Skip to content

GitLab

A
aom-rav1e
Commit 2cdd3bea authored by Parag Salasakar's avatar Parag Salasakar Committed by Gerrit Code Review
Browse files
Options

Merge "mips msa vp9 avg subpel variance optimization rebased"

parents aeee70f9 1d7f1ca7
Expand all Hide whitespace changes
Inline Side-by-side
Showing with 1357 additions and 79 deletions
test/variance_test.cc
View file @ 2cdd3bea
......@@ -1994,5 +1994,47 @@ INSTANTIATE_TEST_CASE_P(
make_tuple(5, 6, subpel_variance32x64_msa, 0),
make_tuple(6, 5, subpel_variance64x32_msa, 0),
make_tuple(6, 6, subpel_variance64x64_msa, 0)));
const SubpixAvgVarMxNFunc subpel_avg_variance64x64_msa =
vpx_sub_pixel_avg_variance64x64_msa;
const SubpixAvgVarMxNFunc subpel_avg_variance64x32_msa =
vpx_sub_pixel_avg_variance64x32_msa;
const SubpixAvgVarMxNFunc subpel_avg_variance32x64_msa =
vpx_sub_pixel_avg_variance32x64_msa;
const SubpixAvgVarMxNFunc subpel_avg_variance32x32_msa =
vpx_sub_pixel_avg_variance32x32_msa;
const SubpixAvgVarMxNFunc subpel_avg_variance32x16_msa =
vpx_sub_pixel_avg_variance32x16_msa;
const SubpixAvgVarMxNFunc subpel_avg_variance16x32_msa =
vpx_sub_pixel_avg_variance16x32_msa;
const SubpixAvgVarMxNFunc subpel_avg_variance16x16_msa =
vpx_sub_pixel_avg_variance16x16_msa;
const SubpixAvgVarMxNFunc subpel_avg_variance16x8_msa =
vpx_sub_pixel_avg_variance16x8_msa;
const SubpixAvgVarMxNFunc subpel_avg_variance8x16_msa =
vpx_sub_pixel_avg_variance8x16_msa;
const SubpixAvgVarMxNFunc subpel_avg_variance8x8_msa =
vpx_sub_pixel_avg_variance8x8_msa;
const SubpixAvgVarMxNFunc subpel_avg_variance8x4_msa =
vpx_sub_pixel_avg_variance8x4_msa;
const SubpixAvgVarMxNFunc subpel_avg_variance4x8_msa =
vpx_sub_pixel_avg_variance4x8_msa;
const SubpixAvgVarMxNFunc subpel_avg_variance4x4_msa =
vpx_sub_pixel_avg_variance4x4_msa;
INSTANTIATE_TEST_CASE_P(
MSA, VpxSubpelAvgVarianceTest,
::testing::Values(make_tuple(6, 6, subpel_avg_variance64x64_msa, 0),
make_tuple(6, 5, subpel_avg_variance64x32_msa, 0),
make_tuple(5, 6, subpel_avg_variance32x64_msa, 0),
make_tuple(5, 5, subpel_avg_variance32x32_msa, 0),
make_tuple(5, 4, subpel_avg_variance32x16_msa, 0),
make_tuple(4, 5, subpel_avg_variance16x32_msa, 0),
make_tuple(4, 4, subpel_avg_variance16x16_msa, 0),
make_tuple(4, 3, subpel_avg_variance16x8_msa, 0),
make_tuple(3, 4, subpel_avg_variance8x16_msa, 0),
make_tuple(3, 3, subpel_avg_variance8x8_msa, 0),
make_tuple(3, 2, subpel_avg_variance8x4_msa, 0),
make_tuple(2, 3, subpel_avg_variance4x8_msa, 0),
make_tuple(2, 2, subpel_avg_variance4x4_msa, 0)));
#endif // HAVE_MSA
} // namespace
vpx_dsp/mips/sub_pixel_variance_msa.c
View file @ 2cdd3bea
This diff is collapsed.
vpx_dsp/vpx_dsp_rtcd_defs.pl
View file @ 2cdd3bea
......@@ -636,43 +636,43 @@ add_proto qw/uint32_t vpx_sub_pixel_variance4x4/, "const uint8_t *src_ptr, int s
specialize qw/vpx_sub_pixel_variance4x4 mmx msa/, "$sse_x86inc", "$ssse3_x86inc";
add_proto qw/uint32_t vpx_sub_pixel_avg_variance64x64/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred";
specialize qw/vpx_sub_pixel_avg_variance64x64 avx2/, "$sse2_x86inc", "$ssse3_x86inc";
specialize qw/vpx_sub_pixel_avg_variance64x64 avx2 msa/, "$sse2_x86inc", "$ssse3_x86inc";
add_proto qw/uint32_t vpx_sub_pixel_avg_variance64x32/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred";
specialize qw/vpx_sub_pixel_avg_variance64x32/, "$sse2_x86inc", "$ssse3_x86inc";
specialize qw/vpx_sub_pixel_avg_variance64x32 msa/, "$sse2_x86inc", "$ssse3_x86inc";
add_proto qw/uint32_t vpx_sub_pixel_avg_variance32x64/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred";
specialize qw/vpx_sub_pixel_avg_variance32x64/, "$sse2_x86inc", "$ssse3_x86inc";
specialize qw/vpx_sub_pixel_avg_variance32x64 msa/, "$sse2_x86inc", "$ssse3_x86inc";
add_proto qw/uint32_t vpx_sub_pixel_avg_variance32x32/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred";
specialize qw/vpx_sub_pixel_avg_variance32x32 avx2/, "$sse2_x86inc", "$ssse3_x86inc";
specialize qw/vpx_sub_pixel_avg_variance32x32 avx2 msa/, "$sse2_x86inc", "$ssse3_x86inc";
add_proto qw/uint32_t vpx_sub_pixel_avg_variance32x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred";
specialize qw/vpx_sub_pixel_avg_variance32x16/, "$sse2_x86inc", "$ssse3_x86inc";
specialize qw/vpx_sub_pixel_avg_variance32x16 msa/, "$sse2_x86inc", "$ssse3_x86inc";
add_proto qw/uint32_t vpx_sub_pixel_avg_variance16x32/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred";
specialize qw/vpx_sub_pixel_avg_variance16x32/, "$sse2_x86inc", "$ssse3_x86inc";
specialize qw/vpx_sub_pixel_avg_variance16x32 msa/, "$sse2_x86inc", "$ssse3_x86inc";
add_proto qw/uint32_t vpx_sub_pixel_avg_variance16x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred";
specialize qw/vpx_sub_pixel_avg_variance16x16/, "$sse2_x86inc", "$ssse3_x86inc";
specialize qw/vpx_sub_pixel_avg_variance16x16 msa/, "$sse2_x86inc", "$ssse3_x86inc";
add_proto qw/uint32_t vpx_sub_pixel_avg_variance16x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred";
specialize qw/vpx_sub_pixel_avg_variance16x8/, "$sse2_x86inc", "$ssse3_x86inc";
specialize qw/vpx_sub_pixel_avg_variance16x8 msa/, "$sse2_x86inc", "$ssse3_x86inc";
add_proto qw/uint32_t vpx_sub_pixel_avg_variance8x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred";
specialize qw/vpx_sub_pixel_avg_variance8x16/, "$sse2_x86inc", "$ssse3_x86inc";
specialize qw/vpx_sub_pixel_avg_variance8x16 msa/, "$sse2_x86inc", "$ssse3_x86inc";
add_proto qw/uint32_t vpx_sub_pixel_avg_variance8x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred";
specialize qw/vpx_sub_pixel_avg_variance8x8/, "$sse2_x86inc", "$ssse3_x86inc";
specialize qw/vpx_sub_pixel_avg_variance8x8 msa/, "$sse2_x86inc", "$ssse3_x86inc";
add_proto qw/uint32_t vpx_sub_pixel_avg_variance8x4/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred";
specialize qw/vpx_sub_pixel_avg_variance8x4/, "$sse2_x86inc", "$ssse3_x86inc";
specialize qw/vpx_sub_pixel_avg_variance8x4 msa/, "$sse2_x86inc", "$ssse3_x86inc";
add_proto qw/uint32_t vpx_sub_pixel_avg_variance4x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred";
specialize qw/vpx_sub_pixel_avg_variance4x8/, "$sse_x86inc", "$ssse3_x86inc";
specialize qw/vpx_sub_pixel_avg_variance4x8 msa/, "$sse_x86inc", "$ssse3_x86inc";
add_proto qw/uint32_t vpx_sub_pixel_avg_variance4x4/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred";
specialize qw/vpx_sub_pixel_avg_variance4x4/, "$sse_x86inc", "$ssse3_x86inc";
specialize qw/vpx_sub_pixel_avg_variance4x4 msa/, "$sse_x86inc", "$ssse3_x86inc";
#
# Specialty Subpixel
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment