1 /*
2 * Copyright (c) 2023, Alliance for Open Media. All rights reserved
3 *
4 * This source code is subject to the terms of the BSD 2 Clause License and
5 * the Alliance for Open Media Patent License 1.0. If the BSD 2 Clause License
6 * was not distributed with this source code in the LICENSE file, you can
7 * obtain it at www.aomedia.org/license/software. If the Alliance for Open
8 * Media Patent License 1.0 was not distributed with this source code in the
9 * PATENTS file, you can obtain it at www.aomedia.org/license/patent.
10 */
11
12 #include <arm_neon.h>
13
14 #include "aom/aom_integer.h"
15 #include "aom_dsp/arm/mem_neon.h"
16 #include "aom_dsp/arm/sum_neon.h"
17 #include "aom_ports/mem.h"
18 #include "config/aom_config.h"
19 #include "config/aom_dsp_rtcd.h"
20
variance_4xh_neon_dotprod(const uint8_t * src,int src_stride,const uint8_t * ref,int ref_stride,int h,uint32_t * sse,int * sum)21 static INLINE void variance_4xh_neon_dotprod(const uint8_t *src, int src_stride,
22 const uint8_t *ref, int ref_stride,
23 int h, uint32_t *sse, int *sum) {
24 uint32x4_t src_sum = vdupq_n_u32(0);
25 uint32x4_t ref_sum = vdupq_n_u32(0);
26 uint32x4_t sse_u32 = vdupq_n_u32(0);
27
28 int i = h;
29 do {
30 uint8x16_t s = load_unaligned_u8q(src, src_stride);
31 uint8x16_t r = load_unaligned_u8q(ref, ref_stride);
32
33 src_sum = vdotq_u32(src_sum, s, vdupq_n_u8(1));
34 ref_sum = vdotq_u32(ref_sum, r, vdupq_n_u8(1));
35
36 uint8x16_t abs_diff = vabdq_u8(s, r);
37 sse_u32 = vdotq_u32(sse_u32, abs_diff, abs_diff);
38
39 src += 4 * src_stride;
40 ref += 4 * ref_stride;
41 i -= 4;
42 } while (i != 0);
43
44 int32x4_t sum_diff =
45 vsubq_s32(vreinterpretq_s32_u32(src_sum), vreinterpretq_s32_u32(ref_sum));
46 *sum = horizontal_add_s32x4(sum_diff);
47 *sse = horizontal_add_u32x4(sse_u32);
48 }
49
variance_8xh_neon_dotprod(const uint8_t * src,int src_stride,const uint8_t * ref,int ref_stride,int h,uint32_t * sse,int * sum)50 static INLINE void variance_8xh_neon_dotprod(const uint8_t *src, int src_stride,
51 const uint8_t *ref, int ref_stride,
52 int h, uint32_t *sse, int *sum) {
53 uint32x4_t src_sum = vdupq_n_u32(0);
54 uint32x4_t ref_sum = vdupq_n_u32(0);
55 uint32x4_t sse_u32 = vdupq_n_u32(0);
56
57 int i = h;
58 do {
59 uint8x16_t s = vcombine_u8(vld1_u8(src), vld1_u8(src + src_stride));
60 uint8x16_t r = vcombine_u8(vld1_u8(ref), vld1_u8(ref + ref_stride));
61
62 src_sum = vdotq_u32(src_sum, s, vdupq_n_u8(1));
63 ref_sum = vdotq_u32(ref_sum, r, vdupq_n_u8(1));
64
65 uint8x16_t abs_diff = vabdq_u8(s, r);
66 sse_u32 = vdotq_u32(sse_u32, abs_diff, abs_diff);
67
68 src += 2 * src_stride;
69 ref += 2 * ref_stride;
70 i -= 2;
71 } while (i != 0);
72
73 int32x4_t sum_diff =
74 vsubq_s32(vreinterpretq_s32_u32(src_sum), vreinterpretq_s32_u32(ref_sum));
75 *sum = horizontal_add_s32x4(sum_diff);
76 *sse = horizontal_add_u32x4(sse_u32);
77 }
78
variance_16xh_neon_dotprod(const uint8_t * src,int src_stride,const uint8_t * ref,int ref_stride,int h,uint32_t * sse,int * sum)79 static INLINE void variance_16xh_neon_dotprod(const uint8_t *src,
80 int src_stride,
81 const uint8_t *ref,
82 int ref_stride, int h,
83 uint32_t *sse, int *sum) {
84 uint32x4_t src_sum = vdupq_n_u32(0);
85 uint32x4_t ref_sum = vdupq_n_u32(0);
86 uint32x4_t sse_u32 = vdupq_n_u32(0);
87
88 int i = h;
89 do {
90 uint8x16_t s = vld1q_u8(src);
91 uint8x16_t r = vld1q_u8(ref);
92
93 src_sum = vdotq_u32(src_sum, s, vdupq_n_u8(1));
94 ref_sum = vdotq_u32(ref_sum, r, vdupq_n_u8(1));
95
96 uint8x16_t abs_diff = vabdq_u8(s, r);
97 sse_u32 = vdotq_u32(sse_u32, abs_diff, abs_diff);
98
99 src += src_stride;
100 ref += ref_stride;
101 } while (--i != 0);
102
103 int32x4_t sum_diff =
104 vsubq_s32(vreinterpretq_s32_u32(src_sum), vreinterpretq_s32_u32(ref_sum));
105 *sum = horizontal_add_s32x4(sum_diff);
106 *sse = horizontal_add_u32x4(sse_u32);
107 }
108
variance_large_neon_dotprod(const uint8_t * src,int src_stride,const uint8_t * ref,int ref_stride,int w,int h,uint32_t * sse,int * sum)109 static INLINE void variance_large_neon_dotprod(const uint8_t *src,
110 int src_stride,
111 const uint8_t *ref,
112 int ref_stride, int w, int h,
113 uint32_t *sse, int *sum) {
114 uint32x4_t src_sum = vdupq_n_u32(0);
115 uint32x4_t ref_sum = vdupq_n_u32(0);
116 uint32x4_t sse_u32 = vdupq_n_u32(0);
117
118 int i = h;
119 do {
120 int j = 0;
121 do {
122 uint8x16_t s = vld1q_u8(src + j);
123 uint8x16_t r = vld1q_u8(ref + j);
124
125 src_sum = vdotq_u32(src_sum, s, vdupq_n_u8(1));
126 ref_sum = vdotq_u32(ref_sum, r, vdupq_n_u8(1));
127
128 uint8x16_t abs_diff = vabdq_u8(s, r);
129 sse_u32 = vdotq_u32(sse_u32, abs_diff, abs_diff);
130
131 j += 16;
132 } while (j < w);
133
134 src += src_stride;
135 ref += ref_stride;
136 } while (--i != 0);
137
138 int32x4_t sum_diff =
139 vsubq_s32(vreinterpretq_s32_u32(src_sum), vreinterpretq_s32_u32(ref_sum));
140 *sum = horizontal_add_s32x4(sum_diff);
141 *sse = horizontal_add_u32x4(sse_u32);
142 }
143
variance_32xh_neon_dotprod(const uint8_t * src,int src_stride,const uint8_t * ref,int ref_stride,int h,uint32_t * sse,int * sum)144 static INLINE void variance_32xh_neon_dotprod(const uint8_t *src,
145 int src_stride,
146 const uint8_t *ref,
147 int ref_stride, int h,
148 uint32_t *sse, int *sum) {
149 variance_large_neon_dotprod(src, src_stride, ref, ref_stride, 32, h, sse,
150 sum);
151 }
152
variance_64xh_neon_dotprod(const uint8_t * src,int src_stride,const uint8_t * ref,int ref_stride,int h,uint32_t * sse,int * sum)153 static INLINE void variance_64xh_neon_dotprod(const uint8_t *src,
154 int src_stride,
155 const uint8_t *ref,
156 int ref_stride, int h,
157 uint32_t *sse, int *sum) {
158 variance_large_neon_dotprod(src, src_stride, ref, ref_stride, 64, h, sse,
159 sum);
160 }
161
variance_128xh_neon_dotprod(const uint8_t * src,int src_stride,const uint8_t * ref,int ref_stride,int h,uint32_t * sse,int * sum)162 static INLINE void variance_128xh_neon_dotprod(const uint8_t *src,
163 int src_stride,
164 const uint8_t *ref,
165 int ref_stride, int h,
166 uint32_t *sse, int *sum) {
167 variance_large_neon_dotprod(src, src_stride, ref, ref_stride, 128, h, sse,
168 sum);
169 }
170
171 #define VARIANCE_WXH_NEON_DOTPROD(w, h, shift) \
172 unsigned int aom_variance##w##x##h##_neon_dotprod( \
173 const uint8_t *src, int src_stride, const uint8_t *ref, int ref_stride, \
174 unsigned int *sse) { \
175 int sum; \
176 variance_##w##xh_neon_dotprod(src, src_stride, ref, ref_stride, h, sse, \
177 &sum); \
178 return *sse - (uint32_t)(((int64_t)sum * sum) >> shift); \
179 }
180
181 VARIANCE_WXH_NEON_DOTPROD(4, 4, 4)
182 VARIANCE_WXH_NEON_DOTPROD(4, 8, 5)
183 VARIANCE_WXH_NEON_DOTPROD(4, 16, 6)
184
185 VARIANCE_WXH_NEON_DOTPROD(8, 4, 5)
186 VARIANCE_WXH_NEON_DOTPROD(8, 8, 6)
187 VARIANCE_WXH_NEON_DOTPROD(8, 16, 7)
188 VARIANCE_WXH_NEON_DOTPROD(8, 32, 8)
189
190 VARIANCE_WXH_NEON_DOTPROD(16, 4, 6)
191 VARIANCE_WXH_NEON_DOTPROD(16, 8, 7)
192 VARIANCE_WXH_NEON_DOTPROD(16, 16, 8)
193 VARIANCE_WXH_NEON_DOTPROD(16, 32, 9)
194 VARIANCE_WXH_NEON_DOTPROD(16, 64, 10)
195
196 VARIANCE_WXH_NEON_DOTPROD(32, 8, 8)
197 VARIANCE_WXH_NEON_DOTPROD(32, 16, 9)
198 VARIANCE_WXH_NEON_DOTPROD(32, 32, 10)
199 VARIANCE_WXH_NEON_DOTPROD(32, 64, 11)
200
201 VARIANCE_WXH_NEON_DOTPROD(64, 16, 10)
202 VARIANCE_WXH_NEON_DOTPROD(64, 32, 11)
203 VARIANCE_WXH_NEON_DOTPROD(64, 64, 12)
204 VARIANCE_WXH_NEON_DOTPROD(64, 128, 13)
205
206 VARIANCE_WXH_NEON_DOTPROD(128, 64, 13)
207 VARIANCE_WXH_NEON_DOTPROD(128, 128, 14)
208
209 #undef VARIANCE_WXH_NEON_DOTPROD
210
aom_get_var_sse_sum_8x8_quad_neon_dotprod(const uint8_t * src,int src_stride,const uint8_t * ref,int ref_stride,uint32_t * sse8x8,int * sum8x8,unsigned int * tot_sse,int * tot_sum,uint32_t * var8x8)211 void aom_get_var_sse_sum_8x8_quad_neon_dotprod(
212 const uint8_t *src, int src_stride, const uint8_t *ref, int ref_stride,
213 uint32_t *sse8x8, int *sum8x8, unsigned int *tot_sse, int *tot_sum,
214 uint32_t *var8x8) {
215 // Loop over four 8x8 blocks. Process one 8x32 block.
216 for (int k = 0; k < 4; k++) {
217 variance_8xh_neon_dotprod(src + (k * 8), src_stride, ref + (k * 8),
218 ref_stride, 8, &sse8x8[k], &sum8x8[k]);
219 }
220
221 *tot_sse += sse8x8[0] + sse8x8[1] + sse8x8[2] + sse8x8[3];
222 *tot_sum += sum8x8[0] + sum8x8[1] + sum8x8[2] + sum8x8[3];
223 for (int i = 0; i < 4; i++) {
224 var8x8[i] = sse8x8[i] - (uint32_t)(((int64_t)sum8x8[i] * sum8x8[i]) >> 6);
225 }
226 }
227
aom_get_var_sse_sum_16x16_dual_neon_dotprod(const uint8_t * src,int src_stride,const uint8_t * ref,int ref_stride,uint32_t * sse16x16,unsigned int * tot_sse,int * tot_sum,uint32_t * var16x16)228 void aom_get_var_sse_sum_16x16_dual_neon_dotprod(
229 const uint8_t *src, int src_stride, const uint8_t *ref, int ref_stride,
230 uint32_t *sse16x16, unsigned int *tot_sse, int *tot_sum,
231 uint32_t *var16x16) {
232 int sum16x16[2] = { 0 };
233 // Loop over two 16x16 blocks. Process one 16x32 block.
234 for (int k = 0; k < 2; k++) {
235 variance_16xh_neon_dotprod(src + (k * 16), src_stride, ref + (k * 16),
236 ref_stride, 16, &sse16x16[k], &sum16x16[k]);
237 }
238
239 *tot_sse += sse16x16[0] + sse16x16[1];
240 *tot_sum += sum16x16[0] + sum16x16[1];
241 for (int i = 0; i < 2; i++) {
242 var16x16[i] =
243 sse16x16[i] - (uint32_t)(((int64_t)sum16x16[i] * sum16x16[i]) >> 8);
244 }
245 }
246
mse8xh_neon_dotprod(const uint8_t * src,int src_stride,const uint8_t * ref,int ref_stride,unsigned int * sse,int h)247 static INLINE unsigned int mse8xh_neon_dotprod(const uint8_t *src,
248 int src_stride,
249 const uint8_t *ref,
250 int ref_stride,
251 unsigned int *sse, int h) {
252 uint32x4_t sse_u32 = vdupq_n_u32(0);
253
254 int i = h;
255 do {
256 uint8x16_t s = vcombine_u8(vld1_u8(src), vld1_u8(src + src_stride));
257 uint8x16_t r = vcombine_u8(vld1_u8(ref), vld1_u8(ref + ref_stride));
258
259 uint8x16_t abs_diff = vabdq_u8(s, r);
260
261 sse_u32 = vdotq_u32(sse_u32, abs_diff, abs_diff);
262
263 src += 2 * src_stride;
264 ref += 2 * ref_stride;
265 i -= 2;
266 } while (i != 0);
267
268 *sse = horizontal_add_u32x4(sse_u32);
269 return horizontal_add_u32x4(sse_u32);
270 }
271
mse16xh_neon_dotprod(const uint8_t * src,int src_stride,const uint8_t * ref,int ref_stride,unsigned int * sse,int h)272 static INLINE unsigned int mse16xh_neon_dotprod(const uint8_t *src,
273 int src_stride,
274 const uint8_t *ref,
275 int ref_stride,
276 unsigned int *sse, int h) {
277 uint32x4_t sse_u32[2] = { vdupq_n_u32(0), vdupq_n_u32(0) };
278
279 int i = h;
280 do {
281 uint8x16_t s0 = vld1q_u8(src);
282 uint8x16_t s1 = vld1q_u8(src + src_stride);
283 uint8x16_t r0 = vld1q_u8(ref);
284 uint8x16_t r1 = vld1q_u8(ref + ref_stride);
285
286 uint8x16_t abs_diff0 = vabdq_u8(s0, r0);
287 uint8x16_t abs_diff1 = vabdq_u8(s1, r1);
288
289 sse_u32[0] = vdotq_u32(sse_u32[0], abs_diff0, abs_diff0);
290 sse_u32[1] = vdotq_u32(sse_u32[1], abs_diff1, abs_diff1);
291
292 src += 2 * src_stride;
293 ref += 2 * ref_stride;
294 i -= 2;
295 } while (i != 0);
296
297 *sse = horizontal_add_u32x4(vaddq_u32(sse_u32[0], sse_u32[1]));
298 return horizontal_add_u32x4(vaddq_u32(sse_u32[0], sse_u32[1]));
299 }
300
301 #define MSE_WXH_NEON_DOTPROD(w, h) \
302 unsigned int aom_mse##w##x##h##_neon_dotprod( \
303 const uint8_t *src, int src_stride, const uint8_t *ref, int ref_stride, \
304 unsigned int *sse) { \
305 return mse##w##xh_neon_dotprod(src, src_stride, ref, ref_stride, sse, h); \
306 }
307
308 MSE_WXH_NEON_DOTPROD(8, 8)
309 MSE_WXH_NEON_DOTPROD(8, 16)
310
311 MSE_WXH_NEON_DOTPROD(16, 8)
312 MSE_WXH_NEON_DOTPROD(16, 16)
313
314 #undef MSE_WXH_NEON_DOTPROD
315