linux/arch/x86/lib/copy_user_64.S
Jan Beulich 1b1d925818 x86-64: Modify copy_user_generic() alternatives mechanism
In order to avoid unnecessary chains of branches, rather than
implementing copy_user_generic() as a function consisting of
just a single (possibly patched) branch, instead properly deal
with patching call instructions in the alternative instructions
framework, and move the patching into the callers.

As a follow-on, one could also introduce something like
__EXPORT_SYMBOL_ALT() to avoid patching call sites in modules.

Signed-off-by: Jan Beulich <jbeulich@novell.com>
Cc: Nick Piggin <npiggin@suse.de>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
LKML-Reference: <4B2BB8180200007800026AE7@vpn.id2.novell.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-12-30 11:57:31 +01:00

258 lines
5.0 KiB
ArmAsm

/*
* Copyright 2008 Vitaly Mayatskikh <vmayatsk@redhat.com>
* Copyright 2002 Andi Kleen, SuSE Labs.
* Subject to the GNU Public License v2.
*
* Functions to copy from and to user space.
*/
#include <linux/linkage.h>
#include <asm/dwarf2.h>
#define FIX_ALIGNMENT 1
#include <asm/current.h>
#include <asm/asm-offsets.h>
#include <asm/thread_info.h>
#include <asm/cpufeature.h>
.macro ALTERNATIVE_JUMP feature,orig,alt
0:
.byte 0xe9 /* 32bit jump */
.long \orig-1f /* by default jump to orig */
1:
.section .altinstr_replacement,"ax"
2: .byte 0xe9 /* near jump with 32bit immediate */
.long \alt-1b /* offset */ /* or alternatively to alt */
.previous
.section .altinstructions,"a"
.align 8
.quad 0b
.quad 2b
.byte \feature /* when feature is set */
.byte 5
.byte 5
.previous
.endm
.macro ALIGN_DESTINATION
#ifdef FIX_ALIGNMENT
/* check for bad alignment of destination */
movl %edi,%ecx
andl $7,%ecx
jz 102f /* already aligned */
subl $8,%ecx
negl %ecx
subl %ecx,%edx
100: movb (%rsi),%al
101: movb %al,(%rdi)
incq %rsi
incq %rdi
decl %ecx
jnz 100b
102:
.section .fixup,"ax"
103: addl %ecx,%edx /* ecx is zerorest also */
jmp copy_user_handle_tail
.previous
.section __ex_table,"a"
.align 8
.quad 100b,103b
.quad 101b,103b
.previous
#endif
.endm
/* Standard copy_to_user with segment limit checking */
ENTRY(_copy_to_user)
CFI_STARTPROC
GET_THREAD_INFO(%rax)
movq %rdi,%rcx
addq %rdx,%rcx
jc bad_to_user
cmpq TI_addr_limit(%rax),%rcx
jae bad_to_user
ALTERNATIVE_JUMP X86_FEATURE_REP_GOOD,copy_user_generic_unrolled,copy_user_generic_string
CFI_ENDPROC
ENDPROC(_copy_to_user)
/* Standard copy_from_user with segment limit checking */
ENTRY(_copy_from_user)
CFI_STARTPROC
GET_THREAD_INFO(%rax)
movq %rsi,%rcx
addq %rdx,%rcx
jc bad_from_user
cmpq TI_addr_limit(%rax),%rcx
jae bad_from_user
ALTERNATIVE_JUMP X86_FEATURE_REP_GOOD,copy_user_generic_unrolled,copy_user_generic_string
CFI_ENDPROC
ENDPROC(_copy_from_user)
.section .fixup,"ax"
/* must zero dest */
ENTRY(bad_from_user)
bad_from_user:
CFI_STARTPROC
movl %edx,%ecx
xorl %eax,%eax
rep
stosb
bad_to_user:
movl %edx,%eax
ret
CFI_ENDPROC
ENDPROC(bad_from_user)
.previous
/*
* copy_user_generic_unrolled - memory copy with exception handling.
* This version is for CPUs like P4 that don't have efficient micro
* code for rep movsq
*
* Input:
* rdi destination
* rsi source
* rdx count
*
* Output:
* eax uncopied bytes or 0 if successfull.
*/
ENTRY(copy_user_generic_unrolled)
CFI_STARTPROC
cmpl $8,%edx
jb 20f /* less then 8 bytes, go to byte copy loop */
ALIGN_DESTINATION
movl %edx,%ecx
andl $63,%edx
shrl $6,%ecx
jz 17f
1: movq (%rsi),%r8
2: movq 1*8(%rsi),%r9
3: movq 2*8(%rsi),%r10
4: movq 3*8(%rsi),%r11
5: movq %r8,(%rdi)
6: movq %r9,1*8(%rdi)
7: movq %r10,2*8(%rdi)
8: movq %r11,3*8(%rdi)
9: movq 4*8(%rsi),%r8
10: movq 5*8(%rsi),%r9
11: movq 6*8(%rsi),%r10
12: movq 7*8(%rsi),%r11
13: movq %r8,4*8(%rdi)
14: movq %r9,5*8(%rdi)
15: movq %r10,6*8(%rdi)
16: movq %r11,7*8(%rdi)
leaq 64(%rsi),%rsi
leaq 64(%rdi),%rdi
decl %ecx
jnz 1b
17: movl %edx,%ecx
andl $7,%edx
shrl $3,%ecx
jz 20f
18: movq (%rsi),%r8
19: movq %r8,(%rdi)
leaq 8(%rsi),%rsi
leaq 8(%rdi),%rdi
decl %ecx
jnz 18b
20: andl %edx,%edx
jz 23f
movl %edx,%ecx
21: movb (%rsi),%al
22: movb %al,(%rdi)
incq %rsi
incq %rdi
decl %ecx
jnz 21b
23: xor %eax,%eax
ret
.section .fixup,"ax"
30: shll $6,%ecx
addl %ecx,%edx
jmp 60f
40: lea (%rdx,%rcx,8),%rdx
jmp 60f
50: movl %ecx,%edx
60: jmp copy_user_handle_tail /* ecx is zerorest also */
.previous
.section __ex_table,"a"
.align 8
.quad 1b,30b
.quad 2b,30b
.quad 3b,30b
.quad 4b,30b
.quad 5b,30b
.quad 6b,30b
.quad 7b,30b
.quad 8b,30b
.quad 9b,30b
.quad 10b,30b
.quad 11b,30b
.quad 12b,30b
.quad 13b,30b
.quad 14b,30b
.quad 15b,30b
.quad 16b,30b
.quad 18b,40b
.quad 19b,40b
.quad 21b,50b
.quad 22b,50b
.previous
CFI_ENDPROC
ENDPROC(copy_user_generic_unrolled)
/* Some CPUs run faster using the string copy instructions.
* This is also a lot simpler. Use them when possible.
*
* Only 4GB of copy is supported. This shouldn't be a problem
* because the kernel normally only writes from/to page sized chunks
* even if user space passed a longer buffer.
* And more would be dangerous because both Intel and AMD have
* errata with rep movsq > 4GB. If someone feels the need to fix
* this please consider this.
*
* Input:
* rdi destination
* rsi source
* rdx count
*
* Output:
* eax uncopied bytes or 0 if successful.
*/
ENTRY(copy_user_generic_string)
CFI_STARTPROC
andl %edx,%edx
jz 4f
cmpl $8,%edx
jb 2f /* less than 8 bytes, go to byte copy loop */
ALIGN_DESTINATION
movl %edx,%ecx
shrl $3,%ecx
andl $7,%edx
1: rep
movsq
2: movl %edx,%ecx
3: rep
movsb
4: xorl %eax,%eax
ret
.section .fixup,"ax"
11: lea (%rdx,%rcx,8),%rcx
12: movl %ecx,%edx /* ecx is zerorest also */
jmp copy_user_handle_tail
.previous
.section __ex_table,"a"
.align 8
.quad 1b,11b
.quad 3b,12b
.previous
CFI_ENDPROC
ENDPROC(copy_user_generic_string)