[FFmpeg-cvslog] r10165 - trunk/libavcodec/i386/snowdsp_mmx.c

michael subversion
Tue Aug 21 02:03:18 CEST 2007


Author: michael
Date: Tue Aug 21 02:03:18 2007
New Revision: 10165

Log:
Simplify and optimize the 4th vertical lifting step of the SSE2 code (untested)
This also reduces the needed headroom in that step by 1 bit


Modified:
   trunk/libavcodec/i386/snowdsp_mmx.c

Modified: trunk/libavcodec/i386/snowdsp_mmx.c
==============================================================================
--- trunk/libavcodec/i386/snowdsp_mmx.c	(original)
+++ trunk/libavcodec/i386/snowdsp_mmx.c	Tue Aug 21 02:03:18 2007
@@ -498,9 +498,8 @@ void ff_snow_vertical_compose97i_sse2(DW
         snow_vertical_compose_sse2_store(REG_c,"xmm0","xmm2","xmm4","xmm6")
         snow_vertical_compose_sse2_add(REG_S,"xmm0","xmm2","xmm4","xmm6")
         snow_vertical_compose_sse2_move("xmm0","xmm2","xmm4","xmm6","xmm1","xmm3","xmm5","xmm7")
-        snow_vertical_compose_sse2_r2r_add("xmm0","xmm2","xmm4","xmm6","xmm0","xmm2","xmm4","xmm6")
-        snow_vertical_compose_sse2_r2r_add("xmm1","xmm3","xmm5","xmm7","xmm0","xmm2","xmm4","xmm6")
         snow_vertical_compose_sse2_sra("1","xmm0","xmm2","xmm4","xmm6")
+        snow_vertical_compose_sse2_r2r_add("xmm1","xmm3","xmm5","xmm7","xmm0","xmm2","xmm4","xmm6")
         snow_vertical_compose_sse2_add(REG_a,"xmm0","xmm2","xmm4","xmm6")
         snow_vertical_compose_sse2_store(REG_a,"xmm0","xmm2","xmm4","xmm6")
 




More information about the ffmpeg-cvslog mailing list