| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
|
|
Conflicts:
include/asm-powerpc/unistd.h
include/asm-sparc/unistd.h
include/asm-sparc64/unistd.h
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
|
|
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Acked-by: Paul Mackerras <paulus@samba.org>
Cc: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
|
|
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
|
|
We have a case where __get_user and __put_user can validly be used
on kernel addresses in interrupt context - namely, the alignment
exception handler, as our get/put_unaligned just do a single access
and rely on the alignment exception handler to fix things up in the
rare cases where the cpu can't handle it in hardware. Thus we can
get alignment exceptions in the network stack at interrupt level.
The alignment exception handler does a __get_user to read the
instruction and blows up in might_sleep().
Since a __get_user on a kernel address won't actually ever sleep,
this makes the might_sleep conditional on the address being less
than PAGE_OFFSET.
Signed-off-by: Paul Mackerras <paulus@samba.org>
|
|
Add an nid member to the spu structure, and store the numa id of the spu there
on creation.
Signed-off-by: Arnd Bergmann <arnd.bergmann@de.ibm.com>
Cc: Paul Mackerras <paulus@samba.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
Change of_node_to_nid() to traverse the device tree, looking for a numa id.
Cell uses this to assign ids to SPUs, which are children of the CPU node.
Existing users of of_node_to_nid() are altered to use of_node_to_nid_single(),
which doesn't do the traversal.
Export an attach_sysdev_to_node() function, allowing system devices (eg.
SPUs) to link themselves into the numa topology in sysfs.
Signed-off-by: Arnd Bergmann <arnd.bergmann@de.ibm.com>
Cc: Paul Mackerras <paulus@samba.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
|
|
Add a cputable entry for the POWER6 processor.
The SIHV and SIPR bits in the mmcra have moved in POWER6, so disable
support for that until oprofile is fixed.
Also tell firmware that we know about POWER6.
Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
|
|
These aren't needed by glibc or klibc, and they're broken in some cases
anyway. The uClibc folks are apparently switching over to stop using
them too (now that we agreed that they should be dropped, at least).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
|
|
|
|
Wire up *at syscalls.
This patch has been tested on ppc64 (using glibc's testsuite, both 32bit
and 64bit), and compile-tested for ppc32 (I have currently no ppc32 system
available, but I expect no problems).
Signed-off-by: Andreas Schwab <schwab@suse.de>
Signed-off-by: Paul Mackerras <paulus@samba.org>
|
|
Some people report that we die on some Macs when we are expecting to
catch machine checks after poking at some random I/O address. I'd seen
it happen on my dual G4 with serial ports until we fixed those to use
OF, but now other users are reporting it with i8042.
This expands the use of check_legacy_ioport() to avoid that situation
even on 32-bit kernels.
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
|
|
At present, ARCH=powerpc kernels can waste considerable space in
pagetables when making large hugepage mappings. Hugepage PTEs go in
PMD pages, but each PMD page maps 256M and so contains only 16
hugepage PTEs (128 bytes of data), but takes up a 1024 byte
allocation. With CONFIG_PPC_64K_PAGES enabled (64k base page size),
the situation is worse. Now hugepage PTEs are at the PTE page level
(also mapping 256M), so we store 16 hugepage PTEs in a 64k allocation.
The PowerPC MMU already means that any 256M region is either all
hugepage, or all normal pages. Thus, with some care, we can use a
different allocation for the hugepage PTE tables and only allocate the
128 bytes necessary.
Signed-off-by: Paul Mackerras <paulus@samba.org>
|
|
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
|
|
sys_splice() moves data to/from pipes with a file input/output. sys_vmsplice()
moves data to a pipe, with the input being a user address range instead.
This uses an approach suggested by Linus, where we can hold partial ranges
inside the pages[] map. Hopefully this will be useful for network
receive support as well.
Signed-off-by: Jens Axboe <axboe@suse.de>
|
|
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
|
|
machine_is() was always returning 0 when used in a module, because
we weren't exporting the machine definitions. This was why sound
wasn't working on powermacs when CONFIG_SND_POWERMAC=m. Original
fix from Ben Herrenschmidt, further fixed by me.
Signed-off-by: Paul Mackerras <paulus@samba.org>
|
|
Some devices don't support full 32-bit DMA address space, which we currently
assume. Add the required mask-passing to the IOMMU allocators.
Signed-off-by: Olof Johansson <olof@lixom.net>
Signed-off-by: Paul Mackerras <paulus@samba.org>
|
|
* git://git.kernel.org/pub/scm/linux/kernel/git/paulus/powerpc-merge:
powerpc: Use correct sequence for putting CPU into nap mode
[PATCH] spufs: fix context-switch decrementer code
[PATCH] powerpc32: Set cpu explicitly in kernel compiles
[PATCH] powerpc/pseries: bugfix: balance calls to pci_device_put
[PATCH] powerpc: Fix machine detection in prom_init.c
[PATCH] ppc32: Fix string comparing in platform_notify_map
[PATCH] powerpc: Avoid __initcall warnings
[PATCH] powerpc: Ensure runlatch is off in the idle loop
powerpc: Fix CHRP booting - needs a define_machine call
powerpc: iSeries has only 256 IRQs
|
|
We weren't using the recommended sequence for putting the CPU into
nap mode. When I changed the idle loop, for some reason 7447A cpus
started hanging when we put them into nap mode. Changing to the
recommended sequence fixes that.
The complexity here is that the recommended sequence is a loop that
keeps putting the cpu back into nap mode. Clearly we need some way
to break out of the loop when an interrupt (external interrupt,
decrementer, performance monitor) occurs. Here we use a bit in
the thread_info struct to indicate that we need this, and the exception
entry code notices this and arranges for the exception to return
to the value in the link register, thus breaking out of the loop.
We use a new `local_flags' field in the thread_info which we can
alter without needing to use an atomic update sequence.
The PPC970 has the same recommended sequence, so we do the same thing
there too.
This also fixes a bug in the kernel stack overflow handling code on
32-bit, since it was causing a value that we needed in a register to
get trashed.
Signed-off-by: Paul Mackerras <paulus@samba.org>
|
|
Basically an in-kernel implementation of tee, which uses splice and the
pipe buffers as an intelligent way to pass data around by reference.
Where the user space tee consumes the input and produces a stdout and
file output, this syscall merely duplicates the data inside a pipe to
another pipe. No data is copied, the output just grabs a reference to the
input pipe data.
Signed-off-by: Jens Axboe <axboe@suse.de>
|
|
Current implementations define NODES_SHIFT in include/asm-xxx/numnodes.h for
each arch. Its definition is sometimes configurable. Indeed, ia64 defines 5
NODES_SHIFT values in the current git tree. But it looks a bit messy.
SGI-SN2(ia64) system requires 1024 nodes, and the number of nodes already has
been changeable by config. Suitable node's number may be changed in the
future even if it is other architecture. So, I wrote configurable node's
number.
This patch set defines just default value for each arch which needs multi
nodes except ia64. But, it is easy to change to configurable if necessary.
On ia64 the number of nodes can be already configured in generic ia64 and SN2
config. But, NODES_SHIFT is defined for DIG64 and HP'S machine too. So, I
changed it so that all platforms can be configured via CONFIG_NODES_SHIFT. It
would be simpler.
See also: http://marc.theaimsgroup.com/?l=linux-kernel&m=114358010523896&w=2
Signed-off-by: Yasunori Goto <y-goto@jp.fujitsu.com>
Cc: Hirokazu Takata <takata@linux-m32r.org>
Cc: "Luck, Tony" <tony.luck@intel.com>
Cc: Andi Kleen <ak@muc.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Cc: Richard Henderson <rth@twiddle.net>
Cc: Kyle McMartin <kyle@mcmartin.ca>
Cc: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: Jack Steiner <steiner@sgi.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
The iSeries Hypervisor only allows us to specify IRQ numbers up to 255 (it
has a u8 field to pass it in). This patch allows platforms to specify a
maximum to the virtual IRQ numbers we will use and has iSeries set that
to 255. If not set, the maximum is NR_IRQS - 1 (as before).
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
|
|
This patch removes unnecessary exports, marks functions as static when
possible, and simplifies some list-related code.
Signed-off-by: Nathan Fontenot <nfont@austin.ibm.com>
Signed-off-by: Linas Vepstas <linas@austin.ibm.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
|
|
This extends the HCALL interface for InfiniBand usage. I've
made the patch against the linux-2.6 git tree and Segher's patch:
[PATCH] Change H_StudlyCaps to H_SHOUTING_CAPS
We moved this into the common powerpc code based on comments we
got after posting the first eHCA InfiniBand device driver patch.
Signed-off-by: Heiko j Schick <schickhj@de.ibm.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
|
|
Also cleans up some nearby whitespace problems.
Signed-off-by: Segher Boessenkool <segher@kernel.crashing.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
|
|
* master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6:
[NET]: Allow skb headroom to be overridden
[TCP]: Kill unused extern decl for tcp_v4_hash_connecting()
[NET]: add SO_RCVBUF comment
[NET]: Deinline some larger functions from netdevice.h
[DCCP]: Use NULL for pointers, comfort sparse.
[DECNET]: Fix refcount
|
|
Previously we added NET_IP_ALIGN so an architecture can override the
padding done to align headers. The next step is to allow the skb
headroom to be overridden.
We currently always reserve 16 bytes to grow into, meaning all DMAs
start 16 bytes into a cacheline. On ppc64 we really want DMA writes to
start on a cacheline boundary, so we increase that headroom to one
cacheline.
Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
This adds support for the sys_splice system call. Using a pipe as a
transport, it can connect to files or sockets (latter as output only).
From the splice.c comments:
"splice": joining two ropes together by interweaving their strands.
This is the "extended pipe" functionality, where a pipe is used as
an arbitrary in-memory buffer. Think of a pipe as a small kernel
buffer that you can use to transfer data from one end to the other.
The traditional unix read/write is extended with a "splice()" operation
that transfers data buffers to or from a pipe buffer.
Named by Larry McVoy, original implementation from Linus, extended by
Jens to support splicing to files and fixing the initial implementation
bugs.
Signed-off-by: Jens Axboe <axboe@suse.de>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
Remove oprofile spinlock backtrace code now we have proper calltrace
support. Also make MMCRA sihv and sipr bits a variable since they may
change in future cpus. Finally, MMCRA should be a 64bit quantity.
Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
|
|
Add oprofile calltrace support to powerpc. Disable spinlock backtracing
now we can use calltrace info.
(Updated to work on both 32bit and 64bit by me).
Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
|
|
for_each_cpu() actually iterates across all possible CPUs. We've had mistakes
in the past where people were using for_each_cpu() where they should have been
iterating across only online or present CPUs. This is inefficient and
possibly buggy.
We're renaming for_each_cpu() to for_each_possible_cpu() to avoid this in the
future.
This patch replaces for_each_cpu with for_each_possible_cpu.
Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
|
|
|
|
This removes statically assigned platform numbers and reworks the
powerpc platform probe code to use a better mechanism. With this,
board support files can simply declare a new machine type with a
macro, and implement a probe() function that uses the flattened
device-tree to detect if they apply for a given machine.
We now have a machine_is() macro that replaces the comparisons of
_machine with the various PLATFORM_* constants. This commit also
changes various drivers to use the new macro instead of looking at
_machine.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
|
|
There are at least 14 different implementations of WARN() in the tree already.
The build fails all over the place.
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <michael@ellerman.id.au>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
|
|
We used to assume that a DMA mapping request with a NULL dev was for
ISA DMA. This assumption was broken at some point. Now we explicitly
pass the detected ISA PCI device in the floppy setup.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Paul Mackerras <paulus@samba.org>
|
|
These are some updates from both Ryan and Arnd for the hvc_console
driver:
The main point is to enable the inclusion of a console driver
for rtas, which is currrently needed for the cell platform.
Also shuffle around some data-type declarations and moves some
functions out of include/asm-ppc64/hvconsole.h and into a new
drivers/char/hvc_console.h file.
Signed-off-by: "Ryan S. Arnold" <rsa@us.ibm.com>
Signed-off-by: Arnd Bergmann <abergman@de.ibm.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
|
|
We need to export ppc64_firmware_features for modules. Before we do that
I think we should probably rename it to powerpc_firmware_features.
Signed-off-by: Michael Ellerman <michael@ellerman.id.au>
Signed-off-by: Paul Mackerras <paulus@samba.org>
|
|
Export validate_sp so we can use it in the oprofile calltrace code.
Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
|
|
- No one uses op_counter_config.valid, so remove it
- No need to ifdef around function protypes.
Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
|
|
32-bit CHRP machines are now supported only in arch/powerpc, as are
all 64-bit PowerPC processors. This means that we don't use
Open Firmware on any platform in arch/ppc any more.
This makes PReP support a single-platform option like every other
platform support option in arch/ppc now, thus CONFIG_PPC_MULTIPLATFORM
is gone from arch/ppc. CONFIG_PPC_PREP is the option that selects
PReP support and is generally what has replaced
CONFIG_PPC_MULTIPLATFORM within arch/ppc.
_machine is all but dead now, being #defined to 0.
Updated Makefiles, comments and Kconfig options generally to reflect
these changes.
Signed-off-by: Paul Mackerras <paulus@samba.org>
|
|
The kernel's implementation of notifier chains is unsafe. There is no
protection against entries being added to or removed from a chain while the
chain is in use. The issues were discussed in this thread:
http://marc.theaimsgroup.com/?l=linux-kernel&m=113018709002036&w=2
We noticed that notifier chains in the kernel fall into two basic usage
classes:
"Blocking" chains are always called from a process context
and the callout routines are allowed to sleep;
"Atomic" chains can be called from an atomic context and
the callout routines are not allowed to sleep.
We decided to codify this distinction and make it part of the API. Therefore
this set of patches introduces three new, parallel APIs: one for blocking
notifiers, one for atomic notifiers, and one for "raw" notifiers (which is
really just the old API under a new name). New kinds of data structures are
used for the heads of the chains, and new routines are defined for
registration, unregistration, and calling a chain. The three APIs are
explained in include/linux/notifier.h and their implementation is in
kernel/sys.c.
With atomic and blocking chains, the implementation guarantees that the chain
links will not be corrupted and that chain callers will not get messed up by
entries being added or removed. For raw chains the implementation provides no
guarantees at all; users of this API must provide their own protections. (The
idea was that situations may come up where the assumptions of the atomic and
blocking APIs are not appropriate, so it should be possible for users to
handle these things in their own way.)
There are some limitations, which should not be too hard to live with. For
atomic/blocking chains, registration and unregistration must always be done in
a process context since the chain is protected by a mutex/rwsem. Also, a
callout routine for a non-raw chain must not try to register or unregister
entries on its own chain. (This did happen in a couple of places and the code
had to be changed to avoid it.)
Since atomic chains may be called from within an NMI handler, they cannot use
spinlocks for synchronization. Instead we use RCU. The overhead falls almost
entirely in the unregister routine, which is okay since unregistration is much
less frequent that calling a chain.
Here is the list of chains that we adjusted and their classifications. None
of them use the raw API, so for the moment it is only a placeholder.
ATOMIC CHAINS
-------------
arch/i386/kernel/traps.c: i386die_chain
arch/ia64/kernel/traps.c: ia64die_chain
arch/powerpc/kernel/traps.c: powerpc_die_chain
arch/sparc64/kernel/traps.c: sparc64die_chain
arch/x86_64/kernel/traps.c: die_chain
drivers/char/ipmi/ipmi_si_intf.c: xaction_notifier_list
kernel/panic.c: panic_notifier_list
kernel/profile.c: task_free_notifier
net/bluetooth/hci_core.c: hci_notifier
net/ipv4/netfilter/ip_conntrack_core.c: ip_conntrack_chain
net/ipv4/netfilter/ip_conntrack_core.c: ip_conntrack_expect_chain
net/ipv6/addrconf.c: inet6addr_chain
net/netfilter/nf_conntrack_core.c: nf_conntrack_chain
net/netfilter/nf_conntrack_core.c: nf_conntrack_expect_chain
net/netlink/af_netlink.c: netlink_chain
BLOCKING CHAINS
---------------
arch/powerpc/platforms/pseries/reconfig.c: pSeries_reconfig_chain
arch/s390/kernel/process.c: idle_chain
arch/x86_64/kernel/process.c idle_notifier
drivers/base/memory.c: memory_chain
drivers/cpufreq/cpufreq.c cpufreq_policy_notifier_list
drivers/cpufreq/cpufreq.c cpufreq_transition_notifier_list
drivers/macintosh/adb.c: adb_client_list
drivers/macintosh/via-pmu.c sleep_notifier_list
drivers/macintosh/via-pmu68k.c sleep_notifier_list
drivers/macintosh/windfarm_core.c wf_client_list
drivers/usb/core/notify.c usb_notifier_list
drivers/video/fbmem.c fb_notifier_list
kernel/cpu.c cpu_chain
kernel/module.c module_notify_list
kernel/profile.c munmap_notifier
kernel/profile.c task_exit_notifier
kernel/sys.c reboot_notifier_list
net/core/dev.c netdev_chain
net/decnet/dn_dev.c: dnaddr_chain
net/ipv4/devinet.c: inetaddr_chain
It's possible that some of these classifications are wrong. If they are,
please let us know or submit a patch to fix them. Note that any chain that
gets called very frequently should be atomic, because the rwsem read-locking
used for blocking chains is very likely to incur cache misses on SMP systems.
(However, if the chain's callout routines may sleep then the chain cannot be
atomic.)
The patch set was written by Alan Stern and Chandra Seetharaman, incorporating
material written by Keith Owens and suggestions from Paul McKenney and Andrew
Morton.
[jes@sgi.com: restructure the notifier chain initialization macros]
Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Chandra Seetharaman <sekharan@us.ibm.com>
Signed-off-by: Jes Sorensen <jes@sgi.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
- fix: initialize the robust list(s) to NULL in copy_process.
- doc update
- cleanup: rename _inuser to _inatomic
- __user cleanups and other small cleanups
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Arjan van de Ven <arjan@infradead.org>
Cc: Ulrich Drepper <drepper@redhat.com>
Cc: Andi Kleen <ak@muc.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
This patchset provides a new (written from scratch) implementation of robust
futexes, called "lightweight robust futexes". We believe this new
implementation is faster and simpler than the vma-based robust futex solutions
presented before, and we'd like this patchset to be adopted in the upstream
kernel. This is version 1 of the patchset.
Background
----------
What are robust futexes? To answer that, we first need to understand what
futexes are: normal futexes are special types of locks that in the
noncontended case can be acquired/released from userspace without having to
enter the kernel.
A futex is in essence a user-space address, e.g. a 32-bit lock variable
field. If userspace notices contention (the lock is already owned and someone
else wants to grab it too) then the lock is marked with a value that says
"there's a waiter pending", and the sys_futex(FUTEX_WAIT) syscall is used to
wait for the other guy to release it. The kernel creates a 'futex queue'
internally, so that it can later on match up the waiter with the waker -
without them having to know about each other. When the owner thread releases
the futex, it notices (via the variable value) that there were waiter(s)
pending, and does the sys_futex(FUTEX_WAKE) syscall to wake them up. Once all
waiters have taken and released the lock, the futex is again back to
'uncontended' state, and there's no in-kernel state associated with it. The
kernel completely forgets that there ever was a futex at that address. This
method makes futexes very lightweight and scalable.
"Robustness" is about dealing with crashes while holding a lock: if a process
exits prematurely while holding a pthread_mutex_t lock that is also shared
with some other process (e.g. yum segfaults while holding a pthread_mutex_t,
or yum is kill -9-ed), then waiters for that lock need to be notified that the
last owner of the lock exited in some irregular way.
To solve such types of problems, "robust mutex" userspace APIs were created:
pthread_mutex_lock() returns an error value if the owner exits prematurely -
and the new owner can decide whether the data protected by the lock can be
recovered safely.
There is a big conceptual problem with futex based mutexes though: it is the
kernel that destroys the owner task (e.g. due to a SEGFAULT), but the kernel
cannot help with the cleanup: if there is no 'futex queue' (and in most cases
there is none, futexes being fast lightweight locks) then the kernel has no
information to clean up after the held lock! Userspace has no chance to clean
up after the lock either - userspace is the one that crashes, so it has no
opportunity to clean up. Catch-22.
In practice, when e.g. yum is kill -9-ed (or segfaults), a system reboot is
needed to release that futex based lock. This is one of the leading
bugreports against yum.
To solve this problem, 'Robust Futex' patches were created and presented on
lkml: the one written by Todd Kneisel and David Singleton is the most advanced
at the moment. These patches all tried to extend the futex abstraction by
registering futex-based locks in the kernel - and thus give the kernel a
chance to clean up.
E.g. in David Singleton's robust-futex-6.patch, there are 3 new syscall
variants to sys_futex(): FUTEX_REGISTER, FUTEX_DEREGISTER and FUTEX_RECOVER.
The kernel attaches such robust futexes to vmas (via
vma->vm_file->f_mapping->robust_head), and at do_exit() time, all vmas are
searched to see whether they have a robust_head set.
Lots of work went into the vma-based robust-futex patch, and recently it has
improved significantly, but unfortunately it still has two fundamental
problems left:
- they have quite complex locking and race scenarios. The vma-based
patches had been pending for years, but they are still not completely
reliable.
- they have to scan _every_ vma at sys_exit() time, per thread!
The second disadvantage is a real killer: pthread_exit() takes around 1
microsecond on Linux, but with thousands (or tens of thousands) of vmas every
pthread_exit() takes a millisecond or more, also totally destroying the CPU's
L1 and L2 caches!
This is very much noticeable even for normal process sys_exit_group() calls:
the kernel has to do the vma scanning unconditionally! (this is because the
kernel has no knowledge about how many robust futexes there are to be cleaned
up, because a robust futex might have been registered in another task, and the
futex variable might have been simply mmap()-ed into this process's address
space).
This huge overhead forced the creation of CONFIG_FUTEX_ROBUST, but worse than
that: the overhead makes robust futexes impractical for any type of generic
Linux distribution.
So it became clear to us, something had to be done. Last week, when Thomas
Gleixner tried to fix up the vma-based robust futex patch in the -rt tree, he
found a handful of new races and we were talking about it and were analyzing
the situation. At that point a fundamentally different solution occured to
me. This patchset (written in the past couple of days) implements that new
solution. Be warned though - the patchset does things we normally dont do in
Linux, so some might find the approach disturbing. Parental advice
recommended ;-)
New approach to robust futexes
------------------------------
At the heart of this new approach there is a per-thread private list of robust
locks that userspace is holding (maintained by glibc) - which userspace list
is registered with the kernel via a new syscall [this registration happens at
most once per thread lifetime]. At do_exit() time, the kernel checks this
user-space list: are there any robust futex locks to be cleaned up?
In the common case, at do_exit() time, there is no list registered, so the
cost of robust futexes is just a simple current->robust_list != NULL
comparison. If the thread has registered a list, then normally the list is
empty. If the thread/process crashed or terminated in some incorrect way then
the list might be non-empty: in this case the kernel carefully walks the list
[not trusting it], and marks all locks that are owned by this thread with the
FUTEX_OWNER_DEAD bit, and wakes up one waiter (if any).
The list is guaranteed to be private and per-thread, so it's lockless. There
is one race possible though: since adding to and removing from the list is
done after the futex is acquired by glibc, there is a few instructions window
for the thread (or process) to die there, leaving the futex hung. To protect
against this possibility, userspace (glibc) also maintains a simple per-thread
'list_op_pending' field, to allow the kernel to clean up if the thread dies
after acquiring the lock, but just before it could have added itself to the
list. Glibc sets this list_op_pending field before it tries to acquire the
futex, and clears it after the list-add (or list-remove) has finished.
That's all that is needed - all the rest of robust-futex cleanup is done in
userspace [just like with the previous patches].
Ulrich Drepper has implemented the necessary glibc support for this new
mechanism, which fully enables robust mutexes. (Ulrich plans to commit these
changes to glibc-HEAD later today.)
Key differences of this userspace-list based approach, compared to the vma
based method:
- it's much, much faster: at thread exit time, there's no need to loop
over every vma (!), which the VM-based method has to do. Only a very
simple 'is the list empty' op is done.
- no VM changes are needed - 'struct address_space' is left alone.
- no registration of individual locks is needed: robust mutexes dont need
any extra per-lock syscalls. Robust mutexes thus become a very lightweight
primitive - so they dont force the application designer to do a hard choice
between performance and robustness - robust mutexes are just as fast.
- no per-lock kernel allocation happens.
- no resource limits are needed.
- no kernel-space recovery call (FUTEX_RECOVER) is needed.
- the implementation and the locking is "obvious", and there are no
interactions with the VM.
Performance
-----------
I have benchmarked the time needed for the kernel to process a list of 1
million (!) held locks, using the new method [on a 2GHz CPU]:
- with FUTEX_WAIT set [contended mutex]: 130 msecs
- without FUTEX_WAIT set [uncontended mutex]: 30 msecs
I have also measured an approach where glibc does the lock notification [which
it currently does for !pshared robust mutexes], and that took 256 msecs -
clearly slower, due to the 1 million FUTEX_WAKE syscalls userspace had to do.
(1 million held locks are unheard of - we expect at most a handful of locks to
be held at a time. Nevertheless it's nice to know that this approach scales
nicely.)
Implementation details
----------------------
The patch adds two new syscalls: one to register the userspace list, and one
to query the registered list pointer:
asmlinkage long
sys_set_robust_list(struct robust_list_head __user *head,
size_t len);
asmlinkage long
sys_get_robust_list(int pid, struct robust_list_head __user **head_ptr,
size_t __user *len_ptr);
List registration is very fast: the pointer is simply stored in
current->robust_list. [Note that in the future, if robust futexes become
widespread, we could extend sys_clone() to register a robust-list head for new
threads, without the need of another syscall.]
So there is virtually zero overhead for tasks not using robust futexes, and
even for robust futex users, there is only one extra syscall per thread
lifetime, and the cleanup operation, if it happens, is fast and
straightforward. The kernel doesnt have any internal distinction between
robust and normal futexes.
If a futex is found to be held at exit time, the kernel sets the highest bit
of the futex word:
#define FUTEX_OWNER_DIED 0x40000000
and wakes up the next futex waiter (if any). User-space does the rest of
the cleanup.
Otherwise, robust futexes are acquired by glibc by putting the TID into the
futex field atomically. Waiters set the FUTEX_WAITERS bit:
#define FUTEX_WAITERS 0x80000000
and the remaining bits are for the TID.
Testing, architecture support
-----------------------------
I've tested the new syscalls on x86 and x86_64, and have made sure the parsing
of the userspace list is robust [ ;-) ] even if the list is deliberately
corrupted.
i386 and x86_64 syscalls are wired up at the moment, and Ulrich has tested the
new glibc code (on x86_64 and i386), and it works for his robust-mutex
testcases.
All other architectures should build just fine too - but they wont have the
new syscalls yet.
Architectures need to implement the new futex_atomic_cmpxchg_inuser() inline
function before writing up the syscalls (that function returns -ENOSYS right
now).
This patch:
Add placeholder futex_atomic_cmpxchg_inuser() implementations to every
architecture that supports futexes. It returns -ENOSYS.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Arjan van de Ven <arjan@infradead.org>
Acked-by: Ulrich Drepper <drepper@redhat.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
PowerPC can use generic ones.
Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
|
|
On CHRP machines we are supposed to call into firmware (RTAS)
periodically, to give it a chance to check for errors and other
events. Under ppc we had some special code in timer_interrupt
to do this, but that didn't get transferred over to arch/powerpc.
Instead, we use an array of timer_list structs, one per CPU,
and use add_timer_on to make sure each one gets called on the
appropriate CPU.
With this we can remove the heartbeat_* elements of the ppc_md
struct.
Signed-off-by: Paul Mackerras <paulus@samba.org>
|
|
Since pSeries only wants to do something different in the idle loop when
there is no work to do, we can simplify the code by implementing
ppc_md.power_save functions instead of complete idle loops. There are
two versions: one for shared-processor partitions and one for dedicated-
processor partitions.
With this we also do a cede_processor() call on dedicated processor
partitions if the poll_pending() call indicates that the hypervisor
has work it wants to do.
Signed-off-by: Paul Mackerras <paulus@samba.org>
|
|
This unifies the 32-bit (ARCH=ppc and ARCH=powerpc) and 64-bit idle
loops. It brings over the concept of having a ppc_md.power_save
function from 32-bit to ARCH=powerpc, which lets us get rid of
native_idle(). With this we will also be able to simplify the idle
handling for pSeries and cell.
Signed-off-by: Paul Mackerras <paulus@samba.org>
|
David Woodhouse
Paul Mackerras
Jeremy Kerr
David Woodhouse
Anton Blanchard
Andreas Schwab
David Gibson
Jens Axboe
Olof Johansson
Linus Torvalds
Yasunori Goto
Stephen Rothwell
Nathan Fontenot
Heiko J Schick
Segher Boessenkool
Brian Rogan
KAMEZAWA Hiroyuki
Benjamin Herrenschmidt
Andrew Morton
Ryan S. Arnold
Michael Ellerman
Alan Stern
Ingo Molnar