Blame view
mm/migrate.c
26.5 KB
b20a35035
|
1 2 3 4 5 6 7 8 9 10 11 |
/* * Memory Migration functionality - linux/mm/migration.c * * Copyright (C) 2006 Silicon Graphics, Inc., Christoph Lameter * * Page migration was first developed in the context of the memory hotplug * project. The main authors of the migration code are: * * IWAMOTO Toshihiro <iwamoto@valinux.co.jp> * Hirokazu Takahashi <taka@valinux.co.jp> * Dave Hansen <haveblue@us.ibm.com> |
cde535359
|
12 |
* Christoph Lameter |
b20a35035
|
13 14 15 16 17 |
*/ #include <linux/migrate.h> #include <linux/module.h> #include <linux/swap.h> |
0697212a4
|
18 |
#include <linux/swapops.h> |
b20a35035
|
19 |
#include <linux/pagemap.h> |
e23ca00bf
|
20 |
#include <linux/buffer_head.h> |
b20a35035
|
21 |
#include <linux/mm_inline.h> |
b488893a3
|
22 |
#include <linux/nsproxy.h> |
b20a35035
|
23 24 25 26 27 |
#include <linux/pagevec.h> #include <linux/rmap.h> #include <linux/topology.h> #include <linux/cpu.h> #include <linux/cpuset.h> |
04e62a29b
|
28 |
#include <linux/writeback.h> |
742755a1d
|
29 30 |
#include <linux/mempolicy.h> #include <linux/vmalloc.h> |
86c3a7645
|
31 |
#include <linux/security.h> |
8a9f3ccd2
|
32 |
#include <linux/memcontrol.h> |
4f5ca2657
|
33 |
#include <linux/syscalls.h> |
b20a35035
|
34 35 |
#include "internal.h" |
b20a35035
|
36 37 38 |
#define lru_to_page(_head) (list_entry((_head)->prev, struct page, lru)) /* |
742755a1d
|
39 40 |
* migrate_prep() needs to be called before we start compiling a list of pages * to be migrated using isolate_lru_page(). |
b20a35035
|
41 42 43 |
*/ int migrate_prep(void) { |
b20a35035
|
44 45 46 47 48 49 50 51 52 53 |
/* * Clear the LRU lists so pages can be isolated. * Note that pages may be moved off the LRU after we have * drained them. Those pages will fail to migrate like other * pages that may be busy. */ lru_add_drain_all(); return 0; } |
b20a35035
|
54 |
/* |
894bc3104
|
55 56 |
* Add isolated pages on the list back to the LRU under page lock * to avoid leaking evictable pages back onto unevictable list. |
b20a35035
|
57 58 59 60 61 62 63 64 65 66 |
* * returns the number of pages put back. */ int putback_lru_pages(struct list_head *l) { struct page *page; struct page *page2; int count = 0; list_for_each_entry_safe(page, page2, l, lru) { |
e24f0b8f7
|
67 |
list_del(&page->lru); |
894bc3104
|
68 |
putback_lru_page(page); |
b20a35035
|
69 70 71 72 |
count++; } return count; } |
0697212a4
|
73 74 75 |
/* * Restore a potential migration pte to a working pte entry */ |
04e62a29b
|
76 |
static void remove_migration_pte(struct vm_area_struct *vma, |
0697212a4
|
77 78 79 80 81 82 83 84 85 |
struct page *old, struct page *new) { struct mm_struct *mm = vma->vm_mm; swp_entry_t entry; pgd_t *pgd; pud_t *pud; pmd_t *pmd; pte_t *ptep, pte; spinlock_t *ptl; |
04e62a29b
|
86 87 88 89 |
unsigned long addr = page_address_in_vma(new, vma); if (addr == -EFAULT) return; |
0697212a4
|
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 |
pgd = pgd_offset(mm, addr); if (!pgd_present(*pgd)) return; pud = pud_offset(pgd, addr); if (!pud_present(*pud)) return; pmd = pmd_offset(pud, addr); if (!pmd_present(*pmd)) return; ptep = pte_offset_map(pmd, addr); if (!is_swap_pte(*ptep)) { pte_unmap(ptep); return; } ptl = pte_lockptr(mm, pmd); spin_lock(ptl); pte = *ptep; if (!is_swap_pte(pte)) goto out; entry = pte_to_swp_entry(pte); if (!is_migration_entry(entry) || migration_entry_to_page(entry) != old) goto out; |
98837c7f8
|
120 121 122 123 124 125 126 127 128 129 130 131 132 |
/* * Yes, ignore the return value from a GFP_ATOMIC mem_cgroup_charge. * Failure is not an option here: we're now expected to remove every * migration pte, and will cause crashes otherwise. Normally this * is not an issue: mem_cgroup_prepare_migration bumped up the old * page_cgroup count for safety, that's now attached to the new page, * so this charge should just be another incrementation of the count, * to keep in balance with rmap.c's mem_cgroup_uncharging. But if * there's been a force_empty, those reference counts may no longer * be reliable, and this charge can actually fail: oh well, we don't * make the situation any worse by proceeding as if it had succeeded. */ mem_cgroup_charge(new, mm, GFP_ATOMIC); |
0697212a4
|
133 134 135 136 |
get_page(new); pte = pte_mkold(mk_pte(new, vma->vm_page_prot)); if (is_write_migration_entry(entry)) pte = pte_mkwrite(pte); |
97ee05246
|
137 |
flush_cache_page(vma, addr, pte_pfn(pte)); |
0697212a4
|
138 |
set_pte_at(mm, addr, ptep, pte); |
04e62a29b
|
139 140 141 142 143 144 145 146 |
if (PageAnon(new)) page_add_anon_rmap(new, vma, addr); else page_add_file_rmap(new); /* No need to invalidate - it was non-present before */ update_mmu_cache(vma, addr, pte); |
04e62a29b
|
147 |
|
0697212a4
|
148 149 150 151 152 |
out: pte_unmap_unlock(ptep, ptl); } /* |
04e62a29b
|
153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 |
* Note that remove_file_migration_ptes will only work on regular mappings, * Nonlinear mappings do not use migration entries. */ static void remove_file_migration_ptes(struct page *old, struct page *new) { struct vm_area_struct *vma; struct address_space *mapping = page_mapping(new); struct prio_tree_iter iter; pgoff_t pgoff = new->index << (PAGE_CACHE_SHIFT - PAGE_SHIFT); if (!mapping) return; spin_lock(&mapping->i_mmap_lock); vma_prio_tree_foreach(vma, &iter, &mapping->i_mmap, pgoff, pgoff) remove_migration_pte(vma, old, new); spin_unlock(&mapping->i_mmap_lock); } /* |
0697212a4
|
175 176 177 |
* Must hold mmap_sem lock on at least one of the vmas containing * the page so that the anon_vma cannot vanish. */ |
04e62a29b
|
178 |
static void remove_anon_migration_ptes(struct page *old, struct page *new) |
0697212a4
|
179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 |
{ struct anon_vma *anon_vma; struct vm_area_struct *vma; unsigned long mapping; mapping = (unsigned long)new->mapping; if (!mapping || (mapping & PAGE_MAPPING_ANON) == 0) return; /* * We hold the mmap_sem lock. So no need to call page_lock_anon_vma. */ anon_vma = (struct anon_vma *) (mapping - PAGE_MAPPING_ANON); spin_lock(&anon_vma->lock); list_for_each_entry(vma, &anon_vma->head, anon_vma_node) |
04e62a29b
|
196 |
remove_migration_pte(vma, old, new); |
0697212a4
|
197 198 199 200 201 |
spin_unlock(&anon_vma->lock); } /* |
04e62a29b
|
202 203 204 205 206 207 208 209 210 211 212 213 |
* Get rid of all migration entries and replace them by * references to the indicated page. */ static void remove_migration_ptes(struct page *old, struct page *new) { if (PageAnon(new)) remove_anon_migration_ptes(old, new); else remove_file_migration_ptes(old, new); } /* |
0697212a4
|
214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 |
* Something used the pte of a page under migration. We need to * get to the page and wait until migration is finished. * When we return from this function the fault will be retried. * * This function is called from do_swap_page(). */ void migration_entry_wait(struct mm_struct *mm, pmd_t *pmd, unsigned long address) { pte_t *ptep, pte; spinlock_t *ptl; swp_entry_t entry; struct page *page; ptep = pte_offset_map_lock(mm, pmd, address, &ptl); pte = *ptep; if (!is_swap_pte(pte)) goto out; entry = pte_to_swp_entry(pte); if (!is_migration_entry(entry)) goto out; page = migration_entry_to_page(entry); |
e286781d5
|
238 239 240 241 242 243 244 245 246 |
/* * Once radix-tree replacement of page migration started, page_count * *must* be zero. And, we don't want to call wait_on_page_locked() * against a page without get_page(). * So, we use get_page_unless_zero(), here. Even failed, page fault * will occur again. */ if (!get_page_unless_zero(page)) goto out; |
0697212a4
|
247 248 249 250 251 252 253 |
pte_unmap_unlock(ptep, ptl); wait_on_page_locked(page); put_page(page); return; out: pte_unmap_unlock(ptep, ptl); } |
b20a35035
|
254 |
/* |
c3fcf8a5d
|
255 |
* Replace the page in the mapping. |
5b5c7120e
|
256 257 258 259 260 |
* * The number of remaining references must be: * 1 for anonymous pages without a mapping * 2 for pages with a mapping * 3 for pages with a mapping and PagePrivate set. |
b20a35035
|
261 |
*/ |
2d1db3b11
|
262 263 |
static int migrate_page_move_mapping(struct address_space *mapping, struct page *newpage, struct page *page) |
b20a35035
|
264 |
{ |
e286781d5
|
265 |
int expected_count; |
7cf9c2c76
|
266 |
void **pslot; |
b20a35035
|
267 |
|
6c5240ae7
|
268 |
if (!mapping) { |
0e8c7d0fd
|
269 |
/* Anonymous page without mapping */ |
6c5240ae7
|
270 271 272 273 |
if (page_count(page) != 1) return -EAGAIN; return 0; } |
19fd62312
|
274 |
spin_lock_irq(&mapping->tree_lock); |
b20a35035
|
275 |
|
7cf9c2c76
|
276 277 |
pslot = radix_tree_lookup_slot(&mapping->page_tree, page_index(page)); |
b20a35035
|
278 |
|
e286781d5
|
279 280 |
expected_count = 2 + !!PagePrivate(page); if (page_count(page) != expected_count || |
7cf9c2c76
|
281 |
(struct page *)radix_tree_deref_slot(pslot) != page) { |
19fd62312
|
282 |
spin_unlock_irq(&mapping->tree_lock); |
e23ca00bf
|
283 |
return -EAGAIN; |
b20a35035
|
284 |
} |
e286781d5
|
285 |
if (!page_freeze_refs(page, expected_count)) { |
19fd62312
|
286 |
spin_unlock_irq(&mapping->tree_lock); |
e286781d5
|
287 288 |
return -EAGAIN; } |
b20a35035
|
289 290 |
/* * Now we know that no one else is looking at the page. |
b20a35035
|
291 |
*/ |
7cf9c2c76
|
292 |
get_page(newpage); /* add cache reference */ |
b20a35035
|
293 294 295 296 |
if (PageSwapCache(page)) { SetPageSwapCache(newpage); set_page_private(newpage, page_private(page)); } |
7cf9c2c76
|
297 |
radix_tree_replace_slot(pslot, newpage); |
e286781d5
|
298 |
page_unfreeze_refs(page, expected_count); |
7cf9c2c76
|
299 300 301 302 |
/* * Drop cache reference from old page. * We know this isn't the last reference. */ |
b20a35035
|
303 |
__put_page(page); |
7cf9c2c76
|
304 |
|
0e8c7d0fd
|
305 306 307 308 309 310 311 312 313 314 315 316 |
/* * If moved to a different zone then also account * the page for that zone. Other VM counters will be * taken care of when we establish references to the * new page and drop references to the old page. * * Note that anonymous pages are accounted for * via NR_FILE_PAGES and NR_ANON_PAGES if they * are mapped to swap space. */ __dec_zone_page_state(page, NR_FILE_PAGES); __inc_zone_page_state(newpage, NR_FILE_PAGES); |
19fd62312
|
317 |
spin_unlock_irq(&mapping->tree_lock); |
b20a35035
|
318 319 320 |
return 0; } |
b20a35035
|
321 322 323 324 |
/* * Copy the page to its new location */ |
e7340f733
|
325 |
static void migrate_page_copy(struct page *newpage, struct page *page) |
b20a35035
|
326 |
{ |
b7abea963
|
327 |
int anon; |
b20a35035
|
328 329 330 331 332 333 334 335 |
copy_highpage(newpage, page); if (PageError(page)) SetPageError(newpage); if (PageReferenced(page)) SetPageReferenced(newpage); if (PageUptodate(page)) SetPageUptodate(newpage); |
894bc3104
|
336 337 |
if (TestClearPageActive(page)) { VM_BUG_ON(PageUnevictable(page)); |
b20a35035
|
338 |
SetPageActive(newpage); |
894bc3104
|
339 340 |
} else unevictable_migrate_page(newpage, page); |
b20a35035
|
341 342 343 344 345 346 347 |
if (PageChecked(page)) SetPageChecked(newpage); if (PageMappedToDisk(page)) SetPageMappedToDisk(newpage); if (PageDirty(page)) { clear_page_dirty_for_io(page); |
3a902c5f6
|
348 349 350 351 352 353 354 355 |
/* * Want to mark the page and the radix tree as dirty, and * redo the accounting that clear_page_dirty_for_io undid, * but we can't use set_page_dirty because that function * is actually a signal that all of the page has become dirty. * Wheras only part of our page may be dirty. */ __set_page_dirty_nobuffers(newpage); |
b20a35035
|
356 |
} |
b291f0003
|
357 |
mlock_migrate_page(newpage, page); |
b20a35035
|
358 |
ClearPageSwapCache(page); |
b20a35035
|
359 360 |
ClearPagePrivate(page); set_page_private(page, 0); |
b7abea963
|
361 362 |
/* page->mapping contains a flag for PageAnon() */ anon = PageAnon(page); |
b20a35035
|
363 |
page->mapping = NULL; |
b7abea963
|
364 365 |
if (!anon) /* This page was removed from radix-tree. */ mem_cgroup_uncharge_cache_page(page); |
b20a35035
|
366 367 368 369 370 371 372 |
/* * If any waiters have accumulated on the new page then * wake them up. */ if (PageWriteback(newpage)) end_page_writeback(newpage); } |
b20a35035
|
373 |
|
1d8b85ccf
|
374 375 376 377 378 |
/************************************************************ * Migration functions ***********************************************************/ /* Always fail migration. Used for mappings that are not movable */ |
2d1db3b11
|
379 380 |
int fail_migrate_page(struct address_space *mapping, struct page *newpage, struct page *page) |
1d8b85ccf
|
381 382 383 384 |
{ return -EIO; } EXPORT_SYMBOL(fail_migrate_page); |
b20a35035
|
385 386 387 388 389 390 |
/* * Common logic to directly migrate a single page suitable for * pages that do not use PagePrivate. * * Pages are locked upon entry and exit. */ |
2d1db3b11
|
391 392 |
int migrate_page(struct address_space *mapping, struct page *newpage, struct page *page) |
b20a35035
|
393 394 395 396 |
{ int rc; BUG_ON(PageWriteback(page)); /* Writeback must be complete */ |
2d1db3b11
|
397 |
rc = migrate_page_move_mapping(mapping, newpage, page); |
b20a35035
|
398 399 400 401 402 |
if (rc) return rc; migrate_page_copy(newpage, page); |
b20a35035
|
403 404 405 |
return 0; } EXPORT_SYMBOL(migrate_page); |
9361401eb
|
406 |
#ifdef CONFIG_BLOCK |
b20a35035
|
407 |
/* |
1d8b85ccf
|
408 409 410 411 |
* Migration function for pages with buffers. This function can only be used * if the underlying filesystem guarantees that no other references to "page" * exist. */ |
2d1db3b11
|
412 413 |
int buffer_migrate_page(struct address_space *mapping, struct page *newpage, struct page *page) |
1d8b85ccf
|
414 |
{ |
1d8b85ccf
|
415 416 |
struct buffer_head *bh, *head; int rc; |
1d8b85ccf
|
417 |
if (!page_has_buffers(page)) |
2d1db3b11
|
418 |
return migrate_page(mapping, newpage, page); |
1d8b85ccf
|
419 420 |
head = page_buffers(page); |
2d1db3b11
|
421 |
rc = migrate_page_move_mapping(mapping, newpage, page); |
1d8b85ccf
|
422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 |
if (rc) return rc; bh = head; do { get_bh(bh); lock_buffer(bh); bh = bh->b_this_page; } while (bh != head); ClearPagePrivate(page); set_page_private(newpage, page_private(page)); set_page_private(page, 0); put_page(page); get_page(newpage); bh = head; do { set_bh_page(bh, newpage, bh_offset(bh)); bh = bh->b_this_page; } while (bh != head); SetPagePrivate(newpage); migrate_page_copy(newpage, page); bh = head; do { unlock_buffer(bh); put_bh(bh); bh = bh->b_this_page; } while (bh != head); return 0; } EXPORT_SYMBOL(buffer_migrate_page); |
9361401eb
|
462 |
#endif |
1d8b85ccf
|
463 |
|
04e62a29b
|
464 465 466 467 |
/* * Writeback a page to clean the dirty state */ static int writeout(struct address_space *mapping, struct page *page) |
8351a6e47
|
468 |
{ |
04e62a29b
|
469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 |
struct writeback_control wbc = { .sync_mode = WB_SYNC_NONE, .nr_to_write = 1, .range_start = 0, .range_end = LLONG_MAX, .nonblocking = 1, .for_reclaim = 1 }; int rc; if (!mapping->a_ops->writepage) /* No write method for the address space */ return -EINVAL; if (!clear_page_dirty_for_io(page)) /* Someone else already triggered a write */ return -EAGAIN; |
8351a6e47
|
486 |
/* |
04e62a29b
|
487 488 489 490 491 492 |
* A dirty page may imply that the underlying filesystem has * the page on some queue. So the page must be clean for * migration. Writeout may mean we loose the lock and the * page state is no longer what we checked for earlier. * At this point we know that the migration attempt cannot * be successful. |
8351a6e47
|
493 |
*/ |
04e62a29b
|
494 |
remove_migration_ptes(page, page); |
8351a6e47
|
495 |
|
04e62a29b
|
496 |
rc = mapping->a_ops->writepage(page, &wbc); |
8351a6e47
|
497 |
|
04e62a29b
|
498 499 500 |
if (rc != AOP_WRITEPAGE_ACTIVATE) /* unlocked. Relock */ lock_page(page); |
bda8550de
|
501 |
return (rc < 0) ? -EIO : -EAGAIN; |
04e62a29b
|
502 503 504 505 506 507 508 509 510 511 |
} /* * Default handling if a filesystem does not provide a migration function. */ static int fallback_migrate_page(struct address_space *mapping, struct page *newpage, struct page *page) { if (PageDirty(page)) return writeout(mapping, page); |
8351a6e47
|
512 513 514 515 516 |
/* * Buffers may be managed in a filesystem specific way. * We must have no buffers or drop them. */ |
b398f6bff
|
517 |
if (PagePrivate(page) && |
8351a6e47
|
518 519 520 521 522 |
!try_to_release_page(page, GFP_KERNEL)) return -EAGAIN; return migrate_page(mapping, newpage, page); } |
1d8b85ccf
|
523 |
/* |
e24f0b8f7
|
524 525 526 527 528 |
* Move a page to a newly allocated page * The page is locked and all ptes have been successfully removed. * * The new page will have replaced the old page if this function * is successful. |
894bc3104
|
529 530 531 532 |
* * Return value: * < 0 - error code * == 0 - success |
e24f0b8f7
|
533 534 535 536 537 538 539 540 541 542 543 |
*/ static int move_to_new_page(struct page *newpage, struct page *page) { struct address_space *mapping; int rc; /* * Block others from accessing the page when we get around to * establishing additional references. We are the only one * holding a reference to the new page at this point. */ |
529ae9aaa
|
544 |
if (!trylock_page(newpage)) |
e24f0b8f7
|
545 546 547 548 549 |
BUG(); /* Prepare mapping for the new page.*/ newpage->index = page->index; newpage->mapping = page->mapping; |
b2e185384
|
550 551 |
if (PageSwapBacked(page)) SetPageSwapBacked(newpage); |
e24f0b8f7
|
552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 |
mapping = page_mapping(page); if (!mapping) rc = migrate_page(mapping, newpage, page); else if (mapping->a_ops->migratepage) /* * Most pages have a mapping and most filesystems * should provide a migration function. Anonymous * pages are part of swap space which also has its * own migration function. This is the most common * path for page migration. */ rc = mapping->a_ops->migratepage(mapping, newpage, page); else rc = fallback_migrate_page(mapping, newpage, page); |
ae41be374
|
568 |
if (!rc) { |
e24f0b8f7
|
569 |
remove_migration_ptes(page, newpage); |
ae41be374
|
570 |
} else |
e24f0b8f7
|
571 572 573 574 575 576 577 578 579 580 581 |
newpage->mapping = NULL; unlock_page(newpage); return rc; } /* * Obtain the lock on page, remove all ptes and migrate the page * to the newly allocated page in newpage. */ |
95a402c38
|
582 583 |
static int unmap_and_move(new_page_t get_new_page, unsigned long private, struct page *page, int force) |
e24f0b8f7
|
584 585 |
{ int rc = 0; |
742755a1d
|
586 587 |
int *result = NULL; struct page *newpage = get_new_page(page, private, &result); |
989f89c57
|
588 |
int rcu_locked = 0; |
ae41be374
|
589 |
int charge = 0; |
95a402c38
|
590 591 592 |
if (!newpage) return -ENOMEM; |
e24f0b8f7
|
593 |
|
894bc3104
|
594 |
if (page_count(page) == 1) { |
e24f0b8f7
|
595 |
/* page was freed from under us. So we are done. */ |
95a402c38
|
596 |
goto move_newpage; |
894bc3104
|
597 |
} |
e24f0b8f7
|
598 |
|
e8589cc18
|
599 600 601 602 603 604 605 |
charge = mem_cgroup_prepare_migration(page, newpage); if (charge == -ENOMEM) { rc = -ENOMEM; goto move_newpage; } /* prepare cgroup just returns 0 or -ENOMEM */ BUG_ON(charge); |
e24f0b8f7
|
606 |
rc = -EAGAIN; |
529ae9aaa
|
607 |
if (!trylock_page(page)) { |
e24f0b8f7
|
608 |
if (!force) |
95a402c38
|
609 |
goto move_newpage; |
e24f0b8f7
|
610 611 612 613 614 615 616 617 |
lock_page(page); } if (PageWriteback(page)) { if (!force) goto unlock; wait_on_page_writeback(page); } |
e24f0b8f7
|
618 |
/* |
dc386d4d1
|
619 620 621 622 |
* By try_to_unmap(), page->mapcount goes down to 0 here. In this case, * we cannot notice that anon_vma is freed while we migrates a page. * This rcu_read_lock() delays freeing anon_vma pointer until the end * of migration. File cache pages are no problem because of page_lock() |
989f89c57
|
623 624 |
* File Caches may use write_page() or lock_page() in migration, then, * just care Anon page here. |
dc386d4d1
|
625 |
*/ |
989f89c57
|
626 627 628 629 |
if (PageAnon(page)) { rcu_read_lock(); rcu_locked = 1; } |
62e1c5530
|
630 |
|
dc386d4d1
|
631 |
/* |
62e1c5530
|
632 633 634 635 636 637 638 639 640 641 |
* Corner case handling: * 1. When a new swap-cache page is read into, it is added to the LRU * and treated as swapcache but it has no rmap yet. * Calling try_to_unmap() against a page->mapping==NULL page will * trigger a BUG. So handle it here. * 2. An orphaned page (see truncate_complete_page) might have * fs-private metadata. The page can be picked up due to memory * offlining. Everywhere else except page reclaim, the page is * invisible to the vm, so the page can not be migrated. So try to * free the metadata, so the page can be freed. |
e24f0b8f7
|
642 |
*/ |
62e1c5530
|
643 644 645 646 647 648 649 650 651 652 653 |
if (!page->mapping) { if (!PageAnon(page) && PagePrivate(page)) { /* * Go direct to try_to_free_buffers() here because * a) that's what try_to_release_page() would do anyway * b) we may be under rcu_read_lock() here, so we can't * use GFP_KERNEL which is what try_to_release_page() * needs to be effective. */ try_to_free_buffers(page); } |
dc386d4d1
|
654 |
goto rcu_unlock; |
62e1c5530
|
655 |
} |
dc386d4d1
|
656 |
/* Establish migration ptes or remove ptes */ |
e6a1530d6
|
657 |
try_to_unmap(page, 1); |
dc386d4d1
|
658 |
|
e6a1530d6
|
659 660 |
if (!page_mapped(page)) rc = move_to_new_page(newpage, page); |
e24f0b8f7
|
661 |
|
e8589cc18
|
662 |
if (rc) |
e24f0b8f7
|
663 |
remove_migration_ptes(page, page); |
dc386d4d1
|
664 |
rcu_unlock: |
989f89c57
|
665 666 |
if (rcu_locked) rcu_read_unlock(); |
e6a1530d6
|
667 |
|
e24f0b8f7
|
668 669 |
unlock: unlock_page(page); |
95a402c38
|
670 |
|
e24f0b8f7
|
671 |
if (rc != -EAGAIN) { |
aaa994b30
|
672 673 674 675 676 677 678 |
/* * A page that has been migrated has all references * removed and will be freed. A page that has not been * migrated will have kepts its references and be * restored. */ list_del(&page->lru); |
894bc3104
|
679 |
putback_lru_page(page); |
e24f0b8f7
|
680 |
} |
95a402c38
|
681 682 |
move_newpage: |
e8589cc18
|
683 684 |
if (!charge) mem_cgroup_end_migration(newpage); |
894bc3104
|
685 |
|
95a402c38
|
686 687 688 689 |
/* * Move the new page to the LRU. If migration was not successful * then this will free the page. */ |
894bc3104
|
690 |
putback_lru_page(newpage); |
742755a1d
|
691 692 693 694 695 696 |
if (result) { if (rc) *result = rc; else *result = page_to_nid(newpage); } |
e24f0b8f7
|
697 698 699 700 |
return rc; } /* |
b20a35035
|
701 702 |
* migrate_pages * |
95a402c38
|
703 704 705 |
* The function takes one list of pages to migrate and a function * that determines from the page to be migrated and the private data * the target of the move and allocates the page. |
b20a35035
|
706 707 708 |
* * The function returns after 10 attempts or if no pages * are movable anymore because to has become empty |
aaa994b30
|
709 |
* or no retryable pages exist anymore. All pages will be |
e9534b3fd
|
710 |
* returned to the LRU or freed. |
b20a35035
|
711 |
* |
95a402c38
|
712 |
* Return: Number of pages not migrated or error code. |
b20a35035
|
713 |
*/ |
95a402c38
|
714 715 |
int migrate_pages(struct list_head *from, new_page_t get_new_page, unsigned long private) |
b20a35035
|
716 |
{ |
e24f0b8f7
|
717 |
int retry = 1; |
b20a35035
|
718 719 720 721 722 723 724 725 726 |
int nr_failed = 0; int pass = 0; struct page *page; struct page *page2; int swapwrite = current->flags & PF_SWAPWRITE; int rc; if (!swapwrite) current->flags |= PF_SWAPWRITE; |
e24f0b8f7
|
727 728 |
for(pass = 0; pass < 10 && retry; pass++) { retry = 0; |
b20a35035
|
729 |
|
e24f0b8f7
|
730 |
list_for_each_entry_safe(page, page2, from, lru) { |
e24f0b8f7
|
731 |
cond_resched(); |
2d1db3b11
|
732 |
|
95a402c38
|
733 734 |
rc = unmap_and_move(get_new_page, private, page, pass > 2); |
2d1db3b11
|
735 |
|
e24f0b8f7
|
736 |
switch(rc) { |
95a402c38
|
737 738 |
case -ENOMEM: goto out; |
e24f0b8f7
|
739 |
case -EAGAIN: |
2d1db3b11
|
740 |
retry++; |
e24f0b8f7
|
741 742 |
break; case 0: |
e24f0b8f7
|
743 744 |
break; default: |
2d1db3b11
|
745 |
/* Permanent failure */ |
2d1db3b11
|
746 |
nr_failed++; |
e24f0b8f7
|
747 |
break; |
2d1db3b11
|
748 |
} |
b20a35035
|
749 750 |
} } |
95a402c38
|
751 752 |
rc = 0; out: |
b20a35035
|
753 754 |
if (!swapwrite) current->flags &= ~PF_SWAPWRITE; |
aaa994b30
|
755 |
putback_lru_pages(from); |
b20a35035
|
756 |
|
95a402c38
|
757 758 |
if (rc) return rc; |
b20a35035
|
759 |
|
95a402c38
|
760 |
return nr_failed + retry; |
b20a35035
|
761 |
} |
95a402c38
|
762 |
|
742755a1d
|
763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 |
#ifdef CONFIG_NUMA /* * Move a list of individual pages */ struct page_to_node { unsigned long addr; struct page *page; int node; int status; }; static struct page *new_page_node(struct page *p, unsigned long private, int **result) { struct page_to_node *pm = (struct page_to_node *)private; while (pm->node != MAX_NUMNODES && pm->page != p) pm++; if (pm->node == MAX_NUMNODES) return NULL; *result = &pm->status; |
769848c03
|
786 787 |
return alloc_pages_node(pm->node, GFP_HIGHUSER_MOVABLE | GFP_THISNODE, 0); |
742755a1d
|
788 789 790 791 792 793 |
} /* * Move a set of pages as indicated in the pm array. The addr * field must be set to the virtual address of the page to be moved * and the node number must contain a valid target node. |
5e9a0f023
|
794 |
* The pm array ends with node = MAX_NUMNODES. |
742755a1d
|
795 |
*/ |
5e9a0f023
|
796 797 798 |
static int do_move_page_to_node_array(struct mm_struct *mm, struct page_to_node *pm, int migrate_all) |
742755a1d
|
799 800 801 802 |
{ int err; struct page_to_node *pp; LIST_HEAD(pagelist); |
0aedadf91
|
803 |
migrate_prep(); |
742755a1d
|
804 805 806 807 808 |
down_read(&mm->mmap_sem); /* * Build a list of pages to migrate */ |
742755a1d
|
809 810 811 |
for (pp = pm; pp->node != MAX_NUMNODES; pp++) { struct vm_area_struct *vma; struct page *page; |
742755a1d
|
812 813 |
err = -EFAULT; vma = find_vma(mm, pp->addr); |
0dc952dc3
|
814 |
if (!vma || !vma_migratable(vma)) |
742755a1d
|
815 816 817 |
goto set_status; page = follow_page(vma, pp->addr, FOLL_GET); |
89f5b7da2
|
818 819 820 821 |
err = PTR_ERR(page); if (IS_ERR(page)) goto set_status; |
742755a1d
|
822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 |
err = -ENOENT; if (!page) goto set_status; if (PageReserved(page)) /* Check for zero page */ goto put_and_set; pp->page = page; err = page_to_nid(page); if (err == pp->node) /* * Node already in the right place */ goto put_and_set; err = -EACCES; if (page_mapcount(page) > 1 && !migrate_all) goto put_and_set; |
62695a84e
|
842 843 844 |
err = isolate_lru_page(page); if (!err) list_add_tail(&page->lru, &pagelist); |
742755a1d
|
845 846 847 848 849 850 851 852 853 854 |
put_and_set: /* * Either remove the duplicate refcount from * isolate_lru_page() or drop the page ref if it was * not isolated. */ put_page(page); set_status: pp->status = err; } |
e78bbfa82
|
855 |
err = 0; |
742755a1d
|
856 857 858 |
if (!list_empty(&pagelist)) err = migrate_pages(&pagelist, new_page_node, (unsigned long)pm); |
742755a1d
|
859 860 861 862 863 864 |
up_read(&mm->mmap_sem); return err; } /* |
5e9a0f023
|
865 866 867 868 869 870 871 872 873 |
* Migrate an array of page address onto an array of nodes and fill * the corresponding array of status. */ static int do_pages_move(struct mm_struct *mm, struct task_struct *task, unsigned long nr_pages, const void __user * __user *pages, const int __user *nodes, int __user *status, int flags) { |
3140a2273
|
874 |
struct page_to_node *pm; |
5e9a0f023
|
875 |
nodemask_t task_nodes; |
3140a2273
|
876 877 878 |
unsigned long chunk_nr_pages; unsigned long chunk_start; int err; |
5e9a0f023
|
879 880 |
task_nodes = cpuset_mems_allowed(task); |
3140a2273
|
881 882 883 |
err = -ENOMEM; pm = (struct page_to_node *)__get_free_page(GFP_KERNEL); if (!pm) |
5e9a0f023
|
884 |
goto out; |
5e9a0f023
|
885 |
/* |
3140a2273
|
886 887 |
* Store a chunk of page_to_node array in a page, * but keep the last one as a marker |
5e9a0f023
|
888 |
*/ |
3140a2273
|
889 |
chunk_nr_pages = (PAGE_SIZE / sizeof(struct page_to_node)) - 1; |
5e9a0f023
|
890 |
|
3140a2273
|
891 892 893 894 |
for (chunk_start = 0; chunk_start < nr_pages; chunk_start += chunk_nr_pages) { int j; |
5e9a0f023
|
895 |
|
3140a2273
|
896 897 898 899 900 901 |
if (chunk_start + chunk_nr_pages > nr_pages) chunk_nr_pages = nr_pages - chunk_start; /* fill the chunk pm with addrs and nodes from user-space */ for (j = 0; j < chunk_nr_pages; j++) { const void __user *p; |
5e9a0f023
|
902 |
int node; |
3140a2273
|
903 904 905 906 907 908 |
err = -EFAULT; if (get_user(p, pages + j + chunk_start)) goto out_pm; pm[j].addr = (unsigned long) p; if (get_user(node, nodes + j + chunk_start)) |
5e9a0f023
|
909 910 911 912 913 914 915 916 917 |
goto out_pm; err = -ENODEV; if (!node_state(node, N_HIGH_MEMORY)) goto out_pm; err = -EACCES; if (!node_isset(node, task_nodes)) goto out_pm; |
3140a2273
|
918 919 920 921 922 923 924 925 926 927 928 |
pm[j].node = node; } /* End marker for this chunk */ pm[chunk_nr_pages].node = MAX_NUMNODES; /* Migrate this chunk */ err = do_move_page_to_node_array(mm, pm, flags & MPOL_MF_MOVE_ALL); if (err < 0) goto out_pm; |
5e9a0f023
|
929 |
|
5e9a0f023
|
930 |
/* Return status information */ |
3140a2273
|
931 932 |
for (j = 0; j < chunk_nr_pages; j++) if (put_user(pm[j].status, status + j + chunk_start)) { |
5e9a0f023
|
933 |
err = -EFAULT; |
3140a2273
|
934 935 936 937 |
goto out_pm; } } err = 0; |
5e9a0f023
|
938 939 |
out_pm: |
3140a2273
|
940 |
free_page((unsigned long)pm); |
5e9a0f023
|
941 942 943 944 945 |
out: return err; } /* |
2f007e74b
|
946 |
* Determine the nodes of an array of pages and store it in an array of status. |
742755a1d
|
947 |
*/ |
80bba1290
|
948 949 |
static void do_pages_stat_array(struct mm_struct *mm, unsigned long nr_pages, const void __user **pages, int *status) |
742755a1d
|
950 |
{ |
2f007e74b
|
951 |
unsigned long i; |
2f007e74b
|
952 |
|
742755a1d
|
953 |
down_read(&mm->mmap_sem); |
2f007e74b
|
954 |
for (i = 0; i < nr_pages; i++) { |
80bba1290
|
955 |
unsigned long addr = (unsigned long)(*pages); |
742755a1d
|
956 957 |
struct vm_area_struct *vma; struct page *page; |
c095adbc2
|
958 |
int err = -EFAULT; |
2f007e74b
|
959 960 |
vma = find_vma(mm, addr); |
742755a1d
|
961 962 |
if (!vma) goto set_status; |
2f007e74b
|
963 |
page = follow_page(vma, addr, 0); |
89f5b7da2
|
964 965 966 967 |
err = PTR_ERR(page); if (IS_ERR(page)) goto set_status; |
742755a1d
|
968 969 970 971 972 973 974 |
err = -ENOENT; /* Use PageReserved to check for zero page */ if (!page || PageReserved(page)) goto set_status; err = page_to_nid(page); set_status: |
80bba1290
|
975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 |
*status = err; pages++; status++; } up_read(&mm->mmap_sem); } /* * Determine the nodes of a user array of pages and store it in * a user array of status. */ static int do_pages_stat(struct mm_struct *mm, unsigned long nr_pages, const void __user * __user *pages, int __user *status) { #define DO_PAGES_STAT_CHUNK_NR 16 const void __user *chunk_pages[DO_PAGES_STAT_CHUNK_NR]; int chunk_status[DO_PAGES_STAT_CHUNK_NR]; unsigned long i, chunk_nr = DO_PAGES_STAT_CHUNK_NR; int err; for (i = 0; i < nr_pages; i += chunk_nr) { if (chunk_nr + i > nr_pages) chunk_nr = nr_pages - i; err = copy_from_user(chunk_pages, &pages[i], chunk_nr * sizeof(*chunk_pages)); if (err) { err = -EFAULT; goto out; } do_pages_stat_array(mm, chunk_nr, chunk_pages, chunk_status); err = copy_to_user(&status[i], chunk_status, chunk_nr * sizeof(*chunk_status)); if (err) { err = -EFAULT; goto out; } |
742755a1d
|
1017 |
} |
2f007e74b
|
1018 |
err = 0; |
742755a1d
|
1019 |
|
2f007e74b
|
1020 |
out: |
2f007e74b
|
1021 |
return err; |
742755a1d
|
1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 |
} /* * Move a list of pages in the address space of the currently executing * process. */ asmlinkage long sys_move_pages(pid_t pid, unsigned long nr_pages, const void __user * __user *pages, const int __user *nodes, int __user *status, int flags) { |
c69e8d9c0
|
1033 |
const struct cred *cred = current_cred(), *tcred; |
742755a1d
|
1034 |
struct task_struct *task; |
742755a1d
|
1035 |
struct mm_struct *mm; |
5e9a0f023
|
1036 |
int err; |
742755a1d
|
1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 |
/* Check flags */ if (flags & ~(MPOL_MF_MOVE|MPOL_MF_MOVE_ALL)) return -EINVAL; if ((flags & MPOL_MF_MOVE_ALL) && !capable(CAP_SYS_NICE)) return -EPERM; /* Find the mm_struct */ read_lock(&tasklist_lock); |
228ebcbe6
|
1047 |
task = pid ? find_task_by_vpid(pid) : current; |
742755a1d
|
1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 |
if (!task) { read_unlock(&tasklist_lock); return -ESRCH; } mm = get_task_mm(task); read_unlock(&tasklist_lock); if (!mm) return -EINVAL; /* * Check if this process has the right to modify the specified * process. The right exists if the process has administrative * capabilities, superuser privileges or the same * userid as the target process. */ |
c69e8d9c0
|
1064 1065 |
rcu_read_lock(); tcred = __task_cred(task); |
b6dff3ec5
|
1066 1067 |
if (cred->euid != tcred->suid && cred->euid != tcred->uid && cred->uid != tcred->suid && cred->uid != tcred->uid && |
742755a1d
|
1068 |
!capable(CAP_SYS_NICE)) { |
c69e8d9c0
|
1069 |
rcu_read_unlock(); |
742755a1d
|
1070 |
err = -EPERM; |
5e9a0f023
|
1071 |
goto out; |
742755a1d
|
1072 |
} |
c69e8d9c0
|
1073 |
rcu_read_unlock(); |
742755a1d
|
1074 |
|
86c3a7645
|
1075 1076 |
err = security_task_movememory(task); if (err) |
5e9a0f023
|
1077 |
goto out; |
86c3a7645
|
1078 |
|
5e9a0f023
|
1079 1080 1081 1082 |
if (nodes) { err = do_pages_move(mm, task, nr_pages, pages, nodes, status, flags); } else { |
2f007e74b
|
1083 |
err = do_pages_stat(mm, nr_pages, pages, status); |
742755a1d
|
1084 |
} |
742755a1d
|
1085 |
out: |
742755a1d
|
1086 1087 1088 |
mmput(mm); return err; } |
742755a1d
|
1089 |
|
7b2259b3e
|
1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 |
/* * Call migration functions in the vma_ops that may prepare * memory in a vm for migration. migration functions may perform * the migration for vmas that do not have an underlying page struct. */ int migrate_vmas(struct mm_struct *mm, const nodemask_t *to, const nodemask_t *from, unsigned long flags) { struct vm_area_struct *vma; int err = 0; for(vma = mm->mmap; vma->vm_next && !err; vma = vma->vm_next) { if (vma->vm_ops && vma->vm_ops->migrate) { err = vma->vm_ops->migrate(vma, to, from, flags); if (err) break; } } return err; } |
83d1674a9
|
1110 |
#endif |