Blame view
fs/buffer.c
88.8 KB
1da177e4c Linux-2.6.12-rc2 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 |
/* * linux/fs/buffer.c * * Copyright (C) 1991, 1992, 2002 Linus Torvalds */ /* * Start bdflush() with kernel_thread not syscall - Paul Gortmaker, 12/95 * * Removed a lot of unnecessary code and simplified things now that * the buffer cache isn't our primary cache - Andrew Tridgell 12/96 * * Speed up hash, lru, and free list operations. Use gfp() for allocating * hash table, use SLAB cache for buffer heads. SMP threading. -DaveM * * Added 32k buffer block sizes - these are required older ARM systems. - RMK * * async buffer flushing, 1999 Andrea Arcangeli <andrea@suse.de> */ |
1da177e4c Linux-2.6.12-rc2 |
20 21 22 23 24 25 |
#include <linux/kernel.h> #include <linux/syscalls.h> #include <linux/fs.h> #include <linux/mm.h> #include <linux/percpu.h> #include <linux/slab.h> |
16f7e0fe2 [PATCH] capable/c... |
26 |
#include <linux/capability.h> |
1da177e4c Linux-2.6.12-rc2 |
27 28 29 30 |
#include <linux/blkdev.h> #include <linux/file.h> #include <linux/quotaops.h> #include <linux/highmem.h> |
630d9c472 fs: reduce the us... |
31 |
#include <linux/export.h> |
1da177e4c Linux-2.6.12-rc2 |
32 33 34 35 |
#include <linux/writeback.h> #include <linux/hash.h> #include <linux/suspend.h> #include <linux/buffer_head.h> |
55e829af0 [PATCH] io-accoun... |
36 |
#include <linux/task_io_accounting_ops.h> |
1da177e4c Linux-2.6.12-rc2 |
37 38 39 40 41 |
#include <linux/bio.h> #include <linux/notifier.h> #include <linux/cpu.h> #include <linux/bitops.h> #include <linux/mpage.h> |
fb1c8f93d [PATCH] spinlock ... |
42 |
#include <linux/bit_spinlock.h> |
5305cb830 block: add block_... |
43 |
#include <trace/events/block.h> |
1da177e4c Linux-2.6.12-rc2 |
44 45 |
static int fsync_buffers_list(spinlock_t *lock, struct list_head *list); |
1da177e4c Linux-2.6.12-rc2 |
46 47 |
#define BH_ENTRY(list) list_entry((list), struct buffer_head, b_assoc_buffers) |
a3f3c29cb fs/buffer.c: do n... |
48 |
void init_buffer(struct buffer_head *bh, bh_end_io_t *handler, void *private) |
1da177e4c Linux-2.6.12-rc2 |
49 50 51 52 |
{ bh->b_end_io = handler; bh->b_private = private; } |
1fe72eaa0 fs/buffer.c: clea... |
53 |
EXPORT_SYMBOL(init_buffer); |
1da177e4c Linux-2.6.12-rc2 |
54 |
|
f0059afd3 buffer: make touc... |
55 56 |
inline void touch_buffer(struct buffer_head *bh) { |
5305cb830 block: add block_... |
57 |
trace_block_touch_buffer(bh); |
f0059afd3 buffer: make touc... |
58 59 60 |
mark_page_accessed(bh->b_page); } EXPORT_SYMBOL(touch_buffer); |
7eaceacca block: remove per... |
61 |
static int sleep_on_buffer(void *word) |
1da177e4c Linux-2.6.12-rc2 |
62 |
{ |
1da177e4c Linux-2.6.12-rc2 |
63 64 65 |
io_schedule(); return 0; } |
fc9b52cd8 fs: remove fastca... |
66 |
void __lock_buffer(struct buffer_head *bh) |
1da177e4c Linux-2.6.12-rc2 |
67 |
{ |
7eaceacca block: remove per... |
68 |
wait_on_bit_lock(&bh->b_state, BH_Lock, sleep_on_buffer, |
1da177e4c Linux-2.6.12-rc2 |
69 70 71 |
TASK_UNINTERRUPTIBLE); } EXPORT_SYMBOL(__lock_buffer); |
fc9b52cd8 fs: remove fastca... |
72 |
void unlock_buffer(struct buffer_head *bh) |
1da177e4c Linux-2.6.12-rc2 |
73 |
{ |
51b07fc3c fs: buffer lock u... |
74 |
clear_bit_unlock(BH_Lock, &bh->b_state); |
1da177e4c Linux-2.6.12-rc2 |
75 76 77 |
smp_mb__after_clear_bit(); wake_up_bit(&bh->b_state, BH_Lock); } |
1fe72eaa0 fs/buffer.c: clea... |
78 |
EXPORT_SYMBOL(unlock_buffer); |
1da177e4c Linux-2.6.12-rc2 |
79 80 |
/* |
b45972265 mm: vmscan: take ... |
81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 |
* Returns if the page has dirty or writeback buffers. If all the buffers * are unlocked and clean then the PageDirty information is stale. If * any of the pages are locked, it is assumed they are locked for IO. */ void buffer_check_dirty_writeback(struct page *page, bool *dirty, bool *writeback) { struct buffer_head *head, *bh; *dirty = false; *writeback = false; BUG_ON(!PageLocked(page)); if (!page_has_buffers(page)) return; if (PageWriteback(page)) *writeback = true; head = page_buffers(page); bh = head; do { if (buffer_locked(bh)) *writeback = true; if (buffer_dirty(bh)) *dirty = true; bh = bh->b_this_page; } while (bh != head); } EXPORT_SYMBOL(buffer_check_dirty_writeback); /* |
1da177e4c Linux-2.6.12-rc2 |
115 116 117 118 119 120 |
* Block until a buffer comes unlocked. This doesn't stop it * from becoming locked again - you have to lock it yourself * if you want to preserve its state. */ void __wait_on_buffer(struct buffer_head * bh) { |
7eaceacca block: remove per... |
121 |
wait_on_bit(&bh->b_state, BH_Lock, sleep_on_buffer, TASK_UNINTERRUPTIBLE); |
1da177e4c Linux-2.6.12-rc2 |
122 |
} |
1fe72eaa0 fs/buffer.c: clea... |
123 |
EXPORT_SYMBOL(__wait_on_buffer); |
1da177e4c Linux-2.6.12-rc2 |
124 125 126 127 128 |
static void __clear_page_buffers(struct page *page) { ClearPagePrivate(page); |
4c21e2f24 [PATCH] mm: split... |
129 |
set_page_private(page, 0); |
1da177e4c Linux-2.6.12-rc2 |
130 131 |
page_cache_release(page); } |
08bafc034 block: Supress Bu... |
132 133 134 135 136 137 138 |
static int quiet_error(struct buffer_head *bh) { if (!test_bit(BH_Quiet, &bh->b_state) && printk_ratelimit()) return 0; return 1; } |
1da177e4c Linux-2.6.12-rc2 |
139 140 141 |
static void buffer_io_error(struct buffer_head *bh) { char b[BDEVNAME_SIZE]; |
1da177e4c Linux-2.6.12-rc2 |
142 143 144 145 146 147 148 |
printk(KERN_ERR "Buffer I/O error on device %s, logical block %Lu ", bdevname(bh->b_bdev, b), (unsigned long long)bh->b_blocknr); } /* |
68671f35f mm: add end_buffe... |
149 150 151 152 153 154 |
* End-of-IO handler helper function which does not touch the bh after * unlocking it. * Note: unlock_buffer() sort-of does touch the bh after unlocking it, but * a race there is benign: unlock_buffer() only use the bh's address for * hashing after unlocking the buffer, so it doesn't actually touch the bh * itself. |
1da177e4c Linux-2.6.12-rc2 |
155 |
*/ |
68671f35f mm: add end_buffe... |
156 |
static void __end_buffer_read_notouch(struct buffer_head *bh, int uptodate) |
1da177e4c Linux-2.6.12-rc2 |
157 158 159 160 161 162 163 164 |
{ if (uptodate) { set_buffer_uptodate(bh); } else { /* This happens, due to failed READA attempts. */ clear_buffer_uptodate(bh); } unlock_buffer(bh); |
68671f35f mm: add end_buffe... |
165 166 167 168 169 170 171 172 173 |
} /* * Default synchronous end-of-IO handler.. Just mark it up-to-date and * unlock the buffer. This is what ll_rw_block uses too. */ void end_buffer_read_sync(struct buffer_head *bh, int uptodate) { __end_buffer_read_notouch(bh, uptodate); |
1da177e4c Linux-2.6.12-rc2 |
174 175 |
put_bh(bh); } |
1fe72eaa0 fs/buffer.c: clea... |
176 |
EXPORT_SYMBOL(end_buffer_read_sync); |
1da177e4c Linux-2.6.12-rc2 |
177 178 179 180 181 182 183 184 |
void end_buffer_write_sync(struct buffer_head *bh, int uptodate) { char b[BDEVNAME_SIZE]; if (uptodate) { set_buffer_uptodate(bh); } else { |
0edd55fae block: remove the... |
185 |
if (!quiet_error(bh)) { |
1da177e4c Linux-2.6.12-rc2 |
186 187 188 189 190 191 192 193 194 195 196 197 |
buffer_io_error(bh); printk(KERN_WARNING "lost page write due to " "I/O error on %s ", bdevname(bh->b_bdev, b)); } set_buffer_write_io_error(bh); clear_buffer_uptodate(bh); } unlock_buffer(bh); put_bh(bh); } |
1fe72eaa0 fs/buffer.c: clea... |
198 |
EXPORT_SYMBOL(end_buffer_write_sync); |
1da177e4c Linux-2.6.12-rc2 |
199 200 |
/* |
1da177e4c Linux-2.6.12-rc2 |
201 202 203 204 205 206 207 208 209 210 211 |
* Various filesystems appear to want __find_get_block to be non-blocking. * But it's the page lock which protects the buffers. To get around this, * we get exclusion from try_to_free_buffers with the blockdev mapping's * private_lock. * * Hack idea: for the blockdev mapping, i_bufferlist_lock contention * may be quite high. This code could TryLock the page, and if that * succeeds, there is no need to take private_lock. (But if * private_lock is contended then so is mapping->tree_lock). */ static struct buffer_head * |
385fd4c59 [PATCH] __find_ge... |
212 |
__find_get_block_slow(struct block_device *bdev, sector_t block) |
1da177e4c Linux-2.6.12-rc2 |
213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 |
{ struct inode *bd_inode = bdev->bd_inode; struct address_space *bd_mapping = bd_inode->i_mapping; struct buffer_head *ret = NULL; pgoff_t index; struct buffer_head *bh; struct buffer_head *head; struct page *page; int all_mapped = 1; index = block >> (PAGE_CACHE_SHIFT - bd_inode->i_blkbits); page = find_get_page(bd_mapping, index); if (!page) goto out; spin_lock(&bd_mapping->private_lock); if (!page_has_buffers(page)) goto out_unlock; head = page_buffers(page); bh = head; do { |
97f76d3d1 vfs: check bh->b_... |
234 235 236 |
if (!buffer_mapped(bh)) all_mapped = 0; else if (bh->b_blocknr == block) { |
1da177e4c Linux-2.6.12-rc2 |
237 238 239 240 |
ret = bh; get_bh(bh); goto out_unlock; } |
1da177e4c Linux-2.6.12-rc2 |
241 242 243 244 245 246 247 248 249 |
bh = bh->b_this_page; } while (bh != head); /* we might be here because some of the buffers on this page are * not mapped. This is due to various races between * file io on the block device and getblk. It gets dealt with * elsewhere, don't buffer_error if we had some unmapped buffers */ if (all_mapped) { |
72a2ebd8b fs/buffer.c: add ... |
250 |
char b[BDEVNAME_SIZE]; |
1da177e4c Linux-2.6.12-rc2 |
251 252 253 |
printk("__find_get_block_slow() failed. " "block=%llu, b_blocknr=%llu ", |
205f87f6b [PATCH] change bu... |
254 255 256 257 258 |
(unsigned long long)block, (unsigned long long)bh->b_blocknr); printk("b_state=0x%08lx, b_size=%zu ", bh->b_state, bh->b_size); |
72a2ebd8b fs/buffer.c: add ... |
259 260 261 |
printk("device %s blocksize: %d ", bdevname(bdev, b), 1 << bd_inode->i_blkbits); |
1da177e4c Linux-2.6.12-rc2 |
262 263 264 265 266 267 268 |
} out_unlock: spin_unlock(&bd_mapping->private_lock); page_cache_release(page); out: return ret; } |
1da177e4c Linux-2.6.12-rc2 |
269 |
/* |
5b0830cb9 writeback: get ri... |
270 |
* Kick the writeback threads then try to free up some ZONE_NORMAL memory. |
1da177e4c Linux-2.6.12-rc2 |
271 272 273 |
*/ static void free_more_memory(void) { |
19770b326 mm: filter based ... |
274 |
struct zone *zone; |
0e88460da mm: introduce nod... |
275 |
int nid; |
1da177e4c Linux-2.6.12-rc2 |
276 |
|
0e175a183 writeback: Add a ... |
277 |
wakeup_flusher_threads(1024, WB_REASON_FREE_MORE_MEM); |
1da177e4c Linux-2.6.12-rc2 |
278 |
yield(); |
0e88460da mm: introduce nod... |
279 |
for_each_online_node(nid) { |
19770b326 mm: filter based ... |
280 281 282 283 |
(void)first_zones_zonelist(node_zonelist(nid, GFP_NOFS), gfp_zone(GFP_NOFS), NULL, &zone); if (zone) |
54a6eb5c4 mm: use two zonel... |
284 |
try_to_free_pages(node_zonelist(nid, GFP_NOFS), 0, |
327c0e968 vmscan: fix it to... |
285 |
GFP_NOFS, NULL); |
1da177e4c Linux-2.6.12-rc2 |
286 287 288 289 290 291 292 293 294 |
} } /* * I/O completion handler for block_read_full_page() - pages * which come unlocked at the end of I/O. */ static void end_buffer_async_read(struct buffer_head *bh, int uptodate) { |
1da177e4c Linux-2.6.12-rc2 |
295 |
unsigned long flags; |
a39722034 [PATCH] page_upto... |
296 |
struct buffer_head *first; |
1da177e4c Linux-2.6.12-rc2 |
297 298 299 300 301 302 303 304 305 306 307 |
struct buffer_head *tmp; struct page *page; int page_uptodate = 1; BUG_ON(!buffer_async_read(bh)); page = bh->b_page; if (uptodate) { set_buffer_uptodate(bh); } else { clear_buffer_uptodate(bh); |
08bafc034 block: Supress Bu... |
308 |
if (!quiet_error(bh)) |
1da177e4c Linux-2.6.12-rc2 |
309 310 311 312 313 314 315 316 317 |
buffer_io_error(bh); SetPageError(page); } /* * Be _very_ careful from here on. Bad things can happen if * two buffer heads end IO at almost the same time and both * decide that the page is now completely done. */ |
a39722034 [PATCH] page_upto... |
318 319 320 |
first = page_buffers(page); local_irq_save(flags); bit_spin_lock(BH_Uptodate_Lock, &first->b_state); |
1da177e4c Linux-2.6.12-rc2 |
321 322 323 324 325 326 327 328 329 330 331 332 |
clear_buffer_async_read(bh); unlock_buffer(bh); tmp = bh; do { if (!buffer_uptodate(tmp)) page_uptodate = 0; if (buffer_async_read(tmp)) { BUG_ON(!buffer_locked(tmp)); goto still_busy; } tmp = tmp->b_this_page; } while (tmp != bh); |
a39722034 [PATCH] page_upto... |
333 334 |
bit_spin_unlock(BH_Uptodate_Lock, &first->b_state); local_irq_restore(flags); |
1da177e4c Linux-2.6.12-rc2 |
335 336 337 338 339 340 341 342 343 344 345 |
/* * If none of the buffers had errors and they are all * uptodate then we can set the page uptodate. */ if (page_uptodate && !PageError(page)) SetPageUptodate(page); unlock_page(page); return; still_busy: |
a39722034 [PATCH] page_upto... |
346 347 |
bit_spin_unlock(BH_Uptodate_Lock, &first->b_state); local_irq_restore(flags); |
1da177e4c Linux-2.6.12-rc2 |
348 349 350 351 352 353 354 |
return; } /* * Completion handler for block_write_full_page() - pages which are unlocked * during I/O, and which have PageWriteback cleared upon I/O completion. */ |
35c80d5f4 Add block_write_f... |
355 |
void end_buffer_async_write(struct buffer_head *bh, int uptodate) |
1da177e4c Linux-2.6.12-rc2 |
356 357 |
{ char b[BDEVNAME_SIZE]; |
1da177e4c Linux-2.6.12-rc2 |
358 |
unsigned long flags; |
a39722034 [PATCH] page_upto... |
359 |
struct buffer_head *first; |
1da177e4c Linux-2.6.12-rc2 |
360 361 362 363 364 365 366 367 368 |
struct buffer_head *tmp; struct page *page; BUG_ON(!buffer_async_write(bh)); page = bh->b_page; if (uptodate) { set_buffer_uptodate(bh); } else { |
08bafc034 block: Supress Bu... |
369 |
if (!quiet_error(bh)) { |
1da177e4c Linux-2.6.12-rc2 |
370 371 372 373 374 375 376 |
buffer_io_error(bh); printk(KERN_WARNING "lost page write due to " "I/O error on %s ", bdevname(bh->b_bdev, b)); } set_bit(AS_EIO, &page->mapping->flags); |
58ff407be [PATCH] Fix IO er... |
377 |
set_buffer_write_io_error(bh); |
1da177e4c Linux-2.6.12-rc2 |
378 379 380 |
clear_buffer_uptodate(bh); SetPageError(page); } |
a39722034 [PATCH] page_upto... |
381 382 383 |
first = page_buffers(page); local_irq_save(flags); bit_spin_lock(BH_Uptodate_Lock, &first->b_state); |
1da177e4c Linux-2.6.12-rc2 |
384 385 386 387 388 389 390 391 392 393 |
clear_buffer_async_write(bh); unlock_buffer(bh); tmp = bh->b_this_page; while (tmp != bh) { if (buffer_async_write(tmp)) { BUG_ON(!buffer_locked(tmp)); goto still_busy; } tmp = tmp->b_this_page; } |
a39722034 [PATCH] page_upto... |
394 395 |
bit_spin_unlock(BH_Uptodate_Lock, &first->b_state); local_irq_restore(flags); |
1da177e4c Linux-2.6.12-rc2 |
396 397 398 399 |
end_page_writeback(page); return; still_busy: |
a39722034 [PATCH] page_upto... |
400 401 |
bit_spin_unlock(BH_Uptodate_Lock, &first->b_state); local_irq_restore(flags); |
1da177e4c Linux-2.6.12-rc2 |
402 403 |
return; } |
1fe72eaa0 fs/buffer.c: clea... |
404 |
EXPORT_SYMBOL(end_buffer_async_write); |
1da177e4c Linux-2.6.12-rc2 |
405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 |
/* * If a page's buffers are under async readin (end_buffer_async_read * completion) then there is a possibility that another thread of * control could lock one of the buffers after it has completed * but while some of the other buffers have not completed. This * locked buffer would confuse end_buffer_async_read() into not unlocking * the page. So the absence of BH_Async_Read tells end_buffer_async_read() * that this buffer is not under async I/O. * * The page comes unlocked when it has no locked buffer_async buffers * left. * * PageLocked prevents anyone starting new async I/O reads any of * the buffers. * * PageWriteback is used to prevent simultaneous writeout of the same * page. * * PageLocked prevents anyone from starting writeback of a page which is * under read I/O (PageWriteback is only ever set against a locked page). */ static void mark_buffer_async_read(struct buffer_head *bh) { bh->b_end_io = end_buffer_async_read; set_buffer_async_read(bh); } |
1fe72eaa0 fs/buffer.c: clea... |
432 433 |
static void mark_buffer_async_write_endio(struct buffer_head *bh, bh_end_io_t *handler) |
1da177e4c Linux-2.6.12-rc2 |
434 |
{ |
35c80d5f4 Add block_write_f... |
435 |
bh->b_end_io = handler; |
1da177e4c Linux-2.6.12-rc2 |
436 437 |
set_buffer_async_write(bh); } |
35c80d5f4 Add block_write_f... |
438 439 440 441 442 |
void mark_buffer_async_write(struct buffer_head *bh) { mark_buffer_async_write_endio(bh, end_buffer_async_write); } |
1da177e4c Linux-2.6.12-rc2 |
443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 |
EXPORT_SYMBOL(mark_buffer_async_write); /* * fs/buffer.c contains helper functions for buffer-backed address space's * fsync functions. A common requirement for buffer-based filesystems is * that certain data from the backing blockdev needs to be written out for * a successful fsync(). For example, ext2 indirect blocks need to be * written back and waited upon before fsync() returns. * * The functions mark_buffer_inode_dirty(), fsync_inode_buffers(), * inode_has_buffers() and invalidate_inode_buffers() are provided for the * management of a list of dependent buffers at ->i_mapping->private_list. * * Locking is a little subtle: try_to_free_buffers() will remove buffers * from their controlling inode's queue when they are being freed. But * try_to_free_buffers() will be operating against the *blockdev* mapping * at the time, not against the S_ISREG file which depends on those buffers. * So the locking for private_list is via the private_lock in the address_space * which backs the buffers. Which is different from the address_space * against which the buffers are listed. So for a particular address_space, * mapping->private_lock does *not* protect mapping->private_list! In fact, * mapping->private_list will always be protected by the backing blockdev's * ->private_lock. * * Which introduces a requirement: all buffers on an address_space's * ->private_list must be from the same address_space: the blockdev's. * * address_spaces which do not place buffers at ->private_list via these * utility functions are free to use private_lock and private_list for * whatever they want. The only requirement is that list_empty(private_list) * be true at clear_inode() time. * * FIXME: clear_inode should not call invalidate_inode_buffers(). The * filesystems should do that. invalidate_inode_buffers() should just go * BUG_ON(!list_empty). * * FIXME: mark_buffer_dirty_inode() is a data-plane operation. It should * take an address_space, not an inode. And it should be called * mark_buffer_dirty_fsync() to clearly define why those buffers are being * queued up. * * FIXME: mark_buffer_dirty_inode() doesn't need to add the buffer to the * list if it is already on a list. Because if the buffer is on a list, * it *must* already be on the right one. If not, the filesystem is being * silly. This will save a ton of locking. But first we have to ensure * that buffers are taken *off* the old inode's list when they are freed * (presumably in truncate). That requires careful auditing of all * filesystems (do it inside bforget()). It could also be done by bringing * b_inode back. */ /* * The buffer's backing address_space's private_lock must be held */ |
dbacefc9c fs/buffer.c: unin... |
498 |
static void __remove_assoc_queue(struct buffer_head *bh) |
1da177e4c Linux-2.6.12-rc2 |
499 500 |
{ list_del_init(&bh->b_assoc_buffers); |
58ff407be [PATCH] Fix IO er... |
501 502 503 504 |
WARN_ON(!bh->b_assoc_map); if (buffer_write_io_error(bh)) set_bit(AS_EIO, &bh->b_assoc_map->flags); bh->b_assoc_map = NULL; |
1da177e4c Linux-2.6.12-rc2 |
505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 |
} int inode_has_buffers(struct inode *inode) { return !list_empty(&inode->i_data.private_list); } /* * osync is designed to support O_SYNC io. It waits synchronously for * all already-submitted IO to complete, but does not queue any new * writes to the disk. * * To do O_SYNC writes, just queue the buffer writes with ll_rw_block as * you dirty the buffers, and then use osync_inode_buffers to wait for * completion. Any other dirty buffers which are not yet queued for * write will not be flushed to disk by the osync. */ static int osync_buffers_list(spinlock_t *lock, struct list_head *list) { struct buffer_head *bh; struct list_head *p; int err = 0; spin_lock(lock); repeat: list_for_each_prev(p, list) { bh = BH_ENTRY(p); if (buffer_locked(bh)) { get_bh(bh); spin_unlock(lock); wait_on_buffer(bh); if (!buffer_uptodate(bh)) err = -EIO; brelse(bh); spin_lock(lock); goto repeat; } } spin_unlock(lock); return err; } |
01a05b337 new helper: itera... |
546 |
static void do_thaw_one(struct super_block *sb, void *unused) |
c2d754385 filesystem freeze... |
547 |
{ |
c2d754385 filesystem freeze... |
548 |
char b[BDEVNAME_SIZE]; |
01a05b337 new helper: itera... |
549 550 551 552 553 |
while (sb->s_bdev && !thaw_bdev(sb->s_bdev, sb)) printk(KERN_WARNING "Emergency Thaw on %s ", bdevname(sb->s_bdev, b)); } |
c2d754385 filesystem freeze... |
554 |
|
01a05b337 new helper: itera... |
555 556 557 |
static void do_thaw_all(struct work_struct *work) { iterate_supers(do_thaw_one, NULL); |
053c525fc buffer: switch do... |
558 |
kfree(work); |
c2d754385 filesystem freeze... |
559 560 561 562 563 564 565 566 567 568 569 |
printk(KERN_WARNING "Emergency Thaw complete "); } /** * emergency_thaw_all -- forcibly thaw every frozen filesystem * * Used for emergency unfreeze of all filesystems via SysRq */ void emergency_thaw_all(void) { |
053c525fc buffer: switch do... |
570 571 572 573 574 575 576 |
struct work_struct *work; work = kmalloc(sizeof(*work), GFP_ATOMIC); if (work) { INIT_WORK(work, do_thaw_all); schedule_work(work); } |
c2d754385 filesystem freeze... |
577 |
} |
1da177e4c Linux-2.6.12-rc2 |
578 |
/** |
78a4a50a8 docbook: fix file... |
579 |
* sync_mapping_buffers - write out & wait upon a mapping's "associated" buffers |
67be2dd1b [PATCH] DocBook: ... |
580 |
* @mapping: the mapping which wants those buffers written |
1da177e4c Linux-2.6.12-rc2 |
581 582 583 584 |
* * Starts I/O against the buffers at mapping->private_list, and waits upon * that I/O. * |
67be2dd1b [PATCH] DocBook: ... |
585 586 587 |
* Basically, this is a convenience function for fsync(). * @mapping is a file or directory which needs those buffers to be written for * a successful fsync(). |
1da177e4c Linux-2.6.12-rc2 |
588 589 590 |
*/ int sync_mapping_buffers(struct address_space *mapping) { |
252aa6f5b mm: redefine addr... |
591 |
struct address_space *buffer_mapping = mapping->private_data; |
1da177e4c Linux-2.6.12-rc2 |
592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 |
if (buffer_mapping == NULL || list_empty(&mapping->private_list)) return 0; return fsync_buffers_list(&buffer_mapping->private_lock, &mapping->private_list); } EXPORT_SYMBOL(sync_mapping_buffers); /* * Called when we've recently written block `bblock', and it is known that * `bblock' was for a buffer_boundary() buffer. This means that the block at * `bblock + 1' is probably a dirty indirect block. Hunt it down and, if it's * dirty, schedule it for IO. So that indirects merge nicely with their data. */ void write_boundary_block(struct block_device *bdev, sector_t bblock, unsigned blocksize) { struct buffer_head *bh = __find_get_block(bdev, bblock + 1, blocksize); if (bh) { if (buffer_dirty(bh)) ll_rw_block(WRITE, 1, &bh); put_bh(bh); } } void mark_buffer_dirty_inode(struct buffer_head *bh, struct inode *inode) { struct address_space *mapping = inode->i_mapping; struct address_space *buffer_mapping = bh->b_page->mapping; mark_buffer_dirty(bh); |
252aa6f5b mm: redefine addr... |
624 625 |
if (!mapping->private_data) { mapping->private_data = buffer_mapping; |
1da177e4c Linux-2.6.12-rc2 |
626 |
} else { |
252aa6f5b mm: redefine addr... |
627 |
BUG_ON(mapping->private_data != buffer_mapping); |
1da177e4c Linux-2.6.12-rc2 |
628 |
} |
535ee2fbf buffer_head: fix ... |
629 |
if (!bh->b_assoc_map) { |
1da177e4c Linux-2.6.12-rc2 |
630 631 632 |
spin_lock(&buffer_mapping->private_lock); list_move_tail(&bh->b_assoc_buffers, &mapping->private_list); |
58ff407be [PATCH] Fix IO er... |
633 |
bh->b_assoc_map = mapping; |
1da177e4c Linux-2.6.12-rc2 |
634 635 636 637 638 639 |
spin_unlock(&buffer_mapping->private_lock); } } EXPORT_SYMBOL(mark_buffer_dirty_inode); /* |
787d2214c fs: introduce som... |
640 641 642 643 644 645 |
* Mark the page dirty, and set it dirty in the radix tree, and mark the inode * dirty. * * If warn is true, then emit a warning if the page is not uptodate and has * not been truncated. */ |
a8e7d49aa Fix race in creat... |
646 |
static void __set_page_dirty(struct page *page, |
787d2214c fs: introduce som... |
647 648 |
struct address_space *mapping, int warn) { |
227d53b39 mm: __set_page_di... |
649 650 651 |
unsigned long flags; spin_lock_irqsave(&mapping->tree_lock, flags); |
787d2214c fs: introduce som... |
652 653 |
if (page->mapping) { /* Race with truncate? */ WARN_ON_ONCE(warn && !PageUptodate(page)); |
e3a7cca1e vfs: add/use acco... |
654 |
account_page_dirtied(page, mapping); |
787d2214c fs: introduce som... |
655 656 657 |
radix_tree_tag_set(&mapping->page_tree, page_index(page), PAGECACHE_TAG_DIRTY); } |
227d53b39 mm: __set_page_di... |
658 |
spin_unlock_irqrestore(&mapping->tree_lock, flags); |
787d2214c fs: introduce som... |
659 |
__mark_inode_dirty(mapping->host, I_DIRTY_PAGES); |
787d2214c fs: introduce som... |
660 661 662 |
} /* |
1da177e4c Linux-2.6.12-rc2 |
663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 |
* Add a page to the dirty page list. * * It is a sad fact of life that this function is called from several places * deeply under spinlocking. It may not sleep. * * If the page has buffers, the uptodate buffers are set dirty, to preserve * dirty-state coherency between the page and the buffers. It the page does * not have buffers then when they are later attached they will all be set * dirty. * * The buffers are dirtied before the page is dirtied. There's a small race * window in which a writepage caller may see the page cleanness but not the * buffer dirtiness. That's fine. If this code were to set the page dirty * before the buffers, a concurrent writepage caller could clear the page dirty * bit, see a bunch of clean buffers and we'd end up with dirty buffers/clean * page on the dirty page list. * * We use private_lock to lock against try_to_free_buffers while using the * page's buffer list. Also use this to protect against clean buffers being * added to the page after it was set dirty. * * FIXME: may need to call ->reservepage here as well. That's rather up to the * address_space though. */ int __set_page_dirty_buffers(struct page *page) { |
a8e7d49aa Fix race in creat... |
689 |
int newly_dirty; |
787d2214c fs: introduce som... |
690 |
struct address_space *mapping = page_mapping(page); |
ebf7a227d [PATCH] mm: bug i... |
691 692 693 |
if (unlikely(!mapping)) return !TestSetPageDirty(page); |
1da177e4c Linux-2.6.12-rc2 |
694 695 696 697 698 699 700 701 702 703 704 |
spin_lock(&mapping->private_lock); if (page_has_buffers(page)) { struct buffer_head *head = page_buffers(page); struct buffer_head *bh = head; do { set_buffer_dirty(bh); bh = bh->b_this_page; } while (bh != head); } |
a8e7d49aa Fix race in creat... |
705 |
newly_dirty = !TestSetPageDirty(page); |
1da177e4c Linux-2.6.12-rc2 |
706 |
spin_unlock(&mapping->private_lock); |
a8e7d49aa Fix race in creat... |
707 708 709 |
if (newly_dirty) __set_page_dirty(page, mapping, 1); return newly_dirty; |
1da177e4c Linux-2.6.12-rc2 |
710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 |
} EXPORT_SYMBOL(__set_page_dirty_buffers); /* * Write out and wait upon a list of buffers. * * We have conflicting pressures: we want to make sure that all * initially dirty buffers get waited on, but that any subsequently * dirtied buffers don't. After all, we don't want fsync to last * forever if somebody is actively writing to the file. * * Do this in two main stages: first we copy dirty buffers to a * temporary inode list, queueing the writes as we go. Then we clean * up, waiting for those writes to complete. * * During this second stage, any subsequent updates to the file may end * up refiling the buffer on the original inode's dirty list again, so * there is a chance we will end up with a buffer queued for write but * not yet completed on that list. So, as a final cleanup we go through * the osync code to catch these locked, dirty buffers without requeuing * any newly dirty buffers for write. */ static int fsync_buffers_list(spinlock_t *lock, struct list_head *list) { struct buffer_head *bh; struct list_head tmp; |
7eaceacca block: remove per... |
736 |
struct address_space *mapping; |
1da177e4c Linux-2.6.12-rc2 |
737 |
int err = 0, err2; |
4ee2491ed fs: make fsync_bu... |
738 |
struct blk_plug plug; |
1da177e4c Linux-2.6.12-rc2 |
739 740 |
INIT_LIST_HEAD(&tmp); |
4ee2491ed fs: make fsync_bu... |
741 |
blk_start_plug(&plug); |
1da177e4c Linux-2.6.12-rc2 |
742 743 744 745 |
spin_lock(lock); while (!list_empty(list)) { bh = BH_ENTRY(list->next); |
535ee2fbf buffer_head: fix ... |
746 |
mapping = bh->b_assoc_map; |
58ff407be [PATCH] Fix IO er... |
747 |
__remove_assoc_queue(bh); |
535ee2fbf buffer_head: fix ... |
748 749 750 |
/* Avoid race with mark_buffer_dirty_inode() which does * a lockless check and we rely on seeing the dirty bit */ smp_mb(); |
1da177e4c Linux-2.6.12-rc2 |
751 752 |
if (buffer_dirty(bh) || buffer_locked(bh)) { list_add(&bh->b_assoc_buffers, &tmp); |
535ee2fbf buffer_head: fix ... |
753 |
bh->b_assoc_map = mapping; |
1da177e4c Linux-2.6.12-rc2 |
754 755 756 757 758 |
if (buffer_dirty(bh)) { get_bh(bh); spin_unlock(lock); /* * Ensure any pending I/O completes so that |
9cb569d60 remove SWRITE* I/... |
759 760 761 762 |
* write_dirty_buffer() actually writes the * current contents - it is a noop if I/O is * still in flight on potentially older * contents. |
1da177e4c Linux-2.6.12-rc2 |
763 |
*/ |
721a9602e block: kill off R... |
764 |
write_dirty_buffer(bh, WRITE_SYNC); |
9cf6b720f block: fsync_buff... |
765 766 767 768 769 770 771 |
/* * Kick off IO for the previous mapping. Note * that we will not run the very last mapping, * wait_on_buffer() will do that for us * through sync_buffer(). */ |
1da177e4c Linux-2.6.12-rc2 |
772 773 774 775 776 |
brelse(bh); spin_lock(lock); } } } |
4ee2491ed fs: make fsync_bu... |
777 778 779 |
spin_unlock(lock); blk_finish_plug(&plug); spin_lock(lock); |
1da177e4c Linux-2.6.12-rc2 |
780 781 |
while (!list_empty(&tmp)) { bh = BH_ENTRY(tmp.prev); |
1da177e4c Linux-2.6.12-rc2 |
782 |
get_bh(bh); |
535ee2fbf buffer_head: fix ... |
783 784 785 786 787 788 789 |
mapping = bh->b_assoc_map; __remove_assoc_queue(bh); /* Avoid race with mark_buffer_dirty_inode() which does * a lockless check and we rely on seeing the dirty bit */ smp_mb(); if (buffer_dirty(bh)) { list_add(&bh->b_assoc_buffers, |
e3892296d vfs: fix NULL poi... |
790 |
&mapping->private_list); |
535ee2fbf buffer_head: fix ... |
791 792 |
bh->b_assoc_map = mapping; } |
1da177e4c Linux-2.6.12-rc2 |
793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 |
spin_unlock(lock); wait_on_buffer(bh); if (!buffer_uptodate(bh)) err = -EIO; brelse(bh); spin_lock(lock); } spin_unlock(lock); err2 = osync_buffers_list(lock, list); if (err) return err; else return err2; } /* * Invalidate any and all dirty buffers on a given inode. We are * probably unmounting the fs, but that doesn't mean we have already * done a sync(). Just drop the buffers from the inode list. * * NOTE: we take the inode's blockdev's mapping's private_lock. Which * assumes that all the buffers are against the blockdev. Not true * for reiserfs. */ void invalidate_inode_buffers(struct inode *inode) { if (inode_has_buffers(inode)) { struct address_space *mapping = &inode->i_data; struct list_head *list = &mapping->private_list; |
252aa6f5b mm: redefine addr... |
823 |
struct address_space *buffer_mapping = mapping->private_data; |
1da177e4c Linux-2.6.12-rc2 |
824 825 826 827 828 829 830 |
spin_lock(&buffer_mapping->private_lock); while (!list_empty(list)) __remove_assoc_queue(BH_ENTRY(list->next)); spin_unlock(&buffer_mapping->private_lock); } } |
52b19ac99 udf: Fix BUG_ON()... |
831 |
EXPORT_SYMBOL(invalidate_inode_buffers); |
1da177e4c Linux-2.6.12-rc2 |
832 833 834 835 836 837 838 839 840 841 842 843 844 845 |
/* * Remove any clean buffers from the inode's buffer list. This is called * when we're trying to free the inode itself. Those buffers can pin it. * * Returns true if all buffers were removed. */ int remove_inode_buffers(struct inode *inode) { int ret = 1; if (inode_has_buffers(inode)) { struct address_space *mapping = &inode->i_data; struct list_head *list = &mapping->private_list; |
252aa6f5b mm: redefine addr... |
846 |
struct address_space *buffer_mapping = mapping->private_data; |
1da177e4c Linux-2.6.12-rc2 |
847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 |
spin_lock(&buffer_mapping->private_lock); while (!list_empty(list)) { struct buffer_head *bh = BH_ENTRY(list->next); if (buffer_dirty(bh)) { ret = 0; break; } __remove_assoc_queue(bh); } spin_unlock(&buffer_mapping->private_lock); } return ret; } /* * Create the appropriate buffers when given a page for data area and * the size of each buffer.. Use the bh->b_this_page linked list to * follow the buffers created. Return NULL if unable to create more * buffers. * * The retry flag is used to differentiate async IO (paging, swapping) * which may not fail from ordinary buffer allocations. */ struct buffer_head *alloc_page_buffers(struct page *page, unsigned long size, int retry) { struct buffer_head *bh, *head; long offset; try_again: head = NULL; offset = PAGE_SIZE; while ((offset -= size) >= 0) { bh = alloc_buffer_head(GFP_NOFS); if (!bh) goto no_grow; |
1da177e4c Linux-2.6.12-rc2 |
884 885 886 |
bh->b_this_page = head; bh->b_blocknr = -1; head = bh; |
1da177e4c Linux-2.6.12-rc2 |
887 888 889 890 |
bh->b_size = size; /* Link the buffer to its page */ set_bh_page(bh, page, offset); |
1da177e4c Linux-2.6.12-rc2 |
891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 |
} return head; /* * In case anything failed, we just free everything we got. */ no_grow: if (head) { do { bh = head; head = head->b_this_page; free_buffer_head(bh); } while (head); } /* * Return failure for non-async IO requests. Async IO requests * are not allowed to fail, so we have to wait until buffer heads * become available. But we don't want tasks sleeping with * partially complete buffers, so all were released above. */ if (!retry) return NULL; /* We're _really_ low on memory. Now we just * wait for old buffer heads to become free due to * finishing IO. Since this is an async request and * the reserve list is empty, we're sure there are * async buffer heads in use. */ free_more_memory(); goto try_again; } EXPORT_SYMBOL_GPL(alloc_page_buffers); static inline void link_dev_buffers(struct page *page, struct buffer_head *head) { struct buffer_head *bh, *tail; bh = head; do { tail = bh; bh = bh->b_this_page; } while (bh); tail->b_this_page = head; attach_page_buffers(page, head); } |
bbec0270b blkdev_max_block:... |
938 939 940 941 942 943 944 945 946 947 948 |
static sector_t blkdev_max_block(struct block_device *bdev, unsigned int size) { sector_t retval = ~((sector_t)0); loff_t sz = i_size_read(bdev->bd_inode); if (sz) { unsigned int sizebits = blksize_bits(size); retval = (sz >> sizebits); } return retval; } |
1da177e4c Linux-2.6.12-rc2 |
949 950 951 |
/* * Initialise the state of a blockdev page's buffers. */ |
676ce6d5c block: replace __... |
952 |
static sector_t |
1da177e4c Linux-2.6.12-rc2 |
953 954 955 956 957 958 |
init_page_buffers(struct page *page, struct block_device *bdev, sector_t block, int size) { struct buffer_head *head = page_buffers(page); struct buffer_head *bh = head; int uptodate = PageUptodate(page); |
bbec0270b blkdev_max_block:... |
959 |
sector_t end_block = blkdev_max_block(I_BDEV(bdev->bd_inode), size); |
1da177e4c Linux-2.6.12-rc2 |
960 961 962 963 964 965 966 967 |
do { if (!buffer_mapped(bh)) { init_buffer(bh, NULL, NULL); bh->b_bdev = bdev; bh->b_blocknr = block; if (uptodate) set_buffer_uptodate(bh); |
080399aaa block: don't mark... |
968 969 |
if (block < end_block) set_buffer_mapped(bh); |
1da177e4c Linux-2.6.12-rc2 |
970 971 972 973 |
} block++; bh = bh->b_this_page; } while (bh != head); |
676ce6d5c block: replace __... |
974 975 976 977 978 |
/* * Caller needs to validate requested block against end of device. */ return end_block; |
1da177e4c Linux-2.6.12-rc2 |
979 980 981 982 983 |
} /* * Create the page-cache page that contains the requested block. * |
676ce6d5c block: replace __... |
984 |
* This is used purely for blockdev mappings. |
1da177e4c Linux-2.6.12-rc2 |
985 |
*/ |
676ce6d5c block: replace __... |
986 |
static int |
1da177e4c Linux-2.6.12-rc2 |
987 |
grow_dev_page(struct block_device *bdev, sector_t block, |
676ce6d5c block: replace __... |
988 |
pgoff_t index, int size, int sizebits) |
1da177e4c Linux-2.6.12-rc2 |
989 990 991 992 |
{ struct inode *inode = bdev->bd_inode; struct page *page; struct buffer_head *bh; |
676ce6d5c block: replace __... |
993 994 |
sector_t end_block; int ret = 0; /* Will call free_more_memory() */ |
84235de39 fs: buffer: move ... |
995 |
gfp_t gfp_mask; |
1da177e4c Linux-2.6.12-rc2 |
996 |
|
84235de39 fs: buffer: move ... |
997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 |
gfp_mask = mapping_gfp_mask(inode->i_mapping) & ~__GFP_FS; gfp_mask |= __GFP_MOVABLE; /* * XXX: __getblk_slow() can not really deal with failure and * will endlessly loop on improvised global reclaim. Prefer * looping in the allocator rather than here, at least that * code knows what it's doing. */ gfp_mask |= __GFP_NOFAIL; page = find_or_create_page(inode->i_mapping, index, gfp_mask); |
1da177e4c Linux-2.6.12-rc2 |
1008 |
if (!page) |
676ce6d5c block: replace __... |
1009 |
return ret; |
1da177e4c Linux-2.6.12-rc2 |
1010 |
|
e827f9235 BUG_ON() Conversi... |
1011 |
BUG_ON(!PageLocked(page)); |
1da177e4c Linux-2.6.12-rc2 |
1012 1013 1014 1015 |
if (page_has_buffers(page)) { bh = page_buffers(page); if (bh->b_size == size) { |
676ce6d5c block: replace __... |
1016 1017 1018 |
end_block = init_page_buffers(page, bdev, index << sizebits, size); goto done; |
1da177e4c Linux-2.6.12-rc2 |
1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 |
} if (!try_to_free_buffers(page)) goto failed; } /* * Allocate some buffers for this page */ bh = alloc_page_buffers(page, size, 0); if (!bh) goto failed; /* * Link the page to the buffers and initialise them. Take the * lock to be atomic wrt __find_get_block(), which does not * run under the page lock. */ spin_lock(&inode->i_mapping->private_lock); link_dev_buffers(page, bh); |
676ce6d5c block: replace __... |
1038 |
end_block = init_page_buffers(page, bdev, index << sizebits, size); |
1da177e4c Linux-2.6.12-rc2 |
1039 |
spin_unlock(&inode->i_mapping->private_lock); |
676ce6d5c block: replace __... |
1040 1041 |
done: ret = (block < end_block) ? 1 : -ENXIO; |
1da177e4c Linux-2.6.12-rc2 |
1042 |
failed: |
1da177e4c Linux-2.6.12-rc2 |
1043 1044 |
unlock_page(page); page_cache_release(page); |
676ce6d5c block: replace __... |
1045 |
return ret; |
1da177e4c Linux-2.6.12-rc2 |
1046 1047 1048 1049 1050 |
} /* * Create buffers for the specified block device block's page. If * that page was dirty, the buffers are set dirty also. |
1da177e4c Linux-2.6.12-rc2 |
1051 |
*/ |
858119e15 [PATCH] Unlinline... |
1052 |
static int |
1da177e4c Linux-2.6.12-rc2 |
1053 1054 |
grow_buffers(struct block_device *bdev, sector_t block, int size) { |
1da177e4c Linux-2.6.12-rc2 |
1055 1056 1057 1058 1059 1060 1061 1062 1063 |
pgoff_t index; int sizebits; sizebits = -1; do { sizebits++; } while ((size << sizebits) < PAGE_SIZE); index = block >> sizebits; |
1da177e4c Linux-2.6.12-rc2 |
1064 |
|
e56579338 [PATCH] grow_buff... |
1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 |
/* * Check for a block which wants to lie outside our maximum possible * pagecache index. (this comparison is done using sector_t types). */ if (unlikely(index != block >> sizebits)) { char b[BDEVNAME_SIZE]; printk(KERN_ERR "%s: requested out-of-range block %llu for " "device %s ", |
8e24eea72 fs: replace remai... |
1075 |
__func__, (unsigned long long)block, |
e56579338 [PATCH] grow_buff... |
1076 1077 1078 |
bdevname(bdev, b)); return -EIO; } |
676ce6d5c block: replace __... |
1079 |
|
1da177e4c Linux-2.6.12-rc2 |
1080 |
/* Create a page with the proper size buffers.. */ |
676ce6d5c block: replace __... |
1081 |
return grow_dev_page(bdev, block, index, size, sizebits); |
1da177e4c Linux-2.6.12-rc2 |
1082 |
} |
75c96f858 [PATCH] make some... |
1083 |
static struct buffer_head * |
1da177e4c Linux-2.6.12-rc2 |
1084 1085 1086 |
__getblk_slow(struct block_device *bdev, sector_t block, int size) { /* Size must be multiple of hard sectorsize */ |
e1defc4ff block: Do away wi... |
1087 |
if (unlikely(size & (bdev_logical_block_size(bdev)-1) || |
1da177e4c Linux-2.6.12-rc2 |
1088 1089 1090 1091 |
(size < 512 || size > PAGE_SIZE))) { printk(KERN_ERR "getblk(): invalid block size %d requested ", size); |
e1defc4ff block: Do away wi... |
1092 1093 1094 |
printk(KERN_ERR "logical block size: %d ", bdev_logical_block_size(bdev)); |
1da177e4c Linux-2.6.12-rc2 |
1095 1096 1097 1098 |
dump_stack(); return NULL; } |
676ce6d5c block: replace __... |
1099 1100 1101 |
for (;;) { struct buffer_head *bh; int ret; |
1da177e4c Linux-2.6.12-rc2 |
1102 1103 1104 1105 |
bh = __find_get_block(bdev, block, size); if (bh) return bh; |
676ce6d5c block: replace __... |
1106 1107 1108 1109 1110 1111 |
ret = grow_buffers(bdev, block, size); if (ret < 0) return NULL; if (ret == 0) free_more_memory(); |
1da177e4c Linux-2.6.12-rc2 |
1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 |
} } /* * The relationship between dirty buffers and dirty pages: * * Whenever a page has any dirty buffers, the page's dirty bit is set, and * the page is tagged dirty in its radix tree. * * At all times, the dirtiness of the buffers represents the dirtiness of * subsections of the page. If the page has buffers, the page dirty bit is * merely a hint about the true dirty state. * * When a page is set dirty in its entirety, all its buffers are marked dirty * (if the page has buffers). * * When a buffer is marked dirty, its page is dirtied, but the page's other * buffers are not. * * Also. When blockdev buffers are explicitly read with bread(), they * individually become uptodate. But their backing page remains not * uptodate - even if all of its buffers are uptodate. A subsequent * block_read_full_page() against that page will discover all the uptodate * buffers, will set the page uptodate and will perform no I/O. */ /** * mark_buffer_dirty - mark a buffer_head as needing writeout |
67be2dd1b [PATCH] DocBook: ... |
1140 |
* @bh: the buffer_head to mark dirty |
1da177e4c Linux-2.6.12-rc2 |
1141 1142 1143 1144 1145 1146 1147 |
* * mark_buffer_dirty() will set the dirty bit against the buffer, then set its * backing page dirty, then tag the page as dirty in its address_space's radix * tree and then attach the address_space's inode to its superblock's dirty * inode list. * * mark_buffer_dirty() is atomic. It takes bh->b_page->mapping->private_lock, |
250df6ed2 fs: protect inode... |
1148 |
* mapping->tree_lock and mapping->host->i_lock. |
1da177e4c Linux-2.6.12-rc2 |
1149 |
*/ |
fc9b52cd8 fs: remove fastca... |
1150 |
void mark_buffer_dirty(struct buffer_head *bh) |
1da177e4c Linux-2.6.12-rc2 |
1151 |
{ |
787d2214c fs: introduce som... |
1152 |
WARN_ON_ONCE(!buffer_uptodate(bh)); |
1be62dc19 Be more careful a... |
1153 |
|
5305cb830 block: add block_... |
1154 |
trace_block_dirty_buffer(bh); |
1be62dc19 Be more careful a... |
1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 |
/* * Very *carefully* optimize the it-is-already-dirty case. * * Don't let the final "is it dirty" escape to before we * perhaps modified the buffer. */ if (buffer_dirty(bh)) { smp_mb(); if (buffer_dirty(bh)) return; } |
a8e7d49aa Fix race in creat... |
1166 1167 |
if (!test_set_buffer_dirty(bh)) { struct page *page = bh->b_page; |
8e9d78ede Re-introduce page... |
1168 1169 1170 1171 1172 |
if (!TestSetPageDirty(page)) { struct address_space *mapping = page_mapping(page); if (mapping) __set_page_dirty(page, mapping, 0); } |
a8e7d49aa Fix race in creat... |
1173 |
} |
1da177e4c Linux-2.6.12-rc2 |
1174 |
} |
1fe72eaa0 fs/buffer.c: clea... |
1175 |
EXPORT_SYMBOL(mark_buffer_dirty); |
1da177e4c Linux-2.6.12-rc2 |
1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 |
/* * Decrement a buffer_head's reference count. If all buffers against a page * have zero reference count, are clean and unlocked, and if the page is clean * and unlocked then try_to_free_buffers() may strip the buffers from the page * in preparation for freeing it (sometimes, rarely, buffers are removed from * a page but it ends up not being freed, and buffers may later be reattached). */ void __brelse(struct buffer_head * buf) { if (atomic_read(&buf->b_count)) { put_bh(buf); return; } |
5c752ad9f Use WARN() in fs/ |
1190 1191 |
WARN(1, KERN_ERR "VFS: brelse: Trying to free free buffer "); |
1da177e4c Linux-2.6.12-rc2 |
1192 |
} |
1fe72eaa0 fs/buffer.c: clea... |
1193 |
EXPORT_SYMBOL(__brelse); |
1da177e4c Linux-2.6.12-rc2 |
1194 1195 1196 1197 1198 1199 1200 1201 |
/* * bforget() is like brelse(), except it discards any * potentially dirty data. */ void __bforget(struct buffer_head *bh) { clear_buffer_dirty(bh); |
535ee2fbf buffer_head: fix ... |
1202 |
if (bh->b_assoc_map) { |
1da177e4c Linux-2.6.12-rc2 |
1203 1204 1205 1206 |
struct address_space *buffer_mapping = bh->b_page->mapping; spin_lock(&buffer_mapping->private_lock); list_del_init(&bh->b_assoc_buffers); |
58ff407be [PATCH] Fix IO er... |
1207 |
bh->b_assoc_map = NULL; |
1da177e4c Linux-2.6.12-rc2 |
1208 1209 1210 1211 |
spin_unlock(&buffer_mapping->private_lock); } __brelse(bh); } |
1fe72eaa0 fs/buffer.c: clea... |
1212 |
EXPORT_SYMBOL(__bforget); |
1da177e4c Linux-2.6.12-rc2 |
1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 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 1267 1268 1269 1270 1271 1272 1273 1274 |
static struct buffer_head *__bread_slow(struct buffer_head *bh) { lock_buffer(bh); if (buffer_uptodate(bh)) { unlock_buffer(bh); return bh; } else { get_bh(bh); bh->b_end_io = end_buffer_read_sync; submit_bh(READ, bh); wait_on_buffer(bh); if (buffer_uptodate(bh)) return bh; } brelse(bh); return NULL; } /* * Per-cpu buffer LRU implementation. To reduce the cost of __find_get_block(). * The bhs[] array is sorted - newest buffer is at bhs[0]. Buffers have their * refcount elevated by one when they're in an LRU. A buffer can only appear * once in a particular CPU's LRU. A single buffer can be present in multiple * CPU's LRUs at the same time. * * This is a transparent caching front-end to sb_bread(), sb_getblk() and * sb_find_get_block(). * * The LRUs themselves only need locking against invalidate_bh_lrus. We use * a local interrupt disable for that. */ #define BH_LRU_SIZE 8 struct bh_lru { struct buffer_head *bhs[BH_LRU_SIZE]; }; static DEFINE_PER_CPU(struct bh_lru, bh_lrus) = {{ NULL }}; #ifdef CONFIG_SMP #define bh_lru_lock() local_irq_disable() #define bh_lru_unlock() local_irq_enable() #else #define bh_lru_lock() preempt_disable() #define bh_lru_unlock() preempt_enable() #endif static inline void check_irqs_on(void) { #ifdef irqs_disabled BUG_ON(irqs_disabled()); #endif } /* * The LRU management algorithm is dopey-but-simple. Sorry. */ static void bh_lru_install(struct buffer_head *bh) { struct buffer_head *evictee = NULL; |
1da177e4c Linux-2.6.12-rc2 |
1275 1276 1277 |
check_irqs_on(); bh_lru_lock(); |
c7b92516a fs: Use this_cpu_... |
1278 |
if (__this_cpu_read(bh_lrus.bhs[0]) != bh) { |
1da177e4c Linux-2.6.12-rc2 |
1279 1280 1281 1282 1283 1284 1285 |
struct buffer_head *bhs[BH_LRU_SIZE]; int in; int out = 0; get_bh(bh); bhs[out++] = bh; for (in = 0; in < BH_LRU_SIZE; in++) { |
c7b92516a fs: Use this_cpu_... |
1286 1287 |
struct buffer_head *bh2 = __this_cpu_read(bh_lrus.bhs[in]); |
1da177e4c Linux-2.6.12-rc2 |
1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 |
if (bh2 == bh) { __brelse(bh2); } else { if (out >= BH_LRU_SIZE) { BUG_ON(evictee != NULL); evictee = bh2; } else { bhs[out++] = bh2; } } } while (out < BH_LRU_SIZE) bhs[out++] = NULL; |
ca6673b02 block: Replace __... |
1302 |
memcpy(this_cpu_ptr(&bh_lrus.bhs), bhs, sizeof(bhs)); |
1da177e4c Linux-2.6.12-rc2 |
1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 |
} bh_lru_unlock(); if (evictee) __brelse(evictee); } /* * Look up the bh in this cpu's LRU. If it's there, move it to the head. */ |
858119e15 [PATCH] Unlinline... |
1313 |
static struct buffer_head * |
3991d3bd1 [PATCH] warning f... |
1314 |
lookup_bh_lru(struct block_device *bdev, sector_t block, unsigned size) |
1da177e4c Linux-2.6.12-rc2 |
1315 1316 |
{ struct buffer_head *ret = NULL; |
3991d3bd1 [PATCH] warning f... |
1317 |
unsigned int i; |
1da177e4c Linux-2.6.12-rc2 |
1318 1319 1320 |
check_irqs_on(); bh_lru_lock(); |
1da177e4c Linux-2.6.12-rc2 |
1321 |
for (i = 0; i < BH_LRU_SIZE; i++) { |
c7b92516a fs: Use this_cpu_... |
1322 |
struct buffer_head *bh = __this_cpu_read(bh_lrus.bhs[i]); |
1da177e4c Linux-2.6.12-rc2 |
1323 1324 1325 1326 1327 |
if (bh && bh->b_bdev == bdev && bh->b_blocknr == block && bh->b_size == size) { if (i) { while (i) { |
c7b92516a fs: Use this_cpu_... |
1328 1329 |
__this_cpu_write(bh_lrus.bhs[i], __this_cpu_read(bh_lrus.bhs[i - 1])); |
1da177e4c Linux-2.6.12-rc2 |
1330 1331 |
i--; } |
c7b92516a fs: Use this_cpu_... |
1332 |
__this_cpu_write(bh_lrus.bhs[0], bh); |
1da177e4c Linux-2.6.12-rc2 |
1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 |
} get_bh(bh); ret = bh; break; } } bh_lru_unlock(); return ret; } /* * Perform a pagecache lookup for the matching buffer. If it's there, refresh * it in the LRU and mark it as accessed. If it is not present then return * NULL */ struct buffer_head * |
3991d3bd1 [PATCH] warning f... |
1349 |
__find_get_block(struct block_device *bdev, sector_t block, unsigned size) |
1da177e4c Linux-2.6.12-rc2 |
1350 1351 1352 1353 |
{ struct buffer_head *bh = lookup_bh_lru(bdev, block, size); if (bh == NULL) { |
385fd4c59 [PATCH] __find_ge... |
1354 |
bh = __find_get_block_slow(bdev, block); |
1da177e4c Linux-2.6.12-rc2 |
1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 |
if (bh) bh_lru_install(bh); } if (bh) touch_buffer(bh); return bh; } EXPORT_SYMBOL(__find_get_block); /* * __getblk will locate (and, if necessary, create) the buffer_head * which corresponds to the passed block_device, block and size. The * returned buffer has its reference count incremented. * |
1da177e4c Linux-2.6.12-rc2 |
1369 1370 1371 1372 |
* __getblk() will lock up the machine if grow_dev_page's try_to_free_buffers() * attempt is failing. FIXME, perhaps? */ struct buffer_head * |
3991d3bd1 [PATCH] warning f... |
1373 |
__getblk(struct block_device *bdev, sector_t block, unsigned size) |
1da177e4c Linux-2.6.12-rc2 |
1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 |
{ struct buffer_head *bh = __find_get_block(bdev, block, size); might_sleep(); if (bh == NULL) bh = __getblk_slow(bdev, block, size); return bh; } EXPORT_SYMBOL(__getblk); /* * Do async read-ahead on a buffer.. */ |
3991d3bd1 [PATCH] warning f... |
1387 |
void __breadahead(struct block_device *bdev, sector_t block, unsigned size) |
1da177e4c Linux-2.6.12-rc2 |
1388 1389 |
{ struct buffer_head *bh = __getblk(bdev, block, size); |
a3e713b5f [PATCH] __bread o... |
1390 1391 1392 1393 |
if (likely(bh)) { ll_rw_block(READA, 1, &bh); brelse(bh); } |
1da177e4c Linux-2.6.12-rc2 |
1394 1395 1396 1397 1398 |
} EXPORT_SYMBOL(__breadahead); /** * __bread() - reads a specified block and returns the bh |
67be2dd1b [PATCH] DocBook: ... |
1399 |
* @bdev: the block_device to read from |
1da177e4c Linux-2.6.12-rc2 |
1400 1401 1402 1403 1404 1405 1406 |
* @block: number of block * @size: size (in bytes) to read * * Reads a specified block, and returns buffer head that contains it. * It returns NULL if the block was unreadable. */ struct buffer_head * |
3991d3bd1 [PATCH] warning f... |
1407 |
__bread(struct block_device *bdev, sector_t block, unsigned size) |
1da177e4c Linux-2.6.12-rc2 |
1408 1409 |
{ struct buffer_head *bh = __getblk(bdev, block, size); |
a3e713b5f [PATCH] __bread o... |
1410 |
if (likely(bh) && !buffer_uptodate(bh)) |
1da177e4c Linux-2.6.12-rc2 |
1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 |
bh = __bread_slow(bh); return bh; } EXPORT_SYMBOL(__bread); /* * invalidate_bh_lrus() is called rarely - but not only at unmount. * This doesn't race because it runs in each cpu either in irq * or with preempt disabled. */ static void invalidate_bh_lru(void *arg) { struct bh_lru *b = &get_cpu_var(bh_lrus); int i; for (i = 0; i < BH_LRU_SIZE; i++) { brelse(b->bhs[i]); b->bhs[i] = NULL; } put_cpu_var(bh_lrus); } |
42be35d03 fs: only send IPI... |
1432 1433 1434 1435 1436 |
static bool has_bh_in_lru(int cpu, void *dummy) { struct bh_lru *b = per_cpu_ptr(&bh_lrus, cpu); int i; |
1da177e4c Linux-2.6.12-rc2 |
1437 |
|
42be35d03 fs: only send IPI... |
1438 1439 1440 1441 1442 1443 1444 |
for (i = 0; i < BH_LRU_SIZE; i++) { if (b->bhs[i]) return 1; } return 0; } |
f9a14399a mm: optimize kill... |
1445 |
void invalidate_bh_lrus(void) |
1da177e4c Linux-2.6.12-rc2 |
1446 |
{ |
42be35d03 fs: only send IPI... |
1447 |
on_each_cpu_cond(has_bh_in_lru, invalidate_bh_lru, NULL, 1, GFP_KERNEL); |
1da177e4c Linux-2.6.12-rc2 |
1448 |
} |
9db5579be rewrite rd |
1449 |
EXPORT_SYMBOL_GPL(invalidate_bh_lrus); |
1da177e4c Linux-2.6.12-rc2 |
1450 1451 1452 1453 1454 |
void set_bh_page(struct buffer_head *bh, struct page *page, unsigned long offset) { bh->b_page = page; |
e827f9235 BUG_ON() Conversi... |
1455 |
BUG_ON(offset >= PAGE_SIZE); |
1da177e4c Linux-2.6.12-rc2 |
1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 |
if (PageHighMem(page)) /* * This catches illegal uses and preserves the offset: */ bh->b_data = (char *)(0 + offset); else bh->b_data = page_address(page) + offset; } EXPORT_SYMBOL(set_bh_page); /* * Called when truncating a buffer on a page completely. */ |
858119e15 [PATCH] Unlinline... |
1469 |
static void discard_buffer(struct buffer_head * bh) |
1da177e4c Linux-2.6.12-rc2 |
1470 1471 1472 1473 1474 1475 1476 1477 |
{ lock_buffer(bh); clear_buffer_dirty(bh); bh->b_bdev = NULL; clear_buffer_mapped(bh); clear_buffer_req(bh); clear_buffer_new(bh); clear_buffer_delay(bh); |
33a266dda [PATCH] Make BH_U... |
1478 |
clear_buffer_unwritten(bh); |
1da177e4c Linux-2.6.12-rc2 |
1479 1480 1481 1482 |
unlock_buffer(bh); } /** |
814e1d25a cleanup: vfs: sma... |
1483 |
* block_invalidatepage - invalidate part or all of a buffer-backed page |
1da177e4c Linux-2.6.12-rc2 |
1484 1485 |
* * @page: the page which is affected |
d47992f86 mm: change invali... |
1486 1487 |
* @offset: start of the range to invalidate * @length: length of the range to invalidate |
1da177e4c Linux-2.6.12-rc2 |
1488 1489 |
* * block_invalidatepage() is called when all or part of the page has become |
814e1d25a cleanup: vfs: sma... |
1490 |
* invalidated by a truncate operation. |
1da177e4c Linux-2.6.12-rc2 |
1491 1492 1493 1494 1495 1496 1497 |
* * block_invalidatepage() does not have to release all buffers, but it must * ensure that no dirty buffer is left outside @offset and that no I/O * is underway against any of the blocks which are outside the truncation * point. Because the caller is about to free (and possibly reuse) those * blocks on-disk. */ |
d47992f86 mm: change invali... |
1498 1499 |
void block_invalidatepage(struct page *page, unsigned int offset, unsigned int length) |
1da177e4c Linux-2.6.12-rc2 |
1500 1501 1502 |
{ struct buffer_head *head, *bh, *next; unsigned int curr_off = 0; |
d47992f86 mm: change invali... |
1503 |
unsigned int stop = length + offset; |
1da177e4c Linux-2.6.12-rc2 |
1504 1505 1506 1507 |
BUG_ON(!PageLocked(page)); if (!page_has_buffers(page)) goto out; |
d47992f86 mm: change invali... |
1508 1509 1510 1511 |
/* * Check for overflow */ BUG_ON(stop > PAGE_CACHE_SIZE || stop < length); |
1da177e4c Linux-2.6.12-rc2 |
1512 1513 1514 1515 1516 1517 1518 |
head = page_buffers(page); bh = head; do { unsigned int next_off = curr_off + bh->b_size; next = bh->b_this_page; /* |
d47992f86 mm: change invali... |
1519 1520 1521 1522 1523 1524 |
* Are we still fully in range ? */ if (next_off > stop) goto out; /* |
1da177e4c Linux-2.6.12-rc2 |
1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 |
* is this block fully invalidated? */ if (offset <= curr_off) discard_buffer(bh); curr_off = next_off; bh = next; } while (bh != head); /* * We release buffers only if the entire page is being invalidated. * The get_block cached value has been unconditionally invalidated, * so real IO is not possible anymore. */ if (offset == 0) |
2ff28e22b [PATCH] Make addr... |
1539 |
try_to_release_page(page, 0); |
1da177e4c Linux-2.6.12-rc2 |
1540 |
out: |
2ff28e22b [PATCH] Make addr... |
1541 |
return; |
1da177e4c Linux-2.6.12-rc2 |
1542 1543 |
} EXPORT_SYMBOL(block_invalidatepage); |
d47992f86 mm: change invali... |
1544 |
|
1da177e4c Linux-2.6.12-rc2 |
1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 |
/* * We attach and possibly dirty the buffers atomically wrt * __set_page_dirty_buffers() via private_lock. try_to_free_buffers * is already excluded via the page lock. */ void create_empty_buffers(struct page *page, unsigned long blocksize, unsigned long b_state) { struct buffer_head *bh, *head, *tail; head = alloc_page_buffers(page, blocksize, 1); bh = head; do { bh->b_state |= b_state; tail = bh; bh = bh->b_this_page; } while (bh); tail->b_this_page = head; spin_lock(&page->mapping->private_lock); if (PageUptodate(page) || PageDirty(page)) { bh = head; do { if (PageDirty(page)) set_buffer_dirty(bh); if (PageUptodate(page)) set_buffer_uptodate(bh); bh = bh->b_this_page; } while (bh != head); } attach_page_buffers(page, head); spin_unlock(&page->mapping->private_lock); } EXPORT_SYMBOL(create_empty_buffers); /* * We are taking a block for data and we don't want any output from any * buffer-cache aliases starting from return from that function and * until the moment when something will explicitly mark the buffer * dirty (hopefully that will not happen until we will free that block ;-) * We don't even need to mark it not-uptodate - nobody can expect * anything from a newly allocated buffer anyway. We used to used * unmap_buffer() for such invalidation, but that was wrong. We definitely * don't want to mark the alias unmapped, for example - it would confuse * anyone who might pick it with bread() afterwards... * * Also.. Note that bforget() doesn't lock the buffer. So there can * be writeout I/O going on against recently-freed buffers. We don't * wait on that I/O in bforget() - it's more efficient to wait on the I/O * only if we really need to. That happens here. */ void unmap_underlying_metadata(struct block_device *bdev, sector_t block) { struct buffer_head *old_bh; might_sleep(); |
385fd4c59 [PATCH] __find_ge... |
1601 |
old_bh = __find_get_block_slow(bdev, block); |
1da177e4c Linux-2.6.12-rc2 |
1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 |
if (old_bh) { clear_buffer_dirty(old_bh); wait_on_buffer(old_bh); clear_buffer_req(old_bh); __brelse(old_bh); } } EXPORT_SYMBOL(unmap_underlying_metadata); /* |
45bce8f3e fs/buffer.c: make... |
1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 |
* Size is a power-of-two in the range 512..PAGE_SIZE, * and the case we care about most is PAGE_SIZE. * * So this *could* possibly be written with those * constraints in mind (relevant mostly if some * architecture has a slow bit-scan instruction) */ static inline int block_size_bits(unsigned int blocksize) { return ilog2(blocksize); } static struct buffer_head *create_page_buffers(struct page *page, struct inode *inode, unsigned int b_state) { BUG_ON(!PageLocked(page)); if (!page_has_buffers(page)) create_empty_buffers(page, 1 << ACCESS_ONCE(inode->i_blkbits), b_state); return page_buffers(page); } /* |
1da177e4c Linux-2.6.12-rc2 |
1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 |
* NOTE! All mapped/uptodate combinations are valid: * * Mapped Uptodate Meaning * * No No "unknown" - must do get_block() * No Yes "hole" - zero-filled * Yes No "allocated" - allocated on disk, not read in * Yes Yes "valid" - allocated and up-to-date in memory. * * "Dirty" is valid only with the last case (mapped+uptodate). */ /* * While block_write_full_page is writing back the dirty buffers under * the page lock, whoever dirtied the buffers may decide to clean them * again at any time. We handle that by only looking at the buffer * state inside lock_buffer(). * * If block_write_full_page() is called for regular writeback * (wbc->sync_mode == WB_SYNC_NONE) then it will redirty a page which has a * locked buffer. This only can happen if someone has written the buffer * directly, with submit_bh(). At the address_space level PageWriteback * prevents this contention from occurring. |
6e34eeddf block_write_full_... |
1657 1658 |
* * If block_write_full_page() is called with wbc->sync_mode == |
721a9602e block: kill off R... |
1659 1660 |
* WB_SYNC_ALL, the writes are posted using WRITE_SYNC; this * causes the writes to be flagged as synchronous writes. |
1da177e4c Linux-2.6.12-rc2 |
1661 1662 |
*/ static int __block_write_full_page(struct inode *inode, struct page *page, |
35c80d5f4 Add block_write_f... |
1663 1664 |
get_block_t *get_block, struct writeback_control *wbc, bh_end_io_t *handler) |
1da177e4c Linux-2.6.12-rc2 |
1665 1666 1667 1668 |
{ int err; sector_t block; sector_t last_block; |
f0fbd5fc0 [PATCH] __block_w... |
1669 |
struct buffer_head *bh, *head; |
45bce8f3e fs/buffer.c: make... |
1670 |
unsigned int blocksize, bbits; |
1da177e4c Linux-2.6.12-rc2 |
1671 |
int nr_underway = 0; |
6e34eeddf block_write_full_... |
1672 |
int write_op = (wbc->sync_mode == WB_SYNC_ALL ? |
721a9602e block: kill off R... |
1673 |
WRITE_SYNC : WRITE); |
1da177e4c Linux-2.6.12-rc2 |
1674 |
|
45bce8f3e fs/buffer.c: make... |
1675 |
head = create_page_buffers(page, inode, |
1da177e4c Linux-2.6.12-rc2 |
1676 |
(1 << BH_Dirty)|(1 << BH_Uptodate)); |
1da177e4c Linux-2.6.12-rc2 |
1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 |
/* * Be very careful. We have no exclusion from __set_page_dirty_buffers * here, and the (potentially unmapped) buffers may become dirty at * any time. If a buffer becomes dirty here after we've inspected it * then we just miss that fact, and the page stays dirty. * * Buffers outside i_size may be dirtied by __set_page_dirty_buffers; * handle that here by just cleaning them. */ |
1da177e4c Linux-2.6.12-rc2 |
1687 |
bh = head; |
45bce8f3e fs/buffer.c: make... |
1688 1689 1690 1691 1692 |
blocksize = bh->b_size; bbits = block_size_bits(blocksize); block = (sector_t)page->index << (PAGE_CACHE_SHIFT - bbits); last_block = (i_size_read(inode) - 1) >> bbits; |
1da177e4c Linux-2.6.12-rc2 |
1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 |
/* * Get all the dirty buffers mapped to disk addresses and * handle any aliases from the underlying blockdev's mapping. */ do { if (block > last_block) { /* * mapped buffers outside i_size will occur, because * this page can be outside i_size when there is a * truncate in progress. */ /* * The buffer was zeroed by block_write_full_page() */ clear_buffer_dirty(bh); set_buffer_uptodate(bh); |
29a814d2e vfs: add hooks fo... |
1710 1711 |
} else if ((!buffer_mapped(bh) || buffer_delay(bh)) && buffer_dirty(bh)) { |
b0cf2321c [PATCH] pass b_si... |
1712 |
WARN_ON(bh->b_size != blocksize); |
1da177e4c Linux-2.6.12-rc2 |
1713 1714 1715 |
err = get_block(inode, block, bh, 1); if (err) goto recover; |
29a814d2e vfs: add hooks fo... |
1716 |
clear_buffer_delay(bh); |
1da177e4c Linux-2.6.12-rc2 |
1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 |
if (buffer_new(bh)) { /* blockdev mappings never come here */ clear_buffer_new(bh); unmap_underlying_metadata(bh->b_bdev, bh->b_blocknr); } } bh = bh->b_this_page; block++; } while (bh != head); do { |
1da177e4c Linux-2.6.12-rc2 |
1729 1730 1731 1732 1733 |
if (!buffer_mapped(bh)) continue; /* * If it's a fully non-blocking write attempt and we cannot * lock the buffer then redirty the page. Note that this can |
5b0830cb9 writeback: get ri... |
1734 1735 1736 |
* potentially cause a busy-wait loop from writeback threads * and kswapd activity, but those code paths have their own * higher-level throttling. |
1da177e4c Linux-2.6.12-rc2 |
1737 |
*/ |
1b430beee writeback: remove... |
1738 |
if (wbc->sync_mode != WB_SYNC_NONE) { |
1da177e4c Linux-2.6.12-rc2 |
1739 |
lock_buffer(bh); |
ca5de404f fs: rename buffer... |
1740 |
} else if (!trylock_buffer(bh)) { |
1da177e4c Linux-2.6.12-rc2 |
1741 1742 1743 1744 |
redirty_page_for_writepage(wbc, page); continue; } if (test_clear_buffer_dirty(bh)) { |
35c80d5f4 Add block_write_f... |
1745 |
mark_buffer_async_write_endio(bh, handler); |
1da177e4c Linux-2.6.12-rc2 |
1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 |
} else { unlock_buffer(bh); } } while ((bh = bh->b_this_page) != head); /* * The page and its buffers are protected by PageWriteback(), so we can * drop the bh refcounts early. */ BUG_ON(PageWriteback(page)); set_page_writeback(page); |
1da177e4c Linux-2.6.12-rc2 |
1757 1758 1759 1760 |
do { struct buffer_head *next = bh->b_this_page; if (buffer_async_write(bh)) { |
a64c8610b block_write_full_... |
1761 |
submit_bh(write_op, bh); |
1da177e4c Linux-2.6.12-rc2 |
1762 1763 |
nr_underway++; } |
1da177e4c Linux-2.6.12-rc2 |
1764 1765 |
bh = next; } while (bh != head); |
05937baae [PATCH] __block_w... |
1766 |
unlock_page(page); |
1da177e4c Linux-2.6.12-rc2 |
1767 1768 1769 1770 1771 1772 1773 1774 1775 |
err = 0; done: if (nr_underway == 0) { /* * The page was marked dirty, but the buffers were * clean. Someone wrote them back by hand with * ll_rw_block/submit_bh. A rare case. */ |
1da177e4c Linux-2.6.12-rc2 |
1776 |
end_page_writeback(page); |
3d67f2d7c fs: buffer don't ... |
1777 |
|
1da177e4c Linux-2.6.12-rc2 |
1778 1779 1780 1781 |
/* * The page and buffer_heads can be released at any time from * here on. */ |
1da177e4c Linux-2.6.12-rc2 |
1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 |
} return err; recover: /* * ENOSPC, or some other error. We may already have added some * blocks to the file, so we need to write these out to avoid * exposing stale data. * The page is currently locked and not marked for writeback */ bh = head; /* Recovery: lock and submit the mapped buffers */ do { |
29a814d2e vfs: add hooks fo... |
1795 1796 |
if (buffer_mapped(bh) && buffer_dirty(bh) && !buffer_delay(bh)) { |
1da177e4c Linux-2.6.12-rc2 |
1797 |
lock_buffer(bh); |
35c80d5f4 Add block_write_f... |
1798 |
mark_buffer_async_write_endio(bh, handler); |
1da177e4c Linux-2.6.12-rc2 |
1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 |
} else { /* * The buffer may have been set dirty during * attachment to a dirty page. */ clear_buffer_dirty(bh); } } while ((bh = bh->b_this_page) != head); SetPageError(page); BUG_ON(PageWriteback(page)); |
7e4c3690b block_write_full_... |
1809 |
mapping_set_error(page->mapping, err); |
1da177e4c Linux-2.6.12-rc2 |
1810 |
set_page_writeback(page); |
1da177e4c Linux-2.6.12-rc2 |
1811 1812 1813 1814 |
do { struct buffer_head *next = bh->b_this_page; if (buffer_async_write(bh)) { clear_buffer_dirty(bh); |
a64c8610b block_write_full_... |
1815 |
submit_bh(write_op, bh); |
1da177e4c Linux-2.6.12-rc2 |
1816 1817 |
nr_underway++; } |
1da177e4c Linux-2.6.12-rc2 |
1818 1819 |
bh = next; } while (bh != head); |
ffda9d302 [PATCH] fs: fix _... |
1820 |
unlock_page(page); |
1da177e4c Linux-2.6.12-rc2 |
1821 1822 |
goto done; } |
afddba49d fs: introduce wri... |
1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 |
/* * If a page has any new buffers, zero them out here, and mark them uptodate * and dirty so they'll be written out (in order to prevent uninitialised * block data from leaking). And clear the new bit. */ void page_zero_new_buffers(struct page *page, unsigned from, unsigned to) { unsigned int block_start, block_end; struct buffer_head *head, *bh; BUG_ON(!PageLocked(page)); if (!page_has_buffers(page)) return; bh = head = page_buffers(page); block_start = 0; do { block_end = block_start + bh->b_size; if (buffer_new(bh)) { if (block_end > from && block_start < to) { if (!PageUptodate(page)) { unsigned start, size; start = max(from, block_start); size = min(to, block_end) - start; |
eebd2aa35 Pagecache zeroing... |
1849 |
zero_user(page, start, size); |
afddba49d fs: introduce wri... |
1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 |
set_buffer_uptodate(bh); } clear_buffer_new(bh); mark_buffer_dirty(bh); } } block_start = block_end; bh = bh->b_this_page; } while (bh != head); } EXPORT_SYMBOL(page_zero_new_buffers); |
ebdec241d fs: kill block_pr... |
1863 |
int __block_write_begin(struct page *page, loff_t pos, unsigned len, |
6e1db88d5 introduce __block... |
1864 |
get_block_t *get_block) |
1da177e4c Linux-2.6.12-rc2 |
1865 |
{ |
ebdec241d fs: kill block_pr... |
1866 1867 |
unsigned from = pos & (PAGE_CACHE_SIZE - 1); unsigned to = from + len; |
6e1db88d5 introduce __block... |
1868 |
struct inode *inode = page->mapping->host; |
1da177e4c Linux-2.6.12-rc2 |
1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 |
unsigned block_start, block_end; sector_t block; int err = 0; unsigned blocksize, bbits; struct buffer_head *bh, *head, *wait[2], **wait_bh=wait; BUG_ON(!PageLocked(page)); BUG_ON(from > PAGE_CACHE_SIZE); BUG_ON(to > PAGE_CACHE_SIZE); BUG_ON(from > to); |
45bce8f3e fs/buffer.c: make... |
1879 1880 1881 |
head = create_page_buffers(page, inode, 0); blocksize = head->b_size; bbits = block_size_bits(blocksize); |
1da177e4c Linux-2.6.12-rc2 |
1882 |
|
1da177e4c Linux-2.6.12-rc2 |
1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 |
block = (sector_t)page->index << (PAGE_CACHE_SHIFT - bbits); for(bh = head, block_start = 0; bh != head || !block_start; block++, block_start=block_end, bh = bh->b_this_page) { block_end = block_start + blocksize; if (block_end <= from || block_start >= to) { if (PageUptodate(page)) { if (!buffer_uptodate(bh)) set_buffer_uptodate(bh); } continue; } if (buffer_new(bh)) clear_buffer_new(bh); if (!buffer_mapped(bh)) { |
b0cf2321c [PATCH] pass b_si... |
1898 |
WARN_ON(bh->b_size != blocksize); |
1da177e4c Linux-2.6.12-rc2 |
1899 1900 |
err = get_block(inode, block, bh, 1); if (err) |
f3ddbdc62 [PATCH] fix race ... |
1901 |
break; |
1da177e4c Linux-2.6.12-rc2 |
1902 |
if (buffer_new(bh)) { |
1da177e4c Linux-2.6.12-rc2 |
1903 1904 1905 |
unmap_underlying_metadata(bh->b_bdev, bh->b_blocknr); if (PageUptodate(page)) { |
637aff46f fs: fix data-loss... |
1906 |
clear_buffer_new(bh); |
1da177e4c Linux-2.6.12-rc2 |
1907 |
set_buffer_uptodate(bh); |
637aff46f fs: fix data-loss... |
1908 |
mark_buffer_dirty(bh); |
1da177e4c Linux-2.6.12-rc2 |
1909 1910 |
continue; } |
eebd2aa35 Pagecache zeroing... |
1911 1912 1913 1914 |
if (block_end > to || block_start < from) zero_user_segments(page, to, block_end, block_start, from); |
1da177e4c Linux-2.6.12-rc2 |
1915 1916 1917 1918 1919 1920 1921 1922 1923 |
continue; } } if (PageUptodate(page)) { if (!buffer_uptodate(bh)) set_buffer_uptodate(bh); continue; } if (!buffer_uptodate(bh) && !buffer_delay(bh) && |
33a266dda [PATCH] Make BH_U... |
1924 |
!buffer_unwritten(bh) && |
1da177e4c Linux-2.6.12-rc2 |
1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 |
(block_start < from || block_end > to)) { ll_rw_block(READ, 1, &bh); *wait_bh++=bh; } } /* * If we issued read requests - let them complete. */ while(wait_bh > wait) { wait_on_buffer(*--wait_bh); if (!buffer_uptodate(*wait_bh)) |
f3ddbdc62 [PATCH] fix race ... |
1936 |
err = -EIO; |
1da177e4c Linux-2.6.12-rc2 |
1937 |
} |
f9f07b6c1 vfs: Fix data cor... |
1938 |
if (unlikely(err)) |
afddba49d fs: introduce wri... |
1939 |
page_zero_new_buffers(page, from, to); |
1da177e4c Linux-2.6.12-rc2 |
1940 1941 |
return err; } |
ebdec241d fs: kill block_pr... |
1942 |
EXPORT_SYMBOL(__block_write_begin); |
1da177e4c Linux-2.6.12-rc2 |
1943 1944 1945 1946 1947 1948 1949 1950 |
static int __block_commit_write(struct inode *inode, struct page *page, unsigned from, unsigned to) { unsigned block_start, block_end; int partial = 0; unsigned blocksize; struct buffer_head *bh, *head; |
45bce8f3e fs/buffer.c: make... |
1951 1952 |
bh = head = page_buffers(page); blocksize = bh->b_size; |
1da177e4c Linux-2.6.12-rc2 |
1953 |
|
45bce8f3e fs/buffer.c: make... |
1954 1955 |
block_start = 0; do { |
1da177e4c Linux-2.6.12-rc2 |
1956 1957 1958 1959 1960 1961 1962 1963 |
block_end = block_start + blocksize; if (block_end <= from || block_start >= to) { if (!buffer_uptodate(bh)) partial = 1; } else { set_buffer_uptodate(bh); mark_buffer_dirty(bh); } |
afddba49d fs: introduce wri... |
1964 |
clear_buffer_new(bh); |
45bce8f3e fs/buffer.c: make... |
1965 1966 1967 1968 |
block_start = block_end; bh = bh->b_this_page; } while (bh != head); |
1da177e4c Linux-2.6.12-rc2 |
1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 |
/* * If this is a partial write which happened to make all buffers * uptodate then we can optimize away a bogus readpage() for * the next read(). Here we 'discover' whether the page went * uptodate as a result of this (potentially partial) write. */ if (!partial) SetPageUptodate(page); return 0; } /* |
155130a4f get rid of block_... |
1982 1983 1984 |
* block_write_begin takes care of the basic task of block allocation and * bringing partial write blocks uptodate first. * |
7bb46a673 fs: introduce new... |
1985 |
* The filesystem needs to handle block truncation upon failure. |
afddba49d fs: introduce wri... |
1986 |
*/ |
155130a4f get rid of block_... |
1987 1988 |
int block_write_begin(struct address_space *mapping, loff_t pos, unsigned len, unsigned flags, struct page **pagep, get_block_t *get_block) |
afddba49d fs: introduce wri... |
1989 |
{ |
6e1db88d5 introduce __block... |
1990 |
pgoff_t index = pos >> PAGE_CACHE_SHIFT; |
afddba49d fs: introduce wri... |
1991 |
struct page *page; |
6e1db88d5 introduce __block... |
1992 |
int status; |
afddba49d fs: introduce wri... |
1993 |
|
6e1db88d5 introduce __block... |
1994 1995 1996 |
page = grab_cache_page_write_begin(mapping, index, flags); if (!page) return -ENOMEM; |
afddba49d fs: introduce wri... |
1997 |
|
6e1db88d5 introduce __block... |
1998 |
status = __block_write_begin(page, pos, len, get_block); |
afddba49d fs: introduce wri... |
1999 |
if (unlikely(status)) { |
6e1db88d5 introduce __block... |
2000 2001 2002 |
unlock_page(page); page_cache_release(page); page = NULL; |
afddba49d fs: introduce wri... |
2003 |
} |
6e1db88d5 introduce __block... |
2004 |
*pagep = page; |
afddba49d fs: introduce wri... |
2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 |
return status; } EXPORT_SYMBOL(block_write_begin); int block_write_end(struct file *file, struct address_space *mapping, loff_t pos, unsigned len, unsigned copied, struct page *page, void *fsdata) { struct inode *inode = mapping->host; unsigned start; start = pos & (PAGE_CACHE_SIZE - 1); if (unlikely(copied < len)) { /* * The buffers that were written will now be uptodate, so we * don't have to worry about a readpage reading them and * overwriting a partial write. However if we have encountered * a short write and only partially written into a buffer, it * will not be marked uptodate, so a readpage might come in and * destroy our partial write. * * Do the simplest thing, and just treat any short write to a * non uptodate page as a zero-length write, and force the * caller to redo the whole thing. */ if (!PageUptodate(page)) copied = 0; page_zero_new_buffers(page, start+copied, start+len); } flush_dcache_page(page); /* This could be a short (even 0-length) commit */ __block_commit_write(inode, page, start, start+copied); return copied; } EXPORT_SYMBOL(block_write_end); int generic_write_end(struct file *file, struct address_space *mapping, loff_t pos, unsigned len, unsigned copied, struct page *page, void *fsdata) { struct inode *inode = mapping->host; |
c7d206b33 vfs: Move mark_in... |
2050 |
int i_size_changed = 0; |
afddba49d fs: introduce wri... |
2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 |
copied = block_write_end(file, mapping, pos, len, copied, page, fsdata); /* * No need to use i_size_read() here, the i_size * cannot change under us because we hold i_mutex. * * But it's important to update i_size while still holding page lock: * page writeout could otherwise come in and zero beyond i_size. */ if (pos+copied > inode->i_size) { i_size_write(inode, pos+copied); |
c7d206b33 vfs: Move mark_in... |
2063 |
i_size_changed = 1; |
afddba49d fs: introduce wri... |
2064 2065 2066 2067 |
} unlock_page(page); page_cache_release(page); |
c7d206b33 vfs: Move mark_in... |
2068 2069 2070 2071 2072 2073 2074 2075 |
/* * Don't mark the inode dirty under page lock. First, it unnecessarily * makes the holding time of page lock longer. Second, it forces lock * ordering of page lock and transaction start for journaling * filesystems. */ if (i_size_changed) mark_inode_dirty(inode); |
afddba49d fs: introduce wri... |
2076 2077 2078 2079 2080 |
return copied; } EXPORT_SYMBOL(generic_write_end); /* |
8ab22b9ab vfs: pagecache us... |
2081 2082 2083 2084 2085 2086 |
* block_is_partially_uptodate checks whether buffers within a page are * uptodate or not. * * Returns true if all buffers which correspond to a file portion * we want to read are uptodate. */ |
c186afb4d switch ->is_parti... |
2087 2088 |
int block_is_partially_uptodate(struct page *page, unsigned long from, unsigned long count) |
8ab22b9ab vfs: pagecache us... |
2089 |
{ |
8ab22b9ab vfs: pagecache us... |
2090 2091 2092 2093 2094 2095 2096 |
unsigned block_start, block_end, blocksize; unsigned to; struct buffer_head *bh, *head; int ret = 1; if (!page_has_buffers(page)) return 0; |
45bce8f3e fs/buffer.c: make... |
2097 2098 |
head = page_buffers(page); blocksize = head->b_size; |
c186afb4d switch ->is_parti... |
2099 |
to = min_t(unsigned, PAGE_CACHE_SIZE - from, count); |
8ab22b9ab vfs: pagecache us... |
2100 2101 2102 |
to = from + to; if (from < blocksize && to > PAGE_CACHE_SIZE - blocksize) return 0; |
8ab22b9ab vfs: pagecache us... |
2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 |
bh = head; block_start = 0; do { block_end = block_start + blocksize; if (block_end > from && block_start < to) { if (!buffer_uptodate(bh)) { ret = 0; break; } if (block_end >= to) break; } block_start = block_end; bh = bh->b_this_page; } while (bh != head); return ret; } EXPORT_SYMBOL(block_is_partially_uptodate); /* |
1da177e4c Linux-2.6.12-rc2 |
2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 |
* Generic "read page" function for block devices that have the normal * get_block functionality. This is most of the block device filesystems. * Reads the page asynchronously --- the unlock_buffer() and * set/clear_buffer_uptodate() functions propagate buffer state into the * page struct once IO has completed. */ int block_read_full_page(struct page *page, get_block_t *get_block) { struct inode *inode = page->mapping->host; sector_t iblock, lblock; struct buffer_head *bh, *head, *arr[MAX_BUF_PER_PAGE]; |
45bce8f3e fs/buffer.c: make... |
2135 |
unsigned int blocksize, bbits; |
1da177e4c Linux-2.6.12-rc2 |
2136 2137 |
int nr, i; int fully_mapped = 1; |
45bce8f3e fs/buffer.c: make... |
2138 2139 2140 |
head = create_page_buffers(page, inode, 0); blocksize = head->b_size; bbits = block_size_bits(blocksize); |
1da177e4c Linux-2.6.12-rc2 |
2141 |
|
45bce8f3e fs/buffer.c: make... |
2142 2143 |
iblock = (sector_t)page->index << (PAGE_CACHE_SHIFT - bbits); lblock = (i_size_read(inode)+blocksize-1) >> bbits; |
1da177e4c Linux-2.6.12-rc2 |
2144 2145 2146 2147 2148 2149 2150 2151 2152 |
bh = head; nr = 0; i = 0; do { if (buffer_uptodate(bh)) continue; if (!buffer_mapped(bh)) { |
c64610ba5 [PATCH] block_rea... |
2153 |
int err = 0; |
1da177e4c Linux-2.6.12-rc2 |
2154 2155 |
fully_mapped = 0; if (iblock < lblock) { |
b0cf2321c [PATCH] pass b_si... |
2156 |
WARN_ON(bh->b_size != blocksize); |
c64610ba5 [PATCH] block_rea... |
2157 2158 |
err = get_block(inode, iblock, bh, 0); if (err) |
1da177e4c Linux-2.6.12-rc2 |
2159 2160 2161 |
SetPageError(page); } if (!buffer_mapped(bh)) { |
eebd2aa35 Pagecache zeroing... |
2162 |
zero_user(page, i * blocksize, blocksize); |
c64610ba5 [PATCH] block_rea... |
2163 2164 |
if (!err) set_buffer_uptodate(bh); |
1da177e4c Linux-2.6.12-rc2 |
2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 |
continue; } /* * get_block() might have updated the buffer * synchronously */ if (buffer_uptodate(bh)) continue; } arr[nr++] = bh; } while (i++, iblock++, (bh = bh->b_this_page) != head); if (fully_mapped) SetPageMappedToDisk(page); if (!nr) { /* * All buffers are uptodate - we can set the page uptodate * as well. But not if get_block() returned an error. */ if (!PageError(page)) SetPageUptodate(page); unlock_page(page); return 0; } /* Stage two: lock the buffers */ for (i = 0; i < nr; i++) { bh = arr[i]; lock_buffer(bh); mark_buffer_async_read(bh); } /* * Stage 3: start the IO. Check for uptodateness * inside the buffer lock in case another process reading * the underlying blockdev brought it uptodate (the sct fix). */ for (i = 0; i < nr; i++) { bh = arr[i]; if (buffer_uptodate(bh)) end_buffer_async_read(bh, 1); else submit_bh(READ, bh); } return 0; } |
1fe72eaa0 fs/buffer.c: clea... |
2212 |
EXPORT_SYMBOL(block_read_full_page); |
1da177e4c Linux-2.6.12-rc2 |
2213 2214 |
/* utility function for filesystems that need to do work on expanding |
89e107877 fs: new cont helpers |
2215 |
* truncates. Uses filesystem pagecache writes to allow the filesystem to |
1da177e4c Linux-2.6.12-rc2 |
2216 2217 |
* deal with the hole. */ |
89e107877 fs: new cont helpers |
2218 |
int generic_cont_expand_simple(struct inode *inode, loff_t size) |
1da177e4c Linux-2.6.12-rc2 |
2219 2220 2221 |
{ struct address_space *mapping = inode->i_mapping; struct page *page; |
89e107877 fs: new cont helpers |
2222 |
void *fsdata; |
1da177e4c Linux-2.6.12-rc2 |
2223 |
int err; |
c08d3b0e3 truncate: use new... |
2224 2225 |
err = inode_newsize_ok(inode, size); if (err) |
1da177e4c Linux-2.6.12-rc2 |
2226 |
goto out; |
89e107877 fs: new cont helpers |
2227 2228 2229 2230 |
err = pagecache_write_begin(NULL, mapping, size, 0, AOP_FLAG_UNINTERRUPTIBLE|AOP_FLAG_CONT_EXPAND, &page, &fsdata); if (err) |
05eb0b51f [PATCH] fat: supp... |
2231 |
goto out; |
05eb0b51f [PATCH] fat: supp... |
2232 |
|
89e107877 fs: new cont helpers |
2233 2234 |
err = pagecache_write_end(NULL, mapping, size, 0, 0, page, fsdata); BUG_ON(err > 0); |
05eb0b51f [PATCH] fat: supp... |
2235 |
|
1da177e4c Linux-2.6.12-rc2 |
2236 2237 2238 |
out: return err; } |
1fe72eaa0 fs/buffer.c: clea... |
2239 |
EXPORT_SYMBOL(generic_cont_expand_simple); |
1da177e4c Linux-2.6.12-rc2 |
2240 |
|
f1e3af72c make fs/buffer.c:... |
2241 2242 |
static int cont_expand_zero(struct file *file, struct address_space *mapping, loff_t pos, loff_t *bytes) |
1da177e4c Linux-2.6.12-rc2 |
2243 |
{ |
1da177e4c Linux-2.6.12-rc2 |
2244 |
struct inode *inode = mapping->host; |
1da177e4c Linux-2.6.12-rc2 |
2245 |
unsigned blocksize = 1 << inode->i_blkbits; |
89e107877 fs: new cont helpers |
2246 2247 2248 2249 2250 2251 |
struct page *page; void *fsdata; pgoff_t index, curidx; loff_t curpos; unsigned zerofrom, offset, len; int err = 0; |
1da177e4c Linux-2.6.12-rc2 |
2252 |
|
89e107877 fs: new cont helpers |
2253 2254 2255 2256 2257 |
index = pos >> PAGE_CACHE_SHIFT; offset = pos & ~PAGE_CACHE_MASK; while (index > (curidx = (curpos = *bytes)>>PAGE_CACHE_SHIFT)) { zerofrom = curpos & ~PAGE_CACHE_MASK; |
1da177e4c Linux-2.6.12-rc2 |
2258 2259 2260 2261 |
if (zerofrom & (blocksize-1)) { *bytes |= (blocksize-1); (*bytes)++; } |
89e107877 fs: new cont helpers |
2262 |
len = PAGE_CACHE_SIZE - zerofrom; |
1da177e4c Linux-2.6.12-rc2 |
2263 |
|
89e107877 fs: new cont helpers |
2264 2265 2266 2267 2268 |
err = pagecache_write_begin(file, mapping, curpos, len, AOP_FLAG_UNINTERRUPTIBLE, &page, &fsdata); if (err) goto out; |
eebd2aa35 Pagecache zeroing... |
2269 |
zero_user(page, zerofrom, len); |
89e107877 fs: new cont helpers |
2270 2271 2272 2273 2274 2275 |
err = pagecache_write_end(file, mapping, curpos, len, len, page, fsdata); if (err < 0) goto out; BUG_ON(err != len); err = 0; |
061e97469 Add balance_dirty... |
2276 2277 |
balance_dirty_pages_ratelimited(mapping); |
89e107877 fs: new cont helpers |
2278 |
} |
1da177e4c Linux-2.6.12-rc2 |
2279 |
|
89e107877 fs: new cont helpers |
2280 2281 2282 |
/* page covers the boundary, find the boundary offset */ if (index == curidx) { zerofrom = curpos & ~PAGE_CACHE_MASK; |
1da177e4c Linux-2.6.12-rc2 |
2283 |
/* if we will expand the thing last block will be filled */ |
89e107877 fs: new cont helpers |
2284 2285 2286 2287 |
if (offset <= zerofrom) { goto out; } if (zerofrom & (blocksize-1)) { |
1da177e4c Linux-2.6.12-rc2 |
2288 2289 2290 |
*bytes |= (blocksize-1); (*bytes)++; } |
89e107877 fs: new cont helpers |
2291 |
len = offset - zerofrom; |
1da177e4c Linux-2.6.12-rc2 |
2292 |
|
89e107877 fs: new cont helpers |
2293 2294 2295 2296 2297 |
err = pagecache_write_begin(file, mapping, curpos, len, AOP_FLAG_UNINTERRUPTIBLE, &page, &fsdata); if (err) goto out; |
eebd2aa35 Pagecache zeroing... |
2298 |
zero_user(page, zerofrom, len); |
89e107877 fs: new cont helpers |
2299 2300 2301 2302 2303 2304 |
err = pagecache_write_end(file, mapping, curpos, len, len, page, fsdata); if (err < 0) goto out; BUG_ON(err != len); err = 0; |
1da177e4c Linux-2.6.12-rc2 |
2305 |
} |
89e107877 fs: new cont helpers |
2306 2307 2308 2309 2310 2311 2312 2313 |
out: return err; } /* * For moronic filesystems that do not allow holes in file. * We may have to extend the file. */ |
282dc1788 get rid of cont_w... |
2314 |
int cont_write_begin(struct file *file, struct address_space *mapping, |
89e107877 fs: new cont helpers |
2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 |
loff_t pos, unsigned len, unsigned flags, struct page **pagep, void **fsdata, get_block_t *get_block, loff_t *bytes) { struct inode *inode = mapping->host; unsigned blocksize = 1 << inode->i_blkbits; unsigned zerofrom; int err; err = cont_expand_zero(file, mapping, pos, bytes); if (err) |
155130a4f get rid of block_... |
2326 |
return err; |
89e107877 fs: new cont helpers |
2327 2328 2329 2330 2331 |
zerofrom = *bytes & ~PAGE_CACHE_MASK; if (pos+len > *bytes && zerofrom & (blocksize-1)) { *bytes |= (blocksize-1); (*bytes)++; |
1da177e4c Linux-2.6.12-rc2 |
2332 |
} |
1da177e4c Linux-2.6.12-rc2 |
2333 |
|
155130a4f get rid of block_... |
2334 |
return block_write_begin(mapping, pos, len, flags, pagep, get_block); |
1da177e4c Linux-2.6.12-rc2 |
2335 |
} |
1fe72eaa0 fs/buffer.c: clea... |
2336 |
EXPORT_SYMBOL(cont_write_begin); |
1da177e4c Linux-2.6.12-rc2 |
2337 |
|
1da177e4c Linux-2.6.12-rc2 |
2338 2339 2340 2341 2342 2343 |
int block_commit_write(struct page *page, unsigned from, unsigned to) { struct inode *inode = page->mapping->host; __block_commit_write(inode,page,from,to); return 0; } |
1fe72eaa0 fs/buffer.c: clea... |
2344 |
EXPORT_SYMBOL(block_commit_write); |
1da177e4c Linux-2.6.12-rc2 |
2345 |
|
541716902 [FS] Implement bl... |
2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 |
/* * block_page_mkwrite() is not allowed to change the file size as it gets * called from a page fault handler when a page is first dirtied. Hence we must * be careful to check for EOF conditions here. We set the page up correctly * for a written page which means we get ENOSPC checking when writing into * holes and correct delalloc and unwritten extent mapping on filesystems that * support these features. * * We are not allowed to take the i_mutex here so we have to play games to * protect against truncate races as the page could now be beyond EOF. Because |
7bb46a673 fs: introduce new... |
2356 |
* truncate writes the inode size before removing pages, once we have the |
541716902 [FS] Implement bl... |
2357 2358 2359 |
* page lock we can determine safely if the page is beyond EOF. If it is not * beyond EOF, then the page is guaranteed safe against truncation until we * unlock the page. |
ea13a8646 vfs: Block mmappe... |
2360 |
* |
14da92001 fs: Protect write... |
2361 2362 |
* Direct callers of this function should protect against filesystem freezing * using sb_start_write() - sb_end_write() functions. |
541716902 [FS] Implement bl... |
2363 |
*/ |
24da4fab5 vfs: Create __blo... |
2364 2365 |
int __block_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf, get_block_t get_block) |
541716902 [FS] Implement bl... |
2366 |
{ |
c2ec175c3 mm: page_mkwrite ... |
2367 |
struct page *page = vmf->page; |
496ad9aa8 new helper: file_... |
2368 |
struct inode *inode = file_inode(vma->vm_file); |
541716902 [FS] Implement bl... |
2369 2370 |
unsigned long end; loff_t size; |
24da4fab5 vfs: Create __blo... |
2371 |
int ret; |
541716902 [FS] Implement bl... |
2372 2373 2374 2375 |
lock_page(page); size = i_size_read(inode); if ((page->mapping != inode->i_mapping) || |
183363380 fix some conversi... |
2376 |
(page_offset(page) > size)) { |
24da4fab5 vfs: Create __blo... |
2377 2378 2379 |
/* We overload EFAULT to mean page got truncated */ ret = -EFAULT; goto out_unlock; |
541716902 [FS] Implement bl... |
2380 2381 2382 2383 2384 2385 2386 |
} /* page is wholly or partially inside EOF */ if (((page->index + 1) << PAGE_CACHE_SHIFT) > size) end = size & ~PAGE_CACHE_MASK; else end = PAGE_CACHE_SIZE; |
ebdec241d fs: kill block_pr... |
2387 |
ret = __block_write_begin(page, 0, end, get_block); |
541716902 [FS] Implement bl... |
2388 2389 |
if (!ret) ret = block_commit_write(page, 0, end); |
24da4fab5 vfs: Create __blo... |
2390 2391 |
if (unlikely(ret < 0)) goto out_unlock; |
ea13a8646 vfs: Block mmappe... |
2392 |
set_page_dirty(page); |
1d1d1a767 mm: only enforce ... |
2393 |
wait_for_stable_page(page); |
24da4fab5 vfs: Create __blo... |
2394 2395 2396 |
return 0; out_unlock: unlock_page(page); |
541716902 [FS] Implement bl... |
2397 |
return ret; |
24da4fab5 vfs: Create __blo... |
2398 2399 2400 2401 2402 2403 |
} EXPORT_SYMBOL(__block_page_mkwrite); int block_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf, get_block_t get_block) { |
ea13a8646 vfs: Block mmappe... |
2404 |
int ret; |
496ad9aa8 new helper: file_... |
2405 |
struct super_block *sb = file_inode(vma->vm_file)->i_sb; |
24da4fab5 vfs: Create __blo... |
2406 |
|
14da92001 fs: Protect write... |
2407 |
sb_start_pagefault(sb); |
041bbb6d3 ext4: fix mtime u... |
2408 2409 2410 2411 2412 2413 |
/* * Update file times before taking page lock. We may end up failing the * fault so this update may be superfluous but who really cares... */ file_update_time(vma->vm_file); |
ea13a8646 vfs: Block mmappe... |
2414 |
ret = __block_page_mkwrite(vma, vmf, get_block); |
14da92001 fs: Protect write... |
2415 |
sb_end_pagefault(sb); |
24da4fab5 vfs: Create __blo... |
2416 |
return block_page_mkwrite_return(ret); |
541716902 [FS] Implement bl... |
2417 |
} |
1fe72eaa0 fs/buffer.c: clea... |
2418 |
EXPORT_SYMBOL(block_page_mkwrite); |
1da177e4c Linux-2.6.12-rc2 |
2419 2420 |
/* |
03158cd7e fs: restore nobh |
2421 |
* nobh_write_begin()'s prereads are special: the buffer_heads are freed |
1da177e4c Linux-2.6.12-rc2 |
2422 2423 |
* immediately, while under the page lock. So it needs a special end_io * handler which does not touch the bh after unlocking it. |
1da177e4c Linux-2.6.12-rc2 |
2424 2425 2426 |
*/ static void end_buffer_read_nobh(struct buffer_head *bh, int uptodate) { |
68671f35f mm: add end_buffe... |
2427 |
__end_buffer_read_notouch(bh, uptodate); |
1da177e4c Linux-2.6.12-rc2 |
2428 2429 2430 |
} /* |
03158cd7e fs: restore nobh |
2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 |
* Attach the singly-linked list of buffers created by nobh_write_begin, to * the page (converting it to circular linked list and taking care of page * dirty races). */ static void attach_nobh_buffers(struct page *page, struct buffer_head *head) { struct buffer_head *bh; BUG_ON(!PageLocked(page)); spin_lock(&page->mapping->private_lock); bh = head; do { if (PageDirty(page)) set_buffer_dirty(bh); if (!bh->b_this_page) bh->b_this_page = head; bh = bh->b_this_page; } while (bh != head); attach_page_buffers(page, head); spin_unlock(&page->mapping->private_lock); } /* |
ea0f04e59 get rid of nobh_w... |
2455 2456 |
* On entry, the page is fully not uptodate. * On exit the page is fully uptodate in the areas outside (from,to) |
7bb46a673 fs: introduce new... |
2457 |
* The filesystem needs to handle block truncation upon failure. |
1da177e4c Linux-2.6.12-rc2 |
2458 |
*/ |
ea0f04e59 get rid of nobh_w... |
2459 |
int nobh_write_begin(struct address_space *mapping, |
03158cd7e fs: restore nobh |
2460 2461 |
loff_t pos, unsigned len, unsigned flags, struct page **pagep, void **fsdata, |
1da177e4c Linux-2.6.12-rc2 |
2462 2463 |
get_block_t *get_block) { |
03158cd7e fs: restore nobh |
2464 |
struct inode *inode = mapping->host; |
1da177e4c Linux-2.6.12-rc2 |
2465 2466 |
const unsigned blkbits = inode->i_blkbits; const unsigned blocksize = 1 << blkbits; |
a4b0672db fs: fix nobh erro... |
2467 |
struct buffer_head *head, *bh; |
03158cd7e fs: restore nobh |
2468 2469 2470 |
struct page *page; pgoff_t index; unsigned from, to; |
1da177e4c Linux-2.6.12-rc2 |
2471 |
unsigned block_in_page; |
a4b0672db fs: fix nobh erro... |
2472 |
unsigned block_start, block_end; |
1da177e4c Linux-2.6.12-rc2 |
2473 |
sector_t block_in_file; |
1da177e4c Linux-2.6.12-rc2 |
2474 |
int nr_reads = 0; |
1da177e4c Linux-2.6.12-rc2 |
2475 2476 |
int ret = 0; int is_mapped_to_disk = 1; |
1da177e4c Linux-2.6.12-rc2 |
2477 |
|
03158cd7e fs: restore nobh |
2478 2479 2480 |
index = pos >> PAGE_CACHE_SHIFT; from = pos & (PAGE_CACHE_SIZE - 1); to = from + len; |
54566b2c1 fs: symlink write... |
2481 |
page = grab_cache_page_write_begin(mapping, index, flags); |
03158cd7e fs: restore nobh |
2482 2483 2484 2485 2486 2487 |
if (!page) return -ENOMEM; *pagep = page; *fsdata = NULL; if (page_has_buffers(page)) { |
309f77ad9 fs/buffer.c: call... |
2488 2489 2490 2491 |
ret = __block_write_begin(page, pos, len, get_block); if (unlikely(ret)) goto out_release; return ret; |
03158cd7e fs: restore nobh |
2492 |
} |
a4b0672db fs: fix nobh erro... |
2493 |
|
1da177e4c Linux-2.6.12-rc2 |
2494 2495 |
if (PageMappedToDisk(page)) return 0; |
a4b0672db fs: fix nobh erro... |
2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 |
/* * Allocate buffers so that we can keep track of state, and potentially * attach them to the page if an error occurs. In the common case of * no error, they will just be freed again without ever being attached * to the page (which is all OK, because we're under the page lock). * * Be careful: the buffer linked list is a NULL terminated one, rather * than the circular one we're used to. */ head = alloc_page_buffers(page, blocksize, 0); |
03158cd7e fs: restore nobh |
2506 2507 2508 2509 |
if (!head) { ret = -ENOMEM; goto out_release; } |
a4b0672db fs: fix nobh erro... |
2510 |
|
1da177e4c Linux-2.6.12-rc2 |
2511 |
block_in_file = (sector_t)page->index << (PAGE_CACHE_SHIFT - blkbits); |
1da177e4c Linux-2.6.12-rc2 |
2512 2513 2514 2515 2516 2517 |
/* * We loop across all blocks in the page, whether or not they are * part of the affected region. This is so we can discover if the * page is fully mapped-to-disk. */ |
a4b0672db fs: fix nobh erro... |
2518 |
for (block_start = 0, block_in_page = 0, bh = head; |
1da177e4c Linux-2.6.12-rc2 |
2519 |
block_start < PAGE_CACHE_SIZE; |
a4b0672db fs: fix nobh erro... |
2520 |
block_in_page++, block_start += blocksize, bh = bh->b_this_page) { |
1da177e4c Linux-2.6.12-rc2 |
2521 |
int create; |
a4b0672db fs: fix nobh erro... |
2522 2523 |
block_end = block_start + blocksize; bh->b_state = 0; |
1da177e4c Linux-2.6.12-rc2 |
2524 2525 2526 2527 |
create = 1; if (block_start >= to) create = 0; ret = get_block(inode, block_in_file + block_in_page, |
a4b0672db fs: fix nobh erro... |
2528 |
bh, create); |
1da177e4c Linux-2.6.12-rc2 |
2529 2530 |
if (ret) goto failed; |
a4b0672db fs: fix nobh erro... |
2531 |
if (!buffer_mapped(bh)) |
1da177e4c Linux-2.6.12-rc2 |
2532 |
is_mapped_to_disk = 0; |
a4b0672db fs: fix nobh erro... |
2533 2534 2535 2536 |
if (buffer_new(bh)) unmap_underlying_metadata(bh->b_bdev, bh->b_blocknr); if (PageUptodate(page)) { set_buffer_uptodate(bh); |
1da177e4c Linux-2.6.12-rc2 |
2537 |
continue; |
a4b0672db fs: fix nobh erro... |
2538 2539 |
} if (buffer_new(bh) || !buffer_mapped(bh)) { |
eebd2aa35 Pagecache zeroing... |
2540 2541 |
zero_user_segments(page, block_start, from, to, block_end); |
1da177e4c Linux-2.6.12-rc2 |
2542 2543 |
continue; } |
a4b0672db fs: fix nobh erro... |
2544 |
if (buffer_uptodate(bh)) |
1da177e4c Linux-2.6.12-rc2 |
2545 2546 |
continue; /* reiserfs does this */ if (block_start < from || block_end > to) { |
a4b0672db fs: fix nobh erro... |
2547 2548 2549 2550 |
lock_buffer(bh); bh->b_end_io = end_buffer_read_nobh; submit_bh(READ, bh); nr_reads++; |
1da177e4c Linux-2.6.12-rc2 |
2551 2552 2553 2554 |
} } if (nr_reads) { |
1da177e4c Linux-2.6.12-rc2 |
2555 2556 2557 2558 2559 |
/* * The page is locked, so these buffers are protected from * any VM or truncate activity. Hence we don't need to care * for the buffer_head refcounts. */ |
a4b0672db fs: fix nobh erro... |
2560 |
for (bh = head; bh; bh = bh->b_this_page) { |
1da177e4c Linux-2.6.12-rc2 |
2561 2562 2563 |
wait_on_buffer(bh); if (!buffer_uptodate(bh)) ret = -EIO; |
1da177e4c Linux-2.6.12-rc2 |
2564 2565 2566 2567 2568 2569 2570 |
} if (ret) goto failed; } if (is_mapped_to_disk) SetPageMappedToDisk(page); |
1da177e4c Linux-2.6.12-rc2 |
2571 |
|
03158cd7e fs: restore nobh |
2572 |
*fsdata = head; /* to be released by nobh_write_end */ |
a4b0672db fs: fix nobh erro... |
2573 |
|
1da177e4c Linux-2.6.12-rc2 |
2574 2575 2576 |
return 0; failed: |
03158cd7e fs: restore nobh |
2577 |
BUG_ON(!ret); |
1da177e4c Linux-2.6.12-rc2 |
2578 |
/* |
a4b0672db fs: fix nobh erro... |
2579 2580 2581 2582 2583 |
* Error recovery is a bit difficult. We need to zero out blocks that * were newly allocated, and dirty them to ensure they get written out. * Buffers need to be attached to the page at this point, otherwise * the handling of potential IO errors during writeout would be hard * (could try doing synchronous writeout, but what if that fails too?) |
1da177e4c Linux-2.6.12-rc2 |
2584 |
*/ |
03158cd7e fs: restore nobh |
2585 2586 |
attach_nobh_buffers(page, head); page_zero_new_buffers(page, from, to); |
a4b0672db fs: fix nobh erro... |
2587 |
|
03158cd7e fs: restore nobh |
2588 2589 2590 2591 |
out_release: unlock_page(page); page_cache_release(page); *pagep = NULL; |
a4b0672db fs: fix nobh erro... |
2592 |
|
7bb46a673 fs: introduce new... |
2593 2594 |
return ret; } |
03158cd7e fs: restore nobh |
2595 |
EXPORT_SYMBOL(nobh_write_begin); |
1da177e4c Linux-2.6.12-rc2 |
2596 |
|
03158cd7e fs: restore nobh |
2597 2598 2599 |
int nobh_write_end(struct file *file, struct address_space *mapping, loff_t pos, unsigned len, unsigned copied, struct page *page, void *fsdata) |
1da177e4c Linux-2.6.12-rc2 |
2600 2601 |
{ struct inode *inode = page->mapping->host; |
efdc31319 nobh: nobh_write_... |
2602 |
struct buffer_head *head = fsdata; |
03158cd7e fs: restore nobh |
2603 |
struct buffer_head *bh; |
5b41e74ad vfs: fix data lea... |
2604 |
BUG_ON(fsdata != NULL && page_has_buffers(page)); |
1da177e4c Linux-2.6.12-rc2 |
2605 |
|
d4cf109f0 vfs: Don't call a... |
2606 |
if (unlikely(copied < len) && head) |
5b41e74ad vfs: fix data lea... |
2607 2608 2609 2610 |
attach_nobh_buffers(page, head); if (page_has_buffers(page)) return generic_write_end(file, mapping, pos, len, copied, page, fsdata); |
a4b0672db fs: fix nobh erro... |
2611 |
|
22c8ca78f [PATCH] fs: fix n... |
2612 |
SetPageUptodate(page); |
1da177e4c Linux-2.6.12-rc2 |
2613 |
set_page_dirty(page); |
03158cd7e fs: restore nobh |
2614 2615 |
if (pos+copied > inode->i_size) { i_size_write(inode, pos+copied); |
1da177e4c Linux-2.6.12-rc2 |
2616 2617 |
mark_inode_dirty(inode); } |
03158cd7e fs: restore nobh |
2618 2619 2620 |
unlock_page(page); page_cache_release(page); |
03158cd7e fs: restore nobh |
2621 2622 2623 2624 2625 2626 2627 |
while (head) { bh = head; head = head->b_this_page; free_buffer_head(bh); } return copied; |
1da177e4c Linux-2.6.12-rc2 |
2628 |
} |
03158cd7e fs: restore nobh |
2629 |
EXPORT_SYMBOL(nobh_write_end); |
1da177e4c Linux-2.6.12-rc2 |
2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 |
/* * nobh_writepage() - based on block_full_write_page() except * that it tries to operate without attaching bufferheads to * the page. */ int nobh_writepage(struct page *page, get_block_t *get_block, struct writeback_control *wbc) { struct inode * const inode = page->mapping->host; loff_t i_size = i_size_read(inode); const pgoff_t end_index = i_size >> PAGE_CACHE_SHIFT; unsigned offset; |
1da177e4c Linux-2.6.12-rc2 |
2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 |
int ret; /* Is the page fully inside i_size? */ if (page->index < end_index) goto out; /* Is the page fully outside i_size? (truncate in progress) */ offset = i_size & (PAGE_CACHE_SIZE-1); if (page->index >= end_index+1 || !offset) { /* * The page may have dirty, unmapped buffers. For example, * they may have been added in ext3_writepage(). Make them * freeable here, so the page does not leak. */ #if 0 /* Not really sure about this - do we need this ? */ if (page->mapping->a_ops->invalidatepage) page->mapping->a_ops->invalidatepage(page, offset); #endif unlock_page(page); return 0; /* don't care */ } /* * The page straddles i_size. It must be zeroed out on each and every * writepage invocation because it may be mmapped. "A file is mapped * in multiples of the page size. For a file that is not a multiple of * the page size, the remaining memory is zeroed when mapped, and * writes to that region are not written out to the file." */ |
eebd2aa35 Pagecache zeroing... |
2673 |
zero_user_segment(page, offset, PAGE_CACHE_SIZE); |
1da177e4c Linux-2.6.12-rc2 |
2674 2675 2676 |
out: ret = mpage_writepage(page, get_block, wbc); if (ret == -EAGAIN) |
35c80d5f4 Add block_write_f... |
2677 2678 |
ret = __block_write_full_page(inode, page, get_block, wbc, end_buffer_async_write); |
1da177e4c Linux-2.6.12-rc2 |
2679 2680 2681 |
return ret; } EXPORT_SYMBOL(nobh_writepage); |
03158cd7e fs: restore nobh |
2682 2683 |
int nobh_truncate_page(struct address_space *mapping, loff_t from, get_block_t *get_block) |
1da177e4c Linux-2.6.12-rc2 |
2684 |
{ |
1da177e4c Linux-2.6.12-rc2 |
2685 2686 |
pgoff_t index = from >> PAGE_CACHE_SHIFT; unsigned offset = from & (PAGE_CACHE_SIZE-1); |
03158cd7e fs: restore nobh |
2687 2688 2689 2690 |
unsigned blocksize; sector_t iblock; unsigned length, pos; struct inode *inode = mapping->host; |
1da177e4c Linux-2.6.12-rc2 |
2691 |
struct page *page; |
03158cd7e fs: restore nobh |
2692 2693 |
struct buffer_head map_bh; int err; |
1da177e4c Linux-2.6.12-rc2 |
2694 |
|
03158cd7e fs: restore nobh |
2695 2696 2697 2698 2699 2700 2701 2702 2703 |
blocksize = 1 << inode->i_blkbits; length = offset & (blocksize - 1); /* Block boundary? Nothing to do */ if (!length) return 0; length = blocksize - length; iblock = (sector_t)index << (PAGE_CACHE_SHIFT - inode->i_blkbits); |
1da177e4c Linux-2.6.12-rc2 |
2704 |
|
1da177e4c Linux-2.6.12-rc2 |
2705 |
page = grab_cache_page(mapping, index); |
03158cd7e fs: restore nobh |
2706 |
err = -ENOMEM; |
1da177e4c Linux-2.6.12-rc2 |
2707 2708 |
if (!page) goto out; |
03158cd7e fs: restore nobh |
2709 2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 |
if (page_has_buffers(page)) { has_buffers: unlock_page(page); page_cache_release(page); return block_truncate_page(mapping, from, get_block); } /* Find the buffer that contains "offset" */ pos = blocksize; while (offset >= pos) { iblock++; pos += blocksize; } |
460bcf57b Fix nobh_truncate... |
2722 2723 |
map_bh.b_size = blocksize; map_bh.b_state = 0; |
03158cd7e fs: restore nobh |
2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 |
err = get_block(inode, iblock, &map_bh, 0); if (err) goto unlock; /* unmapped? It's a hole - nothing to do */ if (!buffer_mapped(&map_bh)) goto unlock; /* Ok, it's mapped. Make sure it's up-to-date */ if (!PageUptodate(page)) { err = mapping->a_ops->readpage(NULL, page); if (err) { page_cache_release(page); goto out; } lock_page(page); if (!PageUptodate(page)) { err = -EIO; goto unlock; } if (page_has_buffers(page)) goto has_buffers; |
1da177e4c Linux-2.6.12-rc2 |
2745 |
} |
eebd2aa35 Pagecache zeroing... |
2746 |
zero_user(page, offset, length); |
03158cd7e fs: restore nobh |
2747 2748 2749 2750 |
set_page_dirty(page); err = 0; unlock: |
1da177e4c Linux-2.6.12-rc2 |
2751 2752 2753 |
unlock_page(page); page_cache_release(page); out: |
03158cd7e fs: restore nobh |
2754 |
return err; |
1da177e4c Linux-2.6.12-rc2 |
2755 2756 2757 2758 2759 2760 2761 2762 2763 |
} EXPORT_SYMBOL(nobh_truncate_page); int block_truncate_page(struct address_space *mapping, loff_t from, get_block_t *get_block) { pgoff_t index = from >> PAGE_CACHE_SHIFT; unsigned offset = from & (PAGE_CACHE_SIZE-1); unsigned blocksize; |
54b21a799 [PATCH] fix possi... |
2764 |
sector_t iblock; |
1da177e4c Linux-2.6.12-rc2 |
2765 2766 2767 2768 |
unsigned length, pos; struct inode *inode = mapping->host; struct page *page; struct buffer_head *bh; |
1da177e4c Linux-2.6.12-rc2 |
2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 |
int err; blocksize = 1 << inode->i_blkbits; length = offset & (blocksize - 1); /* Block boundary? Nothing to do */ if (!length) return 0; length = blocksize - length; |
54b21a799 [PATCH] fix possi... |
2779 |
iblock = (sector_t)index << (PAGE_CACHE_SHIFT - inode->i_blkbits); |
1da177e4c Linux-2.6.12-rc2 |
2780 2781 2782 2783 2784 2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 2795 2796 2797 2798 2799 |
page = grab_cache_page(mapping, index); err = -ENOMEM; if (!page) goto out; if (!page_has_buffers(page)) create_empty_buffers(page, blocksize, 0); /* Find the buffer that contains "offset" */ bh = page_buffers(page); pos = blocksize; while (offset >= pos) { bh = bh->b_this_page; iblock++; pos += blocksize; } err = 0; if (!buffer_mapped(bh)) { |
b0cf2321c [PATCH] pass b_si... |
2800 |
WARN_ON(bh->b_size != blocksize); |
1da177e4c Linux-2.6.12-rc2 |
2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811 |
err = get_block(inode, iblock, bh, 0); if (err) goto unlock; /* unmapped? It's a hole - nothing to do */ if (!buffer_mapped(bh)) goto unlock; } /* Ok, it's mapped. Make sure it's up-to-date */ if (PageUptodate(page)) set_buffer_uptodate(bh); |
33a266dda [PATCH] Make BH_U... |
2812 |
if (!buffer_uptodate(bh) && !buffer_delay(bh) && !buffer_unwritten(bh)) { |
1da177e4c Linux-2.6.12-rc2 |
2813 2814 2815 2816 2817 2818 2819 |
err = -EIO; ll_rw_block(READ, 1, &bh); wait_on_buffer(bh); /* Uhhuh. Read error. Complain and punt. */ if (!buffer_uptodate(bh)) goto unlock; } |
eebd2aa35 Pagecache zeroing... |
2820 |
zero_user(page, offset, length); |
1da177e4c Linux-2.6.12-rc2 |
2821 2822 2823 2824 2825 2826 2827 2828 2829 |
mark_buffer_dirty(bh); err = 0; unlock: unlock_page(page); page_cache_release(page); out: return err; } |
1fe72eaa0 fs/buffer.c: clea... |
2830 |
EXPORT_SYMBOL(block_truncate_page); |
1da177e4c Linux-2.6.12-rc2 |
2831 2832 2833 |
/* * The generic ->writepage function for buffer-backed address_spaces |
35c80d5f4 Add block_write_f... |
2834 |
* this form passes in the end_io handler used to finish the IO. |
1da177e4c Linux-2.6.12-rc2 |
2835 |
*/ |
35c80d5f4 Add block_write_f... |
2836 2837 |
int block_write_full_page_endio(struct page *page, get_block_t *get_block, struct writeback_control *wbc, bh_end_io_t *handler) |
1da177e4c Linux-2.6.12-rc2 |
2838 2839 2840 2841 2842 |
{ struct inode * const inode = page->mapping->host; loff_t i_size = i_size_read(inode); const pgoff_t end_index = i_size >> PAGE_CACHE_SHIFT; unsigned offset; |
1da177e4c Linux-2.6.12-rc2 |
2843 2844 2845 |
/* Is the page fully inside i_size? */ if (page->index < end_index) |
35c80d5f4 Add block_write_f... |
2846 2847 |
return __block_write_full_page(inode, page, get_block, wbc, handler); |
1da177e4c Linux-2.6.12-rc2 |
2848 2849 2850 2851 2852 2853 2854 2855 2856 |
/* Is the page fully outside i_size? (truncate in progress) */ offset = i_size & (PAGE_CACHE_SIZE-1); if (page->index >= end_index+1 || !offset) { /* * The page may have dirty, unmapped buffers. For example, * they may have been added in ext3_writepage(). Make them * freeable here, so the page does not leak. */ |
d47992f86 mm: change invali... |
2857 |
do_invalidatepage(page, 0, PAGE_CACHE_SIZE); |
1da177e4c Linux-2.6.12-rc2 |
2858 2859 2860 2861 2862 2863 |
unlock_page(page); return 0; /* don't care */ } /* * The page straddles i_size. It must be zeroed out on each and every |
2a61aa401 Fix misspellings ... |
2864 |
* writepage invocation because it may be mmapped. "A file is mapped |
1da177e4c Linux-2.6.12-rc2 |
2865 2866 2867 2868 |
* in multiples of the page size. For a file that is not a multiple of * the page size, the remaining memory is zeroed when mapped, and * writes to that region are not written out to the file." */ |
eebd2aa35 Pagecache zeroing... |
2869 |
zero_user_segment(page, offset, PAGE_CACHE_SIZE); |
35c80d5f4 Add block_write_f... |
2870 |
return __block_write_full_page(inode, page, get_block, wbc, handler); |
1da177e4c Linux-2.6.12-rc2 |
2871 |
} |
1fe72eaa0 fs/buffer.c: clea... |
2872 |
EXPORT_SYMBOL(block_write_full_page_endio); |
1da177e4c Linux-2.6.12-rc2 |
2873 |
|
35c80d5f4 Add block_write_f... |
2874 2875 2876 2877 2878 2879 2880 2881 2882 |
/* * The generic ->writepage function for buffer-backed address_spaces */ int block_write_full_page(struct page *page, get_block_t *get_block, struct writeback_control *wbc) { return block_write_full_page_endio(page, get_block, wbc, end_buffer_async_write); } |
1fe72eaa0 fs/buffer.c: clea... |
2883 |
EXPORT_SYMBOL(block_write_full_page); |
35c80d5f4 Add block_write_f... |
2884 |
|
1da177e4c Linux-2.6.12-rc2 |
2885 2886 2887 2888 2889 2890 2891 |
sector_t generic_block_bmap(struct address_space *mapping, sector_t block, get_block_t *get_block) { struct buffer_head tmp; struct inode *inode = mapping->host; tmp.b_state = 0; tmp.b_blocknr = 0; |
b0cf2321c [PATCH] pass b_si... |
2892 |
tmp.b_size = 1 << inode->i_blkbits; |
1da177e4c Linux-2.6.12-rc2 |
2893 2894 2895 |
get_block(inode, block, &tmp, 0); return tmp.b_blocknr; } |
1fe72eaa0 fs/buffer.c: clea... |
2896 |
EXPORT_SYMBOL(generic_block_bmap); |
1da177e4c Linux-2.6.12-rc2 |
2897 |
|
6712ecf8f Drop 'size' argum... |
2898 |
static void end_bio_bh_io_sync(struct bio *bio, int err) |
1da177e4c Linux-2.6.12-rc2 |
2899 2900 |
{ struct buffer_head *bh = bio->bi_private; |
1da177e4c Linux-2.6.12-rc2 |
2901 2902 |
if (err == -EOPNOTSUPP) { set_bit(BIO_EOPNOTSUPP, &bio->bi_flags); |
1da177e4c Linux-2.6.12-rc2 |
2903 |
} |
08bafc034 block: Supress Bu... |
2904 2905 |
if (unlikely (test_bit(BIO_QUIET,&bio->bi_flags))) set_bit(BH_Quiet, &bh->b_state); |
1da177e4c Linux-2.6.12-rc2 |
2906 2907 |
bh->b_end_io(bh, test_bit(BIO_UPTODATE, &bio->bi_flags)); bio_put(bio); |
1da177e4c Linux-2.6.12-rc2 |
2908 |
} |
57302e0dd vfs: avoid "attem... |
2909 2910 2911 2912 2913 2914 2915 2916 2917 2918 2919 2920 2921 2922 2923 2924 2925 2926 2927 2928 2929 2930 2931 2932 2933 2934 |
/* * This allows us to do IO even on the odd last sectors * of a device, even if the bh block size is some multiple * of the physical sector size. * * We'll just truncate the bio to the size of the device, * and clear the end of the buffer head manually. * * Truly out-of-range accesses will turn into actual IO * errors, this only handles the "we need to be able to * do IO at the final sector" case. */ static void guard_bh_eod(int rw, struct bio *bio, struct buffer_head *bh) { sector_t maxsector; unsigned bytes; maxsector = i_size_read(bio->bi_bdev->bd_inode) >> 9; if (!maxsector) return; /* * If the *whole* IO is past the end of the device, * let it through, and the IO layer will turn it into * an EIO. */ |
4f024f379 block: Abstract o... |
2935 |
if (unlikely(bio->bi_iter.bi_sector >= maxsector)) |
57302e0dd vfs: avoid "attem... |
2936 |
return; |
4f024f379 block: Abstract o... |
2937 2938 |
maxsector -= bio->bi_iter.bi_sector; bytes = bio->bi_iter.bi_size; |
57302e0dd vfs: avoid "attem... |
2939 2940 2941 2942 2943 2944 2945 |
if (likely((bytes >> 9) <= maxsector)) return; /* Uhhuh. We've got a bh that straddles the device size! */ bytes = maxsector << 9; /* Truncate the bio.. */ |
4f024f379 block: Abstract o... |
2946 |
bio->bi_iter.bi_size = bytes; |
57302e0dd vfs: avoid "attem... |
2947 2948 2949 |
bio->bi_io_vec[0].bv_len = bytes; /* ..and clear the end of the buffer for reads */ |
27d7c2a00 vfs: clear to the... |
2950 |
if ((rw & RW_MASK) == READ) { |
57302e0dd vfs: avoid "attem... |
2951 2952 2953 |
void *kaddr = kmap_atomic(bh->b_page); memset(kaddr + bh_offset(bh) + bytes, 0, bh->b_size - bytes); kunmap_atomic(kaddr); |
6d283dba3 vfs: add missing ... |
2954 |
flush_dcache_page(bh->b_page); |
57302e0dd vfs: avoid "attem... |
2955 2956 |
} } |
713685111 mm: make snapshot... |
2957 |
int _submit_bh(int rw, struct buffer_head *bh, unsigned long bio_flags) |
1da177e4c Linux-2.6.12-rc2 |
2958 2959 2960 2961 2962 2963 2964 |
{ struct bio *bio; int ret = 0; BUG_ON(!buffer_locked(bh)); BUG_ON(!buffer_mapped(bh)); BUG_ON(!bh->b_end_io); |
8fb0e3424 vfs: Add BUG_ON f... |
2965 2966 |
BUG_ON(buffer_delay(bh)); BUG_ON(buffer_unwritten(bh)); |
1da177e4c Linux-2.6.12-rc2 |
2967 |
|
48fd4f93a block: submit_bh(... |
2968 |
/* |
48fd4f93a block: submit_bh(... |
2969 |
* Only clear out a write error when rewriting |
1da177e4c Linux-2.6.12-rc2 |
2970 |
*/ |
48fd4f93a block: submit_bh(... |
2971 |
if (test_set_buffer_req(bh) && (rw & WRITE)) |
1da177e4c Linux-2.6.12-rc2 |
2972 2973 2974 2975 2976 2977 2978 |
clear_buffer_write_io_error(bh); /* * from here on down, it's all bio -- do the initial mapping, * submit_bio -> generic_make_request may further map this bio around */ bio = bio_alloc(GFP_NOIO, 1); |
4f024f379 block: Abstract o... |
2979 |
bio->bi_iter.bi_sector = bh->b_blocknr * (bh->b_size >> 9); |
1da177e4c Linux-2.6.12-rc2 |
2980 2981 2982 2983 2984 2985 |
bio->bi_bdev = bh->b_bdev; bio->bi_io_vec[0].bv_page = bh->b_page; bio->bi_io_vec[0].bv_len = bh->b_size; bio->bi_io_vec[0].bv_offset = bh_offset(bh); bio->bi_vcnt = 1; |
4f024f379 block: Abstract o... |
2986 |
bio->bi_iter.bi_size = bh->b_size; |
1da177e4c Linux-2.6.12-rc2 |
2987 2988 2989 |
bio->bi_end_io = end_bio_bh_io_sync; bio->bi_private = bh; |
713685111 mm: make snapshot... |
2990 |
bio->bi_flags |= bio_flags; |
1da177e4c Linux-2.6.12-rc2 |
2991 |
|
57302e0dd vfs: avoid "attem... |
2992 2993 |
/* Take care of bh's that straddle the end of the device */ guard_bh_eod(rw, bio, bh); |
877f962c5 buffer: add BH_Pr... |
2994 2995 2996 2997 |
if (buffer_meta(bh)) rw |= REQ_META; if (buffer_prio(bh)) rw |= REQ_PRIO; |
1da177e4c Linux-2.6.12-rc2 |
2998 2999 3000 3001 3002 3003 3004 3005 3006 |
bio_get(bio); submit_bio(rw, bio); if (bio_flagged(bio, BIO_EOPNOTSUPP)) ret = -EOPNOTSUPP; bio_put(bio); return ret; } |
713685111 mm: make snapshot... |
3007 3008 3009 3010 3011 3012 |
EXPORT_SYMBOL_GPL(_submit_bh); int submit_bh(int rw, struct buffer_head *bh) { return _submit_bh(rw, bh, 0); } |
1fe72eaa0 fs/buffer.c: clea... |
3013 |
EXPORT_SYMBOL(submit_bh); |
1da177e4c Linux-2.6.12-rc2 |
3014 3015 3016 |
/** * ll_rw_block: low-level access to block devices (DEPRECATED) |
9cb569d60 remove SWRITE* I/... |
3017 |
* @rw: whether to %READ or %WRITE or maybe %READA (readahead) |
1da177e4c Linux-2.6.12-rc2 |
3018 3019 3020 |
* @nr: number of &struct buffer_heads in the array * @bhs: array of pointers to &struct buffer_head * |
a76622362 [PATCH] Make ll_r... |
3021 3022 |
* ll_rw_block() takes an array of pointers to &struct buffer_heads, and * requests an I/O operation on them, either a %READ or a %WRITE. The third |
9cb569d60 remove SWRITE* I/... |
3023 3024 |
* %READA option is described in the documentation for generic_make_request() * which ll_rw_block() calls. |
1da177e4c Linux-2.6.12-rc2 |
3025 3026 |
* * This function drops any buffer that it cannot get a lock on (with the |
9cb569d60 remove SWRITE* I/... |
3027 3028 3029 3030 3031 |
* BH_Lock state bit), any buffer that appears to be clean when doing a write * request, and any buffer that appears to be up-to-date when doing read * request. Further it marks as clean buffers that are processed for * writing (the buffer cache won't assume that they are actually clean * until the buffer gets unlocked). |
1da177e4c Linux-2.6.12-rc2 |
3032 3033 |
* * ll_rw_block sets b_end_io to simple completion handler that marks |
e227867f1 treewide: Fix typ... |
3034 |
* the buffer up-to-date (if appropriate), unlocks the buffer and wakes |
1da177e4c Linux-2.6.12-rc2 |
3035 3036 3037 3038 3039 3040 3041 3042 3043 3044 3045 |
* any waiters. * * All of the buffers must be for the same device, and must also be a * multiple of the current approved size for the device. */ void ll_rw_block(int rw, int nr, struct buffer_head *bhs[]) { int i; for (i = 0; i < nr; i++) { struct buffer_head *bh = bhs[i]; |
9cb569d60 remove SWRITE* I/... |
3046 |
if (!trylock_buffer(bh)) |
1da177e4c Linux-2.6.12-rc2 |
3047 |
continue; |
9cb569d60 remove SWRITE* I/... |
3048 |
if (rw == WRITE) { |
1da177e4c Linux-2.6.12-rc2 |
3049 |
if (test_clear_buffer_dirty(bh)) { |
76c3073a8 [PATCH] end_buffe... |
3050 |
bh->b_end_io = end_buffer_write_sync; |
e60e5c50a [PATCH] Trivial o... |
3051 |
get_bh(bh); |
9cb569d60 remove SWRITE* I/... |
3052 |
submit_bh(WRITE, bh); |
1da177e4c Linux-2.6.12-rc2 |
3053 3054 3055 |
continue; } } else { |
1da177e4c Linux-2.6.12-rc2 |
3056 |
if (!buffer_uptodate(bh)) { |
76c3073a8 [PATCH] end_buffe... |
3057 |
bh->b_end_io = end_buffer_read_sync; |
e60e5c50a [PATCH] Trivial o... |
3058 |
get_bh(bh); |
1da177e4c Linux-2.6.12-rc2 |
3059 3060 3061 3062 3063 |
submit_bh(rw, bh); continue; } } unlock_buffer(bh); |
1da177e4c Linux-2.6.12-rc2 |
3064 3065 |
} } |
1fe72eaa0 fs/buffer.c: clea... |
3066 |
EXPORT_SYMBOL(ll_rw_block); |
1da177e4c Linux-2.6.12-rc2 |
3067 |
|
9cb569d60 remove SWRITE* I/... |
3068 3069 3070 3071 3072 3073 3074 3075 3076 3077 3078 3079 |
void write_dirty_buffer(struct buffer_head *bh, int rw) { lock_buffer(bh); if (!test_clear_buffer_dirty(bh)) { unlock_buffer(bh); return; } bh->b_end_io = end_buffer_write_sync; get_bh(bh); submit_bh(rw, bh); } EXPORT_SYMBOL(write_dirty_buffer); |
1da177e4c Linux-2.6.12-rc2 |
3080 3081 3082 3083 3084 |
/* * For a data-integrity writeout, we need to wait upon any in-progress I/O * and then start new I/O and then wait upon it. The caller must have a ref on * the buffer_head. */ |
87e99511e kill BH_Ordered flag |
3085 |
int __sync_dirty_buffer(struct buffer_head *bh, int rw) |
1da177e4c Linux-2.6.12-rc2 |
3086 3087 3088 3089 3090 3091 3092 3093 |
{ int ret = 0; WARN_ON(atomic_read(&bh->b_count) < 1); lock_buffer(bh); if (test_clear_buffer_dirty(bh)) { get_bh(bh); bh->b_end_io = end_buffer_write_sync; |
87e99511e kill BH_Ordered flag |
3094 |
ret = submit_bh(rw, bh); |
1da177e4c Linux-2.6.12-rc2 |
3095 |
wait_on_buffer(bh); |
1da177e4c Linux-2.6.12-rc2 |
3096 3097 3098 3099 3100 3101 3102 |
if (!ret && !buffer_uptodate(bh)) ret = -EIO; } else { unlock_buffer(bh); } return ret; } |
87e99511e kill BH_Ordered flag |
3103 3104 3105 3106 3107 3108 |
EXPORT_SYMBOL(__sync_dirty_buffer); int sync_dirty_buffer(struct buffer_head *bh) { return __sync_dirty_buffer(bh, WRITE_SYNC); } |
1fe72eaa0 fs/buffer.c: clea... |
3109 |
EXPORT_SYMBOL(sync_dirty_buffer); |
1da177e4c Linux-2.6.12-rc2 |
3110 3111 3112 3113 3114 3115 3116 3117 3118 3119 3120 3121 3122 3123 3124 3125 3126 3127 3128 3129 3130 3131 3132 3133 3134 3135 3136 3137 3138 3139 3140 3141 3142 3143 3144 |
/* * try_to_free_buffers() checks if all the buffers on this particular page * are unused, and releases them if so. * * Exclusion against try_to_free_buffers may be obtained by either * locking the page or by holding its mapping's private_lock. * * If the page is dirty but all the buffers are clean then we need to * be sure to mark the page clean as well. This is because the page * may be against a block device, and a later reattachment of buffers * to a dirty page will set *all* buffers dirty. Which would corrupt * filesystem data on the same device. * * The same applies to regular filesystem pages: if all the buffers are * clean then we set the page clean and proceed. To do that, we require * total exclusion from __set_page_dirty_buffers(). That is obtained with * private_lock. * * try_to_free_buffers() is non-blocking. */ static inline int buffer_busy(struct buffer_head *bh) { return atomic_read(&bh->b_count) | (bh->b_state & ((1 << BH_Dirty) | (1 << BH_Lock))); } static int drop_buffers(struct page *page, struct buffer_head **buffers_to_free) { struct buffer_head *head = page_buffers(page); struct buffer_head *bh; bh = head; do { |
de7d5a3b6 [PATCH] drop_buff... |
3145 |
if (buffer_write_io_error(bh) && page->mapping) |
1da177e4c Linux-2.6.12-rc2 |
3146 3147 3148 3149 3150 3151 3152 3153 |
set_bit(AS_EIO, &page->mapping->flags); if (buffer_busy(bh)) goto failed; bh = bh->b_this_page; } while (bh != head); do { struct buffer_head *next = bh->b_this_page; |
535ee2fbf buffer_head: fix ... |
3154 |
if (bh->b_assoc_map) |
1da177e4c Linux-2.6.12-rc2 |
3155 3156 3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 3167 3168 3169 3170 3171 |
__remove_assoc_queue(bh); bh = next; } while (bh != head); *buffers_to_free = head; __clear_page_buffers(page); return 1; failed: return 0; } int try_to_free_buffers(struct page *page) { struct address_space * const mapping = page->mapping; struct buffer_head *buffers_to_free = NULL; int ret = 0; BUG_ON(!PageLocked(page)); |
ecdfc9787 Resurrect 'try_to... |
3172 |
if (PageWriteback(page)) |
1da177e4c Linux-2.6.12-rc2 |
3173 3174 3175 3176 3177 3178 3179 3180 3181 |
return 0; if (mapping == NULL) { /* can this still happen? */ ret = drop_buffers(page, &buffers_to_free); goto out; } spin_lock(&mapping->private_lock); ret = drop_buffers(page, &buffers_to_free); |
ecdfc9787 Resurrect 'try_to... |
3182 3183 3184 3185 3186 3187 3188 3189 3190 3191 |
/* * If the filesystem writes its buffers by hand (eg ext3) * then we can have clean buffers against a dirty page. We * clean the page here; otherwise the VM will never notice * that the filesystem did any IO at all. * * Also, during truncate, discard_buffer will have marked all * the page's buffers clean. We discover that here and clean * the page also. |
87df7241b [PATCH] Fix try_t... |
3192 3193 3194 3195 |
* * private_lock must be held over this entire operation in order * to synchronise against __set_page_dirty_buffers and prevent the * dirty bit from being lost. |
ecdfc9787 Resurrect 'try_to... |
3196 3197 3198 |
*/ if (ret) cancel_dirty_page(page, PAGE_CACHE_SIZE); |
87df7241b [PATCH] Fix try_t... |
3199 |
spin_unlock(&mapping->private_lock); |
1da177e4c Linux-2.6.12-rc2 |
3200 3201 3202 3203 3204 3205 3206 3207 3208 3209 3210 3211 3212 |
out: if (buffers_to_free) { struct buffer_head *bh = buffers_to_free; do { struct buffer_head *next = bh->b_this_page; free_buffer_head(bh); bh = next; } while (bh != buffers_to_free); } return ret; } EXPORT_SYMBOL(try_to_free_buffers); |
1da177e4c Linux-2.6.12-rc2 |
3213 3214 3215 3216 3217 |
/* * There are no bdflush tunables left. But distributions are * still running obsolete flush daemons, so we terminate them here. * * Use of bdflush() is deprecated and will be removed in a future kernel. |
5b0830cb9 writeback: get ri... |
3218 |
* The `flush-X' kernel threads fully replace bdflush daemons and this call. |
1da177e4c Linux-2.6.12-rc2 |
3219 |
*/ |
bdc480e3b [CVE-2009-0029] S... |
3220 |
SYSCALL_DEFINE2(bdflush, int, func, long, data) |
1da177e4c Linux-2.6.12-rc2 |
3221 3222 3223 3224 3225 3226 3227 3228 3229 3230 3231 3232 3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 |
{ static int msg_count; if (!capable(CAP_SYS_ADMIN)) return -EPERM; if (msg_count < 5) { msg_count++; printk(KERN_INFO "warning: process `%s' used the obsolete bdflush" " system call ", current->comm); printk(KERN_INFO "Fix your initscripts? "); } if (func == 1) do_exit(0); return 0; } /* * Buffer-head allocation */ |
a0a9b0433 fs: Move bh_cache... |
3245 |
static struct kmem_cache *bh_cachep __read_mostly; |
1da177e4c Linux-2.6.12-rc2 |
3246 3247 3248 3249 3250 |
/* * Once the number of bh's in the machine exceeds this level, we start * stripping them in writeback. */ |
43be594a6 fs/buffer.c: chan... |
3251 |
static unsigned long max_buffer_heads; |
1da177e4c Linux-2.6.12-rc2 |
3252 3253 3254 3255 3256 3257 3258 3259 3260 3261 3262 3263 3264 3265 |
int buffer_heads_over_limit; struct bh_accounting { int nr; /* Number of live bh's */ int ratelimit; /* Limit cacheline bouncing */ }; static DEFINE_PER_CPU(struct bh_accounting, bh_accounting) = {0, 0}; static void recalc_bh_state(void) { int i; int tot = 0; |
ee1be8626 fs: Use this_cpu_... |
3266 |
if (__this_cpu_inc_return(bh_accounting.ratelimit) - 1 < 4096) |
1da177e4c Linux-2.6.12-rc2 |
3267 |
return; |
c7b92516a fs: Use this_cpu_... |
3268 |
__this_cpu_write(bh_accounting.ratelimit, 0); |
8a1434268 [PATCH] HOTPLUG_C... |
3269 |
for_each_online_cpu(i) |
1da177e4c Linux-2.6.12-rc2 |
3270 3271 3272 |
tot += per_cpu(bh_accounting, i).nr; buffer_heads_over_limit = (tot > max_buffer_heads); } |
c7b92516a fs: Use this_cpu_... |
3273 |
|
dd0fc66fb [PATCH] gfp flags... |
3274 |
struct buffer_head *alloc_buffer_head(gfp_t gfp_flags) |
1da177e4c Linux-2.6.12-rc2 |
3275 |
{ |
019b4d123 fs: buffer_head: ... |
3276 |
struct buffer_head *ret = kmem_cache_zalloc(bh_cachep, gfp_flags); |
1da177e4c Linux-2.6.12-rc2 |
3277 |
if (ret) { |
a35afb830 Remove SLAB_CTOR_... |
3278 |
INIT_LIST_HEAD(&ret->b_assoc_buffers); |
c7b92516a fs: Use this_cpu_... |
3279 3280 |
preempt_disable(); __this_cpu_inc(bh_accounting.nr); |
1da177e4c Linux-2.6.12-rc2 |
3281 |
recalc_bh_state(); |
c7b92516a fs: Use this_cpu_... |
3282 |
preempt_enable(); |
1da177e4c Linux-2.6.12-rc2 |
3283 3284 3285 3286 3287 3288 3289 3290 3291 |
} return ret; } EXPORT_SYMBOL(alloc_buffer_head); void free_buffer_head(struct buffer_head *bh) { BUG_ON(!list_empty(&bh->b_assoc_buffers)); kmem_cache_free(bh_cachep, bh); |
c7b92516a fs: Use this_cpu_... |
3292 3293 |
preempt_disable(); __this_cpu_dec(bh_accounting.nr); |
1da177e4c Linux-2.6.12-rc2 |
3294 |
recalc_bh_state(); |
c7b92516a fs: Use this_cpu_... |
3295 |
preempt_enable(); |
1da177e4c Linux-2.6.12-rc2 |
3296 3297 |
} EXPORT_SYMBOL(free_buffer_head); |
1da177e4c Linux-2.6.12-rc2 |
3298 3299 3300 3301 3302 3303 3304 3305 3306 |
static void buffer_exit_cpu(int cpu) { int i; struct bh_lru *b = &per_cpu(bh_lrus, cpu); for (i = 0; i < BH_LRU_SIZE; i++) { brelse(b->bhs[i]); b->bhs[i] = NULL; } |
c7b92516a fs: Use this_cpu_... |
3307 |
this_cpu_add(bh_accounting.nr, per_cpu(bh_accounting, cpu).nr); |
8a1434268 [PATCH] HOTPLUG_C... |
3308 |
per_cpu(bh_accounting, cpu).nr = 0; |
1da177e4c Linux-2.6.12-rc2 |
3309 3310 3311 3312 3313 |
} static int buffer_cpu_notify(struct notifier_block *self, unsigned long action, void *hcpu) { |
8bb784428 Add suspend-relat... |
3314 |
if (action == CPU_DEAD || action == CPU_DEAD_FROZEN) |
1da177e4c Linux-2.6.12-rc2 |
3315 3316 3317 |
buffer_exit_cpu((unsigned long)hcpu); return NOTIFY_OK; } |
1da177e4c Linux-2.6.12-rc2 |
3318 |
|
389d1b083 Add buffer head r... |
3319 |
/** |
a6b91919e fs: fix kernel-do... |
3320 |
* bh_uptodate_or_lock - Test whether the buffer is uptodate |
389d1b083 Add buffer head r... |
3321 3322 3323 3324 3325 3326 3327 3328 3329 3330 3331 3332 3333 3334 3335 3336 3337 3338 |
* @bh: struct buffer_head * * Return true if the buffer is up-to-date and false, * with the buffer locked, if not. */ int bh_uptodate_or_lock(struct buffer_head *bh) { if (!buffer_uptodate(bh)) { lock_buffer(bh); if (!buffer_uptodate(bh)) return 0; unlock_buffer(bh); } return 1; } EXPORT_SYMBOL(bh_uptodate_or_lock); /** |
a6b91919e fs: fix kernel-do... |
3339 |
* bh_submit_read - Submit a locked buffer for reading |
389d1b083 Add buffer head r... |
3340 3341 3342 3343 3344 3345 3346 3347 3348 3349 3350 3351 3352 3353 3354 3355 3356 3357 3358 3359 3360 3361 |
* @bh: struct buffer_head * * Returns zero on success and -EIO on error. */ int bh_submit_read(struct buffer_head *bh) { BUG_ON(!buffer_locked(bh)); if (buffer_uptodate(bh)) { unlock_buffer(bh); return 0; } get_bh(bh); bh->b_end_io = end_buffer_read_sync; submit_bh(READ, bh); wait_on_buffer(bh); if (buffer_uptodate(bh)) return 0; return -EIO; } EXPORT_SYMBOL(bh_submit_read); |
1da177e4c Linux-2.6.12-rc2 |
3362 3363 |
void __init buffer_init(void) { |
43be594a6 fs/buffer.c: chan... |
3364 |
unsigned long nrpages; |
1da177e4c Linux-2.6.12-rc2 |
3365 |
|
b98938c37 bufferhead: rever... |
3366 3367 3368 3369 |
bh_cachep = kmem_cache_create("buffer_head", sizeof(struct buffer_head), 0, (SLAB_RECLAIM_ACCOUNT|SLAB_PANIC| SLAB_MEM_SPREAD), |
019b4d123 fs: buffer_head: ... |
3370 |
NULL); |
1da177e4c Linux-2.6.12-rc2 |
3371 3372 3373 3374 3375 3376 3377 3378 |
/* * Limit the bh occupancy to 10% of ZONE_NORMAL */ nrpages = (nr_free_buffer_pages() * 10) / 100; max_buffer_heads = nrpages * (PAGE_SIZE / sizeof(struct buffer_head)); hotcpu_notifier(buffer_cpu_notify, 0); } |