summaryrefslogtreecommitdiffstats
path: root/fs/btrfs/async-thread.c
blob: 04fb9702d14c999493c345dc976c73f3bb812a2a (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
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
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
/*
 * Copyright (C) 2007 Oracle.  All rights reserved.
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public
 * License v2 as published by the Free Software Foundation.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * General Public License for more details.
 *
 * You should have received a copy of the GNU General Public
 * License along with this program; if not, write to the
 * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
 * Boston, MA 021110-1307, USA.
 */

#include <linux/version.h>
#include <linux/kthread.h>
#include <linux/list.h>
#include <linux/spinlock.h>
# include <linux/freezer.h>
#include "async-thread.h"

/*
 * container for the kthread task pointer and the list of pending work
 * One of these is allocated per thread.
 */
struct btrfs_worker_thread {
	/* pool we belong to */
	struct btrfs_workers *workers;

	/* list of struct btrfs_work that are waiting for service */
	struct list_head pending;

	/* list of worker threads from struct btrfs_workers */
	struct list_head worker_list;

	/* kthread */
	struct task_struct *task;

	/* number of things on the pending list */
	atomic_t num_pending;

	unsigned long sequence;

	/* protects the pending list. */
	spinlock_t lock;

	/* set to non-zero when this thread is already awake and kicking */
	int working;

	/* are we currently idle */
	int idle;
};

/*
 * helper function to move a thread onto the idle list after it
 * has finished some requests.
 */
static void check_idle_worker(struct btrfs_worker_thread *worker)
{
	if (!worker->idle && atomic_read(&worker->num_pending) <
	    worker->workers->idle_thresh / 2) {
		unsigned long flags;
		spin_lock_irqsave(&worker->workers->lock, flags);
		worker->idle = 1;
		list_move(&worker->worker_list, &worker->workers->idle_list);
		spin_unlock_irqrestore(&worker->workers->lock, flags);
	}
}

/*
 * helper function to move a thread off the idle list after new
 * pending work is added.
 */
static void check_busy_worker(struct btrfs_worker_thread *worker)
{
	if (worker->idle && atomic_read(&worker->num_pending) >=
	    worker->workers->idle_thresh) {
		unsigned long flags;
		spin_lock_irqsave(&worker->workers->lock, flags);
		worker->idle = 0;
		list_move_tail(&worker->worker_list,
			       &worker->workers->worker_list);
		spin_unlock_irqrestore(&worker->workers->lock, flags);
	}
}

/*
 * main loop for servicing work items
 */
static int worker_loop(void *arg)
{
	struct btrfs_worker_thread *worker = arg;
	struct list_head *cur;
	struct btrfs_work *work;
	do {
		spin_lock_irq(&worker->lock);
		while(!list_empty(&worker->pending)) {
			cur = worker->pending.next;
			work = list_entry(cur, struct btrfs_work, list);
			list_del(&work->list);
			clear_bit(0, &work->flags);

			work->worker = worker;
			spin_unlock_irq(&worker->lock);

			work->func(work);

			atomic_dec(&worker->num_pending);
			spin_lock_irq(&worker->lock);
			check_idle_worker(worker);
		}
		worker->working = 0;
		if (freezing(current)) {
			refrigerator();
		} else {
			set_current_state(TASK_INTERRUPTIBLE);
			spin_unlock_irq(&worker->lock);
			schedule();
			__set_current_state(TASK_RUNNING);
		}
	} while (!kthread_should_stop());
	return 0;
}

/*
 * this will wait for all the worker threads to shutdown
 */
int btrfs_stop_workers(struct btrfs_workers *workers)
{
	struct list_head *cur;
	struct btrfs_worker_thread *worker;

	list_splice_init(&workers->idle_list, &workers->worker_list);
	while(!list_empty(&workers->worker_list)) {
		cur = workers->worker_list.next;
		worker = list_entry(cur, struct btrfs_worker_thread,
				    worker_list);
		kthread_stop(worker->task);
		list_del(&worker->worker_list);
		kfree(worker);
	}
	return 0;
}

/*
 * simple init on struct btrfs_workers
 */
void btrfs_init_workers(struct btrfs_workers *workers, char *name, int max)
{
	workers->num_workers = 0;
	INIT_LIST_HEAD(&workers->worker_list);
	INIT_LIST_HEAD(&workers->idle_list);
	spin_lock_init(&workers->lock);
	workers->max_workers = max;
	workers->idle_thresh = 32;
	workers->name = name;
}

/*
 * starts new worker threads.  This does not enforce the max worker
 * count in case you need to temporarily go past it.
 */
int btrfs_start_workers(struct btrfs_workers *workers, int num_workers)
{
	struct btrfs_worker_thread *worker;
	int ret = 0;
	int i;

	for (i = 0; i < num_workers; i++) {
		worker = kzalloc(sizeof(*worker), GFP_NOFS);
		if (!worker) {
			ret = -ENOMEM;
			goto fail;
		}

		INIT_LIST_HEAD(&worker->pending);
		INIT_LIST_HEAD(&worker->worker_list);
		spin_lock_init(&worker->lock);
		atomic_set(&worker->num_pending, 0);
		worker->task = kthread_run(worker_loop, worker,
					   "btrfs-%s-%d", workers->name,
					   workers->num_workers + i);
		worker->workers = workers;
		if (IS_ERR(worker->task)) {
			kfree(worker);
			ret = PTR_ERR(worker->task);
			goto fail;
		}

		spin_lock_irq(&workers->lock);
		list_add_tail(&worker->worker_list, &workers->idle_list);
		worker->idle = 1;
		workers->num_workers++;
		spin_unlock_irq(&workers->lock);
	}
	return 0;
fail:
	btrfs_stop_workers(workers);
	return ret;
}

/*
 * run through the list and find a worker thread that doesn't have a lot
 * to do right now.  This can return null if we aren't yet at the thread
 * count limit and all of the threads are busy.
 */
static struct btrfs_worker_thread *next_worker(struct btrfs_workers *workers)
{
	struct btrfs_worker_thread *worker;
	struct list_head *next;
	int enforce_min = workers->num_workers < workers->max_workers;

	/*
	 * if we find an idle thread, don't move it to the end of the
	 * idle list.  This improves the chance that the next submission
	 * will reuse the same thread, and maybe catch it while it is still
	 * working
	 */
	if (!list_empty(&workers->idle_list)) {
		next = workers->idle_list.next;
		worker = list_entry(next, struct btrfs_worker_thread,
				    worker_list);
		return worker;
	}
	if (enforce_min || list_empty(&workers->worker_list))
		return NULL;

	/*
	 * if we pick a busy task, move the task to the end of the list.
	 * hopefully this will keep things somewhat evenly balanced.
	 * Do the move in batches based on the sequence number.  This groups
	 * requests submitted at roughly the same time onto the same worker.
	 */
	next = workers->worker_list.next;
	worker = list_entry(next, struct btrfs_worker_thread, worker_list);
	atomic_inc(&worker->num_pending);
	worker->sequence++;

	if (worker->sequence % workers->idle_thresh == 0)
		list_move_tail(next, &workers->worker_list);
	return worker;
}

/*
 * selects a worker thread to take the next job.  This will either find
 * an idle worker, start a new worker up to the max count, or just return
 * one of the existing busy workers.
 */
static struct btrfs_worker_thread *find_worker(struct btrfs_workers *workers)
{
	struct btrfs_worker_thread *worker;
	unsigned long flags;

again:
	spin_lock_irqsave(&workers->lock, flags);
	worker = next_worker(workers);
	spin_unlock_irqrestore(&workers->lock, flags);

	if (!worker) {
		spin_lock_irqsave(&workers->lock, flags);
		if (workers->num_workers >= workers->max_workers) {
			struct list_head *fallback = NULL;
			/*
			 * we have failed to find any workers, just
			 * return the force one
			 */
			if (!list_empty(&workers->worker_list))
				fallback = workers->worker_list.next;
			if (!list_empty(&workers->idle_list))
				fallback = workers->idle_list.next;
			BUG_ON(!fallback);
			worker = list_entry(fallback,
				  struct btrfs_worker_thread, worker_list);
			spin_unlock_irqrestore(&workers->lock, flags);
		} else {
			spin_unlock_irqrestore(&workers->lock, flags);
			/* we're below the limit, start another worker */
			btrfs_start_workers(workers, 1);
			goto again;
		}
	}
	return worker;
}

/*
 * btrfs_requeue_work just puts the work item back on the tail of the list
 * it was taken from.  It is intended for use with long running work functions
 * that make some progress and want to give the cpu up for others.
 */
int btrfs_requeue_work(struct btrfs_work *work)
{
	struct btrfs_worker_thread *worker = work->worker;
	unsigned long flags;

	if (test_and_set_bit(0, &work->flags))
		goto out;

	spin_lock_irqsave(&worker->lock, flags);
	atomic_inc(&worker->num_pending);
	list_add_tail(&work->list, &worker->pending);
	check_busy_worker(worker);
	spin_unlock_irqrestore(&worker->lock, flags);
out:
	return 0;
}

/*
 * places a struct btrfs_work into the pending queue of one of the kthreads
 */
int btrfs_queue_worker(struct btrfs_workers *workers, struct btrfs_work *work)
{
	struct btrfs_worker_thread *worker;
	unsigned long flags;
	int wake = 0;

	/* don't requeue something already on a list */
	if (test_and_set_bit(0, &work->flags))
		goto out;

	worker = find_worker(workers);

	spin_lock_irqsave(&worker->lock, flags);
	atomic_inc(&worker->num_pending);
	check_busy_worker(worker);
	list_add_tail(&work->list, &worker->pending);

	/*
	 * avoid calling into wake_up_process if this thread has already
	 * been kicked
	 */
	if (!worker->working)
		wake = 1;
	worker->working = 1;

	spin_unlock_irqrestore(&worker->lock, flags);

	if (wake)
		wake_up_process(worker->task);
out:
	return 0;
}