6229 Commits (03c09e43993a7c8e4db7930a0e6be736457af494)

Author SHA1 Message Date
Loren Merritt ec199cc94c asm argument that might be in memory needs a size 17 years ago
Michael Niedermayer a3689bf7b4 Remove pointless code. 17 years ago
Diego Biurrun 7bab3a26e4 Use HAVE_ALTIVEC preprocessor conditionals like in the rest of FFmpeg. 17 years ago
Michael Niedermayer dc6340e048 Remove pointless code. 17 years ago
Michael Niedermayer a4195e68b6 assert() and note about odd len. 17 years ago
Loren Merritt e1813a2f1e 30% faster ff_vorbis_floor1_render_list, 3% faster overall 17 years ago
Loren Merritt 2c70770e33 use fewer registers in apply_welch_window_sse2 17 years ago
Michael Niedermayer fa4ed8c10a optimize apply_welch_window() 17 years ago
Reimar Döffinger cb4ddf7765 Add support for speex in ogg 17 years ago
Loren Merritt 1d67b037f7 sse2 h264 motion compensation. not new code, just separate out the cases that didn't need ssse3. 17 years ago
Loren Merritt 20d565be6d put loop counter in a register if possible. makes some of the qpel functions 3% faster. 17 years ago
Loren Merritt 7080ec2937 fix aliasing warnings. simpler too. 17 years ago
Loren Merritt a2b7bc8e71 constant was excessively aligned 17 years ago
Reimar Döffinger 63f2670a53 Check that we have enough input data in IMC decoder. 17 years ago
Loren Merritt ddf969705f ssse3 h264 motion compensation. 17 years ago
Loren Merritt b64dfbb8d2 add qpel rounder once during hv rather than twice during hv and whatever it's averaged with 17 years ago
Loren Merritt fa9b873e08 clean up an ugliness introduced in r11826. this syntax will require fewer changes when adding future sse2 code. 17 years ago
Michael Niedermayer 9a7871f77d Deprecate old and inefficient per instruction asm(). 17 years ago
Loren Merritt b2f775860b reduce code duplication 17 years ago
Diego Biurrun 1db38dad33 Make luma_dc_quant_i and luma_dc_quant_p const, fixes a couple of 17 years ago
Michael Niedermayer 1701cbfafc const 17 years ago
Michael Niedermayer 20e4beaef7 const 17 years ago
Uoti Urpala 4bdd3b76fe mpegvideo.h has two function declarations with the 'inline' specifier 17 years ago
Loren Merritt b313e8159c avg_pixels4_mmx2 17 years ago
Loren Merritt 6c01d0069d use mmx2/3dnow avg functions in avg_qpel*_mc00 17 years ago
Loren Merritt ed5d7a531c ff_h264_idct8_add_sse2. 17 years ago
Loren Merritt 51f0ac65df remove some movq in ff_h264_idct8_add_mmx. 225->217 cycles on core2. 17 years ago
Michael Niedermayer 2efa7fd10d Make avcodec_decode_* functions take const input buffers. 17 years ago
Reimar Döffinger 909063f745 100l, since we already check for buf_size == 769 we should also 17 years ago
Reimar Döffinger 3e368d726a Use bytestream_get_be24 to simplify palette parsing. 17 years ago
Reimar Döffinger 04deea9ad2 mark qtrle input data as const. 17 years ago
Reimar Döffinger b3d5468d87 Add some const where appropriate 17 years ago
Reimar Döffinger 6be4042c52 Fix warning: 17 years ago
Reimar Döffinger eee7e7998b Some consts for cscd decoder helper functions 17 years ago
Loren Merritt ffbc5e04ce "swap first two elements of L1 if L0 and L1 are identical" applies after splitting the lists into fields 17 years ago
Loren Merritt 02f7695b0b fix mbaff diagonal neighbor mv 17 years ago
Loren Merritt 50b3ab0fa5 mbaff spatial direct 17 years ago
Kostya Shishkov 4990eb5f49 Do not modify input data 17 years ago
Kostya Shishkov 0fe04628b9 IMC decoder always operates on 64-byte blocks 17 years ago
Aurelien Jacobs 2755e41d26 use AV_RL* instead of le2me_* 17 years ago
Aurelien Jacobs 4e56af8253 don't cast const away 17 years ago
Aurelien Jacobs 60b6dbe716 const 17 years ago
Aurelien Jacobs 9f5bd89506 const 17 years ago
Michael Niedermayer c67102f902 const 17 years ago
Michael Niedermayer 31d889ec0a const 17 years ago
Michael Niedermayer 896b438915 const 17 years ago
Michael Niedermayer 1786784259 const 17 years ago
Reimar Döffinger b034f72b17 Add some const keywords in rtjpeg decoder 17 years ago
Michael Niedermayer 7e5cdb5b1b const 17 years ago
Michael Niedermayer 74dc744850 const 17 years ago