sha/asm/sha512p8-ppc.pl: improve POWER9 performance by ~10%.
authorAndy Polyakov <appro@openssl.org>
Sat, 2 Jun 2018 13:25:50 +0000 (15:25 +0200)
committerAndy Polyakov <appro@openssl.org>
Sun, 3 Jun 2018 19:20:40 +0000 (21:20 +0200)
commitc4d9ef4cc5bf1c48a74b64879622ae9fd6f26b03
tree0efd8ba216093d0029a697208136fcbe76fda7d5
parent1a467bd12f20928f3d5e6809b5f9394dbe606541
sha/asm/sha512p8-ppc.pl: improve POWER9 performance by ~10%.

Biggest part, ~7%, of improvement resulted from omitting constants'
table index increment in each round. And minor part from rescheduling
instructions. Apparently POWER9 (and POWER8) manage to dispatch
instructions more efficiently if they are laid down as if they have
no latency...

Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6406)
crypto/sha/asm/sha512p8-ppc.pl