Blame view
mm/swapfile.c
97 KB
457c89965 treewide: Add SPD... |
1 |
// SPDX-License-Identifier: GPL-2.0-only |
1da177e4c Linux-2.6.12-rc2 |
2 3 4 5 6 7 |
/* * linux/mm/swapfile.c * * Copyright (C) 1991, 1992, 1993, 1994 Linus Torvalds * Swap reorganised 29.12.95, Stephen Tweedie */ |
1da177e4c Linux-2.6.12-rc2 |
8 |
#include <linux/mm.h> |
6e84f3152 sched/headers: Pr... |
9 |
#include <linux/sched/mm.h> |
299300258 sched/headers: Pr... |
10 |
#include <linux/sched/task.h> |
1da177e4c Linux-2.6.12-rc2 |
11 12 13 14 15 16 17 18 |
#include <linux/hugetlb.h> #include <linux/mman.h> #include <linux/slab.h> #include <linux/kernel_stat.h> #include <linux/swap.h> #include <linux/vmalloc.h> #include <linux/pagemap.h> #include <linux/namei.h> |
072441e21 mm: move shmem pr... |
19 |
#include <linux/shmem_fs.h> |
1da177e4c Linux-2.6.12-rc2 |
20 |
#include <linux/blkdev.h> |
20137a490 swapfile: swapon ... |
21 |
#include <linux/random.h> |
1da177e4c Linux-2.6.12-rc2 |
22 23 24 25 |
#include <linux/writeback.h> #include <linux/proc_fs.h> #include <linux/seq_file.h> #include <linux/init.h> |
5ad646880 ksm: let shared p... |
26 |
#include <linux/ksm.h> |
1da177e4c Linux-2.6.12-rc2 |
27 28 29 |
#include <linux/rmap.h> #include <linux/security.h> #include <linux/backing-dev.h> |
fc0abb145 [PATCH] sem2mutex... |
30 |
#include <linux/mutex.h> |
c59ede7b7 [PATCH] move capa... |
31 |
#include <linux/capability.h> |
1da177e4c Linux-2.6.12-rc2 |
32 |
#include <linux/syscalls.h> |
8a9f3ccd2 Memory controller... |
33 |
#include <linux/memcontrol.h> |
66d7dd518 /proc/swaps: supp... |
34 |
#include <linux/poll.h> |
72788c385 oom: replace PF_O... |
35 |
#include <linux/oom.h> |
38b5faf4b mm: frontswap: co... |
36 37 |
#include <linux/frontswap.h> #include <linux/swapfile.h> |
f981c5950 mm: methods for t... |
38 |
#include <linux/export.h> |
67afa38e0 mm/swap: add cach... |
39 |
#include <linux/swap_slots.h> |
155b5f88e mm/swapfile.c: so... |
40 |
#include <linux/sort.h> |
1da177e4c Linux-2.6.12-rc2 |
41 |
|
1da177e4c Linux-2.6.12-rc2 |
42 43 |
#include <asm/tlbflush.h> #include <linux/swapops.h> |
5d1ea48bd mm: page_cgroup: ... |
44 |
#include <linux/swap_cgroup.h> |
1da177e4c Linux-2.6.12-rc2 |
45 |
|
570a335b8 swap_info: swap c... |
46 47 48 |
static bool swap_count_continued(struct swap_info_struct *, pgoff_t, unsigned char); static void free_swap_count_continuations(struct swap_info_struct *); |
d4906e1aa swap: rework map_... |
49 |
static sector_t map_swap_entry(swp_entry_t, struct block_device**); |
570a335b8 swap_info: swap c... |
50 |
|
38b5faf4b mm: frontswap: co... |
51 |
DEFINE_SPINLOCK(swap_lock); |
7c363b8c6 mm/swapfile.c: ma... |
52 |
static unsigned int nr_swapfiles; |
ec8acf20a swap: add per-par... |
53 |
atomic_long_t nr_swap_pages; |
fb0fec501 mm: Export nr_swa... |
54 55 56 57 58 59 |
/* * Some modules use swappable objects and may try to swap them out under * memory pressure (via the shrinker). Before doing so, they may wish to * check to see if any swap space is available. */ EXPORT_SYMBOL_GPL(nr_swap_pages); |
ec8acf20a swap: add per-par... |
60 |
/* protected with swap_lock. reading in vm_swap_full() doesn't need lock */ |
1da177e4c Linux-2.6.12-rc2 |
61 |
long total_swap_pages; |
a2468cc9b swap: choose swap... |
62 |
static int least_priority = -1; |
1da177e4c Linux-2.6.12-rc2 |
63 |
|
1da177e4c Linux-2.6.12-rc2 |
64 65 66 67 |
static const char Bad_file[] = "Bad swap file entry "; static const char Unused_file[] = "Unused swap file entry "; static const char Bad_offset[] = "Bad swap offset entry "; static const char Unused_offset[] = "Unused swap offset entry "; |
adfab836f swap: change swap... |
68 69 70 71 |
/* * all active swap_info_structs * protected with swap_lock, and ordered by priority. */ |
18ab4d4ce swap: change swap... |
72 73 74 75 76 77 78 79 80 81 82 83 84 85 |
PLIST_HEAD(swap_active_head); /* * all available (active, not full) swap_info_structs * protected with swap_avail_lock, ordered by priority. * This is used by get_swap_page() instead of swap_active_head * because swap_active_head includes all swap_info_structs, * but get_swap_page() doesn't need to look at full ones. * This uses its own lock instead of swap_lock because when a * swap_info_struct changes between not-full/full, it needs to * add/remove itself to/from this list, but the swap_info_struct->lock * is held and the locking order requires swap_lock to be taken * before any swap_info_struct->lock. */ |
bfc6b1cab mm/swapfile.c: ma... |
86 |
static struct plist_head *swap_avail_heads; |
18ab4d4ce swap: change swap... |
87 |
static DEFINE_SPINLOCK(swap_avail_lock); |
1da177e4c Linux-2.6.12-rc2 |
88 |
|
38b5faf4b mm: frontswap: co... |
89 |
struct swap_info_struct *swap_info[MAX_SWAPFILES]; |
1da177e4c Linux-2.6.12-rc2 |
90 |
|
fc0abb145 [PATCH] sem2mutex... |
91 |
static DEFINE_MUTEX(swapon_mutex); |
1da177e4c Linux-2.6.12-rc2 |
92 |
|
66d7dd518 /proc/swaps: supp... |
93 94 95 |
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); |
81a0298bd mm, swap: don't u... |
96 |
atomic_t nr_rotate_swap = ATOMIC_INIT(0); |
c10d38cc8 mm, swap: bounds ... |
97 98 99 100 101 102 103 104 |
static struct swap_info_struct *swap_type_to_swap_info(int type) { if (type >= READ_ONCE(nr_swapfiles)) return NULL; smp_rmb(); /* Pairs with smp_wmb in alloc_swap_info. */ return READ_ONCE(swap_info[type]); } |
8d69aaee8 swap_info: swap_m... |
105 |
static inline unsigned char swap_count(unsigned char ent) |
355cfa73d mm: modify swap_m... |
106 |
{ |
955c97f08 mm/swapfile.c: fi... |
107 |
return ent & ~SWAP_HAS_CACHE; /* may include COUNT_CONTINUED flag */ |
355cfa73d mm: modify swap_m... |
108 |
} |
bcd49e867 mm/swapfile.c: us... |
109 110 111 112 113 114 115 116 117 |
/* Reclaim the swap entry anyway if possible */ #define TTRS_ANYWAY 0x1 /* * Reclaim the swap entry if there are no more mappings of the * corresponding page */ #define TTRS_UNMAPPED 0x2 /* Reclaim the swap entry if swap is getting full*/ #define TTRS_FULL 0x4 |
efa90a981 swap_info: change... |
118 |
/* returns 1 if swap entry is freed */ |
bcd49e867 mm/swapfile.c: us... |
119 120 |
static int __try_to_reclaim_swap(struct swap_info_struct *si, unsigned long offset, unsigned long flags) |
c9e444103 mm: reuse unused ... |
121 |
{ |
efa90a981 swap_info: change... |
122 |
swp_entry_t entry = swp_entry(si->type, offset); |
c9e444103 mm: reuse unused ... |
123 124 |
struct page *page; int ret = 0; |
bcd49e867 mm/swapfile.c: us... |
125 |
page = find_get_page(swap_address_space(entry), offset); |
c9e444103 mm: reuse unused ... |
126 127 128 |
if (!page) return 0; /* |
bcd49e867 mm/swapfile.c: us... |
129 130 131 |
* When this function is called from scan_swap_map_slots() 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 |
c9e444103 mm: reuse unused ... |
132 133 134 135 |
* case and you should use try_to_free_swap() with explicit lock_page() * in usual operations. */ if (trylock_page(page)) { |
bcd49e867 mm/swapfile.c: us... |
136 137 138 139 |
if ((flags & TTRS_ANYWAY) || ((flags & TTRS_UNMAPPED) && !page_mapped(page)) || ((flags & TTRS_FULL) && mem_cgroup_swap_full(page))) ret = try_to_free_swap(page); |
c9e444103 mm: reuse unused ... |
140 141 |
unlock_page(page); } |
09cbfeaf1 mm, fs: get rid o... |
142 |
put_page(page); |
c9e444103 mm: reuse unused ... |
143 144 |
return ret; } |
355cfa73d mm: modify swap_m... |
145 |
|
4efaceb1c mm, swap: use rbt... |
146 147 148 149 150 151 152 153 154 155 156 |
static inline struct swap_extent *first_se(struct swap_info_struct *sis) { struct rb_node *rb = rb_first(&sis->swap_extent_root); return rb_entry(rb, struct swap_extent, rb_node); } static inline struct swap_extent *next_se(struct swap_extent *se) { struct rb_node *rb = rb_next(&se->rb_node); return rb ? rb_entry(rb, struct swap_extent, rb_node) : NULL; } |
1da177e4c Linux-2.6.12-rc2 |
157 |
/* |
6a6ba8317 swapfile: swapon ... |
158 159 160 161 162 163 |
* 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 swap_info: includ... |
164 165 |
sector_t start_block; sector_t nr_blocks; |
6a6ba8317 swapfile: swapon ... |
166 |
int err = 0; |
9625a5f28 swap_info: includ... |
167 |
/* Do not discard the swap header page! */ |
4efaceb1c mm, swap: use rbt... |
168 |
se = first_se(si); |
9625a5f28 swap_info: includ... |
169 170 171 172 |
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 block: remove BLK... |
173 |
nr_blocks, GFP_KERNEL, 0); |
9625a5f28 swap_info: includ... |
174 175 176 177 |
if (err) return err; cond_resched(); } |
6a6ba8317 swapfile: swapon ... |
178 |
|
4efaceb1c mm, swap: use rbt... |
179 |
for (se = next_se(se); se; se = next_se(se)) { |
9625a5f28 swap_info: includ... |
180 181 |
start_block = se->start_block << (PAGE_SHIFT - 9); nr_blocks = (sector_t)se->nr_pages << (PAGE_SHIFT - 9); |
6a6ba8317 swapfile: swapon ... |
182 183 |
err = blkdev_issue_discard(si->bdev, start_block, |
dd3932edd block: remove BLK... |
184 |
nr_blocks, GFP_KERNEL, 0); |
6a6ba8317 swapfile: swapon ... |
185 186 187 188 189 190 191 |
if (err) break; cond_resched(); } return err; /* That will often be -EOPNOTSUPP */ } |
4efaceb1c mm, swap: use rbt... |
192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 |
static struct swap_extent * offset_to_swap_extent(struct swap_info_struct *sis, unsigned long offset) { struct swap_extent *se; struct rb_node *rb; rb = sis->swap_extent_root.rb_node; while (rb) { se = rb_entry(rb, struct swap_extent, rb_node); if (offset < se->start_page) rb = rb->rb_left; else if (offset >= se->start_page + se->nr_pages) rb = rb->rb_right; else return se; } /* It *must* be present */ BUG(); } |
7992fde72 swapfile: swap al... |
211 212 213 214 215 216 217 |
/* * 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) { |
4efaceb1c mm, swap: use rbt... |
218 |
struct swap_extent *se = offset_to_swap_extent(si, start_page); |
7992fde72 swapfile: swap al... |
219 220 |
while (nr_pages) { |
4efaceb1c mm, swap: use rbt... |
221 222 223 224 225 226 227 228 229 230 231 232 233 234 |
pgoff_t offset = start_page - se->start_page; sector_t start_block = se->start_block + offset; sector_t nr_blocks = se->nr_pages - offset; if (nr_blocks > nr_pages) nr_blocks = nr_pages; start_page += nr_blocks; nr_pages -= nr_blocks; start_block <<= PAGE_SHIFT - 9; nr_blocks <<= PAGE_SHIFT - 9; if (blkdev_issue_discard(si->bdev, start_block, nr_blocks, GFP_NOIO, 0)) break; |
7992fde72 swapfile: swap al... |
235 |
|
4efaceb1c mm, swap: use rbt... |
236 |
se = next_se(se); |
7992fde72 swapfile: swap al... |
237 238 |
} } |
38d8b4e6b mm, THP, swap: de... |
239 240 |
#ifdef CONFIG_THP_SWAP #define SWAPFILE_CLUSTER HPAGE_PMD_NR |
a448f2d07 mm/swapfile.c: un... |
241 242 |
#define swap_entry_size(size) (size) |
38d8b4e6b mm, THP, swap: de... |
243 |
#else |
048c27fd7 [PATCH] swap: sca... |
244 |
#define SWAPFILE_CLUSTER 256 |
a448f2d07 mm/swapfile.c: un... |
245 246 247 248 249 250 |
/* * Define swap_entry_size() as constant to let compiler to optimize * out some code if !CONFIG_THP_SWAP */ #define swap_entry_size(size) 1 |
38d8b4e6b mm, THP, swap: de... |
251 |
#endif |
048c27fd7 [PATCH] swap: sca... |
252 |
#define LATENCY_LIMIT 256 |
2a8f94493 swap: change bloc... |
253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 |
static inline void cluster_set_flag(struct swap_cluster_info *info, unsigned int flag) { info->flags = flag; } static inline unsigned int cluster_count(struct swap_cluster_info *info) { return info->data; } static inline void cluster_set_count(struct swap_cluster_info *info, unsigned int c) { info->data = c; } static inline void cluster_set_count_flag(struct swap_cluster_info *info, unsigned int c, unsigned int f) { info->flags = f; info->data = c; } static inline unsigned int cluster_next(struct swap_cluster_info *info) { return info->data; } static inline void cluster_set_next(struct swap_cluster_info *info, unsigned int n) { info->data = n; } static inline void cluster_set_next_flag(struct swap_cluster_info *info, unsigned int n, unsigned int f) { info->flags = f; info->data = n; } static inline bool cluster_is_free(struct swap_cluster_info *info) { return info->flags & CLUSTER_FLAG_FREE; } static inline bool cluster_is_null(struct swap_cluster_info *info) { return info->flags & CLUSTER_FLAG_NEXT_NULL; } static inline void cluster_set_null(struct swap_cluster_info *info) { info->flags = CLUSTER_FLAG_NEXT_NULL; info->data = 0; } |
e07098294 mm, THP, swap: su... |
310 311 |
static inline bool cluster_is_huge(struct swap_cluster_info *info) { |
33ee011e5 mm/swapfile.c: un... |
312 313 314 |
if (IS_ENABLED(CONFIG_THP_SWAP)) return info->flags & CLUSTER_FLAG_HUGE; return false; |
e07098294 mm, THP, swap: su... |
315 316 317 318 319 320 |
} static inline void cluster_clear_huge(struct swap_cluster_info *info) { info->flags &= ~CLUSTER_FLAG_HUGE; } |
235b62176 mm/swap: add clus... |
321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 |
static inline struct swap_cluster_info *lock_cluster(struct swap_info_struct *si, unsigned long offset) { struct swap_cluster_info *ci; ci = si->cluster_info; if (ci) { ci += offset / SWAPFILE_CLUSTER; spin_lock(&ci->lock); } return ci; } static inline void unlock_cluster(struct swap_cluster_info *ci) { if (ci) spin_unlock(&ci->lock); } |
59d98bf3c mm: swap: add com... |
339 340 341 342 |
/* * Determine the locking method in use for this device. Return * swap_cluster_info if SSD-style cluster-based locking is in place. */ |
235b62176 mm/swap: add clus... |
343 |
static inline struct swap_cluster_info *lock_cluster_or_swap_info( |
59d98bf3c mm: swap: add com... |
344 |
struct swap_info_struct *si, unsigned long offset) |
235b62176 mm/swap: add clus... |
345 346 |
{ struct swap_cluster_info *ci; |
59d98bf3c mm: swap: add com... |
347 |
/* Try to use fine-grained SSD-style locking if available: */ |
235b62176 mm/swap: add clus... |
348 |
ci = lock_cluster(si, offset); |
59d98bf3c mm: swap: add com... |
349 |
/* Otherwise, fall back to traditional, coarse locking: */ |
235b62176 mm/swap: add clus... |
350 351 352 353 354 355 356 357 358 359 360 361 362 363 |
if (!ci) spin_lock(&si->lock); return ci; } static inline void unlock_cluster_or_swap_info(struct swap_info_struct *si, struct swap_cluster_info *ci) { if (ci) unlock_cluster(ci); else spin_unlock(&si->lock); } |
6b5349159 mm, swap: add swa... |
364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 |
static inline bool cluster_list_empty(struct swap_cluster_list *list) { return cluster_is_null(&list->head); } static inline unsigned int cluster_list_first(struct swap_cluster_list *list) { return cluster_next(&list->head); } static void cluster_list_init(struct swap_cluster_list *list) { cluster_set_null(&list->head); cluster_set_null(&list->tail); } static void cluster_list_add_tail(struct swap_cluster_list *list, struct swap_cluster_info *ci, unsigned int idx) { if (cluster_list_empty(list)) { cluster_set_next_flag(&list->head, idx, 0); cluster_set_next_flag(&list->tail, idx, 0); } else { |
235b62176 mm/swap: add clus... |
388 |
struct swap_cluster_info *ci_tail; |
6b5349159 mm, swap: add swa... |
389 |
unsigned int tail = cluster_next(&list->tail); |
235b62176 mm/swap: add clus... |
390 391 392 393 394 395 396 |
/* * Nested cluster lock, but both cluster locks are * only acquired when we held swap_info_struct->lock */ ci_tail = ci + tail; spin_lock_nested(&ci_tail->lock, SINGLE_DEPTH_NESTING); cluster_set_next(ci_tail, idx); |
0ef017d11 mm, swap: improve... |
397 |
spin_unlock(&ci_tail->lock); |
6b5349159 mm, swap: add swa... |
398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 |
cluster_set_next_flag(&list->tail, idx, 0); } } static unsigned int cluster_list_del_first(struct swap_cluster_list *list, struct swap_cluster_info *ci) { unsigned int idx; idx = cluster_next(&list->head); if (cluster_next(&list->tail) == idx) { cluster_set_null(&list->head); cluster_set_null(&list->tail); } else cluster_set_next_flag(&list->head, cluster_next(&ci[idx]), 0); return idx; } |
815c2c543 swap: make swap d... |
417 418 419 420 421 422 423 424 425 426 427 428 |
/* Add a cluster to discard list and schedule it to do discard */ static void swap_cluster_schedule_discard(struct swap_info_struct *si, unsigned int idx) { /* * If scan_swap_map() can't find a free cluster, it will check * si->swap_map directly. To make sure the discarding cluster isn't * taken by scan_swap_map(), mark the swap entries bad (occupied). It * will be cleared after discard */ memset(si->swap_map + idx * SWAPFILE_CLUSTER, SWAP_MAP_BAD, SWAPFILE_CLUSTER); |
6b5349159 mm, swap: add swa... |
429 |
cluster_list_add_tail(&si->discard_clusters, si->cluster_info, idx); |
815c2c543 swap: make swap d... |
430 431 432 |
schedule_work(&si->discard_work); } |
38d8b4e6b mm, THP, swap: de... |
433 434 435 436 437 438 439 |
static void __free_cluster(struct swap_info_struct *si, unsigned long idx) { struct swap_cluster_info *ci = si->cluster_info; cluster_set_flag(ci + idx, CLUSTER_FLAG_FREE); cluster_list_add_tail(&si->free_clusters, ci, idx); } |
815c2c543 swap: make swap d... |
440 441 442 443 444 445 |
/* * Doing discard actually. After a cluster discard is finished, the cluster * will be added to free cluster list. caller should hold si->lock. */ static void swap_do_scheduled_discard(struct swap_info_struct *si) { |
235b62176 mm/swap: add clus... |
446 |
struct swap_cluster_info *info, *ci; |
815c2c543 swap: make swap d... |
447 448 449 |
unsigned int idx; info = si->cluster_info; |
6b5349159 mm, swap: add swa... |
450 451 |
while (!cluster_list_empty(&si->discard_clusters)) { idx = cluster_list_del_first(&si->discard_clusters, info); |
815c2c543 swap: make swap d... |
452 453 454 455 456 457 |
spin_unlock(&si->lock); discard_swap_cluster(si, idx * SWAPFILE_CLUSTER, SWAPFILE_CLUSTER); spin_lock(&si->lock); |
235b62176 mm/swap: add clus... |
458 |
ci = lock_cluster(si, idx * SWAPFILE_CLUSTER); |
38d8b4e6b mm, THP, swap: de... |
459 |
__free_cluster(si, idx); |
815c2c543 swap: make swap d... |
460 461 |
memset(si->swap_map + idx * SWAPFILE_CLUSTER, 0, SWAPFILE_CLUSTER); |
235b62176 mm/swap: add clus... |
462 |
unlock_cluster(ci); |
815c2c543 swap: make swap d... |
463 464 465 466 467 468 469 470 471 472 473 474 475 |
} } static void swap_discard_work(struct work_struct *work) { struct swap_info_struct *si; si = container_of(work, struct swap_info_struct, discard_work); spin_lock(&si->lock); swap_do_scheduled_discard(si); spin_unlock(&si->lock); } |
38d8b4e6b mm, THP, swap: de... |
476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 |
static void alloc_cluster(struct swap_info_struct *si, unsigned long idx) { struct swap_cluster_info *ci = si->cluster_info; VM_BUG_ON(cluster_list_first(&si->free_clusters) != idx); cluster_list_del_first(&si->free_clusters, ci); cluster_set_count_flag(ci + idx, 0, 0); } static void free_cluster(struct swap_info_struct *si, unsigned long idx) { struct swap_cluster_info *ci = si->cluster_info + idx; VM_BUG_ON(cluster_count(ci) != 0); /* * If the swap is discardable, prepare discard the cluster * instead of free it immediately. The cluster will be freed * after discard. */ if ((si->flags & (SWP_WRITEOK | SWP_PAGE_DISCARD)) == (SWP_WRITEOK | SWP_PAGE_DISCARD)) { swap_cluster_schedule_discard(si, idx); return; } __free_cluster(si, idx); } |
2a8f94493 swap: change bloc... |
503 504 505 506 507 508 509 510 511 512 513 |
/* * The cluster corresponding to page_nr will be used. The cluster will be * removed from free cluster list and its usage counter will be increased. */ static void inc_cluster_info_page(struct swap_info_struct *p, struct swap_cluster_info *cluster_info, unsigned long page_nr) { unsigned long idx = page_nr / SWAPFILE_CLUSTER; if (!cluster_info) return; |
38d8b4e6b mm, THP, swap: de... |
514 515 |
if (cluster_is_free(&cluster_info[idx])) alloc_cluster(p, idx); |
2a8f94493 swap: change bloc... |
516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 |
VM_BUG_ON(cluster_count(&cluster_info[idx]) >= SWAPFILE_CLUSTER); cluster_set_count(&cluster_info[idx], cluster_count(&cluster_info[idx]) + 1); } /* * The cluster corresponding to page_nr decreases one usage. If the usage * counter becomes 0, which means no page in the cluster is in using, we can * optionally discard the cluster and add it to free cluster list. */ static void dec_cluster_info_page(struct swap_info_struct *p, struct swap_cluster_info *cluster_info, unsigned long page_nr) { unsigned long idx = page_nr / SWAPFILE_CLUSTER; if (!cluster_info) return; VM_BUG_ON(cluster_count(&cluster_info[idx]) == 0); cluster_set_count(&cluster_info[idx], cluster_count(&cluster_info[idx]) - 1); |
38d8b4e6b mm, THP, swap: de... |
538 539 |
if (cluster_count(&cluster_info[idx]) == 0) free_cluster(p, idx); |
2a8f94493 swap: change bloc... |
540 541 542 543 544 545 |
} /* * It's possible scan_swap_map() uses a free cluster in the middle of free * cluster list. Avoiding such abuse to avoid list corruption. */ |
ebc2a1a69 swap: make cluste... |
546 547 |
static bool scan_swap_map_ssd_cluster_conflict(struct swap_info_struct *si, |
2a8f94493 swap: change bloc... |
548 549 |
unsigned long offset) { |
ebc2a1a69 swap: make cluste... |
550 551 |
struct percpu_cluster *percpu_cluster; bool conflict; |
2a8f94493 swap: change bloc... |
552 |
offset /= SWAPFILE_CLUSTER; |
6b5349159 mm, swap: add swa... |
553 554 |
conflict = !cluster_list_empty(&si->free_clusters) && offset != cluster_list_first(&si->free_clusters) && |
2a8f94493 swap: change bloc... |
555 |
cluster_is_free(&si->cluster_info[offset]); |
ebc2a1a69 swap: make cluste... |
556 557 558 559 560 561 562 563 564 565 566 567 568 |
if (!conflict) return false; percpu_cluster = this_cpu_ptr(si->percpu_cluster); cluster_set_null(&percpu_cluster->index); return true; } /* * Try to get a swap entry from current cpu's swap entry pool (a cluster). This * might involve allocating a new cluster for current CPU too. */ |
36005bae2 mm/swap: allocate... |
569 |
static bool scan_swap_map_try_ssd_cluster(struct swap_info_struct *si, |
ebc2a1a69 swap: make cluste... |
570 571 572 |
unsigned long *offset, unsigned long *scan_base) { struct percpu_cluster *cluster; |
235b62176 mm/swap: add clus... |
573 |
struct swap_cluster_info *ci; |
235b62176 mm/swap: add clus... |
574 |
unsigned long tmp, max; |
ebc2a1a69 swap: make cluste... |
575 576 577 578 |
new_cluster: cluster = this_cpu_ptr(si->percpu_cluster); if (cluster_is_null(&cluster->index)) { |
6b5349159 mm, swap: add swa... |
579 580 |
if (!cluster_list_empty(&si->free_clusters)) { cluster->index = si->free_clusters.head; |
ebc2a1a69 swap: make cluste... |
581 582 |
cluster->next = cluster_next(&cluster->index) * SWAPFILE_CLUSTER; |
6b5349159 mm, swap: add swa... |
583 |
} else if (!cluster_list_empty(&si->discard_clusters)) { |
ebc2a1a69 swap: make cluste... |
584 585 |
/* * we don't have free cluster but have some clusters in |
490705888 swap: reduce lock... |
586 587 |
* discarding, do discard now and reclaim them, then * reread cluster_next_cpu since we dropped si->lock |
ebc2a1a69 swap: make cluste... |
588 589 |
*/ swap_do_scheduled_discard(si); |
490705888 swap: reduce lock... |
590 591 |
*scan_base = this_cpu_read(*si->cluster_next_cpu); *offset = *scan_base; |
ebc2a1a69 swap: make cluste... |
592 593 |
goto new_cluster; } else |
36005bae2 mm/swap: allocate... |
594 |
return false; |
ebc2a1a69 swap: make cluste... |
595 |
} |
ebc2a1a69 swap: make cluste... |
596 597 598 599 600 |
/* * Other CPUs can use our cluster if they can't find a free cluster, * check if there is still free entry in the cluster */ tmp = cluster->next; |
235b62176 mm/swap: add clus... |
601 602 |
max = min_t(unsigned long, si->max, (cluster_next(&cluster->index) + 1) * SWAPFILE_CLUSTER); |
7b9e2de13 mm/swapfile.c: om... |
603 604 605 606 607 608 609 610 |
if (tmp < max) { ci = lock_cluster(si, tmp); while (tmp < max) { if (!si->swap_map[tmp]) break; tmp++; } unlock_cluster(ci); |
ebc2a1a69 swap: make cluste... |
611 |
} |
0fd0e19e4 mm/swapfile.c: fo... |
612 |
if (tmp >= max) { |
ebc2a1a69 swap: make cluste... |
613 614 615 616 617 618 |
cluster_set_null(&cluster->index); goto new_cluster; } cluster->next = tmp + 1; *offset = tmp; *scan_base = tmp; |
fdff1debb mm/swapfile.c: tm... |
619 |
return true; |
2a8f94493 swap: change bloc... |
620 |
} |
a2468cc9b swap: choose swap... |
621 622 623 624 625 626 627 628 629 630 631 632 633 634 |
static void __del_from_avail_list(struct swap_info_struct *p) { int nid; for_each_node(nid) plist_del(&p->avail_lists[nid], &swap_avail_heads[nid]); } static void del_from_avail_list(struct swap_info_struct *p) { spin_lock(&swap_avail_lock); __del_from_avail_list(p); spin_unlock(&swap_avail_lock); } |
38d8b4e6b mm, THP, swap: de... |
635 636 637 638 639 640 641 642 |
static void swap_range_alloc(struct swap_info_struct *si, unsigned long offset, unsigned int nr_entries) { unsigned int end = offset + nr_entries - 1; if (offset == si->lowest_bit) si->lowest_bit += nr_entries; if (end == si->highest_bit) |
a449bf58e mm/swapfile: fix ... |
643 |
WRITE_ONCE(si->highest_bit, si->highest_bit - nr_entries); |
38d8b4e6b mm, THP, swap: de... |
644 645 646 647 |
si->inuse_pages += nr_entries; if (si->inuse_pages == si->pages) { si->lowest_bit = si->max; si->highest_bit = 0; |
a2468cc9b swap: choose swap... |
648 |
del_from_avail_list(si); |
38d8b4e6b mm, THP, swap: de... |
649 650 |
} } |
a2468cc9b swap: choose swap... |
651 652 653 654 655 656 657 658 659 660 661 |
static void add_to_avail_list(struct swap_info_struct *p) { int nid; spin_lock(&swap_avail_lock); for_each_node(nid) { WARN_ON(!plist_node_empty(&p->avail_lists[nid])); plist_add(&p->avail_lists[nid], &swap_avail_heads[nid]); } spin_unlock(&swap_avail_lock); } |
38d8b4e6b mm, THP, swap: de... |
662 663 664 |
static void swap_range_free(struct swap_info_struct *si, unsigned long offset, unsigned int nr_entries) { |
3852f6768 mm/swapcache: sup... |
665 |
unsigned long begin = offset; |
38d8b4e6b mm, THP, swap: de... |
666 667 668 669 670 671 672 |
unsigned long end = offset + nr_entries - 1; void (*swap_slot_free_notify)(struct block_device *, unsigned long); if (offset < si->lowest_bit) si->lowest_bit = offset; if (end > si->highest_bit) { bool was_full = !si->highest_bit; |
a449bf58e mm/swapfile: fix ... |
673 |
WRITE_ONCE(si->highest_bit, end); |
a2468cc9b swap: choose swap... |
674 675 |
if (was_full && (si->flags & SWP_WRITEOK)) add_to_avail_list(si); |
38d8b4e6b mm, THP, swap: de... |
676 677 678 679 680 681 682 683 684 |
} atomic_long_add(nr_entries, &nr_swap_pages); si->inuse_pages -= nr_entries; if (si->flags & SWP_BLKDEV) swap_slot_free_notify = si->bdev->bd_disk->fops->swap_slot_free_notify; else swap_slot_free_notify = NULL; while (offset <= end) { |
8a84802e2 mm: Add arch hook... |
685 |
arch_swap_invalidate_page(si->type, offset); |
38d8b4e6b mm, THP, swap: de... |
686 687 688 689 690 |
frontswap_invalidate_page(si->type, offset); if (swap_slot_free_notify) swap_slot_free_notify(si->bdev, offset); offset++; } |
3852f6768 mm/swapcache: sup... |
691 |
clear_shadow_from_swap_cache(si->type, begin, end); |
38d8b4e6b mm, THP, swap: de... |
692 |
} |
490705888 swap: reduce lock... |
693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 |
static void set_cluster_next(struct swap_info_struct *si, unsigned long next) { unsigned long prev; if (!(si->flags & SWP_SOLIDSTATE)) { si->cluster_next = next; return; } prev = this_cpu_read(*si->cluster_next_cpu); /* * Cross the swap address space size aligned trunk, choose * another trunk randomly to avoid lock contention on swap * address space if possible. */ if ((prev >> SWAP_ADDRESS_SPACE_SHIFT) != (next >> SWAP_ADDRESS_SPACE_SHIFT)) { /* No free swap slots available */ if (si->highest_bit <= si->lowest_bit) return; next = si->lowest_bit + prandom_u32_max(si->highest_bit - si->lowest_bit + 1); next = ALIGN_DOWN(next, SWAP_ADDRESS_SPACE_PAGES); next = max_t(unsigned int, next, si->lowest_bit); } this_cpu_write(*si->cluster_next_cpu, next); } |
36005bae2 mm/swap: allocate... |
720 721 722 |
static int scan_swap_map_slots(struct swap_info_struct *si, unsigned char usage, int nr, swp_entry_t slots[]) |
1da177e4c Linux-2.6.12-rc2 |
723 |
{ |
235b62176 mm/swap: add clus... |
724 |
struct swap_cluster_info *ci; |
ebebbbe90 swapfile: rearran... |
725 |
unsigned long offset; |
c60aa176c swapfile: swap al... |
726 |
unsigned long scan_base; |
7992fde72 swapfile: swap al... |
727 |
unsigned long last_in_cluster = 0; |
048c27fd7 [PATCH] swap: sca... |
728 |
int latency_ration = LATENCY_LIMIT; |
36005bae2 mm/swap: allocate... |
729 |
int n_ret = 0; |
ed43af109 swap: try to scan... |
730 |
bool scanned_many = false; |
36005bae2 mm/swap: allocate... |
731 |
|
886bb7e9c swapfile: remove ... |
732 |
/* |
7dfad4183 [PATCH] swap: sca... |
733 734 735 736 737 738 739 |
* 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 swapfile: swap al... |
740 |
* And we let swap pages go all over an SSD partition. Hugh |
7dfad4183 [PATCH] swap: sca... |
741 |
*/ |
52b7efdbe [PATCH] swap: sca... |
742 |
si->flags += SWP_SCANNING; |
490705888 swap: reduce lock... |
743 744 745 746 747 748 749 750 751 752 |
/* * Use percpu scan base for SSD to reduce lock contention on * cluster and swap cache. For HDD, sequential access is more * important. */ if (si->flags & SWP_SOLIDSTATE) scan_base = this_cpu_read(*si->cluster_next_cpu); else scan_base = si->cluster_next; offset = scan_base; |
ebebbbe90 swapfile: rearran... |
753 |
|
ebc2a1a69 swap: make cluste... |
754 755 |
/* SSD algorithm */ if (si->cluster_info) { |
bd2d18da4 mm/swapfile.c: re... |
756 |
if (!scan_swap_map_try_ssd_cluster(si, &offset, &scan_base)) |
36005bae2 mm/swap: allocate... |
757 |
goto scan; |
f4eaf51a7 mm/swapfile.c: ex... |
758 |
} else if (unlikely(!si->cluster_nr--)) { |
ebebbbe90 swapfile: rearran... |
759 760 761 762 |
if (si->pages - si->inuse_pages < SWAPFILE_CLUSTER) { si->cluster_nr = SWAPFILE_CLUSTER - 1; goto checks; } |
2a8f94493 swap: change bloc... |
763 |
|
ec8acf20a swap: add per-par... |
764 |
spin_unlock(&si->lock); |
7dfad4183 [PATCH] swap: sca... |
765 |
|
c60aa176c swapfile: swap al... |
766 767 768 |
/* * If seek is expensive, start searching for new cluster from * start of partition, to minimize the span of allocated swap. |
50088c440 mm/swapfile.c: de... |
769 770 |
* If seek is cheap, that is the SWP_SOLIDSTATE si->cluster_info * case, just handled by scan_swap_map_try_ssd_cluster() above. |
c60aa176c swapfile: swap al... |
771 |
*/ |
50088c440 mm/swapfile.c: de... |
772 |
scan_base = offset = si->lowest_bit; |
7dfad4183 [PATCH] swap: sca... |
773 774 775 776 |
last_in_cluster = offset + SWAPFILE_CLUSTER - 1; /* Locate the first empty (unaligned) cluster */ for (; last_in_cluster <= si->highest_bit; offset++) { |
1da177e4c Linux-2.6.12-rc2 |
777 |
if (si->swap_map[offset]) |
7dfad4183 [PATCH] swap: sca... |
778 779 |
last_in_cluster = offset + SWAPFILE_CLUSTER; else if (offset == last_in_cluster) { |
ec8acf20a swap: add per-par... |
780 |
spin_lock(&si->lock); |
ebebbbe90 swapfile: rearran... |
781 782 783 |
offset -= SWAPFILE_CLUSTER - 1; si->cluster_next = offset; si->cluster_nr = SWAPFILE_CLUSTER - 1; |
c60aa176c swapfile: swap al... |
784 785 786 787 788 789 790 791 792 |
goto checks; } if (unlikely(--latency_ration < 0)) { cond_resched(); latency_ration = LATENCY_LIMIT; } } offset = scan_base; |
ec8acf20a swap: add per-par... |
793 |
spin_lock(&si->lock); |
ebebbbe90 swapfile: rearran... |
794 |
si->cluster_nr = SWAPFILE_CLUSTER - 1; |
1da177e4c Linux-2.6.12-rc2 |
795 |
} |
7dfad4183 [PATCH] swap: sca... |
796 |
|
ebebbbe90 swapfile: rearran... |
797 |
checks: |
ebc2a1a69 swap: make cluste... |
798 |
if (si->cluster_info) { |
36005bae2 mm/swap: allocate... |
799 800 801 802 803 804 805 806 |
while (scan_swap_map_ssd_cluster_conflict(si, offset)) { /* take a break if we already got some slots */ if (n_ret) goto done; if (!scan_swap_map_try_ssd_cluster(si, &offset, &scan_base)) goto scan; } |
ebc2a1a69 swap: make cluste... |
807 |
} |
ebebbbe90 swapfile: rearran... |
808 |
if (!(si->flags & SWP_WRITEOK)) |
52b7efdbe [PATCH] swap: sca... |
809 |
goto no_page; |
7dfad4183 [PATCH] swap: sca... |
810 811 |
if (!si->highest_bit) goto no_page; |
ebebbbe90 swapfile: rearran... |
812 |
if (offset > si->highest_bit) |
c60aa176c swapfile: swap al... |
813 |
scan_base = offset = si->lowest_bit; |
c9e444103 mm: reuse unused ... |
814 |
|
235b62176 mm/swap: add clus... |
815 |
ci = lock_cluster(si, offset); |
b73d7fcec swap: prevent reu... |
816 817 |
/* reuse swap entry of cache-only swap if not busy. */ if (vm_swap_full() && si->swap_map[offset] == SWAP_HAS_CACHE) { |
c9e444103 mm: reuse unused ... |
818 |
int swap_was_freed; |
235b62176 mm/swap: add clus... |
819 |
unlock_cluster(ci); |
ec8acf20a swap: add per-par... |
820 |
spin_unlock(&si->lock); |
bcd49e867 mm/swapfile.c: us... |
821 |
swap_was_freed = __try_to_reclaim_swap(si, offset, TTRS_ANYWAY); |
ec8acf20a swap: add per-par... |
822 |
spin_lock(&si->lock); |
c9e444103 mm: reuse unused ... |
823 824 825 826 827 |
/* entry was freed successfully, try to use this again */ if (swap_was_freed) goto checks; goto scan; /* check next one */ } |
235b62176 mm/swap: add clus... |
828 829 |
if (si->swap_map[offset]) { unlock_cluster(ci); |
36005bae2 mm/swap: allocate... |
830 831 832 833 |
if (!n_ret) goto scan; else goto done; |
235b62176 mm/swap: add clus... |
834 |
} |
a449bf58e mm/swapfile: fix ... |
835 |
WRITE_ONCE(si->swap_map[offset], usage); |
2872bb2d0 mm, swap: avoid l... |
836 837 |
inc_cluster_info_page(si, si->cluster_info, offset); unlock_cluster(ci); |
ebebbbe90 swapfile: rearran... |
838 |
|
38d8b4e6b mm, THP, swap: de... |
839 |
swap_range_alloc(si, offset, 1); |
36005bae2 mm/swap: allocate... |
840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 |
slots[n_ret++] = swp_entry(si->type, offset); /* got enough slots or reach max slots? */ if ((n_ret == nr) || (offset >= si->highest_bit)) goto done; /* search for next available slot */ /* time to take a break? */ if (unlikely(--latency_ration < 0)) { if (n_ret) goto done; spin_unlock(&si->lock); cond_resched(); spin_lock(&si->lock); latency_ration = LATENCY_LIMIT; } /* try to get more slots in cluster */ if (si->cluster_info) { if (scan_swap_map_try_ssd_cluster(si, &offset, &scan_base)) goto checks; |
f4eaf51a7 mm/swapfile.c: ex... |
862 863 |
} else if (si->cluster_nr && !si->swap_map[++offset]) { /* non-ssd case, still more slots in cluster? */ |
36005bae2 mm/swap: allocate... |
864 865 866 |
--si->cluster_nr; goto checks; } |
7992fde72 swapfile: swap al... |
867 |
|
ed43af109 swap: try to scan... |
868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 |
/* * Even if there's no free clusters available (fragmented), * try to scan a little more quickly with lock held unless we * have scanned too many slots already. */ if (!scanned_many) { unsigned long scan_limit; if (offset < scan_base) scan_limit = scan_base; else scan_limit = si->highest_bit; for (; offset <= scan_limit && --latency_ration > 0; offset++) { if (!si->swap_map[offset]) goto checks; } } |
36005bae2 mm/swap: allocate... |
886 |
done: |
490705888 swap: reduce lock... |
887 |
set_cluster_next(si, offset + 1); |
36005bae2 mm/swap: allocate... |
888 889 |
si->flags -= SWP_SCANNING; return n_ret; |
7dfad4183 [PATCH] swap: sca... |
890 |
|
ebebbbe90 swapfile: rearran... |
891 |
scan: |
ec8acf20a swap: add per-par... |
892 |
spin_unlock(&si->lock); |
a449bf58e mm/swapfile: fix ... |
893 894 |
while (++offset <= READ_ONCE(si->highest_bit)) { if (data_race(!si->swap_map[offset])) { |
ec8acf20a swap: add per-par... |
895 |
spin_lock(&si->lock); |
52b7efdbe [PATCH] swap: sca... |
896 897 |
goto checks; } |
a449bf58e mm/swapfile: fix ... |
898 899 |
if (vm_swap_full() && READ_ONCE(si->swap_map[offset]) == SWAP_HAS_CACHE) { |
ec8acf20a swap: add per-par... |
900 |
spin_lock(&si->lock); |
c9e444103 mm: reuse unused ... |
901 902 |
goto checks; } |
048c27fd7 [PATCH] swap: sca... |
903 904 905 |
if (unlikely(--latency_ration < 0)) { cond_resched(); latency_ration = LATENCY_LIMIT; |
ed43af109 swap: try to scan... |
906 |
scanned_many = true; |
048c27fd7 [PATCH] swap: sca... |
907 |
} |
7dfad4183 [PATCH] swap: sca... |
908 |
} |
c60aa176c swapfile: swap al... |
909 |
offset = si->lowest_bit; |
a5998061d mm/swapfile.c: do... |
910 |
while (offset < scan_base) { |
a449bf58e mm/swapfile: fix ... |
911 |
if (data_race(!si->swap_map[offset])) { |
ec8acf20a swap: add per-par... |
912 |
spin_lock(&si->lock); |
c60aa176c swapfile: swap al... |
913 914 |
goto checks; } |
a449bf58e mm/swapfile: fix ... |
915 916 |
if (vm_swap_full() && READ_ONCE(si->swap_map[offset]) == SWAP_HAS_CACHE) { |
ec8acf20a swap: add per-par... |
917 |
spin_lock(&si->lock); |
c9e444103 mm: reuse unused ... |
918 919 |
goto checks; } |
c60aa176c swapfile: swap al... |
920 921 922 |
if (unlikely(--latency_ration < 0)) { cond_resched(); latency_ration = LATENCY_LIMIT; |
ed43af109 swap: try to scan... |
923 |
scanned_many = true; |
c60aa176c swapfile: swap al... |
924 |
} |
a5998061d mm/swapfile.c: do... |
925 |
offset++; |
c60aa176c swapfile: swap al... |
926 |
} |
ec8acf20a swap: add per-par... |
927 |
spin_lock(&si->lock); |
7dfad4183 [PATCH] swap: sca... |
928 929 |
no_page: |
52b7efdbe [PATCH] swap: sca... |
930 |
si->flags -= SWP_SCANNING; |
36005bae2 mm/swap: allocate... |
931 |
return n_ret; |
1da177e4c Linux-2.6.12-rc2 |
932 |
} |
38d8b4e6b mm, THP, swap: de... |
933 934 935 936 937 938 |
static int swap_alloc_cluster(struct swap_info_struct *si, swp_entry_t *slot) { unsigned long idx; struct swap_cluster_info *ci; unsigned long offset, i; unsigned char *map; |
fe5266d5d mm/swapfile.c: re... |
939 940 941 942 943 944 945 946 |
/* * Should not even be attempting cluster allocations when huge * page swap is disabled. Warn and fail the allocation. */ if (!IS_ENABLED(CONFIG_THP_SWAP)) { VM_WARN_ON_ONCE(1); return 0; } |
38d8b4e6b mm, THP, swap: de... |
947 948 949 950 951 952 953 |
if (cluster_list_empty(&si->free_clusters)) return 0; idx = cluster_list_first(&si->free_clusters); offset = idx * SWAPFILE_CLUSTER; ci = lock_cluster(si, offset); alloc_cluster(si, idx); |
e07098294 mm, THP, swap: su... |
954 |
cluster_set_count_flag(ci, SWAPFILE_CLUSTER, CLUSTER_FLAG_HUGE); |
38d8b4e6b mm, THP, swap: de... |
955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 |
map = si->swap_map + offset; for (i = 0; i < SWAPFILE_CLUSTER; i++) map[i] = SWAP_HAS_CACHE; unlock_cluster(ci); swap_range_alloc(si, offset, SWAPFILE_CLUSTER); *slot = swp_entry(si->type, offset); return 1; } static void swap_free_cluster(struct swap_info_struct *si, unsigned long idx) { unsigned long offset = idx * SWAPFILE_CLUSTER; struct swap_cluster_info *ci; ci = lock_cluster(si, offset); |
979aafa59 mm/swapfile.c: cl... |
972 |
memset(si->swap_map + offset, 0, SWAPFILE_CLUSTER); |
38d8b4e6b mm, THP, swap: de... |
973 974 975 976 977 |
cluster_set_count_flag(ci, 0, 0); free_cluster(si, idx); unlock_cluster(ci); swap_range_free(si, offset, SWAPFILE_CLUSTER); } |
38d8b4e6b mm, THP, swap: de... |
978 |
|
36005bae2 mm/swap: allocate... |
979 980 981 982 983 984 985 986 987 988 989 990 991 992 |
static unsigned long scan_swap_map(struct swap_info_struct *si, unsigned char usage) { swp_entry_t entry; int n_ret; n_ret = scan_swap_map_slots(si, usage, 1, &entry); if (n_ret) return swp_offset(entry); else return 0; } |
5d5e8f195 mm, swap, get_swa... |
993 |
int get_swap_pages(int n_goal, swp_entry_t swp_entries[], int entry_size) |
1da177e4c Linux-2.6.12-rc2 |
994 |
{ |
5d5e8f195 mm, swap, get_swa... |
995 |
unsigned long size = swap_entry_size(entry_size); |
adfab836f swap: change swap... |
996 |
struct swap_info_struct *si, *next; |
36005bae2 mm/swap: allocate... |
997 998 |
long avail_pgs; int n_ret = 0; |
a2468cc9b swap: choose swap... |
999 |
int node; |
1da177e4c Linux-2.6.12-rc2 |
1000 |
|
38d8b4e6b mm, THP, swap: de... |
1001 |
/* Only single cluster request supported */ |
5d5e8f195 mm, swap, get_swa... |
1002 |
WARN_ON_ONCE(n_goal > 1 && size == SWAPFILE_CLUSTER); |
38d8b4e6b mm, THP, swap: de... |
1003 |
|
5d5e8f195 mm, swap, get_swa... |
1004 |
avail_pgs = atomic_long_read(&nr_swap_pages) / size; |
36005bae2 mm/swap: allocate... |
1005 |
if (avail_pgs <= 0) |
fb4f88dca [PATCH] swap: get... |
1006 |
goto noswap; |
36005bae2 mm/swap: allocate... |
1007 |
|
08d3090fc mm/swapfile.c: si... |
1008 |
n_goal = min3((long)n_goal, (long)SWAP_BATCH, avail_pgs); |
36005bae2 mm/swap: allocate... |
1009 |
|
5d5e8f195 mm, swap, get_swa... |
1010 |
atomic_long_sub(n_goal * size, &nr_swap_pages); |
fb4f88dca [PATCH] swap: get... |
1011 |
|
18ab4d4ce swap: change swap... |
1012 1013 1014 |
spin_lock(&swap_avail_lock); start_over: |
a2468cc9b swap: choose swap... |
1015 1016 |
node = numa_node_id(); plist_for_each_entry_safe(si, next, &swap_avail_heads[node], avail_lists[node]) { |
18ab4d4ce swap: change swap... |
1017 |
/* requeue si to after same-priority siblings */ |
a2468cc9b swap: choose swap... |
1018 |
plist_requeue(&si->avail_lists[node], &swap_avail_heads[node]); |
18ab4d4ce swap: change swap... |
1019 |
spin_unlock(&swap_avail_lock); |
ec8acf20a swap: add per-par... |
1020 |
spin_lock(&si->lock); |
adfab836f swap: change swap... |
1021 |
if (!si->highest_bit || !(si->flags & SWP_WRITEOK)) { |
18ab4d4ce swap: change swap... |
1022 |
spin_lock(&swap_avail_lock); |
a2468cc9b swap: choose swap... |
1023 |
if (plist_node_empty(&si->avail_lists[node])) { |
18ab4d4ce swap: change swap... |
1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 |
spin_unlock(&si->lock); goto nextsi; } WARN(!si->highest_bit, "swap_info %d in list but !highest_bit ", si->type); WARN(!(si->flags & SWP_WRITEOK), "swap_info %d in list but !SWP_WRITEOK ", si->type); |
a2468cc9b swap: choose swap... |
1035 |
__del_from_avail_list(si); |
ec8acf20a swap: add per-par... |
1036 |
spin_unlock(&si->lock); |
18ab4d4ce swap: change swap... |
1037 |
goto nextsi; |
ec8acf20a swap: add per-par... |
1038 |
} |
5d5e8f195 mm, swap, get_swa... |
1039 |
if (size == SWAPFILE_CLUSTER) { |
416634305 mm, THP, swap: fi... |
1040 |
if (si->flags & SWP_BLKDEV) |
f0eea189e mm, THP, swap: do... |
1041 1042 |
n_ret = swap_alloc_cluster(si, swp_entries); } else |
38d8b4e6b mm, THP, swap: de... |
1043 1044 |
n_ret = scan_swap_map_slots(si, SWAP_HAS_CACHE, n_goal, swp_entries); |
ec8acf20a swap: add per-par... |
1045 |
spin_unlock(&si->lock); |
5d5e8f195 mm, swap, get_swa... |
1046 |
if (n_ret || size == SWAPFILE_CLUSTER) |
36005bae2 mm/swap: allocate... |
1047 |
goto check_out; |
18ab4d4ce swap: change swap... |
1048 1049 |
pr_debug("scan_swap_map of si %d failed to find offset ", |
36005bae2 mm/swap: allocate... |
1050 |
si->type); |
18ab4d4ce swap: change swap... |
1051 1052 |
spin_lock(&swap_avail_lock); nextsi: |
adfab836f swap: change swap... |
1053 1054 1055 1056 |
/* * if we got here, it's likely that si was almost full before, * and since scan_swap_map() can drop the si->lock, multiple * callers probably all tried to get a page from the same si |
18ab4d4ce swap: change swap... |
1057 1058 1059 1060 |
* and it filled up before we could get one; or, the si filled * up between us dropping swap_avail_lock and taking si->lock. * Since we dropped the swap_avail_lock, the swap_avail_head * list may have been modified; so if next is still in the |
36005bae2 mm/swap: allocate... |
1061 1062 |
* swap_avail_head list then try it, otherwise start over * if we have not gotten any slots. |
adfab836f swap: change swap... |
1063 |
*/ |
a2468cc9b swap: choose swap... |
1064 |
if (plist_node_empty(&next->avail_lists[node])) |
18ab4d4ce swap: change swap... |
1065 |
goto start_over; |
1da177e4c Linux-2.6.12-rc2 |
1066 |
} |
fb4f88dca [PATCH] swap: get... |
1067 |
|
18ab4d4ce swap: change swap... |
1068 |
spin_unlock(&swap_avail_lock); |
36005bae2 mm/swap: allocate... |
1069 1070 |
check_out: if (n_ret < n_goal) |
5d5e8f195 mm, swap, get_swa... |
1071 |
atomic_long_add((long)(n_goal - n_ret) * size, |
38d8b4e6b mm, THP, swap: de... |
1072 |
&nr_swap_pages); |
fb4f88dca [PATCH] swap: get... |
1073 |
noswap: |
36005bae2 mm/swap: allocate... |
1074 1075 |
return n_ret; } |
2de1a7e40 mm/swapfile.c: fi... |
1076 |
/* The only caller of this function is now suspend routine */ |
910321ea8 swap: revert spec... |
1077 1078 |
swp_entry_t get_swap_page_of_type(int type) { |
c10d38cc8 mm, swap: bounds ... |
1079 |
struct swap_info_struct *si = swap_type_to_swap_info(type); |
910321ea8 swap: revert spec... |
1080 |
pgoff_t offset; |
c10d38cc8 mm, swap: bounds ... |
1081 1082 |
if (!si) goto fail; |
ec8acf20a swap: add per-par... |
1083 |
spin_lock(&si->lock); |
c10d38cc8 mm, swap: bounds ... |
1084 |
if (si->flags & SWP_WRITEOK) { |
ec8acf20a swap: add per-par... |
1085 |
atomic_long_dec(&nr_swap_pages); |
910321ea8 swap: revert spec... |
1086 1087 1088 |
/* This is called for allocating swap entry, not cache */ offset = scan_swap_map(si, 1); if (offset) { |
ec8acf20a swap: add per-par... |
1089 |
spin_unlock(&si->lock); |
910321ea8 swap: revert spec... |
1090 1091 |
return swp_entry(type, offset); } |
ec8acf20a swap: add per-par... |
1092 |
atomic_long_inc(&nr_swap_pages); |
910321ea8 swap: revert spec... |
1093 |
} |
ec8acf20a swap: add per-par... |
1094 |
spin_unlock(&si->lock); |
c10d38cc8 mm, swap: bounds ... |
1095 |
fail: |
910321ea8 swap: revert spec... |
1096 1097 |
return (swp_entry_t) {0}; } |
e8c26ab60 mm/swap: skip rea... |
1098 |
static struct swap_info_struct *__swap_info_get(swp_entry_t entry) |
1da177e4c Linux-2.6.12-rc2 |
1099 |
{ |
73c34b6ac swap_info: miscel... |
1100 |
struct swap_info_struct *p; |
eb085574a mm, swap: fix rac... |
1101 |
unsigned long offset; |
1da177e4c Linux-2.6.12-rc2 |
1102 1103 1104 |
if (!entry.val) goto out; |
eb085574a mm, swap: fix rac... |
1105 |
p = swp_swap_info(entry); |
c10d38cc8 mm, swap: bounds ... |
1106 |
if (!p) |
1da177e4c Linux-2.6.12-rc2 |
1107 |
goto bad_nofile; |
a449bf58e mm/swapfile: fix ... |
1108 |
if (data_race(!(p->flags & SWP_USED))) |
1da177e4c Linux-2.6.12-rc2 |
1109 1110 1111 1112 |
goto bad_device; offset = swp_offset(entry); if (offset >= p->max) goto bad_offset; |
1da177e4c Linux-2.6.12-rc2 |
1113 |
return p; |
1da177e4c Linux-2.6.12-rc2 |
1114 |
bad_offset: |
6a991fc72 mm/swap: fix kern... |
1115 1116 |
pr_err("swap_info_get: %s%08lx ", Bad_offset, entry.val); |
1da177e4c Linux-2.6.12-rc2 |
1117 1118 |
goto out; bad_device: |
6a991fc72 mm/swap: fix kern... |
1119 1120 |
pr_err("swap_info_get: %s%08lx ", Unused_file, entry.val); |
1da177e4c Linux-2.6.12-rc2 |
1121 1122 |
goto out; bad_nofile: |
6a991fc72 mm/swap: fix kern... |
1123 1124 |
pr_err("swap_info_get: %s%08lx ", Bad_file, entry.val); |
1da177e4c Linux-2.6.12-rc2 |
1125 1126 |
out: return NULL; |
886bb7e9c swapfile: remove ... |
1127 |
} |
1da177e4c Linux-2.6.12-rc2 |
1128 |
|
e8c26ab60 mm/swap: skip rea... |
1129 1130 1131 1132 1133 1134 1135 |
static struct swap_info_struct *_swap_info_get(swp_entry_t entry) { struct swap_info_struct *p; p = __swap_info_get(entry); if (!p) goto out; |
a449bf58e mm/swapfile: fix ... |
1136 |
if (data_race(!p->swap_map[swp_offset(entry)])) |
e8c26ab60 mm/swap: skip rea... |
1137 1138 1139 1140 1141 1142 |
goto bad_free; return p; bad_free: pr_err("swap_info_get: %s%08lx ", Unused_offset, entry.val); |
e8c26ab60 mm/swap: skip rea... |
1143 1144 1145 |
out: return NULL; } |
235b62176 mm/swap: add clus... |
1146 1147 1148 1149 1150 1151 1152 1153 1154 |
static struct swap_info_struct *swap_info_get(swp_entry_t entry) { struct swap_info_struct *p; p = _swap_info_get(entry); if (p) spin_lock(&p->lock); return p; } |
7c00bafee mm/swap: free swa... |
1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 |
static struct swap_info_struct *swap_info_get_cont(swp_entry_t entry, struct swap_info_struct *q) { struct swap_info_struct *p; p = _swap_info_get(entry); if (p != q) { if (q != NULL) spin_unlock(&q->lock); if (p != NULL) spin_lock(&p->lock); } return p; } |
b32d5f32b mm/swapfile.c: ad... |
1170 1171 1172 |
static unsigned char __swap_entry_free_locked(struct swap_info_struct *p, unsigned long offset, unsigned char usage) |
1da177e4c Linux-2.6.12-rc2 |
1173 |
{ |
8d69aaee8 swap_info: swap_m... |
1174 1175 |
unsigned char count; unsigned char has_cache; |
235b62176 mm/swap: add clus... |
1176 |
|
253d553ba swap_info: SWAP_H... |
1177 |
count = p->swap_map[offset]; |
235b62176 mm/swap: add clus... |
1178 |
|
253d553ba swap_info: SWAP_H... |
1179 1180 |
has_cache = count & SWAP_HAS_CACHE; count &= ~SWAP_HAS_CACHE; |
355cfa73d mm: modify swap_m... |
1181 |
|
253d553ba swap_info: SWAP_H... |
1182 |
if (usage == SWAP_HAS_CACHE) { |
355cfa73d mm: modify swap_m... |
1183 |
VM_BUG_ON(!has_cache); |
253d553ba swap_info: SWAP_H... |
1184 |
has_cache = 0; |
aaa468653 swap_info: note S... |
1185 1186 1187 1188 1189 1190 |
} 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 swap_info: swap c... |
1191 1192 1193 1194 1195 1196 1197 1198 1199 |
} 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 swap_info: SWAP_H... |
1200 |
|
253d553ba swap_info: SWAP_H... |
1201 |
usage = count | has_cache; |
a449bf58e mm/swapfile: fix ... |
1202 1203 1204 1205 |
if (usage) WRITE_ONCE(p->swap_map[offset], usage); else WRITE_ONCE(p->swap_map[offset], SWAP_HAS_CACHE); |
7c00bafee mm/swap: free swa... |
1206 |
|
b32d5f32b mm/swapfile.c: ad... |
1207 1208 |
return usage; } |
eb085574a mm, swap: fix rac... |
1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 |
/* * Check whether swap entry is valid in the swap device. If so, * return pointer to swap_info_struct, and keep the swap entry valid * via preventing the swap device from being swapoff, until * put_swap_device() is called. Otherwise return NULL. * * The entirety of the RCU read critical section must come before the * return from or after the call to synchronize_rcu() in * enable_swap_info() or swapoff(). So if "si->flags & SWP_VALID" is * true, the si->map, si->cluster_info, etc. must be valid in the * critical section. * * Notice that swapoff or swapoff+swapon can still happen before the * rcu_read_lock() in get_swap_device() or after the rcu_read_unlock() * in put_swap_device() if there isn't any other way to prevent * swapoff, such as page lock, page table lock, etc. The caller must * be prepared for that. For example, the following situation is * possible. * * CPU1 CPU2 * do_swap_page() * ... swapoff+swapon * __read_swap_cache_async() * swapcache_prepare() * __swap_duplicate() * // check swap_map * // verify PTE not changed * * In __swap_duplicate(), the swap_map need to be checked before * changing partly because the specified swap entry may be for another * swap device which has been swapoff. And in do_swap_page(), after * the page is read from the swap device, the PTE is verified not * changed with the page table locked to check whether the swap device * has been swapoff or swapoff+swapon. */ struct swap_info_struct *get_swap_device(swp_entry_t entry) { struct swap_info_struct *si; unsigned long offset; if (!entry.val) goto out; si = swp_swap_info(entry); if (!si) goto bad_nofile; rcu_read_lock(); |
a449bf58e mm/swapfile: fix ... |
1256 |
if (data_race(!(si->flags & SWP_VALID))) |
eb085574a mm, swap: fix rac... |
1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 |
goto unlock_out; offset = swp_offset(entry); if (offset >= si->max) goto unlock_out; return si; bad_nofile: pr_err("%s: %s%08lx ", __func__, Bad_file, entry.val); out: return NULL; unlock_out: rcu_read_unlock(); return NULL; } |
b32d5f32b mm/swapfile.c: ad... |
1272 |
static unsigned char __swap_entry_free(struct swap_info_struct *p, |
33e16272f mm/swapfile.c: __... |
1273 |
swp_entry_t entry) |
b32d5f32b mm/swapfile.c: ad... |
1274 1275 1276 |
{ struct swap_cluster_info *ci; unsigned long offset = swp_offset(entry); |
33e16272f mm/swapfile.c: __... |
1277 |
unsigned char usage; |
b32d5f32b mm/swapfile.c: ad... |
1278 1279 |
ci = lock_cluster_or_swap_info(p, offset); |
33e16272f mm/swapfile.c: __... |
1280 |
usage = __swap_entry_free_locked(p, offset, 1); |
7c00bafee mm/swap: free swa... |
1281 |
unlock_cluster_or_swap_info(p, ci); |
10e364da1 mm/swapfile.c: ca... |
1282 1283 |
if (!usage) free_swap_slot(entry); |
7c00bafee mm/swap: free swa... |
1284 1285 1286 |
return usage; } |
355cfa73d mm: modify swap_m... |
1287 |
|
7c00bafee mm/swap: free swa... |
1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 |
static void swap_entry_free(struct swap_info_struct *p, swp_entry_t entry) { struct swap_cluster_info *ci; unsigned long offset = swp_offset(entry); unsigned char count; ci = lock_cluster(p, offset); count = p->swap_map[offset]; VM_BUG_ON(count != SWAP_HAS_CACHE); p->swap_map[offset] = 0; dec_cluster_info_page(p, p->cluster_info, offset); |
235b62176 mm/swap: add clus... |
1299 |
unlock_cluster(ci); |
38d8b4e6b mm, THP, swap: de... |
1300 1301 |
mem_cgroup_uncharge_swap(entry, 1); swap_range_free(p, offset, 1); |
1da177e4c Linux-2.6.12-rc2 |
1302 1303 1304 |
} /* |
2de1a7e40 mm/swapfile.c: fi... |
1305 |
* Caller has made sure that the swap device corresponding to entry |
1da177e4c Linux-2.6.12-rc2 |
1306 1307 1308 1309 |
* is still around or has not been recycled. */ void swap_free(swp_entry_t entry) { |
73c34b6ac swap_info: miscel... |
1310 |
struct swap_info_struct *p; |
1da177e4c Linux-2.6.12-rc2 |
1311 |
|
235b62176 mm/swap: add clus... |
1312 |
p = _swap_info_get(entry); |
10e364da1 mm/swapfile.c: ca... |
1313 |
if (p) |
33e16272f mm/swapfile.c: __... |
1314 |
__swap_entry_free(p, entry); |
1da177e4c Linux-2.6.12-rc2 |
1315 1316 1317 |
} /* |
cb4b86ba4 mm: add swap cach... |
1318 1319 |
* Called after dropping swapcache to decrease refcnt to swap entries. */ |
a448f2d07 mm/swapfile.c: un... |
1320 |
void put_swap_page(struct page *page, swp_entry_t entry) |
38d8b4e6b mm, THP, swap: de... |
1321 1322 1323 1324 1325 1326 |
{ unsigned long offset = swp_offset(entry); unsigned long idx = offset / SWAPFILE_CLUSTER; struct swap_cluster_info *ci; struct swap_info_struct *si; unsigned char *map; |
a3aea839e mm, THP, swap: su... |
1327 1328 |
unsigned int i, free_entries = 0; unsigned char val; |
6c357848b mm: replace hpage... |
1329 |
int size = swap_entry_size(thp_nr_pages(page)); |
fe5266d5d mm/swapfile.c: re... |
1330 |
|
a3aea839e mm, THP, swap: su... |
1331 |
si = _swap_info_get(entry); |
38d8b4e6b mm, THP, swap: de... |
1332 1333 |
if (!si) return; |
c2343d276 mm/swapfile.c: pu... |
1334 |
ci = lock_cluster_or_swap_info(si, offset); |
a448f2d07 mm/swapfile.c: un... |
1335 |
if (size == SWAPFILE_CLUSTER) { |
a448f2d07 mm/swapfile.c: un... |
1336 1337 1338 1339 1340 1341 1342 1343 |
VM_BUG_ON(!cluster_is_huge(ci)); map = si->swap_map + offset; for (i = 0; i < SWAPFILE_CLUSTER; i++) { val = map[i]; VM_BUG_ON(!(val & SWAP_HAS_CACHE)); if (val == SWAP_HAS_CACHE) free_entries++; } |
a448f2d07 mm/swapfile.c: un... |
1344 |
cluster_clear_huge(ci); |
a448f2d07 mm/swapfile.c: un... |
1345 |
if (free_entries == SWAPFILE_CLUSTER) { |
c2343d276 mm/swapfile.c: pu... |
1346 |
unlock_cluster_or_swap_info(si, ci); |
a448f2d07 mm/swapfile.c: un... |
1347 |
spin_lock(&si->lock); |
a448f2d07 mm/swapfile.c: un... |
1348 1349 1350 1351 1352 1353 |
mem_cgroup_uncharge_swap(entry, SWAPFILE_CLUSTER); swap_free_cluster(si, idx); spin_unlock(&si->lock); return; } } |
c2343d276 mm/swapfile.c: pu... |
1354 1355 1356 1357 1358 1359 1360 |
for (i = 0; i < size; i++, entry.val++) { if (!__swap_entry_free_locked(si, offset + i, SWAP_HAS_CACHE)) { unlock_cluster_or_swap_info(si, ci); free_swap_slot(entry); if (i == size - 1) return; lock_cluster_or_swap_info(si, offset); |
a3aea839e mm, THP, swap: su... |
1361 1362 |
} } |
c2343d276 mm/swapfile.c: pu... |
1363 |
unlock_cluster_or_swap_info(si, ci); |
38d8b4e6b mm, THP, swap: de... |
1364 |
} |
59807685a mm, THP, swap: su... |
1365 |
|
fe5266d5d mm/swapfile.c: re... |
1366 |
#ifdef CONFIG_THP_SWAP |
59807685a mm, THP, swap: su... |
1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 |
int split_swap_cluster(swp_entry_t entry) { struct swap_info_struct *si; struct swap_cluster_info *ci; unsigned long offset = swp_offset(entry); si = _swap_info_get(entry); if (!si) return -EBUSY; ci = lock_cluster(si, offset); cluster_clear_huge(ci); unlock_cluster(ci); return 0; } |
fe5266d5d mm/swapfile.c: re... |
1381 |
#endif |
38d8b4e6b mm, THP, swap: de... |
1382 |
|
155b5f88e mm/swapfile.c: so... |
1383 1384 1385 1386 1387 1388 |
static int swp_entry_cmp(const void *ent1, const void *ent2) { const swp_entry_t *e1 = ent1, *e2 = ent2; return (int)swp_type(*e1) - (int)swp_type(*e2); } |
7c00bafee mm/swap: free swa... |
1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 |
void swapcache_free_entries(swp_entry_t *entries, int n) { struct swap_info_struct *p, *prev; int i; if (n <= 0) return; prev = NULL; p = NULL; |
155b5f88e mm/swapfile.c: so... |
1399 1400 1401 1402 1403 1404 1405 1406 |
/* * Sort swap entries by swap device, so each lock is only taken once. * nr_swapfiles isn't absolutely correct, but the overhead of sort() is * so low that it isn't necessary to optimize further. */ if (nr_swapfiles > 1) sort(entries, n, sizeof(entries[0]), swp_entry_cmp, NULL); |
7c00bafee mm/swap: free swa... |
1407 1408 1409 1410 |
for (i = 0; i < n; ++i) { p = swap_info_get_cont(entries[i], prev); if (p) swap_entry_free(p, entries[i]); |
7c00bafee mm/swap: free swa... |
1411 1412 |
prev = p; } |
235b62176 mm/swap: add clus... |
1413 |
if (p) |
7c00bafee mm/swap: free swa... |
1414 |
spin_unlock(&p->lock); |
cb4b86ba4 mm: add swap cach... |
1415 1416 1417 |
} /* |
c475a8ab6 [PATCH] can_share... |
1418 |
* How many references to page are currently swapped out? |
570a335b8 swap_info: swap c... |
1419 1420 |
* 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 Linux-2.6.12-rc2 |
1421 |
*/ |
bde05d1cc shmem: replace pa... |
1422 |
int page_swapcount(struct page *page) |
1da177e4c Linux-2.6.12-rc2 |
1423 |
{ |
c475a8ab6 [PATCH] can_share... |
1424 1425 |
int count = 0; struct swap_info_struct *p; |
235b62176 mm/swap: add clus... |
1426 |
struct swap_cluster_info *ci; |
1da177e4c Linux-2.6.12-rc2 |
1427 |
swp_entry_t entry; |
235b62176 mm/swap: add clus... |
1428 |
unsigned long offset; |
1da177e4c Linux-2.6.12-rc2 |
1429 |
|
4c21e2f24 [PATCH] mm: split... |
1430 |
entry.val = page_private(page); |
235b62176 mm/swap: add clus... |
1431 |
p = _swap_info_get(entry); |
1da177e4c Linux-2.6.12-rc2 |
1432 |
if (p) { |
235b62176 mm/swap: add clus... |
1433 1434 1435 1436 |
offset = swp_offset(entry); ci = lock_cluster_or_swap_info(p, offset); count = swap_count(p->swap_map[offset]); unlock_cluster_or_swap_info(p, ci); |
1da177e4c Linux-2.6.12-rc2 |
1437 |
} |
c475a8ab6 [PATCH] can_share... |
1438 |
return count; |
1da177e4c Linux-2.6.12-rc2 |
1439 |
} |
eb085574a mm, swap: fix rac... |
1440 |
int __swap_count(swp_entry_t entry) |
aa8d22a11 mm: swap: SWP_SYN... |
1441 |
{ |
eb085574a mm, swap: fix rac... |
1442 |
struct swap_info_struct *si; |
aa8d22a11 mm: swap: SWP_SYN... |
1443 |
pgoff_t offset = swp_offset(entry); |
eb085574a mm, swap: fix rac... |
1444 |
int count = 0; |
aa8d22a11 mm: swap: SWP_SYN... |
1445 |
|
eb085574a mm, swap: fix rac... |
1446 1447 1448 1449 1450 1451 |
si = get_swap_device(entry); if (si) { count = swap_count(si->swap_map[offset]); put_swap_device(si); } return count; |
aa8d22a11 mm: swap: SWP_SYN... |
1452 |
} |
322b8afe4 mm, swap: Fix a r... |
1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 |
static int swap_swapcount(struct swap_info_struct *si, swp_entry_t entry) { int count = 0; pgoff_t offset = swp_offset(entry); struct swap_cluster_info *ci; ci = lock_cluster_or_swap_info(si, offset); count = swap_count(si->swap_map[offset]); unlock_cluster_or_swap_info(si, ci); return count; } |
1da177e4c Linux-2.6.12-rc2 |
1464 |
/* |
8334b9622 mm: /proc/pid/sma... |
1465 |
* How many references to @entry are currently swapped out? |
e8c26ab60 mm/swap: skip rea... |
1466 1467 1468 1469 1470 1471 |
* This does not give an exact answer when swap count is continued, * but does include the high COUNT_CONTINUED flag to allow for that. */ int __swp_swapcount(swp_entry_t entry) { int count = 0; |
e8c26ab60 mm/swap: skip rea... |
1472 |
struct swap_info_struct *si; |
e8c26ab60 mm/swap: skip rea... |
1473 |
|
eb085574a mm, swap: fix rac... |
1474 1475 |
si = get_swap_device(entry); if (si) { |
322b8afe4 mm, swap: Fix a r... |
1476 |
count = swap_swapcount(si, entry); |
eb085574a mm, swap: fix rac... |
1477 1478 |
put_swap_device(si); } |
e8c26ab60 mm/swap: skip rea... |
1479 1480 1481 1482 1483 |
return count; } /* * How many references to @entry are currently swapped out? |
8334b9622 mm: /proc/pid/sma... |
1484 1485 1486 1487 1488 1489 |
* This considers COUNT_CONTINUED so it returns exact answer. */ int swp_swapcount(swp_entry_t entry) { int count, tmp_count, n; struct swap_info_struct *p; |
235b62176 mm/swap: add clus... |
1490 |
struct swap_cluster_info *ci; |
8334b9622 mm: /proc/pid/sma... |
1491 1492 1493 |
struct page *page; pgoff_t offset; unsigned char *map; |
235b62176 mm/swap: add clus... |
1494 |
p = _swap_info_get(entry); |
8334b9622 mm: /proc/pid/sma... |
1495 1496 |
if (!p) return 0; |
235b62176 mm/swap: add clus... |
1497 1498 1499 1500 1501 |
offset = swp_offset(entry); ci = lock_cluster_or_swap_info(p, offset); count = swap_count(p->swap_map[offset]); |
8334b9622 mm: /proc/pid/sma... |
1502 1503 1504 1505 1506 |
if (!(count & COUNT_CONTINUED)) goto out; count &= ~COUNT_CONTINUED; n = SWAP_MAP_MAX + 1; |
8334b9622 mm: /proc/pid/sma... |
1507 1508 1509 1510 1511 |
page = vmalloc_to_page(p->swap_map + offset); offset &= ~PAGE_MASK; VM_BUG_ON(page_private(page) != SWP_CONTINUED); do { |
a8ae49917 mm/swapfile.c: us... |
1512 |
page = list_next_entry(page, lru); |
8334b9622 mm: /proc/pid/sma... |
1513 1514 1515 1516 1517 1518 1519 1520 |
map = kmap_atomic(page); tmp_count = map[offset]; kunmap_atomic(map); count += (tmp_count & ~COUNT_CONTINUED) * n; n *= (SWAP_CONT_MAX + 1); } while (tmp_count & COUNT_CONTINUED); out: |
235b62176 mm/swap: add clus... |
1521 |
unlock_cluster_or_swap_info(p, ci); |
8334b9622 mm: /proc/pid/sma... |
1522 1523 |
return count; } |
e07098294 mm, THP, swap: su... |
1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 |
static bool swap_page_trans_huge_swapped(struct swap_info_struct *si, swp_entry_t entry) { struct swap_cluster_info *ci; unsigned char *map = si->swap_map; unsigned long roffset = swp_offset(entry); unsigned long offset = round_down(roffset, SWAPFILE_CLUSTER); int i; bool ret = false; ci = lock_cluster_or_swap_info(si, offset); if (!ci || !cluster_is_huge(ci)) { |
afa4711ef mm/swapfile.c: us... |
1536 |
if (swap_count(map[roffset])) |
e07098294 mm, THP, swap: su... |
1537 1538 1539 1540 |
ret = true; goto unlock_out; } for (i = 0; i < SWAPFILE_CLUSTER; i++) { |
afa4711ef mm/swapfile.c: us... |
1541 |
if (swap_count(map[offset + i])) { |
e07098294 mm, THP, swap: su... |
1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 |
ret = true; break; } } unlock_out: unlock_cluster_or_swap_info(si, ci); return ret; } static bool page_swapped(struct page *page) { swp_entry_t entry; struct swap_info_struct *si; |
fe5266d5d mm/swapfile.c: re... |
1555 |
if (!IS_ENABLED(CONFIG_THP_SWAP) || likely(!PageTransCompound(page))) |
e07098294 mm, THP, swap: su... |
1556 1557 1558 1559 1560 1561 1562 1563 1564 |
return page_swapcount(page) != 0; page = compound_head(page); entry.val = page_private(page); si = _swap_info_get(entry); if (si) return swap_page_trans_huge_swapped(si, entry); return false; } |
ba3c4ce6d mm, THP, swap: ma... |
1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 |
static int page_trans_huge_map_swapcount(struct page *page, int *total_mapcount, int *total_swapcount) { int i, map_swapcount, _total_mapcount, _total_swapcount; unsigned long offset = 0; struct swap_info_struct *si; struct swap_cluster_info *ci = NULL; unsigned char *map = NULL; int mapcount, swapcount = 0; /* hugetlbfs shouldn't call it */ VM_BUG_ON_PAGE(PageHuge(page), page); |
fe5266d5d mm/swapfile.c: re... |
1578 1579 |
if (!IS_ENABLED(CONFIG_THP_SWAP) || likely(!PageTransCompound(page))) { mapcount = page_trans_huge_mapcount(page, total_mapcount); |
ba3c4ce6d mm, THP, swap: ma... |
1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 |
if (PageSwapCache(page)) swapcount = page_swapcount(page); if (total_swapcount) *total_swapcount = swapcount; return mapcount + swapcount; } page = compound_head(page); _total_mapcount = _total_swapcount = map_swapcount = 0; if (PageSwapCache(page)) { swp_entry_t entry; entry.val = page_private(page); si = _swap_info_get(entry); if (si) { map = si->swap_map; offset = swp_offset(entry); } } if (map) ci = lock_cluster(si, offset); for (i = 0; i < HPAGE_PMD_NR; i++) { mapcount = atomic_read(&page[i]._mapcount) + 1; _total_mapcount += mapcount; if (map) { swapcount = swap_count(map[offset + i]); _total_swapcount += swapcount; } map_swapcount = max(map_swapcount, mapcount + swapcount); } unlock_cluster(ci); if (PageDoubleMap(page)) { map_swapcount -= 1; _total_mapcount -= HPAGE_PMD_NR; } mapcount = compound_mapcount(page); map_swapcount += mapcount; _total_mapcount += mapcount; if (total_mapcount) *total_mapcount = _total_mapcount; if (total_swapcount) *total_swapcount = _total_swapcount; return map_swapcount; } |
e07098294 mm, THP, swap: su... |
1626 |
|
8334b9622 mm: /proc/pid/sma... |
1627 |
/* |
7b1fe5979 mm: reuse_swap_pa... |
1628 1629 1630 1631 |
* 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. |
6d0a07edd mm: thp: calculat... |
1632 |
* |
ba3c4ce6d mm, THP, swap: ma... |
1633 |
* NOTE: total_map_swapcount should not be relied upon by the caller if |
6d0a07edd mm: thp: calculat... |
1634 1635 |
* reuse_swap_page() returns false, but it may be always overwritten * (see the other implementation for CONFIG_SWAP=n). |
1da177e4c Linux-2.6.12-rc2 |
1636 |
*/ |
ba3c4ce6d mm, THP, swap: ma... |
1637 |
bool reuse_swap_page(struct page *page, int *total_map_swapcount) |
1da177e4c Linux-2.6.12-rc2 |
1638 |
{ |
ba3c4ce6d mm, THP, swap: ma... |
1639 |
int count, total_mapcount, total_swapcount; |
c475a8ab6 [PATCH] can_share... |
1640 |
|
309381fea mm: dump page whe... |
1641 |
VM_BUG_ON_PAGE(!PageLocked(page), page); |
5ad646880 ksm: let shared p... |
1642 |
if (unlikely(PageKsm(page))) |
6d0a07edd mm: thp: calculat... |
1643 |
return false; |
ba3c4ce6d mm, THP, swap: ma... |
1644 1645 1646 1647 1648 1649 1650 1651 |
count = page_trans_huge_map_swapcount(page, &total_mapcount, &total_swapcount); if (total_map_swapcount) *total_map_swapcount = total_mapcount + total_swapcount; if (count == 1 && PageSwapCache(page) && (likely(!PageTransCompound(page)) || /* The remaining swap count will be freed soon */ total_swapcount == page_swapcount(page))) { |
f05714293 mm: support anony... |
1652 |
if (!PageWriteback(page)) { |
ba3c4ce6d mm, THP, swap: ma... |
1653 |
page = compound_head(page); |
7b1fe5979 mm: reuse_swap_pa... |
1654 1655 |
delete_from_swap_cache(page); SetPageDirty(page); |
f05714293 mm: support anony... |
1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 |
} else { swp_entry_t entry; struct swap_info_struct *p; entry.val = page_private(page); p = swap_info_get(entry); if (p->flags & SWP_STABLE_WRITES) { spin_unlock(&p->lock); return false; } spin_unlock(&p->lock); |
7b1fe5979 mm: reuse_swap_pa... |
1667 1668 |
} } |
ba3c4ce6d mm, THP, swap: ma... |
1669 |
|
5ad646880 ksm: let shared p... |
1670 |
return count <= 1; |
1da177e4c Linux-2.6.12-rc2 |
1671 1672 1673 |
} /* |
a2c43eed8 mm: try_to_free_s... |
1674 1675 |
* 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 Linux-2.6.12-rc2 |
1676 |
*/ |
a2c43eed8 mm: try_to_free_s... |
1677 |
int try_to_free_swap(struct page *page) |
1da177e4c Linux-2.6.12-rc2 |
1678 |
{ |
309381fea mm: dump page whe... |
1679 |
VM_BUG_ON_PAGE(!PageLocked(page), page); |
1da177e4c Linux-2.6.12-rc2 |
1680 1681 1682 1683 1684 |
if (!PageSwapCache(page)) return 0; if (PageWriteback(page)) return 0; |
e07098294 mm, THP, swap: su... |
1685 |
if (page_swapped(page)) |
1da177e4c Linux-2.6.12-rc2 |
1686 |
return 0; |
b73d7fcec swap: prevent reu... |
1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 |
/* * Once hibernation has begun to create its image of memory, * there's a danger that one of the calls to try_to_free_swap() * - most probably a call from __try_to_reclaim_swap() while * hibernation is allocating its own swap pages for the image, * but conceivably even a call from memory reclaim - will free * the swap from a page which has already been recorded in the * image as a clean swapcache page, and then reuse its swap for * another page of the image. On waking from hibernation, the * original page might be freed under memory pressure, then * later read back in from swap, now with the wrong data. * |
2de1a7e40 mm/swapfile.c: fi... |
1699 |
* Hibernation suspends storage while it is writing the image |
f90ac3982 mm: avoid liveloc... |
1700 |
* to disk so check that here. |
b73d7fcec swap: prevent reu... |
1701 |
*/ |
f90ac3982 mm: avoid liveloc... |
1702 |
if (pm_suspended_storage()) |
b73d7fcec swap: prevent reu... |
1703 |
return 0; |
e07098294 mm, THP, swap: su... |
1704 |
page = compound_head(page); |
a2c43eed8 mm: try_to_free_s... |
1705 1706 1707 |
delete_from_swap_cache(page); SetPageDirty(page); return 1; |
68a22394c vmscan: free swap... |
1708 1709 1710 |
} /* |
1da177e4c Linux-2.6.12-rc2 |
1711 1712 1713 |
* Free the swap entry like above, but also try to * free the page cache entry if it is the last user. */ |
2509ef26d badpage: zap prin... |
1714 |
int free_swap_and_cache(swp_entry_t entry) |
1da177e4c Linux-2.6.12-rc2 |
1715 |
{ |
2509ef26d badpage: zap prin... |
1716 |
struct swap_info_struct *p; |
7c00bafee mm/swap: free swa... |
1717 |
unsigned char count; |
1da177e4c Linux-2.6.12-rc2 |
1718 |
|
a7420aa54 HWPOISON: Add sup... |
1719 |
if (non_swap_entry(entry)) |
2509ef26d badpage: zap prin... |
1720 |
return 1; |
0697212a4 [PATCH] Swapless ... |
1721 |
|
7c00bafee mm/swap: free swa... |
1722 |
p = _swap_info_get(entry); |
1da177e4c Linux-2.6.12-rc2 |
1723 |
if (p) { |
33e16272f mm/swapfile.c: __... |
1724 |
count = __swap_entry_free(p, entry); |
e07098294 mm, THP, swap: su... |
1725 |
if (count == SWAP_HAS_CACHE && |
bcd49e867 mm/swapfile.c: us... |
1726 1727 1728 |
!swap_page_trans_huge_swapped(p, entry)) __try_to_reclaim_swap(p, swp_offset(entry), TTRS_UNMAPPED | TTRS_FULL); |
1da177e4c Linux-2.6.12-rc2 |
1729 |
} |
2509ef26d badpage: zap prin... |
1730 |
return p != NULL; |
1da177e4c Linux-2.6.12-rc2 |
1731 |
} |
b0cb1a19d Replace CONFIG_SO... |
1732 |
#ifdef CONFIG_HIBERNATION |
f577eb30a [PATCH] swsusp: l... |
1733 |
/* |
915bae9eb [PATCH] swsusp: u... |
1734 |
* Find the swap type that corresponds to given device (if any). |
f577eb30a [PATCH] swsusp: l... |
1735 |
* |
915bae9eb [PATCH] swsusp: u... |
1736 1737 1738 1739 |
* @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 [PATCH] swsusp: l... |
1740 |
*/ |
21bd90057 mm: split swap_ty... |
1741 |
int swap_type_of(dev_t device, sector_t offset) |
f577eb30a [PATCH] swsusp: l... |
1742 |
{ |
efa90a981 swap_info: change... |
1743 |
int type; |
f577eb30a [PATCH] swsusp: l... |
1744 |
|
21bd90057 mm: split swap_ty... |
1745 1746 |
if (!device) return -1; |
915bae9eb [PATCH] swsusp: u... |
1747 |
|
f577eb30a [PATCH] swsusp: l... |
1748 |
spin_lock(&swap_lock); |
efa90a981 swap_info: change... |
1749 1750 |
for (type = 0; type < nr_swapfiles; type++) { struct swap_info_struct *sis = swap_info[type]; |
f577eb30a [PATCH] swsusp: l... |
1751 |
|
915bae9eb [PATCH] swsusp: u... |
1752 |
if (!(sis->flags & SWP_WRITEOK)) |
f577eb30a [PATCH] swsusp: l... |
1753 |
continue; |
b6b5bce35 [PATCH] swsusp: F... |
1754 |
|
21bd90057 mm: split swap_ty... |
1755 |
if (device == sis->bdev->bd_dev) { |
4efaceb1c mm, swap: use rbt... |
1756 |
struct swap_extent *se = first_se(sis); |
915bae9eb [PATCH] swsusp: u... |
1757 |
|
915bae9eb [PATCH] swsusp: u... |
1758 1759 |
if (se->start_block == offset) { spin_unlock(&swap_lock); |
efa90a981 swap_info: change... |
1760 |
return type; |
915bae9eb [PATCH] swsusp: u... |
1761 |
} |
f577eb30a [PATCH] swsusp: l... |
1762 1763 1764 |
} } spin_unlock(&swap_lock); |
21bd90057 mm: split swap_ty... |
1765 1766 |
return -ENODEV; } |
915bae9eb [PATCH] swsusp: u... |
1767 |
|
21bd90057 mm: split swap_ty... |
1768 1769 1770 |
int find_first_swap(dev_t *device) { int type; |
915bae9eb [PATCH] swsusp: u... |
1771 |
|
21bd90057 mm: split swap_ty... |
1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 |
spin_lock(&swap_lock); for (type = 0; type < nr_swapfiles; type++) { struct swap_info_struct *sis = swap_info[type]; if (!(sis->flags & SWP_WRITEOK)) continue; *device = sis->bdev->bd_dev; spin_unlock(&swap_lock); return type; } spin_unlock(&swap_lock); |
f577eb30a [PATCH] swsusp: l... |
1783 1784 1785 1786 |
return -ENODEV; } /* |
73c34b6ac swap_info: miscel... |
1787 1788 1789 1790 1791 1792 |
* 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; |
c10d38cc8 mm, swap: bounds ... |
1793 |
struct swap_info_struct *si = swap_type_to_swap_info(type); |
73c34b6ac swap_info: miscel... |
1794 |
|
c10d38cc8 mm, swap: bounds ... |
1795 |
if (!si || !(si->flags & SWP_WRITEOK)) |
73c34b6ac swap_info: miscel... |
1796 |
return 0; |
d4906e1aa swap: rework map_... |
1797 |
return map_swap_entry(swp_entry(type, offset), &bdev); |
73c34b6ac swap_info: miscel... |
1798 1799 1800 |
} /* |
f577eb30a [PATCH] swsusp: l... |
1801 1802 1803 1804 1805 1806 1807 1808 |
* 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 swap_info: change... |
1809 1810 1811 |
spin_lock(&swap_lock); if ((unsigned int)type < nr_swapfiles) { struct swap_info_struct *sis = swap_info[type]; |
ec8acf20a swap: add per-par... |
1812 |
spin_lock(&sis->lock); |
efa90a981 swap_info: change... |
1813 1814 |
if (sis->flags & SWP_WRITEOK) { n = sis->pages; |
f577eb30a [PATCH] swsusp: l... |
1815 |
if (free) |
efa90a981 swap_info: change... |
1816 |
n -= sis->inuse_pages; |
f577eb30a [PATCH] swsusp: l... |
1817 |
} |
ec8acf20a swap: add per-par... |
1818 |
spin_unlock(&sis->lock); |
f577eb30a [PATCH] swsusp: l... |
1819 |
} |
efa90a981 swap_info: change... |
1820 |
spin_unlock(&swap_lock); |
f577eb30a [PATCH] swsusp: l... |
1821 1822 |
return n; } |
73c34b6ac swap_info: miscel... |
1823 |
#endif /* CONFIG_HIBERNATION */ |
f577eb30a [PATCH] swsusp: l... |
1824 |
|
9f8bdb3f3 mm: make swapoff ... |
1825 |
static inline int pte_same_as_swp(pte_t pte, pte_t swp_pte) |
179ef71cb mm: save soft-dir... |
1826 |
{ |
9f8bdb3f3 mm: make swapoff ... |
1827 |
return pte_same(pte_swp_clear_soft_dirty(pte), swp_pte); |
179ef71cb mm: save soft-dir... |
1828 |
} |
1da177e4c Linux-2.6.12-rc2 |
1829 |
/* |
72866f6f2 [PATCH] mm: anon ... |
1830 1831 1832 |
* 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 Linux-2.6.12-rc2 |
1833 |
*/ |
044d66c1d memcgroup: reinst... |
1834 |
static int unuse_pte(struct vm_area_struct *vma, pmd_t *pmd, |
1da177e4c Linux-2.6.12-rc2 |
1835 1836 |
unsigned long addr, swp_entry_t entry, struct page *page) { |
9e16b7fb1 mm,ksm: swapoff m... |
1837 |
struct page *swapcache; |
044d66c1d memcgroup: reinst... |
1838 1839 1840 |
spinlock_t *ptl; pte_t *pte; int ret = 1; |
9e16b7fb1 mm,ksm: swapoff m... |
1841 1842 1843 1844 |
swapcache = page; page = ksm_might_need_to_copy(page, vma, addr); if (unlikely(!page)) return -ENOMEM; |
044d66c1d memcgroup: reinst... |
1845 |
pte = pte_offset_map_lock(vma->vm_mm, pmd, addr, &ptl); |
9f8bdb3f3 mm: make swapoff ... |
1846 |
if (unlikely(!pte_same_as_swp(*pte, swp_entry_to_pte(entry)))) { |
044d66c1d memcgroup: reinst... |
1847 1848 1849 |
ret = 0; goto out; } |
8a9f3ccd2 Memory controller... |
1850 |
|
b084d4353 mm: count swap usage |
1851 |
dec_mm_counter(vma->vm_mm, MM_SWAPENTS); |
d559db086 mm: clean up mm_c... |
1852 |
inc_mm_counter(vma->vm_mm, MM_ANONPAGES); |
1da177e4c Linux-2.6.12-rc2 |
1853 1854 1855 |
get_page(page); set_pte_at(vma->vm_mm, addr, pte, pte_mkold(mk_pte(page, vma->vm_page_prot))); |
00501b531 mm: memcontrol: r... |
1856 |
if (page == swapcache) { |
be5d0a74c mm: memcontrol: s... |
1857 |
page_add_anon_rmap(page, vma, addr, false); |
00501b531 mm: memcontrol: r... |
1858 |
} else { /* ksm created a completely new copy */ |
be5d0a74c mm: memcontrol: s... |
1859 |
page_add_new_anon_rmap(page, vma, addr, false); |
b518154e5 mm/vmscan: protec... |
1860 |
lru_cache_add_inactive_or_unevictable(page, vma); |
00501b531 mm: memcontrol: r... |
1861 |
} |
1da177e4c Linux-2.6.12-rc2 |
1862 |
swap_free(entry); |
044d66c1d memcgroup: reinst... |
1863 1864 |
out: pte_unmap_unlock(pte, ptl); |
9e16b7fb1 mm,ksm: swapoff m... |
1865 1866 1867 1868 |
if (page != swapcache) { unlock_page(page); put_page(page); } |
044d66c1d memcgroup: reinst... |
1869 |
return ret; |
1da177e4c Linux-2.6.12-rc2 |
1870 1871 1872 |
} static int unuse_pte_range(struct vm_area_struct *vma, pmd_t *pmd, |
b56a2d8af mm: rid swapoff o... |
1873 1874 1875 |
unsigned long addr, unsigned long end, unsigned int type, bool frontswap, unsigned long *fs_pages_to_unuse) |
1da177e4c Linux-2.6.12-rc2 |
1876 |
{ |
b56a2d8af mm: rid swapoff o... |
1877 1878 |
struct page *page; swp_entry_t entry; |
705e87c0c [PATCH] mm: pte_o... |
1879 |
pte_t *pte; |
b56a2d8af mm: rid swapoff o... |
1880 1881 |
struct swap_info_struct *si; unsigned long offset; |
8a9f3ccd2 Memory controller... |
1882 |
int ret = 0; |
b56a2d8af mm: rid swapoff o... |
1883 |
volatile unsigned char *swap_map; |
1da177e4c Linux-2.6.12-rc2 |
1884 |
|
b56a2d8af mm: rid swapoff o... |
1885 |
si = swap_info[type]; |
044d66c1d memcgroup: reinst... |
1886 |
pte = pte_offset_map(pmd, addr); |
1da177e4c Linux-2.6.12-rc2 |
1887 |
do { |
b56a2d8af mm: rid swapoff o... |
1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 |
struct vm_fault vmf; if (!is_swap_pte(*pte)) continue; entry = pte_to_swp_entry(*pte); if (swp_type(entry) != type) continue; offset = swp_offset(entry); if (frontswap && !frontswap_test(si, offset)) continue; pte_unmap(pte); swap_map = &si->swap_map[offset]; |
ebc5951ee mm: swap: properl... |
1903 1904 1905 1906 1907 1908 1909 1910 |
page = lookup_swap_cache(entry, vma, addr); if (!page) { vmf.vma = vma; vmf.address = addr; vmf.pmd = pmd; page = swapin_readahead(entry, GFP_HIGHUSER_MOVABLE, &vmf); } |
b56a2d8af mm: rid swapoff o... |
1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 |
if (!page) { if (*swap_map == 0 || *swap_map == SWAP_MAP_BAD) goto try_next; return -ENOMEM; } lock_page(page); wait_on_page_writeback(page); ret = unuse_pte(vma, pmd, addr, entry, page); if (ret < 0) { unlock_page(page); put_page(page); goto out; } try_to_free_swap(page); unlock_page(page); put_page(page); if (*fs_pages_to_unuse && !--(*fs_pages_to_unuse)) { ret = FRONTSWAP_PAGES_UNUSED; goto out; |
1da177e4c Linux-2.6.12-rc2 |
1933 |
} |
b56a2d8af mm: rid swapoff o... |
1934 1935 |
try_next: pte = pte_offset_map(pmd, addr); |
1da177e4c Linux-2.6.12-rc2 |
1936 |
} while (pte++, addr += PAGE_SIZE, addr != end); |
044d66c1d memcgroup: reinst... |
1937 |
pte_unmap(pte - 1); |
b56a2d8af mm: rid swapoff o... |
1938 1939 |
ret = 0; |
044d66c1d memcgroup: reinst... |
1940 |
out: |
8a9f3ccd2 Memory controller... |
1941 |
return ret; |
1da177e4c Linux-2.6.12-rc2 |
1942 1943 1944 1945 |
} static inline int unuse_pmd_range(struct vm_area_struct *vma, pud_t *pud, unsigned long addr, unsigned long end, |
b56a2d8af mm: rid swapoff o... |
1946 1947 |
unsigned int type, bool frontswap, unsigned long *fs_pages_to_unuse) |
1da177e4c Linux-2.6.12-rc2 |
1948 1949 1950 |
{ pmd_t *pmd; unsigned long next; |
8a9f3ccd2 Memory controller... |
1951 |
int ret; |
1da177e4c Linux-2.6.12-rc2 |
1952 1953 1954 |
pmd = pmd_offset(pud, addr); do { |
dc644a073 mm: add three mor... |
1955 |
cond_resched(); |
1da177e4c Linux-2.6.12-rc2 |
1956 |
next = pmd_addr_end(addr, end); |
1a5a9906d mm: thp: fix pmd_... |
1957 |
if (pmd_none_or_trans_huge_or_clear_bad(pmd)) |
1da177e4c Linux-2.6.12-rc2 |
1958 |
continue; |
b56a2d8af mm: rid swapoff o... |
1959 1960 |
ret = unuse_pte_range(vma, pmd, addr, next, type, frontswap, fs_pages_to_unuse); |
8a9f3ccd2 Memory controller... |
1961 1962 |
if (ret) return ret; |
1da177e4c Linux-2.6.12-rc2 |
1963 1964 1965 |
} while (pmd++, addr = next, addr != end); return 0; } |
c2febafc6 mm: convert gener... |
1966 |
static inline int unuse_pud_range(struct vm_area_struct *vma, p4d_t *p4d, |
1da177e4c Linux-2.6.12-rc2 |
1967 |
unsigned long addr, unsigned long end, |
b56a2d8af mm: rid swapoff o... |
1968 1969 |
unsigned int type, bool frontswap, unsigned long *fs_pages_to_unuse) |
1da177e4c Linux-2.6.12-rc2 |
1970 1971 1972 |
{ pud_t *pud; unsigned long next; |
8a9f3ccd2 Memory controller... |
1973 |
int ret; |
1da177e4c Linux-2.6.12-rc2 |
1974 |
|
c2febafc6 mm: convert gener... |
1975 |
pud = pud_offset(p4d, addr); |
1da177e4c Linux-2.6.12-rc2 |
1976 1977 1978 1979 |
do { next = pud_addr_end(addr, end); if (pud_none_or_clear_bad(pud)) continue; |
b56a2d8af mm: rid swapoff o... |
1980 1981 |
ret = unuse_pmd_range(vma, pud, addr, next, type, frontswap, fs_pages_to_unuse); |
8a9f3ccd2 Memory controller... |
1982 1983 |
if (ret) return ret; |
1da177e4c Linux-2.6.12-rc2 |
1984 1985 1986 |
} while (pud++, addr = next, addr != end); return 0; } |
c2febafc6 mm: convert gener... |
1987 1988 |
static inline int unuse_p4d_range(struct vm_area_struct *vma, pgd_t *pgd, unsigned long addr, unsigned long end, |
b56a2d8af mm: rid swapoff o... |
1989 1990 |
unsigned int type, bool frontswap, unsigned long *fs_pages_to_unuse) |
c2febafc6 mm: convert gener... |
1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 |
{ p4d_t *p4d; unsigned long next; int ret; p4d = p4d_offset(pgd, addr); do { next = p4d_addr_end(addr, end); if (p4d_none_or_clear_bad(p4d)) continue; |
b56a2d8af mm: rid swapoff o... |
2001 2002 |
ret = unuse_pud_range(vma, p4d, addr, next, type, frontswap, fs_pages_to_unuse); |
c2febafc6 mm: convert gener... |
2003 2004 2005 2006 2007 |
if (ret) return ret; } while (p4d++, addr = next, addr != end); return 0; } |
b56a2d8af mm: rid swapoff o... |
2008 2009 |
static int unuse_vma(struct vm_area_struct *vma, unsigned int type, bool frontswap, unsigned long *fs_pages_to_unuse) |
1da177e4c Linux-2.6.12-rc2 |
2010 2011 2012 |
{ pgd_t *pgd; unsigned long addr, end, next; |
8a9f3ccd2 Memory controller... |
2013 |
int ret; |
1da177e4c Linux-2.6.12-rc2 |
2014 |
|
b56a2d8af mm: rid swapoff o... |
2015 2016 |
addr = vma->vm_start; end = vma->vm_end; |
1da177e4c Linux-2.6.12-rc2 |
2017 2018 2019 2020 2021 2022 |
pgd = pgd_offset(vma->vm_mm, addr); do { next = pgd_addr_end(addr, end); if (pgd_none_or_clear_bad(pgd)) continue; |
b56a2d8af mm: rid swapoff o... |
2023 2024 |
ret = unuse_p4d_range(vma, pgd, addr, next, type, frontswap, fs_pages_to_unuse); |
8a9f3ccd2 Memory controller... |
2025 2026 |
if (ret) return ret; |
1da177e4c Linux-2.6.12-rc2 |
2027 2028 2029 |
} while (pgd++, addr = next, addr != end); return 0; } |
b56a2d8af mm: rid swapoff o... |
2030 2031 |
static int unuse_mm(struct mm_struct *mm, unsigned int type, bool frontswap, unsigned long *fs_pages_to_unuse) |
1da177e4c Linux-2.6.12-rc2 |
2032 2033 |
{ struct vm_area_struct *vma; |
8a9f3ccd2 Memory controller... |
2034 |
int ret = 0; |
1da177e4c Linux-2.6.12-rc2 |
2035 |
|
d8ed45c5d mmap locking API:... |
2036 |
mmap_read_lock(mm); |
1da177e4c Linux-2.6.12-rc2 |
2037 |
for (vma = mm->mmap; vma; vma = vma->vm_next) { |
b56a2d8af mm: rid swapoff o... |
2038 2039 2040 2041 2042 2043 |
if (vma->anon_vma) { ret = unuse_vma(vma, type, frontswap, fs_pages_to_unuse); if (ret) break; } |
dc644a073 mm: add three mor... |
2044 |
cond_resched(); |
1da177e4c Linux-2.6.12-rc2 |
2045 |
} |
d8ed45c5d mmap locking API:... |
2046 |
mmap_read_unlock(mm); |
b56a2d8af mm: rid swapoff o... |
2047 |
return ret; |
1da177e4c Linux-2.6.12-rc2 |
2048 2049 2050 |
} /* |
38b5faf4b mm: frontswap: co... |
2051 |
* Scan swap_map (or frontswap_map if frontswap parameter is true) |
b56a2d8af mm: rid swapoff o... |
2052 2053 |
* from current position to next entry still in use. Return 0 * if there are no inuse entries after prev till end of the map. |
1da177e4c Linux-2.6.12-rc2 |
2054 |
*/ |
6eb396dc4 [PATCH] swap: swa... |
2055 |
static unsigned int find_next_to_unuse(struct swap_info_struct *si, |
38b5faf4b mm: frontswap: co... |
2056 |
unsigned int prev, bool frontswap) |
1da177e4c Linux-2.6.12-rc2 |
2057 |
{ |
b56a2d8af mm: rid swapoff o... |
2058 |
unsigned int i; |
8d69aaee8 swap_info: swap_m... |
2059 |
unsigned char count; |
1da177e4c Linux-2.6.12-rc2 |
2060 2061 |
/* |
5d337b919 [PATCH] swap: swa... |
2062 |
* No need for swap_lock here: we're just looking |
1da177e4c Linux-2.6.12-rc2 |
2063 2064 |
* for whether an entry is in use, not modifying it; false * hits are okay, and sys_swapoff() has already prevented new |
5d337b919 [PATCH] swap: swa... |
2065 |
* allocations from this area (while holding swap_lock). |
1da177e4c Linux-2.6.12-rc2 |
2066 |
*/ |
b56a2d8af mm: rid swapoff o... |
2067 |
for (i = prev + 1; i < si->max; i++) { |
4db0c3c29 mm: remove rest o... |
2068 |
count = READ_ONCE(si->swap_map[i]); |
355cfa73d mm: modify swap_m... |
2069 |
if (count && swap_count(count) != SWAP_MAP_BAD) |
dc644a073 mm: add three mor... |
2070 2071 2072 2073 |
if (!frontswap || frontswap_test(si, i)) break; if ((i % LATENCY_LIMIT) == 0) cond_resched(); |
1da177e4c Linux-2.6.12-rc2 |
2074 |
} |
b56a2d8af mm: rid swapoff o... |
2075 2076 2077 |
if (i == si->max) i = 0; |
1da177e4c Linux-2.6.12-rc2 |
2078 2079 2080 2081 |
return i; } /* |
b56a2d8af mm: rid swapoff o... |
2082 |
* If the boolean frontswap is true, only unuse pages_to_unuse pages; |
38b5faf4b mm: frontswap: co... |
2083 |
* pages_to_unuse==0 means all pages; ignored if frontswap is false |
1da177e4c Linux-2.6.12-rc2 |
2084 |
*/ |
38b5faf4b mm: frontswap: co... |
2085 2086 |
int try_to_unuse(unsigned int type, bool frontswap, unsigned long pages_to_unuse) |
1da177e4c Linux-2.6.12-rc2 |
2087 |
{ |
b56a2d8af mm: rid swapoff o... |
2088 2089 2090 2091 |
struct mm_struct *prev_mm; struct mm_struct *mm; struct list_head *p; int retval = 0; |
efa90a981 swap_info: change... |
2092 |
struct swap_info_struct *si = swap_info[type]; |
1da177e4c Linux-2.6.12-rc2 |
2093 2094 |
struct page *page; swp_entry_t entry; |
b56a2d8af mm: rid swapoff o... |
2095 |
unsigned int i; |
1da177e4c Linux-2.6.12-rc2 |
2096 |
|
218209487 mm/swapfile: fix ... |
2097 |
if (!READ_ONCE(si->inuse_pages)) |
b56a2d8af mm: rid swapoff o... |
2098 |
return 0; |
1da177e4c Linux-2.6.12-rc2 |
2099 |
|
b56a2d8af mm: rid swapoff o... |
2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 |
if (!frontswap) pages_to_unuse = 0; retry: retval = shmem_unuse(type, frontswap, &pages_to_unuse); if (retval) goto out; prev_mm = &init_mm; mmget(prev_mm); spin_lock(&mmlist_lock); p = &init_mm.mmlist; |
218209487 mm/swapfile: fix ... |
2113 |
while (READ_ONCE(si->inuse_pages) && |
64165b1af mm: swapoff: take... |
2114 2115 |
!signal_pending(current) && (p = p->next) != &init_mm.mmlist) { |
1da177e4c Linux-2.6.12-rc2 |
2116 |
|
b56a2d8af mm: rid swapoff o... |
2117 2118 2119 2120 2121 2122 2123 |
mm = list_entry(p, struct mm_struct, mmlist); if (!mmget_not_zero(mm)) continue; spin_unlock(&mmlist_lock); mmput(prev_mm); prev_mm = mm; retval = unuse_mm(mm, type, frontswap, &pages_to_unuse); |
1da177e4c Linux-2.6.12-rc2 |
2124 |
|
b56a2d8af mm: rid swapoff o... |
2125 2126 2127 |
if (retval) { mmput(prev_mm); goto out; |
1da177e4c Linux-2.6.12-rc2 |
2128 2129 2130 |
} /* |
b56a2d8af mm: rid swapoff o... |
2131 2132 |
* Make sure that we aren't completely killing * interactive performance. |
1da177e4c Linux-2.6.12-rc2 |
2133 |
*/ |
b56a2d8af mm: rid swapoff o... |
2134 2135 2136 2137 |
cond_resched(); spin_lock(&mmlist_lock); } spin_unlock(&mmlist_lock); |
1da177e4c Linux-2.6.12-rc2 |
2138 |
|
b56a2d8af mm: rid swapoff o... |
2139 |
mmput(prev_mm); |
1da177e4c Linux-2.6.12-rc2 |
2140 |
|
b56a2d8af mm: rid swapoff o... |
2141 |
i = 0; |
218209487 mm/swapfile: fix ... |
2142 |
while (READ_ONCE(si->inuse_pages) && |
64165b1af mm: swapoff: take... |
2143 2144 |
!signal_pending(current) && (i = find_next_to_unuse(si, i, frontswap)) != 0) { |
1da177e4c Linux-2.6.12-rc2 |
2145 |
|
b56a2d8af mm: rid swapoff o... |
2146 2147 2148 2149 |
entry = swp_entry(type, i); page = find_get_page(swap_address_space(entry), i); if (!page) continue; |
68bdc8d64 mm: try_to_unuse ... |
2150 2151 2152 |
/* * It is conceivable that a racing task removed this page from |
b56a2d8af mm: rid swapoff o... |
2153 2154 2155 |
* swap cache just before we acquired the page lock. The page * might even be back in swap cache on another swap area. But * that is okay, try_to_free_swap() only removes stale pages. |
1da177e4c Linux-2.6.12-rc2 |
2156 |
*/ |
b56a2d8af mm: rid swapoff o... |
2157 2158 2159 |
lock_page(page); wait_on_page_writeback(page); try_to_free_swap(page); |
1da177e4c Linux-2.6.12-rc2 |
2160 |
unlock_page(page); |
09cbfeaf1 mm, fs: get rid o... |
2161 |
put_page(page); |
1da177e4c Linux-2.6.12-rc2 |
2162 2163 |
/* |
b56a2d8af mm: rid swapoff o... |
2164 2165 2166 |
* For frontswap, we just need to unuse pages_to_unuse, if * it was specified. Need not check frontswap again here as * we already zeroed out pages_to_unuse if not frontswap. |
1da177e4c Linux-2.6.12-rc2 |
2167 |
*/ |
b56a2d8af mm: rid swapoff o... |
2168 2169 |
if (pages_to_unuse && --pages_to_unuse == 0) goto out; |
1da177e4c Linux-2.6.12-rc2 |
2170 |
} |
b56a2d8af mm: rid swapoff o... |
2171 2172 2173 2174 2175 |
/* * Lets check again to see if there are still swap entries in the map. * If yes, we would need to do retry the unuse logic again. * Under global memory pressure, swap entries can be reinserted back * into process space after the mmlist loop above passes over them. |
dd862deb1 mm: swapoff: remo... |
2176 |
* |
af53d3e9e mm: swapoff: shme... |
2177 2178 2179 2180 2181 |
* Limit the number of retries? No: when mmget_not_zero() above fails, * that mm is likely to be freeing swap from exit_mmap(), which proceeds * at its own independent pace; and even shmem_writepage() could have * been preempted after get_swap_page(), temporarily hiding that swap. * It's easy and robust (though cpu-intensive) just to keep retrying. |
b56a2d8af mm: rid swapoff o... |
2182 |
*/ |
218209487 mm/swapfile: fix ... |
2183 |
if (READ_ONCE(si->inuse_pages)) { |
64165b1af mm: swapoff: take... |
2184 2185 2186 2187 |
if (!signal_pending(current)) goto retry; retval = -EINTR; } |
b56a2d8af mm: rid swapoff o... |
2188 2189 |
out: return (retval == FRONTSWAP_PAGES_UNUSED) ? 0 : retval; |
1da177e4c Linux-2.6.12-rc2 |
2190 2191 2192 |
} /* |
5d337b919 [PATCH] swap: swa... |
2193 2194 2195 |
* 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 Linux-2.6.12-rc2 |
2196 2197 2198 2199 2200 |
* added to the mmlist just after page_duplicate - before would be racy. */ static void drain_mmlist(void) { struct list_head *p, *next; |
efa90a981 swap_info: change... |
2201 |
unsigned int type; |
1da177e4c Linux-2.6.12-rc2 |
2202 |
|
efa90a981 swap_info: change... |
2203 2204 |
for (type = 0; type < nr_swapfiles; type++) if (swap_info[type]->inuse_pages) |
1da177e4c Linux-2.6.12-rc2 |
2205 2206 2207 2208 2209 2210 2211 2212 2213 |
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 swap: rework map_... |
2214 2215 2216 |
* 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 Linux-2.6.12-rc2 |
2217 |
*/ |
d4906e1aa swap: rework map_... |
2218 |
static sector_t map_swap_entry(swp_entry_t entry, struct block_device **bdev) |
1da177e4c Linux-2.6.12-rc2 |
2219 |
{ |
f29ad6a99 swap_info: privat... |
2220 |
struct swap_info_struct *sis; |
f29ad6a99 swap_info: privat... |
2221 2222 |
struct swap_extent *se; pgoff_t offset; |
c10d38cc8 mm, swap: bounds ... |
2223 |
sis = swp_swap_info(entry); |
f29ad6a99 swap_info: privat... |
2224 2225 2226 |
*bdev = sis->bdev; offset = swp_offset(entry); |
4efaceb1c mm, swap: use rbt... |
2227 2228 |
se = offset_to_swap_extent(sis, offset); return se->start_block + (offset - se->start_page); |
1da177e4c Linux-2.6.12-rc2 |
2229 2230 2231 |
} /* |
d4906e1aa swap: rework map_... |
2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 |
* 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 Linux-2.6.12-rc2 |
2242 2243 2244 2245 |
* Free all of a swapdev's extent information */ static void destroy_swap_extents(struct swap_info_struct *sis) { |
4efaceb1c mm, swap: use rbt... |
2246 2247 2248 |
while (!RB_EMPTY_ROOT(&sis->swap_extent_root)) { struct rb_node *rb = sis->swap_extent_root.rb_node; struct swap_extent *se = rb_entry(rb, struct swap_extent, rb_node); |
1da177e4c Linux-2.6.12-rc2 |
2249 |
|
4efaceb1c mm, swap: use rbt... |
2250 |
rb_erase(rb, &sis->swap_extent_root); |
1da177e4c Linux-2.6.12-rc2 |
2251 2252 |
kfree(se); } |
62c230bc1 mm: add support f... |
2253 |
|
bc4ae27d8 mm: split SWP_FIL... |
2254 |
if (sis->flags & SWP_ACTIVATED) { |
62c230bc1 mm: add support f... |
2255 2256 |
struct file *swap_file = sis->swap_file; struct address_space *mapping = swap_file->f_mapping; |
bc4ae27d8 mm: split SWP_FIL... |
2257 2258 2259 |
sis->flags &= ~SWP_ACTIVATED; if (mapping->a_ops->swap_deactivate) mapping->a_ops->swap_deactivate(swap_file); |
62c230bc1 mm: add support f... |
2260 |
} |
1da177e4c Linux-2.6.12-rc2 |
2261 2262 2263 2264 |
} /* * Add a block range (and the corresponding page range) into this swapdev's |
4efaceb1c mm, swap: use rbt... |
2265 |
* extent tree. |
1da177e4c Linux-2.6.12-rc2 |
2266 |
* |
11d31886d [PATCH] swap: swa... |
2267 |
* This function rather assumes that it is called in ascending page order. |
1da177e4c Linux-2.6.12-rc2 |
2268 |
*/ |
a509bc1a9 mm: swap: impleme... |
2269 |
int |
1da177e4c Linux-2.6.12-rc2 |
2270 2271 2272 |
add_swap_extent(struct swap_info_struct *sis, unsigned long start_page, unsigned long nr_pages, sector_t start_block) { |
4efaceb1c mm, swap: use rbt... |
2273 |
struct rb_node **link = &sis->swap_extent_root.rb_node, *parent = NULL; |
1da177e4c Linux-2.6.12-rc2 |
2274 2275 |
struct swap_extent *se; struct swap_extent *new_se; |
4efaceb1c mm, swap: use rbt... |
2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 |
/* * place the new node at the right most since the * function is called in ascending page order. */ while (*link) { parent = *link; link = &parent->rb_right; } if (parent) { se = rb_entry(parent, struct swap_extent, rb_node); |
11d31886d [PATCH] swap: swa... |
2288 2289 |
BUG_ON(se->start_page + se->nr_pages != start_page); if (se->start_block + se->nr_pages == start_block) { |
1da177e4c Linux-2.6.12-rc2 |
2290 2291 2292 2293 |
/* Merge it */ se->nr_pages += nr_pages; return 0; } |
1da177e4c Linux-2.6.12-rc2 |
2294 |
} |
4efaceb1c mm, swap: use rbt... |
2295 |
/* No merge, insert a new extent. */ |
1da177e4c Linux-2.6.12-rc2 |
2296 2297 2298 2299 2300 2301 |
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; |
4efaceb1c mm, swap: use rbt... |
2302 2303 |
rb_link_node(&new_se->rb_node, parent, link); rb_insert_color(&new_se->rb_node, &sis->swap_extent_root); |
53092a740 [PATCH] swap: sho... |
2304 |
return 1; |
1da177e4c Linux-2.6.12-rc2 |
2305 |
} |
aa8aa8a33 mm: export add_sw... |
2306 |
EXPORT_SYMBOL_GPL(add_swap_extent); |
1da177e4c Linux-2.6.12-rc2 |
2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 |
/* * 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. * |
1638045c3 mm: set S_SWAPFIL... |
2328 2329 |
* For all swap devices we set S_SWAPFILE across the life of the swapon. This * prevents users from writing to the swap device, which will corrupt memory. |
1da177e4c Linux-2.6.12-rc2 |
2330 2331 2332 2333 2334 2335 2336 2337 |
* * 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 [PATCH] swap: sho... |
2338 |
static int setup_swap_extents(struct swap_info_struct *sis, sector_t *span) |
1da177e4c Linux-2.6.12-rc2 |
2339 |
{ |
62c230bc1 mm: add support f... |
2340 2341 2342 |
struct file *swap_file = sis->swap_file; struct address_space *mapping = swap_file->f_mapping; struct inode *inode = mapping->host; |
1da177e4c Linux-2.6.12-rc2 |
2343 |
int ret; |
1da177e4c Linux-2.6.12-rc2 |
2344 2345 |
if (S_ISBLK(inode->i_mode)) { ret = add_swap_extent(sis, 0, sis->max, 0); |
53092a740 [PATCH] swap: sho... |
2346 |
*span = sis->pages; |
a509bc1a9 mm: swap: impleme... |
2347 |
return ret; |
1da177e4c Linux-2.6.12-rc2 |
2348 |
} |
62c230bc1 mm: add support f... |
2349 |
if (mapping->a_ops->swap_activate) { |
a509bc1a9 mm: swap: impleme... |
2350 |
ret = mapping->a_ops->swap_activate(sis, swap_file, span); |
bc4ae27d8 mm: split SWP_FIL... |
2351 2352 |
if (ret >= 0) sis->flags |= SWP_ACTIVATED; |
62c230bc1 mm: add support f... |
2353 |
if (!ret) { |
326463154 swap: rename SWP_... |
2354 |
sis->flags |= SWP_FS_OPS; |
62c230bc1 mm: add support f... |
2355 2356 2357 |
ret = add_swap_extent(sis, 0, sis->max, 0); *span = sis->pages; } |
a509bc1a9 mm: swap: impleme... |
2358 |
return ret; |
62c230bc1 mm: add support f... |
2359 |
} |
a509bc1a9 mm: swap: impleme... |
2360 |
return generic_swapfile_activate(sis, swap_file, span); |
1da177e4c Linux-2.6.12-rc2 |
2361 |
} |
a2468cc9b swap: choose swap... |
2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 |
static int swap_node(struct swap_info_struct *p) { struct block_device *bdev; if (p->bdev) bdev = p->bdev; else bdev = p->swap_file->f_inode->i_sb->s_bdev; return bdev ? bdev->bd_disk->node_id : NUMA_NO_NODE; } |
eb085574a mm, swap: fix rac... |
2373 2374 2375 |
static void setup_swap_info(struct swap_info_struct *p, int prio, unsigned char *swap_map, struct swap_cluster_info *cluster_info) |
40531542e sys_swapon: separ... |
2376 |
{ |
a2468cc9b swap: choose swap... |
2377 |
int i; |
40531542e sys_swapon: separ... |
2378 2379 2380 2381 |
if (prio >= 0) p->prio = prio; else p->prio = --least_priority; |
18ab4d4ce swap: change swap... |
2382 2383 2384 2385 2386 |
/* * the plist prio is negated because plist ordering is * low-to-high, while swap ordering is high-to-low */ p->list.prio = -p->prio; |
a2468cc9b swap: choose swap... |
2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 |
for_each_node(i) { if (p->prio >= 0) p->avail_lists[i].prio = -p->prio; else { if (swap_node(p) == i) p->avail_lists[i].prio = 1; else p->avail_lists[i].prio = -p->prio; } } |
40531542e sys_swapon: separ... |
2397 |
p->swap_map = swap_map; |
2a8f94493 swap: change bloc... |
2398 |
p->cluster_info = cluster_info; |
eb085574a mm, swap: fix rac... |
2399 2400 2401 2402 2403 |
} static void _enable_swap_info(struct swap_info_struct *p) { p->flags |= SWP_WRITEOK | SWP_VALID; |
ec8acf20a swap: add per-par... |
2404 |
atomic_long_add(p->pages, &nr_swap_pages); |
40531542e sys_swapon: separ... |
2405 |
total_swap_pages += p->pages; |
adfab836f swap: change swap... |
2406 |
assert_spin_locked(&swap_lock); |
adfab836f swap: change swap... |
2407 |
/* |
18ab4d4ce swap: change swap... |
2408 2409 2410 2411 2412 2413 2414 2415 |
* both lists are plists, and thus priority ordered. * swap_active_head needs to be priority ordered for swapoff(), * which on removal of any swap_info_struct with an auto-assigned * (i.e. negative) priority increments the auto-assigned priority * of any lower-priority swap_info_structs. * swap_avail_head needs to be priority ordered for get_swap_page(), * which allocates swap pages from the highest available priority * swap_info_struct. |
adfab836f swap: change swap... |
2416 |
*/ |
18ab4d4ce swap: change swap... |
2417 |
plist_add(&p->list, &swap_active_head); |
a2468cc9b swap: choose swap... |
2418 |
add_to_avail_list(p); |
cf0cac0a0 mm: refactor rein... |
2419 2420 2421 2422 |
} static void enable_swap_info(struct swap_info_struct *p, int prio, unsigned char *swap_map, |
2a8f94493 swap: change bloc... |
2423 |
struct swap_cluster_info *cluster_info, |
cf0cac0a0 mm: refactor rein... |
2424 2425 |
unsigned long *frontswap_map) { |
4f89849da frontswap: get ri... |
2426 |
frontswap_init(p->type, frontswap_map); |
cf0cac0a0 mm: refactor rein... |
2427 |
spin_lock(&swap_lock); |
ec8acf20a swap: add per-par... |
2428 |
spin_lock(&p->lock); |
eb085574a mm, swap: fix rac... |
2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 |
setup_swap_info(p, prio, swap_map, cluster_info); spin_unlock(&p->lock); spin_unlock(&swap_lock); /* * Guarantee swap_map, cluster_info, etc. fields are valid * between get/put_swap_device() if SWP_VALID bit is set */ synchronize_rcu(); spin_lock(&swap_lock); spin_lock(&p->lock); _enable_swap_info(p); |
ec8acf20a swap: add per-par... |
2440 |
spin_unlock(&p->lock); |
cf0cac0a0 mm: refactor rein... |
2441 2442 2443 2444 2445 2446 |
spin_unlock(&swap_lock); } static void reinsert_swap_info(struct swap_info_struct *p) { spin_lock(&swap_lock); |
ec8acf20a swap: add per-par... |
2447 |
spin_lock(&p->lock); |
eb085574a mm, swap: fix rac... |
2448 2449 |
setup_swap_info(p, p->prio, p->swap_map, p->cluster_info); _enable_swap_info(p); |
ec8acf20a swap: add per-par... |
2450 |
spin_unlock(&p->lock); |
40531542e sys_swapon: separ... |
2451 2452 |
spin_unlock(&swap_lock); } |
67afa38e0 mm/swap: add cach... |
2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 |
bool has_usable_swap(void) { bool ret = true; spin_lock(&swap_lock); if (plist_head_empty(&swap_active_head)) ret = false; spin_unlock(&swap_lock); return ret; } |
c4ea37c26 [CVE-2009-0029] S... |
2463 |
SYSCALL_DEFINE1(swapoff, const char __user *, specialfile) |
1da177e4c Linux-2.6.12-rc2 |
2464 |
{ |
73c34b6ac swap_info: miscel... |
2465 |
struct swap_info_struct *p = NULL; |
8d69aaee8 swap_info: swap_m... |
2466 |
unsigned char *swap_map; |
2a8f94493 swap: change bloc... |
2467 |
struct swap_cluster_info *cluster_info; |
4f89849da frontswap: get ri... |
2468 |
unsigned long *frontswap_map; |
1da177e4c Linux-2.6.12-rc2 |
2469 2470 2471 |
struct file *swap_file, *victim; struct address_space *mapping; struct inode *inode; |
91a27b2a7 vfs: define struc... |
2472 |
struct filename *pathname; |
adfab836f swap: change swap... |
2473 |
int err, found = 0; |
5b808a230 swap: fix set_blo... |
2474 |
unsigned int old_block_size; |
886bb7e9c swapfile: remove ... |
2475 |
|
1da177e4c Linux-2.6.12-rc2 |
2476 2477 |
if (!capable(CAP_SYS_ADMIN)) return -EPERM; |
191c54244 mm: collapse secu... |
2478 |
BUG_ON(!current->mm); |
1da177e4c Linux-2.6.12-rc2 |
2479 |
pathname = getname(specialfile); |
1da177e4c Linux-2.6.12-rc2 |
2480 |
if (IS_ERR(pathname)) |
f58b59c1d swapfile: fix nam... |
2481 |
return PTR_ERR(pathname); |
1da177e4c Linux-2.6.12-rc2 |
2482 |
|
669abf4e5 vfs: make path_op... |
2483 |
victim = file_open_name(pathname, O_RDWR|O_LARGEFILE, 0); |
1da177e4c Linux-2.6.12-rc2 |
2484 2485 2486 2487 2488 |
err = PTR_ERR(victim); if (IS_ERR(victim)) goto out; mapping = victim->f_mapping; |
5d337b919 [PATCH] swap: swa... |
2489 |
spin_lock(&swap_lock); |
18ab4d4ce swap: change swap... |
2490 |
plist_for_each_entry(p, &swap_active_head, list) { |
22c6f8fdb swapfile: remove ... |
2491 |
if (p->flags & SWP_WRITEOK) { |
adfab836f swap: change swap... |
2492 2493 |
if (p->swap_file->f_mapping == mapping) { found = 1; |
1da177e4c Linux-2.6.12-rc2 |
2494 |
break; |
adfab836f swap: change swap... |
2495 |
} |
1da177e4c Linux-2.6.12-rc2 |
2496 |
} |
1da177e4c Linux-2.6.12-rc2 |
2497 |
} |
adfab836f swap: change swap... |
2498 |
if (!found) { |
1da177e4c Linux-2.6.12-rc2 |
2499 |
err = -EINVAL; |
5d337b919 [PATCH] swap: swa... |
2500 |
spin_unlock(&swap_lock); |
1da177e4c Linux-2.6.12-rc2 |
2501 2502 |
goto out_dput; } |
191c54244 mm: collapse secu... |
2503 |
if (!security_vm_enough_memory_mm(current->mm, p->pages)) |
1da177e4c Linux-2.6.12-rc2 |
2504 2505 2506 |
vm_unacct_memory(p->pages); else { err = -ENOMEM; |
5d337b919 [PATCH] swap: swa... |
2507 |
spin_unlock(&swap_lock); |
1da177e4c Linux-2.6.12-rc2 |
2508 2509 |
goto out_dput; } |
a2468cc9b swap: choose swap... |
2510 |
del_from_avail_list(p); |
ec8acf20a swap: add per-par... |
2511 |
spin_lock(&p->lock); |
78ecba081 mm: fix ever-decr... |
2512 |
if (p->prio < 0) { |
adfab836f swap: change swap... |
2513 |
struct swap_info_struct *si = p; |
a2468cc9b swap: choose swap... |
2514 |
int nid; |
adfab836f swap: change swap... |
2515 |
|
18ab4d4ce swap: change swap... |
2516 |
plist_for_each_entry_continue(si, &swap_active_head, list) { |
adfab836f swap: change swap... |
2517 |
si->prio++; |
18ab4d4ce swap: change swap... |
2518 |
si->list.prio--; |
a2468cc9b swap: choose swap... |
2519 2520 2521 2522 |
for_each_node(nid) { if (si->avail_lists[nid].prio != 1) si->avail_lists[nid].prio--; } |
adfab836f swap: change swap... |
2523 |
} |
78ecba081 mm: fix ever-decr... |
2524 2525 |
least_priority++; } |
18ab4d4ce swap: change swap... |
2526 |
plist_del(&p->list, &swap_active_head); |
ec8acf20a swap: add per-par... |
2527 |
atomic_long_sub(p->pages, &nr_swap_pages); |
1da177e4c Linux-2.6.12-rc2 |
2528 2529 |
total_swap_pages -= p->pages; p->flags &= ~SWP_WRITEOK; |
ec8acf20a swap: add per-par... |
2530 |
spin_unlock(&p->lock); |
5d337b919 [PATCH] swap: swa... |
2531 |
spin_unlock(&swap_lock); |
fb4f88dca [PATCH] swap: get... |
2532 |
|
039939a65 mm/swap: enable s... |
2533 |
disable_swap_slots_cache_lock(); |
e1e12d2f3 mm, oom: fix race... |
2534 |
set_current_oom_origin(); |
adfab836f swap: change swap... |
2535 |
err = try_to_unuse(p->type, false, 0); /* force unuse all pages */ |
e1e12d2f3 mm, oom: fix race... |
2536 |
clear_current_oom_origin(); |
1da177e4c Linux-2.6.12-rc2 |
2537 |
|
1da177e4c Linux-2.6.12-rc2 |
2538 2539 |
if (err) { /* re-insert swap space back into swap_list */ |
cf0cac0a0 mm: refactor rein... |
2540 |
reinsert_swap_info(p); |
039939a65 mm/swap: enable s... |
2541 |
reenable_swap_slots_cache_unlock(); |
1da177e4c Linux-2.6.12-rc2 |
2542 2543 |
goto out_dput; } |
52b7efdbe [PATCH] swap: sca... |
2544 |
|
039939a65 mm/swap: enable s... |
2545 |
reenable_swap_slots_cache_unlock(); |
eb085574a mm, swap: fix rac... |
2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 |
spin_lock(&swap_lock); spin_lock(&p->lock); p->flags &= ~SWP_VALID; /* mark swap device as invalid */ spin_unlock(&p->lock); spin_unlock(&swap_lock); /* * wait for swap operations protected by get/put_swap_device() * to complete */ synchronize_rcu(); |
815c2c543 swap: make swap d... |
2556 |
flush_work(&p->discard_work); |
5d337b919 [PATCH] swap: swa... |
2557 |
destroy_swap_extents(p); |
570a335b8 swap_info: swap c... |
2558 2559 |
if (p->flags & SWP_CONTINUED) free_swap_count_continuations(p); |
81a0298bd mm, swap: don't u... |
2560 2561 |
if (!p->bdev || !blk_queue_nonrot(bdev_get_queue(p->bdev))) atomic_dec(&nr_rotate_swap); |
fc0abb145 [PATCH] sem2mutex... |
2562 |
mutex_lock(&swapon_mutex); |
5d337b919 [PATCH] swap: swa... |
2563 |
spin_lock(&swap_lock); |
ec8acf20a swap: add per-par... |
2564 |
spin_lock(&p->lock); |
5d337b919 [PATCH] swap: swa... |
2565 |
drain_mmlist(); |
52b7efdbe [PATCH] swap: sca... |
2566 |
/* wait for anyone still in scan_swap_map */ |
52b7efdbe [PATCH] swap: sca... |
2567 2568 |
p->highest_bit = 0; /* cuts scans short */ while (p->flags >= SWP_SCANNING) { |
ec8acf20a swap: add per-par... |
2569 |
spin_unlock(&p->lock); |
5d337b919 [PATCH] swap: swa... |
2570 |
spin_unlock(&swap_lock); |
13e4b57f6 [PATCH] mm: fix-u... |
2571 |
schedule_timeout_uninterruptible(1); |
5d337b919 [PATCH] swap: swa... |
2572 |
spin_lock(&swap_lock); |
ec8acf20a swap: add per-par... |
2573 |
spin_lock(&p->lock); |
52b7efdbe [PATCH] swap: sca... |
2574 |
} |
52b7efdbe [PATCH] swap: sca... |
2575 |
|
1da177e4c Linux-2.6.12-rc2 |
2576 |
swap_file = p->swap_file; |
5b808a230 swap: fix set_blo... |
2577 |
old_block_size = p->old_block_size; |
1da177e4c Linux-2.6.12-rc2 |
2578 2579 2580 2581 |
p->swap_file = NULL; p->max = 0; swap_map = p->swap_map; p->swap_map = NULL; |
2a8f94493 swap: change bloc... |
2582 2583 |
cluster_info = p->cluster_info; p->cluster_info = NULL; |
4f89849da frontswap: get ri... |
2584 |
frontswap_map = frontswap_map_get(p); |
ec8acf20a swap: add per-par... |
2585 |
spin_unlock(&p->lock); |
5d337b919 [PATCH] swap: swa... |
2586 |
spin_unlock(&swap_lock); |
8a84802e2 mm: Add arch hook... |
2587 |
arch_swap_invalidate_area(p->type); |
adfab836f swap: change swap... |
2588 |
frontswap_invalidate_area(p->type); |
58e97ba6b frontswap: enable... |
2589 |
frontswap_map_set(p, NULL); |
fc0abb145 [PATCH] sem2mutex... |
2590 |
mutex_unlock(&swapon_mutex); |
ebc2a1a69 swap: make cluste... |
2591 2592 |
free_percpu(p->percpu_cluster); p->percpu_cluster = NULL; |
490705888 swap: reduce lock... |
2593 2594 |
free_percpu(p->cluster_next_cpu); p->cluster_next_cpu = NULL; |
1da177e4c Linux-2.6.12-rc2 |
2595 |
vfree(swap_map); |
54f180d3c mm, swap: use kvz... |
2596 2597 |
kvfree(cluster_info); kvfree(frontswap_map); |
2de1a7e40 mm/swapfile.c: fi... |
2598 |
/* Destroy swap account information */ |
adfab836f swap: change swap... |
2599 |
swap_cgroup_swapoff(p->type); |
4b3ef9daa mm/swap: split sw... |
2600 |
exit_swap_address_space(p->type); |
27a7faa07 memcg: swap cgrou... |
2601 |
|
1da177e4c Linux-2.6.12-rc2 |
2602 2603 2604 |
inode = mapping->host; if (S_ISBLK(inode->i_mode)) { struct block_device *bdev = I_BDEV(inode); |
1638045c3 mm: set S_SWAPFIL... |
2605 |
|
5b808a230 swap: fix set_blo... |
2606 |
set_blocksize(bdev, old_block_size); |
e525fd89d block: make blkde... |
2607 |
blkdev_put(bdev, FMODE_READ | FMODE_WRITE | FMODE_EXCL); |
1da177e4c Linux-2.6.12-rc2 |
2608 |
} |
1638045c3 mm: set S_SWAPFIL... |
2609 2610 2611 2612 |
inode_lock(inode); inode->i_flags &= ~S_SWAPFILE; inode_unlock(inode); |
1da177e4c Linux-2.6.12-rc2 |
2613 |
filp_close(swap_file, NULL); |
f893ab41e mm/swap: fix race... |
2614 2615 2616 2617 2618 2619 2620 2621 2622 |
/* * Clear the SWP_USED flag after all resources are freed so that swapon * can reuse this swap_info in alloc_swap_info() safely. It is ok to * not hold p->lock after we cleared its SWP_WRITEOK. */ spin_lock(&swap_lock); p->flags = 0; spin_unlock(&swap_lock); |
1da177e4c Linux-2.6.12-rc2 |
2623 |
err = 0; |
66d7dd518 /proc/swaps: supp... |
2624 2625 |
atomic_inc(&proc_poll_event); wake_up_interruptible(&proc_poll_wait); |
1da177e4c Linux-2.6.12-rc2 |
2626 2627 2628 2629 |
out_dput: filp_close(victim, NULL); out: |
f58b59c1d swapfile: fix nam... |
2630 |
putname(pathname); |
1da177e4c Linux-2.6.12-rc2 |
2631 2632 2633 2634 |
return err; } #ifdef CONFIG_PROC_FS |
9dd957485 ipc, kernel, mm: ... |
2635 |
static __poll_t swaps_poll(struct file *file, poll_table *wait) |
66d7dd518 /proc/swaps: supp... |
2636 |
{ |
f15146380 fs: seq_file - ad... |
2637 |
struct seq_file *seq = file->private_data; |
66d7dd518 /proc/swaps: supp... |
2638 2639 |
poll_wait(file, &proc_poll_wait, wait); |
f15146380 fs: seq_file - ad... |
2640 2641 |
if (seq->poll_event != atomic_read(&proc_poll_event)) { seq->poll_event = atomic_read(&proc_poll_event); |
a9a08845e vfs: do bulk POLL... |
2642 |
return EPOLLIN | EPOLLRDNORM | EPOLLERR | EPOLLPRI; |
66d7dd518 /proc/swaps: supp... |
2643 |
} |
a9a08845e vfs: do bulk POLL... |
2644 |
return EPOLLIN | EPOLLRDNORM; |
66d7dd518 /proc/swaps: supp... |
2645 |
} |
1da177e4c Linux-2.6.12-rc2 |
2646 2647 2648 |
/* iterator */ static void *swap_start(struct seq_file *swap, loff_t *pos) { |
efa90a981 swap_info: change... |
2649 2650 |
struct swap_info_struct *si; int type; |
1da177e4c Linux-2.6.12-rc2 |
2651 |
loff_t l = *pos; |
fc0abb145 [PATCH] sem2mutex... |
2652 |
mutex_lock(&swapon_mutex); |
1da177e4c Linux-2.6.12-rc2 |
2653 |
|
881e4aabe [PATCH] Always pr... |
2654 2655 |
if (!l) return SEQ_START_TOKEN; |
c10d38cc8 mm, swap: bounds ... |
2656 |
for (type = 0; (si = swap_type_to_swap_info(type)); type++) { |
efa90a981 swap_info: change... |
2657 |
if (!(si->flags & SWP_USED) || !si->swap_map) |
1da177e4c Linux-2.6.12-rc2 |
2658 |
continue; |
881e4aabe [PATCH] Always pr... |
2659 |
if (!--l) |
efa90a981 swap_info: change... |
2660 |
return si; |
1da177e4c Linux-2.6.12-rc2 |
2661 2662 2663 2664 2665 2666 2667 |
} return NULL; } static void *swap_next(struct seq_file *swap, void *v, loff_t *pos) { |
efa90a981 swap_info: change... |
2668 2669 |
struct swap_info_struct *si = v; int type; |
1da177e4c Linux-2.6.12-rc2 |
2670 |
|
881e4aabe [PATCH] Always pr... |
2671 |
if (v == SEQ_START_TOKEN) |
efa90a981 swap_info: change... |
2672 2673 2674 |
type = 0; else type = si->type + 1; |
881e4aabe [PATCH] Always pr... |
2675 |
|
10c8d69f3 mm/swapfile.c: sw... |
2676 |
++(*pos); |
c10d38cc8 mm, swap: bounds ... |
2677 |
for (; (si = swap_type_to_swap_info(type)); type++) { |
efa90a981 swap_info: change... |
2678 |
if (!(si->flags & SWP_USED) || !si->swap_map) |
1da177e4c Linux-2.6.12-rc2 |
2679 |
continue; |
efa90a981 swap_info: change... |
2680 |
return si; |
1da177e4c Linux-2.6.12-rc2 |
2681 2682 2683 2684 2685 2686 2687 |
} return NULL; } static void swap_stop(struct seq_file *swap, void *v) { |
fc0abb145 [PATCH] sem2mutex... |
2688 |
mutex_unlock(&swapon_mutex); |
1da177e4c Linux-2.6.12-rc2 |
2689 2690 2691 2692 |
} static int swap_show(struct seq_file *swap, void *v) { |
efa90a981 swap_info: change... |
2693 |
struct swap_info_struct *si = v; |
1da177e4c Linux-2.6.12-rc2 |
2694 2695 |
struct file *file; int len; |
6f7939405 mm: swapfile: fix... |
2696 |
unsigned int bytes, inuse; |
1da177e4c Linux-2.6.12-rc2 |
2697 |
|
efa90a981 swap_info: change... |
2698 |
if (si == SEQ_START_TOKEN) { |
6f7939405 mm: swapfile: fix... |
2699 2700 |
seq_puts(swap,"Filename\t\t\t\tType\t\tSize\t\tUsed\t\tPriority "); |
881e4aabe [PATCH] Always pr... |
2701 2702 |
return 0; } |
1da177e4c Linux-2.6.12-rc2 |
2703 |
|
6f7939405 mm: swapfile: fix... |
2704 2705 |
bytes = si->pages << (PAGE_SHIFT - 10); inuse = si->inuse_pages << (PAGE_SHIFT - 10); |
efa90a981 swap_info: change... |
2706 |
file = si->swap_file; |
2726d5662 vfs: add seq_file... |
2707 2708 |
len = seq_file_path(swap, file, " \t \\"); |
6f7939405 mm: swapfile: fix... |
2709 2710 |
seq_printf(swap, "%*s%s\t%u\t%s%u\t%s%d ", |
886bb7e9c swapfile: remove ... |
2711 |
len < 40 ? 40 - len : 1, " ", |
496ad9aa8 new helper: file_... |
2712 |
S_ISBLK(file_inode(file)->i_mode) ? |
1da177e4c Linux-2.6.12-rc2 |
2713 |
"partition" : "file\t", |
6f7939405 mm: swapfile: fix... |
2714 2715 |
bytes, bytes < 10000000 ? "\t" : "", inuse, inuse < 10000000 ? "\t" : "", |
efa90a981 swap_info: change... |
2716 |
si->prio); |
1da177e4c Linux-2.6.12-rc2 |
2717 2718 |
return 0; } |
15ad7cdcf [PATCH] struct se... |
2719 |
static const struct seq_operations swaps_op = { |
1da177e4c Linux-2.6.12-rc2 |
2720 2721 2722 2723 2724 2725 2726 2727 |
.start = swap_start, .next = swap_next, .stop = swap_stop, .show = swap_show }; static int swaps_open(struct inode *inode, struct file *file) { |
f15146380 fs: seq_file - ad... |
2728 |
struct seq_file *seq; |
66d7dd518 /proc/swaps: supp... |
2729 |
int ret; |
66d7dd518 /proc/swaps: supp... |
2730 |
ret = seq_open(file, &swaps_op); |
f15146380 fs: seq_file - ad... |
2731 |
if (ret) |
66d7dd518 /proc/swaps: supp... |
2732 |
return ret; |
66d7dd518 /proc/swaps: supp... |
2733 |
|
f15146380 fs: seq_file - ad... |
2734 2735 2736 |
seq = file->private_data; seq->poll_event = atomic_read(&proc_poll_event); return 0; |
1da177e4c Linux-2.6.12-rc2 |
2737 |
} |
97a32539b proc: convert eve... |
2738 |
static const struct proc_ops swaps_proc_ops = { |
d919b33da proc: faster open... |
2739 |
.proc_flags = PROC_ENTRY_PERMANENT, |
97a32539b proc: convert eve... |
2740 2741 2742 2743 2744 |
.proc_open = swaps_open, .proc_read = seq_read, .proc_lseek = seq_lseek, .proc_release = seq_release, .proc_poll = swaps_poll, |
1da177e4c Linux-2.6.12-rc2 |
2745 2746 2747 2748 |
}; static int __init procswaps_init(void) { |
97a32539b proc: convert eve... |
2749 |
proc_create("swaps", 0, NULL, &swaps_proc_ops); |
1da177e4c Linux-2.6.12-rc2 |
2750 2751 2752 2753 |
return 0; } __initcall(procswaps_init); #endif /* CONFIG_PROC_FS */ |
1796316a8 x86: consolidate ... |
2754 2755 2756 2757 2758 2759 2760 2761 |
#ifdef MAX_SWAPFILES_CHECK static int __init max_swapfiles_check(void) { MAX_SWAPFILES_CHECK(); return 0; } late_initcall(max_swapfiles_check); #endif |
53cbb2435 sys_swapon: separ... |
2762 |
static struct swap_info_struct *alloc_swap_info(void) |
1da177e4c Linux-2.6.12-rc2 |
2763 |
{ |
73c34b6ac swap_info: miscel... |
2764 |
struct swap_info_struct *p; |
b11a76b37 mm/swapfile: do n... |
2765 |
struct swap_info_struct *defer = NULL; |
1da177e4c Linux-2.6.12-rc2 |
2766 |
unsigned int type; |
a2468cc9b swap: choose swap... |
2767 |
int i; |
efa90a981 swap_info: change... |
2768 |
|
960087445 mm/swapfile.c: us... |
2769 |
p = kvzalloc(struct_size(p, avail_lists, nr_node_ids), GFP_KERNEL); |
efa90a981 swap_info: change... |
2770 |
if (!p) |
53cbb2435 sys_swapon: separ... |
2771 |
return ERR_PTR(-ENOMEM); |
efa90a981 swap_info: change... |
2772 |
|
5d337b919 [PATCH] swap: swa... |
2773 |
spin_lock(&swap_lock); |
efa90a981 swap_info: change... |
2774 2775 |
for (type = 0; type < nr_swapfiles; type++) { if (!(swap_info[type]->flags & SWP_USED)) |
1da177e4c Linux-2.6.12-rc2 |
2776 |
break; |
efa90a981 swap_info: change... |
2777 |
} |
0697212a4 [PATCH] Swapless ... |
2778 |
if (type >= MAX_SWAPFILES) { |
5d337b919 [PATCH] swap: swa... |
2779 |
spin_unlock(&swap_lock); |
873d7bcfd mm/swapfile.c: us... |
2780 |
kvfree(p); |
730c0581c sys_swapon: simpl... |
2781 |
return ERR_PTR(-EPERM); |
1da177e4c Linux-2.6.12-rc2 |
2782 |
} |
efa90a981 swap_info: change... |
2783 2784 |
if (type >= nr_swapfiles) { p->type = type; |
c10d38cc8 mm, swap: bounds ... |
2785 |
WRITE_ONCE(swap_info[type], p); |
efa90a981 swap_info: change... |
2786 2787 2788 2789 2790 2791 |
/* * 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(); |
c10d38cc8 mm, swap: bounds ... |
2792 |
WRITE_ONCE(nr_swapfiles, nr_swapfiles + 1); |
efa90a981 swap_info: change... |
2793 |
} else { |
b11a76b37 mm/swapfile: do n... |
2794 |
defer = p; |
efa90a981 swap_info: change... |
2795 2796 2797 2798 2799 2800 |
p = swap_info[type]; /* * Do not memset this entry: a racing procfs swap_next() * would be relying on p->type to remain valid. */ } |
4efaceb1c mm, swap: use rbt... |
2801 |
p->swap_extent_root = RB_ROOT; |
18ab4d4ce swap: change swap... |
2802 |
plist_node_init(&p->list, 0); |
a2468cc9b swap: choose swap... |
2803 2804 |
for_each_node(i) plist_node_init(&p->avail_lists[i], 0); |
1da177e4c Linux-2.6.12-rc2 |
2805 |
p->flags = SWP_USED; |
5d337b919 [PATCH] swap: swa... |
2806 |
spin_unlock(&swap_lock); |
b11a76b37 mm/swapfile: do n... |
2807 |
kvfree(defer); |
ec8acf20a swap: add per-par... |
2808 |
spin_lock_init(&p->lock); |
2628bd6fc mm, swap: fix rac... |
2809 |
spin_lock_init(&p->cont_lock); |
efa90a981 swap_info: change... |
2810 |
|
53cbb2435 sys_swapon: separ... |
2811 |
return p; |
53cbb2435 sys_swapon: separ... |
2812 |
} |
4d0e1e107 sys_swapon: separ... |
2813 2814 2815 2816 2817 |
static int claim_swapfile(struct swap_info_struct *p, struct inode *inode) { int error; if (S_ISBLK(inode->i_mode)) { |
ef16e1d98 mm: cleanup claim... |
2818 |
p->bdev = blkdev_get_by_dev(inode->i_rdev, |
6f179af88 mm: fix potential... |
2819 |
FMODE_READ | FMODE_WRITE | FMODE_EXCL, p); |
ef16e1d98 mm: cleanup claim... |
2820 2821 |
if (IS_ERR(p->bdev)) { error = PTR_ERR(p->bdev); |
4d0e1e107 sys_swapon: separ... |
2822 |
p->bdev = NULL; |
6f179af88 mm: fix potential... |
2823 |
return error; |
4d0e1e107 sys_swapon: separ... |
2824 2825 2826 2827 |
} p->old_block_size = block_size(p->bdev); error = set_blocksize(p->bdev, PAGE_SIZE); if (error < 0) |
87ade72a7 sys_swapon: simpl... |
2828 |
return error; |
12d2966d8 mm, swap: disallo... |
2829 2830 2831 2832 2833 |
/* * Zoned block devices contain zones that have a sequential * write only restriction. Hence zoned block devices are not * suitable for swapping. Disallow them here. */ |
e556f6ba1 block: remove the... |
2834 |
if (blk_queue_is_zoned(p->bdev->bd_disk->queue)) |
12d2966d8 mm, swap: disallo... |
2835 |
return -EINVAL; |
4d0e1e107 sys_swapon: separ... |
2836 2837 2838 |
p->flags |= SWP_BLKDEV; } else if (S_ISREG(inode->i_mode)) { p->bdev = inode->i_sb->s_bdev; |
1638045c3 mm: set S_SWAPFIL... |
2839 |
} |
4d0e1e107 sys_swapon: separ... |
2840 |
return 0; |
4d0e1e107 sys_swapon: separ... |
2841 |
} |
377eeaa8e x86/speculation/l... |
2842 2843 2844 2845 2846 2847 2848 2849 2850 2851 2852 2853 2854 2855 2856 2857 2858 2859 2860 2861 2862 2863 2864 2865 2866 2867 2868 2869 |
/* * 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 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. */ unsigned long generic_max_swapfile_size(void) { return swp_offset(pte_to_swp_entry( swp_entry_to_pte(swp_entry(0, ~0UL)))) + 1; } /* Can be overridden by an architecture for additional checks. */ __weak unsigned long max_swapfile_size(void) { return generic_max_swapfile_size(); } |
ca8bd38bf sys_swapon: separ... |
2870 2871 2872 2873 2874 2875 2876 |
static unsigned long read_swap_header(struct swap_info_struct *p, union swap_header *swap_header, struct inode *inode) { int i; unsigned long maxpages; unsigned long swapfilepages; |
d6bbbd29b swap: warn when a... |
2877 |
unsigned long last_page; |
ca8bd38bf sys_swapon: separ... |
2878 2879 |
if (memcmp("SWAPSPACE2", swap_header->magic.magic, 10)) { |
465c47fd8 mm/swapfile.c: co... |
2880 2881 |
pr_err("Unable to find swap-space signature "); |
387190253 sys_swapon: simpl... |
2882 |
return 0; |
ca8bd38bf sys_swapon: separ... |
2883 2884 2885 2886 2887 2888 2889 |
} /* 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); |
dd111be69 swapfile: fix mem... |
2890 2891 |
if (swap_header->info.nr_badpages > MAX_SWAP_BADPAGES) return 0; |
ca8bd38bf sys_swapon: separ... |
2892 2893 2894 2895 2896 |
for (i = 0; i < swap_header->info.nr_badpages; i++) swab32s(&swap_header->info.badpages[i]); } /* Check the swap header's sub-version */ if (swap_header->info.version != 1) { |
465c47fd8 mm/swapfile.c: co... |
2897 2898 2899 |
pr_warn("Unable to handle swap header version %d ", swap_header->info.version); |
387190253 sys_swapon: simpl... |
2900 |
return 0; |
ca8bd38bf sys_swapon: separ... |
2901 2902 2903 2904 2905 |
} p->lowest_bit = 1; p->cluster_next = 1; p->cluster_nr = 0; |
377eeaa8e x86/speculation/l... |
2906 |
maxpages = max_swapfile_size(); |
d6bbbd29b swap: warn when a... |
2907 |
last_page = swap_header->info.last_page; |
a06ad633a swap: divide-by-z... |
2908 2909 2910 2911 2912 |
if (!last_page) { pr_warn("Empty swap-file "); return 0; } |
d6bbbd29b swap: warn when a... |
2913 |
if (last_page > maxpages) { |
465c47fd8 mm/swapfile.c: co... |
2914 2915 |
pr_warn("Truncating oversized swap area, only using %luk out of %luk ", |
d6bbbd29b swap: warn when a... |
2916 2917 2918 2919 2920 |
maxpages << (PAGE_SHIFT - 10), last_page << (PAGE_SHIFT - 10)); } if (maxpages > last_page) { maxpages = last_page + 1; |
ca8bd38bf sys_swapon: separ... |
2921 2922 2923 2924 2925 2926 2927 |
/* 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 sys_swapon: simpl... |
2928 |
return 0; |
ca8bd38bf sys_swapon: separ... |
2929 2930 |
swapfilepages = i_size_read(inode) >> PAGE_SHIFT; if (swapfilepages && maxpages > swapfilepages) { |
465c47fd8 mm/swapfile.c: co... |
2931 2932 |
pr_warn("Swap area shorter than signature indicates "); |
387190253 sys_swapon: simpl... |
2933 |
return 0; |
ca8bd38bf sys_swapon: separ... |
2934 2935 |
} if (swap_header->info.nr_badpages && S_ISREG(inode->i_mode)) |
387190253 sys_swapon: simpl... |
2936 |
return 0; |
ca8bd38bf sys_swapon: separ... |
2937 |
if (swap_header->info.nr_badpages > MAX_SWAP_BADPAGES) |
387190253 sys_swapon: simpl... |
2938 |
return 0; |
ca8bd38bf sys_swapon: separ... |
2939 2940 |
return maxpages; |
ca8bd38bf sys_swapon: separ... |
2941 |
} |
4b3ef9daa mm/swap: split sw... |
2942 |
#define SWAP_CLUSTER_INFO_COLS \ |
235b62176 mm/swap: add clus... |
2943 |
DIV_ROUND_UP(L1_CACHE_BYTES, sizeof(struct swap_cluster_info)) |
4b3ef9daa mm/swap: split sw... |
2944 2945 2946 2947 |
#define SWAP_CLUSTER_SPACE_COLS \ DIV_ROUND_UP(SWAP_ADDRESS_SPACE_PAGES, SWAPFILE_CLUSTER) #define SWAP_CLUSTER_COLS \ max_t(unsigned int, SWAP_CLUSTER_INFO_COLS, SWAP_CLUSTER_SPACE_COLS) |
235b62176 mm/swap: add clus... |
2948 |
|
915d4d7bc sys_swapon: separ... |
2949 2950 2951 |
static int setup_swap_map_and_extents(struct swap_info_struct *p, union swap_header *swap_header, unsigned char *swap_map, |
2a8f94493 swap: change bloc... |
2952 |
struct swap_cluster_info *cluster_info, |
915d4d7bc sys_swapon: separ... |
2953 2954 2955 |
unsigned long maxpages, sector_t *span) { |
235b62176 mm/swap: add clus... |
2956 |
unsigned int j, k; |
915d4d7bc sys_swapon: separ... |
2957 2958 |
unsigned int nr_good_pages; int nr_extents; |
2a8f94493 swap: change bloc... |
2959 |
unsigned long nr_clusters = DIV_ROUND_UP(maxpages, SWAPFILE_CLUSTER); |
235b62176 mm/swap: add clus... |
2960 2961 |
unsigned long col = p->cluster_next / SWAPFILE_CLUSTER % SWAP_CLUSTER_COLS; unsigned long i, idx; |
915d4d7bc sys_swapon: separ... |
2962 2963 |
nr_good_pages = maxpages - 1; /* omit header page */ |
6b5349159 mm, swap: add swa... |
2964 2965 |
cluster_list_init(&p->free_clusters); cluster_list_init(&p->discard_clusters); |
2a8f94493 swap: change bloc... |
2966 |
|
915d4d7bc sys_swapon: separ... |
2967 2968 |
for (i = 0; i < swap_header->info.nr_badpages; i++) { unsigned int page_nr = swap_header->info.badpages[i]; |
bdb8e3f68 sys_swapon: simpl... |
2969 2970 |
if (page_nr == 0 || page_nr > swap_header->info.last_page) return -EINVAL; |
915d4d7bc sys_swapon: separ... |
2971 2972 2973 |
if (page_nr < maxpages) { swap_map[page_nr] = SWAP_MAP_BAD; nr_good_pages--; |
2a8f94493 swap: change bloc... |
2974 2975 2976 2977 2978 |
/* * Haven't marked the cluster free yet, no list * operation involved */ inc_cluster_info_page(p, cluster_info, page_nr); |
915d4d7bc sys_swapon: separ... |
2979 2980 |
} } |
2a8f94493 swap: change bloc... |
2981 2982 2983 |
/* Haven't marked the cluster free yet, no list operation involved */ for (i = maxpages; i < round_up(maxpages, SWAPFILE_CLUSTER); i++) inc_cluster_info_page(p, cluster_info, i); |
915d4d7bc sys_swapon: separ... |
2984 2985 |
if (nr_good_pages) { swap_map[0] = SWAP_MAP_BAD; |
2a8f94493 swap: change bloc... |
2986 2987 2988 2989 2990 |
/* * Not mark the cluster free yet, no list * operation involved */ inc_cluster_info_page(p, cluster_info, 0); |
915d4d7bc sys_swapon: separ... |
2991 2992 2993 |
p->max = maxpages; p->pages = nr_good_pages; nr_extents = setup_swap_extents(p, span); |
bdb8e3f68 sys_swapon: simpl... |
2994 2995 |
if (nr_extents < 0) return nr_extents; |
915d4d7bc sys_swapon: separ... |
2996 2997 2998 |
nr_good_pages = p->pages; } if (!nr_good_pages) { |
465c47fd8 mm/swapfile.c: co... |
2999 3000 |
pr_warn("Empty swap-file "); |
bdb8e3f68 sys_swapon: simpl... |
3001 |
return -EINVAL; |
915d4d7bc sys_swapon: separ... |
3002 |
} |
2a8f94493 swap: change bloc... |
3003 3004 |
if (!cluster_info) return nr_extents; |
235b62176 mm/swap: add clus... |
3005 |
|
4b3ef9daa mm/swap: split sw... |
3006 3007 3008 3009 |
/* * Reduce false cache line sharing between cluster_info and * sharing same address space. */ |
235b62176 mm/swap: add clus... |
3010 3011 3012 3013 3014 3015 3016 3017 |
for (k = 0; k < SWAP_CLUSTER_COLS; k++) { j = (k + col) % SWAP_CLUSTER_COLS; for (i = 0; i < DIV_ROUND_UP(nr_clusters, SWAP_CLUSTER_COLS); i++) { idx = i * SWAP_CLUSTER_COLS + j; if (idx >= nr_clusters) continue; if (cluster_count(&cluster_info[idx])) continue; |
2a8f94493 swap: change bloc... |
3018 |
cluster_set_flag(&cluster_info[idx], CLUSTER_FLAG_FREE); |
6b5349159 mm, swap: add swa... |
3019 3020 |
cluster_list_add_tail(&p->free_clusters, cluster_info, idx); |
2a8f94493 swap: change bloc... |
3021 |
} |
2a8f94493 swap: change bloc... |
3022 |
} |
915d4d7bc sys_swapon: separ... |
3023 |
return nr_extents; |
915d4d7bc sys_swapon: separ... |
3024 |
} |
dcf6b7ddd swap: discard whi... |
3025 3026 3027 3028 3029 3030 3031 3032 3033 3034 3035 3036 3037 |
/* * Helper to sys_swapon determining if a given swap * backing device queue supports DISCARD operations. */ static bool swap_discardable(struct swap_info_struct *si) { struct request_queue *q = bdev_get_queue(si->bdev); if (!q || !blk_queue_discard(q)) return false; return true; } |
53cbb2435 sys_swapon: separ... |
3038 3039 3040 |
SYSCALL_DEFINE2(swapon, const char __user *, specialfile, int, swap_flags) { struct swap_info_struct *p; |
91a27b2a7 vfs: define struc... |
3041 |
struct filename *name; |
53cbb2435 sys_swapon: separ... |
3042 3043 |
struct file *swap_file = NULL; struct address_space *mapping; |
40531542e sys_swapon: separ... |
3044 |
int prio; |
53cbb2435 sys_swapon: separ... |
3045 3046 |
int error; union swap_header *swap_header; |
915d4d7bc sys_swapon: separ... |
3047 |
int nr_extents; |
53cbb2435 sys_swapon: separ... |
3048 3049 |
sector_t span; unsigned long maxpages; |
53cbb2435 sys_swapon: separ... |
3050 |
unsigned char *swap_map = NULL; |
2a8f94493 swap: change bloc... |
3051 |
struct swap_cluster_info *cluster_info = NULL; |
38b5faf4b mm: frontswap: co... |
3052 |
unsigned long *frontswap_map = NULL; |
53cbb2435 sys_swapon: separ... |
3053 3054 |
struct page *page = NULL; struct inode *inode = NULL; |
7cbf31923 mm: fix nr_rotate... |
3055 |
bool inced_nr_rotate_swap = false; |
53cbb2435 sys_swapon: separ... |
3056 |
|
d15cab975 swapon: check val... |
3057 3058 |
if (swap_flags & ~SWAP_FLAGS_VALID) return -EINVAL; |
53cbb2435 sys_swapon: separ... |
3059 3060 |
if (!capable(CAP_SYS_ADMIN)) return -EPERM; |
a2468cc9b swap: choose swap... |
3061 3062 |
if (!swap_avail_heads) return -ENOMEM; |
53cbb2435 sys_swapon: separ... |
3063 |
p = alloc_swap_info(); |
2542e5134 sys_swapon: simpl... |
3064 3065 |
if (IS_ERR(p)) return PTR_ERR(p); |
53cbb2435 sys_swapon: separ... |
3066 |
|
815c2c543 swap: make swap d... |
3067 |
INIT_WORK(&p->discard_work, swap_discard_work); |
1da177e4c Linux-2.6.12-rc2 |
3068 |
name = getname(specialfile); |
1da177e4c Linux-2.6.12-rc2 |
3069 |
if (IS_ERR(name)) { |
7de7fb6b3 sys_swapon: move ... |
3070 |
error = PTR_ERR(name); |
1da177e4c Linux-2.6.12-rc2 |
3071 |
name = NULL; |
bd69010b0 sys_swapon: use a... |
3072 |
goto bad_swap; |
1da177e4c Linux-2.6.12-rc2 |
3073 |
} |
669abf4e5 vfs: make path_op... |
3074 |
swap_file = file_open_name(name, O_RDWR|O_LARGEFILE, 0); |
1da177e4c Linux-2.6.12-rc2 |
3075 |
if (IS_ERR(swap_file)) { |
7de7fb6b3 sys_swapon: move ... |
3076 |
error = PTR_ERR(swap_file); |
1da177e4c Linux-2.6.12-rc2 |
3077 |
swap_file = NULL; |
bd69010b0 sys_swapon: use a... |
3078 |
goto bad_swap; |
1da177e4c Linux-2.6.12-rc2 |
3079 3080 3081 3082 |
} p->swap_file = swap_file; mapping = swap_file->f_mapping; |
2130781e2 sys_swapon: fix i... |
3083 |
inode = mapping->host; |
6f179af88 mm: fix potential... |
3084 |
|
4d0e1e107 sys_swapon: separ... |
3085 3086 |
error = claim_swapfile(p, inode); if (unlikely(error)) |
1da177e4c Linux-2.6.12-rc2 |
3087 |
goto bad_swap; |
1da177e4c Linux-2.6.12-rc2 |
3088 |
|
d795a90e2 mm/swapfile.c: mo... |
3089 3090 3091 3092 3093 |
inode_lock(inode); if (IS_SWAPFILE(inode)) { error = -EBUSY; goto bad_swap_unlock_inode; } |
1da177e4c Linux-2.6.12-rc2 |
3094 3095 3096 3097 3098 |
/* * Read the swap header. */ if (!mapping->a_ops->readpage) { error = -EINVAL; |
d795a90e2 mm/swapfile.c: mo... |
3099 |
goto bad_swap_unlock_inode; |
1da177e4c Linux-2.6.12-rc2 |
3100 |
} |
090d2b185 [PATCH] read_mapp... |
3101 |
page = read_mapping_page(mapping, 0, swap_file); |
1da177e4c Linux-2.6.12-rc2 |
3102 3103 |
if (IS_ERR(page)) { error = PTR_ERR(page); |
d795a90e2 mm/swapfile.c: mo... |
3104 |
goto bad_swap_unlock_inode; |
1da177e4c Linux-2.6.12-rc2 |
3105 |
} |
81e339712 swapfile: remove ... |
3106 |
swap_header = kmap(page); |
1da177e4c Linux-2.6.12-rc2 |
3107 |
|
ca8bd38bf sys_swapon: separ... |
3108 3109 |
maxpages = read_swap_header(p, swap_header, inode); if (unlikely(!maxpages)) { |
1da177e4c Linux-2.6.12-rc2 |
3110 |
error = -EINVAL; |
d795a90e2 mm/swapfile.c: mo... |
3111 |
goto bad_swap_unlock_inode; |
1da177e4c Linux-2.6.12-rc2 |
3112 |
} |
886bb7e9c swapfile: remove ... |
3113 |
|
81e339712 swapfile: remove ... |
3114 |
/* OK, set up the swap map and apply the bad block list */ |
803d0c835 sys_swapon: use v... |
3115 |
swap_map = vzalloc(maxpages); |
81e339712 swapfile: remove ... |
3116 3117 |
if (!swap_map) { error = -ENOMEM; |
d795a90e2 mm/swapfile.c: mo... |
3118 |
goto bad_swap_unlock_inode; |
81e339712 swapfile: remove ... |
3119 |
} |
f05714293 mm: support anony... |
3120 |
|
1cb039f3d bdi: replace BDI_... |
3121 |
if (p->bdev && blk_queue_stable_writes(p->bdev->bd_disk->queue)) |
f05714293 mm: support anony... |
3122 |
p->flags |= SWP_STABLE_WRITES; |
a8b456d01 bdi: remove BDI_C... |
3123 |
if (p->bdev && p->bdev->bd_disk->fops->rw_page) |
539a6fea7 mm, swap: introdu... |
3124 |
p->flags |= SWP_SYNCHRONOUS_IO; |
2a8f94493 swap: change bloc... |
3125 |
if (p->bdev && blk_queue_nonrot(bdev_get_queue(p->bdev))) { |
6f179af88 mm: fix potential... |
3126 |
int cpu; |
235b62176 mm/swap: add clus... |
3127 |
unsigned long ci, nr_cluster; |
6f179af88 mm: fix potential... |
3128 |
|
2a8f94493 swap: change bloc... |
3129 |
p->flags |= SWP_SOLIDSTATE; |
490705888 swap: reduce lock... |
3130 3131 3132 3133 3134 |
p->cluster_next_cpu = alloc_percpu(unsigned int); if (!p->cluster_next_cpu) { error = -ENOMEM; goto bad_swap_unlock_inode; } |
2a8f94493 swap: change bloc... |
3135 3136 3137 3138 |
/* * select a random position to start with to help wear leveling * SSD */ |
490705888 swap: reduce lock... |
3139 3140 3141 3142 |
for_each_possible_cpu(cpu) { per_cpu(*p->cluster_next_cpu, cpu) = 1 + prandom_u32_max(p->highest_bit); } |
235b62176 mm/swap: add clus... |
3143 |
nr_cluster = DIV_ROUND_UP(maxpages, SWAPFILE_CLUSTER); |
2a8f94493 swap: change bloc... |
3144 |
|
778e1cdd8 treewide: kvzallo... |
3145 |
cluster_info = kvcalloc(nr_cluster, sizeof(*cluster_info), |
54f180d3c mm, swap: use kvz... |
3146 |
GFP_KERNEL); |
2a8f94493 swap: change bloc... |
3147 3148 |
if (!cluster_info) { error = -ENOMEM; |
d795a90e2 mm/swapfile.c: mo... |
3149 |
goto bad_swap_unlock_inode; |
2a8f94493 swap: change bloc... |
3150 |
} |
235b62176 mm/swap: add clus... |
3151 3152 3153 |
for (ci = 0; ci < nr_cluster; ci++) spin_lock_init(&((cluster_info + ci)->lock)); |
ebc2a1a69 swap: make cluste... |
3154 3155 3156 |
p->percpu_cluster = alloc_percpu(struct percpu_cluster); if (!p->percpu_cluster) { error = -ENOMEM; |
d795a90e2 mm/swapfile.c: mo... |
3157 |
goto bad_swap_unlock_inode; |
ebc2a1a69 swap: make cluste... |
3158 |
} |
6f179af88 mm: fix potential... |
3159 |
for_each_possible_cpu(cpu) { |
ebc2a1a69 swap: make cluste... |
3160 |
struct percpu_cluster *cluster; |
6f179af88 mm: fix potential... |
3161 |
cluster = per_cpu_ptr(p->percpu_cluster, cpu); |
ebc2a1a69 swap: make cluste... |
3162 3163 |
cluster_set_null(&cluster->index); } |
7cbf31923 mm: fix nr_rotate... |
3164 |
} else { |
81a0298bd mm, swap: don't u... |
3165 |
atomic_inc(&nr_rotate_swap); |
7cbf31923 mm: fix nr_rotate... |
3166 3167 |
inced_nr_rotate_swap = true; } |
1da177e4c Linux-2.6.12-rc2 |
3168 |
|
1421ef3cd sys_swapon: call ... |
3169 3170 |
error = swap_cgroup_swapon(p->type, maxpages); if (error) |
d795a90e2 mm/swapfile.c: mo... |
3171 |
goto bad_swap_unlock_inode; |
1421ef3cd sys_swapon: call ... |
3172 |
|
915d4d7bc sys_swapon: separ... |
3173 |
nr_extents = setup_swap_map_and_extents(p, swap_header, swap_map, |
2a8f94493 swap: change bloc... |
3174 |
cluster_info, maxpages, &span); |
915d4d7bc sys_swapon: separ... |
3175 3176 |
if (unlikely(nr_extents < 0)) { error = nr_extents; |
d795a90e2 mm/swapfile.c: mo... |
3177 |
goto bad_swap_unlock_inode; |
1da177e4c Linux-2.6.12-rc2 |
3178 |
} |
38b5faf4b mm: frontswap: co... |
3179 |
/* frontswap enabled? set up bit-per-page map for frontswap */ |
8ea1d2a19 mm, frontswap: co... |
3180 |
if (IS_ENABLED(CONFIG_FRONTSWAP)) |
778e1cdd8 treewide: kvzallo... |
3181 3182 |
frontswap_map = kvcalloc(BITS_TO_LONGS(maxpages), sizeof(long), |
54f180d3c mm, swap: use kvz... |
3183 |
GFP_KERNEL); |
1da177e4c Linux-2.6.12-rc2 |
3184 |
|
2a8f94493 swap: change bloc... |
3185 3186 3187 3188 3189 3190 3191 3192 3193 |
if (p->bdev &&(swap_flags & SWAP_FLAG_DISCARD) && swap_discardable(p)) { /* * When discard is enabled for swap with no particular * policy flagged, we set all swap discard flags here in * order to sustain backward compatibility with older * swapon(8) releases. */ p->flags |= (SWP_DISCARDABLE | SWP_AREA_DISCARD | SWP_PAGE_DISCARD); |
dcf6b7ddd swap: discard whi... |
3194 |
|
2a8f94493 swap: change bloc... |
3195 3196 3197 3198 3199 3200 3201 3202 3203 3204 3205 3206 3207 3208 3209 3210 3211 3212 |
/* * By flagging sys_swapon, a sysadmin can tell us to * either do single-time area discards only, or to just * perform discards for released swap page-clusters. * Now it's time to adjust the p->flags accordingly. */ if (swap_flags & SWAP_FLAG_DISCARD_ONCE) p->flags &= ~SWP_PAGE_DISCARD; else if (swap_flags & SWAP_FLAG_DISCARD_PAGES) p->flags &= ~SWP_AREA_DISCARD; /* issue a swapon-time discard if it's still required */ if (p->flags & SWP_AREA_DISCARD) { int err = discard_swap(p); if (unlikely(err)) pr_err("swapon: discard_swap(%p): %d ", p, err); |
dcf6b7ddd swap: discard whi... |
3213 |
} |
20137a490 swapfile: swapon ... |
3214 |
} |
6a6ba8317 swapfile: swapon ... |
3215 |
|
4b3ef9daa mm/swap: split sw... |
3216 3217 |
error = init_swap_address_space(p->type, maxpages); if (error) |
d795a90e2 mm/swapfile.c: mo... |
3218 |
goto bad_swap_unlock_inode; |
4b3ef9daa mm/swap: split sw... |
3219 |
|
dc617f29d vfs: don't allow ... |
3220 3221 3222 3223 3224 3225 3226 3227 |
/* * Flush any pending IO and dirty mappings before we start using this * swap device. */ inode->i_flags |= S_SWAPFILE; error = inode_drain_writes(inode); if (error) { inode->i_flags &= ~S_SWAPFILE; |
822bca52e mm/swapfile.c: fi... |
3228 |
goto free_swap_address_space; |
dc617f29d vfs: don't allow ... |
3229 |
} |
fc0abb145 [PATCH] sem2mutex... |
3230 |
mutex_lock(&swapon_mutex); |
40531542e sys_swapon: separ... |
3231 |
prio = -1; |
78ecba081 mm: fix ever-decr... |
3232 |
if (swap_flags & SWAP_FLAG_PREFER) |
40531542e sys_swapon: separ... |
3233 |
prio = |
78ecba081 mm: fix ever-decr... |
3234 |
(swap_flags & SWAP_FLAG_PRIO_MASK) >> SWAP_FLAG_PRIO_SHIFT; |
2a8f94493 swap: change bloc... |
3235 |
enable_swap_info(p, prio, swap_map, cluster_info, frontswap_map); |
c69dbfb84 sys_swapon: move ... |
3236 |
|
756a025f0 mm: coalesce spli... |
3237 3238 |
pr_info("Adding %uk swap on %s. Priority:%d extents:%d across:%lluk %s%s%s%s%s ", |
91a27b2a7 vfs: define struc... |
3239 |
p->pages<<(PAGE_SHIFT-10), name->name, p->prio, |
c69dbfb84 sys_swapon: move ... |
3240 3241 |
nr_extents, (unsigned long long)span<<(PAGE_SHIFT-10), (p->flags & SWP_SOLIDSTATE) ? "SS" : "", |
38b5faf4b mm: frontswap: co... |
3242 |
(p->flags & SWP_DISCARDABLE) ? "D" : "", |
dcf6b7ddd swap: discard whi... |
3243 3244 |
(p->flags & SWP_AREA_DISCARD) ? "s" : "", (p->flags & SWP_PAGE_DISCARD) ? "c" : "", |
38b5faf4b mm: frontswap: co... |
3245 |
(frontswap_map) ? "FS" : ""); |
c69dbfb84 sys_swapon: move ... |
3246 |
|
fc0abb145 [PATCH] sem2mutex... |
3247 |
mutex_unlock(&swapon_mutex); |
66d7dd518 /proc/swaps: supp... |
3248 3249 |
atomic_inc(&proc_poll_event); wake_up_interruptible(&proc_poll_wait); |
1da177e4c Linux-2.6.12-rc2 |
3250 3251 |
error = 0; goto out; |
822bca52e mm/swapfile.c: fi... |
3252 3253 |
free_swap_address_space: exit_swap_address_space(p->type); |
d795a90e2 mm/swapfile.c: mo... |
3254 3255 |
bad_swap_unlock_inode: inode_unlock(inode); |
1da177e4c Linux-2.6.12-rc2 |
3256 |
bad_swap: |
ebc2a1a69 swap: make cluste... |
3257 3258 |
free_percpu(p->percpu_cluster); p->percpu_cluster = NULL; |
490705888 swap: reduce lock... |
3259 3260 |
free_percpu(p->cluster_next_cpu); p->cluster_next_cpu = NULL; |
bd69010b0 sys_swapon: use a... |
3261 |
if (inode && S_ISBLK(inode->i_mode) && p->bdev) { |
f2090d2df sys_swapon: remov... |
3262 3263 |
set_blocksize(p->bdev, p->old_block_size); blkdev_put(p->bdev, FMODE_READ | FMODE_WRITE | FMODE_EXCL); |
1da177e4c Linux-2.6.12-rc2 |
3264 |
} |
d795a90e2 mm/swapfile.c: mo... |
3265 |
inode = NULL; |
4cd3bb10f [PATCH] swap: mov... |
3266 |
destroy_swap_extents(p); |
e8e6c2ec4 sys_swapon: do no... |
3267 |
swap_cgroup_swapoff(p->type); |
5d337b919 [PATCH] swap: swa... |
3268 |
spin_lock(&swap_lock); |
1da177e4c Linux-2.6.12-rc2 |
3269 |
p->swap_file = NULL; |
1da177e4c Linux-2.6.12-rc2 |
3270 |
p->flags = 0; |
5d337b919 [PATCH] swap: swa... |
3271 |
spin_unlock(&swap_lock); |
1da177e4c Linux-2.6.12-rc2 |
3272 |
vfree(swap_map); |
8606a1a94 mm: kvfree the sw... |
3273 |
kvfree(cluster_info); |
b6b1fd2a6 mm/swapfile.c: fi... |
3274 |
kvfree(frontswap_map); |
7cbf31923 mm: fix nr_rotate... |
3275 3276 |
if (inced_nr_rotate_swap) atomic_dec(&nr_rotate_swap); |
d795a90e2 mm/swapfile.c: mo... |
3277 |
if (swap_file) |
1da177e4c Linux-2.6.12-rc2 |
3278 3279 3280 3281 |
filp_close(swap_file, NULL); out: if (page && !IS_ERR(page)) { kunmap(page); |
09cbfeaf1 mm, fs: get rid o... |
3282 |
put_page(page); |
1da177e4c Linux-2.6.12-rc2 |
3283 3284 3285 |
} if (name) putname(name); |
1638045c3 mm: set S_SWAPFIL... |
3286 |
if (inode) |
5955102c9 wrappers for ->i_... |
3287 |
inode_unlock(inode); |
039939a65 mm/swap: enable s... |
3288 3289 |
if (!error) enable_swap_slots_cache(); |
1da177e4c Linux-2.6.12-rc2 |
3290 3291 3292 3293 3294 |
return error; } void si_swapinfo(struct sysinfo *val) { |
efa90a981 swap_info: change... |
3295 |
unsigned int type; |
1da177e4c Linux-2.6.12-rc2 |
3296 |
unsigned long nr_to_be_unused = 0; |
5d337b919 [PATCH] swap: swa... |
3297 |
spin_lock(&swap_lock); |
efa90a981 swap_info: change... |
3298 3299 3300 3301 3302 |
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 Linux-2.6.12-rc2 |
3303 |
} |
ec8acf20a swap: add per-par... |
3304 |
val->freeswap = atomic_long_read(&nr_swap_pages) + nr_to_be_unused; |
1da177e4c Linux-2.6.12-rc2 |
3305 |
val->totalswap = total_swap_pages + nr_to_be_unused; |
5d337b919 [PATCH] swap: swa... |
3306 |
spin_unlock(&swap_lock); |
1da177e4c Linux-2.6.12-rc2 |
3307 |
} |
5e07d2eb0 ANDROID: mm: Expo... |
3308 |
EXPORT_SYMBOL_GPL(si_swapinfo); |
1da177e4c Linux-2.6.12-rc2 |
3309 3310 3311 3312 |
/* * Verify that a swap entry is valid and increment its swap map count. * |
355cfa73d mm: modify swap_m... |
3313 3314 3315 3316 3317 3318 |
* 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 swap_info: swap c... |
3319 |
* - swap-mapped reference requested but needs continued swap count. -> ENOMEM |
1da177e4c Linux-2.6.12-rc2 |
3320 |
*/ |
8d69aaee8 swap_info: swap_m... |
3321 |
static int __swap_duplicate(swp_entry_t entry, unsigned char usage) |
1da177e4c Linux-2.6.12-rc2 |
3322 |
{ |
73c34b6ac swap_info: miscel... |
3323 |
struct swap_info_struct *p; |
235b62176 mm/swap: add clus... |
3324 |
struct swap_cluster_info *ci; |
c10d38cc8 mm, swap: bounds ... |
3325 |
unsigned long offset; |
8d69aaee8 swap_info: swap_m... |
3326 3327 |
unsigned char count; unsigned char has_cache; |
253d553ba swap_info: SWAP_H... |
3328 |
int err = -EINVAL; |
1da177e4c Linux-2.6.12-rc2 |
3329 |
|
eb085574a mm, swap: fix rac... |
3330 |
p = get_swap_device(entry); |
c10d38cc8 mm, swap: bounds ... |
3331 |
if (!p) |
235b62176 mm/swap: add clus... |
3332 |
goto out; |
eb085574a mm, swap: fix rac... |
3333 |
offset = swp_offset(entry); |
235b62176 mm/swap: add clus... |
3334 |
ci = lock_cluster_or_swap_info(p, offset); |
355cfa73d mm: modify swap_m... |
3335 |
|
253d553ba swap_info: SWAP_H... |
3336 |
count = p->swap_map[offset]; |
edfe23dac swap: fix races e... |
3337 3338 3339 3340 3341 3342 3343 3344 3345 |
/* * swapin_readahead() doesn't check if a swap entry is valid, so the * swap entry could be SWAP_MAP_BAD. Check here with lock held. */ if (unlikely(swap_count(count) == SWAP_MAP_BAD)) { err = -ENOENT; goto unlock_out; } |
253d553ba swap_info: SWAP_H... |
3346 3347 3348 |
has_cache = count & SWAP_HAS_CACHE; count &= ~SWAP_HAS_CACHE; err = 0; |
355cfa73d mm: modify swap_m... |
3349 |
|
253d553ba swap_info: SWAP_H... |
3350 |
if (usage == SWAP_HAS_CACHE) { |
355cfa73d mm: modify swap_m... |
3351 3352 |
/* set SWAP_HAS_CACHE if there is no cache and entry is used */ |
253d553ba swap_info: SWAP_H... |
3353 3354 3355 3356 3357 3358 |
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 mm: modify swap_m... |
3359 3360 |
} else if (count || has_cache) { |
253d553ba swap_info: SWAP_H... |
3361 |
|
570a335b8 swap_info: swap c... |
3362 3363 3364 |
if ((count & ~COUNT_CONTINUED) < SWAP_MAP_MAX) count += usage; else if ((count & ~COUNT_CONTINUED) > SWAP_MAP_MAX) |
253d553ba swap_info: SWAP_H... |
3365 |
err = -EINVAL; |
570a335b8 swap_info: swap c... |
3366 3367 3368 3369 |
else if (swap_count_continued(p, offset, count)) count = COUNT_CONTINUED; else err = -ENOMEM; |
355cfa73d mm: modify swap_m... |
3370 |
} else |
253d553ba swap_info: SWAP_H... |
3371 |
err = -ENOENT; /* unused swap entry */ |
a449bf58e mm/swapfile: fix ... |
3372 |
WRITE_ONCE(p->swap_map[offset], count | has_cache); |
253d553ba swap_info: SWAP_H... |
3373 |
|
355cfa73d mm: modify swap_m... |
3374 |
unlock_out: |
235b62176 mm/swap: add clus... |
3375 |
unlock_cluster_or_swap_info(p, ci); |
1da177e4c Linux-2.6.12-rc2 |
3376 |
out: |
eb085574a mm, swap: fix rac... |
3377 3378 |
if (p) put_swap_device(p); |
253d553ba swap_info: SWAP_H... |
3379 |
return err; |
1da177e4c Linux-2.6.12-rc2 |
3380 |
} |
253d553ba swap_info: SWAP_H... |
3381 |
|
355cfa73d mm: modify swap_m... |
3382 |
/* |
aaa468653 swap_info: note S... |
3383 3384 3385 3386 3387 3388 3389 3390 3391 |
* 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 mm: add comment o... |
3392 3393 3394 3395 3396 |
* 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 mm: modify swap_m... |
3397 |
*/ |
570a335b8 swap_info: swap c... |
3398 |
int swap_duplicate(swp_entry_t entry) |
355cfa73d mm: modify swap_m... |
3399 |
{ |
570a335b8 swap_info: swap c... |
3400 3401 3402 3403 3404 |
int err = 0; while (!err && __swap_duplicate(entry, 1) == -ENOMEM) err = add_swap_count_continuation(entry, GFP_ATOMIC); return err; |
355cfa73d mm: modify swap_m... |
3405 |
} |
1da177e4c Linux-2.6.12-rc2 |
3406 |
|
cb4b86ba4 mm: add swap cach... |
3407 |
/* |
355cfa73d mm: modify swap_m... |
3408 3409 |
* @entry: swap entry for which we allocate swap cache. * |
73c34b6ac swap_info: miscel... |
3410 |
* Called when allocating swap cache for existing swap entry, |
355cfa73d mm: modify swap_m... |
3411 |
* This can return error codes. Returns 0 at success. |
3eeba1356 mm/swapfile.c: fi... |
3412 |
* -EEXIST means there is a swap cache. |
355cfa73d mm: modify swap_m... |
3413 |
* Note: return code is different from swap_duplicate(). |
cb4b86ba4 mm: add swap cach... |
3414 3415 3416 |
*/ int swapcache_prepare(swp_entry_t entry) { |
253d553ba swap_info: SWAP_H... |
3417 |
return __swap_duplicate(entry, SWAP_HAS_CACHE); |
cb4b86ba4 mm: add swap cach... |
3418 |
} |
0bcac06f2 mm, swap: skip sw... |
3419 3420 |
struct swap_info_struct *swp_swap_info(swp_entry_t entry) { |
c10d38cc8 mm, swap: bounds ... |
3421 |
return swap_type_to_swap_info(swp_type(entry)); |
0bcac06f2 mm, swap: skip sw... |
3422 |
} |
f981c5950 mm: methods for t... |
3423 3424 |
struct swap_info_struct *page_swap_info(struct page *page) { |
0bcac06f2 mm, swap: skip sw... |
3425 3426 |
swp_entry_t entry = { .val = page_private(page) }; return swp_swap_info(entry); |
f981c5950 mm: methods for t... |
3427 3428 3429 3430 3431 3432 3433 |
} /* * out-of-line __page_file_ methods to avoid include hell. */ struct address_space *__page_file_mapping(struct page *page) { |
f981c5950 mm: methods for t... |
3434 3435 3436 3437 3438 3439 3440 |
return page_swap_info(page)->swap_file->f_mapping; } EXPORT_SYMBOL_GPL(__page_file_mapping); pgoff_t __page_file_index(struct page *page) { swp_entry_t swap = { .val = page_private(page) }; |
f981c5950 mm: methods for t... |
3441 3442 3443 |
return swp_offset(swap); } EXPORT_SYMBOL_GPL(__page_file_index); |
1da177e4c Linux-2.6.12-rc2 |
3444 |
/* |
570a335b8 swap_info: swap c... |
3445 3446 3447 3448 3449 3450 3451 3452 3453 3454 3455 3456 3457 3458 3459 3460 3461 |
* 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; |
235b62176 mm/swap: add clus... |
3462 |
struct swap_cluster_info *ci; |
570a335b8 swap_info: swap c... |
3463 3464 3465 3466 3467 |
struct page *head; struct page *page; struct page *list_page; pgoff_t offset; unsigned char count; |
eb085574a mm, swap: fix rac... |
3468 |
int ret = 0; |
570a335b8 swap_info: swap c... |
3469 3470 3471 3472 3473 3474 |
/* * 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); |
eb085574a mm, swap: fix rac... |
3475 |
si = get_swap_device(entry); |
570a335b8 swap_info: swap c... |
3476 3477 3478 |
if (!si) { /* * An acceptable race has occurred since the failing |
eb085574a mm, swap: fix rac... |
3479 |
* __swap_duplicate(): the swap device may be swapoff |
570a335b8 swap_info: swap c... |
3480 3481 3482 |
*/ goto outer; } |
eb085574a mm, swap: fix rac... |
3483 |
spin_lock(&si->lock); |
570a335b8 swap_info: swap c... |
3484 3485 |
offset = swp_offset(entry); |
235b62176 mm/swap: add clus... |
3486 3487 |
ci = lock_cluster(si, offset); |
570a335b8 swap_info: swap c... |
3488 3489 3490 3491 3492 3493 3494 3495 3496 3497 3498 3499 |
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) { |
eb085574a mm, swap: fix rac... |
3500 3501 |
ret = -ENOMEM; goto out; |
570a335b8 swap_info: swap c... |
3502 3503 3504 3505 |
} /* * We are fortunate that although vmalloc_to_page uses pte_offset_map, |
2de1a7e40 mm/swapfile.c: fi... |
3506 3507 |
* no architecture is using highmem pages for kernel page tables: so it * will not corrupt the GFP_ATOMIC caller's atomic page table kmaps. |
570a335b8 swap_info: swap c... |
3508 3509 3510 |
*/ head = vmalloc_to_page(si->swap_map + offset); offset &= ~PAGE_MASK; |
2628bd6fc mm, swap: fix rac... |
3511 |
spin_lock(&si->cont_lock); |
570a335b8 swap_info: swap c... |
3512 3513 3514 3515 3516 3517 3518 3519 3520 3521 3522 3523 3524 3525 3526 3527 3528 3529 3530 |
/* * 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)) |
2628bd6fc mm, swap: fix rac... |
3531 |
goto out_unlock_cont; |
570a335b8 swap_info: swap c... |
3532 |
|
9b04c5fec mm: remove the se... |
3533 |
map = kmap_atomic(list_page) + offset; |
570a335b8 swap_info: swap c... |
3534 |
count = *map; |
9b04c5fec mm: remove the se... |
3535 |
kunmap_atomic(map); |
570a335b8 swap_info: swap c... |
3536 3537 3538 3539 3540 3541 |
/* * If this continuation count now has some space in it, * free our allocation and use this one. */ if ((count & ~COUNT_CONTINUED) != SWAP_CONT_MAX) |
2628bd6fc mm, swap: fix rac... |
3542 |
goto out_unlock_cont; |
570a335b8 swap_info: swap c... |
3543 3544 3545 3546 |
} list_add_tail(&page->lru, &head->lru); page = NULL; /* now it's attached, don't free it */ |
2628bd6fc mm, swap: fix rac... |
3547 3548 |
out_unlock_cont: spin_unlock(&si->cont_lock); |
570a335b8 swap_info: swap c... |
3549 |
out: |
235b62176 mm/swap: add clus... |
3550 |
unlock_cluster(ci); |
ec8acf20a swap: add per-par... |
3551 |
spin_unlock(&si->lock); |
eb085574a mm, swap: fix rac... |
3552 |
put_swap_device(si); |
570a335b8 swap_info: swap c... |
3553 3554 3555 |
outer: if (page) __free_page(page); |
eb085574a mm, swap: fix rac... |
3556 |
return ret; |
570a335b8 swap_info: swap c... |
3557 3558 3559 3560 3561 3562 3563 3564 |
} /* * 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. |
235b62176 mm/swap: add clus... |
3565 3566 |
* Called while __swap_duplicate() or swap_entry_free() holds swap or cluster * lock. |
570a335b8 swap_info: swap c... |
3567 3568 3569 3570 3571 3572 3573 |
*/ 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; |
2628bd6fc mm, swap: fix rac... |
3574 |
bool ret; |
570a335b8 swap_info: swap c... |
3575 3576 3577 3578 3579 3580 |
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 */ } |
2628bd6fc mm, swap: fix rac... |
3581 |
spin_lock(&si->cont_lock); |
570a335b8 swap_info: swap c... |
3582 |
offset &= ~PAGE_MASK; |
213516ac0 mm/swapfile: use ... |
3583 |
page = list_next_entry(head, lru); |
9b04c5fec mm: remove the se... |
3584 |
map = kmap_atomic(page) + offset; |
570a335b8 swap_info: swap c... |
3585 3586 3587 3588 3589 3590 3591 3592 3593 |
if (count == SWAP_MAP_MAX) /* initial increment from swap_map */ goto init_map; /* jump over SWAP_CONT_MAX checks */ if (count == (SWAP_MAP_MAX | COUNT_CONTINUED)) { /* incrementing */ /* * Think of how you add 1 to 999 */ while (*map == (SWAP_CONT_MAX | COUNT_CONTINUED)) { |
9b04c5fec mm: remove the se... |
3594 |
kunmap_atomic(map); |
213516ac0 mm/swapfile: use ... |
3595 |
page = list_next_entry(page, lru); |
570a335b8 swap_info: swap c... |
3596 |
BUG_ON(page == head); |
9b04c5fec mm: remove the se... |
3597 |
map = kmap_atomic(page) + offset; |
570a335b8 swap_info: swap c... |
3598 3599 |
} if (*map == SWAP_CONT_MAX) { |
9b04c5fec mm: remove the se... |
3600 |
kunmap_atomic(map); |
213516ac0 mm/swapfile: use ... |
3601 |
page = list_next_entry(page, lru); |
2628bd6fc mm, swap: fix rac... |
3602 3603 3604 3605 |
if (page == head) { ret = false; /* add count continuation */ goto out; } |
9b04c5fec mm: remove the se... |
3606 |
map = kmap_atomic(page) + offset; |
570a335b8 swap_info: swap c... |
3607 3608 3609 |
init_map: *map = 0; /* we didn't zero the page */ } *map += 1; |
9b04c5fec mm: remove the se... |
3610 |
kunmap_atomic(map); |
213516ac0 mm/swapfile: use ... |
3611 |
while ((page = list_prev_entry(page, lru)) != head) { |
9b04c5fec mm: remove the se... |
3612 |
map = kmap_atomic(page) + offset; |
570a335b8 swap_info: swap c... |
3613 |
*map = COUNT_CONTINUED; |
9b04c5fec mm: remove the se... |
3614 |
kunmap_atomic(map); |
570a335b8 swap_info: swap c... |
3615 |
} |
2628bd6fc mm, swap: fix rac... |
3616 |
ret = true; /* incremented */ |
570a335b8 swap_info: swap c... |
3617 3618 3619 3620 3621 3622 3623 |
} else { /* decrementing */ /* * Think of how you subtract 1 from 1000 */ BUG_ON(count != COUNT_CONTINUED); while (*map == COUNT_CONTINUED) { |
9b04c5fec mm: remove the se... |
3624 |
kunmap_atomic(map); |
213516ac0 mm/swapfile: use ... |
3625 |
page = list_next_entry(page, lru); |
570a335b8 swap_info: swap c... |
3626 |
BUG_ON(page == head); |
9b04c5fec mm: remove the se... |
3627 |
map = kmap_atomic(page) + offset; |
570a335b8 swap_info: swap c... |
3628 3629 3630 3631 3632 |
} BUG_ON(*map == 0); *map -= 1; if (*map == 0) count = 0; |
9b04c5fec mm: remove the se... |
3633 |
kunmap_atomic(map); |
213516ac0 mm/swapfile: use ... |
3634 |
while ((page = list_prev_entry(page, lru)) != head) { |
9b04c5fec mm: remove the se... |
3635 |
map = kmap_atomic(page) + offset; |
570a335b8 swap_info: swap c... |
3636 3637 |
*map = SWAP_CONT_MAX | count; count = COUNT_CONTINUED; |
9b04c5fec mm: remove the se... |
3638 |
kunmap_atomic(map); |
570a335b8 swap_info: swap c... |
3639 |
} |
2628bd6fc mm, swap: fix rac... |
3640 |
ret = count == COUNT_CONTINUED; |
570a335b8 swap_info: swap c... |
3641 |
} |
2628bd6fc mm, swap: fix rac... |
3642 3643 3644 |
out: spin_unlock(&si->cont_lock); return ret; |
570a335b8 swap_info: swap c... |
3645 3646 3647 3648 3649 3650 3651 3652 3653 3654 3655 3656 3657 3658 |
} /* * free_swap_count_continuations - swapoff free all the continuation pages * appended to the swap_map, after swap_map is quiesced, before vfree'ing it. */ static void free_swap_count_continuations(struct swap_info_struct *si) { pgoff_t offset; for (offset = 0; offset < si->max; offset += PAGE_SIZE) { struct page *head; head = vmalloc_to_page(si->swap_map + offset); if (page_private(head)) { |
0d576d20c mm/swapfile.c: us... |
3659 3660 3661 3662 |
struct page *page, *next; list_for_each_entry_safe(page, next, &head->lru, lru) { list_del(&page->lru); |
570a335b8 swap_info: swap c... |
3663 3664 3665 3666 3667 |
__free_page(page); } } } } |
a2468cc9b swap: choose swap... |
3668 |
|
2cf855837 memcontrol: sched... |
3669 |
#if defined(CONFIG_MEMCG) && defined(CONFIG_BLK_CGROUP) |
6caa6a070 mm: memcontrol: m... |
3670 |
void cgroup_throttle_swaprate(struct page *page, gfp_t gfp_mask) |
2cf855837 memcontrol: sched... |
3671 3672 |
{ struct swap_info_struct *si, *next; |
6caa6a070 mm: memcontrol: m... |
3673 3674 3675 |
int nid = page_to_nid(page); if (!(gfp_mask & __GFP_IO)) |
2cf855837 memcontrol: sched... |
3676 3677 3678 3679 3680 3681 3682 3683 3684 3685 3686 3687 3688 |
return; if (!blk_cgroup_congested()) return; /* * We've already scheduled a throttle, avoid taking the global swap * lock. */ if (current->throttle_queue) return; spin_lock(&swap_avail_lock); |
6caa6a070 mm: memcontrol: m... |
3689 3690 |
plist_for_each_entry_safe(si, next, &swap_avail_heads[nid], avail_lists[nid]) { |
2cf855837 memcontrol: sched... |
3691 |
if (si->bdev) { |
6caa6a070 mm: memcontrol: m... |
3692 |
blkcg_schedule_throttle(bdev_get_queue(si->bdev), true); |
2cf855837 memcontrol: sched... |
3693 3694 3695 3696 3697 3698 |
break; } } spin_unlock(&swap_avail_lock); } #endif |
a2468cc9b swap: choose swap... |
3699 3700 3701 3702 3703 3704 3705 3706 3707 3708 3709 3710 3711 3712 3713 3714 3715 3716 |
static int __init swapfile_init(void) { int nid; swap_avail_heads = kmalloc_array(nr_node_ids, sizeof(struct plist_head), GFP_KERNEL); if (!swap_avail_heads) { pr_emerg("Not enough memory for swap heads, swap is disabled "); return -ENOMEM; } for_each_node(nid) plist_head_init(&swap_avail_heads[nid]); return 0; } subsys_initcall(swapfile_init); |