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
|
/* Optimized memchr with sse2
Copyright (C) 2011-2020 Free Software Foundation, Inc.
Contributed by Intel Corporation.
This file is part of the GNU C Library.
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/>. */
#if IS_IN (libc)
# include <sysdep.h>
# define CFI_PUSH(REG) \
cfi_adjust_cfa_offset (4); \
cfi_rel_offset (REG, 0)
# define CFI_POP(REG) \
cfi_adjust_cfa_offset (-4); \
cfi_restore (REG)
# define PUSH(REG) pushl REG; CFI_PUSH (REG)
# define POP(REG) popl REG; CFI_POP (REG)
# define PARMS 4
# define STR1 PARMS
# define STR2 STR1+4
# ifndef USE_AS_RAWMEMCHR
# define LEN STR2+4
# define RETURN POP(%edi); ret; CFI_PUSH(%edi);
# endif
# ifndef MEMCHR
# define MEMCHR __memchr_sse2_bsf
# endif
.text
ENTRY (MEMCHR)
mov STR1(%esp), %ecx
movd STR2(%esp), %xmm1
# ifndef USE_AS_RAWMEMCHR
mov LEN(%esp), %edx
test %edx, %edx
jz L(return_null_1)
# endif
mov %ecx, %eax
punpcklbw %xmm1, %xmm1
punpcklbw %xmm1, %xmm1
and $63, %ecx
pshufd $0, %xmm1, %xmm1
cmp $48, %ecx
ja L(crosscache)
movdqu (%eax), %xmm0
pcmpeqb %xmm1, %xmm0
/* Check if there is a match. */
pmovmskb %xmm0, %ecx
test %ecx, %ecx
je L(unaligned_no_match_1)
/* Check which byte is a match. */
bsf %ecx, %ecx
# ifndef USE_AS_RAWMEMCHR
sub %ecx, %edx
jbe L(return_null_1)
# endif
add %ecx, %eax
ret
.p2align 4
L(unaligned_no_match_1):
# ifndef USE_AS_RAWMEMCHR
sub $16, %edx
jbe L(return_null_1)
PUSH (%edi)
lea 16(%eax), %edi
and $15, %eax
and $-16, %edi
add %eax, %edx
# else
lea 16(%eax), %edx
and $-16, %edx
# endif
jmp L(loop_prolog)
.p2align 4
L(return_null_1):
xor %eax, %eax
ret
# ifndef USE_AS_RAWMEMCHR
CFI_POP (%edi)
# endif
.p2align 4
L(crosscache):
/* Handle unaligned string. */
# ifndef USE_AS_RAWMEMCHR
PUSH (%edi)
mov %eax, %edi
and $15, %ecx
and $-16, %edi
movdqa (%edi), %xmm0
# else
mov %eax, %edx
and $15, %ecx
and $-16, %edx
movdqa (%edx), %xmm0
# endif
pcmpeqb %xmm1, %xmm0
/* Check if there is a match. */
pmovmskb %xmm0, %eax
/* Remove the leading bytes. */
sar %cl, %eax
test %eax, %eax
je L(unaligned_no_match)
/* Check which byte is a match. */
bsf %eax, %eax
# ifndef USE_AS_RAWMEMCHR
sub %eax, %edx
jbe L(return_null)
add %edi, %eax
add %ecx, %eax
RETURN
# else
add %edx, %eax
add %ecx, %eax
ret
# endif
.p2align 4
L(unaligned_no_match):
# ifndef USE_AS_RAWMEMCHR
/* Calculate the last acceptable address and check for possible
addition overflow by using satured math:
edx = ecx + edx
edx |= -(edx < ecx) */
add %ecx, %edx
sbb %eax, %eax
or %eax, %edx
sub $16, %edx
jbe L(return_null)
add $16, %edi
# else
add $16, %edx
# endif
.p2align 4
/* Loop start on aligned string. */
L(loop_prolog):
# ifndef USE_AS_RAWMEMCHR
sub $64, %edx
jbe L(exit_loop)
movdqa (%edi), %xmm0
# else
movdqa (%edx), %xmm0
# endif
pcmpeqb %xmm1, %xmm0
pmovmskb %xmm0, %eax
test %eax, %eax
jnz L(matches)
# ifndef USE_AS_RAWMEMCHR
movdqa 16(%edi), %xmm2
# else
movdqa 16(%edx), %xmm2
# endif
pcmpeqb %xmm1, %xmm2
pmovmskb %xmm2, %eax
test %eax, %eax
jnz L(matches16)
# ifndef USE_AS_RAWMEMCHR
movdqa 32(%edi), %xmm3
# else
movdqa 32(%edx), %xmm3
# endif
pcmpeqb %xmm1, %xmm3
pmovmskb %xmm3, %eax
test %eax, %eax
jnz L(matches32)
# ifndef USE_AS_RAWMEMCHR
movdqa 48(%edi), %xmm4
# else
movdqa 48(%edx), %xmm4
# endif
pcmpeqb %xmm1, %xmm4
# ifndef USE_AS_RAWMEMCHR
add $64, %edi
# else
add $64, %edx
# endif
pmovmskb %xmm4, %eax
test %eax, %eax
jnz L(matches0)
# ifndef USE_AS_RAWMEMCHR
test $0x3f, %edi
# else
test $0x3f, %edx
# endif
jz L(align64_loop)
# ifndef USE_AS_RAWMEMCHR
sub $64, %edx
jbe L(exit_loop)
movdqa (%edi), %xmm0
# else
movdqa (%edx), %xmm0
# endif
pcmpeqb %xmm1, %xmm0
pmovmskb %xmm0, %eax
test %eax, %eax
jnz L(matches)
# ifndef USE_AS_RAWMEMCHR
movdqa 16(%edi), %xmm2
# else
movdqa 16(%edx), %xmm2
# endif
pcmpeqb %xmm1, %xmm2
pmovmskb %xmm2, %eax
test %eax, %eax
jnz L(matches16)
# ifndef USE_AS_RAWMEMCHR
movdqa 32(%edi), %xmm3
# else
movdqa 32(%edx), %xmm3
# endif
pcmpeqb %xmm1, %xmm3
pmovmskb %xmm3, %eax
test %eax, %eax
jnz L(matches32)
# ifndef USE_AS_RAWMEMCHR
movdqa 48(%edi), %xmm3
# else
movdqa 48(%edx), %xmm3
# endif
pcmpeqb %xmm1, %xmm3
pmovmskb %xmm3, %eax
# ifndef USE_AS_RAWMEMCHR
add $64, %edi
# else
add $64, %edx
# endif
test %eax, %eax
jnz L(matches0)
# ifndef USE_AS_RAWMEMCHR
mov %edi, %ecx
and $-64, %edi
and $63, %ecx
add %ecx, %edx
# else
and $-64, %edx
# endif
.p2align 4
L(align64_loop):
# ifndef USE_AS_RAWMEMCHR
sub $64, %edx
jbe L(exit_loop)
movdqa (%edi), %xmm0
movdqa 16(%edi), %xmm2
movdqa 32(%edi), %xmm3
movdqa 48(%edi), %xmm4
# else
movdqa (%edx), %xmm0
movdqa 16(%edx), %xmm2
movdqa 32(%edx), %xmm3
movdqa 48(%edx), %xmm4
# endif
pcmpeqb %xmm1, %xmm0
pcmpeqb %xmm1, %xmm2
pcmpeqb %xmm1, %xmm3
pcmpeqb %xmm1, %xmm4
pmaxub %xmm0, %xmm3
pmaxub %xmm2, %xmm4
pmaxub %xmm3, %xmm4
pmovmskb %xmm4, %eax
# ifndef USE_AS_RAWMEMCHR
add $64, %edi
# else
add $64, %edx
# endif
test %eax, %eax
jz L(align64_loop)
# ifndef USE_AS_RAWMEMCHR
sub $64, %edi
# else
sub $64, %edx
# endif
pmovmskb %xmm0, %eax
test %eax, %eax
jnz L(matches)
pmovmskb %xmm2, %eax
test %eax, %eax
jnz L(matches16)
# ifndef USE_AS_RAWMEMCHR
movdqa 32(%edi), %xmm3
# else
movdqa 32(%edx), %xmm3
# endif
pcmpeqb %xmm1, %xmm3
# ifndef USE_AS_RAWMEMCHR
pcmpeqb 48(%edi), %xmm1
# else
pcmpeqb 48(%edx), %xmm1
# endif
pmovmskb %xmm3, %eax
test %eax, %eax
jnz L(matches32)
pmovmskb %xmm1, %eax
bsf %eax, %eax
# ifndef USE_AS_RAWMEMCHR
lea 48(%edi, %eax), %eax
RETURN
# else
lea 48(%edx, %eax), %eax
ret
# endif
# ifndef USE_AS_RAWMEMCHR
.p2align 4
L(exit_loop):
add $64, %edx
cmp $32, %edx
jbe L(exit_loop_32)
movdqa (%edi), %xmm0
pcmpeqb %xmm1, %xmm0
pmovmskb %xmm0, %eax
test %eax, %eax
jnz L(matches)
movdqa 16(%edi), %xmm2
pcmpeqb %xmm1, %xmm2
pmovmskb %xmm2, %eax
test %eax, %eax
jnz L(matches16)
movdqa 32(%edi), %xmm3
pcmpeqb %xmm1, %xmm3
pmovmskb %xmm3, %eax
test %eax, %eax
jnz L(matches32_1)
cmp $48, %edx
jbe L(return_null)
pcmpeqb 48(%edi), %xmm1
pmovmskb %xmm1, %eax
test %eax, %eax
jnz L(matches48_1)
xor %eax, %eax
RETURN
.p2align 4
L(exit_loop_32):
movdqa (%edi), %xmm0
pcmpeqb %xmm1, %xmm0
pmovmskb %xmm0, %eax
test %eax, %eax
jnz L(matches_1)
cmp $16, %edx
jbe L(return_null)
pcmpeqb 16(%edi), %xmm1
pmovmskb %xmm1, %eax
test %eax, %eax
jnz L(matches16_1)
xor %eax, %eax
RETURN
# endif
.p2align 4
L(matches0):
bsf %eax, %eax
# ifndef USE_AS_RAWMEMCHR
lea -16(%eax, %edi), %eax
RETURN
# else
lea -16(%eax, %edx), %eax
ret
# endif
.p2align 4
L(matches):
bsf %eax, %eax
# ifndef USE_AS_RAWMEMCHR
add %edi, %eax
RETURN
# else
add %edx, %eax
ret
# endif
.p2align 4
L(matches16):
bsf %eax, %eax
# ifndef USE_AS_RAWMEMCHR
lea 16(%eax, %edi), %eax
RETURN
# else
lea 16(%eax, %edx), %eax
ret
# endif
.p2align 4
L(matches32):
bsf %eax, %eax
# ifndef USE_AS_RAWMEMCHR
lea 32(%eax, %edi), %eax
RETURN
# else
lea 32(%eax, %edx), %eax
ret
# endif
# ifndef USE_AS_RAWMEMCHR
.p2align 4
L(matches_1):
bsf %eax, %eax
sub %eax, %edx
jbe L(return_null)
add %edi, %eax
RETURN
.p2align 4
L(matches16_1):
sub $16, %edx
bsf %eax, %eax
sub %eax, %edx
jbe L(return_null)
lea 16(%edi, %eax), %eax
RETURN
.p2align 4
L(matches32_1):
sub $32, %edx
bsf %eax, %eax
sub %eax, %edx
jbe L(return_null)
lea 32(%edi, %eax), %eax
RETURN
.p2align 4
L(matches48_1):
sub $48, %edx
bsf %eax, %eax
sub %eax, %edx
jbe L(return_null)
lea 48(%edi, %eax), %eax
RETURN
# endif
.p2align 4
L(return_null):
xor %eax, %eax
# ifndef USE_AS_RAWMEMCHR
RETURN
# else
ret
# endif
END (MEMCHR)
#endif
|
