Blame view
fs/fs-writeback.c
75 KB
457c89965 treewide: Add SPD... |
1 |
// SPDX-License-Identifier: GPL-2.0-only |
1da177e4c Linux-2.6.12-rc2 |
2 3 4 5 6 7 8 9 10 11 |
/* * 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 Remove Andrew Mor... |
12 |
* 10Apr2002 Andrew Morton |
1da177e4c Linux-2.6.12-rc2 |
13 14 15 16 17 |
* Split out of fs/inode.c * Additions for address_space-based writeback */ #include <linux/kernel.h> |
630d9c472 fs: reduce the us... |
18 |
#include <linux/export.h> |
1da177e4c Linux-2.6.12-rc2 |
19 |
#include <linux/spinlock.h> |
5a0e3ad6a include cleanup: ... |
20 |
#include <linux/slab.h> |
1da177e4c Linux-2.6.12-rc2 |
21 22 23 |
#include <linux/sched.h> #include <linux/fs.h> #include <linux/mm.h> |
bc31b86a5 writeback: move M... |
24 |
#include <linux/pagemap.h> |
03ba3782e writeback: switch... |
25 |
#include <linux/kthread.h> |
1da177e4c Linux-2.6.12-rc2 |
26 27 28 |
#include <linux/writeback.h> #include <linux/blkdev.h> #include <linux/backing-dev.h> |
455b28646 writeback: Initia... |
29 |
#include <linux/tracepoint.h> |
719ea2fbb new helpers: lock... |
30 |
#include <linux/device.h> |
21c6321fb writeback: reloca... |
31 |
#include <linux/memcontrol.h> |
07f3f05c1 [PATCH] BLOCK: Mo... |
32 |
#include "internal.h" |
1da177e4c Linux-2.6.12-rc2 |
33 |
|
d0bceac74 writeback: get ri... |
34 |
/* |
bc31b86a5 writeback: move M... |
35 36 |
* 4MB minimal write chunk size */ |
09cbfeaf1 mm, fs: get rid o... |
37 |
#define MIN_WRITEBACK_PAGES (4096UL >> (PAGE_SHIFT - 10)) |
bc31b86a5 writeback: move M... |
38 39 |
/* |
c4a77a6c7 writeback: make w... |
40 41 |
* Passed into wb_writeback(), essentially a subset of writeback_control */ |
83ba7b071 writeback: simpli... |
42 |
struct wb_writeback_work { |
c4a77a6c7 writeback: make w... |
43 44 45 |
long nr_pages; struct super_block *sb; enum writeback_sync_modes sync_mode; |
6e6938b6d writeback: introd... |
46 |
unsigned int tagged_writepages:1; |
52957fe1c fs-writeback.c: b... |
47 48 49 |
unsigned int for_kupdate:1; unsigned int range_cyclic:1; unsigned int for_background:1; |
7747bd4bc sync: don't block... |
50 |
unsigned int for_sync:1; /* sync(2) WB_SYNC_ALL writeback */ |
ac7b19a34 writeback: add wb... |
51 |
unsigned int auto_free:1; /* free on completion */ |
0e175a183 writeback: Add a ... |
52 |
enum wb_reason reason; /* why was writeback initiated? */ |
c4a77a6c7 writeback: make w... |
53 |
|
8010c3b63 writeback: add co... |
54 |
struct list_head list; /* pending work list */ |
cc395d7f1 writeback: implem... |
55 |
struct wb_completion *done; /* set if the caller waits */ |
03ba3782e writeback: switch... |
56 |
}; |
a2f487069 fs: make sure the... |
57 58 59 60 61 62 63 64 65 66 67 |
/* * 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 fs: inode split I... |
68 69 |
static inline struct inode *wb_inode(struct list_head *head) { |
c7f540849 inode: rename i_w... |
70 |
return list_entry(head, struct inode, i_io_list); |
7ccf19a80 fs: inode split I... |
71 |
} |
15eb77a07 writeback: fix NU... |
72 73 74 75 76 77 78 |
/* * 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 GFS2: journal dat... |
79 |
EXPORT_TRACEPOINT_SYMBOL_GPL(wbc_writepage); |
d6c10f1fc writeback: implem... |
80 81 82 83 84 85 |
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 writeback: make b... |
86 |
WARN_ON_ONCE(!wb->avg_write_bandwidth); |
766a9d6e6 writeback: implem... |
87 88 |
atomic_long_add(wb->avg_write_bandwidth, &wb->bdi->tot_write_bandwidth); |
d6c10f1fc writeback: implem... |
89 90 91 92 93 94 95 |
return true; } } static void wb_io_lists_depopulated(struct bdi_writeback *wb) { if (wb_has_dirty_io(wb) && list_empty(&wb->b_dirty) && |
766a9d6e6 writeback: implem... |
96 |
list_empty(&wb->b_io) && list_empty(&wb->b_more_io)) { |
d6c10f1fc writeback: implem... |
97 |
clear_bit(WB_has_dirty_io, &wb->state); |
95a46c65e writeback: make b... |
98 99 |
WARN_ON_ONCE(atomic_long_sub_return(wb->avg_write_bandwidth, &wb->bdi->tot_write_bandwidth) < 0); |
766a9d6e6 writeback: implem... |
100 |
} |
d6c10f1fc writeback: implem... |
101 102 103 |
} /** |
c7f540849 inode: rename i_w... |
104 |
* inode_io_list_move_locked - move an inode onto a bdi_writeback IO list |
d6c10f1fc writeback: implem... |
105 106 |
* @inode: inode to be moved * @wb: target bdi_writeback |
bbbc3c1cf writeback: update... |
107 |
* @head: one of @wb->b_{dirty|io|more_io|dirty_time} |
d6c10f1fc writeback: implem... |
108 |
* |
c7f540849 inode: rename i_w... |
109 |
* Move @inode->i_io_list to @list of @wb and set %WB_has_dirty_io. |
d6c10f1fc writeback: implem... |
110 111 112 |
* Returns %true if @inode is the first occupant of the !dirty_time IO * lists; otherwise, %false. */ |
c7f540849 inode: rename i_w... |
113 |
static bool inode_io_list_move_locked(struct inode *inode, |
d6c10f1fc writeback: implem... |
114 115 116 117 |
struct bdi_writeback *wb, struct list_head *head) { assert_spin_locked(&wb->list_lock); |
c7f540849 inode: rename i_w... |
118 |
list_move(&inode->i_io_list, head); |
d6c10f1fc writeback: implem... |
119 120 121 122 123 124 125 126 127 128 |
/* 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 inode: rename i_w... |
129 |
* inode_io_list_del_locked - remove an inode from its bdi_writeback IO list |
d6c10f1fc writeback: implem... |
130 131 132 133 134 135 |
* @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 inode: rename i_w... |
136 |
static void inode_io_list_del_locked(struct inode *inode, |
d6c10f1fc writeback: implem... |
137 138 139 |
struct bdi_writeback *wb) { assert_spin_locked(&wb->list_lock); |
8eab2b531 writeback: Protec... |
140 |
assert_spin_locked(&inode->i_lock); |
d6c10f1fc writeback: implem... |
141 |
|
cb0c74450 writeback: Avoid ... |
142 |
inode->i_state &= ~I_SYNC_QUEUED; |
c7f540849 inode: rename i_w... |
143 |
list_del_init(&inode->i_io_list); |
d6c10f1fc writeback: implem... |
144 145 |
wb_io_lists_depopulated(wb); } |
f0054bb1e writeback: move b... |
146 |
static void wb_wakeup(struct bdi_writeback *wb) |
5acda9d12 bdi: avoid oops o... |
147 |
{ |
f0054bb1e writeback: move b... |
148 149 150 151 |
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 bdi: avoid oops o... |
152 |
} |
4a3a485b1 writeback: fix me... |
153 154 155 156 157 158 159 |
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); |
8e00c4e9d writeback: fix us... |
160 161 162 163 164 165 166 |
if (done) { wait_queue_head_t *waitq = done->waitq; /* @done can't be accessed after the following dec */ if (atomic_dec_and_test(&done->cnt)) wake_up_all(waitq); } |
4a3a485b1 writeback: fix me... |
167 |
} |
f0054bb1e writeback: move b... |
168 169 |
static void wb_queue_work(struct bdi_writeback *wb, struct wb_writeback_work *work) |
6585027a5 writeback: integr... |
170 |
{ |
5634cc2aa writeback: update... |
171 |
trace_writeback_queue(wb, work); |
6585027a5 writeback: integr... |
172 |
|
cc395d7f1 writeback: implem... |
173 174 |
if (work->done) atomic_inc(&work->done->cnt); |
4a3a485b1 writeback: fix me... |
175 176 177 178 179 180 181 182 |
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 writeback: move b... |
183 |
spin_unlock_bh(&wb->work_lock); |
1da177e4c Linux-2.6.12-rc2 |
184 |
} |
cc395d7f1 writeback: implem... |
185 186 |
/** * wb_wait_for_completion - wait for completion of bdi_writeback_works |
cc395d7f1 writeback: implem... |
187 188 189 |
* @done: target wb_completion * * Wait for one or more work items issued to @bdi with their ->done field |
5b9cce4c7 writeback: Genera... |
190 191 192 |
* set to @done, which should have been initialized with * DEFINE_WB_COMPLETION(). This function returns after all such work items * are completed. Work items which are waited upon aren't freed |
cc395d7f1 writeback: implem... |
193 194 |
* automatically on completion. */ |
5b9cce4c7 writeback: Genera... |
195 |
void wb_wait_for_completion(struct wb_completion *done) |
cc395d7f1 writeback: implem... |
196 197 |
{ atomic_dec(&done->cnt); /* put down the initial count */ |
5b9cce4c7 writeback: Genera... |
198 |
wait_event(*done->waitq, !atomic_read(&done->cnt)); |
cc395d7f1 writeback: implem... |
199 |
} |
703c27088 writeback: implem... |
200 |
#ifdef CONFIG_CGROUP_WRITEBACK |
55a694dff writeback, cgroup... |
201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 |
/* * Parameters for foreign inode detection, see wbc_detach_inode() to see * how they're used. * * These paramters are inherently heuristical as the detection target * itself is fuzzy. All we want to do is detaching an inode from the * current owner if it's being written to by some other cgroups too much. * * The current cgroup writeback is built on the assumption that multiple * cgroups writing to the same inode concurrently is very rare and a mode * of operation which isn't well supported. As such, the goal is not * taking too long when a different cgroup takes over an inode while * avoiding too aggressive flip-flops from occasional foreign writes. * * We record, very roughly, 2s worth of IO time history and if more than * half of that is foreign, trigger the switch. The recording is quantized * to 16 slots. To avoid tiny writes from swinging the decision too much, * writes smaller than 1/8 of avg size are ignored. */ |
2a8149081 writeback: implem... |
220 221 |
#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 */ |
55a694dff writeback, cgroup... |
222 |
#define WB_FRN_TIME_CUT_DIV 8 /* ignore rounds < avg / 8 */ |
2a8149081 writeback: implem... |
223 224 225 226 227 228 229 230 231 |
#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 */ |
6444f47eb writeback, cgroup... |
232 |
#define WB_FRN_MAX_IN_FLIGHT 1024 /* don't queue too many concurrently */ |
2a8149081 writeback: implem... |
233 |
|
a1a0e23e4 writeback: flush ... |
234 235 |
static atomic_t isw_nr_in_flight = ATOMIC_INIT(0); static struct workqueue_struct *isw_wq; |
21c6321fb writeback: reloca... |
236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 |
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); } |
9b0eb69b7 cgroup, blkcg: Pr... |
265 |
EXPORT_SYMBOL_GPL(__inode_attach_wb); |
21c6321fb writeback: reloca... |
266 |
|
703c27088 writeback: implem... |
267 |
/** |
87e1d789b writeback: implem... |
268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 |
* 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 writeback: implem... |
292 |
|
aaa2cacf8 writeback: add lo... |
293 |
/* i_wb may have changed inbetween, can't use inode_to_wb() */ |
614a4e377 writeback, cgroup... |
294 295 296 297 |
if (likely(wb == inode->i_wb)) { wb_put(wb); /* @inode already has ref */ return wb; } |
87e1d789b writeback: implem... |
298 299 |
spin_unlock(&wb->list_lock); |
614a4e377 writeback, cgroup... |
300 |
wb_put(wb); |
87e1d789b writeback: implem... |
301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 |
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 writeback: implem... |
319 320 321 322 323 324 325 |
struct inode_switch_wbs_context { struct inode *inode; struct bdi_writeback *new_wb; struct rcu_head rcu_head; struct work_struct work; }; |
7fc5854f8 writeback: synchr... |
326 327 328 329 330 331 332 333 334 |
static void bdi_down_write_wb_switch_rwsem(struct backing_dev_info *bdi) { down_write(&bdi->wb_switch_rwsem); } static void bdi_up_write_wb_switch_rwsem(struct backing_dev_info *bdi) { up_write(&bdi->wb_switch_rwsem); } |
682aa8e1a writeback: implem... |
335 336 337 338 339 |
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; |
7fc5854f8 writeback: synchr... |
340 |
struct backing_dev_info *bdi = inode_to_bdi(inode); |
d10c80955 writeback: implem... |
341 342 |
struct address_space *mapping = inode->i_mapping; struct bdi_writeback *old_wb = inode->i_wb; |
682aa8e1a writeback: implem... |
343 |
struct bdi_writeback *new_wb = isw->new_wb; |
04edf02cd fs: Convert write... |
344 345 |
XA_STATE(xas, &mapping->i_pages, 0); struct page *page; |
d10c80955 writeback: implem... |
346 |
bool switched = false; |
682aa8e1a writeback: implem... |
347 348 |
/* |
7fc5854f8 writeback: synchr... |
349 350 351 352 353 354 |
* If @inode switches cgwb membership while sync_inodes_sb() is * being issued, sync_inodes_sb() might miss it. Synchronize. */ down_read(&bdi->wb_switch_rwsem); /* |
682aa8e1a writeback: implem... |
355 356 357 |
* 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 |
b93b01631 page cache: use x... |
358 |
* synchronizing against the i_pages lock. |
d10c80955 writeback: implem... |
359 |
* |
b93b01631 page cache: use x... |
360 |
* Grabbing old_wb->list_lock, inode->i_lock and the i_pages lock |
d10c80955 writeback: implem... |
361 362 |
* gives us exclusion against all wb related operations on @inode * including IO list manipulations and stat updates. |
682aa8e1a writeback: implem... |
363 |
*/ |
d10c80955 writeback: implem... |
364 365 366 367 368 369 370 |
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 writeback: implem... |
371 |
spin_lock(&inode->i_lock); |
b93b01631 page cache: use x... |
372 |
xa_lock_irq(&mapping->i_pages); |
d10c80955 writeback: implem... |
373 374 375 |
/* * Once I_FREEING is visible under i_lock, the eviction path owns |
c7f540849 inode: rename i_w... |
376 |
* the inode and we shouldn't modify ->i_io_list. |
d10c80955 writeback: implem... |
377 378 379 |
*/ if (unlikely(inode->i_state & I_FREEING)) goto skip_switch; |
3a8e9ac89 writeback: add tr... |
380 |
trace_inode_switch_wbs(inode, old_wb, new_wb); |
d10c80955 writeback: implem... |
381 382 383 |
/* * Count and transfer stats. Note that PAGECACHE_TAG_DIRTY points * to possibly dirty pages while PAGECACHE_TAG_WRITEBACK points to |
b93b01631 page cache: use x... |
384 |
* pages actually under writeback. |
d10c80955 writeback: implem... |
385 |
*/ |
04edf02cd fs: Convert write... |
386 387 |
xas_for_each_marked(&xas, page, ULONG_MAX, PAGECACHE_TAG_DIRTY) { if (PageDirty(page)) { |
3e8f399da writeback: rework... |
388 389 |
dec_wb_stat(old_wb, WB_RECLAIMABLE); inc_wb_stat(new_wb, WB_RECLAIMABLE); |
d10c80955 writeback: implem... |
390 391 |
} } |
04edf02cd fs: Convert write... |
392 393 394 395 396 |
xas_set(&xas, 0); xas_for_each_marked(&xas, page, ULONG_MAX, PAGECACHE_TAG_WRITEBACK) { WARN_ON_ONCE(!PageWriteback(page)); dec_wb_stat(old_wb, WB_WRITEBACK); inc_wb_stat(new_wb, WB_WRITEBACK); |
d10c80955 writeback: implem... |
397 398 399 400 401 402 403 404 405 406 |
} 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 inode: rename i_w... |
407 |
if (!list_empty(&inode->i_io_list)) { |
d10c80955 writeback: implem... |
408 |
struct inode *pos; |
c7f540849 inode: rename i_w... |
409 |
inode_io_list_del_locked(inode, old_wb); |
d10c80955 writeback: implem... |
410 |
inode->i_wb = new_wb; |
c7f540849 inode: rename i_w... |
411 |
list_for_each_entry(pos, &new_wb->b_dirty, i_io_list) |
d10c80955 writeback: implem... |
412 413 414 |
if (time_after_eq(inode->dirtied_when, pos->dirtied_when)) break; |
c7f540849 inode: rename i_w... |
415 |
inode_io_list_move_locked(inode, new_wb, pos->i_io_list.prev); |
d10c80955 writeback: implem... |
416 417 418 |
} else { inode->i_wb = new_wb; } |
682aa8e1a writeback: implem... |
419 |
|
d10c80955 writeback: implem... |
420 |
/* ->i_wb_frn updates may race wbc_detach_inode() but doesn't matter */ |
682aa8e1a writeback: implem... |
421 422 423 |
inode->i_wb_frn_winner = 0; inode->i_wb_frn_avg_time = 0; inode->i_wb_frn_history = 0; |
d10c80955 writeback: implem... |
424 425 |
switched = true; skip_switch: |
682aa8e1a writeback: implem... |
426 427 428 429 430 |
/* * 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); |
b93b01631 page cache: use x... |
431 |
xa_unlock_irq(&mapping->i_pages); |
682aa8e1a writeback: implem... |
432 |
spin_unlock(&inode->i_lock); |
d10c80955 writeback: implem... |
433 434 |
spin_unlock(&new_wb->list_lock); spin_unlock(&old_wb->list_lock); |
682aa8e1a writeback: implem... |
435 |
|
7fc5854f8 writeback: synchr... |
436 |
up_read(&bdi->wb_switch_rwsem); |
d10c80955 writeback: implem... |
437 438 439 440 |
if (switched) { wb_wakeup(new_wb); wb_put(old_wb); } |
682aa8e1a writeback: implem... |
441 |
wb_put(new_wb); |
d10c80955 writeback: implem... |
442 443 |
iput(inode); |
682aa8e1a writeback: implem... |
444 |
kfree(isw); |
a1a0e23e4 writeback: flush ... |
445 446 |
atomic_dec(&isw_nr_in_flight); |
682aa8e1a writeback: implem... |
447 448 449 450 451 452 453 454 455 |
} 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 writeback: flush ... |
456 |
queue_work(isw_wq, &isw->work); |
682aa8e1a writeback: implem... |
457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 |
} /** * 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; |
6444f47eb writeback, cgroup... |
476 477 |
/* avoid queueing a new switch if too many are already in flight */ if (atomic_read(&isw_nr_in_flight) > WB_FRN_MAX_IN_FLIGHT) |
7fc5854f8 writeback: synchr... |
478 |
return; |
682aa8e1a writeback: implem... |
479 480 |
isw = kzalloc(sizeof(*isw), GFP_ATOMIC); if (!isw) |
6444f47eb writeback, cgroup... |
481 |
return; |
682aa8e1a writeback: implem... |
482 483 484 485 486 487 488 489 490 491 492 493 |
/* 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); |
1751e8a6c Rename superblock... |
494 |
if (!(inode->i_sb->s_flags & SB_ACTIVE) || |
a1a0e23e4 writeback: flush ... |
495 496 497 498 499 |
inode->i_state & (I_WB_SWITCH | I_FREEING) || inode_to_wb(inode) == isw->new_wb) { spin_unlock(&inode->i_lock); goto out_free; } |
682aa8e1a writeback: implem... |
500 |
inode->i_state |= I_WB_SWITCH; |
745249555 writeback: inode ... |
501 |
__iget(inode); |
682aa8e1a writeback: implem... |
502 |
spin_unlock(&inode->i_lock); |
682aa8e1a writeback: implem... |
503 504 505 506 |
isw->inode = inode; /* * In addition to synchronizing among switchers, I_WB_SWITCH tells |
b93b01631 page cache: use x... |
507 508 |
* the RCU protected stat update paths to grab the i_page * lock so that stat transfer can synchronize against them. |
682aa8e1a writeback: implem... |
509 510 511 |
* Let's continue after I_WB_SWITCH is guaranteed to be visible. */ call_rcu(&isw->rcu_head, inode_switch_wbs_rcu_fn); |
ec084de92 fs/writeback.c: u... |
512 513 |
atomic_inc(&isw_nr_in_flight); |
6444f47eb writeback, cgroup... |
514 |
return; |
682aa8e1a writeback: implem... |
515 516 517 518 519 520 |
out_free: if (isw->new_wb) wb_put(isw->new_wb); kfree(isw); } |
87e1d789b writeback: implem... |
521 |
/** |
b16b1deb5 writeback: make w... |
522 523 524 525 526 527 528 529 530 531 532 533 |
* 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 writeback: do for... |
534 535 536 537 |
if (!inode_cgwb_enabled(inode)) { spin_unlock(&inode->i_lock); return; } |
b16b1deb5 writeback: make w... |
538 |
wbc->wb = inode_to_wb(inode); |
2a8149081 writeback: implem... |
539 540 541 542 543 544 545 546 |
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 writeback: make w... |
547 548 |
wb_get(wbc->wb); spin_unlock(&inode->i_lock); |
e8a7abf5a writeback: disass... |
549 550 |
/* |
65de03e25 cgroup,writeback:... |
551 552 553 554 555 |
* A dying wb indicates that either the blkcg associated with the * memcg changed or the associated memcg is dying. In the first * case, a replacement wb should already be available and we should * refresh the wb immediately. In the second case, trying to * refresh will keep failing. |
e8a7abf5a writeback: disass... |
556 |
*/ |
65de03e25 cgroup,writeback:... |
557 |
if (unlikely(wb_dying(wbc->wb) && !css_is_dying(wbc->wb->memcg_css))) |
e8a7abf5a writeback: disass... |
558 |
inode_switch_wbs(inode, wbc->wb_id); |
b16b1deb5 writeback: make w... |
559 |
} |
9b0eb69b7 cgroup, blkcg: Pr... |
560 |
EXPORT_SYMBOL_GPL(wbc_attach_and_unlock_inode); |
b16b1deb5 writeback: make w... |
561 562 |
/** |
2a8149081 writeback: implem... |
563 564 |
* wbc_detach_inode - disassociate wbc from inode and perform foreign detection * @wbc: writeback_control of the just finished writeback |
b16b1deb5 writeback: make w... |
565 566 567 |
* * 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 writeback: implem... |
568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 |
* * 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 writeback: make w... |
598 599 600 |
*/ void wbc_detach_inode(struct writeback_control *wbc) { |
2a8149081 writeback: implem... |
601 602 |
struct bdi_writeback *wb = wbc->wb; struct inode *inode = wbc->inode; |
dd73e4b7d writeback: do for... |
603 604 |
unsigned long avg_time, max_bytes, max_time; u16 history; |
2a8149081 writeback: implem... |
605 |
int max_id; |
dd73e4b7d writeback: do for... |
606 607 608 609 610 |
if (!wb) return; history = inode->i_wb_frn_history; avg_time = inode->i_wb_frn_avg_time; |
2a8149081 writeback: implem... |
611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 |
/* 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; |
3a8e9ac89 writeback: add tr... |
655 656 |
if (history) trace_inode_foreign_history(inode, wbc, history); |
2a8149081 writeback: implem... |
657 658 659 660 661 662 663 |
/* * 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 writeback: implem... |
664 665 |
if (hweight32(history) > WB_FRN_HIST_THR_SLOTS) inode_switch_wbs(inode, max_id); |
2a8149081 writeback: implem... |
666 667 668 669 670 671 672 673 674 |
} /* * 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 writeback: make w... |
675 676 677 |
wb_put(wbc->wb); wbc->wb = NULL; } |
9b0eb69b7 cgroup, blkcg: Pr... |
678 |
EXPORT_SYMBOL_GPL(wbc_detach_inode); |
b16b1deb5 writeback: make w... |
679 680 |
/** |
34e51a5e1 blkcg, writeback:... |
681 |
* wbc_account_cgroup_owner - account writeback to update inode cgroup ownership |
2a8149081 writeback: implem... |
682 683 684 685 686 687 688 689 |
* @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(). */ |
34e51a5e1 blkcg, writeback:... |
690 691 |
void wbc_account_cgroup_owner(struct writeback_control *wbc, struct page *page, size_t bytes) |
2a8149081 writeback: implem... |
692 |
{ |
663114222 blkcg, writeback:... |
693 |
struct cgroup_subsys_state *css; |
2a8149081 writeback: implem... |
694 695 696 697 698 699 700 701 |
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. */ |
27b36d8fa blkcg, writeback:... |
702 |
if (!wbc->wb || wbc->no_cgroup_owner) |
2a8149081 writeback: implem... |
703 |
return; |
663114222 blkcg, writeback:... |
704 705 706 707 708 709 |
css = mem_cgroup_css_from_page(page); /* dead cgroups shouldn't contribute to inode ownership arbitration */ if (!(css->flags & CSS_ONLINE)) return; id = css->id; |
2a8149081 writeback: implem... |
710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 |
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); } |
34e51a5e1 blkcg, writeback:... |
727 |
EXPORT_SYMBOL_GPL(wbc_account_cgroup_owner); |
2a8149081 writeback: implem... |
728 729 |
/** |
703c27088 writeback: implem... |
730 |
* inode_congested - test whether an inode is congested |
60292bcc1 writeback: explai... |
731 |
* @inode: inode to test for congestion (may be NULL) |
703c27088 writeback: implem... |
732 733 734 735 736 737 738 739 740 |
* @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 writeback: explai... |
741 742 743 |
* * @inode is allowed to be NULL as this function is often called on * mapping->host which is NULL for the swapper space. |
703c27088 writeback: implem... |
744 745 746 |
*/ int inode_congested(struct inode *inode, int cong_bits) { |
5cb8b8241 writeback: use un... |
747 748 749 750 |
/* * Once set, ->i_wb never becomes NULL while the inode is alive. * Start transaction iff ->i_wb is visible. */ |
aaa2cacf8 writeback: add lo... |
751 |
if (inode && inode_to_wb_is_valid(inode)) { |
5cb8b8241 writeback: use un... |
752 |
struct bdi_writeback *wb; |
2e898e4c0 writeback: safer ... |
753 754 |
struct wb_lock_cookie lock_cookie = {}; bool congested; |
5cb8b8241 writeback: use un... |
755 |
|
2e898e4c0 writeback: safer ... |
756 |
wb = unlocked_inode_to_wb_begin(inode, &lock_cookie); |
5cb8b8241 writeback: use un... |
757 |
congested = wb_congested(wb, cong_bits); |
2e898e4c0 writeback: safer ... |
758 |
unlocked_inode_to_wb_end(inode, &lock_cookie); |
5cb8b8241 writeback: use un... |
759 |
return congested; |
703c27088 writeback: implem... |
760 761 762 763 764 |
} return wb_congested(&inode_to_bdi(inode)->wb, cong_bits); } EXPORT_SYMBOL_GPL(inode_congested); |
f2b651216 writeback: make w... |
765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 |
/** * 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 writeback: make w... |
792 |
/** |
db1253604 writeback: make w... |
793 794 795 796 797 798 799 800 801 802 803 804 805 806 |
* 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 writeback: bdi_wr... |
807 |
struct bdi_writeback *last_wb = NULL; |
b33e18f61 fs/writeback, rcu... |
808 809 |
struct bdi_writeback *wb = list_entry(&bdi->wb_list, struct bdi_writeback, bdi_node); |
db1253604 writeback: make w... |
810 811 |
might_sleep(); |
db1253604 writeback: make w... |
812 813 |
restart: rcu_read_lock(); |
b817525a4 writeback: bdi_wr... |
814 |
list_for_each_entry_continue_rcu(wb, &bdi->wb_list, bdi_node) { |
5b9cce4c7 writeback: Genera... |
815 |
DEFINE_WB_COMPLETION(fallback_work_done, bdi); |
8a1270cda writeback: remove... |
816 817 818 |
struct wb_writeback_work fallback_work; struct wb_writeback_work *work; long nr_pages; |
b817525a4 writeback: bdi_wr... |
819 820 821 822 |
if (last_wb) { wb_put(last_wb); last_wb = NULL; } |
006a0973e writeback: sync_i... |
823 824 825 826 827 828 |
/* 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 writeback: make w... |
829 |
continue; |
8a1270cda writeback: remove... |
830 831 832 833 834 835 836 837 838 |
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 writeback: make w... |
839 |
} |
8a1270cda writeback: remove... |
840 841 842 843 844 845 846 847 848 |
/* 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 writeback: bdi_wr... |
849 850 851 852 853 854 855 |
/* * 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 writeback: remove... |
856 |
rcu_read_unlock(); |
5b9cce4c7 writeback: Genera... |
857 |
wb_wait_for_completion(&fallback_work_done); |
8a1270cda writeback: remove... |
858 |
goto restart; |
db1253604 writeback: make w... |
859 860 |
} rcu_read_unlock(); |
b817525a4 writeback: bdi_wr... |
861 862 863 |
if (last_wb) wb_put(last_wb); |
db1253604 writeback: make w... |
864 |
} |
a1a0e23e4 writeback: flush ... |
865 |
/** |
d62241c7a writeback, memcg:... |
866 867 868 |
* cgroup_writeback_by_id - initiate cgroup writeback from bdi and memcg IDs * @bdi_id: target bdi id * @memcg_id: target memcg css id |
b46ec1da5 fs/fs-writeback.c... |
869 |
* @nr: number of pages to write, 0 for best-effort dirty flushing |
d62241c7a writeback, memcg:... |
870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 |
* @reason: reason why some writeback work initiated * @done: target wb_completion * * Initiate flush of the bdi_writeback identified by @bdi_id and @memcg_id * with the specified parameters. */ int cgroup_writeback_by_id(u64 bdi_id, int memcg_id, unsigned long nr, enum wb_reason reason, struct wb_completion *done) { struct backing_dev_info *bdi; struct cgroup_subsys_state *memcg_css; struct bdi_writeback *wb; struct wb_writeback_work *work; int ret; /* lookup bdi and memcg */ bdi = bdi_get_by_id(bdi_id); if (!bdi) return -ENOENT; rcu_read_lock(); memcg_css = css_from_id(memcg_id, &memory_cgrp_subsys); if (memcg_css && !css_tryget(memcg_css)) memcg_css = NULL; rcu_read_unlock(); if (!memcg_css) { ret = -ENOENT; goto out_bdi_put; } /* * And find the associated wb. If the wb isn't there already * there's nothing to flush, don't create one. */ wb = wb_get_lookup(bdi, memcg_css); if (!wb) { ret = -ENOENT; goto out_css_put; } /* * If @nr is zero, the caller is attempting to write out most of * the currently dirty pages. Let's take the current dirty page * count and inflate it by 25% which should be large enough to * flush out most dirty pages while avoiding getting livelocked by * concurrent dirtiers. */ if (!nr) { unsigned long filepages, headroom, dirty, writeback; mem_cgroup_wb_stats(wb, &filepages, &headroom, &dirty, &writeback); nr = dirty * 10 / 8; } /* issue the writeback work */ work = kzalloc(sizeof(*work), GFP_NOWAIT | __GFP_NOWARN); if (work) { work->nr_pages = nr; work->sync_mode = WB_SYNC_NONE; work->range_cyclic = 1; work->reason = reason; work->done = done; work->auto_free = 1; wb_queue_work(wb, work); ret = 0; } else { ret = -ENOMEM; } wb_put(wb); out_css_put: css_put(memcg_css); out_bdi_put: bdi_put(bdi); return ret; } /** |
a1a0e23e4 writeback: flush ... |
949 950 951 952 953 954 955 956 957 958 959 960 |
* 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)) { |
ec084de92 fs/writeback.c: u... |
961 962 963 964 965 |
/* * Use rcu_barrier() to wait for all pending callbacks to * ensure that all in-flight wb switches are in the workqueue. */ rcu_barrier(); |
a1a0e23e4 writeback: flush ... |
966 967 968 969 970 971 972 973 974 975 976 977 |
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 writeback: make w... |
978 |
#else /* CONFIG_CGROUP_WRITEBACK */ |
7fc5854f8 writeback: synchr... |
979 980 |
static void bdi_down_write_wb_switch_rwsem(struct backing_dev_info *bdi) { } static void bdi_up_write_wb_switch_rwsem(struct backing_dev_info *bdi) { } |
87e1d789b writeback: implem... |
981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 |
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 writeback: make w... |
1001 1002 1003 1004 |
static long wb_split_bdi_pages(struct bdi_writeback *wb, long nr_pages) { return nr_pages; } |
db1253604 writeback: make w... |
1005 1006 1007 1008 1009 |
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 writeback: sync_i... |
1010 |
if (!skip_if_busy || !writeback_in_progress(&bdi->wb)) { |
db1253604 writeback: make w... |
1011 |
base_work->auto_free = 0; |
db1253604 writeback: make w... |
1012 1013 1014 |
wb_queue_work(&bdi->wb, base_work); } } |
703c27088 writeback: implem... |
1015 |
#endif /* CONFIG_CGROUP_WRITEBACK */ |
e8e8a0c6c writeback: move n... |
1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 |
/* * Add in the number of potentially dirty inodes, because each inode * write can dirty pagecache in the underlying blockdev. */ static unsigned long get_nr_dirty_pages(void) { return global_node_page_state(NR_FILE_DIRTY) + global_node_page_state(NR_UNSTABLE_NFS) + get_nr_dirty_inodes(); } static void wb_start_writeback(struct bdi_writeback *wb, enum wb_reason reason) |
b6e51316d writeback: separa... |
1028 |
{ |
c00ddad39 writeback: remove... |
1029 1030 1031 1032 |
if (!wb_has_dirty_io(wb)) return; /* |
aac8d41cd writeback: only a... |
1033 1034 1035 1036 1037 |
* All callers of this function want to start writeback of all * dirty pages. Places like vmscan can call this at a very * high frequency, causing pointless allocations of tons of * work items and keeping the flusher threads busy retrieving * that work. Ensure that we only allow one of them pending and |
85009b4f5 writeback: elimin... |
1038 |
* inflight at the time. |
aac8d41cd writeback: only a... |
1039 |
*/ |
85009b4f5 writeback: elimin... |
1040 1041 |
if (test_bit(WB_start_all, &wb->state) || test_and_set_bit(WB_start_all, &wb->state)) |
aac8d41cd writeback: only a... |
1042 |
return; |
85009b4f5 writeback: elimin... |
1043 1044 |
wb->start_all_reason = reason; wb_wakeup(wb); |
c5444198c writeback: simpli... |
1045 |
} |
d3ddec763 writeback: stop b... |
1046 |
|
c5444198c writeback: simpli... |
1047 |
/** |
9ecf4866c writeback: make b... |
1048 1049 |
* wb_start_background_writeback - start background writeback * @wb: bdi_writback to write from |
c5444198c writeback: simpli... |
1050 1051 |
* * Description: |
6585027a5 writeback: integr... |
1052 |
* This makes sure WB_SYNC_NONE background writeback happens. When |
9ecf4866c writeback: make b... |
1053 |
* this function returns, it is only guaranteed that for given wb |
6585027a5 writeback: integr... |
1054 1055 |
* some IO is happening if we are over background dirty threshold. * Caller need not hold sb s_umount semaphore. |
c5444198c writeback: simpli... |
1056 |
*/ |
9ecf4866c writeback: make b... |
1057 |
void wb_start_background_writeback(struct bdi_writeback *wb) |
c5444198c writeback: simpli... |
1058 |
{ |
6585027a5 writeback: integr... |
1059 1060 1061 1062 |
/* * We just wake up the flusher thread. It will perform background * writeback as soon as there is no other work to do. */ |
5634cc2aa writeback: update... |
1063 |
trace_writeback_wake_background(wb); |
9ecf4866c writeback: make b... |
1064 |
wb_wakeup(wb); |
1da177e4c Linux-2.6.12-rc2 |
1065 1066 1067 |
} /* |
a66979aba fs: move i_wb_lis... |
1068 1069 |
* Remove the inode from the writeback list it is on. */ |
c7f540849 inode: rename i_w... |
1070 |
void inode_io_list_del(struct inode *inode) |
a66979aba fs: move i_wb_lis... |
1071 |
{ |
87e1d789b writeback: implem... |
1072 |
struct bdi_writeback *wb; |
f758eeabe writeback: split ... |
1073 |
|
87e1d789b writeback: implem... |
1074 |
wb = inode_to_wb_and_lock_list(inode); |
8eab2b531 writeback: Protec... |
1075 |
spin_lock(&inode->i_lock); |
c7f540849 inode: rename i_w... |
1076 |
inode_io_list_del_locked(inode, wb); |
8eab2b531 writeback: Protec... |
1077 |
spin_unlock(&inode->i_lock); |
52ebea749 writeback: make b... |
1078 |
spin_unlock(&wb->list_lock); |
a66979aba fs: move i_wb_lis... |
1079 |
} |
a66979aba fs: move i_wb_lis... |
1080 |
/* |
6c60d2b57 fs/fs-writeback.c... |
1081 1082 1083 1084 1085 1086 1087 1088 1089 |
* 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 fs/fs-writeback.c... |
1090 |
if (list_empty(&inode->i_wb_list)) { |
6c60d2b57 fs/fs-writeback.c... |
1091 |
list_add_tail(&inode->i_wb_list, &sb->s_inodes_wb); |
9a46b04f1 fs/fs-writeback.c... |
1092 1093 |
trace_sb_mark_inode_writeback(inode); } |
6c60d2b57 fs/fs-writeback.c... |
1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 |
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 fs/fs-writeback.c... |
1108 1109 1110 1111 |
if (!list_empty(&inode->i_wb_list)) { list_del_init(&inode->i_wb_list); trace_sb_clear_inode_writeback(inode); } |
6c60d2b57 fs/fs-writeback.c... |
1112 1113 1114 1115 1116 |
spin_unlock_irqrestore(&sb->s_inode_wblist_lock, flags); } } /* |
6610a0bc8 writeback: fix ti... |
1117 1118 1119 1120 |
* 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 writeback: move d... |
1121 |
* already the most-recently-dirtied inode on the b_dirty list. If that is |
6610a0bc8 writeback: fix ti... |
1122 1123 1124 |
* the case then the inode must have been redirtied while it was being written * out and we don't reset its dirtied_when. */ |
8eab2b531 writeback: Protec... |
1125 |
static void redirty_tail_locked(struct inode *inode, struct bdi_writeback *wb) |
6610a0bc8 writeback: fix ti... |
1126 |
{ |
8eab2b531 writeback: Protec... |
1127 |
assert_spin_locked(&inode->i_lock); |
03ba3782e writeback: switch... |
1128 |
if (!list_empty(&wb->b_dirty)) { |
66f3b8e2e writeback: move d... |
1129 |
struct inode *tail; |
6610a0bc8 writeback: fix ti... |
1130 |
|
7ccf19a80 fs: inode split I... |
1131 |
tail = wb_inode(wb->b_dirty.next); |
66f3b8e2e writeback: move d... |
1132 |
if (time_before(inode->dirtied_when, tail->dirtied_when)) |
6610a0bc8 writeback: fix ti... |
1133 1134 |
inode->dirtied_when = jiffies; } |
c7f540849 inode: rename i_w... |
1135 |
inode_io_list_move_locked(inode, wb, &wb->b_dirty); |
cb0c74450 writeback: Avoid ... |
1136 |
inode->i_state &= ~I_SYNC_QUEUED; |
6610a0bc8 writeback: fix ti... |
1137 |
} |
8eab2b531 writeback: Protec... |
1138 1139 1140 1141 1142 1143 |
static void redirty_tail(struct inode *inode, struct bdi_writeback *wb) { spin_lock(&inode->i_lock); redirty_tail_locked(inode, wb); spin_unlock(&inode->i_lock); } |
6610a0bc8 writeback: fix ti... |
1144 |
/* |
66f3b8e2e writeback: move d... |
1145 |
* requeue inode for re-scanning after bdi->b_io list is exhausted. |
c986d1e2a writeback: fix ti... |
1146 |
*/ |
f758eeabe writeback: split ... |
1147 |
static void requeue_io(struct inode *inode, struct bdi_writeback *wb) |
c986d1e2a writeback: fix ti... |
1148 |
{ |
c7f540849 inode: rename i_w... |
1149 |
inode_io_list_move_locked(inode, wb, &wb->b_more_io); |
c986d1e2a writeback: fix ti... |
1150 |
} |
1c0eeaf56 introduce I_SYNC |
1151 1152 |
static void inode_sync_complete(struct inode *inode) { |
365b94ae6 writeback: Move c... |
1153 |
inode->i_state &= ~I_SYNC; |
4eff96dd5 writeback: put un... |
1154 1155 |
/* If inode is clean an unused, put it into LRU now... */ inode_add_lru(inode); |
365b94ae6 writeback: Move c... |
1156 |
/* Waiters must see I_SYNC cleared before being woken up */ |
1c0eeaf56 introduce I_SYNC |
1157 1158 1159 |
smp_mb(); wake_up_bit(&inode->i_state, __I_SYNC); } |
d2caa3c54 writeback: guard ... |
1160 1161 1162 1163 1164 1165 1166 1167 |
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 writeback: get ri... |
1168 |
* from permanently stopping the whole bdi writeback. |
d2caa3c54 writeback: guard ... |
1169 1170 1171 1172 1173 |
*/ ret = ret && time_before_eq(inode->dirtied_when, jiffies); #endif return ret; } |
0ae45f63d vfs: add support ... |
1174 |
#define EXPIRE_DIRTY_ATIME 0x0001 |
c986d1e2a writeback: fix ti... |
1175 |
/* |
6623c1904 writeback: Fix sy... |
1176 |
* Move expired (dirtied before dirtied_before) dirty inodes from |
697e6fed9 writeback: Remove... |
1177 |
* @delaying_queue to @dispatch_queue. |
2c1365791 writeback: fix ti... |
1178 |
*/ |
e84d0a4f8 writeback: trace ... |
1179 |
static int move_expired_inodes(struct list_head *delaying_queue, |
2c1365791 writeback: fix ti... |
1180 |
struct list_head *dispatch_queue, |
6623c1904 writeback: Fix sy... |
1181 |
int flags, unsigned long dirtied_before) |
2c1365791 writeback: fix ti... |
1182 |
{ |
5c03449d3 writeback: move i... |
1183 1184 |
LIST_HEAD(tmp); struct list_head *pos, *node; |
cf137307c writeback: don't ... |
1185 |
struct super_block *sb = NULL; |
5c03449d3 writeback: move i... |
1186 |
struct inode *inode; |
cf137307c writeback: don't ... |
1187 |
int do_sb_sort = 0; |
e84d0a4f8 writeback: trace ... |
1188 |
int moved = 0; |
5c03449d3 writeback: move i... |
1189 |
|
2c1365791 writeback: fix ti... |
1190 |
while (!list_empty(delaying_queue)) { |
7ccf19a80 fs: inode split I... |
1191 |
inode = wb_inode(delaying_queue->prev); |
6623c1904 writeback: Fix sy... |
1192 |
if (inode_dirtied_after(inode, dirtied_before)) |
2c1365791 writeback: fix ti... |
1193 |
break; |
c7f540849 inode: rename i_w... |
1194 |
list_move(&inode->i_io_list, &tmp); |
a8855990e writeback: Do not... |
1195 |
moved++; |
cb0c74450 writeback: Avoid ... |
1196 |
spin_lock(&inode->i_lock); |
0ae45f63d vfs: add support ... |
1197 |
if (flags & EXPIRE_DIRTY_ATIME) |
cb0c74450 writeback: Avoid ... |
1198 1199 1200 |
inode->i_state |= I_DIRTY_TIME_EXPIRED; inode->i_state |= I_SYNC_QUEUED; spin_unlock(&inode->i_lock); |
a8855990e writeback: Do not... |
1201 1202 |
if (sb_is_blkdev_sb(inode->i_sb)) continue; |
cf137307c writeback: don't ... |
1203 1204 1205 |
if (sb && sb != inode->i_sb) do_sb_sort = 1; sb = inode->i_sb; |
5c03449d3 writeback: move i... |
1206 |
} |
cf137307c writeback: don't ... |
1207 1208 1209 |
/* just one sb in list, splice to dispatch_queue and we're done */ if (!do_sb_sort) { list_splice(&tmp, dispatch_queue); |
e84d0a4f8 writeback: trace ... |
1210 |
goto out; |
cf137307c writeback: don't ... |
1211 |
} |
5c03449d3 writeback: move i... |
1212 1213 |
/* Move inodes from one superblock together */ while (!list_empty(&tmp)) { |
7ccf19a80 fs: inode split I... |
1214 |
sb = wb_inode(tmp.prev)->i_sb; |
5c03449d3 writeback: move i... |
1215 |
list_for_each_prev_safe(pos, node, &tmp) { |
7ccf19a80 fs: inode split I... |
1216 |
inode = wb_inode(pos); |
5c03449d3 writeback: move i... |
1217 |
if (inode->i_sb == sb) |
c7f540849 inode: rename i_w... |
1218 |
list_move(&inode->i_io_list, dispatch_queue); |
5c03449d3 writeback: move i... |
1219 |
} |
2c1365791 writeback: fix ti... |
1220 |
} |
e84d0a4f8 writeback: trace ... |
1221 1222 |
out: return moved; |
2c1365791 writeback: fix ti... |
1223 1224 1225 1226 |
} /* * Queue all expired dirty inodes for io, eldest first. |
4ea879b96 writeback: fix qu... |
1227 1228 1229 1230 1231 1232 1233 1234 |
* 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 writeback: fix ti... |
1235 |
*/ |
6623c1904 writeback: Fix sy... |
1236 1237 |
static void queue_io(struct bdi_writeback *wb, struct wb_writeback_work *work, unsigned long dirtied_before) |
66f3b8e2e writeback: move d... |
1238 |
{ |
e84d0a4f8 writeback: trace ... |
1239 |
int moved; |
6623c1904 writeback: Fix sy... |
1240 |
unsigned long time_expire_jif = dirtied_before; |
0ae45f63d vfs: add support ... |
1241 |
|
f758eeabe writeback: split ... |
1242 |
assert_spin_locked(&wb->list_lock); |
4ea879b96 writeback: fix qu... |
1243 |
list_splice_init(&wb->b_more_io, &wb->b_io); |
6623c1904 writeback: Fix sy... |
1244 1245 1246 |
moved = move_expired_inodes(&wb->b_dirty, &wb->b_io, 0, dirtied_before); if (!work->for_sync) time_expire_jif = jiffies - dirtytime_expire_interval * HZ; |
0ae45f63d vfs: add support ... |
1247 |
moved += move_expired_inodes(&wb->b_dirty_time, &wb->b_io, |
6623c1904 writeback: Fix sy... |
1248 |
EXPIRE_DIRTY_ATIME, time_expire_jif); |
d6c10f1fc writeback: implem... |
1249 1250 |
if (moved) wb_io_lists_populated(wb); |
6623c1904 writeback: Fix sy... |
1251 |
trace_writeback_queue_io(wb, work, dirtied_before, moved); |
66f3b8e2e writeback: move d... |
1252 |
} |
a9185b41a pass writeback_co... |
1253 |
static int write_inode(struct inode *inode, struct writeback_control *wbc) |
08d8e9749 writeback: fix nt... |
1254 |
{ |
9fb0a7da0 writeback: add mo... |
1255 1256 1257 1258 1259 1260 1261 1262 |
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 writeback: switch... |
1263 |
return 0; |
08d8e9749 writeback: fix nt... |
1264 |
} |
08d8e9749 writeback: fix nt... |
1265 |
|
2c1365791 writeback: fix ti... |
1266 |
/* |
169ebd901 writeback: Avoid ... |
1267 1268 |
* 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 cleanup __writeba... |
1269 |
*/ |
169ebd901 writeback: Avoid ... |
1270 1271 1272 |
static void __inode_wait_for_writeback(struct inode *inode) __releases(inode->i_lock) __acquires(inode->i_lock) |
01c031945 cleanup __writeba... |
1273 1274 1275 1276 1277 |
{ DEFINE_WAIT_BIT(wq, &inode->i_state, __I_SYNC); wait_queue_head_t *wqh; wqh = bit_waitqueue(&inode->i_state, __I_SYNC); |
250df6ed2 fs: protect inode... |
1278 1279 |
while (inode->i_state & I_SYNC) { spin_unlock(&inode->i_lock); |
743162013 sched: Remove pro... |
1280 1281 |
__wait_on_bit(wqh, &wq, bit_wait, TASK_UNINTERRUPTIBLE); |
250df6ed2 fs: protect inode... |
1282 |
spin_lock(&inode->i_lock); |
58a9d3d8d fs-writeback: che... |
1283 |
} |
01c031945 cleanup __writeba... |
1284 1285 1286 |
} /* |
169ebd901 writeback: Avoid ... |
1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 |
* 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 writeback: Separa... |
1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 |
* 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 writeback: Refact... |
1338 1339 1340 1341 1342 |
if (wbc->pages_skipped) { /* * writeback is not making progress due to locked * buffers. Skip this inode for now. */ |
8eab2b531 writeback: Protec... |
1343 |
redirty_tail_locked(inode, wb); |
4f8ad655d writeback: Refact... |
1344 1345 |
return; } |
ccb26b5a6 writeback: Separa... |
1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 |
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. */ |
8eab2b531 writeback: Protec... |
1362 |
redirty_tail_locked(inode, wb); |
ccb26b5a6 writeback: Separa... |
1363 1364 1365 1366 1367 1368 1369 |
} } 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. */ |
8eab2b531 writeback: Protec... |
1370 |
redirty_tail_locked(inode, wb); |
0ae45f63d vfs: add support ... |
1371 |
} else if (inode->i_state & I_DIRTY_TIME) { |
a2f487069 fs: make sure the... |
1372 |
inode->dirtied_when = jiffies; |
c7f540849 inode: rename i_w... |
1373 |
inode_io_list_move_locked(inode, wb, &wb->b_dirty_time); |
cb0c74450 writeback: Avoid ... |
1374 |
inode->i_state &= ~I_SYNC_QUEUED; |
ccb26b5a6 writeback: Separa... |
1375 1376 |
} else { /* The inode is clean. Remove from writeback lists. */ |
c7f540849 inode: rename i_w... |
1377 |
inode_io_list_del_locked(inode, wb); |
ccb26b5a6 writeback: Separa... |
1378 1379 1380 1381 |
} } /* |
4f8ad655d writeback: Refact... |
1382 1383 1384 |
* 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 Linux-2.6.12-rc2 |
1385 1386 |
*/ static int |
cd8ed2a45 fs/fs-writeback.c... |
1387 |
__writeback_single_inode(struct inode *inode, struct writeback_control *wbc) |
1da177e4c Linux-2.6.12-rc2 |
1388 |
{ |
1da177e4c Linux-2.6.12-rc2 |
1389 |
struct address_space *mapping = inode->i_mapping; |
251d6a471 writeback: trace ... |
1390 |
long nr_to_write = wbc->nr_to_write; |
01c031945 cleanup __writeba... |
1391 |
unsigned dirty; |
1da177e4c Linux-2.6.12-rc2 |
1392 |
int ret; |
4f8ad655d writeback: Refact... |
1393 |
WARN_ON(!(inode->i_state & I_SYNC)); |
1da177e4c Linux-2.6.12-rc2 |
1394 |
|
9fb0a7da0 writeback: add mo... |
1395 |
trace_writeback_single_inode_start(inode, wbc, nr_to_write); |
1da177e4c Linux-2.6.12-rc2 |
1396 |
ret = do_writepages(mapping, wbc); |
26821ed40 make sure data is... |
1397 1398 1399 |
/* * Make sure to wait on the data before writing out the metadata. * This is important for filesystems that modify metadata on data |
7747bd4bc sync: don't block... |
1400 1401 1402 |
* 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 make sure data is... |
1403 |
*/ |
7747bd4bc sync: don't block... |
1404 |
if (wbc->sync_mode == WB_SYNC_ALL && !wbc->for_sync) { |
26821ed40 make sure data is... |
1405 |
int err = filemap_fdatawait(mapping); |
1da177e4c Linux-2.6.12-rc2 |
1406 1407 1408 |
if (ret == 0) ret = err; } |
5547e8aac writeback: Update... |
1409 1410 1411 1412 1413 |
/* * Some filesystems may redirty the inode during the writeback * due to delalloc, clear dirty metadata flags right before * write_inode() */ |
250df6ed2 fs: protect inode... |
1414 |
spin_lock(&inode->i_lock); |
9c6ac78eb writeback: fix a ... |
1415 |
|
5547e8aac writeback: Update... |
1416 |
dirty = inode->i_state & I_DIRTY; |
a2f487069 fs: make sure the... |
1417 |
if (inode->i_state & I_DIRTY_TIME) { |
0e11f6443 fs: move I_DIRTY_... |
1418 |
if ((dirty & I_DIRTY_INODE) || |
dc5ff2b1d writeback: Write ... |
1419 |
wbc->sync_mode == WB_SYNC_ALL || |
a2f487069 fs: make sure the... |
1420 1421 1422 1423 1424 1425 1426 1427 1428 |
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 vfs: add support ... |
1429 |
inode->i_state &= ~dirty; |
9c6ac78eb writeback: fix a ... |
1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 |
/* * 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 fs: protect inode... |
1446 |
spin_unlock(&inode->i_lock); |
9c6ac78eb writeback: fix a ... |
1447 |
|
0ae45f63d vfs: add support ... |
1448 1449 |
if (dirty & I_DIRTY_TIME) mark_inode_dirty_sync(inode); |
26821ed40 make sure data is... |
1450 |
/* Don't write the inode if only I_DIRTY_PAGES was set */ |
0ae45f63d vfs: add support ... |
1451 |
if (dirty & ~I_DIRTY_PAGES) { |
a9185b41a pass writeback_co... |
1452 |
int err = write_inode(inode, wbc); |
1da177e4c Linux-2.6.12-rc2 |
1453 1454 1455 |
if (ret == 0) ret = err; } |
4f8ad655d writeback: Refact... |
1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 |
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 writeback, cgroup... |
1468 1469 |
static int writeback_single_inode(struct inode *inode, struct writeback_control *wbc) |
4f8ad655d writeback: Refact... |
1470 |
{ |
aaf255933 writeback, cgroup... |
1471 |
struct bdi_writeback *wb; |
4f8ad655d writeback: Refact... |
1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 |
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 writeback: Avoid ... |
1484 1485 1486 |
* 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 writeback: Refact... |
1487 |
*/ |
169ebd901 writeback: Avoid ... |
1488 |
__inode_wait_for_writeback(inode); |
4f8ad655d writeback: Refact... |
1489 1490 1491 |
} WARN_ON(inode->i_state & I_SYNC); /* |
f9b0e058c writeback: Fix da... |
1492 1493 1494 1495 1496 1497 |
* 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 writeback: Refact... |
1498 |
*/ |
0ae45f63d vfs: add support ... |
1499 |
if (!(inode->i_state & I_DIRTY_ALL) && |
f9b0e058c writeback: Fix da... |
1500 1501 |
(wbc->sync_mode != WB_SYNC_ALL || !mapping_tagged(inode->i_mapping, PAGECACHE_TAG_WRITEBACK))) |
4f8ad655d writeback: Refact... |
1502 1503 |
goto out; inode->i_state |= I_SYNC; |
b16b1deb5 writeback: make w... |
1504 |
wbc_attach_and_unlock_inode(wbc, inode); |
4f8ad655d writeback: Refact... |
1505 |
|
cd8ed2a45 fs/fs-writeback.c... |
1506 |
ret = __writeback_single_inode(inode, wbc); |
1da177e4c Linux-2.6.12-rc2 |
1507 |
|
b16b1deb5 writeback: make w... |
1508 |
wbc_detach_inode(wbc); |
aaf255933 writeback, cgroup... |
1509 1510 |
wb = inode_to_wb_and_lock_list(inode); |
250df6ed2 fs: protect inode... |
1511 |
spin_lock(&inode->i_lock); |
4f8ad655d writeback: Refact... |
1512 1513 1514 1515 |
/* * If inode is clean, remove it from writeback lists. Otherwise don't * touch it. See comment above for explanation. */ |
0ae45f63d vfs: add support ... |
1516 |
if (!(inode->i_state & I_DIRTY_ALL)) |
c7f540849 inode: rename i_w... |
1517 |
inode_io_list_del_locked(inode, wb); |
4f8ad655d writeback: Refact... |
1518 |
spin_unlock(&wb->list_lock); |
1c0eeaf56 introduce I_SYNC |
1519 |
inode_sync_complete(inode); |
4f8ad655d writeback: Refact... |
1520 1521 |
out: spin_unlock(&inode->i_lock); |
1da177e4c Linux-2.6.12-rc2 |
1522 1523 |
return ret; } |
a88a341a7 writeback: move b... |
1524 |
static long writeback_chunk_size(struct bdi_writeback *wb, |
1a12d8bd7 writeback: scale ... |
1525 |
struct wb_writeback_work *work) |
d46db3d58 writeback: make w... |
1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 |
{ 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 writeback: scale ... |
1544 |
else { |
a88a341a7 writeback: move b... |
1545 |
pages = min(wb->avg_write_bandwidth / 2, |
dcc25ae76 writeback: move g... |
1546 |
global_wb_domain.dirty_limit / DIRTY_SCOPE); |
1a12d8bd7 writeback: scale ... |
1547 1548 1549 1550 |
pages = min(pages, work->nr_pages); pages = round_down(pages + MIN_WRITEBACK_PAGES, MIN_WRITEBACK_PAGES); } |
d46db3d58 writeback: make w... |
1551 1552 1553 |
return pages; } |
03ba3782e writeback: switch... |
1554 |
/* |
f11c9c5c2 vfs: improve writ... |
1555 |
* Write a portion of b_io inodes which belong to @sb. |
edadfb10b writeback: split ... |
1556 |
* |
d46db3d58 writeback: make w... |
1557 |
* Return the number of pages and/or inodes written. |
0ba13fd19 Revert "writeback... |
1558 1559 1560 1561 |
* * 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 vfs: improve writ... |
1562 |
*/ |
d46db3d58 writeback: make w... |
1563 1564 1565 |
static long writeback_sb_inodes(struct super_block *sb, struct bdi_writeback *wb, struct wb_writeback_work *work) |
1da177e4c Linux-2.6.12-rc2 |
1566 |
{ |
d46db3d58 writeback: make w... |
1567 1568 1569 1570 1571 |
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 sync: don't block... |
1572 |
.for_sync = work->for_sync, |
d46db3d58 writeback: make w... |
1573 1574 1575 1576 1577 1578 1579 |
.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 writeback: switch... |
1580 |
while (!list_empty(&wb->b_io)) { |
7ccf19a80 fs: inode split I... |
1581 |
struct inode *inode = wb_inode(wb->b_io.prev); |
aaf255933 writeback, cgroup... |
1582 |
struct bdi_writeback *tmp_wb; |
edadfb10b writeback: split ... |
1583 1584 |
if (inode->i_sb != sb) { |
d46db3d58 writeback: make w... |
1585 |
if (work->sb) { |
edadfb10b writeback: split ... |
1586 1587 1588 1589 1590 |
/* * We only want to write back data for this * superblock, move all inodes not belonging * to it back onto the dirty list. */ |
f758eeabe writeback: split ... |
1591 |
redirty_tail(inode, wb); |
edadfb10b writeback: split ... |
1592 1593 1594 1595 1596 1597 1598 1599 |
continue; } /* * The inode belongs to a different superblock. * Bounce back to the caller to unpin this and * pin the next superblock. */ |
d46db3d58 writeback: make w... |
1600 |
break; |
edadfb10b writeback: split ... |
1601 |
} |
9843b76aa fs: skip I_FREEIN... |
1602 |
/* |
331cbdeed writeback: Fix so... |
1603 1604 |
* Don't bother with new inodes or inodes being freed, first * kind does not need periodic writeout yet, and for the latter |
9843b76aa fs: skip I_FREEIN... |
1605 1606 |
* kind writeout is handled by the freer. */ |
250df6ed2 fs: protect inode... |
1607 |
spin_lock(&inode->i_lock); |
9843b76aa fs: skip I_FREEIN... |
1608 |
if (inode->i_state & (I_NEW | I_FREEING | I_WILL_FREE)) { |
8eab2b531 writeback: Protec... |
1609 |
redirty_tail_locked(inode, wb); |
250df6ed2 fs: protect inode... |
1610 |
spin_unlock(&inode->i_lock); |
7ef0d7377 fs: new inode i_s... |
1611 1612 |
continue; } |
cc1676d91 writeback: Move r... |
1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 |
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 writeback: Remove... |
1628 |
spin_unlock(&wb->list_lock); |
4f8ad655d writeback: Refact... |
1629 1630 1631 1632 1633 |
/* * 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 writeback: Avoid ... |
1634 1635 1636 1637 |
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 writeback: Fix lo... |
1638 |
spin_lock(&wb->list_lock); |
169ebd901 writeback: Avoid ... |
1639 1640 |
continue; } |
4f8ad655d writeback: Refact... |
1641 |
inode->i_state |= I_SYNC; |
b16b1deb5 writeback: make w... |
1642 |
wbc_attach_and_unlock_inode(&wbc, inode); |
169ebd901 writeback: Avoid ... |
1643 |
|
a88a341a7 writeback: move b... |
1644 |
write_chunk = writeback_chunk_size(wb, work); |
d46db3d58 writeback: make w... |
1645 1646 |
wbc.nr_to_write = write_chunk; wbc.pages_skipped = 0; |
250df6ed2 fs: protect inode... |
1647 |
|
169ebd901 writeback: Avoid ... |
1648 1649 1650 1651 |
/* * 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 fs/fs-writeback.c... |
1652 |
__writeback_single_inode(inode, &wbc); |
250df6ed2 fs: protect inode... |
1653 |
|
b16b1deb5 writeback: make w... |
1654 |
wbc_detach_inode(&wbc); |
d46db3d58 writeback: make w... |
1655 1656 |
work->nr_pages -= write_chunk - wbc.nr_to_write; wrote += write_chunk - wbc.nr_to_write; |
590dca3a7 fs-writeback: unp... |
1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 |
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 writeback, cgroup... |
1670 1671 1672 1673 1674 |
/* * 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 writeback: Refact... |
1675 |
spin_lock(&inode->i_lock); |
0ae45f63d vfs: add support ... |
1676 |
if (!(inode->i_state & I_DIRTY_ALL)) |
d46db3d58 writeback: make w... |
1677 |
wrote++; |
aaf255933 writeback, cgroup... |
1678 |
requeue_inode(inode, tmp_wb, &wbc); |
4f8ad655d writeback: Refact... |
1679 |
inode_sync_complete(inode); |
0f1b1fd86 fs: pull inode->i... |
1680 |
spin_unlock(&inode->i_lock); |
590dca3a7 fs-writeback: unp... |
1681 |
|
aaf255933 writeback, cgroup... |
1682 1683 1684 1685 |
if (unlikely(tmp_wb != wb)) { spin_unlock(&tmp_wb->list_lock); spin_lock(&wb->list_lock); } |
d46db3d58 writeback: make w... |
1686 1687 1688 1689 1690 1691 1692 1693 1694 |
/* * 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 writeback: speed ... |
1695 |
} |
1da177e4c Linux-2.6.12-rc2 |
1696 |
} |
d46db3d58 writeback: make w... |
1697 |
return wrote; |
f11c9c5c2 vfs: improve writ... |
1698 |
} |
d46db3d58 writeback: make w... |
1699 1700 |
static long __writeback_inodes_wb(struct bdi_writeback *wb, struct wb_writeback_work *work) |
f11c9c5c2 vfs: improve writ... |
1701 |
{ |
d46db3d58 writeback: make w... |
1702 1703 |
unsigned long start_time = jiffies; long wrote = 0; |
38f219776 fs: sync_sb_inode... |
1704 |
|
f11c9c5c2 vfs: improve writ... |
1705 |
while (!list_empty(&wb->b_io)) { |
7ccf19a80 fs: inode split I... |
1706 |
struct inode *inode = wb_inode(wb->b_io.prev); |
f11c9c5c2 vfs: improve writ... |
1707 |
struct super_block *sb = inode->i_sb; |
9ecc2738a writeback: make t... |
1708 |
|
eb6ef3df4 trylock_super(): ... |
1709 |
if (!trylock_super(sb)) { |
0e995816f don't busy retry ... |
1710 |
/* |
eb6ef3df4 trylock_super(): ... |
1711 |
* trylock_super() may fail consistently due to |
0e995816f don't busy retry ... |
1712 1713 1714 1715 |
* s_umount being grabbed by someone else. Don't use * requeue_io() to avoid busy retrying the inode/sb. */ redirty_tail(inode, wb); |
edadfb10b writeback: split ... |
1716 |
continue; |
f11c9c5c2 vfs: improve writ... |
1717 |
} |
d46db3d58 writeback: make w... |
1718 |
wrote += writeback_sb_inodes(sb, wb, work); |
eb6ef3df4 trylock_super(): ... |
1719 |
up_read(&sb->s_umount); |
f11c9c5c2 vfs: improve writ... |
1720 |
|
d46db3d58 writeback: make w... |
1721 1722 1723 1724 1725 1726 1727 |
/* 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 vfs: improve writ... |
1728 |
} |
66f3b8e2e writeback: move d... |
1729 |
/* Leave any unwritten inodes on b_io */ |
d46db3d58 writeback: make w... |
1730 |
return wrote; |
66f3b8e2e writeback: move d... |
1731 |
} |
7d9f073b8 mm/writeback: mak... |
1732 |
static long writeback_inodes_wb(struct bdi_writeback *wb, long nr_pages, |
0e175a183 writeback: Add a ... |
1733 |
enum wb_reason reason) |
edadfb10b writeback: split ... |
1734 |
{ |
d46db3d58 writeback: make w... |
1735 1736 1737 1738 |
struct wb_writeback_work work = { .nr_pages = nr_pages, .sync_mode = WB_SYNC_NONE, .range_cyclic = 1, |
0e175a183 writeback: Add a ... |
1739 |
.reason = reason, |
d46db3d58 writeback: make w... |
1740 |
}; |
505a666ee writeback: plug w... |
1741 |
struct blk_plug plug; |
edadfb10b writeback: split ... |
1742 |
|
505a666ee writeback: plug w... |
1743 |
blk_start_plug(&plug); |
f758eeabe writeback: split ... |
1744 |
spin_lock(&wb->list_lock); |
424b351fe writeback: refill... |
1745 |
if (list_empty(&wb->b_io)) |
6623c1904 writeback: Fix sy... |
1746 |
queue_io(wb, &work, jiffies); |
d46db3d58 writeback: make w... |
1747 |
__writeback_inodes_wb(wb, &work); |
f758eeabe writeback: split ... |
1748 |
spin_unlock(&wb->list_lock); |
505a666ee writeback: plug w... |
1749 |
blk_finish_plug(&plug); |
edadfb10b writeback: split ... |
1750 |
|
d46db3d58 writeback: make w... |
1751 1752 |
return nr_pages - work.nr_pages; } |
03ba3782e writeback: switch... |
1753 |
|
03ba3782e writeback: switch... |
1754 1755 |
/* * Explicit flushing or periodic writeback of "old" data. |
66f3b8e2e writeback: move d... |
1756 |
* |
03ba3782e writeback: switch... |
1757 1758 1759 1760 |
* 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 writeback: move d... |
1761 |
* |
03ba3782e writeback: switch... |
1762 1763 1764 |
* 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 writeback: move d... |
1765 |
* |
6623c1904 writeback: Fix sy... |
1766 |
* dirtied_before takes precedence over nr_to_write. So we'll only write back |
03ba3782e writeback: switch... |
1767 |
* all dirty pages if they are all attached to "old" mappings. |
66f3b8e2e writeback: move d... |
1768 |
*/ |
c4a77a6c7 writeback: make w... |
1769 |
static long wb_writeback(struct bdi_writeback *wb, |
83ba7b071 writeback: simpli... |
1770 |
struct wb_writeback_work *work) |
66f3b8e2e writeback: move d... |
1771 |
{ |
e98be2d59 writeback: bdi wr... |
1772 |
unsigned long wb_start = jiffies; |
d46db3d58 writeback: make w... |
1773 |
long nr_pages = work->nr_pages; |
6623c1904 writeback: Fix sy... |
1774 |
unsigned long dirtied_before = jiffies; |
a5989bdc9 fs: Fix busyloop ... |
1775 |
struct inode *inode; |
d46db3d58 writeback: make w... |
1776 |
long progress; |
505a666ee writeback: plug w... |
1777 |
struct blk_plug plug; |
66f3b8e2e writeback: move d... |
1778 |
|
505a666ee writeback: plug w... |
1779 |
blk_start_plug(&plug); |
e8dfc3058 writeback: elevat... |
1780 |
spin_lock(&wb->list_lock); |
03ba3782e writeback: switch... |
1781 1782 |
for (;;) { /* |
d3ddec763 writeback: stop b... |
1783 |
* Stop writeback when nr_pages has been consumed |
03ba3782e writeback: switch... |
1784 |
*/ |
83ba7b071 writeback: simpli... |
1785 |
if (work->nr_pages <= 0) |
03ba3782e writeback: switch... |
1786 |
break; |
66f3b8e2e writeback: move d... |
1787 |
|
38f219776 fs: sync_sb_inode... |
1788 |
/* |
aa373cf55 writeback: stop b... |
1789 1790 1791 1792 1793 1794 |
* 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 writeback: move b... |
1795 |
!list_empty(&wb->work_list)) |
aa373cf55 writeback: stop b... |
1796 1797 1798 |
break; /* |
d3ddec763 writeback: stop b... |
1799 1800 |
* For background writeout, stop when we are below the * background dirty threshold |
38f219776 fs: sync_sb_inode... |
1801 |
*/ |
aa661bbe1 writeback: move o... |
1802 |
if (work->for_background && !wb_over_bg_thresh(wb)) |
03ba3782e writeback: switch... |
1803 |
break; |
38f219776 fs: sync_sb_inode... |
1804 |
|
1bc36b642 writeback: Includ... |
1805 1806 1807 1808 1809 1810 |
/* * 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 writeback: the ku... |
1811 |
if (work->for_kupdate) { |
6623c1904 writeback: Fix sy... |
1812 |
dirtied_before = jiffies - |
ba9aa8399 writeback: the ku... |
1813 |
msecs_to_jiffies(dirty_expire_interval * 10); |
1bc36b642 writeback: Includ... |
1814 |
} else if (work->for_background) |
6623c1904 writeback: Fix sy... |
1815 |
dirtied_before = jiffies; |
028c2dd18 writeback: Add tr... |
1816 |
|
5634cc2aa writeback: update... |
1817 |
trace_writeback_start(wb, work); |
e8dfc3058 writeback: elevat... |
1818 |
if (list_empty(&wb->b_io)) |
6623c1904 writeback: Fix sy... |
1819 |
queue_io(wb, work, dirtied_before); |
83ba7b071 writeback: simpli... |
1820 |
if (work->sb) |
d46db3d58 writeback: make w... |
1821 |
progress = writeback_sb_inodes(work->sb, wb, work); |
edadfb10b writeback: split ... |
1822 |
else |
d46db3d58 writeback: make w... |
1823 |
progress = __writeback_inodes_wb(wb, work); |
5634cc2aa writeback: update... |
1824 |
trace_writeback_written(wb, work); |
028c2dd18 writeback: Add tr... |
1825 |
|
e98be2d59 writeback: bdi wr... |
1826 |
wb_update_bandwidth(wb, wb_start); |
03ba3782e writeback: switch... |
1827 1828 |
/* |
e6fb6da2e writeback: try mo... |
1829 1830 1831 1832 1833 1834 |
* 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 writeback: switch... |
1835 |
*/ |
d46db3d58 writeback: make w... |
1836 |
if (progress) |
71fd05a88 writeback: improv... |
1837 1838 |
continue; /* |
e6fb6da2e writeback: try mo... |
1839 |
* No more inodes for IO, bail |
71fd05a88 writeback: improv... |
1840 |
*/ |
b7a2441f9 writeback: remove... |
1841 |
if (list_empty(&wb->b_more_io)) |
03ba3782e writeback: switch... |
1842 |
break; |
71fd05a88 writeback: improv... |
1843 |
/* |
71fd05a88 writeback: improv... |
1844 1845 1846 1847 |
* Nothing written. Wait for some inode to * become available for writeback. Otherwise * we'll just busyloop. */ |
bace92481 fs/fs-writeback.c... |
1848 1849 1850 1851 1852 1853 1854 |
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 writeback: switch... |
1855 |
} |
e8dfc3058 writeback: elevat... |
1856 |
spin_unlock(&wb->list_lock); |
505a666ee writeback: plug w... |
1857 |
blk_finish_plug(&plug); |
03ba3782e writeback: switch... |
1858 |
|
d46db3d58 writeback: make w... |
1859 |
return nr_pages - work->nr_pages; |
03ba3782e writeback: switch... |
1860 1861 1862 |
} /* |
83ba7b071 writeback: simpli... |
1863 |
* Return the next wb_writeback_work struct that hasn't been processed yet. |
03ba3782e writeback: switch... |
1864 |
*/ |
f0054bb1e writeback: move b... |
1865 |
static struct wb_writeback_work *get_next_work_item(struct bdi_writeback *wb) |
03ba3782e writeback: switch... |
1866 |
{ |
83ba7b071 writeback: simpli... |
1867 |
struct wb_writeback_work *work = NULL; |
03ba3782e writeback: switch... |
1868 |
|
f0054bb1e writeback: move b... |
1869 1870 1871 |
spin_lock_bh(&wb->work_lock); if (!list_empty(&wb->work_list)) { work = list_entry(wb->work_list.next, |
83ba7b071 writeback: simpli... |
1872 1873 |
struct wb_writeback_work, list); list_del_init(&work->list); |
03ba3782e writeback: switch... |
1874 |
} |
f0054bb1e writeback: move b... |
1875 |
spin_unlock_bh(&wb->work_lock); |
83ba7b071 writeback: simpli... |
1876 |
return work; |
03ba3782e writeback: switch... |
1877 |
} |
6585027a5 writeback: integr... |
1878 1879 |
static long wb_check_background_flush(struct bdi_writeback *wb) { |
aa661bbe1 writeback: move o... |
1880 |
if (wb_over_bg_thresh(wb)) { |
6585027a5 writeback: integr... |
1881 1882 1883 1884 1885 1886 |
struct wb_writeback_work work = { .nr_pages = LONG_MAX, .sync_mode = WB_SYNC_NONE, .for_background = 1, .range_cyclic = 1, |
0e175a183 writeback: Add a ... |
1887 |
.reason = WB_REASON_BACKGROUND, |
6585027a5 writeback: integr... |
1888 1889 1890 1891 1892 1893 1894 |
}; return wb_writeback(wb, &work); } return 0; } |
03ba3782e writeback: switch... |
1895 1896 1897 1898 |
static long wb_check_old_data_flush(struct bdi_writeback *wb) { unsigned long expired; long nr_pages; |
69b62d01e writeback: disabl... |
1899 1900 1901 1902 1903 |
/* * When set to zero, disable periodic writeback */ if (!dirty_writeback_interval) return 0; |
03ba3782e writeback: switch... |
1904 1905 1906 1907 1908 1909 |
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 fs-writeback.c: u... |
1910 |
nr_pages = get_nr_dirty_pages(); |
03ba3782e writeback: switch... |
1911 |
|
c4a77a6c7 writeback: make w... |
1912 |
if (nr_pages) { |
83ba7b071 writeback: simpli... |
1913 |
struct wb_writeback_work work = { |
c4a77a6c7 writeback: make w... |
1914 1915 1916 1917 |
.nr_pages = nr_pages, .sync_mode = WB_SYNC_NONE, .for_kupdate = 1, .range_cyclic = 1, |
0e175a183 writeback: Add a ... |
1918 |
.reason = WB_REASON_PERIODIC, |
c4a77a6c7 writeback: make w... |
1919 |
}; |
83ba7b071 writeback: simpli... |
1920 |
return wb_writeback(wb, &work); |
c4a77a6c7 writeback: make w... |
1921 |
} |
03ba3782e writeback: switch... |
1922 1923 1924 |
return 0; } |
85009b4f5 writeback: elimin... |
1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 |
static long wb_check_start_all(struct bdi_writeback *wb) { long nr_pages; if (!test_bit(WB_start_all, &wb->state)) return 0; nr_pages = get_nr_dirty_pages(); if (nr_pages) { struct wb_writeback_work work = { .nr_pages = wb_split_bdi_pages(wb, nr_pages), .sync_mode = WB_SYNC_NONE, .range_cyclic = 1, .reason = wb->start_all_reason, }; nr_pages = wb_writeback(wb, &work); } clear_bit(WB_start_all, &wb->state); return nr_pages; } |
03ba3782e writeback: switch... |
1947 1948 1949 |
/* * Retrieve work items and do the writeback they describe */ |
25d130ba2 mm/writeback: don... |
1950 |
static long wb_do_writeback(struct bdi_writeback *wb) |
03ba3782e writeback: switch... |
1951 |
{ |
83ba7b071 writeback: simpli... |
1952 |
struct wb_writeback_work *work; |
c4a77a6c7 writeback: make w... |
1953 |
long wrote = 0; |
03ba3782e writeback: switch... |
1954 |
|
4452226ea writeback: move b... |
1955 |
set_bit(WB_writeback_running, &wb->state); |
f0054bb1e writeback: move b... |
1956 |
while ((work = get_next_work_item(wb)) != NULL) { |
5634cc2aa writeback: update... |
1957 |
trace_writeback_exec(wb, work); |
83ba7b071 writeback: simpli... |
1958 |
wrote += wb_writeback(wb, work); |
4a3a485b1 writeback: fix me... |
1959 |
finish_writeback_work(wb, work); |
03ba3782e writeback: switch... |
1960 1961 1962 |
} /* |
85009b4f5 writeback: elimin... |
1963 1964 1965 1966 1967 |
* Check for a flush-everything request */ wrote += wb_check_start_all(wb); /* |
03ba3782e writeback: switch... |
1968 1969 1970 |
* Check for periodic writeback, kupdated() style */ wrote += wb_check_old_data_flush(wb); |
6585027a5 writeback: integr... |
1971 |
wrote += wb_check_background_flush(wb); |
4452226ea writeback: move b... |
1972 |
clear_bit(WB_writeback_running, &wb->state); |
03ba3782e writeback: switch... |
1973 1974 1975 1976 1977 1978 |
return wrote; } /* * Handle writeback of dirty data for the device backed by this bdi. Also |
839a8e866 writeback: replac... |
1979 |
* reschedules periodically and does kupdated style flushing. |
03ba3782e writeback: switch... |
1980 |
*/ |
f0054bb1e writeback: move b... |
1981 |
void wb_workfn(struct work_struct *work) |
03ba3782e writeback: switch... |
1982 |
{ |
839a8e866 writeback: replac... |
1983 1984 |
struct bdi_writeback *wb = container_of(to_delayed_work(work), struct bdi_writeback, dwork); |
03ba3782e writeback: switch... |
1985 |
long pages_written; |
c2c814fc9 memcg: fix a cras... |
1986 |
set_worker_desc("flush-%s", bdi_dev_name(wb->bdi)); |
766f91641 kernel: remove PF... |
1987 |
current->flags |= PF_SWAPWRITE; |
455b28646 writeback: Initia... |
1988 |
|
839a8e866 writeback: replac... |
1989 |
if (likely(!current_is_workqueue_rescuer() || |
4452226ea writeback: move b... |
1990 |
!test_bit(WB_registered, &wb->state))) { |
6467716a3 writeback: optimi... |
1991 |
/* |
f0054bb1e writeback: move b... |
1992 |
* The normal path. Keep writing back @wb until its |
839a8e866 writeback: replac... |
1993 |
* work_list is empty. Note that this path is also taken |
f0054bb1e writeback: move b... |
1994 |
* if @wb is shutting down even when we're running off the |
839a8e866 writeback: replac... |
1995 |
* rescuer as work_list needs to be drained. |
6467716a3 writeback: optimi... |
1996 |
*/ |
839a8e866 writeback: replac... |
1997 |
do { |
25d130ba2 mm/writeback: don... |
1998 |
pages_written = wb_do_writeback(wb); |
839a8e866 writeback: replac... |
1999 |
trace_writeback_pages_written(pages_written); |
f0054bb1e writeback: move b... |
2000 |
} while (!list_empty(&wb->work_list)); |
839a8e866 writeback: replac... |
2001 2002 2003 2004 2005 2006 |
} 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 writeback: move b... |
2007 |
pages_written = writeback_inodes_wb(wb, 1024, |
839a8e866 writeback: replac... |
2008 |
WB_REASON_FORKER_THREAD); |
455b28646 writeback: Initia... |
2009 |
trace_writeback_pages_written(pages_written); |
03ba3782e writeback: switch... |
2010 |
} |
f0054bb1e writeback: move b... |
2011 |
if (!list_empty(&wb->work_list)) |
b8b784958 bdi: Fix oops in ... |
2012 |
wb_wakeup(wb); |
6ca738d60 backing_dev: fix ... |
2013 |
else if (wb_has_dirty_io(wb) && dirty_writeback_interval) |
f0054bb1e writeback: move b... |
2014 |
wb_wakeup_delayed(wb); |
455b28646 writeback: Initia... |
2015 |
|
839a8e866 writeback: replac... |
2016 |
current->flags &= ~PF_SWAPWRITE; |
03ba3782e writeback: switch... |
2017 2018 2019 |
} /* |
595043e5f writeback: provid... |
2020 2021 2022 2023 |
* Start writeback of `nr_pages' pages on this bdi. If `nr_pages' is zero, * write back the whole world. */ static void __wakeup_flusher_threads_bdi(struct backing_dev_info *bdi, |
e8e8a0c6c writeback: move n... |
2024 |
enum wb_reason reason) |
595043e5f writeback: provid... |
2025 2026 2027 2028 2029 2030 2031 |
{ struct bdi_writeback *wb; if (!bdi_has_dirty_io(bdi)) return; list_for_each_entry_rcu(wb, &bdi->wb_list, bdi_node) |
e8e8a0c6c writeback: move n... |
2032 |
wb_start_writeback(wb, reason); |
595043e5f writeback: provid... |
2033 2034 2035 2036 2037 |
} void wakeup_flusher_threads_bdi(struct backing_dev_info *bdi, enum wb_reason reason) { |
595043e5f writeback: provid... |
2038 |
rcu_read_lock(); |
e8e8a0c6c writeback: move n... |
2039 |
__wakeup_flusher_threads_bdi(bdi, reason); |
595043e5f writeback: provid... |
2040 2041 2042 2043 |
rcu_read_unlock(); } /* |
9ba4b2dfa fs: kill 'nr_page... |
2044 |
* Wakeup the flusher threads to start writeback of all currently dirty pages |
03ba3782e writeback: switch... |
2045 |
*/ |
9ba4b2dfa fs: kill 'nr_page... |
2046 |
void wakeup_flusher_threads(enum wb_reason reason) |
03ba3782e writeback: switch... |
2047 |
{ |
b8c2f3474 writeback: simpli... |
2048 |
struct backing_dev_info *bdi; |
03ba3782e writeback: switch... |
2049 |
|
51350ea0d mm, writeback: fl... |
2050 2051 2052 2053 2054 |
/* * If we are expecting writeback progress we must submit plugged IO. */ if (blk_needs_flush_plug(current)) blk_schedule_flush_plug(current); |
b8c2f3474 writeback: simpli... |
2055 |
rcu_read_lock(); |
595043e5f writeback: provid... |
2056 |
list_for_each_entry_rcu(bdi, &bdi_list, bdi_list) |
e8e8a0c6c writeback: move n... |
2057 |
__wakeup_flusher_threads_bdi(bdi, reason); |
cfc4ba536 writeback: use RC... |
2058 |
rcu_read_unlock(); |
1da177e4c Linux-2.6.12-rc2 |
2059 |
} |
a2f487069 fs: make sure the... |
2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 |
/* * 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 writeback: make w... |
2084 |
struct bdi_writeback *wb; |
001fe6f61 writeback: make w... |
2085 |
|
b817525a4 writeback: bdi_wr... |
2086 |
list_for_each_entry_rcu(wb, &bdi->wb_list, bdi_node) |
6fdf860f1 writeback: fix bd... |
2087 2088 |
if (!list_empty(&wb->b_dirty_time)) wb_wakeup(wb); |
a2f487069 fs: make sure the... |
2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 |
} 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 fs: add dirtytime... |
2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 |
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 writeback: switch... |
2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 |
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 fs: add a blank l... |
2134 2135 2136 2137 2138 2139 2140 |
* __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 Linux-2.6.12-rc2 |
2141 |
* |
03ba3782e writeback: switch... |
2142 2143 2144 2145 2146 2147 2148 2149 2150 |
* 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 Linux-2.6.12-rc2 |
2151 |
* |
03ba3782e writeback: switch... |
2152 2153 2154 2155 2156 2157 |
* 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 Linux-2.6.12-rc2 |
2158 |
*/ |
03ba3782e writeback: switch... |
2159 |
void __mark_inode_dirty(struct inode *inode, int flags) |
1da177e4c Linux-2.6.12-rc2 |
2160 |
{ |
03ba3782e writeback: switch... |
2161 |
struct super_block *sb = inode->i_sb; |
0ae45f63d vfs: add support ... |
2162 2163 2164 |
int dirtytime; trace_writeback_mark_inode_dirty(inode, flags); |
1da177e4c Linux-2.6.12-rc2 |
2165 |
|
03ba3782e writeback: switch... |
2166 2167 2168 2169 |
/* * Don't do this for I_DIRTY_PAGES - that doesn't actually * dirty the inode itself */ |
0e11f6443 fs: move I_DIRTY_... |
2170 |
if (flags & (I_DIRTY_INODE | I_DIRTY_TIME)) { |
9fb0a7da0 writeback: add mo... |
2171 |
trace_writeback_dirty_inode_start(inode, flags); |
03ba3782e writeback: switch... |
2172 |
if (sb->s_op->dirty_inode) |
aa3857295 fs: pass exact ty... |
2173 |
sb->s_op->dirty_inode(inode, flags); |
9fb0a7da0 writeback: add mo... |
2174 2175 |
trace_writeback_dirty_inode(inode, flags); |
03ba3782e writeback: switch... |
2176 |
} |
0ae45f63d vfs: add support ... |
2177 2178 2179 |
if (flags & I_DIRTY_INODE) flags &= ~I_DIRTY_TIME; dirtytime = flags & I_DIRTY_TIME; |
03ba3782e writeback: switch... |
2180 2181 |
/* |
9c6ac78eb writeback: fix a ... |
2182 2183 |
* Paired with smp_mb() in __writeback_single_inode() for the * following lockless i_state test. See there for details. |
03ba3782e writeback: switch... |
2184 2185 |
*/ smp_mb(); |
0ae45f63d vfs: add support ... |
2186 2187 |
if (((inode->i_state & flags) == flags) || (dirtytime && (inode->i_state & I_DIRTY_INODE))) |
03ba3782e writeback: switch... |
2188 2189 2190 2191 |
return; if (unlikely(block_dump)) block_dump___mark_inode_dirty(inode); |
250df6ed2 fs: protect inode... |
2192 |
spin_lock(&inode->i_lock); |
0ae45f63d vfs: add support ... |
2193 2194 |
if (dirtytime && (inode->i_state & I_DIRTY_INODE)) goto out_unlock_inode; |
03ba3782e writeback: switch... |
2195 2196 |
if ((inode->i_state & flags) != flags) { const int was_dirty = inode->i_state & I_DIRTY; |
52ebea749 writeback: make b... |
2197 |
inode_attach_wb(inode, NULL); |
0ae45f63d vfs: add support ... |
2198 2199 |
if (flags & I_DIRTY_INODE) inode->i_state &= ~I_DIRTY_TIME; |
03ba3782e writeback: switch... |
2200 2201 2202 |
inode->i_state |= flags; /* |
cb0c74450 writeback: Avoid ... |
2203 2204 2205 2206 |
* If the inode is queued for writeback by flush worker, just * update its dirty state. Once the flush worker is done with * the inode it will place it on the appropriate superblock * list, based upon its state. |
03ba3782e writeback: switch... |
2207 |
*/ |
cb0c74450 writeback: Avoid ... |
2208 |
if (inode->i_state & I_SYNC_QUEUED) |
250df6ed2 fs: protect inode... |
2209 |
goto out_unlock_inode; |
03ba3782e writeback: switch... |
2210 2211 2212 2213 2214 2215 |
/* * Only add valid (hashed) inodes to the superblock's * dirty list. Add blockdev inodes as well. */ if (!S_ISBLK(inode->i_mode)) { |
1d3382cbf new helper: inode... |
2216 |
if (inode_unhashed(inode)) |
250df6ed2 fs: protect inode... |
2217 |
goto out_unlock_inode; |
03ba3782e writeback: switch... |
2218 |
} |
a4ffdde6e simplify checks f... |
2219 |
if (inode->i_state & I_FREEING) |
250df6ed2 fs: protect inode... |
2220 |
goto out_unlock_inode; |
03ba3782e writeback: switch... |
2221 2222 2223 2224 2225 2226 |
/* * 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 writeback: implem... |
2227 |
struct bdi_writeback *wb; |
d6c10f1fc writeback: implem... |
2228 |
struct list_head *dirty_list; |
a66979aba fs: move i_wb_lis... |
2229 |
bool wakeup_bdi = false; |
253c34e9b writeback: preven... |
2230 |
|
87e1d789b writeback: implem... |
2231 |
wb = locked_inode_to_wb_and_lock_list(inode); |
253c34e9b writeback: preven... |
2232 |
|
0747259d1 writeback: dirty ... |
2233 2234 2235 2236 |
WARN(bdi_cap_writeback_dirty(wb->bdi) && !test_bit(WB_registered, &wb->state), "bdi-%s not registered ", wb->bdi->name); |
03ba3782e writeback: switch... |
2237 2238 |
inode->dirtied_when = jiffies; |
a2f487069 fs: make sure the... |
2239 2240 |
if (dirtytime) inode->dirtied_time_when = jiffies; |
d6c10f1fc writeback: implem... |
2241 |
|
0e11f6443 fs: move I_DIRTY_... |
2242 |
if (inode->i_state & I_DIRTY) |
0747259d1 writeback: dirty ... |
2243 |
dirty_list = &wb->b_dirty; |
a2f487069 fs: make sure the... |
2244 |
else |
0747259d1 writeback: dirty ... |
2245 |
dirty_list = &wb->b_dirty_time; |
d6c10f1fc writeback: implem... |
2246 |
|
c7f540849 inode: rename i_w... |
2247 |
wakeup_bdi = inode_io_list_move_locked(inode, wb, |
d6c10f1fc writeback: implem... |
2248 |
dirty_list); |
0747259d1 writeback: dirty ... |
2249 |
spin_unlock(&wb->list_lock); |
0ae45f63d vfs: add support ... |
2250 |
trace_writeback_dirty_inode_enqueue(inode); |
a66979aba fs: move i_wb_lis... |
2251 |
|
d6c10f1fc writeback: implem... |
2252 2253 2254 2255 2256 2257 |
/* * 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 writeback: dirty ... |
2258 2259 |
if (bdi_cap_writeback_dirty(wb->bdi) && wakeup_bdi) wb_wakeup_delayed(wb); |
a66979aba fs: move i_wb_lis... |
2260 |
return; |
1da177e4c Linux-2.6.12-rc2 |
2261 |
} |
1da177e4c Linux-2.6.12-rc2 |
2262 |
} |
250df6ed2 fs: protect inode... |
2263 2264 |
out_unlock_inode: spin_unlock(&inode->i_lock); |
03ba3782e writeback: switch... |
2265 2266 |
} EXPORT_SYMBOL(__mark_inode_dirty); |
e97fedb9e sync: serialise p... |
2267 2268 2269 2270 2271 2272 2273 2274 2275 |
/* * 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 writeback: separa... |
2276 |
static void wait_sb_inodes(struct super_block *sb) |
03ba3782e writeback: switch... |
2277 |
{ |
6c60d2b57 fs/fs-writeback.c... |
2278 |
LIST_HEAD(sync_list); |
03ba3782e writeback: switch... |
2279 2280 2281 2282 2283 |
/* * We need to be protected against the filesystem going from * r/o to r/w or vice versa. */ |
b6e51316d writeback: separa... |
2284 |
WARN_ON(!rwsem_is_locked(&sb->s_umount)); |
03ba3782e writeback: switch... |
2285 |
|
e97fedb9e sync: serialise p... |
2286 |
mutex_lock(&sb->s_sync_lock); |
03ba3782e writeback: switch... |
2287 2288 |
/* |
6c60d2b57 fs/fs-writeback.c... |
2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 |
* 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 writeback: switch... |
2307 |
*/ |
6c60d2b57 fs/fs-writeback.c... |
2308 2309 2310 |
while (!list_empty(&sync_list)) { struct inode *inode = list_first_entry(&sync_list, struct inode, i_wb_list); |
250df6ed2 fs: protect inode... |
2311 |
struct address_space *mapping = inode->i_mapping; |
03ba3782e writeback: switch... |
2312 |
|
6c60d2b57 fs/fs-writeback.c... |
2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 |
/* * 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 fs: protect inode... |
2330 |
spin_lock(&inode->i_lock); |
6c60d2b57 fs/fs-writeback.c... |
2331 |
if (inode->i_state & (I_FREEING|I_WILL_FREE|I_NEW)) { |
250df6ed2 fs: protect inode... |
2332 |
spin_unlock(&inode->i_lock); |
6c60d2b57 fs/fs-writeback.c... |
2333 2334 |
spin_lock_irq(&sb->s_inode_wblist_lock); |
03ba3782e writeback: switch... |
2335 |
continue; |
250df6ed2 fs: protect inode... |
2336 |
} |
03ba3782e writeback: switch... |
2337 |
__iget(inode); |
250df6ed2 fs: protect inode... |
2338 |
spin_unlock(&inode->i_lock); |
6c60d2b57 fs/fs-writeback.c... |
2339 |
rcu_read_unlock(); |
03ba3782e writeback: switch... |
2340 |
|
aa750fd71 mm/filemap.c: mak... |
2341 2342 2343 2344 2345 2346 |
/* * 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 writeback: switch... |
2347 2348 |
cond_resched(); |
6c60d2b57 fs/fs-writeback.c... |
2349 2350 2351 2352 |
iput(inode); rcu_read_lock(); spin_lock_irq(&sb->s_inode_wblist_lock); |
03ba3782e writeback: switch... |
2353 |
} |
6c60d2b57 fs/fs-writeback.c... |
2354 2355 |
spin_unlock_irq(&sb->s_inode_wblist_lock); rcu_read_unlock(); |
e97fedb9e sync: serialise p... |
2356 |
mutex_unlock(&sb->s_sync_lock); |
1da177e4c Linux-2.6.12-rc2 |
2357 |
} |
f30a7d0cc writeback: restru... |
2358 2359 |
static void __writeback_inodes_sb_nr(struct super_block *sb, unsigned long nr, enum wb_reason reason, bool skip_if_busy) |
1da177e4c Linux-2.6.12-rc2 |
2360 |
{ |
5b9cce4c7 writeback: Genera... |
2361 2362 |
struct backing_dev_info *bdi = sb->s_bdi; DEFINE_WB_COMPLETION(done, bdi); |
83ba7b071 writeback: simpli... |
2363 |
struct wb_writeback_work work = { |
6e6938b6d writeback: introd... |
2364 2365 2366 2367 2368 |
.sb = sb, .sync_mode = WB_SYNC_NONE, .tagged_writepages = 1, .done = &done, .nr_pages = nr, |
0e175a183 writeback: Add a ... |
2369 |
.reason = reason, |
3c4d71653 writeback: queue ... |
2370 |
}; |
d8a8559cd writeback: get ri... |
2371 |
|
e79729123 writeback: don't ... |
2372 |
if (!bdi_has_dirty_io(bdi) || bdi == &noop_backing_dev_info) |
6eedc7015 vfs: Move noop_ba... |
2373 |
return; |
cf37e9724 writeback: enforc... |
2374 |
WARN_ON(!rwsem_is_locked(&sb->s_umount)); |
f30a7d0cc writeback: restru... |
2375 |
|
db1253604 writeback: make w... |
2376 |
bdi_split_work_to_wbs(sb->s_bdi, &work, skip_if_busy); |
5b9cce4c7 writeback: Genera... |
2377 |
wb_wait_for_completion(&done); |
e913fc825 writeback: fix WB... |
2378 |
} |
f30a7d0cc writeback: restru... |
2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 |
/** * 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 Add new functions... |
2396 2397 2398 2399 2400 |
EXPORT_SYMBOL(writeback_inodes_sb_nr); /** * writeback_inodes_sb - writeback dirty inodes from given super_block * @sb: the superblock |
786228ab3 writeback: Fix is... |
2401 |
* @reason: reason why some writeback work was initiated |
3259f8bed Add new functions... |
2402 2403 2404 2405 2406 |
* * 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 writeback: Add a ... |
2407 |
void writeback_inodes_sb(struct super_block *sb, enum wb_reason reason) |
3259f8bed Add new functions... |
2408 |
{ |
0e175a183 writeback: Add a ... |
2409 |
return writeback_inodes_sb_nr(sb, get_nr_dirty_pages(), reason); |
3259f8bed Add new functions... |
2410 |
} |
0e3c9a228 Revert "writeback... |
2411 |
EXPORT_SYMBOL(writeback_inodes_sb); |
e913fc825 writeback: fix WB... |
2412 2413 |
/** |
8264c3214 writeback: merge ... |
2414 |
* try_to_writeback_inodes_sb - try to start writeback if none underway |
17bd55d03 fs-writeback: Add... |
2415 |
* @sb: the superblock |
8264c3214 writeback: merge ... |
2416 |
* @reason: reason why some writeback work was initiated |
17bd55d03 fs-writeback: Add... |
2417 |
* |
8264c3214 writeback: merge ... |
2418 |
* Invoke __writeback_inodes_sb_nr if no writeback is currently underway. |
17bd55d03 fs-writeback: Add... |
2419 |
*/ |
8264c3214 writeback: merge ... |
2420 |
void try_to_writeback_inodes_sb(struct super_block *sb, enum wb_reason reason) |
17bd55d03 fs-writeback: Add... |
2421 |
{ |
10ee27a06 vfs: re-implement... |
2422 |
if (!down_read_trylock(&sb->s_umount)) |
8264c3214 writeback: merge ... |
2423 |
return; |
10ee27a06 vfs: re-implement... |
2424 |
|
8264c3214 writeback: merge ... |
2425 |
__writeback_inodes_sb_nr(sb, get_nr_dirty_pages(), reason, true); |
10ee27a06 vfs: re-implement... |
2426 |
up_read(&sb->s_umount); |
3259f8bed Add new functions... |
2427 |
} |
10ee27a06 vfs: re-implement... |
2428 |
EXPORT_SYMBOL(try_to_writeback_inodes_sb); |
3259f8bed Add new functions... |
2429 2430 |
/** |
d8a8559cd writeback: get ri... |
2431 |
* sync_inodes_sb - sync sb inode pages |
0dc83bd30 Revert "writeback... |
2432 |
* @sb: the superblock |
d8a8559cd writeback: get ri... |
2433 2434 |
* * This function writes and waits on any dirty inode belonging to this |
0dc83bd30 Revert "writeback... |
2435 |
* super_block. |
d8a8559cd writeback: get ri... |
2436 |
*/ |
0dc83bd30 Revert "writeback... |
2437 |
void sync_inodes_sb(struct super_block *sb) |
d8a8559cd writeback: get ri... |
2438 |
{ |
5b9cce4c7 writeback: Genera... |
2439 2440 |
struct backing_dev_info *bdi = sb->s_bdi; DEFINE_WB_COMPLETION(done, bdi); |
83ba7b071 writeback: simpli... |
2441 |
struct wb_writeback_work work = { |
3c4d71653 writeback: queue ... |
2442 2443 2444 2445 |
.sb = sb, .sync_mode = WB_SYNC_ALL, .nr_pages = LONG_MAX, .range_cyclic = 0, |
83ba7b071 writeback: simpli... |
2446 |
.done = &done, |
0e175a183 writeback: Add a ... |
2447 |
.reason = WB_REASON_SYNC, |
7747bd4bc sync: don't block... |
2448 |
.for_sync = 1, |
3c4d71653 writeback: queue ... |
2449 |
}; |
006a0973e writeback: sync_i... |
2450 2451 2452 2453 2454 2455 |
/* * 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 vfs: Move noop_ba... |
2456 |
return; |
cf37e9724 writeback: enforc... |
2457 |
WARN_ON(!rwsem_is_locked(&sb->s_umount)); |
7fc5854f8 writeback: synchr... |
2458 2459 |
/* protect against inode wb switch, see inode_switch_wbs_work_fn() */ bdi_down_write_wb_switch_rwsem(bdi); |
db1253604 writeback: make w... |
2460 |
bdi_split_work_to_wbs(bdi, &work, false); |
5b9cce4c7 writeback: Genera... |
2461 |
wb_wait_for_completion(&done); |
7fc5854f8 writeback: synchr... |
2462 |
bdi_up_write_wb_switch_rwsem(bdi); |
83ba7b071 writeback: simpli... |
2463 |
|
b6e51316d writeback: separa... |
2464 |
wait_sb_inodes(sb); |
1da177e4c Linux-2.6.12-rc2 |
2465 |
} |
d8a8559cd writeback: get ri... |
2466 |
EXPORT_SYMBOL(sync_inodes_sb); |
1da177e4c Linux-2.6.12-rc2 |
2467 |
|
1da177e4c Linux-2.6.12-rc2 |
2468 |
/** |
7f04c26d7 [PATCH] fix nr_un... |
2469 2470 2471 2472 2473 2474 |
* 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 Linux-2.6.12-rc2 |
2475 |
* |
7f04c26d7 [PATCH] fix nr_un... |
2476 |
* The caller must either have a ref on the inode or must have set I_WILL_FREE. |
1da177e4c Linux-2.6.12-rc2 |
2477 |
*/ |
1da177e4c Linux-2.6.12-rc2 |
2478 2479 |
int write_inode_now(struct inode *inode, int sync) { |
1da177e4c Linux-2.6.12-rc2 |
2480 2481 |
struct writeback_control wbc = { .nr_to_write = LONG_MAX, |
18914b188 write_inode_now()... |
2482 |
.sync_mode = sync ? WB_SYNC_ALL : WB_SYNC_NONE, |
111ebb6e6 [PATCH] writeback... |
2483 2484 |
.range_start = 0, .range_end = LLONG_MAX, |
1da177e4c Linux-2.6.12-rc2 |
2485 2486 2487 |
}; if (!mapping_cap_writeback_dirty(inode->i_mapping)) |
49364ce25 [PATCH] write_ino... |
2488 |
wbc.nr_to_write = 0; |
1da177e4c Linux-2.6.12-rc2 |
2489 2490 |
might_sleep(); |
aaf255933 writeback, cgroup... |
2491 |
return writeback_single_inode(inode, &wbc); |
1da177e4c Linux-2.6.12-rc2 |
2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 |
} 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 writeback, cgroup... |
2508 |
return writeback_single_inode(inode, wbc); |
1da177e4c Linux-2.6.12-rc2 |
2509 2510 |
} EXPORT_SYMBOL(sync_inode); |
c37650161 fs: add sync_inod... |
2511 2512 |
/** |
c691b9d98 sync_inode_metada... |
2513 |
* sync_inode_metadata - write an inode to disk |
c37650161 fs: add sync_inod... |
2514 2515 2516 |
* @inode: the inode to sync * @wait: wait for I/O to complete. * |
c691b9d98 sync_inode_metada... |
2517 |
* Write an inode to disk and adjust its dirty state after completion. |
c37650161 fs: add sync_inod... |
2518 2519 2520 2521 2522 2523 2524 2525 2526 2527 2528 2529 2530 |
* * 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); |