Blame view
mm/memcontrol.c
120 KB
8cdea7c05 Memory controller... |
1 2 3 4 5 |
/* memcontrol.c - Memory Controller * * Copyright IBM Corporation, 2007 * Author Balbir Singh <balbir@linux.vnet.ibm.com> * |
78fb74669 Memory controller... |
6 7 8 |
* Copyright 2007 OpenVZ SWsoft Inc * Author: Pavel Emelianov <xemul@openvz.org> * |
2e72b6347 memcg: implement ... |
9 10 11 12 |
* Memory thresholds * Copyright (C) 2009 Nokia Corporation * Author: Kirill A. Shutemov * |
8cdea7c05 Memory controller... |
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 Memory controller... |
27 |
#include <linux/mm.h> |
4ffef5fef memcg: move charg... |
28 |
#include <linux/hugetlb.h> |
d13d14430 memcg: handle swa... |
29 |
#include <linux/pagemap.h> |
d52aa412d memory cgroup enh... |
30 |
#include <linux/smp.h> |
8a9f3ccd2 Memory controller... |
31 |
#include <linux/page-flags.h> |
66e1707bc Memory controller... |
32 |
#include <linux/backing-dev.h> |
8a9f3ccd2 Memory controller... |
33 34 |
#include <linux/bit_spinlock.h> #include <linux/rcupdate.h> |
e222432bf memcg: show memcg... |
35 |
#include <linux/limits.h> |
8c7c6e34a memcg: mem+swap c... |
36 |
#include <linux/mutex.h> |
f64c3f549 memory controller... |
37 |
#include <linux/rbtree.h> |
b6ac57d50 memcgroup: move m... |
38 |
#include <linux/slab.h> |
66e1707bc Memory controller... |
39 |
#include <linux/swap.h> |
024914477 memcg: move charg... |
40 |
#include <linux/swapops.h> |
66e1707bc Memory controller... |
41 |
#include <linux/spinlock.h> |
2e72b6347 memcg: implement ... |
42 43 |
#include <linux/eventfd.h> #include <linux/sort.h> |
66e1707bc Memory controller... |
44 |
#include <linux/fs.h> |
d2ceb9b7d memory cgroup enh... |
45 |
#include <linux/seq_file.h> |
333279487 memcgroup: use vm... |
46 |
#include <linux/vmalloc.h> |
b69408e88 vmscan: Use an in... |
47 |
#include <linux/mm_inline.h> |
52d4b9ac0 memcg: allocate a... |
48 |
#include <linux/page_cgroup.h> |
cdec2e426 memcg: coalesce c... |
49 |
#include <linux/cpu.h> |
158e0a2d1 memcg: use find_l... |
50 |
#include <linux/oom.h> |
08e552c69 memcg: synchroniz... |
51 |
#include "internal.h" |
8cdea7c05 Memory controller... |
52 |
|
8697d3319 Memory controller... |
53 |
#include <asm/uaccess.h> |
cc8e970c3 memcg: add mm_vms... |
54 |
#include <trace/events/vmscan.h> |
a181b0e88 memcg: make globa... |
55 |
struct cgroup_subsys mem_cgroup_subsys __read_mostly; |
a181b0e88 memcg: make globa... |
56 |
#define MEM_CGROUP_RECLAIM_RETRIES 5 |
4b3bde4c9 memcg: remove the... |
57 |
struct mem_cgroup *root_mem_cgroup __read_mostly; |
8cdea7c05 Memory controller... |
58 |
|
c077719be memcg: mem+swap c... |
59 |
#ifdef CONFIG_CGROUP_MEM_RES_CTLR_SWAP |
338c84310 memcg: remove som... |
60 |
/* Turned on only when memory cgroup is enabled && really_do_swap_account = 1 */ |
c077719be memcg: mem+swap c... |
61 62 63 64 65 |
int do_swap_account __read_mostly; static int really_do_swap_account __initdata = 1; /* for remember boot option*/ #else #define do_swap_account (0) #endif |
d2265e6fa memcg : share eve... |
66 67 68 69 70 71 72 73 74 |
/* * 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 memcg: mem+swap c... |
75 |
|
8cdea7c05 Memory controller... |
76 |
/* |
d52aa412d memory cgroup enh... |
77 78 79 80 81 82 83 |
* 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 memcg: add file-b... |
84 |
MEM_CGROUP_STAT_RSS, /* # of pages charged as anon rss */ |
d8046582d memcg: make memcg... |
85 |
MEM_CGROUP_STAT_FILE_MAPPED, /* # of pages charged as file rss */ |
55e462b05 memcg: simple sta... |
86 87 |
MEM_CGROUP_STAT_PGPGIN_COUNT, /* # of pages paged in */ MEM_CGROUP_STAT_PGPGOUT_COUNT, /* # of pages paged out */ |
0c3e73e84 memcg: improve re... |
88 |
MEM_CGROUP_STAT_SWAPOUT, /* # of pages, swapped out */ |
d2265e6fa memcg : share eve... |
89 |
MEM_CGROUP_EVENTS, /* incremented at every pagein/pageout */ |
d52aa412d memory cgroup enh... |
90 91 92 93 94 95 |
MEM_CGROUP_STAT_NSTATS, }; struct mem_cgroup_stat_cpu { s64 count[MEM_CGROUP_STAT_NSTATS]; |
d52aa412d memory cgroup enh... |
96 |
}; |
d52aa412d memory cgroup enh... |
97 |
/* |
6d12e2d8d per-zone and recl... |
98 99 |
* per-zone information in memory controller. */ |
6d12e2d8d per-zone and recl... |
100 |
struct mem_cgroup_per_zone { |
072c56c13 per-zone and recl... |
101 102 103 |
/* * spin_lock to protect the per cgroup LRU */ |
b69408e88 vmscan: Use an in... |
104 105 |
struct list_head lists[NR_LRU_LISTS]; unsigned long count[NR_LRU_LISTS]; |
3e2f41f1f memcg: add zone_r... |
106 107 |
struct zone_reclaim_stat reclaim_stat; |
f64c3f549 memory controller... |
108 109 110 111 |
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 memory controller... |
112 113 |
struct mem_cgroup *mem; /* Back pointer, we cannot */ /* use container_of */ |
6d12e2d8d per-zone and recl... |
114 115 116 117 118 119 120 121 122 123 124 125 126 |
}; /* 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 memory controller... |
127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 |
* 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 memcg: implement ... |
145 146 147 148 |
struct mem_cgroup_threshold { struct eventfd_ctx *eventfd; u64 threshold; }; |
9490ff275 memcg: oom notifier |
149 |
/* For threshold */ |
2e72b6347 memcg: implement ... |
150 151 |
struct mem_cgroup_threshold_ary { /* An array index points to threshold just below usage. */ |
5407a5625 mm: remove unnece... |
152 |
int current_threshold; |
2e72b6347 memcg: implement ... |
153 154 155 156 157 |
/* Size of entries[] */ unsigned int size; /* Array of thresholds */ struct mem_cgroup_threshold entries[0]; }; |
2c488db27 memcg: clean up m... |
158 159 160 161 162 163 164 165 166 167 168 |
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 memcg: oom notifier |
169 170 171 172 173 |
/* for OOM */ struct mem_cgroup_eventfd_list { struct list_head list; struct eventfd_ctx *eventfd; }; |
2e72b6347 memcg: implement ... |
174 |
|
2e72b6347 memcg: implement ... |
175 |
static void mem_cgroup_threshold(struct mem_cgroup *mem); |
9490ff275 memcg: oom notifier |
176 |
static void mem_cgroup_oom_notify(struct mem_cgroup *mem); |
2e72b6347 memcg: implement ... |
177 |
|
f64c3f549 memory controller... |
178 |
/* |
8cdea7c05 Memory controller... |
179 180 181 182 183 184 |
* 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 Memory controller... |
185 186 187 |
* 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 Memory controller... |
188 189 190 191 192 193 194 |
*/ struct mem_cgroup { struct cgroup_subsys_state css; /* * the counter to account for memory usage */ struct res_counter res; |
78fb74669 Memory controller... |
195 |
/* |
8c7c6e34a memcg: mem+swap c... |
196 197 198 199 |
* the counter to account for mem+swap usage. */ struct res_counter memsw; /* |
78fb74669 Memory controller... |
200 201 |
* Per cgroup active and inactive list, similar to the * per zone LRU lists. |
78fb74669 Memory controller... |
202 |
*/ |
6d12e2d8d per-zone and recl... |
203 |
struct mem_cgroup_lru_info info; |
072c56c13 per-zone and recl... |
204 |
|
2733c06ac memcg: protect pr... |
205 206 207 208 |
/* protect against reclaim related member. */ spinlock_t reclaim_param_lock; |
6d61ef409 memcg: memory cgr... |
209 |
/* |
af901ca18 tree-wide: fix as... |
210 |
* While reclaiming in a hierarchy, we cache the last child we |
04046e1a0 memcg: use CSS ID |
211 |
* reclaimed from. |
6d61ef409 memcg: memory cgr... |
212 |
*/ |
04046e1a0 memcg: use CSS ID |
213 |
int last_scanned_child; |
18f59ea7d memcg: memory cgr... |
214 215 216 217 |
/* * Should the accounting and control be hierarchical, per subtree? */ bool use_hierarchy; |
867578cbc memcg: fix oom ki... |
218 |
atomic_t oom_lock; |
8c7c6e34a memcg: mem+swap c... |
219 |
atomic_t refcnt; |
14797e236 memcg: add inacti... |
220 |
|
a7885eb8a memcg: swappiness |
221 |
unsigned int swappiness; |
3c11ecf44 memcg: oom kill d... |
222 223 |
/* OOM-Killer disable */ int oom_kill_disable; |
a7885eb8a memcg: swappiness |
224 |
|
22a668d7c memcg: fix behavi... |
225 226 |
/* set when res.limit == memsw.limit */ bool memsw_is_minimum; |
2e72b6347 memcg: implement ... |
227 228 229 230 |
/* protect arrays of thresholds */ struct mutex thresholds_lock; /* thresholds for memory usage. RCU-protected */ |
2c488db27 memcg: clean up m... |
231 |
struct mem_cgroup_thresholds thresholds; |
907860ed3 cgroups: make cft... |
232 |
|
2e72b6347 memcg: implement ... |
233 |
/* thresholds for mem+swap usage. RCU-protected */ |
2c488db27 memcg: clean up m... |
234 |
struct mem_cgroup_thresholds memsw_thresholds; |
907860ed3 cgroups: make cft... |
235 |
|
9490ff275 memcg: oom notifier |
236 237 |
/* For oom notifier event fd */ struct list_head oom_notify; |
d52aa412d memory cgroup enh... |
238 |
/* |
7dc74be03 memcg: add interf... |
239 240 241 242 |
* 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 memcg: add interf... |
243 |
/* |
c62b1a3b3 memcg: use generi... |
244 |
* percpu counter. |
d52aa412d memory cgroup enh... |
245 |
*/ |
c62b1a3b3 memcg: use generi... |
246 |
struct mem_cgroup_stat_cpu *stat; |
8cdea7c05 Memory controller... |
247 |
}; |
7dc74be03 memcg: add interf... |
248 249 250 251 252 253 |
/* 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 memcg: move charg... |
254 |
MOVE_CHARGE_TYPE_ANON, /* private anonymous page and swap of it */ |
87946a722 memcg: move charg... |
255 |
MOVE_CHARGE_TYPE_FILE, /* file page(including tmpfs) and swap of it */ |
7dc74be03 memcg: add interf... |
256 257 |
NR_MOVE_TYPE, }; |
4ffef5fef memcg: move charg... |
258 259 |
/* "mc" and its members are protected by cgroup_mutex */ static struct move_charge_struct { |
2bd9bb206 memcg: clean up w... |
260 |
spinlock_t lock; /* for from, to, moving_task */ |
4ffef5fef memcg: move charg... |
261 262 263 |
struct mem_cgroup *from; struct mem_cgroup *to; unsigned long precharge; |
854ffa8d1 memcg: improve pe... |
264 |
unsigned long moved_charge; |
483c30b51 memcg: improve pe... |
265 |
unsigned long moved_swap; |
8033b97c9 memcg: avoid oom ... |
266 267 268 |
struct task_struct *moving_task; /* a task moving charges */ wait_queue_head_t waitq; /* a waitq for other context */ } mc = { |
2bd9bb206 memcg: clean up w... |
269 |
.lock = __SPIN_LOCK_UNLOCKED(mc.lock), |
8033b97c9 memcg: avoid oom ... |
270 271 |
.waitq = __WAIT_QUEUE_HEAD_INITIALIZER(mc.waitq), }; |
4ffef5fef memcg: move charg... |
272 |
|
90254a658 memcg: clean up m... |
273 274 275 276 277 |
static bool move_anon(void) { return test_bit(MOVE_CHARGE_TYPE_ANON, &mc.to->move_charge_at_immigrate); } |
87946a722 memcg: move charg... |
278 279 280 281 282 |
static bool move_file(void) { return test_bit(MOVE_CHARGE_TYPE_FILE, &mc.to->move_charge_at_immigrate); } |
4e4169535 memory controller... |
283 284 285 286 287 288 |
/* * 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 memory cgroup enh... |
289 290 291 |
enum charge_type { MEM_CGROUP_CHARGE_TYPE_CACHE = 0, MEM_CGROUP_CHARGE_TYPE_MAPPED, |
4f98a2fee vmscan: split LRU... |
292 |
MEM_CGROUP_CHARGE_TYPE_SHMEM, /* used by page migration of shmem */ |
c05555b57 memcg: atomic ops... |
293 |
MEM_CGROUP_CHARGE_TYPE_FORCE, /* used by force_empty */ |
d13d14430 memcg: handle swa... |
294 |
MEM_CGROUP_CHARGE_TYPE_SWAPOUT, /* for accounting swapcache */ |
8a9478ca7 memcg: fix swap a... |
295 |
MEM_CGROUP_CHARGE_TYPE_DROP, /* a page was unused swap cache */ |
c05555b57 memcg: atomic ops... |
296 297 |
NR_CHARGE_TYPE, }; |
52d4b9ac0 memcg: allocate a... |
298 299 300 |
/* only for here (for easy reading.) */ #define PCGF_CACHE (1UL << PCG_CACHE) #define PCGF_USED (1UL << PCG_USED) |
52d4b9ac0 memcg: allocate a... |
301 |
#define PCGF_LOCK (1UL << PCG_LOCK) |
4b3bde4c9 memcg: remove the... |
302 303 |
/* Not used, but added here for completeness */ #define PCGF_ACCT (1UL << PCG_ACCT) |
217bc3194 memory cgroup enh... |
304 |
|
8c7c6e34a memcg: mem+swap c... |
305 306 307 |
/* for encoding cft->private value on file */ #define _MEM (0) #define _MEMSWAP (1) |
9490ff275 memcg: oom notifier |
308 |
#define _OOM_TYPE (2) |
8c7c6e34a memcg: mem+swap c... |
309 310 311 |
#define MEMFILE_PRIVATE(x, val) (((x) << 16) | (val)) #define MEMFILE_TYPE(val) (((val) >> 16) & 0xffff) #define MEMFILE_ATTR(val) ((val) & 0xffff) |
9490ff275 memcg: oom notifier |
312 313 |
/* Used for OOM nofiier */ #define OOM_CONTROL (0) |
8c7c6e34a memcg: mem+swap c... |
314 |
|
75822b449 memory controller... |
315 316 317 318 319 320 321 |
/* * 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 memory controller... |
322 323 |
#define MEM_CGROUP_RECLAIM_SOFT_BIT 0x2 #define MEM_CGROUP_RECLAIM_SOFT (1 << MEM_CGROUP_RECLAIM_SOFT_BIT) |
75822b449 memory controller... |
324 |
|
8c7c6e34a memcg: mem+swap c... |
325 326 |
static void mem_cgroup_get(struct mem_cgroup *mem); static void mem_cgroup_put(struct mem_cgroup *mem); |
7bcc1bb12 memcg: get/put pa... |
327 |
static struct mem_cgroup *parent_mem_cgroup(struct mem_cgroup *mem); |
cdec2e426 memcg: coalesce c... |
328 |
static void drain_all_stock_async(void); |
8c7c6e34a memcg: mem+swap c... |
329 |
|
f64c3f549 memory controller... |
330 331 332 333 334 |
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 memcg: add access... |
335 336 337 338 |
struct cgroup_subsys_state *mem_cgroup_css(struct mem_cgroup *mem) { return &mem->css; } |
f64c3f549 memory controller... |
339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 |
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 memory controller... |
368 |
__mem_cgroup_insert_exceeded(struct mem_cgroup *mem, |
f64c3f549 memory controller... |
369 |
struct mem_cgroup_per_zone *mz, |
ef8745c1e memcg: reduce che... |
370 371 |
struct mem_cgroup_tree_per_zone *mctz, unsigned long long new_usage_in_excess) |
f64c3f549 memory controller... |
372 373 374 375 376 377 378 |
{ 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 memcg: reduce che... |
379 380 381 |
mz->usage_in_excess = new_usage_in_excess; if (!mz->usage_in_excess) return; |
f64c3f549 memory controller... |
382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 |
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 memory controller... |
398 399 400 401 402 403 404 405 406 407 408 409 410 411 |
} 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 memory controller... |
412 413 414 415 416 |
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 memory controller... |
417 |
__mem_cgroup_remove_exceeded(mem, mz, mctz); |
f64c3f549 memory controller... |
418 419 |
spin_unlock(&mctz->lock); } |
f64c3f549 memory controller... |
420 421 422 |
static void mem_cgroup_update_tree(struct mem_cgroup *mem, struct page *page) { |
ef8745c1e memcg: reduce che... |
423 |
unsigned long long excess; |
f64c3f549 memory controller... |
424 425 |
struct mem_cgroup_per_zone *mz; struct mem_cgroup_tree_per_zone *mctz; |
4e649152c memcg: some modif... |
426 427 |
int nid = page_to_nid(page); int zid = page_zonenum(page); |
f64c3f549 memory controller... |
428 429 430 |
mctz = soft_limit_tree_from_page(page); /* |
4e649152c memcg: some modif... |
431 432 |
* Necessary to update all ancestors when hierarchy is used. * because their event counter is not touched. |
f64c3f549 memory controller... |
433 |
*/ |
4e649152c memcg: some modif... |
434 435 |
for (; mem; mem = parent_mem_cgroup(mem)) { mz = mem_cgroup_zoneinfo(mem, nid, zid); |
ef8745c1e memcg: reduce che... |
436 |
excess = res_counter_soft_limit_excess(&mem->res); |
4e649152c memcg: some modif... |
437 438 439 440 |
/* * We have to update the tree if mz is on RB-tree or * mem is over its softlimit. */ |
ef8745c1e memcg: reduce che... |
441 |
if (excess || mz->on_tree) { |
4e649152c memcg: some modif... |
442 443 444 445 446 |
spin_lock(&mctz->lock); /* if on-tree, remove it */ if (mz->on_tree) __mem_cgroup_remove_exceeded(mem, mz, mctz); /* |
ef8745c1e memcg: reduce che... |
447 448 |
* Insert again. mz->usage_in_excess will be updated. * If excess is 0, no tree ops. |
4e649152c memcg: some modif... |
449 |
*/ |
ef8745c1e memcg: reduce che... |
450 |
__mem_cgroup_insert_exceeded(mem, mz, mctz, excess); |
4e649152c memcg: some modif... |
451 452 |
spin_unlock(&mctz->lock); } |
f64c3f549 memory controller... |
453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 |
} } 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 memory controller... |
470 471 472 473 474 475 476 477 478 |
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 memcg: fix refcnt... |
479 |
struct mem_cgroup_per_zone *mz; |
4e4169535 memory controller... |
480 481 |
retry: |
26251eaf9 memcg: fix refcnt... |
482 |
mz = NULL; |
4e4169535 memory controller... |
483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 |
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; } |
c62b1a3b3 memcg: use generi... |
511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 |
static s64 mem_cgroup_read_stat(struct mem_cgroup *mem, enum mem_cgroup_stat_index idx) { int cpu; s64 val = 0; for_each_possible_cpu(cpu) val += per_cpu(mem->stat->count[idx], cpu); 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 memcg: improve re... |
530 531 532 533 |
static void mem_cgroup_swap_statistics(struct mem_cgroup *mem, bool charge) { int val = (charge) ? 1 : -1; |
c62b1a3b3 memcg: use generi... |
534 |
this_cpu_add(mem->stat->count[MEM_CGROUP_STAT_SWAPOUT], val); |
0c3e73e84 memcg: improve re... |
535 |
} |
c05555b57 memcg: atomic ops... |
536 537 538 |
static void mem_cgroup_charge_statistics(struct mem_cgroup *mem, struct page_cgroup *pc, bool charge) |
d52aa412d memory cgroup enh... |
539 |
{ |
0c3e73e84 memcg: improve re... |
540 |
int val = (charge) ? 1 : -1; |
d52aa412d memory cgroup enh... |
541 |
|
c62b1a3b3 memcg: use generi... |
542 |
preempt_disable(); |
c05555b57 memcg: atomic ops... |
543 |
if (PageCgroupCache(pc)) |
c62b1a3b3 memcg: use generi... |
544 |
__this_cpu_add(mem->stat->count[MEM_CGROUP_STAT_CACHE], val); |
d52aa412d memory cgroup enh... |
545 |
else |
c62b1a3b3 memcg: use generi... |
546 |
__this_cpu_add(mem->stat->count[MEM_CGROUP_STAT_RSS], val); |
55e462b05 memcg: simple sta... |
547 548 |
if (charge) |
c62b1a3b3 memcg: use generi... |
549 |
__this_cpu_inc(mem->stat->count[MEM_CGROUP_STAT_PGPGIN_COUNT]); |
55e462b05 memcg: simple sta... |
550 |
else |
c62b1a3b3 memcg: use generi... |
551 |
__this_cpu_inc(mem->stat->count[MEM_CGROUP_STAT_PGPGOUT_COUNT]); |
d2265e6fa memcg : share eve... |
552 |
__this_cpu_inc(mem->stat->count[MEM_CGROUP_EVENTS]); |
2e72b6347 memcg: implement ... |
553 |
|
c62b1a3b3 memcg: use generi... |
554 |
preempt_enable(); |
6d12e2d8d per-zone and recl... |
555 |
} |
14067bb3e memcg: hierarchic... |
556 |
static unsigned long mem_cgroup_get_local_zonestat(struct mem_cgroup *mem, |
b69408e88 vmscan: Use an in... |
557 |
enum lru_list idx) |
6d12e2d8d per-zone and recl... |
558 559 560 561 562 563 564 565 566 567 568 |
{ 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 memory cgroup enh... |
569 |
} |
d2265e6fa memcg : share eve... |
570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 |
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 memcg: memcontrol... |
592 |
static struct mem_cgroup *mem_cgroup_from_cont(struct cgroup *cont) |
8cdea7c05 Memory controller... |
593 594 595 596 597 |
{ return container_of(cgroup_subsys_state(cont, mem_cgroup_subsys_id), struct mem_cgroup, css); } |
cf475ad28 cgroups: add an o... |
598 |
struct mem_cgroup *mem_cgroup_from_task(struct task_struct *p) |
78fb74669 Memory controller... |
599 |
{ |
31a78f23b mm owner: fix rac... |
600 601 602 603 604 605 606 |
/* * 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 Memory controller... |
607 608 609 |
return container_of(task_subsys_state(p, mem_cgroup_subsys_id), struct mem_cgroup, css); } |
54595fe26 memcg: use css_tr... |
610 611 612 |
static struct mem_cgroup *try_get_mem_cgroup_from_mm(struct mm_struct *mm) { struct mem_cgroup *mem = NULL; |
0b7f569e4 memcg: fix OOM ki... |
613 614 615 |
if (!mm) return NULL; |
54595fe26 memcg: use css_tr... |
616 617 618 619 620 621 622 623 624 625 626 627 628 629 |
/* * 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; } |
14067bb3e memcg: hierarchic... |
630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 |
/* * Call callback function against all cgroup under hierarchy tree. */ static int mem_cgroup_walk_tree(struct mem_cgroup *root, void *data, int (*func)(struct mem_cgroup *, void *)) { int found, ret, nextid; struct cgroup_subsys_state *css; struct mem_cgroup *mem; if (!root->use_hierarchy) return (*func)(root, data); nextid = 1; do { ret = 0; mem = NULL; rcu_read_lock(); css = css_get_next(&mem_cgroup_subsys, nextid, &root->css, &found); if (css && css_tryget(css)) mem = container_of(css, struct mem_cgroup, css); rcu_read_unlock(); if (mem) { ret = (*func)(mem, data); css_put(&mem->css); } nextid = found + 1; } while (!ret && css); return ret; } |
4b3bde4c9 memcg: remove the... |
664 665 666 667 |
static inline bool mem_cgroup_is_root(struct mem_cgroup *mem) { return (mem == root_mem_cgroup); } |
08e552c69 memcg: synchroniz... |
668 669 670 671 672 673 674 675 676 677 678 679 680 |
/* * 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 vmscan: split LRU... |
681 |
|
08e552c69 memcg: synchroniz... |
682 683 684 |
void mem_cgroup_del_lru_list(struct page *page, enum lru_list lru) { struct page_cgroup *pc; |
08e552c69 memcg: synchroniz... |
685 |
struct mem_cgroup_per_zone *mz; |
6d12e2d8d per-zone and recl... |
686 |
|
f8d665422 memcg: add mem_cg... |
687 |
if (mem_cgroup_disabled()) |
08e552c69 memcg: synchroniz... |
688 689 690 |
return; pc = lookup_page_cgroup(page); /* can happen while we handle swapcache. */ |
4b3bde4c9 memcg: remove the... |
691 |
if (!TestClearPageCgroupAcctLRU(pc)) |
08e552c69 memcg: synchroniz... |
692 |
return; |
4b3bde4c9 memcg: remove the... |
693 |
VM_BUG_ON(!pc->mem_cgroup); |
544122e5e memcg: fix LRU ac... |
694 695 696 697 |
/* * We don't check PCG_USED bit. It's cleared when the "page" is finally * removed from global LRU. */ |
08e552c69 memcg: synchroniz... |
698 |
mz = page_cgroup_zoneinfo(pc); |
b69408e88 vmscan: Use an in... |
699 |
MEM_CGROUP_ZSTAT(mz, lru) -= 1; |
4b3bde4c9 memcg: remove the... |
700 701 702 |
if (mem_cgroup_is_root(pc->mem_cgroup)) return; VM_BUG_ON(list_empty(&pc->lru)); |
08e552c69 memcg: synchroniz... |
703 704 |
list_del_init(&pc->lru); return; |
6d12e2d8d per-zone and recl... |
705 |
} |
08e552c69 memcg: synchroniz... |
706 |
void mem_cgroup_del_lru(struct page *page) |
6d12e2d8d per-zone and recl... |
707 |
{ |
08e552c69 memcg: synchroniz... |
708 709 |
mem_cgroup_del_lru_list(page, page_lru(page)); } |
b69408e88 vmscan: Use an in... |
710 |
|
08e552c69 memcg: synchroniz... |
711 712 713 714 |
void mem_cgroup_rotate_lru_list(struct page *page, enum lru_list lru) { struct mem_cgroup_per_zone *mz; struct page_cgroup *pc; |
b69408e88 vmscan: Use an in... |
715 |
|
f8d665422 memcg: add mem_cg... |
716 |
if (mem_cgroup_disabled()) |
08e552c69 memcg: synchroniz... |
717 |
return; |
6d12e2d8d per-zone and recl... |
718 |
|
08e552c69 memcg: synchroniz... |
719 |
pc = lookup_page_cgroup(page); |
bd112db87 memcg: fix mem_cg... |
720 721 722 723 |
/* * Used bit is set without atomic ops but after smp_wmb(). * For making pc->mem_cgroup visible, insert smp_rmb() here. */ |
08e552c69 memcg: synchroniz... |
724 |
smp_rmb(); |
4b3bde4c9 memcg: remove the... |
725 726 |
/* unused or root page is not rotated. */ if (!PageCgroupUsed(pc) || mem_cgroup_is_root(pc->mem_cgroup)) |
08e552c69 memcg: synchroniz... |
727 728 729 |
return; mz = page_cgroup_zoneinfo(pc); list_move(&pc->lru, &mz->lists[lru]); |
6d12e2d8d per-zone and recl... |
730 |
} |
08e552c69 memcg: synchroniz... |
731 |
void mem_cgroup_add_lru_list(struct page *page, enum lru_list lru) |
66e1707bc Memory controller... |
732 |
{ |
08e552c69 memcg: synchroniz... |
733 734 |
struct page_cgroup *pc; struct mem_cgroup_per_zone *mz; |
6d12e2d8d per-zone and recl... |
735 |
|
f8d665422 memcg: add mem_cg... |
736 |
if (mem_cgroup_disabled()) |
08e552c69 memcg: synchroniz... |
737 738 |
return; pc = lookup_page_cgroup(page); |
4b3bde4c9 memcg: remove the... |
739 |
VM_BUG_ON(PageCgroupAcctLRU(pc)); |
bd112db87 memcg: fix mem_cg... |
740 741 742 743 |
/* * Used bit is set without atomic ops but after smp_wmb(). * For making pc->mem_cgroup visible, insert smp_rmb() here. */ |
08e552c69 memcg: synchroniz... |
744 745 |
smp_rmb(); if (!PageCgroupUsed(pc)) |
894bc3104 Unevictable LRU I... |
746 |
return; |
b69408e88 vmscan: Use an in... |
747 |
|
08e552c69 memcg: synchroniz... |
748 |
mz = page_cgroup_zoneinfo(pc); |
b69408e88 vmscan: Use an in... |
749 |
MEM_CGROUP_ZSTAT(mz, lru) += 1; |
4b3bde4c9 memcg: remove the... |
750 751 752 |
SetPageCgroupAcctLRU(pc); if (mem_cgroup_is_root(pc->mem_cgroup)) return; |
08e552c69 memcg: synchroniz... |
753 754 |
list_add(&pc->lru, &mz->lists[lru]); } |
544122e5e memcg: fix LRU ac... |
755 |
|
08e552c69 memcg: synchroniz... |
756 |
/* |
544122e5e memcg: fix LRU ac... |
757 758 759 760 761 |
* 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 memcg: synchroniz... |
762 |
*/ |
544122e5e memcg: fix LRU ac... |
763 |
static void mem_cgroup_lru_del_before_commit_swapcache(struct page *page) |
08e552c69 memcg: synchroniz... |
764 |
{ |
544122e5e memcg: fix LRU ac... |
765 766 767 768 769 770 771 772 773 774 775 776 |
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 memcg: synchroniz... |
777 |
} |
544122e5e memcg: fix LRU ac... |
778 779 780 781 782 783 784 785 |
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 memcg: remove the... |
786 |
if (PageLRU(page) && !PageCgroupAcctLRU(pc)) |
544122e5e memcg: fix LRU ac... |
787 788 789 |
mem_cgroup_add_lru_list(page, page_lru(page)); spin_unlock_irqrestore(&zone->lru_lock, flags); } |
08e552c69 memcg: synchroniz... |
790 791 792 |
void mem_cgroup_move_lists(struct page *page, enum lru_list from, enum lru_list to) { |
f8d665422 memcg: add mem_cg... |
793 |
if (mem_cgroup_disabled()) |
08e552c69 memcg: synchroniz... |
794 795 796 |
return; mem_cgroup_del_lru_list(page, from); mem_cgroup_add_lru_list(page, to); |
66e1707bc Memory controller... |
797 |
} |
4c4a22148 memcontrol: move ... |
798 799 800 |
int task_in_mem_cgroup(struct task_struct *task, const struct mem_cgroup *mem) { int ret; |
0b7f569e4 memcg: fix OOM ki... |
801 |
struct mem_cgroup *curr = NULL; |
158e0a2d1 memcg: use find_l... |
802 |
struct task_struct *p; |
4c4a22148 memcontrol: move ... |
803 |
|
158e0a2d1 memcg: use find_l... |
804 805 806 807 808 |
p = find_lock_task_mm(task); if (!p) return 0; curr = try_get_mem_cgroup_from_mm(p->mm); task_unlock(p); |
0b7f569e4 memcg: fix OOM ki... |
809 810 |
if (!curr) return 0; |
d31f56dbf memcg: avoid oom-... |
811 812 813 814 815 816 817 |
/* * 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 memcg: fix OOM ki... |
818 819 820 821 |
ret = css_is_ancestor(&curr->css, &mem->css); else ret = (curr == mem); css_put(&curr->css); |
4c4a22148 memcontrol: move ... |
822 823 |
return ret; } |
c772be939 memcg: fix calcul... |
824 |
static int calc_inactive_ratio(struct mem_cgroup *memcg, unsigned long *present_pages) |
14797e236 memcg: add inacti... |
825 826 827 |
{ unsigned long active; unsigned long inactive; |
c772be939 memcg: fix calcul... |
828 829 |
unsigned long gb; unsigned long inactive_ratio; |
14797e236 memcg: add inacti... |
830 |
|
14067bb3e memcg: hierarchic... |
831 832 |
inactive = mem_cgroup_get_local_zonestat(memcg, LRU_INACTIVE_ANON); active = mem_cgroup_get_local_zonestat(memcg, LRU_ACTIVE_ANON); |
14797e236 memcg: add inacti... |
833 |
|
c772be939 memcg: fix calcul... |
834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 |
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 memcg: add inacti... |
861 862 863 864 |
return 1; return 0; } |
56e49d218 vmscan: evict use... |
865 866 867 868 869 870 871 872 873 874 |
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 memcg: add mem_cg... |
875 876 877 878 |
unsigned long mem_cgroup_zone_nr_pages(struct mem_cgroup *memcg, struct zone *zone, enum lru_list lru) { |
13d7e3a2d memcg: convert to... |
879 |
int nid = zone_to_nid(zone); |
a3d8e0549 memcg: add mem_cg... |
880 881 882 883 884 |
int zid = zone_idx(zone); struct mem_cgroup_per_zone *mz = mem_cgroup_zoneinfo(memcg, nid, zid); return MEM_CGROUP_ZSTAT(mz, lru); } |
3e2f41f1f memcg: add zone_r... |
885 886 887 |
struct zone_reclaim_stat *mem_cgroup_get_reclaim_stat(struct mem_cgroup *memcg, struct zone *zone) { |
13d7e3a2d memcg: convert to... |
888 |
int nid = zone_to_nid(zone); |
3e2f41f1f memcg: add zone_r... |
889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 |
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 memcg: fix mem_cg... |
905 906 907 908 909 910 911 |
/* * 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 memcg: add zone_r... |
912 913 914 915 916 917 |
mz = page_cgroup_zoneinfo(pc); if (!mz) return NULL; return &mz->reclaim_stat; } |
66e1707bc Memory controller... |
918 919 920 921 922 |
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 vmscan: split LRU... |
923 |
int active, int file) |
66e1707bc Memory controller... |
924 925 926 927 928 929 |
{ unsigned long nr_taken = 0; struct page *page; unsigned long scan; LIST_HEAD(pc_list); struct list_head *src; |
ff7283fa3 bugfix for memory... |
930 |
struct page_cgroup *pc, *tmp; |
13d7e3a2d memcg: convert to... |
931 |
int nid = zone_to_nid(z); |
1ecaab2bd per-zone and recl... |
932 933 |
int zid = zone_idx(z); struct mem_cgroup_per_zone *mz; |
b7c46d151 mm: drop unneeded... |
934 |
int lru = LRU_FILE * file + active; |
2ffebca6a memcg: fix lru ro... |
935 |
int ret; |
66e1707bc Memory controller... |
936 |
|
cf475ad28 cgroups: add an o... |
937 |
BUG_ON(!mem_cont); |
1ecaab2bd per-zone and recl... |
938 |
mz = mem_cgroup_zoneinfo(mem_cont, nid, zid); |
b69408e88 vmscan: Use an in... |
939 |
src = &mz->lists[lru]; |
66e1707bc Memory controller... |
940 |
|
ff7283fa3 bugfix for memory... |
941 942 |
scan = 0; list_for_each_entry_safe_reverse(pc, tmp, src, lru) { |
436c6541b memcgroup: fix zo... |
943 |
if (scan >= nr_to_scan) |
ff7283fa3 bugfix for memory... |
944 |
break; |
08e552c69 memcg: synchroniz... |
945 946 |
page = pc->page; |
52d4b9ac0 memcg: allocate a... |
947 948 |
if (unlikely(!PageCgroupUsed(pc))) continue; |
436c6541b memcgroup: fix zo... |
949 |
if (unlikely(!PageLRU(page))) |
ff7283fa3 bugfix for memory... |
950 |
continue; |
ff7283fa3 bugfix for memory... |
951 |
|
436c6541b memcgroup: fix zo... |
952 |
scan++; |
2ffebca6a memcg: fix lru ro... |
953 954 955 |
ret = __isolate_lru_page(page, mode, file); switch (ret) { case 0: |
66e1707bc Memory controller... |
956 |
list_move(&page->lru, dst); |
2ffebca6a memcg: fix lru ro... |
957 |
mem_cgroup_del_lru(page); |
66e1707bc Memory controller... |
958 |
nr_taken++; |
2ffebca6a memcg: fix lru ro... |
959 960 961 962 963 964 965 |
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 Memory controller... |
966 967 |
} } |
66e1707bc Memory controller... |
968 |
*scanned = scan; |
cc8e970c3 memcg: add mm_vms... |
969 970 971 |
trace_mm_vmscan_memcg_isolate(0, nr_to_scan, scan, nr_taken, 0, 0, 0, mode); |
66e1707bc Memory controller... |
972 973 |
return nr_taken; } |
6d61ef409 memcg: memory cgr... |
974 975 |
#define mem_cgroup_from_res_counter(counter, member) \ container_of(counter, struct mem_cgroup, member) |
b85a96c0b memcg: memory swa... |
976 977 978 979 980 981 982 983 984 985 986 |
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 memcg: swappiness |
987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 |
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; } |
4b5343346 memcg: clean up t... |
1002 1003 1004 1005 |
/* A routine for testing mem is not under move_account */ static bool mem_cgroup_under_move(struct mem_cgroup *mem) { |
2bd9bb206 memcg: clean up w... |
1006 1007 |
struct mem_cgroup *from; struct mem_cgroup *to; |
4b5343346 memcg: clean up t... |
1008 |
bool ret = false; |
2bd9bb206 memcg: clean up w... |
1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 |
/* * 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 memcg: clean up t... |
1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 |
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; } |
81d39c20f memcg: fix shrink... |
1042 1043 1044 1045 1046 1047 |
static int mem_cgroup_count_children_cb(struct mem_cgroup *mem, void *data) { int *val = data; (*val)++; return 0; } |
e222432bf memcg: show memcg... |
1048 1049 |
/** |
6a6135b64 memcg: typo in co... |
1050 |
* mem_cgroup_print_oom_info: Called from OOM with tasklist_lock held in read mode. |
e222432bf memcg: show memcg... |
1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 |
* @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 memcg: avoid oom-... |
1068 |
if (!memcg || !p) |
e222432bf memcg: show memcg... |
1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 |
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 memcg: fix shrink... |
1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 |
/* * 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; mem_cgroup_walk_tree(mem, &num, mem_cgroup_count_children_cb); return num; } |
6d61ef409 memcg: memory cgr... |
1127 |
/* |
a63d83f42 oom: badness heur... |
1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 |
* Return the memory (and swap, if configured) limit for a memcg. */ u64 mem_cgroup_get_limit(struct mem_cgroup *memcg) { u64 limit; u64 memsw; limit = res_counter_read_u64(&memcg->res, RES_LIMIT) + total_swap_pages; 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 memcg: use CSS ID |
1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 |
* 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 memcg: memory cgr... |
1188 1189 |
* * root_mem is the original ancestor that we've been reclaim from. |
04046e1a0 memcg: use CSS ID |
1190 1191 1192 |
* * We give up and return to the caller when we visit root_mem twice. * (other groups can be removed while we're walking....) |
81d39c20f memcg: fix shrink... |
1193 1194 |
* * If shrink==true, for avoiding to free too much, this returns immedieately. |
6d61ef409 memcg: memory cgr... |
1195 1196 |
*/ static int mem_cgroup_hierarchical_reclaim(struct mem_cgroup *root_mem, |
4e4169535 memory controller... |
1197 |
struct zone *zone, |
75822b449 memory controller... |
1198 1199 |
gfp_t gfp_mask, unsigned long reclaim_options) |
6d61ef409 memcg: memory cgr... |
1200 |
{ |
04046e1a0 memcg: use CSS ID |
1201 1202 1203 |
struct mem_cgroup *victim; int ret, total = 0; int loop = 0; |
75822b449 memory controller... |
1204 1205 |
bool noswap = reclaim_options & MEM_CGROUP_RECLAIM_NOSWAP; bool shrink = reclaim_options & MEM_CGROUP_RECLAIM_SHRINK; |
4e4169535 memory controller... |
1206 1207 |
bool check_soft = reclaim_options & MEM_CGROUP_RECLAIM_SOFT; unsigned long excess = mem_cgroup_get_excess(root_mem); |
04046e1a0 memcg: use CSS ID |
1208 |
|
22a668d7c memcg: fix behavi... |
1209 1210 1211 |
/* If memsw_is_minimum==1, swap-out is of-no-use. */ if (root_mem->memsw_is_minimum) noswap = true; |
4e4169535 memory controller... |
1212 |
while (1) { |
04046e1a0 memcg: use CSS ID |
1213 |
victim = mem_cgroup_select_victim(root_mem); |
4e4169535 memory controller... |
1214 |
if (victim == root_mem) { |
04046e1a0 memcg: use CSS ID |
1215 |
loop++; |
cdec2e426 memcg: coalesce c... |
1216 1217 |
if (loop >= 1) drain_all_stock_async(); |
4e4169535 memory controller... |
1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 |
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 memcg: use generi... |
1241 |
if (!mem_cgroup_local_usage(victim)) { |
04046e1a0 memcg: use CSS ID |
1242 1243 |
/* this cgroup's local usage == 0 */ css_put(&victim->css); |
6d61ef409 memcg: memory cgr... |
1244 1245 |
continue; } |
04046e1a0 memcg: use CSS ID |
1246 |
/* we use swappiness of local cgroup */ |
4e4169535 memory controller... |
1247 1248 |
if (check_soft) ret = mem_cgroup_shrink_node_zone(victim, gfp_mask, |
14fec7968 memcg: mem_cgroup... |
1249 |
noswap, get_swappiness(victim), zone); |
4e4169535 memory controller... |
1250 1251 1252 |
else ret = try_to_free_mem_cgroup_pages(victim, gfp_mask, noswap, get_swappiness(victim)); |
04046e1a0 memcg: use CSS ID |
1253 |
css_put(&victim->css); |
81d39c20f memcg: fix shrink... |
1254 1255 1256 1257 1258 1259 1260 |
/* * 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 memcg: use CSS ID |
1261 |
total += ret; |
4e4169535 memory controller... |
1262 1263 1264 1265 |
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 memcg: use CSS ID |
1266 |
return 1 + total; |
6d61ef409 memcg: memory cgr... |
1267 |
} |
04046e1a0 memcg: use CSS ID |
1268 |
return total; |
6d61ef409 memcg: memory cgr... |
1269 |
} |
867578cbc memcg: fix oom ki... |
1270 |
static int mem_cgroup_oom_lock_cb(struct mem_cgroup *mem, void *data) |
a636b327f memcg: avoid unne... |
1271 |
{ |
867578cbc memcg: fix oom ki... |
1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 |
int *val = (int *)data; int x; /* * Logically, we can stop scanning immediately when we find * a memcg is already locked. But condidering unlock ops and * creation/removal of memcg, scan-all is simple operation. */ x = atomic_inc_return(&mem->oom_lock); *val = max(x, *val); return 0; } /* * 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) { int lock_count = 0; |
a636b327f memcg: avoid unne... |
1290 |
|
867578cbc memcg: fix oom ki... |
1291 1292 1293 1294 1295 |
mem_cgroup_walk_tree(mem, &lock_count, mem_cgroup_oom_lock_cb); if (lock_count == 1) return true; return false; |
a636b327f memcg: avoid unne... |
1296 |
} |
0b7f569e4 memcg: fix OOM ki... |
1297 |
|
867578cbc memcg: fix oom ki... |
1298 |
static int mem_cgroup_oom_unlock_cb(struct mem_cgroup *mem, void *data) |
0b7f569e4 memcg: fix OOM ki... |
1299 |
{ |
867578cbc memcg: fix oom ki... |
1300 1301 1302 1303 1304 1305 |
/* * 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. */ atomic_add_unless(&mem->oom_lock, -1, 0); |
0b7f569e4 memcg: fix OOM ki... |
1306 1307 |
return 0; } |
867578cbc memcg: fix oom ki... |
1308 1309 1310 1311 1312 1313 1314 |
static void mem_cgroup_oom_unlock(struct mem_cgroup *mem) { mem_cgroup_walk_tree(mem, NULL, mem_cgroup_oom_unlock_cb); } static DEFINE_MUTEX(memcg_oom_mutex); static DECLARE_WAIT_QUEUE_HEAD(memcg_oom_waitq); |
dc98df5a1 memcg: oom wakeup... |
1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 |
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 memcg: oom kill d... |
1350 1351 |
static void memcg_oom_recover(struct mem_cgroup *mem) { |
2bd9bb206 memcg: clean up w... |
1352 |
if (mem && atomic_read(&mem->oom_lock)) |
3c11ecf44 memcg: oom kill d... |
1353 1354 |
memcg_wakeup_oom(mem); } |
867578cbc memcg: fix oom ki... |
1355 1356 1357 1358 |
/* * 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 memcg: fix OOM ki... |
1359 |
{ |
dc98df5a1 memcg: oom wakeup... |
1360 |
struct oom_wait_info owait; |
3c11ecf44 memcg: oom kill d... |
1361 |
bool locked, need_to_kill; |
867578cbc memcg: fix oom ki... |
1362 |
|
dc98df5a1 memcg: oom wakeup... |
1363 1364 1365 1366 1367 |
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 memcg: oom kill d... |
1368 |
need_to_kill = true; |
867578cbc memcg: fix oom ki... |
1369 1370 1371 1372 1373 1374 1375 1376 |
/* 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 memcg: oom kill d... |
1377 1378 1379 1380 |
prepare_to_wait(&memcg_oom_waitq, &owait.wait, TASK_KILLABLE); if (!locked || mem->oom_kill_disable) need_to_kill = false; if (locked) |
9490ff275 memcg: oom notifier |
1381 |
mem_cgroup_oom_notify(mem); |
867578cbc memcg: fix oom ki... |
1382 |
mutex_unlock(&memcg_oom_mutex); |
3c11ecf44 memcg: oom kill d... |
1383 1384 |
if (need_to_kill) { finish_wait(&memcg_oom_waitq, &owait.wait); |
867578cbc memcg: fix oom ki... |
1385 |
mem_cgroup_out_of_memory(mem, mask); |
3c11ecf44 memcg: oom kill d... |
1386 |
} else { |
867578cbc memcg: fix oom ki... |
1387 |
schedule(); |
dc98df5a1 memcg: oom wakeup... |
1388 |
finish_wait(&memcg_oom_waitq, &owait.wait); |
867578cbc memcg: fix oom ki... |
1389 1390 1391 |
} mutex_lock(&memcg_oom_mutex); mem_cgroup_oom_unlock(mem); |
dc98df5a1 memcg: oom wakeup... |
1392 |
memcg_wakeup_oom(mem); |
867578cbc memcg: fix oom ki... |
1393 1394 1395 1396 1397 1398 1399 |
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 memcg: fix OOM ki... |
1400 |
} |
d69b042f3 memcg: add file-b... |
1401 1402 1403 1404 |
/* * Currently used to update mapped file statistics, but the routine can be * generalized to update other statistics as well. */ |
d8046582d memcg: make memcg... |
1405 |
void mem_cgroup_update_file_mapped(struct page *page, int val) |
d69b042f3 memcg: add file-b... |
1406 1407 |
{ struct mem_cgroup *mem; |
d69b042f3 memcg: add file-b... |
1408 |
struct page_cgroup *pc; |
d69b042f3 memcg: add file-b... |
1409 1410 1411 1412 1413 1414 |
pc = lookup_page_cgroup(page); if (unlikely(!pc)) return; lock_page_cgroup(pc); mem = pc->mem_cgroup; |
8725d5416 memcg: fix race i... |
1415 |
if (!mem || !PageCgroupUsed(pc)) |
d69b042f3 memcg: add file-b... |
1416 1417 1418 |
goto done; /* |
c62b1a3b3 memcg: use generi... |
1419 |
* Preemption is already disabled. We can use __this_cpu_xxx |
d69b042f3 memcg: add file-b... |
1420 |
*/ |
8725d5416 memcg: fix race i... |
1421 1422 1423 1424 1425 1426 1427 |
if (val > 0) { __this_cpu_inc(mem->stat->count[MEM_CGROUP_STAT_FILE_MAPPED]); SetPageCgroupFileMapped(pc); } else { __this_cpu_dec(mem->stat->count[MEM_CGROUP_STAT_FILE_MAPPED]); ClearPageCgroupFileMapped(pc); } |
d69b042f3 memcg: add file-b... |
1428 |
|
d69b042f3 memcg: add file-b... |
1429 1430 1431 |
done: unlock_page_cgroup(pc); } |
0b7f569e4 memcg: fix OOM ki... |
1432 |
|
f817ed485 memcg: move all a... |
1433 |
/* |
cdec2e426 memcg: coalesce c... |
1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 |
* 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 mm: fix typo in r... |
1494 |
* This will be consumed by consume_stock() function, later. |
cdec2e426 memcg: coalesce c... |
1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 |
*/ 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); } static int __cpuinit memcg_stock_cpu_callback(struct notifier_block *nb, unsigned long action, void *hcpu) { int cpu = (unsigned long)hcpu; struct memcg_stock_pcp *stock; if (action != CPU_DEAD) return NOTIFY_OK; stock = &per_cpu(memcg_stock, cpu); drain_stock(stock); return NOTIFY_OK; } |
4b5343346 memcg: clean up t... |
1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 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 |
/* 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 memcg: coalesce c... |
1625 |
/* |
f817ed485 memcg: move all a... |
1626 1627 |
* Unlike exported interface, "oom" parameter is added. if oom==true, * oom-killer can be invoked. |
8a9f3ccd2 Memory controller... |
1628 |
*/ |
f817ed485 memcg: move all a... |
1629 |
static int __mem_cgroup_try_charge(struct mm_struct *mm, |
4b5343346 memcg: clean up t... |
1630 |
gfp_t gfp_mask, struct mem_cgroup **memcg, bool oom) |
8a9f3ccd2 Memory controller... |
1631 |
{ |
4b5343346 memcg: clean up t... |
1632 1633 1634 |
int nr_oom_retries = MEM_CGROUP_RECLAIM_RETRIES; struct mem_cgroup *mem = NULL; int ret; |
cdec2e426 memcg: coalesce c... |
1635 |
int csize = CHARGE_SIZE; |
a636b327f memcg: avoid unne... |
1636 |
|
867578cbc memcg: fix oom ki... |
1637 1638 1639 1640 1641 1642 1643 1644 |
/* * 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 memcg: avoid unne... |
1645 |
|
8a9f3ccd2 Memory controller... |
1646 |
/* |
3be91277e memcgroup: tidy u... |
1647 1648 |
* We always charge the cgroup the mm_struct belongs to. * The mm_struct's mem_cgroup changes on task migration if the |
8a9f3ccd2 Memory controller... |
1649 1650 1651 |
* thread group leader migrates. It's possible that mm is not * set, if so charge the init_mm (happens for pagecache usage). */ |
f75ca9620 memcg: avoid css_... |
1652 1653 1654 1655 |
if (!*memcg && !mm) goto bypass; again: if (*memcg) { /* css should be a valid one */ |
4b5343346 memcg: clean up t... |
1656 |
mem = *memcg; |
f75ca9620 memcg: avoid css_... |
1657 1658 1659 1660 1661 |
VM_BUG_ON(css_is_removed(&mem->css)); if (mem_cgroup_is_root(mem)) goto done; if (consume_stock(mem)) goto done; |
4b5343346 memcg: clean up t... |
1662 1663 |
css_get(&mem->css); } else { |
f75ca9620 memcg: avoid css_... |
1664 |
struct task_struct *p; |
54595fe26 memcg: use css_tr... |
1665 |
|
f75ca9620 memcg: avoid css_... |
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 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 |
rcu_read_lock(); p = rcu_dereference(mm->owner); VM_BUG_ON(!p); /* * because we don't have task_lock(), "p" can exit while * we're here. In that case, "mem" can point to root * cgroup but never be NULL. (and task_struct itself is freed * by RCU, cgroup itself is RCU safe.) Then, we have small * risk here to get wrong cgroup. But such kind of mis-account * by race always happens because we don't have cgroup_mutex(). * It's overkill and we allow that small race, here. */ mem = mem_cgroup_from_task(p); VM_BUG_ON(!mem); if (mem_cgroup_is_root(mem)) { rcu_read_unlock(); goto done; } if (consume_stock(mem)) { /* * 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 Memory controller... |
1703 |
|
4b5343346 memcg: clean up t... |
1704 1705 |
do { bool oom_check; |
7a81b88cb memcg: introduce ... |
1706 |
|
4b5343346 memcg: clean up t... |
1707 |
/* If killed, bypass charge */ |
f75ca9620 memcg: avoid css_... |
1708 1709 |
if (fatal_signal_pending(current)) { css_put(&mem->css); |
4b5343346 memcg: clean up t... |
1710 |
goto bypass; |
f75ca9620 memcg: avoid css_... |
1711 |
} |
6d61ef409 memcg: memory cgr... |
1712 |
|
4b5343346 memcg: clean up t... |
1713 1714 1715 1716 |
oom_check = false; if (oom && !nr_oom_retries) { oom_check = true; nr_oom_retries = MEM_CGROUP_RECLAIM_RETRIES; |
cdec2e426 memcg: coalesce c... |
1717 |
} |
66e1707bc Memory controller... |
1718 |
|
4b5343346 memcg: clean up t... |
1719 |
ret = __mem_cgroup_do_charge(mem, gfp_mask, csize, oom_check); |
8033b97c9 memcg: avoid oom ... |
1720 |
|
4b5343346 memcg: clean up t... |
1721 1722 1723 1724 1725 |
switch (ret) { case CHARGE_OK: break; case CHARGE_RETRY: /* not in OOM situation but retry */ csize = PAGE_SIZE; |
f75ca9620 memcg: avoid css_... |
1726 1727 1728 |
css_put(&mem->css); mem = NULL; goto again; |
4b5343346 memcg: clean up t... |
1729 |
case CHARGE_WOULDBLOCK: /* !__GFP_WAIT */ |
f75ca9620 memcg: avoid css_... |
1730 |
css_put(&mem->css); |
4b5343346 memcg: clean up t... |
1731 1732 |
goto nomem; case CHARGE_NOMEM: /* OOM routine works */ |
f75ca9620 memcg: avoid css_... |
1733 1734 |
if (!oom) { css_put(&mem->css); |
867578cbc memcg: fix oom ki... |
1735 |
goto nomem; |
f75ca9620 memcg: avoid css_... |
1736 |
} |
4b5343346 memcg: clean up t... |
1737 1738 1739 1740 |
/* If oom, we never return -ENOMEM */ nr_oom_retries--; break; case CHARGE_OOM_DIE: /* Killed by OOM Killer */ |
f75ca9620 memcg: avoid css_... |
1741 |
css_put(&mem->css); |
867578cbc memcg: fix oom ki... |
1742 |
goto bypass; |
66e1707bc Memory controller... |
1743 |
} |
4b5343346 memcg: clean up t... |
1744 |
} while (ret != CHARGE_OK); |
cdec2e426 memcg: coalesce c... |
1745 1746 |
if (csize > PAGE_SIZE) refill_stock(mem, csize - PAGE_SIZE); |
f75ca9620 memcg: avoid css_... |
1747 |
css_put(&mem->css); |
0c3e73e84 memcg: improve re... |
1748 |
done: |
f75ca9620 memcg: avoid css_... |
1749 |
*memcg = mem; |
7a81b88cb memcg: introduce ... |
1750 1751 |
return 0; nomem: |
f75ca9620 memcg: avoid css_... |
1752 |
*memcg = NULL; |
7a81b88cb memcg: introduce ... |
1753 |
return -ENOMEM; |
867578cbc memcg: fix oom ki... |
1754 1755 1756 |
bypass: *memcg = NULL; return 0; |
7a81b88cb memcg: introduce ... |
1757 |
} |
8a9f3ccd2 Memory controller... |
1758 |
|
a3b2d6926 cgroups: use css ... |
1759 |
/* |
a3032a2c1 memcg: add mem_cg... |
1760 1761 1762 1763 |
* 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 memcg: improve pe... |
1764 1765 |
static void __mem_cgroup_cancel_charge(struct mem_cgroup *mem, unsigned long count) |
a3032a2c1 memcg: add mem_cg... |
1766 1767 |
{ if (!mem_cgroup_is_root(mem)) { |
854ffa8d1 memcg: improve pe... |
1768 |
res_counter_uncharge(&mem->res, PAGE_SIZE * count); |
a3032a2c1 memcg: add mem_cg... |
1769 |
if (do_swap_account) |
854ffa8d1 memcg: improve pe... |
1770 |
res_counter_uncharge(&mem->memsw, PAGE_SIZE * count); |
a3032a2c1 memcg: add mem_cg... |
1771 |
} |
854ffa8d1 memcg: improve pe... |
1772 1773 1774 1775 1776 |
} static void mem_cgroup_cancel_charge(struct mem_cgroup *mem) { __mem_cgroup_cancel_charge(mem, 1); |
a3032a2c1 memcg: add mem_cg... |
1777 1778 1779 |
} /* |
a3b2d6926 cgroups: use css ... |
1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 |
* 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 memcg: rename and... |
1797 |
struct mem_cgroup *try_get_mem_cgroup_from_page(struct page *page) |
b5a84319a memcg: fix shmem'... |
1798 |
{ |
e42d9d5d4 memcg: rename and... |
1799 |
struct mem_cgroup *mem = NULL; |
3c776e646 memcg: charge swa... |
1800 |
struct page_cgroup *pc; |
a3b2d6926 cgroups: use css ... |
1801 |
unsigned short id; |
b5a84319a memcg: fix shmem'... |
1802 |
swp_entry_t ent; |
3c776e646 memcg: charge swa... |
1803 |
VM_BUG_ON(!PageLocked(page)); |
3c776e646 memcg: charge swa... |
1804 |
pc = lookup_page_cgroup(page); |
c0bd3f63c memcg: fix try_ge... |
1805 |
lock_page_cgroup(pc); |
a3b2d6926 cgroups: use css ... |
1806 |
if (PageCgroupUsed(pc)) { |
3c776e646 memcg: charge swa... |
1807 |
mem = pc->mem_cgroup; |
a3b2d6926 cgroups: use css ... |
1808 1809 |
if (mem && !css_tryget(&mem->css)) mem = NULL; |
e42d9d5d4 memcg: rename and... |
1810 |
} else if (PageSwapCache(page)) { |
3c776e646 memcg: charge swa... |
1811 |
ent.val = page_private(page); |
a3b2d6926 cgroups: use css ... |
1812 1813 1814 1815 1816 1817 |
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 memcg: charge swa... |
1818 |
} |
c0bd3f63c memcg: fix try_ge... |
1819 |
unlock_page_cgroup(pc); |
b5a84319a memcg: fix shmem'... |
1820 1821 |
return mem; } |
7a81b88cb memcg: introduce ... |
1822 |
/* |
a5e924f5f memcg: remove mem... |
1823 |
* commit a charge got by __mem_cgroup_try_charge() and makes page_cgroup to be |
7a81b88cb memcg: introduce ... |
1824 1825 1826 1827 1828 1829 1830 |
* USED state. If already USED, uncharge and return. */ static void __mem_cgroup_commit_charge(struct mem_cgroup *mem, struct page_cgroup *pc, enum charge_type ctype) { |
7a81b88cb memcg: introduce ... |
1831 1832 1833 |
/* try_charge() can return NULL to *memcg, taking care of it. */ if (!mem) return; |
52d4b9ac0 memcg: allocate a... |
1834 1835 1836 1837 |
lock_page_cgroup(pc); if (unlikely(PageCgroupUsed(pc))) { unlock_page_cgroup(pc); |
a3032a2c1 memcg: add mem_cg... |
1838 |
mem_cgroup_cancel_charge(mem); |
7a81b88cb memcg: introduce ... |
1839 |
return; |
52d4b9ac0 memcg: allocate a... |
1840 |
} |
4b3bde4c9 memcg: remove the... |
1841 |
|
8a9f3ccd2 Memory controller... |
1842 |
pc->mem_cgroup = mem; |
261fb61a8 memcg: add commen... |
1843 1844 1845 1846 1847 1848 1849 |
/* * 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 memcg: synchroniz... |
1850 |
smp_wmb(); |
4b3bde4c9 memcg: remove the... |
1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 |
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 memcgroup: tidy u... |
1864 |
|
08e552c69 memcg: synchroniz... |
1865 |
mem_cgroup_charge_statistics(mem, pc, true); |
52d4b9ac0 memcg: allocate a... |
1866 |
|
52d4b9ac0 memcg: allocate a... |
1867 |
unlock_page_cgroup(pc); |
430e48631 memcg: update thr... |
1868 1869 1870 1871 1872 |
/* * "charge_statistics" updated event counter. Then, check it. * Insert ancestor (and ancestor's ancestors), to softlimit RB-tree. * if they exceeds softlimit. */ |
d2265e6fa memcg : share eve... |
1873 |
memcg_check_events(mem, pc->page); |
7a81b88cb memcg: introduce ... |
1874 |
} |
66e1707bc Memory controller... |
1875 |
|
f817ed485 memcg: move all a... |
1876 |
/** |
57f9fd7d2 memcg: cleanup me... |
1877 |
* __mem_cgroup_move_account - move account of the page |
f817ed485 memcg: move all a... |
1878 1879 1880 |
* @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 memcg: improve pe... |
1881 |
* @uncharge: whether we should call uncharge and css_put against @from. |
f817ed485 memcg: move all a... |
1882 1883 |
* * The caller must confirm following. |
08e552c69 memcg: synchroniz... |
1884 |
* - page is not on LRU (isolate_page() is useful.) |
57f9fd7d2 memcg: cleanup me... |
1885 |
* - the pc is locked, used, and ->mem_cgroup points to @from. |
f817ed485 memcg: move all a... |
1886 |
* |
854ffa8d1 memcg: improve pe... |
1887 1888 1889 1890 |
* 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 memcg: move all a... |
1891 |
*/ |
57f9fd7d2 memcg: cleanup me... |
1892 |
static void __mem_cgroup_move_account(struct page_cgroup *pc, |
854ffa8d1 memcg: improve pe... |
1893 |
struct mem_cgroup *from, struct mem_cgroup *to, bool uncharge) |
f817ed485 memcg: move all a... |
1894 |
{ |
f817ed485 memcg: move all a... |
1895 |
VM_BUG_ON(from == to); |
08e552c69 memcg: synchroniz... |
1896 |
VM_BUG_ON(PageLRU(pc->page)); |
57f9fd7d2 memcg: cleanup me... |
1897 1898 1899 |
VM_BUG_ON(!PageCgroupLocked(pc)); VM_BUG_ON(!PageCgroupUsed(pc)); VM_BUG_ON(pc->mem_cgroup != from); |
f817ed485 memcg: move all a... |
1900 |
|
8725d5416 memcg: fix race i... |
1901 |
if (PageCgroupFileMapped(pc)) { |
c62b1a3b3 memcg: use generi... |
1902 1903 1904 1905 1906 |
/* 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 memcg: add file-b... |
1907 |
} |
854ffa8d1 memcg: improve pe... |
1908 1909 1910 1911 |
mem_cgroup_charge_statistics(from, pc, false); if (uncharge) /* This is not "cancel", but cancel_charge does all we need. */ mem_cgroup_cancel_charge(from); |
d69b042f3 memcg: add file-b... |
1912 |
|
854ffa8d1 memcg: improve pe... |
1913 |
/* caller should have done css_get */ |
08e552c69 memcg: synchroniz... |
1914 1915 |
pc->mem_cgroup = to; mem_cgroup_charge_statistics(to, pc, true); |
887032670 cgroup avoid perm... |
1916 1917 1918 |
/* * We charges against "to" which may not have any tasks. Then, "to" * can be under rmdir(). But in current implementation, caller of |
4ffef5fef memcg: move charg... |
1919 1920 1921 |
* 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 cgroup avoid perm... |
1922 |
*/ |
57f9fd7d2 memcg: cleanup me... |
1923 1924 1925 1926 1927 1928 1929 |
} /* * 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 memcg: improve pe... |
1930 |
struct mem_cgroup *from, struct mem_cgroup *to, bool uncharge) |
57f9fd7d2 memcg: cleanup me... |
1931 1932 1933 1934 |
{ int ret = -EINVAL; lock_page_cgroup(pc); if (PageCgroupUsed(pc) && pc->mem_cgroup == from) { |
854ffa8d1 memcg: improve pe... |
1935 |
__mem_cgroup_move_account(pc, from, to, uncharge); |
57f9fd7d2 memcg: cleanup me... |
1936 1937 1938 |
ret = 0; } unlock_page_cgroup(pc); |
d2265e6fa memcg : share eve... |
1939 1940 1941 1942 1943 |
/* * check events */ memcg_check_events(to, pc->page); memcg_check_events(from, pc->page); |
f817ed485 memcg: move all a... |
1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 |
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 memcg: synchroniz... |
1955 |
struct page *page = pc->page; |
f817ed485 memcg: move all a... |
1956 1957 1958 |
struct cgroup *cg = child->css.cgroup; struct cgroup *pcg = cg->parent; struct mem_cgroup *parent; |
f817ed485 memcg: move all a... |
1959 1960 1961 1962 1963 |
int ret; /* Is ROOT ? */ if (!pcg) return -EINVAL; |
57f9fd7d2 memcg: cleanup me... |
1964 1965 1966 1967 1968 |
ret = -EBUSY; if (!get_page_unless_zero(page)) goto out; if (isolate_lru_page(page)) goto put; |
08e552c69 memcg: synchroniz... |
1969 |
|
f817ed485 memcg: move all a... |
1970 |
parent = mem_cgroup_from_cont(pcg); |
430e48631 memcg: update thr... |
1971 |
ret = __mem_cgroup_try_charge(NULL, gfp_mask, &parent, false); |
a636b327f memcg: avoid unne... |
1972 |
if (ret || !parent) |
57f9fd7d2 memcg: cleanup me... |
1973 |
goto put_back; |
f817ed485 memcg: move all a... |
1974 |
|
854ffa8d1 memcg: improve pe... |
1975 1976 1977 |
ret = mem_cgroup_move_account(pc, child, parent, true); if (ret) mem_cgroup_cancel_charge(parent); |
57f9fd7d2 memcg: cleanup me... |
1978 |
put_back: |
08e552c69 memcg: synchroniz... |
1979 |
putback_lru_page(page); |
57f9fd7d2 memcg: cleanup me... |
1980 |
put: |
40d58138f memcg: fix error ... |
1981 |
put_page(page); |
57f9fd7d2 memcg: cleanup me... |
1982 |
out: |
f817ed485 memcg: move all a... |
1983 1984 |
return ret; } |
7a81b88cb memcg: introduce ... |
1985 1986 1987 1988 1989 1990 1991 |
/* * 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 memcg: remove mem... |
1992 |
gfp_t gfp_mask, enum charge_type ctype) |
7a81b88cb memcg: introduce ... |
1993 |
{ |
73045c47b memcg: remove mem... |
1994 |
struct mem_cgroup *mem = NULL; |
7a81b88cb memcg: introduce ... |
1995 1996 1997 1998 1999 2000 2001 2002 |
struct page_cgroup *pc; int ret; pc = lookup_page_cgroup(page); /* can happen at boot */ if (unlikely(!pc)) return 0; prefetchw(pc); |
430e48631 memcg: update thr... |
2003 |
ret = __mem_cgroup_try_charge(mm, gfp_mask, &mem, true); |
a636b327f memcg: avoid unne... |
2004 |
if (ret || !mem) |
7a81b88cb memcg: introduce ... |
2005 2006 2007 |
return ret; __mem_cgroup_commit_charge(mem, pc, ctype); |
8a9f3ccd2 Memory controller... |
2008 |
return 0; |
8a9f3ccd2 Memory controller... |
2009 |
} |
7a81b88cb memcg: introduce ... |
2010 2011 |
int mem_cgroup_newpage_charge(struct page *page, struct mm_struct *mm, gfp_t gfp_mask) |
217bc3194 memory cgroup enh... |
2012 |
{ |
f8d665422 memcg: add mem_cg... |
2013 |
if (mem_cgroup_disabled()) |
cede86acd memcg: clean up c... |
2014 |
return 0; |
52d4b9ac0 memcg: allocate a... |
2015 2016 |
if (PageCompound(page)) return 0; |
69029cd55 memcg: remove ref... |
2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 |
/* * 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 memory cgroup enh... |
2028 |
return mem_cgroup_charge_common(page, mm, gfp_mask, |
73045c47b memcg: remove mem... |
2029 |
MEM_CGROUP_CHARGE_TYPE_MAPPED); |
217bc3194 memory cgroup enh... |
2030 |
} |
83aae4c73 memcg: cleanup ca... |
2031 2032 2033 |
static void __mem_cgroup_commit_charge_swapin(struct page *page, struct mem_cgroup *ptr, enum charge_type ctype); |
e1a1cd590 Memory controller... |
2034 2035 |
int mem_cgroup_cache_charge(struct page *page, struct mm_struct *mm, gfp_t gfp_mask) |
8697d3319 Memory controller... |
2036 |
{ |
b5a84319a memcg: fix shmem'... |
2037 |
int ret; |
f8d665422 memcg: add mem_cg... |
2038 |
if (mem_cgroup_disabled()) |
cede86acd memcg: clean up c... |
2039 |
return 0; |
52d4b9ac0 memcg: allocate a... |
2040 2041 |
if (PageCompound(page)) return 0; |
accf163e6 memcg: remove a r... |
2042 2043 2044 2045 2046 2047 2048 2049 |
/* * 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 memcg: fix shmem'... |
2050 2051 |
* And when the page is SwapCache, it should take swap information * into account. This is under lock_page() now. |
accf163e6 memcg: remove a r... |
2052 2053 2054 |
*/ if (!(gfp_mask & __GFP_WAIT)) { struct page_cgroup *pc; |
52d4b9ac0 memcg: allocate a... |
2055 2056 2057 2058 2059 2060 |
pc = lookup_page_cgroup(page); if (!pc) return 0; lock_page_cgroup(pc); if (PageCgroupUsed(pc)) { unlock_page_cgroup(pc); |
accf163e6 memcg: remove a r... |
2061 2062 |
return 0; } |
52d4b9ac0 memcg: allocate a... |
2063 |
unlock_page_cgroup(pc); |
accf163e6 memcg: remove a r... |
2064 |
} |
73045c47b memcg: remove mem... |
2065 |
if (unlikely(!mm)) |
8697d3319 Memory controller... |
2066 |
mm = &init_mm; |
accf163e6 memcg: remove a r... |
2067 |
|
c05555b57 memcg: atomic ops... |
2068 2069 |
if (page_is_file_cache(page)) return mem_cgroup_charge_common(page, mm, gfp_mask, |
73045c47b memcg: remove mem... |
2070 |
MEM_CGROUP_CHARGE_TYPE_CACHE); |
b5a84319a memcg: fix shmem'... |
2071 |
|
83aae4c73 memcg: cleanup ca... |
2072 2073 |
/* shmem */ if (PageSwapCache(page)) { |
73045c47b memcg: remove mem... |
2074 |
struct mem_cgroup *mem = NULL; |
83aae4c73 memcg: cleanup ca... |
2075 2076 2077 2078 2079 2080 |
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 memcg: remove mem... |
2081 |
MEM_CGROUP_CHARGE_TYPE_SHMEM); |
b5a84319a memcg: fix shmem'... |
2082 |
|
b5a84319a memcg: fix shmem'... |
2083 |
return ret; |
e8589cc18 memcg: better mig... |
2084 |
} |
54595fe26 memcg: use css_tr... |
2085 2086 2087 |
/* * While swap-in, try_charge -> commit or cancel, the page is locked. * And when try_charge() successfully returns, one refcnt to memcg without |
21ae2956c tree-wide: fix ty... |
2088 |
* struct page_cgroup is acquired. This refcnt will be consumed by |
54595fe26 memcg: use css_tr... |
2089 2090 |
* "commit()" or removed by "cancel()" */ |
8c7c6e34a memcg: mem+swap c... |
2091 2092 2093 2094 2095 |
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 memcg: use css_tr... |
2096 |
int ret; |
8c7c6e34a memcg: mem+swap c... |
2097 |
|
f8d665422 memcg: add mem_cg... |
2098 |
if (mem_cgroup_disabled()) |
8c7c6e34a memcg: mem+swap c... |
2099 2100 2101 2102 |
return 0; if (!do_swap_account) goto charge_cur_mm; |
8c7c6e34a memcg: mem+swap c... |
2103 2104 |
/* * A racing thread's fault, or swapoff, may have already updated |
407f9c8b0 ksm: mem cgroup c... |
2105 2106 2107 |
* 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 memcg: mem+swap c... |
2108 2109 |
*/ if (!PageSwapCache(page)) |
407f9c8b0 ksm: mem cgroup c... |
2110 |
goto charge_cur_mm; |
e42d9d5d4 memcg: rename and... |
2111 |
mem = try_get_mem_cgroup_from_page(page); |
54595fe26 memcg: use css_tr... |
2112 2113 |
if (!mem) goto charge_cur_mm; |
8c7c6e34a memcg: mem+swap c... |
2114 |
*ptr = mem; |
430e48631 memcg: update thr... |
2115 |
ret = __mem_cgroup_try_charge(NULL, mask, ptr, true); |
54595fe26 memcg: use css_tr... |
2116 2117 |
css_put(&mem->css); return ret; |
8c7c6e34a memcg: mem+swap c... |
2118 2119 2120 |
charge_cur_mm: if (unlikely(!mm)) mm = &init_mm; |
430e48631 memcg: update thr... |
2121 |
return __mem_cgroup_try_charge(mm, mask, ptr, true); |
8c7c6e34a memcg: mem+swap c... |
2122 |
} |
83aae4c73 memcg: cleanup ca... |
2123 2124 2125 |
static void __mem_cgroup_commit_charge_swapin(struct page *page, struct mem_cgroup *ptr, enum charge_type ctype) |
7a81b88cb memcg: introduce ... |
2126 2127 |
{ struct page_cgroup *pc; |
f8d665422 memcg: add mem_cg... |
2128 |
if (mem_cgroup_disabled()) |
7a81b88cb memcg: introduce ... |
2129 2130 2131 |
return; if (!ptr) return; |
887032670 cgroup avoid perm... |
2132 |
cgroup_exclude_rmdir(&ptr->css); |
7a81b88cb memcg: introduce ... |
2133 |
pc = lookup_page_cgroup(page); |
544122e5e memcg: fix LRU ac... |
2134 |
mem_cgroup_lru_del_before_commit_swapcache(page); |
83aae4c73 memcg: cleanup ca... |
2135 |
__mem_cgroup_commit_charge(ptr, pc, ctype); |
544122e5e memcg: fix LRU ac... |
2136 |
mem_cgroup_lru_add_after_commit_swapcache(page); |
8c7c6e34a memcg: mem+swap c... |
2137 2138 2139 |
/* * Now swap is on-memory. This means this page may be * counted both as mem and swap....double count. |
03f3c4336 memcg: fix swap a... |
2140 2141 2142 |
* 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 memcg: mem+swap c... |
2143 |
*/ |
03f3c4336 memcg: fix swap a... |
2144 |
if (do_swap_account && PageSwapCache(page)) { |
8c7c6e34a memcg: mem+swap c... |
2145 |
swp_entry_t ent = {.val = page_private(page)}; |
a3b2d6926 cgroups: use css ... |
2146 |
unsigned short id; |
8c7c6e34a memcg: mem+swap c... |
2147 |
struct mem_cgroup *memcg; |
a3b2d6926 cgroups: use css ... |
2148 2149 2150 2151 |
id = swap_cgroup_record(ent, 0); rcu_read_lock(); memcg = mem_cgroup_lookup(id); |
8c7c6e34a memcg: mem+swap c... |
2152 |
if (memcg) { |
a3b2d6926 cgroups: use css ... |
2153 2154 2155 2156 |
/* * This recorded memcg can be obsolete one. So, avoid * calling css_tryget */ |
0c3e73e84 memcg: improve re... |
2157 |
if (!mem_cgroup_is_root(memcg)) |
4e649152c memcg: some modif... |
2158 |
res_counter_uncharge(&memcg->memsw, PAGE_SIZE); |
0c3e73e84 memcg: improve re... |
2159 |
mem_cgroup_swap_statistics(memcg, false); |
8c7c6e34a memcg: mem+swap c... |
2160 2161 |
mem_cgroup_put(memcg); } |
a3b2d6926 cgroups: use css ... |
2162 |
rcu_read_unlock(); |
8c7c6e34a memcg: mem+swap c... |
2163 |
} |
887032670 cgroup avoid perm... |
2164 2165 2166 2167 2168 2169 |
/* * 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 memcg: introduce ... |
2170 |
} |
83aae4c73 memcg: cleanup ca... |
2171 2172 2173 2174 2175 |
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 memcg: introduce ... |
2176 2177 |
void mem_cgroup_cancel_charge_swapin(struct mem_cgroup *mem) { |
f8d665422 memcg: add mem_cg... |
2178 |
if (mem_cgroup_disabled()) |
7a81b88cb memcg: introduce ... |
2179 2180 2181 |
return; if (!mem) return; |
a3032a2c1 memcg: add mem_cg... |
2182 |
mem_cgroup_cancel_charge(mem); |
7a81b88cb memcg: introduce ... |
2183 |
} |
569b846df memcg: coalesce u... |
2184 2185 2186 2187 2188 2189 2190 2191 |
static void __do_uncharge(struct mem_cgroup *mem, const enum charge_type ctype) { 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 memcg: coalesce u... |
2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 |
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 memcg: oom kill d... |
2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 |
* 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; /* |
569b846df memcg: coalesce u... |
2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 |
* 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: res_counter_uncharge(&mem->res, PAGE_SIZE); if (uncharge_memsw) res_counter_uncharge(&mem->memsw, PAGE_SIZE); |
3c11ecf44 memcg: oom kill d... |
2228 2229 |
if (unlikely(batch->memcg != mem)) memcg_oom_recover(mem); |
569b846df memcg: coalesce u... |
2230 2231 |
return; } |
7a81b88cb memcg: introduce ... |
2232 |
|
8697d3319 Memory controller... |
2233 |
/* |
69029cd55 memcg: remove ref... |
2234 |
* uncharge if !page_mapped(page) |
8a9f3ccd2 Memory controller... |
2235 |
*/ |
8c7c6e34a memcg: mem+swap c... |
2236 |
static struct mem_cgroup * |
69029cd55 memcg: remove ref... |
2237 |
__mem_cgroup_uncharge_common(struct page *page, enum charge_type ctype) |
8a9f3ccd2 Memory controller... |
2238 |
{ |
8289546e5 memcg: remove mem... |
2239 |
struct page_cgroup *pc; |
8c7c6e34a memcg: mem+swap c... |
2240 |
struct mem_cgroup *mem = NULL; |
8a9f3ccd2 Memory controller... |
2241 |
|
f8d665422 memcg: add mem_cg... |
2242 |
if (mem_cgroup_disabled()) |
8c7c6e34a memcg: mem+swap c... |
2243 |
return NULL; |
4077960e2 memory controller... |
2244 |
|
d13d14430 memcg: handle swa... |
2245 |
if (PageSwapCache(page)) |
8c7c6e34a memcg: mem+swap c... |
2246 |
return NULL; |
d13d14430 memcg: handle swa... |
2247 |
|
8697d3319 Memory controller... |
2248 |
/* |
3c541e14b Memory controller... |
2249 |
* Check if our page_cgroup is valid |
8697d3319 Memory controller... |
2250 |
*/ |
52d4b9ac0 memcg: allocate a... |
2251 2252 |
pc = lookup_page_cgroup(page); if (unlikely(!pc || !PageCgroupUsed(pc))) |
8c7c6e34a memcg: mem+swap c... |
2253 |
return NULL; |
b9c565d5a memcg: remove cle... |
2254 |
|
52d4b9ac0 memcg: allocate a... |
2255 |
lock_page_cgroup(pc); |
d13d14430 memcg: handle swa... |
2256 |
|
8c7c6e34a memcg: mem+swap c... |
2257 |
mem = pc->mem_cgroup; |
d13d14430 memcg: handle swa... |
2258 2259 2260 2261 2262 |
if (!PageCgroupUsed(pc)) goto unlock_out; switch (ctype) { case MEM_CGROUP_CHARGE_TYPE_MAPPED: |
8a9478ca7 memcg: fix swap a... |
2263 |
case MEM_CGROUP_CHARGE_TYPE_DROP: |
ac39cf8cb memcg: fix mis-ac... |
2264 2265 |
/* See mem_cgroup_prepare_migration() */ if (page_mapped(page) || PageCgroupMigration(pc)) |
d13d14430 memcg: handle swa... |
2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 |
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 memcg: allocate a... |
2277 |
} |
d13d14430 memcg: handle swa... |
2278 |
|
08e552c69 memcg: synchroniz... |
2279 |
mem_cgroup_charge_statistics(mem, pc, false); |
04046e1a0 memcg: use CSS ID |
2280 |
|
52d4b9ac0 memcg: allocate a... |
2281 |
ClearPageCgroupUsed(pc); |
544122e5e memcg: fix LRU ac... |
2282 2283 2284 2285 2286 2287 |
/* * 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 memcg: remove cle... |
2288 |
|
52d4b9ac0 memcg: allocate a... |
2289 |
unlock_page_cgroup(pc); |
f75ca9620 memcg: avoid css_... |
2290 2291 2292 2293 |
/* * even after unlock, we have mem->res.usage here and this memcg * will never be freed. */ |
d2265e6fa memcg : share eve... |
2294 |
memcg_check_events(mem, page); |
f75ca9620 memcg: avoid css_... |
2295 2296 2297 2298 2299 2300 |
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)) __do_uncharge(mem, ctype); |
6d12e2d8d per-zone and recl... |
2301 |
|
8c7c6e34a memcg: mem+swap c... |
2302 |
return mem; |
d13d14430 memcg: handle swa... |
2303 2304 2305 |
unlock_out: unlock_page_cgroup(pc); |
8c7c6e34a memcg: mem+swap c... |
2306 |
return NULL; |
3c541e14b Memory controller... |
2307 |
} |
69029cd55 memcg: remove ref... |
2308 2309 |
void mem_cgroup_uncharge_page(struct page *page) { |
52d4b9ac0 memcg: allocate a... |
2310 2311 2312 2313 2314 |
/* early check. */ if (page_mapped(page)) return; if (page->mapping && !PageAnon(page)) return; |
69029cd55 memcg: remove ref... |
2315 2316 2317 2318 2319 2320 |
__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 memcg: make page-... |
2321 |
VM_BUG_ON(page->mapping); |
69029cd55 memcg: remove ref... |
2322 2323 |
__mem_cgroup_uncharge_common(page, MEM_CGROUP_CHARGE_TYPE_CACHE); } |
569b846df memcg: coalesce u... |
2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 |
/* * 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 memcg: oom kill d... |
2364 |
memcg_oom_recover(batch->memcg); |
569b846df memcg: coalesce u... |
2365 2366 2367 |
/* forget this pointer (for sanity check) */ batch->memcg = NULL; } |
e767e0561 memcg: fix deadlo... |
2368 |
#ifdef CONFIG_SWAP |
8c7c6e34a memcg: mem+swap c... |
2369 |
/* |
e767e0561 memcg: fix deadlo... |
2370 |
* called after __delete_from_swap_cache() and drop "page" account. |
8c7c6e34a memcg: mem+swap c... |
2371 2372 |
* memcg information is recorded to swap_cgroup of "ent" */ |
8a9478ca7 memcg: fix swap a... |
2373 2374 |
void mem_cgroup_uncharge_swapcache(struct page *page, swp_entry_t ent, bool swapout) |
8c7c6e34a memcg: mem+swap c... |
2375 2376 |
{ struct mem_cgroup *memcg; |
8a9478ca7 memcg: fix swap a... |
2377 2378 2379 2380 2381 2382 |
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 memcg: mem+swap c... |
2383 |
|
f75ca9620 memcg: avoid css_... |
2384 2385 2386 2387 2388 |
/* * record memcg information, if swapout && memcg != NULL, * mem_cgroup_get() was called in uncharge(). */ if (do_swap_account && swapout && memcg) |
a3b2d6926 cgroups: use css ... |
2389 |
swap_cgroup_record(ent, css_id(&memcg->css)); |
8c7c6e34a memcg: mem+swap c... |
2390 |
} |
e767e0561 memcg: fix deadlo... |
2391 |
#endif |
8c7c6e34a memcg: mem+swap c... |
2392 2393 2394 2395 2396 2397 2398 |
#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 memcg: handle swa... |
2399 |
{ |
8c7c6e34a memcg: mem+swap c... |
2400 |
struct mem_cgroup *memcg; |
a3b2d6926 cgroups: use css ... |
2401 |
unsigned short id; |
8c7c6e34a memcg: mem+swap c... |
2402 2403 2404 |
if (!do_swap_account) return; |
a3b2d6926 cgroups: use css ... |
2405 2406 2407 |
id = swap_cgroup_record(ent, 0); rcu_read_lock(); memcg = mem_cgroup_lookup(id); |
8c7c6e34a memcg: mem+swap c... |
2408 |
if (memcg) { |
a3b2d6926 cgroups: use css ... |
2409 2410 2411 2412 |
/* * We uncharge this because swap is freed. * This memcg can be obsolete one. We avoid calling css_tryget */ |
0c3e73e84 memcg: improve re... |
2413 |
if (!mem_cgroup_is_root(memcg)) |
4e649152c memcg: some modif... |
2414 |
res_counter_uncharge(&memcg->memsw, PAGE_SIZE); |
0c3e73e84 memcg: improve re... |
2415 |
mem_cgroup_swap_statistics(memcg, false); |
8c7c6e34a memcg: mem+swap c... |
2416 2417 |
mem_cgroup_put(memcg); } |
a3b2d6926 cgroups: use css ... |
2418 |
rcu_read_unlock(); |
d13d14430 memcg: handle swa... |
2419 |
} |
024914477 memcg: move charg... |
2420 2421 2422 2423 2424 2425 |
/** * 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 memcg: improve pe... |
2426 |
* @need_fixup: whether we should fixup res_counters and refcounts. |
024914477 memcg: move charg... |
2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 |
* * 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 memcg: improve pe... |
2437 |
struct mem_cgroup *from, struct mem_cgroup *to, bool need_fixup) |
024914477 memcg: move charg... |
2438 2439 2440 2441 2442 2443 2444 |
{ 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 memcg: move charg... |
2445 |
mem_cgroup_swap_statistics(from, false); |
483c30b51 memcg: improve pe... |
2446 |
mem_cgroup_swap_statistics(to, true); |
024914477 memcg: move charg... |
2447 |
/* |
483c30b51 memcg: improve pe... |
2448 2449 2450 2451 2452 2453 |
* 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 memcg: move charg... |
2454 |
*/ |
024914477 memcg: move charg... |
2455 |
mem_cgroup_get(to); |
483c30b51 memcg: improve pe... |
2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 |
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 memcg: improve pe... |
2466 |
} |
024914477 memcg: move charg... |
2467 2468 2469 2470 2471 2472 |
return 0; } return -EINVAL; } #else static inline int mem_cgroup_move_swap_account(swp_entry_t entry, |
483c30b51 memcg: improve pe... |
2473 |
struct mem_cgroup *from, struct mem_cgroup *to, bool need_fixup) |
024914477 memcg: move charg... |
2474 2475 2476 |
{ return -EINVAL; } |
8c7c6e34a memcg: mem+swap c... |
2477 |
#endif |
d13d14430 memcg: handle swa... |
2478 |
|
ae41be374 bugfix for memory... |
2479 |
/* |
01b1ae63c memcg: simple mig... |
2480 2481 |
* Before starting migration, account PAGE_SIZE to mem_cgroup that the old * page belongs to. |
ae41be374 bugfix for memory... |
2482 |
*/ |
ac39cf8cb memcg: fix mis-ac... |
2483 2484 |
int mem_cgroup_prepare_migration(struct page *page, struct page *newpage, struct mem_cgroup **ptr) |
ae41be374 bugfix for memory... |
2485 2486 |
{ struct page_cgroup *pc; |
e8589cc18 memcg: better mig... |
2487 |
struct mem_cgroup *mem = NULL; |
ac39cf8cb memcg: fix mis-ac... |
2488 |
enum charge_type ctype; |
e8589cc18 memcg: better mig... |
2489 |
int ret = 0; |
8869b8f6e memcg: memcontrol... |
2490 |
|
f8d665422 memcg: add mem_cg... |
2491 |
if (mem_cgroup_disabled()) |
4077960e2 memory controller... |
2492 |
return 0; |
52d4b9ac0 memcg: allocate a... |
2493 2494 2495 |
pc = lookup_page_cgroup(page); lock_page_cgroup(pc); if (PageCgroupUsed(pc)) { |
e8589cc18 memcg: better mig... |
2496 2497 |
mem = pc->mem_cgroup; css_get(&mem->css); |
ac39cf8cb memcg: fix mis-ac... |
2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526 2527 2528 |
/* * 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 memcg: better mig... |
2529 |
} |
52d4b9ac0 memcg: allocate a... |
2530 |
unlock_page_cgroup(pc); |
ac39cf8cb memcg: fix mis-ac... |
2531 2532 2533 2534 2535 2536 |
/* * If the page is not charged at this point, * we return here. */ if (!mem) return 0; |
01b1ae63c memcg: simple mig... |
2537 |
|
93d5c9be1 memcg: fix prepar... |
2538 |
*ptr = mem; |
ac39cf8cb memcg: fix mis-ac... |
2539 2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 |
ret = __mem_cgroup_try_charge(NULL, GFP_KERNEL, ptr, false); 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 memcg: better mig... |
2552 |
} |
ac39cf8cb memcg: fix mis-ac... |
2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564 2565 2566 |
/* * 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; __mem_cgroup_commit_charge(mem, pc, ctype); |
e8589cc18 memcg: better mig... |
2567 |
return ret; |
ae41be374 bugfix for memory... |
2568 |
} |
8869b8f6e memcg: memcontrol... |
2569 |
|
69029cd55 memcg: remove ref... |
2570 |
/* remove redundant charge if migration failed*/ |
01b1ae63c memcg: simple mig... |
2571 |
void mem_cgroup_end_migration(struct mem_cgroup *mem, |
ac39cf8cb memcg: fix mis-ac... |
2572 |
struct page *oldpage, struct page *newpage) |
ae41be374 bugfix for memory... |
2573 |
{ |
ac39cf8cb memcg: fix mis-ac... |
2574 |
struct page *used, *unused; |
01b1ae63c memcg: simple mig... |
2575 |
struct page_cgroup *pc; |
01b1ae63c memcg: simple mig... |
2576 2577 2578 |
if (!mem) return; |
ac39cf8cb memcg: fix mis-ac... |
2579 |
/* blocks rmdir() */ |
887032670 cgroup avoid perm... |
2580 |
cgroup_exclude_rmdir(&mem->css); |
01b1ae63c memcg: simple mig... |
2581 2582 |
/* at migration success, oldpage->mapping is NULL. */ if (oldpage->mapping) { |
ac39cf8cb memcg: fix mis-ac... |
2583 2584 |
used = oldpage; unused = newpage; |
01b1ae63c memcg: simple mig... |
2585 |
} else { |
ac39cf8cb memcg: fix mis-ac... |
2586 |
used = newpage; |
01b1ae63c memcg: simple mig... |
2587 2588 |
unused = oldpage; } |
69029cd55 memcg: remove ref... |
2589 |
/* |
ac39cf8cb memcg: fix mis-ac... |
2590 2591 2592 |
* 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 memcg: simple mig... |
2593 |
*/ |
ac39cf8cb memcg: fix mis-ac... |
2594 2595 2596 2597 |
pc = lookup_page_cgroup(oldpage); lock_page_cgroup(pc); ClearPageCgroupMigration(pc); unlock_page_cgroup(pc); |
01b1ae63c memcg: simple mig... |
2598 |
|
ac39cf8cb memcg: fix mis-ac... |
2599 |
__mem_cgroup_uncharge_common(unused, MEM_CGROUP_CHARGE_TYPE_FORCE); |
01b1ae63c memcg: simple mig... |
2600 |
/* |
ac39cf8cb memcg: fix mis-ac... |
2601 2602 2603 2604 2605 2606 |
* 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 memcg: remove ref... |
2607 |
*/ |
ac39cf8cb memcg: fix mis-ac... |
2608 2609 |
if (PageAnon(used)) mem_cgroup_uncharge_page(used); |
887032670 cgroup avoid perm... |
2610 |
/* |
ac39cf8cb memcg: fix mis-ac... |
2611 2612 |
* At migration, we may charge account against cgroup which has no * tasks. |
887032670 cgroup avoid perm... |
2613 2614 2615 2616 |
* 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 bugfix for memory... |
2617 |
} |
78fb74669 Memory controller... |
2618 |
|
cc8475822 memory cgroup enh... |
2619 |
/* |
ae3abae64 memcg: fix mem_cg... |
2620 2621 2622 2623 2624 2625 |
* 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 memcg: helper fun... |
2626 |
*/ |
ae3abae64 memcg: fix mem_cg... |
2627 |
int mem_cgroup_shmem_charge_fallback(struct page *page, |
b5a84319a memcg: fix shmem'... |
2628 2629 |
struct mm_struct *mm, gfp_t gfp_mask) |
c9b0ed514 memcg: helper fun... |
2630 |
{ |
b5a84319a memcg: fix shmem'... |
2631 |
struct mem_cgroup *mem = NULL; |
ae3abae64 memcg: fix mem_cg... |
2632 |
int ret; |
c9b0ed514 memcg: helper fun... |
2633 |
|
f8d665422 memcg: add mem_cg... |
2634 |
if (mem_cgroup_disabled()) |
cede86acd memcg: clean up c... |
2635 |
return 0; |
c9b0ed514 memcg: helper fun... |
2636 |
|
ae3abae64 memcg: fix mem_cg... |
2637 2638 2639 |
ret = mem_cgroup_try_charge_swapin(mm, page, gfp_mask, &mem); if (!ret) mem_cgroup_cancel_charge_swapin(mem); /* it does !mem check */ |
c9b0ed514 memcg: helper fun... |
2640 |
|
ae3abae64 memcg: fix mem_cg... |
2641 |
return ret; |
c9b0ed514 memcg: helper fun... |
2642 |
} |
8c7c6e34a memcg: mem+swap c... |
2643 |
static DEFINE_MUTEX(set_limit_mutex); |
d38d2a758 mm: make mem_cgro... |
2644 |
static int mem_cgroup_resize_limit(struct mem_cgroup *memcg, |
8c7c6e34a memcg: mem+swap c... |
2645 |
unsigned long long val) |
628f42355 memcg: limit chan... |
2646 |
{ |
81d39c20f memcg: fix shrink... |
2647 |
int retry_count; |
3c11ecf44 memcg: oom kill d... |
2648 |
u64 memswlimit, memlimit; |
628f42355 memcg: limit chan... |
2649 |
int ret = 0; |
81d39c20f memcg: fix shrink... |
2650 2651 |
int children = mem_cgroup_count_children(memcg); u64 curusage, oldusage; |
3c11ecf44 memcg: oom kill d... |
2652 |
int enlarge; |
81d39c20f memcg: fix shrink... |
2653 2654 2655 2656 2657 2658 2659 2660 2661 |
/* * 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 memcg: limit chan... |
2662 |
|
3c11ecf44 memcg: oom kill d... |
2663 |
enlarge = 0; |
8c7c6e34a memcg: mem+swap c... |
2664 |
while (retry_count) { |
628f42355 memcg: limit chan... |
2665 2666 2667 2668 |
if (signal_pending(current)) { ret = -EINTR; break; } |
8c7c6e34a memcg: mem+swap c... |
2669 2670 2671 2672 2673 2674 2675 2676 2677 2678 |
/* * 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 memcg: limit chan... |
2679 2680 |
break; } |
3c11ecf44 memcg: oom kill d... |
2681 2682 2683 2684 |
memlimit = res_counter_read_u64(&memcg->res, RES_LIMIT); if (memlimit < val) enlarge = 1; |
8c7c6e34a memcg: mem+swap c... |
2685 |
ret = res_counter_set_limit(&memcg->res, val); |
22a668d7c memcg: fix behavi... |
2686 2687 2688 2689 2690 2691 |
if (!ret) { if (memswlimit == val) memcg->memsw_is_minimum = true; else memcg->memsw_is_minimum = false; } |
8c7c6e34a memcg: mem+swap c... |
2692 2693 2694 2695 |
mutex_unlock(&set_limit_mutex); if (!ret) break; |
aa20d489c memcg: code clean... |
2696 |
mem_cgroup_hierarchical_reclaim(memcg, NULL, GFP_KERNEL, |
4e4169535 memory controller... |
2697 |
MEM_CGROUP_RECLAIM_SHRINK); |
81d39c20f memcg: fix shrink... |
2698 2699 2700 2701 2702 2703 |
curusage = res_counter_read_u64(&memcg->res, RES_USAGE); /* Usage is reduced ? */ if (curusage >= oldusage) retry_count--; else oldusage = curusage; |
8c7c6e34a memcg: mem+swap c... |
2704 |
} |
3c11ecf44 memcg: oom kill d... |
2705 2706 |
if (!ret && enlarge) memcg_oom_recover(memcg); |
14797e236 memcg: add inacti... |
2707 |
|
8c7c6e34a memcg: mem+swap c... |
2708 2709 |
return ret; } |
338c84310 memcg: remove som... |
2710 2711 |
static int mem_cgroup_resize_memsw_limit(struct mem_cgroup *memcg, unsigned long long val) |
8c7c6e34a memcg: mem+swap c... |
2712 |
{ |
81d39c20f memcg: fix shrink... |
2713 |
int retry_count; |
3c11ecf44 memcg: oom kill d... |
2714 |
u64 memlimit, memswlimit, oldusage, curusage; |
81d39c20f memcg: fix shrink... |
2715 2716 |
int children = mem_cgroup_count_children(memcg); int ret = -EBUSY; |
3c11ecf44 memcg: oom kill d... |
2717 |
int enlarge = 0; |
8c7c6e34a memcg: mem+swap c... |
2718 |
|
81d39c20f memcg: fix shrink... |
2719 2720 2721 |
/* see mem_cgroup_resize_res_limit */ retry_count = children * MEM_CGROUP_RECLAIM_RETRIES; oldusage = res_counter_read_u64(&memcg->memsw, RES_USAGE); |
8c7c6e34a memcg: mem+swap c... |
2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 |
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 memcg: oom kill d... |
2739 2740 2741 |
memswlimit = res_counter_read_u64(&memcg->memsw, RES_LIMIT); if (memswlimit < val) enlarge = 1; |
8c7c6e34a memcg: mem+swap c... |
2742 |
ret = res_counter_set_limit(&memcg->memsw, val); |
22a668d7c memcg: fix behavi... |
2743 2744 2745 2746 2747 2748 |
if (!ret) { if (memlimit == val) memcg->memsw_is_minimum = true; else memcg->memsw_is_minimum = false; } |
8c7c6e34a memcg: mem+swap c... |
2749 2750 2751 2752 |
mutex_unlock(&set_limit_mutex); if (!ret) break; |
4e4169535 memory controller... |
2753 |
mem_cgroup_hierarchical_reclaim(memcg, NULL, GFP_KERNEL, |
75822b449 memory controller... |
2754 2755 |
MEM_CGROUP_RECLAIM_NOSWAP | MEM_CGROUP_RECLAIM_SHRINK); |
8c7c6e34a memcg: mem+swap c... |
2756 |
curusage = res_counter_read_u64(&memcg->memsw, RES_USAGE); |
81d39c20f memcg: fix shrink... |
2757 |
/* Usage is reduced ? */ |
8c7c6e34a memcg: mem+swap c... |
2758 |
if (curusage >= oldusage) |
628f42355 memcg: limit chan... |
2759 |
retry_count--; |
81d39c20f memcg: fix shrink... |
2760 2761 |
else oldusage = curusage; |
628f42355 memcg: limit chan... |
2762 |
} |
3c11ecf44 memcg: oom kill d... |
2763 2764 |
if (!ret && enlarge) memcg_oom_recover(memcg); |
628f42355 memcg: limit chan... |
2765 2766 |
return ret; } |
4e4169535 memory controller... |
2767 |
unsigned long mem_cgroup_soft_limit_reclaim(struct zone *zone, int order, |
00918b6ab memcg: remove nid... |
2768 |
gfp_t gfp_mask) |
4e4169535 memory controller... |
2769 2770 2771 2772 2773 2774 |
{ 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 memcg: reduce che... |
2775 |
unsigned long long excess; |
4e4169535 memory controller... |
2776 2777 2778 |
if (order > 0) return 0; |
00918b6ab memcg: remove nid... |
2779 |
mctz = soft_limit_tree_node_zone(zone_to_nid(zone), zone_idx(zone)); |
4e4169535 memory controller... |
2780 2781 2782 2783 2784 2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 2795 2796 2797 2798 2799 2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811 2812 2813 2814 2815 2816 2817 2818 2819 2820 2821 2822 2823 2824 2825 |
/* * 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 memory controller... |
2826 |
__mem_cgroup_remove_exceeded(mz->mem, mz, mctz); |
ef8745c1e memcg: reduce che... |
2827 |
excess = res_counter_soft_limit_excess(&mz->mem->res); |
4e4169535 memory controller... |
2828 2829 2830 2831 2832 2833 2834 2835 |
/* * 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 memcg: reduce che... |
2836 2837 |
/* If excess == 0, no tree ops */ __mem_cgroup_insert_exceeded(mz->mem, mz, mctz, excess); |
4e4169535 memory controller... |
2838 2839 2840 2841 2842 2843 2844 2845 2846 2847 2848 2849 2850 2851 2852 2853 2854 |
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 memcg: helper fun... |
2855 |
/* |
cc8475822 memory cgroup enh... |
2856 |
* This routine traverse page_cgroup in given list and drop them all. |
cc8475822 memory cgroup enh... |
2857 2858 |
* *And* this routine doesn't reclaim page itself, just removes page_cgroup. */ |
f817ed485 memcg: move all a... |
2859 |
static int mem_cgroup_force_empty_list(struct mem_cgroup *mem, |
08e552c69 memcg: synchroniz... |
2860 |
int node, int zid, enum lru_list lru) |
cc8475822 memory cgroup enh... |
2861 |
{ |
08e552c69 memcg: synchroniz... |
2862 2863 |
struct zone *zone; struct mem_cgroup_per_zone *mz; |
f817ed485 memcg: move all a... |
2864 |
struct page_cgroup *pc, *busy; |
08e552c69 memcg: synchroniz... |
2865 |
unsigned long flags, loop; |
072c56c13 per-zone and recl... |
2866 |
struct list_head *list; |
f817ed485 memcg: move all a... |
2867 |
int ret = 0; |
072c56c13 per-zone and recl... |
2868 |
|
08e552c69 memcg: synchroniz... |
2869 2870 |
zone = &NODE_DATA(node)->node_zones[zid]; mz = mem_cgroup_zoneinfo(mem, node, zid); |
b69408e88 vmscan: Use an in... |
2871 |
list = &mz->lists[lru]; |
cc8475822 memory cgroup enh... |
2872 |
|
f817ed485 memcg: move all a... |
2873 2874 2875 2876 2877 2878 |
loop = MEM_CGROUP_ZSTAT(mz, lru); /* give some margin against EBUSY etc...*/ loop += 256; busy = NULL; while (loop--) { ret = 0; |
08e552c69 memcg: synchroniz... |
2879 |
spin_lock_irqsave(&zone->lru_lock, flags); |
f817ed485 memcg: move all a... |
2880 |
if (list_empty(list)) { |
08e552c69 memcg: synchroniz... |
2881 |
spin_unlock_irqrestore(&zone->lru_lock, flags); |
52d4b9ac0 memcg: allocate a... |
2882 |
break; |
f817ed485 memcg: move all a... |
2883 2884 2885 2886 |
} pc = list_entry(list->prev, struct page_cgroup, lru); if (busy == pc) { list_move(&pc->lru, list); |
648bcc771 mm/memcontrol.c: ... |
2887 |
busy = NULL; |
08e552c69 memcg: synchroniz... |
2888 |
spin_unlock_irqrestore(&zone->lru_lock, flags); |
f817ed485 memcg: move all a... |
2889 2890 |
continue; } |
08e552c69 memcg: synchroniz... |
2891 |
spin_unlock_irqrestore(&zone->lru_lock, flags); |
f817ed485 memcg: move all a... |
2892 |
|
2c26fdd70 memcg: revert gfp... |
2893 |
ret = mem_cgroup_move_parent(pc, mem, GFP_KERNEL); |
f817ed485 memcg: move all a... |
2894 |
if (ret == -ENOMEM) |
52d4b9ac0 memcg: allocate a... |
2895 |
break; |
f817ed485 memcg: move all a... |
2896 2897 2898 2899 2900 2901 2902 |
if (ret == -EBUSY || ret == -EINVAL) { /* found lock contention or "pc" is obsolete. */ busy = pc; cond_resched(); } else busy = NULL; |
cc8475822 memory cgroup enh... |
2903 |
} |
08e552c69 memcg: synchroniz... |
2904 |
|
f817ed485 memcg: move all a... |
2905 2906 2907 |
if (!ret && !list_empty(list)) return -EBUSY; return ret; |
cc8475822 memory cgroup enh... |
2908 2909 2910 2911 2912 2913 |
} /* * make mem_cgroup's charge to be 0 if there is no task. * This enables deleting this mem_cgroup. */ |
c1e862c1f memcg: new force_... |
2914 |
static int mem_cgroup_force_empty(struct mem_cgroup *mem, bool free_all) |
cc8475822 memory cgroup enh... |
2915 |
{ |
f817ed485 memcg: move all a... |
2916 2917 2918 |
int ret; int node, zid, shrink; int nr_retries = MEM_CGROUP_RECLAIM_RETRIES; |
c1e862c1f memcg: new force_... |
2919 |
struct cgroup *cgrp = mem->css.cgroup; |
8869b8f6e memcg: memcontrol... |
2920 |
|
cc8475822 memory cgroup enh... |
2921 |
css_get(&mem->css); |
f817ed485 memcg: move all a... |
2922 2923 |
shrink = 0; |
c1e862c1f memcg: new force_... |
2924 2925 2926 |
/* should free all ? */ if (free_all) goto try_to_free; |
f817ed485 memcg: move all a... |
2927 |
move_account: |
fce664775 memcg: ensure lis... |
2928 |
do { |
f817ed485 memcg: move all a... |
2929 |
ret = -EBUSY; |
c1e862c1f memcg: new force_... |
2930 2931 2932 2933 |
if (cgroup_task_count(cgrp) || !list_empty(&cgrp->children)) goto out; ret = -EINTR; if (signal_pending(current)) |
cc8475822 memory cgroup enh... |
2934 |
goto out; |
52d4b9ac0 memcg: allocate a... |
2935 2936 |
/* This is for making all *used* pages to be on LRU. */ lru_add_drain_all(); |
cdec2e426 memcg: coalesce c... |
2937 |
drain_all_stock_sync(); |
f817ed485 memcg: move all a... |
2938 |
ret = 0; |
299b4eaa3 memcg: NULL point... |
2939 |
for_each_node_state(node, N_HIGH_MEMORY) { |
f817ed485 memcg: move all a... |
2940 |
for (zid = 0; !ret && zid < MAX_NR_ZONES; zid++) { |
b69408e88 vmscan: Use an in... |
2941 |
enum lru_list l; |
f817ed485 memcg: move all a... |
2942 2943 |
for_each_lru(l) { ret = mem_cgroup_force_empty_list(mem, |
08e552c69 memcg: synchroniz... |
2944 |
node, zid, l); |
f817ed485 memcg: move all a... |
2945 2946 2947 |
if (ret) break; } |
1ecaab2bd per-zone and recl... |
2948 |
} |
f817ed485 memcg: move all a... |
2949 2950 2951 |
if (ret) break; } |
3c11ecf44 memcg: oom kill d... |
2952 |
memcg_oom_recover(mem); |
f817ed485 memcg: move all a... |
2953 2954 2955 |
/* it seems parent cgroup doesn't have enough mem */ if (ret == -ENOMEM) goto try_to_free; |
52d4b9ac0 memcg: allocate a... |
2956 |
cond_resched(); |
fce664775 memcg: ensure lis... |
2957 2958 |
/* "ret" should also be checked to ensure all lists are empty. */ } while (mem->res.usage > 0 || ret); |
cc8475822 memory cgroup enh... |
2959 2960 2961 |
out: css_put(&mem->css); return ret; |
f817ed485 memcg: move all a... |
2962 2963 |
try_to_free: |
c1e862c1f memcg: new force_... |
2964 2965 |
/* returns EBUSY if there is a task or if we come here twice. */ if (cgroup_task_count(cgrp) || !list_empty(&cgrp->children) || shrink) { |
f817ed485 memcg: move all a... |
2966 2967 2968 |
ret = -EBUSY; goto out; } |
c1e862c1f memcg: new force_... |
2969 2970 |
/* we call try-to-free pages for make this cgroup empty */ lru_add_drain_all(); |
f817ed485 memcg: move all a... |
2971 2972 2973 2974 |
/* try to free all pages in this cgroup */ shrink = 1; while (nr_retries && mem->res.usage > 0) { int progress; |
c1e862c1f memcg: new force_... |
2975 2976 2977 2978 2979 |
if (signal_pending(current)) { ret = -EINTR; goto out; } |
a7885eb8a memcg: swappiness |
2980 2981 |
progress = try_to_free_mem_cgroup_pages(mem, GFP_KERNEL, false, get_swappiness(mem)); |
c1e862c1f memcg: new force_... |
2982 |
if (!progress) { |
f817ed485 memcg: move all a... |
2983 |
nr_retries--; |
c1e862c1f memcg: new force_... |
2984 |
/* maybe some writeback is necessary */ |
8aa7e847d Fix congestion_wa... |
2985 |
congestion_wait(BLK_RW_ASYNC, HZ/10); |
c1e862c1f memcg: new force_... |
2986 |
} |
f817ed485 memcg: move all a... |
2987 2988 |
} |
08e552c69 memcg: synchroniz... |
2989 |
lru_add_drain(); |
f817ed485 memcg: move all a... |
2990 |
/* try move_account...there may be some *locked* pages. */ |
fce664775 memcg: ensure lis... |
2991 |
goto move_account; |
cc8475822 memory cgroup enh... |
2992 |
} |
c1e862c1f memcg: new force_... |
2993 2994 2995 2996 |
int mem_cgroup_force_empty_write(struct cgroup *cont, unsigned int event) { return mem_cgroup_force_empty(mem_cgroup_from_cont(cont), true); } |
18f59ea7d memcg: memory cgr... |
2997 2998 2999 3000 3001 3002 3003 3004 3005 3006 3007 3008 3009 3010 3011 3012 3013 3014 |
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 tree-wide: fix as... |
3015 |
* If parent's use_hierarchy is set, we can't make any modifications |
18f59ea7d memcg: memory cgr... |
3016 3017 3018 3019 3020 3021 3022 3023 3024 3025 3026 3027 3028 3029 3030 3031 3032 3033 |
* 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 memcg: improve re... |
3034 3035 3036 3037 3038 3039 3040 3041 3042 |
struct mem_cgroup_idx_data { s64 val; enum mem_cgroup_stat_index idx; }; static int mem_cgroup_get_idx_stat(struct mem_cgroup *mem, void *data) { struct mem_cgroup_idx_data *d = data; |
c62b1a3b3 memcg: use generi... |
3043 |
d->val += mem_cgroup_read_stat(mem, d->idx); |
0c3e73e84 memcg: improve re... |
3044 3045 3046 3047 3048 3049 3050 3051 3052 3053 3054 3055 3056 |
return 0; } static void mem_cgroup_get_recursive_idx_stat(struct mem_cgroup *mem, enum mem_cgroup_stat_index idx, s64 *val) { struct mem_cgroup_idx_data d; d.idx = idx; d.val = 0; mem_cgroup_walk_tree(mem, &d, mem_cgroup_get_idx_stat); *val = d.val; } |
104f39284 memcg: extract me... |
3057 3058 3059 3060 3061 3062 3063 3064 3065 3066 3067 3068 3069 3070 3071 3072 3073 3074 3075 3076 3077 3078 3079 3080 |
static inline u64 mem_cgroup_usage(struct mem_cgroup *mem, bool swap) { u64 idx_val, val; 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); } mem_cgroup_get_recursive_idx_stat(mem, MEM_CGROUP_STAT_CACHE, &idx_val); val = idx_val; mem_cgroup_get_recursive_idx_stat(mem, MEM_CGROUP_STAT_RSS, &idx_val); val += idx_val; if (swap) { mem_cgroup_get_recursive_idx_stat(mem, MEM_CGROUP_STAT_SWAPOUT, &idx_val); val += idx_val; } return val << PAGE_SHIFT; } |
2c3daa722 CGroup API files:... |
3081 |
static u64 mem_cgroup_read(struct cgroup *cont, struct cftype *cft) |
8cdea7c05 Memory controller... |
3082 |
{ |
8c7c6e34a memcg: mem+swap c... |
3083 |
struct mem_cgroup *mem = mem_cgroup_from_cont(cont); |
104f39284 memcg: extract me... |
3084 |
u64 val; |
8c7c6e34a memcg: mem+swap c... |
3085 3086 3087 3088 3089 3090 |
int type, name; type = MEMFILE_TYPE(cft->private); name = MEMFILE_ATTR(cft->private); switch (type) { case _MEM: |
104f39284 memcg: extract me... |
3091 3092 3093 |
if (name == RES_USAGE) val = mem_cgroup_usage(mem, false); else |
0c3e73e84 memcg: improve re... |
3094 |
val = res_counter_read_u64(&mem->res, name); |
8c7c6e34a memcg: mem+swap c... |
3095 3096 |
break; case _MEMSWAP: |
104f39284 memcg: extract me... |
3097 3098 3099 |
if (name == RES_USAGE) val = mem_cgroup_usage(mem, true); else |
0c3e73e84 memcg: improve re... |
3100 |
val = res_counter_read_u64(&mem->memsw, name); |
8c7c6e34a memcg: mem+swap c... |
3101 3102 3103 3104 3105 3106 |
break; default: BUG(); break; } return val; |
8cdea7c05 Memory controller... |
3107 |
} |
628f42355 memcg: limit chan... |
3108 3109 3110 3111 |
/* * The user of this function is... * RES_LIMIT. */ |
856c13aa1 cgroup files: con... |
3112 3113 |
static int mem_cgroup_write(struct cgroup *cont, struct cftype *cft, const char *buffer) |
8cdea7c05 Memory controller... |
3114 |
{ |
628f42355 memcg: limit chan... |
3115 |
struct mem_cgroup *memcg = mem_cgroup_from_cont(cont); |
8c7c6e34a memcg: mem+swap c... |
3116 |
int type, name; |
628f42355 memcg: limit chan... |
3117 3118 |
unsigned long long val; int ret; |
8c7c6e34a memcg: mem+swap c... |
3119 3120 3121 |
type = MEMFILE_TYPE(cft->private); name = MEMFILE_ATTR(cft->private); switch (name) { |
628f42355 memcg: limit chan... |
3122 |
case RES_LIMIT: |
4b3bde4c9 memcg: remove the... |
3123 3124 3125 3126 |
if (mem_cgroup_is_root(memcg)) { /* Can't set limit on root */ ret = -EINVAL; break; } |
628f42355 memcg: limit chan... |
3127 3128 |
/* This function does all necessary parse...reuse it */ ret = res_counter_memparse_write_strategy(buffer, &val); |
8c7c6e34a memcg: mem+swap c... |
3129 3130 3131 |
if (ret) break; if (type == _MEM) |
628f42355 memcg: limit chan... |
3132 |
ret = mem_cgroup_resize_limit(memcg, val); |
8c7c6e34a memcg: mem+swap c... |
3133 3134 |
else ret = mem_cgroup_resize_memsw_limit(memcg, val); |
628f42355 memcg: limit chan... |
3135 |
break; |
296c81d89 memory controller... |
3136 3137 3138 3139 3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 |
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 memcg: limit chan... |
3150 3151 3152 3153 3154 |
default: ret = -EINVAL; /* should be BUG() ? */ break; } return ret; |
8cdea7c05 Memory controller... |
3155 |
} |
fee7b548e memcg: show real ... |
3156 3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 3167 3168 3169 3170 3171 3172 3173 3174 3175 3176 3177 3178 3179 3180 3181 3182 |
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 memcgroup: implem... |
3183 |
static int mem_cgroup_reset(struct cgroup *cont, unsigned int event) |
c84872e16 memcgroup: add th... |
3184 3185 |
{ struct mem_cgroup *mem; |
8c7c6e34a memcg: mem+swap c... |
3186 |
int type, name; |
c84872e16 memcgroup: add th... |
3187 3188 |
mem = mem_cgroup_from_cont(cont); |
8c7c6e34a memcg: mem+swap c... |
3189 3190 3191 |
type = MEMFILE_TYPE(event); name = MEMFILE_ATTR(event); switch (name) { |
29f2a4dac memcgroup: implem... |
3192 |
case RES_MAX_USAGE: |
8c7c6e34a memcg: mem+swap c... |
3193 3194 3195 3196 |
if (type == _MEM) res_counter_reset_max(&mem->res); else res_counter_reset_max(&mem->memsw); |
29f2a4dac memcgroup: implem... |
3197 3198 |
break; case RES_FAILCNT: |
8c7c6e34a memcg: mem+swap c... |
3199 3200 3201 3202 |
if (type == _MEM) res_counter_reset_failcnt(&mem->res); else res_counter_reset_failcnt(&mem->memsw); |
29f2a4dac memcgroup: implem... |
3203 3204 |
break; } |
f64c3f549 memory controller... |
3205 |
|
85cc59db1 memcgroup: use tr... |
3206 |
return 0; |
c84872e16 memcgroup: add th... |
3207 |
} |
7dc74be03 memcg: add interf... |
3208 3209 3210 3211 3212 |
static u64 mem_cgroup_move_charge_read(struct cgroup *cgrp, struct cftype *cft) { return mem_cgroup_from_cont(cgrp)->move_charge_at_immigrate; } |
024914477 memcg: move charg... |
3213 |
#ifdef CONFIG_MMU |
7dc74be03 memcg: add interf... |
3214 3215 3216 3217 3218 3219 3220 3221 3222 3223 3224 3225 3226 3227 3228 3229 3230 3231 |
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 memcg: move charg... |
3232 3233 3234 3235 3236 3237 3238 |
#else static int mem_cgroup_move_charge_write(struct cgroup *cgrp, struct cftype *cft, u64 val) { return -ENOSYS; } #endif |
7dc74be03 memcg: add interf... |
3239 |
|
14067bb3e memcg: hierarchic... |
3240 3241 3242 3243 3244 |
/* For read statistics */ enum { MCS_CACHE, MCS_RSS, |
d8046582d memcg: make memcg... |
3245 |
MCS_FILE_MAPPED, |
14067bb3e memcg: hierarchic... |
3246 3247 |
MCS_PGPGIN, MCS_PGPGOUT, |
1dd3a2732 memcg: show swap ... |
3248 |
MCS_SWAP, |
14067bb3e memcg: hierarchic... |
3249 3250 3251 3252 3253 3254 3255 3256 3257 3258 |
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 memory cgroup enh... |
3259 |
}; |
14067bb3e memcg: hierarchic... |
3260 3261 3262 3263 3264 3265 |
struct { char *local_name; char *total_name; } memcg_stat_strings[NR_MCS_STAT] = { {"cache", "total_cache"}, {"rss", "total_rss"}, |
d69b042f3 memcg: add file-b... |
3266 |
{"mapped_file", "total_mapped_file"}, |
14067bb3e memcg: hierarchic... |
3267 3268 |
{"pgpgin", "total_pgpgin"}, {"pgpgout", "total_pgpgout"}, |
1dd3a2732 memcg: show swap ... |
3269 |
{"swap", "total_swap"}, |
14067bb3e memcg: hierarchic... |
3270 3271 3272 3273 3274 3275 3276 3277 3278 3279 3280 3281 3282 3283 |
{"inactive_anon", "total_inactive_anon"}, {"active_anon", "total_active_anon"}, {"inactive_file", "total_inactive_file"}, {"active_file", "total_active_file"}, {"unevictable", "total_unevictable"} }; static int mem_cgroup_get_local_stat(struct mem_cgroup *mem, void *data) { struct mcs_total_stat *s = data; s64 val; /* per cpu stat */ |
c62b1a3b3 memcg: use generi... |
3284 |
val = mem_cgroup_read_stat(mem, MEM_CGROUP_STAT_CACHE); |
14067bb3e memcg: hierarchic... |
3285 |
s->stat[MCS_CACHE] += val * PAGE_SIZE; |
c62b1a3b3 memcg: use generi... |
3286 |
val = mem_cgroup_read_stat(mem, MEM_CGROUP_STAT_RSS); |
14067bb3e memcg: hierarchic... |
3287 |
s->stat[MCS_RSS] += val * PAGE_SIZE; |
c62b1a3b3 memcg: use generi... |
3288 |
val = mem_cgroup_read_stat(mem, MEM_CGROUP_STAT_FILE_MAPPED); |
d8046582d memcg: make memcg... |
3289 |
s->stat[MCS_FILE_MAPPED] += val * PAGE_SIZE; |
c62b1a3b3 memcg: use generi... |
3290 |
val = mem_cgroup_read_stat(mem, MEM_CGROUP_STAT_PGPGIN_COUNT); |
14067bb3e memcg: hierarchic... |
3291 |
s->stat[MCS_PGPGIN] += val; |
c62b1a3b3 memcg: use generi... |
3292 |
val = mem_cgroup_read_stat(mem, MEM_CGROUP_STAT_PGPGOUT_COUNT); |
14067bb3e memcg: hierarchic... |
3293 |
s->stat[MCS_PGPGOUT] += val; |
1dd3a2732 memcg: show swap ... |
3294 |
if (do_swap_account) { |
c62b1a3b3 memcg: use generi... |
3295 |
val = mem_cgroup_read_stat(mem, MEM_CGROUP_STAT_SWAPOUT); |
1dd3a2732 memcg: show swap ... |
3296 3297 |
s->stat[MCS_SWAP] += val * PAGE_SIZE; } |
14067bb3e memcg: hierarchic... |
3298 3299 3300 3301 3302 3303 3304 3305 3306 3307 3308 3309 3310 3311 3312 3313 3314 3315 3316 3317 |
/* 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; return 0; } static void mem_cgroup_get_total_stat(struct mem_cgroup *mem, struct mcs_total_stat *s) { mem_cgroup_walk_tree(mem, s, mem_cgroup_get_local_stat); } |
c64745cf0 CGroup API files:... |
3318 3319 |
static int mem_control_stat_show(struct cgroup *cont, struct cftype *cft, struct cgroup_map_cb *cb) |
d2ceb9b7d memory cgroup enh... |
3320 |
{ |
d2ceb9b7d memory cgroup enh... |
3321 |
struct mem_cgroup *mem_cont = mem_cgroup_from_cont(cont); |
14067bb3e memcg: hierarchic... |
3322 |
struct mcs_total_stat mystat; |
d2ceb9b7d memory cgroup enh... |
3323 |
int i; |
14067bb3e memcg: hierarchic... |
3324 3325 |
memset(&mystat, 0, sizeof(mystat)); mem_cgroup_get_local_stat(mem_cont, &mystat); |
d2ceb9b7d memory cgroup enh... |
3326 |
|
1dd3a2732 memcg: show swap ... |
3327 3328 3329 |
for (i = 0; i < NR_MCS_STAT; i++) { if (i == MCS_SWAP && !do_swap_account) continue; |
14067bb3e memcg: hierarchic... |
3330 |
cb->fill(cb, memcg_stat_strings[i].local_name, mystat.stat[i]); |
1dd3a2732 memcg: show swap ... |
3331 |
} |
7b854121e Unevictable LRU P... |
3332 |
|
14067bb3e memcg: hierarchic... |
3333 |
/* Hierarchical information */ |
fee7b548e memcg: show real ... |
3334 3335 3336 3337 3338 3339 3340 |
{ 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 memcg: show recla... |
3341 |
|
14067bb3e memcg: hierarchic... |
3342 3343 |
memset(&mystat, 0, sizeof(mystat)); mem_cgroup_get_total_stat(mem_cont, &mystat); |
1dd3a2732 memcg: show swap ... |
3344 3345 3346 |
for (i = 0; i < NR_MCS_STAT; i++) { if (i == MCS_SWAP && !do_swap_account) continue; |
14067bb3e memcg: hierarchic... |
3347 |
cb->fill(cb, memcg_stat_strings[i].total_name, mystat.stat[i]); |
1dd3a2732 memcg: show swap ... |
3348 |
} |
14067bb3e memcg: hierarchic... |
3349 |
|
7f016ee8b memcg: show recla... |
3350 |
#ifdef CONFIG_DEBUG_VM |
c772be939 memcg: fix calcul... |
3351 |
cb->fill(cb, "inactive_ratio", calc_inactive_ratio(mem_cont, NULL)); |
7f016ee8b memcg: show recla... |
3352 3353 3354 3355 3356 3357 3358 3359 3360 3361 3362 3363 3364 3365 3366 3367 3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 |
{ 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 memory cgroup enh... |
3378 3379 |
return 0; } |
a7885eb8a memcg: swappiness |
3380 3381 3382 3383 3384 3385 3386 3387 3388 3389 3390 3391 |
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 memcg: fix a race... |
3392 |
|
a7885eb8a memcg: swappiness |
3393 3394 3395 3396 3397 3398 3399 |
if (val > 100) return -EINVAL; if (cgrp->parent == NULL) return -EINVAL; parent = mem_cgroup_from_cont(cgrp->parent); |
068b38c1f memcg: fix a race... |
3400 3401 |
cgroup_lock(); |
a7885eb8a memcg: swappiness |
3402 3403 |
/* If under hierarchy, only empty-root can set this value */ if ((parent->use_hierarchy) || |
068b38c1f memcg: fix a race... |
3404 3405 |
(memcg->use_hierarchy && !list_empty(&cgrp->children))) { cgroup_unlock(); |
a7885eb8a memcg: swappiness |
3406 |
return -EINVAL; |
068b38c1f memcg: fix a race... |
3407 |
} |
a7885eb8a memcg: swappiness |
3408 3409 3410 3411 |
spin_lock(&memcg->reclaim_param_lock); memcg->swappiness = val; spin_unlock(&memcg->reclaim_param_lock); |
068b38c1f memcg: fix a race... |
3412 |
cgroup_unlock(); |
a7885eb8a memcg: swappiness |
3413 3414 |
return 0; } |
2e72b6347 memcg: implement ... |
3415 3416 3417 3418 3419 3420 3421 3422 |
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 memcg: clean up m... |
3423 |
t = rcu_dereference(memcg->thresholds.primary); |
2e72b6347 memcg: implement ... |
3424 |
else |
2c488db27 memcg: clean up m... |
3425 |
t = rcu_dereference(memcg->memsw_thresholds.primary); |
2e72b6347 memcg: implement ... |
3426 3427 3428 3429 3430 3431 3432 3433 3434 3435 3436 |
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 mm: remove unnece... |
3437 |
i = t->current_threshold; |
2e72b6347 memcg: implement ... |
3438 3439 3440 3441 3442 3443 3444 3445 3446 3447 3448 3449 3450 3451 3452 3453 3454 3455 3456 3457 3458 3459 3460 |
/* * 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 mm: remove unnece... |
3461 |
t->current_threshold = i - 1; |
2e72b6347 memcg: implement ... |
3462 3463 3464 3465 3466 3467 3468 3469 3470 3471 3472 3473 3474 3475 3476 3477 3478 3479 |
unlock: rcu_read_unlock(); } static void mem_cgroup_threshold(struct mem_cgroup *memcg) { __mem_cgroup_threshold(memcg, false); if (do_swap_account) __mem_cgroup_threshold(memcg, true); } 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; } |
9490ff275 memcg: oom notifier |
3480 3481 3482 3483 3484 3485 3486 3487 3488 3489 3490 3491 3492 3493 3494 3495 |
static int mem_cgroup_oom_notify_cb(struct mem_cgroup *mem, void *data) { 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) { mem_cgroup_walk_tree(mem, NULL, mem_cgroup_oom_notify_cb); } static int mem_cgroup_usage_register_event(struct cgroup *cgrp, struct cftype *cft, struct eventfd_ctx *eventfd, const char *args) |
2e72b6347 memcg: implement ... |
3496 3497 |
{ struct mem_cgroup *memcg = mem_cgroup_from_cont(cgrp); |
2c488db27 memcg: clean up m... |
3498 3499 |
struct mem_cgroup_thresholds *thresholds; struct mem_cgroup_threshold_ary *new; |
2e72b6347 memcg: implement ... |
3500 3501 |
int type = MEMFILE_TYPE(cft->private); u64 threshold, usage; |
2c488db27 memcg: clean up m... |
3502 |
int i, size, ret; |
2e72b6347 memcg: implement ... |
3503 3504 3505 3506 3507 3508 |
ret = res_counter_memparse_write_strategy(args, &threshold); if (ret) return ret; mutex_lock(&memcg->thresholds_lock); |
2c488db27 memcg: clean up m... |
3509 |
|
2e72b6347 memcg: implement ... |
3510 |
if (type == _MEM) |
2c488db27 memcg: clean up m... |
3511 |
thresholds = &memcg->thresholds; |
2e72b6347 memcg: implement ... |
3512 |
else if (type == _MEMSWAP) |
2c488db27 memcg: clean up m... |
3513 |
thresholds = &memcg->memsw_thresholds; |
2e72b6347 memcg: implement ... |
3514 3515 3516 3517 3518 3519 |
else BUG(); usage = mem_cgroup_usage(memcg, type == _MEMSWAP); /* Check if a threshold crossed before adding a new one */ |
2c488db27 memcg: clean up m... |
3520 |
if (thresholds->primary) |
2e72b6347 memcg: implement ... |
3521 |
__mem_cgroup_threshold(memcg, type == _MEMSWAP); |
2c488db27 memcg: clean up m... |
3522 |
size = thresholds->primary ? thresholds->primary->size + 1 : 1; |
2e72b6347 memcg: implement ... |
3523 3524 |
/* Allocate memory for new array of thresholds */ |
2c488db27 memcg: clean up m... |
3525 |
new = kmalloc(sizeof(*new) + size * sizeof(struct mem_cgroup_threshold), |
2e72b6347 memcg: implement ... |
3526 |
GFP_KERNEL); |
2c488db27 memcg: clean up m... |
3527 |
if (!new) { |
2e72b6347 memcg: implement ... |
3528 3529 3530 |
ret = -ENOMEM; goto unlock; } |
2c488db27 memcg: clean up m... |
3531 |
new->size = size; |
2e72b6347 memcg: implement ... |
3532 3533 |
/* Copy thresholds (if any) to new array */ |
2c488db27 memcg: clean up m... |
3534 3535 |
if (thresholds->primary) { memcpy(new->entries, thresholds->primary->entries, (size - 1) * |
2e72b6347 memcg: implement ... |
3536 |
sizeof(struct mem_cgroup_threshold)); |
2c488db27 memcg: clean up m... |
3537 |
} |
2e72b6347 memcg: implement ... |
3538 |
/* Add new threshold */ |
2c488db27 memcg: clean up m... |
3539 3540 |
new->entries[size - 1].eventfd = eventfd; new->entries[size - 1].threshold = threshold; |
2e72b6347 memcg: implement ... |
3541 3542 |
/* Sort thresholds. Registering of new threshold isn't time-critical */ |
2c488db27 memcg: clean up m... |
3543 |
sort(new->entries, size, sizeof(struct mem_cgroup_threshold), |
2e72b6347 memcg: implement ... |
3544 3545 3546 |
compare_thresholds, NULL); /* Find current threshold */ |
2c488db27 memcg: clean up m... |
3547 |
new->current_threshold = -1; |
2e72b6347 memcg: implement ... |
3548 |
for (i = 0; i < size; i++) { |
2c488db27 memcg: clean up m... |
3549 |
if (new->entries[i].threshold < usage) { |
2e72b6347 memcg: implement ... |
3550 |
/* |
2c488db27 memcg: clean up m... |
3551 3552 |
* new->current_threshold will not be used until * rcu_assign_pointer(), so it's safe to increment |
2e72b6347 memcg: implement ... |
3553 3554 |
* it here. */ |
2c488db27 memcg: clean up m... |
3555 |
++new->current_threshold; |
2e72b6347 memcg: implement ... |
3556 3557 |
} } |
2c488db27 memcg: clean up m... |
3558 3559 3560 3561 3562 |
/* 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 memcg: implement ... |
3563 |
|
907860ed3 cgroups: make cft... |
3564 |
/* To be sure that nobody uses thresholds */ |
2e72b6347 memcg: implement ... |
3565 |
synchronize_rcu(); |
2e72b6347 memcg: implement ... |
3566 3567 3568 3569 3570 |
unlock: mutex_unlock(&memcg->thresholds_lock); return ret; } |
907860ed3 cgroups: make cft... |
3571 |
static void mem_cgroup_usage_unregister_event(struct cgroup *cgrp, |
9490ff275 memcg: oom notifier |
3572 |
struct cftype *cft, struct eventfd_ctx *eventfd) |
2e72b6347 memcg: implement ... |
3573 3574 |
{ struct mem_cgroup *memcg = mem_cgroup_from_cont(cgrp); |
2c488db27 memcg: clean up m... |
3575 3576 |
struct mem_cgroup_thresholds *thresholds; struct mem_cgroup_threshold_ary *new; |
2e72b6347 memcg: implement ... |
3577 3578 |
int type = MEMFILE_TYPE(cft->private); u64 usage; |
2c488db27 memcg: clean up m... |
3579 |
int i, j, size; |
2e72b6347 memcg: implement ... |
3580 3581 3582 |
mutex_lock(&memcg->thresholds_lock); if (type == _MEM) |
2c488db27 memcg: clean up m... |
3583 |
thresholds = &memcg->thresholds; |
2e72b6347 memcg: implement ... |
3584 |
else if (type == _MEMSWAP) |
2c488db27 memcg: clean up m... |
3585 |
thresholds = &memcg->memsw_thresholds; |
2e72b6347 memcg: implement ... |
3586 3587 3588 3589 3590 3591 3592 3593 3594 3595 3596 3597 3598 3599 3600 |
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 memcg: clean up m... |
3601 3602 3603 |
size = 0; for (i = 0; i < thresholds->primary->size; i++) { if (thresholds->primary->entries[i].eventfd != eventfd) |
2e72b6347 memcg: implement ... |
3604 3605 |
size++; } |
2c488db27 memcg: clean up m... |
3606 |
new = thresholds->spare; |
907860ed3 cgroups: make cft... |
3607 |
|
2e72b6347 memcg: implement ... |
3608 3609 |
/* Set thresholds array to NULL if we don't have thresholds */ if (!size) { |
2c488db27 memcg: clean up m... |
3610 3611 |
kfree(new); new = NULL; |
907860ed3 cgroups: make cft... |
3612 |
goto swap_buffers; |
2e72b6347 memcg: implement ... |
3613 |
} |
2c488db27 memcg: clean up m... |
3614 |
new->size = size; |
2e72b6347 memcg: implement ... |
3615 3616 |
/* Copy thresholds and find current threshold */ |
2c488db27 memcg: clean up m... |
3617 3618 3619 |
new->current_threshold = -1; for (i = 0, j = 0; i < thresholds->primary->size; i++) { if (thresholds->primary->entries[i].eventfd == eventfd) |
2e72b6347 memcg: implement ... |
3620 |
continue; |
2c488db27 memcg: clean up m... |
3621 3622 |
new->entries[j] = thresholds->primary->entries[i]; if (new->entries[j].threshold < usage) { |
2e72b6347 memcg: implement ... |
3623 |
/* |
2c488db27 memcg: clean up m... |
3624 |
* new->current_threshold will not be used |
2e72b6347 memcg: implement ... |
3625 3626 3627 |
* until rcu_assign_pointer(), so it's safe to increment * it here. */ |
2c488db27 memcg: clean up m... |
3628 |
++new->current_threshold; |
2e72b6347 memcg: implement ... |
3629 3630 3631 |
} j++; } |
907860ed3 cgroups: make cft... |
3632 |
swap_buffers: |
2c488db27 memcg: clean up m... |
3633 3634 3635 |
/* Swap primary and spare array */ thresholds->spare = thresholds->primary; rcu_assign_pointer(thresholds->primary, new); |
2e72b6347 memcg: implement ... |
3636 |
|
907860ed3 cgroups: make cft... |
3637 |
/* To be sure that nobody uses thresholds */ |
2e72b6347 memcg: implement ... |
3638 |
synchronize_rcu(); |
2e72b6347 memcg: implement ... |
3639 |
mutex_unlock(&memcg->thresholds_lock); |
2e72b6347 memcg: implement ... |
3640 |
} |
c1e862c1f memcg: new force_... |
3641 |
|
9490ff275 memcg: oom notifier |
3642 3643 3644 3645 3646 3647 3648 3649 3650 3651 3652 3653 3654 3655 3656 3657 3658 3659 3660 3661 3662 3663 3664 3665 |
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 cgroups: make cft... |
3666 |
static void mem_cgroup_oom_unregister_event(struct cgroup *cgrp, |
9490ff275 memcg: oom notifier |
3667 3668 3669 3670 3671 3672 3673 3674 3675 3676 3677 3678 3679 3680 3681 3682 3683 3684 |
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 memcg: oom notifier |
3685 |
} |
3c11ecf44 memcg: oom kill d... |
3686 3687 3688 3689 3690 3691 3692 3693 3694 3695 3696 3697 3698 |
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 memcg: oom kill d... |
3699 3700 3701 3702 3703 3704 3705 3706 3707 3708 3709 3710 3711 3712 3713 3714 3715 3716 3717 3718 |
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 memcg: fix wake u... |
3719 3720 |
if (!val) memcg_oom_recover(mem); |
3c11ecf44 memcg: oom kill d... |
3721 3722 3723 |
cgroup_unlock(); return 0; } |
8cdea7c05 Memory controller... |
3724 3725 |
static struct cftype mem_cgroup_files[] = { { |
0eea10301 Memory controller... |
3726 |
.name = "usage_in_bytes", |
8c7c6e34a memcg: mem+swap c... |
3727 |
.private = MEMFILE_PRIVATE(_MEM, RES_USAGE), |
2c3daa722 CGroup API files:... |
3728 |
.read_u64 = mem_cgroup_read, |
9490ff275 memcg: oom notifier |
3729 3730 |
.register_event = mem_cgroup_usage_register_event, .unregister_event = mem_cgroup_usage_unregister_event, |
8cdea7c05 Memory controller... |
3731 3732 |
}, { |
c84872e16 memcgroup: add th... |
3733 |
.name = "max_usage_in_bytes", |
8c7c6e34a memcg: mem+swap c... |
3734 |
.private = MEMFILE_PRIVATE(_MEM, RES_MAX_USAGE), |
29f2a4dac memcgroup: implem... |
3735 |
.trigger = mem_cgroup_reset, |
c84872e16 memcgroup: add th... |
3736 3737 3738 |
.read_u64 = mem_cgroup_read, }, { |
0eea10301 Memory controller... |
3739 |
.name = "limit_in_bytes", |
8c7c6e34a memcg: mem+swap c... |
3740 |
.private = MEMFILE_PRIVATE(_MEM, RES_LIMIT), |
856c13aa1 cgroup files: con... |
3741 |
.write_string = mem_cgroup_write, |
2c3daa722 CGroup API files:... |
3742 |
.read_u64 = mem_cgroup_read, |
8cdea7c05 Memory controller... |
3743 3744 |
}, { |
296c81d89 memory controller... |
3745 3746 3747 3748 3749 3750 |
.name = "soft_limit_in_bytes", .private = MEMFILE_PRIVATE(_MEM, RES_SOFT_LIMIT), .write_string = mem_cgroup_write, .read_u64 = mem_cgroup_read, }, { |
8cdea7c05 Memory controller... |
3751 |
.name = "failcnt", |
8c7c6e34a memcg: mem+swap c... |
3752 |
.private = MEMFILE_PRIVATE(_MEM, RES_FAILCNT), |
29f2a4dac memcgroup: implem... |
3753 |
.trigger = mem_cgroup_reset, |
2c3daa722 CGroup API files:... |
3754 |
.read_u64 = mem_cgroup_read, |
8cdea7c05 Memory controller... |
3755 |
}, |
8697d3319 Memory controller... |
3756 |
{ |
d2ceb9b7d memory cgroup enh... |
3757 |
.name = "stat", |
c64745cf0 CGroup API files:... |
3758 |
.read_map = mem_control_stat_show, |
d2ceb9b7d memory cgroup enh... |
3759 |
}, |
c1e862c1f memcg: new force_... |
3760 3761 3762 3763 |
{ .name = "force_empty", .trigger = mem_cgroup_force_empty_write, }, |
18f59ea7d memcg: memory cgr... |
3764 3765 3766 3767 3768 |
{ .name = "use_hierarchy", .write_u64 = mem_cgroup_hierarchy_write, .read_u64 = mem_cgroup_hierarchy_read, }, |
a7885eb8a memcg: swappiness |
3769 3770 3771 3772 3773 |
{ .name = "swappiness", .read_u64 = mem_cgroup_swappiness_read, .write_u64 = mem_cgroup_swappiness_write, }, |
7dc74be03 memcg: add interf... |
3774 3775 3776 3777 3778 |
{ .name = "move_charge_at_immigrate", .read_u64 = mem_cgroup_move_charge_read, .write_u64 = mem_cgroup_move_charge_write, }, |
9490ff275 memcg: oom notifier |
3779 3780 |
{ .name = "oom_control", |
3c11ecf44 memcg: oom kill d... |
3781 3782 |
.read_map = mem_cgroup_oom_control_read, .write_u64 = mem_cgroup_oom_control_write, |
9490ff275 memcg: oom notifier |
3783 3784 3785 3786 |
.register_event = mem_cgroup_oom_register_event, .unregister_event = mem_cgroup_oom_unregister_event, .private = MEMFILE_PRIVATE(_OOM_TYPE, OOM_CONTROL), }, |
8cdea7c05 Memory controller... |
3787 |
}; |
8c7c6e34a memcg: mem+swap c... |
3788 3789 3790 3791 3792 3793 |
#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 memcg: oom notifier |
3794 3795 |
.register_event = mem_cgroup_usage_register_event, .unregister_event = mem_cgroup_usage_unregister_event, |
8c7c6e34a memcg: mem+swap c... |
3796 3797 3798 3799 3800 3801 3802 3803 3804 3805 3806 3807 3808 3809 3810 3811 3812 3813 3814 3815 3816 3817 3818 3819 3820 3821 3822 3823 3824 3825 3826 3827 3828 3829 |
}, { .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 per-zone and recl... |
3830 3831 3832 |
static int alloc_mem_cgroup_per_zone_info(struct mem_cgroup *mem, int node) { struct mem_cgroup_per_node *pn; |
1ecaab2bd per-zone and recl... |
3833 |
struct mem_cgroup_per_zone *mz; |
b69408e88 vmscan: Use an in... |
3834 |
enum lru_list l; |
41e3355de memcg: fix node_s... |
3835 |
int zone, tmp = node; |
1ecaab2bd per-zone and recl... |
3836 3837 3838 3839 3840 3841 3842 3843 |
/* * 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 memcg: fix node_s... |
3844 3845 3846 |
if (!node_state(node, N_NORMAL_MEMORY)) tmp = -1; pn = kmalloc_node(sizeof(*pn), GFP_KERNEL, tmp); |
6d12e2d8d per-zone and recl... |
3847 3848 |
if (!pn) return 1; |
1ecaab2bd per-zone and recl... |
3849 |
|
6d12e2d8d per-zone and recl... |
3850 3851 |
mem->info.nodeinfo[node] = pn; memset(pn, 0, sizeof(*pn)); |
1ecaab2bd per-zone and recl... |
3852 3853 3854 |
for (zone = 0; zone < MAX_NR_ZONES; zone++) { mz = &pn->zoneinfo[zone]; |
b69408e88 vmscan: Use an in... |
3855 3856 |
for_each_lru(l) INIT_LIST_HEAD(&mz->lists[l]); |
f64c3f549 memory controller... |
3857 |
mz->usage_in_excess = 0; |
4e4169535 memory controller... |
3858 3859 |
mz->on_tree = false; mz->mem = mem; |
1ecaab2bd per-zone and recl... |
3860 |
} |
6d12e2d8d per-zone and recl... |
3861 3862 |
return 0; } |
1ecaab2bd per-zone and recl... |
3863 3864 3865 3866 |
static void free_mem_cgroup_per_zone_info(struct mem_cgroup *mem, int node) { kfree(mem->info.nodeinfo[node]); } |
333279487 memcgroup: use vm... |
3867 3868 3869 |
static struct mem_cgroup *mem_cgroup_alloc(void) { struct mem_cgroup *mem; |
c62b1a3b3 memcg: use generi... |
3870 |
int size = sizeof(struct mem_cgroup); |
333279487 memcgroup: use vm... |
3871 |
|
c62b1a3b3 memcg: use generi... |
3872 |
/* Can be very big if MAX_NUMNODES is very big */ |
c8dad2bb6 memcg: reduce siz... |
3873 3874 |
if (size < PAGE_SIZE) mem = kmalloc(size, GFP_KERNEL); |
333279487 memcgroup: use vm... |
3875 |
else |
c8dad2bb6 memcg: reduce siz... |
3876 |
mem = vmalloc(size); |
333279487 memcgroup: use vm... |
3877 |
|
e7bbcdf37 memcontrol: fix p... |
3878 3879 3880 3881 |
if (!mem) return NULL; memset(mem, 0, size); |
c62b1a3b3 memcg: use generi... |
3882 3883 3884 3885 3886 3887 3888 3889 |
mem->stat = alloc_percpu(struct mem_cgroup_stat_cpu); if (!mem->stat) { if (size < PAGE_SIZE) kfree(mem); else vfree(mem); mem = NULL; } |
333279487 memcgroup: use vm... |
3890 3891 |
return mem; } |
8c7c6e34a memcg: mem+swap c... |
3892 3893 3894 3895 3896 3897 3898 3899 |
/* * 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 memcg: mem+swap c... |
3900 3901 |
* Removal of cgroup itself succeeds regardless of refs from swap. */ |
a7ba0eef3 memcg: fix double... |
3902 |
static void __mem_cgroup_free(struct mem_cgroup *mem) |
333279487 memcgroup: use vm... |
3903 |
{ |
08e552c69 memcg: synchroniz... |
3904 |
int node; |
f64c3f549 memory controller... |
3905 |
mem_cgroup_remove_from_trees(mem); |
04046e1a0 memcg: use CSS ID |
3906 |
free_css_id(&mem_cgroup_subsys, &mem->css); |
08e552c69 memcg: synchroniz... |
3907 3908 |
for_each_node_state(node, N_POSSIBLE) free_mem_cgroup_per_zone_info(mem, node); |
c62b1a3b3 memcg: use generi... |
3909 3910 |
free_percpu(mem->stat); if (sizeof(struct mem_cgroup) < PAGE_SIZE) |
333279487 memcgroup: use vm... |
3911 3912 3913 3914 |
kfree(mem); else vfree(mem); } |
8c7c6e34a memcg: mem+swap c... |
3915 3916 3917 3918 |
static void mem_cgroup_get(struct mem_cgroup *mem) { atomic_inc(&mem->refcnt); } |
483c30b51 memcg: improve pe... |
3919 |
static void __mem_cgroup_put(struct mem_cgroup *mem, int count) |
8c7c6e34a memcg: mem+swap c... |
3920 |
{ |
483c30b51 memcg: improve pe... |
3921 |
if (atomic_sub_and_test(count, &mem->refcnt)) { |
7bcc1bb12 memcg: get/put pa... |
3922 |
struct mem_cgroup *parent = parent_mem_cgroup(mem); |
a7ba0eef3 memcg: fix double... |
3923 |
__mem_cgroup_free(mem); |
7bcc1bb12 memcg: get/put pa... |
3924 3925 3926 |
if (parent) mem_cgroup_put(parent); } |
8c7c6e34a memcg: mem+swap c... |
3927 |
} |
483c30b51 memcg: improve pe... |
3928 3929 3930 3931 |
static void mem_cgroup_put(struct mem_cgroup *mem) { __mem_cgroup_put(mem, 1); } |
7bcc1bb12 memcg: get/put pa... |
3932 3933 3934 3935 3936 3937 3938 3939 3940 |
/* * 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 memcgroup: use vm... |
3941 |
|
c077719be memcg: mem+swap c... |
3942 3943 3944 |
#ifdef CONFIG_CGROUP_MEM_RES_CTLR_SWAP static void __init enable_swap_cgroup(void) { |
f8d665422 memcg: add mem_cg... |
3945 |
if (!mem_cgroup_disabled() && really_do_swap_account) |
c077719be memcg: mem+swap c... |
3946 3947 3948 3949 3950 3951 3952 |
do_swap_account = 1; } #else static void __init enable_swap_cgroup(void) { } #endif |
f64c3f549 memory controller... |
3953 3954 3955 3956 3957 3958 3959 3960 3961 3962 3963 3964 3965 3966 3967 3968 3969 3970 3971 3972 3973 3974 3975 3976 |
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 memcg: fix sectio... |
3977 |
static struct cgroup_subsys_state * __ref |
8cdea7c05 Memory controller... |
3978 3979 |
mem_cgroup_create(struct cgroup_subsys *ss, struct cgroup *cont) { |
28dbc4b6a memcg: memory cgr... |
3980 |
struct mem_cgroup *mem, *parent; |
04046e1a0 memcg: use CSS ID |
3981 |
long error = -ENOMEM; |
6d12e2d8d per-zone and recl... |
3982 |
int node; |
8cdea7c05 Memory controller... |
3983 |
|
c8dad2bb6 memcg: reduce siz... |
3984 3985 |
mem = mem_cgroup_alloc(); if (!mem) |
04046e1a0 memcg: use CSS ID |
3986 |
return ERR_PTR(error); |
78fb74669 Memory controller... |
3987 |
|
6d12e2d8d per-zone and recl... |
3988 3989 3990 |
for_each_node_state(node, N_POSSIBLE) if (alloc_mem_cgroup_per_zone_info(mem, node)) goto free_out; |
f64c3f549 memory controller... |
3991 |
|
c077719be memcg: mem+swap c... |
3992 |
/* root ? */ |
28dbc4b6a memcg: memory cgr... |
3993 |
if (cont->parent == NULL) { |
cdec2e426 memcg: coalesce c... |
3994 |
int cpu; |
c077719be memcg: mem+swap c... |
3995 |
enable_swap_cgroup(); |
28dbc4b6a memcg: memory cgr... |
3996 |
parent = NULL; |
4b3bde4c9 memcg: remove the... |
3997 |
root_mem_cgroup = mem; |
f64c3f549 memory controller... |
3998 3999 |
if (mem_cgroup_soft_limit_tree_init()) goto free_out; |
cdec2e426 memcg: coalesce c... |
4000 4001 4002 4003 4004 4005 |
for_each_possible_cpu(cpu) { struct memcg_stock_pcp *stock = &per_cpu(memcg_stock, cpu); INIT_WORK(&stock->work, drain_local_stock); } hotcpu_notifier(memcg_stock_cpu_callback, 0); |
18f59ea7d memcg: memory cgr... |
4006 |
} else { |
28dbc4b6a memcg: memory cgr... |
4007 |
parent = mem_cgroup_from_cont(cont->parent); |
18f59ea7d memcg: memory cgr... |
4008 |
mem->use_hierarchy = parent->use_hierarchy; |
3c11ecf44 memcg: oom kill d... |
4009 |
mem->oom_kill_disable = parent->oom_kill_disable; |
18f59ea7d memcg: memory cgr... |
4010 |
} |
28dbc4b6a memcg: memory cgr... |
4011 |
|
18f59ea7d memcg: memory cgr... |
4012 4013 4014 |
if (parent && parent->use_hierarchy) { res_counter_init(&mem->res, &parent->res); res_counter_init(&mem->memsw, &parent->memsw); |
7bcc1bb12 memcg: get/put pa... |
4015 4016 4017 4018 4019 4020 4021 |
/* * 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 memcg: memory cgr... |
4022 4023 4024 4025 |
} else { res_counter_init(&mem->res, NULL); res_counter_init(&mem->memsw, NULL); } |
04046e1a0 memcg: use CSS ID |
4026 |
mem->last_scanned_child = 0; |
2733c06ac memcg: protect pr... |
4027 |
spin_lock_init(&mem->reclaim_param_lock); |
9490ff275 memcg: oom notifier |
4028 |
INIT_LIST_HEAD(&mem->oom_notify); |
6d61ef409 memcg: memory cgr... |
4029 |
|
a7885eb8a memcg: swappiness |
4030 4031 |
if (parent) mem->swappiness = get_swappiness(parent); |
a7ba0eef3 memcg: fix double... |
4032 |
atomic_set(&mem->refcnt, 1); |
7dc74be03 memcg: add interf... |
4033 |
mem->move_charge_at_immigrate = 0; |
2e72b6347 memcg: implement ... |
4034 |
mutex_init(&mem->thresholds_lock); |
8cdea7c05 Memory controller... |
4035 |
return &mem->css; |
6d12e2d8d per-zone and recl... |
4036 |
free_out: |
a7ba0eef3 memcg: fix double... |
4037 |
__mem_cgroup_free(mem); |
4b3bde4c9 memcg: remove the... |
4038 |
root_mem_cgroup = NULL; |
04046e1a0 memcg: use CSS ID |
4039 |
return ERR_PTR(error); |
8cdea7c05 Memory controller... |
4040 |
} |
ec64f5154 cgroup: fix frequ... |
4041 |
static int mem_cgroup_pre_destroy(struct cgroup_subsys *ss, |
df878fb04 memory cgroup enh... |
4042 4043 4044 |
struct cgroup *cont) { struct mem_cgroup *mem = mem_cgroup_from_cont(cont); |
ec64f5154 cgroup: fix frequ... |
4045 4046 |
return mem_cgroup_force_empty(mem, false); |
df878fb04 memory cgroup enh... |
4047 |
} |
8cdea7c05 Memory controller... |
4048 4049 4050 |
static void mem_cgroup_destroy(struct cgroup_subsys *ss, struct cgroup *cont) { |
c268e9946 memcg: fix hierar... |
4051 |
struct mem_cgroup *mem = mem_cgroup_from_cont(cont); |
c268e9946 memcg: fix hierar... |
4052 |
|
c268e9946 memcg: fix hierar... |
4053 |
mem_cgroup_put(mem); |
8cdea7c05 Memory controller... |
4054 4055 4056 4057 4058 |
} static int mem_cgroup_populate(struct cgroup_subsys *ss, struct cgroup *cont) { |
8c7c6e34a memcg: mem+swap c... |
4059 4060 4061 4062 4063 4064 4065 4066 |
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 Memory controller... |
4067 |
} |
024914477 memcg: move charg... |
4068 |
#ifdef CONFIG_MMU |
7dc74be03 memcg: add interf... |
4069 |
/* Handlers for move charge at task migration. */ |
854ffa8d1 memcg: improve pe... |
4070 4071 |
#define PRECHARGE_COUNT_AT_ONCE 256 static int mem_cgroup_do_precharge(unsigned long count) |
7dc74be03 memcg: add interf... |
4072 |
{ |
854ffa8d1 memcg: improve pe... |
4073 4074 |
int ret = 0; int batch_count = PRECHARGE_COUNT_AT_ONCE; |
4ffef5fef memcg: move charg... |
4075 |
struct mem_cgroup *mem = mc.to; |
854ffa8d1 memcg: improve pe... |
4076 4077 4078 4079 4080 4081 4082 4083 4084 4085 4086 4087 4088 4089 4090 4091 4092 4093 4094 4095 4096 4097 |
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 memcg: improve pe... |
4098 4099 4100 4101 4102 4103 4104 4105 4106 4107 4108 4109 4110 |
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(); } |
430e48631 memcg: update thr... |
4111 |
ret = __mem_cgroup_try_charge(NULL, GFP_KERNEL, &mem, false); |
854ffa8d1 memcg: improve pe... |
4112 4113 4114 4115 4116 |
if (ret || !mem) /* mem_cgroup_clear_mc() will do uncharge later */ return -ENOMEM; mc.precharge++; } |
4ffef5fef memcg: move charg... |
4117 4118 4119 4120 4121 4122 4123 4124 |
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 memcg: move charg... |
4125 |
* @target: the pointer the target page or swap ent will be stored(can be NULL) |
4ffef5fef memcg: move charg... |
4126 4127 4128 4129 4130 4131 |
* * 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 memcg: move charg... |
4132 4133 4134 |
* 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 memcg: move charg... |
4135 4136 4137 |
* * Called with pte lock held. */ |
4ffef5fef memcg: move charg... |
4138 4139 |
union mc_target { struct page *page; |
024914477 memcg: move charg... |
4140 |
swp_entry_t ent; |
4ffef5fef memcg: move charg... |
4141 |
}; |
4ffef5fef memcg: move charg... |
4142 4143 4144 |
enum mc_target_type { MC_TARGET_NONE, /* not used */ MC_TARGET_PAGE, |
024914477 memcg: move charg... |
4145 |
MC_TARGET_SWAP, |
4ffef5fef memcg: move charg... |
4146 |
}; |
90254a658 memcg: clean up m... |
4147 4148 |
static struct page *mc_handle_present_pte(struct vm_area_struct *vma, unsigned long addr, pte_t ptent) |
4ffef5fef memcg: move charg... |
4149 |
{ |
90254a658 memcg: clean up m... |
4150 |
struct page *page = vm_normal_page(vma, addr, ptent); |
4ffef5fef memcg: move charg... |
4151 |
|
90254a658 memcg: clean up m... |
4152 4153 4154 4155 4156 4157 |
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 memcg: move charg... |
4158 4159 |
} else if (!move_file()) /* we ignore mapcount for file pages */ |
90254a658 memcg: clean up m... |
4160 4161 4162 4163 4164 4165 4166 4167 4168 4169 4170 4171 4172 4173 4174 4175 4176 4177 |
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 memcg: move charg... |
4178 4179 |
if (page) put_page(page); |
90254a658 memcg: clean up m... |
4180 |
return NULL; |
024914477 memcg: move charg... |
4181 |
} |
90254a658 memcg: clean up m... |
4182 4183 4184 4185 4186 |
if (do_swap_account) entry->val = ent.val; return page; } |
87946a722 memcg: move charg... |
4187 4188 4189 4190 4191 4192 4193 4194 4195 4196 4197 4198 4199 4200 4201 4202 4203 4204 4205 4206 4207 4208 4209 4210 4211 4212 4213 4214 4215 4216 4217 4218 |
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 memcg: clean up m... |
4219 4220 4221 4222 4223 4224 4225 4226 4227 4228 4229 4230 |
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 memcg: move charg... |
4231 4232 |
else if (pte_none(ptent) || pte_file(ptent)) page = mc_handle_file_pte(vma, addr, ptent, &ent); |
90254a658 memcg: clean up m... |
4233 4234 4235 |
if (!page && !ent.val) return 0; |
024914477 memcg: move charg... |
4236 4237 4238 4239 4240 4241 4242 4243 4244 4245 4246 4247 4248 4249 4250 |
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 memcg: clean up m... |
4251 4252 |
/* There is a swap entry and a page doesn't exist or isn't charged */ if (ent.val && !ret && |
7f0f15464 memcg: fix css_id... |
4253 4254 4255 4256 |
css_id(&mc.from->css) == lookup_swap_cgroup(ent)) { ret = MC_TARGET_SWAP; if (target) target->ent = ent; |
4ffef5fef memcg: move charg... |
4257 |
} |
4ffef5fef memcg: move charg... |
4258 4259 4260 4261 4262 4263 4264 4265 4266 4267 4268 4269 4270 4271 4272 4273 4274 |
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; 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 memcg: add interf... |
4275 4276 |
return 0; } |
4ffef5fef memcg: move charg... |
4277 4278 4279 4280 4281 4282 4283 4284 4285 4286 4287 4288 4289 4290 |
static unsigned long mem_cgroup_count_precharge(struct mm_struct *mm) { unsigned long precharge; struct vm_area_struct *vma; down_read(&mm->mmap_sem); 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 memcg: move charg... |
4291 4292 4293 4294 4295 4296 4297 4298 4299 4300 |
walk_page_range(vma->vm_start, vma->vm_end, &mem_cgroup_count_precharge_walk); } up_read(&mm->mmap_sem); precharge = mc.precharge; mc.precharge = 0; return precharge; } |
4ffef5fef memcg: move charg... |
4301 4302 |
static int mem_cgroup_precharge_mc(struct mm_struct *mm) { |
854ffa8d1 memcg: improve pe... |
4303 |
return mem_cgroup_do_precharge(mem_cgroup_count_precharge(mm)); |
4ffef5fef memcg: move charg... |
4304 4305 4306 4307 |
} static void mem_cgroup_clear_mc(void) { |
2bd9bb206 memcg: clean up w... |
4308 4309 |
struct mem_cgroup *from = mc.from; struct mem_cgroup *to = mc.to; |
4ffef5fef memcg: move charg... |
4310 |
/* we must uncharge all the leftover precharges from mc.to */ |
854ffa8d1 memcg: improve pe... |
4311 4312 4313 4314 4315 4316 4317 4318 4319 4320 4321 |
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 memcg: move charg... |
4322 |
} |
483c30b51 memcg: improve pe... |
4323 4324 |
/* we must fixup refcnts and charges */ if (mc.moved_swap) { |
483c30b51 memcg: improve pe... |
4325 4326 4327 4328 4329 4330 4331 4332 4333 4334 4335 4336 4337 |
/* 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 memcg: improve pe... |
4338 4339 4340 4341 4342 |
} /* we've already done mem_cgroup_get(mc.to) */ mc.moved_swap = 0; } |
2bd9bb206 memcg: clean up w... |
4343 |
spin_lock(&mc.lock); |
4ffef5fef memcg: move charg... |
4344 4345 |
mc.from = NULL; mc.to = NULL; |
8033b97c9 memcg: avoid oom ... |
4346 |
mc.moving_task = NULL; |
2bd9bb206 memcg: clean up w... |
4347 4348 4349 |
spin_unlock(&mc.lock); memcg_oom_recover(from); memcg_oom_recover(to); |
8033b97c9 memcg: avoid oom ... |
4350 |
wake_up_all(&mc.waitq); |
4ffef5fef memcg: move charg... |
4351 |
} |
7dc74be03 memcg: add interf... |
4352 4353 4354 4355 4356 4357 4358 4359 4360 4361 4362 4363 4364 4365 4366 4367 4368 |
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 memcg: add interf... |
4369 |
/* We move charges only when we move a owner of the mm */ |
4ffef5fef memcg: move charg... |
4370 4371 4372 4373 |
if (mm->owner == p) { VM_BUG_ON(mc.from); VM_BUG_ON(mc.to); VM_BUG_ON(mc.precharge); |
854ffa8d1 memcg: improve pe... |
4374 |
VM_BUG_ON(mc.moved_charge); |
483c30b51 memcg: improve pe... |
4375 |
VM_BUG_ON(mc.moved_swap); |
8033b97c9 memcg: avoid oom ... |
4376 |
VM_BUG_ON(mc.moving_task); |
2bd9bb206 memcg: clean up w... |
4377 |
spin_lock(&mc.lock); |
4ffef5fef memcg: move charg... |
4378 4379 4380 |
mc.from = from; mc.to = mem; mc.precharge = 0; |
854ffa8d1 memcg: improve pe... |
4381 |
mc.moved_charge = 0; |
483c30b51 memcg: improve pe... |
4382 |
mc.moved_swap = 0; |
8033b97c9 memcg: avoid oom ... |
4383 |
mc.moving_task = current; |
2bd9bb206 memcg: clean up w... |
4384 |
spin_unlock(&mc.lock); |
4ffef5fef memcg: move charg... |
4385 4386 4387 4388 4389 |
ret = mem_cgroup_precharge_mc(mm); if (ret) mem_cgroup_clear_mc(); } |
7dc74be03 memcg: add interf... |
4390 4391 4392 4393 4394 4395 4396 4397 4398 4399 |
mmput(mm); } return ret; } static void mem_cgroup_cancel_attach(struct cgroup_subsys *ss, struct cgroup *cgroup, struct task_struct *p, bool threadgroup) { |
4ffef5fef memcg: move charg... |
4400 |
mem_cgroup_clear_mc(); |
7dc74be03 memcg: add interf... |
4401 |
} |
4ffef5fef memcg: move charg... |
4402 4403 4404 |
static int mem_cgroup_move_charge_pte_range(pmd_t *pmd, unsigned long addr, unsigned long end, struct mm_walk *walk) |
7dc74be03 memcg: add interf... |
4405 |
{ |
4ffef5fef memcg: move charg... |
4406 4407 4408 4409 4410 4411 4412 4413 4414 4415 4416 4417 4418 |
int ret = 0; struct vm_area_struct *vma = walk->private; pte_t *pte; spinlock_t *ptl; retry: 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 memcg: move charg... |
4419 |
swp_entry_t ent; |
4ffef5fef memcg: move charg... |
4420 4421 4422 4423 4424 4425 4426 4427 4428 4429 4430 |
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 memcg: improve pe... |
4431 4432 |
if (!mem_cgroup_move_account(pc, mc.from, mc.to, false)) { |
4ffef5fef memcg: move charg... |
4433 |
mc.precharge--; |
854ffa8d1 memcg: improve pe... |
4434 4435 |
/* we uncharge from mc.from later. */ mc.moved_charge++; |
4ffef5fef memcg: move charg... |
4436 4437 4438 4439 4440 |
} putback_lru_page(page); put: /* is_target_pte_for_mc() gets the page */ put_page(page); break; |
024914477 memcg: move charg... |
4441 4442 |
case MC_TARGET_SWAP: ent = target.ent; |
483c30b51 memcg: improve pe... |
4443 4444 |
if (!mem_cgroup_move_swap_account(ent, mc.from, mc.to, false)) { |
024914477 memcg: move charg... |
4445 |
mc.precharge--; |
483c30b51 memcg: improve pe... |
4446 4447 4448 |
/* we fixup refcnts and charges later. */ mc.moved_swap++; } |
024914477 memcg: move charg... |
4449 |
break; |
4ffef5fef memcg: move charg... |
4450 4451 4452 4453 4454 4455 4456 4457 4458 4459 4460 4461 4462 4463 |
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 memcg: improve pe... |
4464 |
ret = mem_cgroup_do_precharge(1); |
4ffef5fef memcg: move charg... |
4465 4466 4467 4468 4469 4470 4471 4472 4473 4474 4475 4476 4477 4478 4479 4480 4481 4482 4483 4484 4485 4486 |
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(); down_read(&mm->mmap_sem); 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 memcg: move charg... |
4487 4488 4489 4490 4491 4492 4493 4494 4495 4496 |
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; } up_read(&mm->mmap_sem); |
7dc74be03 memcg: add interf... |
4497 |
} |
67e465a77 Memory controller... |
4498 4499 4500 |
static void mem_cgroup_move_task(struct cgroup_subsys *ss, struct cgroup *cont, struct cgroup *old_cont, |
be367d099 cgroups: let ss->... |
4501 4502 |
struct task_struct *p, bool threadgroup) |
67e465a77 Memory controller... |
4503 |
{ |
4ffef5fef memcg: move charg... |
4504 4505 4506 4507 4508 4509 4510 4511 4512 4513 4514 4515 |
struct mm_struct *mm; if (!mc.to) /* no need to move charge */ return; mm = get_task_mm(p); if (mm) { mem_cgroup_move_charge(mm); mmput(mm); } mem_cgroup_clear_mc(); |
67e465a77 Memory controller... |
4516 |
} |
5cfb80a73 memcg: disable mo... |
4517 4518 4519 4520 4521 4522 4523 4524 4525 4526 4527 4528 4529 4530 4531 4532 4533 4534 4535 4536 4537 4538 |
#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 Memory controller... |
4539 |
|
8cdea7c05 Memory controller... |
4540 4541 4542 4543 |
struct cgroup_subsys mem_cgroup_subsys = { .name = "memory", .subsys_id = mem_cgroup_subsys_id, .create = mem_cgroup_create, |
df878fb04 memory cgroup enh... |
4544 |
.pre_destroy = mem_cgroup_pre_destroy, |
8cdea7c05 Memory controller... |
4545 4546 |
.destroy = mem_cgroup_destroy, .populate = mem_cgroup_populate, |
7dc74be03 memcg: add interf... |
4547 4548 |
.can_attach = mem_cgroup_can_attach, .cancel_attach = mem_cgroup_cancel_attach, |
67e465a77 Memory controller... |
4549 |
.attach = mem_cgroup_move_task, |
6d12e2d8d per-zone and recl... |
4550 |
.early_init = 0, |
04046e1a0 memcg: use CSS ID |
4551 |
.use_id = 1, |
8cdea7c05 Memory controller... |
4552 |
}; |
c077719be memcg: mem+swap c... |
4553 4554 4555 4556 4557 4558 4559 4560 4561 4562 |
#ifdef CONFIG_CGROUP_MEM_RES_CTLR_SWAP static int __init disable_swap_account(char *s) { really_do_swap_account = 0; return 1; } __setup("noswapaccount", disable_swap_account); #endif |