This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
[PATCH] Speedup first memmem match
- From: Rajalakshmi Srinivasaraghavan <raji at linux dot vnet dot ibm dot com>
- To: libc-alpha at sourceware dot org
- Cc: Rajalakshmi Srinivasaraghavan <raji at linux dot vnet dot ibm dot com>
- Date: Fri, 24 Aug 2018 16:58:46 +0530
- Subject: [PATCH] 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
more than 40% for first match.
Tested on powerpc64le and x86_64.
2018-08-24 Rajalakshmi Srinivasaraghavan <raji@linux.vnet.ibm.com>
* string/memmem.c: Use memcmp for first match.
---
string/memmem.c | 4 ++++
1 file changed, 4 insertions(+)
diff --git a/string/memmem.c b/string/memmem.c
index 43efaa3fb7..d72b8249e6 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
--
2.18.0