Blame view
mm/migrate.c
31.6 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 |
#include <linux/pagevec.h> |
e9995ef97
|
24 |
#include <linux/ksm.h> |
b20a35035
|
25 26 27 28 |
#include <linux/rmap.h> #include <linux/topology.h> #include <linux/cpu.h> #include <linux/cpuset.h> |
04e62a29b
|
29 |
#include <linux/writeback.h> |
742755a1d
|
30 31 |
#include <linux/mempolicy.h> #include <linux/vmalloc.h> |
86c3a7645
|
32 |
#include <linux/security.h> |
8a9f3ccd2
|
33 |
#include <linux/memcontrol.h> |
4f5ca2657
|
34 |
#include <linux/syscalls.h> |
290408d4a
|
35 |
#include <linux/hugetlb.h> |
5a0e3ad6a
|
36 |
#include <linux/gfp.h> |
b20a35035
|
37 |
|
0d1836c36
|
38 |
#include <asm/tlbflush.h> |
b20a35035
|
39 |
#include "internal.h" |
b20a35035
|
40 41 42 |
#define lru_to_page(_head) (list_entry((_head)->prev, struct page, lru)) /* |
742755a1d
|
43 |
* migrate_prep() needs to be called before we start compiling a list of pages |
748446bb6
|
44 45 |
* to be migrated using isolate_lru_page(). If scheduling work on other CPUs is * undesirable, use migrate_prep_local() |
b20a35035
|
46 47 48 |
*/ int migrate_prep(void) { |
b20a35035
|
49 50 51 52 53 54 55 56 57 58 |
/* * 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; } |
748446bb6
|
59 60 61 62 63 64 65 |
/* Do the necessary work of migrate_prep but not if it involves other CPUs */ int migrate_prep_local(void) { lru_add_drain(); return 0; } |
b20a35035
|
66 |
/* |
894bc3104
|
67 68 |
* Add isolated pages on the list back to the LRU under page lock * to avoid leaking evictable pages back onto unevictable list. |
b20a35035
|
69 |
*/ |
e13861d82
|
70 |
void putback_lru_pages(struct list_head *l) |
b20a35035
|
71 72 73 |
{ struct page *page; struct page *page2; |
b20a35035
|
74 75 |
list_for_each_entry_safe(page, page2, l, lru) { |
e24f0b8f7
|
76 |
list_del(&page->lru); |
a731286de
|
77 |
dec_zone_page_state(page, NR_ISOLATED_ANON + |
6c0b13519
|
78 |
page_is_file_cache(page)); |
894bc3104
|
79 |
putback_lru_page(page); |
b20a35035
|
80 |
} |
b20a35035
|
81 |
} |
0697212a4
|
82 83 84 |
/* * Restore a potential migration pte to a working pte entry */ |
e9995ef97
|
85 86 |
static int remove_migration_pte(struct page *new, struct vm_area_struct *vma, unsigned long addr, void *old) |
0697212a4
|
87 88 89 90 91 92 93 94 |
{ 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; |
290408d4a
|
95 96 97 98 99 100 101 102 103 |
if (unlikely(PageHuge(new))) { ptep = huge_pte_offset(mm, addr); if (!ptep) goto out; ptl = &mm->page_table_lock; } else { pgd = pgd_offset(mm, addr); if (!pgd_present(*pgd)) goto out; |
0697212a4
|
104 |
|
290408d4a
|
105 106 107 |
pud = pud_offset(pgd, addr); if (!pud_present(*pud)) goto out; |
0697212a4
|
108 |
|
290408d4a
|
109 |
pmd = pmd_offset(pud, addr); |
500d65d47
|
110 111 |
if (pmd_trans_huge(*pmd)) goto out; |
290408d4a
|
112 113 |
if (!pmd_present(*pmd)) goto out; |
0697212a4
|
114 |
|
290408d4a
|
115 |
ptep = pte_offset_map(pmd, addr); |
0697212a4
|
116 |
|
290408d4a
|
117 118 119 120 121 122 123 |
if (!is_swap_pte(*ptep)) { pte_unmap(ptep); goto out; } ptl = pte_lockptr(mm, pmd); } |
0697212a4
|
124 |
|
0697212a4
|
125 126 127 |
spin_lock(ptl); pte = *ptep; if (!is_swap_pte(pte)) |
e9995ef97
|
128 |
goto unlock; |
0697212a4
|
129 130 |
entry = pte_to_swp_entry(pte); |
e9995ef97
|
131 132 133 |
if (!is_migration_entry(entry) || migration_entry_to_page(entry) != old) goto unlock; |
0697212a4
|
134 |
|
0697212a4
|
135 136 137 138 |
get_page(new); pte = pte_mkold(mk_pte(new, vma->vm_page_prot)); if (is_write_migration_entry(entry)) pte = pte_mkwrite(pte); |
3ef8fd7f7
|
139 |
#ifdef CONFIG_HUGETLB_PAGE |
290408d4a
|
140 141 |
if (PageHuge(new)) pte = pte_mkhuge(pte); |
3ef8fd7f7
|
142 |
#endif |
97ee05246
|
143 |
flush_cache_page(vma, addr, pte_pfn(pte)); |
0697212a4
|
144 |
set_pte_at(mm, addr, ptep, pte); |
04e62a29b
|
145 |
|
290408d4a
|
146 147 148 149 150 151 |
if (PageHuge(new)) { if (PageAnon(new)) hugepage_add_anon_rmap(new, vma, addr); else page_dup_rmap(new); } else if (PageAnon(new)) |
04e62a29b
|
152 153 154 155 156 |
page_add_anon_rmap(new, vma, addr); else page_add_file_rmap(new); /* No need to invalidate - it was non-present before */ |
4b3073e1c
|
157 |
update_mmu_cache(vma, addr, ptep); |
e9995ef97
|
158 |
unlock: |
0697212a4
|
159 |
pte_unmap_unlock(ptep, ptl); |
e9995ef97
|
160 161 |
out: return SWAP_AGAIN; |
0697212a4
|
162 163 164 |
} /* |
04e62a29b
|
165 166 167 168 169 |
* 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) { |
e9995ef97
|
170 |
rmap_walk(new, remove_migration_pte, old); |
04e62a29b
|
171 172 173 |
} /* |
0697212a4
|
174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 |
* 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
|
198 199 200 201 202 203 204 205 206 |
/* * 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
|
207 208 209 210 211 212 213 |
pte_unmap_unlock(ptep, ptl); wait_on_page_locked(page); put_page(page); return; out: pte_unmap_unlock(ptep, ptl); } |
b20a35035
|
214 |
/* |
c3fcf8a5d
|
215 |
* Replace the page in the mapping. |
5b5c7120e
|
216 217 218 219 |
* * The number of remaining references must be: * 1 for anonymous pages without a mapping * 2 for pages with a mapping |
266cf658e
|
220 |
* 3 for pages with a mapping and PagePrivate/PagePrivate2 set. |
b20a35035
|
221 |
*/ |
2d1db3b11
|
222 223 |
static int migrate_page_move_mapping(struct address_space *mapping, struct page *newpage, struct page *page) |
b20a35035
|
224 |
{ |
e286781d5
|
225 |
int expected_count; |
7cf9c2c76
|
226 |
void **pslot; |
b20a35035
|
227 |
|
6c5240ae7
|
228 |
if (!mapping) { |
0e8c7d0fd
|
229 |
/* Anonymous page without mapping */ |
6c5240ae7
|
230 231 232 233 |
if (page_count(page) != 1) return -EAGAIN; return 0; } |
19fd62312
|
234 |
spin_lock_irq(&mapping->tree_lock); |
b20a35035
|
235 |
|
7cf9c2c76
|
236 237 |
pslot = radix_tree_lookup_slot(&mapping->page_tree, page_index(page)); |
b20a35035
|
238 |
|
edcf4748c
|
239 |
expected_count = 2 + page_has_private(page); |
e286781d5
|
240 |
if (page_count(page) != expected_count || |
29c1f677d
|
241 |
radix_tree_deref_slot_protected(pslot, &mapping->tree_lock) != page) { |
19fd62312
|
242 |
spin_unlock_irq(&mapping->tree_lock); |
e23ca00bf
|
243 |
return -EAGAIN; |
b20a35035
|
244 |
} |
e286781d5
|
245 |
if (!page_freeze_refs(page, expected_count)) { |
19fd62312
|
246 |
spin_unlock_irq(&mapping->tree_lock); |
e286781d5
|
247 248 |
return -EAGAIN; } |
b20a35035
|
249 250 |
/* * Now we know that no one else is looking at the page. |
b20a35035
|
251 |
*/ |
7cf9c2c76
|
252 |
get_page(newpage); /* add cache reference */ |
b20a35035
|
253 254 255 256 |
if (PageSwapCache(page)) { SetPageSwapCache(newpage); set_page_private(newpage, page_private(page)); } |
7cf9c2c76
|
257 |
radix_tree_replace_slot(pslot, newpage); |
e286781d5
|
258 |
page_unfreeze_refs(page, expected_count); |
7cf9c2c76
|
259 260 261 262 |
/* * Drop cache reference from old page. * We know this isn't the last reference. */ |
b20a35035
|
263 |
__put_page(page); |
7cf9c2c76
|
264 |
|
0e8c7d0fd
|
265 266 267 268 269 270 271 272 273 274 275 276 |
/* * 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); |
4b02108ac
|
277 278 279 280 |
if (PageSwapBacked(page)) { __dec_zone_page_state(page, NR_SHMEM); __inc_zone_page_state(newpage, NR_SHMEM); } |
19fd62312
|
281 |
spin_unlock_irq(&mapping->tree_lock); |
b20a35035
|
282 283 284 |
return 0; } |
b20a35035
|
285 286 |
/* |
290408d4a
|
287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 |
* The expected number of remaining references is the same as that * of migrate_page_move_mapping(). */ int migrate_huge_page_move_mapping(struct address_space *mapping, struct page *newpage, struct page *page) { int expected_count; void **pslot; if (!mapping) { if (page_count(page) != 1) return -EAGAIN; return 0; } spin_lock_irq(&mapping->tree_lock); pslot = radix_tree_lookup_slot(&mapping->page_tree, page_index(page)); expected_count = 2 + page_has_private(page); if (page_count(page) != expected_count || |
29c1f677d
|
309 |
radix_tree_deref_slot_protected(pslot, &mapping->tree_lock) != page) { |
290408d4a
|
310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 |
spin_unlock_irq(&mapping->tree_lock); return -EAGAIN; } if (!page_freeze_refs(page, expected_count)) { spin_unlock_irq(&mapping->tree_lock); return -EAGAIN; } get_page(newpage); radix_tree_replace_slot(pslot, newpage); page_unfreeze_refs(page, expected_count); __put_page(page); spin_unlock_irq(&mapping->tree_lock); return 0; } /* |
b20a35035
|
332 333 |
* Copy the page to its new location */ |
290408d4a
|
334 |
void migrate_page_copy(struct page *newpage, struct page *page) |
b20a35035
|
335 |
{ |
290408d4a
|
336 337 338 339 |
if (PageHuge(page)) copy_huge_page(newpage, page); else copy_highpage(newpage, page); |
b20a35035
|
340 341 342 343 344 345 346 |
if (PageError(page)) SetPageError(newpage); if (PageReferenced(page)) SetPageReferenced(newpage); if (PageUptodate(page)) SetPageUptodate(newpage); |
894bc3104
|
347 348 |
if (TestClearPageActive(page)) { VM_BUG_ON(PageUnevictable(page)); |
b20a35035
|
349 |
SetPageActive(newpage); |
418b27ef5
|
350 351 |
} else if (TestClearPageUnevictable(page)) SetPageUnevictable(newpage); |
b20a35035
|
352 353 354 355 356 357 358 |
if (PageChecked(page)) SetPageChecked(newpage); if (PageMappedToDisk(page)) SetPageMappedToDisk(newpage); if (PageDirty(page)) { clear_page_dirty_for_io(page); |
3a902c5f6
|
359 360 361 362 363 |
/* * 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. |
25985edce
|
364 |
* Whereas only part of our page may be dirty. |
3a902c5f6
|
365 366 |
*/ __set_page_dirty_nobuffers(newpage); |
b20a35035
|
367 |
} |
b291f0003
|
368 |
mlock_migrate_page(newpage, page); |
e9995ef97
|
369 |
ksm_migrate_page(newpage, page); |
b291f0003
|
370 |
|
b20a35035
|
371 |
ClearPageSwapCache(page); |
b20a35035
|
372 373 374 375 376 377 378 379 380 381 382 |
ClearPagePrivate(page); set_page_private(page, 0); page->mapping = NULL; /* * If any waiters have accumulated on the new page then * wake them up. */ if (PageWriteback(newpage)) end_page_writeback(newpage); } |
b20a35035
|
383 |
|
1d8b85ccf
|
384 385 386 387 388 |
/************************************************************ * Migration functions ***********************************************************/ /* Always fail migration. Used for mappings that are not movable */ |
2d1db3b11
|
389 390 |
int fail_migrate_page(struct address_space *mapping, struct page *newpage, struct page *page) |
1d8b85ccf
|
391 392 393 394 |
{ return -EIO; } EXPORT_SYMBOL(fail_migrate_page); |
b20a35035
|
395 396 |
/* * Common logic to directly migrate a single page suitable for |
266cf658e
|
397 |
* pages that do not use PagePrivate/PagePrivate2. |
b20a35035
|
398 399 400 |
* * Pages are locked upon entry and exit. */ |
2d1db3b11
|
401 402 |
int migrate_page(struct address_space *mapping, struct page *newpage, struct page *page) |
b20a35035
|
403 404 405 406 |
{ int rc; BUG_ON(PageWriteback(page)); /* Writeback must be complete */ |
2d1db3b11
|
407 |
rc = migrate_page_move_mapping(mapping, newpage, page); |
b20a35035
|
408 409 410 411 412 |
if (rc) return rc; migrate_page_copy(newpage, page); |
b20a35035
|
413 414 415 |
return 0; } EXPORT_SYMBOL(migrate_page); |
9361401eb
|
416 |
#ifdef CONFIG_BLOCK |
b20a35035
|
417 |
/* |
1d8b85ccf
|
418 419 420 421 |
* 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
|
422 423 |
int buffer_migrate_page(struct address_space *mapping, struct page *newpage, struct page *page) |
1d8b85ccf
|
424 |
{ |
1d8b85ccf
|
425 426 |
struct buffer_head *bh, *head; int rc; |
1d8b85ccf
|
427 |
if (!page_has_buffers(page)) |
2d1db3b11
|
428 |
return migrate_page(mapping, newpage, page); |
1d8b85ccf
|
429 430 |
head = page_buffers(page); |
2d1db3b11
|
431 |
rc = migrate_page_move_mapping(mapping, newpage, page); |
1d8b85ccf
|
432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 |
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
|
472 |
#endif |
1d8b85ccf
|
473 |
|
04e62a29b
|
474 475 476 477 |
/* * Writeback a page to clean the dirty state */ static int writeout(struct address_space *mapping, struct page *page) |
8351a6e47
|
478 |
{ |
04e62a29b
|
479 480 481 482 483 |
struct writeback_control wbc = { .sync_mode = WB_SYNC_NONE, .nr_to_write = 1, .range_start = 0, .range_end = LLONG_MAX, |
04e62a29b
|
484 485 486 487 488 489 490 491 492 493 494 |
.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
|
495 |
/* |
04e62a29b
|
496 497 498 499 500 501 |
* 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
|
502 |
*/ |
04e62a29b
|
503 |
remove_migration_ptes(page, page); |
8351a6e47
|
504 |
|
04e62a29b
|
505 |
rc = mapping->a_ops->writepage(page, &wbc); |
8351a6e47
|
506 |
|
04e62a29b
|
507 508 509 |
if (rc != AOP_WRITEPAGE_ACTIVATE) /* unlocked. Relock */ lock_page(page); |
bda8550de
|
510 |
return (rc < 0) ? -EIO : -EAGAIN; |
04e62a29b
|
511 512 513 514 515 516 517 518 519 520 |
} /* * 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
|
521 522 523 524 525 |
/* * Buffers may be managed in a filesystem specific way. * We must have no buffers or drop them. */ |
266cf658e
|
526 |
if (page_has_private(page) && |
8351a6e47
|
527 528 529 530 531 |
!try_to_release_page(page, GFP_KERNEL)) return -EAGAIN; return migrate_page(mapping, newpage, page); } |
1d8b85ccf
|
532 |
/* |
e24f0b8f7
|
533 534 535 536 537 |
* 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
|
538 539 540 541 |
* * Return value: * < 0 - error code * == 0 - success |
e24f0b8f7
|
542 |
*/ |
3fe2011ff
|
543 |
static int move_to_new_page(struct page *newpage, struct page *page, |
11bc82d67
|
544 |
int remap_swapcache, bool sync) |
e24f0b8f7
|
545 546 547 548 549 550 551 552 553 |
{ 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
|
554 |
if (!trylock_page(newpage)) |
e24f0b8f7
|
555 556 557 558 559 |
BUG(); /* Prepare mapping for the new page.*/ newpage->index = page->index; newpage->mapping = page->mapping; |
b2e185384
|
560 561 |
if (PageSwapBacked(page)) SetPageSwapBacked(newpage); |
e24f0b8f7
|
562 563 564 565 |
mapping = page_mapping(page); if (!mapping) rc = migrate_page(mapping, newpage, page); |
11bc82d67
|
566 |
else { |
e24f0b8f7
|
567 |
/* |
11bc82d67
|
568 569 570 |
* Do not writeback pages if !sync and migratepage is * not pointing to migrate_page() which is nonblocking * (swapcache/tmpfs uses migratepage = migrate_page). |
e24f0b8f7
|
571 |
*/ |
11bc82d67
|
572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 |
if (PageDirty(page) && !sync && mapping->a_ops->migratepage != migrate_page) rc = -EBUSY; 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); } |
e24f0b8f7
|
588 |
|
3fe2011ff
|
589 |
if (rc) { |
e24f0b8f7
|
590 |
newpage->mapping = NULL; |
3fe2011ff
|
591 592 593 594 |
} else { if (remap_swapcache) remove_migration_ptes(page, newpage); } |
e24f0b8f7
|
595 596 597 598 599 600 601 602 603 604 |
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
|
605 |
static int unmap_and_move(new_page_t get_new_page, unsigned long private, |
7f0f24967
|
606 |
struct page *page, int force, bool offlining, bool sync) |
e24f0b8f7
|
607 608 |
{ int rc = 0; |
742755a1d
|
609 610 |
int *result = NULL; struct page *newpage = get_new_page(page, private, &result); |
3fe2011ff
|
611 |
int remap_swapcache = 1; |
ae41be374
|
612 |
int charge = 0; |
56039efa1
|
613 |
struct mem_cgroup *mem; |
3f6c82728
|
614 |
struct anon_vma *anon_vma = NULL; |
95a402c38
|
615 616 617 |
if (!newpage) return -ENOMEM; |
e24f0b8f7
|
618 |
|
894bc3104
|
619 |
if (page_count(page) == 1) { |
e24f0b8f7
|
620 |
/* page was freed from under us. So we are done. */ |
95a402c38
|
621 |
goto move_newpage; |
894bc3104
|
622 |
} |
500d65d47
|
623 624 625 |
if (unlikely(PageTransHuge(page))) if (unlikely(split_huge_page(page))) goto move_newpage; |
e24f0b8f7
|
626 |
|
e8589cc18
|
627 |
/* prepare cgroup just returns 0 or -ENOMEM */ |
e24f0b8f7
|
628 |
rc = -EAGAIN; |
01b1ae63c
|
629 |
|
529ae9aaa
|
630 |
if (!trylock_page(page)) { |
11bc82d67
|
631 |
if (!force || !sync) |
95a402c38
|
632 |
goto move_newpage; |
3e7d34497
|
633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 |
/* * It's not safe for direct compaction to call lock_page. * For example, during page readahead pages are added locked * to the LRU. Later, when the IO completes the pages are * marked uptodate and unlocked. However, the queueing * could be merging multiple pages for one bio (e.g. * mpage_readpages). If an allocation happens for the * second or third page, the process can end up locking * the same page twice and deadlocking. Rather than * trying to be clever about what pages can be locked, * avoid the use of lock_page for direct compaction * altogether. */ if (current->flags & PF_MEMALLOC) goto move_newpage; |
e24f0b8f7
|
649 650 |
lock_page(page); } |
62b61f611
|
651 652 653 654 655 656 657 658 659 660 661 662 663 |
/* * Only memory hotplug's offline_pages() caller has locked out KSM, * and can safely migrate a KSM page. The other cases have skipped * PageKsm along with PageReserved - but it is only now when we have * the page lock that we can be certain it will not go KSM beneath us * (KSM will not upgrade a page from PageAnon to PageKsm when it sees * its pagecount raised, but only here do we take the page lock which * serializes that). */ if (PageKsm(page) && !offlining) { rc = -EBUSY; goto unlock; } |
01b1ae63c
|
664 |
/* charge against new page */ |
ef6a3c631
|
665 |
charge = mem_cgroup_prepare_migration(page, newpage, &mem, GFP_KERNEL); |
01b1ae63c
|
666 667 668 669 670 |
if (charge == -ENOMEM) { rc = -ENOMEM; goto unlock; } BUG_ON(charge); |
e24f0b8f7
|
671 |
if (PageWriteback(page)) { |
11bc82d67
|
672 673 674 675 676 677 678 679 680 |
/* * For !sync, there is no point retrying as the retry loop * is expected to be too short for PageWriteback to be cleared */ if (!sync) { rc = -EBUSY; goto uncharge; } if (!force) |
01b1ae63c
|
681 |
goto uncharge; |
e24f0b8f7
|
682 683 |
wait_on_page_writeback(page); } |
e24f0b8f7
|
684 |
/* |
dc386d4d1
|
685 686 |
* 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. |
1ce82b69e
|
687 |
* This get_anon_vma() delays freeing anon_vma pointer until the end |
dc386d4d1
|
688 |
* of migration. File cache pages are no problem because of page_lock() |
989f89c57
|
689 690 |
* File Caches may use write_page() or lock_page() in migration, then, * just care Anon page here. |
dc386d4d1
|
691 |
*/ |
989f89c57
|
692 |
if (PageAnon(page)) { |
1ce82b69e
|
693 694 695 696 |
/* * Only page_lock_anon_vma() understands the subtleties of * getting a hold on an anon_vma from outside one of its mms. */ |
746b18d42
|
697 |
anon_vma = page_get_anon_vma(page); |
1ce82b69e
|
698 699 |
if (anon_vma) { /* |
746b18d42
|
700 |
* Anon page |
1ce82b69e
|
701 |
*/ |
1ce82b69e
|
702 |
} else if (PageSwapCache(page)) { |
3fe2011ff
|
703 704 705 706 707 708 709 710 711 712 713 714 715 716 |
/* * We cannot be sure that the anon_vma of an unmapped * swapcache page is safe to use because we don't * know in advance if the VMA that this page belonged * to still exists. If the VMA and others sharing the * data have been freed, then the anon_vma could * already be invalid. * * To avoid this possibility, swapcache pages get * migrated but are not remapped when migration * completes */ remap_swapcache = 0; } else { |
1ce82b69e
|
717 |
goto uncharge; |
3fe2011ff
|
718 |
} |
989f89c57
|
719 |
} |
62e1c5530
|
720 |
|
dc386d4d1
|
721 |
/* |
62e1c5530
|
722 723 724 725 726 727 728 729 730 731 |
* 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
|
732 |
*/ |
62e1c5530
|
733 |
if (!page->mapping) { |
1ce82b69e
|
734 735 |
VM_BUG_ON(PageAnon(page)); if (page_has_private(page)) { |
62e1c5530
|
736 |
try_to_free_buffers(page); |
1ce82b69e
|
737 |
goto uncharge; |
62e1c5530
|
738 |
} |
abfc34881
|
739 |
goto skip_unmap; |
62e1c5530
|
740 |
} |
dc386d4d1
|
741 |
/* Establish migration ptes or remove ptes */ |
14fa31b89
|
742 |
try_to_unmap(page, TTU_MIGRATION|TTU_IGNORE_MLOCK|TTU_IGNORE_ACCESS); |
dc386d4d1
|
743 |
|
abfc34881
|
744 |
skip_unmap: |
e6a1530d6
|
745 |
if (!page_mapped(page)) |
11bc82d67
|
746 |
rc = move_to_new_page(newpage, page, remap_swapcache, sync); |
e24f0b8f7
|
747 |
|
3fe2011ff
|
748 |
if (rc && remap_swapcache) |
e24f0b8f7
|
749 |
remove_migration_ptes(page, page); |
3f6c82728
|
750 751 |
/* Drop an anon_vma reference if we took one */ |
76545066c
|
752 |
if (anon_vma) |
9e60109f1
|
753 |
put_anon_vma(anon_vma); |
3f6c82728
|
754 |
|
01b1ae63c
|
755 756 |
uncharge: if (!charge) |
50de1dd96
|
757 |
mem_cgroup_end_migration(mem, page, newpage, rc == 0); |
e24f0b8f7
|
758 759 |
unlock: unlock_page(page); |
95a402c38
|
760 |
|
57fc4a5ee
|
761 |
move_newpage: |
e24f0b8f7
|
762 |
if (rc != -EAGAIN) { |
aaa994b30
|
763 764 765 766 767 768 769 |
/* * 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); |
a731286de
|
770 |
dec_zone_page_state(page, NR_ISOLATED_ANON + |
6c0b13519
|
771 |
page_is_file_cache(page)); |
894bc3104
|
772 |
putback_lru_page(page); |
e24f0b8f7
|
773 |
} |
95a402c38
|
774 |
|
95a402c38
|
775 776 777 778 |
/* * Move the new page to the LRU. If migration was not successful * then this will free the page. */ |
894bc3104
|
779 |
putback_lru_page(newpage); |
742755a1d
|
780 781 782 783 784 785 |
if (result) { if (rc) *result = rc; else *result = page_to_nid(newpage); } |
e24f0b8f7
|
786 787 788 789 |
return rc; } /* |
290408d4a
|
790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 |
* Counterpart of unmap_and_move_page() for hugepage migration. * * This function doesn't wait the completion of hugepage I/O * because there is no race between I/O and migration for hugepage. * Note that currently hugepage I/O occurs only in direct I/O * where no lock is held and PG_writeback is irrelevant, * and writeback status of all subpages are counted in the reference * count of the head page (i.e. if all subpages of a 2MB hugepage are * under direct I/O, the reference of the head page is 512 and a bit more.) * This means that when we try to migrate hugepage whose subpages are * doing direct I/O, some references remain after try_to_unmap() and * hugepage migration fails without data corruption. * * There is also no race when direct I/O is issued on the page under migration, * because then pte is replaced with migration swap entry and direct I/O code * will wait in the page fault for migration to complete. */ static int unmap_and_move_huge_page(new_page_t get_new_page, unsigned long private, struct page *hpage, |
7f0f24967
|
809 |
int force, bool offlining, bool sync) |
290408d4a
|
810 811 812 813 |
{ int rc = 0; int *result = NULL; struct page *new_hpage = get_new_page(hpage, private, &result); |
290408d4a
|
814 815 816 817 818 819 820 821 |
struct anon_vma *anon_vma = NULL; if (!new_hpage) return -ENOMEM; rc = -EAGAIN; if (!trylock_page(hpage)) { |
77f1fe6b0
|
822 |
if (!force || !sync) |
290408d4a
|
823 824 825 |
goto out; lock_page(hpage); } |
746b18d42
|
826 827 |
if (PageAnon(hpage)) anon_vma = page_get_anon_vma(hpage); |
290408d4a
|
828 829 830 831 |
try_to_unmap(hpage, TTU_MIGRATION|TTU_IGNORE_MLOCK|TTU_IGNORE_ACCESS); if (!page_mapped(hpage)) |
11bc82d67
|
832 |
rc = move_to_new_page(new_hpage, hpage, 1, sync); |
290408d4a
|
833 834 835 |
if (rc) remove_migration_ptes(hpage, hpage); |
fd4a4663d
|
836 |
if (anon_vma) |
9e60109f1
|
837 |
put_anon_vma(anon_vma); |
290408d4a
|
838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 |
out: unlock_page(hpage); if (rc != -EAGAIN) { list_del(&hpage->lru); put_page(hpage); } put_page(new_hpage); if (result) { if (rc) *result = rc; else *result = page_to_nid(new_hpage); } return rc; } /* |
b20a35035
|
858 859 |
* migrate_pages * |
95a402c38
|
860 861 862 |
* 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
|
863 864 865 |
* * The function returns after 10 attempts or if no pages * are movable anymore because to has become empty |
cf608ac19
|
866 867 |
* or no retryable pages exist anymore. * Caller should call putback_lru_pages to return pages to the LRU |
28bd65781
|
868 |
* or free list only if ret != 0. |
b20a35035
|
869 |
* |
95a402c38
|
870 |
* Return: Number of pages not migrated or error code. |
b20a35035
|
871 |
*/ |
95a402c38
|
872 |
int migrate_pages(struct list_head *from, |
7f0f24967
|
873 |
new_page_t get_new_page, unsigned long private, bool offlining, |
77f1fe6b0
|
874 |
bool sync) |
b20a35035
|
875 |
{ |
e24f0b8f7
|
876 |
int retry = 1; |
b20a35035
|
877 878 879 880 881 882 883 884 885 |
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
|
886 887 |
for(pass = 0; pass < 10 && retry; pass++) { retry = 0; |
b20a35035
|
888 |
|
e24f0b8f7
|
889 |
list_for_each_entry_safe(page, page2, from, lru) { |
e24f0b8f7
|
890 |
cond_resched(); |
2d1db3b11
|
891 |
|
95a402c38
|
892 |
rc = unmap_and_move(get_new_page, private, |
77f1fe6b0
|
893 894 |
page, pass > 2, offlining, sync); |
2d1db3b11
|
895 |
|
e24f0b8f7
|
896 |
switch(rc) { |
95a402c38
|
897 898 |
case -ENOMEM: goto out; |
e24f0b8f7
|
899 |
case -EAGAIN: |
2d1db3b11
|
900 |
retry++; |
e24f0b8f7
|
901 902 |
break; case 0: |
e24f0b8f7
|
903 904 |
break; default: |
2d1db3b11
|
905 |
/* Permanent failure */ |
2d1db3b11
|
906 |
nr_failed++; |
e24f0b8f7
|
907 |
break; |
2d1db3b11
|
908 |
} |
b20a35035
|
909 910 |
} } |
95a402c38
|
911 912 |
rc = 0; out: |
b20a35035
|
913 914 |
if (!swapwrite) current->flags &= ~PF_SWAPWRITE; |
95a402c38
|
915 916 |
if (rc) return rc; |
b20a35035
|
917 |
|
95a402c38
|
918 |
return nr_failed + retry; |
b20a35035
|
919 |
} |
95a402c38
|
920 |
|
290408d4a
|
921 |
int migrate_huge_pages(struct list_head *from, |
7f0f24967
|
922 |
new_page_t get_new_page, unsigned long private, bool offlining, |
77f1fe6b0
|
923 |
bool sync) |
290408d4a
|
924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 |
{ int retry = 1; int nr_failed = 0; int pass = 0; struct page *page; struct page *page2; int rc; for (pass = 0; pass < 10 && retry; pass++) { retry = 0; list_for_each_entry_safe(page, page2, from, lru) { cond_resched(); rc = unmap_and_move_huge_page(get_new_page, |
77f1fe6b0
|
939 940 |
private, page, pass > 2, offlining, sync); |
290408d4a
|
941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 |
switch(rc) { case -ENOMEM: goto out; case -EAGAIN: retry++; break; case 0: break; default: /* Permanent failure */ nr_failed++; break; } } } rc = 0; out: |
290408d4a
|
959 960 961 962 963 |
if (rc) return rc; return nr_failed + retry; } |
742755a1d
|
964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 |
#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; |
6484eb3e2
|
987 |
return alloc_pages_exact_node(pm->node, |
769848c03
|
988 |
GFP_HIGHUSER_MOVABLE | GFP_THISNODE, 0); |
742755a1d
|
989 990 991 992 993 994 |
} /* * 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
|
995 |
* The pm array ends with node = MAX_NUMNODES. |
742755a1d
|
996 |
*/ |
5e9a0f023
|
997 998 999 |
static int do_move_page_to_node_array(struct mm_struct *mm, struct page_to_node *pm, int migrate_all) |
742755a1d
|
1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 |
{ int err; struct page_to_node *pp; LIST_HEAD(pagelist); down_read(&mm->mmap_sem); /* * Build a list of pages to migrate */ |
742755a1d
|
1010 1011 1012 |
for (pp = pm; pp->node != MAX_NUMNODES; pp++) { struct vm_area_struct *vma; struct page *page; |
742755a1d
|
1013 1014 |
err = -EFAULT; vma = find_vma(mm, pp->addr); |
70384dc6d
|
1015 |
if (!vma || pp->addr < vma->vm_start || !vma_migratable(vma)) |
742755a1d
|
1016 |
goto set_status; |
500d65d47
|
1017 |
page = follow_page(vma, pp->addr, FOLL_GET|FOLL_SPLIT); |
89f5b7da2
|
1018 1019 1020 1021 |
err = PTR_ERR(page); if (IS_ERR(page)) goto set_status; |
742755a1d
|
1022 1023 1024 |
err = -ENOENT; if (!page) goto set_status; |
62b61f611
|
1025 1026 |
/* Use PageReserved to check for zero page */ if (PageReserved(page) || PageKsm(page)) |
742755a1d
|
1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 |
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
|
1042 |
err = isolate_lru_page(page); |
6d9c285a6
|
1043 |
if (!err) { |
62695a84e
|
1044 |
list_add_tail(&page->lru, &pagelist); |
6d9c285a6
|
1045 1046 1047 |
inc_zone_page_state(page, NR_ISOLATED_ANON + page_is_file_cache(page)); } |
742755a1d
|
1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 |
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
|
1058 |
err = 0; |
cf608ac19
|
1059 |
if (!list_empty(&pagelist)) { |
742755a1d
|
1060 |
err = migrate_pages(&pagelist, new_page_node, |
77f1fe6b0
|
1061 |
(unsigned long)pm, 0, true); |
cf608ac19
|
1062 1063 1064 |
if (err) putback_lru_pages(&pagelist); } |
742755a1d
|
1065 1066 1067 1068 1069 1070 |
up_read(&mm->mmap_sem); return err; } /* |
5e9a0f023
|
1071 1072 1073 1074 1075 1076 1077 1078 1079 |
* 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
|
1080 |
struct page_to_node *pm; |
5e9a0f023
|
1081 |
nodemask_t task_nodes; |
3140a2273
|
1082 1083 1084 |
unsigned long chunk_nr_pages; unsigned long chunk_start; int err; |
5e9a0f023
|
1085 1086 |
task_nodes = cpuset_mems_allowed(task); |
3140a2273
|
1087 1088 1089 |
err = -ENOMEM; pm = (struct page_to_node *)__get_free_page(GFP_KERNEL); if (!pm) |
5e9a0f023
|
1090 |
goto out; |
35282a2de
|
1091 1092 |
migrate_prep(); |
5e9a0f023
|
1093 |
/* |
3140a2273
|
1094 1095 |
* Store a chunk of page_to_node array in a page, * but keep the last one as a marker |
5e9a0f023
|
1096 |
*/ |
3140a2273
|
1097 |
chunk_nr_pages = (PAGE_SIZE / sizeof(struct page_to_node)) - 1; |
5e9a0f023
|
1098 |
|
3140a2273
|
1099 1100 1101 1102 |
for (chunk_start = 0; chunk_start < nr_pages; chunk_start += chunk_nr_pages) { int j; |
5e9a0f023
|
1103 |
|
3140a2273
|
1104 1105 1106 1107 1108 1109 |
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
|
1110 |
int node; |
3140a2273
|
1111 1112 1113 1114 1115 1116 |
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
|
1117 1118 1119 |
goto out_pm; err = -ENODEV; |
6f5a55f1a
|
1120 1121 |
if (node < 0 || node >= MAX_NUMNODES) goto out_pm; |
5e9a0f023
|
1122 1123 1124 1125 1126 1127 |
if (!node_state(node, N_HIGH_MEMORY)) goto out_pm; err = -EACCES; if (!node_isset(node, task_nodes)) goto out_pm; |
3140a2273
|
1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 |
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
|
1139 |
|
5e9a0f023
|
1140 |
/* Return status information */ |
3140a2273
|
1141 1142 |
for (j = 0; j < chunk_nr_pages; j++) if (put_user(pm[j].status, status + j + chunk_start)) { |
5e9a0f023
|
1143 |
err = -EFAULT; |
3140a2273
|
1144 1145 1146 1147 |
goto out_pm; } } err = 0; |
5e9a0f023
|
1148 1149 |
out_pm: |
3140a2273
|
1150 |
free_page((unsigned long)pm); |
5e9a0f023
|
1151 1152 1153 1154 1155 |
out: return err; } /* |
2f007e74b
|
1156 |
* Determine the nodes of an array of pages and store it in an array of status. |
742755a1d
|
1157 |
*/ |
80bba1290
|
1158 1159 |
static void do_pages_stat_array(struct mm_struct *mm, unsigned long nr_pages, const void __user **pages, int *status) |
742755a1d
|
1160 |
{ |
2f007e74b
|
1161 |
unsigned long i; |
2f007e74b
|
1162 |
|
742755a1d
|
1163 |
down_read(&mm->mmap_sem); |
2f007e74b
|
1164 |
for (i = 0; i < nr_pages; i++) { |
80bba1290
|
1165 |
unsigned long addr = (unsigned long)(*pages); |
742755a1d
|
1166 1167 |
struct vm_area_struct *vma; struct page *page; |
c095adbc2
|
1168 |
int err = -EFAULT; |
2f007e74b
|
1169 1170 |
vma = find_vma(mm, addr); |
70384dc6d
|
1171 |
if (!vma || addr < vma->vm_start) |
742755a1d
|
1172 |
goto set_status; |
2f007e74b
|
1173 |
page = follow_page(vma, addr, 0); |
89f5b7da2
|
1174 1175 1176 1177 |
err = PTR_ERR(page); if (IS_ERR(page)) goto set_status; |
742755a1d
|
1178 1179 |
err = -ENOENT; /* Use PageReserved to check for zero page */ |
62b61f611
|
1180 |
if (!page || PageReserved(page) || PageKsm(page)) |
742755a1d
|
1181 1182 1183 1184 |
goto set_status; err = page_to_nid(page); set_status: |
80bba1290
|
1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 |
*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]; |
80bba1290
|
1205 |
|
87b8d1ade
|
1206 1207 |
while (nr_pages) { unsigned long chunk_nr; |
80bba1290
|
1208 |
|
87b8d1ade
|
1209 1210 1211 1212 1213 1214 |
chunk_nr = nr_pages; if (chunk_nr > DO_PAGES_STAT_CHUNK_NR) chunk_nr = DO_PAGES_STAT_CHUNK_NR; if (copy_from_user(chunk_pages, pages, chunk_nr * sizeof(*chunk_pages))) break; |
80bba1290
|
1215 1216 |
do_pages_stat_array(mm, chunk_nr, chunk_pages, chunk_status); |
87b8d1ade
|
1217 1218 |
if (copy_to_user(status, chunk_status, chunk_nr * sizeof(*status))) break; |
742755a1d
|
1219 |
|
87b8d1ade
|
1220 1221 1222 1223 1224 |
pages += chunk_nr; status += chunk_nr; nr_pages -= chunk_nr; } return nr_pages ? -EFAULT : 0; |
742755a1d
|
1225 1226 1227 1228 1229 1230 |
} /* * Move a list of pages in the address space of the currently executing * process. */ |
938bb9f5e
|
1231 1232 1233 1234 |
SYSCALL_DEFINE6(move_pages, pid_t, pid, unsigned long, nr_pages, const void __user * __user *, pages, const int __user *, nodes, int __user *, status, int, flags) |
742755a1d
|
1235 |
{ |
c69e8d9c0
|
1236 |
const struct cred *cred = current_cred(), *tcred; |
742755a1d
|
1237 |
struct task_struct *task; |
742755a1d
|
1238 |
struct mm_struct *mm; |
5e9a0f023
|
1239 |
int err; |
742755a1d
|
1240 1241 1242 1243 1244 1245 1246 1247 1248 |
/* 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 */ |
a879bf582
|
1249 |
rcu_read_lock(); |
228ebcbe6
|
1250 |
task = pid ? find_task_by_vpid(pid) : current; |
742755a1d
|
1251 |
if (!task) { |
a879bf582
|
1252 |
rcu_read_unlock(); |
742755a1d
|
1253 1254 1255 |
return -ESRCH; } mm = get_task_mm(task); |
a879bf582
|
1256 |
rcu_read_unlock(); |
742755a1d
|
1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 |
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
|
1267 1268 |
rcu_read_lock(); tcred = __task_cred(task); |
b6dff3ec5
|
1269 1270 |
if (cred->euid != tcred->suid && cred->euid != tcred->uid && cred->uid != tcred->suid && cred->uid != tcred->uid && |
742755a1d
|
1271 |
!capable(CAP_SYS_NICE)) { |
c69e8d9c0
|
1272 |
rcu_read_unlock(); |
742755a1d
|
1273 |
err = -EPERM; |
5e9a0f023
|
1274 |
goto out; |
742755a1d
|
1275 |
} |
c69e8d9c0
|
1276 |
rcu_read_unlock(); |
742755a1d
|
1277 |
|
86c3a7645
|
1278 1279 |
err = security_task_movememory(task); if (err) |
5e9a0f023
|
1280 |
goto out; |
86c3a7645
|
1281 |
|
5e9a0f023
|
1282 1283 1284 1285 |
if (nodes) { err = do_pages_move(mm, task, nr_pages, pages, nodes, status, flags); } else { |
2f007e74b
|
1286 |
err = do_pages_stat(mm, nr_pages, pages, status); |
742755a1d
|
1287 |
} |
742755a1d
|
1288 |
out: |
742755a1d
|
1289 1290 1291 |
mmput(mm); return err; } |
742755a1d
|
1292 |
|
7b2259b3e
|
1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 |
/* * 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; |
1001c9fb8
|
1303 |
for (vma = mm->mmap; vma && !err; vma = vma->vm_next) { |
7b2259b3e
|
1304 1305 1306 1307 1308 1309 1310 1311 |
if (vma->vm_ops && vma->vm_ops->migrate) { err = vma->vm_ops->migrate(vma, to, from, flags); if (err) break; } } return err; } |
83d1674a9
|
1312 |
#endif |