Blame view
mm/shmem.c
106 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> |
250297edf
|
27 |
#include <linux/ramfs.h> |
caefba174
|
28 |
#include <linux/pagemap.h> |
853ac43ab
|
29 30 |
#include <linux/file.h> #include <linux/mm.h> |
b95f1b31b
|
31 |
#include <linux/export.h> |
853ac43ab
|
32 |
#include <linux/swap.h> |
e2e40f2c1
|
33 |
#include <linux/uio.h> |
f3f0e1d21
|
34 |
#include <linux/khugepaged.h> |
853ac43ab
|
35 36 37 38 |
static struct vfsmount *shm_mnt; #ifdef CONFIG_SHMEM |
1da177e4c
|
39 40 41 42 43 |
/* * 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
|
44 |
#include <linux/xattr.h> |
a56942551
|
45 |
#include <linux/exportfs.h> |
1c7c474c3
|
46 |
#include <linux/posix_acl.h> |
feda821e7
|
47 |
#include <linux/posix_acl_xattr.h> |
1da177e4c
|
48 |
#include <linux/mman.h> |
1da177e4c
|
49 50 51 52 |
#include <linux/string.h> #include <linux/slab.h> #include <linux/backing-dev.h> #include <linux/shmem_fs.h> |
1da177e4c
|
53 |
#include <linux/writeback.h> |
1da177e4c
|
54 |
#include <linux/blkdev.h> |
bda97eab0
|
55 |
#include <linux/pagevec.h> |
41ffe5d5c
|
56 |
#include <linux/percpu_counter.h> |
83e4fa9c1
|
57 |
#include <linux/falloc.h> |
708e3508c
|
58 |
#include <linux/splice.h> |
1da177e4c
|
59 60 61 62 |
#include <linux/security.h> #include <linux/swapops.h> #include <linux/mempolicy.h> #include <linux/namei.h> |
b00dc3ad7
|
63 |
#include <linux/ctype.h> |
304dbdb7a
|
64 |
#include <linux/migrate.h> |
c1f60a5a4
|
65 |
#include <linux/highmem.h> |
680d794ba
|
66 |
#include <linux/seq_file.h> |
925629278
|
67 |
#include <linux/magic.h> |
9183df25f
|
68 |
#include <linux/syscalls.h> |
40e041a2c
|
69 |
#include <linux/fcntl.h> |
9183df25f
|
70 |
#include <uapi/linux/memfd.h> |
304dbdb7a
|
71 |
|
1da177e4c
|
72 |
#include <asm/uaccess.h> |
1da177e4c
|
73 |
#include <asm/pgtable.h> |
dd56b0464
|
74 |
#include "internal.h" |
09cbfeaf1
|
75 76 |
#define BLOCKS_PER_PAGE (PAGE_SIZE/512) #define VM_ACCT(size) (PAGE_ALIGN(size) >> PAGE_SHIFT) |
1da177e4c
|
77 |
|
1da177e4c
|
78 79 |
/* Pretend that each entry is of this size in directory's i_size */ #define BOGO_DIRENT_SIZE 20 |
69f07ec93
|
80 81 |
/* Symlink up to this size is kmalloc'ed instead of using a swappable page */ #define SHORT_SYMLINK_LEN 128 |
1aac14003
|
82 |
/* |
f00cdc6df
|
83 84 85 |
* shmem_fallocate communicates with shmem_fault or shmem_writepage via * inode->i_private (with i_mutex making sure that it has only one user at * a time): we would prefer not to enlarge the shmem inode just for that. |
1aac14003
|
86 87 |
*/ struct shmem_falloc { |
8e205f779
|
88 |
wait_queue_head_t *waitq; /* faults into hole wait for punch to end */ |
1aac14003
|
89 90 91 92 93 |
pgoff_t start; /* start of range currently being fallocated */ pgoff_t next; /* the next page offset to be fallocated */ pgoff_t nr_falloced; /* how many new pages have been fallocated */ pgoff_t nr_unswapped; /* how often writepage refused to swap out */ }; |
b76db7354
|
94 |
#ifdef CONFIG_TMPFS |
680d794ba
|
95 96 97 98 99 100 101 102 103 |
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
|
104 |
#endif |
680d794ba
|
105 |
|
bde05d1cc
|
106 107 108 |
static bool shmem_should_replace_page(struct page *page, gfp_t gfp); static int shmem_replace_page(struct page **pagep, gfp_t gfp, struct shmem_inode_info *info, pgoff_t index); |
68da9f055
|
109 |
static int shmem_getpage_gfp(struct inode *inode, pgoff_t index, |
9e18eb293
|
110 111 |
struct page **pagep, enum sgp_type sgp, gfp_t gfp, struct mm_struct *fault_mm, int *fault_type); |
68da9f055
|
112 |
|
f3f0e1d21
|
113 |
int shmem_getpage(struct inode *inode, pgoff_t index, |
9e18eb293
|
114 |
struct page **pagep, enum sgp_type sgp) |
68da9f055
|
115 116 |
{ return shmem_getpage_gfp(inode, index, pagep, sgp, |
9e18eb293
|
117 |
mapping_gfp_mask(inode->i_mapping), NULL, NULL); |
68da9f055
|
118 |
} |
1da177e4c
|
119 |
|
1da177e4c
|
120 121 122 123 124 125 126 127 128 129 130 131 132 |
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
|
133 |
return (flags & VM_NORESERVE) ? |
191c54244
|
134 |
0 : security_vm_enough_memory_mm(current->mm, VM_ACCT(size)); |
1da177e4c
|
135 136 137 138 |
} static inline void shmem_unacct_size(unsigned long flags, loff_t size) { |
0b0a0806b
|
139 |
if (!(flags & VM_NORESERVE)) |
1da177e4c
|
140 141 |
vm_unacct_memory(VM_ACCT(size)); } |
771425179
|
142 143 144 145 146 147 148 149 150 151 152 153 |
static inline int shmem_reacct_size(unsigned long flags, loff_t oldsize, loff_t newsize) { if (!(flags & VM_NORESERVE)) { if (VM_ACCT(newsize) > VM_ACCT(oldsize)) return security_vm_enough_memory_mm(current->mm, VM_ACCT(newsize) - VM_ACCT(oldsize)); else if (VM_ACCT(newsize) < VM_ACCT(oldsize)) vm_unacct_memory(VM_ACCT(oldsize) - VM_ACCT(newsize)); } return 0; } |
1da177e4c
|
154 155 |
/* * ... whereas tmpfs objects are accounted incrementally as |
75edd345e
|
156 |
* pages are allocated, in order to allow large sparse files. |
1da177e4c
|
157 158 159 |
* 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. */ |
800d8c63b
|
160 |
static inline int shmem_acct_block(unsigned long flags, long pages) |
1da177e4c
|
161 |
{ |
800d8c63b
|
162 163 164 165 166 |
if (!(flags & VM_NORESERVE)) return 0; return security_vm_enough_memory_mm(current->mm, pages * VM_ACCT(PAGE_SIZE)); |
1da177e4c
|
167 168 169 170 |
} static inline void shmem_unacct_blocks(unsigned long flags, long pages) { |
0b0a0806b
|
171 |
if (flags & VM_NORESERVE) |
09cbfeaf1
|
172 |
vm_unacct_memory(pages * VM_ACCT(PAGE_SIZE)); |
1da177e4c
|
173 |
} |
759b9775c
|
174 |
static const struct super_operations shmem_ops; |
f5e54d6e5
|
175 |
static const struct address_space_operations shmem_aops; |
15ad7cdcf
|
176 |
static const struct file_operations shmem_file_operations; |
92e1d5be9
|
177 178 179 |
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
|
180 |
static const struct vm_operations_struct shmem_vm_ops; |
779750d20
|
181 |
static struct file_system_type shmem_fs_type; |
1da177e4c
|
182 |
|
1da177e4c
|
183 |
static LIST_HEAD(shmem_swaplist); |
cb5f7b9a4
|
184 |
static DEFINE_MUTEX(shmem_swaplist_mutex); |
1da177e4c
|
185 |
|
5b04c6890
|
186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 |
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
|
210 |
/** |
41ffe5d5c
|
211 |
* shmem_recalc_inode - recalculate the block usage of an inode |
1da177e4c
|
212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 |
* @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
|
229 230 231 |
struct shmem_sb_info *sbinfo = SHMEM_SB(inode->i_sb); if (sbinfo->max_blocks) percpu_counter_add(&sbinfo->used_blocks, -freed); |
1da177e4c
|
232 |
info->alloced -= freed; |
54af60421
|
233 |
inode->i_blocks -= freed * BLOCKS_PER_PAGE; |
1da177e4c
|
234 |
shmem_unacct_blocks(info->flags, freed); |
1da177e4c
|
235 236 |
} } |
800d8c63b
|
237 238 239 240 |
bool shmem_charge(struct inode *inode, long pages) { struct shmem_inode_info *info = SHMEM_I(inode); struct shmem_sb_info *sbinfo = SHMEM_SB(inode->i_sb); |
4595ef88d
|
241 |
unsigned long flags; |
800d8c63b
|
242 243 244 |
if (shmem_acct_block(info->flags, pages)) return false; |
4595ef88d
|
245 |
spin_lock_irqsave(&info->lock, flags); |
800d8c63b
|
246 247 248 |
info->alloced += pages; inode->i_blocks += pages * BLOCKS_PER_PAGE; shmem_recalc_inode(inode); |
4595ef88d
|
249 |
spin_unlock_irqrestore(&info->lock, flags); |
800d8c63b
|
250 251 252 253 254 255 256 |
inode->i_mapping->nrpages += pages; if (!sbinfo->max_blocks) return true; if (percpu_counter_compare(&sbinfo->used_blocks, sbinfo->max_blocks - pages) > 0) { inode->i_mapping->nrpages -= pages; |
4595ef88d
|
257 |
spin_lock_irqsave(&info->lock, flags); |
800d8c63b
|
258 259 |
info->alloced -= pages; shmem_recalc_inode(inode); |
4595ef88d
|
260 |
spin_unlock_irqrestore(&info->lock, flags); |
71664665c
|
261 |
shmem_unacct_blocks(info->flags, pages); |
800d8c63b
|
262 263 264 265 266 267 268 269 270 271 |
return false; } percpu_counter_add(&sbinfo->used_blocks, pages); return true; } void shmem_uncharge(struct inode *inode, long pages) { struct shmem_inode_info *info = SHMEM_I(inode); struct shmem_sb_info *sbinfo = SHMEM_SB(inode->i_sb); |
4595ef88d
|
272 |
unsigned long flags; |
800d8c63b
|
273 |
|
4595ef88d
|
274 |
spin_lock_irqsave(&info->lock, flags); |
800d8c63b
|
275 276 277 |
info->alloced -= pages; inode->i_blocks -= pages * BLOCKS_PER_PAGE; shmem_recalc_inode(inode); |
4595ef88d
|
278 |
spin_unlock_irqrestore(&info->lock, flags); |
800d8c63b
|
279 280 281 |
if (sbinfo->max_blocks) percpu_counter_sub(&sbinfo->used_blocks, pages); |
71664665c
|
282 |
shmem_unacct_blocks(info->flags, pages); |
800d8c63b
|
283 |
} |
7a5d0fbb2
|
284 285 286 287 288 289 290 |
/* * 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; |
6dbaf22ce
|
291 |
void *item; |
7a5d0fbb2
|
292 293 |
VM_BUG_ON(!expected); |
6dbaf22ce
|
294 |
VM_BUG_ON(!replacement); |
7a5d0fbb2
|
295 |
pslot = radix_tree_lookup_slot(&mapping->page_tree, index); |
6dbaf22ce
|
296 297 298 |
if (!pslot) return -ENOENT; item = radix_tree_deref_slot_protected(pslot, &mapping->tree_lock); |
7a5d0fbb2
|
299 300 |
if (item != expected) return -ENOENT; |
6dbaf22ce
|
301 |
radix_tree_replace_slot(pslot, replacement); |
7a5d0fbb2
|
302 303 304 305 |
return 0; } /* |
d18992286
|
306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 |
* Sometimes, before we decide whether to proceed or to fail, we must check * that an entry was not already brought back from swap by a racing thread. * * Checking page is not enough: by the time a SwapCache page is locked, it * might be reused, and again be SwapCache, using the same swap as before. */ static bool shmem_confirm_swap(struct address_space *mapping, pgoff_t index, swp_entry_t swap) { void *item; rcu_read_lock(); item = radix_tree_lookup(&mapping->page_tree, index); rcu_read_unlock(); return item == swp_to_radix_entry(swap); } /* |
5a6e75f81
|
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 |
* Definitions for "huge tmpfs": tmpfs mounted with the huge= option * * SHMEM_HUGE_NEVER: * disables huge pages for the mount; * SHMEM_HUGE_ALWAYS: * enables huge pages for the mount; * SHMEM_HUGE_WITHIN_SIZE: * only allocate huge pages if the page will be fully within i_size, * also respect fadvise()/madvise() hints; * SHMEM_HUGE_ADVISE: * only allocate huge pages if requested with fadvise()/madvise(); */ #define SHMEM_HUGE_NEVER 0 #define SHMEM_HUGE_ALWAYS 1 #define SHMEM_HUGE_WITHIN_SIZE 2 #define SHMEM_HUGE_ADVISE 3 /* * Special values. * Only can be set via /sys/kernel/mm/transparent_hugepage/shmem_enabled: * * SHMEM_HUGE_DENY: * disables huge on shm_mnt and all mounts, for emergency use; * SHMEM_HUGE_FORCE: * enables huge on shm_mnt and all mounts, w/o needing option, for testing; * */ #define SHMEM_HUGE_DENY (-1) #define SHMEM_HUGE_FORCE (-2) |
e496cf3d7
|
354 |
#ifdef CONFIG_TRANSPARENT_HUGE_PAGECACHE |
5a6e75f81
|
355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 |
/* ifdef here to avoid bloating shmem.o when not necessary */ int shmem_huge __read_mostly; static int shmem_parse_huge(const char *str) { if (!strcmp(str, "never")) return SHMEM_HUGE_NEVER; if (!strcmp(str, "always")) return SHMEM_HUGE_ALWAYS; if (!strcmp(str, "within_size")) return SHMEM_HUGE_WITHIN_SIZE; if (!strcmp(str, "advise")) return SHMEM_HUGE_ADVISE; if (!strcmp(str, "deny")) return SHMEM_HUGE_DENY; if (!strcmp(str, "force")) return SHMEM_HUGE_FORCE; return -EINVAL; } static const char *shmem_format_huge(int huge) { switch (huge) { case SHMEM_HUGE_NEVER: return "never"; case SHMEM_HUGE_ALWAYS: return "always"; case SHMEM_HUGE_WITHIN_SIZE: return "within_size"; case SHMEM_HUGE_ADVISE: return "advise"; case SHMEM_HUGE_DENY: return "deny"; case SHMEM_HUGE_FORCE: return "force"; default: VM_BUG_ON(1); return "bad_val"; } } |
779750d20
|
396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 |
static unsigned long shmem_unused_huge_shrink(struct shmem_sb_info *sbinfo, struct shrink_control *sc, unsigned long nr_to_split) { LIST_HEAD(list), *pos, *next; struct inode *inode; struct shmem_inode_info *info; struct page *page; unsigned long batch = sc ? sc->nr_to_scan : 128; int removed = 0, split = 0; if (list_empty(&sbinfo->shrinklist)) return SHRINK_STOP; spin_lock(&sbinfo->shrinklist_lock); list_for_each_safe(pos, next, &sbinfo->shrinklist) { info = list_entry(pos, struct shmem_inode_info, shrinklist); /* pin the inode */ inode = igrab(&info->vfs_inode); /* inode is about to be evicted */ if (!inode) { list_del_init(&info->shrinklist); removed++; goto next; } /* Check if there's anything to gain */ if (round_up(inode->i_size, PAGE_SIZE) == round_up(inode->i_size, HPAGE_PMD_SIZE)) { list_del_init(&info->shrinklist); removed++; iput(inode); goto next; } list_move(&info->shrinklist, &list); next: if (!--batch) break; } spin_unlock(&sbinfo->shrinklist_lock); list_for_each_safe(pos, next, &list) { int ret; info = list_entry(pos, struct shmem_inode_info, shrinklist); inode = &info->vfs_inode; if (nr_to_split && split >= nr_to_split) { iput(inode); continue; } page = find_lock_page(inode->i_mapping, (inode->i_size & HPAGE_PMD_MASK) >> PAGE_SHIFT); if (!page) goto drop; if (!PageTransHuge(page)) { unlock_page(page); put_page(page); goto drop; } ret = split_huge_page(page); unlock_page(page); put_page(page); if (ret) { /* split failed: leave it on the list */ iput(inode); continue; } split++; drop: list_del_init(&info->shrinklist); removed++; iput(inode); } spin_lock(&sbinfo->shrinklist_lock); list_splice_tail(&list, &sbinfo->shrinklist); sbinfo->shrinklist_len -= removed; spin_unlock(&sbinfo->shrinklist_lock); return split; } static long shmem_unused_huge_scan(struct super_block *sb, struct shrink_control *sc) { struct shmem_sb_info *sbinfo = SHMEM_SB(sb); if (!READ_ONCE(sbinfo->shrinklist_len)) return SHRINK_STOP; return shmem_unused_huge_shrink(sbinfo, sc, 0); } static long shmem_unused_huge_count(struct super_block *sb, struct shrink_control *sc) { struct shmem_sb_info *sbinfo = SHMEM_SB(sb); return READ_ONCE(sbinfo->shrinklist_len); } |
e496cf3d7
|
503 |
#else /* !CONFIG_TRANSPARENT_HUGE_PAGECACHE */ |
5a6e75f81
|
504 505 |
#define shmem_huge SHMEM_HUGE_DENY |
779750d20
|
506 507 508 509 510 |
static unsigned long shmem_unused_huge_shrink(struct shmem_sb_info *sbinfo, struct shrink_control *sc, unsigned long nr_to_split) { return 0; } |
e496cf3d7
|
511 |
#endif /* CONFIG_TRANSPARENT_HUGE_PAGECACHE */ |
5a6e75f81
|
512 513 |
/* |
46f65ec15
|
514 515 516 517 |
* 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, |
fed400a18
|
518 |
pgoff_t index, void *expected) |
46f65ec15
|
519 |
{ |
800d8c63b
|
520 |
int error, nr = hpage_nr_pages(page); |
46f65ec15
|
521 |
|
800d8c63b
|
522 523 |
VM_BUG_ON_PAGE(PageTail(page), page); VM_BUG_ON_PAGE(index != round_down(index, nr), page); |
309381fea
|
524 525 |
VM_BUG_ON_PAGE(!PageLocked(page), page); VM_BUG_ON_PAGE(!PageSwapBacked(page), page); |
800d8c63b
|
526 |
VM_BUG_ON(expected && PageTransHuge(page)); |
46f65ec15
|
527 |
|
800d8c63b
|
528 |
page_ref_add(page, nr); |
b065b4321
|
529 530 531 532 |
page->mapping = mapping; page->index = index; spin_lock_irq(&mapping->tree_lock); |
800d8c63b
|
533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 |
if (PageTransHuge(page)) { void __rcu **results; pgoff_t idx; int i; error = 0; if (radix_tree_gang_lookup_slot(&mapping->page_tree, &results, &idx, index, 1) && idx < index + HPAGE_PMD_NR) { error = -EEXIST; } if (!error) { for (i = 0; i < HPAGE_PMD_NR; i++) { error = radix_tree_insert(&mapping->page_tree, index + i, page + i); VM_BUG_ON(error); } count_vm_event(THP_FILE_ALLOC); } } else if (!expected) { |
b065b4321
|
554 |
error = radix_tree_insert(&mapping->page_tree, index, page); |
800d8c63b
|
555 |
} else { |
b065b4321
|
556 557 |
error = shmem_radix_tree_replace(mapping, index, expected, page); |
800d8c63b
|
558 |
} |
46f65ec15
|
559 |
if (!error) { |
800d8c63b
|
560 561 |
mapping->nrpages += nr; if (PageTransHuge(page)) |
11fb99898
|
562 563 564 |
__inc_node_page_state(page, NR_SHMEM_THPS); __mod_node_page_state(page_pgdat(page), NR_FILE_PAGES, nr); __mod_node_page_state(page_pgdat(page), NR_SHMEM, nr); |
b065b4321
|
565 566 567 568 |
spin_unlock_irq(&mapping->tree_lock); } else { page->mapping = NULL; spin_unlock_irq(&mapping->tree_lock); |
800d8c63b
|
569 |
page_ref_sub(page, nr); |
46f65ec15
|
570 |
} |
46f65ec15
|
571 572 573 574 |
return error; } /* |
6922c0c7a
|
575 576 577 578 579 580 |
* 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; |
800d8c63b
|
581 |
VM_BUG_ON_PAGE(PageCompound(page), page); |
6922c0c7a
|
582 583 584 585 |
spin_lock_irq(&mapping->tree_lock); error = shmem_radix_tree_replace(mapping, page->index, page, radswap); page->mapping = NULL; mapping->nrpages--; |
11fb99898
|
586 587 |
__dec_node_page_state(page, NR_FILE_PAGES); __dec_node_page_state(page, NR_SHMEM); |
6922c0c7a
|
588 |
spin_unlock_irq(&mapping->tree_lock); |
09cbfeaf1
|
589 |
put_page(page); |
6922c0c7a
|
590 591 592 593 |
BUG_ON(error); } /* |
7a5d0fbb2
|
594 595 596 597 598 |
* 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) { |
6dbaf22ce
|
599 |
void *old; |
7a5d0fbb2
|
600 601 |
spin_lock_irq(&mapping->tree_lock); |
6dbaf22ce
|
602 |
old = radix_tree_delete_item(&mapping->page_tree, index, radswap); |
7a5d0fbb2
|
603 |
spin_unlock_irq(&mapping->tree_lock); |
6dbaf22ce
|
604 605 606 607 |
if (old != radswap) return -ENOENT; free_swap_and_cache(radix_to_swp_entry(radswap)); return 0; |
7a5d0fbb2
|
608 609 610 |
} /* |
6a15a3709
|
611 |
* Determine (in bytes) how many of the shmem object's pages mapped by the |
48131e03c
|
612 |
* given offsets are swapped out. |
6a15a3709
|
613 614 615 616 |
* * This is safe to call without i_mutex or mapping->tree_lock thanks to RCU, * as long as the inode doesn't go away and racy results are not a problem. */ |
48131e03c
|
617 618 |
unsigned long shmem_partial_swap_usage(struct address_space *mapping, pgoff_t start, pgoff_t end) |
6a15a3709
|
619 |
{ |
6a15a3709
|
620 621 622 |
struct radix_tree_iter iter; void **slot; struct page *page; |
48131e03c
|
623 |
unsigned long swapped = 0; |
6a15a3709
|
624 625 |
rcu_read_lock(); |
6a15a3709
|
626 627 628 629 630 |
radix_tree_for_each_slot(slot, &mapping->page_tree, &iter, start) { if (iter.index >= end) break; page = radix_tree_deref_slot(slot); |
2cf938aae
|
631 632 633 634 |
if (radix_tree_deref_retry(page)) { slot = radix_tree_iter_retry(&iter); continue; } |
6a15a3709
|
635 636 637 638 639 640 |
if (radix_tree_exceptional_entry(page)) swapped++; if (need_resched()) { cond_resched_rcu(); |
7165092fe
|
641 |
slot = radix_tree_iter_next(&iter); |
6a15a3709
|
642 643 644 645 646 647 648 649 650 |
} } rcu_read_unlock(); return swapped << PAGE_SHIFT; } /* |
48131e03c
|
651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 |
* Determine (in bytes) how many of the shmem object's pages mapped by the * given vma is swapped out. * * This is safe to call without i_mutex or mapping->tree_lock thanks to RCU, * as long as the inode doesn't go away and racy results are not a problem. */ unsigned long shmem_swap_usage(struct vm_area_struct *vma) { struct inode *inode = file_inode(vma->vm_file); struct shmem_inode_info *info = SHMEM_I(inode); struct address_space *mapping = inode->i_mapping; unsigned long swapped; /* Be careful as we don't hold info->lock */ swapped = READ_ONCE(info->swapped); /* * The easier cases are when the shmem object has nothing in swap, or * the vma maps it whole. Then we can simply use the stats that we * already track. */ if (!swapped) return 0; if (!vma->vm_pgoff && vma->vm_end - vma->vm_start >= inode->i_size) return swapped << PAGE_SHIFT; /* Here comes the more involved part */ return shmem_partial_swap_usage(mapping, linear_page_index(vma, vma->vm_start), linear_page_index(vma, vma->vm_end)); } /* |
245132643
|
685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 |
* SysV IPC SHM_UNLOCK restore Unevictable pages to their evictable lists. */ void shmem_unlock_mapping(struct address_space *mapping) { struct pagevec pvec; pgoff_t indices[PAGEVEC_SIZE]; pgoff_t index = 0; pagevec_init(&pvec, 0); /* * Minor point, but we might as well stop if someone else SHM_LOCKs it. */ while (!mapping_unevictable(mapping)) { /* * Avoid pagevec_lookup(): find_get_pages() returns 0 as if it * has finished, if it hits a row of PAGEVEC_SIZE swap entries. */ |
0cd6144aa
|
702 703 |
pvec.nr = find_get_entries(mapping, index, PAGEVEC_SIZE, pvec.pages, indices); |
245132643
|
704 705 706 |
if (!pvec.nr) break; index = indices[pvec.nr - 1] + 1; |
0cd6144aa
|
707 |
pagevec_remove_exceptionals(&pvec); |
245132643
|
708 709 710 711 |
check_move_unevictable_pages(pvec.pages, pvec.nr); pagevec_release(&pvec); cond_resched(); } |
7a5d0fbb2
|
712 713 714 715 |
} /* * Remove range of pages and swap entries from radix tree, and free them. |
1635f6a74
|
716 |
* If !unfalloc, truncate or punch hole; if unfalloc, undo failed fallocate. |
7a5d0fbb2
|
717 |
*/ |
1635f6a74
|
718 719 |
static void shmem_undo_range(struct inode *inode, loff_t lstart, loff_t lend, bool unfalloc) |
1da177e4c
|
720 |
{ |
285b2c4fd
|
721 |
struct address_space *mapping = inode->i_mapping; |
1da177e4c
|
722 |
struct shmem_inode_info *info = SHMEM_I(inode); |
09cbfeaf1
|
723 724 725 726 |
pgoff_t start = (lstart + PAGE_SIZE - 1) >> PAGE_SHIFT; pgoff_t end = (lend + 1) >> PAGE_SHIFT; unsigned int partial_start = lstart & (PAGE_SIZE - 1); unsigned int partial_end = (lend + 1) & (PAGE_SIZE - 1); |
bda97eab0
|
727 |
struct pagevec pvec; |
7a5d0fbb2
|
728 729 |
pgoff_t indices[PAGEVEC_SIZE]; long nr_swaps_freed = 0; |
285b2c4fd
|
730 |
pgoff_t index; |
bda97eab0
|
731 |
int i; |
83e4fa9c1
|
732 733 |
if (lend == -1) end = -1; /* unsigned, so actually very big */ |
bda97eab0
|
734 735 736 |
pagevec_init(&pvec, 0); index = start; |
83e4fa9c1
|
737 |
while (index < end) { |
0cd6144aa
|
738 739 740 |
pvec.nr = find_get_entries(mapping, index, min(end - index, (pgoff_t)PAGEVEC_SIZE), pvec.pages, indices); |
7a5d0fbb2
|
741 742 |
if (!pvec.nr) break; |
bda97eab0
|
743 744 |
for (i = 0; i < pagevec_count(&pvec); i++) { struct page *page = pvec.pages[i]; |
7a5d0fbb2
|
745 |
index = indices[i]; |
83e4fa9c1
|
746 |
if (index >= end) |
bda97eab0
|
747 |
break; |
7a5d0fbb2
|
748 |
if (radix_tree_exceptional_entry(page)) { |
1635f6a74
|
749 750 |
if (unfalloc) continue; |
7a5d0fbb2
|
751 752 |
nr_swaps_freed += !shmem_free_swap(mapping, index, page); |
bda97eab0
|
753 |
continue; |
7a5d0fbb2
|
754 |
} |
800d8c63b
|
755 |
VM_BUG_ON_PAGE(page_to_pgoff(page) != index, page); |
7a5d0fbb2
|
756 |
if (!trylock_page(page)) |
bda97eab0
|
757 |
continue; |
800d8c63b
|
758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 |
if (PageTransTail(page)) { /* Middle of THP: zero out the page */ clear_highpage(page); unlock_page(page); continue; } else if (PageTransHuge(page)) { if (index == round_down(end, HPAGE_PMD_NR)) { /* * Range ends in the middle of THP: * zero out the page */ clear_highpage(page); unlock_page(page); continue; } index += HPAGE_PMD_NR - 1; i += HPAGE_PMD_NR - 1; } |
1635f6a74
|
777 |
if (!unfalloc || !PageUptodate(page)) { |
800d8c63b
|
778 779 |
VM_BUG_ON_PAGE(PageTail(page), page); if (page_mapping(page) == mapping) { |
309381fea
|
780 |
VM_BUG_ON_PAGE(PageWriteback(page), page); |
1635f6a74
|
781 782 |
truncate_inode_page(mapping, page); } |
bda97eab0
|
783 |
} |
bda97eab0
|
784 785 |
unlock_page(page); } |
0cd6144aa
|
786 |
pagevec_remove_exceptionals(&pvec); |
245132643
|
787 |
pagevec_release(&pvec); |
bda97eab0
|
788 789 790 |
cond_resched(); index++; } |
1da177e4c
|
791 |
|
83e4fa9c1
|
792 |
if (partial_start) { |
bda97eab0
|
793 |
struct page *page = NULL; |
9e18eb293
|
794 |
shmem_getpage(inode, start - 1, &page, SGP_READ); |
bda97eab0
|
795 |
if (page) { |
09cbfeaf1
|
796 |
unsigned int top = PAGE_SIZE; |
83e4fa9c1
|
797 798 799 800 801 802 803 |
if (start > end) { top = partial_end; partial_end = 0; } zero_user_segment(page, partial_start, top); set_page_dirty(page); unlock_page(page); |
09cbfeaf1
|
804 |
put_page(page); |
83e4fa9c1
|
805 806 807 808 |
} } if (partial_end) { struct page *page = NULL; |
9e18eb293
|
809 |
shmem_getpage(inode, end, &page, SGP_READ); |
83e4fa9c1
|
810 811 |
if (page) { zero_user_segment(page, 0, partial_end); |
bda97eab0
|
812 813 |
set_page_dirty(page); unlock_page(page); |
09cbfeaf1
|
814 |
put_page(page); |
bda97eab0
|
815 816 |
} } |
83e4fa9c1
|
817 818 |
if (start >= end) return; |
bda97eab0
|
819 820 |
index = start; |
b1a366500
|
821 |
while (index < end) { |
bda97eab0
|
822 |
cond_resched(); |
0cd6144aa
|
823 824 |
pvec.nr = find_get_entries(mapping, index, |
83e4fa9c1
|
825 |
min(end - index, (pgoff_t)PAGEVEC_SIZE), |
0cd6144aa
|
826 |
pvec.pages, indices); |
7a5d0fbb2
|
827 |
if (!pvec.nr) { |
b1a366500
|
828 829 |
/* If all gone or hole-punch or unfalloc, we're done */ if (index == start || end != -1) |
bda97eab0
|
830 |
break; |
b1a366500
|
831 |
/* But if truncating, restart to make sure all gone */ |
bda97eab0
|
832 833 834 |
index = start; continue; } |
bda97eab0
|
835 836 |
for (i = 0; i < pagevec_count(&pvec); i++) { struct page *page = pvec.pages[i]; |
7a5d0fbb2
|
837 |
index = indices[i]; |
83e4fa9c1
|
838 |
if (index >= end) |
bda97eab0
|
839 |
break; |
7a5d0fbb2
|
840 |
if (radix_tree_exceptional_entry(page)) { |
1635f6a74
|
841 842 |
if (unfalloc) continue; |
b1a366500
|
843 844 845 846 847 848 |
if (shmem_free_swap(mapping, index, page)) { /* Swap was replaced by page: retry */ index--; break; } nr_swaps_freed++; |
7a5d0fbb2
|
849 850 |
continue; } |
bda97eab0
|
851 |
lock_page(page); |
800d8c63b
|
852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 |
if (PageTransTail(page)) { /* Middle of THP: zero out the page */ clear_highpage(page); unlock_page(page); /* * Partial thp truncate due 'start' in middle * of THP: don't need to look on these pages * again on !pvec.nr restart. */ if (index != round_down(end, HPAGE_PMD_NR)) start++; continue; } else if (PageTransHuge(page)) { if (index == round_down(end, HPAGE_PMD_NR)) { /* * Range ends in the middle of THP: * zero out the page */ clear_highpage(page); unlock_page(page); continue; } index += HPAGE_PMD_NR - 1; i += HPAGE_PMD_NR - 1; } |
1635f6a74
|
878 |
if (!unfalloc || !PageUptodate(page)) { |
800d8c63b
|
879 880 |
VM_BUG_ON_PAGE(PageTail(page), page); if (page_mapping(page) == mapping) { |
309381fea
|
881 |
VM_BUG_ON_PAGE(PageWriteback(page), page); |
1635f6a74
|
882 |
truncate_inode_page(mapping, page); |
b1a366500
|
883 884 885 886 887 |
} else { /* Page was replaced by swap: retry */ unlock_page(page); index--; break; |
1635f6a74
|
888 |
} |
7a5d0fbb2
|
889 |
} |
bda97eab0
|
890 891 |
unlock_page(page); } |
0cd6144aa
|
892 |
pagevec_remove_exceptionals(&pvec); |
245132643
|
893 |
pagevec_release(&pvec); |
bda97eab0
|
894 895 |
index++; } |
94c1e62df
|
896 |
|
4595ef88d
|
897 |
spin_lock_irq(&info->lock); |
7a5d0fbb2
|
898 |
info->swapped -= nr_swaps_freed; |
1da177e4c
|
899 |
shmem_recalc_inode(inode); |
4595ef88d
|
900 |
spin_unlock_irq(&info->lock); |
1635f6a74
|
901 |
} |
1da177e4c
|
902 |
|
1635f6a74
|
903 904 905 |
void shmem_truncate_range(struct inode *inode, loff_t lstart, loff_t lend) { shmem_undo_range(inode, lstart, lend, false); |
078cd8279
|
906 |
inode->i_ctime = inode->i_mtime = current_time(inode); |
1da177e4c
|
907 |
} |
94c1e62df
|
908 |
EXPORT_SYMBOL_GPL(shmem_truncate_range); |
1da177e4c
|
909 |
|
44a30220b
|
910 911 912 913 914 |
static int shmem_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat) { struct inode *inode = dentry->d_inode; struct shmem_inode_info *info = SHMEM_I(inode); |
d0424c429
|
915 |
if (info->alloced - info->swapped != inode->i_mapping->nrpages) { |
4595ef88d
|
916 |
spin_lock_irq(&info->lock); |
d0424c429
|
917 |
shmem_recalc_inode(inode); |
4595ef88d
|
918 |
spin_unlock_irq(&info->lock); |
d0424c429
|
919 |
} |
44a30220b
|
920 |
generic_fillattr(inode, stat); |
44a30220b
|
921 922 |
return 0; } |
94c1e62df
|
923 |
static int shmem_setattr(struct dentry *dentry, struct iattr *attr) |
1da177e4c
|
924 |
{ |
75c3cfa85
|
925 |
struct inode *inode = d_inode(dentry); |
40e041a2c
|
926 |
struct shmem_inode_info *info = SHMEM_I(inode); |
779750d20
|
927 |
struct shmem_sb_info *sbinfo = SHMEM_SB(inode->i_sb); |
1da177e4c
|
928 |
int error; |
31051c85b
|
929 |
error = setattr_prepare(dentry, attr); |
db78b877f
|
930 931 |
if (error) return error; |
94c1e62df
|
932 933 934 |
if (S_ISREG(inode->i_mode) && (attr->ia_valid & ATTR_SIZE)) { loff_t oldsize = inode->i_size; loff_t newsize = attr->ia_size; |
3889e6e76
|
935 |
|
40e041a2c
|
936 937 938 939 |
/* protected by i_mutex */ if ((newsize < oldsize && (info->seals & F_SEAL_SHRINK)) || (newsize > oldsize && (info->seals & F_SEAL_GROW))) return -EPERM; |
94c1e62df
|
940 |
if (newsize != oldsize) { |
771425179
|
941 942 943 944 |
error = shmem_reacct_size(SHMEM_I(inode)->flags, oldsize, newsize); if (error) return error; |
94c1e62df
|
945 |
i_size_write(inode, newsize); |
078cd8279
|
946 |
inode->i_ctime = inode->i_mtime = current_time(inode); |
94c1e62df
|
947 |
} |
afa2db2fb
|
948 |
if (newsize <= oldsize) { |
94c1e62df
|
949 |
loff_t holebegin = round_up(newsize, PAGE_SIZE); |
d0424c429
|
950 951 952 953 954 955 |
if (oldsize > holebegin) unmap_mapping_range(inode->i_mapping, holebegin, 0, 1); if (info->alloced) shmem_truncate_range(inode, newsize, (loff_t)-1); |
94c1e62df
|
956 |
/* unmap again to remove racily COWed private pages */ |
d0424c429
|
957 958 959 |
if (oldsize > holebegin) unmap_mapping_range(inode->i_mapping, holebegin, 0, 1); |
779750d20
|
960 961 962 963 964 965 966 967 968 969 970 971 972 973 |
/* * Part of the huge page can be beyond i_size: subject * to shrink under memory pressure. */ if (IS_ENABLED(CONFIG_TRANSPARENT_HUGE_PAGECACHE)) { spin_lock(&sbinfo->shrinklist_lock); if (list_empty(&info->shrinklist)) { list_add_tail(&info->shrinklist, &sbinfo->shrinklist); sbinfo->shrinklist_len++; } spin_unlock(&sbinfo->shrinklist_lock); } |
94c1e62df
|
974 |
} |
1da177e4c
|
975 |
} |
db78b877f
|
976 |
setattr_copy(inode, attr); |
db78b877f
|
977 |
if (attr->ia_valid & ATTR_MODE) |
feda821e7
|
978 |
error = posix_acl_chmod(inode, inode->i_mode); |
1da177e4c
|
979 980 |
return error; } |
1f895f75d
|
981 |
static void shmem_evict_inode(struct inode *inode) |
1da177e4c
|
982 |
{ |
1da177e4c
|
983 |
struct shmem_inode_info *info = SHMEM_I(inode); |
779750d20
|
984 |
struct shmem_sb_info *sbinfo = SHMEM_SB(inode->i_sb); |
1da177e4c
|
985 |
|
3889e6e76
|
986 |
if (inode->i_mapping->a_ops == &shmem_aops) { |
1da177e4c
|
987 988 |
shmem_unacct_size(info->flags, inode->i_size); inode->i_size = 0; |
3889e6e76
|
989 |
shmem_truncate_range(inode, 0, (loff_t)-1); |
779750d20
|
990 991 992 993 994 995 996 997 |
if (!list_empty(&info->shrinklist)) { spin_lock(&sbinfo->shrinklist_lock); if (!list_empty(&info->shrinklist)) { list_del_init(&info->shrinklist); sbinfo->shrinklist_len--; } spin_unlock(&sbinfo->shrinklist_lock); } |
1da177e4c
|
998 |
if (!list_empty(&info->swaplist)) { |
cb5f7b9a4
|
999 |
mutex_lock(&shmem_swaplist_mutex); |
1da177e4c
|
1000 |
list_del_init(&info->swaplist); |
cb5f7b9a4
|
1001 |
mutex_unlock(&shmem_swaplist_mutex); |
1da177e4c
|
1002 |
} |
3ed47db34
|
1003 |
} |
b09e0fa4b
|
1004 |
|
38f386574
|
1005 |
simple_xattrs_free(&info->xattrs); |
0f3c42f52
|
1006 |
WARN_ON(inode->i_blocks); |
5b04c6890
|
1007 |
shmem_free_inode(inode->i_sb); |
dbd5768f8
|
1008 |
clear_inode(inode); |
1da177e4c
|
1009 |
} |
46f65ec15
|
1010 1011 1012 |
/* * If swap found in inode, free it and move page from swapcache to filecache. */ |
41ffe5d5c
|
1013 |
static int shmem_unuse_inode(struct shmem_inode_info *info, |
bde05d1cc
|
1014 |
swp_entry_t swap, struct page **pagep) |
1da177e4c
|
1015 |
{ |
285b2c4fd
|
1016 |
struct address_space *mapping = info->vfs_inode.i_mapping; |
46f65ec15
|
1017 |
void *radswap; |
41ffe5d5c
|
1018 |
pgoff_t index; |
bde05d1cc
|
1019 1020 |
gfp_t gfp; int error = 0; |
1da177e4c
|
1021 |
|
46f65ec15
|
1022 |
radswap = swp_to_radix_entry(swap); |
e504f3fdd
|
1023 |
index = radix_tree_locate_item(&mapping->page_tree, radswap); |
46f65ec15
|
1024 |
if (index == -1) |
00501b531
|
1025 |
return -EAGAIN; /* tell shmem_unuse we found nothing */ |
2e0e26c76
|
1026 |
|
1b1b32f2c
|
1027 1028 |
/* * Move _head_ to start search for next from here. |
1f895f75d
|
1029 |
* But be careful: shmem_evict_inode checks list_empty without taking |
1b1b32f2c
|
1030 |
* mutex, and there's an instant in list_move_tail when info->swaplist |
285b2c4fd
|
1031 |
* would appear empty, if it were the only one on shmem_swaplist. |
1b1b32f2c
|
1032 1033 1034 |
*/ if (shmem_swaplist.next != &info->swaplist) list_move_tail(&shmem_swaplist, &info->swaplist); |
2e0e26c76
|
1035 |
|
bde05d1cc
|
1036 1037 1038 1039 1040 1041 1042 |
gfp = mapping_gfp_mask(mapping); if (shmem_should_replace_page(*pagep, gfp)) { mutex_unlock(&shmem_swaplist_mutex); error = shmem_replace_page(pagep, gfp, info, index); mutex_lock(&shmem_swaplist_mutex); /* * We needed to drop mutex to make that restrictive page |
0142ef6cd
|
1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 |
* allocation, but the inode might have been freed while we * dropped it: although a racing shmem_evict_inode() cannot * complete without emptying the radix_tree, our page lock * on this swapcache page is not enough to prevent that - * free_swap_and_cache() of our swap entry will only * trylock_page(), removing swap from radix_tree whatever. * * We must not proceed to shmem_add_to_page_cache() if the * inode has been freed, but of course we cannot rely on * inode or mapping or info to check that. However, we can * safely check if our swap entry is still in use (and here * it can't have got reused for another page): if it's still * in use, then the inode cannot have been freed yet, and we * can safely proceed (if it's no longer in use, that tells * nothing about the inode, but we don't need to unuse swap). |
bde05d1cc
|
1058 1059 1060 1061 |
*/ if (!page_swapcount(*pagep)) error = -ENOENT; } |
d13d14430
|
1062 |
/* |
778dd893a
|
1063 1064 1065 |
* 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
|
1066 |
*/ |
bde05d1cc
|
1067 1068 |
if (!error) error = shmem_add_to_page_cache(*pagep, mapping, index, |
fed400a18
|
1069 |
radswap); |
48f170fb7
|
1070 |
if (error != -ENOMEM) { |
46f65ec15
|
1071 1072 1073 1074 |
/* * Truncation and eviction use free_swap_and_cache(), which * only does trylock page: if we raced, best clean up here. */ |
bde05d1cc
|
1075 1076 |
delete_from_swap_cache(*pagep); set_page_dirty(*pagep); |
46f65ec15
|
1077 |
if (!error) { |
4595ef88d
|
1078 |
spin_lock_irq(&info->lock); |
46f65ec15
|
1079 |
info->swapped--; |
4595ef88d
|
1080 |
spin_unlock_irq(&info->lock); |
46f65ec15
|
1081 1082 |
swap_free(swap); } |
1da177e4c
|
1083 |
} |
2e0e26c76
|
1084 |
return error; |
1da177e4c
|
1085 1086 1087 |
} /* |
46f65ec15
|
1088 |
* Search through swapped inodes to find and replace swap by page. |
1da177e4c
|
1089 |
*/ |
41ffe5d5c
|
1090 |
int shmem_unuse(swp_entry_t swap, struct page *page) |
1da177e4c
|
1091 |
{ |
41ffe5d5c
|
1092 |
struct list_head *this, *next; |
1da177e4c
|
1093 |
struct shmem_inode_info *info; |
00501b531
|
1094 |
struct mem_cgroup *memcg; |
bde05d1cc
|
1095 1096 1097 1098 |
int error = 0; /* * There's a faint possibility that swap page was replaced before |
0142ef6cd
|
1099 |
* caller locked it: caller will come back later with the right page. |
bde05d1cc
|
1100 |
*/ |
0142ef6cd
|
1101 |
if (unlikely(!PageSwapCache(page) || page_private(page) != swap.val)) |
bde05d1cc
|
1102 |
goto out; |
778dd893a
|
1103 1104 1105 1106 1107 |
/* * 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
|
1108 |
*/ |
f627c2f53
|
1109 1110 |
error = mem_cgroup_try_charge(page, current->mm, GFP_KERNEL, &memcg, false); |
778dd893a
|
1111 1112 |
if (error) goto out; |
46f65ec15
|
1113 |
/* No radix_tree_preload: swap entry keeps a place for page in tree */ |
00501b531
|
1114 |
error = -EAGAIN; |
1da177e4c
|
1115 |
|
cb5f7b9a4
|
1116 |
mutex_lock(&shmem_swaplist_mutex); |
41ffe5d5c
|
1117 1118 |
list_for_each_safe(this, next, &shmem_swaplist) { info = list_entry(this, struct shmem_inode_info, swaplist); |
285b2c4fd
|
1119 |
if (info->swapped) |
00501b531
|
1120 |
error = shmem_unuse_inode(info, swap, &page); |
6922c0c7a
|
1121 1122 |
else list_del_init(&info->swaplist); |
cb5f7b9a4
|
1123 |
cond_resched(); |
00501b531
|
1124 |
if (error != -EAGAIN) |
778dd893a
|
1125 |
break; |
00501b531
|
1126 |
/* found nothing in this: move on to search the next */ |
1da177e4c
|
1127 |
} |
cb5f7b9a4
|
1128 |
mutex_unlock(&shmem_swaplist_mutex); |
778dd893a
|
1129 |
|
00501b531
|
1130 1131 1132 |
if (error) { if (error != -ENOMEM) error = 0; |
f627c2f53
|
1133 |
mem_cgroup_cancel_charge(page, memcg, false); |
00501b531
|
1134 |
} else |
f627c2f53
|
1135 |
mem_cgroup_commit_charge(page, memcg, true, false); |
778dd893a
|
1136 |
out: |
aaa468653
|
1137 |
unlock_page(page); |
09cbfeaf1
|
1138 |
put_page(page); |
778dd893a
|
1139 |
return error; |
1da177e4c
|
1140 1141 1142 1143 1144 1145 1146 1147 |
} /* * 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
|
1148 |
struct address_space *mapping; |
1da177e4c
|
1149 |
struct inode *inode; |
6922c0c7a
|
1150 1151 |
swp_entry_t swap; pgoff_t index; |
1da177e4c
|
1152 |
|
800d8c63b
|
1153 |
VM_BUG_ON_PAGE(PageCompound(page), page); |
1da177e4c
|
1154 |
BUG_ON(!PageLocked(page)); |
1da177e4c
|
1155 1156 1157 1158 1159 1160 |
mapping = page->mapping; index = page->index; inode = mapping->host; info = SHMEM_I(inode); if (info->flags & VM_LOCKED) goto redirty; |
d9fe526a8
|
1161 |
if (!total_swap_pages) |
1da177e4c
|
1162 |
goto redirty; |
d9fe526a8
|
1163 |
/* |
97b713ba3
|
1164 1165 1166 1167 1168 |
* Our capabilities prevent regular writeback or sync from ever calling * shmem_writepage; but a stacking filesystem might use ->writepage of * its underlying filesystem, in which case tmpfs should write out to * swap only in response to memory pressure, and not for the writeback * threads or sync. |
d9fe526a8
|
1169 |
*/ |
48f170fb7
|
1170 1171 1172 1173 |
if (!wbc->for_reclaim) { WARN_ON_ONCE(1); /* Still happens? Tell us about it! */ goto redirty; } |
1635f6a74
|
1174 1175 1176 1177 1178 |
/* * This is somewhat ridiculous, but without plumbing a SWAP_MAP_FALLOC * value into swapfile.c, the only way we can correctly account for a * fallocated page arriving here is now to initialize it and write it. |
1aac14003
|
1179 1180 1181 1182 1183 1184 |
* * That's okay for a page already fallocated earlier, but if we have * not yet completed the fallocation, then (a) we want to keep track * of this page in case we have to undo it, and (b) it may not be a * good idea to continue anyway, once we're pushing into swap. So * reactivate the page, and let shmem_fallocate() quit when too many. |
1635f6a74
|
1185 1186 |
*/ if (!PageUptodate(page)) { |
1aac14003
|
1187 1188 1189 1190 1191 |
if (inode->i_private) { struct shmem_falloc *shmem_falloc; spin_lock(&inode->i_lock); shmem_falloc = inode->i_private; if (shmem_falloc && |
8e205f779
|
1192 |
!shmem_falloc->waitq && |
1aac14003
|
1193 1194 1195 1196 1197 1198 1199 1200 1201 |
index >= shmem_falloc->start && index < shmem_falloc->next) shmem_falloc->nr_unswapped++; else shmem_falloc = NULL; spin_unlock(&inode->i_lock); if (shmem_falloc) goto redirty; } |
1635f6a74
|
1202 1203 1204 1205 |
clear_highpage(page); flush_dcache_page(page); SetPageUptodate(page); } |
48f170fb7
|
1206 1207 1208 |
swap = get_swap_page(); if (!swap.val) goto redirty; |
d9fe526a8
|
1209 |
|
37e843511
|
1210 1211 |
if (mem_cgroup_try_charge_swap(page, swap)) goto free_swap; |
b1dea800a
|
1212 1213 |
/* * Add inode to shmem_unuse()'s list of swapped-out inodes, |
6922c0c7a
|
1214 1215 |
* if it's not already there. Do it now before the page is * moved to swap cache, when its pagelock no longer protects |
b1dea800a
|
1216 |
* the inode from eviction. But don't unlock the mutex until |
6922c0c7a
|
1217 1218 |
* we've incremented swapped, because shmem_unuse_inode() will * prune a !swapped inode from the swaplist under this mutex. |
b1dea800a
|
1219 |
*/ |
48f170fb7
|
1220 1221 1222 |
mutex_lock(&shmem_swaplist_mutex); if (list_empty(&info->swaplist)) list_add_tail(&info->swaplist, &shmem_swaplist); |
b1dea800a
|
1223 |
|
48f170fb7
|
1224 |
if (add_to_swap_cache(page, swap, GFP_ATOMIC) == 0) { |
4595ef88d
|
1225 |
spin_lock_irq(&info->lock); |
6922c0c7a
|
1226 |
shmem_recalc_inode(inode); |
267a4c76b
|
1227 |
info->swapped++; |
4595ef88d
|
1228 |
spin_unlock_irq(&info->lock); |
6922c0c7a
|
1229 |
|
267a4c76b
|
1230 1231 |
swap_shmem_alloc(swap); shmem_delete_from_page_cache(page, swp_to_radix_entry(swap)); |
6922c0c7a
|
1232 |
mutex_unlock(&shmem_swaplist_mutex); |
d9fe526a8
|
1233 |
BUG_ON(page_mapped(page)); |
9fab5619b
|
1234 |
swap_writepage(page, wbc); |
1da177e4c
|
1235 1236 |
return 0; } |
6922c0c7a
|
1237 |
mutex_unlock(&shmem_swaplist_mutex); |
37e843511
|
1238 |
free_swap: |
0a31bc97c
|
1239 |
swapcache_free(swap); |
1da177e4c
|
1240 1241 |
redirty: set_page_dirty(page); |
d9fe526a8
|
1242 1243 1244 1245 |
if (wbc->for_reclaim) return AOP_WRITEPAGE_ACTIVATE; /* Return with page locked */ unlock_page(page); return 0; |
1da177e4c
|
1246 |
} |
75edd345e
|
1247 |
#if defined(CONFIG_NUMA) && defined(CONFIG_TMPFS) |
71fe804b6
|
1248 |
static void shmem_show_mpol(struct seq_file *seq, struct mempolicy *mpol) |
680d794ba
|
1249 |
{ |
095f1fc4e
|
1250 |
char buffer[64]; |
680d794ba
|
1251 |
|
71fe804b6
|
1252 |
if (!mpol || mpol->mode == MPOL_DEFAULT) |
095f1fc4e
|
1253 |
return; /* show nothing */ |
680d794ba
|
1254 |
|
a7a88b237
|
1255 |
mpol_to_str(buffer, sizeof(buffer), mpol); |
095f1fc4e
|
1256 1257 |
seq_printf(seq, ",mpol=%s", buffer); |
680d794ba
|
1258 |
} |
71fe804b6
|
1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 |
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; } |
75edd345e
|
1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 |
#else /* !CONFIG_NUMA || !CONFIG_TMPFS */ static inline void shmem_show_mpol(struct seq_file *seq, struct mempolicy *mpol) { } static inline struct mempolicy *shmem_get_sbmpol(struct shmem_sb_info *sbinfo) { return NULL; } #endif /* CONFIG_NUMA && CONFIG_TMPFS */ #ifndef CONFIG_NUMA #define vm_policy vm_private_data #endif |
680d794ba
|
1283 |
|
800d8c63b
|
1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 |
static void shmem_pseudo_vma_init(struct vm_area_struct *vma, struct shmem_inode_info *info, pgoff_t index) { /* Create a pseudo vma that just contains the policy */ vma->vm_start = 0; /* Bias interleave by inode number to distribute better across nodes */ vma->vm_pgoff = index + info->vfs_inode.i_ino; vma->vm_ops = NULL; vma->vm_policy = mpol_shared_policy_lookup(&info->policy, index); } static void shmem_pseudo_vma_destroy(struct vm_area_struct *vma) { /* Drop reference taken by mpol_shared_policy_lookup() */ mpol_cond_put(vma->vm_policy); } |
41ffe5d5c
|
1300 1301 |
static struct page *shmem_swapin(swp_entry_t swap, gfp_t gfp, struct shmem_inode_info *info, pgoff_t index) |
1da177e4c
|
1302 |
{ |
1da177e4c
|
1303 |
struct vm_area_struct pvma; |
18a2f371f
|
1304 |
struct page *page; |
52cd3b074
|
1305 |
|
800d8c63b
|
1306 |
shmem_pseudo_vma_init(&pvma, info, index); |
18a2f371f
|
1307 |
page = swapin_readahead(swap, gfp, &pvma, 0); |
800d8c63b
|
1308 |
shmem_pseudo_vma_destroy(&pvma); |
18a2f371f
|
1309 |
|
800d8c63b
|
1310 1311 1312 1313 1314 1315 1316 1317 1318 |
return page; } static struct page *shmem_alloc_hugepage(gfp_t gfp, struct shmem_inode_info *info, pgoff_t index) { struct vm_area_struct pvma; struct inode *inode = &info->vfs_inode; struct address_space *mapping = inode->i_mapping; |
4620a06e4
|
1319 |
pgoff_t idx, hindex; |
800d8c63b
|
1320 1321 |
void __rcu **results; struct page *page; |
e496cf3d7
|
1322 |
if (!IS_ENABLED(CONFIG_TRANSPARENT_HUGE_PAGECACHE)) |
800d8c63b
|
1323 |
return NULL; |
4620a06e4
|
1324 |
hindex = round_down(index, HPAGE_PMD_NR); |
800d8c63b
|
1325 1326 1327 1328 1329 1330 1331 |
rcu_read_lock(); if (radix_tree_gang_lookup_slot(&mapping->page_tree, &results, &idx, hindex, 1) && idx < hindex + HPAGE_PMD_NR) { rcu_read_unlock(); return NULL; } rcu_read_unlock(); |
18a2f371f
|
1332 |
|
800d8c63b
|
1333 1334 1335 1336 1337 1338 |
shmem_pseudo_vma_init(&pvma, info, hindex); page = alloc_pages_vma(gfp | __GFP_COMP | __GFP_NORETRY | __GFP_NOWARN, HPAGE_PMD_ORDER, &pvma, 0, numa_node_id(), true); shmem_pseudo_vma_destroy(&pvma); if (page) prep_transhuge_page(page); |
18a2f371f
|
1339 |
return page; |
1da177e4c
|
1340 |
} |
02098feaa
|
1341 |
static struct page *shmem_alloc_page(gfp_t gfp, |
41ffe5d5c
|
1342 |
struct shmem_inode_info *info, pgoff_t index) |
1da177e4c
|
1343 1344 |
{ struct vm_area_struct pvma; |
18a2f371f
|
1345 |
struct page *page; |
1da177e4c
|
1346 |
|
800d8c63b
|
1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 |
shmem_pseudo_vma_init(&pvma, info, index); page = alloc_page_vma(gfp, &pvma, 0); shmem_pseudo_vma_destroy(&pvma); return page; } static struct page *shmem_alloc_and_acct_page(gfp_t gfp, struct shmem_inode_info *info, struct shmem_sb_info *sbinfo, pgoff_t index, bool huge) { struct page *page; int nr; int err = -ENOSPC; |
52cd3b074
|
1361 |
|
e496cf3d7
|
1362 |
if (!IS_ENABLED(CONFIG_TRANSPARENT_HUGE_PAGECACHE)) |
800d8c63b
|
1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 |
huge = false; nr = huge ? HPAGE_PMD_NR : 1; if (shmem_acct_block(info->flags, nr)) goto failed; if (sbinfo->max_blocks) { if (percpu_counter_compare(&sbinfo->used_blocks, sbinfo->max_blocks - nr) > 0) goto unacct; percpu_counter_add(&sbinfo->used_blocks, nr); } if (huge) page = shmem_alloc_hugepage(gfp, info, index); else page = shmem_alloc_page(gfp, info, index); |
75edd345e
|
1379 1380 1381 |
if (page) { __SetPageLocked(page); __SetPageSwapBacked(page); |
800d8c63b
|
1382 |
return page; |
75edd345e
|
1383 |
} |
18a2f371f
|
1384 |
|
800d8c63b
|
1385 1386 1387 1388 1389 1390 1391 |
err = -ENOMEM; if (sbinfo->max_blocks) percpu_counter_add(&sbinfo->used_blocks, -nr); unacct: shmem_unacct_blocks(info->flags, nr); failed: return ERR_PTR(err); |
1da177e4c
|
1392 |
} |
71fe804b6
|
1393 |
|
1da177e4c
|
1394 |
/* |
bde05d1cc
|
1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 |
* When a page is moved from swapcache to shmem filecache (either by the * usual swapin of shmem_getpage_gfp(), or by the less common swapoff of * shmem_unuse_inode()), it may have been read in earlier from swap, in * ignorance of the mapping it belongs to. If that mapping has special * constraints (like the gma500 GEM driver, which requires RAM below 4GB), * we may need to copy to a suitable page before moving to filecache. * * In a future release, this may well be extended to respect cpuset and * NUMA mempolicy, and applied also to anonymous pages in do_swap_page(); * but for now it is a simple matter of zone. */ static bool shmem_should_replace_page(struct page *page, gfp_t gfp) { return page_zonenum(page) > gfp_zone(gfp); } static int shmem_replace_page(struct page **pagep, gfp_t gfp, struct shmem_inode_info *info, pgoff_t index) { struct page *oldpage, *newpage; struct address_space *swap_mapping; pgoff_t swap_index; int error; oldpage = *pagep; swap_index = page_private(oldpage); swap_mapping = page_mapping(oldpage); /* * We have arrived here because our zones are constrained, so don't * limit chance of success by further cpuset and node constraints. */ gfp &= ~GFP_CONSTRAINT_MASK; newpage = shmem_alloc_page(gfp, info, index); if (!newpage) return -ENOMEM; |
bde05d1cc
|
1431 |
|
09cbfeaf1
|
1432 |
get_page(newpage); |
bde05d1cc
|
1433 |
copy_highpage(newpage, oldpage); |
0142ef6cd
|
1434 |
flush_dcache_page(newpage); |
bde05d1cc
|
1435 |
|
9956edf37
|
1436 1437 |
__SetPageLocked(newpage); __SetPageSwapBacked(newpage); |
bde05d1cc
|
1438 |
SetPageUptodate(newpage); |
bde05d1cc
|
1439 |
set_page_private(newpage, swap_index); |
bde05d1cc
|
1440 1441 1442 1443 1444 1445 1446 1447 1448 |
SetPageSwapCache(newpage); /* * Our caller will very soon move newpage out of swapcache, but it's * a nice clean interface for us to replace oldpage by newpage there. */ spin_lock_irq(&swap_mapping->tree_lock); error = shmem_radix_tree_replace(swap_mapping, swap_index, oldpage, newpage); |
0142ef6cd
|
1449 |
if (!error) { |
11fb99898
|
1450 1451 |
__inc_node_page_state(newpage, NR_FILE_PAGES); __dec_node_page_state(oldpage, NR_FILE_PAGES); |
0142ef6cd
|
1452 |
} |
bde05d1cc
|
1453 |
spin_unlock_irq(&swap_mapping->tree_lock); |
bde05d1cc
|
1454 |
|
0142ef6cd
|
1455 1456 1457 1458 1459 1460 1461 1462 |
if (unlikely(error)) { /* * Is this possible? I think not, now that our callers check * both PageSwapCache and page_private after getting page lock; * but be defensive. Reverse old to newpage for clear and free. */ oldpage = newpage; } else { |
6a93ca8fd
|
1463 |
mem_cgroup_migrate(oldpage, newpage); |
0142ef6cd
|
1464 1465 1466 |
lru_cache_add_anon(newpage); *pagep = newpage; } |
bde05d1cc
|
1467 1468 1469 1470 1471 |
ClearPageSwapCache(oldpage); set_page_private(oldpage, 0); unlock_page(oldpage); |
09cbfeaf1
|
1472 1473 |
put_page(oldpage); put_page(oldpage); |
0142ef6cd
|
1474 |
return error; |
bde05d1cc
|
1475 1476 1477 |
} /* |
68da9f055
|
1478 |
* shmem_getpage_gfp - find page in cache, or get from swap, or allocate |
1da177e4c
|
1479 1480 1481 |
* * 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 |
9e18eb293
|
1482 1483 1484 1485 |
* entry since a page cannot live in both the swap and page cache. * * fault_mm and fault_type are only supplied by shmem_fault: * otherwise they are NULL. |
1da177e4c
|
1486 |
*/ |
41ffe5d5c
|
1487 |
static int shmem_getpage_gfp(struct inode *inode, pgoff_t index, |
9e18eb293
|
1488 1489 |
struct page **pagep, enum sgp_type sgp, gfp_t gfp, struct mm_struct *fault_mm, int *fault_type) |
1da177e4c
|
1490 1491 |
{ struct address_space *mapping = inode->i_mapping; |
54af60421
|
1492 |
struct shmem_inode_info *info; |
1da177e4c
|
1493 |
struct shmem_sb_info *sbinfo; |
9e18eb293
|
1494 |
struct mm_struct *charge_mm; |
00501b531
|
1495 |
struct mem_cgroup *memcg; |
27ab70062
|
1496 |
struct page *page; |
1da177e4c
|
1497 |
swp_entry_t swap; |
657e3038c
|
1498 |
enum sgp_type sgp_huge = sgp; |
800d8c63b
|
1499 |
pgoff_t hindex = index; |
1da177e4c
|
1500 |
int error; |
54af60421
|
1501 |
int once = 0; |
1635f6a74
|
1502 |
int alloced = 0; |
1da177e4c
|
1503 |
|
09cbfeaf1
|
1504 |
if (index > (MAX_LFS_FILESIZE >> PAGE_SHIFT)) |
1da177e4c
|
1505 |
return -EFBIG; |
657e3038c
|
1506 1507 |
if (sgp == SGP_NOHUGE || sgp == SGP_HUGE) sgp = SGP_CACHE; |
1da177e4c
|
1508 |
repeat: |
54af60421
|
1509 |
swap.val = 0; |
0cd6144aa
|
1510 |
page = find_lock_entry(mapping, index); |
54af60421
|
1511 1512 1513 1514 |
if (radix_tree_exceptional_entry(page)) { swap = radix_to_swp_entry(page); page = NULL; } |
75edd345e
|
1515 |
if (sgp <= SGP_CACHE && |
09cbfeaf1
|
1516 |
((loff_t)index << PAGE_SHIFT) >= i_size_read(inode)) { |
54af60421
|
1517 |
error = -EINVAL; |
267a4c76b
|
1518 |
goto unlock; |
54af60421
|
1519 |
} |
66d2f4d28
|
1520 1521 |
if (page && sgp == SGP_WRITE) mark_page_accessed(page); |
1635f6a74
|
1522 1523 1524 1525 1526 |
/* fallocated page? */ if (page && !PageUptodate(page)) { if (sgp != SGP_READ) goto clear; unlock_page(page); |
09cbfeaf1
|
1527 |
put_page(page); |
1635f6a74
|
1528 1529 |
page = NULL; } |
54af60421
|
1530 |
if (page || (sgp == SGP_READ && !swap.val)) { |
54af60421
|
1531 1532 |
*pagep = page; return 0; |
27ab70062
|
1533 1534 1535 |
} /* |
54af60421
|
1536 1537 |
* Fast cache lookup did not find it: * bring it back from swap or allocate. |
27ab70062
|
1538 |
*/ |
54af60421
|
1539 1540 |
info = SHMEM_I(inode); sbinfo = SHMEM_SB(inode->i_sb); |
9e18eb293
|
1541 |
charge_mm = fault_mm ? : current->mm; |
1da177e4c
|
1542 |
|
1da177e4c
|
1543 1544 |
if (swap.val) { /* Look it up and read it in.. */ |
27ab70062
|
1545 1546 |
page = lookup_swap_cache(swap); if (!page) { |
9e18eb293
|
1547 1548 |
/* Or update major stats only when swapin succeeds?? */ if (fault_type) { |
68da9f055
|
1549 |
*fault_type |= VM_FAULT_MAJOR; |
9e18eb293
|
1550 1551 1552 1553 |
count_vm_event(PGMAJFAULT); mem_cgroup_count_vm_event(fault_mm, PGMAJFAULT); } /* Here we actually start the io */ |
41ffe5d5c
|
1554 |
page = shmem_swapin(swap, gfp, info, index); |
27ab70062
|
1555 |
if (!page) { |
54af60421
|
1556 1557 |
error = -ENOMEM; goto failed; |
1da177e4c
|
1558 |
} |
1da177e4c
|
1559 1560 1561 |
} /* We have to do this with page locked to prevent races */ |
54af60421
|
1562 |
lock_page(page); |
0142ef6cd
|
1563 |
if (!PageSwapCache(page) || page_private(page) != swap.val || |
d18992286
|
1564 |
!shmem_confirm_swap(mapping, index, swap)) { |
bde05d1cc
|
1565 |
error = -EEXIST; /* try again */ |
d18992286
|
1566 |
goto unlock; |
bde05d1cc
|
1567 |
} |
27ab70062
|
1568 |
if (!PageUptodate(page)) { |
1da177e4c
|
1569 |
error = -EIO; |
54af60421
|
1570 |
goto failed; |
1da177e4c
|
1571 |
} |
54af60421
|
1572 |
wait_on_page_writeback(page); |
bde05d1cc
|
1573 1574 1575 1576 |
if (shmem_should_replace_page(page, gfp)) { error = shmem_replace_page(&page, gfp, info, index); if (error) goto failed; |
1da177e4c
|
1577 |
} |
27ab70062
|
1578 |
|
9e18eb293
|
1579 |
error = mem_cgroup_try_charge(page, charge_mm, gfp, &memcg, |
f627c2f53
|
1580 |
false); |
d18992286
|
1581 |
if (!error) { |
aa3b18955
|
1582 |
error = shmem_add_to_page_cache(page, mapping, index, |
fed400a18
|
1583 |
swp_to_radix_entry(swap)); |
215c02bc3
|
1584 1585 1586 1587 1588 1589 1590 1591 |
/* * We already confirmed swap under page lock, and make * no memory allocation here, so usually no possibility * of error; but free_swap_and_cache() only trylocks a * page, so it is just possible that the entry has been * truncated or holepunched since swap was confirmed. * shmem_undo_range() will have done some of the * unaccounting, now delete_from_swap_cache() will do |
93aa7d952
|
1592 |
* the rest. |
215c02bc3
|
1593 1594 1595 |
* Reset swap.val? No, leave it so "failed" goes back to * "repeat": reading a hole and writing should succeed. */ |
00501b531
|
1596 |
if (error) { |
f627c2f53
|
1597 |
mem_cgroup_cancel_charge(page, memcg, false); |
215c02bc3
|
1598 |
delete_from_swap_cache(page); |
00501b531
|
1599 |
} |
d18992286
|
1600 |
} |
54af60421
|
1601 1602 |
if (error) goto failed; |
f627c2f53
|
1603 |
mem_cgroup_commit_charge(page, memcg, true, false); |
00501b531
|
1604 |
|
4595ef88d
|
1605 |
spin_lock_irq(&info->lock); |
285b2c4fd
|
1606 |
info->swapped--; |
54af60421
|
1607 |
shmem_recalc_inode(inode); |
4595ef88d
|
1608 |
spin_unlock_irq(&info->lock); |
54af60421
|
1609 |
|
66d2f4d28
|
1610 1611 |
if (sgp == SGP_WRITE) mark_page_accessed(page); |
54af60421
|
1612 |
delete_from_swap_cache(page); |
27ab70062
|
1613 1614 |
set_page_dirty(page); swap_free(swap); |
54af60421
|
1615 |
} else { |
800d8c63b
|
1616 1617 1618 |
/* shmem_symlink() */ if (mapping->a_ops != &shmem_aops) goto alloc_nohuge; |
657e3038c
|
1619 |
if (shmem_huge == SHMEM_HUGE_DENY || sgp_huge == SGP_NOHUGE) |
800d8c63b
|
1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 |
goto alloc_nohuge; if (shmem_huge == SHMEM_HUGE_FORCE) goto alloc_huge; switch (sbinfo->huge) { loff_t i_size; pgoff_t off; case SHMEM_HUGE_NEVER: goto alloc_nohuge; case SHMEM_HUGE_WITHIN_SIZE: off = round_up(index, HPAGE_PMD_NR); i_size = round_up(i_size_read(inode), PAGE_SIZE); if (i_size >= HPAGE_PMD_SIZE && i_size >> PAGE_SHIFT >= off) goto alloc_huge; /* fallthrough */ case SHMEM_HUGE_ADVISE: |
657e3038c
|
1636 1637 1638 |
if (sgp_huge == SGP_HUGE) goto alloc_huge; /* TODO: implement fadvise() hints */ |
800d8c63b
|
1639 |
goto alloc_nohuge; |
54af60421
|
1640 |
} |
1da177e4c
|
1641 |
|
800d8c63b
|
1642 1643 1644 1645 1646 1647 |
alloc_huge: page = shmem_alloc_and_acct_page(gfp, info, sbinfo, index, true); if (IS_ERR(page)) { alloc_nohuge: page = shmem_alloc_and_acct_page(gfp, info, sbinfo, index, false); |
1da177e4c
|
1648 |
} |
800d8c63b
|
1649 |
if (IS_ERR(page)) { |
779750d20
|
1650 |
int retry = 5; |
800d8c63b
|
1651 1652 |
error = PTR_ERR(page); page = NULL; |
779750d20
|
1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 |
if (error != -ENOSPC) goto failed; /* * Try to reclaim some spece by splitting a huge page * beyond i_size on the filesystem. */ while (retry--) { int ret; ret = shmem_unused_huge_shrink(sbinfo, NULL, 1); if (ret == SHRINK_STOP) break; if (ret) goto alloc_nohuge; } |
800d8c63b
|
1667 1668 1669 1670 1671 1672 1673 |
goto failed; } if (PageTransHuge(page)) hindex = round_down(index, HPAGE_PMD_NR); else hindex = index; |
66d2f4d28
|
1674 |
if (sgp == SGP_WRITE) |
eb39d618f
|
1675 |
__SetPageReferenced(page); |
66d2f4d28
|
1676 |
|
9e18eb293
|
1677 |
error = mem_cgroup_try_charge(page, charge_mm, gfp, &memcg, |
800d8c63b
|
1678 |
PageTransHuge(page)); |
54af60421
|
1679 |
if (error) |
800d8c63b
|
1680 1681 1682 |
goto unacct; error = radix_tree_maybe_preload_order(gfp & GFP_RECLAIM_MASK, compound_order(page)); |
b065b4321
|
1683 |
if (!error) { |
800d8c63b
|
1684 |
error = shmem_add_to_page_cache(page, mapping, hindex, |
fed400a18
|
1685 |
NULL); |
b065b4321
|
1686 1687 1688 |
radix_tree_preload_end(); } if (error) { |
800d8c63b
|
1689 1690 1691 |
mem_cgroup_cancel_charge(page, memcg, PageTransHuge(page)); goto unacct; |
b065b4321
|
1692 |
} |
800d8c63b
|
1693 1694 |
mem_cgroup_commit_charge(page, memcg, false, PageTransHuge(page)); |
54af60421
|
1695 |
lru_cache_add_anon(page); |
4595ef88d
|
1696 |
spin_lock_irq(&info->lock); |
800d8c63b
|
1697 1698 |
info->alloced += 1 << compound_order(page); inode->i_blocks += BLOCKS_PER_PAGE << compound_order(page); |
54af60421
|
1699 |
shmem_recalc_inode(inode); |
4595ef88d
|
1700 |
spin_unlock_irq(&info->lock); |
1635f6a74
|
1701 |
alloced = true; |
54af60421
|
1702 |
|
779750d20
|
1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 |
if (PageTransHuge(page) && DIV_ROUND_UP(i_size_read(inode), PAGE_SIZE) < hindex + HPAGE_PMD_NR - 1) { /* * Part of the huge page is beyond i_size: subject * to shrink under memory pressure. */ spin_lock(&sbinfo->shrinklist_lock); if (list_empty(&info->shrinklist)) { list_add_tail(&info->shrinklist, &sbinfo->shrinklist); sbinfo->shrinklist_len++; } spin_unlock(&sbinfo->shrinklist_lock); } |
ec9516fbc
|
1718 |
/* |
1635f6a74
|
1719 1720 1721 1722 1723 1724 1725 1726 1727 |
* Let SGP_FALLOC use the SGP_WRITE optimization on a new page. */ if (sgp == SGP_FALLOC) sgp = SGP_WRITE; clear: /* * Let SGP_WRITE caller clear ends if write does not fill page; * but SGP_FALLOC on a page fallocated earlier must initialize * it now, lest undo on failure cancel our earlier guarantee. |
ec9516fbc
|
1728 |
*/ |
800d8c63b
|
1729 1730 1731 1732 1733 1734 1735 1736 1737 |
if (sgp != SGP_WRITE && !PageUptodate(page)) { struct page *head = compound_head(page); int i; for (i = 0; i < (1 << compound_order(head)); i++) { clear_highpage(head + i); flush_dcache_page(head + i); } SetPageUptodate(head); |
ec9516fbc
|
1738 |
} |
1da177e4c
|
1739 |
} |
bde05d1cc
|
1740 |
|
54af60421
|
1741 |
/* Perhaps the file has been truncated since we checked */ |
75edd345e
|
1742 |
if (sgp <= SGP_CACHE && |
09cbfeaf1
|
1743 |
((loff_t)index << PAGE_SHIFT) >= i_size_read(inode)) { |
267a4c76b
|
1744 1745 1746 |
if (alloced) { ClearPageDirty(page); delete_from_page_cache(page); |
4595ef88d
|
1747 |
spin_lock_irq(&info->lock); |
267a4c76b
|
1748 |
shmem_recalc_inode(inode); |
4595ef88d
|
1749 |
spin_unlock_irq(&info->lock); |
267a4c76b
|
1750 |
} |
54af60421
|
1751 |
error = -EINVAL; |
267a4c76b
|
1752 |
goto unlock; |
e83c32e8f
|
1753 |
} |
800d8c63b
|
1754 |
*pagep = page + index - hindex; |
54af60421
|
1755 |
return 0; |
1da177e4c
|
1756 |
|
59a16ead5
|
1757 |
/* |
54af60421
|
1758 |
* Error recovery. |
59a16ead5
|
1759 |
*/ |
54af60421
|
1760 |
unacct: |
800d8c63b
|
1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 |
if (sbinfo->max_blocks) percpu_counter_sub(&sbinfo->used_blocks, 1 << compound_order(page)); shmem_unacct_blocks(info->flags, 1 << compound_order(page)); if (PageTransHuge(page)) { unlock_page(page); put_page(page); goto alloc_nohuge; } |
54af60421
|
1771 |
failed: |
267a4c76b
|
1772 |
if (swap.val && !shmem_confirm_swap(mapping, index, swap)) |
d18992286
|
1773 1774 |
error = -EEXIST; unlock: |
27ab70062
|
1775 |
if (page) { |
54af60421
|
1776 |
unlock_page(page); |
09cbfeaf1
|
1777 |
put_page(page); |
54af60421
|
1778 1779 1780 |
} if (error == -ENOSPC && !once++) { info = SHMEM_I(inode); |
4595ef88d
|
1781 |
spin_lock_irq(&info->lock); |
54af60421
|
1782 |
shmem_recalc_inode(inode); |
4595ef88d
|
1783 |
spin_unlock_irq(&info->lock); |
27ab70062
|
1784 |
goto repeat; |
ff36b8016
|
1785 |
} |
d18992286
|
1786 |
if (error == -EEXIST) /* from above or from radix_tree_insert */ |
54af60421
|
1787 1788 |
goto repeat; return error; |
1da177e4c
|
1789 |
} |
10d20bd25
|
1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 |
/* * This is like autoremove_wake_function, but it removes the wait queue * entry unconditionally - even if something else had already woken the * target. */ static int synchronous_wake_function(wait_queue_t *wait, unsigned mode, int sync, void *key) { int ret = default_wake_function(wait, mode, sync, key); list_del_init(&wait->task_list); return ret; } |
d0217ac04
|
1801 |
static int shmem_fault(struct vm_area_struct *vma, struct vm_fault *vmf) |
1da177e4c
|
1802 |
{ |
496ad9aa8
|
1803 |
struct inode *inode = file_inode(vma->vm_file); |
9e18eb293
|
1804 |
gfp_t gfp = mapping_gfp_mask(inode->i_mapping); |
657e3038c
|
1805 |
enum sgp_type sgp; |
1da177e4c
|
1806 |
int error; |
68da9f055
|
1807 |
int ret = VM_FAULT_LOCKED; |
1da177e4c
|
1808 |
|
f00cdc6df
|
1809 1810 1811 1812 |
/* * Trinity finds that probing a hole which tmpfs is punching can * prevent the hole-punch from ever completing: which in turn * locks writers out with its hold on i_mutex. So refrain from |
8e205f779
|
1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 |
* faulting pages into the hole while it's being punched. Although * shmem_undo_range() does remove the additions, it may be unable to * keep up, as each new page needs its own unmap_mapping_range() call, * and the i_mmap tree grows ever slower to scan if new vmas are added. * * It does not matter if we sometimes reach this check just before the * hole-punch begins, so that one fault then races with the punch: * we just need to make racing faults a rare case. * * The implementation below would be much simpler if we just used a * standard mutex or completion: but we cannot take i_mutex in fault, * and bloating every shmem inode for this unlikely case would be sad. |
f00cdc6df
|
1825 1826 1827 1828 1829 1830 |
*/ if (unlikely(inode->i_private)) { struct shmem_falloc *shmem_falloc; spin_lock(&inode->i_lock); shmem_falloc = inode->i_private; |
8e205f779
|
1831 1832 1833 1834 1835 |
if (shmem_falloc && shmem_falloc->waitq && vmf->pgoff >= shmem_falloc->start && vmf->pgoff < shmem_falloc->next) { wait_queue_head_t *shmem_falloc_waitq; |
10d20bd25
|
1836 |
DEFINE_WAIT_FUNC(shmem_fault_wait, synchronous_wake_function); |
8e205f779
|
1837 1838 |
ret = VM_FAULT_NOPAGE; |
f00cdc6df
|
1839 1840 |
if ((vmf->flags & FAULT_FLAG_ALLOW_RETRY) && !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT)) { |
8e205f779
|
1841 |
/* It's polite to up mmap_sem if we can */ |
f00cdc6df
|
1842 |
up_read(&vma->vm_mm->mmap_sem); |
8e205f779
|
1843 |
ret = VM_FAULT_RETRY; |
f00cdc6df
|
1844 |
} |
8e205f779
|
1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 |
shmem_falloc_waitq = shmem_falloc->waitq; prepare_to_wait(shmem_falloc_waitq, &shmem_fault_wait, TASK_UNINTERRUPTIBLE); spin_unlock(&inode->i_lock); schedule(); /* * shmem_falloc_waitq points into the shmem_fallocate() * stack of the hole-punching task: shmem_falloc_waitq * is usually invalid by the time we reach here, but * finish_wait() does not dereference it in that case; * though i_lock needed lest racing with wake_up_all(). */ spin_lock(&inode->i_lock); finish_wait(shmem_falloc_waitq, &shmem_fault_wait); spin_unlock(&inode->i_lock); return ret; |
f00cdc6df
|
1863 |
} |
8e205f779
|
1864 |
spin_unlock(&inode->i_lock); |
f00cdc6df
|
1865 |
} |
657e3038c
|
1866 1867 1868 1869 1870 1871 1872 |
sgp = SGP_CACHE; if (vma->vm_flags & VM_HUGEPAGE) sgp = SGP_HUGE; else if (vma->vm_flags & VM_NOHUGEPAGE) sgp = SGP_NOHUGE; error = shmem_getpage_gfp(inode, vmf->pgoff, &vmf->page, sgp, |
9e18eb293
|
1873 |
gfp, vma->vm_mm, &ret); |
d0217ac04
|
1874 1875 |
if (error) return ((error == -ENOMEM) ? VM_FAULT_OOM : VM_FAULT_SIGBUS); |
68da9f055
|
1876 |
return ret; |
1da177e4c
|
1877 |
} |
c01d5b300
|
1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 |
unsigned long shmem_get_unmapped_area(struct file *file, unsigned long uaddr, unsigned long len, unsigned long pgoff, unsigned long flags) { unsigned long (*get_area)(struct file *, unsigned long, unsigned long, unsigned long, unsigned long); unsigned long addr; unsigned long offset; unsigned long inflated_len; unsigned long inflated_addr; unsigned long inflated_offset; if (len > TASK_SIZE) return -ENOMEM; get_area = current->mm->get_unmapped_area; addr = get_area(file, uaddr, len, pgoff, flags); |
e496cf3d7
|
1895 |
if (!IS_ENABLED(CONFIG_TRANSPARENT_HUGE_PAGECACHE)) |
c01d5b300
|
1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 |
return addr; if (IS_ERR_VALUE(addr)) return addr; if (addr & ~PAGE_MASK) return addr; if (addr > TASK_SIZE - len) return addr; if (shmem_huge == SHMEM_HUGE_DENY) return addr; if (len < HPAGE_PMD_SIZE) return addr; if (flags & MAP_FIXED) return addr; /* * Our priority is to support MAP_SHARED mapped hugely; * and support MAP_PRIVATE mapped hugely too, until it is COWed. * But if caller specified an address hint, respect that as before. */ if (uaddr) return addr; if (shmem_huge != SHMEM_HUGE_FORCE) { struct super_block *sb; if (file) { VM_BUG_ON(file->f_op != &shmem_file_operations); sb = file_inode(file)->i_sb; } else { /* * Called directly from mm/mmap.c, or drivers/char/mem.c * for "/dev/zero", to create a shared anonymous object. */ if (IS_ERR(shm_mnt)) return addr; sb = shm_mnt->mnt_sb; } |
3089bf614
|
1933 |
if (SHMEM_SB(sb)->huge == SHMEM_HUGE_NEVER) |
c01d5b300
|
1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 |
return addr; } offset = (pgoff << PAGE_SHIFT) & (HPAGE_PMD_SIZE-1); if (offset && offset + len < 2 * HPAGE_PMD_SIZE) return addr; if ((addr & (HPAGE_PMD_SIZE-1)) == offset) return addr; inflated_len = len + HPAGE_PMD_SIZE - PAGE_SIZE; if (inflated_len > TASK_SIZE) return addr; if (inflated_len < len) return addr; inflated_addr = get_area(NULL, 0, inflated_len, 0, flags); if (IS_ERR_VALUE(inflated_addr)) return addr; if (inflated_addr & ~PAGE_MASK) return addr; inflated_offset = inflated_addr & (HPAGE_PMD_SIZE-1); inflated_addr += offset - inflated_offset; if (inflated_offset > offset) inflated_addr += HPAGE_PMD_SIZE; if (inflated_addr > TASK_SIZE - len) return addr; return inflated_addr; } |
1da177e4c
|
1964 |
#ifdef CONFIG_NUMA |
41ffe5d5c
|
1965 |
static int shmem_set_policy(struct vm_area_struct *vma, struct mempolicy *mpol) |
1da177e4c
|
1966 |
{ |
496ad9aa8
|
1967 |
struct inode *inode = file_inode(vma->vm_file); |
41ffe5d5c
|
1968 |
return mpol_set_shared_policy(&SHMEM_I(inode)->policy, vma, mpol); |
1da177e4c
|
1969 |
} |
d8dc74f21
|
1970 1971 |
static struct mempolicy *shmem_get_policy(struct vm_area_struct *vma, unsigned long addr) |
1da177e4c
|
1972 |
{ |
496ad9aa8
|
1973 |
struct inode *inode = file_inode(vma->vm_file); |
41ffe5d5c
|
1974 |
pgoff_t index; |
1da177e4c
|
1975 |
|
41ffe5d5c
|
1976 1977 |
index = ((addr - vma->vm_start) >> PAGE_SHIFT) + vma->vm_pgoff; return mpol_shared_policy_lookup(&SHMEM_I(inode)->policy, index); |
1da177e4c
|
1978 1979 1980 1981 1982 |
} #endif int shmem_lock(struct file *file, int lock, struct user_struct *user) { |
496ad9aa8
|
1983 |
struct inode *inode = file_inode(file); |
1da177e4c
|
1984 1985 |
struct shmem_inode_info *info = SHMEM_I(inode); int retval = -ENOMEM; |
4595ef88d
|
1986 |
spin_lock_irq(&info->lock); |
1da177e4c
|
1987 1988 1989 1990 |
if (lock && !(info->flags & VM_LOCKED)) { if (!user_shm_lock(inode->i_size, user)) goto out_nomem; info->flags |= VM_LOCKED; |
89e004ea5
|
1991 |
mapping_set_unevictable(file->f_mapping); |
1da177e4c
|
1992 1993 1994 1995 |
} if (!lock && (info->flags & VM_LOCKED) && user) { user_shm_unlock(inode->i_size, user); info->flags &= ~VM_LOCKED; |
89e004ea5
|
1996 |
mapping_clear_unevictable(file->f_mapping); |
1da177e4c
|
1997 1998 |
} retval = 0; |
89e004ea5
|
1999 |
|
1da177e4c
|
2000 |
out_nomem: |
4595ef88d
|
2001 |
spin_unlock_irq(&info->lock); |
1da177e4c
|
2002 2003 |
return retval; } |
9b83a6a85
|
2004 |
static int shmem_mmap(struct file *file, struct vm_area_struct *vma) |
1da177e4c
|
2005 2006 2007 |
{ file_accessed(file); vma->vm_ops = &shmem_vm_ops; |
e496cf3d7
|
2008 |
if (IS_ENABLED(CONFIG_TRANSPARENT_HUGE_PAGECACHE) && |
f3f0e1d21
|
2009 2010 2011 2012 |
((vma->vm_start + ~HPAGE_PMD_MASK) & HPAGE_PMD_MASK) < (vma->vm_end & HPAGE_PMD_MASK)) { khugepaged_enter(vma, vma->vm_flags); } |
1da177e4c
|
2013 2014 |
return 0; } |
454abafe9
|
2015 |
static struct inode *shmem_get_inode(struct super_block *sb, const struct inode *dir, |
09208d150
|
2016 |
umode_t mode, dev_t dev, unsigned long flags) |
1da177e4c
|
2017 2018 2019 2020 |
{ struct inode *inode; struct shmem_inode_info *info; struct shmem_sb_info *sbinfo = SHMEM_SB(sb); |
5b04c6890
|
2021 2022 |
if (shmem_reserve_inode(sb)) return NULL; |
1da177e4c
|
2023 2024 2025 |
inode = new_inode(sb); if (inode) { |
85fe4025c
|
2026 |
inode->i_ino = get_next_ino(); |
454abafe9
|
2027 |
inode_init_owner(inode, dir, mode); |
1da177e4c
|
2028 |
inode->i_blocks = 0; |
078cd8279
|
2029 |
inode->i_atime = inode->i_mtime = inode->i_ctime = current_time(inode); |
91828a405
|
2030 |
inode->i_generation = get_seconds(); |
1da177e4c
|
2031 2032 2033 |
info = SHMEM_I(inode); memset(info, 0, (char *)inode - (char *)info); spin_lock_init(&info->lock); |
40e041a2c
|
2034 |
info->seals = F_SEAL_SEAL; |
0b0a0806b
|
2035 |
info->flags = flags & VM_NORESERVE; |
779750d20
|
2036 |
INIT_LIST_HEAD(&info->shrinklist); |
1da177e4c
|
2037 |
INIT_LIST_HEAD(&info->swaplist); |
38f386574
|
2038 |
simple_xattrs_init(&info->xattrs); |
72c04902d
|
2039 |
cache_no_acl(inode); |
1da177e4c
|
2040 2041 2042 |
switch (mode & S_IFMT) { default: |
39f0247d3
|
2043 |
inode->i_op = &shmem_special_inode_operations; |
1da177e4c
|
2044 2045 2046 |
init_special_inode(inode, mode, dev); break; case S_IFREG: |
14fcc23fd
|
2047 |
inode->i_mapping->a_ops = &shmem_aops; |
1da177e4c
|
2048 2049 |
inode->i_op = &shmem_inode_operations; inode->i_fop = &shmem_file_operations; |
71fe804b6
|
2050 2051 |
mpol_shared_policy_init(&info->policy, shmem_get_sbmpol(sbinfo)); |
1da177e4c
|
2052 2053 |
break; case S_IFDIR: |
d8c76e6f4
|
2054 |
inc_nlink(inode); |
1da177e4c
|
2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 |
/* 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
|
2065 |
mpol_shared_policy_init(&info->policy, NULL); |
1da177e4c
|
2066 2067 |
break; } |
5b04c6890
|
2068 2069 |
} else shmem_free_inode(sb); |
1da177e4c
|
2070 2071 |
return inode; } |
0cd6144aa
|
2072 2073 |
bool shmem_mapping(struct address_space *mapping) { |
f0774d884
|
2074 2075 |
if (!mapping->host) return false; |
97b713ba3
|
2076 |
return mapping->host->i_sb->s_op == &shmem_ops; |
0cd6144aa
|
2077 |
} |
1da177e4c
|
2078 |
#ifdef CONFIG_TMPFS |
92e1d5be9
|
2079 |
static const struct inode_operations shmem_symlink_inode_operations; |
69f07ec93
|
2080 |
static const struct inode_operations shmem_short_symlink_operations; |
1da177e4c
|
2081 |
|
6d9d88d07
|
2082 2083 2084 2085 2086 |
#ifdef CONFIG_TMPFS_XATTR static int shmem_initxattrs(struct inode *, const struct xattr *, void *); #else #define shmem_initxattrs NULL #endif |
1da177e4c
|
2087 |
static int |
800d15a53
|
2088 2089 2090 |
shmem_write_begin(struct file *file, struct address_space *mapping, loff_t pos, unsigned len, unsigned flags, struct page **pagep, void **fsdata) |
1da177e4c
|
2091 |
{ |
800d15a53
|
2092 |
struct inode *inode = mapping->host; |
40e041a2c
|
2093 |
struct shmem_inode_info *info = SHMEM_I(inode); |
09cbfeaf1
|
2094 |
pgoff_t index = pos >> PAGE_SHIFT; |
40e041a2c
|
2095 2096 2097 2098 2099 2100 2101 2102 |
/* i_mutex is held by caller */ if (unlikely(info->seals)) { if (info->seals & F_SEAL_WRITE) return -EPERM; if ((info->seals & F_SEAL_GROW) && pos + len > inode->i_size) return -EPERM; } |
9e18eb293
|
2103 |
return shmem_getpage(inode, index, pagep, SGP_WRITE); |
800d15a53
|
2104 2105 2106 2107 2108 2109 2110 2111 |
} 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
|
2112 2113 |
if (pos + copied > inode->i_size) i_size_write(inode, pos + copied); |
ec9516fbc
|
2114 |
if (!PageUptodate(page)) { |
800d8c63b
|
2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 |
struct page *head = compound_head(page); if (PageTransCompound(page)) { int i; for (i = 0; i < HPAGE_PMD_NR; i++) { if (head + i == page) continue; clear_highpage(head + i); flush_dcache_page(head + i); } } |
09cbfeaf1
|
2126 2127 |
if (copied < PAGE_SIZE) { unsigned from = pos & (PAGE_SIZE - 1); |
ec9516fbc
|
2128 |
zero_user_segments(page, 0, from, |
09cbfeaf1
|
2129 |
from + copied, PAGE_SIZE); |
ec9516fbc
|
2130 |
} |
800d8c63b
|
2131 |
SetPageUptodate(head); |
ec9516fbc
|
2132 |
} |
800d15a53
|
2133 |
set_page_dirty(page); |
6746aff74
|
2134 |
unlock_page(page); |
09cbfeaf1
|
2135 |
put_page(page); |
800d15a53
|
2136 |
|
800d15a53
|
2137 |
return copied; |
1da177e4c
|
2138 |
} |
2ba5bbed0
|
2139 |
static ssize_t shmem_file_read_iter(struct kiocb *iocb, struct iov_iter *to) |
1da177e4c
|
2140 |
{ |
6e58e79db
|
2141 2142 |
struct file *file = iocb->ki_filp; struct inode *inode = file_inode(file); |
1da177e4c
|
2143 |
struct address_space *mapping = inode->i_mapping; |
41ffe5d5c
|
2144 2145 |
pgoff_t index; unsigned long offset; |
a0ee5ec52
|
2146 |
enum sgp_type sgp = SGP_READ; |
f7c1d0742
|
2147 |
int error = 0; |
cb66a7a1f
|
2148 |
ssize_t retval = 0; |
6e58e79db
|
2149 |
loff_t *ppos = &iocb->ki_pos; |
a0ee5ec52
|
2150 2151 2152 2153 2154 2155 |
/* * 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. */ |
777eda2c5
|
2156 |
if (!iter_is_iovec(to)) |
75edd345e
|
2157 |
sgp = SGP_CACHE; |
1da177e4c
|
2158 |
|
09cbfeaf1
|
2159 2160 |
index = *ppos >> PAGE_SHIFT; offset = *ppos & ~PAGE_MASK; |
1da177e4c
|
2161 2162 2163 |
for (;;) { struct page *page = NULL; |
41ffe5d5c
|
2164 2165 |
pgoff_t end_index; unsigned long nr, ret; |
1da177e4c
|
2166 |
loff_t i_size = i_size_read(inode); |
09cbfeaf1
|
2167 |
end_index = i_size >> PAGE_SHIFT; |
1da177e4c
|
2168 2169 2170 |
if (index > end_index) break; if (index == end_index) { |
09cbfeaf1
|
2171 |
nr = i_size & ~PAGE_MASK; |
1da177e4c
|
2172 2173 2174 |
if (nr <= offset) break; } |
9e18eb293
|
2175 |
error = shmem_getpage(inode, index, &page, sgp); |
6e58e79db
|
2176 2177 2178 |
if (error) { if (error == -EINVAL) error = 0; |
1da177e4c
|
2179 2180 |
break; } |
75edd345e
|
2181 2182 2183 |
if (page) { if (sgp == SGP_CACHE) set_page_dirty(page); |
d3602444e
|
2184 |
unlock_page(page); |
75edd345e
|
2185 |
} |
1da177e4c
|
2186 2187 2188 |
/* * We must evaluate after, since reads (unlike writes) |
1b1dcc1b5
|
2189 |
* are called without i_mutex protection against truncate |
1da177e4c
|
2190 |
*/ |
09cbfeaf1
|
2191 |
nr = PAGE_SIZE; |
1da177e4c
|
2192 |
i_size = i_size_read(inode); |
09cbfeaf1
|
2193 |
end_index = i_size >> PAGE_SHIFT; |
1da177e4c
|
2194 |
if (index == end_index) { |
09cbfeaf1
|
2195 |
nr = i_size & ~PAGE_MASK; |
1da177e4c
|
2196 2197 |
if (nr <= offset) { if (page) |
09cbfeaf1
|
2198 |
put_page(page); |
1da177e4c
|
2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 |
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
|
2217 |
} else { |
1da177e4c
|
2218 |
page = ZERO_PAGE(0); |
09cbfeaf1
|
2219 |
get_page(page); |
b5810039a
|
2220 |
} |
1da177e4c
|
2221 2222 2223 2224 |
/* * Ok, we have the page, and it's up-to-date, so * now we can copy it to user space... |
1da177e4c
|
2225 |
*/ |
2ba5bbed0
|
2226 |
ret = copy_page_to_iter(page, offset, nr, to); |
6e58e79db
|
2227 |
retval += ret; |
1da177e4c
|
2228 |
offset += ret; |
09cbfeaf1
|
2229 2230 |
index += offset >> PAGE_SHIFT; offset &= ~PAGE_MASK; |
1da177e4c
|
2231 |
|
09cbfeaf1
|
2232 |
put_page(page); |
2ba5bbed0
|
2233 |
if (!iov_iter_count(to)) |
1da177e4c
|
2234 |
break; |
6e58e79db
|
2235 2236 2237 2238 |
if (ret < nr) { error = -EFAULT; break; } |
1da177e4c
|
2239 2240 |
cond_resched(); } |
09cbfeaf1
|
2241 |
*ppos = ((loff_t) index << PAGE_SHIFT) + offset; |
6e58e79db
|
2242 2243 |
file_accessed(file); return retval ? retval : error; |
1da177e4c
|
2244 |
} |
220f2ac91
|
2245 2246 2247 2248 |
/* * llseek SEEK_DATA or SEEK_HOLE through the radix_tree. */ static pgoff_t shmem_seek_hole_data(struct address_space *mapping, |
965c8e59c
|
2249 |
pgoff_t index, pgoff_t end, int whence) |
220f2ac91
|
2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 |
{ struct page *page; struct pagevec pvec; pgoff_t indices[PAGEVEC_SIZE]; bool done = false; int i; pagevec_init(&pvec, 0); pvec.nr = 1; /* start small: we may be there already */ while (!done) { |
0cd6144aa
|
2260 |
pvec.nr = find_get_entries(mapping, index, |
220f2ac91
|
2261 2262 |
pvec.nr, pvec.pages, indices); if (!pvec.nr) { |
965c8e59c
|
2263 |
if (whence == SEEK_DATA) |
220f2ac91
|
2264 2265 2266 2267 2268 |
index = end; break; } for (i = 0; i < pvec.nr; i++, index++) { if (index < indices[i]) { |
965c8e59c
|
2269 |
if (whence == SEEK_HOLE) { |
220f2ac91
|
2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 |
done = true; break; } index = indices[i]; } page = pvec.pages[i]; if (page && !radix_tree_exceptional_entry(page)) { if (!PageUptodate(page)) page = NULL; } if (index >= end || |
965c8e59c
|
2281 2282 |
(page && whence == SEEK_DATA) || (!page && whence == SEEK_HOLE)) { |
220f2ac91
|
2283 2284 2285 2286 |
done = true; break; } } |
0cd6144aa
|
2287 |
pagevec_remove_exceptionals(&pvec); |
220f2ac91
|
2288 2289 2290 2291 2292 2293 |
pagevec_release(&pvec); pvec.nr = PAGEVEC_SIZE; cond_resched(); } return index; } |
965c8e59c
|
2294 |
static loff_t shmem_file_llseek(struct file *file, loff_t offset, int whence) |
220f2ac91
|
2295 2296 2297 2298 2299 |
{ struct address_space *mapping = file->f_mapping; struct inode *inode = mapping->host; pgoff_t start, end; loff_t new_offset; |
965c8e59c
|
2300 2301 |
if (whence != SEEK_DATA && whence != SEEK_HOLE) return generic_file_llseek_size(file, offset, whence, |
220f2ac91
|
2302 |
MAX_LFS_FILESIZE, i_size_read(inode)); |
5955102c9
|
2303 |
inode_lock(inode); |
220f2ac91
|
2304 2305 2306 2307 2308 2309 2310 |
/* We're holding i_mutex so we can access i_size directly */ if (offset < 0) offset = -EINVAL; else if (offset >= inode->i_size) offset = -ENXIO; else { |
09cbfeaf1
|
2311 2312 |
start = offset >> PAGE_SHIFT; end = (inode->i_size + PAGE_SIZE - 1) >> PAGE_SHIFT; |
965c8e59c
|
2313 |
new_offset = shmem_seek_hole_data(mapping, start, end, whence); |
09cbfeaf1
|
2314 |
new_offset <<= PAGE_SHIFT; |
220f2ac91
|
2315 2316 2317 |
if (new_offset > offset) { if (new_offset < inode->i_size) offset = new_offset; |
965c8e59c
|
2318 |
else if (whence == SEEK_DATA) |
220f2ac91
|
2319 2320 2321 2322 2323 |
offset = -ENXIO; else offset = inode->i_size; } } |
387aae6fd
|
2324 2325 |
if (offset >= 0) offset = vfs_setpos(file, offset, MAX_LFS_FILESIZE); |
5955102c9
|
2326 |
inode_unlock(inode); |
220f2ac91
|
2327 2328 |
return offset; } |
05f65b5c7
|
2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 |
/* * We need a tag: a new tag would expand every radix_tree_node by 8 bytes, * so reuse a tag which we firmly believe is never set or cleared on shmem. */ #define SHMEM_TAG_PINNED PAGECACHE_TAG_TOWRITE #define LAST_SCAN 4 /* about 150ms max */ static void shmem_tag_pins(struct address_space *mapping) { struct radix_tree_iter iter; void **slot; pgoff_t start; struct page *page; lru_add_drain(); start = 0; rcu_read_lock(); |
05f65b5c7
|
2346 2347 2348 |
radix_tree_for_each_slot(slot, &mapping->page_tree, &iter, start) { page = radix_tree_deref_slot(slot); if (!page || radix_tree_exception(page)) { |
2cf938aae
|
2349 2350 2351 2352 |
if (radix_tree_deref_retry(page)) { slot = radix_tree_iter_retry(&iter); continue; } |
05f65b5c7
|
2353 2354 2355 2356 2357 2358 2359 2360 2361 |
} else if (page_count(page) - page_mapcount(page) > 1) { spin_lock_irq(&mapping->tree_lock); radix_tree_tag_set(&mapping->page_tree, iter.index, SHMEM_TAG_PINNED); spin_unlock_irq(&mapping->tree_lock); } if (need_resched()) { cond_resched_rcu(); |
7165092fe
|
2362 |
slot = radix_tree_iter_next(&iter); |
05f65b5c7
|
2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 |
} } rcu_read_unlock(); } /* * Setting SEAL_WRITE requires us to verify there's no pending writer. However, * via get_user_pages(), drivers might have some pending I/O without any active * user-space mappings (eg., direct-IO, AIO). Therefore, we look at all pages * and see whether it has an elevated ref-count. If so, we tag them and wait for * them to be dropped. * The caller must guarantee that no new user will acquire writable references * to those pages to avoid races. */ |
40e041a2c
|
2377 2378 |
static int shmem_wait_for_pins(struct address_space *mapping) { |
05f65b5c7
|
2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 |
struct radix_tree_iter iter; void **slot; pgoff_t start; struct page *page; int error, scan; shmem_tag_pins(mapping); error = 0; for (scan = 0; scan <= LAST_SCAN; scan++) { if (!radix_tree_tagged(&mapping->page_tree, SHMEM_TAG_PINNED)) break; if (!scan) lru_add_drain_all(); else if (schedule_timeout_killable((HZ << scan) / 200)) scan = LAST_SCAN; start = 0; rcu_read_lock(); |
05f65b5c7
|
2399 2400 2401 2402 2403 |
radix_tree_for_each_tagged(slot, &mapping->page_tree, &iter, start, SHMEM_TAG_PINNED) { page = radix_tree_deref_slot(slot); if (radix_tree_exception(page)) { |
2cf938aae
|
2404 2405 2406 2407 |
if (radix_tree_deref_retry(page)) { slot = radix_tree_iter_retry(&iter); continue; } |
05f65b5c7
|
2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 |
page = NULL; } if (page && page_count(page) - page_mapcount(page) != 1) { if (scan < LAST_SCAN) goto continue_resched; /* * On the last scan, we clean up all those tags * we inserted; but make a note that we still * found pages pinned. */ error = -EBUSY; } spin_lock_irq(&mapping->tree_lock); radix_tree_tag_clear(&mapping->page_tree, iter.index, SHMEM_TAG_PINNED); spin_unlock_irq(&mapping->tree_lock); continue_resched: if (need_resched()) { cond_resched_rcu(); |
7165092fe
|
2432 |
slot = radix_tree_iter_next(&iter); |
05f65b5c7
|
2433 2434 2435 2436 2437 2438 |
} } rcu_read_unlock(); } return error; |
40e041a2c
|
2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 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 |
} #define F_ALL_SEALS (F_SEAL_SEAL | \ F_SEAL_SHRINK | \ F_SEAL_GROW | \ F_SEAL_WRITE) int shmem_add_seals(struct file *file, unsigned int seals) { struct inode *inode = file_inode(file); struct shmem_inode_info *info = SHMEM_I(inode); int error; /* * SEALING * Sealing allows multiple parties to share a shmem-file but restrict * access to a specific subset of file operations. Seals can only be * added, but never removed. This way, mutually untrusted parties can * share common memory regions with a well-defined policy. A malicious * peer can thus never perform unwanted operations on a shared object. * * Seals are only supported on special shmem-files and always affect * the whole underlying inode. Once a seal is set, it may prevent some * kinds of access to the file. Currently, the following seals are * defined: * SEAL_SEAL: Prevent further seals from being set on this file * SEAL_SHRINK: Prevent the file from shrinking * SEAL_GROW: Prevent the file from growing * SEAL_WRITE: Prevent write access to the file * * As we don't require any trust relationship between two parties, we * must prevent seals from being removed. Therefore, sealing a file * only adds a given set of seals to the file, it never touches * existing seals. Furthermore, the "setting seals"-operation can be * sealed itself, which basically prevents any further seal from being * added. * * Semantics of sealing are only defined on volatile files. Only * anonymous shmem files support sealing. More importantly, seals are * never written to disk. Therefore, there's no plan to support it on * other file types. */ if (file->f_op != &shmem_file_operations) return -EINVAL; if (!(file->f_mode & FMODE_WRITE)) return -EPERM; if (seals & ~(unsigned int)F_ALL_SEALS) return -EINVAL; |
5955102c9
|
2488 |
inode_lock(inode); |
40e041a2c
|
2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 |
if (info->seals & F_SEAL_SEAL) { error = -EPERM; goto unlock; } if ((seals & F_SEAL_WRITE) && !(info->seals & F_SEAL_WRITE)) { error = mapping_deny_writable(file->f_mapping); if (error) goto unlock; error = shmem_wait_for_pins(file->f_mapping); if (error) { mapping_allow_writable(file->f_mapping); goto unlock; } } info->seals |= seals; error = 0; unlock: |
5955102c9
|
2511 |
inode_unlock(inode); |
40e041a2c
|
2512 2513 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 2545 2546 |
return error; } EXPORT_SYMBOL_GPL(shmem_add_seals); int shmem_get_seals(struct file *file) { if (file->f_op != &shmem_file_operations) return -EINVAL; return SHMEM_I(file_inode(file))->seals; } EXPORT_SYMBOL_GPL(shmem_get_seals); long shmem_fcntl(struct file *file, unsigned int cmd, unsigned long arg) { long error; switch (cmd) { case F_ADD_SEALS: /* disallow upper 32bit */ if (arg > UINT_MAX) return -EINVAL; error = shmem_add_seals(file, arg); break; case F_GET_SEALS: error = shmem_get_seals(file); break; default: error = -EINVAL; break; } return error; } |
83e4fa9c1
|
2547 2548 2549 |
static long shmem_fallocate(struct file *file, int mode, loff_t offset, loff_t len) { |
496ad9aa8
|
2550 |
struct inode *inode = file_inode(file); |
e2d12e22c
|
2551 |
struct shmem_sb_info *sbinfo = SHMEM_SB(inode->i_sb); |
40e041a2c
|
2552 |
struct shmem_inode_info *info = SHMEM_I(inode); |
1aac14003
|
2553 |
struct shmem_falloc shmem_falloc; |
e2d12e22c
|
2554 2555 |
pgoff_t start, index, end; int error; |
83e4fa9c1
|
2556 |
|
13ace4d0d
|
2557 2558 |
if (mode & ~(FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE)) return -EOPNOTSUPP; |
5955102c9
|
2559 |
inode_lock(inode); |
83e4fa9c1
|
2560 2561 2562 2563 2564 |
if (mode & FALLOC_FL_PUNCH_HOLE) { struct address_space *mapping = file->f_mapping; loff_t unmap_start = round_up(offset, PAGE_SIZE); loff_t unmap_end = round_down(offset + len, PAGE_SIZE) - 1; |
8e205f779
|
2565 |
DECLARE_WAIT_QUEUE_HEAD_ONSTACK(shmem_falloc_waitq); |
83e4fa9c1
|
2566 |
|
40e041a2c
|
2567 2568 2569 2570 2571 |
/* protected by i_mutex */ if (info->seals & F_SEAL_WRITE) { error = -EPERM; goto out; } |
8e205f779
|
2572 |
shmem_falloc.waitq = &shmem_falloc_waitq; |
f00cdc6df
|
2573 2574 2575 2576 2577 |
shmem_falloc.start = unmap_start >> PAGE_SHIFT; shmem_falloc.next = (unmap_end + 1) >> PAGE_SHIFT; spin_lock(&inode->i_lock); inode->i_private = &shmem_falloc; spin_unlock(&inode->i_lock); |
83e4fa9c1
|
2578 2579 2580 2581 2582 |
if ((u64)unmap_end > (u64)unmap_start) unmap_mapping_range(mapping, unmap_start, 1 + unmap_end - unmap_start, 0); shmem_truncate_range(inode, offset, offset + len - 1); /* No need to unmap again: hole-punching leaves COWed pages */ |
8e205f779
|
2583 2584 2585 2586 |
spin_lock(&inode->i_lock); inode->i_private = NULL; wake_up_all(&shmem_falloc_waitq); |
10d20bd25
|
2587 |
WARN_ON_ONCE(!list_empty(&shmem_falloc_waitq.task_list)); |
8e205f779
|
2588 |
spin_unlock(&inode->i_lock); |
83e4fa9c1
|
2589 |
error = 0; |
8e205f779
|
2590 |
goto out; |
e2d12e22c
|
2591 2592 2593 2594 2595 2596 |
} /* We need to check rlimit even when FALLOC_FL_KEEP_SIZE */ error = inode_newsize_ok(inode, offset + len); if (error) goto out; |
40e041a2c
|
2597 2598 2599 2600 |
if ((info->seals & F_SEAL_GROW) && offset + len > inode->i_size) { error = -EPERM; goto out; } |
09cbfeaf1
|
2601 2602 |
start = offset >> PAGE_SHIFT; end = (offset + len + PAGE_SIZE - 1) >> PAGE_SHIFT; |
e2d12e22c
|
2603 2604 2605 2606 |
/* Try to avoid a swapstorm if len is impossible to satisfy */ if (sbinfo->max_blocks && end - start > sbinfo->max_blocks) { error = -ENOSPC; goto out; |
83e4fa9c1
|
2607 |
} |
8e205f779
|
2608 |
shmem_falloc.waitq = NULL; |
1aac14003
|
2609 2610 2611 2612 2613 2614 2615 |
shmem_falloc.start = start; shmem_falloc.next = start; shmem_falloc.nr_falloced = 0; shmem_falloc.nr_unswapped = 0; spin_lock(&inode->i_lock); inode->i_private = &shmem_falloc; spin_unlock(&inode->i_lock); |
e2d12e22c
|
2616 2617 2618 2619 2620 2621 2622 2623 2624 |
for (index = start; index < end; index++) { struct page *page; /* * Good, the fallocate(2) manpage permits EINTR: we may have * been interrupted because we are using up too much memory. */ if (signal_pending(current)) error = -EINTR; |
1aac14003
|
2625 2626 |
else if (shmem_falloc.nr_unswapped > shmem_falloc.nr_falloced) error = -ENOMEM; |
e2d12e22c
|
2627 |
else |
9e18eb293
|
2628 |
error = shmem_getpage(inode, index, &page, SGP_FALLOC); |
e2d12e22c
|
2629 |
if (error) { |
1635f6a74
|
2630 |
/* Remove the !PageUptodate pages we added */ |
7f5565670
|
2631 2632 2633 2634 2635 |
if (index > start) { shmem_undo_range(inode, (loff_t)start << PAGE_SHIFT, ((loff_t)index << PAGE_SHIFT) - 1, true); } |
1aac14003
|
2636 |
goto undone; |
e2d12e22c
|
2637 |
} |
e2d12e22c
|
2638 |
/* |
1aac14003
|
2639 2640 2641 2642 2643 2644 2645 2646 |
* Inform shmem_writepage() how far we have reached. * No need for lock or barrier: we have the page lock. */ shmem_falloc.next++; if (!PageUptodate(page)) shmem_falloc.nr_falloced++; /* |
1635f6a74
|
2647 2648 2649 |
* If !PageUptodate, leave it that way so that freeable pages * can be recognized if we need to rollback on error later. * But set_page_dirty so that memory pressure will swap rather |
e2d12e22c
|
2650 2651 2652 2653 2654 |
* than free the pages we are allocating (and SGP_CACHE pages * might still be clean: we now need to mark those dirty too). */ set_page_dirty(page); unlock_page(page); |
09cbfeaf1
|
2655 |
put_page(page); |
e2d12e22c
|
2656 2657 2658 2659 2660 |
cond_resched(); } if (!(mode & FALLOC_FL_KEEP_SIZE) && offset + len > inode->i_size) i_size_write(inode, offset + len); |
078cd8279
|
2661 |
inode->i_ctime = current_time(inode); |
1aac14003
|
2662 2663 2664 2665 |
undone: spin_lock(&inode->i_lock); inode->i_private = NULL; spin_unlock(&inode->i_lock); |
e2d12e22c
|
2666 |
out: |
5955102c9
|
2667 |
inode_unlock(inode); |
83e4fa9c1
|
2668 2669 |
return error; } |
726c33422
|
2670 |
static int shmem_statfs(struct dentry *dentry, struct kstatfs *buf) |
1da177e4c
|
2671 |
{ |
726c33422
|
2672 |
struct shmem_sb_info *sbinfo = SHMEM_SB(dentry->d_sb); |
1da177e4c
|
2673 2674 |
buf->f_type = TMPFS_MAGIC; |
09cbfeaf1
|
2675 |
buf->f_bsize = PAGE_SIZE; |
1da177e4c
|
2676 |
buf->f_namelen = NAME_MAX; |
0edd73b33
|
2677 |
if (sbinfo->max_blocks) { |
1da177e4c
|
2678 |
buf->f_blocks = sbinfo->max_blocks; |
41ffe5d5c
|
2679 2680 2681 |
buf->f_bavail = buf->f_bfree = sbinfo->max_blocks - percpu_counter_sum(&sbinfo->used_blocks); |
0edd73b33
|
2682 2683 |
} if (sbinfo->max_inodes) { |
1da177e4c
|
2684 2685 |
buf->f_files = sbinfo->max_inodes; buf->f_ffree = sbinfo->free_inodes; |
1da177e4c
|
2686 2687 2688 2689 2690 2691 2692 2693 2694 |
} /* else leave those fields 0 like simple_statfs */ return 0; } /* * File creation. Allocate an inode, and we're done.. */ static int |
1a67aafb5
|
2695 |
shmem_mknod(struct inode *dir, struct dentry *dentry, umode_t mode, dev_t dev) |
1da177e4c
|
2696 |
{ |
0b0a0806b
|
2697 |
struct inode *inode; |
1da177e4c
|
2698 |
int error = -ENOSPC; |
454abafe9
|
2699 |
inode = shmem_get_inode(dir->i_sb, dir, mode, dev, VM_NORESERVE); |
1da177e4c
|
2700 |
if (inode) { |
feda821e7
|
2701 2702 2703 |
error = simple_acl_create(dir, inode); if (error) goto out_iput; |
2a7dba391
|
2704 |
error = security_inode_init_security(inode, dir, |
9d8f13ba3
|
2705 |
&dentry->d_name, |
6d9d88d07
|
2706 |
shmem_initxattrs, NULL); |
feda821e7
|
2707 2708 |
if (error && error != -EOPNOTSUPP) goto out_iput; |
37ec43cdc
|
2709 |
|
718deb6b6
|
2710 |
error = 0; |
1da177e4c
|
2711 |
dir->i_size += BOGO_DIRENT_SIZE; |
078cd8279
|
2712 |
dir->i_ctime = dir->i_mtime = current_time(dir); |
1da177e4c
|
2713 2714 |
d_instantiate(dentry, inode); dget(dentry); /* Extra count - pin the dentry in core */ |
1da177e4c
|
2715 2716 |
} return error; |
feda821e7
|
2717 2718 2719 |
out_iput: iput(inode); return error; |
1da177e4c
|
2720 |
} |
60545d0d4
|
2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 |
static int shmem_tmpfile(struct inode *dir, struct dentry *dentry, umode_t mode) { struct inode *inode; int error = -ENOSPC; inode = shmem_get_inode(dir->i_sb, dir, mode, 0, VM_NORESERVE); if (inode) { error = security_inode_init_security(inode, dir, NULL, shmem_initxattrs, NULL); |
feda821e7
|
2732 2733 2734 2735 2736 |
if (error && error != -EOPNOTSUPP) goto out_iput; error = simple_acl_create(dir, inode); if (error) goto out_iput; |
60545d0d4
|
2737 2738 2739 |
d_tmpfile(dentry, inode); } return error; |
feda821e7
|
2740 2741 2742 |
out_iput: iput(inode); return error; |
60545d0d4
|
2743 |
} |
18bb1db3e
|
2744 |
static int shmem_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode) |
1da177e4c
|
2745 2746 2747 2748 2749 |
{ int error; if ((error = shmem_mknod(dir, dentry, mode | S_IFDIR, 0))) return error; |
d8c76e6f4
|
2750 |
inc_nlink(dir); |
1da177e4c
|
2751 2752 |
return 0; } |
4acdaf27e
|
2753 |
static int shmem_create(struct inode *dir, struct dentry *dentry, umode_t mode, |
ebfc3b49a
|
2754 |
bool excl) |
1da177e4c
|
2755 2756 2757 2758 2759 2760 2761 2762 2763 |
{ 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) { |
75c3cfa85
|
2764 |
struct inode *inode = d_inode(old_dentry); |
5b04c6890
|
2765 |
int ret; |
1da177e4c
|
2766 2767 2768 2769 2770 2771 |
/* * 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
|
2772 2773 2774 |
ret = shmem_reserve_inode(inode->i_sb); if (ret) goto out; |
1da177e4c
|
2775 2776 |
dir->i_size += BOGO_DIRENT_SIZE; |
078cd8279
|
2777 |
inode->i_ctime = dir->i_ctime = dir->i_mtime = current_time(inode); |
d8c76e6f4
|
2778 |
inc_nlink(inode); |
7de9c6ee3
|
2779 |
ihold(inode); /* New dentry reference */ |
1da177e4c
|
2780 2781 |
dget(dentry); /* Extra pinning count for the created dentry */ d_instantiate(dentry, inode); |
5b04c6890
|
2782 2783 |
out: return ret; |
1da177e4c
|
2784 2785 2786 2787 |
} static int shmem_unlink(struct inode *dir, struct dentry *dentry) { |
75c3cfa85
|
2788 |
struct inode *inode = d_inode(dentry); |
1da177e4c
|
2789 |
|
5b04c6890
|
2790 2791 |
if (inode->i_nlink > 1 && !S_ISDIR(inode->i_mode)) shmem_free_inode(inode->i_sb); |
1da177e4c
|
2792 2793 |
dir->i_size -= BOGO_DIRENT_SIZE; |
078cd8279
|
2794 |
inode->i_ctime = dir->i_ctime = dir->i_mtime = current_time(inode); |
9a53c3a78
|
2795 |
drop_nlink(inode); |
1da177e4c
|
2796 2797 2798 2799 2800 2801 2802 2803 |
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; |
75c3cfa85
|
2804 |
drop_nlink(d_inode(dentry)); |
9a53c3a78
|
2805 |
drop_nlink(dir); |
1da177e4c
|
2806 2807 |
return shmem_unlink(dir, dentry); } |
37456771c
|
2808 2809 |
static int shmem_exchange(struct inode *old_dir, struct dentry *old_dentry, struct inode *new_dir, struct dentry *new_dentry) { |
e36cb0b89
|
2810 2811 |
bool old_is_dir = d_is_dir(old_dentry); bool new_is_dir = d_is_dir(new_dentry); |
37456771c
|
2812 2813 2814 2815 2816 2817 2818 2819 2820 2821 2822 2823 |
if (old_dir != new_dir && old_is_dir != new_is_dir) { if (old_is_dir) { drop_nlink(old_dir); inc_nlink(new_dir); } else { drop_nlink(new_dir); inc_nlink(old_dir); } } old_dir->i_ctime = old_dir->i_mtime = new_dir->i_ctime = new_dir->i_mtime = |
75c3cfa85
|
2824 |
d_inode(old_dentry)->i_ctime = |
078cd8279
|
2825 |
d_inode(new_dentry)->i_ctime = current_time(old_dir); |
37456771c
|
2826 2827 2828 |
return 0; } |
46fdb794e
|
2829 2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846 2847 2848 2849 2850 2851 2852 2853 |
static int shmem_whiteout(struct inode *old_dir, struct dentry *old_dentry) { struct dentry *whiteout; int error; whiteout = d_alloc(old_dentry->d_parent, &old_dentry->d_name); if (!whiteout) return -ENOMEM; error = shmem_mknod(old_dir, whiteout, S_IFCHR | WHITEOUT_MODE, WHITEOUT_DEV); dput(whiteout); if (error) return error; /* * Cheat and hash the whiteout while the old dentry is still in * place, instead of playing games with FS_RENAME_DOES_D_MOVE. * * d_lookup() will consistently find one of them at this point, * not sure which one, but that isn't even important. */ d_rehash(whiteout); return 0; } |
1da177e4c
|
2854 2855 2856 2857 2858 2859 |
/* * 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. */ |
3b69ff51d
|
2860 |
static int shmem_rename2(struct inode *old_dir, struct dentry *old_dentry, struct inode *new_dir, struct dentry *new_dentry, unsigned int flags) |
1da177e4c
|
2861 |
{ |
75c3cfa85
|
2862 |
struct inode *inode = d_inode(old_dentry); |
1da177e4c
|
2863 |
int they_are_dirs = S_ISDIR(inode->i_mode); |
46fdb794e
|
2864 |
if (flags & ~(RENAME_NOREPLACE | RENAME_EXCHANGE | RENAME_WHITEOUT)) |
3b69ff51d
|
2865 |
return -EINVAL; |
37456771c
|
2866 2867 |
if (flags & RENAME_EXCHANGE) return shmem_exchange(old_dir, old_dentry, new_dir, new_dentry); |
1da177e4c
|
2868 2869 |
if (!simple_empty(new_dentry)) return -ENOTEMPTY; |
46fdb794e
|
2870 2871 2872 2873 2874 2875 2876 |
if (flags & RENAME_WHITEOUT) { int error; error = shmem_whiteout(old_dir, old_dentry); if (error) return error; } |
75c3cfa85
|
2877 |
if (d_really_is_positive(new_dentry)) { |
1da177e4c
|
2878 |
(void) shmem_unlink(new_dir, new_dentry); |
b928095b0
|
2879 |
if (they_are_dirs) { |
75c3cfa85
|
2880 |
drop_nlink(d_inode(new_dentry)); |
9a53c3a78
|
2881 |
drop_nlink(old_dir); |
b928095b0
|
2882 |
} |
1da177e4c
|
2883 |
} else if (they_are_dirs) { |
9a53c3a78
|
2884 |
drop_nlink(old_dir); |
d8c76e6f4
|
2885 |
inc_nlink(new_dir); |
1da177e4c
|
2886 2887 2888 2889 2890 2891 |
} 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 = |
078cd8279
|
2892 |
inode->i_ctime = current_time(old_dir); |
1da177e4c
|
2893 2894 2895 2896 2897 2898 2899 2900 |
return 0; } static int shmem_symlink(struct inode *dir, struct dentry *dentry, const char *symname) { int error; int len; struct inode *inode; |
9276aad6c
|
2901 |
struct page *page; |
1da177e4c
|
2902 2903 2904 |
struct shmem_inode_info *info; len = strlen(symname) + 1; |
09cbfeaf1
|
2905 |
if (len > PAGE_SIZE) |
1da177e4c
|
2906 |
return -ENAMETOOLONG; |
454abafe9
|
2907 |
inode = shmem_get_inode(dir->i_sb, dir, S_IFLNK|S_IRWXUGO, 0, VM_NORESERVE); |
1da177e4c
|
2908 2909 |
if (!inode) return -ENOSPC; |
9d8f13ba3
|
2910 |
error = security_inode_init_security(inode, dir, &dentry->d_name, |
6d9d88d07
|
2911 |
shmem_initxattrs, NULL); |
570bc1c2e
|
2912 2913 2914 2915 2916 2917 2918 |
if (error) { if (error != -EOPNOTSUPP) { iput(inode); return error; } error = 0; } |
1da177e4c
|
2919 2920 |
info = SHMEM_I(inode); inode->i_size = len-1; |
69f07ec93
|
2921 |
if (len <= SHORT_SYMLINK_LEN) { |
3ed47db34
|
2922 2923 |
inode->i_link = kmemdup(symname, len, GFP_KERNEL); if (!inode->i_link) { |
69f07ec93
|
2924 2925 2926 2927 |
iput(inode); return -ENOMEM; } inode->i_op = &shmem_short_symlink_operations; |
1da177e4c
|
2928 |
} else { |
e8ecde25f
|
2929 |
inode_nohighmem(inode); |
9e18eb293
|
2930 |
error = shmem_getpage(inode, 0, &page, SGP_WRITE); |
1da177e4c
|
2931 2932 2933 2934 |
if (error) { iput(inode); return error; } |
14fcc23fd
|
2935 |
inode->i_mapping->a_ops = &shmem_aops; |
1da177e4c
|
2936 |
inode->i_op = &shmem_symlink_inode_operations; |
21fc61c73
|
2937 |
memcpy(page_address(page), symname, len); |
ec9516fbc
|
2938 |
SetPageUptodate(page); |
1da177e4c
|
2939 |
set_page_dirty(page); |
6746aff74
|
2940 |
unlock_page(page); |
09cbfeaf1
|
2941 |
put_page(page); |
1da177e4c
|
2942 |
} |
1da177e4c
|
2943 |
dir->i_size += BOGO_DIRENT_SIZE; |
078cd8279
|
2944 |
dir->i_ctime = dir->i_mtime = current_time(dir); |
1da177e4c
|
2945 2946 2947 2948 |
d_instantiate(dentry, inode); dget(dentry); return 0; } |
fceef393a
|
2949 |
static void shmem_put_link(void *arg) |
1da177e4c
|
2950 |
{ |
fceef393a
|
2951 2952 |
mark_page_accessed(arg); put_page(arg); |
1da177e4c
|
2953 |
} |
6b2553918
|
2954 |
static const char *shmem_get_link(struct dentry *dentry, |
fceef393a
|
2955 2956 |
struct inode *inode, struct delayed_call *done) |
1da177e4c
|
2957 |
{ |
1da177e4c
|
2958 |
struct page *page = NULL; |
6b2553918
|
2959 |
int error; |
6a6c99049
|
2960 2961 2962 2963 2964 2965 2966 2967 2968 |
if (!dentry) { page = find_get_page(inode->i_mapping, 0); if (!page) return ERR_PTR(-ECHILD); if (!PageUptodate(page)) { put_page(page); return ERR_PTR(-ECHILD); } } else { |
9e18eb293
|
2969 |
error = shmem_getpage(inode, 0, &page, SGP_READ); |
6a6c99049
|
2970 2971 2972 2973 |
if (error) return ERR_PTR(error); unlock_page(page); } |
fceef393a
|
2974 |
set_delayed_call(done, shmem_put_link, page); |
21fc61c73
|
2975 |
return page_address(page); |
1da177e4c
|
2976 |
} |
b09e0fa4b
|
2977 |
#ifdef CONFIG_TMPFS_XATTR |
467118102
|
2978 |
/* |
b09e0fa4b
|
2979 2980 |
* 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
|
2981 2982 2983 |
* like ACLs, we also need to implement the security.* handlers at * filesystem level, though. */ |
6d9d88d07
|
2984 |
/* |
6d9d88d07
|
2985 2986 2987 2988 2989 2990 2991 2992 |
* Callback for security_inode_init_security() for acquiring xattrs. */ static int shmem_initxattrs(struct inode *inode, const struct xattr *xattr_array, void *fs_info) { struct shmem_inode_info *info = SHMEM_I(inode); const struct xattr *xattr; |
38f386574
|
2993 |
struct simple_xattr *new_xattr; |
6d9d88d07
|
2994 2995 2996 |
size_t len; for (xattr = xattr_array; xattr->name != NULL; xattr++) { |
38f386574
|
2997 |
new_xattr = simple_xattr_alloc(xattr->value, xattr->value_len); |
6d9d88d07
|
2998 2999 3000 3001 3002 3003 3004 3005 3006 3007 3008 3009 3010 3011 3012 |
if (!new_xattr) return -ENOMEM; len = strlen(xattr->name) + 1; new_xattr->name = kmalloc(XATTR_SECURITY_PREFIX_LEN + len, GFP_KERNEL); if (!new_xattr->name) { kfree(new_xattr); return -ENOMEM; } memcpy(new_xattr->name, XATTR_SECURITY_PREFIX, XATTR_SECURITY_PREFIX_LEN); memcpy(new_xattr->name + XATTR_SECURITY_PREFIX_LEN, xattr->name, len); |
38f386574
|
3013 |
simple_xattr_list_add(&info->xattrs, new_xattr); |
6d9d88d07
|
3014 3015 3016 3017 |
} return 0; } |
aa7c5241c
|
3018 |
static int shmem_xattr_handler_get(const struct xattr_handler *handler, |
b296821a7
|
3019 3020 |
struct dentry *unused, struct inode *inode, const char *name, void *buffer, size_t size) |
b09e0fa4b
|
3021 |
{ |
b296821a7
|
3022 |
struct shmem_inode_info *info = SHMEM_I(inode); |
b09e0fa4b
|
3023 |
|
aa7c5241c
|
3024 |
name = xattr_full_name(handler, name); |
38f386574
|
3025 |
return simple_xattr_get(&info->xattrs, name, buffer, size); |
b09e0fa4b
|
3026 |
} |
aa7c5241c
|
3027 |
static int shmem_xattr_handler_set(const struct xattr_handler *handler, |
593012268
|
3028 3029 3030 |
struct dentry *unused, struct inode *inode, const char *name, const void *value, size_t size, int flags) |
b09e0fa4b
|
3031 |
{ |
593012268
|
3032 |
struct shmem_inode_info *info = SHMEM_I(inode); |
b09e0fa4b
|
3033 |
|
aa7c5241c
|
3034 |
name = xattr_full_name(handler, name); |
38f386574
|
3035 |
return simple_xattr_set(&info->xattrs, name, value, size, flags); |
b09e0fa4b
|
3036 |
} |
aa7c5241c
|
3037 3038 3039 3040 3041 |
static const struct xattr_handler shmem_security_xattr_handler = { .prefix = XATTR_SECURITY_PREFIX, .get = shmem_xattr_handler_get, .set = shmem_xattr_handler_set, }; |
b09e0fa4b
|
3042 |
|
aa7c5241c
|
3043 3044 3045 3046 3047 |
static const struct xattr_handler shmem_trusted_xattr_handler = { .prefix = XATTR_TRUSTED_PREFIX, .get = shmem_xattr_handler_get, .set = shmem_xattr_handler_set, }; |
b09e0fa4b
|
3048 |
|
aa7c5241c
|
3049 3050 3051 3052 3053 3054 3055 3056 3057 |
static const struct xattr_handler *shmem_xattr_handlers[] = { #ifdef CONFIG_TMPFS_POSIX_ACL &posix_acl_access_xattr_handler, &posix_acl_default_xattr_handler, #endif &shmem_security_xattr_handler, &shmem_trusted_xattr_handler, NULL }; |
b09e0fa4b
|
3058 3059 3060 |
static ssize_t shmem_listxattr(struct dentry *dentry, char *buffer, size_t size) { |
75c3cfa85
|
3061 |
struct shmem_inode_info *info = SHMEM_I(d_inode(dentry)); |
786534b92
|
3062 |
return simple_xattr_list(d_inode(dentry), &info->xattrs, buffer, size); |
b09e0fa4b
|
3063 3064 |
} #endif /* CONFIG_TMPFS_XATTR */ |
69f07ec93
|
3065 |
static const struct inode_operations shmem_short_symlink_operations = { |
b09e0fa4b
|
3066 |
.readlink = generic_readlink, |
6b2553918
|
3067 |
.get_link = simple_get_link, |
b09e0fa4b
|
3068 |
#ifdef CONFIG_TMPFS_XATTR |
b09e0fa4b
|
3069 |
.listxattr = shmem_listxattr, |
b09e0fa4b
|
3070 3071 3072 3073 3074 |
#endif }; static const struct inode_operations shmem_symlink_inode_operations = { .readlink = generic_readlink, |
6b2553918
|
3075 |
.get_link = shmem_get_link, |
b09e0fa4b
|
3076 |
#ifdef CONFIG_TMPFS_XATTR |
b09e0fa4b
|
3077 |
.listxattr = shmem_listxattr, |
39f0247d3
|
3078 |
#endif |
b09e0fa4b
|
3079 |
}; |
39f0247d3
|
3080 |
|
91828a405
|
3081 3082 3083 3084 3085 3086 3087 3088 3089 3090 3091 3092 |
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
|
3093 3094 |
static struct dentry *shmem_fh_to_dentry(struct super_block *sb, struct fid *fid, int fh_len, int fh_type) |
91828a405
|
3095 |
{ |
91828a405
|
3096 |
struct inode *inode; |
480b116c9
|
3097 |
struct dentry *dentry = NULL; |
35c2a7f49
|
3098 |
u64 inum; |
480b116c9
|
3099 3100 3101 |
if (fh_len < 3) return NULL; |
91828a405
|
3102 |
|
35c2a7f49
|
3103 3104 |
inum = fid->raw[2]; inum = (inum << 32) | fid->raw[1]; |
480b116c9
|
3105 3106 |
inode = ilookup5(sb, (unsigned long)(inum + fid->raw[0]), shmem_match, fid->raw); |
91828a405
|
3107 |
if (inode) { |
480b116c9
|
3108 |
dentry = d_find_alias(inode); |
91828a405
|
3109 3110 |
iput(inode); } |
480b116c9
|
3111 |
return dentry; |
91828a405
|
3112 |
} |
b0b0382bb
|
3113 3114 |
static int shmem_encode_fh(struct inode *inode, __u32 *fh, int *len, struct inode *parent) |
91828a405
|
3115 |
{ |
5fe0c2378
|
3116 3117 |
if (*len < 3) { *len = 3; |
94e07a759
|
3118 |
return FILEID_INVALID; |
5fe0c2378
|
3119 |
} |
91828a405
|
3120 |
|
1d3382cbf
|
3121 |
if (inode_unhashed(inode)) { |
91828a405
|
3122 3123 3124 3125 3126 3127 3128 |
/* 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
|
3129 |
if (inode_unhashed(inode)) |
91828a405
|
3130 3131 3132 3133 3134 3135 3136 3137 3138 3139 3140 3141 |
__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
|
3142 |
static const struct export_operations shmem_export_ops = { |
91828a405
|
3143 |
.get_parent = shmem_get_parent, |
91828a405
|
3144 |
.encode_fh = shmem_encode_fh, |
480b116c9
|
3145 |
.fh_to_dentry = shmem_fh_to_dentry, |
91828a405
|
3146 |
}; |
680d794ba
|
3147 3148 |
static int shmem_parse_options(char *options, struct shmem_sb_info *sbinfo, bool remount) |
1da177e4c
|
3149 3150 |
{ char *this_char, *value, *rest; |
49cd0a5c2
|
3151 |
struct mempolicy *mpol = NULL; |
8751e0395
|
3152 3153 |
uid_t uid; gid_t gid; |
1da177e4c
|
3154 |
|
b00dc3ad7
|
3155 3156 3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 3167 3168 3169 3170 3171 |
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
|
3172 3173 3174 3175 3176 |
if (!*this_char) continue; if ((value = strchr(this_char,'=')) != NULL) { *value++ = 0; } else { |
1170532bb
|
3177 3178 3179 |
pr_err("tmpfs: No value for mount option '%s' ", this_char); |
49cd0a5c2
|
3180 |
goto error; |
1da177e4c
|
3181 3182 3183 3184 3185 3186 3187 3188 3189 3190 3191 3192 3193 |
} 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
|
3194 |
sbinfo->max_blocks = |
09cbfeaf1
|
3195 |
DIV_ROUND_UP(size, PAGE_SIZE); |
1da177e4c
|
3196 |
} else if (!strcmp(this_char,"nr_blocks")) { |
680d794ba
|
3197 |
sbinfo->max_blocks = memparse(value, &rest); |
1da177e4c
|
3198 3199 3200 |
if (*rest) goto bad_val; } else if (!strcmp(this_char,"nr_inodes")) { |
680d794ba
|
3201 |
sbinfo->max_inodes = memparse(value, &rest); |
1da177e4c
|
3202 3203 3204 |
if (*rest) goto bad_val; } else if (!strcmp(this_char,"mode")) { |
680d794ba
|
3205 |
if (remount) |
1da177e4c
|
3206 |
continue; |
680d794ba
|
3207 |
sbinfo->mode = simple_strtoul(value, &rest, 8) & 07777; |
1da177e4c
|
3208 3209 3210 |
if (*rest) goto bad_val; } else if (!strcmp(this_char,"uid")) { |
680d794ba
|
3211 |
if (remount) |
1da177e4c
|
3212 |
continue; |
8751e0395
|
3213 |
uid = simple_strtoul(value, &rest, 0); |
1da177e4c
|
3214 3215 |
if (*rest) goto bad_val; |
8751e0395
|
3216 3217 3218 |
sbinfo->uid = make_kuid(current_user_ns(), uid); if (!uid_valid(sbinfo->uid)) goto bad_val; |
1da177e4c
|
3219 |
} else if (!strcmp(this_char,"gid")) { |
680d794ba
|
3220 |
if (remount) |
1da177e4c
|
3221 |
continue; |
8751e0395
|
3222 |
gid = simple_strtoul(value, &rest, 0); |
1da177e4c
|
3223 3224 |
if (*rest) goto bad_val; |
8751e0395
|
3225 3226 3227 |
sbinfo->gid = make_kgid(current_user_ns(), gid); if (!gid_valid(sbinfo->gid)) goto bad_val; |
e496cf3d7
|
3228 |
#ifdef CONFIG_TRANSPARENT_HUGE_PAGECACHE |
5a6e75f81
|
3229 3230 3231 3232 3233 3234 3235 3236 3237 3238 3239 |
} else if (!strcmp(this_char, "huge")) { int huge; huge = shmem_parse_huge(value); if (huge < 0) goto bad_val; if (!has_transparent_hugepage() && huge != SHMEM_HUGE_NEVER) goto bad_val; sbinfo->huge = huge; #endif #ifdef CONFIG_NUMA |
7339ff830
|
3240 |
} else if (!strcmp(this_char,"mpol")) { |
49cd0a5c2
|
3241 3242 3243 |
mpol_put(mpol); mpol = NULL; if (mpol_parse_str(value, &mpol)) |
7339ff830
|
3244 |
goto bad_val; |
5a6e75f81
|
3245 |
#endif |
1da177e4c
|
3246 |
} else { |
1170532bb
|
3247 3248 |
pr_err("tmpfs: Bad mount option %s ", this_char); |
49cd0a5c2
|
3249 |
goto error; |
1da177e4c
|
3250 3251 |
} } |
49cd0a5c2
|
3252 |
sbinfo->mpol = mpol; |
1da177e4c
|
3253 3254 3255 |
return 0; bad_val: |
1170532bb
|
3256 3257 |
pr_err("tmpfs: Bad value '%s' for mount option '%s' ", |
1da177e4c
|
3258 |
value, this_char); |
49cd0a5c2
|
3259 3260 |
error: mpol_put(mpol); |
1da177e4c
|
3261 3262 3263 3264 3265 3266 3267 |
return 1; } static int shmem_remount_fs(struct super_block *sb, int *flags, char *data) { struct shmem_sb_info *sbinfo = SHMEM_SB(sb); |
680d794ba
|
3268 |
struct shmem_sb_info config = *sbinfo; |
0edd73b33
|
3269 3270 |
unsigned long inodes; int error = -EINVAL; |
5f00110f7
|
3271 |
config.mpol = NULL; |
680d794ba
|
3272 |
if (shmem_parse_options(data, &config, true)) |
0edd73b33
|
3273 |
return error; |
1da177e4c
|
3274 |
|
0edd73b33
|
3275 |
spin_lock(&sbinfo->stat_lock); |
0edd73b33
|
3276 |
inodes = sbinfo->max_inodes - sbinfo->free_inodes; |
7e496299d
|
3277 |
if (percpu_counter_compare(&sbinfo->used_blocks, config.max_blocks) > 0) |
0edd73b33
|
3278 |
goto out; |
680d794ba
|
3279 |
if (config.max_inodes < inodes) |
0edd73b33
|
3280 3281 |
goto out; /* |
54af60421
|
3282 |
* Those tests disallow limited->unlimited while any are in use; |
0edd73b33
|
3283 3284 3285 |
* but we must separately disallow unlimited->limited, because * in that case we have no record of how much is already in use. */ |
680d794ba
|
3286 |
if (config.max_blocks && !sbinfo->max_blocks) |
0edd73b33
|
3287 |
goto out; |
680d794ba
|
3288 |
if (config.max_inodes && !sbinfo->max_inodes) |
0edd73b33
|
3289 3290 3291 |
goto out; error = 0; |
5a6e75f81
|
3292 |
sbinfo->huge = config.huge; |
680d794ba
|
3293 |
sbinfo->max_blocks = config.max_blocks; |
680d794ba
|
3294 3295 |
sbinfo->max_inodes = config.max_inodes; sbinfo->free_inodes = config.max_inodes - inodes; |
71fe804b6
|
3296 |
|
5f00110f7
|
3297 3298 3299 3300 3301 3302 3303 |
/* * Preserve previous mempolicy unless mpol remount option was specified. */ if (config.mpol) { mpol_put(sbinfo->mpol); sbinfo->mpol = config.mpol; /* transfers initial ref */ } |
0edd73b33
|
3304 3305 3306 |
out: spin_unlock(&sbinfo->stat_lock); return error; |
1da177e4c
|
3307 |
} |
680d794ba
|
3308 |
|
34c80b1d9
|
3309 |
static int shmem_show_options(struct seq_file *seq, struct dentry *root) |
680d794ba
|
3310 |
{ |
34c80b1d9
|
3311 |
struct shmem_sb_info *sbinfo = SHMEM_SB(root->d_sb); |
680d794ba
|
3312 3313 3314 |
if (sbinfo->max_blocks != shmem_default_max_blocks()) seq_printf(seq, ",size=%luk", |
09cbfeaf1
|
3315 |
sbinfo->max_blocks << (PAGE_SHIFT - 10)); |
680d794ba
|
3316 3317 3318 |
if (sbinfo->max_inodes != shmem_default_max_inodes()) seq_printf(seq, ",nr_inodes=%lu", sbinfo->max_inodes); if (sbinfo->mode != (S_IRWXUGO | S_ISVTX)) |
09208d150
|
3319 |
seq_printf(seq, ",mode=%03ho", sbinfo->mode); |
8751e0395
|
3320 3321 3322 3323 3324 3325 |
if (!uid_eq(sbinfo->uid, GLOBAL_ROOT_UID)) seq_printf(seq, ",uid=%u", from_kuid_munged(&init_user_ns, sbinfo->uid)); if (!gid_eq(sbinfo->gid, GLOBAL_ROOT_GID)) seq_printf(seq, ",gid=%u", from_kgid_munged(&init_user_ns, sbinfo->gid)); |
e496cf3d7
|
3326 |
#ifdef CONFIG_TRANSPARENT_HUGE_PAGECACHE |
5a6e75f81
|
3327 3328 3329 3330 |
/* Rightly or wrongly, show huge mount option unmasked by shmem_huge */ if (sbinfo->huge) seq_printf(seq, ",huge=%s", shmem_format_huge(sbinfo->huge)); #endif |
71fe804b6
|
3331 |
shmem_show_mpol(seq, sbinfo->mpol); |
680d794ba
|
3332 3333 |
return 0; } |
9183df25f
|
3334 3335 3336 3337 3338 3339 3340 3341 3342 3343 3344 3345 3346 3347 3348 3349 3350 3351 3352 3353 3354 3355 3356 3357 3358 3359 3360 3361 3362 3363 3364 3365 3366 3367 3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 3378 3379 3380 3381 3382 3383 3384 3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397 3398 3399 3400 3401 3402 3403 |
#define MFD_NAME_PREFIX "memfd:" #define MFD_NAME_PREFIX_LEN (sizeof(MFD_NAME_PREFIX) - 1) #define MFD_NAME_MAX_LEN (NAME_MAX - MFD_NAME_PREFIX_LEN) #define MFD_ALL_FLAGS (MFD_CLOEXEC | MFD_ALLOW_SEALING) SYSCALL_DEFINE2(memfd_create, const char __user *, uname, unsigned int, flags) { struct shmem_inode_info *info; struct file *file; int fd, error; char *name; long len; if (flags & ~(unsigned int)MFD_ALL_FLAGS) return -EINVAL; /* length includes terminating zero */ len = strnlen_user(uname, MFD_NAME_MAX_LEN + 1); if (len <= 0) return -EFAULT; if (len > MFD_NAME_MAX_LEN + 1) return -EINVAL; name = kmalloc(len + MFD_NAME_PREFIX_LEN, GFP_TEMPORARY); if (!name) return -ENOMEM; strcpy(name, MFD_NAME_PREFIX); if (copy_from_user(&name[MFD_NAME_PREFIX_LEN], uname, len)) { error = -EFAULT; goto err_name; } /* terminating-zero may have changed after strnlen_user() returned */ if (name[len + MFD_NAME_PREFIX_LEN - 1]) { error = -EFAULT; goto err_name; } fd = get_unused_fd_flags((flags & MFD_CLOEXEC) ? O_CLOEXEC : 0); if (fd < 0) { error = fd; goto err_name; } file = shmem_file_setup(name, 0, VM_NORESERVE); if (IS_ERR(file)) { error = PTR_ERR(file); goto err_fd; } info = SHMEM_I(file_inode(file)); file->f_mode |= FMODE_LSEEK | FMODE_PREAD | FMODE_PWRITE; file->f_flags |= O_RDWR | O_LARGEFILE; if (flags & MFD_ALLOW_SEALING) info->seals &= ~F_SEAL_SEAL; fd_install(fd, file); kfree(name); return fd; err_fd: put_unused_fd(fd); err_name: kfree(name); return error; } |
680d794ba
|
3404 |
#endif /* CONFIG_TMPFS */ |
1da177e4c
|
3405 3406 3407 |
static void shmem_put_super(struct super_block *sb) { |
602586a83
|
3408 3409 3410 |
struct shmem_sb_info *sbinfo = SHMEM_SB(sb); percpu_counter_destroy(&sbinfo->used_blocks); |
49cd0a5c2
|
3411 |
mpol_put(sbinfo->mpol); |
602586a83
|
3412 |
kfree(sbinfo); |
1da177e4c
|
3413 3414 |
sb->s_fs_info = NULL; } |
2b2af54a5
|
3415 |
int shmem_fill_super(struct super_block *sb, void *data, int silent) |
1da177e4c
|
3416 3417 |
{ struct inode *inode; |
0edd73b33
|
3418 |
struct shmem_sb_info *sbinfo; |
680d794ba
|
3419 3420 3421 |
int err = -ENOMEM; /* Round up to L1_CACHE_BYTES to resist false sharing */ |
425fbf047
|
3422 |
sbinfo = kzalloc(max((int)sizeof(struct shmem_sb_info), |
680d794ba
|
3423 3424 3425 |
L1_CACHE_BYTES), GFP_KERNEL); if (!sbinfo) return -ENOMEM; |
680d794ba
|
3426 |
sbinfo->mode = S_IRWXUGO | S_ISVTX; |
76aac0e9a
|
3427 3428 |
sbinfo->uid = current_fsuid(); sbinfo->gid = current_fsgid(); |
680d794ba
|
3429 |
sb->s_fs_info = sbinfo; |
1da177e4c
|
3430 |
|
0edd73b33
|
3431 |
#ifdef CONFIG_TMPFS |
1da177e4c
|
3432 3433 3434 3435 3436 |
/* * 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. */ |
ca4e05195
|
3437 |
if (!(sb->s_flags & MS_KERNMOUNT)) { |
680d794ba
|
3438 3439 3440 3441 3442 3443 |
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; } |
ca4e05195
|
3444 3445 |
} else { sb->s_flags |= MS_NOUSER; |
1da177e4c
|
3446 |
} |
91828a405
|
3447 |
sb->s_export_op = &shmem_export_ops; |
2f6e38f3c
|
3448 |
sb->s_flags |= MS_NOSEC; |
1da177e4c
|
3449 3450 3451 |
#else sb->s_flags |= MS_NOUSER; #endif |
0edd73b33
|
3452 |
spin_lock_init(&sbinfo->stat_lock); |
908c7f194
|
3453 |
if (percpu_counter_init(&sbinfo->used_blocks, 0, GFP_KERNEL)) |
602586a83
|
3454 |
goto failed; |
680d794ba
|
3455 |
sbinfo->free_inodes = sbinfo->max_inodes; |
779750d20
|
3456 3457 |
spin_lock_init(&sbinfo->shrinklist_lock); INIT_LIST_HEAD(&sbinfo->shrinklist); |
0edd73b33
|
3458 |
|
285b2c4fd
|
3459 |
sb->s_maxbytes = MAX_LFS_FILESIZE; |
09cbfeaf1
|
3460 3461 |
sb->s_blocksize = PAGE_SIZE; sb->s_blocksize_bits = PAGE_SHIFT; |
1da177e4c
|
3462 3463 |
sb->s_magic = TMPFS_MAGIC; sb->s_op = &shmem_ops; |
cfd95a9cf
|
3464 |
sb->s_time_gran = 1; |
b09e0fa4b
|
3465 |
#ifdef CONFIG_TMPFS_XATTR |
39f0247d3
|
3466 |
sb->s_xattr = shmem_xattr_handlers; |
b09e0fa4b
|
3467 3468 |
#endif #ifdef CONFIG_TMPFS_POSIX_ACL |
39f0247d3
|
3469 3470 |
sb->s_flags |= MS_POSIXACL; #endif |
0edd73b33
|
3471 |
|
454abafe9
|
3472 |
inode = shmem_get_inode(sb, NULL, S_IFDIR | sbinfo->mode, 0, VM_NORESERVE); |
1da177e4c
|
3473 3474 |
if (!inode) goto failed; |
680d794ba
|
3475 3476 |
inode->i_uid = sbinfo->uid; inode->i_gid = sbinfo->gid; |
318ceed08
|
3477 3478 |
sb->s_root = d_make_root(inode); if (!sb->s_root) |
48fde701a
|
3479 |
goto failed; |
1da177e4c
|
3480 |
return 0; |
1da177e4c
|
3481 3482 3483 3484 |
failed: shmem_put_super(sb); return err; } |
fcc234f88
|
3485 |
static struct kmem_cache *shmem_inode_cachep; |
1da177e4c
|
3486 3487 3488 |
static struct inode *shmem_alloc_inode(struct super_block *sb) { |
41ffe5d5c
|
3489 3490 3491 |
struct shmem_inode_info *info; info = kmem_cache_alloc(shmem_inode_cachep, GFP_KERNEL); if (!info) |
1da177e4c
|
3492 |
return NULL; |
41ffe5d5c
|
3493 |
return &info->vfs_inode; |
1da177e4c
|
3494 |
} |
41ffe5d5c
|
3495 |
static void shmem_destroy_callback(struct rcu_head *head) |
fa0d7e3de
|
3496 3497 |
{ struct inode *inode = container_of(head, struct inode, i_rcu); |
84e710da2
|
3498 3499 |
if (S_ISLNK(inode->i_mode)) kfree(inode->i_link); |
fa0d7e3de
|
3500 3501 |
kmem_cache_free(shmem_inode_cachep, SHMEM_I(inode)); } |
1da177e4c
|
3502 3503 |
static void shmem_destroy_inode(struct inode *inode) { |
09208d150
|
3504 |
if (S_ISREG(inode->i_mode)) |
1da177e4c
|
3505 |
mpol_free_shared_policy(&SHMEM_I(inode)->policy); |
41ffe5d5c
|
3506 |
call_rcu(&inode->i_rcu, shmem_destroy_callback); |
1da177e4c
|
3507 |
} |
41ffe5d5c
|
3508 |
static void shmem_init_inode(void *foo) |
1da177e4c
|
3509 |
{ |
41ffe5d5c
|
3510 3511 |
struct shmem_inode_info *info = foo; inode_init_once(&info->vfs_inode); |
1da177e4c
|
3512 |
} |
41ffe5d5c
|
3513 |
static int shmem_init_inodecache(void) |
1da177e4c
|
3514 3515 3516 |
{ shmem_inode_cachep = kmem_cache_create("shmem_inode_cache", sizeof(struct shmem_inode_info), |
5d097056c
|
3517 |
0, SLAB_PANIC|SLAB_ACCOUNT, shmem_init_inode); |
1da177e4c
|
3518 3519 |
return 0; } |
41ffe5d5c
|
3520 |
static void shmem_destroy_inodecache(void) |
1da177e4c
|
3521 |
{ |
1a1d92c10
|
3522 |
kmem_cache_destroy(shmem_inode_cachep); |
1da177e4c
|
3523 |
} |
f5e54d6e5
|
3524 |
static const struct address_space_operations shmem_aops = { |
1da177e4c
|
3525 |
.writepage = shmem_writepage, |
767193253
|
3526 |
.set_page_dirty = __set_page_dirty_no_writeback, |
1da177e4c
|
3527 |
#ifdef CONFIG_TMPFS |
800d15a53
|
3528 3529 |
.write_begin = shmem_write_begin, .write_end = shmem_write_end, |
1da177e4c
|
3530 |
#endif |
1c93923cc
|
3531 |
#ifdef CONFIG_MIGRATION |
304dbdb7a
|
3532 |
.migratepage = migrate_page, |
1c93923cc
|
3533 |
#endif |
aa261f549
|
3534 |
.error_remove_page = generic_error_remove_page, |
1da177e4c
|
3535 |
}; |
15ad7cdcf
|
3536 |
static const struct file_operations shmem_file_operations = { |
1da177e4c
|
3537 |
.mmap = shmem_mmap, |
c01d5b300
|
3538 |
.get_unmapped_area = shmem_get_unmapped_area, |
1da177e4c
|
3539 |
#ifdef CONFIG_TMPFS |
220f2ac91
|
3540 |
.llseek = shmem_file_llseek, |
2ba5bbed0
|
3541 |
.read_iter = shmem_file_read_iter, |
8174202b3
|
3542 |
.write_iter = generic_file_write_iter, |
1b061d924
|
3543 |
.fsync = noop_fsync, |
82c156f85
|
3544 |
.splice_read = generic_file_splice_read, |
f6cb85d00
|
3545 |
.splice_write = iter_file_splice_write, |
83e4fa9c1
|
3546 |
.fallocate = shmem_fallocate, |
1da177e4c
|
3547 3548 |
#endif }; |
92e1d5be9
|
3549 |
static const struct inode_operations shmem_inode_operations = { |
44a30220b
|
3550 |
.getattr = shmem_getattr, |
94c1e62df
|
3551 |
.setattr = shmem_setattr, |
b09e0fa4b
|
3552 |
#ifdef CONFIG_TMPFS_XATTR |
b09e0fa4b
|
3553 |
.listxattr = shmem_listxattr, |
feda821e7
|
3554 |
.set_acl = simple_set_acl, |
b09e0fa4b
|
3555 |
#endif |
1da177e4c
|
3556 |
}; |
92e1d5be9
|
3557 |
static const struct inode_operations shmem_dir_inode_operations = { |
1da177e4c
|
3558 3559 3560 3561 3562 3563 3564 3565 3566 |
#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, |
2773bf00a
|
3567 |
.rename = shmem_rename2, |
60545d0d4
|
3568 |
.tmpfile = shmem_tmpfile, |
1da177e4c
|
3569 |
#endif |
b09e0fa4b
|
3570 |
#ifdef CONFIG_TMPFS_XATTR |
b09e0fa4b
|
3571 |
.listxattr = shmem_listxattr, |
b09e0fa4b
|
3572 |
#endif |
39f0247d3
|
3573 |
#ifdef CONFIG_TMPFS_POSIX_ACL |
94c1e62df
|
3574 |
.setattr = shmem_setattr, |
feda821e7
|
3575 |
.set_acl = simple_set_acl, |
39f0247d3
|
3576 3577 |
#endif }; |
92e1d5be9
|
3578 |
static const struct inode_operations shmem_special_inode_operations = { |
b09e0fa4b
|
3579 |
#ifdef CONFIG_TMPFS_XATTR |
b09e0fa4b
|
3580 |
.listxattr = shmem_listxattr, |
b09e0fa4b
|
3581 |
#endif |
39f0247d3
|
3582 |
#ifdef CONFIG_TMPFS_POSIX_ACL |
94c1e62df
|
3583 |
.setattr = shmem_setattr, |
feda821e7
|
3584 |
.set_acl = simple_set_acl, |
39f0247d3
|
3585 |
#endif |
1da177e4c
|
3586 |
}; |
759b9775c
|
3587 |
static const struct super_operations shmem_ops = { |
1da177e4c
|
3588 3589 3590 3591 3592 |
.alloc_inode = shmem_alloc_inode, .destroy_inode = shmem_destroy_inode, #ifdef CONFIG_TMPFS .statfs = shmem_statfs, .remount_fs = shmem_remount_fs, |
680d794ba
|
3593 |
.show_options = shmem_show_options, |
1da177e4c
|
3594 |
#endif |
1f895f75d
|
3595 |
.evict_inode = shmem_evict_inode, |
1da177e4c
|
3596 3597 |
.drop_inode = generic_delete_inode, .put_super = shmem_put_super, |
779750d20
|
3598 3599 3600 3601 |
#ifdef CONFIG_TRANSPARENT_HUGE_PAGECACHE .nr_cached_objects = shmem_unused_huge_count, .free_cached_objects = shmem_unused_huge_scan, #endif |
1da177e4c
|
3602 |
}; |
f0f37e2f7
|
3603 |
static const struct vm_operations_struct shmem_vm_ops = { |
54cb8821d
|
3604 |
.fault = shmem_fault, |
d7c175517
|
3605 |
.map_pages = filemap_map_pages, |
1da177e4c
|
3606 3607 3608 3609 3610 |
#ifdef CONFIG_NUMA .set_policy = shmem_set_policy, .get_policy = shmem_get_policy, #endif }; |
3c26ff6e4
|
3611 3612 |
static struct dentry *shmem_mount(struct file_system_type *fs_type, int flags, const char *dev_name, void *data) |
1da177e4c
|
3613 |
{ |
3c26ff6e4
|
3614 |
return mount_nodev(fs_type, flags, data, shmem_fill_super); |
1da177e4c
|
3615 |
} |
41ffe5d5c
|
3616 |
static struct file_system_type shmem_fs_type = { |
1da177e4c
|
3617 3618 |
.owner = THIS_MODULE, .name = "tmpfs", |
3c26ff6e4
|
3619 |
.mount = shmem_mount, |
1da177e4c
|
3620 |
.kill_sb = kill_litter_super, |
2b8576cb0
|
3621 |
.fs_flags = FS_USERNS_MOUNT, |
1da177e4c
|
3622 |
}; |
1da177e4c
|
3623 |
|
41ffe5d5c
|
3624 |
int __init shmem_init(void) |
1da177e4c
|
3625 3626 |
{ int error; |
16203a7a9
|
3627 3628 3629 |
/* If rootfs called this, don't re-init */ if (shmem_inode_cachep) return 0; |
41ffe5d5c
|
3630 |
error = shmem_init_inodecache(); |
1da177e4c
|
3631 3632 |
if (error) goto out3; |
41ffe5d5c
|
3633 |
error = register_filesystem(&shmem_fs_type); |
1da177e4c
|
3634 |
if (error) { |
1170532bb
|
3635 3636 |
pr_err("Could not register tmpfs "); |
1da177e4c
|
3637 3638 |
goto out2; } |
95dc112a5
|
3639 |
|
ca4e05195
|
3640 |
shm_mnt = kern_mount(&shmem_fs_type); |
1da177e4c
|
3641 3642 |
if (IS_ERR(shm_mnt)) { error = PTR_ERR(shm_mnt); |
1170532bb
|
3643 3644 |
pr_err("Could not kern_mount tmpfs "); |
1da177e4c
|
3645 3646 |
goto out1; } |
5a6e75f81
|
3647 |
|
e496cf3d7
|
3648 |
#ifdef CONFIG_TRANSPARENT_HUGE_PAGECACHE |
5a6e75f81
|
3649 3650 3651 3652 3653 |
if (has_transparent_hugepage() && shmem_huge < SHMEM_HUGE_DENY) SHMEM_SB(shm_mnt->mnt_sb)->huge = shmem_huge; else shmem_huge = 0; /* just in case it was patched */ #endif |
1da177e4c
|
3654 3655 3656 |
return 0; out1: |
41ffe5d5c
|
3657 |
unregister_filesystem(&shmem_fs_type); |
1da177e4c
|
3658 |
out2: |
41ffe5d5c
|
3659 |
shmem_destroy_inodecache(); |
1da177e4c
|
3660 3661 3662 3663 |
out3: shm_mnt = ERR_PTR(error); return error; } |
853ac43ab
|
3664 |
|
e496cf3d7
|
3665 |
#if defined(CONFIG_TRANSPARENT_HUGE_PAGECACHE) && defined(CONFIG_SYSFS) |
5a6e75f81
|
3666 3667 3668 3669 3670 3671 3672 3673 3674 3675 3676 3677 3678 3679 3680 3681 3682 3683 3684 3685 3686 3687 3688 3689 3690 3691 3692 3693 3694 3695 3696 3697 3698 3699 3700 3701 3702 3703 3704 3705 3706 3707 3708 3709 3710 3711 3712 3713 3714 3715 3716 3717 3718 |
static ssize_t shmem_enabled_show(struct kobject *kobj, struct kobj_attribute *attr, char *buf) { int values[] = { SHMEM_HUGE_ALWAYS, SHMEM_HUGE_WITHIN_SIZE, SHMEM_HUGE_ADVISE, SHMEM_HUGE_NEVER, SHMEM_HUGE_DENY, SHMEM_HUGE_FORCE, }; int i, count; for (i = 0, count = 0; i < ARRAY_SIZE(values); i++) { const char *fmt = shmem_huge == values[i] ? "[%s] " : "%s "; count += sprintf(buf + count, fmt, shmem_format_huge(values[i])); } buf[count - 1] = ' '; return count; } static ssize_t shmem_enabled_store(struct kobject *kobj, struct kobj_attribute *attr, const char *buf, size_t count) { char tmp[16]; int huge; if (count + 1 > sizeof(tmp)) return -EINVAL; memcpy(tmp, buf, count); tmp[count] = '\0'; if (count && tmp[count - 1] == ' ') tmp[count - 1] = '\0'; huge = shmem_parse_huge(tmp); if (huge == -EINVAL) return -EINVAL; if (!has_transparent_hugepage() && huge != SHMEM_HUGE_NEVER && huge != SHMEM_HUGE_DENY) return -EINVAL; shmem_huge = huge; if (shmem_huge < SHMEM_HUGE_DENY) SHMEM_SB(shm_mnt->mnt_sb)->huge = shmem_huge; return count; } struct kobj_attribute shmem_enabled_attr = __ATTR(shmem_enabled, 0644, shmem_enabled_show, shmem_enabled_store); |
3b33719c9
|
3719 |
#endif /* CONFIG_TRANSPARENT_HUGE_PAGECACHE && CONFIG_SYSFS */ |
f3f0e1d21
|
3720 |
|
3b33719c9
|
3721 |
#ifdef CONFIG_TRANSPARENT_HUGE_PAGECACHE |
f3f0e1d21
|
3722 3723 3724 3725 3726 3727 3728 3729 3730 3731 3732 3733 3734 3735 3736 3737 3738 3739 3740 3741 3742 3743 3744 3745 3746 3747 3748 3749 3750 3751 |
bool shmem_huge_enabled(struct vm_area_struct *vma) { struct inode *inode = file_inode(vma->vm_file); struct shmem_sb_info *sbinfo = SHMEM_SB(inode->i_sb); loff_t i_size; pgoff_t off; if (shmem_huge == SHMEM_HUGE_FORCE) return true; if (shmem_huge == SHMEM_HUGE_DENY) return false; switch (sbinfo->huge) { case SHMEM_HUGE_NEVER: return false; case SHMEM_HUGE_ALWAYS: return true; case SHMEM_HUGE_WITHIN_SIZE: off = round_up(vma->vm_pgoff, HPAGE_PMD_NR); i_size = round_up(i_size_read(inode), PAGE_SIZE); if (i_size >= HPAGE_PMD_SIZE && i_size >> PAGE_SHIFT >= off) return true; case SHMEM_HUGE_ADVISE: /* TODO: implement fadvise() hints */ return (vma->vm_flags & VM_HUGEPAGE); default: VM_BUG_ON(1); return false; } } |
3b33719c9
|
3752 |
#endif /* CONFIG_TRANSPARENT_HUGE_PAGECACHE */ |
5a6e75f81
|
3753 |
|
853ac43ab
|
3754 3755 3756 3757 3758 3759 3760 3761 3762 3763 |
#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. */ |
41ffe5d5c
|
3764 |
static struct file_system_type shmem_fs_type = { |
853ac43ab
|
3765 |
.name = "tmpfs", |
3c26ff6e4
|
3766 |
.mount = ramfs_mount, |
853ac43ab
|
3767 |
.kill_sb = kill_litter_super, |
2b8576cb0
|
3768 |
.fs_flags = FS_USERNS_MOUNT, |
853ac43ab
|
3769 |
}; |
41ffe5d5c
|
3770 |
int __init shmem_init(void) |
853ac43ab
|
3771 |
{ |
41ffe5d5c
|
3772 |
BUG_ON(register_filesystem(&shmem_fs_type) != 0); |
853ac43ab
|
3773 |
|
41ffe5d5c
|
3774 |
shm_mnt = kern_mount(&shmem_fs_type); |
853ac43ab
|
3775 3776 3777 3778 |
BUG_ON(IS_ERR(shm_mnt)); return 0; } |
41ffe5d5c
|
3779 |
int shmem_unuse(swp_entry_t swap, struct page *page) |
853ac43ab
|
3780 3781 3782 |
{ return 0; } |
3f96b79ad
|
3783 3784 3785 3786 |
int shmem_lock(struct file *file, int lock, struct user_struct *user) { return 0; } |
245132643
|
3787 3788 3789 |
void shmem_unlock_mapping(struct address_space *mapping) { } |
c01d5b300
|
3790 3791 3792 3793 3794 3795 3796 3797 |
#ifdef CONFIG_MMU unsigned long shmem_get_unmapped_area(struct file *file, unsigned long addr, unsigned long len, unsigned long pgoff, unsigned long flags) { return current->mm->get_unmapped_area(file, addr, len, pgoff, flags); } #endif |
41ffe5d5c
|
3798 |
void shmem_truncate_range(struct inode *inode, loff_t lstart, loff_t lend) |
94c1e62df
|
3799 |
{ |
41ffe5d5c
|
3800 |
truncate_inode_pages_range(inode->i_mapping, lstart, lend); |
94c1e62df
|
3801 3802 |
} EXPORT_SYMBOL_GPL(shmem_truncate_range); |
0b0a0806b
|
3803 3804 |
#define shmem_vm_ops generic_file_vm_ops #define shmem_file_operations ramfs_file_operations |
454abafe9
|
3805 |
#define shmem_get_inode(sb, dir, mode, dev, flags) ramfs_get_inode(sb, dir, mode, dev) |
0b0a0806b
|
3806 3807 |
#define shmem_acct_size(flags, size) 0 #define shmem_unacct_size(flags, size) do {} while (0) |
853ac43ab
|
3808 3809 3810 3811 |
#endif /* CONFIG_SHMEM */ /* common code */ |
1da177e4c
|
3812 |
|
19938e350
|
3813 |
static const struct dentry_operations anon_ops = { |
118b23022
|
3814 |
.d_dname = simple_dname |
3451538a1
|
3815 |
}; |
c72770909
|
3816 3817 |
static struct file *__shmem_file_setup(const char *name, loff_t size, unsigned long flags, unsigned int i_flags) |
1da177e4c
|
3818 |
{ |
6b4d0b279
|
3819 |
struct file *res; |
1da177e4c
|
3820 |
struct inode *inode; |
2c48b9c45
|
3821 |
struct path path; |
3451538a1
|
3822 |
struct super_block *sb; |
1da177e4c
|
3823 3824 3825 |
struct qstr this; if (IS_ERR(shm_mnt)) |
6b4d0b279
|
3826 |
return ERR_CAST(shm_mnt); |
1da177e4c
|
3827 |
|
285b2c4fd
|
3828 |
if (size < 0 || size > MAX_LFS_FILESIZE) |
1da177e4c
|
3829 3830 3831 3832 |
return ERR_PTR(-EINVAL); if (shmem_acct_size(flags, size)) return ERR_PTR(-ENOMEM); |
6b4d0b279
|
3833 |
res = ERR_PTR(-ENOMEM); |
1da177e4c
|
3834 3835 3836 |
this.name = name; this.len = strlen(name); this.hash = 0; /* will go */ |
3451538a1
|
3837 |
sb = shm_mnt->mnt_sb; |
66ee4b888
|
3838 |
path.mnt = mntget(shm_mnt); |
3451538a1
|
3839 |
path.dentry = d_alloc_pseudo(sb, &this); |
2c48b9c45
|
3840 |
if (!path.dentry) |
1da177e4c
|
3841 |
goto put_memory; |
3451538a1
|
3842 |
d_set_d_op(path.dentry, &anon_ops); |
1da177e4c
|
3843 |
|
6b4d0b279
|
3844 |
res = ERR_PTR(-ENOSPC); |
3451538a1
|
3845 |
inode = shmem_get_inode(sb, NULL, S_IFREG | S_IRWXUGO, 0, flags); |
1da177e4c
|
3846 |
if (!inode) |
66ee4b888
|
3847 |
goto put_memory; |
1da177e4c
|
3848 |
|
c72770909
|
3849 |
inode->i_flags |= i_flags; |
2c48b9c45
|
3850 |
d_instantiate(path.dentry, inode); |
1da177e4c
|
3851 |
inode->i_size = size; |
6d6b77f16
|
3852 |
clear_nlink(inode); /* It is unlinked */ |
26567cdbb
|
3853 3854 |
res = ERR_PTR(ramfs_nommu_expand_for_mapping(inode, size)); if (IS_ERR(res)) |
66ee4b888
|
3855 |
goto put_path; |
4b42af81f
|
3856 |
|
6b4d0b279
|
3857 |
res = alloc_file(&path, FMODE_WRITE | FMODE_READ, |
4b42af81f
|
3858 |
&shmem_file_operations); |
6b4d0b279
|
3859 |
if (IS_ERR(res)) |
66ee4b888
|
3860 |
goto put_path; |
4b42af81f
|
3861 |
|
6b4d0b279
|
3862 |
return res; |
1da177e4c
|
3863 |
|
1da177e4c
|
3864 3865 |
put_memory: shmem_unacct_size(flags, size); |
66ee4b888
|
3866 3867 |
put_path: path_put(&path); |
6b4d0b279
|
3868 |
return res; |
1da177e4c
|
3869 |
} |
c72770909
|
3870 3871 3872 3873 3874 |
/** * shmem_kernel_file_setup - get an unlinked file living in tmpfs which must be * kernel internal. There will be NO LSM permission checks against the * underlying inode. So users of this interface must do LSM checks at a |
e1832f292
|
3875 3876 |
* higher layer. The users are the big_key and shm implementations. LSM * checks are provided at the key or shm level rather than the inode. |
c72770909
|
3877 3878 3879 3880 3881 3882 3883 3884 3885 3886 3887 3888 3889 3890 3891 3892 3893 3894 3895 |
* @name: name for dentry (to be seen in /proc/<pid>/maps * @size: size to be set for the file * @flags: VM_NORESERVE suppresses pre-accounting of the entire object size */ struct file *shmem_kernel_file_setup(const char *name, loff_t size, unsigned long flags) { return __shmem_file_setup(name, size, flags, S_PRIVATE); } /** * 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 * @flags: VM_NORESERVE suppresses pre-accounting of the entire object size */ struct file *shmem_file_setup(const char *name, loff_t size, unsigned long flags) { return __shmem_file_setup(name, size, flags, 0); } |
395e0ddc4
|
3896 |
EXPORT_SYMBOL_GPL(shmem_file_setup); |
1da177e4c
|
3897 |
|
467118102
|
3898 |
/** |
1da177e4c
|
3899 |
* shmem_zero_setup - setup a shared anonymous mapping |
1da177e4c
|
3900 3901 3902 3903 3904 3905 |
* @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; |
66fc13039
|
3906 3907 3908 3909 3910 3911 3912 |
/* * Cloning a new file under mmap_sem leads to a lock ordering conflict * between XFS directory reading and selinux: since this file is only * accessible to the user through its mapping, use S_PRIVATE flag to * bypass file security, in the same way as shmem_kernel_file_setup(). */ file = __shmem_file_setup("dev/zero", size, vma->vm_flags, S_PRIVATE); |
1da177e4c
|
3913 3914 3915 3916 3917 3918 3919 |
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; |
f3f0e1d21
|
3920 |
|
e496cf3d7
|
3921 |
if (IS_ENABLED(CONFIG_TRANSPARENT_HUGE_PAGECACHE) && |
f3f0e1d21
|
3922 3923 3924 3925 |
((vma->vm_start + ~HPAGE_PMD_MASK) & HPAGE_PMD_MASK) < (vma->vm_end & HPAGE_PMD_MASK)) { khugepaged_enter(vma, vma->vm_flags); } |
1da177e4c
|
3926 3927 |
return 0; } |
d9d90e5eb
|
3928 3929 3930 3931 3932 3933 3934 3935 3936 3937 3938 3939 3940 |
/** * 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
|
3941 3942 |
* i915_gem_object_get_pages_gtt() mixes __GFP_NORETRY | __GFP_NOWARN in * with the mapping_gfp_mask(), to avoid OOMing the machine unnecessarily. |
d9d90e5eb
|
3943 3944 3945 3946 |
*/ struct page *shmem_read_mapping_page_gfp(struct address_space *mapping, pgoff_t index, gfp_t gfp) { |
68da9f055
|
3947 3948 |
#ifdef CONFIG_SHMEM struct inode *inode = mapping->host; |
9276aad6c
|
3949 |
struct page *page; |
68da9f055
|
3950 3951 3952 |
int error; BUG_ON(mapping->a_ops != &shmem_aops); |
9e18eb293
|
3953 3954 |
error = shmem_getpage_gfp(inode, index, &page, SGP_CACHE, gfp, NULL, NULL); |
68da9f055
|
3955 3956 3957 3958 3959 3960 3961 3962 3963 |
if (error) page = ERR_PTR(error); else unlock_page(page); return page; #else /* * The tiny !SHMEM case uses ramfs without swap */ |
d9d90e5eb
|
3964 |
return read_cache_page_gfp(mapping, index, gfp); |
68da9f055
|
3965 |
#endif |
d9d90e5eb
|
3966 3967 |
} EXPORT_SYMBOL_GPL(shmem_read_mapping_page_gfp); |