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
|
dnl X86-64 mpn_redc_1 optimised for Intel Nehalem and Westmere.
dnl Contributed to the GNU project by Torbjörn Granlund.
dnl Copyright 2003-2005, 2007, 2008, 2011-2013 Free Software Foundation, Inc.
dnl This file is part of the GNU MP Library.
dnl
dnl The GNU MP Library is free software; you can redistribute it and/or modify
dnl it under the terms of either:
dnl
dnl * the GNU Lesser General Public License as published by the Free
dnl Software Foundation; either version 3 of the License, or (at your
dnl option) any later version.
dnl
dnl or
dnl
dnl * the GNU General Public License as published by the Free Software
dnl Foundation; either version 2 of the License, or (at your option) any
dnl later version.
dnl
dnl or both in parallel, as here.
dnl
dnl The GNU MP Library is distributed in the hope that it will be useful, but
dnl WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
dnl or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
dnl for more details.
dnl
dnl You should have received copies of the GNU General Public License and the
dnl GNU Lesser General Public License along with the GNU MP Library. If not,
dnl see https://www.gnu.org/licenses/.
include(`../config.m4')
C cycles/limb
C AMD K8,K9 ?
C AMD K10 ?
C AMD bull ?
C AMD pile ?
C AMD steam ?
C AMD bobcat ?
C AMD jaguar ?
C Intel P4 ?
C Intel core ?
C Intel NHM ?
C Intel SBR ?
C Intel IBR ?
C Intel HWL ?
C Intel BWL ?
C Intel atom ?
C VIA nano ?
C The inner loops of this code are the result of running a code generation and
C optimisation tool suite written by David Harvey and Torbjörn Granlund.
C TODO
C * Micro-optimise, none performed thus far.
C * Consider inlining mpn_add_n.
C * Single basecases out before the pushes.
C When playing with pointers, set this to $2 to fall back to conservative
C indexing in wind-down code.
define(`I',`$1')
define(`rp', `%rdi') C rcx
define(`up', `%rsi') C rdx
define(`mp_param', `%rdx') C r8
define(`n', `%rcx') C r9
define(`u0inv', `%r8') C stack
define(`i', `%r14')
define(`j', `%r15')
define(`mp', `%r12')
define(`q0', `%r13')
C rax rbx rcx rdx rdi rsi rbp r8 r9 r10 r11 r12 r13 r14 r15
ABI_SUPPORT(DOS64)
ABI_SUPPORT(STD64)
define(`ALIGNx', `ALIGN(16)')
ASM_START()
TEXT
ALIGN(32)
PROLOGUE(mpn_redc_1)
FUNC_ENTRY(4)
IFDOS(` mov 56(%rsp), %r8 ')
push %rbx
push %rbp
push %r12
push %r13
push %r14
push %r15
mov (up), q0
mov n, j C outer loop induction var
lea (mp_param,n,8), mp
lea (up,n,8), up
neg n
imul u0inv, q0 C first iteration q0
test $1, R8(n)
jz L(bx0)
L(bx1): test $2, R8(n)
jz L(b3)
L(b1): cmp $-1, R32(n)
jz L(n1)
L(otp1):lea 3(n), i
mov (mp,n,8), %rax
mov (up,n,8), %rbp
mul q0
add %rax, %rbp
mov $0, R32(%r9)
mov 8(mp,n,8), %rax
adc %rdx, %r9
mul q0
mov $0, R32(%r11)
mov 8(up,n,8), %rbx
add %rax, %rbx
mov 16(mp,n,8), %rax
adc %rdx, %r11
add %r9, %rbx
adc $0, %r11
mov 16(up,n,8), %rbp
mul q0
add %rax, %rbp
mov $0, R32(%r9)
mov 24(mp,n,8), %rax
adc %rdx, %r9
mov %rbx, 8(up,n,8)
imul u0inv, %rbx C next q limb
jmp L(e1)
ALIGNx
L(tp1): mul q0
add %rax, %rbp
mov $0, R32(%r9)
mov -16(mp,i,8), %rax
adc %rdx, %r9
mul q0
add %r11, %rbp
mov $0, R32(%r11)
mov -16(up,i,8), %r10
adc $0, %r9
add %rax, %r10
mov -8(mp,i,8), %rax
adc %rdx, %r11
mov %rbp, -24(up,i,8)
add %r9, %r10
adc $0, %r11
mov -8(up,i,8), %rbp
mul q0
add %rax, %rbp
mov $0, R32(%r9)
mov (mp,i,8), %rax
adc %rdx, %r9
mov %r10, -16(up,i,8)
L(e1): add %r11, %rbp
adc $0, %r9
mul q0
mov (up,i,8), %r10
mov $0, R32(%r11)
add %rax, %r10
mov 8(mp,i,8), %rax
adc %rdx, %r11
mov %rbp, -8(up,i,8)
add %r9, %r10
adc $0, %r11
mov 8(up,i,8), %rbp
mov %r10, (up,i,8)
add $4, i
jnc L(tp1)
L(ed1): mul q0
add %rax, %rbp
adc $0, %rdx
add %r11, %rbp
adc $0, %rdx
mov %rbp, I(-8(up),-24(up,i,8))
mov %rdx, (up,n,8) C up[0]
mov %rbx, q0 C previously computed q limb -> q0
lea 8(up), up C up++
dec j
jnz L(otp1)
jmp L(cj)
L(b3): cmp $-3, R32(n)
jz L(n3)
L(otp3):lea 5(n), i
mov (mp,n,8), %rax
mov (up,n,8), %rbp
mul q0
add %rax, %rbp
mov $0, R32(%r9)
mov 8(mp,n,8), %rax
adc %rdx, %r9
mul q0
mov 8(up,n,8), %rbx
mov $0, R32(%r11)
add %rax, %rbx
mov 16(mp,n,8), %rax
adc %rdx, %r11
add %r9, %rbx
adc $0, %r11
mov 16(up,n,8), %rbp
mov %rbx, 8(up,n,8)
imul u0inv, %rbx C next q limb
C jmp L(tp3)
ALIGNx
L(tp3): mul q0
add %rax, %rbp
mov $0, R32(%r9)
mov -16(mp,i,8), %rax
adc %rdx, %r9
mul q0
add %r11, %rbp
mov $0, R32(%r11)
mov -16(up,i,8), %r10
adc $0, %r9
add %rax, %r10
mov -8(mp,i,8), %rax
adc %rdx, %r11
mov %rbp, -24(up,i,8)
add %r9, %r10
adc $0, %r11
mov -8(up,i,8), %rbp
mul q0
add %rax, %rbp
mov $0, R32(%r9)
mov (mp,i,8), %rax
adc %rdx, %r9
mov %r10, -16(up,i,8)
add %r11, %rbp
adc $0, %r9
mul q0
mov (up,i,8), %r10
mov $0, R32(%r11)
add %rax, %r10
mov 8(mp,i,8), %rax
adc %rdx, %r11
mov %rbp, -8(up,i,8)
add %r9, %r10
adc $0, %r11
mov 8(up,i,8), %rbp
mov %r10, (up,i,8)
add $4, i
jnc L(tp3)
L(ed3): mul q0
add %rax, %rbp
adc $0, %rdx
add %r11, %rbp
adc $0, %rdx
mov %rbp, I(-8(up),-24(up,i,8))
mov %rdx, (up,n,8) C up[0]
mov %rbx, q0 C previously computed q limb -> q0
lea 8(up), up C up++
dec j
jnz L(otp3)
C jmp L(cj)
L(cj):
IFSTD(` lea (up,n,8), up C param 2: up
lea (up,n,8), %rdx C param 3: up - n
neg R32(n) ') C param 4: n
IFDOS(` lea (up,n,8), %rdx C param 2: up
lea (%rdx,n,8), %r8 C param 3: up - n
neg R32(n)
mov n, %r9 C param 4: n
mov rp, %rcx ') C param 1: rp
CALL( mpn_add_n)
L(ret): pop %r15
pop %r14
pop %r13
pop %r12
pop %rbp
pop %rbx
FUNC_EXIT()
ret
L(bx0): test $2, R8(n)
jnz L(b2)
L(b0):
L(otp0):lea 2(n), i
mov (mp,n,8), %rax
mul q0
mov $0, R32(%r11)
mov (up,n,8), %r10
add %rax, %r10
mov 8(mp,n,8), %rax
adc %rdx, %r11
mov 8(up,n,8), %rbx
mul q0
add %rax, %rbx
mov $0, R32(%r9)
mov 16(mp,n,8), %rax
adc %rdx, %r9
add %r11, %rbx
adc $0, %r9
mul q0
mov 16(up,n,8), %r10
mov $0, R32(%r11)
add %rax, %r10
mov 24(mp,n,8), %rax
adc %rdx, %r11
mov %rbx, 8(up,n,8)
imul u0inv, %rbx C next q limb
jmp L(e0)
ALIGNx
L(tp0): mul q0
add %rax, %rbp
mov $0, R32(%r9)
mov -16(mp,i,8), %rax
adc %rdx, %r9
mul q0
add %r11, %rbp
mov $0, R32(%r11)
mov -16(up,i,8), %r10
adc $0, %r9
add %rax, %r10
mov -8(mp,i,8), %rax
adc %rdx, %r11
mov %rbp, -24(up,i,8)
add %r9, %r10
adc $0, %r11
mov -8(up,i,8), %rbp
mul q0
add %rax, %rbp
mov $0, R32(%r9)
mov (mp,i,8), %rax
adc %rdx, %r9
mov %r10, -16(up,i,8)
add %r11, %rbp
adc $0, %r9
mul q0
mov (up,i,8), %r10
mov $0, R32(%r11)
add %rax, %r10
mov 8(mp,i,8), %rax
adc %rdx, %r11
mov %rbp, -8(up,i,8)
L(e0): add %r9, %r10
adc $0, %r11
mov 8(up,i,8), %rbp
mov %r10, (up,i,8)
add $4, i
jnc L(tp0)
L(ed0): mul q0
add %rax, %rbp
adc $0, %rdx
add %r11, %rbp
adc $0, %rdx
mov %rbp, I(-8(up),-24(up,i,8))
mov %rdx, (up,n,8) C up[0]
mov %rbx, q0 C previously computed q limb -> q0
lea 8(up), up C up++
dec j
jnz L(otp0)
jmp L(cj)
L(b2): cmp $-2, R32(n)
jz L(n2)
L(otp2):lea 4(n), i
mov (mp,n,8), %rax
mul q0
mov (up,n,8), %r10
mov $0, R32(%r11)
add %rax, %r10
mov 8(mp,n,8), %rax
adc %rdx, %r11
mov 8(up,n,8), %rbx
mul q0
add %rax, %rbx
mov $0, R32(%r9)
mov 16(mp,n,8), %rax
adc %rdx, %r9
mul q0
add %r11, %rbx
mov $0, R32(%r11)
mov 16(up,n,8), %r10
adc $0, %r9
add %rax, %r10
mov 24(mp,n,8), %rax
adc %rdx, %r11
mov %rbx, 8(up,n,8)
imul u0inv, %rbx C next q limb
jmp L(e2)
ALIGNx
L(tp2): mul q0
add %rax, %rbp
mov $0, R32(%r9)
mov -16(mp,i,8), %rax
adc %rdx, %r9
mul q0
add %r11, %rbp
mov $0, R32(%r11)
mov -16(up,i,8), %r10
adc $0, %r9
add %rax, %r10
mov -8(mp,i,8), %rax
adc %rdx, %r11
mov %rbp, -24(up,i,8)
L(e2): add %r9, %r10
adc $0, %r11
mov -8(up,i,8), %rbp
mul q0
add %rax, %rbp
mov $0, R32(%r9)
mov (mp,i,8), %rax
adc %rdx, %r9
mov %r10, -16(up,i,8)
add %r11, %rbp
adc $0, %r9
mul q0
mov (up,i,8), %r10
mov $0, R32(%r11)
add %rax, %r10
mov 8(mp,i,8), %rax
adc %rdx, %r11
mov %rbp, -8(up,i,8)
add %r9, %r10
adc $0, %r11
mov 8(up,i,8), %rbp
mov %r10, (up,i,8)
add $4, i
jnc L(tp2)
L(ed2): mul q0
add %rax, %rbp
adc $0, %rdx
add %r11, %rbp
adc $0, %rdx
mov %rbp, I(-8(up),-24(up,i,8))
mov %rdx, (up,n,8) C up[0]
mov %rbx, q0 C previously computed q limb -> q0
lea 8(up), up C up++
dec j
jnz L(otp2)
jmp L(cj)
L(n1): mov (mp_param), %rax
mul q0
add -8(up), %rax
adc (up), %rdx
mov %rdx, (rp)
mov $0, R32(%rax)
adc R32(%rax), R32(%rax)
jmp L(ret)
L(n2): mov (mp_param), %rax
mov -16(up), %rbp
mul q0
add %rax, %rbp
mov %rdx, %r9
adc $0, %r9
mov -8(mp), %rax
mov -8(up), %r10
mul q0
add %rax, %r10
mov %rdx, %r11
adc $0, %r11
add %r9, %r10
adc $0, %r11
mov %r10, q0
imul u0inv, q0 C next q0
mov -16(mp), %rax
mul q0
add %rax, %r10
mov %rdx, %r9
adc $0, %r9
mov -8(mp), %rax
mov (up), %r14
mul q0
add %rax, %r14
adc $0, %rdx
add %r9, %r14
adc $0, %rdx
xor R32(%rax), R32(%rax)
add %r11, %r14
adc 8(up), %rdx
mov %r14, (rp)
mov %rdx, 8(rp)
adc R32(%rax), R32(%rax)
jmp L(ret)
ALIGNx
L(n3): mov -24(mp), %rax
mov -24(up), %r10
mul q0
add %rax, %r10
mov -16(mp), %rax
mov %rdx, %r11
adc $0, %r11
mov -16(up), %rbp
mul q0
add %rax, %rbp
mov %rdx, %r9
adc $0, %r9
mov -8(mp), %rax
add %r11, %rbp
mov -8(up), %r10
adc $0, %r9
mul q0
mov %rbp, q0
imul u0inv, q0 C next q0
add %rax, %r10
mov %rdx, %r11
adc $0, %r11
mov %rbp, -16(up)
add %r9, %r10
adc $0, %r11
mov %r10, -8(up)
mov %r11, -24(up) C up[0]
lea 8(up), up C up++
dec j
jnz L(n3)
mov -48(up), %rdx
mov -40(up), %rbx
xor R32(%rax), R32(%rax)
add %rbp, %rdx
adc %r10, %rbx
adc -8(up), %r11
mov %rdx, (rp)
mov %rbx, 8(rp)
mov %r11, 16(rp)
adc R32(%rax), R32(%rax)
jmp L(ret)
EPILOGUE()
ASM_END()
|
