git.libav.org Git - libav.git/rss log

Les articles publiés sur le site

• audiodsp : fix vector_clipf documentation

  9 août 2016, par Anton Khirnov

  audiodsp: fix vector_clipf documentation
  
  The x86 version processes 16 floats per iteration, so len must be a
  multiple of 16.
  

  • [DBH] libavcodec/audiodsp.h

• checkasm : add tests for audiodsp

  9 août 2016, par Anton Khirnov

  checkasm: add tests for audiodsp
  

  • [DBH] tests/checkasm/Makefile
  • [DBH] tests/checkasm/audiodsp.c
  • [DBH] tests/checkasm/checkasm.c
  • [DBH] tests/checkasm/checkasm.h

• audiodsp/x86 : clear the high bits of the order parameter on 64bit

  9 août 2016, par Anton Khirnov

  audiodsp/x86: clear the high bits of the order parameter on 64bit
  
  Also change shl to add, since it can be faster on some CPUs.
  
  CC: libav-stable@libav.org
  

  • [DBH] libavcodec/x86/audiodsp.asm

• audiodsp/x86 : fix ff_vector_clip_int32_sse2

  9 août 2016, par Anton Khirnov

  audiodsp/x86: fix ff_vector_clip_int32_sse2
  
  This version, which is the only one doing two processing cycles per loop
  iteration, computes the load/store indices incorrectly for the second
  cycle.
  
  CC: libav-stable@libav.org
  

  • [DBH] libavcodec/x86/audiodsp.asm

• qsv{dec,enc} : always use an internal mfxFrameSurface1

  9 août 2016, par Anton Khirnov

  qsv{dec,enc}: always use an internal mfxFrameSurface1
  
  For encoding, this avoids modifying the input surface, which we are not
  allowed to do.
  This will also be useful in the following commits.
  
  Signed-off-by: Maxym Dmytrychenko <maxym.dmytrychenko@intel.com>
  

  • [DBH] libavcodec/qsv_internal.h
  • [DBH] libavcodec/qsvdec.c
  • [DBH] libavcodec/qsvenc.c