]> git.ipfire.org Git - thirdparty/glibc.git/commit
aarch64: Optimized implementation of memcmp
authorXuelei Zhang <zhangxuelei4@huawei.com>
Thu, 19 Dec 2019 12:31:59 +0000 (12:31 +0000)
committerAdhemerval Zanella <adhemerval.zanella@linaro.org>
Thu, 19 Dec 2019 19:31:04 +0000 (16:31 -0300)
commit233efd433d847e69480fe587c4c29a32fe554174
treeef91338851e143c4dcc296a59ba2bacae3424211
parent442d9c9c677804287a54b10d5fa5e58a9cdd338d
aarch64: Optimized implementation of memcmp

The loop body is expanded from a 16-byte comparison to a 64-byte
comparison, and the usage of ldp is replaced by the Post-index
mode to the Base plus offset mode. Hence, compare can faster 18%
around > 128 bytes in all.

Checked on aarch64-linux-gnu.

Reviewed-by: Wilco Dijkstra <Wilco.Dijkstra@arm.com>
sysdeps/aarch64/memcmp.S