This is the mail archive of the
glibc-cvs@sourceware.org
mailing list for the glibc project.
[glibc/release/2.27/master] Speedup first memmem match
- From: Wilco Dijkstra <wilco at sourceware dot org>
- To: glibc-cvs at sourceware dot org
- Date: 13 Sep 2019 15:29:01 -0000
- Subject: [glibc/release/2.27/master] Speedup first memmem match
https://sourceware.org/git/gitweb.cgi?p=glibc.git;h=ecd6271ed80e71b3b1f286e2977f534d54c33af4
commit ecd6271ed80e71b3b1f286e2977f534d54c33af4
Author: Rajalakshmi Srinivasaraghavan <raji@linux.vnet.ibm.com>
Date: Tue Aug 28 12:42:19 2018 +0530
Speedup first memmem match
As done in commit 284f42bc778e487dfd5dff5c01959f93b9e0c4f5, memcmp
can be used after memchr to avoid the initialization overhead of the
two-way algorithm for the first match. This has shown improvement
>40% for first match.
(cherry picked from commit c8dd67e7c958de04c3783cbea7c384431707b5f8)
Diff:
---
ChangeLog | 4 ++++
string/memmem.c | 4 ++++
2 files changed, 8 insertions(+)
diff --git a/ChangeLog b/ChangeLog
index 00e666b..d4bd101 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,7 @@
+2019-09-13 Rajalakshmi Srinivasaraghavan <raji@linux.vnet.ibm.com>
+
+ * string/memmem.c: Use memcmp for first match.
+
2019-09-13 Wilco Dijkstra <wdijkstr@arm.com>
* string/strcasestr.c (STRCASESTR): Simplify and speedup first match.
diff --git a/string/memmem.c b/string/memmem.c
index 43efaa3..d72b824 100644
--- a/string/memmem.c
+++ b/string/memmem.c
@@ -70,6 +70,10 @@ __memmem (const void *haystack_start, size_t haystack_len,
haystack_len -= haystack - (const unsigned char *) haystack_start;
if (haystack_len < needle_len)
return NULL;
+ /* Check whether we have a match. This improves performance since we
+ avoid the initialization overhead of the two-way algorithm. */
+ if (memcmp (haystack, needle, needle_len) == 0)
+ return (void *) haystack;
return two_way_short_needle (haystack, haystack_len, needle, needle_len);
}
else