Blame view
mm/swapfile.c
65.3 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> |
72788c385
|
33 |
#include <linux/oom.h> |
1da177e4c
|
34 35 36 37 |
#include <asm/pgtable.h> #include <asm/tlbflush.h> #include <linux/swapops.h> |
27a7faa07
|
38 |
#include <linux/page_cgroup.h> |
1da177e4c
|
39 |
|
570a335b8
|
40 41 42 |
static bool swap_count_continued(struct swap_info_struct *, pgoff_t, unsigned char); static void free_swap_count_continuations(struct swap_info_struct *); |
d4906e1aa
|
43 |
static sector_t map_swap_entry(swp_entry_t, struct block_device**); |
570a335b8
|
44 |
|
7c363b8c6
|
45 46 |
static DEFINE_SPINLOCK(swap_lock); static unsigned int nr_swapfiles; |
b962716b4
|
47 |
long nr_swap_pages; |
1da177e4c
|
48 |
long total_swap_pages; |
78ecba081
|
49 |
static int least_priority; |
1da177e4c
|
50 |
|
1da177e4c
|
51 52 53 54 |
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
|
55 |
static struct swap_list_t swap_list = {-1, -1}; |
1da177e4c
|
56 |
|
efa90a981
|
57 |
static struct swap_info_struct *swap_info[MAX_SWAPFILES]; |
1da177e4c
|
58 |
|
fc0abb145
|
59 |
static DEFINE_MUTEX(swapon_mutex); |
1da177e4c
|
60 |
|
66d7dd518
|
61 62 63 |
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
|
64 |
static inline unsigned char swap_count(unsigned char ent) |
355cfa73d
|
65 |
{ |
570a335b8
|
66 |
return ent & ~SWAP_HAS_CACHE; /* may include SWAP_HAS_CONT flag */ |
355cfa73d
|
67 |
} |
efa90a981
|
68 |
/* returns 1 if swap entry is freed */ |
c9e444103
|
69 70 71 |
static int __try_to_reclaim_swap(struct swap_info_struct *si, unsigned long offset) { |
efa90a981
|
72 |
swp_entry_t entry = swp_entry(si->type, offset); |
c9e444103
|
73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 |
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
|
93 |
|
1da177e4c
|
94 |
/* |
6a6ba8317
|
95 96 97 98 99 100 |
* 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
|
101 102 |
sector_t start_block; sector_t nr_blocks; |
6a6ba8317
|
103 |
int err = 0; |
9625a5f28
|
104 105 106 107 108 109 |
/* 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
|
110 |
nr_blocks, GFP_KERNEL, 0); |
9625a5f28
|
111 112 113 114 |
if (err) return err; cond_resched(); } |
6a6ba8317
|
115 |
|
9625a5f28
|
116 117 118 |
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
|
119 120 |
err = blkdev_issue_discard(si->bdev, start_block, |
dd3932edd
|
121 |
nr_blocks, GFP_KERNEL, 0); |
6a6ba8317
|
122 123 124 125 126 127 128 |
if (err) break; cond_resched(); } return err; /* That will often be -EOPNOTSUPP */ } |
7992fde72
|
129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 |
/* * 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
|
146 |
sector_t nr_blocks = se->nr_pages - offset; |
7992fde72
|
147 148 149 150 151 152 153 154 155 156 157 158 |
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
|
159 |
nr_blocks, GFP_NOIO, 0)) |
7992fde72
|
160 161 162 163 |
break; } lh = se->list.next; |
7992fde72
|
164 165 166 167 168 169 170 171 172 |
se = list_entry(lh, struct swap_extent, list); } } static int wait_for_discard(void *word) { schedule(); return 0; } |
048c27fd7
|
173 174 |
#define SWAPFILE_CLUSTER 256 #define LATENCY_LIMIT 256 |
24b8ff7c2
|
175 176 |
static unsigned long scan_swap_map(struct swap_info_struct *si, unsigned char usage) |
1da177e4c
|
177 |
{ |
ebebbbe90
|
178 |
unsigned long offset; |
c60aa176c
|
179 |
unsigned long scan_base; |
7992fde72
|
180 |
unsigned long last_in_cluster = 0; |
048c27fd7
|
181 |
int latency_ration = LATENCY_LIMIT; |
7992fde72
|
182 |
int found_free_cluster = 0; |
7dfad4183
|
183 |
|
886bb7e9c
|
184 |
/* |
7dfad4183
|
185 186 187 188 189 190 191 |
* 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
|
192 |
* And we let swap pages go all over an SSD partition. Hugh |
7dfad4183
|
193 |
*/ |
52b7efdbe
|
194 |
si->flags += SWP_SCANNING; |
c60aa176c
|
195 |
scan_base = offset = si->cluster_next; |
ebebbbe90
|
196 197 198 199 200 201 |
if (unlikely(!si->cluster_nr--)) { if (si->pages - si->inuse_pages < SWAPFILE_CLUSTER) { si->cluster_nr = SWAPFILE_CLUSTER - 1; goto checks; } |
7992fde72
|
202 203 204 205 206 207 208 209 210 211 212 213 214 |
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
|
215 |
spin_unlock(&swap_lock); |
7dfad4183
|
216 |
|
c60aa176c
|
217 218 219 220 221 222 223 224 225 226 |
/* * 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
|
227 228 229 230 |
last_in_cluster = offset + SWAPFILE_CLUSTER - 1; /* Locate the first empty (unaligned) cluster */ for (; last_in_cluster <= si->highest_bit; offset++) { |
1da177e4c
|
231 |
if (si->swap_map[offset]) |
7dfad4183
|
232 233 |
last_in_cluster = offset + SWAPFILE_CLUSTER; else if (offset == last_in_cluster) { |
5d337b919
|
234 |
spin_lock(&swap_lock); |
ebebbbe90
|
235 236 237 |
offset -= SWAPFILE_CLUSTER - 1; si->cluster_next = offset; si->cluster_nr = SWAPFILE_CLUSTER - 1; |
7992fde72
|
238 |
found_free_cluster = 1; |
ebebbbe90
|
239 |
goto checks; |
1da177e4c
|
240 |
} |
048c27fd7
|
241 242 243 244 |
if (unlikely(--latency_ration < 0)) { cond_resched(); latency_ration = LATENCY_LIMIT; } |
7dfad4183
|
245 |
} |
ebebbbe90
|
246 247 |
offset = si->lowest_bit; |
c60aa176c
|
248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 |
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
|
269 |
spin_lock(&swap_lock); |
ebebbbe90
|
270 |
si->cluster_nr = SWAPFILE_CLUSTER - 1; |
7992fde72
|
271 |
si->lowest_alloc = 0; |
1da177e4c
|
272 |
} |
7dfad4183
|
273 |
|
ebebbbe90
|
274 275 |
checks: if (!(si->flags & SWP_WRITEOK)) |
52b7efdbe
|
276 |
goto no_page; |
7dfad4183
|
277 278 |
if (!si->highest_bit) goto no_page; |
ebebbbe90
|
279 |
if (offset > si->highest_bit) |
c60aa176c
|
280 |
scan_base = offset = si->lowest_bit; |
c9e444103
|
281 |
|
b73d7fcec
|
282 283 |
/* reuse swap entry of cache-only swap if not busy. */ if (vm_swap_full() && si->swap_map[offset] == SWAP_HAS_CACHE) { |
c9e444103
|
284 285 286 287 288 289 290 291 292 |
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
|
293 294 295 296 297 298 299 300 301 302 303 |
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
|
304 |
} |
253d553ba
|
305 |
si->swap_map[offset] = usage; |
ebebbbe90
|
306 307 |
si->cluster_next = offset + 1; si->flags -= SWP_SCANNING; |
7992fde72
|
308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 |
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
|
362 |
return offset; |
7dfad4183
|
363 |
|
ebebbbe90
|
364 |
scan: |
5d337b919
|
365 |
spin_unlock(&swap_lock); |
7dfad4183
|
366 |
while (++offset <= si->highest_bit) { |
52b7efdbe
|
367 |
if (!si->swap_map[offset]) { |
5d337b919
|
368 |
spin_lock(&swap_lock); |
52b7efdbe
|
369 370 |
goto checks; } |
c9e444103
|
371 372 373 374 |
if (vm_swap_full() && si->swap_map[offset] == SWAP_HAS_CACHE) { spin_lock(&swap_lock); goto checks; } |
048c27fd7
|
375 376 377 378 |
if (unlikely(--latency_ration < 0)) { cond_resched(); latency_ration = LATENCY_LIMIT; } |
7dfad4183
|
379 |
} |
c60aa176c
|
380 381 382 383 384 385 |
offset = si->lowest_bit; while (++offset < scan_base) { if (!si->swap_map[offset]) { spin_lock(&swap_lock); goto checks; } |
c9e444103
|
386 387 388 389 |
if (vm_swap_full() && si->swap_map[offset] == SWAP_HAS_CACHE) { spin_lock(&swap_lock); goto checks; } |
c60aa176c
|
390 391 392 393 394 |
if (unlikely(--latency_ration < 0)) { cond_resched(); latency_ration = LATENCY_LIMIT; } } |
5d337b919
|
395 |
spin_lock(&swap_lock); |
7dfad4183
|
396 397 |
no_page: |
52b7efdbe
|
398 |
si->flags -= SWP_SCANNING; |
1da177e4c
|
399 400 401 402 403 |
return 0; } swp_entry_t get_swap_page(void) { |
fb4f88dca
|
404 405 406 407 |
struct swap_info_struct *si; pgoff_t offset; int type, next; int wrapped = 0; |
1da177e4c
|
408 |
|
5d337b919
|
409 |
spin_lock(&swap_lock); |
1da177e4c
|
410 |
if (nr_swap_pages <= 0) |
fb4f88dca
|
411 412 413 414 |
goto noswap; nr_swap_pages--; for (type = swap_list.next; type >= 0 && wrapped < 2; type = next) { |
efa90a981
|
415 |
si = swap_info[type]; |
fb4f88dca
|
416 417 |
next = si->next; if (next < 0 || |
efa90a981
|
418 |
(!wrapped && si->prio != swap_info[next]->prio)) { |
fb4f88dca
|
419 420 |
next = swap_list.head; wrapped++; |
1da177e4c
|
421 |
} |
fb4f88dca
|
422 423 424 425 426 427 428 |
if (!si->highest_bit) continue; if (!(si->flags & SWP_WRITEOK)) continue; swap_list.next = next; |
355cfa73d
|
429 |
/* This is called for allocating swap entry for cache */ |
253d553ba
|
430 |
offset = scan_swap_map(si, SWAP_HAS_CACHE); |
5d337b919
|
431 432 |
if (offset) { spin_unlock(&swap_lock); |
fb4f88dca
|
433 |
return swp_entry(type, offset); |
5d337b919
|
434 |
} |
fb4f88dca
|
435 |
next = swap_list.next; |
1da177e4c
|
436 |
} |
fb4f88dca
|
437 438 439 |
nr_swap_pages++; noswap: |
5d337b919
|
440 |
spin_unlock(&swap_lock); |
fb4f88dca
|
441 |
return (swp_entry_t) {0}; |
1da177e4c
|
442 |
} |
910321ea8
|
443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 |
/* 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
|
464 |
static struct swap_info_struct *swap_info_get(swp_entry_t entry) |
1da177e4c
|
465 |
{ |
73c34b6ac
|
466 |
struct swap_info_struct *p; |
1da177e4c
|
467 468 469 470 471 472 473 |
unsigned long offset, type; if (!entry.val) goto out; type = swp_type(entry); if (type >= nr_swapfiles) goto bad_nofile; |
efa90a981
|
474 |
p = swap_info[type]; |
1da177e4c
|
475 476 477 478 479 480 481 |
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
|
482 |
spin_lock(&swap_lock); |
1da177e4c
|
483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 |
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
|
502 |
} |
1da177e4c
|
503 |
|
8d69aaee8
|
504 505 |
static unsigned char swap_entry_free(struct swap_info_struct *p, swp_entry_t entry, unsigned char usage) |
1da177e4c
|
506 |
{ |
253d553ba
|
507 |
unsigned long offset = swp_offset(entry); |
8d69aaee8
|
508 509 |
unsigned char count; unsigned char has_cache; |
355cfa73d
|
510 |
|
253d553ba
|
511 512 513 |
count = p->swap_map[offset]; has_cache = count & SWAP_HAS_CACHE; count &= ~SWAP_HAS_CACHE; |
355cfa73d
|
514 |
|
253d553ba
|
515 |
if (usage == SWAP_HAS_CACHE) { |
355cfa73d
|
516 |
VM_BUG_ON(!has_cache); |
253d553ba
|
517 |
has_cache = 0; |
aaa468653
|
518 519 520 521 522 523 |
} 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
|
524 525 526 527 528 529 530 531 532 |
} 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
|
533 534 535 536 537 538 |
if (!count) mem_cgroup_uncharge_swap(entry); usage = count | has_cache; p->swap_map[offset] = usage; |
355cfa73d
|
539 |
|
355cfa73d
|
540 |
/* free if no reference */ |
253d553ba
|
541 |
if (!usage) { |
b3a27d052
|
542 |
struct gendisk *disk = p->bdev->bd_disk; |
355cfa73d
|
543 544 545 546 |
if (offset < p->lowest_bit) p->lowest_bit = offset; if (offset > p->highest_bit) p->highest_bit = offset; |
efa90a981
|
547 548 549 |
if (swap_list.next >= 0 && p->prio > swap_info[swap_list.next]->prio) swap_list.next = p->type; |
355cfa73d
|
550 551 |
nr_swap_pages++; p->inuse_pages--; |
b3a27d052
|
552 553 554 |
if ((p->flags & SWP_BLKDEV) && disk->fops->swap_slot_free_notify) disk->fops->swap_slot_free_notify(p->bdev, offset); |
1da177e4c
|
555 |
} |
253d553ba
|
556 557 |
return usage; |
1da177e4c
|
558 559 560 561 562 563 564 565 |
} /* * 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
|
566 |
struct swap_info_struct *p; |
1da177e4c
|
567 568 569 |
p = swap_info_get(entry); if (p) { |
253d553ba
|
570 |
swap_entry_free(p, entry, 1); |
5d337b919
|
571 |
spin_unlock(&swap_lock); |
1da177e4c
|
572 573 574 575 |
} } /* |
cb4b86ba4
|
576 577 578 579 |
* Called after dropping swapcache to decrease refcnt to swap entries. */ void swapcache_free(swp_entry_t entry, struct page *page) { |
355cfa73d
|
580 |
struct swap_info_struct *p; |
8d69aaee8
|
581 |
unsigned char count; |
355cfa73d
|
582 |
|
355cfa73d
|
583 584 |
p = swap_info_get(entry); if (p) { |
253d553ba
|
585 586 587 |
count = swap_entry_free(p, entry, SWAP_HAS_CACHE); if (page) mem_cgroup_uncharge_swapcache(page, entry, count != 0); |
355cfa73d
|
588 589 |
spin_unlock(&swap_lock); } |
cb4b86ba4
|
590 591 592 |
} /* |
c475a8ab6
|
593 |
* How many references to page are currently swapped out? |
570a335b8
|
594 595 |
* 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
|
596 |
*/ |
c475a8ab6
|
597 |
static inline int page_swapcount(struct page *page) |
1da177e4c
|
598 |
{ |
c475a8ab6
|
599 600 |
int count = 0; struct swap_info_struct *p; |
1da177e4c
|
601 |
swp_entry_t entry; |
4c21e2f24
|
602 |
entry.val = page_private(page); |
1da177e4c
|
603 604 |
p = swap_info_get(entry); if (p) { |
355cfa73d
|
605 |
count = swap_count(p->swap_map[swp_offset(entry)]); |
5d337b919
|
606 |
spin_unlock(&swap_lock); |
1da177e4c
|
607 |
} |
c475a8ab6
|
608 |
return count; |
1da177e4c
|
609 610 611 |
} /* |
7b1fe5979
|
612 613 614 615 |
* 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
|
616 |
*/ |
7b1fe5979
|
617 |
int reuse_swap_page(struct page *page) |
1da177e4c
|
618 |
{ |
c475a8ab6
|
619 |
int count; |
51726b122
|
620 |
VM_BUG_ON(!PageLocked(page)); |
5ad646880
|
621 622 |
if (unlikely(PageKsm(page))) return 0; |
c475a8ab6
|
623 |
count = page_mapcount(page); |
7b1fe5979
|
624 |
if (count <= 1 && PageSwapCache(page)) { |
c475a8ab6
|
625 |
count += page_swapcount(page); |
7b1fe5979
|
626 627 628 629 630 |
if (count == 1 && !PageWriteback(page)) { delete_from_swap_cache(page); SetPageDirty(page); } } |
5ad646880
|
631 |
return count <= 1; |
1da177e4c
|
632 633 634 |
} /* |
a2c43eed8
|
635 636 |
* 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
|
637 |
*/ |
a2c43eed8
|
638 |
int try_to_free_swap(struct page *page) |
1da177e4c
|
639 |
{ |
51726b122
|
640 |
VM_BUG_ON(!PageLocked(page)); |
1da177e4c
|
641 642 643 644 645 |
if (!PageSwapCache(page)) return 0; if (PageWriteback(page)) return 0; |
a2c43eed8
|
646 |
if (page_swapcount(page)) |
1da177e4c
|
647 |
return 0; |
b73d7fcec
|
648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 |
/* * 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
|
665 666 667 |
delete_from_swap_cache(page); SetPageDirty(page); return 1; |
68a22394c
|
668 669 670 |
} /* |
1da177e4c
|
671 672 673 |
* Free the swap entry like above, but also try to * free the page cache entry if it is the last user. */ |
2509ef26d
|
674 |
int free_swap_and_cache(swp_entry_t entry) |
1da177e4c
|
675 |
{ |
2509ef26d
|
676 |
struct swap_info_struct *p; |
1da177e4c
|
677 |
struct page *page = NULL; |
a7420aa54
|
678 |
if (non_swap_entry(entry)) |
2509ef26d
|
679 |
return 1; |
0697212a4
|
680 |
|
1da177e4c
|
681 682 |
p = swap_info_get(entry); if (p) { |
253d553ba
|
683 |
if (swap_entry_free(p, entry, 1) == SWAP_HAS_CACHE) { |
93fac7041
|
684 |
page = find_get_page(&swapper_space, entry.val); |
8413ac9d8
|
685 |
if (page && !trylock_page(page)) { |
93fac7041
|
686 687 688 689 |
page_cache_release(page); page = NULL; } } |
5d337b919
|
690 |
spin_unlock(&swap_lock); |
1da177e4c
|
691 692 |
} if (page) { |
a2c43eed8
|
693 694 695 696 |
/* * Not mapped elsewhere, or swap space full? Free it! * Also recheck PageSwapCache now page is locked (above). */ |
93fac7041
|
697 |
if (PageSwapCache(page) && !PageWriteback(page) && |
a2c43eed8
|
698 |
(!page_mapped(page) || vm_swap_full())) { |
1da177e4c
|
699 700 701 702 703 704 |
delete_from_swap_cache(page); SetPageDirty(page); } unlock_page(page); page_cache_release(page); } |
2509ef26d
|
705 |
return p != NULL; |
1da177e4c
|
706 |
} |
024914477
|
707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 |
#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
|
737 |
#ifdef CONFIG_HIBERNATION |
f577eb30a
|
738 |
/* |
915bae9eb
|
739 |
* Find the swap type that corresponds to given device (if any). |
f577eb30a
|
740 |
* |
915bae9eb
|
741 742 743 744 |
* @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
|
745 |
*/ |
7bf236874
|
746 |
int swap_type_of(dev_t device, sector_t offset, struct block_device **bdev_p) |
f577eb30a
|
747 |
{ |
915bae9eb
|
748 |
struct block_device *bdev = NULL; |
efa90a981
|
749 |
int type; |
f577eb30a
|
750 |
|
915bae9eb
|
751 752 |
if (device) bdev = bdget(device); |
f577eb30a
|
753 |
spin_lock(&swap_lock); |
efa90a981
|
754 755 |
for (type = 0; type < nr_swapfiles; type++) { struct swap_info_struct *sis = swap_info[type]; |
f577eb30a
|
756 |
|
915bae9eb
|
757 |
if (!(sis->flags & SWP_WRITEOK)) |
f577eb30a
|
758 |
continue; |
b6b5bce35
|
759 |
|
915bae9eb
|
760 |
if (!bdev) { |
7bf236874
|
761 |
if (bdev_p) |
dddac6a7b
|
762 |
*bdev_p = bdgrab(sis->bdev); |
7bf236874
|
763 |
|
6e1819d61
|
764 |
spin_unlock(&swap_lock); |
efa90a981
|
765 |
return type; |
6e1819d61
|
766 |
} |
915bae9eb
|
767 |
if (bdev == sis->bdev) { |
9625a5f28
|
768 |
struct swap_extent *se = &sis->first_swap_extent; |
915bae9eb
|
769 |
|
915bae9eb
|
770 |
if (se->start_block == offset) { |
7bf236874
|
771 |
if (bdev_p) |
dddac6a7b
|
772 |
*bdev_p = bdgrab(sis->bdev); |
7bf236874
|
773 |
|
915bae9eb
|
774 775 |
spin_unlock(&swap_lock); bdput(bdev); |
efa90a981
|
776 |
return type; |
915bae9eb
|
777 |
} |
f577eb30a
|
778 779 780 |
} } spin_unlock(&swap_lock); |
915bae9eb
|
781 782 |
if (bdev) bdput(bdev); |
f577eb30a
|
783 784 785 786 |
return -ENODEV; } /* |
73c34b6ac
|
787 788 789 790 791 792 793 794 795 796 797 |
* 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
|
798 |
return map_swap_entry(swp_entry(type, offset), &bdev); |
73c34b6ac
|
799 800 801 |
} /* |
f577eb30a
|
802 803 804 805 806 807 808 809 |
* 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
|
810 811 812 813 814 815 |
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
|
816 |
if (free) |
efa90a981
|
817 |
n -= sis->inuse_pages; |
f577eb30a
|
818 |
} |
f577eb30a
|
819 |
} |
efa90a981
|
820 |
spin_unlock(&swap_lock); |
f577eb30a
|
821 822 |
return n; } |
73c34b6ac
|
823 |
#endif /* CONFIG_HIBERNATION */ |
f577eb30a
|
824 |
|
1da177e4c
|
825 |
/* |
72866f6f2
|
826 827 828 |
* 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
|
829 |
*/ |
044d66c1d
|
830 |
static int unuse_pte(struct vm_area_struct *vma, pmd_t *pmd, |
1da177e4c
|
831 832 |
unsigned long addr, swp_entry_t entry, struct page *page) { |
56039efa1
|
833 |
struct mem_cgroup *ptr; |
044d66c1d
|
834 835 836 |
spinlock_t *ptl; pte_t *pte; int ret = 1; |
85d9fc89f
|
837 |
if (mem_cgroup_try_charge_swapin(vma->vm_mm, page, GFP_KERNEL, &ptr)) { |
044d66c1d
|
838 |
ret = -ENOMEM; |
85d9fc89f
|
839 840 |
goto out_nolock; } |
044d66c1d
|
841 842 843 844 |
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
|
845 |
mem_cgroup_cancel_charge_swapin(ptr); |
044d66c1d
|
846 847 848 |
ret = 0; goto out; } |
8a9f3ccd2
|
849 |
|
b084d4353
|
850 |
dec_mm_counter(vma->vm_mm, MM_SWAPENTS); |
d559db086
|
851 |
inc_mm_counter(vma->vm_mm, MM_ANONPAGES); |
1da177e4c
|
852 853 854 855 |
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
|
856 |
mem_cgroup_commit_charge_swapin(page, ptr); |
1da177e4c
|
857 858 859 860 861 862 |
swap_free(entry); /* * Move the page to the active list so it is not * immediately swapped out again after swapon. */ activate_page(page); |
044d66c1d
|
863 864 |
out: pte_unmap_unlock(pte, ptl); |
85d9fc89f
|
865 |
out_nolock: |
044d66c1d
|
866 |
return ret; |
1da177e4c
|
867 868 869 870 871 872 |
} 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
|
873 |
pte_t swp_pte = swp_entry_to_pte(entry); |
705e87c0c
|
874 |
pte_t *pte; |
8a9f3ccd2
|
875 |
int ret = 0; |
1da177e4c
|
876 |
|
044d66c1d
|
877 878 879 880 881 882 883 884 885 886 |
/* * 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
|
887 888 889 890 891 892 |
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
|
893 894 895 896 897 |
pte_unmap(pte); ret = unuse_pte(vma, pmd, addr, entry, page); if (ret) goto out; pte = pte_offset_map(pmd, addr); |
1da177e4c
|
898 899 |
} } while (pte++, addr += PAGE_SIZE, addr != end); |
044d66c1d
|
900 901 |
pte_unmap(pte - 1); out: |
8a9f3ccd2
|
902 |
return ret; |
1da177e4c
|
903 904 905 906 907 908 909 910 |
} 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
|
911 |
int ret; |
1da177e4c
|
912 913 914 915 |
pmd = pmd_offset(pud, addr); do { next = pmd_addr_end(addr, end); |
3f04f62f9
|
916 917 |
if (unlikely(pmd_trans_huge(*pmd))) continue; |
1da177e4c
|
918 919 |
if (pmd_none_or_clear_bad(pmd)) continue; |
8a9f3ccd2
|
920 921 922 |
ret = unuse_pte_range(vma, pmd, addr, next, entry, page); if (ret) return ret; |
1da177e4c
|
923 924 925 926 927 928 929 930 931 932 |
} 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
|
933 |
int ret; |
1da177e4c
|
934 935 936 937 938 939 |
pud = pud_offset(pgd, addr); do { next = pud_addr_end(addr, end); if (pud_none_or_clear_bad(pud)) continue; |
8a9f3ccd2
|
940 941 942 |
ret = unuse_pmd_range(vma, pud, addr, next, entry, page); if (ret) return ret; |
1da177e4c
|
943 944 945 946 947 948 949 950 951 |
} 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
|
952 |
int ret; |
1da177e4c
|
953 |
|
3ca7b3c5b
|
954 |
if (page_anon_vma(page)) { |
1da177e4c
|
955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 |
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
|
970 971 972 |
ret = unuse_pud_range(vma, pgd, addr, next, entry, page); if (ret) return ret; |
1da177e4c
|
973 974 975 976 977 978 979 980 |
} 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
|
981 |
int ret = 0; |
1da177e4c
|
982 983 984 |
if (!down_read_trylock(&mm->mmap_sem)) { /* |
7d03431cf
|
985 986 |
* Activate page so shrink_inactive_list is unlikely to unmap * its ptes while lock is dropped, so swapoff can make progress. |
1da177e4c
|
987 |
*/ |
c475a8ab6
|
988 |
activate_page(page); |
1da177e4c
|
989 990 991 992 |
unlock_page(page); down_read(&mm->mmap_sem); lock_page(page); } |
1da177e4c
|
993 |
for (vma = mm->mmap; vma; vma = vma->vm_next) { |
8a9f3ccd2
|
994 |
if (vma->anon_vma && (ret = unuse_vma(vma, entry, page))) |
1da177e4c
|
995 996 |
break; } |
1da177e4c
|
997 |
up_read(&mm->mmap_sem); |
8a9f3ccd2
|
998 |
return (ret < 0)? ret: 0; |
1da177e4c
|
999 1000 1001 1002 1003 1004 |
} /* * Scan swap_map from current position to next entry still in use. * Recycle to start on reaching the end, returning 0 when empty. */ |
6eb396dc4
|
1005 1006 |
static unsigned int find_next_to_unuse(struct swap_info_struct *si, unsigned int prev) |
1da177e4c
|
1007 |
{ |
6eb396dc4
|
1008 1009 |
unsigned int max = si->max; unsigned int i = prev; |
8d69aaee8
|
1010 |
unsigned char count; |
1da177e4c
|
1011 1012 |
/* |
5d337b919
|
1013 |
* No need for swap_lock here: we're just looking |
1da177e4c
|
1014 1015 |
* for whether an entry is in use, not modifying it; false * hits are okay, and sys_swapoff() has already prevented new |
5d337b919
|
1016 |
* allocations from this area (while holding swap_lock). |
1da177e4c
|
1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 |
*/ 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
|
1033 |
if (count && swap_count(count) != SWAP_MAP_BAD) |
1da177e4c
|
1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 |
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
|
1046 |
struct swap_info_struct *si = swap_info[type]; |
1da177e4c
|
1047 |
struct mm_struct *start_mm; |
8d69aaee8
|
1048 1049 |
unsigned char *swap_map; unsigned char swcount; |
1da177e4c
|
1050 1051 |
struct page *page; swp_entry_t entry; |
6eb396dc4
|
1052 |
unsigned int i = 0; |
1da177e4c
|
1053 |
int retval = 0; |
1da177e4c
|
1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 |
/* * 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
|
1067 |
* that. |
1da177e4c
|
1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 |
*/ 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
|
1082 |
/* |
1da177e4c
|
1083 1084 |
* Get a page for the entry, using the existing swap * cache page if there is one. Otherwise, get a clean |
886bb7e9c
|
1085 |
* page and read the swap into it. |
1da177e4c
|
1086 1087 1088 |
*/ swap_map = &si->swap_map[i]; entry = swp_entry(type, i); |
02098feaa
|
1089 1090 |
page = read_swap_cache_async(entry, GFP_HIGHUSER_MOVABLE, NULL, 0); |
1da177e4c
|
1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 |
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
|
1128 |
*/ |
1da177e4c
|
1129 |
swcount = *swap_map; |
aaa468653
|
1130 1131 1132 1133 1134 1135 |
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
|
1136 |
} |
aaa468653
|
1137 1138 |
if (swap_count(swcount) && start_mm != &init_mm) retval = unuse_mm(start_mm, entry, page); |
355cfa73d
|
1139 |
if (swap_count(*swap_map)) { |
1da177e4c
|
1140 1141 1142 1143 1144 1145 1146 1147 1148 |
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
|
1149 |
while (swap_count(*swap_map) && !retval && |
1da177e4c
|
1150 1151 |
(p = p->next) != &start_mm->mmlist) { mm = list_entry(p, struct mm_struct, mmlist); |
70af7c5c6
|
1152 |
if (!atomic_inc_not_zero(&mm->mm_users)) |
1da177e4c
|
1153 |
continue; |
1da177e4c
|
1154 1155 1156 1157 1158 1159 1160 |
spin_unlock(&mmlist_lock); mmput(prev_mm); prev_mm = mm; cond_resched(); swcount = *swap_map; |
355cfa73d
|
1161 |
if (!swap_count(swcount)) /* any usage ? */ |
1da177e4c
|
1162 |
; |
aaa468653
|
1163 |
else if (mm == &init_mm) |
1da177e4c
|
1164 |
set_start_mm = 1; |
aaa468653
|
1165 |
else |
1da177e4c
|
1166 |
retval = unuse_mm(mm, entry, page); |
355cfa73d
|
1167 |
|
32c5fc10e
|
1168 |
if (set_start_mm && *swap_map < swcount) { |
1da177e4c
|
1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 |
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
|
1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 |
* 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
|
1199 1200 1201 1202 1203 1204 |
* * 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
|
1205 |
*/ |
355cfa73d
|
1206 1207 |
if (swap_count(*swap_map) && PageDirty(page) && PageSwapCache(page)) { |
1da177e4c
|
1208 1209 1210 1211 1212 1213 1214 1215 |
struct writeback_control wbc = { .sync_mode = WB_SYNC_NONE, }; swap_writepage(page, &wbc); lock_page(page); wait_on_page_writeback(page); } |
68bdc8d64
|
1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 |
/* * 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
|
1226 |
delete_from_swap_cache(page); |
1da177e4c
|
1227 1228 1229 1230 |
/* * So we could skip searching mms once swap count went * to 1, we did not mark any present ptes as dirty: must |
2706a1b89
|
1231 |
* mark page dirty so shrink_page_list will preserve it. |
1da177e4c
|
1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 |
*/ 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
|
1245 1246 1247 1248 |
return retval; } /* |
5d337b919
|
1249 1250 1251 |
* 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
|
1252 1253 1254 1255 1256 |
* added to the mmlist just after page_duplicate - before would be racy. */ static void drain_mmlist(void) { struct list_head *p, *next; |
efa90a981
|
1257 |
unsigned int type; |
1da177e4c
|
1258 |
|
efa90a981
|
1259 1260 |
for (type = 0; type < nr_swapfiles; type++) if (swap_info[type]->inuse_pages) |
1da177e4c
|
1261 1262 1263 1264 1265 1266 1267 1268 1269 |
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
|
1270 1271 1272 |
* 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
|
1273 |
*/ |
d4906e1aa
|
1274 |
static sector_t map_swap_entry(swp_entry_t entry, struct block_device **bdev) |
1da177e4c
|
1275 |
{ |
f29ad6a99
|
1276 1277 1278 1279 |
struct swap_info_struct *sis; struct swap_extent *start_se; struct swap_extent *se; pgoff_t offset; |
efa90a981
|
1280 |
sis = swap_info[swp_type(entry)]; |
f29ad6a99
|
1281 1282 1283 1284 1285 |
*bdev = sis->bdev; offset = swp_offset(entry); start_se = sis->curr_swap_extent; se = start_se; |
1da177e4c
|
1286 1287 1288 1289 1290 1291 1292 1293 |
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
|
1294 |
lh = se->list.next; |
1da177e4c
|
1295 1296 1297 1298 1299 1300 1301 |
se = list_entry(lh, struct swap_extent, list); sis->curr_swap_extent = se; BUG_ON(se == start_se); /* It *must* be present */ } } /* |
d4906e1aa
|
1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 |
* 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
|
1312 1313 1314 1315 |
* Free all of a swapdev's extent information */ static void destroy_swap_extents(struct swap_info_struct *sis) { |
9625a5f28
|
1316 |
while (!list_empty(&sis->first_swap_extent.list)) { |
1da177e4c
|
1317 |
struct swap_extent *se; |
9625a5f28
|
1318 |
se = list_entry(sis->first_swap_extent.list.next, |
1da177e4c
|
1319 1320 1321 1322 |
struct swap_extent, list); list_del(&se->list); kfree(se); } |
1da177e4c
|
1323 1324 1325 1326 |
} /* * Add a block range (and the corresponding page range) into this swapdev's |
11d31886d
|
1327 |
* extent list. The extent list is kept sorted in page order. |
1da177e4c
|
1328 |
* |
11d31886d
|
1329 |
* This function rather assumes that it is called in ascending page order. |
1da177e4c
|
1330 1331 1332 1333 1334 1335 1336 1337 |
*/ 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
|
1338 1339 1340 1341 1342 1343 1344 1345 1346 |
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
|
1347 |
se = list_entry(lh, struct swap_extent, list); |
11d31886d
|
1348 1349 |
BUG_ON(se->start_page + se->nr_pages != start_page); if (se->start_block + se->nr_pages == start_block) { |
1da177e4c
|
1350 1351 1352 1353 |
/* Merge it */ se->nr_pages += nr_pages; return 0; } |
1da177e4c
|
1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 |
} /* * 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
|
1365 |
list_add_tail(&new_se->list, &sis->first_swap_extent.list); |
53092a740
|
1366 |
return 1; |
1da177e4c
|
1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 |
} /* * 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
|
1389 |
* For S_ISREG swapfiles we set S_SWAPFILE across the life of the swapon. This |
1da177e4c
|
1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 |
* 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
|
1400 |
static int setup_swap_extents(struct swap_info_struct *sis, sector_t *span) |
1da177e4c
|
1401 1402 1403 1404 1405 1406 1407 |
{ struct inode *inode; unsigned blocks_per_page; unsigned long page_no; unsigned blkbits; sector_t probe_block; sector_t last_block; |
53092a740
|
1408 1409 1410 |
sector_t lowest_block = -1; sector_t highest_block = 0; int nr_extents = 0; |
1da177e4c
|
1411 1412 1413 1414 1415 |
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
|
1416 |
*span = sis->pages; |
9625a5f28
|
1417 |
goto out; |
1da177e4c
|
1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 |
} 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
|
1460 1461 1462 1463 1464 1465 1466 |
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
|
1467 1468 1469 |
/* * We found a PAGE_SIZE-length, PAGE_SIZE-aligned run of blocks */ |
53092a740
|
1470 1471 |
ret = add_swap_extent(sis, page_no, 1, first_block); if (ret < 0) |
1da177e4c
|
1472 |
goto out; |
53092a740
|
1473 |
nr_extents += ret; |
1da177e4c
|
1474 1475 1476 1477 1478 |
page_no++; probe_block += blocks_per_page; reprobe: continue; } |
53092a740
|
1479 1480 |
ret = nr_extents; *span = 1 + highest_block - lowest_block; |
1da177e4c
|
1481 |
if (page_no == 0) |
e2244ec2e
|
1482 |
page_no = 1; /* force Empty message */ |
1da177e4c
|
1483 |
sis->max = page_no; |
e2244ec2e
|
1484 |
sis->pages = page_no - 1; |
1da177e4c
|
1485 |
sis->highest_bit = page_no - 1; |
9625a5f28
|
1486 1487 |
out: return ret; |
1da177e4c
|
1488 1489 1490 1491 |
bad_bmap: printk(KERN_ERR "swapon: swapfile has holes "); ret = -EINVAL; |
9625a5f28
|
1492 |
goto out; |
1da177e4c
|
1493 |
} |
40531542e
|
1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 |
static void enable_swap_info(struct swap_info_struct *p, int prio, unsigned char *swap_map) { int i, prev; spin_lock(&swap_lock); if (prio >= 0) p->prio = prio; else p->prio = --least_priority; p->swap_map = swap_map; p->flags |= SWP_WRITEOK; nr_swap_pages += p->pages; total_swap_pages += p->pages; /* insert swap space into swap_list: */ prev = -1; for (i = swap_list.head; i >= 0; i = swap_info[i]->next) { if (p->prio >= swap_info[i]->prio) break; prev = i; } p->next = i; if (prev < 0) swap_list.head = swap_list.next = p->type; else swap_info[prev]->next = p->type; spin_unlock(&swap_lock); } |
c4ea37c26
|
1523 |
SYSCALL_DEFINE1(swapoff, const char __user *, specialfile) |
1da177e4c
|
1524 |
{ |
73c34b6ac
|
1525 |
struct swap_info_struct *p = NULL; |
8d69aaee8
|
1526 |
unsigned char *swap_map; |
1da177e4c
|
1527 1528 1529 |
struct file *swap_file, *victim; struct address_space *mapping; struct inode *inode; |
73c34b6ac
|
1530 |
char *pathname; |
72788c385
|
1531 |
int oom_score_adj; |
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 |
|
72788c385
|
1590 |
oom_score_adj = test_set_oom_score_adj(OOM_SCORE_ADJ_MAX); |
1da177e4c
|
1591 |
err = try_to_unuse(type); |
72788c385
|
1592 |
test_set_oom_score_adj(oom_score_adj); |
1da177e4c
|
1593 |
|
1da177e4c
|
1594 |
if (err) { |
40531542e
|
1595 1596 1597 1598 1599 1600 |
/* * reading p->prio and p->swap_map outside the lock is * safe here because only sys_swapon and sys_swapoff * change them, and there can be no other sys_swapon or * sys_swapoff for this swap_info_struct at this point. */ |
1da177e4c
|
1601 |
/* re-insert swap space back into swap_list */ |
40531542e
|
1602 |
enable_swap_info(p, p->prio, p->swap_map); |
1da177e4c
|
1603 1604 |
goto out_dput; } |
52b7efdbe
|
1605 |
|
5d337b919
|
1606 |
destroy_swap_extents(p); |
570a335b8
|
1607 1608 |
if (p->flags & SWP_CONTINUED) free_swap_count_continuations(p); |
fc0abb145
|
1609 |
mutex_lock(&swapon_mutex); |
5d337b919
|
1610 1611 |
spin_lock(&swap_lock); drain_mmlist(); |
52b7efdbe
|
1612 |
/* wait for anyone still in scan_swap_map */ |
52b7efdbe
|
1613 1614 |
p->highest_bit = 0; /* cuts scans short */ while (p->flags >= SWP_SCANNING) { |
5d337b919
|
1615 |
spin_unlock(&swap_lock); |
13e4b57f6
|
1616 |
schedule_timeout_uninterruptible(1); |
5d337b919
|
1617 |
spin_lock(&swap_lock); |
52b7efdbe
|
1618 |
} |
52b7efdbe
|
1619 |
|
1da177e4c
|
1620 1621 1622 1623 1624 1625 |
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
|
1626 |
spin_unlock(&swap_lock); |
fc0abb145
|
1627 |
mutex_unlock(&swapon_mutex); |
1da177e4c
|
1628 |
vfree(swap_map); |
27a7faa07
|
1629 1630 |
/* Destroy swap account informatin */ swap_cgroup_swapoff(type); |
1da177e4c
|
1631 1632 1633 1634 |
inode = mapping->host; if (S_ISBLK(inode->i_mode)) { struct block_device *bdev = I_BDEV(inode); set_blocksize(bdev, p->old_block_size); |
e525fd89d
|
1635 |
blkdev_put(bdev, FMODE_READ | FMODE_WRITE | FMODE_EXCL); |
1da177e4c
|
1636 |
} else { |
1b1dcc1b5
|
1637 |
mutex_lock(&inode->i_mutex); |
1da177e4c
|
1638 |
inode->i_flags &= ~S_SWAPFILE; |
1b1dcc1b5
|
1639 |
mutex_unlock(&inode->i_mutex); |
1da177e4c
|
1640 1641 1642 |
} filp_close(swap_file, NULL); err = 0; |
66d7dd518
|
1643 1644 |
atomic_inc(&proc_poll_event); wake_up_interruptible(&proc_poll_wait); |
1da177e4c
|
1645 1646 1647 1648 1649 1650 1651 1652 |
out_dput: filp_close(victim, NULL); out: return err; } #ifdef CONFIG_PROC_FS |
66d7dd518
|
1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 |
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
|
1671 1672 1673 |
/* iterator */ static void *swap_start(struct seq_file *swap, loff_t *pos) { |
efa90a981
|
1674 1675 |
struct swap_info_struct *si; int type; |
1da177e4c
|
1676 |
loff_t l = *pos; |
fc0abb145
|
1677 |
mutex_lock(&swapon_mutex); |
1da177e4c
|
1678 |
|
881e4aabe
|
1679 1680 |
if (!l) return SEQ_START_TOKEN; |
efa90a981
|
1681 1682 1683 1684 |
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
|
1685 |
continue; |
881e4aabe
|
1686 |
if (!--l) |
efa90a981
|
1687 |
return si; |
1da177e4c
|
1688 1689 1690 1691 1692 1693 1694 |
} return NULL; } static void *swap_next(struct seq_file *swap, void *v, loff_t *pos) { |
efa90a981
|
1695 1696 |
struct swap_info_struct *si = v; int type; |
1da177e4c
|
1697 |
|
881e4aabe
|
1698 |
if (v == SEQ_START_TOKEN) |
efa90a981
|
1699 1700 1701 |
type = 0; else type = si->type + 1; |
881e4aabe
|
1702 |
|
efa90a981
|
1703 1704 1705 1706 |
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
|
1707 1708 |
continue; ++*pos; |
efa90a981
|
1709 |
return si; |
1da177e4c
|
1710 1711 1712 1713 1714 1715 1716 |
} return NULL; } static void swap_stop(struct seq_file *swap, void *v) { |
fc0abb145
|
1717 |
mutex_unlock(&swapon_mutex); |
1da177e4c
|
1718 1719 1720 1721 |
} static int swap_show(struct seq_file *swap, void *v) { |
efa90a981
|
1722 |
struct swap_info_struct *si = v; |
1da177e4c
|
1723 1724 |
struct file *file; int len; |
efa90a981
|
1725 |
if (si == SEQ_START_TOKEN) { |
881e4aabe
|
1726 1727 1728 1729 |
seq_puts(swap,"Filename\t\t\t\tType\t\tSize\tUsed\tPriority "); return 0; } |
1da177e4c
|
1730 |
|
efa90a981
|
1731 |
file = si->swap_file; |
c32c2f63a
|
1732 1733 |
len = seq_path(swap, &file->f_path, " \t \\"); |
6eb396dc4
|
1734 1735 |
seq_printf(swap, "%*s%s\t%u\t%u\t%d ", |
886bb7e9c
|
1736 1737 |
len < 40 ? 40 - len : 1, " ", S_ISBLK(file->f_path.dentry->d_inode->i_mode) ? |
1da177e4c
|
1738 |
"partition" : "file\t", |
efa90a981
|
1739 1740 1741 |
si->pages << (PAGE_SHIFT - 10), si->inuse_pages << (PAGE_SHIFT - 10), si->prio); |
1da177e4c
|
1742 1743 |
return 0; } |
15ad7cdcf
|
1744 |
static const struct seq_operations swaps_op = { |
1da177e4c
|
1745 1746 1747 1748 1749 1750 1751 1752 |
.start = swap_start, .next = swap_next, .stop = swap_stop, .show = swap_show }; static int swaps_open(struct inode *inode, struct file *file) { |
66d7dd518
|
1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 |
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
|
1771 |
} |
15ad7cdcf
|
1772 |
static const struct file_operations proc_swaps_operations = { |
1da177e4c
|
1773 1774 1775 1776 |
.open = swaps_open, .read = seq_read, .llseek = seq_lseek, .release = seq_release, |
66d7dd518
|
1777 |
.poll = swaps_poll, |
1da177e4c
|
1778 1779 1780 1781 |
}; static int __init procswaps_init(void) { |
3d71f86f4
|
1782 |
proc_create("swaps", 0, NULL, &proc_swaps_operations); |
1da177e4c
|
1783 1784 1785 1786 |
return 0; } __initcall(procswaps_init); #endif /* CONFIG_PROC_FS */ |
1796316a8
|
1787 1788 1789 1790 1791 1792 1793 1794 |
#ifdef MAX_SWAPFILES_CHECK static int __init max_swapfiles_check(void) { MAX_SWAPFILES_CHECK(); return 0; } late_initcall(max_swapfiles_check); #endif |
53cbb2435
|
1795 |
static struct swap_info_struct *alloc_swap_info(void) |
1da177e4c
|
1796 |
{ |
73c34b6ac
|
1797 |
struct swap_info_struct *p; |
1da177e4c
|
1798 |
unsigned int type; |
efa90a981
|
1799 1800 1801 |
p = kzalloc(sizeof(*p), GFP_KERNEL); if (!p) |
53cbb2435
|
1802 |
return ERR_PTR(-ENOMEM); |
efa90a981
|
1803 |
|
5d337b919
|
1804 |
spin_lock(&swap_lock); |
efa90a981
|
1805 1806 |
for (type = 0; type < nr_swapfiles; type++) { if (!(swap_info[type]->flags & SWP_USED)) |
1da177e4c
|
1807 |
break; |
efa90a981
|
1808 |
} |
0697212a4
|
1809 |
if (type >= MAX_SWAPFILES) { |
5d337b919
|
1810 |
spin_unlock(&swap_lock); |
efa90a981
|
1811 |
kfree(p); |
730c0581c
|
1812 |
return ERR_PTR(-EPERM); |
1da177e4c
|
1813 |
} |
efa90a981
|
1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 |
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
|
1832 |
INIT_LIST_HEAD(&p->first_swap_extent.list); |
1da177e4c
|
1833 |
p->flags = SWP_USED; |
1da177e4c
|
1834 |
p->next = -1; |
5d337b919
|
1835 |
spin_unlock(&swap_lock); |
efa90a981
|
1836 |
|
53cbb2435
|
1837 |
return p; |
53cbb2435
|
1838 |
} |
4d0e1e107
|
1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 |
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, FMODE_READ | FMODE_WRITE | FMODE_EXCL, sys_swapon); if (error < 0) { p->bdev = NULL; |
87ade72a7
|
1850 |
return -EINVAL; |
4d0e1e107
|
1851 1852 1853 1854 |
} p->old_block_size = block_size(p->bdev); error = set_blocksize(p->bdev, PAGE_SIZE); if (error < 0) |
87ade72a7
|
1855 |
return error; |
4d0e1e107
|
1856 1857 1858 1859 |
p->flags |= SWP_BLKDEV; } else if (S_ISREG(inode->i_mode)) { p->bdev = inode->i_sb->s_bdev; mutex_lock(&inode->i_mutex); |
87ade72a7
|
1860 1861 1862 1863 |
if (IS_SWAPFILE(inode)) return -EBUSY; } else return -EINVAL; |
4d0e1e107
|
1864 1865 |
return 0; |
4d0e1e107
|
1866 |
} |
ca8bd38bf
|
1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 |
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; if (memcmp("SWAPSPACE2", swap_header->magic.magic, 10)) { printk(KERN_ERR "Unable to find swap-space signature "); |
387190253
|
1878 |
return 0; |
ca8bd38bf
|
1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 |
} /* 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); |
387190253
|
1895 |
return 0; |
ca8bd38bf
|
1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 |
} p->lowest_bit = 1; p->cluster_next = 1; p->cluster_nr = 0; /* * 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( 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; } p->highest_bit = maxpages - 1; if (!maxpages) |
387190253
|
1927 |
return 0; |
ca8bd38bf
|
1928 1929 1930 1931 1932 |
swapfilepages = i_size_read(inode) >> PAGE_SHIFT; if (swapfilepages && maxpages > swapfilepages) { printk(KERN_WARNING "Swap area shorter than signature indicates "); |
387190253
|
1933 |
return 0; |
ca8bd38bf
|
1934 1935 |
} if (swap_header->info.nr_badpages && S_ISREG(inode->i_mode)) |
387190253
|
1936 |
return 0; |
ca8bd38bf
|
1937 |
if (swap_header->info.nr_badpages > MAX_SWAP_BADPAGES) |
387190253
|
1938 |
return 0; |
ca8bd38bf
|
1939 1940 |
return maxpages; |
ca8bd38bf
|
1941 |
} |
915d4d7bc
|
1942 1943 1944 1945 1946 1947 1948 |
static int setup_swap_map_and_extents(struct swap_info_struct *p, union swap_header *swap_header, unsigned char *swap_map, unsigned long maxpages, sector_t *span) { int i; |
915d4d7bc
|
1949 1950 1951 1952 1953 1954 1955 |
unsigned int nr_good_pages; int nr_extents; nr_good_pages = maxpages - 1; /* omit header page */ for (i = 0; i < swap_header->info.nr_badpages; i++) { unsigned int page_nr = swap_header->info.badpages[i]; |
bdb8e3f68
|
1956 1957 |
if (page_nr == 0 || page_nr > swap_header->info.last_page) return -EINVAL; |
915d4d7bc
|
1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 |
if (page_nr < maxpages) { swap_map[page_nr] = SWAP_MAP_BAD; nr_good_pages--; } } if (nr_good_pages) { swap_map[0] = SWAP_MAP_BAD; p->max = maxpages; p->pages = nr_good_pages; nr_extents = setup_swap_extents(p, span); |
bdb8e3f68
|
1969 1970 |
if (nr_extents < 0) return nr_extents; |
915d4d7bc
|
1971 1972 1973 1974 1975 |
nr_good_pages = p->pages; } if (!nr_good_pages) { printk(KERN_WARNING "Empty swap-file "); |
bdb8e3f68
|
1976 |
return -EINVAL; |
915d4d7bc
|
1977 1978 1979 |
} return nr_extents; |
915d4d7bc
|
1980 |
} |
53cbb2435
|
1981 1982 1983 |
SYSCALL_DEFINE2(swapon, const char __user *, specialfile, int, swap_flags) { struct swap_info_struct *p; |
28b36bd74
|
1984 |
char *name; |
53cbb2435
|
1985 1986 |
struct file *swap_file = NULL; struct address_space *mapping; |
40531542e
|
1987 1988 |
int i; int prio; |
53cbb2435
|
1989 1990 |
int error; union swap_header *swap_header; |
915d4d7bc
|
1991 |
int nr_extents; |
53cbb2435
|
1992 1993 |
sector_t span; unsigned long maxpages; |
53cbb2435
|
1994 1995 1996 |
unsigned char *swap_map = NULL; struct page *page = NULL; struct inode *inode = NULL; |
53cbb2435
|
1997 1998 1999 2000 2001 |
if (!capable(CAP_SYS_ADMIN)) return -EPERM; p = alloc_swap_info(); |
2542e5134
|
2002 2003 |
if (IS_ERR(p)) return PTR_ERR(p); |
53cbb2435
|
2004 |
|
1da177e4c
|
2005 |
name = getname(specialfile); |
1da177e4c
|
2006 |
if (IS_ERR(name)) { |
7de7fb6b3
|
2007 |
error = PTR_ERR(name); |
1da177e4c
|
2008 |
name = NULL; |
bd69010b0
|
2009 |
goto bad_swap; |
1da177e4c
|
2010 2011 |
} swap_file = filp_open(name, O_RDWR|O_LARGEFILE, 0); |
1da177e4c
|
2012 |
if (IS_ERR(swap_file)) { |
7de7fb6b3
|
2013 |
error = PTR_ERR(swap_file); |
1da177e4c
|
2014 |
swap_file = NULL; |
bd69010b0
|
2015 |
goto bad_swap; |
1da177e4c
|
2016 2017 2018 2019 |
} p->swap_file = swap_file; mapping = swap_file->f_mapping; |
1da177e4c
|
2020 |
|
1da177e4c
|
2021 |
for (i = 0; i < nr_swapfiles; i++) { |
efa90a981
|
2022 |
struct swap_info_struct *q = swap_info[i]; |
1da177e4c
|
2023 |
|
e8e6c2ec4
|
2024 |
if (q == p || !q->swap_file) |
1da177e4c
|
2025 |
continue; |
7de7fb6b3
|
2026 2027 |
if (mapping == q->swap_file->f_mapping) { error = -EBUSY; |
1da177e4c
|
2028 |
goto bad_swap; |
7de7fb6b3
|
2029 |
} |
1da177e4c
|
2030 |
} |
2130781e2
|
2031 2032 |
inode = mapping->host; /* If S_ISREG(inode->i_mode) will do mutex_lock(&inode->i_mutex); */ |
4d0e1e107
|
2033 2034 |
error = claim_swapfile(p, inode); if (unlikely(error)) |
1da177e4c
|
2035 |
goto bad_swap; |
1da177e4c
|
2036 |
|
1da177e4c
|
2037 2038 2039 2040 2041 2042 2043 |
/* * Read the swap header. */ if (!mapping->a_ops->readpage) { error = -EINVAL; goto bad_swap; } |
090d2b185
|
2044 |
page = read_mapping_page(mapping, 0, swap_file); |
1da177e4c
|
2045 2046 2047 2048 |
if (IS_ERR(page)) { error = PTR_ERR(page); goto bad_swap; } |
81e339712
|
2049 |
swap_header = kmap(page); |
1da177e4c
|
2050 |
|
ca8bd38bf
|
2051 2052 |
maxpages = read_swap_header(p, swap_header, inode); if (unlikely(!maxpages)) { |
1da177e4c
|
2053 2054 2055 |
error = -EINVAL; goto bad_swap; } |
886bb7e9c
|
2056 |
|
81e339712
|
2057 |
/* OK, set up the swap map and apply the bad block list */ |
803d0c835
|
2058 |
swap_map = vzalloc(maxpages); |
81e339712
|
2059 2060 2061 2062 |
if (!swap_map) { error = -ENOMEM; goto bad_swap; } |
1da177e4c
|
2063 |
|
1421ef3cd
|
2064 2065 2066 |
error = swap_cgroup_swapon(p->type, maxpages); if (error) goto bad_swap; |
915d4d7bc
|
2067 2068 2069 2070 |
nr_extents = setup_swap_map_and_extents(p, swap_header, swap_map, maxpages, &span); if (unlikely(nr_extents < 0)) { error = nr_extents; |
1da177e4c
|
2071 2072 |
goto bad_swap; } |
1da177e4c
|
2073 |
|
3bd0f0c76
|
2074 2075 2076 2077 2078 |
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
|
2079 |
if (discard_swap(p) == 0 && (swap_flags & SWAP_FLAG_DISCARD)) |
3bd0f0c76
|
2080 |
p->flags |= SWP_DISCARDABLE; |
20137a490
|
2081 |
} |
6a6ba8317
|
2082 |
|
fc0abb145
|
2083 |
mutex_lock(&swapon_mutex); |
40531542e
|
2084 |
prio = -1; |
78ecba081
|
2085 |
if (swap_flags & SWAP_FLAG_PREFER) |
40531542e
|
2086 |
prio = |
78ecba081
|
2087 |
(swap_flags & SWAP_FLAG_PRIO_MASK) >> SWAP_FLAG_PRIO_SHIFT; |
40531542e
|
2088 |
enable_swap_info(p, prio, swap_map); |
c69dbfb84
|
2089 2090 2091 2092 2093 2094 2095 2096 |
printk(KERN_INFO "Adding %uk swap on %s. " "Priority:%d extents:%d across:%lluk %s%s ", p->pages<<(PAGE_SHIFT-10), name, p->prio, nr_extents, (unsigned long long)span<<(PAGE_SHIFT-10), (p->flags & SWP_SOLIDSTATE) ? "SS" : "", (p->flags & SWP_DISCARDABLE) ? "D" : ""); |
fc0abb145
|
2097 |
mutex_unlock(&swapon_mutex); |
66d7dd518
|
2098 2099 |
atomic_inc(&proc_poll_event); wake_up_interruptible(&proc_poll_wait); |
9b01c350a
|
2100 2101 |
if (S_ISREG(inode->i_mode)) inode->i_flags |= S_SWAPFILE; |
1da177e4c
|
2102 2103 2104 |
error = 0; goto out; bad_swap: |
bd69010b0
|
2105 |
if (inode && S_ISBLK(inode->i_mode) && p->bdev) { |
f2090d2df
|
2106 2107 |
set_blocksize(p->bdev, p->old_block_size); blkdev_put(p->bdev, FMODE_READ | FMODE_WRITE | FMODE_EXCL); |
1da177e4c
|
2108 |
} |
4cd3bb10f
|
2109 |
destroy_swap_extents(p); |
e8e6c2ec4
|
2110 |
swap_cgroup_swapoff(p->type); |
5d337b919
|
2111 |
spin_lock(&swap_lock); |
1da177e4c
|
2112 |
p->swap_file = NULL; |
1da177e4c
|
2113 |
p->flags = 0; |
5d337b919
|
2114 |
spin_unlock(&swap_lock); |
1da177e4c
|
2115 |
vfree(swap_map); |
52c50567d
|
2116 |
if (swap_file) { |
2130781e2
|
2117 |
if (inode && S_ISREG(inode->i_mode)) { |
52c50567d
|
2118 |
mutex_unlock(&inode->i_mutex); |
2130781e2
|
2119 2120 |
inode = NULL; } |
1da177e4c
|
2121 |
filp_close(swap_file, NULL); |
52c50567d
|
2122 |
} |
1da177e4c
|
2123 2124 2125 2126 2127 2128 2129 |
out: if (page && !IS_ERR(page)) { kunmap(page); page_cache_release(page); } if (name) putname(name); |
9b01c350a
|
2130 |
if (inode && S_ISREG(inode->i_mode)) |
1b1dcc1b5
|
2131 |
mutex_unlock(&inode->i_mutex); |
1da177e4c
|
2132 2133 2134 2135 2136 |
return error; } void si_swapinfo(struct sysinfo *val) { |
efa90a981
|
2137 |
unsigned int type; |
1da177e4c
|
2138 |
unsigned long nr_to_be_unused = 0; |
5d337b919
|
2139 |
spin_lock(&swap_lock); |
efa90a981
|
2140 2141 2142 2143 2144 |
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
|
2145 2146 2147 |
} val->freeswap = nr_swap_pages + nr_to_be_unused; val->totalswap = total_swap_pages + nr_to_be_unused; |
5d337b919
|
2148 |
spin_unlock(&swap_lock); |
1da177e4c
|
2149 2150 2151 2152 2153 |
} /* * Verify that a swap entry is valid and increment its swap map count. * |
355cfa73d
|
2154 2155 2156 2157 2158 2159 |
* 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
|
2160 |
* - swap-mapped reference requested but needs continued swap count. -> ENOMEM |
1da177e4c
|
2161 |
*/ |
8d69aaee8
|
2162 |
static int __swap_duplicate(swp_entry_t entry, unsigned char usage) |
1da177e4c
|
2163 |
{ |
73c34b6ac
|
2164 |
struct swap_info_struct *p; |
1da177e4c
|
2165 |
unsigned long offset, type; |
8d69aaee8
|
2166 2167 |
unsigned char count; unsigned char has_cache; |
253d553ba
|
2168 |
int err = -EINVAL; |
1da177e4c
|
2169 |
|
a7420aa54
|
2170 |
if (non_swap_entry(entry)) |
253d553ba
|
2171 |
goto out; |
0697212a4
|
2172 |
|
1da177e4c
|
2173 2174 2175 |
type = swp_type(entry); if (type >= nr_swapfiles) goto bad_file; |
efa90a981
|
2176 |
p = swap_info[type]; |
1da177e4c
|
2177 |
offset = swp_offset(entry); |
5d337b919
|
2178 |
spin_lock(&swap_lock); |
355cfa73d
|
2179 2180 |
if (unlikely(offset >= p->max)) goto unlock_out; |
253d553ba
|
2181 2182 2183 2184 |
count = p->swap_map[offset]; has_cache = count & SWAP_HAS_CACHE; count &= ~SWAP_HAS_CACHE; err = 0; |
355cfa73d
|
2185 |
|
253d553ba
|
2186 |
if (usage == SWAP_HAS_CACHE) { |
355cfa73d
|
2187 2188 |
/* set SWAP_HAS_CACHE if there is no cache and entry is used */ |
253d553ba
|
2189 2190 2191 2192 2193 2194 |
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
|
2195 2196 |
} else if (count || has_cache) { |
253d553ba
|
2197 |
|
570a335b8
|
2198 2199 2200 |
if ((count & ~COUNT_CONTINUED) < SWAP_MAP_MAX) count += usage; else if ((count & ~COUNT_CONTINUED) > SWAP_MAP_MAX) |
253d553ba
|
2201 |
err = -EINVAL; |
570a335b8
|
2202 2203 2204 2205 |
else if (swap_count_continued(p, offset, count)) count = COUNT_CONTINUED; else err = -ENOMEM; |
355cfa73d
|
2206 |
} else |
253d553ba
|
2207 2208 2209 |
err = -ENOENT; /* unused swap entry */ p->swap_map[offset] = count | has_cache; |
355cfa73d
|
2210 |
unlock_out: |
5d337b919
|
2211 |
spin_unlock(&swap_lock); |
1da177e4c
|
2212 |
out: |
253d553ba
|
2213 |
return err; |
1da177e4c
|
2214 2215 2216 2217 2218 2219 |
bad_file: printk(KERN_ERR "swap_dup: %s%08lx ", Bad_file, entry.val); goto out; } |
253d553ba
|
2220 |
|
355cfa73d
|
2221 |
/* |
aaa468653
|
2222 2223 2224 2225 2226 2227 2228 2229 2230 |
* 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
|
2231 2232 2233 2234 2235 |
* 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
|
2236 |
*/ |
570a335b8
|
2237 |
int swap_duplicate(swp_entry_t entry) |
355cfa73d
|
2238 |
{ |
570a335b8
|
2239 2240 2241 2242 2243 |
int err = 0; while (!err && __swap_duplicate(entry, 1) == -ENOMEM) err = add_swap_count_continuation(entry, GFP_ATOMIC); return err; |
355cfa73d
|
2244 |
} |
1da177e4c
|
2245 |
|
cb4b86ba4
|
2246 |
/* |
355cfa73d
|
2247 2248 |
* @entry: swap entry for which we allocate swap cache. * |
73c34b6ac
|
2249 |
* Called when allocating swap cache for existing swap entry, |
355cfa73d
|
2250 2251 2252 |
* 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
|
2253 2254 2255 |
*/ int swapcache_prepare(swp_entry_t entry) { |
253d553ba
|
2256 |
return __swap_duplicate(entry, SWAP_HAS_CACHE); |
cb4b86ba4
|
2257 |
} |
1da177e4c
|
2258 |
/* |
5d337b919
|
2259 |
* swap_lock prevents swap_map being freed. Don't grab an extra |
1da177e4c
|
2260 2261 2262 2263 |
* reference on the swaphandle, it doesn't matter if it becomes unused. */ int valid_swaphandles(swp_entry_t entry, unsigned long *offset) { |
8952898b0
|
2264 |
struct swap_info_struct *si; |
3f9e7949f
|
2265 |
int our_page_cluster = page_cluster; |
8952898b0
|
2266 2267 2268 |
pgoff_t target, toff; pgoff_t base, end; int nr_pages = 0; |
1da177e4c
|
2269 |
|
3f9e7949f
|
2270 |
if (!our_page_cluster) /* no readahead */ |
1da177e4c
|
2271 |
return 0; |
8952898b0
|
2272 |
|
efa90a981
|
2273 |
si = swap_info[swp_type(entry)]; |
8952898b0
|
2274 2275 2276 2277 2278 |
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
|
2279 |
|
5d337b919
|
2280 |
spin_lock(&swap_lock); |
8952898b0
|
2281 2282 2283 2284 2285 2286 2287 2288 |
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
|
2289 |
if (swap_count(si->swap_map[toff]) == SWAP_MAP_BAD) |
1da177e4c
|
2290 |
break; |
8952898b0
|
2291 2292 2293 |
} /* Count contiguous allocated slots below our target */ for (toff = target; --toff >= base; nr_pages++) { |
1da177e4c
|
2294 |
/* Don't read in free or bad pages */ |
8952898b0
|
2295 |
if (!si->swap_map[toff]) |
1da177e4c
|
2296 |
break; |
355cfa73d
|
2297 |
if (swap_count(si->swap_map[toff]) == SWAP_MAP_BAD) |
1da177e4c
|
2298 |
break; |
8952898b0
|
2299 |
} |
5d337b919
|
2300 |
spin_unlock(&swap_lock); |
8952898b0
|
2301 2302 2303 2304 2305 2306 2307 |
/* * 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
|
2308 |
} |
570a335b8
|
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 2521 2522 2523 2524 |
/* * 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); } } } } |