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
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
|
;
; Copyright (c) 2010 The VP8 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 "vpx_ports/x86_abi_support.asm"
%define BLOCK_HEIGHT_WIDTH 4
%define VP8_FILTER_WEIGHT 128
%define VP8_FILTER_SHIFT 7
;/************************************************************************************
; Notes: filter_block1d_h6 applies a 6 tap filter horizontally to the input pixels. The
; input pixel array has output_height rows. This routine assumes that output_height is an
; even number. This function handles 8 pixels in horizontal direction, calculating ONE
; rows each iteration to take advantage of the 128 bits operations.
;
; This is an implementation of some of the SSE optimizations first seen in ffvp8
;
;*************************************************************************************/
;void vp8_filter_block1d8_h6_ssse3
;(
; unsigned char *src_ptr,
; unsigned int src_pixels_per_line,
; unsigned char *output_ptr,
; unsigned int output_pitch,
; unsigned int output_height,
; unsigned int vp8_filter_index
;)
global sym(vp8_filter_block1d8_h6_ssse3)
sym(vp8_filter_block1d8_h6_ssse3):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 6
GET_GOT rbx
push rsi
push rdi
; end prolog
movsxd rdx, DWORD PTR arg(5) ;table index
xor rsi, rsi
shl rdx, 4
movdqa xmm7, [rd GLOBAL]
lea rax, [k0_k5 GLOBAL]
add rax, rdx
mov rdi, arg(2) ;output_ptr
cmp esi, DWORD PTR [rax]
je vp8_filter_block1d8_h4_ssse3
movdqa xmm4, XMMWORD PTR [rax] ;k0_k5
movdqa xmm5, XMMWORD PTR [rax+256] ;k2_k4
movdqa xmm6, XMMWORD PTR [rax+128] ;k1_k3
mov rsi, arg(0) ;src_ptr
movsxd rax, dword ptr arg(1) ;src_pixels_per_line
movsxd rcx, dword ptr arg(4) ;output_height
movsxd rdx, dword ptr arg(3) ;output_pitch
sub rdi, rdx
;xmm3 free
filter_block1d8_h6_rowloop_ssse3:
movdqu xmm0, XMMWORD PTR [rsi - 2]
movdqa xmm1, xmm0
pshufb xmm0, [shuf1b GLOBAL]
movdqa xmm2, xmm1
pshufb xmm1, [shuf2b GLOBAL]
pmaddubsw xmm0, xmm4
pmaddubsw xmm1, xmm5
pshufb xmm2, [shuf3b GLOBAL]
add rdi, rdx
pmaddubsw xmm2, xmm6
lea rsi, [rsi + rax]
dec rcx
paddsw xmm0, xmm1
paddsw xmm0, xmm7
paddsw xmm0, xmm2
psraw xmm0, 7
packuswb xmm0, xmm0
movq MMWORD Ptr [rdi], xmm0
jnz filter_block1d8_h6_rowloop_ssse3
; begin epilog
pop rdi
pop rsi
RESTORE_GOT
UNSHADOW_ARGS
pop rbp
ret
vp8_filter_block1d8_h4_ssse3:
movdqa xmm5, XMMWORD PTR [rax+256] ;k2_k4
movdqa xmm6, XMMWORD PTR [rax+128] ;k1_k3
movdqa xmm3, XMMWORD PTR [shuf2b GLOBAL]
movdqa xmm4, XMMWORD PTR [shuf3b GLOBAL]
mov rsi, arg(0) ;src_ptr
movsxd rax, dword ptr arg(1) ;src_pixels_per_line
movsxd rcx, dword ptr arg(4) ;output_height
movsxd rdx, dword ptr arg(3) ;output_pitch
sub rdi, rdx
;xmm3 free
filter_block1d8_h4_rowloop_ssse3:
movdqu xmm0, XMMWORD PTR [rsi - 2]
movdqa xmm2, xmm0
pshufb xmm0, xmm3 ;[shuf2b GLOBAL]
pshufb xmm2, xmm4 ;[shuf3b GLOBAL]
pmaddubsw xmm0, xmm5
add rdi, rdx
pmaddubsw xmm2, xmm6
lea rsi, [rsi + rax]
dec rcx
paddsw xmm0, xmm7
paddsw xmm0, xmm2
psraw xmm0, 7
packuswb xmm0, xmm0
movq MMWORD Ptr [rdi], xmm0
jnz filter_block1d8_h4_rowloop_ssse3
; begin epilog
pop rdi
pop rsi
RESTORE_GOT
UNSHADOW_ARGS
pop rbp
ret
;void vp8_filter_block1d16_h6_ssse3
;(
; unsigned char *src_ptr,
; unsigned int src_pixels_per_line,
; unsigned char *output_ptr,
; unsigned int output_pitch,
; unsigned int output_height,
; unsigned int vp8_filter_index
;)
global sym(vp8_filter_block1d16_h6_ssse3)
sym(vp8_filter_block1d16_h6_ssse3):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 6
SAVE_XMM
GET_GOT rbx
push rsi
push rdi
; end prolog
movsxd rdx, DWORD PTR arg(5) ;table index
xor rsi, rsi
shl rdx, 4 ;
lea rax, [k0_k5 GLOBAL]
add rax, rdx
mov rdi, arg(2) ;output_ptr
movdqa xmm7, [rd GLOBAL]
;;
;; cmp esi, DWORD PTR [rax]
;; je vp8_filter_block1d16_h4_ssse3
mov rsi, arg(0) ;src_ptr
movdqa xmm4, XMMWORD PTR [rax] ;k0_k5
movdqa xmm5, XMMWORD PTR [rax+256] ;k2_k4
movdqa xmm6, XMMWORD PTR [rax+128] ;k1_k3
movsxd rax, dword ptr arg(1) ;src_pixels_per_line
movsxd rcx, dword ptr arg(4) ;output_height
movsxd rdx, dword ptr arg(3) ;output_pitch
filter_block1d16_h6_rowloop_ssse3:
movdqu xmm0, XMMWORD PTR [rsi - 2]
movdqa xmm1, xmm0
pshufb xmm0, [shuf1b GLOBAL]
movdqa xmm2, xmm1
pmaddubsw xmm0, xmm4
pshufb xmm1, [shuf2b GLOBAL]
pshufb xmm2, [shuf3b GLOBAL]
pmaddubsw xmm1, xmm5
movdqu xmm3, XMMWORD PTR [rsi + 6]
pmaddubsw xmm2, xmm6
paddsw xmm0, xmm1
movdqa xmm1, xmm3
pshufb xmm3, [shuf1b GLOBAL]
paddsw xmm0, xmm7
pmaddubsw xmm3, xmm4
paddsw xmm0, xmm2
movdqa xmm2, xmm1
pshufb xmm1, [shuf2b GLOBAL]
pshufb xmm2, [shuf3b GLOBAL]
pmaddubsw xmm1, xmm5
pmaddubsw xmm2, xmm6
psraw xmm0, 7
packuswb xmm0, xmm0
lea rsi, [rsi + rax]
paddsw xmm3, xmm1
paddsw xmm3, xmm7
paddsw xmm3, xmm2
psraw xmm3, 7
packuswb xmm3, xmm3
punpcklqdq xmm0, xmm3
movdqa XMMWORD Ptr [rdi], xmm0
add rdi, rdx
dec rcx
jnz filter_block1d16_h6_rowloop_ssse3
; begin epilog
pop rdi
pop rsi
RESTORE_GOT
UNSHADOW_ARGS
pop rbp
ret
vp8_filter_block1d16_h4_ssse3:
movdqa xmm5, XMMWORD PTR [rax+256] ;k2_k4
movdqa xmm6, XMMWORD PTR [rax+128] ;k1_k3
mov rsi, arg(0) ;src_ptr
movsxd rax, dword ptr arg(1) ;src_pixels_per_line
movsxd rcx, dword ptr arg(4) ;output_height
movsxd rdx, dword ptr arg(3) ;output_pitch
filter_block1d16_h4_rowloop_ssse3:
movdqu xmm1, XMMWORD PTR [rsi - 2]
movdqa xmm2, xmm1
pshufb xmm1, [shuf2b GLOBAL]
pshufb xmm2, [shuf3b GLOBAL]
pmaddubsw xmm1, xmm5
movdqu xmm3, XMMWORD PTR [rsi + 6]
pmaddubsw xmm2, xmm6
movdqa xmm0, xmm3
pshufb xmm3, [shuf3b GLOBAL]
pshufb xmm0, [shuf2b GLOBAL]
paddsw xmm1, xmm7
paddsw xmm1, xmm2
pmaddubsw xmm0, xmm5
pmaddubsw xmm3, xmm6
psraw xmm1, 7
packuswb xmm1, xmm1
lea rsi, [rsi + rax]
paddsw xmm3, xmm0
paddsw xmm3, xmm7
psraw xmm3, 7
packuswb xmm3, xmm3
punpcklqdq xmm1, xmm3
movdqa XMMWORD Ptr [rdi], xmm1
add rdi, rdx
dec rcx
jnz filter_block1d16_h4_rowloop_ssse3
; begin epilog
pop rdi
pop rsi
RESTORE_GOT
UNSHADOW_ARGS
pop rbp
ret
;void vp8_filter_block1d4_h6_ssse3
;(
; unsigned char *src_ptr,
; unsigned int src_pixels_per_line,
; unsigned char *output_ptr,
; unsigned int output_pitch,
; unsigned int output_height,
; unsigned int vp8_filter_index
;)
global sym(vp8_filter_block1d4_h6_ssse3)
sym(vp8_filter_block1d4_h6_ssse3):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 6
GET_GOT rbx
push rsi
push rdi
; end prolog
movsxd rdx, DWORD PTR arg(5) ;table index
xor rsi, rsi
shl rdx, 4 ;
lea rax, [k0_k5 GLOBAL]
add rax, rdx
movdqa xmm7, [rd GLOBAL]
cmp esi, DWORD PTR [rax]
je vp8_filter_block1d4_h4_ssse3
movdqa xmm4, XMMWORD PTR [rax] ;k0_k5
movdqa xmm5, XMMWORD PTR [rax+256] ;k2_k4
movdqa xmm6, XMMWORD PTR [rax+128] ;k1_k3
mov rsi, arg(0) ;src_ptr
mov rdi, arg(2) ;output_ptr
movsxd rax, dword ptr arg(1) ;src_pixels_per_line
movsxd rcx, dword ptr arg(4) ;output_height
movsxd rdx, dword ptr arg(3) ;output_pitch
;xmm3 free
filter_block1d4_h6_rowloop_ssse3:
movdqu xmm0, XMMWORD PTR [rsi - 2]
movdqa xmm1, xmm0
pshufb xmm0, [shuf1b GLOBAL]
movdqa xmm2, xmm1
pshufb xmm1, [shuf2b GLOBAL]
pmaddubsw xmm0, xmm4
pshufb xmm2, [shuf3b GLOBAL]
pmaddubsw xmm1, xmm5
;--
pmaddubsw xmm2, xmm6
lea rsi, [rsi + rax]
;--
paddsw xmm0, xmm1
paddsw xmm0, xmm7
pxor xmm1, xmm1
paddsw xmm0, xmm2
psraw xmm0, 7
packuswb xmm0, xmm0
movd DWORD PTR [rdi], xmm0
add rdi, rdx
dec rcx
jnz filter_block1d4_h6_rowloop_ssse3
; begin epilog
pop rdi
pop rsi
RESTORE_GOT
UNSHADOW_ARGS
pop rbp
ret
vp8_filter_block1d4_h4_ssse3:
movdqa xmm5, XMMWORD PTR [rax+256] ;k2_k4
movdqa xmm6, XMMWORD PTR [rax+128] ;k1_k3
movdqa xmm0, XMMWORD PTR [shuf2b GLOBAL]
movdqa xmm3, XMMWORD PTR [shuf3b GLOBAL]
mov rsi, arg(0) ;src_ptr
mov rdi, arg(2) ;output_ptr
movsxd rax, dword ptr arg(1) ;src_pixels_per_line
movsxd rcx, dword ptr arg(4) ;output_height
movsxd rdx, dword ptr arg(3) ;output_pitch
filter_block1d4_h4_rowloop_ssse3:
movdqu xmm1, XMMWORD PTR [rsi - 2]
movdqa xmm2, xmm1
pshufb xmm1, xmm0 ;;[shuf2b GLOBAL]
pshufb xmm2, xmm3 ;;[shuf3b GLOBAL]
pmaddubsw xmm1, xmm5
;--
pmaddubsw xmm2, xmm6
lea rsi, [rsi + rax]
;--
paddsw xmm1, xmm7
paddsw xmm1, xmm2
psraw xmm1, 7
packuswb xmm1, xmm1
movd DWORD PTR [rdi], xmm1
add rdi, rdx
dec rcx
jnz filter_block1d4_h4_rowloop_ssse3
; begin epilog
pop rdi
pop rsi
RESTORE_GOT
UNSHADOW_ARGS
pop rbp
ret
;void vp8_filter_block1d16_v6_ssse3
;(
; unsigned char *src_ptr,
; unsigned int src_pitch,
; unsigned char *output_ptr,
; unsigned int out_pitch,
; unsigned int output_height,
; unsigned int vp8_filter_index
;)
global sym(vp8_filter_block1d16_v6_ssse3)
sym(vp8_filter_block1d16_v6_ssse3):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 6
GET_GOT rbx
push rsi
push rdi
; end prolog
movsxd rdx, DWORD PTR arg(5) ;table index
xor rsi, rsi
shl rdx, 4 ;
lea rax, [k0_k5 GLOBAL]
add rax, rdx
cmp esi, DWORD PTR [rax]
je vp8_filter_block1d16_v4_ssse3
movdqa xmm5, XMMWORD PTR [rax] ;k0_k5
movdqa xmm6, XMMWORD PTR [rax+256] ;k2_k4
movdqa xmm7, XMMWORD PTR [rax+128] ;k1_k3
mov rsi, arg(0) ;src_ptr
movsxd rdx, DWORD PTR arg(1) ;pixels_per_line
mov rdi, arg(2) ;output_ptr
%if ABI_IS_32BIT=0
movsxd r8, DWORD PTR arg(3) ;out_pitch
%endif
mov rax, rsi
movsxd rcx, DWORD PTR arg(4) ;output_height
add rax, rdx
vp8_filter_block1d16_v6_ssse3_loop:
movq xmm1, MMWORD PTR [rsi] ;A
movq xmm2, MMWORD PTR [rsi + rdx] ;B
movq xmm3, MMWORD PTR [rsi + rdx * 2] ;C
movq xmm4, MMWORD PTR [rax + rdx * 2] ;D
movq xmm0, MMWORD PTR [rsi + rdx * 4] ;E
punpcklbw xmm2, xmm4 ;B D
punpcklbw xmm3, xmm0 ;C E
movq xmm0, MMWORD PTR [rax + rdx * 4] ;F
pmaddubsw xmm3, xmm6
punpcklbw xmm1, xmm0 ;A F
pmaddubsw xmm2, xmm7
pmaddubsw xmm1, xmm5
paddsw xmm2, xmm3
paddsw xmm2, xmm1
paddsw xmm2, [rd GLOBAL]
psraw xmm2, 7
packuswb xmm2, xmm2
movq MMWORD PTR [rdi], xmm2 ;store the results
movq xmm1, MMWORD PTR [rsi + 8] ;A
movq xmm2, MMWORD PTR [rsi + rdx + 8] ;B
movq xmm3, MMWORD PTR [rsi + rdx * 2 + 8] ;C
movq xmm4, MMWORD PTR [rax + rdx * 2 + 8] ;D
movq xmm0, MMWORD PTR [rsi + rdx * 4 + 8] ;E
punpcklbw xmm2, xmm4 ;B D
punpcklbw xmm3, xmm0 ;C E
movq xmm0, MMWORD PTR [rax + rdx * 4 + 8] ;F
pmaddubsw xmm3, xmm6
punpcklbw xmm1, xmm0 ;A F
pmaddubsw xmm2, xmm7
pmaddubsw xmm1, xmm5
add rsi, rdx
add rax, rdx
;--
;--
paddsw xmm2, xmm3
paddsw xmm2, xmm1
paddsw xmm2, [rd GLOBAL]
psraw xmm2, 7
packuswb xmm2, xmm2
movq MMWORD PTR [rdi+8], xmm2
%if ABI_IS_32BIT
add rdi, DWORD PTR arg(3) ;out_pitch
%else
add rdi, r8
%endif
dec rcx
jnz vp8_filter_block1d16_v6_ssse3_loop
; begin epilog
pop rdi
pop rsi
RESTORE_GOT
UNSHADOW_ARGS
pop rbp
ret
vp8_filter_block1d16_v4_ssse3:
movdqa xmm6, XMMWORD PTR [rax+256] ;k2_k4
movdqa xmm7, XMMWORD PTR [rax+128] ;k1_k3
mov rsi, arg(0) ;src_ptr
movsxd rdx, DWORD PTR arg(1) ;pixels_per_line
mov rdi, arg(2) ;output_ptr
%if ABI_IS_32BIT=0
movsxd r8, DWORD PTR arg(3) ;out_pitch
%endif
mov rax, rsi
movsxd rcx, DWORD PTR arg(4) ;output_height
add rax, rdx
vp8_filter_block1d16_v4_ssse3_loop:
movq xmm2, MMWORD PTR [rsi + rdx] ;B
movq xmm3, MMWORD PTR [rsi + rdx * 2] ;C
movq xmm4, MMWORD PTR [rax + rdx * 2] ;D
movq xmm0, MMWORD PTR [rsi + rdx * 4] ;E
punpcklbw xmm2, xmm4 ;B D
punpcklbw xmm3, xmm0 ;C E
pmaddubsw xmm3, xmm6
pmaddubsw xmm2, xmm7
movq xmm5, MMWORD PTR [rsi + rdx + 8] ;B
movq xmm1, MMWORD PTR [rsi + rdx * 2 + 8] ;C
movq xmm4, MMWORD PTR [rax + rdx * 2 + 8] ;D
movq xmm0, MMWORD PTR [rsi + rdx * 4 + 8] ;E
paddsw xmm2, [rd GLOBAL]
paddsw xmm2, xmm3
psraw xmm2, 7
packuswb xmm2, xmm2
punpcklbw xmm5, xmm4 ;B D
punpcklbw xmm1, xmm0 ;C E
pmaddubsw xmm1, xmm6
pmaddubsw xmm5, xmm7
movdqa xmm4, [rd GLOBAL]
add rsi, rdx
add rax, rdx
;--
;--
paddsw xmm5, xmm1
paddsw xmm5, xmm4
psraw xmm5, 7
packuswb xmm5, xmm5
punpcklqdq xmm2, xmm5
movdqa XMMWORD PTR [rdi], xmm2
%if ABI_IS_32BIT
add rdi, DWORD PTR arg(3) ;out_pitch
%else
add rdi, r8
%endif
dec rcx
jnz vp8_filter_block1d16_v4_ssse3_loop
; begin epilog
pop rdi
pop rsi
RESTORE_GOT
UNSHADOW_ARGS
pop rbp
ret
;void vp8_filter_block1d8_v6_ssse3
;(
; unsigned char *src_ptr,
; unsigned int src_pitch,
; unsigned char *output_ptr,
; unsigned int out_pitch,
; unsigned int output_height,
; unsigned int vp8_filter_index
;)
global sym(vp8_filter_block1d8_v6_ssse3)
sym(vp8_filter_block1d8_v6_ssse3):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 6
GET_GOT rbx
push rsi
push rdi
; end prolog
movsxd rdx, DWORD PTR arg(5) ;table index
xor rsi, rsi
shl rdx, 4 ;
lea rax, [k0_k5 GLOBAL]
add rax, rdx
movsxd rdx, DWORD PTR arg(1) ;pixels_per_line
mov rdi, arg(2) ;output_ptr
%if ABI_IS_32BIT=0
movsxd r8, DWORD PTR arg(3) ; out_pitch
%endif
movsxd rcx, DWORD PTR arg(4) ;[output_height]
cmp esi, DWORD PTR [rax]
je vp8_filter_block1d8_v4_ssse3
movdqa xmm5, XMMWORD PTR [rax] ;k0_k5
movdqa xmm6, XMMWORD PTR [rax+256] ;k2_k4
movdqa xmm7, XMMWORD PTR [rax+128] ;k1_k3
mov rsi, arg(0) ;src_ptr
mov rax, rsi
add rax, rdx
vp8_filter_block1d8_v6_ssse3_loop:
movq xmm1, MMWORD PTR [rsi] ;A
movq xmm2, MMWORD PTR [rsi + rdx] ;B
movq xmm3, MMWORD PTR [rsi + rdx * 2] ;C
movq xmm4, MMWORD PTR [rax + rdx * 2] ;D
movq xmm0, MMWORD PTR [rsi + rdx * 4] ;E
punpcklbw xmm2, xmm4 ;B D
punpcklbw xmm3, xmm0 ;C E
movq xmm0, MMWORD PTR [rax + rdx * 4] ;F
movdqa xmm4, [rd GLOBAL]
pmaddubsw xmm3, xmm6
punpcklbw xmm1, xmm0 ;A F
pmaddubsw xmm2, xmm7
pmaddubsw xmm1, xmm5
add rsi, rdx
add rax, rdx
;--
;--
paddsw xmm2, xmm3
paddsw xmm2, xmm1
paddsw xmm2, xmm4
psraw xmm2, 7
packuswb xmm2, xmm2
movq MMWORD PTR [rdi], xmm2
%if ABI_IS_32BIT
add rdi, DWORD PTR arg(3) ;[out_pitch]
%else
add rdi, r8
%endif
dec rcx
jnz vp8_filter_block1d8_v6_ssse3_loop
; begin epilog
pop rdi
pop rsi
RESTORE_GOT
UNSHADOW_ARGS
pop rbp
ret
vp8_filter_block1d8_v4_ssse3:
movdqa xmm6, XMMWORD PTR [rax+256] ;k2_k4
movdqa xmm7, XMMWORD PTR [rax+128] ;k1_k3
movdqa xmm5, [rd GLOBAL]
mov rsi, arg(0) ;src_ptr
mov rax, rsi
add rax, rdx
vp8_filter_block1d8_v4_ssse3_loop:
movq xmm2, MMWORD PTR [rsi + rdx] ;B
movq xmm3, MMWORD PTR [rsi + rdx * 2] ;C
movq xmm4, MMWORD PTR [rax + rdx * 2] ;D
movq xmm0, MMWORD PTR [rsi + rdx * 4] ;E
punpcklbw xmm2, xmm4 ;B D
punpcklbw xmm3, xmm0 ;C E
pmaddubsw xmm3, xmm6
pmaddubsw xmm2, xmm7
add rsi, rdx
add rax, rdx
;--
;--
paddsw xmm2, xmm3
paddsw xmm2, xmm5
psraw xmm2, 7
packuswb xmm2, xmm2
movq MMWORD PTR [rdi], xmm2
%if ABI_IS_32BIT
add rdi, DWORD PTR arg(3) ;[out_pitch]
%else
add rdi, r8
%endif
dec rcx
jnz vp8_filter_block1d8_v4_ssse3_loop
; begin epilog
pop rdi
pop rsi
RESTORE_GOT
UNSHADOW_ARGS
pop rbp
ret
;void vp8_filter_block1d4_v6_ssse3
;(
; unsigned char *src_ptr,
; unsigned int src_pitch,
; unsigned char *output_ptr,
; unsigned int out_pitch,
; unsigned int output_height,
; unsigned int vp8_filter_index
;)
global sym(vp8_filter_block1d4_v6_ssse3)
sym(vp8_filter_block1d4_v6_ssse3):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 6
GET_GOT rbx
push rsi
push rdi
; end prolog
movsxd rdx, DWORD PTR arg(5) ;table index
xor rsi, rsi
shl rdx, 4 ;
lea rax, [k0_k5 GLOBAL]
add rax, rdx
movsxd rdx, DWORD PTR arg(1) ;pixels_per_line
mov rdi, arg(2) ;output_ptr
%if ABI_IS_32BIT=0
movsxd r8, DWORD PTR arg(3) ; out_pitch
%endif
movsxd rcx, DWORD PTR arg(4) ;[output_height]
cmp esi, DWORD PTR [rax]
je vp8_filter_block1d4_v4_ssse3
movq mm5, MMWORD PTR [rax] ;k0_k5
movq mm6, MMWORD PTR [rax+256] ;k2_k4
movq mm7, MMWORD PTR [rax+128] ;k1_k3
mov rsi, arg(0) ;src_ptr
mov rax, rsi
add rax, rdx
vp8_filter_block1d4_v6_ssse3_loop:
movd mm1, DWORD PTR [rsi] ;A
movd mm2, DWORD PTR [rsi + rdx] ;B
movd mm3, DWORD PTR [rsi + rdx * 2] ;C
movd mm4, DWORD PTR [rax + rdx * 2] ;D
movd mm0, DWORD PTR [rsi + rdx * 4] ;E
punpcklbw mm2, mm4 ;B D
punpcklbw mm3, mm0 ;C E
movd mm0, DWORD PTR [rax + rdx * 4] ;F
movq mm4, [rd GLOBAL]
pmaddubsw mm3, mm6
punpcklbw mm1, mm0 ;A F
pmaddubsw mm2, mm7
pmaddubsw mm1, mm5
add rsi, rdx
add rax, rdx
;--
;--
paddsw mm2, mm3
paddsw mm2, mm1
paddsw mm2, mm4
psraw mm2, 7
packuswb mm2, mm2
movd DWORD PTR [rdi], mm2
%if ABI_IS_32BIT
add rdi, DWORD PTR arg(3) ;[out_pitch]
%else
add rdi, r8
%endif
dec rcx
jnz vp8_filter_block1d4_v6_ssse3_loop
; begin epilog
pop rdi
pop rsi
RESTORE_GOT
UNSHADOW_ARGS
pop rbp
ret
vp8_filter_block1d4_v4_ssse3:
movq mm6, MMWORD PTR [rax+256] ;k2_k4
movq mm7, MMWORD PTR [rax+128] ;k1_k3
movq mm5, MMWORD PTR [rd GLOBAL]
mov rsi, arg(0) ;src_ptr
mov rax, rsi
add rax, rdx
vp8_filter_block1d4_v4_ssse3_loop:
movd mm2, DWORD PTR [rsi + rdx] ;B
movd mm3, DWORD PTR [rsi + rdx * 2] ;C
movd mm4, DWORD PTR [rax + rdx * 2] ;D
movd mm0, DWORD PTR [rsi + rdx * 4] ;E
punpcklbw mm2, mm4 ;B D
punpcklbw mm3, mm0 ;C E
pmaddubsw mm3, mm6
pmaddubsw mm2, mm7
add rsi, rdx
add rax, rdx
;--
;--
paddsw mm2, mm3
paddsw mm2, mm5
psraw mm2, 7
packuswb mm2, mm2
movd DWORD PTR [rdi], mm2
%if ABI_IS_32BIT
add rdi, DWORD PTR arg(3) ;[out_pitch]
%else
add rdi, r8
%endif
dec rcx
jnz vp8_filter_block1d4_v4_ssse3_loop
; begin epilog
pop rdi
pop rsi
RESTORE_GOT
UNSHADOW_ARGS
pop rbp
ret
SECTION_RODATA
align 16
shuf1b:
db 0, 5, 1, 6, 2, 7, 3, 8, 4, 9, 5, 10, 6, 11, 7, 12
shuf2b:
db 2, 4, 3, 5, 4, 6, 5, 7, 6, 8, 7, 9, 8, 10, 9, 11
shuf3b:
db 1, 3, 2, 4, 3, 5, 4, 6, 5, 7, 6, 8, 7, 9, 8, 10
align 16
rd:
times 8 dw 0x40
align 16
k0_k5:
times 8 db 0, 0 ;placeholder
times 8 db 0, 0
times 8 db 2, 1
times 8 db 0, 0
times 8 db 3, 3
times 8 db 0, 0
times 8 db 1, 2
times 8 db 0, 0
k1_k3:
times 8 db 0, 0 ;placeholder
times 8 db -6, 12
times 8 db -11, 36
times 8 db -9, 50
times 8 db -16, 77
times 8 db -6, 93
times 8 db -8, 108
times 8 db -1, 123
k2_k4:
times 8 db 128, 0 ;placeholder
times 8 db 123, -1
times 8 db 108, -8
times 8 db 93, -6
times 8 db 77, -16
times 8 db 50, -9
times 8 db 36, -11
times 8 db 12, -6
|
