summaryrefslogtreecommitdiffstats
path: root/include/linux/blkdev_compat.h
blob: 89e888c30025b763557eaa6b68661152b69f8f09 (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
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
/*
 * CDDL HEADER START
 *
 * The contents of this file are subject to the terms of the
 * Common Development and Distribution License (the "License").
 * You may not use this file except in compliance with the License.
 *
 * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
 * or http://www.opensolaris.org/os/licensing.
 * See the License for the specific language governing permissions
 * and limitations under the License.
 *
 * When distributing Covered Code, include this CDDL HEADER in each
 * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
 * If applicable, add the following below this CDDL HEADER, with the
 * fields enclosed by brackets "[]" replaced with your own identifying
 * information: Portions Copyright [yyyy] [name of copyright owner]
 *
 * CDDL HEADER END
 */

/*
 * Copyright (C) 2011 Lawrence Livermore National Security, LLC.
 * Produced at Lawrence Livermore National Laboratory (cf, DISCLAIMER).
 * Written by Brian Behlendorf <behlendorf1@llnl.gov>.
 * LLNL-CODE-403049.
 */

#ifndef _ZFS_BLKDEV_H
#define	_ZFS_BLKDEV_H

#include <linux/blkdev.h>
#include <linux/elevator.h>

#ifndef HAVE_FMODE_T
typedef unsigned __bitwise__ fmode_t;
#endif /* HAVE_FMODE_T */

#ifndef HAVE_BLK_FETCH_REQUEST
static inline struct request *
blk_fetch_request(struct request_queue *q)
{
	struct request *req;

	req = elv_next_request(q);
	if (req)
		blkdev_dequeue_request(req);

	return (req);
}
#endif /* HAVE_BLK_FETCH_REQUEST */

#ifndef HAVE_BLK_REQUEUE_REQUEST
static inline void
blk_requeue_request(request_queue_t *q, struct request *req)
{
	elv_requeue_request(q, req);
}
#endif /* HAVE_BLK_REQUEUE_REQUEST */

#ifndef HAVE_BLK_END_REQUEST
static inline bool
__blk_end_request(struct request *req, int error, unsigned int nr_bytes)
{
	LIST_HEAD(list);

	/*
	 * Request has already been dequeued but 2.6.18 version of
	 * end_request() unconditionally dequeues the request so we
	 * add it to a local list to prevent hitting the BUG_ON.
	 */
	list_add(&req->queuelist, &list);

	/*
	 * The old API required the driver to end each segment and not
	 * the entire request.  In our case we always need to end the
	 * entire request partial requests are not supported.
	 */
	req->hard_cur_sectors = nr_bytes >> 9;
	end_request(req, ((error == 0) ? 1 : error));

	return (0);
}

static inline bool
blk_end_request(struct request *req, int error, unsigned int nr_bytes)
{
	struct request_queue *q = req->q;
	bool rc;

	spin_lock_irq(q->queue_lock);
	rc = __blk_end_request(req, error, nr_bytes);
	spin_unlock_irq(q->queue_lock);

	return (rc);
}
#else
#ifdef HAVE_BLK_END_REQUEST_GPL_ONLY
/*
 * Define required to avoid conflicting 2.6.29 non-static prototype for a
 * GPL-only version of the helper.  As of 2.6.31 the helper is available
 * to non-GPL modules and is not explicitly exported GPL-only.
 */
#define	__blk_end_request __blk_end_request_x
#define	blk_end_request blk_end_request_x

static inline bool
__blk_end_request_x(struct request *req, int error, unsigned int nr_bytes)
{
	/*
	 * The old API required the driver to end each segment and not
	 * the entire request.  In our case we always need to end the
	 * entire request partial requests are not supported.
	 */
	req->hard_cur_sectors = nr_bytes >> 9;
	end_request(req, ((error == 0) ? 1 : error));

	return (0);
}
static inline bool
blk_end_request_x(struct request *req, int error, unsigned int nr_bytes)
{
	struct request_queue *q = req->q;
	bool rc;

	spin_lock_irq(q->queue_lock);
	rc = __blk_end_request_x(req, error, nr_bytes);
	spin_unlock_irq(q->queue_lock);

	return (rc);
}
#endif /* HAVE_BLK_END_REQUEST_GPL_ONLY */
#endif /* HAVE_BLK_END_REQUEST */

/*
 * 2.6.36 API change,
 * The blk_queue_flush() interface has replaced blk_queue_ordered()
 * interface.  However, while the old interface was available to all the
 * new one is GPL-only.   Thus if the GPL-only version is detected we
 * implement our own trivial helper compatibility funcion.   The hope is
 * that long term this function will be opened up.
 */
#if defined(HAVE_BLK_QUEUE_FLUSH) && defined(HAVE_BLK_QUEUE_FLUSH_GPL_ONLY)
#define	blk_queue_flush __blk_queue_flush
static inline void
__blk_queue_flush(struct request_queue *q, unsigned int flags)
{
	q->flush_flags = flags & (REQ_FLUSH | REQ_FUA);
}
#endif /* HAVE_BLK_QUEUE_FLUSH && HAVE_BLK_QUEUE_FLUSH_GPL_ONLY */

#ifndef HAVE_BLK_RQ_POS
static inline sector_t
blk_rq_pos(struct request *req)
{
	return (req->sector);
}
#endif /* HAVE_BLK_RQ_POS */

#ifndef HAVE_BLK_RQ_SECTORS
static inline unsigned int
blk_rq_sectors(struct request *req)
{
	return (req->nr_sectors);
}
#endif /* HAVE_BLK_RQ_SECTORS */

#if !defined(HAVE_BLK_RQ_BYTES) || defined(HAVE_BLK_RQ_BYTES_GPL_ONLY)
/*
 * Define required to avoid conflicting 2.6.29 non-static prototype for a
 * GPL-only version of the helper.  As of 2.6.31 the helper is available
 * to non-GPL modules in the form of a static inline in the header.
 */
#define	blk_rq_bytes __blk_rq_bytes
static inline unsigned int
__blk_rq_bytes(struct request *req)
{
	return (blk_rq_sectors(req) << 9);
}
#endif /* !HAVE_BLK_RQ_BYTES || HAVE_BLK_RQ_BYTES_GPL_ONLY */

/*
 * Most of the blk_* macros were removed in 2.6.36.  Ostensibly this was
 * done to improve readability and allow easier grepping.  However, from
 * a portability stand point the macros are helpful.  Therefore the needed
 * macros are redefined here if they are missing from the kernel.
 */
#ifndef blk_fs_request
#define	blk_fs_request(rq)	((rq)->cmd_type == REQ_TYPE_FS)
#endif

/*
 * 2.6.27 API change,
 * The blk_queue_stackable() queue flag was added in 2.6.27 to handle dm
 * stacking drivers.  Prior to this request stacking drivers were detected
 * by checking (q->request_fn == NULL), for earlier kernels we revert to
 * this legacy behavior.
 */
#ifndef blk_queue_stackable
#define	blk_queue_stackable(q)	((q)->request_fn == NULL)
#endif

/*
 * 2.6.34 API change,
 * The blk_queue_max_hw_sectors() function replaces blk_queue_max_sectors().
 */
#ifndef HAVE_BLK_QUEUE_MAX_HW_SECTORS
#define	blk_queue_max_hw_sectors __blk_queue_max_hw_sectors
static inline void
__blk_queue_max_hw_sectors(struct request_queue *q, unsigned int max_hw_sectors)
{
	blk_queue_max_sectors(q, max_hw_sectors);
}
#endif

/*
 * 2.6.34 API change,
 * The blk_queue_max_segments() function consolidates
 * blk_queue_max_hw_segments() and blk_queue_max_phys_segments().
 */
#ifndef HAVE_BLK_QUEUE_MAX_SEGMENTS
#define	blk_queue_max_segments __blk_queue_max_segments
static inline void
__blk_queue_max_segments(struct request_queue *q, unsigned short max_segments)
{
	blk_queue_max_phys_segments(q, max_segments);
	blk_queue_max_hw_segments(q, max_segments);
}
#endif

/*
 * 2.6.30 API change,
 * The blk_queue_physical_block_size() function was introduced to
 * indicate the smallest I/O the device can write without incurring
 * a read-modify-write penalty.  For older kernels this is a no-op.
 */
#ifndef HAVE_BLK_QUEUE_PHYSICAL_BLOCK_SIZE
#define	blk_queue_physical_block_size(q, x)	((void)(0))
#endif

/*
 * 2.6.30 API change,
 * The blk_queue_io_opt() function was added to indicate the optimal
 * I/O size for the device.  For older kernels this is a no-op.
 */
#ifndef HAVE_BLK_QUEUE_IO_OPT
#define	blk_queue_io_opt(q, x)			((void)(0))
#endif

#ifndef HAVE_GET_DISK_RO
static inline int
get_disk_ro(struct gendisk *disk)
{
	int policy = 0;

	if (disk->part[0])
		policy = disk->part[0]->policy;

	return (policy);
}
#endif /* HAVE_GET_DISK_RO */

#ifndef HAVE_RQ_IS_SYNC
static inline bool
rq_is_sync(struct request *req)
{
	return (req->flags & REQ_RW_SYNC);
}
#endif /* HAVE_RQ_IS_SYNC */

#ifndef HAVE_RQ_FOR_EACH_SEGMENT
struct req_iterator {
	int i;
	struct bio *bio;
};

#define	for_each_bio(_bio)              \
	for (; _bio; _bio = _bio->bi_next)

#define	__rq_for_each_bio(_bio, rq)     \
	if ((rq->bio))                  \
		for (_bio = (rq)->bio; _bio; _bio = _bio->bi_next)

#define	rq_for_each_segment(bvl, _rq, _iter)                    \
	__rq_for_each_bio(_iter.bio, _rq)                       \
		bio_for_each_segment(bvl, _iter.bio, _iter.i)

#define	HAVE_RQ_FOR_EACH_SEGMENT_BVP 1
#endif /* HAVE_RQ_FOR_EACH_SEGMENT */

/*
 * 3.14 API change
 * rq_for_each_segment changed from taking bio_vec * to taking bio_vec.
 * We provide rq_for_each_segment4 which takes both.
 * You should not modify the fields in @bv and @bvp.
 *
 * Note: the if-else is just to inject the assignment before the loop body.
 */
#ifdef HAVE_RQ_FOR_EACH_SEGMENT_BVP
#define	rq_for_each_segment4(bv, bvp, rq, iter)	\
	rq_for_each_segment(bvp, rq, iter)	\
		if ((bv = *bvp), 0)		\
			;			\
		else
#else
#define	rq_for_each_segment4(bv, bvp, rq, iter)	\
	rq_for_each_segment(bv, rq, iter)	\
		if ((bvp = &bv), 0)		\
			;			\
		else
#endif

#ifdef HAVE_BIO_BVEC_ITER
#define	BIO_BI_SECTOR(bio)	(bio)->bi_iter.bi_sector
#define	BIO_BI_SIZE(bio)	(bio)->bi_iter.bi_size
#define	BIO_BI_IDX(bio)		(bio)->bi_iter.bi_idx
#else
#define	BIO_BI_SECTOR(bio)	(bio)->bi_sector
#define	BIO_BI_SIZE(bio)	(bio)->bi_size
#define	BIO_BI_IDX(bio)		(bio)->bi_idx
#endif

/*
 * Portable helper for correctly setting the FAILFAST flags.  The
 * correct usage has changed 3 times from 2.6.12 to 2.6.38.
 */
static inline void
bio_set_flags_failfast(struct block_device *bdev, int *flags)
{
#ifdef CONFIG_BUG
	/*
	 * Disable FAILFAST for loopback devices because of the
	 * following incorrect BUG_ON() in loop_make_request().
	 * This support is also disabled for md devices because the
	 * test suite layers md devices on top of loopback devices.
	 * This may be removed when the loopback driver is fixed.
	 *
	 *   BUG_ON(!lo || (rw != READ && rw != WRITE));
	 */
	if ((MAJOR(bdev->bd_dev) == LOOP_MAJOR) ||
	    (MAJOR(bdev->bd_dev) == MD_MAJOR))
		return;

#ifdef BLOCK_EXT_MAJOR
	if (MAJOR(bdev->bd_dev) == BLOCK_EXT_MAJOR)
		return;
#endif /* BLOCK_EXT_MAJOR */
#endif /* CONFIG_BUG */

#if defined(HAVE_BIO_RW_FAILFAST_DTD)
	/* BIO_RW_FAILFAST_* preferred interface from 2.6.28 - 2.6.35 */
	*flags |= (
	    (1 << BIO_RW_FAILFAST_DEV) |
	    (1 << BIO_RW_FAILFAST_TRANSPORT) |
	    (1 << BIO_RW_FAILFAST_DRIVER));
#elif defined(HAVE_REQ_FAILFAST_MASK)
	/*
	 * REQ_FAILFAST_* preferred interface from 2.6.36 - 2.6.xx,
	 * the BIO_* and REQ_* flags were unified under REQ_* flags.
	 */
	*flags |= REQ_FAILFAST_MASK;
#else
#error "Undefined block IO FAILFAST interface."
#endif
}

/*
 * Maximum disk label length, it may be undefined for some kernels.
 */
#ifndef DISK_NAME_LEN
#define	DISK_NAME_LEN	32
#endif /* DISK_NAME_LEN */

/*
 * 2.6.24 API change,
 * The bio_end_io() prototype changed slightly.  These are helper
 * macro's to ensure the prototype and return value are handled.
 */
#ifdef HAVE_2ARGS_BIO_END_IO_T
#define	BIO_END_IO_PROTO(fn, x, y, z)	static void fn(struct bio *x, int z)
#define	BIO_END_IO_RETURN(rc)		return
#else
#define	BIO_END_IO_PROTO(fn, x, y, z)	static int fn( \
					    struct bio *x, \
					    unsigned int y, \
					    int z)
#define	BIO_END_IO_RETURN(rc)		return rc
#endif /* HAVE_2ARGS_BIO_END_IO_T */

/*
 * 2.6.38 - 2.6.x API,
 *   blkdev_get_by_path()
 *   blkdev_put()
 *
 * 2.6.28 - 2.6.37 API,
 *   open_bdev_exclusive()
 *   close_bdev_exclusive()
 *
 * 2.6.12 - 2.6.27 API,
 *   open_bdev_excl()
 *   close_bdev_excl()
 *
 * Used to exclusively open a block device from within the kernel.
 */
#if defined(HAVE_BLKDEV_GET_BY_PATH)
#define	vdev_bdev_open(path, md, hld)	blkdev_get_by_path(path, \
					    (md) | FMODE_EXCL, hld)
#define	vdev_bdev_close(bdev, md)	blkdev_put(bdev, (md) | FMODE_EXCL)
#elif defined(HAVE_OPEN_BDEV_EXCLUSIVE)
#define	vdev_bdev_open(path, md, hld)	open_bdev_exclusive(path, md, hld)
#define	vdev_bdev_close(bdev, md)	close_bdev_exclusive(bdev, md)
#else
#define	vdev_bdev_open(path, md, hld)	open_bdev_excl(path, md, hld)
#define	vdev_bdev_close(bdev, md)	close_bdev_excl(bdev)
#endif /* HAVE_BLKDEV_GET_BY_PATH | HAVE_OPEN_BDEV_EXCLUSIVE */

/*
 * 2.6.22 API change
 * The function invalidate_bdev() lost it's second argument because
 * it was unused.
 */
#ifdef HAVE_1ARG_INVALIDATE_BDEV
#define	vdev_bdev_invalidate(bdev)	invalidate_bdev(bdev)
#else
#define	vdev_bdev_invalidate(bdev)	invalidate_bdev(bdev, 1)
#endif /* HAVE_1ARG_INVALIDATE_BDEV */

/*
 * 2.6.27 API change
 * The function was exported for use, prior to this it existed by the
 * symbol was not exported.
 */
#ifndef HAVE_LOOKUP_BDEV
#define	lookup_bdev(path)		ERR_PTR(-ENOTSUP)
#endif

/*
 * 2.6.30 API change
 * To ensure good performance preferentially use the physical block size
 * for proper alignment.  The physical size is supposed to be the internal
 * sector size used by the device.  This is often 4096 byte for AF devices,
 * while a smaller 512 byte logical size is supported for compatibility.
 *
 * Unfortunately, many drives still misreport their physical sector size.
 * For devices which are known to lie you may need to manually set this
 * at pool creation time with 'zpool create -o ashift=12 ...'.
 *
 * When the physical block size interface isn't available, we fall back to
 * the logical block size interface and then the older hard sector size.
 */
#ifdef HAVE_BDEV_PHYSICAL_BLOCK_SIZE
#define	vdev_bdev_block_size(bdev)	bdev_physical_block_size(bdev)
#else
#ifdef HAVE_BDEV_LOGICAL_BLOCK_SIZE
#define	vdev_bdev_block_size(bdev)	bdev_logical_block_size(bdev)
#else
#define	vdev_bdev_block_size(bdev)	bdev_hardsect_size(bdev)
#endif /* HAVE_BDEV_LOGICAL_BLOCK_SIZE */
#endif /* HAVE_BDEV_PHYSICAL_BLOCK_SIZE */

/*
 * 2.6.37 API change
 * The WRITE_FLUSH, WRITE_FUA, and WRITE_FLUSH_FUA flags have been
 * introduced as a replacement for WRITE_BARRIER.  This was done to
 * allow richer semantics to be expressed to the block layer.  It is
 * the block layers responsibility to choose the correct way to
 * implement these semantics.
 *
 * The existence of these flags implies that REQ_FLUSH an REQ_FUA are
 * defined.  Thus we can safely define VDEV_REQ_FLUSH and VDEV_REQ_FUA
 * compatibility macros.
 */
#ifdef WRITE_FLUSH_FUA
#define	VDEV_WRITE_FLUSH_FUA		WRITE_FLUSH_FUA
#define	VDEV_REQ_FLUSH			REQ_FLUSH
#define	VDEV_REQ_FUA			REQ_FUA
#else
#define	VDEV_WRITE_FLUSH_FUA		WRITE_BARRIER
#define	VDEV_REQ_FLUSH			REQ_HARDBARRIER
#define	VDEV_REQ_FUA			REQ_HARDBARRIER
#endif

/*
 * 2.6.32 API change
 * Use the normal I/O patch for discards.
 */
#ifdef REQ_DISCARD
#define	VDEV_REQ_DISCARD		REQ_DISCARD
#endif

/*
 * 2.6.33 API change
 * Discard granularity and alignment restrictions may now be set.  For
 * older kernels which do not support this it is safe to skip it.
 */
#ifdef HAVE_DISCARD_GRANULARITY
static inline void
blk_queue_discard_granularity(struct request_queue *q, unsigned int dg)
{
	q->limits.discard_granularity = dg;
}
#else
#define	blk_queue_discard_granularity(x, dg)	((void)0)
#endif /* HAVE_DISCARD_GRANULARITY */

/*
 * Default Linux IO Scheduler,
 * Setting the scheduler to noop will allow the Linux IO scheduler to
 * still perform front and back merging, while leaving the request
 * ordering and prioritization to the ZFS IO scheduler.
 */
#define	VDEV_SCHEDULER			"noop"

/*
 * A common holder for vdev_bdev_open() is used to relax the exclusive open
 * semantics slightly.  Internal vdev disk callers may pass VDEV_HOLDER to
 * allow them to open the device multiple times.  Other kernel callers and
 * user space processes which don't pass this value will get EBUSY.  This is
 * currently required for the correct operation of hot spares.
 */
#define	VDEV_HOLDER			((void *)0x2401de7)

#endif /* _ZFS_BLKDEV_H */