Fix avx2 strncmp offset compare condition check [BZ #25933]
Commit Message
From: Sunil K Pandey <skpgkp1@gmail.com>
strcmp-avx2.S: In avx2 strncmp function, strings are compared in
chunks of 4 vector size(i.e. 32x4=128 byte for avx2). After first 4
vector size comparison, code must check whether it already passed
the given offset. This patch implement avx2 offset check condition
for strncmp function, if both string compare same for first 4 vector
size.
---
sysdeps/x86_64/multiarch/strcmp-avx2.S | 15 +++++++++++++++
1 file changed, 15 insertions(+)
Comments
On Sat, Jun 13, 2020 at 4:23 AM H.J. Lu <hjl.tools@gmail.com> wrote:
>
> From: Sunil K Pandey <skpgkp1@gmail.com>
>
> strcmp-avx2.S: In avx2 strncmp function, strings are compared in
> chunks of 4 vector size(i.e. 32x4=128 byte for avx2). After first 4
> vector size comparison, code must check whether it already passed
> the given offset. This patch implement avx2 offset check condition
> for strncmp function, if both string compare same for first 4 vector
> size.
> ---
> sysdeps/x86_64/multiarch/strcmp-avx2.S | 15 +++++++++++++++
> 1 file changed, 15 insertions(+)
>
> diff --git a/sysdeps/x86_64/multiarch/strcmp-avx2.S b/sysdeps/x86_64/multiarch/strcmp-avx2.S
> index 48d03a9f46..ee82fa3e19 100644
> --- a/sysdeps/x86_64/multiarch/strcmp-avx2.S
> +++ b/sysdeps/x86_64/multiarch/strcmp-avx2.S
> @@ -591,7 +591,14 @@ L(loop_cross_page_2_vec):
> movl $(PAGE_SIZE / (VEC_SIZE * 4) - 1), %esi
>
> testq %rdi, %rdi
> +# ifdef USE_AS_STRNCMP
> + /* At this point, if %rdi value is 0, it already tested
> + VEC_SIZE*4+%r10 byte starting from %rax. This label
> + checks whether strncmp maximum offset reached or not. */
> + je L(string_nbyte_offset_check)
> +# else
> je L(back_to_loop)
> +# endif
> tzcntq %rdi, %rcx
> addq %r10, %rcx
> /* Adjust for number of bytes skipped. */
> @@ -627,6 +634,14 @@ L(loop_cross_page_2_vec):
> VZEROUPPER
> ret
>
> +# ifdef USE_AS_STRNCMP
> +L(string_nbyte_offset_check):
> + leaq (VEC_SIZE * 4)(%r10), %r10
> + cmpq %r10, %r11
> + jbe L(zero)
> + jmp L(back_to_loop)
> +# endif
> +
> .p2align 4
> L(cross_page_loop):
> /* Check one byte/dword at a time. */
> --
> 2.26.2
>
I am checking it in and backport it to the active branches.
@@ -591,7 +591,14 @@ L(loop_cross_page_2_vec):
movl $(PAGE_SIZE / (VEC_SIZE * 4) - 1), %esi
testq %rdi, %rdi
+# ifdef USE_AS_STRNCMP
+ /* At this point, if %rdi value is 0, it already tested
+ VEC_SIZE*4+%r10 byte starting from %rax. This label
+ checks whether strncmp maximum offset reached or not. */
+ je L(string_nbyte_offset_check)
+# else
je L(back_to_loop)
+# endif
tzcntq %rdi, %rcx
addq %r10, %rcx
/* Adjust for number of bytes skipped. */
@@ -627,6 +634,14 @@ L(loop_cross_page_2_vec):
VZEROUPPER
ret
+# ifdef USE_AS_STRNCMP
+L(string_nbyte_offset_check):
+ leaq (VEC_SIZE * 4)(%r10), %r10
+ cmpq %r10, %r11
+ jbe L(zero)
+ jmp L(back_to_loop)
+# endif
+
.p2align 4
L(cross_page_loop):
/* Check one byte/dword at a time. */