From de50494505b53b32a0b127525b18e259ee0f9033 Mon Sep 17 00:00:00 2001
From: Andy Polyakov <appro@openssl.org>
Date: Tue, 20 Mar 2007 09:13:07 +0000
Subject: [PATCH] Two extra instructions in RC4 character loop give 80%
 performance improvement on Core2. I still need to detect Core2 and choose
 this path...

---
 crypto/rc4/asm/rc4-x86_64.pl | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/crypto/rc4/asm/rc4-x86_64.pl b/crypto/rc4/asm/rc4-x86_64.pl
index 4b990cba07..5236afec12 100755
--- a/crypto/rc4/asm/rc4-x86_64.pl
+++ b/crypto/rc4/asm/rc4-x86_64.pl
@@ -221,6 +221,8 @@ $code.=<<___;
 	movb	$TY#b,($dat,$XX[0])
 	add	$TX[0]#b,$TY#b
 	add	\$1,$XX[0]#b
+	movzb	$TY#b,$TY#d
+	movzb	$XX[0]#b,$XX[0]#d
 	movzb	($dat,$TY),$TY#d
 	movzb	($dat,$XX[0]),$TX[0]#d
 	xorb	($inp),$TY#b
-- 
GitLab