linux/crypto
Martin Willi c70f4abef0 crypto: poly1305 - Add a SSE2 SIMD variant for x86_64
Implements an x86_64 assembler driver for the Poly1305 authenticator. This
single block variant holds the 130-bit integer in 5 32-bit words, but uses
SSE to do two multiplications/additions in parallel.

When calling updates with small blocks, the overhead for kernel_fpu_begin/
kernel_fpu_end() negates the perfmance gain. We therefore use the
poly1305-generic fallback for small updates.

For large messages, throughput increases by ~5-10% compared to
poly1305-generic:

testing speed of poly1305 (poly1305-generic)
test  0 (   96 byte blocks,   16 bytes per update,   6 updates): 4080026 opers/sec,  391682496 bytes/sec
test  1 (   96 byte blocks,   32 bytes per update,   3 updates): 6221094 opers/sec,  597225024 bytes/sec
test  2 (   96 byte blocks,   96 bytes per update,   1 updates): 9609750 opers/sec,  922536057 bytes/sec
test  3 (  288 byte blocks,   16 bytes per update,  18 updates): 1459379 opers/sec,  420301267 bytes/sec
test  4 (  288 byte blocks,   32 bytes per update,   9 updates): 2115179 opers/sec,  609171609 bytes/sec
test  5 (  288 byte blocks,  288 bytes per update,   1 updates): 3729874 opers/sec, 1074203856 bytes/sec
test  6 ( 1056 byte blocks,   32 bytes per update,  33 updates):  593000 opers/sec,  626208000 bytes/sec
test  7 ( 1056 byte blocks, 1056 bytes per update,   1 updates): 1081536 opers/sec, 1142102332 bytes/sec
test  8 ( 2080 byte blocks,   32 bytes per update,  65 updates):  302077 opers/sec,  628320576 bytes/sec
test  9 ( 2080 byte blocks, 2080 bytes per update,   1 updates):  554384 opers/sec, 1153120176 bytes/sec
test 10 ( 4128 byte blocks, 4128 bytes per update,   1 updates):  278715 opers/sec, 1150536345 bytes/sec
test 11 ( 8224 byte blocks, 8224 bytes per update,   1 updates):  140202 opers/sec, 1153022070 bytes/sec

testing speed of poly1305 (poly1305-simd)
test  0 (   96 byte blocks,   16 bytes per update,   6 updates): 3790063 opers/sec,  363846076 bytes/sec
test  1 (   96 byte blocks,   32 bytes per update,   3 updates): 5913378 opers/sec,  567684355 bytes/sec
test  2 (   96 byte blocks,   96 bytes per update,   1 updates): 9352574 opers/sec,  897847104 bytes/sec
test  3 (  288 byte blocks,   16 bytes per update,  18 updates): 1362145 opers/sec,  392297990 bytes/sec
test  4 (  288 byte blocks,   32 bytes per update,   9 updates): 2007075 opers/sec,  578037628 bytes/sec
test  5 (  288 byte blocks,  288 bytes per update,   1 updates): 3709811 opers/sec, 1068425798 bytes/sec
test  6 ( 1056 byte blocks,   32 bytes per update,  33 updates):  566272 opers/sec,  597984182 bytes/sec
test  7 ( 1056 byte blocks, 1056 bytes per update,   1 updates): 1111657 opers/sec, 1173910108 bytes/sec
test  8 ( 2080 byte blocks,   32 bytes per update,  65 updates):  288857 opers/sec,  600823808 bytes/sec
test  9 ( 2080 byte blocks, 2080 bytes per update,   1 updates):  590746 opers/sec, 1228751888 bytes/sec
test 10 ( 4128 byte blocks, 4128 bytes per update,   1 updates):  301825 opers/sec, 1245936902 bytes/sec
test 11 ( 8224 byte blocks, 8224 bytes per update,   1 updates):  153075 opers/sec, 1258896201 bytes/sec

Benchmark results from a Core i5-4670T.

Signed-off-by: Martin Willi <martin@strongswan.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2015-07-17 21:20:27 +08:00
..
asymmetric_keys Fix up implicit <module.h> users that will break later. 2015-07-02 10:25:22 -07:00
async_tx md/raid5: activate raid6 rmw feature 2015-04-22 08:00:42 +10:00
.gitignore crypto: rsa - add .gitignore for crypto/*.-asn1.[ch] files 2015-06-25 23:29:24 +08:00
842.c crypto: 842 - change 842 alg to use software 2015-05-11 15:06:43 +08:00
ablk_helper.c crypto: cryptd - process CRYPTO_ALG_INTERNAL 2015-03-31 21:21:04 +08:00
ablkcipher.c crypto: skcipher - Allow givencrypt to be NULL 2015-06-22 15:49:16 +08:00
aead.c crypto: aead - Propagate new AEAD implementation flag for IV generators 2015-07-14 14:56:46 +08:00
aes_generic.c crypto: add missing crypto module aliases 2015-01-13 22:29:11 +11:00
af_alg.c Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next 2015-06-24 16:49:49 -07:00
ahash.c crypto: replace scatterwalk_sg_next with sg_next 2015-01-26 11:34:22 +11:00
akcipher.c crypto: akcipher - fix spelling cihper -> cipher 2015-06-25 23:18:32 +08:00
algapi.c crypto: api - Add instance free function to crypto_type 2015-07-14 14:56:45 +08:00
algboss.c crypto: algboss - Hold ref count on larval 2013-06-25 19:15:17 +08:00
algif_aead.c crypto: algif_aead - Temporarily disable all AEAD algorithms 2015-06-22 15:49:30 +08:00
algif_hash.c new helper: msg_data_left() 2015-04-11 15:53:35 -04:00
algif_rng.c crypto: algif_rng - Remove obsolete const-removal cast 2015-04-22 09:30:21 +08:00
algif_skcipher.c new helper: msg_data_left() 2015-04-11 15:53:35 -04:00
ansi_cprng.c crypto: ansi_cprng - Convert to new rng interface 2015-04-22 09:30:18 +08:00
anubis.c crypto: prefix module autoloading with "crypto-" 2014-11-24 22:43:57 +08:00
api.c crypto: api - prevent helper ciphers from being used 2015-03-31 21:21:03 +08:00
arc4.c crypto: prefix module autoloading with "crypto-" 2014-11-24 22:43:57 +08:00
authenc.c crypto: authenc - Use crypto_aead_set_reqsize helper 2015-05-13 10:31:37 +08:00
authencesn.c crypto: authencesn - Use crypto_aead_set_reqsize helper 2015-05-13 10:31:38 +08:00
blkcipher.c crypto: blkcipher - Include crypto/aead.h 2015-05-13 10:31:34 +08:00
blowfish_common.c crypto: blowfish - split generic and common c code 2011-09-22 21:25:25 +10:00
blowfish_generic.c crypto: add missing crypto module aliases 2015-01-13 22:29:11 +11:00
camellia_generic.c crypto: add missing crypto module aliases 2015-01-13 22:29:11 +11:00
cast5_generic.c crypto: add missing crypto module aliases 2015-01-13 22:29:11 +11:00
cast6_generic.c crypto: add missing crypto module aliases 2015-01-13 22:29:11 +11:00
cast_common.c crypto: make tables used from assembler __visible 2013-08-14 20:42:03 +10:00
cbc.c crypto: include crypto- module prefix in template 2014-11-26 20:06:30 +08:00
ccm.c crypto: ccm - Convert to new AEAD interface 2015-07-17 21:20:09 +08:00
chacha20_generic.c crypto: chacha20 - Export common ChaCha20 helpers 2015-07-17 21:20:21 +08:00
chacha20poly1305.c crypto: poly1305 - Export common Poly1305 helpers 2015-07-17 21:20:26 +08:00
chainiv.c crypto: chainiv - Offer normal cipher functionality without RNG 2015-06-22 15:49:28 +08:00
cipher.c crypto: cipher - Fix checkpatch errors 2010-02-16 20:31:37 +08:00
cmac.c crypto: include crypto- module prefix in template 2014-11-26 20:06:30 +08:00
compress.c
crc32.c crypto: prefix module autoloading with "crypto-" 2014-11-24 22:43:57 +08:00
crc32c_generic.c crypto: add missing crypto module aliases 2015-01-13 22:29:11 +11:00
crct10dif_common.c crypto: crct10dif - Add fallback for broken initrds 2013-09-12 15:31:34 +10:00
crct10dif_generic.c crypto: add missing crypto module aliases 2015-01-13 22:29:11 +11:00
cryptd.c crypto: cryptd - Propagate new AEAD implementation flag 2015-07-14 14:56:45 +08:00
crypto_null.c crypto: null - Add default null skcipher 2015-05-22 11:25:55 +08:00
crypto_user.c crypto: user - Add CRYPTO_MSG_DELRNG 2015-06-22 15:49:27 +08:00
crypto_wq.c crypto: crypto_wq - Fix late crypto work queue initialization 2014-03-21 21:54:28 +08:00
ctr.c crypto: include crypto- module prefix in template 2014-11-26 20:06:30 +08:00
cts.c crypto: cts - Weed out non-CBC algorithms 2015-01-20 14:44:15 +11:00
deflate.c crypto: prefix module autoloading with "crypto-" 2014-11-24 22:43:57 +08:00
des_generic.c crypto: add missing crypto module aliases 2015-01-13 22:29:11 +11:00
drbg.c crypto: drbg - report backend_cra_name when allocation fails 2015-06-11 21:55:28 +08:00
ecb.c crypto: include crypto- module prefix in template 2014-11-26 20:06:30 +08:00
echainiv.c crypto: echainiv - Fix encryption convention 2015-07-14 14:56:46 +08:00
eseqiv.c crypto: eseqiv - Offer normal cipher functionality without RNG 2015-06-22 15:49:28 +08:00
fcrypt.c crypto: prefix module autoloading with "crypto-" 2014-11-24 22:43:57 +08:00
fips.c crypto: fips - Move fips_enabled sysctl into fips.c 2015-04-23 14:18:09 +08:00
gcm.c crypto: gcm - Use new IV convention 2015-07-14 14:56:47 +08:00
gf128mul.c crypto: gf128mul - fix call to memset() 2011-07-08 17:21:21 +08:00
ghash-generic.c crypto: add missing crypto module aliases 2015-01-13 22:29:11 +11:00
hash_info.c crypto: provide single place for hash algo information 2013-10-25 17:14:03 -04:00
hmac.c crypto: include crypto- module prefix in template 2014-11-26 20:06:30 +08:00
internal.h crypto: api - Remove linux/fips.h from internal.h 2015-04-23 14:18:10 +08:00
jitterentropy-kcapi.c crypto: jitterentropy - avoid compiler warnings 2015-06-25 23:18:32 +08:00
jitterentropy.c crypto: jitterentropy - Delete unnecessary checks before the function call "kzfree" 2015-06-25 23:18:33 +08:00
Kconfig crypto: poly1305 - Add a SSE2 SIMD variant for x86_64 2015-07-17 21:20:27 +08:00
khazad.c crypto: prefix module autoloading with "crypto-" 2014-11-24 22:43:57 +08:00
lrw.c crypto: include crypto- module prefix in template 2014-11-26 20:06:30 +08:00
lz4.c crypto: prefix module autoloading with "crypto-" 2014-11-24 22:43:57 +08:00
lz4hc.c crypto: prefix module autoloading with "crypto-" 2014-11-24 22:43:57 +08:00
lzo.c crypto: prefix module autoloading with "crypto-" 2014-11-24 22:43:57 +08:00
Makefile crypto: jitterentropy - avoid compiler warnings 2015-06-25 23:18:32 +08:00
mcryptd.c crypto: mcryptd - process CRYPTO_ALG_INTERNAL 2015-03-31 21:21:13 +08:00
md4.c crypto: prefix module autoloading with "crypto-" 2014-11-24 22:43:57 +08:00
md5.c crypto: md5 - use md5 IV MD5_HX instead of their raw value 2015-05-18 12:20:18 +08:00
memneq.c crypto: memneq - fix for archs without efficient unaligned access 2013-12-09 20:09:12 +08:00
michael_mic.c crypto: prefix module autoloading with "crypto-" 2014-11-24 22:43:57 +08:00
pcbc.c crypto: include crypto- module prefix in template 2014-11-26 20:06:30 +08:00
pcompress.c crypto: pcomp - Use crypto_alg_extsize helper 2015-04-21 10:19:55 +08:00
pcrypt.c crypto: pcrypt - Propagate new AEAD implementation flag 2015-07-14 14:56:45 +08:00
poly1305_generic.c crypto: poly1305 - Export common Poly1305 helpers 2015-07-17 21:20:26 +08:00
proc.c crypto: fips - Move fips_enabled sysctl into fips.c 2015-04-23 14:18:09 +08:00
ripemd.h
rmd128.c crypto: prefix module autoloading with "crypto-" 2014-11-24 22:43:57 +08:00
rmd160.c crypto: prefix module autoloading with "crypto-" 2014-11-24 22:43:57 +08:00
rmd256.c crypto: prefix module autoloading with "crypto-" 2014-11-24 22:43:57 +08:00
rmd320.c crypto: prefix module autoloading with "crypto-" 2014-11-24 22:43:57 +08:00
rng.c crypto: rng - Do not free default RNG when it becomes unused 2015-06-22 15:49:18 +08:00
rsa_helper.c crypto: rsa - add a new rsa generic implementation 2015-06-17 17:03:53 +08:00
rsa.c crypto: rsa - limit supported key lengths 2015-07-17 21:20:19 +08:00
rsakey.asn1 crypto: rsa - add a new rsa generic implementation 2015-06-17 17:03:53 +08:00
salsa20_generic.c crypto: add missing crypto module aliases 2015-01-13 22:29:11 +11:00
scatterwalk.c crypto: scatterwalk - Hide PageSlab call to optimise away flush_dcache_page 2015-06-03 10:51:25 +08:00
seed.c crypto: prefix module autoloading with "crypto-" 2014-11-24 22:43:57 +08:00
seqiv.c crypto: seqiv - Replace seqniv with seqiv 2015-07-14 14:56:46 +08:00
serpent_generic.c crypto: add missing crypto module aliases 2015-01-13 22:29:11 +11:00
sha1_generic.c crypto: sha1-generic - move to generic glue implementation 2015-04-10 21:39:40 +08:00
sha256_generic.c crypto: sha256-generic - move to generic glue implementation 2015-04-10 21:39:41 +08:00
sha512_generic.c crypto: sha512-generic - move to generic glue implementation 2015-04-10 21:39:41 +08:00
shash.c crypto: shash - Use crypto_alg_extsize helper 2015-04-21 10:19:54 +08:00
tcrypt.c crypto: tcrypt - Add ChaCha20/Poly1305 speed tests 2015-07-17 21:20:20 +08:00
tcrypt.h crypto: tcrypt - Add ChaCha20/Poly1305 speed tests 2015-07-17 21:20:20 +08:00
tea.c crypto: add missing crypto module aliases 2015-01-13 22:29:11 +11:00
testmgr.c crypto: testmgr - Reenable rfc4309 test 2015-07-17 21:20:14 +08:00
testmgr.h crypto: testmgr - Add a longer ChaCha20 test vector 2015-07-17 21:20:26 +08:00
tgr192.c crypto: add missing crypto module aliases 2015-01-13 22:29:11 +11:00
twofish_common.c crypto: twofish-x86_64-3way - add lrw support 2011-11-09 11:53:32 +08:00
twofish_generic.c crypto: add missing crypto module aliases 2015-01-13 22:29:11 +11:00
vmac.c crypto: include crypto- module prefix in template 2014-11-26 20:06:30 +08:00
wp512.c crypto: add missing crypto module aliases 2015-01-13 22:29:11 +11:00
xcbc.c crypto: include crypto- module prefix in template 2014-11-26 20:06:30 +08:00
xor.c add further __init annotations to crypto/xor.c 2012-10-11 13:42:32 +11:00
xts.c crypto: include crypto- module prefix in template 2014-11-26 20:06:30 +08:00
zlib.c crypto: pcomp - Constify (de)compression parameters 2015-05-01 11:16:37 +08:00