Blame view
mm/shmem.c
112 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> |
174cd4b1e
|
31 |
#include <linux/sched/signal.h> |
b95f1b31b
|
32 |
#include <linux/export.h> |
853ac43ab
|
33 |
#include <linux/swap.h> |
e2e40f2c1
|
34 |
#include <linux/uio.h> |
f3f0e1d21
|
35 |
#include <linux/khugepaged.h> |
749df87bd
|
36 |
#include <linux/hugetlb.h> |
853ac43ab
|
37 |
|
95cc09d66
|
38 |
#include <asm/tlbflush.h> /* for arch/microblaze update_mmu_cache() */ |
853ac43ab
|
39 40 41 |
static struct vfsmount *shm_mnt; #ifdef CONFIG_SHMEM |
1da177e4c
|
42 43 44 45 46 |
/* * 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
|
47 |
#include <linux/xattr.h> |
a56942551
|
48 |
#include <linux/exportfs.h> |
1c7c474c3
|
49 |
#include <linux/posix_acl.h> |
feda821e7
|
50 |
#include <linux/posix_acl_xattr.h> |
1da177e4c
|
51 |
#include <linux/mman.h> |
1da177e4c
|
52 53 54 55 |
#include <linux/string.h> #include <linux/slab.h> #include <linux/backing-dev.h> #include <linux/shmem_fs.h> |
1da177e4c
|
56 |
#include <linux/writeback.h> |
1da177e4c
|
57 |
#include <linux/blkdev.h> |
bda97eab0
|
58 |
#include <linux/pagevec.h> |
41ffe5d5c
|
59 |
#include <linux/percpu_counter.h> |
83e4fa9c1
|
60 |
#include <linux/falloc.h> |
708e3508c
|
61 |
#include <linux/splice.h> |
1da177e4c
|
62 63 64 65 |
#include <linux/security.h> #include <linux/swapops.h> #include <linux/mempolicy.h> #include <linux/namei.h> |
b00dc3ad7
|
66 |
#include <linux/ctype.h> |
304dbdb7a
|
67 |
#include <linux/migrate.h> |
c1f60a5a4
|
68 |
#include <linux/highmem.h> |
680d794ba
|
69 |
#include <linux/seq_file.h> |
925629278
|
70 |
#include <linux/magic.h> |
9183df25f
|
71 |
#include <linux/syscalls.h> |
40e041a2c
|
72 |
#include <linux/fcntl.h> |
9183df25f
|
73 |
#include <uapi/linux/memfd.h> |
cfda05267
|
74 |
#include <linux/userfaultfd_k.h> |
4c27fe4c4
|
75 |
#include <linux/rmap.h> |
2b4db7961
|
76 |
#include <linux/uuid.h> |
304dbdb7a
|
77 |
|
7c0f6ba68
|
78 |
#include <linux/uaccess.h> |
1da177e4c
|
79 |
#include <asm/pgtable.h> |
dd56b0464
|
80 |
#include "internal.h" |
09cbfeaf1
|
81 82 |
#define BLOCKS_PER_PAGE (PAGE_SIZE/512) #define VM_ACCT(size) (PAGE_ALIGN(size) >> PAGE_SHIFT) |
1da177e4c
|
83 |
|
1da177e4c
|
84 85 |
/* Pretend that each entry is of this size in directory's i_size */ #define BOGO_DIRENT_SIZE 20 |
69f07ec93
|
86 87 |
/* Symlink up to this size is kmalloc'ed instead of using a swappable page */ #define SHORT_SYMLINK_LEN 128 |
1aac14003
|
88 |
/* |
f00cdc6df
|
89 90 91 |
* 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
|
92 93 |
*/ struct shmem_falloc { |
8e205f779
|
94 |
wait_queue_head_t *waitq; /* faults into hole wait for punch to end */ |
1aac14003
|
95 96 97 98 99 |
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
|
100 |
#ifdef CONFIG_TMPFS |
680d794ba
|
101 102 103 104 105 106 107 108 109 |
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
|
110 |
#endif |
680d794ba
|
111 |
|
bde05d1cc
|
112 113 114 |
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
|
115 |
static int shmem_getpage_gfp(struct inode *inode, pgoff_t index, |
9e18eb293
|
116 |
struct page **pagep, enum sgp_type sgp, |
cfda05267
|
117 118 |
gfp_t gfp, struct vm_area_struct *vma, struct vm_fault *vmf, int *fault_type); |
68da9f055
|
119 |
|
f3f0e1d21
|
120 |
int shmem_getpage(struct inode *inode, pgoff_t index, |
9e18eb293
|
121 |
struct page **pagep, enum sgp_type sgp) |
68da9f055
|
122 123 |
{ return shmem_getpage_gfp(inode, index, pagep, sgp, |
cfda05267
|
124 |
mapping_gfp_mask(inode->i_mapping), NULL, NULL, NULL); |
68da9f055
|
125 |
} |
1da177e4c
|
126 |
|
1da177e4c
|
127 128 129 130 131 132 133 134 135 136 137 138 139 |
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
|
140 |
return (flags & VM_NORESERVE) ? |
191c54244
|
141 |
0 : security_vm_enough_memory_mm(current->mm, VM_ACCT(size)); |
1da177e4c
|
142 143 144 145 |
} static inline void shmem_unacct_size(unsigned long flags, loff_t size) { |
0b0a0806b
|
146 |
if (!(flags & VM_NORESERVE)) |
1da177e4c
|
147 148 |
vm_unacct_memory(VM_ACCT(size)); } |
771425179
|
149 150 151 152 153 154 155 156 157 158 159 160 |
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
|
161 162 |
/* * ... whereas tmpfs objects are accounted incrementally as |
75edd345e
|
163 |
* pages are allocated, in order to allow large sparse files. |
1da177e4c
|
164 165 166 |
* 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
|
167 |
static inline int shmem_acct_block(unsigned long flags, long pages) |
1da177e4c
|
168 |
{ |
800d8c63b
|
169 170 171 172 173 |
if (!(flags & VM_NORESERVE)) return 0; return security_vm_enough_memory_mm(current->mm, pages * VM_ACCT(PAGE_SIZE)); |
1da177e4c
|
174 175 176 177 |
} static inline void shmem_unacct_blocks(unsigned long flags, long pages) { |
0b0a0806b
|
178 |
if (flags & VM_NORESERVE) |
09cbfeaf1
|
179 |
vm_unacct_memory(pages * VM_ACCT(PAGE_SIZE)); |
1da177e4c
|
180 |
} |
0f0796945
|
181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 |
static inline bool shmem_inode_acct_block(struct inode *inode, long pages) { struct shmem_inode_info *info = SHMEM_I(inode); struct shmem_sb_info *sbinfo = SHMEM_SB(inode->i_sb); if (shmem_acct_block(info->flags, pages)) return false; if (sbinfo->max_blocks) { if (percpu_counter_compare(&sbinfo->used_blocks, sbinfo->max_blocks - pages) > 0) goto unacct; percpu_counter_add(&sbinfo->used_blocks, pages); } return true; unacct: shmem_unacct_blocks(info->flags, pages); return false; } static inline void shmem_inode_unacct_blocks(struct inode *inode, long pages) { struct shmem_inode_info *info = SHMEM_I(inode); struct shmem_sb_info *sbinfo = SHMEM_SB(inode->i_sb); if (sbinfo->max_blocks) percpu_counter_sub(&sbinfo->used_blocks, pages); shmem_unacct_blocks(info->flags, pages); } |
759b9775c
|
212 |
static const struct super_operations shmem_ops; |
f5e54d6e5
|
213 |
static const struct address_space_operations shmem_aops; |
15ad7cdcf
|
214 |
static const struct file_operations shmem_file_operations; |
92e1d5be9
|
215 216 217 |
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
|
218 |
static const struct vm_operations_struct shmem_vm_ops; |
779750d20
|
219 |
static struct file_system_type shmem_fs_type; |
1da177e4c
|
220 |
|
b0506e488
|
221 222 223 224 |
bool vma_is_shmem(struct vm_area_struct *vma) { return vma->vm_ops == &shmem_vm_ops; } |
1da177e4c
|
225 |
static LIST_HEAD(shmem_swaplist); |
cb5f7b9a4
|
226 |
static DEFINE_MUTEX(shmem_swaplist_mutex); |
1da177e4c
|
227 |
|
5b04c6890
|
228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 |
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
|
252 |
/** |
41ffe5d5c
|
253 |
* shmem_recalc_inode - recalculate the block usage of an inode |
1da177e4c
|
254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 |
* @inode: inode to recalc * * We have to calculate the free blocks since the mm can drop * undirtied hole pages behind our back. * * But normally info->alloced == inode->i_mapping->nrpages + info->swapped * So mm freed is info->alloced - (inode->i_mapping->nrpages + info->swapped) * * It has to be called with the spinlock held. */ static void shmem_recalc_inode(struct inode *inode) { struct shmem_inode_info *info = SHMEM_I(inode); long freed; freed = info->alloced - info->swapped - inode->i_mapping->nrpages; if (freed > 0) { info->alloced -= freed; |
54af60421
|
272 |
inode->i_blocks -= freed * BLOCKS_PER_PAGE; |
0f0796945
|
273 |
shmem_inode_unacct_blocks(inode, freed); |
1da177e4c
|
274 275 |
} } |
800d8c63b
|
276 277 278 |
bool shmem_charge(struct inode *inode, long pages) { struct shmem_inode_info *info = SHMEM_I(inode); |
4595ef88d
|
279 |
unsigned long flags; |
800d8c63b
|
280 |
|
0f0796945
|
281 |
if (!shmem_inode_acct_block(inode, pages)) |
800d8c63b
|
282 |
return false; |
b1cc94ab2
|
283 |
|
98f1ae169
|
284 285 |
/* nrpages adjustment first, then shmem_recalc_inode() when balanced */ inode->i_mapping->nrpages += pages; |
4595ef88d
|
286 |
spin_lock_irqsave(&info->lock, flags); |
800d8c63b
|
287 288 289 |
info->alloced += pages; inode->i_blocks += pages * BLOCKS_PER_PAGE; shmem_recalc_inode(inode); |
4595ef88d
|
290 |
spin_unlock_irqrestore(&info->lock, flags); |
800d8c63b
|
291 |
|
800d8c63b
|
292 293 294 295 296 297 |
return true; } void shmem_uncharge(struct inode *inode, long pages) { struct shmem_inode_info *info = SHMEM_I(inode); |
4595ef88d
|
298 |
unsigned long flags; |
800d8c63b
|
299 |
|
98f1ae169
|
300 |
/* nrpages adjustment done by __delete_from_page_cache() or caller */ |
4595ef88d
|
301 |
spin_lock_irqsave(&info->lock, flags); |
800d8c63b
|
302 303 304 |
info->alloced -= pages; inode->i_blocks -= pages * BLOCKS_PER_PAGE; shmem_recalc_inode(inode); |
4595ef88d
|
305 |
spin_unlock_irqrestore(&info->lock, flags); |
800d8c63b
|
306 |
|
0f0796945
|
307 |
shmem_inode_unacct_blocks(inode, pages); |
800d8c63b
|
308 |
} |
7a5d0fbb2
|
309 310 311 312 313 314 |
/* * 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) { |
f7942430e
|
315 |
struct radix_tree_node *node; |
7a5d0fbb2
|
316 |
void **pslot; |
6dbaf22ce
|
317 |
void *item; |
7a5d0fbb2
|
318 319 |
VM_BUG_ON(!expected); |
6dbaf22ce
|
320 |
VM_BUG_ON(!replacement); |
f7942430e
|
321 322 |
item = __radix_tree_lookup(&mapping->page_tree, index, &node, &pslot); if (!item) |
6dbaf22ce
|
323 |
return -ENOENT; |
7a5d0fbb2
|
324 325 |
if (item != expected) return -ENOENT; |
4d693d086
|
326 327 |
__radix_tree_replace(&mapping->page_tree, node, pslot, replacement, NULL, NULL); |
7a5d0fbb2
|
328 329 330 331 |
return 0; } /* |
d18992286
|
332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 |
* 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
|
350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 |
* 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
|
380 |
#ifdef CONFIG_TRANSPARENT_HUGE_PAGECACHE |
5a6e75f81
|
381 382 383 |
/* ifdef here to avoid bloating shmem.o when not necessary */ int shmem_huge __read_mostly; |
f1f5929cd
|
384 |
#if defined(CONFIG_SYSFS) || defined(CONFIG_TMPFS) |
5a6e75f81
|
385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 |
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"; } } |
f1f5929cd
|
422 |
#endif |
5a6e75f81
|
423 |
|
779750d20
|
424 425 426 427 |
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; |
253fd0f02
|
428 |
LIST_HEAD(to_remove); |
779750d20
|
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 |
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)) { |
253fd0f02
|
455 |
list_move(&info->shrinklist, &to_remove); |
779750d20
|
456 |
removed++; |
779750d20
|
457 458 459 460 461 462 463 464 465 |
goto next; } list_move(&info->shrinklist, &list); next: if (!--batch) break; } spin_unlock(&sbinfo->shrinklist_lock); |
253fd0f02
|
466 467 468 469 470 471 |
list_for_each_safe(pos, next, &to_remove) { info = list_entry(pos, struct shmem_inode_info, shrinklist); inode = &info->vfs_inode; list_del_init(&info->shrinklist); iput(inode); } |
779750d20
|
472 473 474 475 476 |
list_for_each_safe(pos, next, &list) { int ret; info = list_entry(pos, struct shmem_inode_info, shrinklist); inode = &info->vfs_inode; |
f5dad0409
|
477 478 |
if (nr_to_split && split >= nr_to_split) goto leave; |
779750d20
|
479 |
|
f5dad0409
|
480 |
page = find_get_page(inode->i_mapping, |
779750d20
|
481 482 483 |
(inode->i_size & HPAGE_PMD_MASK) >> PAGE_SHIFT); if (!page) goto drop; |
f5dad0409
|
484 |
/* No huge page at the end of the file: nothing to split */ |
779750d20
|
485 |
if (!PageTransHuge(page)) { |
779750d20
|
486 487 488 |
put_page(page); goto drop; } |
f5dad0409
|
489 490 491 492 493 494 495 496 497 498 499 |
/* * Leave the inode on the list if we failed to lock * the page at this time. * * Waiting for the lock may lead to deadlock in the * reclaim path. */ if (!trylock_page(page)) { put_page(page); goto leave; } |
779750d20
|
500 501 502 |
ret = split_huge_page(page); unlock_page(page); put_page(page); |
f5dad0409
|
503 504 505 |
/* If split failed leave the inode on the list */ if (ret) goto leave; |
779750d20
|
506 507 508 509 510 |
split++; drop: list_del_init(&info->shrinklist); removed++; |
f5dad0409
|
511 |
leave: |
779750d20
|
512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 |
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
|
540 |
#else /* !CONFIG_TRANSPARENT_HUGE_PAGECACHE */ |
5a6e75f81
|
541 542 |
#define shmem_huge SHMEM_HUGE_DENY |
779750d20
|
543 544 545 546 547 |
static unsigned long shmem_unused_huge_shrink(struct shmem_sb_info *sbinfo, struct shrink_control *sc, unsigned long nr_to_split) { return 0; } |
e496cf3d7
|
548 |
#endif /* CONFIG_TRANSPARENT_HUGE_PAGECACHE */ |
5a6e75f81
|
549 550 |
/* |
46f65ec15
|
551 552 553 554 |
* 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
|
555 |
pgoff_t index, void *expected) |
46f65ec15
|
556 |
{ |
800d8c63b
|
557 |
int error, nr = hpage_nr_pages(page); |
46f65ec15
|
558 |
|
800d8c63b
|
559 560 |
VM_BUG_ON_PAGE(PageTail(page), page); VM_BUG_ON_PAGE(index != round_down(index, nr), page); |
309381fea
|
561 562 |
VM_BUG_ON_PAGE(!PageLocked(page), page); VM_BUG_ON_PAGE(!PageSwapBacked(page), page); |
800d8c63b
|
563 |
VM_BUG_ON(expected && PageTransHuge(page)); |
46f65ec15
|
564 |
|
800d8c63b
|
565 |
page_ref_add(page, nr); |
b065b4321
|
566 567 568 569 |
page->mapping = mapping; page->index = index; spin_lock_irq(&mapping->tree_lock); |
800d8c63b
|
570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 |
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
|
591 |
error = radix_tree_insert(&mapping->page_tree, index, page); |
800d8c63b
|
592 |
} else { |
b065b4321
|
593 594 |
error = shmem_radix_tree_replace(mapping, index, expected, page); |
800d8c63b
|
595 |
} |
46f65ec15
|
596 |
if (!error) { |
800d8c63b
|
597 598 |
mapping->nrpages += nr; if (PageTransHuge(page)) |
11fb99898
|
599 600 601 |
__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
|
602 603 604 605 |
spin_unlock_irq(&mapping->tree_lock); } else { page->mapping = NULL; spin_unlock_irq(&mapping->tree_lock); |
800d8c63b
|
606 |
page_ref_sub(page, nr); |
46f65ec15
|
607 |
} |
46f65ec15
|
608 609 610 611 |
return error; } /* |
6922c0c7a
|
612 613 614 615 616 617 |
* 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
|
618 |
VM_BUG_ON_PAGE(PageCompound(page), page); |
6922c0c7a
|
619 620 621 622 |
spin_lock_irq(&mapping->tree_lock); error = shmem_radix_tree_replace(mapping, page->index, page, radswap); page->mapping = NULL; mapping->nrpages--; |
11fb99898
|
623 624 |
__dec_node_page_state(page, NR_FILE_PAGES); __dec_node_page_state(page, NR_SHMEM); |
6922c0c7a
|
625 |
spin_unlock_irq(&mapping->tree_lock); |
09cbfeaf1
|
626 |
put_page(page); |
6922c0c7a
|
627 628 629 630 |
BUG_ON(error); } /* |
7a5d0fbb2
|
631 632 633 634 635 |
* 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
|
636 |
void *old; |
7a5d0fbb2
|
637 638 |
spin_lock_irq(&mapping->tree_lock); |
6dbaf22ce
|
639 |
old = radix_tree_delete_item(&mapping->page_tree, index, radswap); |
7a5d0fbb2
|
640 |
spin_unlock_irq(&mapping->tree_lock); |
6dbaf22ce
|
641 642 643 644 |
if (old != radswap) return -ENOENT; free_swap_and_cache(radix_to_swp_entry(radswap)); return 0; |
7a5d0fbb2
|
645 646 647 |
} /* |
6a15a3709
|
648 |
* Determine (in bytes) how many of the shmem object's pages mapped by the |
48131e03c
|
649 |
* given offsets are swapped out. |
6a15a3709
|
650 651 652 653 |
* * 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
|
654 655 |
unsigned long shmem_partial_swap_usage(struct address_space *mapping, pgoff_t start, pgoff_t end) |
6a15a3709
|
656 |
{ |
6a15a3709
|
657 658 659 |
struct radix_tree_iter iter; void **slot; struct page *page; |
48131e03c
|
660 |
unsigned long swapped = 0; |
6a15a3709
|
661 662 |
rcu_read_lock(); |
6a15a3709
|
663 664 665 666 667 |
radix_tree_for_each_slot(slot, &mapping->page_tree, &iter, start) { if (iter.index >= end) break; page = radix_tree_deref_slot(slot); |
2cf938aae
|
668 669 670 671 |
if (radix_tree_deref_retry(page)) { slot = radix_tree_iter_retry(&iter); continue; } |
6a15a3709
|
672 673 674 675 676 |
if (radix_tree_exceptional_entry(page)) swapped++; if (need_resched()) { |
148deab22
|
677 |
slot = radix_tree_iter_resume(slot, &iter); |
6a15a3709
|
678 |
cond_resched_rcu(); |
6a15a3709
|
679 680 681 682 683 684 685 686 687 |
} } rcu_read_unlock(); return swapped << PAGE_SHIFT; } /* |
48131e03c
|
688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 |
* 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
|
722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 |
* 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
|
739 740 |
pvec.nr = find_get_entries(mapping, index, PAGEVEC_SIZE, pvec.pages, indices); |
245132643
|
741 742 743 |
if (!pvec.nr) break; index = indices[pvec.nr - 1] + 1; |
0cd6144aa
|
744 |
pagevec_remove_exceptionals(&pvec); |
245132643
|
745 746 747 748 |
check_move_unevictable_pages(pvec.pages, pvec.nr); pagevec_release(&pvec); cond_resched(); } |
7a5d0fbb2
|
749 750 751 752 |
} /* * Remove range of pages and swap entries from radix tree, and free them. |
1635f6a74
|
753 |
* If !unfalloc, truncate or punch hole; if unfalloc, undo failed fallocate. |
7a5d0fbb2
|
754 |
*/ |
1635f6a74
|
755 756 |
static void shmem_undo_range(struct inode *inode, loff_t lstart, loff_t lend, bool unfalloc) |
1da177e4c
|
757 |
{ |
285b2c4fd
|
758 |
struct address_space *mapping = inode->i_mapping; |
1da177e4c
|
759 |
struct shmem_inode_info *info = SHMEM_I(inode); |
09cbfeaf1
|
760 761 762 763 |
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
|
764 |
struct pagevec pvec; |
7a5d0fbb2
|
765 766 |
pgoff_t indices[PAGEVEC_SIZE]; long nr_swaps_freed = 0; |
285b2c4fd
|
767 |
pgoff_t index; |
bda97eab0
|
768 |
int i; |
83e4fa9c1
|
769 770 |
if (lend == -1) end = -1; /* unsigned, so actually very big */ |
bda97eab0
|
771 772 773 |
pagevec_init(&pvec, 0); index = start; |
83e4fa9c1
|
774 |
while (index < end) { |
0cd6144aa
|
775 776 777 |
pvec.nr = find_get_entries(mapping, index, min(end - index, (pgoff_t)PAGEVEC_SIZE), pvec.pages, indices); |
7a5d0fbb2
|
778 779 |
if (!pvec.nr) break; |
bda97eab0
|
780 781 |
for (i = 0; i < pagevec_count(&pvec); i++) { struct page *page = pvec.pages[i]; |
7a5d0fbb2
|
782 |
index = indices[i]; |
83e4fa9c1
|
783 |
if (index >= end) |
bda97eab0
|
784 |
break; |
7a5d0fbb2
|
785 |
if (radix_tree_exceptional_entry(page)) { |
1635f6a74
|
786 787 |
if (unfalloc) continue; |
7a5d0fbb2
|
788 789 |
nr_swaps_freed += !shmem_free_swap(mapping, index, page); |
bda97eab0
|
790 |
continue; |
7a5d0fbb2
|
791 |
} |
800d8c63b
|
792 |
VM_BUG_ON_PAGE(page_to_pgoff(page) != index, page); |
7a5d0fbb2
|
793 |
if (!trylock_page(page)) |
bda97eab0
|
794 |
continue; |
800d8c63b
|
795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 |
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
|
814 |
if (!unfalloc || !PageUptodate(page)) { |
800d8c63b
|
815 816 |
VM_BUG_ON_PAGE(PageTail(page), page); if (page_mapping(page) == mapping) { |
309381fea
|
817 |
VM_BUG_ON_PAGE(PageWriteback(page), page); |
1635f6a74
|
818 819 |
truncate_inode_page(mapping, page); } |
bda97eab0
|
820 |
} |
bda97eab0
|
821 822 |
unlock_page(page); } |
0cd6144aa
|
823 |
pagevec_remove_exceptionals(&pvec); |
245132643
|
824 |
pagevec_release(&pvec); |
bda97eab0
|
825 826 827 |
cond_resched(); index++; } |
1da177e4c
|
828 |
|
83e4fa9c1
|
829 |
if (partial_start) { |
bda97eab0
|
830 |
struct page *page = NULL; |
9e18eb293
|
831 |
shmem_getpage(inode, start - 1, &page, SGP_READ); |
bda97eab0
|
832 |
if (page) { |
09cbfeaf1
|
833 |
unsigned int top = PAGE_SIZE; |
83e4fa9c1
|
834 835 836 837 838 839 840 |
if (start > end) { top = partial_end; partial_end = 0; } zero_user_segment(page, partial_start, top); set_page_dirty(page); unlock_page(page); |
09cbfeaf1
|
841 |
put_page(page); |
83e4fa9c1
|
842 843 844 845 |
} } if (partial_end) { struct page *page = NULL; |
9e18eb293
|
846 |
shmem_getpage(inode, end, &page, SGP_READ); |
83e4fa9c1
|
847 848 |
if (page) { zero_user_segment(page, 0, partial_end); |
bda97eab0
|
849 850 |
set_page_dirty(page); unlock_page(page); |
09cbfeaf1
|
851 |
put_page(page); |
bda97eab0
|
852 853 |
} } |
83e4fa9c1
|
854 855 |
if (start >= end) return; |
bda97eab0
|
856 857 |
index = start; |
b1a366500
|
858 |
while (index < end) { |
bda97eab0
|
859 |
cond_resched(); |
0cd6144aa
|
860 861 |
pvec.nr = find_get_entries(mapping, index, |
83e4fa9c1
|
862 |
min(end - index, (pgoff_t)PAGEVEC_SIZE), |
0cd6144aa
|
863 |
pvec.pages, indices); |
7a5d0fbb2
|
864 |
if (!pvec.nr) { |
b1a366500
|
865 866 |
/* If all gone or hole-punch or unfalloc, we're done */ if (index == start || end != -1) |
bda97eab0
|
867 |
break; |
b1a366500
|
868 |
/* But if truncating, restart to make sure all gone */ |
bda97eab0
|
869 870 871 |
index = start; continue; } |
bda97eab0
|
872 873 |
for (i = 0; i < pagevec_count(&pvec); i++) { struct page *page = pvec.pages[i]; |
7a5d0fbb2
|
874 |
index = indices[i]; |
83e4fa9c1
|
875 |
if (index >= end) |
bda97eab0
|
876 |
break; |
7a5d0fbb2
|
877 |
if (radix_tree_exceptional_entry(page)) { |
1635f6a74
|
878 879 |
if (unfalloc) continue; |
b1a366500
|
880 881 882 883 884 885 |
if (shmem_free_swap(mapping, index, page)) { /* Swap was replaced by page: retry */ index--; break; } nr_swaps_freed++; |
7a5d0fbb2
|
886 887 |
continue; } |
bda97eab0
|
888 |
lock_page(page); |
800d8c63b
|
889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 |
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
|
915 |
if (!unfalloc || !PageUptodate(page)) { |
800d8c63b
|
916 917 |
VM_BUG_ON_PAGE(PageTail(page), page); if (page_mapping(page) == mapping) { |
309381fea
|
918 |
VM_BUG_ON_PAGE(PageWriteback(page), page); |
1635f6a74
|
919 |
truncate_inode_page(mapping, page); |
b1a366500
|
920 921 922 923 924 |
} else { /* Page was replaced by swap: retry */ unlock_page(page); index--; break; |
1635f6a74
|
925 |
} |
7a5d0fbb2
|
926 |
} |
bda97eab0
|
927 928 |
unlock_page(page); } |
0cd6144aa
|
929 |
pagevec_remove_exceptionals(&pvec); |
245132643
|
930 |
pagevec_release(&pvec); |
bda97eab0
|
931 932 |
index++; } |
94c1e62df
|
933 |
|
4595ef88d
|
934 |
spin_lock_irq(&info->lock); |
7a5d0fbb2
|
935 |
info->swapped -= nr_swaps_freed; |
1da177e4c
|
936 |
shmem_recalc_inode(inode); |
4595ef88d
|
937 |
spin_unlock_irq(&info->lock); |
1635f6a74
|
938 |
} |
1da177e4c
|
939 |
|
1635f6a74
|
940 941 942 |
void shmem_truncate_range(struct inode *inode, loff_t lstart, loff_t lend) { shmem_undo_range(inode, lstart, lend, false); |
078cd8279
|
943 |
inode->i_ctime = inode->i_mtime = current_time(inode); |
1da177e4c
|
944 |
} |
94c1e62df
|
945 |
EXPORT_SYMBOL_GPL(shmem_truncate_range); |
1da177e4c
|
946 |
|
a528d35e8
|
947 948 |
static int shmem_getattr(const struct path *path, struct kstat *stat, u32 request_mask, unsigned int query_flags) |
44a30220b
|
949 |
{ |
a528d35e8
|
950 |
struct inode *inode = path->dentry->d_inode; |
44a30220b
|
951 |
struct shmem_inode_info *info = SHMEM_I(inode); |
d0424c429
|
952 |
if (info->alloced - info->swapped != inode->i_mapping->nrpages) { |
4595ef88d
|
953 |
spin_lock_irq(&info->lock); |
d0424c429
|
954 |
shmem_recalc_inode(inode); |
4595ef88d
|
955 |
spin_unlock_irq(&info->lock); |
d0424c429
|
956 |
} |
44a30220b
|
957 |
generic_fillattr(inode, stat); |
44a30220b
|
958 959 |
return 0; } |
94c1e62df
|
960 |
static int shmem_setattr(struct dentry *dentry, struct iattr *attr) |
1da177e4c
|
961 |
{ |
75c3cfa85
|
962 |
struct inode *inode = d_inode(dentry); |
40e041a2c
|
963 |
struct shmem_inode_info *info = SHMEM_I(inode); |
779750d20
|
964 |
struct shmem_sb_info *sbinfo = SHMEM_SB(inode->i_sb); |
1da177e4c
|
965 |
int error; |
31051c85b
|
966 |
error = setattr_prepare(dentry, attr); |
db78b877f
|
967 968 |
if (error) return error; |
94c1e62df
|
969 970 971 |
if (S_ISREG(inode->i_mode) && (attr->ia_valid & ATTR_SIZE)) { loff_t oldsize = inode->i_size; loff_t newsize = attr->ia_size; |
3889e6e76
|
972 |
|
40e041a2c
|
973 974 975 976 |
/* protected by i_mutex */ if ((newsize < oldsize && (info->seals & F_SEAL_SHRINK)) || (newsize > oldsize && (info->seals & F_SEAL_GROW))) return -EPERM; |
94c1e62df
|
977 |
if (newsize != oldsize) { |
771425179
|
978 979 980 981 |
error = shmem_reacct_size(SHMEM_I(inode)->flags, oldsize, newsize); if (error) return error; |
94c1e62df
|
982 |
i_size_write(inode, newsize); |
078cd8279
|
983 |
inode->i_ctime = inode->i_mtime = current_time(inode); |
94c1e62df
|
984 |
} |
afa2db2fb
|
985 |
if (newsize <= oldsize) { |
94c1e62df
|
986 |
loff_t holebegin = round_up(newsize, PAGE_SIZE); |
d0424c429
|
987 988 989 990 991 992 |
if (oldsize > holebegin) unmap_mapping_range(inode->i_mapping, holebegin, 0, 1); if (info->alloced) shmem_truncate_range(inode, newsize, (loff_t)-1); |
94c1e62df
|
993 |
/* unmap again to remove racily COWed private pages */ |
d0424c429
|
994 995 996 |
if (oldsize > holebegin) unmap_mapping_range(inode->i_mapping, holebegin, 0, 1); |
779750d20
|
997 998 999 1000 1001 1002 1003 |
/* * 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); |
d041353dc
|
1004 1005 1006 1007 1008 |
/* * _careful to defend against unlocked access to * ->shrink_list in shmem_unused_huge_shrink() */ if (list_empty_careful(&info->shrinklist)) { |
779750d20
|
1009 1010 1011 1012 1013 1014 |
list_add_tail(&info->shrinklist, &sbinfo->shrinklist); sbinfo->shrinklist_len++; } spin_unlock(&sbinfo->shrinklist_lock); } |
94c1e62df
|
1015 |
} |
1da177e4c
|
1016 |
} |
db78b877f
|
1017 |
setattr_copy(inode, attr); |
db78b877f
|
1018 |
if (attr->ia_valid & ATTR_MODE) |
feda821e7
|
1019 |
error = posix_acl_chmod(inode, inode->i_mode); |
1da177e4c
|
1020 1021 |
return error; } |
1f895f75d
|
1022 |
static void shmem_evict_inode(struct inode *inode) |
1da177e4c
|
1023 |
{ |
1da177e4c
|
1024 |
struct shmem_inode_info *info = SHMEM_I(inode); |
779750d20
|
1025 |
struct shmem_sb_info *sbinfo = SHMEM_SB(inode->i_sb); |
1da177e4c
|
1026 |
|
3889e6e76
|
1027 |
if (inode->i_mapping->a_ops == &shmem_aops) { |
1da177e4c
|
1028 1029 |
shmem_unacct_size(info->flags, inode->i_size); inode->i_size = 0; |
3889e6e76
|
1030 |
shmem_truncate_range(inode, 0, (loff_t)-1); |
779750d20
|
1031 1032 1033 1034 1035 1036 1037 1038 |
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
|
1039 |
if (!list_empty(&info->swaplist)) { |
cb5f7b9a4
|
1040 |
mutex_lock(&shmem_swaplist_mutex); |
1da177e4c
|
1041 |
list_del_init(&info->swaplist); |
cb5f7b9a4
|
1042 |
mutex_unlock(&shmem_swaplist_mutex); |
1da177e4c
|
1043 |
} |
3ed47db34
|
1044 |
} |
b09e0fa4b
|
1045 |
|
38f386574
|
1046 |
simple_xattrs_free(&info->xattrs); |
0f3c42f52
|
1047 |
WARN_ON(inode->i_blocks); |
5b04c6890
|
1048 |
shmem_free_inode(inode->i_sb); |
dbd5768f8
|
1049 |
clear_inode(inode); |
1da177e4c
|
1050 |
} |
478922e2b
|
1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 |
static unsigned long find_swap_entry(struct radix_tree_root *root, void *item) { struct radix_tree_iter iter; void **slot; unsigned long found = -1; unsigned int checked = 0; rcu_read_lock(); radix_tree_for_each_slot(slot, root, &iter, 0) { if (*slot == item) { found = iter.index; break; } checked++; if ((checked % 4096) != 0) continue; slot = radix_tree_iter_resume(slot, &iter); cond_resched_rcu(); } rcu_read_unlock(); return found; } |
46f65ec15
|
1074 1075 1076 |
/* * If swap found in inode, free it and move page from swapcache to filecache. */ |
41ffe5d5c
|
1077 |
static int shmem_unuse_inode(struct shmem_inode_info *info, |
bde05d1cc
|
1078 |
swp_entry_t swap, struct page **pagep) |
1da177e4c
|
1079 |
{ |
285b2c4fd
|
1080 |
struct address_space *mapping = info->vfs_inode.i_mapping; |
46f65ec15
|
1081 |
void *radswap; |
41ffe5d5c
|
1082 |
pgoff_t index; |
bde05d1cc
|
1083 1084 |
gfp_t gfp; int error = 0; |
1da177e4c
|
1085 |
|
46f65ec15
|
1086 |
radswap = swp_to_radix_entry(swap); |
478922e2b
|
1087 |
index = find_swap_entry(&mapping->page_tree, radswap); |
46f65ec15
|
1088 |
if (index == -1) |
00501b531
|
1089 |
return -EAGAIN; /* tell shmem_unuse we found nothing */ |
2e0e26c76
|
1090 |
|
1b1b32f2c
|
1091 1092 |
/* * Move _head_ to start search for next from here. |
1f895f75d
|
1093 |
* But be careful: shmem_evict_inode checks list_empty without taking |
1b1b32f2c
|
1094 |
* mutex, and there's an instant in list_move_tail when info->swaplist |
285b2c4fd
|
1095 |
* would appear empty, if it were the only one on shmem_swaplist. |
1b1b32f2c
|
1096 1097 1098 |
*/ if (shmem_swaplist.next != &info->swaplist) list_move_tail(&shmem_swaplist, &info->swaplist); |
2e0e26c76
|
1099 |
|
bde05d1cc
|
1100 1101 1102 1103 1104 1105 1106 |
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
|
1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 |
* 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
|
1122 1123 1124 1125 |
*/ if (!page_swapcount(*pagep)) error = -ENOENT; } |
d13d14430
|
1126 |
/* |
778dd893a
|
1127 1128 1129 |
* 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
|
1130 |
*/ |
bde05d1cc
|
1131 1132 |
if (!error) error = shmem_add_to_page_cache(*pagep, mapping, index, |
fed400a18
|
1133 |
radswap); |
48f170fb7
|
1134 |
if (error != -ENOMEM) { |
46f65ec15
|
1135 1136 1137 1138 |
/* * Truncation and eviction use free_swap_and_cache(), which * only does trylock page: if we raced, best clean up here. */ |
bde05d1cc
|
1139 1140 |
delete_from_swap_cache(*pagep); set_page_dirty(*pagep); |
46f65ec15
|
1141 |
if (!error) { |
4595ef88d
|
1142 |
spin_lock_irq(&info->lock); |
46f65ec15
|
1143 |
info->swapped--; |
4595ef88d
|
1144 |
spin_unlock_irq(&info->lock); |
46f65ec15
|
1145 1146 |
swap_free(swap); } |
1da177e4c
|
1147 |
} |
2e0e26c76
|
1148 |
return error; |
1da177e4c
|
1149 1150 1151 |
} /* |
46f65ec15
|
1152 |
* Search through swapped inodes to find and replace swap by page. |
1da177e4c
|
1153 |
*/ |
41ffe5d5c
|
1154 |
int shmem_unuse(swp_entry_t swap, struct page *page) |
1da177e4c
|
1155 |
{ |
41ffe5d5c
|
1156 |
struct list_head *this, *next; |
1da177e4c
|
1157 |
struct shmem_inode_info *info; |
00501b531
|
1158 |
struct mem_cgroup *memcg; |
bde05d1cc
|
1159 1160 1161 1162 |
int error = 0; /* * There's a faint possibility that swap page was replaced before |
0142ef6cd
|
1163 |
* caller locked it: caller will come back later with the right page. |
bde05d1cc
|
1164 |
*/ |
0142ef6cd
|
1165 |
if (unlikely(!PageSwapCache(page) || page_private(page) != swap.val)) |
bde05d1cc
|
1166 |
goto out; |
778dd893a
|
1167 1168 1169 1170 1171 |
/* * 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
|
1172 |
*/ |
f627c2f53
|
1173 1174 |
error = mem_cgroup_try_charge(page, current->mm, GFP_KERNEL, &memcg, false); |
778dd893a
|
1175 1176 |
if (error) goto out; |
46f65ec15
|
1177 |
/* No radix_tree_preload: swap entry keeps a place for page in tree */ |
00501b531
|
1178 |
error = -EAGAIN; |
1da177e4c
|
1179 |
|
cb5f7b9a4
|
1180 |
mutex_lock(&shmem_swaplist_mutex); |
41ffe5d5c
|
1181 1182 |
list_for_each_safe(this, next, &shmem_swaplist) { info = list_entry(this, struct shmem_inode_info, swaplist); |
285b2c4fd
|
1183 |
if (info->swapped) |
00501b531
|
1184 |
error = shmem_unuse_inode(info, swap, &page); |
6922c0c7a
|
1185 1186 |
else list_del_init(&info->swaplist); |
cb5f7b9a4
|
1187 |
cond_resched(); |
00501b531
|
1188 |
if (error != -EAGAIN) |
778dd893a
|
1189 |
break; |
00501b531
|
1190 |
/* found nothing in this: move on to search the next */ |
1da177e4c
|
1191 |
} |
cb5f7b9a4
|
1192 |
mutex_unlock(&shmem_swaplist_mutex); |
778dd893a
|
1193 |
|
00501b531
|
1194 1195 1196 |
if (error) { if (error != -ENOMEM) error = 0; |
f627c2f53
|
1197 |
mem_cgroup_cancel_charge(page, memcg, false); |
00501b531
|
1198 |
} else |
f627c2f53
|
1199 |
mem_cgroup_commit_charge(page, memcg, true, false); |
778dd893a
|
1200 |
out: |
aaa468653
|
1201 |
unlock_page(page); |
09cbfeaf1
|
1202 |
put_page(page); |
778dd893a
|
1203 |
return error; |
1da177e4c
|
1204 1205 1206 1207 1208 1209 1210 1211 |
} /* * 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
|
1212 |
struct address_space *mapping; |
1da177e4c
|
1213 |
struct inode *inode; |
6922c0c7a
|
1214 1215 |
swp_entry_t swap; pgoff_t index; |
1da177e4c
|
1216 |
|
800d8c63b
|
1217 |
VM_BUG_ON_PAGE(PageCompound(page), page); |
1da177e4c
|
1218 |
BUG_ON(!PageLocked(page)); |
1da177e4c
|
1219 1220 1221 1222 1223 1224 |
mapping = page->mapping; index = page->index; inode = mapping->host; info = SHMEM_I(inode); if (info->flags & VM_LOCKED) goto redirty; |
d9fe526a8
|
1225 |
if (!total_swap_pages) |
1da177e4c
|
1226 |
goto redirty; |
d9fe526a8
|
1227 |
/* |
97b713ba3
|
1228 1229 1230 1231 1232 |
* 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
|
1233 |
*/ |
48f170fb7
|
1234 1235 1236 1237 |
if (!wbc->for_reclaim) { WARN_ON_ONCE(1); /* Still happens? Tell us about it! */ goto redirty; } |
1635f6a74
|
1238 1239 1240 1241 1242 |
/* * 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
|
1243 1244 1245 1246 1247 1248 |
* * 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
|
1249 1250 |
*/ if (!PageUptodate(page)) { |
1aac14003
|
1251 1252 1253 1254 1255 |
if (inode->i_private) { struct shmem_falloc *shmem_falloc; spin_lock(&inode->i_lock); shmem_falloc = inode->i_private; if (shmem_falloc && |
8e205f779
|
1256 |
!shmem_falloc->waitq && |
1aac14003
|
1257 1258 1259 1260 1261 1262 1263 1264 1265 |
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
|
1266 1267 1268 1269 |
clear_highpage(page); flush_dcache_page(page); SetPageUptodate(page); } |
38d8b4e6b
|
1270 |
swap = get_swap_page(page); |
48f170fb7
|
1271 1272 |
if (!swap.val) goto redirty; |
d9fe526a8
|
1273 |
|
37e843511
|
1274 1275 |
if (mem_cgroup_try_charge_swap(page, swap)) goto free_swap; |
b1dea800a
|
1276 1277 |
/* * Add inode to shmem_unuse()'s list of swapped-out inodes, |
6922c0c7a
|
1278 1279 |
* if it's not already there. Do it now before the page is * moved to swap cache, when its pagelock no longer protects |
b1dea800a
|
1280 |
* the inode from eviction. But don't unlock the mutex until |
6922c0c7a
|
1281 1282 |
* we've incremented swapped, because shmem_unuse_inode() will * prune a !swapped inode from the swaplist under this mutex. |
b1dea800a
|
1283 |
*/ |
48f170fb7
|
1284 1285 1286 |
mutex_lock(&shmem_swaplist_mutex); if (list_empty(&info->swaplist)) list_add_tail(&info->swaplist, &shmem_swaplist); |
b1dea800a
|
1287 |
|
48f170fb7
|
1288 |
if (add_to_swap_cache(page, swap, GFP_ATOMIC) == 0) { |
4595ef88d
|
1289 |
spin_lock_irq(&info->lock); |
6922c0c7a
|
1290 |
shmem_recalc_inode(inode); |
267a4c76b
|
1291 |
info->swapped++; |
4595ef88d
|
1292 |
spin_unlock_irq(&info->lock); |
6922c0c7a
|
1293 |
|
267a4c76b
|
1294 1295 |
swap_shmem_alloc(swap); shmem_delete_from_page_cache(page, swp_to_radix_entry(swap)); |
6922c0c7a
|
1296 |
mutex_unlock(&shmem_swaplist_mutex); |
d9fe526a8
|
1297 |
BUG_ON(page_mapped(page)); |
9fab5619b
|
1298 |
swap_writepage(page, wbc); |
1da177e4c
|
1299 1300 |
return 0; } |
6922c0c7a
|
1301 |
mutex_unlock(&shmem_swaplist_mutex); |
37e843511
|
1302 |
free_swap: |
75f6d6d29
|
1303 |
put_swap_page(page, swap); |
1da177e4c
|
1304 1305 |
redirty: set_page_dirty(page); |
d9fe526a8
|
1306 1307 1308 1309 |
if (wbc->for_reclaim) return AOP_WRITEPAGE_ACTIVATE; /* Return with page locked */ unlock_page(page); return 0; |
1da177e4c
|
1310 |
} |
75edd345e
|
1311 |
#if defined(CONFIG_NUMA) && defined(CONFIG_TMPFS) |
71fe804b6
|
1312 |
static void shmem_show_mpol(struct seq_file *seq, struct mempolicy *mpol) |
680d794ba
|
1313 |
{ |
095f1fc4e
|
1314 |
char buffer[64]; |
680d794ba
|
1315 |
|
71fe804b6
|
1316 |
if (!mpol || mpol->mode == MPOL_DEFAULT) |
095f1fc4e
|
1317 |
return; /* show nothing */ |
680d794ba
|
1318 |
|
a7a88b237
|
1319 |
mpol_to_str(buffer, sizeof(buffer), mpol); |
095f1fc4e
|
1320 1321 |
seq_printf(seq, ",mpol=%s", buffer); |
680d794ba
|
1322 |
} |
71fe804b6
|
1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 |
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
|
1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 |
#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
|
1347 |
|
800d8c63b
|
1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 |
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
|
1364 1365 |
static struct page *shmem_swapin(swp_entry_t swap, gfp_t gfp, struct shmem_inode_info *info, pgoff_t index) |
1da177e4c
|
1366 |
{ |
1da177e4c
|
1367 |
struct vm_area_struct pvma; |
18a2f371f
|
1368 |
struct page *page; |
52cd3b074
|
1369 |
|
800d8c63b
|
1370 |
shmem_pseudo_vma_init(&pvma, info, index); |
18a2f371f
|
1371 |
page = swapin_readahead(swap, gfp, &pvma, 0); |
800d8c63b
|
1372 |
shmem_pseudo_vma_destroy(&pvma); |
18a2f371f
|
1373 |
|
800d8c63b
|
1374 1375 1376 1377 1378 1379 1380 1381 1382 |
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
|
1383 |
pgoff_t idx, hindex; |
800d8c63b
|
1384 1385 |
void __rcu **results; struct page *page; |
e496cf3d7
|
1386 |
if (!IS_ENABLED(CONFIG_TRANSPARENT_HUGE_PAGECACHE)) |
800d8c63b
|
1387 |
return NULL; |
4620a06e4
|
1388 |
hindex = round_down(index, HPAGE_PMD_NR); |
800d8c63b
|
1389 1390 1391 1392 1393 1394 1395 |
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
|
1396 |
|
800d8c63b
|
1397 1398 1399 1400 1401 1402 |
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
|
1403 |
return page; |
1da177e4c
|
1404 |
} |
02098feaa
|
1405 |
static struct page *shmem_alloc_page(gfp_t gfp, |
41ffe5d5c
|
1406 |
struct shmem_inode_info *info, pgoff_t index) |
1da177e4c
|
1407 1408 |
{ struct vm_area_struct pvma; |
18a2f371f
|
1409 |
struct page *page; |
1da177e4c
|
1410 |
|
800d8c63b
|
1411 1412 1413 1414 1415 1416 1417 1418 |
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, |
0f0796945
|
1419 |
struct inode *inode, |
800d8c63b
|
1420 1421 |
pgoff_t index, bool huge) { |
0f0796945
|
1422 |
struct shmem_inode_info *info = SHMEM_I(inode); |
800d8c63b
|
1423 1424 1425 |
struct page *page; int nr; int err = -ENOSPC; |
52cd3b074
|
1426 |
|
e496cf3d7
|
1427 |
if (!IS_ENABLED(CONFIG_TRANSPARENT_HUGE_PAGECACHE)) |
800d8c63b
|
1428 1429 |
huge = false; nr = huge ? HPAGE_PMD_NR : 1; |
0f0796945
|
1430 |
if (!shmem_inode_acct_block(inode, nr)) |
800d8c63b
|
1431 |
goto failed; |
800d8c63b
|
1432 1433 1434 1435 1436 |
if (huge) page = shmem_alloc_hugepage(gfp, info, index); else page = shmem_alloc_page(gfp, info, index); |
75edd345e
|
1437 1438 1439 |
if (page) { __SetPageLocked(page); __SetPageSwapBacked(page); |
800d8c63b
|
1440 |
return page; |
75edd345e
|
1441 |
} |
18a2f371f
|
1442 |
|
800d8c63b
|
1443 |
err = -ENOMEM; |
0f0796945
|
1444 |
shmem_inode_unacct_blocks(inode, nr); |
800d8c63b
|
1445 1446 |
failed: return ERR_PTR(err); |
1da177e4c
|
1447 |
} |
71fe804b6
|
1448 |
|
1da177e4c
|
1449 |
/* |
bde05d1cc
|
1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 |
* 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; |
7c1ba1a1b
|
1471 |
swp_entry_t entry; |
bde05d1cc
|
1472 1473 1474 1475 |
pgoff_t swap_index; int error; oldpage = *pagep; |
7c1ba1a1b
|
1476 1477 |
entry.val = page_private(oldpage); swap_index = swp_offset(entry); |
bde05d1cc
|
1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 |
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
|
1488 |
|
09cbfeaf1
|
1489 |
get_page(newpage); |
bde05d1cc
|
1490 |
copy_highpage(newpage, oldpage); |
0142ef6cd
|
1491 |
flush_dcache_page(newpage); |
bde05d1cc
|
1492 |
|
9956edf37
|
1493 1494 |
__SetPageLocked(newpage); __SetPageSwapBacked(newpage); |
bde05d1cc
|
1495 |
SetPageUptodate(newpage); |
7c1ba1a1b
|
1496 |
set_page_private(newpage, entry.val); |
bde05d1cc
|
1497 1498 1499 1500 1501 1502 1503 1504 1505 |
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
|
1506 |
if (!error) { |
11fb99898
|
1507 1508 |
__inc_node_page_state(newpage, NR_FILE_PAGES); __dec_node_page_state(oldpage, NR_FILE_PAGES); |
0142ef6cd
|
1509 |
} |
bde05d1cc
|
1510 |
spin_unlock_irq(&swap_mapping->tree_lock); |
bde05d1cc
|
1511 |
|
0142ef6cd
|
1512 1513 1514 1515 1516 1517 1518 1519 |
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
|
1520 |
mem_cgroup_migrate(oldpage, newpage); |
0142ef6cd
|
1521 1522 1523 |
lru_cache_add_anon(newpage); *pagep = newpage; } |
bde05d1cc
|
1524 1525 1526 1527 1528 |
ClearPageSwapCache(oldpage); set_page_private(oldpage, 0); unlock_page(oldpage); |
09cbfeaf1
|
1529 1530 |
put_page(oldpage); put_page(oldpage); |
0142ef6cd
|
1531 |
return error; |
bde05d1cc
|
1532 1533 1534 |
} /* |
68da9f055
|
1535 |
* shmem_getpage_gfp - find page in cache, or get from swap, or allocate |
1da177e4c
|
1536 1537 1538 |
* * 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
|
1539 1540 1541 1542 |
* 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
|
1543 |
*/ |
41ffe5d5c
|
1544 |
static int shmem_getpage_gfp(struct inode *inode, pgoff_t index, |
9e18eb293
|
1545 |
struct page **pagep, enum sgp_type sgp, gfp_t gfp, |
cfda05267
|
1546 |
struct vm_area_struct *vma, struct vm_fault *vmf, int *fault_type) |
1da177e4c
|
1547 1548 |
{ struct address_space *mapping = inode->i_mapping; |
23f919d4a
|
1549 |
struct shmem_inode_info *info = SHMEM_I(inode); |
1da177e4c
|
1550 |
struct shmem_sb_info *sbinfo; |
9e18eb293
|
1551 |
struct mm_struct *charge_mm; |
00501b531
|
1552 |
struct mem_cgroup *memcg; |
27ab70062
|
1553 |
struct page *page; |
1da177e4c
|
1554 |
swp_entry_t swap; |
657e3038c
|
1555 |
enum sgp_type sgp_huge = sgp; |
800d8c63b
|
1556 |
pgoff_t hindex = index; |
1da177e4c
|
1557 |
int error; |
54af60421
|
1558 |
int once = 0; |
1635f6a74
|
1559 |
int alloced = 0; |
1da177e4c
|
1560 |
|
09cbfeaf1
|
1561 |
if (index > (MAX_LFS_FILESIZE >> PAGE_SHIFT)) |
1da177e4c
|
1562 |
return -EFBIG; |
657e3038c
|
1563 1564 |
if (sgp == SGP_NOHUGE || sgp == SGP_HUGE) sgp = SGP_CACHE; |
1da177e4c
|
1565 |
repeat: |
54af60421
|
1566 |
swap.val = 0; |
0cd6144aa
|
1567 |
page = find_lock_entry(mapping, index); |
54af60421
|
1568 1569 1570 1571 |
if (radix_tree_exceptional_entry(page)) { swap = radix_to_swp_entry(page); page = NULL; } |
75edd345e
|
1572 |
if (sgp <= SGP_CACHE && |
09cbfeaf1
|
1573 |
((loff_t)index << PAGE_SHIFT) >= i_size_read(inode)) { |
54af60421
|
1574 |
error = -EINVAL; |
267a4c76b
|
1575 |
goto unlock; |
54af60421
|
1576 |
} |
66d2f4d28
|
1577 1578 |
if (page && sgp == SGP_WRITE) mark_page_accessed(page); |
1635f6a74
|
1579 1580 1581 1582 1583 |
/* fallocated page? */ if (page && !PageUptodate(page)) { if (sgp != SGP_READ) goto clear; unlock_page(page); |
09cbfeaf1
|
1584 |
put_page(page); |
1635f6a74
|
1585 1586 |
page = NULL; } |
54af60421
|
1587 |
if (page || (sgp == SGP_READ && !swap.val)) { |
54af60421
|
1588 1589 |
*pagep = page; return 0; |
27ab70062
|
1590 1591 1592 |
} /* |
54af60421
|
1593 1594 |
* Fast cache lookup did not find it: * bring it back from swap or allocate. |
27ab70062
|
1595 |
*/ |
54af60421
|
1596 |
sbinfo = SHMEM_SB(inode->i_sb); |
cfda05267
|
1597 |
charge_mm = vma ? vma->vm_mm : current->mm; |
1da177e4c
|
1598 |
|
1da177e4c
|
1599 1600 |
if (swap.val) { /* Look it up and read it in.. */ |
ec560175c
|
1601 |
page = lookup_swap_cache(swap, NULL, 0); |
27ab70062
|
1602 |
if (!page) { |
9e18eb293
|
1603 1604 |
/* Or update major stats only when swapin succeeds?? */ if (fault_type) { |
68da9f055
|
1605 |
*fault_type |= VM_FAULT_MAJOR; |
9e18eb293
|
1606 |
count_vm_event(PGMAJFAULT); |
2262185c5
|
1607 |
count_memcg_event_mm(charge_mm, PGMAJFAULT); |
9e18eb293
|
1608 1609 |
} /* Here we actually start the io */ |
41ffe5d5c
|
1610 |
page = shmem_swapin(swap, gfp, info, index); |
27ab70062
|
1611 |
if (!page) { |
54af60421
|
1612 1613 |
error = -ENOMEM; goto failed; |
1da177e4c
|
1614 |
} |
1da177e4c
|
1615 1616 1617 |
} /* We have to do this with page locked to prevent races */ |
54af60421
|
1618 |
lock_page(page); |
0142ef6cd
|
1619 |
if (!PageSwapCache(page) || page_private(page) != swap.val || |
d18992286
|
1620 |
!shmem_confirm_swap(mapping, index, swap)) { |
bde05d1cc
|
1621 |
error = -EEXIST; /* try again */ |
d18992286
|
1622 |
goto unlock; |
bde05d1cc
|
1623 |
} |
27ab70062
|
1624 |
if (!PageUptodate(page)) { |
1da177e4c
|
1625 |
error = -EIO; |
54af60421
|
1626 |
goto failed; |
1da177e4c
|
1627 |
} |
54af60421
|
1628 |
wait_on_page_writeback(page); |
bde05d1cc
|
1629 1630 1631 1632 |
if (shmem_should_replace_page(page, gfp)) { error = shmem_replace_page(&page, gfp, info, index); if (error) goto failed; |
1da177e4c
|
1633 |
} |
27ab70062
|
1634 |
|
9e18eb293
|
1635 |
error = mem_cgroup_try_charge(page, charge_mm, gfp, &memcg, |
f627c2f53
|
1636 |
false); |
d18992286
|
1637 |
if (!error) { |
aa3b18955
|
1638 |
error = shmem_add_to_page_cache(page, mapping, index, |
fed400a18
|
1639 |
swp_to_radix_entry(swap)); |
215c02bc3
|
1640 1641 1642 1643 1644 1645 1646 1647 |
/* * 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
|
1648 |
* the rest. |
215c02bc3
|
1649 1650 1651 |
* Reset swap.val? No, leave it so "failed" goes back to * "repeat": reading a hole and writing should succeed. */ |
00501b531
|
1652 |
if (error) { |
f627c2f53
|
1653 |
mem_cgroup_cancel_charge(page, memcg, false); |
215c02bc3
|
1654 |
delete_from_swap_cache(page); |
00501b531
|
1655 |
} |
d18992286
|
1656 |
} |
54af60421
|
1657 1658 |
if (error) goto failed; |
f627c2f53
|
1659 |
mem_cgroup_commit_charge(page, memcg, true, false); |
00501b531
|
1660 |
|
4595ef88d
|
1661 |
spin_lock_irq(&info->lock); |
285b2c4fd
|
1662 |
info->swapped--; |
54af60421
|
1663 |
shmem_recalc_inode(inode); |
4595ef88d
|
1664 |
spin_unlock_irq(&info->lock); |
54af60421
|
1665 |
|
66d2f4d28
|
1666 1667 |
if (sgp == SGP_WRITE) mark_page_accessed(page); |
54af60421
|
1668 |
delete_from_swap_cache(page); |
27ab70062
|
1669 1670 |
set_page_dirty(page); swap_free(swap); |
54af60421
|
1671 |
} else { |
cfda05267
|
1672 1673 1674 1675 |
if (vma && userfaultfd_missing(vma)) { *fault_type = handle_userfault(vmf, VM_UFFD_MISSING); return 0; } |
800d8c63b
|
1676 1677 1678 |
/* shmem_symlink() */ if (mapping->a_ops != &shmem_aops) goto alloc_nohuge; |
657e3038c
|
1679 |
if (shmem_huge == SHMEM_HUGE_DENY || sgp_huge == SGP_NOHUGE) |
800d8c63b
|
1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 |
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
|
1696 1697 1698 |
if (sgp_huge == SGP_HUGE) goto alloc_huge; /* TODO: implement fadvise() hints */ |
800d8c63b
|
1699 |
goto alloc_nohuge; |
54af60421
|
1700 |
} |
1da177e4c
|
1701 |
|
800d8c63b
|
1702 |
alloc_huge: |
0f0796945
|
1703 |
page = shmem_alloc_and_acct_page(gfp, inode, index, true); |
800d8c63b
|
1704 |
if (IS_ERR(page)) { |
0f0796945
|
1705 |
alloc_nohuge: page = shmem_alloc_and_acct_page(gfp, inode, |
800d8c63b
|
1706 |
index, false); |
1da177e4c
|
1707 |
} |
800d8c63b
|
1708 |
if (IS_ERR(page)) { |
779750d20
|
1709 |
int retry = 5; |
800d8c63b
|
1710 1711 |
error = PTR_ERR(page); page = NULL; |
779750d20
|
1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 |
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
|
1726 1727 1728 1729 1730 1731 1732 |
goto failed; } if (PageTransHuge(page)) hindex = round_down(index, HPAGE_PMD_NR); else hindex = index; |
66d2f4d28
|
1733 |
if (sgp == SGP_WRITE) |
eb39d618f
|
1734 |
__SetPageReferenced(page); |
66d2f4d28
|
1735 |
|
9e18eb293
|
1736 |
error = mem_cgroup_try_charge(page, charge_mm, gfp, &memcg, |
800d8c63b
|
1737 |
PageTransHuge(page)); |
54af60421
|
1738 |
if (error) |
800d8c63b
|
1739 1740 1741 |
goto unacct; error = radix_tree_maybe_preload_order(gfp & GFP_RECLAIM_MASK, compound_order(page)); |
b065b4321
|
1742 |
if (!error) { |
800d8c63b
|
1743 |
error = shmem_add_to_page_cache(page, mapping, hindex, |
fed400a18
|
1744 |
NULL); |
b065b4321
|
1745 1746 1747 |
radix_tree_preload_end(); } if (error) { |
800d8c63b
|
1748 1749 1750 |
mem_cgroup_cancel_charge(page, memcg, PageTransHuge(page)); goto unacct; |
b065b4321
|
1751 |
} |
800d8c63b
|
1752 1753 |
mem_cgroup_commit_charge(page, memcg, false, PageTransHuge(page)); |
54af60421
|
1754 |
lru_cache_add_anon(page); |
4595ef88d
|
1755 |
spin_lock_irq(&info->lock); |
800d8c63b
|
1756 1757 |
info->alloced += 1 << compound_order(page); inode->i_blocks += BLOCKS_PER_PAGE << compound_order(page); |
54af60421
|
1758 |
shmem_recalc_inode(inode); |
4595ef88d
|
1759 |
spin_unlock_irq(&info->lock); |
1635f6a74
|
1760 |
alloced = true; |
54af60421
|
1761 |
|
779750d20
|
1762 1763 1764 1765 1766 1767 1768 1769 |
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); |
d041353dc
|
1770 1771 1772 1773 1774 |
/* * _careful to defend against unlocked access to * ->shrink_list in shmem_unused_huge_shrink() */ if (list_empty_careful(&info->shrinklist)) { |
779750d20
|
1775 1776 1777 1778 1779 1780 |
list_add_tail(&info->shrinklist, &sbinfo->shrinklist); sbinfo->shrinklist_len++; } spin_unlock(&sbinfo->shrinklist_lock); } |
ec9516fbc
|
1781 |
/* |
1635f6a74
|
1782 1783 1784 1785 1786 1787 1788 1789 1790 |
* 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
|
1791 |
*/ |
800d8c63b
|
1792 1793 1794 1795 1796 1797 1798 1799 1800 |
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
|
1801 |
} |
1da177e4c
|
1802 |
} |
bde05d1cc
|
1803 |
|
54af60421
|
1804 |
/* Perhaps the file has been truncated since we checked */ |
75edd345e
|
1805 |
if (sgp <= SGP_CACHE && |
09cbfeaf1
|
1806 |
((loff_t)index << PAGE_SHIFT) >= i_size_read(inode)) { |
267a4c76b
|
1807 1808 1809 |
if (alloced) { ClearPageDirty(page); delete_from_page_cache(page); |
4595ef88d
|
1810 |
spin_lock_irq(&info->lock); |
267a4c76b
|
1811 |
shmem_recalc_inode(inode); |
4595ef88d
|
1812 |
spin_unlock_irq(&info->lock); |
267a4c76b
|
1813 |
} |
54af60421
|
1814 |
error = -EINVAL; |
267a4c76b
|
1815 |
goto unlock; |
e83c32e8f
|
1816 |
} |
800d8c63b
|
1817 |
*pagep = page + index - hindex; |
54af60421
|
1818 |
return 0; |
1da177e4c
|
1819 |
|
59a16ead5
|
1820 |
/* |
54af60421
|
1821 |
* Error recovery. |
59a16ead5
|
1822 |
*/ |
54af60421
|
1823 |
unacct: |
0f0796945
|
1824 |
shmem_inode_unacct_blocks(inode, 1 << compound_order(page)); |
800d8c63b
|
1825 1826 1827 1828 1829 1830 |
if (PageTransHuge(page)) { unlock_page(page); put_page(page); goto alloc_nohuge; } |
54af60421
|
1831 |
failed: |
267a4c76b
|
1832 |
if (swap.val && !shmem_confirm_swap(mapping, index, swap)) |
d18992286
|
1833 1834 |
error = -EEXIST; unlock: |
27ab70062
|
1835 |
if (page) { |
54af60421
|
1836 |
unlock_page(page); |
09cbfeaf1
|
1837 |
put_page(page); |
54af60421
|
1838 1839 |
} if (error == -ENOSPC && !once++) { |
4595ef88d
|
1840 |
spin_lock_irq(&info->lock); |
54af60421
|
1841 |
shmem_recalc_inode(inode); |
4595ef88d
|
1842 |
spin_unlock_irq(&info->lock); |
27ab70062
|
1843 |
goto repeat; |
ff36b8016
|
1844 |
} |
d18992286
|
1845 |
if (error == -EEXIST) /* from above or from radix_tree_insert */ |
54af60421
|
1846 1847 |
goto repeat; return error; |
1da177e4c
|
1848 |
} |
10d20bd25
|
1849 1850 1851 1852 1853 |
/* * This is like autoremove_wake_function, but it removes the wait queue * entry unconditionally - even if something else had already woken the * target. */ |
ac6424b98
|
1854 |
static int synchronous_wake_function(wait_queue_entry_t *wait, unsigned mode, int sync, void *key) |
10d20bd25
|
1855 1856 |
{ int ret = default_wake_function(wait, mode, sync, key); |
2055da973
|
1857 |
list_del_init(&wait->entry); |
10d20bd25
|
1858 1859 |
return ret; } |
11bac8000
|
1860 |
static int shmem_fault(struct vm_fault *vmf) |
1da177e4c
|
1861 |
{ |
11bac8000
|
1862 |
struct vm_area_struct *vma = vmf->vma; |
496ad9aa8
|
1863 |
struct inode *inode = file_inode(vma->vm_file); |
9e18eb293
|
1864 |
gfp_t gfp = mapping_gfp_mask(inode->i_mapping); |
657e3038c
|
1865 |
enum sgp_type sgp; |
1da177e4c
|
1866 |
int error; |
68da9f055
|
1867 |
int ret = VM_FAULT_LOCKED; |
1da177e4c
|
1868 |
|
f00cdc6df
|
1869 1870 1871 1872 |
/* * 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
|
1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 |
* 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
|
1885 1886 1887 1888 1889 1890 |
*/ if (unlikely(inode->i_private)) { struct shmem_falloc *shmem_falloc; spin_lock(&inode->i_lock); shmem_falloc = inode->i_private; |
8e205f779
|
1891 1892 1893 1894 1895 |
if (shmem_falloc && shmem_falloc->waitq && vmf->pgoff >= shmem_falloc->start && vmf->pgoff < shmem_falloc->next) { wait_queue_head_t *shmem_falloc_waitq; |
10d20bd25
|
1896 |
DEFINE_WAIT_FUNC(shmem_fault_wait, synchronous_wake_function); |
8e205f779
|
1897 1898 |
ret = VM_FAULT_NOPAGE; |
f00cdc6df
|
1899 1900 |
if ((vmf->flags & FAULT_FLAG_ALLOW_RETRY) && !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT)) { |
8e205f779
|
1901 |
/* It's polite to up mmap_sem if we can */ |
f00cdc6df
|
1902 |
up_read(&vma->vm_mm->mmap_sem); |
8e205f779
|
1903 |
ret = VM_FAULT_RETRY; |
f00cdc6df
|
1904 |
} |
8e205f779
|
1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 |
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
|
1923 |
} |
8e205f779
|
1924 |
spin_unlock(&inode->i_lock); |
f00cdc6df
|
1925 |
} |
657e3038c
|
1926 |
sgp = SGP_CACHE; |
186003323
|
1927 1928 1929 |
if ((vma->vm_flags & VM_NOHUGEPAGE) || test_bit(MMF_DISABLE_THP, &vma->vm_mm->flags)) |
657e3038c
|
1930 |
sgp = SGP_NOHUGE; |
186003323
|
1931 1932 |
else if (vma->vm_flags & VM_HUGEPAGE) sgp = SGP_HUGE; |
657e3038c
|
1933 1934 |
error = shmem_getpage_gfp(inode, vmf->pgoff, &vmf->page, sgp, |
cfda05267
|
1935 |
gfp, vma, vmf, &ret); |
d0217ac04
|
1936 1937 |
if (error) return ((error == -ENOMEM) ? VM_FAULT_OOM : VM_FAULT_SIGBUS); |
68da9f055
|
1938 |
return ret; |
1da177e4c
|
1939 |
} |
c01d5b300
|
1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 |
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
|
1957 |
if (!IS_ENABLED(CONFIG_TRANSPARENT_HUGE_PAGECACHE)) |
c01d5b300
|
1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 |
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
|
1995 |
if (SHMEM_SB(sb)->huge == SHMEM_HUGE_NEVER) |
c01d5b300
|
1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 |
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
|
2026 |
#ifdef CONFIG_NUMA |
41ffe5d5c
|
2027 |
static int shmem_set_policy(struct vm_area_struct *vma, struct mempolicy *mpol) |
1da177e4c
|
2028 |
{ |
496ad9aa8
|
2029 |
struct inode *inode = file_inode(vma->vm_file); |
41ffe5d5c
|
2030 |
return mpol_set_shared_policy(&SHMEM_I(inode)->policy, vma, mpol); |
1da177e4c
|
2031 |
} |
d8dc74f21
|
2032 2033 |
static struct mempolicy *shmem_get_policy(struct vm_area_struct *vma, unsigned long addr) |
1da177e4c
|
2034 |
{ |
496ad9aa8
|
2035 |
struct inode *inode = file_inode(vma->vm_file); |
41ffe5d5c
|
2036 |
pgoff_t index; |
1da177e4c
|
2037 |
|
41ffe5d5c
|
2038 2039 |
index = ((addr - vma->vm_start) >> PAGE_SHIFT) + vma->vm_pgoff; return mpol_shared_policy_lookup(&SHMEM_I(inode)->policy, index); |
1da177e4c
|
2040 2041 2042 2043 2044 |
} #endif int shmem_lock(struct file *file, int lock, struct user_struct *user) { |
496ad9aa8
|
2045 |
struct inode *inode = file_inode(file); |
1da177e4c
|
2046 2047 |
struct shmem_inode_info *info = SHMEM_I(inode); int retval = -ENOMEM; |
4595ef88d
|
2048 |
spin_lock_irq(&info->lock); |
1da177e4c
|
2049 2050 2051 2052 |
if (lock && !(info->flags & VM_LOCKED)) { if (!user_shm_lock(inode->i_size, user)) goto out_nomem; info->flags |= VM_LOCKED; |
89e004ea5
|
2053 |
mapping_set_unevictable(file->f_mapping); |
1da177e4c
|
2054 2055 2056 2057 |
} if (!lock && (info->flags & VM_LOCKED) && user) { user_shm_unlock(inode->i_size, user); info->flags &= ~VM_LOCKED; |
89e004ea5
|
2058 |
mapping_clear_unevictable(file->f_mapping); |
1da177e4c
|
2059 2060 |
} retval = 0; |
89e004ea5
|
2061 |
|
1da177e4c
|
2062 |
out_nomem: |
4595ef88d
|
2063 |
spin_unlock_irq(&info->lock); |
1da177e4c
|
2064 2065 |
return retval; } |
9b83a6a85
|
2066 |
static int shmem_mmap(struct file *file, struct vm_area_struct *vma) |
1da177e4c
|
2067 2068 2069 |
{ file_accessed(file); vma->vm_ops = &shmem_vm_ops; |
e496cf3d7
|
2070 |
if (IS_ENABLED(CONFIG_TRANSPARENT_HUGE_PAGECACHE) && |
f3f0e1d21
|
2071 2072 2073 2074 |
((vma->vm_start + ~HPAGE_PMD_MASK) & HPAGE_PMD_MASK) < (vma->vm_end & HPAGE_PMD_MASK)) { khugepaged_enter(vma, vma->vm_flags); } |
1da177e4c
|
2075 2076 |
return 0; } |
454abafe9
|
2077 |
static struct inode *shmem_get_inode(struct super_block *sb, const struct inode *dir, |
09208d150
|
2078 |
umode_t mode, dev_t dev, unsigned long flags) |
1da177e4c
|
2079 2080 2081 2082 |
{ struct inode *inode; struct shmem_inode_info *info; struct shmem_sb_info *sbinfo = SHMEM_SB(sb); |
5b04c6890
|
2083 2084 |
if (shmem_reserve_inode(sb)) return NULL; |
1da177e4c
|
2085 2086 2087 |
inode = new_inode(sb); if (inode) { |
85fe4025c
|
2088 |
inode->i_ino = get_next_ino(); |
454abafe9
|
2089 |
inode_init_owner(inode, dir, mode); |
1da177e4c
|
2090 |
inode->i_blocks = 0; |
078cd8279
|
2091 |
inode->i_atime = inode->i_mtime = inode->i_ctime = current_time(inode); |
91828a405
|
2092 |
inode->i_generation = get_seconds(); |
1da177e4c
|
2093 2094 2095 |
info = SHMEM_I(inode); memset(info, 0, (char *)inode - (char *)info); spin_lock_init(&info->lock); |
40e041a2c
|
2096 |
info->seals = F_SEAL_SEAL; |
0b0a0806b
|
2097 |
info->flags = flags & VM_NORESERVE; |
779750d20
|
2098 |
INIT_LIST_HEAD(&info->shrinklist); |
1da177e4c
|
2099 |
INIT_LIST_HEAD(&info->swaplist); |
38f386574
|
2100 |
simple_xattrs_init(&info->xattrs); |
72c04902d
|
2101 |
cache_no_acl(inode); |
1da177e4c
|
2102 2103 2104 |
switch (mode & S_IFMT) { default: |
39f0247d3
|
2105 |
inode->i_op = &shmem_special_inode_operations; |
1da177e4c
|
2106 2107 2108 |
init_special_inode(inode, mode, dev); break; case S_IFREG: |
14fcc23fd
|
2109 |
inode->i_mapping->a_ops = &shmem_aops; |
1da177e4c
|
2110 2111 |
inode->i_op = &shmem_inode_operations; inode->i_fop = &shmem_file_operations; |
71fe804b6
|
2112 2113 |
mpol_shared_policy_init(&info->policy, shmem_get_sbmpol(sbinfo)); |
1da177e4c
|
2114 2115 |
break; case S_IFDIR: |
d8c76e6f4
|
2116 |
inc_nlink(inode); |
1da177e4c
|
2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 |
/* 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
|
2127 |
mpol_shared_policy_init(&info->policy, NULL); |
1da177e4c
|
2128 2129 |
break; } |
6447b34fc
|
2130 2131 |
lockdep_annotate_inode_mutex_key(inode); |
5b04c6890
|
2132 2133 |
} else shmem_free_inode(sb); |
1da177e4c
|
2134 2135 |
return inode; } |
0cd6144aa
|
2136 2137 |
bool shmem_mapping(struct address_space *mapping) { |
f8005451d
|
2138 |
return mapping->a_ops == &shmem_aops; |
0cd6144aa
|
2139 |
} |
8d1039634
|
2140 2141 2142 2143 2144 2145 2146 |
static int shmem_mfill_atomic_pte(struct mm_struct *dst_mm, pmd_t *dst_pmd, struct vm_area_struct *dst_vma, unsigned long dst_addr, unsigned long src_addr, bool zeropage, struct page **pagep) |
4c27fe4c4
|
2147 2148 2149 |
{ struct inode *inode = file_inode(dst_vma->vm_file); struct shmem_inode_info *info = SHMEM_I(inode); |
4c27fe4c4
|
2150 2151 2152 2153 2154 2155 2156 2157 2158 |
struct address_space *mapping = inode->i_mapping; gfp_t gfp = mapping_gfp_mask(mapping); pgoff_t pgoff = linear_page_index(dst_vma, dst_addr); struct mem_cgroup *memcg; spinlock_t *ptl; void *page_kaddr; struct page *page; pte_t _dst_pte, *dst_pte; int ret; |
af3edb30c
|
2159 |
pgoff_t offset, max_off; |
4c27fe4c4
|
2160 |
|
cb658a453
|
2161 |
ret = -ENOMEM; |
0f0796945
|
2162 |
if (!shmem_inode_acct_block(inode, 1)) |
cb658a453
|
2163 |
goto out; |
4c27fe4c4
|
2164 |
|
cb658a453
|
2165 |
if (!*pagep) { |
4c27fe4c4
|
2166 2167 |
page = shmem_alloc_page(gfp, info, pgoff); if (!page) |
0f0796945
|
2168 |
goto out_unacct_blocks; |
4c27fe4c4
|
2169 |
|
8d1039634
|
2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 |
if (!zeropage) { /* mcopy_atomic */ page_kaddr = kmap_atomic(page); ret = copy_from_user(page_kaddr, (const void __user *)src_addr, PAGE_SIZE); kunmap_atomic(page_kaddr); /* fallback to copy_from_user outside mmap_sem */ if (unlikely(ret)) { *pagep = page; shmem_inode_unacct_blocks(inode, 1); /* don't free the page */ |
82c5a8c0d
|
2182 |
return -ENOENT; |
8d1039634
|
2183 2184 2185 |
} } else { /* mfill_zeropage_atomic */ clear_highpage(page); |
4c27fe4c4
|
2186 2187 2188 2189 2190 |
} } else { page = *pagep; *pagep = NULL; } |
9cc90c664
|
2191 2192 2193 |
VM_BUG_ON(PageLocked(page) || PageSwapBacked(page)); __SetPageLocked(page); __SetPageSwapBacked(page); |
a425d3584
|
2194 |
__SetPageUptodate(page); |
9cc90c664
|
2195 |
|
af3edb30c
|
2196 2197 2198 2199 2200 |
ret = -EFAULT; offset = linear_page_index(dst_vma, dst_addr); max_off = DIV_ROUND_UP(i_size_read(inode), PAGE_SIZE); if (unlikely(offset >= max_off)) goto out_release; |
4c27fe4c4
|
2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 |
ret = mem_cgroup_try_charge(page, dst_mm, gfp, &memcg, false); if (ret) goto out_release; ret = radix_tree_maybe_preload(gfp & GFP_RECLAIM_MASK); if (!ret) { ret = shmem_add_to_page_cache(page, mapping, pgoff, NULL); radix_tree_preload_end(); } if (ret) goto out_release_uncharge; mem_cgroup_commit_charge(page, memcg, false, false); _dst_pte = mk_pte(page, dst_vma->vm_page_prot); if (dst_vma->vm_flags & VM_WRITE) _dst_pte = pte_mkwrite(pte_mkdirty(_dst_pte)); |
46466e23b
|
2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 |
else { /* * We don't set the pte dirty if the vma has no * VM_WRITE permission, so mark the page dirty or it * could be freed from under us. We could do it * unconditionally before unlock_page(), but doing it * only if VM_WRITE is not set is faster. */ set_page_dirty(page); } |
4c27fe4c4
|
2228 |
|
4c27fe4c4
|
2229 |
dst_pte = pte_offset_map_lock(dst_mm, dst_pmd, dst_addr, &ptl); |
af3edb30c
|
2230 2231 2232 2233 2234 2235 2236 |
ret = -EFAULT; max_off = DIV_ROUND_UP(i_size_read(inode), PAGE_SIZE); if (unlikely(offset >= max_off)) goto out_release_uncharge_unlock; ret = -EEXIST; |
4c27fe4c4
|
2237 2238 |
if (!pte_none(*dst_pte)) goto out_release_uncharge_unlock; |
4c27fe4c4
|
2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 |
lru_cache_add_anon(page); spin_lock(&info->lock); info->alloced++; inode->i_blocks += BLOCKS_PER_PAGE; shmem_recalc_inode(inode); spin_unlock(&info->lock); inc_mm_counter(dst_mm, mm_counter_file(page)); page_add_file_rmap(page, false); set_pte_at(dst_mm, dst_addr, dst_pte, _dst_pte); /* No need to invalidate - it was non-present before */ update_mmu_cache(dst_vma, dst_addr, dst_pte); |
4c27fe4c4
|
2253 |
pte_unmap_unlock(dst_pte, ptl); |
af3edb30c
|
2254 |
unlock_page(page); |
4c27fe4c4
|
2255 2256 2257 2258 2259 |
ret = 0; out: return ret; out_release_uncharge_unlock: pte_unmap_unlock(dst_pte, ptl); |
46466e23b
|
2260 |
ClearPageDirty(page); |
af3edb30c
|
2261 |
delete_from_page_cache(page); |
4c27fe4c4
|
2262 2263 2264 |
out_release_uncharge: mem_cgroup_cancel_charge(page, memcg, false); out_release: |
9cc90c664
|
2265 |
unlock_page(page); |
4c27fe4c4
|
2266 |
put_page(page); |
4c27fe4c4
|
2267 |
out_unacct_blocks: |
0f0796945
|
2268 |
shmem_inode_unacct_blocks(inode, 1); |
4c27fe4c4
|
2269 2270 |
goto out; } |
8d1039634
|
2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 |
int shmem_mcopy_atomic_pte(struct mm_struct *dst_mm, pmd_t *dst_pmd, struct vm_area_struct *dst_vma, unsigned long dst_addr, unsigned long src_addr, struct page **pagep) { return shmem_mfill_atomic_pte(dst_mm, dst_pmd, dst_vma, dst_addr, src_addr, false, pagep); } int shmem_mfill_zeropage_pte(struct mm_struct *dst_mm, pmd_t *dst_pmd, struct vm_area_struct *dst_vma, unsigned long dst_addr) { struct page *page = NULL; return shmem_mfill_atomic_pte(dst_mm, dst_pmd, dst_vma, dst_addr, 0, true, &page); } |
1da177e4c
|
2292 |
#ifdef CONFIG_TMPFS |
92e1d5be9
|
2293 |
static const struct inode_operations shmem_symlink_inode_operations; |
69f07ec93
|
2294 |
static const struct inode_operations shmem_short_symlink_operations; |
1da177e4c
|
2295 |
|
6d9d88d07
|
2296 2297 2298 2299 2300 |
#ifdef CONFIG_TMPFS_XATTR static int shmem_initxattrs(struct inode *, const struct xattr *, void *); #else #define shmem_initxattrs NULL #endif |
1da177e4c
|
2301 |
static int |
800d15a53
|
2302 2303 2304 |
shmem_write_begin(struct file *file, struct address_space *mapping, loff_t pos, unsigned len, unsigned flags, struct page **pagep, void **fsdata) |
1da177e4c
|
2305 |
{ |
800d15a53
|
2306 |
struct inode *inode = mapping->host; |
40e041a2c
|
2307 |
struct shmem_inode_info *info = SHMEM_I(inode); |
09cbfeaf1
|
2308 |
pgoff_t index = pos >> PAGE_SHIFT; |
40e041a2c
|
2309 2310 |
/* i_mutex is held by caller */ |
3f472cc97
|
2311 |
if (unlikely(info->seals & (F_SEAL_WRITE | F_SEAL_GROW))) { |
40e041a2c
|
2312 2313 2314 2315 2316 |
if (info->seals & F_SEAL_WRITE) return -EPERM; if ((info->seals & F_SEAL_GROW) && pos + len > inode->i_size) return -EPERM; } |
9e18eb293
|
2317 |
return shmem_getpage(inode, index, pagep, SGP_WRITE); |
800d15a53
|
2318 2319 2320 2321 2322 2323 2324 2325 |
} 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
|
2326 2327 |
if (pos + copied > inode->i_size) i_size_write(inode, pos + copied); |
ec9516fbc
|
2328 |
if (!PageUptodate(page)) { |
800d8c63b
|
2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 |
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
|
2340 2341 |
if (copied < PAGE_SIZE) { unsigned from = pos & (PAGE_SIZE - 1); |
ec9516fbc
|
2342 |
zero_user_segments(page, 0, from, |
09cbfeaf1
|
2343 |
from + copied, PAGE_SIZE); |
ec9516fbc
|
2344 |
} |
800d8c63b
|
2345 |
SetPageUptodate(head); |
ec9516fbc
|
2346 |
} |
800d15a53
|
2347 |
set_page_dirty(page); |
6746aff74
|
2348 |
unlock_page(page); |
09cbfeaf1
|
2349 |
put_page(page); |
800d15a53
|
2350 |
|
800d15a53
|
2351 |
return copied; |
1da177e4c
|
2352 |
} |
2ba5bbed0
|
2353 |
static ssize_t shmem_file_read_iter(struct kiocb *iocb, struct iov_iter *to) |
1da177e4c
|
2354 |
{ |
6e58e79db
|
2355 2356 |
struct file *file = iocb->ki_filp; struct inode *inode = file_inode(file); |
1da177e4c
|
2357 |
struct address_space *mapping = inode->i_mapping; |
41ffe5d5c
|
2358 2359 |
pgoff_t index; unsigned long offset; |
a0ee5ec52
|
2360 |
enum sgp_type sgp = SGP_READ; |
f7c1d0742
|
2361 |
int error = 0; |
cb66a7a1f
|
2362 |
ssize_t retval = 0; |
6e58e79db
|
2363 |
loff_t *ppos = &iocb->ki_pos; |
a0ee5ec52
|
2364 2365 2366 2367 2368 2369 |
/* * 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
|
2370 |
if (!iter_is_iovec(to)) |
75edd345e
|
2371 |
sgp = SGP_CACHE; |
1da177e4c
|
2372 |
|
09cbfeaf1
|
2373 2374 |
index = *ppos >> PAGE_SHIFT; offset = *ppos & ~PAGE_MASK; |
1da177e4c
|
2375 2376 2377 |
for (;;) { struct page *page = NULL; |
41ffe5d5c
|
2378 2379 |
pgoff_t end_index; unsigned long nr, ret; |
1da177e4c
|
2380 |
loff_t i_size = i_size_read(inode); |
09cbfeaf1
|
2381 |
end_index = i_size >> PAGE_SHIFT; |
1da177e4c
|
2382 2383 2384 |
if (index > end_index) break; if (index == end_index) { |
09cbfeaf1
|
2385 |
nr = i_size & ~PAGE_MASK; |
1da177e4c
|
2386 2387 2388 |
if (nr <= offset) break; } |
9e18eb293
|
2389 |
error = shmem_getpage(inode, index, &page, sgp); |
6e58e79db
|
2390 2391 2392 |
if (error) { if (error == -EINVAL) error = 0; |
1da177e4c
|
2393 2394 |
break; } |
75edd345e
|
2395 2396 2397 |
if (page) { if (sgp == SGP_CACHE) set_page_dirty(page); |
d3602444e
|
2398 |
unlock_page(page); |
75edd345e
|
2399 |
} |
1da177e4c
|
2400 2401 2402 |
/* * We must evaluate after, since reads (unlike writes) |
1b1dcc1b5
|
2403 |
* are called without i_mutex protection against truncate |
1da177e4c
|
2404 |
*/ |
09cbfeaf1
|
2405 |
nr = PAGE_SIZE; |
1da177e4c
|
2406 |
i_size = i_size_read(inode); |
09cbfeaf1
|
2407 |
end_index = i_size >> PAGE_SHIFT; |
1da177e4c
|
2408 |
if (index == end_index) { |
09cbfeaf1
|
2409 |
nr = i_size & ~PAGE_MASK; |
1da177e4c
|
2410 2411 |
if (nr <= offset) { if (page) |
09cbfeaf1
|
2412 |
put_page(page); |
1da177e4c
|
2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 |
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
|
2431 |
} else { |
1da177e4c
|
2432 |
page = ZERO_PAGE(0); |
09cbfeaf1
|
2433 |
get_page(page); |
b5810039a
|
2434 |
} |
1da177e4c
|
2435 2436 2437 2438 |
/* * Ok, we have the page, and it's up-to-date, so * now we can copy it to user space... |
1da177e4c
|
2439 |
*/ |
2ba5bbed0
|
2440 |
ret = copy_page_to_iter(page, offset, nr, to); |
6e58e79db
|
2441 |
retval += ret; |
1da177e4c
|
2442 |
offset += ret; |
09cbfeaf1
|
2443 2444 |
index += offset >> PAGE_SHIFT; offset &= ~PAGE_MASK; |
1da177e4c
|
2445 |
|
09cbfeaf1
|
2446 |
put_page(page); |
2ba5bbed0
|
2447 |
if (!iov_iter_count(to)) |
1da177e4c
|
2448 |
break; |
6e58e79db
|
2449 2450 2451 2452 |
if (ret < nr) { error = -EFAULT; break; } |
1da177e4c
|
2453 2454 |
cond_resched(); } |
09cbfeaf1
|
2455 |
*ppos = ((loff_t) index << PAGE_SHIFT) + offset; |
6e58e79db
|
2456 2457 |
file_accessed(file); return retval ? retval : error; |
1da177e4c
|
2458 |
} |
220f2ac91
|
2459 2460 2461 2462 |
/* * llseek SEEK_DATA or SEEK_HOLE through the radix_tree. */ static pgoff_t shmem_seek_hole_data(struct address_space *mapping, |
965c8e59c
|
2463 |
pgoff_t index, pgoff_t end, int whence) |
220f2ac91
|
2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 |
{ 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
|
2474 |
pvec.nr = find_get_entries(mapping, index, |
220f2ac91
|
2475 2476 |
pvec.nr, pvec.pages, indices); if (!pvec.nr) { |
965c8e59c
|
2477 |
if (whence == SEEK_DATA) |
220f2ac91
|
2478 2479 2480 2481 2482 |
index = end; break; } for (i = 0; i < pvec.nr; i++, index++) { if (index < indices[i]) { |
965c8e59c
|
2483 |
if (whence == SEEK_HOLE) { |
220f2ac91
|
2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 |
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
|
2495 2496 |
(page && whence == SEEK_DATA) || (!page && whence == SEEK_HOLE)) { |
220f2ac91
|
2497 2498 2499 2500 |
done = true; break; } } |
0cd6144aa
|
2501 |
pagevec_remove_exceptionals(&pvec); |
220f2ac91
|
2502 2503 2504 2505 2506 2507 |
pagevec_release(&pvec); pvec.nr = PAGEVEC_SIZE; cond_resched(); } return index; } |
965c8e59c
|
2508 |
static loff_t shmem_file_llseek(struct file *file, loff_t offset, int whence) |
220f2ac91
|
2509 2510 2511 2512 2513 |
{ struct address_space *mapping = file->f_mapping; struct inode *inode = mapping->host; pgoff_t start, end; loff_t new_offset; |
965c8e59c
|
2514 2515 |
if (whence != SEEK_DATA && whence != SEEK_HOLE) return generic_file_llseek_size(file, offset, whence, |
220f2ac91
|
2516 |
MAX_LFS_FILESIZE, i_size_read(inode)); |
5955102c9
|
2517 |
inode_lock(inode); |
220f2ac91
|
2518 |
/* We're holding i_mutex so we can access i_size directly */ |
a84c872c8
|
2519 |
if (offset < 0 || offset >= inode->i_size) |
220f2ac91
|
2520 2521 |
offset = -ENXIO; else { |
09cbfeaf1
|
2522 2523 |
start = offset >> PAGE_SHIFT; end = (inode->i_size + PAGE_SIZE - 1) >> PAGE_SHIFT; |
965c8e59c
|
2524 |
new_offset = shmem_seek_hole_data(mapping, start, end, whence); |
09cbfeaf1
|
2525 |
new_offset <<= PAGE_SHIFT; |
220f2ac91
|
2526 2527 2528 |
if (new_offset > offset) { if (new_offset < inode->i_size) offset = new_offset; |
965c8e59c
|
2529 |
else if (whence == SEEK_DATA) |
220f2ac91
|
2530 2531 2532 2533 2534 |
offset = -ENXIO; else offset = inode->i_size; } } |
387aae6fd
|
2535 2536 |
if (offset >= 0) offset = vfs_setpos(file, offset, MAX_LFS_FILESIZE); |
5955102c9
|
2537 |
inode_unlock(inode); |
220f2ac91
|
2538 2539 |
return offset; } |
05f65b5c7
|
2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 |
/* * 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
|
2557 2558 2559 |
radix_tree_for_each_slot(slot, &mapping->page_tree, &iter, start) { page = radix_tree_deref_slot(slot); if (!page || radix_tree_exception(page)) { |
2cf938aae
|
2560 2561 2562 2563 |
if (radix_tree_deref_retry(page)) { slot = radix_tree_iter_retry(&iter); continue; } |
05f65b5c7
|
2564 2565 2566 2567 2568 2569 2570 2571 |
} 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()) { |
148deab22
|
2572 |
slot = radix_tree_iter_resume(slot, &iter); |
05f65b5c7
|
2573 |
cond_resched_rcu(); |
05f65b5c7
|
2574 2575 2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 2586 2587 |
} } 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
|
2588 2589 |
static int shmem_wait_for_pins(struct address_space *mapping) { |
05f65b5c7
|
2590 2591 2592 2593 2594 2595 2596 2597 2598 2599 2600 2601 2602 2603 2604 2605 2606 2607 2608 2609 |
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
|
2610 2611 2612 2613 2614 |
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
|
2615 2616 2617 2618 |
if (radix_tree_deref_retry(page)) { slot = radix_tree_iter_retry(&iter); continue; } |
05f65b5c7
|
2619 2620 2621 2622 2623 2624 2625 2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 |
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()) { |
148deab22
|
2642 |
slot = radix_tree_iter_resume(slot, &iter); |
05f65b5c7
|
2643 |
cond_resched_rcu(); |
05f65b5c7
|
2644 2645 2646 2647 2648 2649 |
} } rcu_read_unlock(); } return error; |
40e041a2c
|
2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 |
} #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
|
2699 |
inode_lock(inode); |
40e041a2c
|
2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 |
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
|
2722 |
inode_unlock(inode); |
40e041a2c
|
2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751 2752 2753 2754 2755 2756 2757 |
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
|
2758 2759 2760 |
static long shmem_fallocate(struct file *file, int mode, loff_t offset, loff_t len) { |
496ad9aa8
|
2761 |
struct inode *inode = file_inode(file); |
e2d12e22c
|
2762 |
struct shmem_sb_info *sbinfo = SHMEM_SB(inode->i_sb); |
40e041a2c
|
2763 |
struct shmem_inode_info *info = SHMEM_I(inode); |
1aac14003
|
2764 |
struct shmem_falloc shmem_falloc; |
e2d12e22c
|
2765 2766 |
pgoff_t start, index, end; int error; |
83e4fa9c1
|
2767 |
|
13ace4d0d
|
2768 2769 |
if (mode & ~(FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE)) return -EOPNOTSUPP; |
5955102c9
|
2770 |
inode_lock(inode); |
83e4fa9c1
|
2771 2772 2773 2774 2775 |
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
|
2776 |
DECLARE_WAIT_QUEUE_HEAD_ONSTACK(shmem_falloc_waitq); |
83e4fa9c1
|
2777 |
|
40e041a2c
|
2778 2779 2780 2781 2782 |
/* protected by i_mutex */ if (info->seals & F_SEAL_WRITE) { error = -EPERM; goto out; } |
8e205f779
|
2783 |
shmem_falloc.waitq = &shmem_falloc_waitq; |
f00cdc6df
|
2784 2785 2786 2787 2788 |
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
|
2789 2790 2791 2792 2793 |
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
|
2794 2795 2796 2797 |
spin_lock(&inode->i_lock); inode->i_private = NULL; wake_up_all(&shmem_falloc_waitq); |
2055da973
|
2798 |
WARN_ON_ONCE(!list_empty(&shmem_falloc_waitq.head)); |
8e205f779
|
2799 |
spin_unlock(&inode->i_lock); |
83e4fa9c1
|
2800 |
error = 0; |
8e205f779
|
2801 |
goto out; |
e2d12e22c
|
2802 2803 2804 2805 2806 2807 |
} /* We need to check rlimit even when FALLOC_FL_KEEP_SIZE */ error = inode_newsize_ok(inode, offset + len); if (error) goto out; |
40e041a2c
|
2808 2809 2810 2811 |
if ((info->seals & F_SEAL_GROW) && offset + len > inode->i_size) { error = -EPERM; goto out; } |
09cbfeaf1
|
2812 2813 |
start = offset >> PAGE_SHIFT; end = (offset + len + PAGE_SIZE - 1) >> PAGE_SHIFT; |
e2d12e22c
|
2814 2815 2816 2817 |
/* 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
|
2818 |
} |
8e205f779
|
2819 |
shmem_falloc.waitq = NULL; |
1aac14003
|
2820 2821 2822 2823 2824 2825 2826 |
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
|
2827 2828 2829 2830 2831 2832 2833 2834 2835 |
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
|
2836 2837 |
else if (shmem_falloc.nr_unswapped > shmem_falloc.nr_falloced) error = -ENOMEM; |
e2d12e22c
|
2838 |
else |
9e18eb293
|
2839 |
error = shmem_getpage(inode, index, &page, SGP_FALLOC); |
e2d12e22c
|
2840 |
if (error) { |
1635f6a74
|
2841 |
/* Remove the !PageUptodate pages we added */ |
7f5565670
|
2842 2843 2844 2845 2846 |
if (index > start) { shmem_undo_range(inode, (loff_t)start << PAGE_SHIFT, ((loff_t)index << PAGE_SHIFT) - 1, true); } |
1aac14003
|
2847 |
goto undone; |
e2d12e22c
|
2848 |
} |
e2d12e22c
|
2849 |
/* |
1aac14003
|
2850 2851 2852 2853 2854 2855 2856 2857 |
* 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
|
2858 2859 2860 |
* 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
|
2861 2862 2863 2864 2865 |
* 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
|
2866 |
put_page(page); |
e2d12e22c
|
2867 2868 2869 2870 2871 |
cond_resched(); } if (!(mode & FALLOC_FL_KEEP_SIZE) && offset + len > inode->i_size) i_size_write(inode, offset + len); |
078cd8279
|
2872 |
inode->i_ctime = current_time(inode); |
1aac14003
|
2873 2874 2875 2876 |
undone: spin_lock(&inode->i_lock); inode->i_private = NULL; spin_unlock(&inode->i_lock); |
e2d12e22c
|
2877 |
out: |
5955102c9
|
2878 |
inode_unlock(inode); |
83e4fa9c1
|
2879 2880 |
return error; } |
726c33422
|
2881 |
static int shmem_statfs(struct dentry *dentry, struct kstatfs *buf) |
1da177e4c
|
2882 |
{ |
726c33422
|
2883 |
struct shmem_sb_info *sbinfo = SHMEM_SB(dentry->d_sb); |
1da177e4c
|
2884 2885 |
buf->f_type = TMPFS_MAGIC; |
09cbfeaf1
|
2886 |
buf->f_bsize = PAGE_SIZE; |
1da177e4c
|
2887 |
buf->f_namelen = NAME_MAX; |
0edd73b33
|
2888 |
if (sbinfo->max_blocks) { |
1da177e4c
|
2889 |
buf->f_blocks = sbinfo->max_blocks; |
41ffe5d5c
|
2890 2891 2892 |
buf->f_bavail = buf->f_bfree = sbinfo->max_blocks - percpu_counter_sum(&sbinfo->used_blocks); |
0edd73b33
|
2893 2894 |
} if (sbinfo->max_inodes) { |
1da177e4c
|
2895 2896 |
buf->f_files = sbinfo->max_inodes; buf->f_ffree = sbinfo->free_inodes; |
1da177e4c
|
2897 2898 2899 2900 2901 2902 2903 2904 2905 |
} /* else leave those fields 0 like simple_statfs */ return 0; } /* * File creation. Allocate an inode, and we're done.. */ static int |
1a67aafb5
|
2906 |
shmem_mknod(struct inode *dir, struct dentry *dentry, umode_t mode, dev_t dev) |
1da177e4c
|
2907 |
{ |
0b0a0806b
|
2908 |
struct inode *inode; |
1da177e4c
|
2909 |
int error = -ENOSPC; |
454abafe9
|
2910 |
inode = shmem_get_inode(dir->i_sb, dir, mode, dev, VM_NORESERVE); |
1da177e4c
|
2911 |
if (inode) { |
feda821e7
|
2912 2913 2914 |
error = simple_acl_create(dir, inode); if (error) goto out_iput; |
2a7dba391
|
2915 |
error = security_inode_init_security(inode, dir, |
9d8f13ba3
|
2916 |
&dentry->d_name, |
6d9d88d07
|
2917 |
shmem_initxattrs, NULL); |
feda821e7
|
2918 2919 |
if (error && error != -EOPNOTSUPP) goto out_iput; |
37ec43cdc
|
2920 |
|
718deb6b6
|
2921 |
error = 0; |
1da177e4c
|
2922 |
dir->i_size += BOGO_DIRENT_SIZE; |
078cd8279
|
2923 |
dir->i_ctime = dir->i_mtime = current_time(dir); |
1da177e4c
|
2924 2925 |
d_instantiate(dentry, inode); dget(dentry); /* Extra count - pin the dentry in core */ |
1da177e4c
|
2926 2927 |
} return error; |
feda821e7
|
2928 2929 2930 |
out_iput: iput(inode); return error; |
1da177e4c
|
2931 |
} |
60545d0d4
|
2932 2933 2934 2935 2936 2937 2938 2939 2940 2941 2942 |
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
|
2943 2944 2945 2946 2947 |
if (error && error != -EOPNOTSUPP) goto out_iput; error = simple_acl_create(dir, inode); if (error) goto out_iput; |
60545d0d4
|
2948 2949 2950 |
d_tmpfile(dentry, inode); } return error; |
feda821e7
|
2951 2952 2953 |
out_iput: iput(inode); return error; |
60545d0d4
|
2954 |
} |
18bb1db3e
|
2955 |
static int shmem_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode) |
1da177e4c
|
2956 2957 2958 2959 2960 |
{ int error; if ((error = shmem_mknod(dir, dentry, mode | S_IFDIR, 0))) return error; |
d8c76e6f4
|
2961 |
inc_nlink(dir); |
1da177e4c
|
2962 2963 |
return 0; } |
4acdaf27e
|
2964 |
static int shmem_create(struct inode *dir, struct dentry *dentry, umode_t mode, |
ebfc3b49a
|
2965 |
bool excl) |
1da177e4c
|
2966 2967 2968 2969 2970 2971 2972 2973 2974 |
{ 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
|
2975 |
struct inode *inode = d_inode(old_dentry); |
5b04c6890
|
2976 |
int ret; |
1da177e4c
|
2977 2978 2979 2980 2981 2982 |
/* * 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
|
2983 2984 2985 |
ret = shmem_reserve_inode(inode->i_sb); if (ret) goto out; |
1da177e4c
|
2986 2987 |
dir->i_size += BOGO_DIRENT_SIZE; |
078cd8279
|
2988 |
inode->i_ctime = dir->i_ctime = dir->i_mtime = current_time(inode); |
d8c76e6f4
|
2989 |
inc_nlink(inode); |
7de9c6ee3
|
2990 |
ihold(inode); /* New dentry reference */ |
1da177e4c
|
2991 2992 |
dget(dentry); /* Extra pinning count for the created dentry */ d_instantiate(dentry, inode); |
5b04c6890
|
2993 2994 |
out: return ret; |
1da177e4c
|
2995 2996 2997 2998 |
} static int shmem_unlink(struct inode *dir, struct dentry *dentry) { |
75c3cfa85
|
2999 |
struct inode *inode = d_inode(dentry); |
1da177e4c
|
3000 |
|
5b04c6890
|
3001 3002 |
if (inode->i_nlink > 1 && !S_ISDIR(inode->i_mode)) shmem_free_inode(inode->i_sb); |
1da177e4c
|
3003 3004 |
dir->i_size -= BOGO_DIRENT_SIZE; |
078cd8279
|
3005 |
inode->i_ctime = dir->i_ctime = dir->i_mtime = current_time(inode); |
9a53c3a78
|
3006 |
drop_nlink(inode); |
1da177e4c
|
3007 3008 3009 3010 3011 3012 3013 3014 |
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
|
3015 |
drop_nlink(d_inode(dentry)); |
9a53c3a78
|
3016 |
drop_nlink(dir); |
1da177e4c
|
3017 3018 |
return shmem_unlink(dir, dentry); } |
37456771c
|
3019 3020 |
static int shmem_exchange(struct inode *old_dir, struct dentry *old_dentry, struct inode *new_dir, struct dentry *new_dentry) { |
e36cb0b89
|
3021 3022 |
bool old_is_dir = d_is_dir(old_dentry); bool new_is_dir = d_is_dir(new_dentry); |
37456771c
|
3023 3024 3025 3026 3027 3028 3029 3030 3031 3032 3033 3034 |
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
|
3035 |
d_inode(old_dentry)->i_ctime = |
078cd8279
|
3036 |
d_inode(new_dentry)->i_ctime = current_time(old_dir); |
37456771c
|
3037 3038 3039 |
return 0; } |
46fdb794e
|
3040 3041 3042 3043 3044 3045 3046 3047 3048 3049 3050 3051 3052 3053 3054 3055 3056 3057 3058 3059 3060 3061 3062 3063 3064 |
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
|
3065 3066 3067 3068 3069 3070 |
/* * 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
|
3071 |
static int shmem_rename2(struct inode *old_dir, struct dentry *old_dentry, struct inode *new_dir, struct dentry *new_dentry, unsigned int flags) |
1da177e4c
|
3072 |
{ |
75c3cfa85
|
3073 |
struct inode *inode = d_inode(old_dentry); |
1da177e4c
|
3074 |
int they_are_dirs = S_ISDIR(inode->i_mode); |
46fdb794e
|
3075 |
if (flags & ~(RENAME_NOREPLACE | RENAME_EXCHANGE | RENAME_WHITEOUT)) |
3b69ff51d
|
3076 |
return -EINVAL; |
37456771c
|
3077 3078 |
if (flags & RENAME_EXCHANGE) return shmem_exchange(old_dir, old_dentry, new_dir, new_dentry); |
1da177e4c
|
3079 3080 |
if (!simple_empty(new_dentry)) return -ENOTEMPTY; |
46fdb794e
|
3081 3082 3083 3084 3085 3086 3087 |
if (flags & RENAME_WHITEOUT) { int error; error = shmem_whiteout(old_dir, old_dentry); if (error) return error; } |
75c3cfa85
|
3088 |
if (d_really_is_positive(new_dentry)) { |
1da177e4c
|
3089 |
(void) shmem_unlink(new_dir, new_dentry); |
b928095b0
|
3090 |
if (they_are_dirs) { |
75c3cfa85
|
3091 |
drop_nlink(d_inode(new_dentry)); |
9a53c3a78
|
3092 |
drop_nlink(old_dir); |
b928095b0
|
3093 |
} |
1da177e4c
|
3094 |
} else if (they_are_dirs) { |
9a53c3a78
|
3095 |
drop_nlink(old_dir); |
d8c76e6f4
|
3096 |
inc_nlink(new_dir); |
1da177e4c
|
3097 3098 3099 3100 3101 3102 |
} 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
|
3103 |
inode->i_ctime = current_time(old_dir); |
1da177e4c
|
3104 3105 3106 3107 3108 3109 3110 3111 |
return 0; } static int shmem_symlink(struct inode *dir, struct dentry *dentry, const char *symname) { int error; int len; struct inode *inode; |
9276aad6c
|
3112 |
struct page *page; |
1da177e4c
|
3113 3114 3115 |
struct shmem_inode_info *info; len = strlen(symname) + 1; |
09cbfeaf1
|
3116 |
if (len > PAGE_SIZE) |
1da177e4c
|
3117 |
return -ENAMETOOLONG; |
454abafe9
|
3118 |
inode = shmem_get_inode(dir->i_sb, dir, S_IFLNK|S_IRWXUGO, 0, VM_NORESERVE); |
1da177e4c
|
3119 3120 |
if (!inode) return -ENOSPC; |
9d8f13ba3
|
3121 |
error = security_inode_init_security(inode, dir, &dentry->d_name, |
6d9d88d07
|
3122 |
shmem_initxattrs, NULL); |
570bc1c2e
|
3123 3124 3125 3126 3127 3128 3129 |
if (error) { if (error != -EOPNOTSUPP) { iput(inode); return error; } error = 0; } |
1da177e4c
|
3130 3131 |
info = SHMEM_I(inode); inode->i_size = len-1; |
69f07ec93
|
3132 |
if (len <= SHORT_SYMLINK_LEN) { |
3ed47db34
|
3133 3134 |
inode->i_link = kmemdup(symname, len, GFP_KERNEL); if (!inode->i_link) { |
69f07ec93
|
3135 3136 3137 3138 |
iput(inode); return -ENOMEM; } inode->i_op = &shmem_short_symlink_operations; |
1da177e4c
|
3139 |
} else { |
e8ecde25f
|
3140 |
inode_nohighmem(inode); |
9e18eb293
|
3141 |
error = shmem_getpage(inode, 0, &page, SGP_WRITE); |
1da177e4c
|
3142 3143 3144 3145 |
if (error) { iput(inode); return error; } |
14fcc23fd
|
3146 |
inode->i_mapping->a_ops = &shmem_aops; |
1da177e4c
|
3147 |
inode->i_op = &shmem_symlink_inode_operations; |
21fc61c73
|
3148 |
memcpy(page_address(page), symname, len); |
ec9516fbc
|
3149 |
SetPageUptodate(page); |
1da177e4c
|
3150 |
set_page_dirty(page); |
6746aff74
|
3151 |
unlock_page(page); |
09cbfeaf1
|
3152 |
put_page(page); |
1da177e4c
|
3153 |
} |
1da177e4c
|
3154 |
dir->i_size += BOGO_DIRENT_SIZE; |
078cd8279
|
3155 |
dir->i_ctime = dir->i_mtime = current_time(dir); |
1da177e4c
|
3156 3157 3158 3159 |
d_instantiate(dentry, inode); dget(dentry); return 0; } |
fceef393a
|
3160 |
static void shmem_put_link(void *arg) |
1da177e4c
|
3161 |
{ |
fceef393a
|
3162 3163 |
mark_page_accessed(arg); put_page(arg); |
1da177e4c
|
3164 |
} |
6b2553918
|
3165 |
static const char *shmem_get_link(struct dentry *dentry, |
fceef393a
|
3166 3167 |
struct inode *inode, struct delayed_call *done) |
1da177e4c
|
3168 |
{ |
1da177e4c
|
3169 |
struct page *page = NULL; |
6b2553918
|
3170 |
int error; |
6a6c99049
|
3171 3172 3173 3174 3175 3176 3177 3178 3179 |
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
|
3180 |
error = shmem_getpage(inode, 0, &page, SGP_READ); |
6a6c99049
|
3181 3182 3183 3184 |
if (error) return ERR_PTR(error); unlock_page(page); } |
fceef393a
|
3185 |
set_delayed_call(done, shmem_put_link, page); |
21fc61c73
|
3186 |
return page_address(page); |
1da177e4c
|
3187 |
} |
b09e0fa4b
|
3188 |
#ifdef CONFIG_TMPFS_XATTR |
467118102
|
3189 |
/* |
b09e0fa4b
|
3190 3191 |
* 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
|
3192 3193 3194 |
* like ACLs, we also need to implement the security.* handlers at * filesystem level, though. */ |
6d9d88d07
|
3195 |
/* |
6d9d88d07
|
3196 3197 3198 3199 3200 3201 3202 3203 |
* 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
|
3204 |
struct simple_xattr *new_xattr; |
6d9d88d07
|
3205 3206 3207 |
size_t len; for (xattr = xattr_array; xattr->name != NULL; xattr++) { |
38f386574
|
3208 |
new_xattr = simple_xattr_alloc(xattr->value, xattr->value_len); |
6d9d88d07
|
3209 3210 3211 3212 3213 3214 3215 3216 3217 3218 3219 3220 3221 3222 3223 |
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
|
3224 |
simple_xattr_list_add(&info->xattrs, new_xattr); |
6d9d88d07
|
3225 3226 3227 3228 |
} return 0; } |
aa7c5241c
|
3229 |
static int shmem_xattr_handler_get(const struct xattr_handler *handler, |
b296821a7
|
3230 3231 |
struct dentry *unused, struct inode *inode, const char *name, void *buffer, size_t size) |
b09e0fa4b
|
3232 |
{ |
b296821a7
|
3233 |
struct shmem_inode_info *info = SHMEM_I(inode); |
b09e0fa4b
|
3234 |
|
aa7c5241c
|
3235 |
name = xattr_full_name(handler, name); |
38f386574
|
3236 |
return simple_xattr_get(&info->xattrs, name, buffer, size); |
b09e0fa4b
|
3237 |
} |
aa7c5241c
|
3238 |
static int shmem_xattr_handler_set(const struct xattr_handler *handler, |
593012268
|
3239 3240 3241 |
struct dentry *unused, struct inode *inode, const char *name, const void *value, size_t size, int flags) |
b09e0fa4b
|
3242 |
{ |
593012268
|
3243 |
struct shmem_inode_info *info = SHMEM_I(inode); |
b09e0fa4b
|
3244 |
|
aa7c5241c
|
3245 |
name = xattr_full_name(handler, name); |
38f386574
|
3246 |
return simple_xattr_set(&info->xattrs, name, value, size, flags); |
b09e0fa4b
|
3247 |
} |
aa7c5241c
|
3248 3249 3250 3251 3252 |
static const struct xattr_handler shmem_security_xattr_handler = { .prefix = XATTR_SECURITY_PREFIX, .get = shmem_xattr_handler_get, .set = shmem_xattr_handler_set, }; |
b09e0fa4b
|
3253 |
|
aa7c5241c
|
3254 3255 3256 3257 3258 |
static const struct xattr_handler shmem_trusted_xattr_handler = { .prefix = XATTR_TRUSTED_PREFIX, .get = shmem_xattr_handler_get, .set = shmem_xattr_handler_set, }; |
b09e0fa4b
|
3259 |
|
aa7c5241c
|
3260 3261 3262 3263 3264 3265 3266 3267 3268 |
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
|
3269 3270 3271 |
static ssize_t shmem_listxattr(struct dentry *dentry, char *buffer, size_t size) { |
75c3cfa85
|
3272 |
struct shmem_inode_info *info = SHMEM_I(d_inode(dentry)); |
786534b92
|
3273 |
return simple_xattr_list(d_inode(dentry), &info->xattrs, buffer, size); |
b09e0fa4b
|
3274 3275 |
} #endif /* CONFIG_TMPFS_XATTR */ |
69f07ec93
|
3276 |
static const struct inode_operations shmem_short_symlink_operations = { |
6b2553918
|
3277 |
.get_link = simple_get_link, |
b09e0fa4b
|
3278 |
#ifdef CONFIG_TMPFS_XATTR |
b09e0fa4b
|
3279 |
.listxattr = shmem_listxattr, |
b09e0fa4b
|
3280 3281 3282 3283 |
#endif }; static const struct inode_operations shmem_symlink_inode_operations = { |
6b2553918
|
3284 |
.get_link = shmem_get_link, |
b09e0fa4b
|
3285 |
#ifdef CONFIG_TMPFS_XATTR |
b09e0fa4b
|
3286 |
.listxattr = shmem_listxattr, |
39f0247d3
|
3287 |
#endif |
b09e0fa4b
|
3288 |
}; |
39f0247d3
|
3289 |
|
91828a405
|
3290 3291 3292 3293 3294 3295 3296 3297 3298 3299 3300 3301 |
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
|
3302 3303 |
static struct dentry *shmem_fh_to_dentry(struct super_block *sb, struct fid *fid, int fh_len, int fh_type) |
91828a405
|
3304 |
{ |
91828a405
|
3305 |
struct inode *inode; |
480b116c9
|
3306 |
struct dentry *dentry = NULL; |
35c2a7f49
|
3307 |
u64 inum; |
480b116c9
|
3308 3309 3310 |
if (fh_len < 3) return NULL; |
91828a405
|
3311 |
|
35c2a7f49
|
3312 3313 |
inum = fid->raw[2]; inum = (inum << 32) | fid->raw[1]; |
480b116c9
|
3314 3315 |
inode = ilookup5(sb, (unsigned long)(inum + fid->raw[0]), shmem_match, fid->raw); |
91828a405
|
3316 |
if (inode) { |
480b116c9
|
3317 |
dentry = d_find_alias(inode); |
91828a405
|
3318 3319 |
iput(inode); } |
480b116c9
|
3320 |
return dentry; |
91828a405
|
3321 |
} |
b0b0382bb
|
3322 3323 |
static int shmem_encode_fh(struct inode *inode, __u32 *fh, int *len, struct inode *parent) |
91828a405
|
3324 |
{ |
5fe0c2378
|
3325 3326 |
if (*len < 3) { *len = 3; |
94e07a759
|
3327 |
return FILEID_INVALID; |
5fe0c2378
|
3328 |
} |
91828a405
|
3329 |
|
1d3382cbf
|
3330 |
if (inode_unhashed(inode)) { |
91828a405
|
3331 3332 3333 3334 3335 3336 3337 |
/* 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
|
3338 |
if (inode_unhashed(inode)) |
91828a405
|
3339 3340 3341 3342 3343 3344 3345 3346 3347 3348 3349 3350 |
__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
|
3351 |
static const struct export_operations shmem_export_ops = { |
91828a405
|
3352 |
.get_parent = shmem_get_parent, |
91828a405
|
3353 |
.encode_fh = shmem_encode_fh, |
480b116c9
|
3354 |
.fh_to_dentry = shmem_fh_to_dentry, |
91828a405
|
3355 |
}; |
680d794ba
|
3356 3357 |
static int shmem_parse_options(char *options, struct shmem_sb_info *sbinfo, bool remount) |
1da177e4c
|
3358 3359 |
{ char *this_char, *value, *rest; |
49cd0a5c2
|
3360 |
struct mempolicy *mpol = NULL; |
8751e0395
|
3361 3362 |
uid_t uid; gid_t gid; |
1da177e4c
|
3363 |
|
b00dc3ad7
|
3364 3365 3366 3367 3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 3378 3379 3380 |
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
|
3381 3382 3383 3384 3385 |
if (!*this_char) continue; if ((value = strchr(this_char,'=')) != NULL) { *value++ = 0; } else { |
1170532bb
|
3386 3387 3388 |
pr_err("tmpfs: No value for mount option '%s' ", this_char); |
49cd0a5c2
|
3389 |
goto error; |
1da177e4c
|
3390 3391 3392 3393 3394 3395 3396 3397 3398 3399 3400 3401 3402 |
} 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
|
3403 |
sbinfo->max_blocks = |
09cbfeaf1
|
3404 |
DIV_ROUND_UP(size, PAGE_SIZE); |
1da177e4c
|
3405 |
} else if (!strcmp(this_char,"nr_blocks")) { |
680d794ba
|
3406 |
sbinfo->max_blocks = memparse(value, &rest); |
1da177e4c
|
3407 3408 3409 |
if (*rest) goto bad_val; } else if (!strcmp(this_char,"nr_inodes")) { |
680d794ba
|
3410 |
sbinfo->max_inodes = memparse(value, &rest); |
1da177e4c
|
3411 3412 3413 |
if (*rest) goto bad_val; } else if (!strcmp(this_char,"mode")) { |
680d794ba
|
3414 |
if (remount) |
1da177e4c
|
3415 |
continue; |
680d794ba
|
3416 |
sbinfo->mode = simple_strtoul(value, &rest, 8) & 07777; |
1da177e4c
|
3417 3418 3419 |
if (*rest) goto bad_val; } else if (!strcmp(this_char,"uid")) { |
680d794ba
|
3420 |
if (remount) |
1da177e4c
|
3421 |
continue; |
8751e0395
|
3422 |
uid = simple_strtoul(value, &rest, 0); |
1da177e4c
|
3423 3424 |
if (*rest) goto bad_val; |
8751e0395
|
3425 3426 3427 |
sbinfo->uid = make_kuid(current_user_ns(), uid); if (!uid_valid(sbinfo->uid)) goto bad_val; |
1da177e4c
|
3428 |
} else if (!strcmp(this_char,"gid")) { |
680d794ba
|
3429 |
if (remount) |
1da177e4c
|
3430 |
continue; |
8751e0395
|
3431 |
gid = simple_strtoul(value, &rest, 0); |
1da177e4c
|
3432 3433 |
if (*rest) goto bad_val; |
8751e0395
|
3434 3435 3436 |
sbinfo->gid = make_kgid(current_user_ns(), gid); if (!gid_valid(sbinfo->gid)) goto bad_val; |
e496cf3d7
|
3437 |
#ifdef CONFIG_TRANSPARENT_HUGE_PAGECACHE |
5a6e75f81
|
3438 3439 3440 3441 3442 3443 3444 3445 3446 3447 3448 |
} 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
|
3449 |
} else if (!strcmp(this_char,"mpol")) { |
49cd0a5c2
|
3450 3451 3452 |
mpol_put(mpol); mpol = NULL; if (mpol_parse_str(value, &mpol)) |
7339ff830
|
3453 |
goto bad_val; |
5a6e75f81
|
3454 |
#endif |
1da177e4c
|
3455 |
} else { |
1170532bb
|
3456 3457 |
pr_err("tmpfs: Bad mount option %s ", this_char); |
49cd0a5c2
|
3458 |
goto error; |
1da177e4c
|
3459 3460 |
} } |
49cd0a5c2
|
3461 |
sbinfo->mpol = mpol; |
1da177e4c
|
3462 3463 3464 |
return 0; bad_val: |
1170532bb
|
3465 3466 |
pr_err("tmpfs: Bad value '%s' for mount option '%s' ", |
1da177e4c
|
3467 |
value, this_char); |
49cd0a5c2
|
3468 3469 |
error: mpol_put(mpol); |
1da177e4c
|
3470 3471 3472 3473 3474 3475 3476 |
return 1; } static int shmem_remount_fs(struct super_block *sb, int *flags, char *data) { struct shmem_sb_info *sbinfo = SHMEM_SB(sb); |
680d794ba
|
3477 |
struct shmem_sb_info config = *sbinfo; |
0edd73b33
|
3478 3479 |
unsigned long inodes; int error = -EINVAL; |
5f00110f7
|
3480 |
config.mpol = NULL; |
680d794ba
|
3481 |
if (shmem_parse_options(data, &config, true)) |
0edd73b33
|
3482 |
return error; |
1da177e4c
|
3483 |
|
0edd73b33
|
3484 |
spin_lock(&sbinfo->stat_lock); |
0edd73b33
|
3485 |
inodes = sbinfo->max_inodes - sbinfo->free_inodes; |
7e496299d
|
3486 |
if (percpu_counter_compare(&sbinfo->used_blocks, config.max_blocks) > 0) |
0edd73b33
|
3487 |
goto out; |
680d794ba
|
3488 |
if (config.max_inodes < inodes) |
0edd73b33
|
3489 3490 |
goto out; /* |
54af60421
|
3491 |
* Those tests disallow limited->unlimited while any are in use; |
0edd73b33
|
3492 3493 3494 |
* but we must separately disallow unlimited->limited, because * in that case we have no record of how much is already in use. */ |
680d794ba
|
3495 |
if (config.max_blocks && !sbinfo->max_blocks) |
0edd73b33
|
3496 |
goto out; |
680d794ba
|
3497 |
if (config.max_inodes && !sbinfo->max_inodes) |
0edd73b33
|
3498 3499 3500 |
goto out; error = 0; |
5a6e75f81
|
3501 |
sbinfo->huge = config.huge; |
680d794ba
|
3502 |
sbinfo->max_blocks = config.max_blocks; |
680d794ba
|
3503 3504 |
sbinfo->max_inodes = config.max_inodes; sbinfo->free_inodes = config.max_inodes - inodes; |
71fe804b6
|
3505 |
|
5f00110f7
|
3506 3507 3508 3509 3510 3511 3512 |
/* * Preserve previous mempolicy unless mpol remount option was specified. */ if (config.mpol) { mpol_put(sbinfo->mpol); sbinfo->mpol = config.mpol; /* transfers initial ref */ } |
0edd73b33
|
3513 3514 3515 |
out: spin_unlock(&sbinfo->stat_lock); return error; |
1da177e4c
|
3516 |
} |
680d794ba
|
3517 |
|
34c80b1d9
|
3518 |
static int shmem_show_options(struct seq_file *seq, struct dentry *root) |
680d794ba
|
3519 |
{ |
34c80b1d9
|
3520 |
struct shmem_sb_info *sbinfo = SHMEM_SB(root->d_sb); |
680d794ba
|
3521 3522 3523 |
if (sbinfo->max_blocks != shmem_default_max_blocks()) seq_printf(seq, ",size=%luk", |
09cbfeaf1
|
3524 |
sbinfo->max_blocks << (PAGE_SHIFT - 10)); |
680d794ba
|
3525 3526 3527 |
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
|
3528 |
seq_printf(seq, ",mode=%03ho", sbinfo->mode); |
8751e0395
|
3529 3530 3531 3532 3533 3534 |
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
|
3535 |
#ifdef CONFIG_TRANSPARENT_HUGE_PAGECACHE |
5a6e75f81
|
3536 3537 3538 3539 |
/* 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
|
3540 |
shmem_show_mpol(seq, sbinfo->mpol); |
680d794ba
|
3541 3542 |
return 0; } |
9183df25f
|
3543 3544 3545 3546 |
#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) |
749df87bd
|
3547 |
#define MFD_ALL_FLAGS (MFD_CLOEXEC | MFD_ALLOW_SEALING | MFD_HUGETLB) |
9183df25f
|
3548 3549 3550 3551 3552 3553 3554 3555 3556 3557 |
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; |
749df87bd
|
3558 3559 3560 3561 3562 3563 3564 3565 3566 3567 3568 3569 |
if (!(flags & MFD_HUGETLB)) { if (flags & ~(unsigned int)MFD_ALL_FLAGS) return -EINVAL; } else { /* Sealing not supported in hugetlbfs (MFD_HUGETLB) */ if (flags & MFD_ALLOW_SEALING) return -EINVAL; /* Allow huge page size encoding in flags. */ if (flags & ~(unsigned int)(MFD_ALL_FLAGS | (MFD_HUGE_MASK << MFD_HUGE_SHIFT))) return -EINVAL; } |
9183df25f
|
3570 3571 3572 3573 3574 3575 3576 |
/* 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; |
0ee931c4e
|
3577 |
name = kmalloc(len + MFD_NAME_PREFIX_LEN, GFP_KERNEL); |
9183df25f
|
3578 3579 3580 3581 3582 3583 3584 3585 3586 3587 3588 3589 3590 3591 3592 3593 3594 3595 3596 3597 |
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; } |
749df87bd
|
3598 3599 3600 3601 3602 3603 3604 3605 3606 |
if (flags & MFD_HUGETLB) { struct user_struct *user = NULL; file = hugetlb_file_setup(name, 0, VM_NORESERVE, &user, HUGETLB_ANONHUGE_INODE, (flags >> MFD_HUGE_SHIFT) & MFD_HUGE_MASK); } else file = shmem_file_setup(name, 0, VM_NORESERVE); |
9183df25f
|
3607 3608 3609 3610 |
if (IS_ERR(file)) { error = PTR_ERR(file); goto err_fd; } |
9183df25f
|
3611 3612 |
file->f_mode |= FMODE_LSEEK | FMODE_PREAD | FMODE_PWRITE; file->f_flags |= O_RDWR | O_LARGEFILE; |
749df87bd
|
3613 3614 3615 3616 3617 3618 3619 |
if (flags & MFD_ALLOW_SEALING) { /* * flags check at beginning of function ensures * this is not a hugetlbfs (MFD_HUGETLB) file. */ info = SHMEM_I(file_inode(file)); |
9183df25f
|
3620 |
info->seals &= ~F_SEAL_SEAL; |
749df87bd
|
3621 |
} |
9183df25f
|
3622 3623 3624 3625 3626 3627 3628 3629 3630 3631 3632 |
fd_install(fd, file); kfree(name); return fd; err_fd: put_unused_fd(fd); err_name: kfree(name); return error; } |
680d794ba
|
3633 |
#endif /* CONFIG_TMPFS */ |
1da177e4c
|
3634 3635 3636 |
static void shmem_put_super(struct super_block *sb) { |
602586a83
|
3637 3638 3639 |
struct shmem_sb_info *sbinfo = SHMEM_SB(sb); percpu_counter_destroy(&sbinfo->used_blocks); |
49cd0a5c2
|
3640 |
mpol_put(sbinfo->mpol); |
602586a83
|
3641 |
kfree(sbinfo); |
1da177e4c
|
3642 3643 |
sb->s_fs_info = NULL; } |
2b2af54a5
|
3644 |
int shmem_fill_super(struct super_block *sb, void *data, int silent) |
1da177e4c
|
3645 3646 |
{ struct inode *inode; |
0edd73b33
|
3647 |
struct shmem_sb_info *sbinfo; |
680d794ba
|
3648 3649 3650 |
int err = -ENOMEM; /* Round up to L1_CACHE_BYTES to resist false sharing */ |
425fbf047
|
3651 |
sbinfo = kzalloc(max((int)sizeof(struct shmem_sb_info), |
680d794ba
|
3652 3653 3654 |
L1_CACHE_BYTES), GFP_KERNEL); if (!sbinfo) return -ENOMEM; |
680d794ba
|
3655 |
sbinfo->mode = S_IRWXUGO | S_ISVTX; |
76aac0e9a
|
3656 3657 |
sbinfo->uid = current_fsuid(); sbinfo->gid = current_fsgid(); |
680d794ba
|
3658 |
sb->s_fs_info = sbinfo; |
1da177e4c
|
3659 |
|
0edd73b33
|
3660 |
#ifdef CONFIG_TMPFS |
1da177e4c
|
3661 3662 3663 3664 3665 |
/* * 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
|
3666 |
if (!(sb->s_flags & MS_KERNMOUNT)) { |
680d794ba
|
3667 3668 3669 3670 3671 3672 |
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
|
3673 3674 |
} else { sb->s_flags |= MS_NOUSER; |
1da177e4c
|
3675 |
} |
91828a405
|
3676 |
sb->s_export_op = &shmem_export_ops; |
2f6e38f3c
|
3677 |
sb->s_flags |= MS_NOSEC; |
1da177e4c
|
3678 3679 3680 |
#else sb->s_flags |= MS_NOUSER; #endif |
0edd73b33
|
3681 |
spin_lock_init(&sbinfo->stat_lock); |
908c7f194
|
3682 |
if (percpu_counter_init(&sbinfo->used_blocks, 0, GFP_KERNEL)) |
602586a83
|
3683 |
goto failed; |
680d794ba
|
3684 |
sbinfo->free_inodes = sbinfo->max_inodes; |
779750d20
|
3685 3686 |
spin_lock_init(&sbinfo->shrinklist_lock); INIT_LIST_HEAD(&sbinfo->shrinklist); |
0edd73b33
|
3687 |
|
285b2c4fd
|
3688 |
sb->s_maxbytes = MAX_LFS_FILESIZE; |
09cbfeaf1
|
3689 3690 |
sb->s_blocksize = PAGE_SIZE; sb->s_blocksize_bits = PAGE_SHIFT; |
1da177e4c
|
3691 3692 |
sb->s_magic = TMPFS_MAGIC; sb->s_op = &shmem_ops; |
cfd95a9cf
|
3693 |
sb->s_time_gran = 1; |
b09e0fa4b
|
3694 |
#ifdef CONFIG_TMPFS_XATTR |
39f0247d3
|
3695 |
sb->s_xattr = shmem_xattr_handlers; |
b09e0fa4b
|
3696 3697 |
#endif #ifdef CONFIG_TMPFS_POSIX_ACL |
39f0247d3
|
3698 3699 |
sb->s_flags |= MS_POSIXACL; #endif |
2b4db7961
|
3700 |
uuid_gen(&sb->s_uuid); |
0edd73b33
|
3701 |
|
454abafe9
|
3702 |
inode = shmem_get_inode(sb, NULL, S_IFDIR | sbinfo->mode, 0, VM_NORESERVE); |
1da177e4c
|
3703 3704 |
if (!inode) goto failed; |
680d794ba
|
3705 3706 |
inode->i_uid = sbinfo->uid; inode->i_gid = sbinfo->gid; |
318ceed08
|
3707 3708 |
sb->s_root = d_make_root(inode); if (!sb->s_root) |
48fde701a
|
3709 |
goto failed; |
1da177e4c
|
3710 |
return 0; |
1da177e4c
|
3711 3712 3713 3714 |
failed: shmem_put_super(sb); return err; } |
fcc234f88
|
3715 |
static struct kmem_cache *shmem_inode_cachep; |
1da177e4c
|
3716 3717 3718 |
static struct inode *shmem_alloc_inode(struct super_block *sb) { |
41ffe5d5c
|
3719 3720 3721 |
struct shmem_inode_info *info; info = kmem_cache_alloc(shmem_inode_cachep, GFP_KERNEL); if (!info) |
1da177e4c
|
3722 |
return NULL; |
41ffe5d5c
|
3723 |
return &info->vfs_inode; |
1da177e4c
|
3724 |
} |
41ffe5d5c
|
3725 |
static void shmem_destroy_callback(struct rcu_head *head) |
fa0d7e3de
|
3726 3727 |
{ struct inode *inode = container_of(head, struct inode, i_rcu); |
84e710da2
|
3728 3729 |
if (S_ISLNK(inode->i_mode)) kfree(inode->i_link); |
fa0d7e3de
|
3730 3731 |
kmem_cache_free(shmem_inode_cachep, SHMEM_I(inode)); } |
1da177e4c
|
3732 3733 |
static void shmem_destroy_inode(struct inode *inode) { |
09208d150
|
3734 |
if (S_ISREG(inode->i_mode)) |
1da177e4c
|
3735 |
mpol_free_shared_policy(&SHMEM_I(inode)->policy); |
41ffe5d5c
|
3736 |
call_rcu(&inode->i_rcu, shmem_destroy_callback); |
1da177e4c
|
3737 |
} |
41ffe5d5c
|
3738 |
static void shmem_init_inode(void *foo) |
1da177e4c
|
3739 |
{ |
41ffe5d5c
|
3740 3741 |
struct shmem_inode_info *info = foo; inode_init_once(&info->vfs_inode); |
1da177e4c
|
3742 |
} |
41ffe5d5c
|
3743 |
static int shmem_init_inodecache(void) |
1da177e4c
|
3744 3745 3746 |
{ shmem_inode_cachep = kmem_cache_create("shmem_inode_cache", sizeof(struct shmem_inode_info), |
5d097056c
|
3747 |
0, SLAB_PANIC|SLAB_ACCOUNT, shmem_init_inode); |
1da177e4c
|
3748 3749 |
return 0; } |
41ffe5d5c
|
3750 |
static void shmem_destroy_inodecache(void) |
1da177e4c
|
3751 |
{ |
1a1d92c10
|
3752 |
kmem_cache_destroy(shmem_inode_cachep); |
1da177e4c
|
3753 |
} |
f5e54d6e5
|
3754 |
static const struct address_space_operations shmem_aops = { |
1da177e4c
|
3755 |
.writepage = shmem_writepage, |
767193253
|
3756 |
.set_page_dirty = __set_page_dirty_no_writeback, |
1da177e4c
|
3757 |
#ifdef CONFIG_TMPFS |
800d15a53
|
3758 3759 |
.write_begin = shmem_write_begin, .write_end = shmem_write_end, |
1da177e4c
|
3760 |
#endif |
1c93923cc
|
3761 |
#ifdef CONFIG_MIGRATION |
304dbdb7a
|
3762 |
.migratepage = migrate_page, |
1c93923cc
|
3763 |
#endif |
aa261f549
|
3764 |
.error_remove_page = generic_error_remove_page, |
1da177e4c
|
3765 |
}; |
15ad7cdcf
|
3766 |
static const struct file_operations shmem_file_operations = { |
1da177e4c
|
3767 |
.mmap = shmem_mmap, |
c01d5b300
|
3768 |
.get_unmapped_area = shmem_get_unmapped_area, |
1da177e4c
|
3769 |
#ifdef CONFIG_TMPFS |
220f2ac91
|
3770 |
.llseek = shmem_file_llseek, |
2ba5bbed0
|
3771 |
.read_iter = shmem_file_read_iter, |
8174202b3
|
3772 |
.write_iter = generic_file_write_iter, |
1b061d924
|
3773 |
.fsync = noop_fsync, |
82c156f85
|
3774 |
.splice_read = generic_file_splice_read, |
f6cb85d00
|
3775 |
.splice_write = iter_file_splice_write, |
83e4fa9c1
|
3776 |
.fallocate = shmem_fallocate, |
1da177e4c
|
3777 3778 |
#endif }; |
92e1d5be9
|
3779 |
static const struct inode_operations shmem_inode_operations = { |
44a30220b
|
3780 |
.getattr = shmem_getattr, |
94c1e62df
|
3781 |
.setattr = shmem_setattr, |
b09e0fa4b
|
3782 |
#ifdef CONFIG_TMPFS_XATTR |
b09e0fa4b
|
3783 |
.listxattr = shmem_listxattr, |
feda821e7
|
3784 |
.set_acl = simple_set_acl, |
b09e0fa4b
|
3785 |
#endif |
1da177e4c
|
3786 |
}; |
92e1d5be9
|
3787 |
static const struct inode_operations shmem_dir_inode_operations = { |
1da177e4c
|
3788 3789 3790 3791 3792 3793 3794 3795 3796 |
#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
|
3797 |
.rename = shmem_rename2, |
60545d0d4
|
3798 |
.tmpfile = shmem_tmpfile, |
1da177e4c
|
3799 |
#endif |
b09e0fa4b
|
3800 |
#ifdef CONFIG_TMPFS_XATTR |
b09e0fa4b
|
3801 |
.listxattr = shmem_listxattr, |
b09e0fa4b
|
3802 |
#endif |
39f0247d3
|
3803 |
#ifdef CONFIG_TMPFS_POSIX_ACL |
94c1e62df
|
3804 |
.setattr = shmem_setattr, |
feda821e7
|
3805 |
.set_acl = simple_set_acl, |
39f0247d3
|
3806 3807 |
#endif }; |
92e1d5be9
|
3808 |
static const struct inode_operations shmem_special_inode_operations = { |
b09e0fa4b
|
3809 |
#ifdef CONFIG_TMPFS_XATTR |
b09e0fa4b
|
3810 |
.listxattr = shmem_listxattr, |
b09e0fa4b
|
3811 |
#endif |
39f0247d3
|
3812 |
#ifdef CONFIG_TMPFS_POSIX_ACL |
94c1e62df
|
3813 |
.setattr = shmem_setattr, |
feda821e7
|
3814 |
.set_acl = simple_set_acl, |
39f0247d3
|
3815 |
#endif |
1da177e4c
|
3816 |
}; |
759b9775c
|
3817 |
static const struct super_operations shmem_ops = { |
1da177e4c
|
3818 3819 3820 3821 3822 |
.alloc_inode = shmem_alloc_inode, .destroy_inode = shmem_destroy_inode, #ifdef CONFIG_TMPFS .statfs = shmem_statfs, .remount_fs = shmem_remount_fs, |
680d794ba
|
3823 |
.show_options = shmem_show_options, |
1da177e4c
|
3824 |
#endif |
1f895f75d
|
3825 |
.evict_inode = shmem_evict_inode, |
1da177e4c
|
3826 3827 |
.drop_inode = generic_delete_inode, .put_super = shmem_put_super, |
779750d20
|
3828 3829 3830 3831 |
#ifdef CONFIG_TRANSPARENT_HUGE_PAGECACHE .nr_cached_objects = shmem_unused_huge_count, .free_cached_objects = shmem_unused_huge_scan, #endif |
1da177e4c
|
3832 |
}; |
f0f37e2f7
|
3833 |
static const struct vm_operations_struct shmem_vm_ops = { |
54cb8821d
|
3834 |
.fault = shmem_fault, |
d7c175517
|
3835 |
.map_pages = filemap_map_pages, |
1da177e4c
|
3836 3837 3838 3839 3840 |
#ifdef CONFIG_NUMA .set_policy = shmem_set_policy, .get_policy = shmem_get_policy, #endif }; |
3c26ff6e4
|
3841 3842 |
static struct dentry *shmem_mount(struct file_system_type *fs_type, int flags, const char *dev_name, void *data) |
1da177e4c
|
3843 |
{ |
3c26ff6e4
|
3844 |
return mount_nodev(fs_type, flags, data, shmem_fill_super); |
1da177e4c
|
3845 |
} |
41ffe5d5c
|
3846 |
static struct file_system_type shmem_fs_type = { |
1da177e4c
|
3847 3848 |
.owner = THIS_MODULE, .name = "tmpfs", |
3c26ff6e4
|
3849 |
.mount = shmem_mount, |
1da177e4c
|
3850 |
.kill_sb = kill_litter_super, |
2b8576cb0
|
3851 |
.fs_flags = FS_USERNS_MOUNT, |
1da177e4c
|
3852 |
}; |
1da177e4c
|
3853 |
|
41ffe5d5c
|
3854 |
int __init shmem_init(void) |
1da177e4c
|
3855 3856 |
{ int error; |
16203a7a9
|
3857 3858 3859 |
/* If rootfs called this, don't re-init */ if (shmem_inode_cachep) return 0; |
41ffe5d5c
|
3860 |
error = shmem_init_inodecache(); |
1da177e4c
|
3861 3862 |
if (error) goto out3; |
41ffe5d5c
|
3863 |
error = register_filesystem(&shmem_fs_type); |
1da177e4c
|
3864 |
if (error) { |
1170532bb
|
3865 3866 |
pr_err("Could not register tmpfs "); |
1da177e4c
|
3867 3868 |
goto out2; } |
95dc112a5
|
3869 |
|
ca4e05195
|
3870 |
shm_mnt = kern_mount(&shmem_fs_type); |
1da177e4c
|
3871 3872 |
if (IS_ERR(shm_mnt)) { error = PTR_ERR(shm_mnt); |
1170532bb
|
3873 3874 |
pr_err("Could not kern_mount tmpfs "); |
1da177e4c
|
3875 3876 |
goto out1; } |
5a6e75f81
|
3877 |
|
e496cf3d7
|
3878 |
#ifdef CONFIG_TRANSPARENT_HUGE_PAGECACHE |
435c0b87d
|
3879 |
if (has_transparent_hugepage() && shmem_huge > SHMEM_HUGE_DENY) |
5a6e75f81
|
3880 3881 3882 3883 |
SHMEM_SB(shm_mnt->mnt_sb)->huge = shmem_huge; else shmem_huge = 0; /* just in case it was patched */ #endif |
1da177e4c
|
3884 3885 3886 |
return 0; out1: |
41ffe5d5c
|
3887 |
unregister_filesystem(&shmem_fs_type); |
1da177e4c
|
3888 |
out2: |
41ffe5d5c
|
3889 |
shmem_destroy_inodecache(); |
1da177e4c
|
3890 3891 3892 3893 |
out3: shm_mnt = ERR_PTR(error); return error; } |
853ac43ab
|
3894 |
|
e496cf3d7
|
3895 |
#if defined(CONFIG_TRANSPARENT_HUGE_PAGECACHE) && defined(CONFIG_SYSFS) |
5a6e75f81
|
3896 3897 3898 3899 3900 3901 3902 3903 3904 3905 3906 3907 3908 3909 3910 3911 3912 3913 3914 3915 3916 3917 3918 3919 3920 3921 3922 3923 3924 3925 3926 3927 3928 3929 3930 3931 3932 3933 3934 3935 3936 3937 3938 3939 3940 3941 |
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; |
435c0b87d
|
3942 |
if (shmem_huge > SHMEM_HUGE_DENY) |
5a6e75f81
|
3943 3944 3945 3946 3947 3948 |
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
|
3949 |
#endif /* CONFIG_TRANSPARENT_HUGE_PAGECACHE && CONFIG_SYSFS */ |
f3f0e1d21
|
3950 |
|
3b33719c9
|
3951 |
#ifdef CONFIG_TRANSPARENT_HUGE_PAGECACHE |
f3f0e1d21
|
3952 3953 3954 3955 3956 3957 3958 3959 3960 3961 3962 3963 3964 3965 3966 3967 3968 3969 3970 3971 3972 3973 3974 3975 3976 3977 3978 3979 3980 3981 |
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
|
3982 |
#endif /* CONFIG_TRANSPARENT_HUGE_PAGECACHE */ |
5a6e75f81
|
3983 |
|
853ac43ab
|
3984 3985 3986 3987 3988 3989 3990 3991 3992 3993 |
#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
|
3994 |
static struct file_system_type shmem_fs_type = { |
853ac43ab
|
3995 |
.name = "tmpfs", |
3c26ff6e4
|
3996 |
.mount = ramfs_mount, |
853ac43ab
|
3997 |
.kill_sb = kill_litter_super, |
2b8576cb0
|
3998 |
.fs_flags = FS_USERNS_MOUNT, |
853ac43ab
|
3999 |
}; |
41ffe5d5c
|
4000 |
int __init shmem_init(void) |
853ac43ab
|
4001 |
{ |
41ffe5d5c
|
4002 |
BUG_ON(register_filesystem(&shmem_fs_type) != 0); |
853ac43ab
|
4003 |
|
41ffe5d5c
|
4004 |
shm_mnt = kern_mount(&shmem_fs_type); |
853ac43ab
|
4005 4006 4007 4008 |
BUG_ON(IS_ERR(shm_mnt)); return 0; } |
41ffe5d5c
|
4009 |
int shmem_unuse(swp_entry_t swap, struct page *page) |
853ac43ab
|
4010 4011 4012 |
{ return 0; } |
3f96b79ad
|
4013 4014 4015 4016 |
int shmem_lock(struct file *file, int lock, struct user_struct *user) { return 0; } |
245132643
|
4017 4018 4019 |
void shmem_unlock_mapping(struct address_space *mapping) { } |
c01d5b300
|
4020 4021 4022 4023 4024 4025 4026 4027 |
#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
|
4028 |
void shmem_truncate_range(struct inode *inode, loff_t lstart, loff_t lend) |
94c1e62df
|
4029 |
{ |
41ffe5d5c
|
4030 |
truncate_inode_pages_range(inode->i_mapping, lstart, lend); |
94c1e62df
|
4031 4032 |
} EXPORT_SYMBOL_GPL(shmem_truncate_range); |
0b0a0806b
|
4033 4034 |
#define shmem_vm_ops generic_file_vm_ops #define shmem_file_operations ramfs_file_operations |
454abafe9
|
4035 |
#define shmem_get_inode(sb, dir, mode, dev, flags) ramfs_get_inode(sb, dir, mode, dev) |
0b0a0806b
|
4036 4037 |
#define shmem_acct_size(flags, size) 0 #define shmem_unacct_size(flags, size) do {} while (0) |
853ac43ab
|
4038 4039 4040 4041 |
#endif /* CONFIG_SHMEM */ /* common code */ |
1da177e4c
|
4042 |
|
19938e350
|
4043 |
static const struct dentry_operations anon_ops = { |
118b23022
|
4044 |
.d_dname = simple_dname |
3451538a1
|
4045 |
}; |
c72770909
|
4046 4047 |
static struct file *__shmem_file_setup(const char *name, loff_t size, unsigned long flags, unsigned int i_flags) |
1da177e4c
|
4048 |
{ |
6b4d0b279
|
4049 |
struct file *res; |
1da177e4c
|
4050 |
struct inode *inode; |
2c48b9c45
|
4051 |
struct path path; |
3451538a1
|
4052 |
struct super_block *sb; |
1da177e4c
|
4053 4054 4055 |
struct qstr this; if (IS_ERR(shm_mnt)) |
6b4d0b279
|
4056 |
return ERR_CAST(shm_mnt); |
1da177e4c
|
4057 |
|
285b2c4fd
|
4058 |
if (size < 0 || size > MAX_LFS_FILESIZE) |
1da177e4c
|
4059 4060 4061 4062 |
return ERR_PTR(-EINVAL); if (shmem_acct_size(flags, size)) return ERR_PTR(-ENOMEM); |
6b4d0b279
|
4063 |
res = ERR_PTR(-ENOMEM); |
1da177e4c
|
4064 4065 4066 |
this.name = name; this.len = strlen(name); this.hash = 0; /* will go */ |
3451538a1
|
4067 |
sb = shm_mnt->mnt_sb; |
66ee4b888
|
4068 |
path.mnt = mntget(shm_mnt); |
3451538a1
|
4069 |
path.dentry = d_alloc_pseudo(sb, &this); |
2c48b9c45
|
4070 |
if (!path.dentry) |
1da177e4c
|
4071 |
goto put_memory; |
3451538a1
|
4072 |
d_set_d_op(path.dentry, &anon_ops); |
1da177e4c
|
4073 |
|
6b4d0b279
|
4074 |
res = ERR_PTR(-ENOSPC); |
3451538a1
|
4075 |
inode = shmem_get_inode(sb, NULL, S_IFREG | S_IRWXUGO, 0, flags); |
1da177e4c
|
4076 |
if (!inode) |
66ee4b888
|
4077 |
goto put_memory; |
1da177e4c
|
4078 |
|
c72770909
|
4079 |
inode->i_flags |= i_flags; |
2c48b9c45
|
4080 |
d_instantiate(path.dentry, inode); |
1da177e4c
|
4081 |
inode->i_size = size; |
6d6b77f16
|
4082 |
clear_nlink(inode); /* It is unlinked */ |
26567cdbb
|
4083 4084 |
res = ERR_PTR(ramfs_nommu_expand_for_mapping(inode, size)); if (IS_ERR(res)) |
66ee4b888
|
4085 |
goto put_path; |
4b42af81f
|
4086 |
|
6b4d0b279
|
4087 |
res = alloc_file(&path, FMODE_WRITE | FMODE_READ, |
4b42af81f
|
4088 |
&shmem_file_operations); |
6b4d0b279
|
4089 |
if (IS_ERR(res)) |
66ee4b888
|
4090 |
goto put_path; |
4b42af81f
|
4091 |
|
6b4d0b279
|
4092 |
return res; |
1da177e4c
|
4093 |
|
1da177e4c
|
4094 4095 |
put_memory: shmem_unacct_size(flags, size); |
66ee4b888
|
4096 4097 |
put_path: path_put(&path); |
6b4d0b279
|
4098 |
return res; |
1da177e4c
|
4099 |
} |
c72770909
|
4100 4101 4102 4103 4104 |
/** * 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
|
4105 4106 |
* 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
|
4107 4108 4109 4110 4111 4112 4113 4114 4115 4116 4117 4118 4119 4120 4121 4122 4123 4124 4125 |
* @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
|
4126 |
EXPORT_SYMBOL_GPL(shmem_file_setup); |
1da177e4c
|
4127 |
|
467118102
|
4128 |
/** |
1da177e4c
|
4129 |
* shmem_zero_setup - setup a shared anonymous mapping |
1da177e4c
|
4130 4131 4132 4133 4134 4135 |
* @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
|
4136 4137 4138 4139 4140 4141 4142 |
/* * 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
|
4143 4144 4145 4146 4147 4148 4149 |
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
|
4150 |
|
e496cf3d7
|
4151 |
if (IS_ENABLED(CONFIG_TRANSPARENT_HUGE_PAGECACHE) && |
f3f0e1d21
|
4152 4153 4154 4155 |
((vma->vm_start + ~HPAGE_PMD_MASK) & HPAGE_PMD_MASK) < (vma->vm_end & HPAGE_PMD_MASK)) { khugepaged_enter(vma, vma->vm_flags); } |
1da177e4c
|
4156 4157 |
return 0; } |
d9d90e5eb
|
4158 4159 4160 4161 4162 4163 4164 4165 4166 4167 4168 4169 4170 |
/** * 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
|
4171 4172 |
* i915_gem_object_get_pages_gtt() mixes __GFP_NORETRY | __GFP_NOWARN in * with the mapping_gfp_mask(), to avoid OOMing the machine unnecessarily. |
d9d90e5eb
|
4173 4174 4175 4176 |
*/ struct page *shmem_read_mapping_page_gfp(struct address_space *mapping, pgoff_t index, gfp_t gfp) { |
68da9f055
|
4177 4178 |
#ifdef CONFIG_SHMEM struct inode *inode = mapping->host; |
9276aad6c
|
4179 |
struct page *page; |
68da9f055
|
4180 4181 4182 |
int error; BUG_ON(mapping->a_ops != &shmem_aops); |
9e18eb293
|
4183 |
error = shmem_getpage_gfp(inode, index, &page, SGP_CACHE, |
cfda05267
|
4184 |
gfp, NULL, NULL, NULL); |
68da9f055
|
4185 4186 4187 4188 4189 4190 4191 4192 4193 |
if (error) page = ERR_PTR(error); else unlock_page(page); return page; #else /* * The tiny !SHMEM case uses ramfs without swap */ |
d9d90e5eb
|
4194 |
return read_cache_page_gfp(mapping, index, gfp); |
68da9f055
|
4195 |
#endif |
d9d90e5eb
|
4196 4197 |
} EXPORT_SYMBOL_GPL(shmem_read_mapping_page_gfp); |