Blame view
mm/compaction.c
56.1 KB
748446bb6
|
1 2 3 4 5 6 7 8 9 |
/* * linux/mm/compaction.c * * Memory compaction for the reduction of external fragmentation. Note that * this heavily depends upon page migration to do all the real heavy * lifting * * Copyright IBM Corp. 2007-2010 Mel Gorman <mel@csn.ul.ie> */ |
698b1b306
|
10 |
#include <linux/cpu.h> |
748446bb6
|
11 12 13 14 15 |
#include <linux/swap.h> #include <linux/migrate.h> #include <linux/compaction.h> #include <linux/mm_inline.h> #include <linux/backing-dev.h> |
76ab0f530
|
16 |
#include <linux/sysctl.h> |
ed4a6d7f0
|
17 |
#include <linux/sysfs.h> |
194159fbc
|
18 |
#include <linux/page-isolation.h> |
b8c73fc24
|
19 |
#include <linux/kasan.h> |
698b1b306
|
20 21 |
#include <linux/kthread.h> #include <linux/freezer.h> |
83358ece2
|
22 |
#include <linux/page_owner.h> |
748446bb6
|
23 |
#include "internal.h" |
010fc29a4
|
24 25 26 27 28 29 30 31 32 33 34 35 36 37 |
#ifdef CONFIG_COMPACTION static inline void count_compact_event(enum vm_event_item item) { count_vm_event(item); } static inline void count_compact_events(enum vm_event_item item, long delta) { count_vm_events(item, delta); } #else #define count_compact_event(item) do { } while (0) #define count_compact_events(item, delta) do { } while (0) #endif |
ff9543fd3
|
38 |
#if defined CONFIG_COMPACTION || defined CONFIG_CMA |
b7aba6984
|
39 40 |
#define CREATE_TRACE_POINTS #include <trace/events/compaction.h> |
06b6640a3
|
41 42 43 44 |
#define block_start_pfn(pfn, order) round_down(pfn, 1UL << (order)) #define block_end_pfn(pfn, order) ALIGN((pfn) + 1, 1UL << (order)) #define pageblock_start_pfn(pfn) block_start_pfn(pfn, pageblock_order) #define pageblock_end_pfn(pfn) block_end_pfn(pfn, pageblock_order) |
748446bb6
|
45 46 47 |
static unsigned long release_freepages(struct list_head *freelist) { struct page *page, *next; |
6bace090a
|
48 |
unsigned long high_pfn = 0; |
748446bb6
|
49 50 |
list_for_each_entry_safe(page, next, freelist, lru) { |
6bace090a
|
51 |
unsigned long pfn = page_to_pfn(page); |
748446bb6
|
52 53 |
list_del(&page->lru); __free_page(page); |
6bace090a
|
54 55 |
if (pfn > high_pfn) high_pfn = pfn; |
748446bb6
|
56 |
} |
6bace090a
|
57 |
return high_pfn; |
748446bb6
|
58 |
} |
ff9543fd3
|
59 60 |
static void map_pages(struct list_head *list) { |
66c64223a
|
61 62 63 64 65 66 67 68 69 |
unsigned int i, order, nr_pages; struct page *page, *next; LIST_HEAD(tmp_list); list_for_each_entry_safe(page, next, list, lru) { list_del(&page->lru); order = page_private(page); nr_pages = 1 << order; |
66c64223a
|
70 |
|
46f24fd85
|
71 |
post_alloc_hook(page, order, __GFP_MOVABLE); |
66c64223a
|
72 73 |
if (order) split_page(page, order); |
ff9543fd3
|
74 |
|
66c64223a
|
75 76 77 78 |
for (i = 0; i < nr_pages; i++) { list_add(&page->lru, &tmp_list); page++; } |
ff9543fd3
|
79 |
} |
66c64223a
|
80 81 |
list_splice(&tmp_list, list); |
ff9543fd3
|
82 |
} |
47118af07
|
83 84 85 86 |
static inline bool migrate_async_suitable(int migratetype) { return is_migrate_cma(migratetype) || migratetype == MIGRATE_MOVABLE; } |
bb13ffeb9
|
87 |
#ifdef CONFIG_COMPACTION |
24e2716f6
|
88 |
|
bda807d44
|
89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 |
int PageMovable(struct page *page) { struct address_space *mapping; VM_BUG_ON_PAGE(!PageLocked(page), page); if (!__PageMovable(page)) return 0; mapping = page_mapping(page); if (mapping && mapping->a_ops && mapping->a_ops->isolate_page) return 1; return 0; } EXPORT_SYMBOL(PageMovable); void __SetPageMovable(struct page *page, struct address_space *mapping) { VM_BUG_ON_PAGE(!PageLocked(page), page); VM_BUG_ON_PAGE((unsigned long)mapping & PAGE_MAPPING_MOVABLE, page); page->mapping = (void *)((unsigned long)mapping | PAGE_MAPPING_MOVABLE); } EXPORT_SYMBOL(__SetPageMovable); void __ClearPageMovable(struct page *page) { VM_BUG_ON_PAGE(!PageLocked(page), page); VM_BUG_ON_PAGE(!PageMovable(page), page); /* * Clear registered address_space val with keeping PAGE_MAPPING_MOVABLE * flag so that VM can catch up released page by driver after isolation. * With it, VM migration doesn't try to put it back. */ page->mapping = (void *)((unsigned long)page->mapping & PAGE_MAPPING_MOVABLE); } EXPORT_SYMBOL(__ClearPageMovable); |
24e2716f6
|
126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 |
/* Do not skip compaction more than 64 times */ #define COMPACT_MAX_DEFER_SHIFT 6 /* * Compaction is deferred when compaction fails to result in a page * allocation success. 1 << compact_defer_limit compactions are skipped up * to a limit of 1 << COMPACT_MAX_DEFER_SHIFT */ void defer_compaction(struct zone *zone, int order) { zone->compact_considered = 0; zone->compact_defer_shift++; if (order < zone->compact_order_failed) zone->compact_order_failed = order; if (zone->compact_defer_shift > COMPACT_MAX_DEFER_SHIFT) zone->compact_defer_shift = COMPACT_MAX_DEFER_SHIFT; trace_mm_compaction_defer_compaction(zone, order); } /* Returns true if compaction should be skipped this time */ bool compaction_deferred(struct zone *zone, int order) { unsigned long defer_limit = 1UL << zone->compact_defer_shift; if (order < zone->compact_order_failed) return false; /* Avoid possible overflow */ if (++zone->compact_considered > defer_limit) zone->compact_considered = defer_limit; if (zone->compact_considered >= defer_limit) return false; trace_mm_compaction_deferred(zone, order); return true; } /* * Update defer tracking counters after successful compaction of given order, * which means an allocation either succeeded (alloc_success == true) or is * expected to succeed. */ void compaction_defer_reset(struct zone *zone, int order, bool alloc_success) { if (alloc_success) { zone->compact_considered = 0; zone->compact_defer_shift = 0; } if (order >= zone->compact_order_failed) zone->compact_order_failed = order + 1; trace_mm_compaction_defer_reset(zone, order); } /* Returns true if restarting compaction after many failures */ bool compaction_restarting(struct zone *zone, int order) { if (order < zone->compact_order_failed) return false; return zone->compact_defer_shift == COMPACT_MAX_DEFER_SHIFT && zone->compact_considered >= 1UL << zone->compact_defer_shift; } |
bb13ffeb9
|
195 196 197 198 199 200 201 202 203 |
/* Returns true if the pageblock should be scanned for pages to isolate. */ static inline bool isolation_suitable(struct compact_control *cc, struct page *page) { if (cc->ignore_skip_hint) return true; return !get_pageblock_skip(page); } |
02333641e
|
204 205 206 207 |
static void reset_cached_positions(struct zone *zone) { zone->compact_cached_migrate_pfn[0] = zone->zone_start_pfn; zone->compact_cached_migrate_pfn[1] = zone->zone_start_pfn; |
623446e4d
|
208 |
zone->compact_cached_free_pfn = |
06b6640a3
|
209 |
pageblock_start_pfn(zone_end_pfn(zone) - 1); |
02333641e
|
210 |
} |
bb13ffeb9
|
211 212 213 214 215 |
/* * This function is called to clear all cached information on pageblocks that * should be skipped for page isolation when the migrate and free page scanner * meet. */ |
62997027c
|
216 |
static void __reset_isolation_suitable(struct zone *zone) |
bb13ffeb9
|
217 218 |
{ unsigned long start_pfn = zone->zone_start_pfn; |
108bcc96e
|
219 |
unsigned long end_pfn = zone_end_pfn(zone); |
bb13ffeb9
|
220 |
unsigned long pfn; |
62997027c
|
221 |
zone->compact_blockskip_flush = false; |
bb13ffeb9
|
222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 |
/* Walk the zone and mark every pageblock as suitable for isolation */ for (pfn = start_pfn; pfn < end_pfn; pfn += pageblock_nr_pages) { struct page *page; cond_resched(); if (!pfn_valid(pfn)) continue; page = pfn_to_page(pfn); if (zone != page_zone(page)) continue; clear_pageblock_skip(page); } |
02333641e
|
238 239 |
reset_cached_positions(zone); |
bb13ffeb9
|
240 |
} |
62997027c
|
241 242 243 244 245 246 247 248 249 250 251 252 253 254 |
void reset_isolation_suitable(pg_data_t *pgdat) { int zoneid; for (zoneid = 0; zoneid < MAX_NR_ZONES; zoneid++) { struct zone *zone = &pgdat->node_zones[zoneid]; if (!populated_zone(zone)) continue; /* Only flush if a full compaction finished recently */ if (zone->compact_blockskip_flush) __reset_isolation_suitable(zone); } } |
bb13ffeb9
|
255 256 |
/* * If no pages were isolated then mark this pageblock to be skipped in the |
62997027c
|
257 |
* future. The information is later cleared by __reset_isolation_suitable(). |
bb13ffeb9
|
258 |
*/ |
c89511ab2
|
259 260 |
static void update_pageblock_skip(struct compact_control *cc, struct page *page, unsigned long nr_isolated, |
edc2ca612
|
261 |
bool migrate_scanner) |
bb13ffeb9
|
262 |
{ |
c89511ab2
|
263 |
struct zone *zone = cc->zone; |
35979ef33
|
264 |
unsigned long pfn; |
6815bf3f2
|
265 266 267 |
if (cc->ignore_skip_hint) return; |
bb13ffeb9
|
268 269 |
if (!page) return; |
35979ef33
|
270 271 |
if (nr_isolated) return; |
edc2ca612
|
272 |
set_pageblock_skip(page); |
c89511ab2
|
273 |
|
35979ef33
|
274 275 276 277 |
pfn = page_to_pfn(page); /* Update where async and sync compaction should restart */ if (migrate_scanner) { |
35979ef33
|
278 279 |
if (pfn > zone->compact_cached_migrate_pfn[0]) zone->compact_cached_migrate_pfn[0] = pfn; |
e0b9daeb4
|
280 281 |
if (cc->mode != MIGRATE_ASYNC && pfn > zone->compact_cached_migrate_pfn[1]) |
35979ef33
|
282 283 |
zone->compact_cached_migrate_pfn[1] = pfn; } else { |
35979ef33
|
284 285 |
if (pfn < zone->compact_cached_free_pfn) zone->compact_cached_free_pfn = pfn; |
c89511ab2
|
286 |
} |
bb13ffeb9
|
287 288 289 290 291 292 293 |
} #else static inline bool isolation_suitable(struct compact_control *cc, struct page *page) { return true; } |
c89511ab2
|
294 295 |
static void update_pageblock_skip(struct compact_control *cc, struct page *page, unsigned long nr_isolated, |
edc2ca612
|
296 |
bool migrate_scanner) |
bb13ffeb9
|
297 298 299 |
{ } #endif /* CONFIG_COMPACTION */ |
8b44d2791
|
300 301 302 303 304 305 306 307 308 309 |
/* * Compaction requires the taking of some coarse locks that are potentially * very heavily contended. For async compaction, back out if the lock cannot * be taken immediately. For sync compaction, spin on the lock if needed. * * Returns true if the lock is held * Returns false if the lock is not held and compaction should abort */ static bool compact_trylock_irqsave(spinlock_t *lock, unsigned long *flags, struct compact_control *cc) |
2a1402aa0
|
310 |
{ |
8b44d2791
|
311 312 |
if (cc->mode == MIGRATE_ASYNC) { if (!spin_trylock_irqsave(lock, *flags)) { |
c3486f537
|
313 |
cc->contended = true; |
8b44d2791
|
314 315 316 317 318 |
return false; } } else { spin_lock_irqsave(lock, *flags); } |
1f9efdef4
|
319 |
|
8b44d2791
|
320 |
return true; |
2a1402aa0
|
321 |
} |
85aa125f0
|
322 |
/* |
c67fe3752
|
323 |
* Compaction requires the taking of some coarse locks that are potentially |
8b44d2791
|
324 325 326 327 328 329 330 |
* very heavily contended. The lock should be periodically unlocked to avoid * having disabled IRQs for a long time, even when there is nobody waiting on * the lock. It might also be that allowing the IRQs will result in * need_resched() becoming true. If scheduling is needed, async compaction * aborts. Sync compaction schedules. * Either compaction type will also abort if a fatal signal is pending. * In either case if the lock was locked, it is dropped and not regained. |
c67fe3752
|
331 |
* |
8b44d2791
|
332 333 334 335 |
* Returns true if compaction should abort due to fatal signal pending, or * async compaction due to need_resched() * Returns false when compaction can continue (sync compaction might have * scheduled) |
c67fe3752
|
336 |
*/ |
8b44d2791
|
337 338 |
static bool compact_unlock_should_abort(spinlock_t *lock, unsigned long flags, bool *locked, struct compact_control *cc) |
c67fe3752
|
339 |
{ |
8b44d2791
|
340 341 342 343 |
if (*locked) { spin_unlock_irqrestore(lock, flags); *locked = false; } |
1f9efdef4
|
344 |
|
8b44d2791
|
345 |
if (fatal_signal_pending(current)) { |
c3486f537
|
346 |
cc->contended = true; |
8b44d2791
|
347 348 |
return true; } |
c67fe3752
|
349 |
|
8b44d2791
|
350 |
if (need_resched()) { |
e0b9daeb4
|
351 |
if (cc->mode == MIGRATE_ASYNC) { |
c3486f537
|
352 |
cc->contended = true; |
8b44d2791
|
353 |
return true; |
c67fe3752
|
354 |
} |
c67fe3752
|
355 |
cond_resched(); |
c67fe3752
|
356 |
} |
8b44d2791
|
357 |
return false; |
c67fe3752
|
358 |
} |
be9765722
|
359 360 361 |
/* * Aside from avoiding lock contention, compaction also periodically checks * need_resched() and either schedules in sync compaction or aborts async |
8b44d2791
|
362 |
* compaction. This is similar to what compact_unlock_should_abort() does, but |
be9765722
|
363 364 365 366 367 368 369 370 371 372 |
* is used where no lock is concerned. * * Returns false when no scheduling was needed, or sync compaction scheduled. * Returns true when async compaction should abort. */ static inline bool compact_should_abort(struct compact_control *cc) { /* async compaction aborts if contended */ if (need_resched()) { if (cc->mode == MIGRATE_ASYNC) { |
c3486f537
|
373 |
cc->contended = true; |
be9765722
|
374 375 376 377 378 379 380 381 |
return true; } cond_resched(); } return false; } |
c67fe3752
|
382 |
/* |
9e4be4708
|
383 384 385 |
* Isolate free pages onto a private freelist. If @strict is true, will abort * returning 0 on any invalid PFNs or non-free pages inside of the pageblock * (even though it may still end up isolating some pages). |
85aa125f0
|
386 |
*/ |
f40d1e42b
|
387 |
static unsigned long isolate_freepages_block(struct compact_control *cc, |
e14c720ef
|
388 |
unsigned long *start_pfn, |
85aa125f0
|
389 390 391 |
unsigned long end_pfn, struct list_head *freelist, bool strict) |
748446bb6
|
392 |
{ |
b7aba6984
|
393 |
int nr_scanned = 0, total_isolated = 0; |
bb13ffeb9
|
394 |
struct page *cursor, *valid_page = NULL; |
b8b2d8253
|
395 |
unsigned long flags = 0; |
f40d1e42b
|
396 |
bool locked = false; |
e14c720ef
|
397 |
unsigned long blockpfn = *start_pfn; |
66c64223a
|
398 |
unsigned int order; |
748446bb6
|
399 |
|
748446bb6
|
400 |
cursor = pfn_to_page(blockpfn); |
f40d1e42b
|
401 |
/* Isolate free pages. */ |
748446bb6
|
402 |
for (; blockpfn < end_pfn; blockpfn++, cursor++) { |
66c64223a
|
403 |
int isolated; |
748446bb6
|
404 |
struct page *page = cursor; |
8b44d2791
|
405 406 407 408 409 410 411 412 413 |
/* * Periodically drop the lock (if held) regardless of its * contention, to give chance to IRQs. Abort if fatal signal * pending or async compaction detects need_resched() */ if (!(blockpfn % SWAP_CLUSTER_MAX) && compact_unlock_should_abort(&cc->zone->lock, flags, &locked, cc)) break; |
b7aba6984
|
414 |
nr_scanned++; |
f40d1e42b
|
415 |
if (!pfn_valid_within(blockpfn)) |
2af120bc0
|
416 |
goto isolate_fail; |
bb13ffeb9
|
417 418 |
if (!valid_page) valid_page = page; |
9fcd6d2e0
|
419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 |
/* * For compound pages such as THP and hugetlbfs, we can save * potentially a lot of iterations if we skip them at once. * The check is racy, but we can consider only valid values * and the only danger is skipping too much. */ if (PageCompound(page)) { unsigned int comp_order = compound_order(page); if (likely(comp_order < MAX_ORDER)) { blockpfn += (1UL << comp_order) - 1; cursor += (1UL << comp_order) - 1; } goto isolate_fail; } |
f40d1e42b
|
436 |
if (!PageBuddy(page)) |
2af120bc0
|
437 |
goto isolate_fail; |
f40d1e42b
|
438 439 |
/* |
69b7189f1
|
440 441 442 443 444 |
* If we already hold the lock, we can skip some rechecking. * Note that if we hold the lock now, checked_pageblock was * already set in some previous iteration (or strict is true), * so it is correct to skip the suitable migration target * recheck as well. |
f40d1e42b
|
445 |
*/ |
69b7189f1
|
446 447 448 449 450 451 452 453 454 |
if (!locked) { /* * The zone lock must be held to isolate freepages. * Unfortunately this is a very coarse lock and can be * heavily contended if there are parallel allocations * or parallel compactions. For async compaction do not * spin on the lock and we acquire the lock as late as * possible. */ |
8b44d2791
|
455 456 |
locked = compact_trylock_irqsave(&cc->zone->lock, &flags, cc); |
69b7189f1
|
457 458 |
if (!locked) break; |
f40d1e42b
|
459 |
|
69b7189f1
|
460 461 462 463 |
/* Recheck this is a buddy page under lock */ if (!PageBuddy(page)) goto isolate_fail; } |
748446bb6
|
464 |
|
66c64223a
|
465 466 467 |
/* Found a free page, will break it into order-0 pages */ order = page_order(page); isolated = __isolate_free_page(page, order); |
a4f04f2c6
|
468 469 |
if (!isolated) break; |
66c64223a
|
470 |
set_page_private(page, order); |
a4f04f2c6
|
471 |
|
748446bb6
|
472 |
total_isolated += isolated; |
a4f04f2c6
|
473 |
cc->nr_freepages += isolated; |
66c64223a
|
474 |
list_add_tail(&page->lru, freelist); |
a4f04f2c6
|
475 476 477 |
if (!strict && cc->nr_migratepages <= cc->nr_freepages) { blockpfn += isolated; break; |
748446bb6
|
478 |
} |
a4f04f2c6
|
479 480 481 482 |
/* Advance to the end of split page */ blockpfn += isolated - 1; cursor += isolated - 1; continue; |
2af120bc0
|
483 484 485 486 487 488 |
isolate_fail: if (strict) break; else continue; |
748446bb6
|
489 |
} |
a4f04f2c6
|
490 491 |
if (locked) spin_unlock_irqrestore(&cc->zone->lock, flags); |
9fcd6d2e0
|
492 493 494 495 496 497 |
/* * There is a tiny chance that we have read bogus compound_order(), * so be careful to not go outside of the pageblock. */ if (unlikely(blockpfn > end_pfn)) blockpfn = end_pfn; |
e34d85f0e
|
498 499 |
trace_mm_compaction_isolate_freepages(*start_pfn, blockpfn, nr_scanned, total_isolated); |
e14c720ef
|
500 501 |
/* Record how far we have got within the block */ *start_pfn = blockpfn; |
f40d1e42b
|
502 503 504 505 506 |
/* * If strict isolation is requested by CMA then check that all the * pages requested were isolated. If there were any failures, 0 is * returned and CMA will fail. */ |
2af120bc0
|
507 |
if (strict && blockpfn < end_pfn) |
f40d1e42b
|
508 |
total_isolated = 0; |
bb13ffeb9
|
509 510 |
/* Update the pageblock-skip if the whole pageblock was scanned */ if (blockpfn == end_pfn) |
edc2ca612
|
511 |
update_pageblock_skip(cc, valid_page, total_isolated, false); |
bb13ffeb9
|
512 |
|
010fc29a4
|
513 |
count_compact_events(COMPACTFREE_SCANNED, nr_scanned); |
397487db6
|
514 |
if (total_isolated) |
010fc29a4
|
515 |
count_compact_events(COMPACTISOLATED, total_isolated); |
748446bb6
|
516 517 |
return total_isolated; } |
85aa125f0
|
518 519 520 521 522 523 524 525 526 527 528 529 530 |
/** * isolate_freepages_range() - isolate free pages. * @start_pfn: The first PFN to start isolating. * @end_pfn: The one-past-last PFN. * * Non-free pages, invalid PFNs, or zone boundaries within the * [start_pfn, end_pfn) range are considered errors, cause function to * undo its actions and return zero. * * Otherwise, function returns one-past-the-last PFN of isolated page * (which may be greater then end_pfn if end fell in a middle of * a free page). */ |
ff9543fd3
|
531 |
unsigned long |
bb13ffeb9
|
532 533 |
isolate_freepages_range(struct compact_control *cc, unsigned long start_pfn, unsigned long end_pfn) |
85aa125f0
|
534 |
{ |
e1409c325
|
535 |
unsigned long isolated, pfn, block_start_pfn, block_end_pfn; |
85aa125f0
|
536 |
LIST_HEAD(freelist); |
7d49d8868
|
537 |
pfn = start_pfn; |
06b6640a3
|
538 |
block_start_pfn = pageblock_start_pfn(pfn); |
e1409c325
|
539 540 |
if (block_start_pfn < cc->zone->zone_start_pfn) block_start_pfn = cc->zone->zone_start_pfn; |
06b6640a3
|
541 |
block_end_pfn = pageblock_end_pfn(pfn); |
7d49d8868
|
542 543 |
for (; pfn < end_pfn; pfn += isolated, |
e1409c325
|
544 |
block_start_pfn = block_end_pfn, |
7d49d8868
|
545 |
block_end_pfn += pageblock_nr_pages) { |
e14c720ef
|
546 547 |
/* Protect pfn from changing by isolate_freepages_block */ unsigned long isolate_start_pfn = pfn; |
85aa125f0
|
548 |
|
85aa125f0
|
549 |
block_end_pfn = min(block_end_pfn, end_pfn); |
584200163
|
550 551 552 553 554 555 |
/* * pfn could pass the block_end_pfn if isolated freepage * is more than pageblock order. In this case, we adjust * scanning range to right one. */ if (pfn >= block_end_pfn) { |
06b6640a3
|
556 557 |
block_start_pfn = pageblock_start_pfn(pfn); block_end_pfn = pageblock_end_pfn(pfn); |
584200163
|
558 559 |
block_end_pfn = min(block_end_pfn, end_pfn); } |
e1409c325
|
560 561 |
if (!pageblock_pfn_to_page(block_start_pfn, block_end_pfn, cc->zone)) |
7d49d8868
|
562 |
break; |
e14c720ef
|
563 564 |
isolated = isolate_freepages_block(cc, &isolate_start_pfn, block_end_pfn, &freelist, true); |
85aa125f0
|
565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 |
/* * In strict mode, isolate_freepages_block() returns 0 if * there are any holes in the block (ie. invalid PFNs or * non-free pages). */ if (!isolated) break; /* * If we managed to isolate pages, it is always (1 << n) * * pageblock_nr_pages for some non-negative n. (Max order * page may span two pageblocks). */ } |
66c64223a
|
580 |
/* __isolate_free_page() does not map the pages */ |
85aa125f0
|
581 582 583 584 585 586 587 588 589 590 591 |
map_pages(&freelist); if (pfn < end_pfn) { /* Loop terminated early, cleanup. */ release_freepages(&freelist); return 0; } /* We don't use freelists for anything. */ return pfn; } |
748446bb6
|
592 |
/* Update the number of anon and file isolated pages in the zone */ |
edc2ca612
|
593 |
static void acct_isolated(struct zone *zone, struct compact_control *cc) |
748446bb6
|
594 595 |
{ struct page *page; |
b9e84ac15
|
596 |
unsigned int count[2] = { 0, }; |
748446bb6
|
597 |
|
edc2ca612
|
598 599 |
if (list_empty(&cc->migratepages)) return; |
b9e84ac15
|
600 601 |
list_for_each_entry(page, &cc->migratepages, lru) count[!!page_is_file_cache(page)]++; |
748446bb6
|
602 |
|
599d0c954
|
603 604 |
mod_node_page_state(zone->zone_pgdat, NR_ISOLATED_ANON, count[0]); mod_node_page_state(zone->zone_pgdat, NR_ISOLATED_FILE, count[1]); |
748446bb6
|
605 606 607 608 609 |
} /* Similar to reclaim, but different enough that they don't share logic */ static bool too_many_isolated(struct zone *zone) { |
bc6930457
|
610 |
unsigned long active, inactive, isolated; |
748446bb6
|
611 |
|
599d0c954
|
612 613 614 615 616 617 |
inactive = node_page_state(zone->zone_pgdat, NR_INACTIVE_FILE) + node_page_state(zone->zone_pgdat, NR_INACTIVE_ANON); active = node_page_state(zone->zone_pgdat, NR_ACTIVE_FILE) + node_page_state(zone->zone_pgdat, NR_ACTIVE_ANON); isolated = node_page_state(zone->zone_pgdat, NR_ISOLATED_FILE) + node_page_state(zone->zone_pgdat, NR_ISOLATED_ANON); |
748446bb6
|
618 |
|
bc6930457
|
619 |
return isolated > (inactive + active) / 2; |
748446bb6
|
620 |
} |
2fe86e000
|
621 |
/** |
edc2ca612
|
622 623 |
* isolate_migratepages_block() - isolate all migrate-able pages within * a single pageblock |
2fe86e000
|
624 |
* @cc: Compaction control structure. |
edc2ca612
|
625 626 627 |
* @low_pfn: The first PFN to isolate * @end_pfn: The one-past-the-last PFN to isolate, within same pageblock * @isolate_mode: Isolation mode to be used. |
2fe86e000
|
628 629 |
* * Isolate all pages that can be migrated from the range specified by |
edc2ca612
|
630 631 632 633 |
* [low_pfn, end_pfn). The range is expected to be within same pageblock. * Returns zero if there is a fatal signal pending, otherwise PFN of the * first page that was not scanned (which may be both less, equal to or more * than end_pfn). |
2fe86e000
|
634 |
* |
edc2ca612
|
635 636 637 |
* The pages are isolated on cc->migratepages list (not required to be empty), * and cc->nr_migratepages is updated accordingly. The cc->migrate_pfn field * is neither read nor updated. |
748446bb6
|
638 |
*/ |
edc2ca612
|
639 640 641 |
static unsigned long isolate_migratepages_block(struct compact_control *cc, unsigned long low_pfn, unsigned long end_pfn, isolate_mode_t isolate_mode) |
748446bb6
|
642 |
{ |
edc2ca612
|
643 |
struct zone *zone = cc->zone; |
b7aba6984
|
644 |
unsigned long nr_scanned = 0, nr_isolated = 0; |
fa9add641
|
645 |
struct lruvec *lruvec; |
b8b2d8253
|
646 |
unsigned long flags = 0; |
2a1402aa0
|
647 |
bool locked = false; |
bb13ffeb9
|
648 |
struct page *page = NULL, *valid_page = NULL; |
e34d85f0e
|
649 |
unsigned long start_pfn = low_pfn; |
fdd048e12
|
650 651 |
bool skip_on_failure = false; unsigned long next_skip_pfn = 0; |
748446bb6
|
652 |
|
748446bb6
|
653 654 655 656 657 658 |
/* * Ensure that there are not too many pages isolated from the LRU * list by either parallel reclaimers or compaction. If there are, * delay for some time until fewer pages are isolated */ while (unlikely(too_many_isolated(zone))) { |
f9e35b3b4
|
659 |
/* async migration should just abort */ |
e0b9daeb4
|
660 |
if (cc->mode == MIGRATE_ASYNC) |
2fe86e000
|
661 |
return 0; |
f9e35b3b4
|
662 |
|
748446bb6
|
663 664 665 |
congestion_wait(BLK_RW_ASYNC, HZ/10); if (fatal_signal_pending(current)) |
2fe86e000
|
666 |
return 0; |
748446bb6
|
667 |
} |
be9765722
|
668 669 |
if (compact_should_abort(cc)) return 0; |
aeef4b838
|
670 |
|
fdd048e12
|
671 672 673 674 |
if (cc->direct_compaction && (cc->mode == MIGRATE_ASYNC)) { skip_on_failure = true; next_skip_pfn = block_end_pfn(low_pfn, cc->order); } |
748446bb6
|
675 |
/* Time to isolate some pages for migration */ |
748446bb6
|
676 |
for (; low_pfn < end_pfn; low_pfn++) { |
29c0dde83
|
677 |
|
fdd048e12
|
678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 |
if (skip_on_failure && low_pfn >= next_skip_pfn) { /* * We have isolated all migration candidates in the * previous order-aligned block, and did not skip it due * to failure. We should migrate the pages now and * hopefully succeed compaction. */ if (nr_isolated) break; /* * We failed to isolate in the previous order-aligned * block. Set the new boundary to the end of the * current block. Note we can't simply increase * next_skip_pfn by 1 << order, as low_pfn might have * been incremented by a higher number due to skipping * a compound or a high-order buddy page in the * previous loop iteration. */ next_skip_pfn = block_end_pfn(low_pfn, cc->order); } |
8b44d2791
|
699 700 701 702 703 704 |
/* * Periodically drop the lock (if held) regardless of its * contention, to give chance to IRQs. Abort async compaction * if contended. */ if (!(low_pfn % SWAP_CLUSTER_MAX) |
a52633d8e
|
705 |
&& compact_unlock_should_abort(zone_lru_lock(zone), flags, |
8b44d2791
|
706 707 |
&locked, cc)) break; |
c67fe3752
|
708 |
|
748446bb6
|
709 |
if (!pfn_valid_within(low_pfn)) |
fdd048e12
|
710 |
goto isolate_fail; |
b7aba6984
|
711 |
nr_scanned++; |
748446bb6
|
712 |
|
748446bb6
|
713 |
page = pfn_to_page(low_pfn); |
dc9086004
|
714 |
|
bb13ffeb9
|
715 716 |
if (!valid_page) valid_page = page; |
6c14466cc
|
717 |
/* |
99c0fd5e5
|
718 719 720 721 |
* Skip if free. We read page order here without zone lock * which is generally unsafe, but the race window is small and * the worst thing that can happen is that we skip some * potential isolation targets. |
6c14466cc
|
722 |
*/ |
99c0fd5e5
|
723 724 725 726 727 728 729 730 731 732 |
if (PageBuddy(page)) { unsigned long freepage_order = page_order_unsafe(page); /* * Without lock, we cannot be sure that what we got is * a valid page order. Consider only values in the * valid order range to prevent low_pfn overflow. */ if (freepage_order > 0 && freepage_order < MAX_ORDER) low_pfn += (1UL << freepage_order) - 1; |
748446bb6
|
733 |
continue; |
99c0fd5e5
|
734 |
} |
748446bb6
|
735 |
|
9927af740
|
736 |
/* |
29c0dde83
|
737 738 739 740 741 |
* Regardless of being on LRU, compound pages such as THP and * hugetlbfs are not to be compacted. We can potentially save * a lot of iterations if we skip them at once. The check is * racy, but we can consider only valid values and the only * danger is skipping too much. |
bc835011a
|
742 |
*/ |
29c0dde83
|
743 744 745 746 747 |
if (PageCompound(page)) { unsigned int comp_order = compound_order(page); if (likely(comp_order < MAX_ORDER)) low_pfn += (1UL << comp_order) - 1; |
edc2ca612
|
748 |
|
fdd048e12
|
749 |
goto isolate_fail; |
2a1402aa0
|
750 |
} |
bda807d44
|
751 752 753 754 755 756 |
/* * Check may be lockless but that's ok as we recheck later. * It's possible to migrate LRU and non-lru movable pages. * Skip any other type of page */ if (!PageLRU(page)) { |
bda807d44
|
757 758 759 760 761 762 763 |
/* * __PageMovable can return false positive so we need * to verify it under page_lock. */ if (unlikely(__PageMovable(page)) && !PageIsolated(page)) { if (locked) { |
a52633d8e
|
764 |
spin_unlock_irqrestore(zone_lru_lock(zone), |
bda807d44
|
765 766 767 768 769 770 771 |
flags); locked = false; } if (isolate_movable_page(page, isolate_mode)) goto isolate_success; } |
fdd048e12
|
772 |
goto isolate_fail; |
bda807d44
|
773 |
} |
29c0dde83
|
774 |
|
119d6d59d
|
775 776 777 778 779 780 781 |
/* * Migration will fail if an anonymous page is pinned in memory, * so avoid taking lru_lock and isolating it unnecessarily in an * admittedly racy check. */ if (!page_mapping(page) && page_count(page) > page_mapcount(page)) |
fdd048e12
|
782 |
goto isolate_fail; |
119d6d59d
|
783 |
|
69b7189f1
|
784 785 |
/* If we already hold the lock, we can skip some rechecking */ if (!locked) { |
a52633d8e
|
786 |
locked = compact_trylock_irqsave(zone_lru_lock(zone), |
8b44d2791
|
787 |
&flags, cc); |
69b7189f1
|
788 789 |
if (!locked) break; |
2a1402aa0
|
790 |
|
29c0dde83
|
791 |
/* Recheck PageLRU and PageCompound under lock */ |
69b7189f1
|
792 |
if (!PageLRU(page)) |
fdd048e12
|
793 |
goto isolate_fail; |
29c0dde83
|
794 795 796 797 798 799 800 801 |
/* * Page become compound since the non-locked check, * and it's on LRU. It can only be a THP so the order * is safe to read and it's 0 for tail pages. */ if (unlikely(PageCompound(page))) { low_pfn += (1UL << compound_order(page)) - 1; |
fdd048e12
|
802 |
goto isolate_fail; |
69b7189f1
|
803 |
} |
bc835011a
|
804 |
} |
599d0c954
|
805 |
lruvec = mem_cgroup_page_lruvec(page, zone->zone_pgdat); |
fa9add641
|
806 |
|
748446bb6
|
807 |
/* Try isolate the page */ |
edc2ca612
|
808 |
if (__isolate_lru_page(page, isolate_mode) != 0) |
fdd048e12
|
809 |
goto isolate_fail; |
748446bb6
|
810 |
|
29c0dde83
|
811 |
VM_BUG_ON_PAGE(PageCompound(page), page); |
bc835011a
|
812 |
|
748446bb6
|
813 |
/* Successfully isolated */ |
fa9add641
|
814 |
del_page_from_lru_list(page, lruvec, page_lru(page)); |
b6c750163
|
815 816 |
isolate_success: |
fdd048e12
|
817 |
list_add(&page->lru, &cc->migratepages); |
748446bb6
|
818 |
cc->nr_migratepages++; |
b7aba6984
|
819 |
nr_isolated++; |
748446bb6
|
820 |
|
a34753d27
|
821 822 823 824 825 826 827 828 |
/* * Record where we could have freed pages by migration and not * yet flushed them to buddy allocator. * - this is the lowest page that was isolated and likely be * then freed by migration. */ if (!cc->last_migrated_pfn) cc->last_migrated_pfn = low_pfn; |
748446bb6
|
829 |
/* Avoid isolating too much */ |
31b8384a5
|
830 831 |
if (cc->nr_migratepages == COMPACT_CLUSTER_MAX) { ++low_pfn; |
748446bb6
|
832 |
break; |
31b8384a5
|
833 |
} |
fdd048e12
|
834 835 836 837 838 839 840 841 842 843 844 845 846 |
continue; isolate_fail: if (!skip_on_failure) continue; /* * We have isolated some pages, but then failed. Release them * instead of migrating, as we cannot form the cc->order buddy * page anyway. */ if (nr_isolated) { if (locked) { |
a52633d8e
|
847 |
spin_unlock_irqrestore(zone_lru_lock(zone), flags); |
fdd048e12
|
848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 |
locked = false; } acct_isolated(zone, cc); putback_movable_pages(&cc->migratepages); cc->nr_migratepages = 0; cc->last_migrated_pfn = 0; nr_isolated = 0; } if (low_pfn < next_skip_pfn) { low_pfn = next_skip_pfn - 1; /* * The check near the loop beginning would have updated * next_skip_pfn too, but this is a bit simpler. */ next_skip_pfn += 1UL << cc->order; } |
748446bb6
|
865 |
} |
99c0fd5e5
|
866 867 868 869 870 871 |
/* * The PageBuddy() check could have potentially brought us outside * the range to be scanned. */ if (unlikely(low_pfn > end_pfn)) low_pfn = end_pfn; |
c67fe3752
|
872 |
if (locked) |
a52633d8e
|
873 |
spin_unlock_irqrestore(zone_lru_lock(zone), flags); |
748446bb6
|
874 |
|
50b5b094e
|
875 876 877 |
/* * Update the pageblock-skip information and cached scanner pfn, * if the whole pageblock was scanned without isolating any page. |
50b5b094e
|
878 |
*/ |
35979ef33
|
879 |
if (low_pfn == end_pfn) |
edc2ca612
|
880 |
update_pageblock_skip(cc, valid_page, nr_isolated, true); |
bb13ffeb9
|
881 |
|
e34d85f0e
|
882 883 |
trace_mm_compaction_isolate_migratepages(start_pfn, low_pfn, nr_scanned, nr_isolated); |
b7aba6984
|
884 |
|
010fc29a4
|
885 |
count_compact_events(COMPACTMIGRATE_SCANNED, nr_scanned); |
397487db6
|
886 |
if (nr_isolated) |
010fc29a4
|
887 |
count_compact_events(COMPACTISOLATED, nr_isolated); |
397487db6
|
888 |
|
2fe86e000
|
889 890 |
return low_pfn; } |
edc2ca612
|
891 892 893 894 895 896 897 898 899 900 901 902 903 904 |
/** * isolate_migratepages_range() - isolate migrate-able pages in a PFN range * @cc: Compaction control structure. * @start_pfn: The first PFN to start isolating. * @end_pfn: The one-past-last PFN. * * Returns zero if isolation fails fatally due to e.g. pending signal. * Otherwise, function returns one-past-the-last PFN of isolated page * (which may be greater than end_pfn if end fell in a middle of a THP page). */ unsigned long isolate_migratepages_range(struct compact_control *cc, unsigned long start_pfn, unsigned long end_pfn) { |
e1409c325
|
905 |
unsigned long pfn, block_start_pfn, block_end_pfn; |
edc2ca612
|
906 907 908 |
/* Scan block by block. First and last block may be incomplete */ pfn = start_pfn; |
06b6640a3
|
909 |
block_start_pfn = pageblock_start_pfn(pfn); |
e1409c325
|
910 911 |
if (block_start_pfn < cc->zone->zone_start_pfn) block_start_pfn = cc->zone->zone_start_pfn; |
06b6640a3
|
912 |
block_end_pfn = pageblock_end_pfn(pfn); |
edc2ca612
|
913 914 |
for (; pfn < end_pfn; pfn = block_end_pfn, |
e1409c325
|
915 |
block_start_pfn = block_end_pfn, |
edc2ca612
|
916 917 918 |
block_end_pfn += pageblock_nr_pages) { block_end_pfn = min(block_end_pfn, end_pfn); |
e1409c325
|
919 920 |
if (!pageblock_pfn_to_page(block_start_pfn, block_end_pfn, cc->zone)) |
edc2ca612
|
921 922 923 924 |
continue; pfn = isolate_migratepages_block(cc, pfn, block_end_pfn, ISOLATE_UNEVICTABLE); |
14af4a5e9
|
925 |
if (!pfn) |
edc2ca612
|
926 |
break; |
6ea41c0c0
|
927 928 929 |
if (cc->nr_migratepages == COMPACT_CLUSTER_MAX) break; |
edc2ca612
|
930 931 932 933 934 |
} acct_isolated(cc->zone, cc); return pfn; } |
ff9543fd3
|
935 936 |
#endif /* CONFIG_COMPACTION || CONFIG_CMA */ #ifdef CONFIG_COMPACTION |
018e9a49a
|
937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 |
/* Returns true if the page is within a block suitable for migration to */ static bool suitable_migration_target(struct page *page) { /* If the page is a large free page, then disallow migration */ if (PageBuddy(page)) { /* * We are checking page_order without zone->lock taken. But * the only small danger is that we skip a potentially suitable * pageblock, so it's not worth to check order for valid range. */ if (page_order_unsafe(page) >= pageblock_order) return false; } /* If the block is MIGRATE_MOVABLE or MIGRATE_CMA, allow migration */ if (migrate_async_suitable(get_pageblock_migratetype(page))) return true; /* Otherwise skip the block */ return false; } |
2fe86e000
|
959 |
/* |
f2849aa09
|
960 961 962 963 964 965 966 967 968 969 |
* Test whether the free scanner has reached the same or lower pageblock than * the migration scanner, and compaction should thus terminate. */ static inline bool compact_scanners_met(struct compact_control *cc) { return (cc->free_pfn >> pageblock_order) <= (cc->migrate_pfn >> pageblock_order); } /* |
ff9543fd3
|
970 971 |
* Based on information in the current compact_control, find blocks * suitable for isolating free pages from and then isolate them. |
2fe86e000
|
972 |
*/ |
edc2ca612
|
973 |
static void isolate_freepages(struct compact_control *cc) |
2fe86e000
|
974 |
{ |
edc2ca612
|
975 |
struct zone *zone = cc->zone; |
ff9543fd3
|
976 |
struct page *page; |
c96b9e508
|
977 |
unsigned long block_start_pfn; /* start of current pageblock */ |
e14c720ef
|
978 |
unsigned long isolate_start_pfn; /* exact pfn we start at */ |
c96b9e508
|
979 980 |
unsigned long block_end_pfn; /* end of current pageblock */ unsigned long low_pfn; /* lowest pfn scanner is able to scan */ |
ff9543fd3
|
981 |
struct list_head *freelist = &cc->freepages; |
2fe86e000
|
982 |
|
ff9543fd3
|
983 984 |
/* * Initialise the free scanner. The starting point is where we last |
49e068f0b
|
985 |
* successfully isolated from, zone-cached value, or the end of the |
e14c720ef
|
986 987 |
* zone when isolating for the first time. For looping we also need * this pfn aligned down to the pageblock boundary, because we do |
c96b9e508
|
988 989 990 |
* block_start_pfn -= pageblock_nr_pages in the for loop. * For ending point, take care when isolating in last pageblock of a * a zone which ends in the middle of a pageblock. |
49e068f0b
|
991 992 |
* The low boundary is the end of the pageblock the migration scanner * is using. |
ff9543fd3
|
993 |
*/ |
e14c720ef
|
994 |
isolate_start_pfn = cc->free_pfn; |
06b6640a3
|
995 |
block_start_pfn = pageblock_start_pfn(cc->free_pfn); |
c96b9e508
|
996 997 |
block_end_pfn = min(block_start_pfn + pageblock_nr_pages, zone_end_pfn(zone)); |
06b6640a3
|
998 |
low_pfn = pageblock_end_pfn(cc->migrate_pfn); |
2fe86e000
|
999 |
|
ff9543fd3
|
1000 |
/* |
ff9543fd3
|
1001 1002 1003 1004 |
* Isolate free pages until enough are available to migrate the * pages on cc->migratepages. We stop searching if the migrate * and free page scanners meet or enough free pages are isolated. */ |
f5f61a320
|
1005 |
for (; block_start_pfn >= low_pfn; |
c96b9e508
|
1006 |
block_end_pfn = block_start_pfn, |
e14c720ef
|
1007 1008 |
block_start_pfn -= pageblock_nr_pages, isolate_start_pfn = block_start_pfn) { |
f6ea3adb7
|
1009 1010 1011 |
/* * This can iterate a massively long zone without finding any * suitable migration targets, so periodically check if we need |
be9765722
|
1012 |
* to schedule, or even abort async compaction. |
f6ea3adb7
|
1013 |
*/ |
be9765722
|
1014 1015 1016 |
if (!(block_start_pfn % (SWAP_CLUSTER_MAX * pageblock_nr_pages)) && compact_should_abort(cc)) break; |
f6ea3adb7
|
1017 |
|
7d49d8868
|
1018 1019 1020 |
page = pageblock_pfn_to_page(block_start_pfn, block_end_pfn, zone); if (!page) |
ff9543fd3
|
1021 1022 1023 |
continue; /* Check the block is suitable for migration */ |
68e3e9262
|
1024 |
if (!suitable_migration_target(page)) |
ff9543fd3
|
1025 |
continue; |
68e3e9262
|
1026 |
|
bb13ffeb9
|
1027 1028 1029 |
/* If isolation recently failed, do not retry */ if (!isolation_suitable(cc, page)) continue; |
e14c720ef
|
1030 |
/* Found a block suitable for isolating free pages from. */ |
a46cbf3bc
|
1031 1032 |
isolate_freepages_block(cc, &isolate_start_pfn, block_end_pfn, freelist, false); |
ff9543fd3
|
1033 1034 |
/* |
a46cbf3bc
|
1035 1036 |
* If we isolated enough freepages, or aborted due to lock * contention, terminate. |
e14c720ef
|
1037 |
*/ |
f5f61a320
|
1038 1039 |
if ((cc->nr_freepages >= cc->nr_migratepages) || cc->contended) { |
a46cbf3bc
|
1040 1041 1042 1043 1044 |
if (isolate_start_pfn >= block_end_pfn) { /* * Restart at previous pageblock if more * freepages can be isolated next time. */ |
f5f61a320
|
1045 1046 |
isolate_start_pfn = block_start_pfn - pageblock_nr_pages; |
a46cbf3bc
|
1047 |
} |
be9765722
|
1048 |
break; |
a46cbf3bc
|
1049 |
} else if (isolate_start_pfn < block_end_pfn) { |
f5f61a320
|
1050 |
/* |
a46cbf3bc
|
1051 1052 |
* If isolation failed early, do not continue * needlessly. |
f5f61a320
|
1053 |
*/ |
a46cbf3bc
|
1054 |
break; |
f5f61a320
|
1055 |
} |
ff9543fd3
|
1056 |
} |
66c64223a
|
1057 |
/* __isolate_free_page() does not map the pages */ |
ff9543fd3
|
1058 |
map_pages(freelist); |
7ed695e06
|
1059 |
/* |
f5f61a320
|
1060 1061 1062 1063 |
* Record where the free scanner will restart next time. Either we * broke from the loop and set isolate_start_pfn based on the last * call to isolate_freepages_block(), or we met the migration scanner * and the loop terminated due to isolate_start_pfn < low_pfn |
7ed695e06
|
1064 |
*/ |
f5f61a320
|
1065 |
cc->free_pfn = isolate_start_pfn; |
748446bb6
|
1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 |
} /* * This is a migrate-callback that "allocates" freepages by taking pages * from the isolated freelists in the block we are migrating to. */ static struct page *compaction_alloc(struct page *migratepage, unsigned long data, int **result) { struct compact_control *cc = (struct compact_control *)data; struct page *freepage; |
be9765722
|
1078 1079 1080 1081 |
/* * Isolate free pages if necessary, and if we are not aborting due to * contention. */ |
748446bb6
|
1082 |
if (list_empty(&cc->freepages)) { |
be9765722
|
1083 |
if (!cc->contended) |
edc2ca612
|
1084 |
isolate_freepages(cc); |
748446bb6
|
1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 |
if (list_empty(&cc->freepages)) return NULL; } freepage = list_entry(cc->freepages.next, struct page, lru); list_del(&freepage->lru); cc->nr_freepages--; return freepage; } /* |
d53aea3d4
|
1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 |
* This is a migrate-callback that "frees" freepages back to the isolated * freelist. All pages on the freelist are from the same zone, so there is no * special handling needed for NUMA. */ static void compaction_free(struct page *page, unsigned long data) { struct compact_control *cc = (struct compact_control *)data; list_add(&page->lru, &cc->freepages); cc->nr_freepages++; } |
ff9543fd3
|
1109 1110 1111 1112 1113 1114 1115 1116 |
/* possible outcome of isolate_migratepages */ typedef enum { ISOLATE_ABORT, /* Abort compaction now */ ISOLATE_NONE, /* No pages isolated, continue scanning */ ISOLATE_SUCCESS, /* Pages isolated, migrate */ } isolate_migrate_t; /* |
5bbe3547a
|
1117 1118 1119 1120 1121 1122 |
* Allow userspace to control policy on scanning the unevictable LRU for * compactable pages. */ int sysctl_compact_unevictable_allowed __read_mostly = 1; /* |
edc2ca612
|
1123 1124 1125 |
* Isolate all pages that can be migrated from the first suitable block, * starting at the block pointed to by the migrate scanner pfn within * compact_control. |
ff9543fd3
|
1126 1127 1128 1129 |
*/ static isolate_migrate_t isolate_migratepages(struct zone *zone, struct compact_control *cc) { |
e1409c325
|
1130 1131 1132 |
unsigned long block_start_pfn; unsigned long block_end_pfn; unsigned long low_pfn; |
edc2ca612
|
1133 1134 |
struct page *page; const isolate_mode_t isolate_mode = |
5bbe3547a
|
1135 |
(sysctl_compact_unevictable_allowed ? ISOLATE_UNEVICTABLE : 0) | |
1d2047fef
|
1136 |
(cc->mode != MIGRATE_SYNC ? ISOLATE_ASYNC_MIGRATE : 0); |
ff9543fd3
|
1137 |
|
edc2ca612
|
1138 1139 1140 1141 1142 |
/* * Start at where we last stopped, or beginning of the zone as * initialized by compact_zone() */ low_pfn = cc->migrate_pfn; |
06b6640a3
|
1143 |
block_start_pfn = pageblock_start_pfn(low_pfn); |
e1409c325
|
1144 1145 |
if (block_start_pfn < zone->zone_start_pfn) block_start_pfn = zone->zone_start_pfn; |
ff9543fd3
|
1146 1147 |
/* Only scan within a pageblock boundary */ |
06b6640a3
|
1148 |
block_end_pfn = pageblock_end_pfn(low_pfn); |
ff9543fd3
|
1149 |
|
edc2ca612
|
1150 1151 1152 1153 |
/* * Iterate over whole pageblocks until we find the first suitable. * Do not cross the free scanner. */ |
e1409c325
|
1154 1155 1156 1157 |
for (; block_end_pfn <= cc->free_pfn; low_pfn = block_end_pfn, block_start_pfn = block_end_pfn, block_end_pfn += pageblock_nr_pages) { |
ff9543fd3
|
1158 |
|
edc2ca612
|
1159 1160 1161 1162 1163 1164 1165 1166 |
/* * This can potentially iterate a massively long zone with * many pageblocks unsuitable, so periodically check if we * need to schedule, or even abort async compaction. */ if (!(low_pfn % (SWAP_CLUSTER_MAX * pageblock_nr_pages)) && compact_should_abort(cc)) break; |
ff9543fd3
|
1167 |
|
e1409c325
|
1168 1169 |
page = pageblock_pfn_to_page(block_start_pfn, block_end_pfn, zone); |
7d49d8868
|
1170 |
if (!page) |
edc2ca612
|
1171 |
continue; |
edc2ca612
|
1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 |
/* If isolation recently failed, do not retry */ if (!isolation_suitable(cc, page)) continue; /* * For async compaction, also only scan in MOVABLE blocks. * Async compaction is optimistic to see if the minimum amount * of work satisfies the allocation. */ if (cc->mode == MIGRATE_ASYNC && !migrate_async_suitable(get_pageblock_migratetype(page))) continue; /* Perform the isolation */ |
e1409c325
|
1186 1187 |
low_pfn = isolate_migratepages_block(cc, low_pfn, block_end_pfn, isolate_mode); |
edc2ca612
|
1188 |
|
ff59909a0
|
1189 1190 |
if (!low_pfn || cc->contended) { acct_isolated(zone, cc); |
edc2ca612
|
1191 |
return ISOLATE_ABORT; |
ff59909a0
|
1192 |
} |
edc2ca612
|
1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 |
/* * Either we isolated something and proceed with migration. Or * we failed and compact_zone should decide if we should * continue or not. */ break; } acct_isolated(zone, cc); |
f2849aa09
|
1203 1204 |
/* Record where migration scanner will be restarted. */ cc->migrate_pfn = low_pfn; |
ff9543fd3
|
1205 |
|
edc2ca612
|
1206 |
return cc->nr_migratepages ? ISOLATE_SUCCESS : ISOLATE_NONE; |
ff9543fd3
|
1207 |
} |
21c527a3c
|
1208 1209 1210 1211 1212 1213 1214 1215 |
/* * order == -1 is expected when compacting via * /proc/sys/vm/compact_memory */ static inline bool is_via_compact_memory(int order) { return order == -1; } |
ea7ab982b
|
1216 |
static enum compact_result __compact_finished(struct zone *zone, struct compact_control *cc, |
6d7ce5594
|
1217 |
const int migratetype) |
748446bb6
|
1218 |
{ |
8fb74b9fb
|
1219 |
unsigned int order; |
5a03b051e
|
1220 |
unsigned long watermark; |
56de7263f
|
1221 |
|
be9765722
|
1222 |
if (cc->contended || fatal_signal_pending(current)) |
2d1e10412
|
1223 |
return COMPACT_CONTENDED; |
748446bb6
|
1224 |
|
753341a4b
|
1225 |
/* Compaction run completes if the migrate and free scanner meet */ |
f2849aa09
|
1226 |
if (compact_scanners_met(cc)) { |
55b7c4c99
|
1227 |
/* Let the next compaction start anew. */ |
02333641e
|
1228 |
reset_cached_positions(zone); |
55b7c4c99
|
1229 |
|
62997027c
|
1230 1231 |
/* * Mark that the PG_migrate_skip information should be cleared |
accf62422
|
1232 |
* by kswapd when it goes to sleep. kcompactd does not set the |
62997027c
|
1233 1234 1235 |
* flag itself as the decision to be clear should be directly * based on an allocation request. */ |
accf62422
|
1236 |
if (cc->direct_compaction) |
62997027c
|
1237 |
zone->compact_blockskip_flush = true; |
c8f7de0bf
|
1238 1239 1240 1241 |
if (cc->whole_zone) return COMPACT_COMPLETE; else return COMPACT_PARTIAL_SKIPPED; |
bb13ffeb9
|
1242 |
} |
748446bb6
|
1243 |
|
21c527a3c
|
1244 |
if (is_via_compact_memory(cc->order)) |
56de7263f
|
1245 |
return COMPACT_CONTINUE; |
3957c7768
|
1246 1247 |
/* Compaction run is not finished if the watermark is not met */ watermark = low_wmark_pages(zone); |
3957c7768
|
1248 |
|
ebff39801
|
1249 1250 |
if (!zone_watermark_ok(zone, cc->order, watermark, cc->classzone_idx, cc->alloc_flags)) |
3957c7768
|
1251 |
return COMPACT_CONTINUE; |
56de7263f
|
1252 |
/* Direct compactor: Is a suitable page free? */ |
8fb74b9fb
|
1253 1254 |
for (order = cc->order; order < MAX_ORDER; order++) { struct free_area *area = &zone->free_area[order]; |
2149cdaef
|
1255 |
bool can_steal; |
8fb74b9fb
|
1256 1257 |
/* Job done if page is free of the right migratetype */ |
6d7ce5594
|
1258 |
if (!list_empty(&area->free_list[migratetype])) |
8fb74b9fb
|
1259 |
return COMPACT_PARTIAL; |
2149cdaef
|
1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 |
#ifdef CONFIG_CMA /* MIGRATE_MOVABLE can fallback on MIGRATE_CMA */ if (migratetype == MIGRATE_MOVABLE && !list_empty(&area->free_list[MIGRATE_CMA])) return COMPACT_PARTIAL; #endif /* * Job done if allocation would steal freepages from * other migratetype buddy lists. */ if (find_suitable_fallback(area, order, migratetype, true, &can_steal) != -1) |
56de7263f
|
1272 1273 |
return COMPACT_PARTIAL; } |
837d026d5
|
1274 1275 |
return COMPACT_NO_SUITABLE_PAGE; } |
ea7ab982b
|
1276 1277 1278 |
static enum compact_result compact_finished(struct zone *zone, struct compact_control *cc, const int migratetype) |
837d026d5
|
1279 1280 1281 1282 1283 1284 1285 1286 1287 |
{ int ret; ret = __compact_finished(zone, cc, migratetype); trace_mm_compaction_finished(zone, cc->order, ret); if (ret == COMPACT_NO_SUITABLE_PAGE) ret = COMPACT_CONTINUE; return ret; |
748446bb6
|
1288 |
} |
3e7d34497
|
1289 1290 1291 1292 1293 1294 1295 |
/* * compaction_suitable: Is this suitable to run compaction on this zone now? * Returns * COMPACT_SKIPPED - If there are too few free pages for compaction * COMPACT_PARTIAL - If the allocation would succeed without compaction * COMPACT_CONTINUE - If compaction should run now */ |
ea7ab982b
|
1296 |
static enum compact_result __compaction_suitable(struct zone *zone, int order, |
c603844bd
|
1297 |
unsigned int alloc_flags, |
86a294a81
|
1298 1299 |
int classzone_idx, unsigned long wmark_target) |
3e7d34497
|
1300 1301 1302 |
{ int fragindex; unsigned long watermark; |
21c527a3c
|
1303 |
if (is_via_compact_memory(order)) |
3957c7768
|
1304 |
return COMPACT_CONTINUE; |
ebff39801
|
1305 1306 1307 1308 1309 1310 1311 1312 |
watermark = low_wmark_pages(zone); /* * If watermarks for high-order allocation are already met, there * should be no need for compaction at all. */ if (zone_watermark_ok(zone, order, watermark, classzone_idx, alloc_flags)) return COMPACT_PARTIAL; |
3957c7768
|
1313 |
/* |
3e7d34497
|
1314 1315 1316 1317 |
* Watermarks for order-0 must be met for compaction. Note the 2UL. * This is because during migration, copies of pages need to be * allocated and for a short time, the footprint is higher */ |
ebff39801
|
1318 |
watermark += (2UL << order); |
86a294a81
|
1319 1320 |
if (!__zone_watermark_ok(zone, 0, watermark, classzone_idx, alloc_flags, wmark_target)) |
3e7d34497
|
1321 1322 1323 1324 1325 1326 |
return COMPACT_SKIPPED; /* * fragmentation index determines if allocation failures are due to * low memory or external fragmentation * |
ebff39801
|
1327 1328 |
* index of -1000 would imply allocations might succeed depending on * watermarks, but we already failed the high-order watermark check |
3e7d34497
|
1329 1330 1331 1332 1333 1334 1335 |
* index towards 0 implies failure is due to lack of memory * index towards 1000 implies failure is due to fragmentation * * Only compact if a failure would be due to fragmentation. */ fragindex = fragmentation_index(zone, order); if (fragindex >= 0 && fragindex <= sysctl_extfrag_threshold) |
837d026d5
|
1336 |
return COMPACT_NOT_SUITABLE_ZONE; |
3e7d34497
|
1337 |
|
3e7d34497
|
1338 1339 |
return COMPACT_CONTINUE; } |
ea7ab982b
|
1340 |
enum compact_result compaction_suitable(struct zone *zone, int order, |
c603844bd
|
1341 1342 |
unsigned int alloc_flags, int classzone_idx) |
837d026d5
|
1343 |
{ |
ea7ab982b
|
1344 |
enum compact_result ret; |
837d026d5
|
1345 |
|
86a294a81
|
1346 1347 |
ret = __compaction_suitable(zone, order, alloc_flags, classzone_idx, zone_page_state(zone, NR_FREE_PAGES)); |
837d026d5
|
1348 1349 1350 1351 1352 1353 |
trace_mm_compaction_suitable(zone, order, ret); if (ret == COMPACT_NOT_SUITABLE_ZONE) ret = COMPACT_SKIPPED; return ret; } |
86a294a81
|
1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 |
bool compaction_zonelist_suitable(struct alloc_context *ac, int order, int alloc_flags) { struct zone *zone; struct zoneref *z; /* * Make sure at least one zone would pass __compaction_suitable if we continue * retrying the reclaim. */ for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, ac->high_zoneidx, ac->nodemask) { unsigned long available; enum compact_result compact_result; /* * Do not consider all the reclaimable memory because we do not * want to trash just for a single high order allocation which * is even not guaranteed to appear even if __compaction_suitable * is happy about the watermark check. */ |
5a1c84b40
|
1375 |
available = zone_reclaimable_pages(zone) / order; |
86a294a81
|
1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 |
available += zone_page_state_snapshot(zone, NR_FREE_PAGES); compact_result = __compaction_suitable(zone, order, alloc_flags, ac_classzone_idx(ac), available); if (compact_result != COMPACT_SKIPPED && compact_result != COMPACT_NOT_SUITABLE_ZONE) return true; } return false; } |
ea7ab982b
|
1386 |
static enum compact_result compact_zone(struct zone *zone, struct compact_control *cc) |
748446bb6
|
1387 |
{ |
ea7ab982b
|
1388 |
enum compact_result ret; |
c89511ab2
|
1389 |
unsigned long start_pfn = zone->zone_start_pfn; |
108bcc96e
|
1390 |
unsigned long end_pfn = zone_end_pfn(zone); |
6d7ce5594
|
1391 |
const int migratetype = gfpflags_to_migratetype(cc->gfp_mask); |
e0b9daeb4
|
1392 |
const bool sync = cc->mode != MIGRATE_ASYNC; |
748446bb6
|
1393 |
|
ebff39801
|
1394 1395 |
ret = compaction_suitable(zone, cc->order, cc->alloc_flags, cc->classzone_idx); |
c46649dea
|
1396 1397 |
/* Compaction is likely to fail */ if (ret == COMPACT_PARTIAL || ret == COMPACT_SKIPPED) |
3e7d34497
|
1398 |
return ret; |
c46649dea
|
1399 1400 1401 |
/* huh, compaction_suitable is returning something unexpected */ VM_BUG_ON(ret != COMPACT_CONTINUE); |
3e7d34497
|
1402 |
|
c89511ab2
|
1403 |
/* |
d3132e4b8
|
1404 |
* Clear pageblock skip if there were failures recently and compaction |
accf62422
|
1405 |
* is about to be retried after being deferred. |
d3132e4b8
|
1406 |
*/ |
accf62422
|
1407 |
if (compaction_restarting(zone, cc->order)) |
d3132e4b8
|
1408 1409 1410 |
__reset_isolation_suitable(zone); /* |
c89511ab2
|
1411 1412 1413 1414 |
* Setup to move all movable pages to the end of the zone. Used cached * information on where the scanners should start but check that it * is initialised by ensuring the values are within zone boundaries. */ |
e0b9daeb4
|
1415 |
cc->migrate_pfn = zone->compact_cached_migrate_pfn[sync]; |
c89511ab2
|
1416 |
cc->free_pfn = zone->compact_cached_free_pfn; |
623446e4d
|
1417 |
if (cc->free_pfn < start_pfn || cc->free_pfn >= end_pfn) { |
06b6640a3
|
1418 |
cc->free_pfn = pageblock_start_pfn(end_pfn - 1); |
c89511ab2
|
1419 1420 |
zone->compact_cached_free_pfn = cc->free_pfn; } |
623446e4d
|
1421 |
if (cc->migrate_pfn < start_pfn || cc->migrate_pfn >= end_pfn) { |
c89511ab2
|
1422 |
cc->migrate_pfn = start_pfn; |
35979ef33
|
1423 1424 |
zone->compact_cached_migrate_pfn[0] = cc->migrate_pfn; zone->compact_cached_migrate_pfn[1] = cc->migrate_pfn; |
c89511ab2
|
1425 |
} |
c8f7de0bf
|
1426 1427 1428 |
if (cc->migrate_pfn == start_pfn) cc->whole_zone = true; |
1a16718cf
|
1429 |
cc->last_migrated_pfn = 0; |
748446bb6
|
1430 |
|
16c4a097a
|
1431 1432 |
trace_mm_compaction_begin(start_pfn, cc->migrate_pfn, cc->free_pfn, end_pfn, sync); |
0eb927c0a
|
1433 |
|
748446bb6
|
1434 |
migrate_prep_local(); |
6d7ce5594
|
1435 1436 |
while ((ret = compact_finished(zone, cc, migratetype)) == COMPACT_CONTINUE) { |
9d502c1c8
|
1437 |
int err; |
748446bb6
|
1438 |
|
f9e35b3b4
|
1439 1440 |
switch (isolate_migratepages(zone, cc)) { case ISOLATE_ABORT: |
2d1e10412
|
1441 |
ret = COMPACT_CONTENDED; |
5733c7d11
|
1442 |
putback_movable_pages(&cc->migratepages); |
e64c5237c
|
1443 |
cc->nr_migratepages = 0; |
f9e35b3b4
|
1444 1445 |
goto out; case ISOLATE_NONE: |
fdaf7f5c4
|
1446 1447 1448 1449 1450 1451 |
/* * We haven't isolated and migrated anything, but * there might still be unflushed migrations from * previous cc->order aligned block. */ goto check_drain; |
f9e35b3b4
|
1452 1453 1454 |
case ISOLATE_SUCCESS: ; } |
748446bb6
|
1455 |
|
d53aea3d4
|
1456 |
err = migrate_pages(&cc->migratepages, compaction_alloc, |
e0b9daeb4
|
1457 |
compaction_free, (unsigned long)cc, cc->mode, |
7b2a2d4a1
|
1458 |
MR_COMPACTION); |
748446bb6
|
1459 |
|
f8c9301fa
|
1460 1461 |
trace_mm_compaction_migratepages(cc->nr_migratepages, err, &cc->migratepages); |
748446bb6
|
1462 |
|
f8c9301fa
|
1463 1464 |
/* All pages were either migrated or will be released */ cc->nr_migratepages = 0; |
9d502c1c8
|
1465 |
if (err) { |
5733c7d11
|
1466 |
putback_movable_pages(&cc->migratepages); |
7ed695e06
|
1467 1468 1469 1470 |
/* * migrate_pages() may return -ENOMEM when scanners meet * and we want compact_finished() to detect it */ |
f2849aa09
|
1471 |
if (err == -ENOMEM && !compact_scanners_met(cc)) { |
2d1e10412
|
1472 |
ret = COMPACT_CONTENDED; |
4bf2bba37
|
1473 1474 |
goto out; } |
fdd048e12
|
1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 |
/* * We failed to migrate at least one page in the current * order-aligned block, so skip the rest of it. */ if (cc->direct_compaction && (cc->mode == MIGRATE_ASYNC)) { cc->migrate_pfn = block_end_pfn( cc->migrate_pfn - 1, cc->order); /* Draining pcplists is useless in this case */ cc->last_migrated_pfn = 0; } |
748446bb6
|
1487 |
} |
fdaf7f5c4
|
1488 |
|
fdaf7f5c4
|
1489 1490 1491 1492 1493 1494 1495 1496 |
check_drain: /* * Has the migration scanner moved away from the previous * cc->order aligned block where we migrated from? If yes, * flush the pages that were freed, so that they can merge and * compact_finished() can detect immediately if allocation * would succeed. */ |
1a16718cf
|
1497 |
if (cc->order > 0 && cc->last_migrated_pfn) { |
fdaf7f5c4
|
1498 1499 |
int cpu; unsigned long current_block_start = |
06b6640a3
|
1500 |
block_start_pfn(cc->migrate_pfn, cc->order); |
fdaf7f5c4
|
1501 |
|
1a16718cf
|
1502 |
if (cc->last_migrated_pfn < current_block_start) { |
fdaf7f5c4
|
1503 1504 1505 1506 1507 |
cpu = get_cpu(); lru_add_drain_cpu(cpu); drain_local_pages(zone); put_cpu(); /* No more flushing until we migrate again */ |
1a16718cf
|
1508 |
cc->last_migrated_pfn = 0; |
fdaf7f5c4
|
1509 1510 |
} } |
748446bb6
|
1511 |
} |
f9e35b3b4
|
1512 |
out: |
6bace090a
|
1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 |
/* * Release free pages and update where the free scanner should restart, * so we don't leave any returned pages behind in the next attempt. */ if (cc->nr_freepages > 0) { unsigned long free_pfn = release_freepages(&cc->freepages); cc->nr_freepages = 0; VM_BUG_ON(free_pfn == 0); /* The cached pfn is always the first in a pageblock */ |
06b6640a3
|
1523 |
free_pfn = pageblock_start_pfn(free_pfn); |
6bace090a
|
1524 1525 1526 1527 1528 1529 1530 |
/* * Only go back, not forward. The cached pfn might have been * already reset to zone end in compact_finished() */ if (free_pfn > zone->compact_cached_free_pfn) zone->compact_cached_free_pfn = free_pfn; } |
748446bb6
|
1531 |
|
16c4a097a
|
1532 1533 |
trace_mm_compaction_end(start_pfn, cc->migrate_pfn, cc->free_pfn, end_pfn, sync, ret); |
0eb927c0a
|
1534 |
|
748446bb6
|
1535 1536 |
return ret; } |
76ab0f530
|
1537 |
|
ea7ab982b
|
1538 |
static enum compact_result compact_zone_order(struct zone *zone, int order, |
c3486f537
|
1539 |
gfp_t gfp_mask, enum compact_priority prio, |
c603844bd
|
1540 |
unsigned int alloc_flags, int classzone_idx) |
56de7263f
|
1541 |
{ |
ea7ab982b
|
1542 |
enum compact_result ret; |
56de7263f
|
1543 1544 1545 1546 |
struct compact_control cc = { .nr_freepages = 0, .nr_migratepages = 0, .order = order, |
6d7ce5594
|
1547 |
.gfp_mask = gfp_mask, |
56de7263f
|
1548 |
.zone = zone, |
a5508cd83
|
1549 1550 |
.mode = (prio == COMPACT_PRIO_ASYNC) ? MIGRATE_ASYNC : MIGRATE_SYNC_LIGHT, |
ebff39801
|
1551 1552 |
.alloc_flags = alloc_flags, .classzone_idx = classzone_idx, |
accf62422
|
1553 |
.direct_compaction = true, |
56de7263f
|
1554 1555 1556 |
}; INIT_LIST_HEAD(&cc.freepages); INIT_LIST_HEAD(&cc.migratepages); |
e64c5237c
|
1557 1558 1559 1560 |
ret = compact_zone(zone, &cc); VM_BUG_ON(!list_empty(&cc.freepages)); VM_BUG_ON(!list_empty(&cc.migratepages)); |
e64c5237c
|
1561 |
return ret; |
56de7263f
|
1562 |
} |
5e7719058
|
1563 |
int sysctl_extfrag_threshold = 500; |
56de7263f
|
1564 1565 |
/** * try_to_compact_pages - Direct compact to satisfy a high-order allocation |
56de7263f
|
1566 |
* @gfp_mask: The GFP mask of the current allocation |
1a6d53a10
|
1567 1568 1569 |
* @order: The order of the current allocation * @alloc_flags: The allocation flags of the current allocation * @ac: The context of current allocation |
e0b9daeb4
|
1570 |
* @mode: The migration mode for async, sync light, or sync migration |
56de7263f
|
1571 1572 1573 |
* * This is the main entry point for direct page compaction. */ |
ea7ab982b
|
1574 |
enum compact_result try_to_compact_pages(gfp_t gfp_mask, unsigned int order, |
c603844bd
|
1575 |
unsigned int alloc_flags, const struct alloc_context *ac, |
c3486f537
|
1576 |
enum compact_priority prio) |
56de7263f
|
1577 |
{ |
56de7263f
|
1578 1579 |
int may_enter_fs = gfp_mask & __GFP_FS; int may_perform_io = gfp_mask & __GFP_IO; |
56de7263f
|
1580 1581 |
struct zoneref *z; struct zone *zone; |
1d4746d39
|
1582 |
enum compact_result rc = COMPACT_SKIPPED; |
56de7263f
|
1583 |
|
4ffb6335d
|
1584 |
/* Check if the GFP flags allow compaction */ |
b2b331f96
|
1585 |
if (!may_enter_fs || !may_perform_io) |
53853e2d2
|
1586 |
return COMPACT_SKIPPED; |
56de7263f
|
1587 |
|
a5508cd83
|
1588 |
trace_mm_compaction_try_to_compact_pages(order, gfp_mask, prio); |
837d026d5
|
1589 |
|
56de7263f
|
1590 |
/* Compact each zone in the list */ |
1a6d53a10
|
1591 1592 |
for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, ac->high_zoneidx, ac->nodemask) { |
ea7ab982b
|
1593 |
enum compact_result status; |
56de7263f
|
1594 |
|
1d4746d39
|
1595 1596 |
if (compaction_deferred(zone, order)) { rc = max_t(enum compact_result, COMPACT_DEFERRED, rc); |
53853e2d2
|
1597 |
continue; |
1d4746d39
|
1598 |
} |
53853e2d2
|
1599 |
|
a5508cd83
|
1600 |
status = compact_zone_order(zone, order, gfp_mask, prio, |
c3486f537
|
1601 |
alloc_flags, ac_classzone_idx(ac)); |
56de7263f
|
1602 |
rc = max(status, rc); |
3e7d34497
|
1603 |
/* If a normal allocation would succeed, stop compacting */ |
ebff39801
|
1604 |
if (zone_watermark_ok(zone, order, low_wmark_pages(zone), |
93ea9964d
|
1605 |
ac_classzone_idx(ac), alloc_flags)) { |
53853e2d2
|
1606 1607 1608 1609 1610 1611 1612 |
/* * We think the allocation will succeed in this zone, * but it is not certain, hence the false. The caller * will repeat this with true if allocation indeed * succeeds in this zone. */ compaction_defer_reset(zone, order, false); |
1f9efdef4
|
1613 |
|
c3486f537
|
1614 |
break; |
1f9efdef4
|
1615 |
} |
a5508cd83
|
1616 |
if (prio != COMPACT_PRIO_ASYNC && (status == COMPACT_COMPLETE || |
c3486f537
|
1617 |
status == COMPACT_PARTIAL_SKIPPED)) |
53853e2d2
|
1618 1619 1620 1621 1622 1623 |
/* * We think that allocation won't succeed in this zone * so we defer compaction there. If it ends up * succeeding after all, it will be reset. */ defer_compaction(zone, order); |
1f9efdef4
|
1624 1625 1626 1627 |
/* * We might have stopped compacting due to need_resched() in * async compaction, or due to a fatal signal detected. In that |
c3486f537
|
1628 |
* case do not try further zones |
1f9efdef4
|
1629 |
*/ |
c3486f537
|
1630 1631 1632 |
if ((prio == COMPACT_PRIO_ASYNC && need_resched()) || fatal_signal_pending(current)) break; |
56de7263f
|
1633 1634 1635 1636 |
} return rc; } |
76ab0f530
|
1637 |
/* Compact all zones within a node */ |
7103f16db
|
1638 |
static void __compact_pgdat(pg_data_t *pgdat, struct compact_control *cc) |
76ab0f530
|
1639 1640 |
{ int zoneid; |
76ab0f530
|
1641 |
struct zone *zone; |
76ab0f530
|
1642 |
for (zoneid = 0; zoneid < MAX_NR_ZONES; zoneid++) { |
76ab0f530
|
1643 1644 1645 1646 |
zone = &pgdat->node_zones[zoneid]; if (!populated_zone(zone)) continue; |
7be62de99
|
1647 1648 1649 1650 1651 |
cc->nr_freepages = 0; cc->nr_migratepages = 0; cc->zone = zone; INIT_LIST_HEAD(&cc->freepages); INIT_LIST_HEAD(&cc->migratepages); |
76ab0f530
|
1652 |
|
195b0c608
|
1653 1654 1655 1656 1657 |
/* * When called via /proc/sys/vm/compact_memory * this makes sure we compact the whole zone regardless of * cached scanner positions. */ |
21c527a3c
|
1658 |
if (is_via_compact_memory(cc->order)) |
195b0c608
|
1659 |
__reset_isolation_suitable(zone); |
21c527a3c
|
1660 1661 |
if (is_via_compact_memory(cc->order) || !compaction_deferred(zone, cc->order)) |
7be62de99
|
1662 |
compact_zone(zone, cc); |
76ab0f530
|
1663 |
|
7be62de99
|
1664 1665 |
VM_BUG_ON(!list_empty(&cc->freepages)); VM_BUG_ON(!list_empty(&cc->migratepages)); |
754693457
|
1666 1667 1668 1669 1670 1671 1672 |
if (is_via_compact_memory(cc->order)) continue; if (zone_watermark_ok(zone, cc->order, low_wmark_pages(zone), 0, 0)) compaction_defer_reset(zone, cc->order, false); |
76ab0f530
|
1673 |
} |
76ab0f530
|
1674 |
} |
7103f16db
|
1675 |
void compact_pgdat(pg_data_t *pgdat, int order) |
7be62de99
|
1676 1677 1678 |
{ struct compact_control cc = { .order = order, |
e0b9daeb4
|
1679 |
.mode = MIGRATE_ASYNC, |
7be62de99
|
1680 |
}; |
3a7200af3
|
1681 1682 |
if (!order) return; |
7103f16db
|
1683 |
__compact_pgdat(pgdat, &cc); |
7be62de99
|
1684 |
} |
7103f16db
|
1685 |
static void compact_node(int nid) |
7be62de99
|
1686 |
{ |
7be62de99
|
1687 1688 |
struct compact_control cc = { .order = -1, |
e0b9daeb4
|
1689 |
.mode = MIGRATE_SYNC, |
91ca91864
|
1690 |
.ignore_skip_hint = true, |
7be62de99
|
1691 |
}; |
7103f16db
|
1692 |
__compact_pgdat(NODE_DATA(nid), &cc); |
7be62de99
|
1693 |
} |
76ab0f530
|
1694 |
/* Compact all nodes in the system */ |
7964c06d6
|
1695 |
static void compact_nodes(void) |
76ab0f530
|
1696 1697 |
{ int nid; |
8575ec29f
|
1698 1699 |
/* Flush pending updates to the LRU lists */ lru_add_drain_all(); |
76ab0f530
|
1700 1701 |
for_each_online_node(nid) compact_node(nid); |
76ab0f530
|
1702 1703 1704 1705 |
} /* The written value is actually unused, all memory is compacted */ int sysctl_compact_memory; |
fec4eb2c8
|
1706 1707 1708 1709 |
/* * This is the entry point for compacting all nodes via * /proc/sys/vm/compact_memory */ |
76ab0f530
|
1710 1711 1712 1713 |
int sysctl_compaction_handler(struct ctl_table *table, int write, void __user *buffer, size_t *length, loff_t *ppos) { if (write) |
7964c06d6
|
1714 |
compact_nodes(); |
76ab0f530
|
1715 1716 1717 |
return 0; } |
ed4a6d7f0
|
1718 |
|
5e7719058
|
1719 1720 1721 1722 1723 1724 1725 |
int sysctl_extfrag_handler(struct ctl_table *table, int write, void __user *buffer, size_t *length, loff_t *ppos) { proc_dointvec_minmax(table, write, buffer, length, ppos); return 0; } |
ed4a6d7f0
|
1726 |
#if defined(CONFIG_SYSFS) && defined(CONFIG_NUMA) |
74e77fb9a
|
1727 |
static ssize_t sysfs_compact_node(struct device *dev, |
10fbcf4c6
|
1728 |
struct device_attribute *attr, |
ed4a6d7f0
|
1729 1730 |
const char *buf, size_t count) { |
8575ec29f
|
1731 1732 1733 1734 1735 1736 1737 1738 |
int nid = dev->id; if (nid >= 0 && nid < nr_node_ids && node_online(nid)) { /* Flush pending updates to the LRU lists */ lru_add_drain_all(); compact_node(nid); } |
ed4a6d7f0
|
1739 1740 1741 |
return count; } |
10fbcf4c6
|
1742 |
static DEVICE_ATTR(compact, S_IWUSR, NULL, sysfs_compact_node); |
ed4a6d7f0
|
1743 1744 1745 |
int compaction_register_node(struct node *node) { |
10fbcf4c6
|
1746 |
return device_create_file(&node->dev, &dev_attr_compact); |
ed4a6d7f0
|
1747 1748 1749 1750 |
} void compaction_unregister_node(struct node *node) { |
10fbcf4c6
|
1751 |
return device_remove_file(&node->dev, &dev_attr_compact); |
ed4a6d7f0
|
1752 1753 |
} #endif /* CONFIG_SYSFS && CONFIG_NUMA */ |
ff9543fd3
|
1754 |
|
698b1b306
|
1755 1756 |
static inline bool kcompactd_work_requested(pg_data_t *pgdat) { |
172400c69
|
1757 |
return pgdat->kcompactd_max_order > 0 || kthread_should_stop(); |
698b1b306
|
1758 1759 1760 1761 1762 1763 1764 |
} static bool kcompactd_node_suitable(pg_data_t *pgdat) { int zoneid; struct zone *zone; enum zone_type classzone_idx = pgdat->kcompactd_classzone_idx; |
6cd9dc3e7
|
1765 |
for (zoneid = 0; zoneid <= classzone_idx; zoneid++) { |
698b1b306
|
1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 |
zone = &pgdat->node_zones[zoneid]; if (!populated_zone(zone)) continue; if (compaction_suitable(zone, pgdat->kcompactd_max_order, 0, classzone_idx) == COMPACT_CONTINUE) return true; } return false; } static void kcompactd_do_work(pg_data_t *pgdat) { /* * With no special task, compact all zones so that a page of requested * order is allocatable. */ int zoneid; struct zone *zone; struct compact_control cc = { .order = pgdat->kcompactd_max_order, .classzone_idx = pgdat->kcompactd_classzone_idx, .mode = MIGRATE_SYNC_LIGHT, .ignore_skip_hint = true, }; bool success = false; trace_mm_compaction_kcompactd_wake(pgdat->node_id, cc.order, cc.classzone_idx); count_vm_event(KCOMPACTD_WAKE); |
6cd9dc3e7
|
1799 |
for (zoneid = 0; zoneid <= cc.classzone_idx; zoneid++) { |
698b1b306
|
1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 |
int status; zone = &pgdat->node_zones[zoneid]; if (!populated_zone(zone)) continue; if (compaction_deferred(zone, cc.order)) continue; if (compaction_suitable(zone, cc.order, 0, zoneid) != COMPACT_CONTINUE) continue; cc.nr_freepages = 0; cc.nr_migratepages = 0; cc.zone = zone; INIT_LIST_HEAD(&cc.freepages); INIT_LIST_HEAD(&cc.migratepages); |
172400c69
|
1818 1819 |
if (kthread_should_stop()) return; |
698b1b306
|
1820 1821 1822 1823 1824 1825 |
status = compact_zone(zone, &cc); if (zone_watermark_ok(zone, cc.order, low_wmark_pages(zone), cc.classzone_idx, 0)) { success = true; compaction_defer_reset(zone, cc.order, false); |
c8f7de0bf
|
1826 |
} else if (status == COMPACT_PARTIAL_SKIPPED || status == COMPACT_COMPLETE) { |
698b1b306
|
1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 |
/* * We use sync migration mode here, so we defer like * sync direct compaction does. */ defer_compaction(zone, cc.order); } VM_BUG_ON(!list_empty(&cc.freepages)); VM_BUG_ON(!list_empty(&cc.migratepages)); } /* * Regardless of success, we are done until woken up next. But remember * the requested order/classzone_idx in case it was higher/tighter than * our current ones */ if (pgdat->kcompactd_max_order <= cc.order) pgdat->kcompactd_max_order = 0; if (pgdat->kcompactd_classzone_idx >= cc.classzone_idx) pgdat->kcompactd_classzone_idx = pgdat->nr_zones - 1; } void wakeup_kcompactd(pg_data_t *pgdat, int order, int classzone_idx) { if (!order) return; if (pgdat->kcompactd_max_order < order) pgdat->kcompactd_max_order = order; if (pgdat->kcompactd_classzone_idx > classzone_idx) pgdat->kcompactd_classzone_idx = classzone_idx; if (!waitqueue_active(&pgdat->kcompactd_wait)) return; if (!kcompactd_node_suitable(pgdat)) return; trace_mm_compaction_wakeup_kcompactd(pgdat->node_id, order, classzone_idx); wake_up_interruptible(&pgdat->kcompactd_wait); } /* * The background compaction daemon, started as a kernel thread * from the init process. */ static int kcompactd(void *p) { pg_data_t *pgdat = (pg_data_t*)p; struct task_struct *tsk = current; const struct cpumask *cpumask = cpumask_of_node(pgdat->node_id); if (!cpumask_empty(cpumask)) set_cpus_allowed_ptr(tsk, cpumask); set_freezable(); pgdat->kcompactd_max_order = 0; pgdat->kcompactd_classzone_idx = pgdat->nr_zones - 1; while (!kthread_should_stop()) { trace_mm_compaction_kcompactd_sleep(pgdat->node_id); wait_event_freezable(pgdat->kcompactd_wait, kcompactd_work_requested(pgdat)); kcompactd_do_work(pgdat); } return 0; } /* * This kcompactd start function will be called by init and node-hot-add. * On node-hot-add, kcompactd will moved to proper cpus if cpus are hot-added. */ int kcompactd_run(int nid) { pg_data_t *pgdat = NODE_DATA(nid); int ret = 0; if (pgdat->kcompactd) return 0; pgdat->kcompactd = kthread_run(kcompactd, pgdat, "kcompactd%d", nid); if (IS_ERR(pgdat->kcompactd)) { pr_err("Failed to start kcompactd on node %d ", nid); ret = PTR_ERR(pgdat->kcompactd); pgdat->kcompactd = NULL; } return ret; } /* * Called by memory hotplug when all memory in a node is offlined. Caller must * hold mem_hotplug_begin/end(). */ void kcompactd_stop(int nid) { struct task_struct *kcompactd = NODE_DATA(nid)->kcompactd; if (kcompactd) { kthread_stop(kcompactd); NODE_DATA(nid)->kcompactd = NULL; } } /* * It's optimal to keep kcompactd on the same CPUs as their memory, but * not required for correctness. So if the last cpu in a node goes * away, we get changed to run anywhere: as the first one comes back, * restore their cpu bindings. */ static int cpu_callback(struct notifier_block *nfb, unsigned long action, void *hcpu) { int nid; if (action == CPU_ONLINE || action == CPU_ONLINE_FROZEN) { for_each_node_state(nid, N_MEMORY) { pg_data_t *pgdat = NODE_DATA(nid); const struct cpumask *mask; mask = cpumask_of_node(pgdat->node_id); if (cpumask_any_and(cpu_online_mask, mask) < nr_cpu_ids) /* One of our CPUs online: restore mask */ set_cpus_allowed_ptr(pgdat->kcompactd, mask); } } return NOTIFY_OK; } static int __init kcompactd_init(void) { int nid; for_each_node_state(nid, N_MEMORY) kcompactd_run(nid); hotcpu_notifier(cpu_callback, 0); return 0; } subsys_initcall(kcompactd_init) |
ff9543fd3
|
1973 |
#endif /* CONFIG_COMPACTION */ |