Blame view
mm/swapfile.c
94.8 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 42 43 44 |
#include <asm/pgtable.h> #include <asm/tlbflush.h> #include <linux/swapops.h> |
5d1ea48bd mm: page_cgroup: ... |
45 |
#include <linux/swap_cgroup.h> |
1da177e4c Linux-2.6.12-rc2 |
46 |
|
570a335b8 swap_info: swap c... |
47 48 49 |
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_... |
50 |
static sector_t map_swap_entry(swp_entry_t, struct block_device**); |
570a335b8 swap_info: swap c... |
51 |
|
38b5faf4b mm: frontswap: co... |
52 |
DEFINE_SPINLOCK(swap_lock); |
7c363b8c6 mm/swapfile.c: ma... |
53 |
static unsigned int nr_swapfiles; |
ec8acf20a swap: add per-par... |
54 |
atomic_long_t nr_swap_pages; |
fb0fec501 mm: Export nr_swa... |
55 56 57 58 59 60 |
/* * 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... |
61 |
/* protected with swap_lock. reading in vm_swap_full() doesn't need lock */ |
1da177e4c Linux-2.6.12-rc2 |
62 |
long total_swap_pages; |
a2468cc9b swap: choose swap... |
63 |
static int least_priority = -1; |
1da177e4c Linux-2.6.12-rc2 |
64 |
|
1da177e4c Linux-2.6.12-rc2 |
65 66 67 68 |
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... |
69 70 71 72 |
/* * all active swap_info_structs * protected with swap_lock, and ordered by priority. */ |
18ab4d4ce swap: change swap... |
73 74 75 76 77 78 79 80 81 82 83 84 85 86 |
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... |
87 |
static struct plist_head *swap_avail_heads; |
18ab4d4ce swap: change swap... |
88 |
static DEFINE_SPINLOCK(swap_avail_lock); |
1da177e4c Linux-2.6.12-rc2 |
89 |
|
38b5faf4b mm: frontswap: co... |
90 |
struct swap_info_struct *swap_info[MAX_SWAPFILES]; |
1da177e4c Linux-2.6.12-rc2 |
91 |
|
fc0abb145 [PATCH] sem2mutex... |
92 |
static DEFINE_MUTEX(swapon_mutex); |
1da177e4c Linux-2.6.12-rc2 |
93 |
|
66d7dd518 /proc/swaps: supp... |
94 95 96 |
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... |
97 |
atomic_t nr_rotate_swap = ATOMIC_INIT(0); |
c10d38cc8 mm, swap: bounds ... |
98 99 100 101 102 103 104 105 |
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... |
106 |
static inline unsigned char swap_count(unsigned char ent) |
355cfa73d mm: modify swap_m... |
107 |
{ |
955c97f08 mm/swapfile.c: fi... |
108 |
return ent & ~SWAP_HAS_CACHE; /* may include COUNT_CONTINUED flag */ |
355cfa73d mm: modify swap_m... |
109 |
} |
bcd49e867 mm/swapfile.c: us... |
110 111 112 113 114 115 116 117 118 |
/* 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... |
119 |
/* returns 1 if swap entry is freed */ |
bcd49e867 mm/swapfile.c: us... |
120 121 |
static int __try_to_reclaim_swap(struct swap_info_struct *si, unsigned long offset, unsigned long flags) |
c9e444103 mm: reuse unused ... |
122 |
{ |
efa90a981 swap_info: change... |
123 |
swp_entry_t entry = swp_entry(si->type, offset); |
c9e444103 mm: reuse unused ... |
124 125 |
struct page *page; int ret = 0; |
bcd49e867 mm/swapfile.c: us... |
126 |
page = find_get_page(swap_address_space(entry), offset); |
c9e444103 mm: reuse unused ... |
127 128 129 |
if (!page) return 0; /* |
bcd49e867 mm/swapfile.c: us... |
130 131 132 |
* 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 ... |
133 134 135 136 |
* 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... |
137 138 139 140 |
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 ... |
141 142 |
unlock_page(page); } |
09cbfeaf1 mm, fs: get rid o... |
143 |
put_page(page); |
c9e444103 mm: reuse unused ... |
144 145 |
return ret; } |
355cfa73d mm: modify swap_m... |
146 |
|
4efaceb1c mm, swap: use rbt... |
147 148 149 150 151 152 153 154 155 156 157 |
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 |
158 |
/* |
6a6ba8317 swapfile: swapon ... |
159 160 161 162 163 164 |
* 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... |
165 166 |
sector_t start_block; sector_t nr_blocks; |
6a6ba8317 swapfile: swapon ... |
167 |
int err = 0; |
9625a5f28 swap_info: includ... |
168 |
/* Do not discard the swap header page! */ |
4efaceb1c mm, swap: use rbt... |
169 |
se = first_se(si); |
9625a5f28 swap_info: includ... |
170 171 172 173 |
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... |
174 |
nr_blocks, GFP_KERNEL, 0); |
9625a5f28 swap_info: includ... |
175 176 177 178 |
if (err) return err; cond_resched(); } |
6a6ba8317 swapfile: swapon ... |
179 |
|
4efaceb1c mm, swap: use rbt... |
180 |
for (se = next_se(se); se; se = next_se(se)) { |
9625a5f28 swap_info: includ... |
181 182 |
start_block = se->start_block << (PAGE_SHIFT - 9); nr_blocks = (sector_t)se->nr_pages << (PAGE_SHIFT - 9); |
6a6ba8317 swapfile: swapon ... |
183 184 |
err = blkdev_issue_discard(si->bdev, start_block, |
dd3932edd block: remove BLK... |
185 |
nr_blocks, GFP_KERNEL, 0); |
6a6ba8317 swapfile: swapon ... |
186 187 188 189 190 191 192 |
if (err) break; cond_resched(); } return err; /* That will often be -EOPNOTSUPP */ } |
4efaceb1c mm, swap: use rbt... |
193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 |
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... |
212 213 214 215 216 217 218 |
/* * 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... |
219 |
struct swap_extent *se = offset_to_swap_extent(si, start_page); |
7992fde72 swapfile: swap al... |
220 221 |
while (nr_pages) { |
4efaceb1c mm, swap: use rbt... |
222 223 224 225 226 227 228 229 230 231 232 233 234 235 |
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... |
236 |
|
4efaceb1c mm, swap: use rbt... |
237 |
se = next_se(se); |
7992fde72 swapfile: swap al... |
238 239 |
} } |
38d8b4e6b mm, THP, swap: de... |
240 241 |
#ifdef CONFIG_THP_SWAP #define SWAPFILE_CLUSTER HPAGE_PMD_NR |
a448f2d07 mm/swapfile.c: un... |
242 243 |
#define swap_entry_size(size) (size) |
38d8b4e6b mm, THP, swap: de... |
244 |
#else |
048c27fd7 [PATCH] swap: sca... |
245 |
#define SWAPFILE_CLUSTER 256 |
a448f2d07 mm/swapfile.c: un... |
246 247 248 249 250 251 |
/* * 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... |
252 |
#endif |
048c27fd7 [PATCH] swap: sca... |
253 |
#define LATENCY_LIMIT 256 |
2a8f94493 swap: change bloc... |
254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 |
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... |
311 312 |
static inline bool cluster_is_huge(struct swap_cluster_info *info) { |
33ee011e5 mm/swapfile.c: un... |
313 314 315 |
if (IS_ENABLED(CONFIG_THP_SWAP)) return info->flags & CLUSTER_FLAG_HUGE; return false; |
e07098294 mm, THP, swap: su... |
316 317 318 319 320 321 |
} static inline void cluster_clear_huge(struct swap_cluster_info *info) { info->flags &= ~CLUSTER_FLAG_HUGE; } |
235b62176 mm/swap: add clus... |
322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 |
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... |
340 341 342 343 |
/* * 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... |
344 |
static inline struct swap_cluster_info *lock_cluster_or_swap_info( |
59d98bf3c mm: swap: add com... |
345 |
struct swap_info_struct *si, unsigned long offset) |
235b62176 mm/swap: add clus... |
346 347 |
{ struct swap_cluster_info *ci; |
59d98bf3c mm: swap: add com... |
348 |
/* Try to use fine-grained SSD-style locking if available: */ |
235b62176 mm/swap: add clus... |
349 |
ci = lock_cluster(si, offset); |
59d98bf3c mm: swap: add com... |
350 |
/* Otherwise, fall back to traditional, coarse locking: */ |
235b62176 mm/swap: add clus... |
351 352 353 354 355 356 357 358 359 360 361 362 363 364 |
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... |
365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 |
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... |
389 |
struct swap_cluster_info *ci_tail; |
6b5349159 mm, swap: add swa... |
390 |
unsigned int tail = cluster_next(&list->tail); |
235b62176 mm/swap: add clus... |
391 392 393 394 395 396 397 |
/* * 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... |
398 |
spin_unlock(&ci_tail->lock); |
6b5349159 mm, swap: add swa... |
399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 |
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... |
418 419 420 421 422 423 424 425 426 427 428 429 |
/* 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... |
430 |
cluster_list_add_tail(&si->discard_clusters, si->cluster_info, idx); |
815c2c543 swap: make swap d... |
431 432 433 |
schedule_work(&si->discard_work); } |
38d8b4e6b mm, THP, swap: de... |
434 435 436 437 438 439 440 |
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... |
441 442 443 444 445 446 |
/* * 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... |
447 |
struct swap_cluster_info *info, *ci; |
815c2c543 swap: make swap d... |
448 449 450 |
unsigned int idx; info = si->cluster_info; |
6b5349159 mm, swap: add swa... |
451 452 |
while (!cluster_list_empty(&si->discard_clusters)) { idx = cluster_list_del_first(&si->discard_clusters, info); |
815c2c543 swap: make swap d... |
453 454 455 456 457 458 |
spin_unlock(&si->lock); discard_swap_cluster(si, idx * SWAPFILE_CLUSTER, SWAPFILE_CLUSTER); spin_lock(&si->lock); |
235b62176 mm/swap: add clus... |
459 |
ci = lock_cluster(si, idx * SWAPFILE_CLUSTER); |
38d8b4e6b mm, THP, swap: de... |
460 |
__free_cluster(si, idx); |
815c2c543 swap: make swap d... |
461 462 |
memset(si->swap_map + idx * SWAPFILE_CLUSTER, 0, SWAPFILE_CLUSTER); |
235b62176 mm/swap: add clus... |
463 |
unlock_cluster(ci); |
815c2c543 swap: make swap d... |
464 465 466 467 468 469 470 471 472 473 474 475 476 |
} } 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... |
477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 |
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... |
504 505 506 507 508 509 510 511 512 513 514 |
/* * 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... |
515 516 |
if (cluster_is_free(&cluster_info[idx])) alloc_cluster(p, idx); |
2a8f94493 swap: change bloc... |
517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 |
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... |
539 540 |
if (cluster_count(&cluster_info[idx]) == 0) free_cluster(p, idx); |
2a8f94493 swap: change bloc... |
541 542 543 544 545 546 |
} /* * 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... |
547 548 |
static bool scan_swap_map_ssd_cluster_conflict(struct swap_info_struct *si, |
2a8f94493 swap: change bloc... |
549 550 |
unsigned long offset) { |
ebc2a1a69 swap: make cluste... |
551 552 |
struct percpu_cluster *percpu_cluster; bool conflict; |
2a8f94493 swap: change bloc... |
553 |
offset /= SWAPFILE_CLUSTER; |
6b5349159 mm, swap: add swa... |
554 555 |
conflict = !cluster_list_empty(&si->free_clusters) && offset != cluster_list_first(&si->free_clusters) && |
2a8f94493 swap: change bloc... |
556 |
cluster_is_free(&si->cluster_info[offset]); |
ebc2a1a69 swap: make cluste... |
557 558 559 560 561 562 563 564 565 566 567 568 569 |
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... |
570 |
static bool scan_swap_map_try_ssd_cluster(struct swap_info_struct *si, |
ebc2a1a69 swap: make cluste... |
571 572 573 |
unsigned long *offset, unsigned long *scan_base) { struct percpu_cluster *cluster; |
235b62176 mm/swap: add clus... |
574 |
struct swap_cluster_info *ci; |
ebc2a1a69 swap: make cluste... |
575 |
bool found_free; |
235b62176 mm/swap: add clus... |
576 |
unsigned long tmp, max; |
ebc2a1a69 swap: make cluste... |
577 578 579 580 |
new_cluster: cluster = this_cpu_ptr(si->percpu_cluster); if (cluster_is_null(&cluster->index)) { |
6b5349159 mm, swap: add swa... |
581 582 |
if (!cluster_list_empty(&si->free_clusters)) { cluster->index = si->free_clusters.head; |
ebc2a1a69 swap: make cluste... |
583 584 |
cluster->next = cluster_next(&cluster->index) * SWAPFILE_CLUSTER; |
6b5349159 mm, swap: add swa... |
585 |
} else if (!cluster_list_empty(&si->discard_clusters)) { |
ebc2a1a69 swap: make cluste... |
586 587 588 589 590 591 592 593 |
/* * we don't have free cluster but have some clusters in * discarding, do discard now and reclaim them */ swap_do_scheduled_discard(si); *scan_base = *offset = si->cluster_next; goto new_cluster; } else |
36005bae2 mm/swap: allocate... |
594 |
return false; |
ebc2a1a69 swap: make cluste... |
595 596 597 598 599 600 601 602 603 |
} found_free = false; /* * Other CPUs can use our cluster if they can't find a free cluster, * check if there is still free entry in the cluster */ tmp = cluster->next; |
235b62176 mm/swap: add clus... |
604 605 606 607 608 609 610 611 |
max = min_t(unsigned long, si->max, (cluster_next(&cluster->index) + 1) * SWAPFILE_CLUSTER); if (tmp >= max) { cluster_set_null(&cluster->index); goto new_cluster; } ci = lock_cluster(si, tmp); while (tmp < max) { |
ebc2a1a69 swap: make cluste... |
612 613 614 615 616 617 |
if (!si->swap_map[tmp]) { found_free = true; break; } tmp++; } |
235b62176 mm/swap: add clus... |
618 |
unlock_cluster(ci); |
ebc2a1a69 swap: make cluste... |
619 620 621 622 623 624 625 |
if (!found_free) { cluster_set_null(&cluster->index); goto new_cluster; } cluster->next = tmp + 1; *offset = tmp; *scan_base = tmp; |
36005bae2 mm/swap: allocate... |
626 |
return found_free; |
2a8f94493 swap: change bloc... |
627 |
} |
a2468cc9b swap: choose swap... |
628 629 630 631 632 633 634 635 636 637 638 639 640 641 |
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... |
642 643 644 645 646 647 648 649 650 651 652 653 654 |
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) si->highest_bit -= nr_entries; si->inuse_pages += nr_entries; if (si->inuse_pages == si->pages) { si->lowest_bit = si->max; si->highest_bit = 0; |
a2468cc9b swap: choose swap... |
655 |
del_from_avail_list(si); |
38d8b4e6b mm, THP, swap: de... |
656 657 |
} } |
a2468cc9b swap: choose swap... |
658 659 660 661 662 663 664 665 666 667 668 |
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... |
669 670 671 672 673 674 675 676 677 678 679 680 |
static void swap_range_free(struct swap_info_struct *si, unsigned long offset, unsigned int nr_entries) { 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; si->highest_bit = end; |
a2468cc9b swap: choose swap... |
681 682 |
if (was_full && (si->flags & SWP_WRITEOK)) add_to_avail_list(si); |
38d8b4e6b mm, THP, swap: de... |
683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 |
} 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) { frontswap_invalidate_page(si->type, offset); if (swap_slot_free_notify) swap_slot_free_notify(si->bdev, offset); offset++; } } |
36005bae2 mm/swap: allocate... |
698 699 700 |
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 |
701 |
{ |
235b62176 mm/swap: add clus... |
702 |
struct swap_cluster_info *ci; |
ebebbbe90 swapfile: rearran... |
703 |
unsigned long offset; |
c60aa176c swapfile: swap al... |
704 |
unsigned long scan_base; |
7992fde72 swapfile: swap al... |
705 |
unsigned long last_in_cluster = 0; |
048c27fd7 [PATCH] swap: sca... |
706 |
int latency_ration = LATENCY_LIMIT; |
36005bae2 mm/swap: allocate... |
707 708 709 710 |
int n_ret = 0; if (nr > SWAP_BATCH) nr = SWAP_BATCH; |
7dfad4183 [PATCH] swap: sca... |
711 |
|
886bb7e9c swapfile: remove ... |
712 |
/* |
7dfad4183 [PATCH] swap: sca... |
713 714 715 716 717 718 719 |
* 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... |
720 |
* And we let swap pages go all over an SSD partition. Hugh |
7dfad4183 [PATCH] swap: sca... |
721 |
*/ |
52b7efdbe [PATCH] swap: sca... |
722 |
si->flags += SWP_SCANNING; |
c60aa176c swapfile: swap al... |
723 |
scan_base = offset = si->cluster_next; |
ebebbbe90 swapfile: rearran... |
724 |
|
ebc2a1a69 swap: make cluste... |
725 726 |
/* SSD algorithm */ if (si->cluster_info) { |
36005bae2 mm/swap: allocate... |
727 728 729 730 |
if (scan_swap_map_try_ssd_cluster(si, &offset, &scan_base)) goto checks; else goto scan; |
ebc2a1a69 swap: make cluste... |
731 |
} |
ebebbbe90 swapfile: rearran... |
732 733 734 735 736 |
if (unlikely(!si->cluster_nr--)) { if (si->pages - si->inuse_pages < SWAPFILE_CLUSTER) { si->cluster_nr = SWAPFILE_CLUSTER - 1; goto checks; } |
2a8f94493 swap: change bloc... |
737 |
|
ec8acf20a swap: add per-par... |
738 |
spin_unlock(&si->lock); |
7dfad4183 [PATCH] swap: sca... |
739 |
|
c60aa176c swapfile: swap al... |
740 741 742 |
/* * 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... |
743 744 |
* 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... |
745 |
*/ |
50088c440 mm/swapfile.c: de... |
746 |
scan_base = offset = si->lowest_bit; |
7dfad4183 [PATCH] swap: sca... |
747 748 749 750 |
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 |
751 |
if (si->swap_map[offset]) |
7dfad4183 [PATCH] swap: sca... |
752 753 |
last_in_cluster = offset + SWAPFILE_CLUSTER; else if (offset == last_in_cluster) { |
ec8acf20a swap: add per-par... |
754 |
spin_lock(&si->lock); |
ebebbbe90 swapfile: rearran... |
755 756 757 |
offset -= SWAPFILE_CLUSTER - 1; si->cluster_next = offset; si->cluster_nr = SWAPFILE_CLUSTER - 1; |
c60aa176c swapfile: swap al... |
758 759 760 761 762 763 764 765 766 |
goto checks; } if (unlikely(--latency_ration < 0)) { cond_resched(); latency_ration = LATENCY_LIMIT; } } offset = scan_base; |
ec8acf20a swap: add per-par... |
767 |
spin_lock(&si->lock); |
ebebbbe90 swapfile: rearran... |
768 |
si->cluster_nr = SWAPFILE_CLUSTER - 1; |
1da177e4c Linux-2.6.12-rc2 |
769 |
} |
7dfad4183 [PATCH] swap: sca... |
770 |
|
ebebbbe90 swapfile: rearran... |
771 |
checks: |
ebc2a1a69 swap: make cluste... |
772 |
if (si->cluster_info) { |
36005bae2 mm/swap: allocate... |
773 774 775 776 777 778 779 780 |
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... |
781 |
} |
ebebbbe90 swapfile: rearran... |
782 |
if (!(si->flags & SWP_WRITEOK)) |
52b7efdbe [PATCH] swap: sca... |
783 |
goto no_page; |
7dfad4183 [PATCH] swap: sca... |
784 785 |
if (!si->highest_bit) goto no_page; |
ebebbbe90 swapfile: rearran... |
786 |
if (offset > si->highest_bit) |
c60aa176c swapfile: swap al... |
787 |
scan_base = offset = si->lowest_bit; |
c9e444103 mm: reuse unused ... |
788 |
|
235b62176 mm/swap: add clus... |
789 |
ci = lock_cluster(si, offset); |
b73d7fcec swap: prevent reu... |
790 791 |
/* 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 ... |
792 |
int swap_was_freed; |
235b62176 mm/swap: add clus... |
793 |
unlock_cluster(ci); |
ec8acf20a swap: add per-par... |
794 |
spin_unlock(&si->lock); |
bcd49e867 mm/swapfile.c: us... |
795 |
swap_was_freed = __try_to_reclaim_swap(si, offset, TTRS_ANYWAY); |
ec8acf20a swap: add per-par... |
796 |
spin_lock(&si->lock); |
c9e444103 mm: reuse unused ... |
797 798 799 800 801 |
/* 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... |
802 803 |
if (si->swap_map[offset]) { unlock_cluster(ci); |
36005bae2 mm/swap: allocate... |
804 805 806 807 |
if (!n_ret) goto scan; else goto done; |
235b62176 mm/swap: add clus... |
808 |
} |
2872bb2d0 mm, swap: avoid l... |
809 810 811 |
si->swap_map[offset] = usage; inc_cluster_info_page(si, si->cluster_info, offset); unlock_cluster(ci); |
ebebbbe90 swapfile: rearran... |
812 |
|
38d8b4e6b mm, THP, swap: de... |
813 |
swap_range_alloc(si, offset, 1); |
ebebbbe90 swapfile: rearran... |
814 |
si->cluster_next = offset + 1; |
36005bae2 mm/swap: allocate... |
815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 |
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; else goto done; } /* non-ssd case */ ++offset; /* non-ssd case, still more slots in cluster? */ if (si->cluster_nr && !si->swap_map[offset]) { --si->cluster_nr; goto checks; } |
7992fde72 swapfile: swap al... |
848 |
|
36005bae2 mm/swap: allocate... |
849 850 851 |
done: si->flags -= SWP_SCANNING; return n_ret; |
7dfad4183 [PATCH] swap: sca... |
852 |
|
ebebbbe90 swapfile: rearran... |
853 |
scan: |
ec8acf20a swap: add per-par... |
854 |
spin_unlock(&si->lock); |
7dfad4183 [PATCH] swap: sca... |
855 |
while (++offset <= si->highest_bit) { |
52b7efdbe [PATCH] swap: sca... |
856 |
if (!si->swap_map[offset]) { |
ec8acf20a swap: add per-par... |
857 |
spin_lock(&si->lock); |
52b7efdbe [PATCH] swap: sca... |
858 859 |
goto checks; } |
c9e444103 mm: reuse unused ... |
860 |
if (vm_swap_full() && si->swap_map[offset] == SWAP_HAS_CACHE) { |
ec8acf20a swap: add per-par... |
861 |
spin_lock(&si->lock); |
c9e444103 mm: reuse unused ... |
862 863 |
goto checks; } |
048c27fd7 [PATCH] swap: sca... |
864 865 866 867 |
if (unlikely(--latency_ration < 0)) { cond_resched(); latency_ration = LATENCY_LIMIT; } |
7dfad4183 [PATCH] swap: sca... |
868 |
} |
c60aa176c swapfile: swap al... |
869 |
offset = si->lowest_bit; |
a5998061d mm/swapfile.c: do... |
870 |
while (offset < scan_base) { |
c60aa176c swapfile: swap al... |
871 |
if (!si->swap_map[offset]) { |
ec8acf20a swap: add per-par... |
872 |
spin_lock(&si->lock); |
c60aa176c swapfile: swap al... |
873 874 |
goto checks; } |
c9e444103 mm: reuse unused ... |
875 |
if (vm_swap_full() && si->swap_map[offset] == SWAP_HAS_CACHE) { |
ec8acf20a swap: add per-par... |
876 |
spin_lock(&si->lock); |
c9e444103 mm: reuse unused ... |
877 878 |
goto checks; } |
c60aa176c swapfile: swap al... |
879 880 881 882 |
if (unlikely(--latency_ration < 0)) { cond_resched(); latency_ration = LATENCY_LIMIT; } |
a5998061d mm/swapfile.c: do... |
883 |
offset++; |
c60aa176c swapfile: swap al... |
884 |
} |
ec8acf20a swap: add per-par... |
885 |
spin_lock(&si->lock); |
7dfad4183 [PATCH] swap: sca... |
886 887 |
no_page: |
52b7efdbe [PATCH] swap: sca... |
888 |
si->flags -= SWP_SCANNING; |
36005bae2 mm/swap: allocate... |
889 |
return n_ret; |
1da177e4c Linux-2.6.12-rc2 |
890 |
} |
38d8b4e6b mm, THP, swap: de... |
891 892 893 894 895 896 |
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... |
897 898 899 900 901 902 903 904 |
/* * 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... |
905 906 907 908 909 910 911 |
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... |
912 |
cluster_set_count_flag(ci, SWAPFILE_CLUSTER, CLUSTER_FLAG_HUGE); |
38d8b4e6b mm, THP, swap: de... |
913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 |
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... |
930 |
memset(si->swap_map + offset, 0, SWAPFILE_CLUSTER); |
38d8b4e6b mm, THP, swap: de... |
931 932 933 934 935 |
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... |
936 |
|
36005bae2 mm/swap: allocate... |
937 938 939 940 941 942 943 944 945 946 947 948 949 950 |
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... |
951 |
int get_swap_pages(int n_goal, swp_entry_t swp_entries[], int entry_size) |
1da177e4c Linux-2.6.12-rc2 |
952 |
{ |
5d5e8f195 mm, swap, get_swa... |
953 |
unsigned long size = swap_entry_size(entry_size); |
adfab836f swap: change swap... |
954 |
struct swap_info_struct *si, *next; |
36005bae2 mm/swap: allocate... |
955 956 |
long avail_pgs; int n_ret = 0; |
a2468cc9b swap: choose swap... |
957 |
int node; |
1da177e4c Linux-2.6.12-rc2 |
958 |
|
38d8b4e6b mm, THP, swap: de... |
959 |
/* Only single cluster request supported */ |
5d5e8f195 mm, swap, get_swa... |
960 |
WARN_ON_ONCE(n_goal > 1 && size == SWAPFILE_CLUSTER); |
38d8b4e6b mm, THP, swap: de... |
961 |
|
5d5e8f195 mm, swap, get_swa... |
962 |
avail_pgs = atomic_long_read(&nr_swap_pages) / size; |
36005bae2 mm/swap: allocate... |
963 |
if (avail_pgs <= 0) |
fb4f88dca [PATCH] swap: get... |
964 |
goto noswap; |
36005bae2 mm/swap: allocate... |
965 966 967 968 969 970 |
if (n_goal > SWAP_BATCH) n_goal = SWAP_BATCH; if (n_goal > avail_pgs) n_goal = avail_pgs; |
5d5e8f195 mm, swap, get_swa... |
971 |
atomic_long_sub(n_goal * size, &nr_swap_pages); |
fb4f88dca [PATCH] swap: get... |
972 |
|
18ab4d4ce swap: change swap... |
973 974 975 |
spin_lock(&swap_avail_lock); start_over: |
a2468cc9b swap: choose swap... |
976 977 |
node = numa_node_id(); plist_for_each_entry_safe(si, next, &swap_avail_heads[node], avail_lists[node]) { |
18ab4d4ce swap: change swap... |
978 |
/* requeue si to after same-priority siblings */ |
a2468cc9b swap: choose swap... |
979 |
plist_requeue(&si->avail_lists[node], &swap_avail_heads[node]); |
18ab4d4ce swap: change swap... |
980 |
spin_unlock(&swap_avail_lock); |
ec8acf20a swap: add per-par... |
981 |
spin_lock(&si->lock); |
adfab836f swap: change swap... |
982 |
if (!si->highest_bit || !(si->flags & SWP_WRITEOK)) { |
18ab4d4ce swap: change swap... |
983 |
spin_lock(&swap_avail_lock); |
a2468cc9b swap: choose swap... |
984 |
if (plist_node_empty(&si->avail_lists[node])) { |
18ab4d4ce swap: change swap... |
985 986 987 988 989 990 991 992 993 994 995 |
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... |
996 |
__del_from_avail_list(si); |
ec8acf20a swap: add per-par... |
997 |
spin_unlock(&si->lock); |
18ab4d4ce swap: change swap... |
998 |
goto nextsi; |
ec8acf20a swap: add per-par... |
999 |
} |
5d5e8f195 mm, swap, get_swa... |
1000 |
if (size == SWAPFILE_CLUSTER) { |
bc4ae27d8 mm: split SWP_FIL... |
1001 |
if (!(si->flags & SWP_FS)) |
f0eea189e mm, THP, swap: do... |
1002 1003 |
n_ret = swap_alloc_cluster(si, swp_entries); } else |
38d8b4e6b mm, THP, swap: de... |
1004 1005 |
n_ret = scan_swap_map_slots(si, SWAP_HAS_CACHE, n_goal, swp_entries); |
ec8acf20a swap: add per-par... |
1006 |
spin_unlock(&si->lock); |
5d5e8f195 mm, swap, get_swa... |
1007 |
if (n_ret || size == SWAPFILE_CLUSTER) |
36005bae2 mm/swap: allocate... |
1008 |
goto check_out; |
18ab4d4ce swap: change swap... |
1009 1010 |
pr_debug("scan_swap_map of si %d failed to find offset ", |
36005bae2 mm/swap: allocate... |
1011 |
si->type); |
18ab4d4ce swap: change swap... |
1012 1013 |
spin_lock(&swap_avail_lock); nextsi: |
adfab836f swap: change swap... |
1014 1015 1016 1017 |
/* * 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... |
1018 1019 1020 1021 |
* 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... |
1022 1023 |
* swap_avail_head list then try it, otherwise start over * if we have not gotten any slots. |
adfab836f swap: change swap... |
1024 |
*/ |
a2468cc9b swap: choose swap... |
1025 |
if (plist_node_empty(&next->avail_lists[node])) |
18ab4d4ce swap: change swap... |
1026 |
goto start_over; |
1da177e4c Linux-2.6.12-rc2 |
1027 |
} |
fb4f88dca [PATCH] swap: get... |
1028 |
|
18ab4d4ce swap: change swap... |
1029 |
spin_unlock(&swap_avail_lock); |
36005bae2 mm/swap: allocate... |
1030 1031 |
check_out: if (n_ret < n_goal) |
5d5e8f195 mm, swap, get_swa... |
1032 |
atomic_long_add((long)(n_goal - n_ret) * size, |
38d8b4e6b mm, THP, swap: de... |
1033 |
&nr_swap_pages); |
fb4f88dca [PATCH] swap: get... |
1034 |
noswap: |
36005bae2 mm/swap: allocate... |
1035 1036 |
return n_ret; } |
2de1a7e40 mm/swapfile.c: fi... |
1037 |
/* The only caller of this function is now suspend routine */ |
910321ea8 swap: revert spec... |
1038 1039 |
swp_entry_t get_swap_page_of_type(int type) { |
c10d38cc8 mm, swap: bounds ... |
1040 |
struct swap_info_struct *si = swap_type_to_swap_info(type); |
910321ea8 swap: revert spec... |
1041 |
pgoff_t offset; |
c10d38cc8 mm, swap: bounds ... |
1042 1043 |
if (!si) goto fail; |
ec8acf20a swap: add per-par... |
1044 |
spin_lock(&si->lock); |
c10d38cc8 mm, swap: bounds ... |
1045 |
if (si->flags & SWP_WRITEOK) { |
ec8acf20a swap: add per-par... |
1046 |
atomic_long_dec(&nr_swap_pages); |
910321ea8 swap: revert spec... |
1047 1048 1049 |
/* This is called for allocating swap entry, not cache */ offset = scan_swap_map(si, 1); if (offset) { |
ec8acf20a swap: add per-par... |
1050 |
spin_unlock(&si->lock); |
910321ea8 swap: revert spec... |
1051 1052 |
return swp_entry(type, offset); } |
ec8acf20a swap: add per-par... |
1053 |
atomic_long_inc(&nr_swap_pages); |
910321ea8 swap: revert spec... |
1054 |
} |
ec8acf20a swap: add per-par... |
1055 |
spin_unlock(&si->lock); |
c10d38cc8 mm, swap: bounds ... |
1056 |
fail: |
910321ea8 swap: revert spec... |
1057 1058 |
return (swp_entry_t) {0}; } |
e8c26ab60 mm/swap: skip rea... |
1059 |
static struct swap_info_struct *__swap_info_get(swp_entry_t entry) |
1da177e4c Linux-2.6.12-rc2 |
1060 |
{ |
73c34b6ac swap_info: miscel... |
1061 |
struct swap_info_struct *p; |
eb085574a mm, swap: fix rac... |
1062 |
unsigned long offset; |
1da177e4c Linux-2.6.12-rc2 |
1063 1064 1065 |
if (!entry.val) goto out; |
eb085574a mm, swap: fix rac... |
1066 |
p = swp_swap_info(entry); |
c10d38cc8 mm, swap: bounds ... |
1067 |
if (!p) |
1da177e4c Linux-2.6.12-rc2 |
1068 |
goto bad_nofile; |
1da177e4c Linux-2.6.12-rc2 |
1069 1070 1071 1072 1073 |
if (!(p->flags & SWP_USED)) goto bad_device; offset = swp_offset(entry); if (offset >= p->max) goto bad_offset; |
1da177e4c Linux-2.6.12-rc2 |
1074 |
return p; |
1da177e4c Linux-2.6.12-rc2 |
1075 |
bad_offset: |
6a991fc72 mm/swap: fix kern... |
1076 1077 |
pr_err("swap_info_get: %s%08lx ", Bad_offset, entry.val); |
1da177e4c Linux-2.6.12-rc2 |
1078 1079 |
goto out; bad_device: |
6a991fc72 mm/swap: fix kern... |
1080 1081 |
pr_err("swap_info_get: %s%08lx ", Unused_file, entry.val); |
1da177e4c Linux-2.6.12-rc2 |
1082 1083 |
goto out; bad_nofile: |
6a991fc72 mm/swap: fix kern... |
1084 1085 |
pr_err("swap_info_get: %s%08lx ", Bad_file, entry.val); |
1da177e4c Linux-2.6.12-rc2 |
1086 1087 |
out: return NULL; |
886bb7e9c swapfile: remove ... |
1088 |
} |
1da177e4c Linux-2.6.12-rc2 |
1089 |
|
e8c26ab60 mm/swap: skip rea... |
1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 |
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; if (!p->swap_map[swp_offset(entry)]) goto bad_free; return p; bad_free: pr_err("swap_info_get: %s%08lx ", Unused_offset, entry.val); goto out; out: return NULL; } |
235b62176 mm/swap: add clus... |
1108 1109 1110 1111 1112 1113 1114 1115 1116 |
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... |
1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 |
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... |
1132 1133 1134 |
static unsigned char __swap_entry_free_locked(struct swap_info_struct *p, unsigned long offset, unsigned char usage) |
1da177e4c Linux-2.6.12-rc2 |
1135 |
{ |
8d69aaee8 swap_info: swap_m... |
1136 1137 |
unsigned char count; unsigned char has_cache; |
235b62176 mm/swap: add clus... |
1138 |
|
253d553ba swap_info: SWAP_H... |
1139 |
count = p->swap_map[offset]; |
235b62176 mm/swap: add clus... |
1140 |
|
253d553ba swap_info: SWAP_H... |
1141 1142 |
has_cache = count & SWAP_HAS_CACHE; count &= ~SWAP_HAS_CACHE; |
355cfa73d mm: modify swap_m... |
1143 |
|
253d553ba swap_info: SWAP_H... |
1144 |
if (usage == SWAP_HAS_CACHE) { |
355cfa73d mm: modify swap_m... |
1145 |
VM_BUG_ON(!has_cache); |
253d553ba swap_info: SWAP_H... |
1146 |
has_cache = 0; |
aaa468653 swap_info: note S... |
1147 1148 1149 1150 1151 1152 |
} 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... |
1153 1154 1155 1156 1157 1158 1159 1160 1161 |
} 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... |
1162 |
|
253d553ba swap_info: SWAP_H... |
1163 |
usage = count | has_cache; |
7c00bafee mm/swap: free swa... |
1164 |
p->swap_map[offset] = usage ? : SWAP_HAS_CACHE; |
b32d5f32b mm/swapfile.c: ad... |
1165 1166 |
return usage; } |
eb085574a mm, swap: fix rac... |
1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 |
/* * 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(); if (!(si->flags & SWP_VALID)) 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... |
1230 1231 1232 1233 1234 1235 1236 1237 |
static unsigned char __swap_entry_free(struct swap_info_struct *p, swp_entry_t entry, unsigned char usage) { struct swap_cluster_info *ci; unsigned long offset = swp_offset(entry); ci = lock_cluster_or_swap_info(p, offset); usage = __swap_entry_free_locked(p, offset, usage); |
7c00bafee mm/swap: free swa... |
1238 |
unlock_cluster_or_swap_info(p, ci); |
10e364da1 mm/swapfile.c: ca... |
1239 1240 |
if (!usage) free_swap_slot(entry); |
7c00bafee mm/swap: free swa... |
1241 1242 1243 |
return usage; } |
355cfa73d mm: modify swap_m... |
1244 |
|
7c00bafee mm/swap: free swa... |
1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 |
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... |
1256 |
unlock_cluster(ci); |
38d8b4e6b mm, THP, swap: de... |
1257 1258 |
mem_cgroup_uncharge_swap(entry, 1); swap_range_free(p, offset, 1); |
1da177e4c Linux-2.6.12-rc2 |
1259 1260 1261 |
} /* |
2de1a7e40 mm/swapfile.c: fi... |
1262 |
* Caller has made sure that the swap device corresponding to entry |
1da177e4c Linux-2.6.12-rc2 |
1263 1264 1265 1266 |
* is still around or has not been recycled. */ void swap_free(swp_entry_t entry) { |
73c34b6ac swap_info: miscel... |
1267 |
struct swap_info_struct *p; |
1da177e4c Linux-2.6.12-rc2 |
1268 |
|
235b62176 mm/swap: add clus... |
1269 |
p = _swap_info_get(entry); |
10e364da1 mm/swapfile.c: ca... |
1270 1271 |
if (p) __swap_entry_free(p, entry, 1); |
1da177e4c Linux-2.6.12-rc2 |
1272 1273 1274 |
} /* |
cb4b86ba4 mm: add swap cach... |
1275 1276 |
* Called after dropping swapcache to decrease refcnt to swap entries. */ |
a448f2d07 mm/swapfile.c: un... |
1277 |
void put_swap_page(struct page *page, swp_entry_t entry) |
38d8b4e6b mm, THP, swap: de... |
1278 1279 1280 1281 1282 1283 |
{ 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... |
1284 1285 |
unsigned int i, free_entries = 0; unsigned char val; |
a448f2d07 mm/swapfile.c: un... |
1286 |
int size = swap_entry_size(hpage_nr_pages(page)); |
fe5266d5d mm/swapfile.c: re... |
1287 |
|
a3aea839e mm, THP, swap: su... |
1288 |
si = _swap_info_get(entry); |
38d8b4e6b mm, THP, swap: de... |
1289 1290 |
if (!si) return; |
c2343d276 mm/swapfile.c: pu... |
1291 |
ci = lock_cluster_or_swap_info(si, offset); |
a448f2d07 mm/swapfile.c: un... |
1292 |
if (size == SWAPFILE_CLUSTER) { |
a448f2d07 mm/swapfile.c: un... |
1293 1294 1295 1296 1297 1298 1299 1300 |
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... |
1301 |
cluster_clear_huge(ci); |
a448f2d07 mm/swapfile.c: un... |
1302 |
if (free_entries == SWAPFILE_CLUSTER) { |
c2343d276 mm/swapfile.c: pu... |
1303 |
unlock_cluster_or_swap_info(si, ci); |
a448f2d07 mm/swapfile.c: un... |
1304 |
spin_lock(&si->lock); |
a448f2d07 mm/swapfile.c: un... |
1305 1306 1307 1308 1309 1310 |
mem_cgroup_uncharge_swap(entry, SWAPFILE_CLUSTER); swap_free_cluster(si, idx); spin_unlock(&si->lock); return; } } |
c2343d276 mm/swapfile.c: pu... |
1311 1312 1313 1314 1315 1316 1317 |
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... |
1318 1319 |
} } |
c2343d276 mm/swapfile.c: pu... |
1320 |
unlock_cluster_or_swap_info(si, ci); |
38d8b4e6b mm, THP, swap: de... |
1321 |
} |
59807685a mm, THP, swap: su... |
1322 |
|
fe5266d5d mm/swapfile.c: re... |
1323 |
#ifdef CONFIG_THP_SWAP |
59807685a mm, THP, swap: su... |
1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 |
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... |
1338 |
#endif |
38d8b4e6b mm, THP, swap: de... |
1339 |
|
155b5f88e mm/swapfile.c: so... |
1340 1341 1342 1343 1344 1345 |
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... |
1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 |
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... |
1356 1357 1358 1359 1360 1361 1362 1363 |
/* * 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... |
1364 1365 1366 1367 |
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... |
1368 1369 |
prev = p; } |
235b62176 mm/swap: add clus... |
1370 |
if (p) |
7c00bafee mm/swap: free swa... |
1371 |
spin_unlock(&p->lock); |
cb4b86ba4 mm: add swap cach... |
1372 1373 1374 |
} /* |
c475a8ab6 [PATCH] can_share... |
1375 |
* How many references to page are currently swapped out? |
570a335b8 swap_info: swap c... |
1376 1377 |
* 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 |
1378 |
*/ |
bde05d1cc shmem: replace pa... |
1379 |
int page_swapcount(struct page *page) |
1da177e4c Linux-2.6.12-rc2 |
1380 |
{ |
c475a8ab6 [PATCH] can_share... |
1381 1382 |
int count = 0; struct swap_info_struct *p; |
235b62176 mm/swap: add clus... |
1383 |
struct swap_cluster_info *ci; |
1da177e4c Linux-2.6.12-rc2 |
1384 |
swp_entry_t entry; |
235b62176 mm/swap: add clus... |
1385 |
unsigned long offset; |
1da177e4c Linux-2.6.12-rc2 |
1386 |
|
4c21e2f24 [PATCH] mm: split... |
1387 |
entry.val = page_private(page); |
235b62176 mm/swap: add clus... |
1388 |
p = _swap_info_get(entry); |
1da177e4c Linux-2.6.12-rc2 |
1389 |
if (p) { |
235b62176 mm/swap: add clus... |
1390 1391 1392 1393 |
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 |
1394 |
} |
c475a8ab6 [PATCH] can_share... |
1395 |
return count; |
1da177e4c Linux-2.6.12-rc2 |
1396 |
} |
eb085574a mm, swap: fix rac... |
1397 |
int __swap_count(swp_entry_t entry) |
aa8d22a11 mm: swap: SWP_SYN... |
1398 |
{ |
eb085574a mm, swap: fix rac... |
1399 |
struct swap_info_struct *si; |
aa8d22a11 mm: swap: SWP_SYN... |
1400 |
pgoff_t offset = swp_offset(entry); |
eb085574a mm, swap: fix rac... |
1401 |
int count = 0; |
aa8d22a11 mm: swap: SWP_SYN... |
1402 |
|
eb085574a mm, swap: fix rac... |
1403 1404 1405 1406 1407 1408 |
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... |
1409 |
} |
322b8afe4 mm, swap: Fix a r... |
1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 |
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 |
1421 |
/* |
8334b9622 mm: /proc/pid/sma... |
1422 |
* How many references to @entry are currently swapped out? |
e8c26ab60 mm/swap: skip rea... |
1423 1424 1425 1426 1427 1428 |
* 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... |
1429 |
struct swap_info_struct *si; |
e8c26ab60 mm/swap: skip rea... |
1430 |
|
eb085574a mm, swap: fix rac... |
1431 1432 |
si = get_swap_device(entry); if (si) { |
322b8afe4 mm, swap: Fix a r... |
1433 |
count = swap_swapcount(si, entry); |
eb085574a mm, swap: fix rac... |
1434 1435 |
put_swap_device(si); } |
e8c26ab60 mm/swap: skip rea... |
1436 1437 1438 1439 1440 |
return count; } /* * How many references to @entry are currently swapped out? |
8334b9622 mm: /proc/pid/sma... |
1441 1442 1443 1444 1445 1446 |
* 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... |
1447 |
struct swap_cluster_info *ci; |
8334b9622 mm: /proc/pid/sma... |
1448 1449 1450 |
struct page *page; pgoff_t offset; unsigned char *map; |
235b62176 mm/swap: add clus... |
1451 |
p = _swap_info_get(entry); |
8334b9622 mm: /proc/pid/sma... |
1452 1453 |
if (!p) return 0; |
235b62176 mm/swap: add clus... |
1454 1455 1456 1457 1458 |
offset = swp_offset(entry); ci = lock_cluster_or_swap_info(p, offset); count = swap_count(p->swap_map[offset]); |
8334b9622 mm: /proc/pid/sma... |
1459 1460 1461 1462 1463 |
if (!(count & COUNT_CONTINUED)) goto out; count &= ~COUNT_CONTINUED; n = SWAP_MAP_MAX + 1; |
8334b9622 mm: /proc/pid/sma... |
1464 1465 1466 1467 1468 |
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... |
1469 |
page = list_next_entry(page, lru); |
8334b9622 mm: /proc/pid/sma... |
1470 1471 1472 1473 1474 1475 1476 1477 |
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... |
1478 |
unlock_cluster_or_swap_info(p, ci); |
8334b9622 mm: /proc/pid/sma... |
1479 1480 |
return count; } |
e07098294 mm, THP, swap: su... |
1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 |
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... |
1493 |
if (swap_count(map[roffset])) |
e07098294 mm, THP, swap: su... |
1494 1495 1496 1497 |
ret = true; goto unlock_out; } for (i = 0; i < SWAPFILE_CLUSTER; i++) { |
afa4711ef mm/swapfile.c: us... |
1498 |
if (swap_count(map[offset + i])) { |
e07098294 mm, THP, swap: su... |
1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 |
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... |
1512 |
if (!IS_ENABLED(CONFIG_THP_SWAP) || likely(!PageTransCompound(page))) |
e07098294 mm, THP, swap: su... |
1513 1514 1515 1516 1517 1518 1519 1520 1521 |
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... |
1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 |
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... |
1535 1536 |
if (!IS_ENABLED(CONFIG_THP_SWAP) || likely(!PageTransCompound(page))) { mapcount = page_trans_huge_mapcount(page, total_mapcount); |
ba3c4ce6d mm, THP, swap: ma... |
1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 |
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... |
1583 |
|
8334b9622 mm: /proc/pid/sma... |
1584 |
/* |
7b1fe5979 mm: reuse_swap_pa... |
1585 1586 1587 1588 |
* 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... |
1589 |
* |
ba3c4ce6d mm, THP, swap: ma... |
1590 |
* NOTE: total_map_swapcount should not be relied upon by the caller if |
6d0a07edd mm: thp: calculat... |
1591 1592 |
* 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 |
1593 |
*/ |
ba3c4ce6d mm, THP, swap: ma... |
1594 |
bool reuse_swap_page(struct page *page, int *total_map_swapcount) |
1da177e4c Linux-2.6.12-rc2 |
1595 |
{ |
ba3c4ce6d mm, THP, swap: ma... |
1596 |
int count, total_mapcount, total_swapcount; |
c475a8ab6 [PATCH] can_share... |
1597 |
|
309381fea mm: dump page whe... |
1598 |
VM_BUG_ON_PAGE(!PageLocked(page), page); |
5ad646880 ksm: let shared p... |
1599 |
if (unlikely(PageKsm(page))) |
6d0a07edd mm: thp: calculat... |
1600 |
return false; |
ba3c4ce6d mm, THP, swap: ma... |
1601 1602 1603 1604 1605 1606 1607 1608 |
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... |
1609 |
if (!PageWriteback(page)) { |
ba3c4ce6d mm, THP, swap: ma... |
1610 |
page = compound_head(page); |
7b1fe5979 mm: reuse_swap_pa... |
1611 1612 |
delete_from_swap_cache(page); SetPageDirty(page); |
f05714293 mm: support anony... |
1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 |
} 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... |
1624 1625 |
} } |
ba3c4ce6d mm, THP, swap: ma... |
1626 |
|
5ad646880 ksm: let shared p... |
1627 |
return count <= 1; |
1da177e4c Linux-2.6.12-rc2 |
1628 1629 1630 |
} /* |
a2c43eed8 mm: try_to_free_s... |
1631 1632 |
* 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 |
1633 |
*/ |
a2c43eed8 mm: try_to_free_s... |
1634 |
int try_to_free_swap(struct page *page) |
1da177e4c Linux-2.6.12-rc2 |
1635 |
{ |
309381fea mm: dump page whe... |
1636 |
VM_BUG_ON_PAGE(!PageLocked(page), page); |
1da177e4c Linux-2.6.12-rc2 |
1637 1638 1639 1640 1641 |
if (!PageSwapCache(page)) return 0; if (PageWriteback(page)) return 0; |
e07098294 mm, THP, swap: su... |
1642 |
if (page_swapped(page)) |
1da177e4c Linux-2.6.12-rc2 |
1643 |
return 0; |
b73d7fcec swap: prevent reu... |
1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 |
/* * 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... |
1656 |
* Hibernation suspends storage while it is writing the image |
f90ac3982 mm: avoid liveloc... |
1657 |
* to disk so check that here. |
b73d7fcec swap: prevent reu... |
1658 |
*/ |
f90ac3982 mm: avoid liveloc... |
1659 |
if (pm_suspended_storage()) |
b73d7fcec swap: prevent reu... |
1660 |
return 0; |
e07098294 mm, THP, swap: su... |
1661 |
page = compound_head(page); |
a2c43eed8 mm: try_to_free_s... |
1662 1663 1664 |
delete_from_swap_cache(page); SetPageDirty(page); return 1; |
68a22394c vmscan: free swap... |
1665 1666 1667 |
} /* |
1da177e4c Linux-2.6.12-rc2 |
1668 1669 1670 |
* 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... |
1671 |
int free_swap_and_cache(swp_entry_t entry) |
1da177e4c Linux-2.6.12-rc2 |
1672 |
{ |
2509ef26d badpage: zap prin... |
1673 |
struct swap_info_struct *p; |
7c00bafee mm/swap: free swa... |
1674 |
unsigned char count; |
1da177e4c Linux-2.6.12-rc2 |
1675 |
|
a7420aa54 HWPOISON: Add sup... |
1676 |
if (non_swap_entry(entry)) |
2509ef26d badpage: zap prin... |
1677 |
return 1; |
0697212a4 [PATCH] Swapless ... |
1678 |
|
7c00bafee mm/swap: free swa... |
1679 |
p = _swap_info_get(entry); |
1da177e4c Linux-2.6.12-rc2 |
1680 |
if (p) { |
7c00bafee mm/swap: free swa... |
1681 |
count = __swap_entry_free(p, entry, 1); |
e07098294 mm, THP, swap: su... |
1682 |
if (count == SWAP_HAS_CACHE && |
bcd49e867 mm/swapfile.c: us... |
1683 1684 1685 |
!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 |
1686 |
} |
2509ef26d badpage: zap prin... |
1687 |
return p != NULL; |
1da177e4c Linux-2.6.12-rc2 |
1688 |
} |
b0cb1a19d Replace CONFIG_SO... |
1689 |
#ifdef CONFIG_HIBERNATION |
f577eb30a [PATCH] swsusp: l... |
1690 |
/* |
915bae9eb [PATCH] swsusp: u... |
1691 |
* Find the swap type that corresponds to given device (if any). |
f577eb30a [PATCH] swsusp: l... |
1692 |
* |
915bae9eb [PATCH] swsusp: u... |
1693 1694 1695 1696 |
* @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... |
1697 |
*/ |
7bf236874 [PATCH] swsusp: D... |
1698 |
int swap_type_of(dev_t device, sector_t offset, struct block_device **bdev_p) |
f577eb30a [PATCH] swsusp: l... |
1699 |
{ |
915bae9eb [PATCH] swsusp: u... |
1700 |
struct block_device *bdev = NULL; |
efa90a981 swap_info: change... |
1701 |
int type; |
f577eb30a [PATCH] swsusp: l... |
1702 |
|
915bae9eb [PATCH] swsusp: u... |
1703 1704 |
if (device) bdev = bdget(device); |
f577eb30a [PATCH] swsusp: l... |
1705 |
spin_lock(&swap_lock); |
efa90a981 swap_info: change... |
1706 1707 |
for (type = 0; type < nr_swapfiles; type++) { struct swap_info_struct *sis = swap_info[type]; |
f577eb30a [PATCH] swsusp: l... |
1708 |
|
915bae9eb [PATCH] swsusp: u... |
1709 |
if (!(sis->flags & SWP_WRITEOK)) |
f577eb30a [PATCH] swsusp: l... |
1710 |
continue; |
b6b5bce35 [PATCH] swsusp: F... |
1711 |
|
915bae9eb [PATCH] swsusp: u... |
1712 |
if (!bdev) { |
7bf236874 [PATCH] swsusp: D... |
1713 |
if (bdev_p) |
dddac6a7b PM / Hibernate: R... |
1714 |
*bdev_p = bdgrab(sis->bdev); |
7bf236874 [PATCH] swsusp: D... |
1715 |
|
6e1819d61 [PATCH] swsusp: u... |
1716 |
spin_unlock(&swap_lock); |
efa90a981 swap_info: change... |
1717 |
return type; |
6e1819d61 [PATCH] swsusp: u... |
1718 |
} |
915bae9eb [PATCH] swsusp: u... |
1719 |
if (bdev == sis->bdev) { |
4efaceb1c mm, swap: use rbt... |
1720 |
struct swap_extent *se = first_se(sis); |
915bae9eb [PATCH] swsusp: u... |
1721 |
|
915bae9eb [PATCH] swsusp: u... |
1722 |
if (se->start_block == offset) { |
7bf236874 [PATCH] swsusp: D... |
1723 |
if (bdev_p) |
dddac6a7b PM / Hibernate: R... |
1724 |
*bdev_p = bdgrab(sis->bdev); |
7bf236874 [PATCH] swsusp: D... |
1725 |
|
915bae9eb [PATCH] swsusp: u... |
1726 1727 |
spin_unlock(&swap_lock); bdput(bdev); |
efa90a981 swap_info: change... |
1728 |
return type; |
915bae9eb [PATCH] swsusp: u... |
1729 |
} |
f577eb30a [PATCH] swsusp: l... |
1730 1731 1732 |
} } spin_unlock(&swap_lock); |
915bae9eb [PATCH] swsusp: u... |
1733 1734 |
if (bdev) bdput(bdev); |
f577eb30a [PATCH] swsusp: l... |
1735 1736 1737 1738 |
return -ENODEV; } /* |
73c34b6ac swap_info: miscel... |
1739 1740 1741 1742 1743 1744 |
* 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 ... |
1745 |
struct swap_info_struct *si = swap_type_to_swap_info(type); |
73c34b6ac swap_info: miscel... |
1746 |
|
c10d38cc8 mm, swap: bounds ... |
1747 |
if (!si || !(si->flags & SWP_WRITEOK)) |
73c34b6ac swap_info: miscel... |
1748 |
return 0; |
d4906e1aa swap: rework map_... |
1749 |
return map_swap_entry(swp_entry(type, offset), &bdev); |
73c34b6ac swap_info: miscel... |
1750 1751 1752 |
} /* |
f577eb30a [PATCH] swsusp: l... |
1753 1754 1755 1756 1757 1758 1759 1760 |
* 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... |
1761 1762 1763 |
spin_lock(&swap_lock); if ((unsigned int)type < nr_swapfiles) { struct swap_info_struct *sis = swap_info[type]; |
ec8acf20a swap: add per-par... |
1764 |
spin_lock(&sis->lock); |
efa90a981 swap_info: change... |
1765 1766 |
if (sis->flags & SWP_WRITEOK) { n = sis->pages; |
f577eb30a [PATCH] swsusp: l... |
1767 |
if (free) |
efa90a981 swap_info: change... |
1768 |
n -= sis->inuse_pages; |
f577eb30a [PATCH] swsusp: l... |
1769 |
} |
ec8acf20a swap: add per-par... |
1770 |
spin_unlock(&sis->lock); |
f577eb30a [PATCH] swsusp: l... |
1771 |
} |
efa90a981 swap_info: change... |
1772 |
spin_unlock(&swap_lock); |
f577eb30a [PATCH] swsusp: l... |
1773 1774 |
return n; } |
73c34b6ac swap_info: miscel... |
1775 |
#endif /* CONFIG_HIBERNATION */ |
f577eb30a [PATCH] swsusp: l... |
1776 |
|
9f8bdb3f3 mm: make swapoff ... |
1777 |
static inline int pte_same_as_swp(pte_t pte, pte_t swp_pte) |
179ef71cb mm: save soft-dir... |
1778 |
{ |
9f8bdb3f3 mm: make swapoff ... |
1779 |
return pte_same(pte_swp_clear_soft_dirty(pte), swp_pte); |
179ef71cb mm: save soft-dir... |
1780 |
} |
1da177e4c Linux-2.6.12-rc2 |
1781 |
/* |
72866f6f2 [PATCH] mm: anon ... |
1782 1783 1784 |
* 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 |
1785 |
*/ |
044d66c1d memcgroup: reinst... |
1786 |
static int unuse_pte(struct vm_area_struct *vma, pmd_t *pmd, |
1da177e4c Linux-2.6.12-rc2 |
1787 1788 |
unsigned long addr, swp_entry_t entry, struct page *page) { |
9e16b7fb1 mm,ksm: swapoff m... |
1789 |
struct page *swapcache; |
72835c86c mm: unify remaini... |
1790 |
struct mem_cgroup *memcg; |
044d66c1d memcgroup: reinst... |
1791 1792 1793 |
spinlock_t *ptl; pte_t *pte; int ret = 1; |
9e16b7fb1 mm,ksm: swapoff m... |
1794 1795 1796 1797 |
swapcache = page; page = ksm_might_need_to_copy(page, vma, addr); if (unlikely(!page)) return -ENOMEM; |
f627c2f53 memcg: adjust to ... |
1798 1799 |
if (mem_cgroup_try_charge(page, vma->vm_mm, GFP_KERNEL, &memcg, false)) { |
044d66c1d memcgroup: reinst... |
1800 |
ret = -ENOMEM; |
85d9fc89f memcg: fix refcnt... |
1801 1802 |
goto out_nolock; } |
044d66c1d memcgroup: reinst... |
1803 1804 |
pte = pte_offset_map_lock(vma->vm_mm, pmd, addr, &ptl); |
9f8bdb3f3 mm: make swapoff ... |
1805 |
if (unlikely(!pte_same_as_swp(*pte, swp_entry_to_pte(entry)))) { |
f627c2f53 memcg: adjust to ... |
1806 |
mem_cgroup_cancel_charge(page, memcg, false); |
044d66c1d memcgroup: reinst... |
1807 1808 1809 |
ret = 0; goto out; } |
8a9f3ccd2 Memory controller... |
1810 |
|
b084d4353 mm: count swap usage |
1811 |
dec_mm_counter(vma->vm_mm, MM_SWAPENTS); |
d559db086 mm: clean up mm_c... |
1812 |
inc_mm_counter(vma->vm_mm, MM_ANONPAGES); |
1da177e4c Linux-2.6.12-rc2 |
1813 1814 1815 |
get_page(page); set_pte_at(vma->vm_mm, addr, pte, pte_mkold(mk_pte(page, vma->vm_page_prot))); |
00501b531 mm: memcontrol: r... |
1816 |
if (page == swapcache) { |
d281ee614 rmap: add argumen... |
1817 |
page_add_anon_rmap(page, vma, addr, false); |
f627c2f53 memcg: adjust to ... |
1818 |
mem_cgroup_commit_charge(page, memcg, true, false); |
00501b531 mm: memcontrol: r... |
1819 |
} else { /* ksm created a completely new copy */ |
d281ee614 rmap: add argumen... |
1820 |
page_add_new_anon_rmap(page, vma, addr, false); |
f627c2f53 memcg: adjust to ... |
1821 |
mem_cgroup_commit_charge(page, memcg, false, false); |
00501b531 mm: memcontrol: r... |
1822 1823 |
lru_cache_add_active_or_unevictable(page, vma); } |
1da177e4c Linux-2.6.12-rc2 |
1824 1825 1826 1827 1828 1829 |
swap_free(entry); /* * Move the page to the active list so it is not * immediately swapped out again after swapon. */ activate_page(page); |
044d66c1d memcgroup: reinst... |
1830 1831 |
out: pte_unmap_unlock(pte, ptl); |
85d9fc89f memcg: fix refcnt... |
1832 |
out_nolock: |
9e16b7fb1 mm,ksm: swapoff m... |
1833 1834 1835 1836 |
if (page != swapcache) { unlock_page(page); put_page(page); } |
044d66c1d memcgroup: reinst... |
1837 |
return ret; |
1da177e4c Linux-2.6.12-rc2 |
1838 1839 1840 |
} static int unuse_pte_range(struct vm_area_struct *vma, pmd_t *pmd, |
b56a2d8af mm: rid swapoff o... |
1841 1842 1843 |
unsigned long addr, unsigned long end, unsigned int type, bool frontswap, unsigned long *fs_pages_to_unuse) |
1da177e4c Linux-2.6.12-rc2 |
1844 |
{ |
b56a2d8af mm: rid swapoff o... |
1845 1846 |
struct page *page; swp_entry_t entry; |
705e87c0c [PATCH] mm: pte_o... |
1847 |
pte_t *pte; |
b56a2d8af mm: rid swapoff o... |
1848 1849 |
struct swap_info_struct *si; unsigned long offset; |
8a9f3ccd2 Memory controller... |
1850 |
int ret = 0; |
b56a2d8af mm: rid swapoff o... |
1851 |
volatile unsigned char *swap_map; |
1da177e4c Linux-2.6.12-rc2 |
1852 |
|
b56a2d8af mm: rid swapoff o... |
1853 |
si = swap_info[type]; |
044d66c1d memcgroup: reinst... |
1854 |
pte = pte_offset_map(pmd, addr); |
1da177e4c Linux-2.6.12-rc2 |
1855 |
do { |
b56a2d8af mm: rid swapoff o... |
1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 |
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]; vmf.vma = vma; vmf.address = addr; vmf.pmd = pmd; page = swapin_readahead(entry, GFP_HIGHUSER_MOVABLE, &vmf); 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 |
1897 |
} |
b56a2d8af mm: rid swapoff o... |
1898 1899 |
try_next: pte = pte_offset_map(pmd, addr); |
1da177e4c Linux-2.6.12-rc2 |
1900 |
} while (pte++, addr += PAGE_SIZE, addr != end); |
044d66c1d memcgroup: reinst... |
1901 |
pte_unmap(pte - 1); |
b56a2d8af mm: rid swapoff o... |
1902 1903 |
ret = 0; |
044d66c1d memcgroup: reinst... |
1904 |
out: |
8a9f3ccd2 Memory controller... |
1905 |
return ret; |
1da177e4c Linux-2.6.12-rc2 |
1906 1907 1908 1909 |
} 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... |
1910 1911 |
unsigned int type, bool frontswap, unsigned long *fs_pages_to_unuse) |
1da177e4c Linux-2.6.12-rc2 |
1912 1913 1914 |
{ pmd_t *pmd; unsigned long next; |
8a9f3ccd2 Memory controller... |
1915 |
int ret; |
1da177e4c Linux-2.6.12-rc2 |
1916 1917 1918 |
pmd = pmd_offset(pud, addr); do { |
dc644a073 mm: add three mor... |
1919 |
cond_resched(); |
1da177e4c Linux-2.6.12-rc2 |
1920 |
next = pmd_addr_end(addr, end); |
1a5a9906d mm: thp: fix pmd_... |
1921 |
if (pmd_none_or_trans_huge_or_clear_bad(pmd)) |
1da177e4c Linux-2.6.12-rc2 |
1922 |
continue; |
b56a2d8af mm: rid swapoff o... |
1923 1924 |
ret = unuse_pte_range(vma, pmd, addr, next, type, frontswap, fs_pages_to_unuse); |
8a9f3ccd2 Memory controller... |
1925 1926 |
if (ret) return ret; |
1da177e4c Linux-2.6.12-rc2 |
1927 1928 1929 |
} while (pmd++, addr = next, addr != end); return 0; } |
c2febafc6 mm: convert gener... |
1930 |
static inline int unuse_pud_range(struct vm_area_struct *vma, p4d_t *p4d, |
1da177e4c Linux-2.6.12-rc2 |
1931 |
unsigned long addr, unsigned long end, |
b56a2d8af mm: rid swapoff o... |
1932 1933 |
unsigned int type, bool frontswap, unsigned long *fs_pages_to_unuse) |
1da177e4c Linux-2.6.12-rc2 |
1934 1935 1936 |
{ pud_t *pud; unsigned long next; |
8a9f3ccd2 Memory controller... |
1937 |
int ret; |
1da177e4c Linux-2.6.12-rc2 |
1938 |
|
c2febafc6 mm: convert gener... |
1939 |
pud = pud_offset(p4d, addr); |
1da177e4c Linux-2.6.12-rc2 |
1940 1941 1942 1943 |
do { next = pud_addr_end(addr, end); if (pud_none_or_clear_bad(pud)) continue; |
b56a2d8af mm: rid swapoff o... |
1944 1945 |
ret = unuse_pmd_range(vma, pud, addr, next, type, frontswap, fs_pages_to_unuse); |
8a9f3ccd2 Memory controller... |
1946 1947 |
if (ret) return ret; |
1da177e4c Linux-2.6.12-rc2 |
1948 1949 1950 |
} while (pud++, addr = next, addr != end); return 0; } |
c2febafc6 mm: convert gener... |
1951 1952 |
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... |
1953 1954 |
unsigned int type, bool frontswap, unsigned long *fs_pages_to_unuse) |
c2febafc6 mm: convert gener... |
1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 |
{ 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... |
1965 1966 |
ret = unuse_pud_range(vma, p4d, addr, next, type, frontswap, fs_pages_to_unuse); |
c2febafc6 mm: convert gener... |
1967 1968 1969 1970 1971 |
if (ret) return ret; } while (p4d++, addr = next, addr != end); return 0; } |
b56a2d8af mm: rid swapoff o... |
1972 1973 |
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 |
1974 1975 1976 |
{ pgd_t *pgd; unsigned long addr, end, next; |
8a9f3ccd2 Memory controller... |
1977 |
int ret; |
1da177e4c Linux-2.6.12-rc2 |
1978 |
|
b56a2d8af mm: rid swapoff o... |
1979 1980 |
addr = vma->vm_start; end = vma->vm_end; |
1da177e4c Linux-2.6.12-rc2 |
1981 1982 1983 1984 1985 1986 |
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... |
1987 1988 |
ret = unuse_p4d_range(vma, pgd, addr, next, type, frontswap, fs_pages_to_unuse); |
8a9f3ccd2 Memory controller... |
1989 1990 |
if (ret) return ret; |
1da177e4c Linux-2.6.12-rc2 |
1991 1992 1993 |
} while (pgd++, addr = next, addr != end); return 0; } |
b56a2d8af mm: rid swapoff o... |
1994 1995 |
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 |
1996 1997 |
{ struct vm_area_struct *vma; |
8a9f3ccd2 Memory controller... |
1998 |
int ret = 0; |
1da177e4c Linux-2.6.12-rc2 |
1999 |
|
b56a2d8af mm: rid swapoff o... |
2000 |
down_read(&mm->mmap_sem); |
1da177e4c Linux-2.6.12-rc2 |
2001 |
for (vma = mm->mmap; vma; vma = vma->vm_next) { |
b56a2d8af mm: rid swapoff o... |
2002 2003 2004 2005 2006 2007 |
if (vma->anon_vma) { ret = unuse_vma(vma, type, frontswap, fs_pages_to_unuse); if (ret) break; } |
dc644a073 mm: add three mor... |
2008 |
cond_resched(); |
1da177e4c Linux-2.6.12-rc2 |
2009 |
} |
1da177e4c Linux-2.6.12-rc2 |
2010 |
up_read(&mm->mmap_sem); |
b56a2d8af mm: rid swapoff o... |
2011 |
return ret; |
1da177e4c Linux-2.6.12-rc2 |
2012 2013 2014 |
} /* |
38b5faf4b mm: frontswap: co... |
2015 |
* Scan swap_map (or frontswap_map if frontswap parameter is true) |
b56a2d8af mm: rid swapoff o... |
2016 2017 |
* 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 |
2018 |
*/ |
6eb396dc4 [PATCH] swap: swa... |
2019 |
static unsigned int find_next_to_unuse(struct swap_info_struct *si, |
38b5faf4b mm: frontswap: co... |
2020 |
unsigned int prev, bool frontswap) |
1da177e4c Linux-2.6.12-rc2 |
2021 |
{ |
b56a2d8af mm: rid swapoff o... |
2022 |
unsigned int i; |
8d69aaee8 swap_info: swap_m... |
2023 |
unsigned char count; |
1da177e4c Linux-2.6.12-rc2 |
2024 2025 |
/* |
5d337b919 [PATCH] swap: swa... |
2026 |
* No need for swap_lock here: we're just looking |
1da177e4c Linux-2.6.12-rc2 |
2027 2028 |
* 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... |
2029 |
* allocations from this area (while holding swap_lock). |
1da177e4c Linux-2.6.12-rc2 |
2030 |
*/ |
b56a2d8af mm: rid swapoff o... |
2031 |
for (i = prev + 1; i < si->max; i++) { |
4db0c3c29 mm: remove rest o... |
2032 |
count = READ_ONCE(si->swap_map[i]); |
355cfa73d mm: modify swap_m... |
2033 |
if (count && swap_count(count) != SWAP_MAP_BAD) |
dc644a073 mm: add three mor... |
2034 2035 2036 2037 |
if (!frontswap || frontswap_test(si, i)) break; if ((i % LATENCY_LIMIT) == 0) cond_resched(); |
1da177e4c Linux-2.6.12-rc2 |
2038 |
} |
b56a2d8af mm: rid swapoff o... |
2039 2040 2041 |
if (i == si->max) i = 0; |
1da177e4c Linux-2.6.12-rc2 |
2042 2043 2044 2045 |
return i; } /* |
b56a2d8af mm: rid swapoff o... |
2046 |
* If the boolean frontswap is true, only unuse pages_to_unuse pages; |
38b5faf4b mm: frontswap: co... |
2047 |
* pages_to_unuse==0 means all pages; ignored if frontswap is false |
1da177e4c Linux-2.6.12-rc2 |
2048 |
*/ |
38b5faf4b mm: frontswap: co... |
2049 2050 |
int try_to_unuse(unsigned int type, bool frontswap, unsigned long pages_to_unuse) |
1da177e4c Linux-2.6.12-rc2 |
2051 |
{ |
b56a2d8af mm: rid swapoff o... |
2052 2053 2054 2055 |
struct mm_struct *prev_mm; struct mm_struct *mm; struct list_head *p; int retval = 0; |
efa90a981 swap_info: change... |
2056 |
struct swap_info_struct *si = swap_info[type]; |
1da177e4c Linux-2.6.12-rc2 |
2057 2058 |
struct page *page; swp_entry_t entry; |
b56a2d8af mm: rid swapoff o... |
2059 |
unsigned int i; |
1da177e4c Linux-2.6.12-rc2 |
2060 |
|
b56a2d8af mm: rid swapoff o... |
2061 2062 |
if (!si->inuse_pages) return 0; |
1da177e4c Linux-2.6.12-rc2 |
2063 |
|
b56a2d8af mm: rid swapoff o... |
2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 |
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; |
64165b1af mm: swapoff: take... |
2077 2078 2079 |
while (si->inuse_pages && !signal_pending(current) && (p = p->next) != &init_mm.mmlist) { |
1da177e4c Linux-2.6.12-rc2 |
2080 |
|
b56a2d8af mm: rid swapoff o... |
2081 2082 2083 2084 2085 2086 2087 |
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 |
2088 |
|
b56a2d8af mm: rid swapoff o... |
2089 2090 2091 |
if (retval) { mmput(prev_mm); goto out; |
1da177e4c Linux-2.6.12-rc2 |
2092 2093 2094 |
} /* |
b56a2d8af mm: rid swapoff o... |
2095 2096 |
* Make sure that we aren't completely killing * interactive performance. |
1da177e4c Linux-2.6.12-rc2 |
2097 |
*/ |
b56a2d8af mm: rid swapoff o... |
2098 2099 2100 2101 |
cond_resched(); spin_lock(&mmlist_lock); } spin_unlock(&mmlist_lock); |
1da177e4c Linux-2.6.12-rc2 |
2102 |
|
b56a2d8af mm: rid swapoff o... |
2103 |
mmput(prev_mm); |
1da177e4c Linux-2.6.12-rc2 |
2104 |
|
b56a2d8af mm: rid swapoff o... |
2105 |
i = 0; |
64165b1af mm: swapoff: take... |
2106 2107 2108 |
while (si->inuse_pages && !signal_pending(current) && (i = find_next_to_unuse(si, i, frontswap)) != 0) { |
1da177e4c Linux-2.6.12-rc2 |
2109 |
|
b56a2d8af mm: rid swapoff o... |
2110 2111 2112 2113 |
entry = swp_entry(type, i); page = find_get_page(swap_address_space(entry), i); if (!page) continue; |
68bdc8d64 mm: try_to_unuse ... |
2114 2115 2116 |
/* * It is conceivable that a racing task removed this page from |
b56a2d8af mm: rid swapoff o... |
2117 2118 2119 |
* 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 |
2120 |
*/ |
b56a2d8af mm: rid swapoff o... |
2121 2122 2123 |
lock_page(page); wait_on_page_writeback(page); try_to_free_swap(page); |
1da177e4c Linux-2.6.12-rc2 |
2124 |
unlock_page(page); |
09cbfeaf1 mm, fs: get rid o... |
2125 |
put_page(page); |
1da177e4c Linux-2.6.12-rc2 |
2126 2127 |
/* |
b56a2d8af mm: rid swapoff o... |
2128 2129 2130 |
* 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 |
2131 |
*/ |
b56a2d8af mm: rid swapoff o... |
2132 2133 |
if (pages_to_unuse && --pages_to_unuse == 0) goto out; |
1da177e4c Linux-2.6.12-rc2 |
2134 |
} |
b56a2d8af mm: rid swapoff o... |
2135 2136 2137 2138 2139 |
/* * 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... |
2140 |
* |
af53d3e9e mm: swapoff: shme... |
2141 2142 2143 2144 2145 |
* 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... |
2146 |
*/ |
64165b1af mm: swapoff: take... |
2147 2148 2149 2150 2151 |
if (si->inuse_pages) { if (!signal_pending(current)) goto retry; retval = -EINTR; } |
b56a2d8af mm: rid swapoff o... |
2152 2153 |
out: return (retval == FRONTSWAP_PAGES_UNUSED) ? 0 : retval; |
1da177e4c Linux-2.6.12-rc2 |
2154 2155 2156 |
} /* |
5d337b919 [PATCH] swap: swa... |
2157 2158 2159 |
* 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 |
2160 2161 2162 2163 2164 |
* 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... |
2165 |
unsigned int type; |
1da177e4c Linux-2.6.12-rc2 |
2166 |
|
efa90a981 swap_info: change... |
2167 2168 |
for (type = 0; type < nr_swapfiles; type++) if (swap_info[type]->inuse_pages) |
1da177e4c Linux-2.6.12-rc2 |
2169 2170 2171 2172 2173 2174 2175 2176 2177 |
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_... |
2178 2179 2180 |
* 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 |
2181 |
*/ |
d4906e1aa swap: rework map_... |
2182 |
static sector_t map_swap_entry(swp_entry_t entry, struct block_device **bdev) |
1da177e4c Linux-2.6.12-rc2 |
2183 |
{ |
f29ad6a99 swap_info: privat... |
2184 |
struct swap_info_struct *sis; |
f29ad6a99 swap_info: privat... |
2185 2186 |
struct swap_extent *se; pgoff_t offset; |
c10d38cc8 mm, swap: bounds ... |
2187 |
sis = swp_swap_info(entry); |
f29ad6a99 swap_info: privat... |
2188 2189 2190 |
*bdev = sis->bdev; offset = swp_offset(entry); |
4efaceb1c mm, swap: use rbt... |
2191 2192 |
se = offset_to_swap_extent(sis, offset); return se->start_block + (offset - se->start_page); |
1da177e4c Linux-2.6.12-rc2 |
2193 2194 2195 |
} /* |
d4906e1aa swap: rework map_... |
2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 |
* 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 |
2206 2207 2208 2209 |
* Free all of a swapdev's extent information */ static void destroy_swap_extents(struct swap_info_struct *sis) { |
4efaceb1c mm, swap: use rbt... |
2210 2211 2212 |
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 |
2213 |
|
4efaceb1c mm, swap: use rbt... |
2214 |
rb_erase(rb, &sis->swap_extent_root); |
1da177e4c Linux-2.6.12-rc2 |
2215 2216 |
kfree(se); } |
62c230bc1 mm: add support f... |
2217 |
|
bc4ae27d8 mm: split SWP_FIL... |
2218 |
if (sis->flags & SWP_ACTIVATED) { |
62c230bc1 mm: add support f... |
2219 2220 |
struct file *swap_file = sis->swap_file; struct address_space *mapping = swap_file->f_mapping; |
bc4ae27d8 mm: split SWP_FIL... |
2221 2222 2223 |
sis->flags &= ~SWP_ACTIVATED; if (mapping->a_ops->swap_deactivate) mapping->a_ops->swap_deactivate(swap_file); |
62c230bc1 mm: add support f... |
2224 |
} |
1da177e4c Linux-2.6.12-rc2 |
2225 2226 2227 2228 |
} /* * Add a block range (and the corresponding page range) into this swapdev's |
4efaceb1c mm, swap: use rbt... |
2229 |
* extent tree. |
1da177e4c Linux-2.6.12-rc2 |
2230 |
* |
11d31886d [PATCH] swap: swa... |
2231 |
* This function rather assumes that it is called in ascending page order. |
1da177e4c Linux-2.6.12-rc2 |
2232 |
*/ |
a509bc1a9 mm: swap: impleme... |
2233 |
int |
1da177e4c Linux-2.6.12-rc2 |
2234 2235 2236 |
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... |
2237 |
struct rb_node **link = &sis->swap_extent_root.rb_node, *parent = NULL; |
1da177e4c Linux-2.6.12-rc2 |
2238 2239 |
struct swap_extent *se; struct swap_extent *new_se; |
4efaceb1c mm, swap: use rbt... |
2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 |
/* * 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... |
2252 2253 |
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 |
2254 2255 2256 2257 |
/* Merge it */ se->nr_pages += nr_pages; return 0; } |
1da177e4c Linux-2.6.12-rc2 |
2258 |
} |
4efaceb1c mm, swap: use rbt... |
2259 |
/* No merge, insert a new extent. */ |
1da177e4c Linux-2.6.12-rc2 |
2260 2261 2262 2263 2264 2265 |
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... |
2266 2267 |
rb_link_node(&new_se->rb_node, parent, link); rb_insert_color(&new_se->rb_node, &sis->swap_extent_root); |
53092a740 [PATCH] swap: sho... |
2268 |
return 1; |
1da177e4c Linux-2.6.12-rc2 |
2269 |
} |
aa8aa8a33 mm: export add_sw... |
2270 |
EXPORT_SYMBOL_GPL(add_swap_extent); |
1da177e4c Linux-2.6.12-rc2 |
2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 |
/* * 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... |
2292 2293 |
* 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 |
2294 2295 2296 2297 2298 2299 2300 2301 |
* * 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... |
2302 |
static int setup_swap_extents(struct swap_info_struct *sis, sector_t *span) |
1da177e4c Linux-2.6.12-rc2 |
2303 |
{ |
62c230bc1 mm: add support f... |
2304 2305 2306 |
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 |
2307 |
int ret; |
1da177e4c Linux-2.6.12-rc2 |
2308 2309 |
if (S_ISBLK(inode->i_mode)) { ret = add_swap_extent(sis, 0, sis->max, 0); |
53092a740 [PATCH] swap: sho... |
2310 |
*span = sis->pages; |
a509bc1a9 mm: swap: impleme... |
2311 |
return ret; |
1da177e4c Linux-2.6.12-rc2 |
2312 |
} |
62c230bc1 mm: add support f... |
2313 |
if (mapping->a_ops->swap_activate) { |
a509bc1a9 mm: swap: impleme... |
2314 |
ret = mapping->a_ops->swap_activate(sis, swap_file, span); |
bc4ae27d8 mm: split SWP_FIL... |
2315 2316 |
if (ret >= 0) sis->flags |= SWP_ACTIVATED; |
62c230bc1 mm: add support f... |
2317 |
if (!ret) { |
bc4ae27d8 mm: split SWP_FIL... |
2318 |
sis->flags |= SWP_FS; |
62c230bc1 mm: add support f... |
2319 2320 2321 |
ret = add_swap_extent(sis, 0, sis->max, 0); *span = sis->pages; } |
a509bc1a9 mm: swap: impleme... |
2322 |
return ret; |
62c230bc1 mm: add support f... |
2323 |
} |
a509bc1a9 mm: swap: impleme... |
2324 |
return generic_swapfile_activate(sis, swap_file, span); |
1da177e4c Linux-2.6.12-rc2 |
2325 |
} |
a2468cc9b swap: choose swap... |
2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 |
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... |
2337 2338 2339 |
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... |
2340 |
{ |
a2468cc9b swap: choose swap... |
2341 |
int i; |
40531542e sys_swapon: separ... |
2342 2343 2344 2345 |
if (prio >= 0) p->prio = prio; else p->prio = --least_priority; |
18ab4d4ce swap: change swap... |
2346 2347 2348 2349 2350 |
/* * 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... |
2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 |
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... |
2361 |
p->swap_map = swap_map; |
2a8f94493 swap: change bloc... |
2362 |
p->cluster_info = cluster_info; |
eb085574a mm, swap: fix rac... |
2363 2364 2365 2366 2367 |
} static void _enable_swap_info(struct swap_info_struct *p) { p->flags |= SWP_WRITEOK | SWP_VALID; |
ec8acf20a swap: add per-par... |
2368 |
atomic_long_add(p->pages, &nr_swap_pages); |
40531542e sys_swapon: separ... |
2369 |
total_swap_pages += p->pages; |
adfab836f swap: change swap... |
2370 |
assert_spin_locked(&swap_lock); |
adfab836f swap: change swap... |
2371 |
/* |
18ab4d4ce swap: change swap... |
2372 2373 2374 2375 2376 2377 2378 2379 |
* 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... |
2380 |
*/ |
18ab4d4ce swap: change swap... |
2381 |
plist_add(&p->list, &swap_active_head); |
a2468cc9b swap: choose swap... |
2382 |
add_to_avail_list(p); |
cf0cac0a0 mm: refactor rein... |
2383 2384 2385 2386 |
} static void enable_swap_info(struct swap_info_struct *p, int prio, unsigned char *swap_map, |
2a8f94493 swap: change bloc... |
2387 |
struct swap_cluster_info *cluster_info, |
cf0cac0a0 mm: refactor rein... |
2388 2389 |
unsigned long *frontswap_map) { |
4f89849da frontswap: get ri... |
2390 |
frontswap_init(p->type, frontswap_map); |
cf0cac0a0 mm: refactor rein... |
2391 |
spin_lock(&swap_lock); |
ec8acf20a swap: add per-par... |
2392 |
spin_lock(&p->lock); |
eb085574a mm, swap: fix rac... |
2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 |
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... |
2404 |
spin_unlock(&p->lock); |
cf0cac0a0 mm: refactor rein... |
2405 2406 2407 2408 2409 2410 |
spin_unlock(&swap_lock); } static void reinsert_swap_info(struct swap_info_struct *p) { spin_lock(&swap_lock); |
ec8acf20a swap: add per-par... |
2411 |
spin_lock(&p->lock); |
eb085574a mm, swap: fix rac... |
2412 2413 |
setup_swap_info(p, p->prio, p->swap_map, p->cluster_info); _enable_swap_info(p); |
ec8acf20a swap: add per-par... |
2414 |
spin_unlock(&p->lock); |
40531542e sys_swapon: separ... |
2415 2416 |
spin_unlock(&swap_lock); } |
67afa38e0 mm/swap: add cach... |
2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 |
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... |
2427 |
SYSCALL_DEFINE1(swapoff, const char __user *, specialfile) |
1da177e4c Linux-2.6.12-rc2 |
2428 |
{ |
73c34b6ac swap_info: miscel... |
2429 |
struct swap_info_struct *p = NULL; |
8d69aaee8 swap_info: swap_m... |
2430 |
unsigned char *swap_map; |
2a8f94493 swap: change bloc... |
2431 |
struct swap_cluster_info *cluster_info; |
4f89849da frontswap: get ri... |
2432 |
unsigned long *frontswap_map; |
1da177e4c Linux-2.6.12-rc2 |
2433 2434 2435 |
struct file *swap_file, *victim; struct address_space *mapping; struct inode *inode; |
91a27b2a7 vfs: define struc... |
2436 |
struct filename *pathname; |
adfab836f swap: change swap... |
2437 |
int err, found = 0; |
5b808a230 swap: fix set_blo... |
2438 |
unsigned int old_block_size; |
886bb7e9c swapfile: remove ... |
2439 |
|
1da177e4c Linux-2.6.12-rc2 |
2440 2441 |
if (!capable(CAP_SYS_ADMIN)) return -EPERM; |
191c54244 mm: collapse secu... |
2442 |
BUG_ON(!current->mm); |
1da177e4c Linux-2.6.12-rc2 |
2443 |
pathname = getname(specialfile); |
1da177e4c Linux-2.6.12-rc2 |
2444 |
if (IS_ERR(pathname)) |
f58b59c1d swapfile: fix nam... |
2445 |
return PTR_ERR(pathname); |
1da177e4c Linux-2.6.12-rc2 |
2446 |
|
669abf4e5 vfs: make path_op... |
2447 |
victim = file_open_name(pathname, O_RDWR|O_LARGEFILE, 0); |
1da177e4c Linux-2.6.12-rc2 |
2448 2449 2450 2451 2452 |
err = PTR_ERR(victim); if (IS_ERR(victim)) goto out; mapping = victim->f_mapping; |
5d337b919 [PATCH] swap: swa... |
2453 |
spin_lock(&swap_lock); |
18ab4d4ce swap: change swap... |
2454 |
plist_for_each_entry(p, &swap_active_head, list) { |
22c6f8fdb swapfile: remove ... |
2455 |
if (p->flags & SWP_WRITEOK) { |
adfab836f swap: change swap... |
2456 2457 |
if (p->swap_file->f_mapping == mapping) { found = 1; |
1da177e4c Linux-2.6.12-rc2 |
2458 |
break; |
adfab836f swap: change swap... |
2459 |
} |
1da177e4c Linux-2.6.12-rc2 |
2460 |
} |
1da177e4c Linux-2.6.12-rc2 |
2461 |
} |
adfab836f swap: change swap... |
2462 |
if (!found) { |
1da177e4c Linux-2.6.12-rc2 |
2463 |
err = -EINVAL; |
5d337b919 [PATCH] swap: swa... |
2464 |
spin_unlock(&swap_lock); |
1da177e4c Linux-2.6.12-rc2 |
2465 2466 |
goto out_dput; } |
191c54244 mm: collapse secu... |
2467 |
if (!security_vm_enough_memory_mm(current->mm, p->pages)) |
1da177e4c Linux-2.6.12-rc2 |
2468 2469 2470 |
vm_unacct_memory(p->pages); else { err = -ENOMEM; |
5d337b919 [PATCH] swap: swa... |
2471 |
spin_unlock(&swap_lock); |
1da177e4c Linux-2.6.12-rc2 |
2472 2473 |
goto out_dput; } |
a2468cc9b swap: choose swap... |
2474 |
del_from_avail_list(p); |
ec8acf20a swap: add per-par... |
2475 |
spin_lock(&p->lock); |
78ecba081 mm: fix ever-decr... |
2476 |
if (p->prio < 0) { |
adfab836f swap: change swap... |
2477 |
struct swap_info_struct *si = p; |
a2468cc9b swap: choose swap... |
2478 |
int nid; |
adfab836f swap: change swap... |
2479 |
|
18ab4d4ce swap: change swap... |
2480 |
plist_for_each_entry_continue(si, &swap_active_head, list) { |
adfab836f swap: change swap... |
2481 |
si->prio++; |
18ab4d4ce swap: change swap... |
2482 |
si->list.prio--; |
a2468cc9b swap: choose swap... |
2483 2484 2485 2486 |
for_each_node(nid) { if (si->avail_lists[nid].prio != 1) si->avail_lists[nid].prio--; } |
adfab836f swap: change swap... |
2487 |
} |
78ecba081 mm: fix ever-decr... |
2488 2489 |
least_priority++; } |
18ab4d4ce swap: change swap... |
2490 |
plist_del(&p->list, &swap_active_head); |
ec8acf20a swap: add per-par... |
2491 |
atomic_long_sub(p->pages, &nr_swap_pages); |
1da177e4c Linux-2.6.12-rc2 |
2492 2493 |
total_swap_pages -= p->pages; p->flags &= ~SWP_WRITEOK; |
ec8acf20a swap: add per-par... |
2494 |
spin_unlock(&p->lock); |
5d337b919 [PATCH] swap: swa... |
2495 |
spin_unlock(&swap_lock); |
fb4f88dca [PATCH] swap: get... |
2496 |
|
039939a65 mm/swap: enable s... |
2497 |
disable_swap_slots_cache_lock(); |
e1e12d2f3 mm, oom: fix race... |
2498 |
set_current_oom_origin(); |
adfab836f swap: change swap... |
2499 |
err = try_to_unuse(p->type, false, 0); /* force unuse all pages */ |
e1e12d2f3 mm, oom: fix race... |
2500 |
clear_current_oom_origin(); |
1da177e4c Linux-2.6.12-rc2 |
2501 |
|
1da177e4c Linux-2.6.12-rc2 |
2502 2503 |
if (err) { /* re-insert swap space back into swap_list */ |
cf0cac0a0 mm: refactor rein... |
2504 |
reinsert_swap_info(p); |
039939a65 mm/swap: enable s... |
2505 |
reenable_swap_slots_cache_unlock(); |
1da177e4c Linux-2.6.12-rc2 |
2506 2507 |
goto out_dput; } |
52b7efdbe [PATCH] swap: sca... |
2508 |
|
039939a65 mm/swap: enable s... |
2509 |
reenable_swap_slots_cache_unlock(); |
eb085574a mm, swap: fix rac... |
2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 |
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... |
2520 |
flush_work(&p->discard_work); |
5d337b919 [PATCH] swap: swa... |
2521 |
destroy_swap_extents(p); |
570a335b8 swap_info: swap c... |
2522 2523 |
if (p->flags & SWP_CONTINUED) free_swap_count_continuations(p); |
81a0298bd mm, swap: don't u... |
2524 2525 |
if (!p->bdev || !blk_queue_nonrot(bdev_get_queue(p->bdev))) atomic_dec(&nr_rotate_swap); |
fc0abb145 [PATCH] sem2mutex... |
2526 |
mutex_lock(&swapon_mutex); |
5d337b919 [PATCH] swap: swa... |
2527 |
spin_lock(&swap_lock); |
ec8acf20a swap: add per-par... |
2528 |
spin_lock(&p->lock); |
5d337b919 [PATCH] swap: swa... |
2529 |
drain_mmlist(); |
52b7efdbe [PATCH] swap: sca... |
2530 |
/* wait for anyone still in scan_swap_map */ |
52b7efdbe [PATCH] swap: sca... |
2531 2532 |
p->highest_bit = 0; /* cuts scans short */ while (p->flags >= SWP_SCANNING) { |
ec8acf20a swap: add per-par... |
2533 |
spin_unlock(&p->lock); |
5d337b919 [PATCH] swap: swa... |
2534 |
spin_unlock(&swap_lock); |
13e4b57f6 [PATCH] mm: fix-u... |
2535 |
schedule_timeout_uninterruptible(1); |
5d337b919 [PATCH] swap: swa... |
2536 |
spin_lock(&swap_lock); |
ec8acf20a swap: add per-par... |
2537 |
spin_lock(&p->lock); |
52b7efdbe [PATCH] swap: sca... |
2538 |
} |
52b7efdbe [PATCH] swap: sca... |
2539 |
|
1da177e4c Linux-2.6.12-rc2 |
2540 |
swap_file = p->swap_file; |
5b808a230 swap: fix set_blo... |
2541 |
old_block_size = p->old_block_size; |
1da177e4c Linux-2.6.12-rc2 |
2542 2543 2544 2545 |
p->swap_file = NULL; p->max = 0; swap_map = p->swap_map; p->swap_map = NULL; |
2a8f94493 swap: change bloc... |
2546 2547 |
cluster_info = p->cluster_info; p->cluster_info = NULL; |
4f89849da frontswap: get ri... |
2548 |
frontswap_map = frontswap_map_get(p); |
ec8acf20a swap: add per-par... |
2549 |
spin_unlock(&p->lock); |
5d337b919 [PATCH] swap: swa... |
2550 |
spin_unlock(&swap_lock); |
adfab836f swap: change swap... |
2551 |
frontswap_invalidate_area(p->type); |
58e97ba6b frontswap: enable... |
2552 |
frontswap_map_set(p, NULL); |
fc0abb145 [PATCH] sem2mutex... |
2553 |
mutex_unlock(&swapon_mutex); |
ebc2a1a69 swap: make cluste... |
2554 2555 |
free_percpu(p->percpu_cluster); p->percpu_cluster = NULL; |
1da177e4c Linux-2.6.12-rc2 |
2556 |
vfree(swap_map); |
54f180d3c mm, swap: use kvz... |
2557 2558 |
kvfree(cluster_info); kvfree(frontswap_map); |
2de1a7e40 mm/swapfile.c: fi... |
2559 |
/* Destroy swap account information */ |
adfab836f swap: change swap... |
2560 |
swap_cgroup_swapoff(p->type); |
4b3ef9daa mm/swap: split sw... |
2561 |
exit_swap_address_space(p->type); |
27a7faa07 memcg: swap cgrou... |
2562 |
|
1da177e4c Linux-2.6.12-rc2 |
2563 2564 2565 |
inode = mapping->host; if (S_ISBLK(inode->i_mode)) { struct block_device *bdev = I_BDEV(inode); |
1638045c3 mm: set S_SWAPFIL... |
2566 |
|
5b808a230 swap: fix set_blo... |
2567 |
set_blocksize(bdev, old_block_size); |
e525fd89d block: make blkde... |
2568 |
blkdev_put(bdev, FMODE_READ | FMODE_WRITE | FMODE_EXCL); |
1da177e4c Linux-2.6.12-rc2 |
2569 |
} |
1638045c3 mm: set S_SWAPFIL... |
2570 2571 2572 2573 |
inode_lock(inode); inode->i_flags &= ~S_SWAPFILE; inode_unlock(inode); |
1da177e4c Linux-2.6.12-rc2 |
2574 |
filp_close(swap_file, NULL); |
f893ab41e mm/swap: fix race... |
2575 2576 2577 2578 2579 2580 2581 2582 2583 |
/* * 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 |
2584 |
err = 0; |
66d7dd518 /proc/swaps: supp... |
2585 2586 |
atomic_inc(&proc_poll_event); wake_up_interruptible(&proc_poll_wait); |
1da177e4c Linux-2.6.12-rc2 |
2587 2588 2589 2590 |
out_dput: filp_close(victim, NULL); out: |
f58b59c1d swapfile: fix nam... |
2591 |
putname(pathname); |
1da177e4c Linux-2.6.12-rc2 |
2592 2593 2594 2595 |
return err; } #ifdef CONFIG_PROC_FS |
9dd957485 ipc, kernel, mm: ... |
2596 |
static __poll_t swaps_poll(struct file *file, poll_table *wait) |
66d7dd518 /proc/swaps: supp... |
2597 |
{ |
f15146380 fs: seq_file - ad... |
2598 |
struct seq_file *seq = file->private_data; |
66d7dd518 /proc/swaps: supp... |
2599 2600 |
poll_wait(file, &proc_poll_wait, wait); |
f15146380 fs: seq_file - ad... |
2601 2602 |
if (seq->poll_event != atomic_read(&proc_poll_event)) { seq->poll_event = atomic_read(&proc_poll_event); |
a9a08845e vfs: do bulk POLL... |
2603 |
return EPOLLIN | EPOLLRDNORM | EPOLLERR | EPOLLPRI; |
66d7dd518 /proc/swaps: supp... |
2604 |
} |
a9a08845e vfs: do bulk POLL... |
2605 |
return EPOLLIN | EPOLLRDNORM; |
66d7dd518 /proc/swaps: supp... |
2606 |
} |
1da177e4c Linux-2.6.12-rc2 |
2607 2608 2609 |
/* iterator */ static void *swap_start(struct seq_file *swap, loff_t *pos) { |
efa90a981 swap_info: change... |
2610 2611 |
struct swap_info_struct *si; int type; |
1da177e4c Linux-2.6.12-rc2 |
2612 |
loff_t l = *pos; |
fc0abb145 [PATCH] sem2mutex... |
2613 |
mutex_lock(&swapon_mutex); |
1da177e4c Linux-2.6.12-rc2 |
2614 |
|
881e4aabe [PATCH] Always pr... |
2615 2616 |
if (!l) return SEQ_START_TOKEN; |
c10d38cc8 mm, swap: bounds ... |
2617 |
for (type = 0; (si = swap_type_to_swap_info(type)); type++) { |
efa90a981 swap_info: change... |
2618 |
if (!(si->flags & SWP_USED) || !si->swap_map) |
1da177e4c Linux-2.6.12-rc2 |
2619 |
continue; |
881e4aabe [PATCH] Always pr... |
2620 |
if (!--l) |
efa90a981 swap_info: change... |
2621 |
return si; |
1da177e4c Linux-2.6.12-rc2 |
2622 2623 2624 2625 2626 2627 2628 |
} return NULL; } static void *swap_next(struct seq_file *swap, void *v, loff_t *pos) { |
efa90a981 swap_info: change... |
2629 2630 |
struct swap_info_struct *si = v; int type; |
1da177e4c Linux-2.6.12-rc2 |
2631 |
|
881e4aabe [PATCH] Always pr... |
2632 |
if (v == SEQ_START_TOKEN) |
efa90a981 swap_info: change... |
2633 2634 2635 |
type = 0; else type = si->type + 1; |
881e4aabe [PATCH] Always pr... |
2636 |
|
c10d38cc8 mm, swap: bounds ... |
2637 |
for (; (si = swap_type_to_swap_info(type)); type++) { |
efa90a981 swap_info: change... |
2638 |
if (!(si->flags & SWP_USED) || !si->swap_map) |
1da177e4c Linux-2.6.12-rc2 |
2639 2640 |
continue; ++*pos; |
efa90a981 swap_info: change... |
2641 |
return si; |
1da177e4c Linux-2.6.12-rc2 |
2642 2643 2644 2645 2646 2647 2648 |
} return NULL; } static void swap_stop(struct seq_file *swap, void *v) { |
fc0abb145 [PATCH] sem2mutex... |
2649 |
mutex_unlock(&swapon_mutex); |
1da177e4c Linux-2.6.12-rc2 |
2650 2651 2652 2653 |
} static int swap_show(struct seq_file *swap, void *v) { |
efa90a981 swap_info: change... |
2654 |
struct swap_info_struct *si = v; |
1da177e4c Linux-2.6.12-rc2 |
2655 2656 |
struct file *file; int len; |
efa90a981 swap_info: change... |
2657 |
if (si == SEQ_START_TOKEN) { |
881e4aabe [PATCH] Always pr... |
2658 2659 2660 2661 |
seq_puts(swap,"Filename\t\t\t\tType\t\tSize\tUsed\tPriority "); return 0; } |
1da177e4c Linux-2.6.12-rc2 |
2662 |
|
efa90a981 swap_info: change... |
2663 |
file = si->swap_file; |
2726d5662 vfs: add seq_file... |
2664 2665 |
len = seq_file_path(swap, file, " \t \\"); |
6eb396dc4 [PATCH] swap: swa... |
2666 2667 |
seq_printf(swap, "%*s%s\t%u\t%u\t%d ", |
886bb7e9c swapfile: remove ... |
2668 |
len < 40 ? 40 - len : 1, " ", |
496ad9aa8 new helper: file_... |
2669 |
S_ISBLK(file_inode(file)->i_mode) ? |
1da177e4c Linux-2.6.12-rc2 |
2670 |
"partition" : "file\t", |
efa90a981 swap_info: change... |
2671 2672 2673 |
si->pages << (PAGE_SHIFT - 10), si->inuse_pages << (PAGE_SHIFT - 10), si->prio); |
1da177e4c Linux-2.6.12-rc2 |
2674 2675 |
return 0; } |
15ad7cdcf [PATCH] struct se... |
2676 |
static const struct seq_operations swaps_op = { |
1da177e4c Linux-2.6.12-rc2 |
2677 2678 2679 2680 2681 2682 2683 2684 |
.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... |
2685 |
struct seq_file *seq; |
66d7dd518 /proc/swaps: supp... |
2686 |
int ret; |
66d7dd518 /proc/swaps: supp... |
2687 |
ret = seq_open(file, &swaps_op); |
f15146380 fs: seq_file - ad... |
2688 |
if (ret) |
66d7dd518 /proc/swaps: supp... |
2689 |
return ret; |
66d7dd518 /proc/swaps: supp... |
2690 |
|
f15146380 fs: seq_file - ad... |
2691 2692 2693 |
seq = file->private_data; seq->poll_event = atomic_read(&proc_poll_event); return 0; |
1da177e4c Linux-2.6.12-rc2 |
2694 |
} |
15ad7cdcf [PATCH] struct se... |
2695 |
static const struct file_operations proc_swaps_operations = { |
1da177e4c Linux-2.6.12-rc2 |
2696 2697 2698 2699 |
.open = swaps_open, .read = seq_read, .llseek = seq_lseek, .release = seq_release, |
66d7dd518 /proc/swaps: supp... |
2700 |
.poll = swaps_poll, |
1da177e4c Linux-2.6.12-rc2 |
2701 2702 2703 2704 |
}; static int __init procswaps_init(void) { |
3d71f86f4 mm: use non-racy ... |
2705 |
proc_create("swaps", 0, NULL, &proc_swaps_operations); |
1da177e4c Linux-2.6.12-rc2 |
2706 2707 2708 2709 |
return 0; } __initcall(procswaps_init); #endif /* CONFIG_PROC_FS */ |
1796316a8 x86: consolidate ... |
2710 2711 2712 2713 2714 2715 2716 2717 |
#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... |
2718 |
static struct swap_info_struct *alloc_swap_info(void) |
1da177e4c Linux-2.6.12-rc2 |
2719 |
{ |
73c34b6ac swap_info: miscel... |
2720 |
struct swap_info_struct *p; |
1da177e4c Linux-2.6.12-rc2 |
2721 |
unsigned int type; |
a2468cc9b swap: choose swap... |
2722 |
int i; |
efa90a981 swap_info: change... |
2723 |
|
960087445 mm/swapfile.c: us... |
2724 |
p = kvzalloc(struct_size(p, avail_lists, nr_node_ids), GFP_KERNEL); |
efa90a981 swap_info: change... |
2725 |
if (!p) |
53cbb2435 sys_swapon: separ... |
2726 |
return ERR_PTR(-ENOMEM); |
efa90a981 swap_info: change... |
2727 |
|
5d337b919 [PATCH] swap: swa... |
2728 |
spin_lock(&swap_lock); |
efa90a981 swap_info: change... |
2729 2730 |
for (type = 0; type < nr_swapfiles; type++) { if (!(swap_info[type]->flags & SWP_USED)) |
1da177e4c Linux-2.6.12-rc2 |
2731 |
break; |
efa90a981 swap_info: change... |
2732 |
} |
0697212a4 [PATCH] Swapless ... |
2733 |
if (type >= MAX_SWAPFILES) { |
5d337b919 [PATCH] swap: swa... |
2734 |
spin_unlock(&swap_lock); |
873d7bcfd mm/swapfile.c: us... |
2735 |
kvfree(p); |
730c0581c sys_swapon: simpl... |
2736 |
return ERR_PTR(-EPERM); |
1da177e4c Linux-2.6.12-rc2 |
2737 |
} |
efa90a981 swap_info: change... |
2738 2739 |
if (type >= nr_swapfiles) { p->type = type; |
c10d38cc8 mm, swap: bounds ... |
2740 |
WRITE_ONCE(swap_info[type], p); |
efa90a981 swap_info: change... |
2741 2742 2743 2744 2745 2746 |
/* * 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 ... |
2747 |
WRITE_ONCE(nr_swapfiles, nr_swapfiles + 1); |
efa90a981 swap_info: change... |
2748 |
} else { |
873d7bcfd mm/swapfile.c: us... |
2749 |
kvfree(p); |
efa90a981 swap_info: change... |
2750 2751 2752 2753 2754 2755 |
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... |
2756 |
p->swap_extent_root = RB_ROOT; |
18ab4d4ce swap: change swap... |
2757 |
plist_node_init(&p->list, 0); |
a2468cc9b swap: choose swap... |
2758 2759 |
for_each_node(i) plist_node_init(&p->avail_lists[i], 0); |
1da177e4c Linux-2.6.12-rc2 |
2760 |
p->flags = SWP_USED; |
5d337b919 [PATCH] swap: swa... |
2761 |
spin_unlock(&swap_lock); |
ec8acf20a swap: add per-par... |
2762 |
spin_lock_init(&p->lock); |
2628bd6fc mm, swap: fix rac... |
2763 |
spin_lock_init(&p->cont_lock); |
efa90a981 swap_info: change... |
2764 |
|
53cbb2435 sys_swapon: separ... |
2765 |
return p; |
53cbb2435 sys_swapon: separ... |
2766 |
} |
4d0e1e107 sys_swapon: separ... |
2767 2768 2769 2770 2771 2772 2773 |
static int claim_swapfile(struct swap_info_struct *p, struct inode *inode) { int error; if (S_ISBLK(inode->i_mode)) { p->bdev = bdgrab(I_BDEV(inode)); error = blkdev_get(p->bdev, |
6f179af88 mm: fix potential... |
2774 |
FMODE_READ | FMODE_WRITE | FMODE_EXCL, p); |
4d0e1e107 sys_swapon: separ... |
2775 2776 |
if (error < 0) { p->bdev = NULL; |
6f179af88 mm: fix potential... |
2777 |
return error; |
4d0e1e107 sys_swapon: separ... |
2778 2779 2780 2781 |
} p->old_block_size = block_size(p->bdev); error = set_blocksize(p->bdev, PAGE_SIZE); if (error < 0) |
87ade72a7 sys_swapon: simpl... |
2782 |
return error; |
4d0e1e107 sys_swapon: separ... |
2783 2784 2785 |
p->flags |= SWP_BLKDEV; } else if (S_ISREG(inode->i_mode)) { p->bdev = inode->i_sb->s_bdev; |
1638045c3 mm: set S_SWAPFIL... |
2786 2787 2788 2789 2790 |
} inode_lock(inode); if (IS_SWAPFILE(inode)) return -EBUSY; |
4d0e1e107 sys_swapon: separ... |
2791 2792 |
return 0; |
4d0e1e107 sys_swapon: separ... |
2793 |
} |
377eeaa8e x86/speculation/l... |
2794 2795 2796 2797 2798 2799 2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811 2812 2813 2814 2815 2816 2817 2818 2819 2820 2821 |
/* * 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... |
2822 2823 2824 2825 2826 2827 2828 |
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... |
2829 |
unsigned long last_page; |
ca8bd38bf sys_swapon: separ... |
2830 2831 |
if (memcmp("SWAPSPACE2", swap_header->magic.magic, 10)) { |
465c47fd8 mm/swapfile.c: co... |
2832 2833 |
pr_err("Unable to find swap-space signature "); |
387190253 sys_swapon: simpl... |
2834 |
return 0; |
ca8bd38bf sys_swapon: separ... |
2835 2836 2837 2838 2839 2840 2841 |
} /* 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... |
2842 2843 |
if (swap_header->info.nr_badpages > MAX_SWAP_BADPAGES) return 0; |
ca8bd38bf sys_swapon: separ... |
2844 2845 2846 2847 2848 |
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... |
2849 2850 2851 |
pr_warn("Unable to handle swap header version %d ", swap_header->info.version); |
387190253 sys_swapon: simpl... |
2852 |
return 0; |
ca8bd38bf sys_swapon: separ... |
2853 2854 2855 2856 2857 |
} p->lowest_bit = 1; p->cluster_next = 1; p->cluster_nr = 0; |
377eeaa8e x86/speculation/l... |
2858 |
maxpages = max_swapfile_size(); |
d6bbbd29b swap: warn when a... |
2859 |
last_page = swap_header->info.last_page; |
a06ad633a swap: divide-by-z... |
2860 2861 2862 2863 2864 |
if (!last_page) { pr_warn("Empty swap-file "); return 0; } |
d6bbbd29b swap: warn when a... |
2865 |
if (last_page > maxpages) { |
465c47fd8 mm/swapfile.c: co... |
2866 2867 |
pr_warn("Truncating oversized swap area, only using %luk out of %luk ", |
d6bbbd29b swap: warn when a... |
2868 2869 2870 2871 2872 |
maxpages << (PAGE_SHIFT - 10), last_page << (PAGE_SHIFT - 10)); } if (maxpages > last_page) { maxpages = last_page + 1; |
ca8bd38bf sys_swapon: separ... |
2873 2874 2875 2876 2877 2878 2879 |
/* 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... |
2880 |
return 0; |
ca8bd38bf sys_swapon: separ... |
2881 2882 |
swapfilepages = i_size_read(inode) >> PAGE_SHIFT; if (swapfilepages && maxpages > swapfilepages) { |
465c47fd8 mm/swapfile.c: co... |
2883 2884 |
pr_warn("Swap area shorter than signature indicates "); |
387190253 sys_swapon: simpl... |
2885 |
return 0; |
ca8bd38bf sys_swapon: separ... |
2886 2887 |
} if (swap_header->info.nr_badpages && S_ISREG(inode->i_mode)) |
387190253 sys_swapon: simpl... |
2888 |
return 0; |
ca8bd38bf sys_swapon: separ... |
2889 |
if (swap_header->info.nr_badpages > MAX_SWAP_BADPAGES) |
387190253 sys_swapon: simpl... |
2890 |
return 0; |
ca8bd38bf sys_swapon: separ... |
2891 2892 |
return maxpages; |
ca8bd38bf sys_swapon: separ... |
2893 |
} |
4b3ef9daa mm/swap: split sw... |
2894 |
#define SWAP_CLUSTER_INFO_COLS \ |
235b62176 mm/swap: add clus... |
2895 |
DIV_ROUND_UP(L1_CACHE_BYTES, sizeof(struct swap_cluster_info)) |
4b3ef9daa mm/swap: split sw... |
2896 2897 2898 2899 |
#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... |
2900 |
|
915d4d7bc sys_swapon: separ... |
2901 2902 2903 |
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... |
2904 |
struct swap_cluster_info *cluster_info, |
915d4d7bc sys_swapon: separ... |
2905 2906 2907 |
unsigned long maxpages, sector_t *span) { |
235b62176 mm/swap: add clus... |
2908 |
unsigned int j, k; |
915d4d7bc sys_swapon: separ... |
2909 2910 |
unsigned int nr_good_pages; int nr_extents; |
2a8f94493 swap: change bloc... |
2911 |
unsigned long nr_clusters = DIV_ROUND_UP(maxpages, SWAPFILE_CLUSTER); |
235b62176 mm/swap: add clus... |
2912 2913 |
unsigned long col = p->cluster_next / SWAPFILE_CLUSTER % SWAP_CLUSTER_COLS; unsigned long i, idx; |
915d4d7bc sys_swapon: separ... |
2914 2915 |
nr_good_pages = maxpages - 1; /* omit header page */ |
6b5349159 mm, swap: add swa... |
2916 2917 |
cluster_list_init(&p->free_clusters); cluster_list_init(&p->discard_clusters); |
2a8f94493 swap: change bloc... |
2918 |
|
915d4d7bc sys_swapon: separ... |
2919 2920 |
for (i = 0; i < swap_header->info.nr_badpages; i++) { unsigned int page_nr = swap_header->info.badpages[i]; |
bdb8e3f68 sys_swapon: simpl... |
2921 2922 |
if (page_nr == 0 || page_nr > swap_header->info.last_page) return -EINVAL; |
915d4d7bc sys_swapon: separ... |
2923 2924 2925 |
if (page_nr < maxpages) { swap_map[page_nr] = SWAP_MAP_BAD; nr_good_pages--; |
2a8f94493 swap: change bloc... |
2926 2927 2928 2929 2930 |
/* * Haven't marked the cluster free yet, no list * operation involved */ inc_cluster_info_page(p, cluster_info, page_nr); |
915d4d7bc sys_swapon: separ... |
2931 2932 |
} } |
2a8f94493 swap: change bloc... |
2933 2934 2935 |
/* 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... |
2936 2937 |
if (nr_good_pages) { swap_map[0] = SWAP_MAP_BAD; |
2a8f94493 swap: change bloc... |
2938 2939 2940 2941 2942 |
/* * Not mark the cluster free yet, no list * operation involved */ inc_cluster_info_page(p, cluster_info, 0); |
915d4d7bc sys_swapon: separ... |
2943 2944 2945 |
p->max = maxpages; p->pages = nr_good_pages; nr_extents = setup_swap_extents(p, span); |
bdb8e3f68 sys_swapon: simpl... |
2946 2947 |
if (nr_extents < 0) return nr_extents; |
915d4d7bc sys_swapon: separ... |
2948 2949 2950 |
nr_good_pages = p->pages; } if (!nr_good_pages) { |
465c47fd8 mm/swapfile.c: co... |
2951 2952 |
pr_warn("Empty swap-file "); |
bdb8e3f68 sys_swapon: simpl... |
2953 |
return -EINVAL; |
915d4d7bc sys_swapon: separ... |
2954 |
} |
2a8f94493 swap: change bloc... |
2955 2956 |
if (!cluster_info) return nr_extents; |
235b62176 mm/swap: add clus... |
2957 |
|
4b3ef9daa mm/swap: split sw... |
2958 2959 2960 2961 |
/* * Reduce false cache line sharing between cluster_info and * sharing same address space. */ |
235b62176 mm/swap: add clus... |
2962 2963 2964 2965 2966 2967 2968 2969 |
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... |
2970 |
cluster_set_flag(&cluster_info[idx], CLUSTER_FLAG_FREE); |
6b5349159 mm, swap: add swa... |
2971 2972 |
cluster_list_add_tail(&p->free_clusters, cluster_info, idx); |
2a8f94493 swap: change bloc... |
2973 |
} |
2a8f94493 swap: change bloc... |
2974 |
} |
915d4d7bc sys_swapon: separ... |
2975 |
return nr_extents; |
915d4d7bc sys_swapon: separ... |
2976 |
} |
dcf6b7ddd swap: discard whi... |
2977 2978 2979 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 |
/* * 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... |
2990 2991 2992 |
SYSCALL_DEFINE2(swapon, const char __user *, specialfile, int, swap_flags) { struct swap_info_struct *p; |
91a27b2a7 vfs: define struc... |
2993 |
struct filename *name; |
53cbb2435 sys_swapon: separ... |
2994 2995 |
struct file *swap_file = NULL; struct address_space *mapping; |
40531542e sys_swapon: separ... |
2996 |
int prio; |
53cbb2435 sys_swapon: separ... |
2997 2998 |
int error; union swap_header *swap_header; |
915d4d7bc sys_swapon: separ... |
2999 |
int nr_extents; |
53cbb2435 sys_swapon: separ... |
3000 3001 |
sector_t span; unsigned long maxpages; |
53cbb2435 sys_swapon: separ... |
3002 |
unsigned char *swap_map = NULL; |
2a8f94493 swap: change bloc... |
3003 |
struct swap_cluster_info *cluster_info = NULL; |
38b5faf4b mm: frontswap: co... |
3004 |
unsigned long *frontswap_map = NULL; |
53cbb2435 sys_swapon: separ... |
3005 3006 |
struct page *page = NULL; struct inode *inode = NULL; |
7cbf31923 mm: fix nr_rotate... |
3007 |
bool inced_nr_rotate_swap = false; |
53cbb2435 sys_swapon: separ... |
3008 |
|
d15cab975 swapon: check val... |
3009 3010 |
if (swap_flags & ~SWAP_FLAGS_VALID) return -EINVAL; |
53cbb2435 sys_swapon: separ... |
3011 3012 |
if (!capable(CAP_SYS_ADMIN)) return -EPERM; |
a2468cc9b swap: choose swap... |
3013 3014 |
if (!swap_avail_heads) return -ENOMEM; |
53cbb2435 sys_swapon: separ... |
3015 |
p = alloc_swap_info(); |
2542e5134 sys_swapon: simpl... |
3016 3017 |
if (IS_ERR(p)) return PTR_ERR(p); |
53cbb2435 sys_swapon: separ... |
3018 |
|
815c2c543 swap: make swap d... |
3019 |
INIT_WORK(&p->discard_work, swap_discard_work); |
1da177e4c Linux-2.6.12-rc2 |
3020 |
name = getname(specialfile); |
1da177e4c Linux-2.6.12-rc2 |
3021 |
if (IS_ERR(name)) { |
7de7fb6b3 sys_swapon: move ... |
3022 |
error = PTR_ERR(name); |
1da177e4c Linux-2.6.12-rc2 |
3023 |
name = NULL; |
bd69010b0 sys_swapon: use a... |
3024 |
goto bad_swap; |
1da177e4c Linux-2.6.12-rc2 |
3025 |
} |
669abf4e5 vfs: make path_op... |
3026 |
swap_file = file_open_name(name, O_RDWR|O_LARGEFILE, 0); |
1da177e4c Linux-2.6.12-rc2 |
3027 |
if (IS_ERR(swap_file)) { |
7de7fb6b3 sys_swapon: move ... |
3028 |
error = PTR_ERR(swap_file); |
1da177e4c Linux-2.6.12-rc2 |
3029 |
swap_file = NULL; |
bd69010b0 sys_swapon: use a... |
3030 |
goto bad_swap; |
1da177e4c Linux-2.6.12-rc2 |
3031 3032 3033 3034 |
} p->swap_file = swap_file; mapping = swap_file->f_mapping; |
2130781e2 sys_swapon: fix i... |
3035 |
inode = mapping->host; |
6f179af88 mm: fix potential... |
3036 |
|
5955102c9 wrappers for ->i_... |
3037 |
/* If S_ISREG(inode->i_mode) will do inode_lock(inode); */ |
4d0e1e107 sys_swapon: separ... |
3038 3039 |
error = claim_swapfile(p, inode); if (unlikely(error)) |
1da177e4c Linux-2.6.12-rc2 |
3040 |
goto bad_swap; |
1da177e4c Linux-2.6.12-rc2 |
3041 |
|
1da177e4c Linux-2.6.12-rc2 |
3042 3043 3044 3045 3046 3047 3048 |
/* * Read the swap header. */ if (!mapping->a_ops->readpage) { error = -EINVAL; goto bad_swap; } |
090d2b185 [PATCH] read_mapp... |
3049 |
page = read_mapping_page(mapping, 0, swap_file); |
1da177e4c Linux-2.6.12-rc2 |
3050 3051 3052 3053 |
if (IS_ERR(page)) { error = PTR_ERR(page); goto bad_swap; } |
81e339712 swapfile: remove ... |
3054 |
swap_header = kmap(page); |
1da177e4c Linux-2.6.12-rc2 |
3055 |
|
ca8bd38bf sys_swapon: separ... |
3056 3057 |
maxpages = read_swap_header(p, swap_header, inode); if (unlikely(!maxpages)) { |
1da177e4c Linux-2.6.12-rc2 |
3058 3059 3060 |
error = -EINVAL; goto bad_swap; } |
886bb7e9c swapfile: remove ... |
3061 |
|
81e339712 swapfile: remove ... |
3062 |
/* OK, set up the swap map and apply the bad block list */ |
803d0c835 sys_swapon: use v... |
3063 |
swap_map = vzalloc(maxpages); |
81e339712 swapfile: remove ... |
3064 3065 3066 3067 |
if (!swap_map) { error = -ENOMEM; goto bad_swap; } |
f05714293 mm: support anony... |
3068 3069 3070 |
if (bdi_cap_stable_pages_required(inode_to_bdi(inode))) p->flags |= SWP_STABLE_WRITES; |
539a6fea7 mm, swap: introdu... |
3071 3072 |
if (bdi_cap_synchronous_io(inode_to_bdi(inode))) p->flags |= SWP_SYNCHRONOUS_IO; |
2a8f94493 swap: change bloc... |
3073 |
if (p->bdev && blk_queue_nonrot(bdev_get_queue(p->bdev))) { |
6f179af88 mm: fix potential... |
3074 |
int cpu; |
235b62176 mm/swap: add clus... |
3075 |
unsigned long ci, nr_cluster; |
6f179af88 mm: fix potential... |
3076 |
|
2a8f94493 swap: change bloc... |
3077 3078 3079 3080 3081 3082 |
p->flags |= SWP_SOLIDSTATE; /* * select a random position to start with to help wear leveling * SSD */ p->cluster_next = 1 + (prandom_u32() % p->highest_bit); |
235b62176 mm/swap: add clus... |
3083 |
nr_cluster = DIV_ROUND_UP(maxpages, SWAPFILE_CLUSTER); |
2a8f94493 swap: change bloc... |
3084 |
|
778e1cdd8 treewide: kvzallo... |
3085 |
cluster_info = kvcalloc(nr_cluster, sizeof(*cluster_info), |
54f180d3c mm, swap: use kvz... |
3086 |
GFP_KERNEL); |
2a8f94493 swap: change bloc... |
3087 3088 3089 3090 |
if (!cluster_info) { error = -ENOMEM; goto bad_swap; } |
235b62176 mm/swap: add clus... |
3091 3092 3093 |
for (ci = 0; ci < nr_cluster; ci++) spin_lock_init(&((cluster_info + ci)->lock)); |
ebc2a1a69 swap: make cluste... |
3094 3095 3096 3097 3098 |
p->percpu_cluster = alloc_percpu(struct percpu_cluster); if (!p->percpu_cluster) { error = -ENOMEM; goto bad_swap; } |
6f179af88 mm: fix potential... |
3099 |
for_each_possible_cpu(cpu) { |
ebc2a1a69 swap: make cluste... |
3100 |
struct percpu_cluster *cluster; |
6f179af88 mm: fix potential... |
3101 |
cluster = per_cpu_ptr(p->percpu_cluster, cpu); |
ebc2a1a69 swap: make cluste... |
3102 3103 |
cluster_set_null(&cluster->index); } |
7cbf31923 mm: fix nr_rotate... |
3104 |
} else { |
81a0298bd mm, swap: don't u... |
3105 |
atomic_inc(&nr_rotate_swap); |
7cbf31923 mm: fix nr_rotate... |
3106 3107 |
inced_nr_rotate_swap = true; } |
1da177e4c Linux-2.6.12-rc2 |
3108 |
|
1421ef3cd sys_swapon: call ... |
3109 3110 3111 |
error = swap_cgroup_swapon(p->type, maxpages); if (error) goto bad_swap; |
915d4d7bc sys_swapon: separ... |
3112 |
nr_extents = setup_swap_map_and_extents(p, swap_header, swap_map, |
2a8f94493 swap: change bloc... |
3113 |
cluster_info, maxpages, &span); |
915d4d7bc sys_swapon: separ... |
3114 3115 |
if (unlikely(nr_extents < 0)) { error = nr_extents; |
1da177e4c Linux-2.6.12-rc2 |
3116 3117 |
goto bad_swap; } |
38b5faf4b mm: frontswap: co... |
3118 |
/* frontswap enabled? set up bit-per-page map for frontswap */ |
8ea1d2a19 mm, frontswap: co... |
3119 |
if (IS_ENABLED(CONFIG_FRONTSWAP)) |
778e1cdd8 treewide: kvzallo... |
3120 3121 |
frontswap_map = kvcalloc(BITS_TO_LONGS(maxpages), sizeof(long), |
54f180d3c mm, swap: use kvz... |
3122 |
GFP_KERNEL); |
1da177e4c Linux-2.6.12-rc2 |
3123 |
|
2a8f94493 swap: change bloc... |
3124 3125 3126 3127 3128 3129 3130 3131 3132 |
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... |
3133 |
|
2a8f94493 swap: change bloc... |
3134 3135 3136 3137 3138 3139 3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 3150 3151 |
/* * 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... |
3152 |
} |
20137a490 swapfile: swapon ... |
3153 |
} |
6a6ba8317 swapfile: swapon ... |
3154 |
|
4b3ef9daa mm/swap: split sw... |
3155 3156 3157 |
error = init_swap_address_space(p->type, maxpages); if (error) goto bad_swap; |
dc617f29d vfs: don't allow ... |
3158 3159 3160 3161 3162 3163 3164 3165 3166 3167 |
/* * 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; goto bad_swap; } |
fc0abb145 [PATCH] sem2mutex... |
3168 |
mutex_lock(&swapon_mutex); |
40531542e sys_swapon: separ... |
3169 |
prio = -1; |
78ecba081 mm: fix ever-decr... |
3170 |
if (swap_flags & SWAP_FLAG_PREFER) |
40531542e sys_swapon: separ... |
3171 |
prio = |
78ecba081 mm: fix ever-decr... |
3172 |
(swap_flags & SWAP_FLAG_PRIO_MASK) >> SWAP_FLAG_PRIO_SHIFT; |
2a8f94493 swap: change bloc... |
3173 |
enable_swap_info(p, prio, swap_map, cluster_info, frontswap_map); |
c69dbfb84 sys_swapon: move ... |
3174 |
|
756a025f0 mm: coalesce spli... |
3175 3176 |
pr_info("Adding %uk swap on %s. Priority:%d extents:%d across:%lluk %s%s%s%s%s ", |
91a27b2a7 vfs: define struc... |
3177 |
p->pages<<(PAGE_SHIFT-10), name->name, p->prio, |
c69dbfb84 sys_swapon: move ... |
3178 3179 |
nr_extents, (unsigned long long)span<<(PAGE_SHIFT-10), (p->flags & SWP_SOLIDSTATE) ? "SS" : "", |
38b5faf4b mm: frontswap: co... |
3180 |
(p->flags & SWP_DISCARDABLE) ? "D" : "", |
dcf6b7ddd swap: discard whi... |
3181 3182 |
(p->flags & SWP_AREA_DISCARD) ? "s" : "", (p->flags & SWP_PAGE_DISCARD) ? "c" : "", |
38b5faf4b mm: frontswap: co... |
3183 |
(frontswap_map) ? "FS" : ""); |
c69dbfb84 sys_swapon: move ... |
3184 |
|
fc0abb145 [PATCH] sem2mutex... |
3185 |
mutex_unlock(&swapon_mutex); |
66d7dd518 /proc/swaps: supp... |
3186 3187 |
atomic_inc(&proc_poll_event); wake_up_interruptible(&proc_poll_wait); |
1da177e4c Linux-2.6.12-rc2 |
3188 3189 3190 |
error = 0; goto out; bad_swap: |
ebc2a1a69 swap: make cluste... |
3191 3192 |
free_percpu(p->percpu_cluster); p->percpu_cluster = NULL; |
bd69010b0 sys_swapon: use a... |
3193 |
if (inode && S_ISBLK(inode->i_mode) && p->bdev) { |
f2090d2df sys_swapon: remov... |
3194 3195 |
set_blocksize(p->bdev, p->old_block_size); blkdev_put(p->bdev, FMODE_READ | FMODE_WRITE | FMODE_EXCL); |
1da177e4c Linux-2.6.12-rc2 |
3196 |
} |
4cd3bb10f [PATCH] swap: mov... |
3197 |
destroy_swap_extents(p); |
e8e6c2ec4 sys_swapon: do no... |
3198 |
swap_cgroup_swapoff(p->type); |
5d337b919 [PATCH] swap: swa... |
3199 |
spin_lock(&swap_lock); |
1da177e4c Linux-2.6.12-rc2 |
3200 |
p->swap_file = NULL; |
1da177e4c Linux-2.6.12-rc2 |
3201 |
p->flags = 0; |
5d337b919 [PATCH] swap: swa... |
3202 |
spin_unlock(&swap_lock); |
1da177e4c Linux-2.6.12-rc2 |
3203 |
vfree(swap_map); |
8606a1a94 mm: kvfree the sw... |
3204 |
kvfree(cluster_info); |
b6b1fd2a6 mm/swapfile.c: fi... |
3205 |
kvfree(frontswap_map); |
7cbf31923 mm: fix nr_rotate... |
3206 3207 |
if (inced_nr_rotate_swap) atomic_dec(&nr_rotate_swap); |
52c50567d mm: swap: unlock ... |
3208 |
if (swap_file) { |
1638045c3 mm: set S_SWAPFIL... |
3209 |
if (inode) { |
5955102c9 wrappers for ->i_... |
3210 |
inode_unlock(inode); |
2130781e2 sys_swapon: fix i... |
3211 3212 |
inode = NULL; } |
1da177e4c Linux-2.6.12-rc2 |
3213 |
filp_close(swap_file, NULL); |
52c50567d mm: swap: unlock ... |
3214 |
} |
1da177e4c Linux-2.6.12-rc2 |
3215 3216 3217 |
out: if (page && !IS_ERR(page)) { kunmap(page); |
09cbfeaf1 mm, fs: get rid o... |
3218 |
put_page(page); |
1da177e4c Linux-2.6.12-rc2 |
3219 3220 3221 |
} if (name) putname(name); |
1638045c3 mm: set S_SWAPFIL... |
3222 |
if (inode) |
5955102c9 wrappers for ->i_... |
3223 |
inode_unlock(inode); |
039939a65 mm/swap: enable s... |
3224 3225 |
if (!error) enable_swap_slots_cache(); |
1da177e4c Linux-2.6.12-rc2 |
3226 3227 3228 3229 3230 |
return error; } void si_swapinfo(struct sysinfo *val) { |
efa90a981 swap_info: change... |
3231 |
unsigned int type; |
1da177e4c Linux-2.6.12-rc2 |
3232 |
unsigned long nr_to_be_unused = 0; |
5d337b919 [PATCH] swap: swa... |
3233 |
spin_lock(&swap_lock); |
efa90a981 swap_info: change... |
3234 3235 3236 3237 3238 |
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 |
3239 |
} |
ec8acf20a swap: add per-par... |
3240 |
val->freeswap = atomic_long_read(&nr_swap_pages) + nr_to_be_unused; |
1da177e4c Linux-2.6.12-rc2 |
3241 |
val->totalswap = total_swap_pages + nr_to_be_unused; |
5d337b919 [PATCH] swap: swa... |
3242 |
spin_unlock(&swap_lock); |
1da177e4c Linux-2.6.12-rc2 |
3243 3244 3245 3246 3247 |
} /* * Verify that a swap entry is valid and increment its swap map count. * |
355cfa73d mm: modify swap_m... |
3248 3249 3250 3251 3252 3253 |
* 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... |
3254 |
* - swap-mapped reference requested but needs continued swap count. -> ENOMEM |
1da177e4c Linux-2.6.12-rc2 |
3255 |
*/ |
8d69aaee8 swap_info: swap_m... |
3256 |
static int __swap_duplicate(swp_entry_t entry, unsigned char usage) |
1da177e4c Linux-2.6.12-rc2 |
3257 |
{ |
73c34b6ac swap_info: miscel... |
3258 |
struct swap_info_struct *p; |
235b62176 mm/swap: add clus... |
3259 |
struct swap_cluster_info *ci; |
c10d38cc8 mm, swap: bounds ... |
3260 |
unsigned long offset; |
8d69aaee8 swap_info: swap_m... |
3261 3262 |
unsigned char count; unsigned char has_cache; |
253d553ba swap_info: SWAP_H... |
3263 |
int err = -EINVAL; |
1da177e4c Linux-2.6.12-rc2 |
3264 |
|
eb085574a mm, swap: fix rac... |
3265 |
p = get_swap_device(entry); |
c10d38cc8 mm, swap: bounds ... |
3266 |
if (!p) |
235b62176 mm/swap: add clus... |
3267 |
goto out; |
eb085574a mm, swap: fix rac... |
3268 |
offset = swp_offset(entry); |
235b62176 mm/swap: add clus... |
3269 |
ci = lock_cluster_or_swap_info(p, offset); |
355cfa73d mm: modify swap_m... |
3270 |
|
253d553ba swap_info: SWAP_H... |
3271 |
count = p->swap_map[offset]; |
edfe23dac swap: fix races e... |
3272 3273 3274 3275 3276 3277 3278 3279 3280 |
/* * 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... |
3281 3282 3283 |
has_cache = count & SWAP_HAS_CACHE; count &= ~SWAP_HAS_CACHE; err = 0; |
355cfa73d mm: modify swap_m... |
3284 |
|
253d553ba swap_info: SWAP_H... |
3285 |
if (usage == SWAP_HAS_CACHE) { |
355cfa73d mm: modify swap_m... |
3286 3287 |
/* set SWAP_HAS_CACHE if there is no cache and entry is used */ |
253d553ba swap_info: SWAP_H... |
3288 3289 3290 3291 3292 3293 |
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... |
3294 3295 |
} else if (count || has_cache) { |
253d553ba swap_info: SWAP_H... |
3296 |
|
570a335b8 swap_info: swap c... |
3297 3298 3299 |
if ((count & ~COUNT_CONTINUED) < SWAP_MAP_MAX) count += usage; else if ((count & ~COUNT_CONTINUED) > SWAP_MAP_MAX) |
253d553ba swap_info: SWAP_H... |
3300 |
err = -EINVAL; |
570a335b8 swap_info: swap c... |
3301 3302 3303 3304 |
else if (swap_count_continued(p, offset, count)) count = COUNT_CONTINUED; else err = -ENOMEM; |
355cfa73d mm: modify swap_m... |
3305 |
} else |
253d553ba swap_info: SWAP_H... |
3306 3307 3308 |
err = -ENOENT; /* unused swap entry */ p->swap_map[offset] = count | has_cache; |
355cfa73d mm: modify swap_m... |
3309 |
unlock_out: |
235b62176 mm/swap: add clus... |
3310 |
unlock_cluster_or_swap_info(p, ci); |
1da177e4c Linux-2.6.12-rc2 |
3311 |
out: |
eb085574a mm, swap: fix rac... |
3312 3313 |
if (p) put_swap_device(p); |
253d553ba swap_info: SWAP_H... |
3314 |
return err; |
1da177e4c Linux-2.6.12-rc2 |
3315 |
} |
253d553ba swap_info: SWAP_H... |
3316 |
|
355cfa73d mm: modify swap_m... |
3317 |
/* |
aaa468653 swap_info: note S... |
3318 3319 3320 3321 3322 3323 3324 3325 3326 |
* 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... |
3327 3328 3329 3330 3331 |
* 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... |
3332 |
*/ |
570a335b8 swap_info: swap c... |
3333 |
int swap_duplicate(swp_entry_t entry) |
355cfa73d mm: modify swap_m... |
3334 |
{ |
570a335b8 swap_info: swap c... |
3335 3336 3337 3338 3339 |
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... |
3340 |
} |
1da177e4c Linux-2.6.12-rc2 |
3341 |
|
cb4b86ba4 mm: add swap cach... |
3342 |
/* |
355cfa73d mm: modify swap_m... |
3343 3344 |
* @entry: swap entry for which we allocate swap cache. * |
73c34b6ac swap_info: miscel... |
3345 |
* Called when allocating swap cache for existing swap entry, |
355cfa73d mm: modify swap_m... |
3346 3347 3348 |
* This can return error codes. Returns 0 at success. * -EBUSY means there is a swap cache. * Note: return code is different from swap_duplicate(). |
cb4b86ba4 mm: add swap cach... |
3349 3350 3351 |
*/ int swapcache_prepare(swp_entry_t entry) { |
253d553ba swap_info: SWAP_H... |
3352 |
return __swap_duplicate(entry, SWAP_HAS_CACHE); |
cb4b86ba4 mm: add swap cach... |
3353 |
} |
0bcac06f2 mm, swap: skip sw... |
3354 3355 |
struct swap_info_struct *swp_swap_info(swp_entry_t entry) { |
c10d38cc8 mm, swap: bounds ... |
3356 |
return swap_type_to_swap_info(swp_type(entry)); |
0bcac06f2 mm, swap: skip sw... |
3357 |
} |
f981c5950 mm: methods for t... |
3358 3359 |
struct swap_info_struct *page_swap_info(struct page *page) { |
0bcac06f2 mm, swap: skip sw... |
3360 3361 |
swp_entry_t entry = { .val = page_private(page) }; return swp_swap_info(entry); |
f981c5950 mm: methods for t... |
3362 3363 3364 3365 3366 3367 3368 |
} /* * out-of-line __page_file_ methods to avoid include hell. */ struct address_space *__page_file_mapping(struct page *page) { |
f981c5950 mm: methods for t... |
3369 3370 3371 3372 3373 3374 3375 |
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... |
3376 3377 3378 |
return swp_offset(swap); } EXPORT_SYMBOL_GPL(__page_file_index); |
1da177e4c Linux-2.6.12-rc2 |
3379 |
/* |
570a335b8 swap_info: swap c... |
3380 3381 3382 3383 3384 3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 |
* 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... |
3397 |
struct swap_cluster_info *ci; |
570a335b8 swap_info: swap c... |
3398 3399 3400 3401 3402 |
struct page *head; struct page *page; struct page *list_page; pgoff_t offset; unsigned char count; |
eb085574a mm, swap: fix rac... |
3403 |
int ret = 0; |
570a335b8 swap_info: swap c... |
3404 3405 3406 3407 3408 3409 |
/* * 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... |
3410 |
si = get_swap_device(entry); |
570a335b8 swap_info: swap c... |
3411 3412 3413 |
if (!si) { /* * An acceptable race has occurred since the failing |
eb085574a mm, swap: fix rac... |
3414 |
* __swap_duplicate(): the swap device may be swapoff |
570a335b8 swap_info: swap c... |
3415 3416 3417 |
*/ goto outer; } |
eb085574a mm, swap: fix rac... |
3418 |
spin_lock(&si->lock); |
570a335b8 swap_info: swap c... |
3419 3420 |
offset = swp_offset(entry); |
235b62176 mm/swap: add clus... |
3421 3422 |
ci = lock_cluster(si, offset); |
570a335b8 swap_info: swap c... |
3423 3424 3425 3426 3427 3428 3429 3430 3431 3432 3433 3434 |
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... |
3435 3436 |
ret = -ENOMEM; goto out; |
570a335b8 swap_info: swap c... |
3437 3438 3439 3440 |
} /* * We are fortunate that although vmalloc_to_page uses pte_offset_map, |
2de1a7e40 mm/swapfile.c: fi... |
3441 3442 |
* 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... |
3443 3444 3445 |
*/ head = vmalloc_to_page(si->swap_map + offset); offset &= ~PAGE_MASK; |
2628bd6fc mm, swap: fix rac... |
3446 |
spin_lock(&si->cont_lock); |
570a335b8 swap_info: swap c... |
3447 3448 3449 3450 3451 3452 3453 3454 3455 3456 3457 3458 3459 3460 3461 3462 3463 3464 3465 |
/* * 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... |
3466 |
goto out_unlock_cont; |
570a335b8 swap_info: swap c... |
3467 |
|
9b04c5fec mm: remove the se... |
3468 |
map = kmap_atomic(list_page) + offset; |
570a335b8 swap_info: swap c... |
3469 |
count = *map; |
9b04c5fec mm: remove the se... |
3470 |
kunmap_atomic(map); |
570a335b8 swap_info: swap c... |
3471 3472 3473 3474 3475 3476 |
/* * 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... |
3477 |
goto out_unlock_cont; |
570a335b8 swap_info: swap c... |
3478 3479 3480 3481 |
} list_add_tail(&page->lru, &head->lru); page = NULL; /* now it's attached, don't free it */ |
2628bd6fc mm, swap: fix rac... |
3482 3483 |
out_unlock_cont: spin_unlock(&si->cont_lock); |
570a335b8 swap_info: swap c... |
3484 |
out: |
235b62176 mm/swap: add clus... |
3485 |
unlock_cluster(ci); |
ec8acf20a swap: add per-par... |
3486 |
spin_unlock(&si->lock); |
eb085574a mm, swap: fix rac... |
3487 |
put_swap_device(si); |
570a335b8 swap_info: swap c... |
3488 3489 3490 |
outer: if (page) __free_page(page); |
eb085574a mm, swap: fix rac... |
3491 |
return ret; |
570a335b8 swap_info: swap c... |
3492 3493 3494 3495 3496 3497 3498 3499 |
} /* * 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... |
3500 3501 |
* Called while __swap_duplicate() or swap_entry_free() holds swap or cluster * lock. |
570a335b8 swap_info: swap c... |
3502 3503 3504 3505 3506 3507 3508 |
*/ 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... |
3509 |
bool ret; |
570a335b8 swap_info: swap c... |
3510 3511 3512 3513 3514 3515 |
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... |
3516 |
spin_lock(&si->cont_lock); |
570a335b8 swap_info: swap c... |
3517 3518 |
offset &= ~PAGE_MASK; page = list_entry(head->lru.next, struct page, lru); |
9b04c5fec mm: remove the se... |
3519 |
map = kmap_atomic(page) + offset; |
570a335b8 swap_info: swap c... |
3520 3521 3522 3523 3524 3525 3526 3527 3528 |
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... |
3529 |
kunmap_atomic(map); |
570a335b8 swap_info: swap c... |
3530 3531 |
page = list_entry(page->lru.next, struct page, lru); BUG_ON(page == head); |
9b04c5fec mm: remove the se... |
3532 |
map = kmap_atomic(page) + offset; |
570a335b8 swap_info: swap c... |
3533 3534 |
} if (*map == SWAP_CONT_MAX) { |
9b04c5fec mm: remove the se... |
3535 |
kunmap_atomic(map); |
570a335b8 swap_info: swap c... |
3536 |
page = list_entry(page->lru.next, struct page, lru); |
2628bd6fc mm, swap: fix rac... |
3537 3538 3539 3540 |
if (page == head) { ret = false; /* add count continuation */ goto out; } |
9b04c5fec mm: remove the se... |
3541 |
map = kmap_atomic(page) + offset; |
570a335b8 swap_info: swap c... |
3542 3543 3544 |
init_map: *map = 0; /* we didn't zero the page */ } *map += 1; |
9b04c5fec mm: remove the se... |
3545 |
kunmap_atomic(map); |
570a335b8 swap_info: swap c... |
3546 3547 |
page = list_entry(page->lru.prev, struct page, lru); while (page != head) { |
9b04c5fec mm: remove the se... |
3548 |
map = kmap_atomic(page) + offset; |
570a335b8 swap_info: swap c... |
3549 |
*map = COUNT_CONTINUED; |
9b04c5fec mm: remove the se... |
3550 |
kunmap_atomic(map); |
570a335b8 swap_info: swap c... |
3551 3552 |
page = list_entry(page->lru.prev, struct page, lru); } |
2628bd6fc mm, swap: fix rac... |
3553 |
ret = true; /* incremented */ |
570a335b8 swap_info: swap c... |
3554 3555 3556 3557 3558 3559 3560 |
} else { /* decrementing */ /* * Think of how you subtract 1 from 1000 */ BUG_ON(count != COUNT_CONTINUED); while (*map == COUNT_CONTINUED) { |
9b04c5fec mm: remove the se... |
3561 |
kunmap_atomic(map); |
570a335b8 swap_info: swap c... |
3562 3563 |
page = list_entry(page->lru.next, struct page, lru); BUG_ON(page == head); |
9b04c5fec mm: remove the se... |
3564 |
map = kmap_atomic(page) + offset; |
570a335b8 swap_info: swap c... |
3565 3566 3567 3568 3569 |
} BUG_ON(*map == 0); *map -= 1; if (*map == 0) count = 0; |
9b04c5fec mm: remove the se... |
3570 |
kunmap_atomic(map); |
570a335b8 swap_info: swap c... |
3571 3572 |
page = list_entry(page->lru.prev, struct page, lru); while (page != head) { |
9b04c5fec mm: remove the se... |
3573 |
map = kmap_atomic(page) + offset; |
570a335b8 swap_info: swap c... |
3574 3575 |
*map = SWAP_CONT_MAX | count; count = COUNT_CONTINUED; |
9b04c5fec mm: remove the se... |
3576 |
kunmap_atomic(map); |
570a335b8 swap_info: swap c... |
3577 3578 |
page = list_entry(page->lru.prev, struct page, lru); } |
2628bd6fc mm, swap: fix rac... |
3579 |
ret = count == COUNT_CONTINUED; |
570a335b8 swap_info: swap c... |
3580 |
} |
2628bd6fc mm, swap: fix rac... |
3581 3582 3583 |
out: spin_unlock(&si->cont_lock); return ret; |
570a335b8 swap_info: swap c... |
3584 3585 3586 3587 3588 3589 3590 3591 3592 3593 3594 3595 3596 3597 |
} /* * 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... |
3598 3599 3600 3601 |
struct page *page, *next; list_for_each_entry_safe(page, next, &head->lru, lru) { list_del(&page->lru); |
570a335b8 swap_info: swap c... |
3602 3603 3604 3605 3606 |
__free_page(page); } } } } |
a2468cc9b swap: choose swap... |
3607 |
|
2cf855837 memcontrol: sched... |
3608 3609 3610 3611 3612 3613 3614 3615 3616 3617 3618 3619 3620 3621 3622 3623 3624 3625 3626 3627 3628 3629 3630 3631 3632 3633 3634 3635 3636 3637 |
#if defined(CONFIG_MEMCG) && defined(CONFIG_BLK_CGROUP) void mem_cgroup_throttle_swaprate(struct mem_cgroup *memcg, int node, gfp_t gfp_mask) { struct swap_info_struct *si, *next; if (!(gfp_mask & __GFP_IO) || !memcg) 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); plist_for_each_entry_safe(si, next, &swap_avail_heads[node], avail_lists[node]) { if (si->bdev) { blkcg_schedule_throttle(bdev_get_queue(si->bdev), true); break; } } spin_unlock(&swap_avail_lock); } #endif |
a2468cc9b swap: choose swap... |
3638 3639 3640 3641 3642 3643 3644 3645 3646 3647 3648 3649 3650 3651 3652 3653 3654 3655 |
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); |