Blame view
mm/swapfile.c
77.4 KB
1da177e4c
|
1 2 3 4 5 6 |
/* * linux/mm/swapfile.c * * Copyright (C) 1991, 1992, 1993, 1994 Linus Torvalds * Swap reorganised 29.12.95, Stephen Tweedie */ |
1da177e4c
|
7 8 9 10 11 12 13 14 15 |
#include <linux/mm.h> #include <linux/hugetlb.h> #include <linux/mman.h> #include <linux/slab.h> #include <linux/kernel_stat.h> #include <linux/swap.h> #include <linux/vmalloc.h> #include <linux/pagemap.h> #include <linux/namei.h> |
072441e21
|
16 |
#include <linux/shmem_fs.h> |
1da177e4c
|
17 |
#include <linux/blkdev.h> |
20137a490
|
18 |
#include <linux/random.h> |
1da177e4c
|
19 20 21 22 |
#include <linux/writeback.h> #include <linux/proc_fs.h> #include <linux/seq_file.h> #include <linux/init.h> |
5ad646880
|
23 |
#include <linux/ksm.h> |
1da177e4c
|
24 25 26 |
#include <linux/rmap.h> #include <linux/security.h> #include <linux/backing-dev.h> |
fc0abb145
|
27 |
#include <linux/mutex.h> |
c59ede7b7
|
28 |
#include <linux/capability.h> |
1da177e4c
|
29 |
#include <linux/syscalls.h> |
8a9f3ccd2
|
30 |
#include <linux/memcontrol.h> |
66d7dd518
|
31 |
#include <linux/poll.h> |
72788c385
|
32 |
#include <linux/oom.h> |
38b5faf4b
|
33 34 |
#include <linux/frontswap.h> #include <linux/swapfile.h> |
f981c5950
|
35 |
#include <linux/export.h> |
1da177e4c
|
36 37 38 39 |
#include <asm/pgtable.h> #include <asm/tlbflush.h> #include <linux/swapops.h> |
5d1ea48bd
|
40 |
#include <linux/swap_cgroup.h> |
1da177e4c
|
41 |
|
570a335b8
|
42 43 44 |
static bool swap_count_continued(struct swap_info_struct *, pgoff_t, unsigned char); static void free_swap_count_continuations(struct swap_info_struct *); |
d4906e1aa
|
45 |
static sector_t map_swap_entry(swp_entry_t, struct block_device**); |
570a335b8
|
46 |
|
38b5faf4b
|
47 |
DEFINE_SPINLOCK(swap_lock); |
7c363b8c6
|
48 |
static unsigned int nr_swapfiles; |
ec8acf20a
|
49 |
atomic_long_t nr_swap_pages; |
fb0fec501
|
50 51 52 53 54 55 |
/* * Some modules use swappable objects and may try to swap them out under * memory pressure (via the shrinker). Before doing so, they may wish to * check to see if any swap space is available. */ EXPORT_SYMBOL_GPL(nr_swap_pages); |
ec8acf20a
|
56 |
/* protected with swap_lock. reading in vm_swap_full() doesn't need lock */ |
1da177e4c
|
57 |
long total_swap_pages; |
78ecba081
|
58 |
static int least_priority; |
1da177e4c
|
59 |
|
1da177e4c
|
60 61 62 63 |
static const char Bad_file[] = "Bad swap file entry "; static const char Unused_file[] = "Unused swap file entry "; static const char Bad_offset[] = "Bad swap offset entry "; static const char Unused_offset[] = "Unused swap offset entry "; |
adfab836f
|
64 65 66 67 |
/* * all active swap_info_structs * protected with swap_lock, and ordered by priority. */ |
18ab4d4ce
|
68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 |
PLIST_HEAD(swap_active_head); /* * all available (active, not full) swap_info_structs * protected with swap_avail_lock, ordered by priority. * This is used by get_swap_page() instead of swap_active_head * because swap_active_head includes all swap_info_structs, * but get_swap_page() doesn't need to look at full ones. * This uses its own lock instead of swap_lock because when a * swap_info_struct changes between not-full/full, it needs to * add/remove itself to/from this list, but the swap_info_struct->lock * is held and the locking order requires swap_lock to be taken * before any swap_info_struct->lock. */ static PLIST_HEAD(swap_avail_head); static DEFINE_SPINLOCK(swap_avail_lock); |
1da177e4c
|
84 |
|
38b5faf4b
|
85 |
struct swap_info_struct *swap_info[MAX_SWAPFILES]; |
1da177e4c
|
86 |
|
fc0abb145
|
87 |
static DEFINE_MUTEX(swapon_mutex); |
1da177e4c
|
88 |
|
66d7dd518
|
89 90 91 |
static DECLARE_WAIT_QUEUE_HEAD(proc_poll_wait); /* Activity counter to indicate that a swapon or swapoff has occurred */ static atomic_t proc_poll_event = ATOMIC_INIT(0); |
8d69aaee8
|
92 |
static inline unsigned char swap_count(unsigned char ent) |
355cfa73d
|
93 |
{ |
570a335b8
|
94 |
return ent & ~SWAP_HAS_CACHE; /* may include SWAP_HAS_CONT flag */ |
355cfa73d
|
95 |
} |
efa90a981
|
96 |
/* returns 1 if swap entry is freed */ |
c9e444103
|
97 98 99 |
static int __try_to_reclaim_swap(struct swap_info_struct *si, unsigned long offset) { |
efa90a981
|
100 |
swp_entry_t entry = swp_entry(si->type, offset); |
c9e444103
|
101 102 |
struct page *page; int ret = 0; |
33806f06d
|
103 |
page = find_get_page(swap_address_space(entry), entry.val); |
c9e444103
|
104 105 106 107 108 109 110 111 112 113 114 115 116 |
if (!page) return 0; /* * This function is called from scan_swap_map() and it's called * by vmscan.c at reclaiming pages. So, we hold a lock on a page, here. * We have to use trylock for avoiding deadlock. This is a special * case and you should use try_to_free_swap() with explicit lock_page() * in usual operations. */ if (trylock_page(page)) { ret = try_to_free_swap(page); unlock_page(page); } |
09cbfeaf1
|
117 |
put_page(page); |
c9e444103
|
118 119 |
return ret; } |
355cfa73d
|
120 |
|
1da177e4c
|
121 |
/* |
6a6ba8317
|
122 123 124 125 126 127 |
* swapon tell device that all the old swap contents can be discarded, * to allow the swap device to optimize its wear-levelling. */ static int discard_swap(struct swap_info_struct *si) { struct swap_extent *se; |
9625a5f28
|
128 129 |
sector_t start_block; sector_t nr_blocks; |
6a6ba8317
|
130 |
int err = 0; |
9625a5f28
|
131 132 133 134 135 136 |
/* Do not discard the swap header page! */ se = &si->first_swap_extent; start_block = (se->start_block + 1) << (PAGE_SHIFT - 9); nr_blocks = ((sector_t)se->nr_pages - 1) << (PAGE_SHIFT - 9); if (nr_blocks) { err = blkdev_issue_discard(si->bdev, start_block, |
dd3932edd
|
137 |
nr_blocks, GFP_KERNEL, 0); |
9625a5f28
|
138 139 140 141 |
if (err) return err; cond_resched(); } |
6a6ba8317
|
142 |
|
9625a5f28
|
143 144 145 |
list_for_each_entry(se, &si->first_swap_extent.list, list) { start_block = se->start_block << (PAGE_SHIFT - 9); nr_blocks = (sector_t)se->nr_pages << (PAGE_SHIFT - 9); |
6a6ba8317
|
146 147 |
err = blkdev_issue_discard(si->bdev, start_block, |
dd3932edd
|
148 |
nr_blocks, GFP_KERNEL, 0); |
6a6ba8317
|
149 150 151 152 153 154 155 |
if (err) break; cond_resched(); } return err; /* That will often be -EOPNOTSUPP */ } |
7992fde72
|
156 157 158 159 160 161 162 163 164 165 166 |
/* * swap allocation tell device that a cluster of swap can now be discarded, * to allow the swap device to optimize its wear-levelling. */ static void discard_swap_cluster(struct swap_info_struct *si, pgoff_t start_page, pgoff_t nr_pages) { struct swap_extent *se = si->curr_swap_extent; int found_extent = 0; while (nr_pages) { |
7992fde72
|
167 168 169 170 |
if (se->start_page <= start_page && start_page < se->start_page + se->nr_pages) { pgoff_t offset = start_page - se->start_page; sector_t start_block = se->start_block + offset; |
858a29900
|
171 |
sector_t nr_blocks = se->nr_pages - offset; |
7992fde72
|
172 173 174 175 176 177 178 179 180 181 182 183 |
if (nr_blocks > nr_pages) nr_blocks = nr_pages; start_page += nr_blocks; nr_pages -= nr_blocks; if (!found_extent++) si->curr_swap_extent = se; start_block <<= PAGE_SHIFT - 9; nr_blocks <<= PAGE_SHIFT - 9; if (blkdev_issue_discard(si->bdev, start_block, |
dd3932edd
|
184 |
nr_blocks, GFP_NOIO, 0)) |
7992fde72
|
185 186 |
break; } |
a8ae49917
|
187 |
se = list_next_entry(se, list); |
7992fde72
|
188 189 |
} } |
048c27fd7
|
190 191 |
#define SWAPFILE_CLUSTER 256 #define LATENCY_LIMIT 256 |
2a8f94493
|
192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 |
static inline void cluster_set_flag(struct swap_cluster_info *info, unsigned int flag) { info->flags = flag; } static inline unsigned int cluster_count(struct swap_cluster_info *info) { return info->data; } static inline void cluster_set_count(struct swap_cluster_info *info, unsigned int c) { info->data = c; } static inline void cluster_set_count_flag(struct swap_cluster_info *info, unsigned int c, unsigned int f) { info->flags = f; info->data = c; } static inline unsigned int cluster_next(struct swap_cluster_info *info) { return info->data; } static inline void cluster_set_next(struct swap_cluster_info *info, unsigned int n) { info->data = n; } static inline void cluster_set_next_flag(struct swap_cluster_info *info, unsigned int n, unsigned int f) { info->flags = f; info->data = n; } static inline bool cluster_is_free(struct swap_cluster_info *info) { return info->flags & CLUSTER_FLAG_FREE; } static inline bool cluster_is_null(struct swap_cluster_info *info) { return info->flags & CLUSTER_FLAG_NEXT_NULL; } static inline void cluster_set_null(struct swap_cluster_info *info) { info->flags = CLUSTER_FLAG_NEXT_NULL; info->data = 0; } |
815c2c543
|
249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 |
/* Add a cluster to discard list and schedule it to do discard */ static void swap_cluster_schedule_discard(struct swap_info_struct *si, unsigned int idx) { /* * If scan_swap_map() can't find a free cluster, it will check * si->swap_map directly. To make sure the discarding cluster isn't * taken by scan_swap_map(), mark the swap entries bad (occupied). It * will be cleared after discard */ memset(si->swap_map + idx * SWAPFILE_CLUSTER, SWAP_MAP_BAD, SWAPFILE_CLUSTER); if (cluster_is_null(&si->discard_cluster_head)) { cluster_set_next_flag(&si->discard_cluster_head, idx, 0); cluster_set_next_flag(&si->discard_cluster_tail, idx, 0); } else { unsigned int tail = cluster_next(&si->discard_cluster_tail); cluster_set_next(&si->cluster_info[tail], idx); cluster_set_next_flag(&si->discard_cluster_tail, idx, 0); } schedule_work(&si->discard_work); } /* * Doing discard actually. After a cluster discard is finished, the cluster * will be added to free cluster list. caller should hold si->lock. */ static void swap_do_scheduled_discard(struct swap_info_struct *si) { struct swap_cluster_info *info; unsigned int idx; info = si->cluster_info; while (!cluster_is_null(&si->discard_cluster_head)) { idx = cluster_next(&si->discard_cluster_head); cluster_set_next_flag(&si->discard_cluster_head, cluster_next(&info[idx]), 0); if (cluster_next(&si->discard_cluster_tail) == idx) { cluster_set_null(&si->discard_cluster_head); cluster_set_null(&si->discard_cluster_tail); } spin_unlock(&si->lock); discard_swap_cluster(si, idx * SWAPFILE_CLUSTER, SWAPFILE_CLUSTER); spin_lock(&si->lock); cluster_set_flag(&info[idx], CLUSTER_FLAG_FREE); if (cluster_is_null(&si->free_cluster_head)) { cluster_set_next_flag(&si->free_cluster_head, idx, 0); cluster_set_next_flag(&si->free_cluster_tail, idx, 0); } else { unsigned int tail; tail = cluster_next(&si->free_cluster_tail); cluster_set_next(&info[tail], idx); cluster_set_next_flag(&si->free_cluster_tail, idx, 0); } memset(si->swap_map + idx * SWAPFILE_CLUSTER, 0, SWAPFILE_CLUSTER); } } static void swap_discard_work(struct work_struct *work) { struct swap_info_struct *si; si = container_of(work, struct swap_info_struct, discard_work); spin_lock(&si->lock); swap_do_scheduled_discard(si); spin_unlock(&si->lock); } |
2a8f94493
|
332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 |
/* * The cluster corresponding to page_nr will be used. The cluster will be * removed from free cluster list and its usage counter will be increased. */ static void inc_cluster_info_page(struct swap_info_struct *p, struct swap_cluster_info *cluster_info, unsigned long page_nr) { unsigned long idx = page_nr / SWAPFILE_CLUSTER; if (!cluster_info) return; if (cluster_is_free(&cluster_info[idx])) { VM_BUG_ON(cluster_next(&p->free_cluster_head) != idx); cluster_set_next_flag(&p->free_cluster_head, cluster_next(&cluster_info[idx]), 0); if (cluster_next(&p->free_cluster_tail) == idx) { cluster_set_null(&p->free_cluster_tail); cluster_set_null(&p->free_cluster_head); } cluster_set_count_flag(&cluster_info[idx], 0, 0); } VM_BUG_ON(cluster_count(&cluster_info[idx]) >= SWAPFILE_CLUSTER); cluster_set_count(&cluster_info[idx], cluster_count(&cluster_info[idx]) + 1); } /* * The cluster corresponding to page_nr decreases one usage. If the usage * counter becomes 0, which means no page in the cluster is in using, we can * optionally discard the cluster and add it to free cluster list. */ static void dec_cluster_info_page(struct swap_info_struct *p, struct swap_cluster_info *cluster_info, unsigned long page_nr) { unsigned long idx = page_nr / SWAPFILE_CLUSTER; if (!cluster_info) return; VM_BUG_ON(cluster_count(&cluster_info[idx]) == 0); cluster_set_count(&cluster_info[idx], cluster_count(&cluster_info[idx]) - 1); if (cluster_count(&cluster_info[idx]) == 0) { |
815c2c543
|
377 378 379 380 381 |
/* * If the swap is discardable, prepare discard the cluster * instead of free it immediately. The cluster will be freed * after discard. */ |
edfe23dac
|
382 383 |
if ((p->flags & (SWP_WRITEOK | SWP_PAGE_DISCARD)) == (SWP_WRITEOK | SWP_PAGE_DISCARD)) { |
815c2c543
|
384 385 386 |
swap_cluster_schedule_discard(p, idx); return; } |
2a8f94493
|
387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 |
cluster_set_flag(&cluster_info[idx], CLUSTER_FLAG_FREE); if (cluster_is_null(&p->free_cluster_head)) { cluster_set_next_flag(&p->free_cluster_head, idx, 0); cluster_set_next_flag(&p->free_cluster_tail, idx, 0); } else { unsigned int tail = cluster_next(&p->free_cluster_tail); cluster_set_next(&cluster_info[tail], idx); cluster_set_next_flag(&p->free_cluster_tail, idx, 0); } } } /* * It's possible scan_swap_map() uses a free cluster in the middle of free * cluster list. Avoiding such abuse to avoid list corruption. */ |
ebc2a1a69
|
403 404 |
static bool scan_swap_map_ssd_cluster_conflict(struct swap_info_struct *si, |
2a8f94493
|
405 406 |
unsigned long offset) { |
ebc2a1a69
|
407 408 |
struct percpu_cluster *percpu_cluster; bool conflict; |
2a8f94493
|
409 |
offset /= SWAPFILE_CLUSTER; |
ebc2a1a69
|
410 |
conflict = !cluster_is_null(&si->free_cluster_head) && |
2a8f94493
|
411 412 |
offset != cluster_next(&si->free_cluster_head) && cluster_is_free(&si->cluster_info[offset]); |
ebc2a1a69
|
413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 |
if (!conflict) return false; percpu_cluster = this_cpu_ptr(si->percpu_cluster); cluster_set_null(&percpu_cluster->index); return true; } /* * Try to get a swap entry from current cpu's swap entry pool (a cluster). This * might involve allocating a new cluster for current CPU too. */ static void scan_swap_map_try_ssd_cluster(struct swap_info_struct *si, unsigned long *offset, unsigned long *scan_base) { struct percpu_cluster *cluster; bool found_free; unsigned long tmp; new_cluster: cluster = this_cpu_ptr(si->percpu_cluster); if (cluster_is_null(&cluster->index)) { if (!cluster_is_null(&si->free_cluster_head)) { cluster->index = si->free_cluster_head; cluster->next = cluster_next(&cluster->index) * SWAPFILE_CLUSTER; } else if (!cluster_is_null(&si->discard_cluster_head)) { /* * we don't have free cluster but have some clusters in * discarding, do discard now and reclaim them */ swap_do_scheduled_discard(si); *scan_base = *offset = si->cluster_next; goto new_cluster; } else return; } found_free = false; /* * Other CPUs can use our cluster if they can't find a free cluster, * check if there is still free entry in the cluster */ tmp = cluster->next; while (tmp < si->max && tmp < (cluster_next(&cluster->index) + 1) * SWAPFILE_CLUSTER) { if (!si->swap_map[tmp]) { found_free = true; break; } tmp++; } if (!found_free) { cluster_set_null(&cluster->index); goto new_cluster; } cluster->next = tmp + 1; *offset = tmp; *scan_base = tmp; |
2a8f94493
|
474 |
} |
24b8ff7c2
|
475 476 |
static unsigned long scan_swap_map(struct swap_info_struct *si, unsigned char usage) |
1da177e4c
|
477 |
{ |
ebebbbe90
|
478 |
unsigned long offset; |
c60aa176c
|
479 |
unsigned long scan_base; |
7992fde72
|
480 |
unsigned long last_in_cluster = 0; |
048c27fd7
|
481 |
int latency_ration = LATENCY_LIMIT; |
7dfad4183
|
482 |
|
886bb7e9c
|
483 |
/* |
7dfad4183
|
484 485 486 487 488 489 490 |
* We try to cluster swap pages by allocating them sequentially * in swap. Once we've allocated SWAPFILE_CLUSTER pages this * way, however, we resort to first-free allocation, starting * a new cluster. This prevents us from scattering swap pages * all over the entire swap partition, so that we reduce * overall disk seek times between swap pages. -- sct * But we do now try to find an empty cluster. -Andrea |
c60aa176c
|
491 |
* And we let swap pages go all over an SSD partition. Hugh |
7dfad4183
|
492 |
*/ |
52b7efdbe
|
493 |
si->flags += SWP_SCANNING; |
c60aa176c
|
494 |
scan_base = offset = si->cluster_next; |
ebebbbe90
|
495 |
|
ebc2a1a69
|
496 497 498 499 500 |
/* SSD algorithm */ if (si->cluster_info) { scan_swap_map_try_ssd_cluster(si, &offset, &scan_base); goto checks; } |
ebebbbe90
|
501 502 503 504 505 |
if (unlikely(!si->cluster_nr--)) { if (si->pages - si->inuse_pages < SWAPFILE_CLUSTER) { si->cluster_nr = SWAPFILE_CLUSTER - 1; goto checks; } |
2a8f94493
|
506 |
|
ec8acf20a
|
507 |
spin_unlock(&si->lock); |
7dfad4183
|
508 |
|
c60aa176c
|
509 510 511 |
/* * If seek is expensive, start searching for new cluster from * start of partition, to minimize the span of allocated swap. |
50088c440
|
512 513 |
* If seek is cheap, that is the SWP_SOLIDSTATE si->cluster_info * case, just handled by scan_swap_map_try_ssd_cluster() above. |
c60aa176c
|
514 |
*/ |
50088c440
|
515 |
scan_base = offset = si->lowest_bit; |
7dfad4183
|
516 517 518 519 |
last_in_cluster = offset + SWAPFILE_CLUSTER - 1; /* Locate the first empty (unaligned) cluster */ for (; last_in_cluster <= si->highest_bit; offset++) { |
1da177e4c
|
520 |
if (si->swap_map[offset]) |
7dfad4183
|
521 522 |
last_in_cluster = offset + SWAPFILE_CLUSTER; else if (offset == last_in_cluster) { |
ec8acf20a
|
523 |
spin_lock(&si->lock); |
ebebbbe90
|
524 525 526 |
offset -= SWAPFILE_CLUSTER - 1; si->cluster_next = offset; si->cluster_nr = SWAPFILE_CLUSTER - 1; |
c60aa176c
|
527 528 529 530 531 532 533 534 535 |
goto checks; } if (unlikely(--latency_ration < 0)) { cond_resched(); latency_ration = LATENCY_LIMIT; } } offset = scan_base; |
ec8acf20a
|
536 |
spin_lock(&si->lock); |
ebebbbe90
|
537 |
si->cluster_nr = SWAPFILE_CLUSTER - 1; |
1da177e4c
|
538 |
} |
7dfad4183
|
539 |
|
ebebbbe90
|
540 |
checks: |
ebc2a1a69
|
541 542 543 544 |
if (si->cluster_info) { while (scan_swap_map_ssd_cluster_conflict(si, offset)) scan_swap_map_try_ssd_cluster(si, &offset, &scan_base); } |
ebebbbe90
|
545 |
if (!(si->flags & SWP_WRITEOK)) |
52b7efdbe
|
546 |
goto no_page; |
7dfad4183
|
547 548 |
if (!si->highest_bit) goto no_page; |
ebebbbe90
|
549 |
if (offset > si->highest_bit) |
c60aa176c
|
550 |
scan_base = offset = si->lowest_bit; |
c9e444103
|
551 |
|
b73d7fcec
|
552 553 |
/* reuse swap entry of cache-only swap if not busy. */ if (vm_swap_full() && si->swap_map[offset] == SWAP_HAS_CACHE) { |
c9e444103
|
554 |
int swap_was_freed; |
ec8acf20a
|
555 |
spin_unlock(&si->lock); |
c9e444103
|
556 |
swap_was_freed = __try_to_reclaim_swap(si, offset); |
ec8acf20a
|
557 |
spin_lock(&si->lock); |
c9e444103
|
558 559 560 561 562 |
/* entry was freed successfully, try to use this again */ if (swap_was_freed) goto checks; goto scan; /* check next one */ } |
ebebbbe90
|
563 564 565 566 567 568 569 570 571 572 573 |
if (si->swap_map[offset]) goto scan; if (offset == si->lowest_bit) si->lowest_bit++; if (offset == si->highest_bit) si->highest_bit--; si->inuse_pages++; if (si->inuse_pages == si->pages) { si->lowest_bit = si->max; si->highest_bit = 0; |
18ab4d4ce
|
574 575 576 |
spin_lock(&swap_avail_lock); plist_del(&si->avail_list, &swap_avail_head); spin_unlock(&swap_avail_lock); |
1da177e4c
|
577 |
} |
253d553ba
|
578 |
si->swap_map[offset] = usage; |
2a8f94493
|
579 |
inc_cluster_info_page(si, si->cluster_info, offset); |
ebebbbe90
|
580 581 |
si->cluster_next = offset + 1; si->flags -= SWP_SCANNING; |
7992fde72
|
582 |
|
ebebbbe90
|
583 |
return offset; |
7dfad4183
|
584 |
|
ebebbbe90
|
585 |
scan: |
ec8acf20a
|
586 |
spin_unlock(&si->lock); |
7dfad4183
|
587 |
while (++offset <= si->highest_bit) { |
52b7efdbe
|
588 |
if (!si->swap_map[offset]) { |
ec8acf20a
|
589 |
spin_lock(&si->lock); |
52b7efdbe
|
590 591 |
goto checks; } |
c9e444103
|
592 |
if (vm_swap_full() && si->swap_map[offset] == SWAP_HAS_CACHE) { |
ec8acf20a
|
593 |
spin_lock(&si->lock); |
c9e444103
|
594 595 |
goto checks; } |
048c27fd7
|
596 597 598 599 |
if (unlikely(--latency_ration < 0)) { cond_resched(); latency_ration = LATENCY_LIMIT; } |
7dfad4183
|
600 |
} |
c60aa176c
|
601 |
offset = si->lowest_bit; |
a5998061d
|
602 |
while (offset < scan_base) { |
c60aa176c
|
603 |
if (!si->swap_map[offset]) { |
ec8acf20a
|
604 |
spin_lock(&si->lock); |
c60aa176c
|
605 606 |
goto checks; } |
c9e444103
|
607 |
if (vm_swap_full() && si->swap_map[offset] == SWAP_HAS_CACHE) { |
ec8acf20a
|
608 |
spin_lock(&si->lock); |
c9e444103
|
609 610 |
goto checks; } |
c60aa176c
|
611 612 613 614 |
if (unlikely(--latency_ration < 0)) { cond_resched(); latency_ration = LATENCY_LIMIT; } |
a5998061d
|
615 |
offset++; |
c60aa176c
|
616 |
} |
ec8acf20a
|
617 |
spin_lock(&si->lock); |
7dfad4183
|
618 619 |
no_page: |
52b7efdbe
|
620 |
si->flags -= SWP_SCANNING; |
1da177e4c
|
621 622 623 624 625 |
return 0; } swp_entry_t get_swap_page(void) { |
adfab836f
|
626 |
struct swap_info_struct *si, *next; |
fb4f88dca
|
627 |
pgoff_t offset; |
1da177e4c
|
628 |
|
ec8acf20a
|
629 |
if (atomic_long_read(&nr_swap_pages) <= 0) |
fb4f88dca
|
630 |
goto noswap; |
ec8acf20a
|
631 |
atomic_long_dec(&nr_swap_pages); |
fb4f88dca
|
632 |
|
18ab4d4ce
|
633 634 635 636 637 638 639 |
spin_lock(&swap_avail_lock); start_over: plist_for_each_entry_safe(si, next, &swap_avail_head, avail_list) { /* requeue si to after same-priority siblings */ plist_requeue(&si->avail_list, &swap_avail_head); spin_unlock(&swap_avail_lock); |
ec8acf20a
|
640 |
spin_lock(&si->lock); |
adfab836f
|
641 |
if (!si->highest_bit || !(si->flags & SWP_WRITEOK)) { |
18ab4d4ce
|
642 643 644 645 646 647 648 649 650 651 652 653 654 655 |
spin_lock(&swap_avail_lock); if (plist_node_empty(&si->avail_list)) { spin_unlock(&si->lock); goto nextsi; } WARN(!si->highest_bit, "swap_info %d in list but !highest_bit ", si->type); WARN(!(si->flags & SWP_WRITEOK), "swap_info %d in list but !SWP_WRITEOK ", si->type); plist_del(&si->avail_list, &swap_avail_head); |
ec8acf20a
|
656 |
spin_unlock(&si->lock); |
18ab4d4ce
|
657 |
goto nextsi; |
ec8acf20a
|
658 |
} |
fb4f88dca
|
659 |
|
355cfa73d
|
660 |
/* This is called for allocating swap entry for cache */ |
253d553ba
|
661 |
offset = scan_swap_map(si, SWAP_HAS_CACHE); |
ec8acf20a
|
662 663 |
spin_unlock(&si->lock); if (offset) |
adfab836f
|
664 |
return swp_entry(si->type, offset); |
18ab4d4ce
|
665 666 667 668 669 |
pr_debug("scan_swap_map of si %d failed to find offset ", si->type); spin_lock(&swap_avail_lock); nextsi: |
adfab836f
|
670 671 672 673 |
/* * if we got here, it's likely that si was almost full before, * and since scan_swap_map() can drop the si->lock, multiple * callers probably all tried to get a page from the same si |
18ab4d4ce
|
674 675 676 677 678 |
* and it filled up before we could get one; or, the si filled * up between us dropping swap_avail_lock and taking si->lock. * Since we dropped the swap_avail_lock, the swap_avail_head * list may have been modified; so if next is still in the * swap_avail_head list then try it, otherwise start over. |
adfab836f
|
679 |
*/ |
18ab4d4ce
|
680 681 |
if (plist_node_empty(&next->avail_list)) goto start_over; |
1da177e4c
|
682 |
} |
fb4f88dca
|
683 |
|
18ab4d4ce
|
684 |
spin_unlock(&swap_avail_lock); |
ec8acf20a
|
685 |
atomic_long_inc(&nr_swap_pages); |
fb4f88dca
|
686 |
noswap: |
fb4f88dca
|
687 |
return (swp_entry_t) {0}; |
1da177e4c
|
688 |
} |
2de1a7e40
|
689 |
/* The only caller of this function is now suspend routine */ |
910321ea8
|
690 691 692 693 |
swp_entry_t get_swap_page_of_type(int type) { struct swap_info_struct *si; pgoff_t offset; |
910321ea8
|
694 |
si = swap_info[type]; |
ec8acf20a
|
695 |
spin_lock(&si->lock); |
910321ea8
|
696 |
if (si && (si->flags & SWP_WRITEOK)) { |
ec8acf20a
|
697 |
atomic_long_dec(&nr_swap_pages); |
910321ea8
|
698 699 700 |
/* This is called for allocating swap entry, not cache */ offset = scan_swap_map(si, 1); if (offset) { |
ec8acf20a
|
701 |
spin_unlock(&si->lock); |
910321ea8
|
702 703 |
return swp_entry(type, offset); } |
ec8acf20a
|
704 |
atomic_long_inc(&nr_swap_pages); |
910321ea8
|
705 |
} |
ec8acf20a
|
706 |
spin_unlock(&si->lock); |
910321ea8
|
707 708 |
return (swp_entry_t) {0}; } |
73c34b6ac
|
709 |
static struct swap_info_struct *swap_info_get(swp_entry_t entry) |
1da177e4c
|
710 |
{ |
73c34b6ac
|
711 |
struct swap_info_struct *p; |
1da177e4c
|
712 713 714 715 716 717 718 |
unsigned long offset, type; if (!entry.val) goto out; type = swp_type(entry); if (type >= nr_swapfiles) goto bad_nofile; |
efa90a981
|
719 |
p = swap_info[type]; |
1da177e4c
|
720 721 722 723 724 725 726 |
if (!(p->flags & SWP_USED)) goto bad_device; offset = swp_offset(entry); if (offset >= p->max) goto bad_offset; if (!p->swap_map[offset]) goto bad_free; |
ec8acf20a
|
727 |
spin_lock(&p->lock); |
1da177e4c
|
728 729 730 |
return p; bad_free: |
465c47fd8
|
731 732 |
pr_err("swap_free: %s%08lx ", Unused_offset, entry.val); |
1da177e4c
|
733 734 |
goto out; bad_offset: |
465c47fd8
|
735 736 |
pr_err("swap_free: %s%08lx ", Bad_offset, entry.val); |
1da177e4c
|
737 738 |
goto out; bad_device: |
465c47fd8
|
739 740 |
pr_err("swap_free: %s%08lx ", Unused_file, entry.val); |
1da177e4c
|
741 742 |
goto out; bad_nofile: |
465c47fd8
|
743 744 |
pr_err("swap_free: %s%08lx ", Bad_file, entry.val); |
1da177e4c
|
745 746 |
out: return NULL; |
886bb7e9c
|
747 |
} |
1da177e4c
|
748 |
|
8d69aaee8
|
749 750 |
static unsigned char swap_entry_free(struct swap_info_struct *p, swp_entry_t entry, unsigned char usage) |
1da177e4c
|
751 |
{ |
253d553ba
|
752 |
unsigned long offset = swp_offset(entry); |
8d69aaee8
|
753 754 |
unsigned char count; unsigned char has_cache; |
355cfa73d
|
755 |
|
253d553ba
|
756 757 758 |
count = p->swap_map[offset]; has_cache = count & SWAP_HAS_CACHE; count &= ~SWAP_HAS_CACHE; |
355cfa73d
|
759 |
|
253d553ba
|
760 |
if (usage == SWAP_HAS_CACHE) { |
355cfa73d
|
761 |
VM_BUG_ON(!has_cache); |
253d553ba
|
762 |
has_cache = 0; |
aaa468653
|
763 764 765 766 767 768 |
} else if (count == SWAP_MAP_SHMEM) { /* * Or we could insist on shmem.c using a special * swap_shmem_free() and free_shmem_swap_and_cache()... */ count = 0; |
570a335b8
|
769 770 771 772 773 774 775 776 777 |
} else if ((count & ~COUNT_CONTINUED) <= SWAP_MAP_MAX) { if (count == COUNT_CONTINUED) { if (swap_count_continued(p, offset, count)) count = SWAP_MAP_MAX | COUNT_CONTINUED; else count = SWAP_MAP_MAX; } else count--; } |
253d553ba
|
778 |
|
253d553ba
|
779 780 |
usage = count | has_cache; p->swap_map[offset] = usage; |
355cfa73d
|
781 |
|
355cfa73d
|
782 |
/* free if no reference */ |
253d553ba
|
783 |
if (!usage) { |
37e843511
|
784 |
mem_cgroup_uncharge_swap(entry); |
2a8f94493
|
785 |
dec_cluster_info_page(p, p->cluster_info, offset); |
355cfa73d
|
786 787 |
if (offset < p->lowest_bit) p->lowest_bit = offset; |
18ab4d4ce
|
788 789 |
if (offset > p->highest_bit) { bool was_full = !p->highest_bit; |
355cfa73d
|
790 |
p->highest_bit = offset; |
18ab4d4ce
|
791 792 793 794 795 796 797 798 799 |
if (was_full && (p->flags & SWP_WRITEOK)) { spin_lock(&swap_avail_lock); WARN_ON(!plist_node_empty(&p->avail_list)); if (plist_node_empty(&p->avail_list)) plist_add(&p->avail_list, &swap_avail_head); spin_unlock(&swap_avail_lock); } } |
ec8acf20a
|
800 |
atomic_long_inc(&nr_swap_pages); |
355cfa73d
|
801 |
p->inuse_pages--; |
38b5faf4b
|
802 |
frontswap_invalidate_page(p->type, offset); |
737449236
|
803 804 805 806 807 808 |
if (p->flags & SWP_BLKDEV) { struct gendisk *disk = p->bdev->bd_disk; if (disk->fops->swap_slot_free_notify) disk->fops->swap_slot_free_notify(p->bdev, offset); } |
1da177e4c
|
809 |
} |
253d553ba
|
810 811 |
return usage; |
1da177e4c
|
812 813 814 |
} /* |
2de1a7e40
|
815 |
* Caller has made sure that the swap device corresponding to entry |
1da177e4c
|
816 817 818 819 |
* is still around or has not been recycled. */ void swap_free(swp_entry_t entry) { |
73c34b6ac
|
820 |
struct swap_info_struct *p; |
1da177e4c
|
821 822 823 |
p = swap_info_get(entry); if (p) { |
253d553ba
|
824 |
swap_entry_free(p, entry, 1); |
ec8acf20a
|
825 |
spin_unlock(&p->lock); |
1da177e4c
|
826 827 828 829 |
} } /* |
cb4b86ba4
|
830 831 |
* Called after dropping swapcache to decrease refcnt to swap entries. */ |
0a31bc97c
|
832 |
void swapcache_free(swp_entry_t entry) |
cb4b86ba4
|
833 |
{ |
355cfa73d
|
834 |
struct swap_info_struct *p; |
355cfa73d
|
835 836 |
p = swap_info_get(entry); if (p) { |
0a31bc97c
|
837 |
swap_entry_free(p, entry, SWAP_HAS_CACHE); |
ec8acf20a
|
838 |
spin_unlock(&p->lock); |
355cfa73d
|
839 |
} |
cb4b86ba4
|
840 841 842 |
} /* |
c475a8ab6
|
843 |
* How many references to page are currently swapped out? |
570a335b8
|
844 845 |
* This does not give an exact answer when swap count is continued, * but does include the high COUNT_CONTINUED flag to allow for that. |
1da177e4c
|
846 |
*/ |
bde05d1cc
|
847 |
int page_swapcount(struct page *page) |
1da177e4c
|
848 |
{ |
c475a8ab6
|
849 850 |
int count = 0; struct swap_info_struct *p; |
1da177e4c
|
851 |
swp_entry_t entry; |
4c21e2f24
|
852 |
entry.val = page_private(page); |
1da177e4c
|
853 854 |
p = swap_info_get(entry); if (p) { |
355cfa73d
|
855 |
count = swap_count(p->swap_map[swp_offset(entry)]); |
ec8acf20a
|
856 |
spin_unlock(&p->lock); |
1da177e4c
|
857 |
} |
c475a8ab6
|
858 |
return count; |
1da177e4c
|
859 860 861 |
} /* |
8334b9622
|
862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 |
* How many references to @entry are currently swapped out? * This considers COUNT_CONTINUED so it returns exact answer. */ int swp_swapcount(swp_entry_t entry) { int count, tmp_count, n; struct swap_info_struct *p; struct page *page; pgoff_t offset; unsigned char *map; p = swap_info_get(entry); if (!p) return 0; count = swap_count(p->swap_map[swp_offset(entry)]); if (!(count & COUNT_CONTINUED)) goto out; count &= ~COUNT_CONTINUED; n = SWAP_MAP_MAX + 1; offset = swp_offset(entry); page = vmalloc_to_page(p->swap_map + offset); offset &= ~PAGE_MASK; VM_BUG_ON(page_private(page) != SWP_CONTINUED); do { |
a8ae49917
|
890 |
page = list_next_entry(page, lru); |
8334b9622
|
891 892 893 894 895 896 897 898 899 900 901 902 903 |
map = kmap_atomic(page); tmp_count = map[offset]; kunmap_atomic(map); count += (tmp_count & ~COUNT_CONTINUED) * n; n *= (SWAP_CONT_MAX + 1); } while (tmp_count & COUNT_CONTINUED); out: spin_unlock(&p->lock); return count; } /* |
7b1fe5979
|
904 905 906 907 |
* We can write to an anon page without COW if there are no other references * to it. And as a side-effect, free up its swap: because the old content * on disk will never be read, and seeking back there to write new content * later would only waste time away from clustering. |
1da177e4c
|
908 |
*/ |
7b1fe5979
|
909 |
int reuse_swap_page(struct page *page) |
1da177e4c
|
910 |
{ |
c475a8ab6
|
911 |
int count; |
309381fea
|
912 |
VM_BUG_ON_PAGE(!PageLocked(page), page); |
5ad646880
|
913 914 |
if (unlikely(PageKsm(page))) return 0; |
1f25fe20a
|
915 916 917 |
/* The page is part of THP and cannot be reused */ if (PageTransCompound(page)) return 0; |
c475a8ab6
|
918 |
count = page_mapcount(page); |
7b1fe5979
|
919 |
if (count <= 1 && PageSwapCache(page)) { |
c475a8ab6
|
920 |
count += page_swapcount(page); |
7b1fe5979
|
921 922 923 924 925 |
if (count == 1 && !PageWriteback(page)) { delete_from_swap_cache(page); SetPageDirty(page); } } |
5ad646880
|
926 |
return count <= 1; |
1da177e4c
|
927 928 929 |
} /* |
a2c43eed8
|
930 931 |
* If swap is getting full, or if there are no more mappings of this page, * then try_to_free_swap is called to free its swap space. |
1da177e4c
|
932 |
*/ |
a2c43eed8
|
933 |
int try_to_free_swap(struct page *page) |
1da177e4c
|
934 |
{ |
309381fea
|
935 |
VM_BUG_ON_PAGE(!PageLocked(page), page); |
1da177e4c
|
936 937 938 939 940 |
if (!PageSwapCache(page)) return 0; if (PageWriteback(page)) return 0; |
a2c43eed8
|
941 |
if (page_swapcount(page)) |
1da177e4c
|
942 |
return 0; |
b73d7fcec
|
943 944 945 946 947 948 949 950 951 952 953 954 |
/* * Once hibernation has begun to create its image of memory, * there's a danger that one of the calls to try_to_free_swap() * - most probably a call from __try_to_reclaim_swap() while * hibernation is allocating its own swap pages for the image, * but conceivably even a call from memory reclaim - will free * the swap from a page which has already been recorded in the * image as a clean swapcache page, and then reuse its swap for * another page of the image. On waking from hibernation, the * original page might be freed under memory pressure, then * later read back in from swap, now with the wrong data. * |
2de1a7e40
|
955 |
* Hibernation suspends storage while it is writing the image |
f90ac3982
|
956 |
* to disk so check that here. |
b73d7fcec
|
957 |
*/ |
f90ac3982
|
958 |
if (pm_suspended_storage()) |
b73d7fcec
|
959 |
return 0; |
a2c43eed8
|
960 961 962 |
delete_from_swap_cache(page); SetPageDirty(page); return 1; |
68a22394c
|
963 964 965 |
} /* |
1da177e4c
|
966 967 968 |
* Free the swap entry like above, but also try to * free the page cache entry if it is the last user. */ |
2509ef26d
|
969 |
int free_swap_and_cache(swp_entry_t entry) |
1da177e4c
|
970 |
{ |
2509ef26d
|
971 |
struct swap_info_struct *p; |
1da177e4c
|
972 |
struct page *page = NULL; |
a7420aa54
|
973 |
if (non_swap_entry(entry)) |
2509ef26d
|
974 |
return 1; |
0697212a4
|
975 |
|
1da177e4c
|
976 977 |
p = swap_info_get(entry); if (p) { |
253d553ba
|
978 |
if (swap_entry_free(p, entry, 1) == SWAP_HAS_CACHE) { |
33806f06d
|
979 980 |
page = find_get_page(swap_address_space(entry), entry.val); |
8413ac9d8
|
981 |
if (page && !trylock_page(page)) { |
09cbfeaf1
|
982 |
put_page(page); |
93fac7041
|
983 984 985 |
page = NULL; } } |
ec8acf20a
|
986 |
spin_unlock(&p->lock); |
1da177e4c
|
987 988 |
} if (page) { |
a2c43eed8
|
989 990 991 992 |
/* * Not mapped elsewhere, or swap space full? Free it! * Also recheck PageSwapCache now page is locked (above). */ |
93fac7041
|
993 |
if (PageSwapCache(page) && !PageWriteback(page) && |
5ccc5abaa
|
994 |
(!page_mapped(page) || mem_cgroup_swap_full(page))) { |
1da177e4c
|
995 996 997 998 |
delete_from_swap_cache(page); SetPageDirty(page); } unlock_page(page); |
09cbfeaf1
|
999 |
put_page(page); |
1da177e4c
|
1000 |
} |
2509ef26d
|
1001 |
return p != NULL; |
1da177e4c
|
1002 |
} |
b0cb1a19d
|
1003 |
#ifdef CONFIG_HIBERNATION |
f577eb30a
|
1004 |
/* |
915bae9eb
|
1005 |
* Find the swap type that corresponds to given device (if any). |
f577eb30a
|
1006 |
* |
915bae9eb
|
1007 1008 1009 1010 |
* @offset - number of the PAGE_SIZE-sized block of the device, starting * from 0, in which the swap header is expected to be located. * * This is needed for the suspend to disk (aka swsusp). |
f577eb30a
|
1011 |
*/ |
7bf236874
|
1012 |
int swap_type_of(dev_t device, sector_t offset, struct block_device **bdev_p) |
f577eb30a
|
1013 |
{ |
915bae9eb
|
1014 |
struct block_device *bdev = NULL; |
efa90a981
|
1015 |
int type; |
f577eb30a
|
1016 |
|
915bae9eb
|
1017 1018 |
if (device) bdev = bdget(device); |
f577eb30a
|
1019 |
spin_lock(&swap_lock); |
efa90a981
|
1020 1021 |
for (type = 0; type < nr_swapfiles; type++) { struct swap_info_struct *sis = swap_info[type]; |
f577eb30a
|
1022 |
|
915bae9eb
|
1023 |
if (!(sis->flags & SWP_WRITEOK)) |
f577eb30a
|
1024 |
continue; |
b6b5bce35
|
1025 |
|
915bae9eb
|
1026 |
if (!bdev) { |
7bf236874
|
1027 |
if (bdev_p) |
dddac6a7b
|
1028 |
*bdev_p = bdgrab(sis->bdev); |
7bf236874
|
1029 |
|
6e1819d61
|
1030 |
spin_unlock(&swap_lock); |
efa90a981
|
1031 |
return type; |
6e1819d61
|
1032 |
} |
915bae9eb
|
1033 |
if (bdev == sis->bdev) { |
9625a5f28
|
1034 |
struct swap_extent *se = &sis->first_swap_extent; |
915bae9eb
|
1035 |
|
915bae9eb
|
1036 |
if (se->start_block == offset) { |
7bf236874
|
1037 |
if (bdev_p) |
dddac6a7b
|
1038 |
*bdev_p = bdgrab(sis->bdev); |
7bf236874
|
1039 |
|
915bae9eb
|
1040 1041 |
spin_unlock(&swap_lock); bdput(bdev); |
efa90a981
|
1042 |
return type; |
915bae9eb
|
1043 |
} |
f577eb30a
|
1044 1045 1046 |
} } spin_unlock(&swap_lock); |
915bae9eb
|
1047 1048 |
if (bdev) bdput(bdev); |
f577eb30a
|
1049 1050 1051 1052 |
return -ENODEV; } /* |
73c34b6ac
|
1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 |
* Get the (PAGE_SIZE) block corresponding to given offset on the swapdev * corresponding to given index in swap_info (swap type). */ sector_t swapdev_block(int type, pgoff_t offset) { struct block_device *bdev; if ((unsigned int)type >= nr_swapfiles) return 0; if (!(swap_info[type]->flags & SWP_WRITEOK)) return 0; |
d4906e1aa
|
1064 |
return map_swap_entry(swp_entry(type, offset), &bdev); |
73c34b6ac
|
1065 1066 1067 |
} /* |
f577eb30a
|
1068 1069 1070 1071 1072 1073 1074 1075 |
* Return either the total number of swap pages of given type, or the number * of free pages of that type (depending on @free) * * This is needed for software suspend */ unsigned int count_swap_pages(int type, int free) { unsigned int n = 0; |
efa90a981
|
1076 1077 1078 |
spin_lock(&swap_lock); if ((unsigned int)type < nr_swapfiles) { struct swap_info_struct *sis = swap_info[type]; |
ec8acf20a
|
1079 |
spin_lock(&sis->lock); |
efa90a981
|
1080 1081 |
if (sis->flags & SWP_WRITEOK) { n = sis->pages; |
f577eb30a
|
1082 |
if (free) |
efa90a981
|
1083 |
n -= sis->inuse_pages; |
f577eb30a
|
1084 |
} |
ec8acf20a
|
1085 |
spin_unlock(&sis->lock); |
f577eb30a
|
1086 |
} |
efa90a981
|
1087 |
spin_unlock(&swap_lock); |
f577eb30a
|
1088 1089 |
return n; } |
73c34b6ac
|
1090 |
#endif /* CONFIG_HIBERNATION */ |
f577eb30a
|
1091 |
|
9f8bdb3f3
|
1092 |
static inline int pte_same_as_swp(pte_t pte, pte_t swp_pte) |
179ef71cb
|
1093 |
{ |
9f8bdb3f3
|
1094 |
return pte_same(pte_swp_clear_soft_dirty(pte), swp_pte); |
179ef71cb
|
1095 |
} |
1da177e4c
|
1096 |
/* |
72866f6f2
|
1097 1098 1099 |
* No need to decide whether this PTE shares the swap entry with others, * just let do_wp_page work it out if a write is requested later - to * force COW, vm_page_prot omits write permission from any private vma. |
1da177e4c
|
1100 |
*/ |
044d66c1d
|
1101 |
static int unuse_pte(struct vm_area_struct *vma, pmd_t *pmd, |
1da177e4c
|
1102 1103 |
unsigned long addr, swp_entry_t entry, struct page *page) { |
9e16b7fb1
|
1104 |
struct page *swapcache; |
72835c86c
|
1105 |
struct mem_cgroup *memcg; |
044d66c1d
|
1106 1107 1108 |
spinlock_t *ptl; pte_t *pte; int ret = 1; |
9e16b7fb1
|
1109 1110 1111 1112 |
swapcache = page; page = ksm_might_need_to_copy(page, vma, addr); if (unlikely(!page)) return -ENOMEM; |
f627c2f53
|
1113 1114 |
if (mem_cgroup_try_charge(page, vma->vm_mm, GFP_KERNEL, &memcg, false)) { |
044d66c1d
|
1115 |
ret = -ENOMEM; |
85d9fc89f
|
1116 1117 |
goto out_nolock; } |
044d66c1d
|
1118 1119 |
pte = pte_offset_map_lock(vma->vm_mm, pmd, addr, &ptl); |
9f8bdb3f3
|
1120 |
if (unlikely(!pte_same_as_swp(*pte, swp_entry_to_pte(entry)))) { |
f627c2f53
|
1121 |
mem_cgroup_cancel_charge(page, memcg, false); |
044d66c1d
|
1122 1123 1124 |
ret = 0; goto out; } |
8a9f3ccd2
|
1125 |
|
b084d4353
|
1126 |
dec_mm_counter(vma->vm_mm, MM_SWAPENTS); |
d559db086
|
1127 |
inc_mm_counter(vma->vm_mm, MM_ANONPAGES); |
1da177e4c
|
1128 1129 1130 |
get_page(page); set_pte_at(vma->vm_mm, addr, pte, pte_mkold(mk_pte(page, vma->vm_page_prot))); |
00501b531
|
1131 |
if (page == swapcache) { |
d281ee614
|
1132 |
page_add_anon_rmap(page, vma, addr, false); |
f627c2f53
|
1133 |
mem_cgroup_commit_charge(page, memcg, true, false); |
00501b531
|
1134 |
} else { /* ksm created a completely new copy */ |
d281ee614
|
1135 |
page_add_new_anon_rmap(page, vma, addr, false); |
f627c2f53
|
1136 |
mem_cgroup_commit_charge(page, memcg, false, false); |
00501b531
|
1137 1138 |
lru_cache_add_active_or_unevictable(page, vma); } |
1da177e4c
|
1139 1140 1141 1142 1143 1144 |
swap_free(entry); /* * Move the page to the active list so it is not * immediately swapped out again after swapon. */ activate_page(page); |
044d66c1d
|
1145 1146 |
out: pte_unmap_unlock(pte, ptl); |
85d9fc89f
|
1147 |
out_nolock: |
9e16b7fb1
|
1148 1149 1150 1151 |
if (page != swapcache) { unlock_page(page); put_page(page); } |
044d66c1d
|
1152 |
return ret; |
1da177e4c
|
1153 1154 1155 1156 1157 1158 |
} static int unuse_pte_range(struct vm_area_struct *vma, pmd_t *pmd, unsigned long addr, unsigned long end, swp_entry_t entry, struct page *page) { |
1da177e4c
|
1159 |
pte_t swp_pte = swp_entry_to_pte(entry); |
705e87c0c
|
1160 |
pte_t *pte; |
8a9f3ccd2
|
1161 |
int ret = 0; |
1da177e4c
|
1162 |
|
044d66c1d
|
1163 1164 1165 1166 1167 1168 1169 |
/* * We don't actually need pte lock while scanning for swp_pte: since * we hold page lock and mmap_sem, swp_pte cannot be inserted into the * page table while we're scanning; though it could get zapped, and on * some architectures (e.g. x86_32 with PAE) we might catch a glimpse * of unmatched parts which look like swp_pte, so unuse_pte must * recheck under pte lock. Scanning without pte lock lets it be |
2de1a7e40
|
1170 |
* preemptable whenever CONFIG_PREEMPT but not CONFIG_HIGHPTE. |
044d66c1d
|
1171 1172 |
*/ pte = pte_offset_map(pmd, addr); |
1da177e4c
|
1173 1174 1175 1176 1177 |
do { /* * swapoff spends a _lot_ of time in this loop! * Test inline before going to call unuse_pte. */ |
9f8bdb3f3
|
1178 |
if (unlikely(pte_same_as_swp(*pte, swp_pte))) { |
044d66c1d
|
1179 1180 1181 1182 1183 |
pte_unmap(pte); ret = unuse_pte(vma, pmd, addr, entry, page); if (ret) goto out; pte = pte_offset_map(pmd, addr); |
1da177e4c
|
1184 1185 |
} } while (pte++, addr += PAGE_SIZE, addr != end); |
044d66c1d
|
1186 1187 |
pte_unmap(pte - 1); out: |
8a9f3ccd2
|
1188 |
return ret; |
1da177e4c
|
1189 1190 1191 1192 1193 1194 1195 1196 |
} static inline int unuse_pmd_range(struct vm_area_struct *vma, pud_t *pud, unsigned long addr, unsigned long end, swp_entry_t entry, struct page *page) { pmd_t *pmd; unsigned long next; |
8a9f3ccd2
|
1197 |
int ret; |
1da177e4c
|
1198 1199 1200 1201 |
pmd = pmd_offset(pud, addr); do { next = pmd_addr_end(addr, end); |
1a5a9906d
|
1202 |
if (pmd_none_or_trans_huge_or_clear_bad(pmd)) |
1da177e4c
|
1203 |
continue; |
8a9f3ccd2
|
1204 1205 1206 |
ret = unuse_pte_range(vma, pmd, addr, next, entry, page); if (ret) return ret; |
1da177e4c
|
1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 |
} while (pmd++, addr = next, addr != end); return 0; } static inline int unuse_pud_range(struct vm_area_struct *vma, pgd_t *pgd, unsigned long addr, unsigned long end, swp_entry_t entry, struct page *page) { pud_t *pud; unsigned long next; |
8a9f3ccd2
|
1217 |
int ret; |
1da177e4c
|
1218 1219 1220 1221 1222 1223 |
pud = pud_offset(pgd, addr); do { next = pud_addr_end(addr, end); if (pud_none_or_clear_bad(pud)) continue; |
8a9f3ccd2
|
1224 1225 1226 |
ret = unuse_pmd_range(vma, pud, addr, next, entry, page); if (ret) return ret; |
1da177e4c
|
1227 1228 1229 1230 1231 1232 1233 1234 1235 |
} while (pud++, addr = next, addr != end); return 0; } static int unuse_vma(struct vm_area_struct *vma, swp_entry_t entry, struct page *page) { pgd_t *pgd; unsigned long addr, end, next; |
8a9f3ccd2
|
1236 |
int ret; |
1da177e4c
|
1237 |
|
3ca7b3c5b
|
1238 |
if (page_anon_vma(page)) { |
1da177e4c
|
1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 |
addr = page_address_in_vma(page, vma); if (addr == -EFAULT) return 0; else end = addr + PAGE_SIZE; } else { addr = vma->vm_start; end = vma->vm_end; } pgd = pgd_offset(vma->vm_mm, addr); do { next = pgd_addr_end(addr, end); if (pgd_none_or_clear_bad(pgd)) continue; |
8a9f3ccd2
|
1254 1255 1256 |
ret = unuse_pud_range(vma, pgd, addr, next, entry, page); if (ret) return ret; |
1da177e4c
|
1257 1258 1259 1260 1261 1262 1263 1264 |
} while (pgd++, addr = next, addr != end); return 0; } static int unuse_mm(struct mm_struct *mm, swp_entry_t entry, struct page *page) { struct vm_area_struct *vma; |
8a9f3ccd2
|
1265 |
int ret = 0; |
1da177e4c
|
1266 1267 1268 |
if (!down_read_trylock(&mm->mmap_sem)) { /* |
7d03431cf
|
1269 1270 |
* Activate page so shrink_inactive_list is unlikely to unmap * its ptes while lock is dropped, so swapoff can make progress. |
1da177e4c
|
1271 |
*/ |
c475a8ab6
|
1272 |
activate_page(page); |
1da177e4c
|
1273 1274 1275 1276 |
unlock_page(page); down_read(&mm->mmap_sem); lock_page(page); } |
1da177e4c
|
1277 |
for (vma = mm->mmap; vma; vma = vma->vm_next) { |
8a9f3ccd2
|
1278 |
if (vma->anon_vma && (ret = unuse_vma(vma, entry, page))) |
1da177e4c
|
1279 1280 |
break; } |
1da177e4c
|
1281 |
up_read(&mm->mmap_sem); |
8a9f3ccd2
|
1282 |
return (ret < 0)? ret: 0; |
1da177e4c
|
1283 1284 1285 |
} /* |
38b5faf4b
|
1286 1287 |
* Scan swap_map (or frontswap_map if frontswap parameter is true) * from current position to next entry still in use. |
1da177e4c
|
1288 1289 |
* Recycle to start on reaching the end, returning 0 when empty. */ |
6eb396dc4
|
1290 |
static unsigned int find_next_to_unuse(struct swap_info_struct *si, |
38b5faf4b
|
1291 |
unsigned int prev, bool frontswap) |
1da177e4c
|
1292 |
{ |
6eb396dc4
|
1293 1294 |
unsigned int max = si->max; unsigned int i = prev; |
8d69aaee8
|
1295 |
unsigned char count; |
1da177e4c
|
1296 1297 |
/* |
5d337b919
|
1298 |
* No need for swap_lock here: we're just looking |
1da177e4c
|
1299 1300 |
* for whether an entry is in use, not modifying it; false * hits are okay, and sys_swapoff() has already prevented new |
5d337b919
|
1301 |
* allocations from this area (while holding swap_lock). |
1da177e4c
|
1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 |
*/ for (;;) { if (++i >= max) { if (!prev) { i = 0; break; } /* * No entries in use at top of swap_map, * loop back to start and recheck there. */ max = prev + 1; prev = 0; i = 1; } |
38b5faf4b
|
1317 1318 1319 1320 1321 1322 |
if (frontswap) { if (frontswap_test(si, i)) break; else continue; } |
4db0c3c29
|
1323 |
count = READ_ONCE(si->swap_map[i]); |
355cfa73d
|
1324 |
if (count && swap_count(count) != SWAP_MAP_BAD) |
1da177e4c
|
1325 1326 1327 1328 1329 1330 1331 1332 1333 |
break; } return i; } /* * We completely avoid races by reading each swap page in advance, * and then search for the process using it. All the necessary * page table adjustments can then be made atomically. |
38b5faf4b
|
1334 1335 1336 |
* * if the boolean frontswap is true, only unuse pages_to_unuse pages; * pages_to_unuse==0 means all pages; ignored if frontswap is false |
1da177e4c
|
1337 |
*/ |
38b5faf4b
|
1338 1339 |
int try_to_unuse(unsigned int type, bool frontswap, unsigned long pages_to_unuse) |
1da177e4c
|
1340 |
{ |
efa90a981
|
1341 |
struct swap_info_struct *si = swap_info[type]; |
1da177e4c
|
1342 |
struct mm_struct *start_mm; |
edfe23dac
|
1343 1344 1345 1346 1347 |
volatile unsigned char *swap_map; /* swap_map is accessed without * locking. Mark it as volatile * to prevent compiler doing * something odd. */ |
8d69aaee8
|
1348 |
unsigned char swcount; |
1da177e4c
|
1349 1350 |
struct page *page; swp_entry_t entry; |
6eb396dc4
|
1351 |
unsigned int i = 0; |
1da177e4c
|
1352 |
int retval = 0; |
1da177e4c
|
1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 |
/* * When searching mms for an entry, a good strategy is to * start at the first mm we freed the previous entry from * (though actually we don't notice whether we or coincidence * freed the entry). Initialize this start_mm with a hold. * * A simpler strategy would be to start at the last mm we * freed the previous entry from; but that would take less * advantage of mmlist ordering, which clusters forked mms * together, child after parent. If we race with dup_mmap(), we * prefer to resolve parent before child, lest we miss entries * duplicated after we scanned child: using last mm would invert |
570a335b8
|
1366 |
* that. |
1da177e4c
|
1367 1368 1369 1370 1371 1372 1373 1374 1375 |
*/ start_mm = &init_mm; atomic_inc(&init_mm.mm_users); /* * Keep on scanning until all entries have gone. Usually, * one pass through swap_map is enough, but not necessarily: * there are races when an instance of an entry might be missed. */ |
38b5faf4b
|
1376 |
while ((i = find_next_to_unuse(si, i, frontswap)) != 0) { |
1da177e4c
|
1377 1378 1379 1380 |
if (signal_pending(current)) { retval = -EINTR; break; } |
886bb7e9c
|
1381 |
/* |
1da177e4c
|
1382 1383 |
* Get a page for the entry, using the existing swap * cache page if there is one. Otherwise, get a clean |
886bb7e9c
|
1384 |
* page and read the swap into it. |
1da177e4c
|
1385 1386 1387 |
*/ swap_map = &si->swap_map[i]; entry = swp_entry(type, i); |
02098feaa
|
1388 1389 |
page = read_swap_cache_async(entry, GFP_HIGHUSER_MOVABLE, NULL, 0); |
1da177e4c
|
1390 1391 1392 1393 1394 1395 1396 |
if (!page) { /* * Either swap_duplicate() failed because entry * has been freed independently, and will not be * reused since sys_swapoff() already disabled * allocation from here, or alloc_page() failed. */ |
edfe23dac
|
1397 1398 1399 1400 1401 1402 1403 1404 1405 |
swcount = *swap_map; /* * We don't hold lock here, so the swap entry could be * SWAP_MAP_BAD (when the cluster is discarding). * Instead of fail out, We can just skip the swap * entry because swapoff will wait for discarding * finish anyway. */ if (!swcount || swcount == SWAP_MAP_BAD) |
1da177e4c
|
1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 |
continue; retval = -ENOMEM; break; } /* * Don't hold on to start_mm if it looks like exiting. */ if (atomic_read(&start_mm->mm_users) == 1) { mmput(start_mm); start_mm = &init_mm; atomic_inc(&init_mm.mm_users); } /* * Wait for and lock page. When do_swap_page races with * try_to_unuse, do_swap_page can handle the fault much * faster than try_to_unuse can locate the entry. This * apparently redundant "wait_on_page_locked" lets try_to_unuse * defer to do_swap_page in such a case - in some tests, * do_swap_page and try_to_unuse repeatedly compete. */ wait_on_page_locked(page); wait_on_page_writeback(page); lock_page(page); wait_on_page_writeback(page); /* * Remove all references to entry. |
1da177e4c
|
1435 |
*/ |
1da177e4c
|
1436 |
swcount = *swap_map; |
aaa468653
|
1437 1438 1439 1440 1441 1442 |
if (swap_count(swcount) == SWAP_MAP_SHMEM) { retval = shmem_unuse(entry, page); /* page has already been unlocked and released */ if (retval < 0) break; continue; |
1da177e4c
|
1443 |
} |
aaa468653
|
1444 1445 |
if (swap_count(swcount) && start_mm != &init_mm) retval = unuse_mm(start_mm, entry, page); |
355cfa73d
|
1446 |
if (swap_count(*swap_map)) { |
1da177e4c
|
1447 1448 1449 1450 1451 1452 1453 1454 1455 |
int set_start_mm = (*swap_map >= swcount); struct list_head *p = &start_mm->mmlist; struct mm_struct *new_start_mm = start_mm; struct mm_struct *prev_mm = start_mm; struct mm_struct *mm; atomic_inc(&new_start_mm->mm_users); atomic_inc(&prev_mm->mm_users); spin_lock(&mmlist_lock); |
aaa468653
|
1456 |
while (swap_count(*swap_map) && !retval && |
1da177e4c
|
1457 1458 |
(p = p->next) != &start_mm->mmlist) { mm = list_entry(p, struct mm_struct, mmlist); |
70af7c5c6
|
1459 |
if (!atomic_inc_not_zero(&mm->mm_users)) |
1da177e4c
|
1460 |
continue; |
1da177e4c
|
1461 1462 1463 1464 1465 1466 1467 |
spin_unlock(&mmlist_lock); mmput(prev_mm); prev_mm = mm; cond_resched(); swcount = *swap_map; |
355cfa73d
|
1468 |
if (!swap_count(swcount)) /* any usage ? */ |
1da177e4c
|
1469 |
; |
aaa468653
|
1470 |
else if (mm == &init_mm) |
1da177e4c
|
1471 |
set_start_mm = 1; |
aaa468653
|
1472 |
else |
1da177e4c
|
1473 |
retval = unuse_mm(mm, entry, page); |
355cfa73d
|
1474 |
|
32c5fc10e
|
1475 |
if (set_start_mm && *swap_map < swcount) { |
1da177e4c
|
1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 |
mmput(new_start_mm); atomic_inc(&mm->mm_users); new_start_mm = mm; set_start_mm = 0; } spin_lock(&mmlist_lock); } spin_unlock(&mmlist_lock); mmput(prev_mm); mmput(start_mm); start_mm = new_start_mm; } if (retval) { unlock_page(page); |
09cbfeaf1
|
1490 |
put_page(page); |
1da177e4c
|
1491 1492 1493 1494 |
break; } /* |
1da177e4c
|
1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 |
* If a reference remains (rare), we would like to leave * the page in the swap cache; but try_to_unmap could * then re-duplicate the entry once we drop page lock, * so we might loop indefinitely; also, that page could * not be swapped out to other storage meanwhile. So: * delete from cache even if there's another reference, * after ensuring that the data has been saved to disk - * since if the reference remains (rarer), it will be * read from disk into another page. Splitting into two * pages would be incorrect if swap supported "shared * private" pages, but they are handled by tmpfs files. |
5ad646880
|
1506 1507 1508 1509 1510 1511 |
* * Given how unuse_vma() targets one particular offset * in an anon_vma, once the anon_vma has been determined, * this splitting happens to be just what is needed to * handle where KSM pages have been swapped out: re-reading * is unnecessarily slow, but we can fix that later on. |
1da177e4c
|
1512 |
*/ |
355cfa73d
|
1513 1514 |
if (swap_count(*swap_map) && PageDirty(page) && PageSwapCache(page)) { |
1da177e4c
|
1515 1516 1517 1518 1519 1520 1521 1522 |
struct writeback_control wbc = { .sync_mode = WB_SYNC_NONE, }; swap_writepage(page, &wbc); lock_page(page); wait_on_page_writeback(page); } |
68bdc8d64
|
1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 |
/* * It is conceivable that a racing task removed this page from * swap cache just before we acquired the page lock at the top, * or while we dropped it in unuse_mm(). The page might even * be back in swap cache on another swap area: that we must not * delete, since it may not have been written out to swap yet. */ if (PageSwapCache(page) && likely(page_private(page) == entry.val)) |
2e0e26c76
|
1533 |
delete_from_swap_cache(page); |
1da177e4c
|
1534 1535 1536 1537 |
/* * So we could skip searching mms once swap count went * to 1, we did not mark any present ptes as dirty: must |
2706a1b89
|
1538 |
* mark page dirty so shrink_page_list will preserve it. |
1da177e4c
|
1539 1540 1541 |
*/ SetPageDirty(page); unlock_page(page); |
09cbfeaf1
|
1542 |
put_page(page); |
1da177e4c
|
1543 1544 1545 1546 1547 1548 |
/* * Make sure that we aren't completely killing * interactive performance. */ cond_resched(); |
38b5faf4b
|
1549 1550 1551 1552 |
if (frontswap && pages_to_unuse > 0) { if (!--pages_to_unuse) break; } |
1da177e4c
|
1553 1554 1555 |
} mmput(start_mm); |
1da177e4c
|
1556 1557 1558 1559 |
return retval; } /* |
5d337b919
|
1560 1561 1562 |
* After a successful try_to_unuse, if no swap is now in use, we know * we can empty the mmlist. swap_lock must be held on entry and exit. * Note that mmlist_lock nests inside swap_lock, and an mm must be |
1da177e4c
|
1563 1564 1565 1566 1567 |
* added to the mmlist just after page_duplicate - before would be racy. */ static void drain_mmlist(void) { struct list_head *p, *next; |
efa90a981
|
1568 |
unsigned int type; |
1da177e4c
|
1569 |
|
efa90a981
|
1570 1571 |
for (type = 0; type < nr_swapfiles; type++) if (swap_info[type]->inuse_pages) |
1da177e4c
|
1572 1573 1574 1575 1576 1577 1578 1579 1580 |
return; spin_lock(&mmlist_lock); list_for_each_safe(p, next, &init_mm.mmlist) list_del_init(p); spin_unlock(&mmlist_lock); } /* * Use this swapdev's extent info to locate the (PAGE_SIZE) block which |
d4906e1aa
|
1581 1582 1583 |
* corresponds to page offset for the specified swap entry. * Note that the type of this function is sector_t, but it returns page offset * into the bdev, not sector offset. |
1da177e4c
|
1584 |
*/ |
d4906e1aa
|
1585 |
static sector_t map_swap_entry(swp_entry_t entry, struct block_device **bdev) |
1da177e4c
|
1586 |
{ |
f29ad6a99
|
1587 1588 1589 1590 |
struct swap_info_struct *sis; struct swap_extent *start_se; struct swap_extent *se; pgoff_t offset; |
efa90a981
|
1591 |
sis = swap_info[swp_type(entry)]; |
f29ad6a99
|
1592 1593 1594 1595 1596 |
*bdev = sis->bdev; offset = swp_offset(entry); start_se = sis->curr_swap_extent; se = start_se; |
1da177e4c
|
1597 1598 |
for ( ; ; ) { |
1da177e4c
|
1599 1600 1601 1602 |
if (se->start_page <= offset && offset < (se->start_page + se->nr_pages)) { return se->start_block + (offset - se->start_page); } |
a8ae49917
|
1603 |
se = list_next_entry(se, list); |
1da177e4c
|
1604 1605 1606 1607 1608 1609 |
sis->curr_swap_extent = se; BUG_ON(se == start_se); /* It *must* be present */ } } /* |
d4906e1aa
|
1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 |
* Returns the page offset into bdev for the specified page's swap entry. */ sector_t map_swap_page(struct page *page, struct block_device **bdev) { swp_entry_t entry; entry.val = page_private(page); return map_swap_entry(entry, bdev); } /* |
1da177e4c
|
1620 1621 1622 1623 |
* Free all of a swapdev's extent information */ static void destroy_swap_extents(struct swap_info_struct *sis) { |
9625a5f28
|
1624 |
while (!list_empty(&sis->first_swap_extent.list)) { |
1da177e4c
|
1625 |
struct swap_extent *se; |
a8ae49917
|
1626 |
se = list_first_entry(&sis->first_swap_extent.list, |
1da177e4c
|
1627 1628 1629 1630 |
struct swap_extent, list); list_del(&se->list); kfree(se); } |
62c230bc1
|
1631 1632 1633 1634 1635 1636 1637 1638 |
if (sis->flags & SWP_FILE) { struct file *swap_file = sis->swap_file; struct address_space *mapping = swap_file->f_mapping; sis->flags &= ~SWP_FILE; mapping->a_ops->swap_deactivate(swap_file); } |
1da177e4c
|
1639 1640 1641 1642 |
} /* * Add a block range (and the corresponding page range) into this swapdev's |
11d31886d
|
1643 |
* extent list. The extent list is kept sorted in page order. |
1da177e4c
|
1644 |
* |
11d31886d
|
1645 |
* This function rather assumes that it is called in ascending page order. |
1da177e4c
|
1646 |
*/ |
a509bc1a9
|
1647 |
int |
1da177e4c
|
1648 1649 1650 1651 1652 1653 |
add_swap_extent(struct swap_info_struct *sis, unsigned long start_page, unsigned long nr_pages, sector_t start_block) { struct swap_extent *se; struct swap_extent *new_se; struct list_head *lh; |
9625a5f28
|
1654 1655 1656 1657 1658 1659 1660 1661 1662 |
if (start_page == 0) { se = &sis->first_swap_extent; sis->curr_swap_extent = se; se->start_page = 0; se->nr_pages = nr_pages; se->start_block = start_block; return 1; } else { lh = sis->first_swap_extent.list.prev; /* Highest extent */ |
1da177e4c
|
1663 |
se = list_entry(lh, struct swap_extent, list); |
11d31886d
|
1664 1665 |
BUG_ON(se->start_page + se->nr_pages != start_page); if (se->start_block + se->nr_pages == start_block) { |
1da177e4c
|
1666 1667 1668 1669 |
/* Merge it */ se->nr_pages += nr_pages; return 0; } |
1da177e4c
|
1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 |
} /* * No merge. Insert a new extent, preserving ordering. */ new_se = kmalloc(sizeof(*se), GFP_KERNEL); if (new_se == NULL) return -ENOMEM; new_se->start_page = start_page; new_se->nr_pages = nr_pages; new_se->start_block = start_block; |
9625a5f28
|
1681 |
list_add_tail(&new_se->list, &sis->first_swap_extent.list); |
53092a740
|
1682 |
return 1; |
1da177e4c
|
1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 |
} /* * A `swap extent' is a simple thing which maps a contiguous range of pages * onto a contiguous range of disk blocks. An ordered list of swap extents * is built at swapon time and is then used at swap_writepage/swap_readpage * time for locating where on disk a page belongs. * * If the swapfile is an S_ISBLK block device, a single extent is installed. * This is done so that the main operating code can treat S_ISBLK and S_ISREG * swap files identically. * * Whether the swapdev is an S_ISREG file or an S_ISBLK blockdev, the swap * extent list operates in PAGE_SIZE disk blocks. Both S_ISREG and S_ISBLK * swapfiles are handled *identically* after swapon time. * * For S_ISREG swapfiles, setup_swap_extents() will walk all the file's blocks * and will parse them into an ordered extent list, in PAGE_SIZE chunks. If * some stray blocks are found which do not fall within the PAGE_SIZE alignment * requirements, they are simply tossed out - we will never use those blocks * for swapping. * |
b0d9bcd4b
|
1705 |
* For S_ISREG swapfiles we set S_SWAPFILE across the life of the swapon. This |
1da177e4c
|
1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 |
* prevents root from shooting her foot off by ftruncating an in-use swapfile, * which will scribble on the fs. * * The amount of disk space which a single swap extent represents varies. * Typically it is in the 1-4 megabyte range. So we can have hundreds of * extents in the list. To avoid much list walking, we cache the previous * search location in `curr_swap_extent', and start new searches from there. * This is extremely effective. The average number of iterations in * map_swap_page() has been measured at about 0.3 per page. - akpm. */ |
53092a740
|
1716 |
static int setup_swap_extents(struct swap_info_struct *sis, sector_t *span) |
1da177e4c
|
1717 |
{ |
62c230bc1
|
1718 1719 1720 |
struct file *swap_file = sis->swap_file; struct address_space *mapping = swap_file->f_mapping; struct inode *inode = mapping->host; |
1da177e4c
|
1721 |
int ret; |
1da177e4c
|
1722 1723 |
if (S_ISBLK(inode->i_mode)) { ret = add_swap_extent(sis, 0, sis->max, 0); |
53092a740
|
1724 |
*span = sis->pages; |
a509bc1a9
|
1725 |
return ret; |
1da177e4c
|
1726 |
} |
62c230bc1
|
1727 |
if (mapping->a_ops->swap_activate) { |
a509bc1a9
|
1728 |
ret = mapping->a_ops->swap_activate(sis, swap_file, span); |
62c230bc1
|
1729 1730 1731 1732 1733 |
if (!ret) { sis->flags |= SWP_FILE; ret = add_swap_extent(sis, 0, sis->max, 0); *span = sis->pages; } |
a509bc1a9
|
1734 |
return ret; |
62c230bc1
|
1735 |
} |
a509bc1a9
|
1736 |
return generic_swapfile_activate(sis, swap_file, span); |
1da177e4c
|
1737 |
} |
cf0cac0a0
|
1738 |
static void _enable_swap_info(struct swap_info_struct *p, int prio, |
2a8f94493
|
1739 1740 |
unsigned char *swap_map, struct swap_cluster_info *cluster_info) |
40531542e
|
1741 |
{ |
40531542e
|
1742 1743 1744 1745 |
if (prio >= 0) p->prio = prio; else p->prio = --least_priority; |
18ab4d4ce
|
1746 1747 1748 1749 1750 1751 |
/* * the plist prio is negated because plist ordering is * low-to-high, while swap ordering is high-to-low */ p->list.prio = -p->prio; p->avail_list.prio = -p->prio; |
40531542e
|
1752 |
p->swap_map = swap_map; |
2a8f94493
|
1753 |
p->cluster_info = cluster_info; |
40531542e
|
1754 |
p->flags |= SWP_WRITEOK; |
ec8acf20a
|
1755 |
atomic_long_add(p->pages, &nr_swap_pages); |
40531542e
|
1756 |
total_swap_pages += p->pages; |
adfab836f
|
1757 |
assert_spin_locked(&swap_lock); |
adfab836f
|
1758 |
/* |
18ab4d4ce
|
1759 1760 1761 1762 1763 1764 1765 1766 |
* both lists are plists, and thus priority ordered. * swap_active_head needs to be priority ordered for swapoff(), * which on removal of any swap_info_struct with an auto-assigned * (i.e. negative) priority increments the auto-assigned priority * of any lower-priority swap_info_structs. * swap_avail_head needs to be priority ordered for get_swap_page(), * which allocates swap pages from the highest available priority * swap_info_struct. |
adfab836f
|
1767 |
*/ |
18ab4d4ce
|
1768 1769 1770 1771 |
plist_add(&p->list, &swap_active_head); spin_lock(&swap_avail_lock); plist_add(&p->avail_list, &swap_avail_head); spin_unlock(&swap_avail_lock); |
cf0cac0a0
|
1772 1773 1774 1775 |
} static void enable_swap_info(struct swap_info_struct *p, int prio, unsigned char *swap_map, |
2a8f94493
|
1776 |
struct swap_cluster_info *cluster_info, |
cf0cac0a0
|
1777 1778 |
unsigned long *frontswap_map) { |
4f89849da
|
1779 |
frontswap_init(p->type, frontswap_map); |
cf0cac0a0
|
1780 |
spin_lock(&swap_lock); |
ec8acf20a
|
1781 |
spin_lock(&p->lock); |
2a8f94493
|
1782 |
_enable_swap_info(p, prio, swap_map, cluster_info); |
ec8acf20a
|
1783 |
spin_unlock(&p->lock); |
cf0cac0a0
|
1784 1785 1786 1787 1788 1789 |
spin_unlock(&swap_lock); } static void reinsert_swap_info(struct swap_info_struct *p) { spin_lock(&swap_lock); |
ec8acf20a
|
1790 |
spin_lock(&p->lock); |
2a8f94493
|
1791 |
_enable_swap_info(p, p->prio, p->swap_map, p->cluster_info); |
ec8acf20a
|
1792 |
spin_unlock(&p->lock); |
40531542e
|
1793 1794 |
spin_unlock(&swap_lock); } |
c4ea37c26
|
1795 |
SYSCALL_DEFINE1(swapoff, const char __user *, specialfile) |
1da177e4c
|
1796 |
{ |
73c34b6ac
|
1797 |
struct swap_info_struct *p = NULL; |
8d69aaee8
|
1798 |
unsigned char *swap_map; |
2a8f94493
|
1799 |
struct swap_cluster_info *cluster_info; |
4f89849da
|
1800 |
unsigned long *frontswap_map; |
1da177e4c
|
1801 1802 1803 |
struct file *swap_file, *victim; struct address_space *mapping; struct inode *inode; |
91a27b2a7
|
1804 |
struct filename *pathname; |
adfab836f
|
1805 |
int err, found = 0; |
5b808a230
|
1806 |
unsigned int old_block_size; |
886bb7e9c
|
1807 |
|
1da177e4c
|
1808 1809 |
if (!capable(CAP_SYS_ADMIN)) return -EPERM; |
191c54244
|
1810 |
BUG_ON(!current->mm); |
1da177e4c
|
1811 |
pathname = getname(specialfile); |
1da177e4c
|
1812 |
if (IS_ERR(pathname)) |
f58b59c1d
|
1813 |
return PTR_ERR(pathname); |
1da177e4c
|
1814 |
|
669abf4e5
|
1815 |
victim = file_open_name(pathname, O_RDWR|O_LARGEFILE, 0); |
1da177e4c
|
1816 1817 1818 1819 1820 |
err = PTR_ERR(victim); if (IS_ERR(victim)) goto out; mapping = victim->f_mapping; |
5d337b919
|
1821 |
spin_lock(&swap_lock); |
18ab4d4ce
|
1822 |
plist_for_each_entry(p, &swap_active_head, list) { |
22c6f8fdb
|
1823 |
if (p->flags & SWP_WRITEOK) { |
adfab836f
|
1824 1825 |
if (p->swap_file->f_mapping == mapping) { found = 1; |
1da177e4c
|
1826 |
break; |
adfab836f
|
1827 |
} |
1da177e4c
|
1828 |
} |
1da177e4c
|
1829 |
} |
adfab836f
|
1830 |
if (!found) { |
1da177e4c
|
1831 |
err = -EINVAL; |
5d337b919
|
1832 |
spin_unlock(&swap_lock); |
1da177e4c
|
1833 1834 |
goto out_dput; } |
191c54244
|
1835 |
if (!security_vm_enough_memory_mm(current->mm, p->pages)) |
1da177e4c
|
1836 1837 1838 |
vm_unacct_memory(p->pages); else { err = -ENOMEM; |
5d337b919
|
1839 |
spin_unlock(&swap_lock); |
1da177e4c
|
1840 1841 |
goto out_dput; } |
18ab4d4ce
|
1842 1843 1844 |
spin_lock(&swap_avail_lock); plist_del(&p->avail_list, &swap_avail_head); spin_unlock(&swap_avail_lock); |
ec8acf20a
|
1845 |
spin_lock(&p->lock); |
78ecba081
|
1846 |
if (p->prio < 0) { |
adfab836f
|
1847 |
struct swap_info_struct *si = p; |
18ab4d4ce
|
1848 |
plist_for_each_entry_continue(si, &swap_active_head, list) { |
adfab836f
|
1849 |
si->prio++; |
18ab4d4ce
|
1850 1851 |
si->list.prio--; si->avail_list.prio--; |
adfab836f
|
1852 |
} |
78ecba081
|
1853 1854 |
least_priority++; } |
18ab4d4ce
|
1855 |
plist_del(&p->list, &swap_active_head); |
ec8acf20a
|
1856 |
atomic_long_sub(p->pages, &nr_swap_pages); |
1da177e4c
|
1857 1858 |
total_swap_pages -= p->pages; p->flags &= ~SWP_WRITEOK; |
ec8acf20a
|
1859 |
spin_unlock(&p->lock); |
5d337b919
|
1860 |
spin_unlock(&swap_lock); |
fb4f88dca
|
1861 |
|
e1e12d2f3
|
1862 |
set_current_oom_origin(); |
adfab836f
|
1863 |
err = try_to_unuse(p->type, false, 0); /* force unuse all pages */ |
e1e12d2f3
|
1864 |
clear_current_oom_origin(); |
1da177e4c
|
1865 |
|
1da177e4c
|
1866 1867 |
if (err) { /* re-insert swap space back into swap_list */ |
cf0cac0a0
|
1868 |
reinsert_swap_info(p); |
1da177e4c
|
1869 1870 |
goto out_dput; } |
52b7efdbe
|
1871 |
|
815c2c543
|
1872 |
flush_work(&p->discard_work); |
5d337b919
|
1873 |
destroy_swap_extents(p); |
570a335b8
|
1874 1875 |
if (p->flags & SWP_CONTINUED) free_swap_count_continuations(p); |
fc0abb145
|
1876 |
mutex_lock(&swapon_mutex); |
5d337b919
|
1877 |
spin_lock(&swap_lock); |
ec8acf20a
|
1878 |
spin_lock(&p->lock); |
5d337b919
|
1879 |
drain_mmlist(); |
52b7efdbe
|
1880 |
/* wait for anyone still in scan_swap_map */ |
52b7efdbe
|
1881 1882 |
p->highest_bit = 0; /* cuts scans short */ while (p->flags >= SWP_SCANNING) { |
ec8acf20a
|
1883 |
spin_unlock(&p->lock); |
5d337b919
|
1884 |
spin_unlock(&swap_lock); |
13e4b57f6
|
1885 |
schedule_timeout_uninterruptible(1); |
5d337b919
|
1886 |
spin_lock(&swap_lock); |
ec8acf20a
|
1887 |
spin_lock(&p->lock); |
52b7efdbe
|
1888 |
} |
52b7efdbe
|
1889 |
|
1da177e4c
|
1890 |
swap_file = p->swap_file; |
5b808a230
|
1891 |
old_block_size = p->old_block_size; |
1da177e4c
|
1892 1893 1894 1895 |
p->swap_file = NULL; p->max = 0; swap_map = p->swap_map; p->swap_map = NULL; |
2a8f94493
|
1896 1897 |
cluster_info = p->cluster_info; p->cluster_info = NULL; |
4f89849da
|
1898 |
frontswap_map = frontswap_map_get(p); |
ec8acf20a
|
1899 |
spin_unlock(&p->lock); |
5d337b919
|
1900 |
spin_unlock(&swap_lock); |
adfab836f
|
1901 |
frontswap_invalidate_area(p->type); |
58e97ba6b
|
1902 |
frontswap_map_set(p, NULL); |
fc0abb145
|
1903 |
mutex_unlock(&swapon_mutex); |
ebc2a1a69
|
1904 1905 |
free_percpu(p->percpu_cluster); p->percpu_cluster = NULL; |
1da177e4c
|
1906 |
vfree(swap_map); |
2a8f94493
|
1907 |
vfree(cluster_info); |
4f89849da
|
1908 |
vfree(frontswap_map); |
2de1a7e40
|
1909 |
/* Destroy swap account information */ |
adfab836f
|
1910 |
swap_cgroup_swapoff(p->type); |
27a7faa07
|
1911 |
|
1da177e4c
|
1912 1913 1914 |
inode = mapping->host; if (S_ISBLK(inode->i_mode)) { struct block_device *bdev = I_BDEV(inode); |
5b808a230
|
1915 |
set_blocksize(bdev, old_block_size); |
e525fd89d
|
1916 |
blkdev_put(bdev, FMODE_READ | FMODE_WRITE | FMODE_EXCL); |
1da177e4c
|
1917 |
} else { |
5955102c9
|
1918 |
inode_lock(inode); |
1da177e4c
|
1919 |
inode->i_flags &= ~S_SWAPFILE; |
5955102c9
|
1920 |
inode_unlock(inode); |
1da177e4c
|
1921 1922 |
} filp_close(swap_file, NULL); |
f893ab41e
|
1923 1924 1925 1926 1927 1928 1929 1930 1931 |
/* * Clear the SWP_USED flag after all resources are freed so that swapon * can reuse this swap_info in alloc_swap_info() safely. It is ok to * not hold p->lock after we cleared its SWP_WRITEOK. */ spin_lock(&swap_lock); p->flags = 0; spin_unlock(&swap_lock); |
1da177e4c
|
1932 |
err = 0; |
66d7dd518
|
1933 1934 |
atomic_inc(&proc_poll_event); wake_up_interruptible(&proc_poll_wait); |
1da177e4c
|
1935 1936 1937 1938 |
out_dput: filp_close(victim, NULL); out: |
f58b59c1d
|
1939 |
putname(pathname); |
1da177e4c
|
1940 1941 1942 1943 |
return err; } #ifdef CONFIG_PROC_FS |
66d7dd518
|
1944 1945 |
static unsigned swaps_poll(struct file *file, poll_table *wait) { |
f15146380
|
1946 |
struct seq_file *seq = file->private_data; |
66d7dd518
|
1947 1948 |
poll_wait(file, &proc_poll_wait, wait); |
f15146380
|
1949 1950 |
if (seq->poll_event != atomic_read(&proc_poll_event)) { seq->poll_event = atomic_read(&proc_poll_event); |
66d7dd518
|
1951 1952 1953 1954 1955 |
return POLLIN | POLLRDNORM | POLLERR | POLLPRI; } return POLLIN | POLLRDNORM; } |
1da177e4c
|
1956 1957 1958 |
/* iterator */ static void *swap_start(struct seq_file *swap, loff_t *pos) { |
efa90a981
|
1959 1960 |
struct swap_info_struct *si; int type; |
1da177e4c
|
1961 |
loff_t l = *pos; |
fc0abb145
|
1962 |
mutex_lock(&swapon_mutex); |
1da177e4c
|
1963 |
|
881e4aabe
|
1964 1965 |
if (!l) return SEQ_START_TOKEN; |
efa90a981
|
1966 1967 1968 1969 |
for (type = 0; type < nr_swapfiles; type++) { smp_rmb(); /* read nr_swapfiles before swap_info[type] */ si = swap_info[type]; if (!(si->flags & SWP_USED) || !si->swap_map) |
1da177e4c
|
1970 |
continue; |
881e4aabe
|
1971 |
if (!--l) |
efa90a981
|
1972 |
return si; |
1da177e4c
|
1973 1974 1975 1976 1977 1978 1979 |
} return NULL; } static void *swap_next(struct seq_file *swap, void *v, loff_t *pos) { |
efa90a981
|
1980 1981 |
struct swap_info_struct *si = v; int type; |
1da177e4c
|
1982 |
|
881e4aabe
|
1983 |
if (v == SEQ_START_TOKEN) |
efa90a981
|
1984 1985 1986 |
type = 0; else type = si->type + 1; |
881e4aabe
|
1987 |
|
efa90a981
|
1988 1989 1990 1991 |
for (; type < nr_swapfiles; type++) { smp_rmb(); /* read nr_swapfiles before swap_info[type] */ si = swap_info[type]; if (!(si->flags & SWP_USED) || !si->swap_map) |
1da177e4c
|
1992 1993 |
continue; ++*pos; |
efa90a981
|
1994 |
return si; |
1da177e4c
|
1995 1996 1997 1998 1999 2000 2001 |
} return NULL; } static void swap_stop(struct seq_file *swap, void *v) { |
fc0abb145
|
2002 |
mutex_unlock(&swapon_mutex); |
1da177e4c
|
2003 2004 2005 2006 |
} static int swap_show(struct seq_file *swap, void *v) { |
efa90a981
|
2007 |
struct swap_info_struct *si = v; |
1da177e4c
|
2008 2009 |
struct file *file; int len; |
efa90a981
|
2010 |
if (si == SEQ_START_TOKEN) { |
881e4aabe
|
2011 2012 2013 2014 |
seq_puts(swap,"Filename\t\t\t\tType\t\tSize\tUsed\tPriority "); return 0; } |
1da177e4c
|
2015 |
|
efa90a981
|
2016 |
file = si->swap_file; |
2726d5662
|
2017 2018 |
len = seq_file_path(swap, file, " \t \\"); |
6eb396dc4
|
2019 2020 |
seq_printf(swap, "%*s%s\t%u\t%u\t%d ", |
886bb7e9c
|
2021 |
len < 40 ? 40 - len : 1, " ", |
496ad9aa8
|
2022 |
S_ISBLK(file_inode(file)->i_mode) ? |
1da177e4c
|
2023 |
"partition" : "file\t", |
efa90a981
|
2024 2025 2026 |
si->pages << (PAGE_SHIFT - 10), si->inuse_pages << (PAGE_SHIFT - 10), si->prio); |
1da177e4c
|
2027 2028 |
return 0; } |
15ad7cdcf
|
2029 |
static const struct seq_operations swaps_op = { |
1da177e4c
|
2030 2031 2032 2033 2034 2035 2036 2037 |
.start = swap_start, .next = swap_next, .stop = swap_stop, .show = swap_show }; static int swaps_open(struct inode *inode, struct file *file) { |
f15146380
|
2038 |
struct seq_file *seq; |
66d7dd518
|
2039 |
int ret; |
66d7dd518
|
2040 |
ret = seq_open(file, &swaps_op); |
f15146380
|
2041 |
if (ret) |
66d7dd518
|
2042 |
return ret; |
66d7dd518
|
2043 |
|
f15146380
|
2044 2045 2046 |
seq = file->private_data; seq->poll_event = atomic_read(&proc_poll_event); return 0; |
1da177e4c
|
2047 |
} |
15ad7cdcf
|
2048 |
static const struct file_operations proc_swaps_operations = { |
1da177e4c
|
2049 2050 2051 2052 |
.open = swaps_open, .read = seq_read, .llseek = seq_lseek, .release = seq_release, |
66d7dd518
|
2053 |
.poll = swaps_poll, |
1da177e4c
|
2054 2055 2056 2057 |
}; static int __init procswaps_init(void) { |
3d71f86f4
|
2058 |
proc_create("swaps", 0, NULL, &proc_swaps_operations); |
1da177e4c
|
2059 2060 2061 2062 |
return 0; } __initcall(procswaps_init); #endif /* CONFIG_PROC_FS */ |
1796316a8
|
2063 2064 2065 2066 2067 2068 2069 2070 |
#ifdef MAX_SWAPFILES_CHECK static int __init max_swapfiles_check(void) { MAX_SWAPFILES_CHECK(); return 0; } late_initcall(max_swapfiles_check); #endif |
53cbb2435
|
2071 |
static struct swap_info_struct *alloc_swap_info(void) |
1da177e4c
|
2072 |
{ |
73c34b6ac
|
2073 |
struct swap_info_struct *p; |
1da177e4c
|
2074 |
unsigned int type; |
efa90a981
|
2075 2076 2077 |
p = kzalloc(sizeof(*p), GFP_KERNEL); if (!p) |
53cbb2435
|
2078 |
return ERR_PTR(-ENOMEM); |
efa90a981
|
2079 |
|
5d337b919
|
2080 |
spin_lock(&swap_lock); |
efa90a981
|
2081 2082 |
for (type = 0; type < nr_swapfiles; type++) { if (!(swap_info[type]->flags & SWP_USED)) |
1da177e4c
|
2083 |
break; |
efa90a981
|
2084 |
} |
0697212a4
|
2085 |
if (type >= MAX_SWAPFILES) { |
5d337b919
|
2086 |
spin_unlock(&swap_lock); |
efa90a981
|
2087 |
kfree(p); |
730c0581c
|
2088 |
return ERR_PTR(-EPERM); |
1da177e4c
|
2089 |
} |
efa90a981
|
2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 |
if (type >= nr_swapfiles) { p->type = type; swap_info[type] = p; /* * Write swap_info[type] before nr_swapfiles, in case a * racing procfs swap_start() or swap_next() is reading them. * (We never shrink nr_swapfiles, we never free this entry.) */ smp_wmb(); nr_swapfiles++; } else { kfree(p); p = swap_info[type]; /* * Do not memset this entry: a racing procfs swap_next() * would be relying on p->type to remain valid. */ } |
9625a5f28
|
2108 |
INIT_LIST_HEAD(&p->first_swap_extent.list); |
18ab4d4ce
|
2109 2110 |
plist_node_init(&p->list, 0); plist_node_init(&p->avail_list, 0); |
1da177e4c
|
2111 |
p->flags = SWP_USED; |
5d337b919
|
2112 |
spin_unlock(&swap_lock); |
ec8acf20a
|
2113 |
spin_lock_init(&p->lock); |
efa90a981
|
2114 |
|
53cbb2435
|
2115 |
return p; |
53cbb2435
|
2116 |
} |
4d0e1e107
|
2117 2118 2119 2120 2121 2122 2123 |
static int claim_swapfile(struct swap_info_struct *p, struct inode *inode) { int error; if (S_ISBLK(inode->i_mode)) { p->bdev = bdgrab(I_BDEV(inode)); error = blkdev_get(p->bdev, |
6f179af88
|
2124 |
FMODE_READ | FMODE_WRITE | FMODE_EXCL, p); |
4d0e1e107
|
2125 2126 |
if (error < 0) { p->bdev = NULL; |
6f179af88
|
2127 |
return error; |
4d0e1e107
|
2128 2129 2130 2131 |
} p->old_block_size = block_size(p->bdev); error = set_blocksize(p->bdev, PAGE_SIZE); if (error < 0) |
87ade72a7
|
2132 |
return error; |
4d0e1e107
|
2133 2134 2135 |
p->flags |= SWP_BLKDEV; } else if (S_ISREG(inode->i_mode)) { p->bdev = inode->i_sb->s_bdev; |
5955102c9
|
2136 |
inode_lock(inode); |
87ade72a7
|
2137 2138 2139 2140 |
if (IS_SWAPFILE(inode)) return -EBUSY; } else return -EINVAL; |
4d0e1e107
|
2141 2142 |
return 0; |
4d0e1e107
|
2143 |
} |
ca8bd38bf
|
2144 2145 2146 2147 2148 2149 2150 |
static unsigned long read_swap_header(struct swap_info_struct *p, union swap_header *swap_header, struct inode *inode) { int i; unsigned long maxpages; unsigned long swapfilepages; |
d6bbbd29b
|
2151 |
unsigned long last_page; |
ca8bd38bf
|
2152 2153 |
if (memcmp("SWAPSPACE2", swap_header->magic.magic, 10)) { |
465c47fd8
|
2154 2155 |
pr_err("Unable to find swap-space signature "); |
387190253
|
2156 |
return 0; |
ca8bd38bf
|
2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 |
} /* swap partition endianess hack... */ if (swab32(swap_header->info.version) == 1) { swab32s(&swap_header->info.version); swab32s(&swap_header->info.last_page); swab32s(&swap_header->info.nr_badpages); for (i = 0; i < swap_header->info.nr_badpages; i++) swab32s(&swap_header->info.badpages[i]); } /* Check the swap header's sub-version */ if (swap_header->info.version != 1) { |
465c47fd8
|
2169 2170 2171 |
pr_warn("Unable to handle swap header version %d ", swap_header->info.version); |
387190253
|
2172 |
return 0; |
ca8bd38bf
|
2173 2174 2175 2176 2177 2178 2179 2180 |
} p->lowest_bit = 1; p->cluster_next = 1; p->cluster_nr = 0; /* * Find out how many pages are allowed for a single swap |
9b15b817f
|
2181 |
* device. There are two limiting factors: 1) the number |
a2c16d6cb
|
2182 2183 |
* of bits for the swap offset in the swp_entry_t type, and * 2) the number of bits in the swap pte as defined by the |
9b15b817f
|
2184 |
* different architectures. In order to find the |
a2c16d6cb
|
2185 |
* largest possible bit mask, a swap entry with swap type 0 |
ca8bd38bf
|
2186 |
* and swap offset ~0UL is created, encoded to a swap pte, |
a2c16d6cb
|
2187 |
* decoded to a swp_entry_t again, and finally the swap |
ca8bd38bf
|
2188 2189 2190 |
* offset is extracted. This will mask all the bits from * the initial ~0UL mask that can't be encoded in either * the swp_entry_t or the architecture definition of a |
9b15b817f
|
2191 |
* swap pte. |
ca8bd38bf
|
2192 2193 |
*/ maxpages = swp_offset(pte_to_swp_entry( |
9b15b817f
|
2194 |
swp_entry_to_pte(swp_entry(0, ~0UL)))) + 1; |
d6bbbd29b
|
2195 2196 |
last_page = swap_header->info.last_page; if (last_page > maxpages) { |
465c47fd8
|
2197 2198 |
pr_warn("Truncating oversized swap area, only using %luk out of %luk ", |
d6bbbd29b
|
2199 2200 2201 2202 2203 |
maxpages << (PAGE_SHIFT - 10), last_page << (PAGE_SHIFT - 10)); } if (maxpages > last_page) { maxpages = last_page + 1; |
ca8bd38bf
|
2204 2205 2206 2207 2208 2209 2210 |
/* p->max is an unsigned int: don't overflow it */ if ((unsigned int)maxpages == 0) maxpages = UINT_MAX; } p->highest_bit = maxpages - 1; if (!maxpages) |
387190253
|
2211 |
return 0; |
ca8bd38bf
|
2212 2213 |
swapfilepages = i_size_read(inode) >> PAGE_SHIFT; if (swapfilepages && maxpages > swapfilepages) { |
465c47fd8
|
2214 2215 |
pr_warn("Swap area shorter than signature indicates "); |
387190253
|
2216 |
return 0; |
ca8bd38bf
|
2217 2218 |
} if (swap_header->info.nr_badpages && S_ISREG(inode->i_mode)) |
387190253
|
2219 |
return 0; |
ca8bd38bf
|
2220 |
if (swap_header->info.nr_badpages > MAX_SWAP_BADPAGES) |
387190253
|
2221 |
return 0; |
ca8bd38bf
|
2222 2223 |
return maxpages; |
ca8bd38bf
|
2224 |
} |
915d4d7bc
|
2225 2226 2227 |
static int setup_swap_map_and_extents(struct swap_info_struct *p, union swap_header *swap_header, unsigned char *swap_map, |
2a8f94493
|
2228 |
struct swap_cluster_info *cluster_info, |
915d4d7bc
|
2229 2230 2231 2232 |
unsigned long maxpages, sector_t *span) { int i; |
915d4d7bc
|
2233 2234 |
unsigned int nr_good_pages; int nr_extents; |
2a8f94493
|
2235 2236 |
unsigned long nr_clusters = DIV_ROUND_UP(maxpages, SWAPFILE_CLUSTER); unsigned long idx = p->cluster_next / SWAPFILE_CLUSTER; |
915d4d7bc
|
2237 2238 |
nr_good_pages = maxpages - 1; /* omit header page */ |
2a8f94493
|
2239 2240 |
cluster_set_null(&p->free_cluster_head); cluster_set_null(&p->free_cluster_tail); |
815c2c543
|
2241 2242 |
cluster_set_null(&p->discard_cluster_head); cluster_set_null(&p->discard_cluster_tail); |
2a8f94493
|
2243 |
|
915d4d7bc
|
2244 2245 |
for (i = 0; i < swap_header->info.nr_badpages; i++) { unsigned int page_nr = swap_header->info.badpages[i]; |
bdb8e3f68
|
2246 2247 |
if (page_nr == 0 || page_nr > swap_header->info.last_page) return -EINVAL; |
915d4d7bc
|
2248 2249 2250 |
if (page_nr < maxpages) { swap_map[page_nr] = SWAP_MAP_BAD; nr_good_pages--; |
2a8f94493
|
2251 2252 2253 2254 2255 |
/* * Haven't marked the cluster free yet, no list * operation involved */ inc_cluster_info_page(p, cluster_info, page_nr); |
915d4d7bc
|
2256 2257 |
} } |
2a8f94493
|
2258 2259 2260 |
/* Haven't marked the cluster free yet, no list operation involved */ for (i = maxpages; i < round_up(maxpages, SWAPFILE_CLUSTER); i++) inc_cluster_info_page(p, cluster_info, i); |
915d4d7bc
|
2261 2262 |
if (nr_good_pages) { swap_map[0] = SWAP_MAP_BAD; |
2a8f94493
|
2263 2264 2265 2266 2267 |
/* * Not mark the cluster free yet, no list * operation involved */ inc_cluster_info_page(p, cluster_info, 0); |
915d4d7bc
|
2268 2269 2270 |
p->max = maxpages; p->pages = nr_good_pages; nr_extents = setup_swap_extents(p, span); |
bdb8e3f68
|
2271 2272 |
if (nr_extents < 0) return nr_extents; |
915d4d7bc
|
2273 2274 2275 |
nr_good_pages = p->pages; } if (!nr_good_pages) { |
465c47fd8
|
2276 2277 |
pr_warn("Empty swap-file "); |
bdb8e3f68
|
2278 |
return -EINVAL; |
915d4d7bc
|
2279 |
} |
2a8f94493
|
2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 |
if (!cluster_info) return nr_extents; for (i = 0; i < nr_clusters; i++) { if (!cluster_count(&cluster_info[idx])) { cluster_set_flag(&cluster_info[idx], CLUSTER_FLAG_FREE); if (cluster_is_null(&p->free_cluster_head)) { cluster_set_next_flag(&p->free_cluster_head, idx, 0); cluster_set_next_flag(&p->free_cluster_tail, idx, 0); } else { unsigned int tail; tail = cluster_next(&p->free_cluster_tail); cluster_set_next(&cluster_info[tail], idx); cluster_set_next_flag(&p->free_cluster_tail, idx, 0); } } idx++; if (idx == nr_clusters) idx = 0; } |
915d4d7bc
|
2304 |
return nr_extents; |
915d4d7bc
|
2305 |
} |
dcf6b7ddd
|
2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 |
/* * Helper to sys_swapon determining if a given swap * backing device queue supports DISCARD operations. */ static bool swap_discardable(struct swap_info_struct *si) { struct request_queue *q = bdev_get_queue(si->bdev); if (!q || !blk_queue_discard(q)) return false; return true; } |
53cbb2435
|
2319 2320 2321 |
SYSCALL_DEFINE2(swapon, const char __user *, specialfile, int, swap_flags) { struct swap_info_struct *p; |
91a27b2a7
|
2322 |
struct filename *name; |
53cbb2435
|
2323 2324 |
struct file *swap_file = NULL; struct address_space *mapping; |
40531542e
|
2325 |
int prio; |
53cbb2435
|
2326 2327 |
int error; union swap_header *swap_header; |
915d4d7bc
|
2328 |
int nr_extents; |
53cbb2435
|
2329 2330 |
sector_t span; unsigned long maxpages; |
53cbb2435
|
2331 |
unsigned char *swap_map = NULL; |
2a8f94493
|
2332 |
struct swap_cluster_info *cluster_info = NULL; |
38b5faf4b
|
2333 |
unsigned long *frontswap_map = NULL; |
53cbb2435
|
2334 2335 |
struct page *page = NULL; struct inode *inode = NULL; |
53cbb2435
|
2336 |
|
d15cab975
|
2337 2338 |
if (swap_flags & ~SWAP_FLAGS_VALID) return -EINVAL; |
53cbb2435
|
2339 2340 2341 2342 |
if (!capable(CAP_SYS_ADMIN)) return -EPERM; p = alloc_swap_info(); |
2542e5134
|
2343 2344 |
if (IS_ERR(p)) return PTR_ERR(p); |
53cbb2435
|
2345 |
|
815c2c543
|
2346 |
INIT_WORK(&p->discard_work, swap_discard_work); |
1da177e4c
|
2347 |
name = getname(specialfile); |
1da177e4c
|
2348 |
if (IS_ERR(name)) { |
7de7fb6b3
|
2349 |
error = PTR_ERR(name); |
1da177e4c
|
2350 |
name = NULL; |
bd69010b0
|
2351 |
goto bad_swap; |
1da177e4c
|
2352 |
} |
669abf4e5
|
2353 |
swap_file = file_open_name(name, O_RDWR|O_LARGEFILE, 0); |
1da177e4c
|
2354 |
if (IS_ERR(swap_file)) { |
7de7fb6b3
|
2355 |
error = PTR_ERR(swap_file); |
1da177e4c
|
2356 |
swap_file = NULL; |
bd69010b0
|
2357 |
goto bad_swap; |
1da177e4c
|
2358 2359 2360 2361 |
} p->swap_file = swap_file; mapping = swap_file->f_mapping; |
2130781e2
|
2362 |
inode = mapping->host; |
6f179af88
|
2363 |
|
5955102c9
|
2364 |
/* If S_ISREG(inode->i_mode) will do inode_lock(inode); */ |
4d0e1e107
|
2365 2366 |
error = claim_swapfile(p, inode); if (unlikely(error)) |
1da177e4c
|
2367 |
goto bad_swap; |
1da177e4c
|
2368 |
|
1da177e4c
|
2369 2370 2371 2372 2373 2374 2375 |
/* * Read the swap header. */ if (!mapping->a_ops->readpage) { error = -EINVAL; goto bad_swap; } |
090d2b185
|
2376 |
page = read_mapping_page(mapping, 0, swap_file); |
1da177e4c
|
2377 2378 2379 2380 |
if (IS_ERR(page)) { error = PTR_ERR(page); goto bad_swap; } |
81e339712
|
2381 |
swap_header = kmap(page); |
1da177e4c
|
2382 |
|
ca8bd38bf
|
2383 2384 |
maxpages = read_swap_header(p, swap_header, inode); if (unlikely(!maxpages)) { |
1da177e4c
|
2385 2386 2387 |
error = -EINVAL; goto bad_swap; } |
886bb7e9c
|
2388 |
|
81e339712
|
2389 |
/* OK, set up the swap map and apply the bad block list */ |
803d0c835
|
2390 |
swap_map = vzalloc(maxpages); |
81e339712
|
2391 2392 2393 2394 |
if (!swap_map) { error = -ENOMEM; goto bad_swap; } |
2a8f94493
|
2395 |
if (p->bdev && blk_queue_nonrot(bdev_get_queue(p->bdev))) { |
6f179af88
|
2396 |
int cpu; |
2a8f94493
|
2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 |
p->flags |= SWP_SOLIDSTATE; /* * select a random position to start with to help wear leveling * SSD */ p->cluster_next = 1 + (prandom_u32() % p->highest_bit); cluster_info = vzalloc(DIV_ROUND_UP(maxpages, SWAPFILE_CLUSTER) * sizeof(*cluster_info)); if (!cluster_info) { error = -ENOMEM; goto bad_swap; } |
ebc2a1a69
|
2410 2411 2412 2413 2414 |
p->percpu_cluster = alloc_percpu(struct percpu_cluster); if (!p->percpu_cluster) { error = -ENOMEM; goto bad_swap; } |
6f179af88
|
2415 |
for_each_possible_cpu(cpu) { |
ebc2a1a69
|
2416 |
struct percpu_cluster *cluster; |
6f179af88
|
2417 |
cluster = per_cpu_ptr(p->percpu_cluster, cpu); |
ebc2a1a69
|
2418 2419 |
cluster_set_null(&cluster->index); } |
2a8f94493
|
2420 |
} |
1da177e4c
|
2421 |
|
1421ef3cd
|
2422 2423 2424 |
error = swap_cgroup_swapon(p->type, maxpages); if (error) goto bad_swap; |
915d4d7bc
|
2425 |
nr_extents = setup_swap_map_and_extents(p, swap_header, swap_map, |
2a8f94493
|
2426 |
cluster_info, maxpages, &span); |
915d4d7bc
|
2427 2428 |
if (unlikely(nr_extents < 0)) { error = nr_extents; |
1da177e4c
|
2429 2430 |
goto bad_swap; } |
38b5faf4b
|
2431 2432 |
/* frontswap enabled? set up bit-per-page map for frontswap */ if (frontswap_enabled) |
7b57976da
|
2433 |
frontswap_map = vzalloc(BITS_TO_LONGS(maxpages) * sizeof(long)); |
1da177e4c
|
2434 |
|
2a8f94493
|
2435 2436 2437 2438 2439 2440 2441 2442 2443 |
if (p->bdev &&(swap_flags & SWAP_FLAG_DISCARD) && swap_discardable(p)) { /* * When discard is enabled for swap with no particular * policy flagged, we set all swap discard flags here in * order to sustain backward compatibility with older * swapon(8) releases. */ p->flags |= (SWP_DISCARDABLE | SWP_AREA_DISCARD | SWP_PAGE_DISCARD); |
dcf6b7ddd
|
2444 |
|
2a8f94493
|
2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 |
/* * By flagging sys_swapon, a sysadmin can tell us to * either do single-time area discards only, or to just * perform discards for released swap page-clusters. * Now it's time to adjust the p->flags accordingly. */ if (swap_flags & SWAP_FLAG_DISCARD_ONCE) p->flags &= ~SWP_PAGE_DISCARD; else if (swap_flags & SWAP_FLAG_DISCARD_PAGES) p->flags &= ~SWP_AREA_DISCARD; /* issue a swapon-time discard if it's still required */ if (p->flags & SWP_AREA_DISCARD) { int err = discard_swap(p); if (unlikely(err)) pr_err("swapon: discard_swap(%p): %d ", p, err); |
dcf6b7ddd
|
2463 |
} |
20137a490
|
2464 |
} |
6a6ba8317
|
2465 |
|
fc0abb145
|
2466 |
mutex_lock(&swapon_mutex); |
40531542e
|
2467 |
prio = -1; |
78ecba081
|
2468 |
if (swap_flags & SWAP_FLAG_PREFER) |
40531542e
|
2469 |
prio = |
78ecba081
|
2470 |
(swap_flags & SWAP_FLAG_PRIO_MASK) >> SWAP_FLAG_PRIO_SHIFT; |
2a8f94493
|
2471 |
enable_swap_info(p, prio, swap_map, cluster_info, frontswap_map); |
c69dbfb84
|
2472 |
|
756a025f0
|
2473 2474 |
pr_info("Adding %uk swap on %s. Priority:%d extents:%d across:%lluk %s%s%s%s%s ", |
91a27b2a7
|
2475 |
p->pages<<(PAGE_SHIFT-10), name->name, p->prio, |
c69dbfb84
|
2476 2477 |
nr_extents, (unsigned long long)span<<(PAGE_SHIFT-10), (p->flags & SWP_SOLIDSTATE) ? "SS" : "", |
38b5faf4b
|
2478 |
(p->flags & SWP_DISCARDABLE) ? "D" : "", |
dcf6b7ddd
|
2479 2480 |
(p->flags & SWP_AREA_DISCARD) ? "s" : "", (p->flags & SWP_PAGE_DISCARD) ? "c" : "", |
38b5faf4b
|
2481 |
(frontswap_map) ? "FS" : ""); |
c69dbfb84
|
2482 |
|
fc0abb145
|
2483 |
mutex_unlock(&swapon_mutex); |
66d7dd518
|
2484 2485 |
atomic_inc(&proc_poll_event); wake_up_interruptible(&proc_poll_wait); |
9b01c350a
|
2486 2487 |
if (S_ISREG(inode->i_mode)) inode->i_flags |= S_SWAPFILE; |
1da177e4c
|
2488 2489 2490 |
error = 0; goto out; bad_swap: |
ebc2a1a69
|
2491 2492 |
free_percpu(p->percpu_cluster); p->percpu_cluster = NULL; |
bd69010b0
|
2493 |
if (inode && S_ISBLK(inode->i_mode) && p->bdev) { |
f2090d2df
|
2494 2495 |
set_blocksize(p->bdev, p->old_block_size); blkdev_put(p->bdev, FMODE_READ | FMODE_WRITE | FMODE_EXCL); |
1da177e4c
|
2496 |
} |
4cd3bb10f
|
2497 |
destroy_swap_extents(p); |
e8e6c2ec4
|
2498 |
swap_cgroup_swapoff(p->type); |
5d337b919
|
2499 |
spin_lock(&swap_lock); |
1da177e4c
|
2500 |
p->swap_file = NULL; |
1da177e4c
|
2501 |
p->flags = 0; |
5d337b919
|
2502 |
spin_unlock(&swap_lock); |
1da177e4c
|
2503 |
vfree(swap_map); |
2a8f94493
|
2504 |
vfree(cluster_info); |
52c50567d
|
2505 |
if (swap_file) { |
2130781e2
|
2506 |
if (inode && S_ISREG(inode->i_mode)) { |
5955102c9
|
2507 |
inode_unlock(inode); |
2130781e2
|
2508 2509 |
inode = NULL; } |
1da177e4c
|
2510 |
filp_close(swap_file, NULL); |
52c50567d
|
2511 |
} |
1da177e4c
|
2512 2513 2514 |
out: if (page && !IS_ERR(page)) { kunmap(page); |
09cbfeaf1
|
2515 |
put_page(page); |
1da177e4c
|
2516 2517 2518 |
} if (name) putname(name); |
9b01c350a
|
2519 |
if (inode && S_ISREG(inode->i_mode)) |
5955102c9
|
2520 |
inode_unlock(inode); |
1da177e4c
|
2521 2522 2523 2524 2525 |
return error; } void si_swapinfo(struct sysinfo *val) { |
efa90a981
|
2526 |
unsigned int type; |
1da177e4c
|
2527 |
unsigned long nr_to_be_unused = 0; |
5d337b919
|
2528 |
spin_lock(&swap_lock); |
efa90a981
|
2529 2530 2531 2532 2533 |
for (type = 0; type < nr_swapfiles; type++) { struct swap_info_struct *si = swap_info[type]; if ((si->flags & SWP_USED) && !(si->flags & SWP_WRITEOK)) nr_to_be_unused += si->inuse_pages; |
1da177e4c
|
2534 |
} |
ec8acf20a
|
2535 |
val->freeswap = atomic_long_read(&nr_swap_pages) + nr_to_be_unused; |
1da177e4c
|
2536 |
val->totalswap = total_swap_pages + nr_to_be_unused; |
5d337b919
|
2537 |
spin_unlock(&swap_lock); |
1da177e4c
|
2538 2539 2540 2541 2542 |
} /* * Verify that a swap entry is valid and increment its swap map count. * |
355cfa73d
|
2543 2544 2545 2546 2547 2548 |
* Returns error code in following case. * - success -> 0 * - swp_entry is invalid -> EINVAL * - swp_entry is migration entry -> EINVAL * - swap-cache reference is requested but there is already one. -> EEXIST * - swap-cache reference is requested but the entry is not used. -> ENOENT |
570a335b8
|
2549 |
* - swap-mapped reference requested but needs continued swap count. -> ENOMEM |
1da177e4c
|
2550 |
*/ |
8d69aaee8
|
2551 |
static int __swap_duplicate(swp_entry_t entry, unsigned char usage) |
1da177e4c
|
2552 |
{ |
73c34b6ac
|
2553 |
struct swap_info_struct *p; |
1da177e4c
|
2554 |
unsigned long offset, type; |
8d69aaee8
|
2555 2556 |
unsigned char count; unsigned char has_cache; |
253d553ba
|
2557 |
int err = -EINVAL; |
1da177e4c
|
2558 |
|
a7420aa54
|
2559 |
if (non_swap_entry(entry)) |
253d553ba
|
2560 |
goto out; |
0697212a4
|
2561 |
|
1da177e4c
|
2562 2563 2564 |
type = swp_type(entry); if (type >= nr_swapfiles) goto bad_file; |
efa90a981
|
2565 |
p = swap_info[type]; |
1da177e4c
|
2566 |
offset = swp_offset(entry); |
ec8acf20a
|
2567 |
spin_lock(&p->lock); |
355cfa73d
|
2568 2569 |
if (unlikely(offset >= p->max)) goto unlock_out; |
253d553ba
|
2570 |
count = p->swap_map[offset]; |
edfe23dac
|
2571 2572 2573 2574 2575 2576 2577 2578 2579 |
/* * swapin_readahead() doesn't check if a swap entry is valid, so the * swap entry could be SWAP_MAP_BAD. Check here with lock held. */ if (unlikely(swap_count(count) == SWAP_MAP_BAD)) { err = -ENOENT; goto unlock_out; } |
253d553ba
|
2580 2581 2582 |
has_cache = count & SWAP_HAS_CACHE; count &= ~SWAP_HAS_CACHE; err = 0; |
355cfa73d
|
2583 |
|
253d553ba
|
2584 |
if (usage == SWAP_HAS_CACHE) { |
355cfa73d
|
2585 2586 |
/* set SWAP_HAS_CACHE if there is no cache and entry is used */ |
253d553ba
|
2587 2588 2589 2590 2591 2592 |
if (!has_cache && count) has_cache = SWAP_HAS_CACHE; else if (has_cache) /* someone else added cache */ err = -EEXIST; else /* no users remaining */ err = -ENOENT; |
355cfa73d
|
2593 2594 |
} else if (count || has_cache) { |
253d553ba
|
2595 |
|
570a335b8
|
2596 2597 2598 |
if ((count & ~COUNT_CONTINUED) < SWAP_MAP_MAX) count += usage; else if ((count & ~COUNT_CONTINUED) > SWAP_MAP_MAX) |
253d553ba
|
2599 |
err = -EINVAL; |
570a335b8
|
2600 2601 2602 2603 |
else if (swap_count_continued(p, offset, count)) count = COUNT_CONTINUED; else err = -ENOMEM; |
355cfa73d
|
2604 |
} else |
253d553ba
|
2605 2606 2607 |
err = -ENOENT; /* unused swap entry */ p->swap_map[offset] = count | has_cache; |
355cfa73d
|
2608 |
unlock_out: |
ec8acf20a
|
2609 |
spin_unlock(&p->lock); |
1da177e4c
|
2610 |
out: |
253d553ba
|
2611 |
return err; |
1da177e4c
|
2612 2613 |
bad_file: |
465c47fd8
|
2614 2615 |
pr_err("swap_dup: %s%08lx ", Bad_file, entry.val); |
1da177e4c
|
2616 2617 |
goto out; } |
253d553ba
|
2618 |
|
355cfa73d
|
2619 |
/* |
aaa468653
|
2620 2621 2622 2623 2624 2625 2626 2627 2628 |
* Help swapoff by noting that swap entry belongs to shmem/tmpfs * (in which case its reference count is never incremented). */ void swap_shmem_alloc(swp_entry_t entry) { __swap_duplicate(entry, SWAP_MAP_SHMEM); } /* |
08259d58e
|
2629 2630 2631 2632 2633 |
* Increase reference count of swap entry by 1. * Returns 0 for success, or -ENOMEM if a swap_count_continuation is required * but could not be atomically allocated. Returns 0, just as if it succeeded, * if __swap_duplicate() fails for another reason (-EINVAL or -ENOENT), which * might occur if a page table entry has got corrupted. |
355cfa73d
|
2634 |
*/ |
570a335b8
|
2635 |
int swap_duplicate(swp_entry_t entry) |
355cfa73d
|
2636 |
{ |
570a335b8
|
2637 2638 2639 2640 2641 |
int err = 0; while (!err && __swap_duplicate(entry, 1) == -ENOMEM) err = add_swap_count_continuation(entry, GFP_ATOMIC); return err; |
355cfa73d
|
2642 |
} |
1da177e4c
|
2643 |
|
cb4b86ba4
|
2644 |
/* |
355cfa73d
|
2645 2646 |
* @entry: swap entry for which we allocate swap cache. * |
73c34b6ac
|
2647 |
* Called when allocating swap cache for existing swap entry, |
355cfa73d
|
2648 2649 2650 |
* This can return error codes. Returns 0 at success. * -EBUSY means there is a swap cache. * Note: return code is different from swap_duplicate(). |
cb4b86ba4
|
2651 2652 2653 |
*/ int swapcache_prepare(swp_entry_t entry) { |
253d553ba
|
2654 |
return __swap_duplicate(entry, SWAP_HAS_CACHE); |
cb4b86ba4
|
2655 |
} |
f981c5950
|
2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666 2667 |
struct swap_info_struct *page_swap_info(struct page *page) { swp_entry_t swap = { .val = page_private(page) }; BUG_ON(!PageSwapCache(page)); return swap_info[swp_type(swap)]; } /* * out-of-line __page_file_ methods to avoid include hell. */ struct address_space *__page_file_mapping(struct page *page) { |
309381fea
|
2668 |
VM_BUG_ON_PAGE(!PageSwapCache(page), page); |
f981c5950
|
2669 2670 2671 2672 2673 2674 2675 |
return page_swap_info(page)->swap_file->f_mapping; } EXPORT_SYMBOL_GPL(__page_file_mapping); pgoff_t __page_file_index(struct page *page) { swp_entry_t swap = { .val = page_private(page) }; |
309381fea
|
2676 |
VM_BUG_ON_PAGE(!PageSwapCache(page), page); |
f981c5950
|
2677 2678 2679 |
return swp_offset(swap); } EXPORT_SYMBOL_GPL(__page_file_index); |
1da177e4c
|
2680 |
/* |
570a335b8
|
2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 2699 2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 |
* add_swap_count_continuation - called when a swap count is duplicated * beyond SWAP_MAP_MAX, it allocates a new page and links that to the entry's * page of the original vmalloc'ed swap_map, to hold the continuation count * (for that entry and for its neighbouring PAGE_SIZE swap entries). Called * again when count is duplicated beyond SWAP_MAP_MAX * SWAP_CONT_MAX, etc. * * These continuation pages are seldom referenced: the common paths all work * on the original swap_map, only referring to a continuation page when the * low "digit" of a count is incremented or decremented through SWAP_MAP_MAX. * * add_swap_count_continuation(, GFP_ATOMIC) can be called while holding * page table locks; if it fails, add_swap_count_continuation(, GFP_KERNEL) * can be called after dropping locks. */ int add_swap_count_continuation(swp_entry_t entry, gfp_t gfp_mask) { struct swap_info_struct *si; struct page *head; struct page *page; struct page *list_page; pgoff_t offset; unsigned char count; /* * When debugging, it's easier to use __GFP_ZERO here; but it's better * for latency not to zero a page while GFP_ATOMIC and holding locks. */ page = alloc_page(gfp_mask | __GFP_HIGHMEM); si = swap_info_get(entry); if (!si) { /* * An acceptable race has occurred since the failing * __swap_duplicate(): the swap entry has been freed, * perhaps even the whole swap_map cleared for swapoff. */ goto outer; } offset = swp_offset(entry); count = si->swap_map[offset] & ~SWAP_HAS_CACHE; if ((count & ~COUNT_CONTINUED) != SWAP_MAP_MAX) { /* * The higher the swap count, the more likely it is that tasks * will race to add swap count continuation: we need to avoid * over-provisioning. */ goto out; } if (!page) { |
ec8acf20a
|
2733 |
spin_unlock(&si->lock); |
570a335b8
|
2734 2735 2736 2737 2738 |
return -ENOMEM; } /* * We are fortunate that although vmalloc_to_page uses pte_offset_map, |
2de1a7e40
|
2739 2740 |
* no architecture is using highmem pages for kernel page tables: so it * will not corrupt the GFP_ATOMIC caller's atomic page table kmaps. |
570a335b8
|
2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751 2752 2753 2754 2755 2756 2757 2758 2759 2760 2761 2762 2763 2764 |
*/ head = vmalloc_to_page(si->swap_map + offset); offset &= ~PAGE_MASK; /* * Page allocation does not initialize the page's lru field, * but it does always reset its private field. */ if (!page_private(head)) { BUG_ON(count & COUNT_CONTINUED); INIT_LIST_HEAD(&head->lru); set_page_private(head, SWP_CONTINUED); si->flags |= SWP_CONTINUED; } list_for_each_entry(list_page, &head->lru, lru) { unsigned char *map; /* * If the previous map said no continuation, but we've found * a continuation page, free our allocation and use this one. */ if (!(count & COUNT_CONTINUED)) goto out; |
9b04c5fec
|
2765 |
map = kmap_atomic(list_page) + offset; |
570a335b8
|
2766 |
count = *map; |
9b04c5fec
|
2767 |
kunmap_atomic(map); |
570a335b8
|
2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779 |
/* * If this continuation count now has some space in it, * free our allocation and use this one. */ if ((count & ~COUNT_CONTINUED) != SWAP_CONT_MAX) goto out; } list_add_tail(&page->lru, &head->lru); page = NULL; /* now it's attached, don't free it */ out: |
ec8acf20a
|
2780 |
spin_unlock(&si->lock); |
570a335b8
|
2781 2782 2783 2784 2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 2795 2796 2797 2798 2799 2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 |
outer: if (page) __free_page(page); return 0; } /* * swap_count_continued - when the original swap_map count is incremented * from SWAP_MAP_MAX, check if there is already a continuation page to carry * into, carry if so, or else fail until a new continuation page is allocated; * when the original swap_map count is decremented from 0 with continuation, * borrow from the continuation and report whether it still holds more. * Called while __swap_duplicate() or swap_entry_free() holds swap_lock. */ static bool swap_count_continued(struct swap_info_struct *si, pgoff_t offset, unsigned char count) { struct page *head; struct page *page; unsigned char *map; head = vmalloc_to_page(si->swap_map + offset); if (page_private(head) != SWP_CONTINUED) { BUG_ON(count & COUNT_CONTINUED); return false; /* need to add count continuation */ } offset &= ~PAGE_MASK; page = list_entry(head->lru.next, struct page, lru); |
9b04c5fec
|
2810 |
map = kmap_atomic(page) + offset; |
570a335b8
|
2811 2812 2813 2814 2815 2816 2817 2818 2819 |
if (count == SWAP_MAP_MAX) /* initial increment from swap_map */ goto init_map; /* jump over SWAP_CONT_MAX checks */ if (count == (SWAP_MAP_MAX | COUNT_CONTINUED)) { /* incrementing */ /* * Think of how you add 1 to 999 */ while (*map == (SWAP_CONT_MAX | COUNT_CONTINUED)) { |
9b04c5fec
|
2820 |
kunmap_atomic(map); |
570a335b8
|
2821 2822 |
page = list_entry(page->lru.next, struct page, lru); BUG_ON(page == head); |
9b04c5fec
|
2823 |
map = kmap_atomic(page) + offset; |
570a335b8
|
2824 2825 |
} if (*map == SWAP_CONT_MAX) { |
9b04c5fec
|
2826 |
kunmap_atomic(map); |
570a335b8
|
2827 2828 2829 |
page = list_entry(page->lru.next, struct page, lru); if (page == head) return false; /* add count continuation */ |
9b04c5fec
|
2830 |
map = kmap_atomic(page) + offset; |
570a335b8
|
2831 2832 2833 |
init_map: *map = 0; /* we didn't zero the page */ } *map += 1; |
9b04c5fec
|
2834 |
kunmap_atomic(map); |
570a335b8
|
2835 2836 |
page = list_entry(page->lru.prev, struct page, lru); while (page != head) { |
9b04c5fec
|
2837 |
map = kmap_atomic(page) + offset; |
570a335b8
|
2838 |
*map = COUNT_CONTINUED; |
9b04c5fec
|
2839 |
kunmap_atomic(map); |
570a335b8
|
2840 2841 2842 2843 2844 2845 2846 2847 2848 2849 |
page = list_entry(page->lru.prev, struct page, lru); } return true; /* incremented */ } else { /* decrementing */ /* * Think of how you subtract 1 from 1000 */ BUG_ON(count != COUNT_CONTINUED); while (*map == COUNT_CONTINUED) { |
9b04c5fec
|
2850 |
kunmap_atomic(map); |
570a335b8
|
2851 2852 |
page = list_entry(page->lru.next, struct page, lru); BUG_ON(page == head); |
9b04c5fec
|
2853 |
map = kmap_atomic(page) + offset; |
570a335b8
|
2854 2855 2856 2857 2858 |
} BUG_ON(*map == 0); *map -= 1; if (*map == 0) count = 0; |
9b04c5fec
|
2859 |
kunmap_atomic(map); |
570a335b8
|
2860 2861 |
page = list_entry(page->lru.prev, struct page, lru); while (page != head) { |
9b04c5fec
|
2862 |
map = kmap_atomic(page) + offset; |
570a335b8
|
2863 2864 |
*map = SWAP_CONT_MAX | count; count = COUNT_CONTINUED; |
9b04c5fec
|
2865 |
kunmap_atomic(map); |
570a335b8
|
2866 2867 2868 2869 2870 2871 2872 2873 2874 2875 2876 2877 2878 2879 2880 2881 2882 2883 |
page = list_entry(page->lru.prev, struct page, lru); } return count == COUNT_CONTINUED; } } /* * free_swap_count_continuations - swapoff free all the continuation pages * appended to the swap_map, after swap_map is quiesced, before vfree'ing it. */ static void free_swap_count_continuations(struct swap_info_struct *si) { pgoff_t offset; for (offset = 0; offset < si->max; offset += PAGE_SIZE) { struct page *head; head = vmalloc_to_page(si->swap_map + offset); if (page_private(head)) { |
0d576d20c
|
2884 2885 2886 2887 |
struct page *page, *next; list_for_each_entry_safe(page, next, &head->lru, lru) { list_del(&page->lru); |
570a335b8
|
2888 2889 2890 2891 2892 |
__free_page(page); } } } } |