Blame view
fs/fs-writeback.c
70.6 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> |
630d9c472
|
17 |
#include <linux/export.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> |
bc31b86a5
|
23 |
#include <linux/pagemap.h> |
03ba3782e
|
24 |
#include <linux/kthread.h> |
1da177e4c
|
25 26 27 |
#include <linux/writeback.h> #include <linux/blkdev.h> #include <linux/backing-dev.h> |
455b28646
|
28 |
#include <linux/tracepoint.h> |
719ea2fbb
|
29 |
#include <linux/device.h> |
21c6321fb
|
30 |
#include <linux/memcontrol.h> |
07f3f05c1
|
31 |
#include "internal.h" |
1da177e4c
|
32 |
|
d0bceac74
|
33 |
/* |
bc31b86a5
|
34 35 |
* 4MB minimal write chunk size */ |
09cbfeaf1
|
36 |
#define MIN_WRITEBACK_PAGES (4096UL >> (PAGE_SHIFT - 10)) |
bc31b86a5
|
37 |
|
cc395d7f1
|
38 39 40 |
struct wb_completion { atomic_t cnt; }; |
bc31b86a5
|
41 |
/* |
c4a77a6c7
|
42 43 |
* Passed into wb_writeback(), essentially a subset of writeback_control */ |
83ba7b071
|
44 |
struct wb_writeback_work { |
c4a77a6c7
|
45 46 |
long nr_pages; struct super_block *sb; |
0dc83bd30
|
47 |
unsigned long *older_than_this; |
c4a77a6c7
|
48 |
enum writeback_sync_modes sync_mode; |
6e6938b6d
|
49 |
unsigned int tagged_writepages:1; |
52957fe1c
|
50 51 52 |
unsigned int for_kupdate:1; unsigned int range_cyclic:1; unsigned int for_background:1; |
7747bd4bc
|
53 |
unsigned int for_sync:1; /* sync(2) WB_SYNC_ALL writeback */ |
ac7b19a34
|
54 |
unsigned int auto_free:1; /* free on completion */ |
0e175a183
|
55 |
enum wb_reason reason; /* why was writeback initiated? */ |
c4a77a6c7
|
56 |
|
8010c3b63
|
57 |
struct list_head list; /* pending work list */ |
cc395d7f1
|
58 |
struct wb_completion *done; /* set if the caller waits */ |
03ba3782e
|
59 |
}; |
a2f487069
|
60 |
/* |
cc395d7f1
|
61 62 63 64 65 66 67 68 69 70 71 72 73 |
* If one wants to wait for one or more wb_writeback_works, each work's * ->done should be set to a wb_completion defined using the following * macro. Once all work items are issued with wb_queue_work(), the caller * can wait for the completion of all using wb_wait_for_completion(). Work * items which are waited upon aren't freed automatically on completion. */ #define DEFINE_WB_COMPLETION_ONSTACK(cmpl) \ struct wb_completion cmpl = { \ .cnt = ATOMIC_INIT(1), \ } /* |
a2f487069
|
74 75 76 77 78 79 80 81 82 83 |
* If an inode is constantly having its pages dirtied, but then the * updates stop dirtytime_expire_interval seconds in the past, it's * possible for the worst case time between when an inode has its * timestamps updated and when they finally get written out to be two * dirtytime_expire_intervals. We set the default to 12 hours (in * seconds), which means most of the time inodes will have their * timestamps written to disk after 12 hours, but in the worst case a * few inodes might not their timestamps updated for 24 hours. */ unsigned int dirtytime_expire_interval = 12 * 60 * 60; |
7ccf19a80
|
84 85 |
static inline struct inode *wb_inode(struct list_head *head) { |
c7f540849
|
86 |
return list_entry(head, struct inode, i_io_list); |
7ccf19a80
|
87 |
} |
15eb77a07
|
88 89 90 91 92 93 94 |
/* * Include the creation of the trace points after defining the * wb_writeback_work structure and inline functions so that the definition * remains local to this file. */ #define CREATE_TRACE_POINTS #include <trace/events/writeback.h> |
774016b2d
|
95 |
EXPORT_TRACEPOINT_SYMBOL_GPL(wbc_writepage); |
d6c10f1fc
|
96 97 98 99 100 101 |
static bool wb_io_lists_populated(struct bdi_writeback *wb) { if (wb_has_dirty_io(wb)) { return false; } else { set_bit(WB_has_dirty_io, &wb->state); |
95a46c65e
|
102 |
WARN_ON_ONCE(!wb->avg_write_bandwidth); |
766a9d6e6
|
103 104 |
atomic_long_add(wb->avg_write_bandwidth, &wb->bdi->tot_write_bandwidth); |
d6c10f1fc
|
105 106 107 108 109 110 111 |
return true; } } static void wb_io_lists_depopulated(struct bdi_writeback *wb) { if (wb_has_dirty_io(wb) && list_empty(&wb->b_dirty) && |
766a9d6e6
|
112 |
list_empty(&wb->b_io) && list_empty(&wb->b_more_io)) { |
d6c10f1fc
|
113 |
clear_bit(WB_has_dirty_io, &wb->state); |
95a46c65e
|
114 115 |
WARN_ON_ONCE(atomic_long_sub_return(wb->avg_write_bandwidth, &wb->bdi->tot_write_bandwidth) < 0); |
766a9d6e6
|
116 |
} |
d6c10f1fc
|
117 118 119 |
} /** |
c7f540849
|
120 |
* inode_io_list_move_locked - move an inode onto a bdi_writeback IO list |
d6c10f1fc
|
121 122 123 124 |
* @inode: inode to be moved * @wb: target bdi_writeback * @head: one of @wb->b_{dirty|io|more_io} * |
c7f540849
|
125 |
* Move @inode->i_io_list to @list of @wb and set %WB_has_dirty_io. |
d6c10f1fc
|
126 127 128 |
* Returns %true if @inode is the first occupant of the !dirty_time IO * lists; otherwise, %false. */ |
c7f540849
|
129 |
static bool inode_io_list_move_locked(struct inode *inode, |
d6c10f1fc
|
130 131 132 133 |
struct bdi_writeback *wb, struct list_head *head) { assert_spin_locked(&wb->list_lock); |
c7f540849
|
134 |
list_move(&inode->i_io_list, head); |
d6c10f1fc
|
135 136 137 138 139 140 141 142 143 144 |
/* dirty_time doesn't count as dirty_io until expiration */ if (head != &wb->b_dirty_time) return wb_io_lists_populated(wb); wb_io_lists_depopulated(wb); return false; } /** |
c7f540849
|
145 |
* inode_io_list_del_locked - remove an inode from its bdi_writeback IO list |
d6c10f1fc
|
146 147 148 149 150 151 |
* @inode: inode to be removed * @wb: bdi_writeback @inode is being removed from * * Remove @inode which may be on one of @wb->b_{dirty|io|more_io} lists and * clear %WB_has_dirty_io if all are empty afterwards. */ |
c7f540849
|
152 |
static void inode_io_list_del_locked(struct inode *inode, |
d6c10f1fc
|
153 154 155 |
struct bdi_writeback *wb) { assert_spin_locked(&wb->list_lock); |
c7f540849
|
156 |
list_del_init(&inode->i_io_list); |
d6c10f1fc
|
157 158 |
wb_io_lists_depopulated(wb); } |
f0054bb1e
|
159 |
static void wb_wakeup(struct bdi_writeback *wb) |
5acda9d12
|
160 |
{ |
f0054bb1e
|
161 162 163 164 |
spin_lock_bh(&wb->work_lock); if (test_bit(WB_registered, &wb->state)) mod_delayed_work(bdi_wq, &wb->dwork, 0); spin_unlock_bh(&wb->work_lock); |
5acda9d12
|
165 |
} |
4a3a485b1
|
166 167 168 169 170 171 172 173 174 175 |
static void finish_writeback_work(struct bdi_writeback *wb, struct wb_writeback_work *work) { struct wb_completion *done = work->done; if (work->auto_free) kfree(work); if (done && atomic_dec_and_test(&done->cnt)) wake_up_all(&wb->bdi->wb_waitq); } |
f0054bb1e
|
176 177 |
static void wb_queue_work(struct bdi_writeback *wb, struct wb_writeback_work *work) |
6585027a5
|
178 |
{ |
5634cc2aa
|
179 |
trace_writeback_queue(wb, work); |
6585027a5
|
180 |
|
cc395d7f1
|
181 182 |
if (work->done) atomic_inc(&work->done->cnt); |
4a3a485b1
|
183 184 185 186 187 188 189 190 |
spin_lock_bh(&wb->work_lock); if (test_bit(WB_registered, &wb->state)) { list_add_tail(&work->list, &wb->work_list); mod_delayed_work(bdi_wq, &wb->dwork, 0); } else finish_writeback_work(wb, work); |
f0054bb1e
|
191 |
spin_unlock_bh(&wb->work_lock); |
1da177e4c
|
192 |
} |
cc395d7f1
|
193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 |
/** * wb_wait_for_completion - wait for completion of bdi_writeback_works * @bdi: bdi work items were issued to * @done: target wb_completion * * Wait for one or more work items issued to @bdi with their ->done field * set to @done, which should have been defined with * DEFINE_WB_COMPLETION_ONSTACK(). This function returns after all such * work items are completed. Work items which are waited upon aren't freed * automatically on completion. */ static void wb_wait_for_completion(struct backing_dev_info *bdi, struct wb_completion *done) { atomic_dec(&done->cnt); /* put down the initial count */ wait_event(bdi->wb_waitq, !atomic_read(&done->cnt)); } |
703c27088
|
210 |
#ifdef CONFIG_CGROUP_WRITEBACK |
2a8149081
|
211 212 213 214 215 216 217 218 219 220 221 222 223 |
/* parameters for foreign inode detection, see wb_detach_inode() */ #define WB_FRN_TIME_SHIFT 13 /* 1s = 2^13, upto 8 secs w/ 16bit */ #define WB_FRN_TIME_AVG_SHIFT 3 /* avg = avg * 7/8 + new * 1/8 */ #define WB_FRN_TIME_CUT_DIV 2 /* ignore rounds < avg / 2 */ #define WB_FRN_TIME_PERIOD (2 * (1 << WB_FRN_TIME_SHIFT)) /* 2s */ #define WB_FRN_HIST_SLOTS 16 /* inode->i_wb_frn_history is 16bit */ #define WB_FRN_HIST_UNIT (WB_FRN_TIME_PERIOD / WB_FRN_HIST_SLOTS) /* each slot's duration is 2s / 16 */ #define WB_FRN_HIST_THR_SLOTS (WB_FRN_HIST_SLOTS / 2) /* if foreign slots >= 8, switch */ #define WB_FRN_HIST_MAX_SLOTS (WB_FRN_HIST_THR_SLOTS / 2 + 1) /* one round can affect upto 5 slots */ |
a1a0e23e4
|
224 225 |
static atomic_t isw_nr_in_flight = ATOMIC_INIT(0); static struct workqueue_struct *isw_wq; |
21c6321fb
|
226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 |
void __inode_attach_wb(struct inode *inode, struct page *page) { struct backing_dev_info *bdi = inode_to_bdi(inode); struct bdi_writeback *wb = NULL; if (inode_cgwb_enabled(inode)) { struct cgroup_subsys_state *memcg_css; if (page) { memcg_css = mem_cgroup_css_from_page(page); wb = wb_get_create(bdi, memcg_css, GFP_ATOMIC); } else { /* must pin memcg_css, see wb_get_create() */ memcg_css = task_get_css(current, memory_cgrp_id); wb = wb_get_create(bdi, memcg_css, GFP_ATOMIC); css_put(memcg_css); } } if (!wb) wb = &bdi->wb; /* * There may be multiple instances of this function racing to * update the same inode. Use cmpxchg() to tell the winner. */ if (unlikely(cmpxchg(&inode->i_wb, NULL, wb))) wb_put(wb); } |
703c27088
|
255 |
/** |
87e1d789b
|
256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 |
* locked_inode_to_wb_and_lock_list - determine a locked inode's wb and lock it * @inode: inode of interest with i_lock held * * Returns @inode's wb with its list_lock held. @inode->i_lock must be * held on entry and is released on return. The returned wb is guaranteed * to stay @inode's associated wb until its list_lock is released. */ static struct bdi_writeback * locked_inode_to_wb_and_lock_list(struct inode *inode) __releases(&inode->i_lock) __acquires(&wb->list_lock) { while (true) { struct bdi_writeback *wb = inode_to_wb(inode); /* * inode_to_wb() association is protected by both * @inode->i_lock and @wb->list_lock but list_lock nests * outside i_lock. Drop i_lock and verify that the * association hasn't changed after acquiring list_lock. */ wb_get(wb); spin_unlock(&inode->i_lock); spin_lock(&wb->list_lock); |
87e1d789b
|
280 |
|
aaa2cacf8
|
281 |
/* i_wb may have changed inbetween, can't use inode_to_wb() */ |
614a4e377
|
282 283 284 285 |
if (likely(wb == inode->i_wb)) { wb_put(wb); /* @inode already has ref */ return wb; } |
87e1d789b
|
286 287 |
spin_unlock(&wb->list_lock); |
614a4e377
|
288 |
wb_put(wb); |
87e1d789b
|
289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 |
cpu_relax(); spin_lock(&inode->i_lock); } } /** * inode_to_wb_and_lock_list - determine an inode's wb and lock it * @inode: inode of interest * * Same as locked_inode_to_wb_and_lock_list() but @inode->i_lock isn't held * on entry. */ static struct bdi_writeback *inode_to_wb_and_lock_list(struct inode *inode) __acquires(&wb->list_lock) { spin_lock(&inode->i_lock); return locked_inode_to_wb_and_lock_list(inode); } |
682aa8e1a
|
307 308 309 310 311 312 313 314 315 316 317 318 319 |
struct inode_switch_wbs_context { struct inode *inode; struct bdi_writeback *new_wb; struct rcu_head rcu_head; struct work_struct work; }; static void inode_switch_wbs_work_fn(struct work_struct *work) { struct inode_switch_wbs_context *isw = container_of(work, struct inode_switch_wbs_context, work); struct inode *inode = isw->inode; |
d10c80955
|
320 321 |
struct address_space *mapping = inode->i_mapping; struct bdi_writeback *old_wb = inode->i_wb; |
682aa8e1a
|
322 |
struct bdi_writeback *new_wb = isw->new_wb; |
d10c80955
|
323 324 325 |
struct radix_tree_iter iter; bool switched = false; void **slot; |
682aa8e1a
|
326 327 328 329 330 331 |
/* * By the time control reaches here, RCU grace period has passed * since I_WB_SWITCH assertion and all wb stat update transactions * between unlocked_inode_to_wb_begin/end() are guaranteed to be * synchronizing against mapping->tree_lock. |
d10c80955
|
332 333 334 335 |
* * Grabbing old_wb->list_lock, inode->i_lock and mapping->tree_lock * gives us exclusion against all wb related operations on @inode * including IO list manipulations and stat updates. |
682aa8e1a
|
336 |
*/ |
d10c80955
|
337 338 339 340 341 342 343 |
if (old_wb < new_wb) { spin_lock(&old_wb->list_lock); spin_lock_nested(&new_wb->list_lock, SINGLE_DEPTH_NESTING); } else { spin_lock(&new_wb->list_lock); spin_lock_nested(&old_wb->list_lock, SINGLE_DEPTH_NESTING); } |
682aa8e1a
|
344 |
spin_lock(&inode->i_lock); |
d10c80955
|
345 346 347 348 |
spin_lock_irq(&mapping->tree_lock); /* * Once I_FREEING is visible under i_lock, the eviction path owns |
c7f540849
|
349 |
* the inode and we shouldn't modify ->i_io_list. |
d10c80955
|
350 351 352 353 354 355 356 357 358 359 360 361 362 363 |
*/ if (unlikely(inode->i_state & I_FREEING)) goto skip_switch; /* * Count and transfer stats. Note that PAGECACHE_TAG_DIRTY points * to possibly dirty pages while PAGECACHE_TAG_WRITEBACK points to * pages actually under underwriteback. */ radix_tree_for_each_tagged(slot, &mapping->page_tree, &iter, 0, PAGECACHE_TAG_DIRTY) { struct page *page = radix_tree_deref_slot_protected(slot, &mapping->tree_lock); if (likely(page) && PageDirty(page)) { |
3e8f399da
|
364 365 |
dec_wb_stat(old_wb, WB_RECLAIMABLE); inc_wb_stat(new_wb, WB_RECLAIMABLE); |
d10c80955
|
366 367 368 369 370 371 372 373 374 |
} } radix_tree_for_each_tagged(slot, &mapping->page_tree, &iter, 0, PAGECACHE_TAG_WRITEBACK) { struct page *page = radix_tree_deref_slot_protected(slot, &mapping->tree_lock); if (likely(page)) { WARN_ON_ONCE(!PageWriteback(page)); |
3e8f399da
|
375 376 |
dec_wb_stat(old_wb, WB_WRITEBACK); inc_wb_stat(new_wb, WB_WRITEBACK); |
d10c80955
|
377 378 379 380 381 382 383 384 385 386 387 |
} } wb_get(new_wb); /* * Transfer to @new_wb's IO list if necessary. The specific list * @inode was on is ignored and the inode is put on ->b_dirty which * is always correct including from ->b_dirty_time. The transfer * preserves @inode->dirtied_when ordering. */ |
c7f540849
|
388 |
if (!list_empty(&inode->i_io_list)) { |
d10c80955
|
389 |
struct inode *pos; |
c7f540849
|
390 |
inode_io_list_del_locked(inode, old_wb); |
d10c80955
|
391 |
inode->i_wb = new_wb; |
c7f540849
|
392 |
list_for_each_entry(pos, &new_wb->b_dirty, i_io_list) |
d10c80955
|
393 394 395 |
if (time_after_eq(inode->dirtied_when, pos->dirtied_when)) break; |
c7f540849
|
396 |
inode_io_list_move_locked(inode, new_wb, pos->i_io_list.prev); |
d10c80955
|
397 398 399 |
} else { inode->i_wb = new_wb; } |
682aa8e1a
|
400 |
|
d10c80955
|
401 |
/* ->i_wb_frn updates may race wbc_detach_inode() but doesn't matter */ |
682aa8e1a
|
402 403 404 |
inode->i_wb_frn_winner = 0; inode->i_wb_frn_avg_time = 0; inode->i_wb_frn_history = 0; |
d10c80955
|
405 406 |
switched = true; skip_switch: |
682aa8e1a
|
407 408 409 410 411 |
/* * Paired with load_acquire in unlocked_inode_to_wb_begin() and * ensures that the new wb is visible if they see !I_WB_SWITCH. */ smp_store_release(&inode->i_state, inode->i_state & ~I_WB_SWITCH); |
d10c80955
|
412 |
spin_unlock_irq(&mapping->tree_lock); |
682aa8e1a
|
413 |
spin_unlock(&inode->i_lock); |
d10c80955
|
414 415 |
spin_unlock(&new_wb->list_lock); spin_unlock(&old_wb->list_lock); |
682aa8e1a
|
416 |
|
d10c80955
|
417 418 419 420 |
if (switched) { wb_wakeup(new_wb); wb_put(old_wb); } |
682aa8e1a
|
421 |
wb_put(new_wb); |
d10c80955
|
422 423 |
iput(inode); |
682aa8e1a
|
424 |
kfree(isw); |
a1a0e23e4
|
425 426 |
atomic_dec(&isw_nr_in_flight); |
682aa8e1a
|
427 428 429 430 431 432 433 434 435 |
} static void inode_switch_wbs_rcu_fn(struct rcu_head *rcu_head) { struct inode_switch_wbs_context *isw = container_of(rcu_head, struct inode_switch_wbs_context, rcu_head); /* needs to grab bh-unsafe locks, bounce to work item */ INIT_WORK(&isw->work, inode_switch_wbs_work_fn); |
a1a0e23e4
|
436 |
queue_work(isw_wq, &isw->work); |
682aa8e1a
|
437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 |
} /** * inode_switch_wbs - change the wb association of an inode * @inode: target inode * @new_wb_id: ID of the new wb * * Switch @inode's wb association to the wb identified by @new_wb_id. The * switching is performed asynchronously and may fail silently. */ static void inode_switch_wbs(struct inode *inode, int new_wb_id) { struct backing_dev_info *bdi = inode_to_bdi(inode); struct cgroup_subsys_state *memcg_css; struct inode_switch_wbs_context *isw; /* noop if seems to be already in progress */ if (inode->i_state & I_WB_SWITCH) return; isw = kzalloc(sizeof(*isw), GFP_ATOMIC); if (!isw) return; /* find and pin the new wb */ rcu_read_lock(); memcg_css = css_from_id(new_wb_id, &memory_cgrp_subsys); if (memcg_css) isw->new_wb = wb_get_create(bdi, memcg_css, GFP_ATOMIC); rcu_read_unlock(); if (!isw->new_wb) goto out_free; /* while holding I_WB_SWITCH, no one else can update the association */ spin_lock(&inode->i_lock); |
a1a0e23e4
|
472 473 474 475 476 477 |
if (!(inode->i_sb->s_flags & MS_ACTIVE) || inode->i_state & (I_WB_SWITCH | I_FREEING) || inode_to_wb(inode) == isw->new_wb) { spin_unlock(&inode->i_lock); goto out_free; } |
682aa8e1a
|
478 |
inode->i_state |= I_WB_SWITCH; |
745249555
|
479 |
__iget(inode); |
682aa8e1a
|
480 |
spin_unlock(&inode->i_lock); |
682aa8e1a
|
481 |
isw->inode = inode; |
a1a0e23e4
|
482 |
atomic_inc(&isw_nr_in_flight); |
682aa8e1a
|
483 484 485 486 487 488 489 490 491 492 493 494 495 496 |
/* * In addition to synchronizing among switchers, I_WB_SWITCH tells * the RCU protected stat update paths to grab the mapping's * tree_lock so that stat transfer can synchronize against them. * Let's continue after I_WB_SWITCH is guaranteed to be visible. */ call_rcu(&isw->rcu_head, inode_switch_wbs_rcu_fn); return; out_free: if (isw->new_wb) wb_put(isw->new_wb); kfree(isw); } |
87e1d789b
|
497 |
/** |
b16b1deb5
|
498 499 500 501 502 503 504 505 506 507 508 509 |
* wbc_attach_and_unlock_inode - associate wbc with target inode and unlock it * @wbc: writeback_control of interest * @inode: target inode * * @inode is locked and about to be written back under the control of @wbc. * Record @inode's writeback context into @wbc and unlock the i_lock. On * writeback completion, wbc_detach_inode() should be called. This is used * to track the cgroup writeback context. */ void wbc_attach_and_unlock_inode(struct writeback_control *wbc, struct inode *inode) { |
dd73e4b7d
|
510 511 512 513 |
if (!inode_cgwb_enabled(inode)) { spin_unlock(&inode->i_lock); return; } |
b16b1deb5
|
514 |
wbc->wb = inode_to_wb(inode); |
2a8149081
|
515 516 517 518 519 520 521 522 |
wbc->inode = inode; wbc->wb_id = wbc->wb->memcg_css->id; wbc->wb_lcand_id = inode->i_wb_frn_winner; wbc->wb_tcand_id = 0; wbc->wb_bytes = 0; wbc->wb_lcand_bytes = 0; wbc->wb_tcand_bytes = 0; |
b16b1deb5
|
523 524 |
wb_get(wbc->wb); spin_unlock(&inode->i_lock); |
e8a7abf5a
|
525 526 527 528 529 530 531 |
/* * A dying wb indicates that the memcg-blkcg mapping has changed * and a new wb is already serving the memcg. Switch immediately. */ if (unlikely(wb_dying(wbc->wb))) inode_switch_wbs(inode, wbc->wb_id); |
b16b1deb5
|
532 533 534 |
} /** |
2a8149081
|
535 536 |
* wbc_detach_inode - disassociate wbc from inode and perform foreign detection * @wbc: writeback_control of the just finished writeback |
b16b1deb5
|
537 538 539 |
* * To be called after a writeback attempt of an inode finishes and undoes * wbc_attach_and_unlock_inode(). Can be called under any context. |
2a8149081
|
540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 |
* * As concurrent write sharing of an inode is expected to be very rare and * memcg only tracks page ownership on first-use basis severely confining * the usefulness of such sharing, cgroup writeback tracks ownership * per-inode. While the support for concurrent write sharing of an inode * is deemed unnecessary, an inode being written to by different cgroups at * different points in time is a lot more common, and, more importantly, * charging only by first-use can too readily lead to grossly incorrect * behaviors (single foreign page can lead to gigabytes of writeback to be * incorrectly attributed). * * To resolve this issue, cgroup writeback detects the majority dirtier of * an inode and transfers the ownership to it. To avoid unnnecessary * oscillation, the detection mechanism keeps track of history and gives * out the switch verdict only if the foreign usage pattern is stable over * a certain amount of time and/or writeback attempts. * * On each writeback attempt, @wbc tries to detect the majority writer * using Boyer-Moore majority vote algorithm. In addition to the byte * count from the majority voting, it also counts the bytes written for the * current wb and the last round's winner wb (max of last round's current * wb, the winner from two rounds ago, and the last round's majority * candidate). Keeping track of the historical winner helps the algorithm * to semi-reliably detect the most active writer even when it's not the * absolute majority. * * Once the winner of the round is determined, whether the winner is * foreign or not and how much IO time the round consumed is recorded in * inode->i_wb_frn_history. If the amount of recorded foreign IO time is * over a certain threshold, the switch verdict is given. |
b16b1deb5
|
570 571 572 |
*/ void wbc_detach_inode(struct writeback_control *wbc) { |
2a8149081
|
573 574 |
struct bdi_writeback *wb = wbc->wb; struct inode *inode = wbc->inode; |
dd73e4b7d
|
575 576 |
unsigned long avg_time, max_bytes, max_time; u16 history; |
2a8149081
|
577 |
int max_id; |
dd73e4b7d
|
578 579 580 581 582 |
if (!wb) return; history = inode->i_wb_frn_history; avg_time = inode->i_wb_frn_avg_time; |
2a8149081
|
583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 |
/* pick the winner of this round */ if (wbc->wb_bytes >= wbc->wb_lcand_bytes && wbc->wb_bytes >= wbc->wb_tcand_bytes) { max_id = wbc->wb_id; max_bytes = wbc->wb_bytes; } else if (wbc->wb_lcand_bytes >= wbc->wb_tcand_bytes) { max_id = wbc->wb_lcand_id; max_bytes = wbc->wb_lcand_bytes; } else { max_id = wbc->wb_tcand_id; max_bytes = wbc->wb_tcand_bytes; } /* * Calculate the amount of IO time the winner consumed and fold it * into the running average kept per inode. If the consumed IO * time is lower than avag / WB_FRN_TIME_CUT_DIV, ignore it for * deciding whether to switch or not. This is to prevent one-off * small dirtiers from skewing the verdict. */ max_time = DIV_ROUND_UP((max_bytes >> PAGE_SHIFT) << WB_FRN_TIME_SHIFT, wb->avg_write_bandwidth); if (avg_time) avg_time += (max_time >> WB_FRN_TIME_AVG_SHIFT) - (avg_time >> WB_FRN_TIME_AVG_SHIFT); else avg_time = max_time; /* immediate catch up on first run */ if (max_time >= avg_time / WB_FRN_TIME_CUT_DIV) { int slots; /* * The switch verdict is reached if foreign wb's consume * more than a certain proportion of IO time in a * WB_FRN_TIME_PERIOD. This is loosely tracked by 16 slot * history mask where each bit represents one sixteenth of * the period. Determine the number of slots to shift into * history from @max_time. */ slots = min(DIV_ROUND_UP(max_time, WB_FRN_HIST_UNIT), (unsigned long)WB_FRN_HIST_MAX_SLOTS); history <<= slots; if (wbc->wb_id != max_id) history |= (1U << slots) - 1; /* * Switch if the current wb isn't the consistent winner. * If there are multiple closely competing dirtiers, the * inode may switch across them repeatedly over time, which * is okay. The main goal is avoiding keeping an inode on * the wrong wb for an extended period of time. */ |
682aa8e1a
|
635 636 |
if (hweight32(history) > WB_FRN_HIST_THR_SLOTS) inode_switch_wbs(inode, max_id); |
2a8149081
|
637 638 639 640 641 642 643 644 645 |
} /* * Multiple instances of this function may race to update the * following fields but we don't mind occassional inaccuracies. */ inode->i_wb_frn_winner = max_id; inode->i_wb_frn_avg_time = min(avg_time, (unsigned long)U16_MAX); inode->i_wb_frn_history = history; |
b16b1deb5
|
646 647 648 649 650 |
wb_put(wbc->wb); wbc->wb = NULL; } /** |
2a8149081
|
651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 |
* wbc_account_io - account IO issued during writeback * @wbc: writeback_control of the writeback in progress * @page: page being written out * @bytes: number of bytes being written out * * @bytes from @page are about to written out during the writeback * controlled by @wbc. Keep the book for foreign inode detection. See * wbc_detach_inode(). */ void wbc_account_io(struct writeback_control *wbc, struct page *page, size_t bytes) { int id; /* * pageout() path doesn't attach @wbc to the inode being written * out. This is intentional as we don't want the function to block * behind a slow cgroup. Ultimately, we want pageout() to kick off * regular writeback instead of writing things out itself. */ if (!wbc->wb) return; |
2a8149081
|
673 |
id = mem_cgroup_css_from_page(page)->id; |
2a8149081
|
674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 |
if (id == wbc->wb_id) { wbc->wb_bytes += bytes; return; } if (id == wbc->wb_lcand_id) wbc->wb_lcand_bytes += bytes; /* Boyer-Moore majority vote algorithm */ if (!wbc->wb_tcand_bytes) wbc->wb_tcand_id = id; if (id == wbc->wb_tcand_id) wbc->wb_tcand_bytes += bytes; else wbc->wb_tcand_bytes -= min(bytes, wbc->wb_tcand_bytes); } |
5aa2a96b3
|
691 |
EXPORT_SYMBOL_GPL(wbc_account_io); |
2a8149081
|
692 693 |
/** |
703c27088
|
694 |
* inode_congested - test whether an inode is congested |
60292bcc1
|
695 |
* @inode: inode to test for congestion (may be NULL) |
703c27088
|
696 697 698 699 700 701 702 703 704 |
* @cong_bits: mask of WB_[a]sync_congested bits to test * * Tests whether @inode is congested. @cong_bits is the mask of congestion * bits to test and the return value is the mask of set bits. * * If cgroup writeback is enabled for @inode, the congestion state is * determined by whether the cgwb (cgroup bdi_writeback) for the blkcg * associated with @inode is congested; otherwise, the root wb's congestion * state is used. |
60292bcc1
|
705 706 707 |
* * @inode is allowed to be NULL as this function is often called on * mapping->host which is NULL for the swapper space. |
703c27088
|
708 709 710 |
*/ int inode_congested(struct inode *inode, int cong_bits) { |
5cb8b8241
|
711 712 713 714 |
/* * Once set, ->i_wb never becomes NULL while the inode is alive. * Start transaction iff ->i_wb is visible. */ |
aaa2cacf8
|
715 |
if (inode && inode_to_wb_is_valid(inode)) { |
5cb8b8241
|
716 |
struct bdi_writeback *wb; |
7c9b87a78
|
717 718 |
struct wb_lock_cookie lock_cookie = {}; bool congested; |
5cb8b8241
|
719 |
|
7c9b87a78
|
720 |
wb = unlocked_inode_to_wb_begin(inode, &lock_cookie); |
5cb8b8241
|
721 |
congested = wb_congested(wb, cong_bits); |
7c9b87a78
|
722 |
unlocked_inode_to_wb_end(inode, &lock_cookie); |
5cb8b8241
|
723 |
return congested; |
703c27088
|
724 725 726 727 728 |
} return wb_congested(&inode_to_bdi(inode)->wb, cong_bits); } EXPORT_SYMBOL_GPL(inode_congested); |
f2b651216
|
729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 |
/** * wb_split_bdi_pages - split nr_pages to write according to bandwidth * @wb: target bdi_writeback to split @nr_pages to * @nr_pages: number of pages to write for the whole bdi * * Split @wb's portion of @nr_pages according to @wb's write bandwidth in * relation to the total write bandwidth of all wb's w/ dirty inodes on * @wb->bdi. */ static long wb_split_bdi_pages(struct bdi_writeback *wb, long nr_pages) { unsigned long this_bw = wb->avg_write_bandwidth; unsigned long tot_bw = atomic_long_read(&wb->bdi->tot_write_bandwidth); if (nr_pages == LONG_MAX) return LONG_MAX; /* * This may be called on clean wb's and proportional distribution * may not make sense, just use the original @nr_pages in those * cases. In general, we wanna err on the side of writing more. */ if (!tot_bw || this_bw >= tot_bw) return nr_pages; else return DIV_ROUND_UP_ULL((u64)nr_pages * this_bw, tot_bw); } |
db1253604
|
756 |
/** |
db1253604
|
757 758 759 760 761 762 763 764 765 766 767 768 769 770 |
* bdi_split_work_to_wbs - split a wb_writeback_work to all wb's of a bdi * @bdi: target backing_dev_info * @base_work: wb_writeback_work to issue * @skip_if_busy: skip wb's which already have writeback in progress * * Split and issue @base_work to all wb's (bdi_writeback's) of @bdi which * have dirty inodes. If @base_work->nr_page isn't %LONG_MAX, it's * distributed to the busy wbs according to each wb's proportion in the * total active write bandwidth of @bdi. */ static void bdi_split_work_to_wbs(struct backing_dev_info *bdi, struct wb_writeback_work *base_work, bool skip_if_busy) { |
b817525a4
|
771 |
struct bdi_writeback *last_wb = NULL; |
b33e18f61
|
772 773 |
struct bdi_writeback *wb = list_entry(&bdi->wb_list, struct bdi_writeback, bdi_node); |
db1253604
|
774 775 |
might_sleep(); |
db1253604
|
776 777 |
restart: rcu_read_lock(); |
b817525a4
|
778 |
list_for_each_entry_continue_rcu(wb, &bdi->wb_list, bdi_node) { |
8a1270cda
|
779 780 781 782 |
DEFINE_WB_COMPLETION_ONSTACK(fallback_work_done); struct wb_writeback_work fallback_work; struct wb_writeback_work *work; long nr_pages; |
b817525a4
|
783 784 785 786 |
if (last_wb) { wb_put(last_wb); last_wb = NULL; } |
006a0973e
|
787 788 789 790 791 792 |
/* SYNC_ALL writes out I_DIRTY_TIME too */ if (!wb_has_dirty_io(wb) && (base_work->sync_mode == WB_SYNC_NONE || list_empty(&wb->b_dirty_time))) continue; if (skip_if_busy && writeback_in_progress(wb)) |
db1253604
|
793 |
continue; |
8a1270cda
|
794 795 796 797 798 799 800 801 802 |
nr_pages = wb_split_bdi_pages(wb, base_work->nr_pages); work = kmalloc(sizeof(*work), GFP_ATOMIC); if (work) { *work = *base_work; work->nr_pages = nr_pages; work->auto_free = 1; wb_queue_work(wb, work); continue; |
db1253604
|
803 |
} |
8a1270cda
|
804 805 806 807 808 809 810 811 812 |
/* alloc failed, execute synchronously using on-stack fallback */ work = &fallback_work; *work = *base_work; work->nr_pages = nr_pages; work->auto_free = 0; work->done = &fallback_work_done; wb_queue_work(wb, work); |
b817525a4
|
813 814 815 816 817 818 819 |
/* * Pin @wb so that it stays on @bdi->wb_list. This allows * continuing iteration from @wb after dropping and * regrabbing rcu read lock. */ wb_get(wb); last_wb = wb; |
8a1270cda
|
820 821 822 |
rcu_read_unlock(); wb_wait_for_completion(bdi, &fallback_work_done); goto restart; |
db1253604
|
823 824 |
} rcu_read_unlock(); |
b817525a4
|
825 826 827 |
if (last_wb) wb_put(last_wb); |
db1253604
|
828 |
} |
a1a0e23e4
|
829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 |
/** * cgroup_writeback_umount - flush inode wb switches for umount * * This function is called when a super_block is about to be destroyed and * flushes in-flight inode wb switches. An inode wb switch goes through * RCU and then workqueue, so the two need to be flushed in order to ensure * that all previously scheduled switches are finished. As wb switches are * rare occurrences and synchronize_rcu() can take a while, perform * flushing iff wb switches are in flight. */ void cgroup_writeback_umount(void) { if (atomic_read(&isw_nr_in_flight)) { synchronize_rcu(); flush_workqueue(isw_wq); } } static int __init cgroup_writeback_init(void) { isw_wq = alloc_workqueue("inode_switch_wbs", 0, 0); if (!isw_wq) return -ENOMEM; return 0; } fs_initcall(cgroup_writeback_init); |
f2b651216
|
855 |
#else /* CONFIG_CGROUP_WRITEBACK */ |
87e1d789b
|
856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 |
static struct bdi_writeback * locked_inode_to_wb_and_lock_list(struct inode *inode) __releases(&inode->i_lock) __acquires(&wb->list_lock) { struct bdi_writeback *wb = inode_to_wb(inode); spin_unlock(&inode->i_lock); spin_lock(&wb->list_lock); return wb; } static struct bdi_writeback *inode_to_wb_and_lock_list(struct inode *inode) __acquires(&wb->list_lock) { struct bdi_writeback *wb = inode_to_wb(inode); spin_lock(&wb->list_lock); return wb; } |
f2b651216
|
876 877 878 879 |
static long wb_split_bdi_pages(struct bdi_writeback *wb, long nr_pages) { return nr_pages; } |
db1253604
|
880 881 882 883 884 |
static void bdi_split_work_to_wbs(struct backing_dev_info *bdi, struct wb_writeback_work *base_work, bool skip_if_busy) { might_sleep(); |
006a0973e
|
885 |
if (!skip_if_busy || !writeback_in_progress(&bdi->wb)) { |
db1253604
|
886 |
base_work->auto_free = 0; |
db1253604
|
887 888 889 |
wb_queue_work(&bdi->wb, base_work); } } |
703c27088
|
890 |
#endif /* CONFIG_CGROUP_WRITEBACK */ |
c00ddad39
|
891 892 |
void wb_start_writeback(struct bdi_writeback *wb, long nr_pages, bool range_cyclic, enum wb_reason reason) |
b6e51316d
|
893 |
{ |
c00ddad39
|
894 895 896 897 898 899 900 901 902 |
struct wb_writeback_work *work; if (!wb_has_dirty_io(wb)) return; /* * This is WB_SYNC_NONE writeback, so if allocation fails just * wakeup the thread for old dirty data writeback */ |
78ebc2f71
|
903 904 |
work = kzalloc(sizeof(*work), GFP_NOWAIT | __GFP_NOMEMALLOC | __GFP_NOWARN); |
c00ddad39
|
905 |
if (!work) { |
5634cc2aa
|
906 |
trace_writeback_nowork(wb); |
c00ddad39
|
907 908 909 910 911 912 913 914 |
wb_wakeup(wb); return; } work->sync_mode = WB_SYNC_NONE; work->nr_pages = nr_pages; work->range_cyclic = range_cyclic; work->reason = reason; |
ac7b19a34
|
915 |
work->auto_free = 1; |
c00ddad39
|
916 917 |
wb_queue_work(wb, work); |
c5444198c
|
918 |
} |
d3ddec763
|
919 |
|
c5444198c
|
920 |
/** |
9ecf4866c
|
921 922 |
* wb_start_background_writeback - start background writeback * @wb: bdi_writback to write from |
c5444198c
|
923 924 |
* * Description: |
6585027a5
|
925 |
* This makes sure WB_SYNC_NONE background writeback happens. When |
9ecf4866c
|
926 |
* this function returns, it is only guaranteed that for given wb |
6585027a5
|
927 928 |
* some IO is happening if we are over background dirty threshold. * Caller need not hold sb s_umount semaphore. |
c5444198c
|
929 |
*/ |
9ecf4866c
|
930 |
void wb_start_background_writeback(struct bdi_writeback *wb) |
c5444198c
|
931 |
{ |
6585027a5
|
932 933 934 935 |
/* * We just wake up the flusher thread. It will perform background * writeback as soon as there is no other work to do. */ |
5634cc2aa
|
936 |
trace_writeback_wake_background(wb); |
9ecf4866c
|
937 |
wb_wakeup(wb); |
1da177e4c
|
938 939 940 |
} /* |
a66979aba
|
941 942 |
* Remove the inode from the writeback list it is on. */ |
c7f540849
|
943 |
void inode_io_list_del(struct inode *inode) |
a66979aba
|
944 |
{ |
87e1d789b
|
945 |
struct bdi_writeback *wb; |
f758eeabe
|
946 |
|
87e1d789b
|
947 |
wb = inode_to_wb_and_lock_list(inode); |
c7f540849
|
948 |
inode_io_list_del_locked(inode, wb); |
52ebea749
|
949 |
spin_unlock(&wb->list_lock); |
a66979aba
|
950 |
} |
a66979aba
|
951 |
/* |
6c60d2b57
|
952 953 954 955 956 957 958 959 960 |
* mark an inode as under writeback on the sb */ void sb_mark_inode_writeback(struct inode *inode) { struct super_block *sb = inode->i_sb; unsigned long flags; if (list_empty(&inode->i_wb_list)) { spin_lock_irqsave(&sb->s_inode_wblist_lock, flags); |
9a46b04f1
|
961 |
if (list_empty(&inode->i_wb_list)) { |
6c60d2b57
|
962 |
list_add_tail(&inode->i_wb_list, &sb->s_inodes_wb); |
9a46b04f1
|
963 964 |
trace_sb_mark_inode_writeback(inode); } |
6c60d2b57
|
965 966 967 968 969 970 971 972 973 974 975 976 977 978 |
spin_unlock_irqrestore(&sb->s_inode_wblist_lock, flags); } } /* * clear an inode as under writeback on the sb */ void sb_clear_inode_writeback(struct inode *inode) { struct super_block *sb = inode->i_sb; unsigned long flags; if (!list_empty(&inode->i_wb_list)) { spin_lock_irqsave(&sb->s_inode_wblist_lock, flags); |
9a46b04f1
|
979 980 981 982 |
if (!list_empty(&inode->i_wb_list)) { list_del_init(&inode->i_wb_list); trace_sb_clear_inode_writeback(inode); } |
6c60d2b57
|
983 984 985 986 987 |
spin_unlock_irqrestore(&sb->s_inode_wblist_lock, flags); } } /* |
6610a0bc8
|
988 989 990 991 |
* 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
|
992 |
* already the most-recently-dirtied inode on the b_dirty list. If that is |
6610a0bc8
|
993 994 995 |
* the case then the inode must have been redirtied while it was being written * out and we don't reset its dirtied_when. */ |
f758eeabe
|
996 |
static void redirty_tail(struct inode *inode, struct bdi_writeback *wb) |
6610a0bc8
|
997 |
{ |
03ba3782e
|
998 |
if (!list_empty(&wb->b_dirty)) { |
66f3b8e2e
|
999 |
struct inode *tail; |
6610a0bc8
|
1000 |
|
7ccf19a80
|
1001 |
tail = wb_inode(wb->b_dirty.next); |
66f3b8e2e
|
1002 |
if (time_before(inode->dirtied_when, tail->dirtied_when)) |
6610a0bc8
|
1003 1004 |
inode->dirtied_when = jiffies; } |
c7f540849
|
1005 |
inode_io_list_move_locked(inode, wb, &wb->b_dirty); |
6610a0bc8
|
1006 1007 1008 |
} /* |
66f3b8e2e
|
1009 |
* requeue inode for re-scanning after bdi->b_io list is exhausted. |
c986d1e2a
|
1010 |
*/ |
f758eeabe
|
1011 |
static void requeue_io(struct inode *inode, struct bdi_writeback *wb) |
c986d1e2a
|
1012 |
{ |
c7f540849
|
1013 |
inode_io_list_move_locked(inode, wb, &wb->b_more_io); |
c986d1e2a
|
1014 |
} |
1c0eeaf56
|
1015 1016 |
static void inode_sync_complete(struct inode *inode) { |
365b94ae6
|
1017 |
inode->i_state &= ~I_SYNC; |
4eff96dd5
|
1018 1019 |
/* If inode is clean an unused, put it into LRU now... */ inode_add_lru(inode); |
365b94ae6
|
1020 |
/* Waiters must see I_SYNC cleared before being woken up */ |
1c0eeaf56
|
1021 1022 1023 |
smp_mb(); wake_up_bit(&inode->i_state, __I_SYNC); } |
d2caa3c54
|
1024 1025 1026 1027 1028 1029 1030 1031 |
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
|
1032 |
* from permanently stopping the whole bdi writeback. |
d2caa3c54
|
1033 1034 1035 1036 1037 |
*/ ret = ret && time_before_eq(inode->dirtied_when, jiffies); #endif return ret; } |
0ae45f63d
|
1038 |
#define EXPIRE_DIRTY_ATIME 0x0001 |
c986d1e2a
|
1039 |
/* |
0e2f2b236
|
1040 |
* Move expired (dirtied before work->older_than_this) dirty inodes from |
697e6fed9
|
1041 |
* @delaying_queue to @dispatch_queue. |
2c1365791
|
1042 |
*/ |
e84d0a4f8
|
1043 |
static int move_expired_inodes(struct list_head *delaying_queue, |
2c1365791
|
1044 |
struct list_head *dispatch_queue, |
0ae45f63d
|
1045 |
int flags, |
ad4e38dd6
|
1046 |
struct wb_writeback_work *work) |
2c1365791
|
1047 |
{ |
0ae45f63d
|
1048 1049 |
unsigned long *older_than_this = NULL; unsigned long expire_time; |
5c03449d3
|
1050 1051 |
LIST_HEAD(tmp); struct list_head *pos, *node; |
cf137307c
|
1052 |
struct super_block *sb = NULL; |
5c03449d3
|
1053 |
struct inode *inode; |
cf137307c
|
1054 |
int do_sb_sort = 0; |
e84d0a4f8
|
1055 |
int moved = 0; |
5c03449d3
|
1056 |
|
0ae45f63d
|
1057 1058 |
if ((flags & EXPIRE_DIRTY_ATIME) == 0) older_than_this = work->older_than_this; |
a2f487069
|
1059 1060 |
else if (!work->for_sync) { expire_time = jiffies - (dirtytime_expire_interval * HZ); |
0ae45f63d
|
1061 1062 |
older_than_this = &expire_time; } |
2c1365791
|
1063 |
while (!list_empty(delaying_queue)) { |
7ccf19a80
|
1064 |
inode = wb_inode(delaying_queue->prev); |
0ae45f63d
|
1065 1066 |
if (older_than_this && inode_dirtied_after(inode, *older_than_this)) |
2c1365791
|
1067 |
break; |
c7f540849
|
1068 |
list_move(&inode->i_io_list, &tmp); |
a8855990e
|
1069 |
moved++; |
0ae45f63d
|
1070 1071 |
if (flags & EXPIRE_DIRTY_ATIME) set_bit(__I_DIRTY_TIME_EXPIRED, &inode->i_state); |
a8855990e
|
1072 1073 |
if (sb_is_blkdev_sb(inode->i_sb)) continue; |
cf137307c
|
1074 1075 1076 |
if (sb && sb != inode->i_sb) do_sb_sort = 1; sb = inode->i_sb; |
5c03449d3
|
1077 |
} |
cf137307c
|
1078 1079 1080 |
/* just one sb in list, splice to dispatch_queue and we're done */ if (!do_sb_sort) { list_splice(&tmp, dispatch_queue); |
e84d0a4f8
|
1081 |
goto out; |
cf137307c
|
1082 |
} |
5c03449d3
|
1083 1084 |
/* Move inodes from one superblock together */ while (!list_empty(&tmp)) { |
7ccf19a80
|
1085 |
sb = wb_inode(tmp.prev)->i_sb; |
5c03449d3
|
1086 |
list_for_each_prev_safe(pos, node, &tmp) { |
7ccf19a80
|
1087 |
inode = wb_inode(pos); |
5c03449d3
|
1088 |
if (inode->i_sb == sb) |
c7f540849
|
1089 |
list_move(&inode->i_io_list, dispatch_queue); |
5c03449d3
|
1090 |
} |
2c1365791
|
1091 |
} |
e84d0a4f8
|
1092 1093 |
out: return moved; |
2c1365791
|
1094 1095 1096 1097 |
} /* * Queue all expired dirty inodes for io, eldest first. |
4ea879b96
|
1098 1099 1100 1101 1102 1103 1104 1105 |
* 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
|
1106 |
*/ |
ad4e38dd6
|
1107 |
static void queue_io(struct bdi_writeback *wb, struct wb_writeback_work *work) |
66f3b8e2e
|
1108 |
{ |
e84d0a4f8
|
1109 |
int moved; |
0ae45f63d
|
1110 |
|
f758eeabe
|
1111 |
assert_spin_locked(&wb->list_lock); |
4ea879b96
|
1112 |
list_splice_init(&wb->b_more_io, &wb->b_io); |
0ae45f63d
|
1113 1114 1115 |
moved = move_expired_inodes(&wb->b_dirty, &wb->b_io, 0, work); moved += move_expired_inodes(&wb->b_dirty_time, &wb->b_io, EXPIRE_DIRTY_ATIME, work); |
d6c10f1fc
|
1116 1117 |
if (moved) wb_io_lists_populated(wb); |
ad4e38dd6
|
1118 |
trace_writeback_queue_io(wb, work, moved); |
66f3b8e2e
|
1119 |
} |
a9185b41a
|
1120 |
static int write_inode(struct inode *inode, struct writeback_control *wbc) |
08d8e9749
|
1121 |
{ |
9fb0a7da0
|
1122 1123 1124 1125 1126 1127 1128 1129 |
int ret; if (inode->i_sb->s_op->write_inode && !is_bad_inode(inode)) { trace_writeback_write_inode_start(inode, wbc); ret = inode->i_sb->s_op->write_inode(inode, wbc); trace_writeback_write_inode(inode, wbc); return ret; } |
03ba3782e
|
1130 |
return 0; |
08d8e9749
|
1131 |
} |
08d8e9749
|
1132 |
|
2c1365791
|
1133 |
/* |
169ebd901
|
1134 1135 |
* Wait for writeback on an inode to complete. Called with i_lock held. * Caller must make sure inode cannot go away when we drop i_lock. |
01c031945
|
1136 |
*/ |
169ebd901
|
1137 1138 1139 |
static void __inode_wait_for_writeback(struct inode *inode) __releases(inode->i_lock) __acquires(inode->i_lock) |
01c031945
|
1140 1141 1142 1143 1144 |
{ DEFINE_WAIT_BIT(wq, &inode->i_state, __I_SYNC); wait_queue_head_t *wqh; wqh = bit_waitqueue(&inode->i_state, __I_SYNC); |
250df6ed2
|
1145 1146 |
while (inode->i_state & I_SYNC) { spin_unlock(&inode->i_lock); |
743162013
|
1147 1148 |
__wait_on_bit(wqh, &wq, bit_wait, TASK_UNINTERRUPTIBLE); |
250df6ed2
|
1149 |
spin_lock(&inode->i_lock); |
58a9d3d8d
|
1150 |
} |
01c031945
|
1151 1152 1153 |
} /* |
169ebd901
|
1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 |
* Wait for writeback on an inode to complete. Caller must have inode pinned. */ void inode_wait_for_writeback(struct inode *inode) { spin_lock(&inode->i_lock); __inode_wait_for_writeback(inode); spin_unlock(&inode->i_lock); } /* * Sleep until I_SYNC is cleared. This function must be called with i_lock * held and drops it. It is aimed for callers not holding any inode reference * so once i_lock is dropped, inode can go away. */ static void inode_sleep_on_writeback(struct inode *inode) __releases(inode->i_lock) { DEFINE_WAIT(wait); wait_queue_head_t *wqh = bit_waitqueue(&inode->i_state, __I_SYNC); int sleep; prepare_to_wait(wqh, &wait, TASK_UNINTERRUPTIBLE); sleep = inode->i_state & I_SYNC; spin_unlock(&inode->i_lock); if (sleep) schedule(); finish_wait(wqh, &wait); } /* |
ccb26b5a6
|
1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 |
* Find proper writeback list for the inode depending on its current state and * possibly also change of its state while we were doing writeback. Here we * handle things such as livelock prevention or fairness of writeback among * inodes. This function can be called only by flusher thread - noone else * processes all inodes in writeback lists and requeueing inodes behind flusher * thread's back can have unexpected consequences. */ static void requeue_inode(struct inode *inode, struct bdi_writeback *wb, struct writeback_control *wbc) { if (inode->i_state & I_FREEING) return; /* * Sync livelock prevention. Each inode is tagged and synced in one * shot. If still dirty, it will be redirty_tail()'ed below. Update * the dirty time to prevent enqueue and sync it again. */ if ((inode->i_state & I_DIRTY) && (wbc->sync_mode == WB_SYNC_ALL || wbc->tagged_writepages)) inode->dirtied_when = jiffies; |
4f8ad655d
|
1205 1206 1207 1208 1209 1210 1211 1212 |
if (wbc->pages_skipped) { /* * writeback is not making progress due to locked * buffers. Skip this inode for now. */ redirty_tail(inode, wb); return; } |
ccb26b5a6
|
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 |
if (mapping_tagged(inode->i_mapping, PAGECACHE_TAG_DIRTY)) { /* * We didn't write back all the pages. nfs_writepages() * sometimes bales out without doing anything. */ if (wbc->nr_to_write <= 0) { /* Slice used up. Queue for next turn. */ requeue_io(inode, wb); } else { /* * 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. */ redirty_tail(inode, wb); } } 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, wb); |
0ae45f63d
|
1238 |
} else if (inode->i_state & I_DIRTY_TIME) { |
a2f487069
|
1239 |
inode->dirtied_when = jiffies; |
c7f540849
|
1240 |
inode_io_list_move_locked(inode, wb, &wb->b_dirty_time); |
ccb26b5a6
|
1241 1242 |
} else { /* The inode is clean. Remove from writeback lists. */ |
c7f540849
|
1243 |
inode_io_list_del_locked(inode, wb); |
ccb26b5a6
|
1244 1245 1246 1247 |
} } /* |
4f8ad655d
|
1248 1249 1250 |
* Write out an inode and its dirty pages. Do not update the writeback list * linkage. That is left to the caller. The caller is also responsible for * setting I_SYNC flag and calling inode_sync_complete() to clear it. |
1da177e4c
|
1251 1252 |
*/ static int |
cd8ed2a45
|
1253 |
__writeback_single_inode(struct inode *inode, struct writeback_control *wbc) |
1da177e4c
|
1254 |
{ |
1da177e4c
|
1255 |
struct address_space *mapping = inode->i_mapping; |
251d6a471
|
1256 |
long nr_to_write = wbc->nr_to_write; |
01c031945
|
1257 |
unsigned dirty; |
1da177e4c
|
1258 |
int ret; |
4f8ad655d
|
1259 |
WARN_ON(!(inode->i_state & I_SYNC)); |
1da177e4c
|
1260 |
|
9fb0a7da0
|
1261 |
trace_writeback_single_inode_start(inode, wbc, nr_to_write); |
1da177e4c
|
1262 |
ret = do_writepages(mapping, wbc); |
26821ed40
|
1263 1264 1265 |
/* * Make sure to wait on the data before writing out the metadata. * This is important for filesystems that modify metadata on data |
7747bd4bc
|
1266 1267 1268 |
* I/O completion. We don't do it for sync(2) writeback because it has a * separate, external IO completion path and ->sync_fs for guaranteeing * inode metadata is written back correctly. |
26821ed40
|
1269 |
*/ |
7747bd4bc
|
1270 |
if (wbc->sync_mode == WB_SYNC_ALL && !wbc->for_sync) { |
26821ed40
|
1271 |
int err = filemap_fdatawait(mapping); |
1da177e4c
|
1272 1273 1274 |
if (ret == 0) ret = err; } |
5547e8aac
|
1275 1276 1277 1278 1279 |
/* * Some filesystems may redirty the inode during the writeback * due to delalloc, clear dirty metadata flags right before * write_inode() */ |
250df6ed2
|
1280 |
spin_lock(&inode->i_lock); |
9c6ac78eb
|
1281 |
|
5547e8aac
|
1282 |
dirty = inode->i_state & I_DIRTY; |
a2f487069
|
1283 1284 |
if (inode->i_state & I_DIRTY_TIME) { if ((dirty & (I_DIRTY_SYNC | I_DIRTY_DATASYNC)) || |
dc5ff2b1d
|
1285 |
wbc->sync_mode == WB_SYNC_ALL || |
a2f487069
|
1286 1287 1288 1289 1290 1291 1292 1293 1294 |
unlikely(inode->i_state & I_DIRTY_TIME_EXPIRED) || unlikely(time_after(jiffies, (inode->dirtied_time_when + dirtytime_expire_interval * HZ)))) { dirty |= I_DIRTY_TIME | I_DIRTY_TIME_EXPIRED; trace_writeback_lazytime(inode); } } else inode->i_state &= ~I_DIRTY_TIME_EXPIRED; |
0ae45f63d
|
1295 |
inode->i_state &= ~dirty; |
9c6ac78eb
|
1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 |
/* * Paired with smp_mb() in __mark_inode_dirty(). This allows * __mark_inode_dirty() to test i_state without grabbing i_lock - * either they see the I_DIRTY bits cleared or we see the dirtied * inode. * * I_DIRTY_PAGES is always cleared together above even if @mapping * still has dirty pages. The flag is reinstated after smp_mb() if * necessary. This guarantees that either __mark_inode_dirty() * sees clear I_DIRTY_PAGES or we see PAGECACHE_TAG_DIRTY. */ smp_mb(); if (mapping_tagged(mapping, PAGECACHE_TAG_DIRTY)) inode->i_state |= I_DIRTY_PAGES; |
250df6ed2
|
1312 |
spin_unlock(&inode->i_lock); |
9c6ac78eb
|
1313 |
|
0ae45f63d
|
1314 1315 |
if (dirty & I_DIRTY_TIME) mark_inode_dirty_sync(inode); |
26821ed40
|
1316 |
/* Don't write the inode if only I_DIRTY_PAGES was set */ |
0ae45f63d
|
1317 |
if (dirty & ~I_DIRTY_PAGES) { |
a9185b41a
|
1318 |
int err = write_inode(inode, wbc); |
1da177e4c
|
1319 1320 1321 |
if (ret == 0) ret = err; } |
4f8ad655d
|
1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 |
trace_writeback_single_inode(inode, wbc, nr_to_write); return ret; } /* * Write out an inode's dirty pages. Either the caller has an active reference * on the inode or the inode has I_WILL_FREE set. * * This function is designed to be called for writing back one inode which * we go e.g. from filesystem. Flusher thread uses __writeback_single_inode() * and does more profound writeback list handling in writeback_sb_inodes(). */ |
aaf255933
|
1334 1335 |
static int writeback_single_inode(struct inode *inode, struct writeback_control *wbc) |
4f8ad655d
|
1336 |
{ |
aaf255933
|
1337 |
struct bdi_writeback *wb; |
4f8ad655d
|
1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 |
int ret = 0; spin_lock(&inode->i_lock); 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 (wbc->sync_mode != WB_SYNC_ALL) goto out; /* |
169ebd901
|
1350 1351 1352 |
* It's a data-integrity sync. We must wait. Since callers hold * inode reference or inode has I_WILL_FREE set, it cannot go * away under us. |
4f8ad655d
|
1353 |
*/ |
169ebd901
|
1354 |
__inode_wait_for_writeback(inode); |
4f8ad655d
|
1355 1356 1357 |
} WARN_ON(inode->i_state & I_SYNC); /* |
f9b0e058c
|
1358 1359 1360 1361 1362 1363 |
* Skip inode if it is clean and we have no outstanding writeback in * WB_SYNC_ALL mode. We don't want to mess with writeback lists in this * function since flusher thread may be doing for example sync in * parallel and if we move the inode, it could get skipped. So here we * make sure inode is on some writeback list and leave it there unless * we have completely cleaned the inode. |
4f8ad655d
|
1364 |
*/ |
0ae45f63d
|
1365 |
if (!(inode->i_state & I_DIRTY_ALL) && |
f9b0e058c
|
1366 1367 |
(wbc->sync_mode != WB_SYNC_ALL || !mapping_tagged(inode->i_mapping, PAGECACHE_TAG_WRITEBACK))) |
4f8ad655d
|
1368 1369 |
goto out; inode->i_state |= I_SYNC; |
b16b1deb5
|
1370 |
wbc_attach_and_unlock_inode(wbc, inode); |
4f8ad655d
|
1371 |
|
cd8ed2a45
|
1372 |
ret = __writeback_single_inode(inode, wbc); |
1da177e4c
|
1373 |
|
b16b1deb5
|
1374 |
wbc_detach_inode(wbc); |
aaf255933
|
1375 1376 |
wb = inode_to_wb_and_lock_list(inode); |
250df6ed2
|
1377 |
spin_lock(&inode->i_lock); |
4f8ad655d
|
1378 1379 1380 1381 |
/* * If inode is clean, remove it from writeback lists. Otherwise don't * touch it. See comment above for explanation. */ |
0ae45f63d
|
1382 |
if (!(inode->i_state & I_DIRTY_ALL)) |
c7f540849
|
1383 |
inode_io_list_del_locked(inode, wb); |
4f8ad655d
|
1384 |
spin_unlock(&wb->list_lock); |
1c0eeaf56
|
1385 |
inode_sync_complete(inode); |
4f8ad655d
|
1386 1387 |
out: spin_unlock(&inode->i_lock); |
1da177e4c
|
1388 1389 |
return ret; } |
a88a341a7
|
1390 |
static long writeback_chunk_size(struct bdi_writeback *wb, |
1a12d8bd7
|
1391 |
struct wb_writeback_work *work) |
d46db3d58
|
1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 |
{ long pages; /* * 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_sb_inodes() <== called only once * write_cache_pages() <== called once for each inode * (quickly) tag currently dirty pages * (maybe slowly) sync all tagged pages */ if (work->sync_mode == WB_SYNC_ALL || work->tagged_writepages) pages = LONG_MAX; |
1a12d8bd7
|
1410 |
else { |
a88a341a7
|
1411 |
pages = min(wb->avg_write_bandwidth / 2, |
dcc25ae76
|
1412 |
global_wb_domain.dirty_limit / DIRTY_SCOPE); |
1a12d8bd7
|
1413 1414 1415 1416 |
pages = min(pages, work->nr_pages); pages = round_down(pages + MIN_WRITEBACK_PAGES, MIN_WRITEBACK_PAGES); } |
d46db3d58
|
1417 1418 1419 |
return pages; } |
03ba3782e
|
1420 |
/* |
f11c9c5c2
|
1421 |
* Write a portion of b_io inodes which belong to @sb. |
edadfb10b
|
1422 |
* |
d46db3d58
|
1423 |
* Return the number of pages and/or inodes written. |
0ba13fd19
|
1424 1425 1426 1427 |
* * NOTE! This is called with wb->list_lock held, and will * unlock and relock that for each inode it ends up doing * IO for. |
f11c9c5c2
|
1428 |
*/ |
d46db3d58
|
1429 1430 1431 |
static long writeback_sb_inodes(struct super_block *sb, struct bdi_writeback *wb, struct wb_writeback_work *work) |
1da177e4c
|
1432 |
{ |
d46db3d58
|
1433 1434 1435 1436 1437 |
struct writeback_control wbc = { .sync_mode = work->sync_mode, .tagged_writepages = work->tagged_writepages, .for_kupdate = work->for_kupdate, .for_background = work->for_background, |
7747bd4bc
|
1438 |
.for_sync = work->for_sync, |
d46db3d58
|
1439 1440 1441 1442 1443 1444 1445 |
.range_cyclic = work->range_cyclic, .range_start = 0, .range_end = LLONG_MAX, }; unsigned long start_time = jiffies; long write_chunk; long wrote = 0; /* count both pages and inodes */ |
03ba3782e
|
1446 |
while (!list_empty(&wb->b_io)) { |
7ccf19a80
|
1447 |
struct inode *inode = wb_inode(wb->b_io.prev); |
aaf255933
|
1448 |
struct bdi_writeback *tmp_wb; |
edadfb10b
|
1449 1450 |
if (inode->i_sb != sb) { |
d46db3d58
|
1451 |
if (work->sb) { |
edadfb10b
|
1452 1453 1454 1455 1456 |
/* * We only want to write back data for this * superblock, move all inodes not belonging * to it back onto the dirty list. */ |
f758eeabe
|
1457 |
redirty_tail(inode, wb); |
edadfb10b
|
1458 1459 1460 1461 1462 1463 1464 1465 |
continue; } /* * The inode belongs to a different superblock. * Bounce back to the caller to unpin this and * pin the next superblock. */ |
d46db3d58
|
1466 |
break; |
edadfb10b
|
1467 |
} |
9843b76aa
|
1468 |
/* |
331cbdeed
|
1469 1470 |
* Don't bother with new inodes or inodes being freed, first * kind does not need periodic writeout yet, and for the latter |
9843b76aa
|
1471 1472 |
* kind writeout is handled by the freer. */ |
250df6ed2
|
1473 |
spin_lock(&inode->i_lock); |
9843b76aa
|
1474 |
if (inode->i_state & (I_NEW | I_FREEING | I_WILL_FREE)) { |
250df6ed2
|
1475 |
spin_unlock(&inode->i_lock); |
fcc5c2221
|
1476 |
redirty_tail(inode, wb); |
7ef0d7377
|
1477 1478 |
continue; } |
cc1676d91
|
1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 |
if ((inode->i_state & I_SYNC) && wbc.sync_mode != WB_SYNC_ALL) { /* * If this inode is locked for writeback and we are not * doing writeback-for-data-integrity, move it to * b_more_io so that writeback can proceed with the * other inodes on s_io. * * We'll have another go at writing back this inode * when we completed a full scan of b_io. */ spin_unlock(&inode->i_lock); requeue_io(inode, wb); trace_writeback_sb_inodes_requeue(inode); continue; } |
f0d07b7ff
|
1494 |
spin_unlock(&wb->list_lock); |
4f8ad655d
|
1495 1496 1497 1498 1499 |
/* * We already requeued the inode if it had I_SYNC set and we * are doing WB_SYNC_NONE writeback. So this catches only the * WB_SYNC_ALL case. */ |
169ebd901
|
1500 1501 1502 1503 |
if (inode->i_state & I_SYNC) { /* Wait for I_SYNC. This function drops i_lock... */ inode_sleep_on_writeback(inode); /* Inode may be gone, start again */ |
ead188f9f
|
1504 |
spin_lock(&wb->list_lock); |
169ebd901
|
1505 1506 |
continue; } |
4f8ad655d
|
1507 |
inode->i_state |= I_SYNC; |
b16b1deb5
|
1508 |
wbc_attach_and_unlock_inode(&wbc, inode); |
169ebd901
|
1509 |
|
a88a341a7
|
1510 |
write_chunk = writeback_chunk_size(wb, work); |
d46db3d58
|
1511 1512 |
wbc.nr_to_write = write_chunk; wbc.pages_skipped = 0; |
250df6ed2
|
1513 |
|
169ebd901
|
1514 1515 1516 1517 |
/* * We use I_SYNC to pin the inode in memory. While it is set * evict_inode() will wait so the inode cannot be freed. */ |
cd8ed2a45
|
1518 |
__writeback_single_inode(inode, &wbc); |
250df6ed2
|
1519 |
|
b16b1deb5
|
1520 |
wbc_detach_inode(&wbc); |
d46db3d58
|
1521 1522 |
work->nr_pages -= write_chunk - wbc.nr_to_write; wrote += write_chunk - wbc.nr_to_write; |
590dca3a7
|
1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 |
if (need_resched()) { /* * We're trying to balance between building up a nice * long list of IOs to improve our merge rate, and * getting those IOs out quickly for anyone throttling * in balance_dirty_pages(). cond_resched() doesn't * unplug, so get our IOs out the door before we * give up the CPU. */ blk_flush_plug(current); cond_resched(); } |
aaf255933
|
1536 1537 1538 1539 1540 |
/* * Requeue @inode if still dirty. Be careful as @inode may * have been switched to another wb in the meantime. */ tmp_wb = inode_to_wb_and_lock_list(inode); |
4f8ad655d
|
1541 |
spin_lock(&inode->i_lock); |
0ae45f63d
|
1542 |
if (!(inode->i_state & I_DIRTY_ALL)) |
d46db3d58
|
1543 |
wrote++; |
aaf255933
|
1544 |
requeue_inode(inode, tmp_wb, &wbc); |
4f8ad655d
|
1545 |
inode_sync_complete(inode); |
0f1b1fd86
|
1546 |
spin_unlock(&inode->i_lock); |
590dca3a7
|
1547 |
|
aaf255933
|
1548 1549 1550 1551 |
if (unlikely(tmp_wb != wb)) { spin_unlock(&tmp_wb->list_lock); spin_lock(&wb->list_lock); } |
d46db3d58
|
1552 1553 1554 1555 1556 1557 1558 1559 1560 |
/* * bail out to wb_writeback() often enough to check * background threshold and other termination conditions. */ if (wrote) { if (time_is_before_jiffies(start_time + HZ / 10UL)) break; if (work->nr_pages <= 0) break; |
8bc3be275
|
1561 |
} |
1da177e4c
|
1562 |
} |
d46db3d58
|
1563 |
return wrote; |
f11c9c5c2
|
1564 |
} |
d46db3d58
|
1565 1566 |
static long __writeback_inodes_wb(struct bdi_writeback *wb, struct wb_writeback_work *work) |
f11c9c5c2
|
1567 |
{ |
d46db3d58
|
1568 1569 |
unsigned long start_time = jiffies; long wrote = 0; |
38f219776
|
1570 |
|
f11c9c5c2
|
1571 |
while (!list_empty(&wb->b_io)) { |
7ccf19a80
|
1572 |
struct inode *inode = wb_inode(wb->b_io.prev); |
f11c9c5c2
|
1573 |
struct super_block *sb = inode->i_sb; |
9ecc2738a
|
1574 |
|
eb6ef3df4
|
1575 |
if (!trylock_super(sb)) { |
0e995816f
|
1576 |
/* |
eb6ef3df4
|
1577 |
* trylock_super() may fail consistently due to |
0e995816f
|
1578 1579 1580 1581 |
* s_umount being grabbed by someone else. Don't use * requeue_io() to avoid busy retrying the inode/sb. */ redirty_tail(inode, wb); |
edadfb10b
|
1582 |
continue; |
f11c9c5c2
|
1583 |
} |
d46db3d58
|
1584 |
wrote += writeback_sb_inodes(sb, wb, work); |
eb6ef3df4
|
1585 |
up_read(&sb->s_umount); |
f11c9c5c2
|
1586 |
|
d46db3d58
|
1587 1588 1589 1590 1591 1592 1593 |
/* refer to the same tests at the end of writeback_sb_inodes */ if (wrote) { if (time_is_before_jiffies(start_time + HZ / 10UL)) break; if (work->nr_pages <= 0) break; } |
f11c9c5c2
|
1594 |
} |
66f3b8e2e
|
1595 |
/* Leave any unwritten inodes on b_io */ |
d46db3d58
|
1596 |
return wrote; |
66f3b8e2e
|
1597 |
} |
7d9f073b8
|
1598 |
static long writeback_inodes_wb(struct bdi_writeback *wb, long nr_pages, |
0e175a183
|
1599 |
enum wb_reason reason) |
edadfb10b
|
1600 |
{ |
d46db3d58
|
1601 1602 1603 1604 |
struct wb_writeback_work work = { .nr_pages = nr_pages, .sync_mode = WB_SYNC_NONE, .range_cyclic = 1, |
0e175a183
|
1605 |
.reason = reason, |
d46db3d58
|
1606 |
}; |
505a666ee
|
1607 |
struct blk_plug plug; |
edadfb10b
|
1608 |
|
505a666ee
|
1609 |
blk_start_plug(&plug); |
f758eeabe
|
1610 |
spin_lock(&wb->list_lock); |
424b351fe
|
1611 |
if (list_empty(&wb->b_io)) |
ad4e38dd6
|
1612 |
queue_io(wb, &work); |
d46db3d58
|
1613 |
__writeback_inodes_wb(wb, &work); |
f758eeabe
|
1614 |
spin_unlock(&wb->list_lock); |
505a666ee
|
1615 |
blk_finish_plug(&plug); |
edadfb10b
|
1616 |
|
d46db3d58
|
1617 1618 |
return nr_pages - work.nr_pages; } |
03ba3782e
|
1619 |
|
03ba3782e
|
1620 1621 |
/* * Explicit flushing or periodic writeback of "old" data. |
66f3b8e2e
|
1622 |
* |
03ba3782e
|
1623 1624 1625 1626 |
* 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
|
1627 |
* |
03ba3782e
|
1628 1629 1630 |
* 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
|
1631 |
* |
03ba3782e
|
1632 1633 |
* 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
|
1634 |
*/ |
c4a77a6c7
|
1635 |
static long wb_writeback(struct bdi_writeback *wb, |
83ba7b071
|
1636 |
struct wb_writeback_work *work) |
66f3b8e2e
|
1637 |
{ |
e98be2d59
|
1638 |
unsigned long wb_start = jiffies; |
d46db3d58
|
1639 |
long nr_pages = work->nr_pages; |
0dc83bd30
|
1640 |
unsigned long oldest_jif; |
a5989bdc9
|
1641 |
struct inode *inode; |
d46db3d58
|
1642 |
long progress; |
505a666ee
|
1643 |
struct blk_plug plug; |
66f3b8e2e
|
1644 |
|
0dc83bd30
|
1645 1646 |
oldest_jif = jiffies; work->older_than_this = &oldest_jif; |
38f219776
|
1647 |
|
505a666ee
|
1648 |
blk_start_plug(&plug); |
e8dfc3058
|
1649 |
spin_lock(&wb->list_lock); |
03ba3782e
|
1650 1651 |
for (;;) { /* |
d3ddec763
|
1652 |
* Stop writeback when nr_pages has been consumed |
03ba3782e
|
1653 |
*/ |
83ba7b071
|
1654 |
if (work->nr_pages <= 0) |
03ba3782e
|
1655 |
break; |
66f3b8e2e
|
1656 |
|
38f219776
|
1657 |
/* |
aa373cf55
|
1658 1659 1660 1661 1662 1663 |
* 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) && |
f0054bb1e
|
1664 |
!list_empty(&wb->work_list)) |
aa373cf55
|
1665 1666 1667 |
break; /* |
d3ddec763
|
1668 1669 |
* For background writeout, stop when we are below the * background dirty threshold |
38f219776
|
1670 |
*/ |
aa661bbe1
|
1671 |
if (work->for_background && !wb_over_bg_thresh(wb)) |
03ba3782e
|
1672 |
break; |
38f219776
|
1673 |
|
1bc36b642
|
1674 1675 1676 1677 1678 1679 |
/* * Kupdate and background works are special and we want to * include all inodes that need writing. Livelock avoidance is * handled by these works yielding to any other work so we are * safe. */ |
ba9aa8399
|
1680 |
if (work->for_kupdate) { |
0dc83bd30
|
1681 |
oldest_jif = jiffies - |
ba9aa8399
|
1682 |
msecs_to_jiffies(dirty_expire_interval * 10); |
1bc36b642
|
1683 |
} else if (work->for_background) |
0dc83bd30
|
1684 |
oldest_jif = jiffies; |
028c2dd18
|
1685 |
|
5634cc2aa
|
1686 |
trace_writeback_start(wb, work); |
e8dfc3058
|
1687 |
if (list_empty(&wb->b_io)) |
ad4e38dd6
|
1688 |
queue_io(wb, work); |
83ba7b071
|
1689 |
if (work->sb) |
d46db3d58
|
1690 |
progress = writeback_sb_inodes(work->sb, wb, work); |
edadfb10b
|
1691 |
else |
d46db3d58
|
1692 |
progress = __writeback_inodes_wb(wb, work); |
5634cc2aa
|
1693 |
trace_writeback_written(wb, work); |
028c2dd18
|
1694 |
|
e98be2d59
|
1695 |
wb_update_bandwidth(wb, wb_start); |
03ba3782e
|
1696 1697 |
/* |
e6fb6da2e
|
1698 1699 1700 1701 1702 1703 |
* Did we write something? Try for more * * Dirty inodes are moved to b_io for writeback in batches. * The completion of the current batch does not necessarily * mean the overall work is done. So we keep looping as long * as made some progress on cleaning pages or inodes. |
03ba3782e
|
1704 |
*/ |
d46db3d58
|
1705 |
if (progress) |
71fd05a88
|
1706 1707 |
continue; /* |
e6fb6da2e
|
1708 |
* No more inodes for IO, bail |
71fd05a88
|
1709 |
*/ |
b7a2441f9
|
1710 |
if (list_empty(&wb->b_more_io)) |
03ba3782e
|
1711 |
break; |
71fd05a88
|
1712 |
/* |
71fd05a88
|
1713 1714 1715 1716 |
* Nothing written. Wait for some inode to * become available for writeback. Otherwise * we'll just busyloop. */ |
bace92481
|
1717 1718 1719 1720 1721 1722 1723 |
trace_writeback_wait(wb, work); inode = wb_inode(wb->b_more_io.prev); spin_lock(&inode->i_lock); spin_unlock(&wb->list_lock); /* This function drops i_lock... */ inode_sleep_on_writeback(inode); spin_lock(&wb->list_lock); |
03ba3782e
|
1724 |
} |
e8dfc3058
|
1725 |
spin_unlock(&wb->list_lock); |
505a666ee
|
1726 |
blk_finish_plug(&plug); |
03ba3782e
|
1727 |
|
d46db3d58
|
1728 |
return nr_pages - work->nr_pages; |
03ba3782e
|
1729 1730 1731 |
} /* |
83ba7b071
|
1732 |
* Return the next wb_writeback_work struct that hasn't been processed yet. |
03ba3782e
|
1733 |
*/ |
f0054bb1e
|
1734 |
static struct wb_writeback_work *get_next_work_item(struct bdi_writeback *wb) |
03ba3782e
|
1735 |
{ |
83ba7b071
|
1736 |
struct wb_writeback_work *work = NULL; |
03ba3782e
|
1737 |
|
f0054bb1e
|
1738 1739 1740 |
spin_lock_bh(&wb->work_lock); if (!list_empty(&wb->work_list)) { work = list_entry(wb->work_list.next, |
83ba7b071
|
1741 1742 |
struct wb_writeback_work, list); list_del_init(&work->list); |
03ba3782e
|
1743 |
} |
f0054bb1e
|
1744 |
spin_unlock_bh(&wb->work_lock); |
83ba7b071
|
1745 |
return work; |
03ba3782e
|
1746 |
} |
cdf01dd54
|
1747 1748 1749 1750 1751 1752 |
/* * 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) { |
11fb99898
|
1753 1754 |
return global_node_page_state(NR_FILE_DIRTY) + global_node_page_state(NR_UNSTABLE_NFS) + |
cdf01dd54
|
1755 1756 |
get_nr_dirty_inodes(); } |
6585027a5
|
1757 1758 |
static long wb_check_background_flush(struct bdi_writeback *wb) { |
aa661bbe1
|
1759 |
if (wb_over_bg_thresh(wb)) { |
6585027a5
|
1760 1761 1762 1763 1764 1765 |
struct wb_writeback_work work = { .nr_pages = LONG_MAX, .sync_mode = WB_SYNC_NONE, .for_background = 1, .range_cyclic = 1, |
0e175a183
|
1766 |
.reason = WB_REASON_BACKGROUND, |
6585027a5
|
1767 1768 1769 1770 1771 1772 1773 |
}; return wb_writeback(wb, &work); } return 0; } |
03ba3782e
|
1774 1775 1776 1777 |
static long wb_check_old_data_flush(struct bdi_writeback *wb) { unsigned long expired; long nr_pages; |
69b62d01e
|
1778 1779 1780 1781 1782 |
/* * When set to zero, disable periodic writeback */ if (!dirty_writeback_interval) return 0; |
03ba3782e
|
1783 1784 1785 1786 1787 1788 |
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
|
1789 |
nr_pages = get_nr_dirty_pages(); |
03ba3782e
|
1790 |
|
c4a77a6c7
|
1791 |
if (nr_pages) { |
83ba7b071
|
1792 |
struct wb_writeback_work work = { |
c4a77a6c7
|
1793 1794 1795 1796 |
.nr_pages = nr_pages, .sync_mode = WB_SYNC_NONE, .for_kupdate = 1, .range_cyclic = 1, |
0e175a183
|
1797 |
.reason = WB_REASON_PERIODIC, |
c4a77a6c7
|
1798 |
}; |
83ba7b071
|
1799 |
return wb_writeback(wb, &work); |
c4a77a6c7
|
1800 |
} |
03ba3782e
|
1801 1802 1803 1804 1805 1806 1807 |
return 0; } /* * Retrieve work items and do the writeback they describe */ |
25d130ba2
|
1808 |
static long wb_do_writeback(struct bdi_writeback *wb) |
03ba3782e
|
1809 |
{ |
83ba7b071
|
1810 |
struct wb_writeback_work *work; |
c4a77a6c7
|
1811 |
long wrote = 0; |
03ba3782e
|
1812 |
|
4452226ea
|
1813 |
set_bit(WB_writeback_running, &wb->state); |
f0054bb1e
|
1814 |
while ((work = get_next_work_item(wb)) != NULL) { |
5634cc2aa
|
1815 |
trace_writeback_exec(wb, work); |
83ba7b071
|
1816 |
wrote += wb_writeback(wb, work); |
4a3a485b1
|
1817 |
finish_writeback_work(wb, work); |
03ba3782e
|
1818 1819 1820 1821 1822 1823 |
} /* * Check for periodic writeback, kupdated() style */ wrote += wb_check_old_data_flush(wb); |
6585027a5
|
1824 |
wrote += wb_check_background_flush(wb); |
4452226ea
|
1825 |
clear_bit(WB_writeback_running, &wb->state); |
03ba3782e
|
1826 1827 1828 1829 1830 1831 |
return wrote; } /* * Handle writeback of dirty data for the device backed by this bdi. Also |
839a8e866
|
1832 |
* reschedules periodically and does kupdated style flushing. |
03ba3782e
|
1833 |
*/ |
f0054bb1e
|
1834 |
void wb_workfn(struct work_struct *work) |
03ba3782e
|
1835 |
{ |
839a8e866
|
1836 1837 |
struct bdi_writeback *wb = container_of(to_delayed_work(work), struct bdi_writeback, dwork); |
03ba3782e
|
1838 |
long pages_written; |
f0054bb1e
|
1839 |
set_worker_desc("flush-%s", dev_name(wb->bdi->dev)); |
766f91641
|
1840 |
current->flags |= PF_SWAPWRITE; |
455b28646
|
1841 |
|
839a8e866
|
1842 |
if (likely(!current_is_workqueue_rescuer() || |
4452226ea
|
1843 |
!test_bit(WB_registered, &wb->state))) { |
6467716a3
|
1844 |
/* |
f0054bb1e
|
1845 |
* The normal path. Keep writing back @wb until its |
839a8e866
|
1846 |
* work_list is empty. Note that this path is also taken |
f0054bb1e
|
1847 |
* if @wb is shutting down even when we're running off the |
839a8e866
|
1848 |
* rescuer as work_list needs to be drained. |
6467716a3
|
1849 |
*/ |
839a8e866
|
1850 |
do { |
25d130ba2
|
1851 |
pages_written = wb_do_writeback(wb); |
839a8e866
|
1852 |
trace_writeback_pages_written(pages_written); |
f0054bb1e
|
1853 |
} while (!list_empty(&wb->work_list)); |
839a8e866
|
1854 1855 1856 1857 1858 1859 |
} else { /* * bdi_wq can't get enough workers and we're running off * the emergency worker. Don't hog it. Hopefully, 1024 is * enough for efficient IO. */ |
f0054bb1e
|
1860 |
pages_written = writeback_inodes_wb(wb, 1024, |
839a8e866
|
1861 |
WB_REASON_FORKER_THREAD); |
455b28646
|
1862 |
trace_writeback_pages_written(pages_written); |
03ba3782e
|
1863 |
} |
f0054bb1e
|
1864 |
if (!list_empty(&wb->work_list)) |
683b4520d
|
1865 |
wb_wakeup(wb); |
6ca738d60
|
1866 |
else if (wb_has_dirty_io(wb) && dirty_writeback_interval) |
f0054bb1e
|
1867 |
wb_wakeup_delayed(wb); |
455b28646
|
1868 |
|
839a8e866
|
1869 |
current->flags &= ~PF_SWAPWRITE; |
03ba3782e
|
1870 1871 1872 |
} /* |
b8c2f3474
|
1873 1874 |
* Start writeback of `nr_pages' pages. If `nr_pages' is zero, write back * the whole world. |
03ba3782e
|
1875 |
*/ |
0e175a183
|
1876 |
void wakeup_flusher_threads(long nr_pages, enum wb_reason reason) |
03ba3782e
|
1877 |
{ |
b8c2f3474
|
1878 |
struct backing_dev_info *bdi; |
03ba3782e
|
1879 |
|
51350ea0d
|
1880 1881 1882 1883 1884 |
/* * If we are expecting writeback progress we must submit plugged IO. */ if (blk_needs_flush_plug(current)) blk_schedule_flush_plug(current); |
47df3dded
|
1885 1886 |
if (!nr_pages) nr_pages = get_nr_dirty_pages(); |
03ba3782e
|
1887 |
|
b8c2f3474
|
1888 |
rcu_read_lock(); |
f2b651216
|
1889 1890 |
list_for_each_entry_rcu(bdi, &bdi_list, bdi_list) { struct bdi_writeback *wb; |
f2b651216
|
1891 1892 1893 |
if (!bdi_has_dirty_io(bdi)) continue; |
b817525a4
|
1894 |
list_for_each_entry_rcu(wb, &bdi->wb_list, bdi_node) |
f2b651216
|
1895 1896 1897 |
wb_start_writeback(wb, wb_split_bdi_pages(wb, nr_pages), false, reason); } |
cfc4ba536
|
1898 |
rcu_read_unlock(); |
1da177e4c
|
1899 |
} |
a2f487069
|
1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 |
/* * Wake up bdi's periodically to make sure dirtytime inodes gets * written back periodically. We deliberately do *not* check the * b_dirtytime list in wb_has_dirty_io(), since this would cause the * kernel to be constantly waking up once there are any dirtytime * inodes on the system. So instead we define a separate delayed work * function which gets called much more rarely. (By default, only * once every 12 hours.) * * If there is any other write activity going on in the file system, * this function won't be necessary. But if the only thing that has * happened on the file system is a dirtytime inode caused by an atime * update, we need this infrastructure below to make sure that inode * eventually gets pushed out to disk. */ static void wakeup_dirtytime_writeback(struct work_struct *w); static DECLARE_DELAYED_WORK(dirtytime_work, wakeup_dirtytime_writeback); static void wakeup_dirtytime_writeback(struct work_struct *w) { struct backing_dev_info *bdi; rcu_read_lock(); list_for_each_entry_rcu(bdi, &bdi_list, bdi_list) { |
001fe6f61
|
1924 |
struct bdi_writeback *wb; |
001fe6f61
|
1925 |
|
b817525a4
|
1926 |
list_for_each_entry_rcu(wb, &bdi->wb_list, bdi_node) |
6fdf860f1
|
1927 1928 |
if (!list_empty(&wb->b_dirty_time)) wb_wakeup(wb); |
a2f487069
|
1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 |
} rcu_read_unlock(); schedule_delayed_work(&dirtytime_work, dirtytime_expire_interval * HZ); } static int __init start_dirtytime_writeback(void) { schedule_delayed_work(&dirtytime_work, dirtytime_expire_interval * HZ); return 0; } __initcall(start_dirtytime_writeback); |
1efff914a
|
1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 |
int dirtytime_interval_handler(struct ctl_table *table, int write, void __user *buffer, size_t *lenp, loff_t *ppos) { int ret; ret = proc_dointvec_minmax(table, write, buffer, lenp, ppos); if (ret == 0 && write) mod_delayed_work(system_wq, &dirtytime_work, 0); return ret; } |
03ba3782e
|
1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 |
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); } } } /** |
0117d4272
|
1974 1975 1976 1977 1978 1979 1980 |
* __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
|
1981 |
* |
03ba3782e
|
1982 1983 1984 1985 1986 1987 1988 1989 1990 |
* 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
|
1991 |
* |
03ba3782e
|
1992 1993 1994 1995 1996 1997 |
* 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
|
1998 |
*/ |
03ba3782e
|
1999 |
void __mark_inode_dirty(struct inode *inode, int flags) |
1da177e4c
|
2000 |
{ |
dbce03b9e
|
2001 |
#define I_DIRTY_INODE (I_DIRTY_SYNC | I_DIRTY_DATASYNC) |
03ba3782e
|
2002 |
struct super_block *sb = inode->i_sb; |
0ae45f63d
|
2003 2004 2005 |
int dirtytime; trace_writeback_mark_inode_dirty(inode, flags); |
1da177e4c
|
2006 |
|
03ba3782e
|
2007 2008 2009 2010 |
/* * Don't do this for I_DIRTY_PAGES - that doesn't actually * dirty the inode itself */ |
0ae45f63d
|
2011 |
if (flags & (I_DIRTY_SYNC | I_DIRTY_DATASYNC | I_DIRTY_TIME)) { |
9fb0a7da0
|
2012 |
trace_writeback_dirty_inode_start(inode, flags); |
03ba3782e
|
2013 |
if (sb->s_op->dirty_inode) |
aa3857295
|
2014 |
sb->s_op->dirty_inode(inode, flags); |
9fb0a7da0
|
2015 2016 |
trace_writeback_dirty_inode(inode, flags); |
03ba3782e
|
2017 |
} |
0ae45f63d
|
2018 2019 2020 |
if (flags & I_DIRTY_INODE) flags &= ~I_DIRTY_TIME; dirtytime = flags & I_DIRTY_TIME; |
03ba3782e
|
2021 2022 |
/* |
9c6ac78eb
|
2023 2024 |
* Paired with smp_mb() in __writeback_single_inode() for the * following lockless i_state test. See there for details. |
03ba3782e
|
2025 2026 |
*/ smp_mb(); |
0ae45f63d
|
2027 2028 |
if (((inode->i_state & flags) == flags) || (dirtytime && (inode->i_state & I_DIRTY_INODE))) |
03ba3782e
|
2029 2030 2031 2032 |
return; if (unlikely(block_dump)) block_dump___mark_inode_dirty(inode); |
250df6ed2
|
2033 |
spin_lock(&inode->i_lock); |
0ae45f63d
|
2034 2035 |
if (dirtytime && (inode->i_state & I_DIRTY_INODE)) goto out_unlock_inode; |
03ba3782e
|
2036 2037 |
if ((inode->i_state & flags) != flags) { const int was_dirty = inode->i_state & I_DIRTY; |
52ebea749
|
2038 |
inode_attach_wb(inode, NULL); |
0ae45f63d
|
2039 2040 |
if (flags & I_DIRTY_INODE) inode->i_state &= ~I_DIRTY_TIME; |
03ba3782e
|
2041 2042 2043 2044 2045 2046 2047 2048 |
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
|
2049 |
goto out_unlock_inode; |
03ba3782e
|
2050 2051 2052 2053 2054 2055 |
/* * Only add valid (hashed) inodes to the superblock's * dirty list. Add blockdev inodes as well. */ if (!S_ISBLK(inode->i_mode)) { |
1d3382cbf
|
2056 |
if (inode_unhashed(inode)) |
250df6ed2
|
2057 |
goto out_unlock_inode; |
03ba3782e
|
2058 |
} |
a4ffdde6e
|
2059 |
if (inode->i_state & I_FREEING) |
250df6ed2
|
2060 |
goto out_unlock_inode; |
03ba3782e
|
2061 2062 2063 2064 2065 2066 |
/* * 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) { |
87e1d789b
|
2067 |
struct bdi_writeback *wb; |
d6c10f1fc
|
2068 |
struct list_head *dirty_list; |
a66979aba
|
2069 |
bool wakeup_bdi = false; |
253c34e9b
|
2070 |
|
87e1d789b
|
2071 |
wb = locked_inode_to_wb_and_lock_list(inode); |
253c34e9b
|
2072 |
|
0747259d1
|
2073 2074 2075 2076 |
WARN(bdi_cap_writeback_dirty(wb->bdi) && !test_bit(WB_registered, &wb->state), "bdi-%s not registered ", wb->bdi->name); |
03ba3782e
|
2077 2078 |
inode->dirtied_when = jiffies; |
a2f487069
|
2079 2080 |
if (dirtytime) inode->dirtied_time_when = jiffies; |
d6c10f1fc
|
2081 |
|
a2f487069
|
2082 |
if (inode->i_state & (I_DIRTY_INODE | I_DIRTY_PAGES)) |
0747259d1
|
2083 |
dirty_list = &wb->b_dirty; |
a2f487069
|
2084 |
else |
0747259d1
|
2085 |
dirty_list = &wb->b_dirty_time; |
d6c10f1fc
|
2086 |
|
c7f540849
|
2087 |
wakeup_bdi = inode_io_list_move_locked(inode, wb, |
d6c10f1fc
|
2088 |
dirty_list); |
0747259d1
|
2089 |
spin_unlock(&wb->list_lock); |
0ae45f63d
|
2090 |
trace_writeback_dirty_inode_enqueue(inode); |
a66979aba
|
2091 |
|
d6c10f1fc
|
2092 2093 2094 2095 2096 2097 |
/* * 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. */ |
0747259d1
|
2098 2099 |
if (bdi_cap_writeback_dirty(wb->bdi) && wakeup_bdi) wb_wakeup_delayed(wb); |
a66979aba
|
2100 |
return; |
1da177e4c
|
2101 |
} |
1da177e4c
|
2102 |
} |
250df6ed2
|
2103 2104 |
out_unlock_inode: spin_unlock(&inode->i_lock); |
253c34e9b
|
2105 |
|
dbce03b9e
|
2106 |
#undef I_DIRTY_INODE |
03ba3782e
|
2107 2108 |
} EXPORT_SYMBOL(__mark_inode_dirty); |
e97fedb9e
|
2109 2110 2111 2112 2113 2114 2115 2116 2117 |
/* * The @s_sync_lock is used to serialise concurrent sync operations * to avoid lock contention problems with concurrent wait_sb_inodes() calls. * Concurrent callers will block on the s_sync_lock rather than doing contending * walks. The queueing maintains sync(2) required behaviour as all the IO that * has been issued up to the time this function is enter is guaranteed to be * completed by the time we have gained the lock and waited for all IO that is * in progress regardless of the order callers are granted the lock. */ |
b6e51316d
|
2118 |
static void wait_sb_inodes(struct super_block *sb) |
03ba3782e
|
2119 |
{ |
6c60d2b57
|
2120 |
LIST_HEAD(sync_list); |
03ba3782e
|
2121 2122 2123 2124 2125 |
/* * We need to be protected against the filesystem going from * r/o to r/w or vice versa. */ |
b6e51316d
|
2126 |
WARN_ON(!rwsem_is_locked(&sb->s_umount)); |
03ba3782e
|
2127 |
|
e97fedb9e
|
2128 |
mutex_lock(&sb->s_sync_lock); |
03ba3782e
|
2129 2130 |
/* |
6c60d2b57
|
2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 |
* Splice the writeback list onto a temporary list to avoid waiting on * inodes that have started writeback after this point. * * Use rcu_read_lock() to keep the inodes around until we have a * reference. s_inode_wblist_lock protects sb->s_inodes_wb as well as * the local list because inodes can be dropped from either by writeback * completion. */ rcu_read_lock(); spin_lock_irq(&sb->s_inode_wblist_lock); list_splice_init(&sb->s_inodes_wb, &sync_list); /* * 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. |
03ba3782e
|
2149 |
*/ |
6c60d2b57
|
2150 2151 2152 |
while (!list_empty(&sync_list)) { struct inode *inode = list_first_entry(&sync_list, struct inode, i_wb_list); |
250df6ed2
|
2153 |
struct address_space *mapping = inode->i_mapping; |
03ba3782e
|
2154 |
|
6c60d2b57
|
2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 |
/* * Move each inode back to the wb list before we drop the lock * to preserve consistency between i_wb_list and the mapping * writeback tag. Writeback completion is responsible to remove * the inode from either list once the writeback tag is cleared. */ list_move_tail(&inode->i_wb_list, &sb->s_inodes_wb); /* * The mapping can appear untagged while still on-list since we * do not have the mapping lock. Skip it here, wb completion * will remove it. */ if (!mapping_tagged(mapping, PAGECACHE_TAG_WRITEBACK)) continue; spin_unlock_irq(&sb->s_inode_wblist_lock); |
250df6ed2
|
2172 |
spin_lock(&inode->i_lock); |
6c60d2b57
|
2173 |
if (inode->i_state & (I_FREEING|I_WILL_FREE|I_NEW)) { |
250df6ed2
|
2174 |
spin_unlock(&inode->i_lock); |
6c60d2b57
|
2175 2176 |
spin_lock_irq(&sb->s_inode_wblist_lock); |
03ba3782e
|
2177 |
continue; |
250df6ed2
|
2178 |
} |
03ba3782e
|
2179 |
__iget(inode); |
250df6ed2
|
2180 |
spin_unlock(&inode->i_lock); |
6c60d2b57
|
2181 |
rcu_read_unlock(); |
03ba3782e
|
2182 |
|
aa750fd71
|
2183 2184 2185 2186 2187 2188 |
/* * We keep the error status of individual mapping so that * applications can catch the writeback error using fsync(2). * See filemap_fdatawait_keep_errors() for details. */ filemap_fdatawait_keep_errors(mapping); |
03ba3782e
|
2189 2190 |
cond_resched(); |
6c60d2b57
|
2191 2192 2193 2194 |
iput(inode); rcu_read_lock(); spin_lock_irq(&sb->s_inode_wblist_lock); |
03ba3782e
|
2195 |
} |
6c60d2b57
|
2196 2197 |
spin_unlock_irq(&sb->s_inode_wblist_lock); rcu_read_unlock(); |
e97fedb9e
|
2198 |
mutex_unlock(&sb->s_sync_lock); |
1da177e4c
|
2199 |
} |
f30a7d0cc
|
2200 2201 |
static void __writeback_inodes_sb_nr(struct super_block *sb, unsigned long nr, enum wb_reason reason, bool skip_if_busy) |
1da177e4c
|
2202 |
{ |
cc395d7f1
|
2203 |
DEFINE_WB_COMPLETION_ONSTACK(done); |
83ba7b071
|
2204 |
struct wb_writeback_work work = { |
6e6938b6d
|
2205 2206 2207 2208 2209 |
.sb = sb, .sync_mode = WB_SYNC_NONE, .tagged_writepages = 1, .done = &done, .nr_pages = nr, |
0e175a183
|
2210 |
.reason = reason, |
3c4d71653
|
2211 |
}; |
e79729123
|
2212 |
struct backing_dev_info *bdi = sb->s_bdi; |
d8a8559cd
|
2213 |
|
e79729123
|
2214 |
if (!bdi_has_dirty_io(bdi) || bdi == &noop_backing_dev_info) |
6eedc7015
|
2215 |
return; |
cf37e9724
|
2216 |
WARN_ON(!rwsem_is_locked(&sb->s_umount)); |
f30a7d0cc
|
2217 |
|
db1253604
|
2218 |
bdi_split_work_to_wbs(sb->s_bdi, &work, skip_if_busy); |
cc395d7f1
|
2219 |
wb_wait_for_completion(bdi, &done); |
e913fc825
|
2220 |
} |
f30a7d0cc
|
2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 |
/** * writeback_inodes_sb_nr - writeback dirty inodes from given super_block * @sb: the superblock * @nr: the number of pages to write * @reason: reason why some writeback work initiated * * 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_nr(struct super_block *sb, unsigned long nr, enum wb_reason reason) { __writeback_inodes_sb_nr(sb, nr, reason, false); } |
3259f8bed
|
2238 2239 2240 2241 2242 |
EXPORT_SYMBOL(writeback_inodes_sb_nr); /** * writeback_inodes_sb - writeback dirty inodes from given super_block * @sb: the superblock |
786228ab3
|
2243 |
* @reason: reason why some writeback work was initiated |
3259f8bed
|
2244 2245 2246 2247 2248 |
* * 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. */ |
0e175a183
|
2249 |
void writeback_inodes_sb(struct super_block *sb, enum wb_reason reason) |
3259f8bed
|
2250 |
{ |
0e175a183
|
2251 |
return writeback_inodes_sb_nr(sb, get_nr_dirty_pages(), reason); |
3259f8bed
|
2252 |
} |
0e3c9a228
|
2253 |
EXPORT_SYMBOL(writeback_inodes_sb); |
e913fc825
|
2254 2255 |
/** |
10ee27a06
|
2256 |
* try_to_writeback_inodes_sb_nr - try to start writeback if none underway |
17bd55d03
|
2257 |
* @sb: the superblock |
10ee27a06
|
2258 2259 |
* @nr: the number of pages to write * @reason: the reason of writeback |
17bd55d03
|
2260 |
* |
10ee27a06
|
2261 |
* Invoke writeback_inodes_sb_nr if no writeback is currently underway. |
17bd55d03
|
2262 2263 |
* Returns 1 if writeback was started, 0 if not. */ |
f30a7d0cc
|
2264 2265 |
bool try_to_writeback_inodes_sb_nr(struct super_block *sb, unsigned long nr, enum wb_reason reason) |
17bd55d03
|
2266 |
{ |
10ee27a06
|
2267 |
if (!down_read_trylock(&sb->s_umount)) |
f30a7d0cc
|
2268 |
return false; |
10ee27a06
|
2269 |
|
f30a7d0cc
|
2270 |
__writeback_inodes_sb_nr(sb, nr, reason, true); |
10ee27a06
|
2271 |
up_read(&sb->s_umount); |
f30a7d0cc
|
2272 |
return true; |
17bd55d03
|
2273 |
} |
10ee27a06
|
2274 |
EXPORT_SYMBOL(try_to_writeback_inodes_sb_nr); |
17bd55d03
|
2275 2276 |
/** |
10ee27a06
|
2277 |
* try_to_writeback_inodes_sb - try to start writeback if none underway |
3259f8bed
|
2278 |
* @sb: the superblock |
786228ab3
|
2279 |
* @reason: reason why some writeback work was initiated |
3259f8bed
|
2280 |
* |
10ee27a06
|
2281 |
* Implement by try_to_writeback_inodes_sb_nr() |
3259f8bed
|
2282 2283 |
* Returns 1 if writeback was started, 0 if not. */ |
f30a7d0cc
|
2284 |
bool try_to_writeback_inodes_sb(struct super_block *sb, enum wb_reason reason) |
3259f8bed
|
2285 |
{ |
10ee27a06
|
2286 |
return try_to_writeback_inodes_sb_nr(sb, get_nr_dirty_pages(), reason); |
3259f8bed
|
2287 |
} |
10ee27a06
|
2288 |
EXPORT_SYMBOL(try_to_writeback_inodes_sb); |
3259f8bed
|
2289 2290 |
/** |
d8a8559cd
|
2291 |
* sync_inodes_sb - sync sb inode pages |
0dc83bd30
|
2292 |
* @sb: the superblock |
d8a8559cd
|
2293 2294 |
* * This function writes and waits on any dirty inode belonging to this |
0dc83bd30
|
2295 |
* super_block. |
d8a8559cd
|
2296 |
*/ |
0dc83bd30
|
2297 |
void sync_inodes_sb(struct super_block *sb) |
d8a8559cd
|
2298 |
{ |
cc395d7f1
|
2299 |
DEFINE_WB_COMPLETION_ONSTACK(done); |
83ba7b071
|
2300 |
struct wb_writeback_work work = { |
3c4d71653
|
2301 2302 2303 2304 |
.sb = sb, .sync_mode = WB_SYNC_ALL, .nr_pages = LONG_MAX, .range_cyclic = 0, |
83ba7b071
|
2305 |
.done = &done, |
0e175a183
|
2306 |
.reason = WB_REASON_SYNC, |
7747bd4bc
|
2307 |
.for_sync = 1, |
3c4d71653
|
2308 |
}; |
e79729123
|
2309 |
struct backing_dev_info *bdi = sb->s_bdi; |
3c4d71653
|
2310 |
|
006a0973e
|
2311 2312 2313 2314 2315 2316 |
/* * Can't skip on !bdi_has_dirty() because we should wait for !dirty * inodes under writeback and I_DIRTY_TIME inodes ignored by * bdi_has_dirty() need to be written out too. */ if (bdi == &noop_backing_dev_info) |
6eedc7015
|
2317 |
return; |
cf37e9724
|
2318 |
WARN_ON(!rwsem_is_locked(&sb->s_umount)); |
db1253604
|
2319 |
bdi_split_work_to_wbs(bdi, &work, false); |
cc395d7f1
|
2320 |
wb_wait_for_completion(bdi, &done); |
83ba7b071
|
2321 |
|
b6e51316d
|
2322 |
wait_sb_inodes(sb); |
1da177e4c
|
2323 |
} |
d8a8559cd
|
2324 |
EXPORT_SYMBOL(sync_inodes_sb); |
1da177e4c
|
2325 |
|
1da177e4c
|
2326 |
/** |
7f04c26d7
|
2327 2328 2329 2330 2331 2332 |
* 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
|
2333 |
* |
7f04c26d7
|
2334 |
* The caller must either have a ref on the inode or must have set I_WILL_FREE. |
1da177e4c
|
2335 |
*/ |
1da177e4c
|
2336 2337 |
int write_inode_now(struct inode *inode, int sync) { |
1da177e4c
|
2338 2339 |
struct writeback_control wbc = { .nr_to_write = LONG_MAX, |
18914b188
|
2340 |
.sync_mode = sync ? WB_SYNC_ALL : WB_SYNC_NONE, |
111ebb6e6
|
2341 2342 |
.range_start = 0, .range_end = LLONG_MAX, |
1da177e4c
|
2343 2344 2345 |
}; if (!mapping_cap_writeback_dirty(inode->i_mapping)) |
49364ce25
|
2346 |
wbc.nr_to_write = 0; |
1da177e4c
|
2347 2348 |
might_sleep(); |
aaf255933
|
2349 |
return writeback_single_inode(inode, &wbc); |
1da177e4c
|
2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 |
} 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) { |
aaf255933
|
2366 |
return writeback_single_inode(inode, wbc); |
1da177e4c
|
2367 2368 |
} EXPORT_SYMBOL(sync_inode); |
c37650161
|
2369 2370 |
/** |
c691b9d98
|
2371 |
* sync_inode_metadata - write an inode to disk |
c37650161
|
2372 2373 2374 |
* @inode: the inode to sync * @wait: wait for I/O to complete. * |
c691b9d98
|
2375 |
* Write an inode to disk and adjust its dirty state after completion. |
c37650161
|
2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 |
* * 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); |