From: Linus Torvalds <torvalds@linux-foundation.org>
To: Nicolas Pitre <nico@cam.org>
Cc: George Spelvin <linux@horizon.com>,
Junio C Hamano <gitster@pobox.com>,
git@vger.kernel.org
Subject: Re: x86 SHA1: Faster than OpenSSL
Date: Wed, 5 Aug 2009 19:08:03 -0700 (PDT) [thread overview]
Message-ID: <alpine.LFD.2.01.0908051902580.3390@localhost.localdomain> (raw)
In-Reply-To: <alpine.LFD.2.00.0908052144430.16073@xanadu.home>
On Wed, 5 Aug 2009, Nicolas Pitre wrote:
>
> Even better is to not shift len at all in SHA_update() but shift
> ctx->size only at the end in SHA_final(). It is not like if
> SHA_update() could operate on partial bytes, so counting total bytes
> instead of total bits is all you need. This way you need no cast there
> and make the code slightly faster.
Yeah, I tried it, but it's not noticeable.
The bigger issue seems to be that it's shifter-limited, or that's what I
take away from my profiles. I suspect it's even _more_ shifter-limited on
some other micro-architectures, because gcc is being stupid, and generates
ror $31,%eax
from the "left shift + right shift" combination. It seems to -always-
generate a "ror", rather than trying to generate 'rot' if the shift count
would be smaller that way.
And I know _some_ old micro-architectures will literally internally loop
on the rol/ror counts, so "ror $31" can be _much_ more expensive than "rol
$1".
That isn't the case on my Nehalem, though. But I can't seem to get gcc to
generate better code without actually using inline asm..
(So to clarify: this patch makes no difference that I can see to
performance, but I suspect it could matter on other CPU's like an old
Pentium or maybe an Atom).
Linus
---
block-sha1/sha1.c | 36 ++++++++++++++++++++++++------------
block-sha1/sha1.h | 2 +-
2 files changed, 25 insertions(+), 13 deletions(-)
diff --git a/block-sha1/sha1.c b/block-sha1/sha1.c
index 8fd90b0..a45a3de 100644
--- a/block-sha1/sha1.c
+++ b/block-sha1/sha1.c
@@ -80,7 +80,19 @@ void blk_SHA1_Final(unsigned char hashout[20], blk_SHA_CTX *ctx)
((unsigned int *)hashout)[i] = htonl(ctx->H[i]);
}
-#define SHA_ROT(X,n) (((X) << (n)) | ((X) >> (32-(n))))
+#if defined(__i386__) || defined(__x86_64__)
+
+#define SHA_ASM(op, x, n) ({ unsigned int __res; asm(op " %1,%0":"=r" (__res):"i" (n), "0" (x)); __res; })
+#define SHA_ROL(x,n) SHA_ASM("rol", x, n)
+#define SHA_ROR(x,n) SHA_ASM("ror", x, n)
+
+#else
+
+#define SHA_ROT(X,n) (((X) << (l)) | ((X) >> (r)))
+#define SHA_ROL(X,n) SHA_ROT(X,n,32-(n))
+#define SHA_ROR(X,n) SHA_ROT(X,32-(n),n)
+
+#endif
static void blk_SHA1Block(blk_SHA_CTX *ctx, const unsigned int *data)
{
@@ -93,7 +105,7 @@ static void blk_SHA1Block(blk_SHA_CTX *ctx, const unsigned int *data)
/* Unroll it? */
for (t = 16; t <= 79; t++)
- W[t] = SHA_ROT(W[t-3] ^ W[t-8] ^ W[t-14] ^ W[t-16], 1);
+ W[t] = SHA_ROL(W[t-3] ^ W[t-8] ^ W[t-14] ^ W[t-16], 1);
A = ctx->H[0];
B = ctx->H[1];
@@ -102,8 +114,8 @@ static void blk_SHA1Block(blk_SHA_CTX *ctx, const unsigned int *data)
E = ctx->H[4];
#define T_0_19(t) \
- TEMP = SHA_ROT(A,5) + (((C^D)&B)^D) + E + W[t] + 0x5a827999; \
- E = D; D = C; C = SHA_ROT(B, 30); B = A; A = TEMP;
+ TEMP = SHA_ROL(A,5) + (((C^D)&B)^D) + E + W[t] + 0x5a827999; \
+ E = D; D = C; C = SHA_ROR(B, 2); B = A; A = TEMP;
T_0_19( 0); T_0_19( 1); T_0_19( 2); T_0_19( 3); T_0_19( 4);
T_0_19( 5); T_0_19( 6); T_0_19( 7); T_0_19( 8); T_0_19( 9);
@@ -111,8 +123,8 @@ static void blk_SHA1Block(blk_SHA_CTX *ctx, const unsigned int *data)
T_0_19(15); T_0_19(16); T_0_19(17); T_0_19(18); T_0_19(19);
#define T_20_39(t) \
- TEMP = SHA_ROT(A,5) + (B^C^D) + E + W[t] + 0x6ed9eba1; \
- E = D; D = C; C = SHA_ROT(B, 30); B = A; A = TEMP;
+ TEMP = SHA_ROL(A,5) + (B^C^D) + E + W[t] + 0x6ed9eba1; \
+ E = D; D = C; C = SHA_ROR(B, 2); B = A; A = TEMP;
T_20_39(20); T_20_39(21); T_20_39(22); T_20_39(23); T_20_39(24);
T_20_39(25); T_20_39(26); T_20_39(27); T_20_39(28); T_20_39(29);
@@ -120,8 +132,8 @@ static void blk_SHA1Block(blk_SHA_CTX *ctx, const unsigned int *data)
T_20_39(35); T_20_39(36); T_20_39(37); T_20_39(38); T_20_39(39);
#define T_40_59(t) \
- TEMP = SHA_ROT(A,5) + ((B&C)|(D&(B|C))) + E + W[t] + 0x8f1bbcdc; \
- E = D; D = C; C = SHA_ROT(B, 30); B = A; A = TEMP;
+ TEMP = SHA_ROL(A,5) + ((B&C)|(D&(B|C))) + E + W[t] + 0x8f1bbcdc; \
+ E = D; D = C; C = SHA_ROR(B, 2); B = A; A = TEMP;
T_40_59(40); T_40_59(41); T_40_59(42); T_40_59(43); T_40_59(44);
T_40_59(45); T_40_59(46); T_40_59(47); T_40_59(48); T_40_59(49);
@@ -129,8 +141,8 @@ static void blk_SHA1Block(blk_SHA_CTX *ctx, const unsigned int *data)
T_40_59(55); T_40_59(56); T_40_59(57); T_40_59(58); T_40_59(59);
#define T_60_79(t) \
- TEMP = SHA_ROT(A,5) + (B^C^D) + E + W[t] + 0xca62c1d6; \
- E = D; D = C; C = SHA_ROT(B, 30); B = A; A = TEMP;
+ TEMP = SHA_ROL(A,5) + (B^C^D) + E + W[t] + 0xca62c1d6; \
+ E = D; D = C; C = SHA_ROR(B, 2); B = A; A = TEMP;
T_60_79(60); T_60_79(61); T_60_79(62); T_60_79(63); T_60_79(64);
T_60_79(65); T_60_79(66); T_60_79(67); T_60_79(68); T_60_79(69);
next prev parent reply other threads:[~2009-08-06 2:09 UTC|newest]
Thread overview: 60+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-07-26 23:21 Performance issue of 'git branch' George Spelvin
2009-07-31 10:46 ` Request for benchmarking: x86 SHA1 code George Spelvin
2009-07-31 11:11 ` Erik Faye-Lund
2009-07-31 11:31 ` George Spelvin
2009-07-31 11:37 ` Michael J Gruber
2009-07-31 12:24 ` Erik Faye-Lund
2009-07-31 12:29 ` Johannes Schindelin
2009-07-31 12:32 ` George Spelvin
2009-07-31 12:45 ` Erik Faye-Lund
2009-07-31 13:02 ` George Spelvin
2009-07-31 11:21 ` Michael J Gruber
2009-07-31 11:26 ` Michael J Gruber
2009-07-31 12:31 ` Carlos R. Mafra
2009-07-31 13:27 ` Brian Ristuccia
2009-07-31 14:05 ` George Spelvin
2009-07-31 13:27 ` Jakub Narebski
2009-07-31 15:05 ` Peter Harris
2009-07-31 15:22 ` Peter Harris
2009-08-03 3:47 ` x86 SHA1: Faster than OpenSSL George Spelvin
2009-08-03 7:36 ` Jonathan del Strother
2009-08-04 1:40 ` Mark Lodato
2009-08-04 2:30 ` Linus Torvalds
2009-08-04 2:51 ` Linus Torvalds
2009-08-04 3:07 ` Jon Smirl
2009-08-04 5:01 ` George Spelvin
2009-08-04 12:56 ` Jon Smirl
2009-08-04 14:29 ` Dmitry Potapov
2009-08-18 21:50 ` Andy Polyakov
2009-08-04 4:48 ` George Spelvin
2009-08-04 6:30 ` Linus Torvalds
2009-08-04 8:01 ` George Spelvin
2009-08-04 20:41 ` Junio C Hamano
2009-08-05 18:17 ` George Spelvin
2009-08-05 20:36 ` Johannes Schindelin
2009-08-05 20:44 ` Junio C Hamano
2009-08-05 20:55 ` Linus Torvalds
2009-08-05 23:13 ` Linus Torvalds
2009-08-06 1:18 ` Linus Torvalds
2009-08-06 1:52 ` Nicolas Pitre
2009-08-06 2:04 ` Junio C Hamano
2009-08-06 2:10 ` Linus Torvalds
2009-08-06 2:20 ` Nicolas Pitre
2009-08-06 2:08 ` Linus Torvalds [this message]
2009-08-06 3:19 ` Artur Skawina
2009-08-06 3:31 ` Linus Torvalds
2009-08-06 3:48 ` Linus Torvalds
2009-08-06 4:01 ` Linus Torvalds
2009-08-06 4:28 ` Artur Skawina
2009-08-06 4:50 ` Linus Torvalds
2009-08-06 5:19 ` Artur Skawina
2009-08-06 7:03 ` George Spelvin
2009-08-06 4:52 ` George Spelvin
2009-08-06 4:08 ` Artur Skawina
2009-08-06 4:27 ` Linus Torvalds
2009-08-06 5:44 ` Artur Skawina
2009-08-06 5:56 ` Artur Skawina
2009-08-06 7:45 ` Artur Skawina
2009-08-06 18:49 ` Erik Faye-Lund
2009-08-04 6:40 ` Linus Torvalds
2009-08-18 21:26 ` Andy Polyakov
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=alpine.LFD.2.01.0908051902580.3390@localhost.localdomain \
--to=torvalds@linux-foundation.org \
--cc=git@vger.kernel.org \
--cc=gitster@pobox.com \
--cc=linux@horizon.com \
--cc=nico@cam.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).