Blame view
mm/compaction.c
56.2 KB
748446bb6 mm: compaction: m... |
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 mm, compaction: i... |
10 |
#include <linux/cpu.h> |
748446bb6 mm: compaction: m... |
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 mm: compaction: a... |
16 |
#include <linux/sysctl.h> |
ed4a6d7f0 mm: compaction: a... |
17 |
#include <linux/sysfs.h> |
194159fbc mm: remove MIGRAT... |
18 |
#include <linux/page-isolation.h> |
b8c73fc24 mm: page_alloc: a... |
19 |
#include <linux/kasan.h> |
698b1b306 mm, compaction: i... |
20 21 |
#include <linux/kthread.h> #include <linux/freezer.h> |
83358ece2 mm/page_owner: in... |
22 |
#include <linux/page_owner.h> |
748446bb6 mm: compaction: m... |
23 |
#include "internal.h" |
010fc29a4 compaction: fix b... |
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 mm: compaction: e... |
38 |
#if defined CONFIG_COMPACTION || defined CONFIG_CMA |
b7aba6984 mm: compaction: a... |
39 40 |
#define CREATE_TRACE_POINTS #include <trace/events/compaction.h> |
06b6640a3 mm, compaction: w... |
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 mm: compaction: m... |
45 46 47 |
static unsigned long release_freepages(struct list_head *freelist) { struct page *page, *next; |
6bace090a mm, compaction: a... |
48 |
unsigned long high_pfn = 0; |
748446bb6 mm: compaction: m... |
49 50 |
list_for_each_entry_safe(page, next, freelist, lru) { |
6bace090a mm, compaction: a... |
51 |
unsigned long pfn = page_to_pfn(page); |
748446bb6 mm: compaction: m... |
52 53 |
list_del(&page->lru); __free_page(page); |
6bace090a mm, compaction: a... |
54 55 |
if (pfn > high_pfn) high_pfn = pfn; |
748446bb6 mm: compaction: m... |
56 |
} |
6bace090a mm, compaction: a... |
57 |
return high_pfn; |
748446bb6 mm: compaction: m... |
58 |
} |
ff9543fd3 mm: compaction: e... |
59 60 |
static void map_pages(struct list_head *list) { |
66c64223a mm/compaction: sp... |
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 mm/compaction: sp... |
70 |
|
46f24fd85 mm/page_alloc: in... |
71 |
post_alloc_hook(page, order, __GFP_MOVABLE); |
66c64223a mm/compaction: sp... |
72 73 |
if (order) split_page(page, order); |
ff9543fd3 mm: compaction: e... |
74 |
|
66c64223a mm/compaction: sp... |
75 76 77 78 |
for (i = 0; i < nr_pages; i++) { list_add(&page->lru, &tmp_list); page++; } |
ff9543fd3 mm: compaction: e... |
79 |
} |
66c64223a mm/compaction: sp... |
80 81 |
list_splice(&tmp_list, list); |
ff9543fd3 mm: compaction: e... |
82 |
} |
47118af07 mm: mmzone: MIGRA... |
83 84 85 86 |
static inline bool migrate_async_suitable(int migratetype) { return is_migrate_cma(migratetype) || migratetype == MIGRATE_MOVABLE; } |
bb13ffeb9 mm: compaction: c... |
87 |
#ifdef CONFIG_COMPACTION |
24e2716f6 mm/compaction: ad... |
88 |
|
bda807d44 mm: migrate: supp... |
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 mm/compaction: ad... |
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 mm: compaction: c... |
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 mm, compaction: e... |
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 mm/compaction: fi... |
208 |
zone->compact_cached_free_pfn = |
06b6640a3 mm, compaction: w... |
209 |
pageblock_start_pfn(zone_end_pfn(zone) - 1); |
02333641e mm, compaction: e... |
210 |
} |
bb13ffeb9 mm: compaction: c... |
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 mm: compaction: c... |
216 |
static void __reset_isolation_suitable(struct zone *zone) |
bb13ffeb9 mm: compaction: c... |
217 218 |
{ unsigned long start_pfn = zone->zone_start_pfn; |
108bcc96e mm: add & use zon... |
219 |
unsigned long end_pfn = zone_end_pfn(zone); |
bb13ffeb9 mm: compaction: c... |
220 |
unsigned long pfn; |
62997027c mm: compaction: c... |
221 |
zone->compact_blockskip_flush = false; |
bb13ffeb9 mm: compaction: c... |
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 mm, compaction: e... |
238 239 |
reset_cached_positions(zone); |
bb13ffeb9 mm: compaction: c... |
240 |
} |
62997027c mm: compaction: c... |
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 mm: compaction: c... |
255 256 |
/* * If no pages were isolated then mark this pageblock to be skipped in the |
62997027c mm: compaction: c... |
257 |
* future. The information is later cleared by __reset_isolation_suitable(). |
bb13ffeb9 mm: compaction: c... |
258 |
*/ |
c89511ab2 mm: compaction: R... |
259 260 |
static void update_pageblock_skip(struct compact_control *cc, struct page *page, unsigned long nr_isolated, |
edc2ca612 mm, compaction: m... |
261 |
bool migrate_scanner) |
bb13ffeb9 mm: compaction: c... |
262 |
{ |
c89511ab2 mm: compaction: R... |
263 |
struct zone *zone = cc->zone; |
35979ef33 mm, compaction: a... |
264 |
unsigned long pfn; |
6815bf3f2 mm/compaction: re... |
265 266 267 |
if (cc->ignore_skip_hint) return; |
bb13ffeb9 mm: compaction: c... |
268 269 |
if (!page) return; |
35979ef33 mm, compaction: a... |
270 271 |
if (nr_isolated) return; |
edc2ca612 mm, compaction: m... |
272 |
set_pageblock_skip(page); |
c89511ab2 mm: compaction: R... |
273 |
|
35979ef33 mm, compaction: a... |
274 275 276 277 |
pfn = page_to_pfn(page); /* Update where async and sync compaction should restart */ if (migrate_scanner) { |
35979ef33 mm, compaction: a... |
278 279 |
if (pfn > zone->compact_cached_migrate_pfn[0]) zone->compact_cached_migrate_pfn[0] = pfn; |
e0b9daeb4 mm, compaction: e... |
280 281 |
if (cc->mode != MIGRATE_ASYNC && pfn > zone->compact_cached_migrate_pfn[1]) |
35979ef33 mm, compaction: a... |
282 283 |
zone->compact_cached_migrate_pfn[1] = pfn; } else { |
35979ef33 mm, compaction: a... |
284 285 |
if (pfn < zone->compact_cached_free_pfn) zone->compact_cached_free_pfn = pfn; |
c89511ab2 mm: compaction: R... |
286 |
} |
bb13ffeb9 mm: compaction: c... |
287 288 289 290 291 292 293 |
} #else static inline bool isolation_suitable(struct compact_control *cc, struct page *page) { return true; } |
c89511ab2 mm: compaction: R... |
294 295 |
static void update_pageblock_skip(struct compact_control *cc, struct page *page, unsigned long nr_isolated, |
edc2ca612 mm, compaction: m... |
296 |
bool migrate_scanner) |
bb13ffeb9 mm: compaction: c... |
297 298 299 |
{ } #endif /* CONFIG_COMPACTION */ |
8b44d2791 mm, compaction: p... |
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 mm: compaction: a... |
310 |
{ |
8b44d2791 mm, compaction: p... |
311 312 |
if (cc->mode == MIGRATE_ASYNC) { if (!spin_trylock_irqsave(lock, *flags)) { |
c3486f537 mm, compaction: s... |
313 |
cc->contended = true; |
8b44d2791 mm, compaction: p... |
314 315 316 317 318 |
return false; } } else { spin_lock_irqsave(lock, *flags); } |
1f9efdef4 mm, compaction: k... |
319 |
|
8b44d2791 mm, compaction: p... |
320 |
return true; |
2a1402aa0 mm: compaction: a... |
321 |
} |
85aa125f0 mm: compaction: i... |
322 |
/* |
c67fe3752 mm: compaction: A... |
323 |
* Compaction requires the taking of some coarse locks that are potentially |
8b44d2791 mm, compaction: p... |
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 mm: compaction: A... |
331 |
* |
8b44d2791 mm, compaction: p... |
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 mm: compaction: A... |
336 |
*/ |
8b44d2791 mm, compaction: p... |
337 338 |
static bool compact_unlock_should_abort(spinlock_t *lock, unsigned long flags, bool *locked, struct compact_control *cc) |
c67fe3752 mm: compaction: A... |
339 |
{ |
8b44d2791 mm, compaction: p... |
340 341 342 343 |
if (*locked) { spin_unlock_irqrestore(lock, flags); *locked = false; } |
1f9efdef4 mm, compaction: k... |
344 |
|
8b44d2791 mm, compaction: p... |
345 |
if (fatal_signal_pending(current)) { |
c3486f537 mm, compaction: s... |
346 |
cc->contended = true; |
8b44d2791 mm, compaction: p... |
347 348 |
return true; } |
c67fe3752 mm: compaction: A... |
349 |
|
8b44d2791 mm, compaction: p... |
350 |
if (need_resched()) { |
e0b9daeb4 mm, compaction: e... |
351 |
if (cc->mode == MIGRATE_ASYNC) { |
c3486f537 mm, compaction: s... |
352 |
cc->contended = true; |
8b44d2791 mm, compaction: p... |
353 |
return true; |
c67fe3752 mm: compaction: A... |
354 |
} |
c67fe3752 mm: compaction: A... |
355 |
cond_resched(); |
c67fe3752 mm: compaction: A... |
356 |
} |
8b44d2791 mm, compaction: p... |
357 |
return false; |
c67fe3752 mm: compaction: A... |
358 |
} |
be9765722 mm, compaction: p... |
359 360 361 |
/* * Aside from avoiding lock contention, compaction also periodically checks * need_resched() and either schedules in sync compaction or aborts async |
8b44d2791 mm, compaction: p... |
362 |
* compaction. This is similar to what compact_unlock_should_abort() does, but |
be9765722 mm, compaction: p... |
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 mm, compaction: s... |
373 |
cc->contended = true; |
be9765722 mm, compaction: p... |
374 375 376 377 378 379 380 381 |
return true; } cond_resched(); } return false; } |
c67fe3752 mm: compaction: A... |
382 |
/* |
9e4be4708 mm/compaction.c: ... |
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 mm: compaction: i... |
386 |
*/ |
f40d1e42b mm: compaction: a... |
387 |
static unsigned long isolate_freepages_block(struct compact_control *cc, |
e14c720ef mm, compaction: r... |
388 |
unsigned long *start_pfn, |
85aa125f0 mm: compaction: i... |
389 390 391 |
unsigned long end_pfn, struct list_head *freelist, bool strict) |
748446bb6 mm: compaction: m... |
392 |
{ |
b7aba6984 mm: compaction: a... |
393 |
int nr_scanned = 0, total_isolated = 0; |
bb13ffeb9 mm: compaction: c... |
394 |
struct page *cursor, *valid_page = NULL; |
b8b2d8253 mm/compaction.c: ... |
395 |
unsigned long flags = 0; |
f40d1e42b mm: compaction: a... |
396 |
bool locked = false; |
e14c720ef mm, compaction: r... |
397 |
unsigned long blockpfn = *start_pfn; |
66c64223a mm/compaction: sp... |
398 |
unsigned int order; |
748446bb6 mm: compaction: m... |
399 |
|
748446bb6 mm: compaction: m... |
400 |
cursor = pfn_to_page(blockpfn); |
f40d1e42b mm: compaction: a... |
401 |
/* Isolate free pages. */ |
748446bb6 mm: compaction: m... |
402 |
for (; blockpfn < end_pfn; blockpfn++, cursor++) { |
66c64223a mm/compaction: sp... |
403 |
int isolated; |
748446bb6 mm: compaction: m... |
404 |
struct page *page = cursor; |
8b44d2791 mm, compaction: p... |
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 mm: compaction: a... |
414 |
nr_scanned++; |
f40d1e42b mm: compaction: a... |
415 |
if (!pfn_valid_within(blockpfn)) |
2af120bc0 mm/compaction: br... |
416 |
goto isolate_fail; |
bb13ffeb9 mm: compaction: c... |
417 418 |
if (!valid_page) valid_page = page; |
9fcd6d2e0 mm, compaction: s... |
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 mm: compaction: a... |
436 |
if (!PageBuddy(page)) |
2af120bc0 mm/compaction: br... |
437 |
goto isolate_fail; |
f40d1e42b mm: compaction: a... |
438 439 |
/* |
69b7189f1 mm, compaction: s... |
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 mm: compaction: a... |
445 |
*/ |
69b7189f1 mm, compaction: s... |
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 mm, compaction: p... |
455 456 |
locked = compact_trylock_irqsave(&cc->zone->lock, &flags, cc); |
69b7189f1 mm, compaction: s... |
457 458 |
if (!locked) break; |
f40d1e42b mm: compaction: a... |
459 |
|
69b7189f1 mm, compaction: s... |
460 461 462 463 |
/* Recheck this is a buddy page under lock */ if (!PageBuddy(page)) goto isolate_fail; } |
748446bb6 mm: compaction: m... |
464 |
|
66c64223a mm/compaction: sp... |
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 mm, compaction: a... |
468 469 |
if (!isolated) break; |
66c64223a mm/compaction: sp... |
470 |
set_page_private(page, order); |
a4f04f2c6 mm, compaction: a... |
471 |
|
748446bb6 mm: compaction: m... |
472 |
total_isolated += isolated; |
a4f04f2c6 mm, compaction: a... |
473 |
cc->nr_freepages += isolated; |
66c64223a mm/compaction: sp... |
474 |
list_add_tail(&page->lru, freelist); |
a4f04f2c6 mm, compaction: a... |
475 476 477 |
if (!strict && cc->nr_migratepages <= cc->nr_freepages) { blockpfn += isolated; break; |
748446bb6 mm: compaction: m... |
478 |
} |
a4f04f2c6 mm, compaction: a... |
479 480 481 482 |
/* Advance to the end of split page */ blockpfn += isolated - 1; cursor += isolated - 1; continue; |
2af120bc0 mm/compaction: br... |
483 484 485 486 487 488 |
isolate_fail: if (strict) break; else continue; |
748446bb6 mm: compaction: m... |
489 |
} |
a4f04f2c6 mm, compaction: a... |
490 491 |
if (locked) spin_unlock_irqrestore(&cc->zone->lock, flags); |
9fcd6d2e0 mm, compaction: s... |
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 mm/compaction: pr... |
498 499 |
trace_mm_compaction_isolate_freepages(*start_pfn, blockpfn, nr_scanned, total_isolated); |
e14c720ef mm, compaction: r... |
500 501 |
/* Record how far we have got within the block */ *start_pfn = blockpfn; |
f40d1e42b mm: compaction: a... |
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 mm/compaction: br... |
507 |
if (strict && blockpfn < end_pfn) |
f40d1e42b mm: compaction: a... |
508 |
total_isolated = 0; |
bb13ffeb9 mm: compaction: c... |
509 510 |
/* Update the pageblock-skip if the whole pageblock was scanned */ if (blockpfn == end_pfn) |
edc2ca612 mm, compaction: m... |
511 |
update_pageblock_skip(cc, valid_page, total_isolated, false); |
bb13ffeb9 mm: compaction: c... |
512 |
|
010fc29a4 compaction: fix b... |
513 |
count_compact_events(COMPACTFREE_SCANNED, nr_scanned); |
397487db6 mm: compaction: A... |
514 |
if (total_isolated) |
010fc29a4 compaction: fix b... |
515 |
count_compact_events(COMPACTISOLATED, total_isolated); |
748446bb6 mm: compaction: m... |
516 517 |
return total_isolated; } |
85aa125f0 mm: compaction: i... |
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 mm: compaction: e... |
531 |
unsigned long |
bb13ffeb9 mm: compaction: c... |
532 533 |
isolate_freepages_range(struct compact_control *cc, unsigned long start_pfn, unsigned long end_pfn) |
85aa125f0 mm: compaction: i... |
534 |
{ |
e1409c325 mm/compaction: pa... |
535 |
unsigned long isolated, pfn, block_start_pfn, block_end_pfn; |
85aa125f0 mm: compaction: i... |
536 |
LIST_HEAD(freelist); |
7d49d8868 mm, compaction: r... |
537 |
pfn = start_pfn; |
06b6640a3 mm, compaction: w... |
538 |
block_start_pfn = pageblock_start_pfn(pfn); |
e1409c325 mm/compaction: pa... |
539 540 |
if (block_start_pfn < cc->zone->zone_start_pfn) block_start_pfn = cc->zone->zone_start_pfn; |
06b6640a3 mm, compaction: w... |
541 |
block_end_pfn = pageblock_end_pfn(pfn); |
7d49d8868 mm, compaction: r... |
542 543 |
for (; pfn < end_pfn; pfn += isolated, |
e1409c325 mm/compaction: pa... |
544 |
block_start_pfn = block_end_pfn, |
7d49d8868 mm, compaction: r... |
545 |
block_end_pfn += pageblock_nr_pages) { |
e14c720ef mm, compaction: r... |
546 547 |
/* Protect pfn from changing by isolate_freepages_block */ unsigned long isolate_start_pfn = pfn; |
85aa125f0 mm: compaction: i... |
548 |
|
85aa125f0 mm: compaction: i... |
549 |
block_end_pfn = min(block_end_pfn, end_pfn); |
584200163 mm/compaction: sk... |
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 mm, compaction: w... |
556 557 |
block_start_pfn = pageblock_start_pfn(pfn); block_end_pfn = pageblock_end_pfn(pfn); |
584200163 mm/compaction: sk... |
558 559 |
block_end_pfn = min(block_end_pfn, end_pfn); } |
e1409c325 mm/compaction: pa... |
560 561 |
if (!pageblock_pfn_to_page(block_start_pfn, block_end_pfn, cc->zone)) |
7d49d8868 mm, compaction: r... |
562 |
break; |
e14c720ef mm, compaction: r... |
563 564 |
isolated = isolate_freepages_block(cc, &isolate_start_pfn, block_end_pfn, &freelist, true); |
85aa125f0 mm: compaction: i... |
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 mm/compaction: sp... |
580 |
/* __isolate_free_page() does not map the pages */ |
85aa125f0 mm: compaction: i... |
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 mm: compaction: m... |
592 593 594 |
/* Similar to reclaim, but different enough that they don't share logic */ static bool too_many_isolated(struct zone *zone) { |
bc6930457 mm: compaction: h... |
595 |
unsigned long active, inactive, isolated; |
748446bb6 mm: compaction: m... |
596 |
|
599d0c954 mm, vmscan: move ... |
597 598 599 600 601 602 |
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 mm: compaction: m... |
603 |
|
bc6930457 mm: compaction: h... |
604 |
return isolated > (inactive + active) / 2; |
748446bb6 mm: compaction: m... |
605 |
} |
2fe86e000 mm: compaction: i... |
606 |
/** |
edc2ca612 mm, compaction: m... |
607 608 |
* isolate_migratepages_block() - isolate all migrate-able pages within * a single pageblock |
2fe86e000 mm: compaction: i... |
609 |
* @cc: Compaction control structure. |
edc2ca612 mm, compaction: m... |
610 611 612 |
* @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 mm: compaction: i... |
613 614 |
* * Isolate all pages that can be migrated from the range specified by |
edc2ca612 mm, compaction: m... |
615 616 617 618 |
* [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 mm: compaction: i... |
619 |
* |
edc2ca612 mm, compaction: m... |
620 621 622 |
* 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 mm: compaction: m... |
623 |
*/ |
edc2ca612 mm, compaction: m... |
624 625 626 |
static unsigned long isolate_migratepages_block(struct compact_control *cc, unsigned long low_pfn, unsigned long end_pfn, isolate_mode_t isolate_mode) |
748446bb6 mm: compaction: m... |
627 |
{ |
edc2ca612 mm, compaction: m... |
628 |
struct zone *zone = cc->zone; |
b7aba6984 mm: compaction: a... |
629 |
unsigned long nr_scanned = 0, nr_isolated = 0; |
fa9add641 mm/memcg: apply a... |
630 |
struct lruvec *lruvec; |
b8b2d8253 mm/compaction.c: ... |
631 |
unsigned long flags = 0; |
2a1402aa0 mm: compaction: a... |
632 |
bool locked = false; |
bb13ffeb9 mm: compaction: c... |
633 |
struct page *page = NULL, *valid_page = NULL; |
e34d85f0e mm/compaction: pr... |
634 |
unsigned long start_pfn = low_pfn; |
fdd048e12 mm, compaction: s... |
635 636 |
bool skip_on_failure = false; unsigned long next_skip_pfn = 0; |
748446bb6 mm: compaction: m... |
637 |
|
748446bb6 mm: compaction: m... |
638 639 640 641 642 643 |
/* * 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 mm: compaction: a... |
644 |
/* async migration should just abort */ |
e0b9daeb4 mm, compaction: e... |
645 |
if (cc->mode == MIGRATE_ASYNC) |
2fe86e000 mm: compaction: i... |
646 |
return 0; |
f9e35b3b4 mm: compaction: a... |
647 |
|
748446bb6 mm: compaction: m... |
648 649 650 |
congestion_wait(BLK_RW_ASYNC, HZ/10); if (fatal_signal_pending(current)) |
2fe86e000 mm: compaction: i... |
651 |
return 0; |
748446bb6 mm: compaction: m... |
652 |
} |
be9765722 mm, compaction: p... |
653 654 |
if (compact_should_abort(cc)) return 0; |
aeef4b838 mm, compaction: t... |
655 |
|
fdd048e12 mm, compaction: s... |
656 657 658 659 |
if (cc->direct_compaction && (cc->mode == MIGRATE_ASYNC)) { skip_on_failure = true; next_skip_pfn = block_end_pfn(low_pfn, cc->order); } |
748446bb6 mm: compaction: m... |
660 |
/* Time to isolate some pages for migration */ |
748446bb6 mm: compaction: m... |
661 |
for (; low_pfn < end_pfn; low_pfn++) { |
29c0dde83 mm, compaction: a... |
662 |
|
fdd048e12 mm, compaction: s... |
663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 |
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 mm, compaction: p... |
684 685 686 687 688 689 |
/* * 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 mm, vmscan: move ... |
690 |
&& compact_unlock_should_abort(zone_lru_lock(zone), flags, |
8b44d2791 mm, compaction: p... |
691 692 |
&locked, cc)) break; |
c67fe3752 mm: compaction: A... |
693 |
|
748446bb6 mm: compaction: m... |
694 |
if (!pfn_valid_within(low_pfn)) |
fdd048e12 mm, compaction: s... |
695 |
goto isolate_fail; |
b7aba6984 mm: compaction: a... |
696 |
nr_scanned++; |
748446bb6 mm: compaction: m... |
697 |
|
748446bb6 mm: compaction: m... |
698 |
page = pfn_to_page(low_pfn); |
dc9086004 mm: compaction: c... |
699 |
|
bb13ffeb9 mm: compaction: c... |
700 701 |
if (!valid_page) valid_page = page; |
6c14466cc mm: improve docum... |
702 |
/* |
99c0fd5e5 mm, compaction: s... |
703 704 705 706 |
* 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 mm: improve docum... |
707 |
*/ |
99c0fd5e5 mm, compaction: s... |
708 709 710 711 712 713 714 715 716 717 |
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 mm: compaction: m... |
718 |
continue; |
99c0fd5e5 mm, compaction: s... |
719 |
} |
748446bb6 mm: compaction: m... |
720 |
|
9927af740 mm: compaction: p... |
721 |
/* |
29c0dde83 mm, compaction: a... |
722 723 724 725 726 |
* 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 thp: transhuge is... |
727 |
*/ |
29c0dde83 mm, compaction: a... |
728 729 730 731 732 |
if (PageCompound(page)) { unsigned int comp_order = compound_order(page); if (likely(comp_order < MAX_ORDER)) low_pfn += (1UL << comp_order) - 1; |
edc2ca612 mm, compaction: m... |
733 |
|
fdd048e12 mm, compaction: s... |
734 |
goto isolate_fail; |
2a1402aa0 mm: compaction: a... |
735 |
} |
bda807d44 mm: migrate: supp... |
736 737 738 739 740 741 |
/* * 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 mm: migrate: supp... |
742 743 744 745 746 747 748 |
/* * __PageMovable can return false positive so we need * to verify it under page_lock. */ if (unlikely(__PageMovable(page)) && !PageIsolated(page)) { if (locked) { |
a52633d8e mm, vmscan: move ... |
749 |
spin_unlock_irqrestore(zone_lru_lock(zone), |
bda807d44 mm: migrate: supp... |
750 751 752 753 754 755 756 |
flags); locked = false; } if (isolate_movable_page(page, isolate_mode)) goto isolate_success; } |
fdd048e12 mm, compaction: s... |
757 |
goto isolate_fail; |
bda807d44 mm: migrate: supp... |
758 |
} |
29c0dde83 mm, compaction: a... |
759 |
|
119d6d59d mm, compaction: a... |
760 761 762 763 764 765 766 |
/* * 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 mm, compaction: s... |
767 |
goto isolate_fail; |
119d6d59d mm, compaction: a... |
768 |
|
69b7189f1 mm, compaction: s... |
769 770 |
/* If we already hold the lock, we can skip some rechecking */ if (!locked) { |
a52633d8e mm, vmscan: move ... |
771 |
locked = compact_trylock_irqsave(zone_lru_lock(zone), |
8b44d2791 mm, compaction: p... |
772 |
&flags, cc); |
69b7189f1 mm, compaction: s... |
773 774 |
if (!locked) break; |
2a1402aa0 mm: compaction: a... |
775 |
|
29c0dde83 mm, compaction: a... |
776 |
/* Recheck PageLRU and PageCompound under lock */ |
69b7189f1 mm, compaction: s... |
777 |
if (!PageLRU(page)) |
fdd048e12 mm, compaction: s... |
778 |
goto isolate_fail; |
29c0dde83 mm, compaction: a... |
779 780 781 782 783 784 785 786 |
/* * 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 mm, compaction: s... |
787 |
goto isolate_fail; |
69b7189f1 mm, compaction: s... |
788 |
} |
bc835011a thp: transhuge is... |
789 |
} |
599d0c954 mm, vmscan: move ... |
790 |
lruvec = mem_cgroup_page_lruvec(page, zone->zone_pgdat); |
fa9add641 mm/memcg: apply a... |
791 |
|
748446bb6 mm: compaction: m... |
792 |
/* Try isolate the page */ |
edc2ca612 mm, compaction: m... |
793 |
if (__isolate_lru_page(page, isolate_mode) != 0) |
fdd048e12 mm, compaction: s... |
794 |
goto isolate_fail; |
748446bb6 mm: compaction: m... |
795 |
|
29c0dde83 mm, compaction: a... |
796 |
VM_BUG_ON_PAGE(PageCompound(page), page); |
bc835011a thp: transhuge is... |
797 |
|
748446bb6 mm: compaction: m... |
798 |
/* Successfully isolated */ |
fa9add641 mm/memcg: apply a... |
799 |
del_page_from_lru_list(page, lruvec, page_lru(page)); |
5d7d362ab mm, compaction: f... |
800 801 |
inc_node_page_state(page, NR_ISOLATED_ANON + page_is_file_cache(page)); |
b6c750163 mm/compaction: cl... |
802 803 |
isolate_success: |
fdd048e12 mm, compaction: s... |
804 |
list_add(&page->lru, &cc->migratepages); |
748446bb6 mm: compaction: m... |
805 |
cc->nr_migratepages++; |
b7aba6984 mm: compaction: a... |
806 |
nr_isolated++; |
748446bb6 mm: compaction: m... |
807 |
|
a34753d27 mm, compaction: r... |
808 809 810 811 812 813 814 815 |
/* * 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 mm: compaction: m... |
816 |
/* Avoid isolating too much */ |
31b8384a5 mm: compaction: p... |
817 818 |
if (cc->nr_migratepages == COMPACT_CLUSTER_MAX) { ++low_pfn; |
748446bb6 mm: compaction: m... |
819 |
break; |
31b8384a5 mm: compaction: p... |
820 |
} |
fdd048e12 mm, compaction: s... |
821 822 823 824 825 826 827 828 829 830 831 832 833 |
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 mm, vmscan: move ... |
834 |
spin_unlock_irqrestore(zone_lru_lock(zone), flags); |
fdd048e12 mm, compaction: s... |
835 836 |
locked = false; } |
fdd048e12 mm, compaction: s... |
837 838 839 840 841 842 843 844 845 846 847 848 849 850 |
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 mm: compaction: m... |
851 |
} |
99c0fd5e5 mm, compaction: s... |
852 853 854 855 856 857 |
/* * 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 mm: compaction: A... |
858 |
if (locked) |
a52633d8e mm, vmscan: move ... |
859 |
spin_unlock_irqrestore(zone_lru_lock(zone), flags); |
748446bb6 mm: compaction: m... |
860 |
|
50b5b094e mm: compaction: d... |
861 862 863 |
/* * Update the pageblock-skip information and cached scanner pfn, * if the whole pageblock was scanned without isolating any page. |
50b5b094e mm: compaction: d... |
864 |
*/ |
35979ef33 mm, compaction: a... |
865 |
if (low_pfn == end_pfn) |
edc2ca612 mm, compaction: m... |
866 |
update_pageblock_skip(cc, valid_page, nr_isolated, true); |
bb13ffeb9 mm: compaction: c... |
867 |
|
e34d85f0e mm/compaction: pr... |
868 869 |
trace_mm_compaction_isolate_migratepages(start_pfn, low_pfn, nr_scanned, nr_isolated); |
b7aba6984 mm: compaction: a... |
870 |
|
010fc29a4 compaction: fix b... |
871 |
count_compact_events(COMPACTMIGRATE_SCANNED, nr_scanned); |
397487db6 mm: compaction: A... |
872 |
if (nr_isolated) |
010fc29a4 compaction: fix b... |
873 |
count_compact_events(COMPACTISOLATED, nr_isolated); |
397487db6 mm: compaction: A... |
874 |
|
2fe86e000 mm: compaction: i... |
875 876 |
return low_pfn; } |
edc2ca612 mm, compaction: m... |
877 878 879 880 881 882 883 884 885 886 887 888 889 890 |
/** * 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 mm/compaction: pa... |
891 |
unsigned long pfn, block_start_pfn, block_end_pfn; |
edc2ca612 mm, compaction: m... |
892 893 894 |
/* Scan block by block. First and last block may be incomplete */ pfn = start_pfn; |
06b6640a3 mm, compaction: w... |
895 |
block_start_pfn = pageblock_start_pfn(pfn); |
e1409c325 mm/compaction: pa... |
896 897 |
if (block_start_pfn < cc->zone->zone_start_pfn) block_start_pfn = cc->zone->zone_start_pfn; |
06b6640a3 mm, compaction: w... |
898 |
block_end_pfn = pageblock_end_pfn(pfn); |
edc2ca612 mm, compaction: m... |
899 900 |
for (; pfn < end_pfn; pfn = block_end_pfn, |
e1409c325 mm/compaction: pa... |
901 |
block_start_pfn = block_end_pfn, |
edc2ca612 mm, compaction: m... |
902 903 904 |
block_end_pfn += pageblock_nr_pages) { block_end_pfn = min(block_end_pfn, end_pfn); |
e1409c325 mm/compaction: pa... |
905 906 |
if (!pageblock_pfn_to_page(block_start_pfn, block_end_pfn, cc->zone)) |
edc2ca612 mm, compaction: m... |
907 908 909 910 |
continue; pfn = isolate_migratepages_block(cc, pfn, block_end_pfn, ISOLATE_UNEVICTABLE); |
14af4a5e9 mm, cma: prevent ... |
911 |
if (!pfn) |
edc2ca612 mm, compaction: m... |
912 |
break; |
6ea41c0c0 mm/compaction.c: ... |
913 914 915 |
if (cc->nr_migratepages == COMPACT_CLUSTER_MAX) break; |
edc2ca612 mm, compaction: m... |
916 |
} |
edc2ca612 mm, compaction: m... |
917 918 919 |
return pfn; } |
ff9543fd3 mm: compaction: e... |
920 921 |
#endif /* CONFIG_COMPACTION || CONFIG_CMA */ #ifdef CONFIG_COMPACTION |
018e9a49a mm/compaction.c: ... |
922 923 |
/* Returns true if the page is within a block suitable for migration to */ |
9f7e33879 mm, compaction: m... |
924 925 |
static bool suitable_migration_target(struct compact_control *cc, struct page *page) |
018e9a49a mm/compaction.c: ... |
926 |
{ |
9f7e33879 mm, compaction: m... |
927 928 |
if (cc->ignore_block_suitable) return true; |
018e9a49a mm/compaction.c: ... |
929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 |
/* 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 mm: compaction: i... |
947 |
/* |
f2849aa09 mm, compaction: m... |
948 949 950 951 952 953 954 955 956 957 |
* 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 mm: compaction: e... |
958 959 |
* Based on information in the current compact_control, find blocks * suitable for isolating free pages from and then isolate them. |
2fe86e000 mm: compaction: i... |
960 |
*/ |
edc2ca612 mm, compaction: m... |
961 |
static void isolate_freepages(struct compact_control *cc) |
2fe86e000 mm: compaction: i... |
962 |
{ |
edc2ca612 mm, compaction: m... |
963 |
struct zone *zone = cc->zone; |
ff9543fd3 mm: compaction: e... |
964 |
struct page *page; |
c96b9e508 mm/compaction: cl... |
965 |
unsigned long block_start_pfn; /* start of current pageblock */ |
e14c720ef mm, compaction: r... |
966 |
unsigned long isolate_start_pfn; /* exact pfn we start at */ |
c96b9e508 mm/compaction: cl... |
967 968 |
unsigned long block_end_pfn; /* end of current pageblock */ unsigned long low_pfn; /* lowest pfn scanner is able to scan */ |
ff9543fd3 mm: compaction: e... |
969 |
struct list_head *freelist = &cc->freepages; |
2fe86e000 mm: compaction: i... |
970 |
|
ff9543fd3 mm: compaction: e... |
971 972 |
/* * Initialise the free scanner. The starting point is where we last |
49e068f0b mm/compaction: ma... |
973 |
* successfully isolated from, zone-cached value, or the end of the |
e14c720ef mm, compaction: r... |
974 975 |
* zone when isolating for the first time. For looping we also need * this pfn aligned down to the pageblock boundary, because we do |
c96b9e508 mm/compaction: cl... |
976 977 978 |
* 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 mm/compaction: ma... |
979 980 |
* The low boundary is the end of the pageblock the migration scanner * is using. |
ff9543fd3 mm: compaction: e... |
981 |
*/ |
e14c720ef mm, compaction: r... |
982 |
isolate_start_pfn = cc->free_pfn; |
06b6640a3 mm, compaction: w... |
983 |
block_start_pfn = pageblock_start_pfn(cc->free_pfn); |
c96b9e508 mm/compaction: cl... |
984 985 |
block_end_pfn = min(block_start_pfn + pageblock_nr_pages, zone_end_pfn(zone)); |
06b6640a3 mm, compaction: w... |
986 |
low_pfn = pageblock_end_pfn(cc->migrate_pfn); |
2fe86e000 mm: compaction: i... |
987 |
|
ff9543fd3 mm: compaction: e... |
988 |
/* |
ff9543fd3 mm: compaction: e... |
989 990 991 992 |
* 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 mm, compaction: s... |
993 |
for (; block_start_pfn >= low_pfn; |
c96b9e508 mm/compaction: cl... |
994 |
block_end_pfn = block_start_pfn, |
e14c720ef mm, compaction: r... |
995 996 |
block_start_pfn -= pageblock_nr_pages, isolate_start_pfn = block_start_pfn) { |
f6ea3adb7 mm/compaction.c: ... |
997 998 999 |
/* * This can iterate a massively long zone without finding any * suitable migration targets, so periodically check if we need |
be9765722 mm, compaction: p... |
1000 |
* to schedule, or even abort async compaction. |
f6ea3adb7 mm/compaction.c: ... |
1001 |
*/ |
be9765722 mm, compaction: p... |
1002 1003 1004 |
if (!(block_start_pfn % (SWAP_CLUSTER_MAX * pageblock_nr_pages)) && compact_should_abort(cc)) break; |
f6ea3adb7 mm/compaction.c: ... |
1005 |
|
7d49d8868 mm, compaction: r... |
1006 1007 1008 |
page = pageblock_pfn_to_page(block_start_pfn, block_end_pfn, zone); if (!page) |
ff9543fd3 mm: compaction: e... |
1009 1010 1011 |
continue; /* Check the block is suitable for migration */ |
9f7e33879 mm, compaction: m... |
1012 |
if (!suitable_migration_target(cc, page)) |
ff9543fd3 mm: compaction: e... |
1013 |
continue; |
68e3e9262 Revert "mm: compa... |
1014 |
|
bb13ffeb9 mm: compaction: c... |
1015 1016 1017 |
/* If isolation recently failed, do not retry */ if (!isolation_suitable(cc, page)) continue; |
e14c720ef mm, compaction: r... |
1018 |
/* Found a block suitable for isolating free pages from. */ |
a46cbf3bc mm, compaction: p... |
1019 1020 |
isolate_freepages_block(cc, &isolate_start_pfn, block_end_pfn, freelist, false); |
ff9543fd3 mm: compaction: e... |
1021 1022 |
/* |
a46cbf3bc mm, compaction: p... |
1023 1024 |
* If we isolated enough freepages, or aborted due to lock * contention, terminate. |
e14c720ef mm, compaction: r... |
1025 |
*/ |
f5f61a320 mm, compaction: s... |
1026 1027 |
if ((cc->nr_freepages >= cc->nr_migratepages) || cc->contended) { |
a46cbf3bc mm, compaction: p... |
1028 1029 1030 1031 1032 |
if (isolate_start_pfn >= block_end_pfn) { /* * Restart at previous pageblock if more * freepages can be isolated next time. */ |
f5f61a320 mm, compaction: s... |
1033 1034 |
isolate_start_pfn = block_start_pfn - pageblock_nr_pages; |
a46cbf3bc mm, compaction: p... |
1035 |
} |
be9765722 mm, compaction: p... |
1036 |
break; |
a46cbf3bc mm, compaction: p... |
1037 |
} else if (isolate_start_pfn < block_end_pfn) { |
f5f61a320 mm, compaction: s... |
1038 |
/* |
a46cbf3bc mm, compaction: p... |
1039 1040 |
* If isolation failed early, do not continue * needlessly. |
f5f61a320 mm, compaction: s... |
1041 |
*/ |
a46cbf3bc mm, compaction: p... |
1042 |
break; |
f5f61a320 mm, compaction: s... |
1043 |
} |
ff9543fd3 mm: compaction: e... |
1044 |
} |
66c64223a mm/compaction: sp... |
1045 |
/* __isolate_free_page() does not map the pages */ |
ff9543fd3 mm: compaction: e... |
1046 |
map_pages(freelist); |
7ed695e06 mm: compaction: d... |
1047 |
/* |
f5f61a320 mm, compaction: s... |
1048 1049 1050 1051 |
* 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 mm: compaction: d... |
1052 |
*/ |
f5f61a320 mm, compaction: s... |
1053 |
cc->free_pfn = isolate_start_pfn; |
748446bb6 mm: compaction: m... |
1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 |
} /* * 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 mm, compaction: p... |
1066 1067 1068 1069 |
/* * Isolate free pages if necessary, and if we are not aborting due to * contention. */ |
748446bb6 mm: compaction: m... |
1070 |
if (list_empty(&cc->freepages)) { |
be9765722 mm, compaction: p... |
1071 |
if (!cc->contended) |
edc2ca612 mm, compaction: m... |
1072 |
isolate_freepages(cc); |
748446bb6 mm: compaction: m... |
1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 |
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 mm, compaction: r... |
1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 |
* 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 mm: compaction: e... |
1097 1098 1099 1100 1101 1102 1103 1104 |
/* 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 mm: allow compact... |
1105 1106 1107 1108 1109 1110 |
* Allow userspace to control policy on scanning the unevictable LRU for * compactable pages. */ int sysctl_compact_unevictable_allowed __read_mostly = 1; /* |
edc2ca612 mm, compaction: m... |
1111 1112 1113 |
* 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 mm: compaction: e... |
1114 1115 1116 1117 |
*/ static isolate_migrate_t isolate_migratepages(struct zone *zone, struct compact_control *cc) { |
e1409c325 mm/compaction: pa... |
1118 1119 1120 |
unsigned long block_start_pfn; unsigned long block_end_pfn; unsigned long low_pfn; |
edc2ca612 mm, compaction: m... |
1121 1122 |
struct page *page; const isolate_mode_t isolate_mode = |
5bbe3547a mm: allow compact... |
1123 |
(sysctl_compact_unevictable_allowed ? ISOLATE_UNEVICTABLE : 0) | |
1d2047fef mm, compaction: d... |
1124 |
(cc->mode != MIGRATE_SYNC ? ISOLATE_ASYNC_MIGRATE : 0); |
ff9543fd3 mm: compaction: e... |
1125 |
|
edc2ca612 mm, compaction: m... |
1126 1127 1128 1129 1130 |
/* * Start at where we last stopped, or beginning of the zone as * initialized by compact_zone() */ low_pfn = cc->migrate_pfn; |
06b6640a3 mm, compaction: w... |
1131 |
block_start_pfn = pageblock_start_pfn(low_pfn); |
e1409c325 mm/compaction: pa... |
1132 1133 |
if (block_start_pfn < zone->zone_start_pfn) block_start_pfn = zone->zone_start_pfn; |
ff9543fd3 mm: compaction: e... |
1134 1135 |
/* Only scan within a pageblock boundary */ |
06b6640a3 mm, compaction: w... |
1136 |
block_end_pfn = pageblock_end_pfn(low_pfn); |
ff9543fd3 mm: compaction: e... |
1137 |
|
edc2ca612 mm, compaction: m... |
1138 1139 1140 1141 |
/* * Iterate over whole pageblocks until we find the first suitable. * Do not cross the free scanner. */ |
e1409c325 mm/compaction: pa... |
1142 1143 1144 1145 |
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 mm: compaction: e... |
1146 |
|
edc2ca612 mm, compaction: m... |
1147 1148 1149 1150 1151 1152 1153 1154 |
/* * 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 mm: compaction: e... |
1155 |
|
e1409c325 mm/compaction: pa... |
1156 1157 |
page = pageblock_pfn_to_page(block_start_pfn, block_end_pfn, zone); |
7d49d8868 mm, compaction: r... |
1158 |
if (!page) |
edc2ca612 mm, compaction: m... |
1159 |
continue; |
edc2ca612 mm, compaction: m... |
1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 |
/* 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 mm/compaction: pa... |
1174 1175 |
low_pfn = isolate_migratepages_block(cc, low_pfn, block_end_pfn, isolate_mode); |
edc2ca612 mm, compaction: m... |
1176 |
|
5d7d362ab mm, compaction: f... |
1177 |
if (!low_pfn || cc->contended) |
edc2ca612 mm, compaction: m... |
1178 1179 1180 1181 1182 1183 1184 1185 1186 |
return ISOLATE_ABORT; /* * Either we isolated something and proceed with migration. Or * we failed and compact_zone should decide if we should * continue or not. */ break; } |
f2849aa09 mm, compaction: m... |
1187 1188 |
/* Record where migration scanner will be restarted. */ cc->migrate_pfn = low_pfn; |
ff9543fd3 mm: compaction: e... |
1189 |
|
edc2ca612 mm, compaction: m... |
1190 |
return cc->nr_migratepages ? ISOLATE_SUCCESS : ISOLATE_NONE; |
ff9543fd3 mm: compaction: e... |
1191 |
} |
21c527a3c mm/compaction.c: ... |
1192 1193 1194 1195 1196 1197 1198 1199 |
/* * 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 mm, compaction: c... |
1200 |
static enum compact_result __compact_finished(struct zone *zone, struct compact_control *cc, |
6d7ce5594 mm, compaction: p... |
1201 |
const int migratetype) |
748446bb6 mm: compaction: m... |
1202 |
{ |
8fb74b9fb mm: compaction: p... |
1203 |
unsigned int order; |
5a03b051e thp: use compacti... |
1204 |
unsigned long watermark; |
56de7263f mm: compaction: d... |
1205 |
|
be9765722 mm, compaction: p... |
1206 |
if (cc->contended || fatal_signal_pending(current)) |
2d1e10412 mm, compaction: d... |
1207 |
return COMPACT_CONTENDED; |
748446bb6 mm: compaction: m... |
1208 |
|
753341a4b revert "mm: have ... |
1209 |
/* Compaction run completes if the migrate and free scanner meet */ |
f2849aa09 mm, compaction: m... |
1210 |
if (compact_scanners_met(cc)) { |
55b7c4c99 mm: compaction: r... |
1211 |
/* Let the next compaction start anew. */ |
02333641e mm, compaction: e... |
1212 |
reset_cached_positions(zone); |
55b7c4c99 mm: compaction: r... |
1213 |
|
62997027c mm: compaction: c... |
1214 1215 |
/* * Mark that the PG_migrate_skip information should be cleared |
accf62422 mm, kswapd: repla... |
1216 |
* by kswapd when it goes to sleep. kcompactd does not set the |
62997027c mm: compaction: c... |
1217 1218 1219 |
* flag itself as the decision to be clear should be directly * based on an allocation request. */ |
accf62422 mm, kswapd: repla... |
1220 |
if (cc->direct_compaction) |
62997027c mm: compaction: c... |
1221 |
zone->compact_blockskip_flush = true; |
c8f7de0bf mm, compaction: d... |
1222 1223 1224 1225 |
if (cc->whole_zone) return COMPACT_COMPLETE; else return COMPACT_PARTIAL_SKIPPED; |
bb13ffeb9 mm: compaction: c... |
1226 |
} |
748446bb6 mm: compaction: m... |
1227 |
|
21c527a3c mm/compaction.c: ... |
1228 |
if (is_via_compact_memory(cc->order)) |
56de7263f mm: compaction: d... |
1229 |
return COMPACT_CONTINUE; |
3957c7768 mm: compaction: f... |
1230 |
/* Compaction run is not finished if the watermark is not met */ |
f2b8228c5 mm, compaction: u... |
1231 |
watermark = zone->watermark[cc->alloc_flags & ALLOC_WMARK_MASK]; |
3957c7768 mm: compaction: f... |
1232 |
|
ebff39801 mm, compaction: p... |
1233 1234 |
if (!zone_watermark_ok(zone, cc->order, watermark, cc->classzone_idx, cc->alloc_flags)) |
3957c7768 mm: compaction: f... |
1235 |
return COMPACT_CONTINUE; |
56de7263f mm: compaction: d... |
1236 |
/* Direct compactor: Is a suitable page free? */ |
8fb74b9fb mm: compaction: p... |
1237 1238 |
for (order = cc->order; order < MAX_ORDER; order++) { struct free_area *area = &zone->free_area[order]; |
2149cdaef mm/compaction: en... |
1239 |
bool can_steal; |
8fb74b9fb mm: compaction: p... |
1240 1241 |
/* Job done if page is free of the right migratetype */ |
6d7ce5594 mm, compaction: p... |
1242 |
if (!list_empty(&area->free_list[migratetype])) |
cf378319d mm, compaction: r... |
1243 |
return COMPACT_SUCCESS; |
8fb74b9fb mm: compaction: p... |
1244 |
|
2149cdaef mm/compaction: en... |
1245 1246 1247 1248 |
#ifdef CONFIG_CMA /* MIGRATE_MOVABLE can fallback on MIGRATE_CMA */ if (migratetype == MIGRATE_MOVABLE && !list_empty(&area->free_list[MIGRATE_CMA])) |
cf378319d mm, compaction: r... |
1249 |
return COMPACT_SUCCESS; |
2149cdaef mm/compaction: en... |
1250 1251 1252 1253 1254 1255 1256 |
#endif /* * Job done if allocation would steal freepages from * other migratetype buddy lists. */ if (find_suitable_fallback(area, order, migratetype, true, &can_steal) != -1) |
cf378319d mm, compaction: r... |
1257 |
return COMPACT_SUCCESS; |
56de7263f mm: compaction: d... |
1258 |
} |
837d026d5 mm/compaction: mo... |
1259 1260 |
return COMPACT_NO_SUITABLE_PAGE; } |
ea7ab982b mm, compaction: c... |
1261 1262 1263 |
static enum compact_result compact_finished(struct zone *zone, struct compact_control *cc, const int migratetype) |
837d026d5 mm/compaction: mo... |
1264 1265 1266 1267 1268 1269 1270 1271 1272 |
{ 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 mm: compaction: m... |
1273 |
} |
3e7d34497 mm: vmscan: recla... |
1274 1275 1276 1277 |
/* * compaction_suitable: Is this suitable to run compaction on this zone now? * Returns * COMPACT_SKIPPED - If there are too few free pages for compaction |
cf378319d mm, compaction: r... |
1278 |
* COMPACT_SUCCESS - If the allocation would succeed without compaction |
3e7d34497 mm: vmscan: recla... |
1279 1280 |
* COMPACT_CONTINUE - If compaction should run now */ |
ea7ab982b mm, compaction: c... |
1281 |
static enum compact_result __compaction_suitable(struct zone *zone, int order, |
c603844bd mm, page_alloc: c... |
1282 |
unsigned int alloc_flags, |
86a294a81 mm, oom, compacti... |
1283 1284 |
int classzone_idx, unsigned long wmark_target) |
3e7d34497 mm: vmscan: recla... |
1285 |
{ |
3e7d34497 mm: vmscan: recla... |
1286 |
unsigned long watermark; |
21c527a3c mm/compaction.c: ... |
1287 |
if (is_via_compact_memory(order)) |
3957c7768 mm: compaction: f... |
1288 |
return COMPACT_CONTINUE; |
f2b8228c5 mm, compaction: u... |
1289 |
watermark = zone->watermark[alloc_flags & ALLOC_WMARK_MASK]; |
ebff39801 mm, compaction: p... |
1290 1291 1292 1293 1294 1295 |
/* * 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)) |
cf378319d mm, compaction: r... |
1296 |
return COMPACT_SUCCESS; |
ebff39801 mm, compaction: p... |
1297 |
|
3957c7768 mm: compaction: f... |
1298 |
/* |
9861a62c3 mm, compaction: c... |
1299 |
* Watermarks for order-0 must be met for compaction to be able to |
984fdba6a mm, compaction: u... |
1300 1301 1302 1303 1304 1305 1306 |
* isolate free pages for migration targets. This means that the * watermark and alloc_flags have to match, or be more pessimistic than * the check in __isolate_free_page(). We don't use the direct * compactor's alloc_flags, as they are not relevant for freepage * isolation. We however do use the direct compactor's classzone_idx to * skip over zones where lowmem reserves would prevent allocation even * if compaction succeeds. |
8348faf91 mm, compaction: r... |
1307 1308 |
* For costly orders, we require low watermark instead of min for * compaction to proceed to increase its chances. |
984fdba6a mm, compaction: u... |
1309 1310 |
* ALLOC_CMA is used, as pages in CMA pageblocks are considered * suitable migration targets |
3e7d34497 mm: vmscan: recla... |
1311 |
*/ |
8348faf91 mm, compaction: r... |
1312 1313 1314 |
watermark = (order > PAGE_ALLOC_COSTLY_ORDER) ? low_wmark_pages(zone) : min_wmark_pages(zone); watermark += compact_gap(order); |
86a294a81 mm, oom, compacti... |
1315 |
if (!__zone_watermark_ok(zone, 0, watermark, classzone_idx, |
984fdba6a mm, compaction: u... |
1316 |
ALLOC_CMA, wmark_target)) |
3e7d34497 mm: vmscan: recla... |
1317 |
return COMPACT_SKIPPED; |
cc5c9f098 mm, compaction: i... |
1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 |
return COMPACT_CONTINUE; } enum compact_result compaction_suitable(struct zone *zone, int order, unsigned int alloc_flags, int classzone_idx) { enum compact_result ret; int fragindex; ret = __compaction_suitable(zone, order, alloc_flags, classzone_idx, zone_page_state(zone, NR_FREE_PAGES)); |
3e7d34497 mm: vmscan: recla... |
1330 1331 1332 1333 |
/* * fragmentation index determines if allocation failures are due to * low memory or external fragmentation * |
ebff39801 mm, compaction: p... |
1334 1335 |
* index of -1000 would imply allocations might succeed depending on * watermarks, but we already failed the high-order watermark check |
3e7d34497 mm: vmscan: recla... |
1336 1337 1338 |
* index towards 0 implies failure is due to lack of memory * index towards 1000 implies failure is due to fragmentation * |
203114202 mm, compaction: r... |
1339 1340 1341 1342 1343 1344 |
* Only compact if a failure would be due to fragmentation. Also * ignore fragindex for non-costly orders where the alternative to * a successful reclaim/compaction is OOM. Fragindex and the * vm.extfrag_threshold sysctl is meant as a heuristic to prevent * excessive compaction for costly orders, but it should not be at the * expense of system stability. |
3e7d34497 mm: vmscan: recla... |
1345 |
*/ |
203114202 mm, compaction: r... |
1346 |
if (ret == COMPACT_CONTINUE && (order > PAGE_ALLOC_COSTLY_ORDER)) { |
cc5c9f098 mm, compaction: i... |
1347 1348 1349 1350 |
fragindex = fragmentation_index(zone, order); if (fragindex >= 0 && fragindex <= sysctl_extfrag_threshold) ret = COMPACT_NOT_SUITABLE_ZONE; } |
837d026d5 mm/compaction: mo... |
1351 |
|
837d026d5 mm/compaction: mo... |
1352 1353 1354 1355 1356 1357 |
trace_mm_compaction_suitable(zone, order, ret); if (ret == COMPACT_NOT_SUITABLE_ZONE) ret = COMPACT_SKIPPED; return ret; } |
86a294a81 mm, oom, compacti... |
1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 |
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 mm: remove reclai... |
1379 |
available = zone_reclaimable_pages(zone) / order; |
86a294a81 mm, oom, compacti... |
1380 1381 1382 |
available += zone_page_state_snapshot(zone, NR_FREE_PAGES); compact_result = __compaction_suitable(zone, order, alloc_flags, ac_classzone_idx(ac), available); |
cc5c9f098 mm, compaction: i... |
1383 |
if (compact_result != COMPACT_SKIPPED) |
86a294a81 mm, oom, compacti... |
1384 1385 1386 1387 1388 |
return true; } return false; } |
ea7ab982b mm, compaction: c... |
1389 |
static enum compact_result compact_zone(struct zone *zone, struct compact_control *cc) |
748446bb6 mm: compaction: m... |
1390 |
{ |
ea7ab982b mm, compaction: c... |
1391 |
enum compact_result ret; |
c89511ab2 mm: compaction: R... |
1392 |
unsigned long start_pfn = zone->zone_start_pfn; |
108bcc96e mm: add & use zon... |
1393 |
unsigned long end_pfn = zone_end_pfn(zone); |
6d7ce5594 mm, compaction: p... |
1394 |
const int migratetype = gfpflags_to_migratetype(cc->gfp_mask); |
e0b9daeb4 mm, compaction: e... |
1395 |
const bool sync = cc->mode != MIGRATE_ASYNC; |
748446bb6 mm: compaction: m... |
1396 |
|
ebff39801 mm, compaction: p... |
1397 1398 |
ret = compaction_suitable(zone, cc->order, cc->alloc_flags, cc->classzone_idx); |
c46649dea mm, compaction: c... |
1399 |
/* Compaction is likely to fail */ |
cf378319d mm, compaction: r... |
1400 |
if (ret == COMPACT_SUCCESS || ret == COMPACT_SKIPPED) |
3e7d34497 mm: vmscan: recla... |
1401 |
return ret; |
c46649dea mm, compaction: c... |
1402 1403 1404 |
/* huh, compaction_suitable is returning something unexpected */ VM_BUG_ON(ret != COMPACT_CONTINUE); |
3e7d34497 mm: vmscan: recla... |
1405 |
|
c89511ab2 mm: compaction: R... |
1406 |
/* |
d3132e4b8 mm: compaction: r... |
1407 |
* Clear pageblock skip if there were failures recently and compaction |
accf62422 mm, kswapd: repla... |
1408 |
* is about to be retried after being deferred. |
d3132e4b8 mm: compaction: r... |
1409 |
*/ |
accf62422 mm, kswapd: repla... |
1410 |
if (compaction_restarting(zone, cc->order)) |
d3132e4b8 mm: compaction: r... |
1411 1412 1413 |
__reset_isolation_suitable(zone); /* |
c89511ab2 mm: compaction: R... |
1414 |
* Setup to move all movable pages to the end of the zone. Used cached |
06ed29989 mm, compaction: m... |
1415 1416 1417 |
* information on where the scanners should start (unless we explicitly * want to compact the whole zone), but check that it is initialised * by ensuring the values are within zone boundaries. |
c89511ab2 mm: compaction: R... |
1418 |
*/ |
06ed29989 mm, compaction: m... |
1419 |
if (cc->whole_zone) { |
c89511ab2 mm: compaction: R... |
1420 |
cc->migrate_pfn = start_pfn; |
06ed29989 mm, compaction: m... |
1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 |
cc->free_pfn = pageblock_start_pfn(end_pfn - 1); } else { cc->migrate_pfn = zone->compact_cached_migrate_pfn[sync]; cc->free_pfn = zone->compact_cached_free_pfn; if (cc->free_pfn < start_pfn || cc->free_pfn >= end_pfn) { cc->free_pfn = pageblock_start_pfn(end_pfn - 1); zone->compact_cached_free_pfn = cc->free_pfn; } if (cc->migrate_pfn < start_pfn || cc->migrate_pfn >= end_pfn) { cc->migrate_pfn = start_pfn; zone->compact_cached_migrate_pfn[0] = cc->migrate_pfn; zone->compact_cached_migrate_pfn[1] = cc->migrate_pfn; } |
c8f7de0bf mm, compaction: d... |
1434 |
|
06ed29989 mm, compaction: m... |
1435 1436 1437 |
if (cc->migrate_pfn == start_pfn) cc->whole_zone = true; } |
c8f7de0bf mm, compaction: d... |
1438 |
|
1a16718cf mm/compaction: co... |
1439 |
cc->last_migrated_pfn = 0; |
748446bb6 mm: compaction: m... |
1440 |
|
16c4a097a mm/compaction: en... |
1441 1442 |
trace_mm_compaction_begin(start_pfn, cc->migrate_pfn, cc->free_pfn, end_pfn, sync); |
0eb927c0a mm: compaction: t... |
1443 |
|
748446bb6 mm: compaction: m... |
1444 |
migrate_prep_local(); |
6d7ce5594 mm, compaction: p... |
1445 1446 |
while ((ret = compact_finished(zone, cc, migratetype)) == COMPACT_CONTINUE) { |
9d502c1c8 mm/compaction: ch... |
1447 |
int err; |
748446bb6 mm: compaction: m... |
1448 |
|
f9e35b3b4 mm: compaction: a... |
1449 1450 |
switch (isolate_migratepages(zone, cc)) { case ISOLATE_ABORT: |
2d1e10412 mm, compaction: d... |
1451 |
ret = COMPACT_CONTENDED; |
5733c7d11 mm: introduce put... |
1452 |
putback_movable_pages(&cc->migratepages); |
e64c5237c mm: compaction: a... |
1453 |
cc->nr_migratepages = 0; |
f9e35b3b4 mm: compaction: a... |
1454 1455 |
goto out; case ISOLATE_NONE: |
fdaf7f5c4 mm, compaction: m... |
1456 1457 1458 1459 1460 1461 |
/* * We haven't isolated and migrated anything, but * there might still be unflushed migrations from * previous cc->order aligned block. */ goto check_drain; |
f9e35b3b4 mm: compaction: a... |
1462 1463 1464 |
case ISOLATE_SUCCESS: ; } |
748446bb6 mm: compaction: m... |
1465 |
|
d53aea3d4 mm, compaction: r... |
1466 |
err = migrate_pages(&cc->migratepages, compaction_alloc, |
e0b9daeb4 mm, compaction: e... |
1467 |
compaction_free, (unsigned long)cc, cc->mode, |
7b2a2d4a1 mm: migrate: Add ... |
1468 |
MR_COMPACTION); |
748446bb6 mm: compaction: m... |
1469 |
|
f8c9301fa mm/compaction: do... |
1470 1471 |
trace_mm_compaction_migratepages(cc->nr_migratepages, err, &cc->migratepages); |
748446bb6 mm: compaction: m... |
1472 |
|
f8c9301fa mm/compaction: do... |
1473 1474 |
/* All pages were either migrated or will be released */ cc->nr_migratepages = 0; |
9d502c1c8 mm/compaction: ch... |
1475 |
if (err) { |
5733c7d11 mm: introduce put... |
1476 |
putback_movable_pages(&cc->migratepages); |
7ed695e06 mm: compaction: d... |
1477 1478 1479 1480 |
/* * migrate_pages() may return -ENOMEM when scanners meet * and we want compact_finished() to detect it */ |
f2849aa09 mm, compaction: m... |
1481 |
if (err == -ENOMEM && !compact_scanners_met(cc)) { |
2d1e10412 mm, compaction: d... |
1482 |
ret = COMPACT_CONTENDED; |
4bf2bba37 mm, thp: abort co... |
1483 1484 |
goto out; } |
fdd048e12 mm, compaction: s... |
1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 |
/* * 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 mm: compaction: m... |
1497 |
} |
fdaf7f5c4 mm, compaction: m... |
1498 |
|
fdaf7f5c4 mm, compaction: m... |
1499 1500 1501 1502 1503 1504 1505 1506 |
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 mm/compaction: co... |
1507 |
if (cc->order > 0 && cc->last_migrated_pfn) { |
fdaf7f5c4 mm, compaction: m... |
1508 1509 |
int cpu; unsigned long current_block_start = |
06b6640a3 mm, compaction: w... |
1510 |
block_start_pfn(cc->migrate_pfn, cc->order); |
fdaf7f5c4 mm, compaction: m... |
1511 |
|
1a16718cf mm/compaction: co... |
1512 |
if (cc->last_migrated_pfn < current_block_start) { |
fdaf7f5c4 mm, compaction: m... |
1513 1514 1515 1516 1517 |
cpu = get_cpu(); lru_add_drain_cpu(cpu); drain_local_pages(zone); put_cpu(); /* No more flushing until we migrate again */ |
1a16718cf mm/compaction: co... |
1518 |
cc->last_migrated_pfn = 0; |
fdaf7f5c4 mm, compaction: m... |
1519 1520 |
} } |
748446bb6 mm: compaction: m... |
1521 |
} |
f9e35b3b4 mm: compaction: a... |
1522 |
out: |
6bace090a mm, compaction: a... |
1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 |
/* * 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 mm, compaction: w... |
1533 |
free_pfn = pageblock_start_pfn(free_pfn); |
6bace090a mm, compaction: a... |
1534 1535 1536 1537 1538 1539 1540 |
/* * 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 mm: compaction: m... |
1541 |
|
16c4a097a mm/compaction: en... |
1542 1543 |
trace_mm_compaction_end(start_pfn, cc->migrate_pfn, cc->free_pfn, end_pfn, sync, ret); |
0eb927c0a mm: compaction: t... |
1544 |
|
748446bb6 mm: compaction: m... |
1545 1546 |
return ret; } |
76ab0f530 mm: compaction: a... |
1547 |
|
ea7ab982b mm, compaction: c... |
1548 |
static enum compact_result compact_zone_order(struct zone *zone, int order, |
c3486f537 mm, compaction: s... |
1549 |
gfp_t gfp_mask, enum compact_priority prio, |
c603844bd mm, page_alloc: c... |
1550 |
unsigned int alloc_flags, int classzone_idx) |
56de7263f mm: compaction: d... |
1551 |
{ |
ea7ab982b mm, compaction: c... |
1552 |
enum compact_result ret; |
56de7263f mm: compaction: d... |
1553 1554 1555 1556 |
struct compact_control cc = { .nr_freepages = 0, .nr_migratepages = 0, .order = order, |
6d7ce5594 mm, compaction: p... |
1557 |
.gfp_mask = gfp_mask, |
56de7263f mm: compaction: d... |
1558 |
.zone = zone, |
a5508cd83 mm, compaction: i... |
1559 1560 |
.mode = (prio == COMPACT_PRIO_ASYNC) ? MIGRATE_ASYNC : MIGRATE_SYNC_LIGHT, |
ebff39801 mm, compaction: p... |
1561 1562 |
.alloc_flags = alloc_flags, .classzone_idx = classzone_idx, |
accf62422 mm, kswapd: repla... |
1563 |
.direct_compaction = true, |
a8e025e55 mm, compaction: a... |
1564 |
.whole_zone = (prio == MIN_COMPACT_PRIORITY), |
9f7e33879 mm, compaction: m... |
1565 1566 |
.ignore_skip_hint = (prio == MIN_COMPACT_PRIORITY), .ignore_block_suitable = (prio == MIN_COMPACT_PRIORITY) |
56de7263f mm: compaction: d... |
1567 1568 1569 |
}; INIT_LIST_HEAD(&cc.freepages); INIT_LIST_HEAD(&cc.migratepages); |
e64c5237c mm: compaction: a... |
1570 1571 1572 1573 |
ret = compact_zone(zone, &cc); VM_BUG_ON(!list_empty(&cc.freepages)); VM_BUG_ON(!list_empty(&cc.migratepages)); |
e64c5237c mm: compaction: a... |
1574 |
return ret; |
56de7263f mm: compaction: d... |
1575 |
} |
5e7719058 mm: compaction: a... |
1576 |
int sysctl_extfrag_threshold = 500; |
56de7263f mm: compaction: d... |
1577 1578 |
/** * try_to_compact_pages - Direct compact to satisfy a high-order allocation |
56de7263f mm: compaction: d... |
1579 |
* @gfp_mask: The GFP mask of the current allocation |
1a6d53a10 mm: reduce try_to... |
1580 1581 1582 |
* @order: The order of the current allocation * @alloc_flags: The allocation flags of the current allocation * @ac: The context of current allocation |
e0b9daeb4 mm, compaction: e... |
1583 |
* @mode: The migration mode for async, sync light, or sync migration |
56de7263f mm: compaction: d... |
1584 1585 1586 |
* * This is the main entry point for direct page compaction. */ |
ea7ab982b mm, compaction: c... |
1587 |
enum compact_result try_to_compact_pages(gfp_t gfp_mask, unsigned int order, |
c603844bd mm, page_alloc: c... |
1588 |
unsigned int alloc_flags, const struct alloc_context *ac, |
c3486f537 mm, compaction: s... |
1589 |
enum compact_priority prio) |
56de7263f mm: compaction: d... |
1590 |
{ |
56de7263f mm: compaction: d... |
1591 1592 |
int may_enter_fs = gfp_mask & __GFP_FS; int may_perform_io = gfp_mask & __GFP_IO; |
56de7263f mm: compaction: d... |
1593 1594 |
struct zoneref *z; struct zone *zone; |
1d4746d39 mm, compaction: d... |
1595 |
enum compact_result rc = COMPACT_SKIPPED; |
56de7263f mm: compaction: d... |
1596 |
|
4ffb6335d mm: compaction: u... |
1597 |
/* Check if the GFP flags allow compaction */ |
b2b331f96 mm/compaction: re... |
1598 |
if (!may_enter_fs || !may_perform_io) |
53853e2d2 mm, compaction: d... |
1599 |
return COMPACT_SKIPPED; |
56de7263f mm: compaction: d... |
1600 |
|
a5508cd83 mm, compaction: i... |
1601 |
trace_mm_compaction_try_to_compact_pages(order, gfp_mask, prio); |
837d026d5 mm/compaction: mo... |
1602 |
|
56de7263f mm: compaction: d... |
1603 |
/* Compact each zone in the list */ |
1a6d53a10 mm: reduce try_to... |
1604 1605 |
for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, ac->high_zoneidx, ac->nodemask) { |
ea7ab982b mm, compaction: c... |
1606 |
enum compact_result status; |
56de7263f mm: compaction: d... |
1607 |
|
a8e025e55 mm, compaction: a... |
1608 1609 |
if (prio > MIN_COMPACT_PRIORITY && compaction_deferred(zone, order)) { |
1d4746d39 mm, compaction: d... |
1610 |
rc = max_t(enum compact_result, COMPACT_DEFERRED, rc); |
53853e2d2 mm, compaction: d... |
1611 |
continue; |
1d4746d39 mm, compaction: d... |
1612 |
} |
53853e2d2 mm, compaction: d... |
1613 |
|
a5508cd83 mm, compaction: i... |
1614 |
status = compact_zone_order(zone, order, gfp_mask, prio, |
c3486f537 mm, compaction: s... |
1615 |
alloc_flags, ac_classzone_idx(ac)); |
56de7263f mm: compaction: d... |
1616 |
rc = max(status, rc); |
7ceb009a2 mm, compaction: d... |
1617 1618 |
/* The allocation should succeed, stop compacting */ if (status == COMPACT_SUCCESS) { |
53853e2d2 mm, compaction: d... |
1619 1620 1621 1622 1623 1624 1625 |
/* * 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 mm, compaction: k... |
1626 |
|
c3486f537 mm, compaction: s... |
1627 |
break; |
1f9efdef4 mm, compaction: k... |
1628 |
} |
a5508cd83 mm, compaction: i... |
1629 |
if (prio != COMPACT_PRIO_ASYNC && (status == COMPACT_COMPLETE || |
c3486f537 mm, compaction: s... |
1630 |
status == COMPACT_PARTIAL_SKIPPED)) |
53853e2d2 mm, compaction: d... |
1631 1632 1633 1634 1635 1636 |
/* * 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 mm, compaction: k... |
1637 1638 1639 1640 |
/* * We might have stopped compacting due to need_resched() in * async compaction, or due to a fatal signal detected. In that |
c3486f537 mm, compaction: s... |
1641 |
* case do not try further zones |
1f9efdef4 mm, compaction: k... |
1642 |
*/ |
c3486f537 mm, compaction: s... |
1643 1644 1645 |
if ((prio == COMPACT_PRIO_ASYNC && need_resched()) || fatal_signal_pending(current)) break; |
56de7263f mm: compaction: d... |
1646 1647 1648 1649 |
} return rc; } |
76ab0f530 mm: compaction: a... |
1650 |
/* Compact all zones within a node */ |
791cae962 mm, compaction: c... |
1651 |
static void compact_node(int nid) |
76ab0f530 mm: compaction: a... |
1652 |
{ |
791cae962 mm, compaction: c... |
1653 |
pg_data_t *pgdat = NODE_DATA(nid); |
76ab0f530 mm: compaction: a... |
1654 |
int zoneid; |
76ab0f530 mm: compaction: a... |
1655 |
struct zone *zone; |
791cae962 mm, compaction: c... |
1656 1657 1658 1659 1660 1661 |
struct compact_control cc = { .order = -1, .mode = MIGRATE_SYNC, .ignore_skip_hint = true, .whole_zone = true, }; |
76ab0f530 mm: compaction: a... |
1662 |
|
76ab0f530 mm: compaction: a... |
1663 |
for (zoneid = 0; zoneid < MAX_NR_ZONES; zoneid++) { |
76ab0f530 mm: compaction: a... |
1664 1665 1666 1667 |
zone = &pgdat->node_zones[zoneid]; if (!populated_zone(zone)) continue; |
791cae962 mm, compaction: c... |
1668 1669 1670 1671 1672 |
cc.nr_freepages = 0; cc.nr_migratepages = 0; cc.zone = zone; INIT_LIST_HEAD(&cc.freepages); INIT_LIST_HEAD(&cc.migratepages); |
76ab0f530 mm: compaction: a... |
1673 |
|
791cae962 mm, compaction: c... |
1674 |
compact_zone(zone, &cc); |
754693457 mm/compaction.c: ... |
1675 |
|
791cae962 mm, compaction: c... |
1676 1677 |
VM_BUG_ON(!list_empty(&cc.freepages)); VM_BUG_ON(!list_empty(&cc.migratepages)); |
76ab0f530 mm: compaction: a... |
1678 |
} |
76ab0f530 mm: compaction: a... |
1679 1680 1681 |
} /* Compact all nodes in the system */ |
7964c06d6 mm: compaction: f... |
1682 |
static void compact_nodes(void) |
76ab0f530 mm: compaction: a... |
1683 1684 |
{ int nid; |
8575ec29f compact_pgdat: wo... |
1685 1686 |
/* Flush pending updates to the LRU lists */ lru_add_drain_all(); |
76ab0f530 mm: compaction: a... |
1687 1688 |
for_each_online_node(nid) compact_node(nid); |
76ab0f530 mm: compaction: a... |
1689 1690 1691 1692 |
} /* The written value is actually unused, all memory is compacted */ int sysctl_compact_memory; |
fec4eb2c8 mm/compaction: im... |
1693 1694 1695 1696 |
/* * This is the entry point for compacting all nodes via * /proc/sys/vm/compact_memory */ |
76ab0f530 mm: compaction: a... |
1697 1698 1699 1700 |
int sysctl_compaction_handler(struct ctl_table *table, int write, void __user *buffer, size_t *length, loff_t *ppos) { if (write) |
7964c06d6 mm: compaction: f... |
1701 |
compact_nodes(); |
76ab0f530 mm: compaction: a... |
1702 1703 1704 |
return 0; } |
ed4a6d7f0 mm: compaction: a... |
1705 |
|
5e7719058 mm: compaction: a... |
1706 1707 1708 1709 1710 1711 1712 |
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 mm: compaction: a... |
1713 |
#if defined(CONFIG_SYSFS) && defined(CONFIG_NUMA) |
74e77fb9a mm/compaction.c: ... |
1714 |
static ssize_t sysfs_compact_node(struct device *dev, |
10fbcf4c6 convert 'memory' ... |
1715 |
struct device_attribute *attr, |
ed4a6d7f0 mm: compaction: a... |
1716 1717 |
const char *buf, size_t count) { |
8575ec29f compact_pgdat: wo... |
1718 1719 1720 1721 1722 1723 1724 1725 |
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 mm: compaction: a... |
1726 1727 1728 |
return count; } |
10fbcf4c6 convert 'memory' ... |
1729 |
static DEVICE_ATTR(compact, S_IWUSR, NULL, sysfs_compact_node); |
ed4a6d7f0 mm: compaction: a... |
1730 1731 1732 |
int compaction_register_node(struct node *node) { |
10fbcf4c6 convert 'memory' ... |
1733 |
return device_create_file(&node->dev, &dev_attr_compact); |
ed4a6d7f0 mm: compaction: a... |
1734 1735 1736 1737 |
} void compaction_unregister_node(struct node *node) { |
10fbcf4c6 convert 'memory' ... |
1738 |
return device_remove_file(&node->dev, &dev_attr_compact); |
ed4a6d7f0 mm: compaction: a... |
1739 1740 |
} #endif /* CONFIG_SYSFS && CONFIG_NUMA */ |
ff9543fd3 mm: compaction: e... |
1741 |
|
698b1b306 mm, compaction: i... |
1742 1743 |
static inline bool kcompactd_work_requested(pg_data_t *pgdat) { |
172400c69 mm: fix kcompactd... |
1744 |
return pgdat->kcompactd_max_order > 0 || kthread_should_stop(); |
698b1b306 mm, compaction: i... |
1745 1746 1747 1748 1749 1750 1751 |
} static bool kcompactd_node_suitable(pg_data_t *pgdat) { int zoneid; struct zone *zone; enum zone_type classzone_idx = pgdat->kcompactd_classzone_idx; |
6cd9dc3e7 mm/compaction.c: ... |
1752 |
for (zoneid = 0; zoneid <= classzone_idx; zoneid++) { |
698b1b306 mm, compaction: i... |
1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 |
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, }; |
698b1b306 mm, compaction: i... |
1781 1782 1783 |
trace_mm_compaction_kcompactd_wake(pgdat->node_id, cc.order, cc.classzone_idx); count_vm_event(KCOMPACTD_WAKE); |
6cd9dc3e7 mm/compaction.c: ... |
1784 |
for (zoneid = 0; zoneid <= cc.classzone_idx; zoneid++) { |
698b1b306 mm, compaction: i... |
1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 |
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 mm: fix kcompactd... |
1803 1804 |
if (kthread_should_stop()) return; |
698b1b306 mm, compaction: i... |
1805 |
status = compact_zone(zone, &cc); |
7ceb009a2 mm, compaction: d... |
1806 |
if (status == COMPACT_SUCCESS) { |
698b1b306 mm, compaction: i... |
1807 |
compaction_defer_reset(zone, cc.order, false); |
c8f7de0bf mm, compaction: d... |
1808 |
} else if (status == COMPACT_PARTIAL_SKIPPED || status == COMPACT_COMPLETE) { |
698b1b306 mm, compaction: i... |
1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 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 |
/* * 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 mm: compaction: e... |
1955 |
#endif /* CONFIG_COMPACTION */ |