blk-cgroup.h 21.8 KB
Newer Older
1
/* SPDX-License-Identifier: GPL-2.0 */
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
#ifndef _BLK_CGROUP_H
#define _BLK_CGROUP_H
/*
 * 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>
 */

#include <linux/cgroup.h>
T
Tejun Heo 已提交
18
#include <linux/percpu_counter.h>
19
#include <linux/seq_file.h>
20
#include <linux/radix-tree.h>
21
#include <linux/blkdev.h>
22
#include <linux/atomic.h>
23
#include <linux/kthread.h>
24

T
Tejun Heo 已提交
25 26 27
/* percpu_counter batch for blkg_[rw]stats, per-cpu drift doesn't matter */
#define BLKG_STAT_CPU_BATCH	(INT_MAX / 2)

28 29 30
/* Max limits for throttle policy */
#define THROTL_IOPS_MAX		UINT_MAX

T
Tejun Heo 已提交
31 32
#ifdef CONFIG_BLK_CGROUP

33 34 35 36 37 38 39 40
enum blkg_rwstat_type {
	BLKG_RWSTAT_READ,
	BLKG_RWSTAT_WRITE,
	BLKG_RWSTAT_SYNC,
	BLKG_RWSTAT_ASYNC,

	BLKG_RWSTAT_NR,
	BLKG_RWSTAT_TOTAL = BLKG_RWSTAT_NR,
41 42
};

43 44
struct blkcg_gq;

T
Tejun Heo 已提交
45
struct blkcg {
46 47
	struct cgroup_subsys_state	css;
	spinlock_t			lock;
48 49

	struct radix_tree_root		blkg_tree;
50
	struct blkcg_gq	__rcu		*blkg_hint;
51
	struct hlist_head		blkg_list;
T
Tejun Heo 已提交
52

53
	struct blkcg_policy_data	*cpd[BLKCG_MAX_POLS];
54

T
Tejun Heo 已提交
55
	struct list_head		all_blkcgs_node;
56 57 58
#ifdef CONFIG_CGROUP_WRITEBACK
	struct list_head		cgwb_list;
#endif
59 60
};

61 62
/*
 * blkg_[rw]stat->aux_cnt is excluded for local stats but included for
T
Tejun Heo 已提交
63 64
 * recursive.  Used to carry stats of dead children, and, for blkg_rwstat,
 * to carry result values from read and sum operations.
65
 */
66
struct blkg_stat {
T
Tejun Heo 已提交
67
	struct percpu_counter		cpu_cnt;
68
	atomic64_t			aux_cnt;
69 70 71
};

struct blkg_rwstat {
T
Tejun Heo 已提交
72
	struct percpu_counter		cpu_cnt[BLKG_RWSTAT_NR];
73
	atomic64_t			aux_cnt[BLKG_RWSTAT_NR];
74 75
};

76 77 78 79 80
/*
 * A blkcg_gq (blkg) is association between a block cgroup (blkcg) and a
 * request_queue (q).  This is used by blkcg policies which need to track
 * information per blkcg - q pair.
 *
81 82 83 84 85
 * There can be multiple active blkcg policies and each blkg:policy pair is
 * represented by a blkg_policy_data which is allocated and freed by each
 * policy's pd_alloc/free_fn() methods.  A policy can allocate private data
 * area by allocating larger data structure which embeds blkg_policy_data
 * at the beginning.
86
 */
87
struct blkg_policy_data {
T
Tejun Heo 已提交
88
	/* the blkg and policy id this per-policy data belongs to */
T
Tejun Heo 已提交
89
	struct blkcg_gq			*blkg;
T
Tejun Heo 已提交
90
	int				plid;
91 92
};

93
/*
94 95 96 97 98
 * Policies that need to keep per-blkcg data which is independent from any
 * request_queue associated to it should implement cpd_alloc/free_fn()
 * methods.  A policy can allocate private data area by allocating larger
 * data structure which embeds blkcg_policy_data at the beginning.
 * cpd_init() is invoked to let each policy handle per-blkcg data.
99 100
 */
struct blkcg_policy_data {
101 102
	/* the blkcg and policy id this per-policy data belongs to */
	struct blkcg			*blkcg;
103 104 105
	int				plid;
};

T
Tejun Heo 已提交
106 107
/* association between a blk cgroup and a request queue */
struct blkcg_gq {
T
Tejun Heo 已提交
108
	/* Pointer to the associated request_queue */
109 110 111
	struct request_queue		*q;
	struct list_head		q_node;
	struct hlist_node		blkcg_node;
T
Tejun Heo 已提交
112
	struct blkcg			*blkcg;
T
Tejun Heo 已提交
113

114 115 116 117 118 119
	/*
	 * Each blkg gets congested separately and the congestion state is
	 * propagated to the matching bdi_writeback_congested.
	 */
	struct bdi_writeback_congested	*wb_congested;

T
Tejun Heo 已提交
120 121 122
	/* all non-root blkcg_gq's are guaranteed to have access to parent */
	struct blkcg_gq			*parent;

123 124
	/* request allocation list for this blkcg-q pair */
	struct request_list		rl;
T
Tejun Heo 已提交
125

T
Tejun Heo 已提交
126
	/* reference count */
127
	atomic_t			refcnt;
128

129 130 131
	/* is this blkg online? protected by both blkcg and q locks */
	bool				online;

132 133 134
	struct blkg_rwstat		stat_bytes;
	struct blkg_rwstat		stat_ios;

135
	struct blkg_policy_data		*pd[BLKCG_MAX_POLS];
T
Tejun Heo 已提交
136

137
	struct rcu_head			rcu_head;
138 139
};

140
typedef struct blkcg_policy_data *(blkcg_pol_alloc_cpd_fn)(gfp_t gfp);
141
typedef void (blkcg_pol_init_cpd_fn)(struct blkcg_policy_data *cpd);
142
typedef void (blkcg_pol_free_cpd_fn)(struct blkcg_policy_data *cpd);
143
typedef void (blkcg_pol_bind_cpd_fn)(struct blkcg_policy_data *cpd);
144
typedef struct blkg_policy_data *(blkcg_pol_alloc_pd_fn)(gfp_t gfp, int node);
145 146 147
typedef void (blkcg_pol_init_pd_fn)(struct blkg_policy_data *pd);
typedef void (blkcg_pol_online_pd_fn)(struct blkg_policy_data *pd);
typedef void (blkcg_pol_offline_pd_fn)(struct blkg_policy_data *pd);
148
typedef void (blkcg_pol_free_pd_fn)(struct blkg_policy_data *pd);
149
typedef void (blkcg_pol_reset_pd_stats_fn)(struct blkg_policy_data *pd);
150

T
Tejun Heo 已提交
151
struct blkcg_policy {
152 153
	int				plid;
	/* cgroup files for the policy */
154
	struct cftype			*dfl_cftypes;
155
	struct cftype			*legacy_cftypes;
156 157

	/* operations */
158
	blkcg_pol_alloc_cpd_fn		*cpd_alloc_fn;
159
	blkcg_pol_init_cpd_fn		*cpd_init_fn;
160
	blkcg_pol_free_cpd_fn		*cpd_free_fn;
161
	blkcg_pol_bind_cpd_fn		*cpd_bind_fn;
162

163
	blkcg_pol_alloc_pd_fn		*pd_alloc_fn;
164
	blkcg_pol_init_pd_fn		*pd_init_fn;
165 166
	blkcg_pol_online_pd_fn		*pd_online_fn;
	blkcg_pol_offline_pd_fn		*pd_offline_fn;
167
	blkcg_pol_free_pd_fn		*pd_free_fn;
168
	blkcg_pol_reset_pd_stats_fn	*pd_reset_stats_fn;
169 170
};

T
Tejun Heo 已提交
171
extern struct blkcg blkcg_root;
T
Tejun Heo 已提交
172
extern struct cgroup_subsys_state * const blkcg_root_css;
173

T
Tejun Heo 已提交
174 175
struct blkcg_gq *blkg_lookup_slowpath(struct blkcg *blkcg,
				      struct request_queue *q, bool update_hint);
T
Tejun Heo 已提交
176 177
struct blkcg_gq *blkg_lookup_create(struct blkcg *blkcg,
				    struct request_queue *q);
178 179 180
int blkcg_init_queue(struct request_queue *q);
void blkcg_drain_queue(struct request_queue *q);
void blkcg_exit_queue(struct request_queue *q);
181

182
/* Blkio controller policy registration */
183
int blkcg_policy_register(struct blkcg_policy *pol);
T
Tejun Heo 已提交
184
void blkcg_policy_unregister(struct blkcg_policy *pol);
185
int blkcg_activate_policy(struct request_queue *q,
T
Tejun Heo 已提交
186
			  const struct blkcg_policy *pol);
187
void blkcg_deactivate_policy(struct request_queue *q,
T
Tejun Heo 已提交
188
			     const struct blkcg_policy *pol);
189

190
const char *blkg_dev_name(struct blkcg_gq *blkg);
T
Tejun Heo 已提交
191
void blkcg_print_blkgs(struct seq_file *sf, struct blkcg *blkcg,
192 193
		       u64 (*prfill)(struct seq_file *,
				     struct blkg_policy_data *, int),
T
Tejun Heo 已提交
194
		       const struct blkcg_policy *pol, int data,
195
		       bool show_total);
196 197
u64 __blkg_prfill_u64(struct seq_file *sf, struct blkg_policy_data *pd, u64 v);
u64 __blkg_prfill_rwstat(struct seq_file *sf, struct blkg_policy_data *pd,
198
			 const struct blkg_rwstat *rwstat);
199 200 201
u64 blkg_prfill_stat(struct seq_file *sf, struct blkg_policy_data *pd, int off);
u64 blkg_prfill_rwstat(struct seq_file *sf, struct blkg_policy_data *pd,
		       int off);
202 203 204 205
int blkg_print_stat_bytes(struct seq_file *sf, void *v);
int blkg_print_stat_ios(struct seq_file *sf, void *v);
int blkg_print_stat_bytes_recursive(struct seq_file *sf, void *v);
int blkg_print_stat_ios_recursive(struct seq_file *sf, void *v);
206

207 208 209 210
u64 blkg_stat_recursive_sum(struct blkcg_gq *blkg,
			    struct blkcg_policy *pol, int off);
struct blkg_rwstat blkg_rwstat_recursive_sum(struct blkcg_gq *blkg,
					     struct blkcg_policy *pol, int off);
211

212
struct blkg_conf_ctx {
213
	struct gendisk			*disk;
T
Tejun Heo 已提交
214
	struct blkcg_gq			*blkg;
215
	char				*body;
216 217
};

T
Tejun Heo 已提交
218
int blkg_conf_prep(struct blkcg *blkcg, const struct blkcg_policy *pol,
219
		   char *input, struct blkg_conf_ctx *ctx);
220 221 222
void blkg_conf_finish(struct blkg_conf_ctx *ctx);


223 224 225 226 227
static inline struct blkcg *css_to_blkcg(struct cgroup_subsys_state *css)
{
	return css ? container_of(css, struct blkcg, css) : NULL;
}

228 229
static inline struct blkcg *bio_blkcg(struct bio *bio)
{
230 231
	struct cgroup_subsys_state *css;

232
	if (bio && bio->bi_css)
233
		return css_to_blkcg(bio->bi_css);
234 235 236 237
	css = kthread_blkcg();
	if (css)
		return css_to_blkcg(css);
	return css_to_blkcg(task_css(current, io_cgrp_id));
238 239
}

T
Tejun Heo 已提交
240 241 242 243 244 245 246 247
/**
 * blkcg_parent - get the parent of a blkcg
 * @blkcg: blkcg of interest
 *
 * Return the parent blkcg of @blkcg.  Can be called anytime.
 */
static inline struct blkcg *blkcg_parent(struct blkcg *blkcg)
{
T
Tejun Heo 已提交
248
	return css_to_blkcg(blkcg->css.parent);
T
Tejun Heo 已提交
249 250
}

T
Tejun Heo 已提交
251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267
/**
 * __blkg_lookup - internal version of blkg_lookup()
 * @blkcg: blkcg of interest
 * @q: request_queue of interest
 * @update_hint: whether to update lookup hint with the result or not
 *
 * This is internal version and shouldn't be used by policy
 * implementations.  Looks up blkgs for the @blkcg - @q pair regardless of
 * @q's bypass state.  If @update_hint is %true, the caller should be
 * holding @q->queue_lock and lookup hint is updated on success.
 */
static inline struct blkcg_gq *__blkg_lookup(struct blkcg *blkcg,
					     struct request_queue *q,
					     bool update_hint)
{
	struct blkcg_gq *blkg;

268 269 270
	if (blkcg == &blkcg_root)
		return q->root_blkg;

T
Tejun Heo 已提交
271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296
	blkg = rcu_dereference(blkcg->blkg_hint);
	if (blkg && blkg->q == q)
		return blkg;

	return blkg_lookup_slowpath(blkcg, q, update_hint);
}

/**
 * blkg_lookup - lookup blkg for the specified blkcg - q pair
 * @blkcg: blkcg of interest
 * @q: request_queue of interest
 *
 * Lookup blkg for the @blkcg - @q pair.  This function should be called
 * under RCU read lock and is guaranteed to return %NULL if @q is bypassing
 * - see blk_queue_bypass_start() for details.
 */
static inline struct blkcg_gq *blkg_lookup(struct blkcg *blkcg,
					   struct request_queue *q)
{
	WARN_ON_ONCE(!rcu_read_lock_held());

	if (unlikely(blk_queue_bypass(q)))
		return NULL;
	return __blkg_lookup(blkcg, q, false);
}

297 298 299 300 301 302 303
/**
 * blkg_to_pdata - get policy private data
 * @blkg: blkg of interest
 * @pol: policy of interest
 *
 * Return pointer to private data associated with the @blkg-@pol pair.
 */
304 305
static inline struct blkg_policy_data *blkg_to_pd(struct blkcg_gq *blkg,
						  struct blkcg_policy *pol)
306
{
307
	return blkg ? blkg->pd[pol->plid] : NULL;
308 309
}

310 311 312
static inline struct blkcg_policy_data *blkcg_to_cpd(struct blkcg *blkcg,
						     struct blkcg_policy *pol)
{
313
	return blkcg ? blkcg->cpd[pol->plid] : NULL;
314 315
}

316 317
/**
 * pdata_to_blkg - get blkg associated with policy private data
318
 * @pd: policy private data of interest
319
 *
320
 * @pd is policy private data.  Determine the blkg it's associated with.
321
 */
322
static inline struct blkcg_gq *pd_to_blkg(struct blkg_policy_data *pd)
323
{
324
	return pd ? pd->blkg : NULL;
325 326
}

327 328 329 330 331
static inline struct blkcg *cpd_to_blkcg(struct blkcg_policy_data *cpd)
{
	return cpd ? cpd->blkcg : NULL;
}

T
Tejun Heo 已提交
332 333 334 335 336 337 338 339
/**
 * blkg_path - format cgroup path of blkg
 * @blkg: blkg of interest
 * @buf: target buffer
 * @buflen: target buffer length
 *
 * Format the path of the cgroup of @blkg into @buf.
 */
T
Tejun Heo 已提交
340
static inline int blkg_path(struct blkcg_gq *blkg, char *buf, int buflen)
341
{
342
	return cgroup_path(blkg->blkcg->css.cgroup, buf, buflen);
343 344
}

T
Tejun Heo 已提交
345 346 347 348
/**
 * blkg_get - get a blkg reference
 * @blkg: blkg to get
 *
349
 * The caller should be holding an existing reference.
T
Tejun Heo 已提交
350
 */
T
Tejun Heo 已提交
351
static inline void blkg_get(struct blkcg_gq *blkg)
T
Tejun Heo 已提交
352
{
353 354
	WARN_ON_ONCE(atomic_read(&blkg->refcnt) <= 0);
	atomic_inc(&blkg->refcnt);
T
Tejun Heo 已提交
355 356
}

357
void __blkg_release_rcu(struct rcu_head *rcu);
T
Tejun Heo 已提交
358 359 360 361 362

/**
 * blkg_put - put a blkg reference
 * @blkg: blkg to put
 */
T
Tejun Heo 已提交
363
static inline void blkg_put(struct blkcg_gq *blkg)
T
Tejun Heo 已提交
364
{
365 366
	WARN_ON_ONCE(atomic_read(&blkg->refcnt) <= 0);
	if (atomic_dec_and_test(&blkg->refcnt))
367
		call_rcu(&blkg->rcu_head, __blkg_release_rcu);
T
Tejun Heo 已提交
368 369
}

370 371 372
/**
 * blkg_for_each_descendant_pre - pre-order walk of a blkg's descendants
 * @d_blkg: loop cursor pointing to the current descendant
373
 * @pos_css: used for iteration
374 375 376 377 378
 * @p_blkg: target blkg to walk descendants of
 *
 * Walk @c_blkg through the descendants of @p_blkg.  Must be used with RCU
 * read locked.  If called under either blkcg or queue lock, the iteration
 * is guaranteed to include all and only online blkgs.  The caller may
379
 * update @pos_css by calling css_rightmost_descendant() to skip subtree.
380
 * @p_blkg is included in the iteration and the first node to be visited.
381
 */
382 383 384
#define blkg_for_each_descendant_pre(d_blkg, pos_css, p_blkg)		\
	css_for_each_descendant_pre((pos_css), &(p_blkg)->blkcg->css)	\
		if (((d_blkg) = __blkg_lookup(css_to_blkcg(pos_css),	\
385 386
					      (p_blkg)->q, false)))

387 388 389
/**
 * blkg_for_each_descendant_post - post-order walk of a blkg's descendants
 * @d_blkg: loop cursor pointing to the current descendant
390
 * @pos_css: used for iteration
391 392 393
 * @p_blkg: target blkg to walk descendants of
 *
 * Similar to blkg_for_each_descendant_pre() but performs post-order
394 395
 * traversal instead.  Synchronization rules are the same.  @p_blkg is
 * included in the iteration and the last node to be visited.
396
 */
397 398 399
#define blkg_for_each_descendant_post(d_blkg, pos_css, p_blkg)		\
	css_for_each_descendant_post((pos_css), &(p_blkg)->blkcg->css)	\
		if (((d_blkg) = __blkg_lookup(css_to_blkcg(pos_css),	\
400 401
					      (p_blkg)->q, false)))

402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430
/**
 * blk_get_rl - get request_list to use
 * @q: request_queue of interest
 * @bio: bio which will be attached to the allocated request (may be %NULL)
 *
 * The caller wants to allocate a request from @q to use for @bio.  Find
 * the request_list to use and obtain a reference on it.  Should be called
 * under queue_lock.  This function is guaranteed to return non-%NULL
 * request_list.
 */
static inline struct request_list *blk_get_rl(struct request_queue *q,
					      struct bio *bio)
{
	struct blkcg *blkcg;
	struct blkcg_gq *blkg;

	rcu_read_lock();

	blkcg = bio_blkcg(bio);

	/* bypass blkg lookup and use @q->root_rl directly for root */
	if (blkcg == &blkcg_root)
		goto root_rl;

	/*
	 * Try to use blkg->rl.  blkg lookup may fail under memory pressure
	 * or if either the blkcg or queue is going away.  Fall back to
	 * root_rl in such cases.
	 */
431 432
	blkg = blkg_lookup(blkcg, q);
	if (unlikely(!blkg))
433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451
		goto root_rl;

	blkg_get(blkg);
	rcu_read_unlock();
	return &blkg->rl;
root_rl:
	rcu_read_unlock();
	return &q->root_rl;
}

/**
 * blk_put_rl - put request_list
 * @rl: request_list to put
 *
 * Put the reference acquired by blk_get_rl().  Should be called under
 * queue_lock.
 */
static inline void blk_put_rl(struct request_list *rl)
{
452
	if (rl->blkg->blkcg != &blkcg_root)
453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489
		blkg_put(rl->blkg);
}

/**
 * blk_rq_set_rl - associate a request with a request_list
 * @rq: request of interest
 * @rl: target request_list
 *
 * Associate @rq with @rl so that accounting and freeing can know the
 * request_list @rq came from.
 */
static inline void blk_rq_set_rl(struct request *rq, struct request_list *rl)
{
	rq->rl = rl;
}

/**
 * blk_rq_rl - return the request_list a request came from
 * @rq: request of interest
 *
 * Return the request_list @rq is allocated from.
 */
static inline struct request_list *blk_rq_rl(struct request *rq)
{
	return rq->rl;
}

struct request_list *__blk_queue_next_rl(struct request_list *rl,
					 struct request_queue *q);
/**
 * blk_queue_for_each_rl - iterate through all request_lists of a request_queue
 *
 * Should be used under queue_lock.
 */
#define blk_queue_for_each_rl(rl, q)	\
	for ((rl) = &(q)->root_rl; (rl); (rl) = __blk_queue_next_rl((rl), (q)))

T
Tejun Heo 已提交
490
static inline int blkg_stat_init(struct blkg_stat *stat, gfp_t gfp)
491
{
T
Tejun Heo 已提交
492 493 494 495 496 497
	int ret;

	ret = percpu_counter_init(&stat->cpu_cnt, 0, gfp);
	if (ret)
		return ret;

498
	atomic64_set(&stat->aux_cnt, 0);
T
Tejun Heo 已提交
499 500 501 502 503 504
	return 0;
}

static inline void blkg_stat_exit(struct blkg_stat *stat)
{
	percpu_counter_destroy(&stat->cpu_cnt);
505 506
}

507 508 509 510 511
/**
 * blkg_stat_add - add a value to a blkg_stat
 * @stat: target blkg_stat
 * @val: value to add
 *
T
Tejun Heo 已提交
512 513
 * Add @val to @stat.  The caller must ensure that IRQ on the same CPU
 * don't re-enter this function for the same counter.
514 515 516
 */
static inline void blkg_stat_add(struct blkg_stat *stat, uint64_t val)
{
517
	percpu_counter_add_batch(&stat->cpu_cnt, val, BLKG_STAT_CPU_BATCH);
518 519 520 521 522 523 524 525
}

/**
 * blkg_stat_read - read the current value of a blkg_stat
 * @stat: blkg_stat to read
 */
static inline uint64_t blkg_stat_read(struct blkg_stat *stat)
{
T
Tejun Heo 已提交
526
	return percpu_counter_sum_positive(&stat->cpu_cnt);
527 528 529 530 531 532 533 534
}

/**
 * blkg_stat_reset - reset a blkg_stat
 * @stat: blkg_stat to reset
 */
static inline void blkg_stat_reset(struct blkg_stat *stat)
{
T
Tejun Heo 已提交
535
	percpu_counter_set(&stat->cpu_cnt, 0);
536
	atomic64_set(&stat->aux_cnt, 0);
537 538
}

539
/**
540
 * blkg_stat_add_aux - add a blkg_stat into another's aux count
541 542 543
 * @to: the destination blkg_stat
 * @from: the source
 *
544
 * Add @from's count including the aux one to @to's aux count.
545
 */
546 547
static inline void blkg_stat_add_aux(struct blkg_stat *to,
				     struct blkg_stat *from)
548
{
549 550
	atomic64_add(blkg_stat_read(from) + atomic64_read(&from->aux_cnt),
		     &to->aux_cnt);
551 552
}

T
Tejun Heo 已提交
553
static inline int blkg_rwstat_init(struct blkg_rwstat *rwstat, gfp_t gfp)
554
{
T
Tejun Heo 已提交
555 556 557 558 559 560 561 562 563 564 565 566 567
	int i, ret;

	for (i = 0; i < BLKG_RWSTAT_NR; i++) {
		ret = percpu_counter_init(&rwstat->cpu_cnt[i], 0, gfp);
		if (ret) {
			while (--i >= 0)
				percpu_counter_destroy(&rwstat->cpu_cnt[i]);
			return ret;
		}
		atomic64_set(&rwstat->aux_cnt[i], 0);
	}
	return 0;
}
568

T
Tejun Heo 已提交
569 570 571
static inline void blkg_rwstat_exit(struct blkg_rwstat *rwstat)
{
	int i;
572 573

	for (i = 0; i < BLKG_RWSTAT_NR; i++)
T
Tejun Heo 已提交
574
		percpu_counter_destroy(&rwstat->cpu_cnt[i]);
575 576
}

577 578 579
/**
 * blkg_rwstat_add - add a value to a blkg_rwstat
 * @rwstat: target blkg_rwstat
580
 * @op: REQ_OP and flags
581 582 583 584 585 586
 * @val: value to add
 *
 * Add @val to @rwstat.  The counters are chosen according to @rw.  The
 * caller is responsible for synchronizing calls to this function.
 */
static inline void blkg_rwstat_add(struct blkg_rwstat *rwstat,
587
				   unsigned int op, uint64_t val)
588
{
T
Tejun Heo 已提交
589
	struct percpu_counter *cnt;
590

591
	if (op_is_write(op))
T
Tejun Heo 已提交
592
		cnt = &rwstat->cpu_cnt[BLKG_RWSTAT_WRITE];
593
	else
T
Tejun Heo 已提交
594 595
		cnt = &rwstat->cpu_cnt[BLKG_RWSTAT_READ];

596
	percpu_counter_add_batch(cnt, val, BLKG_STAT_CPU_BATCH);
T
Tejun Heo 已提交
597

598
	if (op_is_sync(op))
T
Tejun Heo 已提交
599
		cnt = &rwstat->cpu_cnt[BLKG_RWSTAT_SYNC];
600
	else
T
Tejun Heo 已提交
601
		cnt = &rwstat->cpu_cnt[BLKG_RWSTAT_ASYNC];
602

603
	percpu_counter_add_batch(cnt, val, BLKG_STAT_CPU_BATCH);
604 605 606 607 608 609
}

/**
 * blkg_rwstat_read - read the current values of a blkg_rwstat
 * @rwstat: blkg_rwstat to read
 *
T
Tejun Heo 已提交
610
 * Read the current snapshot of @rwstat and return it in the aux counts.
611
 */
612
static inline struct blkg_rwstat blkg_rwstat_read(struct blkg_rwstat *rwstat)
613
{
T
Tejun Heo 已提交
614 615
	struct blkg_rwstat result;
	int i;
616

T
Tejun Heo 已提交
617 618 619 620
	for (i = 0; i < BLKG_RWSTAT_NR; i++)
		atomic64_set(&result.aux_cnt[i],
			     percpu_counter_sum_positive(&rwstat->cpu_cnt[i]));
	return result;
621 622 623
}

/**
624
 * blkg_rwstat_total - read the total count of a blkg_rwstat
625 626 627 628 629 630
 * @rwstat: blkg_rwstat to read
 *
 * Return the total count of @rwstat regardless of the IO direction.  This
 * function can be called without synchronization and takes care of u64
 * atomicity.
 */
631
static inline uint64_t blkg_rwstat_total(struct blkg_rwstat *rwstat)
632 633 634
{
	struct blkg_rwstat tmp = blkg_rwstat_read(rwstat);

T
Tejun Heo 已提交
635 636
	return atomic64_read(&tmp.aux_cnt[BLKG_RWSTAT_READ]) +
		atomic64_read(&tmp.aux_cnt[BLKG_RWSTAT_WRITE]);
637 638 639 640 641 642 643 644
}

/**
 * blkg_rwstat_reset - reset a blkg_rwstat
 * @rwstat: blkg_rwstat to reset
 */
static inline void blkg_rwstat_reset(struct blkg_rwstat *rwstat)
{
645 646
	int i;

T
Tejun Heo 已提交
647 648
	for (i = 0; i < BLKG_RWSTAT_NR; i++) {
		percpu_counter_set(&rwstat->cpu_cnt[i], 0);
649
		atomic64_set(&rwstat->aux_cnt[i], 0);
T
Tejun Heo 已提交
650
	}
651 652
}

653
/**
654
 * blkg_rwstat_add_aux - add a blkg_rwstat into another's aux count
655 656 657
 * @to: the destination blkg_rwstat
 * @from: the source
 *
658
 * Add @from's count including the aux one to @to's aux count.
659
 */
660 661
static inline void blkg_rwstat_add_aux(struct blkg_rwstat *to,
				       struct blkg_rwstat *from)
662
{
663
	u64 sum[BLKG_RWSTAT_NR];
664 665 666
	int i;

	for (i = 0; i < BLKG_RWSTAT_NR; i++)
667 668 669 670
		sum[i] = percpu_counter_sum_positive(&from->cpu_cnt[i]);

	for (i = 0; i < BLKG_RWSTAT_NR; i++)
		atomic64_add(sum[i] + atomic64_read(&from->aux_cnt[i]),
671
			     &to->aux_cnt[i]);
672 673
}

674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691
#ifdef CONFIG_BLK_DEV_THROTTLING
extern bool blk_throtl_bio(struct request_queue *q, struct blkcg_gq *blkg,
			   struct bio *bio);
#else
static inline bool blk_throtl_bio(struct request_queue *q, struct blkcg_gq *blkg,
				  struct bio *bio) { return false; }
#endif

static inline bool blkcg_bio_issue_check(struct request_queue *q,
					 struct bio *bio)
{
	struct blkcg *blkcg;
	struct blkcg_gq *blkg;
	bool throtl = false;

	rcu_read_lock();
	blkcg = bio_blkcg(bio);

692 693 694
	/* associate blkcg if bio hasn't attached one */
	bio_associate_blkcg(bio, &blkcg->css);

695 696 697 698 699 700 701 702 703 704 705
	blkg = blkg_lookup(blkcg, q);
	if (unlikely(!blkg)) {
		spin_lock_irq(q->queue_lock);
		blkg = blkg_lookup_create(blkcg, q);
		if (IS_ERR(blkg))
			blkg = NULL;
		spin_unlock_irq(q->queue_lock);
	}

	throtl = blk_throtl_bio(q, blkg, bio);

706 707
	if (!throtl) {
		blkg = blkg ?: q->root_blkg;
708
		blkg_rwstat_add(&blkg->stat_bytes, bio->bi_opf,
709
				bio->bi_iter.bi_size);
710
		blkg_rwstat_add(&blkg->stat_ios, bio->bi_opf, 1);
711 712
	}

713 714 715 716
	rcu_read_unlock();
	return !throtl;
}

717 718
#else	/* CONFIG_BLK_CGROUP */

719 720
struct blkcg {
};
721

722 723 724
struct blkg_policy_data {
};

725 726 727
struct blkcg_policy_data {
};

T
Tejun Heo 已提交
728
struct blkcg_gq {
729 730
};

T
Tejun Heo 已提交
731
struct blkcg_policy {
732 733
};

T
Tejun Heo 已提交
734 735
#define blkcg_root_css	((struct cgroup_subsys_state *)ERR_PTR(-EINVAL))

736 737
#ifdef CONFIG_BLOCK

T
Tejun Heo 已提交
738
static inline struct blkcg_gq *blkg_lookup(struct blkcg *blkcg, void *key) { return NULL; }
739 740 741
static inline int blkcg_init_queue(struct request_queue *q) { return 0; }
static inline void blkcg_drain_queue(struct request_queue *q) { }
static inline void blkcg_exit_queue(struct request_queue *q) { }
742
static inline int blkcg_policy_register(struct blkcg_policy *pol) { return 0; }
T
Tejun Heo 已提交
743
static inline void blkcg_policy_unregister(struct blkcg_policy *pol) { }
744
static inline int blkcg_activate_policy(struct request_queue *q,
T
Tejun Heo 已提交
745
					const struct blkcg_policy *pol) { return 0; }
746
static inline void blkcg_deactivate_policy(struct request_queue *q,
T
Tejun Heo 已提交
747 748
					   const struct blkcg_policy *pol) { }

749
static inline struct blkcg *bio_blkcg(struct bio *bio) { return NULL; }
750

751 752 753
static inline struct blkg_policy_data *blkg_to_pd(struct blkcg_gq *blkg,
						  struct blkcg_policy *pol) { return NULL; }
static inline struct blkcg_gq *pd_to_blkg(struct blkg_policy_data *pd) { return NULL; }
T
Tejun Heo 已提交
754 755 756
static inline char *blkg_path(struct blkcg_gq *blkg) { return NULL; }
static inline void blkg_get(struct blkcg_gq *blkg) { }
static inline void blkg_put(struct blkcg_gq *blkg) { }
757

758 759 760 761 762 763
static inline struct request_list *blk_get_rl(struct request_queue *q,
					      struct bio *bio) { return &q->root_rl; }
static inline void blk_put_rl(struct request_list *rl) { }
static inline void blk_rq_set_rl(struct request *rq, struct request_list *rl) { }
static inline struct request_list *blk_rq_rl(struct request *rq) { return &rq->q->root_rl; }

764 765 766
static inline bool blkcg_bio_issue_check(struct request_queue *q,
					 struct bio *bio) { return true; }

767 768 769
#define blk_queue_for_each_rl(rl, q)	\
	for ((rl) = &(q)->root_rl; (rl); (rl) = NULL)

770
#endif	/* CONFIG_BLOCK */
771 772
#endif	/* CONFIG_BLK_CGROUP */
#endif	/* _BLK_CGROUP_H */