Revisions : libvpx

Les articles publiés sur le site

• Revision 41a0a0cb35 : Use newer x86inc.asm Rename updated version of x86inc.asm Use "private_prefix"

  30 juillet 2015, par Johann
  Changed Paths:
      Modify    /third_party/x86inc/x86inc.asm
      Delete    /third_party/x86inc/x86inc.asm.new
      Modify    /vp9/encoder/x86/vp9_dct_mmx.asm
      Modify    /vp9/encoder/x86/vp9_dct_ssse3_x86_64.asm
      Modify    /vp9/encoder/x86/vp9_error_sse2.asm
      Modify    /vp9/encoder/x86/vp9_quantize_ssse3_x86_64.asm
      Modify    /vpx_dsp/x86/fwd_txfm_ssse3_x86_64.asm
      Modify    /vpx_dsp/x86/highbd_intrapred_sse2.asm
      Modify    /vpx_dsp/x86/highbd_sad4d_sse2.asm
      Modify    /vpx_dsp/x86/highbd_sad_sse2.asm
      Modify    /vpx_dsp/x86/highbd_subpel_variance_impl_sse2.asm
      Modify    /vpx_dsp/x86/intrapred_sse2.asm
      Modify    /vpx_dsp/x86/intrapred_ssse3.asm
      Modify    /vpx_dsp/x86/inv_txfm_sse2.asm
      Modify    /vpx_dsp/x86/inv_txfm_ssse3_x86_64.asm
      Modify    /vpx_dsp/x86/quantize_ssse3_x86_64.asm
      Modify    /vpx_dsp/x86/sad4d_sse2.asm
      Modify    /vpx_dsp/x86/sad_sse2.asm
      Modify    /vpx_dsp/x86/subpel_variance_sse2.asm
      Modify    /vpx_dsp/x86/subtract_sse2.asm
      Modify    /vpx_dsp/x86/vpx_convolve_copy_sse2.asm
  
  Use newer x86inc.asm Rename updated version of x86inc.asm Use "private_prefix" instead of "program_name" and make vpx the default prefix. Change-Id: I4883a99b2aee8e5dc9f2c16a2e6f4b5d6e4de458

• Revision c7e9e486b8 : Use .text instead of .rodata on macho The read only sections are getting stripp

  30 juillet 2015, par Johann
  Changed Paths:
      Modify    /third_party/x86inc/README.libvpx
      Modify    /third_party/x86inc/x86inc.asm.new
  
  Use .text instead of .rodata on macho The read only sections are getting stripped on some OS X builds. As a result, random data is used in place of the intended tables. Change-Id: I58c18a53e503f093ee268451698c5761e6c32540

• Revision a5e97d874b : VP9_COPY_CONVOLVE_SSE2 optimization This function suffers from a couple problem

  30 juillet 2015, par Scott LaVarnway
  Changed Paths:
      Modify    /vpx_dsp/x86/vpx_convolve_copy_sse2.asm
  
  VP9_COPY_CONVOLVE_SSE2 optimization This function suffers from a couple problems in small core(tablets): -The load of the next iteration is blocked by the store of previous iteration -4k aliasing (between future store and older loads) -current small core machine are in-order machine and because of it the store will spin the rehabQ until the load is finished fixed by: - prefetching 2 lines ahead - unroll copy of 2 rows of block - pre-load all xmm regiters before the loop, final stores after the loop The function is optimized by: copy_convolve_sse2 64x64 - 16% copy_convolve_sse2 32x32 - 52% copy_convolve_sse2 16x16 - 6% copy_convolve_sse2 8x8 - 2.5% copy_convolve_sse2 4x4 - 2.7% credit goes to Tom Craver(tom.r.craver@intel.com) and Ilya Albrekht(ilya.albrekht@intel.com) Change-Id: I63d3428799c50b2bf7b5677c8268bacb9fc29671

• Revision 91feec1452 : Merge "Cosmetics - Fix header file order in unit tests"

  30 juillet 2015, par Jingning Han
  
  Merge "Cosmetics - Fix header file order in unit tests"

• Revision 56aa0da405 : mips msa vp8 quantize optimization average improvement 2x-3x Change-Id : I6fc3

  30 juillet 2015, par Parag Salasakar
  Changed Paths:
      Modify    /test/quantize_test.cc
      Modify    /vp8/common/mips/msa/vp8_macros_msa.h
      Modify    /vp8/common/rtcd_defs.pl
      Add    /vp8/encoder/mips/msa/quantize_msa.c
      Modify    /vp8/vp8cx.mk
  
  mips msa vp8 quantize optimization average improvement ~2x-3x Change-Id: I6fc37191bf9cb5a67e1af9787d0d27659c17bdba