diff options
Diffstat (limited to 'arch/x86/lib/copy_user_64.S')
| -rw-r--r-- | arch/x86/lib/copy_user_64.S | 139 |
1 files changed, 88 insertions, 51 deletions
diff --git a/arch/x86/lib/copy_user_64.S b/arch/x86/lib/copy_user_64.S index a460158b5ac..dee945d5559 100644 --- a/arch/x86/lib/copy_user_64.S +++ b/arch/x86/lib/copy_user_64.S @@ -15,23 +15,32 @@ #include <asm/asm-offsets.h> #include <asm/thread_info.h> #include <asm/cpufeature.h> +#include <asm/alternative-asm.h> +#include <asm/asm.h> +#include <asm/smap.h> - .macro ALTERNATIVE_JUMP feature,orig,alt +/* + * By placing feature2 after feature1 in altinstructions section, we logically + * implement: + * If CPU has feature2, jmp to alt2 is used + * else if CPU has feature1, jmp to alt1 is used + * else jmp to orig is used. + */ + .macro ALTERNATIVE_JUMP feature1,feature2,orig,alt1,alt2 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 */ + .long \alt1-1b /* offset */ /* or alternatively to alt1 */ +3: .byte 0xe9 /* near jump with 32bit immediate */ + .long \alt2-1b /* offset */ /* or alternatively to alt2 */ .previous + .section .altinstructions,"a" - .align 8 - .quad 0b - .quad 2b - .word \feature /* when feature is set */ - .byte 5 - .byte 5 + altinstruction_entry 0b,2b,\feature1,5,5 + altinstruction_entry 0b,3b,\feature2,5,5 .previous .endm @@ -56,11 +65,8 @@ jmp copy_user_handle_tail .previous - .section __ex_table,"a" - .align 8 - .quad 100b,103b - .quad 101b,103b - .previous + _ASM_EXTABLE(100b,103b) + _ASM_EXTABLE(101b,103b) #endif .endm @@ -72,8 +78,10 @@ ENTRY(_copy_to_user) 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 + ja bad_to_user + ALTERNATIVE_JUMP X86_FEATURE_REP_GOOD,X86_FEATURE_ERMS, \ + copy_user_generic_unrolled,copy_user_generic_string, \ + copy_user_enhanced_fast_string CFI_ENDPROC ENDPROC(_copy_to_user) @@ -85,8 +93,10 @@ ENTRY(_copy_from_user) 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 + ja bad_from_user + ALTERNATIVE_JUMP X86_FEATURE_REP_GOOD,X86_FEATURE_ERMS, \ + copy_user_generic_unrolled,copy_user_generic_string, \ + copy_user_enhanced_fast_string CFI_ENDPROC ENDPROC(_copy_from_user) @@ -117,10 +127,11 @@ ENDPROC(bad_from_user) * rdx count * * Output: - * eax uncopied bytes or 0 if successfull. + * eax uncopied bytes or 0 if successful. */ ENTRY(copy_user_generic_unrolled) CFI_STARTPROC + ASM_STAC cmpl $8,%edx jb 20f /* less then 8 bytes, go to byte copy loop */ ALIGN_DESTINATION @@ -168,41 +179,39 @@ ENTRY(copy_user_generic_unrolled) decl %ecx jnz 21b 23: xor %eax,%eax + ASM_CLAC ret .section .fixup,"ax" 30: shll $6,%ecx addl %ecx,%edx jmp 60f -40: lea (%rdx,%rcx,8),%rdx +40: leal (%rdx,%rcx,8),%edx 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 + _ASM_EXTABLE(1b,30b) + _ASM_EXTABLE(2b,30b) + _ASM_EXTABLE(3b,30b) + _ASM_EXTABLE(4b,30b) + _ASM_EXTABLE(5b,30b) + _ASM_EXTABLE(6b,30b) + _ASM_EXTABLE(7b,30b) + _ASM_EXTABLE(8b,30b) + _ASM_EXTABLE(9b,30b) + _ASM_EXTABLE(10b,30b) + _ASM_EXTABLE(11b,30b) + _ASM_EXTABLE(12b,30b) + _ASM_EXTABLE(13b,30b) + _ASM_EXTABLE(14b,30b) + _ASM_EXTABLE(15b,30b) + _ASM_EXTABLE(16b,30b) + _ASM_EXTABLE(18b,40b) + _ASM_EXTABLE(19b,40b) + _ASM_EXTABLE(21b,50b) + _ASM_EXTABLE(22b,50b) CFI_ENDPROC ENDPROC(copy_user_generic_unrolled) @@ -226,8 +235,7 @@ ENDPROC(copy_user_generic_unrolled) */ ENTRY(copy_user_generic_string) CFI_STARTPROC - andl %edx,%edx - jz 4f + ASM_STAC cmpl $8,%edx jb 2f /* less than 8 bytes, go to byte copy loop */ ALIGN_DESTINATION @@ -239,19 +247,48 @@ ENTRY(copy_user_generic_string) 2: movl %edx,%ecx 3: rep movsb -4: xorl %eax,%eax + xorl %eax,%eax + ASM_CLAC ret .section .fixup,"ax" -11: lea (%rdx,%rcx,8),%rcx +11: leal (%rdx,%rcx,8),%ecx 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 + _ASM_EXTABLE(1b,11b) + _ASM_EXTABLE(3b,12b) CFI_ENDPROC ENDPROC(copy_user_generic_string) + +/* + * Some CPUs are adding enhanced REP MOVSB/STOSB instructions. + * It's recommended to use enhanced REP MOVSB/STOSB if it's enabled. + * + * Input: + * rdi destination + * rsi source + * rdx count + * + * Output: + * eax uncopied bytes or 0 if successful. + */ +ENTRY(copy_user_enhanced_fast_string) + CFI_STARTPROC + ASM_STAC + movl %edx,%ecx +1: rep + movsb + xorl %eax,%eax + ASM_CLAC + ret + + .section .fixup,"ax" +12: movl %ecx,%edx /* ecx is zerorest also */ + jmp copy_user_handle_tail + .previous + + _ASM_EXTABLE(1b,12b) + CFI_ENDPROC +ENDPROC(copy_user_enhanced_fast_string) |