Blame view
mm/swapfile.c
65.5 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 16 17 |
#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> #include <linux/shm.h> #include <linux/blkdev.h> |
20137a490
|
18 |
#include <linux/random.h> |
1da177e4c
|
19 20 21 22 23 |
#include <linux/writeback.h> #include <linux/proc_fs.h> #include <linux/seq_file.h> #include <linux/init.h> #include <linux/module.h> |
5ad646880
|
24 |
#include <linux/ksm.h> |
1da177e4c
|
25 26 27 |
#include <linux/rmap.h> #include <linux/security.h> #include <linux/backing-dev.h> |
fc0abb145
|
28 |
#include <linux/mutex.h> |
c59ede7b7
|
29 |
#include <linux/capability.h> |
1da177e4c
|
30 |
#include <linux/syscalls.h> |
8a9f3ccd2
|
31 |
#include <linux/memcontrol.h> |
66d7dd518
|
32 |
#include <linux/poll.h> |
1da177e4c
|
33 34 35 36 |
#include <asm/pgtable.h> #include <asm/tlbflush.h> #include <linux/swapops.h> |
27a7faa07
|
37 |
#include <linux/page_cgroup.h> |
1da177e4c
|
38 |
|
570a335b8
|
39 40 41 |
static bool swap_count_continued(struct swap_info_struct *, pgoff_t, unsigned char); static void free_swap_count_continuations(struct swap_info_struct *); |
d4906e1aa
|
42 |
static sector_t map_swap_entry(swp_entry_t, struct block_device**); |
570a335b8
|
43 |
|
7c363b8c6
|
44 45 |
static DEFINE_SPINLOCK(swap_lock); static unsigned int nr_swapfiles; |
b962716b4
|
46 |
long nr_swap_pages; |
1da177e4c
|
47 |
long total_swap_pages; |
78ecba081
|
48 |
static int least_priority; |
1da177e4c
|
49 |
|
1da177e4c
|
50 51 52 53 |
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 "; |
7c363b8c6
|
54 |
static struct swap_list_t swap_list = {-1, -1}; |
1da177e4c
|
55 |
|
efa90a981
|
56 |
static struct swap_info_struct *swap_info[MAX_SWAPFILES]; |
1da177e4c
|
57 |
|
fc0abb145
|
58 |
static DEFINE_MUTEX(swapon_mutex); |
1da177e4c
|
59 |
|
66d7dd518
|
60 61 62 |
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
|
63 |
static inline unsigned char swap_count(unsigned char ent) |
355cfa73d
|
64 |
{ |
570a335b8
|
65 |
return ent & ~SWAP_HAS_CACHE; /* may include SWAP_HAS_CONT flag */ |
355cfa73d
|
66 |
} |
efa90a981
|
67 |
/* returns 1 if swap entry is freed */ |
c9e444103
|
68 69 70 |
static int __try_to_reclaim_swap(struct swap_info_struct *si, unsigned long offset) { |
efa90a981
|
71 |
swp_entry_t entry = swp_entry(si->type, offset); |
c9e444103
|
72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 |
struct page *page; int ret = 0; page = find_get_page(&swapper_space, entry.val); 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); } page_cache_release(page); return ret; } |
355cfa73d
|
92 |
|
1da177e4c
|
93 94 |
/* * We need this because the bdev->unplug_fn can sleep and we cannot |
5d337b919
|
95 |
* hold swap_lock while calling the unplug_fn. And swap_lock |
fc0abb145
|
96 |
* cannot be turned into a mutex. |
1da177e4c
|
97 98 |
*/ static DECLARE_RWSEM(swap_unplug_sem); |
1da177e4c
|
99 100 101 102 103 |
void swap_unplug_io_fn(struct backing_dev_info *unused_bdi, struct page *page) { swp_entry_t entry; down_read(&swap_unplug_sem); |
4c21e2f24
|
104 |
entry.val = page_private(page); |
1da177e4c
|
105 |
if (PageSwapCache(page)) { |
efa90a981
|
106 |
struct block_device *bdev = swap_info[swp_type(entry)]->bdev; |
1da177e4c
|
107 108 109 110 111 |
struct backing_dev_info *bdi; /* * If the page is removed from swapcache from under us (with a * racy try_to_unuse/swapoff) we need an additional reference |
4c21e2f24
|
112 113 |
* count to avoid reading garbage from page_private(page) above. * If the WARN_ON triggers during a swapoff it maybe the race |
1da177e4c
|
114 115 116 117 118 119 |
* condition and it's harmless. However if it triggers without * swapoff it signals a problem. */ WARN_ON(page_count(page) <= 1); bdi = bdev->bd_inode->i_mapping->backing_dev_info; |
ba32311eb
|
120 |
blk_run_backing_dev(bdi, page); |
1da177e4c
|
121 122 123 |
} up_read(&swap_unplug_sem); } |
6a6ba8317
|
124 125 126 127 128 129 130 |
/* * 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
|
131 132 |
sector_t start_block; sector_t nr_blocks; |
6a6ba8317
|
133 |
int err = 0; |
9625a5f28
|
134 135 136 137 138 139 |
/* 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
|
140 |
nr_blocks, GFP_KERNEL, 0); |
9625a5f28
|
141 142 143 144 |
if (err) return err; cond_resched(); } |
6a6ba8317
|
145 |
|
9625a5f28
|
146 147 148 |
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
|
149 150 |
err = blkdev_issue_discard(si->bdev, start_block, |
dd3932edd
|
151 |
nr_blocks, GFP_KERNEL, 0); |
6a6ba8317
|
152 153 154 155 156 157 158 |
if (err) break; cond_resched(); } return err; /* That will often be -EOPNOTSUPP */ } |
7992fde72
|
159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 |
/* * 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) { struct list_head *lh; 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
|
176 |
sector_t nr_blocks = se->nr_pages - offset; |
7992fde72
|
177 178 179 180 181 182 183 184 185 186 187 188 |
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
|
189 |
nr_blocks, GFP_NOIO, 0)) |
7992fde72
|
190 191 192 193 |
break; } lh = se->list.next; |
7992fde72
|
194 195 196 197 198 199 200 201 202 |
se = list_entry(lh, struct swap_extent, list); } } static int wait_for_discard(void *word) { schedule(); return 0; } |
048c27fd7
|
203 204 |
#define SWAPFILE_CLUSTER 256 #define LATENCY_LIMIT 256 |
355cfa73d
|
205 |
static inline unsigned long scan_swap_map(struct swap_info_struct *si, |
8d69aaee8
|
206 |
unsigned char usage) |
1da177e4c
|
207 |
{ |
ebebbbe90
|
208 |
unsigned long offset; |
c60aa176c
|
209 |
unsigned long scan_base; |
7992fde72
|
210 |
unsigned long last_in_cluster = 0; |
048c27fd7
|
211 |
int latency_ration = LATENCY_LIMIT; |
7992fde72
|
212 |
int found_free_cluster = 0; |
7dfad4183
|
213 |
|
886bb7e9c
|
214 |
/* |
7dfad4183
|
215 216 217 218 219 220 221 |
* 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
|
222 |
* And we let swap pages go all over an SSD partition. Hugh |
7dfad4183
|
223 |
*/ |
52b7efdbe
|
224 |
si->flags += SWP_SCANNING; |
c60aa176c
|
225 |
scan_base = offset = si->cluster_next; |
ebebbbe90
|
226 227 228 229 230 231 |
if (unlikely(!si->cluster_nr--)) { if (si->pages - si->inuse_pages < SWAPFILE_CLUSTER) { si->cluster_nr = SWAPFILE_CLUSTER - 1; goto checks; } |
7992fde72
|
232 233 234 235 236 237 238 239 240 241 242 243 244 |
if (si->flags & SWP_DISCARDABLE) { /* * Start range check on racing allocations, in case * they overlap the cluster we eventually decide on * (we scan without swap_lock to allow preemption). * It's hardly conceivable that cluster_nr could be * wrapped during our scan, but don't depend on it. */ if (si->lowest_alloc) goto checks; si->lowest_alloc = si->max; si->highest_alloc = 0; } |
5d337b919
|
245 |
spin_unlock(&swap_lock); |
7dfad4183
|
246 |
|
c60aa176c
|
247 248 249 250 251 252 253 254 255 256 |
/* * If seek is expensive, start searching for new cluster from * start of partition, to minimize the span of allocated swap. * But if seek is cheap, search from our current position, so * that swap is allocated from all over the partition: if the * Flash Translation Layer only remaps within limited zones, * we don't want to wear out the first zone too quickly. */ if (!(si->flags & SWP_SOLIDSTATE)) scan_base = offset = si->lowest_bit; |
7dfad4183
|
257 258 259 260 |
last_in_cluster = offset + SWAPFILE_CLUSTER - 1; /* Locate the first empty (unaligned) cluster */ for (; last_in_cluster <= si->highest_bit; offset++) { |
1da177e4c
|
261 |
if (si->swap_map[offset]) |
7dfad4183
|
262 263 |
last_in_cluster = offset + SWAPFILE_CLUSTER; else if (offset == last_in_cluster) { |
5d337b919
|
264 |
spin_lock(&swap_lock); |
ebebbbe90
|
265 266 267 |
offset -= SWAPFILE_CLUSTER - 1; si->cluster_next = offset; si->cluster_nr = SWAPFILE_CLUSTER - 1; |
7992fde72
|
268 |
found_free_cluster = 1; |
ebebbbe90
|
269 |
goto checks; |
1da177e4c
|
270 |
} |
048c27fd7
|
271 272 273 274 |
if (unlikely(--latency_ration < 0)) { cond_resched(); latency_ration = LATENCY_LIMIT; } |
7dfad4183
|
275 |
} |
ebebbbe90
|
276 277 |
offset = si->lowest_bit; |
c60aa176c
|
278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 |
last_in_cluster = offset + SWAPFILE_CLUSTER - 1; /* Locate the first empty (unaligned) cluster */ for (; last_in_cluster < scan_base; offset++) { if (si->swap_map[offset]) last_in_cluster = offset + SWAPFILE_CLUSTER; else if (offset == last_in_cluster) { spin_lock(&swap_lock); offset -= SWAPFILE_CLUSTER - 1; si->cluster_next = offset; si->cluster_nr = SWAPFILE_CLUSTER - 1; found_free_cluster = 1; goto checks; } if (unlikely(--latency_ration < 0)) { cond_resched(); latency_ration = LATENCY_LIMIT; } } offset = scan_base; |
5d337b919
|
299 |
spin_lock(&swap_lock); |
ebebbbe90
|
300 |
si->cluster_nr = SWAPFILE_CLUSTER - 1; |
7992fde72
|
301 |
si->lowest_alloc = 0; |
1da177e4c
|
302 |
} |
7dfad4183
|
303 |
|
ebebbbe90
|
304 305 |
checks: if (!(si->flags & SWP_WRITEOK)) |
52b7efdbe
|
306 |
goto no_page; |
7dfad4183
|
307 308 |
if (!si->highest_bit) goto no_page; |
ebebbbe90
|
309 |
if (offset > si->highest_bit) |
c60aa176c
|
310 |
scan_base = offset = si->lowest_bit; |
c9e444103
|
311 |
|
b73d7fcec
|
312 313 |
/* reuse swap entry of cache-only swap if not busy. */ if (vm_swap_full() && si->swap_map[offset] == SWAP_HAS_CACHE) { |
c9e444103
|
314 315 316 317 318 319 320 321 322 |
int swap_was_freed; spin_unlock(&swap_lock); swap_was_freed = __try_to_reclaim_swap(si, offset); spin_lock(&swap_lock); /* entry was freed successfully, try to use this again */ if (swap_was_freed) goto checks; goto scan; /* check next one */ } |
ebebbbe90
|
323 324 325 326 327 328 329 330 331 332 333 |
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; |
1da177e4c
|
334 |
} |
253d553ba
|
335 |
si->swap_map[offset] = usage; |
ebebbbe90
|
336 337 |
si->cluster_next = offset + 1; si->flags -= SWP_SCANNING; |
7992fde72
|
338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 |
if (si->lowest_alloc) { /* * Only set when SWP_DISCARDABLE, and there's a scan * for a free cluster in progress or just completed. */ if (found_free_cluster) { /* * To optimize wear-levelling, discard the * old data of the cluster, taking care not to * discard any of its pages that have already * been allocated by racing tasks (offset has * already stepped over any at the beginning). */ if (offset < si->highest_alloc && si->lowest_alloc <= last_in_cluster) last_in_cluster = si->lowest_alloc - 1; si->flags |= SWP_DISCARDING; spin_unlock(&swap_lock); if (offset < last_in_cluster) discard_swap_cluster(si, offset, last_in_cluster - offset + 1); spin_lock(&swap_lock); si->lowest_alloc = 0; si->flags &= ~SWP_DISCARDING; smp_mb(); /* wake_up_bit advises this */ wake_up_bit(&si->flags, ilog2(SWP_DISCARDING)); } else if (si->flags & SWP_DISCARDING) { /* * Delay using pages allocated by racing tasks * until the whole discard has been issued. We * could defer that delay until swap_writepage, * but it's easier to keep this self-contained. */ spin_unlock(&swap_lock); wait_on_bit(&si->flags, ilog2(SWP_DISCARDING), wait_for_discard, TASK_UNINTERRUPTIBLE); spin_lock(&swap_lock); } else { /* * Note pages allocated by racing tasks while * scan for a free cluster is in progress, so * that its final discard can exclude them. */ if (offset < si->lowest_alloc) si->lowest_alloc = offset; if (offset > si->highest_alloc) si->highest_alloc = offset; } } |
ebebbbe90
|
392 |
return offset; |
7dfad4183
|
393 |
|
ebebbbe90
|
394 |
scan: |
5d337b919
|
395 |
spin_unlock(&swap_lock); |
7dfad4183
|
396 |
while (++offset <= si->highest_bit) { |
52b7efdbe
|
397 |
if (!si->swap_map[offset]) { |
5d337b919
|
398 |
spin_lock(&swap_lock); |
52b7efdbe
|
399 400 |
goto checks; } |
c9e444103
|
401 402 403 404 |
if (vm_swap_full() && si->swap_map[offset] == SWAP_HAS_CACHE) { spin_lock(&swap_lock); goto checks; } |
048c27fd7
|
405 406 407 408 |
if (unlikely(--latency_ration < 0)) { cond_resched(); latency_ration = LATENCY_LIMIT; } |
7dfad4183
|
409 |
} |
c60aa176c
|
410 411 412 413 414 415 |
offset = si->lowest_bit; while (++offset < scan_base) { if (!si->swap_map[offset]) { spin_lock(&swap_lock); goto checks; } |
c9e444103
|
416 417 418 419 |
if (vm_swap_full() && si->swap_map[offset] == SWAP_HAS_CACHE) { spin_lock(&swap_lock); goto checks; } |
c60aa176c
|
420 421 422 423 424 |
if (unlikely(--latency_ration < 0)) { cond_resched(); latency_ration = LATENCY_LIMIT; } } |
5d337b919
|
425 |
spin_lock(&swap_lock); |
7dfad4183
|
426 427 |
no_page: |
52b7efdbe
|
428 |
si->flags -= SWP_SCANNING; |
1da177e4c
|
429 430 431 432 433 |
return 0; } swp_entry_t get_swap_page(void) { |
fb4f88dca
|
434 435 436 437 |
struct swap_info_struct *si; pgoff_t offset; int type, next; int wrapped = 0; |
1da177e4c
|
438 |
|
5d337b919
|
439 |
spin_lock(&swap_lock); |
1da177e4c
|
440 |
if (nr_swap_pages <= 0) |
fb4f88dca
|
441 442 443 444 |
goto noswap; nr_swap_pages--; for (type = swap_list.next; type >= 0 && wrapped < 2; type = next) { |
efa90a981
|
445 |
si = swap_info[type]; |
fb4f88dca
|
446 447 |
next = si->next; if (next < 0 || |
efa90a981
|
448 |
(!wrapped && si->prio != swap_info[next]->prio)) { |
fb4f88dca
|
449 450 |
next = swap_list.head; wrapped++; |
1da177e4c
|
451 |
} |
fb4f88dca
|
452 453 454 455 456 457 458 |
if (!si->highest_bit) continue; if (!(si->flags & SWP_WRITEOK)) continue; swap_list.next = next; |
355cfa73d
|
459 |
/* This is called for allocating swap entry for cache */ |
253d553ba
|
460 |
offset = scan_swap_map(si, SWAP_HAS_CACHE); |
5d337b919
|
461 462 |
if (offset) { spin_unlock(&swap_lock); |
fb4f88dca
|
463 |
return swp_entry(type, offset); |
5d337b919
|
464 |
} |
fb4f88dca
|
465 |
next = swap_list.next; |
1da177e4c
|
466 |
} |
fb4f88dca
|
467 468 469 |
nr_swap_pages++; noswap: |
5d337b919
|
470 |
spin_unlock(&swap_lock); |
fb4f88dca
|
471 |
return (swp_entry_t) {0}; |
1da177e4c
|
472 |
} |
910321ea8
|
473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 |
/* The only caller of this function is now susupend routine */ swp_entry_t get_swap_page_of_type(int type) { struct swap_info_struct *si; pgoff_t offset; spin_lock(&swap_lock); si = swap_info[type]; if (si && (si->flags & SWP_WRITEOK)) { nr_swap_pages--; /* This is called for allocating swap entry, not cache */ offset = scan_swap_map(si, 1); if (offset) { spin_unlock(&swap_lock); return swp_entry(type, offset); } nr_swap_pages++; } spin_unlock(&swap_lock); return (swp_entry_t) {0}; } |
73c34b6ac
|
494 |
static struct swap_info_struct *swap_info_get(swp_entry_t entry) |
1da177e4c
|
495 |
{ |
73c34b6ac
|
496 |
struct swap_info_struct *p; |
1da177e4c
|
497 498 499 500 501 502 503 |
unsigned long offset, type; if (!entry.val) goto out; type = swp_type(entry); if (type >= nr_swapfiles) goto bad_nofile; |
efa90a981
|
504 |
p = swap_info[type]; |
1da177e4c
|
505 506 507 508 509 510 511 |
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; |
5d337b919
|
512 |
spin_lock(&swap_lock); |
1da177e4c
|
513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 |
return p; bad_free: printk(KERN_ERR "swap_free: %s%08lx ", Unused_offset, entry.val); goto out; bad_offset: printk(KERN_ERR "swap_free: %s%08lx ", Bad_offset, entry.val); goto out; bad_device: printk(KERN_ERR "swap_free: %s%08lx ", Unused_file, entry.val); goto out; bad_nofile: printk(KERN_ERR "swap_free: %s%08lx ", Bad_file, entry.val); out: return NULL; |
886bb7e9c
|
532 |
} |
1da177e4c
|
533 |
|
8d69aaee8
|
534 535 |
static unsigned char swap_entry_free(struct swap_info_struct *p, swp_entry_t entry, unsigned char usage) |
1da177e4c
|
536 |
{ |
253d553ba
|
537 |
unsigned long offset = swp_offset(entry); |
8d69aaee8
|
538 539 |
unsigned char count; unsigned char has_cache; |
355cfa73d
|
540 |
|
253d553ba
|
541 542 543 |
count = p->swap_map[offset]; has_cache = count & SWAP_HAS_CACHE; count &= ~SWAP_HAS_CACHE; |
355cfa73d
|
544 |
|
253d553ba
|
545 |
if (usage == SWAP_HAS_CACHE) { |
355cfa73d
|
546 |
VM_BUG_ON(!has_cache); |
253d553ba
|
547 |
has_cache = 0; |
aaa468653
|
548 549 550 551 552 553 |
} 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
|
554 555 556 557 558 559 560 561 562 |
} 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
|
563 564 565 566 567 568 |
if (!count) mem_cgroup_uncharge_swap(entry); usage = count | has_cache; p->swap_map[offset] = usage; |
355cfa73d
|
569 |
|
355cfa73d
|
570 |
/* free if no reference */ |
253d553ba
|
571 |
if (!usage) { |
b3a27d052
|
572 |
struct gendisk *disk = p->bdev->bd_disk; |
355cfa73d
|
573 574 575 576 |
if (offset < p->lowest_bit) p->lowest_bit = offset; if (offset > p->highest_bit) p->highest_bit = offset; |
efa90a981
|
577 578 579 |
if (swap_list.next >= 0 && p->prio > swap_info[swap_list.next]->prio) swap_list.next = p->type; |
355cfa73d
|
580 581 |
nr_swap_pages++; p->inuse_pages--; |
b3a27d052
|
582 583 584 |
if ((p->flags & SWP_BLKDEV) && disk->fops->swap_slot_free_notify) disk->fops->swap_slot_free_notify(p->bdev, offset); |
1da177e4c
|
585 |
} |
253d553ba
|
586 587 |
return usage; |
1da177e4c
|
588 589 590 591 592 593 594 595 |
} /* * Caller has made sure that the swapdevice corresponding to entry * is still around or has not been recycled. */ void swap_free(swp_entry_t entry) { |
73c34b6ac
|
596 |
struct swap_info_struct *p; |
1da177e4c
|
597 598 599 |
p = swap_info_get(entry); if (p) { |
253d553ba
|
600 |
swap_entry_free(p, entry, 1); |
5d337b919
|
601 |
spin_unlock(&swap_lock); |
1da177e4c
|
602 603 604 605 |
} } /* |
cb4b86ba4
|
606 607 608 609 |
* Called after dropping swapcache to decrease refcnt to swap entries. */ void swapcache_free(swp_entry_t entry, struct page *page) { |
355cfa73d
|
610 |
struct swap_info_struct *p; |
8d69aaee8
|
611 |
unsigned char count; |
355cfa73d
|
612 |
|
355cfa73d
|
613 614 |
p = swap_info_get(entry); if (p) { |
253d553ba
|
615 616 617 |
count = swap_entry_free(p, entry, SWAP_HAS_CACHE); if (page) mem_cgroup_uncharge_swapcache(page, entry, count != 0); |
355cfa73d
|
618 619 |
spin_unlock(&swap_lock); } |
cb4b86ba4
|
620 621 622 |
} /* |
c475a8ab6
|
623 |
* How many references to page are currently swapped out? |
570a335b8
|
624 625 |
* 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
|
626 |
*/ |
c475a8ab6
|
627 |
static inline int page_swapcount(struct page *page) |
1da177e4c
|
628 |
{ |
c475a8ab6
|
629 630 |
int count = 0; struct swap_info_struct *p; |
1da177e4c
|
631 |
swp_entry_t entry; |
4c21e2f24
|
632 |
entry.val = page_private(page); |
1da177e4c
|
633 634 |
p = swap_info_get(entry); if (p) { |
355cfa73d
|
635 |
count = swap_count(p->swap_map[swp_offset(entry)]); |
5d337b919
|
636 |
spin_unlock(&swap_lock); |
1da177e4c
|
637 |
} |
c475a8ab6
|
638 |
return count; |
1da177e4c
|
639 640 641 |
} /* |
7b1fe5979
|
642 643 644 645 |
* 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
|
646 |
*/ |
7b1fe5979
|
647 |
int reuse_swap_page(struct page *page) |
1da177e4c
|
648 |
{ |
c475a8ab6
|
649 |
int count; |
51726b122
|
650 |
VM_BUG_ON(!PageLocked(page)); |
5ad646880
|
651 652 |
if (unlikely(PageKsm(page))) return 0; |
c475a8ab6
|
653 |
count = page_mapcount(page); |
7b1fe5979
|
654 |
if (count <= 1 && PageSwapCache(page)) { |
c475a8ab6
|
655 |
count += page_swapcount(page); |
7b1fe5979
|
656 657 658 659 660 |
if (count == 1 && !PageWriteback(page)) { delete_from_swap_cache(page); SetPageDirty(page); } } |
5ad646880
|
661 |
return count <= 1; |
1da177e4c
|
662 663 664 |
} /* |
a2c43eed8
|
665 666 |
* 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
|
667 |
*/ |
a2c43eed8
|
668 |
int try_to_free_swap(struct page *page) |
1da177e4c
|
669 |
{ |
51726b122
|
670 |
VM_BUG_ON(!PageLocked(page)); |
1da177e4c
|
671 672 673 674 675 |
if (!PageSwapCache(page)) return 0; if (PageWriteback(page)) return 0; |
a2c43eed8
|
676 |
if (page_swapcount(page)) |
1da177e4c
|
677 |
return 0; |
b73d7fcec
|
678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 |
/* * 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. * * Hibernation clears bits from gfp_allowed_mask to prevent * memory reclaim from writing to disk, so check that here. */ if (!(gfp_allowed_mask & __GFP_IO)) return 0; |
a2c43eed8
|
695 696 697 |
delete_from_swap_cache(page); SetPageDirty(page); return 1; |
68a22394c
|
698 699 700 |
} /* |
1da177e4c
|
701 702 703 |
* Free the swap entry like above, but also try to * free the page cache entry if it is the last user. */ |
2509ef26d
|
704 |
int free_swap_and_cache(swp_entry_t entry) |
1da177e4c
|
705 |
{ |
2509ef26d
|
706 |
struct swap_info_struct *p; |
1da177e4c
|
707 |
struct page *page = NULL; |
a7420aa54
|
708 |
if (non_swap_entry(entry)) |
2509ef26d
|
709 |
return 1; |
0697212a4
|
710 |
|
1da177e4c
|
711 712 |
p = swap_info_get(entry); if (p) { |
253d553ba
|
713 |
if (swap_entry_free(p, entry, 1) == SWAP_HAS_CACHE) { |
93fac7041
|
714 |
page = find_get_page(&swapper_space, entry.val); |
8413ac9d8
|
715 |
if (page && !trylock_page(page)) { |
93fac7041
|
716 717 718 719 |
page_cache_release(page); page = NULL; } } |
5d337b919
|
720 |
spin_unlock(&swap_lock); |
1da177e4c
|
721 722 |
} if (page) { |
a2c43eed8
|
723 724 725 726 |
/* * Not mapped elsewhere, or swap space full? Free it! * Also recheck PageSwapCache now page is locked (above). */ |
93fac7041
|
727 |
if (PageSwapCache(page) && !PageWriteback(page) && |
a2c43eed8
|
728 |
(!page_mapped(page) || vm_swap_full())) { |
1da177e4c
|
729 730 731 732 733 734 |
delete_from_swap_cache(page); SetPageDirty(page); } unlock_page(page); page_cache_release(page); } |
2509ef26d
|
735 |
return p != NULL; |
1da177e4c
|
736 |
} |
024914477
|
737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 |
#ifdef CONFIG_CGROUP_MEM_RES_CTLR /** * mem_cgroup_count_swap_user - count the user of a swap entry * @ent: the swap entry to be checked * @pagep: the pointer for the swap cache page of the entry to be stored * * Returns the number of the user of the swap entry. The number is valid only * for swaps of anonymous pages. * If the entry is found on swap cache, the page is stored to pagep with * refcount of it being incremented. */ int mem_cgroup_count_swap_user(swp_entry_t ent, struct page **pagep) { struct page *page; struct swap_info_struct *p; int count = 0; page = find_get_page(&swapper_space, ent.val); if (page) count += page_mapcount(page); p = swap_info_get(ent); if (p) { count += swap_count(p->swap_map[swp_offset(ent)]); spin_unlock(&swap_lock); } *pagep = page; return count; } #endif |
b0cb1a19d
|
767 |
#ifdef CONFIG_HIBERNATION |
f577eb30a
|
768 |
/* |
915bae9eb
|
769 |
* Find the swap type that corresponds to given device (if any). |
f577eb30a
|
770 |
* |
915bae9eb
|
771 772 773 774 |
* @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
|
775 |
*/ |
7bf236874
|
776 |
int swap_type_of(dev_t device, sector_t offset, struct block_device **bdev_p) |
f577eb30a
|
777 |
{ |
915bae9eb
|
778 |
struct block_device *bdev = NULL; |
efa90a981
|
779 |
int type; |
f577eb30a
|
780 |
|
915bae9eb
|
781 782 |
if (device) bdev = bdget(device); |
f577eb30a
|
783 |
spin_lock(&swap_lock); |
efa90a981
|
784 785 |
for (type = 0; type < nr_swapfiles; type++) { struct swap_info_struct *sis = swap_info[type]; |
f577eb30a
|
786 |
|
915bae9eb
|
787 |
if (!(sis->flags & SWP_WRITEOK)) |
f577eb30a
|
788 |
continue; |
b6b5bce35
|
789 |
|
915bae9eb
|
790 |
if (!bdev) { |
7bf236874
|
791 |
if (bdev_p) |
dddac6a7b
|
792 |
*bdev_p = bdgrab(sis->bdev); |
7bf236874
|
793 |
|
6e1819d61
|
794 |
spin_unlock(&swap_lock); |
efa90a981
|
795 |
return type; |
6e1819d61
|
796 |
} |
915bae9eb
|
797 |
if (bdev == sis->bdev) { |
9625a5f28
|
798 |
struct swap_extent *se = &sis->first_swap_extent; |
915bae9eb
|
799 |
|
915bae9eb
|
800 |
if (se->start_block == offset) { |
7bf236874
|
801 |
if (bdev_p) |
dddac6a7b
|
802 |
*bdev_p = bdgrab(sis->bdev); |
7bf236874
|
803 |
|
915bae9eb
|
804 805 |
spin_unlock(&swap_lock); bdput(bdev); |
efa90a981
|
806 |
return type; |
915bae9eb
|
807 |
} |
f577eb30a
|
808 809 810 |
} } spin_unlock(&swap_lock); |
915bae9eb
|
811 812 |
if (bdev) bdput(bdev); |
f577eb30a
|
813 814 815 816 |
return -ENODEV; } /* |
73c34b6ac
|
817 818 819 820 821 822 823 824 825 826 827 |
* 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
|
828 |
return map_swap_entry(swp_entry(type, offset), &bdev); |
73c34b6ac
|
829 830 831 |
} /* |
f577eb30a
|
832 833 834 835 836 837 838 839 |
* 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
|
840 841 842 843 844 845 |
spin_lock(&swap_lock); if ((unsigned int)type < nr_swapfiles) { struct swap_info_struct *sis = swap_info[type]; if (sis->flags & SWP_WRITEOK) { n = sis->pages; |
f577eb30a
|
846 |
if (free) |
efa90a981
|
847 |
n -= sis->inuse_pages; |
f577eb30a
|
848 |
} |
f577eb30a
|
849 |
} |
efa90a981
|
850 |
spin_unlock(&swap_lock); |
f577eb30a
|
851 852 |
return n; } |
73c34b6ac
|
853 |
#endif /* CONFIG_HIBERNATION */ |
f577eb30a
|
854 |
|
1da177e4c
|
855 |
/* |
72866f6f2
|
856 857 858 |
* 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
|
859 |
*/ |
044d66c1d
|
860 |
static int unuse_pte(struct vm_area_struct *vma, pmd_t *pmd, |
1da177e4c
|
861 862 |
unsigned long addr, swp_entry_t entry, struct page *page) { |
7a81b88cb
|
863 |
struct mem_cgroup *ptr = NULL; |
044d66c1d
|
864 865 866 |
spinlock_t *ptl; pte_t *pte; int ret = 1; |
85d9fc89f
|
867 |
if (mem_cgroup_try_charge_swapin(vma->vm_mm, page, GFP_KERNEL, &ptr)) { |
044d66c1d
|
868 |
ret = -ENOMEM; |
85d9fc89f
|
869 870 |
goto out_nolock; } |
044d66c1d
|
871 872 873 874 |
pte = pte_offset_map_lock(vma->vm_mm, pmd, addr, &ptl); if (unlikely(!pte_same(*pte, swp_entry_to_pte(entry)))) { if (ret > 0) |
7a81b88cb
|
875 |
mem_cgroup_cancel_charge_swapin(ptr); |
044d66c1d
|
876 877 878 |
ret = 0; goto out; } |
8a9f3ccd2
|
879 |
|
b084d4353
|
880 |
dec_mm_counter(vma->vm_mm, MM_SWAPENTS); |
d559db086
|
881 |
inc_mm_counter(vma->vm_mm, MM_ANONPAGES); |
1da177e4c
|
882 883 884 885 |
get_page(page); set_pte_at(vma->vm_mm, addr, pte, pte_mkold(mk_pte(page, vma->vm_page_prot))); page_add_anon_rmap(page, vma, addr); |
7a81b88cb
|
886 |
mem_cgroup_commit_charge_swapin(page, ptr); |
1da177e4c
|
887 888 889 890 891 892 |
swap_free(entry); /* * Move the page to the active list so it is not * immediately swapped out again after swapon. */ activate_page(page); |
044d66c1d
|
893 894 |
out: pte_unmap_unlock(pte, ptl); |
85d9fc89f
|
895 |
out_nolock: |
044d66c1d
|
896 |
return ret; |
1da177e4c
|
897 898 899 900 901 902 |
} 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
|
903 |
pte_t swp_pte = swp_entry_to_pte(entry); |
705e87c0c
|
904 |
pte_t *pte; |
8a9f3ccd2
|
905 |
int ret = 0; |
1da177e4c
|
906 |
|
044d66c1d
|
907 908 909 910 911 912 913 914 915 916 |
/* * 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 * preemptible whenever CONFIG_PREEMPT but not CONFIG_HIGHPTE. */ pte = pte_offset_map(pmd, addr); |
1da177e4c
|
917 918 919 920 921 922 |
do { /* * swapoff spends a _lot_ of time in this loop! * Test inline before going to call unuse_pte. */ if (unlikely(pte_same(*pte, swp_pte))) { |
044d66c1d
|
923 924 925 926 927 |
pte_unmap(pte); ret = unuse_pte(vma, pmd, addr, entry, page); if (ret) goto out; pte = pte_offset_map(pmd, addr); |
1da177e4c
|
928 929 |
} } while (pte++, addr += PAGE_SIZE, addr != end); |
044d66c1d
|
930 931 |
pte_unmap(pte - 1); out: |
8a9f3ccd2
|
932 |
return ret; |
1da177e4c
|
933 934 935 936 937 938 939 940 |
} 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
|
941 |
int ret; |
1da177e4c
|
942 943 944 945 |
pmd = pmd_offset(pud, addr); do { next = pmd_addr_end(addr, end); |
3f04f62f9
|
946 947 |
if (unlikely(pmd_trans_huge(*pmd))) continue; |
1da177e4c
|
948 949 |
if (pmd_none_or_clear_bad(pmd)) continue; |
8a9f3ccd2
|
950 951 952 |
ret = unuse_pte_range(vma, pmd, addr, next, entry, page); if (ret) return ret; |
1da177e4c
|
953 954 955 956 957 958 959 960 961 962 |
} 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
|
963 |
int ret; |
1da177e4c
|
964 965 966 967 968 969 |
pud = pud_offset(pgd, addr); do { next = pud_addr_end(addr, end); if (pud_none_or_clear_bad(pud)) continue; |
8a9f3ccd2
|
970 971 972 |
ret = unuse_pmd_range(vma, pud, addr, next, entry, page); if (ret) return ret; |
1da177e4c
|
973 974 975 976 977 978 979 980 981 |
} 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
|
982 |
int ret; |
1da177e4c
|
983 |
|
3ca7b3c5b
|
984 |
if (page_anon_vma(page)) { |
1da177e4c
|
985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 |
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
|
1000 1001 1002 |
ret = unuse_pud_range(vma, pgd, addr, next, entry, page); if (ret) return ret; |
1da177e4c
|
1003 1004 1005 1006 1007 1008 1009 1010 |
} 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
|
1011 |
int ret = 0; |
1da177e4c
|
1012 1013 1014 |
if (!down_read_trylock(&mm->mmap_sem)) { /* |
7d03431cf
|
1015 1016 |
* Activate page so shrink_inactive_list is unlikely to unmap * its ptes while lock is dropped, so swapoff can make progress. |
1da177e4c
|
1017 |
*/ |
c475a8ab6
|
1018 |
activate_page(page); |
1da177e4c
|
1019 1020 1021 1022 |
unlock_page(page); down_read(&mm->mmap_sem); lock_page(page); } |
1da177e4c
|
1023 |
for (vma = mm->mmap; vma; vma = vma->vm_next) { |
8a9f3ccd2
|
1024 |
if (vma->anon_vma && (ret = unuse_vma(vma, entry, page))) |
1da177e4c
|
1025 1026 |
break; } |
1da177e4c
|
1027 |
up_read(&mm->mmap_sem); |
8a9f3ccd2
|
1028 |
return (ret < 0)? ret: 0; |
1da177e4c
|
1029 1030 1031 1032 1033 1034 |
} /* * Scan swap_map from current position to next entry still in use. * Recycle to start on reaching the end, returning 0 when empty. */ |
6eb396dc4
|
1035 1036 |
static unsigned int find_next_to_unuse(struct swap_info_struct *si, unsigned int prev) |
1da177e4c
|
1037 |
{ |
6eb396dc4
|
1038 1039 |
unsigned int max = si->max; unsigned int i = prev; |
8d69aaee8
|
1040 |
unsigned char count; |
1da177e4c
|
1041 1042 |
/* |
5d337b919
|
1043 |
* No need for swap_lock here: we're just looking |
1da177e4c
|
1044 1045 |
* for whether an entry is in use, not modifying it; false * hits are okay, and sys_swapoff() has already prevented new |
5d337b919
|
1046 |
* allocations from this area (while holding swap_lock). |
1da177e4c
|
1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 |
*/ 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; } count = si->swap_map[i]; |
355cfa73d
|
1063 |
if (count && swap_count(count) != SWAP_MAP_BAD) |
1da177e4c
|
1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 |
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. */ static int try_to_unuse(unsigned int type) { |
efa90a981
|
1076 |
struct swap_info_struct *si = swap_info[type]; |
1da177e4c
|
1077 |
struct mm_struct *start_mm; |
8d69aaee8
|
1078 1079 |
unsigned char *swap_map; unsigned char swcount; |
1da177e4c
|
1080 1081 |
struct page *page; swp_entry_t entry; |
6eb396dc4
|
1082 |
unsigned int i = 0; |
1da177e4c
|
1083 |
int retval = 0; |
1da177e4c
|
1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 |
/* * 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
|
1097 |
* that. |
1da177e4c
|
1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 |
*/ 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. */ while ((i = find_next_to_unuse(si, i)) != 0) { if (signal_pending(current)) { retval = -EINTR; break; } |
886bb7e9c
|
1112 |
/* |
1da177e4c
|
1113 1114 |
* Get a page for the entry, using the existing swap * cache page if there is one. Otherwise, get a clean |
886bb7e9c
|
1115 |
* page and read the swap into it. |
1da177e4c
|
1116 1117 1118 |
*/ swap_map = &si->swap_map[i]; entry = swp_entry(type, i); |
02098feaa
|
1119 1120 |
page = read_swap_cache_async(entry, GFP_HIGHUSER_MOVABLE, NULL, 0); |
1da177e4c
|
1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 |
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. */ if (!*swap_map) 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
|
1158 |
*/ |
1da177e4c
|
1159 |
swcount = *swap_map; |
aaa468653
|
1160 1161 1162 1163 1164 1165 |
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
|
1166 |
} |
aaa468653
|
1167 1168 |
if (swap_count(swcount) && start_mm != &init_mm) retval = unuse_mm(start_mm, entry, page); |
355cfa73d
|
1169 |
if (swap_count(*swap_map)) { |
1da177e4c
|
1170 1171 1172 1173 1174 1175 1176 1177 1178 |
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
|
1179 |
while (swap_count(*swap_map) && !retval && |
1da177e4c
|
1180 1181 |
(p = p->next) != &start_mm->mmlist) { mm = list_entry(p, struct mm_struct, mmlist); |
70af7c5c6
|
1182 |
if (!atomic_inc_not_zero(&mm->mm_users)) |
1da177e4c
|
1183 |
continue; |
1da177e4c
|
1184 1185 1186 1187 1188 1189 1190 |
spin_unlock(&mmlist_lock); mmput(prev_mm); prev_mm = mm; cond_resched(); swcount = *swap_map; |
355cfa73d
|
1191 |
if (!swap_count(swcount)) /* any usage ? */ |
1da177e4c
|
1192 |
; |
aaa468653
|
1193 |
else if (mm == &init_mm) |
1da177e4c
|
1194 |
set_start_mm = 1; |
aaa468653
|
1195 |
else |
1da177e4c
|
1196 |
retval = unuse_mm(mm, entry, page); |
355cfa73d
|
1197 |
|
32c5fc10e
|
1198 |
if (set_start_mm && *swap_map < swcount) { |
1da177e4c
|
1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 |
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); page_cache_release(page); break; } /* |
1da177e4c
|
1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 |
* 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
|
1229 1230 1231 1232 1233 1234 |
* * 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
|
1235 |
*/ |
355cfa73d
|
1236 1237 |
if (swap_count(*swap_map) && PageDirty(page) && PageSwapCache(page)) { |
1da177e4c
|
1238 1239 1240 1241 1242 1243 1244 1245 |
struct writeback_control wbc = { .sync_mode = WB_SYNC_NONE, }; swap_writepage(page, &wbc); lock_page(page); wait_on_page_writeback(page); } |
68bdc8d64
|
1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 |
/* * 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
|
1256 |
delete_from_swap_cache(page); |
1da177e4c
|
1257 1258 1259 1260 |
/* * So we could skip searching mms once swap count went * to 1, we did not mark any present ptes as dirty: must |
2706a1b89
|
1261 |
* mark page dirty so shrink_page_list will preserve it. |
1da177e4c
|
1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 |
*/ SetPageDirty(page); unlock_page(page); page_cache_release(page); /* * Make sure that we aren't completely killing * interactive performance. */ cond_resched(); } mmput(start_mm); |
1da177e4c
|
1275 1276 1277 1278 |
return retval; } /* |
5d337b919
|
1279 1280 1281 |
* 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
|
1282 1283 1284 1285 1286 |
* added to the mmlist just after page_duplicate - before would be racy. */ static void drain_mmlist(void) { struct list_head *p, *next; |
efa90a981
|
1287 |
unsigned int type; |
1da177e4c
|
1288 |
|
efa90a981
|
1289 1290 |
for (type = 0; type < nr_swapfiles; type++) if (swap_info[type]->inuse_pages) |
1da177e4c
|
1291 1292 1293 1294 1295 1296 1297 1298 1299 |
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
|
1300 1301 1302 |
* 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
|
1303 |
*/ |
d4906e1aa
|
1304 |
static sector_t map_swap_entry(swp_entry_t entry, struct block_device **bdev) |
1da177e4c
|
1305 |
{ |
f29ad6a99
|
1306 1307 1308 1309 |
struct swap_info_struct *sis; struct swap_extent *start_se; struct swap_extent *se; pgoff_t offset; |
efa90a981
|
1310 |
sis = swap_info[swp_type(entry)]; |
f29ad6a99
|
1311 1312 1313 1314 1315 |
*bdev = sis->bdev; offset = swp_offset(entry); start_se = sis->curr_swap_extent; se = start_se; |
1da177e4c
|
1316 1317 1318 1319 1320 1321 1322 1323 |
for ( ; ; ) { struct list_head *lh; if (se->start_page <= offset && offset < (se->start_page + se->nr_pages)) { return se->start_block + (offset - se->start_page); } |
11d31886d
|
1324 |
lh = se->list.next; |
1da177e4c
|
1325 1326 1327 1328 1329 1330 1331 |
se = list_entry(lh, struct swap_extent, list); sis->curr_swap_extent = se; BUG_ON(se == start_se); /* It *must* be present */ } } /* |
d4906e1aa
|
1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 |
* 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
|
1342 1343 1344 1345 |
* Free all of a swapdev's extent information */ static void destroy_swap_extents(struct swap_info_struct *sis) { |
9625a5f28
|
1346 |
while (!list_empty(&sis->first_swap_extent.list)) { |
1da177e4c
|
1347 |
struct swap_extent *se; |
9625a5f28
|
1348 |
se = list_entry(sis->first_swap_extent.list.next, |
1da177e4c
|
1349 1350 1351 1352 |
struct swap_extent, list); list_del(&se->list); kfree(se); } |
1da177e4c
|
1353 1354 1355 1356 |
} /* * Add a block range (and the corresponding page range) into this swapdev's |
11d31886d
|
1357 |
* extent list. The extent list is kept sorted in page order. |
1da177e4c
|
1358 |
* |
11d31886d
|
1359 |
* This function rather assumes that it is called in ascending page order. |
1da177e4c
|
1360 1361 1362 1363 1364 1365 1366 1367 |
*/ static int 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
|
1368 1369 1370 1371 1372 1373 1374 1375 1376 |
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
|
1377 |
se = list_entry(lh, struct swap_extent, list); |
11d31886d
|
1378 1379 |
BUG_ON(se->start_page + se->nr_pages != start_page); if (se->start_block + se->nr_pages == start_block) { |
1da177e4c
|
1380 1381 1382 1383 |
/* Merge it */ se->nr_pages += nr_pages; return 0; } |
1da177e4c
|
1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 |
} /* * 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
|
1395 |
list_add_tail(&new_se->list, &sis->first_swap_extent.list); |
53092a740
|
1396 |
return 1; |
1da177e4c
|
1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 |
} /* * 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
|
1419 |
* For S_ISREG swapfiles we set S_SWAPFILE across the life of the swapon. This |
1da177e4c
|
1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 |
* 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
|
1430 |
static int setup_swap_extents(struct swap_info_struct *sis, sector_t *span) |
1da177e4c
|
1431 1432 1433 1434 1435 1436 1437 |
{ struct inode *inode; unsigned blocks_per_page; unsigned long page_no; unsigned blkbits; sector_t probe_block; sector_t last_block; |
53092a740
|
1438 1439 1440 |
sector_t lowest_block = -1; sector_t highest_block = 0; int nr_extents = 0; |
1da177e4c
|
1441 1442 1443 1444 1445 |
int ret; inode = sis->swap_file->f_mapping->host; if (S_ISBLK(inode->i_mode)) { ret = add_swap_extent(sis, 0, sis->max, 0); |
53092a740
|
1446 |
*span = sis->pages; |
9625a5f28
|
1447 |
goto out; |
1da177e4c
|
1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 |
} blkbits = inode->i_blkbits; blocks_per_page = PAGE_SIZE >> blkbits; /* * Map all the blocks into the extent list. This code doesn't try * to be very smart. */ probe_block = 0; page_no = 0; last_block = i_size_read(inode) >> blkbits; while ((probe_block + blocks_per_page) <= last_block && page_no < sis->max) { unsigned block_in_page; sector_t first_block; first_block = bmap(inode, probe_block); if (first_block == 0) goto bad_bmap; /* * It must be PAGE_SIZE aligned on-disk */ if (first_block & (blocks_per_page - 1)) { probe_block++; goto reprobe; } for (block_in_page = 1; block_in_page < blocks_per_page; block_in_page++) { sector_t block; block = bmap(inode, probe_block + block_in_page); if (block == 0) goto bad_bmap; if (block != first_block + block_in_page) { /* Discontiguity */ probe_block++; goto reprobe; } } |
53092a740
|
1490 1491 1492 1493 1494 1495 1496 |
first_block >>= (PAGE_SHIFT - blkbits); if (page_no) { /* exclude the header page */ if (first_block < lowest_block) lowest_block = first_block; if (first_block > highest_block) highest_block = first_block; } |
1da177e4c
|
1497 1498 1499 |
/* * We found a PAGE_SIZE-length, PAGE_SIZE-aligned run of blocks */ |
53092a740
|
1500 1501 |
ret = add_swap_extent(sis, page_no, 1, first_block); if (ret < 0) |
1da177e4c
|
1502 |
goto out; |
53092a740
|
1503 |
nr_extents += ret; |
1da177e4c
|
1504 1505 1506 1507 1508 |
page_no++; probe_block += blocks_per_page; reprobe: continue; } |
53092a740
|
1509 1510 |
ret = nr_extents; *span = 1 + highest_block - lowest_block; |
1da177e4c
|
1511 |
if (page_no == 0) |
e2244ec2e
|
1512 |
page_no = 1; /* force Empty message */ |
1da177e4c
|
1513 |
sis->max = page_no; |
e2244ec2e
|
1514 |
sis->pages = page_no - 1; |
1da177e4c
|
1515 |
sis->highest_bit = page_no - 1; |
9625a5f28
|
1516 1517 |
out: return ret; |
1da177e4c
|
1518 1519 1520 1521 |
bad_bmap: printk(KERN_ERR "swapon: swapfile has holes "); ret = -EINVAL; |
9625a5f28
|
1522 |
goto out; |
1da177e4c
|
1523 |
} |
c4ea37c26
|
1524 |
SYSCALL_DEFINE1(swapoff, const char __user *, specialfile) |
1da177e4c
|
1525 |
{ |
73c34b6ac
|
1526 |
struct swap_info_struct *p = NULL; |
8d69aaee8
|
1527 |
unsigned char *swap_map; |
1da177e4c
|
1528 1529 1530 |
struct file *swap_file, *victim; struct address_space *mapping; struct inode *inode; |
73c34b6ac
|
1531 |
char *pathname; |
1da177e4c
|
1532 1533 |
int i, type, prev; int err; |
886bb7e9c
|
1534 |
|
1da177e4c
|
1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 |
if (!capable(CAP_SYS_ADMIN)) return -EPERM; pathname = getname(specialfile); err = PTR_ERR(pathname); if (IS_ERR(pathname)) goto out; victim = filp_open(pathname, O_RDWR|O_LARGEFILE, 0); putname(pathname); err = PTR_ERR(victim); if (IS_ERR(victim)) goto out; mapping = victim->f_mapping; prev = -1; |
5d337b919
|
1551 |
spin_lock(&swap_lock); |
efa90a981
|
1552 1553 |
for (type = swap_list.head; type >= 0; type = swap_info[type]->next) { p = swap_info[type]; |
22c6f8fdb
|
1554 |
if (p->flags & SWP_WRITEOK) { |
1da177e4c
|
1555 1556 1557 1558 1559 1560 1561 |
if (p->swap_file->f_mapping == mapping) break; } prev = type; } if (type < 0) { err = -EINVAL; |
5d337b919
|
1562 |
spin_unlock(&swap_lock); |
1da177e4c
|
1563 1564 1565 1566 1567 1568 |
goto out_dput; } if (!security_vm_enough_memory(p->pages)) vm_unacct_memory(p->pages); else { err = -ENOMEM; |
5d337b919
|
1569 |
spin_unlock(&swap_lock); |
1da177e4c
|
1570 1571 |
goto out_dput; } |
efa90a981
|
1572 |
if (prev < 0) |
1da177e4c
|
1573 |
swap_list.head = p->next; |
efa90a981
|
1574 1575 |
else swap_info[prev]->next = p->next; |
1da177e4c
|
1576 1577 1578 1579 |
if (type == swap_list.next) { /* just pick something that's safe... */ swap_list.next = swap_list.head; } |
78ecba081
|
1580 |
if (p->prio < 0) { |
efa90a981
|
1581 1582 |
for (i = p->next; i >= 0; i = swap_info[i]->next) swap_info[i]->prio = p->prio--; |
78ecba081
|
1583 1584 |
least_priority++; } |
1da177e4c
|
1585 1586 1587 |
nr_swap_pages -= p->pages; total_swap_pages -= p->pages; p->flags &= ~SWP_WRITEOK; |
5d337b919
|
1588 |
spin_unlock(&swap_lock); |
fb4f88dca
|
1589 |
|
35451beec
|
1590 |
current->flags |= PF_OOM_ORIGIN; |
1da177e4c
|
1591 |
err = try_to_unuse(type); |
35451beec
|
1592 |
current->flags &= ~PF_OOM_ORIGIN; |
1da177e4c
|
1593 |
|
1da177e4c
|
1594 1595 |
if (err) { /* re-insert swap space back into swap_list */ |
5d337b919
|
1596 |
spin_lock(&swap_lock); |
78ecba081
|
1597 1598 1599 |
if (p->prio < 0) p->prio = --least_priority; prev = -1; |
efa90a981
|
1600 1601 |
for (i = swap_list.head; i >= 0; i = swap_info[i]->next) { if (p->prio >= swap_info[i]->prio) |
1da177e4c
|
1602 |
break; |
78ecba081
|
1603 1604 |
prev = i; } |
1da177e4c
|
1605 1606 |
p->next = i; if (prev < 0) |
efa90a981
|
1607 |
swap_list.head = swap_list.next = type; |
1da177e4c
|
1608 |
else |
efa90a981
|
1609 |
swap_info[prev]->next = type; |
1da177e4c
|
1610 1611 1612 |
nr_swap_pages += p->pages; total_swap_pages += p->pages; p->flags |= SWP_WRITEOK; |
5d337b919
|
1613 |
spin_unlock(&swap_lock); |
1da177e4c
|
1614 1615 |
goto out_dput; } |
52b7efdbe
|
1616 1617 1618 1619 |
/* wait for any unplug function to finish */ down_write(&swap_unplug_sem); up_write(&swap_unplug_sem); |
5d337b919
|
1620 |
destroy_swap_extents(p); |
570a335b8
|
1621 1622 |
if (p->flags & SWP_CONTINUED) free_swap_count_continuations(p); |
fc0abb145
|
1623 |
mutex_lock(&swapon_mutex); |
5d337b919
|
1624 1625 |
spin_lock(&swap_lock); drain_mmlist(); |
52b7efdbe
|
1626 |
/* wait for anyone still in scan_swap_map */ |
52b7efdbe
|
1627 1628 |
p->highest_bit = 0; /* cuts scans short */ while (p->flags >= SWP_SCANNING) { |
5d337b919
|
1629 |
spin_unlock(&swap_lock); |
13e4b57f6
|
1630 |
schedule_timeout_uninterruptible(1); |
5d337b919
|
1631 |
spin_lock(&swap_lock); |
52b7efdbe
|
1632 |
} |
52b7efdbe
|
1633 |
|
1da177e4c
|
1634 1635 1636 1637 1638 1639 |
swap_file = p->swap_file; p->swap_file = NULL; p->max = 0; swap_map = p->swap_map; p->swap_map = NULL; p->flags = 0; |
5d337b919
|
1640 |
spin_unlock(&swap_lock); |
fc0abb145
|
1641 |
mutex_unlock(&swapon_mutex); |
1da177e4c
|
1642 |
vfree(swap_map); |
27a7faa07
|
1643 1644 |
/* Destroy swap account informatin */ swap_cgroup_swapoff(type); |
1da177e4c
|
1645 1646 1647 1648 |
inode = mapping->host; if (S_ISBLK(inode->i_mode)) { struct block_device *bdev = I_BDEV(inode); set_blocksize(bdev, p->old_block_size); |
e525fd89d
|
1649 |
blkdev_put(bdev, FMODE_READ | FMODE_WRITE | FMODE_EXCL); |
1da177e4c
|
1650 |
} else { |
1b1dcc1b5
|
1651 |
mutex_lock(&inode->i_mutex); |
1da177e4c
|
1652 |
inode->i_flags &= ~S_SWAPFILE; |
1b1dcc1b5
|
1653 |
mutex_unlock(&inode->i_mutex); |
1da177e4c
|
1654 1655 1656 |
} filp_close(swap_file, NULL); err = 0; |
66d7dd518
|
1657 1658 |
atomic_inc(&proc_poll_event); wake_up_interruptible(&proc_poll_wait); |
1da177e4c
|
1659 1660 1661 1662 1663 1664 1665 1666 |
out_dput: filp_close(victim, NULL); out: return err; } #ifdef CONFIG_PROC_FS |
66d7dd518
|
1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 |
struct proc_swaps { struct seq_file seq; int event; }; static unsigned swaps_poll(struct file *file, poll_table *wait) { struct proc_swaps *s = file->private_data; poll_wait(file, &proc_poll_wait, wait); if (s->event != atomic_read(&proc_poll_event)) { s->event = atomic_read(&proc_poll_event); return POLLIN | POLLRDNORM | POLLERR | POLLPRI; } return POLLIN | POLLRDNORM; } |
1da177e4c
|
1685 1686 1687 |
/* iterator */ static void *swap_start(struct seq_file *swap, loff_t *pos) { |
efa90a981
|
1688 1689 |
struct swap_info_struct *si; int type; |
1da177e4c
|
1690 |
loff_t l = *pos; |
fc0abb145
|
1691 |
mutex_lock(&swapon_mutex); |
1da177e4c
|
1692 |
|
881e4aabe
|
1693 1694 |
if (!l) return SEQ_START_TOKEN; |
efa90a981
|
1695 1696 1697 1698 |
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
|
1699 |
continue; |
881e4aabe
|
1700 |
if (!--l) |
efa90a981
|
1701 |
return si; |
1da177e4c
|
1702 1703 1704 1705 1706 1707 1708 |
} return NULL; } static void *swap_next(struct seq_file *swap, void *v, loff_t *pos) { |
efa90a981
|
1709 1710 |
struct swap_info_struct *si = v; int type; |
1da177e4c
|
1711 |
|
881e4aabe
|
1712 |
if (v == SEQ_START_TOKEN) |
efa90a981
|
1713 1714 1715 |
type = 0; else type = si->type + 1; |
881e4aabe
|
1716 |
|
efa90a981
|
1717 1718 1719 1720 |
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
|
1721 1722 |
continue; ++*pos; |
efa90a981
|
1723 |
return si; |
1da177e4c
|
1724 1725 1726 1727 1728 1729 1730 |
} return NULL; } static void swap_stop(struct seq_file *swap, void *v) { |
fc0abb145
|
1731 |
mutex_unlock(&swapon_mutex); |
1da177e4c
|
1732 1733 1734 1735 |
} static int swap_show(struct seq_file *swap, void *v) { |
efa90a981
|
1736 |
struct swap_info_struct *si = v; |
1da177e4c
|
1737 1738 |
struct file *file; int len; |
efa90a981
|
1739 |
if (si == SEQ_START_TOKEN) { |
881e4aabe
|
1740 1741 1742 1743 |
seq_puts(swap,"Filename\t\t\t\tType\t\tSize\tUsed\tPriority "); return 0; } |
1da177e4c
|
1744 |
|
efa90a981
|
1745 |
file = si->swap_file; |
c32c2f63a
|
1746 1747 |
len = seq_path(swap, &file->f_path, " \t \\"); |
6eb396dc4
|
1748 1749 |
seq_printf(swap, "%*s%s\t%u\t%u\t%d ", |
886bb7e9c
|
1750 1751 |
len < 40 ? 40 - len : 1, " ", S_ISBLK(file->f_path.dentry->d_inode->i_mode) ? |
1da177e4c
|
1752 |
"partition" : "file\t", |
efa90a981
|
1753 1754 1755 |
si->pages << (PAGE_SHIFT - 10), si->inuse_pages << (PAGE_SHIFT - 10), si->prio); |
1da177e4c
|
1756 1757 |
return 0; } |
15ad7cdcf
|
1758 |
static const struct seq_operations swaps_op = { |
1da177e4c
|
1759 1760 1761 1762 1763 1764 1765 1766 |
.start = swap_start, .next = swap_next, .stop = swap_stop, .show = swap_show }; static int swaps_open(struct inode *inode, struct file *file) { |
66d7dd518
|
1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 |
struct proc_swaps *s; int ret; s = kmalloc(sizeof(struct proc_swaps), GFP_KERNEL); if (!s) return -ENOMEM; file->private_data = s; ret = seq_open(file, &swaps_op); if (ret) { kfree(s); return ret; } s->seq.private = s; s->event = atomic_read(&proc_poll_event); return ret; |
1da177e4c
|
1785 |
} |
15ad7cdcf
|
1786 |
static const struct file_operations proc_swaps_operations = { |
1da177e4c
|
1787 1788 1789 1790 |
.open = swaps_open, .read = seq_read, .llseek = seq_lseek, .release = seq_release, |
66d7dd518
|
1791 |
.poll = swaps_poll, |
1da177e4c
|
1792 1793 1794 1795 |
}; static int __init procswaps_init(void) { |
3d71f86f4
|
1796 |
proc_create("swaps", 0, NULL, &proc_swaps_operations); |
1da177e4c
|
1797 1798 1799 1800 |
return 0; } __initcall(procswaps_init); #endif /* CONFIG_PROC_FS */ |
1796316a8
|
1801 1802 1803 1804 1805 1806 1807 1808 |
#ifdef MAX_SWAPFILES_CHECK static int __init max_swapfiles_check(void) { MAX_SWAPFILES_CHECK(); return 0; } late_initcall(max_swapfiles_check); #endif |
1da177e4c
|
1809 1810 1811 1812 1813 |
/* * Written 01/25/92 by Simmule Turner, heavily changed by Linus. * * The swapon system call */ |
c4ea37c26
|
1814 |
SYSCALL_DEFINE2(swapon, const char __user *, specialfile, int, swap_flags) |
1da177e4c
|
1815 |
{ |
73c34b6ac
|
1816 |
struct swap_info_struct *p; |
1da177e4c
|
1817 1818 1819 1820 1821 1822 1823 |
char *name = NULL; struct block_device *bdev = NULL; struct file *swap_file = NULL; struct address_space *mapping; unsigned int type; int i, prev; int error; |
ad2bd7e0e
|
1824 1825 |
union swap_header *swap_header; unsigned int nr_good_pages; |
6eb396dc4
|
1826 |
int nr_extents = 0; |
53092a740
|
1827 |
sector_t span; |
ad2bd7e0e
|
1828 |
unsigned long maxpages; |
73fd8748a
|
1829 |
unsigned long swapfilepages; |
8d69aaee8
|
1830 |
unsigned char *swap_map = NULL; |
1da177e4c
|
1831 1832 1833 1834 1835 1836 |
struct page *page = NULL; struct inode *inode = NULL; int did_down = 0; if (!capable(CAP_SYS_ADMIN)) return -EPERM; |
efa90a981
|
1837 1838 1839 1840 |
p = kzalloc(sizeof(*p), GFP_KERNEL); if (!p) return -ENOMEM; |
5d337b919
|
1841 |
spin_lock(&swap_lock); |
efa90a981
|
1842 1843 |
for (type = 0; type < nr_swapfiles; type++) { if (!(swap_info[type]->flags & SWP_USED)) |
1da177e4c
|
1844 |
break; |
efa90a981
|
1845 |
} |
1da177e4c
|
1846 |
error = -EPERM; |
0697212a4
|
1847 |
if (type >= MAX_SWAPFILES) { |
5d337b919
|
1848 |
spin_unlock(&swap_lock); |
efa90a981
|
1849 |
kfree(p); |
1da177e4c
|
1850 1851 |
goto out; } |
efa90a981
|
1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 |
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
|
1870 |
INIT_LIST_HEAD(&p->first_swap_extent.list); |
1da177e4c
|
1871 |
p->flags = SWP_USED; |
1da177e4c
|
1872 |
p->next = -1; |
5d337b919
|
1873 |
spin_unlock(&swap_lock); |
efa90a981
|
1874 |
|
1da177e4c
|
1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 |
name = getname(specialfile); error = PTR_ERR(name); if (IS_ERR(name)) { name = NULL; goto bad_swap_2; } swap_file = filp_open(name, O_RDWR|O_LARGEFILE, 0); error = PTR_ERR(swap_file); if (IS_ERR(swap_file)) { swap_file = NULL; goto bad_swap_2; } p->swap_file = swap_file; mapping = swap_file->f_mapping; inode = mapping->host; error = -EBUSY; for (i = 0; i < nr_swapfiles; i++) { |
efa90a981
|
1894 |
struct swap_info_struct *q = swap_info[i]; |
1da177e4c
|
1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 |
if (i == type || !q->swap_file) continue; if (mapping == q->swap_file->f_mapping) goto bad_swap; } error = -EINVAL; if (S_ISBLK(inode->i_mode)) { bdev = I_BDEV(inode); |
e525fd89d
|
1905 1906 |
error = blkdev_get(bdev, FMODE_READ | FMODE_WRITE | FMODE_EXCL, sys_swapon); |
1da177e4c
|
1907 1908 |
if (error < 0) { bdev = NULL; |
f7b3a4359
|
1909 |
error = -EINVAL; |
1da177e4c
|
1910 1911 1912 1913 1914 1915 1916 |
goto bad_swap; } p->old_block_size = block_size(bdev); error = set_blocksize(bdev, PAGE_SIZE); if (error < 0) goto bad_swap; p->bdev = bdev; |
b27256439
|
1917 |
p->flags |= SWP_BLKDEV; |
1da177e4c
|
1918 1919 |
} else if (S_ISREG(inode->i_mode)) { p->bdev = inode->i_sb->s_bdev; |
1b1dcc1b5
|
1920 |
mutex_lock(&inode->i_mutex); |
1da177e4c
|
1921 1922 1923 1924 1925 1926 1927 1928 |
did_down = 1; if (IS_SWAPFILE(inode)) { error = -EBUSY; goto bad_swap; } } else { goto bad_swap; } |
73fd8748a
|
1929 |
swapfilepages = i_size_read(inode) >> PAGE_SHIFT; |
1da177e4c
|
1930 1931 1932 1933 1934 1935 1936 1937 |
/* * Read the swap header. */ if (!mapping->a_ops->readpage) { error = -EINVAL; goto bad_swap; } |
090d2b185
|
1938 |
page = read_mapping_page(mapping, 0, swap_file); |
1da177e4c
|
1939 1940 1941 1942 |
if (IS_ERR(page)) { error = PTR_ERR(page); goto bad_swap; } |
81e339712
|
1943 |
swap_header = kmap(page); |
1da177e4c
|
1944 |
|
81e339712
|
1945 |
if (memcmp("SWAPSPACE2", swap_header->magic.magic, 10)) { |
e97a31117
|
1946 1947 |
printk(KERN_ERR "Unable to find swap-space signature "); |
1da177e4c
|
1948 1949 1950 |
error = -EINVAL; goto bad_swap; } |
886bb7e9c
|
1951 |
|
81e339712
|
1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 |
/* 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) { printk(KERN_WARNING "Unable to handle swap header version %d ", swap_header->info.version); |
1da177e4c
|
1966 1967 |
error = -EINVAL; goto bad_swap; |
81e339712
|
1968 |
} |
1da177e4c
|
1969 |
|
81e339712
|
1970 1971 |
p->lowest_bit = 1; p->cluster_next = 1; |
efa90a981
|
1972 |
p->cluster_nr = 0; |
52b7efdbe
|
1973 |
|
81e339712
|
1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 |
/* * Find out how many pages are allowed for a single swap * device. There are two limiting factors: 1) the number of * bits for the swap offset in the swp_entry_t type and * 2) the number of bits in the a swap pte as defined by * the different architectures. In order to find the * largest possible bit mask a swap entry with swap type 0 * and swap offset ~0UL is created, encoded to a swap pte, * decoded to a swp_entry_t again and finally the swap * 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 * swap pte. */ maxpages = swp_offset(pte_to_swp_entry( |
ad2bd7e0e
|
1989 1990 1991 1992 1993 1994 1995 |
swp_entry_to_pte(swp_entry(0, ~0UL)))) + 1; if (maxpages > swap_header->info.last_page) { maxpages = swap_header->info.last_page + 1; /* p->max is an unsigned int: don't overflow it */ if ((unsigned int)maxpages == 0) maxpages = UINT_MAX; } |
81e339712
|
1996 |
p->highest_bit = maxpages - 1; |
1da177e4c
|
1997 |
|
81e339712
|
1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 |
error = -EINVAL; if (!maxpages) goto bad_swap; if (swapfilepages && maxpages > swapfilepages) { printk(KERN_WARNING "Swap area shorter than signature indicates "); goto bad_swap; } if (swap_header->info.nr_badpages && S_ISREG(inode->i_mode)) goto bad_swap; if (swap_header->info.nr_badpages > MAX_SWAP_BADPAGES) goto bad_swap; |
cd105df45
|
2011 |
|
81e339712
|
2012 |
/* OK, set up the swap map and apply the bad block list */ |
8d69aaee8
|
2013 |
swap_map = vmalloc(maxpages); |
81e339712
|
2014 2015 2016 2017 |
if (!swap_map) { error = -ENOMEM; goto bad_swap; } |
1da177e4c
|
2018 |
|
8d69aaee8
|
2019 |
memset(swap_map, 0, maxpages); |
ad2bd7e0e
|
2020 |
nr_good_pages = maxpages - 1; /* omit header page */ |
81e339712
|
2021 |
for (i = 0; i < swap_header->info.nr_badpages; i++) { |
ad2bd7e0e
|
2022 2023 |
unsigned int page_nr = swap_header->info.badpages[i]; if (page_nr == 0 || page_nr > swap_header->info.last_page) { |
81e339712
|
2024 |
error = -EINVAL; |
1da177e4c
|
2025 |
goto bad_swap; |
81e339712
|
2026 |
} |
ad2bd7e0e
|
2027 2028 2029 2030 |
if (page_nr < maxpages) { swap_map[page_nr] = SWAP_MAP_BAD; nr_good_pages--; } |
1da177e4c
|
2031 |
} |
27a7faa07
|
2032 2033 2034 2035 |
error = swap_cgroup_swapon(type, maxpages); if (error) goto bad_swap; |
e2244ec2e
|
2036 |
if (nr_good_pages) { |
78ecba081
|
2037 |
swap_map[0] = SWAP_MAP_BAD; |
e2244ec2e
|
2038 2039 |
p->max = maxpages; p->pages = nr_good_pages; |
53092a740
|
2040 2041 2042 |
nr_extents = setup_swap_extents(p, &span); if (nr_extents < 0) { error = nr_extents; |
e2244ec2e
|
2043 |
goto bad_swap; |
53092a740
|
2044 |
} |
e2244ec2e
|
2045 2046 |
nr_good_pages = p->pages; } |
1da177e4c
|
2047 2048 2049 2050 2051 2052 |
if (!nr_good_pages) { printk(KERN_WARNING "Empty swap-file "); error = -EINVAL; goto bad_swap; } |
1da177e4c
|
2053 |
|
3bd0f0c76
|
2054 2055 2056 2057 2058 |
if (p->bdev) { if (blk_queue_nonrot(bdev_get_queue(p->bdev))) { p->flags |= SWP_SOLIDSTATE; p->cluster_next = 1 + (random32() % p->highest_bit); } |
339944663
|
2059 |
if (discard_swap(p) == 0 && (swap_flags & SWAP_FLAG_DISCARD)) |
3bd0f0c76
|
2060 |
p->flags |= SWP_DISCARDABLE; |
20137a490
|
2061 |
} |
6a6ba8317
|
2062 |
|
fc0abb145
|
2063 |
mutex_lock(&swapon_mutex); |
5d337b919
|
2064 |
spin_lock(&swap_lock); |
78ecba081
|
2065 2066 2067 2068 2069 2070 |
if (swap_flags & SWAP_FLAG_PREFER) p->prio = (swap_flags & SWAP_FLAG_PRIO_MASK) >> SWAP_FLAG_PRIO_SHIFT; else p->prio = --least_priority; p->swap_map = swap_map; |
22c6f8fdb
|
2071 |
p->flags |= SWP_WRITEOK; |
1da177e4c
|
2072 2073 |
nr_swap_pages += nr_good_pages; total_swap_pages += nr_good_pages; |
53092a740
|
2074 |
|
6eb396dc4
|
2075 |
printk(KERN_INFO "Adding %uk swap on %s. " |
20137a490
|
2076 2077 |
"Priority:%d extents:%d across:%lluk %s%s ", |
53092a740
|
2078 |
nr_good_pages<<(PAGE_SHIFT-10), name, p->prio, |
6a6ba8317
|
2079 |
nr_extents, (unsigned long long)span<<(PAGE_SHIFT-10), |
20137a490
|
2080 2081 |
(p->flags & SWP_SOLIDSTATE) ? "SS" : "", (p->flags & SWP_DISCARDABLE) ? "D" : ""); |
1da177e4c
|
2082 2083 2084 |
/* insert swap space into swap_list: */ prev = -1; |
efa90a981
|
2085 2086 |
for (i = swap_list.head; i >= 0; i = swap_info[i]->next) { if (p->prio >= swap_info[i]->prio) |
1da177e4c
|
2087 |
break; |
1da177e4c
|
2088 2089 2090 |
prev = i; } p->next = i; |
efa90a981
|
2091 2092 2093 2094 |
if (prev < 0) swap_list.head = swap_list.next = type; else swap_info[prev]->next = type; |
5d337b919
|
2095 |
spin_unlock(&swap_lock); |
fc0abb145
|
2096 |
mutex_unlock(&swapon_mutex); |
66d7dd518
|
2097 2098 |
atomic_inc(&proc_poll_event); wake_up_interruptible(&proc_poll_wait); |
1da177e4c
|
2099 2100 2101 2102 2103 |
error = 0; goto out; bad_swap: if (bdev) { set_blocksize(bdev, p->old_block_size); |
e525fd89d
|
2104 |
blkdev_put(bdev, FMODE_READ | FMODE_WRITE | FMODE_EXCL); |
1da177e4c
|
2105 |
} |
4cd3bb10f
|
2106 |
destroy_swap_extents(p); |
27a7faa07
|
2107 |
swap_cgroup_swapoff(type); |
1da177e4c
|
2108 |
bad_swap_2: |
5d337b919
|
2109 |
spin_lock(&swap_lock); |
1da177e4c
|
2110 |
p->swap_file = NULL; |
1da177e4c
|
2111 |
p->flags = 0; |
5d337b919
|
2112 |
spin_unlock(&swap_lock); |
1da177e4c
|
2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 |
vfree(swap_map); if (swap_file) filp_close(swap_file, NULL); out: if (page && !IS_ERR(page)) { kunmap(page); page_cache_release(page); } if (name) putname(name); if (did_down) { if (!error) inode->i_flags |= S_SWAPFILE; |
1b1dcc1b5
|
2126 |
mutex_unlock(&inode->i_mutex); |
1da177e4c
|
2127 2128 2129 2130 2131 2132 |
} return error; } void si_swapinfo(struct sysinfo *val) { |
efa90a981
|
2133 |
unsigned int type; |
1da177e4c
|
2134 |
unsigned long nr_to_be_unused = 0; |
5d337b919
|
2135 |
spin_lock(&swap_lock); |
efa90a981
|
2136 2137 2138 2139 2140 |
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
|
2141 2142 2143 |
} val->freeswap = nr_swap_pages + nr_to_be_unused; val->totalswap = total_swap_pages + nr_to_be_unused; |
5d337b919
|
2144 |
spin_unlock(&swap_lock); |
1da177e4c
|
2145 2146 2147 2148 2149 |
} /* * Verify that a swap entry is valid and increment its swap map count. * |
355cfa73d
|
2150 2151 2152 2153 2154 2155 |
* 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
|
2156 |
* - swap-mapped reference requested but needs continued swap count. -> ENOMEM |
1da177e4c
|
2157 |
*/ |
8d69aaee8
|
2158 |
static int __swap_duplicate(swp_entry_t entry, unsigned char usage) |
1da177e4c
|
2159 |
{ |
73c34b6ac
|
2160 |
struct swap_info_struct *p; |
1da177e4c
|
2161 |
unsigned long offset, type; |
8d69aaee8
|
2162 2163 |
unsigned char count; unsigned char has_cache; |
253d553ba
|
2164 |
int err = -EINVAL; |
1da177e4c
|
2165 |
|
a7420aa54
|
2166 |
if (non_swap_entry(entry)) |
253d553ba
|
2167 |
goto out; |
0697212a4
|
2168 |
|
1da177e4c
|
2169 2170 2171 |
type = swp_type(entry); if (type >= nr_swapfiles) goto bad_file; |
efa90a981
|
2172 |
p = swap_info[type]; |
1da177e4c
|
2173 |
offset = swp_offset(entry); |
5d337b919
|
2174 |
spin_lock(&swap_lock); |
355cfa73d
|
2175 2176 |
if (unlikely(offset >= p->max)) goto unlock_out; |
253d553ba
|
2177 2178 2179 2180 |
count = p->swap_map[offset]; has_cache = count & SWAP_HAS_CACHE; count &= ~SWAP_HAS_CACHE; err = 0; |
355cfa73d
|
2181 |
|
253d553ba
|
2182 |
if (usage == SWAP_HAS_CACHE) { |
355cfa73d
|
2183 2184 |
/* set SWAP_HAS_CACHE if there is no cache and entry is used */ |
253d553ba
|
2185 2186 2187 2188 2189 2190 |
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
|
2191 2192 |
} else if (count || has_cache) { |
253d553ba
|
2193 |
|
570a335b8
|
2194 2195 2196 |
if ((count & ~COUNT_CONTINUED) < SWAP_MAP_MAX) count += usage; else if ((count & ~COUNT_CONTINUED) > SWAP_MAP_MAX) |
253d553ba
|
2197 |
err = -EINVAL; |
570a335b8
|
2198 2199 2200 2201 |
else if (swap_count_continued(p, offset, count)) count = COUNT_CONTINUED; else err = -ENOMEM; |
355cfa73d
|
2202 |
} else |
253d553ba
|
2203 2204 2205 |
err = -ENOENT; /* unused swap entry */ p->swap_map[offset] = count | has_cache; |
355cfa73d
|
2206 |
unlock_out: |
5d337b919
|
2207 |
spin_unlock(&swap_lock); |
1da177e4c
|
2208 |
out: |
253d553ba
|
2209 |
return err; |
1da177e4c
|
2210 2211 2212 2213 2214 2215 |
bad_file: printk(KERN_ERR "swap_dup: %s%08lx ", Bad_file, entry.val); goto out; } |
253d553ba
|
2216 |
|
355cfa73d
|
2217 |
/* |
aaa468653
|
2218 2219 2220 2221 2222 2223 2224 2225 2226 |
* 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
|
2227 2228 2229 2230 2231 |
* 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
|
2232 |
*/ |
570a335b8
|
2233 |
int swap_duplicate(swp_entry_t entry) |
355cfa73d
|
2234 |
{ |
570a335b8
|
2235 2236 2237 2238 2239 |
int err = 0; while (!err && __swap_duplicate(entry, 1) == -ENOMEM) err = add_swap_count_continuation(entry, GFP_ATOMIC); return err; |
355cfa73d
|
2240 |
} |
1da177e4c
|
2241 |
|
cb4b86ba4
|
2242 |
/* |
355cfa73d
|
2243 2244 |
* @entry: swap entry for which we allocate swap cache. * |
73c34b6ac
|
2245 |
* Called when allocating swap cache for existing swap entry, |
355cfa73d
|
2246 2247 2248 |
* 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
|
2249 2250 2251 |
*/ int swapcache_prepare(swp_entry_t entry) { |
253d553ba
|
2252 |
return __swap_duplicate(entry, SWAP_HAS_CACHE); |
cb4b86ba4
|
2253 |
} |
1da177e4c
|
2254 |
/* |
5d337b919
|
2255 |
* swap_lock prevents swap_map being freed. Don't grab an extra |
1da177e4c
|
2256 2257 2258 2259 |
* reference on the swaphandle, it doesn't matter if it becomes unused. */ int valid_swaphandles(swp_entry_t entry, unsigned long *offset) { |
8952898b0
|
2260 |
struct swap_info_struct *si; |
3f9e7949f
|
2261 |
int our_page_cluster = page_cluster; |
8952898b0
|
2262 2263 2264 |
pgoff_t target, toff; pgoff_t base, end; int nr_pages = 0; |
1da177e4c
|
2265 |
|
3f9e7949f
|
2266 |
if (!our_page_cluster) /* no readahead */ |
1da177e4c
|
2267 |
return 0; |
8952898b0
|
2268 |
|
efa90a981
|
2269 |
si = swap_info[swp_type(entry)]; |
8952898b0
|
2270 2271 2272 2273 2274 |
target = swp_offset(entry); base = (target >> our_page_cluster) << our_page_cluster; end = base + (1 << our_page_cluster); if (!base) /* first page is swap header */ base++; |
1da177e4c
|
2275 |
|
5d337b919
|
2276 |
spin_lock(&swap_lock); |
8952898b0
|
2277 2278 2279 2280 2281 2282 2283 2284 |
if (end > si->max) /* don't go beyond end of map */ end = si->max; /* Count contiguous allocated slots above our target */ for (toff = target; ++toff < end; nr_pages++) { /* Don't read in free or bad pages */ if (!si->swap_map[toff]) break; |
355cfa73d
|
2285 |
if (swap_count(si->swap_map[toff]) == SWAP_MAP_BAD) |
1da177e4c
|
2286 |
break; |
8952898b0
|
2287 2288 2289 |
} /* Count contiguous allocated slots below our target */ for (toff = target; --toff >= base; nr_pages++) { |
1da177e4c
|
2290 |
/* Don't read in free or bad pages */ |
8952898b0
|
2291 |
if (!si->swap_map[toff]) |
1da177e4c
|
2292 |
break; |
355cfa73d
|
2293 |
if (swap_count(si->swap_map[toff]) == SWAP_MAP_BAD) |
1da177e4c
|
2294 |
break; |
8952898b0
|
2295 |
} |
5d337b919
|
2296 |
spin_unlock(&swap_lock); |
8952898b0
|
2297 2298 2299 2300 2301 2302 2303 |
/* * Indicate starting offset, and return number of pages to get: * if only 1, say 0, since there's then no readahead to be done. */ *offset = ++toff; return nr_pages? ++nr_pages: 0; |
1da177e4c
|
2304 |
} |
570a335b8
|
2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520 |
/* * 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) { spin_unlock(&swap_lock); return -ENOMEM; } /* * We are fortunate that although vmalloc_to_page uses pte_offset_map, * no architecture is using highmem pages for kernel pagetables: so it * will not corrupt the GFP_ATOMIC caller's atomic pagetable kmaps. */ 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; map = kmap_atomic(list_page, KM_USER0) + offset; count = *map; kunmap_atomic(map, KM_USER0); /* * 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: spin_unlock(&swap_lock); 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); map = kmap_atomic(page, KM_USER0) + offset; 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)) { kunmap_atomic(map, KM_USER0); page = list_entry(page->lru.next, struct page, lru); BUG_ON(page == head); map = kmap_atomic(page, KM_USER0) + offset; } if (*map == SWAP_CONT_MAX) { kunmap_atomic(map, KM_USER0); page = list_entry(page->lru.next, struct page, lru); if (page == head) return false; /* add count continuation */ map = kmap_atomic(page, KM_USER0) + offset; init_map: *map = 0; /* we didn't zero the page */ } *map += 1; kunmap_atomic(map, KM_USER0); page = list_entry(page->lru.prev, struct page, lru); while (page != head) { map = kmap_atomic(page, KM_USER0) + offset; *map = COUNT_CONTINUED; kunmap_atomic(map, KM_USER0); 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) { kunmap_atomic(map, KM_USER0); page = list_entry(page->lru.next, struct page, lru); BUG_ON(page == head); map = kmap_atomic(page, KM_USER0) + offset; } BUG_ON(*map == 0); *map -= 1; if (*map == 0) count = 0; kunmap_atomic(map, KM_USER0); page = list_entry(page->lru.prev, struct page, lru); while (page != head) { map = kmap_atomic(page, KM_USER0) + offset; *map = SWAP_CONT_MAX | count; count = COUNT_CONTINUED; kunmap_atomic(map, KM_USER0); 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)) { struct list_head *this, *next; list_for_each_safe(this, next, &head->lru) { struct page *page; page = list_entry(this, struct page, lru); list_del(this); __free_page(page); } } } } |