summaryrefslogtreecommitdiff
path: root/libavcodec/i386
Commit message (Expand)AuthorAge
* 3DNow! & Extended 3DNow! versions of FFTZuxy Meng2006-03-08
* h264_idct8_add_mmxLoren Merritt2006-03-07
* h264_idct_add only needs mmx1Loren Merritt2006-03-07
* use xorps instead of mulps to toggle the sign of a float, as suggested by Sof...Zuxy Meng2006-03-05
* gcc2.95 workaroundLoren Merritt2006-03-05
* remove some useless instructionsLoren Merritt2006-03-04
* 4% faster h264_qpel_mcLoren Merritt2006-03-02
* h264: special case dc-only idct. ~1% faster overallLoren Merritt2006-02-10
* 10l in 1.12Loren Merritt2006-02-09
* 10l (x86_64)Loren Merritt2006-02-07
* 18% faster put_h264_qpel16_mc[13]2_mmx2Loren Merritt2006-02-07
* 11% faster put_h264_qpel16_v_lowpass_mmx2Loren Merritt2006-02-07
* 15% faster put_h264_qpel16_hv_lowpass_mmx2Loren Merritt2006-02-07
* MSVC-compatible __align8/__align16 declarationSteve L'Homme2006-02-05
* Update licensing information: The FSF changed postal address.Diego Biurrun2006-01-12
* tweak h264_biweightLoren Merritt2006-01-09
* fix some potential arithmetic overflows in pred_direct_motion() andLoren Merritt2005-12-30
* COSMETICS: tabs --> spaces, some prettyprintingDiego Biurrun2005-12-22
* COSMETICS: Remove all trailing whitespace.Diego Biurrun2005-12-17
* Add the rest of missing Reg_* macros to support both AMD-64 style regs and IA...Guillaume Poirier2005-12-10
* use sse16_sse2() in nsseLoren Merritt2005-11-12
* faster h264_chroma_mc8_mmx, added h264_chroma_mc4_mmx.Loren Merritt2005-10-27
* h264 mmx weighted prediction. up to 3% overall speedup.Loren Merritt2005-10-09
* sse2 16x16 sum squared diff (306=>268 cycles on a K8)Loren Merritt2005-09-30
* replace a few mov + psrlq with pshufw, there are more cases which could benef...Michael Niedermayer2005-09-21
* Fix compile without CONFIG_GPL, misplaced #endif caused a missing }.Reimar Döffinger2005-09-10
* remove unused table entriesMichael Niedermayer2005-09-10
* xvids mmx&mmx2 idctsMichael Niedermayer2005-09-10
* Kill some compiler warnings. Compiled code verified identical after changes.Måns Rullgård2005-09-06
* simplify (d&a) and (d&~a) calculation, hint by skalMichael Niedermayer2005-09-03
* add consts (this was in my local tree, dunno where it came from, probably for...Michael Niedermayer2005-09-02
* kill a bunch of compiler warningsMåns Rullgård2005-08-14
* libavutil: Utility code from libavcodec moved to a separate library.Alexander Strasser2005-08-01
* sort H.264 mmx dsp functions into their own fileLoren Merritt2005-06-02
* adding a few constMichael Niedermayer2005-06-01
* 100l for myself (breaking amd64)Michael Niedermayer2005-06-01
* merge a few asm blocks so gcc cant unoptimize it (658->631 dezicycles on duron)Michael Niedermayer2005-06-01
* get rid of 2 movq (680 -> 658 dezicycles on duron)Michael Niedermayer2005-06-01
* avoid one transpose (730->680 dezicycles on duron)Michael Niedermayer2005-06-01
* 10l (symbol mangling)Loren Merritt2005-06-01
* add rounding bias before the horizontal idct (765->730 dezicyles on duron)Michael Niedermayer2005-06-01
* MMX for H.264 iDCT (adapted from x264)Loren Merritt2005-05-31
* MMX code for (put|avg)_h264_chroma_mc8Zoltán Hidvégi2005-05-25
* H.264 deblocking optimizations (mmx for chroma_bS4 case, convert existing cas...Loren Merritt2005-05-18
* porting the mmx&sse2 (sse2 untested) vp3 idcts to the lavc idct APIMichael Niedermayer2005-05-17
* disabling vp3 mmx&mmx2 idcts, they must be ported over to the lavc idct API, ...Michael Niedermayer2005-05-17
* fix PICMichael Niedermayer2005-05-08
* MMX for H.264 deblocking filterLoren Merritt2005-04-25
* read 32bit instead of 64bit to avoid overreading and missalignmentsMichael Niedermayer2005-04-17
* optimization and gcc 4.0 bug workaround patch by (Martin Drab >drab kepler.fj...Martin Drab2005-02-07
generated by cgit v1.2.3 (git 2.39.1) at 2024-07-12 15:21:12 +0000