Commit e9c19617 authored by Yaowu Xu's avatar Yaowu Xu Committed by Gerrit Code Review
Browse files

Merge "vp9_short_fdct32x32_rd vp9_short_fdct32x32 optimized for AVX2"

parents d3a2e55a 8def766d
...@@ -260,4 +260,14 @@ INSTANTIATE_TEST_CASE_P( ...@@ -260,4 +260,14 @@ INSTANTIATE_TEST_CASE_P(
make_tuple(&vp9_fdct32x32_rd_sse2, make_tuple(&vp9_fdct32x32_rd_sse2,
&vp9_idct32x32_1024_add_sse2, 1))); &vp9_idct32x32_1024_add_sse2, 1)));
#endif #endif
#if HAVE_AVX2
INSTANTIATE_TEST_CASE_P(
AVX2, Trans32x32Test,
::testing::Values(
make_tuple(&vp9_fdct32x32_avx2,
&vp9_idct32x32_1024_add_sse2, 0),
make_tuple(&vp9_fdct32x32_rd_avx2,
&vp9_idct32x32_1024_add_sse2, 1)));
#endif
} // namespace } // namespace
...@@ -707,31 +707,31 @@ fi ...@@ -707,31 +707,31 @@ fi
# fdct functions # fdct functions
prototype void vp9_short_fht4x4 "const int16_t *input, int16_t *output, int stride, int tx_type" prototype void vp9_short_fht4x4 "const int16_t *input, int16_t *output, int stride, int tx_type"
specialize vp9_short_fht4x4 sse2 specialize vp9_short_fht4x4 sse2 avx2
prototype void vp9_short_fht8x8 "const int16_t *input, int16_t *output, int stride, int tx_type" prototype void vp9_short_fht8x8 "const int16_t *input, int16_t *output, int stride, int tx_type"
specialize vp9_short_fht8x8 sse2 specialize vp9_short_fht8x8 sse2 avx2
prototype void vp9_short_fht16x16 "const int16_t *input, int16_t *output, int stride, int tx_type" prototype void vp9_short_fht16x16 "const int16_t *input, int16_t *output, int stride, int tx_type"
specialize vp9_short_fht16x16 sse2 specialize vp9_short_fht16x16 sse2 avx2
prototype void vp9_fwht4x4 "const int16_t *input, int16_t *output, int stride" prototype void vp9_fwht4x4 "const int16_t *input, int16_t *output, int stride"
specialize vp9_fwht4x4 specialize vp9_fwht4x4
prototype void vp9_fdct4x4 "const int16_t *input, int16_t *output, int stride" prototype void vp9_fdct4x4 "const int16_t *input, int16_t *output, int stride"
specialize vp9_fdct4x4 sse2 specialize vp9_fdct4x4 sse2 avx2
prototype void vp9_fdct8x8 "const int16_t *input, int16_t *output, int stride" prototype void vp9_fdct8x8 "const int16_t *input, int16_t *output, int stride"
specialize vp9_fdct8x8 sse2 specialize vp9_fdct8x8 sse2 avx2
prototype void vp9_fdct16x16 "const int16_t *input, int16_t *output, int stride" prototype void vp9_fdct16x16 "const int16_t *input, int16_t *output, int stride"
specialize vp9_fdct16x16 sse2 specialize vp9_fdct16x16 sse2 avx2
prototype void vp9_fdct32x32 "const int16_t *input, int16_t *output, int stride" prototype void vp9_fdct32x32 "const int16_t *input, int16_t *output, int stride"
specialize vp9_fdct32x32 sse2 specialize vp9_fdct32x32 sse2 avx2
prototype void vp9_fdct32x32_rd "const int16_t *input, int16_t *output, int stride" prototype void vp9_fdct32x32_rd "const int16_t *input, int16_t *output, int stride"
specialize vp9_fdct32x32_rd sse2 specialize vp9_fdct32x32_rd sse2 avx2
# #
# Motion search # Motion search
......
This diff is collapsed.
This diff is collapsed.
...@@ -104,4 +104,7 @@ VP9_CX_SRCS-$(ARCH_X86_64) += encoder/x86/vp9_ssim_opt.asm ...@@ -104,4 +104,7 @@ VP9_CX_SRCS-$(ARCH_X86_64) += encoder/x86/vp9_ssim_opt.asm
VP9_CX_SRCS-$(HAVE_SSE2) += encoder/x86/vp9_dct_sse2.c VP9_CX_SRCS-$(HAVE_SSE2) += encoder/x86/vp9_dct_sse2.c
VP9_CX_SRCS-$(HAVE_SSE2) += encoder/x86/vp9_dct32x32_sse2.c VP9_CX_SRCS-$(HAVE_SSE2) += encoder/x86/vp9_dct32x32_sse2.c
VP9_CX_SRCS-$(HAVE_AVX2) += encoder/x86/vp9_dct_avx2.c
VP9_CX_SRCS-$(HAVE_AVX2) += encoder/x86/vp9_dct32x32_avx2.c
VP9_CX_SRCS-yes := $(filter-out $(VP9_CX_SRCS_REMOVE-yes),$(VP9_CX_SRCS-yes)) VP9_CX_SRCS-yes := $(filter-out $(VP9_CX_SRCS_REMOVE-yes),$(VP9_CX_SRCS-yes))
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