ref: 806d05e1a81b3a04bdab7af08fb032f57b815886
dir: /vp9/common/x86/vp9_filter_sse4.c/
/* * Copyright (c) 2012 The WebM project authors. All Rights Reserved. * * Use of this source code is governed by a BSD-style license * that can be found in the LICENSE file in the root of the source * tree. An additional intellectual property rights grant can be found * in the file PATENTS. All contributing project authors may * be found in the AUTHORS file in the root of the source tree. */ #include <assert.h> // for alignment checks #include <smmintrin.h> // SSE4.1 #include "vp9/common/vp9_filter.h" #include "vpx_ports/mem.h" // for DECLARE_ALIGNED #include "vp9_rtcd.h" // TODO(cd): After cleanup, commit faster versions for non 4x4 size. This is // just a quick partial snapshot so that other can already use some // speedup. // TODO(cd): Use vectorized 8 tap filtering code as speedup to pure C 6 tap // filtering. // TODO(cd): Reduce source size by using macros instead of current code // duplication. // TODO(cd): Add some comments, better variable naming. // TODO(cd): Maybe use _mm_maddubs_epi16 if smaller filter coeficients (no sum // of positive above 128), or have higher precision filter // coefficients. DECLARE_ALIGNED(16, static const unsigned char, mask0123_c[16]) = { 0x00, 0x01, 0x01, 0x02, 0x02, 0x03, 0x03, 0x04, 0x02, 0x03, 0x03, 0x04, 0x04, 0x05, 0x05, 0x06, }; DECLARE_ALIGNED(16, static const unsigned char, mask4567_c[16]) = { 0x04, 0x05, 0x05, 0x06, 0x06, 0x07, 0x07, 0x08, 0x06, 0x07, 0x07, 0x08, 0x08, 0x09, 0x09, 0x0A, }; DECLARE_ALIGNED(16, static const unsigned int, rounding_c[4]) = { VP9_FILTER_WEIGHT >> 1, VP9_FILTER_WEIGHT >> 1, VP9_FILTER_WEIGHT >> 1, VP9_FILTER_WEIGHT >> 1, }; DECLARE_ALIGNED(16, static const unsigned char, transpose_c[16]) = { 0, 4, 8, 12, 1, 5, 9, 13, 2, 6, 10, 14, 3, 7, 11, 15 }; // Creating a macro to do more than four pixels at once to hide instruction // latency is actually slower :-( #define DO_FOUR_PIXELS(result, offset) \ { \ /*load pixels*/ \ __m128i src = _mm_loadu_si128((const __m128i *)(src_ptr + offset)); \ /* extract the ones used for first column */ \ __m128i src0123 = _mm_shuffle_epi8(src, mask0123); \ __m128i src4567 = _mm_shuffle_epi8(src, mask4567); \ __m128i src01_16 = _mm_unpacklo_epi8(src0123, zero); \ __m128i src23_16 = _mm_unpackhi_epi8(src0123, zero); \ __m128i src45_16 = _mm_unpacklo_epi8(src4567, zero); \ __m128i src67_16 = _mm_unpackhi_epi8(src4567, zero); \ /* multiply accumulate them */ \ __m128i mad01 = _mm_madd_epi16(src01_16, fil01); \ __m128i mad23 = _mm_madd_epi16(src23_16, fil23); \ __m128i mad45 = _mm_madd_epi16(src45_16, fil45); \ __m128i mad67 = _mm_madd_epi16(src67_16, fil67); \ __m128i mad0123 = _mm_add_epi32(mad01, mad23); \ __m128i mad4567 = _mm_add_epi32(mad45, mad67); \ __m128i mad_all = _mm_add_epi32(mad0123, mad4567); \ mad_all = _mm_add_epi32(mad_all, rounding); \ result = _mm_srai_epi32(mad_all, VP9_FILTER_SHIFT); \ } void vp9_filter_block2d_4x4_8_sse4_1 ( const unsigned char *src_ptr, const unsigned int src_stride, const short *HFilter_aligned16, const short *VFilter_aligned16, unsigned char *dst_ptr, unsigned int dst_stride ) { __m128i intermediateA, intermediateB, intermediateC; const int kInterp_Extend = 4; const __m128i zero = _mm_set1_epi16(0); const __m128i mask0123 = _mm_load_si128((const __m128i *)mask0123_c); const __m128i mask4567 = _mm_load_si128((const __m128i *)mask4567_c); const __m128i rounding = _mm_load_si128((const __m128i *)rounding_c); const __m128i transpose = _mm_load_si128((const __m128i *)transpose_c); // check alignment assert(0 == ((long)HFilter_aligned16)%16); assert(0 == ((long)VFilter_aligned16)%16); { __m128i transpose3_0; __m128i transpose3_1; __m128i transpose3_2; __m128i transpose3_3; // Horizontal pass (src -> intermediate). { const __m128i HFilter = _mm_load_si128((const __m128i *)HFilter_aligned16); // get first two columns filter coefficients __m128i fil01 = _mm_shuffle_epi32(HFilter, _MM_SHUFFLE(0, 0, 0, 0)); __m128i fil23 = _mm_shuffle_epi32(HFilter, _MM_SHUFFLE(1, 1, 1, 1)); __m128i fil45 = _mm_shuffle_epi32(HFilter, _MM_SHUFFLE(2, 2, 2, 2)); __m128i fil67 = _mm_shuffle_epi32(HFilter, _MM_SHUFFLE(3, 3, 3, 3)); src_ptr -= (kInterp_Extend - 1) * src_stride + (kInterp_Extend - 1); { __m128i mad_all0; __m128i mad_all1; __m128i mad_all2; __m128i mad_all3; DO_FOUR_PIXELS(mad_all0, 0*src_stride) DO_FOUR_PIXELS(mad_all1, 1*src_stride) DO_FOUR_PIXELS(mad_all2, 2*src_stride) DO_FOUR_PIXELS(mad_all3, 3*src_stride) mad_all0 = _mm_packs_epi32(mad_all0, mad_all1); mad_all2 = _mm_packs_epi32(mad_all2, mad_all3); intermediateA = _mm_packus_epi16(mad_all0, mad_all2); // -- src_ptr += src_stride*4; // -- DO_FOUR_PIXELS(mad_all0, 0*src_stride) DO_FOUR_PIXELS(mad_all1, 1*src_stride) DO_FOUR_PIXELS(mad_all2, 2*src_stride) DO_FOUR_PIXELS(mad_all3, 3*src_stride) mad_all0 = _mm_packs_epi32(mad_all0, mad_all1); mad_all2 = _mm_packs_epi32(mad_all2, mad_all3); intermediateB = _mm_packus_epi16(mad_all0, mad_all2); // -- src_ptr += src_stride*4; // -- DO_FOUR_PIXELS(mad_all0, 0*src_stride) DO_FOUR_PIXELS(mad_all1, 1*src_stride) DO_FOUR_PIXELS(mad_all2, 2*src_stride) mad_all0 = _mm_packs_epi32(mad_all0, mad_all1); mad_all2 = _mm_packs_epi32(mad_all2, mad_all2); intermediateC = _mm_packus_epi16(mad_all0, mad_all2); } } // Transpose result (intermediate -> transpose3_x) { // 00 01 02 03 10 11 12 13 20 21 22 23 30 31 32 33 // 40 41 42 43 50 51 52 53 60 61 62 63 70 71 72 73 // 80 81 82 83 90 91 92 93 A0 A1 A2 A3 xx xx xx xx const __m128i transpose1_0 = _mm_shuffle_epi8(intermediateA, transpose); const __m128i transpose1_1 = _mm_shuffle_epi8(intermediateB, transpose); const __m128i transpose1_2 = _mm_shuffle_epi8(intermediateC, transpose); // 00 10 20 30 01 11 21 31 02 12 22 32 03 13 23 33 // 40 50 60 70 41 51 61 71 42 52 62 72 43 53 63 73 // 80 90 A0 xx 81 91 A1 xx 82 92 A2 xx 83 93 A3 xx const __m128i transpose2_0 = _mm_unpacklo_epi32(transpose1_0, transpose1_1); const __m128i transpose2_1 = _mm_unpackhi_epi32(transpose1_0, transpose1_1); // 00 10 20 30 40 50 60 70 01 11 21 31 41 51 61 71 // 02 12 22 32 42 52 62 72 03 13 23 33 43 53 63 73 transpose3_0 = _mm_castps_si128( _mm_shuffle_ps(_mm_castsi128_ps(transpose2_0), _mm_castsi128_ps(transpose1_2), _MM_SHUFFLE(0, 0, 1, 0))); transpose3_1 = _mm_castps_si128( _mm_shuffle_ps(_mm_castsi128_ps(transpose2_0), _mm_castsi128_ps(transpose1_2), _MM_SHUFFLE(1, 1, 3, 2))); transpose3_2 = _mm_castps_si128( _mm_shuffle_ps(_mm_castsi128_ps(transpose2_1), _mm_castsi128_ps(transpose1_2), _MM_SHUFFLE(2, 2, 1, 0))); transpose3_3 = _mm_castps_si128( _mm_shuffle_ps(_mm_castsi128_ps(transpose2_1), _mm_castsi128_ps(transpose1_2), _MM_SHUFFLE(3, 3, 3, 2))); // 00 10 20 30 40 50 60 70 80 90 A0 xx xx xx xx xx // 01 11 21 31 41 51 61 71 81 91 A1 xx xx xx xx xx // 02 12 22 32 42 52 62 72 82 92 A2 xx xx xx xx xx // 03 13 23 33 43 53 63 73 83 93 A3 xx xx xx xx xx } // Vertical pass (transpose3_x -> dst). { const __m128i VFilter = _mm_load_si128((const __m128i *)VFilter_aligned16); // get first two columns filter coefficients __m128i fil01 = _mm_shuffle_epi32(VFilter, _MM_SHUFFLE(0, 0, 0, 0)); __m128i fil23 = _mm_shuffle_epi32(VFilter, _MM_SHUFFLE(1, 1, 1, 1)); __m128i fil45 = _mm_shuffle_epi32(VFilter, _MM_SHUFFLE(2, 2, 2, 2)); __m128i fil67 = _mm_shuffle_epi32(VFilter, _MM_SHUFFLE(3, 3, 3, 3)); __m128i col0, col1, col2, col3; { //load pixels __m128i src = transpose3_0; // extract the ones used for first column __m128i src0123 = _mm_shuffle_epi8(src, mask0123); __m128i src4567 = _mm_shuffle_epi8(src, mask4567); __m128i src01_16 = _mm_unpacklo_epi8(src0123, zero); __m128i src23_16 = _mm_unpackhi_epi8(src0123, zero); __m128i src45_16 = _mm_unpacklo_epi8(src4567, zero); __m128i src67_16 = _mm_unpackhi_epi8(src4567, zero); // multiply accumulate them __m128i mad01 = _mm_madd_epi16(src01_16, fil01); __m128i mad23 = _mm_madd_epi16(src23_16, fil23); __m128i mad45 = _mm_madd_epi16(src45_16, fil45); __m128i mad67 = _mm_madd_epi16(src67_16, fil67); __m128i mad0123 = _mm_add_epi32(mad01, mad23); __m128i mad4567 = _mm_add_epi32(mad45, mad67); __m128i mad_all = _mm_add_epi32(mad0123, mad4567); mad_all = _mm_add_epi32(mad_all, rounding); mad_all = _mm_srai_epi32(mad_all, VP9_FILTER_SHIFT); mad_all = _mm_packs_epi32(mad_all, mad_all); col0 = _mm_packus_epi16(mad_all, mad_all); } { //load pixels __m128i src = transpose3_1; // extract the ones used for first column __m128i src0123 = _mm_shuffle_epi8(src, mask0123); __m128i src4567 = _mm_shuffle_epi8(src, mask4567); __m128i src01_16 = _mm_unpacklo_epi8(src0123, zero); __m128i src23_16 = _mm_unpackhi_epi8(src0123, zero); __m128i src45_16 = _mm_unpacklo_epi8(src4567, zero); __m128i src67_16 = _mm_unpackhi_epi8(src4567, zero); // multiply accumulate them __m128i mad01 = _mm_madd_epi16(src01_16, fil01); __m128i mad23 = _mm_madd_epi16(src23_16, fil23); __m128i mad45 = _mm_madd_epi16(src45_16, fil45); __m128i mad67 = _mm_madd_epi16(src67_16, fil67); __m128i mad0123 = _mm_add_epi32(mad01, mad23); __m128i mad4567 = _mm_add_epi32(mad45, mad67); __m128i mad_all = _mm_add_epi32(mad0123, mad4567); mad_all = _mm_add_epi32(mad_all, rounding); mad_all = _mm_srai_epi32(mad_all, VP9_FILTER_SHIFT); mad_all = _mm_packs_epi32(mad_all, mad_all); col1 = _mm_packus_epi16(mad_all, mad_all); } { //load pixels __m128i src = transpose3_2; // extract the ones used for first column __m128i src0123 = _mm_shuffle_epi8(src, mask0123); __m128i src4567 = _mm_shuffle_epi8(src, mask4567); __m128i src01_16 = _mm_unpacklo_epi8(src0123, zero); __m128i src23_16 = _mm_unpackhi_epi8(src0123, zero); __m128i src45_16 = _mm_unpacklo_epi8(src4567, zero); __m128i src67_16 = _mm_unpackhi_epi8(src4567, zero); // multiply accumulate them __m128i mad01 = _mm_madd_epi16(src01_16, fil01); __m128i mad23 = _mm_madd_epi16(src23_16, fil23); __m128i mad45 = _mm_madd_epi16(src45_16, fil45); __m128i mad67 = _mm_madd_epi16(src67_16, fil67); __m128i mad0123 = _mm_add_epi32(mad01, mad23); __m128i mad4567 = _mm_add_epi32(mad45, mad67); __m128i mad_all = _mm_add_epi32(mad0123, mad4567); mad_all = _mm_add_epi32(mad_all, rounding); mad_all = _mm_srai_epi32(mad_all, VP9_FILTER_SHIFT); mad_all = _mm_packs_epi32(mad_all, mad_all); col2 = _mm_packus_epi16(mad_all, mad_all); } { //load pixels __m128i src = transpose3_3; // extract the ones used for first column __m128i src0123 = _mm_shuffle_epi8(src, mask0123); __m128i src4567 = _mm_shuffle_epi8(src, mask4567); __m128i src01_16 = _mm_unpacklo_epi8(src0123, zero); __m128i src23_16 = _mm_unpackhi_epi8(src0123, zero); __m128i src45_16 = _mm_unpacklo_epi8(src4567, zero); __m128i src67_16 = _mm_unpackhi_epi8(src4567, zero); // multiply accumulate them __m128i mad01 = _mm_madd_epi16(src01_16, fil01); __m128i mad23 = _mm_madd_epi16(src23_16, fil23); __m128i mad45 = _mm_madd_epi16(src45_16, fil45); __m128i mad67 = _mm_madd_epi16(src67_16, fil67); __m128i mad0123 = _mm_add_epi32(mad01, mad23); __m128i mad4567 = _mm_add_epi32(mad45, mad67); __m128i mad_all = _mm_add_epi32(mad0123, mad4567); mad_all = _mm_add_epi32(mad_all, rounding); mad_all = _mm_srai_epi32(mad_all, VP9_FILTER_SHIFT); mad_all = _mm_packs_epi32(mad_all, mad_all); col3 = _mm_packus_epi16(mad_all, mad_all); } { __m128i col01 = _mm_unpacklo_epi8(col0, col1); __m128i col23 = _mm_unpacklo_epi8(col2, col3); __m128i col0123 = _mm_unpacklo_epi16(col01, col23); //TODO(cd): look into Ronald's comment: // Future suggestion: I believe here, too, you can merge the // packs_epi32() and pacus_epi16() for the 4 cols above, so that // you get the data in a single register, and then use pshufb // (shuffle_epi8()) instead of the unpacks here. Should be // 2+3+2 instructions faster. *((unsigned int *)&dst_ptr[dst_stride * 0]) = _mm_extract_epi32(col0123, 0); *((unsigned int *)&dst_ptr[dst_stride * 1]) = _mm_extract_epi32(col0123, 1); *((unsigned int *)&dst_ptr[dst_stride * 2]) = _mm_extract_epi32(col0123, 2); *((unsigned int *)&dst_ptr[dst_stride * 3]) = _mm_extract_epi32(col0123, 3); } } } } void vp9_filter_block2d_8x4_8_sse4_1 ( const unsigned char *src_ptr, const unsigned int src_stride, const short *HFilter_aligned16, const short *VFilter_aligned16, unsigned char *dst_ptr, unsigned int dst_stride ) { int j; for (j=0; j<8; j+=4) { vp9_filter_block2d_4x4_8_sse4_1(src_ptr + j, src_stride, HFilter_aligned16, VFilter_aligned16, dst_ptr + j, dst_stride); } } void vp9_filter_block2d_8x8_8_sse4_1 ( const unsigned char *src_ptr, const unsigned int src_stride, const short *HFilter_aligned16, const short *VFilter_aligned16, unsigned char *dst_ptr, unsigned int dst_stride ) { int i, j; for (i=0; i<8; i+=4) { for (j=0; j<8; j+=4) { vp9_filter_block2d_4x4_8_sse4_1(src_ptr + j + i*src_stride, src_stride, HFilter_aligned16, VFilter_aligned16, dst_ptr + j + i*dst_stride, dst_stride); } } } void vp9_filter_block2d_16x16_8_sse4_1 ( const unsigned char *src_ptr, const unsigned int src_stride, const short *HFilter_aligned16, const short *VFilter_aligned16, unsigned char *dst_ptr, unsigned int dst_stride ) { int i, j; for (i=0; i<16; i+=4) { for (j=0; j<16; j+=4) { vp9_filter_block2d_4x4_8_sse4_1(src_ptr + j + i*src_stride, src_stride, HFilter_aligned16, VFilter_aligned16, dst_ptr + j + i*dst_stride, dst_stride); } } }