-
Notifications
You must be signed in to change notification settings - Fork 3
/
Copy pathfft2mmxa.s
785 lines (764 loc) · 22.9 KB
/
fft2mmxa.s
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
; Copyright (c) <2012> <Leif Asbrink>
;
; Permission is hereby granted, free of charge, to any person
; obtaining a copy of this software and associated documentation
; files (the "Software"), to deal in the Software without restriction,
; including without limitation the rights to use, copy, modify,
; merge, publish, distribute, sublicense, and/or sell copies of
; the Software, and to permit persons to whom the Software is
; furnished to do so, subject to the following conditions:
;
; The above copyright notice and this permission notice shall be
; included in all copies or substantial portions of the Software.
;
; THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
; EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES
; OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
; NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT
; HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY,
; WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
; FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE
; OR OTHER DEALINGS IN THE SOFTWARE.
section .note.GNU-stack
extern fft2_short_int
extern fft2_size
extern fft2_bigpermute
extern fft2_na
extern timf2_shi
extern timf2_mask
extern timf2_px
extern fft2_mmxwin
extern mailbox
global fft2mmx_a2_nowin
global fft2mmx_a1_nowin
global fft2mmx_a2_win
global fft2mmx_a1_win
%define NEGODD_XOR ebp-8
%define NEGODD_INC ebp-16
%define NEGEVEN_XOR ebp-24
%define NEGEVEN_INC ebp-32
%define NEGMID_XOR ebp-40
%define NEGMID_INC ebp-48
%define MASK ebp-52
%define P0 ebp-56
%define LIMIT ebp-60
section .text
fft2mmx_a1_win
push ebx
push ecx
push edx
push esi
push edi
push ebp
mov ebp,[fft2_bigpermute]
mov eax,[timf2_mask]
sal eax,1
mov [MASK],eax
mov eax,[timf2_px]
sal eax,1
mov [P0],eax
mov eax,[fft2_size]
sal eax,1
mov [LIMIT],eax
;zxy=&fft2_short_int[2*ui.rx_channels*fft2_na*fft2_size];
sal eax,1
imul eax,[fft2_na]
mov edx,[fft2_short_int]
add edx,eax
mov ebx,[timf2_shi]
; The two first trivial loops are combined and
; run together with the unscrambling process
xor eax,eax
mov ecx,0ffffffffh
mov [edx+4],eax
mov [edx],ecx
movq mm7,[edx]
movq mm6,[edx]
psllq mm6,32
mov ecx,[fft2_mmxwin]
win_one:
;for(j=0; j<siz; j+=4)
; {
;ia=(p0+fft2_bigpermute[j ])&mask;
;ib=(p0+fft2_bigpermute[j+1])&mask;
movq mm5,[ecx+eax] ;mm5 = win4 win3 win2 win1
psrlq mm5,16 ;mm5 = 0 win4 win3 win2
movq mm4,[ecx+eax]
punpcklwd mm4,mm4 ;mm4 = . . win1 win1
movq mm3,mm5 ;mm3 = 0 win4 win3 win2
punpcklwd mm3,mm3 ;mm3 = . . win2 win2
punpckldq mm4,mm3 ;mm4 = win2 win2 win1 win1
mov esi,[2*eax+ebp] ;esi=fft2_bigpermute[j]
mov edi,[2*eax+ebp+4] ;edi=fft2_bigpermute[j+1]
add esi,[P0]
add edi,[P0]
and esi,[MASK]
and edi,[MASK]
; a1=fft2_window[ja]*(timf2_shi[ia ]+timf2_shi[ia+2]);
; a2=fft2_window[ja]*(timf2_shi[ia+1]+timf2_shi[ia+3]);
; b1=fft2_window[jb]*(timf2_shi[ib ]+timf2_shi[ib+2]);
; b2=fft2_window[jb]*(timf2_shi[ib+1]+timf2_shi[ib+3]);
movq mm0,[esi+ebx] ;mm0= ia+3 ia+2 ia+1 ia+0
movq mm1,[esi+ebx] ;mm1= ia+3 ia+2 ia+1 ia+0
movq mm2,[edi+ebx] ;mm2= ib+3 ib+2 ib+1 ib+0
movq mm3,[edi+ebx] ;mm3= ib+3 ib+2 ib+1 ib+0
psrlq mm1,32 ;mm1= 0 0 ia+3 ia+2
psllq mm3,32 ;mm3= ib+1 ib+0 0 0
pand mm0,mm7 ;mm0= 0 0 ia+1 ia+0
pand mm2,mm6 ;mm2= ib+3 ib+2 0 0
paddsw mm0,mm1
paddsw mm0,mm2
paddsw mm0,mm3 ;mm0= sums
movq mm1,mm4
pmulhw mm4,mm0
paddsw mm4,mm4
pmullw mm0,mm1
psraw mm0,13
paddsw mm0,mm4 ;mm0= b2 b1 a2 a1
psrlq mm5,16 ;mm5 = 0 0 win4 win3
movq mm4,mm5
psrlq mm4,16 ;mm4 = 0 0 0 win4
punpcklwd mm5,mm5 ;mm5 = . . win3 win3
punpcklwd mm4,mm4 ;mm3 = . . win4 win4
punpckldq mm5,mm4 ;mm5 = win4 win4 win3 win3
mov esi,[2*eax+ebp+8] ;esi=fft2_bigpermute[j+2]
mov edi,[2*eax+ebp+12] ;edi=fft2_bigpermute[j+3]
;ic=(p0+fft2_bigpermute[j+2])&mask;
;id=(p0+fft2_bigpermute[j+3])&mask;
add esi,[P0]
add edi,[P0]
and esi,[MASK]
and edi,[MASK]
; c1=fft2_window[jc]*(timf2_shi[ic ]+timf2_shi[ic+2]);
; c2=fft2_window[jc]*(timf2_shi[ic+1]+timf2_shi[ic+3]);
; d1=fft2_window[jd]*(timf2_shi[id ]+timf2_shi[id+2]);
; d2=fft2_window[jd]*(timf2_shi[id+1]+timf2_shi[id+3]);
movq mm1,[esi+ebx] ;mm1= ic+3 ic+2 ic+1 ic+0
movq mm2,[esi+ebx] ;mm2= ic+3 ic+2 ic+1 ic+0
movq mm3,[edi+ebx] ;mm3= id+3 id+2 id+1 id+0
movq mm4,[edi+ebx] ;mm4= id+3 id+2 id+1 id+0
psrlq mm2,32 ;mm2= 0 0 ic+3 ic+2
psllq mm4,32 ;mm4= id+1 id+0 0 0
pand mm1,mm7 ;mm1= 0 0 ic+1 ic+0
pand mm3,mm6 ;mm3= id+3 id+2 0 0
paddsw mm1,mm2
paddsw mm1,mm3
paddsw mm1,mm4 ;mm1= sums
movq mm2,mm5
pmulhw mm5,mm1
paddsw mm5,mm5
pmullw mm1,mm2
psraw mm1,13
paddsw mm1,mm5 ;mm1= d2 d1 c2 c1
; t1=a1+b1;
; t2=a2+b2;
; t5=-(a1-b1);
; t7=-(a2-b2);
;mm0= b2 b1 a2 a1
movq mm2,mm0
movq mm3,mm0
psrlq mm2,32 ;mm2= 0 0 b2 b1
psllq mm3,32 ;mm3= a2 a1 0 0
paddsw mm0,mm2 ;mm0= b2 b1 a2+b2 a1+b1
psubsw mm0,mm3 ;mm0= t7 t5 t2 t1
; t3=c1+d1;
; t4=c2+d2;
; t10=-(c1-d1);
; t6=-( c2-d2);
movq mm2,mm1
movq mm3,mm1
psrlq mm2,32 ;mm2= 0 0 d2 d1
psllq mm3,32 ;mm3= c2 c1 0 0
paddsw mm1,mm2 ;mm1= d2 d1 c2+d2 c1+d1
psubsw mm1,mm3 ;mm1= t6 t10 t4 t3
movq mm4,mm1
movq mm3,mm1
pxor mm0,[NEGMID_XOR]
paddsw mm0,[NEGMID_INC] ;mm0= t7 -t5 -t2 t1
;mm1= t6 t10 t4 t3
movq mm2,mm1
movq mm3,mm1
psrlq mm2,16 ;mm2= 0 t6 . .
pslld mm3,16 ;mm3= t10 0 . .
paddsw mm2,mm3
pxor mm1,[NEGODD_XOR]
paddsw mm1,[NEGODD_INC] ;mm1= . . -t4 t3
pand mm1,mm7
pand mm2,mm6
paddsw mm1,mm2
movq mm3,mm1
; x1= t1+t3
; x2=-t2-t4
; x3=-t5+t6
; x4= t7+t10
; y1= t1-t3
; y2=-t2+t4
; y3=-t5-t6
; y4= t7-t10
paddsw mm1,mm0 ;mm1= t10 t6 -t4 t3
movq [edx+2*eax],mm1
psubsw mm0,mm3
movq [edx+2*eax+8],mm0
add eax,8
cmp eax,[LIMIT]
jl near win_one
emms
pop ebp
pop edi
pop esi
pop edx
pop ecx
pop ebx
ret
fft2mmx_a1_nowin
push ebx
push ecx
push edx
push esi
push edi
push ebp
mov ebp,[fft2_bigpermute]
mov eax,[timf2_mask]
sal eax,1
mov [MASK],eax
mov eax,[timf2_px]
sal eax,1
mov [P0],eax
mov eax,[fft2_size]
sal eax,2
mov [LIMIT],eax
;zxy=&fft2_short_int[2*ui.rx_channels*fft2_na*fft2_size];
imul eax,[fft2_na]
mov edx,[fft2_short_int]
add edx,eax
mov ebx,[timf2_shi]
; The two first trivial loops are combined and
; run together with the unscrambling
xor eax,eax
mov ecx,0ffffffffh
mov [edx+4],eax
mov [edx],ecx
movq mm7,[edx]
movq mm6,[edx]
psllq mm6,32
nowin_one:
;for(j=0; j<siz; j+=4)
; {
;ia=(p0+fft2_bigpermute[j ])&mask;
;ib=(p0+fft2_bigpermute[j+1])&mask;
mov esi,[eax+ebp] ;esi=fft2_bigpermute[j]
mov edi,[eax+ebp+4] ;edi=fft2_bigpermute[j+1]
add esi,[P0]
add edi,[P0]
and esi,[MASK]
and edi,[MASK]
; a1=timf2_shi[ia ]+timf2_shi[ia+2];
; a2=timf2_shi[ia+1]+timf2_shi[ia+3];
; b1=timf2_shi[ib ]+timf2_shi[ib+2];
; b2=timf2_shi[ib+1]+timf2_shi[ib+3];
; t1=a1+b1;
; t2=a2+b2;
; t5=a1-b1;
; t7=a2-b2;
movq mm0,[esi+ebx] ;mm0= ia+3 ia+2 ia+1 ia+0
movq mm1,[esi+ebx] ;mm1= ia+3 ia+2 ia+1 ia+0
movq mm2,[edi+ebx] ;mm2= ib+3 ib+2 ib+1 ib+0
movq mm3,[edi+ebx] ;mm3= ib+3 ib+2 ib+1 ib+0
psrlq mm1,32 ;mm1= 0 0 ia+3 ia+2
psrlq mm3,32 ;mm3= 0 0 ib+3 ib+2
paddsw mm0,mm1 ;mm0= ? ? a2 a1
pand mm0,mm7 ;mm0= 0 0 a2 a1
paddsw mm2,mm3 ;mm2= ? ? b2 b1
pand mm2,mm7 ;mm2= 0 0 b2 b1
movq mm1,mm0 ;mm1= 0 0 a2 a1
psllq mm1,32 ;mm1= a2 a1 0 0
paddsw mm0,mm1 ;mm0= a2 a1 a2 a1
paddsw mm0,mm2 ;mm0= a2 a1 a2+b2 a1+b1
psllq mm2,32 ;mm2= b2 b1 0 0
psubsw mm0,mm2 ;mm0= a2-b2 a1-b1 a2+b2 a1+b1
;mm0= t7 t5 t2 t1
mov esi,[eax+ebp+8] ;esi=fft2_bigpermute[j+2]
mov edi,[eax+ebp+12] ;edi=fft2_bigpermute[j+3]
;ic=(p0+fft2_bigpermute[j+2])&mask;
;id=(p0+fft2_bigpermute[j+3])&mask;
add esi,[P0]
add edi,[P0]
and esi,[MASK]
and edi,[MASK]
movq mm1,[esi+ebx] ;mm1= ic+3 ic+2 ic+1 ic+0
movq mm2,[esi+ebx] ;mm2= ic+3 ic+2 ic+1 ic+0
movq mm3,[edi+ebx] ;mm3= id+3 id+2 id+1 id+0
movq mm4,[edi+ebx] ;mm4= id+3 id+2 id+1 id+0
psrlq mm2,32 ;mm2= 0 0 ic+3 ic+2
psrlq mm4,32 ;mm4= 0 0 id+3 id+2
paddsw mm1,mm2 ;mm1= ? ? c2 c1
pand mm1,mm7 ;mm1= 0 0 c2 c1
paddsw mm3,mm4 ;mm3= ? ? d2 d1
pand mm3,mm7 ;mm3= 0 0 d2 d1
movq mm2,mm1 ;mm2= 0 0 c2 c1
psllq mm2,32 ;mm2= c2 c1 0 0
paddsw mm1,mm2 ;mm1= c2 c1 c2 c1
paddsw mm1,mm3 ;mm1= c2 c1 c2+d2 c1+d1
psllq mm3,32 ;mm3= d2 d1 0 0
psubsw mm1,mm3 ;mm1= c2-d2 c1-d1 c2+d2 c1+d1
;mm1= t6 t10 t4 t3
movq mm2,mm1 ;mm1= t6 t10 t4 t3
psrlq mm2,16 ;mm2= 0 t6 ? ?
movq mm3,mm1 ;mm1= t6 t10 t4 t3
psllq mm3,16 ;mm3= t10 ? ? ?
pand mm3,[NEGODD_XOR] ;mm3= t10 0 ? 0
paddsw mm2,mm3 ;mm2= t10 t6 ? ?
pand mm2,mm6 ;mm2= t10 t6 0 0
pxor mm1,[NEGEVEN_XOR]
paddsw mm1,[NEGEVEN_INC]
pand mm1,mm7 ;mm1= 0 0 t4 t3
paddsw mm1,mm2 ;mm1= t10 t6 t4 t3
pxor mm0,[NEGODD_XOR]
paddsw mm0,[NEGODD_INC]
movq mm2,mm0
psubsw mm0,mm1
paddsw mm2,mm1
movq [edx+eax],mm0
movq [edx+eax+8],mm2
add eax,16
cmp eax,[LIMIT]
jl near nowin_one
emms
pop ebp
pop edi
pop esi
pop edx
pop ecx
pop ebx
ret
fft2mmx_a2_nowin
push ebx
push ecx
push edx
push esi
push edi
push ebp
mov ebp,[fft2_bigpermute]
mov eax,[timf2_mask]
sal eax,1
mov [MASK],eax
mov eax,[timf2_px]
sal eax,1
mov [P0],eax
mov eax,[fft2_size]
sal eax,2
mov [LIMIT],eax
;zxy=&fft2_short_int[2*ui.rx_channels*fft2_na*fft2_size];
sal eax,1
imul eax,[fft2_na]
mov edx,[fft2_short_int]
add edx,eax
mov ebx,[timf2_shi]
; The two first trivial loops are combined and
; run together with the unscrambling
xor eax,eax
nowin_two:
;for(j=0; j<siz; j+=4)
; {
;ia=(p0+fft2_bigpermute[j ])&mask;
;ib=(p0+fft2_bigpermute[j+1])&mask;
mov esi,[eax+ebp] ;esi=fft2_bigpermute[j]
mov edi,[eax+ebp+4] ;edi=fft2_bigpermute[j+1]
add esi,[P0]
add edi,[P0]
and esi,[MASK]
and edi,[MASK]
; t1=timf2_shi[ia ]+timf2_shi[ib ]+
; timf2_shi[ia+4]+timf2_shi[ib+4];
; t2=timf2_shi[ia+1]+timf2_shi[ib+1]+
; timf2_shi[ia+5]+timf2_shi[ib+5];
; r1=timf2_shi[ia+2]+timf2_shi[ib+2]+
; timf2_shi[ia+6]+timf2_shi[ib+6];
; r2=timf2_shi[ia+3]+timf2_shi[ib+3]+
; timf2_shi[ia+7]+timf2_shi[ib+7];
; t5=timf2_shi[ia ]-timf2_shi[ib ]+
; timf2_shi[ia+4]-timf2_shi[ib+4];
; t7=timf2_shi[ia+1]-timf2_shi[ib+1]+
; timf2_shi[ia+5]-timf2_shi[ib+5];
; r5=timf2_shi[ia+2]-timf2_shi[ib+2]+
; timf2_shi[ia+6]-timf2_shi[ib+6];
; r7=timf2_shi[ia+3]-timf2_shi[ib+3]+
; timf2_shi[ia+7]-timf2_shi[ib+7];
movq mm0,[esi+ebx]
paddsw mm0,[esi+ebx+8]
movq mm1,[edi+ebx]
paddsw mm1,[edi+ebx+8]
mov esi,[eax+ebp+8] ;esi=fft2_bigpermute[j+2]
mov edi,[eax+ebp+12] ;edi=fft2_bigpermute[j+3]
movq mm2,mm0
paddsw mm0,mm1 ;mm0= r2 r1 t2 t1
psubsw mm2,mm1 ;mm2= r7 r5 t7 t5
movq mm5,mm0
;ic=(p0+fft2_bigpermute[j+2])&mask;
;id=(p0+fft2_bigpermute[j+3])&mask;
add esi,[P0]
add edi,[P0]
and esi,[MASK]
and edi,[MASK]
; t3=timf2_shi[ic ]+timf2_shi[id ]+
; timf2_shi[ic+4]+timf2_shi[id+4];
; t4=timf2_shi[ic+1]+timf2_shi[id+1]+
; timf2_shi[ic+5]+timf2_shi[id+5];
; r3=timf2_shi[ic+2]+timf2_shi[id+2]+
; timf2_shi[ic+6]+timf2_shi[id+6];
; r4=timf2_shi[ic+3]+timf2_shi[id+3]+
; timf2_shi[ic+7]+timf2_shi[id+7];
;
;
; t10=timf2_shi[ic ]-timf2_shi[id ]+
; timf2_shi[ic+4]-timf2_shi[id+4];
; t6= timf2_shi[ic+1]-timf2_shi[id+1]+
; timf2_shi[ic+5]-timf2_shi[id+5];
; r10=timf2_shi[ic+2]-timf2_shi[id+2]+
; timf2_shi[ic+6]-timf2_shi[id+6];
; r6= timf2_shi[ic+3]-timf2_shi[id+3]+
; timf2_shi[ic+7]-timf2_shi[id+7];
movq mm3,[esi+ebx]
paddsw mm3,[esi+ebx+8]
movq mm1,[edi+ebx]
paddsw mm1,[edi+ebx+8]
movq mm4,mm3
paddsw mm3,mm1 ;mm3= r4 r3 t4 t3
psubsw mm4,mm1 ;mm4= r6 r10 t6 t10
;mm0= r2 r1 t2 t1
;mm2= r7 r5 t7 t5
;mm5= r2 r1 t2 t1
; a1=t1+t3
; a2=t2+t4
; a3=r1+r3
; a4=r2+r4
; b1=t1-t3
; b2=t2-t4
; b3=r1-r3
; b4=r2-r4
; a2^=0xffff
; a2++
; a4^=0xffff
; a4++
; b2^=0xffff
; b2++
; b4^=0xffff
; b4++
paddsw mm0,mm3
psubsw mm5,mm3
pxor mm0,[NEGODD_XOR]
pxor mm5,[NEGODD_XOR]
paddsw mm0,[NEGODD_INC]
paddsw mm5,[NEGODD_INC]
; zxy[4*j ]=a1;
; zxy[4*j+1]=a2;
; zxy[4*j+2]=a3;
; zxy[4*j+3]=a4;
; zxy[4*j+8]=b1;
; zxy[4*j+9]=b2;
; zxy[4*j+10]=b3;
; zxy[4*j+11]=b4;
movq mm1,mm4
movq [edx+2*eax],mm0
movq [edx+2*eax+16],mm5
;mm1= r6 r10 t6 t10
;mm4= r6 r10 t6 t10
;mm2= r7 r5 t7 t5
psllq mm1,16 ;mm1= r10 t6 t10 0
psrlq mm4,16 ;mm4= 0 r6 r10 t6
pand mm1,[NEGODD_XOR] ;mm1= r10 0 t10 0
pand mm4,[NEGEVEN_XOR] ;mm4= 0 r6 0 t6
movq mm5,mm2 ;mm5= r7 r5 t7 t5
paddsw mm1,mm4
; c1=t5+t6
; c2=t7+t10
; c3=r5+r6
; c4=r7+r10
; d1=t5-t6
; d2=t7-t10
; d3=r5-r6
; d4=r7-r10
paddsw mm2,mm1
psubsw mm5,mm1
; c2^=0xffff
; c2++
; c4^=0xffff
; c4++
; d2^=0xffff
; d2++
; d4^=0xffff
; d4++
pxor mm2,[NEGODD_XOR]
pxor mm5,[NEGODD_XOR]
paddsw mm2,[NEGODD_INC]
paddsw mm5,[NEGODD_INC]
;mm2= c4 c3 c2 c1
;mm5= d4 d3 d2 d1
movq mm0,mm2 ;mm0= c4 c3 c2 c1
pand mm0,[NEGODD_XOR] ;mm0= c4 0 c2 0
movq mm1,mm5 ;mm1= d4 d3 d2 d1
pand mm1,[NEGODD_XOR] ;mm1= d4 0 d2 0
pand mm2,[NEGEVEN_XOR] ;mm2= 0 c3 0 c1
pand mm5,[NEGEVEN_XOR] ;mm5= 0 d3 0 d1
paddsw mm2,mm1
paddsw mm5,mm0
; zxy[4*j+4]=d1;
; zxy[4*j+5]=c2;
; zxy[4*j+6]=d3;
; zxy[4*j+7]=c4;
; zxy[4*j+12]=c1;
; zxy[4*j+13]=d2;
; zxy[4*j+14]=c3;
; zxy[4*j+15]=d4;
movq [edx+2*eax+8],mm5
movq [edx+2*eax+24],mm2
add eax,16
cmp eax,[LIMIT]
jl near nowin_two
emms
pop ebp
pop edi
pop esi
pop edx
pop ecx
pop ebx
ret
fft2mmx_a2_win
push ebx
push ecx
push edx
push esi
push edi
push ebp
mov ebp,[fft2_bigpermute]
mov eax,[timf2_mask]
sal eax,1
mov [MASK],eax
mov eax,[timf2_px]
sal eax,1
mov [P0],eax
mov eax,[fft2_size]
sal eax,1
mov [LIMIT],eax
;zxy=&fft2_short_int[2*ui.rx_channels*fft2_na*fft2_size];
sal eax,2
imul eax,[fft2_na]
mov edx,[fft2_short_int]
add edx,eax
mov ebx,[timf2_shi]
mov ecx,[fft2_mmxwin]
; The two first trivial loops are combined and
; run together with the unscrambling
xor eax,eax
win_two:
;for(j=0; j<siz; j+=4)
; {
movq mm7,[ecx+eax] ;mm7 = win4 win3 win2 win1
movq mm6,[ecx+eax]
punpcklwd mm6,mm6 ;mm6 = . . win1 win1
punpckldq mm6,mm6 ;mm6 = win1 win1 win1 win1
;ia=(p0+fft2_bigpermute[j ])&mask;
;ib=(p0+fft2_bigpermute[j+1])&mask;
mov esi,[2*eax+ebp] ;esi=fft2_bigpermute[j]
mov edi,[2*eax+ebp+4] ;edi=fft2_bigpermute[j+1]
add esi,[P0]
add edi,[P0]
and esi,[MASK]
and edi,[MASK]
; t1=timf2_shi[ia ]+timf2_shi[ib ]+
; timf2_shi[ia+4]+timf2_shi[ib+4];
; t2=timf2_shi[ia+1]+timf2_shi[ib+1]+
; timf2_shi[ia+5]+timf2_shi[ib+5];
; r1=timf2_shi[ia+2]+timf2_shi[ib+2]+
; timf2_shi[ia+6]+timf2_shi[ib+6];
; r2=timf2_shi[ia+3]+timf2_shi[ib+3]+
; timf2_shi[ia+7]+timf2_shi[ib+7];
; t5=timf2_shi[ia ]-timf2_shi[ib ]+
; timf2_shi[ia+4]-timf2_shi[ib+4];
; t7=timf2_shi[ia+1]-timf2_shi[ib+1]+
; timf2_shi[ia+5]-timf2_shi[ib+5];
; r5=timf2_shi[ia+2]-timf2_shi[ib+2]+
; timf2_shi[ia+6]-timf2_shi[ib+6];
; r7=timf2_shi[ia+3]-timf2_shi[ib+3]+
; timf2_shi[ia+7]-timf2_shi[ib+7];
movq mm0,[esi+ebx]
paddsw mm0,[esi+ebx+8]
; Multiply by first window value in mm6
movq mm1,mm6
pmulhw mm1,mm0
paddsw mm1,mm1 ; shift by addition so we saturate properly!!
paddsw mm1,mm1
pmullw mm0,mm6
psraw mm0,14
paddsw mm0,mm1
psrlq mm7,16
movq mm6,mm7
punpcklwd mm6,mm6
punpckldq mm6,mm6 ;mm6 = win2
movq mm1,[edi+ebx]
paddsw mm1,[edi+ebx+8]
; multiply by second window value
movq mm2,mm6
pmulhw mm2,mm1
paddsw mm2,mm2
paddsw mm2,mm2
pmullw mm1,mm6
psraw mm1,14
paddsw mm1,mm2
mov esi,[2*eax+ebp+8] ;esi=fft2_bigpermute[j+2]
mov edi,[2*eax+ebp+12] ;edi=fft2_bigpermute[j+3]
movq mm2,mm0
paddsw mm0,mm1 ;mm0= r2 r1 t2 t1
psubsw mm2,mm1 ;mm2= r7 r5 t7 t5
movq mm5,mm0 ;mm5= r2 r1 t2 t1
;ic=(p0+fft2_bigpermute[j+2])&mask;
;id=(p0+fft2_bigpermute[j+3])&mask;
add esi,[P0]
add edi,[P0]
and esi,[MASK]
and edi,[MASK]
; t3=timf2_shi[ic ]+timf2_shi[id ]+
; timf2_shi[ic+4]+timf2_shi[id+4];
; t4=timf2_shi[ic+1]+timf2_shi[id+1]+
; timf2_shi[ic+5]+timf2_shi[id+5];
; r3=timf2_shi[ic+2]+timf2_shi[id+2]+
; timf2_shi[ic+6]+timf2_shi[id+6];
; r4=timf2_shi[ic+3]+timf2_shi[id+3]+
; timf2_shi[ic+7]+timf2_shi[id+7];
;
;
; t10=timf2_shi[ic ]-timf2_shi[id ]+
; timf2_shi[ic+4]-timf2_shi[id+4];
; t6= timf2_shi[ic+1]-timf2_shi[id+1]+
; timf2_shi[ic+5]-timf2_shi[id+5];
; r10=timf2_shi[ic+2]-timf2_shi[id+2]+
; timf2_shi[ic+6]-timf2_shi[id+6];
; r6= timf2_shi[ic+3]-timf2_shi[id+3]+
; timf2_shi[ic+7]-timf2_shi[id+7];
psrlq mm7,16
movq mm6,mm7
punpcklwd mm6,mm6
punpckldq mm6,mm6 ;mm6 = win3
movq mm3,[esi+ebx]
paddsw mm3,[esi+ebx+8]
movq mm4,mm6
pmulhw mm4,mm3
paddsw mm4,mm4
paddsw mm4,mm4
pmullw mm3,mm6
psraw mm3,14
paddsw mm3,mm4
psrlq mm7,16
punpcklwd mm7,mm7
punpckldq mm7,mm7 ;mm7 = win4
movq mm1,[edi+ebx]
paddsw mm1,[edi+ebx+8]
movq mm6,mm7
pmulhw mm6,mm1
paddsw mm6,mm6
paddsw mm6,mm6
pmullw mm1,mm7
psraw mm1,14
paddsw mm1,mm6
movq mm4,mm3
paddsw mm3,mm1 ;mm3= r4 r3 t4 t3
psubsw mm4,mm1 ;mm4= r6 r10 t6 t10
;mm0= r2 r1 t2 t1
;mm2= r7 r5 t7 t5
;mm5= r2 r1 t2 t1
; a1=t1+t3
; a2=t2+t4
; a3=r1+r3
; a4=r2+r4
; b1=t1-t3
; b2=t2-t4
; b3=r1-r3
; b4=r2-r4
; a2^=0xffff
; a2++
; a4^=0xffff
; a4++
; b2^=0xffff
; b2++
; b4^=0xffff
; b4++
paddsw mm0,mm3
psubsw mm5,mm3
pxor mm0,[NEGODD_XOR]
pxor mm5,[NEGODD_XOR]
paddsw mm0,[NEGODD_INC]
paddsw mm5,[NEGODD_INC]
; zxy[4*j ]=a1;
; zxy[4*j+1]=a2;
; zxy[4*j+2]=a3;
; zxy[4*j+3]=a4;
; zxy[4*j+8]=b1;
; zxy[4*j+9]=b2;
; zxy[4*j+10]=b3;
; zxy[4*j+11]=b4;
movq mm1,mm4
movq [edx+4*eax],mm0
movq [edx+4*eax+16],mm5
;mm1= r6 r10 t6 t10
;mm4= r6 r10 t6 t10
;mm2= r7 r5 t7 t5
psllq mm1,16 ;mm1= r10 t6 t10 0
psrlq mm4,16 ;mm4= 0 r6 r10 t6
pand mm1,[NEGODD_XOR] ;mm1= r10 0 t10 0
pand mm4,[NEGEVEN_XOR] ;mm4= 0 r6 0 t6
movq mm5,mm2 ;mm5= r7 r5 t7 t5
paddsw mm1,mm4
; c1=t5+t6
; c2=t7+t10
; c3=r5+r6
; c4=r7+r10
; d1=t5-t6
; d2=t7-t10
; d3=r5-r6
; d4=r7-r10
paddsw mm2,mm1
psubsw mm5,mm1
; c2^=0xffff
; c2++
; c4^=0xffff
; c4++
; d2^=0xffff
; d2++
; d4^=0xffff
; d4++
pxor mm2,[NEGODD_XOR]
pxor mm5,[NEGODD_XOR]
paddsw mm2,[NEGODD_INC]
paddsw mm5,[NEGODD_INC]
;mm2= c4 c3 c2 c1
;mm5= d4 d3 d2 d1
movq mm0,mm2 ;mm0= c4 c3 c2 c1
pand mm0,[NEGODD_XOR] ;mm0= c4 0 c2 0
movq mm1,mm5 ;mm1= d4 d3 d2 d1
pand mm1,[NEGODD_XOR] ;mm1= d4 0 d2 0
pand mm2,[NEGEVEN_XOR] ;mm2= 0 c3 0 c1
pand mm5,[NEGEVEN_XOR] ;mm5= 0 d3 0 d1
paddsw mm2,mm1
paddsw mm5,mm0
; zxy[4*j+4]=d1;
; zxy[4*j+5]=c2;
; zxy[4*j+6]=d3;
; zxy[4*j+7]=c4;
; zxy[4*j+12]=c1;
; zxy[4*j+13]=d2;
; zxy[4*j+14]=c3;
; zxy[4*j+15]=d4;
movq [edx+4*eax+8],mm5
movq [edx+4*eax+24],mm2
add eax,8
cmp eax,[LIMIT]
jl near win_two
zz:
emms
pop ebp
pop edi
pop esi
pop edx
pop ecx
pop ebx
ret