diff --git a/build/make/configure.sh b/build/make/configure.sh index aa66d9d42..2e1597779 100644 --- a/build/make/configure.sh +++ b/build/make/configure.sh @@ -1396,10 +1396,6 @@ EOF fi fi - if [ "${tgt_isa}" = "x86_64" ] || [ "${tgt_isa}" = "x86" ]; then - enable_feature use_x86inc - fi - # Position Independent Code (PIC) support, for building relocatable # shared objects enabled gcc && enabled pic && check_add_cflags -fPIC diff --git a/configure b/configure index dbbd81f19..c1ad24bfb 100755 --- a/configure +++ b/configure @@ -267,7 +267,6 @@ CONFIG_LIST=" install_bins install_libs install_srcs - use_x86inc debug gprof gcov @@ -329,7 +328,6 @@ CMDLINE_SELECT=" gprof gcov pic - use_x86inc optimizations ccache runtime_cpu_detect diff --git a/test/convolve_test.cc b/test/convolve_test.cc index 5883efbce..5933a62b0 100644 --- a/test/convolve_test.cc +++ b/test/convolve_test.cc @@ -990,14 +990,12 @@ void wrap_ ## func ## _ ## bd(const uint8_t *src, ptrdiff_t src_stride, \ w, h, bd); \ } #if HAVE_SSE2 && ARCH_X86_64 -#if CONFIG_USE_X86INC WRAP(convolve_copy_sse2, 8) WRAP(convolve_avg_sse2, 8) WRAP(convolve_copy_sse2, 10) WRAP(convolve_avg_sse2, 10) WRAP(convolve_copy_sse2, 12) WRAP(convolve_avg_sse2, 12) -#endif // CONFIG_USE_X86INC WRAP(convolve8_horiz_sse2, 8) WRAP(convolve8_avg_horiz_sse2, 8) WRAP(convolve8_vert_sse2, 8) @@ -1091,11 +1089,7 @@ INSTANTIATE_TEST_CASE_P(C, ConvolveTest, #if HAVE_SSE2 && ARCH_X86_64 #if CONFIG_VP9_HIGHBITDEPTH const ConvolveFunctions convolve8_sse2( -#if CONFIG_USE_X86INC wrap_convolve_copy_sse2_8, wrap_convolve_avg_sse2_8, -#else - wrap_convolve_copy_c_8, wrap_convolve_avg_c_8, -#endif // CONFIG_USE_X86INC wrap_convolve8_horiz_sse2_8, wrap_convolve8_avg_horiz_sse2_8, wrap_convolve8_vert_sse2_8, wrap_convolve8_avg_vert_sse2_8, wrap_convolve8_sse2_8, wrap_convolve8_avg_sse2_8, @@ -1103,11 +1097,7 @@ const ConvolveFunctions convolve8_sse2( wrap_convolve8_vert_sse2_8, wrap_convolve8_avg_vert_sse2_8, wrap_convolve8_sse2_8, wrap_convolve8_avg_sse2_8, 8); const ConvolveFunctions convolve10_sse2( -#if CONFIG_USE_X86INC wrap_convolve_copy_sse2_10, wrap_convolve_avg_sse2_10, -#else - wrap_convolve_copy_c_10, wrap_convolve_avg_c_10, -#endif // CONFIG_USE_X86INC wrap_convolve8_horiz_sse2_10, wrap_convolve8_avg_horiz_sse2_10, wrap_convolve8_vert_sse2_10, wrap_convolve8_avg_vert_sse2_10, wrap_convolve8_sse2_10, wrap_convolve8_avg_sse2_10, @@ -1115,11 +1105,7 @@ const ConvolveFunctions convolve10_sse2( wrap_convolve8_vert_sse2_10, wrap_convolve8_avg_vert_sse2_10, wrap_convolve8_sse2_10, wrap_convolve8_avg_sse2_10, 10); const ConvolveFunctions convolve12_sse2( -#if CONFIG_USE_X86INC wrap_convolve_copy_sse2_12, wrap_convolve_avg_sse2_12, -#else - wrap_convolve_copy_c_12, wrap_convolve_avg_c_12, -#endif // CONFIG_USE_X86INC wrap_convolve8_horiz_sse2_12, wrap_convolve8_avg_horiz_sse2_12, wrap_convolve8_vert_sse2_12, wrap_convolve8_avg_vert_sse2_12, wrap_convolve8_sse2_12, wrap_convolve8_avg_sse2_12, @@ -1133,11 +1119,7 @@ const ConvolveParam kArrayConvolve_sse2[] = { }; #else const ConvolveFunctions convolve8_sse2( -#if CONFIG_USE_X86INC vpx_convolve_copy_sse2, vpx_convolve_avg_sse2, -#else - vpx_convolve_copy_c, vpx_convolve_avg_c, -#endif // CONFIG_USE_X86INC vpx_convolve8_horiz_sse2, vpx_convolve8_avg_horiz_sse2, vpx_convolve8_vert_sse2, vpx_convolve8_avg_vert_sse2, vpx_convolve8_sse2, vpx_convolve8_avg_sse2, diff --git a/test/fdct4x4_test.cc b/test/fdct4x4_test.cc index 735cccf8d..a24085606 100644 --- a/test/fdct4x4_test.cc +++ b/test/fdct4x4_test.cc @@ -487,7 +487,7 @@ INSTANTIATE_TEST_CASE_P( make_tuple(&vp9_fht4x4_c, &vp9_iht4x4_16_add_neon, 3, VPX_BITS_8))); #endif // HAVE_NEON && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE -#if CONFIG_USE_X86INC && HAVE_SSE2 && !CONFIG_EMULATE_HARDWARE +#if HAVE_SSE2 && !CONFIG_EMULATE_HARDWARE INSTANTIATE_TEST_CASE_P( SSE2, Trans4x4WHT, ::testing::Values( diff --git a/test/fdct8x8_test.cc b/test/fdct8x8_test.cc index 29f215817..083ee6628 100644 --- a/test/fdct8x8_test.cc +++ b/test/fdct8x8_test.cc @@ -766,7 +766,7 @@ INSTANTIATE_TEST_CASE_P( &idct8x8_64_add_12_sse2, 6225, VPX_BITS_12))); #endif // HAVE_SSE2 && CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE -#if HAVE_SSSE3 && CONFIG_USE_X86INC && ARCH_X86_64 && \ +#if HAVE_SSSE3 && ARCH_X86_64 && \ !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE INSTANTIATE_TEST_CASE_P( SSSE3, FwdTrans8x8DCT, diff --git a/test/hadamard_test.cc b/test/hadamard_test.cc index 7a5bd5b4c..b8eec523f 100644 --- a/test/hadamard_test.cc +++ b/test/hadamard_test.cc @@ -152,10 +152,10 @@ INSTANTIATE_TEST_CASE_P(SSE2, Hadamard8x8Test, ::testing::Values(&vpx_hadamard_8x8_sse2)); #endif // HAVE_SSE2 -#if HAVE_SSSE3 && CONFIG_USE_X86INC && ARCH_X86_64 +#if HAVE_SSSE3 && ARCH_X86_64 INSTANTIATE_TEST_CASE_P(SSSE3, Hadamard8x8Test, ::testing::Values(&vpx_hadamard_8x8_ssse3)); -#endif // HAVE_SSSE3 && CONFIG_USE_X86INC && ARCH_X86_64 +#endif // HAVE_SSSE3 && ARCH_X86_64 #if HAVE_NEON INSTANTIATE_TEST_CASE_P(NEON, Hadamard8x8Test, diff --git a/test/partial_idct_test.cc b/test/partial_idct_test.cc index 6c824128b..1efb1a4eb 100644 --- a/test/partial_idct_test.cc +++ b/test/partial_idct_test.cc @@ -295,7 +295,7 @@ INSTANTIATE_TEST_CASE_P( TX_4X4, 1))); #endif -#if HAVE_SSSE3 && CONFIG_USE_X86INC && ARCH_X86_64 && \ +#if HAVE_SSSE3 && ARCH_X86_64 && \ !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE INSTANTIATE_TEST_CASE_P( SSSE3_64, PartialIDctTest, diff --git a/test/sad_test.cc b/test/sad_test.cc index e6bd0d793..fa7b6c840 100644 --- a/test/sad_test.cc +++ b/test/sad_test.cc @@ -690,7 +690,6 @@ INSTANTIATE_TEST_CASE_P(NEON, SADx4Test, ::testing::ValuesIn(x4d_neon_tests)); //------------------------------------------------------------------------------ // x86 functions #if HAVE_SSE2 -#if CONFIG_USE_X86INC const SadMxNParam sse2_tests[] = { make_tuple(64, 64, &vpx_sad64x64_sse2, -1), make_tuple(64, 32, &vpx_sad64x32_sse2, -1), @@ -852,7 +851,6 @@ const SadMxNx4Param x4d_sse2_tests[] = { #endif // CONFIG_VP9_HIGHBITDEPTH }; INSTANTIATE_TEST_CASE_P(SSE2, SADx4Test, ::testing::ValuesIn(x4d_sse2_tests)); -#endif // CONFIG_USE_X86INC #endif // HAVE_SSE2 #if HAVE_SSE3 diff --git a/test/test_intra_pred_speed.cc b/test/test_intra_pred_speed.cc index 2acf744d5..8928bf87c 100644 --- a/test/test_intra_pred_speed.cc +++ b/test/test_intra_pred_speed.cc @@ -187,21 +187,21 @@ INTRA_PRED_TEST(C, TestIntraPred4, vpx_dc_predictor_4x4_c, vpx_d153_predictor_4x4_c, vpx_d207_predictor_4x4_c, vpx_d63_predictor_4x4_c, vpx_tm_predictor_4x4_c) -#if HAVE_SSE2 && CONFIG_USE_X86INC +#if HAVE_SSE2 INTRA_PRED_TEST(SSE2, TestIntraPred4, vpx_dc_predictor_4x4_sse2, vpx_dc_left_predictor_4x4_sse2, vpx_dc_top_predictor_4x4_sse2, vpx_dc_128_predictor_4x4_sse2, vpx_v_predictor_4x4_sse2, vpx_h_predictor_4x4_sse2, vpx_d45_predictor_4x4_sse2, NULL, NULL, NULL, vpx_d207_predictor_4x4_sse2, NULL, vpx_tm_predictor_4x4_sse2) -#endif // HAVE_SSE2 && CONFIG_USE_X86INC +#endif // HAVE_SSE2 -#if HAVE_SSSE3 && CONFIG_USE_X86INC +#if HAVE_SSSE3 INTRA_PRED_TEST(SSSE3, TestIntraPred4, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, vpx_d153_predictor_4x4_ssse3, NULL, vpx_d63_predictor_4x4_ssse3, NULL) -#endif // HAVE_SSSE3 && CONFIG_USE_X86INC +#endif // HAVE_SSSE3 #if HAVE_DSPR2 INTRA_PRED_TEST(DSPR2, TestIntraPred4, vpx_dc_predictor_4x4_dspr2, NULL, NULL, @@ -237,20 +237,20 @@ INTRA_PRED_TEST(C, TestIntraPred8, vpx_dc_predictor_8x8_c, vpx_d153_predictor_8x8_c, vpx_d207_predictor_8x8_c, vpx_d63_predictor_8x8_c, vpx_tm_predictor_8x8_c) -#if HAVE_SSE2 && CONFIG_USE_X86INC +#if HAVE_SSE2 INTRA_PRED_TEST(SSE2, TestIntraPred8, vpx_dc_predictor_8x8_sse2, vpx_dc_left_predictor_8x8_sse2, vpx_dc_top_predictor_8x8_sse2, vpx_dc_128_predictor_8x8_sse2, vpx_v_predictor_8x8_sse2, vpx_h_predictor_8x8_sse2, vpx_d45_predictor_8x8_sse2, NULL, NULL, NULL, NULL, NULL, vpx_tm_predictor_8x8_sse2) -#endif // HAVE_SSE2 && CONFIG_USE_X86INC +#endif // HAVE_SSE2 -#if HAVE_SSSE3 && CONFIG_USE_X86INC +#if HAVE_SSSE3 INTRA_PRED_TEST(SSSE3, TestIntraPred8, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, vpx_d153_predictor_8x8_ssse3, vpx_d207_predictor_8x8_ssse3, vpx_d63_predictor_8x8_ssse3, NULL) -#endif // HAVE_SSSE3 && CONFIG_USE_X86INC +#endif // HAVE_SSSE3 #if HAVE_DSPR2 INTRA_PRED_TEST(DSPR2, TestIntraPred8, vpx_dc_predictor_8x8_dspr2, NULL, NULL, @@ -286,22 +286,22 @@ INTRA_PRED_TEST(C, TestIntraPred16, vpx_dc_predictor_16x16_c, vpx_d153_predictor_16x16_c, vpx_d207_predictor_16x16_c, vpx_d63_predictor_16x16_c, vpx_tm_predictor_16x16_c) -#if HAVE_SSE2 && CONFIG_USE_X86INC +#if HAVE_SSE2 INTRA_PRED_TEST(SSE2, TestIntraPred16, vpx_dc_predictor_16x16_sse2, vpx_dc_left_predictor_16x16_sse2, vpx_dc_top_predictor_16x16_sse2, vpx_dc_128_predictor_16x16_sse2, vpx_v_predictor_16x16_sse2, vpx_h_predictor_16x16_sse2, NULL, NULL, NULL, NULL, NULL, NULL, vpx_tm_predictor_16x16_sse2) -#endif // HAVE_SSE2 && CONFIG_USE_X86INC +#endif // HAVE_SSE2 -#if HAVE_SSSE3 && CONFIG_USE_X86INC +#if HAVE_SSSE3 INTRA_PRED_TEST(SSSE3, TestIntraPred16, NULL, NULL, NULL, NULL, NULL, NULL, vpx_d45_predictor_16x16_ssse3, NULL, NULL, vpx_d153_predictor_16x16_ssse3, vpx_d207_predictor_16x16_ssse3, vpx_d63_predictor_16x16_ssse3, NULL) -#endif // HAVE_SSSE3 && CONFIG_USE_X86INC +#endif // HAVE_SSSE3 #if HAVE_DSPR2 INTRA_PRED_TEST(DSPR2, TestIntraPred16, vpx_dc_predictor_16x16_dspr2, NULL, @@ -337,21 +337,21 @@ INTRA_PRED_TEST(C, TestIntraPred32, vpx_dc_predictor_32x32_c, vpx_d153_predictor_32x32_c, vpx_d207_predictor_32x32_c, vpx_d63_predictor_32x32_c, vpx_tm_predictor_32x32_c) -#if HAVE_SSE2 && CONFIG_USE_X86INC +#if HAVE_SSE2 INTRA_PRED_TEST(SSE2, TestIntraPred32, vpx_dc_predictor_32x32_sse2, vpx_dc_left_predictor_32x32_sse2, vpx_dc_top_predictor_32x32_sse2, vpx_dc_128_predictor_32x32_sse2, vpx_v_predictor_32x32_sse2, vpx_h_predictor_32x32_sse2, NULL, NULL, NULL, NULL, NULL, NULL, vpx_tm_predictor_32x32_sse2) -#endif // HAVE_SSE2 && CONFIG_USE_X86INC +#endif // HAVE_SSE2 -#if HAVE_SSSE3 && CONFIG_USE_X86INC +#if HAVE_SSSE3 INTRA_PRED_TEST(SSSE3, TestIntraPred32, NULL, NULL, NULL, NULL, NULL, NULL, vpx_d45_predictor_32x32_ssse3, NULL, NULL, vpx_d153_predictor_32x32_ssse3, vpx_d207_predictor_32x32_ssse3, vpx_d63_predictor_32x32_ssse3, NULL) -#endif // HAVE_SSSE3 && CONFIG_USE_X86INC +#endif // HAVE_SSSE3 #if HAVE_NEON INTRA_PRED_TEST(NEON, TestIntraPred32, vpx_dc_predictor_32x32_neon, diff --git a/test/variance_test.cc b/test/variance_test.cc index cb6339041..08c84a613 100644 --- a/test/variance_test.cc +++ b/test/variance_test.cc @@ -1002,7 +1002,6 @@ INSTANTIATE_TEST_CASE_P( make_tuple(2, 3, &vpx_variance4x8_sse2, 0), make_tuple(2, 2, &vpx_variance4x4_sse2, 0))); -#if CONFIG_USE_X86INC INSTANTIATE_TEST_CASE_P( SSE2, VpxSubpelVarianceTest, ::testing::Values(make_tuple(6, 6, &vpx_sub_pixel_variance64x64_sse2, 0), @@ -1035,7 +1034,6 @@ INSTANTIATE_TEST_CASE_P( make_tuple(3, 2, &vpx_sub_pixel_avg_variance8x4_sse2, 0), make_tuple(2, 3, &vpx_sub_pixel_avg_variance4x8_sse2, 0), make_tuple(2, 2, &vpx_sub_pixel_avg_variance4x4_sse2, 0))); -#endif // CONFIG_USE_X86INC #if CONFIG_VP9_HIGHBITDEPTH /* TODO(debargha): This test does not support the highbd version @@ -1088,7 +1086,6 @@ INSTANTIATE_TEST_CASE_P( make_tuple(3, 4, &vpx_highbd_8_variance8x16_sse2, 8), make_tuple(3, 3, &vpx_highbd_8_variance8x8_sse2, 8))); -#if CONFIG_USE_X86INC INSTANTIATE_TEST_CASE_P( SSE2, VpxHBDSubpelVarianceTest, ::testing::Values( @@ -1162,12 +1159,10 @@ INSTANTIATE_TEST_CASE_P( make_tuple(3, 4, &vpx_highbd_8_sub_pixel_avg_variance8x16_sse2, 8), make_tuple(3, 3, &vpx_highbd_8_sub_pixel_avg_variance8x8_sse2, 8), make_tuple(3, 2, &vpx_highbd_8_sub_pixel_avg_variance8x4_sse2, 8))); -#endif // CONFIG_USE_X86INC #endif // CONFIG_VP9_HIGHBITDEPTH #endif // HAVE_SSE2 #if HAVE_SSSE3 -#if CONFIG_USE_X86INC INSTANTIATE_TEST_CASE_P( SSSE3, VpxSubpelVarianceTest, ::testing::Values(make_tuple(6, 6, &vpx_sub_pixel_variance64x64_ssse3, 0), @@ -1200,7 +1195,6 @@ INSTANTIATE_TEST_CASE_P( make_tuple(3, 2, &vpx_sub_pixel_avg_variance8x4_ssse3, 0), make_tuple(2, 3, &vpx_sub_pixel_avg_variance4x8_ssse3, 0), make_tuple(2, 2, &vpx_sub_pixel_avg_variance4x4_ssse3, 0))); -#endif // CONFIG_USE_X86INC #endif // HAVE_SSSE3 #if HAVE_AVX2 diff --git a/test/vp9_error_block_test.cc b/test/vp9_error_block_test.cc index 23a249e2b..341cc19cb 100644 --- a/test/vp9_error_block_test.cc +++ b/test/vp9_error_block_test.cc @@ -157,9 +157,9 @@ TEST_P(ErrorBlockTest, ExtremeValues) { << "First failed at test case " << first_failure; } +#if HAVE_SSE2 || HAVE_AVX using std::tr1::make_tuple; -#if CONFIG_USE_X86INC int64_t wrap_vp9_highbd_block_error_8bit_c(const tran_low_t *coeff, const tran_low_t *dqcoeff, intptr_t block_size, @@ -167,6 +167,7 @@ int64_t wrap_vp9_highbd_block_error_8bit_c(const tran_low_t *coeff, EXPECT_EQ(8, bps); return vp9_highbd_block_error_8bit_c(coeff, dqcoeff, block_size, ssz); } +#endif // HAVE_SSE2 || HAVE_AVX #if HAVE_SSE2 int64_t wrap_vp9_highbd_block_error_8bit_sse2(const tran_low_t *coeff, @@ -206,6 +207,5 @@ INSTANTIATE_TEST_CASE_P( &wrap_vp9_highbd_block_error_8bit_c, VPX_BITS_8))); #endif // HAVE_AVX -#endif // CONFIG_USE_X86INC #endif // CONFIG_VP9_HIGHBITDEPTH } // namespace diff --git a/test/vp9_intrapred_test.cc b/test/vp9_intrapred_test.cc index 416f3c322..ea8abfb77 100644 --- a/test/vp9_intrapred_test.cc +++ b/test/vp9_intrapred_test.cc @@ -131,7 +131,6 @@ using std::tr1::make_tuple; #if HAVE_SSE2 #if CONFIG_VP9_HIGHBITDEPTH -#if CONFIG_USE_X86INC INSTANTIATE_TEST_CASE_P(SSE2_TO_C_8, VP9IntraPredTest, ::testing::Values( make_tuple(&vpx_highbd_dc_predictor_32x32_sse2, @@ -225,7 +224,6 @@ INSTANTIATE_TEST_CASE_P(SSE2_TO_C_12, VP9IntraPredTest, make_tuple(&vpx_highbd_tm_predictor_8x8_sse2, &vpx_highbd_tm_predictor_8x8_c, 8, 12))); -#endif // CONFIG_USE_X86INC #endif // CONFIG_VP9_HIGHBITDEPTH #endif // HAVE_SSE2 } // namespace diff --git a/test/vp9_subtract_test.cc b/test/vp9_subtract_test.cc index 3cad4d7e6..4793a9716 100644 --- a/test/vp9_subtract_test.cc +++ b/test/vp9_subtract_test.cc @@ -93,7 +93,7 @@ TEST_P(VP9SubtractBlockTest, SimpleSubtract) { INSTANTIATE_TEST_CASE_P(C, VP9SubtractBlockTest, ::testing::Values(vpx_subtract_block_c)); -#if HAVE_SSE2 && CONFIG_USE_X86INC +#if HAVE_SSE2 INSTANTIATE_TEST_CASE_P(SSE2, VP9SubtractBlockTest, ::testing::Values(vpx_subtract_block_sse2)); #endif diff --git a/vpx_dsp/vpx_dsp.mk b/vpx_dsp/vpx_dsp.mk index 018126d4b..93855c5ad 100644 --- a/vpx_dsp/vpx_dsp.mk +++ b/vpx_dsp/vpx_dsp.mk @@ -40,18 +40,14 @@ endif # intra predictions DSP_SRCS-yes += intrapred.c -ifeq ($(CONFIG_USE_X86INC),yes) DSP_SRCS-$(HAVE_SSE) += x86/intrapred_sse2.asm DSP_SRCS-$(HAVE_SSE2) += x86/intrapred_sse2.asm DSP_SRCS-$(HAVE_SSSE3) += x86/intrapred_ssse3.asm DSP_SRCS-$(HAVE_SSSE3) += x86/vpx_subpixel_8t_ssse3.asm -endif # CONFIG_USE_X86INC ifeq ($(CONFIG_VP9_HIGHBITDEPTH),yes) -ifeq ($(CONFIG_USE_X86INC),yes) DSP_SRCS-$(HAVE_SSE) += x86/highbd_intrapred_sse2.asm DSP_SRCS-$(HAVE_SSE2) += x86/highbd_intrapred_sse2.asm -endif # CONFIG_USE_X86INC endif # CONFIG_VP9_HIGHBITDEPTH ifneq ($(filter yes,$(CONFIG_POSTPROC) $(CONFIG_VP9_POSTPROC)),) @@ -87,9 +83,8 @@ ifeq ($(CONFIG_VP9_HIGHBITDEPTH),yes) DSP_SRCS-$(HAVE_SSE2) += x86/vpx_high_subpixel_8t_sse2.asm DSP_SRCS-$(HAVE_SSE2) += x86/vpx_high_subpixel_bilinear_sse2.asm endif -ifeq ($(CONFIG_USE_X86INC),yes) + DSP_SRCS-$(HAVE_SSE2) += x86/vpx_convolve_copy_sse2.asm -endif ifeq ($(HAVE_NEON_ASM),yes) DSP_SRCS-yes += arm/vpx_convolve_copy_neon_asm$(ASM) @@ -179,10 +174,8 @@ DSP_SRCS-$(HAVE_SSE2) += x86/fwd_txfm_sse2.c DSP_SRCS-$(HAVE_SSE2) += x86/fwd_txfm_impl_sse2.h DSP_SRCS-$(HAVE_SSE2) += x86/fwd_dct32x32_impl_sse2.h ifeq ($(ARCH_X86_64),yes) -ifeq ($(CONFIG_USE_X86INC),yes) DSP_SRCS-$(HAVE_SSSE3) += x86/fwd_txfm_ssse3_x86_64.asm endif -endif DSP_SRCS-$(HAVE_AVX2) += x86/fwd_txfm_avx2.c DSP_SRCS-$(HAVE_AVX2) += x86/fwd_dct32x32_impl_avx2.h DSP_SRCS-$(HAVE_NEON) += arm/fwd_txfm_neon.c @@ -197,12 +190,10 @@ DSP_SRCS-yes += inv_txfm.h DSP_SRCS-yes += inv_txfm.c DSP_SRCS-$(HAVE_SSE2) += x86/inv_txfm_sse2.h DSP_SRCS-$(HAVE_SSE2) += x86/inv_txfm_sse2.c -ifeq ($(CONFIG_USE_X86INC),yes) DSP_SRCS-$(HAVE_SSE2) += x86/inv_wht_sse2.asm ifeq ($(ARCH_X86_64),yes) DSP_SRCS-$(HAVE_SSSE3) += x86/inv_txfm_ssse3_x86_64.asm endif # ARCH_X86_64 -endif # CONFIG_USE_X86INC ifeq ($(HAVE_NEON_ASM),yes) DSP_SRCS-yes += arm/save_reg_neon$(ASM) @@ -254,11 +245,9 @@ ifeq ($(CONFIG_VP9_HIGHBITDEPTH),yes) DSP_SRCS-$(HAVE_SSE2) += x86/highbd_quantize_intrin_sse2.c endif ifeq ($(ARCH_X86_64),yes) -ifeq ($(CONFIG_USE_X86INC),yes) DSP_SRCS-$(HAVE_SSSE3) += x86/quantize_ssse3_x86_64.asm DSP_SRCS-$(HAVE_AVX) += x86/quantize_avx_x86_64.asm endif -endif # avg DSP_SRCS-yes += avg.c @@ -267,10 +256,8 @@ DSP_SRCS-$(HAVE_NEON) += arm/avg_neon.c DSP_SRCS-$(HAVE_MSA) += mips/avg_msa.c DSP_SRCS-$(HAVE_NEON) += arm/hadamard_neon.c ifeq ($(ARCH_X86_64),yes) -ifeq ($(CONFIG_USE_X86INC),yes) DSP_SRCS-$(HAVE_SSSE3) += x86/avg_ssse3_x86_64.asm endif -endif endif # CONFIG_VP9_ENCODER @@ -292,7 +279,6 @@ DSP_SRCS-$(HAVE_SSE4_1) += x86/sad_sse4.asm DSP_SRCS-$(HAVE_AVX2) += x86/sad4d_avx2.c DSP_SRCS-$(HAVE_AVX2) += x86/sad_avx2.c -ifeq ($(CONFIG_USE_X86INC),yes) DSP_SRCS-$(HAVE_SSE) += x86/sad4d_sse2.asm DSP_SRCS-$(HAVE_SSE) += x86/sad_sse2.asm DSP_SRCS-$(HAVE_SSE2) += x86/sad4d_sse2.asm @@ -303,7 +289,6 @@ ifeq ($(CONFIG_VP9_HIGHBITDEPTH),yes) DSP_SRCS-$(HAVE_SSE2) += x86/highbd_sad4d_sse2.asm DSP_SRCS-$(HAVE_SSE2) += x86/highbd_sad_sse2.asm endif # CONFIG_VP9_HIGHBITDEPTH -endif # CONFIG_USE_X86INC endif # CONFIG_ENCODERS @@ -334,17 +319,13 @@ ifeq ($(ARCH_X86_64),yes) DSP_SRCS-$(HAVE_SSE2) += x86/ssim_opt_x86_64.asm endif # ARCH_X86_64 -ifeq ($(CONFIG_USE_X86INC),yes) DSP_SRCS-$(HAVE_SSE) += x86/subpel_variance_sse2.asm DSP_SRCS-$(HAVE_SSE2) += x86/subpel_variance_sse2.asm # Contains SSE2 and SSSE3 -endif # CONFIG_USE_X86INC ifeq ($(CONFIG_VP9_HIGHBITDEPTH),yes) DSP_SRCS-$(HAVE_SSE2) += x86/highbd_variance_sse2.c DSP_SRCS-$(HAVE_SSE2) += x86/highbd_variance_impl_sse2.asm -ifeq ($(CONFIG_USE_X86INC),yes) DSP_SRCS-$(HAVE_SSE2) += x86/highbd_subpel_variance_impl_sse2.asm -endif # CONFIG_USE_X86INC endif # CONFIG_VP9_HIGHBITDEPTH endif # CONFIG_ENCODERS || CONFIG_POSTPROC || CONFIG_VP9_POSTPROC diff --git a/vpx_dsp/vpx_dsp_rtcd_defs.pl b/vpx_dsp/vpx_dsp_rtcd_defs.pl index 37239a195..7b61415b6 100644 --- a/vpx_dsp/vpx_dsp_rtcd_defs.pl +++ b/vpx_dsp/vpx_dsp_rtcd_defs.pl @@ -11,29 +11,6 @@ EOF } forward_decls qw/vpx_dsp_forward_decls/; -# x86inc.asm had specific constraints. break it out so it's easy to disable. -# zero all the variables to avoid tricky else conditions. -$mmx_x86inc = $sse_x86inc = $sse2_x86inc = $ssse3_x86inc = $avx_x86inc = - $avx2_x86inc = ''; -$mmx_x86_64_x86inc = $sse_x86_64_x86inc = $sse2_x86_64_x86inc = - $ssse3_x86_64_x86inc = $avx_x86_64_x86inc = $avx2_x86_64_x86inc = ''; -if (vpx_config("CONFIG_USE_X86INC") eq "yes") { - $mmx_x86inc = 'mmx'; - $sse_x86inc = 'sse'; - $sse2_x86inc = 'sse2'; - $ssse3_x86inc = 'ssse3'; - $avx_x86inc = 'avx'; - $avx2_x86inc = 'avx2'; - if ($opts{arch} eq "x86_64") { - $mmx_x86_64_x86inc = 'mmx'; - $sse_x86_64_x86inc = 'sse'; - $sse2_x86_64_x86inc = 'sse2'; - $ssse3_x86_64_x86inc = 'ssse3'; - $avx_x86_64_x86inc = 'avx'; - $avx2_x86_64_x86inc = 'avx2'; - } -} - # optimizations which depend on multiple features $avx2_ssse3 = ''; if ((vpx_config("HAVE_AVX2") eq "yes") && (vpx_config("HAVE_SSSE3") eq "yes")) { @@ -55,19 +32,19 @@ if ($opts{arch} eq "x86_64") { # add_proto qw/void vpx_d207_predictor_4x4/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_d207_predictor_4x4/, "$sse2_x86inc"; +specialize qw/vpx_d207_predictor_4x4 sse2/; add_proto qw/void vpx_d207e_predictor_4x4/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; specialize qw/vpx_d207e_predictor_4x4/; add_proto qw/void vpx_d45_predictor_4x4/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_d45_predictor_4x4 neon/, "$sse2_x86inc"; +specialize qw/vpx_d45_predictor_4x4 neon sse2/; add_proto qw/void vpx_d45e_predictor_4x4/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; specialize qw/vpx_d45e_predictor_4x4/; add_proto qw/void vpx_d63_predictor_4x4/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_d63_predictor_4x4/, "$ssse3_x86inc"; +specialize qw/vpx_d63_predictor_4x4 ssse3/; add_proto qw/void vpx_d63e_predictor_4x4/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; specialize qw/vpx_d63e_predictor_4x4/; @@ -76,7 +53,7 @@ add_proto qw/void vpx_d63f_predictor_4x4/, "uint8_t *dst, ptrdiff_t y_stride, co specialize qw/vpx_d63f_predictor_4x4/; add_proto qw/void vpx_h_predictor_4x4/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_h_predictor_4x4 neon dspr2 msa/, "$sse2_x86inc"; +specialize qw/vpx_h_predictor_4x4 neon dspr2 msa sse2/; add_proto qw/void vpx_he_predictor_4x4/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; specialize qw/vpx_he_predictor_4x4/; @@ -88,49 +65,49 @@ add_proto qw/void vpx_d135_predictor_4x4/, "uint8_t *dst, ptrdiff_t y_stride, co specialize qw/vpx_d135_predictor_4x4 neon/; add_proto qw/void vpx_d153_predictor_4x4/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_d153_predictor_4x4/, "$ssse3_x86inc"; +specialize qw/vpx_d153_predictor_4x4 ssse3/; add_proto qw/void vpx_v_predictor_4x4/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_v_predictor_4x4 neon msa/, "$sse2_x86inc"; +specialize qw/vpx_v_predictor_4x4 neon msa sse2/; add_proto qw/void vpx_ve_predictor_4x4/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; specialize qw/vpx_ve_predictor_4x4/; add_proto qw/void vpx_tm_predictor_4x4/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_tm_predictor_4x4 neon dspr2 msa/, "$sse2_x86inc"; +specialize qw/vpx_tm_predictor_4x4 neon dspr2 msa sse2/; add_proto qw/void vpx_dc_predictor_4x4/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_dc_predictor_4x4 dspr2 msa neon/, "$sse2_x86inc"; +specialize qw/vpx_dc_predictor_4x4 dspr2 msa neon sse2/; add_proto qw/void vpx_dc_top_predictor_4x4/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_dc_top_predictor_4x4 msa neon/, "$sse2_x86inc"; +specialize qw/vpx_dc_top_predictor_4x4 msa neon sse2/; add_proto qw/void vpx_dc_left_predictor_4x4/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_dc_left_predictor_4x4 msa neon/, "$sse2_x86inc"; +specialize qw/vpx_dc_left_predictor_4x4 msa neon sse2/; add_proto qw/void vpx_dc_128_predictor_4x4/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_dc_128_predictor_4x4 msa neon/, "$sse2_x86inc"; +specialize qw/vpx_dc_128_predictor_4x4 msa neon sse2/; add_proto qw/void vpx_d207_predictor_8x8/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_d207_predictor_8x8/, "$ssse3_x86inc"; +specialize qw/vpx_d207_predictor_8x8 ssse3/; add_proto qw/void vpx_d207e_predictor_8x8/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; specialize qw/vpx_d207e_predictor_8x8/; add_proto qw/void vpx_d45_predictor_8x8/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_d45_predictor_8x8 neon/, "$sse2_x86inc"; +specialize qw/vpx_d45_predictor_8x8 neon sse2/; add_proto qw/void vpx_d45e_predictor_8x8/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; specialize qw/vpx_d45e_predictor_8x8/; add_proto qw/void vpx_d63_predictor_8x8/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_d63_predictor_8x8/, "$ssse3_x86inc"; +specialize qw/vpx_d63_predictor_8x8 ssse3/; add_proto qw/void vpx_d63e_predictor_8x8/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; specialize qw/vpx_d63e_predictor_8x8/; add_proto qw/void vpx_h_predictor_8x8/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_h_predictor_8x8 neon dspr2 msa/, "$sse2_x86inc"; +specialize qw/vpx_h_predictor_8x8 neon dspr2 msa sse2/; add_proto qw/void vpx_d117_predictor_8x8/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; specialize qw/vpx_d117_predictor_8x8/; @@ -139,46 +116,46 @@ add_proto qw/void vpx_d135_predictor_8x8/, "uint8_t *dst, ptrdiff_t y_stride, co specialize qw/vpx_d135_predictor_8x8/; add_proto qw/void vpx_d153_predictor_8x8/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_d153_predictor_8x8/, "$ssse3_x86inc"; +specialize qw/vpx_d153_predictor_8x8 ssse3/; add_proto qw/void vpx_v_predictor_8x8/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_v_predictor_8x8 neon msa/, "$sse2_x86inc"; +specialize qw/vpx_v_predictor_8x8 neon msa sse2/; add_proto qw/void vpx_tm_predictor_8x8/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_tm_predictor_8x8 neon dspr2 msa/, "$sse2_x86inc"; +specialize qw/vpx_tm_predictor_8x8 neon dspr2 msa sse2/; add_proto qw/void vpx_dc_predictor_8x8/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_dc_predictor_8x8 dspr2 neon msa/, "$sse2_x86inc"; +specialize qw/vpx_dc_predictor_8x8 dspr2 neon msa sse2/; add_proto qw/void vpx_dc_top_predictor_8x8/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_dc_top_predictor_8x8 neon msa/, "$sse2_x86inc"; +specialize qw/vpx_dc_top_predictor_8x8 neon msa sse2/; add_proto qw/void vpx_dc_left_predictor_8x8/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_dc_left_predictor_8x8 neon msa/, "$sse2_x86inc"; +specialize qw/vpx_dc_left_predictor_8x8 neon msa sse2/; add_proto qw/void vpx_dc_128_predictor_8x8/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_dc_128_predictor_8x8 neon msa/, "$sse2_x86inc"; +specialize qw/vpx_dc_128_predictor_8x8 neon msa sse2/; add_proto qw/void vpx_d207_predictor_16x16/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_d207_predictor_16x16/, "$ssse3_x86inc"; +specialize qw/vpx_d207_predictor_16x16 ssse3/; add_proto qw/void vpx_d207e_predictor_16x16/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; specialize qw/vpx_d207e_predictor_16x16/; add_proto qw/void vpx_d45_predictor_16x16/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_d45_predictor_16x16 neon/, "$ssse3_x86inc"; +specialize qw/vpx_d45_predictor_16x16 neon ssse3/; add_proto qw/void vpx_d45e_predictor_16x16/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; specialize qw/vpx_d45e_predictor_16x16/; add_proto qw/void vpx_d63_predictor_16x16/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_d63_predictor_16x16/, "$ssse3_x86inc"; +specialize qw/vpx_d63_predictor_16x16 ssse3/; add_proto qw/void vpx_d63e_predictor_16x16/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; specialize qw/vpx_d63e_predictor_16x16/; add_proto qw/void vpx_h_predictor_16x16/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_h_predictor_16x16 neon dspr2 msa/, "$sse2_x86inc"; +specialize qw/vpx_h_predictor_16x16 neon dspr2 msa sse2/; add_proto qw/void vpx_d117_predictor_16x16/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; specialize qw/vpx_d117_predictor_16x16/; @@ -187,46 +164,46 @@ add_proto qw/void vpx_d135_predictor_16x16/, "uint8_t *dst, ptrdiff_t y_stride, specialize qw/vpx_d135_predictor_16x16/; add_proto qw/void vpx_d153_predictor_16x16/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_d153_predictor_16x16/, "$ssse3_x86inc"; +specialize qw/vpx_d153_predictor_16x16 ssse3/; add_proto qw/void vpx_v_predictor_16x16/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_v_predictor_16x16 neon msa/, "$sse2_x86inc"; +specialize qw/vpx_v_predictor_16x16 neon msa sse2/; add_proto qw/void vpx_tm_predictor_16x16/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_tm_predictor_16x16 neon msa/, "$sse2_x86inc"; +specialize qw/vpx_tm_predictor_16x16 neon msa sse2/; add_proto qw/void vpx_dc_predictor_16x16/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_dc_predictor_16x16 dspr2 neon msa/, "$sse2_x86inc"; +specialize qw/vpx_dc_predictor_16x16 dspr2 neon msa sse2/; add_proto qw/void vpx_dc_top_predictor_16x16/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_dc_top_predictor_16x16 neon msa/, "$sse2_x86inc"; +specialize qw/vpx_dc_top_predictor_16x16 neon msa sse2/; add_proto qw/void vpx_dc_left_predictor_16x16/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_dc_left_predictor_16x16 neon msa/, "$sse2_x86inc"; +specialize qw/vpx_dc_left_predictor_16x16 neon msa sse2/; add_proto qw/void vpx_dc_128_predictor_16x16/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_dc_128_predictor_16x16 neon msa/, "$sse2_x86inc"; +specialize qw/vpx_dc_128_predictor_16x16 neon msa sse2/; add_proto qw/void vpx_d207_predictor_32x32/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_d207_predictor_32x32/, "$ssse3_x86inc"; +specialize qw/vpx_d207_predictor_32x32 ssse3/; add_proto qw/void vpx_d207e_predictor_32x32/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; specialize qw/vpx_d207e_predictor_32x32/; add_proto qw/void vpx_d45_predictor_32x32/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_d45_predictor_32x32/, "$ssse3_x86inc"; +specialize qw/vpx_d45_predictor_32x32 ssse3/; add_proto qw/void vpx_d45e_predictor_32x32/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; specialize qw/vpx_d45e_predictor_32x32/; add_proto qw/void vpx_d63_predictor_32x32/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_d63_predictor_32x32/, "$ssse3_x86inc"; +specialize qw/vpx_d63_predictor_32x32 ssse3/; add_proto qw/void vpx_d63e_predictor_32x32/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; specialize qw/vpx_d63e_predictor_32x32/; add_proto qw/void vpx_h_predictor_32x32/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_h_predictor_32x32 neon msa/, "$sse2_x86inc"; +specialize qw/vpx_h_predictor_32x32 neon msa sse2/; add_proto qw/void vpx_d117_predictor_32x32/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; specialize qw/vpx_d117_predictor_32x32/; @@ -235,25 +212,25 @@ add_proto qw/void vpx_d135_predictor_32x32/, "uint8_t *dst, ptrdiff_t y_stride, specialize qw/vpx_d135_predictor_32x32/; add_proto qw/void vpx_d153_predictor_32x32/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_d153_predictor_32x32/, "$ssse3_x86inc"; +specialize qw/vpx_d153_predictor_32x32 ssse3/; add_proto qw/void vpx_v_predictor_32x32/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_v_predictor_32x32 neon msa/, "$sse2_x86inc"; +specialize qw/vpx_v_predictor_32x32 neon msa sse2/; add_proto qw/void vpx_tm_predictor_32x32/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_tm_predictor_32x32 neon msa/, "$sse2_x86inc"; +specialize qw/vpx_tm_predictor_32x32 neon msa sse2/; add_proto qw/void vpx_dc_predictor_32x32/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_dc_predictor_32x32 msa neon/, "$sse2_x86inc"; +specialize qw/vpx_dc_predictor_32x32 msa neon sse2/; add_proto qw/void vpx_dc_top_predictor_32x32/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_dc_top_predictor_32x32 msa neon/, "$sse2_x86inc"; +specialize qw/vpx_dc_top_predictor_32x32 msa neon sse2/; add_proto qw/void vpx_dc_left_predictor_32x32/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_dc_left_predictor_32x32 msa neon/, "$sse2_x86inc"; +specialize qw/vpx_dc_left_predictor_32x32 msa neon sse2/; add_proto qw/void vpx_dc_128_predictor_32x32/, "uint8_t *dst, ptrdiff_t y_stride, const uint8_t *above, const uint8_t *left"; -specialize qw/vpx_dc_128_predictor_32x32 msa neon/, "$sse2_x86inc"; +specialize qw/vpx_dc_128_predictor_32x32 msa neon sse2/; # High bitdepth functions if (vpx_config("CONFIG_VP9_HIGHBITDEPTH") eq "yes") { @@ -288,13 +265,13 @@ if (vpx_config("CONFIG_VP9_HIGHBITDEPTH") eq "yes") { specialize qw/vpx_highbd_d153_predictor_4x4/; add_proto qw/void vpx_highbd_v_predictor_4x4/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd"; - specialize qw/vpx_highbd_v_predictor_4x4/, "$sse2_x86inc"; + specialize qw/vpx_highbd_v_predictor_4x4 sse2/; add_proto qw/void vpx_highbd_tm_predictor_4x4/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd"; - specialize qw/vpx_highbd_tm_predictor_4x4/, "$sse2_x86inc"; + specialize qw/vpx_highbd_tm_predictor_4x4 sse2/; add_proto qw/void vpx_highbd_dc_predictor_4x4/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd"; - specialize qw/vpx_highbd_dc_predictor_4x4/, "$sse2_x86inc"; + specialize qw/vpx_highbd_dc_predictor_4x4 sse2/; add_proto qw/void vpx_highbd_dc_top_predictor_4x4/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd"; specialize qw/vpx_highbd_dc_top_predictor_4x4/; @@ -336,13 +313,13 @@ if (vpx_config("CONFIG_VP9_HIGHBITDEPTH") eq "yes") { specialize qw/vpx_highbd_d153_predictor_8x8/; add_proto qw/void vpx_highbd_v_predictor_8x8/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd"; - specialize qw/vpx_highbd_v_predictor_8x8/, "$sse2_x86inc"; + specialize qw/vpx_highbd_v_predictor_8x8 sse2/; add_proto qw/void vpx_highbd_tm_predictor_8x8/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd"; - specialize qw/vpx_highbd_tm_predictor_8x8/, "$sse2_x86inc"; + specialize qw/vpx_highbd_tm_predictor_8x8 sse2/; add_proto qw/void vpx_highbd_dc_predictor_8x8/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd"; - specialize qw/vpx_highbd_dc_predictor_8x8/, "$sse2_x86inc";; + specialize qw/vpx_highbd_dc_predictor_8x8 sse2/;; add_proto qw/void vpx_highbd_dc_top_predictor_8x8/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd"; specialize qw/vpx_highbd_dc_top_predictor_8x8/; @@ -384,13 +361,13 @@ if (vpx_config("CONFIG_VP9_HIGHBITDEPTH") eq "yes") { specialize qw/vpx_highbd_d153_predictor_16x16/; add_proto qw/void vpx_highbd_v_predictor_16x16/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd"; - specialize qw/vpx_highbd_v_predictor_16x16/, "$sse2_x86inc"; + specialize qw/vpx_highbd_v_predictor_16x16 sse2/; add_proto qw/void vpx_highbd_tm_predictor_16x16/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd"; - specialize qw/vpx_highbd_tm_predictor_16x16/, "$sse2_x86inc"; + specialize qw/vpx_highbd_tm_predictor_16x16 sse2/; add_proto qw/void vpx_highbd_dc_predictor_16x16/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd"; - specialize qw/vpx_highbd_dc_predictor_16x16/, "$sse2_x86inc"; + specialize qw/vpx_highbd_dc_predictor_16x16 sse2/; add_proto qw/void vpx_highbd_dc_top_predictor_16x16/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd"; specialize qw/vpx_highbd_dc_top_predictor_16x16/; @@ -432,13 +409,13 @@ if (vpx_config("CONFIG_VP9_HIGHBITDEPTH") eq "yes") { specialize qw/vpx_highbd_d153_predictor_32x32/; add_proto qw/void vpx_highbd_v_predictor_32x32/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd"; - specialize qw/vpx_highbd_v_predictor_32x32/, "$sse2_x86inc"; + specialize qw/vpx_highbd_v_predictor_32x32 sse2/; add_proto qw/void vpx_highbd_tm_predictor_32x32/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd"; - specialize qw/vpx_highbd_tm_predictor_32x32/, "$sse2_x86inc"; + specialize qw/vpx_highbd_tm_predictor_32x32 sse2/; add_proto qw/void vpx_highbd_dc_predictor_32x32/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd"; - specialize qw/vpx_highbd_dc_predictor_32x32/, "$sse2_x86inc"; + specialize qw/vpx_highbd_dc_predictor_32x32 sse2/; add_proto qw/void vpx_highbd_dc_top_predictor_32x32/, "uint16_t *dst, ptrdiff_t y_stride, const uint16_t *above, const uint16_t *left, int bd"; specialize qw/vpx_highbd_dc_top_predictor_32x32/; @@ -454,10 +431,10 @@ if (vpx_config("CONFIG_VP9_HIGHBITDEPTH") eq "yes") { # Sub Pixel Filters # add_proto qw/void vpx_convolve_copy/, "const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h"; -specialize qw/vpx_convolve_copy neon dspr2 msa/, "$sse2_x86inc"; +specialize qw/vpx_convolve_copy neon dspr2 msa sse2/; add_proto qw/void vpx_convolve_avg/, "const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h"; -specialize qw/vpx_convolve_avg neon dspr2 msa/, "$sse2_x86inc"; +specialize qw/vpx_convolve_avg neon dspr2 msa sse2/; add_proto qw/void vpx_convolve8/, "const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h"; specialize qw/vpx_convolve8 sse2 ssse3 neon dspr2 msa/, "$avx2_ssse3"; @@ -500,10 +477,10 @@ if (vpx_config("CONFIG_VP9_HIGHBITDEPTH") eq "yes") { # Sub Pixel Filters # add_proto qw/void vpx_highbd_convolve_copy/, "const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps"; - specialize qw/vpx_highbd_convolve_copy/, "$sse2_x86inc"; + specialize qw/vpx_highbd_convolve_copy sse2/; add_proto qw/void vpx_highbd_convolve_avg/, "const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps"; - specialize qw/vpx_highbd_convolve_avg/, "$sse2_x86inc"; + specialize qw/vpx_highbd_convolve_avg sse2/; add_proto qw/void vpx_highbd_convolve8/, "const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h, int bps"; specialize qw/vpx_highbd_convolve8/, "$sse2_x86_64"; @@ -674,7 +651,7 @@ if (vpx_config("CONFIG_VP9_HIGHBITDEPTH") eq "yes") { specialize qw/vpx_fdct4x4_1 sse2/; add_proto qw/void vpx_fdct8x8/, "const int16_t *input, tran_low_t *output, int stride"; - specialize qw/vpx_fdct8x8 sse2 neon msa/, "$ssse3_x86_64_x86inc"; + specialize qw/vpx_fdct8x8 sse2 neon msa/, "$ssse3_x86_64"; add_proto qw/void vpx_fdct8x8_1/, "const int16_t *input, tran_low_t *output, int stride"; specialize qw/vpx_fdct8x8_1 sse2 neon msa/; @@ -706,7 +683,7 @@ if (vpx_config("CONFIG_VP9_HIGHBITDEPTH") eq "yes") { specialize qw/vpx_iwht4x4_1_add/; add_proto qw/void vpx_iwht4x4_16_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride"; - specialize qw/vpx_iwht4x4_16_add/, "$sse2_x86inc"; + specialize qw/vpx_iwht4x4_16_add sse2/; add_proto qw/void vpx_highbd_idct4x4_1_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride, int bd"; specialize qw/vpx_highbd_idct4x4_1_add/; @@ -792,10 +769,10 @@ if (vpx_config("CONFIG_VP9_HIGHBITDEPTH") eq "yes") { specialize qw/vpx_idct4x4_1_add sse2/; add_proto qw/void vpx_idct8x8_64_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride"; - specialize qw/vpx_idct8x8_64_add sse2/, "$ssse3_x86_64_x86inc"; + specialize qw/vpx_idct8x8_64_add sse2/, "$ssse3_x86_64"; add_proto qw/void vpx_idct8x8_12_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride"; - specialize qw/vpx_idct8x8_12_add sse2/, "$ssse3_x86_64_x86inc"; + specialize qw/vpx_idct8x8_12_add sse2/, "$ssse3_x86_64"; add_proto qw/void vpx_idct8x8_1_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride"; specialize qw/vpx_idct8x8_1_add sse2/; @@ -810,15 +787,15 @@ if (vpx_config("CONFIG_VP9_HIGHBITDEPTH") eq "yes") { specialize qw/vpx_idct16x16_1_add sse2/; add_proto qw/void vpx_idct32x32_1024_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride"; - specialize qw/vpx_idct32x32_1024_add sse2/, "$ssse3_x86_64_x86inc"; + specialize qw/vpx_idct32x32_1024_add sse2/, "$ssse3_x86_64"; add_proto qw/void vpx_idct32x32_135_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride"; - specialize qw/vpx_idct32x32_135_add sse2/, "$ssse3_x86_64_x86inc"; + specialize qw/vpx_idct32x32_135_add sse2/, "$ssse3_x86_64"; # Need to add 135 eob idct32x32 implementations. $vpx_idct32x32_135_add_sse2=vpx_idct32x32_1024_add_sse2; add_proto qw/void vpx_idct32x32_34_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride"; - specialize qw/vpx_idct32x32_34_add sse2/, "$ssse3_x86_64_x86inc"; + specialize qw/vpx_idct32x32_34_add sse2/, "$ssse3_x86_64"; add_proto qw/void vpx_idct32x32_1_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride"; specialize qw/vpx_idct32x32_1_add sse2/; @@ -893,10 +870,10 @@ if (vpx_config("CONFIG_VP9_HIGHBITDEPTH") eq "yes") { specialize qw/vpx_idct8x8_1_add sse2 neon dspr2 msa/; add_proto qw/void vpx_idct8x8_64_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride"; - specialize qw/vpx_idct8x8_64_add sse2 neon dspr2 msa/, "$ssse3_x86_64_x86inc"; + specialize qw/vpx_idct8x8_64_add sse2 neon dspr2 msa/, "$ssse3_x86_64"; add_proto qw/void vpx_idct8x8_12_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride"; - specialize qw/vpx_idct8x8_12_add sse2 neon dspr2 msa/, "$ssse3_x86_64_x86inc"; + specialize qw/vpx_idct8x8_12_add sse2 neon dspr2 msa/, "$ssse3_x86_64"; add_proto qw/void vpx_idct16x16_1_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride"; specialize qw/vpx_idct16x16_1_add sse2 neon dspr2 msa/; @@ -908,10 +885,10 @@ if (vpx_config("CONFIG_VP9_HIGHBITDEPTH") eq "yes") { specialize qw/vpx_idct16x16_10_add sse2 neon dspr2 msa/; add_proto qw/void vpx_idct32x32_1024_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride"; - specialize qw/vpx_idct32x32_1024_add sse2 neon dspr2 msa/, "$ssse3_x86_64_x86inc"; + specialize qw/vpx_idct32x32_1024_add sse2 neon dspr2 msa/, "$ssse3_x86_64"; add_proto qw/void vpx_idct32x32_135_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride"; - specialize qw/vpx_idct32x32_135_add sse2 neon dspr2 msa/, "$ssse3_x86_64_x86inc"; + specialize qw/vpx_idct32x32_135_add sse2 neon dspr2 msa/, "$ssse3_x86_64"; # Need to add 135 eob idct32x32 implementations. $vpx_idct32x32_135_add_sse2=vpx_idct32x32_1024_add_sse2; $vpx_idct32x32_135_add_neon=vpx_idct32x32_1024_add_neon; @@ -919,7 +896,7 @@ if (vpx_config("CONFIG_VP9_HIGHBITDEPTH") eq "yes") { $vpx_idct32x32_135_add_msa=vpx_idct32x32_1024_add_msa; add_proto qw/void vpx_idct32x32_34_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride"; - specialize qw/vpx_idct32x32_34_add sse2 neon dspr2 msa/, "$ssse3_x86_64_x86inc"; + specialize qw/vpx_idct32x32_34_add sse2 neon dspr2 msa/, "$ssse3_x86_64"; # Need to add 34 eob idct32x32 neon implementation. $vpx_idct32x32_34_add_neon=vpx_idct32x32_1024_add_neon; @@ -930,7 +907,7 @@ if (vpx_config("CONFIG_VP9_HIGHBITDEPTH") eq "yes") { specialize qw/vpx_iwht4x4_1_add msa/; add_proto qw/void vpx_iwht4x4_16_add/, "const tran_low_t *input, uint8_t *dest, int dest_stride"; - specialize qw/vpx_iwht4x4_16_add msa/, "$sse2_x86inc"; + specialize qw/vpx_iwht4x4_16_add msa sse2/; } # CONFIG_EMULATE_HARDWARE } # CONFIG_VP9_HIGHBITDEPTH } # CONFIG_VP9 @@ -940,10 +917,10 @@ if (vpx_config("CONFIG_VP9_HIGHBITDEPTH") eq "yes") { # if (vpx_config("CONFIG_VP9_ENCODER") eq "yes") { add_proto qw/void vpx_quantize_b/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan"; - specialize qw/vpx_quantize_b sse2/, "$ssse3_x86_64_x86inc", "$avx_x86_64_x86inc"; + specialize qw/vpx_quantize_b sse2/, "$ssse3_x86_64", "$avx_x86_64"; add_proto qw/void vpx_quantize_b_32x32/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan"; - specialize qw/vpx_quantize_b_32x32/, "$ssse3_x86_64_x86inc", "$avx_x86_64_x86inc"; + specialize qw/vpx_quantize_b_32x32/, "$ssse3_x86_64", "$avx_x86_64"; if (vpx_config("CONFIG_VP9_HIGHBITDEPTH") eq "yes") { add_proto qw/void vpx_highbd_quantize_b/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan"; @@ -959,49 +936,49 @@ if (vpx_config("CONFIG_ENCODERS") eq "yes") { # Block subtraction # add_proto qw/void vpx_subtract_block/, "int rows, int cols, int16_t *diff_ptr, ptrdiff_t diff_stride, const uint8_t *src_ptr, ptrdiff_t src_stride, const uint8_t *pred_ptr, ptrdiff_t pred_stride"; -specialize qw/vpx_subtract_block neon msa/, "$sse2_x86inc"; +specialize qw/vpx_subtract_block neon msa sse2/; # # Single block SAD # add_proto qw/unsigned int vpx_sad64x64/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride"; -specialize qw/vpx_sad64x64 avx2 neon msa/, "$sse2_x86inc"; +specialize qw/vpx_sad64x64 avx2 neon msa sse2/; add_proto qw/unsigned int vpx_sad64x32/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride"; -specialize qw/vpx_sad64x32 avx2 msa/, "$sse2_x86inc"; +specialize qw/vpx_sad64x32 avx2 msa sse2/; add_proto qw/unsigned int vpx_sad32x64/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride"; -specialize qw/vpx_sad32x64 avx2 msa/, "$sse2_x86inc"; +specialize qw/vpx_sad32x64 avx2 msa sse2/; add_proto qw/unsigned int vpx_sad32x32/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride"; -specialize qw/vpx_sad32x32 avx2 neon msa/, "$sse2_x86inc"; +specialize qw/vpx_sad32x32 avx2 neon msa sse2/; add_proto qw/unsigned int vpx_sad32x16/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride"; -specialize qw/vpx_sad32x16 avx2 msa/, "$sse2_x86inc"; +specialize qw/vpx_sad32x16 avx2 msa sse2/; add_proto qw/unsigned int vpx_sad16x32/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride"; -specialize qw/vpx_sad16x32 msa/, "$sse2_x86inc"; +specialize qw/vpx_sad16x32 msa sse2/; add_proto qw/unsigned int vpx_sad16x16/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride"; -specialize qw/vpx_sad16x16 media neon msa/, "$sse2_x86inc"; +specialize qw/vpx_sad16x16 media neon msa sse2/; add_proto qw/unsigned int vpx_sad16x8/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride"; -specialize qw/vpx_sad16x8 neon msa/, "$sse2_x86inc"; +specialize qw/vpx_sad16x8 neon msa sse2/; add_proto qw/unsigned int vpx_sad8x16/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride"; -specialize qw/vpx_sad8x16 neon msa/, "$sse2_x86inc"; +specialize qw/vpx_sad8x16 neon msa sse2/; add_proto qw/unsigned int vpx_sad8x8/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride"; -specialize qw/vpx_sad8x8 neon msa/, "$sse2_x86inc"; +specialize qw/vpx_sad8x8 neon msa sse2/; add_proto qw/unsigned int vpx_sad8x4/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride"; -specialize qw/vpx_sad8x4 msa/, "$sse2_x86inc"; +specialize qw/vpx_sad8x4 msa sse2/; add_proto qw/unsigned int vpx_sad4x8/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride"; -specialize qw/vpx_sad4x8 msa/, "$sse2_x86inc"; +specialize qw/vpx_sad4x8 msa sse2/; add_proto qw/unsigned int vpx_sad4x4/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride"; -specialize qw/vpx_sad4x4 neon msa/, "$sse2_x86inc"; +specialize qw/vpx_sad4x4 neon msa sse2/; # # Avg @@ -1017,7 +994,7 @@ if (vpx_config("CONFIG_VP9_ENCODER") eq "yes") { specialize qw/vpx_minmax_8x8 sse2 neon/; add_proto qw/void vpx_hadamard_8x8/, "const int16_t *src_diff, int src_stride, int16_t *coeff"; - specialize qw/vpx_hadamard_8x8 sse2 neon/, "$ssse3_x86_64_x86inc"; + specialize qw/vpx_hadamard_8x8 sse2 neon/, "$ssse3_x86_64"; add_proto qw/void vpx_hadamard_16x16/, "const int16_t *src_diff, int src_stride, int16_t *coeff"; specialize qw/vpx_hadamard_16x16 sse2 neon/; @@ -1036,43 +1013,43 @@ if (vpx_config("CONFIG_VP9_ENCODER") eq "yes") { } # CONFIG_VP9_ENCODER add_proto qw/unsigned int vpx_sad64x64_avg/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred"; -specialize qw/vpx_sad64x64_avg avx2 msa/, "$sse2_x86inc"; +specialize qw/vpx_sad64x64_avg avx2 msa sse2/; add_proto qw/unsigned int vpx_sad64x32_avg/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred"; -specialize qw/vpx_sad64x32_avg avx2 msa/, "$sse2_x86inc"; +specialize qw/vpx_sad64x32_avg avx2 msa sse2/; add_proto qw/unsigned int vpx_sad32x64_avg/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred"; -specialize qw/vpx_sad32x64_avg avx2 msa/, "$sse2_x86inc"; +specialize qw/vpx_sad32x64_avg avx2 msa sse2/; add_proto qw/unsigned int vpx_sad32x32_avg/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred"; -specialize qw/vpx_sad32x32_avg avx2 msa/, "$sse2_x86inc"; +specialize qw/vpx_sad32x32_avg avx2 msa sse2/; add_proto qw/unsigned int vpx_sad32x16_avg/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred"; -specialize qw/vpx_sad32x16_avg avx2 msa/, "$sse2_x86inc"; +specialize qw/vpx_sad32x16_avg avx2 msa sse2/; add_proto qw/unsigned int vpx_sad16x32_avg/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred"; -specialize qw/vpx_sad16x32_avg msa/, "$sse2_x86inc"; +specialize qw/vpx_sad16x32_avg msa sse2/; add_proto qw/unsigned int vpx_sad16x16_avg/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred"; -specialize qw/vpx_sad16x16_avg msa/, "$sse2_x86inc"; +specialize qw/vpx_sad16x16_avg msa sse2/; add_proto qw/unsigned int vpx_sad16x8_avg/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred"; -specialize qw/vpx_sad16x8_avg msa/, "$sse2_x86inc"; +specialize qw/vpx_sad16x8_avg msa sse2/; add_proto qw/unsigned int vpx_sad8x16_avg/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred"; -specialize qw/vpx_sad8x16_avg msa/, "$sse2_x86inc"; +specialize qw/vpx_sad8x16_avg msa sse2/; add_proto qw/unsigned int vpx_sad8x8_avg/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred"; -specialize qw/vpx_sad8x8_avg msa/, "$sse2_x86inc"; +specialize qw/vpx_sad8x8_avg msa sse2/; add_proto qw/unsigned int vpx_sad8x4_avg/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred"; -specialize qw/vpx_sad8x4_avg msa/, "$sse2_x86inc"; +specialize qw/vpx_sad8x4_avg msa sse2/; add_proto qw/unsigned int vpx_sad4x8_avg/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred"; -specialize qw/vpx_sad4x8_avg msa/, "$sse2_x86inc"; +specialize qw/vpx_sad4x8_avg msa sse2/; add_proto qw/unsigned int vpx_sad4x4_avg/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred"; -specialize qw/vpx_sad4x4_avg msa/, "$sse2_x86inc"; +specialize qw/vpx_sad4x4_avg msa sse2/; # # Multi-block SAD, comparing a reference to N blocks 1 pixel apart horizontally @@ -1131,43 +1108,43 @@ specialize qw/vpx_sad4x4x8 sse4_1 msa/; # Multi-block SAD, comparing a reference to N independent blocks # add_proto qw/void vpx_sad64x64x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array"; -specialize qw/vpx_sad64x64x4d avx2 neon msa/, "$sse2_x86inc"; +specialize qw/vpx_sad64x64x4d avx2 neon msa sse2/; add_proto qw/void vpx_sad64x32x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array"; -specialize qw/vpx_sad64x32x4d msa/, "$sse2_x86inc"; +specialize qw/vpx_sad64x32x4d msa sse2/; add_proto qw/void vpx_sad32x64x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array"; -specialize qw/vpx_sad32x64x4d msa/, "$sse2_x86inc"; +specialize qw/vpx_sad32x64x4d msa sse2/; add_proto qw/void vpx_sad32x32x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array"; -specialize qw/vpx_sad32x32x4d avx2 neon msa/, "$sse2_x86inc"; +specialize qw/vpx_sad32x32x4d avx2 neon msa sse2/; add_proto qw/void vpx_sad32x16x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array"; -specialize qw/vpx_sad32x16x4d msa/, "$sse2_x86inc"; +specialize qw/vpx_sad32x16x4d msa sse2/; add_proto qw/void vpx_sad16x32x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array"; -specialize qw/vpx_sad16x32x4d msa/, "$sse2_x86inc"; +specialize qw/vpx_sad16x32x4d msa sse2/; add_proto qw/void vpx_sad16x16x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array"; -specialize qw/vpx_sad16x16x4d neon msa/, "$sse2_x86inc"; +specialize qw/vpx_sad16x16x4d neon msa sse2/; add_proto qw/void vpx_sad16x8x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array"; -specialize qw/vpx_sad16x8x4d msa/, "$sse2_x86inc"; +specialize qw/vpx_sad16x8x4d msa sse2/; add_proto qw/void vpx_sad8x16x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array"; -specialize qw/vpx_sad8x16x4d msa/, "$sse2_x86inc"; +specialize qw/vpx_sad8x16x4d msa sse2/; add_proto qw/void vpx_sad8x8x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array"; -specialize qw/vpx_sad8x8x4d msa/, "$sse2_x86inc"; +specialize qw/vpx_sad8x8x4d msa sse2/; add_proto qw/void vpx_sad8x4x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array"; -specialize qw/vpx_sad8x4x4d msa/, "$sse2_x86inc"; +specialize qw/vpx_sad8x4x4d msa sse2/; add_proto qw/void vpx_sad4x8x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array"; -specialize qw/vpx_sad4x8x4d msa/, "$sse2_x86inc"; +specialize qw/vpx_sad4x8x4d msa sse2/; add_proto qw/void vpx_sad4x4x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t * const ref_ptr[], int ref_stride, uint32_t *sad_array"; -specialize qw/vpx_sad4x4x4d msa/, "$sse2_x86inc"; +specialize qw/vpx_sad4x4x4d msa sse2/; # # Structured Similarity (SSIM) @@ -1191,37 +1168,37 @@ if (vpx_config("CONFIG_VP9_HIGHBITDEPTH") eq "yes") { # Single block SAD # add_proto qw/unsigned int vpx_highbd_sad64x64/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride"; - specialize qw/vpx_highbd_sad64x64/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad64x64 sse2/; add_proto qw/unsigned int vpx_highbd_sad64x32/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride"; - specialize qw/vpx_highbd_sad64x32/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad64x32 sse2/; add_proto qw/unsigned int vpx_highbd_sad32x64/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride"; - specialize qw/vpx_highbd_sad32x64/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad32x64 sse2/; add_proto qw/unsigned int vpx_highbd_sad32x32/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride"; - specialize qw/vpx_highbd_sad32x32/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad32x32 sse2/; add_proto qw/unsigned int vpx_highbd_sad32x16/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride"; - specialize qw/vpx_highbd_sad32x16/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad32x16 sse2/; add_proto qw/unsigned int vpx_highbd_sad16x32/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride"; - specialize qw/vpx_highbd_sad16x32/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad16x32 sse2/; add_proto qw/unsigned int vpx_highbd_sad16x16/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride"; - specialize qw/vpx_highbd_sad16x16/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad16x16 sse2/; add_proto qw/unsigned int vpx_highbd_sad16x8/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride"; - specialize qw/vpx_highbd_sad16x8/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad16x8 sse2/; add_proto qw/unsigned int vpx_highbd_sad8x16/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride"; - specialize qw/vpx_highbd_sad8x16/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad8x16 sse2/; add_proto qw/unsigned int vpx_highbd_sad8x8/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride"; - specialize qw/vpx_highbd_sad8x8/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad8x8 sse2/; add_proto qw/unsigned int vpx_highbd_sad8x4/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride"; - specialize qw/vpx_highbd_sad8x4/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad8x4 sse2/; add_proto qw/unsigned int vpx_highbd_sad4x8/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride"; specialize qw/vpx_highbd_sad4x8/; @@ -1240,37 +1217,37 @@ if (vpx_config("CONFIG_VP9_HIGHBITDEPTH") eq "yes") { specialize qw/vpx_highbd_minmax_8x8/; add_proto qw/unsigned int vpx_highbd_sad64x64_avg/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred"; - specialize qw/vpx_highbd_sad64x64_avg/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad64x64_avg sse2/; add_proto qw/unsigned int vpx_highbd_sad64x32_avg/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred"; - specialize qw/vpx_highbd_sad64x32_avg/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad64x32_avg sse2/; add_proto qw/unsigned int vpx_highbd_sad32x64_avg/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred"; - specialize qw/vpx_highbd_sad32x64_avg/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad32x64_avg sse2/; add_proto qw/unsigned int vpx_highbd_sad32x32_avg/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred"; - specialize qw/vpx_highbd_sad32x32_avg/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad32x32_avg sse2/; add_proto qw/unsigned int vpx_highbd_sad32x16_avg/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred"; - specialize qw/vpx_highbd_sad32x16_avg/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad32x16_avg sse2/; add_proto qw/unsigned int vpx_highbd_sad16x32_avg/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred"; - specialize qw/vpx_highbd_sad16x32_avg/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad16x32_avg sse2/; add_proto qw/unsigned int vpx_highbd_sad16x16_avg/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred"; - specialize qw/vpx_highbd_sad16x16_avg/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad16x16_avg sse2/; add_proto qw/unsigned int vpx_highbd_sad16x8_avg/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred"; - specialize qw/vpx_highbd_sad16x8_avg/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad16x8_avg sse2/; add_proto qw/unsigned int vpx_highbd_sad8x16_avg/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred"; - specialize qw/vpx_highbd_sad8x16_avg/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad8x16_avg sse2/; add_proto qw/unsigned int vpx_highbd_sad8x8_avg/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred"; - specialize qw/vpx_highbd_sad8x8_avg/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad8x8_avg sse2/; add_proto qw/unsigned int vpx_highbd_sad8x4_avg/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred"; - specialize qw/vpx_highbd_sad8x4_avg/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad8x4_avg sse2/; add_proto qw/unsigned int vpx_highbd_sad4x8_avg/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred"; specialize qw/vpx_highbd_sad4x8_avg/; @@ -1335,43 +1312,43 @@ if (vpx_config("CONFIG_VP9_HIGHBITDEPTH") eq "yes") { # Multi-block SAD, comparing a reference to N independent blocks # add_proto qw/void vpx_highbd_sad64x64x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t* const ref_ptr[], int ref_stride, uint32_t *sad_array"; - specialize qw/vpx_highbd_sad64x64x4d/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad64x64x4d sse2/; add_proto qw/void vpx_highbd_sad64x32x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t* const ref_ptr[], int ref_stride, uint32_t *sad_array"; - specialize qw/vpx_highbd_sad64x32x4d/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad64x32x4d sse2/; add_proto qw/void vpx_highbd_sad32x64x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t* const ref_ptr[], int ref_stride, uint32_t *sad_array"; - specialize qw/vpx_highbd_sad32x64x4d/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad32x64x4d sse2/; add_proto qw/void vpx_highbd_sad32x32x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t* const ref_ptr[], int ref_stride, uint32_t *sad_array"; - specialize qw/vpx_highbd_sad32x32x4d/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad32x32x4d sse2/; add_proto qw/void vpx_highbd_sad32x16x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t* const ref_ptr[], int ref_stride, uint32_t *sad_array"; - specialize qw/vpx_highbd_sad32x16x4d/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad32x16x4d sse2/; add_proto qw/void vpx_highbd_sad16x32x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t* const ref_ptr[], int ref_stride, uint32_t *sad_array"; - specialize qw/vpx_highbd_sad16x32x4d/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad16x32x4d sse2/; add_proto qw/void vpx_highbd_sad16x16x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t* const ref_ptr[], int ref_stride, uint32_t *sad_array"; - specialize qw/vpx_highbd_sad16x16x4d/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad16x16x4d sse2/; add_proto qw/void vpx_highbd_sad16x8x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t* const ref_ptr[], int ref_stride, uint32_t *sad_array"; - specialize qw/vpx_highbd_sad16x8x4d/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad16x8x4d sse2/; add_proto qw/void vpx_highbd_sad8x16x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t* const ref_ptr[], int ref_stride, uint32_t *sad_array"; - specialize qw/vpx_highbd_sad8x16x4d/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad8x16x4d sse2/; add_proto qw/void vpx_highbd_sad8x8x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t* const ref_ptr[], int ref_stride, uint32_t *sad_array"; - specialize qw/vpx_highbd_sad8x8x4d/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad8x8x4d sse2/; add_proto qw/void vpx_highbd_sad8x4x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t* const ref_ptr[], int ref_stride, uint32_t *sad_array"; - specialize qw/vpx_highbd_sad8x4x4d/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad8x4x4d sse2/; add_proto qw/void vpx_highbd_sad4x8x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t* const ref_ptr[], int ref_stride, uint32_t *sad_array"; - specialize qw/vpx_highbd_sad4x8x4d/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad4x8x4d sse2/; add_proto qw/void vpx_highbd_sad4x4x4d/, "const uint8_t *src_ptr, int src_stride, const uint8_t* const ref_ptr[], int ref_stride, uint32_t *sad_array"; - specialize qw/vpx_highbd_sad4x4x4d/, "$sse2_x86inc"; + specialize qw/vpx_highbd_sad4x4x4d sse2/; # # Structured Similarity (SSIM) @@ -1460,82 +1437,82 @@ add_proto qw/void vpx_comp_avg_pred/, "uint8_t *comp_pred, const uint8_t *pred, # Subpixel Variance # add_proto qw/uint32_t vpx_sub_pixel_variance64x64/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_sub_pixel_variance64x64 avx2 neon msa/, "$sse2_x86inc", "$ssse3_x86inc"; + specialize qw/vpx_sub_pixel_variance64x64 avx2 neon msa sse2 ssse3/; add_proto qw/uint32_t vpx_sub_pixel_variance64x32/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_sub_pixel_variance64x32 msa/, "$sse2_x86inc", "$ssse3_x86inc"; + specialize qw/vpx_sub_pixel_variance64x32 msa sse2 ssse3/; add_proto qw/uint32_t vpx_sub_pixel_variance32x64/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_sub_pixel_variance32x64 msa/, "$sse2_x86inc", "$ssse3_x86inc"; + specialize qw/vpx_sub_pixel_variance32x64 msa sse2 ssse3/; add_proto qw/uint32_t vpx_sub_pixel_variance32x32/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_sub_pixel_variance32x32 avx2 neon msa/, "$sse2_x86inc", "$ssse3_x86inc"; + specialize qw/vpx_sub_pixel_variance32x32 avx2 neon msa sse2 ssse3/; add_proto qw/uint32_t vpx_sub_pixel_variance32x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_sub_pixel_variance32x16 msa/, "$sse2_x86inc", "$ssse3_x86inc"; + specialize qw/vpx_sub_pixel_variance32x16 msa sse2 ssse3/; add_proto qw/uint32_t vpx_sub_pixel_variance16x32/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_sub_pixel_variance16x32 msa/, "$sse2_x86inc", "$ssse3_x86inc"; + specialize qw/vpx_sub_pixel_variance16x32 msa sse2 ssse3/; add_proto qw/uint32_t vpx_sub_pixel_variance16x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_sub_pixel_variance16x16 media neon msa/, "$sse2_x86inc", "$ssse3_x86inc"; + specialize qw/vpx_sub_pixel_variance16x16 media neon msa sse2 ssse3/; add_proto qw/uint32_t vpx_sub_pixel_variance16x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_sub_pixel_variance16x8 msa/, "$sse2_x86inc", "$ssse3_x86inc"; + specialize qw/vpx_sub_pixel_variance16x8 msa sse2 ssse3/; add_proto qw/uint32_t vpx_sub_pixel_variance8x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_sub_pixel_variance8x16 msa/, "$sse2_x86inc", "$ssse3_x86inc"; + specialize qw/vpx_sub_pixel_variance8x16 msa sse2 ssse3/; add_proto qw/uint32_t vpx_sub_pixel_variance8x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_sub_pixel_variance8x8 media neon msa/, "$sse2_x86inc", "$ssse3_x86inc"; + specialize qw/vpx_sub_pixel_variance8x8 media neon msa sse2 ssse3/; add_proto qw/uint32_t vpx_sub_pixel_variance8x4/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_sub_pixel_variance8x4 msa/, "$sse2_x86inc", "$ssse3_x86inc"; + specialize qw/vpx_sub_pixel_variance8x4 msa sse2 ssse3/; add_proto qw/uint32_t vpx_sub_pixel_variance4x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_sub_pixel_variance4x8 msa/, "$sse2_x86inc", "$ssse3_x86inc"; + specialize qw/vpx_sub_pixel_variance4x8 msa sse2 ssse3/; add_proto qw/uint32_t vpx_sub_pixel_variance4x4/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_sub_pixel_variance4x4 msa/, "$sse2_x86inc", "$ssse3_x86inc"; + specialize qw/vpx_sub_pixel_variance4x4 msa sse2 ssse3/; add_proto qw/uint32_t vpx_sub_pixel_avg_variance64x64/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred"; - specialize qw/vpx_sub_pixel_avg_variance64x64 avx2 msa/, "$sse2_x86inc", "$ssse3_x86inc"; + specialize qw/vpx_sub_pixel_avg_variance64x64 avx2 msa sse2 ssse3/; add_proto qw/uint32_t vpx_sub_pixel_avg_variance64x32/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred"; - specialize qw/vpx_sub_pixel_avg_variance64x32 msa/, "$sse2_x86inc", "$ssse3_x86inc"; + specialize qw/vpx_sub_pixel_avg_variance64x32 msa sse2 ssse3/; add_proto qw/uint32_t vpx_sub_pixel_avg_variance32x64/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred"; - specialize qw/vpx_sub_pixel_avg_variance32x64 msa/, "$sse2_x86inc", "$ssse3_x86inc"; + specialize qw/vpx_sub_pixel_avg_variance32x64 msa sse2 ssse3/; add_proto qw/uint32_t vpx_sub_pixel_avg_variance32x32/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred"; - specialize qw/vpx_sub_pixel_avg_variance32x32 avx2 msa/, "$sse2_x86inc", "$ssse3_x86inc"; + specialize qw/vpx_sub_pixel_avg_variance32x32 avx2 msa sse2 ssse3/; add_proto qw/uint32_t vpx_sub_pixel_avg_variance32x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred"; - specialize qw/vpx_sub_pixel_avg_variance32x16 msa/, "$sse2_x86inc", "$ssse3_x86inc"; + specialize qw/vpx_sub_pixel_avg_variance32x16 msa sse2 ssse3/; add_proto qw/uint32_t vpx_sub_pixel_avg_variance16x32/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred"; - specialize qw/vpx_sub_pixel_avg_variance16x32 msa/, "$sse2_x86inc", "$ssse3_x86inc"; + specialize qw/vpx_sub_pixel_avg_variance16x32 msa sse2 ssse3/; add_proto qw/uint32_t vpx_sub_pixel_avg_variance16x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred"; - specialize qw/vpx_sub_pixel_avg_variance16x16 msa/, "$sse2_x86inc", "$ssse3_x86inc"; + specialize qw/vpx_sub_pixel_avg_variance16x16 msa sse2 ssse3/; add_proto qw/uint32_t vpx_sub_pixel_avg_variance16x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred"; - specialize qw/vpx_sub_pixel_avg_variance16x8 msa/, "$sse2_x86inc", "$ssse3_x86inc"; + specialize qw/vpx_sub_pixel_avg_variance16x8 msa sse2 ssse3/; add_proto qw/uint32_t vpx_sub_pixel_avg_variance8x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred"; - specialize qw/vpx_sub_pixel_avg_variance8x16 msa/, "$sse2_x86inc", "$ssse3_x86inc"; + specialize qw/vpx_sub_pixel_avg_variance8x16 msa sse2 ssse3/; add_proto qw/uint32_t vpx_sub_pixel_avg_variance8x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred"; - specialize qw/vpx_sub_pixel_avg_variance8x8 msa/, "$sse2_x86inc", "$ssse3_x86inc"; + specialize qw/vpx_sub_pixel_avg_variance8x8 msa sse2 ssse3/; add_proto qw/uint32_t vpx_sub_pixel_avg_variance8x4/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred"; - specialize qw/vpx_sub_pixel_avg_variance8x4 msa/, "$sse2_x86inc", "$ssse3_x86inc"; + specialize qw/vpx_sub_pixel_avg_variance8x4 msa sse2 ssse3/; add_proto qw/uint32_t vpx_sub_pixel_avg_variance4x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred"; - specialize qw/vpx_sub_pixel_avg_variance4x8 msa/, "$sse2_x86inc", "$ssse3_x86inc"; + specialize qw/vpx_sub_pixel_avg_variance4x8 msa sse2 ssse3/; add_proto qw/uint32_t vpx_sub_pixel_avg_variance4x4/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred"; - specialize qw/vpx_sub_pixel_avg_variance4x4 msa/, "$sse2_x86inc", "$ssse3_x86inc"; + specialize qw/vpx_sub_pixel_avg_variance4x4 msa sse2 ssse3/; # # Specialty Subpixel @@ -1691,217 +1668,217 @@ if (vpx_config("CONFIG_VP9_HIGHBITDEPTH") eq "yes") { # Subpixel Variance # add_proto qw/uint32_t vpx_highbd_12_sub_pixel_variance64x64/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_highbd_12_sub_pixel_variance64x64/, "$sse2_x86inc"; + specialize qw/vpx_highbd_12_sub_pixel_variance64x64 sse2/; add_proto qw/uint32_t vpx_highbd_12_sub_pixel_variance64x32/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_highbd_12_sub_pixel_variance64x32/, "$sse2_x86inc"; + specialize qw/vpx_highbd_12_sub_pixel_variance64x32 sse2/; add_proto qw/uint32_t vpx_highbd_12_sub_pixel_variance32x64/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_highbd_12_sub_pixel_variance32x64/, "$sse2_x86inc"; + specialize qw/vpx_highbd_12_sub_pixel_variance32x64 sse2/; add_proto qw/uint32_t vpx_highbd_12_sub_pixel_variance32x32/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_highbd_12_sub_pixel_variance32x32/, "$sse2_x86inc"; + specialize qw/vpx_highbd_12_sub_pixel_variance32x32 sse2/; add_proto qw/uint32_t vpx_highbd_12_sub_pixel_variance32x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_highbd_12_sub_pixel_variance32x16/, "$sse2_x86inc"; + specialize qw/vpx_highbd_12_sub_pixel_variance32x16 sse2/; add_proto qw/uint32_t vpx_highbd_12_sub_pixel_variance16x32/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_highbd_12_sub_pixel_variance16x32/, "$sse2_x86inc"; + specialize qw/vpx_highbd_12_sub_pixel_variance16x32 sse2/; add_proto qw/uint32_t vpx_highbd_12_sub_pixel_variance16x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_highbd_12_sub_pixel_variance16x16/, "$sse2_x86inc"; + specialize qw/vpx_highbd_12_sub_pixel_variance16x16 sse2/; add_proto qw/uint32_t vpx_highbd_12_sub_pixel_variance16x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_highbd_12_sub_pixel_variance16x8/, "$sse2_x86inc"; + specialize qw/vpx_highbd_12_sub_pixel_variance16x8 sse2/; add_proto qw/uint32_t vpx_highbd_12_sub_pixel_variance8x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_highbd_12_sub_pixel_variance8x16/, "$sse2_x86inc"; + specialize qw/vpx_highbd_12_sub_pixel_variance8x16 sse2/; add_proto qw/uint32_t vpx_highbd_12_sub_pixel_variance8x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_highbd_12_sub_pixel_variance8x8/, "$sse2_x86inc"; + specialize qw/vpx_highbd_12_sub_pixel_variance8x8 sse2/; add_proto qw/uint32_t vpx_highbd_12_sub_pixel_variance8x4/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_highbd_12_sub_pixel_variance8x4/, "$sse2_x86inc"; + specialize qw/vpx_highbd_12_sub_pixel_variance8x4 sse2/; add_proto qw/uint32_t vpx_highbd_12_sub_pixel_variance4x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; add_proto qw/uint32_t vpx_highbd_12_sub_pixel_variance4x4/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; add_proto qw/uint32_t vpx_highbd_10_sub_pixel_variance64x64/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_highbd_10_sub_pixel_variance64x64/, "$sse2_x86inc"; + specialize qw/vpx_highbd_10_sub_pixel_variance64x64 sse2/; add_proto qw/uint32_t vpx_highbd_10_sub_pixel_variance64x32/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_highbd_10_sub_pixel_variance64x32/, "$sse2_x86inc"; + specialize qw/vpx_highbd_10_sub_pixel_variance64x32 sse2/; add_proto qw/uint32_t vpx_highbd_10_sub_pixel_variance32x64/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_highbd_10_sub_pixel_variance32x64/, "$sse2_x86inc"; + specialize qw/vpx_highbd_10_sub_pixel_variance32x64 sse2/; add_proto qw/uint32_t vpx_highbd_10_sub_pixel_variance32x32/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_highbd_10_sub_pixel_variance32x32/, "$sse2_x86inc"; + specialize qw/vpx_highbd_10_sub_pixel_variance32x32 sse2/; add_proto qw/uint32_t vpx_highbd_10_sub_pixel_variance32x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_highbd_10_sub_pixel_variance32x16/, "$sse2_x86inc"; + specialize qw/vpx_highbd_10_sub_pixel_variance32x16 sse2/; add_proto qw/uint32_t vpx_highbd_10_sub_pixel_variance16x32/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_highbd_10_sub_pixel_variance16x32/, "$sse2_x86inc"; + specialize qw/vpx_highbd_10_sub_pixel_variance16x32 sse2/; add_proto qw/uint32_t vpx_highbd_10_sub_pixel_variance16x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_highbd_10_sub_pixel_variance16x16/, "$sse2_x86inc"; + specialize qw/vpx_highbd_10_sub_pixel_variance16x16 sse2/; add_proto qw/uint32_t vpx_highbd_10_sub_pixel_variance16x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_highbd_10_sub_pixel_variance16x8/, "$sse2_x86inc"; + specialize qw/vpx_highbd_10_sub_pixel_variance16x8 sse2/; add_proto qw/uint32_t vpx_highbd_10_sub_pixel_variance8x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_highbd_10_sub_pixel_variance8x16/, "$sse2_x86inc"; + specialize qw/vpx_highbd_10_sub_pixel_variance8x16 sse2/; add_proto qw/uint32_t vpx_highbd_10_sub_pixel_variance8x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_highbd_10_sub_pixel_variance8x8/, "$sse2_x86inc"; + specialize qw/vpx_highbd_10_sub_pixel_variance8x8 sse2/; add_proto qw/uint32_t vpx_highbd_10_sub_pixel_variance8x4/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_highbd_10_sub_pixel_variance8x4/, "$sse2_x86inc"; + specialize qw/vpx_highbd_10_sub_pixel_variance8x4 sse2/; add_proto qw/uint32_t vpx_highbd_10_sub_pixel_variance4x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; add_proto qw/uint32_t vpx_highbd_10_sub_pixel_variance4x4/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; add_proto qw/uint32_t vpx_highbd_8_sub_pixel_variance64x64/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_highbd_8_sub_pixel_variance64x64/, "$sse2_x86inc"; + specialize qw/vpx_highbd_8_sub_pixel_variance64x64 sse2/; add_proto qw/uint32_t vpx_highbd_8_sub_pixel_variance64x32/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_highbd_8_sub_pixel_variance64x32/, "$sse2_x86inc"; + specialize qw/vpx_highbd_8_sub_pixel_variance64x32 sse2/; add_proto qw/uint32_t vpx_highbd_8_sub_pixel_variance32x64/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_highbd_8_sub_pixel_variance32x64/, "$sse2_x86inc"; + specialize qw/vpx_highbd_8_sub_pixel_variance32x64 sse2/; add_proto qw/uint32_t vpx_highbd_8_sub_pixel_variance32x32/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_highbd_8_sub_pixel_variance32x32/, "$sse2_x86inc"; + specialize qw/vpx_highbd_8_sub_pixel_variance32x32 sse2/; add_proto qw/uint32_t vpx_highbd_8_sub_pixel_variance32x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_highbd_8_sub_pixel_variance32x16/, "$sse2_x86inc"; + specialize qw/vpx_highbd_8_sub_pixel_variance32x16 sse2/; add_proto qw/uint32_t vpx_highbd_8_sub_pixel_variance16x32/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_highbd_8_sub_pixel_variance16x32/, "$sse2_x86inc"; + specialize qw/vpx_highbd_8_sub_pixel_variance16x32 sse2/; add_proto qw/uint32_t vpx_highbd_8_sub_pixel_variance16x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_highbd_8_sub_pixel_variance16x16/, "$sse2_x86inc"; + specialize qw/vpx_highbd_8_sub_pixel_variance16x16 sse2/; add_proto qw/uint32_t vpx_highbd_8_sub_pixel_variance16x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_highbd_8_sub_pixel_variance16x8/, "$sse2_x86inc"; + specialize qw/vpx_highbd_8_sub_pixel_variance16x8 sse2/; add_proto qw/uint32_t vpx_highbd_8_sub_pixel_variance8x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_highbd_8_sub_pixel_variance8x16/, "$sse2_x86inc"; + specialize qw/vpx_highbd_8_sub_pixel_variance8x16 sse2/; add_proto qw/uint32_t vpx_highbd_8_sub_pixel_variance8x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_highbd_8_sub_pixel_variance8x8/, "$sse2_x86inc"; + specialize qw/vpx_highbd_8_sub_pixel_variance8x8 sse2/; add_proto qw/uint32_t vpx_highbd_8_sub_pixel_variance8x4/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; - specialize qw/vpx_highbd_8_sub_pixel_variance8x4/, "$sse2_x86inc"; + specialize qw/vpx_highbd_8_sub_pixel_variance8x4 sse2/; add_proto qw/uint32_t vpx_highbd_8_sub_pixel_variance4x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; add_proto qw/uint32_t vpx_highbd_8_sub_pixel_variance4x4/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse"; add_proto qw/uint32_t vpx_highbd_12_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_highbd_12_sub_pixel_avg_variance64x64/, "$sse2_x86inc"; + specialize qw/vpx_highbd_12_sub_pixel_avg_variance64x64 sse2/; add_proto qw/uint32_t vpx_highbd_12_sub_pixel_avg_variance64x32/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred"; - specialize qw/vpx_highbd_12_sub_pixel_avg_variance64x32/, "$sse2_x86inc"; + specialize qw/vpx_highbd_12_sub_pixel_avg_variance64x32 sse2/; add_proto qw/uint32_t vpx_highbd_12_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_highbd_12_sub_pixel_avg_variance32x64/, "$sse2_x86inc"; + specialize qw/vpx_highbd_12_sub_pixel_avg_variance32x64 sse2/; add_proto qw/uint32_t vpx_highbd_12_sub_pixel_avg_variance32x32/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred"; - specialize qw/vpx_highbd_12_sub_pixel_avg_variance32x32/, "$sse2_x86inc"; + specialize qw/vpx_highbd_12_sub_pixel_avg_variance32x32 sse2/; add_proto qw/uint32_t vpx_highbd_12_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_highbd_12_sub_pixel_avg_variance32x16/, "$sse2_x86inc"; + specialize qw/vpx_highbd_12_sub_pixel_avg_variance32x16 sse2/; add_proto qw/uint32_t vpx_highbd_12_sub_pixel_avg_variance16x32/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred"; - specialize qw/vpx_highbd_12_sub_pixel_avg_variance16x32/, "$sse2_x86inc"; + specialize qw/vpx_highbd_12_sub_pixel_avg_variance16x32 sse2/; add_proto qw/uint32_t vpx_highbd_12_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_highbd_12_sub_pixel_avg_variance16x16/, "$sse2_x86inc"; + specialize qw/vpx_highbd_12_sub_pixel_avg_variance16x16 sse2/; add_proto qw/uint32_t vpx_highbd_12_sub_pixel_avg_variance16x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred"; - specialize qw/vpx_highbd_12_sub_pixel_avg_variance16x8/, "$sse2_x86inc"; + specialize qw/vpx_highbd_12_sub_pixel_avg_variance16x8 sse2/; add_proto qw/uint32_t vpx_highbd_12_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_highbd_12_sub_pixel_avg_variance8x16/, "$sse2_x86inc"; + specialize qw/vpx_highbd_12_sub_pixel_avg_variance8x16 sse2/; add_proto qw/uint32_t vpx_highbd_12_sub_pixel_avg_variance8x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred"; - specialize qw/vpx_highbd_12_sub_pixel_avg_variance8x8/, "$sse2_x86inc"; + specialize qw/vpx_highbd_12_sub_pixel_avg_variance8x8 sse2/; add_proto qw/uint32_t vpx_highbd_12_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_highbd_12_sub_pixel_avg_variance8x4/, "$sse2_x86inc"; + specialize qw/vpx_highbd_12_sub_pixel_avg_variance8x4 sse2/; add_proto qw/uint32_t vpx_highbd_12_sub_pixel_avg_variance4x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred"; add_proto qw/uint32_t vpx_highbd_12_sub_pixel_avg_variance4x4/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred"; add_proto qw/uint32_t vpx_highbd_10_sub_pixel_avg_variance64x64/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred"; - specialize qw/vpx_highbd_10_sub_pixel_avg_variance64x64/, "$sse2_x86inc"; + specialize qw/vpx_highbd_10_sub_pixel_avg_variance64x64 sse2/; add_proto qw/uint32_t vpx_highbd_10_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_highbd_10_sub_pixel_avg_variance64x32/, "$sse2_x86inc"; + specialize qw/vpx_highbd_10_sub_pixel_avg_variance64x32 sse2/; add_proto qw/uint32_t vpx_highbd_10_sub_pixel_avg_variance32x64/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred"; - specialize qw/vpx_highbd_10_sub_pixel_avg_variance32x64/, "$sse2_x86inc"; + specialize qw/vpx_highbd_10_sub_pixel_avg_variance32x64 sse2/; add_proto qw/uint32_t vpx_highbd_10_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_highbd_10_sub_pixel_avg_variance32x32/, "$sse2_x86inc"; + specialize qw/vpx_highbd_10_sub_pixel_avg_variance32x32 sse2/; add_proto qw/uint32_t vpx_highbd_10_sub_pixel_avg_variance32x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred"; - specialize qw/vpx_highbd_10_sub_pixel_avg_variance32x16/, "$sse2_x86inc"; + specialize qw/vpx_highbd_10_sub_pixel_avg_variance32x16 sse2/; add_proto qw/uint32_t vpx_highbd_10_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_highbd_10_sub_pixel_avg_variance16x32/, "$sse2_x86inc"; + specialize qw/vpx_highbd_10_sub_pixel_avg_variance16x32 sse2/; add_proto qw/uint32_t vpx_highbd_10_sub_pixel_avg_variance16x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred"; - specialize qw/vpx_highbd_10_sub_pixel_avg_variance16x16/, "$sse2_x86inc"; + specialize qw/vpx_highbd_10_sub_pixel_avg_variance16x16 sse2/; add_proto qw/uint32_t vpx_highbd_10_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_highbd_10_sub_pixel_avg_variance16x8/, "$sse2_x86inc"; + specialize qw/vpx_highbd_10_sub_pixel_avg_variance16x8 sse2/; add_proto qw/uint32_t vpx_highbd_10_sub_pixel_avg_variance8x16/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred"; - specialize qw/vpx_highbd_10_sub_pixel_avg_variance8x16/, "$sse2_x86inc"; + specialize qw/vpx_highbd_10_sub_pixel_avg_variance8x16 sse2/; add_proto qw/uint32_t vpx_highbd_10_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_highbd_10_sub_pixel_avg_variance8x8/, "$sse2_x86inc"; + specialize qw/vpx_highbd_10_sub_pixel_avg_variance8x8 sse2/; add_proto qw/uint32_t vpx_highbd_10_sub_pixel_avg_variance8x4/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred"; - specialize qw/vpx_highbd_10_sub_pixel_avg_variance8x4/, "$sse2_x86inc"; + specialize qw/vpx_highbd_10_sub_pixel_avg_variance8x4 sse2/; add_proto qw/uint32_t vpx_highbd_10_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"; add_proto qw/uint32_t vpx_highbd_10_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"; add_proto qw/uint32_t vpx_highbd_8_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_highbd_8_sub_pixel_avg_variance64x64/, "$sse2_x86inc"; + specialize qw/vpx_highbd_8_sub_pixel_avg_variance64x64 sse2/; add_proto qw/uint32_t vpx_highbd_8_sub_pixel_avg_variance64x32/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred"; - specialize qw/vpx_highbd_8_sub_pixel_avg_variance64x32/, "$sse2_x86inc"; + specialize qw/vpx_highbd_8_sub_pixel_avg_variance64x32 sse2/; add_proto qw/uint32_t vpx_highbd_8_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_highbd_8_sub_pixel_avg_variance32x64/, "$sse2_x86inc"; + specialize qw/vpx_highbd_8_sub_pixel_avg_variance32x64 sse2/; add_proto qw/uint32_t vpx_highbd_8_sub_pixel_avg_variance32x32/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred"; - specialize qw/vpx_highbd_8_sub_pixel_avg_variance32x32/, "$sse2_x86inc"; + specialize qw/vpx_highbd_8_sub_pixel_avg_variance32x32 sse2/; add_proto qw/uint32_t vpx_highbd_8_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_highbd_8_sub_pixel_avg_variance32x16/, "$sse2_x86inc"; + specialize qw/vpx_highbd_8_sub_pixel_avg_variance32x16 sse2/; add_proto qw/uint32_t vpx_highbd_8_sub_pixel_avg_variance16x32/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred"; - specialize qw/vpx_highbd_8_sub_pixel_avg_variance16x32/, "$sse2_x86inc"; + specialize qw/vpx_highbd_8_sub_pixel_avg_variance16x32 sse2/; add_proto qw/uint32_t vpx_highbd_8_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_highbd_8_sub_pixel_avg_variance16x16/, "$sse2_x86inc"; + specialize qw/vpx_highbd_8_sub_pixel_avg_variance16x16 sse2/; add_proto qw/uint32_t vpx_highbd_8_sub_pixel_avg_variance16x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred"; - specialize qw/vpx_highbd_8_sub_pixel_avg_variance16x8/, "$sse2_x86inc"; + specialize qw/vpx_highbd_8_sub_pixel_avg_variance16x8 sse2/; add_proto qw/uint32_t vpx_highbd_8_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_highbd_8_sub_pixel_avg_variance8x16/, "$sse2_x86inc"; + specialize qw/vpx_highbd_8_sub_pixel_avg_variance8x16 sse2/; add_proto qw/uint32_t vpx_highbd_8_sub_pixel_avg_variance8x8/, "const uint8_t *src_ptr, int source_stride, int xoffset, int yoffset, const uint8_t *ref_ptr, int ref_stride, uint32_t *sse, const uint8_t *second_pred"; - specialize qw/vpx_highbd_8_sub_pixel_avg_variance8x8/, "$sse2_x86inc"; + specialize qw/vpx_highbd_8_sub_pixel_avg_variance8x8 sse2/; add_proto qw/uint32_t vpx_highbd_8_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_highbd_8_sub_pixel_avg_variance8x4/, "$sse2_x86inc"; + specialize qw/vpx_highbd_8_sub_pixel_avg_variance8x4 sse2/; add_proto qw/uint32_t vpx_highbd_8_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"; add_proto qw/uint32_t vpx_highbd_8_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"; diff --git a/vpx_dsp/x86/highbd_variance_sse2.c b/vpx_dsp/x86/highbd_variance_sse2.c index f9ebaf7d3..0d62adf8b 100644 --- a/vpx_dsp/x86/highbd_variance_sse2.c +++ b/vpx_dsp/x86/highbd_variance_sse2.c @@ -246,7 +246,6 @@ unsigned int vpx_highbd_12_mse8x8_sse2(const uint8_t *src8, int src_stride, return *sse; } -#if CONFIG_USE_X86INC // The 2 unused parameters are place holders for PIC enabled build. // These definitions are for functions defined in // highbd_subpel_variance_impl_sse2.asm @@ -593,4 +592,3 @@ FNS(sse2); #undef FNS #undef FN -#endif // CONFIG_USE_X86INC diff --git a/vpx_dsp/x86/variance_sse2.c b/vpx_dsp/x86/variance_sse2.c index 6987c2e24..92dc752c0 100644 --- a/vpx_dsp/x86/variance_sse2.c +++ b/vpx_dsp/x86/variance_sse2.c @@ -308,7 +308,6 @@ unsigned int vpx_mse16x16_sse2(const uint8_t *src, int src_stride, return *sse; } -#if CONFIG_USE_X86INC // The 2 unused parameters are place holders for PIC enabled build. // These definitions are for functions defined in subpel_variance.asm #define DECL(w, opt) \ @@ -474,4 +473,3 @@ FNS(ssse3, ssse3); #undef FNS #undef FN -#endif // CONFIG_USE_X86INC