42 Commits (2b12d6fc79817d14b555b13847e6b328fbc3d2bd)

Author SHA1 Message Date
Christophe Gisquet 28748a9128 Factorize some duplicated code from CAVS and H.264 into a common file. 17 years ago
Christophe Gisquet 9fa3572903 add MMX version for put_no_rnd_h264_chroma_mc8_c, used in VC-1 decoding. 17 years ago
Aurelien Jacobs 43de50659b use ff_ prefix for extern vars 17 years ago
Aurelien Jacobs 182f56cb0c make ff_p* vars extern so that they can be used in various *_mmx.c files 17 years ago
Shane bdb273564c Fix intended order of operations for 4 assert() checks. 17 years ago
Loren Merritt bba5293bb7 cosmetics: remove duplicate transpose macro 18 years ago
Reimar Döffinger a1ce61108b Fix parts missed in clip -> av_clip rename 18 years ago
Michael Niedermayer 2a115873af add a few asserts to ensure alignment 18 years ago
Michael Niedermayer 00e210ddbb prevent h.264 MC related functions from being inlined (yes this is much faster the code just doesnt fit in the code cache otherwise) 18 years ago
Michael Niedermayer 079e61db5d ensure alignment (no speed change) 18 years ago
Michael Niedermayer f5a9e8f33d merging mov & and (no speedchange) 18 years ago
Michael Niedermayer e80cf125a7 2 instructions less (same speed) 18 years ago
Michael Niedermayer 9347118237 comment about failed optimization 18 years ago
Michael Niedermayer 38cfdc83f0 move luma tc0 related init into asm 18 years ago
Michael Niedermayer 25225c3773 2 instructions less in h264_loop_filter_luma_mmx2() 18 years ago
Michael Niedermayer bda2203d56 preempt possible overflow 18 years ago
Michael Niedermayer 5a1553dee3 1 instruction less 18 years ago
Michael Niedermayer e9f1885c21 optimize H264_DEBLOCK_P0_Q0 18 years ago
Diego Biurrun b78e7197a8 Change license headers to say 'FFmpeg' instead of 'this program/this library' 18 years ago
Loren Merritt 3e20143ee7 mmx implementation of deblocking strength decision. 18 years ago
Måns Rullgård 3f8674a902 remove redundant macro definitions 19 years ago
Loren Merritt fdd3057981 added mmx implementation of h264_chroma_mc2 19 years ago
Loren Merritt 548a1c8a35 h264_idct8_add_mmx 19 years ago
Loren Merritt 6da971f160 h264_idct_add only needs mmx1 19 years ago
Loren Merritt d84f7c61ee gcc2.95 workaround 19 years ago
Loren Merritt 7a5b2fa812 remove some useless instructions 19 years ago
Loren Merritt 6a8eb0f45a 4% faster h264_qpel_mc 19 years ago
Loren Merritt ef9d1d1575 h264: special case dc-only idct. ~1% faster overall 19 years ago
Loren Merritt 4e295993ba 10l in 1.12 19 years ago
Loren Merritt 6ee669732d 10l (x86_64) 19 years ago
Loren Merritt e545f37527 18% faster put_h264_qpel16_mc[13]2_mmx2 19 years ago
Loren Merritt c03ce51dfb 11% faster put_h264_qpel16_v_lowpass_mmx2 19 years ago
Loren Merritt 0331f09237 15% faster put_h264_qpel16_hv_lowpass_mmx2 19 years ago
Diego Biurrun 5509bffa88 Update licensing information: The FSF changed postal address. 19 years ago
Loren Merritt e8b562087d tweak h264_biweight 19 years ago
Loren Merritt cec9395977 fix some potential arithmetic overflows in pred_direct_motion() and 19 years ago
Diego Biurrun bb270c0896 COSMETICS: tabs --> spaces, some prettyprinting 19 years ago
Diego Biurrun 115329f160 COSMETICS: Remove all trailing whitespace. 19 years ago
Loren Merritt a6624e21cb faster h264_chroma_mc8_mmx, added h264_chroma_mc4_mmx. 19 years ago
Loren Merritt b926572aa9 h264 mmx weighted prediction. up to 3% overall speedup. 19 years ago
Michael Niedermayer d3a9f79871 simplify (d&a) and (d&~a) calculation, hint by skal 19 years ago
Loren Merritt d2bb7db135 sort H.264 mmx dsp functions into their own file 20 years ago