crypto: x86/chacha20 - Add a 2-block AVX-512VL variant

Browse source

This version uses the same principle as the AVX2 version. It benefits
from the AVX-512VL rotate instructions and the more efficient partial
block handling using "vmovdqu8", resulting in a speedup of ~20%.

Unlike the AVX2 version, it is faster than the single block SSSE3 version
to process a single block. Hence we engage that function for (partial)
single block lengths as well.

Signed-off-by: Martin Willi <martin@strongswan.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

...

This commit is contained in:

Martin Willi 2018-11-20 17:30:49 +01:00 • committed by Herbert Xu

parent cee7a36ecb

commit 29a47b54e0

2 changed files with 178 additions and 0 deletions

Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL

Download patch file Download diff file Expand all files Collapse all files

171

arch/x86/crypto/chacha20-avx512vl-x86_64.S

View file

7

arch/x86/crypto/chacha20_glue.c

View file