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
|
dnl IA-64 mpn_divrem_1 and mpn_preinv_divrem_1 -- Divide an mpn number by an
dnl unnormalized limb.
dnl Contributed to the GNU project by Torbjorn Granlund.
dnl Copyright 2002, 2004, 2005 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 Itanium: 40-42
C Itanium 2: 29-30
C This was generated by gcc, then the loops were optimized. The preinv entry
C point was shoehorned into the file. Lots of things outside the loops could
C be streamlined. It would probably be a good idea to merge the loops for
C normalized and unnormalized divisor, since the shifting stuff is done for
C free in parallel with other operations. It would even be possible to merge
C all loops, if the ld8 were made conditional.
C TODO
C * Consider delaying inversion for normalized mpn_divrem_1 entry till after
C computing leading limb.
C * Inline and interleave limb inversion code with loop setup code.
ASM_START()
C HP's assembler requires these declarations for importing mpn_invert_limb
.global mpn_invert_limb
.type mpn_invert_limb,@function
C INPUT PARAMETERS
C rp = r32
C qxn = r33
C up = r34
C n = r35
C vl = r36
C vlinv = r37 (preinv only)
C cnt = r38 (preinv only)
PROLOGUE(mpn_preinv_divrem_1)
.prologue
.save ar.pfs, r42
alloc r42 = ar.pfs, 7, 8, 1, 0
.save ar.lc, r44
mov r44 = ar.lc
.save rp, r41
mov r41 = b0
.body
ifdef(`HAVE_ABI_32',
` addp4 r32 = 0, r32
sxt4 r33 = r33
addp4 r34 = 0, r34
sxt4 r35 = r35
;;
')
mov r40 = r38
shladd r34 = r35, 3, r34
;;
adds r34 = -8, r34
;;
ld8 r39 = [r34], -8
;;
add r15 = r35, r33
;;
mov r8 = r37
shladd r32 = r15, 3, r32 C r32 = rp + n + qxn
cmp.le p8, p0 = 0, r36
;;
adds r32 = -8, r32 C r32 = rp + n + qxn - 1
cmp.leu p6, p7 = r36, r39
(p8) br.cond.dpnt .Lpunnorm
;;
(p6) addl r15 = 1, r0
(p7) mov r15 = r0
;;
(p6) sub r38 = r39, r36
(p7) mov r38 = r39
st8 [r32] = r15, -8
adds r35 = -2, r35 C un -= 2
br .Lpn
.Lpunnorm:
(p6) add r34 = 8, r34
mov r38 = 0 C r = 0
shl r36 = r36, r40
(p6) br.cond.dptk .Lpu
;;
shl r38 = r39, r40 C r = ahigh << cnt
cmp.ne p8, p0 = 1, r35
st8 [r32] = r0, -8
adds r35 = -1, r35 C un--
(p8) br.cond.dpnt .Lpu
mov r23 = 1
;;
setf.sig f6 = r8
setf.sig f12 = r23
br .L435
EPILOGUE()
PROLOGUE(mpn_divrem_1)
.prologue
.save ar.pfs, r42
alloc r42 = ar.pfs, 5, 8, 1, 0
.save ar.lc, r44
mov r44 = ar.lc
.save rp, r41
mov r41 = b0
.body
ifdef(`HAVE_ABI_32',
` addp4 r32 = 0, r32
sxt4 r33 = r33
addp4 r34 = 0, r34
sxt4 r35 = r35
;;
')
mov r38 = r0
add r15 = r35, r33
;;
cmp.ne p6, p7 = 0, r15
;;
(p7) mov r8 = r0
(p7) br.cond.dpnt .Lret
shladd r14 = r15, 3, r32 C r14 = rp + n + qxn
cmp.le p6, p7 = 0, r36
;;
adds r32 = -8, r14 C r32 = rp + n + qxn - 1
(p6) br.cond.dpnt .Lunnorm
cmp.eq p6, p7 = 0, r35
(p6) br.cond.dpnt .L179
shladd r14 = r35, 3, r34
;;
adds r14 = -8, r14
adds r35 = -1, r35
;;
ld8 r38 = [r14]
;;
cmp.leu p6, p7 = r36, r38
;;
(p6) addl r15 = 1, r0
(p7) mov r15 = r0
;;
st8 [r32] = r15, -8
(p6) sub r38 = r38, r36
.L179:
mov r45 = r36
adds r35 = -1, r35
br.call.sptk.many b0 = mpn_invert_limb
;;
shladd r34 = r35, 3, r34
.Lpn:
mov r23 = 1
;;
setf.sig f6 = r8
setf.sig f12 = r23
cmp.le p6, p7 = 0, r35
mov r40 = 0
(p7) br.cond.dpnt .L435
setf.sig f10 = r36
mov ar.lc = r35
setf.sig f7 = r38
;;
sub r28 = -1, r36
C Develop quotient limbs for normalized divisor
.Loop1: C 00 C q=r18 nh=r38/f7
ld8 r20 = [r34], -8
xma.hu f11 = f7, f6, f0
;; C 04
xma.l f8 = f11, f12, f7 C q = q + nh
;; C 08
getf.sig r18 = f8
xma.hu f9 = f8, f10, f0
xma.l f8 = f8, f10, f0
;; C 12
getf.sig r16 = f9
C 13
getf.sig r15 = f8
;; C 18
cmp.ltu p6, p7 = r20, r15
sub r15 = r20, r15
sub r16 = r38, r16
;; C 19
(p6) cmp.ne p8, p9 = 1, r16 C is rH != 0?
(p7) cmp.ne p8, p9 = 0, r16 C is rH != 0?
(p6) add r16 = -1, r16
(p0) cmp.ne.unc p6, p7 = r0, r0
;; C 20
(p8) cmp.ltu p6, p7 = r15, r36
(p8) sub r15 = r15, r36
(p8) add r18 = 1, r18 C q = q + 1; done if: rH > 0
;; C 21
.pred.rel "mutex",p6,p7
(p6) cmp.ne p8, p9 = 1, r16 C is rH != 0 still?
(p7) cmp.ne p8, p9 = 0, r16 C is rH != 0 still?
cmp.ltu p6, p7 = r15, r36 C speculative
sub r28 = r15, r36 C speculative, just for cmp
;; C 22
(p8) cmp.ltu p6, p7 = r28, r36 C redo last cmp if needed
(p8) mov r15 = r28
(p8) add r18 = 1, r18 C q = q + 1; done if: rH > 0
;; C 23
(p6) setf.sig f7 = r15
(p7) sub r15 = r15, r36
(p7) add r18 = 1, r18 C q = q + 1; done if: rH > 0
;; C 24
(p7) setf.sig f7 = r15
st8 [r32] = r18, -8
mov r38 = r15
br.cloop.dptk .Loop1
C 29/30
br.sptk .L435
;;
.Lunnorm:
mux1 r16 = r36, @rev
cmp.eq p6, p7 = 0, r35
(p6) br.cond.dpnt .L322
shladd r34 = r35, 3, r34
;;
adds r34 = -8, r34
;;
ld8 r39 = [r34]
;;
cmp.leu p6, p7 = r36, r39
(p6) br.cond.dptk .L322
adds r34 = -8, r34
;;
mov r38 = r39
;;
cmp.ne p6, p7 = 1, r15
st8 [r32] = r0, -8
;;
(p7) mov r8 = r38
(p7) br.cond.dpnt .Lret
adds r35 = -1, r35
.L322:
sub r14 = r0, r16
;;
or r14 = r16, r14
;;
mov r16 = -8
czx1.l r14 = r14
;;
shladd r16 = r14, 3, r16
;;
shr.u r14 = r36, r16
;;
cmp.geu p6, p7 = 15, r14
;;
(p7) shr.u r14 = r14, 4
(p7) adds r16 = 4, r16
;;
cmp.geu p6, p7 = 3, r14
;;
(p7) shr.u r14 = r14, 2
(p7) adds r16 = 2, r16
;;
tbit.nz p6, p7 = r14, 1
;;
.pred.rel "mutex",p6,p7
(p6) sub r40 = 62, r16
(p7) sub r40 = 63, r16
;;
shl r45 = r36, r40
shl r36 = r36, r40
shl r38 = r38, r40
br.call.sptk.many b0 = mpn_invert_limb
;;
.Lpu:
mov r23 = 1
;;
setf.sig f6 = r8
setf.sig f12 = r23
cmp.eq p6, p7 = 0, r35
(p6) br.cond.dpnt .L435
sub r16 = 64, r40
adds r35 = -2, r35
;;
ld8 r39 = [r34], -8
cmp.le p6, p7 = 0, r35
;;
shr.u r14 = r39, r16
;;
or r38 = r14, r38
(p7) br.cond.dpnt .Lend3
;;
mov r22 = r16
setf.sig f10 = r36
setf.sig f7 = r38
mov ar.lc = r35
;;
C Develop quotient limbs for unnormalized divisor
.Loop3:
ld8 r14 = [r34], -8
xma.hu f11 = f7, f6, f0
;;
xma.l f8 = f11, f12, f7 C q = q + nh
;;
getf.sig r18 = f8
xma.hu f9 = f8, f10, f0
shl r20 = r39, r40
xma.l f8 = f8, f10, f0
shr.u r24 = r14, r22
;;
getf.sig r16 = f9
getf.sig r15 = f8
or r20 = r24, r20
;;
cmp.ltu p6, p7 = r20, r15
sub r15 = r20, r15
sub r16 = r38, r16
;;
(p6) cmp.ne p8, p9 = 1, r16 C is rH != 0?
(p7) cmp.ne p8, p9 = 0, r16 C is rH != 0?
(p6) add r16 = -1, r16
(p0) cmp.ne.unc p6, p7 = r0, r0
;;
(p8) cmp.ltu p6, p7 = r15, r36
(p8) sub r15 = r15, r36
(p8) add r18 = 1, r18 C q = q + 1; done if: rH > 0
;;
.pred.rel "mutex",p6,p7
(p6) cmp.ne p8, p9 = 1, r16 C is rH != 0 still?
(p7) cmp.ne p8, p9 = 0, r16 C is rH != 0 still?
cmp.ltu p6, p7 = r15, r36 C speculative
sub r28 = r15, r36 C speculative, just for cmp
;;
(p8) cmp.ltu p6, p7 = r28, r36 C redo last cmp if needed
(p8) mov r15 = r28
(p8) add r18 = 1, r18 C q = q + 1; done if: rH > 0
;;
(p6) setf.sig f7 = r15
(p7) sub r15 = r15, r36
(p7) add r18 = 1, r18 C q = q + 1; done if: rH > 0
;;
(p7) setf.sig f7 = r15
st8 [r32] = r18, -8
mov r39 = r14
mov r38 = r15
br.cloop.dptk .Loop3
;;
.Lend3:
setf.sig f10 = r36
setf.sig f7 = r38
;;
xma.hu f11 = f7, f6, f0
;;
xma.l f8 = f11, f12, f7 C q = q + nh
;;
getf.sig r18 = f8
xma.hu f9 = f8, f10, f0
shl r20 = r39, r40
xma.l f8 = f8, f10, f0
;;
getf.sig r16 = f9
getf.sig r15 = f8
;;
cmp.ltu p6, p7 = r20, r15
sub r15 = r20, r15
sub r16 = r38, r16
;;
(p6) cmp.ne p8, p9 = 1, r16 C is rH != 0?
(p7) cmp.ne p8, p9 = 0, r16 C is rH != 0?
(p6) add r16 = -1, r16
(p0) cmp.ne.unc p6, p7 = r0, r0
;;
(p8) cmp.ltu p6, p7 = r15, r36
(p8) sub r15 = r15, r36
(p8) add r18 = 1, r18 C q = q + 1; done if: rH > 0
;;
.pred.rel "mutex",p6,p7
(p6) cmp.ne p8, p9 = 1, r16 C is rH != 0 still?
(p7) cmp.ne p8, p9 = 0, r16 C is rH != 0 still?
;;
(p8) sub r15 = r15, r36
(p8) add r18 = 1, r18 C q = q + 1; done if: rH > 0
;;
cmp.ltu p6, p7 = r15, r36
;;
(p7) sub r15 = r15, r36
(p7) add r18 = 1, r18 C q = q + 1; done if: rH > 0
;;
st8 [r32] = r18, -8
mov r38 = r15
.L435:
adds r35 = -1, r33
cmp.le p6, p7 = 1, r33
(p7) br.cond.dpnt .Lend4
;;
setf.sig f7 = r38
setf.sig f10 = r36
mov ar.lc = r35
;;
.Loop4:
xma.hu f11 = f7, f6, f0
;;
xma.l f8 = f11, f12, f7 C q = q + nh
;;
getf.sig r18 = f8
xma.hu f9 = f8, f10, f0
xma.l f8 = f8, f10, f0
;;
getf.sig r16 = f9
getf.sig r15 = f8
;;
cmp.ltu p6, p7 = 0, r15
sub r15 = 0, r15
sub r16 = r38, r16
;;
(p6) cmp.ne p8, p9 = 1, r16 C is rH != 0?
(p7) cmp.ne p8, p9 = 0, r16 C is rH != 0?
(p6) add r16 = -1, r16
(p0) cmp.ne.unc p6, p7 = r0, r0
;;
(p8) cmp.ltu p6, p7 = r15, r36
(p8) sub r15 = r15, r36
(p8) add r18 = 1, r18 C q = q + 1; done if: rH > 0
;;
.pred.rel "mutex",p6,p7
(p6) cmp.ne p8, p9 = 1, r16 C is rH != 0 still?
(p7) cmp.ne p8, p9 = 0, r16 C is rH != 0 still?
cmp.ltu p6, p7 = r15, r36 C speculative
sub r28 = r15, r36 C speculative, just for cmp
;;
(p8) cmp.ltu p6, p7 = r28, r36 C redo last cmp if needed
(p8) mov r15 = r28
(p8) add r18 = 1, r18 C q = q + 1; done if: rH > 0
;;
(p6) setf.sig f7 = r15
(p7) sub r15 = r15, r36
(p7) add r18 = 1, r18 C q = q + 1; done if: rH > 0
;;
(p7) setf.sig f7 = r15
st8 [r32] = r18, -8
mov r38 = r15
br.cloop.dptk .Loop4
;;
.Lend4:
shr.u r8 = r38, r40
.Lret:
mov ar.pfs = r42
mov ar.lc = r44
mov b0 = r41
br.ret.sptk.many b0
EPILOGUE()
ASM_END()
|