[FFmpeg-cvslog] Replace SSE2 instruction in scalarproduct_float_sse() by SSE equivalent.

ami_stuff git at videolan.org
Wed Apr 4 09:31:53 CEST 2012


ffmpeg | branch: release/0.6 | ami_stuff <ami_stuff at o2.pl> | Thu Mar 22 19:28:52 2012 +0100| [b56606e6bc1156f0edf5fca53216481ebde48630] | committer: Carl Eugen Hoyos

Replace SSE2 instruction in scalarproduct_float_sse() by SSE equivalent.

Fixes an AAC decoding issue with the sample from ticket #213 on machines
with SSE but without SSE2.
Based on 89411a by Reimar.

(cherry picked from commit f6b78638086beae9bcab672d4c9de1790be5a928)

> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=b56606e6bc1156f0edf5fca53216481ebde48630
---

 libavcodec/x86/dsputil_yasm.asm |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/libavcodec/x86/dsputil_yasm.asm b/libavcodec/x86/dsputil_yasm.asm
index e2478a4..1a3bd1a 100644
--- a/libavcodec/x86/dsputil_yasm.asm
+++ b/libavcodec/x86/dsputil_yasm.asm
@@ -417,7 +417,7 @@ cglobal scalarproduct_float_sse, 3,3,2, v1, v2, offset
     shufps  xmm0, xmm0, 1
     addss   xmm0, xmm1
 %ifndef ARCH_X86_64
-    movd    r0m,  xmm0
+    movss   r0m,  xmm0
     fld     dword r0m
 %endif
     RET



More information about the ffmpeg-cvslog mailing list