[FFmpeg-devel] [PATCH 2/5] x86inc: Make REP_RET identical to RET in SSSE3+ functions

James Darnley jdarnley at obe.tv
Fri Jun 9 02:04:59 EEST 2017


From: Henrik Gramner <henrik at gramner.com>

There's no point in emitting a rep prefix before ret on modern CPUs.
---
 libavutil/x86/x86inc.asm | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/libavutil/x86/x86inc.asm b/libavutil/x86/x86inc.asm
index f2a6a3f1db..44069741cc 100644
--- a/libavutil/x86/x86inc.asm
+++ b/libavutil/x86/x86inc.asm
@@ -636,7 +636,7 @@ DECLARE_ARG 7, 8, 9, 10, 11, 12, 13, 14
 ; We can automatically detect "follows a branch", but not a branch target.
 ; (SSSE3 is a sufficient condition to know that your cpu doesn't have this problem.)
 %macro REP_RET 0
-    %if has_epilogue
+    %if has_epilogue || cpuflag(ssse3)
         RET
     %else
         rep ret
-- 
2.13.0



More information about the ffmpeg-devel mailing list