genhd.c 44.4 KB
Newer Older
Linus Torvalds's avatar
Linus Torvalds committed
1 2 3 4 5 6 7
/*
 *  gendisk handling
 */

#include <linux/module.h>
#include <linux/fs.h>
#include <linux/genhd.h>
8
#include <linux/kdev_t.h>
Linus Torvalds's avatar
Linus Torvalds committed
9 10
#include <linux/kernel.h>
#include <linux/blkdev.h>
11
#include <linux/backing-dev.h>
Linus Torvalds's avatar
Linus Torvalds committed
12 13
#include <linux/init.h>
#include <linux/spinlock.h>
14
#include <linux/proc_fs.h>
Linus Torvalds's avatar
Linus Torvalds committed
15 16 17 18
#include <linux/seq_file.h>
#include <linux/slab.h>
#include <linux/kmod.h>
#include <linux/kobj_map.h>
19
#include <linux/mutex.h>
20
#include <linux/idr.h>
21
#include <linux/log2.h>
22
#include <linux/pm_runtime.h>
Linus Torvalds's avatar
Linus Torvalds committed
23

24 25
#include "blk.h"

26 27
static DEFINE_MUTEX(block_class_lock);
struct kobject *block_depr;
Linus Torvalds's avatar
Linus Torvalds committed
28

29
/* for extended dynamic devt allocation, currently only one major is used */
30
#define NR_EXT_DEVT		(1 << MINORBITS)
31

32
/* For extended devt allocation.  ext_devt_lock prevents look up
33 34
 * results from going away underneath its user.
 */
35
static DEFINE_SPINLOCK(ext_devt_lock);
36 37
static DEFINE_IDR(ext_devt_idr);

Adrian Bunk's avatar
Adrian Bunk committed
38 39
static struct device_type disk_type;

Derek Basehore's avatar
Derek Basehore committed
40 41
static void disk_check_events(struct disk_events *ev,
			      unsigned int *clearing_ptr);
42
static void disk_alloc_events(struct gendisk *disk);
43 44 45 46
static void disk_add_events(struct gendisk *disk);
static void disk_del_events(struct gendisk *disk);
static void disk_release_events(struct gendisk *disk);

47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62
/**
 * disk_get_part - get partition
 * @disk: disk to look partition from
 * @partno: partition number
 *
 * Look for partition @partno from @disk.  If found, increment
 * reference count and return it.
 *
 * CONTEXT:
 * Don't care.
 *
 * RETURNS:
 * Pointer to the found partition on success, NULL if not found.
 */
struct hd_struct *disk_get_part(struct gendisk *disk, int partno)
{
63 64
	struct hd_struct *part = NULL;
	struct disk_part_tbl *ptbl;
65

66
	if (unlikely(partno < 0))
67
		return NULL;
68

69
	rcu_read_lock();
70 71 72 73 74 75 76 77

	ptbl = rcu_dereference(disk->part_tbl);
	if (likely(partno < ptbl->len)) {
		part = rcu_dereference(ptbl->part[partno]);
		if (part)
			get_device(part_to_dev(part));
	}

78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97
	rcu_read_unlock();

	return part;
}
EXPORT_SYMBOL_GPL(disk_get_part);

/**
 * disk_part_iter_init - initialize partition iterator
 * @piter: iterator to initialize
 * @disk: disk to iterate over
 * @flags: DISK_PITER_* flags
 *
 * Initialize @piter so that it iterates over partitions of @disk.
 *
 * CONTEXT:
 * Don't care.
 */
void disk_part_iter_init(struct disk_part_iter *piter, struct gendisk *disk,
			  unsigned int flags)
{
98 99 100 101 102
	struct disk_part_tbl *ptbl;

	rcu_read_lock();
	ptbl = rcu_dereference(disk->part_tbl);

103 104 105 106
	piter->disk = disk;
	piter->part = NULL;

	if (flags & DISK_PITER_REVERSE)
107
		piter->idx = ptbl->len - 1;
108
	else if (flags & (DISK_PITER_INCL_PART0 | DISK_PITER_INCL_EMPTY_PART0))
109
		piter->idx = 0;
Tejun Heo's avatar
Tejun Heo committed
110 111
	else
		piter->idx = 1;
112 113

	piter->flags = flags;
114 115

	rcu_read_unlock();
116 117 118 119 120 121 122 123 124 125 126 127 128 129
}
EXPORT_SYMBOL_GPL(disk_part_iter_init);

/**
 * disk_part_iter_next - proceed iterator to the next partition and return it
 * @piter: iterator of interest
 *
 * Proceed @piter to the next partition and return it.
 *
 * CONTEXT:
 * Don't care.
 */
struct hd_struct *disk_part_iter_next(struct disk_part_iter *piter)
{
130
	struct disk_part_tbl *ptbl;
131 132 133 134 135 136
	int inc, end;

	/* put the last partition */
	disk_put_part(piter->part);
	piter->part = NULL;

137
	/* get part_tbl */
138
	rcu_read_lock();
139
	ptbl = rcu_dereference(piter->disk->part_tbl);
140 141 142 143

	/* determine iteration parameters */
	if (piter->flags & DISK_PITER_REVERSE) {
		inc = -1;
144 145
		if (piter->flags & (DISK_PITER_INCL_PART0 |
				    DISK_PITER_INCL_EMPTY_PART0))
Tejun Heo's avatar
Tejun Heo committed
146 147 148
			end = -1;
		else
			end = 0;
149 150
	} else {
		inc = 1;
151
		end = ptbl->len;
152 153 154 155 156 157
	}

	/* iterate to the next partition */
	for (; piter->idx != end; piter->idx += inc) {
		struct hd_struct *part;

158
		part = rcu_dereference(ptbl->part[piter->idx]);
159 160
		if (!part)
			continue;
161
		if (!part_nr_sects_read(part) &&
162 163 164
		    !(piter->flags & DISK_PITER_INCL_EMPTY) &&
		    !(piter->flags & DISK_PITER_INCL_EMPTY_PART0 &&
		      piter->idx == 0))
165 166
			continue;

167
		get_device(part_to_dev(part));
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
		piter->part = part;
		piter->idx += inc;
		break;
	}

	rcu_read_unlock();

	return piter->part;
}
EXPORT_SYMBOL_GPL(disk_part_iter_next);

/**
 * disk_part_iter_exit - finish up partition iteration
 * @piter: iter of interest
 *
 * Called when iteration is over.  Cleans up @piter.
 *
 * CONTEXT:
 * Don't care.
 */
void disk_part_iter_exit(struct disk_part_iter *piter)
{
	disk_put_part(piter->part);
	piter->part = NULL;
}
EXPORT_SYMBOL_GPL(disk_part_iter_exit);

195 196 197
static inline int sector_in_part(struct hd_struct *part, sector_t sector)
{
	return part->start_sect <= sector &&
198
		sector < part->start_sect + part_nr_sects_read(part);
199 200
}

201 202 203 204 205 206 207 208 209 210 211 212 213
/**
 * disk_map_sector_rcu - map sector to partition
 * @disk: gendisk of interest
 * @sector: sector to map
 *
 * Find out which partition @sector maps to on @disk.  This is
 * primarily used for stats accounting.
 *
 * CONTEXT:
 * RCU read locked.  The returned partition pointer is valid only
 * while preemption is disabled.
 *
 * RETURNS:
214
 * Found partition on success, part0 is returned if no partition matches
215 216 217
 */
struct hd_struct *disk_map_sector_rcu(struct gendisk *disk, sector_t sector)
{
218
	struct disk_part_tbl *ptbl;
219
	struct hd_struct *part;
220 221
	int i;

222 223
	ptbl = rcu_dereference(disk->part_tbl);

224 225 226 227
	part = rcu_dereference(ptbl->last_lookup);
	if (part && sector_in_part(part, sector))
		return part;

228
	for (i = 1; i < ptbl->len; i++) {
229
		part = rcu_dereference(ptbl->part[i]);
230

231 232
		if (part && sector_in_part(part, sector)) {
			rcu_assign_pointer(ptbl->last_lookup, part);
233
			return part;
234
		}
235
	}
236
	return &disk->part0;
237 238 239
}
EXPORT_SYMBOL_GPL(disk_map_sector_rcu);

Linus Torvalds's avatar
Linus Torvalds committed
240 241 242 243 244 245 246 247
/*
 * Can be deleted altogether. Later.
 *
 */
static struct blk_major_name {
	struct blk_major_name *next;
	int major;
	char name[16];
248
} *major_names[BLKDEV_MAJOR_HASH_SIZE];
Linus Torvalds's avatar
Linus Torvalds committed
249 250

/* index in the above - for now: assume no multimajor ranges */
251
static inline int major_to_index(unsigned major)
Linus Torvalds's avatar
Linus Torvalds committed
252
{
253
	return major % BLKDEV_MAJOR_HASH_SIZE;
254 255
}

256
#ifdef CONFIG_PROC_FS
257
void blkdev_show(struct seq_file *seqf, off_t offset)
258
{
259
	struct blk_major_name *dp;
260

261
	if (offset < BLKDEV_MAJOR_HASH_SIZE) {
262
		mutex_lock(&block_class_lock);
263
		for (dp = major_names[offset]; dp; dp = dp->next)
264
			seq_printf(seqf, "%3d %s\n", dp->major, dp->name);
265
		mutex_unlock(&block_class_lock);
Linus Torvalds's avatar
Linus Torvalds committed
266 267
	}
}
268
#endif /* CONFIG_PROC_FS */
Linus Torvalds's avatar
Linus Torvalds committed
269

270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285
/**
 * register_blkdev - register a new block device
 *
 * @major: the requested major device number [1..255]. If @major=0, try to
 *         allocate any unused major number.
 * @name: the name of the new block device as a zero terminated string
 *
 * The @name must be unique within the system.
 *
 * The return value depends on the @major input parameter.
 *  - if a major device number was requested in range [1..255] then the
 *    function returns zero on success, or a negative error code
 *  - if any unused major number was requested with @major=0 parameter
 *    then the return value is the allocated major number in range
 *    [1..255] or a negative error code otherwise
 */
Linus Torvalds's avatar
Linus Torvalds committed
286 287 288 289 290
int register_blkdev(unsigned int major, const char *name)
{
	struct blk_major_name **n, *p;
	int index, ret = 0;

291
	mutex_lock(&block_class_lock);
Linus Torvalds's avatar
Linus Torvalds committed
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

	/* temporary */
	if (major == 0) {
		for (index = ARRAY_SIZE(major_names)-1; index > 0; index--) {
			if (major_names[index] == NULL)
				break;
		}

		if (index == 0) {
			printk("register_blkdev: failed to get major for %s\n",
			       name);
			ret = -EBUSY;
			goto out;
		}
		major = index;
		ret = major;
	}

	p = kmalloc(sizeof(struct blk_major_name), GFP_KERNEL);
	if (p == NULL) {
		ret = -ENOMEM;
		goto out;
	}

	p->major = major;
	strlcpy(p->name, name, sizeof(p->name));
	p->next = NULL;
	index = major_to_index(major);

	for (n = &major_names[index]; *n; n = &(*n)->next) {
		if ((*n)->major == major)
			break;
	}
	if (!*n)
		*n = p;
	else
		ret = -EBUSY;

	if (ret < 0) {
		printk("register_blkdev: cannot get major %d for %s\n",
		       major, name);
		kfree(p);
	}
out:
336
	mutex_unlock(&block_class_lock);
Linus Torvalds's avatar
Linus Torvalds committed
337 338 339 340 341
	return ret;
}

EXPORT_SYMBOL(register_blkdev);

342
void unregister_blkdev(unsigned int major, const char *name)
Linus Torvalds's avatar
Linus Torvalds committed
343 344 345 346 347
{
	struct blk_major_name **n;
	struct blk_major_name *p = NULL;
	int index = major_to_index(major);

348
	mutex_lock(&block_class_lock);
Linus Torvalds's avatar
Linus Torvalds committed
349 350 351
	for (n = &major_names[index]; *n; n = &(*n)->next)
		if ((*n)->major == major)
			break;
352 353 354
	if (!*n || strcmp((*n)->name, name)) {
		WARN_ON(1);
	} else {
Linus Torvalds's avatar
Linus Torvalds committed
355 356 357
		p = *n;
		*n = p->next;
	}
358
	mutex_unlock(&block_class_lock);
Linus Torvalds's avatar
Linus Torvalds committed
359 360 361 362 363 364 365
	kfree(p);
}

EXPORT_SYMBOL(unregister_blkdev);

static struct kobj_map *bdev_map;

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
/**
 * blk_mangle_minor - scatter minor numbers apart
 * @minor: minor number to mangle
 *
 * Scatter consecutively allocated @minor number apart if MANGLE_DEVT
 * is enabled.  Mangling twice gives the original value.
 *
 * RETURNS:
 * Mangled value.
 *
 * CONTEXT:
 * Don't care.
 */
static int blk_mangle_minor(int minor)
{
#ifdef CONFIG_DEBUG_BLOCK_EXT_DEVT
	int i;

	for (i = 0; i < MINORBITS / 2; i++) {
		int low = minor & (1 << i);
		int high = minor & (1 << (MINORBITS - 1 - i));
		int distance = MINORBITS - 1 - 2 * i;

		minor ^= low | high;	/* clear both bits */
		low <<= distance;	/* swap the positions */
		high >>= distance;
		minor |= low | high;	/* and set */
	}
#endif
	return minor;
}

398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414
/**
 * blk_alloc_devt - allocate a dev_t for a partition
 * @part: partition to allocate dev_t for
 * @devt: out parameter for resulting dev_t
 *
 * Allocate a dev_t for block device.
 *
 * RETURNS:
 * 0 on success, allocated dev_t is returned in *@devt.  -errno on
 * failure.
 *
 * CONTEXT:
 * Might sleep.
 */
int blk_alloc_devt(struct hd_struct *part, dev_t *devt)
{
	struct gendisk *disk = part_to_disk(part);
Tejun Heo's avatar
Tejun Heo committed
415
	int idx;
416 417 418 419 420 421 422 423

	/* in consecutive minor range? */
	if (part->partno < disk->minors) {
		*devt = MKDEV(disk->major, disk->first_minor + part->partno);
		return 0;
	}

	/* allocate ext devt */
424 425
	idr_preload(GFP_KERNEL);

426
	spin_lock_bh(&ext_devt_lock);
427
	idx = idr_alloc(&ext_devt_idr, part, 0, NR_EXT_DEVT, GFP_NOWAIT);
428
	spin_unlock_bh(&ext_devt_lock);
429 430

	idr_preload_end();
Tejun Heo's avatar
Tejun Heo committed
431 432
	if (idx < 0)
		return idx == -ENOSPC ? -EBUSY : idx;
433

434
	*devt = MKDEV(BLOCK_EXT_MAJOR, blk_mangle_minor(idx));
435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452
	return 0;
}

/**
 * blk_free_devt - free a dev_t
 * @devt: dev_t to free
 *
 * Free @devt which was allocated using blk_alloc_devt().
 *
 * CONTEXT:
 * Might sleep.
 */
void blk_free_devt(dev_t devt)
{
	if (devt == MKDEV(0, 0))
		return;

	if (MAJOR(devt) == BLOCK_EXT_MAJOR) {
453
		spin_lock_bh(&ext_devt_lock);
454
		idr_remove(&ext_devt_idr, blk_mangle_minor(MINOR(devt)));
455
		spin_unlock_bh(&ext_devt_lock);
456 457 458
	}
}

459 460 461 462 463 464 465 466 467 468 469 470
static char *bdevt_str(dev_t devt, char *buf)
{
	if (MAJOR(devt) <= 0xff && MINOR(devt) <= 0xff) {
		char tbuf[BDEVT_SIZE];
		snprintf(tbuf, BDEVT_SIZE, "%02x%02x", MAJOR(devt), MINOR(devt));
		snprintf(buf, BDEVT_SIZE, "%-9s", tbuf);
	} else
		snprintf(buf, BDEVT_SIZE, "%03x:%05x", MAJOR(devt), MINOR(devt));

	return buf;
}

Linus Torvalds's avatar
Linus Torvalds committed
471 472 473 474 475
/*
 * Register device numbers dev..(dev+range-1)
 * range must be nonzero
 * The hash chain is sorted on range, so that subranges can override.
 */
476
void blk_register_region(dev_t devt, unsigned long range, struct module *module,
Linus Torvalds's avatar
Linus Torvalds committed
477 478 479
			 struct kobject *(*probe)(dev_t, int *, void *),
			 int (*lock)(dev_t, void *), void *data)
{
480
	kobj_map(bdev_map, devt, range, module, probe, lock, data);
Linus Torvalds's avatar
Linus Torvalds committed
481 482 483 484
}

EXPORT_SYMBOL(blk_register_region);

485
void blk_unregister_region(dev_t devt, unsigned long range)
Linus Torvalds's avatar
Linus Torvalds committed
486
{
487
	kobj_unmap(bdev_map, devt, range);
Linus Torvalds's avatar
Linus Torvalds committed
488 489 490 491
}

EXPORT_SYMBOL(blk_unregister_region);

492
static struct kobject *exact_match(dev_t devt, int *partno, void *data)
Linus Torvalds's avatar
Linus Torvalds committed
493 494
{
	struct gendisk *p = data;
495

496
	return &disk_to_dev(p)->kobj;
Linus Torvalds's avatar
Linus Torvalds committed
497 498
}

499
static int exact_lock(dev_t devt, void *data)
Linus Torvalds's avatar
Linus Torvalds committed
500 501 502 503 504 505 506 507
{
	struct gendisk *p = data;

	if (!get_disk(p))
		return -1;
	return 0;
}

Al Viro's avatar
Al Viro committed
508
static void register_disk(struct gendisk *disk)
509 510 511 512 513 514 515 516 517
{
	struct device *ddev = disk_to_dev(disk);
	struct block_device *bdev;
	struct disk_part_iter piter;
	struct hd_struct *part;
	int err;

	ddev->parent = disk->driverfs_dev;

518
	dev_set_name(ddev, "%s", disk->disk_name);
519 520 521 522 523 524 525 526 527 528 529 530 531 532

	/* delay uevents, until we scanned partition table */
	dev_set_uevent_suppress(ddev, 1);

	if (device_add(ddev))
		return;
	if (!sysfs_deprecated) {
		err = sysfs_create_link(block_depr, &ddev->kobj,
					kobject_name(&ddev->kobj));
		if (err) {
			device_del(ddev);
			return;
		}
	}
533 534 535 536 537 538 539 540

	/*
	 * avoid probable deadlock caused by allocating memory with
	 * GFP_KERNEL in runtime_resume callback of its all ancestor
	 * devices
	 */
	pm_runtime_set_memalloc_noio(ddev, true);

541 542 543 544
	disk->part0.holder_dir = kobject_create_and_add("holders", &ddev->kobj);
	disk->slave_dir = kobject_create_and_add("slaves", &ddev->kobj);

	/* No minors to use for partitions */
Tejun Heo's avatar
Tejun Heo committed
545
	if (!disk_part_scan_enabled(disk))
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
		goto exit;

	/* No such device (e.g., media were just removed) */
	if (!get_capacity(disk))
		goto exit;

	bdev = bdget_disk(disk, 0);
	if (!bdev)
		goto exit;

	bdev->bd_invalidated = 1;
	err = blkdev_get(bdev, FMODE_READ, NULL);
	if (err < 0)
		goto exit;
	blkdev_put(bdev, FMODE_READ);

exit:
	/* announce disk after possible partitions are created */
	dev_set_uevent_suppress(ddev, 0);
	kobject_uevent(&ddev->kobj, KOBJ_ADD);

	/* announce possible partitions */
	disk_part_iter_init(&piter, disk, 0);
	while ((part = disk_part_iter_next(&piter)))
		kobject_uevent(&part_to_dev(part)->kobj, KOBJ_ADD);
	disk_part_iter_exit(&piter);
}

Linus Torvalds's avatar
Linus Torvalds committed
574 575 576 577 578 579
/**
 * add_disk - add partitioning information to kernel list
 * @disk: per-device partitioning information
 *
 * This function registers the partitioning information in @disk
 * with the kernel.
580 581
 *
 * FIXME: error handling
Linus Torvalds's avatar
Linus Torvalds committed
582 583 584
 */
void add_disk(struct gendisk *disk)
{
585
	struct backing_dev_info *bdi;
586
	dev_t devt;
587
	int retval;
588

589 590 591 592 593 594 595
	/* minors == 0 indicates to use ext devt from part0 and should
	 * be accompanied with EXT_DEVT flag.  Make sure all
	 * parameters make sense.
	 */
	WARN_ON(disk->minors && !(disk->major || disk->first_minor));
	WARN_ON(!disk->minors && !(disk->flags & GENHD_FL_EXT_DEVT));

Linus Torvalds's avatar
Linus Torvalds committed
596
	disk->flags |= GENHD_FL_UP;
597 598 599 600 601 602 603 604 605 606 607 608 609 610

	retval = blk_alloc_devt(&disk->part0, &devt);
	if (retval) {
		WARN_ON(1);
		return;
	}
	disk_to_dev(disk)->devt = devt;

	/* ->major and ->first_minor aren't supposed to be
	 * dereferenced from here on, but set them just in case.
	 */
	disk->major = MAJOR(devt);
	disk->first_minor = MINOR(devt);

611 612
	disk_alloc_events(disk);

613
	/* Register BDI before referencing it from bdev */
614
	bdi = &disk->queue->backing_dev_info;
615
	bdi_register_owner(bdi, disk_to_dev(disk));
616

617 618
	blk_register_region(disk_devt(disk), disk->minors, NULL,
			    exact_match, exact_lock, disk);
Linus Torvalds's avatar
Linus Torvalds committed
619 620
	register_disk(disk);
	blk_register_queue(disk);
621

622 623 624 625
	/*
	 * Take an extra ref on queue which will be put on disk_release()
	 * so that it sticks around as long as @disk is there.
	 */
626
	WARN_ON_ONCE(!blk_get_queue(disk->queue));
627

628 629
	retval = sysfs_create_link(&disk_to_dev(disk)->kobj, &bdi->dev->kobj,
				   "bdi");
630
	WARN_ON(retval);
Linus Torvalds's avatar
Linus Torvalds committed
631

632
	disk_add_events(disk);
633
	blk_integrity_add(disk);
Linus Torvalds's avatar
Linus Torvalds committed
634 635 636
}
EXPORT_SYMBOL(add_disk);

637
void del_gendisk(struct gendisk *disk)
Linus Torvalds's avatar
Linus Torvalds committed
638
{
639 640 641
	struct disk_part_iter piter;
	struct hd_struct *part;

642
	blk_integrity_del(disk);
643 644
	disk_del_events(disk);

645 646 647 648 649 650 651 652 653 654 655 656 657
	/* invalidate stuff */
	disk_part_iter_init(&piter, disk,
			     DISK_PITER_INCL_EMPTY | DISK_PITER_REVERSE);
	while ((part = disk_part_iter_next(&piter))) {
		invalidate_partition(disk, part->partno);
		delete_partition(disk, part->partno);
	}
	disk_part_iter_exit(&piter);

	invalidate_partition(disk, 0);
	set_capacity(disk, 0);
	disk->flags &= ~GENHD_FL_UP;

658
	sysfs_remove_link(&disk_to_dev(disk)->kobj, "bdi");
Linus Torvalds's avatar
Linus Torvalds committed
659
	blk_unregister_queue(disk);
660
	blk_unregister_region(disk_devt(disk), disk->minors);
661 662 663 664 665 666 667 668

	part_stat_set_all(&disk->part0, 0);
	disk->part0.stamp = 0;

	kobject_put(disk->part0.holder_dir);
	kobject_put(disk->slave_dir);
	if (!sysfs_deprecated)
		sysfs_remove_link(block_depr, dev_name(disk_to_dev(disk)));
669
	pm_runtime_set_memalloc_noio(disk_to_dev(disk), false);
670
	device_del(disk_to_dev(disk));
Linus Torvalds's avatar
Linus Torvalds committed
671
}
672
EXPORT_SYMBOL(del_gendisk);
Linus Torvalds's avatar
Linus Torvalds committed
673 674 675

/**
 * get_gendisk - get partitioning information for a given device
676
 * @devt: device to get partitioning information for
677
 * @partno: returned partition index
Linus Torvalds's avatar
Linus Torvalds committed
678 679
 *
 * This function gets the structure containing partitioning
680
 * information for the given device @devt.
Linus Torvalds's avatar
Linus Torvalds committed
681
 */
682
struct gendisk *get_gendisk(dev_t devt, int *partno)
Linus Torvalds's avatar
Linus Torvalds committed
683
{
684 685 686 687 688 689 690 691 692 693 694
	struct gendisk *disk = NULL;

	if (MAJOR(devt) != BLOCK_EXT_MAJOR) {
		struct kobject *kobj;

		kobj = kobj_lookup(bdev_map, devt, partno);
		if (kobj)
			disk = dev_to_disk(kobj_to_dev(kobj));
	} else {
		struct hd_struct *part;

695
		spin_lock_bh(&ext_devt_lock);
696
		part = idr_find(&ext_devt_idr, blk_mangle_minor(MINOR(devt)));
697 698 699 700
		if (part && get_disk(part_to_disk(part))) {
			*partno = part->partno;
			disk = part_to_disk(part);
		}
701
		spin_unlock_bh(&ext_devt_lock);
702
	}
703

704
	return disk;
Linus Torvalds's avatar
Linus Torvalds committed
705
}
706
EXPORT_SYMBOL(get_gendisk);
Linus Torvalds's avatar
Linus Torvalds committed
707

708 709 710 711 712 713 714 715 716 717 718 719 720
/**
 * bdget_disk - do bdget() by gendisk and partition number
 * @disk: gendisk of interest
 * @partno: partition number
 *
 * Find partition @partno from @disk, do bdget() on it.
 *
 * CONTEXT:
 * Don't care.
 *
 * RETURNS:
 * Resulting block_device on success, NULL on failure.
 */
721
struct block_device *bdget_disk(struct gendisk *disk, int partno)
722
{
723 724
	struct hd_struct *part;
	struct block_device *bdev = NULL;
725

726
	part = disk_get_part(disk, partno);
727
	if (part)
728 729
		bdev = bdget(part_devt(part));
	disk_put_part(part);
730

731
	return bdev;
732 733 734
}
EXPORT_SYMBOL(bdget_disk);

735 736 737 738 739 740 741
/*
 * print a full list of all partitions - intended for places where the root
 * filesystem can't be mounted and thus to give the victim some idea of what
 * went wrong
 */
void __init printk_all_partitions(void)
{
742 743 744 745 746 747
	struct class_dev_iter iter;
	struct device *dev;

	class_dev_iter_init(&iter, &block_class, NULL, &disk_type);
	while ((dev = class_dev_iter_next(&iter))) {
		struct gendisk *disk = dev_to_disk(dev);
748 749
		struct disk_part_iter piter;
		struct hd_struct *part;
750 751
		char name_buf[BDEVNAME_SIZE];
		char devt_buf[BDEVT_SIZE];
752 753 754

		/*
		 * Don't show empty devices or things that have been
Lucas De Marchi's avatar
Lucas De Marchi committed
755
		 * suppressed
756 757 758 759 760 761 762 763 764 765
		 */
		if (get_capacity(disk) == 0 ||
		    (disk->flags & GENHD_FL_SUPPRESS_PARTITION_INFO))
			continue;

		/*
		 * Note, unlike /proc/partitions, I am showing the
		 * numbers in hex - the same format as the root=
		 * option takes.
		 */
766 767 768
		disk_part_iter_init(&piter, disk, DISK_PITER_INCL_PART0);
		while ((part = disk_part_iter_next(&piter))) {
			bool is_part0 = part == &disk->part0;
769

770
			printk("%s%s %10llu %s %s", is_part0 ? "" : "  ",
771
			       bdevt_str(part_devt(part), devt_buf),
772 773
			       (unsigned long long)part_nr_sects_read(part) >> 1
			       , disk_name(disk, part->partno, name_buf),
774
			       part->info ? part->info->uuid : "");
775 776 777 778 779 780 781 782 783 784
			if (is_part0) {
				if (disk->driverfs_dev != NULL &&
				    disk->driverfs_dev->driver != NULL)
					printk(" driver: %s\n",
					      disk->driverfs_dev->driver->name);
				else
					printk(" (driver?)\n");
			} else
				printk("\n");
		}
785
		disk_part_iter_exit(&piter);
786 787
	}
	class_dev_iter_exit(&iter);
788 789
}

Linus Torvalds's avatar
Linus Torvalds committed
790 791
#ifdef CONFIG_PROC_FS
/* iterator */
792
static void *disk_seqf_start(struct seq_file *seqf, loff_t *pos)
793
{
794 795 796
	loff_t skip = *pos;
	struct class_dev_iter *iter;
	struct device *dev;
797

798
	iter = kmalloc(sizeof(*iter), GFP_KERNEL);
799 800 801 802 803 804 805 806 807 808 809 810
	if (!iter)
		return ERR_PTR(-ENOMEM);

	seqf->private = iter;
	class_dev_iter_init(iter, &block_class, NULL, &disk_type);
	do {
		dev = class_dev_iter_next(iter);
		if (!dev)
			return NULL;
	} while (skip--);

	return dev_to_disk(dev);
811 812
}

813
static void *disk_seqf_next(struct seq_file *seqf, void *v, loff_t *pos)
Linus Torvalds's avatar
Linus Torvalds committed
814
{
815
	struct device *dev;
Linus Torvalds's avatar
Linus Torvalds committed
816

817 818
	(*pos)++;
	dev = class_dev_iter_next(seqf->private);
819
	if (dev)
820
		return dev_to_disk(dev);
821

Linus Torvalds's avatar
Linus Torvalds committed
822 823 824
	return NULL;
}

825
static void disk_seqf_stop(struct seq_file *seqf, void *v)
826
{
827
	struct class_dev_iter *iter = seqf->private;
828

829 830 831 832
	/* stop is called even after start failed :-( */
	if (iter) {
		class_dev_iter_exit(iter);
		kfree(iter);
833
		seqf->private = NULL;
834
	}
Linus Torvalds's avatar
Linus Torvalds committed
835 836
}

837
static void *show_partition_start(struct seq_file *seqf, loff_t *pos)
Linus Torvalds's avatar
Linus Torvalds committed
838
{
839
	void *p;
840 841

	p = disk_seqf_start(seqf, pos);
842
	if (!IS_ERR_OR_NULL(p) && !*pos)
843 844
		seq_puts(seqf, "major minor  #blocks  name\n\n");
	return p;
Linus Torvalds's avatar
Linus Torvalds committed
845 846
}

847
static int show_partition(struct seq_file *seqf, void *v)
Linus Torvalds's avatar
Linus Torvalds committed
848 849
{
	struct gendisk *sgp = v;
850 851
	struct disk_part_iter piter;
	struct hd_struct *part;
Linus Torvalds's avatar
Linus Torvalds committed
852 853 854
	char buf[BDEVNAME_SIZE];

	/* Don't show non-partitionable removeable devices or empty devices */
Tejun Heo's avatar
Tejun Heo committed
855
	if (!get_capacity(sgp) || (!disk_max_parts(sgp) &&
856
				   (sgp->flags & GENHD_FL_REMOVABLE)))
Linus Torvalds's avatar
Linus Torvalds committed
857 858 859 860 861
		return 0;
	if (sgp->flags & GENHD_FL_SUPPRESS_PARTITION_INFO)
		return 0;

	/* show the full disk and all non-0 size partitions of it */
862
	disk_part_iter_init(&piter, sgp, DISK_PITER_INCL_PART0);
863
	while ((part = disk_part_iter_next(&piter)))
864
		seq_printf(seqf, "%4d  %7d %10llu %s\n",
865
			   MAJOR(part_devt(part)), MINOR(part_devt(part)),
866
			   (unsigned long long)part_nr_sects_read(part) >> 1,
867
			   disk_name(sgp, part->partno, buf));
868
	disk_part_iter_exit(&piter);
Linus Torvalds's avatar
Linus Torvalds committed
869 870 871 872

	return 0;
}

873
static const struct seq_operations partitions_op = {
874 875 876
	.start	= show_partition_start,
	.next	= disk_seqf_next,
	.stop	= disk_seqf_stop,
877
	.show	= show_partition
Linus Torvalds's avatar
Linus Torvalds committed
878
};
879 880 881 882 883 884 885 886 887 888 889 890

static int partitions_open(struct inode *inode, struct file *file)
{
	return seq_open(file, &partitions_op);
}

static const struct file_operations proc_partitions_operations = {
	.open		= partitions_open,
	.read		= seq_read,
	.llseek		= seq_lseek,
	.release	= seq_release,
};
Linus Torvalds's avatar
Linus Torvalds committed
891 892 893
#endif


894
static struct kobject *base_probe(dev_t devt, int *partno, void *data)
Linus Torvalds's avatar
Linus Torvalds committed
895
{
896
	if (request_module("block-major-%d-%d", MAJOR(devt), MINOR(devt)) > 0)
Linus Torvalds's avatar
Linus Torvalds committed
897
		/* Make old-style 2.4 aliases work */
898
		request_module("block-major-%d", MAJOR(devt));
Linus Torvalds's avatar
Linus Torvalds committed
899 900 901 902 903
	return NULL;
}

static int __init genhd_device_init(void)
{
904 905 906 907
	int error;

	block_class.dev_kobj = sysfs_dev_block_kobj;
	error = class_register(&block_class);
908 909
	if (unlikely(error))
		return error;
910
	bdev_map = kobj_map_init(base_probe, &block_class_lock);
Linus Torvalds's avatar
Linus Torvalds committed
911
	blk_dev_init();
912

913 914
	register_blkdev(BLOCK_EXT_MAJOR, "blkext");

915
	/* create top-level block dir */
916 917
	if (!sysfs_deprecated)
		block_depr = kobject_create_and_add("block", NULL);
918
	return 0;
Linus Torvalds's avatar
Linus Torvalds committed
919 920 921 922
}

subsys_initcall(genhd_device_init);

923 924
static ssize_t disk_range_show(struct device *dev,
			       struct device_attribute *attr, char *buf)
Linus Torvalds's avatar
Linus Torvalds committed
925
{
926
	struct gendisk *disk = dev_to_disk(dev);
Linus Torvalds's avatar
Linus Torvalds committed
927

928
	return sprintf(buf, "%d\n", disk->minors);
Linus Torvalds's avatar
Linus Torvalds committed
929 930
}

931 932 933 934 935
static ssize_t disk_ext_range_show(struct device *dev,
				   struct device_attribute *attr, char *buf)
{
	struct gendisk *disk = dev_to_disk(dev);

Tejun Heo's avatar
Tejun Heo committed
936
	return sprintf(buf, "%d\n", disk_max_parts(disk));
937 938
}

939 940
static ssize_t disk_removable_show(struct device *dev,
				   struct device_attribute *attr, char *buf)
941
{
942
	struct gendisk *disk = dev_to_disk(dev);
943

944 945
	return sprintf(buf, "%d\n",
		       (disk->flags & GENHD_FL_REMOVABLE ? 1 : 0));
946 947
}

Kay Sievers's avatar
Kay Sievers committed
948 949 950 951 952
static ssize_t disk_ro_show(struct device *dev,
				   struct device_attribute *attr, char *buf)
{
	struct gendisk *disk = dev_to_disk(dev);

953
	return sprintf(buf, "%d\n", get_disk_ro(disk) ? 1 : 0);
Kay Sievers's avatar
Kay Sievers committed
954 955
}

956 957
static ssize_t disk_capability_show(struct device *dev,
				    struct device_attribute *attr, char *buf)
958
{
959 960 961
	struct gendisk *disk = dev_to_disk(dev);

	return sprintf(buf, "%x\n", disk->flags);
962
}
963

964 965 966 967 968 969 970 971 972
static ssize_t disk_alignment_offset_show(struct device *dev,
					  struct device_attribute *attr,
					  char *buf)
{
	struct gendisk *disk = dev_to_disk(dev);

	return sprintf(buf, "%d\n", queue_alignment_offset(disk->queue));
}

973 974 975 976 977 978
static ssize_t disk_discard_alignment_show(struct device *dev,
					   struct device_attribute *attr,
					   char *buf)
{
	struct gendisk *disk = dev_to_disk(dev);

979
	return sprintf(buf, "%d\n", queue_discard_alignment(disk->queue));
980 981
}

982
static DEVICE_ATTR(range, S_IRUGO, disk_range_show, NULL);
983
static DEVICE_ATTR(ext_range, S_IRUGO, disk_ext_range_show, NULL);
984
static DEVICE_ATTR(removable, S_IRUGO, disk_removable_show, NULL);
Kay Sievers's avatar
Kay Sievers committed
985
static DEVICE_ATTR(ro, S_IRUGO, disk_ro_show, NULL);
986
static DEVICE_ATTR(size, S_IRUGO, part_size_show, NULL);
987
static DEVICE_ATTR(alignment_offset, S_IRUGO, disk_alignment_offset_show, NULL);
988 989
static DEVICE_ATTR(discard_alignment, S_IRUGO, disk_discard_alignment_show,
		   NULL);
990
static DEVICE_ATTR(capability, S_IRUGO, disk_capability_show, NULL);
991
static DEVICE_ATTR(stat, S_IRUGO, part_stat_show, NULL);
992
static DEVICE_ATTR(inflight, S_IRUGO, part_inflight_show, NULL);
993
#ifdef CONFIG_FAIL_MAKE_REQUEST
994
static struct device_attribute dev_attr_fail =
995
	__ATTR(make-it-fail, S_IRUGO|S_IWUSR, part_fail_show, part_fail_store);
996
#endif
997 998 999 1000 1001
#ifdef CONFIG_FAIL_IO_TIMEOUT
static struct device_attribute dev_attr_fail_timeout =
	__ATTR(io-timeout-fail,  S_IRUGO|S_IWUSR, part_timeout_show,
		part_timeout_store);
#endif
1002 1003 1004

static struct attribute *disk_attrs[] = {
	&dev_attr_range.attr,
1005
	&dev_attr_ext_range.attr,
1006
	&dev_attr_removable.attr,
Kay Sievers's avatar
Kay Sievers committed
1007
	&dev_attr_ro.attr,
1008
	&dev_attr_size.attr,
1009
	&dev_attr_alignment_offset.attr,
1010
	&dev_attr_discard_alignment.attr,
1011 1012
	&dev_attr_capability.attr,
	&dev_attr_stat.attr,
1013
	&dev_attr_inflight.attr,
1014 1015
#ifdef CONFIG_FAIL_MAKE_REQUEST
	&dev_attr_fail.attr,
1016 1017 1018
#endif
#ifdef CONFIG_FAIL_IO_TIMEOUT
	&dev_attr_fail_timeout.attr,
1019 1020 1021 1022 1023 1024 1025 1026
#endif
	NULL
};

static struct attribute_group disk_attr_group = {
	.attrs = disk_attrs,
};

1027
static const struct attribute_group *disk_attr_groups[] = {
1028 1029
	&disk_attr_group,
	NULL
Linus Torvalds's avatar
Linus Torvalds committed
1030 1031
};

1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048
/**
 * disk_replace_part_tbl - replace disk->part_tbl in RCU-safe way
 * @disk: disk to replace part_tbl for
 * @new_ptbl: new part_tbl to install
 *
 * Replace disk->part_tbl with @new_ptbl in RCU-safe way.  The
 * original ptbl is freed using RCU callback.
 *
 * LOCKING:
 * Matching bd_mutx locked.
 */
static void disk_replace_part_tbl(struct gendisk *disk,
				  struct disk_part_tbl *new_ptbl)
{
	struct disk_part_tbl *old_ptbl = disk->part_tbl;

	rcu_assign_pointer(disk->part_tbl, new_ptbl);
1049 1050 1051

	if (old_ptbl) {
		rcu_assign_pointer(old_ptbl->last_lookup, NULL);
1052
		kfree_rcu(old_ptbl, rcu_head);
1053
	}
1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074
}

/**
 * disk_expand_part_tbl - expand disk->part_tbl
 * @disk: disk to expand part_tbl for
 * @partno: expand such that this partno can fit in
 *
 * Expand disk->part_tbl such that @partno can fit in.  disk->part_tbl
 * uses RCU to allow unlocked dereferencing for stats and other stuff.
 *
 * LOCKING:
 * Matching bd_mutex locked, might sleep.
 *
 * RETURNS:
 * 0 on success, -errno on failure.
 */
int disk_expand_part_tbl(struct gendisk *disk, int partno)
{
	struct disk_part_tbl *old_ptbl = disk->part_tbl;
	struct disk_part_tbl *new_ptbl;
	int len = old_ptbl ? old_ptbl->len : 0;
1075
	int i, target;
1076
	size_t size;
1077 1078 1079 1080 1081 1082 1083 1084

	/*
	 * check for int overflow, since we can get here from blkpg_ioctl()
	 * with a user passed 'partno'.
	 */
	target = partno + 1;
	if (target < 0)
		return -EINVAL;
1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106

	/* disk_max_parts() is zero during initialization, ignore if so */
	if (disk_max_parts(disk) && target > disk_max_parts(disk))
		return -EINVAL;

	if (target <= len)
		return 0;

	size = sizeof(*new_ptbl) + target * sizeof(new_ptbl->part[0]);
	new_ptbl = kzalloc_node(size, GFP_KERNEL, disk->node_id);
	if (!new_ptbl)
		return -ENOMEM;

	new_ptbl->len = target;

	for (i = 0; i < len; i++)
		rcu_assign_pointer(new_ptbl->part[i], old_ptbl->part[i]);

	disk_replace_part_tbl(disk, new_ptbl);
	return 0;
}

1107
static void disk_release(struct device *dev)
Linus Torvalds's avatar
Linus Torvalds committed
1108
{
1109 1110
	struct gendisk *disk = dev_to_disk(dev);

1111
	blk_free_devt(dev->devt);
1112
	disk_release_events(disk);
Linus Torvalds's avatar
Linus Torvalds committed
1113
	kfree(disk->random);
1114
	disk_replace_part_tbl(disk, NULL);
1115
	hd_free_part(&disk->part0);
1116 1117
	if (disk->queue)
		blk_put_queue(disk->queue);
Linus Torvalds's avatar
Linus Torvalds committed
1118 1119
	kfree(disk);
}
1120 1121
struct class block_class = {
	.name		= "block",
Linus Torvalds's avatar
Linus Torvalds committed
1122 1123
};

1124
static char *block_devnode(struct device *dev, umode_t *mode,
1125
			   kuid_t *uid, kgid_t *gid)
1126 1127 1128
{
	struct gendisk *disk = dev_to_disk(dev);

1129 1130
	if (disk->devnode)
		return disk->devnode(disk, mode);
1131 1132 1133
	return NULL;
}

Adrian Bunk's avatar
Adrian Bunk committed
1134
static struct device_type disk_type = {
1135 1136 1137
	.name		= "disk",
	.groups		= disk_attr_groups,
	.release	= disk_release,
1138
	.devnode	= block_devnode,
Linus Torvalds's avatar
Linus Torvalds committed
1139 1140
};

1141
#ifdef CONFIG_PROC_FS
1142 1143 1144 1145 1146 1147 1148 1149
/*
 * aggregate disk stat collector.  Uses the same stats that the sysfs
 * entries do, above, but makes them available through one seq_file.
 *
 * The output looks suspiciously like /proc/partitions with a bunch of
 * extra fields.
 */
static int diskstats_show(struct seq_file *seqf, void *v)
Linus Torvalds's avatar
Linus Torvalds committed
1150 1151
{
	struct gendisk *gp = v;
1152 1153
	struct disk_part_iter piter;
	struct hd_struct *hd;
Linus Torvalds's avatar
Linus Torvalds committed
1154
	char buf[BDEVNAME_SIZE];
Tejun Heo's avatar
Tejun Heo committed
1155
	int cpu;
Linus Torvalds's avatar
Linus Torvalds committed
1156 1157

	/*
1158
	if (&disk_to_dev(gp)->kobj.entry == block_class.devices.next)
1159
		seq_puts(seqf,	"major minor name"
Linus Torvalds's avatar
Linus Torvalds committed
1160 1161 1162 1163
				"     rio rmerge rsect ruse wio wmerge "
				"wsect wuse running use aveq"
				"\n\n");
	*/
1164

1165
	disk_part_iter_init(&piter, gp, DISK_PITER_INCL_EMPTY_PART0);
1166
	while ((hd = disk_part_iter_next(&piter))) {
1167
		cpu = part_stat_lock();
Tejun Heo's avatar
Tejun Heo committed
1168
		part_round_stats(cpu, hd);
1169
		part_stat_unlock();
1170 1171
		seq_printf(seqf, "%4d %7d %s %lu %lu %lu "
			   "%u %lu %lu %lu %u %u %u %u\n",
1172 1173
			   MAJOR(part_devt(hd)), MINOR(part_devt(hd)),
			   disk_name(gp, hd->partno, buf),
1174 1175
			   part_stat_read(hd, ios[READ]),
			   part_stat_read(hd, merges[READ]),
1176
			   part_stat_read(hd, sectors[READ]),
1177 1178 1179
			   jiffies_to_msecs(part_stat_read(hd, ticks[READ])),
			   part_stat_read(hd, ios[WRITE]),
			   part_stat_read(hd, merges[WRITE]),
1180
			   part_stat_read(hd, sectors[WRITE]),
1181
			   jiffies_to_msecs(part_stat_read(hd, ticks[WRITE])),
1182
			   part_in_flight(hd),
1183 1184 1185
			   jiffies_to_msecs(part_stat_read(hd, io_ticks)),
			   jiffies_to_msecs(part_stat_read(hd, time_in_queue))
			);
Linus Torvalds's avatar
Linus Torvalds committed
1186
	}
1187
	disk_part_iter_exit(&piter);
1188

Linus Torvalds's avatar
Linus Torvalds committed
1189 1190 1191
	return 0;
}

1192
static const struct seq_operations diskstats_op = {
1193 1194 1195
	.start	= disk_seqf_start,
	.next	= disk_seqf_next,
	.stop	= disk_seqf_stop,
Linus Torvalds's avatar
Linus Torvalds committed
1196 1197
	.show	= diskstats_show
};
1198

1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210
static int diskstats_open(struct inode *inode, struct file *file)
{
	return seq_open(file, &diskstats_op);
}

static const struct file_operations proc_diskstats_operations = {
	.open		= diskstats_open,
	.read		= seq_read,
	.llseek		= seq_lseek,
	.release	= seq_release,
};

1211 1212
static int __init proc_genhd_init(void)
{
1213
	proc_create("diskstats", 0, NULL, &proc_diskstats_operations);
1214 1215 1216 1217
	proc_create("partitions", 0, NULL, &proc_partitions_operations);
	return 0;
}
module_init(proc_genhd_init);
1218
#endif /* CONFIG_PROC_FS */
Linus Torvalds's avatar
Linus Torvalds committed
1219

1220
dev_t blk_lookup_devt(const char *name, int partno)
1221
{
1222 1223 1224
	dev_t devt = MKDEV(0, 0);
	struct class_dev_iter iter;
	struct device *dev;
1225

1226 1227
	class_dev_iter_init(&iter, &block_class, NULL, &disk_type);
	while ((dev = class_dev_iter_next(&iter))) {
1228
		struct gendisk *disk = dev_to_disk(dev);
1229
		struct hd_struct *part;
1230

1231
		if (strcmp(dev_name(dev), name))
1232 1233
			continue;

1234 1235 1236 1237 1238 1239 1240 1241
		if (partno < disk->minors) {
			/* We need to return the right devno, even
			 * if the partition doesn't exist yet.
			 */
			devt = MKDEV(MAJOR(dev->devt),
				     MINOR(dev->devt) + partno);
			break;
		}
1242
		part = disk_get_part(disk, partno);
1243
		if (part) {
1244
			devt = part_devt(part);
1245
			disk_put_part(part);
1246
			break;
1247
		}
1248
		disk_put_part(part);
1249
	}