]> sourceware.org Git - glibc.git/commit
x86: Optimize memcmp-evex-movbe.S
authorNoah Goldstein <goldstein.w.n@gmail.com>
Mon, 17 May 2021 17:57:24 +0000 (13:57 -0400)
committerNoah Goldstein <goldstein.w.n@gmail.com>
Wed, 19 May 2021 02:57:51 +0000 (22:57 -0400)
commit4ad473e97acdc5f6d811755b67c09f2128a644ce
tree92dd8d4893487377a5e864d306d2ec160c503f3f
parent16d12015c57701b08d7bbed6ec536641bcafb428
x86: Optimize memcmp-evex-movbe.S

No bug. This commit optimizes memcmp-evex.S. The optimizations include
adding a new vec compare path for small sizes, reorganizing the entry
control flow, removing some unnecissary ALU instructions from the main
loop, and most importantly replacing the heavy use of vpcmp + kand
logic with vpxor + vptern. test-memcmp and test-wmemcmp are both
passing.

Signed-off-by: Noah Goldstein <goldstein.w.n@gmail.com>
Reviewed-by: H.J. Lu <hjl.tools@gmail.com>
sysdeps/x86_64/multiarch/memcmp-evex-movbe.S
This page took 0.041147 seconds and 5 git commands to generate.