projects / linux-block.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 3cae5a3) | patch
crypto: x86/aes-gcm - tune better for AMD CPUs
authorEric Biggers <ebiggers@google.com>
Thu, 12 Dec 2024 21:28:39 +0000 (13:28 -0800)
committerHerbert Xu <herbert@gondor.apana.org.au>
Sat, 21 Dec 2024 14:46:24 +0000 (22:46 +0800)
commit95791ccd112eddd1f31a2c7040122365add52a3f
treeb3cb08075107ec7dc3019b0b5c6f17e82bf0035dtree
parent3cae5a3c05b87917f2f6d58248be8ea4d1a2a705commit | diff
crypto: x86/aes-gcm - tune better for AMD CPUs

Reorganize the main loop to free up the RNDKEYLAST[0-3] registers and
use them for more cached round keys.  This improves performance by about
2% on AMD Zen 4 and Zen 5.  Intel performance remains about the same.

Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
arch/x86/crypto/aes-gcm-avx10-x86_64.S diff | blob | blame | history
Linux 6.x block layer and io_uring tree(s)