Blame view
block/blk-cgroup.c
50 KB
3dcf60bcb
|
1 |
// SPDX-License-Identifier: GPL-2.0 |
31e4c28d9
|
2 3 4 5 6 7 8 9 10 11 12 |
/* * 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
|
13 14 15 16 |
* * For policy-specific per-blkcg data: * Copyright (C) 2015 Paolo Valente <paolo.valente@unimore.it> * Arianna Avanzini <avanzini.arianna@gmail.com> |
31e4c28d9
|
17 18 |
*/ #include <linux/ioprio.h> |
220841906
|
19 |
#include <linux/kdev_t.h> |
9d6a986c0
|
20 |
#include <linux/module.h> |
174cd4b1e
|
21 |
#include <linux/sched/signal.h> |
accee7854
|
22 |
#include <linux/err.h> |
9195291e5
|
23 |
#include <linux/blkdev.h> |
52ebea749
|
24 |
#include <linux/backing-dev.h> |
5a0e3ad6a
|
25 |
#include <linux/slab.h> |
34d0f179d
|
26 |
#include <linux/genhd.h> |
72e06c255
|
27 |
#include <linux/delay.h> |
9a9e8a26d
|
28 |
#include <linux/atomic.h> |
36aa9e5f5
|
29 |
#include <linux/ctype.h> |
eea8f41cc
|
30 |
#include <linux/blk-cgroup.h> |
d09d8df3a
|
31 |
#include <linux/tracehook.h> |
fd112c746
|
32 |
#include <linux/psi.h> |
5efd61135
|
33 |
#include "blk.h" |
3e2520668
|
34 |
|
84c124da9
|
35 |
#define MAX_KEY_LEN 100 |
838f13bf4
|
36 37 38 39 40 41 42 43 |
/* * 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
|
44 |
static DEFINE_MUTEX(blkcg_pol_mutex); |
923adde1b
|
45 |
|
e48453c38
|
46 |
struct blkcg blkcg_root; |
3c798398e
|
47 |
EXPORT_SYMBOL_GPL(blkcg_root); |
9d6a986c0
|
48 |
|
496d5e756
|
49 |
struct cgroup_subsys_state * const blkcg_root_css = &blkcg_root.css; |
9b0eb69b7
|
50 |
EXPORT_SYMBOL_GPL(blkcg_root_css); |
496d5e756
|
51 |
|
3c798398e
|
52 |
static struct blkcg_policy *blkcg_policy[BLKCG_MAX_POLS]; |
035d10b2f
|
53 |
|
7876f930d
|
54 |
static LIST_HEAD(all_blkcgs); /* protected by blkcg_pol_mutex */ |
07b0fdecb
|
55 |
bool blkcg_debug_stats = false; |
d3f77dfdc
|
56 |
static struct workqueue_struct *blkcg_punt_bio_wq; |
903d23f0a
|
57 |
|
a2b1693ba
|
58 |
static bool blkcg_policy_enabled(struct request_queue *q, |
3c798398e
|
59 |
const struct blkcg_policy *pol) |
a2b1693ba
|
60 61 62 |
{ return pol && test_bit(pol->plid, q->blkcg_pols); } |
0381411e4
|
63 64 65 66 67 68 |
/** * blkg_free - free a blkg * @blkg: blkg to free * * Free @blkg which may be partially allocated. */ |
3c798398e
|
69 |
static void blkg_free(struct blkcg_gq *blkg) |
0381411e4
|
70 |
{ |
e8989fae3
|
71 |
int i; |
549d3aa87
|
72 73 74 |
if (!blkg) return; |
db6136703
|
75 |
for (i = 0; i < BLKCG_MAX_POLS; i++) |
001bea73e
|
76 77 |
if (blkg->pd[i]) blkcg_policy[i]->pd_free_fn(blkg->pd[i]); |
e8989fae3
|
78 |
|
f73316482
|
79 |
free_percpu(blkg->iostat_cpu); |
ef069b97f
|
80 |
percpu_ref_exit(&blkg->refcnt); |
549d3aa87
|
81 |
kfree(blkg); |
0381411e4
|
82 |
} |
7fcf2b033
|
83 84 85 |
static void __blkg_release(struct rcu_head *rcu) { struct blkcg_gq *blkg = container_of(rcu, struct blkcg_gq, rcu_head); |
d3f77dfdc
|
86 |
WARN_ON(!bio_list_empty(&blkg->async_bios)); |
7fcf2b033
|
87 88 89 90 |
/* release the blkcg and parent blkg refs this blkg has been holding */ css_put(&blkg->blkcg->css); if (blkg->parent) blkg_put(blkg->parent); |
7fcf2b033
|
91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 |
blkg_free(blkg); } /* * 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. */ static void blkg_release(struct percpu_ref *ref) { struct blkcg_gq *blkg = container_of(ref, struct blkcg_gq, refcnt); call_rcu(&blkg->rcu_head, __blkg_release); } |
d3f77dfdc
|
108 109 110 111 112 113 |
static void blkg_async_bio_workfn(struct work_struct *work) { struct blkcg_gq *blkg = container_of(work, struct blkcg_gq, async_bio_work); struct bio_list bios = BIO_EMPTY_LIST; struct bio *bio; |
192f1c6bc
|
114 115 |
struct blk_plug plug; bool need_plug = false; |
d3f77dfdc
|
116 117 118 119 120 121 |
/* as long as there are pending bios, @blkg can't go away */ spin_lock_bh(&blkg->async_bio_lock); bio_list_merge(&bios, &blkg->async_bios); bio_list_init(&blkg->async_bios); spin_unlock_bh(&blkg->async_bio_lock); |
192f1c6bc
|
122 123 124 125 126 |
/* start plug only when bio_list contains at least 2 bios */ if (bios.head && bios.head->bi_next) { need_plug = true; blk_start_plug(&plug); } |
d3f77dfdc
|
127 128 |
while ((bio = bio_list_pop(&bios))) submit_bio(bio); |
192f1c6bc
|
129 130 |
if (need_plug) blk_finish_plug(&plug); |
d3f77dfdc
|
131 |
} |
0381411e4
|
132 133 134 135 |
/** * blkg_alloc - allocate a blkg * @blkcg: block cgroup the new blkg is associated with * @q: request_queue the new blkg is associated with |
159749937
|
136 |
* @gfp_mask: allocation mask to use |
0381411e4
|
137 |
* |
e8989fae3
|
138 |
* Allocate a new blkg assocating @blkcg and @q. |
0381411e4
|
139 |
*/ |
159749937
|
140 141 |
static struct blkcg_gq *blkg_alloc(struct blkcg *blkcg, struct request_queue *q, gfp_t gfp_mask) |
0381411e4
|
142 |
{ |
3c798398e
|
143 |
struct blkcg_gq *blkg; |
f73316482
|
144 |
int i, cpu; |
0381411e4
|
145 146 |
/* alloc and init base part */ |
159749937
|
147 |
blkg = kzalloc_node(sizeof(*blkg), gfp_mask, q->node); |
0381411e4
|
148 149 |
if (!blkg) return NULL; |
ef069b97f
|
150 151 |
if (percpu_ref_init(&blkg->refcnt, blkg_release, 0, gfp_mask)) goto err_free; |
f73316482
|
152 153 |
blkg->iostat_cpu = alloc_percpu_gfp(struct blkg_iostat_set, gfp_mask); if (!blkg->iostat_cpu) |
77ea73388
|
154 |
goto err_free; |
c875f4d02
|
155 |
blkg->q = q; |
e8989fae3
|
156 |
INIT_LIST_HEAD(&blkg->q_node); |
d3f77dfdc
|
157 158 159 |
spin_lock_init(&blkg->async_bio_lock); bio_list_init(&blkg->async_bios); INIT_WORK(&blkg->async_bio_work, blkg_async_bio_workfn); |
0381411e4
|
160 |
blkg->blkcg = blkcg; |
0381411e4
|
161 |
|
f73316482
|
162 163 164 |
u64_stats_init(&blkg->iostat.sync); for_each_possible_cpu(cpu) u64_stats_init(&per_cpu_ptr(blkg->iostat_cpu, cpu)->sync); |
8bd435b30
|
165 |
for (i = 0; i < BLKCG_MAX_POLS; i++) { |
3c798398e
|
166 |
struct blkcg_policy *pol = blkcg_policy[i]; |
e8989fae3
|
167 |
struct blkg_policy_data *pd; |
0381411e4
|
168 |
|
a2b1693ba
|
169 |
if (!blkcg_policy_enabled(q, pol)) |
e8989fae3
|
170 171 172 |
continue; /* alloc per-policy data and attach it to blkg */ |
cf09a8ee1
|
173 |
pd = pol->pd_alloc_fn(gfp_mask, q, blkcg); |
a051661ca
|
174 175 |
if (!pd) goto err_free; |
549d3aa87
|
176 |
|
e8989fae3
|
177 178 |
blkg->pd[i] = pd; pd->blkg = blkg; |
b276a876a
|
179 |
pd->plid = i; |
e8989fae3
|
180 |
} |
0381411e4
|
181 |
return blkg; |
a051661ca
|
182 183 184 185 |
err_free: blkg_free(blkg); return NULL; |
0381411e4
|
186 |
} |
24f290466
|
187 188 |
struct blkcg_gq *blkg_lookup_slowpath(struct blkcg *blkcg, struct request_queue *q, bool update_hint) |
80fd99792
|
189 |
{ |
3c798398e
|
190 |
struct blkcg_gq *blkg; |
80fd99792
|
191 |
|
a637120e4
|
192 |
/* |
86cde6b62
|
193 194 195 196 |
* 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
|
197 198 |
*/ blkg = radix_tree_lookup(&blkcg->blkg_tree, q->id); |
86cde6b62
|
199 200 |
if (blkg && blkg->q == q) { if (update_hint) { |
0d945c1f9
|
201 |
lockdep_assert_held(&q->queue_lock); |
86cde6b62
|
202 203 |
rcu_assign_pointer(blkcg->blkg_hint, blkg); } |
a637120e4
|
204 |
return blkg; |
86cde6b62
|
205 |
} |
a637120e4
|
206 |
|
80fd99792
|
207 208 |
return NULL; } |
ae1188963
|
209 |
EXPORT_SYMBOL_GPL(blkg_lookup_slowpath); |
80fd99792
|
210 |
|
159749937
|
211 |
/* |
d708f0d50
|
212 213 |
* If @new_blkg is %NULL, this function tries to allocate a new one as * necessary using %GFP_NOWAIT. @new_blkg is always consumed on return. |
159749937
|
214 |
*/ |
86cde6b62
|
215 |
static struct blkcg_gq *blkg_create(struct blkcg *blkcg, |
d708f0d50
|
216 217 |
struct request_queue *q, struct blkcg_gq *new_blkg) |
5624a4e44
|
218 |
{ |
d708f0d50
|
219 |
struct blkcg_gq *blkg; |
f427d9096
|
220 |
int i, ret; |
5624a4e44
|
221 |
|
cd1604fab
|
222 |
WARN_ON_ONCE(!rcu_read_lock_held()); |
0d945c1f9
|
223 |
lockdep_assert_held(&q->queue_lock); |
cd1604fab
|
224 |
|
0273ac349
|
225 226 227 228 229 |
/* request_queue is dying, do not create/recreate a blkg */ if (blk_queue_dying(q)) { ret = -ENODEV; goto err_free_blkg; } |
7ee9c5620
|
230 |
/* blkg holds a reference to blkcg */ |
ec903c0c8
|
231 |
if (!css_tryget_online(&blkcg->css)) { |
20386ce01
|
232 |
ret = -ENODEV; |
93e6d5d8f
|
233 |
goto err_free_blkg; |
159749937
|
234 |
} |
cd1604fab
|
235 |
|
d708f0d50
|
236 237 238 239 240 |
/* allocate */ if (!new_blkg) { new_blkg = blkg_alloc(blkcg, q, GFP_NOWAIT | __GFP_NOWARN); if (unlikely(!new_blkg)) { ret = -ENOMEM; |
8c911f3d4
|
241 |
goto err_put_css; |
159749937
|
242 243 |
} } |
d708f0d50
|
244 |
blkg = new_blkg; |
cd1604fab
|
245 |
|
db6136703
|
246 |
/* link parent */ |
3c5478659
|
247 248 249 |
if (blkcg_parent(blkcg)) { blkg->parent = __blkg_lookup(blkcg_parent(blkcg), q, false); if (WARN_ON_ONCE(!blkg->parent)) { |
20386ce01
|
250 |
ret = -ENODEV; |
8c911f3d4
|
251 |
goto err_put_css; |
3c5478659
|
252 253 254 |
} blkg_get(blkg->parent); } |
db6136703
|
255 256 257 258 259 |
/* 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
|
260 |
pol->pd_init_fn(blkg->pd[i]); |
db6136703
|
261 262 263 |
} /* insert */ |
cd1604fab
|
264 |
spin_lock(&blkcg->lock); |
a637120e4
|
265 266 267 268 |
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
|
269 270 271 272 273 |
for (i = 0; i < BLKCG_MAX_POLS; i++) { struct blkcg_policy *pol = blkcg_policy[i]; if (blkg->pd[i] && pol->pd_online_fn) |
a9520cd6f
|
274 |
pol->pd_online_fn(blkg->pd[i]); |
f427d9096
|
275 |
} |
a637120e4
|
276 |
} |
f427d9096
|
277 |
blkg->online = true; |
cd1604fab
|
278 |
spin_unlock(&blkcg->lock); |
496fb7806
|
279 |
|
ec13b1d6f
|
280 |
if (!ret) |
a637120e4
|
281 |
return blkg; |
159749937
|
282 |
|
3c5478659
|
283 284 285 |
/* @blkg failed fully initialized, use the usual release path */ blkg_put(blkg); return ERR_PTR(ret); |
d708f0d50
|
286 |
err_put_css: |
496fb7806
|
287 |
css_put(&blkcg->css); |
93e6d5d8f
|
288 |
err_free_blkg: |
d708f0d50
|
289 |
blkg_free(new_blkg); |
93e6d5d8f
|
290 |
return ERR_PTR(ret); |
31e4c28d9
|
291 |
} |
3c96cb32d
|
292 |
|
86cde6b62
|
293 |
/** |
8c5462875
|
294 |
* blkg_lookup_create - lookup blkg, try to create one if not there |
86cde6b62
|
295 296 297 298 |
* @blkcg: blkcg of interest * @q: request_queue of interest * * Lookup blkg for the @blkcg - @q pair. If it doesn't exist, try to |
3c5478659
|
299 300 |
* 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 |
8c5462875
|
301 |
* should be called under RCU read lock and takes @q->queue_lock. |
86cde6b62
|
302 |
* |
beea9da07
|
303 304 |
* Returns the blkg or the closest blkg if blkg_create() fails as it walks * down from root. |
86cde6b62
|
305 |
*/ |
8c5462875
|
306 307 |
static struct blkcg_gq *blkg_lookup_create(struct blkcg *blkcg, struct request_queue *q) |
3c96cb32d
|
308 |
{ |
86cde6b62
|
309 |
struct blkcg_gq *blkg; |
8c5462875
|
310 |
unsigned long flags; |
86cde6b62
|
311 312 |
WARN_ON_ONCE(!rcu_read_lock_held()); |
86cde6b62
|
313 |
|
8c5462875
|
314 |
blkg = blkg_lookup(blkcg, q); |
86cde6b62
|
315 316 |
if (blkg) return blkg; |
8c5462875
|
317 318 319 320 |
spin_lock_irqsave(&q->queue_lock, flags); blkg = __blkg_lookup(blkcg, q, true); if (blkg) goto found; |
3c5478659
|
321 322 |
/* * Create blkgs walking down from blkcg_root to @blkcg, so that all |
beea9da07
|
323 324 |
* non-root blkgs have access to their parents. Returns the closest * blkg to the intended blkg should blkg_create() fail. |
3c5478659
|
325 326 327 328 |
*/ while (true) { struct blkcg *pos = blkcg; struct blkcg *parent = blkcg_parent(blkcg); |
beea9da07
|
329 330 331 332 333 334 335 336 337 |
struct blkcg_gq *ret_blkg = q->root_blkg; while (parent) { blkg = __blkg_lookup(parent, q, false); if (blkg) { /* remember closest blkg */ ret_blkg = blkg; break; } |
3c5478659
|
338 339 340 |
pos = parent; parent = blkcg_parent(parent); } |
d708f0d50
|
341 |
blkg = blkg_create(pos, q, NULL); |
8c5462875
|
342 343 344 345 |
if (IS_ERR(blkg)) { blkg = ret_blkg; break; } |
beea9da07
|
346 |
if (pos == blkcg) |
8c5462875
|
347 |
break; |
b978962ad
|
348 |
} |
8c5462875
|
349 350 |
found: spin_unlock_irqrestore(&q->queue_lock, flags); |
b978962ad
|
351 352 |
return blkg; } |
3c798398e
|
353 |
static void blkg_destroy(struct blkcg_gq *blkg) |
03aa264ac
|
354 |
{ |
3c798398e
|
355 |
struct blkcg *blkcg = blkg->blkcg; |
6b0654620
|
356 |
int i; |
03aa264ac
|
357 |
|
0d945c1f9
|
358 |
lockdep_assert_held(&blkg->q->queue_lock); |
9f13ef678
|
359 |
lockdep_assert_held(&blkcg->lock); |
03aa264ac
|
360 361 |
/* Something wrong if we are trying to remove same group twice */ |
e8989fae3
|
362 |
WARN_ON_ONCE(list_empty(&blkg->q_node)); |
9f13ef678
|
363 |
WARN_ON_ONCE(hlist_unhashed(&blkg->blkcg_node)); |
a637120e4
|
364 |
|
6b0654620
|
365 366 367 368 369 370 |
for (i = 0; i < BLKCG_MAX_POLS; i++) { struct blkcg_policy *pol = blkcg_policy[i]; if (blkg->pd[i] && pol->pd_offline_fn) pol->pd_offline_fn(blkg->pd[i]); } |
f427d9096
|
371 |
blkg->online = false; |
a637120e4
|
372 |
radix_tree_delete(&blkcg->blkg_tree, blkg->q->id); |
e8989fae3
|
373 |
list_del_init(&blkg->q_node); |
9f13ef678
|
374 |
hlist_del_init_rcu(&blkg->blkcg_node); |
03aa264ac
|
375 |
|
03aa264ac
|
376 |
/* |
a637120e4
|
377 378 379 380 |
* 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
|
381 |
if (rcu_access_pointer(blkcg->blkg_hint) == blkg) |
a637120e4
|
382 383 384 |
rcu_assign_pointer(blkcg->blkg_hint, NULL); /* |
03aa264ac
|
385 386 387 |
* Put the reference taken at the time of creation so that when all * queues are gone, group can be destroyed. */ |
7fcf2b033
|
388 |
percpu_ref_kill(&blkg->refcnt); |
03aa264ac
|
389 |
} |
9f13ef678
|
390 391 392 |
/** * blkg_destroy_all - destroy all blkgs associated with a request_queue * @q: request_queue of interest |
9f13ef678
|
393 |
* |
3c96cb32d
|
394 |
* Destroy all blkgs associated with @q. |
9f13ef678
|
395 |
*/ |
3c96cb32d
|
396 |
static void blkg_destroy_all(struct request_queue *q) |
72e06c255
|
397 |
{ |
3c798398e
|
398 |
struct blkcg_gq *blkg, *n; |
72e06c255
|
399 |
|
0d945c1f9
|
400 |
spin_lock_irq(&q->queue_lock); |
9f13ef678
|
401 |
list_for_each_entry_safe(blkg, n, &q->blkg_list, q_node) { |
3c798398e
|
402 |
struct blkcg *blkcg = blkg->blkcg; |
72e06c255
|
403 |
|
9f13ef678
|
404 405 406 |
spin_lock(&blkcg->lock); blkg_destroy(blkg); spin_unlock(&blkcg->lock); |
72e06c255
|
407 |
} |
6fe810bda
|
408 409 |
q->root_blkg = NULL; |
0d945c1f9
|
410 |
spin_unlock_irq(&q->queue_lock); |
72e06c255
|
411 |
} |
182446d08
|
412 413 |
static int blkcg_reset_stats(struct cgroup_subsys_state *css, struct cftype *cftype, u64 val) |
303a3acb2
|
414 |
{ |
182446d08
|
415 |
struct blkcg *blkcg = css_to_blkcg(css); |
3c798398e
|
416 |
struct blkcg_gq *blkg; |
f73316482
|
417 |
int i, cpu; |
303a3acb2
|
418 |
|
838f13bf4
|
419 |
mutex_lock(&blkcg_pol_mutex); |
303a3acb2
|
420 |
spin_lock_irq(&blkcg->lock); |
997a026c8
|
421 422 423 424 425 426 |
/* * 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
|
427 |
hlist_for_each_entry(blkg, &blkcg->blkg_list, blkcg_node) { |
f73316482
|
428 429 430 431 432 433 |
for_each_possible_cpu(cpu) { struct blkg_iostat_set *bis = per_cpu_ptr(blkg->iostat_cpu, cpu); memset(bis, 0, sizeof(*bis)); } memset(&blkg->iostat, 0, sizeof(blkg->iostat)); |
77ea73388
|
434 |
|
8bd435b30
|
435 |
for (i = 0; i < BLKCG_MAX_POLS; i++) { |
3c798398e
|
436 |
struct blkcg_policy *pol = blkcg_policy[i]; |
549d3aa87
|
437 |
|
a9520cd6f
|
438 439 |
if (blkg->pd[i] && pol->pd_reset_stats_fn) pol->pd_reset_stats_fn(blkg->pd[i]); |
bc0d6501a
|
440 |
} |
303a3acb2
|
441 |
} |
f0bdc8cdd
|
442 |
|
303a3acb2
|
443 |
spin_unlock_irq(&blkcg->lock); |
bc0d6501a
|
444 |
mutex_unlock(&blkcg_pol_mutex); |
303a3acb2
|
445 446 |
return 0; } |
dd165eb3b
|
447 |
const char *blkg_dev_name(struct blkcg_gq *blkg) |
303a3acb2
|
448 |
{ |
d3d32e69f
|
449 |
/* some drivers (floppy) instantiate a queue w/o disk registered */ |
dc3b17cc8
|
450 |
if (blkg->q->backing_dev_info->dev) |
d51cfc53a
|
451 |
return bdi_dev_name(blkg->q->backing_dev_info); |
d3d32e69f
|
452 |
return NULL; |
303a3acb2
|
453 |
} |
d3d32e69f
|
454 455 456 457 458 459 460 461 462 463 464 |
/** * 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
|
465 466 467 |
* 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
|
468 469 470 471 |
* * This is to be used to construct print functions for * cftype->read_seq_string method. */ |
3c798398e
|
472 |
void blkcg_print_blkgs(struct seq_file *sf, struct blkcg *blkcg, |
f95a04afa
|
473 474 |
u64 (*prfill)(struct seq_file *, struct blkg_policy_data *, int), |
3c798398e
|
475 |
const struct blkcg_policy *pol, int data, |
ec399347d
|
476 |
bool show_total) |
5624a4e44
|
477 |
{ |
3c798398e
|
478 |
struct blkcg_gq *blkg; |
d3d32e69f
|
479 |
u64 total = 0; |
5624a4e44
|
480 |
|
810ecfa76
|
481 |
rcu_read_lock(); |
ee89f8125
|
482 |
hlist_for_each_entry_rcu(blkg, &blkcg->blkg_list, blkcg_node) { |
0d945c1f9
|
483 |
spin_lock_irq(&blkg->q->queue_lock); |
a2b1693ba
|
484 |
if (blkcg_policy_enabled(blkg->q, pol)) |
f95a04afa
|
485 |
total += prfill(sf, blkg->pd[pol->plid], data); |
0d945c1f9
|
486 |
spin_unlock_irq(&blkg->q->queue_lock); |
810ecfa76
|
487 488 |
} rcu_read_unlock(); |
d3d32e69f
|
489 490 491 492 493 |
if (show_total) seq_printf(sf, "Total %llu ", (unsigned long long)total); } |
829fdb500
|
494 |
EXPORT_SYMBOL_GPL(blkcg_print_blkgs); |
d3d32e69f
|
495 496 497 498 |
/** * __blkg_prfill_u64 - prfill helper for a single u64 value * @sf: seq_file to print to |
f95a04afa
|
499 |
* @pd: policy private data of interest |
d3d32e69f
|
500 501 |
* @v: value to print * |
f95a04afa
|
502 |
* Print @v to @sf for the device assocaited with @pd. |
d3d32e69f
|
503 |
*/ |
f95a04afa
|
504 |
u64 __blkg_prfill_u64(struct seq_file *sf, struct blkg_policy_data *pd, u64 v) |
d3d32e69f
|
505 |
{ |
f95a04afa
|
506 |
const char *dname = blkg_dev_name(pd->blkg); |
d3d32e69f
|
507 508 509 510 511 512 513 514 |
if (!dname) return 0; seq_printf(sf, "%s %llu ", dname, (unsigned long long)v); return v; } |
829fdb500
|
515 |
EXPORT_SYMBOL_GPL(__blkg_prfill_u64); |
d3d32e69f
|
516 |
|
457e490f2
|
517 518 519 520 521 522 |
/* Performs queue bypass and policy enabled checks then looks up blkg. */ static struct blkcg_gq *blkg_lookup_check(struct blkcg *blkcg, const struct blkcg_policy *pol, struct request_queue *q) { WARN_ON_ONCE(!rcu_read_lock_held()); |
0d945c1f9
|
523 |
lockdep_assert_held(&q->queue_lock); |
457e490f2
|
524 525 526 |
if (!blkcg_policy_enabled(q, pol)) return ERR_PTR(-EOPNOTSUPP); |
457e490f2
|
527 528 |
return __blkg_lookup(blkcg, q, true /* update_hint */); } |
16b3de665
|
529 |
/** |
3a8b31d39
|
530 |
* blkg_conf_prep - parse and prepare for per-blkg config update |
015d254cb
|
531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 |
* @inputp: input string pointer * * Parse the device node prefix part, MAJ:MIN, of per-blkg config update * from @input and get and return the matching gendisk. *@inputp is * updated to point past the device node prefix. Returns an ERR_PTR() * value on error. * * Use this function iff blkg_conf_prep() can't be used for some reason. */ struct gendisk *blkcg_conf_get_disk(char **inputp) { char *input = *inputp; unsigned int major, minor; struct gendisk *disk; int key_len, part; if (sscanf(input, "%u:%u%n", &major, &minor, &key_len) != 2) return ERR_PTR(-EINVAL); input += key_len; if (!isspace(*input)) return ERR_PTR(-EINVAL); input = skip_spaces(input); disk = get_gendisk(MKDEV(major, minor), &part); if (!disk) return ERR_PTR(-ENODEV); if (part) { put_disk_and_module(disk); return ERR_PTR(-ENODEV); } *inputp = input; return disk; } /** * blkg_conf_prep - parse and prepare for per-blkg config update |
3a8b31d39
|
569 |
* @blkcg: target block cgroup |
da8b06626
|
570 |
* @pol: target policy |
3a8b31d39
|
571 572 573 574 |
* @input: input string * @ctx: blkg_conf_ctx to be filled * * Parse per-blkg config update from @input and initialize @ctx with the |
36aa9e5f5
|
575 576 577 |
* 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
|
578 |
*/ |
3c798398e
|
579 |
int blkg_conf_prep(struct blkcg *blkcg, const struct blkcg_policy *pol, |
36aa9e5f5
|
580 |
char *input, struct blkg_conf_ctx *ctx) |
0d945c1f9
|
581 |
__acquires(rcu) __acquires(&disk->queue->queue_lock) |
34d0f179d
|
582 |
{ |
3a8b31d39
|
583 |
struct gendisk *disk; |
457e490f2
|
584 |
struct request_queue *q; |
3c798398e
|
585 |
struct blkcg_gq *blkg; |
015d254cb
|
586 |
int ret; |
36aa9e5f5
|
587 |
|
015d254cb
|
588 589 590 |
disk = blkcg_conf_get_disk(&input); if (IS_ERR(disk)) return PTR_ERR(disk); |
e56da7e28
|
591 |
|
457e490f2
|
592 |
q = disk->queue; |
da8b06626
|
593 |
|
457e490f2
|
594 |
rcu_read_lock(); |
0d945c1f9
|
595 |
spin_lock_irq(&q->queue_lock); |
e56da7e28
|
596 |
|
457e490f2
|
597 |
blkg = blkg_lookup_check(blkcg, pol, q); |
4bfd482e7
|
598 599 |
if (IS_ERR(blkg)) { ret = PTR_ERR(blkg); |
457e490f2
|
600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 |
goto fail_unlock; } if (blkg) goto success; /* * 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; struct blkcg_gq *new_blkg; parent = blkcg_parent(blkcg); while (parent && !__blkg_lookup(parent, q, false)) { pos = parent; parent = blkcg_parent(parent); } /* Drop locks to do new blkg allocation with GFP_KERNEL. */ |
0d945c1f9
|
622 |
spin_unlock_irq(&q->queue_lock); |
3a8b31d39
|
623 |
rcu_read_unlock(); |
457e490f2
|
624 625 626 627 628 |
new_blkg = blkg_alloc(pos, q, GFP_KERNEL); if (unlikely(!new_blkg)) { ret = -ENOMEM; goto fail; |
7702e8f45
|
629 |
} |
3a8b31d39
|
630 |
|
f255c19b3
|
631 632 633 634 635 |
if (radix_tree_preload(GFP_KERNEL)) { blkg_free(new_blkg); ret = -ENOMEM; goto fail; } |
457e490f2
|
636 |
rcu_read_lock(); |
0d945c1f9
|
637 |
spin_lock_irq(&q->queue_lock); |
457e490f2
|
638 639 640 641 |
blkg = blkg_lookup_check(pos, pol, q); if (IS_ERR(blkg)) { ret = PTR_ERR(blkg); |
52abfcbd5
|
642 |
blkg_free(new_blkg); |
f255c19b3
|
643 |
goto fail_preloaded; |
457e490f2
|
644 645 646 647 648 649 |
} if (blkg) { blkg_free(new_blkg); } else { blkg = blkg_create(pos, q, new_blkg); |
98d669b49
|
650 |
if (IS_ERR(blkg)) { |
457e490f2
|
651 |
ret = PTR_ERR(blkg); |
f255c19b3
|
652 |
goto fail_preloaded; |
457e490f2
|
653 654 |
} } |
f255c19b3
|
655 |
radix_tree_preload_end(); |
457e490f2
|
656 657 658 659 |
if (pos == blkcg) goto success; } success: |
3a8b31d39
|
660 661 |
ctx->disk = disk; ctx->blkg = blkg; |
015d254cb
|
662 |
ctx->body = input; |
726fa6945
|
663 |
return 0; |
457e490f2
|
664 |
|
f255c19b3
|
665 666 |
fail_preloaded: radix_tree_preload_end(); |
457e490f2
|
667 |
fail_unlock: |
0d945c1f9
|
668 |
spin_unlock_irq(&q->queue_lock); |
457e490f2
|
669 670 |
rcu_read_unlock(); fail: |
9df6c2991
|
671 |
put_disk_and_module(disk); |
457e490f2
|
672 673 674 675 676 677 678 679 680 681 682 |
/* * 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(); } return ret; |
34d0f179d
|
683 |
} |
89f3b6d62
|
684 |
EXPORT_SYMBOL_GPL(blkg_conf_prep); |
34d0f179d
|
685 |
|
3a8b31d39
|
686 687 688 689 690 691 692 |
/** * 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
|
693 |
void blkg_conf_finish(struct blkg_conf_ctx *ctx) |
0d945c1f9
|
694 |
__releases(&ctx->disk->queue->queue_lock) __releases(rcu) |
34d0f179d
|
695 |
{ |
0d945c1f9
|
696 |
spin_unlock_irq(&ctx->disk->queue->queue_lock); |
3a8b31d39
|
697 |
rcu_read_unlock(); |
9df6c2991
|
698 |
put_disk_and_module(ctx->disk); |
34d0f179d
|
699 |
} |
89f3b6d62
|
700 |
EXPORT_SYMBOL_GPL(blkg_conf_finish); |
34d0f179d
|
701 |
|
cd1fc4b98
|
702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 |
static void blkg_iostat_set(struct blkg_iostat *dst, struct blkg_iostat *src) { int i; for (i = 0; i < BLKG_IOSTAT_NR; i++) { dst->bytes[i] = src->bytes[i]; dst->ios[i] = src->ios[i]; } } static void blkg_iostat_add(struct blkg_iostat *dst, struct blkg_iostat *src) { int i; for (i = 0; i < BLKG_IOSTAT_NR; i++) { dst->bytes[i] += src->bytes[i]; dst->ios[i] += src->ios[i]; } } static void blkg_iostat_sub(struct blkg_iostat *dst, struct blkg_iostat *src) { int i; for (i = 0; i < BLKG_IOSTAT_NR; i++) { dst->bytes[i] -= src->bytes[i]; dst->ios[i] -= src->ios[i]; } } static void blkcg_rstat_flush(struct cgroup_subsys_state *css, int cpu) { struct blkcg *blkcg = css_to_blkcg(css); struct blkcg_gq *blkg; rcu_read_lock(); hlist_for_each_entry_rcu(blkg, &blkcg->blkg_list, blkcg_node) { struct blkcg_gq *parent = blkg->parent; struct blkg_iostat_set *bisc = per_cpu_ptr(blkg->iostat_cpu, cpu); struct blkg_iostat cur, delta; unsigned int seq; /* fetch the current per-cpu values */ do { seq = u64_stats_fetch_begin(&bisc->sync); blkg_iostat_set(&cur, &bisc->cur); } while (u64_stats_fetch_retry(&bisc->sync, seq)); /* propagate percpu delta to global */ u64_stats_update_begin(&blkg->iostat.sync); blkg_iostat_set(&delta, &cur); blkg_iostat_sub(&delta, &bisc->last); blkg_iostat_add(&blkg->iostat.cur, &delta); blkg_iostat_add(&bisc->last, &delta); u64_stats_update_end(&blkg->iostat.sync); /* propagate global delta to parent */ if (parent) { u64_stats_update_begin(&parent->iostat.sync); blkg_iostat_set(&delta, &blkg->iostat.cur); blkg_iostat_sub(&delta, &blkg->iostat.last); blkg_iostat_add(&parent->iostat.cur, &delta); blkg_iostat_add(&blkg->iostat.last, &delta); u64_stats_update_end(&parent->iostat.sync); } } rcu_read_unlock(); } |
ef45fe470
|
772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 |
/* * The rstat algorithms intentionally don't handle the root cgroup to avoid * incurring overhead when no cgroups are defined. For that reason, * cgroup_rstat_flush in blkcg_print_stat does not actually fill out the * iostat in the root cgroup's blkcg_gq. * * However, we would like to re-use the printing code between the root and * non-root cgroups to the extent possible. For that reason, we simulate * flushing the root cgroup's stats by explicitly filling in the iostat * with disk level statistics. */ static void blkcg_fill_root_iostats(void) { struct class_dev_iter iter; struct device *dev; class_dev_iter_init(&iter, &block_class, NULL, &disk_type); while ((dev = class_dev_iter_next(&iter))) { struct gendisk *disk = dev_to_disk(dev); struct hd_struct *part = disk_get_part(disk, 0); struct blkcg_gq *blkg = blk_queue_root_blkg(disk->queue); struct blkg_iostat tmp; int cpu; memset(&tmp, 0, sizeof(tmp)); for_each_possible_cpu(cpu) { struct disk_stats *cpu_dkstats; cpu_dkstats = per_cpu_ptr(part->dkstats, cpu); tmp.ios[BLKG_IOSTAT_READ] += cpu_dkstats->ios[STAT_READ]; tmp.ios[BLKG_IOSTAT_WRITE] += cpu_dkstats->ios[STAT_WRITE]; tmp.ios[BLKG_IOSTAT_DISCARD] += cpu_dkstats->ios[STAT_DISCARD]; // convert sectors to bytes tmp.bytes[BLKG_IOSTAT_READ] += cpu_dkstats->sectors[STAT_READ] << 9; tmp.bytes[BLKG_IOSTAT_WRITE] += cpu_dkstats->sectors[STAT_WRITE] << 9; tmp.bytes[BLKG_IOSTAT_DISCARD] += cpu_dkstats->sectors[STAT_DISCARD] << 9; u64_stats_update_begin(&blkg->iostat.sync); blkg_iostat_set(&blkg->iostat.cur, &tmp); u64_stats_update_end(&blkg->iostat.sync); } |
b7131ee0b
|
819 |
disk_put_part(part); |
ef45fe470
|
820 821 |
} } |
2ee867dcf
|
822 823 824 825 |
static int blkcg_print_stat(struct seq_file *sf, void *v) { struct blkcg *blkcg = css_to_blkcg(seq_css(sf)); struct blkcg_gq *blkg; |
ef45fe470
|
826 827 828 829 |
if (!seq_css(sf)->parent) blkcg_fill_root_iostats(); else cgroup_rstat_flush(blkcg->css.cgroup); |
2ee867dcf
|
830 831 832 |
rcu_read_lock(); hlist_for_each_entry_rcu(blkg, &blkcg->blkg_list, blkcg_node) { |
f73316482
|
833 |
struct blkg_iostat_set *bis = &blkg->iostat; |
2ee867dcf
|
834 |
const char *dname; |
903d23f0a
|
835 |
char *buf; |
636620b66
|
836 |
u64 rbytes, wbytes, rios, wios, dbytes, dios; |
903d23f0a
|
837 838 839 |
size_t size = seq_get_buf(sf, &buf), off = 0; int i; bool has_stats = false; |
f73316482
|
840 |
unsigned seq; |
2ee867dcf
|
841 |
|
b0814361a
|
842 843 844 845 |
spin_lock_irq(&blkg->q->queue_lock); if (!blkg->online) goto skip; |
2ee867dcf
|
846 847 |
dname = blkg_dev_name(blkg); if (!dname) |
b0814361a
|
848 |
goto skip; |
2ee867dcf
|
849 |
|
903d23f0a
|
850 851 852 853 854 855 856 |
/* * Hooray string manipulation, count is the size written NOT * INCLUDING THE \0, so size is now count+1 less than what we * had before, but we want to start writing the next bit from * the \0 so we only add count to buf. */ off += scnprintf(buf+off, size-off, "%s ", dname); |
f73316482
|
857 858 |
do { seq = u64_stats_fetch_begin(&bis->sync); |
2ee867dcf
|
859 |
|
f73316482
|
860 861 862 863 864 865 866 |
rbytes = bis->cur.bytes[BLKG_IOSTAT_READ]; wbytes = bis->cur.bytes[BLKG_IOSTAT_WRITE]; dbytes = bis->cur.bytes[BLKG_IOSTAT_DISCARD]; rios = bis->cur.ios[BLKG_IOSTAT_READ]; wios = bis->cur.ios[BLKG_IOSTAT_WRITE]; dios = bis->cur.ios[BLKG_IOSTAT_DISCARD]; } while (u64_stats_fetch_retry(&bis->sync, seq)); |
2ee867dcf
|
867 |
|
903d23f0a
|
868 869 870 |
if (rbytes || wbytes || rios || wios) { has_stats = true; off += scnprintf(buf+off, size-off, |
636620b66
|
871 872 873 |
"rbytes=%llu wbytes=%llu rios=%llu wios=%llu dbytes=%llu dios=%llu", rbytes, wbytes, rios, wios, dbytes, dios); |
903d23f0a
|
874 |
} |
07b0fdecb
|
875 |
if (blkcg_debug_stats && atomic_read(&blkg->use_delay)) { |
d09d8df3a
|
876 877 878 879 880 881 |
has_stats = true; off += scnprintf(buf+off, size-off, " use_delay=%d delay_nsec=%llu", atomic_read(&blkg->use_delay), (unsigned long long)atomic64_read(&blkg->delay_nsec)); } |
903d23f0a
|
882 883 884 885 886 887 888 889 890 891 892 893 |
for (i = 0; i < BLKCG_MAX_POLS; i++) { struct blkcg_policy *pol = blkcg_policy[i]; size_t written; if (!blkg->pd[i] || !pol->pd_stat_fn) continue; written = pol->pd_stat_fn(blkg->pd[i], buf+off, size-off); if (written) has_stats = true; off += written; } |
07b0fdecb
|
894 |
|
903d23f0a
|
895 |
if (has_stats) { |
f539da82f
|
896 897 898 899 900 901 902 |
if (off < size - 1) { off += scnprintf(buf+off, size-off, " "); seq_commit(sf, off); } else { seq_commit(sf, -1); } |
903d23f0a
|
903 |
} |
b0814361a
|
904 905 |
skip: spin_unlock_irq(&blkg->q->queue_lock); |
2ee867dcf
|
906 907 908 909 910 |
} rcu_read_unlock(); return 0; } |
e1f3b9412
|
911 |
static struct cftype blkcg_files[] = { |
2ee867dcf
|
912 913 914 915 916 917 |
{ .name = "stat", .seq_show = blkcg_print_stat, }, { } /* terminate */ }; |
e1f3b9412
|
918 |
static struct cftype blkcg_legacy_files[] = { |
31e4c28d9
|
919 |
{ |
84c124da9
|
920 |
.name = "reset_stats", |
3c798398e
|
921 |
.write_u64 = blkcg_reset_stats, |
220841906
|
922 |
}, |
4baf6e332
|
923 |
{ } /* terminate */ |
31e4c28d9
|
924 |
}; |
59b57717f
|
925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 |
/* * blkcg destruction is a three-stage process. * * 1. Destruction starts. The blkcg_css_offline() callback is invoked * which offlines writeback. Here we tie the next stage of blkg destruction * to the completion of writeback associated with the blkcg. This lets us * avoid punting potentially large amounts of outstanding writeback to root * while maintaining any ongoing policies. The next stage is triggered when * the nr_cgwbs count goes to zero. * * 2. When the nr_cgwbs count goes to zero, blkcg_destroy_blkgs() is called * and handles the destruction of blkgs. Here the css reference held by * the blkg is put back eventually allowing blkcg_css_free() to be called. * This work may occur in cgwb_release_workfn() on the cgwb_release * workqueue. Any submitted ios that fail to get the blkg ref will be * punted to the root_blkg. * * 3. Once the blkcg ref count goes to zero, blkcg_css_free() is called. * This finally frees the blkcg. */ |
9f13ef678
|
945 |
/** |
92fb97487
|
946 |
* blkcg_css_offline - cgroup css_offline callback |
eb95419b0
|
947 |
* @css: css of interest |
9f13ef678
|
948 |
* |
59b57717f
|
949 950 951 |
* This function is called when @css is about to go away. Here the cgwbs are * offlined first and only once writeback associated with the blkcg has * finished do we start step 2 (see above). |
9f13ef678
|
952 |
*/ |
eb95419b0
|
953 |
static void blkcg_css_offline(struct cgroup_subsys_state *css) |
31e4c28d9
|
954 |
{ |
eb95419b0
|
955 |
struct blkcg *blkcg = css_to_blkcg(css); |
b1c357696
|
956 |
|
59b57717f
|
957 958 |
/* this prevents anyone from attaching or migrating to this blkcg */ wb_blkcg_offline(blkcg); |
d866dbf61
|
959 960 |
/* put the base online pin allowing step 2 to be triggered */ blkcg_unpin_online(blkcg); |
59b57717f
|
961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 |
} /** * blkcg_destroy_blkgs - responsible for shooting down blkgs * @blkcg: blkcg of interest * * blkgs should be removed while holding both q and blkcg locks. As blkcg lock * is nested inside q lock, this function performs reverse double lock dancing. * Destroying the blkgs releases the reference held on the blkcg's css allowing * blkcg_css_free to eventually be called. * * This is the blkcg counterpart of ioc_release_fn(). */ void blkcg_destroy_blkgs(struct blkcg *blkcg) { |
9f13ef678
|
976 |
spin_lock_irq(&blkcg->lock); |
7ee9c5620
|
977 |
|
4c6994806
|
978 979 |
while (!hlist_empty(&blkcg->blkg_list)) { struct blkcg_gq *blkg = hlist_entry(blkcg->blkg_list.first, |
6b0654620
|
980 |
struct blkcg_gq, blkcg_node); |
4c6994806
|
981 |
struct request_queue *q = blkg->q; |
0d945c1f9
|
982 |
if (spin_trylock(&q->queue_lock)) { |
4c6994806
|
983 |
blkg_destroy(blkg); |
0d945c1f9
|
984 |
spin_unlock(&q->queue_lock); |
4c6994806
|
985 986 987 988 989 990 |
} else { spin_unlock_irq(&blkcg->lock); cpu_relax(); spin_lock_irq(&blkcg->lock); } } |
6b0654620
|
991 |
|
4c6994806
|
992 993 |
spin_unlock_irq(&blkcg->lock); } |
eb95419b0
|
994 |
static void blkcg_css_free(struct cgroup_subsys_state *css) |
7ee9c5620
|
995 |
{ |
eb95419b0
|
996 |
struct blkcg *blkcg = css_to_blkcg(css); |
bc915e61c
|
997 |
int i; |
7ee9c5620
|
998 |
|
7876f930d
|
999 |
mutex_lock(&blkcg_pol_mutex); |
e4a9bde95
|
1000 |
|
7876f930d
|
1001 |
list_del(&blkcg->all_blkcgs_node); |
7876f930d
|
1002 |
|
bc915e61c
|
1003 |
for (i = 0; i < BLKCG_MAX_POLS; i++) |
e4a9bde95
|
1004 1005 1006 1007 |
if (blkcg->cpd[i]) blkcg_policy[i]->cpd_free_fn(blkcg->cpd[i]); mutex_unlock(&blkcg_pol_mutex); |
bc915e61c
|
1008 |
kfree(blkcg); |
31e4c28d9
|
1009 |
} |
eb95419b0
|
1010 1011 |
static struct cgroup_subsys_state * blkcg_css_alloc(struct cgroup_subsys_state *parent_css) |
31e4c28d9
|
1012 |
{ |
3c798398e
|
1013 |
struct blkcg *blkcg; |
e48453c38
|
1014 1015 |
struct cgroup_subsys_state *ret; int i; |
31e4c28d9
|
1016 |
|
7876f930d
|
1017 |
mutex_lock(&blkcg_pol_mutex); |
eb95419b0
|
1018 |
if (!parent_css) { |
3c798398e
|
1019 |
blkcg = &blkcg_root; |
bc915e61c
|
1020 1021 1022 1023 |
} else { blkcg = kzalloc(sizeof(*blkcg), GFP_KERNEL); if (!blkcg) { ret = ERR_PTR(-ENOMEM); |
4c18c9e96
|
1024 |
goto unlock; |
bc915e61c
|
1025 |
} |
e48453c38
|
1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 |
} 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
|
1038 |
if (!pol || !pol->cpd_alloc_fn) |
e48453c38
|
1039 |
continue; |
e4a9bde95
|
1040 |
cpd = pol->cpd_alloc_fn(GFP_KERNEL); |
e48453c38
|
1041 1042 1043 1044 |
if (!cpd) { ret = ERR_PTR(-ENOMEM); goto free_pd_blkcg; } |
814376483
|
1045 1046 |
blkcg->cpd[i] = cpd; cpd->blkcg = blkcg; |
e48453c38
|
1047 |
cpd->plid = i; |
e4a9bde95
|
1048 1049 |
if (pol->cpd_init_fn) pol->cpd_init_fn(cpd); |
e48453c38
|
1050 |
} |
31e4c28d9
|
1051 |
|
31e4c28d9
|
1052 |
spin_lock_init(&blkcg->lock); |
d866dbf61
|
1053 |
refcount_set(&blkcg->online_pin, 1); |
e00f4f4d0
|
1054 |
INIT_RADIX_TREE(&blkcg->blkg_tree, GFP_NOWAIT | __GFP_NOWARN); |
31e4c28d9
|
1055 |
INIT_HLIST_HEAD(&blkcg->blkg_list); |
52ebea749
|
1056 1057 1058 |
#ifdef CONFIG_CGROUP_WRITEBACK INIT_LIST_HEAD(&blkcg->cgwb_list); #endif |
7876f930d
|
1059 1060 1061 |
list_add_tail(&blkcg->all_blkcgs_node, &all_blkcgs); mutex_unlock(&blkcg_pol_mutex); |
31e4c28d9
|
1062 |
return &blkcg->css; |
e48453c38
|
1063 1064 1065 |
free_pd_blkcg: for (i--; i >= 0; i--) |
e4a9bde95
|
1066 1067 |
if (blkcg->cpd[i]) blkcg_policy[i]->cpd_free_fn(blkcg->cpd[i]); |
4c18c9e96
|
1068 1069 1070 1071 |
if (blkcg != &blkcg_root) kfree(blkcg); unlock: |
7876f930d
|
1072 |
mutex_unlock(&blkcg_pol_mutex); |
e48453c38
|
1073 |
return ret; |
31e4c28d9
|
1074 |
} |
4308a434e
|
1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 |
static int blkcg_css_online(struct cgroup_subsys_state *css) { struct blkcg *blkcg = css_to_blkcg(css); struct blkcg *parent = blkcg_parent(blkcg); /* * blkcg_pin_online() is used to delay blkcg offline so that blkgs * don't go offline while cgwbs are still active on them. Pin the * parent so that offline always happens towards the root. */ if (parent) blkcg_pin_online(parent); return 0; } |
5efd61135
|
1089 1090 1091 1092 |
/** * blkcg_init_queue - initialize blkcg part of request queue * @q: request_queue to initialize * |
c62b37d96
|
1093 |
* Called from blk_alloc_queue(). Responsible for initializing blkcg |
5efd61135
|
1094 1095 1096 1097 1098 1099 1100 |
* part of new request_queue @q. * * RETURNS: * 0 on success, -errno on failure. */ int blkcg_init_queue(struct request_queue *q) { |
d708f0d50
|
1101 1102 |
struct blkcg_gq *new_blkg, *blkg; bool preloaded; |
ec13b1d6f
|
1103 |
int ret; |
d708f0d50
|
1104 1105 1106 1107 1108 |
new_blkg = blkg_alloc(&blkcg_root, q, GFP_KERNEL); if (!new_blkg) return -ENOMEM; preloaded = !radix_tree_preload(GFP_KERNEL); |
bea548831
|
1109 |
/* Make sure the root blkg exists. */ |
ec13b1d6f
|
1110 |
rcu_read_lock(); |
0d945c1f9
|
1111 |
spin_lock_irq(&q->queue_lock); |
d708f0d50
|
1112 |
blkg = blkg_create(&blkcg_root, q, new_blkg); |
901932a3f
|
1113 1114 1115 |
if (IS_ERR(blkg)) goto err_unlock; q->root_blkg = blkg; |
0d945c1f9
|
1116 |
spin_unlock_irq(&q->queue_lock); |
ec13b1d6f
|
1117 |
rcu_read_unlock(); |
d708f0d50
|
1118 1119 |
if (preloaded) radix_tree_preload_end(); |
27029b4b1
|
1120 |
ret = blk_throtl_init(q); |
04be60b5e
|
1121 1122 |
if (ret) goto err_destroy_all; |
d70675121
|
1123 |
|
27029b4b1
|
1124 1125 1126 |
ret = blk_iolatency_init(q); if (ret) { blk_throtl_exit(q); |
04be60b5e
|
1127 |
goto err_destroy_all; |
27029b4b1
|
1128 |
} |
04be60b5e
|
1129 |
return 0; |
901932a3f
|
1130 |
|
04be60b5e
|
1131 |
err_destroy_all: |
04be60b5e
|
1132 |
blkg_destroy_all(q); |
04be60b5e
|
1133 |
return ret; |
901932a3f
|
1134 |
err_unlock: |
0d945c1f9
|
1135 |
spin_unlock_irq(&q->queue_lock); |
901932a3f
|
1136 1137 1138 1139 |
rcu_read_unlock(); if (preloaded) radix_tree_preload_end(); return PTR_ERR(blkg); |
5efd61135
|
1140 1141 1142 |
} /** |
5efd61135
|
1143 1144 1145 |
* blkcg_exit_queue - exit and release blkcg part of request_queue * @q: request_queue being released * |
7585d5082
|
1146 |
* Called from blk_exit_queue(). Responsible for exiting blkcg part. |
5efd61135
|
1147 1148 1149 |
*/ void blkcg_exit_queue(struct request_queue *q) { |
3c96cb32d
|
1150 |
blkg_destroy_all(q); |
5efd61135
|
1151 1152 |
blk_throtl_exit(q); } |
31e4c28d9
|
1153 1154 1155 1156 1157 1158 |
/* * 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
|
1159 |
static int blkcg_can_attach(struct cgroup_taskset *tset) |
31e4c28d9
|
1160 |
{ |
bb9d97b6d
|
1161 |
struct task_struct *task; |
1f7dd3e5a
|
1162 |
struct cgroup_subsys_state *dst_css; |
31e4c28d9
|
1163 1164 1165 1166 |
struct io_context *ioc; int ret = 0; /* task_lock() is needed to avoid races with exit_io_context() */ |
1f7dd3e5a
|
1167 |
cgroup_taskset_for_each(task, dst_css, tset) { |
bb9d97b6d
|
1168 1169 1170 1171 1172 1173 1174 1175 |
task_lock(task); ioc = task->io_context; if (ioc && atomic_read(&ioc->nr_tasks) > 1) ret = -EINVAL; task_unlock(task); if (ret) break; } |
31e4c28d9
|
1176 1177 |
return ret; } |
69d7fde59
|
1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 |
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); } |
d09d8df3a
|
1197 1198 1199 1200 1201 1202 |
static void blkcg_exit(struct task_struct *tsk) { if (tsk->throttle_queue) blk_put_queue(tsk->throttle_queue); tsk->throttle_queue = NULL; } |
c165b3e3c
|
1203 |
struct cgroup_subsys io_cgrp_subsys = { |
92fb97487
|
1204 |
.css_alloc = blkcg_css_alloc, |
4308a434e
|
1205 |
.css_online = blkcg_css_online, |
92fb97487
|
1206 1207 |
.css_offline = blkcg_css_offline, .css_free = blkcg_css_free, |
3c798398e
|
1208 |
.can_attach = blkcg_can_attach, |
f73316482
|
1209 |
.css_rstat_flush = blkcg_rstat_flush, |
69d7fde59
|
1210 |
.bind = blkcg_bind, |
2ee867dcf
|
1211 |
.dfl_cftypes = blkcg_files, |
880f50e22
|
1212 |
.legacy_cftypes = blkcg_legacy_files, |
c165b3e3c
|
1213 |
.legacy_name = "blkio", |
d09d8df3a
|
1214 |
.exit = blkcg_exit, |
1ced953b1
|
1215 1216 1217 1218 1219 1220 1221 1222 |
#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
|
1223 |
}; |
c165b3e3c
|
1224 |
EXPORT_SYMBOL_GPL(io_cgrp_subsys); |
676f7c8f8
|
1225 |
|
8bd435b30
|
1226 |
/** |
a2b1693ba
|
1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 |
* 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
|
1243 |
const struct blkcg_policy *pol) |
a2b1693ba
|
1244 |
{ |
4c55f4f9a
|
1245 |
struct blkg_policy_data *pd_prealloc = NULL; |
9d179b865
|
1246 |
struct blkcg_gq *blkg, *pinned_blkg = NULL; |
4c55f4f9a
|
1247 |
int ret; |
a2b1693ba
|
1248 1249 1250 |
if (blkcg_policy_enabled(q, pol)) return 0; |
344e9ffcb
|
1251 |
if (queue_is_mq(q)) |
bd166ef18
|
1252 |
blk_mq_freeze_queue(q); |
9d179b865
|
1253 |
retry: |
0d945c1f9
|
1254 |
spin_lock_irq(&q->queue_lock); |
a2b1693ba
|
1255 |
|
9d179b865
|
1256 |
/* blkg_list is pushed at the head, reverse walk to allocate parents first */ |
71c814077
|
1257 |
list_for_each_entry_reverse(blkg, &q->blkg_list, q_node) { |
4c55f4f9a
|
1258 1259 1260 1261 |
struct blkg_policy_data *pd; if (blkg->pd[pol->plid]) continue; |
a2b1693ba
|
1262 |
|
9d179b865
|
1263 1264 1265 1266 1267 1268 1269 1270 |
/* If prealloc matches, use it; otherwise try GFP_NOWAIT */ if (blkg == pinned_blkg) { pd = pd_prealloc; pd_prealloc = NULL; } else { pd = pol->pd_alloc_fn(GFP_NOWAIT | __GFP_NOWARN, q, blkg->blkcg); } |
4c55f4f9a
|
1271 |
if (!pd) { |
9d179b865
|
1272 1273 1274 1275 1276 1277 1278 1279 |
/* * GFP_NOWAIT failed. Free the existing one and * prealloc for @blkg w/ GFP_KERNEL. */ if (pinned_blkg) blkg_put(pinned_blkg); blkg_get(blkg); pinned_blkg = blkg; |
0d945c1f9
|
1280 |
spin_unlock_irq(&q->queue_lock); |
9d179b865
|
1281 1282 1283 1284 1285 1286 1287 1288 1289 |
if (pd_prealloc) pol->pd_free_fn(pd_prealloc); pd_prealloc = pol->pd_alloc_fn(GFP_KERNEL, q, blkg->blkcg); if (pd_prealloc) goto retry; else goto enomem; |
4c55f4f9a
|
1290 |
} |
a2b1693ba
|
1291 1292 1293 |
blkg->pd[pol->plid] = pd; pd->blkg = blkg; |
b276a876a
|
1294 |
pd->plid = pol->plid; |
a2b1693ba
|
1295 |
} |
9d179b865
|
1296 1297 1298 1299 |
/* all allocated, init in the same order */ if (pol->pd_init_fn) list_for_each_entry_reverse(blkg, &q->blkg_list, q_node) pol->pd_init_fn(blkg->pd[pol->plid]); |
a2b1693ba
|
1300 1301 |
__set_bit(pol->plid, q->blkcg_pols); ret = 0; |
4c55f4f9a
|
1302 |
|
0d945c1f9
|
1303 |
spin_unlock_irq(&q->queue_lock); |
9d179b865
|
1304 |
out: |
344e9ffcb
|
1305 |
if (queue_is_mq(q)) |
bd166ef18
|
1306 |
blk_mq_unfreeze_queue(q); |
9d179b865
|
1307 1308 |
if (pinned_blkg) blkg_put(pinned_blkg); |
001bea73e
|
1309 1310 |
if (pd_prealloc) pol->pd_free_fn(pd_prealloc); |
a2b1693ba
|
1311 |
return ret; |
9d179b865
|
1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 |
enomem: /* alloc failed, nothing's initialized yet, free everything */ spin_lock_irq(&q->queue_lock); list_for_each_entry(blkg, &q->blkg_list, q_node) { if (blkg->pd[pol->plid]) { pol->pd_free_fn(blkg->pd[pol->plid]); blkg->pd[pol->plid] = NULL; } } spin_unlock_irq(&q->queue_lock); ret = -ENOMEM; goto out; |
a2b1693ba
|
1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 |
} 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
|
1337 |
const struct blkcg_policy *pol) |
a2b1693ba
|
1338 |
{ |
3c798398e
|
1339 |
struct blkcg_gq *blkg; |
a2b1693ba
|
1340 1341 1342 |
if (!blkcg_policy_enabled(q, pol)) return; |
344e9ffcb
|
1343 |
if (queue_is_mq(q)) |
bd166ef18
|
1344 |
blk_mq_freeze_queue(q); |
bd166ef18
|
1345 |
|
0d945c1f9
|
1346 |
spin_lock_irq(&q->queue_lock); |
a2b1693ba
|
1347 1348 1349 1350 |
__clear_bit(pol->plid, q->blkcg_pols); list_for_each_entry(blkg, &q->blkg_list, q_node) { |
001bea73e
|
1351 |
if (blkg->pd[pol->plid]) { |
6b0654620
|
1352 |
if (pol->pd_offline_fn) |
a9520cd6f
|
1353 |
pol->pd_offline_fn(blkg->pd[pol->plid]); |
001bea73e
|
1354 1355 1356 |
pol->pd_free_fn(blkg->pd[pol->plid]); blkg->pd[pol->plid] = NULL; } |
a2b1693ba
|
1357 |
} |
0d945c1f9
|
1358 |
spin_unlock_irq(&q->queue_lock); |
bd166ef18
|
1359 |
|
344e9ffcb
|
1360 |
if (queue_is_mq(q)) |
bd166ef18
|
1361 |
blk_mq_unfreeze_queue(q); |
a2b1693ba
|
1362 1363 1364 1365 |
} EXPORT_SYMBOL_GPL(blkcg_deactivate_policy); /** |
3c798398e
|
1366 1367 |
* blkcg_policy_register - register a blkcg policy * @pol: blkcg policy to register |
8bd435b30
|
1368 |
* |
3c798398e
|
1369 1370 |
* Register @pol with blkcg core. Might sleep and @pol may be modified on * successful registration. Returns 0 on success and -errno on failure. |
8bd435b30
|
1371 |
*/ |
d5bf02914
|
1372 |
int blkcg_policy_register(struct blkcg_policy *pol) |
3e2520668
|
1373 |
{ |
06b285bd1
|
1374 |
struct blkcg *blkcg; |
8bd435b30
|
1375 |
int i, ret; |
e8989fae3
|
1376 |
|
838f13bf4
|
1377 |
mutex_lock(&blkcg_pol_register_mutex); |
bc0d6501a
|
1378 |
mutex_lock(&blkcg_pol_mutex); |
8bd435b30
|
1379 1380 1381 |
/* find an empty slot */ ret = -ENOSPC; for (i = 0; i < BLKCG_MAX_POLS; i++) |
3c798398e
|
1382 |
if (!blkcg_policy[i]) |
8bd435b30
|
1383 |
break; |
01c5f85ae
|
1384 1385 1386 |
if (i >= BLKCG_MAX_POLS) { pr_warn("blkcg_policy_register: BLKCG_MAX_POLS too small "); |
838f13bf4
|
1387 |
goto err_unlock; |
01c5f85ae
|
1388 |
} |
035d10b2f
|
1389 |
|
e84010732
|
1390 1391 1392 1393 |
/* Make sure cpd/pd_alloc_fn and cpd/pd_free_fn in pairs */ if ((!pol->cpd_alloc_fn ^ !pol->cpd_free_fn) || (!pol->pd_alloc_fn ^ !pol->pd_free_fn)) goto err_unlock; |
06b285bd1
|
1394 |
/* register @pol */ |
3c798398e
|
1395 |
pol->plid = i; |
06b285bd1
|
1396 1397 1398 |
blkcg_policy[pol->plid] = pol; /* allocate and install cpd's */ |
e4a9bde95
|
1399 |
if (pol->cpd_alloc_fn) { |
06b285bd1
|
1400 1401 |
list_for_each_entry(blkcg, &all_blkcgs, all_blkcgs_node) { struct blkcg_policy_data *cpd; |
e4a9bde95
|
1402 |
cpd = pol->cpd_alloc_fn(GFP_KERNEL); |
bbb427e34
|
1403 |
if (!cpd) |
06b285bd1
|
1404 |
goto err_free_cpds; |
06b285bd1
|
1405 |
|
814376483
|
1406 1407 |
blkcg->cpd[pol->plid] = cpd; cpd->blkcg = blkcg; |
06b285bd1
|
1408 |
cpd->plid = pol->plid; |
86a5bba5c
|
1409 1410 |
if (pol->cpd_init_fn) pol->cpd_init_fn(cpd); |
06b285bd1
|
1411 1412 |
} } |
838f13bf4
|
1413 |
mutex_unlock(&blkcg_pol_mutex); |
8bd435b30
|
1414 |
|
8bd435b30
|
1415 |
/* everything is in place, add intf files for the new policy */ |
2ee867dcf
|
1416 1417 1418 |
if (pol->dfl_cftypes) WARN_ON(cgroup_add_dfl_cftypes(&io_cgrp_subsys, pol->dfl_cftypes)); |
880f50e22
|
1419 |
if (pol->legacy_cftypes) |
c165b3e3c
|
1420 |
WARN_ON(cgroup_add_legacy_cftypes(&io_cgrp_subsys, |
880f50e22
|
1421 |
pol->legacy_cftypes)); |
838f13bf4
|
1422 1423 |
mutex_unlock(&blkcg_pol_register_mutex); return 0; |
06b285bd1
|
1424 |
err_free_cpds: |
58a9edce0
|
1425 |
if (pol->cpd_free_fn) { |
06b285bd1
|
1426 |
list_for_each_entry(blkcg, &all_blkcgs, all_blkcgs_node) { |
e4a9bde95
|
1427 1428 1429 1430 |
if (blkcg->cpd[pol->plid]) { pol->cpd_free_fn(blkcg->cpd[pol->plid]); blkcg->cpd[pol->plid] = NULL; } |
06b285bd1
|
1431 1432 1433 |
} } blkcg_policy[pol->plid] = NULL; |
838f13bf4
|
1434 |
err_unlock: |
bc0d6501a
|
1435 |
mutex_unlock(&blkcg_pol_mutex); |
838f13bf4
|
1436 |
mutex_unlock(&blkcg_pol_register_mutex); |
8bd435b30
|
1437 |
return ret; |
3e2520668
|
1438 |
} |
3c798398e
|
1439 |
EXPORT_SYMBOL_GPL(blkcg_policy_register); |
3e2520668
|
1440 |
|
8bd435b30
|
1441 |
/** |
3c798398e
|
1442 1443 |
* blkcg_policy_unregister - unregister a blkcg policy * @pol: blkcg policy to unregister |
8bd435b30
|
1444 |
* |
3c798398e
|
1445 |
* Undo blkcg_policy_register(@pol). Might sleep. |
8bd435b30
|
1446 |
*/ |
3c798398e
|
1447 |
void blkcg_policy_unregister(struct blkcg_policy *pol) |
3e2520668
|
1448 |
{ |
06b285bd1
|
1449 |
struct blkcg *blkcg; |
838f13bf4
|
1450 |
mutex_lock(&blkcg_pol_register_mutex); |
bc0d6501a
|
1451 |
|
3c798398e
|
1452 |
if (WARN_ON(blkcg_policy[pol->plid] != pol)) |
8bd435b30
|
1453 1454 1455 |
goto out_unlock; /* kill the intf files first */ |
2ee867dcf
|
1456 1457 |
if (pol->dfl_cftypes) cgroup_rm_cftypes(pol->dfl_cftypes); |
880f50e22
|
1458 1459 |
if (pol->legacy_cftypes) cgroup_rm_cftypes(pol->legacy_cftypes); |
44ea53de4
|
1460 |
|
06b285bd1
|
1461 |
/* remove cpds and unregister */ |
838f13bf4
|
1462 |
mutex_lock(&blkcg_pol_mutex); |
06b285bd1
|
1463 |
|
58a9edce0
|
1464 |
if (pol->cpd_free_fn) { |
06b285bd1
|
1465 |
list_for_each_entry(blkcg, &all_blkcgs, all_blkcgs_node) { |
e4a9bde95
|
1466 1467 1468 1469 |
if (blkcg->cpd[pol->plid]) { pol->cpd_free_fn(blkcg->cpd[pol->plid]); blkcg->cpd[pol->plid] = NULL; } |
06b285bd1
|
1470 1471 |
} } |
3c798398e
|
1472 |
blkcg_policy[pol->plid] = NULL; |
06b285bd1
|
1473 |
|
bc0d6501a
|
1474 |
mutex_unlock(&blkcg_pol_mutex); |
838f13bf4
|
1475 1476 |
out_unlock: mutex_unlock(&blkcg_pol_register_mutex); |
3e2520668
|
1477 |
} |
3c798398e
|
1478 |
EXPORT_SYMBOL_GPL(blkcg_policy_unregister); |
903d23f0a
|
1479 |
|
d3f77dfdc
|
1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 |
bool __blkcg_punt_bio_submit(struct bio *bio) { struct blkcg_gq *blkg = bio->bi_blkg; /* consume the flag first */ bio->bi_opf &= ~REQ_CGROUP_PUNT; /* never bounce for the root cgroup */ if (!blkg->parent) return false; spin_lock_bh(&blkg->async_bio_lock); bio_list_add(&blkg->async_bios, bio); spin_unlock_bh(&blkg->async_bio_lock); queue_work(blkcg_punt_bio_wq, &blkg->async_bio_work); return true; } |
d09d8df3a
|
1498 1499 1500 1501 1502 1503 1504 1505 1506 |
/* * Scale the accumulated delay based on how long it has been since we updated * the delay. We only call this when we are adding delay, in case it's been a * while since we added delay, and when we are checking to see if we need to * delay a task, to account for any delays that may have occurred. */ static void blkcg_scale_delay(struct blkcg_gq *blkg, u64 now) { u64 old = atomic64_read(&blkg->delay_start); |
54c52e10d
|
1507 1508 1509 |
/* negative use_delay means no scaling, see blkcg_set_delay() */ if (atomic_read(&blkg->use_delay) < 0) return; |
d09d8df3a
|
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 1559 1560 |
/* * We only want to scale down every second. The idea here is that we * want to delay people for min(delay_nsec, NSEC_PER_SEC) in a certain * time window. We only want to throttle tasks for recent delay that * has occurred, in 1 second time windows since that's the maximum * things can be throttled. We save the current delay window in * blkg->last_delay so we know what amount is still left to be charged * to the blkg from this point onward. blkg->last_use keeps track of * the use_delay counter. The idea is if we're unthrottling the blkg we * are ok with whatever is happening now, and we can take away more of * the accumulated delay as we've already throttled enough that * everybody is happy with their IO latencies. */ if (time_before64(old + NSEC_PER_SEC, now) && atomic64_cmpxchg(&blkg->delay_start, old, now) == old) { u64 cur = atomic64_read(&blkg->delay_nsec); u64 sub = min_t(u64, blkg->last_delay, now - old); int cur_use = atomic_read(&blkg->use_delay); /* * We've been unthrottled, subtract a larger chunk of our * accumulated delay. */ if (cur_use < blkg->last_use) sub = max_t(u64, sub, blkg->last_delay >> 1); /* * This shouldn't happen, but handle it anyway. Our delay_nsec * should only ever be growing except here where we subtract out * min(last_delay, 1 second), but lord knows bugs happen and I'd * rather not end up with negative numbers. */ if (unlikely(cur < sub)) { atomic64_set(&blkg->delay_nsec, 0); blkg->last_delay = 0; } else { atomic64_sub(sub, &blkg->delay_nsec); blkg->last_delay = cur - sub; } blkg->last_use = cur_use; } } /* * This is called when we want to actually walk up the hierarchy and check to * see if we need to throttle, and then actually throttle if there is some * accumulated delay. This should only be called upon return to user space so * we're not holding some lock that would induce a priority inversion. */ static void blkcg_maybe_throttle_blkg(struct blkcg_gq *blkg, bool use_memdelay) { |
fd112c746
|
1561 |
unsigned long pflags; |
5160a5a53
|
1562 |
bool clamp; |
d09d8df3a
|
1563 1564 1565 1566 1567 1568 |
u64 now = ktime_to_ns(ktime_get()); u64 exp; u64 delay_nsec = 0; int tok; while (blkg->parent) { |
5160a5a53
|
1569 1570 1571 1572 |
int use_delay = atomic_read(&blkg->use_delay); if (use_delay) { u64 this_delay; |
d09d8df3a
|
1573 |
blkcg_scale_delay(blkg, now); |
5160a5a53
|
1574 1575 1576 1577 1578 |
this_delay = atomic64_read(&blkg->delay_nsec); if (this_delay > delay_nsec) { delay_nsec = this_delay; clamp = use_delay > 0; } |
d09d8df3a
|
1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 |
} blkg = blkg->parent; } if (!delay_nsec) return; /* * Let's not sleep for all eternity if we've amassed a huge delay. * Swapping or metadata IO can accumulate 10's of seconds worth of * delay, and we want userspace to be able to do _something_ so cap the |
5160a5a53
|
1590 1591 1592 1593 |
* delays at 0.25s. If there's 10's of seconds worth of delay then the * tasks will be delayed for 0.25 second for every syscall. If * blkcg_set_delay() was used as indicated by negative use_delay, the * caller is responsible for regulating the range. |
d09d8df3a
|
1594 |
*/ |
5160a5a53
|
1595 1596 |
if (clamp) delay_nsec = min_t(u64, delay_nsec, 250 * NSEC_PER_MSEC); |
d09d8df3a
|
1597 |
|
fd112c746
|
1598 1599 |
if (use_memdelay) psi_memstall_enter(&pflags); |
d09d8df3a
|
1600 1601 1602 1603 1604 1605 1606 1607 1608 |
exp = ktime_add_ns(now, delay_nsec); tok = io_schedule_prepare(); do { __set_current_state(TASK_KILLABLE); if (!schedule_hrtimeout(&exp, HRTIMER_MODE_ABS)) break; } while (!fatal_signal_pending(current)); io_schedule_finish(tok); |
fd112c746
|
1609 1610 1611 |
if (use_memdelay) psi_memstall_leave(&pflags); |
d09d8df3a
|
1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 |
} /** * blkcg_maybe_throttle_current - throttle the current task if it has been marked * * This is only called if we've been marked with set_notify_resume(). Obviously * we can be set_notify_resume() for reasons other than blkcg throttling, so we * check to see if current->throttle_queue is set and if not this doesn't do * anything. This should only ever be called by the resume code, it's not meant * to be called by people willy-nilly as it will actually do the work to * throttle the task if it is setup for throttling. */ void blkcg_maybe_throttle_current(void) { struct request_queue *q = current->throttle_queue; struct cgroup_subsys_state *css; struct blkcg *blkcg; struct blkcg_gq *blkg; bool use_memdelay = current->use_memdelay; if (!q) return; current->throttle_queue = NULL; current->use_memdelay = false; rcu_read_lock(); css = kthread_blkcg(); if (css) blkcg = css_to_blkcg(css); else blkcg = css_to_blkcg(task_css(current, io_cgrp_id)); if (!blkcg) goto out; blkg = blkg_lookup(blkcg, q); if (!blkg) goto out; |
7754f669f
|
1650 |
if (!blkg_tryget(blkg)) |
d09d8df3a
|
1651 1652 |
goto out; rcu_read_unlock(); |
d09d8df3a
|
1653 1654 1655 |
blkcg_maybe_throttle_blkg(blkg, use_memdelay); blkg_put(blkg); |
cc7ecc258
|
1656 |
blk_put_queue(q); |
d09d8df3a
|
1657 1658 1659 1660 1661 |
return; out: rcu_read_unlock(); blk_put_queue(q); } |
d09d8df3a
|
1662 1663 1664 |
/** * blkcg_schedule_throttle - this task needs to check for throttling |
537d71b3f
|
1665 1666 |
* @q: the request queue IO was submitted on * @use_memdelay: do we charge this to memory delay for PSI |
d09d8df3a
|
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 |
* * This is called by the IO controller when we know there's delay accumulated * for the blkg for this task. We do not pass the blkg because there are places * we call this that may not have that information, the swapping code for * instance will only have a request_queue at that point. This set's the * notify_resume for the task to check and see if it requires throttling before * returning to user space. * * We will only schedule once per syscall. You can call this over and over * again and it will only do the check once upon return to user space, and only * throttle once. If the task needs to be throttled again it'll need to be * re-set at the next time we see the task. */ void blkcg_schedule_throttle(struct request_queue *q, bool use_memdelay) { if (unlikely(current->flags & PF_KTHREAD)) return; if (!blk_get_queue(q)) return; if (current->throttle_queue) blk_put_queue(current->throttle_queue); current->throttle_queue = q; if (use_memdelay) current->use_memdelay = use_memdelay; set_notify_resume(current); } |
d09d8df3a
|
1695 1696 1697 |
/** * blkcg_add_delay - add delay to this blkg |
537d71b3f
|
1698 1699 1700 |
* @blkg: blkg of interest * @now: the current time in nanoseconds * @delta: how many nanoseconds of delay to add |
d09d8df3a
|
1701 1702 1703 1704 1705 1706 |
* * Charge @delta to the blkg's current delay accumulation. This is used to * throttle tasks if an IO controller thinks we need more throttling. */ void blkcg_add_delay(struct blkcg_gq *blkg, u64 now, u64 delta) { |
54c52e10d
|
1707 1708 |
if (WARN_ON_ONCE(atomic_read(&blkg->use_delay) < 0)) return; |
d09d8df3a
|
1709 1710 1711 |
blkcg_scale_delay(blkg, now); atomic64_add(delta, &blkg->delay_nsec); } |
d09d8df3a
|
1712 |
|
28fc591ff
|
1713 1714 |
/** * blkg_tryget_closest - try and get a blkg ref on the closet blkg |
13c7863d4
|
1715 1716 |
* @bio: target bio * @css: target css |
28fc591ff
|
1717 |
* |
13c7863d4
|
1718 1719 1720 |
* As the failure mode here is to walk up the blkg tree, this ensure that the * blkg->parent pointers are always valid. This returns the blkg that it ended * up taking a reference on or %NULL if no reference was taken. |
28fc591ff
|
1721 |
*/ |
13c7863d4
|
1722 1723 |
static inline struct blkcg_gq *blkg_tryget_closest(struct bio *bio, struct cgroup_subsys_state *css) |
28fc591ff
|
1724 |
{ |
13c7863d4
|
1725 |
struct blkcg_gq *blkg, *ret_blkg = NULL; |
28fc591ff
|
1726 |
|
13c7863d4
|
1727 1728 |
rcu_read_lock(); blkg = blkg_lookup_create(css_to_blkcg(css), bio->bi_disk->queue); |
28fc591ff
|
1729 1730 1731 1732 1733 1734 1735 |
while (blkg) { if (blkg_tryget(blkg)) { ret_blkg = blkg; break; } blkg = blkg->parent; } |
13c7863d4
|
1736 |
rcu_read_unlock(); |
28fc591ff
|
1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 |
return ret_blkg; } /** * bio_associate_blkg_from_css - associate a bio with a specified css * @bio: target bio * @css: target css * * Associate @bio with the blkg found by combining the css's blkg and the * request_queue of the @bio. An association failure is handled by walking up * the blkg tree. Therefore, the blkg associated can be anything between @blkg * and q->root_blkg. This situation only happens when a cgroup is dying and * then the remaining bios will spill to the closest alive blkg. * * A reference will be taken on the blkg and will be released when @bio is * freed. */ void bio_associate_blkg_from_css(struct bio *bio, struct cgroup_subsys_state *css) { |
28fc591ff
|
1758 1759 |
if (bio->bi_blkg) blkg_put(bio->bi_blkg); |
a5b97526b
|
1760 |
if (css && css->parent) { |
13c7863d4
|
1761 |
bio->bi_blkg = blkg_tryget_closest(bio, css); |
a5b97526b
|
1762 |
} else { |
13c7863d4
|
1763 1764 |
blkg_get(bio->bi_disk->queue->root_blkg); bio->bi_blkg = bio->bi_disk->queue->root_blkg; |
a5b97526b
|
1765 |
} |
28fc591ff
|
1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 |
} EXPORT_SYMBOL_GPL(bio_associate_blkg_from_css); /** * bio_associate_blkg - associate a bio with a blkg * @bio: target bio * * Associate @bio with the blkg found from the bio's css and request_queue. * If one is not found, bio_lookup_blkg() creates the blkg. If a blkg is * already associated, the css is reused and association redone as the * request_queue may have changed. */ void bio_associate_blkg(struct bio *bio) { struct cgroup_subsys_state *css; rcu_read_lock(); if (bio->bi_blkg) css = &bio_blkcg(bio)->css; else css = blkcg_css(); bio_associate_blkg_from_css(bio, css); rcu_read_unlock(); } EXPORT_SYMBOL_GPL(bio_associate_blkg); /** * bio_clone_blkg_association - clone blkg association from src to dst bio * @dst: destination bio * @src: source bio */ void bio_clone_blkg_association(struct bio *dst, struct bio *src) { if (src->bi_blkg) { if (dst->bi_blkg) blkg_put(dst->bi_blkg); blkg_get(src->bi_blkg); dst->bi_blkg = src->bi_blkg; } } EXPORT_SYMBOL_GPL(bio_clone_blkg_association); |
db18a53e5
|
1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 |
static int blk_cgroup_io_type(struct bio *bio) { if (op_is_discard(bio->bi_opf)) return BLKG_IOSTAT_DISCARD; if (op_is_write(bio->bi_opf)) return BLKG_IOSTAT_WRITE; return BLKG_IOSTAT_READ; } void blk_cgroup_bio_start(struct bio *bio) { int rwd = blk_cgroup_io_type(bio), cpu; struct blkg_iostat_set *bis; cpu = get_cpu(); bis = per_cpu_ptr(bio->bi_blkg->iostat_cpu, cpu); u64_stats_update_begin(&bis->sync); /* * If the bio is flagged with BIO_CGROUP_ACCT it means this is a split * bio and we would have already accounted for the size of the bio. */ if (!bio_flagged(bio, BIO_CGROUP_ACCT)) { bio_set_flag(bio, BIO_CGROUP_ACCT); |
0b8cc25d9
|
1834 |
bis->cur.bytes[rwd] += bio->bi_iter.bi_size; |
db18a53e5
|
1835 1836 1837 1838 1839 1840 1841 1842 |
} bis->cur.ios[rwd]++; u64_stats_update_end(&bis->sync); if (cgroup_subsys_on_dfl(io_cgrp_subsys)) cgroup_rstat_updated(bio->bi_blkg->blkcg->css.cgroup, cpu); put_cpu(); } |
d3f77dfdc
|
1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 |
static int __init blkcg_init(void) { blkcg_punt_bio_wq = alloc_workqueue("blkcg_punt_bio", WQ_MEM_RECLAIM | WQ_FREEZABLE | WQ_UNBOUND | WQ_SYSFS, 0); if (!blkcg_punt_bio_wq) return -ENOMEM; return 0; } subsys_initcall(blkcg_init); |
903d23f0a
|
1853 1854 |
module_param(blkcg_debug_stats, bool, 0644); MODULE_PARM_DESC(blkcg_debug_stats, "True if you want debug stats, false if not"); |