Blame view
fs/block_dev.c
42 KB
1da177e4c
|
1 2 3 4 5 6 |
/* * linux/fs/block_dev.c * * Copyright (C) 1991, 1992 Linus Torvalds * Copyright (C) 2001 Andrea Arcangeli <andrea@suse.de> SuSE */ |
1da177e4c
|
7 8 9 10 11 12 |
#include <linux/init.h> #include <linux/mm.h> #include <linux/fcntl.h> #include <linux/slab.h> #include <linux/kmod.h> #include <linux/major.h> |
7db9cfd38
|
13 |
#include <linux/device_cgroup.h> |
1da177e4c
|
14 15 16 17 |
#include <linux/highmem.h> #include <linux/blkdev.h> #include <linux/module.h> #include <linux/blkpg.h> |
b502bd115
|
18 |
#include <linux/magic.h> |
1da177e4c
|
19 |
#include <linux/buffer_head.h> |
ff01bb483
|
20 |
#include <linux/swap.h> |
585d3bc06
|
21 |
#include <linux/pagevec.h> |
811d736f9
|
22 |
#include <linux/writeback.h> |
1da177e4c
|
23 24 25 26 |
#include <linux/mpage.h> #include <linux/mount.h> #include <linux/uio.h> #include <linux/namei.h> |
1368c4f24
|
27 |
#include <linux/log2.h> |
ff01bb483
|
28 |
#include <linux/cleancache.h> |
a27bb332c
|
29 |
#include <linux/aio.h> |
1da177e4c
|
30 |
#include <asm/uaccess.h> |
07f3f05c1
|
31 |
#include "internal.h" |
1da177e4c
|
32 33 34 35 36 |
struct bdev_inode { struct block_device bdev; struct inode vfs_inode; }; |
4c54ac62d
|
37 |
static const struct address_space_operations def_blk_aops; |
1da177e4c
|
38 39 40 41 42 43 44 45 46 |
static inline struct bdev_inode *BDEV_I(struct inode *inode) { return container_of(inode, struct bdev_inode, vfs_inode); } inline struct block_device *I_BDEV(struct inode *inode) { return &BDEV_I(inode)->bdev; } |
1da177e4c
|
47 |
EXPORT_SYMBOL(I_BDEV); |
a5491e0c7
|
48 |
/* |
f758eeabe
|
49 50 51 |
* Move the inode from its current bdi to a new bdi. If the inode is dirty we * need to move it onto the dirty list of @dst so that the inode is always on * the right list. |
a5491e0c7
|
52 53 54 55 |
*/ static void bdev_inode_switch_bdi(struct inode *inode, struct backing_dev_info *dst) { |
f758eeabe
|
56 |
struct backing_dev_info *old = inode->i_data.backing_dev_info; |
a5faeaf91
|
57 |
bool wakeup_bdi = false; |
f758eeabe
|
58 59 60 61 |
if (unlikely(dst == old)) /* deadlock avoidance */ return; bdi_lock_two(&old->wb, &dst->wb); |
250df6ed2
|
62 |
spin_lock(&inode->i_lock); |
a5491e0c7
|
63 |
inode->i_data.backing_dev_info = dst; |
a5faeaf91
|
64 65 66 |
if (inode->i_state & I_DIRTY) { if (bdi_cap_writeback_dirty(dst) && !wb_has_dirty_io(&dst->wb)) wakeup_bdi = true; |
7ccf19a80
|
67 |
list_move(&inode->i_wb_list, &dst->wb.b_dirty); |
a5faeaf91
|
68 |
} |
250df6ed2
|
69 |
spin_unlock(&inode->i_lock); |
f758eeabe
|
70 71 |
spin_unlock(&old->wb.list_lock); spin_unlock(&dst->wb.list_lock); |
a5faeaf91
|
72 73 74 |
if (wakeup_bdi) bdi_wakeup_thread_delayed(dst); |
a5491e0c7
|
75 |
} |
f9a14399a
|
76 |
/* Kill _all_ buffers and pagecache , dirty or not.. */ |
ff01bb483
|
77 |
void kill_bdev(struct block_device *bdev) |
1da177e4c
|
78 |
{ |
ff01bb483
|
79 80 81 |
struct address_space *mapping = bdev->bd_inode->i_mapping; if (mapping->nrpages == 0) |
f9a14399a
|
82 |
return; |
ff01bb483
|
83 |
|
f9a14399a
|
84 |
invalidate_bh_lrus(); |
ff01bb483
|
85 |
truncate_inode_pages(mapping, 0); |
1da177e4c
|
86 |
} |
ff01bb483
|
87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 |
EXPORT_SYMBOL(kill_bdev); /* Invalidate clean unused buffers and pagecache. */ void invalidate_bdev(struct block_device *bdev) { struct address_space *mapping = bdev->bd_inode->i_mapping; if (mapping->nrpages == 0) return; invalidate_bh_lrus(); lru_add_drain_all(); /* make sure all lru add caches are flushed */ invalidate_mapping_pages(mapping, 0, -1); /* 99% of the time, we don't need to flush the cleancache on the bdev. * But, for the strange corners, lets be cautious */ |
3167760f8
|
103 |
cleancache_invalidate_inode(mapping); |
ff01bb483
|
104 105 |
} EXPORT_SYMBOL(invalidate_bdev); |
1da177e4c
|
106 107 108 109 |
int set_blocksize(struct block_device *bdev, int size) { /* Size must be a power of two, and between 512 and PAGE_SIZE */ |
1368c4f24
|
110 |
if (size > PAGE_SIZE || size < 512 || !is_power_of_2(size)) |
1da177e4c
|
111 112 113 |
return -EINVAL; /* Size cannot be smaller than the size supported by the device */ |
e1defc4ff
|
114 |
if (size < bdev_logical_block_size(bdev)) |
1da177e4c
|
115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 |
return -EINVAL; /* Don't change the size if it is same as current */ if (bdev->bd_block_size != size) { sync_blockdev(bdev); bdev->bd_block_size = size; bdev->bd_inode->i_blkbits = blksize_bits(size); kill_bdev(bdev); } return 0; } EXPORT_SYMBOL(set_blocksize); int sb_set_blocksize(struct super_block *sb, int size) { |
1da177e4c
|
131 132 133 134 135 |
if (set_blocksize(sb->s_bdev, size)) return 0; /* If we get here, we know size is power of two * and it's value is between 512 and PAGE_SIZE */ sb->s_blocksize = size; |
38885bd4c
|
136 |
sb->s_blocksize_bits = blksize_bits(size); |
1da177e4c
|
137 138 139 140 141 142 143 |
return sb->s_blocksize; } EXPORT_SYMBOL(sb_set_blocksize); int sb_min_blocksize(struct super_block *sb, int size) { |
e1defc4ff
|
144 |
int minsize = bdev_logical_block_size(sb->s_bdev); |
1da177e4c
|
145 146 147 148 149 150 151 152 153 154 155 |
if (size < minsize) size = minsize; return sb_set_blocksize(sb, size); } EXPORT_SYMBOL(sb_min_blocksize); static int blkdev_get_block(struct inode *inode, sector_t iblock, struct buffer_head *bh, int create) { |
1da177e4c
|
156 157 158 159 160 |
bh->b_bdev = I_BDEV(inode); bh->b_blocknr = iblock; set_buffer_mapped(bh); return 0; } |
b2e895dbd
|
161 162 163 164 165 166 |
static ssize_t blkdev_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov, loff_t offset, unsigned long nr_segs) { struct file *file = iocb->ki_filp; struct inode *inode = file->f_mapping->host; |
eafdc7d19
|
167 |
return __blockdev_direct_IO(rw, iocb, inode, I_BDEV(inode), iov, offset, |
bbec0270b
|
168 |
nr_segs, blkdev_get_block, NULL, NULL, 0); |
b2e895dbd
|
169 |
} |
5cee5815d
|
170 171 172 173 174 175 176 177 |
int __sync_blockdev(struct block_device *bdev, int wait) { if (!bdev) return 0; if (!wait) return filemap_flush(bdev->bd_inode->i_mapping); return filemap_write_and_wait(bdev->bd_inode->i_mapping); } |
585d3bc06
|
178 179 180 181 182 183 |
/* * Write out and wait upon all the dirty data associated with a block * device via its mapping. Does not take the superblock lock. */ int sync_blockdev(struct block_device *bdev) { |
5cee5815d
|
184 |
return __sync_blockdev(bdev, 1); |
585d3bc06
|
185 186 187 188 189 190 191 192 193 194 195 196 |
} EXPORT_SYMBOL(sync_blockdev); /* * Write out and wait upon all dirty data associated with this * device. Filesystem data as well as the underlying block * device. Takes the superblock lock. */ int fsync_bdev(struct block_device *bdev) { struct super_block *sb = get_super(bdev); if (sb) { |
60b0680fa
|
197 |
int res = sync_filesystem(sb); |
585d3bc06
|
198 199 200 201 202 |
drop_super(sb); return res; } return sync_blockdev(bdev); } |
47e4491b4
|
203 |
EXPORT_SYMBOL(fsync_bdev); |
585d3bc06
|
204 205 206 207 208 |
/** * freeze_bdev -- lock a filesystem and force it into a consistent state * @bdev: blockdevice to lock * |
585d3bc06
|
209 210 211 212 213 214 215 216 217 218 219 220 221 222 |
* If a superblock is found on this device, we take the s_umount semaphore * on it to make sure nobody unmounts until the snapshot creation is done. * The reference counter (bd_fsfreeze_count) guarantees that only the last * unfreeze process can unfreeze the frozen filesystem actually when multiple * freeze requests arrive simultaneously. It counts up in freeze_bdev() and * count down in thaw_bdev(). When it becomes 0, thaw_bdev() will unfreeze * actually. */ struct super_block *freeze_bdev(struct block_device *bdev) { struct super_block *sb; int error = 0; mutex_lock(&bdev->bd_fsfreeze_mutex); |
4504230a7
|
223 224 225 226 227 228 |
if (++bdev->bd_fsfreeze_count > 1) { /* * We don't even need to grab a reference - the first call * to freeze_bdev grab an active reference and only the last * thaw_bdev drops it. */ |
585d3bc06
|
229 |
sb = get_super(bdev); |
4504230a7
|
230 231 232 233 234 235 236 237 |
drop_super(sb); mutex_unlock(&bdev->bd_fsfreeze_mutex); return sb; } sb = get_active_super(bdev); if (!sb) goto out; |
18e9e5104
|
238 239 240 241 |
error = freeze_super(sb); if (error) { deactivate_super(sb); bdev->bd_fsfreeze_count--; |
585d3bc06
|
242 |
mutex_unlock(&bdev->bd_fsfreeze_mutex); |
18e9e5104
|
243 |
return ERR_PTR(error); |
585d3bc06
|
244 |
} |
18e9e5104
|
245 |
deactivate_super(sb); |
4504230a7
|
246 |
out: |
585d3bc06
|
247 248 |
sync_blockdev(bdev); mutex_unlock(&bdev->bd_fsfreeze_mutex); |
4fadd7bb2
|
249 |
return sb; /* thaw_bdev releases s->s_umount */ |
585d3bc06
|
250 251 252 253 254 255 256 257 258 259 260 261 |
} EXPORT_SYMBOL(freeze_bdev); /** * thaw_bdev -- unlock filesystem * @bdev: blockdevice to unlock * @sb: associated superblock * * Unlocks the filesystem and marks it writeable again after freeze_bdev(). */ int thaw_bdev(struct block_device *bdev, struct super_block *sb) { |
4504230a7
|
262 |
int error = -EINVAL; |
585d3bc06
|
263 264 |
mutex_lock(&bdev->bd_fsfreeze_mutex); |
4504230a7
|
265 |
if (!bdev->bd_fsfreeze_count) |
18e9e5104
|
266 |
goto out; |
4504230a7
|
267 268 269 |
error = 0; if (--bdev->bd_fsfreeze_count > 0) |
18e9e5104
|
270 |
goto out; |
4504230a7
|
271 272 |
if (!sb) |
18e9e5104
|
273 |
goto out; |
4504230a7
|
274 |
|
18e9e5104
|
275 276 277 278 279 280 281 |
error = thaw_super(sb); if (error) { bdev->bd_fsfreeze_count++; mutex_unlock(&bdev->bd_fsfreeze_mutex); return error; } out: |
585d3bc06
|
282 283 284 285 |
mutex_unlock(&bdev->bd_fsfreeze_mutex); return 0; } EXPORT_SYMBOL(thaw_bdev); |
1da177e4c
|
286 287 288 289 290 291 292 293 294 |
static int blkdev_writepage(struct page *page, struct writeback_control *wbc) { return block_write_full_page(page, blkdev_get_block, wbc); } static int blkdev_readpage(struct file * file, struct page * page) { return block_read_full_page(page, blkdev_get_block); } |
6272b5a58
|
295 296 297 |
static int blkdev_write_begin(struct file *file, struct address_space *mapping, loff_t pos, unsigned len, unsigned flags, struct page **pagep, void **fsdata) |
1da177e4c
|
298 |
{ |
155130a4f
|
299 300 |
return block_write_begin(mapping, pos, len, flags, pagep, blkdev_get_block); |
1da177e4c
|
301 |
} |
6272b5a58
|
302 303 304 |
static int blkdev_write_end(struct file *file, struct address_space *mapping, loff_t pos, unsigned len, unsigned copied, struct page *page, void *fsdata) |
1da177e4c
|
305 |
{ |
6272b5a58
|
306 307 308 309 310 311 312 |
int ret; ret = block_write_end(file, mapping, pos, len, copied, page, fsdata); unlock_page(page); page_cache_release(page); return ret; |
1da177e4c
|
313 314 315 316 |
} /* * private llseek: |
496ad9aa8
|
317 |
* for a block special file file_inode(file)->i_size is zero |
1da177e4c
|
318 319 |
* so we compute the size by hand (just as in block_read/write above) */ |
965c8e59c
|
320 |
static loff_t block_llseek(struct file *file, loff_t offset, int whence) |
1da177e4c
|
321 322 |
{ struct inode *bd_inode = file->f_mapping->host; |
1da177e4c
|
323 |
loff_t retval; |
1b1dcc1b5
|
324 |
mutex_lock(&bd_inode->i_mutex); |
5d48f3a2d
|
325 |
retval = fixed_size_llseek(file, offset, whence, i_size_read(bd_inode)); |
1b1dcc1b5
|
326 |
mutex_unlock(&bd_inode->i_mutex); |
1da177e4c
|
327 328 329 |
return retval; } |
02c24a821
|
330 |
int blkdev_fsync(struct file *filp, loff_t start, loff_t end, int datasync) |
1da177e4c
|
331 |
{ |
b8af67e26
|
332 333 |
struct inode *bd_inode = filp->f_mapping->host; struct block_device *bdev = I_BDEV(bd_inode); |
ab0a9735e
|
334 |
int error; |
da5aa861b
|
335 336 337 338 |
error = filemap_write_and_wait_range(filp->f_mapping, start, end); if (error) return error; |
ab0a9735e
|
339 |
|
b8af67e26
|
340 341 342 343 344 |
/* * There is no need to serialise calls to blkdev_issue_flush with * i_mutex and doing so causes performance issues with concurrent * O_SYNC writers to a block device. */ |
dd3932edd
|
345 |
error = blkdev_issue_flush(bdev, GFP_KERNEL, NULL); |
ab0a9735e
|
346 347 |
if (error == -EOPNOTSUPP) error = 0; |
b8af67e26
|
348 |
|
ab0a9735e
|
349 |
return error; |
1da177e4c
|
350 |
} |
b1dd3b284
|
351 |
EXPORT_SYMBOL(blkdev_fsync); |
1da177e4c
|
352 353 354 355 356 357 |
/* * pseudo-fs */ static __cacheline_aligned_in_smp DEFINE_SPINLOCK(bdev_lock); |
e18b890bb
|
358 |
static struct kmem_cache * bdev_cachep __read_mostly; |
1da177e4c
|
359 360 361 |
static struct inode *bdev_alloc_inode(struct super_block *sb) { |
e94b17660
|
362 |
struct bdev_inode *ei = kmem_cache_alloc(bdev_cachep, GFP_KERNEL); |
1da177e4c
|
363 364 365 366 |
if (!ei) return NULL; return &ei->vfs_inode; } |
fa0d7e3de
|
367 |
static void bdev_i_callback(struct rcu_head *head) |
1da177e4c
|
368 |
{ |
fa0d7e3de
|
369 |
struct inode *inode = container_of(head, struct inode, i_rcu); |
1da177e4c
|
370 |
struct bdev_inode *bdi = BDEV_I(inode); |
1da177e4c
|
371 372 |
kmem_cache_free(bdev_cachep, bdi); } |
fa0d7e3de
|
373 374 375 376 |
static void bdev_destroy_inode(struct inode *inode) { call_rcu(&inode->i_rcu, bdev_i_callback); } |
51cc50685
|
377 |
static void init_once(void *foo) |
1da177e4c
|
378 379 380 |
{ struct bdev_inode *ei = (struct bdev_inode *) foo; struct block_device *bdev = &ei->bdev; |
a35afb830
|
381 382 |
memset(bdev, 0, sizeof(*bdev)); mutex_init(&bdev->bd_mutex); |
a35afb830
|
383 384 |
INIT_LIST_HEAD(&bdev->bd_inodes); INIT_LIST_HEAD(&bdev->bd_list); |
49731baa4
|
385 386 387 |
#ifdef CONFIG_SYSFS INIT_LIST_HEAD(&bdev->bd_holder_disks); #endif |
a35afb830
|
388 |
inode_init_once(&ei->vfs_inode); |
fcccf5025
|
389 390 |
/* Initialize mutex for freeze. */ mutex_init(&bdev->bd_fsfreeze_mutex); |
1da177e4c
|
391 392 393 394 395 396 397 398 |
} static inline void __bd_forget(struct inode *inode) { list_del_init(&inode->i_devices); inode->i_bdev = NULL; inode->i_mapping = &inode->i_data; } |
b57922d97
|
399 |
static void bdev_evict_inode(struct inode *inode) |
1da177e4c
|
400 401 402 |
{ struct block_device *bdev = &BDEV_I(inode)->bdev; struct list_head *p; |
b57922d97
|
403 404 |
truncate_inode_pages(&inode->i_data, 0); invalidate_inode_buffers(inode); /* is it needed here? */ |
dbd5768f8
|
405 |
clear_inode(inode); |
1da177e4c
|
406 407 408 409 410 411 412 |
spin_lock(&bdev_lock); while ( (p = bdev->bd_inodes.next) != &bdev->bd_inodes ) { __bd_forget(list_entry(p, struct inode, i_devices)); } list_del_init(&bdev->bd_list); spin_unlock(&bdev_lock); } |
ee9b6d61a
|
413 |
static const struct super_operations bdev_sops = { |
1da177e4c
|
414 415 416 417 |
.statfs = simple_statfs, .alloc_inode = bdev_alloc_inode, .destroy_inode = bdev_destroy_inode, .drop_inode = generic_delete_inode, |
b57922d97
|
418 |
.evict_inode = bdev_evict_inode, |
1da177e4c
|
419 |
}; |
51139adac
|
420 421 |
static struct dentry *bd_mount(struct file_system_type *fs_type, int flags, const char *dev_name, void *data) |
1da177e4c
|
422 |
{ |
b502bd115
|
423 |
return mount_pseudo(fs_type, "bdev:", &bdev_sops, NULL, BDEVFS_MAGIC); |
1da177e4c
|
424 425 426 427 |
} static struct file_system_type bd_type = { .name = "bdev", |
51139adac
|
428 |
.mount = bd_mount, |
1da177e4c
|
429 430 |
.kill_sb = kill_anon_super, }; |
f47ec3f28
|
431 |
static struct super_block *blockdev_superblock __read_mostly; |
1da177e4c
|
432 433 434 435 |
void __init bdev_cache_init(void) { int err; |
ace8577ae
|
436 |
static struct vfsmount *bd_mnt; |
c2acf7b90
|
437 |
|
1da177e4c
|
438 |
bdev_cachep = kmem_cache_create("bdev_cache", sizeof(struct bdev_inode), |
fffb60f93
|
439 440 |
0, (SLAB_HWCACHE_ALIGN|SLAB_RECLAIM_ACCOUNT| SLAB_MEM_SPREAD|SLAB_PANIC), |
20c2df83d
|
441 |
init_once); |
1da177e4c
|
442 443 444 445 |
err = register_filesystem(&bd_type); if (err) panic("Cannot register bdev pseudo-fs"); bd_mnt = kern_mount(&bd_type); |
1da177e4c
|
446 447 |
if (IS_ERR(bd_mnt)) panic("Cannot create bdev pseudo-fs"); |
ace8577ae
|
448 |
blockdev_superblock = bd_mnt->mnt_sb; /* For writeback */ |
1da177e4c
|
449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 |
} /* * Most likely _very_ bad one - but then it's hardly critical for small * /dev and can be fixed when somebody will need really large one. * Keep in mind that it will be fed through icache hash function too. */ static inline unsigned long hash(dev_t dev) { return MAJOR(dev)+MINOR(dev); } static int bdev_test(struct inode *inode, void *data) { return BDEV_I(inode)->bdev.bd_dev == *(dev_t *)data; } static int bdev_set(struct inode *inode, void *data) { BDEV_I(inode)->bdev.bd_dev = *(dev_t *)data; return 0; } static LIST_HEAD(all_bdevs); struct block_device *bdget(dev_t dev) { struct block_device *bdev; struct inode *inode; |
c2acf7b90
|
478 |
inode = iget5_locked(blockdev_superblock, hash(dev), |
1da177e4c
|
479 480 481 482 483 484 485 486 487 |
bdev_test, bdev_set, &dev); if (!inode) return NULL; bdev = &BDEV_I(inode)->bdev; if (inode->i_state & I_NEW) { bdev->bd_contains = NULL; |
782b94cdf
|
488 |
bdev->bd_super = NULL; |
1da177e4c
|
489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 |
bdev->bd_inode = inode; bdev->bd_block_size = (1 << inode->i_blkbits); bdev->bd_part_count = 0; bdev->bd_invalidated = 0; inode->i_mode = S_IFBLK; inode->i_rdev = dev; inode->i_bdev = bdev; inode->i_data.a_ops = &def_blk_aops; mapping_set_gfp_mask(&inode->i_data, GFP_USER); inode->i_data.backing_dev_info = &default_backing_dev_info; spin_lock(&bdev_lock); list_add(&bdev->bd_list, &all_bdevs); spin_unlock(&bdev_lock); unlock_new_inode(inode); } return bdev; } EXPORT_SYMBOL(bdget); |
dddac6a7b
|
508 509 510 511 512 513 |
/** * bdgrab -- Grab a reference to an already referenced block device * @bdev: Block device to grab a reference to. */ struct block_device *bdgrab(struct block_device *bdev) { |
7de9c6ee3
|
514 |
ihold(bdev->bd_inode); |
dddac6a7b
|
515 516 |
return bdev; } |
c1681bf8a
|
517 |
EXPORT_SYMBOL(bdgrab); |
dddac6a7b
|
518 |
|
1da177e4c
|
519 520 |
long nr_blockdev_pages(void) { |
203a2935c
|
521 |
struct block_device *bdev; |
1da177e4c
|
522 523 |
long ret = 0; spin_lock(&bdev_lock); |
203a2935c
|
524 |
list_for_each_entry(bdev, &all_bdevs, bd_list) { |
1da177e4c
|
525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 |
ret += bdev->bd_inode->i_mapping->nrpages; } spin_unlock(&bdev_lock); return ret; } void bdput(struct block_device *bdev) { iput(bdev->bd_inode); } EXPORT_SYMBOL(bdput); static struct block_device *bd_acquire(struct inode *inode) { struct block_device *bdev; |
09d967c6f
|
541 |
|
1da177e4c
|
542 543 |
spin_lock(&bdev_lock); bdev = inode->i_bdev; |
09d967c6f
|
544 |
if (bdev) { |
7de9c6ee3
|
545 |
ihold(bdev->bd_inode); |
1da177e4c
|
546 547 548 549 |
spin_unlock(&bdev_lock); return bdev; } spin_unlock(&bdev_lock); |
09d967c6f
|
550 |
|
1da177e4c
|
551 552 553 |
bdev = bdget(inode->i_rdev); if (bdev) { spin_lock(&bdev_lock); |
09d967c6f
|
554 555 |
if (!inode->i_bdev) { /* |
7de9c6ee3
|
556 |
* We take an additional reference to bd_inode, |
09d967c6f
|
557 558 559 560 |
* and it's released in clear_inode() of inode. * So, we can access it via ->i_mapping always * without igrab(). */ |
7de9c6ee3
|
561 |
ihold(bdev->bd_inode); |
09d967c6f
|
562 563 564 565 |
inode->i_bdev = bdev; inode->i_mapping = bdev->bd_inode->i_mapping; list_add(&inode->i_devices, &bdev->bd_inodes); } |
1da177e4c
|
566 567 568 569 |
spin_unlock(&bdev_lock); } return bdev; } |
f47ec3f28
|
570 571 572 573 |
static inline int sb_is_blkdev_sb(struct super_block *sb) { return sb == blockdev_superblock; } |
1da177e4c
|
574 575 576 577 |
/* Call when you free inode */ void bd_forget(struct inode *inode) { |
09d967c6f
|
578 |
struct block_device *bdev = NULL; |
1da177e4c
|
579 |
spin_lock(&bdev_lock); |
b4ea2eaa1
|
580 581 582 |
if (!sb_is_blkdev_sb(inode->i_sb)) bdev = inode->i_bdev; __bd_forget(inode); |
1da177e4c
|
583 |
spin_unlock(&bdev_lock); |
09d967c6f
|
584 585 586 |
if (bdev) iput(bdev->bd_inode); |
1da177e4c
|
587 |
} |
1a3cbbc5a
|
588 589 590 591 592 593 |
/** * bd_may_claim - test whether a block device can be claimed * @bdev: block device of interest * @whole: whole block device containing @bdev, may equal @bdev * @holder: holder trying to claim @bdev * |
25985edce
|
594 |
* Test whether @bdev can be claimed by @holder. |
1a3cbbc5a
|
595 596 597 598 599 600 601 602 603 |
* * CONTEXT: * spin_lock(&bdev_lock). * * RETURNS: * %true if @bdev can be claimed, %false otherwise. */ static bool bd_may_claim(struct block_device *bdev, struct block_device *whole, void *holder) |
1da177e4c
|
604 |
{ |
1da177e4c
|
605 |
if (bdev->bd_holder == holder) |
1a3cbbc5a
|
606 |
return true; /* already a holder */ |
1da177e4c
|
607 |
else if (bdev->bd_holder != NULL) |
1a3cbbc5a
|
608 |
return false; /* held by someone else */ |
1da177e4c
|
609 |
else if (bdev->bd_contains == bdev) |
1a3cbbc5a
|
610 |
return true; /* is a whole device which isn't held */ |
1da177e4c
|
611 |
|
e525fd89d
|
612 |
else if (whole->bd_holder == bd_may_claim) |
1a3cbbc5a
|
613 614 615 |
return true; /* is a partition of a device that is being partitioned */ else if (whole->bd_holder != NULL) return false; /* is a partition of a held device */ |
1da177e4c
|
616 |
else |
1a3cbbc5a
|
617 618 619 620 |
return true; /* is a partition of an un-held device */ } /** |
6b4517a79
|
621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 |
* bd_prepare_to_claim - prepare to claim a block device * @bdev: block device of interest * @whole: the whole device containing @bdev, may equal @bdev * @holder: holder trying to claim @bdev * * Prepare to claim @bdev. This function fails if @bdev is already * claimed by another holder and waits if another claiming is in * progress. This function doesn't actually claim. On successful * return, the caller has ownership of bd_claiming and bd_holder[s]. * * CONTEXT: * spin_lock(&bdev_lock). Might release bdev_lock, sleep and regrab * it multiple times. * * RETURNS: * 0 if @bdev can be claimed, -EBUSY otherwise. */ static int bd_prepare_to_claim(struct block_device *bdev, struct block_device *whole, void *holder) { retry: /* if someone else claimed, fail */ if (!bd_may_claim(bdev, whole, holder)) return -EBUSY; |
e75aa8589
|
645 646 |
/* if claiming is already in progress, wait for it to finish */ if (whole->bd_claiming) { |
6b4517a79
|
647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 |
wait_queue_head_t *wq = bit_waitqueue(&whole->bd_claiming, 0); DEFINE_WAIT(wait); prepare_to_wait(wq, &wait, TASK_UNINTERRUPTIBLE); spin_unlock(&bdev_lock); schedule(); finish_wait(wq, &wait); spin_lock(&bdev_lock); goto retry; } /* yay, all mine */ return 0; } /** * bd_start_claiming - start claiming a block device * @bdev: block device of interest * @holder: holder trying to claim @bdev * * @bdev is about to be opened exclusively. Check @bdev can be opened * exclusively and mark that an exclusive open is in progress. Each * successful call to this function must be matched with a call to |
b0018361c
|
670 671 672 673 674 675 676 |
* either bd_finish_claiming() or bd_abort_claiming() (which do not * fail). * * This function is used to gain exclusive access to the block device * without actually causing other exclusive open attempts to fail. It * should be used when the open sequence itself requires exclusive * access but may subsequently fail. |
6b4517a79
|
677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 |
* * CONTEXT: * Might sleep. * * RETURNS: * Pointer to the block device containing @bdev on success, ERR_PTR() * value on failure. */ static struct block_device *bd_start_claiming(struct block_device *bdev, void *holder) { struct gendisk *disk; struct block_device *whole; int partno, err; might_sleep(); /* * @bdev might not have been initialized properly yet, look up * and grab the outer block device the hard way. */ disk = get_gendisk(bdev->bd_dev, &partno); if (!disk) return ERR_PTR(-ENXIO); |
d4c208b86
|
701 702 703 704 705 706 707 708 709 710 711 712 |
/* * Normally, @bdev should equal what's returned from bdget_disk() * if partno is 0; however, some drivers (floppy) use multiple * bdev's for the same physical device and @bdev may be one of the * aliases. Keep @bdev if partno is 0. This means claimer * tracking is broken for those devices but it has always been that * way. */ if (partno) whole = bdget_disk(disk, 0); else whole = bdgrab(bdev); |
cf3425707
|
713 |
module_put(disk->fops->owner); |
6b4517a79
|
714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 |
put_disk(disk); if (!whole) return ERR_PTR(-ENOMEM); /* prepare to claim, if successful, mark claiming in progress */ spin_lock(&bdev_lock); err = bd_prepare_to_claim(bdev, whole, holder); if (err == 0) { whole->bd_claiming = holder; spin_unlock(&bdev_lock); return whole; } else { spin_unlock(&bdev_lock); bdput(whole); return ERR_PTR(err); } } |
641dc636b
|
732 |
#ifdef CONFIG_SYSFS |
49731baa4
|
733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 |
struct bd_holder_disk { struct list_head list; struct gendisk *disk; int refcnt; }; static struct bd_holder_disk *bd_find_holder_disk(struct block_device *bdev, struct gendisk *disk) { struct bd_holder_disk *holder; list_for_each_entry(holder, &bdev->bd_holder_disks, list) if (holder->disk == disk) return holder; return NULL; } |
4d7dd8fd9
|
749 |
static int add_symlink(struct kobject *from, struct kobject *to) |
641dc636b
|
750 |
{ |
4d7dd8fd9
|
751 |
return sysfs_create_link(from, to, kobject_name(to)); |
641dc636b
|
752 753 754 755 |
} static void del_symlink(struct kobject *from, struct kobject *to) { |
641dc636b
|
756 757 |
sysfs_remove_link(from, kobject_name(to)); } |
641dc636b
|
758 |
/** |
e09b457bd
|
759 760 761 |
* bd_link_disk_holder - create symlinks between holding disk and slave bdev * @bdev: the claimed slave bdev * @disk: the holding disk |
df6c0cd9a
|
762 |
* |
49731baa4
|
763 764 |
* DON'T USE THIS UNLESS YOU'RE ALREADY USING IT. * |
e09b457bd
|
765 |
* This functions creates the following sysfs symlinks. |
641dc636b
|
766 |
* |
e09b457bd
|
767 768 |
* - from "slaves" directory of the holder @disk to the claimed @bdev * - from "holders" directory of the @bdev to the holder @disk |
641dc636b
|
769 |
* |
e09b457bd
|
770 771 |
* For example, if /dev/dm-0 maps to /dev/sda and disk for dm-0 is * passed to bd_link_disk_holder(), then: |
641dc636b
|
772 |
* |
e09b457bd
|
773 774 |
* /sys/block/dm-0/slaves/sda --> /sys/block/sda * /sys/block/sda/holders/dm-0 --> /sys/block/dm-0 |
641dc636b
|
775 |
* |
e09b457bd
|
776 777 778 |
* The caller must have claimed @bdev before calling this function and * ensure that both @bdev and @disk are valid during the creation and * lifetime of these symlinks. |
641dc636b
|
779 |
* |
e09b457bd
|
780 781 |
* CONTEXT: * Might sleep. |
641dc636b
|
782 |
* |
e09b457bd
|
783 784 |
* RETURNS: * 0 on success, -errno on failure. |
641dc636b
|
785 |
*/ |
e09b457bd
|
786 |
int bd_link_disk_holder(struct block_device *bdev, struct gendisk *disk) |
641dc636b
|
787 |
{ |
49731baa4
|
788 |
struct bd_holder_disk *holder; |
e09b457bd
|
789 |
int ret = 0; |
641dc636b
|
790 |
|
2e7b651df
|
791 |
mutex_lock(&bdev->bd_mutex); |
df6c0cd9a
|
792 |
|
49731baa4
|
793 |
WARN_ON_ONCE(!bdev->bd_holder); |
4e91672c7
|
794 |
|
e09b457bd
|
795 796 797 |
/* FIXME: remove the following once add_disk() handles errors */ if (WARN_ON(!disk->slave_dir || !bdev->bd_part->holder_dir)) goto out_unlock; |
4e91672c7
|
798 |
|
49731baa4
|
799 800 801 |
holder = bd_find_holder_disk(bdev, disk); if (holder) { holder->refcnt++; |
e09b457bd
|
802 |
goto out_unlock; |
49731baa4
|
803 |
} |
641dc636b
|
804 |
|
49731baa4
|
805 806 807 |
holder = kzalloc(sizeof(*holder), GFP_KERNEL); if (!holder) { ret = -ENOMEM; |
e09b457bd
|
808 809 |
goto out_unlock; } |
641dc636b
|
810 |
|
49731baa4
|
811 812 813 814 815 816 817 818 819 820 821 |
INIT_LIST_HEAD(&holder->list); holder->disk = disk; holder->refcnt = 1; ret = add_symlink(disk->slave_dir, &part_to_dev(bdev->bd_part)->kobj); if (ret) goto out_free; ret = add_symlink(bdev->bd_part->holder_dir, &disk_to_dev(disk)->kobj); if (ret) goto out_del; |
e7407d161
|
822 823 824 825 826 |
/* * bdev could be deleted beneath us which would implicitly destroy * the holder directory. Hold on to it. */ kobject_get(bdev->bd_part->holder_dir); |
49731baa4
|
827 828 829 830 831 832 833 834 |
list_add(&holder->list, &bdev->bd_holder_disks); goto out_unlock; out_del: del_symlink(disk->slave_dir, &part_to_dev(bdev->bd_part)->kobj); out_free: kfree(holder); |
e09b457bd
|
835 |
out_unlock: |
b4cf1b72e
|
836 |
mutex_unlock(&bdev->bd_mutex); |
e09b457bd
|
837 |
return ret; |
641dc636b
|
838 |
} |
e09b457bd
|
839 |
EXPORT_SYMBOL_GPL(bd_link_disk_holder); |
641dc636b
|
840 |
|
49731baa4
|
841 842 843 844 845 846 847 848 849 850 851 |
/** * bd_unlink_disk_holder - destroy symlinks created by bd_link_disk_holder() * @bdev: the calimed slave bdev * @disk: the holding disk * * DON'T USE THIS UNLESS YOU'RE ALREADY USING IT. * * CONTEXT: * Might sleep. */ void bd_unlink_disk_holder(struct block_device *bdev, struct gendisk *disk) |
641dc636b
|
852 |
{ |
49731baa4
|
853 |
struct bd_holder_disk *holder; |
641dc636b
|
854 |
|
49731baa4
|
855 |
mutex_lock(&bdev->bd_mutex); |
641dc636b
|
856 |
|
49731baa4
|
857 858 859 860 861 862 |
holder = bd_find_holder_disk(bdev, disk); if (!WARN_ON_ONCE(holder == NULL) && !--holder->refcnt) { del_symlink(disk->slave_dir, &part_to_dev(bdev->bd_part)->kobj); del_symlink(bdev->bd_part->holder_dir, &disk_to_dev(disk)->kobj); |
e7407d161
|
863 |
kobject_put(bdev->bd_part->holder_dir); |
49731baa4
|
864 865 866 867 868 |
list_del_init(&holder->list); kfree(holder); } mutex_unlock(&bdev->bd_mutex); |
1da177e4c
|
869 |
} |
49731baa4
|
870 |
EXPORT_SYMBOL_GPL(bd_unlink_disk_holder); |
641dc636b
|
871 |
#endif |
1da177e4c
|
872 |
|
0c002c2f7
|
873 |
/** |
56ade44b4
|
874 875 876 |
* flush_disk - invalidates all buffer-cache entries on a disk * * @bdev: struct block device to be flushed |
e6eb5ce1b
|
877 |
* @kill_dirty: flag to guide handling of dirty inodes |
56ade44b4
|
878 879 880 881 882 |
* * Invalidates all buffer-cache entries on a disk. It should be called * when a disk has been changed -- either by a media change or online * resize. */ |
93b270f76
|
883 |
static void flush_disk(struct block_device *bdev, bool kill_dirty) |
56ade44b4
|
884 |
{ |
93b270f76
|
885 |
if (__invalidate_device(bdev, kill_dirty)) { |
56ade44b4
|
886 887 888 889 890 891 892 893 894 895 896 |
char name[BDEVNAME_SIZE] = ""; if (bdev->bd_disk) disk_name(bdev->bd_disk, 0, name); printk(KERN_WARNING "VFS: busy inodes on changed media or " "resized disk %s ", name); } if (!bdev->bd_disk) return; |
d27769ec3
|
897 |
if (disk_part_scan_enabled(bdev->bd_disk)) |
56ade44b4
|
898 899 900 901 |
bdev->bd_invalidated = 1; } /** |
57d1b5366
|
902 |
* check_disk_size_change - checks for disk size change and adjusts bdev size. |
c3279d145
|
903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 |
* @disk: struct gendisk to check * @bdev: struct bdev to adjust. * * This routine checks to see if the bdev size does not match the disk size * and adjusts it if it differs. */ void check_disk_size_change(struct gendisk *disk, struct block_device *bdev) { loff_t disk_size, bdev_size; disk_size = (loff_t)get_capacity(disk) << 9; bdev_size = i_size_read(bdev->bd_inode); if (disk_size != bdev_size) { char name[BDEVNAME_SIZE]; disk_name(disk, 0, name); printk(KERN_INFO "%s: detected capacity change from %lld to %lld ", name, bdev_size, disk_size); i_size_write(bdev->bd_inode, disk_size); |
93b270f76
|
924 |
flush_disk(bdev, false); |
c3279d145
|
925 926 927 928 929 |
} } EXPORT_SYMBOL(check_disk_size_change); /** |
57d1b5366
|
930 |
* revalidate_disk - wrapper for lower-level driver's revalidate_disk call-back |
0c002c2f7
|
931 932 933 934 935 936 937 938 |
* @disk: struct gendisk to be revalidated * * This routine is a wrapper for lower-level driver's revalidate_disk * call-backs. It is used to do common pre and post operations needed * for all revalidate_disk operations. */ int revalidate_disk(struct gendisk *disk) { |
c3279d145
|
939 |
struct block_device *bdev; |
0c002c2f7
|
940 941 942 943 |
int ret = 0; if (disk->fops->revalidate_disk) ret = disk->fops->revalidate_disk(disk); |
c3279d145
|
944 945 946 947 948 949 |
bdev = bdget_disk(disk, 0); if (!bdev) return ret; mutex_lock(&bdev->bd_mutex); check_disk_size_change(disk, bdev); |
7630b661d
|
950 |
bdev->bd_invalidated = 0; |
c3279d145
|
951 952 |
mutex_unlock(&bdev->bd_mutex); bdput(bdev); |
0c002c2f7
|
953 954 955 |
return ret; } EXPORT_SYMBOL(revalidate_disk); |
1da177e4c
|
956 957 958 959 960 961 962 963 964 965 966 967 |
/* * This routine checks whether a removable media has been changed, * and invalidates all buffer-cache-entries in that case. This * is a relatively slow routine, so we have to try to minimize using * it. Thus it is called only upon a 'mount' or 'open'. This * is the best way of combining speed and utility, I think. * People changing diskettes in the middle of an operation deserve * to lose :-) */ int check_disk_change(struct block_device *bdev) { struct gendisk *disk = bdev->bd_disk; |
83d5cde47
|
968 |
const struct block_device_operations *bdops = disk->fops; |
77ea887e4
|
969 |
unsigned int events; |
1da177e4c
|
970 |
|
77ea887e4
|
971 972 973 |
events = disk_clear_events(disk, DISK_EVENT_MEDIA_CHANGE | DISK_EVENT_EJECT_REQUEST); if (!(events & DISK_EVENT_MEDIA_CHANGE)) |
1da177e4c
|
974 |
return 0; |
93b270f76
|
975 |
flush_disk(bdev, true); |
1da177e4c
|
976 977 |
if (bdops->revalidate_disk) bdops->revalidate_disk(bdev->bd_disk); |
1da177e4c
|
978 979 980 981 982 983 984 |
return 1; } EXPORT_SYMBOL(check_disk_change); void bd_set_size(struct block_device *bdev, loff_t size) { |
e1defc4ff
|
985 |
unsigned bsize = bdev_logical_block_size(bdev); |
1da177e4c
|
986 |
|
d646a02a9
|
987 988 989 |
mutex_lock(&bdev->bd_inode->i_mutex); i_size_write(bdev->bd_inode, size); mutex_unlock(&bdev->bd_inode->i_mutex); |
1da177e4c
|
990 991 992 993 994 995 996 997 998 |
while (bsize < PAGE_CACHE_SIZE) { if (size & bsize) break; bsize <<= 1; } bdev->bd_block_size = bsize; bdev->bd_inode->i_blkbits = blksize_bits(bsize); } EXPORT_SYMBOL(bd_set_size); |
4385bab12
|
999 |
static void __blkdev_put(struct block_device *bdev, fmode_t mode, int for_part); |
37be41241
|
1000 |
|
6d740cd5b
|
1001 1002 1003 1004 1005 1006 |
/* * bd_mutex locking: * * mutex_lock(part->bd_mutex) * mutex_lock_nested(whole->bd_mutex, 1) */ |
572c48921
|
1007 |
static int __blkdev_get(struct block_device *bdev, fmode_t mode, int for_part) |
1da177e4c
|
1008 |
{ |
1da177e4c
|
1009 |
struct gendisk *disk; |
523e1d399
|
1010 |
struct module *owner; |
7db9cfd38
|
1011 |
int ret; |
cf771cb5a
|
1012 |
int partno; |
fe6e9c1f2
|
1013 |
int perm = 0; |
572c48921
|
1014 |
if (mode & FMODE_READ) |
fe6e9c1f2
|
1015 |
perm |= MAY_READ; |
572c48921
|
1016 |
if (mode & FMODE_WRITE) |
fe6e9c1f2
|
1017 1018 1019 1020 |
perm |= MAY_WRITE; /* * hooks: /n/, see "layering violations". */ |
b7300b78d
|
1021 1022 1023 1024 1025 1026 |
if (!for_part) { ret = devcgroup_inode_permission(bdev->bd_inode, perm); if (ret != 0) { bdput(bdev); return ret; } |
826660203
|
1027 |
} |
7db9cfd38
|
1028 |
|
d3374825c
|
1029 |
restart: |
0762b8bde
|
1030 |
|
89f97496e
|
1031 |
ret = -ENXIO; |
cf771cb5a
|
1032 |
disk = get_gendisk(bdev->bd_dev, &partno); |
0762b8bde
|
1033 |
if (!disk) |
6e9624b8c
|
1034 |
goto out; |
523e1d399
|
1035 |
owner = disk->fops->owner; |
1da177e4c
|
1036 |
|
69e02c59a
|
1037 |
disk_block_events(disk); |
6796bf54a
|
1038 |
mutex_lock_nested(&bdev->bd_mutex, for_part); |
1da177e4c
|
1039 1040 |
if (!bdev->bd_openers) { bdev->bd_disk = disk; |
87192a2a4
|
1041 |
bdev->bd_queue = disk->queue; |
1da177e4c
|
1042 |
bdev->bd_contains = bdev; |
cf771cb5a
|
1043 |
if (!partno) { |
1da177e4c
|
1044 |
struct backing_dev_info *bdi; |
89f97496e
|
1045 1046 1047 1048 1049 |
ret = -ENXIO; bdev->bd_part = disk_get_part(disk, partno); if (!bdev->bd_part) goto out_clear; |
1196f8b81
|
1050 |
ret = 0; |
1da177e4c
|
1051 |
if (disk->fops->open) { |
572c48921
|
1052 |
ret = disk->fops->open(bdev, mode); |
d3374825c
|
1053 1054 1055 1056 1057 1058 1059 |
if (ret == -ERESTARTSYS) { /* Lost a race with 'disk' being * deleted, try again. * See md.c */ disk_put_part(bdev->bd_part); bdev->bd_part = NULL; |
d3374825c
|
1060 |
bdev->bd_disk = NULL; |
87192a2a4
|
1061 |
bdev->bd_queue = NULL; |
d3374825c
|
1062 |
mutex_unlock(&bdev->bd_mutex); |
69e02c59a
|
1063 |
disk_unblock_events(disk); |
69e02c59a
|
1064 |
put_disk(disk); |
523e1d399
|
1065 |
module_put(owner); |
d3374825c
|
1066 1067 |
goto restart; } |
1da177e4c
|
1068 |
} |
7e69723fe
|
1069 |
|
de33127d8
|
1070 |
if (!ret) { |
7e69723fe
|
1071 1072 1073 1074 1075 1076 |
bd_set_size(bdev,(loff_t)get_capacity(disk)<<9); bdi = blk_get_backing_dev_info(bdev); if (bdi == NULL) bdi = &default_backing_dev_info; bdev_inode_switch_bdi(bdev->bd_inode, bdi); } |
1196f8b81
|
1077 1078 1079 1080 1081 1082 |
/* * If the device is invalidated, rescan partition * if open succeeded or failed with -ENOMEDIUM. * The latter is necessary to prevent ghost * partitions on a removed medium. */ |
fe316bf2d
|
1083 1084 1085 1086 1087 1088 |
if (bdev->bd_invalidated) { if (!ret) rescan_partitions(disk, bdev); else if (ret == -ENOMEDIUM) invalidate_partitions(disk, bdev); } |
1196f8b81
|
1089 1090 |
if (ret) goto out_clear; |
1da177e4c
|
1091 |
} else { |
1da177e4c
|
1092 1093 1094 1095 |
struct block_device *whole; whole = bdget_disk(disk, 0); ret = -ENOMEM; if (!whole) |
0762b8bde
|
1096 |
goto out_clear; |
37be41241
|
1097 |
BUG_ON(for_part); |
572c48921
|
1098 |
ret = __blkdev_get(whole, mode, 1); |
1da177e4c
|
1099 |
if (ret) |
0762b8bde
|
1100 |
goto out_clear; |
1da177e4c
|
1101 |
bdev->bd_contains = whole; |
a5491e0c7
|
1102 1103 |
bdev_inode_switch_bdi(bdev->bd_inode, whole->bd_inode->i_data.backing_dev_info); |
89f97496e
|
1104 |
bdev->bd_part = disk_get_part(disk, partno); |
e71bf0d0e
|
1105 |
if (!(disk->flags & GENHD_FL_UP) || |
89f97496e
|
1106 |
!bdev->bd_part || !bdev->bd_part->nr_sects) { |
1da177e4c
|
1107 |
ret = -ENXIO; |
0762b8bde
|
1108 |
goto out_clear; |
1da177e4c
|
1109 |
} |
89f97496e
|
1110 |
bd_set_size(bdev, (loff_t)bdev->bd_part->nr_sects << 9); |
1da177e4c
|
1111 1112 |
} } else { |
1da177e4c
|
1113 |
if (bdev->bd_contains == bdev) { |
1196f8b81
|
1114 1115 |
ret = 0; if (bdev->bd_disk->fops->open) |
572c48921
|
1116 |
ret = bdev->bd_disk->fops->open(bdev, mode); |
1196f8b81
|
1117 |
/* the same as first opener case, read comment there */ |
fe316bf2d
|
1118 1119 1120 1121 1122 1123 |
if (bdev->bd_invalidated) { if (!ret) rescan_partitions(bdev->bd_disk, bdev); else if (ret == -ENOMEDIUM) invalidate_partitions(bdev->bd_disk, bdev); } |
1196f8b81
|
1124 1125 |
if (ret) goto out_unlock_bdev; |
1da177e4c
|
1126 |
} |
69e02c59a
|
1127 |
/* only one opener holds refs to the module and disk */ |
69e02c59a
|
1128 |
put_disk(disk); |
523e1d399
|
1129 |
module_put(owner); |
1da177e4c
|
1130 1131 |
} bdev->bd_openers++; |
37be41241
|
1132 1133 |
if (for_part) bdev->bd_part_count++; |
c039e3134
|
1134 |
mutex_unlock(&bdev->bd_mutex); |
69e02c59a
|
1135 |
disk_unblock_events(disk); |
1da177e4c
|
1136 |
return 0; |
0762b8bde
|
1137 |
out_clear: |
89f97496e
|
1138 |
disk_put_part(bdev->bd_part); |
1da177e4c
|
1139 |
bdev->bd_disk = NULL; |
0762b8bde
|
1140 |
bdev->bd_part = NULL; |
87192a2a4
|
1141 |
bdev->bd_queue = NULL; |
a5491e0c7
|
1142 |
bdev_inode_switch_bdi(bdev->bd_inode, &default_backing_dev_info); |
1da177e4c
|
1143 |
if (bdev != bdev->bd_contains) |
572c48921
|
1144 |
__blkdev_put(bdev->bd_contains, mode, 1); |
1da177e4c
|
1145 |
bdev->bd_contains = NULL; |
0762b8bde
|
1146 |
out_unlock_bdev: |
c039e3134
|
1147 |
mutex_unlock(&bdev->bd_mutex); |
69e02c59a
|
1148 |
disk_unblock_events(disk); |
0762b8bde
|
1149 |
put_disk(disk); |
523e1d399
|
1150 |
module_put(owner); |
4345caba3
|
1151 |
out: |
0762b8bde
|
1152 |
bdput(bdev); |
1da177e4c
|
1153 1154 |
return ret; } |
d4d776299
|
1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 |
/** * blkdev_get - open a block device * @bdev: block_device to open * @mode: FMODE_* mask * @holder: exclusive holder identifier * * Open @bdev with @mode. If @mode includes %FMODE_EXCL, @bdev is * open with exclusive access. Specifying %FMODE_EXCL with %NULL * @holder is invalid. Exclusive opens may nest for the same @holder. * * On success, the reference count of @bdev is unchanged. On failure, * @bdev is put. * * CONTEXT: * Might sleep. * * RETURNS: * 0 on success, -errno on failure. */ |
e525fd89d
|
1174 |
int blkdev_get(struct block_device *bdev, fmode_t mode, void *holder) |
1da177e4c
|
1175 |
{ |
e525fd89d
|
1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 |
struct block_device *whole = NULL; int res; WARN_ON_ONCE((mode & FMODE_EXCL) && !holder); if ((mode & FMODE_EXCL) && holder) { whole = bd_start_claiming(bdev, holder); if (IS_ERR(whole)) { bdput(bdev); return PTR_ERR(whole); } } res = __blkdev_get(bdev, mode, 0); if (whole) { |
d4dc210f6
|
1192 |
struct gendisk *disk = whole->bd_disk; |
6a027eff6
|
1193 |
/* finish claiming */ |
77ea887e4
|
1194 |
mutex_lock(&bdev->bd_mutex); |
6a027eff6
|
1195 |
spin_lock(&bdev_lock); |
77ea887e4
|
1196 |
if (!res) { |
6a027eff6
|
1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 |
BUG_ON(!bd_may_claim(bdev, whole, holder)); /* * Note that for a whole device bd_holders * will be incremented twice, and bd_holder * will be set to bd_may_claim before being * set to holder */ whole->bd_holders++; whole->bd_holder = bd_may_claim; bdev->bd_holders++; bdev->bd_holder = holder; } /* tell others that we're done */ BUG_ON(whole->bd_claiming != holder); whole->bd_claiming = NULL; wake_up_bit(&whole->bd_claiming, 0); spin_unlock(&bdev_lock); |
77ea887e4
|
1216 1217 |
/* |
d4dc210f6
|
1218 1219 1220 1221 1222 |
* Block event polling for write claims if requested. Any * write holder makes the write_holder state stick until * all are released. This is good enough and tracking * individual writeable reference is too fragile given the * way @mode is used in blkdev_get/put(). |
77ea887e4
|
1223 |
*/ |
4c49ff3fe
|
1224 1225 |
if (!res && (mode & FMODE_WRITE) && !bdev->bd_write_holder && (disk->flags & GENHD_FL_BLOCK_EVENTS_ON_EXCL_WRITE)) { |
77ea887e4
|
1226 |
bdev->bd_write_holder = true; |
d4dc210f6
|
1227 |
disk_block_events(disk); |
77ea887e4
|
1228 1229 1230 |
} mutex_unlock(&bdev->bd_mutex); |
6a027eff6
|
1231 |
bdput(whole); |
e525fd89d
|
1232 1233 1234 |
} return res; |
37be41241
|
1235 |
} |
1da177e4c
|
1236 |
EXPORT_SYMBOL(blkdev_get); |
d4d776299
|
1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 |
/** * blkdev_get_by_path - open a block device by name * @path: path to the block device to open * @mode: FMODE_* mask * @holder: exclusive holder identifier * * Open the blockdevice described by the device file at @path. @mode * and @holder are identical to blkdev_get(). * * On success, the returned block_device has reference count of one. * * CONTEXT: * Might sleep. * * RETURNS: * Pointer to block_device on success, ERR_PTR(-errno) on failure. */ struct block_device *blkdev_get_by_path(const char *path, fmode_t mode, void *holder) { struct block_device *bdev; int err; bdev = lookup_bdev(path); if (IS_ERR(bdev)) return bdev; err = blkdev_get(bdev, mode, holder); if (err) return ERR_PTR(err); |
e51900f7d
|
1267 1268 1269 1270 |
if ((mode & FMODE_WRITE) && bdev_read_only(bdev)) { blkdev_put(bdev, mode); return ERR_PTR(-EACCES); } |
d4d776299
|
1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 |
return bdev; } EXPORT_SYMBOL(blkdev_get_by_path); /** * blkdev_get_by_dev - open a block device by device number * @dev: device number of block device to open * @mode: FMODE_* mask * @holder: exclusive holder identifier * * Open the blockdevice described by device number @dev. @mode and * @holder are identical to blkdev_get(). * * Use it ONLY if you really do not have anything better - i.e. when * you are behind a truly sucky interface and all you are given is a * device number. _Never_ to be used for internal purposes. If you * ever need it - reconsider your API. * * On success, the returned block_device has reference count of one. * * CONTEXT: * Might sleep. * * RETURNS: * Pointer to block_device on success, ERR_PTR(-errno) on failure. */ struct block_device *blkdev_get_by_dev(dev_t dev, fmode_t mode, void *holder) { struct block_device *bdev; int err; bdev = bdget(dev); if (!bdev) return ERR_PTR(-ENOMEM); err = blkdev_get(bdev, mode, holder); if (err) return ERR_PTR(err); return bdev; } EXPORT_SYMBOL(blkdev_get_by_dev); |
1da177e4c
|
1313 1314 1315 |
static int blkdev_open(struct inode * inode, struct file * filp) { struct block_device *bdev; |
1da177e4c
|
1316 1317 1318 1319 1320 1321 1322 1323 |
/* * Preserve backwards compatibility and allow large file access * even if userspace doesn't ask for it explicitly. Some mkfs * binary needs it. We might want to drop this workaround * during an unstable branch. */ filp->f_flags |= O_LARGEFILE; |
572c48921
|
1324 1325 1326 1327 1328 1329 |
if (filp->f_flags & O_NDELAY) filp->f_mode |= FMODE_NDELAY; if (filp->f_flags & O_EXCL) filp->f_mode |= FMODE_EXCL; if ((filp->f_flags & O_ACCMODE) == 3) filp->f_mode |= FMODE_WRITE_IOCTL; |
1da177e4c
|
1330 |
bdev = bd_acquire(inode); |
6a2aae06c
|
1331 1332 |
if (bdev == NULL) return -ENOMEM; |
1da177e4c
|
1333 |
|
572c48921
|
1334 |
filp->f_mapping = bdev->bd_inode->i_mapping; |
e525fd89d
|
1335 |
return blkdev_get(bdev, filp->f_mode, filp); |
1da177e4c
|
1336 |
} |
4385bab12
|
1337 |
static void __blkdev_put(struct block_device *bdev, fmode_t mode, int for_part) |
2e7b651df
|
1338 |
{ |
2e7b651df
|
1339 |
struct gendisk *disk = bdev->bd_disk; |
37be41241
|
1340 |
struct block_device *victim = NULL; |
2e7b651df
|
1341 |
|
6796bf54a
|
1342 |
mutex_lock_nested(&bdev->bd_mutex, for_part); |
37be41241
|
1343 1344 |
if (for_part) bdev->bd_part_count--; |
2e7b651df
|
1345 |
if (!--bdev->bd_openers) { |
6a027eff6
|
1346 |
WARN_ON_ONCE(bdev->bd_holders); |
2e7b651df
|
1347 1348 |
sync_blockdev(bdev); kill_bdev(bdev); |
94007751b
|
1349 1350 1351 1352 1353 |
/* ->release can cause the old bdi to disappear, * so must switch it out first */ bdev_inode_switch_bdi(bdev->bd_inode, &default_backing_dev_info); |
2e7b651df
|
1354 1355 1356 |
} if (bdev->bd_contains == bdev) { if (disk->fops->release) |
db2a144be
|
1357 |
disk->fops->release(disk, mode); |
2e7b651df
|
1358 1359 1360 |
} if (!bdev->bd_openers) { struct module *owner = disk->fops->owner; |
0762b8bde
|
1361 1362 |
disk_put_part(bdev->bd_part); bdev->bd_part = NULL; |
2e7b651df
|
1363 |
bdev->bd_disk = NULL; |
37be41241
|
1364 1365 |
if (bdev != bdev->bd_contains) victim = bdev->bd_contains; |
2e7b651df
|
1366 |
bdev->bd_contains = NULL; |
523e1d399
|
1367 1368 1369 |
put_disk(disk); module_put(owner); |
2e7b651df
|
1370 |
} |
2e7b651df
|
1371 1372 |
mutex_unlock(&bdev->bd_mutex); bdput(bdev); |
37be41241
|
1373 |
if (victim) |
9a1c35427
|
1374 |
__blkdev_put(victim, mode, 1); |
2e7b651df
|
1375 |
} |
4385bab12
|
1376 |
void blkdev_put(struct block_device *bdev, fmode_t mode) |
37be41241
|
1377 |
{ |
85ef06d1d
|
1378 |
mutex_lock(&bdev->bd_mutex); |
e525fd89d
|
1379 |
if (mode & FMODE_EXCL) { |
6a027eff6
|
1380 1381 1382 1383 1384 1385 1386 |
bool bdev_free; /* * Release a claim on the device. The holder fields * are protected with bdev_lock. bd_mutex is to * synchronize disk_holder unlinking. */ |
6a027eff6
|
1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 |
spin_lock(&bdev_lock); WARN_ON_ONCE(--bdev->bd_holders < 0); WARN_ON_ONCE(--bdev->bd_contains->bd_holders < 0); /* bd_contains might point to self, check in a separate step */ if ((bdev_free = !bdev->bd_holders)) bdev->bd_holder = NULL; if (!bdev->bd_contains->bd_holders) bdev->bd_contains->bd_holder = NULL; spin_unlock(&bdev_lock); |
77ea887e4
|
1399 1400 1401 1402 |
/* * If this was the last claim, remove holder link and * unblock evpoll if it was a write holder. */ |
85ef06d1d
|
1403 1404 1405 |
if (bdev_free && bdev->bd_write_holder) { disk_unblock_events(bdev->bd_disk); bdev->bd_write_holder = false; |
77ea887e4
|
1406 |
} |
6936217cc
|
1407 |
} |
77ea887e4
|
1408 |
|
85ef06d1d
|
1409 1410 1411 1412 1413 1414 1415 1416 |
/* * Trigger event checking and tell drivers to flush MEDIA_CHANGE * event. This is to ensure detection of media removal commanded * from userland - e.g. eject(1). */ disk_flush_events(bdev->bd_disk, DISK_EVENT_MEDIA_CHANGE); mutex_unlock(&bdev->bd_mutex); |
4385bab12
|
1417 |
__blkdev_put(bdev, mode, 0); |
37be41241
|
1418 |
} |
2e7b651df
|
1419 |
EXPORT_SYMBOL(blkdev_put); |
1da177e4c
|
1420 1421 1422 |
static int blkdev_close(struct inode * inode, struct file * filp) { struct block_device *bdev = I_BDEV(filp->f_mapping->host); |
4385bab12
|
1423 1424 |
blkdev_put(bdev, filp->f_mode); return 0; |
1da177e4c
|
1425 |
} |
bb93e3a52
|
1426 |
static long block_ioctl(struct file *file, unsigned cmd, unsigned long arg) |
1da177e4c
|
1427 |
{ |
56b26add0
|
1428 1429 |
struct block_device *bdev = I_BDEV(file->f_mapping->host); fmode_t mode = file->f_mode; |
fd4ce1acd
|
1430 1431 1432 1433 1434 |
/* * O_NDELAY can be altered using fcntl(.., F_SETFL, ..), so we have * to updated it before every ioctl. */ |
56b26add0
|
1435 |
if (file->f_flags & O_NDELAY) |
fd4ce1acd
|
1436 1437 1438 |
mode |= FMODE_NDELAY; else mode &= ~FMODE_NDELAY; |
56b26add0
|
1439 |
return blkdev_ioctl(bdev, mode, cmd, arg); |
1da177e4c
|
1440 |
} |
87d8fe1ee
|
1441 |
/* |
eef993806
|
1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 |
* Write data to the block device. Only intended for the block device itself * and the raw driver which basically is a fake block device. * * Does not take i_mutex for the write and thus is not for general purpose * use. */ ssize_t blkdev_aio_write(struct kiocb *iocb, const struct iovec *iov, unsigned long nr_segs, loff_t pos) { struct file *file = iocb->ki_filp; |
53362a05a
|
1452 |
struct blk_plug plug; |
eef993806
|
1453 1454 1455 |
ssize_t ret; BUG_ON(iocb->ki_pos != pos); |
53362a05a
|
1456 |
blk_start_plug(&plug); |
eef993806
|
1457 1458 1459 1460 1461 1462 1463 1464 |
ret = __generic_file_aio_write(iocb, iov, nr_segs, &iocb->ki_pos); if (ret > 0 || ret == -EIOCBQUEUED) { ssize_t err; err = generic_write_sync(file, pos, ret); if (err < 0 && ret > 0) ret = err; } |
53362a05a
|
1465 |
blk_finish_plug(&plug); |
eef993806
|
1466 1467 1468 |
return ret; } EXPORT_SYMBOL_GPL(blkdev_aio_write); |
684c9aaeb
|
1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 |
static ssize_t blkdev_aio_read(struct kiocb *iocb, const struct iovec *iov, unsigned long nr_segs, loff_t pos) { struct file *file = iocb->ki_filp; struct inode *bd_inode = file->f_mapping->host; loff_t size = i_size_read(bd_inode); if (pos >= size) return 0; size -= pos; |
6f8f5c260
|
1480 |
if (size < iocb->ki_left) |
684c9aaeb
|
1481 1482 1483 |
nr_segs = iov_shorten((struct iovec *)iov, nr_segs, size); return generic_file_aio_read(iocb, iov, nr_segs, pos); } |
eef993806
|
1484 |
/* |
87d8fe1ee
|
1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 |
* Try to release a page associated with block device when the system * is under memory pressure. */ static int blkdev_releasepage(struct page *page, gfp_t wait) { struct super_block *super = BDEV_I(page->mapping->host)->bdev.bd_super; if (super && super->s_op->bdev_try_to_free_page) return super->s_op->bdev_try_to_free_page(super, page, wait); return try_to_free_buffers(page); } |
4c54ac62d
|
1497 |
static const struct address_space_operations def_blk_aops = { |
1da177e4c
|
1498 1499 |
.readpage = blkdev_readpage, .writepage = blkdev_writepage, |
6272b5a58
|
1500 1501 |
.write_begin = blkdev_write_begin, .write_end = blkdev_write_end, |
1da177e4c
|
1502 |
.writepages = generic_writepages, |
87d8fe1ee
|
1503 |
.releasepage = blkdev_releasepage, |
1da177e4c
|
1504 |
.direct_IO = blkdev_direct_IO, |
b45972265
|
1505 |
.is_dirty_writeback = buffer_check_dirty_writeback, |
1da177e4c
|
1506 |
}; |
4b6f5d20b
|
1507 |
const struct file_operations def_blk_fops = { |
1da177e4c
|
1508 1509 1510 |
.open = blkdev_open, .release = blkdev_close, .llseek = block_llseek, |
543ade1fc
|
1511 1512 |
.read = do_sync_read, .write = do_sync_write, |
684c9aaeb
|
1513 |
.aio_read = blkdev_aio_read, |
eef993806
|
1514 |
.aio_write = blkdev_aio_write, |
1e8b33328
|
1515 |
.mmap = generic_file_mmap, |
b1dd3b284
|
1516 |
.fsync = blkdev_fsync, |
bb93e3a52
|
1517 |
.unlocked_ioctl = block_ioctl, |
1da177e4c
|
1518 1519 1520 |
#ifdef CONFIG_COMPAT .compat_ioctl = compat_blkdev_ioctl, #endif |
1e8b33328
|
1521 1522 |
.splice_read = generic_file_splice_read, .splice_write = generic_file_splice_write, |
1da177e4c
|
1523 1524 1525 1526 1527 1528 1529 |
}; int ioctl_by_bdev(struct block_device *bdev, unsigned cmd, unsigned long arg) { int res; mm_segment_t old_fs = get_fs(); set_fs(KERNEL_DS); |
56b26add0
|
1530 |
res = blkdev_ioctl(bdev, 0, cmd, arg); |
1da177e4c
|
1531 1532 1533 1534 1535 1536 1537 1538 |
set_fs(old_fs); return res; } EXPORT_SYMBOL(ioctl_by_bdev); /** * lookup_bdev - lookup a struct block_device by name |
94e2959e7
|
1539 |
* @pathname: special file representing the block device |
1da177e4c
|
1540 |
* |
57d1b5366
|
1541 |
* Get a reference to the blockdevice at @pathname in the current |
1da177e4c
|
1542 1543 1544 |
* namespace if possible and return it. Return ERR_PTR(error) * otherwise. */ |
421748ecd
|
1545 |
struct block_device *lookup_bdev(const char *pathname) |
1da177e4c
|
1546 1547 1548 |
{ struct block_device *bdev; struct inode *inode; |
421748ecd
|
1549 |
struct path path; |
1da177e4c
|
1550 |
int error; |
421748ecd
|
1551 |
if (!pathname || !*pathname) |
1da177e4c
|
1552 |
return ERR_PTR(-EINVAL); |
421748ecd
|
1553 |
error = kern_path(pathname, LOOKUP_FOLLOW, &path); |
1da177e4c
|
1554 1555 |
if (error) return ERR_PTR(error); |
421748ecd
|
1556 |
inode = path.dentry->d_inode; |
1da177e4c
|
1557 1558 1559 1560 |
error = -ENOTBLK; if (!S_ISBLK(inode->i_mode)) goto fail; error = -EACCES; |
421748ecd
|
1561 |
if (path.mnt->mnt_flags & MNT_NODEV) |
1da177e4c
|
1562 1563 1564 1565 1566 1567 |
goto fail; error = -ENOMEM; bdev = bd_acquire(inode); if (!bdev) goto fail; out: |
421748ecd
|
1568 |
path_put(&path); |
1da177e4c
|
1569 1570 1571 1572 1573 |
return bdev; fail: bdev = ERR_PTR(error); goto out; } |
d5686b444
|
1574 |
EXPORT_SYMBOL(lookup_bdev); |
1da177e4c
|
1575 |
|
93b270f76
|
1576 |
int __invalidate_device(struct block_device *bdev, bool kill_dirty) |
b71e8a4ce
|
1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 |
{ struct super_block *sb = get_super(bdev); int res = 0; if (sb) { /* * no need to lock the super, get_super holds the * read mutex so the filesystem cannot go away * under us (->put_super runs with the write lock * hold). */ shrink_dcache_sb(sb); |
93b270f76
|
1589 |
res = invalidate_inodes(sb, kill_dirty); |
b71e8a4ce
|
1590 1591 |
drop_super(sb); } |
f98393a64
|
1592 |
invalidate_bdev(bdev); |
b71e8a4ce
|
1593 1594 1595 |
return res; } EXPORT_SYMBOL(__invalidate_device); |
5c0d6b60a
|
1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 |
void iterate_bdevs(void (*func)(struct block_device *, void *), void *arg) { struct inode *inode, *old_inode = NULL; spin_lock(&inode_sb_list_lock); list_for_each_entry(inode, &blockdev_superblock->s_inodes, i_sb_list) { struct address_space *mapping = inode->i_mapping; spin_lock(&inode->i_lock); if (inode->i_state & (I_FREEING|I_WILL_FREE|I_NEW) || mapping->nrpages == 0) { spin_unlock(&inode->i_lock); continue; } __iget(inode); spin_unlock(&inode->i_lock); spin_unlock(&inode_sb_list_lock); /* * We hold a reference to 'inode' so it couldn't have been * removed from s_inodes list while we dropped the * inode_sb_list_lock. We cannot iput the inode now as we can * be holding the last reference and we cannot iput it under * inode_sb_list_lock. So we keep the reference and iput it * later. */ iput(old_inode); old_inode = inode; func(I_BDEV(inode), arg); spin_lock(&inode_sb_list_lock); } spin_unlock(&inode_sb_list_lock); iput(old_inode); } |