Blame view
mm/memcontrol.c
128 KB
8cdea7c05
|
1 2 3 4 5 |
/* memcontrol.c - Memory Controller * * Copyright IBM Corporation, 2007 * Author Balbir Singh <balbir@linux.vnet.ibm.com> * |
78fb74669
|
6 7 8 |
* Copyright 2007 OpenVZ SWsoft Inc * Author: Pavel Emelianov <xemul@openvz.org> * |
2e72b6347
|
9 10 11 12 |
* Memory thresholds * Copyright (C) 2009 Nokia Corporation * Author: Kirill A. Shutemov * |
8cdea7c05
|
13 14 15 16 17 18 19 20 21 22 23 24 25 26 |
* This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. */ #include <linux/res_counter.h> #include <linux/memcontrol.h> #include <linux/cgroup.h> |
78fb74669
|
27 |
#include <linux/mm.h> |
4ffef5fef
|
28 |
#include <linux/hugetlb.h> |
d13d14430
|
29 |
#include <linux/pagemap.h> |
d52aa412d
|
30 |
#include <linux/smp.h> |
8a9f3ccd2
|
31 |
#include <linux/page-flags.h> |
66e1707bc
|
32 |
#include <linux/backing-dev.h> |
8a9f3ccd2
|
33 34 |
#include <linux/bit_spinlock.h> #include <linux/rcupdate.h> |
e222432bf
|
35 |
#include <linux/limits.h> |
8c7c6e34a
|
36 |
#include <linux/mutex.h> |
f64c3f549
|
37 |
#include <linux/rbtree.h> |
b6ac57d50
|
38 |
#include <linux/slab.h> |
66e1707bc
|
39 |
#include <linux/swap.h> |
024914477
|
40 |
#include <linux/swapops.h> |
66e1707bc
|
41 |
#include <linux/spinlock.h> |
2e72b6347
|
42 43 |
#include <linux/eventfd.h> #include <linux/sort.h> |
66e1707bc
|
44 |
#include <linux/fs.h> |
d2ceb9b7d
|
45 |
#include <linux/seq_file.h> |
333279487
|
46 |
#include <linux/vmalloc.h> |
b69408e88
|
47 |
#include <linux/mm_inline.h> |
52d4b9ac0
|
48 |
#include <linux/page_cgroup.h> |
cdec2e426
|
49 |
#include <linux/cpu.h> |
158e0a2d1
|
50 |
#include <linux/oom.h> |
08e552c69
|
51 |
#include "internal.h" |
8cdea7c05
|
52 |
|
8697d3319
|
53 |
#include <asm/uaccess.h> |
cc8e970c3
|
54 |
#include <trace/events/vmscan.h> |
a181b0e88
|
55 |
struct cgroup_subsys mem_cgroup_subsys __read_mostly; |
a181b0e88
|
56 |
#define MEM_CGROUP_RECLAIM_RETRIES 5 |
4b3bde4c9
|
57 |
struct mem_cgroup *root_mem_cgroup __read_mostly; |
8cdea7c05
|
58 |
|
c077719be
|
59 |
#ifdef CONFIG_CGROUP_MEM_RES_CTLR_SWAP |
338c84310
|
60 |
/* Turned on only when memory cgroup is enabled && really_do_swap_account = 1 */ |
c077719be
|
61 |
int do_swap_account __read_mostly; |
a42c390cf
|
62 63 64 65 66 67 68 |
/* for remember boot option*/ #ifdef CONFIG_CGROUP_MEM_RES_CTLR_SWAP_ENABLED static int really_do_swap_account __initdata = 1; #else static int really_do_swap_account __initdata = 0; #endif |
c077719be
|
69 70 71 |
#else #define do_swap_account (0) #endif |
d2265e6fa
|
72 73 74 75 76 77 78 79 80 |
/* * Per memcg event counter is incremented at every pagein/pageout. This counter * is used for trigger some periodic events. This is straightforward and better * than using jiffies etc. to handle periodic memcg event. * * These values will be used as !((event) & ((1 <<(thresh)) - 1)) */ #define THRESHOLDS_EVENTS_THRESH (7) /* once in 128 */ #define SOFTLIMIT_EVENTS_THRESH (10) /* once in 1024 */ |
c077719be
|
81 |
|
8cdea7c05
|
82 |
/* |
d52aa412d
|
83 84 85 86 87 88 89 |
* Statistics for memory cgroup. */ enum mem_cgroup_stat_index { /* * For MEM_CONTAINER_TYPE_ALL, usage = pagecache + rss. */ MEM_CGROUP_STAT_CACHE, /* # of pages charged as cache */ |
d69b042f3
|
90 |
MEM_CGROUP_STAT_RSS, /* # of pages charged as anon rss */ |
d8046582d
|
91 |
MEM_CGROUP_STAT_FILE_MAPPED, /* # of pages charged as file rss */ |
55e462b05
|
92 93 |
MEM_CGROUP_STAT_PGPGIN_COUNT, /* # of pages paged in */ MEM_CGROUP_STAT_PGPGOUT_COUNT, /* # of pages paged out */ |
0c3e73e84
|
94 |
MEM_CGROUP_STAT_SWAPOUT, /* # of pages, swapped out */ |
711d3d2c9
|
95 96 97 |
MEM_CGROUP_STAT_DATA, /* end of data requires synchronization */ /* incremented at every pagein/pageout */ MEM_CGROUP_EVENTS = MEM_CGROUP_STAT_DATA, |
32047e2a8
|
98 |
MEM_CGROUP_ON_MOVE, /* someone is moving account between groups */ |
d52aa412d
|
99 100 101 102 103 104 |
MEM_CGROUP_STAT_NSTATS, }; struct mem_cgroup_stat_cpu { s64 count[MEM_CGROUP_STAT_NSTATS]; |
d52aa412d
|
105 |
}; |
d52aa412d
|
106 |
/* |
6d12e2d8d
|
107 108 |
* per-zone information in memory controller. */ |
6d12e2d8d
|
109 |
struct mem_cgroup_per_zone { |
072c56c13
|
110 111 112 |
/* * spin_lock to protect the per cgroup LRU */ |
b69408e88
|
113 114 |
struct list_head lists[NR_LRU_LISTS]; unsigned long count[NR_LRU_LISTS]; |
3e2f41f1f
|
115 116 |
struct zone_reclaim_stat reclaim_stat; |
f64c3f549
|
117 118 119 120 |
struct rb_node tree_node; /* RB tree node */ unsigned long long usage_in_excess;/* Set to the value by which */ /* the soft limit is exceeded*/ bool on_tree; |
4e4169535
|
121 122 |
struct mem_cgroup *mem; /* Back pointer, we cannot */ /* use container_of */ |
6d12e2d8d
|
123 124 125 126 127 128 129 130 131 132 133 134 135 |
}; /* Macro for accessing counter */ #define MEM_CGROUP_ZSTAT(mz, idx) ((mz)->count[(idx)]) struct mem_cgroup_per_node { struct mem_cgroup_per_zone zoneinfo[MAX_NR_ZONES]; }; struct mem_cgroup_lru_info { struct mem_cgroup_per_node *nodeinfo[MAX_NUMNODES]; }; /* |
f64c3f549
|
136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 |
* Cgroups above their limits are maintained in a RB-Tree, independent of * their hierarchy representation */ struct mem_cgroup_tree_per_zone { struct rb_root rb_root; spinlock_t lock; }; struct mem_cgroup_tree_per_node { struct mem_cgroup_tree_per_zone rb_tree_per_zone[MAX_NR_ZONES]; }; struct mem_cgroup_tree { struct mem_cgroup_tree_per_node *rb_tree_per_node[MAX_NUMNODES]; }; static struct mem_cgroup_tree soft_limit_tree __read_mostly; |
2e72b6347
|
154 155 156 157 |
struct mem_cgroup_threshold { struct eventfd_ctx *eventfd; u64 threshold; }; |
9490ff275
|
158 |
/* For threshold */ |
2e72b6347
|
159 160 |
struct mem_cgroup_threshold_ary { /* An array index points to threshold just below usage. */ |
5407a5625
|
161 |
int current_threshold; |
2e72b6347
|
162 163 164 165 166 |
/* Size of entries[] */ unsigned int size; /* Array of thresholds */ struct mem_cgroup_threshold entries[0]; }; |
2c488db27
|
167 168 169 170 171 172 173 174 175 176 177 |
struct mem_cgroup_thresholds { /* Primary thresholds array */ struct mem_cgroup_threshold_ary *primary; /* * Spare threshold array. * This is needed to make mem_cgroup_unregister_event() "never fail". * It must be able to store at least primary->size - 1 entries. */ struct mem_cgroup_threshold_ary *spare; }; |
9490ff275
|
178 179 180 181 182 |
/* for OOM */ struct mem_cgroup_eventfd_list { struct list_head list; struct eventfd_ctx *eventfd; }; |
2e72b6347
|
183 |
|
2e72b6347
|
184 |
static void mem_cgroup_threshold(struct mem_cgroup *mem); |
9490ff275
|
185 |
static void mem_cgroup_oom_notify(struct mem_cgroup *mem); |
2e72b6347
|
186 |
|
f64c3f549
|
187 |
/* |
8cdea7c05
|
188 189 190 191 192 193 |
* The memory controller data structure. The memory controller controls both * page cache and RSS per cgroup. We would eventually like to provide * statistics based on the statistics developed by Rik Van Riel for clock-pro, * to help the administrator determine what knobs to tune. * * TODO: Add a water mark for the memory controller. Reclaim will begin when |
8a9f3ccd2
|
194 195 196 |
* we hit the water mark. May be even add a low water mark, such that * no reclaim occurs from a cgroup at it's low water mark, this is * a feature that will be implemented much later in the future. |
8cdea7c05
|
197 198 199 200 201 202 203 |
*/ struct mem_cgroup { struct cgroup_subsys_state css; /* * the counter to account for memory usage */ struct res_counter res; |
78fb74669
|
204 |
/* |
8c7c6e34a
|
205 206 207 208 |
* the counter to account for mem+swap usage. */ struct res_counter memsw; /* |
78fb74669
|
209 210 |
* Per cgroup active and inactive list, similar to the * per zone LRU lists. |
78fb74669
|
211 |
*/ |
6d12e2d8d
|
212 |
struct mem_cgroup_lru_info info; |
072c56c13
|
213 |
|
2733c06ac
|
214 215 216 217 |
/* protect against reclaim related member. */ spinlock_t reclaim_param_lock; |
6d61ef409
|
218 |
/* |
af901ca18
|
219 |
* While reclaiming in a hierarchy, we cache the last child we |
04046e1a0
|
220 |
* reclaimed from. |
6d61ef409
|
221 |
*/ |
04046e1a0
|
222 |
int last_scanned_child; |
18f59ea7d
|
223 224 225 226 |
/* * Should the accounting and control be hierarchical, per subtree? */ bool use_hierarchy; |
867578cbc
|
227 |
atomic_t oom_lock; |
8c7c6e34a
|
228 |
atomic_t refcnt; |
14797e236
|
229 |
|
a7885eb8a
|
230 |
unsigned int swappiness; |
3c11ecf44
|
231 232 |
/* OOM-Killer disable */ int oom_kill_disable; |
a7885eb8a
|
233 |
|
22a668d7c
|
234 235 |
/* set when res.limit == memsw.limit */ bool memsw_is_minimum; |
2e72b6347
|
236 237 238 239 |
/* protect arrays of thresholds */ struct mutex thresholds_lock; /* thresholds for memory usage. RCU-protected */ |
2c488db27
|
240 |
struct mem_cgroup_thresholds thresholds; |
907860ed3
|
241 |
|
2e72b6347
|
242 |
/* thresholds for mem+swap usage. RCU-protected */ |
2c488db27
|
243 |
struct mem_cgroup_thresholds memsw_thresholds; |
907860ed3
|
244 |
|
9490ff275
|
245 246 |
/* For oom notifier event fd */ struct list_head oom_notify; |
d52aa412d
|
247 |
/* |
7dc74be03
|
248 249 250 251 |
* Should we move charges of a task when a task is moved into this * mem_cgroup ? And what type of charges should we move ? */ unsigned long move_charge_at_immigrate; |
7dc74be03
|
252 |
/* |
c62b1a3b3
|
253 |
* percpu counter. |
d52aa412d
|
254 |
*/ |
c62b1a3b3
|
255 |
struct mem_cgroup_stat_cpu *stat; |
711d3d2c9
|
256 257 258 259 260 261 |
/* * used when a cpu is offlined or other synchronizations * See mem_cgroup_read_stat(). */ struct mem_cgroup_stat_cpu nocpu_base; spinlock_t pcp_counter_lock; |
8cdea7c05
|
262 |
}; |
7dc74be03
|
263 264 265 266 267 268 |
/* Stuffs for move charges at task migration. */ /* * Types of charges to be moved. "move_charge_at_immitgrate" is treated as a * left-shifted bitmap of these types. */ enum move_type { |
4ffef5fef
|
269 |
MOVE_CHARGE_TYPE_ANON, /* private anonymous page and swap of it */ |
87946a722
|
270 |
MOVE_CHARGE_TYPE_FILE, /* file page(including tmpfs) and swap of it */ |
7dc74be03
|
271 272 |
NR_MOVE_TYPE, }; |
4ffef5fef
|
273 274 |
/* "mc" and its members are protected by cgroup_mutex */ static struct move_charge_struct { |
b1dd693e5
|
275 |
spinlock_t lock; /* for from, to */ |
4ffef5fef
|
276 277 278 |
struct mem_cgroup *from; struct mem_cgroup *to; unsigned long precharge; |
854ffa8d1
|
279 |
unsigned long moved_charge; |
483c30b51
|
280 |
unsigned long moved_swap; |
8033b97c9
|
281 282 283 |
struct task_struct *moving_task; /* a task moving charges */ wait_queue_head_t waitq; /* a waitq for other context */ } mc = { |
2bd9bb206
|
284 |
.lock = __SPIN_LOCK_UNLOCKED(mc.lock), |
8033b97c9
|
285 286 |
.waitq = __WAIT_QUEUE_HEAD_INITIALIZER(mc.waitq), }; |
4ffef5fef
|
287 |
|
90254a658
|
288 289 290 291 292 |
static bool move_anon(void) { return test_bit(MOVE_CHARGE_TYPE_ANON, &mc.to->move_charge_at_immigrate); } |
87946a722
|
293 294 295 296 297 |
static bool move_file(void) { return test_bit(MOVE_CHARGE_TYPE_FILE, &mc.to->move_charge_at_immigrate); } |
4e4169535
|
298 299 300 301 302 303 |
/* * Maximum loops in mem_cgroup_hierarchical_reclaim(), used for soft * limit reclaim to prevent infinite loops, if they ever occur. */ #define MEM_CGROUP_MAX_RECLAIM_LOOPS (100) #define MEM_CGROUP_MAX_SOFT_LIMIT_RECLAIM_LOOPS (2) |
217bc3194
|
304 305 306 |
enum charge_type { MEM_CGROUP_CHARGE_TYPE_CACHE = 0, MEM_CGROUP_CHARGE_TYPE_MAPPED, |
4f98a2fee
|
307 |
MEM_CGROUP_CHARGE_TYPE_SHMEM, /* used by page migration of shmem */ |
c05555b57
|
308 |
MEM_CGROUP_CHARGE_TYPE_FORCE, /* used by force_empty */ |
d13d14430
|
309 |
MEM_CGROUP_CHARGE_TYPE_SWAPOUT, /* for accounting swapcache */ |
8a9478ca7
|
310 |
MEM_CGROUP_CHARGE_TYPE_DROP, /* a page was unused swap cache */ |
c05555b57
|
311 312 |
NR_CHARGE_TYPE, }; |
52d4b9ac0
|
313 314 315 |
/* only for here (for easy reading.) */ #define PCGF_CACHE (1UL << PCG_CACHE) #define PCGF_USED (1UL << PCG_USED) |
52d4b9ac0
|
316 |
#define PCGF_LOCK (1UL << PCG_LOCK) |
4b3bde4c9
|
317 318 |
/* Not used, but added here for completeness */ #define PCGF_ACCT (1UL << PCG_ACCT) |
217bc3194
|
319 |
|
8c7c6e34a
|
320 321 322 |
/* for encoding cft->private value on file */ #define _MEM (0) #define _MEMSWAP (1) |
9490ff275
|
323 |
#define _OOM_TYPE (2) |
8c7c6e34a
|
324 325 326 |
#define MEMFILE_PRIVATE(x, val) (((x) << 16) | (val)) #define MEMFILE_TYPE(val) (((val) >> 16) & 0xffff) #define MEMFILE_ATTR(val) ((val) & 0xffff) |
9490ff275
|
327 328 |
/* Used for OOM nofiier */ #define OOM_CONTROL (0) |
8c7c6e34a
|
329 |
|
75822b449
|
330 331 332 333 334 335 336 |
/* * Reclaim flags for mem_cgroup_hierarchical_reclaim */ #define MEM_CGROUP_RECLAIM_NOSWAP_BIT 0x0 #define MEM_CGROUP_RECLAIM_NOSWAP (1 << MEM_CGROUP_RECLAIM_NOSWAP_BIT) #define MEM_CGROUP_RECLAIM_SHRINK_BIT 0x1 #define MEM_CGROUP_RECLAIM_SHRINK (1 << MEM_CGROUP_RECLAIM_SHRINK_BIT) |
4e4169535
|
337 338 |
#define MEM_CGROUP_RECLAIM_SOFT_BIT 0x2 #define MEM_CGROUP_RECLAIM_SOFT (1 << MEM_CGROUP_RECLAIM_SOFT_BIT) |
75822b449
|
339 |
|
8c7c6e34a
|
340 341 |
static void mem_cgroup_get(struct mem_cgroup *mem); static void mem_cgroup_put(struct mem_cgroup *mem); |
7bcc1bb12
|
342 |
static struct mem_cgroup *parent_mem_cgroup(struct mem_cgroup *mem); |
cdec2e426
|
343 |
static void drain_all_stock_async(void); |
8c7c6e34a
|
344 |
|
f64c3f549
|
345 346 347 348 349 |
static struct mem_cgroup_per_zone * mem_cgroup_zoneinfo(struct mem_cgroup *mem, int nid, int zid) { return &mem->info.nodeinfo[nid]->zoneinfo[zid]; } |
d324236b3
|
350 351 352 353 |
struct cgroup_subsys_state *mem_cgroup_css(struct mem_cgroup *mem) { return &mem->css; } |
f64c3f549
|
354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 |
static struct mem_cgroup_per_zone * page_cgroup_zoneinfo(struct page_cgroup *pc) { struct mem_cgroup *mem = pc->mem_cgroup; int nid = page_cgroup_nid(pc); int zid = page_cgroup_zid(pc); if (!mem) return NULL; return mem_cgroup_zoneinfo(mem, nid, zid); } static struct mem_cgroup_tree_per_zone * soft_limit_tree_node_zone(int nid, int zid) { return &soft_limit_tree.rb_tree_per_node[nid]->rb_tree_per_zone[zid]; } static struct mem_cgroup_tree_per_zone * soft_limit_tree_from_page(struct page *page) { int nid = page_to_nid(page); int zid = page_zonenum(page); return &soft_limit_tree.rb_tree_per_node[nid]->rb_tree_per_zone[zid]; } static void |
4e4169535
|
383 |
__mem_cgroup_insert_exceeded(struct mem_cgroup *mem, |
f64c3f549
|
384 |
struct mem_cgroup_per_zone *mz, |
ef8745c1e
|
385 386 |
struct mem_cgroup_tree_per_zone *mctz, unsigned long long new_usage_in_excess) |
f64c3f549
|
387 388 389 390 391 392 393 |
{ struct rb_node **p = &mctz->rb_root.rb_node; struct rb_node *parent = NULL; struct mem_cgroup_per_zone *mz_node; if (mz->on_tree) return; |
ef8745c1e
|
394 395 396 |
mz->usage_in_excess = new_usage_in_excess; if (!mz->usage_in_excess) return; |
f64c3f549
|
397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 |
while (*p) { parent = *p; mz_node = rb_entry(parent, struct mem_cgroup_per_zone, tree_node); if (mz->usage_in_excess < mz_node->usage_in_excess) p = &(*p)->rb_left; /* * We can't avoid mem cgroups that are over their soft * limit by the same amount */ else if (mz->usage_in_excess >= mz_node->usage_in_excess) p = &(*p)->rb_right; } rb_link_node(&mz->tree_node, parent, p); rb_insert_color(&mz->tree_node, &mctz->rb_root); mz->on_tree = true; |
4e4169535
|
413 414 415 416 417 418 419 420 421 422 423 424 425 426 |
} static void __mem_cgroup_remove_exceeded(struct mem_cgroup *mem, struct mem_cgroup_per_zone *mz, struct mem_cgroup_tree_per_zone *mctz) { if (!mz->on_tree) return; rb_erase(&mz->tree_node, &mctz->rb_root); mz->on_tree = false; } static void |
f64c3f549
|
427 428 429 430 431 |
mem_cgroup_remove_exceeded(struct mem_cgroup *mem, struct mem_cgroup_per_zone *mz, struct mem_cgroup_tree_per_zone *mctz) { spin_lock(&mctz->lock); |
4e4169535
|
432 |
__mem_cgroup_remove_exceeded(mem, mz, mctz); |
f64c3f549
|
433 434 |
spin_unlock(&mctz->lock); } |
f64c3f549
|
435 436 437 |
static void mem_cgroup_update_tree(struct mem_cgroup *mem, struct page *page) { |
ef8745c1e
|
438 |
unsigned long long excess; |
f64c3f549
|
439 440 |
struct mem_cgroup_per_zone *mz; struct mem_cgroup_tree_per_zone *mctz; |
4e649152c
|
441 442 |
int nid = page_to_nid(page); int zid = page_zonenum(page); |
f64c3f549
|
443 444 445 |
mctz = soft_limit_tree_from_page(page); /* |
4e649152c
|
446 447 |
* Necessary to update all ancestors when hierarchy is used. * because their event counter is not touched. |
f64c3f549
|
448 |
*/ |
4e649152c
|
449 450 |
for (; mem; mem = parent_mem_cgroup(mem)) { mz = mem_cgroup_zoneinfo(mem, nid, zid); |
ef8745c1e
|
451 |
excess = res_counter_soft_limit_excess(&mem->res); |
4e649152c
|
452 453 454 455 |
/* * We have to update the tree if mz is on RB-tree or * mem is over its softlimit. */ |
ef8745c1e
|
456 |
if (excess || mz->on_tree) { |
4e649152c
|
457 458 459 460 461 |
spin_lock(&mctz->lock); /* if on-tree, remove it */ if (mz->on_tree) __mem_cgroup_remove_exceeded(mem, mz, mctz); /* |
ef8745c1e
|
462 463 |
* Insert again. mz->usage_in_excess will be updated. * If excess is 0, no tree ops. |
4e649152c
|
464 |
*/ |
ef8745c1e
|
465 |
__mem_cgroup_insert_exceeded(mem, mz, mctz, excess); |
4e649152c
|
466 467 |
spin_unlock(&mctz->lock); } |
f64c3f549
|
468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 |
} } static void mem_cgroup_remove_from_trees(struct mem_cgroup *mem) { int node, zone; struct mem_cgroup_per_zone *mz; struct mem_cgroup_tree_per_zone *mctz; for_each_node_state(node, N_POSSIBLE) { for (zone = 0; zone < MAX_NR_ZONES; zone++) { mz = mem_cgroup_zoneinfo(mem, node, zone); mctz = soft_limit_tree_node_zone(node, zone); mem_cgroup_remove_exceeded(mem, mz, mctz); } } } |
4e4169535
|
485 486 487 488 489 490 491 492 493 |
static inline unsigned long mem_cgroup_get_excess(struct mem_cgroup *mem) { return res_counter_soft_limit_excess(&mem->res) >> PAGE_SHIFT; } static struct mem_cgroup_per_zone * __mem_cgroup_largest_soft_limit_node(struct mem_cgroup_tree_per_zone *mctz) { struct rb_node *rightmost = NULL; |
26251eaf9
|
494 |
struct mem_cgroup_per_zone *mz; |
4e4169535
|
495 496 |
retry: |
26251eaf9
|
497 |
mz = NULL; |
4e4169535
|
498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 |
rightmost = rb_last(&mctz->rb_root); if (!rightmost) goto done; /* Nothing to reclaim from */ mz = rb_entry(rightmost, struct mem_cgroup_per_zone, tree_node); /* * Remove the node now but someone else can add it back, * we will to add it back at the end of reclaim to its correct * position in the tree. */ __mem_cgroup_remove_exceeded(mz->mem, mz, mctz); if (!res_counter_soft_limit_excess(&mz->mem->res) || !css_tryget(&mz->mem->css)) goto retry; done: return mz; } static struct mem_cgroup_per_zone * mem_cgroup_largest_soft_limit_node(struct mem_cgroup_tree_per_zone *mctz) { struct mem_cgroup_per_zone *mz; spin_lock(&mctz->lock); mz = __mem_cgroup_largest_soft_limit_node(mctz); spin_unlock(&mctz->lock); return mz; } |
711d3d2c9
|
526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 |
/* * Implementation Note: reading percpu statistics for memcg. * * Both of vmstat[] and percpu_counter has threshold and do periodic * synchronization to implement "quick" read. There are trade-off between * reading cost and precision of value. Then, we may have a chance to implement * a periodic synchronizion of counter in memcg's counter. * * But this _read() function is used for user interface now. The user accounts * memory usage by memory cgroup and he _always_ requires exact value because * he accounts memory. Even if we provide quick-and-fuzzy read, we always * have to visit all online cpus and make sum. So, for now, unnecessary * synchronization is not implemented. (just implemented for cpu hotplug) * * If there are kernel internal actions which can make use of some not-exact * value, and reading all cpu value can be performance bottleneck in some * common workload, threashold and synchonization as vmstat[] should be * implemented. */ |
c62b1a3b3
|
545 546 547 548 549 |
static s64 mem_cgroup_read_stat(struct mem_cgroup *mem, enum mem_cgroup_stat_index idx) { int cpu; s64 val = 0; |
711d3d2c9
|
550 551 |
get_online_cpus(); for_each_online_cpu(cpu) |
c62b1a3b3
|
552 |
val += per_cpu(mem->stat->count[idx], cpu); |
711d3d2c9
|
553 554 555 556 557 558 |
#ifdef CONFIG_HOTPLUG_CPU spin_lock(&mem->pcp_counter_lock); val += mem->nocpu_base.count[idx]; spin_unlock(&mem->pcp_counter_lock); #endif put_online_cpus(); |
c62b1a3b3
|
559 560 561 562 563 564 565 566 567 568 569 |
return val; } static s64 mem_cgroup_local_usage(struct mem_cgroup *mem) { s64 ret; ret = mem_cgroup_read_stat(mem, MEM_CGROUP_STAT_RSS); ret += mem_cgroup_read_stat(mem, MEM_CGROUP_STAT_CACHE); return ret; } |
0c3e73e84
|
570 571 572 573 |
static void mem_cgroup_swap_statistics(struct mem_cgroup *mem, bool charge) { int val = (charge) ? 1 : -1; |
c62b1a3b3
|
574 |
this_cpu_add(mem->stat->count[MEM_CGROUP_STAT_SWAPOUT], val); |
0c3e73e84
|
575 |
} |
c05555b57
|
576 577 578 |
static void mem_cgroup_charge_statistics(struct mem_cgroup *mem, struct page_cgroup *pc, bool charge) |
d52aa412d
|
579 |
{ |
0c3e73e84
|
580 |
int val = (charge) ? 1 : -1; |
d52aa412d
|
581 |
|
c62b1a3b3
|
582 |
preempt_disable(); |
c05555b57
|
583 |
if (PageCgroupCache(pc)) |
c62b1a3b3
|
584 |
__this_cpu_add(mem->stat->count[MEM_CGROUP_STAT_CACHE], val); |
d52aa412d
|
585 |
else |
c62b1a3b3
|
586 |
__this_cpu_add(mem->stat->count[MEM_CGROUP_STAT_RSS], val); |
55e462b05
|
587 588 |
if (charge) |
c62b1a3b3
|
589 |
__this_cpu_inc(mem->stat->count[MEM_CGROUP_STAT_PGPGIN_COUNT]); |
55e462b05
|
590 |
else |
c62b1a3b3
|
591 |
__this_cpu_inc(mem->stat->count[MEM_CGROUP_STAT_PGPGOUT_COUNT]); |
d2265e6fa
|
592 |
__this_cpu_inc(mem->stat->count[MEM_CGROUP_EVENTS]); |
2e72b6347
|
593 |
|
c62b1a3b3
|
594 |
preempt_enable(); |
6d12e2d8d
|
595 |
} |
14067bb3e
|
596 |
static unsigned long mem_cgroup_get_local_zonestat(struct mem_cgroup *mem, |
b69408e88
|
597 |
enum lru_list idx) |
6d12e2d8d
|
598 599 600 601 602 603 604 605 606 607 608 |
{ int nid, zid; struct mem_cgroup_per_zone *mz; u64 total = 0; for_each_online_node(nid) for (zid = 0; zid < MAX_NR_ZONES; zid++) { mz = mem_cgroup_zoneinfo(mem, nid, zid); total += MEM_CGROUP_ZSTAT(mz, idx); } return total; |
d52aa412d
|
609 |
} |
d2265e6fa
|
610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 |
static bool __memcg_event_check(struct mem_cgroup *mem, int event_mask_shift) { s64 val; val = this_cpu_read(mem->stat->count[MEM_CGROUP_EVENTS]); return !(val & ((1 << event_mask_shift) - 1)); } /* * Check events in order. * */ static void memcg_check_events(struct mem_cgroup *mem, struct page *page) { /* threshold event is triggered in finer grain than soft limit */ if (unlikely(__memcg_event_check(mem, THRESHOLDS_EVENTS_THRESH))) { mem_cgroup_threshold(mem); if (unlikely(__memcg_event_check(mem, SOFTLIMIT_EVENTS_THRESH))) mem_cgroup_update_tree(mem, page); } } |
d5b69e38f
|
632 |
static struct mem_cgroup *mem_cgroup_from_cont(struct cgroup *cont) |
8cdea7c05
|
633 634 635 636 637 |
{ return container_of(cgroup_subsys_state(cont, mem_cgroup_subsys_id), struct mem_cgroup, css); } |
cf475ad28
|
638 |
struct mem_cgroup *mem_cgroup_from_task(struct task_struct *p) |
78fb74669
|
639 |
{ |
31a78f23b
|
640 641 642 643 644 645 646 |
/* * mm_update_next_owner() may clear mm->owner to NULL * if it races with swapoff, page migration, etc. * So this can be called with p == NULL. */ if (unlikely(!p)) return NULL; |
78fb74669
|
647 648 649 |
return container_of(task_subsys_state(p, mem_cgroup_subsys_id), struct mem_cgroup, css); } |
54595fe26
|
650 651 652 |
static struct mem_cgroup *try_get_mem_cgroup_from_mm(struct mm_struct *mm) { struct mem_cgroup *mem = NULL; |
0b7f569e4
|
653 654 655 |
if (!mm) return NULL; |
54595fe26
|
656 657 658 659 660 661 662 663 664 665 666 667 668 669 |
/* * Because we have no locks, mm->owner's may be being moved to other * cgroup. We use css_tryget() here even if this looks * pessimistic (rather than adding locks here). */ rcu_read_lock(); do { mem = mem_cgroup_from_task(rcu_dereference(mm->owner)); if (unlikely(!mem)) break; } while (!css_tryget(&mem->css)); rcu_read_unlock(); return mem; } |
7d74b06f2
|
670 671 |
/* The caller has to guarantee "mem" exists before calling this */ static struct mem_cgroup *mem_cgroup_start_loop(struct mem_cgroup *mem) |
14067bb3e
|
672 |
{ |
711d3d2c9
|
673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 |
struct cgroup_subsys_state *css; int found; if (!mem) /* ROOT cgroup has the smallest ID */ return root_mem_cgroup; /*css_put/get against root is ignored*/ if (!mem->use_hierarchy) { if (css_tryget(&mem->css)) return mem; return NULL; } rcu_read_lock(); /* * searching a memory cgroup which has the smallest ID under given * ROOT cgroup. (ID >= 1) */ css = css_get_next(&mem_cgroup_subsys, 1, &mem->css, &found); if (css && css_tryget(css)) mem = container_of(css, struct mem_cgroup, css); else mem = NULL; rcu_read_unlock(); return mem; |
7d74b06f2
|
695 696 697 698 699 700 701 702 703 |
} static struct mem_cgroup *mem_cgroup_get_next(struct mem_cgroup *iter, struct mem_cgroup *root, bool cond) { int nextid = css_id(&iter->css) + 1; int found; int hierarchy_used; |
14067bb3e
|
704 |
struct cgroup_subsys_state *css; |
14067bb3e
|
705 |
|
7d74b06f2
|
706 |
hierarchy_used = iter->use_hierarchy; |
14067bb3e
|
707 |
|
7d74b06f2
|
708 |
css_put(&iter->css); |
711d3d2c9
|
709 710 |
/* If no ROOT, walk all, ignore hierarchy */ if (!cond || (root && !hierarchy_used)) |
7d74b06f2
|
711 |
return NULL; |
14067bb3e
|
712 |
|
711d3d2c9
|
713 714 |
if (!root) root = root_mem_cgroup; |
7d74b06f2
|
715 716 |
do { iter = NULL; |
14067bb3e
|
717 |
rcu_read_lock(); |
7d74b06f2
|
718 719 720 |
css = css_get_next(&mem_cgroup_subsys, nextid, &root->css, &found); |
14067bb3e
|
721 |
if (css && css_tryget(css)) |
7d74b06f2
|
722 |
iter = container_of(css, struct mem_cgroup, css); |
14067bb3e
|
723 |
rcu_read_unlock(); |
7d74b06f2
|
724 |
/* If css is NULL, no more cgroups will be found */ |
14067bb3e
|
725 |
nextid = found + 1; |
7d74b06f2
|
726 |
} while (css && !iter); |
14067bb3e
|
727 |
|
7d74b06f2
|
728 |
return iter; |
14067bb3e
|
729 |
} |
7d74b06f2
|
730 731 732 733 734 735 736 737 738 739 740 741 |
/* * for_eacn_mem_cgroup_tree() for visiting all cgroup under tree. Please * be careful that "break" loop is not allowed. We have reference count. * Instead of that modify "cond" to be false and "continue" to exit the loop. */ #define for_each_mem_cgroup_tree_cond(iter, root, cond) \ for (iter = mem_cgroup_start_loop(root);\ iter != NULL;\ iter = mem_cgroup_get_next(iter, root, cond)) #define for_each_mem_cgroup_tree(iter, root) \ for_each_mem_cgroup_tree_cond(iter, root, true) |
711d3d2c9
|
742 743 |
#define for_each_mem_cgroup_all(iter) \ for_each_mem_cgroup_tree_cond(iter, NULL, true) |
14067bb3e
|
744 |
|
4b3bde4c9
|
745 746 747 748 |
static inline bool mem_cgroup_is_root(struct mem_cgroup *mem) { return (mem == root_mem_cgroup); } |
08e552c69
|
749 750 751 752 753 754 755 756 757 758 759 760 761 |
/* * Following LRU functions are allowed to be used without PCG_LOCK. * Operations are called by routine of global LRU independently from memcg. * What we have to take care of here is validness of pc->mem_cgroup. * * Changes to pc->mem_cgroup happens when * 1. charge * 2. moving account * In typical case, "charge" is done before add-to-lru. Exception is SwapCache. * It is added to LRU before charge. * If PCG_USED bit is not set, page_cgroup is not added to this private LRU. * When moving account, the page is not on LRU. It's isolated. */ |
4f98a2fee
|
762 |
|
08e552c69
|
763 764 765 |
void mem_cgroup_del_lru_list(struct page *page, enum lru_list lru) { struct page_cgroup *pc; |
08e552c69
|
766 |
struct mem_cgroup_per_zone *mz; |
6d12e2d8d
|
767 |
|
f8d665422
|
768 |
if (mem_cgroup_disabled()) |
08e552c69
|
769 770 771 |
return; pc = lookup_page_cgroup(page); /* can happen while we handle swapcache. */ |
4b3bde4c9
|
772 |
if (!TestClearPageCgroupAcctLRU(pc)) |
08e552c69
|
773 |
return; |
4b3bde4c9
|
774 |
VM_BUG_ON(!pc->mem_cgroup); |
544122e5e
|
775 776 777 778 |
/* * We don't check PCG_USED bit. It's cleared when the "page" is finally * removed from global LRU. */ |
08e552c69
|
779 |
mz = page_cgroup_zoneinfo(pc); |
b69408e88
|
780 |
MEM_CGROUP_ZSTAT(mz, lru) -= 1; |
4b3bde4c9
|
781 782 783 |
if (mem_cgroup_is_root(pc->mem_cgroup)) return; VM_BUG_ON(list_empty(&pc->lru)); |
08e552c69
|
784 |
list_del_init(&pc->lru); |
6d12e2d8d
|
785 |
} |
08e552c69
|
786 |
void mem_cgroup_del_lru(struct page *page) |
6d12e2d8d
|
787 |
{ |
08e552c69
|
788 789 |
mem_cgroup_del_lru_list(page, page_lru(page)); } |
b69408e88
|
790 |
|
08e552c69
|
791 792 793 794 |
void mem_cgroup_rotate_lru_list(struct page *page, enum lru_list lru) { struct mem_cgroup_per_zone *mz; struct page_cgroup *pc; |
b69408e88
|
795 |
|
f8d665422
|
796 |
if (mem_cgroup_disabled()) |
08e552c69
|
797 |
return; |
6d12e2d8d
|
798 |
|
08e552c69
|
799 |
pc = lookup_page_cgroup(page); |
bd112db87
|
800 801 802 803 |
/* * Used bit is set without atomic ops but after smp_wmb(). * For making pc->mem_cgroup visible, insert smp_rmb() here. */ |
08e552c69
|
804 |
smp_rmb(); |
4b3bde4c9
|
805 806 |
/* unused or root page is not rotated. */ if (!PageCgroupUsed(pc) || mem_cgroup_is_root(pc->mem_cgroup)) |
08e552c69
|
807 808 809 |
return; mz = page_cgroup_zoneinfo(pc); list_move(&pc->lru, &mz->lists[lru]); |
6d12e2d8d
|
810 |
} |
08e552c69
|
811 |
void mem_cgroup_add_lru_list(struct page *page, enum lru_list lru) |
66e1707bc
|
812 |
{ |
08e552c69
|
813 814 |
struct page_cgroup *pc; struct mem_cgroup_per_zone *mz; |
6d12e2d8d
|
815 |
|
f8d665422
|
816 |
if (mem_cgroup_disabled()) |
08e552c69
|
817 818 |
return; pc = lookup_page_cgroup(page); |
4b3bde4c9
|
819 |
VM_BUG_ON(PageCgroupAcctLRU(pc)); |
bd112db87
|
820 821 822 823 |
/* * Used bit is set without atomic ops but after smp_wmb(). * For making pc->mem_cgroup visible, insert smp_rmb() here. */ |
08e552c69
|
824 825 |
smp_rmb(); if (!PageCgroupUsed(pc)) |
894bc3104
|
826 |
return; |
b69408e88
|
827 |
|
08e552c69
|
828 |
mz = page_cgroup_zoneinfo(pc); |
b69408e88
|
829 |
MEM_CGROUP_ZSTAT(mz, lru) += 1; |
4b3bde4c9
|
830 831 832 |
SetPageCgroupAcctLRU(pc); if (mem_cgroup_is_root(pc->mem_cgroup)) return; |
08e552c69
|
833 834 |
list_add(&pc->lru, &mz->lists[lru]); } |
544122e5e
|
835 |
|
08e552c69
|
836 |
/* |
544122e5e
|
837 838 839 840 841 |
* At handling SwapCache, pc->mem_cgroup may be changed while it's linked to * lru because the page may.be reused after it's fully uncharged (because of * SwapCache behavior).To handle that, unlink page_cgroup from LRU when charge * it again. This function is only used to charge SwapCache. It's done under * lock_page and expected that zone->lru_lock is never held. |
08e552c69
|
842 |
*/ |
544122e5e
|
843 |
static void mem_cgroup_lru_del_before_commit_swapcache(struct page *page) |
08e552c69
|
844 |
{ |
544122e5e
|
845 846 847 848 849 850 851 852 853 854 855 856 |
unsigned long flags; struct zone *zone = page_zone(page); struct page_cgroup *pc = lookup_page_cgroup(page); spin_lock_irqsave(&zone->lru_lock, flags); /* * Forget old LRU when this page_cgroup is *not* used. This Used bit * is guarded by lock_page() because the page is SwapCache. */ if (!PageCgroupUsed(pc)) mem_cgroup_del_lru_list(page, page_lru(page)); spin_unlock_irqrestore(&zone->lru_lock, flags); |
08e552c69
|
857 |
} |
544122e5e
|
858 859 860 861 862 863 864 865 |
static void mem_cgroup_lru_add_after_commit_swapcache(struct page *page) { unsigned long flags; struct zone *zone = page_zone(page); struct page_cgroup *pc = lookup_page_cgroup(page); spin_lock_irqsave(&zone->lru_lock, flags); /* link when the page is linked to LRU but page_cgroup isn't */ |
4b3bde4c9
|
866 |
if (PageLRU(page) && !PageCgroupAcctLRU(pc)) |
544122e5e
|
867 868 869 |
mem_cgroup_add_lru_list(page, page_lru(page)); spin_unlock_irqrestore(&zone->lru_lock, flags); } |
08e552c69
|
870 871 872 |
void mem_cgroup_move_lists(struct page *page, enum lru_list from, enum lru_list to) { |
f8d665422
|
873 |
if (mem_cgroup_disabled()) |
08e552c69
|
874 875 876 |
return; mem_cgroup_del_lru_list(page, from); mem_cgroup_add_lru_list(page, to); |
66e1707bc
|
877 |
} |
4c4a22148
|
878 879 880 |
int task_in_mem_cgroup(struct task_struct *task, const struct mem_cgroup *mem) { int ret; |
0b7f569e4
|
881 |
struct mem_cgroup *curr = NULL; |
158e0a2d1
|
882 |
struct task_struct *p; |
4c4a22148
|
883 |
|
158e0a2d1
|
884 885 886 887 888 |
p = find_lock_task_mm(task); if (!p) return 0; curr = try_get_mem_cgroup_from_mm(p->mm); task_unlock(p); |
0b7f569e4
|
889 890 |
if (!curr) return 0; |
d31f56dbf
|
891 892 893 894 895 896 897 |
/* * We should check use_hierarchy of "mem" not "curr". Because checking * use_hierarchy of "curr" here make this function true if hierarchy is * enabled in "curr" and "curr" is a child of "mem" in *cgroup* * hierarchy(even if use_hierarchy is disabled in "mem"). */ if (mem->use_hierarchy) |
0b7f569e4
|
898 899 900 901 |
ret = css_is_ancestor(&curr->css, &mem->css); else ret = (curr == mem); css_put(&curr->css); |
4c4a22148
|
902 903 |
return ret; } |
c772be939
|
904 |
static int calc_inactive_ratio(struct mem_cgroup *memcg, unsigned long *present_pages) |
14797e236
|
905 906 907 |
{ unsigned long active; unsigned long inactive; |
c772be939
|
908 909 |
unsigned long gb; unsigned long inactive_ratio; |
14797e236
|
910 |
|
14067bb3e
|
911 912 |
inactive = mem_cgroup_get_local_zonestat(memcg, LRU_INACTIVE_ANON); active = mem_cgroup_get_local_zonestat(memcg, LRU_ACTIVE_ANON); |
14797e236
|
913 |
|
c772be939
|
914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 |
gb = (inactive + active) >> (30 - PAGE_SHIFT); if (gb) inactive_ratio = int_sqrt(10 * gb); else inactive_ratio = 1; if (present_pages) { present_pages[0] = inactive; present_pages[1] = active; } return inactive_ratio; } int mem_cgroup_inactive_anon_is_low(struct mem_cgroup *memcg) { unsigned long active; unsigned long inactive; unsigned long present_pages[2]; unsigned long inactive_ratio; inactive_ratio = calc_inactive_ratio(memcg, present_pages); inactive = present_pages[0]; active = present_pages[1]; if (inactive * inactive_ratio < active) |
14797e236
|
941 942 943 944 |
return 1; return 0; } |
56e49d218
|
945 946 947 948 949 950 951 952 953 954 |
int mem_cgroup_inactive_file_is_low(struct mem_cgroup *memcg) { unsigned long active; unsigned long inactive; inactive = mem_cgroup_get_local_zonestat(memcg, LRU_INACTIVE_FILE); active = mem_cgroup_get_local_zonestat(memcg, LRU_ACTIVE_FILE); return (active > inactive); } |
a3d8e0549
|
955 956 957 958 |
unsigned long mem_cgroup_zone_nr_pages(struct mem_cgroup *memcg, struct zone *zone, enum lru_list lru) { |
13d7e3a2d
|
959 |
int nid = zone_to_nid(zone); |
a3d8e0549
|
960 961 962 963 964 |
int zid = zone_idx(zone); struct mem_cgroup_per_zone *mz = mem_cgroup_zoneinfo(memcg, nid, zid); return MEM_CGROUP_ZSTAT(mz, lru); } |
3e2f41f1f
|
965 966 967 |
struct zone_reclaim_stat *mem_cgroup_get_reclaim_stat(struct mem_cgroup *memcg, struct zone *zone) { |
13d7e3a2d
|
968 |
int nid = zone_to_nid(zone); |
3e2f41f1f
|
969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 |
int zid = zone_idx(zone); struct mem_cgroup_per_zone *mz = mem_cgroup_zoneinfo(memcg, nid, zid); return &mz->reclaim_stat; } struct zone_reclaim_stat * mem_cgroup_get_reclaim_stat_from_page(struct page *page) { struct page_cgroup *pc; struct mem_cgroup_per_zone *mz; if (mem_cgroup_disabled()) return NULL; pc = lookup_page_cgroup(page); |
bd112db87
|
985 986 987 988 989 990 991 |
/* * Used bit is set without atomic ops but after smp_wmb(). * For making pc->mem_cgroup visible, insert smp_rmb() here. */ smp_rmb(); if (!PageCgroupUsed(pc)) return NULL; |
3e2f41f1f
|
992 993 994 995 996 997 |
mz = page_cgroup_zoneinfo(pc); if (!mz) return NULL; return &mz->reclaim_stat; } |
66e1707bc
|
998 999 1000 1001 1002 |
unsigned long mem_cgroup_isolate_pages(unsigned long nr_to_scan, struct list_head *dst, unsigned long *scanned, int order, int mode, struct zone *z, struct mem_cgroup *mem_cont, |
4f98a2fee
|
1003 |
int active, int file) |
66e1707bc
|
1004 1005 1006 1007 1008 1009 |
{ unsigned long nr_taken = 0; struct page *page; unsigned long scan; LIST_HEAD(pc_list); struct list_head *src; |
ff7283fa3
|
1010 |
struct page_cgroup *pc, *tmp; |
13d7e3a2d
|
1011 |
int nid = zone_to_nid(z); |
1ecaab2bd
|
1012 1013 |
int zid = zone_idx(z); struct mem_cgroup_per_zone *mz; |
b7c46d151
|
1014 |
int lru = LRU_FILE * file + active; |
2ffebca6a
|
1015 |
int ret; |
66e1707bc
|
1016 |
|
cf475ad28
|
1017 |
BUG_ON(!mem_cont); |
1ecaab2bd
|
1018 |
mz = mem_cgroup_zoneinfo(mem_cont, nid, zid); |
b69408e88
|
1019 |
src = &mz->lists[lru]; |
66e1707bc
|
1020 |
|
ff7283fa3
|
1021 1022 |
scan = 0; list_for_each_entry_safe_reverse(pc, tmp, src, lru) { |
436c6541b
|
1023 |
if (scan >= nr_to_scan) |
ff7283fa3
|
1024 |
break; |
08e552c69
|
1025 1026 |
page = pc->page; |
52d4b9ac0
|
1027 1028 |
if (unlikely(!PageCgroupUsed(pc))) continue; |
436c6541b
|
1029 |
if (unlikely(!PageLRU(page))) |
ff7283fa3
|
1030 |
continue; |
ff7283fa3
|
1031 |
|
436c6541b
|
1032 |
scan++; |
2ffebca6a
|
1033 1034 1035 |
ret = __isolate_lru_page(page, mode, file); switch (ret) { case 0: |
66e1707bc
|
1036 |
list_move(&page->lru, dst); |
2ffebca6a
|
1037 |
mem_cgroup_del_lru(page); |
2c888cfbc
|
1038 |
nr_taken += hpage_nr_pages(page); |
2ffebca6a
|
1039 1040 1041 1042 1043 1044 1045 |
break; case -EBUSY: /* we don't affect global LRU but rotate in our LRU */ mem_cgroup_rotate_lru_list(page, page_lru(page)); break; default: break; |
66e1707bc
|
1046 1047 |
} } |
66e1707bc
|
1048 |
*scanned = scan; |
cc8e970c3
|
1049 1050 1051 |
trace_mm_vmscan_memcg_isolate(0, nr_to_scan, scan, nr_taken, 0, 0, 0, mode); |
66e1707bc
|
1052 1053 |
return nr_taken; } |
6d61ef409
|
1054 1055 |
#define mem_cgroup_from_res_counter(counter, member) \ container_of(counter, struct mem_cgroup, member) |
b85a96c0b
|
1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 |
static bool mem_cgroup_check_under_limit(struct mem_cgroup *mem) { if (do_swap_account) { if (res_counter_check_under_limit(&mem->res) && res_counter_check_under_limit(&mem->memsw)) return true; } else if (res_counter_check_under_limit(&mem->res)) return true; return false; } |
a7885eb8a
|
1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 |
static unsigned int get_swappiness(struct mem_cgroup *memcg) { struct cgroup *cgrp = memcg->css.cgroup; unsigned int swappiness; /* root ? */ if (cgrp->parent == NULL) return vm_swappiness; spin_lock(&memcg->reclaim_param_lock); swappiness = memcg->swappiness; spin_unlock(&memcg->reclaim_param_lock); return swappiness; } |
32047e2a8
|
1082 1083 1084 |
static void mem_cgroup_start_move(struct mem_cgroup *mem) { int cpu; |
1489ebad8
|
1085 1086 1087 1088 |
get_online_cpus(); spin_lock(&mem->pcp_counter_lock); for_each_online_cpu(cpu) |
32047e2a8
|
1089 |
per_cpu(mem->stat->count[MEM_CGROUP_ON_MOVE], cpu) += 1; |
1489ebad8
|
1090 1091 1092 |
mem->nocpu_base.count[MEM_CGROUP_ON_MOVE] += 1; spin_unlock(&mem->pcp_counter_lock); put_online_cpus(); |
32047e2a8
|
1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 |
synchronize_rcu(); } static void mem_cgroup_end_move(struct mem_cgroup *mem) { int cpu; if (!mem) return; |
1489ebad8
|
1103 1104 1105 |
get_online_cpus(); spin_lock(&mem->pcp_counter_lock); for_each_online_cpu(cpu) |
32047e2a8
|
1106 |
per_cpu(mem->stat->count[MEM_CGROUP_ON_MOVE], cpu) -= 1; |
1489ebad8
|
1107 1108 1109 |
mem->nocpu_base.count[MEM_CGROUP_ON_MOVE] -= 1; spin_unlock(&mem->pcp_counter_lock); put_online_cpus(); |
32047e2a8
|
1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 |
} /* * 2 routines for checking "mem" is under move_account() or not. * * mem_cgroup_stealed() - checking a cgroup is mc.from or not. This is used * for avoiding race in accounting. If true, * pc->mem_cgroup may be overwritten. * * mem_cgroup_under_move() - checking a cgroup is mc.from or mc.to or * under hierarchy of moving cgroups. This is for * waiting at hith-memory prressure caused by "move". */ static bool mem_cgroup_stealed(struct mem_cgroup *mem) { VM_BUG_ON(!rcu_read_lock_held()); return this_cpu_read(mem->stat->count[MEM_CGROUP_ON_MOVE]) > 0; } |
4b5343346
|
1128 1129 1130 |
static bool mem_cgroup_under_move(struct mem_cgroup *mem) { |
2bd9bb206
|
1131 1132 |
struct mem_cgroup *from; struct mem_cgroup *to; |
4b5343346
|
1133 |
bool ret = false; |
2bd9bb206
|
1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 |
/* * Unlike task_move routines, we access mc.to, mc.from not under * mutual exclusion by cgroup_mutex. Here, we take spinlock instead. */ spin_lock(&mc.lock); from = mc.from; to = mc.to; if (!from) goto unlock; if (from == mem || to == mem || (mem->use_hierarchy && css_is_ancestor(&from->css, &mem->css)) || (mem->use_hierarchy && css_is_ancestor(&to->css, &mem->css))) ret = true; unlock: spin_unlock(&mc.lock); |
4b5343346
|
1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 |
return ret; } static bool mem_cgroup_wait_acct_move(struct mem_cgroup *mem) { if (mc.moving_task && current != mc.moving_task) { if (mem_cgroup_under_move(mem)) { DEFINE_WAIT(wait); prepare_to_wait(&mc.waitq, &wait, TASK_INTERRUPTIBLE); /* moving charge context might have finished. */ if (mc.moving_task) schedule(); finish_wait(&mc.waitq, &wait); return true; } } return false; } |
e222432bf
|
1167 |
/** |
6a6135b64
|
1168 |
* mem_cgroup_print_oom_info: Called from OOM with tasklist_lock held in read mode. |
e222432bf
|
1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 |
* @memcg: The memory cgroup that went over limit * @p: Task that is going to be killed * * NOTE: @memcg and @p's mem_cgroup can be different when hierarchy is * enabled */ void mem_cgroup_print_oom_info(struct mem_cgroup *memcg, struct task_struct *p) { struct cgroup *task_cgrp; struct cgroup *mem_cgrp; /* * Need a buffer in BSS, can't rely on allocations. The code relies * on the assumption that OOM is serialized for memory controller. * If this assumption is broken, revisit this code. */ static char memcg_name[PATH_MAX]; int ret; |
d31f56dbf
|
1186 |
if (!memcg || !p) |
e222432bf
|
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 1230 1231 1232 1233 1234 |
return; rcu_read_lock(); mem_cgrp = memcg->css.cgroup; task_cgrp = task_cgroup(p, mem_cgroup_subsys_id); ret = cgroup_path(task_cgrp, memcg_name, PATH_MAX); if (ret < 0) { /* * Unfortunately, we are unable to convert to a useful name * But we'll still print out the usage information */ rcu_read_unlock(); goto done; } rcu_read_unlock(); printk(KERN_INFO "Task in %s killed", memcg_name); rcu_read_lock(); ret = cgroup_path(mem_cgrp, memcg_name, PATH_MAX); if (ret < 0) { rcu_read_unlock(); goto done; } rcu_read_unlock(); /* * Continues from above, so we don't need an KERN_ level */ printk(KERN_CONT " as a result of limit of %s ", memcg_name); done: printk(KERN_INFO "memory: usage %llukB, limit %llukB, failcnt %llu ", res_counter_read_u64(&memcg->res, RES_USAGE) >> 10, res_counter_read_u64(&memcg->res, RES_LIMIT) >> 10, res_counter_read_u64(&memcg->res, RES_FAILCNT)); printk(KERN_INFO "memory+swap: usage %llukB, limit %llukB, " "failcnt %llu ", res_counter_read_u64(&memcg->memsw, RES_USAGE) >> 10, res_counter_read_u64(&memcg->memsw, RES_LIMIT) >> 10, res_counter_read_u64(&memcg->memsw, RES_FAILCNT)); } |
81d39c20f
|
1235 1236 1237 1238 1239 1240 1241 |
/* * This function returns the number of memcg under hierarchy tree. Returns * 1(self count) if no children. */ static int mem_cgroup_count_children(struct mem_cgroup *mem) { int num = 0; |
7d74b06f2
|
1242 1243 1244 1245 |
struct mem_cgroup *iter; for_each_mem_cgroup_tree(iter, mem) num++; |
81d39c20f
|
1246 1247 |
return num; } |
6d61ef409
|
1248 |
/* |
a63d83f42
|
1249 1250 1251 1252 1253 1254 |
* Return the memory (and swap, if configured) limit for a memcg. */ u64 mem_cgroup_get_limit(struct mem_cgroup *memcg) { u64 limit; u64 memsw; |
f3e8eb70b
|
1255 1256 |
limit = res_counter_read_u64(&memcg->res, RES_LIMIT); limit += total_swap_pages << PAGE_SHIFT; |
a63d83f42
|
1257 1258 1259 1260 1261 1262 1263 1264 1265 |
memsw = res_counter_read_u64(&memcg->memsw, RES_LIMIT); /* * If memsw is finite and limits the amount of swap space available * to this memcg, return that limit. */ return min(limit, memsw); } /* |
04046e1a0
|
1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 |
* Visit the first child (need not be the first child as per the ordering * of the cgroup list, since we track last_scanned_child) of @mem and use * that to reclaim free pages from. */ static struct mem_cgroup * mem_cgroup_select_victim(struct mem_cgroup *root_mem) { struct mem_cgroup *ret = NULL; struct cgroup_subsys_state *css; int nextid, found; if (!root_mem->use_hierarchy) { css_get(&root_mem->css); ret = root_mem; } while (!ret) { rcu_read_lock(); nextid = root_mem->last_scanned_child + 1; css = css_get_next(&mem_cgroup_subsys, nextid, &root_mem->css, &found); if (css && css_tryget(css)) ret = container_of(css, struct mem_cgroup, css); rcu_read_unlock(); /* Updates scanning parameter */ spin_lock(&root_mem->reclaim_param_lock); if (!css) { /* this means start scan from ID:1 */ root_mem->last_scanned_child = 0; } else root_mem->last_scanned_child = found; spin_unlock(&root_mem->reclaim_param_lock); } return ret; } /* * Scan the hierarchy if needed to reclaim memory. We remember the last child * we reclaimed from, so that we don't end up penalizing one child extensively * based on its position in the children list. |
6d61ef409
|
1308 1309 |
* * root_mem is the original ancestor that we've been reclaim from. |
04046e1a0
|
1310 1311 1312 |
* * We give up and return to the caller when we visit root_mem twice. * (other groups can be removed while we're walking....) |
81d39c20f
|
1313 1314 |
* * If shrink==true, for avoiding to free too much, this returns immedieately. |
6d61ef409
|
1315 1316 |
*/ static int mem_cgroup_hierarchical_reclaim(struct mem_cgroup *root_mem, |
4e4169535
|
1317 |
struct zone *zone, |
75822b449
|
1318 1319 |
gfp_t gfp_mask, unsigned long reclaim_options) |
6d61ef409
|
1320 |
{ |
04046e1a0
|
1321 1322 1323 |
struct mem_cgroup *victim; int ret, total = 0; int loop = 0; |
75822b449
|
1324 1325 |
bool noswap = reclaim_options & MEM_CGROUP_RECLAIM_NOSWAP; bool shrink = reclaim_options & MEM_CGROUP_RECLAIM_SHRINK; |
4e4169535
|
1326 1327 |
bool check_soft = reclaim_options & MEM_CGROUP_RECLAIM_SOFT; unsigned long excess = mem_cgroup_get_excess(root_mem); |
04046e1a0
|
1328 |
|
22a668d7c
|
1329 1330 1331 |
/* If memsw_is_minimum==1, swap-out is of-no-use. */ if (root_mem->memsw_is_minimum) noswap = true; |
4e4169535
|
1332 |
while (1) { |
04046e1a0
|
1333 |
victim = mem_cgroup_select_victim(root_mem); |
4e4169535
|
1334 |
if (victim == root_mem) { |
04046e1a0
|
1335 |
loop++; |
cdec2e426
|
1336 1337 |
if (loop >= 1) drain_all_stock_async(); |
4e4169535
|
1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 |
if (loop >= 2) { /* * If we have not been able to reclaim * anything, it might because there are * no reclaimable pages under this hierarchy */ if (!check_soft || !total) { css_put(&victim->css); break; } /* * We want to do more targetted reclaim. * excess >> 2 is not to excessive so as to * reclaim too much, nor too less that we keep * coming back to reclaim from this cgroup */ if (total >= (excess >> 2) || (loop > MEM_CGROUP_MAX_RECLAIM_LOOPS)) { css_put(&victim->css); break; } } } |
c62b1a3b3
|
1361 |
if (!mem_cgroup_local_usage(victim)) { |
04046e1a0
|
1362 1363 |
/* this cgroup's local usage == 0 */ css_put(&victim->css); |
6d61ef409
|
1364 1365 |
continue; } |
04046e1a0
|
1366 |
/* we use swappiness of local cgroup */ |
4e4169535
|
1367 1368 |
if (check_soft) ret = mem_cgroup_shrink_node_zone(victim, gfp_mask, |
14fec7968
|
1369 |
noswap, get_swappiness(victim), zone); |
4e4169535
|
1370 1371 1372 |
else ret = try_to_free_mem_cgroup_pages(victim, gfp_mask, noswap, get_swappiness(victim)); |
04046e1a0
|
1373 |
css_put(&victim->css); |
81d39c20f
|
1374 1375 1376 1377 1378 1379 1380 |
/* * At shrinking usage, we can't check we should stop here or * reclaim more. It's depends on callers. last_scanned_child * will work enough for keeping fairness under tree. */ if (shrink) return ret; |
04046e1a0
|
1381 |
total += ret; |
4e4169535
|
1382 1383 1384 1385 |
if (check_soft) { if (res_counter_check_under_soft_limit(&root_mem->res)) return total; } else if (mem_cgroup_check_under_limit(root_mem)) |
04046e1a0
|
1386 |
return 1 + total; |
6d61ef409
|
1387 |
} |
04046e1a0
|
1388 |
return total; |
6d61ef409
|
1389 |
} |
867578cbc
|
1390 1391 1392 1393 1394 1395 |
/* * Check OOM-Killer is already running under our hierarchy. * If someone is running, return false. */ static bool mem_cgroup_oom_lock(struct mem_cgroup *mem) { |
7d74b06f2
|
1396 1397 |
int x, lock_count = 0; struct mem_cgroup *iter; |
a636b327f
|
1398 |
|
7d74b06f2
|
1399 1400 1401 1402 |
for_each_mem_cgroup_tree(iter, mem) { x = atomic_inc_return(&iter->oom_lock); lock_count = max(x, lock_count); } |
867578cbc
|
1403 1404 1405 1406 |
if (lock_count == 1) return true; return false; |
a636b327f
|
1407 |
} |
0b7f569e4
|
1408 |
|
7d74b06f2
|
1409 |
static int mem_cgroup_oom_unlock(struct mem_cgroup *mem) |
0b7f569e4
|
1410 |
{ |
7d74b06f2
|
1411 |
struct mem_cgroup *iter; |
867578cbc
|
1412 1413 1414 1415 1416 |
/* * When a new child is created while the hierarchy is under oom, * mem_cgroup_oom_lock() may not be called. We have to use * atomic_add_unless() here. */ |
7d74b06f2
|
1417 1418 |
for_each_mem_cgroup_tree(iter, mem) atomic_add_unless(&iter->oom_lock, -1, 0); |
0b7f569e4
|
1419 1420 |
return 0; } |
867578cbc
|
1421 1422 1423 |
static DEFINE_MUTEX(memcg_oom_mutex); static DECLARE_WAIT_QUEUE_HEAD(memcg_oom_waitq); |
dc98df5a1
|
1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 |
struct oom_wait_info { struct mem_cgroup *mem; wait_queue_t wait; }; static int memcg_oom_wake_function(wait_queue_t *wait, unsigned mode, int sync, void *arg) { struct mem_cgroup *wake_mem = (struct mem_cgroup *)arg; struct oom_wait_info *oom_wait_info; oom_wait_info = container_of(wait, struct oom_wait_info, wait); if (oom_wait_info->mem == wake_mem) goto wakeup; /* if no hierarchy, no match */ if (!oom_wait_info->mem->use_hierarchy || !wake_mem->use_hierarchy) return 0; /* * Both of oom_wait_info->mem and wake_mem are stable under us. * Then we can use css_is_ancestor without taking care of RCU. */ if (!css_is_ancestor(&oom_wait_info->mem->css, &wake_mem->css) && !css_is_ancestor(&wake_mem->css, &oom_wait_info->mem->css)) return 0; wakeup: return autoremove_wake_function(wait, mode, sync, arg); } static void memcg_wakeup_oom(struct mem_cgroup *mem) { /* for filtering, pass "mem" as argument. */ __wake_up(&memcg_oom_waitq, TASK_NORMAL, 0, mem); } |
3c11ecf44
|
1459 1460 |
static void memcg_oom_recover(struct mem_cgroup *mem) { |
2bd9bb206
|
1461 |
if (mem && atomic_read(&mem->oom_lock)) |
3c11ecf44
|
1462 1463 |
memcg_wakeup_oom(mem); } |
867578cbc
|
1464 1465 1466 1467 |
/* * try to call OOM killer. returns false if we should exit memory-reclaim loop. */ bool mem_cgroup_handle_oom(struct mem_cgroup *mem, gfp_t mask) |
0b7f569e4
|
1468 |
{ |
dc98df5a1
|
1469 |
struct oom_wait_info owait; |
3c11ecf44
|
1470 |
bool locked, need_to_kill; |
867578cbc
|
1471 |
|
dc98df5a1
|
1472 1473 1474 1475 1476 |
owait.mem = mem; owait.wait.flags = 0; owait.wait.func = memcg_oom_wake_function; owait.wait.private = current; INIT_LIST_HEAD(&owait.wait.task_list); |
3c11ecf44
|
1477 |
need_to_kill = true; |
867578cbc
|
1478 1479 1480 1481 1482 1483 1484 1485 |
/* At first, try to OOM lock hierarchy under mem.*/ mutex_lock(&memcg_oom_mutex); locked = mem_cgroup_oom_lock(mem); /* * Even if signal_pending(), we can't quit charge() loop without * accounting. So, UNINTERRUPTIBLE is appropriate. But SIGKILL * under OOM is always welcomed, use TASK_KILLABLE here. */ |
3c11ecf44
|
1486 1487 1488 1489 |
prepare_to_wait(&memcg_oom_waitq, &owait.wait, TASK_KILLABLE); if (!locked || mem->oom_kill_disable) need_to_kill = false; if (locked) |
9490ff275
|
1490 |
mem_cgroup_oom_notify(mem); |
867578cbc
|
1491 |
mutex_unlock(&memcg_oom_mutex); |
3c11ecf44
|
1492 1493 |
if (need_to_kill) { finish_wait(&memcg_oom_waitq, &owait.wait); |
867578cbc
|
1494 |
mem_cgroup_out_of_memory(mem, mask); |
3c11ecf44
|
1495 |
} else { |
867578cbc
|
1496 |
schedule(); |
dc98df5a1
|
1497 |
finish_wait(&memcg_oom_waitq, &owait.wait); |
867578cbc
|
1498 1499 1500 |
} mutex_lock(&memcg_oom_mutex); mem_cgroup_oom_unlock(mem); |
dc98df5a1
|
1501 |
memcg_wakeup_oom(mem); |
867578cbc
|
1502 1503 1504 1505 1506 1507 1508 |
mutex_unlock(&memcg_oom_mutex); if (test_thread_flag(TIF_MEMDIE) || fatal_signal_pending(current)) return false; /* Give chance to dying process */ schedule_timeout(1); return true; |
0b7f569e4
|
1509 |
} |
d69b042f3
|
1510 1511 1512 |
/* * Currently used to update mapped file statistics, but the routine can be * generalized to update other statistics as well. |
32047e2a8
|
1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 |
* * Notes: Race condition * * We usually use page_cgroup_lock() for accessing page_cgroup member but * it tends to be costly. But considering some conditions, we doesn't need * to do so _always_. * * Considering "charge", lock_page_cgroup() is not required because all * file-stat operations happen after a page is attached to radix-tree. There * are no race with "charge". * * Considering "uncharge", we know that memcg doesn't clear pc->mem_cgroup * at "uncharge" intentionally. So, we always see valid pc->mem_cgroup even * if there are race with "uncharge". Statistics itself is properly handled * by flags. * * Considering "move", this is an only case we see a race. To make the race * small, we check MEM_CGROUP_ON_MOVE percpu value and detect there are * possibility of race condition. If there is, we take a lock. |
d69b042f3
|
1532 |
*/ |
26174efd4
|
1533 |
|
2a7106f2c
|
1534 1535 |
void mem_cgroup_update_page_stat(struct page *page, enum mem_cgroup_page_stat_item idx, int val) |
d69b042f3
|
1536 1537 |
{ struct mem_cgroup *mem; |
32047e2a8
|
1538 1539 |
struct page_cgroup *pc = lookup_page_cgroup(page); bool need_unlock = false; |
dbd4ea78f
|
1540 |
unsigned long uninitialized_var(flags); |
d69b042f3
|
1541 |
|
d69b042f3
|
1542 1543 |
if (unlikely(!pc)) return; |
32047e2a8
|
1544 |
rcu_read_lock(); |
d69b042f3
|
1545 |
mem = pc->mem_cgroup; |
32047e2a8
|
1546 1547 1548 1549 1550 |
if (unlikely(!mem || !PageCgroupUsed(pc))) goto out; /* pc->mem_cgroup is unstable ? */ if (unlikely(mem_cgroup_stealed(mem))) { /* take a lock against to access pc->mem_cgroup */ |
dbd4ea78f
|
1551 |
move_lock_page_cgroup(pc, &flags); |
32047e2a8
|
1552 1553 1554 1555 1556 |
need_unlock = true; mem = pc->mem_cgroup; if (!mem || !PageCgroupUsed(pc)) goto out; } |
26174efd4
|
1557 |
|
26174efd4
|
1558 |
switch (idx) { |
2a7106f2c
|
1559 |
case MEMCG_NR_FILE_MAPPED: |
26174efd4
|
1560 1561 1562 |
if (val > 0) SetPageCgroupFileMapped(pc); else if (!page_mapped(page)) |
0c270f8f9
|
1563 |
ClearPageCgroupFileMapped(pc); |
2a7106f2c
|
1564 |
idx = MEM_CGROUP_STAT_FILE_MAPPED; |
26174efd4
|
1565 1566 1567 |
break; default: BUG(); |
8725d5416
|
1568 |
} |
d69b042f3
|
1569 |
|
2a7106f2c
|
1570 |
this_cpu_add(mem->stat->count[idx], val); |
32047e2a8
|
1571 1572 |
out: if (unlikely(need_unlock)) |
dbd4ea78f
|
1573 |
move_unlock_page_cgroup(pc, &flags); |
32047e2a8
|
1574 1575 |
rcu_read_unlock(); return; |
d69b042f3
|
1576 |
} |
2a7106f2c
|
1577 |
EXPORT_SYMBOL(mem_cgroup_update_page_stat); |
26174efd4
|
1578 |
|
f817ed485
|
1579 |
/* |
cdec2e426
|
1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 |
* size of first charge trial. "32" comes from vmscan.c's magic value. * TODO: maybe necessary to use big numbers in big irons. */ #define CHARGE_SIZE (32 * PAGE_SIZE) struct memcg_stock_pcp { struct mem_cgroup *cached; /* this never be root cgroup */ int charge; struct work_struct work; }; static DEFINE_PER_CPU(struct memcg_stock_pcp, memcg_stock); static atomic_t memcg_drain_count; /* * Try to consume stocked charge on this cpu. If success, PAGE_SIZE is consumed * from local stock and true is returned. If the stock is 0 or charges from a * cgroup which is not current target, returns false. This stock will be * refilled. */ static bool consume_stock(struct mem_cgroup *mem) { struct memcg_stock_pcp *stock; bool ret = true; stock = &get_cpu_var(memcg_stock); if (mem == stock->cached && stock->charge) stock->charge -= PAGE_SIZE; else /* need to call res_counter_charge */ ret = false; put_cpu_var(memcg_stock); return ret; } /* * Returns stocks cached in percpu to res_counter and reset cached information. */ static void drain_stock(struct memcg_stock_pcp *stock) { struct mem_cgroup *old = stock->cached; if (stock->charge) { res_counter_uncharge(&old->res, stock->charge); if (do_swap_account) res_counter_uncharge(&old->memsw, stock->charge); } stock->cached = NULL; stock->charge = 0; } /* * This must be called under preempt disabled or must be called by * a thread which is pinned to local cpu. */ static void drain_local_stock(struct work_struct *dummy) { struct memcg_stock_pcp *stock = &__get_cpu_var(memcg_stock); drain_stock(stock); } /* * Cache charges(val) which is from res_counter, to local per_cpu area. |
320cc51d9
|
1640 |
* This will be consumed by consume_stock() function, later. |
cdec2e426
|
1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 |
*/ static void refill_stock(struct mem_cgroup *mem, int val) { struct memcg_stock_pcp *stock = &get_cpu_var(memcg_stock); if (stock->cached != mem) { /* reset if necessary */ drain_stock(stock); stock->cached = mem; } stock->charge += val; put_cpu_var(memcg_stock); } /* * Tries to drain stocked charges in other cpus. This function is asynchronous * and just put a work per cpu for draining localy on each cpu. Caller can * expects some charges will be back to res_counter later but cannot wait for * it. */ static void drain_all_stock_async(void) { int cpu; /* This function is for scheduling "drain" in asynchronous way. * The result of "drain" is not directly handled by callers. Then, * if someone is calling drain, we don't have to call drain more. * Anyway, WORK_STRUCT_PENDING check in queue_work_on() will catch if * there is a race. We just do loose check here. */ if (atomic_read(&memcg_drain_count)) return; /* Notify other cpus that system-wide "drain" is running */ atomic_inc(&memcg_drain_count); get_online_cpus(); for_each_online_cpu(cpu) { struct memcg_stock_pcp *stock = &per_cpu(memcg_stock, cpu); schedule_work_on(cpu, &stock->work); } put_online_cpus(); atomic_dec(&memcg_drain_count); /* We don't wait for flush_work */ } /* This is a synchronous drain interface. */ static void drain_all_stock_sync(void) { /* called when force_empty is called */ atomic_inc(&memcg_drain_count); schedule_on_each_cpu(drain_local_stock); atomic_dec(&memcg_drain_count); } |
711d3d2c9
|
1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 |
/* * This function drains percpu counter value from DEAD cpu and * move it to local cpu. Note that this function can be preempted. */ static void mem_cgroup_drain_pcp_counter(struct mem_cgroup *mem, int cpu) { int i; spin_lock(&mem->pcp_counter_lock); for (i = 0; i < MEM_CGROUP_STAT_DATA; i++) { s64 x = per_cpu(mem->stat->count[i], cpu); per_cpu(mem->stat->count[i], cpu) = 0; mem->nocpu_base.count[i] += x; } |
1489ebad8
|
1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 |
/* need to clear ON_MOVE value, works as a kind of lock. */ per_cpu(mem->stat->count[MEM_CGROUP_ON_MOVE], cpu) = 0; spin_unlock(&mem->pcp_counter_lock); } static void synchronize_mem_cgroup_on_move(struct mem_cgroup *mem, int cpu) { int idx = MEM_CGROUP_ON_MOVE; spin_lock(&mem->pcp_counter_lock); per_cpu(mem->stat->count[idx], cpu) = mem->nocpu_base.count[idx]; |
711d3d2c9
|
1717 1718 1719 1720 |
spin_unlock(&mem->pcp_counter_lock); } static int __cpuinit memcg_cpu_hotplug_callback(struct notifier_block *nb, |
cdec2e426
|
1721 1722 1723 1724 1725 |
unsigned long action, void *hcpu) { int cpu = (unsigned long)hcpu; struct memcg_stock_pcp *stock; |
711d3d2c9
|
1726 |
struct mem_cgroup *iter; |
cdec2e426
|
1727 |
|
1489ebad8
|
1728 1729 1730 1731 1732 |
if ((action == CPU_ONLINE)) { for_each_mem_cgroup_all(iter) synchronize_mem_cgroup_on_move(iter, cpu); return NOTIFY_OK; } |
711d3d2c9
|
1733 |
if ((action != CPU_DEAD) || action != CPU_DEAD_FROZEN) |
cdec2e426
|
1734 |
return NOTIFY_OK; |
711d3d2c9
|
1735 1736 1737 |
for_each_mem_cgroup_all(iter) mem_cgroup_drain_pcp_counter(iter, cpu); |
cdec2e426
|
1738 1739 1740 1741 |
stock = &per_cpu(memcg_stock, cpu); drain_stock(stock); return NOTIFY_OK; } |
4b5343346
|
1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 |
/* See __mem_cgroup_try_charge() for details */ enum { CHARGE_OK, /* success */ CHARGE_RETRY, /* need to retry but retry is not bad */ CHARGE_NOMEM, /* we can't do more. return -ENOMEM */ CHARGE_WOULDBLOCK, /* GFP_WAIT wasn't set and no enough res. */ CHARGE_OOM_DIE, /* the current is killed because of OOM */ }; static int __mem_cgroup_do_charge(struct mem_cgroup *mem, gfp_t gfp_mask, int csize, bool oom_check) { struct mem_cgroup *mem_over_limit; struct res_counter *fail_res; unsigned long flags = 0; int ret; ret = res_counter_charge(&mem->res, csize, &fail_res); if (likely(!ret)) { if (!do_swap_account) return CHARGE_OK; ret = res_counter_charge(&mem->memsw, csize, &fail_res); if (likely(!ret)) return CHARGE_OK; mem_over_limit = mem_cgroup_from_res_counter(fail_res, memsw); flags |= MEM_CGROUP_RECLAIM_NOSWAP; } else mem_over_limit = mem_cgroup_from_res_counter(fail_res, res); if (csize > PAGE_SIZE) /* change csize and retry */ return CHARGE_RETRY; if (!(gfp_mask & __GFP_WAIT)) return CHARGE_WOULDBLOCK; ret = mem_cgroup_hierarchical_reclaim(mem_over_limit, NULL, gfp_mask, flags); /* * try_to_free_mem_cgroup_pages() might not give us a full * picture of reclaim. Some pages are reclaimed and might be * moved to swap cache or just unmapped from the cgroup. * Check the limit again to see if the reclaim reduced the * current usage of the cgroup before giving up */ if (ret || mem_cgroup_check_under_limit(mem_over_limit)) return CHARGE_RETRY; /* * At task move, charge accounts can be doubly counted. So, it's * better to wait until the end of task_move if something is going on. */ if (mem_cgroup_wait_acct_move(mem_over_limit)) return CHARGE_RETRY; /* If we don't need to call oom-killer at el, return immediately */ if (!oom_check) return CHARGE_NOMEM; /* check OOM */ if (!mem_cgroup_handle_oom(mem_over_limit, gfp_mask)) return CHARGE_OOM_DIE; return CHARGE_RETRY; } |
cdec2e426
|
1808 |
/* |
f817ed485
|
1809 1810 |
* Unlike exported interface, "oom" parameter is added. if oom==true, * oom-killer can be invoked. |
8a9f3ccd2
|
1811 |
*/ |
f817ed485
|
1812 |
static int __mem_cgroup_try_charge(struct mm_struct *mm, |
ec1685109
|
1813 1814 1815 |
gfp_t gfp_mask, struct mem_cgroup **memcg, bool oom, int page_size) |
8a9f3ccd2
|
1816 |
{ |
4b5343346
|
1817 1818 1819 |
int nr_oom_retries = MEM_CGROUP_RECLAIM_RETRIES; struct mem_cgroup *mem = NULL; int ret; |
ec1685109
|
1820 |
int csize = max(CHARGE_SIZE, (unsigned long) page_size); |
a636b327f
|
1821 |
|
867578cbc
|
1822 1823 1824 1825 1826 1827 1828 1829 |
/* * Unlike gloval-vm's OOM-kill, we're not in memory shortage * in system level. So, allow to go ahead dying process in addition to * MEMDIE process. */ if (unlikely(test_thread_flag(TIF_MEMDIE) || fatal_signal_pending(current))) goto bypass; |
a636b327f
|
1830 |
|
8a9f3ccd2
|
1831 |
/* |
3be91277e
|
1832 1833 |
* We always charge the cgroup the mm_struct belongs to. * The mm_struct's mem_cgroup changes on task migration if the |
8a9f3ccd2
|
1834 1835 1836 |
* thread group leader migrates. It's possible that mm is not * set, if so charge the init_mm (happens for pagecache usage). */ |
f75ca9620
|
1837 1838 1839 1840 |
if (!*memcg && !mm) goto bypass; again: if (*memcg) { /* css should be a valid one */ |
4b5343346
|
1841 |
mem = *memcg; |
f75ca9620
|
1842 1843 1844 |
VM_BUG_ON(css_is_removed(&mem->css)); if (mem_cgroup_is_root(mem)) goto done; |
ec1685109
|
1845 |
if (page_size == PAGE_SIZE && consume_stock(mem)) |
f75ca9620
|
1846 |
goto done; |
4b5343346
|
1847 1848 |
css_get(&mem->css); } else { |
f75ca9620
|
1849 |
struct task_struct *p; |
54595fe26
|
1850 |
|
f75ca9620
|
1851 1852 |
rcu_read_lock(); p = rcu_dereference(mm->owner); |
f75ca9620
|
1853 |
/* |
ebb76ce16
|
1854 1855 1856 1857 1858 1859 1860 1861 |
* Because we don't have task_lock(), "p" can exit. * In that case, "mem" can point to root or p can be NULL with * race with swapoff. Then, we have small risk of mis-accouning. * But such kind of mis-account by race always happens because * we don't have cgroup_mutex(). It's overkill and we allo that * small race, here. * (*) swapoff at el will charge against mm-struct not against * task-struct. So, mm->owner can be NULL. |
f75ca9620
|
1862 1863 |
*/ mem = mem_cgroup_from_task(p); |
ebb76ce16
|
1864 |
if (!mem || mem_cgroup_is_root(mem)) { |
f75ca9620
|
1865 1866 1867 |
rcu_read_unlock(); goto done; } |
ec1685109
|
1868 |
if (page_size == PAGE_SIZE && consume_stock(mem)) { |
f75ca9620
|
1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 |
/* * It seems dagerous to access memcg without css_get(). * But considering how consume_stok works, it's not * necessary. If consume_stock success, some charges * from this memcg are cached on this cpu. So, we * don't need to call css_get()/css_tryget() before * calling consume_stock(). */ rcu_read_unlock(); goto done; } /* after here, we may be blocked. we need to get refcnt */ if (!css_tryget(&mem->css)) { rcu_read_unlock(); goto again; } rcu_read_unlock(); } |
8a9f3ccd2
|
1887 |
|
4b5343346
|
1888 1889 |
do { bool oom_check; |
7a81b88cb
|
1890 |
|
4b5343346
|
1891 |
/* If killed, bypass charge */ |
f75ca9620
|
1892 1893 |
if (fatal_signal_pending(current)) { css_put(&mem->css); |
4b5343346
|
1894 |
goto bypass; |
f75ca9620
|
1895 |
} |
6d61ef409
|
1896 |
|
4b5343346
|
1897 1898 1899 1900 |
oom_check = false; if (oom && !nr_oom_retries) { oom_check = true; nr_oom_retries = MEM_CGROUP_RECLAIM_RETRIES; |
cdec2e426
|
1901 |
} |
66e1707bc
|
1902 |
|
4b5343346
|
1903 |
ret = __mem_cgroup_do_charge(mem, gfp_mask, csize, oom_check); |
8033b97c9
|
1904 |
|
4b5343346
|
1905 1906 1907 1908 |
switch (ret) { case CHARGE_OK: break; case CHARGE_RETRY: /* not in OOM situation but retry */ |
ec1685109
|
1909 |
csize = page_size; |
f75ca9620
|
1910 1911 1912 |
css_put(&mem->css); mem = NULL; goto again; |
4b5343346
|
1913 |
case CHARGE_WOULDBLOCK: /* !__GFP_WAIT */ |
f75ca9620
|
1914 |
css_put(&mem->css); |
4b5343346
|
1915 1916 |
goto nomem; case CHARGE_NOMEM: /* OOM routine works */ |
f75ca9620
|
1917 1918 |
if (!oom) { css_put(&mem->css); |
867578cbc
|
1919 |
goto nomem; |
f75ca9620
|
1920 |
} |
4b5343346
|
1921 1922 1923 1924 |
/* If oom, we never return -ENOMEM */ nr_oom_retries--; break; case CHARGE_OOM_DIE: /* Killed by OOM Killer */ |
f75ca9620
|
1925 |
css_put(&mem->css); |
867578cbc
|
1926 |
goto bypass; |
66e1707bc
|
1927 |
} |
4b5343346
|
1928 |
} while (ret != CHARGE_OK); |
ec1685109
|
1929 1930 |
if (csize > page_size) refill_stock(mem, csize - page_size); |
f75ca9620
|
1931 |
css_put(&mem->css); |
0c3e73e84
|
1932 |
done: |
f75ca9620
|
1933 |
*memcg = mem; |
7a81b88cb
|
1934 1935 |
return 0; nomem: |
f75ca9620
|
1936 |
*memcg = NULL; |
7a81b88cb
|
1937 |
return -ENOMEM; |
867578cbc
|
1938 1939 1940 |
bypass: *memcg = NULL; return 0; |
7a81b88cb
|
1941 |
} |
8a9f3ccd2
|
1942 |
|
a3b2d6926
|
1943 |
/* |
a3032a2c1
|
1944 1945 1946 1947 |
* Somemtimes we have to undo a charge we got by try_charge(). * This function is for that and do uncharge, put css's refcnt. * gotten by try_charge(). */ |
854ffa8d1
|
1948 1949 |
static void __mem_cgroup_cancel_charge(struct mem_cgroup *mem, unsigned long count) |
a3032a2c1
|
1950 1951 |
{ if (!mem_cgroup_is_root(mem)) { |
854ffa8d1
|
1952 |
res_counter_uncharge(&mem->res, PAGE_SIZE * count); |
a3032a2c1
|
1953 |
if (do_swap_account) |
854ffa8d1
|
1954 |
res_counter_uncharge(&mem->memsw, PAGE_SIZE * count); |
a3032a2c1
|
1955 |
} |
854ffa8d1
|
1956 |
} |
ec1685109
|
1957 1958 |
static void mem_cgroup_cancel_charge(struct mem_cgroup *mem, int page_size) |
854ffa8d1
|
1959 |
{ |
ec1685109
|
1960 |
__mem_cgroup_cancel_charge(mem, page_size >> PAGE_SHIFT); |
a3032a2c1
|
1961 1962 1963 |
} /* |
a3b2d6926
|
1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 |
* A helper function to get mem_cgroup from ID. must be called under * rcu_read_lock(). The caller must check css_is_removed() or some if * it's concern. (dropping refcnt from swap can be called against removed * memcg.) */ static struct mem_cgroup *mem_cgroup_lookup(unsigned short id) { struct cgroup_subsys_state *css; /* ID 0 is unused ID */ if (!id) return NULL; css = css_lookup(&mem_cgroup_subsys, id); if (!css) return NULL; return container_of(css, struct mem_cgroup, css); } |
e42d9d5d4
|
1981 |
struct mem_cgroup *try_get_mem_cgroup_from_page(struct page *page) |
b5a84319a
|
1982 |
{ |
e42d9d5d4
|
1983 |
struct mem_cgroup *mem = NULL; |
3c776e646
|
1984 |
struct page_cgroup *pc; |
a3b2d6926
|
1985 |
unsigned short id; |
b5a84319a
|
1986 |
swp_entry_t ent; |
3c776e646
|
1987 |
VM_BUG_ON(!PageLocked(page)); |
3c776e646
|
1988 |
pc = lookup_page_cgroup(page); |
c0bd3f63c
|
1989 |
lock_page_cgroup(pc); |
a3b2d6926
|
1990 |
if (PageCgroupUsed(pc)) { |
3c776e646
|
1991 |
mem = pc->mem_cgroup; |
a3b2d6926
|
1992 1993 |
if (mem && !css_tryget(&mem->css)) mem = NULL; |
e42d9d5d4
|
1994 |
} else if (PageSwapCache(page)) { |
3c776e646
|
1995 |
ent.val = page_private(page); |
a3b2d6926
|
1996 1997 1998 1999 2000 2001 |
id = lookup_swap_cgroup(ent); rcu_read_lock(); mem = mem_cgroup_lookup(id); if (mem && !css_tryget(&mem->css)) mem = NULL; rcu_read_unlock(); |
3c776e646
|
2002 |
} |
c0bd3f63c
|
2003 |
unlock_page_cgroup(pc); |
b5a84319a
|
2004 2005 |
return mem; } |
7a81b88cb
|
2006 |
/* |
a5e924f5f
|
2007 |
* commit a charge got by __mem_cgroup_try_charge() and makes page_cgroup to be |
7a81b88cb
|
2008 2009 |
* USED state. If already USED, uncharge and return. */ |
152c9ccb7
|
2010 2011 2012 |
static void ____mem_cgroup_commit_charge(struct mem_cgroup *mem, struct page_cgroup *pc, enum charge_type ctype) |
7a81b88cb
|
2013 |
{ |
8a9f3ccd2
|
2014 |
pc->mem_cgroup = mem; |
261fb61a8
|
2015 2016 2017 2018 2019 2020 2021 |
/* * We access a page_cgroup asynchronously without lock_page_cgroup(). * Especially when a page_cgroup is taken from a page, pc->mem_cgroup * is accessed after testing USED bit. To make pc->mem_cgroup visible * before USED bit, we need memory barrier here. * See mem_cgroup_add_lru_list(), etc. */ |
08e552c69
|
2022 |
smp_wmb(); |
4b3bde4c9
|
2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 |
switch (ctype) { case MEM_CGROUP_CHARGE_TYPE_CACHE: case MEM_CGROUP_CHARGE_TYPE_SHMEM: SetPageCgroupCache(pc); SetPageCgroupUsed(pc); break; case MEM_CGROUP_CHARGE_TYPE_MAPPED: ClearPageCgroupCache(pc); SetPageCgroupUsed(pc); break; default: break; } |
3be91277e
|
2036 |
|
08e552c69
|
2037 |
mem_cgroup_charge_statistics(mem, pc, true); |
152c9ccb7
|
2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 |
} static void __mem_cgroup_commit_charge(struct mem_cgroup *mem, struct page_cgroup *pc, enum charge_type ctype, int page_size) { int i; int count = page_size >> PAGE_SHIFT; /* try_charge() can return NULL to *memcg, taking care of it. */ if (!mem) return; lock_page_cgroup(pc); if (unlikely(PageCgroupUsed(pc))) { unlock_page_cgroup(pc); mem_cgroup_cancel_charge(mem, page_size); return; } /* * we don't need page_cgroup_lock about tail pages, becase they are not * accessed by any other context at this point. */ for (i = 0; i < count; i++) ____mem_cgroup_commit_charge(mem, pc + i, ctype); |
52d4b9ac0
|
2065 |
|
52d4b9ac0
|
2066 |
unlock_page_cgroup(pc); |
430e48631
|
2067 2068 2069 2070 2071 |
/* * "charge_statistics" updated event counter. Then, check it. * Insert ancestor (and ancestor's ancestors), to softlimit RB-tree. * if they exceeds softlimit. */ |
d2265e6fa
|
2072 |
memcg_check_events(mem, pc->page); |
7a81b88cb
|
2073 |
} |
66e1707bc
|
2074 |
|
f817ed485
|
2075 |
/** |
57f9fd7d2
|
2076 |
* __mem_cgroup_move_account - move account of the page |
f817ed485
|
2077 2078 2079 |
* @pc: page_cgroup of the page. * @from: mem_cgroup which the page is moved from. * @to: mem_cgroup which the page is moved to. @from != @to. |
854ffa8d1
|
2080 |
* @uncharge: whether we should call uncharge and css_put against @from. |
f817ed485
|
2081 2082 |
* * The caller must confirm following. |
08e552c69
|
2083 |
* - page is not on LRU (isolate_page() is useful.) |
57f9fd7d2
|
2084 |
* - the pc is locked, used, and ->mem_cgroup points to @from. |
f817ed485
|
2085 |
* |
854ffa8d1
|
2086 2087 2088 2089 |
* This function doesn't do "charge" nor css_get to new cgroup. It should be * done by a caller(__mem_cgroup_try_charge would be usefull). If @uncharge is * true, this function does "uncharge" from old cgroup, but it doesn't if * @uncharge is false, so a caller should do "uncharge". |
f817ed485
|
2090 |
*/ |
57f9fd7d2
|
2091 |
static void __mem_cgroup_move_account(struct page_cgroup *pc, |
854ffa8d1
|
2092 |
struct mem_cgroup *from, struct mem_cgroup *to, bool uncharge) |
f817ed485
|
2093 |
{ |
f817ed485
|
2094 |
VM_BUG_ON(from == to); |
08e552c69
|
2095 |
VM_BUG_ON(PageLRU(pc->page)); |
112bc2e12
|
2096 |
VM_BUG_ON(!page_is_cgroup_locked(pc)); |
57f9fd7d2
|
2097 2098 |
VM_BUG_ON(!PageCgroupUsed(pc)); VM_BUG_ON(pc->mem_cgroup != from); |
f817ed485
|
2099 |
|
8725d5416
|
2100 |
if (PageCgroupFileMapped(pc)) { |
c62b1a3b3
|
2101 2102 2103 2104 2105 |
/* Update mapped_file data for mem_cgroup */ preempt_disable(); __this_cpu_dec(from->stat->count[MEM_CGROUP_STAT_FILE_MAPPED]); __this_cpu_inc(to->stat->count[MEM_CGROUP_STAT_FILE_MAPPED]); preempt_enable(); |
d69b042f3
|
2106 |
} |
854ffa8d1
|
2107 2108 2109 |
mem_cgroup_charge_statistics(from, pc, false); if (uncharge) /* This is not "cancel", but cancel_charge does all we need. */ |
ec1685109
|
2110 |
mem_cgroup_cancel_charge(from, PAGE_SIZE); |
d69b042f3
|
2111 |
|
854ffa8d1
|
2112 |
/* caller should have done css_get */ |
08e552c69
|
2113 2114 |
pc->mem_cgroup = to; mem_cgroup_charge_statistics(to, pc, true); |
887032670
|
2115 2116 2117 |
/* * We charges against "to" which may not have any tasks. Then, "to" * can be under rmdir(). But in current implementation, caller of |
4ffef5fef
|
2118 2119 2120 |
* this function is just force_empty() and move charge, so it's * garanteed that "to" is never removed. So, we don't check rmdir * status here. |
887032670
|
2121 |
*/ |
57f9fd7d2
|
2122 2123 2124 2125 2126 2127 2128 |
} /* * check whether the @pc is valid for moving account and call * __mem_cgroup_move_account() */ static int mem_cgroup_move_account(struct page_cgroup *pc, |
854ffa8d1
|
2129 |
struct mem_cgroup *from, struct mem_cgroup *to, bool uncharge) |
57f9fd7d2
|
2130 2131 |
{ int ret = -EINVAL; |
dbd4ea78f
|
2132 |
unsigned long flags; |
57f9fd7d2
|
2133 2134 |
lock_page_cgroup(pc); if (PageCgroupUsed(pc) && pc->mem_cgroup == from) { |
dbd4ea78f
|
2135 |
move_lock_page_cgroup(pc, &flags); |
854ffa8d1
|
2136 |
__mem_cgroup_move_account(pc, from, to, uncharge); |
dbd4ea78f
|
2137 |
move_unlock_page_cgroup(pc, &flags); |
57f9fd7d2
|
2138 2139 2140 |
ret = 0; } unlock_page_cgroup(pc); |
d2265e6fa
|
2141 2142 2143 2144 2145 |
/* * check events */ memcg_check_events(to, pc->page); memcg_check_events(from, pc->page); |
f817ed485
|
2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 |
return ret; } /* * move charges to its parent. */ static int mem_cgroup_move_parent(struct page_cgroup *pc, struct mem_cgroup *child, gfp_t gfp_mask) { |
08e552c69
|
2157 |
struct page *page = pc->page; |
f817ed485
|
2158 2159 2160 |
struct cgroup *cg = child->css.cgroup; struct cgroup *pcg = cg->parent; struct mem_cgroup *parent; |
f817ed485
|
2161 2162 2163 2164 2165 |
int ret; /* Is ROOT ? */ if (!pcg) return -EINVAL; |
57f9fd7d2
|
2166 2167 2168 2169 2170 |
ret = -EBUSY; if (!get_page_unless_zero(page)) goto out; if (isolate_lru_page(page)) goto put; |
08e552c69
|
2171 |
|
f817ed485
|
2172 |
parent = mem_cgroup_from_cont(pcg); |
ec1685109
|
2173 2174 |
ret = __mem_cgroup_try_charge(NULL, gfp_mask, &parent, false, PAGE_SIZE); |
a636b327f
|
2175 |
if (ret || !parent) |
57f9fd7d2
|
2176 |
goto put_back; |
f817ed485
|
2177 |
|
854ffa8d1
|
2178 2179 |
ret = mem_cgroup_move_account(pc, child, parent, true); if (ret) |
ec1685109
|
2180 |
mem_cgroup_cancel_charge(parent, PAGE_SIZE); |
57f9fd7d2
|
2181 |
put_back: |
08e552c69
|
2182 |
putback_lru_page(page); |
57f9fd7d2
|
2183 |
put: |
40d58138f
|
2184 |
put_page(page); |
57f9fd7d2
|
2185 |
out: |
f817ed485
|
2186 2187 |
return ret; } |
7a81b88cb
|
2188 2189 2190 2191 2192 2193 2194 |
/* * Charge the memory controller for page usage. * Return * 0 if the charge was successful * < 0 if the cgroup is over its limit */ static int mem_cgroup_charge_common(struct page *page, struct mm_struct *mm, |
73045c47b
|
2195 |
gfp_t gfp_mask, enum charge_type ctype) |
7a81b88cb
|
2196 |
{ |
73045c47b
|
2197 |
struct mem_cgroup *mem = NULL; |
7a81b88cb
|
2198 2199 |
struct page_cgroup *pc; int ret; |
ec1685109
|
2200 |
int page_size = PAGE_SIZE; |
37c2ac787
|
2201 |
if (PageTransHuge(page)) { |
ec1685109
|
2202 |
page_size <<= compound_order(page); |
37c2ac787
|
2203 2204 |
VM_BUG_ON(!PageTransHuge(page)); } |
7a81b88cb
|
2205 2206 2207 2208 2209 2210 |
pc = lookup_page_cgroup(page); /* can happen at boot */ if (unlikely(!pc)) return 0; prefetchw(pc); |
ec1685109
|
2211 |
ret = __mem_cgroup_try_charge(mm, gfp_mask, &mem, true, page_size); |
a636b327f
|
2212 |
if (ret || !mem) |
7a81b88cb
|
2213 |
return ret; |
ec1685109
|
2214 |
__mem_cgroup_commit_charge(mem, pc, ctype, page_size); |
8a9f3ccd2
|
2215 |
return 0; |
8a9f3ccd2
|
2216 |
} |
7a81b88cb
|
2217 2218 |
int mem_cgroup_newpage_charge(struct page *page, struct mm_struct *mm, gfp_t gfp_mask) |
217bc3194
|
2219 |
{ |
f8d665422
|
2220 |
if (mem_cgroup_disabled()) |
cede86acd
|
2221 |
return 0; |
69029cd55
|
2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 |
/* * If already mapped, we don't have to account. * If page cache, page->mapping has address_space. * But page->mapping may have out-of-use anon_vma pointer, * detecit it by PageAnon() check. newly-mapped-anon's page->mapping * is NULL. */ if (page_mapped(page) || (page->mapping && !PageAnon(page))) return 0; if (unlikely(!mm)) mm = &init_mm; |
217bc3194
|
2233 |
return mem_cgroup_charge_common(page, mm, gfp_mask, |
73045c47b
|
2234 |
MEM_CGROUP_CHARGE_TYPE_MAPPED); |
217bc3194
|
2235 |
} |
83aae4c73
|
2236 2237 2238 |
static void __mem_cgroup_commit_charge_swapin(struct page *page, struct mem_cgroup *ptr, enum charge_type ctype); |
e1a1cd590
|
2239 2240 |
int mem_cgroup_cache_charge(struct page *page, struct mm_struct *mm, gfp_t gfp_mask) |
8697d3319
|
2241 |
{ |
b5a84319a
|
2242 |
int ret; |
f8d665422
|
2243 |
if (mem_cgroup_disabled()) |
cede86acd
|
2244 |
return 0; |
52d4b9ac0
|
2245 2246 |
if (PageCompound(page)) return 0; |
accf163e6
|
2247 2248 2249 2250 2251 2252 2253 2254 |
/* * Corner case handling. This is called from add_to_page_cache() * in usual. But some FS (shmem) precharges this page before calling it * and call add_to_page_cache() with GFP_NOWAIT. * * For GFP_NOWAIT case, the page may be pre-charged before calling * add_to_page_cache(). (See shmem.c) check it here and avoid to call * charge twice. (It works but has to pay a bit larger cost.) |
b5a84319a
|
2255 2256 |
* And when the page is SwapCache, it should take swap information * into account. This is under lock_page() now. |
accf163e6
|
2257 2258 2259 |
*/ if (!(gfp_mask & __GFP_WAIT)) { struct page_cgroup *pc; |
52d4b9ac0
|
2260 2261 2262 2263 2264 2265 |
pc = lookup_page_cgroup(page); if (!pc) return 0; lock_page_cgroup(pc); if (PageCgroupUsed(pc)) { unlock_page_cgroup(pc); |
accf163e6
|
2266 2267 |
return 0; } |
52d4b9ac0
|
2268 |
unlock_page_cgroup(pc); |
accf163e6
|
2269 |
} |
73045c47b
|
2270 |
if (unlikely(!mm)) |
8697d3319
|
2271 |
mm = &init_mm; |
accf163e6
|
2272 |
|
c05555b57
|
2273 2274 |
if (page_is_file_cache(page)) return mem_cgroup_charge_common(page, mm, gfp_mask, |
73045c47b
|
2275 |
MEM_CGROUP_CHARGE_TYPE_CACHE); |
b5a84319a
|
2276 |
|
83aae4c73
|
2277 2278 |
/* shmem */ if (PageSwapCache(page)) { |
73045c47b
|
2279 |
struct mem_cgroup *mem = NULL; |
83aae4c73
|
2280 2281 2282 2283 2284 2285 |
ret = mem_cgroup_try_charge_swapin(mm, page, gfp_mask, &mem); if (!ret) __mem_cgroup_commit_charge_swapin(page, mem, MEM_CGROUP_CHARGE_TYPE_SHMEM); } else ret = mem_cgroup_charge_common(page, mm, gfp_mask, |
73045c47b
|
2286 |
MEM_CGROUP_CHARGE_TYPE_SHMEM); |
b5a84319a
|
2287 |
|
b5a84319a
|
2288 |
return ret; |
e8589cc18
|
2289 |
} |
54595fe26
|
2290 2291 2292 |
/* * While swap-in, try_charge -> commit or cancel, the page is locked. * And when try_charge() successfully returns, one refcnt to memcg without |
21ae2956c
|
2293 |
* struct page_cgroup is acquired. This refcnt will be consumed by |
54595fe26
|
2294 2295 |
* "commit()" or removed by "cancel()" */ |
8c7c6e34a
|
2296 2297 2298 2299 2300 |
int mem_cgroup_try_charge_swapin(struct mm_struct *mm, struct page *page, gfp_t mask, struct mem_cgroup **ptr) { struct mem_cgroup *mem; |
54595fe26
|
2301 |
int ret; |
8c7c6e34a
|
2302 |
|
f8d665422
|
2303 |
if (mem_cgroup_disabled()) |
8c7c6e34a
|
2304 2305 2306 2307 |
return 0; if (!do_swap_account) goto charge_cur_mm; |
8c7c6e34a
|
2308 2309 |
/* * A racing thread's fault, or swapoff, may have already updated |
407f9c8b0
|
2310 2311 2312 |
* the pte, and even removed page from swap cache: in those cases * do_swap_page()'s pte_same() test will fail; but there's also a * KSM case which does need to charge the page. |
8c7c6e34a
|
2313 2314 |
*/ if (!PageSwapCache(page)) |
407f9c8b0
|
2315 |
goto charge_cur_mm; |
e42d9d5d4
|
2316 |
mem = try_get_mem_cgroup_from_page(page); |
54595fe26
|
2317 2318 |
if (!mem) goto charge_cur_mm; |
8c7c6e34a
|
2319 |
*ptr = mem; |
ec1685109
|
2320 |
ret = __mem_cgroup_try_charge(NULL, mask, ptr, true, PAGE_SIZE); |
54595fe26
|
2321 2322 |
css_put(&mem->css); return ret; |
8c7c6e34a
|
2323 2324 2325 |
charge_cur_mm: if (unlikely(!mm)) mm = &init_mm; |
ec1685109
|
2326 |
return __mem_cgroup_try_charge(mm, mask, ptr, true, PAGE_SIZE); |
8c7c6e34a
|
2327 |
} |
83aae4c73
|
2328 2329 2330 |
static void __mem_cgroup_commit_charge_swapin(struct page *page, struct mem_cgroup *ptr, enum charge_type ctype) |
7a81b88cb
|
2331 2332 |
{ struct page_cgroup *pc; |
f8d665422
|
2333 |
if (mem_cgroup_disabled()) |
7a81b88cb
|
2334 2335 2336 |
return; if (!ptr) return; |
887032670
|
2337 |
cgroup_exclude_rmdir(&ptr->css); |
7a81b88cb
|
2338 |
pc = lookup_page_cgroup(page); |
544122e5e
|
2339 |
mem_cgroup_lru_del_before_commit_swapcache(page); |
ec1685109
|
2340 |
__mem_cgroup_commit_charge(ptr, pc, ctype, PAGE_SIZE); |
544122e5e
|
2341 |
mem_cgroup_lru_add_after_commit_swapcache(page); |
8c7c6e34a
|
2342 2343 2344 |
/* * Now swap is on-memory. This means this page may be * counted both as mem and swap....double count. |
03f3c4336
|
2345 2346 2347 |
* Fix it by uncharging from memsw. Basically, this SwapCache is stable * under lock_page(). But in do_swap_page()::memory.c, reuse_swap_page() * may call delete_from_swap_cache() before reach here. |
8c7c6e34a
|
2348 |
*/ |
03f3c4336
|
2349 |
if (do_swap_account && PageSwapCache(page)) { |
8c7c6e34a
|
2350 |
swp_entry_t ent = {.val = page_private(page)}; |
a3b2d6926
|
2351 |
unsigned short id; |
8c7c6e34a
|
2352 |
struct mem_cgroup *memcg; |
a3b2d6926
|
2353 2354 2355 2356 |
id = swap_cgroup_record(ent, 0); rcu_read_lock(); memcg = mem_cgroup_lookup(id); |
8c7c6e34a
|
2357 |
if (memcg) { |
a3b2d6926
|
2358 2359 2360 2361 |
/* * This recorded memcg can be obsolete one. So, avoid * calling css_tryget */ |
0c3e73e84
|
2362 |
if (!mem_cgroup_is_root(memcg)) |
4e649152c
|
2363 |
res_counter_uncharge(&memcg->memsw, PAGE_SIZE); |
0c3e73e84
|
2364 |
mem_cgroup_swap_statistics(memcg, false); |
8c7c6e34a
|
2365 2366 |
mem_cgroup_put(memcg); } |
a3b2d6926
|
2367 |
rcu_read_unlock(); |
8c7c6e34a
|
2368 |
} |
887032670
|
2369 2370 2371 2372 2373 2374 |
/* * At swapin, we may charge account against cgroup which has no tasks. * So, rmdir()->pre_destroy() can be called while we do this charge. * In that case, we need to call pre_destroy() again. check it here. */ cgroup_release_and_wakeup_rmdir(&ptr->css); |
7a81b88cb
|
2375 |
} |
83aae4c73
|
2376 2377 2378 2379 2380 |
void mem_cgroup_commit_charge_swapin(struct page *page, struct mem_cgroup *ptr) { __mem_cgroup_commit_charge_swapin(page, ptr, MEM_CGROUP_CHARGE_TYPE_MAPPED); } |
7a81b88cb
|
2381 2382 |
void mem_cgroup_cancel_charge_swapin(struct mem_cgroup *mem) { |
f8d665422
|
2383 |
if (mem_cgroup_disabled()) |
7a81b88cb
|
2384 2385 2386 |
return; if (!mem) return; |
ec1685109
|
2387 |
mem_cgroup_cancel_charge(mem, PAGE_SIZE); |
7a81b88cb
|
2388 |
} |
569b846df
|
2389 |
static void |
ec1685109
|
2390 2391 |
__do_uncharge(struct mem_cgroup *mem, const enum charge_type ctype, int page_size) |
569b846df
|
2392 2393 2394 2395 2396 2397 |
{ struct memcg_batch_info *batch = NULL; bool uncharge_memsw = true; /* If swapout, usage of swap doesn't decrease */ if (!do_swap_account || ctype == MEM_CGROUP_CHARGE_TYPE_SWAPOUT) uncharge_memsw = false; |
569b846df
|
2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 |
batch = ¤t->memcg_batch; /* * In usual, we do css_get() when we remember memcg pointer. * But in this case, we keep res->usage until end of a series of * uncharges. Then, it's ok to ignore memcg's refcnt. */ if (!batch->memcg) batch->memcg = mem; /* |
3c11ecf44
|
2408 2409 2410 2411 2412 2413 2414 2415 2416 |
* do_batch > 0 when unmapping pages or inode invalidate/truncate. * In those cases, all pages freed continously can be expected to be in * the same cgroup and we have chance to coalesce uncharges. * But we do uncharge one by one if this is killed by OOM(TIF_MEMDIE) * because we want to do uncharge as soon as possible. */ if (!batch->do_batch || test_thread_flag(TIF_MEMDIE)) goto direct_uncharge; |
ec1685109
|
2417 2418 |
if (page_size != PAGE_SIZE) goto direct_uncharge; |
3c11ecf44
|
2419 |
/* |
569b846df
|
2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 |
* In typical case, batch->memcg == mem. This means we can * merge a series of uncharges to an uncharge of res_counter. * If not, we uncharge res_counter ony by one. */ if (batch->memcg != mem) goto direct_uncharge; /* remember freed charge and uncharge it later */ batch->bytes += PAGE_SIZE; if (uncharge_memsw) batch->memsw_bytes += PAGE_SIZE; return; direct_uncharge: |
ec1685109
|
2432 |
res_counter_uncharge(&mem->res, page_size); |
569b846df
|
2433 |
if (uncharge_memsw) |
ec1685109
|
2434 |
res_counter_uncharge(&mem->memsw, page_size); |
3c11ecf44
|
2435 2436 |
if (unlikely(batch->memcg != mem)) memcg_oom_recover(mem); |
569b846df
|
2437 2438 |
return; } |
7a81b88cb
|
2439 |
|
8697d3319
|
2440 |
/* |
69029cd55
|
2441 |
* uncharge if !page_mapped(page) |
8a9f3ccd2
|
2442 |
*/ |
8c7c6e34a
|
2443 |
static struct mem_cgroup * |
69029cd55
|
2444 |
__mem_cgroup_uncharge_common(struct page *page, enum charge_type ctype) |
8a9f3ccd2
|
2445 |
{ |
152c9ccb7
|
2446 2447 |
int i; int count; |
8289546e5
|
2448 |
struct page_cgroup *pc; |
8c7c6e34a
|
2449 |
struct mem_cgroup *mem = NULL; |
ec1685109
|
2450 |
int page_size = PAGE_SIZE; |
8a9f3ccd2
|
2451 |
|
f8d665422
|
2452 |
if (mem_cgroup_disabled()) |
8c7c6e34a
|
2453 |
return NULL; |
4077960e2
|
2454 |
|
d13d14430
|
2455 |
if (PageSwapCache(page)) |
8c7c6e34a
|
2456 |
return NULL; |
d13d14430
|
2457 |
|
37c2ac787
|
2458 |
if (PageTransHuge(page)) { |
ec1685109
|
2459 |
page_size <<= compound_order(page); |
37c2ac787
|
2460 2461 |
VM_BUG_ON(!PageTransHuge(page)); } |
ec1685109
|
2462 |
|
152c9ccb7
|
2463 |
count = page_size >> PAGE_SHIFT; |
8697d3319
|
2464 |
/* |
3c541e14b
|
2465 |
* Check if our page_cgroup is valid |
8697d3319
|
2466 |
*/ |
52d4b9ac0
|
2467 2468 |
pc = lookup_page_cgroup(page); if (unlikely(!pc || !PageCgroupUsed(pc))) |
8c7c6e34a
|
2469 |
return NULL; |
b9c565d5a
|
2470 |
|
52d4b9ac0
|
2471 |
lock_page_cgroup(pc); |
d13d14430
|
2472 |
|
8c7c6e34a
|
2473 |
mem = pc->mem_cgroup; |
d13d14430
|
2474 2475 2476 2477 2478 |
if (!PageCgroupUsed(pc)) goto unlock_out; switch (ctype) { case MEM_CGROUP_CHARGE_TYPE_MAPPED: |
8a9478ca7
|
2479 |
case MEM_CGROUP_CHARGE_TYPE_DROP: |
ac39cf8cb
|
2480 2481 |
/* See mem_cgroup_prepare_migration() */ if (page_mapped(page) || PageCgroupMigration(pc)) |
d13d14430
|
2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 |
goto unlock_out; break; case MEM_CGROUP_CHARGE_TYPE_SWAPOUT: if (!PageAnon(page)) { /* Shared memory */ if (page->mapping && !page_is_file_cache(page)) goto unlock_out; } else if (page_mapped(page)) /* Anon */ goto unlock_out; break; default: break; |
52d4b9ac0
|
2493 |
} |
d13d14430
|
2494 |
|
152c9ccb7
|
2495 2496 |
for (i = 0; i < count; i++) mem_cgroup_charge_statistics(mem, pc + i, false); |
04046e1a0
|
2497 |
|
52d4b9ac0
|
2498 |
ClearPageCgroupUsed(pc); |
544122e5e
|
2499 2500 2501 2502 2503 2504 |
/* * pc->mem_cgroup is not cleared here. It will be accessed when it's * freed from LRU. This is safe because uncharged page is expected not * to be reused (freed soon). Exception is SwapCache, it's handled by * special functions. */ |
b9c565d5a
|
2505 |
|
52d4b9ac0
|
2506 |
unlock_page_cgroup(pc); |
f75ca9620
|
2507 2508 2509 2510 |
/* * even after unlock, we have mem->res.usage here and this memcg * will never be freed. */ |
d2265e6fa
|
2511 |
memcg_check_events(mem, page); |
f75ca9620
|
2512 2513 2514 2515 2516 |
if (do_swap_account && ctype == MEM_CGROUP_CHARGE_TYPE_SWAPOUT) { mem_cgroup_swap_statistics(mem, true); mem_cgroup_get(mem); } if (!mem_cgroup_is_root(mem)) |
ec1685109
|
2517 |
__do_uncharge(mem, ctype, page_size); |
6d12e2d8d
|
2518 |
|
8c7c6e34a
|
2519 |
return mem; |
d13d14430
|
2520 2521 2522 |
unlock_out: unlock_page_cgroup(pc); |
8c7c6e34a
|
2523 |
return NULL; |
3c541e14b
|
2524 |
} |
69029cd55
|
2525 2526 |
void mem_cgroup_uncharge_page(struct page *page) { |
52d4b9ac0
|
2527 2528 2529 2530 2531 |
/* early check. */ if (page_mapped(page)) return; if (page->mapping && !PageAnon(page)) return; |
69029cd55
|
2532 2533 2534 2535 2536 2537 |
__mem_cgroup_uncharge_common(page, MEM_CGROUP_CHARGE_TYPE_MAPPED); } void mem_cgroup_uncharge_cache_page(struct page *page) { VM_BUG_ON(page_mapped(page)); |
b7abea963
|
2538 |
VM_BUG_ON(page->mapping); |
69029cd55
|
2539 2540 |
__mem_cgroup_uncharge_common(page, MEM_CGROUP_CHARGE_TYPE_CACHE); } |
569b846df
|
2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579 2580 |
/* * Batch_start/batch_end is called in unmap_page_range/invlidate/trucate. * In that cases, pages are freed continuously and we can expect pages * are in the same memcg. All these calls itself limits the number of * pages freed at once, then uncharge_start/end() is called properly. * This may be called prural(2) times in a context, */ void mem_cgroup_uncharge_start(void) { current->memcg_batch.do_batch++; /* We can do nest. */ if (current->memcg_batch.do_batch == 1) { current->memcg_batch.memcg = NULL; current->memcg_batch.bytes = 0; current->memcg_batch.memsw_bytes = 0; } } void mem_cgroup_uncharge_end(void) { struct memcg_batch_info *batch = ¤t->memcg_batch; if (!batch->do_batch) return; batch->do_batch--; if (batch->do_batch) /* If stacked, do nothing. */ return; if (!batch->memcg) return; /* * This "batch->memcg" is valid without any css_get/put etc... * bacause we hide charges behind us. */ if (batch->bytes) res_counter_uncharge(&batch->memcg->res, batch->bytes); if (batch->memsw_bytes) res_counter_uncharge(&batch->memcg->memsw, batch->memsw_bytes); |
3c11ecf44
|
2581 |
memcg_oom_recover(batch->memcg); |
569b846df
|
2582 2583 2584 |
/* forget this pointer (for sanity check) */ batch->memcg = NULL; } |
e767e0561
|
2585 |
#ifdef CONFIG_SWAP |
8c7c6e34a
|
2586 |
/* |
e767e0561
|
2587 |
* called after __delete_from_swap_cache() and drop "page" account. |
8c7c6e34a
|
2588 2589 |
* memcg information is recorded to swap_cgroup of "ent" */ |
8a9478ca7
|
2590 2591 |
void mem_cgroup_uncharge_swapcache(struct page *page, swp_entry_t ent, bool swapout) |
8c7c6e34a
|
2592 2593 |
{ struct mem_cgroup *memcg; |
8a9478ca7
|
2594 2595 2596 2597 2598 2599 |
int ctype = MEM_CGROUP_CHARGE_TYPE_SWAPOUT; if (!swapout) /* this was a swap cache but the swap is unused ! */ ctype = MEM_CGROUP_CHARGE_TYPE_DROP; memcg = __mem_cgroup_uncharge_common(page, ctype); |
8c7c6e34a
|
2600 |
|
f75ca9620
|
2601 2602 2603 2604 2605 |
/* * record memcg information, if swapout && memcg != NULL, * mem_cgroup_get() was called in uncharge(). */ if (do_swap_account && swapout && memcg) |
a3b2d6926
|
2606 |
swap_cgroup_record(ent, css_id(&memcg->css)); |
8c7c6e34a
|
2607 |
} |
e767e0561
|
2608 |
#endif |
8c7c6e34a
|
2609 2610 2611 2612 2613 2614 2615 |
#ifdef CONFIG_CGROUP_MEM_RES_CTLR_SWAP /* * called from swap_entry_free(). remove record in swap_cgroup and * uncharge "memsw" account. */ void mem_cgroup_uncharge_swap(swp_entry_t ent) |
d13d14430
|
2616 |
{ |
8c7c6e34a
|
2617 |
struct mem_cgroup *memcg; |
a3b2d6926
|
2618 |
unsigned short id; |
8c7c6e34a
|
2619 2620 2621 |
if (!do_swap_account) return; |
a3b2d6926
|
2622 2623 2624 |
id = swap_cgroup_record(ent, 0); rcu_read_lock(); memcg = mem_cgroup_lookup(id); |
8c7c6e34a
|
2625 |
if (memcg) { |
a3b2d6926
|
2626 2627 2628 2629 |
/* * We uncharge this because swap is freed. * This memcg can be obsolete one. We avoid calling css_tryget */ |
0c3e73e84
|
2630 |
if (!mem_cgroup_is_root(memcg)) |
4e649152c
|
2631 |
res_counter_uncharge(&memcg->memsw, PAGE_SIZE); |
0c3e73e84
|
2632 |
mem_cgroup_swap_statistics(memcg, false); |
8c7c6e34a
|
2633 2634 |
mem_cgroup_put(memcg); } |
a3b2d6926
|
2635 |
rcu_read_unlock(); |
d13d14430
|
2636 |
} |
024914477
|
2637 2638 2639 2640 2641 2642 |
/** * mem_cgroup_move_swap_account - move swap charge and swap_cgroup's record. * @entry: swap entry to be moved * @from: mem_cgroup which the entry is moved from * @to: mem_cgroup which the entry is moved to |
483c30b51
|
2643 |
* @need_fixup: whether we should fixup res_counters and refcounts. |
024914477
|
2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 |
* * It succeeds only when the swap_cgroup's record for this entry is the same * as the mem_cgroup's id of @from. * * Returns 0 on success, -EINVAL on failure. * * The caller must have charged to @to, IOW, called res_counter_charge() about * both res and memsw, and called css_get(). */ static int mem_cgroup_move_swap_account(swp_entry_t entry, |
483c30b51
|
2654 |
struct mem_cgroup *from, struct mem_cgroup *to, bool need_fixup) |
024914477
|
2655 2656 2657 2658 2659 2660 2661 |
{ unsigned short old_id, new_id; old_id = css_id(&from->css); new_id = css_id(&to->css); if (swap_cgroup_cmpxchg(entry, old_id, new_id) == old_id) { |
024914477
|
2662 |
mem_cgroup_swap_statistics(from, false); |
483c30b51
|
2663 |
mem_cgroup_swap_statistics(to, true); |
024914477
|
2664 |
/* |
483c30b51
|
2665 2666 2667 2668 2669 2670 |
* This function is only called from task migration context now. * It postpones res_counter and refcount handling till the end * of task migration(mem_cgroup_clear_mc()) for performance * improvement. But we cannot postpone mem_cgroup_get(to) * because if the process that has been moved to @to does * swap-in, the refcount of @to might be decreased to 0. |
024914477
|
2671 |
*/ |
024914477
|
2672 |
mem_cgroup_get(to); |
483c30b51
|
2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 |
if (need_fixup) { if (!mem_cgroup_is_root(from)) res_counter_uncharge(&from->memsw, PAGE_SIZE); mem_cgroup_put(from); /* * we charged both to->res and to->memsw, so we should * uncharge to->res. */ if (!mem_cgroup_is_root(to)) res_counter_uncharge(&to->res, PAGE_SIZE); |
483c30b51
|
2683 |
} |
024914477
|
2684 2685 2686 2687 2688 2689 |
return 0; } return -EINVAL; } #else static inline int mem_cgroup_move_swap_account(swp_entry_t entry, |
483c30b51
|
2690 |
struct mem_cgroup *from, struct mem_cgroup *to, bool need_fixup) |
024914477
|
2691 2692 2693 |
{ return -EINVAL; } |
8c7c6e34a
|
2694 |
#endif |
d13d14430
|
2695 |
|
ae41be374
|
2696 |
/* |
01b1ae63c
|
2697 2698 |
* Before starting migration, account PAGE_SIZE to mem_cgroup that the old * page belongs to. |
ae41be374
|
2699 |
*/ |
ac39cf8cb
|
2700 2701 |
int mem_cgroup_prepare_migration(struct page *page, struct page *newpage, struct mem_cgroup **ptr) |
ae41be374
|
2702 2703 |
{ struct page_cgroup *pc; |
e8589cc18
|
2704 |
struct mem_cgroup *mem = NULL; |
ac39cf8cb
|
2705 |
enum charge_type ctype; |
e8589cc18
|
2706 |
int ret = 0; |
8869b8f6e
|
2707 |
|
ec1685109
|
2708 |
VM_BUG_ON(PageTransHuge(page)); |
f8d665422
|
2709 |
if (mem_cgroup_disabled()) |
4077960e2
|
2710 |
return 0; |
52d4b9ac0
|
2711 2712 2713 |
pc = lookup_page_cgroup(page); lock_page_cgroup(pc); if (PageCgroupUsed(pc)) { |
e8589cc18
|
2714 2715 |
mem = pc->mem_cgroup; css_get(&mem->css); |
ac39cf8cb
|
2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 |
/* * At migrating an anonymous page, its mapcount goes down * to 0 and uncharge() will be called. But, even if it's fully * unmapped, migration may fail and this page has to be * charged again. We set MIGRATION flag here and delay uncharge * until end_migration() is called * * Corner Case Thinking * A) * When the old page was mapped as Anon and it's unmap-and-freed * while migration was ongoing. * If unmap finds the old page, uncharge() of it will be delayed * until end_migration(). If unmap finds a new page, it's * uncharged when it make mapcount to be 1->0. If unmap code * finds swap_migration_entry, the new page will not be mapped * and end_migration() will find it(mapcount==0). * * B) * When the old page was mapped but migraion fails, the kernel * remaps it. A charge for it is kept by MIGRATION flag even * if mapcount goes down to 0. We can do remap successfully * without charging it again. * * C) * The "old" page is under lock_page() until the end of * migration, so, the old page itself will not be swapped-out. * If the new page is swapped out before end_migraton, our * hook to usual swap-out path will catch the event. */ if (PageAnon(page)) SetPageCgroupMigration(pc); |
e8589cc18
|
2747 |
} |
52d4b9ac0
|
2748 |
unlock_page_cgroup(pc); |
ac39cf8cb
|
2749 2750 2751 2752 2753 2754 |
/* * If the page is not charged at this point, * we return here. */ if (!mem) return 0; |
01b1ae63c
|
2755 |
|
93d5c9be1
|
2756 |
*ptr = mem; |
ec1685109
|
2757 |
ret = __mem_cgroup_try_charge(NULL, GFP_KERNEL, ptr, false, PAGE_SIZE); |
ac39cf8cb
|
2758 2759 2760 2761 2762 2763 2764 2765 2766 2767 2768 2769 |
css_put(&mem->css);/* drop extra refcnt */ if (ret || *ptr == NULL) { if (PageAnon(page)) { lock_page_cgroup(pc); ClearPageCgroupMigration(pc); unlock_page_cgroup(pc); /* * The old page may be fully unmapped while we kept it. */ mem_cgroup_uncharge_page(page); } return -ENOMEM; |
e8589cc18
|
2770 |
} |
ac39cf8cb
|
2771 2772 2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 |
/* * We charge new page before it's used/mapped. So, even if unlock_page() * is called before end_migration, we can catch all events on this new * page. In the case new page is migrated but not remapped, new page's * mapcount will be finally 0 and we call uncharge in end_migration(). */ pc = lookup_page_cgroup(newpage); if (PageAnon(page)) ctype = MEM_CGROUP_CHARGE_TYPE_MAPPED; else if (page_is_file_cache(page)) ctype = MEM_CGROUP_CHARGE_TYPE_CACHE; else ctype = MEM_CGROUP_CHARGE_TYPE_SHMEM; |
ec1685109
|
2784 |
__mem_cgroup_commit_charge(mem, pc, ctype, PAGE_SIZE); |
e8589cc18
|
2785 |
return ret; |
ae41be374
|
2786 |
} |
8869b8f6e
|
2787 |
|
69029cd55
|
2788 |
/* remove redundant charge if migration failed*/ |
01b1ae63c
|
2789 |
void mem_cgroup_end_migration(struct mem_cgroup *mem, |
50de1dd96
|
2790 |
struct page *oldpage, struct page *newpage, bool migration_ok) |
ae41be374
|
2791 |
{ |
ac39cf8cb
|
2792 |
struct page *used, *unused; |
01b1ae63c
|
2793 |
struct page_cgroup *pc; |
01b1ae63c
|
2794 2795 2796 |
if (!mem) return; |
ac39cf8cb
|
2797 |
/* blocks rmdir() */ |
887032670
|
2798 |
cgroup_exclude_rmdir(&mem->css); |
50de1dd96
|
2799 |
if (!migration_ok) { |
ac39cf8cb
|
2800 2801 |
used = oldpage; unused = newpage; |
01b1ae63c
|
2802 |
} else { |
ac39cf8cb
|
2803 |
used = newpage; |
01b1ae63c
|
2804 2805 |
unused = oldpage; } |
69029cd55
|
2806 |
/* |
ac39cf8cb
|
2807 2808 2809 |
* We disallowed uncharge of pages under migration because mapcount * of the page goes down to zero, temporarly. * Clear the flag and check the page should be charged. |
01b1ae63c
|
2810 |
*/ |
ac39cf8cb
|
2811 2812 2813 2814 |
pc = lookup_page_cgroup(oldpage); lock_page_cgroup(pc); ClearPageCgroupMigration(pc); unlock_page_cgroup(pc); |
01b1ae63c
|
2815 |
|
ac39cf8cb
|
2816 |
__mem_cgroup_uncharge_common(unused, MEM_CGROUP_CHARGE_TYPE_FORCE); |
01b1ae63c
|
2817 |
/* |
ac39cf8cb
|
2818 2819 2820 2821 2822 2823 |
* If a page is a file cache, radix-tree replacement is very atomic * and we can skip this check. When it was an Anon page, its mapcount * goes down to 0. But because we added MIGRATION flage, it's not * uncharged yet. There are several case but page->mapcount check * and USED bit check in mem_cgroup_uncharge_page() will do enough * check. (see prepare_charge() also) |
69029cd55
|
2824 |
*/ |
ac39cf8cb
|
2825 2826 |
if (PageAnon(used)) mem_cgroup_uncharge_page(used); |
887032670
|
2827 |
/* |
ac39cf8cb
|
2828 2829 |
* At migration, we may charge account against cgroup which has no * tasks. |
887032670
|
2830 2831 2832 2833 |
* So, rmdir()->pre_destroy() can be called while we do this charge. * In that case, we need to call pre_destroy() again. check it here. */ cgroup_release_and_wakeup_rmdir(&mem->css); |
ae41be374
|
2834 |
} |
78fb74669
|
2835 |
|
cc8475822
|
2836 |
/* |
ae3abae64
|
2837 2838 2839 2840 2841 2842 |
* A call to try to shrink memory usage on charge failure at shmem's swapin. * Calling hierarchical_reclaim is not enough because we should update * last_oom_jiffies to prevent pagefault_out_of_memory from invoking global OOM. * Moreover considering hierarchy, we should reclaim from the mem_over_limit, * not from the memcg which this page would be charged to. * try_charge_swapin does all of these works properly. |
c9b0ed514
|
2843 |
*/ |
ae3abae64
|
2844 |
int mem_cgroup_shmem_charge_fallback(struct page *page, |
b5a84319a
|
2845 2846 |
struct mm_struct *mm, gfp_t gfp_mask) |
c9b0ed514
|
2847 |
{ |
b5a84319a
|
2848 |
struct mem_cgroup *mem = NULL; |
ae3abae64
|
2849 |
int ret; |
c9b0ed514
|
2850 |
|
f8d665422
|
2851 |
if (mem_cgroup_disabled()) |
cede86acd
|
2852 |
return 0; |
c9b0ed514
|
2853 |
|
ae3abae64
|
2854 2855 2856 |
ret = mem_cgroup_try_charge_swapin(mm, page, gfp_mask, &mem); if (!ret) mem_cgroup_cancel_charge_swapin(mem); /* it does !mem check */ |
c9b0ed514
|
2857 |
|
ae3abae64
|
2858 |
return ret; |
c9b0ed514
|
2859 |
} |
8c7c6e34a
|
2860 |
static DEFINE_MUTEX(set_limit_mutex); |
d38d2a758
|
2861 |
static int mem_cgroup_resize_limit(struct mem_cgroup *memcg, |
8c7c6e34a
|
2862 |
unsigned long long val) |
628f42355
|
2863 |
{ |
81d39c20f
|
2864 |
int retry_count; |
3c11ecf44
|
2865 |
u64 memswlimit, memlimit; |
628f42355
|
2866 |
int ret = 0; |
81d39c20f
|
2867 2868 |
int children = mem_cgroup_count_children(memcg); u64 curusage, oldusage; |
3c11ecf44
|
2869 |
int enlarge; |
81d39c20f
|
2870 2871 2872 2873 2874 2875 2876 2877 2878 |
/* * For keeping hierarchical_reclaim simple, how long we should retry * is depends on callers. We set our retry-count to be function * of # of children which we should visit in this loop. */ retry_count = MEM_CGROUP_RECLAIM_RETRIES * children; oldusage = res_counter_read_u64(&memcg->res, RES_USAGE); |
628f42355
|
2879 |
|
3c11ecf44
|
2880 |
enlarge = 0; |
8c7c6e34a
|
2881 |
while (retry_count) { |
628f42355
|
2882 2883 2884 2885 |
if (signal_pending(current)) { ret = -EINTR; break; } |
8c7c6e34a
|
2886 2887 2888 2889 2890 2891 2892 2893 2894 2895 |
/* * Rather than hide all in some function, I do this in * open coded manner. You see what this really does. * We have to guarantee mem->res.limit < mem->memsw.limit. */ mutex_lock(&set_limit_mutex); memswlimit = res_counter_read_u64(&memcg->memsw, RES_LIMIT); if (memswlimit < val) { ret = -EINVAL; mutex_unlock(&set_limit_mutex); |
628f42355
|
2896 2897 |
break; } |
3c11ecf44
|
2898 2899 2900 2901 |
memlimit = res_counter_read_u64(&memcg->res, RES_LIMIT); if (memlimit < val) enlarge = 1; |
8c7c6e34a
|
2902 |
ret = res_counter_set_limit(&memcg->res, val); |
22a668d7c
|
2903 2904 2905 2906 2907 2908 |
if (!ret) { if (memswlimit == val) memcg->memsw_is_minimum = true; else memcg->memsw_is_minimum = false; } |
8c7c6e34a
|
2909 2910 2911 2912 |
mutex_unlock(&set_limit_mutex); if (!ret) break; |
aa20d489c
|
2913 |
mem_cgroup_hierarchical_reclaim(memcg, NULL, GFP_KERNEL, |
4e4169535
|
2914 |
MEM_CGROUP_RECLAIM_SHRINK); |
81d39c20f
|
2915 2916 2917 2918 2919 2920 |
curusage = res_counter_read_u64(&memcg->res, RES_USAGE); /* Usage is reduced ? */ if (curusage >= oldusage) retry_count--; else oldusage = curusage; |
8c7c6e34a
|
2921 |
} |
3c11ecf44
|
2922 2923 |
if (!ret && enlarge) memcg_oom_recover(memcg); |
14797e236
|
2924 |
|
8c7c6e34a
|
2925 2926 |
return ret; } |
338c84310
|
2927 2928 |
static int mem_cgroup_resize_memsw_limit(struct mem_cgroup *memcg, unsigned long long val) |
8c7c6e34a
|
2929 |
{ |
81d39c20f
|
2930 |
int retry_count; |
3c11ecf44
|
2931 |
u64 memlimit, memswlimit, oldusage, curusage; |
81d39c20f
|
2932 2933 |
int children = mem_cgroup_count_children(memcg); int ret = -EBUSY; |
3c11ecf44
|
2934 |
int enlarge = 0; |
8c7c6e34a
|
2935 |
|
81d39c20f
|
2936 2937 2938 |
/* see mem_cgroup_resize_res_limit */ retry_count = children * MEM_CGROUP_RECLAIM_RETRIES; oldusage = res_counter_read_u64(&memcg->memsw, RES_USAGE); |
8c7c6e34a
|
2939 2940 2941 2942 2943 2944 2945 2946 2947 2948 2949 2950 2951 2952 2953 2954 2955 |
while (retry_count) { if (signal_pending(current)) { ret = -EINTR; break; } /* * Rather than hide all in some function, I do this in * open coded manner. You see what this really does. * We have to guarantee mem->res.limit < mem->memsw.limit. */ mutex_lock(&set_limit_mutex); memlimit = res_counter_read_u64(&memcg->res, RES_LIMIT); if (memlimit > val) { ret = -EINVAL; mutex_unlock(&set_limit_mutex); break; } |
3c11ecf44
|
2956 2957 2958 |
memswlimit = res_counter_read_u64(&memcg->memsw, RES_LIMIT); if (memswlimit < val) enlarge = 1; |
8c7c6e34a
|
2959 |
ret = res_counter_set_limit(&memcg->memsw, val); |
22a668d7c
|
2960 2961 2962 2963 2964 2965 |
if (!ret) { if (memlimit == val) memcg->memsw_is_minimum = true; else memcg->memsw_is_minimum = false; } |
8c7c6e34a
|
2966 2967 2968 2969 |
mutex_unlock(&set_limit_mutex); if (!ret) break; |
4e4169535
|
2970 |
mem_cgroup_hierarchical_reclaim(memcg, NULL, GFP_KERNEL, |
75822b449
|
2971 2972 |
MEM_CGROUP_RECLAIM_NOSWAP | MEM_CGROUP_RECLAIM_SHRINK); |
8c7c6e34a
|
2973 |
curusage = res_counter_read_u64(&memcg->memsw, RES_USAGE); |
81d39c20f
|
2974 |
/* Usage is reduced ? */ |
8c7c6e34a
|
2975 |
if (curusage >= oldusage) |
628f42355
|
2976 |
retry_count--; |
81d39c20f
|
2977 2978 |
else oldusage = curusage; |
628f42355
|
2979 |
} |
3c11ecf44
|
2980 2981 |
if (!ret && enlarge) memcg_oom_recover(memcg); |
628f42355
|
2982 2983 |
return ret; } |
4e4169535
|
2984 |
unsigned long mem_cgroup_soft_limit_reclaim(struct zone *zone, int order, |
00918b6ab
|
2985 |
gfp_t gfp_mask) |
4e4169535
|
2986 2987 2988 2989 2990 2991 |
{ unsigned long nr_reclaimed = 0; struct mem_cgroup_per_zone *mz, *next_mz = NULL; unsigned long reclaimed; int loop = 0; struct mem_cgroup_tree_per_zone *mctz; |
ef8745c1e
|
2992 |
unsigned long long excess; |
4e4169535
|
2993 2994 2995 |
if (order > 0) return 0; |
00918b6ab
|
2996 |
mctz = soft_limit_tree_node_zone(zone_to_nid(zone), zone_idx(zone)); |
4e4169535
|
2997 2998 2999 3000 3001 3002 3003 3004 3005 3006 3007 3008 3009 3010 3011 3012 3013 3014 3015 3016 3017 3018 3019 3020 3021 3022 3023 3024 3025 3026 3027 3028 3029 3030 3031 3032 3033 3034 3035 3036 3037 3038 3039 3040 3041 3042 |
/* * This loop can run a while, specially if mem_cgroup's continuously * keep exceeding their soft limit and putting the system under * pressure */ do { if (next_mz) mz = next_mz; else mz = mem_cgroup_largest_soft_limit_node(mctz); if (!mz) break; reclaimed = mem_cgroup_hierarchical_reclaim(mz->mem, zone, gfp_mask, MEM_CGROUP_RECLAIM_SOFT); nr_reclaimed += reclaimed; spin_lock(&mctz->lock); /* * If we failed to reclaim anything from this memory cgroup * it is time to move on to the next cgroup */ next_mz = NULL; if (!reclaimed) { do { /* * Loop until we find yet another one. * * By the time we get the soft_limit lock * again, someone might have aded the * group back on the RB tree. Iterate to * make sure we get a different mem. * mem_cgroup_largest_soft_limit_node returns * NULL if no other cgroup is present on * the tree */ next_mz = __mem_cgroup_largest_soft_limit_node(mctz); if (next_mz == mz) { css_put(&next_mz->mem->css); next_mz = NULL; } else /* next_mz == NULL or other memcg */ break; } while (1); } |
4e4169535
|
3043 |
__mem_cgroup_remove_exceeded(mz->mem, mz, mctz); |
ef8745c1e
|
3044 |
excess = res_counter_soft_limit_excess(&mz->mem->res); |
4e4169535
|
3045 3046 3047 3048 3049 3050 3051 3052 |
/* * One school of thought says that we should not add * back the node to the tree if reclaim returns 0. * But our reclaim could return 0, simply because due * to priority we are exposing a smaller subset of * memory to reclaim from. Consider this as a longer * term TODO. */ |
ef8745c1e
|
3053 3054 |
/* If excess == 0, no tree ops */ __mem_cgroup_insert_exceeded(mz->mem, mz, mctz, excess); |
4e4169535
|
3055 3056 3057 3058 3059 3060 3061 3062 3063 3064 3065 3066 3067 3068 3069 3070 3071 |
spin_unlock(&mctz->lock); css_put(&mz->mem->css); loop++; /* * Could not reclaim anything and there are no more * mem cgroups to try or we seem to be looping without * reclaiming anything. */ if (!nr_reclaimed && (next_mz == NULL || loop > MEM_CGROUP_MAX_SOFT_LIMIT_RECLAIM_LOOPS)) break; } while (!nr_reclaimed); if (next_mz) css_put(&next_mz->mem->css); return nr_reclaimed; } |
c9b0ed514
|
3072 |
/* |
cc8475822
|
3073 |
* This routine traverse page_cgroup in given list and drop them all. |
cc8475822
|
3074 3075 |
* *And* this routine doesn't reclaim page itself, just removes page_cgroup. */ |
f817ed485
|
3076 |
static int mem_cgroup_force_empty_list(struct mem_cgroup *mem, |
08e552c69
|
3077 |
int node, int zid, enum lru_list lru) |
cc8475822
|
3078 |
{ |
08e552c69
|
3079 3080 |
struct zone *zone; struct mem_cgroup_per_zone *mz; |
f817ed485
|
3081 |
struct page_cgroup *pc, *busy; |
08e552c69
|
3082 |
unsigned long flags, loop; |
072c56c13
|
3083 |
struct list_head *list; |
f817ed485
|
3084 |
int ret = 0; |
072c56c13
|
3085 |
|
08e552c69
|
3086 3087 |
zone = &NODE_DATA(node)->node_zones[zid]; mz = mem_cgroup_zoneinfo(mem, node, zid); |
b69408e88
|
3088 |
list = &mz->lists[lru]; |
cc8475822
|
3089 |
|
f817ed485
|
3090 3091 3092 3093 3094 3095 |
loop = MEM_CGROUP_ZSTAT(mz, lru); /* give some margin against EBUSY etc...*/ loop += 256; busy = NULL; while (loop--) { ret = 0; |
08e552c69
|
3096 |
spin_lock_irqsave(&zone->lru_lock, flags); |
f817ed485
|
3097 |
if (list_empty(list)) { |
08e552c69
|
3098 |
spin_unlock_irqrestore(&zone->lru_lock, flags); |
52d4b9ac0
|
3099 |
break; |
f817ed485
|
3100 3101 3102 3103 |
} pc = list_entry(list->prev, struct page_cgroup, lru); if (busy == pc) { list_move(&pc->lru, list); |
648bcc771
|
3104 |
busy = NULL; |
08e552c69
|
3105 |
spin_unlock_irqrestore(&zone->lru_lock, flags); |
f817ed485
|
3106 3107 |
continue; } |
08e552c69
|
3108 |
spin_unlock_irqrestore(&zone->lru_lock, flags); |
f817ed485
|
3109 |
|
2c26fdd70
|
3110 |
ret = mem_cgroup_move_parent(pc, mem, GFP_KERNEL); |
f817ed485
|
3111 |
if (ret == -ENOMEM) |
52d4b9ac0
|
3112 |
break; |
f817ed485
|
3113 3114 3115 3116 3117 3118 3119 |
if (ret == -EBUSY || ret == -EINVAL) { /* found lock contention or "pc" is obsolete. */ busy = pc; cond_resched(); } else busy = NULL; |
cc8475822
|
3120 |
} |
08e552c69
|
3121 |
|
f817ed485
|
3122 3123 3124 |
if (!ret && !list_empty(list)) return -EBUSY; return ret; |
cc8475822
|
3125 3126 3127 3128 3129 3130 |
} /* * make mem_cgroup's charge to be 0 if there is no task. * This enables deleting this mem_cgroup. */ |
c1e862c1f
|
3131 |
static int mem_cgroup_force_empty(struct mem_cgroup *mem, bool free_all) |
cc8475822
|
3132 |
{ |
f817ed485
|
3133 3134 3135 |
int ret; int node, zid, shrink; int nr_retries = MEM_CGROUP_RECLAIM_RETRIES; |
c1e862c1f
|
3136 |
struct cgroup *cgrp = mem->css.cgroup; |
8869b8f6e
|
3137 |
|
cc8475822
|
3138 |
css_get(&mem->css); |
f817ed485
|
3139 3140 |
shrink = 0; |
c1e862c1f
|
3141 3142 3143 |
/* should free all ? */ if (free_all) goto try_to_free; |
f817ed485
|
3144 |
move_account: |
fce664775
|
3145 |
do { |
f817ed485
|
3146 |
ret = -EBUSY; |
c1e862c1f
|
3147 3148 3149 3150 |
if (cgroup_task_count(cgrp) || !list_empty(&cgrp->children)) goto out; ret = -EINTR; if (signal_pending(current)) |
cc8475822
|
3151 |
goto out; |
52d4b9ac0
|
3152 3153 |
/* This is for making all *used* pages to be on LRU. */ lru_add_drain_all(); |
cdec2e426
|
3154 |
drain_all_stock_sync(); |
f817ed485
|
3155 |
ret = 0; |
32047e2a8
|
3156 |
mem_cgroup_start_move(mem); |
299b4eaa3
|
3157 |
for_each_node_state(node, N_HIGH_MEMORY) { |
f817ed485
|
3158 |
for (zid = 0; !ret && zid < MAX_NR_ZONES; zid++) { |
b69408e88
|
3159 |
enum lru_list l; |
f817ed485
|
3160 3161 |
for_each_lru(l) { ret = mem_cgroup_force_empty_list(mem, |
08e552c69
|
3162 |
node, zid, l); |
f817ed485
|
3163 3164 3165 |
if (ret) break; } |
1ecaab2bd
|
3166 |
} |
f817ed485
|
3167 3168 3169 |
if (ret) break; } |
32047e2a8
|
3170 |
mem_cgroup_end_move(mem); |
3c11ecf44
|
3171 |
memcg_oom_recover(mem); |
f817ed485
|
3172 3173 3174 |
/* it seems parent cgroup doesn't have enough mem */ if (ret == -ENOMEM) goto try_to_free; |
52d4b9ac0
|
3175 |
cond_resched(); |
fce664775
|
3176 3177 |
/* "ret" should also be checked to ensure all lists are empty. */ } while (mem->res.usage > 0 || ret); |
cc8475822
|
3178 3179 3180 |
out: css_put(&mem->css); return ret; |
f817ed485
|
3181 3182 |
try_to_free: |
c1e862c1f
|
3183 3184 |
/* returns EBUSY if there is a task or if we come here twice. */ if (cgroup_task_count(cgrp) || !list_empty(&cgrp->children) || shrink) { |
f817ed485
|
3185 3186 3187 |
ret = -EBUSY; goto out; } |
c1e862c1f
|
3188 3189 |
/* we call try-to-free pages for make this cgroup empty */ lru_add_drain_all(); |
f817ed485
|
3190 3191 3192 3193 |
/* try to free all pages in this cgroup */ shrink = 1; while (nr_retries && mem->res.usage > 0) { int progress; |
c1e862c1f
|
3194 3195 3196 3197 3198 |
if (signal_pending(current)) { ret = -EINTR; goto out; } |
a7885eb8a
|
3199 3200 |
progress = try_to_free_mem_cgroup_pages(mem, GFP_KERNEL, false, get_swappiness(mem)); |
c1e862c1f
|
3201 |
if (!progress) { |
f817ed485
|
3202 |
nr_retries--; |
c1e862c1f
|
3203 |
/* maybe some writeback is necessary */ |
8aa7e847d
|
3204 |
congestion_wait(BLK_RW_ASYNC, HZ/10); |
c1e862c1f
|
3205 |
} |
f817ed485
|
3206 3207 |
} |
08e552c69
|
3208 |
lru_add_drain(); |
f817ed485
|
3209 |
/* try move_account...there may be some *locked* pages. */ |
fce664775
|
3210 |
goto move_account; |
cc8475822
|
3211 |
} |
c1e862c1f
|
3212 3213 3214 3215 |
int mem_cgroup_force_empty_write(struct cgroup *cont, unsigned int event) { return mem_cgroup_force_empty(mem_cgroup_from_cont(cont), true); } |
18f59ea7d
|
3216 3217 3218 3219 3220 3221 3222 3223 3224 3225 3226 3227 3228 3229 3230 3231 3232 3233 |
static u64 mem_cgroup_hierarchy_read(struct cgroup *cont, struct cftype *cft) { return mem_cgroup_from_cont(cont)->use_hierarchy; } static int mem_cgroup_hierarchy_write(struct cgroup *cont, struct cftype *cft, u64 val) { int retval = 0; struct mem_cgroup *mem = mem_cgroup_from_cont(cont); struct cgroup *parent = cont->parent; struct mem_cgroup *parent_mem = NULL; if (parent) parent_mem = mem_cgroup_from_cont(parent); cgroup_lock(); /* |
af901ca18
|
3234 |
* If parent's use_hierarchy is set, we can't make any modifications |
18f59ea7d
|
3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 3245 3246 3247 3248 3249 3250 3251 3252 |
* in the child subtrees. If it is unset, then the change can * occur, provided the current cgroup has no children. * * For the root cgroup, parent_mem is NULL, we allow value to be * set if there are no children. */ if ((!parent_mem || !parent_mem->use_hierarchy) && (val == 1 || val == 0)) { if (list_empty(&cont->children)) mem->use_hierarchy = val; else retval = -EBUSY; } else retval = -EINVAL; cgroup_unlock(); return retval; } |
0c3e73e84
|
3253 |
|
7d74b06f2
|
3254 3255 |
static u64 mem_cgroup_get_recursive_idx_stat(struct mem_cgroup *mem, enum mem_cgroup_stat_index idx) |
0c3e73e84
|
3256 |
{ |
7d74b06f2
|
3257 3258 |
struct mem_cgroup *iter; s64 val = 0; |
0c3e73e84
|
3259 |
|
7d74b06f2
|
3260 3261 3262 3263 3264 3265 3266 |
/* each per cpu's value can be minus.Then, use s64 */ for_each_mem_cgroup_tree(iter, mem) val += mem_cgroup_read_stat(iter, idx); if (val < 0) /* race ? */ val = 0; return val; |
0c3e73e84
|
3267 |
} |
104f39284
|
3268 3269 |
static inline u64 mem_cgroup_usage(struct mem_cgroup *mem, bool swap) { |
7d74b06f2
|
3270 |
u64 val; |
104f39284
|
3271 3272 3273 3274 3275 3276 3277 |
if (!mem_cgroup_is_root(mem)) { if (!swap) return res_counter_read_u64(&mem->res, RES_USAGE); else return res_counter_read_u64(&mem->memsw, RES_USAGE); } |
7d74b06f2
|
3278 3279 |
val = mem_cgroup_get_recursive_idx_stat(mem, MEM_CGROUP_STAT_CACHE); val += mem_cgroup_get_recursive_idx_stat(mem, MEM_CGROUP_STAT_RSS); |
104f39284
|
3280 |
|
7d74b06f2
|
3281 3282 3283 |
if (swap) val += mem_cgroup_get_recursive_idx_stat(mem, MEM_CGROUP_STAT_SWAPOUT); |
104f39284
|
3284 3285 3286 |
return val << PAGE_SHIFT; } |
2c3daa722
|
3287 |
static u64 mem_cgroup_read(struct cgroup *cont, struct cftype *cft) |
8cdea7c05
|
3288 |
{ |
8c7c6e34a
|
3289 |
struct mem_cgroup *mem = mem_cgroup_from_cont(cont); |
104f39284
|
3290 |
u64 val; |
8c7c6e34a
|
3291 3292 3293 3294 3295 3296 |
int type, name; type = MEMFILE_TYPE(cft->private); name = MEMFILE_ATTR(cft->private); switch (type) { case _MEM: |
104f39284
|
3297 3298 3299 |
if (name == RES_USAGE) val = mem_cgroup_usage(mem, false); else |
0c3e73e84
|
3300 |
val = res_counter_read_u64(&mem->res, name); |
8c7c6e34a
|
3301 3302 |
break; case _MEMSWAP: |
104f39284
|
3303 3304 3305 |
if (name == RES_USAGE) val = mem_cgroup_usage(mem, true); else |
0c3e73e84
|
3306 |
val = res_counter_read_u64(&mem->memsw, name); |
8c7c6e34a
|
3307 3308 3309 3310 3311 3312 |
break; default: BUG(); break; } return val; |
8cdea7c05
|
3313 |
} |
628f42355
|
3314 3315 3316 3317 |
/* * The user of this function is... * RES_LIMIT. */ |
856c13aa1
|
3318 3319 |
static int mem_cgroup_write(struct cgroup *cont, struct cftype *cft, const char *buffer) |
8cdea7c05
|
3320 |
{ |
628f42355
|
3321 |
struct mem_cgroup *memcg = mem_cgroup_from_cont(cont); |
8c7c6e34a
|
3322 |
int type, name; |
628f42355
|
3323 3324 |
unsigned long long val; int ret; |
8c7c6e34a
|
3325 3326 3327 |
type = MEMFILE_TYPE(cft->private); name = MEMFILE_ATTR(cft->private); switch (name) { |
628f42355
|
3328 |
case RES_LIMIT: |
4b3bde4c9
|
3329 3330 3331 3332 |
if (mem_cgroup_is_root(memcg)) { /* Can't set limit on root */ ret = -EINVAL; break; } |
628f42355
|
3333 3334 |
/* This function does all necessary parse...reuse it */ ret = res_counter_memparse_write_strategy(buffer, &val); |
8c7c6e34a
|
3335 3336 3337 |
if (ret) break; if (type == _MEM) |
628f42355
|
3338 |
ret = mem_cgroup_resize_limit(memcg, val); |
8c7c6e34a
|
3339 3340 |
else ret = mem_cgroup_resize_memsw_limit(memcg, val); |
628f42355
|
3341 |
break; |
296c81d89
|
3342 3343 3344 3345 3346 3347 3348 3349 3350 3351 3352 3353 3354 3355 |
case RES_SOFT_LIMIT: ret = res_counter_memparse_write_strategy(buffer, &val); if (ret) break; /* * For memsw, soft limits are hard to implement in terms * of semantics, for now, we support soft limits for * control without swap */ if (type == _MEM) ret = res_counter_set_soft_limit(&memcg->res, val); else ret = -EINVAL; break; |
628f42355
|
3356 3357 3358 3359 3360 |
default: ret = -EINVAL; /* should be BUG() ? */ break; } return ret; |
8cdea7c05
|
3361 |
} |
fee7b548e
|
3362 3363 3364 3365 3366 3367 3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 3378 3379 3380 3381 3382 3383 3384 3385 3386 3387 3388 |
static void memcg_get_hierarchical_limit(struct mem_cgroup *memcg, unsigned long long *mem_limit, unsigned long long *memsw_limit) { struct cgroup *cgroup; unsigned long long min_limit, min_memsw_limit, tmp; min_limit = res_counter_read_u64(&memcg->res, RES_LIMIT); min_memsw_limit = res_counter_read_u64(&memcg->memsw, RES_LIMIT); cgroup = memcg->css.cgroup; if (!memcg->use_hierarchy) goto out; while (cgroup->parent) { cgroup = cgroup->parent; memcg = mem_cgroup_from_cont(cgroup); if (!memcg->use_hierarchy) break; tmp = res_counter_read_u64(&memcg->res, RES_LIMIT); min_limit = min(min_limit, tmp); tmp = res_counter_read_u64(&memcg->memsw, RES_LIMIT); min_memsw_limit = min(min_memsw_limit, tmp); } out: *mem_limit = min_limit; *memsw_limit = min_memsw_limit; return; } |
29f2a4dac
|
3389 |
static int mem_cgroup_reset(struct cgroup *cont, unsigned int event) |
c84872e16
|
3390 3391 |
{ struct mem_cgroup *mem; |
8c7c6e34a
|
3392 |
int type, name; |
c84872e16
|
3393 3394 |
mem = mem_cgroup_from_cont(cont); |
8c7c6e34a
|
3395 3396 3397 |
type = MEMFILE_TYPE(event); name = MEMFILE_ATTR(event); switch (name) { |
29f2a4dac
|
3398 |
case RES_MAX_USAGE: |
8c7c6e34a
|
3399 3400 3401 3402 |
if (type == _MEM) res_counter_reset_max(&mem->res); else res_counter_reset_max(&mem->memsw); |
29f2a4dac
|
3403 3404 |
break; case RES_FAILCNT: |
8c7c6e34a
|
3405 3406 3407 3408 |
if (type == _MEM) res_counter_reset_failcnt(&mem->res); else res_counter_reset_failcnt(&mem->memsw); |
29f2a4dac
|
3409 3410 |
break; } |
f64c3f549
|
3411 |
|
85cc59db1
|
3412 |
return 0; |
c84872e16
|
3413 |
} |
7dc74be03
|
3414 3415 3416 3417 3418 |
static u64 mem_cgroup_move_charge_read(struct cgroup *cgrp, struct cftype *cft) { return mem_cgroup_from_cont(cgrp)->move_charge_at_immigrate; } |
024914477
|
3419 |
#ifdef CONFIG_MMU |
7dc74be03
|
3420 3421 3422 3423 3424 3425 3426 3427 3428 3429 3430 3431 3432 3433 3434 3435 3436 3437 |
static int mem_cgroup_move_charge_write(struct cgroup *cgrp, struct cftype *cft, u64 val) { struct mem_cgroup *mem = mem_cgroup_from_cont(cgrp); if (val >= (1 << NR_MOVE_TYPE)) return -EINVAL; /* * We check this value several times in both in can_attach() and * attach(), so we need cgroup lock to prevent this value from being * inconsistent. */ cgroup_lock(); mem->move_charge_at_immigrate = val; cgroup_unlock(); return 0; } |
024914477
|
3438 3439 3440 3441 3442 3443 3444 |
#else static int mem_cgroup_move_charge_write(struct cgroup *cgrp, struct cftype *cft, u64 val) { return -ENOSYS; } #endif |
7dc74be03
|
3445 |
|
14067bb3e
|
3446 3447 3448 3449 3450 |
/* For read statistics */ enum { MCS_CACHE, MCS_RSS, |
d8046582d
|
3451 |
MCS_FILE_MAPPED, |
14067bb3e
|
3452 3453 |
MCS_PGPGIN, MCS_PGPGOUT, |
1dd3a2732
|
3454 |
MCS_SWAP, |
14067bb3e
|
3455 3456 3457 3458 3459 3460 3461 3462 3463 3464 |
MCS_INACTIVE_ANON, MCS_ACTIVE_ANON, MCS_INACTIVE_FILE, MCS_ACTIVE_FILE, MCS_UNEVICTABLE, NR_MCS_STAT, }; struct mcs_total_stat { s64 stat[NR_MCS_STAT]; |
d2ceb9b7d
|
3465 |
}; |
14067bb3e
|
3466 3467 3468 3469 3470 3471 |
struct { char *local_name; char *total_name; } memcg_stat_strings[NR_MCS_STAT] = { {"cache", "total_cache"}, {"rss", "total_rss"}, |
d69b042f3
|
3472 |
{"mapped_file", "total_mapped_file"}, |
14067bb3e
|
3473 3474 |
{"pgpgin", "total_pgpgin"}, {"pgpgout", "total_pgpgout"}, |
1dd3a2732
|
3475 |
{"swap", "total_swap"}, |
14067bb3e
|
3476 3477 3478 3479 3480 3481 |
{"inactive_anon", "total_inactive_anon"}, {"active_anon", "total_active_anon"}, {"inactive_file", "total_inactive_file"}, {"active_file", "total_active_file"}, {"unevictable", "total_unevictable"} }; |
7d74b06f2
|
3482 3483 |
static void mem_cgroup_get_local_stat(struct mem_cgroup *mem, struct mcs_total_stat *s) |
14067bb3e
|
3484 |
{ |
14067bb3e
|
3485 3486 3487 |
s64 val; /* per cpu stat */ |
c62b1a3b3
|
3488 |
val = mem_cgroup_read_stat(mem, MEM_CGROUP_STAT_CACHE); |
14067bb3e
|
3489 |
s->stat[MCS_CACHE] += val * PAGE_SIZE; |
c62b1a3b3
|
3490 |
val = mem_cgroup_read_stat(mem, MEM_CGROUP_STAT_RSS); |
14067bb3e
|
3491 |
s->stat[MCS_RSS] += val * PAGE_SIZE; |
c62b1a3b3
|
3492 |
val = mem_cgroup_read_stat(mem, MEM_CGROUP_STAT_FILE_MAPPED); |
d8046582d
|
3493 |
s->stat[MCS_FILE_MAPPED] += val * PAGE_SIZE; |
c62b1a3b3
|
3494 |
val = mem_cgroup_read_stat(mem, MEM_CGROUP_STAT_PGPGIN_COUNT); |
14067bb3e
|
3495 |
s->stat[MCS_PGPGIN] += val; |
c62b1a3b3
|
3496 |
val = mem_cgroup_read_stat(mem, MEM_CGROUP_STAT_PGPGOUT_COUNT); |
14067bb3e
|
3497 |
s->stat[MCS_PGPGOUT] += val; |
1dd3a2732
|
3498 |
if (do_swap_account) { |
c62b1a3b3
|
3499 |
val = mem_cgroup_read_stat(mem, MEM_CGROUP_STAT_SWAPOUT); |
1dd3a2732
|
3500 3501 |
s->stat[MCS_SWAP] += val * PAGE_SIZE; } |
14067bb3e
|
3502 3503 3504 3505 3506 3507 3508 3509 3510 3511 3512 3513 |
/* per zone stat */ val = mem_cgroup_get_local_zonestat(mem, LRU_INACTIVE_ANON); s->stat[MCS_INACTIVE_ANON] += val * PAGE_SIZE; val = mem_cgroup_get_local_zonestat(mem, LRU_ACTIVE_ANON); s->stat[MCS_ACTIVE_ANON] += val * PAGE_SIZE; val = mem_cgroup_get_local_zonestat(mem, LRU_INACTIVE_FILE); s->stat[MCS_INACTIVE_FILE] += val * PAGE_SIZE; val = mem_cgroup_get_local_zonestat(mem, LRU_ACTIVE_FILE); s->stat[MCS_ACTIVE_FILE] += val * PAGE_SIZE; val = mem_cgroup_get_local_zonestat(mem, LRU_UNEVICTABLE); s->stat[MCS_UNEVICTABLE] += val * PAGE_SIZE; |
14067bb3e
|
3514 3515 3516 3517 3518 |
} static void mem_cgroup_get_total_stat(struct mem_cgroup *mem, struct mcs_total_stat *s) { |
7d74b06f2
|
3519 3520 3521 3522 |
struct mem_cgroup *iter; for_each_mem_cgroup_tree(iter, mem) mem_cgroup_get_local_stat(iter, s); |
14067bb3e
|
3523 |
} |
c64745cf0
|
3524 3525 |
static int mem_control_stat_show(struct cgroup *cont, struct cftype *cft, struct cgroup_map_cb *cb) |
d2ceb9b7d
|
3526 |
{ |
d2ceb9b7d
|
3527 |
struct mem_cgroup *mem_cont = mem_cgroup_from_cont(cont); |
14067bb3e
|
3528 |
struct mcs_total_stat mystat; |
d2ceb9b7d
|
3529 |
int i; |
14067bb3e
|
3530 3531 |
memset(&mystat, 0, sizeof(mystat)); mem_cgroup_get_local_stat(mem_cont, &mystat); |
d2ceb9b7d
|
3532 |
|
1dd3a2732
|
3533 3534 3535 |
for (i = 0; i < NR_MCS_STAT; i++) { if (i == MCS_SWAP && !do_swap_account) continue; |
14067bb3e
|
3536 |
cb->fill(cb, memcg_stat_strings[i].local_name, mystat.stat[i]); |
1dd3a2732
|
3537 |
} |
7b854121e
|
3538 |
|
14067bb3e
|
3539 |
/* Hierarchical information */ |
fee7b548e
|
3540 3541 3542 3543 3544 3545 3546 |
{ unsigned long long limit, memsw_limit; memcg_get_hierarchical_limit(mem_cont, &limit, &memsw_limit); cb->fill(cb, "hierarchical_memory_limit", limit); if (do_swap_account) cb->fill(cb, "hierarchical_memsw_limit", memsw_limit); } |
7f016ee8b
|
3547 |
|
14067bb3e
|
3548 3549 |
memset(&mystat, 0, sizeof(mystat)); mem_cgroup_get_total_stat(mem_cont, &mystat); |
1dd3a2732
|
3550 3551 3552 |
for (i = 0; i < NR_MCS_STAT; i++) { if (i == MCS_SWAP && !do_swap_account) continue; |
14067bb3e
|
3553 |
cb->fill(cb, memcg_stat_strings[i].total_name, mystat.stat[i]); |
1dd3a2732
|
3554 |
} |
14067bb3e
|
3555 |
|
7f016ee8b
|
3556 |
#ifdef CONFIG_DEBUG_VM |
c772be939
|
3557 |
cb->fill(cb, "inactive_ratio", calc_inactive_ratio(mem_cont, NULL)); |
7f016ee8b
|
3558 3559 3560 3561 3562 3563 3564 3565 3566 3567 3568 3569 3570 3571 3572 3573 3574 3575 3576 3577 3578 3579 3580 3581 3582 3583 |
{ int nid, zid; struct mem_cgroup_per_zone *mz; unsigned long recent_rotated[2] = {0, 0}; unsigned long recent_scanned[2] = {0, 0}; for_each_online_node(nid) for (zid = 0; zid < MAX_NR_ZONES; zid++) { mz = mem_cgroup_zoneinfo(mem_cont, nid, zid); recent_rotated[0] += mz->reclaim_stat.recent_rotated[0]; recent_rotated[1] += mz->reclaim_stat.recent_rotated[1]; recent_scanned[0] += mz->reclaim_stat.recent_scanned[0]; recent_scanned[1] += mz->reclaim_stat.recent_scanned[1]; } cb->fill(cb, "recent_rotated_anon", recent_rotated[0]); cb->fill(cb, "recent_rotated_file", recent_rotated[1]); cb->fill(cb, "recent_scanned_anon", recent_scanned[0]); cb->fill(cb, "recent_scanned_file", recent_scanned[1]); } #endif |
d2ceb9b7d
|
3584 3585 |
return 0; } |
a7885eb8a
|
3586 3587 3588 3589 3590 3591 3592 3593 3594 3595 3596 3597 |
static u64 mem_cgroup_swappiness_read(struct cgroup *cgrp, struct cftype *cft) { struct mem_cgroup *memcg = mem_cgroup_from_cont(cgrp); return get_swappiness(memcg); } static int mem_cgroup_swappiness_write(struct cgroup *cgrp, struct cftype *cft, u64 val) { struct mem_cgroup *memcg = mem_cgroup_from_cont(cgrp); struct mem_cgroup *parent; |
068b38c1f
|
3598 |
|
a7885eb8a
|
3599 3600 3601 3602 3603 3604 3605 |
if (val > 100) return -EINVAL; if (cgrp->parent == NULL) return -EINVAL; parent = mem_cgroup_from_cont(cgrp->parent); |
068b38c1f
|
3606 3607 |
cgroup_lock(); |
a7885eb8a
|
3608 3609 |
/* If under hierarchy, only empty-root can set this value */ if ((parent->use_hierarchy) || |
068b38c1f
|
3610 3611 |
(memcg->use_hierarchy && !list_empty(&cgrp->children))) { cgroup_unlock(); |
a7885eb8a
|
3612 |
return -EINVAL; |
068b38c1f
|
3613 |
} |
a7885eb8a
|
3614 3615 3616 3617 |
spin_lock(&memcg->reclaim_param_lock); memcg->swappiness = val; spin_unlock(&memcg->reclaim_param_lock); |
068b38c1f
|
3618 |
cgroup_unlock(); |
a7885eb8a
|
3619 3620 |
return 0; } |
2e72b6347
|
3621 3622 3623 3624 3625 3626 3627 3628 |
static void __mem_cgroup_threshold(struct mem_cgroup *memcg, bool swap) { struct mem_cgroup_threshold_ary *t; u64 usage; int i; rcu_read_lock(); if (!swap) |
2c488db27
|
3629 |
t = rcu_dereference(memcg->thresholds.primary); |
2e72b6347
|
3630 |
else |
2c488db27
|
3631 |
t = rcu_dereference(memcg->memsw_thresholds.primary); |
2e72b6347
|
3632 3633 3634 3635 3636 3637 3638 3639 3640 3641 3642 |
if (!t) goto unlock; usage = mem_cgroup_usage(memcg, swap); /* * current_threshold points to threshold just below usage. * If it's not true, a threshold was crossed after last * call of __mem_cgroup_threshold(). */ |
5407a5625
|
3643 |
i = t->current_threshold; |
2e72b6347
|
3644 3645 3646 3647 3648 3649 3650 3651 3652 3653 3654 3655 3656 3657 3658 3659 3660 3661 3662 3663 3664 3665 3666 |
/* * Iterate backward over array of thresholds starting from * current_threshold and check if a threshold is crossed. * If none of thresholds below usage is crossed, we read * only one element of the array here. */ for (; i >= 0 && unlikely(t->entries[i].threshold > usage); i--) eventfd_signal(t->entries[i].eventfd, 1); /* i = current_threshold + 1 */ i++; /* * Iterate forward over array of thresholds starting from * current_threshold+1 and check if a threshold is crossed. * If none of thresholds above usage is crossed, we read * only one element of the array here. */ for (; i < t->size && unlikely(t->entries[i].threshold <= usage); i++) eventfd_signal(t->entries[i].eventfd, 1); /* Update current_threshold */ |
5407a5625
|
3667 |
t->current_threshold = i - 1; |
2e72b6347
|
3668 3669 3670 3671 3672 3673 |
unlock: rcu_read_unlock(); } static void mem_cgroup_threshold(struct mem_cgroup *memcg) { |
ad4ca5f4b
|
3674 3675 3676 3677 3678 3679 3680 |
while (memcg) { __mem_cgroup_threshold(memcg, false); if (do_swap_account) __mem_cgroup_threshold(memcg, true); memcg = parent_mem_cgroup(memcg); } |
2e72b6347
|
3681 3682 3683 3684 3685 3686 3687 3688 3689 |
} static int compare_thresholds(const void *a, const void *b) { const struct mem_cgroup_threshold *_a = a; const struct mem_cgroup_threshold *_b = b; return _a->threshold - _b->threshold; } |
7d74b06f2
|
3690 |
static int mem_cgroup_oom_notify_cb(struct mem_cgroup *mem) |
9490ff275
|
3691 3692 3693 3694 3695 3696 3697 3698 3699 3700 |
{ struct mem_cgroup_eventfd_list *ev; list_for_each_entry(ev, &mem->oom_notify, list) eventfd_signal(ev->eventfd, 1); return 0; } static void mem_cgroup_oom_notify(struct mem_cgroup *mem) { |
7d74b06f2
|
3701 3702 3703 3704 |
struct mem_cgroup *iter; for_each_mem_cgroup_tree(iter, mem) mem_cgroup_oom_notify_cb(iter); |
9490ff275
|
3705 3706 3707 3708 |
} static int mem_cgroup_usage_register_event(struct cgroup *cgrp, struct cftype *cft, struct eventfd_ctx *eventfd, const char *args) |
2e72b6347
|
3709 3710 |
{ struct mem_cgroup *memcg = mem_cgroup_from_cont(cgrp); |
2c488db27
|
3711 3712 |
struct mem_cgroup_thresholds *thresholds; struct mem_cgroup_threshold_ary *new; |
2e72b6347
|
3713 3714 |
int type = MEMFILE_TYPE(cft->private); u64 threshold, usage; |
2c488db27
|
3715 |
int i, size, ret; |
2e72b6347
|
3716 3717 3718 3719 3720 3721 |
ret = res_counter_memparse_write_strategy(args, &threshold); if (ret) return ret; mutex_lock(&memcg->thresholds_lock); |
2c488db27
|
3722 |
|
2e72b6347
|
3723 |
if (type == _MEM) |
2c488db27
|
3724 |
thresholds = &memcg->thresholds; |
2e72b6347
|
3725 |
else if (type == _MEMSWAP) |
2c488db27
|
3726 |
thresholds = &memcg->memsw_thresholds; |
2e72b6347
|
3727 3728 3729 3730 3731 3732 |
else BUG(); usage = mem_cgroup_usage(memcg, type == _MEMSWAP); /* Check if a threshold crossed before adding a new one */ |
2c488db27
|
3733 |
if (thresholds->primary) |
2e72b6347
|
3734 |
__mem_cgroup_threshold(memcg, type == _MEMSWAP); |
2c488db27
|
3735 |
size = thresholds->primary ? thresholds->primary->size + 1 : 1; |
2e72b6347
|
3736 3737 |
/* Allocate memory for new array of thresholds */ |
2c488db27
|
3738 |
new = kmalloc(sizeof(*new) + size * sizeof(struct mem_cgroup_threshold), |
2e72b6347
|
3739 |
GFP_KERNEL); |
2c488db27
|
3740 |
if (!new) { |
2e72b6347
|
3741 3742 3743 |
ret = -ENOMEM; goto unlock; } |
2c488db27
|
3744 |
new->size = size; |
2e72b6347
|
3745 3746 |
/* Copy thresholds (if any) to new array */ |
2c488db27
|
3747 3748 |
if (thresholds->primary) { memcpy(new->entries, thresholds->primary->entries, (size - 1) * |
2e72b6347
|
3749 |
sizeof(struct mem_cgroup_threshold)); |
2c488db27
|
3750 |
} |
2e72b6347
|
3751 |
/* Add new threshold */ |
2c488db27
|
3752 3753 |
new->entries[size - 1].eventfd = eventfd; new->entries[size - 1].threshold = threshold; |
2e72b6347
|
3754 3755 |
/* Sort thresholds. Registering of new threshold isn't time-critical */ |
2c488db27
|
3756 |
sort(new->entries, size, sizeof(struct mem_cgroup_threshold), |
2e72b6347
|
3757 3758 3759 |
compare_thresholds, NULL); /* Find current threshold */ |
2c488db27
|
3760 |
new->current_threshold = -1; |
2e72b6347
|
3761 |
for (i = 0; i < size; i++) { |
2c488db27
|
3762 |
if (new->entries[i].threshold < usage) { |
2e72b6347
|
3763 |
/* |
2c488db27
|
3764 3765 |
* new->current_threshold will not be used until * rcu_assign_pointer(), so it's safe to increment |
2e72b6347
|
3766 3767 |
* it here. */ |
2c488db27
|
3768 |
++new->current_threshold; |
2e72b6347
|
3769 3770 |
} } |
2c488db27
|
3771 3772 3773 3774 3775 |
/* Free old spare buffer and save old primary buffer as spare */ kfree(thresholds->spare); thresholds->spare = thresholds->primary; rcu_assign_pointer(thresholds->primary, new); |
2e72b6347
|
3776 |
|
907860ed3
|
3777 |
/* To be sure that nobody uses thresholds */ |
2e72b6347
|
3778 |
synchronize_rcu(); |
2e72b6347
|
3779 3780 3781 3782 3783 |
unlock: mutex_unlock(&memcg->thresholds_lock); return ret; } |
907860ed3
|
3784 |
static void mem_cgroup_usage_unregister_event(struct cgroup *cgrp, |
9490ff275
|
3785 |
struct cftype *cft, struct eventfd_ctx *eventfd) |
2e72b6347
|
3786 3787 |
{ struct mem_cgroup *memcg = mem_cgroup_from_cont(cgrp); |
2c488db27
|
3788 3789 |
struct mem_cgroup_thresholds *thresholds; struct mem_cgroup_threshold_ary *new; |
2e72b6347
|
3790 3791 |
int type = MEMFILE_TYPE(cft->private); u64 usage; |
2c488db27
|
3792 |
int i, j, size; |
2e72b6347
|
3793 3794 3795 |
mutex_lock(&memcg->thresholds_lock); if (type == _MEM) |
2c488db27
|
3796 |
thresholds = &memcg->thresholds; |
2e72b6347
|
3797 |
else if (type == _MEMSWAP) |
2c488db27
|
3798 |
thresholds = &memcg->memsw_thresholds; |
2e72b6347
|
3799 3800 3801 3802 3803 3804 3805 3806 3807 3808 3809 3810 3811 3812 3813 |
else BUG(); /* * Something went wrong if we trying to unregister a threshold * if we don't have thresholds */ BUG_ON(!thresholds); usage = mem_cgroup_usage(memcg, type == _MEMSWAP); /* Check if a threshold crossed before removing */ __mem_cgroup_threshold(memcg, type == _MEMSWAP); /* Calculate new number of threshold */ |
2c488db27
|
3814 3815 3816 |
size = 0; for (i = 0; i < thresholds->primary->size; i++) { if (thresholds->primary->entries[i].eventfd != eventfd) |
2e72b6347
|
3817 3818 |
size++; } |
2c488db27
|
3819 |
new = thresholds->spare; |
907860ed3
|
3820 |
|
2e72b6347
|
3821 3822 |
/* Set thresholds array to NULL if we don't have thresholds */ if (!size) { |
2c488db27
|
3823 3824 |
kfree(new); new = NULL; |
907860ed3
|
3825 |
goto swap_buffers; |
2e72b6347
|
3826 |
} |
2c488db27
|
3827 |
new->size = size; |
2e72b6347
|
3828 3829 |
/* Copy thresholds and find current threshold */ |
2c488db27
|
3830 3831 3832 |
new->current_threshold = -1; for (i = 0, j = 0; i < thresholds->primary->size; i++) { if (thresholds->primary->entries[i].eventfd == eventfd) |
2e72b6347
|
3833 |
continue; |
2c488db27
|
3834 3835 |
new->entries[j] = thresholds->primary->entries[i]; if (new->entries[j].threshold < usage) { |
2e72b6347
|
3836 |
/* |
2c488db27
|
3837 |
* new->current_threshold will not be used |
2e72b6347
|
3838 3839 3840 |
* until rcu_assign_pointer(), so it's safe to increment * it here. */ |
2c488db27
|
3841 |
++new->current_threshold; |
2e72b6347
|
3842 3843 3844 |
} j++; } |
907860ed3
|
3845 |
swap_buffers: |
2c488db27
|
3846 3847 3848 |
/* Swap primary and spare array */ thresholds->spare = thresholds->primary; rcu_assign_pointer(thresholds->primary, new); |
2e72b6347
|
3849 |
|
907860ed3
|
3850 |
/* To be sure that nobody uses thresholds */ |
2e72b6347
|
3851 |
synchronize_rcu(); |
2e72b6347
|
3852 |
mutex_unlock(&memcg->thresholds_lock); |
2e72b6347
|
3853 |
} |
c1e862c1f
|
3854 |
|
9490ff275
|
3855 3856 3857 3858 3859 3860 3861 3862 3863 3864 3865 3866 3867 3868 3869 3870 3871 3872 3873 3874 3875 3876 3877 3878 |
static int mem_cgroup_oom_register_event(struct cgroup *cgrp, struct cftype *cft, struct eventfd_ctx *eventfd, const char *args) { struct mem_cgroup *memcg = mem_cgroup_from_cont(cgrp); struct mem_cgroup_eventfd_list *event; int type = MEMFILE_TYPE(cft->private); BUG_ON(type != _OOM_TYPE); event = kmalloc(sizeof(*event), GFP_KERNEL); if (!event) return -ENOMEM; mutex_lock(&memcg_oom_mutex); event->eventfd = eventfd; list_add(&event->list, &memcg->oom_notify); /* already in OOM ? */ if (atomic_read(&memcg->oom_lock)) eventfd_signal(eventfd, 1); mutex_unlock(&memcg_oom_mutex); return 0; } |
907860ed3
|
3879 |
static void mem_cgroup_oom_unregister_event(struct cgroup *cgrp, |
9490ff275
|
3880 3881 3882 3883 3884 3885 3886 3887 3888 3889 3890 3891 3892 3893 3894 3895 3896 3897 |
struct cftype *cft, struct eventfd_ctx *eventfd) { struct mem_cgroup *mem = mem_cgroup_from_cont(cgrp); struct mem_cgroup_eventfd_list *ev, *tmp; int type = MEMFILE_TYPE(cft->private); BUG_ON(type != _OOM_TYPE); mutex_lock(&memcg_oom_mutex); list_for_each_entry_safe(ev, tmp, &mem->oom_notify, list) { if (ev->eventfd == eventfd) { list_del(&ev->list); kfree(ev); } } mutex_unlock(&memcg_oom_mutex); |
9490ff275
|
3898 |
} |
3c11ecf44
|
3899 3900 3901 3902 3903 3904 3905 3906 3907 3908 3909 3910 3911 |
static int mem_cgroup_oom_control_read(struct cgroup *cgrp, struct cftype *cft, struct cgroup_map_cb *cb) { struct mem_cgroup *mem = mem_cgroup_from_cont(cgrp); cb->fill(cb, "oom_kill_disable", mem->oom_kill_disable); if (atomic_read(&mem->oom_lock)) cb->fill(cb, "under_oom", 1); else cb->fill(cb, "under_oom", 0); return 0; } |
3c11ecf44
|
3912 3913 3914 3915 3916 3917 3918 3919 3920 3921 3922 3923 3924 3925 3926 3927 3928 3929 3930 3931 |
static int mem_cgroup_oom_control_write(struct cgroup *cgrp, struct cftype *cft, u64 val) { struct mem_cgroup *mem = mem_cgroup_from_cont(cgrp); struct mem_cgroup *parent; /* cannot set to root cgroup and only 0 and 1 are allowed */ if (!cgrp->parent || !((val == 0) || (val == 1))) return -EINVAL; parent = mem_cgroup_from_cont(cgrp->parent); cgroup_lock(); /* oom-kill-disable is a flag for subhierarchy. */ if ((parent->use_hierarchy) || (mem->use_hierarchy && !list_empty(&cgrp->children))) { cgroup_unlock(); return -EINVAL; } mem->oom_kill_disable = val; |
4d845ebf4
|
3932 3933 |
if (!val) memcg_oom_recover(mem); |
3c11ecf44
|
3934 3935 3936 |
cgroup_unlock(); return 0; } |
8cdea7c05
|
3937 3938 |
static struct cftype mem_cgroup_files[] = { { |
0eea10301
|
3939 |
.name = "usage_in_bytes", |
8c7c6e34a
|
3940 |
.private = MEMFILE_PRIVATE(_MEM, RES_USAGE), |
2c3daa722
|
3941 |
.read_u64 = mem_cgroup_read, |
9490ff275
|
3942 3943 |
.register_event = mem_cgroup_usage_register_event, .unregister_event = mem_cgroup_usage_unregister_event, |
8cdea7c05
|
3944 3945 |
}, { |
c84872e16
|
3946 |
.name = "max_usage_in_bytes", |
8c7c6e34a
|
3947 |
.private = MEMFILE_PRIVATE(_MEM, RES_MAX_USAGE), |
29f2a4dac
|
3948 |
.trigger = mem_cgroup_reset, |
c84872e16
|
3949 3950 3951 |
.read_u64 = mem_cgroup_read, }, { |
0eea10301
|
3952 |
.name = "limit_in_bytes", |
8c7c6e34a
|
3953 |
.private = MEMFILE_PRIVATE(_MEM, RES_LIMIT), |
856c13aa1
|
3954 |
.write_string = mem_cgroup_write, |
2c3daa722
|
3955 |
.read_u64 = mem_cgroup_read, |
8cdea7c05
|
3956 3957 |
}, { |
296c81d89
|
3958 3959 3960 3961 3962 3963 |
.name = "soft_limit_in_bytes", .private = MEMFILE_PRIVATE(_MEM, RES_SOFT_LIMIT), .write_string = mem_cgroup_write, .read_u64 = mem_cgroup_read, }, { |
8cdea7c05
|
3964 |
.name = "failcnt", |
8c7c6e34a
|
3965 |
.private = MEMFILE_PRIVATE(_MEM, RES_FAILCNT), |
29f2a4dac
|
3966 |
.trigger = mem_cgroup_reset, |
2c3daa722
|
3967 |
.read_u64 = mem_cgroup_read, |
8cdea7c05
|
3968 |
}, |
8697d3319
|
3969 |
{ |
d2ceb9b7d
|
3970 |
.name = "stat", |
c64745cf0
|
3971 |
.read_map = mem_control_stat_show, |
d2ceb9b7d
|
3972 |
}, |
c1e862c1f
|
3973 3974 3975 3976 |
{ .name = "force_empty", .trigger = mem_cgroup_force_empty_write, }, |
18f59ea7d
|
3977 3978 3979 3980 3981 |
{ .name = "use_hierarchy", .write_u64 = mem_cgroup_hierarchy_write, .read_u64 = mem_cgroup_hierarchy_read, }, |
a7885eb8a
|
3982 3983 3984 3985 3986 |
{ .name = "swappiness", .read_u64 = mem_cgroup_swappiness_read, .write_u64 = mem_cgroup_swappiness_write, }, |
7dc74be03
|
3987 3988 3989 3990 3991 |
{ .name = "move_charge_at_immigrate", .read_u64 = mem_cgroup_move_charge_read, .write_u64 = mem_cgroup_move_charge_write, }, |
9490ff275
|
3992 3993 |
{ .name = "oom_control", |
3c11ecf44
|
3994 3995 |
.read_map = mem_cgroup_oom_control_read, .write_u64 = mem_cgroup_oom_control_write, |
9490ff275
|
3996 3997 3998 3999 |
.register_event = mem_cgroup_oom_register_event, .unregister_event = mem_cgroup_oom_unregister_event, .private = MEMFILE_PRIVATE(_OOM_TYPE, OOM_CONTROL), }, |
8cdea7c05
|
4000 |
}; |
8c7c6e34a
|
4001 4002 4003 4004 4005 4006 |
#ifdef CONFIG_CGROUP_MEM_RES_CTLR_SWAP static struct cftype memsw_cgroup_files[] = { { .name = "memsw.usage_in_bytes", .private = MEMFILE_PRIVATE(_MEMSWAP, RES_USAGE), .read_u64 = mem_cgroup_read, |
9490ff275
|
4007 4008 |
.register_event = mem_cgroup_usage_register_event, .unregister_event = mem_cgroup_usage_unregister_event, |
8c7c6e34a
|
4009 4010 4011 4012 4013 4014 4015 4016 4017 4018 4019 4020 4021 4022 4023 4024 4025 4026 4027 4028 4029 4030 4031 4032 4033 4034 4035 4036 4037 4038 4039 4040 4041 4042 |
}, { .name = "memsw.max_usage_in_bytes", .private = MEMFILE_PRIVATE(_MEMSWAP, RES_MAX_USAGE), .trigger = mem_cgroup_reset, .read_u64 = mem_cgroup_read, }, { .name = "memsw.limit_in_bytes", .private = MEMFILE_PRIVATE(_MEMSWAP, RES_LIMIT), .write_string = mem_cgroup_write, .read_u64 = mem_cgroup_read, }, { .name = "memsw.failcnt", .private = MEMFILE_PRIVATE(_MEMSWAP, RES_FAILCNT), .trigger = mem_cgroup_reset, .read_u64 = mem_cgroup_read, }, }; static int register_memsw_files(struct cgroup *cont, struct cgroup_subsys *ss) { if (!do_swap_account) return 0; return cgroup_add_files(cont, ss, memsw_cgroup_files, ARRAY_SIZE(memsw_cgroup_files)); }; #else static int register_memsw_files(struct cgroup *cont, struct cgroup_subsys *ss) { return 0; } #endif |
6d12e2d8d
|
4043 4044 4045 |
static int alloc_mem_cgroup_per_zone_info(struct mem_cgroup *mem, int node) { struct mem_cgroup_per_node *pn; |
1ecaab2bd
|
4046 |
struct mem_cgroup_per_zone *mz; |
b69408e88
|
4047 |
enum lru_list l; |
41e3355de
|
4048 |
int zone, tmp = node; |
1ecaab2bd
|
4049 4050 4051 4052 4053 4054 4055 4056 |
/* * This routine is called against possible nodes. * But it's BUG to call kmalloc() against offline node. * * TODO: this routine can waste much memory for nodes which will * never be onlined. It's better to use memory hotplug callback * function. */ |
41e3355de
|
4057 4058 |
if (!node_state(node, N_NORMAL_MEMORY)) tmp = -1; |
17295c88a
|
4059 |
pn = kzalloc_node(sizeof(*pn), GFP_KERNEL, tmp); |
6d12e2d8d
|
4060 4061 |
if (!pn) return 1; |
1ecaab2bd
|
4062 |
|
6d12e2d8d
|
4063 |
mem->info.nodeinfo[node] = pn; |
1ecaab2bd
|
4064 4065 |
for (zone = 0; zone < MAX_NR_ZONES; zone++) { mz = &pn->zoneinfo[zone]; |
b69408e88
|
4066 4067 |
for_each_lru(l) INIT_LIST_HEAD(&mz->lists[l]); |
f64c3f549
|
4068 |
mz->usage_in_excess = 0; |
4e4169535
|
4069 4070 |
mz->on_tree = false; mz->mem = mem; |
1ecaab2bd
|
4071 |
} |
6d12e2d8d
|
4072 4073 |
return 0; } |
1ecaab2bd
|
4074 4075 4076 4077 |
static void free_mem_cgroup_per_zone_info(struct mem_cgroup *mem, int node) { kfree(mem->info.nodeinfo[node]); } |
333279487
|
4078 4079 4080 |
static struct mem_cgroup *mem_cgroup_alloc(void) { struct mem_cgroup *mem; |
c62b1a3b3
|
4081 |
int size = sizeof(struct mem_cgroup); |
333279487
|
4082 |
|
c62b1a3b3
|
4083 |
/* Can be very big if MAX_NUMNODES is very big */ |
c8dad2bb6
|
4084 |
if (size < PAGE_SIZE) |
17295c88a
|
4085 |
mem = kzalloc(size, GFP_KERNEL); |
333279487
|
4086 |
else |
17295c88a
|
4087 |
mem = vzalloc(size); |
333279487
|
4088 |
|
e7bbcdf37
|
4089 4090 |
if (!mem) return NULL; |
c62b1a3b3
|
4091 |
mem->stat = alloc_percpu(struct mem_cgroup_stat_cpu); |
d2e61b8dc
|
4092 4093 |
if (!mem->stat) goto out_free; |
711d3d2c9
|
4094 |
spin_lock_init(&mem->pcp_counter_lock); |
333279487
|
4095 |
return mem; |
d2e61b8dc
|
4096 4097 4098 4099 4100 4101 4102 |
out_free: if (size < PAGE_SIZE) kfree(mem); else vfree(mem); return NULL; |
333279487
|
4103 |
} |
8c7c6e34a
|
4104 4105 4106 4107 4108 4109 4110 4111 |
/* * At destroying mem_cgroup, references from swap_cgroup can remain. * (scanning all at force_empty is too costly...) * * Instead of clearing all references at force_empty, we remember * the number of reference from swap_cgroup and free mem_cgroup when * it goes down to 0. * |
8c7c6e34a
|
4112 4113 |
* Removal of cgroup itself succeeds regardless of refs from swap. */ |
a7ba0eef3
|
4114 |
static void __mem_cgroup_free(struct mem_cgroup *mem) |
333279487
|
4115 |
{ |
08e552c69
|
4116 |
int node; |
f64c3f549
|
4117 |
mem_cgroup_remove_from_trees(mem); |
04046e1a0
|
4118 |
free_css_id(&mem_cgroup_subsys, &mem->css); |
08e552c69
|
4119 4120 |
for_each_node_state(node, N_POSSIBLE) free_mem_cgroup_per_zone_info(mem, node); |
c62b1a3b3
|
4121 4122 |
free_percpu(mem->stat); if (sizeof(struct mem_cgroup) < PAGE_SIZE) |
333279487
|
4123 4124 4125 4126 |
kfree(mem); else vfree(mem); } |
8c7c6e34a
|
4127 4128 4129 4130 |
static void mem_cgroup_get(struct mem_cgroup *mem) { atomic_inc(&mem->refcnt); } |
483c30b51
|
4131 |
static void __mem_cgroup_put(struct mem_cgroup *mem, int count) |
8c7c6e34a
|
4132 |
{ |
483c30b51
|
4133 |
if (atomic_sub_and_test(count, &mem->refcnt)) { |
7bcc1bb12
|
4134 |
struct mem_cgroup *parent = parent_mem_cgroup(mem); |
a7ba0eef3
|
4135 |
__mem_cgroup_free(mem); |
7bcc1bb12
|
4136 4137 4138 |
if (parent) mem_cgroup_put(parent); } |
8c7c6e34a
|
4139 |
} |
483c30b51
|
4140 4141 4142 4143 |
static void mem_cgroup_put(struct mem_cgroup *mem) { __mem_cgroup_put(mem, 1); } |
7bcc1bb12
|
4144 4145 4146 4147 4148 4149 4150 4151 4152 |
/* * Returns the parent mem_cgroup in memcgroup hierarchy with hierarchy enabled. */ static struct mem_cgroup *parent_mem_cgroup(struct mem_cgroup *mem) { if (!mem->res.parent) return NULL; return mem_cgroup_from_res_counter(mem->res.parent, res); } |
333279487
|
4153 |
|
c077719be
|
4154 4155 4156 |
#ifdef CONFIG_CGROUP_MEM_RES_CTLR_SWAP static void __init enable_swap_cgroup(void) { |
f8d665422
|
4157 |
if (!mem_cgroup_disabled() && really_do_swap_account) |
c077719be
|
4158 4159 4160 4161 4162 4163 4164 |
do_swap_account = 1; } #else static void __init enable_swap_cgroup(void) { } #endif |
f64c3f549
|
4165 4166 4167 4168 4169 4170 4171 4172 4173 4174 4175 4176 4177 4178 4179 4180 4181 4182 4183 4184 4185 4186 4187 4188 |
static int mem_cgroup_soft_limit_tree_init(void) { struct mem_cgroup_tree_per_node *rtpn; struct mem_cgroup_tree_per_zone *rtpz; int tmp, node, zone; for_each_node_state(node, N_POSSIBLE) { tmp = node; if (!node_state(node, N_NORMAL_MEMORY)) tmp = -1; rtpn = kzalloc_node(sizeof(*rtpn), GFP_KERNEL, tmp); if (!rtpn) return 1; soft_limit_tree.rb_tree_per_node[node] = rtpn; for (zone = 0; zone < MAX_NR_ZONES; zone++) { rtpz = &rtpn->rb_tree_per_zone[zone]; rtpz->rb_root = RB_ROOT; spin_lock_init(&rtpz->lock); } } return 0; } |
0eb253e22
|
4189 |
static struct cgroup_subsys_state * __ref |
8cdea7c05
|
4190 4191 |
mem_cgroup_create(struct cgroup_subsys *ss, struct cgroup *cont) { |
28dbc4b6a
|
4192 |
struct mem_cgroup *mem, *parent; |
04046e1a0
|
4193 |
long error = -ENOMEM; |
6d12e2d8d
|
4194 |
int node; |
8cdea7c05
|
4195 |
|
c8dad2bb6
|
4196 4197 |
mem = mem_cgroup_alloc(); if (!mem) |
04046e1a0
|
4198 |
return ERR_PTR(error); |
78fb74669
|
4199 |
|
6d12e2d8d
|
4200 4201 4202 |
for_each_node_state(node, N_POSSIBLE) if (alloc_mem_cgroup_per_zone_info(mem, node)) goto free_out; |
f64c3f549
|
4203 |
|
c077719be
|
4204 |
/* root ? */ |
28dbc4b6a
|
4205 |
if (cont->parent == NULL) { |
cdec2e426
|
4206 |
int cpu; |
c077719be
|
4207 |
enable_swap_cgroup(); |
28dbc4b6a
|
4208 |
parent = NULL; |
4b3bde4c9
|
4209 |
root_mem_cgroup = mem; |
f64c3f549
|
4210 4211 |
if (mem_cgroup_soft_limit_tree_init()) goto free_out; |
cdec2e426
|
4212 4213 4214 4215 4216 |
for_each_possible_cpu(cpu) { struct memcg_stock_pcp *stock = &per_cpu(memcg_stock, cpu); INIT_WORK(&stock->work, drain_local_stock); } |
711d3d2c9
|
4217 |
hotcpu_notifier(memcg_cpu_hotplug_callback, 0); |
18f59ea7d
|
4218 |
} else { |
28dbc4b6a
|
4219 |
parent = mem_cgroup_from_cont(cont->parent); |
18f59ea7d
|
4220 |
mem->use_hierarchy = parent->use_hierarchy; |
3c11ecf44
|
4221 |
mem->oom_kill_disable = parent->oom_kill_disable; |
18f59ea7d
|
4222 |
} |
28dbc4b6a
|
4223 |
|
18f59ea7d
|
4224 4225 4226 |
if (parent && parent->use_hierarchy) { res_counter_init(&mem->res, &parent->res); res_counter_init(&mem->memsw, &parent->memsw); |
7bcc1bb12
|
4227 4228 4229 4230 4231 4232 4233 |
/* * We increment refcnt of the parent to ensure that we can * safely access it on res_counter_charge/uncharge. * This refcnt will be decremented when freeing this * mem_cgroup(see mem_cgroup_put). */ mem_cgroup_get(parent); |
18f59ea7d
|
4234 4235 4236 4237 |
} else { res_counter_init(&mem->res, NULL); res_counter_init(&mem->memsw, NULL); } |
04046e1a0
|
4238 |
mem->last_scanned_child = 0; |
2733c06ac
|
4239 |
spin_lock_init(&mem->reclaim_param_lock); |
9490ff275
|
4240 |
INIT_LIST_HEAD(&mem->oom_notify); |
6d61ef409
|
4241 |
|
a7885eb8a
|
4242 4243 |
if (parent) mem->swappiness = get_swappiness(parent); |
a7ba0eef3
|
4244 |
atomic_set(&mem->refcnt, 1); |
7dc74be03
|
4245 |
mem->move_charge_at_immigrate = 0; |
2e72b6347
|
4246 |
mutex_init(&mem->thresholds_lock); |
8cdea7c05
|
4247 |
return &mem->css; |
6d12e2d8d
|
4248 |
free_out: |
a7ba0eef3
|
4249 |
__mem_cgroup_free(mem); |
4b3bde4c9
|
4250 |
root_mem_cgroup = NULL; |
04046e1a0
|
4251 |
return ERR_PTR(error); |
8cdea7c05
|
4252 |
} |
ec64f5154
|
4253 |
static int mem_cgroup_pre_destroy(struct cgroup_subsys *ss, |
df878fb04
|
4254 4255 4256 |
struct cgroup *cont) { struct mem_cgroup *mem = mem_cgroup_from_cont(cont); |
ec64f5154
|
4257 4258 |
return mem_cgroup_force_empty(mem, false); |
df878fb04
|
4259 |
} |
8cdea7c05
|
4260 4261 4262 |
static void mem_cgroup_destroy(struct cgroup_subsys *ss, struct cgroup *cont) { |
c268e9946
|
4263 |
struct mem_cgroup *mem = mem_cgroup_from_cont(cont); |
c268e9946
|
4264 |
|
c268e9946
|
4265 |
mem_cgroup_put(mem); |
8cdea7c05
|
4266 4267 4268 4269 4270 |
} static int mem_cgroup_populate(struct cgroup_subsys *ss, struct cgroup *cont) { |
8c7c6e34a
|
4271 4272 4273 4274 4275 4276 4277 4278 |
int ret; ret = cgroup_add_files(cont, ss, mem_cgroup_files, ARRAY_SIZE(mem_cgroup_files)); if (!ret) ret = register_memsw_files(cont, ss); return ret; |
8cdea7c05
|
4279 |
} |
024914477
|
4280 |
#ifdef CONFIG_MMU |
7dc74be03
|
4281 |
/* Handlers for move charge at task migration. */ |
854ffa8d1
|
4282 4283 |
#define PRECHARGE_COUNT_AT_ONCE 256 static int mem_cgroup_do_precharge(unsigned long count) |
7dc74be03
|
4284 |
{ |
854ffa8d1
|
4285 4286 |
int ret = 0; int batch_count = PRECHARGE_COUNT_AT_ONCE; |
4ffef5fef
|
4287 |
struct mem_cgroup *mem = mc.to; |
854ffa8d1
|
4288 4289 4290 4291 4292 4293 4294 4295 4296 4297 4298 4299 4300 4301 4302 4303 4304 4305 4306 4307 4308 4309 |
if (mem_cgroup_is_root(mem)) { mc.precharge += count; /* we don't need css_get for root */ return ret; } /* try to charge at once */ if (count > 1) { struct res_counter *dummy; /* * "mem" cannot be under rmdir() because we've already checked * by cgroup_lock_live_cgroup() that it is not removed and we * are still under the same cgroup_mutex. So we can postpone * css_get(). */ if (res_counter_charge(&mem->res, PAGE_SIZE * count, &dummy)) goto one_by_one; if (do_swap_account && res_counter_charge(&mem->memsw, PAGE_SIZE * count, &dummy)) { res_counter_uncharge(&mem->res, PAGE_SIZE * count); goto one_by_one; } mc.precharge += count; |
854ffa8d1
|
4310 4311 4312 4313 4314 4315 4316 4317 4318 4319 4320 4321 4322 |
return ret; } one_by_one: /* fall back to one by one charge */ while (count--) { if (signal_pending(current)) { ret = -EINTR; break; } if (!batch_count--) { batch_count = PRECHARGE_COUNT_AT_ONCE; cond_resched(); } |
ec1685109
|
4323 4324 |
ret = __mem_cgroup_try_charge(NULL, GFP_KERNEL, &mem, false, PAGE_SIZE); |
854ffa8d1
|
4325 4326 4327 4328 4329 |
if (ret || !mem) /* mem_cgroup_clear_mc() will do uncharge later */ return -ENOMEM; mc.precharge++; } |
4ffef5fef
|
4330 4331 4332 4333 4334 4335 4336 4337 |
return ret; } /** * is_target_pte_for_mc - check a pte whether it is valid for move charge * @vma: the vma the pte to be checked belongs * @addr: the address corresponding to the pte to be checked * @ptent: the pte to be checked |
024914477
|
4338 |
* @target: the pointer the target page or swap ent will be stored(can be NULL) |
4ffef5fef
|
4339 4340 4341 4342 4343 4344 |
* * Returns * 0(MC_TARGET_NONE): if the pte is not a target for move charge. * 1(MC_TARGET_PAGE): if the page corresponding to this pte is a target for * move charge. if @target is not NULL, the page is stored in target->page * with extra refcnt got(Callers should handle it). |
024914477
|
4345 4346 4347 |
* 2(MC_TARGET_SWAP): if the swap entry corresponding to this pte is a * target for charge migration. if @target is not NULL, the entry is stored * in target->ent. |
4ffef5fef
|
4348 4349 4350 |
* * Called with pte lock held. */ |
4ffef5fef
|
4351 4352 |
union mc_target { struct page *page; |
024914477
|
4353 |
swp_entry_t ent; |
4ffef5fef
|
4354 |
}; |
4ffef5fef
|
4355 4356 4357 |
enum mc_target_type { MC_TARGET_NONE, /* not used */ MC_TARGET_PAGE, |
024914477
|
4358 |
MC_TARGET_SWAP, |
4ffef5fef
|
4359 |
}; |
90254a658
|
4360 4361 |
static struct page *mc_handle_present_pte(struct vm_area_struct *vma, unsigned long addr, pte_t ptent) |
4ffef5fef
|
4362 |
{ |
90254a658
|
4363 |
struct page *page = vm_normal_page(vma, addr, ptent); |
4ffef5fef
|
4364 |
|
90254a658
|
4365 4366 4367 4368 4369 4370 |
if (!page || !page_mapped(page)) return NULL; if (PageAnon(page)) { /* we don't move shared anon */ if (!move_anon() || page_mapcount(page) > 2) return NULL; |
87946a722
|
4371 4372 |
} else if (!move_file()) /* we ignore mapcount for file pages */ |
90254a658
|
4373 4374 4375 4376 4377 4378 4379 4380 4381 4382 4383 4384 4385 4386 4387 4388 4389 4390 |
return NULL; if (!get_page_unless_zero(page)) return NULL; return page; } static struct page *mc_handle_swap_pte(struct vm_area_struct *vma, unsigned long addr, pte_t ptent, swp_entry_t *entry) { int usage_count; struct page *page = NULL; swp_entry_t ent = pte_to_swp_entry(ptent); if (!move_anon() || non_swap_entry(ent)) return NULL; usage_count = mem_cgroup_count_swap_user(ent, &page); if (usage_count > 1) { /* we don't move shared anon */ |
024914477
|
4391 4392 |
if (page) put_page(page); |
90254a658
|
4393 |
return NULL; |
024914477
|
4394 |
} |
90254a658
|
4395 4396 4397 4398 4399 |
if (do_swap_account) entry->val = ent.val; return page; } |
87946a722
|
4400 4401 4402 4403 4404 4405 4406 4407 4408 4409 4410 4411 4412 4413 4414 4415 4416 4417 4418 4419 4420 4421 4422 4423 4424 4425 4426 4427 4428 4429 4430 4431 |
static struct page *mc_handle_file_pte(struct vm_area_struct *vma, unsigned long addr, pte_t ptent, swp_entry_t *entry) { struct page *page = NULL; struct inode *inode; struct address_space *mapping; pgoff_t pgoff; if (!vma->vm_file) /* anonymous vma */ return NULL; if (!move_file()) return NULL; inode = vma->vm_file->f_path.dentry->d_inode; mapping = vma->vm_file->f_mapping; if (pte_none(ptent)) pgoff = linear_page_index(vma, addr); else /* pte_file(ptent) is true */ pgoff = pte_to_pgoff(ptent); /* page is moved even if it's not RSS of this task(page-faulted). */ if (!mapping_cap_swap_backed(mapping)) { /* normal file */ page = find_get_page(mapping, pgoff); } else { /* shmem/tmpfs file. we should take account of swap too. */ swp_entry_t ent; mem_cgroup_get_shmem_target(inode, pgoff, &page, &ent); if (do_swap_account) entry->val = ent.val; } return page; } |
90254a658
|
4432 4433 4434 4435 4436 4437 4438 4439 4440 4441 4442 4443 |
static int is_target_pte_for_mc(struct vm_area_struct *vma, unsigned long addr, pte_t ptent, union mc_target *target) { struct page *page = NULL; struct page_cgroup *pc; int ret = 0; swp_entry_t ent = { .val = 0 }; if (pte_present(ptent)) page = mc_handle_present_pte(vma, addr, ptent); else if (is_swap_pte(ptent)) page = mc_handle_swap_pte(vma, addr, ptent, &ent); |
87946a722
|
4444 4445 |
else if (pte_none(ptent) || pte_file(ptent)) page = mc_handle_file_pte(vma, addr, ptent, &ent); |
90254a658
|
4446 4447 4448 |
if (!page && !ent.val) return 0; |
024914477
|
4449 4450 4451 4452 4453 4454 4455 4456 4457 4458 4459 4460 4461 4462 4463 |
if (page) { pc = lookup_page_cgroup(page); /* * Do only loose check w/o page_cgroup lock. * mem_cgroup_move_account() checks the pc is valid or not under * the lock. */ if (PageCgroupUsed(pc) && pc->mem_cgroup == mc.from) { ret = MC_TARGET_PAGE; if (target) target->page = page; } if (!ret || !target) put_page(page); } |
90254a658
|
4464 4465 |
/* There is a swap entry and a page doesn't exist or isn't charged */ if (ent.val && !ret && |
7f0f15464
|
4466 4467 4468 4469 |
css_id(&mc.from->css) == lookup_swap_cgroup(ent)) { ret = MC_TARGET_SWAP; if (target) target->ent = ent; |
4ffef5fef
|
4470 |
} |
4ffef5fef
|
4471 4472 4473 4474 4475 4476 4477 4478 4479 4480 |
return ret; } static int mem_cgroup_count_precharge_pte_range(pmd_t *pmd, unsigned long addr, unsigned long end, struct mm_walk *walk) { struct vm_area_struct *vma = walk->private; pte_t *pte; spinlock_t *ptl; |
ec1685109
|
4481 |
VM_BUG_ON(pmd_trans_huge(*pmd)); |
4ffef5fef
|
4482 4483 4484 4485 4486 4487 |
pte = pte_offset_map_lock(vma->vm_mm, pmd, addr, &ptl); for (; addr != end; pte++, addr += PAGE_SIZE) if (is_target_pte_for_mc(vma, addr, *pte, NULL)) mc.precharge++; /* increment precharge temporarily */ pte_unmap_unlock(pte - 1, ptl); cond_resched(); |
7dc74be03
|
4488 4489 |
return 0; } |
4ffef5fef
|
4490 4491 4492 4493 |
static unsigned long mem_cgroup_count_precharge(struct mm_struct *mm) { unsigned long precharge; struct vm_area_struct *vma; |
dfe076b09
|
4494 |
down_read(&mm->mmap_sem); |
4ffef5fef
|
4495 4496 4497 4498 4499 4500 4501 4502 |
for (vma = mm->mmap; vma; vma = vma->vm_next) { struct mm_walk mem_cgroup_count_precharge_walk = { .pmd_entry = mem_cgroup_count_precharge_pte_range, .mm = mm, .private = vma, }; if (is_vm_hugetlb_page(vma)) continue; |
4ffef5fef
|
4503 4504 4505 |
walk_page_range(vma->vm_start, vma->vm_end, &mem_cgroup_count_precharge_walk); } |
dfe076b09
|
4506 |
up_read(&mm->mmap_sem); |
4ffef5fef
|
4507 4508 4509 4510 4511 4512 |
precharge = mc.precharge; mc.precharge = 0; return precharge; } |
4ffef5fef
|
4513 4514 |
static int mem_cgroup_precharge_mc(struct mm_struct *mm) { |
dfe076b09
|
4515 4516 4517 4518 4519 |
unsigned long precharge = mem_cgroup_count_precharge(mm); VM_BUG_ON(mc.moving_task); mc.moving_task = current; return mem_cgroup_do_precharge(precharge); |
4ffef5fef
|
4520 |
} |
dfe076b09
|
4521 4522 |
/* cancels all extra charges on mc.from and mc.to, and wakes up all waiters. */ static void __mem_cgroup_clear_mc(void) |
4ffef5fef
|
4523 |
{ |
2bd9bb206
|
4524 4525 |
struct mem_cgroup *from = mc.from; struct mem_cgroup *to = mc.to; |
4ffef5fef
|
4526 |
/* we must uncharge all the leftover precharges from mc.to */ |
854ffa8d1
|
4527 4528 4529 4530 4531 4532 4533 4534 4535 4536 4537 |
if (mc.precharge) { __mem_cgroup_cancel_charge(mc.to, mc.precharge); mc.precharge = 0; } /* * we didn't uncharge from mc.from at mem_cgroup_move_account(), so * we must uncharge here. */ if (mc.moved_charge) { __mem_cgroup_cancel_charge(mc.from, mc.moved_charge); mc.moved_charge = 0; |
4ffef5fef
|
4538 |
} |
483c30b51
|
4539 4540 |
/* we must fixup refcnts and charges */ if (mc.moved_swap) { |
483c30b51
|
4541 4542 4543 4544 4545 4546 4547 4548 4549 4550 4551 4552 4553 |
/* uncharge swap account from the old cgroup */ if (!mem_cgroup_is_root(mc.from)) res_counter_uncharge(&mc.from->memsw, PAGE_SIZE * mc.moved_swap); __mem_cgroup_put(mc.from, mc.moved_swap); if (!mem_cgroup_is_root(mc.to)) { /* * we charged both to->res and to->memsw, so we should * uncharge to->res. */ res_counter_uncharge(&mc.to->res, PAGE_SIZE * mc.moved_swap); |
483c30b51
|
4554 4555 |
} /* we've already done mem_cgroup_get(mc.to) */ |
483c30b51
|
4556 4557 |
mc.moved_swap = 0; } |
dfe076b09
|
4558 4559 4560 4561 4562 4563 4564 4565 4566 4567 4568 4569 4570 4571 4572 |
memcg_oom_recover(from); memcg_oom_recover(to); wake_up_all(&mc.waitq); } static void mem_cgroup_clear_mc(void) { struct mem_cgroup *from = mc.from; /* * we must clear moving_task before waking up waiters at the end of * task migration. */ mc.moving_task = NULL; __mem_cgroup_clear_mc(); |
2bd9bb206
|
4573 |
spin_lock(&mc.lock); |
4ffef5fef
|
4574 4575 |
mc.from = NULL; mc.to = NULL; |
2bd9bb206
|
4576 |
spin_unlock(&mc.lock); |
32047e2a8
|
4577 |
mem_cgroup_end_move(from); |
4ffef5fef
|
4578 |
} |
7dc74be03
|
4579 4580 4581 4582 4583 4584 4585 4586 4587 4588 4589 4590 4591 4592 4593 4594 4595 |
static int mem_cgroup_can_attach(struct cgroup_subsys *ss, struct cgroup *cgroup, struct task_struct *p, bool threadgroup) { int ret = 0; struct mem_cgroup *mem = mem_cgroup_from_cont(cgroup); if (mem->move_charge_at_immigrate) { struct mm_struct *mm; struct mem_cgroup *from = mem_cgroup_from_task(p); VM_BUG_ON(from == mem); mm = get_task_mm(p); if (!mm) return 0; |
7dc74be03
|
4596 |
/* We move charges only when we move a owner of the mm */ |
4ffef5fef
|
4597 4598 4599 4600 |
if (mm->owner == p) { VM_BUG_ON(mc.from); VM_BUG_ON(mc.to); VM_BUG_ON(mc.precharge); |
854ffa8d1
|
4601 |
VM_BUG_ON(mc.moved_charge); |
483c30b51
|
4602 |
VM_BUG_ON(mc.moved_swap); |
32047e2a8
|
4603 |
mem_cgroup_start_move(from); |
2bd9bb206
|
4604 |
spin_lock(&mc.lock); |
4ffef5fef
|
4605 4606 |
mc.from = from; mc.to = mem; |
2bd9bb206
|
4607 |
spin_unlock(&mc.lock); |
dfe076b09
|
4608 |
/* We set mc.moving_task later */ |
4ffef5fef
|
4609 4610 4611 4612 |
ret = mem_cgroup_precharge_mc(mm); if (ret) mem_cgroup_clear_mc(); |
dfe076b09
|
4613 4614 |
} mmput(mm); |
7dc74be03
|
4615 4616 4617 4618 4619 4620 4621 4622 4623 |
} return ret; } static void mem_cgroup_cancel_attach(struct cgroup_subsys *ss, struct cgroup *cgroup, struct task_struct *p, bool threadgroup) { |
4ffef5fef
|
4624 |
mem_cgroup_clear_mc(); |
7dc74be03
|
4625 |
} |
4ffef5fef
|
4626 4627 4628 |
static int mem_cgroup_move_charge_pte_range(pmd_t *pmd, unsigned long addr, unsigned long end, struct mm_walk *walk) |
7dc74be03
|
4629 |
{ |
4ffef5fef
|
4630 4631 4632 4633 4634 4635 |
int ret = 0; struct vm_area_struct *vma = walk->private; pte_t *pte; spinlock_t *ptl; retry: |
ec1685109
|
4636 |
VM_BUG_ON(pmd_trans_huge(*pmd)); |
4ffef5fef
|
4637 4638 4639 4640 4641 4642 4643 |
pte = pte_offset_map_lock(vma->vm_mm, pmd, addr, &ptl); for (; addr != end; addr += PAGE_SIZE) { pte_t ptent = *(pte++); union mc_target target; int type; struct page *page; struct page_cgroup *pc; |
024914477
|
4644 |
swp_entry_t ent; |
4ffef5fef
|
4645 4646 4647 4648 4649 4650 4651 4652 4653 4654 4655 |
if (!mc.precharge) break; type = is_target_pte_for_mc(vma, addr, ptent, &target); switch (type) { case MC_TARGET_PAGE: page = target.page; if (isolate_lru_page(page)) goto put; pc = lookup_page_cgroup(page); |
854ffa8d1
|
4656 4657 |
if (!mem_cgroup_move_account(pc, mc.from, mc.to, false)) { |
4ffef5fef
|
4658 |
mc.precharge--; |
854ffa8d1
|
4659 4660 |
/* we uncharge from mc.from later. */ mc.moved_charge++; |
4ffef5fef
|
4661 4662 4663 4664 4665 |
} putback_lru_page(page); put: /* is_target_pte_for_mc() gets the page */ put_page(page); break; |
024914477
|
4666 4667 |
case MC_TARGET_SWAP: ent = target.ent; |
483c30b51
|
4668 4669 |
if (!mem_cgroup_move_swap_account(ent, mc.from, mc.to, false)) { |
024914477
|
4670 |
mc.precharge--; |
483c30b51
|
4671 4672 4673 |
/* we fixup refcnts and charges later. */ mc.moved_swap++; } |
024914477
|
4674 |
break; |
4ffef5fef
|
4675 4676 4677 4678 4679 4680 4681 4682 4683 4684 4685 4686 4687 4688 |
default: break; } } pte_unmap_unlock(pte - 1, ptl); cond_resched(); if (addr != end) { /* * We have consumed all precharges we got in can_attach(). * We try charge one by one, but don't do any additional * charges to mc.to if we have failed in charge once in attach() * phase. */ |
854ffa8d1
|
4689 |
ret = mem_cgroup_do_precharge(1); |
4ffef5fef
|
4690 4691 4692 4693 4694 4695 4696 4697 4698 4699 4700 4701 |
if (!ret) goto retry; } return ret; } static void mem_cgroup_move_charge(struct mm_struct *mm) { struct vm_area_struct *vma; lru_add_drain_all(); |
dfe076b09
|
4702 4703 4704 4705 4706 4707 4708 4709 4710 4711 4712 4713 4714 |
retry: if (unlikely(!down_read_trylock(&mm->mmap_sem))) { /* * Someone who are holding the mmap_sem might be waiting in * waitq. So we cancel all extra charges, wake up all waiters, * and retry. Because we cancel precharges, we might not be able * to move enough charges, but moving charge is a best-effort * feature anyway, so it wouldn't be a big problem. */ __mem_cgroup_clear_mc(); cond_resched(); goto retry; } |
4ffef5fef
|
4715 4716 4717 4718 4719 4720 4721 4722 4723 |
for (vma = mm->mmap; vma; vma = vma->vm_next) { int ret; struct mm_walk mem_cgroup_move_charge_walk = { .pmd_entry = mem_cgroup_move_charge_pte_range, .mm = mm, .private = vma, }; if (is_vm_hugetlb_page(vma)) continue; |
4ffef5fef
|
4724 4725 4726 4727 4728 4729 4730 4731 4732 |
ret = walk_page_range(vma->vm_start, vma->vm_end, &mem_cgroup_move_charge_walk); if (ret) /* * means we have consumed all precharges and failed in * doing additional charge. Just abandon here. */ break; } |
dfe076b09
|
4733 |
up_read(&mm->mmap_sem); |
7dc74be03
|
4734 |
} |
67e465a77
|
4735 4736 4737 |
static void mem_cgroup_move_task(struct cgroup_subsys *ss, struct cgroup *cont, struct cgroup *old_cont, |
be367d099
|
4738 4739 |
struct task_struct *p, bool threadgroup) |
67e465a77
|
4740 |
{ |
dfe076b09
|
4741 4742 4743 |
struct mm_struct *mm; if (!mc.to) |
4ffef5fef
|
4744 4745 |
/* no need to move charge */ return; |
dfe076b09
|
4746 4747 4748 4749 4750 |
mm = get_task_mm(p); if (mm) { mem_cgroup_move_charge(mm); mmput(mm); } |
4ffef5fef
|
4751 |
mem_cgroup_clear_mc(); |
67e465a77
|
4752 |
} |
5cfb80a73
|
4753 4754 4755 4756 4757 4758 4759 4760 4761 4762 4763 4764 4765 4766 4767 4768 4769 4770 4771 4772 4773 4774 |
#else /* !CONFIG_MMU */ static int mem_cgroup_can_attach(struct cgroup_subsys *ss, struct cgroup *cgroup, struct task_struct *p, bool threadgroup) { return 0; } static void mem_cgroup_cancel_attach(struct cgroup_subsys *ss, struct cgroup *cgroup, struct task_struct *p, bool threadgroup) { } static void mem_cgroup_move_task(struct cgroup_subsys *ss, struct cgroup *cont, struct cgroup *old_cont, struct task_struct *p, bool threadgroup) { } #endif |
67e465a77
|
4775 |
|
8cdea7c05
|
4776 4777 4778 4779 |
struct cgroup_subsys mem_cgroup_subsys = { .name = "memory", .subsys_id = mem_cgroup_subsys_id, .create = mem_cgroup_create, |
df878fb04
|
4780 |
.pre_destroy = mem_cgroup_pre_destroy, |
8cdea7c05
|
4781 4782 |
.destroy = mem_cgroup_destroy, .populate = mem_cgroup_populate, |
7dc74be03
|
4783 4784 |
.can_attach = mem_cgroup_can_attach, .cancel_attach = mem_cgroup_cancel_attach, |
67e465a77
|
4785 |
.attach = mem_cgroup_move_task, |
6d12e2d8d
|
4786 |
.early_init = 0, |
04046e1a0
|
4787 |
.use_id = 1, |
8cdea7c05
|
4788 |
}; |
c077719be
|
4789 4790 |
#ifdef CONFIG_CGROUP_MEM_RES_CTLR_SWAP |
a42c390cf
|
4791 4792 4793 4794 4795 4796 4797 4798 4799 4800 |
static int __init enable_swap_account(char *s) { /* consider enabled if no parameter or 1 is given */ if (!s || !strcmp(s, "1")) really_do_swap_account = 1; else if (!strcmp(s, "0")) really_do_swap_account = 0; return 1; } __setup("swapaccount", enable_swap_account); |
c077719be
|
4801 4802 4803 |
static int __init disable_swap_account(char *s) { |
a42c390cf
|
4804 |
enable_swap_account("0"); |
c077719be
|
4805 4806 4807 4808 |
return 1; } __setup("noswapaccount", disable_swap_account); #endif |