genhd.c 49.0 KB
Newer Older
L
Linus Torvalds 已提交
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>
L
Linus Torvalds 已提交
9 10
#include <linux/kernel.h>
#include <linux/blkdev.h>
11
#include <linux/backing-dev.h>
L
Linus Torvalds 已提交
12 13
#include <linux/init.h>
#include <linux/spinlock.h>
14
#include <linux/proc_fs.h>
L
Linus Torvalds 已提交
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>
23
#include <linux/badblocks.h>
L
Linus Torvalds 已提交
24

25 26
#include "blk.h"

27 28
static DEFINE_MUTEX(block_class_lock);
struct kobject *block_depr;
L
Linus Torvalds 已提交
29

30
/* for extended dynamic devt allocation, currently only one major is used */
31
#define NR_EXT_DEVT		(1 << MINORBITS)
32

33
/* For extended devt allocation.  ext_devt_lock prevents look up
34 35
 * results from going away underneath its user.
 */
36
static DEFINE_SPINLOCK(ext_devt_lock);
37 38
static DEFINE_IDR(ext_devt_idr);

39
static const struct device_type disk_type;
A
Adrian Bunk 已提交
40

41 42
static void disk_check_events(struct disk_events *ev,
			      unsigned int *clearing_ptr);
43
static void disk_alloc_events(struct gendisk *disk);
44 45 46 47
static void disk_add_events(struct gendisk *disk);
static void disk_del_events(struct gendisk *disk);
static void disk_release_events(struct gendisk *disk);

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
void part_inc_in_flight(struct request_queue *q, struct hd_struct *part, int rw)
{
	if (q->mq_ops)
		return;

	atomic_inc(&part->in_flight[rw]);
	if (part->partno)
		atomic_inc(&part_to_disk(part)->part0.in_flight[rw]);
}

void part_dec_in_flight(struct request_queue *q, struct hd_struct *part, int rw)
{
	if (q->mq_ops)
		return;

	atomic_dec(&part->in_flight[rw]);
	if (part->partno)
		atomic_dec(&part_to_disk(part)->part0.in_flight[rw]);
}

void part_in_flight(struct request_queue *q, struct hd_struct *part,
		    unsigned int inflight[2])
{
	if (q->mq_ops) {
		blk_mq_in_flight(q, part, inflight);
		return;
	}

	inflight[0] = atomic_read(&part->in_flight[0]) +
			atomic_read(&part->in_flight[1]);
	if (part->partno) {
		part = &part_to_disk(part)->part0;
		inflight[1] = atomic_read(&part->in_flight[0]) +
				atomic_read(&part->in_flight[1]);
	}
}

85 86 87 88 89 90 91 92 93
struct hd_struct *__disk_get_part(struct gendisk *disk, int partno)
{
	struct disk_part_tbl *ptbl = rcu_dereference(disk->part_tbl);

	if (unlikely(partno < 0 || partno >= ptbl->len))
		return NULL;
	return rcu_dereference(ptbl->part[partno]);
}

94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109
/**
 * 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)
{
110
	struct hd_struct *part;
111

112
	rcu_read_lock();
113 114 115
	part = __disk_get_part(disk, partno);
	if (part)
		get_device(part_to_dev(part));
116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135
	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)
{
136 137 138 139 140
	struct disk_part_tbl *ptbl;

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

141 142 143 144
	piter->disk = disk;
	piter->part = NULL;

	if (flags & DISK_PITER_REVERSE)
145
		piter->idx = ptbl->len - 1;
146
	else if (flags & (DISK_PITER_INCL_PART0 | DISK_PITER_INCL_EMPTY_PART0))
147
		piter->idx = 0;
T
Tejun Heo 已提交
148 149
	else
		piter->idx = 1;
150 151

	piter->flags = flags;
152 153

	rcu_read_unlock();
154 155 156 157 158 159 160 161 162 163 164 165 166 167
}
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)
{
168
	struct disk_part_tbl *ptbl;
169 170 171 172 173 174
	int inc, end;

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

175
	/* get part_tbl */
176
	rcu_read_lock();
177
	ptbl = rcu_dereference(piter->disk->part_tbl);
178 179 180 181

	/* determine iteration parameters */
	if (piter->flags & DISK_PITER_REVERSE) {
		inc = -1;
182 183
		if (piter->flags & (DISK_PITER_INCL_PART0 |
				    DISK_PITER_INCL_EMPTY_PART0))
T
Tejun Heo 已提交
184 185 186
			end = -1;
		else
			end = 0;
187 188
	} else {
		inc = 1;
189
		end = ptbl->len;
190 191 192 193 194 195
	}

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

196
		part = rcu_dereference(ptbl->part[piter->idx]);
197 198
		if (!part)
			continue;
199
		if (!part_nr_sects_read(part) &&
200 201 202
		    !(piter->flags & DISK_PITER_INCL_EMPTY) &&
		    !(piter->flags & DISK_PITER_INCL_EMPTY_PART0 &&
		      piter->idx == 0))
203 204
			continue;

205
		get_device(part_to_dev(part));
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
		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);

233 234 235
static inline int sector_in_part(struct hd_struct *part, sector_t sector)
{
	return part->start_sect <= sector &&
236
		sector < part->start_sect + part_nr_sects_read(part);
237 238
}

239 240 241 242 243 244 245 246 247 248 249 250 251
/**
 * 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:
252
 * Found partition on success, part0 is returned if no partition matches
253 254 255
 */
struct hd_struct *disk_map_sector_rcu(struct gendisk *disk, sector_t sector)
{
256
	struct disk_part_tbl *ptbl;
257
	struct hd_struct *part;
258 259
	int i;

260 261
	ptbl = rcu_dereference(disk->part_tbl);

262 263 264 265
	part = rcu_dereference(ptbl->last_lookup);
	if (part && sector_in_part(part, sector))
		return part;

266
	for (i = 1; i < ptbl->len; i++) {
267
		part = rcu_dereference(ptbl->part[i]);
268

269 270
		if (part && sector_in_part(part, sector)) {
			rcu_assign_pointer(ptbl->last_lookup, part);
271
			return part;
272
		}
273
	}
274
	return &disk->part0;
275 276 277
}
EXPORT_SYMBOL_GPL(disk_map_sector_rcu);

L
Linus Torvalds 已提交
278 279 280 281
/*
 * Can be deleted altogether. Later.
 *
 */
282
#define BLKDEV_MAJOR_HASH_SIZE 255
L
Linus Torvalds 已提交
283 284 285 286
static struct blk_major_name {
	struct blk_major_name *next;
	int major;
	char name[16];
287
} *major_names[BLKDEV_MAJOR_HASH_SIZE];
L
Linus Torvalds 已提交
288 289

/* index in the above - for now: assume no multimajor ranges */
290
static inline int major_to_index(unsigned major)
L
Linus Torvalds 已提交
291
{
292
	return major % BLKDEV_MAJOR_HASH_SIZE;
293 294
}

295
#ifdef CONFIG_PROC_FS
296
void blkdev_show(struct seq_file *seqf, off_t offset)
297
{
298
	struct blk_major_name *dp;
299

300 301 302
	mutex_lock(&block_class_lock);
	for (dp = major_names[major_to_index(offset)]; dp; dp = dp->next)
		if (dp->major == offset)
303
			seq_printf(seqf, "%3d %s\n", dp->major, dp->name);
304
	mutex_unlock(&block_class_lock);
L
Linus Torvalds 已提交
305
}
306
#endif /* CONFIG_PROC_FS */
L
Linus Torvalds 已提交
307

308 309 310
/**
 * register_blkdev - register a new block device
 *
311
 * @major: the requested major device number [1..255]. If @major = 0, try to
312 313 314 315 316
 *         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.
 *
317 318
 * The return value depends on the @major input parameter:
 *
319 320
 *  - if a major device number was requested in range [1..255] then the
 *    function returns zero on success, or a negative error code
321
 *  - if any unused major number was requested with @major = 0 parameter
322 323 324
 *    then the return value is the allocated major number in range
 *    [1..255] or a negative error code otherwise
 */
L
Linus Torvalds 已提交
325 326 327 328 329
int register_blkdev(unsigned int major, const char *name)
{
	struct blk_major_name **n, *p;
	int index, ret = 0;

330
	mutex_lock(&block_class_lock);
L
Linus Torvalds 已提交
331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348

	/* 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;
	}

349 350 351 352 353 354 355 356
	if (major >= BLKDEV_MAJOR_MAX) {
		pr_err("register_blkdev: major requested (%d) is greater than the maximum (%d) for %s\n",
		       major, BLKDEV_MAJOR_MAX, name);

		ret = -EINVAL;
		goto out;
	}

L
Linus Torvalds 已提交
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
	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:
383
	mutex_unlock(&block_class_lock);
L
Linus Torvalds 已提交
384 385 386 387 388
	return ret;
}

EXPORT_SYMBOL(register_blkdev);

389
void unregister_blkdev(unsigned int major, const char *name)
L
Linus Torvalds 已提交
390 391 392 393 394
{
	struct blk_major_name **n;
	struct blk_major_name *p = NULL;
	int index = major_to_index(major);

395
	mutex_lock(&block_class_lock);
L
Linus Torvalds 已提交
396 397 398
	for (n = &major_names[index]; *n; n = &(*n)->next)
		if ((*n)->major == major)
			break;
399 400 401
	if (!*n || strcmp((*n)->name, name)) {
		WARN_ON(1);
	} else {
L
Linus Torvalds 已提交
402 403 404
		p = *n;
		*n = p->next;
	}
405
	mutex_unlock(&block_class_lock);
L
Linus Torvalds 已提交
406 407 408 409 410 411 412
	kfree(p);
}

EXPORT_SYMBOL(unregister_blkdev);

static struct kobj_map *bdev_map;

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
/**
 * 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;
}

445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461
/**
 * 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);
T
Tejun Heo 已提交
462
	int idx;
463 464 465 466 467 468 469 470

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

	/* allocate ext devt */
471 472
	idr_preload(GFP_KERNEL);

473
	spin_lock_bh(&ext_devt_lock);
474
	idx = idr_alloc(&ext_devt_idr, part, 0, NR_EXT_DEVT, GFP_NOWAIT);
475
	spin_unlock_bh(&ext_devt_lock);
476 477

	idr_preload_end();
T
Tejun Heo 已提交
478 479
	if (idx < 0)
		return idx == -ENOSPC ? -EBUSY : idx;
480

481
	*devt = MKDEV(BLOCK_EXT_MAJOR, blk_mangle_minor(idx));
482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499
	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) {
500
		spin_lock_bh(&ext_devt_lock);
501
		idr_remove(&ext_devt_idr, blk_mangle_minor(MINOR(devt)));
502
		spin_unlock_bh(&ext_devt_lock);
503 504 505
	}
}

506 507 508 509 510 511 512 513 514 515 516 517
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;
}

L
Linus Torvalds 已提交
518 519 520 521 522
/*
 * Register device numbers dev..(dev+range-1)
 * range must be nonzero
 * The hash chain is sorted on range, so that subranges can override.
 */
523
void blk_register_region(dev_t devt, unsigned long range, struct module *module,
L
Linus Torvalds 已提交
524 525 526
			 struct kobject *(*probe)(dev_t, int *, void *),
			 int (*lock)(dev_t, void *), void *data)
{
527
	kobj_map(bdev_map, devt, range, module, probe, lock, data);
L
Linus Torvalds 已提交
528 529 530 531
}

EXPORT_SYMBOL(blk_register_region);

532
void blk_unregister_region(dev_t devt, unsigned long range)
L
Linus Torvalds 已提交
533
{
534
	kobj_unmap(bdev_map, devt, range);
L
Linus Torvalds 已提交
535 536 537 538
}

EXPORT_SYMBOL(blk_unregister_region);

539
static struct kobject *exact_match(dev_t devt, int *partno, void *data)
L
Linus Torvalds 已提交
540 541
{
	struct gendisk *p = data;
542

543
	return &disk_to_dev(p)->kobj;
L
Linus Torvalds 已提交
544 545
}

546
static int exact_lock(dev_t devt, void *data)
L
Linus Torvalds 已提交
547 548 549
{
	struct gendisk *p = data;

550
	if (!get_disk_and_module(p))
L
Linus Torvalds 已提交
551 552 553 554
		return -1;
	return 0;
}

555
static void register_disk(struct device *parent, struct gendisk *disk)
556 557 558 559 560 561 562
{
	struct device *ddev = disk_to_dev(disk);
	struct block_device *bdev;
	struct disk_part_iter piter;
	struct hd_struct *part;
	int err;

563
	ddev->parent = parent;
564

565
	dev_set_name(ddev, "%s", disk->disk_name);
566 567 568 569 570 571 572 573 574 575 576 577 578 579

	/* 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;
		}
	}
580 581 582 583 584 585 586 587

	/*
	 * 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);

588 589 590
	disk->part0.holder_dir = kobject_create_and_add("holders", &ddev->kobj);
	disk->slave_dir = kobject_create_and_add("slaves", &ddev->kobj);

591 592 593 594 595
	if (disk->flags & GENHD_FL_HIDDEN) {
		dev_set_uevent_suppress(ddev, 0);
		return;
	}

596
	/* No minors to use for partitions */
597
	if (!disk_part_scan_enabled(disk))
598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623
		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);
624 625 626 627 628

	err = sysfs_create_link(&ddev->kobj,
				&disk->queue->backing_dev_info->dev->kobj,
				"bdi");
	WARN_ON(err);
629 630
}

L
Linus Torvalds 已提交
631
/**
632
 * __device_add_disk - add disk information to kernel list
633
 * @parent: parent device for the disk
L
Linus Torvalds 已提交
634
 * @disk: per-device partitioning information
635
 * @register_queue: register the queue if set to true
L
Linus Torvalds 已提交
636 637 638
 *
 * This function registers the partitioning information in @disk
 * with the kernel.
639 640
 *
 * FIXME: error handling
L
Linus Torvalds 已提交
641
 */
642 643
static void __device_add_disk(struct device *parent, struct gendisk *disk,
			      bool register_queue)
L
Linus Torvalds 已提交
644
{
645
	dev_t devt;
646
	int retval;
647

648 649 650 651 652
	/* 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));
653 654
	WARN_ON(!disk->minors &&
		!(disk->flags & (GENHD_FL_EXT_DEVT | GENHD_FL_HIDDEN)));
655

L
Linus Torvalds 已提交
656
	disk->flags |= GENHD_FL_UP;
657 658 659 660 661 662 663 664 665

	retval = blk_alloc_devt(&disk->part0, &devt);
	if (retval) {
		WARN_ON(1);
		return;
	}
	disk->major = MAJOR(devt);
	disk->first_minor = MINOR(devt);

666 667
	disk_alloc_events(disk);

668 669 670 671 672 673 674 675
	if (disk->flags & GENHD_FL_HIDDEN) {
		/*
		 * Don't let hidden disks show up in /proc/partitions,
		 * and don't bother scanning for partitions either.
		 */
		disk->flags |= GENHD_FL_SUPPRESS_PARTITION_INFO;
		disk->flags |= GENHD_FL_NO_PART_SCAN;
	} else {
676 677
		int ret;

678 679
		/* Register BDI before referencing it from bdev */
		disk_to_dev(disk)->devt = devt;
680 681 682
		ret = bdi_register_owner(disk->queue->backing_dev_info,
						disk_to_dev(disk));
		WARN_ON(ret);
683 684 685
		blk_register_region(disk_devt(disk), disk->minors, NULL,
				    exact_match, exact_lock, disk);
	}
686
	register_disk(parent, disk);
687 688
	if (register_queue)
		blk_register_queue(disk);
689

690 691 692 693
	/*
	 * Take an extra ref on queue which will be put on disk_release()
	 * so that it sticks around as long as @disk is there.
	 */
694
	WARN_ON_ONCE(!blk_get_queue(disk->queue));
695

696
	disk_add_events(disk);
697
	blk_integrity_add(disk);
L
Linus Torvalds 已提交
698
}
699 700 701 702 703

void device_add_disk(struct device *parent, struct gendisk *disk)
{
	__device_add_disk(parent, disk, true);
}
704
EXPORT_SYMBOL(device_add_disk);
L
Linus Torvalds 已提交
705

706 707 708 709 710 711
void device_add_disk_no_queue_reg(struct device *parent, struct gendisk *disk)
{
	__device_add_disk(parent, disk, false);
}
EXPORT_SYMBOL(device_add_disk_no_queue_reg);

712
void del_gendisk(struct gendisk *disk)
L
Linus Torvalds 已提交
713
{
714 715 716
	struct disk_part_iter piter;
	struct hd_struct *part;

717
	blk_integrity_del(disk);
718 719
	disk_del_events(disk);

720 721 722 723 724
	/*
	 * Block lookups of the disk until all bdevs are unhashed and the
	 * disk is marked as dead (GENHD_FL_UP cleared).
	 */
	down_write(&disk->lookup_sem);
725 726 727 728 729
	/* 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);
730
		bdev_unhash_inode(part_devt(part));
731 732 733 734 735
		delete_partition(disk, part->partno);
	}
	disk_part_iter_exit(&piter);

	invalidate_partition(disk, 0);
736
	bdev_unhash_inode(disk_devt(disk));
737 738
	set_capacity(disk, 0);
	disk->flags &= ~GENHD_FL_UP;
739
	up_write(&disk->lookup_sem);
740

741 742
	if (!(disk->flags & GENHD_FL_HIDDEN))
		sysfs_remove_link(&disk_to_dev(disk)->kobj, "bdi");
743 744 745 746 747
	if (disk->queue) {
		/*
		 * Unregister bdi before releasing device numbers (as they can
		 * get reused and we'd get clashes in sysfs).
		 */
748 749
		if (!(disk->flags & GENHD_FL_HIDDEN))
			bdi_unregister(disk->queue->backing_dev_info);
750 751 752 753
		blk_unregister_queue(disk);
	} else {
		WARN_ON(1);
	}
754

755
	if (!(disk->flags & GENHD_FL_HIDDEN))
756
		blk_unregister_region(disk_devt(disk), disk->minors);
757 758 759 760 761 762 763 764

	kobject_put(disk->part0.holder_dir);
	kobject_put(disk->slave_dir);

	part_stat_set_all(&disk->part0, 0);
	disk->part0.stamp = 0;
	if (!sysfs_deprecated)
		sysfs_remove_link(block_depr, dev_name(disk_to_dev(disk)));
765
	pm_runtime_set_memalloc_noio(disk_to_dev(disk), false);
766
	device_del(disk_to_dev(disk));
L
Linus Torvalds 已提交
767
}
768
EXPORT_SYMBOL(del_gendisk);
L
Linus Torvalds 已提交
769

770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794
/* sysfs access to bad-blocks list. */
static ssize_t disk_badblocks_show(struct device *dev,
					struct device_attribute *attr,
					char *page)
{
	struct gendisk *disk = dev_to_disk(dev);

	if (!disk->bb)
		return sprintf(page, "\n");

	return badblocks_show(disk->bb, page, 0);
}

static ssize_t disk_badblocks_store(struct device *dev,
					struct device_attribute *attr,
					const char *page, size_t len)
{
	struct gendisk *disk = dev_to_disk(dev);

	if (!disk->bb)
		return -ENXIO;

	return badblocks_store(disk->bb, page, len, 0);
}

L
Linus Torvalds 已提交
795 796
/**
 * get_gendisk - get partitioning information for a given device
797
 * @devt: device to get partitioning information for
798
 * @partno: returned partition index
L
Linus Torvalds 已提交
799 800
 *
 * This function gets the structure containing partitioning
801
 * information for the given device @devt.
L
Linus Torvalds 已提交
802
 */
803
struct gendisk *get_gendisk(dev_t devt, int *partno)
L
Linus Torvalds 已提交
804
{
805 806 807 808 809 810 811 812 813 814 815
	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;

816
		spin_lock_bh(&ext_devt_lock);
817
		part = idr_find(&ext_devt_idr, blk_mangle_minor(MINOR(devt)));
818
		if (part && get_disk_and_module(part_to_disk(part))) {
819 820 821
			*partno = part->partno;
			disk = part_to_disk(part);
		}
822
		spin_unlock_bh(&ext_devt_lock);
823
	}
824

825 826 827 828 829 830 831 832 833 834 835
	if (!disk)
		return NULL;

	/*
	 * Synchronize with del_gendisk() to not return disk that is being
	 * destroyed.
	 */
	down_read(&disk->lookup_sem);
	if (unlikely((disk->flags & GENHD_FL_HIDDEN) ||
		     !(disk->flags & GENHD_FL_UP))) {
		up_read(&disk->lookup_sem);
836
		put_disk_and_module(disk);
837
		disk = NULL;
838 839
	} else {
		up_read(&disk->lookup_sem);
840
	}
841
	return disk;
L
Linus Torvalds 已提交
842
}
843
EXPORT_SYMBOL(get_gendisk);
L
Linus Torvalds 已提交
844

845 846 847 848 849 850 851 852 853 854 855 856 857
/**
 * 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.
 */
858
struct block_device *bdget_disk(struct gendisk *disk, int partno)
859
{
860 861
	struct hd_struct *part;
	struct block_device *bdev = NULL;
862

863
	part = disk_get_part(disk, partno);
864
	if (part)
865 866
		bdev = bdget(part_devt(part));
	disk_put_part(part);
867

868
	return bdev;
869 870 871
}
EXPORT_SYMBOL(bdget_disk);

872 873 874 875 876 877 878
/*
 * 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)
{
879 880 881 882 883 884
	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);
885 886
		struct disk_part_iter piter;
		struct hd_struct *part;
887 888
		char name_buf[BDEVNAME_SIZE];
		char devt_buf[BDEVT_SIZE];
889 890 891

		/*
		 * Don't show empty devices or things that have been
L
Lucas De Marchi 已提交
892
		 * suppressed
893 894 895 896 897 898 899 900 901 902
		 */
		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.
		 */
903 904 905
		disk_part_iter_init(&piter, disk, DISK_PITER_INCL_PART0);
		while ((part = disk_part_iter_next(&piter))) {
			bool is_part0 = part == &disk->part0;
906

907
			printk("%s%s %10llu %s %s", is_part0 ? "" : "  ",
908
			       bdevt_str(part_devt(part), devt_buf),
909 910
			       (unsigned long long)part_nr_sects_read(part) >> 1
			       , disk_name(disk, part->partno, name_buf),
911
			       part->info ? part->info->uuid : "");
912
			if (is_part0) {
913
				if (dev->parent && dev->parent->driver)
914
					printk(" driver: %s\n",
915
					      dev->parent->driver->name);
916 917 918 919 920
				else
					printk(" (driver?)\n");
			} else
				printk("\n");
		}
921
		disk_part_iter_exit(&piter);
922 923
	}
	class_dev_iter_exit(&iter);
924 925
}

L
Linus Torvalds 已提交
926 927
#ifdef CONFIG_PROC_FS
/* iterator */
928
static void *disk_seqf_start(struct seq_file *seqf, loff_t *pos)
929
{
930 931 932
	loff_t skip = *pos;
	struct class_dev_iter *iter;
	struct device *dev;
933

934
	iter = kmalloc(sizeof(*iter), GFP_KERNEL);
935 936 937 938 939 940 941 942 943 944 945 946
	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);
947 948
}

949
static void *disk_seqf_next(struct seq_file *seqf, void *v, loff_t *pos)
L
Linus Torvalds 已提交
950
{
951
	struct device *dev;
L
Linus Torvalds 已提交
952

953 954
	(*pos)++;
	dev = class_dev_iter_next(seqf->private);
955
	if (dev)
956
		return dev_to_disk(dev);
957

L
Linus Torvalds 已提交
958 959 960
	return NULL;
}

961
static void disk_seqf_stop(struct seq_file *seqf, void *v)
962
{
963
	struct class_dev_iter *iter = seqf->private;
964

965 966 967 968
	/* stop is called even after start failed :-( */
	if (iter) {
		class_dev_iter_exit(iter);
		kfree(iter);
969
		seqf->private = NULL;
970
	}
L
Linus Torvalds 已提交
971 972
}

973
static void *show_partition_start(struct seq_file *seqf, loff_t *pos)
L
Linus Torvalds 已提交
974
{
975
	void *p;
976 977

	p = disk_seqf_start(seqf, pos);
978
	if (!IS_ERR_OR_NULL(p) && !*pos)
979 980
		seq_puts(seqf, "major minor  #blocks  name\n\n");
	return p;
L
Linus Torvalds 已提交
981 982
}

983
static int show_partition(struct seq_file *seqf, void *v)
L
Linus Torvalds 已提交
984 985
{
	struct gendisk *sgp = v;
986 987
	struct disk_part_iter piter;
	struct hd_struct *part;
L
Linus Torvalds 已提交
988 989 990
	char buf[BDEVNAME_SIZE];

	/* Don't show non-partitionable removeable devices or empty devices */
991
	if (!get_capacity(sgp) || (!disk_max_parts(sgp) &&
992
				   (sgp->flags & GENHD_FL_REMOVABLE)))
L
Linus Torvalds 已提交
993 994 995 996 997
		return 0;
	if (sgp->flags & GENHD_FL_SUPPRESS_PARTITION_INFO)
		return 0;

	/* show the full disk and all non-0 size partitions of it */
998
	disk_part_iter_init(&piter, sgp, DISK_PITER_INCL_PART0);
999
	while ((part = disk_part_iter_next(&piter)))
1000
		seq_printf(seqf, "%4d  %7d %10llu %s\n",
1001
			   MAJOR(part_devt(part)), MINOR(part_devt(part)),
1002
			   (unsigned long long)part_nr_sects_read(part) >> 1,
1003
			   disk_name(sgp, part->partno, buf));
1004
	disk_part_iter_exit(&piter);
L
Linus Torvalds 已提交
1005 1006 1007 1008

	return 0;
}

1009
static const struct seq_operations partitions_op = {
1010 1011 1012
	.start	= show_partition_start,
	.next	= disk_seqf_next,
	.stop	= disk_seqf_stop,
1013
	.show	= show_partition
L
Linus Torvalds 已提交
1014
};
1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026

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,
};
L
Linus Torvalds 已提交
1027 1028 1029
#endif


1030
static struct kobject *base_probe(dev_t devt, int *partno, void *data)
L
Linus Torvalds 已提交
1031
{
1032
	if (request_module("block-major-%d-%d", MAJOR(devt), MINOR(devt)) > 0)
L
Linus Torvalds 已提交
1033
		/* Make old-style 2.4 aliases work */
1034
		request_module("block-major-%d", MAJOR(devt));
L
Linus Torvalds 已提交
1035 1036 1037 1038 1039
	return NULL;
}

static int __init genhd_device_init(void)
{
1040 1041 1042 1043
	int error;

	block_class.dev_kobj = sysfs_dev_block_kobj;
	error = class_register(&block_class);
1044 1045
	if (unlikely(error))
		return error;
1046
	bdev_map = kobj_map_init(base_probe, &block_class_lock);
L
Linus Torvalds 已提交
1047
	blk_dev_init();
1048

1049 1050
	register_blkdev(BLOCK_EXT_MAJOR, "blkext");

1051
	/* create top-level block dir */
1052 1053
	if (!sysfs_deprecated)
		block_depr = kobject_create_and_add("block", NULL);
1054
	return 0;
L
Linus Torvalds 已提交
1055 1056 1057 1058
}

subsys_initcall(genhd_device_init);

1059 1060
static ssize_t disk_range_show(struct device *dev,
			       struct device_attribute *attr, char *buf)
L
Linus Torvalds 已提交
1061
{
1062
	struct gendisk *disk = dev_to_disk(dev);
L
Linus Torvalds 已提交
1063

1064
	return sprintf(buf, "%d\n", disk->minors);
L
Linus Torvalds 已提交
1065 1066
}

1067 1068 1069 1070 1071
static ssize_t disk_ext_range_show(struct device *dev,
				   struct device_attribute *attr, char *buf)
{
	struct gendisk *disk = dev_to_disk(dev);

T
Tejun Heo 已提交
1072
	return sprintf(buf, "%d\n", disk_max_parts(disk));
1073 1074
}

1075 1076
static ssize_t disk_removable_show(struct device *dev,
				   struct device_attribute *attr, char *buf)
1077
{
1078
	struct gendisk *disk = dev_to_disk(dev);
1079

1080 1081
	return sprintf(buf, "%d\n",
		       (disk->flags & GENHD_FL_REMOVABLE ? 1 : 0));
1082 1083
}

1084 1085 1086 1087 1088 1089 1090 1091 1092
static ssize_t disk_hidden_show(struct device *dev,
				   struct device_attribute *attr, char *buf)
{
	struct gendisk *disk = dev_to_disk(dev);

	return sprintf(buf, "%d\n",
		       (disk->flags & GENHD_FL_HIDDEN ? 1 : 0));
}

1093 1094 1095 1096 1097
static ssize_t disk_ro_show(struct device *dev,
				   struct device_attribute *attr, char *buf)
{
	struct gendisk *disk = dev_to_disk(dev);

1098
	return sprintf(buf, "%d\n", get_disk_ro(disk) ? 1 : 0);
1099 1100
}

1101 1102
static ssize_t disk_capability_show(struct device *dev,
				    struct device_attribute *attr, char *buf)
1103
{
1104 1105 1106
	struct gendisk *disk = dev_to_disk(dev);

	return sprintf(buf, "%x\n", disk->flags);
1107
}
1108

1109 1110 1111 1112 1113 1114 1115 1116 1117
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));
}

1118 1119 1120 1121 1122 1123
static ssize_t disk_discard_alignment_show(struct device *dev,
					   struct device_attribute *attr,
					   char *buf)
{
	struct gendisk *disk = dev_to_disk(dev);

1124
	return sprintf(buf, "%d\n", queue_discard_alignment(disk->queue));
1125 1126
}

1127
static DEVICE_ATTR(range, S_IRUGO, disk_range_show, NULL);
1128
static DEVICE_ATTR(ext_range, S_IRUGO, disk_ext_range_show, NULL);
1129
static DEVICE_ATTR(removable, S_IRUGO, disk_removable_show, NULL);
1130
static DEVICE_ATTR(hidden, S_IRUGO, disk_hidden_show, NULL);
1131
static DEVICE_ATTR(ro, S_IRUGO, disk_ro_show, NULL);
1132
static DEVICE_ATTR(size, S_IRUGO, part_size_show, NULL);
1133
static DEVICE_ATTR(alignment_offset, S_IRUGO, disk_alignment_offset_show, NULL);
1134 1135
static DEVICE_ATTR(discard_alignment, S_IRUGO, disk_discard_alignment_show,
		   NULL);
1136
static DEVICE_ATTR(capability, S_IRUGO, disk_capability_show, NULL);
1137
static DEVICE_ATTR(stat, S_IRUGO, part_stat_show, NULL);
1138
static DEVICE_ATTR(inflight, S_IRUGO, part_inflight_show, NULL);
1139 1140
static DEVICE_ATTR(badblocks, S_IRUGO | S_IWUSR, disk_badblocks_show,
		disk_badblocks_store);
1141
#ifdef CONFIG_FAIL_MAKE_REQUEST
1142
static struct device_attribute dev_attr_fail =
1143
	__ATTR(make-it-fail, S_IRUGO|S_IWUSR, part_fail_show, part_fail_store);
1144
#endif
1145 1146 1147 1148 1149
#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
1150 1151 1152

static struct attribute *disk_attrs[] = {
	&dev_attr_range.attr,
1153
	&dev_attr_ext_range.attr,
1154
	&dev_attr_removable.attr,
1155
	&dev_attr_hidden.attr,
1156
	&dev_attr_ro.attr,
1157
	&dev_attr_size.attr,
1158
	&dev_attr_alignment_offset.attr,
1159
	&dev_attr_discard_alignment.attr,
1160 1161
	&dev_attr_capability.attr,
	&dev_attr_stat.attr,
1162
	&dev_attr_inflight.attr,
1163
	&dev_attr_badblocks.attr,
1164 1165
#ifdef CONFIG_FAIL_MAKE_REQUEST
	&dev_attr_fail.attr,
1166 1167 1168
#endif
#ifdef CONFIG_FAIL_IO_TIMEOUT
	&dev_attr_fail_timeout.attr,
1169 1170 1171 1172
#endif
	NULL
};

1173 1174 1175 1176 1177 1178 1179 1180 1181 1182
static umode_t disk_visible(struct kobject *kobj, struct attribute *a, int n)
{
	struct device *dev = container_of(kobj, typeof(*dev), kobj);
	struct gendisk *disk = dev_to_disk(dev);

	if (a == &dev_attr_badblocks.attr && !disk->bb)
		return 0;
	return a->mode;
}

1183 1184
static struct attribute_group disk_attr_group = {
	.attrs = disk_attrs,
1185
	.is_visible = disk_visible,
1186 1187
};

1188
static const struct attribute_group *disk_attr_groups[] = {
1189 1190
	&disk_attr_group,
	NULL
L
Linus Torvalds 已提交
1191 1192
};

1193 1194 1195 1196 1197 1198 1199 1200 1201
/**
 * 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:
1202
 * Matching bd_mutex locked or the caller is the only user of @disk.
1203 1204 1205 1206
 */
static void disk_replace_part_tbl(struct gendisk *disk,
				  struct disk_part_tbl *new_ptbl)
{
1207 1208
	struct disk_part_tbl *old_ptbl =
		rcu_dereference_protected(disk->part_tbl, 1);
1209 1210

	rcu_assign_pointer(disk->part_tbl, new_ptbl);
1211 1212 1213

	if (old_ptbl) {
		rcu_assign_pointer(old_ptbl->last_lookup, NULL);
1214
		kfree_rcu(old_ptbl, rcu_head);
1215
	}
1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226
}

/**
 * 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:
1227 1228
 * Matching bd_mutex locked or the caller is the only user of @disk.
 * Might sleep.
1229 1230 1231 1232 1233 1234
 *
 * RETURNS:
 * 0 on success, -errno on failure.
 */
int disk_expand_part_tbl(struct gendisk *disk, int partno)
{
1235 1236
	struct disk_part_tbl *old_ptbl =
		rcu_dereference_protected(disk->part_tbl, 1);
1237 1238
	struct disk_part_tbl *new_ptbl;
	int len = old_ptbl ? old_ptbl->len : 0;
1239
	int i, target;
1240
	size_t size;
1241 1242 1243 1244 1245 1246 1247 1248

	/*
	 * 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;
1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270

	/* 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;
}

1271
static void disk_release(struct device *dev)
L
Linus Torvalds 已提交
1272
{
1273 1274
	struct gendisk *disk = dev_to_disk(dev);

1275
	blk_free_devt(dev->devt);
1276
	disk_release_events(disk);
L
Linus Torvalds 已提交
1277
	kfree(disk->random);
1278
	disk_replace_part_tbl(disk, NULL);
1279
	hd_free_part(&disk->part0);
1280 1281
	if (disk->queue)
		blk_put_queue(disk->queue);
L
Linus Torvalds 已提交
1282 1283
	kfree(disk);
}
1284 1285
struct class block_class = {
	.name		= "block",
L
Linus Torvalds 已提交
1286 1287
};

1288
static char *block_devnode(struct device *dev, umode_t *mode,
1289
			   kuid_t *uid, kgid_t *gid)
1290 1291 1292
{
	struct gendisk *disk = dev_to_disk(dev);

1293 1294
	if (disk->devnode)
		return disk->devnode(disk, mode);
1295 1296 1297
	return NULL;
}

1298
static const struct device_type disk_type = {
1299 1300 1301
	.name		= "disk",
	.groups		= disk_attr_groups,
	.release	= disk_release,
1302
	.devnode	= block_devnode,
L
Linus Torvalds 已提交
1303 1304
};

1305
#ifdef CONFIG_PROC_FS
1306 1307 1308 1309 1310 1311 1312 1313
/*
 * 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)
L
Linus Torvalds 已提交
1314 1315
{
	struct gendisk *gp = v;
1316 1317
	struct disk_part_iter piter;
	struct hd_struct *hd;
L
Linus Torvalds 已提交
1318
	char buf[BDEVNAME_SIZE];
1319
	unsigned int inflight[2];
T
Tejun Heo 已提交
1320
	int cpu;
L
Linus Torvalds 已提交
1321 1322

	/*
1323
	if (&disk_to_dev(gp)->kobj.entry == block_class.devices.next)
1324
		seq_puts(seqf,	"major minor name"
L
Linus Torvalds 已提交
1325 1326 1327 1328
				"     rio rmerge rsect ruse wio wmerge "
				"wsect wuse running use aveq"
				"\n\n");
	*/
1329

1330
	disk_part_iter_init(&piter, gp, DISK_PITER_INCL_EMPTY_PART0);
1331
	while ((hd = disk_part_iter_next(&piter))) {
1332
		cpu = part_stat_lock();
1333
		part_round_stats(gp->queue, cpu, hd);
1334
		part_stat_unlock();
1335
		part_in_flight(gp->queue, hd, inflight);
1336 1337
		seq_printf(seqf, "%4d %7d %s %lu %lu %lu "
			   "%u %lu %lu %lu %u %u %u %u\n",
1338 1339
			   MAJOR(part_devt(hd)), MINOR(part_devt(hd)),
			   disk_name(gp, hd->partno, buf),
1340 1341
			   part_stat_read(hd, ios[READ]),
			   part_stat_read(hd, merges[READ]),
1342
			   part_stat_read(hd, sectors[READ]),
1343 1344 1345
			   jiffies_to_msecs(part_stat_read(hd, ticks[READ])),
			   part_stat_read(hd, ios[WRITE]),
			   part_stat_read(hd, merges[WRITE]),
1346
			   part_stat_read(hd, sectors[WRITE]),
1347
			   jiffies_to_msecs(part_stat_read(hd, ticks[WRITE])),
1348
			   inflight[0],
1349 1350 1351
			   jiffies_to_msecs(part_stat_read(hd, io_ticks)),
			   jiffies_to_msecs(part_stat_read(hd, time_in_queue))
			);
L
Linus Torvalds 已提交
1352
	}
1353
	disk_part_iter_exit(&piter);
1354

L
Linus Torvalds 已提交
1355 1356 1357
	return 0;
}

1358
static const struct seq_operations diskstats_op = {
1359 1360 1361
	.start	= disk_seqf_start,
	.next	= disk_seqf_next,
	.stop	= disk_seqf_stop,
L
Linus Torvalds 已提交
1362 1363
	.show	= diskstats_show
};
1364

1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376
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,
};

1377 1378
static int __init proc_genhd_init(void)
{
1379
	proc_create("diskstats", 0, NULL, &proc_diskstats_operations);
1380 1381 1382 1383
	proc_create("partitions", 0, NULL, &proc_partitions_operations);
	return 0;
}
module_init(proc_genhd_init);
1384
#endif /* CONFIG_PROC_FS */
L
Linus Torvalds 已提交
1385

1386
dev_t blk_lookup_devt(const char *name, int partno)
1387
{
1388 1389 1390
	dev_t devt = MKDEV(0, 0);
	struct class_dev_iter iter;
	struct device *dev;
1391

1392 1393
	class_dev_iter_init(&iter, &block_class, NULL, &disk_type);
	while ((dev = class_dev_iter_next(&iter))) {
1394
		struct gendisk *disk = dev_to_disk(dev);
1395
		struct hd_struct *part;
1396

1397
		if (strcmp(dev_name(dev), name))
1398 1399
			continue;

1400 1401 1402 1403 1404 1405 1406 1407
		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;
		}
1408
		part = disk_get_part(disk, partno);
1409
		if (part) {
1410
			devt = part_devt(part);
1411
			disk_put_part(part);
1412
			break;
1413
		}
1414
		disk_put_part(part);
1415
	}
1416
	class_dev_iter_exit(&iter);
1417 1418 1419 1420
	return devt;
}
EXPORT_SYMBOL(blk_lookup_devt);

1421
struct gendisk *__alloc_disk_node(int minors, int node_id)
1422 1423
{
	struct gendisk *disk;
1424
	struct disk_part_tbl *ptbl;
1425

1426 1427
	if (minors > DISK_MAX_PARTS) {
		printk(KERN_ERR
1428
			"block: can't allocate more than %d partitions\n",
1429 1430 1431
			DISK_MAX_PARTS);
		minors = DISK_MAX_PARTS;
	}
1432

1433
	disk = kzalloc_node(sizeof(struct gendisk), GFP_KERNEL, node_id);
L
Linus Torvalds 已提交
1434
	if (disk) {
1435
		if (!init_part_stats(&disk->part0)) {
L
Linus Torvalds 已提交
1436 1437 1438
			kfree(disk);
			return NULL;
		}
1439
		init_rwsem(&disk->lookup_sem);
1440
		disk->node_id = node_id;
1441 1442
		if (disk_expand_part_tbl(disk, 0)) {
			free_part_stats(&disk->part0);
T
Tejun Heo 已提交
1443 1444
			kfree(disk);
			return NULL;
L
Linus Torvalds 已提交
1445
		}
1446 1447
		ptbl = rcu_dereference_protected(disk->part_tbl, 1);
		rcu_assign_pointer(ptbl->part[0], &disk->part0);
1448

1449 1450 1451 1452 1453 1454 1455 1456 1457 1458
		/*
		 * set_capacity() and get_capacity() currently don't use
		 * seqcounter to read/update the part0->nr_sects. Still init
		 * the counter as we can read the sectors in IO submission
		 * patch using seqence counters.
		 *
		 * TODO: Ideally set_capacity() and get_capacity() should be
		 * converted to make use of bd_mutex and sequence counters.
		 */
		seqcount_init(&disk->part0.nr_sects_seq);
1459 1460 1461 1462 1463
		if (hd_ref_init(&disk->part0)) {
			hd_free_part(&disk->part0);
			kfree(disk);
			return NULL;
		}
T
Tejun Heo 已提交
1464

L
Linus Torvalds 已提交
1465 1466
		disk->minors = minors;
		rand_initialize_disk(disk);
1467 1468 1469
		disk_to_dev(disk)->class = &block_class;
		disk_to_dev(disk)->type = &disk_type;
		device_initialize(disk_to_dev(disk));
L
Linus Torvalds 已提交
1470 1471 1472
	}
	return disk;
}
1473
EXPORT_SYMBOL(__alloc_disk_node);
L
Linus Torvalds 已提交
1474

1475
struct kobject *get_disk_and_module(struct gendisk *disk)
L
Linus Torvalds 已提交
1476 1477 1478 1479 1480 1481 1482 1483 1484
{
	struct module *owner;
	struct kobject *kobj;

	if (!disk->fops)
		return NULL;
	owner = disk->fops->owner;
	if (owner && !try_module_get(owner))
		return NULL;
1485
	kobj = kobject_get_unless_zero(&disk_to_dev(disk)->kobj);
L
Linus Torvalds 已提交
1486 1487 1488 1489 1490 1491 1492
	if (kobj == NULL) {
		module_put(owner);
		return NULL;
	}
	return kobj;

}
1493
EXPORT_SYMBOL(get_disk_and_module);
L
Linus Torvalds 已提交
1494 1495 1496 1497

void put_disk(struct gendisk *disk)
{
	if (disk)
1498
		kobject_put(&disk_to_dev(disk)->kobj);
L
Linus Torvalds 已提交
1499 1500 1501
}
EXPORT_SYMBOL(put_disk);

1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516
/*
 * This is a counterpart of get_disk_and_module() and thus also of
 * get_gendisk().
 */
void put_disk_and_module(struct gendisk *disk)
{
	if (disk) {
		struct module *owner = disk->fops->owner;

		put_disk(disk);
		module_put(owner);
	}
}
EXPORT_SYMBOL(put_disk_and_module);

1517 1518 1519 1520 1521 1522 1523 1524 1525 1526
static void set_disk_ro_uevent(struct gendisk *gd, int ro)
{
	char event[] = "DISK_RO=1";
	char *envp[] = { event, NULL };

	if (!ro)
		event[8] = '0';
	kobject_uevent_env(&disk_to_dev(gd)->kobj, KOBJ_CHANGE, envp);
}

L
Linus Torvalds 已提交
1527 1528
void set_device_ro(struct block_device *bdev, int flag)
{
1529
	bdev->bd_part->policy = flag;
L
Linus Torvalds 已提交
1530 1531 1532 1533 1534 1535
}

EXPORT_SYMBOL(set_device_ro);

void set_disk_ro(struct gendisk *disk, int flag)
{
1536 1537 1538
	struct disk_part_iter piter;
	struct hd_struct *part;

1539 1540 1541 1542 1543 1544
	if (disk->part0.policy != flag) {
		set_disk_ro_uevent(disk, flag);
		disk->part0.policy = flag;
	}

	disk_part_iter_init(&piter, disk, DISK_PITER_INCL_EMPTY);
1545 1546 1547
	while ((part = disk_part_iter_next(&piter)))
		part->policy = flag;
	disk_part_iter_exit(&piter);
L
Linus Torvalds 已提交
1548 1549 1550 1551 1552 1553 1554 1555
}

EXPORT_SYMBOL(set_disk_ro);

int bdev_read_only(struct block_device *bdev)
{
	if (!bdev)
		return 0;
1556
	return bdev->bd_part->policy;
L
Linus Torvalds 已提交
1557 1558 1559 1560
}

EXPORT_SYMBOL(bdev_read_only);

1561
int invalidate_partition(struct gendisk *disk, int partno)
L
Linus Torvalds 已提交
1562 1563
{
	int res = 0;
1564
	struct block_device *bdev = bdget_disk(disk, partno);
L
Linus Torvalds 已提交
1565
	if (bdev) {
1566
		fsync_bdev(bdev);
1567
		res = __invalidate_device(bdev, true);
L
Linus Torvalds 已提交
1568 1569 1570 1571 1572 1573
		bdput(bdev);
	}
	return res;
}

EXPORT_SYMBOL(invalidate_partition);
1574 1575 1576 1577 1578 1579 1580 1581 1582

/*
 * Disk events - monitor disk events like media change and eject request.
 */
struct disk_events {
	struct list_head	node;		/* all disk_event's */
	struct gendisk		*disk;		/* the associated disk */
	spinlock_t		lock;

1583
	struct mutex		block_mutex;	/* protects blocking */
1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606
	int			block;		/* event blocking depth */
	unsigned int		pending;	/* events already sent out */
	unsigned int		clearing;	/* events being cleared */

	long			poll_msecs;	/* interval, -1 for default */
	struct delayed_work	dwork;
};

static const char *disk_events_strs[] = {
	[ilog2(DISK_EVENT_MEDIA_CHANGE)]	= "media_change",
	[ilog2(DISK_EVENT_EJECT_REQUEST)]	= "eject_request",
};

static char *disk_uevents[] = {
	[ilog2(DISK_EVENT_MEDIA_CHANGE)]	= "DISK_MEDIA_CHANGE=1",
	[ilog2(DISK_EVENT_EJECT_REQUEST)]	= "DISK_EJECT_REQUEST=1",
};

/* list of all disk_events */
static DEFINE_MUTEX(disk_events_mutex);
static LIST_HEAD(disk_events);

/* disable in-kernel polling by default */
1607
static unsigned long disk_events_dfl_poll_msecs;
1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626

static unsigned long disk_events_poll_jiffies(struct gendisk *disk)
{
	struct disk_events *ev = disk->ev;
	long intv_msecs = 0;

	/*
	 * If device-specific poll interval is set, always use it.  If
	 * the default is being used, poll iff there are events which
	 * can't be monitored asynchronously.
	 */
	if (ev->poll_msecs >= 0)
		intv_msecs = ev->poll_msecs;
	else if (disk->events & ~disk->async_events)
		intv_msecs = disk_events_dfl_poll_msecs;

	return msecs_to_jiffies(intv_msecs);
}

1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642
/**
 * disk_block_events - block and flush disk event checking
 * @disk: disk to block events for
 *
 * On return from this function, it is guaranteed that event checking
 * isn't in progress and won't happen until unblocked by
 * disk_unblock_events().  Events blocking is counted and the actual
 * unblocking happens after the matching number of unblocks are done.
 *
 * Note that this intentionally does not block event checking from
 * disk_clear_events().
 *
 * CONTEXT:
 * Might sleep.
 */
void disk_block_events(struct gendisk *disk)
1643 1644 1645 1646 1647
{
	struct disk_events *ev = disk->ev;
	unsigned long flags;
	bool cancel;

1648 1649 1650
	if (!ev)
		return;

1651 1652 1653 1654 1655 1656
	/*
	 * Outer mutex ensures that the first blocker completes canceling
	 * the event work before further blockers are allowed to finish.
	 */
	mutex_lock(&ev->block_mutex);

1657 1658 1659 1660
	spin_lock_irqsave(&ev->lock, flags);
	cancel = !ev->block++;
	spin_unlock_irqrestore(&ev->lock, flags);

1661 1662
	if (cancel)
		cancel_delayed_work_sync(&disk->ev->dwork);
1663 1664

	mutex_unlock(&ev->block_mutex);
1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682
}

static void __disk_unblock_events(struct gendisk *disk, bool check_now)
{
	struct disk_events *ev = disk->ev;
	unsigned long intv;
	unsigned long flags;

	spin_lock_irqsave(&ev->lock, flags);

	if (WARN_ON_ONCE(ev->block <= 0))
		goto out_unlock;

	if (--ev->block)
		goto out_unlock;

	intv = disk_events_poll_jiffies(disk);
	if (check_now)
1683 1684
		queue_delayed_work(system_freezable_power_efficient_wq,
				&ev->dwork, 0);
1685
	else if (intv)
1686 1687
		queue_delayed_work(system_freezable_power_efficient_wq,
				&ev->dwork, intv);
1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704
out_unlock:
	spin_unlock_irqrestore(&ev->lock, flags);
}

/**
 * disk_unblock_events - unblock disk event checking
 * @disk: disk to unblock events for
 *
 * Undo disk_block_events().  When the block count reaches zero, it
 * starts events polling if configured.
 *
 * CONTEXT:
 * Don't care.  Safe to call from irq context.
 */
void disk_unblock_events(struct gendisk *disk)
{
	if (disk->ev)
1705
		__disk_unblock_events(disk, false);
1706 1707 1708
}

/**
1709 1710 1711
 * disk_flush_events - schedule immediate event checking and flushing
 * @disk: disk to check and flush events for
 * @mask: events to flush
1712
 *
1713 1714 1715
 * Schedule immediate event checking on @disk if not blocked.  Events in
 * @mask are scheduled to be cleared from the driver.  Note that this
 * doesn't clear the events from @disk->ev.
1716 1717
 *
 * CONTEXT:
1718
 * If @mask is non-zero must be called with bdev->bd_mutex held.
1719
 */
1720
void disk_flush_events(struct gendisk *disk, unsigned int mask)
1721
{
1722 1723 1724 1725 1726
	struct disk_events *ev = disk->ev;

	if (!ev)
		return;

1727 1728
	spin_lock_irq(&ev->lock);
	ev->clearing |= mask;
1729
	if (!ev->block)
1730 1731
		mod_delayed_work(system_freezable_power_efficient_wq,
				&ev->dwork, 0);
1732
	spin_unlock_irq(&ev->lock);
1733 1734 1735 1736 1737
}

/**
 * disk_clear_events - synchronously check, clear and return pending events
 * @disk: disk to fetch and clear events from
1738
 * @mask: mask of events to be fetched and cleared
1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750
 *
 * Disk events are synchronously checked and pending events in @mask
 * are cleared and returned.  This ignores the block count.
 *
 * CONTEXT:
 * Might sleep.
 */
unsigned int disk_clear_events(struct gendisk *disk, unsigned int mask)
{
	const struct block_device_operations *bdops = disk->fops;
	struct disk_events *ev = disk->ev;
	unsigned int pending;
1751
	unsigned int clearing = mask;
1752 1753 1754 1755 1756 1757 1758 1759 1760

	if (!ev) {
		/* for drivers still using the old ->media_changed method */
		if ((mask & DISK_EVENT_MEDIA_CHANGE) &&
		    bdops->media_changed && bdops->media_changed(disk))
			return DISK_EVENT_MEDIA_CHANGE;
		return 0;
	}

1761 1762 1763 1764 1765 1766 1767
	disk_block_events(disk);

	/*
	 * store the union of mask and ev->clearing on the stack so that the
	 * race with disk_flush_events does not cause ambiguity (ev->clearing
	 * can still be modified even if events are blocked).
	 */
1768
	spin_lock_irq(&ev->lock);
1769 1770
	clearing |= ev->clearing;
	ev->clearing = 0;
1771 1772
	spin_unlock_irq(&ev->lock);

1773
	disk_check_events(ev, &clearing);
1774
	/*
1775 1776
	 * if ev->clearing is not 0, the disk_flush_events got called in the
	 * middle of this function, so we want to run the workfn without delay.
1777
	 */
1778
	__disk_unblock_events(disk, ev->clearing ? true : false);
1779 1780 1781 1782 1783 1784

	/* then, fetch and clear pending events */
	spin_lock_irq(&ev->lock);
	pending = ev->pending & mask;
	ev->pending &= ~mask;
	spin_unlock_irq(&ev->lock);
1785
	WARN_ON_ONCE(clearing & mask);
1786 1787 1788 1789

	return pending;
}

1790 1791 1792 1793
/*
 * Separate this part out so that a different pointer for clearing_ptr can be
 * passed in for disk_clear_events.
 */
1794 1795 1796 1797
static void disk_events_workfn(struct work_struct *work)
{
	struct delayed_work *dwork = to_delayed_work(work);
	struct disk_events *ev = container_of(dwork, struct disk_events, dwork);
1798 1799 1800 1801 1802 1803 1804

	disk_check_events(ev, &ev->clearing);
}

static void disk_check_events(struct disk_events *ev,
			      unsigned int *clearing_ptr)
{
1805 1806
	struct gendisk *disk = ev->disk;
	char *envp[ARRAY_SIZE(disk_uevents) + 1] = { };
1807
	unsigned int clearing = *clearing_ptr;
1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819
	unsigned int events;
	unsigned long intv;
	int nr_events = 0, i;

	/* check events */
	events = disk->fops->check_events(disk, clearing);

	/* accumulate pending events and schedule next poll if necessary */
	spin_lock_irq(&ev->lock);

	events &= ~ev->pending;
	ev->pending |= events;
1820
	*clearing_ptr &= ~clearing;
1821 1822 1823

	intv = disk_events_poll_jiffies(disk);
	if (!ev->block && intv)
1824 1825
		queue_delayed_work(system_freezable_power_efficient_wq,
				&ev->dwork, intv);
1826 1827 1828

	spin_unlock_irq(&ev->lock);

1829 1830 1831 1832 1833
	/*
	 * Tell userland about new events.  Only the events listed in
	 * @disk->events are reported.  Unlisted events are processed the
	 * same internally but never get reported to userland.
	 */
1834
	for (i = 0; i < ARRAY_SIZE(disk_uevents); i++)
1835
		if (events & disk->events & (1 << i))
1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904
			envp[nr_events++] = disk_uevents[i];

	if (nr_events)
		kobject_uevent_env(&disk_to_dev(disk)->kobj, KOBJ_CHANGE, envp);
}

/*
 * A disk events enabled device has the following sysfs nodes under
 * its /sys/block/X/ directory.
 *
 * events		: list of all supported events
 * events_async		: list of events which can be detected w/o polling
 * events_poll_msecs	: polling interval, 0: disable, -1: system default
 */
static ssize_t __disk_events_show(unsigned int events, char *buf)
{
	const char *delim = "";
	ssize_t pos = 0;
	int i;

	for (i = 0; i < ARRAY_SIZE(disk_events_strs); i++)
		if (events & (1 << i)) {
			pos += sprintf(buf + pos, "%s%s",
				       delim, disk_events_strs[i]);
			delim = " ";
		}
	if (pos)
		pos += sprintf(buf + pos, "\n");
	return pos;
}

static ssize_t disk_events_show(struct device *dev,
				struct device_attribute *attr, char *buf)
{
	struct gendisk *disk = dev_to_disk(dev);

	return __disk_events_show(disk->events, buf);
}

static ssize_t disk_events_async_show(struct device *dev,
				      struct device_attribute *attr, char *buf)
{
	struct gendisk *disk = dev_to_disk(dev);

	return __disk_events_show(disk->async_events, buf);
}

static ssize_t disk_events_poll_msecs_show(struct device *dev,
					   struct device_attribute *attr,
					   char *buf)
{
	struct gendisk *disk = dev_to_disk(dev);

	return sprintf(buf, "%ld\n", disk->ev->poll_msecs);
}

static ssize_t disk_events_poll_msecs_store(struct device *dev,
					    struct device_attribute *attr,
					    const char *buf, size_t count)
{
	struct gendisk *disk = dev_to_disk(dev);
	long intv;

	if (!count || !sscanf(buf, "%ld", &intv))
		return -EINVAL;

	if (intv < 0 && intv != -1)
		return -EINVAL;

1905
	disk_block_events(disk);
1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943
	disk->ev->poll_msecs = intv;
	__disk_unblock_events(disk, true);

	return count;
}

static const DEVICE_ATTR(events, S_IRUGO, disk_events_show, NULL);
static const DEVICE_ATTR(events_async, S_IRUGO, disk_events_async_show, NULL);
static const DEVICE_ATTR(events_poll_msecs, S_IRUGO|S_IWUSR,
			 disk_events_poll_msecs_show,
			 disk_events_poll_msecs_store);

static const struct attribute *disk_events_attrs[] = {
	&dev_attr_events.attr,
	&dev_attr_events_async.attr,
	&dev_attr_events_poll_msecs.attr,
	NULL,
};

/*
 * The default polling interval can be specified by the kernel
 * parameter block.events_dfl_poll_msecs which defaults to 0
 * (disable).  This can also be modified runtime by writing to
 * /sys/module/block/events_dfl_poll_msecs.
 */
static int disk_events_set_dfl_poll_msecs(const char *val,
					  const struct kernel_param *kp)
{
	struct disk_events *ev;
	int ret;

	ret = param_set_ulong(val, kp);
	if (ret < 0)
		return ret;

	mutex_lock(&disk_events_mutex);

	list_for_each_entry(ev, &disk_events, node)
1944
		disk_flush_events(ev->disk, 0);
1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962

	mutex_unlock(&disk_events_mutex);

	return 0;
}

static const struct kernel_param_ops disk_events_dfl_poll_msecs_param_ops = {
	.set	= disk_events_set_dfl_poll_msecs,
	.get	= param_get_ulong,
};

#undef MODULE_PARAM_PREFIX
#define MODULE_PARAM_PREFIX	"block."

module_param_cb(events_dfl_poll_msecs, &disk_events_dfl_poll_msecs_param_ops,
		&disk_events_dfl_poll_msecs, 0644);

/*
1963
 * disk_{alloc|add|del|release}_events - initialize and destroy disk_events.
1964
 */
1965
static void disk_alloc_events(struct gendisk *disk)
1966 1967 1968
{
	struct disk_events *ev;

1969
	if (!disk->fops->check_events)
1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980
		return;

	ev = kzalloc(sizeof(*ev), GFP_KERNEL);
	if (!ev) {
		pr_warn("%s: failed to initialize events\n", disk->disk_name);
		return;
	}

	INIT_LIST_HEAD(&ev->node);
	ev->disk = disk;
	spin_lock_init(&ev->lock);
1981
	mutex_init(&ev->block_mutex);
1982 1983 1984 1985
	ev->block = 1;
	ev->poll_msecs = -1;
	INIT_DELAYED_WORK(&ev->dwork, disk_events_workfn);

1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998
	disk->ev = ev;
}

static void disk_add_events(struct gendisk *disk)
{
	if (!disk->ev)
		return;

	/* FIXME: error handling */
	if (sysfs_create_files(&disk_to_dev(disk)->kobj, disk_events_attrs) < 0)
		pr_warn("%s: failed to create sysfs files for events\n",
			disk->disk_name);

1999
	mutex_lock(&disk_events_mutex);
2000
	list_add_tail(&disk->ev->node, &disk_events);
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
	mutex_unlock(&disk_events_mutex);

	/*
	 * Block count is initialized to 1 and the following initial
	 * unblock kicks it into action.
	 */
	__disk_unblock_events(disk, true);
}

static void disk_del_events(struct gendisk *disk)
{
	if (!disk->ev)
		return;

2015
	disk_block_events(disk);
2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029

	mutex_lock(&disk_events_mutex);
	list_del_init(&disk->ev->node);
	mutex_unlock(&disk_events_mutex);

	sysfs_remove_files(&disk_to_dev(disk)->kobj, disk_events_attrs);
}

static void disk_release_events(struct gendisk *disk)
{
	/* the block count should be 1 from disk_del_events() */
	WARN_ON_ONCE(disk->ev && disk->ev->block != 1);
	kfree(disk->ev);
}
反馈
建议
客服 返回
顶部