Blame view
mm/page-writeback.c
39.9 KB
1da177e4c Linux-2.6.12-rc2 |
1 |
/* |
f30c22695 fix file specific... |
2 |
* mm/page-writeback.c |
1da177e4c Linux-2.6.12-rc2 |
3 4 |
* * Copyright (C) 2002, Linus Torvalds. |
04fbfdc14 mm: per device di... |
5 |
* Copyright (C) 2007 Red Hat, Inc., Peter Zijlstra <pzijlstr@redhat.com> |
1da177e4c Linux-2.6.12-rc2 |
6 7 8 9 |
* * Contains functions related to writing back dirty pages at the * address_space level. * |
e1f8e8744 Remove Andrew Mor... |
10 |
* 10Apr2002 Andrew Morton |
1da177e4c Linux-2.6.12-rc2 |
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 [PATCH] io-accoun... |
25 |
#include <linux/task_io_accounting_ops.h> |
1da177e4c Linux-2.6.12-rc2 |
26 27 |
#include <linux/blkdev.h> #include <linux/mpage.h> |
d08b3851d [PATCH] mm: track... |
28 |
#include <linux/rmap.h> |
1da177e4c Linux-2.6.12-rc2 |
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 [PATCH] BLOCK: Mo... |
35 |
#include <linux/buffer_head.h> |
811d736f9 [PATCH] BLOCK: Di... |
36 |
#include <linux/pagevec.h> |
1da177e4c Linux-2.6.12-rc2 |
37 38 39 |
/* * The maximum number of pages to writeout in a single bdflush/kupdate |
1c0eeaf56 introduce I_SYNC |
40 |
* operation. We do this so we don't hold I_SYNC against an inode for |
1da177e4c Linux-2.6.12-rc2 |
41 42 43 44 45 46 47 48 49 50 51 |
* enormous amounts of time, which would block a userspace task which has * been forced to throttle against that inode. Also, the code reevaluates * the dirty each time it has written this many pages. */ #define MAX_WRITEBACK_PAGES 1024 /* * 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 Linux-2.6.12-rc2 |
52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 |
/* * 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. * It should be somewhat larger than RATELIMIT_PAGES to ensure that reasonably * large amounts of I/O are submitted. */ static inline long sync_writeback_pages(void) { return ratelimit_pages + ratelimit_pages / 2; } /* The following parameters are exported via /proc/sys/vm */ /* * Start background writeback (via pdflush) at this percentage */ |
1b5e62b42 writeback: double... |
68 |
int dirty_background_ratio = 10; |
1da177e4c Linux-2.6.12-rc2 |
69 70 |
/* |
2da02997e mm: add dirty_bac... |
71 72 73 74 75 76 |
* 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 mm/page-writeback... |
77 78 79 80 81 82 |
* 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 Linux-2.6.12-rc2 |
83 84 |
* The generator of dirty data starts writeback at this percentage */ |
1b5e62b42 writeback: double... |
85 |
int vm_dirty_ratio = 20; |
1da177e4c Linux-2.6.12-rc2 |
86 87 |
/* |
2da02997e mm: add dirty_bac... |
88 89 90 91 92 93 |
* 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 mm: fix proc_doin... |
94 |
* The interval between `kupdate'-style writebacks |
1da177e4c Linux-2.6.12-rc2 |
95 |
*/ |
704503d83 mm: fix proc_doin... |
96 |
unsigned int dirty_writeback_interval = 5 * 100; /* sentiseconds */ |
1da177e4c Linux-2.6.12-rc2 |
97 98 |
/* |
704503d83 mm: fix proc_doin... |
99 |
* The longest time for which data is allowed to remain dirty |
1da177e4c Linux-2.6.12-rc2 |
100 |
*/ |
704503d83 mm: fix proc_doin... |
101 |
unsigned int dirty_expire_interval = 30 * 100; /* sentiseconds */ |
1da177e4c Linux-2.6.12-rc2 |
102 103 104 105 106 107 108 |
/* * Flag that makes the machine dump writes/reads and block dirtyings. */ int block_dump; /* |
ed5b43f15 [PATCH] Represent... |
109 110 |
* 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 Linux-2.6.12-rc2 |
111 112 113 114 115 116 117 118 119 |
*/ int laptop_mode; EXPORT_SYMBOL(laptop_mode); /* End of sysctl-exported parameters */ static void background_writeout(unsigned long _min_pages); |
1da177e4c Linux-2.6.12-rc2 |
120 |
/* |
04fbfdc14 mm: per device di... |
121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 |
* 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 mm: dirty balanci... |
137 |
static struct prop_descriptor vm_dirties; |
04fbfdc14 mm: per device di... |
138 |
|
04fbfdc14 mm: per device di... |
139 140 141 142 143 144 145 146 |
/* * 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 mm: add dirty_bac... |
147 148 149 150 151 |
if (vm_dirty_bytes) dirty_total = vm_dirty_bytes / PAGE_SIZE; else dirty_total = (vm_dirty_ratio * determine_dirtyable_memory()) / 100; |
04fbfdc14 mm: per device di... |
152 153 154 155 |
return 2 + ilog2(dirty_total - 1); } /* |
2da02997e mm: add dirty_bac... |
156 |
* update the period when the dirty threshold changes. |
04fbfdc14 mm: per device di... |
157 |
*/ |
2da02997e mm: add dirty_bac... |
158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 |
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, struct file *filp, void __user *buffer, size_t *lenp, loff_t *ppos) { int ret; ret = proc_dointvec_minmax(table, write, filp, buffer, lenp, ppos); if (ret == 0 && write) dirty_background_bytes = 0; return ret; } int dirty_background_bytes_handler(struct ctl_table *table, int write, struct file *filp, void __user *buffer, size_t *lenp, loff_t *ppos) { int ret; ret = proc_doulongvec_minmax(table, write, filp, buffer, lenp, ppos); if (ret == 0 && write) dirty_background_ratio = 0; return ret; } |
04fbfdc14 mm: per device di... |
188 189 190 191 192 |
int dirty_ratio_handler(struct ctl_table *table, int write, struct file *filp, void __user *buffer, size_t *lenp, loff_t *ppos) { int old_ratio = vm_dirty_ratio; |
2da02997e mm: add dirty_bac... |
193 194 195 |
int ret; ret = proc_dointvec_minmax(table, write, filp, buffer, lenp, ppos); |
04fbfdc14 mm: per device di... |
196 |
if (ret == 0 && write && vm_dirty_ratio != old_ratio) { |
2da02997e mm: add dirty_bac... |
197 198 199 200 201 202 203 204 205 206 207 |
update_completion_period(); vm_dirty_bytes = 0; } return ret; } int dirty_bytes_handler(struct ctl_table *table, int write, struct file *filp, void __user *buffer, size_t *lenp, loff_t *ppos) { |
fc3501d41 mm: fix dirty_byt... |
208 |
unsigned long old_bytes = vm_dirty_bytes; |
2da02997e mm: add dirty_bac... |
209 210 211 212 213 214 |
int ret; ret = proc_doulongvec_minmax(table, write, filp, buffer, lenp, ppos); if (ret == 0 && write && vm_dirty_bytes != old_bytes) { update_completion_period(); vm_dirty_ratio = 0; |
04fbfdc14 mm: per device di... |
215 216 217 218 219 220 221 222 223 224 |
} 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 mm: bdi: allow se... |
225 226 |
__prop_inc_percpu_max(&vm_completions, &bdi->completions, bdi->max_prop_frac); |
04fbfdc14 mm: per device di... |
227 |
} |
dd5656e59 mm: bdi: export b... |
228 229 230 231 232 233 234 235 236 |
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 mm: task dirty ac... |
237 |
void task_dirty_inc(struct task_struct *tsk) |
3e26c149c mm: dirty balanci... |
238 239 240 |
{ prop_inc_single(&vm_dirties, &tsk->dirties); } |
04fbfdc14 mm: per device di... |
241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 |
/* * 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; } } /* * Clip the earned share of dirty pages to that which is actually available. * This avoids exceeding the total dirty_limit when the floating averages * fluctuate too quickly. */ static void clip_bdi_dirty_limit(struct backing_dev_info *bdi, long dirty, long *pbdi_dirty) { long avail_dirty; avail_dirty = dirty - (global_page_state(NR_FILE_DIRTY) + global_page_state(NR_WRITEBACK) + |
fc3ba692a mm: Add NR_WRITEB... |
269 270 |
global_page_state(NR_UNSTABLE_NFS) + global_page_state(NR_WRITEBACK_TEMP)); |
04fbfdc14 mm: per device di... |
271 272 273 274 275 276 277 278 279 |
if (avail_dirty < 0) avail_dirty = 0; avail_dirty += bdi_stat(bdi, BDI_RECLAIMABLE) + bdi_stat(bdi, BDI_WRITEBACK); *pbdi_dirty = min(*pbdi_dirty, avail_dirty); } |
3e26c149c mm: dirty balanci... |
280 281 282 283 284 285 286 287 288 289 290 291 292 293 |
static inline void task_dirties_fraction(struct task_struct *tsk, long *numerator, long *denominator) { prop_fraction_single(&vm_dirties, &tsk->dirties, numerator, denominator); } /* * scale the dirty limit * * task specific dirty limit: * * dirty -= (dirty/8) * p_{t} */ |
f61eaf9fc mm/page-writeback... |
294 |
static void task_dirty_limit(struct task_struct *tsk, long *pdirty) |
3e26c149c mm: dirty balanci... |
295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 |
{ long numerator, denominator; long dirty = *pdirty; u64 inv = dirty >> 3; task_dirties_fraction(tsk, &numerator, &denominator); inv *= numerator; do_div(inv, denominator); dirty -= inv; if (dirty < *pdirty/2) dirty = *pdirty/2; *pdirty = dirty; } |
04fbfdc14 mm: per device di... |
310 |
/* |
189d3c4a9 mm: bdi: allow se... |
311 312 313 314 315 316 317 318 319 320 321 |
* */ static DEFINE_SPINLOCK(bdi_lock); static unsigned int bdi_min_ratio; int bdi_set_min_ratio(struct backing_dev_info *bdi, unsigned int min_ratio) { int ret = 0; unsigned long flags; spin_lock_irqsave(&bdi_lock, flags); |
a42dde041 mm: bdi: allow se... |
322 |
if (min_ratio > bdi->max_ratio) { |
189d3c4a9 mm: bdi: allow se... |
323 |
ret = -EINVAL; |
a42dde041 mm: bdi: allow se... |
324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 |
} 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; } } spin_unlock_irqrestore(&bdi_lock, flags); return ret; } int bdi_set_max_ratio(struct backing_dev_info *bdi, unsigned max_ratio) { unsigned long flags; int ret = 0; if (max_ratio > 100) return -EINVAL; spin_lock_irqsave(&bdi_lock, flags); if (bdi->min_ratio > max_ratio) { ret = -EINVAL; } else { bdi->max_ratio = max_ratio; bdi->max_prop_frac = (PROP_FRAC_BASE * max_ratio) / 100; } |
189d3c4a9 mm: bdi: allow se... |
353 354 355 356 |
spin_unlock_irqrestore(&bdi_lock, flags); return ret; } |
a42dde041 mm: bdi: allow se... |
357 |
EXPORT_SYMBOL(bdi_set_max_ratio); |
189d3c4a9 mm: bdi: allow se... |
358 359 |
/* |
1da177e4c Linux-2.6.12-rc2 |
360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 |
* 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 Use ZVC counters ... |
376 377 378 379 380 381 |
static unsigned long highmem_dirtyable_memory(unsigned long total) { #ifdef CONFIG_HIGHMEM int node; unsigned long x = 0; |
37b07e416 memoryless nodes:... |
382 |
for_each_node_state(node, N_HIGH_MEMORY) { |
1b4244647 Use ZVC counters ... |
383 384 |
struct zone *z = &NODE_DATA(node)->node_zones[ZONE_HIGHMEM]; |
4f98a2fee vmscan: split LRU... |
385 |
x += zone_page_state(z, NR_FREE_PAGES) + zone_lru_pages(z); |
1b4244647 Use ZVC counters ... |
386 387 388 389 390 391 392 393 394 395 396 397 |
} /* * 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 ftrace: limit tra... |
398 399 400 401 402 403 404 |
/** * 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 Use ZVC counters ... |
405 406 |
{ unsigned long x; |
4f98a2fee vmscan: split LRU... |
407 |
x = global_page_state(NR_FREE_PAGES) + global_lru_pages(); |
195cf453d mm/page-writeback... |
408 409 410 |
if (!vm_highmem_is_dirtyable) x -= highmem_dirtyable_memory(x); |
1b4244647 Use ZVC counters ... |
411 412 |
return x + 1; /* Ensure that we never return 0 */ } |
cf0ca9fe5 mm: bdi: export B... |
413 |
void |
364aeb284 mm: change dirty ... |
414 415 |
get_dirty_limits(unsigned long *pbackground, unsigned long *pdirty, unsigned long *pbdi_dirty, struct backing_dev_info *bdi) |
1da177e4c Linux-2.6.12-rc2 |
416 |
{ |
364aeb284 mm: change dirty ... |
417 418 |
unsigned long background; unsigned long dirty; |
1b4244647 Use ZVC counters ... |
419 |
unsigned long available_memory = determine_dirtyable_memory(); |
1da177e4c Linux-2.6.12-rc2 |
420 |
struct task_struct *tsk; |
2da02997e mm: add dirty_bac... |
421 422 423 424 425 426 427 428 429 430 |
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 Linux-2.6.12-rc2 |
431 |
|
2da02997e mm: add dirty_bac... |
432 433 434 435 |
if (dirty_background_bytes) background = DIV_ROUND_UP(dirty_background_bytes, PAGE_SIZE); else background = (dirty_background_ratio * available_memory) / 100; |
1da177e4c Linux-2.6.12-rc2 |
436 |
|
2da02997e mm: add dirty_bac... |
437 438 |
if (background >= dirty) background = dirty / 2; |
1da177e4c Linux-2.6.12-rc2 |
439 440 441 442 443 444 445 |
tsk = current; if (tsk->flags & PF_LESS_THROTTLE || rt_task(tsk)) { background += background / 4; dirty += dirty / 4; } *pbackground = background; *pdirty = dirty; |
04fbfdc14 mm: per device di... |
446 447 |
if (bdi) { |
189d3c4a9 mm: bdi: allow se... |
448 |
u64 bdi_dirty; |
04fbfdc14 mm: per device di... |
449 450 451 452 453 454 |
long numerator, denominator; /* * Calculate this BDI's share of the dirty ratio. */ bdi_writeout_fraction(bdi, &numerator, &denominator); |
189d3c4a9 mm: bdi: allow se... |
455 |
bdi_dirty = (dirty * (100 - bdi_min_ratio)) / 100; |
04fbfdc14 mm: per device di... |
456 457 |
bdi_dirty *= numerator; do_div(bdi_dirty, denominator); |
189d3c4a9 mm: bdi: allow se... |
458 |
bdi_dirty += (dirty * bdi->min_ratio) / 100; |
a42dde041 mm: bdi: allow se... |
459 460 |
if (bdi_dirty > (dirty * bdi->max_ratio) / 100) bdi_dirty = dirty * bdi->max_ratio / 100; |
04fbfdc14 mm: per device di... |
461 462 463 |
*pbdi_dirty = bdi_dirty; clip_bdi_dirty_limit(bdi, dirty, pbdi_dirty); |
3e26c149c mm: dirty balanci... |
464 |
task_dirty_limit(current, pbdi_dirty); |
04fbfdc14 mm: per device di... |
465 |
} |
1da177e4c Linux-2.6.12-rc2 |
466 467 468 469 470 471 472 473 474 475 476 |
} /* * 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'. * If we're over `background_thresh' then pdflush is woken to perform some * writeout. */ static void balance_dirty_pages(struct address_space *mapping) { |
5fce25a9d mm: speed up writ... |
477 478 |
long nr_reclaimable, bdi_nr_reclaimable; long nr_writeback, bdi_nr_writeback; |
364aeb284 mm: change dirty ... |
479 480 481 |
unsigned long background_thresh; unsigned long dirty_thresh; unsigned long bdi_thresh; |
1da177e4c Linux-2.6.12-rc2 |
482 483 484 485 486 487 488 489 490 491 492 |
unsigned long pages_written = 0; unsigned long write_chunk = sync_writeback_pages(); struct backing_dev_info *bdi = mapping->backing_dev_info; for (;;) { struct writeback_control wbc = { .bdi = bdi, .sync_mode = WB_SYNC_NONE, .older_than_this = NULL, .nr_to_write = write_chunk, |
111ebb6e6 [PATCH] writeback... |
493 |
.range_cyclic = 1, |
1da177e4c Linux-2.6.12-rc2 |
494 |
}; |
04fbfdc14 mm: per device di... |
495 496 |
get_dirty_limits(&background_thresh, &dirty_thresh, &bdi_thresh, bdi); |
5fce25a9d mm: speed up writ... |
497 498 499 500 |
nr_reclaimable = global_page_state(NR_FILE_DIRTY) + global_page_state(NR_UNSTABLE_NFS); nr_writeback = global_page_state(NR_WRITEBACK); |
04fbfdc14 mm: per device di... |
501 502 |
bdi_nr_reclaimable = bdi_stat(bdi, BDI_RECLAIMABLE); bdi_nr_writeback = bdi_stat(bdi, BDI_WRITEBACK); |
5fce25a9d mm: speed up writ... |
503 |
|
04fbfdc14 mm: per device di... |
504 505 |
if (bdi_nr_reclaimable + bdi_nr_writeback <= bdi_thresh) break; |
1da177e4c Linux-2.6.12-rc2 |
506 |
|
5fce25a9d mm: speed up writ... |
507 508 509 510 511 512 513 514 |
/* * 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; |
04fbfdc14 mm: per device di... |
515 516 |
if (!bdi->dirty_exceeded) bdi->dirty_exceeded = 1; |
1da177e4c Linux-2.6.12-rc2 |
517 518 519 520 521 522 523 |
/* 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. */ |
04fbfdc14 mm: per device di... |
524 |
if (bdi_nr_reclaimable) { |
1da177e4c Linux-2.6.12-rc2 |
525 |
writeback_inodes(&wbc); |
1da177e4c Linux-2.6.12-rc2 |
526 |
pages_written += write_chunk - wbc.nr_to_write; |
04fbfdc14 mm: per device di... |
527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 |
get_dirty_limits(&background_thresh, &dirty_thresh, &bdi_thresh, bdi); } /* * 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 if (bdi_nr_reclaimable) { bdi_nr_reclaimable = bdi_stat(bdi, BDI_RECLAIMABLE); bdi_nr_writeback = bdi_stat(bdi, BDI_WRITEBACK); |
1da177e4c Linux-2.6.12-rc2 |
547 |
} |
04fbfdc14 mm: per device di... |
548 549 550 551 552 |
if (bdi_nr_reclaimable + bdi_nr_writeback <= bdi_thresh) break; if (pages_written >= write_chunk) break; /* We've done our duty */ |
3fcfab16c [PATCH] separate ... |
553 |
congestion_wait(WRITE, HZ/10); |
1da177e4c Linux-2.6.12-rc2 |
554 |
} |
04fbfdc14 mm: per device di... |
555 556 557 |
if (bdi_nr_reclaimable + bdi_nr_writeback < bdi_thresh && bdi->dirty_exceeded) bdi->dirty_exceeded = 0; |
1da177e4c Linux-2.6.12-rc2 |
558 559 560 561 562 563 564 565 566 567 568 569 570 |
if (writeback_in_progress(bdi)) return; /* pdflush is already working this queue */ /* * 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) || |
04fbfdc14 mm: per device di... |
571 572 573 |
(!laptop_mode && (global_page_state(NR_FILE_DIRTY) + global_page_state(NR_UNSTABLE_NFS) > background_thresh))) |
1da177e4c Linux-2.6.12-rc2 |
574 575 |
pdflush_operation(background_writeout, 0); } |
a200ee182 mm: set_page_dirt... |
576 |
void set_page_dirty_balance(struct page *page, int page_mkwrite) |
edc79b2a4 [PATCH] mm: balan... |
577 |
{ |
a200ee182 mm: set_page_dirt... |
578 |
if (set_page_dirty(page) || page_mkwrite) { |
edc79b2a4 [PATCH] mm: balan... |
579 580 581 582 583 584 |
struct address_space *mapping = page_mapping(page); if (mapping) balance_dirty_pages_ratelimited(mapping); } } |
1da177e4c Linux-2.6.12-rc2 |
585 |
/** |
fa5a734e4 [PATCH] balance_d... |
586 |
* balance_dirty_pages_ratelimited_nr - balance dirty memory state |
67be2dd1b [PATCH] DocBook: ... |
587 |
* @mapping: address_space which was dirtied |
a580290c3 Documentation: fi... |
588 |
* @nr_pages_dirtied: number of pages which the caller has just dirtied |
1da177e4c Linux-2.6.12-rc2 |
589 590 591 592 593 594 595 596 597 598 |
* * 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 [PATCH] balance_d... |
599 600 |
void balance_dirty_pages_ratelimited_nr(struct address_space *mapping, unsigned long nr_pages_dirtied) |
1da177e4c Linux-2.6.12-rc2 |
601 |
{ |
fa5a734e4 [PATCH] balance_d... |
602 603 604 |
static DEFINE_PER_CPU(unsigned long, ratelimits) = 0; unsigned long ratelimit; unsigned long *p; |
1da177e4c Linux-2.6.12-rc2 |
605 606 |
ratelimit = ratelimit_pages; |
04fbfdc14 mm: per device di... |
607 |
if (mapping->backing_dev_info->dirty_exceeded) |
1da177e4c Linux-2.6.12-rc2 |
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 [PATCH] balance_d... |
614 615 616 617 618 619 |
preempt_disable(); p = &__get_cpu_var(ratelimits); *p += nr_pages_dirtied; if (unlikely(*p >= ratelimit)) { *p = 0; preempt_enable(); |
1da177e4c Linux-2.6.12-rc2 |
620 621 622 |
balance_dirty_pages(mapping); return; } |
fa5a734e4 [PATCH] balance_d... |
623 |
preempt_enable(); |
1da177e4c Linux-2.6.12-rc2 |
624 |
} |
fa5a734e4 [PATCH] balance_d... |
625 |
EXPORT_SYMBOL(balance_dirty_pages_ratelimited_nr); |
1da177e4c Linux-2.6.12-rc2 |
626 |
|
232ea4d69 [PATCH] throttle_... |
627 |
void throttle_vm_writeout(gfp_t gfp_mask) |
1da177e4c Linux-2.6.12-rc2 |
628 |
{ |
364aeb284 mm: change dirty ... |
629 630 |
unsigned long background_thresh; unsigned long dirty_thresh; |
1da177e4c Linux-2.6.12-rc2 |
631 632 |
for ( ; ; ) { |
04fbfdc14 mm: per device di... |
633 |
get_dirty_limits(&background_thresh, &dirty_thresh, NULL, NULL); |
1da177e4c Linux-2.6.12-rc2 |
634 635 636 637 638 639 |
/* * 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 [PATCH] zoned vm ... |
640 641 642 |
if (global_page_state(NR_UNSTABLE_NFS) + global_page_state(NR_WRITEBACK) <= dirty_thresh) break; |
3fcfab16c [PATCH] separate ... |
643 |
congestion_wait(WRITE, HZ/10); |
369f2389e writeback: remove... |
644 645 646 647 648 649 650 651 |
/* * 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 Linux-2.6.12-rc2 |
652 653 |
} } |
1da177e4c Linux-2.6.12-rc2 |
654 655 656 657 658 659 660 661 662 663 664 665 666 |
/* * writeback at least _min_pages, and keep writing until the amount of dirty * memory is less than the background threshold, or until we're all clean. */ static void background_writeout(unsigned long _min_pages) { long min_pages = _min_pages; struct writeback_control wbc = { .bdi = NULL, .sync_mode = WB_SYNC_NONE, .older_than_this = NULL, .nr_to_write = 0, .nonblocking = 1, |
111ebb6e6 [PATCH] writeback... |
667 |
.range_cyclic = 1, |
1da177e4c Linux-2.6.12-rc2 |
668 669 670 |
}; for ( ; ; ) { |
364aeb284 mm: change dirty ... |
671 672 |
unsigned long background_thresh; unsigned long dirty_thresh; |
1da177e4c Linux-2.6.12-rc2 |
673 |
|
04fbfdc14 mm: per device di... |
674 |
get_dirty_limits(&background_thresh, &dirty_thresh, NULL, NULL); |
c24f21bda [PATCH] zoned vm ... |
675 676 |
if (global_page_state(NR_FILE_DIRTY) + global_page_state(NR_UNSTABLE_NFS) < background_thresh |
1da177e4c Linux-2.6.12-rc2 |
677 678 |
&& min_pages <= 0) break; |
8bc3be275 writeback: speed ... |
679 |
wbc.more_io = 0; |
1da177e4c Linux-2.6.12-rc2 |
680 681 682 683 684 685 686 |
wbc.encountered_congestion = 0; wbc.nr_to_write = MAX_WRITEBACK_PAGES; wbc.pages_skipped = 0; writeback_inodes(&wbc); min_pages -= MAX_WRITEBACK_PAGES - wbc.nr_to_write; if (wbc.nr_to_write > 0 || wbc.pages_skipped > 0) { /* Wrote less than expected */ |
8bc3be275 writeback: speed ... |
687 688 689 |
if (wbc.encountered_congestion || wbc.more_io) congestion_wait(WRITE, HZ/10); else |
1da177e4c Linux-2.6.12-rc2 |
690 691 692 693 694 695 696 697 698 699 |
break; } } } /* * Start writeback of `nr_pages' pages. If `nr_pages' is zero, write back * the whole world. Returns 0 if a pdflush thread was dispatched. Returns * -1 if all pdflush threads were busy. */ |
687a21cee [PATCH] rename wa... |
700 |
int wakeup_pdflush(long nr_pages) |
1da177e4c Linux-2.6.12-rc2 |
701 |
{ |
c24f21bda [PATCH] zoned vm ... |
702 703 704 |
if (nr_pages == 0) nr_pages = global_page_state(NR_FILE_DIRTY) + global_page_state(NR_UNSTABLE_NFS); |
1da177e4c Linux-2.6.12-rc2 |
705 706 707 708 709 |
return pdflush_operation(background_writeout, nr_pages); } static void wb_timer_fn(unsigned long unused); static void laptop_timer_fn(unsigned long unused); |
8d06afab7 [PATCH] timer ini... |
710 711 |
static DEFINE_TIMER(wb_timer, wb_timer_fn, 0, 0); static DEFINE_TIMER(laptop_mode_wb_timer, laptop_timer_fn, 0, 0); |
1da177e4c Linux-2.6.12-rc2 |
712 713 714 715 716 717 718 719 720 |
/* * Periodic writeback of "old" data. * * 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. * |
f6ef94381 [PATCH] Represent... |
721 722 |
* Try to run once per dirty_writeback_interval. But if a writeback event * takes longer than a dirty_writeback_interval interval, then leave a |
1da177e4c Linux-2.6.12-rc2 |
723 724 725 726 727 728 729 730 731 732 733 |
* one-second gap. * * older_than_this takes precedence over nr_to_write. So we'll only write back * all dirty pages if they are all attached to "old" mappings. */ static void wb_kupdate(unsigned long arg) { unsigned long oldest_jif; unsigned long start_jif; unsigned long next_jif; long nr_to_write; |
1da177e4c Linux-2.6.12-rc2 |
734 735 736 737 738 739 740 |
struct writeback_control wbc = { .bdi = NULL, .sync_mode = WB_SYNC_NONE, .older_than_this = &oldest_jif, .nr_to_write = 0, .nonblocking = 1, .for_kupdate = 1, |
111ebb6e6 [PATCH] writeback... |
741 |
.range_cyclic = 1, |
1da177e4c Linux-2.6.12-rc2 |
742 743 744 |
}; sync_supers(); |
704503d83 mm: fix proc_doin... |
745 |
oldest_jif = jiffies - msecs_to_jiffies(dirty_expire_interval); |
1da177e4c Linux-2.6.12-rc2 |
746 |
start_jif = jiffies; |
704503d83 mm: fix proc_doin... |
747 |
next_jif = start_jif + msecs_to_jiffies(dirty_writeback_interval * 10); |
c24f21bda [PATCH] zoned vm ... |
748 749 |
nr_to_write = global_page_state(NR_FILE_DIRTY) + global_page_state(NR_UNSTABLE_NFS) + |
1da177e4c Linux-2.6.12-rc2 |
750 751 |
(inodes_stat.nr_inodes - inodes_stat.nr_unused); while (nr_to_write > 0) { |
8bc3be275 writeback: speed ... |
752 |
wbc.more_io = 0; |
1da177e4c Linux-2.6.12-rc2 |
753 754 755 756 |
wbc.encountered_congestion = 0; wbc.nr_to_write = MAX_WRITEBACK_PAGES; writeback_inodes(&wbc); if (wbc.nr_to_write > 0) { |
8bc3be275 writeback: speed ... |
757 |
if (wbc.encountered_congestion || wbc.more_io) |
3fcfab16c [PATCH] separate ... |
758 |
congestion_wait(WRITE, HZ/10); |
1da177e4c Linux-2.6.12-rc2 |
759 760 761 762 763 764 765 |
else break; /* All the old data is written */ } nr_to_write -= MAX_WRITEBACK_PAGES - wbc.nr_to_write; } if (time_before(next_jif, jiffies + HZ)) next_jif = jiffies + HZ; |
f6ef94381 [PATCH] Represent... |
766 |
if (dirty_writeback_interval) |
1da177e4c Linux-2.6.12-rc2 |
767 768 769 770 771 772 773 |
mod_timer(&wb_timer, next_jif); } /* * sysctl handler for /proc/sys/vm/dirty_writeback_centisecs */ int dirty_writeback_centisecs_handler(ctl_table *table, int write, |
3e733f071 dirty_writeback_c... |
774 |
struct file *file, void __user *buffer, size_t *length, loff_t *ppos) |
1da177e4c Linux-2.6.12-rc2 |
775 |
{ |
704503d83 mm: fix proc_doin... |
776 |
proc_dointvec(table, write, file, buffer, length, ppos); |
3e733f071 dirty_writeback_c... |
777 |
if (dirty_writeback_interval) |
704503d83 mm: fix proc_doin... |
778 779 |
mod_timer(&wb_timer, jiffies + msecs_to_jiffies(dirty_writeback_interval * 10)); |
3e733f071 dirty_writeback_c... |
780 |
else |
1da177e4c Linux-2.6.12-rc2 |
781 |
del_timer(&wb_timer); |
1da177e4c Linux-2.6.12-rc2 |
782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 |
return 0; } static void wb_timer_fn(unsigned long unused) { if (pdflush_operation(wb_kupdate, 0) < 0) mod_timer(&wb_timer, jiffies + HZ); /* delay 1 second */ } static void laptop_flush(unsigned long unused) { sys_sync(); } static void laptop_timer_fn(unsigned long unused) { pdflush_operation(laptop_flush, 0); } /* * 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. */ void laptop_io_completion(void) { |
ed5b43f15 [PATCH] Represent... |
808 |
mod_timer(&laptop_mode_wb_timer, jiffies + laptop_mode); |
1da177e4c Linux-2.6.12-rc2 |
809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 |
} /* * 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) { del_timer(&laptop_mode_wb_timer); } /* * 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 [PATCH] call mm/p... |
837 |
void writeback_set_ratelimit(void) |
1da177e4c Linux-2.6.12-rc2 |
838 |
{ |
40c99aae2 [PATCH] remove st... |
839 |
ratelimit_pages = vm_total_pages / (num_online_cpus() * 32); |
1da177e4c Linux-2.6.12-rc2 |
840 841 842 843 844 |
if (ratelimit_pages < 16) ratelimit_pages = 16; if (ratelimit_pages * PAGE_CACHE_SIZE > 4096 * 1024) ratelimit_pages = (4096 * 1024) / PAGE_CACHE_SIZE; } |
26c2143b6 [PATCH] cpu hotpl... |
845 |
static int __cpuinit |
1da177e4c Linux-2.6.12-rc2 |
846 847 |
ratelimit_handler(struct notifier_block *self, unsigned long u, void *v) { |
2d1d43f6a [PATCH] call mm/p... |
848 |
writeback_set_ratelimit(); |
aa0f03037 [PATCH] Change co... |
849 |
return NOTIFY_DONE; |
1da177e4c Linux-2.6.12-rc2 |
850 |
} |
74b85f379 [PATCH] cpu hotpl... |
851 |
static struct notifier_block __cpuinitdata ratelimit_nb = { |
1da177e4c Linux-2.6.12-rc2 |
852 853 854 855 856 |
.notifier_call = ratelimit_handler, .next = NULL, }; /* |
dc6e29da9 Fix balance_dirty... |
857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 |
* 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 Linux-2.6.12-rc2 |
873 874 875 |
*/ void __init page_writeback_init(void) { |
04fbfdc14 mm: per device di... |
876 |
int shift; |
704503d83 mm: fix proc_doin... |
877 878 |
mod_timer(&wb_timer, jiffies + msecs_to_jiffies(dirty_writeback_interval * 10)); |
2d1d43f6a [PATCH] call mm/p... |
879 |
writeback_set_ratelimit(); |
1da177e4c Linux-2.6.12-rc2 |
880 |
register_cpu_notifier(&ratelimit_nb); |
04fbfdc14 mm: per device di... |
881 882 883 |
shift = calc_period_shift(); prop_descriptor_init(&vm_completions, shift); |
3e26c149c mm: dirty balanci... |
884 |
prop_descriptor_init(&vm_dirties, shift); |
1da177e4c Linux-2.6.12-rc2 |
885 |
} |
811d736f9 [PATCH] BLOCK: Di... |
886 |
/** |
0ea971801 consolidate gener... |
887 |
* write_cache_pages - walk the list of dirty pages of the given address space and write all of them. |
811d736f9 [PATCH] BLOCK: Di... |
888 889 |
* @mapping: address space structure to write * @wbc: subtract the number of written pages from *@wbc->nr_to_write |
0ea971801 consolidate gener... |
890 891 |
* @writepage: function called for each page * @data: data passed to writepage function |
811d736f9 [PATCH] BLOCK: Di... |
892 |
* |
0ea971801 consolidate gener... |
893 |
* If a page is already under I/O, write_cache_pages() skips it, even |
811d736f9 [PATCH] BLOCK: Di... |
894 895 896 897 898 899 |
* 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. |
811d736f9 [PATCH] BLOCK: Di... |
900 |
*/ |
0ea971801 consolidate gener... |
901 902 903 |
int write_cache_pages(struct address_space *mapping, struct writeback_control *wbc, writepage_t writepage, void *data) |
811d736f9 [PATCH] BLOCK: Di... |
904 905 906 907 |
{ struct backing_dev_info *bdi = mapping->backing_dev_info; int ret = 0; int done = 0; |
811d736f9 [PATCH] BLOCK: Di... |
908 909 |
struct pagevec pvec; int nr_pages; |
31a12666d mm: write_cache_p... |
910 |
pgoff_t uninitialized_var(writeback_index); |
811d736f9 [PATCH] BLOCK: Di... |
911 912 |
pgoff_t index; pgoff_t end; /* Inclusive */ |
bd19e012f mm: write_cache_p... |
913 |
pgoff_t done_index; |
31a12666d mm: write_cache_p... |
914 |
int cycled; |
811d736f9 [PATCH] BLOCK: Di... |
915 |
int range_whole = 0; |
17bc6c30c vfs: Add no_nrwri... |
916 |
long nr_to_write = wbc->nr_to_write; |
811d736f9 [PATCH] BLOCK: Di... |
917 918 919 920 921 |
if (wbc->nonblocking && bdi_write_congested(bdi)) { wbc->encountered_congestion = 1; return 0; } |
811d736f9 [PATCH] BLOCK: Di... |
922 923 |
pagevec_init(&pvec, 0); if (wbc->range_cyclic) { |
31a12666d mm: write_cache_p... |
924 925 926 927 928 929 |
writeback_index = mapping->writeback_index; /* prev offset */ index = writeback_index; if (index == 0) cycled = 1; else cycled = 0; |
811d736f9 [PATCH] BLOCK: Di... |
930 931 932 933 934 935 |
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 mm: write_cache_p... |
936 |
cycled = 1; /* ignore range_cyclic tests */ |
811d736f9 [PATCH] BLOCK: Di... |
937 938 |
} retry: |
bd19e012f mm: write_cache_p... |
939 |
done_index = index; |
5a3d5c981 mm: write_cache_p... |
940 941 942 943 944 945 946 947 |
while (!done && (index <= end)) { int i; nr_pages = pagevec_lookup_tag(&pvec, mapping, &index, PAGECACHE_TAG_DIRTY, min(end - index, (pgoff_t)PAGEVEC_SIZE-1) + 1); if (nr_pages == 0) break; |
811d736f9 [PATCH] BLOCK: Di... |
948 |
|
811d736f9 [PATCH] BLOCK: Di... |
949 950 951 952 |
for (i = 0; i < nr_pages; i++) { struct page *page = pvec.pages[i]; /* |
d5482cdf8 mm: write_cache_p... |
953 954 955 956 957 |
* 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 [PATCH] BLOCK: Di... |
958 |
*/ |
d5482cdf8 mm: write_cache_p... |
959 960 961 962 963 964 965 966 967 968 |
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 [PATCH] BLOCK: Di... |
969 |
lock_page(page); |
5a3d5c981 mm: write_cache_p... |
970 971 972 973 974 975 976 977 |
/* * 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 [PATCH] BLOCK: Di... |
978 |
if (unlikely(page->mapping != mapping)) { |
5a3d5c981 mm: write_cache_p... |
979 |
continue_unlock: |
811d736f9 [PATCH] BLOCK: Di... |
980 981 982 |
unlock_page(page); continue; } |
515f4a037 mm: write_cache_p... |
983 984 985 986 987 988 989 990 991 992 993 |
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 [PATCH] BLOCK: Di... |
994 |
|
515f4a037 mm: write_cache_p... |
995 996 |
BUG_ON(PageWriteback(page)); if (!clear_page_dirty_for_io(page)) |
5a3d5c981 mm: write_cache_p... |
997 |
goto continue_unlock; |
811d736f9 [PATCH] BLOCK: Di... |
998 |
|
0ea971801 consolidate gener... |
999 |
ret = (*writepage)(page, wbc, data); |
00266770b mm: write_cache_p... |
1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 |
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; } } |
89e121900 writeback: fix br... |
1018 |
if (nr_to_write > 0) { |
dcf6a79dd write-back: fix n... |
1019 |
nr_to_write--; |
89e121900 writeback: fix br... |
1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 |
if (nr_to_write == 0 && wbc->sync_mode == WB_SYNC_NONE) { /* * We stop writing back only if we are * not doing integrity sync. In case of * integrity sync we have to keep going * because someone may be concurrently * dirtying pages, and we might have * synced a lot of newly appeared dirty * pages, but have not synced all of the * old dirty pages. */ done = 1; break; } |
05fe478dd mm: write_cache_p... |
1035 |
} |
dcf6a79dd write-back: fix n... |
1036 |
|
811d736f9 [PATCH] BLOCK: Di... |
1037 1038 1039 |
if (wbc->nonblocking && bdi_write_congested(bdi)) { wbc->encountered_congestion = 1; done = 1; |
82fd1a9a8 mm: write_cache_p... |
1040 |
break; |
811d736f9 [PATCH] BLOCK: Di... |
1041 1042 1043 1044 1045 |
} } pagevec_release(&pvec); cond_resched(); } |
3a4c6800f Fix page writebac... |
1046 |
if (!cycled && !done) { |
811d736f9 [PATCH] BLOCK: Di... |
1047 |
/* |
31a12666d mm: write_cache_p... |
1048 |
* range_cyclic: |
811d736f9 [PATCH] BLOCK: Di... |
1049 1050 1051 |
* We hit the last page and there is more work to be done: wrap * back to the start of the file */ |
31a12666d mm: write_cache_p... |
1052 |
cycled = 1; |
811d736f9 [PATCH] BLOCK: Di... |
1053 |
index = 0; |
31a12666d mm: write_cache_p... |
1054 |
end = writeback_index - 1; |
811d736f9 [PATCH] BLOCK: Di... |
1055 1056 |
goto retry; } |
17bc6c30c vfs: Add no_nrwri... |
1057 1058 |
if (!wbc->no_nrwrite_index_update) { if (wbc->range_cyclic || (range_whole && nr_to_write > 0)) |
bd19e012f mm: write_cache_p... |
1059 |
mapping->writeback_index = done_index; |
17bc6c30c vfs: Add no_nrwri... |
1060 1061 |
wbc->nr_to_write = nr_to_write; } |
06d6cf695 mm: Add range_con... |
1062 |
|
811d736f9 [PATCH] BLOCK: Di... |
1063 1064 |
return ret; } |
0ea971801 consolidate gener... |
1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 |
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 [PATCH] BLOCK: Di... |
1097 1098 |
EXPORT_SYMBOL(generic_writepages); |
1da177e4c Linux-2.6.12-rc2 |
1099 1100 |
int do_writepages(struct address_space *mapping, struct writeback_control *wbc) { |
22905f775 identify multipag... |
1101 |
int ret; |
1da177e4c Linux-2.6.12-rc2 |
1102 1103 |
if (wbc->nr_to_write <= 0) return 0; |
22905f775 identify multipag... |
1104 |
wbc->for_writepages = 1; |
1da177e4c Linux-2.6.12-rc2 |
1105 |
if (mapping->a_ops->writepages) |
d08b3851d [PATCH] mm: track... |
1106 |
ret = mapping->a_ops->writepages(mapping, wbc); |
22905f775 identify multipag... |
1107 1108 1109 1110 |
else ret = generic_writepages(mapping, wbc); wbc->for_writepages = 0; return ret; |
1da177e4c Linux-2.6.12-rc2 |
1111 1112 1113 1114 |
} /** * write_one_page - write out a single page and optionally wait on I/O |
67be2dd1b [PATCH] DocBook: ... |
1115 1116 |
* @page: the page to write * @wait: if true, wait on writeout |
1da177e4c Linux-2.6.12-rc2 |
1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 |
* * 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 [PATCH] simplify ... |
1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 |
* 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 vfs: add/use acco... |
1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 |
* 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); } } /* |
1da177e4c Linux-2.6.12-rc2 |
1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 |
* 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 spelling fixes: mm/ |
1189 |
* mapping by re-checking page_mapping() inside tree_lock. |
1da177e4c Linux-2.6.12-rc2 |
1190 1191 1192 |
*/ int __set_page_dirty_nobuffers(struct page *page) { |
1da177e4c Linux-2.6.12-rc2 |
1193 1194 1195 |
if (!TestSetPageDirty(page)) { struct address_space *mapping = page_mapping(page); struct address_space *mapping2; |
8c08540f8 [PATCH] clean up ... |
1196 1197 |
if (!mapping) return 1; |
19fd62312 mm: spinlock tree... |
1198 |
spin_lock_irq(&mapping->tree_lock); |
8c08540f8 [PATCH] clean up ... |
1199 1200 1201 |
mapping2 = page_mapping(page); if (mapping2) { /* Race with truncate? */ BUG_ON(mapping2 != mapping); |
787d2214c fs: introduce som... |
1202 |
WARN_ON_ONCE(!PagePrivate(page) && !PageUptodate(page)); |
e3a7cca1e vfs: add/use acco... |
1203 |
account_page_dirtied(page, mapping); |
8c08540f8 [PATCH] clean up ... |
1204 1205 1206 |
radix_tree_tag_set(&mapping->page_tree, page_index(page), PAGECACHE_TAG_DIRTY); } |
19fd62312 mm: spinlock tree... |
1207 |
spin_unlock_irq(&mapping->tree_lock); |
8c08540f8 [PATCH] clean up ... |
1208 1209 1210 |
if (mapping->host) { /* !PageAnon && !swapper_space */ __mark_inode_dirty(mapping->host, I_DIRTY_PAGES); |
1da177e4c Linux-2.6.12-rc2 |
1211 |
} |
4741c9fd3 [PATCH] set_page_... |
1212 |
return 1; |
1da177e4c Linux-2.6.12-rc2 |
1213 |
} |
4741c9fd3 [PATCH] set_page_... |
1214 |
return 0; |
1da177e4c Linux-2.6.12-rc2 |
1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 |
} 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); /* * If the mapping doesn't provide a set_page_dirty a_op, then * just fall through and assume that it wants buffer_heads. */ |
1cf6e7d83 mm: task dirty ac... |
1234 |
int set_page_dirty(struct page *page) |
1da177e4c Linux-2.6.12-rc2 |
1235 1236 1237 1238 1239 |
{ struct address_space *mapping = page_mapping(page); if (likely(mapping)) { int (*spd)(struct page *) = mapping->a_ops->set_page_dirty; |
9361401eb [PATCH] BLOCK: Ma... |
1240 1241 1242 1243 1244 |
#ifdef CONFIG_BLOCK if (!spd) spd = __set_page_dirty_buffers; #endif return (*spd)(page); |
1da177e4c Linux-2.6.12-rc2 |
1245 |
} |
4741c9fd3 [PATCH] set_page_... |
1246 1247 1248 1249 |
if (!PageDirty(page)) { if (!TestSetPageDirty(page)) return 1; } |
1da177e4c Linux-2.6.12-rc2 |
1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 |
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 [PATCH] mm: non s... |
1267 |
lock_page_nosync(page); |
1da177e4c Linux-2.6.12-rc2 |
1268 1269 1270 1271 1272 1273 1274 |
ret = set_page_dirty(page); unlock_page(page); return ret; } EXPORT_SYMBOL(set_page_dirty_lock); /* |
1da177e4c Linux-2.6.12-rc2 |
1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 |
* 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 mm: fix clear_pag... |
1291 |
BUG_ON(!PageLocked(page)); |
fe3cba17c mm: share PG_read... |
1292 |
ClearPageReclaim(page); |
7658cc289 VM: Fix nasty and... |
1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 |
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 VM: Fix nasty and... |
1318 1319 1320 |
*/ if (page_mkclean(page)) set_page_dirty(page); |
79352894b mm: fix clear_pag... |
1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 |
/* * 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 VM: Fix nasty and... |
1331 |
if (TestClearPageDirty(page)) { |
8c08540f8 [PATCH] clean up ... |
1332 |
dec_zone_page_state(page, NR_FILE_DIRTY); |
c9e51e418 mm: count reclaim... |
1333 1334 |
dec_bdi_stat(mapping->backing_dev_info, BDI_RECLAIMABLE); |
7658cc289 VM: Fix nasty and... |
1335 |
return 1; |
1da177e4c Linux-2.6.12-rc2 |
1336 |
} |
7658cc289 VM: Fix nasty and... |
1337 |
return 0; |
1da177e4c Linux-2.6.12-rc2 |
1338 |
} |
7658cc289 VM: Fix nasty and... |
1339 |
return TestClearPageDirty(page); |
1da177e4c Linux-2.6.12-rc2 |
1340 |
} |
58bb01a9c [PATCH] re-export... |
1341 |
EXPORT_SYMBOL(clear_page_dirty_for_io); |
1da177e4c Linux-2.6.12-rc2 |
1342 1343 1344 1345 1346 1347 1348 |
int test_clear_page_writeback(struct page *page) { struct address_space *mapping = page_mapping(page); int ret; if (mapping) { |
69cb51d18 mm: count writeba... |
1349 |
struct backing_dev_info *bdi = mapping->backing_dev_info; |
1da177e4c Linux-2.6.12-rc2 |
1350 |
unsigned long flags; |
19fd62312 mm: spinlock tree... |
1351 |
spin_lock_irqsave(&mapping->tree_lock, flags); |
1da177e4c Linux-2.6.12-rc2 |
1352 |
ret = TestClearPageWriteback(page); |
69cb51d18 mm: count writeba... |
1353 |
if (ret) { |
1da177e4c Linux-2.6.12-rc2 |
1354 1355 1356 |
radix_tree_tag_clear(&mapping->page_tree, page_index(page), PAGECACHE_TAG_WRITEBACK); |
e4ad08fe6 mm: bdi: add sepa... |
1357 |
if (bdi_cap_account_writeback(bdi)) { |
69cb51d18 mm: count writeba... |
1358 |
__dec_bdi_stat(bdi, BDI_WRITEBACK); |
04fbfdc14 mm: per device di... |
1359 1360 |
__bdi_writeout_inc(bdi); } |
69cb51d18 mm: count writeba... |
1361 |
} |
19fd62312 mm: spinlock tree... |
1362 |
spin_unlock_irqrestore(&mapping->tree_lock, flags); |
1da177e4c Linux-2.6.12-rc2 |
1363 1364 1365 |
} else { ret = TestClearPageWriteback(page); } |
d688abf50 move page writeba... |
1366 1367 |
if (ret) dec_zone_page_state(page, NR_WRITEBACK); |
1da177e4c Linux-2.6.12-rc2 |
1368 1369 1370 1371 1372 1373 1374 1375 1376 |
return ret; } int test_set_page_writeback(struct page *page) { struct address_space *mapping = page_mapping(page); int ret; if (mapping) { |
69cb51d18 mm: count writeba... |
1377 |
struct backing_dev_info *bdi = mapping->backing_dev_info; |
1da177e4c Linux-2.6.12-rc2 |
1378 |
unsigned long flags; |
19fd62312 mm: spinlock tree... |
1379 |
spin_lock_irqsave(&mapping->tree_lock, flags); |
1da177e4c Linux-2.6.12-rc2 |
1380 |
ret = TestSetPageWriteback(page); |
69cb51d18 mm: count writeba... |
1381 |
if (!ret) { |
1da177e4c Linux-2.6.12-rc2 |
1382 1383 1384 |
radix_tree_tag_set(&mapping->page_tree, page_index(page), PAGECACHE_TAG_WRITEBACK); |
e4ad08fe6 mm: bdi: add sepa... |
1385 |
if (bdi_cap_account_writeback(bdi)) |
69cb51d18 mm: count writeba... |
1386 1387 |
__inc_bdi_stat(bdi, BDI_WRITEBACK); } |
1da177e4c Linux-2.6.12-rc2 |
1388 1389 1390 1391 |
if (!PageDirty(page)) radix_tree_tag_clear(&mapping->page_tree, page_index(page), PAGECACHE_TAG_DIRTY); |
19fd62312 mm: spinlock tree... |
1392 |
spin_unlock_irqrestore(&mapping->tree_lock, flags); |
1da177e4c Linux-2.6.12-rc2 |
1393 1394 1395 |
} else { ret = TestSetPageWriteback(page); } |
d688abf50 move page writeba... |
1396 1397 |
if (!ret) inc_zone_page_state(page, NR_WRITEBACK); |
1da177e4c Linux-2.6.12-rc2 |
1398 1399 1400 1401 1402 1403 |
return ret; } EXPORT_SYMBOL(test_set_page_writeback); /* |
001281881 mm: use lockless ... |
1404 |
* Return true if any of the pages in the mapping are marked with the |
1da177e4c Linux-2.6.12-rc2 |
1405 1406 1407 1408 |
* passed tag. */ int mapping_tagged(struct address_space *mapping, int tag) { |
1da177e4c Linux-2.6.12-rc2 |
1409 |
int ret; |
001281881 mm: use lockless ... |
1410 |
rcu_read_lock(); |
1da177e4c Linux-2.6.12-rc2 |
1411 |
ret = radix_tree_tagged(&mapping->page_tree, tag); |
001281881 mm: use lockless ... |
1412 |
rcu_read_unlock(); |
1da177e4c Linux-2.6.12-rc2 |
1413 1414 1415 |
return ret; } EXPORT_SYMBOL(mapping_tagged); |