-
Notifications
You must be signed in to change notification settings - Fork 509
/
file.c
580 lines (478 loc) · 11.3 KB
/
file.c
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
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
// SPDX-License-Identifier: BSD-3-Clause
/* Copyright 2014-2024, Intel Corporation */
/*
* file.c -- file utilities
*/
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <errno.h>
#include <fcntl.h>
#include <unistd.h>
#include <limits.h>
#include <sys/file.h>
#include <sys/mman.h>
#include <sys/sysmacros.h>
#include "../libpmem2/config.h"
#include "../libpmem2/pmem2_utils.h"
#include "file.h"
#include "os.h"
#include "out.h"
#include "mmap.h"
#define DEVICE_DAX_ZERO_LEN (2 * MEGABYTE)
/*
* util_file_exists -- checks whether file exists
*/
int
util_file_exists(const char *path)
{
LOG(3, "path \"%s\"", path);
if (os_access(path, F_OK) == 0)
return 1;
if (errno != ENOENT) {
ERR_W_ERRNO("os_access \"%s\"", path);
return -1;
}
/*
* ENOENT means that some component of a pathname does not exists.
*
* XXX - we should also call os_access on parent directory and
* if this also results in ENOENT -1 should be returned.
*
* The problem is that we would need to use realpath, which fails
* if file does not exist.
*/
return 0;
}
/*
* util_stat_get_type -- checks whether stat structure describes
* device dax or a normal file
*/
enum file_type
util_stat_get_type(const os_stat_t *st)
{
enum pmem2_file_type type;
int ret = pmem2_get_type_from_stat(st, &type);
if (ret) {
errno = pmem2_err_to_errno(ret);
return OTHER_ERROR;
}
if (type == PMEM2_FTYPE_REG || type == PMEM2_FTYPE_DIR)
return TYPE_NORMAL;
if (type == PMEM2_FTYPE_DEVDAX)
return TYPE_DEVDAX;
ASSERTinfo(0, "unhandled file type in util_stat_get_type");
return OTHER_ERROR;
}
/*
* util_fd_get_type -- checks whether a file descriptor is associated
* with a device dax or a normal file
*/
enum file_type
util_fd_get_type(int fd)
{
LOG(3, "fd %d", fd);
os_stat_t st;
if (os_fstat(fd, &st) < 0) {
ERR_W_ERRNO("fstat");
return OTHER_ERROR;
}
return util_stat_get_type(&st);
}
/*
* util_file_get_type -- checks whether the path points to a device dax,
* normal file or non-existent file
*/
enum file_type
util_file_get_type(const char *path)
{
LOG(3, "path \"%s\"", path);
if (path == NULL) {
ERR_WO_ERRNO("invalid (NULL) path");
errno = EINVAL;
return OTHER_ERROR;
}
int exists = util_file_exists(path);
if (exists < 0)
return OTHER_ERROR;
if (!exists)
return NOT_EXISTS;
os_stat_t st;
if (os_stat(path, &st) < 0) {
ERR_W_ERRNO("stat");
return OTHER_ERROR;
}
return util_stat_get_type(&st);
}
/*
* util_file_get_size -- returns size of a file
*/
ssize_t
util_file_get_size(const char *path)
{
LOG(3, "path \"%s\"", path);
int fd = os_open(path, O_RDONLY);
if (fd < 0) {
ERR_W_ERRNO("open");
return -1;
}
ssize_t size = util_fd_get_size(fd);
(void) close(fd);
return size;
}
/*
* util_fd_get_size -- returns size of a file behind a given file descriptor
*/
ssize_t
util_fd_get_size(int fd)
{
LOG(3, "fd %d", fd);
struct pmem2_source *src;
size_t size;
int ret;
if ((ret = pmem2_source_from_fd(&src, fd)) != 0) {
errno = pmem2_err_to_errno(ret);
return -1;
}
ret = pmem2_source_size(src, &size);
pmem2_source_delete(&src);
if (ret) {
errno = pmem2_err_to_errno(ret);
return -1;
}
/* size is unsigned, this function returns signed */
if (size >= INT64_MAX) {
errno = ERANGE;
ERR_WO_ERRNO(
"file size (%ld) too big to be represented in 64-bit signed integer",
size);
return -1;
}
LOG(4, "file length %zu", size);
return (ssize_t)size;
}
/*
* util_file_map_whole -- maps the entire file into memory
*/
void *
util_file_map_whole(const char *path)
{
LOG(3, "path \"%s\"", path);
int fd;
int olderrno;
void *addr = NULL;
int flags = O_RDWR;
if ((fd = os_open(path, flags)) < 0) {
ERR_W_ERRNO("open \"%s\"", path);
return NULL;
}
ssize_t size = util_fd_get_size(fd);
if (size < 0) {
CORE_LOG_ERROR("cannot determine file length \"%s\"", path);
goto out;
}
addr = util_map(fd, 0, (size_t)size, MAP_SHARED, 0, 0, NULL);
if (addr == NULL) {
CORE_LOG_ERROR("failed to map entire file \"%s\"", path);
goto out;
}
out:
olderrno = errno;
(void) os_close(fd);
errno = olderrno;
return addr;
}
/*
* util_file_zero -- zeroes the specified region of the file
*/
int
util_file_zero(const char *path, os_off_t off, size_t len)
{
LOG(3, "path \"%s\" off %ju len %zu", path, off, len);
int fd;
int olderrno;
int ret = 0;
int flags = O_RDWR;
if ((fd = os_open(path, flags)) < 0) {
ERR_W_ERRNO("open \"%s\"", path);
return -1;
}
ssize_t size = util_fd_get_size(fd);
if (size < 0) {
CORE_LOG_ERROR("cannot determine file length \"%s\"", path);
ret = -1;
goto out;
}
if (off > size) {
CORE_LOG_ERROR("offset beyond file length, %ju > %ju", off,
size);
ret = -1;
goto out;
}
if ((size_t)off + len > (size_t)size) {
CORE_LOG_WARNING(
"requested size of write goes beyond the file length, %zu > %zu",
(size_t)off + len, size);
LOG(4, "adjusting len to %zu", size - off);
len = (size_t)(size - off);
}
void *addr = util_map(fd, 0, (size_t)size, MAP_SHARED, 0, 0, NULL);
if (addr == NULL) {
CORE_LOG_ERROR("failed to map entire file \"%s\"", path);
ret = -1;
goto out;
}
/* zero initialize the specified region */
memset((char *)addr + off, 0, len);
util_unmap(addr, (size_t)size);
out:
olderrno = errno;
(void) os_close(fd);
errno = olderrno;
return ret;
}
/*
* util_file_pwrite -- writes to a file with an offset
*/
ssize_t
util_file_pwrite(const char *path, const void *buffer, size_t size,
os_off_t offset)
{
LOG(3, "path \"%s\" buffer %p size %zu offset %ju",
path, buffer, size, offset);
enum file_type type = util_file_get_type(path);
if (type < 0)
return -1;
if (type == TYPE_NORMAL) {
int fd = util_file_open(path, NULL, 0, O_RDWR);
if (fd < 0) {
CORE_LOG_ERROR("failed to open file \"%s\"", path);
return -1;
}
ssize_t write_len = pwrite(fd, buffer, size, offset);
int olderrno = errno;
(void) os_close(fd);
errno = olderrno;
return write_len;
}
ssize_t file_size = util_file_get_size(path);
if (file_size < 0) {
CORE_LOG_ERROR("cannot determine file length \"%s\"", path);
return -1;
}
size_t max_size = (size_t)(file_size - offset);
if (size > max_size) {
CORE_LOG_WARNING(
"requested size of write goes beyond the file length, %zu > %zu",
size, max_size);
LOG(4, "adjusting size to %zu", max_size);
size = max_size;
}
void *addr = util_file_map_whole(path);
if (addr == NULL) {
CORE_LOG_ERROR("failed to map entire file \"%s\"", path);
return -1;
}
memcpy(ADDR_SUM(addr, offset), buffer, size);
util_unmap(addr, (size_t)file_size);
return (ssize_t)size;
}
/*
* util_file_pread -- reads from a file with an offset
*/
ssize_t
util_file_pread(const char *path, void *buffer, size_t size,
os_off_t offset)
{
LOG(3, "path \"%s\" buffer %p size %zu offset %ju",
path, buffer, size, offset);
enum file_type type = util_file_get_type(path);
if (type < 0)
return -1;
if (type == TYPE_NORMAL) {
int fd = util_file_open(path, NULL, 0, O_RDONLY);
if (fd < 0) {
CORE_LOG_ERROR("failed to open file \"%s\"", path);
return -1;
}
ssize_t read_len = pread(fd, buffer, size, offset);
int olderrno = errno;
(void) os_close(fd);
errno = olderrno;
return read_len;
}
ssize_t file_size = util_file_get_size(path);
if (file_size < 0) {
CORE_LOG_ERROR("cannot determine file length \"%s\"", path);
return -1;
}
size_t max_size = (size_t)(file_size - offset);
if (size > max_size) {
CORE_LOG_WARNING(
"requested size of read goes beyond the file length, %zu > %zu",
size, max_size);
LOG(4, "adjusting size to %zu", max_size);
size = max_size;
}
void *addr = util_file_map_whole(path);
if (addr == NULL) {
CORE_LOG_ERROR("failed to map entire file \"%s\"", path);
return -1;
}
memcpy(buffer, ADDR_SUM(addr, offset), size);
util_unmap(addr, (size_t)file_size);
return (ssize_t)size;
}
/*
* util_file_create -- create a new memory pool file
*/
int
util_file_create(const char *path, size_t size, size_t minsize)
{
LOG(3, "path \"%s\" size %zu minsize %zu", path, size, minsize);
ASSERTne(size, 0);
if (size < minsize) {
ERR_WO_ERRNO("size %zu smaller than %zu", size, minsize);
errno = EINVAL;
return -1;
}
if (((os_off_t)size) < 0) {
ERR_WO_ERRNO("invalid size (%zu) for os_off_t", size);
errno = EFBIG;
return -1;
}
int fd;
int mode;
int flags = O_RDWR | O_CREAT | O_EXCL;
mode = 0;
/*
* Create file without any permission. It will be granted once
* initialization completes.
*/
if ((fd = os_open(path, flags, mode)) < 0) {
ERR_W_ERRNO("open \"%s\"", path);
return -1;
}
if ((errno = os_posix_fallocate(fd, 0, (os_off_t)size)) != 0) {
ERR_W_ERRNO("posix_fallocate \"%s\", %zu", path, size);
goto err;
}
if (os_flock(fd, OS_LOCK_EX | OS_LOCK_NB) < 0) {
ERR_W_ERRNO("flock \"%s\"", path);
goto err;
}
return fd;
err:
LOG(4, "error clean up");
int oerrno = errno;
if (fd != -1)
(void) os_close(fd);
os_unlink(path);
errno = oerrno;
return -1;
}
/*
* util_file_open -- open a memory pool file
*/
int
util_file_open(const char *path, size_t *size, size_t minsize, int flags)
{
LOG(3, "path \"%s\" size %p minsize %zu flags %d", path, size, minsize,
flags);
int oerrno;
int fd;
if ((fd = os_open(path, flags)) < 0) {
ERR_W_ERRNO("open \"%s\"", path);
return -1;
}
if (os_flock(fd, OS_LOCK_EX | OS_LOCK_NB) < 0) {
ERR_W_ERRNO("flock \"%s\"", path);
(void) os_close(fd);
return -1;
}
if (size || minsize) {
if (size)
ASSERTeq(*size, 0);
ssize_t actual_size = util_fd_get_size(fd);
if (actual_size < 0) {
ERR_WO_ERRNO("stat \"%s\": negative size", path);
errno = EINVAL;
goto err;
}
if ((size_t)actual_size < minsize) {
ERR_WO_ERRNO("size %zu smaller than %zu",
(size_t)actual_size, minsize);
errno = EINVAL;
goto err;
}
if (size) {
*size = (size_t)actual_size;
LOG(4, "actual file size %zu", *size);
}
}
return fd;
err:
oerrno = errno;
if (os_flock(fd, OS_LOCK_UN))
ERR_W_ERRNO("flock unlock");
(void) os_close(fd);
errno = oerrno;
return -1;
}
/*
* util_unlink -- unlinks a file or zeroes a device dax
*/
int
util_unlink(const char *path)
{
LOG(3, "path \"%s\"", path);
enum file_type type = util_file_get_type(path);
if (type < 0)
return -1;
if (type == TYPE_DEVDAX) {
return util_file_zero(path, 0, DEVICE_DAX_ZERO_LEN);
} else {
return os_unlink(path);
}
}
/*
* util_unlink_flock -- flocks the file and unlinks it
*
* The unlink(2) call on a file which is opened and locked using flock(2)
* by different process works on linux. Thus in order to forbid removing a
* pool when in use by different process we need to flock(2) the pool files
* first before unlinking.
*/
int
util_unlink_flock(const char *path)
{
LOG(3, "path \"%s\"", path);
int fd = util_file_open(path, NULL, 0, O_RDONLY);
if (fd < 0) {
CORE_LOG_ERROR("failed to open file \"%s\"", path);
return -1;
}
int ret = util_unlink(path);
(void) os_close(fd);
return ret;
}
/*
* util_write_all -- a wrapper for util_write
*
* writes exactly count bytes from buf to file referred to by fd
* returns -1 on error, 0 otherwise
*/
int
util_write_all(int fd, const char *buf, size_t count)
{
ssize_t n_wrote = 0;
size_t total = 0;
while (count > total) {
n_wrote = util_write(fd, buf, count - total);
if (n_wrote <= 0)
return -1;
buf += (size_t)n_wrote;
total += (size_t)n_wrote;
}
return 0;
}