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
|
/* Optimized version of the standard memcpy() function.
This file is part of the GNU C Library.
Copyright (C) 2000-2023 Free Software Foundation, Inc.
The GNU C Library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2.1 of the License, or (at your option) any later version.
The GNU C Library is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public
License along with the GNU C Library; if not, see
<https://www.gnu.org/licenses/>. */
/* Return: dest
Inputs:
in0: dest
in1: src
in2: byte count
An assembly implementation of the algorithm used by the generic C
version from glibc. The case when source and sest are aligned is
treated separately, for extra performance.
In this form, memcpy assumes little endian mode. For big endian mode,
sh1 must be computed using an extra instruction: sub sh1 = 64, sh1
and the order of r[MEMLAT] and r[MEMLAT+1] must be reverted in the
shrp instruction. */
#define USE_LFETCH
#define USE_FLP
#include <sysdep.h>
#undef ret
#define LFETCH_DIST 500
#define ALIGN_UNROLL_no 4 // no. of elements
#define ALIGN_UNROLL_sh 2 // (shift amount)
#define MEMLAT 8
#define Nrot ((4*(MEMLAT+2) + 7) & ~7)
#define OP_T_THRES 16
#define OPSIZ 8
#define loopcnt r14
#define elemcnt r15
#define saved_pr r16
#define saved_lc r17
#define adest r18
#define dest r19
#define asrc r20
#define src r21
#define len r22
#define tmp2 r23
#define tmp3 r24
#define tmp4 r25
#define ptable r26
#define ploop56 r27
#define loopaddr r28
#define sh1 r29
#define ptr1 r30
#define ptr2 r31
#define movi0 mov
#define p_scr p6
#define p_xtr p7
#define p_nxtr p8
#define p_few p9
#if defined(USE_FLP)
#define load ldf8
#define store stf8
#define tempreg f6
#define the_r fr
#define the_s fs
#define the_t ft
#define the_q fq
#define the_w fw
#define the_x fx
#define the_y fy
#define the_z fz
#elif defined(USE_INT)
#define load ld8
#define store st8
#define tempreg tmp2
#define the_r r
#define the_s s
#define the_t t
#define the_q q
#define the_w w
#define the_x x
#define the_y y
#define the_z z
#endif
#ifdef GAS_ALIGN_BREAKS_UNWIND_INFO
/* Manually force proper loop-alignment. Note: be sure to
double-check the code-layout after making any changes to
this routine! */
# define ALIGN(n) { nop 0 }
#else
# define ALIGN(n) .align n
#endif
#if defined(USE_LFETCH)
#define LOOP(shift) \
ALIGN(32); \
.loop##shift##: \
{ .mmb \
(p[0]) ld8.nt1 r[0] = [asrc], 8 ; \
(p[0]) lfetch.nt1 [ptr1], 16 ; \
nop.b 0 ; \
} { .mib \
(p[MEMLAT+1]) st8 [dest] = tmp3, 8 ; \
(p[MEMLAT]) shrp tmp3 = r[MEMLAT], s[MEMLAT+1], shift ; \
nop.b 0 ;; \
} { .mmb \
(p[0]) ld8.nt1 s[0] = [asrc], 8 ; \
(p[0]) lfetch.nt1 [ptr2], 16 ; \
nop.b 0 ; \
} { .mib \
(p[MEMLAT+1]) st8 [dest] = tmp4, 8 ; \
(p[MEMLAT]) shrp tmp4 = s[MEMLAT], r[MEMLAT], shift ; \
br.ctop.sptk.many .loop##shift \
;; } \
{ .mib \
br.cond.sptk.many .copy_bytes ; /* deal with the remaining bytes */ \
}
#else
#define LOOP(shift) \
ALIGN(32); \
.loop##shift##: \
{ .mmb \
(p[0]) ld8.nt1 r[0] = [asrc], 8 ; \
nop.b 0 ; \
} { .mib \
(p[MEMLAT+1]) st8 [dest] = tmp3, 8 ; \
(p[MEMLAT]) shrp tmp3 = r[MEMLAT], s[MEMLAT+1], shift ; \
nop.b 0 ;; \
} { .mmb \
(p[0]) ld8.nt1 s[0] = [asrc], 8 ; \
nop.b 0 ; \
} { .mib \
(p[MEMLAT+1]) st8 [dest] = tmp4, 8 ; \
(p[MEMLAT]) shrp tmp4 = s[MEMLAT], r[MEMLAT], shift ; \
br.ctop.sptk.many .loop##shift \
;; } \
{ .mib \
br.cond.sptk.many .copy_bytes ; /* deal with the remaining bytes */ \
}
#endif
ENTRY(memcpy)
{ .mmi
.prologue
alloc r2 = ar.pfs, 3, Nrot - 3, 0, Nrot
.rotr r[MEMLAT+1], s[MEMLAT+2], q[MEMLAT+1], t[MEMLAT+1]
.rotp p[MEMLAT+2]
.rotf fr[MEMLAT+1], fq[MEMLAT+1], fs[MEMLAT+1], ft[MEMLAT+1]
mov ret0 = in0 // return tmp2 = dest
.save pr, saved_pr
movi0 saved_pr = pr // save the predicate registers
} { .mmi
and tmp4 = 7, in0 // check if destination is aligned
mov dest = in0 // dest
mov src = in1 // src
;; }
{ .mii
cmp.eq p_scr, p0 = in2, r0 // if (len == 0)
.save ar.lc, saved_lc
movi0 saved_lc = ar.lc
|
