aboutsummaryrefslogtreecommitdiffstats
path: root/arch/x86/entry
diff options
context:
space:
mode:
Diffstat (limited to 'arch/x86/entry')
-rw-r--r--arch/x86/entry/calling.h2
-rw-r--r--arch/x86/entry/common.c4
-rw-r--r--arch/x86/entry/entry_32.S422
-rw-r--r--arch/x86/entry/entry_64.S112
-rw-r--r--arch/x86/entry/entry_64_compat.S16
-rw-r--r--arch/x86/entry/syscall_32.c8
-rw-r--r--arch/x86/entry/syscall_64.c14
-rw-r--r--arch/x86/entry/syscalls/syscall_32.tbl8
-rw-r--r--arch/x86/entry/thunk_32.S4
-rw-r--r--arch/x86/entry/thunk_64.S7
-rw-r--r--arch/x86/entry/vdso/Makefile2
-rw-r--r--arch/x86/entry/vdso/vclock_gettime.c6
-rw-r--r--arch/x86/entry/vdso/vdso32/system_call.S2
-rw-r--r--arch/x86/entry/vsyscall/vsyscall_64.c6
14 files changed, 361 insertions, 252 deletions
diff --git a/arch/x86/entry/calling.h b/arch/x86/entry/calling.h
index 515c0ceeb4a3..0789e13ece90 100644
--- a/arch/x86/entry/calling.h
+++ b/arch/x86/entry/calling.h
@@ -354,7 +354,7 @@ For 32-bit we have the following conventions - kernel is built with
.macro CALL_enter_from_user_mode
#ifdef CONFIG_CONTEXT_TRACKING
#ifdef CONFIG_JUMP_LABEL
- STATIC_JUMP_IF_FALSE .Lafter_call_\@, context_tracking_enabled, def=0
+ STATIC_JUMP_IF_FALSE .Lafter_call_\@, context_tracking_key, def=0
#endif
call enter_from_user_mode
.Lafter_call_\@:
diff --git a/arch/x86/entry/common.c b/arch/x86/entry/common.c
index 3f8e22615812..9747876980b5 100644
--- a/arch/x86/entry/common.c
+++ b/arch/x86/entry/common.c
@@ -33,6 +33,7 @@
#include <asm/cpufeature.h>
#include <asm/fpu/api.h>
#include <asm/nospec-branch.h>
+#include <asm/io_bitmap.h>
#define CREATE_TRACE_POINTS
#include <trace/events/syscalls.h>
@@ -196,6 +197,9 @@ __visible inline void prepare_exit_to_usermode(struct pt_regs *regs)
/* Reload ti->flags; we may have rescheduled above. */
cached_flags = READ_ONCE(ti->flags);
+ if (unlikely(cached_flags & _TIF_IO_BITMAP))
+ tss_update_io_bitmap();
+
fpregs_assert_state_consistent();
if (unlikely(cached_flags & _TIF_NEED_FPU_LOAD))
switch_fpu_return();
diff --git a/arch/x86/entry/entry_32.S b/arch/x86/entry/entry_32.S
index f83ca5aa8b77..7e0560442538 100644
--- a/arch/x86/entry/entry_32.S
+++ b/arch/x86/entry/entry_32.S
@@ -172,7 +172,7 @@
ALTERNATIVE "jmp .Lend_\@", "", X86_FEATURE_PTI
.if \no_user_check == 0
/* coming from usermode? */
- testl $SEGMENT_RPL_MASK, PT_CS(%esp)
+ testl $USER_SEGMENT_RPL_MASK, PT_CS(%esp)
jz .Lend_\@
.endif
/* On user-cr3? */
@@ -205,64 +205,76 @@
#define CS_FROM_ENTRY_STACK (1 << 31)
#define CS_FROM_USER_CR3 (1 << 30)
#define CS_FROM_KERNEL (1 << 29)
+#define CS_FROM_ESPFIX (1 << 28)
.macro FIXUP_FRAME
/*
* The high bits of the CS dword (__csh) are used for CS_FROM_*.
* Clear them in case hardware didn't do this for us.
*/
- andl $0x0000ffff, 3*4(%esp)
+ andl $0x0000ffff, 4*4(%esp)
#ifdef CONFIG_VM86
- testl $X86_EFLAGS_VM, 4*4(%esp)
+ testl $X86_EFLAGS_VM, 5*4(%esp)
jnz .Lfrom_usermode_no_fixup_\@
#endif
- testl $SEGMENT_RPL_MASK, 3*4(%esp)
+ testl $USER_SEGMENT_RPL_MASK, 4*4(%esp)
jnz .Lfrom_usermode_no_fixup_\@
- orl $CS_FROM_KERNEL, 3*4(%esp)
+ orl $CS_FROM_KERNEL, 4*4(%esp)
/*
* When we're here from kernel mode; the (exception) stack looks like:
*
- * 5*4(%esp) - <previous context>
- * 4*4(%esp) - flags
- * 3*4(%esp) - cs
- * 2*4(%esp) - ip
- * 1*4(%esp) - orig_eax
- * 0*4(%esp) - gs / function
+ * 6*4(%esp) - <previous context>
+ * 5*4(%esp) - flags
+ * 4*4(%esp) - cs
+ * 3*4(%esp) - ip
+ * 2*4(%esp) - orig_eax
+ * 1*4(%esp) - gs / function
+ * 0*4(%esp) - fs
*
* Lets build a 5 entry IRET frame after that, such that struct pt_regs
* is complete and in particular regs->sp is correct. This gives us
- * the original 5 enties as gap:
+ * the original 6 enties as gap:
*
- * 12*4(%esp) - <previous context>
- * 11*4(%esp) - gap / flags
- * 10*4(%esp) - gap / cs
- * 9*4(%esp) - gap / ip
- * 8*4(%esp) - gap / orig_eax
- * 7*4(%esp) - gap / gs / function
- * 6*4(%esp) - ss
- * 5*4(%esp) - sp
- * 4*4(%esp) - flags
- * 3*4(%esp) - cs
- * 2*4(%esp) - ip
- * 1*4(%esp) - orig_eax
- * 0*4(%esp) - gs / function
+ * 14*4(%esp) - <previous context>
+ * 13*4(%esp) - gap / flags
+ * 12*4(%esp) - gap / cs
+ * 11*4(%esp) - gap / ip
+ * 10*4(%esp) - gap / orig_eax
+ * 9*4(%esp) - gap / gs / function
+ * 8*4(%esp) - gap / fs
+ * 7*4(%esp) - ss
+ * 6*4(%esp) - sp
+ * 5*4(%esp) - flags
+ * 4*4(%esp) - cs
+ * 3*4(%esp) - ip
+ * 2*4(%esp) - orig_eax
+ * 1*4(%esp) - gs / function
+ * 0*4(%esp) - fs
*/
pushl %ss # ss
pushl %esp # sp (points at ss)
- addl $6*4, (%esp) # point sp back at the previous context
- pushl 6*4(%esp) # flags
- pushl 6*4(%esp) # cs
- pushl 6*4(%esp) # ip
- pushl 6*4(%esp) # orig_eax
- pushl 6*4(%esp) # gs / function
+ addl $7*4, (%esp) # point sp back at the previous context
+ pushl 7*4(%esp) # flags
+ pushl 7*4(%esp) # cs
+ pushl 7*4(%esp) # ip
+ pushl 7*4(%esp) # orig_eax
+ pushl 7*4(%esp) # gs / function
+ pushl 7*4(%esp) # fs
.Lfrom_usermode_no_fixup_\@:
.endm
.macro IRET_FRAME
+ /*
+ * We're called with %ds, %es, %fs, and %gs from the interrupted
+ * frame, so we shouldn't use them. Also, we may be in ESPFIX
+ * mode and therefore have a nonzero SS base and an offset ESP,
+ * so any attempt to access the stack needs to use SS. (except for
+ * accesses through %esp, which automatically use SS.)
+ */
testl $CS_FROM_KERNEL, 1*4(%esp)
jz .Lfinished_frame_\@
@@ -276,31 +288,40 @@
movl 5*4(%esp), %eax # (modified) regs->sp
movl 4*4(%esp), %ecx # flags
- movl %ecx, -4(%eax)
+ movl %ecx, %ss:-1*4(%eax)
movl 3*4(%esp), %ecx # cs
andl $0x0000ffff, %ecx
- movl %ecx, -8(%eax)
+ movl %ecx, %ss:-2*4(%eax)
movl 2*4(%esp), %ecx # ip
- movl %ecx, -12(%eax)
+ movl %ecx, %ss:-3*4(%eax)
movl 1*4(%esp), %ecx # eax
- movl %ecx, -16(%eax)
+ movl %ecx, %ss:-4*4(%eax)
popl %ecx
- lea -16(%eax), %esp
+ lea -4*4(%eax), %esp
popl %eax
.Lfinished_frame_\@:
.endm
-.macro SAVE_ALL pt_regs_ax=%eax switch_stacks=0 skip_gs=0
+.macro SAVE_ALL pt_regs_ax=%eax switch_stacks=0 skip_gs=0 unwind_espfix=0
cld
.if \skip_gs == 0
PUSH_GS
.endif
- FIXUP_FRAME
pushl %fs
+
+ pushl %eax
+ movl $(__KERNEL_PERCPU), %eax
+ movl %eax, %fs
+.if \unwind_espfix > 0
+ UNWIND_ESPFIX_STACK
+.endif
+ popl %eax
+
+ FIXUP_FRAME
pushl %es
pushl %ds
pushl \pt_regs_ax
@@ -313,8 +334,6 @@
movl $(__USER_DS), %edx
movl %edx, %ds
movl %edx, %es
- movl $(__KERNEL_PERCPU), %edx
- movl %edx, %fs
.if \skip_gs == 0
SET_KERNEL_GS %edx
.endif
@@ -324,8 +343,8 @@
.endif
.endm
-.macro SAVE_ALL_NMI cr3_reg:req
- SAVE_ALL
+.macro SAVE_ALL_NMI cr3_reg:req unwind_espfix=0
+ SAVE_ALL unwind_espfix=\unwind_espfix
BUG_IF_WRONG_CR3
@@ -357,6 +376,7 @@
2: popl %es
3: popl %fs
POP_GS \pop
+ IRET_FRAME
.pushsection .fixup, "ax"
4: movl $0, (%esp)
jmp 1b
@@ -395,7 +415,8 @@
.macro CHECK_AND_APPLY_ESPFIX
#ifdef CONFIG_X86_ESPFIX32
-#define GDT_ESPFIX_SS PER_CPU_VAR(gdt_page) + (GDT_ENTRY_ESPFIX_SS * 8)
+#define GDT_ESPFIX_OFFSET (GDT_ENTRY_ESPFIX_SS * 8)
+#define GDT_ESPFIX_SS PER_CPU_VAR(gdt_page) + GDT_ESPFIX_OFFSET
ALTERNATIVE "jmp .Lend_\@", "", X86_BUG_ESPFIX
@@ -709,7 +730,7 @@
* %eax: prev task
* %edx: next task
*/
-ENTRY(__switch_to_asm)
+SYM_CODE_START(__switch_to_asm)
/*
* Save callee-saved registers
* This must match the order in struct inactive_task_frame
@@ -718,6 +739,11 @@ ENTRY(__switch_to_asm)
pushl %ebx
pushl %edi
pushl %esi
+ /*
+ * Flags are saved to prevent AC leakage. This could go
+ * away if objtool would have 32bit support to verify
+ * the STAC/CLAC correctness.
+ */
pushfl
/* switch stack */
@@ -740,15 +766,16 @@ ENTRY(__switch_to_asm)
FILL_RETURN_BUFFER %ebx, RSB_CLEAR_LOOPS, X86_FEATURE_RSB_CTXSW
#endif
- /* restore callee-saved registers */
+ /* Restore flags or the incoming task to restore AC state. */
popfl
+ /* restore callee-saved registers */
popl %esi
popl %edi
popl %ebx
popl %ebp
jmp __switch_to
-END(__switch_to_asm)
+SYM_CODE_END(__switch_to_asm)
/*
* The unwinder expects the last frame on the stack to always be at the same
@@ -757,7 +784,7 @@ END(__switch_to_asm)
* asmlinkage function so its argument has to be pushed on the stack. This
* wrapper creates a proper "end of stack" frame header before the call.
*/
-ENTRY(schedule_tail_wrapper)
+SYM_FUNC_START(schedule_tail_wrapper)
FRAME_BEGIN
pushl %eax
@@ -766,7 +793,7 @@ ENTRY(schedule_tail_wrapper)
FRAME_END
ret
-ENDPROC(schedule_tail_wrapper)
+SYM_FUNC_END(schedule_tail_wrapper)
/*
* A newly forked process directly context switches into this address.
*
@@ -774,7 +801,7 @@ ENDPROC(schedule_tail_wrapper)
* ebx: kernel thread func (NULL for user thread)
* edi: kernel thread arg
*/
-ENTRY(ret_from_fork)
+SYM_CODE_START(ret_from_fork)
call schedule_tail_wrapper
testl %ebx, %ebx
@@ -797,7 +824,7 @@ ENTRY(ret_from_fork)
*/
movl $0, PT_EAX(%esp)
jmp 2b
-END(ret_from_fork)
+SYM_CODE_END(ret_from_fork)
/*
* Return to user mode is not as complex as all this looks,
@@ -807,8 +834,7 @@ END(ret_from_fork)
*/
# userspace resumption stub bypassing syscall exit tracing
- ALIGN
-ret_from_exception:
+SYM_CODE_START_LOCAL(ret_from_exception)
preempt_stop(CLBR_ANY)
ret_from_intr:
#ifdef CONFIG_VM86
@@ -825,15 +851,14 @@ ret_from_intr:
cmpl $USER_RPL, %eax
jb restore_all_kernel # not returning to v8086 or userspace
-ENTRY(resume_userspace)
DISABLE_INTERRUPTS(CLBR_ANY)
TRACE_IRQS_OFF
movl %esp, %eax
call prepare_exit_to_usermode
jmp restore_all
-END(ret_from_exception)
+SYM_CODE_END(ret_from_exception)
-GLOBAL(__begin_SYSENTER_singlestep_region)
+SYM_ENTRY(__begin_SYSENTER_singlestep_region, SYM_L_GLOBAL, SYM_A_NONE)
/*
* All code from here through __end_SYSENTER_singlestep_region is subject
* to being single-stepped if a user program sets TF and executes SYSENTER.
@@ -848,9 +873,10 @@ GLOBAL(__begin_SYSENTER_singlestep_region)
* Xen doesn't set %esp to be precisely what the normal SYSENTER
* entry point expects, so fix it up before using the normal path.
*/
-ENTRY(xen_sysenter_target)
+SYM_CODE_START(xen_sysenter_target)
addl $5*4, %esp /* remove xen-provided frame */
jmp .Lsysenter_past_esp
+SYM_CODE_END(xen_sysenter_target)
#endif
/*
@@ -885,7 +911,7 @@ ENTRY(xen_sysenter_target)
* ebp user stack
* 0(%ebp) arg6
*/
-ENTRY(entry_SYSENTER_32)
+SYM_FUNC_START(entry_SYSENTER_32)
/*
* On entry-stack with all userspace-regs live - save and
* restore eflags and %eax to use it as scratch-reg for the cr3
@@ -1012,8 +1038,8 @@ ENTRY(entry_SYSENTER_32)
pushl $X86_EFLAGS_FIXED
popfl
jmp .Lsysenter_flags_fixed
-GLOBAL(__end_SYSENTER_singlestep_region)
-ENDPROC(entry_SYSENTER_32)
+SYM_ENTRY(__end_SYSENTER_singlestep_region, SYM_L_GLOBAL, SYM_A_NONE)
+SYM_FUNC_END(entry_SYSENTER_32)
/*
* 32-bit legacy system call entry.
@@ -1043,7 +1069,7 @@ ENDPROC(entry_SYSENTER_32)
* edi arg5
* ebp arg6
*/
-ENTRY(entry_INT80_32)
+SYM_FUNC_START(entry_INT80_32)
ASM_CLAC
pushl %eax /* pt_regs->orig_ax */
@@ -1064,7 +1090,6 @@ ENTRY(entry_INT80_32)
restore_all:
TRACE_IRQS_IRET
SWITCH_TO_ENTRY_STACK
-.Lrestore_all_notrace:
CHECK_AND_APPLY_ESPFIX
.Lrestore_nocheck:
/* Switch back to user CR3 */
@@ -1075,7 +1100,6 @@ restore_all:
/* Restore user state */
RESTORE_REGS pop=4 # skip orig_eax/error_code
.Lirq_return:
- IRET_FRAME
/*
* ARCH_HAS_MEMBARRIER_SYNC_CORE rely on IRET core serialization
* when returning from IPI handler and when returning from
@@ -1100,7 +1124,7 @@ restore_all_kernel:
jmp .Lirq_return
.section .fixup, "ax"
-ENTRY(iret_exc )
+SYM_CODE_START(iret_exc)
pushl $0 # no error code
pushl $do_iret_error
@@ -1117,9 +1141,10 @@ ENTRY(iret_exc )
#endif
jmp common_exception
+SYM_CODE_END(iret_exc)
.previous
_ASM_EXTABLE(.Lirq_return, iret_exc)
-ENDPROC(entry_INT80_32)
+SYM_FUNC_END(entry_INT80_32)
.macro FIXUP_ESPFIX_STACK
/*
@@ -1128,30 +1153,43 @@ ENDPROC(entry_INT80_32)
* We can't call C functions using the ESPFIX stack. This code reads
* the high word of the segment base from the GDT and swiches to the
* normal stack and adjusts ESP with the matching offset.
+ *
+ * We might be on user CR3 here, so percpu data is not mapped and we can't
+ * access the GDT through the percpu segment. Instead, use SGDT to find
+ * the cpu_entry_area alias of the GDT.
*/
#ifdef CONFIG_X86_ESPFIX32
/* fixup the stack */
- mov GDT_ESPFIX_SS + 4, %al /* bits 16..23 */
- mov GDT_ESPFIX_SS + 7, %ah /* bits 24..31 */
+ pushl %ecx
+ subl $2*4, %esp
+ sgdt (%esp)
+ movl 2(%esp), %ecx /* GDT address */
+ /*
+ * Careful: ECX is a linear pointer, so we need to force base
+ * zero. %cs is the only known-linear segment we have right now.
+ */
+ mov %cs:GDT_ESPFIX_OFFSET + 4(%ecx), %al /* bits 16..23 */
+ mov %cs:GDT_ESPFIX_OFFSET + 7(%ecx), %ah /* bits 24..31 */
shl $16, %eax
+ addl $2*4, %esp
+ popl %ecx
addl %esp, %eax /* the adjusted stack pointer */
pushl $__KERNEL_DS
pushl %eax
lss (%esp), %esp /* switch to the normal stack segment */
#endif
.endm
+
.macro UNWIND_ESPFIX_STACK
+ /* It's safe to clobber %eax, all other regs need to be preserved */
#ifdef CONFIG_X86_ESPFIX32
movl %ss, %eax
/* see if on espfix stack */
cmpw $__ESPFIX_SS, %ax
- jne 27f
- movl $__KERNEL_DS, %eax
- movl %eax, %ds
- movl %eax, %es
+ jne .Lno_fixup_\@
/* switch to normal stack */
FIXUP_ESPFIX_STACK
-27:
+.Lno_fixup_\@:
#endif
.endm
@@ -1160,7 +1198,7 @@ ENDPROC(entry_INT80_32)
* We pack 1 stub into every 8-byte block.
*/
.align 8
-ENTRY(irq_entries_start)
+SYM_CODE_START(irq_entries_start)
vector=FIRST_EXTERNAL_VECTOR
.rept (FIRST_SYSTEM_VECTOR - FIRST_EXTERNAL_VECTOR)
pushl $(~vector+0x80) /* Note: always in signed byte range */
@@ -1168,11 +1206,11 @@ ENTRY(irq_entries_start)
jmp common_interrupt
.align 8
.endr
-END(irq_entries_start)
+SYM_CODE_END(irq_entries_start)
#ifdef CONFIG_X86_LOCAL_APIC
.align 8
-ENTRY(spurious_entries_start)
+SYM_CODE_START(spurious_entries_start)
vector=FIRST_SYSTEM_VECTOR
.rept (NR_VECTORS - FIRST_SYSTEM_VECTOR)
pushl $(~vector+0x80) /* Note: always in signed byte range */
@@ -1180,9 +1218,9 @@ ENTRY(spurious_entries_start)
jmp common_spurious
.align 8
.endr
-END(spurious_entries_start)
+SYM_CODE_END(spurious_entries_start)
-common_spurious:
+SYM_CODE_START_LOCAL(common_spurious)
ASM_CLAC
addl $-0x80, (%esp) /* Adjust vector into the [-256, -1] range */
SAVE_ALL switch_stacks=1
@@ -1191,7 +1229,7 @@ common_spurious:
movl %esp, %eax
call smp_spurious_interrupt
jmp ret_from_intr
-ENDPROC(common_spurious)
+SYM_CODE_END(common_spurious)
#endif
/*
@@ -1199,7 +1237,7 @@ ENDPROC(common_spurious)
* so IRQ-flags tracing has to follow that:
*/
.p2align CONFIG_X86_L1_CACHE_SHIFT
-common_interrupt:
+SYM_CODE_START_LOCAL(common_interrupt)
ASM_CLAC
addl $-0x80, (%esp) /* Adjust vector into the [-256, -1] range */
@@ -1209,10 +1247,10 @@ common_interrupt:
movl %esp, %eax
call do_IRQ
jmp ret_from_intr
-ENDPROC(common_interrupt)
+SYM_CODE_END(common_interrupt)
#define BUILD_INTERRUPT3(name, nr, fn) \
-ENTRY(name) \
+SYM_FUNC_START(name) \
ASM_CLAC; \
pushl $~(nr); \
SAVE_ALL switch_stacks=1; \
@@ -1221,7 +1259,7 @@ ENTRY(name) \
movl %esp, %eax; \
call fn; \
jmp ret_from_intr; \
-ENDPROC(name)
+SYM_FUNC_END(name)
#define BUILD_INTERRUPT(name, nr) \
BUILD_INTERRUPT3(name, nr, smp_##name); \
@@ -1229,14 +1267,14 @@ ENDPROC(name)
/* The include is where all of the SMP etc. interrupts come from */
#include <asm/entry_arch.h>
-ENTRY(coprocessor_error)
+SYM_CODE_START(coprocessor_error)
ASM_CLAC
pushl $0
pushl $do_coprocessor_error
jmp common_exception
-END(coprocessor_error)
+SYM_CODE_END(coprocessor_error)
-ENTRY(simd_coprocessor_error)
+SYM_CODE_START(simd_coprocessor_error)
ASM_CLAC
pushl $0
#ifdef CONFIG_X86_INVD_BUG
@@ -1248,104 +1286,99 @@ ENTRY(simd_coprocessor_error)
pushl $do_simd_coprocessor_error
#endif
jmp common_exception
-END(simd_coprocessor_error)
+SYM_CODE_END(simd_coprocessor_error)
-ENTRY(device_not_available)
+SYM_CODE_START(device_not_available)
ASM_CLAC
pushl $-1 # mark this as an int
pushl $do_device_not_available
jmp common_exception
-END(device_not_available)
+SYM_CODE_END(device_not_available)
#ifdef CONFIG_PARAVIRT
-ENTRY(native_iret)
+SYM_CODE_START(native_iret)
iret
_ASM_EXTABLE(native_iret, iret_exc)
-END(native_iret)
+SYM_CODE_END(native_iret)
#endif
-ENTRY(overflow)
+SYM_CODE_START(overflow)
ASM_CLAC
pushl $0
pushl $do_overflow
jmp common_exception
-END(overflow)
+SYM_CODE_END(overflow)
-ENTRY(bounds)
+SYM_CODE_START(bounds)
ASM_CLAC
pushl $0
pushl $do_bounds
jmp common_exception
-END(bounds)
+SYM_CODE_END(bounds)
-ENTRY(invalid_op)
+SYM_CODE_START(invalid_op)
ASM_CLAC
pushl $0
pushl $do_invalid_op
jmp common_exception
-END(invalid_op)
+SYM_CODE_END(invalid_op)
-ENTRY(coprocessor_segment_overrun)
+SYM_CODE_START(coprocessor_segment_overrun)
ASM_CLAC
pushl $0
pushl $do_coprocessor_segment_overrun
jmp common_exception
-END(coprocessor_segment_overrun)
+SYM_CODE_END(coprocessor_segment_overrun)
-ENTRY(invalid_TSS)
+SYM_CODE_START(invalid_TSS)
ASM_CLAC
pushl $do_invalid_TSS
jmp common_exception
-END(invalid_TSS)
+SYM_CODE_END(invalid_TSS)
-ENTRY(segment_not_present)
+SYM_CODE_START(segment_not_present)
ASM_CLAC
pushl $do_segment_not_present
jmp common_exception
-END(segment_not_present)
+SYM_CODE_END(segment_not_present)
-ENTRY(stack_segment)
+SYM_CODE_START(stack_segment)
ASM_CLAC
pushl $do_stack_segment
jmp common_exception
-END(stack_segment)
+SYM_CODE_END(stack_segment)
-ENTRY(alignment_check)
+SYM_CODE_START(alignment_check)
ASM_CLAC
pushl $do_alignment_check
jmp common_exception
-END(alignment_check)
+SYM_CODE_END(alignment_check)
-ENTRY(divide_error)
+SYM_CODE_START(divide_error)
ASM_CLAC
pushl $0 # no error code
pushl $do_divide_error
jmp common_exception
-END(divide_error)
+SYM_CODE_END(divide_error)
#ifdef CONFIG_X86_MCE
-ENTRY(machine_check)
+SYM_CODE_START(machine_check)
ASM_CLAC
pushl $0
pushl machine_check_vector
jmp common_exception
-END(machine_check)
+SYM_CODE_END(machine_check)
#endif
-ENTRY(spurious_interrupt_bug)
+SYM_CODE_START(spurious_interrupt_bug)
ASM_CLAC
pushl $0
pushl $do_spurious_interrupt_bug
jmp common_exception
-END(spurious_interrupt_bug)
+SYM_CODE_END(spurious_interrupt_bug)
#ifdef CONFIG_XEN_PV
-ENTRY(xen_hypervisor_callback)
- pushl $-1 /* orig_ax = -1 => not a system call */
- SAVE_ALL
- ENCODE_FRAME_POINTER
- TRACE_IRQS_OFF
-
+SYM_FUNC_START(xen_hypervisor_callback)
/*
* Check to see if we got the event in the critical
* region in xen_iret_direct, after we've reenabled
@@ -1353,22 +1386,23 @@ ENTRY(xen_hypervisor_callback)
* iret instruction's behaviour where it delivers a
* pending interrupt when enabling interrupts:
*/
- movl PT_EIP(%esp), %eax
- cmpl $xen_iret_start_crit, %eax
+ cmpl $xen_iret_start_crit, (%esp)
jb 1f
- cmpl $xen_iret_end_crit, %eax
+ cmpl $xen_iret_end_crit, (%esp)
jae 1f
-
- jmp xen_iret_crit_fixup
-
-ENTRY(xen_do_upcall)
-1: mov %esp, %eax
+ call xen_iret_crit_fixup
+1:
+ pushl $-1 /* orig_ax = -1 => not a system call */
+ SAVE_ALL
+ ENCODE_FRAME_POINTER
+ TRACE_IRQS_OFF
+ mov %esp, %eax
call xen_evtchn_do_upcall
#ifndef CONFIG_PREEMPTION
call xen_maybe_preempt_hcall
#endif
jmp ret_from_intr
-ENDPROC(xen_hypervisor_callback)
+SYM_FUNC_END(xen_hypervisor_callback)
/*
* Hypervisor uses this for application faults while it executes.
@@ -1382,7 +1416,7 @@ ENDPROC(xen_hypervisor_callback)
* to pop the stack frame we end up in an infinite loop of failsafe callbacks.
* We distinguish between categories by maintaining a status value in EAX.
*/
-ENTRY(xen_failsafe_callback)
+SYM_FUNC_START(xen_failsafe_callback)
pushl %eax
movl $1, %eax
1: mov 4(%esp), %ds
@@ -1419,7 +1453,7 @@ ENTRY(xen_failsafe_callback)
_ASM_EXTABLE(2b, 7b)
_ASM_EXTABLE(3b, 8b)
_ASM_EXTABLE(4b, 9b)
-ENDPROC(xen_failsafe_callback)
+SYM_FUNC_END(xen_failsafe_callback)
#endif /* CONFIG_XEN_PV */
#ifdef CONFIG_XEN_PVHVM
@@ -1441,18 +1475,17 @@ BUILD_INTERRUPT3(hv_stimer0_callback_vector, HYPERV_STIMER0_VECTOR,
#endif /* CONFIG_HYPERV */
-ENTRY(page_fault)
+SYM_CODE_START(page_fault)
ASM_CLAC
pushl $do_page_fault
jmp common_exception_read_cr2
-END(page_fault)
+SYM_CODE_END(page_fault)
-common_exception_read_cr2:
+SYM_CODE_START_LOCAL_NOALIGN(common_exception_read_cr2)
/* the function address is in %gs's slot on the stack */
- SAVE_ALL switch_stacks=1 skip_gs=1
+ SAVE_ALL switch_stacks=1 skip_gs=1 unwind_espfix=1
ENCODE_FRAME_POINTER
- UNWIND_ESPFIX_STACK
/* fixup %gs */
GS_TO_REG %ecx
@@ -1470,13 +1503,12 @@ common_exception_read_cr2:
movl %esp, %eax # pt_regs pointer
CALL_NOSPEC %edi
jmp ret_from_exception
-END(common_exception_read_cr2)
+SYM_CODE_END(common_exception_read_cr2)
-common_exception:
+SYM_CODE_START_LOCAL_NOALIGN(common_exception)
/* the function address is in %gs's slot on the stack */
- SAVE_ALL switch_stacks=1 skip_gs=1
+ SAVE_ALL switch_stacks=1 skip_gs=1 unwind_espfix=1
ENCODE_FRAME_POINTER
- UNWIND_ESPFIX_STACK
/* fixup %gs */
GS_TO_REG %ecx
@@ -1492,9 +1524,9 @@ common_exception:
movl %esp, %eax # pt_regs pointer
CALL_NOSPEC %edi
jmp ret_from_exception
-END(common_exception)
+SYM_CODE_END(common_exception)
-ENTRY(debug)
+SYM_CODE_START(debug)
/*
* Entry from sysenter is now handled in common_exception
*/
@@ -1502,7 +1534,49 @@ ENTRY(debug)
pushl $-1 # mark this as an int
pushl $do_debug
jmp common_exception
-END(debug)
+SYM_CODE_END(debug)
+
+#ifdef CONFIG_DOUBLEFAULT
+SYM_CODE_START(double_fault)
+1:
+ /*
+ * This is a task gate handler, not an interrupt gate handler.
+ * The error code is on the stack, but the stack is otherwise
+ * empty. Interrupts are off. Our state is sane with the following
+ * exceptions:
+ *
+ * - CR0.TS is set. "TS" literally means "task switched".
+ * - EFLAGS.NT is set because we're a "nested task".
+ * - The doublefault TSS has back_link set and has been marked busy.
+ * - TR points to the doublefault TSS and the normal TSS is busy.
+ * - CR3 is the normal kernel PGD. This would be delightful, except
+ * that the CPU didn't bother to save the old CR3 anywhere. This
+ * would make it very awkward to return back to the context we came
+ * from.
+ *
+ * The rest of EFLAGS is sanitized for us, so we don't need to
+ * worry about AC or DF.
+ *
+ * Don't even bother popping the error code. It's always zero,
+ * and ignoring it makes us a bit more robust against buggy
+ * hypervisor task gate implementations.
+ *
+ * We will manually undo the task switch instead of doing a
+ * task-switching IRET.
+ */
+
+ clts /* clear CR0.TS */
+ pushl $X86_EFLAGS_FIXED
+ popfl /* clear EFLAGS.NT */
+
+ call doublefault_shim
+
+ /* We don't support returning, so we have no IRET here. */
+1:
+ hlt
+ jmp 1b
+SYM_CODE_END(double_fault)
+#endif
/*
* NMI is doubly nasty. It can happen on the first instruction of
@@ -1511,10 +1585,14 @@ END(debug)
* switched stacks. We handle both conditions by simply checking whether we
* interrupted kernel code running on the SYSENTER stack.
*/
-ENTRY(nmi)
+SYM_CODE_START(nmi)
ASM_CLAC
#ifdef CONFIG_X86_ESPFIX32
+ /*
+ * ESPFIX_SS is only ever set on the return to user path
+ * after we've switched to the entry stack.
+ */
pushl %eax
movl %ss, %eax
cmpw $__ESPFIX_SS, %ax
@@ -1550,6 +1628,11 @@ ENTRY(nmi)
movl %ebx, %esp
.Lnmi_return:
+#ifdef CONFIG_X86_ESPFIX32
+ testl $CS_FROM_ESPFIX, PT_CS(%esp)
+ jnz .Lnmi_from_espfix
+#endif
+
CHECK_AND_APPLY_ESPFIX
RESTORE_ALL_NMI cr3_reg=%edi pop=4
jmp .Lirq_return
@@ -1557,28 +1640,47 @@ ENTRY(nmi)
#ifdef CONFIG_X86_ESPFIX32
.Lnmi_espfix_stack:
/*
- * create the pointer to lss back
+ * Create the pointer to LSS back
*/
pushl %ss
pushl %esp
addl $4, (%esp)
- /* copy the iret frame of 12 bytes */
- .rept 3
- pushl 16(%esp)
- .endr
- pushl %eax
- SAVE_ALL_NMI cr3_reg=%edi
+
+ /* Copy the (short) IRET frame */
+ pushl 4*4(%esp) # flags
+ pushl 4*4(%esp) # cs
+ pushl 4*4(%esp) # ip
+
+ pushl %eax # orig_ax
+
+ SAVE_ALL_NMI cr3_reg=%edi unwind_espfix=1
ENCODE_FRAME_POINTER
- FIXUP_ESPFIX_STACK # %eax == %esp
+
+ /* clear CS_FROM_KERNEL, set CS_FROM_ESPFIX */
+ xorl $(CS_FROM_ESPFIX | CS_FROM_KERNEL), PT_CS(%esp)
+
xorl %edx, %edx # zero error code
- call do_nmi
+ movl %esp, %eax # pt_regs pointer
+ jmp .Lnmi_from_sysenter_stack
+
+.Lnmi_from_espfix:
RESTORE_ALL_NMI cr3_reg=%edi
- lss 12+4(%esp), %esp # back to espfix stack
+ /*
+ * Because we cleared CS_FROM_KERNEL, IRET_FRAME 'forgot' to
+ * fix up the gap and long frame:
+ *
+ * 3 - original frame (exception)
+ * 2 - ESPFIX block (above)
+ * 6 - gap (FIXUP_FRAME)
+ * 5 - long frame (FIXUP_FRAME)
+ * 1 - orig_ax
+ */
+ lss (1+5+6)*4(%esp), %esp # back to espfix stack
jmp .Lirq_return
#endif
-END(nmi)
+SYM_CODE_END(nmi)
-ENTRY(int3)
+SYM_CODE_START(int3)
ASM_CLAC
pushl $-1 # mark this as an int
@@ -1589,22 +1691,22 @@ ENTRY(int3)
movl %esp, %eax # pt_regs pointer
call do_int3
jmp ret_from_exception
-END(int3)
+SYM_CODE_END(int3)
-ENTRY(general_protection)
+SYM_CODE_START(general_protection)
pushl $do_general_protection
jmp common_exception
-END(general_protection)
+SYM_CODE_END(general_protection)
#ifdef CONFIG_KVM_GUEST
-ENTRY(async_page_fault)
+SYM_CODE_START(async_page_fault)
ASM_CLAC
pushl $do_async_page_fault
jmp common_exception_read_cr2
-END(async_page_fault)
+SYM_CODE_END(async_page_fault)
#endif
-ENTRY(rewind_stack_do_exit)
+SYM_CODE_START(rewind_stack_do_exit)
/* Prevent any naive code from trying to unwind to our caller. */
xorl %ebp, %ebp
@@ -1613,4 +1715,4 @@ ENTRY(rewind_stack_do_exit)
call do_exit
1: jmp 1b
-END(rewind_stack_do_exit)
+SYM_CODE_END(rewind_stack_do_exit)
diff --git a/arch/x86/entry/entry_64.S b/arch/x86/entry/entry_64.S
index b7c3ea4cb19d..76942cbd95a1 100644
--- a/arch/x86/entry/entry_64.S
+++ b/arch/x86/entry/entry_64.S
@@ -15,7 +15,7 @@
* at the top of the kernel process stack.
*
* Some macro usage:
- * - ENTRY/END: Define functions in the symbol table.
+ * - SYM_FUNC_START/END:Define functions in the symbol table.
* - TRACE_IRQ_*: Trace hardirq state for lock debugging.
* - idtentry: Define exception entry points.
*/
@@ -46,11 +46,11 @@
.section .entry.text, "ax"
#ifdef CONFIG_PARAVIRT
-ENTRY(native_usergs_sysret64)
+SYM_CODE_START(native_usergs_sysret64)
UNWIND_HINT_EMPTY
swapgs
sysretq
-END(native_usergs_sysret64)
+SYM_CODE_END(native_usergs_sysret64)
#endif /* CONFIG_PARAVIRT */
.macro TRACE_IRQS_FLAGS flags:req
@@ -142,7 +142,7 @@ END(native_usergs_sysret64)
* with them due to bugs in both AMD and Intel CPUs.
*/
-ENTRY(entry_SYSCALL_64)
+SYM_CODE_START(entry_SYSCALL_64)
UNWIND_HINT_EMPTY
/*
* Interrupts are off on entry.
@@ -162,7 +162,7 @@ ENTRY(entry_SYSCALL_64)
pushq %r11 /* pt_regs->flags */
pushq $__USER_CS /* pt_regs->cs */
pushq %rcx /* pt_regs->ip */
-GLOBAL(entry_SYSCALL_64_after_hwframe)
+SYM_INNER_LABEL(entry_SYSCALL_64_after_hwframe, SYM_L_GLOBAL)
pushq %rax /* pt_regs->orig_ax */
PUSH_AND_CLEAR_REGS rax=$-ENOSYS
@@ -273,13 +273,13 @@ syscall_return_via_sysret:
popq %rdi
popq %rsp
USERGS_SYSRET64
-END(entry_SYSCALL_64)
+SYM_CODE_END(entry_SYSCALL_64)
/*
* %rdi: prev task
* %rsi: next task
*/
-ENTRY(__switch_to_asm)
+SYM_CODE_START(__switch_to_asm)
UNWIND_HINT_FUNC
/*
* Save callee-saved registers
@@ -321,7 +321,7 @@ ENTRY(__switch_to_asm)
popq %rbp
jmp __switch_to
-END(__switch_to_asm)
+SYM_CODE_END(__switch_to_asm)
/*
* A newly forked process directly context switches into this address.
@@ -330,7 +330,7 @@ END(__switch_to_asm)
* rbx: kernel thread func (NULL for user thread)
* r12: kernel thread arg
*/
-ENTRY(ret_from_fork)
+SYM_CODE_START(ret_from_fork)
UNWIND_HINT_EMPTY
movq %rax, %rdi
call schedule_tail /* rdi: 'prev' task parameter */
@@ -357,14 +357,14 @@ ENTRY(ret_from_fork)
*/
movq $0, RAX(%rsp)
jmp 2b
-END(ret_from_fork)
+SYM_CODE_END(ret_from_fork)
/*
* Build the entry stubs with some assembler magic.
* We pack 1 stub into every 8-byte block.
*/
.align 8
-ENTRY(irq_entries_start)
+SYM_CODE_START(irq_entries_start)
vector=FIRST_EXTERNAL_VECTOR
.rept (FIRST_SYSTEM_VECTOR - FIRST_EXTERNAL_VECTOR)
UNWIND_HINT_IRET_REGS
@@ -373,10 +373,10 @@ ENTRY(irq_entries_start)
.align 8
vector=vector+1
.endr
-END(irq_entries_start)
+SYM_CODE_END(irq_entries_start)
.align 8
-ENTRY(spurious_entries_start)
+SYM_CODE_START(spurious_entries_start)
vector=FIRST_SYSTEM_VECTOR
.rept (NR_VECTORS - FIRST_SYSTEM_VECTOR)
UNWIND_HINT_IRET_REGS
@@ -385,7 +385,7 @@ ENTRY(spurious_entries_start)
.align 8
vector=vector+1
.endr
-END(spurious_entries_start)
+SYM_CODE_END(spurious_entries_start)
.macro DEBUG_ENTRY_ASSERT_IRQS_OFF
#ifdef CONFIG_DEBUG_ENTRY
@@ -511,7 +511,7 @@ END(spurious_entries_start)
* | return address |
* +----------------------------------------------------+
*/
-ENTRY(interrupt_entry)
+SYM_CODE_START(interrupt_entry)
UNWIND_HINT_FUNC
ASM_CLAC
cld
@@ -579,7 +579,7 @@ ENTRY(interrupt_entry)
TRACE_IRQS_OFF
ret
-END(interrupt_entry)
+SYM_CODE_END(interrupt_entry)
_ASM_NOKPROBE(interrupt_entry)
@@ -589,18 +589,18 @@ _ASM_NOKPROBE(interrupt_entry)
* The interrupt stubs push (~vector+0x80) onto the stack and
* then jump to common_spurious/interrupt.
*/
-common_spurious:
+SYM_CODE_START_LOCAL(common_spurious)
addq $-0x80, (%rsp) /* Adjust vector to [-256, -1] range */
call interrupt_entry
UNWIND_HINT_REGS indirect=1
call smp_spurious_interrupt /* rdi points to pt_regs */
jmp ret_from_intr
-END(common_spurious)
+SYM_CODE_END(common_spurious)
_ASM_NOKPROBE(common_spurious)
/* common_interrupt is a hotpath. Align it */
.p2align CONFIG_X86_L1_CACHE_SHIFT
-common_interrupt:
+SYM_CODE_START_LOCAL(common_interrupt)
addq $-0x80, (%rsp) /* Adjust vector to [-256, -1] range */
call interrupt_entry
UNWIND_HINT_REGS indirect=1
@@ -616,12 +616,12 @@ ret_from_intr:
jz retint_kernel
/* Interrupt came from user space */
-GLOBAL(retint_user)
+.Lretint_user:
mov %rsp,%rdi
call prepare_exit_to_usermode
TRACE_IRQS_IRETQ
-GLOBAL(swapgs_restore_regs_and_return_to_usermode)
+SYM_INNER_LABEL(swapgs_restore_regs_and_return_to_usermode, SYM_L_GLOBAL)
#ifdef CONFIG_DEBUG_ENTRY
/* Assert that pt_regs indicates user mode. */
testb $3, CS(%rsp)
@@ -679,7 +679,7 @@ retint_kernel:
*/
TRACE_IRQS_IRETQ
-GLOBAL(restore_regs_and_return_to_kernel)
+SYM_INNER_LABEL(restore_regs_and_return_to_kernel, SYM_L_GLOBAL)
#ifdef CONFIG_DEBUG_ENTRY
/* Assert that pt_regs indicates kernel mode. */
testb $3, CS(%rsp)
@@ -695,7 +695,7 @@ GLOBAL(restore_regs_and_return_to_kernel)
*/
INTERRUPT_RETURN
-ENTRY(native_iret)
+SYM_INNER_LABEL_ALIGN(native_iret, SYM_L_GLOBAL)
UNWIND_HINT_IRET_REGS
/*
* Are we returning to a stack segment from the LDT? Note: in
@@ -706,8 +706,7 @@ ENTRY(native_iret)
jnz native_irq_return_ldt
#endif
-.global native_irq_return_iret
-native_irq_return_iret:
+SYM_INNER_LABEL(native_irq_return_iret, SYM_L_GLOBAL)
/*
* This may fault. Non-paranoid faults on return to userspace are
* handled by fixup_bad_iret. These include #SS, #GP, and #NP.
@@ -789,14 +788,14 @@ native_irq_return_ldt:
*/
jmp native_irq_return_iret
#endif
-END(common_interrupt)
+SYM_CODE_END(common_interrupt)
_ASM_NOKPROBE(common_interrupt)
/*
* APIC interrupts.
*/
.macro apicinterrupt3 num sym do_sym
-ENTRY(\sym)
+SYM_CODE_START(\sym)
UNWIND_HINT_IRET_REGS
pushq $~(\num)
.Lcommon_\sym:
@@ -804,7 +803,7 @@ ENTRY(\sym)
UNWIND_HINT_REGS indirect=1
call \do_sym /* rdi points to pt_regs */
jmp ret_from_intr
-END(\sym)
+SYM_CODE_END(\sym)
_ASM_NOKPROBE(\sym)
.endm
@@ -969,7 +968,7 @@ apicinterrupt IRQ_WORK_VECTOR irq_work_interrupt smp_irq_work_interrupt
* #DF: if the thread stack is somehow unusable, we'll still get a useful OOPS.
*/
.macro idtentry sym do_sym has_error_code:req paranoid=0 shift_ist=-1 ist_offset=0 create_gap=0 read_cr2=0
-ENTRY(\sym)
+SYM_CODE_START(\sym)
UNWIND_HINT_IRET_REGS offset=\has_error_code*8
/* Sanity check */
@@ -1019,7 +1018,7 @@ ENTRY(\sym)
.endif
_ASM_NOKPROBE(\sym)
-END(\sym)
+SYM_CODE_END(\sym)
.endm
idtentry divide_error do_divide_error has_error_code=0
@@ -1041,7 +1040,7 @@ idtentry simd_coprocessor_error do_simd_coprocessor_error has_error_code=0
* Reload gs selector with exception handling
* edi: new selector
*/
-ENTRY(native_load_gs_index)
+SYM_FUNC_START(native_load_gs_index)
FRAME_BEGIN
pushfq
DISABLE_INTERRUPTS(CLBR_ANY & ~CLBR_RDI)
@@ -1055,13 +1054,13 @@ ENTRY(native_load_gs_index)
popfq
FRAME_END
ret
-ENDPROC(native_load_gs_index)
+SYM_FUNC_END(native_load_gs_index)
EXPORT_SYMBOL(native_load_gs_index)
_ASM_EXTABLE(.Lgs_change, .Lbad_gs)
.section .fixup, "ax"
/* running with kernelgs */
-.Lbad_gs:
+SYM_CODE_START_LOCAL_NOALIGN(.Lbad_gs)
SWAPGS /* switch back to user gs */
.macro ZAP_GS
/* This can't be a string because the preprocessor needs to see it. */
@@ -1072,10 +1071,11 @@ EXPORT_SYMBOL(native_load_gs_index)
xorl %eax, %eax
movl %eax, %gs
jmp 2b
+SYM_CODE_END(.Lbad_gs)
.previous
/* Call softirq on interrupt stack. Interrupts are off. */
-ENTRY(do_softirq_own_stack)
+SYM_FUNC_START(do_softirq_own_stack)
pushq %rbp
mov %rsp, %rbp
ENTER_IRQ_STACK regs=0 old_rsp=%r11
@@ -1083,7 +1083,7 @@ ENTRY(do_softirq_own_stack)
LEAVE_IRQ_STACK regs=0
leaveq
ret
-ENDPROC(do_softirq_own_stack)
+SYM_FUNC_END(do_softirq_own_stack)
#ifdef CONFIG_XEN_PV
idtentry hypervisor_callback xen_do_hypervisor_callback has_error_code=0
@@ -1101,7 +1101,8 @@ idtentry hypervisor_callback xen_do_hypervisor_callback has_error_code=0
* existing activation in its critical region -- if so, we pop the current
* activation and restart the handler using the previous one.
*/
-ENTRY(xen_do_hypervisor_callback) /* do_hypervisor_callback(struct *pt_regs) */
+/* do_hypervisor_callback(struct *pt_regs) */
+SYM_CODE_START_LOCAL(xen_do_hypervisor_callback)
/*
* Since we don't modify %rdi, evtchn_do_upall(struct *pt_regs) will
@@ -1119,7 +1120,7 @@ ENTRY(xen_do_hypervisor_callback) /* do_hypervisor_callback(struct *pt_regs) */
call xen_maybe_preempt_hcall
#endif
jmp error_exit
-END(xen_do_hypervisor_callback)
+SYM_CODE_END(xen_do_hypervisor_callback)
/*
* Hypervisor uses this for application faults while it executes.
@@ -1134,7 +1135,7 @@ END(xen_do_hypervisor_callback)
* We distinguish between categories by comparing each saved segment register
* with its current contents: any discrepancy means we in category 1.
*/
-ENTRY(xen_failsafe_callback)
+SYM_CODE_START(xen_failsafe_callback)
UNWIND_HINT_EMPTY
movl %ds, %ecx
cmpw %cx, 0x10(%rsp)
@@ -1164,7 +1165,7 @@ ENTRY(xen_failsafe_callback)
PUSH_AND_CLEAR_REGS
ENCODE_FRAME_POINTER
jmp error_exit
-END(xen_failsafe_callback)
+SYM_CODE_END(xen_failsafe_callback)
#endif /* CONFIG_XEN_PV */
#ifdef CONFIG_XEN_PVHVM
@@ -1214,7 +1215,7 @@ idtentry machine_check do_mce has_error_code=0 paranoid=1
* Use slow, but surefire "are we in kernel?" check.
* Return: ebx=0: need swapgs on exit, ebx=1: otherwise
*/
-ENTRY(paranoid_entry)
+SYM_CODE_START_LOCAL(paranoid_entry)
UNWIND_HINT_FUNC
cld
PUSH_AND_CLEAR_REGS save_ret=1
@@ -1248,7 +1249,7 @@ ENTRY(paranoid_entry)
FENCE_SWAPGS_KERNEL_ENTRY
ret
-END(paranoid_entry)
+SYM_CODE_END(paranoid_entry)
/*
* "Paranoid" exit path from exception stack. This is invoked
@@ -1262,7 +1263,7 @@ END(paranoid_entry)
*
* On entry, ebx is "no swapgs" flag (1: don't need swapgs, 0: need it)
*/
-ENTRY(paranoid_exit)
+SYM_CODE_START_LOCAL(paranoid_exit)
UNWIND_HINT_REGS
DISABLE_INTERRUPTS(CLBR_ANY)
TRACE_IRQS_OFF_DEBUG
@@ -1272,19 +1273,18 @@ ENTRY(paranoid_exit)
/* Always restore stashed CR3 value (see paranoid_entry) */
RESTORE_CR3 scratch_reg=%rbx save_reg=%r14
SWAPGS_UNSAFE_STACK
- jmp .Lparanoid_exit_restore
+ jmp restore_regs_and_return_to_kernel
.Lparanoid_exit_no_swapgs:
TRACE_IRQS_IRETQ_DEBUG
/* Always restore stashed CR3 value (see paranoid_entry) */
RESTORE_CR3 scratch_reg=%rbx save_reg=%r14
-.Lparanoid_exit_restore:
jmp restore_regs_and_return_to_kernel
-END(paranoid_exit)
+SYM_CODE_END(paranoid_exit)
/*
* Save all registers in pt_regs, and switch GS if needed.
*/
-ENTRY(error_entry)
+SYM_CODE_START_LOCAL(error_entry)
UNWIND_HINT_FUNC
cld
PUSH_AND_CLEAR_REGS save_ret=1
@@ -1364,16 +1364,16 @@ ENTRY(error_entry)
call fixup_bad_iret
mov %rax, %rsp
jmp .Lerror_entry_from_usermode_after_swapgs
-END(error_entry)
+SYM_CODE_END(error_entry)
-ENTRY(error_exit)
+SYM_CODE_START_LOCAL(error_exit)
UNWIND_HINT_REGS
DISABLE_INTERRUPTS(CLBR_ANY)
TRACE_IRQS_OFF
testb $3, CS(%rsp)
jz retint_kernel
- jmp retint_user
-END(error_exit)
+ jmp .Lretint_user
+SYM_CODE_END(error_exit)
/*
* Runs on exception stack. Xen PV does not go through this path at all,
@@ -1383,7 +1383,7 @@ END(error_exit)
* %r14: Used to save/restore the CR3 of the interrupted context
* when PAGE_TABLE_ISOLATION is in use. Do not clobber.
*/
-ENTRY(nmi)
+SYM_CODE_START(nmi)
UNWIND_HINT_IRET_REGS
/*
@@ -1718,21 +1718,21 @@ nmi_restore:
* about espfix64 on the way back to kernel mode.
*/
iretq
-END(nmi)
+SYM_CODE_END(nmi)
#ifndef CONFIG_IA32_EMULATION
/*
* This handles SYSCALL from 32-bit code. There is no way to program
* MSRs to fully disable 32-bit SYSCALL.
*/
-ENTRY(ignore_sysret)
+SYM_CODE_START(ignore_sysret)
UNWIND_HINT_EMPTY
mov $-ENOSYS, %eax
sysret
-END(ignore_sysret)
+SYM_CODE_END(ignore_sysret)
#endif
-ENTRY(rewind_stack_do_exit)
+SYM_CODE_START(rewind_stack_do_exit)
UNWIND_HINT_FUNC
/* Prevent any naive code from trying to unwind to our caller. */
xorl %ebp, %ebp
@@ -1742,4 +1742,4 @@ ENTRY(rewind_stack_do_exit)
UNWIND_HINT_FUNC sp_offset=PTREGS_SIZE
call do_exit
-END(rewind_stack_do_exit)
+SYM_CODE_END(rewind_stack_do_exit)
diff --git a/arch/x86/entry/entry_64_compat.S b/arch/x86/entry/entry_64_compat.S
index 39913770a44d..f1d3ccae5dd5 100644
--- a/arch/x86/entry/entry_64_compat.S
+++ b/arch/x86/entry/entry_64_compat.S
@@ -46,7 +46,7 @@
* ebp user stack
* 0(%ebp) arg6
*/
-ENTRY(entry_SYSENTER_compat)
+SYM_FUNC_START(entry_SYSENTER_compat)
/* Interrupts are off on entry. */
SWAPGS
@@ -146,8 +146,8 @@ ENTRY(entry_SYSENTER_compat)
pushq $X86_EFLAGS_FIXED
popfq
jmp .Lsysenter_flags_fixed
-GLOBAL(__end_entry_SYSENTER_compat)
-ENDPROC(entry_SYSENTER_compat)
+SYM_INNER_LABEL(__end_entry_SYSENTER_compat, SYM_L_GLOBAL)
+SYM_FUNC_END(entry_SYSENTER_compat)
/*
* 32-bit SYSCALL entry.
@@ -196,7 +196,7 @@ ENDPROC(entry_SYSENTER_compat)
* esp user stack
* 0(%esp) arg6
*/
-ENTRY(entry_SYSCALL_compat)
+SYM_CODE_START(entry_SYSCALL_compat)
/* Interrupts are off on entry. */
swapgs
@@ -215,7 +215,7 @@ ENTRY(entry_SYSCALL_compat)
pushq %r11 /* pt_regs->flags */
pushq $__USER32_CS /* pt_regs->cs */
pushq %rcx /* pt_regs->ip */
-GLOBAL(entry_SYSCALL_compat_after_hwframe)
+SYM_INNER_LABEL(entry_SYSCALL_compat_after_hwframe, SYM_L_GLOBAL)
movl %eax, %eax /* discard orig_ax high bits */
pushq %rax /* pt_regs->orig_ax */
pushq %rdi /* pt_regs->di */
@@ -311,7 +311,7 @@ sysret32_from_system_call:
xorl %r10d, %r10d
swapgs
sysretl
-END(entry_SYSCALL_compat)
+SYM_CODE_END(entry_SYSCALL_compat)
/*
* 32-bit legacy system call entry.
@@ -339,7 +339,7 @@ END(entry_SYSCALL_compat)
* edi arg5
* ebp arg6
*/
-ENTRY(entry_INT80_compat)
+SYM_CODE_START(entry_INT80_compat)
/*
* Interrupts are off on entry.
*/
@@ -416,4 +416,4 @@ ENTRY(entry_INT80_compat)
/* Go back to user mode. */
TRACE_IRQS_ON
jmp swapgs_restore_regs_and_return_to_usermode
-END(entry_INT80_compat)
+SYM_CODE_END(entry_INT80_compat)
diff --git a/arch/x86/entry/syscall_32.c b/arch/x86/entry/syscall_32.c
index aa3336a7cb15..7d17b3addbbb 100644
--- a/arch/x86/entry/syscall_32.c
+++ b/arch/x86/entry/syscall_32.c
@@ -10,13 +10,11 @@
#ifdef CONFIG_IA32_EMULATION
/* On X86_64, we use struct pt_regs * to pass parameters to syscalls */
#define __SYSCALL_I386(nr, sym, qual) extern asmlinkage long sym(const struct pt_regs *);
-
-/* this is a lie, but it does not hurt as sys_ni_syscall just returns -EINVAL */
-extern asmlinkage long sys_ni_syscall(const struct pt_regs *);
-
+#define __sys_ni_syscall __ia32_sys_ni_syscall
#else /* CONFIG_IA32_EMULATION */
#define __SYSCALL_I386(nr, sym, qual) extern asmlinkage long sym(unsigned long, unsigned long, unsigned long, unsigned long, unsigned long, unsigned long);
extern asmlinkage long sys_ni_syscall(unsigned long, unsigned long, unsigned long, unsigned long, unsigned long, unsigned long);
+#define __sys_ni_syscall sys_ni_syscall
#endif /* CONFIG_IA32_EMULATION */
#include <asm/syscalls_32.h>
@@ -29,6 +27,6 @@ __visible const sys_call_ptr_t ia32_sys_call_table[__NR_syscall_compat_max+1] =
* Smells like a compiler bug -- it doesn't work
* when the & below is removed.
*/
- [0 ... __NR_syscall_compat_max] = &sys_ni_syscall,
+ [0 ... __NR_syscall_compat_max] = &__sys_ni_syscall,
#include <asm/syscalls_32.h>
};
diff --git a/arch/x86/entry/syscall_64.c b/arch/x86/entry/syscall_64.c
index b1bf31713374..adf619a856e8 100644
--- a/arch/x86/entry/syscall_64.c
+++ b/arch/x86/entry/syscall_64.c
@@ -4,11 +4,17 @@
#include <linux/linkage.h>
#include <linux/sys.h>
#include <linux/cache.h>
+#include <linux/syscalls.h>
#include <asm/asm-offsets.h>
#include <asm/syscall.h>
-/* this is a lie, but it does not hurt as sys_ni_syscall just returns -EINVAL */
-extern asmlinkage long sys_ni_syscall(const struct pt_regs *);
+extern asmlinkage long sys_ni_syscall(void);
+
+SYSCALL_DEFINE0(ni_syscall)
+{
+ return sys_ni_syscall();
+}
+
#define __SYSCALL_64(nr, sym, qual) extern asmlinkage long sym(const struct pt_regs *);
#define __SYSCALL_X32(nr, sym, qual) __SYSCALL_64(nr, sym, qual)
#include <asm/syscalls_64.h>
@@ -23,7 +29,7 @@ asmlinkage const sys_call_ptr_t sys_call_table[__NR_syscall_max+1] = {
* Smells like a compiler bug -- it doesn't work
* when the & below is removed.
*/
- [0 ... __NR_syscall_max] = &sys_ni_syscall,
+ [0 ... __NR_syscall_max] = &__x64_sys_ni_syscall,
#include <asm/syscalls_64.h>
};
@@ -40,7 +46,7 @@ asmlinkage const sys_call_ptr_t x32_sys_call_table[__NR_syscall_x32_max+1] = {
* Smells like a compiler bug -- it doesn't work
* when the & below is removed.
*/
- [0 ... __NR_syscall_x32_max] = &sys_ni_syscall,
+ [0 ... __NR_syscall_x32_max] = &__x64_sys_ni_syscall,
#include <asm/syscalls_64.h>
};
diff --git a/arch/x86/entry/syscalls/syscall_32.tbl b/arch/x86/entry/syscalls/syscall_32.tbl
index 3fe02546aed3..15908eb9b17e 100644
--- a/arch/x86/entry/syscalls/syscall_32.tbl
+++ b/arch/x86/entry/syscalls/syscall_32.tbl
@@ -124,13 +124,13 @@
110 i386 iopl sys_iopl __ia32_sys_iopl
111 i386 vhangup sys_vhangup __ia32_sys_vhangup
112 i386 idle
-113 i386 vm86old sys_vm86old sys_ni_syscall
+113 i386 vm86old sys_vm86old __ia32_sys_ni_syscall
114 i386 wait4 sys_wait4 __ia32_compat_sys_wait4
115 i386 swapoff sys_swapoff __ia32_sys_swapoff
116 i386 sysinfo sys_sysinfo __ia32_compat_sys_sysinfo
117 i386 ipc sys_ipc __ia32_compat_sys_ipc
118 i386 fsync sys_fsync __ia32_sys_fsync
-119 i386 sigreturn sys_sigreturn sys32_sigreturn
+119 i386 sigreturn sys_sigreturn __ia32_compat_sys_sigreturn
120 i386 clone sys_clone __ia32_compat_sys_x86_clone
121 i386 setdomainname sys_setdomainname __ia32_sys_setdomainname
122 i386 uname sys_newuname __ia32_sys_newuname
@@ -177,14 +177,14 @@
163 i386 mremap sys_mremap __ia32_sys_mremap
164 i386 setresuid sys_setresuid16 __ia32_sys_setresuid16
165 i386 getresuid sys_getresuid16 __ia32_sys_getresuid16
-166 i386 vm86 sys_vm86 sys_ni_syscall
+166 i386 vm86 sys_vm86 __ia32_sys_ni_syscall
167 i386 query_module
168 i386 poll sys_poll __ia32_sys_poll
169 i386 nfsservctl
170 i386 setresgid sys_setresgid16 __ia32_sys_setresgid16
171 i386 getresgid sys_getresgid16 __ia32_sys_getresgid16
172 i386 prctl sys_prctl __ia32_sys_prctl
-173 i386 rt_sigreturn sys_rt_sigreturn sys32_rt_sigreturn
+173 i386 rt_sigreturn sys_rt_sigreturn __ia32_compat_sys_rt_sigreturn
174 i386 rt_sigaction sys_rt_sigaction __ia32_compat_sys_rt_sigaction
175 i386 rt_sigprocmask sys_rt_sigprocmask __ia32_compat_sys_rt_sigprocmask
176 i386 rt_sigpending sys_rt_sigpending __ia32_compat_sys_rt_sigpending
diff --git a/arch/x86/entry/thunk_32.S b/arch/x86/entry/thunk_32.S
index 2713490611a3..e010d4ae11f1 100644
--- a/arch/x86/entry/thunk_32.S
+++ b/arch/x86/entry/thunk_32.S
@@ -10,8 +10,7 @@
/* put return address in eax (arg1) */
.macro THUNK name, func, put_ret_addr_in_eax=0
- .globl \name
-\name:
+SYM_CODE_START_NOALIGN(\name)
pushl %eax
pushl %ecx
pushl %edx
@@ -27,6 +26,7 @@
popl %eax
ret
_ASM_NOKPROBE(\name)
+SYM_CODE_END(\name)
.endm
#ifdef CONFIG_TRACE_IRQFLAGS
diff --git a/arch/x86/entry/thunk_64.S b/arch/x86/entry/thunk_64.S
index ea5c4167086c..c5c3b6e86e62 100644
--- a/arch/x86/entry/thunk_64.S
+++ b/arch/x86/entry/thunk_64.S
@@ -12,7 +12,7 @@
/* rdi: arg1 ... normal C conventions. rax is saved/restored. */
.macro THUNK name, func, put_ret_addr_in_rdi=0
- ENTRY(\name)
+SYM_FUNC_START_NOALIGN(\name)
pushq %rbp
movq %rsp, %rbp
@@ -33,7 +33,7 @@
call \func
jmp .L_restore
- ENDPROC(\name)
+SYM_FUNC_END(\name)
_ASM_NOKPROBE(\name)
.endm
@@ -56,7 +56,7 @@
#if defined(CONFIG_TRACE_IRQFLAGS) \
|| defined(CONFIG_DEBUG_LOCK_ALLOC) \
|| defined(CONFIG_PREEMPTION)
-.L_restore:
+SYM_CODE_START_LOCAL_NOALIGN(.L_restore)
popq %r11
popq %r10
popq %r9
@@ -69,4 +69,5 @@
popq %rbp
ret
_ASM_NOKPROBE(.L_restore)
+SYM_CODE_END(.L_restore)
#endif
diff --git a/arch/x86/entry/vdso/Makefile b/arch/x86/entry/vdso/Makefile
index 0f2154106d01..2b75e80f6b41 100644
--- a/arch/x86/entry/vdso/Makefile
+++ b/arch/x86/entry/vdso/Makefile
@@ -87,11 +87,9 @@ $(vobjs): KBUILD_CFLAGS := $(filter-out $(GCC_PLUGINS_CFLAGS) $(RETPOLINE_CFLAGS
#
# vDSO code runs in userspace and -pg doesn't help with profiling anyway.
#
-CFLAGS_REMOVE_vdso-note.o = -pg
CFLAGS_REMOVE_vclock_gettime.o = -pg
CFLAGS_REMOVE_vdso32/vclock_gettime.o = -pg
CFLAGS_REMOVE_vgetcpu.o = -pg
-CFLAGS_REMOVE_vvar.o = -pg
#
# X32 processes use x32 vDSO to access 64bit kernel data.
diff --git a/arch/x86/entry/vdso/vclock_gettime.c b/arch/x86/entry/vdso/vclock_gettime.c
index d9ff616bb0f6..7d70935b6758 100644
--- a/arch/x86/entry/vdso/vclock_gettime.c
+++ b/arch/x86/entry/vdso/vclock_gettime.c
@@ -15,7 +15,7 @@
#include "../../../../lib/vdso/gettimeofday.c"
extern int __vdso_gettimeofday(struct __kernel_old_timeval *tv, struct timezone *tz);
-extern time_t __vdso_time(time_t *t);
+extern __kernel_old_time_t __vdso_time(__kernel_old_time_t *t);
int __vdso_gettimeofday(struct __kernel_old_timeval *tv, struct timezone *tz)
{
@@ -25,12 +25,12 @@ int __vdso_gettimeofday(struct __kernel_old_timeval *tv, struct timezone *tz)
int gettimeofday(struct __kernel_old_timeval *, struct timezone *)
__attribute__((weak, alias("__vdso_gettimeofday")));
-time_t __vdso_time(time_t *t)
+__kernel_old_time_t __vdso_time(__kernel_old_time_t *t)
{
return __cvdso_time(t);
}
-time_t time(time_t *t) __attribute__((weak, alias("__vdso_time")));
+__kernel_old_time_t time(__kernel_old_time_t *t) __attribute__((weak, alias("__vdso_time")));
#if defined(CONFIG_X86_64) && !defined(BUILD_VDSO32_64)
diff --git a/arch/x86/entry/vdso/vdso32/system_call.S b/arch/x86/entry/vdso/vdso32/system_call.S
index 263d7433dea8..de1fff7188aa 100644
--- a/arch/x86/entry/vdso/vdso32/system_call.S
+++ b/arch/x86/entry/vdso/vdso32/system_call.S
@@ -62,7 +62,7 @@ __kernel_vsyscall:
/* Enter using int $0x80 */
int $0x80
-GLOBAL(int80_landing_pad)
+SYM_INNER_LABEL(int80_landing_pad, SYM_L_GLOBAL)
/*
* Restore EDX and ECX in case they were clobbered. EBP is not
diff --git a/arch/x86/entry/vsyscall/vsyscall_64.c b/arch/x86/entry/vsyscall/vsyscall_64.c
index e7c596dea947..44c33103a955 100644
--- a/arch/x86/entry/vsyscall/vsyscall_64.c
+++ b/arch/x86/entry/vsyscall/vsyscall_64.c
@@ -184,7 +184,7 @@ bool emulate_vsyscall(unsigned long error_code,
*/
switch (vsyscall_nr) {
case 0:
- if (!write_ok_or_segv(regs->di, sizeof(struct timeval)) ||
+ if (!write_ok_or_segv(regs->di, sizeof(struct __kernel_old_timeval)) ||
!write_ok_or_segv(regs->si, sizeof(struct timezone))) {
ret = -EFAULT;
goto check_fault;
@@ -194,7 +194,7 @@ bool emulate_vsyscall(unsigned long error_code,
break;
case 1:
- if (!write_ok_or_segv(regs->di, sizeof(time_t))) {
+ if (!write_ok_or_segv(regs->di, sizeof(__kernel_old_time_t))) {
ret = -EFAULT;
goto check_fault;
}
@@ -222,7 +222,7 @@ bool emulate_vsyscall(unsigned long error_code,
*/
regs->orig_ax = syscall_nr;
regs->ax = -ENOSYS;
- tmp = secure_computing(NULL);
+ tmp = secure_computing();
if ((!tmp && regs->orig_ax != syscall_nr) || regs->ip != address) {
warn_bad_vsyscall(KERN_DEBUG, regs,
"seccomp tried to change syscall nr or ip");
Copyright © 1996 – 2023 Jason A. Donenfeld. All Rights Reverse Engineered.