crypto: cast5-avx - tune assembler code for more performance

Browse source

Patch replaces 'movb' instructions with 'movzbl' to break false register
dependencies, interleaves instructions better for out-of-order scheduling
and merges constant 16-bit rotation with round-key variable rotation.

tcrypt ECB results (128bit key):

Intel Core i5-2450M:

size    old-vs-new      new-vs-generic  old-vs-generic
        enc     dec     enc     dec     enc     dec
256     1.18x   1.18x   2.45x   2.47x   2.08x   2.10x
1k      1.20x   1.20x   2.73x   2.73x   2.28x   2.28x
8k      1.20x   1.19x   2.73x   2.73x   2.28x   2.29x

[v2]
 - Do instruction interleaving another way to avoid adding new FPU<=>CPU
   register moves as these cause performance drop on Bulldozer.
 - Improvements to round-key variable rotation handling.
 - Further interleaving improvements for better out-of-order scheduling.

Cc: Johannes Goetzfried <Johannes.Goetzfried@informatik.stud.uni-erlangen.de>
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

...

This commit is contained in:

Jussi Kivilinna 2012-08-28 14:24:49 +03:00 • committed by Herbert Xu

parent f94a73f8dd

commit ddaea7869d

1 changed files with 160 additions and 106 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

266

arch/x86/crypto/cast5-avx-x86_64-asm_64.S

View file