| OLD | NEW |
| 1 /* | 1 /* |
| 2 * Copyright (c) 2010 The WebM project authors. All Rights Reserved. | 2 * Copyright (c) 2010 The WebM project authors. All Rights Reserved. |
| 3 * | 3 * |
| 4 * Use of this source code is governed by a BSD-style license | 4 * Use of this source code is governed by a BSD-style license |
| 5 * that can be found in the LICENSE file in the root of the source | 5 * that can be found in the LICENSE file in the root of the source |
| 6 * tree. An additional intellectual property rights grant can be found | 6 * tree. An additional intellectual property rights grant can be found |
| 7 * in the file PATENTS. All contributing project authors may | 7 * in the file PATENTS. All contributing project authors may |
| 8 * be found in the AUTHORS file in the root of the source tree. | 8 * be found in the AUTHORS file in the root of the source tree. |
| 9 */ | 9 */ |
| 10 | 10 |
| 11 | 11 #include "vpx_config.h" |
| 12 #include "vp8/encoder/variance.h" | 12 #include "vp8/encoder/variance.h" |
| 13 #include "vp8/common/pragmas.h" | 13 #include "vp8/common/pragmas.h" |
| 14 #include "vpx_ports/mem.h" | 14 #include "vpx_ports/mem.h" |
| 15 | 15 |
| 16 extern void filter_block1d_h6_mmx | 16 extern void filter_block1d_h6_mmx |
| 17 ( | 17 ( |
| 18 const unsigned char *src_ptr, | 18 const unsigned char *src_ptr, |
| 19 unsigned short *output_ptr, | 19 unsigned short *output_ptr, |
| 20 unsigned int src_pixels_per_line, | 20 unsigned int src_pixels_per_line, |
| 21 unsigned int pixel_step, | 21 unsigned int pixel_step, |
| 22 unsigned int output_height, | 22 unsigned int output_height, |
| 23 unsigned int output_width, | 23 unsigned int output_width, |
| 24 short *vp7_filter | 24 short *vp7_filter |
| 25 ); | 25 ); |
| 26 extern void filter_block1d_v6_mmx | 26 extern void filter_block1d_v6_mmx |
| 27 ( | 27 ( |
| 28 const short *src_ptr, | 28 const short *src_ptr, |
| 29 unsigned char *output_ptr, | 29 unsigned char *output_ptr, |
| 30 unsigned int pixels_per_line, | 30 unsigned int pixels_per_line, |
| 31 unsigned int pixel_step, | 31 unsigned int pixel_step, |
| 32 unsigned int output_height, | 32 unsigned int output_height, |
| 33 unsigned int output_width, | 33 unsigned int output_width, |
| 34 short *vp7_filter | 34 short *vp7_filter |
| 35 ); | 35 ); |
| 36 | 36 |
| 37 extern unsigned int vp8_get_mb_ss_mmx(short *src_ptr); | 37 extern unsigned int vp8_get_mb_ss_mmx(const short *src_ptr); |
| 38 extern unsigned int vp8_get8x8var_mmx | 38 extern unsigned int vp8_get8x8var_mmx |
| 39 ( | 39 ( |
| 40 const unsigned char *src_ptr, | 40 const unsigned char *src_ptr, |
| 41 int source_stride, | 41 int source_stride, |
| 42 const unsigned char *ref_ptr, | 42 const unsigned char *ref_ptr, |
| 43 int recon_stride, | 43 int recon_stride, |
| 44 unsigned int *SSE, | 44 unsigned int *SSE, |
| 45 int *Sum | 45 int *Sum |
| 46 ); | 46 ); |
| 47 extern unsigned int vp8_get4x4var_mmx | 47 extern unsigned int vp8_get4x4var_mmx |
| (...skipping 21 matching lines...) Expand all Loading... |
| 69 const unsigned char *ref_ptr, | 69 const unsigned char *ref_ptr, |
| 70 int ref_pixels_per_line, | 70 int ref_pixels_per_line, |
| 71 const unsigned char *src_ptr, | 71 const unsigned char *src_ptr, |
| 72 int src_pixels_per_line, | 72 int src_pixels_per_line, |
| 73 unsigned int Height, | 73 unsigned int Height, |
| 74 const short *HFilter, | 74 const short *HFilter, |
| 75 const short *VFilter, | 75 const short *VFilter, |
| 76 int *sum, | 76 int *sum, |
| 77 unsigned int *sumsquared | 77 unsigned int *sumsquared |
| 78 ); | 78 ); |
| 79 extern unsigned int vp8_get16x16pred_error_mmx | |
| 80 ( | |
| 81 unsigned char *src_ptr, | |
| 82 int src_stride, | |
| 83 unsigned char *ref_ptr, | |
| 84 int ref_stride | |
| 85 ); | |
| 86 | |
| 87 | |
| 88 unsigned int vp8_get16x16var_mmx( | |
| 89 const unsigned char *src_ptr, | |
| 90 int source_stride, | |
| 91 const unsigned char *ref_ptr, | |
| 92 int recon_stride, | |
| 93 unsigned *SSE, | |
| 94 unsigned *SUM | |
| 95 ) | |
| 96 { | |
| 97 unsigned int sse0, sse1, sse2, sse3, var; | |
| 98 int sum0, sum1, sum2, sum3, avg; | |
| 99 | |
| 100 | |
| 101 vp8_get8x8var_mmx(src_ptr, source_stride, ref_ptr, recon_stride, &sse0, &sum
0) ; | |
| 102 vp8_get8x8var_mmx(src_ptr + 8, source_stride, ref_ptr + 8, recon_stride, &ss
e1, &sum1); | |
| 103 vp8_get8x8var_mmx(src_ptr + 8 * source_stride, source_stride, ref_ptr + 8 *
recon_stride, recon_stride, &sse2, &sum2) ; | |
| 104 vp8_get8x8var_mmx(src_ptr + 8 * source_stride + 8, source_stride, ref_ptr +
8 * recon_stride + 8, recon_stride, &sse3, &sum3); | |
| 105 | |
| 106 var = sse0 + sse1 + sse2 + sse3; | |
| 107 avg = sum0 + sum1 + sum2 + sum3; | |
| 108 | |
| 109 *SSE = var; | |
| 110 *SUM = avg; | |
| 111 return (var - ((avg * avg) >> 8)); | |
| 112 | |
| 113 } | |
| 114 | |
| 115 | |
| 116 | |
| 117 | 79 |
| 118 | 80 |
| 119 unsigned int vp8_variance4x4_mmx( | 81 unsigned int vp8_variance4x4_mmx( |
| 120 const unsigned char *src_ptr, | 82 const unsigned char *src_ptr, |
| 121 int source_stride, | 83 int source_stride, |
| 122 const unsigned char *ref_ptr, | 84 const unsigned char *ref_ptr, |
| 123 int recon_stride, | 85 int recon_stride, |
| 124 unsigned int *sse) | 86 unsigned int *sse) |
| 125 { | 87 { |
| 126 unsigned int var; | 88 unsigned int var; |
| (...skipping 42 matching lines...) Expand 10 before | Expand all | Expand 10 after Loading... |
| 169 *sse = var; | 131 *sse = var; |
| 170 return var; | 132 return var; |
| 171 } | 133 } |
| 172 | 134 |
| 173 | 135 |
| 174 unsigned int vp8_variance16x16_mmx( | 136 unsigned int vp8_variance16x16_mmx( |
| 175 const unsigned char *src_ptr, | 137 const unsigned char *src_ptr, |
| 176 int source_stride, | 138 int source_stride, |
| 177 const unsigned char *ref_ptr, | 139 const unsigned char *ref_ptr, |
| 178 int recon_stride, | 140 int recon_stride, |
| 179 int *sse) | 141 unsigned int *sse) |
| 180 { | 142 { |
| 181 unsigned int sse0, sse1, sse2, sse3, var; | 143 unsigned int sse0, sse1, sse2, sse3, var; |
| 182 int sum0, sum1, sum2, sum3, avg; | 144 int sum0, sum1, sum2, sum3, avg; |
| 183 | 145 |
| 184 | 146 |
| 185 vp8_get8x8var_mmx(src_ptr, source_stride, ref_ptr, recon_stride, &sse0, &sum
0) ; | 147 vp8_get8x8var_mmx(src_ptr, source_stride, ref_ptr, recon_stride, &sse0, &sum
0) ; |
| 186 vp8_get8x8var_mmx(src_ptr + 8, source_stride, ref_ptr + 8, recon_stride, &ss
e1, &sum1); | 148 vp8_get8x8var_mmx(src_ptr + 8, source_stride, ref_ptr + 8, recon_stride, &ss
e1, &sum1); |
| 187 vp8_get8x8var_mmx(src_ptr + 8 * source_stride, source_stride, ref_ptr + 8 *
recon_stride, recon_stride, &sse2, &sum2) ; | 149 vp8_get8x8var_mmx(src_ptr + 8 * source_stride, source_stride, ref_ptr + 8 *
recon_stride, recon_stride, &sse2, &sum2) ; |
| 188 vp8_get8x8var_mmx(src_ptr + 8 * source_stride + 8, source_stride, ref_ptr +
8 * recon_stride + 8, recon_stride, &sse3, &sum3); | 150 vp8_get8x8var_mmx(src_ptr + 8 * source_stride + 8, source_stride, ref_ptr +
8 * recon_stride + 8, recon_stride, &sse3, &sum3); |
| 189 | 151 |
| (...skipping 204 matching lines...) Expand 10 before | Expand all | Expand 10 after Loading... |
| 394 } | 356 } |
| 395 | 357 |
| 396 unsigned int vp8_sub_pixel_variance8x16_mmx | 358 unsigned int vp8_sub_pixel_variance8x16_mmx |
| 397 ( | 359 ( |
| 398 const unsigned char *src_ptr, | 360 const unsigned char *src_ptr, |
| 399 int src_pixels_per_line, | 361 int src_pixels_per_line, |
| 400 int xoffset, | 362 int xoffset, |
| 401 int yoffset, | 363 int yoffset, |
| 402 const unsigned char *dst_ptr, | 364 const unsigned char *dst_ptr, |
| 403 int dst_pixels_per_line, | 365 int dst_pixels_per_line, |
| 404 int *sse | 366 unsigned int *sse |
| 405 ) | 367 ) |
| 406 { | 368 { |
| 407 int xsum; | 369 int xsum; |
| 408 unsigned int xxsum; | 370 unsigned int xxsum; |
| 409 vp8_filter_block2d_bil_var_mmx( | 371 vp8_filter_block2d_bil_var_mmx( |
| 410 src_ptr, src_pixels_per_line, | 372 src_ptr, src_pixels_per_line, |
| 411 dst_ptr, dst_pixels_per_line, 16, | 373 dst_ptr, dst_pixels_per_line, 16, |
| 412 vp8_vp7_bilinear_filters_mmx[xoffset], vp8_vp7_bilinear_filters_mmx[yoff
set], | 374 vp8_vp7_bilinear_filters_mmx[xoffset], vp8_vp7_bilinear_filters_mmx[yoff
set], |
| 413 &xsum, &xxsum | 375 &xsum, &xxsum |
| 414 ); | 376 ); |
| (...skipping 29 matching lines...) Expand all Loading... |
| 444 unsigned int vp8_variance_halfpixvar16x16_hv_mmx( | 406 unsigned int vp8_variance_halfpixvar16x16_hv_mmx( |
| 445 const unsigned char *src_ptr, | 407 const unsigned char *src_ptr, |
| 446 int source_stride, | 408 int source_stride, |
| 447 const unsigned char *ref_ptr, | 409 const unsigned char *ref_ptr, |
| 448 int recon_stride, | 410 int recon_stride, |
| 449 unsigned int *sse) | 411 unsigned int *sse) |
| 450 { | 412 { |
| 451 return vp8_sub_pixel_variance16x16_mmx(src_ptr, source_stride, 4, 4, | 413 return vp8_sub_pixel_variance16x16_mmx(src_ptr, source_stride, 4, 4, |
| 452 ref_ptr, recon_stride, sse); | 414 ref_ptr, recon_stride, sse); |
| 453 } | 415 } |
| OLD | NEW |