Blame view
mm/page-writeback.c
38.8 KB
1da177e4c
|
1 |
/* |
f30c22695
|
2 |
* mm/page-writeback.c |
1da177e4c
|
3 4 |
* * Copyright (C) 2002, Linus Torvalds. |
04fbfdc14
|
5 |
* Copyright (C) 2007 Red Hat, Inc., Peter Zijlstra <pzijlstr@redhat.com> |
1da177e4c
|
6 7 8 9 |
* * Contains functions related to writing back dirty pages at the * address_space level. * |
e1f8e8744
|
10 |
* 10Apr2002 Andrew Morton |
1da177e4c
|
11 12 13 14 15 16 17 18 19 20 21 22 23 24 |
* Initial version */ #include <linux/kernel.h> #include <linux/module.h> #include <linux/spinlock.h> #include <linux/fs.h> #include <linux/mm.h> #include <linux/swap.h> #include <linux/slab.h> #include <linux/pagemap.h> #include <linux/writeback.h> #include <linux/init.h> #include <linux/backing-dev.h> |
55e829af0
|
25 |
#include <linux/task_io_accounting_ops.h> |
1da177e4c
|
26 27 |
#include <linux/blkdev.h> #include <linux/mpage.h> |
d08b3851d
|
28 |
#include <linux/rmap.h> |
1da177e4c
|
29 30 31 32 33 34 |
#include <linux/percpu.h> #include <linux/notifier.h> #include <linux/smp.h> #include <linux/sysctl.h> #include <linux/cpu.h> #include <linux/syscalls.h> |
cf9a2ae8d
|
35 |
#include <linux/buffer_head.h> |
811d736f9
|
36 |
#include <linux/pagevec.h> |
028c2dd18
|
37 |
#include <trace/events/writeback.h> |
1da177e4c
|
38 39 |
/* |
1da177e4c
|
40 41 42 43 |
* After a CPU has dirtied this many pages, balance_dirty_pages_ratelimited * will look to see if it needs to force writeback or throttling. */ static long ratelimit_pages = 32; |
1da177e4c
|
44 45 46 |
/* * When balance_dirty_pages decides that the caller needs to perform some * non-background writeback, this is how many pages it will attempt to write. |
3a2e9a5a2
|
47 |
* It should be somewhat larger than dirtied pages to ensure that reasonably |
1da177e4c
|
48 49 |
* large amounts of I/O are submitted. */ |
3a2e9a5a2
|
50 |
static inline long sync_writeback_pages(unsigned long dirtied) |
1da177e4c
|
51 |
{ |
3a2e9a5a2
|
52 53 54 55 |
if (dirtied < ratelimit_pages) dirtied = ratelimit_pages; return dirtied + dirtied / 2; |
1da177e4c
|
56 57 58 59 60 |
} /* The following parameters are exported via /proc/sys/vm */ /* |
5b0830cb9
|
61 |
* Start background writeback (via writeback threads) at this percentage |
1da177e4c
|
62 |
*/ |
1b5e62b42
|
63 |
int dirty_background_ratio = 10; |
1da177e4c
|
64 65 |
/* |
2da02997e
|
66 67 68 69 70 71 |
* dirty_background_bytes starts at 0 (disabled) so that it is a function of * dirty_background_ratio * the amount of dirtyable memory */ unsigned long dirty_background_bytes; /* |
195cf453d
|
72 73 74 75 76 77 |
* free highmem will not be subtracted from the total free memory * for calculating free ratios if vm_highmem_is_dirtyable is true */ int vm_highmem_is_dirtyable; /* |
1da177e4c
|
78 79 |
* The generator of dirty data starts writeback at this percentage */ |
1b5e62b42
|
80 |
int vm_dirty_ratio = 20; |
1da177e4c
|
81 82 |
/* |
2da02997e
|
83 84 85 86 87 88 |
* vm_dirty_bytes starts at 0 (disabled) so that it is a function of * vm_dirty_ratio * the amount of dirtyable memory */ unsigned long vm_dirty_bytes; /* |
704503d83
|
89 |
* The interval between `kupdate'-style writebacks |
1da177e4c
|
90 |
*/ |
22ef37eed
|
91 |
unsigned int dirty_writeback_interval = 5 * 100; /* centiseconds */ |
1da177e4c
|
92 93 |
/* |
704503d83
|
94 |
* The longest time for which data is allowed to remain dirty |
1da177e4c
|
95 |
*/ |
22ef37eed
|
96 |
unsigned int dirty_expire_interval = 30 * 100; /* centiseconds */ |
1da177e4c
|
97 98 99 100 101 102 103 |
/* * Flag that makes the machine dump writes/reads and block dirtyings. */ int block_dump; /* |
ed5b43f15
|
104 105 |
* Flag that puts the machine in "laptop mode". Doubles as a timeout in jiffies: * a full sync is triggered after this time elapses without any disk activity. |
1da177e4c
|
106 107 108 109 110 111 |
*/ int laptop_mode; EXPORT_SYMBOL(laptop_mode); /* End of sysctl-exported parameters */ |
1da177e4c
|
112 |
/* |
04fbfdc14
|
113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 |
* Scale the writeback cache size proportional to the relative writeout speeds. * * We do this by keeping a floating proportion between BDIs, based on page * writeback completions [end_page_writeback()]. Those devices that write out * pages fastest will get the larger share, while the slower will get a smaller * share. * * We use page writeout completions because we are interested in getting rid of * dirty pages. Having them written out is the primary goal. * * We introduce a concept of time, a period over which we measure these events, * because demand can/will vary over time. The length of this period itself is * measured in page writeback completions. * */ static struct prop_descriptor vm_completions; |
3e26c149c
|
129 |
static struct prop_descriptor vm_dirties; |
04fbfdc14
|
130 |
|
04fbfdc14
|
131 132 133 134 135 136 137 138 |
/* * couple the period to the dirty_ratio: * * period/2 ~ roundup_pow_of_two(dirty limit) */ static int calc_period_shift(void) { unsigned long dirty_total; |
2da02997e
|
139 140 141 142 143 |
if (vm_dirty_bytes) dirty_total = vm_dirty_bytes / PAGE_SIZE; else dirty_total = (vm_dirty_ratio * determine_dirtyable_memory()) / 100; |
04fbfdc14
|
144 145 146 147 |
return 2 + ilog2(dirty_total - 1); } /* |
2da02997e
|
148 |
* update the period when the dirty threshold changes. |
04fbfdc14
|
149 |
*/ |
2da02997e
|
150 151 152 153 154 155 156 157 |
static void update_completion_period(void) { int shift = calc_period_shift(); prop_change_shift(&vm_completions, shift); prop_change_shift(&vm_dirties, shift); } int dirty_background_ratio_handler(struct ctl_table *table, int write, |
8d65af789
|
158 |
void __user *buffer, size_t *lenp, |
2da02997e
|
159 160 161 |
loff_t *ppos) { int ret; |
8d65af789
|
162 |
ret = proc_dointvec_minmax(table, write, buffer, lenp, ppos); |
2da02997e
|
163 164 165 166 167 168 |
if (ret == 0 && write) dirty_background_bytes = 0; return ret; } int dirty_background_bytes_handler(struct ctl_table *table, int write, |
8d65af789
|
169 |
void __user *buffer, size_t *lenp, |
2da02997e
|
170 171 172 |
loff_t *ppos) { int ret; |
8d65af789
|
173 |
ret = proc_doulongvec_minmax(table, write, buffer, lenp, ppos); |
2da02997e
|
174 175 176 177 |
if (ret == 0 && write) dirty_background_ratio = 0; return ret; } |
04fbfdc14
|
178 |
int dirty_ratio_handler(struct ctl_table *table, int write, |
8d65af789
|
179 |
void __user *buffer, size_t *lenp, |
04fbfdc14
|
180 181 182 |
loff_t *ppos) { int old_ratio = vm_dirty_ratio; |
2da02997e
|
183 |
int ret; |
8d65af789
|
184 |
ret = proc_dointvec_minmax(table, write, buffer, lenp, ppos); |
04fbfdc14
|
185 |
if (ret == 0 && write && vm_dirty_ratio != old_ratio) { |
2da02997e
|
186 187 188 189 190 191 192 193 |
update_completion_period(); vm_dirty_bytes = 0; } return ret; } int dirty_bytes_handler(struct ctl_table *table, int write, |
8d65af789
|
194 |
void __user *buffer, size_t *lenp, |
2da02997e
|
195 196 |
loff_t *ppos) { |
fc3501d41
|
197 |
unsigned long old_bytes = vm_dirty_bytes; |
2da02997e
|
198 |
int ret; |
8d65af789
|
199 |
ret = proc_doulongvec_minmax(table, write, buffer, lenp, ppos); |
2da02997e
|
200 201 202 |
if (ret == 0 && write && vm_dirty_bytes != old_bytes) { update_completion_period(); vm_dirty_ratio = 0; |
04fbfdc14
|
203 204 205 206 207 208 209 210 211 212 |
} return ret; } /* * Increment the BDI's writeout completion count and the global writeout * completion count. Called from test_clear_page_writeback(). */ static inline void __bdi_writeout_inc(struct backing_dev_info *bdi) { |
a42dde041
|
213 214 |
__prop_inc_percpu_max(&vm_completions, &bdi->completions, bdi->max_prop_frac); |
04fbfdc14
|
215 |
} |
dd5656e59
|
216 217 218 219 220 221 222 223 224 |
void bdi_writeout_inc(struct backing_dev_info *bdi) { unsigned long flags; local_irq_save(flags); __bdi_writeout_inc(bdi); local_irq_restore(flags); } EXPORT_SYMBOL_GPL(bdi_writeout_inc); |
1cf6e7d83
|
225 |
void task_dirty_inc(struct task_struct *tsk) |
3e26c149c
|
226 227 228 |
{ prop_inc_single(&vm_dirties, &tsk->dirties); } |
04fbfdc14
|
229 230 231 232 233 234 235 236 237 238 239 240 241 242 |
/* * Obtain an accurate fraction of the BDI's portion. */ static void bdi_writeout_fraction(struct backing_dev_info *bdi, long *numerator, long *denominator) { if (bdi_cap_writeback_dirty(bdi)) { prop_fraction_percpu(&vm_completions, &bdi->completions, numerator, denominator); } else { *numerator = 0; *denominator = 1; } } |
3e26c149c
|
243 244 245 246 247 248 249 250 |
static inline void task_dirties_fraction(struct task_struct *tsk, long *numerator, long *denominator) { prop_fraction_single(&vm_dirties, &tsk->dirties, numerator, denominator); } /* |
1babe1838
|
251 |
* task_dirty_limit - scale down dirty throttling threshold for one task |
3e26c149c
|
252 253 254 255 |
* * task specific dirty limit: * * dirty -= (dirty/8) * p_{t} |
1babe1838
|
256 257 258 259 260 261 262 |
* * To protect light/slow dirtying tasks from heavier/fast ones, we start * throttling individual tasks before reaching the bdi dirty limit. * Relatively low thresholds will be allocated to heavy dirtiers. So when * dirty pages grow large, heavy dirtiers will be throttled first, which will * effectively curb the growth of dirty pages. Light dirtiers with high enough * dirty threshold may never get throttled. |
3e26c149c
|
263 |
*/ |
16c4042f0
|
264 265 |
static unsigned long task_dirty_limit(struct task_struct *tsk, unsigned long bdi_dirty) |
3e26c149c
|
266 267 |
{ long numerator, denominator; |
16c4042f0
|
268 |
unsigned long dirty = bdi_dirty; |
3e26c149c
|
269 270 271 272 273 274 275 |
u64 inv = dirty >> 3; task_dirties_fraction(tsk, &numerator, &denominator); inv *= numerator; do_div(inv, denominator); dirty -= inv; |
3e26c149c
|
276 |
|
16c4042f0
|
277 |
return max(dirty, bdi_dirty/2); |
3e26c149c
|
278 |
} |
04fbfdc14
|
279 |
/* |
189d3c4a9
|
280 281 |
* */ |
189d3c4a9
|
282 283 284 285 286 |
static unsigned int bdi_min_ratio; int bdi_set_min_ratio(struct backing_dev_info *bdi, unsigned int min_ratio) { int ret = 0; |
189d3c4a9
|
287 |
|
cfc4ba536
|
288 |
spin_lock_bh(&bdi_lock); |
a42dde041
|
289 |
if (min_ratio > bdi->max_ratio) { |
189d3c4a9
|
290 |
ret = -EINVAL; |
a42dde041
|
291 292 293 294 295 296 297 298 299 |
} else { min_ratio -= bdi->min_ratio; if (bdi_min_ratio + min_ratio < 100) { bdi_min_ratio += min_ratio; bdi->min_ratio += min_ratio; } else { ret = -EINVAL; } } |
cfc4ba536
|
300 |
spin_unlock_bh(&bdi_lock); |
a42dde041
|
301 302 303 304 305 306 |
return ret; } int bdi_set_max_ratio(struct backing_dev_info *bdi, unsigned max_ratio) { |
a42dde041
|
307 308 309 310 |
int ret = 0; if (max_ratio > 100) return -EINVAL; |
cfc4ba536
|
311 |
spin_lock_bh(&bdi_lock); |
a42dde041
|
312 313 314 315 316 317 |
if (bdi->min_ratio > max_ratio) { ret = -EINVAL; } else { bdi->max_ratio = max_ratio; bdi->max_prop_frac = (PROP_FRAC_BASE * max_ratio) / 100; } |
cfc4ba536
|
318 |
spin_unlock_bh(&bdi_lock); |
189d3c4a9
|
319 320 321 |
return ret; } |
a42dde041
|
322 |
EXPORT_SYMBOL(bdi_set_max_ratio); |
189d3c4a9
|
323 324 |
/* |
1da177e4c
|
325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 |
* Work out the current dirty-memory clamping and background writeout * thresholds. * * The main aim here is to lower them aggressively if there is a lot of mapped * memory around. To avoid stressing page reclaim with lots of unreclaimable * pages. It is better to clamp down on writers than to start swapping, and * performing lots of scanning. * * We only allow 1/2 of the currently-unmapped memory to be dirtied. * * We don't permit the clamping level to fall below 5% - that is getting rather * excessive. * * We make sure that the background writeout level is below the adjusted * clamping level. */ |
1b4244647
|
341 342 343 344 345 346 |
static unsigned long highmem_dirtyable_memory(unsigned long total) { #ifdef CONFIG_HIGHMEM int node; unsigned long x = 0; |
37b07e416
|
347 |
for_each_node_state(node, N_HIGH_MEMORY) { |
1b4244647
|
348 349 |
struct zone *z = &NODE_DATA(node)->node_zones[ZONE_HIGHMEM]; |
adea02a1b
|
350 351 |
x += zone_page_state(z, NR_FREE_PAGES) + zone_reclaimable_pages(z); |
1b4244647
|
352 353 354 355 356 357 358 359 360 361 362 363 |
} /* * Make sure that the number of highmem pages is never larger * than the number of the total dirtyable memory. This can only * occur in very strange VM situations but we want to make sure * that this does not occur. */ return min(x, total); #else return 0; #endif } |
3eefae994
|
364 365 366 367 368 369 370 |
/** * determine_dirtyable_memory - amount of memory that may be used * * Returns the numebr of pages that can currently be freed and used * by the kernel for direct mappings. */ unsigned long determine_dirtyable_memory(void) |
1b4244647
|
371 372 |
{ unsigned long x; |
adea02a1b
|
373 |
x = global_page_state(NR_FREE_PAGES) + global_reclaimable_pages(); |
195cf453d
|
374 375 376 |
if (!vm_highmem_is_dirtyable) x -= highmem_dirtyable_memory(x); |
1b4244647
|
377 378 |
return x + 1; /* Ensure that we never return 0 */ } |
03ab450f0
|
379 |
/* |
1babe1838
|
380 381 382 383 384 385 386 387 |
* global_dirty_limits - background-writeback and dirty-throttling thresholds * * Calculate the dirty thresholds based on sysctl parameters * - vm.dirty_background_ratio or vm.dirty_background_bytes * - vm.dirty_ratio or vm.dirty_bytes * The dirty limits will be lifted by 1/4 for PF_LESS_THROTTLE (ie. nfsd) and * runtime tasks. */ |
16c4042f0
|
388 |
void global_dirty_limits(unsigned long *pbackground, unsigned long *pdirty) |
1da177e4c
|
389 |
{ |
364aeb284
|
390 391 |
unsigned long background; unsigned long dirty; |
1b4244647
|
392 |
unsigned long available_memory = determine_dirtyable_memory(); |
1da177e4c
|
393 |
struct task_struct *tsk; |
2da02997e
|
394 395 396 397 398 399 400 401 402 403 |
if (vm_dirty_bytes) dirty = DIV_ROUND_UP(vm_dirty_bytes, PAGE_SIZE); else { int dirty_ratio; dirty_ratio = vm_dirty_ratio; if (dirty_ratio < 5) dirty_ratio = 5; dirty = (dirty_ratio * available_memory) / 100; } |
1da177e4c
|
404 |
|
2da02997e
|
405 406 407 408 |
if (dirty_background_bytes) background = DIV_ROUND_UP(dirty_background_bytes, PAGE_SIZE); else background = (dirty_background_ratio * available_memory) / 100; |
1da177e4c
|
409 |
|
2da02997e
|
410 411 |
if (background >= dirty) background = dirty / 2; |
1da177e4c
|
412 413 414 415 416 417 418 |
tsk = current; if (tsk->flags & PF_LESS_THROTTLE || rt_task(tsk)) { background += background / 4; dirty += dirty / 4; } *pbackground = background; *pdirty = dirty; |
16c4042f0
|
419 |
} |
04fbfdc14
|
420 |
|
03ab450f0
|
421 |
/* |
1babe1838
|
422 423 424 425 426 427 428 429 430 431 |
* bdi_dirty_limit - @bdi's share of dirty throttling threshold * * Allocate high/low dirty limits to fast/slow devices, in order to prevent * - starving fast devices * - piling up dirty pages (that will take long time to sync) on slow devices * * The bdi's share of dirty limit will be adapting to its throughput and * bounded by the bdi->min_ratio and/or bdi->max_ratio parameters, if set. */ unsigned long bdi_dirty_limit(struct backing_dev_info *bdi, unsigned long dirty) |
16c4042f0
|
432 433 434 |
{ u64 bdi_dirty; long numerator, denominator; |
04fbfdc14
|
435 |
|
16c4042f0
|
436 437 438 439 |
/* * Calculate this BDI's share of the dirty ratio. */ bdi_writeout_fraction(bdi, &numerator, &denominator); |
04fbfdc14
|
440 |
|
16c4042f0
|
441 442 443 |
bdi_dirty = (dirty * (100 - bdi_min_ratio)) / 100; bdi_dirty *= numerator; do_div(bdi_dirty, denominator); |
04fbfdc14
|
444 |
|
16c4042f0
|
445 446 447 448 449 |
bdi_dirty += (dirty * bdi->min_ratio) / 100; if (bdi_dirty > (dirty * bdi->max_ratio) / 100) bdi_dirty = dirty * bdi->max_ratio / 100; return bdi_dirty; |
1da177e4c
|
450 451 452 453 454 455 |
} /* * balance_dirty_pages() must be called by processes which are generating dirty * data. It looks at the number of dirty pages in the machine and will force * the caller to perform writeback if the system is over `vm_dirty_ratio'. |
5b0830cb9
|
456 457 |
* If we're over `background_thresh' then the writeback threads are woken to * perform some writeout. |
1da177e4c
|
458 |
*/ |
3a2e9a5a2
|
459 460 |
static void balance_dirty_pages(struct address_space *mapping, unsigned long write_chunk) |
1da177e4c
|
461 |
{ |
5fce25a9d
|
462 463 |
long nr_reclaimable, bdi_nr_reclaimable; long nr_writeback, bdi_nr_writeback; |
364aeb284
|
464 465 466 |
unsigned long background_thresh; unsigned long dirty_thresh; unsigned long bdi_thresh; |
1da177e4c
|
467 |
unsigned long pages_written = 0; |
87c6a9b25
|
468 |
unsigned long pause = 1; |
e50e37201
|
469 |
bool dirty_exceeded = false; |
1da177e4c
|
470 471 472 473 |
struct backing_dev_info *bdi = mapping->backing_dev_info; for (;;) { struct writeback_control wbc = { |
1da177e4c
|
474 475 476 |
.sync_mode = WB_SYNC_NONE, .older_than_this = NULL, .nr_to_write = write_chunk, |
111ebb6e6
|
477 |
.range_cyclic = 1, |
1da177e4c
|
478 |
}; |
5fce25a9d
|
479 480 481 |
nr_reclaimable = global_page_state(NR_FILE_DIRTY) + global_page_state(NR_UNSTABLE_NFS); nr_writeback = global_page_state(NR_WRITEBACK); |
16c4042f0
|
482 483 484 485 486 487 488 489 490 491 492 493 494 |
global_dirty_limits(&background_thresh, &dirty_thresh); /* * Throttle it only when the background writeback cannot * catch-up. This avoids (excessively) small writeouts * when the bdi limits are ramping up. */ if (nr_reclaimable + nr_writeback < (background_thresh + dirty_thresh) / 2) break; bdi_thresh = bdi_dirty_limit(bdi, dirty_thresh); bdi_thresh = task_dirty_limit(current, bdi_thresh); |
e50e37201
|
495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 |
/* * In order to avoid the stacked BDI deadlock we need * to ensure we accurately count the 'dirty' pages when * the threshold is low. * * Otherwise it would be possible to get thresh+n pages * reported dirty, even though there are thresh-m pages * actually dirty; with m+n sitting in the percpu * deltas. */ if (bdi_thresh < 2*bdi_stat_error(bdi)) { bdi_nr_reclaimable = bdi_stat_sum(bdi, BDI_RECLAIMABLE); bdi_nr_writeback = bdi_stat_sum(bdi, BDI_WRITEBACK); } else { bdi_nr_reclaimable = bdi_stat(bdi, BDI_RECLAIMABLE); bdi_nr_writeback = bdi_stat(bdi, BDI_WRITEBACK); } |
5fce25a9d
|
512 |
|
e50e37201
|
513 514 515 516 517 518 519 520 521 522 523 |
/* * The bdi thresh is somehow "soft" limit derived from the * global "hard" limit. The former helps to prevent heavy IO * bdi or process from holding back light ones; The latter is * the last resort safeguard. */ dirty_exceeded = (bdi_nr_reclaimable + bdi_nr_writeback >= bdi_thresh) || (nr_reclaimable + nr_writeback >= dirty_thresh); if (!dirty_exceeded) |
04fbfdc14
|
524 |
break; |
1da177e4c
|
525 |
|
04fbfdc14
|
526 527 |
if (!bdi->dirty_exceeded) bdi->dirty_exceeded = 1; |
1da177e4c
|
528 529 530 531 532 533 |
/* Note: nr_reclaimable denotes nr_dirty + nr_unstable. * Unstable writes are a feature of certain networked * filesystems (i.e. NFS) in which data may have been * written to the server's write cache, but has not yet * been flushed to permanent storage. |
d7831a0bd
|
534 535 536 |
* Only move pages to writeback if this bdi is over its * threshold otherwise wait until the disk writes catch * up. |
1da177e4c
|
537 |
*/ |
028c2dd18
|
538 |
trace_wbc_balance_dirty_start(&wbc, bdi); |
d7831a0bd
|
539 |
if (bdi_nr_reclaimable > bdi_thresh) { |
9c3a8ee8a
|
540 |
writeback_inodes_wb(&bdi->wb, &wbc); |
1da177e4c
|
541 |
pages_written += write_chunk - wbc.nr_to_write; |
028c2dd18
|
542 |
trace_wbc_balance_dirty_written(&wbc, bdi); |
e50e37201
|
543 544 |
if (pages_written >= write_chunk) break; /* We've done our duty */ |
04fbfdc14
|
545 |
} |
028c2dd18
|
546 |
trace_wbc_balance_dirty_wait(&wbc, bdi); |
d25105e89
|
547 548 |
__set_current_state(TASK_INTERRUPTIBLE); io_schedule_timeout(pause); |
87c6a9b25
|
549 550 551 552 553 554 555 556 |
/* * Increase the delay for each loop, up to our previous * default of taking a 100ms nap. */ pause <<= 1; if (pause > HZ / 10) pause = HZ / 10; |
1da177e4c
|
557 |
} |
e50e37201
|
558 |
if (!dirty_exceeded && bdi->dirty_exceeded) |
04fbfdc14
|
559 |
bdi->dirty_exceeded = 0; |
1da177e4c
|
560 561 |
if (writeback_in_progress(bdi)) |
5b0830cb9
|
562 |
return; |
1da177e4c
|
563 564 565 566 567 568 569 570 571 572 |
/* * In laptop mode, we wait until hitting the higher threshold before * starting background writeout, and then write out all the way down * to the lower threshold. So slow writers cause minimal disk activity. * * In normal mode, we start background writeout at the lower * background_thresh, to keep the amount of dirty memory low. */ if ((laptop_mode && pages_written) || |
e50e37201
|
573 |
(!laptop_mode && (nr_reclaimable > background_thresh))) |
c5444198c
|
574 |
bdi_start_background_writeback(bdi); |
1da177e4c
|
575 |
} |
a200ee182
|
576 |
void set_page_dirty_balance(struct page *page, int page_mkwrite) |
edc79b2a4
|
577 |
{ |
a200ee182
|
578 |
if (set_page_dirty(page) || page_mkwrite) { |
edc79b2a4
|
579 580 581 582 583 584 |
struct address_space *mapping = page_mapping(page); if (mapping) balance_dirty_pages_ratelimited(mapping); } } |
245b2e70e
|
585 |
static DEFINE_PER_CPU(unsigned long, bdp_ratelimits) = 0; |
1da177e4c
|
586 |
/** |
fa5a734e4
|
587 |
* balance_dirty_pages_ratelimited_nr - balance dirty memory state |
67be2dd1b
|
588 |
* @mapping: address_space which was dirtied |
a580290c3
|
589 |
* @nr_pages_dirtied: number of pages which the caller has just dirtied |
1da177e4c
|
590 591 592 593 594 595 596 597 598 599 |
* * Processes which are dirtying memory should call in here once for each page * which was newly dirtied. The function will periodically check the system's * dirty state and will initiate writeback if needed. * * On really big machines, get_writeback_state is expensive, so try to avoid * calling it too often (ratelimiting). But once we're over the dirty memory * limit we decrease the ratelimiting by a lot, to prevent individual processes * from overshooting the limit by (ratelimit_pages) each. */ |
fa5a734e4
|
600 601 |
void balance_dirty_pages_ratelimited_nr(struct address_space *mapping, unsigned long nr_pages_dirtied) |
1da177e4c
|
602 |
{ |
fa5a734e4
|
603 604 |
unsigned long ratelimit; unsigned long *p; |
1da177e4c
|
605 606 |
ratelimit = ratelimit_pages; |
04fbfdc14
|
607 |
if (mapping->backing_dev_info->dirty_exceeded) |
1da177e4c
|
608 609 610 611 612 613 |
ratelimit = 8; /* * Check the rate limiting. Also, we do not want to throttle real-time * tasks in balance_dirty_pages(). Period. */ |
fa5a734e4
|
614 |
preempt_disable(); |
245b2e70e
|
615 |
p = &__get_cpu_var(bdp_ratelimits); |
fa5a734e4
|
616 617 |
*p += nr_pages_dirtied; if (unlikely(*p >= ratelimit)) { |
3a2e9a5a2
|
618 |
ratelimit = sync_writeback_pages(*p); |
fa5a734e4
|
619 620 |
*p = 0; preempt_enable(); |
3a2e9a5a2
|
621 |
balance_dirty_pages(mapping, ratelimit); |
1da177e4c
|
622 623 |
return; } |
fa5a734e4
|
624 |
preempt_enable(); |
1da177e4c
|
625 |
} |
fa5a734e4
|
626 |
EXPORT_SYMBOL(balance_dirty_pages_ratelimited_nr); |
1da177e4c
|
627 |
|
232ea4d69
|
628 |
void throttle_vm_writeout(gfp_t gfp_mask) |
1da177e4c
|
629 |
{ |
364aeb284
|
630 631 |
unsigned long background_thresh; unsigned long dirty_thresh; |
1da177e4c
|
632 633 |
for ( ; ; ) { |
16c4042f0
|
634 |
global_dirty_limits(&background_thresh, &dirty_thresh); |
1da177e4c
|
635 636 637 638 639 640 |
/* * Boost the allowable dirty threshold a bit for page * allocators so they don't get DoS'ed by heavy writers */ dirty_thresh += dirty_thresh / 10; /* wheeee... */ |
c24f21bda
|
641 642 643 |
if (global_page_state(NR_UNSTABLE_NFS) + global_page_state(NR_WRITEBACK) <= dirty_thresh) break; |
8aa7e847d
|
644 |
congestion_wait(BLK_RW_ASYNC, HZ/10); |
369f2389e
|
645 646 647 648 649 650 651 652 |
/* * The caller might hold locks which can prevent IO completion * or progress in the filesystem. So we cannot just sit here * waiting for IO to complete. */ if ((gfp_mask & (__GFP_FS|__GFP_IO)) != (__GFP_FS|__GFP_IO)) break; |
1da177e4c
|
653 654 |
} } |
1da177e4c
|
655 |
/* |
1da177e4c
|
656 657 658 |
* sysctl handler for /proc/sys/vm/dirty_writeback_centisecs */ int dirty_writeback_centisecs_handler(ctl_table *table, int write, |
8d65af789
|
659 |
void __user *buffer, size_t *length, loff_t *ppos) |
1da177e4c
|
660 |
{ |
8d65af789
|
661 |
proc_dointvec(table, write, buffer, length, ppos); |
6423104b6
|
662 |
bdi_arm_supers_timer(); |
1da177e4c
|
663 664 |
return 0; } |
c2c4986ed
|
665 |
#ifdef CONFIG_BLOCK |
31373d09d
|
666 |
void laptop_mode_timer_fn(unsigned long data) |
1da177e4c
|
667 |
{ |
31373d09d
|
668 669 670 |
struct request_queue *q = (struct request_queue *)data; int nr_pages = global_page_state(NR_FILE_DIRTY) + global_page_state(NR_UNSTABLE_NFS); |
1da177e4c
|
671 |
|
31373d09d
|
672 673 674 675 |
/* * We want to write everything out, not just down to the dirty * threshold */ |
31373d09d
|
676 |
if (bdi_has_dirty_io(&q->backing_dev_info)) |
c5444198c
|
677 |
bdi_start_writeback(&q->backing_dev_info, nr_pages); |
1da177e4c
|
678 679 680 681 682 683 684 |
} /* * We've spun up the disk and we're in laptop mode: schedule writeback * of all dirty data a few seconds from now. If the flush is already scheduled * then push it back - the user is still using the disk. */ |
31373d09d
|
685 |
void laptop_io_completion(struct backing_dev_info *info) |
1da177e4c
|
686 |
{ |
31373d09d
|
687 |
mod_timer(&info->laptop_mode_wb_timer, jiffies + laptop_mode); |
1da177e4c
|
688 689 690 691 692 693 694 695 696 |
} /* * We're in laptop mode and we've just synced. The sync's writes will have * caused another writeback to be scheduled by laptop_io_completion. * Nothing needs to be written back anymore, so we unschedule the writeback. */ void laptop_sync_completion(void) { |
31373d09d
|
697 698 699 700 701 702 703 704 |
struct backing_dev_info *bdi; rcu_read_lock(); list_for_each_entry_rcu(bdi, &bdi_list, bdi_list) del_timer(&bdi->laptop_mode_wb_timer); rcu_read_unlock(); |
1da177e4c
|
705 |
} |
c2c4986ed
|
706 |
#endif |
1da177e4c
|
707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 |
/* * If ratelimit_pages is too high then we can get into dirty-data overload * if a large number of processes all perform writes at the same time. * If it is too low then SMP machines will call the (expensive) * get_writeback_state too often. * * Here we set ratelimit_pages to a level which ensures that when all CPUs are * dirtying in parallel, we cannot go more than 3% (1/32) over the dirty memory * thresholds before writeback cuts in. * * But the limit should not be set too high. Because it also controls the * amount of memory which the balance_dirty_pages() caller has to write back. * If this is too large then the caller will block on the IO queue all the * time. So limit it to four megabytes - the balance_dirty_pages() caller * will write six megabyte chunks, max. */ |
2d1d43f6a
|
724 |
void writeback_set_ratelimit(void) |
1da177e4c
|
725 |
{ |
40c99aae2
|
726 |
ratelimit_pages = vm_total_pages / (num_online_cpus() * 32); |
1da177e4c
|
727 728 729 730 731 |
if (ratelimit_pages < 16) ratelimit_pages = 16; if (ratelimit_pages * PAGE_CACHE_SIZE > 4096 * 1024) ratelimit_pages = (4096 * 1024) / PAGE_CACHE_SIZE; } |
26c2143b6
|
732 |
static int __cpuinit |
1da177e4c
|
733 734 |
ratelimit_handler(struct notifier_block *self, unsigned long u, void *v) { |
2d1d43f6a
|
735 |
writeback_set_ratelimit(); |
aa0f03037
|
736 |
return NOTIFY_DONE; |
1da177e4c
|
737 |
} |
74b85f379
|
738 |
static struct notifier_block __cpuinitdata ratelimit_nb = { |
1da177e4c
|
739 740 741 742 743 |
.notifier_call = ratelimit_handler, .next = NULL, }; /* |
dc6e29da9
|
744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 |
* Called early on to tune the page writeback dirty limits. * * We used to scale dirty pages according to how total memory * related to pages that could be allocated for buffers (by * comparing nr_free_buffer_pages() to vm_total_pages. * * However, that was when we used "dirty_ratio" to scale with * all memory, and we don't do that any more. "dirty_ratio" * is now applied to total non-HIGHPAGE memory (by subtracting * totalhigh_pages from vm_total_pages), and as such we can't * get into the old insane situation any more where we had * large amounts of dirty pages compared to a small amount of * non-HIGHMEM memory. * * But we might still want to scale the dirty_ratio by how * much memory the box has.. |
1da177e4c
|
760 761 762 |
*/ void __init page_writeback_init(void) { |
04fbfdc14
|
763 |
int shift; |
2d1d43f6a
|
764 |
writeback_set_ratelimit(); |
1da177e4c
|
765 |
register_cpu_notifier(&ratelimit_nb); |
04fbfdc14
|
766 767 768 |
shift = calc_period_shift(); prop_descriptor_init(&vm_completions, shift); |
3e26c149c
|
769 |
prop_descriptor_init(&vm_dirties, shift); |
1da177e4c
|
770 |
} |
811d736f9
|
771 |
/** |
f446daaea
|
772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 |
* tag_pages_for_writeback - tag pages to be written by write_cache_pages * @mapping: address space structure to write * @start: starting page index * @end: ending page index (inclusive) * * This function scans the page range from @start to @end (inclusive) and tags * all pages that have DIRTY tag set with a special TOWRITE tag. The idea is * that write_cache_pages (or whoever calls this function) will then use * TOWRITE tag to identify pages eligible for writeback. This mechanism is * used to avoid livelocking of writeback by a process steadily creating new * dirty pages in the file (thus it is important for this function to be quick * so that it can tag pages faster than a dirtying process can create them). */ /* * We tag pages in batches of WRITEBACK_TAG_BATCH to reduce tree_lock latency. */ |
f446daaea
|
788 789 790 |
void tag_pages_for_writeback(struct address_space *mapping, pgoff_t start, pgoff_t end) { |
3c111a071
|
791 |
#define WRITEBACK_TAG_BATCH 4096 |
f446daaea
|
792 793 794 795 796 797 798 799 800 801 |
unsigned long tagged; do { spin_lock_irq(&mapping->tree_lock); tagged = radix_tree_range_tag_if_tagged(&mapping->page_tree, &start, end, WRITEBACK_TAG_BATCH, PAGECACHE_TAG_DIRTY, PAGECACHE_TAG_TOWRITE); spin_unlock_irq(&mapping->tree_lock); WARN_ON_ONCE(tagged > WRITEBACK_TAG_BATCH); cond_resched(); |
d5ed3a4af
|
802 803 |
/* We check 'start' to handle wrapping when end == ~0UL */ } while (tagged >= WRITEBACK_TAG_BATCH && start); |
f446daaea
|
804 805 806 807 |
} EXPORT_SYMBOL(tag_pages_for_writeback); /** |
0ea971801
|
808 |
* write_cache_pages - walk the list of dirty pages of the given address space and write all of them. |
811d736f9
|
809 810 |
* @mapping: address space structure to write * @wbc: subtract the number of written pages from *@wbc->nr_to_write |
0ea971801
|
811 812 |
* @writepage: function called for each page * @data: data passed to writepage function |
811d736f9
|
813 |
* |
0ea971801
|
814 |
* If a page is already under I/O, write_cache_pages() skips it, even |
811d736f9
|
815 816 817 818 819 820 |
* if it's dirty. This is desirable behaviour for memory-cleaning writeback, * but it is INCORRECT for data-integrity system calls such as fsync(). fsync() * and msync() need to guarantee that all the data which was dirty at the time * the call was made get new I/O started against them. If wbc->sync_mode is * WB_SYNC_ALL then we were called for data integrity and we must wait for * existing IO to complete. |
f446daaea
|
821 822 823 824 825 826 827 |
* * To avoid livelocks (when other process dirties new pages), we first tag * pages which should be written back with TOWRITE tag and only then start * writing them. For data-integrity sync we have to be careful so that we do * not miss some pages (e.g., because some other process has cleared TOWRITE * tag we set). The rule we follow is that TOWRITE tag can be cleared only * by the process clearing the DIRTY tag (and submitting the page for IO). |
811d736f9
|
828 |
*/ |
0ea971801
|
829 830 831 |
int write_cache_pages(struct address_space *mapping, struct writeback_control *wbc, writepage_t writepage, void *data) |
811d736f9
|
832 |
{ |
811d736f9
|
833 834 |
int ret = 0; int done = 0; |
811d736f9
|
835 836 |
struct pagevec pvec; int nr_pages; |
31a12666d
|
837 |
pgoff_t uninitialized_var(writeback_index); |
811d736f9
|
838 839 |
pgoff_t index; pgoff_t end; /* Inclusive */ |
bd19e012f
|
840 |
pgoff_t done_index; |
31a12666d
|
841 |
int cycled; |
811d736f9
|
842 |
int range_whole = 0; |
f446daaea
|
843 |
int tag; |
811d736f9
|
844 |
|
811d736f9
|
845 846 |
pagevec_init(&pvec, 0); if (wbc->range_cyclic) { |
31a12666d
|
847 848 849 850 851 852 |
writeback_index = mapping->writeback_index; /* prev offset */ index = writeback_index; if (index == 0) cycled = 1; else cycled = 0; |
811d736f9
|
853 854 855 856 857 858 |
end = -1; } else { index = wbc->range_start >> PAGE_CACHE_SHIFT; end = wbc->range_end >> PAGE_CACHE_SHIFT; if (wbc->range_start == 0 && wbc->range_end == LLONG_MAX) range_whole = 1; |
31a12666d
|
859 |
cycled = 1; /* ignore range_cyclic tests */ |
811d736f9
|
860 |
} |
f446daaea
|
861 862 863 864 |
if (wbc->sync_mode == WB_SYNC_ALL) tag = PAGECACHE_TAG_TOWRITE; else tag = PAGECACHE_TAG_DIRTY; |
811d736f9
|
865 |
retry: |
f446daaea
|
866 867 |
if (wbc->sync_mode == WB_SYNC_ALL) tag_pages_for_writeback(mapping, index, end); |
bd19e012f
|
868 |
done_index = index; |
5a3d5c981
|
869 870 |
while (!done && (index <= end)) { int i; |
f446daaea
|
871 |
nr_pages = pagevec_lookup_tag(&pvec, mapping, &index, tag, |
5a3d5c981
|
872 873 874 |
min(end - index, (pgoff_t)PAGEVEC_SIZE-1) + 1); if (nr_pages == 0) break; |
811d736f9
|
875 |
|
811d736f9
|
876 877 878 879 |
for (i = 0; i < nr_pages; i++) { struct page *page = pvec.pages[i]; /* |
d5482cdf8
|
880 881 882 883 884 |
* At this point, the page may be truncated or * invalidated (changing page->mapping to NULL), or * even swizzled back from swapper_space to tmpfs file * mapping. However, page->index will not change * because we have a reference on the page. |
811d736f9
|
885 |
*/ |
d5482cdf8
|
886 887 888 889 890 891 892 893 894 895 |
if (page->index > end) { /* * can't be range_cyclic (1st pass) because * end == -1 in that case. */ done = 1; break; } done_index = page->index + 1; |
811d736f9
|
896 |
lock_page(page); |
5a3d5c981
|
897 898 899 900 901 902 903 904 |
/* * Page truncated or invalidated. We can freely skip it * then, even for data integrity operations: the page * has disappeared concurrently, so there could be no * real expectation of this data interity operation * even if there is now a new, dirty page at the same * pagecache address. */ |
811d736f9
|
905 |
if (unlikely(page->mapping != mapping)) { |
5a3d5c981
|
906 |
continue_unlock: |
811d736f9
|
907 908 909 |
unlock_page(page); continue; } |
515f4a037
|
910 911 912 913 914 915 916 917 918 919 920 |
if (!PageDirty(page)) { /* someone wrote it for us */ goto continue_unlock; } if (PageWriteback(page)) { if (wbc->sync_mode != WB_SYNC_NONE) wait_on_page_writeback(page); else goto continue_unlock; } |
811d736f9
|
921 |
|
515f4a037
|
922 923 |
BUG_ON(PageWriteback(page)); if (!clear_page_dirty_for_io(page)) |
5a3d5c981
|
924 |
goto continue_unlock; |
811d736f9
|
925 |
|
9e094383b
|
926 |
trace_wbc_writepage(wbc, mapping->backing_dev_info); |
0ea971801
|
927 |
ret = (*writepage)(page, wbc, data); |
00266770b
|
928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 |
if (unlikely(ret)) { if (ret == AOP_WRITEPAGE_ACTIVATE) { unlock_page(page); ret = 0; } else { /* * done_index is set past this page, * so media errors will not choke * background writeout for the entire * file. This has consequences for * range_cyclic semantics (ie. it may * not be suitable for data integrity * writeout). */ done = 1; break; } |
0b5649278
|
945 |
} |
00266770b
|
946 |
|
546a19242
|
947 948 949 950 951 952 953 954 955 956 |
/* * We stop writing back only if we are not doing * integrity sync. In case of integrity sync we have to * keep going until we have written all the pages * we tagged for writeback prior to entering this loop. */ if (--wbc->nr_to_write <= 0 && wbc->sync_mode == WB_SYNC_NONE) { done = 1; break; |
05fe478dd
|
957 |
} |
811d736f9
|
958 959 960 961 |
} pagevec_release(&pvec); cond_resched(); } |
3a4c6800f
|
962 |
if (!cycled && !done) { |
811d736f9
|
963 |
/* |
31a12666d
|
964 |
* range_cyclic: |
811d736f9
|
965 966 967 |
* We hit the last page and there is more work to be done: wrap * back to the start of the file */ |
31a12666d
|
968 |
cycled = 1; |
811d736f9
|
969 |
index = 0; |
31a12666d
|
970 |
end = writeback_index - 1; |
811d736f9
|
971 972 |
goto retry; } |
0b5649278
|
973 974 |
if (wbc->range_cyclic || (range_whole && wbc->nr_to_write > 0)) mapping->writeback_index = done_index; |
06d6cf695
|
975 |
|
811d736f9
|
976 977 |
return ret; } |
0ea971801
|
978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 |
EXPORT_SYMBOL(write_cache_pages); /* * Function used by generic_writepages to call the real writepage * function and set the mapping flags on error */ static int __writepage(struct page *page, struct writeback_control *wbc, void *data) { struct address_space *mapping = data; int ret = mapping->a_ops->writepage(page, wbc); mapping_set_error(mapping, ret); return ret; } /** * generic_writepages - walk the list of dirty pages of the given address space and writepage() all of them. * @mapping: address space structure to write * @wbc: subtract the number of written pages from *@wbc->nr_to_write * * This is a library function, which implements the writepages() * address_space_operation. */ int generic_writepages(struct address_space *mapping, struct writeback_control *wbc) { /* deal with chardevs and other special file */ if (!mapping->a_ops->writepage) return 0; return write_cache_pages(mapping, wbc, __writepage, mapping); } |
811d736f9
|
1010 1011 |
EXPORT_SYMBOL(generic_writepages); |
1da177e4c
|
1012 1013 |
int do_writepages(struct address_space *mapping, struct writeback_control *wbc) { |
22905f775
|
1014 |
int ret; |
1da177e4c
|
1015 1016 1017 |
if (wbc->nr_to_write <= 0) return 0; if (mapping->a_ops->writepages) |
d08b3851d
|
1018 |
ret = mapping->a_ops->writepages(mapping, wbc); |
22905f775
|
1019 1020 |
else ret = generic_writepages(mapping, wbc); |
22905f775
|
1021 |
return ret; |
1da177e4c
|
1022 1023 1024 1025 |
} /** * write_one_page - write out a single page and optionally wait on I/O |
67be2dd1b
|
1026 1027 |
* @page: the page to write * @wait: if true, wait on writeout |
1da177e4c
|
1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 |
* * The page must be locked by the caller and will be unlocked upon return. * * write_one_page() returns a negative error code if I/O failed. */ int write_one_page(struct page *page, int wait) { struct address_space *mapping = page->mapping; int ret = 0; struct writeback_control wbc = { .sync_mode = WB_SYNC_ALL, .nr_to_write = 1, }; BUG_ON(!PageLocked(page)); if (wait) wait_on_page_writeback(page); if (clear_page_dirty_for_io(page)) { page_cache_get(page); ret = mapping->a_ops->writepage(page, &wbc); if (ret == 0 && wait) { wait_on_page_writeback(page); if (PageError(page)) ret = -EIO; } page_cache_release(page); } else { unlock_page(page); } return ret; } EXPORT_SYMBOL(write_one_page); /* |
767193253
|
1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 |
* For address_spaces which do not use buffers nor write back. */ int __set_page_dirty_no_writeback(struct page *page) { if (!PageDirty(page)) SetPageDirty(page); return 0; } /* |
e3a7cca1e
|
1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 |
* Helper function for set_page_dirty family. * NOTE: This relies on being atomic wrt interrupts. */ void account_page_dirtied(struct page *page, struct address_space *mapping) { if (mapping_cap_account_dirty(mapping)) { __inc_zone_page_state(page, NR_FILE_DIRTY); __inc_bdi_stat(mapping->backing_dev_info, BDI_RECLAIMABLE); task_dirty_inc(current); task_io_account_write(PAGE_CACHE_SIZE); } } |
679ceace8
|
1086 |
EXPORT_SYMBOL(account_page_dirtied); |
e3a7cca1e
|
1087 1088 |
/* |
1da177e4c
|
1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 |
* For address_spaces which do not use buffers. Just tag the page as dirty in * its radix tree. * * This is also used when a single buffer is being dirtied: we want to set the * page dirty in that case, but not all the buffers. This is a "bottom-up" * dirtying, whereas __set_page_dirty_buffers() is a "top-down" dirtying. * * Most callers have locked the page, which pins the address_space in memory. * But zap_pte_range() does not lock the page, however in that case the * mapping is pinned by the vma's ->vm_file reference. * * We take care to handle the case where the page was truncated from the |
183ff22bb
|
1101 |
* mapping by re-checking page_mapping() inside tree_lock. |
1da177e4c
|
1102 1103 1104 |
*/ int __set_page_dirty_nobuffers(struct page *page) { |
1da177e4c
|
1105 1106 1107 |
if (!TestSetPageDirty(page)) { struct address_space *mapping = page_mapping(page); struct address_space *mapping2; |
8c08540f8
|
1108 1109 |
if (!mapping) return 1; |
19fd62312
|
1110 |
spin_lock_irq(&mapping->tree_lock); |
8c08540f8
|
1111 1112 1113 |
mapping2 = page_mapping(page); if (mapping2) { /* Race with truncate? */ BUG_ON(mapping2 != mapping); |
787d2214c
|
1114 |
WARN_ON_ONCE(!PagePrivate(page) && !PageUptodate(page)); |
e3a7cca1e
|
1115 |
account_page_dirtied(page, mapping); |
8c08540f8
|
1116 1117 1118 |
radix_tree_tag_set(&mapping->page_tree, page_index(page), PAGECACHE_TAG_DIRTY); } |
19fd62312
|
1119 |
spin_unlock_irq(&mapping->tree_lock); |
8c08540f8
|
1120 1121 1122 |
if (mapping->host) { /* !PageAnon && !swapper_space */ __mark_inode_dirty(mapping->host, I_DIRTY_PAGES); |
1da177e4c
|
1123 |
} |
4741c9fd3
|
1124 |
return 1; |
1da177e4c
|
1125 |
} |
4741c9fd3
|
1126 |
return 0; |
1da177e4c
|
1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 |
} EXPORT_SYMBOL(__set_page_dirty_nobuffers); /* * When a writepage implementation decides that it doesn't want to write this * page for some reason, it should redirty the locked page via * redirty_page_for_writepage() and it should then unlock the page and return 0 */ int redirty_page_for_writepage(struct writeback_control *wbc, struct page *page) { wbc->pages_skipped++; return __set_page_dirty_nobuffers(page); } EXPORT_SYMBOL(redirty_page_for_writepage); /* |
6746aff74
|
1143 1144 1145 1146 1147 1148 1149 |
* Dirty a page. * * For pages with a mapping this should be done under the page lock * for the benefit of asynchronous memory errors who prefer a consistent * dirty state. This rule can be broken in some special cases, * but should be better not to. * |
1da177e4c
|
1150 1151 1152 |
* If the mapping doesn't provide a set_page_dirty a_op, then * just fall through and assume that it wants buffer_heads. */ |
1cf6e7d83
|
1153 |
int set_page_dirty(struct page *page) |
1da177e4c
|
1154 1155 1156 1157 1158 |
{ struct address_space *mapping = page_mapping(page); if (likely(mapping)) { int (*spd)(struct page *) = mapping->a_ops->set_page_dirty; |
9361401eb
|
1159 1160 1161 1162 1163 |
#ifdef CONFIG_BLOCK if (!spd) spd = __set_page_dirty_buffers; #endif return (*spd)(page); |
1da177e4c
|
1164 |
} |
4741c9fd3
|
1165 1166 1167 1168 |
if (!PageDirty(page)) { if (!TestSetPageDirty(page)) return 1; } |
1da177e4c
|
1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 |
return 0; } EXPORT_SYMBOL(set_page_dirty); /* * set_page_dirty() is racy if the caller has no reference against * page->mapping->host, and if the page is unlocked. This is because another * CPU could truncate the page off the mapping and then free the mapping. * * Usually, the page _is_ locked, or the caller is a user-space process which * holds a reference on the inode by having an open file. * * In other cases, the page should be locked before running set_page_dirty(). */ int set_page_dirty_lock(struct page *page) { int ret; |
db37648cd
|
1186 |
lock_page_nosync(page); |
1da177e4c
|
1187 1188 1189 1190 1191 1192 1193 |
ret = set_page_dirty(page); unlock_page(page); return ret; } EXPORT_SYMBOL(set_page_dirty_lock); /* |
1da177e4c
|
1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 |
* Clear a page's dirty flag, while caring for dirty memory accounting. * Returns true if the page was previously dirty. * * This is for preparing to put the page under writeout. We leave the page * tagged as dirty in the radix tree so that a concurrent write-for-sync * can discover it via a PAGECACHE_TAG_DIRTY walk. The ->writepage * implementation will run either set_page_writeback() or set_page_dirty(), * at which stage we bring the page's dirty flag and radix-tree dirty tag * back into sync. * * This incoherency between the page's dirty flag and radix-tree tag is * unfortunate, but it only exists while the page is locked. */ int clear_page_dirty_for_io(struct page *page) { struct address_space *mapping = page_mapping(page); |
79352894b
|
1210 |
BUG_ON(!PageLocked(page)); |
fe3cba17c
|
1211 |
ClearPageReclaim(page); |
7658cc289
|
1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 |
if (mapping && mapping_cap_account_dirty(mapping)) { /* * Yes, Virginia, this is indeed insane. * * We use this sequence to make sure that * (a) we account for dirty stats properly * (b) we tell the low-level filesystem to * mark the whole page dirty if it was * dirty in a pagetable. Only to then * (c) clean the page again and return 1 to * cause the writeback. * * This way we avoid all nasty races with the * dirty bit in multiple places and clearing * them concurrently from different threads. * * Note! Normally the "set_page_dirty(page)" * has no effect on the actual dirty bit - since * that will already usually be set. But we * need the side effects, and it can help us * avoid races. * * We basically use the page "master dirty bit" * as a serialization point for all the different * threads doing their things. |
7658cc289
|
1237 1238 1239 |
*/ if (page_mkclean(page)) set_page_dirty(page); |
79352894b
|
1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 |
/* * We carefully synchronise fault handlers against * installing a dirty pte and marking the page dirty * at this point. We do this by having them hold the * page lock at some point after installing their * pte, but before marking the page dirty. * Pages are always locked coming in here, so we get * the desired exclusion. See mm/memory.c:do_wp_page() * for more comments. */ |
7658cc289
|
1250 |
if (TestClearPageDirty(page)) { |
8c08540f8
|
1251 |
dec_zone_page_state(page, NR_FILE_DIRTY); |
c9e51e418
|
1252 1253 |
dec_bdi_stat(mapping->backing_dev_info, BDI_RECLAIMABLE); |
7658cc289
|
1254 |
return 1; |
1da177e4c
|
1255 |
} |
7658cc289
|
1256 |
return 0; |
1da177e4c
|
1257 |
} |
7658cc289
|
1258 |
return TestClearPageDirty(page); |
1da177e4c
|
1259 |
} |
58bb01a9c
|
1260 |
EXPORT_SYMBOL(clear_page_dirty_for_io); |
1da177e4c
|
1261 1262 1263 1264 1265 1266 1267 |
int test_clear_page_writeback(struct page *page) { struct address_space *mapping = page_mapping(page); int ret; if (mapping) { |
69cb51d18
|
1268 |
struct backing_dev_info *bdi = mapping->backing_dev_info; |
1da177e4c
|
1269 |
unsigned long flags; |
19fd62312
|
1270 |
spin_lock_irqsave(&mapping->tree_lock, flags); |
1da177e4c
|
1271 |
ret = TestClearPageWriteback(page); |
69cb51d18
|
1272 |
if (ret) { |
1da177e4c
|
1273 1274 1275 |
radix_tree_tag_clear(&mapping->page_tree, page_index(page), PAGECACHE_TAG_WRITEBACK); |
e4ad08fe6
|
1276 |
if (bdi_cap_account_writeback(bdi)) { |
69cb51d18
|
1277 |
__dec_bdi_stat(bdi, BDI_WRITEBACK); |
04fbfdc14
|
1278 1279 |
__bdi_writeout_inc(bdi); } |
69cb51d18
|
1280 |
} |
19fd62312
|
1281 |
spin_unlock_irqrestore(&mapping->tree_lock, flags); |
1da177e4c
|
1282 1283 1284 |
} else { ret = TestClearPageWriteback(page); } |
d688abf50
|
1285 1286 |
if (ret) dec_zone_page_state(page, NR_WRITEBACK); |
1da177e4c
|
1287 1288 1289 1290 1291 1292 1293 1294 1295 |
return ret; } int test_set_page_writeback(struct page *page) { struct address_space *mapping = page_mapping(page); int ret; if (mapping) { |
69cb51d18
|
1296 |
struct backing_dev_info *bdi = mapping->backing_dev_info; |
1da177e4c
|
1297 |
unsigned long flags; |
19fd62312
|
1298 |
spin_lock_irqsave(&mapping->tree_lock, flags); |
1da177e4c
|
1299 |
ret = TestSetPageWriteback(page); |
69cb51d18
|
1300 |
if (!ret) { |
1da177e4c
|
1301 1302 1303 |
radix_tree_tag_set(&mapping->page_tree, page_index(page), PAGECACHE_TAG_WRITEBACK); |
e4ad08fe6
|
1304 |
if (bdi_cap_account_writeback(bdi)) |
69cb51d18
|
1305 1306 |
__inc_bdi_stat(bdi, BDI_WRITEBACK); } |
1da177e4c
|
1307 1308 1309 1310 |
if (!PageDirty(page)) radix_tree_tag_clear(&mapping->page_tree, page_index(page), PAGECACHE_TAG_DIRTY); |
f446daaea
|
1311 1312 1313 |
radix_tree_tag_clear(&mapping->page_tree, page_index(page), PAGECACHE_TAG_TOWRITE); |
19fd62312
|
1314 |
spin_unlock_irqrestore(&mapping->tree_lock, flags); |
1da177e4c
|
1315 1316 1317 |
} else { ret = TestSetPageWriteback(page); } |
d688abf50
|
1318 1319 |
if (!ret) inc_zone_page_state(page, NR_WRITEBACK); |
1da177e4c
|
1320 1321 1322 1323 1324 1325 |
return ret; } EXPORT_SYMBOL(test_set_page_writeback); /* |
001281881
|
1326 |
* Return true if any of the pages in the mapping are marked with the |
1da177e4c
|
1327 1328 1329 1330 |
* passed tag. */ int mapping_tagged(struct address_space *mapping, int tag) { |
1da177e4c
|
1331 |
int ret; |
001281881
|
1332 |
rcu_read_lock(); |
1da177e4c
|
1333 |
ret = radix_tree_tagged(&mapping->page_tree, tag); |
001281881
|
1334 |
rcu_read_unlock(); |
1da177e4c
|
1335 1336 1337 |
return ret; } EXPORT_SYMBOL(mapping_tagged); |