shmem.c 64.6 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8
/*
 * Resizable virtual memory filesystem for Linux.
 *
 * Copyright (C) 2000 Linus Torvalds.
 *		 2000 Transmeta Corp.
 *		 2000-2001 Christoph Rohland
 *		 2000-2001 SAP AG
 *		 2002 Red Hat Inc.
9 10
 * Copyright (C) 2002-2005 Hugh Dickins.
 * Copyright (C) 2002-2005 VERITAS Software Corporation.
L
Linus Torvalds 已提交
11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
 * Copyright (C) 2004 Andi Kleen, SuSE Labs
 *
 * Extended attribute support for tmpfs:
 * Copyright (c) 2004, Luke Kenneth Casson Leighton <lkcl@lkcl.net>
 * Copyright (c) 2004 Red Hat, Inc., James Morris <jmorris@redhat.com>
 *
 * This file is released under the GPL.
 */

/*
 * This virtual memory filesystem is heavily based on the ramfs. It
 * extends ramfs by the ability to use swap and honor resource limits
 * which makes it a completely usable filesystem.
 */

#include <linux/module.h>
#include <linux/init.h>
#include <linux/fs.h>
29
#include <linux/xattr.h>
30
#include <linux/exportfs.h>
31
#include <linux/generic_acl.h>
L
Linus Torvalds 已提交
32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
#include <linux/mm.h>
#include <linux/mman.h>
#include <linux/file.h>
#include <linux/swap.h>
#include <linux/pagemap.h>
#include <linux/string.h>
#include <linux/slab.h>
#include <linux/backing-dev.h>
#include <linux/shmem_fs.h>
#include <linux/mount.h>
#include <linux/writeback.h>
#include <linux/vfs.h>
#include <linux/blkdev.h>
#include <linux/security.h>
#include <linux/swapops.h>
#include <linux/mempolicy.h>
#include <linux/namei.h>
49
#include <linux/ctype.h>
50
#include <linux/migrate.h>
51
#include <linux/highmem.h>
52

L
Linus Torvalds 已提交
53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82
#include <asm/uaccess.h>
#include <asm/div64.h>
#include <asm/pgtable.h>

/* This magic number is used in glibc for posix shared memory */
#define TMPFS_MAGIC	0x01021994

#define ENTRIES_PER_PAGE (PAGE_CACHE_SIZE/sizeof(unsigned long))
#define ENTRIES_PER_PAGEPAGE (ENTRIES_PER_PAGE*ENTRIES_PER_PAGE)
#define BLOCKS_PER_PAGE  (PAGE_CACHE_SIZE/512)

#define SHMEM_MAX_INDEX  (SHMEM_NR_DIRECT + (ENTRIES_PER_PAGEPAGE/2) * (ENTRIES_PER_PAGE+1))
#define SHMEM_MAX_BYTES  ((unsigned long long)SHMEM_MAX_INDEX << PAGE_CACHE_SHIFT)

#define VM_ACCT(size)    (PAGE_CACHE_ALIGN(size) >> PAGE_SHIFT)

/* info->flags needs VM_flags to handle pagein/truncate races efficiently */
#define SHMEM_PAGEIN	 VM_READ
#define SHMEM_TRUNCATE	 VM_WRITE

/* Definition to limit shmem_truncate's steps between cond_rescheds */
#define LATENCY_LIMIT	 64

/* Pretend that each entry is of this size in directory's i_size */
#define BOGO_DIRENT_SIZE 20

/* Flag allocation requirements to shmem_getpage and shmem_swp_alloc */
enum sgp_type {
	SGP_READ,	/* don't exceed i_size, don't allocate page */
	SGP_CACHE,	/* don't exceed i_size, may allocate page */
83
	SGP_DIRTY,	/* like SGP_CACHE, but set new page dirty */
L
Linus Torvalds 已提交
84 85 86 87 88 89
	SGP_WRITE,	/* may exceed i_size, may allocate page */
};

static int shmem_getpage(struct inode *inode, unsigned long idx,
			 struct page **pagep, enum sgp_type sgp, int *type);

A
Al Viro 已提交
90
static inline struct page *shmem_dir_alloc(gfp_t gfp_mask)
L
Linus Torvalds 已提交
91 92 93 94 95
{
	/*
	 * The above definition of ENTRIES_PER_PAGE, and the use of
	 * BLOCKS_PER_PAGE on indirect pages, assume PAGE_CACHE_SIZE:
	 * might be reconsidered if it ever diverges from PAGE_SIZE.
96
	 *
97
	 * Mobility flags are masked out as swap vectors cannot move
L
Linus Torvalds 已提交
98
	 */
99
	return alloc_pages((gfp_mask & ~GFP_MOVABLE_MASK) | __GFP_ZERO,
100
				PAGE_CACHE_SHIFT-PAGE_SHIFT);
L
Linus Torvalds 已提交
101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180
}

static inline void shmem_dir_free(struct page *page)
{
	__free_pages(page, PAGE_CACHE_SHIFT-PAGE_SHIFT);
}

static struct page **shmem_dir_map(struct page *page)
{
	return (struct page **)kmap_atomic(page, KM_USER0);
}

static inline void shmem_dir_unmap(struct page **dir)
{
	kunmap_atomic(dir, KM_USER0);
}

static swp_entry_t *shmem_swp_map(struct page *page)
{
	return (swp_entry_t *)kmap_atomic(page, KM_USER1);
}

static inline void shmem_swp_balance_unmap(void)
{
	/*
	 * When passing a pointer to an i_direct entry, to code which
	 * also handles indirect entries and so will shmem_swp_unmap,
	 * we must arrange for the preempt count to remain in balance.
	 * What kmap_atomic of a lowmem page does depends on config
	 * and architecture, so pretend to kmap_atomic some lowmem page.
	 */
	(void) kmap_atomic(ZERO_PAGE(0), KM_USER1);
}

static inline void shmem_swp_unmap(swp_entry_t *entry)
{
	kunmap_atomic(entry, KM_USER1);
}

static inline struct shmem_sb_info *SHMEM_SB(struct super_block *sb)
{
	return sb->s_fs_info;
}

/*
 * shmem_file_setup pre-accounts the whole fixed size of a VM object,
 * for shared memory and for shared anonymous (/dev/zero) mappings
 * (unless MAP_NORESERVE and sysctl_overcommit_memory <= 1),
 * consistent with the pre-accounting of private mappings ...
 */
static inline int shmem_acct_size(unsigned long flags, loff_t size)
{
	return (flags & VM_ACCOUNT)?
		security_vm_enough_memory(VM_ACCT(size)): 0;
}

static inline void shmem_unacct_size(unsigned long flags, loff_t size)
{
	if (flags & VM_ACCOUNT)
		vm_unacct_memory(VM_ACCT(size));
}

/*
 * ... whereas tmpfs objects are accounted incrementally as
 * pages are allocated, in order to allow huge sparse files.
 * shmem_getpage reports shmem_acct_block failure as -ENOSPC not -ENOMEM,
 * so that a failure on a sparse tmpfs mapping will give SIGBUS not OOM.
 */
static inline int shmem_acct_block(unsigned long flags)
{
	return (flags & VM_ACCOUNT)?
		0: security_vm_enough_memory(VM_ACCT(PAGE_CACHE_SIZE));
}

static inline void shmem_unacct_blocks(unsigned long flags, long pages)
{
	if (!(flags & VM_ACCOUNT))
		vm_unacct_memory(pages * VM_ACCT(PAGE_CACHE_SIZE));
}

181
static const struct super_operations shmem_ops;
182
static const struct address_space_operations shmem_aops;
183
static const struct file_operations shmem_file_operations;
184 185 186
static const struct inode_operations shmem_inode_operations;
static const struct inode_operations shmem_dir_inode_operations;
static const struct inode_operations shmem_special_inode_operations;
L
Linus Torvalds 已提交
187 188
static struct vm_operations_struct shmem_vm_ops;

189
static struct backing_dev_info shmem_backing_dev_info  __read_mostly = {
L
Linus Torvalds 已提交
190 191 192 193 194 195
	.ra_pages	= 0,	/* No readahead */
	.capabilities	= BDI_CAP_NO_ACCT_DIRTY | BDI_CAP_NO_WRITEBACK,
	.unplug_io_fn	= default_unplug_io_fn,
};

static LIST_HEAD(shmem_swaplist);
196
static DEFINE_MUTEX(shmem_swaplist_mutex);
L
Linus Torvalds 已提交
197 198 199 200

static void shmem_free_blocks(struct inode *inode, long pages)
{
	struct shmem_sb_info *sbinfo = SHMEM_SB(inode->i_sb);
201
	if (sbinfo->max_blocks) {
L
Linus Torvalds 已提交
202 203 204 205 206 207 208
		spin_lock(&sbinfo->stat_lock);
		sbinfo->free_blocks += pages;
		inode->i_blocks -= pages*BLOCKS_PER_PAGE;
		spin_unlock(&sbinfo->stat_lock);
	}
}

209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233
static int shmem_reserve_inode(struct super_block *sb)
{
	struct shmem_sb_info *sbinfo = SHMEM_SB(sb);
	if (sbinfo->max_inodes) {
		spin_lock(&sbinfo->stat_lock);
		if (!sbinfo->free_inodes) {
			spin_unlock(&sbinfo->stat_lock);
			return -ENOSPC;
		}
		sbinfo->free_inodes--;
		spin_unlock(&sbinfo->stat_lock);
	}
	return 0;
}

static void shmem_free_inode(struct super_block *sb)
{
	struct shmem_sb_info *sbinfo = SHMEM_SB(sb);
	if (sbinfo->max_inodes) {
		spin_lock(&sbinfo->stat_lock);
		sbinfo->free_inodes++;
		spin_unlock(&sbinfo->stat_lock);
	}
}

L
Linus Torvalds 已提交
234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357
/*
 * shmem_recalc_inode - recalculate the size of an inode
 *
 * @inode: inode to recalc
 *
 * We have to calculate the free blocks since the mm can drop
 * undirtied hole pages behind our back.
 *
 * But normally   info->alloced == inode->i_mapping->nrpages + info->swapped
 * So mm freed is info->alloced - (inode->i_mapping->nrpages + info->swapped)
 *
 * It has to be called with the spinlock held.
 */
static void shmem_recalc_inode(struct inode *inode)
{
	struct shmem_inode_info *info = SHMEM_I(inode);
	long freed;

	freed = info->alloced - info->swapped - inode->i_mapping->nrpages;
	if (freed > 0) {
		info->alloced -= freed;
		shmem_unacct_blocks(info->flags, freed);
		shmem_free_blocks(inode, freed);
	}
}

/*
 * shmem_swp_entry - find the swap vector position in the info structure
 *
 * @info:  info structure for the inode
 * @index: index of the page to find
 * @page:  optional page to add to the structure. Has to be preset to
 *         all zeros
 *
 * If there is no space allocated yet it will return NULL when
 * page is NULL, else it will use the page for the needed block,
 * setting it to NULL on return to indicate that it has been used.
 *
 * The swap vector is organized the following way:
 *
 * There are SHMEM_NR_DIRECT entries directly stored in the
 * shmem_inode_info structure. So small files do not need an addional
 * allocation.
 *
 * For pages with index > SHMEM_NR_DIRECT there is the pointer
 * i_indirect which points to a page which holds in the first half
 * doubly indirect blocks, in the second half triple indirect blocks:
 *
 * For an artificial ENTRIES_PER_PAGE = 4 this would lead to the
 * following layout (for SHMEM_NR_DIRECT == 16):
 *
 * i_indirect -> dir --> 16-19
 * 	      |	     +-> 20-23
 * 	      |
 * 	      +-->dir2 --> 24-27
 * 	      |	       +-> 28-31
 * 	      |	       +-> 32-35
 * 	      |	       +-> 36-39
 * 	      |
 * 	      +-->dir3 --> 40-43
 * 	       	       +-> 44-47
 * 	      	       +-> 48-51
 * 	      	       +-> 52-55
 */
static swp_entry_t *shmem_swp_entry(struct shmem_inode_info *info, unsigned long index, struct page **page)
{
	unsigned long offset;
	struct page **dir;
	struct page *subdir;

	if (index < SHMEM_NR_DIRECT) {
		shmem_swp_balance_unmap();
		return info->i_direct+index;
	}
	if (!info->i_indirect) {
		if (page) {
			info->i_indirect = *page;
			*page = NULL;
		}
		return NULL;			/* need another page */
	}

	index -= SHMEM_NR_DIRECT;
	offset = index % ENTRIES_PER_PAGE;
	index /= ENTRIES_PER_PAGE;
	dir = shmem_dir_map(info->i_indirect);

	if (index >= ENTRIES_PER_PAGE/2) {
		index -= ENTRIES_PER_PAGE/2;
		dir += ENTRIES_PER_PAGE/2 + index/ENTRIES_PER_PAGE;
		index %= ENTRIES_PER_PAGE;
		subdir = *dir;
		if (!subdir) {
			if (page) {
				*dir = *page;
				*page = NULL;
			}
			shmem_dir_unmap(dir);
			return NULL;		/* need another page */
		}
		shmem_dir_unmap(dir);
		dir = shmem_dir_map(subdir);
	}

	dir += index;
	subdir = *dir;
	if (!subdir) {
		if (!page || !(subdir = *page)) {
			shmem_dir_unmap(dir);
			return NULL;		/* need a page */
		}
		*dir = subdir;
		*page = NULL;
	}
	shmem_dir_unmap(dir);
	return shmem_swp_map(subdir) + offset;
}

static void shmem_swp_set(struct shmem_inode_info *info, swp_entry_t *entry, unsigned long value)
{
	long incdec = value? 1: -1;

	entry->val = value;
	info->swapped += incdec;
H
Hugh Dickins 已提交
358 359 360 361
	if ((unsigned long)(entry - info->i_direct) >= SHMEM_NR_DIRECT) {
		struct page *page = kmap_atomic_to_page(entry);
		set_page_private(page, page_private(page) + incdec);
	}
L
Linus Torvalds 已提交
362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390
}

/*
 * shmem_swp_alloc - get the position of the swap entry for the page.
 *                   If it does not exist allocate the entry.
 *
 * @info:	info structure for the inode
 * @index:	index of the page to find
 * @sgp:	check and recheck i_size? skip allocation?
 */
static swp_entry_t *shmem_swp_alloc(struct shmem_inode_info *info, unsigned long index, enum sgp_type sgp)
{
	struct inode *inode = &info->vfs_inode;
	struct shmem_sb_info *sbinfo = SHMEM_SB(inode->i_sb);
	struct page *page = NULL;
	swp_entry_t *entry;

	if (sgp != SGP_WRITE &&
	    ((loff_t) index << PAGE_CACHE_SHIFT) >= i_size_read(inode))
		return ERR_PTR(-EINVAL);

	while (!(entry = shmem_swp_entry(info, index, &page))) {
		if (sgp == SGP_READ)
			return shmem_swp_map(ZERO_PAGE(0));
		/*
		 * Test free_blocks against 1 not 0, since we have 1 data
		 * page (and perhaps indirect index pages) yet to allocate:
		 * a waste to allocate index if we cannot allocate data.
		 */
391
		if (sbinfo->max_blocks) {
L
Linus Torvalds 已提交
392 393 394 395 396 397 398 399 400 401 402
			spin_lock(&sbinfo->stat_lock);
			if (sbinfo->free_blocks <= 1) {
				spin_unlock(&sbinfo->stat_lock);
				return ERR_PTR(-ENOSPC);
			}
			sbinfo->free_blocks--;
			inode->i_blocks += BLOCKS_PER_PAGE;
			spin_unlock(&sbinfo->stat_lock);
		}

		spin_unlock(&info->lock);
403
		page = shmem_dir_alloc(mapping_gfp_mask(inode->i_mapping));
H
Hugh Dickins 已提交
404 405
		if (page)
			set_page_private(page, 0);
L
Linus Torvalds 已提交
406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432
		spin_lock(&info->lock);

		if (!page) {
			shmem_free_blocks(inode, 1);
			return ERR_PTR(-ENOMEM);
		}
		if (sgp != SGP_WRITE &&
		    ((loff_t) index << PAGE_CACHE_SHIFT) >= i_size_read(inode)) {
			entry = ERR_PTR(-EINVAL);
			break;
		}
		if (info->next_index <= index)
			info->next_index = index + 1;
	}
	if (page) {
		/* another task gave its page, or truncated the file */
		shmem_free_blocks(inode, 1);
		shmem_dir_free(page);
	}
	if (info->next_index <= index && !IS_ERR(entry))
		info->next_index = index + 1;
	return entry;
}

/*
 * shmem_free_swp - free some swap entries in a directory
 *
433 434 435
 * @dir:        pointer to the directory
 * @edir:       pointer after last entry of the directory
 * @punch_lock: pointer to spinlock when needed for the holepunch case
L
Linus Torvalds 已提交
436
 */
437 438
static int shmem_free_swp(swp_entry_t *dir, swp_entry_t *edir,
						spinlock_t *punch_lock)
L
Linus Torvalds 已提交
439
{
440
	spinlock_t *punch_unlock = NULL;
L
Linus Torvalds 已提交
441 442 443 444 445
	swp_entry_t *ptr;
	int freed = 0;

	for (ptr = dir; ptr < edir; ptr++) {
		if (ptr->val) {
446 447 448 449 450 451 452
			if (unlikely(punch_lock)) {
				punch_unlock = punch_lock;
				punch_lock = NULL;
				spin_lock(punch_unlock);
				if (!ptr->val)
					continue;
			}
L
Linus Torvalds 已提交
453 454 455 456 457
			free_swap_and_cache(*ptr);
			*ptr = (swp_entry_t){0};
			freed++;
		}
	}
458 459
	if (punch_unlock)
		spin_unlock(punch_unlock);
L
Linus Torvalds 已提交
460 461 462
	return freed;
}

463 464
static int shmem_map_and_free_swp(struct page *subdir, int offset,
		int limit, struct page ***dir, spinlock_t *punch_lock)
L
Linus Torvalds 已提交
465 466 467 468 469 470 471 472 473
{
	swp_entry_t *ptr;
	int freed = 0;

	ptr = shmem_swp_map(subdir);
	for (; offset < limit; offset += LATENCY_LIMIT) {
		int size = limit - offset;
		if (size > LATENCY_LIMIT)
			size = LATENCY_LIMIT;
474 475
		freed += shmem_free_swp(ptr+offset, ptr+offset+size,
							punch_lock);
L
Linus Torvalds 已提交
476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506
		if (need_resched()) {
			shmem_swp_unmap(ptr);
			if (*dir) {
				shmem_dir_unmap(*dir);
				*dir = NULL;
			}
			cond_resched();
			ptr = shmem_swp_map(subdir);
		}
	}
	shmem_swp_unmap(ptr);
	return freed;
}

static void shmem_free_pages(struct list_head *next)
{
	struct page *page;
	int freed = 0;

	do {
		page = container_of(next, struct page, lru);
		next = next->next;
		shmem_dir_free(page);
		freed++;
		if (freed >= LATENCY_LIMIT) {
			cond_resched();
			freed = 0;
		}
	} while (next);
}

507
static void shmem_truncate_range(struct inode *inode, loff_t start, loff_t end)
L
Linus Torvalds 已提交
508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524
{
	struct shmem_inode_info *info = SHMEM_I(inode);
	unsigned long idx;
	unsigned long size;
	unsigned long limit;
	unsigned long stage;
	unsigned long diroff;
	struct page **dir;
	struct page *topdir;
	struct page *middir;
	struct page *subdir;
	swp_entry_t *ptr;
	LIST_HEAD(pages_to_free);
	long nr_pages_to_free = 0;
	long nr_swaps_freed = 0;
	int offset;
	int freed;
525
	int punch_hole;
526 527
	spinlock_t *needs_lock;
	spinlock_t *punch_lock;
528
	unsigned long upper_limit;
L
Linus Torvalds 已提交
529 530

	inode->i_ctime = inode->i_mtime = CURRENT_TIME;
531
	idx = (start + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
L
Linus Torvalds 已提交
532 533 534 535 536
	if (idx >= info->next_index)
		return;

	spin_lock(&info->lock);
	info->flags |= SHMEM_TRUNCATE;
537 538
	if (likely(end == (loff_t) -1)) {
		limit = info->next_index;
539
		upper_limit = SHMEM_MAX_INDEX;
540
		info->next_index = idx;
541
		needs_lock = NULL;
542
		punch_hole = 0;
543
	} else {
544 545 546 547 548 549 550 551
		if (end + 1 >= inode->i_size) {	/* we may free a little more */
			limit = (inode->i_size + PAGE_CACHE_SIZE - 1) >>
							PAGE_CACHE_SHIFT;
			upper_limit = SHMEM_MAX_INDEX;
		} else {
			limit = (end + 1) >> PAGE_CACHE_SHIFT;
			upper_limit = limit;
		}
552
		needs_lock = &info->lock;
553 554 555
		punch_hole = 1;
	}

L
Linus Torvalds 已提交
556
	topdir = info->i_indirect;
557
	if (topdir && idx <= SHMEM_NR_DIRECT && !punch_hole) {
L
Linus Torvalds 已提交
558 559 560 561 562 563 564 565 566 567 568
		info->i_indirect = NULL;
		nr_pages_to_free++;
		list_add(&topdir->lru, &pages_to_free);
	}
	spin_unlock(&info->lock);

	if (info->swapped && idx < SHMEM_NR_DIRECT) {
		ptr = info->i_direct;
		size = limit;
		if (size > SHMEM_NR_DIRECT)
			size = SHMEM_NR_DIRECT;
569
		nr_swaps_freed = shmem_free_swp(ptr+idx, ptr+size, needs_lock);
L
Linus Torvalds 已提交
570
	}
571 572 573 574 575

	/*
	 * If there are no indirect blocks or we are punching a hole
	 * below indirect blocks, nothing to be done.
	 */
576
	if (!topdir || limit <= SHMEM_NR_DIRECT)
L
Linus Torvalds 已提交
577 578
		goto done2;

579 580 581 582 583 584 585 586 587 588 589 590 591
	/*
	 * The truncation case has already dropped info->lock, and we're safe
	 * because i_size and next_index have already been lowered, preventing
	 * access beyond.  But in the punch_hole case, we still need to take
	 * the lock when updating the swap directory, because there might be
	 * racing accesses by shmem_getpage(SGP_CACHE), shmem_unuse_inode or
	 * shmem_writepage.  However, whenever we find we can remove a whole
	 * directory page (not at the misaligned start or end of the range),
	 * we first NULLify its pointer in the level above, and then have no
	 * need to take the lock when updating its contents: needs_lock and
	 * punch_lock (either pointing to info->lock or NULL) manage this.
	 */

592
	upper_limit -= SHMEM_NR_DIRECT;
L
Linus Torvalds 已提交
593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611
	limit -= SHMEM_NR_DIRECT;
	idx = (idx > SHMEM_NR_DIRECT)? (idx - SHMEM_NR_DIRECT): 0;
	offset = idx % ENTRIES_PER_PAGE;
	idx -= offset;

	dir = shmem_dir_map(topdir);
	stage = ENTRIES_PER_PAGEPAGE/2;
	if (idx < ENTRIES_PER_PAGEPAGE/2) {
		middir = topdir;
		diroff = idx/ENTRIES_PER_PAGE;
	} else {
		dir += ENTRIES_PER_PAGE/2;
		dir += (idx - ENTRIES_PER_PAGEPAGE/2)/ENTRIES_PER_PAGEPAGE;
		while (stage <= idx)
			stage += ENTRIES_PER_PAGEPAGE;
		middir = *dir;
		if (*dir) {
			diroff = ((idx - ENTRIES_PER_PAGEPAGE/2) %
				ENTRIES_PER_PAGEPAGE) / ENTRIES_PER_PAGE;
612
			if (!diroff && !offset && upper_limit >= stage) {
613 614 615 616 617 618 619
				if (needs_lock) {
					spin_lock(needs_lock);
					*dir = NULL;
					spin_unlock(needs_lock);
					needs_lock = NULL;
				} else
					*dir = NULL;
L
Linus Torvalds 已提交
620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644
				nr_pages_to_free++;
				list_add(&middir->lru, &pages_to_free);
			}
			shmem_dir_unmap(dir);
			dir = shmem_dir_map(middir);
		} else {
			diroff = 0;
			offset = 0;
			idx = stage;
		}
	}

	for (; idx < limit; idx += ENTRIES_PER_PAGE, diroff++) {
		if (unlikely(idx == stage)) {
			shmem_dir_unmap(dir);
			dir = shmem_dir_map(topdir) +
			    ENTRIES_PER_PAGE/2 + idx/ENTRIES_PER_PAGEPAGE;
			while (!*dir) {
				dir++;
				idx += ENTRIES_PER_PAGEPAGE;
				if (idx >= limit)
					goto done1;
			}
			stage = idx + ENTRIES_PER_PAGEPAGE;
			middir = *dir;
645 646
			if (punch_hole)
				needs_lock = &info->lock;
647
			if (upper_limit >= stage) {
648 649 650 651 652 653 654
				if (needs_lock) {
					spin_lock(needs_lock);
					*dir = NULL;
					spin_unlock(needs_lock);
					needs_lock = NULL;
				} else
					*dir = NULL;
655 656 657
				nr_pages_to_free++;
				list_add(&middir->lru, &pages_to_free);
			}
L
Linus Torvalds 已提交
658 659 660 661 662
			shmem_dir_unmap(dir);
			cond_resched();
			dir = shmem_dir_map(middir);
			diroff = 0;
		}
663
		punch_lock = needs_lock;
L
Linus Torvalds 已提交
664
		subdir = dir[diroff];
665 666 667 668 669 670 671 672 673 674 675 676
		if (subdir && !offset && upper_limit-idx >= ENTRIES_PER_PAGE) {
			if (needs_lock) {
				spin_lock(needs_lock);
				dir[diroff] = NULL;
				spin_unlock(needs_lock);
				punch_lock = NULL;
			} else
				dir[diroff] = NULL;
			nr_pages_to_free++;
			list_add(&subdir->lru, &pages_to_free);
		}
		if (subdir && page_private(subdir) /* has swap entries */) {
L
Linus Torvalds 已提交
677 678 679 680
			size = limit - idx;
			if (size > ENTRIES_PER_PAGE)
				size = ENTRIES_PER_PAGE;
			freed = shmem_map_and_free_swp(subdir,
681
					offset, size, &dir, punch_lock);
L
Linus Torvalds 已提交
682 683 684
			if (!dir)
				dir = shmem_dir_map(middir);
			nr_swaps_freed += freed;
685
			if (offset || punch_lock) {
L
Linus Torvalds 已提交
686
				spin_lock(&info->lock);
687 688
				set_page_private(subdir,
					page_private(subdir) - freed);
L
Linus Torvalds 已提交
689
				spin_unlock(&info->lock);
690 691
			} else
				BUG_ON(page_private(subdir) != freed);
L
Linus Torvalds 已提交
692
		}
693
		offset = 0;
L
Linus Torvalds 已提交
694 695 696 697 698 699 700 701 702 703 704
	}
done1:
	shmem_dir_unmap(dir);
done2:
	if (inode->i_mapping->nrpages && (info->flags & SHMEM_PAGEIN)) {
		/*
		 * Call truncate_inode_pages again: racing shmem_unuse_inode
		 * may have swizzled a page in from swap since vmtruncate or
		 * generic_delete_inode did it, before we lowered next_index.
		 * Also, though shmem_getpage checks i_size before adding to
		 * cache, no recheck after: so fix the narrow window there too.
705 706 707 708 709
		 *
		 * Recalling truncate_inode_pages_range and unmap_mapping_range
		 * every time for punch_hole (which never got a chance to clear
		 * SHMEM_PAGEIN at the start of vmtruncate_range) is expensive,
		 * yet hardly ever necessary: try to optimize them out later.
L
Linus Torvalds 已提交
710
		 */
711
		truncate_inode_pages_range(inode->i_mapping, start, end);
712 713 714
		if (punch_hole)
			unmap_mapping_range(inode->i_mapping, start,
							end - start, 1);
L
Linus Torvalds 已提交
715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733
	}

	spin_lock(&info->lock);
	info->flags &= ~SHMEM_TRUNCATE;
	info->swapped -= nr_swaps_freed;
	if (nr_pages_to_free)
		shmem_free_blocks(inode, nr_pages_to_free);
	shmem_recalc_inode(inode);
	spin_unlock(&info->lock);

	/*
	 * Empty swap vector directory pages to be freed?
	 */
	if (!list_empty(&pages_to_free)) {
		pages_to_free.prev->next = NULL;
		shmem_free_pages(pages_to_free.next);
	}
}

734 735 736 737 738
static void shmem_truncate(struct inode *inode)
{
	shmem_truncate_range(inode, inode->i_size, (loff_t)-1);
}

L
Linus Torvalds 已提交
739 740 741 742 743 744
static int shmem_notify_change(struct dentry *dentry, struct iattr *attr)
{
	struct inode *inode = dentry->d_inode;
	struct page *page = NULL;
	int error;

745
	if (S_ISREG(inode->i_mode) && (attr->ia_valid & ATTR_SIZE)) {
L
Linus Torvalds 已提交
746 747 748 749 750 751 752 753 754 755 756 757
		if (attr->ia_size < inode->i_size) {
			/*
			 * If truncating down to a partial page, then
			 * if that page is already allocated, hold it
			 * in memory until the truncation is over, so
			 * truncate_partial_page cannnot miss it were
			 * it assigned to swap.
			 */
			if (attr->ia_size & (PAGE_CACHE_SIZE-1)) {
				(void) shmem_getpage(inode,
					attr->ia_size>>PAGE_CACHE_SHIFT,
						&page, SGP_READ, NULL);
H
Hugh Dickins 已提交
758 759
				if (page)
					unlock_page(page);
L
Linus Torvalds 已提交
760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779
			}
			/*
			 * Reset SHMEM_PAGEIN flag so that shmem_truncate can
			 * detect if any pages might have been added to cache
			 * after truncate_inode_pages.  But we needn't bother
			 * if it's being fully truncated to zero-length: the
			 * nrpages check is efficient enough in that case.
			 */
			if (attr->ia_size) {
				struct shmem_inode_info *info = SHMEM_I(inode);
				spin_lock(&info->lock);
				info->flags &= ~SHMEM_PAGEIN;
				spin_unlock(&info->lock);
			}
		}
	}

	error = inode_change_ok(inode, attr);
	if (!error)
		error = inode_setattr(inode, attr);
780 781 782 783
#ifdef CONFIG_TMPFS_POSIX_ACL
	if (!error && (attr->ia_valid & ATTR_MODE))
		error = generic_acl_chmod(inode, &shmem_acl_ops);
#endif
L
Linus Torvalds 已提交
784 785 786 787 788 789 790 791 792 793
	if (page)
		page_cache_release(page);
	return error;
}

static void shmem_delete_inode(struct inode *inode)
{
	struct shmem_inode_info *info = SHMEM_I(inode);

	if (inode->i_op->truncate == shmem_truncate) {
794
		truncate_inode_pages(inode->i_mapping, 0);
L
Linus Torvalds 已提交
795 796 797 798
		shmem_unacct_size(info->flags, inode->i_size);
		inode->i_size = 0;
		shmem_truncate(inode);
		if (!list_empty(&info->swaplist)) {
799
			mutex_lock(&shmem_swaplist_mutex);
L
Linus Torvalds 已提交
800
			list_del_init(&info->swaplist);
801
			mutex_unlock(&shmem_swaplist_mutex);
L
Linus Torvalds 已提交
802 803
		}
	}
804
	BUG_ON(inode->i_blocks);
805
	shmem_free_inode(inode->i_sb);
L
Linus Torvalds 已提交
806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830
	clear_inode(inode);
}

static inline int shmem_find_swp(swp_entry_t entry, swp_entry_t *dir, swp_entry_t *edir)
{
	swp_entry_t *ptr;

	for (ptr = dir; ptr < edir; ptr++) {
		if (ptr->val == entry.val)
			return ptr - dir;
	}
	return -1;
}

static int shmem_unuse_inode(struct shmem_inode_info *info, swp_entry_t entry, struct page *page)
{
	struct inode *inode;
	unsigned long idx;
	unsigned long size;
	unsigned long limit;
	unsigned long stage;
	struct page **dir;
	struct page *subdir;
	swp_entry_t *ptr;
	int offset;
831
	int error;
L
Linus Torvalds 已提交
832 833 834 835

	idx = 0;
	ptr = info->i_direct;
	spin_lock(&info->lock);
H
Hugh Dickins 已提交
836 837 838 839
	if (!info->swapped) {
		list_del_init(&info->swaplist);
		goto lost2;
	}
L
Linus Torvalds 已提交
840 841 842 843 844
	limit = info->next_index;
	size = limit;
	if (size > SHMEM_NR_DIRECT)
		size = SHMEM_NR_DIRECT;
	offset = shmem_find_swp(entry, ptr, ptr+size);
845
	if (offset >= 0)
L
Linus Torvalds 已提交
846 847 848 849 850 851 852 853 854 855
		goto found;
	if (!info->i_indirect)
		goto lost2;

	dir = shmem_dir_map(info->i_indirect);
	stage = SHMEM_NR_DIRECT + ENTRIES_PER_PAGEPAGE/2;

	for (idx = SHMEM_NR_DIRECT; idx < limit; idx += ENTRIES_PER_PAGE, dir++) {
		if (unlikely(idx == stage)) {
			shmem_dir_unmap(dir-1);
856 857 858 859 860 861 862 863
			if (cond_resched_lock(&info->lock)) {
				/* check it has not been truncated */
				if (limit > info->next_index) {
					limit = info->next_index;
					if (idx >= limit)
						goto lost2;
				}
			}
L
Linus Torvalds 已提交
864 865 866 867 868 869 870 871 872 873 874 875 876 877
			dir = shmem_dir_map(info->i_indirect) +
			    ENTRIES_PER_PAGE/2 + idx/ENTRIES_PER_PAGEPAGE;
			while (!*dir) {
				dir++;
				idx += ENTRIES_PER_PAGEPAGE;
				if (idx >= limit)
					goto lost1;
			}
			stage = idx + ENTRIES_PER_PAGEPAGE;
			subdir = *dir;
			shmem_dir_unmap(dir);
			dir = shmem_dir_map(subdir);
		}
		subdir = *dir;
H
Hugh Dickins 已提交
878
		if (subdir && page_private(subdir)) {
L
Linus Torvalds 已提交
879 880 881 882 883
			ptr = shmem_swp_map(subdir);
			size = limit - idx;
			if (size > ENTRIES_PER_PAGE)
				size = ENTRIES_PER_PAGE;
			offset = shmem_find_swp(entry, ptr, ptr+size);
884
			shmem_swp_unmap(ptr);
L
Linus Torvalds 已提交
885 886 887 888 889 890 891 892 893 894 895 896 897
			if (offset >= 0) {
				shmem_dir_unmap(dir);
				goto found;
			}
		}
	}
lost1:
	shmem_dir_unmap(dir-1);
lost2:
	spin_unlock(&info->lock);
	return 0;
found:
	idx += offset;
898 899 900
	inode = igrab(&info->vfs_inode);
	spin_unlock(&info->lock);

H
Hugh Dickins 已提交
901 902 903 904 905 906 907 908 909
	/*
	 * Move _head_ to start search for next from here.
	 * But be careful: shmem_delete_inode checks list_empty without taking
	 * mutex, and there's an instant in list_move_tail when info->swaplist
	 * would appear empty, if it were the only one on shmem_swaplist.  We
	 * could avoid doing it if inode NULL; or use this minor optimization.
	 */
	if (shmem_swaplist.next != &info->swaplist)
		list_move_tail(&shmem_swaplist, &info->swaplist);
910 911 912 913 914
	mutex_unlock(&shmem_swaplist_mutex);

	error = 1;
	if (!inode)
		goto out;
H
Hugh Dickins 已提交
915 916 917 918
	error = radix_tree_preload(GFP_KERNEL);
	if (error)
		goto out;
	error = 1;
919 920 921 922 923

	spin_lock(&info->lock);
	ptr = shmem_swp_entry(info, idx, NULL);
	if (ptr && ptr->val == entry.val)
		error = add_to_page_cache(page, inode->i_mapping,
H
Hugh Dickins 已提交
924
						idx, GFP_NOWAIT);
925 926
	if (error == -EEXIST) {
		struct page *filepage = find_get_page(inode->i_mapping, idx);
927
		error = 1;
928 929 930 931 932 933 934 935 936 937 938
		if (filepage) {
			/*
			 * There might be a more uptodate page coming down
			 * from a stacked writepage: forget our swappage if so.
			 */
			if (PageUptodate(filepage))
				error = 0;
			page_cache_release(filepage);
		}
	}
	if (!error) {
939 940
		delete_from_swap_cache(page);
		set_page_dirty(page);
L
Linus Torvalds 已提交
941
		info->flags |= SHMEM_PAGEIN;
942 943 944
		shmem_swp_set(info, ptr, 0);
		swap_free(entry);
		error = 1;	/* not an error, but entry was found */
L
Linus Torvalds 已提交
945
	}
946 947
	if (ptr)
		shmem_swp_unmap(ptr);
L
Linus Torvalds 已提交
948
	spin_unlock(&info->lock);
H
Hugh Dickins 已提交
949
	radix_tree_preload_end();
950 951 952 953 954
out:
	unlock_page(page);
	page_cache_release(page);
	iput(inode);		/* allows for NULL */
	return error;
L
Linus Torvalds 已提交
955 956 957 958 959 960 961 962 963 964 965
}

/*
 * shmem_unuse() search for an eventually swapped out shmem page.
 */
int shmem_unuse(swp_entry_t entry, struct page *page)
{
	struct list_head *p, *next;
	struct shmem_inode_info *info;
	int found = 0;

966
	mutex_lock(&shmem_swaplist_mutex);
L
Linus Torvalds 已提交
967 968
	list_for_each_safe(p, next, &shmem_swaplist) {
		info = list_entry(p, struct shmem_inode_info, swaplist);
H
Hugh Dickins 已提交
969
		found = shmem_unuse_inode(info, entry, page);
970
		cond_resched();
971 972
		if (found)
			goto out;
L
Linus Torvalds 已提交
973
	}
974
	mutex_unlock(&shmem_swaplist_mutex);
975
out:	return found;	/* 0 or 1 or -ENOMEM */
L
Linus Torvalds 已提交
976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995
}

/*
 * Move the page from the page cache to the swap cache.
 */
static int shmem_writepage(struct page *page, struct writeback_control *wbc)
{
	struct shmem_inode_info *info;
	swp_entry_t *entry, swap;
	struct address_space *mapping;
	unsigned long index;
	struct inode *inode;

	BUG_ON(!PageLocked(page));
	mapping = page->mapping;
	index = page->index;
	inode = mapping->host;
	info = SHMEM_I(inode);
	if (info->flags & VM_LOCKED)
		goto redirty;
996
	if (!total_swap_pages)
L
Linus Torvalds 已提交
997 998
		goto redirty;

999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011
	/*
	 * shmem_backing_dev_info's capabilities prevent regular writeback or
	 * sync from ever calling shmem_writepage; but a stacking filesystem
	 * may use the ->writepage of its underlying filesystem, in which case
	 * tmpfs should write out to swap only in response to memory pressure,
	 * and not for pdflush or sync.  However, in those cases, we do still
	 * want to check if there's a redundant swappage to be discarded.
	 */
	if (wbc->for_reclaim)
		swap = get_swap_page();
	else
		swap.val = 0;

L
Linus Torvalds 已提交
1012 1013 1014 1015 1016 1017
	spin_lock(&info->lock);
	if (index >= info->next_index) {
		BUG_ON(!(info->flags & SHMEM_TRUNCATE));
		goto unlock;
	}
	entry = shmem_swp_entry(info, index, NULL);
1018 1019 1020 1021 1022 1023 1024 1025 1026
	if (entry->val) {
		/*
		 * The more uptodate page coming down from a stacked
		 * writepage should replace our old swappage.
		 */
		free_swap_and_cache(*entry);
		shmem_swp_set(info, entry, 0);
	}
	shmem_recalc_inode(inode);
L
Linus Torvalds 已提交
1027

1028
	if (swap.val && add_to_swap_cache(page, swap, GFP_ATOMIC) == 0) {
1029
		remove_from_page_cache(page);
L
Linus Torvalds 已提交
1030 1031
		shmem_swp_set(info, entry, swap.val);
		shmem_swp_unmap(entry);
H
Hugh Dickins 已提交
1032 1033 1034 1035
		if (list_empty(&info->swaplist))
			inode = igrab(inode);
		else
			inode = NULL;
L
Linus Torvalds 已提交
1036
		spin_unlock(&info->lock);
1037
		swap_duplicate(swap);
1038
		BUG_ON(page_mapped(page));
1039 1040
		page_cache_release(page);	/* pagecache ref */
		set_page_dirty(page);
L
Linus Torvalds 已提交
1041
		unlock_page(page);
H
Hugh Dickins 已提交
1042 1043 1044 1045 1046 1047 1048
		if (inode) {
			mutex_lock(&shmem_swaplist_mutex);
			/* move instead of add in case we're racing */
			list_move_tail(&info->swaplist, &shmem_swaplist);
			mutex_unlock(&shmem_swaplist_mutex);
			iput(inode);
		}
L
Linus Torvalds 已提交
1049 1050 1051 1052 1053 1054 1055 1056 1057
		return 0;
	}

	shmem_swp_unmap(entry);
unlock:
	spin_unlock(&info->lock);
	swap_free(swap);
redirty:
	set_page_dirty(page);
1058 1059 1060 1061
	if (wbc->for_reclaim)
		return AOP_WRITEPAGE_ACTIVATE;	/* Return with page locked */
	unlock_page(page);
	return 0;
L
Linus Torvalds 已提交
1062 1063 1064
}

#ifdef CONFIG_NUMA
1065
static inline int shmem_parse_mpol(char *value, int *policy, nodemask_t *policy_nodes)
1066 1067 1068 1069 1070 1071 1072 1073 1074
{
	char *nodelist = strchr(value, ':');
	int err = 1;

	if (nodelist) {
		/* NUL-terminate policy string */
		*nodelist++ = '\0';
		if (nodelist_parse(nodelist, *policy_nodes))
			goto out;
1075
		if (!nodes_subset(*policy_nodes, node_states[N_HIGH_MEMORY]))
1076
			goto out;
1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099
	}
	if (!strcmp(value, "default")) {
		*policy = MPOL_DEFAULT;
		/* Don't allow a nodelist */
		if (!nodelist)
			err = 0;
	} else if (!strcmp(value, "prefer")) {
		*policy = MPOL_PREFERRED;
		/* Insist on a nodelist of one node only */
		if (nodelist) {
			char *rest = nodelist;
			while (isdigit(*rest))
				rest++;
			if (!*rest)
				err = 0;
		}
	} else if (!strcmp(value, "bind")) {
		*policy = MPOL_BIND;
		/* Insist on a nodelist */
		if (nodelist)
			err = 0;
	} else if (!strcmp(value, "interleave")) {
		*policy = MPOL_INTERLEAVE;
1100 1101 1102
		/*
		 * Default to online nodes with memory if no nodelist
		 */
1103
		if (!nodelist)
1104
			*policy_nodes = node_states[N_HIGH_MEMORY];
1105 1106 1107 1108 1109 1110 1111 1112 1113
		err = 0;
	}
out:
	/* Restore string for error message */
	if (nodelist)
		*--nodelist = ':';
	return err;
}

1114 1115
static struct page *shmem_swapin(swp_entry_t entry, gfp_t gfp,
			struct shmem_inode_info *info, unsigned long idx)
L
Linus Torvalds 已提交
1116 1117
{
	struct vm_area_struct pvma;
1118
	struct page *page;
L
Linus Torvalds 已提交
1119 1120

	/* Create a pseudo vma that just contains the policy */
1121
	pvma.vm_start = 0;
L
Linus Torvalds 已提交
1122
	pvma.vm_pgoff = idx;
1123 1124
	pvma.vm_ops = NULL;
	pvma.vm_policy = mpol_shared_policy_lookup(&info->policy, idx);
1125
	page = swapin_readahead(entry, gfp, &pvma, 0);
L
Linus Torvalds 已提交
1126 1127 1128 1129
	mpol_free(pvma.vm_policy);
	return page;
}

1130 1131
static struct page *shmem_alloc_page(gfp_t gfp,
			struct shmem_inode_info *info, unsigned long idx)
L
Linus Torvalds 已提交
1132 1133 1134 1135
{
	struct vm_area_struct pvma;
	struct page *page;

1136 1137
	/* Create a pseudo vma that just contains the policy */
	pvma.vm_start = 0;
L
Linus Torvalds 已提交
1138
	pvma.vm_pgoff = idx;
1139 1140
	pvma.vm_ops = NULL;
	pvma.vm_policy = mpol_shared_policy_lookup(&info->policy, idx);
1141
	page = alloc_page_vma(gfp, &pvma, 0);
L
Linus Torvalds 已提交
1142 1143 1144 1145
	mpol_free(pvma.vm_policy);
	return page;
}
#else
1146 1147
static inline int shmem_parse_mpol(char *value, int *policy,
						nodemask_t *policy_nodes)
1148 1149 1150 1151
{
	return 1;
}

1152 1153
static inline struct page *shmem_swapin(swp_entry_t entry, gfp_t gfp,
			struct shmem_inode_info *info, unsigned long idx)
L
Linus Torvalds 已提交
1154
{
1155
	return swapin_readahead(entry, gfp, NULL, 0);
L
Linus Torvalds 已提交
1156 1157
}

1158 1159
static inline struct page *shmem_alloc_page(gfp_t gfp,
			struct shmem_inode_info *info, unsigned long idx)
L
Linus Torvalds 已提交
1160
{
1161
	return alloc_page(gfp);
L
Linus Torvalds 已提交
1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181
}
#endif

/*
 * shmem_getpage - either get the page from swap or allocate a new one
 *
 * If we allocate a new one we do not mark it dirty. That's up to the
 * vm. If we swap it in we mark it dirty since we also free the swap
 * entry since a page cannot live in both the swap and page cache
 */
static int shmem_getpage(struct inode *inode, unsigned long idx,
			struct page **pagep, enum sgp_type sgp, int *type)
{
	struct address_space *mapping = inode->i_mapping;
	struct shmem_inode_info *info = SHMEM_I(inode);
	struct shmem_sb_info *sbinfo;
	struct page *filepage = *pagep;
	struct page *swappage;
	swp_entry_t *entry;
	swp_entry_t swap;
1182
	gfp_t gfp;
L
Linus Torvalds 已提交
1183 1184 1185 1186
	int error;

	if (idx >= SHMEM_MAX_INDEX)
		return -EFBIG;
1187 1188

	if (type)
N
Nick Piggin 已提交
1189
		*type = 0;
1190

L
Linus Torvalds 已提交
1191 1192 1193 1194
	/*
	 * Normally, filepage is NULL on entry, and either found
	 * uptodate immediately, or allocated and zeroed, or read
	 * in under swappage, which is then assigned to filepage.
H
Hugh Dickins 已提交
1195
	 * But shmem_readpage (required for splice) passes in a locked
1196 1197
	 * filepage, which may be found not uptodate by other callers
	 * too, and may need to be copied from the swappage read in.
L
Linus Torvalds 已提交
1198 1199 1200 1201 1202 1203 1204
	 */
repeat:
	if (!filepage)
		filepage = find_lock_page(mapping, idx);
	if (filepage && PageUptodate(filepage))
		goto done;
	error = 0;
1205
	gfp = mapping_gfp_mask(mapping);
H
Hugh Dickins 已提交
1206 1207 1208 1209 1210 1211 1212 1213 1214 1215
	if (!filepage) {
		/*
		 * Try to preload while we can wait, to not make a habit of
		 * draining atomic reserves; but don't latch on to this cpu.
		 */
		error = radix_tree_preload(gfp & ~__GFP_HIGHMEM);
		if (error)
			goto failed;
		radix_tree_preload_end();
	}
L
Linus Torvalds 已提交
1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232

	spin_lock(&info->lock);
	shmem_recalc_inode(inode);
	entry = shmem_swp_alloc(info, idx, sgp);
	if (IS_ERR(entry)) {
		spin_unlock(&info->lock);
		error = PTR_ERR(entry);
		goto failed;
	}
	swap = *entry;

	if (swap.val) {
		/* Look it up and read it in.. */
		swappage = lookup_swap_cache(swap);
		if (!swappage) {
			shmem_swp_unmap(entry);
			/* here we actually do the io */
N
Nick Piggin 已提交
1233
			if (type && !(*type & VM_FAULT_MAJOR)) {
1234
				__count_vm_event(PGMAJFAULT);
N
Nick Piggin 已提交
1235
				*type |= VM_FAULT_MAJOR;
L
Linus Torvalds 已提交
1236
			}
1237
			spin_unlock(&info->lock);
1238
			swappage = shmem_swapin(swap, gfp, info, idx);
L
Linus Torvalds 已提交
1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295
			if (!swappage) {
				spin_lock(&info->lock);
				entry = shmem_swp_alloc(info, idx, sgp);
				if (IS_ERR(entry))
					error = PTR_ERR(entry);
				else {
					if (entry->val == swap.val)
						error = -ENOMEM;
					shmem_swp_unmap(entry);
				}
				spin_unlock(&info->lock);
				if (error)
					goto failed;
				goto repeat;
			}
			wait_on_page_locked(swappage);
			page_cache_release(swappage);
			goto repeat;
		}

		/* We have to do this with page locked to prevent races */
		if (TestSetPageLocked(swappage)) {
			shmem_swp_unmap(entry);
			spin_unlock(&info->lock);
			wait_on_page_locked(swappage);
			page_cache_release(swappage);
			goto repeat;
		}
		if (PageWriteback(swappage)) {
			shmem_swp_unmap(entry);
			spin_unlock(&info->lock);
			wait_on_page_writeback(swappage);
			unlock_page(swappage);
			page_cache_release(swappage);
			goto repeat;
		}
		if (!PageUptodate(swappage)) {
			shmem_swp_unmap(entry);
			spin_unlock(&info->lock);
			unlock_page(swappage);
			page_cache_release(swappage);
			error = -EIO;
			goto failed;
		}

		if (filepage) {
			shmem_swp_set(info, entry, 0);
			shmem_swp_unmap(entry);
			delete_from_swap_cache(swappage);
			spin_unlock(&info->lock);
			copy_highpage(filepage, swappage);
			unlock_page(swappage);
			page_cache_release(swappage);
			flush_dcache_page(filepage);
			SetPageUptodate(filepage);
			set_page_dirty(filepage);
			swap_free(swap);
1296
		} else if (!(error = add_to_page_cache(
H
Hugh Dickins 已提交
1297
				swappage, mapping, idx, GFP_NOWAIT))) {
L
Linus Torvalds 已提交
1298 1299 1300
			info->flags |= SHMEM_PAGEIN;
			shmem_swp_set(info, entry, 0);
			shmem_swp_unmap(entry);
1301
			delete_from_swap_cache(swappage);
L
Linus Torvalds 已提交
1302 1303
			spin_unlock(&info->lock);
			filepage = swappage;
1304
			set_page_dirty(filepage);
L
Linus Torvalds 已提交
1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327
			swap_free(swap);
		} else {
			shmem_swp_unmap(entry);
			spin_unlock(&info->lock);
			unlock_page(swappage);
			page_cache_release(swappage);
			goto repeat;
		}
	} else if (sgp == SGP_READ && !filepage) {
		shmem_swp_unmap(entry);
		filepage = find_get_page(mapping, idx);
		if (filepage &&
		    (!PageUptodate(filepage) || TestSetPageLocked(filepage))) {
			spin_unlock(&info->lock);
			wait_on_page_locked(filepage);
			page_cache_release(filepage);
			filepage = NULL;
			goto repeat;
		}
		spin_unlock(&info->lock);
	} else {
		shmem_swp_unmap(entry);
		sbinfo = SHMEM_SB(inode->i_sb);
1328
		if (sbinfo->max_blocks) {
L
Linus Torvalds 已提交
1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347
			spin_lock(&sbinfo->stat_lock);
			if (sbinfo->free_blocks == 0 ||
			    shmem_acct_block(info->flags)) {
				spin_unlock(&sbinfo->stat_lock);
				spin_unlock(&info->lock);
				error = -ENOSPC;
				goto failed;
			}
			sbinfo->free_blocks--;
			inode->i_blocks += BLOCKS_PER_PAGE;
			spin_unlock(&sbinfo->stat_lock);
		} else if (shmem_acct_block(info->flags)) {
			spin_unlock(&info->lock);
			error = -ENOSPC;
			goto failed;
		}

		if (!filepage) {
			spin_unlock(&info->lock);
1348
			filepage = shmem_alloc_page(gfp, info, idx);
L
Linus Torvalds 已提交
1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364
			if (!filepage) {
				shmem_unacct_blocks(info->flags, 1);
				shmem_free_blocks(inode, 1);
				error = -ENOMEM;
				goto failed;
			}

			spin_lock(&info->lock);
			entry = shmem_swp_alloc(info, idx, sgp);
			if (IS_ERR(entry))
				error = PTR_ERR(entry);
			else {
				swap = *entry;
				shmem_swp_unmap(entry);
			}
			if (error || swap.val || 0 != add_to_page_cache_lru(
H
Hugh Dickins 已提交
1365
					filepage, mapping, idx, GFP_NOWAIT)) {
L
Linus Torvalds 已提交
1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379
				spin_unlock(&info->lock);
				page_cache_release(filepage);
				shmem_unacct_blocks(info->flags, 1);
				shmem_free_blocks(inode, 1);
				filepage = NULL;
				if (error)
					goto failed;
				goto repeat;
			}
			info->flags |= SHMEM_PAGEIN;
		}

		info->alloced++;
		spin_unlock(&info->lock);
1380
		clear_highpage(filepage);
L
Linus Torvalds 已提交
1381 1382
		flush_dcache_page(filepage);
		SetPageUptodate(filepage);
1383 1384
		if (sgp == SGP_DIRTY)
			set_page_dirty(filepage);
L
Linus Torvalds 已提交
1385 1386
	}
done:
H
Hugh Dickins 已提交
1387
	*pagep = filepage;
L
Linus Torvalds 已提交
1388 1389 1390 1391 1392 1393 1394 1395 1396 1397
	return 0;

failed:
	if (*pagep != filepage) {
		unlock_page(filepage);
		page_cache_release(filepage);
	}
	return error;
}

N
Nick Piggin 已提交
1398
static int shmem_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
L
Linus Torvalds 已提交
1399
{
1400
	struct inode *inode = vma->vm_file->f_path.dentry->d_inode;
L
Linus Torvalds 已提交
1401
	int error;
N
Nick Piggin 已提交
1402
	int ret;
L
Linus Torvalds 已提交
1403

N
Nick Piggin 已提交
1404 1405
	if (((loff_t)vmf->pgoff << PAGE_CACHE_SHIFT) >= i_size_read(inode))
		return VM_FAULT_SIGBUS;
1406

1407
	error = shmem_getpage(inode, vmf->pgoff, &vmf->page, SGP_CACHE, &ret);
N
Nick Piggin 已提交
1408 1409
	if (error)
		return ((error == -ENOMEM) ? VM_FAULT_OOM : VM_FAULT_SIGBUS);
L
Linus Torvalds 已提交
1410

N
Nick Piggin 已提交
1411
	mark_page_accessed(vmf->page);
N
Nick Piggin 已提交
1412
	return ret | VM_FAULT_LOCKED;
L
Linus Torvalds 已提交
1413 1414 1415
}

#ifdef CONFIG_NUMA
A
Adrian Bunk 已提交
1416
static int shmem_set_policy(struct vm_area_struct *vma, struct mempolicy *new)
L
Linus Torvalds 已提交
1417
{
1418
	struct inode *i = vma->vm_file->f_path.dentry->d_inode;
L
Linus Torvalds 已提交
1419 1420 1421
	return mpol_set_shared_policy(&SHMEM_I(i)->policy, vma, new);
}

A
Adrian Bunk 已提交
1422 1423
static struct mempolicy *shmem_get_policy(struct vm_area_struct *vma,
					  unsigned long addr)
L
Linus Torvalds 已提交
1424
{
1425
	struct inode *i = vma->vm_file->f_path.dentry->d_inode;
L
Linus Torvalds 已提交
1426 1427 1428 1429 1430 1431 1432 1433 1434
	unsigned long idx;

	idx = ((addr - vma->vm_start) >> PAGE_SHIFT) + vma->vm_pgoff;
	return mpol_shared_policy_lookup(&SHMEM_I(i)->policy, idx);
}
#endif

int shmem_lock(struct file *file, int lock, struct user_struct *user)
{
1435
	struct inode *inode = file->f_path.dentry->d_inode;
L
Linus Torvalds 已提交
1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454
	struct shmem_inode_info *info = SHMEM_I(inode);
	int retval = -ENOMEM;

	spin_lock(&info->lock);
	if (lock && !(info->flags & VM_LOCKED)) {
		if (!user_shm_lock(inode->i_size, user))
			goto out_nomem;
		info->flags |= VM_LOCKED;
	}
	if (!lock && (info->flags & VM_LOCKED) && user) {
		user_shm_unlock(inode->i_size, user);
		info->flags &= ~VM_LOCKED;
	}
	retval = 0;
out_nomem:
	spin_unlock(&info->lock);
	return retval;
}

1455
static int shmem_mmap(struct file *file, struct vm_area_struct *vma)
L
Linus Torvalds 已提交
1456 1457 1458
{
	file_accessed(file);
	vma->vm_ops = &shmem_vm_ops;
N
Nick Piggin 已提交
1459
	vma->vm_flags |= VM_CAN_NONLINEAR;
L
Linus Torvalds 已提交
1460 1461 1462 1463 1464 1465 1466 1467 1468 1469
	return 0;
}

static struct inode *
shmem_get_inode(struct super_block *sb, int mode, dev_t dev)
{
	struct inode *inode;
	struct shmem_inode_info *info;
	struct shmem_sb_info *sbinfo = SHMEM_SB(sb);

1470 1471
	if (shmem_reserve_inode(sb))
		return NULL;
L
Linus Torvalds 已提交
1472 1473 1474 1475 1476 1477 1478 1479 1480 1481

	inode = new_inode(sb);
	if (inode) {
		inode->i_mode = mode;
		inode->i_uid = current->fsuid;
		inode->i_gid = current->fsgid;
		inode->i_blocks = 0;
		inode->i_mapping->a_ops = &shmem_aops;
		inode->i_mapping->backing_dev_info = &shmem_backing_dev_info;
		inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME;
1482
		inode->i_generation = get_seconds();
L
Linus Torvalds 已提交
1483 1484 1485 1486 1487 1488 1489
		info = SHMEM_I(inode);
		memset(info, 0, (char *)inode - (char *)info);
		spin_lock_init(&info->lock);
		INIT_LIST_HEAD(&info->swaplist);

		switch (mode & S_IFMT) {
		default:
1490
			inode->i_op = &shmem_special_inode_operations;
L
Linus Torvalds 已提交
1491 1492 1493 1494 1495
			init_special_inode(inode, mode, dev);
			break;
		case S_IFREG:
			inode->i_op = &shmem_inode_operations;
			inode->i_fop = &shmem_file_operations;
1496 1497
			mpol_shared_policy_init(&info->policy, sbinfo->policy,
							&sbinfo->policy_nodes);
L
Linus Torvalds 已提交
1498 1499
			break;
		case S_IFDIR:
1500
			inc_nlink(inode);
L
Linus Torvalds 已提交
1501 1502 1503 1504 1505 1506 1507 1508 1509 1510
			/* Some things misbehave if size == 0 on a directory */
			inode->i_size = 2 * BOGO_DIRENT_SIZE;
			inode->i_op = &shmem_dir_inode_operations;
			inode->i_fop = &simple_dir_operations;
			break;
		case S_IFLNK:
			/*
			 * Must not load anything in the rbtree,
			 * mpol_free_shared_policy will not be called.
			 */
1511 1512
			mpol_shared_policy_init(&info->policy, MPOL_DEFAULT,
						NULL);
L
Linus Torvalds 已提交
1513 1514
			break;
		}
1515 1516
	} else
		shmem_free_inode(sb);
L
Linus Torvalds 已提交
1517 1518 1519 1520
	return inode;
}

#ifdef CONFIG_TMPFS
1521 1522
static const struct inode_operations shmem_symlink_inode_operations;
static const struct inode_operations shmem_symlink_inline_operations;
L
Linus Torvalds 已提交
1523 1524

/*
N
Nick Piggin 已提交
1525
 * Normally tmpfs avoids the use of shmem_readpage and shmem_write_begin;
1526 1527
 * but providing them allows a tmpfs file to be used for splice, sendfile, and
 * below the loop driver, in the generic fashion that many filesystems support.
L
Linus Torvalds 已提交
1528
 */
1529 1530 1531 1532 1533 1534 1535 1536
static int shmem_readpage(struct file *file, struct page *page)
{
	struct inode *inode = page->mapping->host;
	int error = shmem_getpage(inode, page->index, &page, SGP_CACHE, NULL);
	unlock_page(page);
	return error;
}

L
Linus Torvalds 已提交
1537
static int
N
Nick Piggin 已提交
1538 1539 1540
shmem_write_begin(struct file *file, struct address_space *mapping,
			loff_t pos, unsigned len, unsigned flags,
			struct page **pagep, void **fsdata)
L
Linus Torvalds 已提交
1541
{
N
Nick Piggin 已提交
1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554
	struct inode *inode = mapping->host;
	pgoff_t index = pos >> PAGE_CACHE_SHIFT;
	*pagep = NULL;
	return shmem_getpage(inode, index, pagep, SGP_WRITE, NULL);
}

static int
shmem_write_end(struct file *file, struct address_space *mapping,
			loff_t pos, unsigned len, unsigned copied,
			struct page *page, void *fsdata)
{
	struct inode *inode = mapping->host;

H
Hugh Dickins 已提交
1555 1556 1557 1558
	if (pos + copied > inode->i_size)
		i_size_write(inode, pos + copied);

	unlock_page(page);
N
Nick Piggin 已提交
1559 1560 1561 1562
	set_page_dirty(page);
	page_cache_release(page);

	return copied;
L
Linus Torvalds 已提交
1563 1564 1565 1566
}

static void do_shmem_file_read(struct file *filp, loff_t *ppos, read_descriptor_t *desc, read_actor_t actor)
{
1567
	struct inode *inode = filp->f_path.dentry->d_inode;
L
Linus Torvalds 已提交
1568 1569
	struct address_space *mapping = inode->i_mapping;
	unsigned long index, offset;
1570 1571 1572 1573 1574 1575 1576 1577 1578
	enum sgp_type sgp = SGP_READ;

	/*
	 * Might this read be for a stacking filesystem?  Then when reading
	 * holes of a sparse file, we actually need to allocate those pages,
	 * and even mark them dirty, so it cannot exceed the max_blocks limit.
	 */
	if (segment_eq(get_fs(), KERNEL_DS))
		sgp = SGP_DIRTY;
L
Linus Torvalds 已提交
1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596

	index = *ppos >> PAGE_CACHE_SHIFT;
	offset = *ppos & ~PAGE_CACHE_MASK;

	for (;;) {
		struct page *page = NULL;
		unsigned long end_index, nr, ret;
		loff_t i_size = i_size_read(inode);

		end_index = i_size >> PAGE_CACHE_SHIFT;
		if (index > end_index)
			break;
		if (index == end_index) {
			nr = i_size & ~PAGE_CACHE_MASK;
			if (nr <= offset)
				break;
		}

1597
		desc->error = shmem_getpage(inode, index, &page, sgp, NULL);
L
Linus Torvalds 已提交
1598 1599 1600 1601 1602
		if (desc->error) {
			if (desc->error == -EINVAL)
				desc->error = 0;
			break;
		}
H
Hugh Dickins 已提交
1603 1604
		if (page)
			unlock_page(page);
L
Linus Torvalds 已提交
1605 1606 1607

		/*
		 * We must evaluate after, since reads (unlike writes)
1608
		 * are called without i_mutex protection against truncate
L
Linus Torvalds 已提交
1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635
		 */
		nr = PAGE_CACHE_SIZE;
		i_size = i_size_read(inode);
		end_index = i_size >> PAGE_CACHE_SHIFT;
		if (index == end_index) {
			nr = i_size & ~PAGE_CACHE_MASK;
			if (nr <= offset) {
				if (page)
					page_cache_release(page);
				break;
			}
		}
		nr -= offset;

		if (page) {
			/*
			 * If users can be writing to this page using arbitrary
			 * virtual addresses, take care about potential aliasing
			 * before reading the page on the kernel side.
			 */
			if (mapping_writably_mapped(mapping))
				flush_dcache_page(page);
			/*
			 * Mark the page accessed if we read the beginning.
			 */
			if (!offset)
				mark_page_accessed(page);
N
Nick Piggin 已提交
1636
		} else {
L
Linus Torvalds 已提交
1637
			page = ZERO_PAGE(0);
N
Nick Piggin 已提交
1638 1639
			page_cache_get(page);
		}
L
Linus Torvalds 已提交
1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688

		/*
		 * Ok, we have the page, and it's up-to-date, so
		 * now we can copy it to user space...
		 *
		 * The actor routine returns how many bytes were actually used..
		 * NOTE! This may not be the same as how much of a user buffer
		 * we filled up (we may be padding etc), so we can only update
		 * "pos" here (the actor routine has to update the user buffer
		 * pointers and the remaining count).
		 */
		ret = actor(desc, page, offset, nr);
		offset += ret;
		index += offset >> PAGE_CACHE_SHIFT;
		offset &= ~PAGE_CACHE_MASK;

		page_cache_release(page);
		if (ret != nr || !desc->count)
			break;

		cond_resched();
	}

	*ppos = ((loff_t) index << PAGE_CACHE_SHIFT) + offset;
	file_accessed(filp);
}

static ssize_t shmem_file_read(struct file *filp, char __user *buf, size_t count, loff_t *ppos)
{
	read_descriptor_t desc;

	if ((ssize_t) count < 0)
		return -EINVAL;
	if (!access_ok(VERIFY_WRITE, buf, count))
		return -EFAULT;
	if (!count)
		return 0;

	desc.written = 0;
	desc.count = count;
	desc.arg.buf = buf;
	desc.error = 0;

	do_shmem_file_read(filp, ppos, &desc, file_read_actor);
	if (desc.written)
		return desc.written;
	return desc.error;
}

1689
static int shmem_statfs(struct dentry *dentry, struct kstatfs *buf)
L
Linus Torvalds 已提交
1690
{
1691
	struct shmem_sb_info *sbinfo = SHMEM_SB(dentry->d_sb);
L
Linus Torvalds 已提交
1692 1693 1694 1695

	buf->f_type = TMPFS_MAGIC;
	buf->f_bsize = PAGE_CACHE_SIZE;
	buf->f_namelen = NAME_MAX;
1696 1697
	spin_lock(&sbinfo->stat_lock);
	if (sbinfo->max_blocks) {
L
Linus Torvalds 已提交
1698 1699
		buf->f_blocks = sbinfo->max_blocks;
		buf->f_bavail = buf->f_bfree = sbinfo->free_blocks;
1700 1701
	}
	if (sbinfo->max_inodes) {
L
Linus Torvalds 已提交
1702 1703 1704 1705
		buf->f_files = sbinfo->max_inodes;
		buf->f_ffree = sbinfo->free_inodes;
	}
	/* else leave those fields 0 like simple_statfs */
1706
	spin_unlock(&sbinfo->stat_lock);
L
Linus Torvalds 已提交
1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719
	return 0;
}

/*
 * File creation. Allocate an inode, and we're done..
 */
static int
shmem_mknod(struct inode *dir, struct dentry *dentry, int mode, dev_t dev)
{
	struct inode *inode = shmem_get_inode(dir->i_sb, mode, dev);
	int error = -ENOSPC;

	if (inode) {
1720 1721 1722 1723 1724 1725 1726
		error = security_inode_init_security(inode, dir, NULL, NULL,
						     NULL);
		if (error) {
			if (error != -EOPNOTSUPP) {
				iput(inode);
				return error;
			}
1727 1728 1729 1730 1731
		}
		error = shmem_acl_init(inode, dir);
		if (error) {
			iput(inode);
			return error;
1732
		}
L
Linus Torvalds 已提交
1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751
		if (dir->i_mode & S_ISGID) {
			inode->i_gid = dir->i_gid;
			if (S_ISDIR(mode))
				inode->i_mode |= S_ISGID;
		}
		dir->i_size += BOGO_DIRENT_SIZE;
		dir->i_ctime = dir->i_mtime = CURRENT_TIME;
		d_instantiate(dentry, inode);
		dget(dentry); /* Extra count - pin the dentry in core */
	}
	return error;
}

static int shmem_mkdir(struct inode *dir, struct dentry *dentry, int mode)
{
	int error;

	if ((error = shmem_mknod(dir, dentry, mode | S_IFDIR, 0)))
		return error;
1752
	inc_nlink(dir);
L
Linus Torvalds 已提交
1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767
	return 0;
}

static int shmem_create(struct inode *dir, struct dentry *dentry, int mode,
		struct nameidata *nd)
{
	return shmem_mknod(dir, dentry, mode | S_IFREG, 0);
}

/*
 * Link a file..
 */
static int shmem_link(struct dentry *old_dentry, struct inode *dir, struct dentry *dentry)
{
	struct inode *inode = old_dentry->d_inode;
1768
	int ret;
L
Linus Torvalds 已提交
1769 1770 1771 1772 1773 1774

	/*
	 * No ordinary (disk based) filesystem counts links as inodes;
	 * but each new link needs a new dentry, pinning lowmem, and
	 * tmpfs dentries cannot be pruned until they are unlinked.
	 */
1775 1776 1777
	ret = shmem_reserve_inode(inode->i_sb);
	if (ret)
		goto out;
L
Linus Torvalds 已提交
1778 1779 1780

	dir->i_size += BOGO_DIRENT_SIZE;
	inode->i_ctime = dir->i_ctime = dir->i_mtime = CURRENT_TIME;
1781
	inc_nlink(inode);
L
Linus Torvalds 已提交
1782 1783 1784
	atomic_inc(&inode->i_count);	/* New dentry reference */
	dget(dentry);		/* Extra pinning count for the created dentry */
	d_instantiate(dentry, inode);
1785 1786
out:
	return ret;
L
Linus Torvalds 已提交
1787 1788 1789 1790 1791 1792
}

static int shmem_unlink(struct inode *dir, struct dentry *dentry)
{
	struct inode *inode = dentry->d_inode;

1793 1794
	if (inode->i_nlink > 1 && !S_ISDIR(inode->i_mode))
		shmem_free_inode(inode->i_sb);
L
Linus Torvalds 已提交
1795 1796 1797

	dir->i_size -= BOGO_DIRENT_SIZE;
	inode->i_ctime = dir->i_ctime = dir->i_mtime = CURRENT_TIME;
1798
	drop_nlink(inode);
L
Linus Torvalds 已提交
1799 1800 1801 1802 1803 1804 1805 1806 1807
	dput(dentry);	/* Undo the count from "create" - this does all the work */
	return 0;
}

static int shmem_rmdir(struct inode *dir, struct dentry *dentry)
{
	if (!simple_empty(dentry))
		return -ENOTEMPTY;

1808 1809
	drop_nlink(dentry->d_inode);
	drop_nlink(dir);
L
Linus Torvalds 已提交
1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829
	return shmem_unlink(dir, dentry);
}

/*
 * The VFS layer already does all the dentry stuff for rename,
 * we just have to decrement the usage count for the target if
 * it exists so that the VFS layer correctly free's it when it
 * gets overwritten.
 */
static int shmem_rename(struct inode *old_dir, struct dentry *old_dentry, struct inode *new_dir, struct dentry *new_dentry)
{
	struct inode *inode = old_dentry->d_inode;
	int they_are_dirs = S_ISDIR(inode->i_mode);

	if (!simple_empty(new_dentry))
		return -ENOTEMPTY;

	if (new_dentry->d_inode) {
		(void) shmem_unlink(new_dir, new_dentry);
		if (they_are_dirs)
1830
			drop_nlink(old_dir);
L
Linus Torvalds 已提交
1831
	} else if (they_are_dirs) {
1832
		drop_nlink(old_dir);
1833
		inc_nlink(new_dir);
L
Linus Torvalds 已提交
1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860
	}

	old_dir->i_size -= BOGO_DIRENT_SIZE;
	new_dir->i_size += BOGO_DIRENT_SIZE;
	old_dir->i_ctime = old_dir->i_mtime =
	new_dir->i_ctime = new_dir->i_mtime =
	inode->i_ctime = CURRENT_TIME;
	return 0;
}

static int shmem_symlink(struct inode *dir, struct dentry *dentry, const char *symname)
{
	int error;
	int len;
	struct inode *inode;
	struct page *page = NULL;
	char *kaddr;
	struct shmem_inode_info *info;

	len = strlen(symname) + 1;
	if (len > PAGE_CACHE_SIZE)
		return -ENAMETOOLONG;

	inode = shmem_get_inode(dir->i_sb, S_IFLNK|S_IRWXUGO, 0);
	if (!inode)
		return -ENOSPC;

1861 1862 1863 1864 1865 1866 1867 1868 1869 1870
	error = security_inode_init_security(inode, dir, NULL, NULL,
					     NULL);
	if (error) {
		if (error != -EOPNOTSUPP) {
			iput(inode);
			return error;
		}
		error = 0;
	}

L
Linus Torvalds 已提交
1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882
	info = SHMEM_I(inode);
	inode->i_size = len-1;
	if (len <= (char *)inode - (char *)info) {
		/* do it inline */
		memcpy(info, symname, len);
		inode->i_op = &shmem_symlink_inline_operations;
	} else {
		error = shmem_getpage(inode, 0, &page, SGP_WRITE, NULL);
		if (error) {
			iput(inode);
			return error;
		}
H
Hugh Dickins 已提交
1883
		unlock_page(page);
L
Linus Torvalds 已提交
1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899
		inode->i_op = &shmem_symlink_inode_operations;
		kaddr = kmap_atomic(page, KM_USER0);
		memcpy(kaddr, symname, len);
		kunmap_atomic(kaddr, KM_USER0);
		set_page_dirty(page);
		page_cache_release(page);
	}
	if (dir->i_mode & S_ISGID)
		inode->i_gid = dir->i_gid;
	dir->i_size += BOGO_DIRENT_SIZE;
	dir->i_ctime = dir->i_mtime = CURRENT_TIME;
	d_instantiate(dentry, inode);
	dget(dentry);
	return 0;
}

1900
static void *shmem_follow_link_inline(struct dentry *dentry, struct nameidata *nd)
L
Linus Torvalds 已提交
1901 1902
{
	nd_set_link(nd, (char *)SHMEM_I(dentry->d_inode));
1903
	return NULL;
L
Linus Torvalds 已提交
1904 1905
}

1906
static void *shmem_follow_link(struct dentry *dentry, struct nameidata *nd)
L
Linus Torvalds 已提交
1907 1908 1909 1910
{
	struct page *page = NULL;
	int res = shmem_getpage(dentry->d_inode, 0, &page, SGP_READ, NULL);
	nd_set_link(nd, res ? ERR_PTR(res) : kmap(page));
H
Hugh Dickins 已提交
1911 1912
	if (page)
		unlock_page(page);
1913
	return page;
L
Linus Torvalds 已提交
1914 1915
}

1916
static void shmem_put_link(struct dentry *dentry, struct nameidata *nd, void *cookie)
L
Linus Torvalds 已提交
1917 1918
{
	if (!IS_ERR(nd_get_link(nd))) {
1919
		struct page *page = cookie;
L
Linus Torvalds 已提交
1920 1921 1922 1923 1924 1925
		kunmap(page);
		mark_page_accessed(page);
		page_cache_release(page);
	}
}

1926
static const struct inode_operations shmem_symlink_inline_operations = {
L
Linus Torvalds 已提交
1927 1928 1929 1930
	.readlink	= generic_readlink,
	.follow_link	= shmem_follow_link_inline,
};

1931
static const struct inode_operations shmem_symlink_inode_operations = {
L
Linus Torvalds 已提交
1932 1933 1934 1935 1936 1937
	.truncate	= shmem_truncate,
	.readlink	= generic_readlink,
	.follow_link	= shmem_follow_link,
	.put_link	= shmem_put_link,
};

1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957
#ifdef CONFIG_TMPFS_POSIX_ACL
/**
 * Superblocks without xattr inode operations will get security.* xattr
 * support from the VFS "for free". As soon as we have any other xattrs
 * like ACLs, we also need to implement the security.* handlers at
 * filesystem level, though.
 */

static size_t shmem_xattr_security_list(struct inode *inode, char *list,
					size_t list_len, const char *name,
					size_t name_len)
{
	return security_inode_listsecurity(inode, list, list_len);
}

static int shmem_xattr_security_get(struct inode *inode, const char *name,
				    void *buffer, size_t size)
{
	if (strcmp(name, "") == 0)
		return -EINVAL;
1958
	return xattr_getsecurity(inode, name, buffer, size);
1959 1960 1961 1962 1963 1964 1965 1966 1967 1968
}

static int shmem_xattr_security_set(struct inode *inode, const char *name,
				    const void *value, size_t size, int flags)
{
	if (strcmp(name, "") == 0)
		return -EINVAL;
	return security_inode_setsecurity(inode, name, value, size, flags);
}

1969
static struct xattr_handler shmem_xattr_security_handler = {
1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983
	.prefix = XATTR_SECURITY_PREFIX,
	.list   = shmem_xattr_security_list,
	.get    = shmem_xattr_security_get,
	.set    = shmem_xattr_security_set,
};

static struct xattr_handler *shmem_xattr_handlers[] = {
	&shmem_xattr_acl_access_handler,
	&shmem_xattr_acl_default_handler,
	&shmem_xattr_security_handler,
	NULL
};
#endif

1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996
static struct dentry *shmem_get_parent(struct dentry *child)
{
	return ERR_PTR(-ESTALE);
}

static int shmem_match(struct inode *ino, void *vfh)
{
	__u32 *fh = vfh;
	__u64 inum = fh[2];
	inum = (inum << 32) | fh[1];
	return ino->i_ino == inum && fh[0] == ino->i_generation;
}

C
Christoph Hellwig 已提交
1997 1998
static struct dentry *shmem_fh_to_dentry(struct super_block *sb,
		struct fid *fid, int fh_len, int fh_type)
1999 2000
{
	struct inode *inode;
C
Christoph Hellwig 已提交
2001 2002 2003 2004 2005 2006
	struct dentry *dentry = NULL;
	u64 inum = fid->raw[2];
	inum = (inum << 32) | fid->raw[1];

	if (fh_len < 3)
		return NULL;
2007

C
Christoph Hellwig 已提交
2008 2009
	inode = ilookup5(sb, (unsigned long)(inum + fid->raw[0]),
			shmem_match, fid->raw);
2010
	if (inode) {
C
Christoph Hellwig 已提交
2011
		dentry = d_find_alias(inode);
2012 2013 2014
		iput(inode);
	}

C
Christoph Hellwig 已提交
2015
	return dentry;
2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047
}

static int shmem_encode_fh(struct dentry *dentry, __u32 *fh, int *len,
				int connectable)
{
	struct inode *inode = dentry->d_inode;

	if (*len < 3)
		return 255;

	if (hlist_unhashed(&inode->i_hash)) {
		/* Unfortunately insert_inode_hash is not idempotent,
		 * so as we hash inodes here rather than at creation
		 * time, we need a lock to ensure we only try
		 * to do it once
		 */
		static DEFINE_SPINLOCK(lock);
		spin_lock(&lock);
		if (hlist_unhashed(&inode->i_hash))
			__insert_inode_hash(inode,
					    inode->i_ino + inode->i_generation);
		spin_unlock(&lock);
	}

	fh[0] = inode->i_generation;
	fh[1] = inode->i_ino;
	fh[2] = ((__u64)inode->i_ino) >> 32;

	*len = 3;
	return 1;
}

2048
static const struct export_operations shmem_export_ops = {
2049 2050
	.get_parent     = shmem_get_parent,
	.encode_fh      = shmem_encode_fh,
C
Christoph Hellwig 已提交
2051
	.fh_to_dentry	= shmem_fh_to_dentry,
2052 2053
};

2054 2055 2056
static int shmem_parse_options(char *options, int *mode, uid_t *uid,
	gid_t *gid, unsigned long *blocks, unsigned long *inodes,
	int *policy, nodemask_t *policy_nodes)
L
Linus Torvalds 已提交
2057 2058 2059
{
	char *this_char, *value, *rest;

2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076
	while (options != NULL) {
		this_char = options;
		for (;;) {
			/*
			 * NUL-terminate this option: unfortunately,
			 * mount options form a comma-separated list,
			 * but mpol's nodelist may also contain commas.
			 */
			options = strchr(options, ',');
			if (options == NULL)
				break;
			options++;
			if (!isdigit(*options)) {
				options[-1] = '\0';
				break;
			}
		}
L
Linus Torvalds 已提交
2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098
		if (!*this_char)
			continue;
		if ((value = strchr(this_char,'=')) != NULL) {
			*value++ = 0;
		} else {
			printk(KERN_ERR
			    "tmpfs: No value for mount option '%s'\n",
			    this_char);
			return 1;
		}

		if (!strcmp(this_char,"size")) {
			unsigned long long size;
			size = memparse(value,&rest);
			if (*rest == '%') {
				size <<= PAGE_SHIFT;
				size *= totalram_pages;
				do_div(size, 100);
				rest++;
			}
			if (*rest)
				goto bad_val;
2099
			*blocks = DIV_ROUND_UP(size, PAGE_CACHE_SIZE);
L
Linus Torvalds 已提交
2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125
		} else if (!strcmp(this_char,"nr_blocks")) {
			*blocks = memparse(value,&rest);
			if (*rest)
				goto bad_val;
		} else if (!strcmp(this_char,"nr_inodes")) {
			*inodes = memparse(value,&rest);
			if (*rest)
				goto bad_val;
		} else if (!strcmp(this_char,"mode")) {
			if (!mode)
				continue;
			*mode = simple_strtoul(value,&rest,8);
			if (*rest)
				goto bad_val;
		} else if (!strcmp(this_char,"uid")) {
			if (!uid)
				continue;
			*uid = simple_strtoul(value,&rest,0);
			if (*rest)
				goto bad_val;
		} else if (!strcmp(this_char,"gid")) {
			if (!gid)
				continue;
			*gid = simple_strtoul(value,&rest,0);
			if (*rest)
				goto bad_val;
2126
		} else if (!strcmp(this_char,"mpol")) {
2127
			if (shmem_parse_mpol(value,policy,policy_nodes))
2128
				goto bad_val;
L
Linus Torvalds 已提交
2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146
		} else {
			printk(KERN_ERR "tmpfs: Bad mount option %s\n",
			       this_char);
			return 1;
		}
	}
	return 0;

bad_val:
	printk(KERN_ERR "tmpfs: Bad value '%s' for mount option '%s'\n",
	       value, this_char);
	return 1;

}

static int shmem_remount_fs(struct super_block *sb, int *flags, char *data)
{
	struct shmem_sb_info *sbinfo = SHMEM_SB(sb);
2147 2148
	unsigned long max_blocks = sbinfo->max_blocks;
	unsigned long max_inodes = sbinfo->max_inodes;
2149 2150
	int policy = sbinfo->policy;
	nodemask_t policy_nodes = sbinfo->policy_nodes;
2151 2152 2153 2154
	unsigned long blocks;
	unsigned long inodes;
	int error = -EINVAL;

2155 2156
	if (shmem_parse_options(data, NULL, NULL, NULL, &max_blocks,
				&max_inodes, &policy, &policy_nodes))
2157
		return error;
L
Linus Torvalds 已提交
2158

2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181
	spin_lock(&sbinfo->stat_lock);
	blocks = sbinfo->max_blocks - sbinfo->free_blocks;
	inodes = sbinfo->max_inodes - sbinfo->free_inodes;
	if (max_blocks < blocks)
		goto out;
	if (max_inodes < inodes)
		goto out;
	/*
	 * Those tests also disallow limited->unlimited while any are in
	 * use, so i_blocks will always be zero when max_blocks is zero;
	 * but we must separately disallow unlimited->limited, because
	 * in that case we have no record of how much is already in use.
	 */
	if (max_blocks && !sbinfo->max_blocks)
		goto out;
	if (max_inodes && !sbinfo->max_inodes)
		goto out;

	error = 0;
	sbinfo->max_blocks  = max_blocks;
	sbinfo->free_blocks = max_blocks - blocks;
	sbinfo->max_inodes  = max_inodes;
	sbinfo->free_inodes = max_inodes - inodes;
2182 2183
	sbinfo->policy = policy;
	sbinfo->policy_nodes = policy_nodes;
2184 2185 2186
out:
	spin_unlock(&sbinfo->stat_lock);
	return error;
L
Linus Torvalds 已提交
2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204
}
#endif

static void shmem_put_super(struct super_block *sb)
{
	kfree(sb->s_fs_info);
	sb->s_fs_info = NULL;
}

static int shmem_fill_super(struct super_block *sb,
			    void *data, int silent)
{
	struct inode *inode;
	struct dentry *root;
	int mode   = S_IRWXUGO | S_ISVTX;
	uid_t uid = current->fsuid;
	gid_t gid = current->fsgid;
	int err = -ENOMEM;
2205
	struct shmem_sb_info *sbinfo;
L
Linus Torvalds 已提交
2206 2207
	unsigned long blocks = 0;
	unsigned long inodes = 0;
2208
	int policy = MPOL_DEFAULT;
2209
	nodemask_t policy_nodes = node_states[N_HIGH_MEMORY];
L
Linus Torvalds 已提交
2210

2211
#ifdef CONFIG_TMPFS
L
Linus Torvalds 已提交
2212 2213 2214 2215 2216 2217 2218 2219 2220 2221
	/*
	 * Per default we only allow half of the physical ram per
	 * tmpfs instance, limiting inodes to one per page of lowmem;
	 * but the internal instance is left unlimited.
	 */
	if (!(sb->s_flags & MS_NOUSER)) {
		blocks = totalram_pages / 2;
		inodes = totalram_pages - totalhigh_pages;
		if (inodes > blocks)
			inodes = blocks;
2222 2223
		if (shmem_parse_options(data, &mode, &uid, &gid, &blocks,
					&inodes, &policy, &policy_nodes))
L
Linus Torvalds 已提交
2224 2225
			return -EINVAL;
	}
2226
	sb->s_export_op = &shmem_export_ops;
L
Linus Torvalds 已提交
2227 2228 2229 2230
#else
	sb->s_flags |= MS_NOUSER;
#endif

2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241
	/* Round up to L1_CACHE_BYTES to resist false sharing */
	sbinfo = kmalloc(max((int)sizeof(struct shmem_sb_info),
				L1_CACHE_BYTES), GFP_KERNEL);
	if (!sbinfo)
		return -ENOMEM;

	spin_lock_init(&sbinfo->stat_lock);
	sbinfo->max_blocks = blocks;
	sbinfo->free_blocks = blocks;
	sbinfo->max_inodes = inodes;
	sbinfo->free_inodes = inodes;
2242 2243
	sbinfo->policy = policy;
	sbinfo->policy_nodes = policy_nodes;
2244 2245

	sb->s_fs_info = sbinfo;
L
Linus Torvalds 已提交
2246 2247 2248 2249 2250
	sb->s_maxbytes = SHMEM_MAX_BYTES;
	sb->s_blocksize = PAGE_CACHE_SIZE;
	sb->s_blocksize_bits = PAGE_CACHE_SHIFT;
	sb->s_magic = TMPFS_MAGIC;
	sb->s_op = &shmem_ops;
2251
	sb->s_time_gran = 1;
2252 2253 2254 2255
#ifdef CONFIG_TMPFS_POSIX_ACL
	sb->s_xattr = shmem_xattr_handlers;
	sb->s_flags |= MS_POSIXACL;
#endif
2256

L
Linus Torvalds 已提交
2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274
	inode = shmem_get_inode(sb, S_IFDIR | mode, 0);
	if (!inode)
		goto failed;
	inode->i_uid = uid;
	inode->i_gid = gid;
	root = d_alloc_root(inode);
	if (!root)
		goto failed_iput;
	sb->s_root = root;
	return 0;

failed_iput:
	iput(inode);
failed:
	shmem_put_super(sb);
	return err;
}

2275
static struct kmem_cache *shmem_inode_cachep;
L
Linus Torvalds 已提交
2276 2277 2278 2279

static struct inode *shmem_alloc_inode(struct super_block *sb)
{
	struct shmem_inode_info *p;
2280
	p = (struct shmem_inode_info *)kmem_cache_alloc(shmem_inode_cachep, GFP_KERNEL);
L
Linus Torvalds 已提交
2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291
	if (!p)
		return NULL;
	return &p->vfs_inode;
}

static void shmem_destroy_inode(struct inode *inode)
{
	if ((inode->i_mode & S_IFMT) == S_IFREG) {
		/* only struct inode is valid if it's an inline symlink */
		mpol_free_shared_policy(&SHMEM_I(inode)->policy);
	}
2292
	shmem_acl_destroy_inode(inode);
L
Linus Torvalds 已提交
2293 2294 2295
	kmem_cache_free(shmem_inode_cachep, SHMEM_I(inode));
}

2296
static void init_once(struct kmem_cache *cachep, void *foo)
L
Linus Torvalds 已提交
2297 2298 2299
{
	struct shmem_inode_info *p = (struct shmem_inode_info *) foo;

C
Christoph Lameter 已提交
2300
	inode_init_once(&p->vfs_inode);
2301
#ifdef CONFIG_TMPFS_POSIX_ACL
C
Christoph Lameter 已提交
2302 2303
	p->i_acl = NULL;
	p->i_default_acl = NULL;
2304
#endif
L
Linus Torvalds 已提交
2305 2306 2307 2308 2309 2310
}

static int init_inodecache(void)
{
	shmem_inode_cachep = kmem_cache_create("shmem_inode_cache",
				sizeof(struct shmem_inode_info),
2311
				0, SLAB_PANIC, init_once);
L
Linus Torvalds 已提交
2312 2313 2314 2315 2316
	return 0;
}

static void destroy_inodecache(void)
{
2317
	kmem_cache_destroy(shmem_inode_cachep);
L
Linus Torvalds 已提交
2318 2319
}

2320
static const struct address_space_operations shmem_aops = {
L
Linus Torvalds 已提交
2321
	.writepage	= shmem_writepage,
2322
	.set_page_dirty	= __set_page_dirty_no_writeback,
L
Linus Torvalds 已提交
2323
#ifdef CONFIG_TMPFS
2324
	.readpage	= shmem_readpage,
N
Nick Piggin 已提交
2325 2326
	.write_begin	= shmem_write_begin,
	.write_end	= shmem_write_end,
L
Linus Torvalds 已提交
2327
#endif
2328
	.migratepage	= migrate_page,
L
Linus Torvalds 已提交
2329 2330
};

2331
static const struct file_operations shmem_file_operations = {
L
Linus Torvalds 已提交
2332 2333 2334 2335
	.mmap		= shmem_mmap,
#ifdef CONFIG_TMPFS
	.llseek		= generic_file_llseek,
	.read		= shmem_file_read,
H
Hugh Dickins 已提交
2336 2337
	.write		= do_sync_write,
	.aio_write	= generic_file_aio_write,
L
Linus Torvalds 已提交
2338
	.fsync		= simple_sync_file,
2339 2340
	.splice_read	= generic_file_splice_read,
	.splice_write	= generic_file_splice_write,
L
Linus Torvalds 已提交
2341 2342 2343
#endif
};

2344
static const struct inode_operations shmem_inode_operations = {
L
Linus Torvalds 已提交
2345 2346
	.truncate	= shmem_truncate,
	.setattr	= shmem_notify_change,
2347
	.truncate_range	= shmem_truncate_range,
2348 2349 2350 2351 2352 2353 2354 2355
#ifdef CONFIG_TMPFS_POSIX_ACL
	.setxattr	= generic_setxattr,
	.getxattr	= generic_getxattr,
	.listxattr	= generic_listxattr,
	.removexattr	= generic_removexattr,
	.permission	= shmem_permission,
#endif

L
Linus Torvalds 已提交
2356 2357
};

2358
static const struct inode_operations shmem_dir_inode_operations = {
L
Linus Torvalds 已提交
2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369
#ifdef CONFIG_TMPFS
	.create		= shmem_create,
	.lookup		= simple_lookup,
	.link		= shmem_link,
	.unlink		= shmem_unlink,
	.symlink	= shmem_symlink,
	.mkdir		= shmem_mkdir,
	.rmdir		= shmem_rmdir,
	.mknod		= shmem_mknod,
	.rename		= shmem_rename,
#endif
2370 2371 2372 2373 2374 2375 2376 2377 2378 2379
#ifdef CONFIG_TMPFS_POSIX_ACL
	.setattr	= shmem_notify_change,
	.setxattr	= generic_setxattr,
	.getxattr	= generic_getxattr,
	.listxattr	= generic_listxattr,
	.removexattr	= generic_removexattr,
	.permission	= shmem_permission,
#endif
};

2380
static const struct inode_operations shmem_special_inode_operations = {
2381 2382 2383 2384 2385 2386 2387 2388
#ifdef CONFIG_TMPFS_POSIX_ACL
	.setattr	= shmem_notify_change,
	.setxattr	= generic_setxattr,
	.getxattr	= generic_getxattr,
	.listxattr	= generic_listxattr,
	.removexattr	= generic_removexattr,
	.permission	= shmem_permission,
#endif
L
Linus Torvalds 已提交
2389 2390
};

2391
static const struct super_operations shmem_ops = {
L
Linus Torvalds 已提交
2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403
	.alloc_inode	= shmem_alloc_inode,
	.destroy_inode	= shmem_destroy_inode,
#ifdef CONFIG_TMPFS
	.statfs		= shmem_statfs,
	.remount_fs	= shmem_remount_fs,
#endif
	.delete_inode	= shmem_delete_inode,
	.drop_inode	= generic_delete_inode,
	.put_super	= shmem_put_super,
};

static struct vm_operations_struct shmem_vm_ops = {
2404
	.fault		= shmem_fault,
L
Linus Torvalds 已提交
2405 2406 2407 2408 2409 2410 2411
#ifdef CONFIG_NUMA
	.set_policy     = shmem_set_policy,
	.get_policy     = shmem_get_policy,
#endif
};


2412 2413
static int shmem_get_sb(struct file_system_type *fs_type,
	int flags, const char *dev_name, void *data, struct vfsmount *mnt)
L
Linus Torvalds 已提交
2414
{
2415
	return get_sb_nodev(fs_type, flags, data, shmem_fill_super, mnt);
L
Linus Torvalds 已提交
2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429
}

static struct file_system_type tmpfs_fs_type = {
	.owner		= THIS_MODULE,
	.name		= "tmpfs",
	.get_sb		= shmem_get_sb,
	.kill_sb	= kill_litter_super,
};
static struct vfsmount *shm_mnt;

static int __init init_tmpfs(void)
{
	int error;

P
Peter Zijlstra 已提交
2430 2431 2432 2433
	error = bdi_init(&shmem_backing_dev_info);
	if (error)
		goto out4;

L
Linus Torvalds 已提交
2434 2435 2436 2437 2438 2439 2440 2441 2442
	error = init_inodecache();
	if (error)
		goto out3;

	error = register_filesystem(&tmpfs_fs_type);
	if (error) {
		printk(KERN_ERR "Could not register tmpfs\n");
		goto out2;
	}
2443

2444
	shm_mnt = vfs_kern_mount(&tmpfs_fs_type, MS_NOUSER,
L
Linus Torvalds 已提交
2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457
				tmpfs_fs_type.name, NULL);
	if (IS_ERR(shm_mnt)) {
		error = PTR_ERR(shm_mnt);
		printk(KERN_ERR "Could not kern_mount tmpfs\n");
		goto out1;
	}
	return 0;

out1:
	unregister_filesystem(&tmpfs_fs_type);
out2:
	destroy_inodecache();
out3:
P
Peter Zijlstra 已提交
2458 2459
	bdi_destroy(&shmem_backing_dev_info);
out4:
L
Linus Torvalds 已提交
2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511
	shm_mnt = ERR_PTR(error);
	return error;
}
module_init(init_tmpfs)

/*
 * shmem_file_setup - get an unlinked file living in tmpfs
 *
 * @name: name for dentry (to be seen in /proc/<pid>/maps
 * @size: size to be set for the file
 *
 */
struct file *shmem_file_setup(char *name, loff_t size, unsigned long flags)
{
	int error;
	struct file *file;
	struct inode *inode;
	struct dentry *dentry, *root;
	struct qstr this;

	if (IS_ERR(shm_mnt))
		return (void *)shm_mnt;

	if (size < 0 || size > SHMEM_MAX_BYTES)
		return ERR_PTR(-EINVAL);

	if (shmem_acct_size(flags, size))
		return ERR_PTR(-ENOMEM);

	error = -ENOMEM;
	this.name = name;
	this.len = strlen(name);
	this.hash = 0; /* will go */
	root = shm_mnt->mnt_root;
	dentry = d_alloc(root, &this);
	if (!dentry)
		goto put_memory;

	error = -ENFILE;
	file = get_empty_filp();
	if (!file)
		goto put_dentry;

	error = -ENOSPC;
	inode = shmem_get_inode(root->d_sb, S_IFREG | S_IRWXUGO, 0);
	if (!inode)
		goto close_file;

	SHMEM_I(inode)->flags = flags & VM_ACCOUNT;
	d_instantiate(dentry, inode);
	inode->i_size = size;
	inode->i_nlink = 0;	/* It is unlinked */
2512 2513
	init_file(file, shm_mnt, dentry, FMODE_WRITE | FMODE_READ,
			&shmem_file_operations);
L
Linus Torvalds 已提交
2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543 2544
	return file;

close_file:
	put_filp(file);
put_dentry:
	dput(dentry);
put_memory:
	shmem_unacct_size(flags, size);
	return ERR_PTR(error);
}

/*
 * shmem_zero_setup - setup a shared anonymous mapping
 *
 * @vma: the vma to be mmapped is prepared by do_mmap_pgoff
 */
int shmem_zero_setup(struct vm_area_struct *vma)
{
	struct file *file;
	loff_t size = vma->vm_end - vma->vm_start;

	file = shmem_file_setup("dev/zero", size, vma->vm_flags);
	if (IS_ERR(file))
		return PTR_ERR(file);

	if (vma->vm_file)
		fput(vma->vm_file);
	vma->vm_file = file;
	vma->vm_ops = &shmem_vm_ops;
	return 0;
}