summaryrefslogtreecommitdiff
path: root/fs/ocfs2/buffer_head_io.c
blob: cdb9b9bdea1f6d69a1d122943c48f31b951e1de7 (plain)
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
// SPDX-License-Identifier: GPL-2.0-or-later
/*
 * io.c
 *
 * Buffer cache handling
 *
 * Copyright (C) 2002, 2004 Oracle.  All rights reserved.
 */

#include <linux/fs.h>
#include <linux/types.h>
#include <linux/highmem.h>
#include <linux/bio.h>

#include <cluster/masklog.h>

#include "ocfs2.h"

#include "alloc.h"
#include "inode.h"
#include "journal.h"
#include "uptodate.h"
#include "buffer_head_io.h"
#include "ocfs2_trace.h"

/*
 * Bits on bh->b_state used by ocfs2.
 *
 * These MUST be after the JBD2 bits.  Hence, we use BH_JBDPrivateStart.
 */
enum ocfs2_state_bits {
	BH_NeedsValidate = BH_JBDPrivateStart,
};

/* Expand the magic b_state functions */
BUFFER_FNS(NeedsValidate, needs_validate);

int ocfs2_write_block(struct ocfs2_super *osb, struct buffer_head *bh,
		      struct ocfs2_caching_info *ci)
{
	int ret = 0;

	trace_ocfs2_write_block((unsigned long long)bh->b_blocknr, ci);

	BUG_ON(bh->b_blocknr < OCFS2_SUPER_BLOCK_BLKNO);
	BUG_ON(buffer_jbd(bh));

	/* No need to check for a soft readonly file system here. non
	 * journalled writes are only ever done on system files which
	 * can get modified during recovery even if read-only. */
	if (ocfs2_is_hard_readonly(osb)) {
		ret = -EROFS;
		mlog_errno(ret);
		goto out;
	}

	ocfs2_metadata_cache_io_lock(ci);

	lock_buffer(bh);
	set_buffer_uptodate(bh);

	/* remove from dirty list before I/O. */
	clear_buffer_dirty(bh);

	get_bh(bh); /* for end_buffer_write_sync() */
	bh->b_end_io = end_buffer_write_sync;
	submit_bh(REQ_OP_WRITE, bh);

	wait_on_buffer(bh);

	if (buffer_uptodate(bh)) {
		ocfs2_set_buffer_uptodate(ci, bh);
	} else {
		/* We don't need to remove the clustered uptodate
		 * information for this bh as it's not marked locally
		 * uptodate. */
		ret = -EIO;
		mlog_errno(ret);
	}

	ocfs2_metadata_cache_io_unlock(ci);
out:
	return ret;
}

/* Caller must provide a bhs[] with all NULL or non-NULL entries, so it
 * will be easier to handle read failure.
 */
int ocfs2_read_blocks_sync(struct ocfs2_super *osb, u64 block,
			   unsigned int nr, struct buffer_head *bhs[])
{
	int status = 0;
	unsigned int i;
	struct buffer_head *bh;
	int new_bh = 0;

	trace_ocfs2_read_blocks_sync((unsigned long long)block, nr);

	if (!nr)
		goto bail;

	/* Don't put buffer head and re-assign it to NULL if it is allocated
	 * outside since the caller can't be aware of this alternation!
	 */
	new_bh = (bhs[0] == NULL);

	for (i = 0 ; i < nr ; i++) {
		if (bhs[i] == NULL) {
			bhs[i] = sb_getblk(osb->sb, block++);
			if (bhs[i] == NULL) {
				status = -ENOMEM;
				mlog_errno(status);
				break;
			}
		}
		bh = bhs[i];

		if (buffer_jbd(bh)) {
			trace_ocfs2_read_blocks_sync_jbd(
					(unsigned long long)bh->b_blocknr);
			continue;
		}

		if (buffer_dirty(bh)) {
			/* This should probably be a BUG, or
			 * at least return an error. */
			mlog(ML_ERROR,
			     "trying to sync read a dirty "
			     "buffer! (blocknr = %llu), skipping\n",
			     (unsigned long long)bh->b_blocknr);
			continue;
		}

		lock_buffer(bh);
		if (buffer_jbd(bh)) {
#ifdef CATCH_BH_JBD_RACES
			mlog(ML_ERROR,
			     "block %llu had the JBD bit set "
			     "while I was in lock_buffer!",
			     (unsigned long long)bh->b_blocknr);
			BUG();
#else
			unlock_buffer(bh);
			continue;
#endif
		}

		get_bh(bh); /* for end_buffer_read_sync() */
		bh->b_end_io = end_buffer_read_sync;
		submit_bh(REQ_OP_READ, bh);
	}

read_failure:
	for (i = nr; i > 0; i--) {
		bh = bhs[i - 1];

		if (unlikely(status)) {
			if (new_bh && bh) {
				/* If middle bh fails, let previous bh
				 * finish its read and then put it to
				 * avoid bh leak
				 */
				if (!buffer_jbd(bh))
					wait_on_buffer(bh);
				put_bh(bh);
				bhs[i - 1] = NULL;
			} else if (bh && buffer_uptodate(bh)) {
				clear_buffer_uptodate(bh);
			}
			continue;
		}

		/* No need to wait on the buffer if it's managed by JBD. */
		if (!buffer_jbd(bh))
			wait_on_buffer(bh);

		if (!buffer_uptodate(bh)) {
			/* Status won't be cleared from here on out,
			 * so we can safely record this and loop back
			 * to cleanup the other buffers. */
			status = -EIO;
			goto read_failure;
		}
	}

bail:
	return status;
}

/* Caller must provide a bhs[] with all NULL or non-NULL entries, so it
 * will be easier to handle read failure.
 */
int ocfs2_read_blocks(struct ocfs2_caching_info *ci, u64 block, int nr,
		      struct buffer_head *bhs[], int flags,
		      int (*validate)(struct super_block *sb,
				      struct buffer_head *bh))
{
	int status = 0;
	int i, ignore_cache = 0;
	struct buffer_head *bh;
	struct super_block *sb = ocfs2_metadata_cache_get_super(ci);
	int new_bh = 0;

	trace_ocfs2_read_blocks_begin(ci, (unsigned long long)block, nr, flags);

	BUG_ON(!ci);
	BUG_ON((flags & OCFS2_BH_READAHEAD) &&
	       (flags & OCFS2_BH_IGNORE_CACHE));

	if (bhs == NULL) {
		status = -EINVAL;
		mlog_errno(status);
		goto bail;
	}

	if (nr < 0) {
		mlog(ML_ERROR, "asked to read %d blocks!\n", nr);
		status = -EINVAL;
		mlog_errno(status);
		goto bail;
	}

	if (nr == 0) {
		status = 0;
		goto bail;
	}

	/* Don't put buffer head and re-assign it to NULL if it is allocated
	 * outside since the caller can't be aware of this alternation!
	 */
	new_bh = (bhs[0] == NULL);

	ocfs2_metadata_cache_io_lock(ci);
	for (i = 0 ; i < nr ; i++) {
		if (bhs[i] == NULL) {
			bhs[i] = sb_getblk(sb, block++);
			if (bhs[i] == NULL) {
				ocfs2_metadata_cache_io_unlock(ci);
				status = -ENOMEM;
				mlog_errno(status);
				/* Don't forget to put previous bh! */
				break;
			}
		}
		bh = bhs[i];
		ignore_cache = (flags & OCFS2_BH_IGNORE_CACHE);

		/* There are three read-ahead cases here which we need to
		 * be concerned with. All three assume a buffer has
		 * previously been submitted with OCFS2_BH_READAHEAD
		 * and it hasn't yet completed I/O.
		 *
		 * 1) The current request is sync to disk. This rarely
		 *    happens these days, and never when performance
		 *    matters - the code can just wait on the buffer
		 *    lock and re-submit.
		 *
		 * 2) The current request is cached, but not
		 *    readahead. ocfs2_buffer_uptodate() will return
		 *    false anyway, so we'll wind up waiting on the
		 *    buffer lock to do I/O. We re-check the request
		 *    with after getting the lock to avoid a re-submit.
		 *
		 * 3) The current request is readahead (and so must
		 *    also be a caching one). We short circuit if the
		 *    buffer is locked (under I/O) and if it's in the
		 *    uptodate cache. The re-check from #2 catches the
		 *    case that the previous read-ahead completes just
		 *    before our is-it-in-flight check.
		 */

		if (!ignore_cache && !ocfs2_buffer_uptodate(ci, bh)) {
			trace_ocfs2_read_blocks_from_disk(
			     (unsigned long long)bh->b_blocknr,
			     (unsigned long long)ocfs2_metadata_cache_owner(ci));
			/* We're using ignore_cache here to say
			 * "go to disk" */
			ignore_cache = 1;
		}

		trace_ocfs2_read_blocks_bh((unsigned long long)bh->b_blocknr,
			ignore_cache, buffer_jbd(bh), buffer_dirty(bh));

		if (buffer_jbd(bh)) {
			continue;
		}

		if (ignore_cache) {
			if (buffer_dirty(bh)) {
				/* This should probably be a BUG, or
				 * at least return an error. */
				continue;
			}

			/* A read-ahead request was made - if the
			 * buffer is already under read-ahead from a
			 * previously submitted request than we are
			 * done here. */
			if ((flags & OCFS2_BH_READAHEAD)
			    && ocfs2_buffer_read_ahead(ci, bh))
				continue;

			lock_buffer(bh);
			if (buffer_jbd(bh)) {
#ifdef CATCH_BH_JBD_RACES
				mlog(ML_ERROR, "block %llu had the JBD bit set "
					       "while I was in lock_buffer!",
				     (unsigned long long)bh->b_blocknr);
				BUG();
#else
				unlock_buffer(bh);
				continue;
#endif
			}

			/* Re-check ocfs2_buffer_uptodate() as a
			 * previously read-ahead buffer may have
			 * completed I/O while we were waiting for the
			 * buffer lock. */
			if (!(flags & OCFS2_BH_IGNORE_CACHE)
			    && !(flags & OCFS2_BH_READAHEAD)
			    && ocfs2_buffer_uptodate(ci, bh)) {
				unlock_buffer(bh);
				continue;
			}

			get_bh(bh); /* for end_buffer_read_sync() */
			if (validate)
				set_buffer_needs_validate(bh);
			bh->b_end_io = end_buffer_read_sync;
			submit_bh(REQ_OP_READ, bh);
			continue;
		}
	}

read_failure:
	for (i = (nr - 1); i >= 0; i--) {
		bh = bhs[i];

		if (!(flags & OCFS2_BH_READAHEAD)) {
			if (unlikely(status)) {
				/* Clear the buffers on error including those
				 * ever succeeded in reading
				 */
				if (new_bh && bh) {
					/* If middle bh fails, let previous bh
					 * finish its read and then put it to
					 * avoid bh leak
					 */
					if (!buffer_jbd(bh))
						wait_on_buffer(bh);
					put_bh(bh);
					bhs[i] = NULL;
				} else if (bh && buffer_uptodate(bh)) {
					clear_buffer_uptodate(bh);
				}
				continue;
			}
			/* We know this can't have changed as we hold the
			 * owner sem. Avoid doing any work on the bh if the
			 * journal has it. */
			if (!buffer_jbd(bh))
				wait_on_buffer(bh);

			if (!buffer_uptodate(bh)) {
				/* Status won't be cleared from here on out,
				 * so we can safely record this and loop back
				 * to cleanup the other buffers. Don't need to
				 * remove the clustered uptodate information
				 * for this bh as it's not marked locally
				 * uptodate. */
				status = -EIO;
				clear_buffer_needs_validate(bh);
				goto read_failure;
			}

			if (buffer_needs_validate(bh)) {
				/* We never set NeedsValidate if the
				 * buffer was held by the journal, so
				 * that better not have changed */
				BUG_ON(buffer_jbd(bh));
				clear_buffer_needs_validate(bh);
				status = validate(sb, bh);
				if (status)
					goto read_failure;
			}
		}

		/* Always set the buffer in the cache, even if it was
		 * a forced read, or read-ahead which hasn't yet
		 * completed. */
		ocfs2_set_buffer_uptodate(ci, bh);
	}
	ocfs2_metadata_cache_io_unlock(ci);

	trace_ocfs2_read_blocks_end((unsigned long long)block, nr,
				    flags, ignore_cache);

bail:

	return status;
}

/* Check whether the blkno is the super block or one of the backups. */
static void ocfs2_check_super_or_backup(struct super_block *sb,
					sector_t blkno)
{
	int i;
	u64 backup_blkno;

	if (blkno == OCFS2_SUPER_BLOCK_BLKNO)
		return;

	for (i = 0; i < OCFS2_MAX_BACKUP_SUPERBLOCKS; i++) {
		backup_blkno = ocfs2_backup_super_blkno(sb, i);
		if (backup_blkno == blkno)
			return;
	}

	BUG();
}

/*
 * Write super block and backups doesn't need to collaborate with journal,
 * so we don't need to lock ip_io_mutex and ci doesn't need to bea passed
 * into this function.
 */
int ocfs2_write_super_or_backup(struct ocfs2_super *osb,
				struct buffer_head *bh)
{
	int ret = 0;
	struct ocfs2_dinode *di = (struct ocfs2_dinode *)bh->b_data;

	BUG_ON(buffer_jbd(bh));
	ocfs2_check_super_or_backup(osb->sb, bh->b_blocknr);

	if (ocfs2_is_hard_readonly(osb) || ocfs2_is_soft_readonly(osb)) {
		ret = -EROFS;
		mlog_errno(ret);
		goto out;
	}

	lock_buffer(bh);
	set_buffer_uptodate(bh);

	/* remove from dirty list before I/O. */
	clear_buffer_dirty(bh);

	get_bh(bh); /* for end_buffer_write_sync() */
	bh->b_end_io = end_buffer_write_sync;
	ocfs2_compute_meta_ecc(osb->sb, bh->b_data, &di->i_check);
	submit_bh(REQ_OP_WRITE, bh);

	wait_on_buffer(bh);

	if (!buffer_uptodate(bh)) {
		ret = -EIO;
		mlog_errno(ret);
	}

out:
	return ret;
}