From: Willy Tarreau <w@1wt.eu>
Date: Sat, 30 Dec 2023 15:28:05 +0000 (+0100)
Subject: IMPORT: ebtree: make string_equal_bits turn back to unsigned char
X-Git-Tag: v3.0-dev1~3
X-Git-Url: http://git.ipfire.org/cgi-bin/gitweb.cgi?a=commitdiff_plain;h=2a14f99dbb243df5bb29866285b0ce922b4198a1;p=thirdparty%2Fhaproxy.git

IMPORT: ebtree: make string_equal_bits turn back to unsigned char

With flsnz() instead of flsnz_long() we're now getting a better
performance on both x86 and ARM. The difference is that previously
we were relying on a function that was forcing the use of register
%eax for the 8-bit version and that was preventing the compiler
from keeping the code optimized. The gain is roughly 5% on ARM and
1% on x86.

This is cbtree commit 19cf39b2514bea79fed94d85e421e293be097a0e.
This is ebtree commit a9aaf2d94e2c92fa37aa3152c2ad8220a9533ead.
---

diff --git a/include/import/ebtree.h b/include/import/ebtree.h
index 25eb85cfbe..c7a9a3897a 100644
--- a/include/import/ebtree.h
+++ b/include/import/ebtree.h
@@ -840,7 +840,7 @@ static forceinline ssize_t string_equal_bits(const unsigned char *a,
 					     const unsigned char *b,
 					     size_t ignore)
 {
-	unsigned long c, d;
+	unsigned char c, d;
 	size_t beg;
 
 	beg = ignore >> 3;
@@ -864,7 +864,7 @@ static forceinline ssize_t string_equal_bits(const unsigned char *a,
 	 * identical bits. Note that low bit numbers are assigned to high positions
 	 * in the byte, as we compare them as strings.
 	 */
-	return (beg << 3) - flsnz_long(c);
+	return (beg << 3) - flsnz(c);
 }
 
 static forceinline int cmp_bits(const unsigned char *a, const unsigned char *b, unsigned int pos)