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
|
;
; Copyright (c) 2010 The WebM 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"
;void vp9_recon2b_sse2(unsigned char *s, short *q, unsigned char *d, int stride)
global sym(vp9_recon2b_sse2)
sym(vp9_recon2b_sse2):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 4
push rsi
push rdi
; end prolog
mov rsi, arg(0) ;s
mov rdi, arg(2) ;d
mov rdx, arg(1) ;q
movsxd rax, dword ptr arg(3) ;stride
pxor xmm0, xmm0
movq xmm1, MMWORD PTR [rsi]
punpcklbw xmm1, xmm0
paddsw xmm1, XMMWORD PTR [rdx]
packuswb xmm1, xmm0 ; pack and unpack to saturate
movq MMWORD PTR [rdi], xmm1
movq xmm2, MMWORD PTR [rsi+8]
punpcklbw xmm2, xmm0
paddsw xmm2, XMMWORD PTR [rdx+16]
packuswb xmm2, xmm0 ; pack and unpack to saturate
movq MMWORD PTR [rdi+rax], xmm2
movq xmm3, MMWORD PTR [rsi+16]
punpcklbw xmm3, xmm0
paddsw xmm3, XMMWORD PTR [rdx+32]
packuswb xmm3, xmm0 ; pack and unpack to saturate
movq MMWORD PTR [rdi+rax*2], xmm3
add rdi, rax
movq xmm4, MMWORD PTR [rsi+24]
punpcklbw xmm4, xmm0
paddsw xmm4, XMMWORD PTR [rdx+48]
packuswb xmm4, xmm0 ; pack and unpack to saturate
movq MMWORD PTR [rdi+rax*2], xmm4
; begin epilog
pop rdi
pop rsi
UNSHADOW_ARGS
pop rbp
ret
;void vp9_recon4b_sse2(unsigned char *s, short *q, unsigned char *d, int stride)
global sym(vp9_recon4b_sse2)
sym(vp9_recon4b_sse2):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 4
SAVE_XMM 7
push rsi
push rdi
; end prolog
mov rsi, arg(0) ;s
mov rdi, arg(2) ;d
mov rdx, arg(1) ;q
movsxd rax, dword ptr arg(3) ;stride
pxor xmm0, xmm0
movdqa xmm1, XMMWORD PTR [rsi]
movdqa xmm5, xmm1
punpcklbw xmm1, xmm0
punpckhbw xmm5, xmm0
paddsw xmm1, XMMWORD PTR [rdx]
paddsw xmm5, XMMWORD PTR [rdx+16]
packuswb xmm1, xmm5 ; pack and unpack to saturate
movdqa XMMWORD PTR [rdi], xmm1
movdqa xmm2, XMMWORD PTR [rsi+16]
movdqa xmm6, xmm2
punpcklbw xmm2, xmm0
punpckhbw xmm6, xmm0
paddsw xmm2, XMMWORD PTR [rdx+32]
paddsw xmm6, XMMWORD PTR [rdx+48]
packuswb xmm2, xmm6 ; pack and unpack to saturate
movdqa XMMWORD PTR [rdi+rax], xmm2
movdqa xmm3, XMMWORD PTR [rsi+32]
movdqa xmm7, xmm3
punpcklbw xmm3, xmm0
punpckhbw xmm7, xmm0
paddsw xmm3, XMMWORD PTR [rdx+64]
paddsw xmm7, XMMWORD PTR [rdx+80]
packuswb xmm3, xmm7 ; pack and unpack to saturate
movdqa XMMWORD PTR [rdi+rax*2], xmm3
add rdi, rax
movdqa xmm4, XMMWORD PTR [rsi+48]
movdqa xmm5, xmm4
punpcklbw xmm4, xmm0
punpckhbw xmm5, xmm0
paddsw xmm4, XMMWORD PTR [rdx+96]
paddsw xmm5, XMMWORD PTR [rdx+112]
packuswb xmm4, xmm5 ; pack and unpack to saturate
movdqa XMMWORD PTR [rdi+rax*2], xmm4
; begin epilog
pop rdi
pop rsi
RESTORE_XMM
UNSHADOW_ARGS
pop rbp
ret
;void copy_mem16x16_sse2(
; unsigned char *src,
; int src_stride,
; unsigned char *dst,
; int dst_stride
; )
global sym(vp9_copy_mem16x16_sse2)
sym(vp9_copy_mem16x16_sse2):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 4
push rsi
push rdi
; end prolog
mov rsi, arg(0) ;src;
movdqu xmm0, [rsi]
movsxd rax, dword ptr arg(1) ;src_stride;
mov rdi, arg(2) ;dst;
movdqu xmm1, [rsi+rax]
movdqu xmm2, [rsi+rax*2]
movsxd rcx, dword ptr arg(3) ;dst_stride
lea rsi, [rsi+rax*2]
movdqa [rdi], xmm0
add rsi, rax
movdqa [rdi+rcx], xmm1
movdqa [rdi+rcx*2],xmm2
lea rdi, [rdi+rcx*2]
movdqu xmm3, [rsi]
add rdi, rcx
movdqu xmm4, [rsi+rax]
movdqu xmm5, [rsi+rax*2]
lea rsi, [rsi+rax*2]
movdqa [rdi], xmm3
add rsi, rax
movdqa [rdi+rcx], xmm4
movdqa [rdi+rcx*2],xmm5
lea rdi, [rdi+rcx*2]
movdqu xmm0, [rsi]
add rdi, rcx
movdqu xmm1, [rsi+rax]
movdqu xmm2, [rsi+rax*2]
lea rsi, [rsi+rax*2]
movdqa [rdi], xmm0
add rsi, rax
movdqa [rdi+rcx], xmm1
movdqa [rdi+rcx*2], xmm2
movdqu xmm3, [rsi]
movdqu xmm4, [rsi+rax]
lea rdi, [rdi+rcx*2]
add rdi, rcx
movdqu xmm5, [rsi+rax*2]
lea rsi, [rsi+rax*2]
movdqa [rdi], xmm3
add rsi, rax
movdqa [rdi+rcx], xmm4
movdqa [rdi+rcx*2],xmm5
movdqu xmm0, [rsi]
lea rdi, [rdi+rcx*2]
movdqu xmm1, [rsi+rax]
add rdi, rcx
movdqu xmm2, [rsi+rax*2]
lea rsi, [rsi+rax*2]
movdqa [rdi], xmm0
movdqa [rdi+rcx], xmm1
movdqa [rdi+rcx*2],xmm2
movdqu xmm3, [rsi+rax]
lea rdi, [rdi+rcx*2]
movdqa [rdi+rcx], xmm3
; begin epilog
pop rdi
pop rsi
UNSHADOW_ARGS
pop rbp
ret
;void vp9_intra_pred_uv_dc_mmx2(
; unsigned char *dst,
; int dst_stride
; unsigned char *src,
; int src_stride,
; )
global sym(vp9_intra_pred_uv_dc_mmx2)
sym(vp9_intra_pred_uv_dc_mmx2):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 4
push rsi
push rdi
; end prolog
; from top
mov rsi, arg(2) ;src;
movsxd rax, dword ptr arg(3) ;src_stride;
sub rsi, rax
pxor mm0, mm0
movq mm1, [rsi]
psadbw mm1, mm0
; from left
dec rsi
lea rdi, [rax*3]
movzx ecx, byte [rsi+rax]
movzx edx, byte [rsi+rax*2]
add ecx, edx
movzx edx, byte [rsi+rdi]
add ecx, edx
lea rsi, [rsi+rax*4]
movzx edx, byte [rsi]
add ecx, edx
movzx edx, byte [rsi+rax]
add ecx, edx
movzx edx, byte [rsi+rax*2]
add ecx, edx
movzx edx, byte [rsi+rdi]
add ecx, edx
movzx edx, byte [rsi+rax*4]
add ecx, edx
; add up
pextrw edx, mm1, 0x0
lea edx, [edx+ecx+8]
sar edx, 4
movd mm1, edx
pshufw mm1, mm1, 0x0
packuswb mm1, mm1
; write out
mov rdi, arg(0) ;dst;
movsxd rcx, dword ptr arg(1) ;dst_stride
lea rax, [rcx*3]
movq [rdi ], mm1
movq [rdi+rcx ], mm1
movq [rdi+rcx*2], mm1
movq [rdi+rax ], mm1
lea rdi, [rdi+rcx*4]
movq [rdi ], mm1
movq [rdi+rcx ], mm1
movq [rdi+rcx*2], mm1
movq [rdi+rax ], mm1
; begin epilog
pop rdi
pop rsi
UNSHADOW_ARGS
pop rbp
ret
;void vp9_intra_pred_uv_dctop_mmx2(
; unsigned char *dst,
; int dst_stride
; unsigned char *src,
; int src_stride,
; )
global sym(vp9_intra_pred_uv_dctop_mmx2)
sym(vp9_intra_pred_uv_dctop_mmx2):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 4
GET_GOT rbx
push rsi
push rdi
; end prolog
; from top
mov rsi, arg(2) ;src;
movsxd rax, dword ptr arg(3) ;src_stride;
sub rsi, rax
pxor mm0, mm0
movq mm1, [rsi]
psadbw mm1, mm0
; add up
paddw mm1, [GLOBAL(dc_4)]
psraw mm1, 3
pshufw mm1, mm1, 0x0
packuswb mm1, mm1
; write out
mov rdi, arg(0) ;dst;
movsxd rcx, dword ptr arg(1) ;dst_stride
lea rax, [rcx*3]
movq [rdi ], mm1
movq [rdi+rcx ], mm1
movq [rdi+rcx*2], mm1
movq [rdi+rax ], mm1
lea rdi, [rdi+rcx*4]
movq [rdi ], mm1
movq [rdi+rcx ], mm1
movq [rdi+rcx*2], mm1
movq [rdi+rax ], mm1
; begin epilog
pop rdi
pop rsi
RESTORE_GOT
UNSHADOW_ARGS
pop rbp
ret
;void vp9_intra_pred_uv_dcleft_mmx2(
; unsigned char *dst,
; int dst_stride
; unsigned char *src,
; int src_stride,
; )
global sym(vp9_intra_pred_uv_dcleft_mmx2)
sym(vp9_intra_pred_uv_dcleft_mmx2):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 4
push rsi
push rdi
; end prolog
; from left
mov rsi, arg(2) ;src;
movsxd rax, dword ptr arg(3) ;src_stride;
dec rsi
lea rdi, [rax*3]
movzx ecx, byte [rsi]
movzx edx, byte [rsi+rax]
add ecx, edx
movzx edx, byte [rsi+rax*2]
add ecx, edx
movzx edx, byte [rsi+rdi]
add ecx, edx
lea rsi, [rsi+rax*4]
movzx edx, byte [rsi]
add ecx, edx
movzx edx, byte [rsi+rax]
add ecx, edx
movzx edx, byte [rsi+rax*2]
add ecx, edx
movzx edx, byte [rsi+rdi]
lea edx, [ecx+edx+4]
; add up
shr edx, 3
movd mm1, edx
pshufw mm1, mm1, 0x0
packuswb mm1, mm1
; write out
mov rdi, arg(0) ;dst;
movsxd rcx, dword ptr arg(1) ;dst_stride
lea rax, [rcx*3]
movq [rdi ], mm1
movq [rdi+rcx ], mm1
movq [rdi+rcx*2], mm1
movq [rdi+rax ], mm1
lea rdi, [rdi+rcx*4]
movq [rdi ], mm1
movq [rdi+rcx ], mm1
movq [rdi+rcx*2], mm1
movq [rdi+rax ], mm1
; begin epilog
pop rdi
pop rsi
UNSHADOW_ARGS
pop rbp
ret
;void vp9_intra_pred_uv_dc128_mmx(
; unsigned char *dst,
; int dst_stride
; unsigned char *src,
; int src_stride,
; )
global sym(vp9_intra_pred_uv_dc128_mmx)
sym(vp9_intra_pred_uv_dc128_mmx):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 4
GET_GOT rbx
; end prolog
; write out
movq mm1, [GLOBAL(dc_128)]
mov rax, arg(0) ;dst;
movsxd rdx, dword ptr arg(1) ;dst_stride
lea rcx, [rdx*3]
movq [rax ], mm1
movq [rax+rdx ], mm1
movq [rax+rdx*2], mm1
movq [rax+rcx ], mm1
lea rax, [rax+rdx*4]
movq [rax ], mm1
movq [rax+rdx ], mm1
movq [rax+rdx*2], mm1
movq [rax+rcx ], mm1
; begin epilog
RESTORE_GOT
UNSHADOW_ARGS
pop rbp
ret
;void vp9_intra_pred_uv_tm_sse2(
; unsigned char *dst,
; int dst_stride
; unsigned char *src,
; int src_stride,
; )
%macro vp9_intra_pred_uv_tm 1
global sym(vp9_intra_pred_uv_tm_%1)
sym(vp9_intra_pred_uv_tm_%1):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 4
GET_GOT rbx
push rsi
push rdi
; end prolog
; read top row
mov edx, 4
mov rsi, arg(2) ;src;
movsxd rax, dword ptr arg(3) ;src_stride;
sub rsi, rax
pxor xmm0, xmm0
%ifidn %1, ssse3
movdqa xmm2, [GLOBAL(dc_1024)]
%endif
movq xmm1, [rsi]
punpcklbw xmm1, xmm0
; set up left ptrs ans subtract topleft
movd xmm3, [rsi-1]
lea rsi, [rsi+rax-1]
%ifidn %1, sse2
punpcklbw xmm3, xmm0
pshuflw xmm3, xmm3, 0x0
punpcklqdq xmm3, xmm3
%else
pshufb xmm3, xmm2
%endif
psubw xmm1, xmm3
; set up dest ptrs
mov rdi, arg(0) ;dst;
movsxd rcx, dword ptr arg(1) ;dst_stride
.vp9_intra_pred_uv_tm_%1_loop:
movd xmm3, [rsi]
movd xmm5, [rsi+rax]
%ifidn %1, sse2
punpcklbw xmm3, xmm0
punpcklbw xmm5, xmm0
pshuflw xmm3, xmm3, 0x0
pshuflw xmm5, xmm5, 0x0
punpcklqdq xmm3, xmm3
punpcklqdq xmm5, xmm5
%else
pshufb xmm3, xmm2
pshufb xmm5, xmm2
%endif
paddw xmm3, xmm1
paddw xmm5, xmm1
packuswb xmm3, xmm5
movq [rdi ], xmm3
movhps[rdi+rcx], xmm3
lea rsi, [rsi+rax*2]
lea rdi, [rdi+rcx*2]
dec edx
jnz .vp9_intra_pred_uv_tm_%1_loop
; begin epilog
pop rdi
pop rsi
RESTORE_GOT
UNSHADOW_ARGS
pop rbp
ret
%endmacro
vp9_intra_pred_uv_tm sse2
vp9_intra_pred_uv_tm ssse3
;void vp9_intra_pred_uv_ve_mmx(
; unsigned char *dst,
; int dst_stride
; unsigned char *src,
; int src_stride,
; )
global sym(vp9_intra_pred_uv_ve_mmx)
sym(vp9_intra_pred_uv_ve_mmx):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 4
; end prolog
; read from top
mov rax, arg(2) ;src;
movsxd rdx, dword ptr arg(3) ;src_stride;
sub rax, rdx
movq mm1, [rax]
; write out
mov rax, arg(0) ;dst;
movsxd rdx, dword ptr arg(1) ;dst_stride
lea rcx, [rdx*3]
movq [rax ], mm1
movq [rax+rdx ], mm1
movq [rax+rdx*2], mm1
movq [rax+rcx ], mm1
lea rax, [rax+rdx*4]
movq [rax ], mm1
movq [rax+rdx ], mm1
movq [rax+rdx*2], mm1
movq [rax+rcx ], mm1
; begin epilog
UNSHADOW_ARGS
pop rbp
ret
;void vp9_intra_pred_uv_ho_mmx2(
; unsigned char *dst,
; int dst_stride
; unsigned char *src,
; int src_stride,
; )
%macro vp9_intra_pred_uv_ho 1
global sym(vp9_intra_pred_uv_ho_%1)
sym(vp9_intra_pred_uv_ho_%1):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 4
push rsi
push rdi
%ifidn %1, ssse3
%ifndef GET_GOT_SAVE_ARG
push rbx
%endif
GET_GOT rbx
%endif
; end prolog
; read from left and write out
%ifidn %1, mmx2
mov edx, 4
%endif
mov rsi, arg(2) ;src;
movsxd rax, dword ptr arg(3) ;src_stride;
mov rdi, arg(0) ;dst;
movsxd rcx, dword ptr arg(1) ;dst_stride
%ifidn %1, ssse3
lea rdx, [rcx*3]
movdqa xmm2, [GLOBAL(dc_00001111)]
lea rbx, [rax*3]
%endif
dec rsi
%ifidn %1, mmx2
.vp9_intra_pred_uv_ho_%1_loop:
movd mm0, [rsi]
movd mm1, [rsi+rax]
punpcklbw mm0, mm0
punpcklbw mm1, mm1
pshufw mm0, mm0, 0x0
pshufw mm1, mm1, 0x0
movq [rdi ], mm0
movq [rdi+rcx], mm1
lea rsi, [rsi+rax*2]
lea rdi, [rdi+rcx*2]
dec edx
jnz .vp9_intra_pred_uv_ho_%1_loop
%else
movd xmm0, [rsi]
movd xmm3, [rsi+rax]
movd xmm1, [rsi+rax*2]
movd xmm4, [rsi+rbx]
punpcklbw xmm0, xmm3
punpcklbw xmm1, xmm4
pshufb xmm0, xmm2
pshufb xmm1, xmm2
movq [rdi ], xmm0
movhps [rdi+rcx], xmm0
movq [rdi+rcx*2], xmm1
movhps [rdi+rdx], xmm1
lea rsi, [rsi+rax*4]
lea rdi, [rdi+rcx*4]
movd xmm0, [rsi]
movd xmm3, [rsi+rax]
movd xmm1, [rsi+rax*2]
movd xmm4, [rsi+rbx]
punpcklbw xmm0, xmm3
punpcklbw xmm1, xmm4
pshufb xmm0, xmm2
pshufb xmm1, xmm2
movq [rdi ], xmm0
movhps [rdi+rcx], xmm0
movq [rdi+rcx*2], xmm1
movhps [rdi+rdx], xmm1
%endif
; begin epilog
%ifidn %1, ssse3
RESTORE_GOT
%ifndef GET_GOT_SAVE_ARG
pop rbx
%endif
%endif
pop rdi
pop rsi
UNSHADOW_ARGS
pop rbp
ret
%endmacro
vp9_intra_pred_uv_ho mmx2
vp9_intra_pred_uv_ho ssse3
SECTION_RODATA
dc_128:
times 8 db 128
dc_4:
times 4 dw 4
align 16
dc_1024:
times 8 dw 0x400
align 16
dc_00001111:
times 8 db 0
times 8 db 1
|
