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
|
dnl SPARC v9 64-bit mpn_addmul_2 -- Multiply an n limb number with 2-limb
dnl number and add the result to a n limb vector.
dnl Copyright 2002, 2003 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 UltraSPARC 1&2: 9
C UltraSPARC 3: 10
C Algorithm: We use 16 floating-point multiplies per limb product, with the
C 2-limb v operand split into eight 16-bit pieces, and the n-limb u operand
C split into 32-bit pieces. We sum four 48-bit partial products using
C floating-point add, then convert the resulting four 50-bit quantities and
C transfer them to the integer unit.
C Possible optimizations:
C 1. Align the stack area where we transfer the four 50-bit product-sums
C to a 32-byte boundary. That would minimize the cache collision.
C (UltraSPARC-1/2 use a direct-mapped cache.) (Perhaps even better would
C be to align the area to map to the area immediately before up?)
C 2. Perform two of the fp->int conversions with integer instructions. We
C can get almost ten free IEU slots, if we clean up bookkeeping and the
C silly carry-limb code.
C 3. For an mpn_addmul_1 based on this, we need to fix the silly carry-limb
C code.
C OSP (Overlapping software pipeline) version of mpn_mul_basecase:
C Operand swap will require 8 LDDA and 8 FXTOD, which will mean 8 cycles.
C FI = 20
C L = 9 x un * vn
C WDFI = 10 x vn / 2
C WD = 4
C Instruction classification (as per UltraSPARC functional units).
C Assuming silly carry code is fixed. Includes bookkeeping.
C
C mpn_addmul_X mpn_mul_X
C 1 2 1 2
C ========== ==========
C FM 8 16 8 16
C FA 10 18 10 18
C MEM 12 12 10 10
C ISHIFT 6 6 6 6
C IADDLOG 11 11 10 10
C BRANCH 1 1 1 1
C
C TOTAL IEU 17 17 16 16
C TOTAL 48 64 45 61
C
C IEU cycles 8.5 8.5 8 8
C MEM cycles 12 12 10 10
C ISSUE cycles 12 16 11.25 15.25
C FPU cycles 10 18 10 18
C cycles/loop 12 18 12 18
C cycles/limb 12 9 12 9
C INPUT PARAMETERS
C rp[n + 1] i0
C up[n] i1
C n i2
C vp[2] i3
ASM_START()
REGISTER(%g2,#scratch)
REGISTER(%g3,#scratch)
C Combine registers:
C u00_hi= u32_hi
C u00_lo= u32_lo
C a000 = out000
C a016 = out016
C Free: f52 f54
define(`p000', `%f8') define(`p016',`%f10')
define(`p032',`%f12') define(`p048',`%f14')
define(`p064',`%f16') define(`p080',`%f18')
define(`p096a',`%f20') define(`p112a',`%f22')
define(`p096b',`%f56') define(`p112b',`%f58')
define(`out000',`%f0') define(`out016',`%f6')
define(`v000',`%f24') define(`v016',`%f26')
define(`v032',`%f28') define(`v048',`%f30')
define(`v064',`%f44') define(`v080',`%f46')
define(`v096',`%f48') define(`v112',`%f50')
define(`u00',`%f32') define(`u32', `%f34')
define(`a000',`%f36') define(`a016',`%f38')
define(`a032',`%f40') define(`a048',`%f42')
define(`a064',`%f60') define(`a080',`%f62')
define(`u00_hi',`%f2') define(`u32_hi',`%f4')
define(`u00_lo',`%f3') define(`u32_lo',`%f5')
define(`cy',`%g1')
define(`rlimb',`%g3')
define(`i00',`%l0') define(`i16',`%l1')
define(`r00',`%l2') define(`r32',`%l3')
define(`xffffffff',`%l7')
define(`xffff',`%o0')
PROLOGUE(mpn_addmul_2)
C Initialization. (1) Split v operand into eight 16-bit chunks and store them
C as IEEE double in fp registers. (2) Clear upper 32 bits of fp register pairs
C f2 and f4. (3) Store masks in registers aliased to `xffff' and `xffffffff'.
C This code could be better scheduled.
save %sp, -256, %sp
ifdef(`HAVE_VIS',
` mov -1, %g4
wr %g0, 0xD2, %asi
srlx %g4, 32, xffffffff C store mask in register `xffffffff'
ldda [%i3+6] %asi, v000
ldda [%i3+4] %asi, v016
ldda [%i3+2] %asi, v032
ldda [%i3+0] %asi, v048
fxtod v000, v000
ldda [%i3+14] %asi, v064
fxtod v016, v016
ldda [%i3+12] %asi, v080
fxtod v032, v032
ldda [%i3+10] %asi, v096
fxtod v048, v048
ldda [%i3+8] %asi, v112
fxtod v064, v064
fxtod v080, v080
fxtod v096, v096
fxtod v112, v112
fzero u00_hi
fzero u32_hi
',
` mov -1, %g4
ldx [%i3+0], %l0 C vp[0]
srlx %g4, 48, xffff C store mask in register `xffff'
ldx [%i3+8], %l1 C vp[1]
and %l0, xffff, %g2
stx %g2, [%sp+2223+0]
srlx %l0, 16, %g3
and %g3, xffff, %g3
stx %g3, [%sp+2223+8]
srlx %l0, 32, %g2
and %g2, xffff, %g2
stx %g2, [%sp+2223+16]
srlx %l0, 48, %g3
stx %g3, [%sp+2223+24]
and %l1, xffff, %g2
stx %g2, [%sp+2223+32]
srlx %l1, 16, %g3
and %g3, xffff, %g3
stx %g3, [%sp+2223+40]
srlx %l1, 32, %g2
and %g2, xffff, %g2
stx %g2, [%sp+2223+48]
srlx %l1, 48, %g3
stx %g3, [%sp+2223+56]
srlx %g4, 32, xffffffff C store mask in register `xffffffff'
ldd [%sp+2223+0], v000
ldd [%sp+2223+8], v016
ldd [%sp+2223+16], v032
ldd [%sp+2223+24], v048
fxtod v000, v000
ldd [%sp+2223+32], v064
fxtod v016, v016
ldd [%sp+2223+40], v080
fxtod v032, v032
ldd [%sp+2223+48], v096
fxtod v048, v048
ldd [%sp+2223+56], v112
fxtod v064, v064
ld [%sp+2223+0], u00_hi C zero u00_hi
fxtod v080, v080
ld [%sp+2223+0], u32_hi C zero u32_hi
fxtod v096, v096
fxtod v112, v112
')
C Initialization done.
mov 0, %g2
mov 0, rlimb
mov 0, %g4
add %i0, -8, %i0 C BOOKKEEPING
C Start software pipeline.
ld [%i1+4], u00_lo C read low 32 bits of up[i]
fxtod u00_hi, u00
C mid
ld [%i1+0], u32_lo C read high 32 bits of up[i]
fmuld u00, v000, a000
fmuld u00, v016, a016
fmuld u00, v032, a032
fmuld u00, v048, a048
add %i2, -1, %i2 C BOOKKEEPING
fmuld u00, v064, p064
add %i1, 8, %i1 C BOOKKEEPING
fxtod u32_hi, u32
fmuld u00, v080, p080
fmuld u00, v096, p096a
brnz,pt %i2, .L_2_or_more
fmuld u00, v112, p112a
.L1: fdtox a000, out000
fmuld u32, v000, p000
fdtox a016, out016
fmuld u32, v016, p016
fmovd p064, a064
fmuld u32, v032, p032
fmovd p080, a080
fmuld u32, v048, p048
std out000, [%sp+2223+16]
faddd p000, a032, a000
fmuld u32, v064, p064
std out016, [%sp+2223+24]
fxtod u00_hi, u00
faddd p016, a048, a016
fmuld u32, v080, p080
faddd p032, a064, a032
fmuld u32, v096, p096b
faddd p048, a080, a048
fmuld u32, v112, p112b
C mid
fdtox a000, out000
fdtox a016, out016
faddd p064, p096a, a064
faddd p080, p112a, a080
std out000, [%sp+2223+0]
b .L_wd2
std out016, [%sp+2223+8]
.L_2_or_more:
ld [%i1+4], u00_lo C read low 32 bits of up[i]
fdtox a000, out000
fmuld u32, v000, p000
fdtox a016, out016
fmuld u32, v016, p016
fmovd p064, a064
fmuld u32, v032, p032
fmovd p080, a080
fmuld u32, v048, p048
std out000, [%sp+2223+16]
faddd p000, a032, a000
fmuld u32, v064, p064
std out016, [%sp+2223+24]
fxtod u00_hi, u00
faddd p016, a048, a016
fmuld u32, v080, p080
faddd p032, a064, a032
fmuld u32, v096, p096b
faddd p048, a080, a048
fmuld u32, v112, p112b
C mid
ld [%i1+0], u32_lo C read high 32 bits of up[i]
fdtox a000, out000
fmuld u00, v000, p000
fdtox a016, out016
fmuld u00, v016, p016
faddd p064, p096a, a064
fmuld u00, v032, p032
faddd p080, p112a, a080
fmuld u00, v048, p048
add %i2, -1, %i2 C BOOKKEEPING
std out000, [%sp+2223+0]
faddd p000, a032, a000
fmuld u00, v064, p064
add %i1, 8, %i1 C BOOKKEEPING
std out016, [%sp+2223+8]
fxtod u32_hi, u32
faddd p016, a048, a016
fmuld u00, v080, p080
faddd p032, a064, a032
fmuld u00, v096, p096a
faddd p048, a080, a048
brnz,pt %i2, .L_3_or_more
fmuld u00, v112, p112a
b .Lend
nop
C 64 32 0
C . . .
C . |__rXXX_| 32
C . |___cy___| 34
C . |_______i00__| 50
C |_______i16__| . 50
C BEGIN MAIN LOOP
.align 16
.L_3_or_more:
.Loop: ld [%i1+4], u00_lo C read low 32 bits of up[i]
and %g2, xffffffff, %g2
fdtox a000, out000
fmuld u32, v000, p000
C
lduw [%i0+4+8], r00 C read low 32 bits of rp[i]
add %g2, rlimb, %l5
fdtox a016, out016
fmuld u32, v016, p016
C
srlx %l5, 32, cy
ldx [%sp+2223+16], i00
faddd p064, p096b, a064
fmuld u32, v032, p032
C
add %g4, cy, cy C new cy
ldx [%sp+2223+24], i16
faddd p080, p112b, a080
fmuld u32, v048, p048
C
nop
std out000, [%sp+2223+16]
faddd p000, a032, a000
fmuld u32, v064, p064
C
add i00, r00, rlimb
add %i0, 8, %i0 C BOOKKEEPING
std out016, [%sp+2223+24]
fxtod u00_hi, u00
C
sllx i16, 16, %g2
add cy, rlimb, rlimb
faddd p016, a048, a016
fmuld u32, v080, p080
C
srlx i16, 16, %g4
add %g2, rlimb, %l5
faddd p032, a064, a032
fmuld u32, v096, p096b
C
stw %l5, [%i0+4]
nop
faddd p048, a080, a048
fmuld u32, v112, p112b
C midloop
ld [%i1+0], u32_lo C read high 32 bits of up[i]
and %g2, xffffffff, %g2
fdtox a000, out000
fmuld u00, v000, p000
C
lduw [%i0+0], r32 C read high 32 bits of rp[i]
add %g2, rlimb, %l5
fdtox a016, out016
fmuld u00, v016, p016
C
srlx %l5, 32, cy
ldx [%sp+2223+0], i00
faddd p064, p096a, a064
fmuld u00, v032, p032
C
add %g4, cy, cy C new cy
ldx [%sp+2223+8], i16
faddd p080, p112a, a080
fmuld u00, v048, p048
C
add %i2, -1, %i2 C BOOKKEEPING
std out000, [%sp+2223+0]
faddd p000, a032, a000
fmuld u00, v064, p064
C
add i00, r32, rlimb
add %i1, 8, %i1 C BOOKKEEPING
std out016, [%sp+2223+8]
fxtod u32_hi, u32
C
sllx i16, 16, %g2
add cy, rlimb, rlimb
faddd p016, a048, a016
fmuld u00, v080, p080
C
srlx i16, 16, %g4
add %g2, rlimb, %l5
faddd p032, a064, a032
fmuld u00, v096, p096a
C
stw %l5, [%i0+0]
faddd p048, a080, a048
brnz,pt %i2, .Loop
fmuld u00, v112, p112a
C END MAIN LOOP
C WIND-DOWN PHASE 1
.Lend: and %g2, xffffffff, %g2
fdtox a000, out000
fmuld u32, v000, p000
lduw [%i0+4+8], r00 C read low 32 bits of rp[i]
add %g2, rlimb, %l5
fdtox a016, out016
fmuld u32, v016, p016
srlx %l5, 32, cy
ldx [%sp+2223+16], i00
faddd p064, p096b, a064
fmuld u32, v032, p032
add %g4, cy, cy C new cy
ldx [%sp+2223+24], i16
faddd p080, p112b, a080
fmuld u32, v048, p048
std out000, [%sp+2223+16]
faddd p000, a032, a000
fmuld u32, v064, p064
add i00, r00, rlimb
add %i0, 8, %i0 C BOOKKEEPING
std out016, [%sp+2223+24]
sllx i16, 16, %g2
add cy, rlimb, rlimb
faddd p016, a048, a016
fmuld u32, v080, p080
srlx i16, 16, %g4
add %g2, rlimb, %l5
faddd p032, a064, a032
fmuld u32, v096, p096b
stw %l5, [%i0+4]
faddd p048, a080, a048
fmuld u32, v112, p112b
C mid
and %g2, xffffffff, %g2
fdtox a000, out000
lduw [%i0+0], r32 C read high 32 bits of rp[i]
add %g2, rlimb, %l5
fdtox a016, out016
srlx %l5, 32, cy
ldx [%sp+2223+0], i00
faddd p064, p096a, a064
add %g4, cy, cy C new cy
ldx [%sp+2223+8], i16
faddd p080, p112a, a080
std out000, [%sp+2223+0]
add i00, r32, rlimb
std out016, [%sp+2223+8]
sllx i16, 16, %g2
add cy, rlimb, rlimb
srlx i16, 16, %g4
add %g2, rlimb, %l5
stw %l5, [%i0+0]
C WIND-DOWN PHASE 2
.L_wd2: and %g2, xffffffff, %g2
fdtox a032, out000
lduw [%i0+4+8], r00 C read low 32 bits of rp[i]
add %g2, rlimb, %l5
fdtox a048, out016
srlx %l5, 32, cy
ldx [%sp+2223+16], i00
add %g4, cy, cy C new cy
ldx [%sp+2223+24], i16
std out000, [%sp+2223+16]
add i00, r00, rlimb
add %i0, 8, %i0 C BOOKKEEPING
std out016, [%sp+2223+24]
sllx i16, 16, %g2
add cy, rlimb, rlimb
srlx i16, 16, %g4
add %g2, rlimb, %l5
stw %l5, [%i0+4]
C mid
and %g2, xffffffff, %g2
fdtox a064, out000
lduw [%i0+0], r32 C read high 32 bits of rp[i]
add %g2, rlimb, %l5
fdtox a080, out016
srlx %l5, 32, cy
ldx [%sp+2223+0], i00
add %g4, cy, cy C new cy
ldx [%sp+2223+8], i16
std out000, [%sp+2223+0]
add i00, r32, rlimb
std out016, [%sp+2223+8]
sllx i16, 16, %g2
add cy, rlimb, rlimb
srlx i16, 16, %g4
add %g2, rlimb, %l5
stw %l5, [%i0+0]
C WIND-DOWN PHASE 3
.L_wd3: and %g2, xffffffff, %g2
fdtox p096b, out000
add %g2, rlimb, %l5
fdtox p112b, out016
srlx %l5, 32, cy
ldx [%sp+2223+16], rlimb
add %g4, cy, cy C new cy
ldx [%sp+2223+24], i16
std out000, [%sp+2223+16]
add %i0, 8, %i0 C BOOKKEEPING
std out016, [%sp+2223+24]
sllx i16, 16, %g2
add cy, rlimb, rlimb
srlx i16, 16, %g4
add %g2, rlimb, %l5
stw %l5, [%i0+4]
C mid
and %g2, xffffffff, %g2
add %g2, rlimb, %l5
srlx %l5, 32, cy
ldx [%sp+2223+0], rlimb
add %g4, cy, cy C new cy
ldx [%sp+2223+8], i16
sllx i16, 16, %g2
add cy, rlimb, rlimb
srlx i16, 16, %g4
add %g2, rlimb, %l5
stw %l5, [%i0+0]
and %g2, xffffffff, %g2
add %g2, rlimb, %l5
srlx %l5, 32, cy
ldx [%sp+2223+16], i00
add %g4, cy, cy C new cy
ldx [%sp+2223+24], i16
sllx i16, 16, %g2
add i00, cy, cy
return %i7+8
add %g2, cy, %o0
EPILOGUE(mpn_addmul_2)
|