-
Notifications
You must be signed in to change notification settings - Fork 33
/
AVXMoveOperations.pas
542 lines (384 loc) · 19.7 KB
/
AVXMoveOperations.pas
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
// ###################################################################
// #### This file is part of the mathematics library project, and is
// #### offered under the licence agreement described on
// #### http://www.mrsoft.org/
// ####
// #### Copyright:(c) 2018, Michael R. . All rights reserved.
// ####
// #### Unless required by applicable law or agreed to in writing, software
// #### distributed under the License is distributed on an "AS IS" BASIS,
// #### WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// #### See the License for the specific language governing permissions and
// #### limitations under the License.
// ###################################################################
unit AVXMoveOperations;
interface
{$I 'mrMath_CPU.inc'}
{$IFNDEF x64}
procedure AVXMatrixCopyAligned(Dest : PDouble; const destLineWidth : NativeInt; src : PDouble; const srcLineWidth : NativeInt; Width, Height : NativeInt); {$IFDEF FPC} assembler; {$ELSE} register; {$ENDIF}
procedure AVXMatrixCopyUnAligned(Dest : PDouble; const destLineWidth : NativeInt; src : PDouble; const srcLineWidth : NativeInt; Width, Height : NativeInt); {$IFDEF FPC} assembler; {$ELSE} register; {$ENDIF}
procedure AVXRowSwapAligned(A, B : PDouble; width : NativeInt); {$IFDEF FPC} assembler; {$ELSE} register; {$ENDIF}
procedure AVXRowSwapUnAligned(A, B : PDouble; width : NativeInt); {$IFDEF FPC} assembler; {$ELSE} register; {$ENDIF}
procedure AVXMatrixInitUnAligned( dest : PDouble; const destLineWidth : NativeInt; Width, Height : NativeInt; Value : double); {$IFDEF FPC} assembler; {$ELSE} register; {$ENDIF}
procedure AVXMatrixInitAligned( dest : PDouble; const destLineWidth : NativeInt; Width, Height : NativeInt; Value : double); {$IFDEF FPC} assembler; {$ELSE} register; {$ENDIF}
procedure AVXInitMemAligned(A : PDouble; NumBytes : NativeInt; Value : double); {$IFDEF FPC} assembler; {$ELSE} register; {$ENDIF}
{$ENDIF}
implementation
{$IFNDEF x64}
// eax = dest, edx = destLineWidth, ecx = width
procedure AVXMatrixInitAligned( dest : PDouble; const destLineWidth : NativeInt; Width, Height : NativeInt; Value : double); {$IFDEF FPC} assembler; {$ELSE} register; {$ENDIF}
asm
push ebx;
lea ebx, Value;
{$IFDEF AVXSUP}vbroadcastsd ymm1, [ebx]; {$ELSE}db $C4,$E2,$7D,$19,$0B;{$ENDIF}
//iters := -width*sizeof(double);
imul ecx, -8;
// helper registers for the src and dest pointers
sub eax, ecx;
// for y := 0 to height - 1:
@@addforyloop:
// for x := 0 to w - 1;
// prepare for reverse loop
mov ebx, ecx;
@addforxloop:
add ebx, 128;
jg @loopEnd;
// prefetch data...
// prefetch [ecx + ebx];
// prefetchw [eax + ebx];
// move:
{$IFDEF AVXSUP}vmovapd [eax + ebx - 128], ymm1; {$ELSE}db $C5,$FD,$29,$4C,$18,$80;{$ENDIF}
{$IFDEF AVXSUP}vmovapd [eax + ebx - 96], ymm1; {$ELSE}db $C5,$FD,$29,$4C,$18,$A0;{$ENDIF}
{$IFDEF AVXSUP}vmovapd [eax + ebx - 64], ymm1; {$ELSE}db $C5,$FD,$29,$4C,$18,$C0;{$ENDIF}
{$IFDEF AVXSUP}vmovapd [eax + ebx - 32], ymm1; {$ELSE}db $C5,$FD,$29,$4C,$18,$E0;{$ENDIF}
jmp @addforxloop
@loopEnd:
sub ebx, 128;
jz @nextLine;
@addforxloop2:
add ebx, 16;
jg @loopEnd2;
{$IFDEF AVXSUP}vmovapd [eax + ebx - 16], xmm1; {$ELSE}db $C5,$F9,$29,$4C,$18,$F0;{$ENDIF}
jmp @addforxloop2;
@loopEnd2:
sub ebx, 16;
jz @nextLine;
// last element
{$IFDEF AVXSUP}vmovsd [eax + ebx], xmm1; {$ELSE}db $C5,$FB,$11,$0C,$18;{$ENDIF}
@nextLine:
// next line:
add eax, edx;
// loop y end
dec height;
jnz @@addforyloop;
{$IFDEF AVXSUP}vzeroupper; {$ELSE}db $C5,$F8,$77;{$ENDIF}
// epilog
pop ebx;
end;
procedure AVXMatrixInitUnAligned( dest : PDouble; const destLineWidth : NativeInt; Width, Height : NativeInt; Value : double); {$IFDEF FPC} assembler; {$ELSE} register; {$ENDIF}
asm
push ebx;
lea ebx, Value;
{$IFDEF AVXSUP}vbroadcastsd ymm1, [ebx]; {$ELSE}db $C4,$E2,$7D,$19,$0B;{$ENDIF}
//iters := -width*sizeof(double);
imul ecx, -8;
// helper registers for the src and dest pointers
sub eax, ecx;
// for y := 0 to height - 1:
@@addforyloop:
// for x := 0 to w - 1;
// prepare for reverse loop
mov ebx, ecx;
@addforxloop:
add ebx, 128;
jg @loopEnd;
// prefetch data...
// prefetch [ecx + ebx];
// prefetchw [eax + ebx];
// move:
{$IFDEF AVXSUP}vmovupd [eax + ebx - 128], ymm1; {$ELSE}db $C5,$FD,$11,$4C,$18,$80;{$ENDIF}
{$IFDEF AVXSUP}vmovupd [eax + ebx - 96], ymm1; {$ELSE}db $C5,$FD,$11,$4C,$18,$A0;{$ENDIF}
{$IFDEF AVXSUP}vmovupd [eax + ebx - 64], ymm1; {$ELSE}db $C5,$FD,$11,$4C,$18,$C0;{$ENDIF}
{$IFDEF AVXSUP}vmovupd [eax + ebx - 32], ymm1; {$ELSE}db $C5,$FD,$11,$4C,$18,$E0;{$ENDIF}
jmp @addforxloop
@loopEnd:
sub ebx, 128;
jz @nextLine;
@addforxloop2:
add ebx, 16;
jg @loopEnd2;
{$IFDEF AVXSUP}vmovupd [eax + ebx - 16], xmm1; {$ELSE}db $C5,$F9,$11,$4C,$18,$F0;{$ENDIF}
jmp @addforxloop2;
@loopEnd2:
sub ebx, 16;
jz @nextLine;
// last element
{$IFDEF AVXSUP}vmovsd [eax + ebx], xmm1; {$ELSE}db $C5,$FB,$11,$0C,$18;{$ENDIF}
@nextLine:
// next line:
add eax, edx;
// loop y end
dec height;
jnz @@addforyloop;
{$IFDEF AVXSUP}vzeroupper; {$ELSE}db $C5,$F8,$77;{$ENDIF}
// epilog
pop ebx;
end;
// rcx = A, rdx = NumBytes;
procedure AVXInitMemAligned(A : PDouble; NumBytes : NativeInt; Value : double); {$IFDEF FPC} assembler; {$ELSE} register; {$ENDIF}
asm
lea ecx, Value;
{$IFDEF AVXSUP}vbroadcastsd ymm1, [ecx]; {$ELSE}db $C4,$E2,$7D,$19,$09;{$ENDIF}
imul edx, -1;
sub eax, edx;
@@loopUnrolled:
add edx, 128;
jg @@loopUnrolledEnd;
{$IFDEF AVXSUP}vmovapd [eax + edx - 128], ymm1; {$ELSE}db $C5,$FD,$29,$4C,$10,$80;{$ENDIF}
{$IFDEF AVXSUP}vmovapd [eax + edx - 96], ymm1; {$ELSE}db $C5,$FD,$29,$4C,$10,$A0;{$ENDIF}
{$IFDEF AVXSUP}vmovapd [eax + edx - 64], ymm1; {$ELSE}db $C5,$FD,$29,$4C,$10,$C0;{$ENDIF}
{$IFDEF AVXSUP}vmovapd [eax + edx - 32], ymm1; {$ELSE}db $C5,$FD,$29,$4C,$10,$E0;{$ENDIF}
jmp @@loopUnrolled;
@@loopUnrolledEnd:
sub edx, 128;
jz @@exitProc;
@@loop2:
add edx, 32;
jg @@loopEnd2;
{$IFDEF AVXSUP}vmovapd [eax + edx - 32], xmm1; {$ELSE}db $C5,$F9,$29,$4C,$10,$E0;{$ENDIF}
{$IFDEF AVXSUP}vmovapd [eax + edx - 16], xmm1; {$ELSE}db $C5,$F9,$29,$4C,$10,$F0;{$ENDIF}
jmp @@loop2;
@@loopEnd2:
sub edx, 32;
jz @@exitProc;
@@loop3:
{$IFDEF AVXSUP}vmovsd [eax + edx], xmm1; {$ELSE}db $C5,$FB,$11,$0C,$10;{$ENDIF}
add edx, 8;
jnz @@loop3;
@@exitProc:
{$IFDEF AVXSUP}vzeroupper; {$ELSE}db $C5,$F8,$77;{$ENDIF}
end;
procedure AVXRowSwapAligned(A, B : PDouble; width : NativeInt); {$IFDEF FPC} assembler; {$ELSE} register; {$ENDIF}
asm
// prepare
imul ecx, -8;
sub eax, ecx;
sub edx, ecx;
@unrolloop:
add ecx, 128;
jg @unrolloopend;
// prefetch data...
// prefetchw [ecx + r10];
// prefetchw [edx + r10];
{$IFDEF AVXSUP}vmovdqa ymm0, [eax + ecx - 128]; {$ELSE}db $C5,$FD,$6F,$44,$08,$80;{$ENDIF}
{$IFDEF AVXSUP}vmovdqa ymm1, [edx + ecx - 128]; {$ELSE}db $C5,$FD,$6F,$4C,$0A,$80;{$ENDIF}
{$IFDEF AVXSUP}vmovdqa [eax + ecx - 128], ymm1; {$ELSE}db $C5,$FD,$7F,$4C,$08,$80;{$ENDIF}
{$IFDEF AVXSUP}vmovdqa [edx + ecx - 128], ymm0; {$ELSE}db $C5,$FD,$7F,$44,$0A,$80;{$ENDIF}
{$IFDEF AVXSUP}vmovdqa ymm2, [eax + ecx - 96]; {$ELSE}db $C5,$FD,$6F,$54,$08,$A0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqa ymm3, [edx + ecx - 96]; {$ELSE}db $C5,$FD,$6F,$5C,$0A,$A0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqa [eax + ecx - 96], ymm3; {$ELSE}db $C5,$FD,$7F,$5C,$08,$A0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqa [edx + ecx - 96], ymm2; {$ELSE}db $C5,$FD,$7F,$54,$0A,$A0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqa ymm0, [eax + ecx - 64]; {$ELSE}db $C5,$FD,$6F,$44,$08,$C0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqa ymm1, [edx + ecx - 64]; {$ELSE}db $C5,$FD,$6F,$4C,$0A,$C0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqa [eax + ecx - 64], ymm1; {$ELSE}db $C5,$FD,$7F,$4C,$08,$C0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqa [edx + ecx - 64], ymm0; {$ELSE}db $C5,$FD,$7F,$44,$0A,$C0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqa ymm2, [eax + ecx - 32]; {$ELSE}db $C5,$FD,$6F,$54,$08,$E0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqa ymm3, [edx + ecx - 32]; {$ELSE}db $C5,$FD,$6F,$5C,$0A,$E0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqa [eax + ecx - 32], ymm3; {$ELSE}db $C5,$FD,$7F,$5C,$08,$E0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqa [edx + ecx - 32], ymm2; {$ELSE}db $C5,$FD,$7F,$54,$0A,$E0;{$ENDIF}
jmp @unrolloop;
@unrolloopend:
sub ecx, 128;
jz @endfunc;
@loop2:
add ecx, 16;
jg @loop2End;
{$IFDEF AVXSUP}vmovdqa xmm0, [eax + ecx - 16]; {$ELSE}db $C5,$F9,$6F,$44,$08,$F0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqa xmm1, [edx + ecx - 16]; {$ELSE}db $C5,$F9,$6F,$4C,$0A,$F0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqa [eax + ecx - 16], xmm1; {$ELSE}db $C5,$F9,$7F,$4C,$08,$F0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqa [edx + ecx - 16], xmm0; {$ELSE}db $C5,$F9,$7F,$44,$0A,$F0;{$ENDIF}
jmp @loop2;
@loop2End:
sub ecx, 16;
jz @endfunc;
{$IFDEF AVXSUP}vmovsd xmm0, [eax + ecx]; {$ELSE}db $C5,$FB,$10,$04,$08;{$ENDIF}
{$IFDEF AVXSUP}vmovsd xmm1, [edx + ecx]; {$ELSE}db $C5,$FB,$10,$0C,$0A;{$ENDIF}
{$IFDEF AVXSUP}vmovsd [eax + ecx], xmm1; {$ELSE}db $C5,$FB,$11,$0C,$08;{$ENDIF}
{$IFDEF AVXSUP}vmovsd [edx + ecx], xmm0; {$ELSE}db $C5,$FB,$11,$04,$0A;{$ENDIF}
@endfunc:
{$IFDEF AVXSUP}vzeroupper; {$ELSE}db $C5,$F8,$77;{$ENDIF}
end;
procedure AVXRowSwapUnAligned(A, B : PDouble; width : NativeInt); {$IFDEF FPC} assembler; {$ELSE} register; {$ENDIF}
asm
imul ecx, -8;
sub eax, ecx;
sub edx, ecx;
@unrolloop:
add ecx, 128;
jg @unrolloopend;
// prefetch data...
// prefetchw [ecx + r10];
// prefetchw [edx + r10];
{$IFDEF AVXSUP}vmovdqu ymm0, [eax + ecx - 128]; {$ELSE}db $C5,$FE,$6F,$44,$08,$80;{$ENDIF}
{$IFDEF AVXSUP}vmovdqu ymm1, [edx + ecx - 128]; {$ELSE}db $C5,$FE,$6F,$4C,$0A,$80;{$ENDIF}
{$IFDEF AVXSUP}vmovdqu [eax + ecx - 128], ymm1; {$ELSE}db $C5,$FE,$7F,$4C,$08,$80;{$ENDIF}
{$IFDEF AVXSUP}vmovdqu [edx + ecx - 128], ymm0; {$ELSE}db $C5,$FE,$7F,$44,$0A,$80;{$ENDIF}
{$IFDEF AVXSUP}vmovdqu ymm2, [eax + ecx - 96]; {$ELSE}db $C5,$FE,$6F,$54,$08,$A0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqu ymm3, [edx + ecx - 96]; {$ELSE}db $C5,$FE,$6F,$5C,$0A,$A0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqu [eax + ecx - 96], ymm3; {$ELSE}db $C5,$FE,$7F,$5C,$08,$A0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqu [edx + ecx - 96], ymm2; {$ELSE}db $C5,$FE,$7F,$54,$0A,$A0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqu ymm0, [eax + ecx - 64]; {$ELSE}db $C5,$FE,$6F,$44,$08,$C0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqu ymm1, [edx + ecx - 64]; {$ELSE}db $C5,$FE,$6F,$4C,$0A,$C0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqu [eax + ecx - 64], ymm1; {$ELSE}db $C5,$FE,$7F,$4C,$08,$C0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqu [edx + ecx - 64], ymm0; {$ELSE}db $C5,$FE,$7F,$44,$0A,$C0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqu ymm2, [eax + ecx - 32]; {$ELSE}db $C5,$FE,$6F,$54,$08,$E0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqu ymm3, [edx + ecx - 32]; {$ELSE}db $C5,$FE,$6F,$5C,$0A,$E0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqu [eax + ecx - 32], ymm3; {$ELSE}db $C5,$FE,$7F,$5C,$08,$E0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqu [edx + ecx - 32], ymm2; {$ELSE}db $C5,$FE,$7F,$54,$0A,$E0;{$ENDIF}
jmp @unrolloop;
@unrolloopend:
sub ecx, 128;
jz @endfunc;
@loop2:
add ecx, 16;
jg @loop2End;
{$IFDEF AVXSUP}vmovdqu xmm0, [eax + ecx - 16]; {$ELSE}db $C5,$FA,$6F,$44,$08,$F0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqu xmm1, [edx + ecx - 16]; {$ELSE}db $C5,$FA,$6F,$4C,$0A,$F0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqu [eax + ecx - 16], xmm1; {$ELSE}db $C5,$FA,$7F,$4C,$08,$F0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqu [edx + ecx - 16], xmm0; {$ELSE}db $C5,$FA,$7F,$44,$0A,$F0;{$ENDIF}
jmp @loop2;
@loop2End:
sub ecx, 16;
jz @endfunc;
{$IFDEF AVXSUP}vmovsd xmm0, [eax + ecx]; {$ELSE}db $C5,$FB,$10,$04,$08;{$ENDIF}
{$IFDEF AVXSUP}vmovsd xmm1, [edx + ecx]; {$ELSE}db $C5,$FB,$10,$0C,$0A;{$ENDIF}
{$IFDEF AVXSUP}vmovsd [eax + ecx], xmm1; {$ELSE}db $C5,$FB,$11,$0C,$08;{$ENDIF}
{$IFDEF AVXSUP}vmovsd [edx + ecx], xmm0; {$ELSE}db $C5,$FB,$11,$04,$0A;{$ENDIF}
@endfunc:
{$IFDEF AVXSUP}vzeroupper; {$ELSE}db $C5,$F8,$77;{$ENDIF}
end;
procedure AVXMatrixCopyAligned(Dest : PDouble; const destLineWidth : NativeInt; src : PDouble; const srcLineWidth : NativeInt; Width, Height : NativeInt);
asm
push ebx;
push edi;
push esi;
//iters := -width*sizeof(double);
mov esi, width;
imul esi, -8;
// helper registers for the src and dest pointers
sub ecx, esi;
sub eax, esi;
// for y := 0 to height - 1:
mov edi, Height;
@@addforyloop:
// for x := 0 to w - 1;
// prepare for reverse loop
mov ebx, esi;
@addforxloop:
add ebx, 128;
jg @loopEnd;
// prefetch data...
// prefetch [ecx + ebx];
// prefetchw [eax + ebx];
// move:
{$IFDEF AVXSUP}vmovdqa ymm0, [ecx + ebx - 128]; {$ELSE}db $C5,$FD,$6F,$44,$19,$80;{$ENDIF}
{$IFDEF AVXSUP}vmovdqa [eax + ebx - 128], ymm0; {$ELSE}db $C5,$FD,$7F,$44,$18,$80;{$ENDIF}
{$IFDEF AVXSUP}vmovdqa ymm1, [ecx + ebx - 96]; {$ELSE}db $C5,$FD,$6F,$4C,$19,$A0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqa [eax + ebx - 96], ymm1; {$ELSE}db $C5,$FD,$7F,$4C,$18,$A0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqa ymm2, [ecx + ebx - 64]; {$ELSE}db $C5,$FD,$6F,$54,$19,$C0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqa [eax + ebx - 64], ymm2; {$ELSE}db $C5,$FD,$7F,$54,$18,$C0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqa ymm3, [ecx + ebx - 32]; {$ELSE}db $C5,$FD,$6F,$5C,$19,$E0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqa [eax + ebx - 32], ymm3; {$ELSE}db $C5,$FD,$7F,$5C,$18,$E0;{$ENDIF}
jmp @addforxloop
@loopEnd:
sub ebx, 128;
jz @nextLine;
@addforxloop2:
add ebx, 16;
jg @loopEnd2;
{$IFDEF AVXSUP}vmovdqa xmm0, [ecx + ebx - 16]; {$ELSE}db $C5,$F9,$6F,$44,$19,$F0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqa [eax + ebx - 16], xmm0; {$ELSE}db $C5,$F9,$7F,$44,$18,$F0;{$ENDIF}
jmp @addforxloop2;
@loopEnd2:
sub ebx, 16;
jz @nextLine;
// last element
{$IFDEF AVXSUP}vmovsd xmm0, [ecx + ebx]; {$ELSE}db $C5,$FB,$10,$04,$19;{$ENDIF}
{$IFDEF AVXSUP}vmovsd [eax + ebx], xmm0; {$ELSE}db $C5,$FB,$11,$04,$18;{$ENDIF}
@nextLine:
// next line:
add ecx, srcLineWidth;
add eax, edx;
// loop y end
dec edi;
jnz @@addforyloop;
// epilog
{$IFDEF AVXSUP}vzeroupper; {$ELSE}db $C5,$F8,$77;{$ENDIF}
pop esi;
pop edi;
pop ebx;
end;
procedure AVXMatrixCopyUnAligned(Dest : PDouble; const destLineWidth : NativeInt; src : PDouble; const srcLineWidth : NativeInt; Width, Height : NativeInt);
asm
push ebx;
push edi;
push esi;
// note: ecx = dest, edx = destLineWidth, edx = src, R9 = srcLineWidth
//iters := -width*sizeof(double);
mov esi, width;
imul esi, -8;
// helper registers for the src and dest pointers
sub ecx, esi;
sub eax, esi;
// for y := 0 to height - 1:
mov edi, Height;
@@addforyloop:
// for x := 0 to w - 1;
// prepare for reverse loop
mov ebx, esi;
@addforxloop:
add ebx, 128;
jg @loopEnd;
// prefetch data...
// prefetch [ecx + ebx];
// prefetchw [eax + ebx];
// move:
{$IFDEF AVXSUP}vmovdqu ymm0, [ecx + ebx - 128]; {$ELSE}db $C5,$FE,$6F,$44,$19,$80;{$ENDIF}
{$IFDEF AVXSUP}vmovdqu [eax + ebx - 128], ymm0; {$ELSE}db $C5,$FE,$7F,$44,$18,$80;{$ENDIF}
{$IFDEF AVXSUP}vmovdqu ymm1, [ecx + ebx - 96]; {$ELSE}db $C5,$FE,$6F,$4C,$19,$A0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqu [eax + ebx - 96], ymm1; {$ELSE}db $C5,$FE,$7F,$4C,$18,$A0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqu ymm2, [ecx + ebx - 64]; {$ELSE}db $C5,$FE,$6F,$54,$19,$C0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqu [eax + ebx - 64], ymm2; {$ELSE}db $C5,$FE,$7F,$54,$18,$C0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqu ymm3, [ecx + ebx - 32]; {$ELSE}db $C5,$FE,$6F,$5C,$19,$E0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqu [eax + ebx - 32], ymm3; {$ELSE}db $C5,$FE,$7F,$5C,$18,$E0;{$ENDIF}
jmp @addforxloop
@loopEnd:
sub ebx, 128;
jz @nextLine;
@addforxloop2:
add ebx, 16;
jg @loopEnd2;
{$IFDEF AVXSUP}vmovdqu xmm0, [ecx + ebx - 16]; {$ELSE}db $C5,$FA,$6F,$44,$19,$F0;{$ENDIF}
{$IFDEF AVXSUP}vmovdqu [eax + ebx - 16], xmm0; {$ELSE}db $C5,$FA,$7F,$44,$18,$F0;{$ENDIF}
jmp @addforxloop2;
@loopEnd2:
sub ebx, 16;
jz @nextLine;
// last element
{$IFDEF AVXSUP}vmovsd xmm0, [ecx + ebx]; {$ELSE}db $C5,$FB,$10,$04,$19;{$ENDIF}
{$IFDEF AVXSUP}vmovsd [eax + ebx], xmm0; {$ELSE}db $C5,$FB,$11,$04,$18;{$ENDIF}
@nextLine:
// next line:
add ecx, srcLineWidth;
add eax, edx;
// loop y end
dec edi;
jnz @@addforyloop;
// epilog
{$IFDEF AVXSUP}vzeroupper; {$ELSE}db $C5,$F8,$77;{$ENDIF}
pop esi;
pop edi;
pop ebx;
end;
{$ENDIF}
end.