From: Bob Pearson <rpearson@systemfabricworks.com>
To: linux-kernel@vger.kernel.org
Cc: fzago@systemfabricworks.com, rpearson@systemfabricworks.com,
Joakim Tjernlund <joakim.tjernlund@transmode.se>,
George Spelvin <linux@horizon.com>,
akpm@linux-foundation.org
Subject: [PATCH v6 09/10] crc32-optimize-loops-for-x86.diff
Date: Wed, 31 Aug 2011 17:30:38 -0500 [thread overview]
Message-ID: <4E5EB60E.3030908@systemfabricworks.com> (raw)
In-Reply-To: <20110831213729.395283830@systemfabricworks.com>
Add two changes that improve the performance of x86 systems
1. replace main loop with incrementing counter
this change improves the performance of the selftest
by about 5-6% on Nehalem CPUs. The apparent
reason is that the compiler can use the loop index
to perform an indexed memory access. This is
reported to make the performance of PowerPC CPUs
to get worse.
2. replace the rem_len loop with incrementing counter
this change improves the performance of the selftest,
which has more than the usual number of occurances,
by about 1-2% on x86 CPUs. In actual work loads
the length is most often a multiple of 4 bytes and
this code does not get executed as often if at all.
Again this change is reported to make the performance
of PowerPC get worse.
Signed-off-by: Bob Pearson <rpearson@systemfabricworks.com>
---
lib/crc32.c | 13 +++++++++++++
1 file changed, 13 insertions(+)
Index: for-next/lib/crc32.c
===================================================================
--- for-next.orig/lib/crc32.c
+++ for-next/lib/crc32.c
@@ -66,6 +66,9 @@ crc32_body(u32 crc, unsigned char const
# endif
const u32 *b;
size_t rem_len;
+# ifdef CONFIG_X86
+ size_t i;
+# endif
const u32 *t0 = tab[0], *t1 = tab[1], *t2 = tab[2], *t3 = tab[3];
const u32 *t4 = tab[4], *t5 = tab[5], *t6 = tab[6], *t7 = tab[7];
u32 q;
@@ -86,7 +89,12 @@ crc32_body(u32 crc, unsigned char const
# endif
b = (const u32 *)buf;
+# ifdef CONFIG_X86
+ --b;
+ for (i = 0; i < len; i++) {
+# else
for (--b; len; --len) {
+# endif
q = crc ^ *++b; /* use pre increment for speed */
# if CRC_LE_BITS == 32
crc = DO_CRC4;
@@ -100,9 +108,14 @@ crc32_body(u32 crc, unsigned char const
/* And the last few bytes */
if (len) {
u8 *p = (u8 *)(b + 1) - 1;
+# ifdef CONFIG_X86
+ for (i = 0; i < len; i++)
+ DO_CRC(*++p); /* use pre increment for speed */
+# else
do {
DO_CRC(*++p); /* use pre increment for speed */
} while (--len);
+# endif
}
return crc;
#undef DO_CRC
next prev parent reply other threads:[~2011-08-31 22:30 UTC|newest]
Thread overview: 28+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <20110831213729.395283830@systemfabricworks.com>
2011-08-31 22:29 ` [PATCH v6 01/10] crc32-remove-trailing-whitespace.diff Bob Pearson
2011-08-31 22:29 ` [PATCH v6 02/10] crc32-move-to-documentation.diff Bob Pearson
2011-08-31 22:29 ` [PATCH v6 03/10] crc32-replace-self-test.diff Bob Pearson
2011-09-02 23:51 ` Andrew Morton
2011-09-06 16:14 ` Bob Pearson
2011-08-31 22:30 ` [PATCH v6 04/10] crc32-add-pointer-to-tab.diff Bob Pearson
2011-09-01 8:16 ` Joakim Tjernlund
2011-08-31 22:30 ` [PATCH v6 05/10] crc32-misc-cleanup.diff Bob Pearson
2011-09-02 23:50 ` Andrew Morton
2011-09-03 1:44 ` Stephen Rothwell
2011-09-06 13:40 ` Joakim Tjernlund
2011-09-06 14:50 ` Stephen Rothwell
2011-09-06 19:38 ` Andrew Morton
2011-09-06 20:18 ` Bob Pearson
2011-09-07 7:39 ` Joakim Tjernlund
2011-09-07 16:30 ` Bob Pearson
2011-09-07 17:51 ` Joakim Tjernlund
2011-09-06 16:05 ` Bob Pearson
2011-08-31 22:30 ` [PATCH v6 06/10] crc32-fix-check-endian-warnings.diff Bob Pearson
2011-08-31 22:30 ` [PATCH v6 07/10] crc32-add-real-8-bit.diff Bob Pearson
2011-08-31 22:30 ` [PATCH v6 08/10] crc32-add-slicing-by-8.diff Bob Pearson
2011-09-07 7:31 ` Joakim Tjernlund
2011-09-07 19:44 ` Bob Pearson
[not found] ` <OF3D37A60B.7A33B855-ONC1257904.00276B5B-C1257904.002951AF@LocalDomain>
2011-09-07 8:30 ` Joakim Tjernlund
2011-08-31 22:30 ` Bob Pearson [this message]
2011-08-31 22:30 ` [PATCH v6 10/10] crc32-final.diff Bob Pearson
2011-09-01 3:03 ` [PATCH v6 08/10] crc32-add-slicing-by-8.diff Bob Pearson
2011-09-07 7:32 ` Joakim Tjernlund
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=4E5EB60E.3030908@systemfabricworks.com \
--to=rpearson@systemfabricworks.com \
--cc=akpm@linux-foundation.org \
--cc=fzago@systemfabricworks.com \
--cc=joakim.tjernlund@transmode.se \
--cc=linux-kernel@vger.kernel.org \
--cc=linux@horizon.com \
/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