blk-cgroup.h 21.9 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
#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
17
#include <linux/percpu_counter.h>
18
#include <linux/seq_file.h>
19
#include <linux/radix-tree.h>
20
#include <linux/blkdev.h>
21
#include <linux/atomic.h>
22

T
Tejun Heo 已提交
23 24 25
/* percpu_counter batch for blkg_[rw]stats, per-cpu drift doesn't matter */
#define BLKG_STAT_CPU_BATCH	(INT_MAX / 2)

26 27 28
/* Max limits for throttle policy */
#define THROTL_IOPS_MAX		UINT_MAX

T
Tejun Heo 已提交
29 30
#ifdef CONFIG_BLK_CGROUP

31 32 33 34 35 36 37 38
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,
39 40
};

41 42
struct blkcg_gq;

T
Tejun Heo 已提交
43
struct blkcg {
44 45
	struct cgroup_subsys_state	css;
	spinlock_t			lock;
46 47

	struct radix_tree_root		blkg_tree;
48
	struct blkcg_gq	__rcu		*blkg_hint;
49
	struct hlist_head		blkg_list;
T
Tejun Heo 已提交
50

51
	struct blkcg_policy_data	*cpd[BLKCG_MAX_POLS];
52

T
Tejun Heo 已提交
53
	struct list_head		all_blkcgs_node;
54 55 56
#ifdef CONFIG_CGROUP_WRITEBACK
	struct list_head		cgwb_list;
#endif
57 58
};

59 60
/*
 * blkg_[rw]stat->aux_cnt is excluded for local stats but included for
T
Tejun Heo 已提交
61 62
 * recursive.  Used to carry stats of dead children, and, for blkg_rwstat,
 * to carry result values from read and sum operations.
63
 */
64
struct blkg_stat {
T
Tejun Heo 已提交
65
	struct percpu_counter		cpu_cnt;
66
	atomic64_t			aux_cnt;
67 68 69
};

struct blkg_rwstat {
T
Tejun Heo 已提交
70
	struct percpu_counter		cpu_cnt[BLKG_RWSTAT_NR];
71
	atomic64_t			aux_cnt[BLKG_RWSTAT_NR];
72 73
};

74 75 76 77 78
/*
 * 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.
 *
79 80 81 82 83
 * 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.
84
 */
85
struct blkg_policy_data {
T
Tejun Heo 已提交
86
	/* the blkg and policy id this per-policy data belongs to */
T
Tejun Heo 已提交
87
	struct blkcg_gq			*blkg;
T
Tejun Heo 已提交
88
	int				plid;
89 90
};

91
/*
92 93 94 95 96
 * 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.
97 98
 */
struct blkcg_policy_data {
99 100
	/* the blkcg and policy id this per-policy data belongs to */
	struct blkcg			*blkcg;
101 102 103
	int				plid;
};

T
Tejun Heo 已提交
104 105
/* association between a blk cgroup and a request queue */
struct blkcg_gq {
T
Tejun Heo 已提交
106
	/* Pointer to the associated request_queue */
107 108 109
	struct request_queue		*q;
	struct list_head		q_node;
	struct hlist_node		blkcg_node;
T
Tejun Heo 已提交
110
	struct blkcg			*blkcg;
T
Tejun Heo 已提交
111

112 113 114 115 116 117
	/*
	 * 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
118 119 120
	/* all non-root blkcg_gq's are guaranteed to have access to parent */
	struct blkcg_gq			*parent;

121 122
	/* request allocation list for this blkcg-q pair */
	struct request_list		rl;
T
Tejun Heo 已提交
123

T
Tejun Heo 已提交
124
	/* reference count */
125
	atomic_t			refcnt;
126

127 128 129
	/* is this blkg online? protected by both blkcg and q locks */
	bool				online;

130 131 132
	struct blkg_rwstat		stat_bytes;
	struct blkg_rwstat		stat_ios;

133
	struct blkg_policy_data		*pd[BLKCG_MAX_POLS];
T
Tejun Heo 已提交
134

135
	struct rcu_head			rcu_head;
136 137
};

138
typedef struct blkcg_policy_data *(blkcg_pol_alloc_cpd_fn)(gfp_t gfp);
139
typedef void (blkcg_pol_init_cpd_fn)(struct blkcg_policy_data *cpd);
140
typedef void (blkcg_pol_free_cpd_fn)(struct blkcg_policy_data *cpd);
141
typedef void (blkcg_pol_bind_cpd_fn)(struct blkcg_policy_data *cpd);
142
typedef struct blkg_policy_data *(blkcg_pol_alloc_pd_fn)(gfp_t gfp, int node);
143 144 145
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);
146
typedef void (blkcg_pol_free_pd_fn)(struct blkg_policy_data *pd);
147
typedef void (blkcg_pol_reset_pd_stats_fn)(struct blkg_policy_data *pd);
148

T
Tejun Heo 已提交
149
struct blkcg_policy {
150 151
	int				plid;
	/* cgroup files for the policy */
152
	struct cftype			*dfl_cftypes;
153
	struct cftype			*legacy_cftypes;
154 155

	/* operations */
156
	blkcg_pol_alloc_cpd_fn		*cpd_alloc_fn;
157
	blkcg_pol_init_cpd_fn		*cpd_init_fn;
158
	blkcg_pol_free_cpd_fn		*cpd_free_fn;
159
	blkcg_pol_bind_cpd_fn		*cpd_bind_fn;
160

161
	blkcg_pol_alloc_pd_fn		*pd_alloc_fn;
162
	blkcg_pol_init_pd_fn		*pd_init_fn;
163 164
	blkcg_pol_online_pd_fn		*pd_online_fn;
	blkcg_pol_offline_pd_fn		*pd_offline_fn;
165
	blkcg_pol_free_pd_fn		*pd_free_fn;
166
	blkcg_pol_reset_pd_stats_fn	*pd_reset_stats_fn;
167 168
};

T
Tejun Heo 已提交
169
extern struct blkcg blkcg_root;
T
Tejun Heo 已提交
170
extern struct cgroup_subsys_state * const blkcg_root_css;
171

T
Tejun Heo 已提交
172 173
struct blkcg_gq *blkg_lookup_slowpath(struct blkcg *blkcg,
				      struct request_queue *q, bool update_hint);
T
Tejun Heo 已提交
174
struct blkcg_gq *blkg_lookup_create(struct blkcg *blkcg,
175 176
				    struct request_queue *q, gfp_t gfp,
				    const struct blkcg_policy *pol);
177 178 179
int blkcg_init_queue(struct request_queue *q);
void blkcg_drain_queue(struct request_queue *q);
void blkcg_exit_queue(struct request_queue *q);
180

181
/* Blkio controller policy registration */
182
int blkcg_policy_register(struct blkcg_policy *pol);
T
Tejun Heo 已提交
183
void blkcg_policy_unregister(struct blkcg_policy *pol);
184
int blkcg_activate_policy(struct request_queue *q,
T
Tejun Heo 已提交
185
			  const struct blkcg_policy *pol);
186
void blkcg_deactivate_policy(struct request_queue *q,
T
Tejun Heo 已提交
187
			     const struct blkcg_policy *pol);
188

189
const char *blkg_dev_name(struct blkcg_gq *blkg);
T
Tejun Heo 已提交
190
void blkcg_print_blkgs(struct seq_file *sf, struct blkcg *blkcg,
191 192
		       u64 (*prfill)(struct seq_file *,
				     struct blkg_policy_data *, int),
T
Tejun Heo 已提交
193
		       const struct blkcg_policy *pol, int data,
194
		       bool show_total);
195 196
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,
197
			 const struct blkg_rwstat *rwstat);
198 199 200
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);
201 202 203 204
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);
205

206 207 208 209
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);
210

211
struct blkg_conf_ctx {
212
	struct gendisk			*disk;
T
Tejun Heo 已提交
213
	struct blkcg_gq			*blkg;
214
	char				*body;
215 216
};

T
Tejun Heo 已提交
217
int blkg_conf_prep(struct blkcg *blkcg, const struct blkcg_policy *pol,
218
		   char *input, struct blkg_conf_ctx *ctx);
219 220 221
void blkg_conf_finish(struct blkg_conf_ctx *ctx);


222 223 224 225 226
static inline struct blkcg *css_to_blkcg(struct cgroup_subsys_state *css)
{
	return css ? container_of(css, struct blkcg, css) : NULL;
}

227 228
static inline struct blkcg *task_blkcg(struct task_struct *tsk)
{
229
	return css_to_blkcg(task_css(tsk, io_cgrp_id));
230 231 232 233 234
}

static inline struct blkcg *bio_blkcg(struct bio *bio)
{
	if (bio && bio->bi_css)
235
		return css_to_blkcg(bio->bi_css);
236 237 238
	return task_blkcg(current);
}

239 240 241
static inline struct cgroup_subsys_state *
task_get_blkcg_css(struct task_struct *task)
{
242
	return task_get_css(task, io_cgrp_id);
243 244
}

T
Tejun Heo 已提交
245 246 247 248 249 250 251 252
/**
 * 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
253
	return css_to_blkcg(blkcg->css.parent);
T
Tejun Heo 已提交
254 255
}

T
Tejun Heo 已提交
256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272
/**
 * __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;

273 274 275
	if (blkcg == &blkcg_root)
		return q->root_blkg;

T
Tejun Heo 已提交
276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301
	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);
}

302 303 304 305 306 307 308
/**
 * 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.
 */
309 310
static inline struct blkg_policy_data *blkg_to_pd(struct blkcg_gq *blkg,
						  struct blkcg_policy *pol)
311
{
312
	return blkg ? blkg->pd[pol->plid] : NULL;
313 314
}

315 316 317
static inline struct blkcg_policy_data *blkcg_to_cpd(struct blkcg *blkcg,
						     struct blkcg_policy *pol)
{
318
	return blkcg ? blkcg->cpd[pol->plid] : NULL;
319 320
}

321 322
/**
 * pdata_to_blkg - get blkg associated with policy private data
323
 * @pd: policy private data of interest
324
 *
325
 * @pd is policy private data.  Determine the blkg it's associated with.
326
 */
327
static inline struct blkcg_gq *pd_to_blkg(struct blkg_policy_data *pd)
328
{
329
	return pd ? pd->blkg : NULL;
330 331
}

332 333 334 335 336
static inline struct blkcg *cpd_to_blkcg(struct blkcg_policy_data *cpd)
{
	return cpd ? cpd->blkcg : NULL;
}

T
Tejun Heo 已提交
337 338 339 340 341 342 343 344
/**
 * 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
345
static inline int blkg_path(struct blkcg_gq *blkg, char *buf, int buflen)
346
{
347
	return cgroup_path(blkg->blkcg->css.cgroup, buf, buflen);
348 349
}

T
Tejun Heo 已提交
350 351 352 353
/**
 * blkg_get - get a blkg reference
 * @blkg: blkg to get
 *
354
 * The caller should be holding an existing reference.
T
Tejun Heo 已提交
355
 */
T
Tejun Heo 已提交
356
static inline void blkg_get(struct blkcg_gq *blkg)
T
Tejun Heo 已提交
357
{
358 359
	WARN_ON_ONCE(atomic_read(&blkg->refcnt) <= 0);
	atomic_inc(&blkg->refcnt);
T
Tejun Heo 已提交
360 361
}

362
void __blkg_release_rcu(struct rcu_head *rcu);
T
Tejun Heo 已提交
363 364 365 366 367

/**
 * blkg_put - put a blkg reference
 * @blkg: blkg to put
 */
T
Tejun Heo 已提交
368
static inline void blkg_put(struct blkcg_gq *blkg)
T
Tejun Heo 已提交
369
{
370 371
	WARN_ON_ONCE(atomic_read(&blkg->refcnt) <= 0);
	if (atomic_dec_and_test(&blkg->refcnt))
372
		call_rcu(&blkg->rcu_head, __blkg_release_rcu);
T
Tejun Heo 已提交
373 374
}

375 376 377
/**
 * blkg_for_each_descendant_pre - pre-order walk of a blkg's descendants
 * @d_blkg: loop cursor pointing to the current descendant
378
 * @pos_css: used for iteration
379 380 381 382 383
 * @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
384
 * update @pos_css by calling css_rightmost_descendant() to skip subtree.
385
 * @p_blkg is included in the iteration and the first node to be visited.
386
 */
387 388 389
#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),	\
390 391
					      (p_blkg)->q, false)))

392 393 394
/**
 * blkg_for_each_descendant_post - post-order walk of a blkg's descendants
 * @d_blkg: loop cursor pointing to the current descendant
395
 * @pos_css: used for iteration
396 397 398
 * @p_blkg: target blkg to walk descendants of
 *
 * Similar to blkg_for_each_descendant_pre() but performs post-order
399 400
 * traversal instead.  Synchronization rules are the same.  @p_blkg is
 * included in the iteration and the last node to be visited.
401
 */
402 403 404
#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),	\
405 406
					      (p_blkg)->q, false)))

407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435
/**
 * 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.
	 */
436 437
	blkg = blkg_lookup(blkcg, q);
	if (unlikely(!blkg))
438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456
		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)
{
457
	if (rl->blkg->blkcg != &blkcg_root)
458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494
		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
495
static inline int blkg_stat_init(struct blkg_stat *stat, gfp_t gfp)
496
{
T
Tejun Heo 已提交
497 498 499 500 501 502
	int ret;

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

503
	atomic64_set(&stat->aux_cnt, 0);
T
Tejun Heo 已提交
504 505 506 507 508 509
	return 0;
}

static inline void blkg_stat_exit(struct blkg_stat *stat)
{
	percpu_counter_destroy(&stat->cpu_cnt);
510 511
}

512 513 514 515 516
/**
 * blkg_stat_add - add a value to a blkg_stat
 * @stat: target blkg_stat
 * @val: value to add
 *
T
Tejun Heo 已提交
517 518
 * Add @val to @stat.  The caller must ensure that IRQ on the same CPU
 * don't re-enter this function for the same counter.
519 520 521
 */
static inline void blkg_stat_add(struct blkg_stat *stat, uint64_t val)
{
T
Tejun Heo 已提交
522
	__percpu_counter_add(&stat->cpu_cnt, val, BLKG_STAT_CPU_BATCH);
523 524 525 526 527 528 529 530
}

/**
 * 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
531
	return percpu_counter_sum_positive(&stat->cpu_cnt);
532 533 534 535 536 537 538 539
}

/**
 * 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
540
	percpu_counter_set(&stat->cpu_cnt, 0);
541
	atomic64_set(&stat->aux_cnt, 0);
542 543
}

544
/**
545
 * blkg_stat_add_aux - add a blkg_stat into another's aux count
546 547 548
 * @to: the destination blkg_stat
 * @from: the source
 *
549
 * Add @from's count including the aux one to @to's aux count.
550
 */
551 552
static inline void blkg_stat_add_aux(struct blkg_stat *to,
				     struct blkg_stat *from)
553
{
554 555
	atomic64_add(blkg_stat_read(from) + atomic64_read(&from->aux_cnt),
		     &to->aux_cnt);
556 557
}

T
Tejun Heo 已提交
558
static inline int blkg_rwstat_init(struct blkg_rwstat *rwstat, gfp_t gfp)
559
{
T
Tejun Heo 已提交
560 561 562 563 564 565 566 567 568 569 570 571 572
	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;
}
573

T
Tejun Heo 已提交
574 575 576
static inline void blkg_rwstat_exit(struct blkg_rwstat *rwstat)
{
	int i;
577 578

	for (i = 0; i < BLKG_RWSTAT_NR; i++)
T
Tejun Heo 已提交
579
		percpu_counter_destroy(&rwstat->cpu_cnt[i]);
580 581
}

582 583 584
/**
 * blkg_rwstat_add - add a value to a blkg_rwstat
 * @rwstat: target blkg_rwstat
585
 * @op: REQ_OP and flags
586 587 588 589 590 591
 * @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,
592
				   unsigned int op, uint64_t val)
593
{
T
Tejun Heo 已提交
594
	struct percpu_counter *cnt;
595

596
	if (op_is_write(op))
T
Tejun Heo 已提交
597
		cnt = &rwstat->cpu_cnt[BLKG_RWSTAT_WRITE];
598
	else
T
Tejun Heo 已提交
599 600 601 602
		cnt = &rwstat->cpu_cnt[BLKG_RWSTAT_READ];

	__percpu_counter_add(cnt, val, BLKG_STAT_CPU_BATCH);

603
	if (op_is_sync(op))
T
Tejun Heo 已提交
604
		cnt = &rwstat->cpu_cnt[BLKG_RWSTAT_SYNC];
605
	else
T
Tejun Heo 已提交
606
		cnt = &rwstat->cpu_cnt[BLKG_RWSTAT_ASYNC];
607

T
Tejun Heo 已提交
608
	__percpu_counter_add(cnt, val, BLKG_STAT_CPU_BATCH);
609 610 611 612 613 614
}

/**
 * blkg_rwstat_read - read the current values of a blkg_rwstat
 * @rwstat: blkg_rwstat to read
 *
T
Tejun Heo 已提交
615
 * Read the current snapshot of @rwstat and return it in the aux counts.
616
 */
617
static inline struct blkg_rwstat blkg_rwstat_read(struct blkg_rwstat *rwstat)
618
{
T
Tejun Heo 已提交
619 620
	struct blkg_rwstat result;
	int i;
621

T
Tejun Heo 已提交
622 623 624 625
	for (i = 0; i < BLKG_RWSTAT_NR; i++)
		atomic64_set(&result.aux_cnt[i],
			     percpu_counter_sum_positive(&rwstat->cpu_cnt[i]));
	return result;
626 627 628
}

/**
629
 * blkg_rwstat_total - read the total count of a blkg_rwstat
630 631 632 633 634 635
 * @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.
 */
636
static inline uint64_t blkg_rwstat_total(struct blkg_rwstat *rwstat)
637 638 639
{
	struct blkg_rwstat tmp = blkg_rwstat_read(rwstat);

T
Tejun Heo 已提交
640 641
	return atomic64_read(&tmp.aux_cnt[BLKG_RWSTAT_READ]) +
		atomic64_read(&tmp.aux_cnt[BLKG_RWSTAT_WRITE]);
642 643 644 645 646 647 648 649
}

/**
 * blkg_rwstat_reset - reset a blkg_rwstat
 * @rwstat: blkg_rwstat to reset
 */
static inline void blkg_rwstat_reset(struct blkg_rwstat *rwstat)
{
650 651
	int i;

T
Tejun Heo 已提交
652 653
	for (i = 0; i < BLKG_RWSTAT_NR; i++) {
		percpu_counter_set(&rwstat->cpu_cnt[i], 0);
654
		atomic64_set(&rwstat->aux_cnt[i], 0);
T
Tejun Heo 已提交
655
	}
656 657
}

658
/**
659
 * blkg_rwstat_add_aux - add a blkg_rwstat into another's aux count
660 661 662
 * @to: the destination blkg_rwstat
 * @from: the source
 *
663
 * Add @from's count including the aux one to @to's aux count.
664
 */
665 666
static inline void blkg_rwstat_add_aux(struct blkg_rwstat *to,
				       struct blkg_rwstat *from)
667 668 669 670 671
{
	struct blkg_rwstat v = blkg_rwstat_read(from);
	int i;

	for (i = 0; i < BLKG_RWSTAT_NR; i++)
T
Tejun Heo 已提交
672 673
		atomic64_add(atomic64_read(&v.aux_cnt[i]) +
			     atomic64_read(&from->aux_cnt[i]),
674
			     &to->aux_cnt[i]);
675 676
}

677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697
#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);

	blkg = blkg_lookup(blkcg, q);
	if (unlikely(!blkg)) {
		spin_lock_irq(q->queue_lock);
698 699
		blkg = blkg_lookup_create(blkcg, q, GFP_NOWAIT | __GFP_NOWARN,
					  NULL);
700 701 702 703 704 705 706
		if (IS_ERR(blkg))
			blkg = NULL;
		spin_unlock_irq(q->queue_lock);
	}

	throtl = blk_throtl_bio(q, blkg, bio);

707 708
	if (!throtl) {
		blkg = blkg ?: q->root_blkg;
709
		blkg_rwstat_add(&blkg->stat_bytes, bio->bi_opf,
710
				bio->bi_iter.bi_size);
711
		blkg_rwstat_add(&blkg->stat_ios, bio->bi_opf, 1);
712 713
	}

714 715 716 717
	rcu_read_unlock();
	return !throtl;
}

718 719
#else	/* CONFIG_BLK_CGROUP */

720 721
struct blkcg {
};
722

723 724 725
struct blkg_policy_data {
};

726 727 728
struct blkcg_policy_data {
};

T
Tejun Heo 已提交
729
struct blkcg_gq {
730 731
};

T
Tejun Heo 已提交
732
struct blkcg_policy {
733 734
};

T
Tejun Heo 已提交
735 736
#define blkcg_root_css	((struct cgroup_subsys_state *)ERR_PTR(-EINVAL))

737 738 739 740 741 742
static inline struct cgroup_subsys_state *
task_get_blkcg_css(struct task_struct *task)
{
	return NULL;
}

743 744
#ifdef CONFIG_BLOCK

T
Tejun Heo 已提交
745
static inline struct blkcg_gq *blkg_lookup(struct blkcg *blkcg, void *key) { return NULL; }
746 747 748
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) { }
749
static inline int blkcg_policy_register(struct blkcg_policy *pol) { return 0; }
T
Tejun Heo 已提交
750
static inline void blkcg_policy_unregister(struct blkcg_policy *pol) { }
751
static inline int blkcg_activate_policy(struct request_queue *q,
T
Tejun Heo 已提交
752
					const struct blkcg_policy *pol) { return 0; }
753
static inline void blkcg_deactivate_policy(struct request_queue *q,
T
Tejun Heo 已提交
754 755
					   const struct blkcg_policy *pol) { }

756
static inline struct blkcg *bio_blkcg(struct bio *bio) { return NULL; }
757

758 759 760
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
761 762 763
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) { }
764

765 766 767 768 769 770
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; }

771 772 773
static inline bool blkcg_bio_issue_check(struct request_queue *q,
					 struct bio *bio) { return true; }

774 775 776
#define blk_queue_for_each_rl(rl, q)	\
	for ((rl) = &(q)->root_rl; (rl); (rl) = NULL)

777
#endif	/* CONFIG_BLOCK */
778 779
#endif	/* CONFIG_BLK_CGROUP */
#endif	/* _BLK_CGROUP_H */