Blame view
block/blk-cgroup.c
35.8 KB
31e4c28d9
|
1 2 3 4 5 6 7 8 9 10 11 |
/* * Common Block IO controller cgroup interface * * Based on ideas and code from CFQ, CFS and BFQ: * Copyright (C) 2003 Jens Axboe <axboe@kernel.dk> * * Copyright (C) 2008 Fabio Checconi <fabio@gandalf.sssup.it> * Paolo Valente <paolo.valente@unimore.it> * * Copyright (C) 2009 Vivek Goyal <vgoyal@redhat.com> * Nauman Rafique <nauman@google.com> |
e48453c38
|
12 13 14 15 |
* * For policy-specific per-blkcg data: * Copyright (C) 2015 Paolo Valente <paolo.valente@unimore.it> * Arianna Avanzini <avanzini.arianna@gmail.com> |
31e4c28d9
|
16 17 |
*/ #include <linux/ioprio.h> |
220841906
|
18 |
#include <linux/kdev_t.h> |
9d6a986c0
|
19 |
#include <linux/module.h> |
accee7854
|
20 |
#include <linux/err.h> |
9195291e5
|
21 |
#include <linux/blkdev.h> |
52ebea749
|
22 |
#include <linux/backing-dev.h> |
5a0e3ad6a
|
23 |
#include <linux/slab.h> |
34d0f179d
|
24 |
#include <linux/genhd.h> |
72e06c255
|
25 |
#include <linux/delay.h> |
9a9e8a26d
|
26 |
#include <linux/atomic.h> |
36aa9e5f5
|
27 |
#include <linux/ctype.h> |
eea8f41cc
|
28 |
#include <linux/blk-cgroup.h> |
5efd61135
|
29 |
#include "blk.h" |
3e2520668
|
30 |
|
84c124da9
|
31 |
#define MAX_KEY_LEN 100 |
838f13bf4
|
32 33 34 35 36 37 38 39 |
/* * blkcg_pol_mutex protects blkcg_policy[] and policy [de]activation. * blkcg_pol_register_mutex nests outside of it and synchronizes entire * policy [un]register operations including cgroup file additions / * removals. Putting cgroup file registration outside blkcg_pol_mutex * allows grabbing it from cgroup callbacks. */ static DEFINE_MUTEX(blkcg_pol_register_mutex); |
bc0d6501a
|
40 |
static DEFINE_MUTEX(blkcg_pol_mutex); |
923adde1b
|
41 |
|
e48453c38
|
42 |
struct blkcg blkcg_root; |
3c798398e
|
43 |
EXPORT_SYMBOL_GPL(blkcg_root); |
9d6a986c0
|
44 |
|
496d5e756
|
45 |
struct cgroup_subsys_state * const blkcg_root_css = &blkcg_root.css; |
3c798398e
|
46 |
static struct blkcg_policy *blkcg_policy[BLKCG_MAX_POLS]; |
035d10b2f
|
47 |
|
7876f930d
|
48 |
static LIST_HEAD(all_blkcgs); /* protected by blkcg_pol_mutex */ |
a2b1693ba
|
49 |
static bool blkcg_policy_enabled(struct request_queue *q, |
3c798398e
|
50 |
const struct blkcg_policy *pol) |
a2b1693ba
|
51 52 53 |
{ return pol && test_bit(pol->plid, q->blkcg_pols); } |
0381411e4
|
54 55 56 57 58 59 |
/** * blkg_free - free a blkg * @blkg: blkg to free * * Free @blkg which may be partially allocated. */ |
3c798398e
|
60 |
static void blkg_free(struct blkcg_gq *blkg) |
0381411e4
|
61 |
{ |
e8989fae3
|
62 |
int i; |
549d3aa87
|
63 64 65 |
if (!blkg) return; |
db6136703
|
66 |
for (i = 0; i < BLKCG_MAX_POLS; i++) |
001bea73e
|
67 68 |
if (blkg->pd[i]) blkcg_policy[i]->pd_free_fn(blkg->pd[i]); |
e8989fae3
|
69 |
|
994b78327
|
70 71 |
if (blkg->blkcg != &blkcg_root) blk_exit_rl(&blkg->rl); |
77ea73388
|
72 73 74 |
blkg_rwstat_exit(&blkg->stat_ios); blkg_rwstat_exit(&blkg->stat_bytes); |
549d3aa87
|
75 |
kfree(blkg); |
0381411e4
|
76 77 78 79 80 81 |
} /** * blkg_alloc - allocate a blkg * @blkcg: block cgroup the new blkg is associated with * @q: request_queue the new blkg is associated with |
159749937
|
82 |
* @gfp_mask: allocation mask to use |
0381411e4
|
83 |
* |
e8989fae3
|
84 |
* Allocate a new blkg assocating @blkcg and @q. |
0381411e4
|
85 |
*/ |
159749937
|
86 87 |
static struct blkcg_gq *blkg_alloc(struct blkcg *blkcg, struct request_queue *q, gfp_t gfp_mask) |
0381411e4
|
88 |
{ |
3c798398e
|
89 |
struct blkcg_gq *blkg; |
e8989fae3
|
90 |
int i; |
0381411e4
|
91 92 |
/* alloc and init base part */ |
159749937
|
93 |
blkg = kzalloc_node(sizeof(*blkg), gfp_mask, q->node); |
0381411e4
|
94 95 |
if (!blkg) return NULL; |
77ea73388
|
96 97 98 |
if (blkg_rwstat_init(&blkg->stat_bytes, gfp_mask) || blkg_rwstat_init(&blkg->stat_ios, gfp_mask)) goto err_free; |
c875f4d02
|
99 |
blkg->q = q; |
e8989fae3
|
100 |
INIT_LIST_HEAD(&blkg->q_node); |
0381411e4
|
101 |
blkg->blkcg = blkcg; |
a5049a8ae
|
102 |
atomic_set(&blkg->refcnt, 1); |
0381411e4
|
103 |
|
a051661ca
|
104 105 106 107 108 109 |
/* root blkg uses @q->root_rl, init rl only for !root blkgs */ if (blkcg != &blkcg_root) { if (blk_init_rl(&blkg->rl, q, gfp_mask)) goto err_free; blkg->rl.blkg = blkg; } |
8bd435b30
|
110 |
for (i = 0; i < BLKCG_MAX_POLS; i++) { |
3c798398e
|
111 |
struct blkcg_policy *pol = blkcg_policy[i]; |
e8989fae3
|
112 |
struct blkg_policy_data *pd; |
0381411e4
|
113 |
|
a2b1693ba
|
114 |
if (!blkcg_policy_enabled(q, pol)) |
e8989fae3
|
115 116 117 |
continue; /* alloc per-policy data and attach it to blkg */ |
001bea73e
|
118 |
pd = pol->pd_alloc_fn(gfp_mask, q->node); |
a051661ca
|
119 120 |
if (!pd) goto err_free; |
549d3aa87
|
121 |
|
e8989fae3
|
122 123 |
blkg->pd[i] = pd; pd->blkg = blkg; |
b276a876a
|
124 |
pd->plid = i; |
e8989fae3
|
125 |
} |
0381411e4
|
126 |
return blkg; |
a051661ca
|
127 128 129 130 |
err_free: blkg_free(blkg); return NULL; |
0381411e4
|
131 |
} |
24f290466
|
132 133 |
struct blkcg_gq *blkg_lookup_slowpath(struct blkcg *blkcg, struct request_queue *q, bool update_hint) |
80fd99792
|
134 |
{ |
3c798398e
|
135 |
struct blkcg_gq *blkg; |
80fd99792
|
136 |
|
a637120e4
|
137 |
/* |
86cde6b62
|
138 139 140 141 |
* Hint didn't match. Look up from the radix tree. Note that the * hint can only be updated under queue_lock as otherwise @blkg * could have already been removed from blkg_tree. The caller is * responsible for grabbing queue_lock if @update_hint. |
a637120e4
|
142 143 |
*/ blkg = radix_tree_lookup(&blkcg->blkg_tree, q->id); |
86cde6b62
|
144 145 146 147 148 |
if (blkg && blkg->q == q) { if (update_hint) { lockdep_assert_held(q->queue_lock); rcu_assign_pointer(blkcg->blkg_hint, blkg); } |
a637120e4
|
149 |
return blkg; |
86cde6b62
|
150 |
} |
a637120e4
|
151 |
|
80fd99792
|
152 153 |
return NULL; } |
ae1188963
|
154 |
EXPORT_SYMBOL_GPL(blkg_lookup_slowpath); |
80fd99792
|
155 |
|
159749937
|
156 157 |
/* * If @new_blkg is %NULL, this function tries to allocate a new one as |
d93a11f1c
|
158 |
* necessary using %GFP_NOWAIT. @new_blkg is always consumed on return. |
159749937
|
159 |
*/ |
86cde6b62
|
160 161 162 |
static struct blkcg_gq *blkg_create(struct blkcg *blkcg, struct request_queue *q, struct blkcg_gq *new_blkg) |
5624a4e44
|
163 |
{ |
3c798398e
|
164 |
struct blkcg_gq *blkg; |
ce7acfeaf
|
165 |
struct bdi_writeback_congested *wb_congested; |
f427d9096
|
166 |
int i, ret; |
5624a4e44
|
167 |
|
cd1604fab
|
168 169 |
WARN_ON_ONCE(!rcu_read_lock_held()); lockdep_assert_held(q->queue_lock); |
7ee9c5620
|
170 |
/* blkg holds a reference to blkcg */ |
ec903c0c8
|
171 |
if (!css_tryget_online(&blkcg->css)) { |
20386ce01
|
172 |
ret = -ENODEV; |
93e6d5d8f
|
173 |
goto err_free_blkg; |
159749937
|
174 |
} |
cd1604fab
|
175 |
|
ce7acfeaf
|
176 |
wb_congested = wb_congested_get_create(&q->backing_dev_info, |
d93a11f1c
|
177 |
blkcg->css.id, GFP_NOWAIT); |
ce7acfeaf
|
178 179 180 181 |
if (!wb_congested) { ret = -ENOMEM; goto err_put_css; } |
496fb7806
|
182 |
/* allocate */ |
159749937
|
183 |
if (!new_blkg) { |
d93a11f1c
|
184 |
new_blkg = blkg_alloc(blkcg, q, GFP_NOWAIT); |
159749937
|
185 |
if (unlikely(!new_blkg)) { |
93e6d5d8f
|
186 |
ret = -ENOMEM; |
ce7acfeaf
|
187 |
goto err_put_congested; |
159749937
|
188 189 190 |
} } blkg = new_blkg; |
ce7acfeaf
|
191 |
blkg->wb_congested = wb_congested; |
cd1604fab
|
192 |
|
db6136703
|
193 |
/* link parent */ |
3c5478659
|
194 195 196 |
if (blkcg_parent(blkcg)) { blkg->parent = __blkg_lookup(blkcg_parent(blkcg), q, false); if (WARN_ON_ONCE(!blkg->parent)) { |
20386ce01
|
197 |
ret = -ENODEV; |
ce7acfeaf
|
198 |
goto err_put_congested; |
3c5478659
|
199 200 201 |
} blkg_get(blkg->parent); } |
db6136703
|
202 203 204 205 206 |
/* invoke per-policy init */ for (i = 0; i < BLKCG_MAX_POLS; i++) { struct blkcg_policy *pol = blkcg_policy[i]; if (blkg->pd[i] && pol->pd_init_fn) |
a9520cd6f
|
207 |
pol->pd_init_fn(blkg->pd[i]); |
db6136703
|
208 209 210 |
} /* insert */ |
cd1604fab
|
211 |
spin_lock(&blkcg->lock); |
a637120e4
|
212 213 214 215 |
ret = radix_tree_insert(&blkcg->blkg_tree, q->id, blkg); if (likely(!ret)) { hlist_add_head_rcu(&blkg->blkcg_node, &blkcg->blkg_list); list_add(&blkg->q_node, &q->blkg_list); |
f427d9096
|
216 217 218 219 220 |
for (i = 0; i < BLKCG_MAX_POLS; i++) { struct blkcg_policy *pol = blkcg_policy[i]; if (blkg->pd[i] && pol->pd_online_fn) |
a9520cd6f
|
221 |
pol->pd_online_fn(blkg->pd[i]); |
f427d9096
|
222 |
} |
a637120e4
|
223 |
} |
f427d9096
|
224 |
blkg->online = true; |
cd1604fab
|
225 |
spin_unlock(&blkcg->lock); |
496fb7806
|
226 |
|
ec13b1d6f
|
227 |
if (!ret) |
a637120e4
|
228 |
return blkg; |
159749937
|
229 |
|
3c5478659
|
230 231 232 |
/* @blkg failed fully initialized, use the usual release path */ blkg_put(blkg); return ERR_PTR(ret); |
ce7acfeaf
|
233 234 |
err_put_congested: wb_congested_put(wb_congested); |
93e6d5d8f
|
235 |
err_put_css: |
496fb7806
|
236 |
css_put(&blkcg->css); |
93e6d5d8f
|
237 |
err_free_blkg: |
159749937
|
238 |
blkg_free(new_blkg); |
93e6d5d8f
|
239 |
return ERR_PTR(ret); |
31e4c28d9
|
240 |
} |
3c96cb32d
|
241 |
|
86cde6b62
|
242 243 244 245 246 247 |
/** * blkg_lookup_create - lookup blkg, try to create one if not there * @blkcg: blkcg of interest * @q: request_queue of interest * * Lookup blkg for the @blkcg - @q pair. If it doesn't exist, try to |
3c5478659
|
248 249 250 |
* create one. blkg creation is performed recursively from blkcg_root such * that all non-root blkg's have access to the parent blkg. This function * should be called under RCU read lock and @q->queue_lock. |
86cde6b62
|
251 252 253 254 255 |
* * Returns pointer to the looked up or created blkg on success, ERR_PTR() * value on error. If @q is dead, returns ERR_PTR(-EINVAL). If @q is not * dead and bypassing, returns ERR_PTR(-EBUSY). */ |
3c798398e
|
256 257 |
struct blkcg_gq *blkg_lookup_create(struct blkcg *blkcg, struct request_queue *q) |
3c96cb32d
|
258 |
{ |
86cde6b62
|
259 260 261 262 |
struct blkcg_gq *blkg; WARN_ON_ONCE(!rcu_read_lock_held()); lockdep_assert_held(q->queue_lock); |
3c96cb32d
|
263 264 265 266 267 |
/* * This could be the first entry point of blkcg implementation and * we shouldn't allow anything to go through for a bypassing queue. */ if (unlikely(blk_queue_bypass(q))) |
20386ce01
|
268 |
return ERR_PTR(blk_queue_dying(q) ? -ENODEV : -EBUSY); |
86cde6b62
|
269 270 271 272 |
blkg = __blkg_lookup(blkcg, q, true); if (blkg) return blkg; |
3c5478659
|
273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 |
/* * Create blkgs walking down from blkcg_root to @blkcg, so that all * non-root blkgs have access to their parents. */ while (true) { struct blkcg *pos = blkcg; struct blkcg *parent = blkcg_parent(blkcg); while (parent && !__blkg_lookup(parent, q, false)) { pos = parent; parent = blkcg_parent(parent); } blkg = blkg_create(pos, q, NULL); if (pos == blkcg || IS_ERR(blkg)) return blkg; } |
3c96cb32d
|
290 |
} |
31e4c28d9
|
291 |
|
3c798398e
|
292 |
static void blkg_destroy(struct blkcg_gq *blkg) |
03aa264ac
|
293 |
{ |
3c798398e
|
294 |
struct blkcg *blkcg = blkg->blkcg; |
77ea73388
|
295 |
struct blkcg_gq *parent = blkg->parent; |
f427d9096
|
296 |
int i; |
03aa264ac
|
297 |
|
27e1f9d1c
|
298 |
lockdep_assert_held(blkg->q->queue_lock); |
9f13ef678
|
299 |
lockdep_assert_held(&blkcg->lock); |
03aa264ac
|
300 301 |
/* Something wrong if we are trying to remove same group twice */ |
e8989fae3
|
302 |
WARN_ON_ONCE(list_empty(&blkg->q_node)); |
9f13ef678
|
303 |
WARN_ON_ONCE(hlist_unhashed(&blkg->blkcg_node)); |
a637120e4
|
304 |
|
f427d9096
|
305 306 307 308 |
for (i = 0; i < BLKCG_MAX_POLS; i++) { struct blkcg_policy *pol = blkcg_policy[i]; if (blkg->pd[i] && pol->pd_offline_fn) |
a9520cd6f
|
309 |
pol->pd_offline_fn(blkg->pd[i]); |
f427d9096
|
310 |
} |
77ea73388
|
311 312 313 314 315 |
if (parent) { blkg_rwstat_add_aux(&parent->stat_bytes, &blkg->stat_bytes); blkg_rwstat_add_aux(&parent->stat_ios, &blkg->stat_ios); } |
f427d9096
|
316 |
blkg->online = false; |
a637120e4
|
317 |
radix_tree_delete(&blkcg->blkg_tree, blkg->q->id); |
e8989fae3
|
318 |
list_del_init(&blkg->q_node); |
9f13ef678
|
319 |
hlist_del_init_rcu(&blkg->blkcg_node); |
03aa264ac
|
320 |
|
03aa264ac
|
321 |
/* |
a637120e4
|
322 323 324 325 |
* Both setting lookup hint to and clearing it from @blkg are done * under queue_lock. If it's not pointing to @blkg now, it never * will. Hint assignment itself can race safely. */ |
ec6c676a0
|
326 |
if (rcu_access_pointer(blkcg->blkg_hint) == blkg) |
a637120e4
|
327 328 329 |
rcu_assign_pointer(blkcg->blkg_hint, NULL); /* |
03aa264ac
|
330 331 332 333 334 |
* Put the reference taken at the time of creation so that when all * queues are gone, group can be destroyed. */ blkg_put(blkg); } |
9f13ef678
|
335 336 337 |
/** * blkg_destroy_all - destroy all blkgs associated with a request_queue * @q: request_queue of interest |
9f13ef678
|
338 |
* |
3c96cb32d
|
339 |
* Destroy all blkgs associated with @q. |
9f13ef678
|
340 |
*/ |
3c96cb32d
|
341 |
static void blkg_destroy_all(struct request_queue *q) |
72e06c255
|
342 |
{ |
3c798398e
|
343 |
struct blkcg_gq *blkg, *n; |
72e06c255
|
344 |
|
6d18b008d
|
345 |
lockdep_assert_held(q->queue_lock); |
72e06c255
|
346 |
|
9f13ef678
|
347 |
list_for_each_entry_safe(blkg, n, &q->blkg_list, q_node) { |
3c798398e
|
348 |
struct blkcg *blkcg = blkg->blkcg; |
72e06c255
|
349 |
|
9f13ef678
|
350 351 352 |
spin_lock(&blkcg->lock); blkg_destroy(blkg); spin_unlock(&blkcg->lock); |
72e06c255
|
353 |
} |
6fe810bda
|
354 355 356 |
q->root_blkg = NULL; q->root_rl.blkg = NULL; |
72e06c255
|
357 |
} |
2a4fd070e
|
358 359 360 361 362 363 364 365 366 |
/* * A group is RCU protected, but having an rcu lock does not mean that one * can access all the fields of blkg and assume these are valid. For * example, don't try to follow throtl_data and request queue links. * * Having a reference to blkg under an rcu allows accesses to only values * local to groups like group stats and group rate limits. */ void __blkg_release_rcu(struct rcu_head *rcu_head) |
1adaf3dde
|
367 |
{ |
2a4fd070e
|
368 |
struct blkcg_gq *blkg = container_of(rcu_head, struct blkcg_gq, rcu_head); |
db6136703
|
369 |
|
3c5478659
|
370 |
/* release the blkcg and parent blkg refs this blkg has been holding */ |
1adaf3dde
|
371 |
css_put(&blkg->blkcg->css); |
a5049a8ae
|
372 |
if (blkg->parent) |
3c5478659
|
373 |
blkg_put(blkg->parent); |
1adaf3dde
|
374 |
|
ce7acfeaf
|
375 |
wb_congested_put(blkg->wb_congested); |
2a4fd070e
|
376 |
blkg_free(blkg); |
1adaf3dde
|
377 |
} |
2a4fd070e
|
378 |
EXPORT_SYMBOL_GPL(__blkg_release_rcu); |
1adaf3dde
|
379 |
|
a051661ca
|
380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 |
/* * The next function used by blk_queue_for_each_rl(). It's a bit tricky * because the root blkg uses @q->root_rl instead of its own rl. */ struct request_list *__blk_queue_next_rl(struct request_list *rl, struct request_queue *q) { struct list_head *ent; struct blkcg_gq *blkg; /* * Determine the current blkg list_head. The first entry is * root_rl which is off @q->blkg_list and mapped to the head. */ if (rl == &q->root_rl) { ent = &q->blkg_list; |
65c77fd9e
|
396 397 398 |
/* There are no more block groups, hence no request lists */ if (list_empty(ent)) return NULL; |
a051661ca
|
399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 |
} else { blkg = container_of(rl, struct blkcg_gq, rl); ent = &blkg->q_node; } /* walk to the next list_head, skip root blkcg */ ent = ent->next; if (ent == &q->root_blkg->q_node) ent = ent->next; if (ent == &q->blkg_list) return NULL; blkg = container_of(ent, struct blkcg_gq, q_node); return &blkg->rl; } |
182446d08
|
414 415 |
static int blkcg_reset_stats(struct cgroup_subsys_state *css, struct cftype *cftype, u64 val) |
303a3acb2
|
416 |
{ |
182446d08
|
417 |
struct blkcg *blkcg = css_to_blkcg(css); |
3c798398e
|
418 |
struct blkcg_gq *blkg; |
bc0d6501a
|
419 |
int i; |
303a3acb2
|
420 |
|
838f13bf4
|
421 |
mutex_lock(&blkcg_pol_mutex); |
303a3acb2
|
422 |
spin_lock_irq(&blkcg->lock); |
997a026c8
|
423 424 425 426 427 428 |
/* * Note that stat reset is racy - it doesn't synchronize against * stat updates. This is a debug feature which shouldn't exist * anyway. If you get hit by a race, retry. */ |
b67bfe0d4
|
429 |
hlist_for_each_entry(blkg, &blkcg->blkg_list, blkcg_node) { |
77ea73388
|
430 431 |
blkg_rwstat_reset(&blkg->stat_bytes); blkg_rwstat_reset(&blkg->stat_ios); |
8bd435b30
|
432 |
for (i = 0; i < BLKCG_MAX_POLS; i++) { |
3c798398e
|
433 |
struct blkcg_policy *pol = blkcg_policy[i]; |
549d3aa87
|
434 |
|
a9520cd6f
|
435 436 |
if (blkg->pd[i] && pol->pd_reset_stats_fn) pol->pd_reset_stats_fn(blkg->pd[i]); |
bc0d6501a
|
437 |
} |
303a3acb2
|
438 |
} |
f0bdc8cdd
|
439 |
|
303a3acb2
|
440 |
spin_unlock_irq(&blkcg->lock); |
bc0d6501a
|
441 |
mutex_unlock(&blkcg_pol_mutex); |
303a3acb2
|
442 443 |
return 0; } |
dd165eb3b
|
444 |
const char *blkg_dev_name(struct blkcg_gq *blkg) |
303a3acb2
|
445 |
{ |
d3d32e69f
|
446 447 448 449 |
/* some drivers (floppy) instantiate a queue w/o disk registered */ if (blkg->q->backing_dev_info.dev) return dev_name(blkg->q->backing_dev_info.dev); return NULL; |
303a3acb2
|
450 |
} |
dd165eb3b
|
451 |
EXPORT_SYMBOL_GPL(blkg_dev_name); |
303a3acb2
|
452 |
|
d3d32e69f
|
453 454 455 456 457 458 459 460 461 462 463 |
/** * blkcg_print_blkgs - helper for printing per-blkg data * @sf: seq_file to print to * @blkcg: blkcg of interest * @prfill: fill function to print out a blkg * @pol: policy in question * @data: data to be passed to @prfill * @show_total: to print out sum of prfill return values or not * * This function invokes @prfill on each blkg of @blkcg if pd for the * policy specified by @pol exists. @prfill is invoked with @sf, the |
810ecfa76
|
464 465 466 |
* policy data and @data and the matching queue lock held. If @show_total * is %true, the sum of the return values from @prfill is printed with * "Total" label at the end. |
d3d32e69f
|
467 468 469 470 |
* * This is to be used to construct print functions for * cftype->read_seq_string method. */ |
3c798398e
|
471 |
void blkcg_print_blkgs(struct seq_file *sf, struct blkcg *blkcg, |
f95a04afa
|
472 473 |
u64 (*prfill)(struct seq_file *, struct blkg_policy_data *, int), |
3c798398e
|
474 |
const struct blkcg_policy *pol, int data, |
ec399347d
|
475 |
bool show_total) |
5624a4e44
|
476 |
{ |
3c798398e
|
477 |
struct blkcg_gq *blkg; |
d3d32e69f
|
478 |
u64 total = 0; |
5624a4e44
|
479 |
|
810ecfa76
|
480 |
rcu_read_lock(); |
ee89f8125
|
481 |
hlist_for_each_entry_rcu(blkg, &blkcg->blkg_list, blkcg_node) { |
810ecfa76
|
482 |
spin_lock_irq(blkg->q->queue_lock); |
a2b1693ba
|
483 |
if (blkcg_policy_enabled(blkg->q, pol)) |
f95a04afa
|
484 |
total += prfill(sf, blkg->pd[pol->plid], data); |
810ecfa76
|
485 486 487 |
spin_unlock_irq(blkg->q->queue_lock); } rcu_read_unlock(); |
d3d32e69f
|
488 489 490 491 492 |
if (show_total) seq_printf(sf, "Total %llu ", (unsigned long long)total); } |
829fdb500
|
493 |
EXPORT_SYMBOL_GPL(blkcg_print_blkgs); |
d3d32e69f
|
494 495 496 497 |
/** * __blkg_prfill_u64 - prfill helper for a single u64 value * @sf: seq_file to print to |
f95a04afa
|
498 |
* @pd: policy private data of interest |
d3d32e69f
|
499 500 |
* @v: value to print * |
f95a04afa
|
501 |
* Print @v to @sf for the device assocaited with @pd. |
d3d32e69f
|
502 |
*/ |
f95a04afa
|
503 |
u64 __blkg_prfill_u64(struct seq_file *sf, struct blkg_policy_data *pd, u64 v) |
d3d32e69f
|
504 |
{ |
f95a04afa
|
505 |
const char *dname = blkg_dev_name(pd->blkg); |
d3d32e69f
|
506 507 508 509 510 511 512 513 |
if (!dname) return 0; seq_printf(sf, "%s %llu ", dname, (unsigned long long)v); return v; } |
829fdb500
|
514 |
EXPORT_SYMBOL_GPL(__blkg_prfill_u64); |
d3d32e69f
|
515 516 517 518 |
/** * __blkg_prfill_rwstat - prfill helper for a blkg_rwstat * @sf: seq_file to print to |
f95a04afa
|
519 |
* @pd: policy private data of interest |
d3d32e69f
|
520 521 |
* @rwstat: rwstat to print * |
f95a04afa
|
522 |
* Print @rwstat to @sf for the device assocaited with @pd. |
d3d32e69f
|
523 |
*/ |
f95a04afa
|
524 |
u64 __blkg_prfill_rwstat(struct seq_file *sf, struct blkg_policy_data *pd, |
829fdb500
|
525 |
const struct blkg_rwstat *rwstat) |
d3d32e69f
|
526 527 528 529 530 531 532 |
{ static const char *rwstr[] = { [BLKG_RWSTAT_READ] = "Read", [BLKG_RWSTAT_WRITE] = "Write", [BLKG_RWSTAT_SYNC] = "Sync", [BLKG_RWSTAT_ASYNC] = "Async", }; |
f95a04afa
|
533 |
const char *dname = blkg_dev_name(pd->blkg); |
d3d32e69f
|
534 535 536 537 538 539 540 541 542 |
u64 v; int i; if (!dname) return 0; for (i = 0; i < BLKG_RWSTAT_NR; i++) seq_printf(sf, "%s %s %llu ", dname, rwstr[i], |
24bdb8ef0
|
543 |
(unsigned long long)atomic64_read(&rwstat->aux_cnt[i])); |
d3d32e69f
|
544 |
|
24bdb8ef0
|
545 546 |
v = atomic64_read(&rwstat->aux_cnt[BLKG_RWSTAT_READ]) + atomic64_read(&rwstat->aux_cnt[BLKG_RWSTAT_WRITE]); |
d3d32e69f
|
547 548 549 550 |
seq_printf(sf, "%s Total %llu ", dname, (unsigned long long)v); return v; } |
b50da39f5
|
551 |
EXPORT_SYMBOL_GPL(__blkg_prfill_rwstat); |
d3d32e69f
|
552 |
|
5bc4afb1e
|
553 554 555 |
/** * blkg_prfill_stat - prfill callback for blkg_stat * @sf: seq_file to print to |
f95a04afa
|
556 557 |
* @pd: policy private data of interest * @off: offset to the blkg_stat in @pd |
5bc4afb1e
|
558 559 560 |
* * prfill callback for printing a blkg_stat. */ |
f95a04afa
|
561 |
u64 blkg_prfill_stat(struct seq_file *sf, struct blkg_policy_data *pd, int off) |
d3d32e69f
|
562 |
{ |
f95a04afa
|
563 |
return __blkg_prfill_u64(sf, pd, blkg_stat_read((void *)pd + off)); |
d3d32e69f
|
564 |
} |
5bc4afb1e
|
565 |
EXPORT_SYMBOL_GPL(blkg_prfill_stat); |
d3d32e69f
|
566 |
|
5bc4afb1e
|
567 568 569 |
/** * blkg_prfill_rwstat - prfill callback for blkg_rwstat * @sf: seq_file to print to |
f95a04afa
|
570 571 |
* @pd: policy private data of interest * @off: offset to the blkg_rwstat in @pd |
5bc4afb1e
|
572 573 574 |
* * prfill callback for printing a blkg_rwstat. */ |
f95a04afa
|
575 576 |
u64 blkg_prfill_rwstat(struct seq_file *sf, struct blkg_policy_data *pd, int off) |
d3d32e69f
|
577 |
{ |
f95a04afa
|
578 |
struct blkg_rwstat rwstat = blkg_rwstat_read((void *)pd + off); |
d3d32e69f
|
579 |
|
f95a04afa
|
580 |
return __blkg_prfill_rwstat(sf, pd, &rwstat); |
d3d32e69f
|
581 |
} |
5bc4afb1e
|
582 |
EXPORT_SYMBOL_GPL(blkg_prfill_rwstat); |
d3d32e69f
|
583 |
|
77ea73388
|
584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 |
static u64 blkg_prfill_rwstat_field(struct seq_file *sf, struct blkg_policy_data *pd, int off) { struct blkg_rwstat rwstat = blkg_rwstat_read((void *)pd->blkg + off); return __blkg_prfill_rwstat(sf, pd, &rwstat); } /** * blkg_print_stat_bytes - seq_show callback for blkg->stat_bytes * @sf: seq_file to print to * @v: unused * * To be used as cftype->seq_show to print blkg->stat_bytes. * cftype->private must be set to the blkcg_policy. */ int blkg_print_stat_bytes(struct seq_file *sf, void *v) { blkcg_print_blkgs(sf, css_to_blkcg(seq_css(sf)), blkg_prfill_rwstat_field, (void *)seq_cft(sf)->private, offsetof(struct blkcg_gq, stat_bytes), true); return 0; } EXPORT_SYMBOL_GPL(blkg_print_stat_bytes); /** * blkg_print_stat_bytes - seq_show callback for blkg->stat_ios * @sf: seq_file to print to * @v: unused * * To be used as cftype->seq_show to print blkg->stat_ios. cftype->private * must be set to the blkcg_policy. */ int blkg_print_stat_ios(struct seq_file *sf, void *v) { blkcg_print_blkgs(sf, css_to_blkcg(seq_css(sf)), blkg_prfill_rwstat_field, (void *)seq_cft(sf)->private, offsetof(struct blkcg_gq, stat_ios), true); return 0; } EXPORT_SYMBOL_GPL(blkg_print_stat_ios); static u64 blkg_prfill_rwstat_field_recursive(struct seq_file *sf, struct blkg_policy_data *pd, int off) { struct blkg_rwstat rwstat = blkg_rwstat_recursive_sum(pd->blkg, NULL, off); return __blkg_prfill_rwstat(sf, pd, &rwstat); } /** * blkg_print_stat_bytes_recursive - recursive version of blkg_print_stat_bytes * @sf: seq_file to print to * @v: unused */ int blkg_print_stat_bytes_recursive(struct seq_file *sf, void *v) { blkcg_print_blkgs(sf, css_to_blkcg(seq_css(sf)), blkg_prfill_rwstat_field_recursive, (void *)seq_cft(sf)->private, offsetof(struct blkcg_gq, stat_bytes), true); return 0; } EXPORT_SYMBOL_GPL(blkg_print_stat_bytes_recursive); /** * blkg_print_stat_ios_recursive - recursive version of blkg_print_stat_ios * @sf: seq_file to print to * @v: unused */ int blkg_print_stat_ios_recursive(struct seq_file *sf, void *v) { blkcg_print_blkgs(sf, css_to_blkcg(seq_css(sf)), blkg_prfill_rwstat_field_recursive, (void *)seq_cft(sf)->private, offsetof(struct blkcg_gq, stat_ios), true); return 0; } EXPORT_SYMBOL_GPL(blkg_print_stat_ios_recursive); |
3a8b31d39
|
664 |
/** |
16b3de665
|
665 |
* blkg_stat_recursive_sum - collect hierarchical blkg_stat |
f12c74cab
|
666 667 668 |
* @blkg: blkg of interest * @pol: blkcg_policy which contains the blkg_stat * @off: offset to the blkg_stat in blkg_policy_data or @blkg |
16b3de665
|
669 |
* |
f12c74cab
|
670 671 672 673 674 675 |
* Collect the blkg_stat specified by @blkg, @pol and @off and all its * online descendants and their aux counts. The caller must be holding the * queue lock for online tests. * * If @pol is NULL, blkg_stat is at @off bytes into @blkg; otherwise, it is * at @off bytes into @blkg's blkg_policy_data of the policy. |
16b3de665
|
676 |
*/ |
f12c74cab
|
677 678 |
u64 blkg_stat_recursive_sum(struct blkcg_gq *blkg, struct blkcg_policy *pol, int off) |
16b3de665
|
679 |
{ |
16b3de665
|
680 |
struct blkcg_gq *pos_blkg; |
492eb21b9
|
681 |
struct cgroup_subsys_state *pos_css; |
bd8815a6d
|
682 |
u64 sum = 0; |
16b3de665
|
683 |
|
f12c74cab
|
684 |
lockdep_assert_held(blkg->q->queue_lock); |
16b3de665
|
685 |
|
16b3de665
|
686 |
rcu_read_lock(); |
f12c74cab
|
687 688 689 690 691 |
blkg_for_each_descendant_pre(pos_blkg, pos_css, blkg) { struct blkg_stat *stat; if (!pos_blkg->online) continue; |
16b3de665
|
692 |
|
f12c74cab
|
693 694 695 696 697 698 |
if (pol) stat = (void *)blkg_to_pd(pos_blkg, pol) + off; else stat = (void *)blkg + off; sum += blkg_stat_read(stat) + atomic64_read(&stat->aux_cnt); |
16b3de665
|
699 700 701 702 703 704 705 706 707 |
} rcu_read_unlock(); return sum; } EXPORT_SYMBOL_GPL(blkg_stat_recursive_sum); /** * blkg_rwstat_recursive_sum - collect hierarchical blkg_rwstat |
f12c74cab
|
708 709 710 |
* @blkg: blkg of interest * @pol: blkcg_policy which contains the blkg_rwstat * @off: offset to the blkg_rwstat in blkg_policy_data or @blkg |
16b3de665
|
711 |
* |
f12c74cab
|
712 713 714 715 716 717 |
* Collect the blkg_rwstat specified by @blkg, @pol and @off and all its * online descendants and their aux counts. The caller must be holding the * queue lock for online tests. * * If @pol is NULL, blkg_rwstat is at @off bytes into @blkg; otherwise, it * is at @off bytes into @blkg's blkg_policy_data of the policy. |
16b3de665
|
718 |
*/ |
f12c74cab
|
719 720 |
struct blkg_rwstat blkg_rwstat_recursive_sum(struct blkcg_gq *blkg, struct blkcg_policy *pol, int off) |
16b3de665
|
721 |
{ |
16b3de665
|
722 |
struct blkcg_gq *pos_blkg; |
492eb21b9
|
723 |
struct cgroup_subsys_state *pos_css; |
bd8815a6d
|
724 |
struct blkg_rwstat sum = { }; |
16b3de665
|
725 |
int i; |
f12c74cab
|
726 |
lockdep_assert_held(blkg->q->queue_lock); |
16b3de665
|
727 |
|
16b3de665
|
728 |
rcu_read_lock(); |
f12c74cab
|
729 |
blkg_for_each_descendant_pre(pos_blkg, pos_css, blkg) { |
3a7faeada
|
730 |
struct blkg_rwstat *rwstat; |
16b3de665
|
731 732 733 |
if (!pos_blkg->online) continue; |
f12c74cab
|
734 735 736 737 |
if (pol) rwstat = (void *)blkg_to_pd(pos_blkg, pol) + off; else rwstat = (void *)pos_blkg + off; |
16b3de665
|
738 |
for (i = 0; i < BLKG_RWSTAT_NR; i++) |
3a7faeada
|
739 740 741 |
atomic64_add(atomic64_read(&rwstat->aux_cnt[i]) + percpu_counter_sum_positive(&rwstat->cpu_cnt[i]), &sum.aux_cnt[i]); |
16b3de665
|
742 743 744 745 746 747 748 749 |
} rcu_read_unlock(); return sum; } EXPORT_SYMBOL_GPL(blkg_rwstat_recursive_sum); /** |
3a8b31d39
|
750 751 |
* blkg_conf_prep - parse and prepare for per-blkg config update * @blkcg: target block cgroup |
da8b06626
|
752 |
* @pol: target policy |
3a8b31d39
|
753 754 755 756 |
* @input: input string * @ctx: blkg_conf_ctx to be filled * * Parse per-blkg config update from @input and initialize @ctx with the |
36aa9e5f5
|
757 758 759 |
* result. @ctx->blkg points to the blkg to be updated and @ctx->body the * part of @input following MAJ:MIN. This function returns with RCU read * lock and queue lock held and must be paired with blkg_conf_finish(). |
3a8b31d39
|
760 |
*/ |
3c798398e
|
761 |
int blkg_conf_prep(struct blkcg *blkcg, const struct blkcg_policy *pol, |
36aa9e5f5
|
762 |
char *input, struct blkg_conf_ctx *ctx) |
da8b06626
|
763 |
__acquires(rcu) __acquires(disk->queue->queue_lock) |
34d0f179d
|
764 |
{ |
3a8b31d39
|
765 |
struct gendisk *disk; |
3c798398e
|
766 |
struct blkcg_gq *blkg; |
39a169b62
|
767 |
struct module *owner; |
726fa6945
|
768 |
unsigned int major, minor; |
36aa9e5f5
|
769 770 |
int key_len, part, ret; char *body; |
34d0f179d
|
771 |
|
36aa9e5f5
|
772 |
if (sscanf(input, "%u:%u%n", &major, &minor, &key_len) != 2) |
726fa6945
|
773 |
return -EINVAL; |
3a8b31d39
|
774 |
|
36aa9e5f5
|
775 776 777 778 |
body = input + key_len; if (!isspace(*body)) return -EINVAL; body = skip_spaces(body); |
726fa6945
|
779 |
disk = get_gendisk(MKDEV(major, minor), &part); |
5f6c2d2b7
|
780 |
if (!disk) |
20386ce01
|
781 |
return -ENODEV; |
5f6c2d2b7
|
782 |
if (part) { |
39a169b62
|
783 |
owner = disk->fops->owner; |
5f6c2d2b7
|
784 |
put_disk(disk); |
39a169b62
|
785 |
module_put(owner); |
20386ce01
|
786 |
return -ENODEV; |
5f6c2d2b7
|
787 |
} |
e56da7e28
|
788 789 |
rcu_read_lock(); |
4bfd482e7
|
790 |
spin_lock_irq(disk->queue->queue_lock); |
da8b06626
|
791 |
|
a2b1693ba
|
792 |
if (blkcg_policy_enabled(disk->queue, pol)) |
3c96cb32d
|
793 |
blkg = blkg_lookup_create(blkcg, disk->queue); |
a2b1693ba
|
794 |
else |
20386ce01
|
795 |
blkg = ERR_PTR(-EOPNOTSUPP); |
e56da7e28
|
796 |
|
4bfd482e7
|
797 798 |
if (IS_ERR(blkg)) { ret = PTR_ERR(blkg); |
3a8b31d39
|
799 |
rcu_read_unlock(); |
da8b06626
|
800 |
spin_unlock_irq(disk->queue->queue_lock); |
39a169b62
|
801 |
owner = disk->fops->owner; |
3a8b31d39
|
802 |
put_disk(disk); |
39a169b62
|
803 |
module_put(owner); |
3a8b31d39
|
804 805 806 807 808 809 810 811 812 |
/* * If queue was bypassing, we should retry. Do so after a * short msleep(). It isn't strictly necessary but queue * can be bypassing for some time and it's always nice to * avoid busy looping. */ if (ret == -EBUSY) { msleep(10); ret = restart_syscall(); |
7702e8f45
|
813 |
} |
726fa6945
|
814 |
return ret; |
062a644d6
|
815 |
} |
3a8b31d39
|
816 817 818 |
ctx->disk = disk; ctx->blkg = blkg; |
36aa9e5f5
|
819 |
ctx->body = body; |
726fa6945
|
820 |
return 0; |
34d0f179d
|
821 |
} |
829fdb500
|
822 |
EXPORT_SYMBOL_GPL(blkg_conf_prep); |
34d0f179d
|
823 |
|
3a8b31d39
|
824 825 826 827 828 829 830 |
/** * blkg_conf_finish - finish up per-blkg config update * @ctx: blkg_conf_ctx intiailized by blkg_conf_prep() * * Finish up after per-blkg config update. This function must be paired * with blkg_conf_prep(). */ |
829fdb500
|
831 |
void blkg_conf_finish(struct blkg_conf_ctx *ctx) |
da8b06626
|
832 |
__releases(ctx->disk->queue->queue_lock) __releases(rcu) |
34d0f179d
|
833 |
{ |
39a169b62
|
834 |
struct module *owner; |
da8b06626
|
835 |
spin_unlock_irq(ctx->disk->queue->queue_lock); |
3a8b31d39
|
836 |
rcu_read_unlock(); |
39a169b62
|
837 |
owner = ctx->disk->fops->owner; |
3a8b31d39
|
838 |
put_disk(ctx->disk); |
39a169b62
|
839 |
module_put(owner); |
34d0f179d
|
840 |
} |
829fdb500
|
841 |
EXPORT_SYMBOL_GPL(blkg_conf_finish); |
34d0f179d
|
842 |
|
2ee867dcf
|
843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 |
static int blkcg_print_stat(struct seq_file *sf, void *v) { struct blkcg *blkcg = css_to_blkcg(seq_css(sf)); struct blkcg_gq *blkg; rcu_read_lock(); hlist_for_each_entry_rcu(blkg, &blkcg->blkg_list, blkcg_node) { const char *dname; struct blkg_rwstat rwstat; u64 rbytes, wbytes, rios, wios; dname = blkg_dev_name(blkg); if (!dname) continue; spin_lock_irq(blkg->q->queue_lock); rwstat = blkg_rwstat_recursive_sum(blkg, NULL, offsetof(struct blkcg_gq, stat_bytes)); rbytes = atomic64_read(&rwstat.aux_cnt[BLKG_RWSTAT_READ]); wbytes = atomic64_read(&rwstat.aux_cnt[BLKG_RWSTAT_WRITE]); rwstat = blkg_rwstat_recursive_sum(blkg, NULL, offsetof(struct blkcg_gq, stat_ios)); rios = atomic64_read(&rwstat.aux_cnt[BLKG_RWSTAT_READ]); wios = atomic64_read(&rwstat.aux_cnt[BLKG_RWSTAT_WRITE]); spin_unlock_irq(blkg->q->queue_lock); if (rbytes || wbytes || rios || wios) seq_printf(sf, "%s rbytes=%llu wbytes=%llu rios=%llu wios=%llu ", dname, rbytes, wbytes, rios, wios); } rcu_read_unlock(); return 0; } |
e1f3b9412
|
882 |
static struct cftype blkcg_files[] = { |
2ee867dcf
|
883 884 |
{ .name = "stat", |
ca0752c5e
|
885 |
.flags = CFTYPE_NOT_ON_ROOT, |
2ee867dcf
|
886 887 888 889 |
.seq_show = blkcg_print_stat, }, { } /* terminate */ }; |
e1f3b9412
|
890 |
static struct cftype blkcg_legacy_files[] = { |
31e4c28d9
|
891 |
{ |
84c124da9
|
892 |
.name = "reset_stats", |
3c798398e
|
893 |
.write_u64 = blkcg_reset_stats, |
220841906
|
894 |
}, |
4baf6e332
|
895 |
{ } /* terminate */ |
31e4c28d9
|
896 |
}; |
9f13ef678
|
897 |
/** |
92fb97487
|
898 |
* blkcg_css_offline - cgroup css_offline callback |
eb95419b0
|
899 |
* @css: css of interest |
9f13ef678
|
900 |
* |
eb95419b0
|
901 902 |
* This function is called when @css is about to go away and responsible * for shooting down all blkgs associated with @css. blkgs should be |
9f13ef678
|
903 904 905 906 907 |
* removed while holding both q and blkcg locks. As blkcg lock is nested * inside q lock, this function performs reverse double lock dancing. * * This is the blkcg counterpart of ioc_release_fn(). */ |
eb95419b0
|
908 |
static void blkcg_css_offline(struct cgroup_subsys_state *css) |
31e4c28d9
|
909 |
{ |
eb95419b0
|
910 |
struct blkcg *blkcg = css_to_blkcg(css); |
b1c357696
|
911 |
|
9f13ef678
|
912 |
spin_lock_irq(&blkcg->lock); |
7ee9c5620
|
913 |
|
9f13ef678
|
914 |
while (!hlist_empty(&blkcg->blkg_list)) { |
3c798398e
|
915 916 |
struct blkcg_gq *blkg = hlist_entry(blkcg->blkg_list.first, struct blkcg_gq, blkcg_node); |
c875f4d02
|
917 |
struct request_queue *q = blkg->q; |
b1c357696
|
918 |
|
9f13ef678
|
919 920 921 922 923 |
if (spin_trylock(q->queue_lock)) { blkg_destroy(blkg); spin_unlock(q->queue_lock); } else { spin_unlock_irq(&blkcg->lock); |
9f13ef678
|
924 |
cpu_relax(); |
a5567932f
|
925 |
spin_lock_irq(&blkcg->lock); |
0f3942a39
|
926 |
} |
9f13ef678
|
927 |
} |
b1c357696
|
928 |
|
9f13ef678
|
929 |
spin_unlock_irq(&blkcg->lock); |
52ebea749
|
930 931 |
wb_blkcg_offline(blkcg); |
7ee9c5620
|
932 |
} |
eb95419b0
|
933 |
static void blkcg_css_free(struct cgroup_subsys_state *css) |
7ee9c5620
|
934 |
{ |
eb95419b0
|
935 |
struct blkcg *blkcg = css_to_blkcg(css); |
bc915e61c
|
936 |
int i; |
7ee9c5620
|
937 |
|
7876f930d
|
938 |
mutex_lock(&blkcg_pol_mutex); |
e4a9bde95
|
939 |
|
7876f930d
|
940 |
list_del(&blkcg->all_blkcgs_node); |
7876f930d
|
941 |
|
bc915e61c
|
942 |
for (i = 0; i < BLKCG_MAX_POLS; i++) |
e4a9bde95
|
943 944 945 946 |
if (blkcg->cpd[i]) blkcg_policy[i]->cpd_free_fn(blkcg->cpd[i]); mutex_unlock(&blkcg_pol_mutex); |
bc915e61c
|
947 |
kfree(blkcg); |
31e4c28d9
|
948 |
} |
eb95419b0
|
949 950 |
static struct cgroup_subsys_state * blkcg_css_alloc(struct cgroup_subsys_state *parent_css) |
31e4c28d9
|
951 |
{ |
3c798398e
|
952 |
struct blkcg *blkcg; |
e48453c38
|
953 954 |
struct cgroup_subsys_state *ret; int i; |
31e4c28d9
|
955 |
|
7876f930d
|
956 |
mutex_lock(&blkcg_pol_mutex); |
eb95419b0
|
957 |
if (!parent_css) { |
3c798398e
|
958 |
blkcg = &blkcg_root; |
bc915e61c
|
959 960 961 962 963 964 |
} else { blkcg = kzalloc(sizeof(*blkcg), GFP_KERNEL); if (!blkcg) { ret = ERR_PTR(-ENOMEM); goto free_blkcg; } |
e48453c38
|
965 966 967 968 969 970 971 972 973 974 975 976 |
} for (i = 0; i < BLKCG_MAX_POLS ; i++) { struct blkcg_policy *pol = blkcg_policy[i]; struct blkcg_policy_data *cpd; /* * If the policy hasn't been attached yet, wait for it * to be attached before doing anything else. Otherwise, * check if the policy requires any specific per-cgroup * data: if it does, allocate and initialize it. */ |
e4a9bde95
|
977 |
if (!pol || !pol->cpd_alloc_fn) |
e48453c38
|
978 |
continue; |
e4a9bde95
|
979 |
cpd = pol->cpd_alloc_fn(GFP_KERNEL); |
e48453c38
|
980 981 982 983 |
if (!cpd) { ret = ERR_PTR(-ENOMEM); goto free_pd_blkcg; } |
814376483
|
984 985 |
blkcg->cpd[i] = cpd; cpd->blkcg = blkcg; |
e48453c38
|
986 |
cpd->plid = i; |
e4a9bde95
|
987 988 |
if (pol->cpd_init_fn) pol->cpd_init_fn(cpd); |
e48453c38
|
989 |
} |
31e4c28d9
|
990 |
|
31e4c28d9
|
991 |
spin_lock_init(&blkcg->lock); |
d93a11f1c
|
992 |
INIT_RADIX_TREE(&blkcg->blkg_tree, GFP_NOWAIT); |
31e4c28d9
|
993 |
INIT_HLIST_HEAD(&blkcg->blkg_list); |
52ebea749
|
994 995 996 |
#ifdef CONFIG_CGROUP_WRITEBACK INIT_LIST_HEAD(&blkcg->cgwb_list); #endif |
7876f930d
|
997 998 999 |
list_add_tail(&blkcg->all_blkcgs_node, &all_blkcgs); mutex_unlock(&blkcg_pol_mutex); |
31e4c28d9
|
1000 |
return &blkcg->css; |
e48453c38
|
1001 1002 1003 |
free_pd_blkcg: for (i--; i >= 0; i--) |
e4a9bde95
|
1004 1005 |
if (blkcg->cpd[i]) blkcg_policy[i]->cpd_free_fn(blkcg->cpd[i]); |
e48453c38
|
1006 1007 |
free_blkcg: kfree(blkcg); |
7876f930d
|
1008 |
mutex_unlock(&blkcg_pol_mutex); |
e48453c38
|
1009 |
return ret; |
31e4c28d9
|
1010 |
} |
5efd61135
|
1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 |
/** * blkcg_init_queue - initialize blkcg part of request queue * @q: request_queue to initialize * * Called from blk_alloc_queue_node(). Responsible for initializing blkcg * part of new request_queue @q. * * RETURNS: * 0 on success, -errno on failure. */ int blkcg_init_queue(struct request_queue *q) { |
ec13b1d6f
|
1023 1024 1025 1026 1027 1028 1029 1030 1031 |
struct blkcg_gq *new_blkg, *blkg; bool preloaded; int ret; new_blkg = blkg_alloc(&blkcg_root, q, GFP_KERNEL); if (!new_blkg) return -ENOMEM; preloaded = !radix_tree_preload(GFP_KERNEL); |
5efd61135
|
1032 |
|
ec13b1d6f
|
1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 |
/* * Make sure the root blkg exists and count the existing blkgs. As * @q is bypassing at this point, blkg_lookup_create() can't be * used. Open code insertion. */ rcu_read_lock(); spin_lock_irq(q->queue_lock); blkg = blkg_create(&blkcg_root, q, new_blkg); spin_unlock_irq(q->queue_lock); rcu_read_unlock(); if (preloaded) radix_tree_preload_end(); if (IS_ERR(blkg)) { |
994b78327
|
1048 |
blkg_free(new_blkg); |
ec13b1d6f
|
1049 1050 1051 1052 1053 |
return PTR_ERR(blkg); } q->root_blkg = blkg; q->root_rl.blkg = blkg; |
5efd61135
|
1054 |
|
ec13b1d6f
|
1055 1056 1057 1058 1059 1060 1061 |
ret = blk_throtl_init(q); if (ret) { spin_lock_irq(q->queue_lock); blkg_destroy_all(q); spin_unlock_irq(q->queue_lock); } return ret; |
5efd61135
|
1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 |
} /** * blkcg_drain_queue - drain blkcg part of request_queue * @q: request_queue to drain * * Called from blk_drain_queue(). Responsible for draining blkcg part. */ void blkcg_drain_queue(struct request_queue *q) { lockdep_assert_held(q->queue_lock); |
0b462c89e
|
1073 1074 1075 1076 1077 1078 |
/* * @q could be exiting and already have destroyed all blkgs as * indicated by NULL root_blkg. If so, don't confuse policies. */ if (!q->root_blkg) return; |
5efd61135
|
1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 |
blk_throtl_drain(q); } /** * blkcg_exit_queue - exit and release blkcg part of request_queue * @q: request_queue being released * * Called from blk_release_queue(). Responsible for exiting blkcg part. */ void blkcg_exit_queue(struct request_queue *q) { |
6d18b008d
|
1090 |
spin_lock_irq(q->queue_lock); |
3c96cb32d
|
1091 |
blkg_destroy_all(q); |
6d18b008d
|
1092 |
spin_unlock_irq(q->queue_lock); |
5efd61135
|
1093 1094 |
blk_throtl_exit(q); } |
31e4c28d9
|
1095 1096 1097 1098 1099 1100 |
/* * We cannot support shared io contexts, as we have no mean to support * two tasks with the same ioc in two different groups without major rework * of the main cic data structures. For now we allow a task to change * its cgroup only if it's the only owner of its ioc. */ |
1f7dd3e5a
|
1101 |
static int blkcg_can_attach(struct cgroup_taskset *tset) |
31e4c28d9
|
1102 |
{ |
bb9d97b6d
|
1103 |
struct task_struct *task; |
1f7dd3e5a
|
1104 |
struct cgroup_subsys_state *dst_css; |
31e4c28d9
|
1105 1106 1107 1108 |
struct io_context *ioc; int ret = 0; /* task_lock() is needed to avoid races with exit_io_context() */ |
1f7dd3e5a
|
1109 |
cgroup_taskset_for_each(task, dst_css, tset) { |
bb9d97b6d
|
1110 1111 1112 1113 1114 1115 1116 1117 |
task_lock(task); ioc = task->io_context; if (ioc && atomic_read(&ioc->nr_tasks) > 1) ret = -EINVAL; task_unlock(task); if (ret) break; } |
31e4c28d9
|
1118 1119 |
return ret; } |
69d7fde59
|
1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 |
static void blkcg_bind(struct cgroup_subsys_state *root_css) { int i; mutex_lock(&blkcg_pol_mutex); for (i = 0; i < BLKCG_MAX_POLS; i++) { struct blkcg_policy *pol = blkcg_policy[i]; struct blkcg *blkcg; if (!pol || !pol->cpd_bind_fn) continue; list_for_each_entry(blkcg, &all_blkcgs, all_blkcgs_node) if (blkcg->cpd[pol->plid]) pol->cpd_bind_fn(blkcg->cpd[pol->plid]); } mutex_unlock(&blkcg_pol_mutex); } |
c165b3e3c
|
1139 |
struct cgroup_subsys io_cgrp_subsys = { |
92fb97487
|
1140 1141 1142 |
.css_alloc = blkcg_css_alloc, .css_offline = blkcg_css_offline, .css_free = blkcg_css_free, |
3c798398e
|
1143 |
.can_attach = blkcg_can_attach, |
69d7fde59
|
1144 |
.bind = blkcg_bind, |
2ee867dcf
|
1145 |
.dfl_cftypes = blkcg_files, |
880f50e22
|
1146 |
.legacy_cftypes = blkcg_legacy_files, |
c165b3e3c
|
1147 |
.legacy_name = "blkio", |
1ced953b1
|
1148 1149 1150 1151 1152 1153 1154 1155 |
#ifdef CONFIG_MEMCG /* * This ensures that, if available, memcg is automatically enabled * together on the default hierarchy so that the owner cgroup can * be retrieved from writeback pages. */ .depends_on = 1 << memory_cgrp_id, #endif |
676f7c8f8
|
1156 |
}; |
c165b3e3c
|
1157 |
EXPORT_SYMBOL_GPL(io_cgrp_subsys); |
676f7c8f8
|
1158 |
|
8bd435b30
|
1159 |
/** |
a2b1693ba
|
1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 |
* blkcg_activate_policy - activate a blkcg policy on a request_queue * @q: request_queue of interest * @pol: blkcg policy to activate * * Activate @pol on @q. Requires %GFP_KERNEL context. @q goes through * bypass mode to populate its blkgs with policy_data for @pol. * * Activation happens with @q bypassed, so nobody would be accessing blkgs * from IO path. Update of each blkg is protected by both queue and blkcg * locks so that holding either lock and testing blkcg_policy_enabled() is * always enough for dereferencing policy data. * * The caller is responsible for synchronizing [de]activations and policy * [un]registerations. Returns 0 on success, -errno on failure. */ int blkcg_activate_policy(struct request_queue *q, |
3c798398e
|
1176 |
const struct blkcg_policy *pol) |
a2b1693ba
|
1177 |
{ |
4c55f4f9a
|
1178 |
struct blkg_policy_data *pd_prealloc = NULL; |
ec13b1d6f
|
1179 |
struct blkcg_gq *blkg; |
4c55f4f9a
|
1180 |
int ret; |
a2b1693ba
|
1181 1182 1183 1184 1185 |
if (blkcg_policy_enabled(q, pol)) return 0; blk_queue_bypass_start(q); |
4c55f4f9a
|
1186 1187 |
pd_prealloc: if (!pd_prealloc) { |
001bea73e
|
1188 |
pd_prealloc = pol->pd_alloc_fn(GFP_KERNEL, q->node); |
4c55f4f9a
|
1189 |
if (!pd_prealloc) { |
a2b1693ba
|
1190 |
ret = -ENOMEM; |
4c55f4f9a
|
1191 |
goto out_bypass_end; |
a2b1693ba
|
1192 |
} |
a2b1693ba
|
1193 |
} |
a2b1693ba
|
1194 1195 1196 |
spin_lock_irq(q->queue_lock); list_for_each_entry(blkg, &q->blkg_list, q_node) { |
4c55f4f9a
|
1197 1198 1199 1200 |
struct blkg_policy_data *pd; if (blkg->pd[pol->plid]) continue; |
a2b1693ba
|
1201 |
|
001bea73e
|
1202 |
pd = pol->pd_alloc_fn(GFP_NOWAIT, q->node); |
4c55f4f9a
|
1203 1204 1205 1206 1207 1208 |
if (!pd) swap(pd, pd_prealloc); if (!pd) { spin_unlock_irq(q->queue_lock); goto pd_prealloc; } |
a2b1693ba
|
1209 1210 1211 |
blkg->pd[pol->plid] = pd; pd->blkg = blkg; |
b276a876a
|
1212 |
pd->plid = pol->plid; |
3e4187104
|
1213 |
if (pol->pd_init_fn) |
a9520cd6f
|
1214 |
pol->pd_init_fn(pd); |
a2b1693ba
|
1215 1216 1217 1218 |
} __set_bit(pol->plid, q->blkcg_pols); ret = 0; |
4c55f4f9a
|
1219 |
|
a2b1693ba
|
1220 |
spin_unlock_irq(q->queue_lock); |
4c55f4f9a
|
1221 |
out_bypass_end: |
a2b1693ba
|
1222 |
blk_queue_bypass_end(q); |
001bea73e
|
1223 1224 |
if (pd_prealloc) pol->pd_free_fn(pd_prealloc); |
a2b1693ba
|
1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 |
return ret; } EXPORT_SYMBOL_GPL(blkcg_activate_policy); /** * blkcg_deactivate_policy - deactivate a blkcg policy on a request_queue * @q: request_queue of interest * @pol: blkcg policy to deactivate * * Deactivate @pol on @q. Follows the same synchronization rules as * blkcg_activate_policy(). */ void blkcg_deactivate_policy(struct request_queue *q, |
3c798398e
|
1238 |
const struct blkcg_policy *pol) |
a2b1693ba
|
1239 |
{ |
3c798398e
|
1240 |
struct blkcg_gq *blkg; |
a2b1693ba
|
1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 |
if (!blkcg_policy_enabled(q, pol)) return; blk_queue_bypass_start(q); spin_lock_irq(q->queue_lock); __clear_bit(pol->plid, q->blkcg_pols); list_for_each_entry(blkg, &q->blkg_list, q_node) { /* grab blkcg lock too while removing @pd from @blkg */ spin_lock(&blkg->blkcg->lock); |
001bea73e
|
1253 |
if (blkg->pd[pol->plid]) { |
a9520cd6f
|
1254 1255 |
if (pol->pd_offline_fn) pol->pd_offline_fn(blkg->pd[pol->plid]); |
001bea73e
|
1256 1257 1258 |
pol->pd_free_fn(blkg->pd[pol->plid]); blkg->pd[pol->plid] = NULL; } |
a2b1693ba
|
1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 |
spin_unlock(&blkg->blkcg->lock); } spin_unlock_irq(q->queue_lock); blk_queue_bypass_end(q); } EXPORT_SYMBOL_GPL(blkcg_deactivate_policy); /** |
3c798398e
|
1269 1270 |
* blkcg_policy_register - register a blkcg policy * @pol: blkcg policy to register |
8bd435b30
|
1271 |
* |
3c798398e
|
1272 1273 |
* Register @pol with blkcg core. Might sleep and @pol may be modified on * successful registration. Returns 0 on success and -errno on failure. |
8bd435b30
|
1274 |
*/ |
d5bf02914
|
1275 |
int blkcg_policy_register(struct blkcg_policy *pol) |
3e2520668
|
1276 |
{ |
06b285bd1
|
1277 |
struct blkcg *blkcg; |
8bd435b30
|
1278 |
int i, ret; |
e8989fae3
|
1279 |
|
838f13bf4
|
1280 |
mutex_lock(&blkcg_pol_register_mutex); |
bc0d6501a
|
1281 |
mutex_lock(&blkcg_pol_mutex); |
8bd435b30
|
1282 1283 1284 |
/* find an empty slot */ ret = -ENOSPC; for (i = 0; i < BLKCG_MAX_POLS; i++) |
3c798398e
|
1285 |
if (!blkcg_policy[i]) |
8bd435b30
|
1286 1287 |
break; if (i >= BLKCG_MAX_POLS) |
838f13bf4
|
1288 |
goto err_unlock; |
035d10b2f
|
1289 |
|
06b285bd1
|
1290 |
/* register @pol */ |
3c798398e
|
1291 |
pol->plid = i; |
06b285bd1
|
1292 1293 1294 |
blkcg_policy[pol->plid] = pol; /* allocate and install cpd's */ |
e4a9bde95
|
1295 |
if (pol->cpd_alloc_fn) { |
06b285bd1
|
1296 1297 |
list_for_each_entry(blkcg, &all_blkcgs, all_blkcgs_node) { struct blkcg_policy_data *cpd; |
e4a9bde95
|
1298 |
cpd = pol->cpd_alloc_fn(GFP_KERNEL); |
bbb427e34
|
1299 |
if (!cpd) |
06b285bd1
|
1300 |
goto err_free_cpds; |
06b285bd1
|
1301 |
|
814376483
|
1302 1303 |
blkcg->cpd[pol->plid] = cpd; cpd->blkcg = blkcg; |
06b285bd1
|
1304 |
cpd->plid = pol->plid; |
814376483
|
1305 |
pol->cpd_init_fn(cpd); |
06b285bd1
|
1306 1307 |
} } |
838f13bf4
|
1308 |
mutex_unlock(&blkcg_pol_mutex); |
8bd435b30
|
1309 |
|
8bd435b30
|
1310 |
/* everything is in place, add intf files for the new policy */ |
2ee867dcf
|
1311 1312 1313 |
if (pol->dfl_cftypes) WARN_ON(cgroup_add_dfl_cftypes(&io_cgrp_subsys, pol->dfl_cftypes)); |
880f50e22
|
1314 |
if (pol->legacy_cftypes) |
c165b3e3c
|
1315 |
WARN_ON(cgroup_add_legacy_cftypes(&io_cgrp_subsys, |
880f50e22
|
1316 |
pol->legacy_cftypes)); |
838f13bf4
|
1317 1318 |
mutex_unlock(&blkcg_pol_register_mutex); return 0; |
06b285bd1
|
1319 |
err_free_cpds: |
e4a9bde95
|
1320 |
if (pol->cpd_alloc_fn) { |
06b285bd1
|
1321 |
list_for_each_entry(blkcg, &all_blkcgs, all_blkcgs_node) { |
e4a9bde95
|
1322 1323 1324 1325 |
if (blkcg->cpd[pol->plid]) { pol->cpd_free_fn(blkcg->cpd[pol->plid]); blkcg->cpd[pol->plid] = NULL; } |
06b285bd1
|
1326 1327 1328 |
} } blkcg_policy[pol->plid] = NULL; |
838f13bf4
|
1329 |
err_unlock: |
bc0d6501a
|
1330 |
mutex_unlock(&blkcg_pol_mutex); |
838f13bf4
|
1331 |
mutex_unlock(&blkcg_pol_register_mutex); |
8bd435b30
|
1332 |
return ret; |
3e2520668
|
1333 |
} |
3c798398e
|
1334 |
EXPORT_SYMBOL_GPL(blkcg_policy_register); |
3e2520668
|
1335 |
|
8bd435b30
|
1336 |
/** |
3c798398e
|
1337 1338 |
* blkcg_policy_unregister - unregister a blkcg policy * @pol: blkcg policy to unregister |
8bd435b30
|
1339 |
* |
3c798398e
|
1340 |
* Undo blkcg_policy_register(@pol). Might sleep. |
8bd435b30
|
1341 |
*/ |
3c798398e
|
1342 |
void blkcg_policy_unregister(struct blkcg_policy *pol) |
3e2520668
|
1343 |
{ |
06b285bd1
|
1344 |
struct blkcg *blkcg; |
838f13bf4
|
1345 |
mutex_lock(&blkcg_pol_register_mutex); |
bc0d6501a
|
1346 |
|
3c798398e
|
1347 |
if (WARN_ON(blkcg_policy[pol->plid] != pol)) |
8bd435b30
|
1348 1349 1350 |
goto out_unlock; /* kill the intf files first */ |
2ee867dcf
|
1351 1352 |
if (pol->dfl_cftypes) cgroup_rm_cftypes(pol->dfl_cftypes); |
880f50e22
|
1353 1354 |
if (pol->legacy_cftypes) cgroup_rm_cftypes(pol->legacy_cftypes); |
44ea53de4
|
1355 |
|
06b285bd1
|
1356 |
/* remove cpds and unregister */ |
838f13bf4
|
1357 |
mutex_lock(&blkcg_pol_mutex); |
06b285bd1
|
1358 |
|
e4a9bde95
|
1359 |
if (pol->cpd_alloc_fn) { |
06b285bd1
|
1360 |
list_for_each_entry(blkcg, &all_blkcgs, all_blkcgs_node) { |
e4a9bde95
|
1361 1362 1363 1364 |
if (blkcg->cpd[pol->plid]) { pol->cpd_free_fn(blkcg->cpd[pol->plid]); blkcg->cpd[pol->plid] = NULL; } |
06b285bd1
|
1365 1366 |
} } |
3c798398e
|
1367 |
blkcg_policy[pol->plid] = NULL; |
06b285bd1
|
1368 |
|
bc0d6501a
|
1369 |
mutex_unlock(&blkcg_pol_mutex); |
838f13bf4
|
1370 1371 |
out_unlock: mutex_unlock(&blkcg_pol_register_mutex); |
3e2520668
|
1372 |
} |
3c798398e
|
1373 |
EXPORT_SYMBOL_GPL(blkcg_policy_unregister); |