diff options
author | Allan Jude <allanjude@FreeBSD.org> | 2017-07-01 21:18:06 +0000 |
---|---|---|
committer | Allan Jude <allanjude@FreeBSD.org> | 2017-07-01 21:18:06 +0000 |
commit | 929b476ae6c8ebcda6021593f52d87eadbf706f9 (patch) | |
tree | ac180f5bd3a013811941b2017d6595c7829966f4 /sys/modules | |
parent | e11bad9d2b3254197d6ebe56c51a54910a5274cc (diff) | |
download | src-929b476ae6c8ebcda6021593f52d87eadbf706f9.tar.gz src-929b476ae6c8ebcda6021593f52d87eadbf706f9.zip |
Increase loop unrolling for skein hashes
This patch was inspired by an opposite change made to shrink the code
for the boot loader.
On my i7-4770, it increases the skein1024 speed from 470 to 550 MB/s
Reviewed by: sbruno
MFC after: 1 month
Sponsored by: ScaleEngine Inc.
Differential Revision: https://reviews.freebsd.org/D7824
Notes
Notes:
svn path=/head/; revision=320554
Diffstat (limited to 'sys/modules')
-rw-r--r-- | sys/modules/crypto/Makefile | 4 |
1 files changed, 4 insertions, 0 deletions
diff --git a/sys/modules/crypto/Makefile b/sys/modules/crypto/Makefile index a06d7dacfbe1..930c72792dbb 100644 --- a/sys/modules/crypto/Makefile +++ b/sys/modules/crypto/Makefile @@ -19,11 +19,15 @@ SRCS += camellia.c camellia-api.c SRCS += des_ecb.c des_enc.c des_setkey.c SRCS += sha1.c sha256c.c sha512c.c SRCS += skein.c skein_block.c +# unroll the 256 and 512 loops, half unroll the 1024 +CFLAGS+= -DSKEIN_LOOP=995 .if exists(${MACHINE_ARCH}/skein_block_asm.s) .PATH: ${SRCTOP}/sys/crypto/skein/${MACHINE_ARCH} SRCS += skein_block_asm.s CFLAGS += -DSKEIN_ASM -DSKEIN_USE_ASM=1792 # list of block functions to replace with assembly: 256+512+1024 = 1792 ACFLAGS += -DELF -Wa,--noexecstack +# Fully unroll all loops in the assembly optimized version +AFLAGS+= --defsym SKEIN_LOOP=0 .endif SRCS += siphash.c SRCS += gmac.c gfmult.c |