]> git.sesse.net Git - ffmpeg/history - libavcodec/x86/vp8dsp.asm
Mark recently added pred4x4_down_left_mmxext as CONFIG_GPL. Although Holger
[ffmpeg] / libavcodec / x86 / vp8dsp.asm
2010-09-05 Reimar DöffingerUse "d" suffix for general-purpose registers used with...
2010-08-24 Ronald S. BultjeMark xmm registers as clobbered in simple loopfilter...
2010-08-23 Ronald S. BultjeFix segfaults in VP8 SIMD code on Win64 (and FATE/win64...
2010-08-02 Jason Garrett-GlaserVP8: move zeroing of luma DC block into the WHT
2010-07-31 Ronald S. BultjeUse word-writing instead of dword-writing (with two...
2010-07-26 Ronald S. BultjeUse pmaddubsw for the mbedge_filter (>=ssse3), 6-10...
2010-07-26 Jason Garrett-GlaserVP8: Much faster SSE2 MC
2010-07-26 Ronald S. BultjeEnable no-loop memory/register saving for ssse3/sse4...
2010-07-26 Ronald S. BultjeSave a register (or regsize of stackspace for x86-32...
2010-07-26 Ronald S. BultjeUse nested ifs instead of &&, which appears to not...
2010-07-26 Ronald S. BultjeSplit pextrw macro-spaghetti into several opt-specific...
2010-07-25 Ronald S. BultjeFix obvious bug in assignment. Somehow, the test vector...
2010-07-24 Ronald S. BultjeFix SPLATB_REG mess. Used to be a if/elseif/elseif...
2010-07-23 Jason Garrett-GlaserVP8: optimize DC-only chroma case in the same way as...
2010-07-23 Jason Garrett-GlaserVP8 asm: cosmetics (spacing)
2010-07-23 Jason Garrett-GlaserVP8: 30% faster idct_mb
2010-07-23 Jason Garrett-GlaserVP8: clear DCT blocks in iDCT instead of using clear_bl...
2010-07-22 Ronald S. BultjeUse pextrw for SSE4 mbedge filter result writing, speed...
2010-07-22 Ronald S. BultjeFix and enable horizontal >=SSE2 mbedge loopfilter.
2010-07-21 Jason Garrett-GlaserEliminate one instruction in VP8 dc_add_sse4
2010-07-21 Jason Garrett-GlaserVarious VP8 x86 deblocking speedups
2010-07-21 Jason Garrett-GlaserMake mmx VP8 WHT faster
2010-07-20 Ronald S. BultjeVP8 MBedge loopfilter MMX/MMX2/SSE2 functions for both...
2010-07-20 Ronald S. BultjeChroma (width=8) inner loopfilter MMX/MMX2/SSE2 for...
2010-07-19 Ronald S. BultjeRevert r24339 (it causes fate failures on x86-64) ...
2010-07-19 Ronald S. BultjeImplement chroma (width=8) inner loopfilter MMX/MMX2...
2010-07-19 Ronald S. BultjeBe more efficient with registers or stack memory. Saves...
2010-07-19 Ronald S. BultjeChange function prototypes for width=8 inner and mbedge...
2010-07-16 Ronald S. BultjeAttempt to fix x86-64 testsuite on fate.
2010-07-16 Ronald S. BultjeRemove duplicate define.
2010-07-16 Ronald S. BultjeRevert 24270, it contained some stuff that shouldn...
2010-07-16 Ronald S. BultjeRemove duplicate define.
2010-07-16 Ronald S. BultjeGive x86 r%d registers names, this will simplify implem...
2010-07-16 Ronald S. BultjeChange return statement, the REP_RET is a mistake since...
2010-07-15 Ronald S. BultjeVP8 H/V inner loopfilter MMX/MMXEXT/SSE2 optimizations.
2010-07-03 Ronald S. BultjeSimple H/V loopfilter for VP8 in MMX, MMX2 and SSE2...
2010-07-03 Jason Garrett-GlaserSSSE3 versions of vp8 width4 bilinear MC functions
2010-07-02 Jason Garrett-GlaserSSSE3 versions of width4 VP8 6-tap MC functions
2010-06-29 Jason Garrett-GlaserUse add instead of lshift in mmxext vp8 idct
2010-06-29 Ronald S. BultjeRemove unused macros (duplicates from the now-LGPL...
2010-06-29 Ronald S. BultjeMMX idct_add for VP8.
2010-06-29 Jason Garrett-GlaserAdd mmxext version of VP8 DC Hadamard transform
2010-06-28 Jason Garrett-GlaserFix VP8 bilinear mc on x86_64
2010-06-28 Jason Garrett-GlaserAdd x86 asm functions for VP8 put_pixels
2010-06-28 Jason Garrett-GlaserAdd MMX, SSE2, SSSE3 asm for VP8 bilinear MC
2010-06-27 Jason Garrett-GlaserFirst shot at VP8 optimizations: