1. 26 Sep, 2010 1 commit
  2. 09 Sep, 2010 1 commit
  3. 08 Sep, 2010 1 commit
  4. 10 Jul, 2010 1 commit
  5. 07 Jul, 2010 2 commits
  6. 09 Mar, 2010 1 commit
  7. 08 Mar, 2010 1 commit
  8. 22 Feb, 2010 1 commit
  9. 19 Jan, 2010 1 commit
  10. 18 Jan, 2010 1 commit
  11. 16 Aug, 2009 1 commit
  12. 01 Jul, 2009 1 commit
  13. 18 Apr, 2009 1 commit
    • Måns Rullgård's avatar
      ARM asm for AV_RN*() · 3c55ce03
      Måns Rullgård authored
      ARMv6 and later support unaligned loads and stores for single
      word/halfword but not double/multiple.  GCC is ignorant of this and
      will always use bytewise accesses for unaligned data.  Casting to an
      int32_t pointer is dangerous since a load/store double or multiple
      instruction might be used (this happens with some code in FFmpeg).
      Implementing the AV_[RW]* macros with inline asm using only supported
      instructions gives fast and safe unaligned accesses.  ARM RVCT does
      the right thing with generic code.
      
      This gives an overall speedup of up to 10%.
      
      Originally committed as revision 18601 to svn://svn.ffmpeg.org/ffmpeg/trunk
      3c55ce03
  14. 01 Apr, 2009 1 commit
  15. 05 Mar, 2009 1 commit
  16. 13 Jan, 2009 1 commit
  17. 21 Oct, 2008 1 commit