migrate.c 26.3 KB
Newer Older
C
Christoph Lameter 已提交
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>
C
Christoph Lameter 已提交
12
 * Christoph Lameter
C
Christoph Lameter 已提交
13 14 15 16 17
 */

#include <linux/migrate.h>
#include <linux/module.h>
#include <linux/swap.h>
18
#include <linux/swapops.h>
C
Christoph Lameter 已提交
19
#include <linux/pagemap.h>
20
#include <linux/buffer_head.h>
C
Christoph Lameter 已提交
21
#include <linux/mm_inline.h>
22
#include <linux/nsproxy.h>
C
Christoph Lameter 已提交
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>
28
#include <linux/writeback.h>
29 30
#include <linux/mempolicy.h>
#include <linux/vmalloc.h>
31
#include <linux/security.h>
32
#include <linux/memcontrol.h>
33
#include <linux/syscalls.h>
C
Christoph Lameter 已提交
34 35 36 37 38 39

#include "internal.h"

#define lru_to_page(_head) (list_entry((_head)->prev, struct page, lru))

/*
40 41
 * migrate_prep() needs to be called before we start compiling a list of pages
 * to be migrated using isolate_lru_page().
C
Christoph Lameter 已提交
42 43 44 45 46 47 48 49 50 51 52 53 54 55 56
 */
int migrate_prep(void)
{
	/*
	 * 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;
}

/*
L
Lee Schermerhorn 已提交
57 58
 * Add isolated pages on the list back to the LRU under page lock
 * to avoid leaking evictable pages back onto unevictable list.
C
Christoph Lameter 已提交
59 60 61 62 63 64 65 66 67 68
 *
 * 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) {
69
		list_del(&page->lru);
K
KOSAKI Motohiro 已提交
70
		dec_zone_page_state(page, NR_ISOLATED_ANON +
71
				page_is_file_cache(page));
L
Lee Schermerhorn 已提交
72
		putback_lru_page(page);
C
Christoph Lameter 已提交
73 74 75 76 77
		count++;
	}
	return count;
}

78 79 80
/*
 * Restore a potential migration pte to a working pte entry
 */
81
static void remove_migration_pte(struct vm_area_struct *vma,
82 83 84 85 86 87 88 89 90
		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;
91 92 93 94
	unsigned long addr = page_address_in_vma(new, vma);

	if (addr == -EFAULT)
		return;
95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129

 	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;

	get_page(new);
	pte = pte_mkold(mk_pte(new, vma->vm_page_prot));
	if (is_write_migration_entry(entry))
		pte = pte_mkwrite(pte);
130
	flush_cache_page(vma, addr, pte_pfn(pte));
131
	set_pte_at(mm, addr, ptep, pte);
132 133 134 135 136 137 138 139 140

	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);

141 142 143 144 145
out:
	pte_unmap_unlock(ptep, ptl);
}

/*
146 147 148 149 150 151
 * 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;
152
	struct address_space *mapping = new->mapping;
153 154 155 156 157 158 159 160 161 162 163 164 165 166 167
	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);
}

/*
168 169 170
 * Must hold mmap_sem lock on at least one of the vmas containing
 * the page so that the anon_vma cannot vanish.
 */
171
static void remove_anon_migration_ptes(struct page *old, struct page *new)
172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188
{
	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)
189
		remove_migration_pte(vma, old, new);
190 191 192 193

	spin_unlock(&anon_vma->lock);
}

194 195 196 197 198 199 200 201 202 203 204 205
/*
 * 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);
}

206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231
/*
 * 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);

N
Nick Piggin 已提交
232 233 234 235 236 237 238 239 240
	/*
	 * 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;
241 242 243 244 245 246 247 248
	pte_unmap_unlock(ptep, ptl);
	wait_on_page_locked(page);
	put_page(page);
	return;
out:
	pte_unmap_unlock(ptep, ptl);
}

C
Christoph Lameter 已提交
249
/*
250
 * Replace the page in the mapping.
251 252 253 254
 *
 * The number of remaining references must be:
 * 1 for anonymous pages without a mapping
 * 2 for pages with a mapping
255
 * 3 for pages with a mapping and PagePrivate/PagePrivate2 set.
C
Christoph Lameter 已提交
256
 */
257 258
static int migrate_page_move_mapping(struct address_space *mapping,
		struct page *newpage, struct page *page)
C
Christoph Lameter 已提交
259
{
N
Nick Piggin 已提交
260
	int expected_count;
261
	void **pslot;
C
Christoph Lameter 已提交
262

263
	if (!mapping) {
264
		/* Anonymous page without mapping */
265 266 267 268 269
		if (page_count(page) != 1)
			return -EAGAIN;
		return 0;
	}

N
Nick Piggin 已提交
270
	spin_lock_irq(&mapping->tree_lock);
C
Christoph Lameter 已提交
271

272 273
	pslot = radix_tree_lookup_slot(&mapping->page_tree,
 					page_index(page));
C
Christoph Lameter 已提交
274

275
	expected_count = 2 + !!page_has_private(page);
N
Nick Piggin 已提交
276
	if (page_count(page) != expected_count ||
277
			(struct page *)radix_tree_deref_slot(pslot) != page) {
N
Nick Piggin 已提交
278
		spin_unlock_irq(&mapping->tree_lock);
279
		return -EAGAIN;
C
Christoph Lameter 已提交
280 281
	}

N
Nick Piggin 已提交
282
	if (!page_freeze_refs(page, expected_count)) {
N
Nick Piggin 已提交
283
		spin_unlock_irq(&mapping->tree_lock);
N
Nick Piggin 已提交
284 285 286
		return -EAGAIN;
	}

C
Christoph Lameter 已提交
287 288 289
	/*
	 * Now we know that no one else is looking at the page.
	 */
290
	get_page(newpage);	/* add cache reference */
C
Christoph Lameter 已提交
291 292 293 294 295
	if (PageSwapCache(page)) {
		SetPageSwapCache(newpage);
		set_page_private(newpage, page_private(page));
	}

296 297
	radix_tree_replace_slot(pslot, newpage);

N
Nick Piggin 已提交
298
	page_unfreeze_refs(page, expected_count);
299 300 301 302
	/*
	 * Drop cache reference from old page.
	 * We know this isn't the last reference.
	 */
C
Christoph Lameter 已提交
303
	__put_page(page);
304

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);
317 318 319 320
	if (PageSwapBacked(page)) {
		__dec_zone_page_state(page, NR_SHMEM);
		__inc_zone_page_state(newpage, NR_SHMEM);
	}
N
Nick Piggin 已提交
321
	spin_unlock_irq(&mapping->tree_lock);
C
Christoph Lameter 已提交
322 323 324 325 326 327 328

	return 0;
}

/*
 * Copy the page to its new location
 */
329
static void migrate_page_copy(struct page *newpage, struct page *page)
C
Christoph Lameter 已提交
330
{
331 332
	int anon;

C
Christoph Lameter 已提交
333 334 335 336 337 338 339 340
	copy_highpage(newpage, page);

	if (PageError(page))
		SetPageError(newpage);
	if (PageReferenced(page))
		SetPageReferenced(newpage);
	if (PageUptodate(page))
		SetPageUptodate(newpage);
L
Lee Schermerhorn 已提交
341 342
	if (TestClearPageActive(page)) {
		VM_BUG_ON(PageUnevictable(page));
C
Christoph Lameter 已提交
343
		SetPageActive(newpage);
L
Lee Schermerhorn 已提交
344 345
	} else
		unevictable_migrate_page(newpage, page);
C
Christoph Lameter 已提交
346 347 348 349 350 351 352
	if (PageChecked(page))
		SetPageChecked(newpage);
	if (PageMappedToDisk(page))
		SetPageMappedToDisk(newpage);

	if (PageDirty(page)) {
		clear_page_dirty_for_io(page);
N
Nick Piggin 已提交
353 354 355 356 357 358 359 360
		/*
		 * 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);
C
Christoph Lameter 已提交
361 362
 	}

N
Nick Piggin 已提交
363 364
	mlock_migrate_page(newpage, page);

C
Christoph Lameter 已提交
365 366 367
	ClearPageSwapCache(page);
	ClearPagePrivate(page);
	set_page_private(page, 0);
368 369
	/* page->mapping contains a flag for PageAnon() */
	anon = PageAnon(page);
C
Christoph Lameter 已提交
370 371 372 373 374 375 376 377 378 379
	page->mapping = NULL;

	/*
	 * If any waiters have accumulated on the new page then
	 * wake them up.
	 */
	if (PageWriteback(newpage))
		end_page_writeback(newpage);
}

380 381 382 383 384
/************************************************************
 *                    Migration functions
 ***********************************************************/

/* Always fail migration. Used for mappings that are not movable */
385 386
int fail_migrate_page(struct address_space *mapping,
			struct page *newpage, struct page *page)
387 388 389 390 391
{
	return -EIO;
}
EXPORT_SYMBOL(fail_migrate_page);

C
Christoph Lameter 已提交
392 393
/*
 * Common logic to directly migrate a single page suitable for
394
 * pages that do not use PagePrivate/PagePrivate2.
C
Christoph Lameter 已提交
395 396 397
 *
 * Pages are locked upon entry and exit.
 */
398 399
int migrate_page(struct address_space *mapping,
		struct page *newpage, struct page *page)
C
Christoph Lameter 已提交
400 401 402 403 404
{
	int rc;

	BUG_ON(PageWriteback(page));	/* Writeback must be complete */

405
	rc = migrate_page_move_mapping(mapping, newpage, page);
C
Christoph Lameter 已提交
406 407 408 409 410 411 412 413 414

	if (rc)
		return rc;

	migrate_page_copy(newpage, page);
	return 0;
}
EXPORT_SYMBOL(migrate_page);

415
#ifdef CONFIG_BLOCK
416 417 418 419 420
/*
 * Migration function for pages with buffers. This function can only be used
 * if the underlying filesystem guarantees that no other references to "page"
 * exist.
 */
421 422
int buffer_migrate_page(struct address_space *mapping,
		struct page *newpage, struct page *page)
423 424 425 426 427
{
	struct buffer_head *bh, *head;
	int rc;

	if (!page_has_buffers(page))
428
		return migrate_page(mapping, newpage, page);
429 430 431

	head = page_buffers(page);

432
	rc = migrate_page_move_mapping(mapping, newpage, page);
433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472

	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);
473
#endif
474

475 476 477 478
/*
 * Writeback a page to clean the dirty state
 */
static int writeout(struct address_space *mapping, struct page *page)
479
{
480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497
	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;

498
	/*
499 500 501 502 503 504
	 * 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.
505
	 */
506
	remove_migration_ptes(page, page);
507

508
	rc = mapping->a_ops->writepage(page, &wbc);
509

510 511 512 513
	if (rc != AOP_WRITEPAGE_ACTIVATE)
		/* unlocked. Relock */
		lock_page(page);

H
Hugh Dickins 已提交
514
	return (rc < 0) ? -EIO : -EAGAIN;
515 516 517 518 519 520 521 522 523 524
}

/*
 * 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);
525 526 527 528 529

	/*
	 * Buffers may be managed in a filesystem specific way.
	 * We must have no buffers or drop them.
	 */
530
	if (page_has_private(page) &&
531 532 533 534 535 536
	    !try_to_release_page(page, GFP_KERNEL))
		return -EAGAIN;

	return migrate_page(mapping, newpage, page);
}

537 538 539 540 541 542
/*
 * 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.
L
Lee Schermerhorn 已提交
543 544 545 546
 *
 * Return value:
 *   < 0 - error code
 *  == 0 - success
547 548 549 550 551 552 553 554 555 556 557
 */
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.
	 */
N
Nick Piggin 已提交
558
	if (!trylock_page(newpage))
559 560 561 562 563
		BUG();

	/* Prepare mapping for the new page.*/
	newpage->index = page->index;
	newpage->mapping = page->mapping;
R
Rik van Riel 已提交
564 565
	if (PageSwapBacked(page))
		SetPageSwapBacked(newpage);
566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582

	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);

583
	if (!rc) {
584
		remove_migration_ptes(page, newpage);
585
	} else
586 587 588 589 590 591 592 593 594 595 596
		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.
 */
597 598
static int unmap_and_move(new_page_t get_new_page, unsigned long private,
			struct page *page, int force)
599 600
{
	int rc = 0;
601 602
	int *result = NULL;
	struct page *newpage = get_new_page(page, private, &result);
603
	int rcu_locked = 0;
604
	int charge = 0;
605
	struct mem_cgroup *mem;
606 607 608

	if (!newpage)
		return -ENOMEM;
609

L
Lee Schermerhorn 已提交
610
	if (page_count(page) == 1) {
611
		/* page was freed from under us. So we are done. */
612
		goto move_newpage;
L
Lee Schermerhorn 已提交
613
	}
614

615
	/* prepare cgroup just returns 0 or -ENOMEM */
616
	rc = -EAGAIN;
617

N
Nick Piggin 已提交
618
	if (!trylock_page(page)) {
619
		if (!force)
620
			goto move_newpage;
621 622 623
		lock_page(page);
	}

624 625 626 627 628 629 630 631
	/* charge against new page */
	charge = mem_cgroup_prepare_migration(page, &mem);
	if (charge == -ENOMEM) {
		rc = -ENOMEM;
		goto unlock;
	}
	BUG_ON(charge);

632 633
	if (PageWriteback(page)) {
		if (!force)
634
			goto uncharge;
635 636 637
		wait_on_page_writeback(page);
	}
	/*
638 639 640 641
	 * 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()
642 643
	 * File Caches may use write_page() or lock_page() in migration, then,
	 * just care Anon page here.
644
	 */
645 646 647 648
	if (PageAnon(page)) {
		rcu_read_lock();
		rcu_locked = 1;
	}
649

650
	/*
651 652 653 654 655 656 657 658 659 660
	 * 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.
661
	 */
662
	if (!page->mapping) {
663
		if (!PageAnon(page) && page_has_private(page)) {
664 665 666 667 668 669 670 671
			/*
			 * 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);
672
			goto rcu_unlock;
673
		}
674
		goto skip_unmap;
675 676
	}

677
	/* Establish migration ptes or remove ptes */
678
	try_to_unmap(page, 1);
679

680
skip_unmap:
681 682
	if (!page_mapped(page))
		rc = move_to_new_page(newpage, page);
683

684
	if (rc)
685
		remove_migration_ptes(page, page);
686
rcu_unlock:
687 688
	if (rcu_locked)
		rcu_read_unlock();
689 690 691
uncharge:
	if (!charge)
		mem_cgroup_end_migration(mem, page, newpage);
692 693
unlock:
	unlock_page(page);
694

695
	if (rc != -EAGAIN) {
696 697 698 699 700 701 702
 		/*
 		 * 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);
K
KOSAKI Motohiro 已提交
703
		dec_zone_page_state(page, NR_ISOLATED_ANON +
704
				page_is_file_cache(page));
L
Lee Schermerhorn 已提交
705
		putback_lru_page(page);
706
	}
707 708

move_newpage:
L
Lee Schermerhorn 已提交
709

710 711 712 713
	/*
	 * Move the new page to the LRU. If migration was not successful
	 * then this will free the page.
	 */
L
Lee Schermerhorn 已提交
714 715
	putback_lru_page(newpage);

716 717 718 719 720 721
	if (result) {
		if (rc)
			*result = rc;
		else
			*result = page_to_nid(newpage);
	}
722 723 724
	return rc;
}

C
Christoph Lameter 已提交
725 726 727
/*
 * migrate_pages
 *
728 729 730
 * 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.
C
Christoph Lameter 已提交
731 732 733
 *
 * The function returns after 10 attempts or if no pages
 * are movable anymore because to has become empty
734
 * or no retryable pages exist anymore. All pages will be
G
Gabriel Craciunescu 已提交
735
 * returned to the LRU or freed.
C
Christoph Lameter 已提交
736
 *
737
 * Return: Number of pages not migrated or error code.
C
Christoph Lameter 已提交
738
 */
739 740
int migrate_pages(struct list_head *from,
		new_page_t get_new_page, unsigned long private)
C
Christoph Lameter 已提交
741
{
742
	int retry = 1;
C
Christoph Lameter 已提交
743 744 745 746 747 748
	int nr_failed = 0;
	int pass = 0;
	struct page *page;
	struct page *page2;
	int swapwrite = current->flags & PF_SWAPWRITE;
	int rc;
K
KOSAKI Motohiro 已提交
749 750 751 752 753
	unsigned long flags;

	local_irq_save(flags);
	list_for_each_entry(page, from, lru)
		__inc_zone_page_state(page, NR_ISOLATED_ANON +
754
				page_is_file_cache(page));
K
KOSAKI Motohiro 已提交
755
	local_irq_restore(flags);
C
Christoph Lameter 已提交
756 757 758 759

	if (!swapwrite)
		current->flags |= PF_SWAPWRITE;

760 761
	for(pass = 0; pass < 10 && retry; pass++) {
		retry = 0;
C
Christoph Lameter 已提交
762

763 764
		list_for_each_entry_safe(page, page2, from, lru) {
			cond_resched();
765

766 767
			rc = unmap_and_move(get_new_page, private,
						page, pass > 2);
768

769
			switch(rc) {
770 771
			case -ENOMEM:
				goto out;
772
			case -EAGAIN:
773
				retry++;
774 775 776 777
				break;
			case 0:
				break;
			default:
778 779
				/* Permanent failure */
				nr_failed++;
780
				break;
781
			}
C
Christoph Lameter 已提交
782 783
		}
	}
784 785
	rc = 0;
out:
C
Christoph Lameter 已提交
786 787 788
	if (!swapwrite)
		current->flags &= ~PF_SWAPWRITE;

789
	putback_lru_pages(from);
C
Christoph Lameter 已提交
790

791 792
	if (rc)
		return rc;
C
Christoph Lameter 已提交
793

794
	return nr_failed + retry;
C
Christoph Lameter 已提交
795
}
796

797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820
#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;

821
	return alloc_pages_exact_node(pm->node,
822
				GFP_HIGHUSER_MOVABLE | GFP_THISNODE, 0);
823 824 825 826 827 828
}

/*
 * 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.
829
 * The pm array ends with node = MAX_NUMNODES.
830
 */
831 832 833
static int do_move_page_to_node_array(struct mm_struct *mm,
				      struct page_to_node *pm,
				      int migrate_all)
834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849
{
	int err;
	struct page_to_node *pp;
	LIST_HEAD(pagelist);

	down_read(&mm->mmap_sem);

	/*
	 * Build a list of pages to migrate
	 */
	for (pp = pm; pp->node != MAX_NUMNODES; pp++) {
		struct vm_area_struct *vma;
		struct page *page;

		err = -EFAULT;
		vma = find_vma(mm, pp->addr);
850
		if (!vma || !vma_migratable(vma))
851 852 853
			goto set_status;

		page = follow_page(vma, pp->addr, FOLL_GET);
854 855 856 857 858

		err = PTR_ERR(page);
		if (IS_ERR(page))
			goto set_status;

859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879
		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;

880 881 882
		err = isolate_lru_page(page);
		if (!err)
			list_add_tail(&page->lru, &pagelist);
883 884 885 886 887 888 889 890 891 892 893
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;
	}

894
	err = 0;
895 896 897 898 899 900 901 902
	if (!list_empty(&pagelist))
		err = migrate_pages(&pagelist, new_page_node,
				(unsigned long)pm);

	up_read(&mm->mmap_sem);
	return err;
}

903 904 905 906 907 908 909 910 911 912
/*
 * 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)
{
913
	struct page_to_node *pm;
914
	nodemask_t task_nodes;
915 916 917
	unsigned long chunk_nr_pages;
	unsigned long chunk_start;
	int err;
918 919 920

	task_nodes = cpuset_mems_allowed(task);

921 922 923
	err = -ENOMEM;
	pm = (struct page_to_node *)__get_free_page(GFP_KERNEL);
	if (!pm)
924
		goto out;
925 926 927

	migrate_prep();

928
	/*
929 930
	 * Store a chunk of page_to_node array in a page,
	 * but keep the last one as a marker
931
	 */
932
	chunk_nr_pages = (PAGE_SIZE / sizeof(struct page_to_node)) - 1;
933

934 935 936 937
	for (chunk_start = 0;
	     chunk_start < nr_pages;
	     chunk_start += chunk_nr_pages) {
		int j;
938

939 940 941 942 943 944
		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;
945 946
			int node;

947 948 949 950 951 952
			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))
953 954 955 956 957 958 959 960 961 962
				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;

963 964 965 966 967 968 969 970 971 972 973
			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;
974 975

		/* Return status information */
976 977
		for (j = 0; j < chunk_nr_pages; j++)
			if (put_user(pm[j].status, status + j + chunk_start)) {
978
				err = -EFAULT;
979 980 981 982
				goto out_pm;
			}
	}
	err = 0;
983 984

out_pm:
985
	free_page((unsigned long)pm);
986 987 988 989
out:
	return err;
}

990
/*
991
 * Determine the nodes of an array of pages and store it in an array of status.
992
 */
993 994
static void do_pages_stat_array(struct mm_struct *mm, unsigned long nr_pages,
				const void __user **pages, int *status)
995
{
996 997
	unsigned long i;

998 999
	down_read(&mm->mmap_sem);

1000
	for (i = 0; i < nr_pages; i++) {
1001
		unsigned long addr = (unsigned long)(*pages);
1002 1003
		struct vm_area_struct *vma;
		struct page *page;
1004
		int err = -EFAULT;
1005 1006

		vma = find_vma(mm, addr);
1007 1008 1009
		if (!vma)
			goto set_status;

1010
		page = follow_page(vma, addr, 0);
1011 1012 1013 1014 1015

		err = PTR_ERR(page);
		if (IS_ERR(page))
			goto set_status;

1016 1017 1018 1019 1020 1021 1022
		err = -ENOENT;
		/* Use PageReserved to check for zero page */
		if (!page || PageReserved(page))
			goto set_status;

		err = page_to_nid(page);
set_status:
1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064
		*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;
		}
1065
	}
1066
	err = 0;
1067

1068 1069
out:
	return err;
1070 1071 1072 1073 1074 1075
}

/*
 * Move a list of pages in the address space of the currently executing
 * process.
 */
1076 1077 1078 1079
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)
1080
{
1081
	const struct cred *cred = current_cred(), *tcred;
1082 1083
	struct task_struct *task;
	struct mm_struct *mm;
1084
	int err;
1085 1086 1087 1088 1089 1090 1091 1092 1093 1094

	/* 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);
1095
	task = pid ? find_task_by_vpid(pid) : current;
1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111
	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.
	 */
1112 1113
	rcu_read_lock();
	tcred = __task_cred(task);
1114 1115
	if (cred->euid != tcred->suid && cred->euid != tcred->uid &&
	    cred->uid  != tcred->suid && cred->uid  != tcred->uid &&
1116
	    !capable(CAP_SYS_NICE)) {
1117
		rcu_read_unlock();
1118
		err = -EPERM;
1119
		goto out;
1120
	}
1121
	rcu_read_unlock();
1122

1123 1124
 	err = security_task_movememory(task);
 	if (err)
1125
		goto out;
1126

1127 1128 1129 1130
	if (nodes) {
		err = do_pages_move(mm, task, nr_pages, pages, nodes, status,
				    flags);
	} else {
1131
		err = do_pages_stat(mm, nr_pages, pages, status);
1132 1133 1134 1135 1136 1137 1138
	}

out:
	mmput(mm);
	return err;
}

1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149
/*
 * 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;

1150
	for (vma = mm->mmap; vma && !err; vma = vma->vm_next) {
1151 1152 1153 1154 1155 1156 1157 1158
 		if (vma->vm_ops && vma->vm_ops->migrate) {
 			err = vma->vm_ops->migrate(vma, to, from, flags);
 			if (err)
 				break;
 		}
 	}
 	return err;
}
1159
#endif