forked from mikerabat/mrmath
-
Notifications
You must be signed in to change notification settings - Fork 1
/
AVXMatrixSumOperationsx64.pas
449 lines (346 loc) · 14.2 KB
/
AVXMatrixSumOperationsx64.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
// ###################################################################
// #### 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 AVXMatrixSumOperationsx64;
interface
{$IFDEF CPUX64}
{$DEFINE x64}
{$ENDIF}
{$IFDEF cpux86_64}
{$DEFINE x64}
{$ENDIF}
{$IFDEF x64}
uses MatrixConst;
procedure AVXMatrixSumRowAligned(dest : PDouble; const destLineWidth : TASMNativeInt; Src : PDouble; const srcLineWidth : TASMNativeInt; {$ifdef UNIX}unixWidth{$ELSE}width{$endif}, {$ifdef UNIX}unixHeight{$ELSE}height{$endif} : TASMNativeInt); {$IFDEF FPC}assembler;{$ENDIF}
procedure AVXMatrixSumRowUnAligned(dest : PDouble; const destLineWidth : TASMNativeInt; Src : PDouble; const srcLineWidth : TASMNativeInt; {$ifdef UNIX}unixWidth{$ELSE}width{$endif}, {$ifdef UNIX}unixHeight{$ELSE}height{$endif} : TASMNativeInt); {$IFDEF FPC}assembler;{$ENDIF}
procedure AVXMatrixSumColumnAligned(dest : PDouble; const destLineWidth : TASMNativeInt; Src : PDouble; const srcLineWidth : TASMNativeInt; {$ifdef UNIX}unixWidth{$ELSE}width{$endif}, {$ifdef UNIX}unixHeight{$ELSE}height{$endif} : TASMNativeInt); {$IFDEF FPC}assembler;{$ENDIF}
procedure AVXMatrixSumColumnUnAligned(dest : PDouble; const destLineWidth : TASMNativeInt; Src : PDouble; const srcLineWidth : TASMNativeInt; {$ifdef UNIX}unixWidth{$ELSE}width{$endif}, {$ifdef UNIX}unixHeight{$ELSE}height{$endif} : TASMNativeInt); {$IFDEF FPC}assembler;{$ENDIF}
{$ENDIF}
implementation
{$IFDEF x64}
{$IFDEF FPC} {$ASMMODE intel} {$S-} {$ENDIF}
procedure AVXMatrixSumRowAligned(dest : PDouble; const destLineWidth : TASMNativeInt; Src : PDouble; const srcLineWidth : TASMNativeInt; {$ifdef UNIX}unixWidth{$ELSE}width{$endif}, {$ifdef UNIX}unixHeight{$ELSE}height{$endif} : TASMNativeInt); {$IFDEF FPC}assembler;{$ENDIF}
{$ifdef UNIX}
var width : TASMNativeInt;
height : TASMNativeInt;
{$ENDIF}
asm
{$IFDEF UNIX}
// Linux uses a diffrent ABI -> copy over the registers so they meet with winABI
// (note that the 5th and 6th parameter are are on the stack)
// The parameters are passed in the following order:
// RDI, RSI, RDX, RCX -> mov to RCX, RDX, R8, R9
mov width, r8;
mov height, r9;
mov r8, rdx;
mov r9, rcx;
mov rcx, rdi;
mov rdx, rsi;
{$ENDIF}
// note: RCX = dest, RDX = destLineWidth, R8 = src, R9 = srcLineWidth
// prolog
// iters := -width*sizeof(double)
mov r10, width;
imul r10, -8;
// helper registers for the mt1, mt2 and dest pointers
sub r8, r10;
// for y := 0 to height - 1:
mov r11, Height;
@@addforyloop:
{$IFDEF FPC}vxorpd ymm0, ymm0, ymm0;{$ELSE}db $C5,$FD,$57,$C0;{$ENDIF}
{$IFDEF FPC}vxorpd ymm1, ymm1, ymm1;{$ELSE}db $C5,$F5,$57,$C9;{$ENDIF}
// for x := 0 to w - 1;
// prepare for reverse loop
mov rax, r10;
@addforxloop:
add rax, 128;
jg @loopEnd;
// prefetch data...
// prefetch [r8 + rax];
// addition:
{$IFDEF FPC}vaddpd ymm0, ymm0, [r8 + rax - 128];{$ELSE}db $C4,$C1,$7D,$58,$44,$00,$80;{$ENDIF}
{$IFDEF FPC}vaddpd ymm1, ymm1, [r8 + rax - 96];{$ELSE}db $C4,$C1,$75,$58,$4C,$00,$A0;{$ENDIF}
{$IFDEF FPC}vaddpd ymm0, ymm0, [r8 + rax - 64];{$ELSE}db $C4,$C1,$7D,$58,$44,$00,$C0;{$ENDIF}
{$IFDEF FPC}vaddpd ymm1, ymm1, [r8 + rax - 32];{$ELSE}db $C4,$C1,$75,$58,$4C,$00,$E0;{$ENDIF}
jmp @addforxloop
@loopEnd:
{$IFDEF FPC}vaddpd ymm0, ymm0, ymm1;{$ELSE}db $C5,$FD,$58,$C1;{$ENDIF}
{$IFDEF FPC}vextractf128 xmm2, ymm0, 1;{$ELSE}db $C4,$E3,$7D,$19,$C2,$01;{$ENDIF}
{$IFDEF FPC}vhaddpd xmm0, xmm0, xmm2;{$ELSE}db $C5,$F9,$7C,$C2;{$ENDIF}
sub rax, 128;
jz @buildRes;
@addforxloop2:
add rax, 16;
jg @@addforxloop2end;
{$IFDEF FPC}vaddpd xmm0, xmm0, [r8 + rax - 16];{$ELSE}db $C4,$C1,$79,$58,$44,$00,$F0;{$ENDIF}
jmp @addforxloop2;
@@addforxloop2end:
sub rax, 16;
jz @buildRes;
{$IFDEF FPC}vaddsd xmm0, xmm0, [r8 + rax];{$ELSE}db $C4,$C1,$7B,$58,$04,$00;{$ENDIF}
@buildRes:
// build result
{$IFDEF FPC}vhaddpd xmm0, xmm0, xmm0;{$ELSE}db $C5,$F9,$7C,$C0;{$ENDIF}
// write result
{$IFDEF FPC}vmovsd [rcx], xmm0;{$ELSE}db $C5,$FB,$11,$01;{$ENDIF}
// next line:
add r8, r9;
add rcx, rdx;
// loop y end
dec r11;
jnz @@addforyloop;
// epilog
{$IFDEF FPC}vzeroupper;{$ELSE}db $C5,$F8,$77;{$ENDIF}
end;
procedure AVXMatrixSumRowUnAligned(dest : PDouble; const destLineWidth : TASMNativeInt; Src : PDouble; const srcLineWidth : TASMNativeInt; {$ifdef UNIX}unixWidth{$ELSE}width{$endif}, {$ifdef UNIX}unixHeight{$ELSE}height{$endif} : TASMNativeInt); {$IFDEF FPC}assembler;{$ENDIF}
{$ifdef UNIX}
var width : TASMNativeInt;
height : TASMNativeInt;
{$ENDIF}
asm
{$IFDEF UNIX}
// Linux uses a diffrent ABI -> copy over the registers so they meet with winABI
// (note that the 5th and 6th parameter are are on the stack)
// The parameters are passed in the following order:
// RDI, RSI, RDX, RCX -> mov to RCX, RDX, R8, R9
mov width, r8;
mov height, r9;
mov r8, rdx;
mov r9, rcx;
mov rcx, rdi;
mov rdx, rsi;
{$ENDIF}
// note: RCX = dest, RDX = destLineWidth, R8 = src, R9 = srcLineWidth
// prolog
// iters := -width*sizeof(double)
mov r10, width;
imul r10, -8;
// helper registers for the mt1, mt2 and dest pointers
sub r8, r10;
// for y := 0 to height - 1:
mov r11, Height;
@@addforyloop:
{$IFDEF FPC}vxorpd ymm0, ymm0, ymm0;{$ELSE}db $C5,$FD,$57,$C0;{$ENDIF}
{$IFDEF FPC}vxorpd ymm1, ymm1, ymm1;{$ELSE}db $C5,$F5,$57,$C9;{$ENDIF}
// for x := 0 to w - 1;
// prepare for reverse loop
mov rax, r10;
@addforxloop:
add rax, 128;
jg @loopEnd;
// prefetch data...
// prefetch [r8 + rax];
// addition:
{$IFDEF FPC}vmovupd ymm2, [r8 + rax - 128];{$ELSE}db $C4,$C1,$7D,$10,$54,$00,$80;{$ENDIF}
{$IFDEF FPC}vaddpd ymm0, ymm0, ymm2;{$ELSE}db $C5,$FD,$58,$C2;{$ENDIF}
{$IFDEF FPC}vmovupd ymm2, [r8 + rax - 96];{$ELSE}db $C4,$C1,$7D,$10,$54,$00,$A0;{$ENDIF}
{$IFDEF FPC}vaddpd ymm1, ymm1, ymm2;{$ELSE}db $C5,$F5,$58,$CA;{$ENDIF}
{$IFDEF FPC}vmovupd ymm2, [r8 + rax - 64];{$ELSE}db $C4,$C1,$7D,$10,$54,$00,$C0;{$ENDIF}
{$IFDEF FPC}vaddpd ymm0, ymm0, ymm2;{$ELSE}db $C5,$FD,$58,$C2;{$ENDIF}
{$IFDEF FPC}vmovupd ymm2, [r8 + rax - 32];{$ELSE}db $C4,$C1,$7D,$10,$54,$00,$E0;{$ENDIF}
{$IFDEF FPC}vaddpd ymm1, ymm1, ymm2;{$ELSE}db $C5,$F5,$58,$CA;{$ENDIF}
jmp @addforxloop
@loopEnd:
{$IFDEF FPC}vaddpd ymm0, ymm0, ymm1;{$ELSE}db $C5,$FD,$58,$C1;{$ENDIF}
{$IFDEF FPC}vextractf128 xmm2, ymm0, 1;{$ELSE}db $C4,$E3,$7D,$19,$C2,$01;{$ENDIF}
{$IFDEF FPC}vhaddpd xmm0, xmm0, xmm2;{$ELSE}db $C5,$F9,$7C,$C2;{$ENDIF}
sub rax, 128;
jz @buildRes;
@addforxloop2:
add rax, 16;
jg @@addforxloop2end;
{$IFDEF FPC}vmovupd xmm2, [r8 + rax - 16];{$ELSE}db $C4,$C1,$79,$10,$54,$00,$F0;{$ENDIF}
{$IFDEF FPC}vaddpd xmm0, xmm0, xmm2;{$ELSE}db $C5,$F9,$58,$C2;{$ENDIF}
jmp @addforxloop2;
@@addforxloop2end:
sub rax, 16;
jz @buildRes;
{$IFDEF FPC}vaddsd xmm0, xmm0, [r8 + rax];{$ELSE}db $C4,$C1,$7B,$58,$04,$00;{$ENDIF}
@buildRes:
// build result
{$IFDEF FPC}vhaddpd xmm0, xmm0, xmm0;{$ELSE}db $C5,$F9,$7C,$C0;{$ENDIF}
// write result
{$IFDEF FPC}vmovsd [rcx], xmm0;{$ELSE}db $C5,$FB,$11,$01;{$ENDIF}
// next line:
add r8, r9;
add rcx, rdx;
// loop y end
dec r11;
jnz @@addforyloop;
// epilog
{$IFDEF FPC}vzeroupper;{$ELSE}db $C5,$F8,$77;{$ENDIF}
end;
procedure AVXMatrixSumColumnAligned(dest : PDouble; const destLineWidth : TASMNativeInt; Src : PDouble; const srcLineWidth : TASMNativeInt; {$ifdef UNIX}unixWidth{$ELSE}width{$endif}, {$ifdef UNIX}unixHeight{$ELSE}height{$endif} : TASMNativeInt); {$IFDEF FPC}assembler;{$ENDIF}
{$ifdef UNIX}
var width : TASMNativeInt;
height : TASMNativeInt;
{$ENDIF}
asm
{$IFDEF UNIX}
// Linux uses a diffrent ABI -> copy over the registers so they meet with winABI
// (note that the 5th and 6th parameter are are on the stack)
// The parameters are passed in the following order:
// RDI, RSI, RDX, RCX -> mov to RCX, RDX, R8, R9
mov width, r8;
mov height, r9;
mov r8, rdx;
mov r9, rcx;
mov rcx, rdi;
mov rdx, rsi;
{$ENDIF}
// note: RCX = dest, RDX = destLineWidth, R8 = src, R9 = srcLineWidth
xor r10, r10;
sub r10, height;
imul r10, r9;
// helper registers for the mt1, mt2 and dest pointers
sub r8, r10;
// for x := 0 to width - 1:
mov r11, Width;
sub r11, 4;
jl @@addforxloop4end;
// 4 columns at once
@@addforxloop4:
{$IFDEF FPC}vxorpd ymm1, ymm1, ymm1;{$ELSE}db $C5,$F5,$57,$C9;{$ENDIF}
// for y := 0 to height - 1;
// prepare for reverse loop
mov rax, r10;
@addforyloop4:
{$IFDEF FPC}vaddpd ymm1, ymm1, [r8 + rax];{$ELSE}db $C4,$C1,$75,$58,$0C,$00;{$ENDIF}
add rax, r9;
jnz @addforyloop4;
// build result
{$IFDEF FPC}vmovapd [rcx], ymm1;{$ELSE}db $C5,$FD,$29,$09;{$ENDIF}
// next columns:
add rcx, 32;
add r8, 32;
// loop x end
sub r11, 4;
jge @@addforxloop4;
@@addforxloop4end:
add r11, 4;
jz @@endProc;
sub r11, 2;
jl @@lastcolumn;
{$IFDEF FPC}vxorpd xmm1, xmm1, xmm1;{$ELSE}db $C5,$F1,$57,$C9;{$ENDIF}
// for y := 0 to height - 1;
// prepare for reverse loop
mov rax, r10;
@addforyloop2:
{$IFDEF FPC}vaddpd xmm1, xmm1, [r8 + rax];{$ELSE}db $C4,$C1,$71,$58,$0C,$00;{$ENDIF}
add rax, r9;
jnz @addforyloop2;
// build result
{$IFDEF FPC}vmovapd [rcx], xmm1;{$ELSE}db $C5,$F9,$29,$09;{$ENDIF}
// next columns:
add rcx, 16;
add r8, 16;
dec r11;
jnz @@endProc;
@@lastcolumn:
{$IFDEF FPC}vxorpd xmm1, xmm1, xmm1;{$ELSE}db $C5,$F1,$57,$C9;{$ENDIF}
// for y := 0 to height - 1;
// prepare for reverse loop
mov rax, r10;
@addforyloop:
{$IFDEF FPC}vaddsd xmm1, xmm1, [r8 + rax];{$ELSE}db $C4,$C1,$73,$58,$0C,$00;{$ENDIF}
add rax, r9;
jnz @addforyloop;
// build result
{$IFDEF FPC}vmovsd [rcx], xmm1;{$ELSE}db $C5,$FB,$11,$09;{$ENDIF}
@@endProc:
{$IFDEF FPC}vzeroupper;{$ELSE}db $C5,$F8,$77;{$ENDIF}
end;
procedure AVXMatrixSumColumnUnAligned(dest : PDouble; const destLineWidth : TASMNativeInt; Src : PDouble; const srcLineWidth : TASMNativeInt; {$ifdef UNIX}unixWidth{$ELSE}width{$endif}, {$ifdef UNIX}unixHeight{$ELSE}height{$endif} : TASMNativeInt); {$IFDEF FPC}assembler;{$ENDIF}
{$ifdef UNIX}
var width : TASMNativeInt;
height : TASMNativeInt;
{$ENDIF}
asm
{$IFDEF UNIX}
// Linux uses a diffrent ABI -> copy over the registers so they meet with winABI
// (note that the 5th and 6th parameter are are on the stack)
// The parameters are passed in the following order:
// RDI, RSI, RDX, RCX -> mov to RCX, RDX, R8, R9
mov width, r8;
mov height, r9;
mov r8, rdx;
mov r9, rcx;
mov rcx, rdi;
mov rdx, rsi;
{$ENDIF}
// note: RCX = dest, RDX = destLineWidth, R8 = src, R9 = srcLineWidth
xor r10, r10;
sub r10, height;
imul r10, r9;
// helper registers for the mt1, mt2 and dest pointers
sub r8, r10;
// for x := 0 to width - 1:
mov r11, Width;
sub r11, 4;
jl @@addforxloop4end;
// 4 columns at once
@@addforxloop4:
{$IFDEF FPC}vxorpd ymm1, ymm1, ymm1;{$ELSE}db $C5,$F5,$57,$C9;{$ENDIF}
// for y := 0 to height - 1;
// prepare for reverse loop
mov rax, r10;
@addforyloop4:
{$IFDEF FPC}vmovupd ymm0, [r8 + rax];{$ELSE}db $C4,$C1,$7D,$10,$04,$00;{$ENDIF}
{$IFDEF FPC}vaddpd ymm1, ymm1, ymm0;{$ELSE}db $C5,$F5,$58,$C8;{$ENDIF}
add rax, r9;
jnz @addforyloop4;
// build result
{$IFDEF FPC}vmovupd [rcx], ymm1;{$ELSE}db $C5,$FD,$11,$09;{$ENDIF}
// next columns:
add rcx, 32;
add r8, 32;
// loop x end
sub r11, 4;
jge @@addforxloop4;
@@addforxloop4end:
add r11, 4;
jz @@endProc;
sub r11, 2;
jl @@lastcolumn;
{$IFDEF FPC}vxorpd xmm1, xmm1, xmm1;{$ELSE}db $C5,$F1,$57,$C9;{$ENDIF}
// for y := 0 to height - 1;
// prepare for reverse loop
mov rax, r10;
@addforyloop2:
{$IFDEF FPC}vmovupd xmm0, [r8 + rax];{$ELSE}db $C4,$C1,$79,$10,$04,$00;{$ENDIF}
{$IFDEF FPC}vaddpd xmm1, xmm1, [r8 + rax];{$ELSE}db $C4,$C1,$71,$58,$0C,$00;{$ENDIF}
add rax, r9;
jnz @addforyloop2;
// build result
{$IFDEF FPC}vmovupd [rcx], xmm1;{$ELSE}db $C5,$F9,$11,$09;{$ENDIF}
// next columns:
add rcx, 16;
add r8, 16;
dec r11;
jnz @@endProc;
@@lastcolumn:
{$IFDEF FPC}vxorpd xmm1, xmm1, xmm1;{$ELSE}db $C5,$F1,$57,$C9;{$ENDIF}
// for y := 0 to height - 1;
// prepare for reverse loop
mov rax, r10;
@addforyloop:
{$IFDEF FPC}vaddsd xmm1, xmm1, [r8 + rax];{$ELSE}db $C4,$C1,$73,$58,$0C,$00;{$ENDIF}
add rax, r9;
jnz @addforyloop;
// build result
{$IFDEF FPC}vmovsd [rcx], xmm1;{$ELSE}db $C5,$FB,$11,$09;{$ENDIF}
@@endProc:
{$IFDEF FPC}vzeroupper;{$ELSE}db $C5,$F8,$77;{$ENDIF}
end;
{$ENDIF}
end.