1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
|
;
; Copyright (c) 2016, Alliance for Open Media. All rights reserved
;
; This source code is subject to the terms of the BSD 2 Clause License and
; the Alliance for Open Media Patent License 1.0. If the BSD 2 Clause License
; was not distributed with this source code in the LICENSE file, you can
; obtain it at www.aomedia.org/license/software. If the Alliance for Open
; Media Patent License 1.0 was not distributed with this source code in the
; PATENTS file, you can obtain it at www.aomedia.org/license/patent.
;
;
%include "aom_ports/x86_abi_support.asm"
;Note: tap3 and tap4 have to be applied and added after other taps to avoid
;overflow.
%macro GET_FILTERS_4 0
mov rdx, arg(5) ;filter ptr
mov rcx, 0x0400040
movdqa xmm7, [rdx] ;load filters
pshuflw xmm0, xmm7, 0b ;k0
pshuflw xmm1, xmm7, 01010101b ;k1
pshuflw xmm2, xmm7, 10101010b ;k2
pshuflw xmm3, xmm7, 11111111b ;k3
psrldq xmm7, 8
pshuflw xmm4, xmm7, 0b ;k4
pshuflw xmm5, xmm7, 01010101b ;k5
pshuflw xmm6, xmm7, 10101010b ;k6
pshuflw xmm7, xmm7, 11111111b ;k7
punpcklqdq xmm0, xmm1
punpcklqdq xmm2, xmm3
punpcklqdq xmm5, xmm4
punpcklqdq xmm6, xmm7
movdqa k0k1, xmm0
movdqa k2k3, xmm2
movdqa k5k4, xmm5
movdqa k6k7, xmm6
movq xmm6, rcx
pshufd xmm6, xmm6, 0
movdqa krd, xmm6
pxor xmm7, xmm7
movdqa zero, xmm7
%endm
%macro APPLY_FILTER_4 1
punpckldq xmm0, xmm1 ;two row in one register
punpckldq xmm6, xmm7
punpckldq xmm2, xmm3
punpckldq xmm5, xmm4
punpcklbw xmm0, zero ;unpack to word
punpcklbw xmm6, zero
punpcklbw xmm2, zero
punpcklbw xmm5, zero
pmullw xmm0, k0k1 ;multiply the filter factors
pmullw xmm6, k6k7
pmullw xmm2, k2k3
pmullw xmm5, k5k4
paddsw xmm0, xmm6 ;sum
movdqa xmm1, xmm0
psrldq xmm1, 8
paddsw xmm0, xmm1
paddsw xmm0, xmm2
psrldq xmm2, 8
paddsw xmm0, xmm5
psrldq xmm5, 8
paddsw xmm0, xmm2
paddsw xmm0, xmm5
paddsw xmm0, krd ;rounding
psraw xmm0, 7 ;shift
packuswb xmm0, xmm0 ;pack to byte
%if %1
movd xmm1, [rdi]
pavgb xmm0, xmm1
%endif
movd [rdi], xmm0
%endm
%macro GET_FILTERS 0
mov rdx, arg(5) ;filter ptr
mov rsi, arg(0) ;src_ptr
mov rdi, arg(2) ;output_ptr
mov rcx, 0x0400040
movdqa xmm7, [rdx] ;load filters
pshuflw xmm0, xmm7, 0b ;k0
pshuflw xmm1, xmm7, 01010101b ;k1
pshuflw xmm2, xmm7, 10101010b ;k2
pshuflw xmm3, xmm7, 11111111b ;k3
pshufhw xmm4, xmm7, 0b ;k4
pshufhw xmm5, xmm7, 01010101b ;k5
pshufhw xmm6, xmm7, 10101010b ;k6
pshufhw xmm7, xmm7, 11111111b ;k7
punpcklwd xmm0, xmm0
punpcklwd xmm1, xmm1
punpcklwd xmm2, xmm2
punpcklwd xmm3, xmm3
punpckhwd xmm4, xmm4
punpckhwd xmm5, xmm5
punpckhwd xmm6, xmm6
punpckhwd xmm7, xmm7
movdqa k0, xmm0 ;store filter factors on stack
movdqa k1, xmm1
movdqa k2, xmm2
movdqa k3, xmm3
movdqa k4, xmm4
movdqa k5, xmm5
movdqa k6, xmm6
movdqa k7, xmm7
movq xmm6, rcx
pshufd xmm6, xmm6, 0
movdqa krd, xmm6 ;rounding
pxor xmm7, xmm7
movdqa zero, xmm7
%endm
%macro LOAD_VERT_8 1
movq xmm0, [rsi + %1] ;0
movq xmm1, [rsi + rax + %1] ;1
movq xmm6, [rsi + rdx * 2 + %1] ;6
lea rsi, [rsi + rax]
movq xmm7, [rsi + rdx * 2 + %1] ;7
movq xmm2, [rsi + rax + %1] ;2
movq xmm3, [rsi + rax * 2 + %1] ;3
movq xmm4, [rsi + rdx + %1] ;4
movq xmm5, [rsi + rax * 4 + %1] ;5
%endm
%macro APPLY_FILTER_8 2
punpcklbw xmm0, zero
punpcklbw xmm1, zero
punpcklbw xmm6, zero
punpcklbw xmm7, zero
punpcklbw xmm2, zero
punpcklbw xmm5, zero
punpcklbw xmm3, zero
punpcklbw xmm4, zero
pmullw xmm0, k0
pmullw xmm1, k1
pmullw xmm6, k6
pmullw xmm7, k7
pmullw xmm2, k2
pmullw xmm5, k5
pmullw xmm3, k3
pmullw xmm4, k4
paddsw xmm0, xmm1
paddsw xmm0, xmm6
paddsw xmm0, xmm7
paddsw xmm0, xmm2
paddsw xmm0, xmm5
paddsw xmm0, xmm3
paddsw xmm0, xmm4
paddsw xmm0, krd ;rounding
psraw xmm0, 7 ;shift
packuswb xmm0, xmm0 ;pack back to byte
%if %1
movq xmm1, [rdi + %2]
pavgb xmm0, xmm1
%endif
movq [rdi + %2], xmm0
%endm
SECTION .text
;void aom_filter_block1d4_v8_sse2
;(
; unsigned char *src_ptr,
; unsigned int src_pitch,
; unsigned char *output_ptr,
; unsigned int out_pitch,
; unsigned int output_height,
; short *filter
;)
globalsym(aom_filter_block1d4_v8_sse2)
sym(aom_filter_block1d4_v8_sse2):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 6
SAVE_XMM 7
push rsi
push rdi
push rbx
; end prolog
ALIGN_STACK 16, rax
sub rsp, 16 * 6
%define k0k1 [rsp + 16 * 0]
%define k2k3 [rsp + 16 * 1]
%define k5k4 [rsp + 16 * 2]
%define k6k7 [rsp + 16 * 3]
%define krd [rsp + 16 * 4]
%define zero [rsp + 16 * 5]
GET_FILTERS_4
mov rsi, arg(0) ;src_ptr
mov rdi, arg(2) ;output_ptr
movsxd rax, DWORD PTR arg(1) ;pixels_per_line
movsxd rbx, DWORD PTR arg(3) ;out_pitch
lea rdx, [rax + rax * 2]
movsxd rcx, DWORD PTR arg(4) ;output_height
.loop:
movd xmm0, [rsi] ;load src: row 0
movd xmm1, [rsi + rax] ;1
movd xmm6, [rsi + rdx * 2] ;6
lea rsi, [rsi + rax]
movd xmm7, [rsi + rdx * 2] ;7
movd xmm2, [rsi + rax] ;2
movd xmm3, [rsi + rax * 2] ;3
movd xmm4, [rsi + rdx] ;4
movd xmm5, [rsi + rax * 4] ;5
APPLY_FILTER_4 0
lea rdi, [rdi + rbx]
dec rcx
jnz .loop
add rsp, 16 * 6
pop rsp
pop rbx
; begin epilog
pop rdi
pop rsi
RESTORE_XMM
UNSHADOW_ARGS
pop rbp
ret
;void aom_filter_block1d8_v8_sse2
;(
; unsigned char *src_ptr,
; unsigned int src_pitch,
; unsigned char *output_ptr,
; unsigned int out_pitch,
; unsigned int output_height,
; short *filter
;)
globalsym(aom_filter_block1d8_v8_sse2)
sym(aom_filter_block1d8_v8_sse2):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 6
SAVE_XMM 7
push rsi
push rdi
push rbx
; end prolog
ALIGN_STACK 16, rax
sub rsp, 16 * 10
%define k0 [rsp + 16 * 0]
%define k1 [rsp + 16 * 1]
%define k2 [rsp + 16 * 2]
%define k3 [rsp + 16 * 3]
%define k4 [rsp + 16 * 4]
%define k5 [rsp + 16 * 5]
%define k6 [rsp + 16 * 6]
%define k7 [rsp + 16 * 7]
%define krd [rsp + 16 * 8]
%define zero [rsp + 16 * 9]
GET_FILTERS
movsxd rax, DWORD PTR arg(1) ;pixels_per_line
movsxd rbx, DWORD PTR arg(3) ;out_pitch
lea rdx, [rax + rax * 2]
movsxd rcx, DWORD PTR arg(4) ;output_height
.loop:
LOAD_VERT_8 0
APPLY_FILTER_8 0, 0
lea rdi, [rdi + rbx]
dec rcx
jnz .loop
add rsp, 16 * 10
pop rsp
pop rbx
; begin epilog
pop rdi
pop rsi
RESTORE_XMM
UNSHADOW_ARGS
pop rbp
ret
;void aom_filter_block1d16_v8_sse2
;(
; unsigned char *src_ptr,
; unsigned int src_pitch,
; unsigned char *output_ptr,
; unsigned int out_pitch,
; unsigned int output_height,
; short *filter
;)
globalsym(aom_filter_block1d16_v8_sse2)
sym(aom_filter_block1d16_v8_sse2):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 6
SAVE_XMM 7
push rsi
push rdi
push rbx
; end prolog
ALIGN_STACK 16, rax
sub rsp, 16 * 10
%define k0 [rsp + 16 * 0]
%define k1 [rsp + 16 * 1]
%define k2 [rsp + 16 * 2]
%define k3 [rsp + 16 * 3]
%define k4 [rsp + 16 * 4]
%define k5 [rsp + 16 * 5]
%define k6 [rsp + 16 * 6]
%define k7 [rsp + 16 * 7]
%define krd [rsp + 16 * 8]
%define zero [rsp + 16 * 9]
GET_FILTERS
movsxd rax, DWORD PTR arg(1) ;pixels_per_line
movsxd rbx, DWORD PTR arg(3) ;out_pitch
lea rdx, [rax + rax * 2]
movsxd rcx, DWORD PTR arg(4) ;output_height
.loop:
LOAD_VERT_8 0
APPLY_FILTER_8 0, 0
sub rsi, rax
LOAD_VERT_8 8
APPLY_FILTER_8 0, 8
add rdi, rbx
dec rcx
jnz .loop
add rsp, 16 * 10
pop rsp
pop rbx
; begin epilog
pop rdi
pop rsi
RESTORE_XMM
UNSHADOW_ARGS
pop rbp
ret
;void aom_filter_block1d4_h8_sse2
;(
; unsigned char *src_ptr,
; unsigned int src_pixels_per_line,
; unsigned char *output_ptr,
; unsigned int output_pitch,
; unsigned int output_height,
; short *filter
;)
globalsym(aom_filter_block1d4_h8_sse2)
sym(aom_filter_block1d4_h8_sse2):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 6
SAVE_XMM 7
push rsi
push rdi
; end prolog
ALIGN_STACK 16, rax
sub rsp, 16 * 6
%define k0k1 [rsp + 16 * 0]
%define k2k3 [rsp + 16 * 1]
%define k5k4 [rsp + 16 * 2]
%define k6k7 [rsp + 16 * 3]
%define krd [rsp + 16 * 4]
%define zero [rsp + 16 * 5]
GET_FILTERS_4
mov rsi, arg(0) ;src_ptr
mov rdi, arg(2) ;output_ptr
movsxd rax, DWORD PTR arg(1) ;pixels_per_line
movsxd rdx, DWORD PTR arg(3) ;out_pitch
movsxd rcx, DWORD PTR arg(4) ;output_height
.loop:
movdqu xmm0, [rsi - 3] ;load src
movdqa xmm1, xmm0
movdqa xmm6, xmm0
movdqa xmm7, xmm0
movdqa xmm2, xmm0
movdqa xmm3, xmm0
movdqa xmm5, xmm0
movdqa xmm4, xmm0
psrldq xmm1, 1
psrldq xmm6, 6
psrldq xmm7, 7
psrldq xmm2, 2
psrldq xmm3, 3
psrldq xmm5, 5
psrldq xmm4, 4
APPLY_FILTER_4 0
lea rsi, [rsi + rax]
lea rdi, [rdi + rdx]
dec rcx
jnz .loop
add rsp, 16 * 6
pop rsp
; begin epilog
pop rdi
pop rsi
RESTORE_XMM
UNSHADOW_ARGS
pop rbp
ret
;void aom_filter_block1d8_h8_sse2
;(
; unsigned char *src_ptr,
; unsigned int src_pixels_per_line,
; unsigned char *output_ptr,
; unsigned int output_pitch,
; unsigned int output_height,
; short *filter
;)
globalsym(aom_filter_block1d8_h8_sse2)
sym(aom_filter_block1d8_h8_sse2):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 6
SAVE_XMM 7
push rsi
push rdi
; end prolog
ALIGN_STACK 16, rax
sub rsp, 16 * 10
%define k0 [rsp + 16 * 0]
%define k1 [rsp + 16 * 1]
%define k2 [rsp + 16 * 2]
%define k3 [rsp + 16 * 3]
%define k4 [rsp + 16 * 4]
%define k5 [rsp + 16 * 5]
%define k6 [rsp + 16 * 6]
%define k7 [rsp + 16 * 7]
%define krd [rsp + 16 * 8]
%define zero [rsp + 16 * 9]
GET_FILTERS
movsxd rax, DWORD PTR arg(1) ;pixels_per_line
movsxd rdx, DWORD PTR arg(3) ;out_pitch
movsxd rcx, DWORD PTR arg(4) ;output_height
.loop:
movdqu xmm0, [rsi - 3] ;load src
movdqa xmm1, xmm0
movdqa xmm6, xmm0
movdqa xmm7, xmm0
movdqa xmm2, xmm0
movdqa xmm5, xmm0
movdqa xmm3, xmm0
movdqa xmm4, xmm0
psrldq xmm1, 1
psrldq xmm6, 6
psrldq xmm7, 7
psrldq xmm2, 2
psrldq xmm5, 5
psrldq xmm3, 3
psrldq xmm4, 4
APPLY_FILTER_8 0, 0
lea rsi, [rsi + rax]
lea rdi, [rdi + rdx]
dec rcx
jnz .loop
add rsp, 16 * 10
pop rsp
; begin epilog
pop rdi
pop rsi
RESTORE_XMM
UNSHADOW_ARGS
pop rbp
ret
;void aom_filter_block1d16_h8_sse2
;(
; unsigned char *src_ptr,
; unsigned int src_pixels_per_line,
; unsigned char *output_ptr,
; unsigned int output_pitch,
; unsigned int output_height,
; short *filter
;)
globalsym(aom_filter_block1d16_h8_sse2)
sym(aom_filter_block1d16_h8_sse2):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 6
SAVE_XMM 7
push rsi
push rdi
; end prolog
ALIGN_STACK 16, rax
sub rsp, 16 * 10
%define k0 [rsp + 16 * 0]
%define k1 [rsp + 16 * 1]
%define k2 [rsp + 16 * 2]
%define k3 [rsp + 16 * 3]
%define k4 [rsp + 16 * 4]
%define k5 [rsp + 16 * 5]
%define k6 [rsp + 16 * 6]
%define k7 [rsp + 16 * 7]
%define krd [rsp + 16 * 8]
%define zero [rsp + 16 * 9]
GET_FILTERS
movsxd rax, DWORD PTR arg(1) ;pixels_per_line
movsxd rdx, DWORD PTR arg(3) ;out_pitch
movsxd rcx, DWORD PTR arg(4) ;output_height
.loop:
movdqu xmm0, [rsi - 3] ;load src
movdqa xmm1, xmm0
movdqa xmm6, xmm0
movdqa xmm7, xmm0
movdqa xmm2, xmm0
movdqa xmm5, xmm0
movdqa xmm3, xmm0
movdqa xmm4, xmm0
psrldq xmm1, 1
psrldq xmm6, 6
psrldq xmm7, 7
psrldq xmm2, 2
psrldq xmm5, 5
psrldq xmm3, 3
psrldq xmm4, 4
APPLY_FILTER_8 0, 0
movdqu xmm0, [rsi + 5] ;load src
movdqa xmm1, xmm0
movdqa xmm6, xmm0
movdqa xmm7, xmm0
movdqa xmm2, xmm0
movdqa xmm5, xmm0
movdqa xmm3, xmm0
movdqa xmm4, xmm0
psrldq xmm1, 1
psrldq xmm6, 6
psrldq xmm7, 7
psrldq xmm2, 2
psrldq xmm5, 5
psrldq xmm3, 3
psrldq xmm4, 4
APPLY_FILTER_8 0, 8
lea rsi, [rsi + rax]
lea rdi, [rdi + rdx]
dec rcx
jnz .loop
add rsp, 16 * 10
pop rsp
; begin epilog
pop rdi
pop rsi
RESTORE_XMM
UNSHADOW_ARGS
pop rbp
ret
|