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 #include "./vp8_rtcd.h" |
11 #include "vpx_config.h" | 12 #include "vpx_config.h" |
12 #include "vp8/common/variance.h" | 13 #include "vp8/common/variance.h" |
13 #include "vpx_ports/mem.h" | 14 #include "vpx_ports/mem.h" |
14 #include "vp8/common/x86/filter_x86.h" | 15 #include "vp8/common/x86/filter_x86.h" |
15 | 16 |
16 extern void filter_block1d_h6_mmx | 17 extern void filter_block1d_h6_mmx |
17 ( | 18 ( |
18 const unsigned char *src_ptr, | 19 const unsigned char *src_ptr, |
19 unsigned short *output_ptr, | 20 unsigned short *output_ptr, |
20 unsigned int src_pixels_per_line, | 21 unsigned int src_pixels_per_line, |
21 unsigned int pixel_step, | 22 unsigned int pixel_step, |
22 unsigned int output_height, | 23 unsigned int output_height, |
23 unsigned int output_width, | 24 unsigned int output_width, |
24 short *filter | 25 short *filter |
25 ); | 26 ); |
26 extern void filter_block1d_v6_mmx | 27 extern void filter_block1d_v6_mmx |
27 ( | 28 ( |
28 const short *src_ptr, | 29 const short *src_ptr, |
29 unsigned char *output_ptr, | 30 unsigned char *output_ptr, |
30 unsigned int pixels_per_line, | 31 unsigned int pixels_per_line, |
31 unsigned int pixel_step, | 32 unsigned int pixel_step, |
32 unsigned int output_height, | 33 unsigned int output_height, |
33 unsigned int output_width, | 34 unsigned int output_width, |
34 short *filter | 35 short *filter |
35 ); | 36 ); |
36 | 37 |
37 extern unsigned int vp8_get_mb_ss_mmx(const short *src_ptr); | |
38 extern unsigned int vp8_get8x8var_mmx | |
39 ( | |
40 const unsigned char *src_ptr, | |
41 int source_stride, | |
42 const unsigned char *ref_ptr, | |
43 int recon_stride, | |
44 unsigned int *SSE, | |
45 int *Sum | |
46 ); | |
47 extern unsigned int vp8_get4x4var_mmx | |
48 ( | |
49 const unsigned char *src_ptr, | |
50 int source_stride, | |
51 const unsigned char *ref_ptr, | |
52 int recon_stride, | |
53 unsigned int *SSE, | |
54 int *Sum | |
55 ); | |
56 extern void vp8_filter_block2d_bil4x4_var_mmx | 38 extern void vp8_filter_block2d_bil4x4_var_mmx |
57 ( | 39 ( |
58 const unsigned char *ref_ptr, | 40 const unsigned char *ref_ptr, |
59 int ref_pixels_per_line, | 41 int ref_pixels_per_line, |
60 const unsigned char *src_ptr, | 42 const unsigned char *src_ptr, |
61 int src_pixels_per_line, | 43 int src_pixels_per_line, |
62 const short *HFilter, | 44 const short *HFilter, |
63 const short *VFilter, | 45 const short *VFilter, |
64 int *sum, | 46 int *sum, |
65 unsigned int *sumsquared | 47 unsigned int *sumsquared |
66 ); | 48 ); |
67 extern void vp8_filter_block2d_bil_var_mmx | 49 extern void vp8_filter_block2d_bil_var_mmx |
68 ( | 50 ( |
69 const unsigned char *ref_ptr, | 51 const unsigned char *ref_ptr, |
70 int ref_pixels_per_line, | 52 int ref_pixels_per_line, |
71 const unsigned char *src_ptr, | 53 const unsigned char *src_ptr, |
72 int src_pixels_per_line, | 54 int src_pixels_per_line, |
73 unsigned int Height, | 55 unsigned int Height, |
74 const short *HFilter, | 56 const short *HFilter, |
75 const short *VFilter, | 57 const short *VFilter, |
76 int *sum, | 58 int *sum, |
77 unsigned int *sumsquared | 59 unsigned int *sumsquared |
78 ); | 60 ); |
79 | 61 |
80 | |
81 unsigned int vp8_variance4x4_mmx( | |
82 const unsigned char *src_ptr, | |
83 int source_stride, | |
84 const unsigned char *ref_ptr, | |
85 int recon_stride, | |
86 unsigned int *sse) | |
87 { | |
88 unsigned int var; | |
89 int avg; | |
90 | |
91 vp8_get4x4var_mmx(src_ptr, source_stride, ref_ptr, recon_stride, &var, &avg)
; | |
92 *sse = var; | |
93 return (var - (((unsigned int)avg * avg) >> 4)); | |
94 | |
95 } | |
96 | |
97 unsigned int vp8_variance8x8_mmx( | |
98 const unsigned char *src_ptr, | |
99 int source_stride, | |
100 const unsigned char *ref_ptr, | |
101 int recon_stride, | |
102 unsigned int *sse) | |
103 { | |
104 unsigned int var; | |
105 int avg; | |
106 | |
107 vp8_get8x8var_mmx(src_ptr, source_stride, ref_ptr, recon_stride, &var, &avg)
; | |
108 *sse = var; | |
109 | |
110 return (var - (((unsigned int)avg * avg) >> 6)); | |
111 | |
112 } | |
113 | |
114 unsigned int vp8_mse16x16_mmx( | |
115 const unsigned char *src_ptr, | |
116 int source_stride, | |
117 const unsigned char *ref_ptr, | |
118 int recon_stride, | |
119 unsigned int *sse) | |
120 { | |
121 unsigned int sse0, sse1, sse2, sse3, var; | |
122 int sum0, sum1, sum2, sum3; | |
123 | |
124 | |
125 vp8_get8x8var_mmx(src_ptr, source_stride, ref_ptr, recon_stride, &sse0, &sum
0) ; | |
126 vp8_get8x8var_mmx(src_ptr + 8, source_stride, ref_ptr + 8, recon_stride, &ss
e1, &sum1); | |
127 vp8_get8x8var_mmx(src_ptr + 8 * source_stride, source_stride, ref_ptr + 8 *
recon_stride, recon_stride, &sse2, &sum2) ; | |
128 vp8_get8x8var_mmx(src_ptr + 8 * source_stride + 8, source_stride, ref_ptr +
8 * recon_stride + 8, recon_stride, &sse3, &sum3); | |
129 | |
130 var = sse0 + sse1 + sse2 + sse3; | |
131 *sse = var; | |
132 return var; | |
133 } | |
134 | |
135 | |
136 unsigned int vp8_variance16x16_mmx( | |
137 const unsigned char *src_ptr, | |
138 int source_stride, | |
139 const unsigned char *ref_ptr, | |
140 int recon_stride, | |
141 unsigned int *sse) | |
142 { | |
143 unsigned int sse0, sse1, sse2, sse3, var; | |
144 int sum0, sum1, sum2, sum3, avg; | |
145 | |
146 | |
147 vp8_get8x8var_mmx(src_ptr, source_stride, ref_ptr, recon_stride, &sse0, &sum
0) ; | |
148 vp8_get8x8var_mmx(src_ptr + 8, source_stride, ref_ptr + 8, recon_stride, &ss
e1, &sum1); | |
149 vp8_get8x8var_mmx(src_ptr + 8 * source_stride, source_stride, ref_ptr + 8 *
recon_stride, recon_stride, &sse2, &sum2) ; | |
150 vp8_get8x8var_mmx(src_ptr + 8 * source_stride + 8, source_stride, ref_ptr +
8 * recon_stride + 8, recon_stride, &sse3, &sum3); | |
151 | |
152 var = sse0 + sse1 + sse2 + sse3; | |
153 avg = sum0 + sum1 + sum2 + sum3; | |
154 *sse = var; | |
155 return (var - (((unsigned int)avg * avg) >> 8)); | |
156 } | |
157 | |
158 unsigned int vp8_variance16x8_mmx( | |
159 const unsigned char *src_ptr, | |
160 int source_stride, | |
161 const unsigned char *ref_ptr, | |
162 int recon_stride, | |
163 unsigned int *sse) | |
164 { | |
165 unsigned int sse0, sse1, var; | |
166 int sum0, sum1, avg; | |
167 | |
168 vp8_get8x8var_mmx(src_ptr, source_stride, ref_ptr, recon_stride, &sse0, &sum
0) ; | |
169 vp8_get8x8var_mmx(src_ptr + 8, source_stride, ref_ptr + 8, recon_stride, &ss
e1, &sum1); | |
170 | |
171 var = sse0 + sse1; | |
172 avg = sum0 + sum1; | |
173 *sse = var; | |
174 return (var - (((unsigned int)avg * avg) >> 7)); | |
175 | |
176 } | |
177 | |
178 | |
179 unsigned int vp8_variance8x16_mmx( | |
180 const unsigned char *src_ptr, | |
181 int source_stride, | |
182 const unsigned char *ref_ptr, | |
183 int recon_stride, | |
184 unsigned int *sse) | |
185 { | |
186 unsigned int sse0, sse1, var; | |
187 int sum0, sum1, avg; | |
188 | |
189 vp8_get8x8var_mmx(src_ptr, source_stride, ref_ptr, recon_stride, &sse0, &sum
0) ; | |
190 vp8_get8x8var_mmx(src_ptr + 8 * source_stride, source_stride, ref_ptr + 8 *
recon_stride, recon_stride, &sse1, &sum1) ; | |
191 | |
192 var = sse0 + sse1; | |
193 avg = sum0 + sum1; | |
194 *sse = var; | |
195 | |
196 return (var - (((unsigned int)avg * avg) >> 7)); | |
197 | |
198 } | |
199 | |
200 | |
201 unsigned int vp8_sub_pixel_variance4x4_mmx | 62 unsigned int vp8_sub_pixel_variance4x4_mmx |
202 ( | 63 ( |
203 const unsigned char *src_ptr, | 64 const unsigned char *src_ptr, |
204 int src_pixels_per_line, | 65 int src_pixels_per_line, |
205 int xoffset, | 66 int xoffset, |
206 int yoffset, | 67 int yoffset, |
207 const unsigned char *dst_ptr, | 68 const unsigned char *dst_ptr, |
208 int dst_pixels_per_line, | 69 int dst_pixels_per_line, |
209 unsigned int *sse) | 70 unsigned int *sse) |
210 | 71 |
(...skipping 177 matching lines...) Expand 10 before | Expand all | Expand 10 after Loading... |
388 unsigned int vp8_variance_halfpixvar16x16_hv_mmx( | 249 unsigned int vp8_variance_halfpixvar16x16_hv_mmx( |
389 const unsigned char *src_ptr, | 250 const unsigned char *src_ptr, |
390 int source_stride, | 251 int source_stride, |
391 const unsigned char *ref_ptr, | 252 const unsigned char *ref_ptr, |
392 int recon_stride, | 253 int recon_stride, |
393 unsigned int *sse) | 254 unsigned int *sse) |
394 { | 255 { |
395 return vp8_sub_pixel_variance16x16_mmx(src_ptr, source_stride, 4, 4, | 256 return vp8_sub_pixel_variance16x16_mmx(src_ptr, source_stride, 4, 4, |
396 ref_ptr, recon_stride, sse); | 257 ref_ptr, recon_stride, sse); |
397 } | 258 } |
OLD | NEW |