send_wr.imm_data = 0; if (send_wr.num_sge == 4) /* no need to sync any pad (constant) */ ib_dma_sync_single_for_device(ia->ri_id->device, req->rl_send_iov[3].addr, req->rl_send_iov[3].length, diff --git a/security/Kconfig b/security/Kconfig index 5dfc206748cf..49b51f964897 100644 --- a/security/Kconfig +++ b/security/Kconfig @@ -113,10 +113,12 @@ config SECURITY_DEFAULT_MMAP_MIN_ADDR from userspace allocation. Keeping a user from writing to low pages can help reduce the impact of kernel NULL pointer bugs. - For most users with lots of address space a value of 65536 is - reasonable and should cause no problems. Programs which use vm86 - functionality would either need additional permissions from either - the LSM or the capabilities module or have this protection disabled. + For most ia64, ppc64 and x86 users with lots of address space + a value of 65536 is reasonable and should cause no problems. + On arm and other archs it should not be higher than 32768. + Programs which use vm86 functionality would either need additional + permissions from either the LSM or the capabilities module or have + this protection disabled. This value can be changed after boot using the /proc/sys/vm/mmap_min_addr tunable. diff --git a/security/commoncap.c b/security/commoncap.c index 06d5c9469ba3..852905789caf 100644 --- a/security/commoncap.c +++ b/security/commoncap.c @@ -267,7 +267,7 @@ static int get_file_caps(struct linux_binprm *bprm) rc = cap_from_disk(&vcaps, bprm, rc); if (rc) printk(KERN_NOTICE "%s: cap_from_disk returned %d for %s\n", - __FUNCTION__, rc, bprm->filename); + __func__, rc, bprm->filename); out: dput(dentry); @@ -302,7 +302,7 @@ int cap_bprm_set_security (struct linux_binprm *bprm) ret = get_file_caps(bprm); if (ret) printk(KERN_NOTICE "%s: get_file_caps returned %d for %s\n", - __FUNCTION__, ret, bprm->filename); + __func__, ret, bprm->filename); /* To support inheritance of root-permissions and suid-root * executables under compatibility mode, we raise all three diff --git a/security/keys/internal.h b/security/keys/internal.h index d36d69393356..7d894ef70370 100644 --- a/security/keys/internal.h +++ b/security/keys/internal.h @@ -22,16 +22,16 @@ void no_printk(const char *fmt, ...) #ifdef __KDEBUG #define kenter(FMT, ...) \ - printk(KERN_DEBUG "==> %s("FMT")\n", __FUNCTION__, ##__VA_ARGS__) + printk(KERN_DEBUG "==> %s("FMT")\n", __func__, ##__VA_ARGS__) #define kleave(FMT, ...) \ - printk(KERN_DEBUG "<== %s()"FMT"\n", __FUNCTION__, ##__VA_ARGS__) + printk(KERN_DEBUG "<== %s()"FMT"\n", __func__, ##__VA_ARGS__) #define kdebug(FMT, ...) \ printk(KERN_DEBUG "xxx" FMT"yyy\n", ##__VA_ARGS__) #else #define kenter(FMT, ...) \ - no_printk(KERN_DEBUG "==> %s("FMT")\n", __FUNCTION__, ##__VA_ARGS__) + no_printk(KERN_DEBUG "==> %s("FMT")\n", __func__, ##__VA_ARGS__) #define kleave(FMT, ...) \ - no_printk(KERN_DEBUG "<== %s()"FMT"\n", __FUNCTION__, ##__VA_ARGS__) + no_printk(KERN_DEBUG "<== %s()"FMT"\n", __func__, ##__VA_ARGS__) #define kdebug(FMT, ...) \ no_printk(KERN_DEBUG FMT"\n", ##__VA_ARGS__) #endif diff --git a/security/root_plug.c b/security/root_plug.c index 870f13095bb6..6112d1404c81 100644 --- a/security/root_plug.c +++ b/security/root_plug.c @@ -49,7 +49,7 @@ module_param(debug, bool, 0600); do { \ if (debug) \ printk(KERN_DEBUG "%s: %s: " fmt , \ - MY_NAME , __FUNCTION__ , \ + MY_NAME , __func__ , \ ## arg); \ } while (0) diff --git a/security/security.c b/security/security.c index b1387a6b416d..9beecac933b4 100644 --- a/security/security.c +++ b/security/security.c @@ -57,7 +57,7 @@ int __init security_init(void) if (verify(&dummy_security_ops)) { printk(KERN_ERR "%s could not verify " - "dummy_security_ops structure.\n", __FUNCTION__); + "dummy_security_ops structure.\n", __func__); return -EIO; } @@ -82,7 +82,7 @@ int register_security(struct security_operations *ops) { if (verify(ops)) { printk(KERN_DEBUG "%s could not verify " - "security_operations structure.\n", __FUNCTION__); + "security_operations structure.\n", __func__); return -EINVAL; } @@ -110,13 +110,13 @@ int mod_reg_security(const char *name, struct security_operations *ops) { if (verify(ops)) { printk(KERN_INFO "%s could not verify " - "security operations.\n", __FUNCTION__); + "security operations.\n", __func__); return -EINVAL; } if (ops == security_ops) { printk(KERN_INFO "%s security operations " - "already registered.\n", __FUNCTION__); + "already registered.\n", __func__); return -EINVAL; } diff --git a/security/selinux/Kconfig b/security/selinux/Kconfig index 2b517d618672..a436d1cfa88b 100644 --- a/security/selinux/Kconfig +++ b/security/selinux/Kconfig @@ -145,7 +145,7 @@ config SECURITY_SELINUX_POLICYDB_VERSION_MAX config SECURITY_SELINUX_POLICYDB_VERSION_MAX_VALUE int "NSA SELinux maximum supported policy format version value" depends on SECURITY_SELINUX_POLICYDB_VERSION_MAX - range 15 22 + range 15 23 default 19 help This option sets the value for the maximum policy format version diff --git a/security/selinux/Makefile b/security/selinux/Makefile index 00afd85f1edb..d47fc5e545e0 100644 --- a/security/selinux/Makefile +++ b/security/selinux/Makefile @@ -11,6 +11,7 @@ selinux-y := avc.o \ nlmsgtab.o \ netif.o \ netnode.o \ + netport.o \ exports.o selinux-$(CONFIG_SECURITY_NETWORK_XFRM) += xfrm.o diff --git a/security/selinux/avc.c b/security/selinux/avc.c index 187964e88af1..a4fc6e6d038a 100644 --- a/security/selinux/avc.c +++ b/security/selinux/avc.c @@ -871,6 +871,8 @@ int avc_has_perm_noaudit(u32 ssid, u32 tsid, int rc = 0; u32 denied; + BUG_ON(!requested); + rcu_read_lock(); node = avc_lookup(ssid, tsid, tclass, requested); @@ -890,13 +892,14 @@ int avc_has_perm_noaudit(u32 ssid, u32 tsid, denied = requested & ~(p_ae->avd.allowed); - if (!requested || denied) { - if (selinux_enforcing || (flags & AVC_STRICT)) + if (denied) { + if (flags & AVC_STRICT) rc = -EACCES; + else if (!selinux_enforcing || security_permissive_sid(ssid)) + avc_update_node(AVC_CALLBACK_GRANT, requested, ssid, + tsid, tclass); else - if (node) - avc_update_node(AVC_CALLBACK_GRANT,requested, - ssid,tsid,tclass); + rc = -EACCES; } rcu_read_unlock(); diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c index d39b59cf8a08..34f2d46c7984 100644 --- a/security/selinux/hooks.c +++ b/security/selinux/hooks.c @@ -80,6 +80,7 @@ #include "objsec.h" #include "netif.h" #include "netnode.h" +#include "netport.h" #include "xfrm.h" #include "netlabel.h" @@ -161,8 +162,7 @@ static int task_alloc_security(struct task_struct *task) if (!tsec) return -ENOMEM; - tsec->task = task; - tsec->osid = tsec->sid = tsec->ptrace_sid = SECINITSID_UNLABELED; + tsec->osid = tsec->sid = SECINITSID_UNLABELED; task->security = tsec; return 0; @@ -218,7 +218,6 @@ static int file_alloc_security(struct file *file) if (!fsec) return -ENOMEM; - fsec->file = file; fsec->sid = tsec->sid; fsec->fown_sid = tsec->sid; file->f_security = fsec; @@ -275,12 +274,11 @@ static int sk_alloc_security(struct sock *sk, int family, gfp_t priority) if (!ssec) return -ENOMEM; - ssec->sk = sk; ssec->peer_sid = SECINITSID_UNLABELED; ssec->sid = SECINITSID_UNLABELED; sk->sk_security = ssec; - selinux_netlbl_sk_security_init(ssec, family); + selinux_netlbl_sk_security_reset(ssec, family); return 0; } @@ -324,10 +322,10 @@ enum { }; static match_table_t tokens = { - {Opt_context, "context=%s"}, - {Opt_fscontext, "fscontext=%s"}, - {Opt_defcontext, "defcontext=%s"}, - {Opt_rootcontext, "rootcontext=%s"}, + {Opt_context, CONTEXT_STR "%s"}, + {Opt_fscontext, FSCONTEXT_STR "%s"}, + {Opt_defcontext, DEFCONTEXT_STR "%s"}, + {Opt_rootcontext, ROOTCONTEXT_STR "%s"}, {Opt_error, NULL}, }; @@ -671,7 +669,7 @@ static int selinux_set_mnt_opts(struct super_block *sb, rc = security_fs_use(sb->s_type->name, &sbsec->behavior, &sbsec->sid); if (rc) { printk(KERN_WARNING "%s: security_fs_use(%s) returned %d\n", - __FUNCTION__, sb->s_type->name, rc); + __func__, sb->s_type->name, rc); goto out; } @@ -1137,7 +1135,7 @@ static int inode_doinit_with_dentry(struct inode *inode, struct dentry *opt_dent } if (!dentry) { printk(KERN_WARNING "%s: no dentry for dev=%s " - "ino=%ld\n", __FUNCTION__, inode->i_sb->s_id, + "ino=%ld\n", __func__, inode->i_sb->s_id, inode->i_ino); goto out_unlock; } @@ -1175,7 +1173,7 @@ static int inode_doinit_with_dentry(struct inode *inode, struct dentry *opt_dent if (rc < 0) { if (rc != -ENODATA) { printk(KERN_WARNING "%s: getxattr returned " - "%d for dev=%s ino=%ld\n", __FUNCTION__, + "%d for dev=%s ino=%ld\n", __func__, -rc, inode->i_sb->s_id, inode->i_ino); kfree(context); goto out_unlock; @@ -1190,7 +1188,7 @@ static int inode_doinit_with_dentry(struct inode *inode, struct dentry *opt_dent if (rc) { printk(KERN_WARNING "%s: context_to_sid(%s) " "returned %d for dev=%s ino=%ld\n", - __FUNCTION__, context, -rc, + __func__, context, -rc, inode->i_sb->s_id, inode->i_ino); kfree(context); /* Leave with the unlabeled SID */ @@ -1618,6 +1616,35 @@ static inline u32 file_mask_to_av(int mode, int mask) return av; } +/* + * Convert a file mask to an access vector and include the correct open + * open permission. + */ +static inline u32 open_file_mask_to_av(int mode, int mask) +{ + u32 av = file_mask_to_av(mode, mask); + + if (selinux_policycap_openperm) { + /* + * lnk files and socks do not really have an 'open' + */ + if (S_ISREG(mode)) + av |= FILE__OPEN; + else if (S_ISCHR(mode)) + av |= CHR_FILE__OPEN; + else if (S_ISBLK(mode)) + av |= BLK_FILE__OPEN; + else if (S_ISFIFO(mode)) + av |= FIFO_FILE__OPEN; + else if (S_ISDIR(mode)) + av |= DIR__OPEN; + else + printk(KERN_ERR "SELinux: WARNING: inside open_file_to_av " + "with unknown mode:%x\n", mode); + } + return av; +} + /* Convert a Linux file to an access vector. */ static inline u32 file_to_av(struct file *file) { @@ -1645,19 +1672,13 @@ static inline u32 file_to_av(struct file *file) static int selinux_ptrace(struct task_struct *parent, struct task_struct *child) { - struct task_security_struct *psec = parent->security; - struct task_security_struct *csec = child->security; int rc; rc = secondary_ops->ptrace(parent,child); if (rc) return rc; - rc = task_has_perm(parent, child, PROCESS__PTRACE); - /* Save the SID of the tracing process for later use in apply_creds. */ - if (!(child->ptrace & PT_PTRACED) && !rc) - csec->ptrace_sid = psec->sid; - return rc; + return task_has_perm(parent, child, PROCESS__PTRACE); } static int selinux_capget(struct task_struct *target, kernel_cap_t *effective, @@ -1879,6 +1900,22 @@ static int selinux_vm_enough_memory(struct mm_struct *mm, long pages) return __vm_enough_memory(mm, pages, cap_sys_admin); } +/** + * task_tracer_task - return the task that is tracing the given task + * @task: task to consider + * + * Returns NULL if noone is tracing @task, or the &struct task_struct + * pointer to its tracer. + * + * Must be called under rcu_read_lock(). + */ +static struct task_struct *task_tracer_task(struct task_struct *task) +{ + if (task->ptrace & PT_PTRACED) + return rcu_dereference(task->parent); + return NULL; +} + /* binprm security operations */ static int selinux_bprm_alloc_security(struct linux_binprm *bprm) @@ -1889,7 +1926,6 @@ static int selinux_bprm_alloc_security(struct linux_binprm *bprm) if (!bsec) return -ENOMEM; - bsec->bprm = bprm; bsec->sid = SECINITSID_UNLABELED; bsec->set = 0; @@ -2126,12 +2162,25 @@ static void selinux_bprm_apply_creds(struct linux_binprm *bprm, int unsafe) /* Check for ptracing, and update the task SID if ok. Otherwise, leave SID unchanged and kill. */ if (unsafe & (LSM_UNSAFE_PTRACE | LSM_UNSAFE_PTRACE_CAP)) { - rc = avc_has_perm(tsec->ptrace_sid, sid, - SECCLASS_PROCESS, PROCESS__PTRACE, - NULL); - if (rc) { - bsec->unsafe = 1; - return; + struct task_struct *tracer; + struct task_security_struct *sec; + u32 ptsid = 0; + + rcu_read_lock(); + tracer = task_tracer_task(current); + if (likely(tracer != NULL)) { + sec = tracer->security; + ptsid = sec->sid; + } + rcu_read_unlock(); + + if (ptsid != 0) { + rc = avc_has_perm(ptsid, sid, SECCLASS_PROCESS, + PROCESS__PTRACE, NULL); + if (rc) { + bsec->unsafe = 1; + return; + } } } tsec->sid = sid; @@ -2239,10 +2288,10 @@ static inline int match_prefix(char *prefix, int plen, char *option, int olen) static inline int selinux_option(char *option, int len) { - return (match_prefix("context=", sizeof("context=")-1, option, len) || - match_prefix("fscontext=", sizeof("fscontext=")-1, option, len) || - match_prefix("defcontext=", sizeof("defcontext=")-1, option, len) || - match_prefix("rootcontext=", sizeof("rootcontext=")-1, option, len)); + return (match_prefix(CONTEXT_STR, sizeof(CONTEXT_STR)-1, option, len) || + match_prefix(FSCONTEXT_STR, sizeof(FSCONTEXT_STR)-1, option, len) || + match_prefix(DEFCONTEXT_STR, sizeof(DEFCONTEXT_STR)-1, option, len) || + match_prefix(ROOTCONTEXT_STR, sizeof(ROOTCONTEXT_STR)-1, option, len)); } static inline void take_option(char **to, char *from, int *first, int len) @@ -2412,7 +2461,7 @@ static int selinux_inode_init_security(struct inode *inode, struct inode *dir, printk(KERN_WARNING "%s: " "security_transition_sid failed, rc=%d (dev=%s " "ino=%ld)\n", - __FUNCTION__, + __func__, -rc, inode->i_sb->s_id, inode->i_ino); return rc; } @@ -2536,7 +2585,7 @@ static int selinux_inode_permission(struct inode *inode, int mask, } return inode_has_perm(current, inode, - file_mask_to_av(inode->i_mode, mask), NULL); + open_file_mask_to_av(inode->i_mode, mask), NULL); } static int selinux_inode_setattr(struct dentry *dentry, struct iattr *iattr) @@ -2646,7 +2695,7 @@ static void selinux_inode_post_setxattr(struct dentry *dentry, char *name, rc = security_context_to_sid(value, size, &newsid); if (rc) { printk(KERN_WARNING "%s: unable to obtain SID for context " - "%s, rc=%d\n", __FUNCTION__, (char*)value, -rc); + "%s, rc=%d\n", __func__, (char *)value, -rc); return; } @@ -3087,11 +3136,6 @@ static int selinux_task_alloc_security(struct task_struct *tsk) tsec2->keycreate_sid = tsec1->keycreate_sid; tsec2->sockcreate_sid = tsec1->sockcreate_sid; - /* Retain ptracer SID across fork, if any. - This will be reset by the ptrace hook upon any - subsequent ptrace_attach operations. */ - tsec2->ptrace_sid = tsec1->ptrace_sid; - return 0; } @@ -3627,10 +3671,8 @@ static int selinux_socket_bind(struct socket *sock, struct sockaddr *address, in inet_get_local_port_range(&low, &high); if (snum < max(PROT_SOCK, low) || snum > high) { - err = security_port_sid(sk->sk_family, - sk->sk_type, - sk->sk_protocol, snum, - &sid); + err = sel_netport_sid(sk->sk_protocol, + snum, &sid); if (err) goto out; AVC_AUDIT_DATA_INIT(&ad,NET); @@ -3718,8 +3760,7 @@ static int selinux_socket_connect(struct socket *sock, struct sockaddr *address, snum = ntohs(addr6->sin6_port); } - err = security_port_sid(sk->sk_family, sk->sk_type, - sk->sk_protocol, snum, &sid); + err = sel_netport_sid(sk->sk_protocol, snum, &sid); if (err) goto out; @@ -3950,9 +3991,8 @@ static int selinux_sock_rcv_skb_iptables_compat(struct sock *sk, if (!recv_perm) return 0; - err = security_port_sid(sk->sk_family, sk->sk_type, - sk->sk_protocol, ntohs(ad->u.net.sport), - &port_sid); + err = sel_netport_sid(sk->sk_protocol, + ntohs(ad->u.net.sport), &port_sid); if (unlikely(err)) { printk(KERN_WARNING "SELinux: failure in" @@ -4139,7 +4179,7 @@ static void selinux_sk_clone_security(const struct sock *sk, struct sock *newsk) newssec->peer_sid = ssec->peer_sid; newssec->sclass = ssec->sclass; - selinux_netlbl_sk_security_clone(ssec, newssec); + selinux_netlbl_sk_security_reset(newssec, newsk->sk_family); } static void selinux_sk_getsecid(struct sock *sk, u32 *secid) @@ -4373,9 +4413,8 @@ static int selinux_ip_postroute_iptables_compat(struct sock *sk, if (send_perm != 0) return 0; - err = security_port_sid(sk->sk_family, sk->sk_type, - sk->sk_protocol, ntohs(ad->u.net.dport), - &port_sid); + err = sel_netport_sid(sk->sk_protocol, + ntohs(ad->u.net.dport), &port_sid); if (unlikely(err)) { printk(KERN_WARNING "SELinux: failure in" @@ -4561,7 +4600,6 @@ static int ipc_alloc_security(struct task_struct *task, return -ENOMEM; isec->sclass = sclass; - isec->ipc_perm = perm; isec->sid = tsec->sid; perm->security = isec; @@ -4583,7 +4621,6 @@ static int msg_msg_alloc_security(struct msg_msg *msg) if (!msec) return -ENOMEM; - msec->msg = msg; msec->sid = SECINITSID_UNLABELED; msg->security = msec; @@ -4994,14 +5031,14 @@ static int selinux_register_security (const char *name, struct security_operatio { if (secondary_ops != original_ops) { printk(KERN_ERR "%s: There is already a secondary security " - "module registered.\n", __FUNCTION__); + "module registered.\n", __func__); return -EINVAL; } secondary_ops = ops; printk(KERN_INFO "%s: Registering secondary module %s\n", - __FUNCTION__, + __func__, name); return 0; @@ -5057,6 +5094,7 @@ static int selinux_setprocattr(struct task_struct *p, char *name, void *value, size_t size) { struct task_security_struct *tsec; + struct task_struct *tracer; u32 sid = 0; int error; char *str = value; @@ -5145,18 +5183,24 @@ static int selinux_setprocattr(struct task_struct *p, /* Check for ptracing, and update the task SID if ok. Otherwise, leave SID unchanged and fail. */ task_lock(p); - if (p->ptrace & PT_PTRACED) { - error = avc_has_perm_noaudit(tsec->ptrace_sid, sid, + rcu_read_lock(); + tracer = task_tracer_task(p); + if (tracer != NULL) { + struct task_security_struct *ptsec = tracer->security; + u32 ptsid = ptsec->sid; + rcu_read_unlock(); + error = avc_has_perm_noaudit(ptsid, sid, SECCLASS_PROCESS, PROCESS__PTRACE, 0, &avd); if (!error) tsec->sid = sid; task_unlock(p); - avc_audit(tsec->ptrace_sid, sid, SECCLASS_PROCESS, + avc_audit(ptsid, sid, SECCLASS_PROCESS, PROCESS__PTRACE, &avd, error, NULL); if (error) return error; } else { + rcu_read_unlock(); tsec->sid = sid; task_unlock(p); } @@ -5194,7 +5238,6 @@ static int selinux_key_alloc(struct key *k, struct task_struct *tsk, if (!ksec) return -ENOMEM; - ksec->obj = k; if (tsec->keycreate_sid) ksec->sid = tsec->keycreate_sid; else @@ -5631,5 +5674,3 @@ int selinux_disable(void) return 0; } #endif - - diff --git a/security/selinux/include/av_perm_to_string.h b/security/selinux/include/av_perm_to_string.h index d5696690d3a2..1223b4ff9bee 100644 --- a/security/selinux/include/av_perm_to_string.h +++ b/security/selinux/include/av_perm_to_string.h @@ -14,12 +14,17 @@ S_(SECCLASS_DIR, DIR__REPARENT, "reparent") S_(SECCLASS_DIR, DIR__SEARCH, "search") S_(SECCLASS_DIR, DIR__RMDIR, "rmdir") + S_(SECCLASS_DIR, DIR__OPEN, "open") S_(SECCLASS_FILE, FILE__EXECUTE_NO_TRANS, "execute_no_trans") S_(SECCLASS_FILE, FILE__ENTRYPOINT, "entrypoint") S_(SECCLASS_FILE, FILE__EXECMOD, "execmod") + S_(SECCLASS_FILE, FILE__OPEN, "open") S_(SECCLASS_CHR_FILE, CHR_FILE__EXECUTE_NO_TRANS, "execute_no_trans") S_(SECCLASS_CHR_FILE, CHR_FILE__ENTRYPOINT, "entrypoint") S_(SECCLASS_CHR_FILE, CHR_FILE__EXECMOD, "execmod") + S_(SECCLASS_CHR_FILE, CHR_FILE__OPEN, "open") + S_(SECCLASS_BLK_FILE, BLK_FILE__OPEN, "open") + S_(SECCLASS_FIFO_FILE, FIFO_FILE__OPEN, "open") S_(SECCLASS_FD, FD__USE, "use") S_(SECCLASS_TCP_SOCKET, TCP_SOCKET__CONNECTTO, "connectto") S_(SECCLASS_TCP_SOCKET, TCP_SOCKET__NEWCONN, "newconn") diff --git a/security/selinux/include/av_permissions.h b/security/selinux/include/av_permissions.h index 75b41311ab86..c4c51165c505 100644 --- a/security/selinux/include/av_permissions.h +++ b/security/selinux/include/av_permissions.h @@ -79,6 +79,7 @@ #define DIR__REPARENT 0x00080000UL #define DIR__SEARCH 0x00100000UL #define DIR__RMDIR 0x00200000UL +#define DIR__OPEN 0x00400000UL #define FILE__IOCTL 0x00000001UL #define FILE__READ 0x00000002UL #define FILE__WRITE 0x00000004UL @@ -99,6 +100,7 @@ #define FILE__EXECUTE_NO_TRANS 0x00020000UL #define FILE__ENTRYPOINT 0x00040000UL #define FILE__EXECMOD 0x00080000UL +#define FILE__OPEN 0x00100000UL #define LNK_FILE__IOCTL 0x00000001UL #define LNK_FILE__READ 0x00000002UL #define LNK_FILE__WRITE 0x00000004UL @@ -136,6 +138,7 @@ #define CHR_FILE__EXECUTE_NO_TRANS 0x00020000UL #define CHR_FILE__ENTRYPOINT 0x00040000UL #define CHR_FILE__EXECMOD 0x00080000UL +#define CHR_FILE__OPEN 0x00100000UL #define BLK_FILE__IOCTL 0x00000001UL #define BLK_FILE__READ 0x00000002UL #define BLK_FILE__WRITE 0x00000004UL @@ -153,6 +156,7 @@ #define BLK_FILE__SWAPON 0x00004000UL #define BLK_FILE__QUOTAON 0x00008000UL #define BLK_FILE__MOUNTON 0x00010000UL +#define BLK_FILE__OPEN 0x00020000UL #define SOCK_FILE__IOCTL 0x00000001UL #define SOCK_FILE__READ 0x00000002UL #define SOCK_FILE__WRITE 0x00000004UL @@ -187,6 +191,7 @@ #define FIFO_FILE__SWAPON 0x00004000UL #define FIFO_FILE__QUOTAON 0x00008000UL #define FIFO_FILE__MOUNTON 0x00010000UL +#define FIFO_FILE__OPEN 0x00020000UL #define FD__USE 0x00000001UL #define SOCKET__IOCTL 0x00000001UL #define SOCKET__READ 0x00000002UL diff --git a/security/selinux/include/netlabel.h b/security/selinux/include/netlabel.h index 00a2809c8506..9a9e7cd9a379 100644 --- a/security/selinux/include/netlabel.h +++ b/security/selinux/include/netlabel.h @@ -41,10 +41,6 @@ void selinux_netlbl_cache_invalidate(void); void selinux_netlbl_sk_security_reset(struct sk_security_struct *ssec, int family); -void selinux_netlbl_sk_security_init(struct sk_security_struct *ssec, - int family); -void selinux_netlbl_sk_security_clone(struct sk_security_struct *ssec, - struct sk_security_struct *newssec); int selinux_netlbl_skbuff_getsid(struct sk_buff *skb, u16 family, @@ -73,18 +69,6 @@ static inline void selinux_netlbl_sk_security_reset( { return; } -static inline void selinux_netlbl_sk_security_init( - struct sk_security_struct *ssec, - int family) -{ - return; -} -static inline void selinux_netlbl_sk_security_clone( - struct sk_security_struct *ssec, - struct sk_security_struct *newssec) -{ - return; -} static inline int selinux_netlbl_skbuff_getsid(struct sk_buff *skb, u16 family, diff --git a/security/selinux/include/netport.h b/security/selinux/include/netport.h new file mode 100644 index 000000000000..8991752eaf93 --- /dev/null +++ b/security/selinux/include/netport.h @@ -0,0 +1,31 @@ +/* + * Network port table + * + * SELinux must keep a mapping of network ports to labels/SIDs. This + * mapping is maintained as part of the normal policy but a fast cache is + * needed to reduce the lookup overhead. + * + * Author: Paul Moore <paul.moore@hp.com> + * + */ + +/* + * (c) Copyright Hewlett-Packard Development Company, L.P., 2008 + * + * This program is free software: you can redistribute it and/or modify + * it under the terms of version 2 of the GNU General Public License as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + */ + +#ifndef _SELINUX_NETPORT_H +#define _SELINUX_NETPORT_H + +int sel_netport_sid(u8 protocol, u16 pnum, u32 *sid); + +#endif diff --git a/security/selinux/include/objsec.h b/security/selinux/include/objsec.h index c6c2bb4ebacc..300b61bad7b3 100644 --- a/security/selinux/include/objsec.h +++ b/security/selinux/include/objsec.h @@ -28,14 +28,12 @@ #include "avc.h" struct task_security_struct { - struct task_struct *task; /* back pointer to task object */ u32 osid; /* SID prior to last execve */ u32 sid; /* current SID */ u32 exec_sid; /* exec SID */ u32 create_sid; /* fscreate SID */ u32 keycreate_sid; /* keycreate SID */ u32 sockcreate_sid; /* fscreate SID */ - u32 ptrace_sid; /* SID of ptrace parent */ }; struct inode_security_struct { @@ -50,7 +48,6 @@ struct inode_security_struct { }; struct file_security_struct { - struct file *file; /* back pointer to file object */ u32 sid; /* SID of open file description */ u32 fown_sid; /* SID of file owner (for SIGIO) */ u32 isid; /* SID of inode at the time of file open */ @@ -73,18 +70,15 @@ struct superblock_security_struct { }; struct msg_security_struct { - struct msg_msg *msg; /* back pointer */ u32 sid; /* SID of message */ }; struct ipc_security_struct { - struct kern_ipc_perm *ipc_perm; /* back pointer */ u16 sclass; /* security class of this object */ u32 sid; /* SID of IPC resource */ }; struct bprm_security_struct { - struct linux_binprm *bprm; /* back pointer to bprm object */ u32 sid; /* SID for transformed process */ unsigned char set; @@ -109,8 +103,13 @@ struct netnode_security_struct { u16 family; /* address family */ }; +struct netport_security_struct { + u32 sid; /* SID for this node */ + u16 port; /* port number */ + u8 protocol; /* transport protocol */ +}; + struct sk_security_struct { - struct sock *sk; /* back pointer to sk object */ u32 sid; /* SID of this object */ u32 peer_sid; /* SID of peer */ u16 sclass; /* sock security class */ @@ -120,12 +119,10 @@ struct sk_security_struct { NLBL_REQUIRE, NLBL_LABELED, } nlbl_state; - spinlock_t nlbl_lock; /* protects nlbl_state */ #endif }; struct key_security_struct { - struct key *obj; /* back pointer */ u32 sid; /* SID of key */ }; diff --git a/security/selinux/include/security.h b/security/selinux/include/security.h index 44e12ec88090..1904c462a605 100644 --- a/security/selinux/include/security.h +++ b/security/selinux/include/security.h @@ -26,13 +26,14 @@ #define POLICYDB_VERSION_AVTAB 20 #define POLICYDB_VERSION_RANGETRANS 21 #define POLICYDB_VERSION_POLCAP 22 +#define POLICYDB_VERSION_PERMISSIVE 23 /* Range of policy versions we understand*/ #define POLICYDB_VERSION_MIN POLICYDB_VERSION_BASE #ifdef CONFIG_SECURITY_SELINUX_POLICYDB_VERSION_MAX #define POLICYDB_VERSION_MAX CONFIG_SECURITY_SELINUX_POLICYDB_VERSION_MAX_VALUE #else -#define POLICYDB_VERSION_MAX POLICYDB_VERSION_POLCAP +#define POLICYDB_VERSION_MAX POLICYDB_VERSION_PERMISSIVE #endif #define CONTEXT_MNT 0x01 @@ -40,6 +41,11 @@ #define ROOTCONTEXT_MNT 0x04 #define DEFCONTEXT_MNT 0x08 +#define CONTEXT_STR "context=" +#define FSCONTEXT_STR "fscontext=" +#define ROOTCONTEXT_STR "rootcontext=" +#define DEFCONTEXT_STR "defcontext=" + struct netlbl_lsm_secattr; extern int selinux_enabled; @@ -48,11 +54,13 @@ extern int selinux_mls_enabled; /* Policy capabilities */ enum { POLICYDB_CAPABILITY_NETPEER, + POLICYDB_CAPABILITY_OPENPERM, __POLICYDB_CAPABILITY_MAX }; #define POLICYDB_CAPABILITY_MAX (__POLICYDB_CAPABILITY_MAX - 1) extern int selinux_policycap_netpeer; +extern int selinux_policycap_openperm; int security_load_policy(void * data, size_t len); @@ -67,6 +75,8 @@ struct av_decision { u32 seqno; }; +int security_permissive_sid(u32 sid); + int security_compute_av(u32 ssid, u32 tsid, u16 tclass, u32 requested, struct av_decision *avd); @@ -92,8 +102,7 @@ int security_context_to_sid_default(char *scontext, u32 scontext_len, int security_get_user_sids(u32 callsid, char *username, u32 **sids, u32 *nel); -int security_port_sid(u16 domain, u16 type, u8 protocol, u16 port, - u32 *out_sid); +int security_port_sid(u8 protocol, u16 port, u32 *out_sid); int security_netif_sid(char *name, u32 *if_sid); diff --git a/security/selinux/netlabel.c b/security/selinux/netlabel.c index 0fa2be4149e8..e8ee91ac12ef 100644 --- a/security/selinux/netlabel.c +++ b/security/selinux/netlabel.c @@ -34,6 +34,7 @@ #include "objsec.h" #include "security.h" +#include "netlabel.h" /** * selinux_netlbl_sidlookup_cached - Cache a SID lookup @@ -69,9 +70,7 @@ static int selinux_netlbl_sidlookup_cached(struct sk_buff *skb, * * Description: * Attempt to label a socket using the NetLabel mechanism using the given - * SID. Returns zero values on success, negative values on failure. The - * caller is responsibile for calling rcu_read_lock() before calling this - * this function and rcu_read_unlock() after this function returns. + * SID. Returns zero values on success, negative values on failure. * */ static int selinux_netlbl_sock_setsid(struct sock *sk, u32 sid) @@ -86,11 +85,8 @@ static int selinux_netlbl_sock_setsid(struct sock *sk, u32 sid) if (rc != 0) goto sock_setsid_return; rc = netlbl_sock_setattr(sk, &secattr); - if (rc == 0) { - spin_lock_bh(&sksec->nlbl_lock); + if (rc == 0) sksec->nlbl_state = NLBL_LABELED; - spin_unlock_bh(&sksec->nlbl_lock); - } sock_setsid_return: netlbl_secattr_destroy(&secattr); @@ -129,45 +125,6 @@ void selinux_netlbl_sk_security_reset(struct sk_security_struct *ssec, } /** - * selinux_netlbl_sk_security_init - Setup the NetLabel fields - * @ssec: the sk_security_struct - * @family: the socket family - * - * Description: - * Called when a new sk_security_struct is allocated to initialize the NetLabel - * fields. - * - */ -void selinux_netlbl_sk_security_init(struct sk_security_struct *ssec, - int family) -{ - /* No locking needed, we are the only one who has access to ssec */ - selinux_netlbl_sk_security_reset(ssec, family); - spin_lock_init(&ssec->nlbl_lock); -} - -/** - * selinux_netlbl_sk_security_clone - Copy the NetLabel fields - * @ssec: the original sk_security_struct - * @newssec: the cloned sk_security_struct - * - * Description: - * Clone the NetLabel specific sk_security_struct fields from @ssec to - * @newssec. - * - */ -void selinux_netlbl_sk_security_clone(struct sk_security_struct *ssec, - struct sk_security_struct *newssec) -{ - /* We don't need to take newssec->nlbl_lock because we are the only - * thread with access to newssec, but we do need to take the RCU read - * lock as other threads could have access to ssec */ - rcu_read_lock(); - selinux_netlbl_sk_security_reset(newssec, ssec->sk->sk_family); - rcu_read_unlock(); -} - -/** * selinux_netlbl_skbuff_getsid - Get the sid of a packet using NetLabel * @skb: the packet * @family: protocol family @@ -221,12 +178,8 @@ void selinux_netlbl_sock_graft(struct sock *sk, struct socket *sock) struct netlbl_lsm_secattr secattr; u32 nlbl_peer_sid; - rcu_read_lock(); - - if (sksec->nlbl_state != NLBL_REQUIRE) { - rcu_read_unlock(); + if (sksec->nlbl_state != NLBL_REQUIRE) return; - } netlbl_secattr_init(&secattr); if (netlbl_sock_getattr(sk, &secattr) == 0 && @@ -239,8 +192,6 @@ void selinux_netlbl_sock_graft(struct sock *sk, struct socket *sock) * here we will pick up the pieces in later calls to * selinux_netlbl_inode_permission(). */ selinux_netlbl_sock_setsid(sk, sksec->sid); - - rcu_read_unlock(); } /** @@ -254,16 +205,13 @@ void selinux_netlbl_sock_graft(struct sock *sk, struct socket *sock) */ int selinux_netlbl_socket_post_create(struct socket *sock) { - int rc = 0; struct sock *sk = sock->sk; struct sk_security_struct *sksec = sk->sk_security; - rcu_read_lock(); - if (sksec->nlbl_state == NLBL_REQUIRE) - rc = selinux_netlbl_sock_setsid(sk, sksec->sid); - rcu_read_unlock(); + if (sksec->nlbl_state != NLBL_REQUIRE) + return 0; - return rc; + return selinux_netlbl_sock_setsid(sk, sksec->sid); } /** @@ -288,21 +236,21 @@ int selinux_netlbl_inode_permission(struct inode *inode, int mask) if (!S_ISSOCK(inode->i_mode) || ((mask & (MAY_WRITE | MAY_APPEND)) == 0)) return 0; + sock = SOCKET_I(inode); sk = sock->sk; sksec = sk->sk_security; - - rcu_read_lock(); - if (sksec->nlbl_state != NLBL_REQUIRE) { - rcu_read_unlock(); + if (sksec->nlbl_state != NLBL_REQUIRE) return 0; - } + local_bh_disable(); bh_lock_sock_nested(sk); - rc = selinux_netlbl_sock_setsid(sk, sksec->sid); + if (likely(sksec->nlbl_state == NLBL_REQUIRE)) + rc = selinux_netlbl_sock_setsid(sk, sksec->sid); + else + rc = 0; bh_unlock_sock(sk); local_bh_enable(); - rcu_read_unlock(); return rc; } @@ -385,7 +333,6 @@ int selinux_netlbl_socket_setsockopt(struct socket *sock, struct sk_security_struct *sksec = sk->sk_security; struct netlbl_lsm_secattr secattr; - rcu_read_lock(); if (level == IPPROTO_IP && optname == IP_OPTIONS && sksec->nlbl_state == NLBL_LABELED) { netlbl_secattr_init(&secattr); @@ -396,7 +343,6 @@ int selinux_netlbl_socket_setsockopt(struct socket *sock, rc = -EACCES; netlbl_secattr_destroy(&secattr); } - rcu_read_unlock(); return rc; } diff --git a/security/selinux/netlink.c b/security/selinux/netlink.c index b59871d74dad..6214a7a73149 100644 --- a/security/selinux/netlink.c +++ b/security/selinux/netlink.c @@ -89,7 +89,7 @@ out: nlmsg_failure: kfree_skb(skb); oom: - printk(KERN_ERR "SELinux: OOM in %s\n", __FUNCTION__); + printk(KERN_ERR "SELinux: OOM in %s\n", __func__); goto out; } diff --git a/security/selinux/netport.c b/security/selinux/netport.c new file mode 100644 index 000000000000..68ede3c498ab --- /dev/null +++ b/security/selinux/netport.c @@ -0,0 +1,286 @@ +/* + * Network port table + * + * SELinux must keep a mapping of network ports to labels/SIDs. This + * mapping is maintained as part of the normal policy but a fast cache is + * needed to reduce the lookup overhead. + * + * Author: Paul Moore <paul.moore@hp.com> + * + * This code is heavily based on the "netif" concept originally developed by + * James Morris <jmorris@redhat.com> + * (see security/selinux/netif.c for more information) + * + */ + +/* + * (c) Copyright Hewlett-Packard Development Company, L.P., 2008 + * + * This program is free software: you can redistribute it and/or modify + * it under the terms of version 2 of the GNU General Public License as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + */ + +#include <linux/types.h> +#include <linux/rcupdate.h> +#include <linux/list.h> +#include <linux/spinlock.h> +#include <linux/in.h> +#include <linux/in6.h> +#include <linux/ip.h> +#include <linux/ipv6.h> +#include <net/ip.h> +#include <net/ipv6.h> +#include <asm/bug.h> + +#include "netport.h" +#include "objsec.h" + +#define SEL_NETPORT_HASH_SIZE 256 +#define SEL_NETPORT_HASH_BKT_LIMIT 16 + +struct sel_netport_bkt { + int size; + struct list_head list; +}; + +struct sel_netport { + struct netport_security_struct psec; + + struct list_head list; + struct rcu_head rcu; +}; + +/* NOTE: we are using a combined hash table for both IPv4 and IPv6, the reason + * for this is that I suspect most users will not make heavy use of both + * address families at the same time so one table will usually end up wasted, + * if this becomes a problem we can always add a hash table for each address + * family later */ + +static LIST_HEAD(sel_netport_list); +static DEFINE_SPINLOCK(sel_netport_lock); +static struct sel_netport_bkt sel_netport_hash[SEL_NETPORT_HASH_SIZE]; + +/** + * sel_netport_free - Frees a port entry + * @p: the entry's RCU field + * + * Description: + * This function is designed to be used as a callback to the call_rcu() + * function so that memory allocated to a hash table port entry can be + * released safely. + * + */ +static void sel_netport_free(struct rcu_head *p) +{ + struct sel_netport *port = container_of(p, struct sel_netport, rcu); + kfree(port); +} + +/** + * sel_netport_hashfn - Hashing function for the port table + * @pnum: port number + * + * Description: + * This is the hashing function for the port table, it returns the bucket + * number for the given port. + * + */ +static unsigned int sel_netport_hashfn(u16 pnum) +{ + return (pnum & (SEL_NETPORT_HASH_SIZE - 1)); +} + +/** + * sel_netport_find - Search for a port record + * @protocol: protocol + * @port: pnum + * + * Description: + * Search the network port table and return the matching record. If an entry + * can not be found in the table return NULL. + * + */ +static struct sel_netport *sel_netport_find(u8 protocol, u16 pnum) +{ + unsigned int idx; + struct sel_netport *port; + + idx = sel_netport_hashfn(pnum); + list_for_each_entry_rcu(port, &sel_netport_hash[idx].list, list) + if (port->psec.port == pnum && + port->psec.protocol == protocol) + return port; + + return NULL; +} + +/** + * sel_netport_insert - Insert a new port into the table + * @port: the new port record + * + * Description: + * Add a new port record to the network address hash table. Returns zero on + * success, negative values on failure. + * + */ +static int sel_netport_insert(struct sel_netport *port) +{ + unsigned int idx; + + /* we need to impose a limit on the growth of the hash table so check + * this bucket to make sure it is within the specified bounds */ + idx = sel_netport_hashfn(port->psec.port); + list_add_rcu(&port->list, &sel_netport_hash[idx].list); + if (sel_netport_hash[idx].size == SEL_NETPORT_HASH_BKT_LIMIT) { + struct sel_netport *tail; + tail = list_entry(port->list.prev, struct sel_netport, list); + list_del_rcu(port->list.prev); + call_rcu(&tail->rcu, sel_netport_free); + } else + sel_netport_hash[idx].size++; + + return 0; +} + +/** + * sel_netport_sid_slow - Lookup the SID of a network address using the policy + * @protocol: protocol + * @pnum: port + * @sid: port SID + * + * Description: + * This function determines the SID of a network port by quering the security + * policy. The result is added to the network port table to speedup future + * queries. Returns zero on success, negative values on failure. + * + */ +static int sel_netport_sid_slow(u8 protocol, u16 pnum, u32 *sid) +{ + int ret; + struct sel_netport *port; + struct sel_netport *new = NULL; + + spin_lock_bh(&sel_netport_lock); + port = sel_netport_find(protocol, pnum); + if (port != NULL) { + *sid = port->psec.sid; + ret = 0; + goto out; + } + new = kzalloc(sizeof(*new), GFP_ATOMIC); + if (new == NULL) { + ret = -ENOMEM; + goto out; + } + ret = security_port_sid(protocol, pnum, &new->psec.sid); + if (ret != 0) + goto out; + new->psec.port = pnum; + new->psec.protocol = protocol; + ret = sel_netport_insert(new); + if (ret != 0) + goto out; + *sid = new->psec.sid; + +out: + spin_unlock_bh(&sel_netport_lock); + if (unlikely(ret)) { + printk(KERN_WARNING + "SELinux: failure in sel_netport_sid_slow()," + " unable to determine network port label\n"); + kfree(new); + } + return ret; +} + +/** + * sel_netport_sid - Lookup the SID of a network port + * @protocol: protocol + * @pnum: port + * @sid: port SID + * + * Description: + * This function determines the SID of a network port using the fastest method + * possible. First the port table is queried, but if an entry can't be found + * then the policy is queried and the result is added to the table to speedup + * future queries. Returns zero on success, negative values on failure. + * + */ +int sel_netport_sid(u8 protocol, u16 pnum, u32 *sid) +{ + struct sel_netport *port; + + rcu_read_lock(); + port = sel_netport_find(protocol, pnum); + if (port != NULL) { + *sid = port->psec.sid; + rcu_read_unlock(); + return 0; + } + rcu_read_unlock(); + + return sel_netport_sid_slow(protocol, pnum, sid); +} + +/** + * sel_netport_flush - Flush the entire network port table + * + * Description: + * Remove all entries from the network address table. + * + */ +static void sel_netport_flush(void) +{ + unsigned int idx; + struct sel_netport *port; + + spin_lock_bh(&sel_netport_lock); + for (idx = 0; idx < SEL_NETPORT_HASH_SIZE; idx++) { + list_for_each_entry(port, &sel_netport_hash[idx].list, list) { + list_del_rcu(&port->list); + call_rcu(&port->rcu, sel_netport_free); + } + sel_netport_hash[idx].size = 0; + } + spin_unlock_bh(&sel_netport_lock); +} + +static int sel_netport_avc_callback(u32 event, u32 ssid, u32 tsid, + u16 class, u32 perms, u32 *retained) +{ + if (event == AVC_CALLBACK_RESET) { + sel_netport_flush(); + synchronize_net(); + } + return 0; +} + +static __init int sel_netport_init(void) +{ + int iter; + int ret; + + if (!selinux_enabled) + return 0; + + for (iter = 0; iter < SEL_NETPORT_HASH_SIZE; iter++) { + INIT_LIST_HEAD(&sel_netport_hash[iter].list); + sel_netport_hash[iter].size = 0; + } + + ret = avc_add_callback(sel_netport_avc_callback, AVC_CALLBACK_RESET, + SECSID_NULL, SECSID_NULL, SECCLASS_NULL, 0); + if (ret != 0) + panic("avc_add_callback() failed, error %d\n", ret); + + return ret; +} + +__initcall(sel_netport_init); diff --git a/security/selinux/selinuxfs.c b/security/selinux/selinuxfs.c index 0341567665b3..26fabad09769 100644 --- a/security/selinux/selinuxfs.c +++ b/security/selinux/selinuxfs.c @@ -42,7 +42,8 @@ /* Policy capability filenames */ static char *policycap_names[] = { - "network_peer_controls" + "network_peer_controls", + "open_perms" }; unsigned int selinux_checkreqprot = CONFIG_SECURITY_SELINUX_CHECKREQPROT_VALUE; @@ -391,7 +392,7 @@ static ssize_t sel_write_context(struct file * file, char *buf, size_t size) if (len > SIMPLE_TRANSACTION_LIMIT) { printk(KERN_ERR "%s: context size (%u) exceeds payload " - "max\n", __FUNCTION__, len); + "max\n", __func__, len); length = -ERANGE; goto out; } @@ -644,7 +645,7 @@ static ssize_t sel_write_create(struct file * file, char *buf, size_t size) if (len > SIMPLE_TRANSACTION_LIMIT) { printk(KERN_ERR "%s: context size (%u) exceeds payload " - "max\n", __FUNCTION__, len); + "max\n", __func__, len); length = -ERANGE; goto out3; } @@ -821,7 +822,7 @@ static ssize_t sel_write_member(struct file * file, char *buf, size_t size) if (len > SIMPLE_TRANSACTION_LIMIT) { printk(KERN_ERR "%s: context size (%u) exceeds payload " - "max\n", __FUNCTION__, len); + "max\n", __func__, len); length = -ERANGE; goto out3; } @@ -1760,7 +1761,7 @@ static int sel_fill_super(struct super_block * sb, void * data, int silent) out: return ret; err: - printk(KERN_ERR "%s: failed while creating inodes\n", __FUNCTION__); + printk(KERN_ERR "%s: failed while creating inodes\n", __func__); goto out; } diff --git a/security/selinux/ss/avtab.c b/security/selinux/ss/avtab.c index cd10e27fc9e6..916e73a18bc5 100644 --- a/security/selinux/ss/avtab.c +++ b/security/selinux/ss/avtab.c @@ -280,8 +280,8 @@ int avtab_alloc(struct avtab *h, u32 nrules) h->nel = 0; h->nslot = nslot; h->mask = mask; - printk(KERN_DEBUG "SELinux:%d avtab hash slots allocated. " - "Num of rules:%d\n", h->nslot, nrules); + printk(KERN_DEBUG "SELinux: %d avtab hash slots, %d rules.\n", + h->nslot, nrules); return 0; } @@ -345,18 +345,18 @@ int avtab_read_item(struct avtab *a, void *fp, struct policydb *pol, if (vers < POLICYDB_VERSION_AVTAB) { rc = next_entry(buf32, fp, sizeof(u32)); if (rc < 0) { - printk(KERN_ERR "security: avtab: truncated entry\n"); + printk(KERN_ERR "SELinux: avtab: truncated entry\n"); return -1; } items2 = le32_to_cpu(buf32[0]); if (items2 > ARRAY_SIZE(buf32)) { - printk(KERN_ERR "security: avtab: entry overflow\n"); + printk(KERN_ERR "SELinux: avtab: entry overflow\n"); return -1; } rc = next_entry(buf32, fp, sizeof(u32)*items2); if (rc < 0) { - printk(KERN_ERR "security: avtab: truncated entry\n"); + printk(KERN_ERR "SELinux: avtab: truncated entry\n"); return -1; } items = 0; @@ -364,19 +364,19 @@ int avtab_read_item(struct avtab *a, void *fp, struct policydb *pol, val = le32_to_cpu(buf32[items++]); key.source_type = (u16)val; if (key.source_type != val) { - printk("security: avtab: truncated source type\n"); + printk("SELinux: avtab: truncated source type\n"); return -1; } val = le32_to_cpu(buf32[items++]); key.target_type = (u16)val; if (key.target_type != val) { - printk("security: avtab: truncated target type\n"); + printk("SELinux: avtab: truncated target type\n"); return -1; } val = le32_to_cpu(buf32[items++]); key.target_class = (u16)val; if (key.target_class != val) { - printk("security: avtab: truncated target class\n"); + printk("SELinux: avtab: truncated target class\n"); return -1; } @@ -384,12 +384,12 @@ int avtab_read_item(struct avtab *a, void *fp, struct policydb *pol, enabled = (val & AVTAB_ENABLED_OLD) ? AVTAB_ENABLED : 0; if (!(val & (AVTAB_AV | AVTAB_TYPE))) { - printk("security: avtab: null entry\n"); + printk("SELinux: avtab: null entry\n"); return -1; } if ((val & AVTAB_AV) && (val & AVTAB_TYPE)) { - printk("security: avtab: entry has both access vectors and types\n"); + printk("SELinux: avtab: entry has both access vectors and types\n"); return -1; } @@ -403,7 +403,7 @@ int avtab_read_item(struct avtab *a, void *fp, struct policydb *pol, } if (items != items2) { - printk("security: avtab: entry only had %d items, expected %d\n", items2, items); + printk("SELinux: avtab: entry only had %d items, expected %d\n", items2, items); return -1; } return 0; @@ -411,7 +411,7 @@ int avtab_read_item(struct avtab *a, void *fp, struct policydb *pol, rc = next_entry(buf16, fp, sizeof(u16)*4); if (rc < 0) { - printk("security: avtab: truncated entry\n"); + printk("SELinux: avtab: truncated entry\n"); return -1; } @@ -424,7 +424,7 @@ int avtab_read_item(struct avtab *a, void *fp, struct policydb *pol, if (!policydb_type_isvalid(pol, key.source_type) || !policydb_type_isvalid(pol, key.target_type) || !policydb_class_isvalid(pol, key.target_class)) { - printk(KERN_WARNING "security: avtab: invalid type or class\n"); + printk(KERN_WARNING "SELinux: avtab: invalid type or class\n"); return -1; } @@ -435,19 +435,19 @@ int avtab_read_item(struct avtab *a, void *fp, struct policydb *pol, } if (!set || set > 1) { printk(KERN_WARNING - "security: avtab: more than one specifier\n"); + "SELinux: avtab: more than one specifier\n"); return -1; } rc = next_entry(buf32, fp, sizeof(u32)); if (rc < 0) { - printk("security: avtab: truncated entry\n"); + printk("SELinux: avtab: truncated entry\n"); return -1; } datum.data = le32_to_cpu(*buf32); if ((key.specified & AVTAB_TYPE) && !policydb_type_isvalid(pol, datum.data)) { - printk(KERN_WARNING "security: avtab: invalid type\n"); + printk(KERN_WARNING "SELinux: avtab: invalid type\n"); return -1; } return insertf(a, &key, &datum, p); @@ -468,12 +468,12 @@ int avtab_read(struct avtab *a, void *fp, struct policydb *pol) rc = next_entry(buf, fp, sizeof(u32)); if (rc < 0) { - printk(KERN_ERR "security: avtab: truncated table\n"); + printk(KERN_ERR "SELinux: avtab: truncated table\n"); goto bad; } nel = le32_to_cpu(buf[0]); if (!nel) { - printk(KERN_ERR "security: avtab: table is empty\n"); + printk(KERN_ERR "SELinux: avtab: table is empty\n"); rc = -EINVAL; goto bad; } @@ -486,9 +486,9 @@ int avtab_read(struct avtab *a, void *fp, struct policydb *pol) rc = avtab_read_item(a, fp, pol, avtab_insertf, NULL); if (rc) { if (rc == -ENOMEM) - printk(KERN_ERR "security: avtab: out of memory\n"); + printk(KERN_ERR "SELinux: avtab: out of memory\n"); else if (rc == -EEXIST) - printk(KERN_ERR "security: avtab: duplicate entry\n"); + printk(KERN_ERR "SELinux: avtab: duplicate entry\n"); else rc = -EINVAL; goto bad; diff --git a/security/selinux/ss/conditional.c b/security/selinux/ss/conditional.c index 50ad85d4b77c..a996cf1d378a 100644 --- a/security/selinux/ss/conditional.c +++ b/security/selinux/ss/conditional.c @@ -96,7 +96,7 @@ int evaluate_cond_node(struct policydb *p, struct cond_node *node) if (new_state != node->cur_state) { node->cur_state = new_state; if (new_state == -1) - printk(KERN_ERR "security: expression result was undefined - disabling all rules.\n"); + printk(KERN_ERR "SELinux: expression result was undefined - disabling all rules.\n"); /* turn the rules on or off */ for (cur = node->true_list; cur != NULL; cur = cur->next) { if (new_state <= 0) { @@ -276,7 +276,7 @@ static int cond_insertf(struct avtab *a, struct avtab_key *k, struct avtab_datum */ if (k->specified & AVTAB_TYPE) { if (avtab_search(&p->te_avtab, k)) { - printk("security: type rule already exists outside of a conditional."); + printk("SELinux: type rule already exists outside of a conditional."); goto err; } /* @@ -291,7 +291,7 @@ static int cond_insertf(struct avtab *a, struct avtab_key *k, struct avtab_datum node_ptr = avtab_search_node(&p->te_cond_avtab, k); if (node_ptr) { if (avtab_search_node_next(node_ptr, k->specified)) { - printk("security: too many conflicting type rules."); + printk("SELinux: too many conflicting type rules."); goto err; } found = 0; @@ -302,13 +302,13 @@ static int cond_insertf(struct avtab *a, struct avtab_key *k, struct avtab_datum } } if (!found) { - printk("security: conflicting type rules.\n"); + printk("SELinux: conflicting type rules.\n"); goto err; } } } else { if (avtab_search(&p->te_cond_avtab, k)) { - printk("security: conflicting type rules when adding type rule for true.\n"); + printk("SELinux: conflicting type rules when adding type rule for true.\n"); goto err; } } @@ -316,7 +316,7 @@ static int cond_insertf(struct avtab *a, struct avtab_key *k, struct avtab_datum node_ptr = avtab_insert_nonunique(&p->te_cond_avtab, k, d); if (!node_ptr) { - printk("security: could not insert rule."); + printk("SELinux: could not insert rule."); goto err; } @@ -376,12 +376,12 @@ static int cond_read_av_list(struct policydb *p, void *fp, struct cond_av_list * static int expr_isvalid(struct policydb *p, struct cond_expr *expr) { if (expr->expr_type <= 0 || expr->expr_type > COND_LAST) { - printk("security: conditional expressions uses unknown operator.\n"); + printk("SELinux: conditional expressions uses unknown operator.\n"); return 0; } if (expr->bool > p->p_bools.nprim) { - printk("security: conditional expressions uses unknown bool.\n"); + printk("SELinux: conditional expressions uses unknown bool.\n"); return 0; } return 1; diff --git a/security/selinux/ss/ebitmap.c b/security/selinux/ss/ebitmap.c index 920b5e36a1af..e499af474b35 100644 --- a/security/selinux/ss/ebitmap.c +++ b/security/selinux/ss/ebitmap.c @@ -364,7 +364,7 @@ int ebitmap_read(struct ebitmap *e, void *fp) count = le32_to_cpu(buf[2]); if (mapunit != sizeof(u64) * 8) { - printk(KERN_ERR "security: ebitmap: map size %u does not " + printk(KERN_ERR "SELinux: ebitmap: map size %u does not " "match my size %Zd (high bit was %d)\n", mapunit, sizeof(u64) * 8, e->highbit); goto bad; @@ -382,19 +382,19 @@ int ebitmap_read(struct ebitmap *e, void *fp) for (i = 0; i < count; i++) { rc = next_entry(&startbit, fp, sizeof(u32)); if (rc < 0) { - printk(KERN_ERR "security: ebitmap: truncated map\n"); + printk(KERN_ERR "SELinux: ebitmap: truncated map\n"); goto bad; } startbit = le32_to_cpu(startbit); if (startbit & (mapunit - 1)) { - printk(KERN_ERR "security: ebitmap start bit (%d) is " + printk(KERN_ERR "SELinux: ebitmap start bit (%d) is " "not a multiple of the map unit size (%u)\n", startbit, mapunit); goto bad; } if (startbit > e->highbit - mapunit) { - printk(KERN_ERR "security: ebitmap start bit (%d) is " + printk(KERN_ERR "SELinux: ebitmap start bit (%d) is " "beyond the end of the bitmap (%u)\n", startbit, (e->highbit - mapunit)); goto bad; @@ -405,7 +405,7 @@ int ebitmap_read(struct ebitmap *e, void *fp) tmp = kzalloc(sizeof(*tmp), GFP_KERNEL); if (!tmp) { printk(KERN_ERR - "security: ebitmap: out of memory\n"); + "SELinux: ebitmap: out of memory\n"); rc = -ENOMEM; goto bad; } @@ -418,7 +418,7 @@ int ebitmap_read(struct ebitmap *e, void *fp) } n = tmp; } else if (startbit <= n->startbit) { - printk(KERN_ERR "security: ebitmap: start bit %d" + printk(KERN_ERR "SELinux: ebitmap: start bit %d" " comes after start bit %d\n", startbit, n->startbit); goto bad; @@ -426,7 +426,7 @@ int ebitmap_read(struct ebitmap *e, void *fp) rc = next_entry(&map, fp, sizeof(u64)); if (rc < 0) { - printk(KERN_ERR "security: ebitmap: truncated map\n"); + printk(KERN_ERR "SELinux: ebitmap: truncated map\n"); goto bad; } map = le64_to_cpu(map); diff --git a/security/selinux/ss/policydb.c b/security/selinux/ss/policydb.c index bd7d6a00342d..6bdb0ff6a927 100644 --- a/security/selinux/ss/policydb.c +++ b/security/selinux/ss/policydb.c @@ -111,6 +111,11 @@ static struct policydb_compat_info policydb_compat[] = { .version = POLICYDB_VERSION_POLCAP, .sym_num = SYM_NUM, .ocon_num = OCON_NUM, + }, + { + .version = POLICYDB_VERSION_PERMISSIVE, + .sym_num = SYM_NUM, + .ocon_num = OCON_NUM, } }; @@ -194,6 +199,7 @@ static int policydb_init(struct policydb *p) goto out_free_symtab; ebitmap_init(&p->policycaps); + ebitmap_init(&p->permissive_map); out: return rc; @@ -401,14 +407,14 @@ static int policydb_index_others(struct policydb *p) { int i, rc = 0; - printk(KERN_DEBUG "security: %d users, %d roles, %d types, %d bools", + printk(KERN_DEBUG "SELinux: %d users, %d roles, %d types, %d bools", p->p_users.nprim, p->p_roles.nprim, p->p_types.nprim, p->p_bools.nprim); if (selinux_mls_enabled) printk(", %d sens, %d cats", p->p_levels.nprim, p->p_cats.nprim); printk("\n"); - printk(KERN_DEBUG "security: %d classes, %d rules\n", + printk(KERN_DEBUG "SELinux: %d classes, %d rules\n", p->p_classes.nprim, p->te_avtab.nel); #ifdef DEBUG_HASHES @@ -687,6 +693,7 @@ void policydb_destroy(struct policydb *p) kfree(p->type_attr_map); kfree(p->undefined_perms); ebitmap_destroy(&p->policycaps); + ebitmap_destroy(&p->permissive_map); return; } @@ -702,20 +709,20 @@ int policydb_load_isids(struct policydb *p, struct sidtab *s) rc = sidtab_init(s); if (rc) { - printk(KERN_ERR "security: out of memory on SID table init\n"); + printk(KERN_ERR "SELinux: out of memory on SID table init\n"); goto out; } head = p->ocontexts[OCON_ISID]; for (c = head; c; c = c->next) { if (!c->context[0].user) { - printk(KERN_ERR "security: SID %s was never " + printk(KERN_ERR "SELinux: SID %s was never " "defined.\n", c->u.name); rc = -EINVAL; goto out; } if (sidtab_insert(s, c->sid[0], &c->context[0])) { - printk(KERN_ERR "security: unable to load initial " + printk(KERN_ERR "SELinux: unable to load initial " "SID %s.\n", c->u.name); rc = -EINVAL; goto out; @@ -809,13 +816,13 @@ static int mls_read_range_helper(struct mls_range *r, void *fp) items = le32_to_cpu(buf[0]); if (items > ARRAY_SIZE(buf)) { - printk(KERN_ERR "security: mls: range overflow\n"); + printk(KERN_ERR "SELinux: mls: range overflow\n"); rc = -EINVAL; goto out; } rc = next_entry(buf, fp, sizeof(u32) * items); if (rc < 0) { - printk(KERN_ERR "security: mls: truncated range\n"); + printk(KERN_ERR "SELinux: mls: truncated range\n"); goto out; } r->level[0].sens = le32_to_cpu(buf[0]); @@ -826,21 +833,21 @@ static int mls_read_range_helper(struct mls_range *r, void *fp) rc = ebitmap_read(&r->level[0].cat, fp); if (rc) { - printk(KERN_ERR "security: mls: error reading low " + printk(KERN_ERR "SELinux: mls: error reading low " "categories\n"); goto out; } if (items > 1) { rc = ebitmap_read(&r->level[1].cat, fp); if (rc) { - printk(KERN_ERR "security: mls: error reading high " + printk(KERN_ERR "SELinux: mls: error reading high " "categories\n"); goto bad_high; } } else { rc = ebitmap_cpy(&r->level[1].cat, &r->level[0].cat); if (rc) { - printk(KERN_ERR "security: mls: out of memory\n"); + printk(KERN_ERR "SELinux: mls: out of memory\n"); goto bad_high; } } @@ -866,7 +873,7 @@ static int context_read_and_validate(struct context *c, rc = next_entry(buf, fp, sizeof buf); if (rc < 0) { - printk(KERN_ERR "security: context truncated\n"); + printk(KERN_ERR "SELinux: context truncated\n"); goto out; } c->user = le32_to_cpu(buf[0]); @@ -874,7 +881,7 @@ static int context_read_and_validate(struct context *c, c->type = le32_to_cpu(buf[2]); if (p->policyvers >= POLICYDB_VERSION_MLS) { if (mls_read_range_helper(&c->range, fp)) { - printk(KERN_ERR "security: error reading MLS range of " + printk(KERN_ERR "SELinux: error reading MLS range of " "context\n"); rc = -EINVAL; goto out; @@ -882,7 +889,7 @@ static int context_read_and_validate(struct context *c, } if (!policydb_context_isvalid(p, c)) { - printk(KERN_ERR "security: invalid security context\n"); + printk(KERN_ERR "SELinux: invalid security context\n"); context_destroy(c); rc = -EINVAL; } @@ -1128,7 +1135,7 @@ static int class_read(struct policydb *p, struct hashtab *h, void *fp) cladatum->comdatum = hashtab_search(p->p_commons.table, cladatum->comkey); if (!cladatum->comdatum) { - printk(KERN_ERR "security: unknown common %s\n", + printk(KERN_ERR "SELinux: unknown common %s\n", cladatum->comkey); rc = -EINVAL; goto bad; @@ -1283,13 +1290,13 @@ static int mls_read_level(struct mls_level *lp, void *fp) rc = next_entry(buf, fp, sizeof buf); if (rc < 0) { - printk(KERN_ERR "security: mls: truncated level\n"); + printk(KERN_ERR "SELinux: mls: truncated level\n"); goto bad; } lp->sens = le32_to_cpu(buf[0]); if (ebitmap_read(&lp->cat, fp)) { - printk(KERN_ERR "security: mls: error reading level " + printk(KERN_ERR "SELinux: mls: error reading level " "categories\n"); goto bad; } @@ -1491,7 +1498,7 @@ int policydb_read(struct policydb *p, void *fp) goto bad; if (le32_to_cpu(buf[0]) != POLICYDB_MAGIC) { - printk(KERN_ERR "security: policydb magic number 0x%x does " + printk(KERN_ERR "SELinux: policydb magic number 0x%x does " "not match expected magic number 0x%x\n", le32_to_cpu(buf[0]), POLICYDB_MAGIC); goto bad; @@ -1499,27 +1506,27 @@ int policydb_read(struct policydb *p, void *fp) len = le32_to_cpu(buf[1]); if (len != strlen(POLICYDB_STRING)) { - printk(KERN_ERR "security: policydb string length %d does not " + printk(KERN_ERR "SELinux: policydb string length %d does not " "match expected length %Zu\n", len, strlen(POLICYDB_STRING)); goto bad; } policydb_str = kmalloc(len + 1,GFP_KERNEL); if (!policydb_str) { - printk(KERN_ERR "security: unable to allocate memory for policydb " + printk(KERN_ERR "SELinux: unable to allocate memory for policydb " "string of length %d\n", len); rc = -ENOMEM; goto bad; } rc = next_entry(policydb_str, fp, len); if (rc < 0) { - printk(KERN_ERR "security: truncated policydb string identifier\n"); + printk(KERN_ERR "SELinux: truncated policydb string identifier\n"); kfree(policydb_str); goto bad; } policydb_str[len] = 0; if (strcmp(policydb_str, POLICYDB_STRING)) { - printk(KERN_ERR "security: policydb string %s does not match " + printk(KERN_ERR "SELinux: policydb string %s does not match " "my string %s\n", policydb_str, POLICYDB_STRING); kfree(policydb_str); goto bad; @@ -1536,7 +1543,7 @@ int policydb_read(struct policydb *p, void *fp) p->policyvers = le32_to_cpu(buf[0]); if (p->policyvers < POLICYDB_VERSION_MIN || p->policyvers > POLICYDB_VERSION_MAX) { - printk(KERN_ERR "security: policydb version %d does not match " + printk(KERN_ERR "SELinux: policydb version %d does not match " "my version range %d-%d\n", le32_to_cpu(buf[0]), POLICYDB_VERSION_MIN, POLICYDB_VERSION_MAX); goto bad; @@ -1570,16 +1577,20 @@ int policydb_read(struct policydb *p, void *fp) ebitmap_read(&p->policycaps, fp) != 0) goto bad; + if (p->policyvers >= POLICYDB_VERSION_PERMISSIVE && + ebitmap_read(&p->permissive_map, fp) != 0) + goto bad; + info = policydb_lookup_compat(p->policyvers); if (!info) { - printk(KERN_ERR "security: unable to find policy compat info " + printk(KERN_ERR "SELinux: unable to find policy compat info " "for version %d\n", p->policyvers); goto bad; } if (le32_to_cpu(buf[2]) != info->sym_num || le32_to_cpu(buf[3]) != info->ocon_num) { - printk(KERN_ERR "security: policydb table sizes (%d,%d) do " + printk(KERN_ERR "SELinux: policydb table sizes (%d,%d) do " "not match mine (%d,%d)\n", le32_to_cpu(buf[2]), le32_to_cpu(buf[3]), info->sym_num, info->ocon_num); @@ -1823,7 +1834,7 @@ int policydb_read(struct policydb *p, void *fp) for (genfs_p = NULL, genfs = p->genfs; genfs; genfs_p = genfs, genfs = genfs->next) { if (strcmp(newgenfs->fstype, genfs->fstype) == 0) { - printk(KERN_ERR "security: dup genfs " + printk(KERN_ERR "SELinux: dup genfs " "fstype %s\n", newgenfs->fstype); kfree(newgenfs->fstype); kfree(newgenfs); @@ -1873,7 +1884,7 @@ int policydb_read(struct policydb *p, void *fp) if (!strcmp(newc->u.name, c->u.name) && (!c->v.sclass || !newc->v.sclass || newc->v.sclass == c->v.sclass)) { - printk(KERN_ERR "security: dup genfs " + printk(KERN_ERR "SELinux: dup genfs " "entry (%s,%s)\n", newgenfs->fstype, c->u.name); goto bad_newc; @@ -1931,7 +1942,7 @@ int policydb_read(struct policydb *p, void *fp) if (rc) goto bad; if (!mls_range_isvalid(p, &rt->target_range)) { - printk(KERN_WARNING "security: rangetrans: invalid range\n"); + printk(KERN_WARNING "SELinux: rangetrans: invalid range\n"); goto bad; } lrt = rt; diff --git a/security/selinux/ss/policydb.h b/security/selinux/ss/policydb.h index c4ce996e202c..ba593a3da877 100644 --- a/security/selinux/ss/policydb.h +++ b/security/selinux/ss/policydb.h @@ -243,6 +243,8 @@ struct policydb { struct ebitmap policycaps; + struct ebitmap permissive_map; + unsigned int policyvers; unsigned int reject_unknown : 1; diff --git a/security/selinux/ss/services.c b/security/selinux/ss/services.c index 3f2bad28ee7b..d75050819b06 100644 --- a/security/selinux/ss/services.c +++ b/security/selinux/ss/services.c @@ -40,6 +40,7 @@ #include <linux/sched.h> #include <linux/audit.h> #include <linux/mutex.h> +#include <linux/selinux.h> #include <net/netlabel.h> #include "flask.h" @@ -61,6 +62,7 @@ extern void selnl_notify_policyload(u32 seqno); unsigned int policydb_loaded_version; int selinux_policycap_netpeer; +int selinux_policycap_openperm; /* * This is declared in avc.c @@ -412,10 +414,35 @@ static int context_struct_compute_av(struct context *scontext, return 0; inval_class: - printk(KERN_ERR "%s: unrecognized class %d\n", __FUNCTION__, tclass); + printk(KERN_ERR "%s: unrecognized class %d\n", __func__, tclass); return -EINVAL; } +/* + * Given a sid find if the type has the permissive flag set + */ +int security_permissive_sid(u32 sid) +{ + struct context *context; + u32 type; + int rc; + + POLICY_RDLOCK; + + context = sidtab_search(&sidtab, sid); + BUG_ON(!context); + + type = context->type; + /* + * we are intentionally using type here, not type-1, the 0th bit may + * someday indicate that we are globally setting permissive in policy. + */ + rc = ebitmap_get_bit(&policydb.permissive_map, type); + + POLICY_RDUNLOCK; + return rc; +} + static int security_validtrans_handle_fail(struct context *ocontext, struct context *ncontext, struct context *tcontext, @@ -1096,7 +1123,7 @@ static int validate_classes(struct policydb *p) continue; if (i > p->p_classes.nprim) { printk(KERN_INFO - "security: class %s not defined in policy\n", + "SELinux: class %s not defined in policy\n", def_class); if (p->reject_unknown) return -EINVAL; @@ -1107,7 +1134,7 @@ static int validate_classes(struct policydb *p) pol_class = p->p_class_val_to_name[i-1]; if (strcmp(pol_class, def_class)) { printk(KERN_ERR - "security: class %d is incorrect, found %s but should be %s\n", + "SELinux: class %d is incorrect, found %s but should be %s\n", i, pol_class, def_class); return -EINVAL; } @@ -1125,7 +1152,7 @@ static int validate_classes(struct policydb *p) nprim = 1 << (perms->nprim - 1); if (perm_val > nprim) { printk(KERN_INFO - "security: permission %s in class %s not defined in policy\n", + "SELinux: permission %s in class %s not defined in policy\n", def_perm, pol_class); if (p->reject_unknown) return -EINVAL; @@ -1136,14 +1163,14 @@ static int validate_classes(struct policydb *p) perdatum = hashtab_search(perms->table, def_perm); if (perdatum == NULL) { printk(KERN_ERR - "security: permission %s in class %s not found in policy, bad policy\n", + "SELinux: permission %s in class %s not found in policy, bad policy\n", def_perm, pol_class); return -EINVAL; } pol_val = 1 << (perdatum->value - 1); if (pol_val != perm_val) { printk(KERN_ERR - "security: permission %s in class %s has incorrect value\n", + "SELinux: permission %s in class %s has incorrect value\n", def_perm, pol_class); return -EINVAL; } @@ -1157,7 +1184,7 @@ static int validate_classes(struct policydb *p) BUG_ON(!cladatum); if (!cladatum->comdatum) { printk(KERN_ERR - "security: class %s should have an inherits clause but does not\n", + "SELinux: class %s should have an inherits clause but does not\n", pol_class); return -EINVAL; } @@ -1172,7 +1199,7 @@ static int validate_classes(struct policydb *p) def_perm = kdefs->av_inherit[i].common_pts[j]; if (j >= perms->nprim) { printk(KERN_INFO - "security: permission %s in class %s not defined in policy\n", + "SELinux: permission %s in class %s not defined in policy\n", def_perm, pol_class); if (p->reject_unknown) return -EINVAL; @@ -1183,13 +1210,13 @@ static int validate_classes(struct policydb *p) perdatum = hashtab_search(perms->table, def_perm); if (perdatum == NULL) { printk(KERN_ERR - "security: permission %s in class %s not found in policy, bad policy\n", + "SELinux: permission %s in class %s not found in policy, bad policy\n", def_perm, pol_class); return -EINVAL; } if (perdatum->value != j + 1) { printk(KERN_ERR - "security: permission %s in class %s has incorrect value\n", + "SELinux: permission %s in class %s has incorrect value\n", def_perm, pol_class); return -EINVAL; } @@ -1219,7 +1246,7 @@ static inline int convert_context_handle_invalid_context(struct context *context u32 len; context_struct_to_string(context, &s, &len); - printk(KERN_ERR "security: context %s is invalid\n", s); + printk(KERN_ERR "SELinux: context %s is invalid\n", s); kfree(s); } return rc; @@ -1299,7 +1326,7 @@ out: bad: context_struct_to_string(&oldc, &s, &len); context_destroy(&oldc); - printk(KERN_ERR "security: invalidating context %s\n", s); + printk(KERN_ERR "SELinux: invalidating context %s\n", s); kfree(s); goto out; } @@ -1308,6 +1335,8 @@ static void security_load_policycaps(void) { selinux_policycap_netpeer = ebitmap_get_bit(&policydb.policycaps, POLICYDB_CAPABILITY_NETPEER); + selinux_policycap_openperm = ebitmap_get_bit(&policydb.policycaps, + POLICYDB_CAPABILITY_OPENPERM); } extern void selinux_complete_init(void); @@ -1350,7 +1379,7 @@ int security_load_policy(void *data, size_t len) /* Verify that the kernel defined classes are correct. */ if (validate_classes(&policydb)) { printk(KERN_ERR - "security: the definition of a class is incorrect\n"); + "SELinux: the definition of a class is incorrect\n"); LOAD_UNLOCK; sidtab_destroy(&sidtab); policydb_destroy(&policydb); @@ -1384,14 +1413,14 @@ int security_load_policy(void *data, size_t len) /* Verify that the kernel defined classes are correct. */ if (validate_classes(&newpolicydb)) { printk(KERN_ERR - "security: the definition of a class is incorrect\n"); + "SELinux: the definition of a class is incorrect\n"); rc = -EINVAL; goto err; } rc = security_preserve_bools(&newpolicydb); if (rc) { - printk(KERN_ERR "security: unable to preserve booleans\n"); + printk(KERN_ERR "SELinux: unable to preserve booleans\n"); goto err; } @@ -1443,17 +1472,11 @@ err: /** * security_port_sid - Obtain the SID for a port. - * @domain: communication domain aka address family - * @type: socket type * @protocol: protocol number * @port: port number * @out_sid: security identifier */ -int security_port_sid(u16 domain, - u16 type, - u8 protocol, - u16 port, - u32 *out_sid) +int security_port_sid(u8 protocol, u16 port, u32 *out_sid) { struct ocontext *c; int rc = 0; @@ -2203,7 +2226,7 @@ int security_get_permissions(char *class, char ***perms, int *nperms) match = hashtab_search(policydb.p_classes.table, class); if (!match) { printk(KERN_ERR "%s: unrecognized class %s\n", - __FUNCTION__, class); + __func__, class); rc = -EINVAL; goto out; } |