truncate.c 17.9 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5
/*
 * mm/truncate.c - code for taking down pages from address_spaces
 *
 * Copyright (C) 2002, Linus Torvalds
 *
6
 * 10Sep2002	Andrew Morton
L
Linus Torvalds 已提交
7 8 9 10
 *		Initial version.
 */

#include <linux/kernel.h>
A
Alexey Dobriyan 已提交
11
#include <linux/backing-dev.h>
12
#include <linux/gfp.h>
L
Linus Torvalds 已提交
13
#include <linux/mm.h>
N
Nick Piggin 已提交
14
#include <linux/swap.h>
15
#include <linux/export.h>
L
Linus Torvalds 已提交
16
#include <linux/pagemap.h>
17
#include <linux/highmem.h>
L
Linus Torvalds 已提交
18
#include <linux/pagevec.h>
19
#include <linux/task_io_accounting_ops.h>
L
Linus Torvalds 已提交
20
#include <linux/buffer_head.h>	/* grr. try_to_release_page,
21
				   do_invalidatepage */
22
#include <linux/cleancache.h>
23
#include "internal.h"
L
Linus Torvalds 已提交
24 25


26
/**
27
 * do_invalidatepage - invalidate part or all of a page
28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43
 * @page: the page which is affected
 * @offset: the index of the truncation point
 *
 * do_invalidatepage() is called when all or part of the page has become
 * invalidated by a truncate operation.
 *
 * do_invalidatepage() does not have to release all buffers, but it must
 * ensure that no dirty buffer is left outside @offset and that no I/O
 * is underway against any of the blocks which are outside the truncation
 * point.  Because the caller is about to free (and possibly reuse) those
 * blocks on-disk.
 */
void do_invalidatepage(struct page *page, unsigned long offset)
{
	void (*invalidatepage)(struct page *, unsigned long);
	invalidatepage = page->mapping->a_ops->invalidatepage;
44
#ifdef CONFIG_BLOCK
45 46
	if (!invalidatepage)
		invalidatepage = block_invalidatepage;
47
#endif
48 49 50 51
	if (invalidatepage)
		(*invalidatepage)(page, offset);
}

L
Linus Torvalds 已提交
52 53
static inline void truncate_partial_page(struct page *page, unsigned partial)
{
54
	zero_user_segment(page, partial, PAGE_CACHE_SIZE);
55
	cleancache_flush_page(page->mapping, page);
56
	if (page_has_private(page))
L
Linus Torvalds 已提交
57 58 59
		do_invalidatepage(page, partial);
}

60 61 62 63 64 65 66 67 68 69 70 71 72 73
/*
 * This cancels just the dirty bit on the kernel page itself, it
 * does NOT actually remove dirty bits on any mmap's that may be
 * around. It also leaves the page tagged dirty, so any sync
 * activity will still find it on the dirty lists, and in particular,
 * clear_page_dirty_for_io() will still look at the dirty bits in
 * the VM.
 *
 * Doing this should *normally* only ever be done when a page
 * is truncated, and is not actually mapped anywhere at all. However,
 * fs/buffer.c does this when it notices that somebody has cleaned
 * out all the buffers on a page without actually doing it through
 * the VM. Can you say "ext3 is horribly ugly"? Tought you could.
 */
74 75
void cancel_dirty_page(struct page *page, unsigned int account_size)
{
76 77 78 79
	if (TestClearPageDirty(page)) {
		struct address_space *mapping = page->mapping;
		if (mapping && mapping_cap_account_dirty(mapping)) {
			dec_zone_page_state(page, NR_FILE_DIRTY);
80 81
			dec_bdi_stat(mapping->backing_dev_info,
					BDI_RECLAIMABLE);
82 83 84
			if (account_size)
				task_io_account_cancelled_write(account_size);
		}
85
	}
86
}
87
EXPORT_SYMBOL(cancel_dirty_page);
88

L
Linus Torvalds 已提交
89 90
/*
 * If truncate cannot remove the fs-private metadata from the page, the page
91
 * becomes orphaned.  It will be left on the LRU and may even be mapped into
92
 * user pagetables if we're racing with filemap_fault().
L
Linus Torvalds 已提交
93 94 95
 *
 * We need to bale out if page->mapping is no longer equal to the original
 * mapping.  This happens a) when the VM reclaimed the page while we waited on
96
 * its lock, b) when a concurrent invalidate_mapping_pages got there first and
L
Linus Torvalds 已提交
97 98
 * c) when tmpfs swizzles a page between a tmpfs inode and swapper_space.
 */
99
static int
L
Linus Torvalds 已提交
100 101 102
truncate_complete_page(struct address_space *mapping, struct page *page)
{
	if (page->mapping != mapping)
103
		return -EIO;
L
Linus Torvalds 已提交
104

105
	if (page_has_private(page))
L
Linus Torvalds 已提交
106 107
		do_invalidatepage(page, 0);

108 109
	cancel_dirty_page(page, PAGE_CACHE_SIZE);

110
	clear_page_mlock(page);
L
Linus Torvalds 已提交
111
	ClearPageMappedToDisk(page);
112
	delete_from_page_cache(page);
113
	return 0;
L
Linus Torvalds 已提交
114 115 116
}

/*
117
 * This is for invalidate_mapping_pages().  That function can be called at
L
Linus Torvalds 已提交
118
 * any time, and is not supposed to throw away dirty pages.  But pages can
N
Nick Piggin 已提交
119 120
 * be marked dirty at any time too, so use remove_mapping which safely
 * discards clean, unused pages.
L
Linus Torvalds 已提交
121 122 123 124 125 126
 *
 * Returns non-zero if the page was successfully invalidated.
 */
static int
invalidate_complete_page(struct address_space *mapping, struct page *page)
{
N
Nick Piggin 已提交
127 128
	int ret;

L
Linus Torvalds 已提交
129 130 131
	if (page->mapping != mapping)
		return 0;

132
	if (page_has_private(page) && !try_to_release_page(page, 0))
L
Linus Torvalds 已提交
133 134
		return 0;

135
	clear_page_mlock(page);
N
Nick Piggin 已提交
136 137 138
	ret = remove_mapping(mapping, page);

	return ret;
L
Linus Torvalds 已提交
139 140
}

141 142 143 144 145 146 147 148 149 150
int truncate_inode_page(struct address_space *mapping, struct page *page)
{
	if (page_mapped(page)) {
		unmap_mapping_range(mapping,
				   (loff_t)page->index << PAGE_CACHE_SHIFT,
				   PAGE_CACHE_SIZE, 0);
	}
	return truncate_complete_page(mapping, page);
}

151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167
/*
 * Used to get rid of pages on hardware memory corruption.
 */
int generic_error_remove_page(struct address_space *mapping, struct page *page)
{
	if (!mapping)
		return -EINVAL;
	/*
	 * Only punch for normal data pages for now.
	 * Handling other types like directories would need more auditing.
	 */
	if (!S_ISREG(mapping->host->i_mode))
		return -EIO;
	return truncate_inode_page(mapping, page);
}
EXPORT_SYMBOL(generic_error_remove_page);

168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185
/*
 * Safely invalidate one page from its pagecache mapping.
 * It only drops clean, unused pages. The page must be locked.
 *
 * Returns 1 if the page is successfully invalidated, otherwise 0.
 */
int invalidate_inode_page(struct page *page)
{
	struct address_space *mapping = page_mapping(page);
	if (!mapping)
		return 0;
	if (PageDirty(page) || PageWriteback(page))
		return 0;
	if (page_mapped(page))
		return 0;
	return invalidate_complete_page(mapping, page);
}

L
Linus Torvalds 已提交
186
/**
187
 * truncate_inode_pages - truncate range of pages specified by start & end byte offsets
L
Linus Torvalds 已提交
188 189
 * @mapping: mapping to truncate
 * @lstart: offset from which to truncate
190
 * @lend: offset to which to truncate
L
Linus Torvalds 已提交
191
 *
192 193 194
 * Truncate the page cache, removing the pages that are between
 * specified offsets (and zeroing out partial page
 * (if lstart is not page aligned)).
L
Linus Torvalds 已提交
195 196 197 198 199 200 201 202 203 204 205
 *
 * Truncate takes two passes - the first pass is nonblocking.  It will not
 * block on page locks and it will not block on writeback.  The second pass
 * will wait.  This is to prevent as much IO as possible in the affected region.
 * The first pass will remove most pages, so the search cost of the second pass
 * is low.
 *
 * We pass down the cache-hot hint to the page freeing code.  Even if the
 * mapping is large, it is probably the case that the final pages are the most
 * recently touched, and freeing happens in ascending file offset order.
 */
206 207
void truncate_inode_pages_range(struct address_space *mapping,
				loff_t lstart, loff_t lend)
L
Linus Torvalds 已提交
208 209 210 211
{
	const pgoff_t start = (lstart + PAGE_CACHE_SIZE-1) >> PAGE_CACHE_SHIFT;
	const unsigned partial = lstart & (PAGE_CACHE_SIZE - 1);
	struct pagevec pvec;
212 213
	pgoff_t index;
	pgoff_t end;
L
Linus Torvalds 已提交
214 215
	int i;

216
	cleancache_flush_inode(mapping);
L
Linus Torvalds 已提交
217 218 219
	if (mapping->nrpages == 0)
		return;

220 221 222
	BUG_ON((lend & (PAGE_CACHE_SIZE - 1)) != (PAGE_CACHE_SIZE - 1));
	end = (lend >> PAGE_CACHE_SHIFT);

L
Linus Torvalds 已提交
223
	pagevec_init(&pvec, 0);
224 225 226
	index = start;
	while (index <= end && pagevec_lookup(&pvec, mapping, index,
			min(end - index, (pgoff_t)PAGEVEC_SIZE - 1) + 1)) {
227
		mem_cgroup_uncharge_start();
L
Linus Torvalds 已提交
228 229 230
		for (i = 0; i < pagevec_count(&pvec); i++) {
			struct page *page = pvec.pages[i];

231 232 233
			/* We rely upon deletion not changing page->index */
			index = page->index;
			if (index > end)
234 235
				break;

N
Nick Piggin 已提交
236
			if (!trylock_page(page))
L
Linus Torvalds 已提交
237
				continue;
238
			WARN_ON(page->index != index);
L
Linus Torvalds 已提交
239 240 241 242
			if (PageWriteback(page)) {
				unlock_page(page);
				continue;
			}
243
			truncate_inode_page(mapping, page);
L
Linus Torvalds 已提交
244 245 246
			unlock_page(page);
		}
		pagevec_release(&pvec);
247
		mem_cgroup_uncharge_end();
L
Linus Torvalds 已提交
248
		cond_resched();
249
		index++;
L
Linus Torvalds 已提交
250 251 252 253 254 255 256 257 258 259 260 261
	}

	if (partial) {
		struct page *page = find_lock_page(mapping, start - 1);
		if (page) {
			wait_on_page_writeback(page);
			truncate_partial_page(page, partial);
			unlock_page(page);
			page_cache_release(page);
		}
	}

262
	index = start;
L
Linus Torvalds 已提交
263 264
	for ( ; ; ) {
		cond_resched();
265 266 267
		if (!pagevec_lookup(&pvec, mapping, index,
			min(end - index, (pgoff_t)PAGEVEC_SIZE - 1) + 1)) {
			if (index == start)
L
Linus Torvalds 已提交
268
				break;
269
			index = start;
L
Linus Torvalds 已提交
270 271
			continue;
		}
272
		if (index == start && pvec.pages[0]->index > end) {
273 274 275
			pagevec_release(&pvec);
			break;
		}
276
		mem_cgroup_uncharge_start();
L
Linus Torvalds 已提交
277 278 279
		for (i = 0; i < pagevec_count(&pvec); i++) {
			struct page *page = pvec.pages[i];

280 281 282
			/* We rely upon deletion not changing page->index */
			index = page->index;
			if (index > end)
283
				break;
284

L
Linus Torvalds 已提交
285
			lock_page(page);
286
			WARN_ON(page->index != index);
L
Linus Torvalds 已提交
287
			wait_on_page_writeback(page);
288
			truncate_inode_page(mapping, page);
L
Linus Torvalds 已提交
289 290 291
			unlock_page(page);
		}
		pagevec_release(&pvec);
292
		mem_cgroup_uncharge_end();
293
		index++;
L
Linus Torvalds 已提交
294
	}
295
	cleancache_flush_inode(mapping);
L
Linus Torvalds 已提交
296
}
297
EXPORT_SYMBOL(truncate_inode_pages_range);
L
Linus Torvalds 已提交
298

299 300 301 302 303
/**
 * truncate_inode_pages - truncate *all* the pages from an offset
 * @mapping: mapping to truncate
 * @lstart: offset from which to truncate
 *
304
 * Called under (and serialised by) inode->i_mutex.
305 306 307 308 309
 *
 * Note: When this function returns, there can be a page in the process of
 * deletion (inside __delete_from_page_cache()) in the specified range.  Thus
 * mapping->nrpages can be non-zero when this function returns even after
 * truncation of the whole mapping.
310 311 312 313 314
 */
void truncate_inode_pages(struct address_space *mapping, loff_t lstart)
{
	truncate_inode_pages_range(mapping, lstart, (loff_t)-1);
}
L
Linus Torvalds 已提交
315 316
EXPORT_SYMBOL(truncate_inode_pages);

317 318 319 320 321 322 323 324 325 326 327 328 329 330
/**
 * invalidate_mapping_pages - Invalidate all the unlocked pages of one inode
 * @mapping: the address_space which holds the pages to invalidate
 * @start: the offset 'from' which to invalidate
 * @end: the offset 'to' which to invalidate (inclusive)
 *
 * This function only removes the unlocked pages, if you want to
 * remove all the pages of one inode, you must call truncate_inode_pages.
 *
 * invalidate_mapping_pages() will not block on IO activity. It will not
 * invalidate pages which are dirty, locked, under writeback or mapped into
 * pagetables.
 */
unsigned long invalidate_mapping_pages(struct address_space *mapping,
M
Minchan Kim 已提交
331
		pgoff_t start, pgoff_t end)
L
Linus Torvalds 已提交
332 333
{
	struct pagevec pvec;
334
	pgoff_t index = start;
M
Minchan Kim 已提交
335 336
	unsigned long ret;
	unsigned long count = 0;
L
Linus Torvalds 已提交
337 338
	int i;

339 340 341 342 343 344 345 346
	/*
	 * Note: this function may get called on a shmem/tmpfs mapping:
	 * pagevec_lookup() might then return 0 prematurely (because it
	 * got a gangful of swap entries); but it's hardly worth worrying
	 * about - it can rarely have anything to free from such a mapping
	 * (most pages are dirty), and already skips over any difficulties.
	 */

L
Linus Torvalds 已提交
347
	pagevec_init(&pvec, 0);
348 349
	while (index <= end && pagevec_lookup(&pvec, mapping, index,
			min(end - index, (pgoff_t)PAGEVEC_SIZE - 1) + 1)) {
350
		mem_cgroup_uncharge_start();
L
Linus Torvalds 已提交
351 352
		for (i = 0; i < pagevec_count(&pvec); i++) {
			struct page *page = pvec.pages[i];
353

354
			/* We rely upon deletion not changing page->index */
355
			index = page->index;
356 357
			if (index > end)
				break;
358

359 360 361
			if (!trylock_page(page))
				continue;
			WARN_ON(page->index != index);
M
Minchan Kim 已提交
362
			ret = invalidate_inode_page(page);
L
Linus Torvalds 已提交
363
			unlock_page(page);
M
Minchan Kim 已提交
364 365 366 367 368 369 370
			/*
			 * Invalidation is a hint that the page is no longer
			 * of interest and try to speed up its reclaim.
			 */
			if (!ret)
				deactivate_page(page);
			count += ret;
L
Linus Torvalds 已提交
371 372
		}
		pagevec_release(&pvec);
373
		mem_cgroup_uncharge_end();
374
		cond_resched();
375
		index++;
L
Linus Torvalds 已提交
376
	}
M
Minchan Kim 已提交
377
	return count;
L
Linus Torvalds 已提交
378
}
379
EXPORT_SYMBOL(invalidate_mapping_pages);
L
Linus Torvalds 已提交
380

381 382 383 384
/*
 * This is like invalidate_complete_page(), except it ignores the page's
 * refcount.  We do this because invalidate_inode_pages2() needs stronger
 * invalidation guarantees, and cannot afford to leave pages behind because
385 386
 * shrink_page_list() has a temp ref on them, or because they're transiently
 * sitting in the lru_cache_add() pagevecs.
387 388 389 390 391 392 393
 */
static int
invalidate_complete_page2(struct address_space *mapping, struct page *page)
{
	if (page->mapping != mapping)
		return 0;

394
	if (page_has_private(page) && !try_to_release_page(page, GFP_KERNEL))
395 396
		return 0;

N
Nick Piggin 已提交
397
	spin_lock_irq(&mapping->tree_lock);
398 399 400
	if (PageDirty(page))
		goto failed;

401
	clear_page_mlock(page);
402
	BUG_ON(page_has_private(page));
403
	__delete_from_page_cache(page);
N
Nick Piggin 已提交
404
	spin_unlock_irq(&mapping->tree_lock);
405
	mem_cgroup_uncharge_cache_page(page);
406 407 408 409

	if (mapping->a_ops->freepage)
		mapping->a_ops->freepage(page);

410 411 412
	page_cache_release(page);	/* pagecache ref */
	return 1;
failed:
N
Nick Piggin 已提交
413
	spin_unlock_irq(&mapping->tree_lock);
414 415 416
	return 0;
}

417 418 419 420 421 422 423 424 425
static int do_launder_page(struct address_space *mapping, struct page *page)
{
	if (!PageDirty(page))
		return 0;
	if (page->mapping != mapping || mapping->a_ops->launder_page == NULL)
		return 0;
	return mapping->a_ops->launder_page(page);
}

L
Linus Torvalds 已提交
426 427
/**
 * invalidate_inode_pages2_range - remove range of pages from an address_space
428
 * @mapping: the address_space
L
Linus Torvalds 已提交
429 430 431 432 433 434
 * @start: the page offset 'from' which to invalidate
 * @end: the page offset 'to' which to invalidate (inclusive)
 *
 * Any pages which are found to be mapped into pagetables are unmapped prior to
 * invalidation.
 *
435
 * Returns -EBUSY if any pages could not be invalidated.
L
Linus Torvalds 已提交
436 437 438 439 440
 */
int invalidate_inode_pages2_range(struct address_space *mapping,
				  pgoff_t start, pgoff_t end)
{
	struct pagevec pvec;
441
	pgoff_t index;
L
Linus Torvalds 已提交
442 443
	int i;
	int ret = 0;
444
	int ret2 = 0;
L
Linus Torvalds 已提交
445 446
	int did_range_unmap = 0;

447
	cleancache_flush_inode(mapping);
L
Linus Torvalds 已提交
448
	pagevec_init(&pvec, 0);
449 450 451
	index = start;
	while (index <= end && pagevec_lookup(&pvec, mapping, index,
			min(end - index, (pgoff_t)PAGEVEC_SIZE - 1) + 1)) {
452
		mem_cgroup_uncharge_start();
453
		for (i = 0; i < pagevec_count(&pvec); i++) {
L
Linus Torvalds 已提交
454
			struct page *page = pvec.pages[i];
455 456 457 458 459

			/* We rely upon deletion not changing page->index */
			index = page->index;
			if (index > end)
				break;
L
Linus Torvalds 已提交
460 461

			lock_page(page);
462
			WARN_ON(page->index != index);
L
Linus Torvalds 已提交
463 464 465 466 467
			if (page->mapping != mapping) {
				unlock_page(page);
				continue;
			}
			wait_on_page_writeback(page);
468
			if (page_mapped(page)) {
L
Linus Torvalds 已提交
469 470 471 472 473
				if (!did_range_unmap) {
					/*
					 * Zap the rest of the file in one hit.
					 */
					unmap_mapping_range(mapping,
474 475 476
					   (loff_t)index << PAGE_CACHE_SHIFT,
					   (loff_t)(1 + end - index)
							 << PAGE_CACHE_SHIFT,
L
Linus Torvalds 已提交
477 478 479 480 481 482 483
					    0);
					did_range_unmap = 1;
				} else {
					/*
					 * Just zap this page
					 */
					unmap_mapping_range(mapping,
484 485
					   (loff_t)index << PAGE_CACHE_SHIFT,
					   PAGE_CACHE_SIZE, 0);
L
Linus Torvalds 已提交
486 487
				}
			}
488
			BUG_ON(page_mapped(page));
489 490 491
			ret2 = do_launder_page(mapping, page);
			if (ret2 == 0) {
				if (!invalidate_complete_page2(mapping, page))
492
					ret2 = -EBUSY;
493 494 495
			}
			if (ret2 < 0)
				ret = ret2;
L
Linus Torvalds 已提交
496 497 498
			unlock_page(page);
		}
		pagevec_release(&pvec);
499
		mem_cgroup_uncharge_end();
L
Linus Torvalds 已提交
500
		cond_resched();
501
		index++;
L
Linus Torvalds 已提交
502
	}
503
	cleancache_flush_inode(mapping);
L
Linus Torvalds 已提交
504 505 506 507 508 509
	return ret;
}
EXPORT_SYMBOL_GPL(invalidate_inode_pages2_range);

/**
 * invalidate_inode_pages2 - remove all pages from an address_space
510
 * @mapping: the address_space
L
Linus Torvalds 已提交
511 512 513 514
 *
 * Any pages which are found to be mapped into pagetables are unmapped prior to
 * invalidation.
 *
515
 * Returns -EBUSY if any pages could not be invalidated.
L
Linus Torvalds 已提交
516 517 518 519 520 521
 */
int invalidate_inode_pages2(struct address_space *mapping)
{
	return invalidate_inode_pages2_range(mapping, 0, -1);
}
EXPORT_SYMBOL_GPL(invalidate_inode_pages2);
N
npiggin@suse.de 已提交
522 523 524 525

/**
 * truncate_pagecache - unmap and remove pagecache that has been truncated
 * @inode: inode
526 527
 * @oldsize: old file size
 * @newsize: new file size
N
npiggin@suse.de 已提交
528 529 530 531 532 533 534 535 536 537 538
 *
 * inode's new i_size must already be written before truncate_pagecache
 * is called.
 *
 * This function should typically be called before the filesystem
 * releases resources associated with the freed range (eg. deallocates
 * blocks). This way, pagecache will always stay logically coherent
 * with on-disk format, and the filesystem would not have to deal with
 * situations such as writepage being called for a page that has already
 * had its underlying blocks deallocated.
 */
539
void truncate_pagecache(struct inode *inode, loff_t oldsize, loff_t newsize)
N
npiggin@suse.de 已提交
540
{
541
	struct address_space *mapping = inode->i_mapping;
542
	loff_t holebegin = round_up(newsize, PAGE_SIZE);
543 544 545 546 547 548 549 550 551 552

	/*
	 * unmap_mapping_range is called twice, first simply for
	 * efficiency so that truncate_inode_pages does fewer
	 * single-page unmaps.  However after this first call, and
	 * before truncate_inode_pages finishes, it is possible for
	 * private pages to be COWed, which remain after
	 * truncate_inode_pages finishes, hence the second
	 * unmap_mapping_range call must be made for correctness.
	 */
553 554 555
	unmap_mapping_range(mapping, holebegin, 0, 1);
	truncate_inode_pages(mapping, newsize);
	unmap_mapping_range(mapping, holebegin, 0, 1);
N
npiggin@suse.de 已提交
556 557 558
}
EXPORT_SYMBOL(truncate_pagecache);

559 560 561 562 563
/**
 * truncate_setsize - update inode and pagecache for a new file size
 * @inode: inode
 * @newsize: new file size
 *
J
Jan Kara 已提交
564 565 566
 * truncate_setsize updates i_size and performs pagecache truncation (if
 * necessary) to @newsize. It will be typically be called from the filesystem's
 * setattr function when ATTR_SIZE is passed in.
567
 *
J
Jan Kara 已提交
568 569
 * Must be called with inode_mutex held and before all filesystem specific
 * block truncation has been performed.
570 571 572 573 574 575 576 577 578 579 580 581
 */
void truncate_setsize(struct inode *inode, loff_t newsize)
{
	loff_t oldsize;

	oldsize = inode->i_size;
	i_size_write(inode, newsize);

	truncate_pagecache(inode, oldsize, newsize);
}
EXPORT_SYMBOL(truncate_setsize);

N
npiggin@suse.de 已提交
582 583 584
/**
 * vmtruncate - unmap mappings "freed" by truncate() syscall
 * @inode: inode of the file used
585
 * @newsize: file offset to start truncating
N
npiggin@suse.de 已提交
586
 *
587 588
 * This function is deprecated and truncate_setsize or truncate_pagecache
 * should be used instead, together with filesystem specific block truncation.
N
npiggin@suse.de 已提交
589
 */
590
int vmtruncate(struct inode *inode, loff_t newsize)
N
npiggin@suse.de 已提交
591 592 593
{
	int error;

594
	error = inode_newsize_ok(inode, newsize);
N
npiggin@suse.de 已提交
595 596
	if (error)
		return error;
597

598
	truncate_setsize(inode, newsize);
N
npiggin@suse.de 已提交
599 600
	if (inode->i_op->truncate)
		inode->i_op->truncate(inode);
601
	return 0;
N
npiggin@suse.de 已提交
602 603
}
EXPORT_SYMBOL(vmtruncate);
604

605
int vmtruncate_range(struct inode *inode, loff_t lstart, loff_t lend)
606 607
{
	struct address_space *mapping = inode->i_mapping;
608 609
	loff_t holebegin = round_up(lstart, PAGE_SIZE);
	loff_t holelen = 1 + lend - holebegin;
610 611 612 613 614 615 616 617 618 619

	/*
	 * If the underlying filesystem is not going to provide
	 * a way to truncate a range of blocks (punch a hole) -
	 * we should return failure right now.
	 */
	if (!inode->i_op->truncate_range)
		return -ENOSYS;

	mutex_lock(&inode->i_mutex);
C
Christoph Hellwig 已提交
620
	inode_dio_wait(inode);
621 622
	unmap_mapping_range(mapping, holebegin, holelen, 1);
	inode->i_op->truncate_range(inode, lstart, lend);
623
	/* unmap again to remove racily COWed private pages */
624
	unmap_mapping_range(mapping, holebegin, holelen, 1);
625 626 627 628
	mutex_unlock(&inode->i_mutex);

	return 0;
}