468 Commits (d43cf68898cf577c76edd5483f52a63cefb667fb)

Author SHA1 Message Date
  Reimar Döffinger 00eebe3d6a Fix add_bytes_mmx and add_bytes_l2_mmx for w < 16 17 years ago
  Michael Niedermayer 0bd134abd3 Simplify vsad16_mmx2(). 17 years ago
  Michael Niedermayer 6bf6a9301b Simplify vsad16_mmx(). 17 years ago
  Michael Niedermayer e13810223a Simplify vsad_intra16_mmx2() 17 years ago
  Michael Niedermayer 06bb35f94c Simplify vsad_intra16_mmx() 17 years ago
  Diego Biurrun a12b44d7fb Add missing required header directly. 18 years ago
  Diego Biurrun 20cd685ae8 Add missing path to #include. 18 years ago
  Diego Biurrun 245976da2a Use full path for #includes from another directory. 18 years ago
  Ramiro Polla 40d0e665d0 Do not misuse long as the size of a register in x86. 18 years ago
  Diego Biurrun 57105ddd03 Rename i386/cputest.c --> i386/cpuid.c. 18 years ago
  Diego Biurrun c88c253d8b cosmetics: __asm__ __volatile__ --> asm volatile 18 years ago
  Diego Biurrun 80465c7eed cosmetics: Fix nonstandard indentation. 18 years ago
  Jeff Downs 591d87babe Cosmetics: 18 years ago
  Jeff Downs 52cb7981e2 Redo r12838, this time using svn copy to create h264_i386.h from cabac.h. 18 years ago
  Jeff Downs 3aa9ede400 Revert 12838 to redo it the right way (use svn copy to create new 18 years ago
  Alexander Strange f73a6393e7 Add a new xvid-style IDCT using SSE2. 18 years ago
  Jeff Downs e6cfd8fffb Move decode_significance_x86() and decode_significance_8x8_x86() to 18 years ago
  Luca Barbato 3fbe711832 Eliminate movdqu in vp3dsp_sse2, patch from Alexander Strange astrangeAtithinkswDoTcom 18 years ago
  Alexander Strange 54a0b6e590 Add a header file to declare Xvid IDCT functions. 18 years ago
  Loren Merritt 96275520a3 Fix H.264 interframe decoding when compiling with icc. Patch by Loren 18 years ago
  Loren Merritt ce53144bac h264 chroma mc ssse3 18 years ago
  Diego Biurrun 04932b0d97 cosmetics: typo fixes 18 years ago
  Zuxy Meng 9e8e6d318c Add missed call to ff_cavsdsp_init_3dnow() in dsputil_init_mmx() 18 years ago
  Michael Niedermayer 943032b155 Hardcode register to prevent aparent miscompilation. 18 years ago
  Michael Niedermayer dea00a4623 remove unused temp 18 years ago
  Måns Rullgård b55aa9a904 get register names from x86_cpu.h 18 years ago
  Aurelien Jacobs 5a6a9e78ab move draw_edges() into dsputil 18 years ago
  Aurelien Jacobs 97d1d009e2 split encoding part of dsputil_mmx into its own file 18 years ago
  Reimar Döffinger f2217d6f90 __asm __volatile -> asm volatile part 2 18 years ago
  Reimar Döffinger 78d3d94f14 __asm __volatile -> asm volatile, improves code consistency and works 18 years ago
  Loren Merritt 4a9ca0a279 simd and unroll png_filter_row 18 years ago
  Michael Niedermayer 1435e4ccde Disabling all SSE* code for old gcc to avoid alignment issues. 18 years ago
  Reimar Döffinger 754bf3d8a1 Fix warnings: 18 years ago
  Diego Biurrun 5edac5dc94 cosmetics: Replace // by /* */ comments. 18 years ago
  Loren Merritt ec199cc94c asm argument that might be in memory needs a size 18 years ago
  Loren Merritt 2c70770e33 use fewer registers in apply_welch_window_sse2 18 years ago
  Loren Merritt 1d67b037f7 sse2 h264 motion compensation. not new code, just separate out the cases that didn't need ssse3. 18 years ago
  Loren Merritt 20d565be6d put loop counter in a register if possible. makes some of the qpel functions 3% faster. 18 years ago
  Loren Merritt 7080ec2937 fix aliasing warnings. simpler too. 18 years ago
  Loren Merritt a2b7bc8e71 constant was excessively aligned 18 years ago
  Loren Merritt ddf969705f ssse3 h264 motion compensation. 18 years ago
  Loren Merritt b64dfbb8d2 add qpel rounder once during hv rather than twice during hv and whatever it's averaged with 18 years ago
  Loren Merritt fa9b873e08 clean up an ugliness introduced in r11826. this syntax will require fewer changes when adding future sse2 code. 18 years ago
  Michael Niedermayer 9a7871f77d Deprecate old and inefficient per instruction asm(). 18 years ago
  Loren Merritt b2f775860b reduce code duplication 18 years ago
  Loren Merritt b313e8159c avg_pixels4_mmx2 18 years ago
  Loren Merritt 6c01d0069d use mmx2/3dnow avg functions in avg_qpel*_mc00 18 years ago
  Loren Merritt ed5d7a531c ff_h264_idct8_add_sse2. 18 years ago
  Loren Merritt 51f0ac65df remove some movq in ff_h264_idct8_add_mmx. 225->217 cycles on core2. 18 years ago
  Baptiste Coudurier 066e0cc50d add parenthesis, fix warning: i386/dsputil_mmx.c:2618: warning: suggest parentheses around arithmetic in operand of | 18 years ago