1. 29 Jul, 2017 1 commit
  2. 28 Jul, 2017 1 commit
  3. 27 Jul, 2017 2 commits
  4. 26 Jul, 2017 3 commits
  5. 25 Jul, 2017 1 commit
    • James Zern's avatar
      vpx_variance16x16_sse2: correct cast order · 85736e61
      James Zern authored
      allow the right shift to operate on 64-bits, this matches the rest of
      the implementations
      b0f1ae14 vpx_get16x16var_avx2: correct cast order
      Change-Id: I632ee5e418f3f9b30e79ecd05588eb172b0783aa
  6. 24 Jul, 2017 1 commit
    • James Zern's avatar
      vpx_get16x16var_avx2: correct cast order · b0f1ae14
      James Zern authored
      allow the right shift to operate on 64-bits, this matches the rest of
      the implementations
      missed in:
      6acd061a variance_avx2: sync variance functions with c-code
      Change-Id: Icae436b881251ccb9f9ed64fcbf8d358c58a4617
  7. 22 Jul, 2017 1 commit
  8. 20 Jul, 2017 7 commits
  9. 19 Jul, 2017 5 commits
    • Johann Koenig's avatar
      Merge "Earmark extra space for VSX." · e1809501
      Johann Koenig authored
    • Jerome Jiang's avatar
    • Johann's avatar
      quantize test: eob is output · bde2e4aa
      Johann authored
      eob values are generated by the function.
      Change-Id: I8ce92100e83022bff99888a5a7e6ef378c49fda3
    • Han Shen's avatar
      Earmark extra space for VSX. · b72d3e8a
      Han Shen authored
      Backend specific optimization for PPC VSX reads 16 bytes, whereas arm neon /
      sse2 only reads <= 8 bytes. Although the extra bytes read are actually never
      used, this is not a warrant for groping around.  Fixed by allocating more when
      building for VSX. This is reported by asan.
      Also note - PPC does have assembly that loads 64-bit content from memory - lxsdx
      loads one 64-bit doubleword (whereas lxvd2x loads two 64-bit doubleword) from
      memory. However, we only have "vec_vsx_ld" builtins that mapped to lxvd2x, no
      builtins to lxsdx. The only way to access lxsdx is through inline assembly,
      which does not fit well in the origin paradigm.
          vpx_tm_predictor_4x4_vsx @ third_party/libvpx/git_root/vpx_dsp/ppc/intrapred_vsx.c
          vpx_tm_predictor_4x4_neon @ third_party/libvpx/git_root/vpx_dsp/arm/intrapred_neon_asm.asm
          tm_predictor_4x4 @ third_party/libvpx/git_root/vpx_dsp/x86/intrapred_sse2.asm
        asan tests passed.
      Change-Id: I5f74b56e35c05b67851de8b5530aece213f2ce9d
    • Johann Koenig's avatar
      Merge "variance: call C comp_avg_pred" · 89a116f4
      Johann Koenig authored
  10. 18 Jul, 2017 13 commits
  11. 17 Jul, 2017 2 commits
    • Marco's avatar
      vp9: Fix to setting content_state for real-time mode. · ad563713
      Marco authored
      When content_state_sb is set to LowVarHighSumdiff, don't reset
      it to VeryHighSad. Visually better on clips with strong lighting changes.
      Small/negligible change in RTC metrics and speed.
      Change-Id: I20c383e3c4cf8d1149de5f9260449c0b7cf7c6aa
    • Marco's avatar
      vp9: Reuse motion from choose_partitioning in NEWMV search. · 0c9e2f4c
      Marco authored
      When int_pro_motion_estimation is done for superblock in
      choose_partitioning, use it to avoid the full_pixel_search
      for NEWMV mode, if bsize is >= 32X32.
      For speed > 7.
      Small/neutral change on RTC metrics.
      ~1-2% speedup on arm on high motion clip.
      Change-Id: I3cfe6833ff4bf75d4afa83eaf058ad45729de85b
  12. 15 Jul, 2017 1 commit
  13. 14 Jul, 2017 2 commits