Blame view
fs/fs-writeback.c
35.4 KB
1da177e4c
|
1 2 3 4 5 6 7 8 9 10 |
/* * fs/fs-writeback.c * * Copyright (C) 2002, Linus Torvalds. * * Contains all the functions related to writing back and waiting * upon dirty inodes against superblocks, and writing back dirty * pages against inodes. ie: data writeback. Writeout of the * inode itself is not handled here. * |
e1f8e8744
|
11 |
* 10Apr2002 Andrew Morton |
1da177e4c
|
12 13 14 15 16 |
* Split out of fs/inode.c * Additions for address_space-based writeback */ #include <linux/kernel.h> |
f5ff8422b
|
17 |
#include <linux/module.h> |
1da177e4c
|
18 |
#include <linux/spinlock.h> |
5a0e3ad6a
|
19 |
#include <linux/slab.h> |
1da177e4c
|
20 21 22 |
#include <linux/sched.h> #include <linux/fs.h> #include <linux/mm.h> |
03ba3782e
|
23 24 |
#include <linux/kthread.h> #include <linux/freezer.h> |
1da177e4c
|
25 26 27 28 |
#include <linux/writeback.h> #include <linux/blkdev.h> #include <linux/backing-dev.h> #include <linux/buffer_head.h> |
455b28646
|
29 |
#include <linux/tracepoint.h> |
07f3f05c1
|
30 |
#include "internal.h" |
1da177e4c
|
31 |
|
d0bceac74
|
32 |
/* |
c4a77a6c7
|
33 34 |
* Passed into wb_writeback(), essentially a subset of writeback_control */ |
83ba7b071
|
35 |
struct wb_writeback_work { |
c4a77a6c7
|
36 37 38 |
long nr_pages; struct super_block *sb; enum writeback_sync_modes sync_mode; |
52957fe1c
|
39 40 41 |
unsigned int for_kupdate:1; unsigned int range_cyclic:1; unsigned int for_background:1; |
c4a77a6c7
|
42 |
|
8010c3b63
|
43 |
struct list_head list; /* pending work list */ |
83ba7b071
|
44 |
struct completion *done; /* set if the caller waits */ |
03ba3782e
|
45 |
}; |
455b28646
|
46 47 48 49 50 51 52 |
/* * Include the creation of the trace points after defining the * wb_writeback_work structure so that the definition remains local to this * file. */ #define CREATE_TRACE_POINTS #include <trace/events/writeback.h> |
455b28646
|
53 54 55 56 |
/* * We don't actually have pdflush, but this one is exported though /proc... */ int nr_pdflush_threads; |
f11b00f3b
|
57 58 59 60 |
/** * writeback_in_progress - determine whether there is writeback in progress * @bdi: the device's backing_dev_info structure. * |
03ba3782e
|
61 62 |
* Determine whether there is writeback waiting to be handled against a * backing device. |
f11b00f3b
|
63 64 65 |
*/ int writeback_in_progress(struct backing_dev_info *bdi) { |
81d73a32d
|
66 |
return test_bit(BDI_writeback_running, &bdi->state); |
f11b00f3b
|
67 |
} |
692ebd17c
|
68 69 70 |
static inline struct backing_dev_info *inode_to_bdi(struct inode *inode) { struct super_block *sb = inode->i_sb; |
692ebd17c
|
71 |
|
aaead25b9
|
72 73 74 75 |
if (strcmp(sb->s_type->name, "bdev") == 0) return inode->i_mapping->backing_dev_info; return sb->s_bdi; |
692ebd17c
|
76 |
} |
7ccf19a80
|
77 78 79 80 |
static inline struct inode *wb_inode(struct list_head *head) { return list_entry(head, struct inode, i_wb_list); } |
6585027a5
|
81 82 |
/* Wakeup flusher thread or forker thread to fork it. Requires bdi->wb_lock. */ static void bdi_wakeup_flusher(struct backing_dev_info *bdi) |
03ba3782e
|
83 |
{ |
fff5b85aa
|
84 85 86 87 88 89 90 |
if (bdi->wb.task) { wake_up_process(bdi->wb.task); } else { /* * The bdi thread isn't there, wake up the forker thread which * will create and run it. */ |
03ba3782e
|
91 |
wake_up_process(default_backing_dev_info.wb.task); |
1da177e4c
|
92 |
} |
6585027a5
|
93 94 95 96 97 98 99 100 101 102 103 104 |
} static void bdi_queue_work(struct backing_dev_info *bdi, struct wb_writeback_work *work) { trace_writeback_queue(bdi, work); spin_lock_bh(&bdi->wb_lock); list_add_tail(&work->list, &bdi->work_list); if (!bdi->wb.task) trace_writeback_nothread(bdi, work); bdi_wakeup_flusher(bdi); |
6467716a3
|
105 |
spin_unlock_bh(&bdi->wb_lock); |
1da177e4c
|
106 |
} |
83ba7b071
|
107 108 |
static void __bdi_start_writeback(struct backing_dev_info *bdi, long nr_pages, |
6585027a5
|
109 |
bool range_cyclic) |
1da177e4c
|
110 |
{ |
83ba7b071
|
111 |
struct wb_writeback_work *work; |
03ba3782e
|
112 |
|
bcddc3f01
|
113 114 115 116 |
/* * This is WB_SYNC_NONE writeback, so if allocation fails just * wakeup the thread for old dirty data writeback */ |
83ba7b071
|
117 118 |
work = kzalloc(sizeof(*work), GFP_ATOMIC); if (!work) { |
455b28646
|
119 120 |
if (bdi->wb.task) { trace_writeback_nowork(bdi); |
83ba7b071
|
121 |
wake_up_process(bdi->wb.task); |
455b28646
|
122 |
} |
83ba7b071
|
123 |
return; |
bcddc3f01
|
124 |
} |
03ba3782e
|
125 |
|
83ba7b071
|
126 127 128 |
work->sync_mode = WB_SYNC_NONE; work->nr_pages = nr_pages; work->range_cyclic = range_cyclic; |
03ba3782e
|
129 |
|
83ba7b071
|
130 |
bdi_queue_work(bdi, work); |
b6e51316d
|
131 132 133 134 135 136 137 138 139 |
} /** * bdi_start_writeback - start writeback * @bdi: the backing device to write from * @nr_pages: the number of pages to write * * Description: * This does WB_SYNC_NONE opportunistic writeback. The IO is only |
25985edce
|
140 |
* started when this function returns, we make no guarantees on |
0e3c9a228
|
141 |
* completion. Caller need not hold sb s_umount semaphore. |
b6e51316d
|
142 143 |
* */ |
c5444198c
|
144 |
void bdi_start_writeback(struct backing_dev_info *bdi, long nr_pages) |
b6e51316d
|
145 |
{ |
6585027a5
|
146 |
__bdi_start_writeback(bdi, nr_pages, true); |
c5444198c
|
147 |
} |
d3ddec763
|
148 |
|
c5444198c
|
149 150 151 152 153 |
/** * bdi_start_background_writeback - start background writeback * @bdi: the backing device to write from * * Description: |
6585027a5
|
154 155 156 157 |
* This makes sure WB_SYNC_NONE background writeback happens. When * this function returns, it is only guaranteed that for given BDI * some IO is happening if we are over background dirty threshold. * Caller need not hold sb s_umount semaphore. |
c5444198c
|
158 159 160 |
*/ void bdi_start_background_writeback(struct backing_dev_info *bdi) { |
6585027a5
|
161 162 163 164 |
/* * We just wake up the flusher thread. It will perform background * writeback as soon as there is no other work to do. */ |
71927e84e
|
165 |
trace_writeback_wake_background(bdi); |
6585027a5
|
166 167 168 |
spin_lock_bh(&bdi->wb_lock); bdi_wakeup_flusher(bdi); spin_unlock_bh(&bdi->wb_lock); |
1da177e4c
|
169 170 171 |
} /* |
a66979aba
|
172 173 174 175 176 177 178 179 180 181 182 |
* Remove the inode from the writeback list it is on. */ void inode_wb_list_del(struct inode *inode) { spin_lock(&inode_wb_list_lock); list_del_init(&inode->i_wb_list); spin_unlock(&inode_wb_list_lock); } /* |
6610a0bc8
|
183 184 185 186 |
* Redirty an inode: set its when-it-was dirtied timestamp and move it to the * furthest end of its superblock's dirty-inode list. * * Before stamping the inode's ->dirtied_when, we check to see whether it is |
66f3b8e2e
|
187 |
* already the most-recently-dirtied inode on the b_dirty list. If that is |
6610a0bc8
|
188 189 190 191 192 |
* the case then the inode must have been redirtied while it was being written * out and we don't reset its dirtied_when. */ static void redirty_tail(struct inode *inode) { |
03ba3782e
|
193 |
struct bdi_writeback *wb = &inode_to_bdi(inode)->wb; |
6610a0bc8
|
194 |
|
a66979aba
|
195 |
assert_spin_locked(&inode_wb_list_lock); |
03ba3782e
|
196 |
if (!list_empty(&wb->b_dirty)) { |
66f3b8e2e
|
197 |
struct inode *tail; |
6610a0bc8
|
198 |
|
7ccf19a80
|
199 |
tail = wb_inode(wb->b_dirty.next); |
66f3b8e2e
|
200 |
if (time_before(inode->dirtied_when, tail->dirtied_when)) |
6610a0bc8
|
201 202 |
inode->dirtied_when = jiffies; } |
7ccf19a80
|
203 |
list_move(&inode->i_wb_list, &wb->b_dirty); |
6610a0bc8
|
204 205 206 |
} /* |
66f3b8e2e
|
207 |
* requeue inode for re-scanning after bdi->b_io list is exhausted. |
c986d1e2a
|
208 |
*/ |
0e0f4fc22
|
209 |
static void requeue_io(struct inode *inode) |
c986d1e2a
|
210 |
{ |
03ba3782e
|
211 |
struct bdi_writeback *wb = &inode_to_bdi(inode)->wb; |
a66979aba
|
212 |
assert_spin_locked(&inode_wb_list_lock); |
7ccf19a80
|
213 |
list_move(&inode->i_wb_list, &wb->b_more_io); |
c986d1e2a
|
214 |
} |
1c0eeaf56
|
215 216 217 |
static void inode_sync_complete(struct inode *inode) { /* |
a66979aba
|
218 219 |
* Prevent speculative execution through * spin_unlock(&inode_wb_list_lock); |
1c0eeaf56
|
220 |
*/ |
a66979aba
|
221 |
|
1c0eeaf56
|
222 223 224 |
smp_mb(); wake_up_bit(&inode->i_state, __I_SYNC); } |
d2caa3c54
|
225 226 227 228 229 230 231 232 |
static bool inode_dirtied_after(struct inode *inode, unsigned long t) { bool ret = time_after(inode->dirtied_when, t); #ifndef CONFIG_64BIT /* * For inodes being constantly redirtied, dirtied_when can get stuck. * It _appears_ to be in the future, but is actually in distant past. * This test is necessary to prevent such wrapped-around relative times |
5b0830cb9
|
233 |
* from permanently stopping the whole bdi writeback. |
d2caa3c54
|
234 235 236 237 238 |
*/ ret = ret && time_before_eq(inode->dirtied_when, jiffies); #endif return ret; } |
c986d1e2a
|
239 |
/* |
2c1365791
|
240 241 242 243 244 245 |
* Move expired dirty inodes from @delaying_queue to @dispatch_queue. */ static void move_expired_inodes(struct list_head *delaying_queue, struct list_head *dispatch_queue, unsigned long *older_than_this) { |
5c03449d3
|
246 247 |
LIST_HEAD(tmp); struct list_head *pos, *node; |
cf137307c
|
248 |
struct super_block *sb = NULL; |
5c03449d3
|
249 |
struct inode *inode; |
cf137307c
|
250 |
int do_sb_sort = 0; |
5c03449d3
|
251 |
|
2c1365791
|
252 |
while (!list_empty(delaying_queue)) { |
7ccf19a80
|
253 |
inode = wb_inode(delaying_queue->prev); |
2c1365791
|
254 |
if (older_than_this && |
d2caa3c54
|
255 |
inode_dirtied_after(inode, *older_than_this)) |
2c1365791
|
256 |
break; |
cf137307c
|
257 258 259 |
if (sb && sb != inode->i_sb) do_sb_sort = 1; sb = inode->i_sb; |
7ccf19a80
|
260 |
list_move(&inode->i_wb_list, &tmp); |
5c03449d3
|
261 |
} |
cf137307c
|
262 263 264 265 266 |
/* just one sb in list, splice to dispatch_queue and we're done */ if (!do_sb_sort) { list_splice(&tmp, dispatch_queue); return; } |
5c03449d3
|
267 268 |
/* Move inodes from one superblock together */ while (!list_empty(&tmp)) { |
7ccf19a80
|
269 |
sb = wb_inode(tmp.prev)->i_sb; |
5c03449d3
|
270 |
list_for_each_prev_safe(pos, node, &tmp) { |
7ccf19a80
|
271 |
inode = wb_inode(pos); |
5c03449d3
|
272 |
if (inode->i_sb == sb) |
7ccf19a80
|
273 |
list_move(&inode->i_wb_list, dispatch_queue); |
5c03449d3
|
274 |
} |
2c1365791
|
275 276 277 278 279 |
} } /* * Queue all expired dirty inodes for io, eldest first. |
4ea879b96
|
280 281 282 283 284 285 286 287 |
* Before * newly dirtied b_dirty b_io b_more_io * =============> gf edc BA * After * newly dirtied b_dirty b_io b_more_io * =============> g fBAedc * | * +--> dequeue for IO |
2c1365791
|
288 |
*/ |
03ba3782e
|
289 |
static void queue_io(struct bdi_writeback *wb, unsigned long *older_than_this) |
66f3b8e2e
|
290 |
{ |
a66979aba
|
291 |
assert_spin_locked(&inode_wb_list_lock); |
4ea879b96
|
292 |
list_splice_init(&wb->b_more_io, &wb->b_io); |
03ba3782e
|
293 |
move_expired_inodes(&wb->b_dirty, &wb->b_io, older_than_this); |
66f3b8e2e
|
294 |
} |
a9185b41a
|
295 |
static int write_inode(struct inode *inode, struct writeback_control *wbc) |
08d8e9749
|
296 |
{ |
03ba3782e
|
297 |
if (inode->i_sb->s_op->write_inode && !is_bad_inode(inode)) |
a9185b41a
|
298 |
return inode->i_sb->s_op->write_inode(inode, wbc); |
03ba3782e
|
299 |
return 0; |
08d8e9749
|
300 |
} |
08d8e9749
|
301 |
|
2c1365791
|
302 |
/* |
01c031945
|
303 304 305 306 307 308 309 310 |
* Wait for writeback on an inode to complete. */ static void inode_wait_for_writeback(struct inode *inode) { DEFINE_WAIT_BIT(wq, &inode->i_state, __I_SYNC); wait_queue_head_t *wqh; wqh = bit_waitqueue(&inode->i_state, __I_SYNC); |
250df6ed2
|
311 312 |
while (inode->i_state & I_SYNC) { spin_unlock(&inode->i_lock); |
a66979aba
|
313 |
spin_unlock(&inode_wb_list_lock); |
01c031945
|
314 |
__wait_on_bit(wqh, &wq, inode_wait, TASK_UNINTERRUPTIBLE); |
a66979aba
|
315 |
spin_lock(&inode_wb_list_lock); |
250df6ed2
|
316 |
spin_lock(&inode->i_lock); |
58a9d3d8d
|
317 |
} |
01c031945
|
318 319 320 |
} /* |
0f1b1fd86
|
321 322 323 |
* Write out an inode's dirty pages. Called under inode_wb_list_lock and * inode->i_lock. Either the caller has an active reference on the inode or * the inode has I_WILL_FREE set. |
01c031945
|
324 |
* |
1da177e4c
|
325 326 327 328 329 |
* If `wait' is set, wait on the writeout. * * The whole writeout design is quite complex and fragile. We want to avoid * starvation of particular inodes when others are being redirtied, prevent * livelocks, etc. |
1da177e4c
|
330 331 |
*/ static int |
01c031945
|
332 |
writeback_single_inode(struct inode *inode, struct writeback_control *wbc) |
1da177e4c
|
333 |
{ |
1da177e4c
|
334 |
struct address_space *mapping = inode->i_mapping; |
01c031945
|
335 |
unsigned dirty; |
1da177e4c
|
336 |
int ret; |
0f1b1fd86
|
337 338 |
assert_spin_locked(&inode_wb_list_lock); assert_spin_locked(&inode->i_lock); |
01c031945
|
339 340 341 342 343 344 345 346 |
if (!atomic_read(&inode->i_count)) WARN_ON(!(inode->i_state & (I_WILL_FREE|I_FREEING))); else WARN_ON(inode->i_state & I_WILL_FREE); if (inode->i_state & I_SYNC) { /* * If this inode is locked for writeback and we are not doing |
66f3b8e2e
|
347 |
* writeback-for-data-integrity, move it to b_more_io so that |
01c031945
|
348 349 350 |
* writeback can proceed with the other inodes on s_io. * * We'll have another go at writing back this inode when we |
66f3b8e2e
|
351 |
* completed a full scan of b_io. |
01c031945
|
352 |
*/ |
a9185b41a
|
353 |
if (wbc->sync_mode != WB_SYNC_ALL) { |
01c031945
|
354 355 356 357 358 359 360 361 362 |
requeue_io(inode); return 0; } /* * It's a data-integrity sync. We must wait. */ inode_wait_for_writeback(inode); } |
1c0eeaf56
|
363 |
BUG_ON(inode->i_state & I_SYNC); |
1da177e4c
|
364 |
|
5547e8aac
|
365 |
/* Set I_SYNC, reset I_DIRTY_PAGES */ |
1c0eeaf56
|
366 |
inode->i_state |= I_SYNC; |
5547e8aac
|
367 |
inode->i_state &= ~I_DIRTY_PAGES; |
250df6ed2
|
368 |
spin_unlock(&inode->i_lock); |
a66979aba
|
369 |
spin_unlock(&inode_wb_list_lock); |
1da177e4c
|
370 371 |
ret = do_writepages(mapping, wbc); |
26821ed40
|
372 373 374 375 376 |
/* * Make sure to wait on the data before writing out the metadata. * This is important for filesystems that modify metadata on data * I/O completion. */ |
a9185b41a
|
377 |
if (wbc->sync_mode == WB_SYNC_ALL) { |
26821ed40
|
378 |
int err = filemap_fdatawait(mapping); |
1da177e4c
|
379 380 381 |
if (ret == 0) ret = err; } |
5547e8aac
|
382 383 384 385 386 |
/* * Some filesystems may redirty the inode during the writeback * due to delalloc, clear dirty metadata flags right before * write_inode() */ |
250df6ed2
|
387 |
spin_lock(&inode->i_lock); |
5547e8aac
|
388 389 |
dirty = inode->i_state & I_DIRTY; inode->i_state &= ~(I_DIRTY_SYNC | I_DIRTY_DATASYNC); |
250df6ed2
|
390 |
spin_unlock(&inode->i_lock); |
26821ed40
|
391 392 |
/* Don't write the inode if only I_DIRTY_PAGES was set */ if (dirty & (I_DIRTY_SYNC | I_DIRTY_DATASYNC)) { |
a9185b41a
|
393 |
int err = write_inode(inode, wbc); |
1da177e4c
|
394 395 396 |
if (ret == 0) ret = err; } |
a66979aba
|
397 |
spin_lock(&inode_wb_list_lock); |
250df6ed2
|
398 |
spin_lock(&inode->i_lock); |
1c0eeaf56
|
399 |
inode->i_state &= ~I_SYNC; |
a4ffdde6e
|
400 |
if (!(inode->i_state & I_FREEING)) { |
23539afc7
|
401 |
if (mapping_tagged(mapping, PAGECACHE_TAG_DIRTY)) { |
1da177e4c
|
402 403 |
/* * We didn't write back all the pages. nfs_writepages() |
a50aeb401
|
404 |
* sometimes bales out without doing anything. |
1b43ef91d
|
405 |
*/ |
a50aeb401
|
406 407 |
inode->i_state |= I_DIRTY_PAGES; if (wbc->nr_to_write <= 0) { |
1da177e4c
|
408 |
/* |
a50aeb401
|
409 |
* slice used up: queue for next turn |
1da177e4c
|
410 |
*/ |
a50aeb401
|
411 |
requeue_io(inode); |
1da177e4c
|
412 413 |
} else { /* |
a50aeb401
|
414 415 416 417 418 |
* Writeback blocked by something other than * congestion. Delay the inode for some time to * avoid spinning on the CPU (100% iowait) * retrying writeback of the dirty page/inode * that cannot be performed immediately. |
1da177e4c
|
419 |
*/ |
1b43ef91d
|
420 |
redirty_tail(inode); |
1da177e4c
|
421 |
} |
23539afc7
|
422 423 424 425 426 427 428 429 |
} else if (inode->i_state & I_DIRTY) { /* * Filesystems can dirty the inode during writeback * operations, such as delayed allocation during * submission or metadata updates after data IO * completion. */ redirty_tail(inode); |
1da177e4c
|
430 431 |
} else { /* |
9e38d86ff
|
432 433 434 |
* The inode is clean. At this point we either have * a reference to the inode or it's on it's way out. * No need to add it back to the LRU. |
1da177e4c
|
435 |
*/ |
7ccf19a80
|
436 |
list_del_init(&inode->i_wb_list); |
1da177e4c
|
437 438 |
} } |
1c0eeaf56
|
439 |
inode_sync_complete(inode); |
1da177e4c
|
440 441 |
return ret; } |
03ba3782e
|
442 |
/* |
d19de7edf
|
443 |
* For background writeback the caller does not have the sb pinned |
03ba3782e
|
444 445 |
* before calling writeback. So make sure that we do pin it, so it doesn't * go away while we are writing inodes from it. |
03ba3782e
|
446 |
*/ |
d19de7edf
|
447 |
static bool pin_sb_for_writeback(struct super_block *sb) |
03ba3782e
|
448 |
{ |
03ba3782e
|
449 |
spin_lock(&sb_lock); |
29cb48594
|
450 451 452 453 |
if (list_empty(&sb->s_instances)) { spin_unlock(&sb_lock); return false; } |
03ba3782e
|
454 |
sb->s_count++; |
29cb48594
|
455 |
spin_unlock(&sb_lock); |
03ba3782e
|
456 |
if (down_read_trylock(&sb->s_umount)) { |
29cb48594
|
457 |
if (sb->s_root) |
d19de7edf
|
458 |
return true; |
03ba3782e
|
459 460 |
up_read(&sb->s_umount); } |
29cb48594
|
461 462 |
put_super(sb); |
d19de7edf
|
463 |
return false; |
03ba3782e
|
464 |
} |
f11c9c5c2
|
465 466 |
/* * Write a portion of b_io inodes which belong to @sb. |
edadfb10b
|
467 468 |
* * If @only_this_sb is true, then find and write all such |
f11c9c5c2
|
469 470 |
* inodes. Otherwise write only ones which go sequentially * in reverse order. |
edadfb10b
|
471 |
* |
f11c9c5c2
|
472 473 474 |
* Return 1, if the caller writeback routine should be * interrupted. Otherwise return 0. */ |
edadfb10b
|
475 476 |
static int writeback_sb_inodes(struct super_block *sb, struct bdi_writeback *wb, struct writeback_control *wbc, bool only_this_sb) |
1da177e4c
|
477 |
{ |
03ba3782e
|
478 |
while (!list_empty(&wb->b_io)) { |
1da177e4c
|
479 |
long pages_skipped; |
7ccf19a80
|
480 |
struct inode *inode = wb_inode(wb->b_io.prev); |
edadfb10b
|
481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 |
if (inode->i_sb != sb) { if (only_this_sb) { /* * We only want to write back data for this * superblock, move all inodes not belonging * to it back onto the dirty list. */ redirty_tail(inode); continue; } /* * The inode belongs to a different superblock. * Bounce back to the caller to unpin this and * pin the next superblock. */ |
f11c9c5c2
|
498 |
return 0; |
edadfb10b
|
499 |
} |
9843b76aa
|
500 501 502 503 504 |
/* * Don't bother with new inodes or inodes beeing freed, first * kind does not need peridic writeout yet, and for the latter * kind writeout is handled by the freer. */ |
250df6ed2
|
505 |
spin_lock(&inode->i_lock); |
9843b76aa
|
506 |
if (inode->i_state & (I_NEW | I_FREEING | I_WILL_FREE)) { |
250df6ed2
|
507 |
spin_unlock(&inode->i_lock); |
7ef0d7377
|
508 509 510 |
requeue_io(inode); continue; } |
9843b76aa
|
511 |
|
d2caa3c54
|
512 513 514 515 |
/* * Was this inode dirtied after sync_sb_inodes was called? * This keeps sync from extra jobs and livelock. */ |
250df6ed2
|
516 517 |
if (inode_dirtied_after(inode, wbc->wb_start)) { spin_unlock(&inode->i_lock); |
f11c9c5c2
|
518 |
return 1; |
250df6ed2
|
519 |
} |
1da177e4c
|
520 |
|
1da177e4c
|
521 |
__iget(inode); |
250df6ed2
|
522 |
|
1da177e4c
|
523 |
pages_skipped = wbc->pages_skipped; |
01c031945
|
524 |
writeback_single_inode(inode, wbc); |
1da177e4c
|
525 526 527 528 529 |
if (wbc->pages_skipped != pages_skipped) { /* * writeback is not making progress due to locked * buffers. Skip this inode for now. */ |
f57b9b7b4
|
530 |
redirty_tail(inode); |
1da177e4c
|
531 |
} |
0f1b1fd86
|
532 |
spin_unlock(&inode->i_lock); |
a66979aba
|
533 |
spin_unlock(&inode_wb_list_lock); |
1da177e4c
|
534 |
iput(inode); |
4ffc84442
|
535 |
cond_resched(); |
a66979aba
|
536 |
spin_lock(&inode_wb_list_lock); |
8bc3be275
|
537 538 |
if (wbc->nr_to_write <= 0) { wbc->more_io = 1; |
f11c9c5c2
|
539 |
return 1; |
8bc3be275
|
540 |
} |
03ba3782e
|
541 |
if (!list_empty(&wb->b_more_io)) |
8bc3be275
|
542 |
wbc->more_io = 1; |
1da177e4c
|
543 |
} |
f11c9c5c2
|
544 545 546 |
/* b_io is empty */ return 1; } |
9c3a8ee8a
|
547 548 |
void writeback_inodes_wb(struct bdi_writeback *wb, struct writeback_control *wbc) |
f11c9c5c2
|
549 550 |
{ int ret = 0; |
7624ee72a
|
551 552 |
if (!wbc->wb_start) wbc->wb_start = jiffies; /* livelock avoidance */ |
a66979aba
|
553 |
spin_lock(&inode_wb_list_lock); |
f11c9c5c2
|
554 555 |
if (!wbc->for_kupdate || list_empty(&wb->b_io)) queue_io(wb, wbc->older_than_this); |
38f219776
|
556 |
|
f11c9c5c2
|
557 |
while (!list_empty(&wb->b_io)) { |
7ccf19a80
|
558 |
struct inode *inode = wb_inode(wb->b_io.prev); |
f11c9c5c2
|
559 |
struct super_block *sb = inode->i_sb; |
9ecc2738a
|
560 |
|
edadfb10b
|
561 562 563 |
if (!pin_sb_for_writeback(sb)) { requeue_io(inode); continue; |
f11c9c5c2
|
564 |
} |
edadfb10b
|
565 566 |
ret = writeback_sb_inodes(sb, wb, wbc, false); drop_super(sb); |
f11c9c5c2
|
567 |
|
f11c9c5c2
|
568 569 570 |
if (ret) break; } |
a66979aba
|
571 |
spin_unlock(&inode_wb_list_lock); |
66f3b8e2e
|
572 573 |
/* Leave any unwritten inodes on b_io */ } |
edadfb10b
|
574 575 576 577 |
static void __writeback_inodes_sb(struct super_block *sb, struct bdi_writeback *wb, struct writeback_control *wbc) { WARN_ON(!rwsem_is_locked(&sb->s_umount)); |
a66979aba
|
578 |
spin_lock(&inode_wb_list_lock); |
edadfb10b
|
579 580 581 |
if (!wbc->for_kupdate || list_empty(&wb->b_io)) queue_io(wb, wbc->older_than_this); writeback_sb_inodes(sb, wb, wbc, true); |
a66979aba
|
582 |
spin_unlock(&inode_wb_list_lock); |
edadfb10b
|
583 |
} |
66f3b8e2e
|
584 |
/* |
03ba3782e
|
585 586 587 588 589 590 591 592 593 594 595 |
* The maximum number of pages to writeout in a single bdi flush/kupdate * operation. We do this so we don't hold I_SYNC against an inode for * enormous amounts of time, which would block a userspace task which has * been forced to throttle against that inode. Also, the code reevaluates * the dirty each time it has written this many pages. */ #define MAX_WRITEBACK_PAGES 1024 static inline bool over_bground_thresh(void) { unsigned long background_thresh, dirty_thresh; |
16c4042f0
|
596 |
global_dirty_limits(&background_thresh, &dirty_thresh); |
03ba3782e
|
597 598 |
return (global_page_state(NR_FILE_DIRTY) + |
4cbec4c8b
|
599 |
global_page_state(NR_UNSTABLE_NFS) > background_thresh); |
03ba3782e
|
600 601 602 603 |
} /* * Explicit flushing or periodic writeback of "old" data. |
66f3b8e2e
|
604 |
* |
03ba3782e
|
605 606 607 608 |
* Define "old": the first time one of an inode's pages is dirtied, we mark the * dirtying-time in the inode's address_space. So this periodic writeback code * just walks the superblock inode list, writing back any inodes which are * older than a specific point in time. |
66f3b8e2e
|
609 |
* |
03ba3782e
|
610 611 612 |
* Try to run once per dirty_writeback_interval. But if a writeback event * takes longer than a dirty_writeback_interval interval, then leave a * one-second gap. |
66f3b8e2e
|
613 |
* |
03ba3782e
|
614 615 |
* older_than_this takes precedence over nr_to_write. So we'll only write back * all dirty pages if they are all attached to "old" mappings. |
66f3b8e2e
|
616 |
*/ |
c4a77a6c7
|
617 |
static long wb_writeback(struct bdi_writeback *wb, |
83ba7b071
|
618 |
struct wb_writeback_work *work) |
66f3b8e2e
|
619 |
{ |
03ba3782e
|
620 |
struct writeback_control wbc = { |
83ba7b071
|
621 |
.sync_mode = work->sync_mode, |
03ba3782e
|
622 |
.older_than_this = NULL, |
83ba7b071
|
623 624 625 |
.for_kupdate = work->for_kupdate, .for_background = work->for_background, .range_cyclic = work->range_cyclic, |
03ba3782e
|
626 627 628 |
}; unsigned long oldest_jif; long wrote = 0; |
b9543dac5
|
629 |
long write_chunk; |
a5989bdc9
|
630 |
struct inode *inode; |
66f3b8e2e
|
631 |
|
03ba3782e
|
632 633 634 635 636 |
if (wbc.for_kupdate) { wbc.older_than_this = &oldest_jif; oldest_jif = jiffies - msecs_to_jiffies(dirty_expire_interval * 10); } |
c4a77a6c7
|
637 638 639 640 |
if (!wbc.range_cyclic) { wbc.range_start = 0; wbc.range_end = LLONG_MAX; } |
38f219776
|
641 |
|
b9543dac5
|
642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 |
/* * WB_SYNC_ALL mode does livelock avoidance by syncing dirty * inodes/pages in one big loop. Setting wbc.nr_to_write=LONG_MAX * here avoids calling into writeback_inodes_wb() more than once. * * The intended call sequence for WB_SYNC_ALL writeback is: * * wb_writeback() * __writeback_inodes_sb() <== called only once * write_cache_pages() <== called once for each inode * (quickly) tag currently dirty pages * (maybe slowly) sync all tagged pages */ if (wbc.sync_mode == WB_SYNC_NONE) write_chunk = MAX_WRITEBACK_PAGES; else write_chunk = LONG_MAX; |
7624ee72a
|
659 |
wbc.wb_start = jiffies; /* livelock avoidance */ |
03ba3782e
|
660 661 |
for (;;) { /* |
d3ddec763
|
662 |
* Stop writeback when nr_pages has been consumed |
03ba3782e
|
663 |
*/ |
83ba7b071
|
664 |
if (work->nr_pages <= 0) |
03ba3782e
|
665 |
break; |
66f3b8e2e
|
666 |
|
38f219776
|
667 |
/* |
aa373cf55
|
668 669 670 671 672 673 674 675 676 677 |
* Background writeout and kupdate-style writeback may * run forever. Stop them if there is other work to do * so that e.g. sync can proceed. They'll be restarted * after the other works are all done. */ if ((work->for_background || work->for_kupdate) && !list_empty(&wb->bdi->work_list)) break; /* |
d3ddec763
|
678 679 |
* For background writeout, stop when we are below the * background dirty threshold |
38f219776
|
680 |
*/ |
83ba7b071
|
681 |
if (work->for_background && !over_bground_thresh()) |
03ba3782e
|
682 |
break; |
38f219776
|
683 |
|
03ba3782e
|
684 |
wbc.more_io = 0; |
b9543dac5
|
685 |
wbc.nr_to_write = write_chunk; |
03ba3782e
|
686 |
wbc.pages_skipped = 0; |
028c2dd18
|
687 688 |
trace_wbc_writeback_start(&wbc, wb->bdi); |
83ba7b071
|
689 690 |
if (work->sb) __writeback_inodes_sb(work->sb, wb, &wbc); |
edadfb10b
|
691 692 |
else writeback_inodes_wb(wb, &wbc); |
028c2dd18
|
693 |
trace_wbc_writeback_written(&wbc, wb->bdi); |
b9543dac5
|
694 695 |
work->nr_pages -= write_chunk - wbc.nr_to_write; wrote += write_chunk - wbc.nr_to_write; |
03ba3782e
|
696 697 |
/* |
71fd05a88
|
698 |
* If we consumed everything, see if we have more |
03ba3782e
|
699 |
*/ |
71fd05a88
|
700 701 702 703 704 705 |
if (wbc.nr_to_write <= 0) continue; /* * Didn't write everything and we don't have more IO, bail */ if (!wbc.more_io) |
03ba3782e
|
706 |
break; |
71fd05a88
|
707 708 709 |
/* * Did we write something? Try for more */ |
b9543dac5
|
710 |
if (wbc.nr_to_write < write_chunk) |
71fd05a88
|
711 712 713 714 715 716 |
continue; /* * Nothing written. Wait for some inode to * become available for writeback. Otherwise * we'll just busyloop. */ |
a66979aba
|
717 |
spin_lock(&inode_wb_list_lock); |
71fd05a88
|
718 |
if (!list_empty(&wb->b_more_io)) { |
7ccf19a80
|
719 |
inode = wb_inode(wb->b_more_io.prev); |
028c2dd18
|
720 |
trace_wbc_writeback_wait(&wbc, wb->bdi); |
250df6ed2
|
721 |
spin_lock(&inode->i_lock); |
71fd05a88
|
722 |
inode_wait_for_writeback(inode); |
250df6ed2
|
723 |
spin_unlock(&inode->i_lock); |
03ba3782e
|
724 |
} |
a66979aba
|
725 |
spin_unlock(&inode_wb_list_lock); |
03ba3782e
|
726 727 728 729 730 731 |
} return wrote; } /* |
83ba7b071
|
732 |
* Return the next wb_writeback_work struct that hasn't been processed yet. |
03ba3782e
|
733 |
*/ |
83ba7b071
|
734 |
static struct wb_writeback_work * |
08852b6d6
|
735 |
get_next_work_item(struct backing_dev_info *bdi) |
03ba3782e
|
736 |
{ |
83ba7b071
|
737 |
struct wb_writeback_work *work = NULL; |
03ba3782e
|
738 |
|
6467716a3
|
739 |
spin_lock_bh(&bdi->wb_lock); |
83ba7b071
|
740 741 742 743 |
if (!list_empty(&bdi->work_list)) { work = list_entry(bdi->work_list.next, struct wb_writeback_work, list); list_del_init(&work->list); |
03ba3782e
|
744 |
} |
6467716a3
|
745 |
spin_unlock_bh(&bdi->wb_lock); |
83ba7b071
|
746 |
return work; |
03ba3782e
|
747 |
} |
cdf01dd54
|
748 749 750 751 752 753 754 755 756 757 |
/* * Add in the number of potentially dirty inodes, because each inode * write can dirty pagecache in the underlying blockdev. */ static unsigned long get_nr_dirty_pages(void) { return global_page_state(NR_FILE_DIRTY) + global_page_state(NR_UNSTABLE_NFS) + get_nr_dirty_inodes(); } |
6585027a5
|
758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 |
static long wb_check_background_flush(struct bdi_writeback *wb) { if (over_bground_thresh()) { struct wb_writeback_work work = { .nr_pages = LONG_MAX, .sync_mode = WB_SYNC_NONE, .for_background = 1, .range_cyclic = 1, }; return wb_writeback(wb, &work); } return 0; } |
03ba3782e
|
774 775 776 777 |
static long wb_check_old_data_flush(struct bdi_writeback *wb) { unsigned long expired; long nr_pages; |
69b62d01e
|
778 779 780 781 782 |
/* * When set to zero, disable periodic writeback */ if (!dirty_writeback_interval) return 0; |
03ba3782e
|
783 784 785 786 787 788 |
expired = wb->last_old_flush + msecs_to_jiffies(dirty_writeback_interval * 10); if (time_before(jiffies, expired)) return 0; wb->last_old_flush = jiffies; |
cdf01dd54
|
789 |
nr_pages = get_nr_dirty_pages(); |
03ba3782e
|
790 |
|
c4a77a6c7
|
791 |
if (nr_pages) { |
83ba7b071
|
792 |
struct wb_writeback_work work = { |
c4a77a6c7
|
793 794 795 796 797 |
.nr_pages = nr_pages, .sync_mode = WB_SYNC_NONE, .for_kupdate = 1, .range_cyclic = 1, }; |
83ba7b071
|
798 |
return wb_writeback(wb, &work); |
c4a77a6c7
|
799 |
} |
03ba3782e
|
800 801 802 803 804 805 806 807 808 809 |
return 0; } /* * Retrieve work items and do the writeback they describe */ long wb_do_writeback(struct bdi_writeback *wb, int force_wait) { struct backing_dev_info *bdi = wb->bdi; |
83ba7b071
|
810 |
struct wb_writeback_work *work; |
c4a77a6c7
|
811 |
long wrote = 0; |
03ba3782e
|
812 |
|
81d73a32d
|
813 |
set_bit(BDI_writeback_running, &wb->bdi->state); |
08852b6d6
|
814 |
while ((work = get_next_work_item(bdi)) != NULL) { |
03ba3782e
|
815 816 |
/* * Override sync mode, in case we must wait for completion |
83ba7b071
|
817 |
* because this thread is exiting now. |
03ba3782e
|
818 819 |
*/ if (force_wait) |
83ba7b071
|
820 |
work->sync_mode = WB_SYNC_ALL; |
03ba3782e
|
821 |
|
455b28646
|
822 |
trace_writeback_exec(bdi, work); |
83ba7b071
|
823 |
wrote += wb_writeback(wb, work); |
03ba3782e
|
824 825 |
/* |
83ba7b071
|
826 827 |
* Notify the caller of completion if this is a synchronous * work item, otherwise just free it. |
03ba3782e
|
828 |
*/ |
83ba7b071
|
829 830 831 832 |
if (work->done) complete(work->done); else kfree(work); |
03ba3782e
|
833 834 835 836 837 838 |
} /* * Check for periodic writeback, kupdated() style */ wrote += wb_check_old_data_flush(wb); |
6585027a5
|
839 |
wrote += wb_check_background_flush(wb); |
81d73a32d
|
840 |
clear_bit(BDI_writeback_running, &wb->bdi->state); |
03ba3782e
|
841 842 843 844 845 846 847 848 |
return wrote; } /* * Handle writeback of dirty data for the device backed by this bdi. Also * wakes up periodically and does kupdated style flushing. */ |
082439004
|
849 |
int bdi_writeback_thread(void *data) |
03ba3782e
|
850 |
{ |
082439004
|
851 852 |
struct bdi_writeback *wb = data; struct backing_dev_info *bdi = wb->bdi; |
03ba3782e
|
853 |
long pages_written; |
766f91641
|
854 |
current->flags |= PF_SWAPWRITE; |
082439004
|
855 |
set_freezable(); |
ecd584030
|
856 |
wb->last_active = jiffies; |
082439004
|
857 858 859 860 861 |
/* * Our parent may run at a different priority, just set us to normal */ set_user_nice(current, 0); |
455b28646
|
862 |
trace_writeback_thread_start(bdi); |
03ba3782e
|
863 |
while (!kthread_should_stop()) { |
6467716a3
|
864 865 866 867 868 |
/* * Remove own delayed wake-up timer, since we are already awake * and we'll take care of the preriodic write-back. */ del_timer(&wb->wakeup_timer); |
03ba3782e
|
869 |
pages_written = wb_do_writeback(wb, 0); |
455b28646
|
870 |
trace_writeback_pages_written(pages_written); |
03ba3782e
|
871 |
if (pages_written) |
ecd584030
|
872 |
wb->last_active = jiffies; |
03ba3782e
|
873 |
|
297252c81
|
874 |
set_current_state(TASK_INTERRUPTIBLE); |
b76b4014f
|
875 |
if (!list_empty(&bdi->work_list) || kthread_should_stop()) { |
f9eadbbd4
|
876 |
__set_current_state(TASK_RUNNING); |
297252c81
|
877 |
continue; |
03ba3782e
|
878 |
} |
253c34e9b
|
879 |
if (wb_has_dirty_io(wb) && dirty_writeback_interval) |
fff5b85aa
|
880 |
schedule_timeout(msecs_to_jiffies(dirty_writeback_interval * 10)); |
253c34e9b
|
881 882 883 884 885 886 |
else { /* * We have nothing to do, so can go sleep without any * timeout and save power. When a work is queued or * something is made dirty - we will be woken up. */ |
297252c81
|
887 |
schedule(); |
f9eadbbd4
|
888 |
} |
69b62d01e
|
889 |
|
03ba3782e
|
890 891 |
try_to_freeze(); } |
fff5b85aa
|
892 |
/* Flush any work that raced with us exiting */ |
082439004
|
893 894 |
if (!list_empty(&bdi->work_list)) wb_do_writeback(wb, 1); |
455b28646
|
895 896 |
trace_writeback_thread_stop(bdi); |
03ba3782e
|
897 898 |
return 0; } |
082439004
|
899 |
|
03ba3782e
|
900 |
/* |
b8c2f3474
|
901 902 |
* Start writeback of `nr_pages' pages. If `nr_pages' is zero, write back * the whole world. |
03ba3782e
|
903 |
*/ |
b8c2f3474
|
904 |
void wakeup_flusher_threads(long nr_pages) |
03ba3782e
|
905 |
{ |
b8c2f3474
|
906 |
struct backing_dev_info *bdi; |
03ba3782e
|
907 |
|
83ba7b071
|
908 909 |
if (!nr_pages) { nr_pages = global_page_state(NR_FILE_DIRTY) + |
b8c2f3474
|
910 911 |
global_page_state(NR_UNSTABLE_NFS); } |
03ba3782e
|
912 |
|
b8c2f3474
|
913 |
rcu_read_lock(); |
cfc4ba536
|
914 |
list_for_each_entry_rcu(bdi, &bdi_list, bdi_list) { |
03ba3782e
|
915 916 |
if (!bdi_has_dirty_io(bdi)) continue; |
6585027a5
|
917 |
__bdi_start_writeback(bdi, nr_pages, false); |
03ba3782e
|
918 |
} |
cfc4ba536
|
919 |
rcu_read_unlock(); |
1da177e4c
|
920 |
} |
03ba3782e
|
921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 |
static noinline void block_dump___mark_inode_dirty(struct inode *inode) { if (inode->i_ino || strcmp(inode->i_sb->s_id, "bdev")) { struct dentry *dentry; const char *name = "?"; dentry = d_find_alias(inode); if (dentry) { spin_lock(&dentry->d_lock); name = (const char *) dentry->d_name.name; } printk(KERN_DEBUG "%s(%d): dirtied inode %lu (%s) on %s ", current->comm, task_pid_nr(current), inode->i_ino, name, inode->i_sb->s_id); if (dentry) { spin_unlock(&dentry->d_lock); dput(dentry); } } } /** * __mark_inode_dirty - internal function * @inode: inode to mark * @flags: what kind of dirty (i.e. I_DIRTY_SYNC) * Mark an inode as dirty. Callers should use mark_inode_dirty or * mark_inode_dirty_sync. |
1da177e4c
|
950 |
* |
03ba3782e
|
951 952 953 954 955 956 957 958 959 |
* Put the inode on the super block's dirty list. * * CAREFUL! We mark it dirty unconditionally, but move it onto the * dirty list only if it is hashed or if it refers to a blockdev. * If it was not hashed, it will never be added to the dirty list * even if it is later hashed, as it will have been marked dirty already. * * In short, make sure you hash any inodes _before_ you start marking * them dirty. |
1da177e4c
|
960 |
* |
03ba3782e
|
961 962 |
* This function *must* be atomic for the I_DIRTY_PAGES case - * set_page_dirty() is called under spinlock in several places. |
1da177e4c
|
963 |
* |
03ba3782e
|
964 965 966 967 968 969 |
* Note that for blockdevs, inode->dirtied_when represents the dirtying time of * the block-special inode (/dev/hda1) itself. And the ->dirtied_when field of * the kernel-internal blockdev inode represents the dirtying time of the * blockdev's pages. This is why for I_DIRTY_PAGES we always use * page->mapping->host, so the page-dirtying time is recorded in the internal * blockdev inode. |
1da177e4c
|
970 |
*/ |
03ba3782e
|
971 |
void __mark_inode_dirty(struct inode *inode, int flags) |
1da177e4c
|
972 |
{ |
03ba3782e
|
973 |
struct super_block *sb = inode->i_sb; |
253c34e9b
|
974 |
struct backing_dev_info *bdi = NULL; |
1da177e4c
|
975 |
|
03ba3782e
|
976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 |
/* * Don't do this for I_DIRTY_PAGES - that doesn't actually * dirty the inode itself */ if (flags & (I_DIRTY_SYNC | I_DIRTY_DATASYNC)) { if (sb->s_op->dirty_inode) sb->s_op->dirty_inode(inode); } /* * make sure that changes are seen by all cpus before we test i_state * -- mikulas */ smp_mb(); /* avoid the locking if we can */ if ((inode->i_state & flags) == flags) return; if (unlikely(block_dump)) block_dump___mark_inode_dirty(inode); |
250df6ed2
|
997 |
spin_lock(&inode->i_lock); |
03ba3782e
|
998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 |
if ((inode->i_state & flags) != flags) { const int was_dirty = inode->i_state & I_DIRTY; inode->i_state |= flags; /* * If the inode is being synced, just update its dirty state. * The unlocker will place the inode on the appropriate * superblock list, based upon its state. */ if (inode->i_state & I_SYNC) |
250df6ed2
|
1009 |
goto out_unlock_inode; |
03ba3782e
|
1010 1011 1012 1013 1014 1015 |
/* * Only add valid (hashed) inodes to the superblock's * dirty list. Add blockdev inodes as well. */ if (!S_ISBLK(inode->i_mode)) { |
1d3382cbf
|
1016 |
if (inode_unhashed(inode)) |
250df6ed2
|
1017 |
goto out_unlock_inode; |
03ba3782e
|
1018 |
} |
a4ffdde6e
|
1019 |
if (inode->i_state & I_FREEING) |
250df6ed2
|
1020 |
goto out_unlock_inode; |
03ba3782e
|
1021 1022 1023 1024 1025 1026 |
/* * If the inode was already on b_dirty/b_io/b_more_io, don't * reposition it (that would break b_dirty time-ordering). */ if (!was_dirty) { |
a66979aba
|
1027 |
bool wakeup_bdi = false; |
253c34e9b
|
1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 |
bdi = inode_to_bdi(inode); if (bdi_cap_writeback_dirty(bdi)) { WARN(!test_bit(BDI_registered, &bdi->state), "bdi-%s not registered ", bdi->name); /* * If this is the first dirty inode for this * bdi, we have to wake-up the corresponding * bdi thread to make sure background * write-back happens later. */ if (!wb_has_dirty_io(&bdi->wb)) wakeup_bdi = true; |
500b067c5
|
1043 |
} |
03ba3782e
|
1044 |
|
a66979aba
|
1045 1046 |
spin_unlock(&inode->i_lock); spin_lock(&inode_wb_list_lock); |
03ba3782e
|
1047 |
inode->dirtied_when = jiffies; |
7ccf19a80
|
1048 |
list_move(&inode->i_wb_list, &bdi->wb.b_dirty); |
a66979aba
|
1049 1050 1051 1052 1053 |
spin_unlock(&inode_wb_list_lock); if (wakeup_bdi) bdi_wakeup_thread_delayed(bdi); return; |
1da177e4c
|
1054 |
} |
1da177e4c
|
1055 |
} |
250df6ed2
|
1056 1057 |
out_unlock_inode: spin_unlock(&inode->i_lock); |
253c34e9b
|
1058 |
|
03ba3782e
|
1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 |
} EXPORT_SYMBOL(__mark_inode_dirty); /* * Write out a superblock's list of dirty inodes. A wait will be performed * upon no inodes, all inodes or the final one, depending upon sync_mode. * * If older_than_this is non-NULL, then only write out inodes which * had their first dirtying at a time earlier than *older_than_this. * |
03ba3782e
|
1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 |
* If `bdi' is non-zero then we're being asked to writeback a specific queue. * This function assumes that the blockdev superblock's inodes are backed by * a variety of queues, so all inodes are searched. For other superblocks, * assume that all inodes are backed by the same queue. * * The inodes to be written are parked on bdi->b_io. They are moved back onto * bdi->b_dirty as they are selected for writing. This way, none can be missed * on the writer throttling path, and we get decent balancing between many * throttled threads: we don't want them all piling up on inode_sync_wait. */ |
b6e51316d
|
1079 |
static void wait_sb_inodes(struct super_block *sb) |
03ba3782e
|
1080 1081 1082 1083 1084 1085 1086 |
{ struct inode *inode, *old_inode = NULL; /* * We need to be protected against the filesystem going from * r/o to r/w or vice versa. */ |
b6e51316d
|
1087 |
WARN_ON(!rwsem_is_locked(&sb->s_umount)); |
03ba3782e
|
1088 |
|
55fa6091d
|
1089 |
spin_lock(&inode_sb_list_lock); |
03ba3782e
|
1090 1091 1092 1093 1094 1095 1096 1097 |
/* * Data integrity sync. Must wait for all pages under writeback, * because there may have been pages dirtied before our sync * call, but which had writeout started before we write it out. * In which case, the inode may not be on the dirty list, but * we still have to wait for that writeout. */ |
b6e51316d
|
1098 |
list_for_each_entry(inode, &sb->s_inodes, i_sb_list) { |
250df6ed2
|
1099 |
struct address_space *mapping = inode->i_mapping; |
03ba3782e
|
1100 |
|
250df6ed2
|
1101 1102 1103 1104 |
spin_lock(&inode->i_lock); if ((inode->i_state & (I_FREEING|I_WILL_FREE|I_NEW)) || (mapping->nrpages == 0)) { spin_unlock(&inode->i_lock); |
03ba3782e
|
1105 |
continue; |
250df6ed2
|
1106 |
} |
03ba3782e
|
1107 |
__iget(inode); |
250df6ed2
|
1108 |
spin_unlock(&inode->i_lock); |
55fa6091d
|
1109 |
spin_unlock(&inode_sb_list_lock); |
03ba3782e
|
1110 |
/* |
55fa6091d
|
1111 1112 1113 1114 1115 1116 |
* We hold a reference to 'inode' so it couldn't have been * removed from s_inodes list while we dropped the * inode_sb_list_lock. We cannot iput the inode now as we can * be holding the last reference and we cannot iput it under * inode_sb_list_lock. So we keep the reference and iput it * later. |
03ba3782e
|
1117 1118 1119 1120 1121 1122 1123 |
*/ iput(old_inode); old_inode = inode; filemap_fdatawait(mapping); cond_resched(); |
55fa6091d
|
1124 |
spin_lock(&inode_sb_list_lock); |
03ba3782e
|
1125 |
} |
55fa6091d
|
1126 |
spin_unlock(&inode_sb_list_lock); |
03ba3782e
|
1127 |
iput(old_inode); |
1da177e4c
|
1128 |
} |
d8a8559cd
|
1129 |
/** |
3259f8bed
|
1130 |
* writeback_inodes_sb_nr - writeback dirty inodes from given super_block |
d8a8559cd
|
1131 |
* @sb: the superblock |
3259f8bed
|
1132 |
* @nr: the number of pages to write |
1da177e4c
|
1133 |
* |
d8a8559cd
|
1134 1135 |
* Start writeback on some inodes on this super_block. No guarantees are made * on how many (if any) will be written, and this function does not wait |
3259f8bed
|
1136 |
* for IO completion of submitted IO. |
1da177e4c
|
1137 |
*/ |
3259f8bed
|
1138 |
void writeback_inodes_sb_nr(struct super_block *sb, unsigned long nr) |
1da177e4c
|
1139 |
{ |
83ba7b071
|
1140 1141 |
DECLARE_COMPLETION_ONSTACK(done); struct wb_writeback_work work = { |
3c4d71653
|
1142 1143 |
.sb = sb, .sync_mode = WB_SYNC_NONE, |
83ba7b071
|
1144 |
.done = &done, |
3259f8bed
|
1145 |
.nr_pages = nr, |
3c4d71653
|
1146 |
}; |
d8a8559cd
|
1147 |
|
cf37e9724
|
1148 |
WARN_ON(!rwsem_is_locked(&sb->s_umount)); |
83ba7b071
|
1149 1150 |
bdi_queue_work(sb->s_bdi, &work); wait_for_completion(&done); |
e913fc825
|
1151 |
} |
3259f8bed
|
1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 |
EXPORT_SYMBOL(writeback_inodes_sb_nr); /** * writeback_inodes_sb - writeback dirty inodes from given super_block * @sb: the superblock * * Start writeback on some inodes on this super_block. No guarantees are made * on how many (if any) will be written, and this function does not wait * for IO completion of submitted IO. */ void writeback_inodes_sb(struct super_block *sb) { |
925d169f5
|
1164 |
return writeback_inodes_sb_nr(sb, get_nr_dirty_pages()); |
3259f8bed
|
1165 |
} |
0e3c9a228
|
1166 |
EXPORT_SYMBOL(writeback_inodes_sb); |
e913fc825
|
1167 1168 |
/** |
17bd55d03
|
1169 1170 1171 1172 1173 1174 1175 1176 1177 |
* writeback_inodes_sb_if_idle - start writeback if none underway * @sb: the superblock * * Invoke writeback_inodes_sb if no writeback is currently underway. * Returns 1 if writeback was started, 0 if not. */ int writeback_inodes_sb_if_idle(struct super_block *sb) { if (!writeback_in_progress(sb->s_bdi)) { |
cf37e9724
|
1178 |
down_read(&sb->s_umount); |
17bd55d03
|
1179 |
writeback_inodes_sb(sb); |
cf37e9724
|
1180 |
up_read(&sb->s_umount); |
17bd55d03
|
1181 1182 1183 1184 1185 1186 1187 |
return 1; } else return 0; } EXPORT_SYMBOL(writeback_inodes_sb_if_idle); /** |
3259f8bed
|
1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 |
* writeback_inodes_sb_if_idle - start writeback if none underway * @sb: the superblock * @nr: the number of pages to write * * Invoke writeback_inodes_sb if no writeback is currently underway. * Returns 1 if writeback was started, 0 if not. */ int writeback_inodes_sb_nr_if_idle(struct super_block *sb, unsigned long nr) { if (!writeback_in_progress(sb->s_bdi)) { down_read(&sb->s_umount); writeback_inodes_sb_nr(sb, nr); up_read(&sb->s_umount); return 1; } else return 0; } EXPORT_SYMBOL(writeback_inodes_sb_nr_if_idle); /** |
d8a8559cd
|
1209 1210 1211 1212 |
* sync_inodes_sb - sync sb inode pages * @sb: the superblock * * This function writes and waits on any dirty inode belonging to this |
cb9ef8d5e
|
1213 |
* super_block. |
d8a8559cd
|
1214 |
*/ |
b6e51316d
|
1215 |
void sync_inodes_sb(struct super_block *sb) |
d8a8559cd
|
1216 |
{ |
83ba7b071
|
1217 1218 |
DECLARE_COMPLETION_ONSTACK(done); struct wb_writeback_work work = { |
3c4d71653
|
1219 1220 1221 1222 |
.sb = sb, .sync_mode = WB_SYNC_ALL, .nr_pages = LONG_MAX, .range_cyclic = 0, |
83ba7b071
|
1223 |
.done = &done, |
3c4d71653
|
1224 |
}; |
cf37e9724
|
1225 |
WARN_ON(!rwsem_is_locked(&sb->s_umount)); |
83ba7b071
|
1226 1227 |
bdi_queue_work(sb->s_bdi, &work); wait_for_completion(&done); |
b6e51316d
|
1228 |
wait_sb_inodes(sb); |
1da177e4c
|
1229 |
} |
d8a8559cd
|
1230 |
EXPORT_SYMBOL(sync_inodes_sb); |
1da177e4c
|
1231 |
|
1da177e4c
|
1232 |
/** |
7f04c26d7
|
1233 1234 1235 1236 1237 1238 |
* write_inode_now - write an inode to disk * @inode: inode to write to disk * @sync: whether the write should be synchronous or not * * This function commits an inode to disk immediately if it is dirty. This is * primarily needed by knfsd. |
1da177e4c
|
1239 |
* |
7f04c26d7
|
1240 |
* The caller must either have a ref on the inode or must have set I_WILL_FREE. |
1da177e4c
|
1241 |
*/ |
1da177e4c
|
1242 1243 1244 1245 1246 |
int write_inode_now(struct inode *inode, int sync) { int ret; struct writeback_control wbc = { .nr_to_write = LONG_MAX, |
18914b188
|
1247 |
.sync_mode = sync ? WB_SYNC_ALL : WB_SYNC_NONE, |
111ebb6e6
|
1248 1249 |
.range_start = 0, .range_end = LLONG_MAX, |
1da177e4c
|
1250 1251 1252 |
}; if (!mapping_cap_writeback_dirty(inode->i_mapping)) |
49364ce25
|
1253 |
wbc.nr_to_write = 0; |
1da177e4c
|
1254 1255 |
might_sleep(); |
a66979aba
|
1256 |
spin_lock(&inode_wb_list_lock); |
0f1b1fd86
|
1257 |
spin_lock(&inode->i_lock); |
01c031945
|
1258 |
ret = writeback_single_inode(inode, &wbc); |
0f1b1fd86
|
1259 |
spin_unlock(&inode->i_lock); |
a66979aba
|
1260 |
spin_unlock(&inode_wb_list_lock); |
1da177e4c
|
1261 |
if (sync) |
1c0eeaf56
|
1262 |
inode_sync_wait(inode); |
1da177e4c
|
1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 |
return ret; } EXPORT_SYMBOL(write_inode_now); /** * sync_inode - write an inode and its pages to disk. * @inode: the inode to sync * @wbc: controls the writeback mode * * sync_inode() will write an inode and its pages to disk. It will also * correctly update the inode on its superblock's dirty inode lists and will * update inode->i_state. * * The caller must have a ref on the inode. */ int sync_inode(struct inode *inode, struct writeback_control *wbc) { int ret; |
a66979aba
|
1281 |
spin_lock(&inode_wb_list_lock); |
0f1b1fd86
|
1282 |
spin_lock(&inode->i_lock); |
01c031945
|
1283 |
ret = writeback_single_inode(inode, wbc); |
0f1b1fd86
|
1284 |
spin_unlock(&inode->i_lock); |
a66979aba
|
1285 |
spin_unlock(&inode_wb_list_lock); |
1da177e4c
|
1286 1287 1288 |
return ret; } EXPORT_SYMBOL(sync_inode); |
c37650161
|
1289 1290 |
/** |
c691b9d98
|
1291 |
* sync_inode_metadata - write an inode to disk |
c37650161
|
1292 1293 1294 |
* @inode: the inode to sync * @wait: wait for I/O to complete. * |
c691b9d98
|
1295 |
* Write an inode to disk and adjust its dirty state after completion. |
c37650161
|
1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 |
* * Note: only writes the actual inode, no associated data or other metadata. */ int sync_inode_metadata(struct inode *inode, int wait) { struct writeback_control wbc = { .sync_mode = wait ? WB_SYNC_ALL : WB_SYNC_NONE, .nr_to_write = 0, /* metadata-only */ }; return sync_inode(inode, &wbc); } EXPORT_SYMBOL(sync_inode_metadata); |