]> git.sesse.net Git - ffmpeg/history - libavcodec/i386
flashsv: use skip_bits_long() where required
[ffmpeg] / libavcodec / i386 /
2008-11-24 Måns RullgårdAdd shift argument to MULL() macro
2008-11-14 Baptiste Coudurierconvert fdct_mmx to plain asm
2008-11-03 Dominik MierzejewskiRemove duplicated MM_* macros for CPU capabilities...
2008-10-29 Diego Biurrunasm() --> __asm__() in #warning comment
2008-10-29 Diego BiurrunAdd missing headers to pass 'make checkheaders'.
2008-10-27 Diego PettenòRename template included sources from .h to _template.c.
2008-10-20 Reimar DöffingerUse x86_reg type instead of long in float_to_int16...
2008-10-19 David ConradCosmetics: reindent
2008-10-19 David ConradCombine non-bitexact sections
2008-10-19 David ConradVP3 loop filter is mmx2 not mmx
2008-10-17 David ConradMMX VP3 Loop Filter
2008-10-16 Diego PettenòConvert asm keyword into __asm__.
2008-10-09 Baptiste Coudurierget_pixels_sse2, ~+12% performance compared to mmx
2008-10-07 Diego BiurrunAdd a bunch of const qualifiers to function arguments.
2008-10-07 Diego BiurrunSync cosmetics from upstream.
2008-10-04 Diego PettenòDeclare ff_pw_53 and ff_pw_18 as assembler constants.
2008-09-18 David ConradRemove duplication of TRANSPOSE8 macro, also removing...
2008-09-18 David ConradMake a separate macro for writing an entire block from...
2008-09-18 David ConradFactorize SSE2_(Row|Column)_IDCT into one macro
2008-09-17 David ConradCosmetics: remove excessive blank lines
2008-09-17 David ConradEnsure MMX/SSE2 VP3 IDCT selection isn't disabled when...
2008-09-17 David ConradCosmetics: reindent
2008-09-17 David ConradRewrite SSE2 VP3 IDCT in inline asm
2008-09-17 David ConradRewrite MMX VP3 IDCT in inline asm
2008-09-17 David ConradMMX/SSE2 VP3 IDCT are bitexact now that the dequantizat...
2008-08-31 Stefano SabatiniGlobally rename the header inclusion guard names.
2008-08-31 David ConradUse ff_vp3_idct_data in vp3dsp_mmx.c rather than duplic...
2008-08-31 David ConradDeclare ff_vp3_idct_data to be uint16_t
2008-08-31 David ConradDon't declare SSE vp3 idct data static, so it can be...
2008-08-30 David ConradRemove unused dequantization code from SSE VP3 IDCT
2008-08-30 David ConradUse ff_pw_8 in MMX/SSE VP3 IDCT
2008-08-30 David ConradLet ff_pw_8 be used as an SSE constant
2008-08-26 Vladimir VoroshilovAdd explicit (int) cast to i386 optimized MUL* macros.
2008-08-24 Alexis Ballierstricter constraints of asm() blocks
2008-08-14 Loren Merrittfile which should have been added in r14749
2008-08-14 Loren Merrittmissing prototype
2008-08-14 Loren Merrittgcc chokes on the 7 registers needed for float_to_int16...
2008-08-14 Loren Merrittgcc chokes on xmm constraints, so pessimize int32_to_fl...
2008-08-13 Loren Merrittspecial case 6 channel version of float_to_int16_interleave
2008-08-13 Loren Merrittsimd int->float
2008-08-13 Loren Merrittsimd downmix
2008-08-12 Loren Merrittcosmetics in dsp init
2008-08-12 Loren Merrittremove mdct tmp buffer
2008-08-12 Loren Merrittoptimize imdct_half:
2008-08-12 Loren Merrittsplit-radix FFT
2008-08-11 Loren Merrittimport yasm macros from x264
2008-08-02 Uoti UrpalaMark add_png_paeth_prediction_* functions which are...
2008-07-19 Michael NiedermayerFix h264_loop_filter_strength_mmx2() so it works with...
2008-07-16 Loren Merrittfloat_to_int16_interleave: change src to an array of...
2008-07-15 Loren Merritt10l, float_to_int16_interleave_sse/3dnow wrote the...
2008-07-13 Loren Merrittexploit mdct symmetry
2008-07-13 Loren Merrittsimplify vorbis windowing
2008-07-11 Kostya ShishkovSSE2 optimizations for Monkey's Audio decoder vector...
2008-07-09 Alexander StrangeMake the function prototype visible to comply with...
2008-07-09 Michael Niedermayerfloat_to_int16_sse2()
2008-07-08 Victor PollexMake LOAD4/STORE4 macros more generic.
2008-07-07 Michael Niedermayer1 c-asm loop less and 1x unroll of float_to_int16_sse()
2008-07-07 Michael NiedermayerFix x86-64
2008-07-07 Michael Niedermayerdont use C-asm loops and unroll once float_to_int16_3dnow()
2008-06-28 Alexander StrangeRealign newlines.
2008-06-28 Alexander StrangeUse MANGLE() instead of memory operands to read globals.
2008-06-22 Reimar DöffingerFix add_bytes_mmx and add_bytes_l2_mmx for w < 16
2008-05-17 Michael NiedermayerSimplify vsad16_mmx2().
2008-05-17 Michael NiedermayerSimplify vsad16_mmx().
2008-05-17 Michael NiedermayerSimplify vsad_intra16_mmx2()
2008-05-17 Michael NiedermayerSimplify vsad_intra16_mmx()
2008-05-09 Diego BiurrunAdd missing required header directly.
2008-05-09 Diego BiurrunAdd missing path to #include.
2008-05-09 Diego BiurrunUse full path for #includes from another directory.
2008-05-08 Ramiro PollaDo not misuse long as the size of a register in x86.
2008-04-26 Diego BiurrunRename i386/cputest.c --> i386/cpuid.c.
2008-04-17 Diego Biurruncosmetics: __asm__ __volatile__ --> asm volatile
2008-04-16 Diego Biurruncosmetics: Fix nonstandard indentation.
2008-04-16 Jeff DownsCosmetics:
2008-04-16 Jeff DownsRedo r12838, this time using svn copy to create h264_i3...
2008-04-16 Jeff DownsRevert 12838 to redo it the right way (use svn copy...
2008-04-16 Alexander StrangeAdd a new xvid-style IDCT using SSE2.
2008-04-15 Jeff DownsMove decode_significance_x86() and decode_significance_...
2008-04-14 Luca BarbatoEliminate movdqu in vp3dsp_sse2, patch from Alexander...
2008-04-12 Alexander StrangeAdd a header file to declare Xvid IDCT functions.
2008-04-08 Loren MerrittFix H.264 interframe decoding when compiling with icc...
2008-04-01 Loren Merritth264 chroma mc ssse3
2008-03-22 Diego Biurruncosmetics: typo fixes
2008-03-21 Zuxy MengAdd missed call to ff_cavsdsp_init_3dnow() in dsputil_i...
2008-03-20 Michael NiedermayerHardcode register to prevent aparent miscompilation.
2008-03-20 Michael Niedermayerremove unused temp
2008-03-17 Måns Rullgårdget register names from x86_cpu.h
2008-03-04 Aurelien Jacobsmove draw_edges() into dsputil
2008-02-25 Aurelien Jacobssplit encoding part of dsputil_mmx into its own file
2008-02-24 Reimar Döffinger__asm __volatile -> asm volatile part 2
2008-02-24 Reimar Döffinger__asm __volatile -> asm volatile, improves code consist...
2008-02-21 Loren Merrittsimd and unroll png_filter_row
2008-02-21 Michael NiedermayerDisabling all SSE* code for old gcc to avoid alignment...
2008-02-19 Reimar DöffingerFix warnings:
2008-02-13 Diego Biurruncosmetics: Replace // by /* */ comments.
2008-02-10 Loren Merrittasm argument that might be in memory needs a size
2008-02-09 Loren Merrittuse fewer registers in apply_welch_window_sse2
2008-02-06 Loren Merrittsse2 h264 motion compensation. not new code, just separ...
2008-02-06 Loren Merrittput loop counter in a register if possible. makes some...
2008-02-06 Loren Merrittfix aliasing warnings. simpler too.
next