1. 15 Aug, 2012 1 commit
  2. 08 Aug, 2012 1 commit
  3. 03 Aug, 2012 1 commit
    • Diego Biurrun's avatar
      x86: build: replace mmx2 by mmxext · 239fdf1b
      Diego Biurrun authored
      Refactoring mmx2/mmxext YASM code with cpuflags will force renames.
      So switching to a consistent naming scheme beforehand is sensible.
      The name "mmxext" is more official and widespread and also the name
      of the CPU flag, as reported e.g. by the Linux kernel.
      239fdf1b
  4. 02 Aug, 2012 1 commit
  5. 31 Jul, 2012 3 commits
  6. 28 Jul, 2012 1 commit
  7. 23 Jun, 2012 1 commit
  8. 21 Oct, 2011 3 commits
  9. 11 Jul, 2011 1 commit
  10. 21 Jun, 2011 1 commit
  11. 01 Jun, 2011 1 commit
  12. 31 May, 2011 2 commits
  13. 16 May, 2011 1 commit
  14. 11 May, 2011 3 commits
  15. 10 May, 2011 2 commits
  16. 19 Mar, 2011 1 commit
  17. 14 Jan, 2011 1 commit
  18. 29 Sep, 2010 6 commits
  19. 24 Sep, 2010 1 commit
  20. 21 Sep, 2010 1 commit
  21. 18 Sep, 2010 1 commit
  22. 14 Sep, 2010 1 commit
    • Ronald S. Bultje's avatar
      Rename h264_idct_sse2.asm to h264_idct.asm; move inline IDCT asm from · 1d16a1cf
      Ronald S. Bultje authored
      h264dsp_mmx.c to h264_idct.asm (as yasm code). Because the loops are now
      coded in asm instead of C, this is (depending on the function) up to 50%
      faster for cases where gcc didn't do a great job at looping.
      
      Since h264_idct_add8() is now faster than the manual loop setup in h264.c,
      in-asm idct calling can now be enabled for chroma as well (see r16207). For
      MMX, this is 5% faster. For SSE2 (which isn't done for chroma if h264.c does
      the looping), this makes it up to 50% faster. Speed gain overall is ~0.5-1.0%.
      
      Originally committed as revision 25119 to svn://svn.ffmpeg.org/ffmpeg/trunk
      1d16a1cf
  23. 10 Sep, 2010 1 commit
  24. 08 Sep, 2010 1 commit
  25. 04 Sep, 2010 1 commit
  26. 03 Sep, 2010 1 commit
  27. 01 Sep, 2010 1 commit