Blame view
mm/shmem.c
63.6 KB
1da177e4c
|
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. |
6922c0c7a
|
9 10 |
* Copyright (C) 2002-2011 Hugh Dickins. * Copyright (C) 2011 Google Inc. |
0edd73b33
|
11 |
* Copyright (C) 2002-2005 VERITAS Software Corporation. |
1da177e4c
|
12 13 14 15 16 17 |
* 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> * |
853ac43ab
|
18 19 20 |
* tiny-shmem: * Copyright (c) 2004, 2008 Matt Mackall <mpm@selenic.com> * |
1da177e4c
|
21 22 |
* This file is released under the GPL. */ |
853ac43ab
|
23 24 25 26 |
#include <linux/fs.h> #include <linux/init.h> #include <linux/vfs.h> #include <linux/mount.h> |
caefba174
|
27 |
#include <linux/pagemap.h> |
853ac43ab
|
28 29 30 31 32 33 34 35 |
#include <linux/file.h> #include <linux/mm.h> #include <linux/module.h> #include <linux/swap.h> static struct vfsmount *shm_mnt; #ifdef CONFIG_SHMEM |
1da177e4c
|
36 37 38 39 40 |
/* * 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. */ |
39f0247d3
|
41 |
#include <linux/xattr.h> |
a56942551
|
42 |
#include <linux/exportfs.h> |
1c7c474c3
|
43 |
#include <linux/posix_acl.h> |
39f0247d3
|
44 |
#include <linux/generic_acl.h> |
1da177e4c
|
45 |
#include <linux/mman.h> |
1da177e4c
|
46 47 48 49 |
#include <linux/string.h> #include <linux/slab.h> #include <linux/backing-dev.h> #include <linux/shmem_fs.h> |
1da177e4c
|
50 |
#include <linux/writeback.h> |
1da177e4c
|
51 |
#include <linux/blkdev.h> |
bda97eab0
|
52 |
#include <linux/pagevec.h> |
41ffe5d5c
|
53 |
#include <linux/percpu_counter.h> |
708e3508c
|
54 |
#include <linux/splice.h> |
1da177e4c
|
55 56 57 58 |
#include <linux/security.h> #include <linux/swapops.h> #include <linux/mempolicy.h> #include <linux/namei.h> |
b00dc3ad7
|
59 |
#include <linux/ctype.h> |
304dbdb7a
|
60 |
#include <linux/migrate.h> |
c1f60a5a4
|
61 |
#include <linux/highmem.h> |
680d794ba
|
62 |
#include <linux/seq_file.h> |
925629278
|
63 |
#include <linux/magic.h> |
304dbdb7a
|
64 |
|
1da177e4c
|
65 |
#include <asm/uaccess.h> |
1da177e4c
|
66 |
#include <asm/pgtable.h> |
caefba174
|
67 |
#define BLOCKS_PER_PAGE (PAGE_CACHE_SIZE/512) |
1da177e4c
|
68 |
#define VM_ACCT(size) (PAGE_CACHE_ALIGN(size) >> PAGE_SHIFT) |
1da177e4c
|
69 70 |
/* Pretend that each entry is of this size in directory's i_size */ #define BOGO_DIRENT_SIZE 20 |
69f07ec93
|
71 72 |
/* Symlink up to this size is kmalloc'ed instead of using a swappable page */ #define SHORT_SYMLINK_LEN 128 |
b09e0fa4b
|
73 74 75 76 77 78 |
struct shmem_xattr { struct list_head list; /* anchored by shmem_inode_info->xattr_list */ char *name; /* xattr name */ size_t size; char value[0]; }; |
285b2c4fd
|
79 |
/* Flag allocation requirements to shmem_getpage */ |
1da177e4c
|
80 |
enum sgp_type { |
1da177e4c
|
81 82 |
SGP_READ, /* don't exceed i_size, don't allocate page */ SGP_CACHE, /* don't exceed i_size, may allocate page */ |
a0ee5ec52
|
83 |
SGP_DIRTY, /* like SGP_CACHE, but set new page dirty */ |
1da177e4c
|
84 85 |
SGP_WRITE, /* may exceed i_size, may allocate page */ }; |
b76db7354
|
86 |
#ifdef CONFIG_TMPFS |
680d794ba
|
87 88 89 90 91 92 93 94 95 |
static unsigned long shmem_default_max_blocks(void) { return totalram_pages / 2; } static unsigned long shmem_default_max_inodes(void) { return min(totalram_pages - totalhigh_pages, totalram_pages / 2); } |
b76db7354
|
96 |
#endif |
680d794ba
|
97 |
|
68da9f055
|
98 99 100 101 102 103 104 105 106 |
static int shmem_getpage_gfp(struct inode *inode, pgoff_t index, struct page **pagep, enum sgp_type sgp, gfp_t gfp, int *fault_type); static inline int shmem_getpage(struct inode *inode, pgoff_t index, struct page **pagep, enum sgp_type sgp, int *fault_type) { return shmem_getpage_gfp(inode, index, pagep, sgp, mapping_gfp_mask(inode->i_mapping), fault_type); } |
1da177e4c
|
107 |
|
1da177e4c
|
108 109 110 111 112 113 114 115 116 117 118 119 120 |
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) { |
0b0a0806b
|
121 122 |
return (flags & VM_NORESERVE) ? 0 : security_vm_enough_memory_kern(VM_ACCT(size)); |
1da177e4c
|
123 124 125 126 |
} static inline void shmem_unacct_size(unsigned long flags, loff_t size) { |
0b0a0806b
|
127 |
if (!(flags & VM_NORESERVE)) |
1da177e4c
|
128 129 130 131 132 133 134 135 136 137 138 |
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) { |
0b0a0806b
|
139 140 |
return (flags & VM_NORESERVE) ? security_vm_enough_memory_kern(VM_ACCT(PAGE_CACHE_SIZE)) : 0; |
1da177e4c
|
141 142 143 144 |
} static inline void shmem_unacct_blocks(unsigned long flags, long pages) { |
0b0a0806b
|
145 |
if (flags & VM_NORESERVE) |
1da177e4c
|
146 147 |
vm_unacct_memory(pages * VM_ACCT(PAGE_CACHE_SIZE)); } |
759b9775c
|
148 |
static const struct super_operations shmem_ops; |
f5e54d6e5
|
149 |
static const struct address_space_operations shmem_aops; |
15ad7cdcf
|
150 |
static const struct file_operations shmem_file_operations; |
92e1d5be9
|
151 152 153 |
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; |
f0f37e2f7
|
154 |
static const struct vm_operations_struct shmem_vm_ops; |
1da177e4c
|
155 |
|
6c231b7ba
|
156 |
static struct backing_dev_info shmem_backing_dev_info __read_mostly = { |
1da177e4c
|
157 |
.ra_pages = 0, /* No readahead */ |
4f98a2fee
|
158 |
.capabilities = BDI_CAP_NO_ACCT_AND_WRITEBACK | BDI_CAP_SWAP_BACKED, |
1da177e4c
|
159 160 161 |
}; static LIST_HEAD(shmem_swaplist); |
cb5f7b9a4
|
162 |
static DEFINE_MUTEX(shmem_swaplist_mutex); |
1da177e4c
|
163 |
|
5b04c6890
|
164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 |
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); } } |
467118102
|
188 |
/** |
41ffe5d5c
|
189 |
* shmem_recalc_inode - recalculate the block usage of an inode |
1da177e4c
|
190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 |
* @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) { |
54af60421
|
207 208 209 |
struct shmem_sb_info *sbinfo = SHMEM_SB(inode->i_sb); if (sbinfo->max_blocks) percpu_counter_add(&sbinfo->used_blocks, -freed); |
1da177e4c
|
210 |
info->alloced -= freed; |
54af60421
|
211 |
inode->i_blocks -= freed * BLOCKS_PER_PAGE; |
1da177e4c
|
212 |
shmem_unacct_blocks(info->flags, freed); |
1da177e4c
|
213 214 |
} } |
7a5d0fbb2
|
215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 |
/* * Replace item expected in radix tree by a new item, while holding tree lock. */ static int shmem_radix_tree_replace(struct address_space *mapping, pgoff_t index, void *expected, void *replacement) { void **pslot; void *item = NULL; VM_BUG_ON(!expected); pslot = radix_tree_lookup_slot(&mapping->page_tree, index); if (pslot) item = radix_tree_deref_slot_protected(pslot, &mapping->tree_lock); if (item != expected) return -ENOENT; if (replacement) radix_tree_replace_slot(pslot, replacement); else radix_tree_delete(&mapping->page_tree, index); return 0; } /* |
46f65ec15
|
239 240 241 242 243 244 |
* Like add_to_page_cache_locked, but error if expected item has gone. */ static int shmem_add_to_page_cache(struct page *page, struct address_space *mapping, pgoff_t index, gfp_t gfp, void *expected) { |
aa3b18955
|
245 |
int error = 0; |
46f65ec15
|
246 247 248 |
VM_BUG_ON(!PageLocked(page)); VM_BUG_ON(!PageSwapBacked(page)); |
46f65ec15
|
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 |
if (!expected) error = radix_tree_preload(gfp & GFP_RECLAIM_MASK); if (!error) { page_cache_get(page); page->mapping = mapping; page->index = index; spin_lock_irq(&mapping->tree_lock); if (!expected) error = radix_tree_insert(&mapping->page_tree, index, page); else error = shmem_radix_tree_replace(mapping, index, expected, page); if (!error) { mapping->nrpages++; __inc_zone_page_state(page, NR_FILE_PAGES); __inc_zone_page_state(page, NR_SHMEM); spin_unlock_irq(&mapping->tree_lock); } else { page->mapping = NULL; spin_unlock_irq(&mapping->tree_lock); page_cache_release(page); } if (!expected) radix_tree_preload_end(); } if (error) mem_cgroup_uncharge_cache_page(page); |
46f65ec15
|
278 279 280 281 |
return error; } /* |
6922c0c7a
|
282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 |
* Like delete_from_page_cache, but substitutes swap for page. */ static void shmem_delete_from_page_cache(struct page *page, void *radswap) { struct address_space *mapping = page->mapping; int error; spin_lock_irq(&mapping->tree_lock); error = shmem_radix_tree_replace(mapping, page->index, page, radswap); page->mapping = NULL; mapping->nrpages--; __dec_zone_page_state(page, NR_FILE_PAGES); __dec_zone_page_state(page, NR_SHMEM); spin_unlock_irq(&mapping->tree_lock); page_cache_release(page); BUG_ON(error); } /* |
7a5d0fbb2
|
301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 |
* Like find_get_pages, but collecting swap entries as well as pages. */ static unsigned shmem_find_get_pages_and_swap(struct address_space *mapping, pgoff_t start, unsigned int nr_pages, struct page **pages, pgoff_t *indices) { unsigned int i; unsigned int ret; unsigned int nr_found; rcu_read_lock(); restart: nr_found = radix_tree_gang_lookup_slot(&mapping->page_tree, (void ***)pages, indices, start, nr_pages); ret = 0; for (i = 0; i < nr_found; i++) { struct page *page; repeat: page = radix_tree_deref_slot((void **)pages[i]); if (unlikely(!page)) continue; if (radix_tree_exception(page)) { |
8079b1c85
|
323 324 325 326 327 328 329 330 |
if (radix_tree_deref_retry(page)) goto restart; /* * Otherwise, we must be storing a swap entry * here as an exceptional entry: so return it * without attempting to raise page count. */ goto export; |
7a5d0fbb2
|
331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 |
} if (!page_cache_get_speculative(page)) goto repeat; /* Has the page moved? */ if (unlikely(page != *((void **)pages[i]))) { page_cache_release(page); goto repeat; } export: indices[ret] = indices[i]; pages[ret] = page; ret++; } if (unlikely(!ret && nr_found)) goto restart; rcu_read_unlock(); return ret; } /* * Remove swap entry from radix tree, free the swap and its page cache. */ static int shmem_free_swap(struct address_space *mapping, pgoff_t index, void *radswap) { int error; spin_lock_irq(&mapping->tree_lock); error = shmem_radix_tree_replace(mapping, index, radswap, NULL); spin_unlock_irq(&mapping->tree_lock); if (!error) free_swap_and_cache(radix_to_swp_entry(radswap)); return error; } /* * Pagevec may contain swap entries, so shuffle up pages before releasing. */ static void shmem_pagevec_release(struct pagevec *pvec) { int i, j; for (i = 0, j = 0; i < pagevec_count(pvec); i++) { struct page *page = pvec->pages[i]; if (!radix_tree_exceptional_entry(page)) pvec->pages[j++] = page; } pvec->nr = j; pagevec_release(pvec); } /* * Remove range of pages and swap entries from radix tree, and free them. */ |
285b2c4fd
|
386 |
void shmem_truncate_range(struct inode *inode, loff_t lstart, loff_t lend) |
1da177e4c
|
387 |
{ |
285b2c4fd
|
388 |
struct address_space *mapping = inode->i_mapping; |
1da177e4c
|
389 |
struct shmem_inode_info *info = SHMEM_I(inode); |
285b2c4fd
|
390 |
pgoff_t start = (lstart + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; |
bda97eab0
|
391 |
unsigned partial = lstart & (PAGE_CACHE_SIZE - 1); |
285b2c4fd
|
392 |
pgoff_t end = (lend >> PAGE_CACHE_SHIFT); |
bda97eab0
|
393 |
struct pagevec pvec; |
7a5d0fbb2
|
394 395 |
pgoff_t indices[PAGEVEC_SIZE]; long nr_swaps_freed = 0; |
285b2c4fd
|
396 |
pgoff_t index; |
bda97eab0
|
397 398 399 400 401 402 |
int i; BUG_ON((lend & (PAGE_CACHE_SIZE - 1)) != (PAGE_CACHE_SIZE - 1)); pagevec_init(&pvec, 0); index = start; |
7a5d0fbb2
|
403 404 405 406 407 408 |
while (index <= end) { pvec.nr = shmem_find_get_pages_and_swap(mapping, index, min(end - index, (pgoff_t)PAGEVEC_SIZE - 1) + 1, pvec.pages, indices); if (!pvec.nr) break; |
bda97eab0
|
409 410 411 |
mem_cgroup_uncharge_start(); for (i = 0; i < pagevec_count(&pvec); i++) { struct page *page = pvec.pages[i]; |
7a5d0fbb2
|
412 |
index = indices[i]; |
bda97eab0
|
413 414 |
if (index > end) break; |
7a5d0fbb2
|
415 416 417 |
if (radix_tree_exceptional_entry(page)) { nr_swaps_freed += !shmem_free_swap(mapping, index, page); |
bda97eab0
|
418 |
continue; |
7a5d0fbb2
|
419 420 421 |
} if (!trylock_page(page)) |
bda97eab0
|
422 |
continue; |
7a5d0fbb2
|
423 424 425 |
if (page->mapping == mapping) { VM_BUG_ON(PageWriteback(page)); truncate_inode_page(mapping, page); |
bda97eab0
|
426 |
} |
bda97eab0
|
427 428 |
unlock_page(page); } |
7a5d0fbb2
|
429 |
shmem_pagevec_release(&pvec); |
bda97eab0
|
430 431 432 433 |
mem_cgroup_uncharge_end(); cond_resched(); index++; } |
1da177e4c
|
434 |
|
bda97eab0
|
435 436 437 438 439 440 441 442 443 444 445 446 447 448 |
if (partial) { struct page *page = NULL; shmem_getpage(inode, start - 1, &page, SGP_READ, NULL); if (page) { zero_user_segment(page, partial, PAGE_CACHE_SIZE); set_page_dirty(page); unlock_page(page); page_cache_release(page); } } index = start; for ( ; ; ) { cond_resched(); |
7a5d0fbb2
|
449 450 451 452 |
pvec.nr = shmem_find_get_pages_and_swap(mapping, index, min(end - index, (pgoff_t)PAGEVEC_SIZE - 1) + 1, pvec.pages, indices); if (!pvec.nr) { |
bda97eab0
|
453 454 455 456 457 |
if (index == start) break; index = start; continue; } |
7a5d0fbb2
|
458 459 |
if (index == start && indices[0] > end) { shmem_pagevec_release(&pvec); |
bda97eab0
|
460 461 462 463 464 |
break; } mem_cgroup_uncharge_start(); for (i = 0; i < pagevec_count(&pvec); i++) { struct page *page = pvec.pages[i]; |
7a5d0fbb2
|
465 |
index = indices[i]; |
bda97eab0
|
466 467 |
if (index > end) break; |
7a5d0fbb2
|
468 469 470 471 472 |
if (radix_tree_exceptional_entry(page)) { nr_swaps_freed += !shmem_free_swap(mapping, index, page); continue; } |
bda97eab0
|
473 |
lock_page(page); |
7a5d0fbb2
|
474 475 476 477 |
if (page->mapping == mapping) { VM_BUG_ON(PageWriteback(page)); truncate_inode_page(mapping, page); } |
bda97eab0
|
478 479 |
unlock_page(page); } |
7a5d0fbb2
|
480 |
shmem_pagevec_release(&pvec); |
bda97eab0
|
481 482 483 |
mem_cgroup_uncharge_end(); index++; } |
94c1e62df
|
484 |
|
1da177e4c
|
485 |
spin_lock(&info->lock); |
7a5d0fbb2
|
486 |
info->swapped -= nr_swaps_freed; |
1da177e4c
|
487 488 |
shmem_recalc_inode(inode); spin_unlock(&info->lock); |
285b2c4fd
|
489 |
inode->i_ctime = inode->i_mtime = CURRENT_TIME; |
1da177e4c
|
490 |
} |
94c1e62df
|
491 |
EXPORT_SYMBOL_GPL(shmem_truncate_range); |
1da177e4c
|
492 |
|
94c1e62df
|
493 |
static int shmem_setattr(struct dentry *dentry, struct iattr *attr) |
1da177e4c
|
494 495 |
{ struct inode *inode = dentry->d_inode; |
1da177e4c
|
496 |
int error; |
db78b877f
|
497 498 499 |
error = inode_change_ok(inode, attr); if (error) return error; |
94c1e62df
|
500 501 502 |
if (S_ISREG(inode->i_mode) && (attr->ia_valid & ATTR_SIZE)) { loff_t oldsize = inode->i_size; loff_t newsize = attr->ia_size; |
3889e6e76
|
503 |
|
94c1e62df
|
504 505 506 507 508 509 510 511 512 513 514 |
if (newsize != oldsize) { i_size_write(inode, newsize); inode->i_ctime = inode->i_mtime = CURRENT_TIME; } if (newsize < oldsize) { loff_t holebegin = round_up(newsize, PAGE_SIZE); unmap_mapping_range(inode->i_mapping, holebegin, 0, 1); shmem_truncate_range(inode, newsize, (loff_t)-1); /* unmap again to remove racily COWed private pages */ unmap_mapping_range(inode->i_mapping, holebegin, 0, 1); } |
1da177e4c
|
515 |
} |
db78b877f
|
516 |
setattr_copy(inode, attr); |
39f0247d3
|
517 |
#ifdef CONFIG_TMPFS_POSIX_ACL |
db78b877f
|
518 |
if (attr->ia_valid & ATTR_MODE) |
1c7c474c3
|
519 |
error = generic_acl_chmod(inode); |
39f0247d3
|
520 |
#endif |
1da177e4c
|
521 522 |
return error; } |
1f895f75d
|
523 |
static void shmem_evict_inode(struct inode *inode) |
1da177e4c
|
524 |
{ |
1da177e4c
|
525 |
struct shmem_inode_info *info = SHMEM_I(inode); |
b09e0fa4b
|
526 |
struct shmem_xattr *xattr, *nxattr; |
1da177e4c
|
527 |
|
3889e6e76
|
528 |
if (inode->i_mapping->a_ops == &shmem_aops) { |
1da177e4c
|
529 530 |
shmem_unacct_size(info->flags, inode->i_size); inode->i_size = 0; |
3889e6e76
|
531 |
shmem_truncate_range(inode, 0, (loff_t)-1); |
1da177e4c
|
532 |
if (!list_empty(&info->swaplist)) { |
cb5f7b9a4
|
533 |
mutex_lock(&shmem_swaplist_mutex); |
1da177e4c
|
534 |
list_del_init(&info->swaplist); |
cb5f7b9a4
|
535 |
mutex_unlock(&shmem_swaplist_mutex); |
1da177e4c
|
536 |
} |
69f07ec93
|
537 538 |
} else kfree(info->symlink); |
b09e0fa4b
|
539 540 541 542 543 |
list_for_each_entry_safe(xattr, nxattr, &info->xattr_list, list) { kfree(xattr->name); kfree(xattr); } |
0edd73b33
|
544 |
BUG_ON(inode->i_blocks); |
5b04c6890
|
545 |
shmem_free_inode(inode->i_sb); |
1f895f75d
|
546 |
end_writeback(inode); |
1da177e4c
|
547 |
} |
46f65ec15
|
548 549 550 |
/* * If swap found in inode, free it and move page from swapcache to filecache. */ |
41ffe5d5c
|
551 552 |
static int shmem_unuse_inode(struct shmem_inode_info *info, swp_entry_t swap, struct page *page) |
1da177e4c
|
553 |
{ |
285b2c4fd
|
554 |
struct address_space *mapping = info->vfs_inode.i_mapping; |
46f65ec15
|
555 |
void *radswap; |
41ffe5d5c
|
556 |
pgoff_t index; |
d9fe526a8
|
557 |
int error; |
1da177e4c
|
558 |
|
46f65ec15
|
559 |
radswap = swp_to_radix_entry(swap); |
e504f3fdd
|
560 |
index = radix_tree_locate_item(&mapping->page_tree, radswap); |
46f65ec15
|
561 |
if (index == -1) |
285b2c4fd
|
562 |
return 0; |
2e0e26c76
|
563 |
|
1b1b32f2c
|
564 565 |
/* * Move _head_ to start search for next from here. |
1f895f75d
|
566 |
* But be careful: shmem_evict_inode checks list_empty without taking |
1b1b32f2c
|
567 |
* mutex, and there's an instant in list_move_tail when info->swaplist |
285b2c4fd
|
568 |
* would appear empty, if it were the only one on shmem_swaplist. |
1b1b32f2c
|
569 570 571 |
*/ if (shmem_swaplist.next != &info->swaplist) list_move_tail(&shmem_swaplist, &info->swaplist); |
2e0e26c76
|
572 |
|
d13d14430
|
573 |
/* |
778dd893a
|
574 575 576 |
* We rely on shmem_swaplist_mutex, not only to protect the swaplist, * but also to hold up shmem_evict_inode(): so inode cannot be freed * beneath us (pagelock doesn't help until the page is in pagecache). |
d13d14430
|
577 |
*/ |
46f65ec15
|
578 579 |
error = shmem_add_to_page_cache(page, mapping, index, GFP_NOWAIT, radswap); |
778dd893a
|
580 |
/* which does mem_cgroup_uncharge_cache_page on error */ |
69029cd55
|
581 |
|
48f170fb7
|
582 |
if (error != -ENOMEM) { |
46f65ec15
|
583 584 585 586 |
/* * Truncation and eviction use free_swap_and_cache(), which * only does trylock page: if we raced, best clean up here. */ |
73b1262fa
|
587 588 |
delete_from_swap_cache(page); set_page_dirty(page); |
46f65ec15
|
589 590 591 592 593 594 |
if (!error) { spin_lock(&info->lock); info->swapped--; spin_unlock(&info->lock); swap_free(swap); } |
2e0e26c76
|
595 |
error = 1; /* not an error, but entry was found */ |
1da177e4c
|
596 |
} |
2e0e26c76
|
597 |
return error; |
1da177e4c
|
598 599 600 |
} /* |
46f65ec15
|
601 |
* Search through swapped inodes to find and replace swap by page. |
1da177e4c
|
602 |
*/ |
41ffe5d5c
|
603 |
int shmem_unuse(swp_entry_t swap, struct page *page) |
1da177e4c
|
604 |
{ |
41ffe5d5c
|
605 |
struct list_head *this, *next; |
1da177e4c
|
606 607 |
struct shmem_inode_info *info; int found = 0; |
778dd893a
|
608 609 610 611 612 613 |
int error; /* * Charge page using GFP_KERNEL while we can wait, before taking * the shmem_swaplist_mutex which might hold up shmem_writepage(). * Charged back to the user (not to caller) when swap account is used. |
778dd893a
|
614 615 616 617 |
*/ error = mem_cgroup_cache_charge(page, current->mm, GFP_KERNEL); if (error) goto out; |
46f65ec15
|
618 |
/* No radix_tree_preload: swap entry keeps a place for page in tree */ |
1da177e4c
|
619 |
|
cb5f7b9a4
|
620 |
mutex_lock(&shmem_swaplist_mutex); |
41ffe5d5c
|
621 622 |
list_for_each_safe(this, next, &shmem_swaplist) { info = list_entry(this, struct shmem_inode_info, swaplist); |
285b2c4fd
|
623 |
if (info->swapped) |
41ffe5d5c
|
624 |
found = shmem_unuse_inode(info, swap, page); |
6922c0c7a
|
625 626 |
else list_del_init(&info->swaplist); |
cb5f7b9a4
|
627 |
cond_resched(); |
2e0e26c76
|
628 |
if (found) |
778dd893a
|
629 |
break; |
1da177e4c
|
630 |
} |
cb5f7b9a4
|
631 |
mutex_unlock(&shmem_swaplist_mutex); |
778dd893a
|
632 |
|
778dd893a
|
633 634 635 636 637 |
if (!found) mem_cgroup_uncharge_cache_page(page); if (found < 0) error = found; out: |
aaa468653
|
638 639 |
unlock_page(page); page_cache_release(page); |
778dd893a
|
640 |
return error; |
1da177e4c
|
641 642 643 644 645 646 647 648 |
} /* * 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; |
1da177e4c
|
649 |
struct address_space *mapping; |
1da177e4c
|
650 |
struct inode *inode; |
6922c0c7a
|
651 652 |
swp_entry_t swap; pgoff_t index; |
1da177e4c
|
653 654 |
BUG_ON(!PageLocked(page)); |
1da177e4c
|
655 656 657 658 659 660 |
mapping = page->mapping; index = page->index; inode = mapping->host; info = SHMEM_I(inode); if (info->flags & VM_LOCKED) goto redirty; |
d9fe526a8
|
661 |
if (!total_swap_pages) |
1da177e4c
|
662 |
goto redirty; |
d9fe526a8
|
663 664 665 |
/* * shmem_backing_dev_info's capabilities prevent regular writeback or * sync from ever calling shmem_writepage; but a stacking filesystem |
48f170fb7
|
666 |
* might use ->writepage of its underlying filesystem, in which case |
d9fe526a8
|
667 |
* tmpfs should write out to swap only in response to memory pressure, |
48f170fb7
|
668 |
* and not for the writeback threads or sync. |
d9fe526a8
|
669 |
*/ |
48f170fb7
|
670 671 672 673 674 675 676 |
if (!wbc->for_reclaim) { WARN_ON_ONCE(1); /* Still happens? Tell us about it! */ goto redirty; } swap = get_swap_page(); if (!swap.val) goto redirty; |
d9fe526a8
|
677 |
|
b1dea800a
|
678 679 |
/* * Add inode to shmem_unuse()'s list of swapped-out inodes, |
6922c0c7a
|
680 681 |
* if it's not already there. Do it now before the page is * moved to swap cache, when its pagelock no longer protects |
b1dea800a
|
682 |
* the inode from eviction. But don't unlock the mutex until |
6922c0c7a
|
683 684 |
* we've incremented swapped, because shmem_unuse_inode() will * prune a !swapped inode from the swaplist under this mutex. |
b1dea800a
|
685 |
*/ |
48f170fb7
|
686 687 688 |
mutex_lock(&shmem_swaplist_mutex); if (list_empty(&info->swaplist)) list_add_tail(&info->swaplist, &shmem_swaplist); |
b1dea800a
|
689 |
|
48f170fb7
|
690 |
if (add_to_swap_cache(page, swap, GFP_ATOMIC) == 0) { |
aaa468653
|
691 |
swap_shmem_alloc(swap); |
6922c0c7a
|
692 693 694 695 696 |
shmem_delete_from_page_cache(page, swp_to_radix_entry(swap)); spin_lock(&info->lock); info->swapped++; shmem_recalc_inode(inode); |
826267cf1
|
697 |
spin_unlock(&info->lock); |
6922c0c7a
|
698 699 |
mutex_unlock(&shmem_swaplist_mutex); |
d9fe526a8
|
700 |
BUG_ON(page_mapped(page)); |
9fab5619b
|
701 |
swap_writepage(page, wbc); |
1da177e4c
|
702 703 |
return 0; } |
6922c0c7a
|
704 |
mutex_unlock(&shmem_swaplist_mutex); |
cb4b86ba4
|
705 |
swapcache_free(swap, NULL); |
1da177e4c
|
706 707 |
redirty: set_page_dirty(page); |
d9fe526a8
|
708 709 710 711 |
if (wbc->for_reclaim) return AOP_WRITEPAGE_ACTIVATE; /* Return with page locked */ unlock_page(page); return 0; |
1da177e4c
|
712 713 714 |
} #ifdef CONFIG_NUMA |
680d794ba
|
715 |
#ifdef CONFIG_TMPFS |
71fe804b6
|
716 |
static void shmem_show_mpol(struct seq_file *seq, struct mempolicy *mpol) |
680d794ba
|
717 |
{ |
095f1fc4e
|
718 |
char buffer[64]; |
680d794ba
|
719 |
|
71fe804b6
|
720 |
if (!mpol || mpol->mode == MPOL_DEFAULT) |
095f1fc4e
|
721 |
return; /* show nothing */ |
680d794ba
|
722 |
|
71fe804b6
|
723 |
mpol_to_str(buffer, sizeof(buffer), mpol, 1); |
095f1fc4e
|
724 725 |
seq_printf(seq, ",mpol=%s", buffer); |
680d794ba
|
726 |
} |
71fe804b6
|
727 728 729 730 731 732 733 734 735 736 737 738 |
static struct mempolicy *shmem_get_sbmpol(struct shmem_sb_info *sbinfo) { struct mempolicy *mpol = NULL; if (sbinfo->mpol) { spin_lock(&sbinfo->stat_lock); /* prevent replace/use races */ mpol = sbinfo->mpol; mpol_get(mpol); spin_unlock(&sbinfo->stat_lock); } return mpol; } |
680d794ba
|
739 |
#endif /* CONFIG_TMPFS */ |
41ffe5d5c
|
740 741 |
static struct page *shmem_swapin(swp_entry_t swap, gfp_t gfp, struct shmem_inode_info *info, pgoff_t index) |
1da177e4c
|
742 |
{ |
52cd3b074
|
743 |
struct mempolicy mpol, *spol; |
1da177e4c
|
744 |
struct vm_area_struct pvma; |
52cd3b074
|
745 |
spol = mpol_cond_copy(&mpol, |
41ffe5d5c
|
746 |
mpol_shared_policy_lookup(&info->policy, index)); |
52cd3b074
|
747 |
|
1da177e4c
|
748 |
/* Create a pseudo vma that just contains the policy */ |
c4cc6d07b
|
749 |
pvma.vm_start = 0; |
41ffe5d5c
|
750 |
pvma.vm_pgoff = index; |
c4cc6d07b
|
751 |
pvma.vm_ops = NULL; |
52cd3b074
|
752 |
pvma.vm_policy = spol; |
41ffe5d5c
|
753 |
return swapin_readahead(swap, gfp, &pvma, 0); |
1da177e4c
|
754 |
} |
02098feaa
|
755 |
static struct page *shmem_alloc_page(gfp_t gfp, |
41ffe5d5c
|
756 |
struct shmem_inode_info *info, pgoff_t index) |
1da177e4c
|
757 758 |
{ struct vm_area_struct pvma; |
1da177e4c
|
759 |
|
c4cc6d07b
|
760 761 |
/* Create a pseudo vma that just contains the policy */ pvma.vm_start = 0; |
41ffe5d5c
|
762 |
pvma.vm_pgoff = index; |
c4cc6d07b
|
763 |
pvma.vm_ops = NULL; |
41ffe5d5c
|
764 |
pvma.vm_policy = mpol_shared_policy_lookup(&info->policy, index); |
52cd3b074
|
765 766 767 768 769 |
/* * alloc_page_vma() will drop the shared policy reference */ return alloc_page_vma(gfp, &pvma, 0); |
1da177e4c
|
770 |
} |
680d794ba
|
771 772 |
#else /* !CONFIG_NUMA */ #ifdef CONFIG_TMPFS |
41ffe5d5c
|
773 |
static inline void shmem_show_mpol(struct seq_file *seq, struct mempolicy *mpol) |
680d794ba
|
774 775 776 |
{ } #endif /* CONFIG_TMPFS */ |
41ffe5d5c
|
777 778 |
static inline struct page *shmem_swapin(swp_entry_t swap, gfp_t gfp, struct shmem_inode_info *info, pgoff_t index) |
1da177e4c
|
779 |
{ |
41ffe5d5c
|
780 |
return swapin_readahead(swap, gfp, NULL, 0); |
1da177e4c
|
781 |
} |
02098feaa
|
782 |
static inline struct page *shmem_alloc_page(gfp_t gfp, |
41ffe5d5c
|
783 |
struct shmem_inode_info *info, pgoff_t index) |
1da177e4c
|
784 |
{ |
e84e2e132
|
785 |
return alloc_page(gfp); |
1da177e4c
|
786 |
} |
680d794ba
|
787 |
#endif /* CONFIG_NUMA */ |
1da177e4c
|
788 |
|
71fe804b6
|
789 790 791 792 793 794 |
#if !defined(CONFIG_NUMA) || !defined(CONFIG_TMPFS) static inline struct mempolicy *shmem_get_sbmpol(struct shmem_sb_info *sbinfo) { return NULL; } #endif |
1da177e4c
|
795 |
/* |
68da9f055
|
796 |
* shmem_getpage_gfp - find page in cache, or get from swap, or allocate |
1da177e4c
|
797 798 799 800 801 |
* * 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 */ |
41ffe5d5c
|
802 |
static int shmem_getpage_gfp(struct inode *inode, pgoff_t index, |
68da9f055
|
803 |
struct page **pagep, enum sgp_type sgp, gfp_t gfp, int *fault_type) |
1da177e4c
|
804 805 |
{ struct address_space *mapping = inode->i_mapping; |
54af60421
|
806 |
struct shmem_inode_info *info; |
1da177e4c
|
807 |
struct shmem_sb_info *sbinfo; |
27ab70062
|
808 |
struct page *page; |
1da177e4c
|
809 810 |
swp_entry_t swap; int error; |
54af60421
|
811 |
int once = 0; |
1da177e4c
|
812 |
|
41ffe5d5c
|
813 |
if (index > (MAX_LFS_FILESIZE >> PAGE_CACHE_SHIFT)) |
1da177e4c
|
814 |
return -EFBIG; |
1da177e4c
|
815 |
repeat: |
54af60421
|
816 |
swap.val = 0; |
41ffe5d5c
|
817 |
page = find_lock_page(mapping, index); |
54af60421
|
818 819 820 821 822 823 824 825 826 827 828 829 |
if (radix_tree_exceptional_entry(page)) { swap = radix_to_swp_entry(page); page = NULL; } if (sgp != SGP_WRITE && ((loff_t)index << PAGE_CACHE_SHIFT) >= i_size_read(inode)) { error = -EINVAL; goto failed; } if (page || (sgp == SGP_READ && !swap.val)) { |
b409f9fcf
|
830 |
/* |
27ab70062
|
831 832 833 |
* Once we can get the page lock, it must be uptodate: * if there were an error in reading back from swap, * the page would not be inserted into the filecache. |
b409f9fcf
|
834 |
*/ |
54af60421
|
835 836 837 |
BUG_ON(page && !PageUptodate(page)); *pagep = page; return 0; |
27ab70062
|
838 839 840 |
} /* |
54af60421
|
841 842 |
* Fast cache lookup did not find it: * bring it back from swap or allocate. |
27ab70062
|
843 |
*/ |
54af60421
|
844 845 |
info = SHMEM_I(inode); sbinfo = SHMEM_SB(inode->i_sb); |
1da177e4c
|
846 |
|
1da177e4c
|
847 848 |
if (swap.val) { /* Look it up and read it in.. */ |
27ab70062
|
849 850 |
page = lookup_swap_cache(swap); if (!page) { |
1da177e4c
|
851 |
/* here we actually do the io */ |
68da9f055
|
852 853 |
if (fault_type) *fault_type |= VM_FAULT_MAJOR; |
41ffe5d5c
|
854 |
page = shmem_swapin(swap, gfp, info, index); |
27ab70062
|
855 |
if (!page) { |
54af60421
|
856 857 |
error = -ENOMEM; goto failed; |
1da177e4c
|
858 |
} |
1da177e4c
|
859 860 861 |
} /* We have to do this with page locked to prevent races */ |
54af60421
|
862 |
lock_page(page); |
27ab70062
|
863 |
if (!PageUptodate(page)) { |
1da177e4c
|
864 |
error = -EIO; |
54af60421
|
865 |
goto failed; |
1da177e4c
|
866 |
} |
54af60421
|
867 868 869 870 871 872 873 874 875 |
wait_on_page_writeback(page); /* Someone may have already done it for us */ if (page->mapping) { if (page->mapping == mapping && page->index == index) goto done; error = -EEXIST; goto failed; |
1da177e4c
|
876 |
} |
27ab70062
|
877 |
|
aa3b18955
|
878 879 880 881 882 |
error = mem_cgroup_cache_charge(page, current->mm, gfp & GFP_RECLAIM_MASK); if (!error) error = shmem_add_to_page_cache(page, mapping, index, gfp, swp_to_radix_entry(swap)); |
54af60421
|
883 884 885 886 |
if (error) goto failed; spin_lock(&info->lock); |
285b2c4fd
|
887 |
info->swapped--; |
54af60421
|
888 |
shmem_recalc_inode(inode); |
27ab70062
|
889 |
spin_unlock(&info->lock); |
54af60421
|
890 891 |
delete_from_swap_cache(page); |
27ab70062
|
892 893 |
set_page_dirty(page); swap_free(swap); |
54af60421
|
894 895 896 897 |
} else { if (shmem_acct_block(info->flags)) { error = -ENOSPC; goto failed; |
1da177e4c
|
898 |
} |
0edd73b33
|
899 |
if (sbinfo->max_blocks) { |
fc5da22ae
|
900 |
if (percpu_counter_compare(&sbinfo->used_blocks, |
54af60421
|
901 902 903 904 |
sbinfo->max_blocks) >= 0) { error = -ENOSPC; goto unacct; } |
7e496299d
|
905 |
percpu_counter_inc(&sbinfo->used_blocks); |
54af60421
|
906 |
} |
1da177e4c
|
907 |
|
54af60421
|
908 909 910 911 |
page = shmem_alloc_page(gfp, info, index); if (!page) { error = -ENOMEM; goto decused; |
1da177e4c
|
912 |
} |
54af60421
|
913 914 |
SetPageSwapBacked(page); __set_page_locked(page); |
aa3b18955
|
915 916 917 918 919 |
error = mem_cgroup_cache_charge(page, current->mm, gfp & GFP_RECLAIM_MASK); if (!error) error = shmem_add_to_page_cache(page, mapping, index, gfp, NULL); |
54af60421
|
920 921 922 923 924 |
if (error) goto decused; lru_cache_add_anon(page); spin_lock(&info->lock); |
1da177e4c
|
925 |
info->alloced++; |
54af60421
|
926 927 |
inode->i_blocks += BLOCKS_PER_PAGE; shmem_recalc_inode(inode); |
1da177e4c
|
928 |
spin_unlock(&info->lock); |
54af60421
|
929 |
|
27ab70062
|
930 931 932 |
clear_highpage(page); flush_dcache_page(page); SetPageUptodate(page); |
a0ee5ec52
|
933 |
if (sgp == SGP_DIRTY) |
27ab70062
|
934 |
set_page_dirty(page); |
1da177e4c
|
935 936 |
} done: |
54af60421
|
937 938 939 940 941 |
/* Perhaps the file has been truncated since we checked */ if (sgp != SGP_WRITE && ((loff_t)index << PAGE_CACHE_SHIFT) >= i_size_read(inode)) { error = -EINVAL; goto trunc; |
e83c32e8f
|
942 |
} |
54af60421
|
943 944 |
*pagep = page; return 0; |
1da177e4c
|
945 |
|
59a16ead5
|
946 |
/* |
54af60421
|
947 |
* Error recovery. |
59a16ead5
|
948 |
*/ |
54af60421
|
949 950 951 952 953 954 |
trunc: ClearPageDirty(page); delete_from_page_cache(page); spin_lock(&info->lock); info->alloced--; inode->i_blocks -= BLOCKS_PER_PAGE; |
59a16ead5
|
955 |
spin_unlock(&info->lock); |
54af60421
|
956 957 958 959 960 961 962 963 964 965 966 967 968 969 |
decused: if (sbinfo->max_blocks) percpu_counter_add(&sbinfo->used_blocks, -1); unacct: shmem_unacct_blocks(info->flags, 1); failed: if (swap.val && error != -EINVAL) { struct page *test = find_get_page(mapping, index); if (test && !radix_tree_exceptional_entry(test)) page_cache_release(test); /* Have another try if the entry has changed */ if (test != swp_to_radix_entry(swap)) error = -EEXIST; } |
27ab70062
|
970 |
if (page) { |
54af60421
|
971 |
unlock_page(page); |
27ab70062
|
972 |
page_cache_release(page); |
54af60421
|
973 974 975 976 977 978 |
} if (error == -ENOSPC && !once++) { info = SHMEM_I(inode); spin_lock(&info->lock); shmem_recalc_inode(inode); spin_unlock(&info->lock); |
27ab70062
|
979 |
goto repeat; |
ff36b8016
|
980 |
} |
54af60421
|
981 982 983 |
if (error == -EEXIST) goto repeat; return error; |
1da177e4c
|
984 |
} |
d0217ac04
|
985 |
static int shmem_fault(struct vm_area_struct *vma, struct vm_fault *vmf) |
1da177e4c
|
986 |
{ |
d3ac7f892
|
987 |
struct inode *inode = vma->vm_file->f_path.dentry->d_inode; |
1da177e4c
|
988 |
int error; |
68da9f055
|
989 |
int ret = VM_FAULT_LOCKED; |
1da177e4c
|
990 |
|
27d54b398
|
991 |
error = shmem_getpage(inode, vmf->pgoff, &vmf->page, SGP_CACHE, &ret); |
d0217ac04
|
992 993 |
if (error) return ((error == -ENOMEM) ? VM_FAULT_OOM : VM_FAULT_SIGBUS); |
68da9f055
|
994 |
|
456f998ec
|
995 996 997 998 |
if (ret & VM_FAULT_MAJOR) { count_vm_event(PGMAJFAULT); mem_cgroup_count_vm_event(vma->vm_mm, PGMAJFAULT); } |
68da9f055
|
999 |
return ret; |
1da177e4c
|
1000 |
} |
1da177e4c
|
1001 |
#ifdef CONFIG_NUMA |
41ffe5d5c
|
1002 |
static int shmem_set_policy(struct vm_area_struct *vma, struct mempolicy *mpol) |
1da177e4c
|
1003 |
{ |
41ffe5d5c
|
1004 1005 |
struct inode *inode = vma->vm_file->f_path.dentry->d_inode; return mpol_set_shared_policy(&SHMEM_I(inode)->policy, vma, mpol); |
1da177e4c
|
1006 |
} |
d8dc74f21
|
1007 1008 |
static struct mempolicy *shmem_get_policy(struct vm_area_struct *vma, unsigned long addr) |
1da177e4c
|
1009 |
{ |
41ffe5d5c
|
1010 1011 |
struct inode *inode = vma->vm_file->f_path.dentry->d_inode; pgoff_t index; |
1da177e4c
|
1012 |
|
41ffe5d5c
|
1013 1014 |
index = ((addr - vma->vm_start) >> PAGE_SHIFT) + vma->vm_pgoff; return mpol_shared_policy_lookup(&SHMEM_I(inode)->policy, index); |
1da177e4c
|
1015 1016 1017 1018 1019 |
} #endif int shmem_lock(struct file *file, int lock, struct user_struct *user) { |
d3ac7f892
|
1020 |
struct inode *inode = file->f_path.dentry->d_inode; |
1da177e4c
|
1021 1022 1023 1024 1025 1026 1027 1028 |
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; |
89e004ea5
|
1029 |
mapping_set_unevictable(file->f_mapping); |
1da177e4c
|
1030 1031 1032 1033 |
} if (!lock && (info->flags & VM_LOCKED) && user) { user_shm_unlock(inode->i_size, user); info->flags &= ~VM_LOCKED; |
89e004ea5
|
1034 1035 |
mapping_clear_unevictable(file->f_mapping); scan_mapping_unevictable_pages(file->f_mapping); |
1da177e4c
|
1036 1037 |
} retval = 0; |
89e004ea5
|
1038 |
|
1da177e4c
|
1039 1040 1041 1042 |
out_nomem: spin_unlock(&info->lock); return retval; } |
9b83a6a85
|
1043 |
static int shmem_mmap(struct file *file, struct vm_area_struct *vma) |
1da177e4c
|
1044 1045 1046 |
{ file_accessed(file); vma->vm_ops = &shmem_vm_ops; |
d0217ac04
|
1047 |
vma->vm_flags |= VM_CAN_NONLINEAR; |
1da177e4c
|
1048 1049 |
return 0; } |
454abafe9
|
1050 1051 |
static struct inode *shmem_get_inode(struct super_block *sb, const struct inode *dir, int mode, dev_t dev, unsigned long flags) |
1da177e4c
|
1052 1053 1054 1055 |
{ struct inode *inode; struct shmem_inode_info *info; struct shmem_sb_info *sbinfo = SHMEM_SB(sb); |
5b04c6890
|
1056 1057 |
if (shmem_reserve_inode(sb)) return NULL; |
1da177e4c
|
1058 1059 1060 |
inode = new_inode(sb); if (inode) { |
85fe4025c
|
1061 |
inode->i_ino = get_next_ino(); |
454abafe9
|
1062 |
inode_init_owner(inode, dir, mode); |
1da177e4c
|
1063 |
inode->i_blocks = 0; |
1da177e4c
|
1064 1065 |
inode->i_mapping->backing_dev_info = &shmem_backing_dev_info; inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME; |
91828a405
|
1066 |
inode->i_generation = get_seconds(); |
1da177e4c
|
1067 1068 1069 |
info = SHMEM_I(inode); memset(info, 0, (char *)inode - (char *)info); spin_lock_init(&info->lock); |
0b0a0806b
|
1070 |
info->flags = flags & VM_NORESERVE; |
1da177e4c
|
1071 |
INIT_LIST_HEAD(&info->swaplist); |
b09e0fa4b
|
1072 |
INIT_LIST_HEAD(&info->xattr_list); |
72c04902d
|
1073 |
cache_no_acl(inode); |
1da177e4c
|
1074 1075 1076 |
switch (mode & S_IFMT) { default: |
39f0247d3
|
1077 |
inode->i_op = &shmem_special_inode_operations; |
1da177e4c
|
1078 1079 1080 |
init_special_inode(inode, mode, dev); break; case S_IFREG: |
14fcc23fd
|
1081 |
inode->i_mapping->a_ops = &shmem_aops; |
1da177e4c
|
1082 1083 |
inode->i_op = &shmem_inode_operations; inode->i_fop = &shmem_file_operations; |
71fe804b6
|
1084 1085 |
mpol_shared_policy_init(&info->policy, shmem_get_sbmpol(sbinfo)); |
1da177e4c
|
1086 1087 |
break; case S_IFDIR: |
d8c76e6f4
|
1088 |
inc_nlink(inode); |
1da177e4c
|
1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 |
/* 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. */ |
71fe804b6
|
1099 |
mpol_shared_policy_init(&info->policy, NULL); |
1da177e4c
|
1100 1101 |
break; } |
5b04c6890
|
1102 1103 |
} else shmem_free_inode(sb); |
1da177e4c
|
1104 1105 1106 1107 |
return inode; } #ifdef CONFIG_TMPFS |
92e1d5be9
|
1108 |
static const struct inode_operations shmem_symlink_inode_operations; |
69f07ec93
|
1109 |
static const struct inode_operations shmem_short_symlink_operations; |
1da177e4c
|
1110 |
|
1da177e4c
|
1111 |
static int |
800d15a53
|
1112 1113 1114 |
shmem_write_begin(struct file *file, struct address_space *mapping, loff_t pos, unsigned len, unsigned flags, struct page **pagep, void **fsdata) |
1da177e4c
|
1115 |
{ |
800d15a53
|
1116 1117 |
struct inode *inode = mapping->host; pgoff_t index = pos >> PAGE_CACHE_SHIFT; |
800d15a53
|
1118 1119 1120 1121 1122 1123 1124 1125 1126 |
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; |
d3602444e
|
1127 1128 |
if (pos + copied > inode->i_size) i_size_write(inode, pos + copied); |
800d15a53
|
1129 |
set_page_dirty(page); |
6746aff74
|
1130 |
unlock_page(page); |
800d15a53
|
1131 |
page_cache_release(page); |
800d15a53
|
1132 |
return copied; |
1da177e4c
|
1133 |
} |
1da177e4c
|
1134 1135 |
static void do_shmem_file_read(struct file *filp, loff_t *ppos, read_descriptor_t *desc, read_actor_t actor) { |
d3ac7f892
|
1136 |
struct inode *inode = filp->f_path.dentry->d_inode; |
1da177e4c
|
1137 |
struct address_space *mapping = inode->i_mapping; |
41ffe5d5c
|
1138 1139 |
pgoff_t index; unsigned long offset; |
a0ee5ec52
|
1140 1141 1142 1143 1144 1145 1146 1147 1148 |
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; |
1da177e4c
|
1149 1150 1151 1152 1153 1154 |
index = *ppos >> PAGE_CACHE_SHIFT; offset = *ppos & ~PAGE_CACHE_MASK; for (;;) { struct page *page = NULL; |
41ffe5d5c
|
1155 1156 |
pgoff_t end_index; unsigned long nr, ret; |
1da177e4c
|
1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 |
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; } |
a0ee5ec52
|
1167 |
desc->error = shmem_getpage(inode, index, &page, sgp, NULL); |
1da177e4c
|
1168 1169 1170 1171 1172 |
if (desc->error) { if (desc->error == -EINVAL) desc->error = 0; break; } |
d3602444e
|
1173 1174 |
if (page) unlock_page(page); |
1da177e4c
|
1175 1176 1177 |
/* * We must evaluate after, since reads (unlike writes) |
1b1dcc1b5
|
1178 |
* are called without i_mutex protection against truncate |
1da177e4c
|
1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 |
*/ 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); |
b5810039a
|
1206 |
} else { |
1da177e4c
|
1207 |
page = ZERO_PAGE(0); |
b5810039a
|
1208 1209 |
page_cache_get(page); } |
1da177e4c
|
1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 |
/* * 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); } |
bcd78e496
|
1236 1237 1238 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 |
static ssize_t shmem_file_aio_read(struct kiocb *iocb, const struct iovec *iov, unsigned long nr_segs, loff_t pos) { struct file *filp = iocb->ki_filp; ssize_t retval; unsigned long seg; size_t count; loff_t *ppos = &iocb->ki_pos; retval = generic_segment_checks(iov, &nr_segs, &count, VERIFY_WRITE); if (retval) return retval; for (seg = 0; seg < nr_segs; seg++) { read_descriptor_t desc; desc.written = 0; desc.arg.buf = iov[seg].iov_base; desc.count = iov[seg].iov_len; if (desc.count == 0) continue; desc.error = 0; do_shmem_file_read(filp, ppos, &desc, file_read_actor); retval += desc.written; if (desc.error) { retval = retval ?: desc.error; break; } if (desc.count > 0) break; } return retval; |
1da177e4c
|
1268 |
} |
708e3508c
|
1269 1270 1271 1272 1273 |
static ssize_t shmem_file_splice_read(struct file *in, loff_t *ppos, struct pipe_inode_info *pipe, size_t len, unsigned int flags) { struct address_space *mapping = in->f_mapping; |
71f0e07a6
|
1274 |
struct inode *inode = mapping->host; |
708e3508c
|
1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 |
unsigned int loff, nr_pages, req_pages; struct page *pages[PIPE_DEF_BUFFERS]; struct partial_page partial[PIPE_DEF_BUFFERS]; struct page *page; pgoff_t index, end_index; loff_t isize, left; int error, page_nr; struct splice_pipe_desc spd = { .pages = pages, .partial = partial, .flags = flags, .ops = &page_cache_pipe_buf_ops, .spd_release = spd_release_page, }; |
71f0e07a6
|
1289 |
isize = i_size_read(inode); |
708e3508c
|
1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 |
if (unlikely(*ppos >= isize)) return 0; left = isize - *ppos; if (unlikely(left < len)) len = left; if (splice_grow_spd(pipe, &spd)) return -ENOMEM; index = *ppos >> PAGE_CACHE_SHIFT; loff = *ppos & ~PAGE_CACHE_MASK; req_pages = (len + loff + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; nr_pages = min(req_pages, pipe->buffers); |
708e3508c
|
1304 1305 1306 |
spd.nr_pages = find_get_pages_contig(mapping, index, nr_pages, spd.pages); index += spd.nr_pages; |
708e3508c
|
1307 |
error = 0; |
708e3508c
|
1308 |
|
71f0e07a6
|
1309 |
while (spd.nr_pages < nr_pages) { |
71f0e07a6
|
1310 1311 1312 1313 |
error = shmem_getpage(inode, index, &page, SGP_CACHE, NULL); if (error) break; unlock_page(page); |
708e3508c
|
1314 1315 1316 |
spd.pages[spd.nr_pages++] = page; index++; } |
708e3508c
|
1317 1318 1319 |
index = *ppos >> PAGE_CACHE_SHIFT; nr_pages = spd.nr_pages; spd.nr_pages = 0; |
71f0e07a6
|
1320 |
|
708e3508c
|
1321 1322 1323 1324 1325 |
for (page_nr = 0; page_nr < nr_pages; page_nr++) { unsigned int this_len; if (!len) break; |
708e3508c
|
1326 1327 |
this_len = min_t(unsigned long, len, PAGE_CACHE_SIZE - loff); page = spd.pages[page_nr]; |
71f0e07a6
|
1328 |
if (!PageUptodate(page) || page->mapping != mapping) { |
71f0e07a6
|
1329 1330 1331 |
error = shmem_getpage(inode, index, &page, SGP_CACHE, NULL); if (error) |
708e3508c
|
1332 |
break; |
71f0e07a6
|
1333 1334 1335 |
unlock_page(page); page_cache_release(spd.pages[page_nr]); spd.pages[page_nr] = page; |
708e3508c
|
1336 |
} |
71f0e07a6
|
1337 1338 |
isize = i_size_read(inode); |
708e3508c
|
1339 1340 1341 |
end_index = (isize - 1) >> PAGE_CACHE_SHIFT; if (unlikely(!isize || index > end_index)) break; |
708e3508c
|
1342 1343 |
if (end_index == index) { unsigned int plen; |
708e3508c
|
1344 1345 1346 |
plen = ((isize - 1) & ~PAGE_CACHE_MASK) + 1; if (plen <= loff) break; |
708e3508c
|
1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 |
this_len = min(this_len, plen - loff); len = this_len; } spd.partial[page_nr].offset = loff; spd.partial[page_nr].len = this_len; len -= this_len; loff = 0; spd.nr_pages++; index++; } |
708e3508c
|
1358 1359 |
while (page_nr < nr_pages) page_cache_release(spd.pages[page_nr++]); |
708e3508c
|
1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 |
if (spd.nr_pages) error = splice_to_pipe(pipe, &spd); splice_shrink_spd(pipe, &spd); if (error > 0) { *ppos += error; file_accessed(in); } return error; } |
726c33422
|
1372 |
static int shmem_statfs(struct dentry *dentry, struct kstatfs *buf) |
1da177e4c
|
1373 |
{ |
726c33422
|
1374 |
struct shmem_sb_info *sbinfo = SHMEM_SB(dentry->d_sb); |
1da177e4c
|
1375 1376 1377 1378 |
buf->f_type = TMPFS_MAGIC; buf->f_bsize = PAGE_CACHE_SIZE; buf->f_namelen = NAME_MAX; |
0edd73b33
|
1379 |
if (sbinfo->max_blocks) { |
1da177e4c
|
1380 |
buf->f_blocks = sbinfo->max_blocks; |
41ffe5d5c
|
1381 1382 1383 |
buf->f_bavail = buf->f_bfree = sbinfo->max_blocks - percpu_counter_sum(&sbinfo->used_blocks); |
0edd73b33
|
1384 1385 |
} if (sbinfo->max_inodes) { |
1da177e4c
|
1386 1387 |
buf->f_files = sbinfo->max_inodes; buf->f_ffree = sbinfo->free_inodes; |
1da177e4c
|
1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 |
} /* else leave those fields 0 like simple_statfs */ 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) { |
0b0a0806b
|
1399 |
struct inode *inode; |
1da177e4c
|
1400 |
int error = -ENOSPC; |
454abafe9
|
1401 |
inode = shmem_get_inode(dir->i_sb, dir, mode, dev, VM_NORESERVE); |
1da177e4c
|
1402 |
if (inode) { |
2a7dba391
|
1403 1404 1405 |
error = security_inode_init_security(inode, dir, &dentry->d_name, NULL, NULL, NULL); |
570bc1c2e
|
1406 1407 1408 1409 1410 |
if (error) { if (error != -EOPNOTSUPP) { iput(inode); return error; } |
39f0247d3
|
1411 |
} |
1c7c474c3
|
1412 1413 |
#ifdef CONFIG_TMPFS_POSIX_ACL error = generic_acl_init(inode, dir); |
39f0247d3
|
1414 1415 1416 |
if (error) { iput(inode); return error; |
570bc1c2e
|
1417 |
} |
718deb6b6
|
1418 1419 |
#else error = 0; |
1c7c474c3
|
1420 |
#endif |
1da177e4c
|
1421 1422 1423 1424 |
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 */ |
1da177e4c
|
1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 |
} 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; |
d8c76e6f4
|
1435 |
inc_nlink(dir); |
1da177e4c
|
1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 |
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; |
5b04c6890
|
1451 |
int ret; |
1da177e4c
|
1452 1453 1454 1455 1456 1457 |
/* * 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. */ |
5b04c6890
|
1458 1459 1460 |
ret = shmem_reserve_inode(inode->i_sb); if (ret) goto out; |
1da177e4c
|
1461 1462 1463 |
dir->i_size += BOGO_DIRENT_SIZE; inode->i_ctime = dir->i_ctime = dir->i_mtime = CURRENT_TIME; |
d8c76e6f4
|
1464 |
inc_nlink(inode); |
7de9c6ee3
|
1465 |
ihold(inode); /* New dentry reference */ |
1da177e4c
|
1466 1467 |
dget(dentry); /* Extra pinning count for the created dentry */ d_instantiate(dentry, inode); |
5b04c6890
|
1468 1469 |
out: return ret; |
1da177e4c
|
1470 1471 1472 1473 1474 |
} static int shmem_unlink(struct inode *dir, struct dentry *dentry) { struct inode *inode = dentry->d_inode; |
5b04c6890
|
1475 1476 |
if (inode->i_nlink > 1 && !S_ISDIR(inode->i_mode)) shmem_free_inode(inode->i_sb); |
1da177e4c
|
1477 1478 1479 |
dir->i_size -= BOGO_DIRENT_SIZE; inode->i_ctime = dir->i_ctime = dir->i_mtime = CURRENT_TIME; |
9a53c3a78
|
1480 |
drop_nlink(inode); |
1da177e4c
|
1481 1482 1483 1484 1485 1486 1487 1488 |
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; |
9a53c3a78
|
1489 1490 |
drop_nlink(dentry->d_inode); drop_nlink(dir); |
1da177e4c
|
1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 |
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) |
9a53c3a78
|
1511 |
drop_nlink(old_dir); |
1da177e4c
|
1512 |
} else if (they_are_dirs) { |
9a53c3a78
|
1513 |
drop_nlink(old_dir); |
d8c76e6f4
|
1514 |
inc_nlink(new_dir); |
1da177e4c
|
1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 |
} 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; |
9276aad6c
|
1530 |
struct page *page; |
1da177e4c
|
1531 1532 1533 1534 1535 1536 |
char *kaddr; struct shmem_inode_info *info; len = strlen(symname) + 1; if (len > PAGE_CACHE_SIZE) return -ENAMETOOLONG; |
454abafe9
|
1537 |
inode = shmem_get_inode(dir->i_sb, dir, S_IFLNK|S_IRWXUGO, 0, VM_NORESERVE); |
1da177e4c
|
1538 1539 |
if (!inode) return -ENOSPC; |
2a7dba391
|
1540 1541 |
error = security_inode_init_security(inode, dir, &dentry->d_name, NULL, NULL, NULL); |
570bc1c2e
|
1542 1543 1544 1545 1546 1547 1548 |
if (error) { if (error != -EOPNOTSUPP) { iput(inode); return error; } error = 0; } |
1da177e4c
|
1549 1550 |
info = SHMEM_I(inode); inode->i_size = len-1; |
69f07ec93
|
1551 1552 1553 1554 1555 1556 1557 |
if (len <= SHORT_SYMLINK_LEN) { info->symlink = kmemdup(symname, len, GFP_KERNEL); if (!info->symlink) { iput(inode); return -ENOMEM; } inode->i_op = &shmem_short_symlink_operations; |
1da177e4c
|
1558 1559 1560 1561 1562 1563 |
} else { error = shmem_getpage(inode, 0, &page, SGP_WRITE, NULL); if (error) { iput(inode); return error; } |
14fcc23fd
|
1564 |
inode->i_mapping->a_ops = &shmem_aops; |
1da177e4c
|
1565 1566 1567 1568 1569 |
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); |
6746aff74
|
1570 |
unlock_page(page); |
1da177e4c
|
1571 1572 |
page_cache_release(page); } |
1da177e4c
|
1573 1574 1575 1576 1577 1578 |
dir->i_size += BOGO_DIRENT_SIZE; dir->i_ctime = dir->i_mtime = CURRENT_TIME; d_instantiate(dentry, inode); dget(dentry); return 0; } |
69f07ec93
|
1579 |
static void *shmem_follow_short_symlink(struct dentry *dentry, struct nameidata *nd) |
1da177e4c
|
1580 |
{ |
69f07ec93
|
1581 |
nd_set_link(nd, SHMEM_I(dentry->d_inode)->symlink); |
cc314eef0
|
1582 |
return NULL; |
1da177e4c
|
1583 |
} |
cc314eef0
|
1584 |
static void *shmem_follow_link(struct dentry *dentry, struct nameidata *nd) |
1da177e4c
|
1585 1586 |
{ struct page *page = NULL; |
41ffe5d5c
|
1587 1588 |
int error = shmem_getpage(dentry->d_inode, 0, &page, SGP_READ, NULL); nd_set_link(nd, error ? ERR_PTR(error) : kmap(page)); |
d3602444e
|
1589 1590 |
if (page) unlock_page(page); |
cc314eef0
|
1591 |
return page; |
1da177e4c
|
1592 |
} |
cc314eef0
|
1593 |
static void shmem_put_link(struct dentry *dentry, struct nameidata *nd, void *cookie) |
1da177e4c
|
1594 1595 |
{ if (!IS_ERR(nd_get_link(nd))) { |
cc314eef0
|
1596 |
struct page *page = cookie; |
1da177e4c
|
1597 1598 1599 |
kunmap(page); mark_page_accessed(page); page_cache_release(page); |
1da177e4c
|
1600 1601 |
} } |
b09e0fa4b
|
1602 |
#ifdef CONFIG_TMPFS_XATTR |
467118102
|
1603 |
/* |
b09e0fa4b
|
1604 1605 |
* Superblocks without xattr inode operations may get some security.* xattr * support from the LSM "for free". As soon as we have any other xattrs |
39f0247d3
|
1606 1607 1608 |
* like ACLs, we also need to implement the security.* handlers at * filesystem level, though. */ |
b09e0fa4b
|
1609 1610 |
static int shmem_xattr_get(struct dentry *dentry, const char *name, void *buffer, size_t size) |
39f0247d3
|
1611 |
{ |
b09e0fa4b
|
1612 1613 1614 |
struct shmem_inode_info *info; struct shmem_xattr *xattr; int ret = -ENODATA; |
39f0247d3
|
1615 |
|
b09e0fa4b
|
1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 |
info = SHMEM_I(dentry->d_inode); spin_lock(&info->lock); list_for_each_entry(xattr, &info->xattr_list, list) { if (strcmp(name, xattr->name)) continue; ret = xattr->size; if (buffer) { if (size < xattr->size) ret = -ERANGE; else memcpy(buffer, xattr->value, xattr->size); } break; } spin_unlock(&info->lock); return ret; |
39f0247d3
|
1634 |
} |
b09e0fa4b
|
1635 1636 |
static int shmem_xattr_set(struct dentry *dentry, const char *name, const void *value, size_t size, int flags) |
39f0247d3
|
1637 |
{ |
b09e0fa4b
|
1638 1639 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 1689 1690 1691 1692 |
struct inode *inode = dentry->d_inode; struct shmem_inode_info *info = SHMEM_I(inode); struct shmem_xattr *xattr; struct shmem_xattr *new_xattr = NULL; size_t len; int err = 0; /* value == NULL means remove */ if (value) { /* wrap around? */ len = sizeof(*new_xattr) + size; if (len <= sizeof(*new_xattr)) return -ENOMEM; new_xattr = kmalloc(len, GFP_KERNEL); if (!new_xattr) return -ENOMEM; new_xattr->name = kstrdup(name, GFP_KERNEL); if (!new_xattr->name) { kfree(new_xattr); return -ENOMEM; } new_xattr->size = size; memcpy(new_xattr->value, value, size); } spin_lock(&info->lock); list_for_each_entry(xattr, &info->xattr_list, list) { if (!strcmp(name, xattr->name)) { if (flags & XATTR_CREATE) { xattr = new_xattr; err = -EEXIST; } else if (new_xattr) { list_replace(&xattr->list, &new_xattr->list); } else { list_del(&xattr->list); } goto out; } } if (flags & XATTR_REPLACE) { xattr = new_xattr; err = -ENODATA; } else { list_add(&new_xattr->list, &info->xattr_list); xattr = NULL; } out: spin_unlock(&info->lock); if (xattr) kfree(xattr->name); kfree(xattr); return err; |
39f0247d3
|
1693 |
} |
bb4354538
|
1694 |
static const struct xattr_handler *shmem_xattr_handlers[] = { |
b09e0fa4b
|
1695 |
#ifdef CONFIG_TMPFS_POSIX_ACL |
1c7c474c3
|
1696 1697 |
&generic_acl_access_handler, &generic_acl_default_handler, |
b09e0fa4b
|
1698 |
#endif |
39f0247d3
|
1699 1700 |
NULL }; |
b09e0fa4b
|
1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 |
static int shmem_xattr_validate(const char *name) { struct { const char *prefix; size_t len; } arr[] = { { XATTR_SECURITY_PREFIX, XATTR_SECURITY_PREFIX_LEN }, { XATTR_TRUSTED_PREFIX, XATTR_TRUSTED_PREFIX_LEN } }; int i; for (i = 0; i < ARRAY_SIZE(arr); i++) { size_t preflen = arr[i].len; if (strncmp(name, arr[i].prefix, preflen) == 0) { if (!name[preflen]) return -EINVAL; return 0; } } return -EOPNOTSUPP; } static ssize_t shmem_getxattr(struct dentry *dentry, const char *name, void *buffer, size_t size) { int err; /* * If this is a request for a synthetic attribute in the system.* * namespace use the generic infrastructure to resolve a handler * for it via sb->s_xattr. */ if (!strncmp(name, XATTR_SYSTEM_PREFIX, XATTR_SYSTEM_PREFIX_LEN)) return generic_getxattr(dentry, name, buffer, size); err = shmem_xattr_validate(name); if (err) return err; return shmem_xattr_get(dentry, name, buffer, size); } static int shmem_setxattr(struct dentry *dentry, const char *name, const void *value, size_t size, int flags) { int err; /* * If this is a request for a synthetic attribute in the system.* * namespace use the generic infrastructure to resolve a handler * for it via sb->s_xattr. */ if (!strncmp(name, XATTR_SYSTEM_PREFIX, XATTR_SYSTEM_PREFIX_LEN)) return generic_setxattr(dentry, name, value, size, flags); err = shmem_xattr_validate(name); if (err) return err; if (size == 0) value = ""; /* empty EA, do not remove */ return shmem_xattr_set(dentry, name, value, size, flags); } static int shmem_removexattr(struct dentry *dentry, const char *name) { int err; /* * If this is a request for a synthetic attribute in the system.* * namespace use the generic infrastructure to resolve a handler * for it via sb->s_xattr. */ if (!strncmp(name, XATTR_SYSTEM_PREFIX, XATTR_SYSTEM_PREFIX_LEN)) return generic_removexattr(dentry, name); err = shmem_xattr_validate(name); if (err) return err; return shmem_xattr_set(dentry, name, NULL, 0, XATTR_REPLACE); } static bool xattr_is_trusted(const char *name) { return !strncmp(name, XATTR_TRUSTED_PREFIX, XATTR_TRUSTED_PREFIX_LEN); } static ssize_t shmem_listxattr(struct dentry *dentry, char *buffer, size_t size) { bool trusted = capable(CAP_SYS_ADMIN); struct shmem_xattr *xattr; struct shmem_inode_info *info; size_t used = 0; info = SHMEM_I(dentry->d_inode); spin_lock(&info->lock); list_for_each_entry(xattr, &info->xattr_list, list) { size_t len; /* skip "trusted." attributes for unprivileged callers */ if (!trusted && xattr_is_trusted(xattr->name)) continue; len = strlen(xattr->name) + 1; used += len; if (buffer) { if (size < used) { used = -ERANGE; break; } memcpy(buffer, xattr->name, len); buffer += len; } } spin_unlock(&info->lock); return used; } #endif /* CONFIG_TMPFS_XATTR */ |
69f07ec93
|
1822 |
static const struct inode_operations shmem_short_symlink_operations = { |
b09e0fa4b
|
1823 |
.readlink = generic_readlink, |
69f07ec93
|
1824 |
.follow_link = shmem_follow_short_symlink, |
b09e0fa4b
|
1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 |
#ifdef CONFIG_TMPFS_XATTR .setxattr = shmem_setxattr, .getxattr = shmem_getxattr, .listxattr = shmem_listxattr, .removexattr = shmem_removexattr, #endif }; static const struct inode_operations shmem_symlink_inode_operations = { .readlink = generic_readlink, .follow_link = shmem_follow_link, .put_link = shmem_put_link, #ifdef CONFIG_TMPFS_XATTR .setxattr = shmem_setxattr, .getxattr = shmem_getxattr, .listxattr = shmem_listxattr, .removexattr = shmem_removexattr, |
39f0247d3
|
1842 |
#endif |
b09e0fa4b
|
1843 |
}; |
39f0247d3
|
1844 |
|
91828a405
|
1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 |
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; } |
480b116c9
|
1857 1858 |
static struct dentry *shmem_fh_to_dentry(struct super_block *sb, struct fid *fid, int fh_len, int fh_type) |
91828a405
|
1859 |
{ |
91828a405
|
1860 |
struct inode *inode; |
480b116c9
|
1861 1862 1863 1864 1865 1866 |
struct dentry *dentry = NULL; u64 inum = fid->raw[2]; inum = (inum << 32) | fid->raw[1]; if (fh_len < 3) return NULL; |
91828a405
|
1867 |
|
480b116c9
|
1868 1869 |
inode = ilookup5(sb, (unsigned long)(inum + fid->raw[0]), shmem_match, fid->raw); |
91828a405
|
1870 |
if (inode) { |
480b116c9
|
1871 |
dentry = d_find_alias(inode); |
91828a405
|
1872 1873 |
iput(inode); } |
480b116c9
|
1874 |
return dentry; |
91828a405
|
1875 1876 1877 1878 1879 1880 |
} static int shmem_encode_fh(struct dentry *dentry, __u32 *fh, int *len, int connectable) { struct inode *inode = dentry->d_inode; |
5fe0c2378
|
1881 1882 |
if (*len < 3) { *len = 3; |
91828a405
|
1883 |
return 255; |
5fe0c2378
|
1884 |
} |
91828a405
|
1885 |
|
1d3382cbf
|
1886 |
if (inode_unhashed(inode)) { |
91828a405
|
1887 1888 1889 1890 1891 1892 1893 |
/* 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); |
1d3382cbf
|
1894 |
if (inode_unhashed(inode)) |
91828a405
|
1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 |
__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; } |
396551644
|
1907 |
static const struct export_operations shmem_export_ops = { |
91828a405
|
1908 |
.get_parent = shmem_get_parent, |
91828a405
|
1909 |
.encode_fh = shmem_encode_fh, |
480b116c9
|
1910 |
.fh_to_dentry = shmem_fh_to_dentry, |
91828a405
|
1911 |
}; |
680d794ba
|
1912 1913 |
static int shmem_parse_options(char *options, struct shmem_sb_info *sbinfo, bool remount) |
1da177e4c
|
1914 1915 |
{ char *this_char, *value, *rest; |
b00dc3ad7
|
1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 |
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; } } |
1da177e4c
|
1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 |
if (!*this_char) continue; if ((value = strchr(this_char,'=')) != NULL) { *value++ = 0; } else { printk(KERN_ERR "tmpfs: No value for mount option '%s' ", 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; |
680d794ba
|
1956 1957 |
sbinfo->max_blocks = DIV_ROUND_UP(size, PAGE_CACHE_SIZE); |
1da177e4c
|
1958 |
} else if (!strcmp(this_char,"nr_blocks")) { |
680d794ba
|
1959 |
sbinfo->max_blocks = memparse(value, &rest); |
1da177e4c
|
1960 1961 1962 |
if (*rest) goto bad_val; } else if (!strcmp(this_char,"nr_inodes")) { |
680d794ba
|
1963 |
sbinfo->max_inodes = memparse(value, &rest); |
1da177e4c
|
1964 1965 1966 |
if (*rest) goto bad_val; } else if (!strcmp(this_char,"mode")) { |
680d794ba
|
1967 |
if (remount) |
1da177e4c
|
1968 |
continue; |
680d794ba
|
1969 |
sbinfo->mode = simple_strtoul(value, &rest, 8) & 07777; |
1da177e4c
|
1970 1971 1972 |
if (*rest) goto bad_val; } else if (!strcmp(this_char,"uid")) { |
680d794ba
|
1973 |
if (remount) |
1da177e4c
|
1974 |
continue; |
680d794ba
|
1975 |
sbinfo->uid = simple_strtoul(value, &rest, 0); |
1da177e4c
|
1976 1977 1978 |
if (*rest) goto bad_val; } else if (!strcmp(this_char,"gid")) { |
680d794ba
|
1979 |
if (remount) |
1da177e4c
|
1980 |
continue; |
680d794ba
|
1981 |
sbinfo->gid = simple_strtoul(value, &rest, 0); |
1da177e4c
|
1982 1983 |
if (*rest) goto bad_val; |
7339ff830
|
1984 |
} else if (!strcmp(this_char,"mpol")) { |
71fe804b6
|
1985 |
if (mpol_parse_str(value, &sbinfo->mpol, 1)) |
7339ff830
|
1986 |
goto bad_val; |
1da177e4c
|
1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 |
} else { printk(KERN_ERR "tmpfs: Bad mount option %s ", this_char); return 1; } } return 0; bad_val: printk(KERN_ERR "tmpfs: Bad value '%s' for mount option '%s' ", 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); |
680d794ba
|
2007 |
struct shmem_sb_info config = *sbinfo; |
0edd73b33
|
2008 2009 |
unsigned long inodes; int error = -EINVAL; |
680d794ba
|
2010 |
if (shmem_parse_options(data, &config, true)) |
0edd73b33
|
2011 |
return error; |
1da177e4c
|
2012 |
|
0edd73b33
|
2013 |
spin_lock(&sbinfo->stat_lock); |
0edd73b33
|
2014 |
inodes = sbinfo->max_inodes - sbinfo->free_inodes; |
7e496299d
|
2015 |
if (percpu_counter_compare(&sbinfo->used_blocks, config.max_blocks) > 0) |
0edd73b33
|
2016 |
goto out; |
680d794ba
|
2017 |
if (config.max_inodes < inodes) |
0edd73b33
|
2018 2019 |
goto out; /* |
54af60421
|
2020 |
* Those tests disallow limited->unlimited while any are in use; |
0edd73b33
|
2021 2022 2023 |
* but we must separately disallow unlimited->limited, because * in that case we have no record of how much is already in use. */ |
680d794ba
|
2024 |
if (config.max_blocks && !sbinfo->max_blocks) |
0edd73b33
|
2025 |
goto out; |
680d794ba
|
2026 |
if (config.max_inodes && !sbinfo->max_inodes) |
0edd73b33
|
2027 2028 2029 |
goto out; error = 0; |
680d794ba
|
2030 |
sbinfo->max_blocks = config.max_blocks; |
680d794ba
|
2031 2032 |
sbinfo->max_inodes = config.max_inodes; sbinfo->free_inodes = config.max_inodes - inodes; |
71fe804b6
|
2033 2034 2035 |
mpol_put(sbinfo->mpol); sbinfo->mpol = config.mpol; /* transfers initial ref */ |
0edd73b33
|
2036 2037 2038 |
out: spin_unlock(&sbinfo->stat_lock); return error; |
1da177e4c
|
2039 |
} |
680d794ba
|
2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 |
static int shmem_show_options(struct seq_file *seq, struct vfsmount *vfs) { struct shmem_sb_info *sbinfo = SHMEM_SB(vfs->mnt_sb); if (sbinfo->max_blocks != shmem_default_max_blocks()) seq_printf(seq, ",size=%luk", sbinfo->max_blocks << (PAGE_CACHE_SHIFT - 10)); if (sbinfo->max_inodes != shmem_default_max_inodes()) seq_printf(seq, ",nr_inodes=%lu", sbinfo->max_inodes); if (sbinfo->mode != (S_IRWXUGO | S_ISVTX)) seq_printf(seq, ",mode=%03o", sbinfo->mode); if (sbinfo->uid != 0) seq_printf(seq, ",uid=%u", sbinfo->uid); if (sbinfo->gid != 0) seq_printf(seq, ",gid=%u", sbinfo->gid); |
71fe804b6
|
2056 |
shmem_show_mpol(seq, sbinfo->mpol); |
680d794ba
|
2057 2058 2059 |
return 0; } #endif /* CONFIG_TMPFS */ |
1da177e4c
|
2060 2061 2062 |
static void shmem_put_super(struct super_block *sb) { |
602586a83
|
2063 2064 2065 2066 |
struct shmem_sb_info *sbinfo = SHMEM_SB(sb); percpu_counter_destroy(&sbinfo->used_blocks); kfree(sbinfo); |
1da177e4c
|
2067 2068 |
sb->s_fs_info = NULL; } |
2b2af54a5
|
2069 |
int shmem_fill_super(struct super_block *sb, void *data, int silent) |
1da177e4c
|
2070 2071 2072 |
{ struct inode *inode; struct dentry *root; |
0edd73b33
|
2073 |
struct shmem_sb_info *sbinfo; |
680d794ba
|
2074 2075 2076 |
int err = -ENOMEM; /* Round up to L1_CACHE_BYTES to resist false sharing */ |
425fbf047
|
2077 |
sbinfo = kzalloc(max((int)sizeof(struct shmem_sb_info), |
680d794ba
|
2078 2079 2080 |
L1_CACHE_BYTES), GFP_KERNEL); if (!sbinfo) return -ENOMEM; |
680d794ba
|
2081 |
sbinfo->mode = S_IRWXUGO | S_ISVTX; |
76aac0e9a
|
2082 2083 |
sbinfo->uid = current_fsuid(); sbinfo->gid = current_fsgid(); |
680d794ba
|
2084 |
sb->s_fs_info = sbinfo; |
1da177e4c
|
2085 |
|
0edd73b33
|
2086 |
#ifdef CONFIG_TMPFS |
1da177e4c
|
2087 2088 2089 2090 2091 2092 |
/* * 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)) { |
680d794ba
|
2093 2094 2095 2096 2097 2098 |
sbinfo->max_blocks = shmem_default_max_blocks(); sbinfo->max_inodes = shmem_default_max_inodes(); if (shmem_parse_options(data, sbinfo, false)) { err = -EINVAL; goto failed; } |
1da177e4c
|
2099 |
} |
91828a405
|
2100 |
sb->s_export_op = &shmem_export_ops; |
1da177e4c
|
2101 2102 2103 |
#else sb->s_flags |= MS_NOUSER; #endif |
0edd73b33
|
2104 |
spin_lock_init(&sbinfo->stat_lock); |
602586a83
|
2105 2106 |
if (percpu_counter_init(&sbinfo->used_blocks, 0)) goto failed; |
680d794ba
|
2107 |
sbinfo->free_inodes = sbinfo->max_inodes; |
0edd73b33
|
2108 |
|
285b2c4fd
|
2109 |
sb->s_maxbytes = MAX_LFS_FILESIZE; |
1da177e4c
|
2110 2111 2112 2113 |
sb->s_blocksize = PAGE_CACHE_SIZE; sb->s_blocksize_bits = PAGE_CACHE_SHIFT; sb->s_magic = TMPFS_MAGIC; sb->s_op = &shmem_ops; |
cfd95a9cf
|
2114 |
sb->s_time_gran = 1; |
b09e0fa4b
|
2115 |
#ifdef CONFIG_TMPFS_XATTR |
39f0247d3
|
2116 |
sb->s_xattr = shmem_xattr_handlers; |
b09e0fa4b
|
2117 2118 |
#endif #ifdef CONFIG_TMPFS_POSIX_ACL |
39f0247d3
|
2119 2120 |
sb->s_flags |= MS_POSIXACL; #endif |
0edd73b33
|
2121 |
|
454abafe9
|
2122 |
inode = shmem_get_inode(sb, NULL, S_IFDIR | sbinfo->mode, 0, VM_NORESERVE); |
1da177e4c
|
2123 2124 |
if (!inode) goto failed; |
680d794ba
|
2125 2126 |
inode->i_uid = sbinfo->uid; inode->i_gid = sbinfo->gid; |
1da177e4c
|
2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 |
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; } |
fcc234f88
|
2139 |
static struct kmem_cache *shmem_inode_cachep; |
1da177e4c
|
2140 2141 2142 |
static struct inode *shmem_alloc_inode(struct super_block *sb) { |
41ffe5d5c
|
2143 2144 2145 |
struct shmem_inode_info *info; info = kmem_cache_alloc(shmem_inode_cachep, GFP_KERNEL); if (!info) |
1da177e4c
|
2146 |
return NULL; |
41ffe5d5c
|
2147 |
return &info->vfs_inode; |
1da177e4c
|
2148 |
} |
41ffe5d5c
|
2149 |
static void shmem_destroy_callback(struct rcu_head *head) |
fa0d7e3de
|
2150 2151 2152 2153 2154 |
{ struct inode *inode = container_of(head, struct inode, i_rcu); INIT_LIST_HEAD(&inode->i_dentry); kmem_cache_free(shmem_inode_cachep, SHMEM_I(inode)); } |
1da177e4c
|
2155 2156 |
static void shmem_destroy_inode(struct inode *inode) { |
69f07ec93
|
2157 |
if ((inode->i_mode & S_IFMT) == S_IFREG) |
1da177e4c
|
2158 |
mpol_free_shared_policy(&SHMEM_I(inode)->policy); |
41ffe5d5c
|
2159 |
call_rcu(&inode->i_rcu, shmem_destroy_callback); |
1da177e4c
|
2160 |
} |
41ffe5d5c
|
2161 |
static void shmem_init_inode(void *foo) |
1da177e4c
|
2162 |
{ |
41ffe5d5c
|
2163 2164 |
struct shmem_inode_info *info = foo; inode_init_once(&info->vfs_inode); |
1da177e4c
|
2165 |
} |
41ffe5d5c
|
2166 |
static int shmem_init_inodecache(void) |
1da177e4c
|
2167 2168 2169 |
{ shmem_inode_cachep = kmem_cache_create("shmem_inode_cache", sizeof(struct shmem_inode_info), |
41ffe5d5c
|
2170 |
0, SLAB_PANIC, shmem_init_inode); |
1da177e4c
|
2171 2172 |
return 0; } |
41ffe5d5c
|
2173 |
static void shmem_destroy_inodecache(void) |
1da177e4c
|
2174 |
{ |
1a1d92c10
|
2175 |
kmem_cache_destroy(shmem_inode_cachep); |
1da177e4c
|
2176 |
} |
f5e54d6e5
|
2177 |
static const struct address_space_operations shmem_aops = { |
1da177e4c
|
2178 |
.writepage = shmem_writepage, |
767193253
|
2179 |
.set_page_dirty = __set_page_dirty_no_writeback, |
1da177e4c
|
2180 |
#ifdef CONFIG_TMPFS |
800d15a53
|
2181 2182 |
.write_begin = shmem_write_begin, .write_end = shmem_write_end, |
1da177e4c
|
2183 |
#endif |
304dbdb7a
|
2184 |
.migratepage = migrate_page, |
aa261f549
|
2185 |
.error_remove_page = generic_error_remove_page, |
1da177e4c
|
2186 |
}; |
15ad7cdcf
|
2187 |
static const struct file_operations shmem_file_operations = { |
1da177e4c
|
2188 2189 2190 |
.mmap = shmem_mmap, #ifdef CONFIG_TMPFS .llseek = generic_file_llseek, |
bcd78e496
|
2191 |
.read = do_sync_read, |
5402b976a
|
2192 |
.write = do_sync_write, |
bcd78e496
|
2193 |
.aio_read = shmem_file_aio_read, |
5402b976a
|
2194 |
.aio_write = generic_file_aio_write, |
1b061d924
|
2195 |
.fsync = noop_fsync, |
708e3508c
|
2196 |
.splice_read = shmem_file_splice_read, |
ae9764164
|
2197 |
.splice_write = generic_file_splice_write, |
1da177e4c
|
2198 2199 |
#endif }; |
92e1d5be9
|
2200 |
static const struct inode_operations shmem_inode_operations = { |
94c1e62df
|
2201 |
.setattr = shmem_setattr, |
f6b3ec238
|
2202 |
.truncate_range = shmem_truncate_range, |
b09e0fa4b
|
2203 2204 2205 2206 2207 2208 |
#ifdef CONFIG_TMPFS_XATTR .setxattr = shmem_setxattr, .getxattr = shmem_getxattr, .listxattr = shmem_listxattr, .removexattr = shmem_removexattr, #endif |
1da177e4c
|
2209 |
}; |
92e1d5be9
|
2210 |
static const struct inode_operations shmem_dir_inode_operations = { |
1da177e4c
|
2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 |
#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, |
1da177e4c
|
2221 |
#endif |
b09e0fa4b
|
2222 2223 2224 2225 2226 2227 |
#ifdef CONFIG_TMPFS_XATTR .setxattr = shmem_setxattr, .getxattr = shmem_getxattr, .listxattr = shmem_listxattr, .removexattr = shmem_removexattr, #endif |
39f0247d3
|
2228 |
#ifdef CONFIG_TMPFS_POSIX_ACL |
94c1e62df
|
2229 |
.setattr = shmem_setattr, |
39f0247d3
|
2230 2231 |
#endif }; |
92e1d5be9
|
2232 |
static const struct inode_operations shmem_special_inode_operations = { |
b09e0fa4b
|
2233 2234 2235 2236 2237 2238 |
#ifdef CONFIG_TMPFS_XATTR .setxattr = shmem_setxattr, .getxattr = shmem_getxattr, .listxattr = shmem_listxattr, .removexattr = shmem_removexattr, #endif |
39f0247d3
|
2239 |
#ifdef CONFIG_TMPFS_POSIX_ACL |
94c1e62df
|
2240 |
.setattr = shmem_setattr, |
39f0247d3
|
2241 |
#endif |
1da177e4c
|
2242 |
}; |
759b9775c
|
2243 |
static const struct super_operations shmem_ops = { |
1da177e4c
|
2244 2245 2246 2247 2248 |
.alloc_inode = shmem_alloc_inode, .destroy_inode = shmem_destroy_inode, #ifdef CONFIG_TMPFS .statfs = shmem_statfs, .remount_fs = shmem_remount_fs, |
680d794ba
|
2249 |
.show_options = shmem_show_options, |
1da177e4c
|
2250 |
#endif |
1f895f75d
|
2251 |
.evict_inode = shmem_evict_inode, |
1da177e4c
|
2252 2253 2254 |
.drop_inode = generic_delete_inode, .put_super = shmem_put_super, }; |
f0f37e2f7
|
2255 |
static const struct vm_operations_struct shmem_vm_ops = { |
54cb8821d
|
2256 |
.fault = shmem_fault, |
1da177e4c
|
2257 2258 2259 2260 2261 |
#ifdef CONFIG_NUMA .set_policy = shmem_set_policy, .get_policy = shmem_get_policy, #endif }; |
3c26ff6e4
|
2262 2263 |
static struct dentry *shmem_mount(struct file_system_type *fs_type, int flags, const char *dev_name, void *data) |
1da177e4c
|
2264 |
{ |
3c26ff6e4
|
2265 |
return mount_nodev(fs_type, flags, data, shmem_fill_super); |
1da177e4c
|
2266 |
} |
41ffe5d5c
|
2267 |
static struct file_system_type shmem_fs_type = { |
1da177e4c
|
2268 2269 |
.owner = THIS_MODULE, .name = "tmpfs", |
3c26ff6e4
|
2270 |
.mount = shmem_mount, |
1da177e4c
|
2271 2272 |
.kill_sb = kill_litter_super, }; |
1da177e4c
|
2273 |
|
41ffe5d5c
|
2274 |
int __init shmem_init(void) |
1da177e4c
|
2275 2276 |
{ int error; |
e0bf68dde
|
2277 2278 2279 |
error = bdi_init(&shmem_backing_dev_info); if (error) goto out4; |
41ffe5d5c
|
2280 |
error = shmem_init_inodecache(); |
1da177e4c
|
2281 2282 |
if (error) goto out3; |
41ffe5d5c
|
2283 |
error = register_filesystem(&shmem_fs_type); |
1da177e4c
|
2284 2285 2286 2287 2288 |
if (error) { printk(KERN_ERR "Could not register tmpfs "); goto out2; } |
95dc112a5
|
2289 |
|
41ffe5d5c
|
2290 2291 |
shm_mnt = vfs_kern_mount(&shmem_fs_type, MS_NOUSER, shmem_fs_type.name, NULL); |
1da177e4c
|
2292 2293 2294 2295 2296 2297 2298 2299 2300 |
if (IS_ERR(shm_mnt)) { error = PTR_ERR(shm_mnt); printk(KERN_ERR "Could not kern_mount tmpfs "); goto out1; } return 0; out1: |
41ffe5d5c
|
2301 |
unregister_filesystem(&shmem_fs_type); |
1da177e4c
|
2302 |
out2: |
41ffe5d5c
|
2303 |
shmem_destroy_inodecache(); |
1da177e4c
|
2304 |
out3: |
e0bf68dde
|
2305 2306 |
bdi_destroy(&shmem_backing_dev_info); out4: |
1da177e4c
|
2307 2308 2309 |
shm_mnt = ERR_PTR(error); return error; } |
853ac43ab
|
2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 |
#else /* !CONFIG_SHMEM */ /* * tiny-shmem: simple shmemfs and tmpfs using ramfs code * * This is intended for small system where the benefits of the full * shmem code (swap-backed and resource-limited) are outweighed by * their complexity. On systems without swap this code should be * effectively equivalent, but much lighter weight. */ #include <linux/ramfs.h> |
41ffe5d5c
|
2323 |
static struct file_system_type shmem_fs_type = { |
853ac43ab
|
2324 |
.name = "tmpfs", |
3c26ff6e4
|
2325 |
.mount = ramfs_mount, |
853ac43ab
|
2326 2327 |
.kill_sb = kill_litter_super, }; |
41ffe5d5c
|
2328 |
int __init shmem_init(void) |
853ac43ab
|
2329 |
{ |
41ffe5d5c
|
2330 |
BUG_ON(register_filesystem(&shmem_fs_type) != 0); |
853ac43ab
|
2331 |
|
41ffe5d5c
|
2332 |
shm_mnt = kern_mount(&shmem_fs_type); |
853ac43ab
|
2333 2334 2335 2336 |
BUG_ON(IS_ERR(shm_mnt)); return 0; } |
41ffe5d5c
|
2337 |
int shmem_unuse(swp_entry_t swap, struct page *page) |
853ac43ab
|
2338 2339 2340 |
{ return 0; } |
3f96b79ad
|
2341 2342 2343 2344 |
int shmem_lock(struct file *file, int lock, struct user_struct *user) { return 0; } |
41ffe5d5c
|
2345 |
void shmem_truncate_range(struct inode *inode, loff_t lstart, loff_t lend) |
94c1e62df
|
2346 |
{ |
41ffe5d5c
|
2347 |
truncate_inode_pages_range(inode->i_mapping, lstart, lend); |
94c1e62df
|
2348 2349 |
} EXPORT_SYMBOL_GPL(shmem_truncate_range); |
0b0a0806b
|
2350 2351 |
#define shmem_vm_ops generic_file_vm_ops #define shmem_file_operations ramfs_file_operations |
454abafe9
|
2352 |
#define shmem_get_inode(sb, dir, mode, dev, flags) ramfs_get_inode(sb, dir, mode, dev) |
0b0a0806b
|
2353 2354 |
#define shmem_acct_size(flags, size) 0 #define shmem_unacct_size(flags, size) do {} while (0) |
853ac43ab
|
2355 2356 2357 2358 |
#endif /* CONFIG_SHMEM */ /* common code */ |
1da177e4c
|
2359 |
|
467118102
|
2360 |
/** |
1da177e4c
|
2361 |
* shmem_file_setup - get an unlinked file living in tmpfs |
1da177e4c
|
2362 2363 |
* @name: name for dentry (to be seen in /proc/<pid>/maps * @size: size to be set for the file |
0b0a0806b
|
2364 |
* @flags: VM_NORESERVE suppresses pre-accounting of the entire object size |
1da177e4c
|
2365 |
*/ |
168f5ac66
|
2366 |
struct file *shmem_file_setup(const char *name, loff_t size, unsigned long flags) |
1da177e4c
|
2367 2368 2369 2370 |
{ int error; struct file *file; struct inode *inode; |
2c48b9c45
|
2371 2372 |
struct path path; struct dentry *root; |
1da177e4c
|
2373 2374 2375 2376 |
struct qstr this; if (IS_ERR(shm_mnt)) return (void *)shm_mnt; |
285b2c4fd
|
2377 |
if (size < 0 || size > MAX_LFS_FILESIZE) |
1da177e4c
|
2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 |
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; |
2c48b9c45
|
2388 2389 |
path.dentry = d_alloc(root, &this); if (!path.dentry) |
1da177e4c
|
2390 |
goto put_memory; |
2c48b9c45
|
2391 |
path.mnt = mntget(shm_mnt); |
1da177e4c
|
2392 |
|
1da177e4c
|
2393 |
error = -ENOSPC; |
454abafe9
|
2394 |
inode = shmem_get_inode(root->d_sb, NULL, S_IFREG | S_IRWXUGO, 0, flags); |
1da177e4c
|
2395 |
if (!inode) |
4b42af81f
|
2396 |
goto put_dentry; |
1da177e4c
|
2397 |
|
2c48b9c45
|
2398 |
d_instantiate(path.dentry, inode); |
1da177e4c
|
2399 2400 |
inode->i_size = size; inode->i_nlink = 0; /* It is unlinked */ |
853ac43ab
|
2401 2402 2403 |
#ifndef CONFIG_MMU error = ramfs_nommu_expand_for_mapping(inode, size); if (error) |
4b42af81f
|
2404 |
goto put_dentry; |
853ac43ab
|
2405 |
#endif |
4b42af81f
|
2406 2407 |
error = -ENFILE; |
2c48b9c45
|
2408 |
file = alloc_file(&path, FMODE_WRITE | FMODE_READ, |
4b42af81f
|
2409 2410 2411 |
&shmem_file_operations); if (!file) goto put_dentry; |
1da177e4c
|
2412 |
return file; |
1da177e4c
|
2413 |
put_dentry: |
2c48b9c45
|
2414 |
path_put(&path); |
1da177e4c
|
2415 2416 2417 2418 |
put_memory: shmem_unacct_size(flags, size); return ERR_PTR(error); } |
395e0ddc4
|
2419 |
EXPORT_SYMBOL_GPL(shmem_file_setup); |
1da177e4c
|
2420 |
|
467118102
|
2421 |
/** |
1da177e4c
|
2422 |
* shmem_zero_setup - setup a shared anonymous mapping |
1da177e4c
|
2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 |
* @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; |
bee4c36a5
|
2438 |
vma->vm_flags |= VM_CAN_NONLINEAR; |
1da177e4c
|
2439 2440 |
return 0; } |
d9d90e5eb
|
2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 |
/** * shmem_read_mapping_page_gfp - read into page cache, using specified page allocation flags. * @mapping: the page's address_space * @index: the page index * @gfp: the page allocator flags to use if allocating * * This behaves as a tmpfs "read_cache_page_gfp(mapping, index, gfp)", * with any new page allocations done using the specified allocation flags. * But read_cache_page_gfp() uses the ->readpage() method: which does not * suit tmpfs, since it may have pages in swapcache, and needs to find those * for itself; although drivers/gpu/drm i915 and ttm rely upon this support. * |
68da9f055
|
2454 2455 |
* i915_gem_object_get_pages_gtt() mixes __GFP_NORETRY | __GFP_NOWARN in * with the mapping_gfp_mask(), to avoid OOMing the machine unnecessarily. |
d9d90e5eb
|
2456 2457 2458 2459 |
*/ struct page *shmem_read_mapping_page_gfp(struct address_space *mapping, pgoff_t index, gfp_t gfp) { |
68da9f055
|
2460 2461 |
#ifdef CONFIG_SHMEM struct inode *inode = mapping->host; |
9276aad6c
|
2462 |
struct page *page; |
68da9f055
|
2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 |
int error; BUG_ON(mapping->a_ops != &shmem_aops); error = shmem_getpage_gfp(inode, index, &page, SGP_CACHE, gfp, NULL); if (error) page = ERR_PTR(error); else unlock_page(page); return page; #else /* * The tiny !SHMEM case uses ramfs without swap */ |
d9d90e5eb
|
2476 |
return read_cache_page_gfp(mapping, index, gfp); |
68da9f055
|
2477 |
#endif |
d9d90e5eb
|
2478 2479 |
} EXPORT_SYMBOL_GPL(shmem_read_mapping_page_gfp); |