blk-cgroup.h 10.4 KB
Newer Older
1 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>

18 19
enum blkio_policy_id {
	BLKIO_POLICY_PROP = 0,		/* Proportional Bandwidth division */
20
	BLKIO_POLICY_THROTL,		/* Throttling */
21 22
};

23 24 25
/* Max limits for throttle policy */
#define THROTL_IOPS_MAX		UINT_MAX

B
Ben Blum 已提交
26 27 28 29 30 31 32
#if defined(CONFIG_BLK_CGROUP) || defined(CONFIG_BLK_CGROUP_MODULE)

#ifndef CONFIG_BLK_CGROUP
/* When blk-cgroup is a module, its subsys_id isn't a compile-time constant */
extern struct cgroup_subsys blkio_subsys;
#define blkio_subsys_id blkio_subsys.subsys_id
#endif
33

34 35 36 37 38 39 40 41 42 43 44
enum stat_type {
	/* Total time spent (in ns) between request dispatch to the driver and
	 * request completion for IOs doen by this cgroup. This may not be
	 * accurate when NCQ is turned on. */
	BLKIO_STAT_SERVICE_TIME = 0,
	/* Total bytes transferred */
	BLKIO_STAT_SERVICE_BYTES,
	/* Total IOs serviced, post merge */
	BLKIO_STAT_SERVICED,
	/* Total time spent waiting in scheduler queue in ns */
	BLKIO_STAT_WAIT_TIME,
D
Divyesh Shah 已提交
45 46
	/* Number of IOs merged */
	BLKIO_STAT_MERGED,
47 48
	/* Number of IOs queued up */
	BLKIO_STAT_QUEUED,
49 50 51 52
	/* All the single valued stats go below this */
	BLKIO_STAT_TIME,
	BLKIO_STAT_SECTORS,
#ifdef CONFIG_DEBUG_BLK_CGROUP
53
	BLKIO_STAT_AVG_QUEUE_SIZE,
54 55 56
	BLKIO_STAT_IDLE_TIME,
	BLKIO_STAT_EMPTY_TIME,
	BLKIO_STAT_GROUP_WAIT_TIME,
57 58 59 60 61 62 63 64 65 66
	BLKIO_STAT_DEQUEUE
#endif
};

enum stat_sub_type {
	BLKIO_STAT_READ = 0,
	BLKIO_STAT_WRITE,
	BLKIO_STAT_SYNC,
	BLKIO_STAT_ASYNC,
	BLKIO_STAT_TOTAL
67 68
};

69 70 71 72 73 74 75
/* blkg state flags */
enum blkg_state_flags {
	BLKG_waiting = 0,
	BLKG_idling,
	BLKG_empty,
};

76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94
/* cgroup files owned by proportional weight policy */
enum blkcg_file_name_prop {
	BLKIO_PROP_weight = 1,
	BLKIO_PROP_weight_device,
	BLKIO_PROP_io_service_bytes,
	BLKIO_PROP_io_serviced,
	BLKIO_PROP_time,
	BLKIO_PROP_sectors,
	BLKIO_PROP_io_service_time,
	BLKIO_PROP_io_wait_time,
	BLKIO_PROP_io_merged,
	BLKIO_PROP_io_queued,
	BLKIO_PROP_avg_queue_size,
	BLKIO_PROP_group_wait_time,
	BLKIO_PROP_idle_time,
	BLKIO_PROP_empty_time,
	BLKIO_PROP_dequeue,
};

95 96 97 98
/* cgroup files owned by throttle policy */
enum blkcg_file_name_throtl {
	BLKIO_THROTL_read_bps_device,
	BLKIO_THROTL_write_bps_device,
99 100
	BLKIO_THROTL_read_iops_device,
	BLKIO_THROTL_write_iops_device,
101 102 103 104
	BLKIO_THROTL_io_service_bytes,
	BLKIO_THROTL_io_serviced,
};

105 106 107 108 109
struct blkio_cgroup {
	struct cgroup_subsys_state css;
	unsigned int weight;
	spinlock_t lock;
	struct hlist_head blkg_list;
110
	struct list_head policy_list; /* list of blkio_policy_node */
111 112
};

113 114 115 116
struct blkio_group_stats {
	/* total disk time and nr sectors dispatched by this group */
	uint64_t time;
	uint64_t sectors;
117
	uint64_t stat_arr[BLKIO_STAT_QUEUED + 1][BLKIO_STAT_TOTAL];
118
#ifdef CONFIG_DEBUG_BLK_CGROUP
119 120 121 122
	/* Sum of number of IOs queued across all samples */
	uint64_t avg_queue_size_sum;
	/* Count of samples taken for average */
	uint64_t avg_queue_size_samples;
123 124
	/* How many times this group has been removed from service tree */
	unsigned long dequeue;
125 126 127 128 129 130 131 132 133 134 135 136 137 138 139

	/* Total time spent waiting for it to be assigned a timeslice. */
	uint64_t group_wait_time;
	uint64_t start_group_wait_time;

	/* Time spent idling for this blkio_group */
	uint64_t idle_time;
	uint64_t start_idle_time;
	/*
	 * Total time when we have requests queued and do not contain the
	 * current active queue.
	 */
	uint64_t empty_time;
	uint64_t start_empty_time;
	uint16_t flags;
140 141 142
#endif
};

143 144 145 146
struct blkio_group {
	/* An rcu protected unique identifier for the group */
	void *key;
	struct hlist_node blkcg_node;
147
	unsigned short blkcg_id;
V
Vivek Goyal 已提交
148 149
	/* Store cgroup path */
	char path[128];
150
	/* The device MKDEV(major, minor), this group has been created for */
151
	dev_t dev;
152 153
	/* policy which owns this blk group */
	enum blkio_policy_id plid;
154

155 156 157
	/* Need to serialize the stats in the case of reset/update */
	spinlock_t stats_lock;
	struct blkio_group_stats stats;
158 159
};

160 161 162
struct blkio_policy_node {
	struct list_head node;
	dev_t dev;
163 164 165 166
	/* This node belongs to max bw policy or porportional weight policy */
	enum blkio_policy_id plid;
	/* cgroup file to which this rule belongs to */
	int fileid;
167 168 169 170 171 172 173 174 175

	union {
		unsigned int weight;
		/*
		 * Rate read/write in terms of byptes per second
		 * Whether this rate represents read or write is determined
		 * by file type "fileid".
		 */
		u64 bps;
176
		unsigned int iops;
177
	} val;
178 179 180 181
};

extern unsigned int blkcg_get_weight(struct blkio_cgroup *blkcg,
				     dev_t dev);
182 183 184 185
extern uint64_t blkcg_get_read_bps(struct blkio_cgroup *blkcg,
				     dev_t dev);
extern uint64_t blkcg_get_write_bps(struct blkio_cgroup *blkcg,
				     dev_t dev);
186 187 188 189
extern unsigned int blkcg_get_read_iops(struct blkio_cgroup *blkcg,
				     dev_t dev);
extern unsigned int blkcg_get_write_iops(struct blkio_cgroup *blkcg,
				     dev_t dev);
190

191
typedef void (blkio_unlink_group_fn) (void *key, struct blkio_group *blkg);
192 193 194 195 196 197 198 199 200 201 202

typedef void (blkio_update_group_weight_fn) (void *key,
			struct blkio_group *blkg, unsigned int weight);
typedef void (blkio_update_group_read_bps_fn) (void * key,
			struct blkio_group *blkg, u64 read_bps);
typedef void (blkio_update_group_write_bps_fn) (void *key,
			struct blkio_group *blkg, u64 write_bps);
typedef void (blkio_update_group_read_iops_fn) (void *key,
			struct blkio_group *blkg, unsigned int read_iops);
typedef void (blkio_update_group_write_iops_fn) (void *key,
			struct blkio_group *blkg, unsigned int write_iops);
203 204 205 206

struct blkio_policy_ops {
	blkio_unlink_group_fn *blkio_unlink_group_fn;
	blkio_update_group_weight_fn *blkio_update_group_weight_fn;
207 208
	blkio_update_group_read_bps_fn *blkio_update_group_read_bps_fn;
	blkio_update_group_write_bps_fn *blkio_update_group_write_bps_fn;
209 210
	blkio_update_group_read_iops_fn *blkio_update_group_read_iops_fn;
	blkio_update_group_write_iops_fn *blkio_update_group_write_iops_fn;
211 212 213 214 215
};

struct blkio_policy_type {
	struct list_head list;
	struct blkio_policy_ops ops;
216
	enum blkio_policy_id plid;
217 218 219 220 221 222
};

/* Blkio controller policy registration */
extern void blkio_policy_register(struct blkio_policy_type *);
extern void blkio_policy_unregister(struct blkio_policy_type *);

223 224 225 226 227
static inline char *blkg_path(struct blkio_group *blkg)
{
	return blkg->path;
}

228 229 230 231 232
#else

struct blkio_group {
};

233 234 235 236 237 238
struct blkio_policy_type {
};

static inline void blkio_policy_register(struct blkio_policy_type *blkiop) { }
static inline void blkio_policy_unregister(struct blkio_policy_type *blkiop) { }

239 240
static inline char *blkg_path(struct blkio_group *blkg) { return NULL; }

241 242
#endif

243 244 245 246
#define BLKIO_WEIGHT_MIN	100
#define BLKIO_WEIGHT_MAX	1000
#define BLKIO_WEIGHT_DEFAULT	500

V
Vivek Goyal 已提交
247
#ifdef CONFIG_DEBUG_BLK_CGROUP
248
void blkiocg_update_avg_queue_size_stats(struct blkio_group *blkg);
249
void blkiocg_update_dequeue_stats(struct blkio_group *blkg,
250
				unsigned long dequeue);
251 252
void blkiocg_update_set_idle_time_stats(struct blkio_group *blkg);
void blkiocg_update_idle_time_stats(struct blkio_group *blkg);
253
void blkiocg_set_start_empty_time(struct blkio_group *blkg);
254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274

#define BLKG_FLAG_FNS(name)						\
static inline void blkio_mark_blkg_##name(				\
		struct blkio_group_stats *stats)			\
{									\
	stats->flags |= (1 << BLKG_##name);				\
}									\
static inline void blkio_clear_blkg_##name(				\
		struct blkio_group_stats *stats)			\
{									\
	stats->flags &= ~(1 << BLKG_##name);				\
}									\
static inline int blkio_blkg_##name(struct blkio_group_stats *stats)	\
{									\
	return (stats->flags & (1 << BLKG_##name)) != 0;		\
}									\

BLKG_FLAG_FNS(waiting)
BLKG_FLAG_FNS(idling)
BLKG_FLAG_FNS(empty)
#undef BLKG_FLAG_FNS
V
Vivek Goyal 已提交
275
#else
276
static inline void blkiocg_update_avg_queue_size_stats(
277
						struct blkio_group *blkg) {}
278 279
static inline void blkiocg_update_dequeue_stats(struct blkio_group *blkg,
						unsigned long dequeue) {}
280 281 282
static inline void blkiocg_update_set_idle_time_stats(struct blkio_group *blkg)
{}
static inline void blkiocg_update_idle_time_stats(struct blkio_group *blkg) {}
283
static inline void blkiocg_set_start_empty_time(struct blkio_group *blkg) {}
V
Vivek Goyal 已提交
284 285
#endif

B
Ben Blum 已提交
286
#if defined(CONFIG_BLK_CGROUP) || defined(CONFIG_BLK_CGROUP_MODULE)
287 288 289
extern struct blkio_cgroup blkio_root_cgroup;
extern struct blkio_cgroup *cgroup_to_blkio_cgroup(struct cgroup *cgroup);
extern void blkiocg_add_blkio_group(struct blkio_cgroup *blkcg,
290 291
	struct blkio_group *blkg, void *key, dev_t dev,
	enum blkio_policy_id plid);
292 293 294
extern int blkiocg_del_blkio_group(struct blkio_group *blkg);
extern struct blkio_group *blkiocg_lookup_group(struct blkio_cgroup *blkcg,
						void *key);
295 296
void blkiocg_update_timeslice_used(struct blkio_group *blkg,
					unsigned long time);
297 298 299 300
void blkiocg_update_dispatch_stats(struct blkio_group *blkg, uint64_t bytes,
						bool direction, bool sync);
void blkiocg_update_completion_stats(struct blkio_group *blkg,
	uint64_t start_time, uint64_t io_start_time, bool direction, bool sync);
D
Divyesh Shah 已提交
301 302
void blkiocg_update_io_merged_stats(struct blkio_group *blkg, bool direction,
					bool sync);
303
void blkiocg_update_io_add_stats(struct blkio_group *blkg,
304
		struct blkio_group *curr_blkg, bool direction, bool sync);
305
void blkiocg_update_io_remove_stats(struct blkio_group *blkg,
306
					bool direction, bool sync);
307
#else
308
struct cgroup;
309 310 311 312
static inline struct blkio_cgroup *
cgroup_to_blkio_cgroup(struct cgroup *cgroup) { return NULL; }

static inline void blkiocg_add_blkio_group(struct blkio_cgroup *blkcg,
313 314
		struct blkio_group *blkg, void *key, dev_t dev,
		enum blkio_policy_id plid) {}
315 316 317 318 319 320

static inline int
blkiocg_del_blkio_group(struct blkio_group *blkg) { return 0; }

static inline struct blkio_group *
blkiocg_lookup_group(struct blkio_cgroup *blkcg, void *key) { return NULL; }
321
static inline void blkiocg_update_timeslice_used(struct blkio_group *blkg,
322
						unsigned long time) {}
323 324 325 326 327
static inline void blkiocg_update_dispatch_stats(struct blkio_group *blkg,
				uint64_t bytes, bool direction, bool sync) {}
static inline void blkiocg_update_completion_stats(struct blkio_group *blkg,
		uint64_t start_time, uint64_t io_start_time, bool direction,
		bool sync) {}
D
Divyesh Shah 已提交
328 329
static inline void blkiocg_update_io_merged_stats(struct blkio_group *blkg,
						bool direction, bool sync) {}
330
static inline void blkiocg_update_io_add_stats(struct blkio_group *blkg,
331
		struct blkio_group *curr_blkg, bool direction, bool sync) {}
332
static inline void blkiocg_update_io_remove_stats(struct blkio_group *blkg,
333
						bool direction, bool sync) {}
334 335
#endif
#endif /* _BLK_CGROUP_H */