| Commit message (Expand) | Author | Age | Files | Lines |
* | VP8: move zeroing of luma DC block into the WHT | Jason Garrett-Glaser | 2010-08-02 | 2 | -2/+20 |
* | Use word-writing instead of dword-writing (with two cached but otherwise | Ronald S. Bultje | 2010-07-31 | 2 | -105/+98 |
* | Remove x86/mmx.h. It is not used anymore and has been deprecated for years. | Vitor Sessak | 2010-07-31 | 1 | -267/+0 |
* | Convert deinterlacing MMX code to YASM | Vitor Sessak | 2010-07-31 | 3 | -0/+95 |
* | Fix compilation in x86_64. I broke it with r24580. | Vitor Sessak | 2010-07-29 | 1 | -2/+2 |
* | Translate libmpeg2 MMX IDCT to plain asm | Vitor Sessak | 2010-07-29 | 1 | -208/+237 |
* | Use pmaddubsw for the mbedge_filter (>=ssse3), 6-10 cycles faster. | Ronald S. Bultje | 2010-07-26 | 1 | -2/+78 |
* | VP8: Much faster SSE2 MC | Jason Garrett-Glaser | 2010-07-26 | 1 | -88/+78 |
* | Enable no-loop memory/register saving for ssse3/sse4 also. | Ronald S. Bultje | 2010-07-26 | 1 | -2/+2 |
* | Save a register (or regsize of stackspace for x86-32) for the no-loop | Ronald S. Bultje | 2010-07-26 | 1 | -16/+24 |
* | Use nested ifs instead of &&, which appears to not work with %ifidn (i.e. this | Ronald S. Bultje | 2010-07-26 | 1 | -3/+9 |
* | Split pextrw macro-spaghetti into several opt-specific macros, this will make | Ronald S. Bultje | 2010-07-26 | 1 | -30/+49 |
* | Fix obvious bug in assignment. Somehow, the test vectors don't test this... | Ronald S. Bultje | 2010-07-25 | 1 | -1/+1 |
* | Fix SPLATB_REG mess. Used to be a if/elseif/elseif/elseif spaghetti, so this | Ronald S. Bultje | 2010-07-24 | 1 | -33/+52 |
* | Inline asm for VP56 arith coder | Eli Friedman | 2010-07-23 | 1 | -0/+54 |
* | VP8: optimize DC-only chroma case in the same way as luma. | Jason Garrett-Glaser | 2010-07-23 | 2 | -10/+53 |
* | VP8 asm: cosmetics (spacing) | Jason Garrett-Glaser | 2010-07-23 | 1 | -2/+2 |
* | VP8: 30% faster idct_mb | Jason Garrett-Glaser | 2010-07-23 | 2 | -54/+132 |
* | VP8: clear DCT blocks in iDCT instead of using clear_blocks. | Jason Garrett-Glaser | 2010-07-23 | 2 | -4/+24 |
* | Use pextrw for SSE4 mbedge filter result writing, speedup 5-10cycles on | Ronald S. Bultje | 2010-07-22 | 2 | -5/+34 |
* | Fix and enable horizontal >=SSE2 mbedge loopfilter. | Ronald S. Bultje | 2010-07-22 | 2 | -8/+8 |
* | relicense h264 deblock sse2 to lgpl | Loren Merritt | 2010-07-22 | 3 | -15/+19 |
* | sync yasm macros from x264 | Loren Merritt | 2010-07-21 | 1 | -12/+23 |
* | Eliminate one instruction in VP8 dc_add_sse4 | Jason Garrett-Glaser | 2010-07-21 | 1 | -2/+1 |
* | Various VP8 x86 deblocking speedups | Jason Garrett-Glaser | 2010-07-21 | 2 | -92/+107 |
* | Make mmx VP8 WHT faster | Jason Garrett-Glaser | 2010-07-21 | 2 | -19/+24 |
* | Add header declarations for mmx/sse constants missing them | David Conrad | 2010-07-21 | 1 | -0/+6 |
* | Move ff_pw_* from vc1dsp_mmx.c to dsputil_mmx.c | David Conrad | 2010-07-21 | 2 | -7/+1 |
* | VP8 MBedge loopfilter MMX/MMX2/SSE2 functions for both luma (width=16) | Ronald S. Bultje | 2010-07-20 | 4 | -2/+687 |
* | Chroma (width=8) inner loopfilter MMX/MMX2/SSE2 for VP8 decoder. | Ronald S. Bultje | 2010-07-20 | 2 | -77/+150 |
* | Revert r24339 (it causes fate failures on x86-64) - I'll figure out what's | Ronald S. Bultje | 2010-07-19 | 2 | -127/+32 |
* | Remove FF_MM_SSE2/3 flags for CPUs where this is generally not faster than | Ronald S. Bultje | 2010-07-19 | 3 | -6/+25 |
* | Implement chroma (width=8) inner loopfilter MMX/MMX2/SSE2 functions. | Ronald S. Bultje | 2010-07-19 | 2 | -32/+127 |
* | Be more efficient with registers or stack memory. Saves 8/16 bytes stack | Ronald S. Bultje | 2010-07-19 | 1 | -16/+16 |
* | Change function prototypes for width=8 inner and mbedge loopfilter functions | Ronald S. Bultje | 2010-07-19 | 2 | -19/+19 |
* | more credits to D. J. Bernstein for fft | Loren Merritt | 2010-07-18 | 1 | -0/+3 |
* | Attempt to fix x86-64 testsuite on fate. | Ronald S. Bultje | 2010-07-16 | 1 | -1/+1 |
* | Remove duplicate define. | Ronald S. Bultje | 2010-07-16 | 1 | -1/+0 |
* | Revert 24270, it contained some stuff that shouldn't have been in there. | Ronald S. Bultje | 2010-07-16 | 1 | -1/+2 |
* | Remove duplicate define. | Ronald S. Bultje | 2010-07-16 | 1 | -2/+1 |
* | Give x86 r%d registers names, this will simplify implementation of the chroma | Ronald S. Bultje | 2010-07-16 | 1 | -58/+81 |
* | Change return statement, the REP_RET is a mistake since the else case (x86-64, | Ronald S. Bultje | 2010-07-16 | 1 | -3/+1 |
* | VP8 H/V inner loopfilter MMX/MMXEXT/SSE2 optimizations. | Ronald S. Bultje | 2010-07-15 | 4 | -15/+488 |
* | MMX/SSE VC1 loop filter | David Conrad | 2010-07-11 | 4 | -0/+424 |
* | Make ff_pw_4 128 bits | David Conrad | 2010-07-11 | 4 | -4/+4 |
* | Move SSE optimized 32-point DCT to its own file. Should fix breakage with YASM | Vitor Sessak | 2010-07-06 | 4 | -266/+298 |
* | SSE optimized 32-point DCT | Vitor Sessak | 2010-07-06 | 3 | -0/+275 |
* | Simple H/V loopfilter for VP8 in MMX, MMX2 and SSE2 (yay for yasm macros). | Ronald S. Bultje | 2010-07-03 | 4 | -0/+334 |
* | SSSE3 versions of vp8 width4 bilinear MC functions | Jason Garrett-Glaser | 2010-07-03 | 2 | -4/+34 |
* | SSSE3 versions of width4 VP8 6-tap MC functions | Jason Garrett-Glaser | 2010-07-02 | 2 | -161/+192 |