blob: 58b7d14b08ee1d47ec66a06243ee16f63854b676 [file] [log] [blame]
Chris Mason8b712842008-06-11 16:50:36 -04001/*
2 * Copyright (C) 2007 Oracle. All rights reserved.
3 *
4 * This program is free software; you can redistribute it and/or
5 * modify it under the terms of the GNU General Public
6 * License v2 as published by the Free Software Foundation.
7 *
8 * This program is distributed in the hope that it will be useful,
9 * but WITHOUT ANY WARRANTY; without even the implied warranty of
10 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
11 * General Public License for more details.
12 *
13 * You should have received a copy of the GNU General Public
14 * License along with this program; if not, write to the
15 * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
16 * Boston, MA 021110-1307, USA.
17 */
18
19#include <linux/kthread.h>
Tejun Heo5a0e3ad2010-03-24 17:04:11 +090020#include <linux/slab.h>
Chris Mason8b712842008-06-11 16:50:36 -040021#include <linux/list.h>
22#include <linux/spinlock.h>
Chris Masonb51912c2009-02-04 09:23:24 -050023#include <linux/freezer.h>
Chris Mason8b712842008-06-11 16:50:36 -040024#include "async-thread.h"
25
Chris Mason4a69a412008-11-06 22:03:00 -050026#define WORK_QUEUED_BIT 0
27#define WORK_DONE_BIT 1
28#define WORK_ORDER_DONE_BIT 2
Chris Masond313d7a2009-04-20 15:50:09 -040029#define WORK_HIGH_PRIO_BIT 3
Chris Mason4a69a412008-11-06 22:03:00 -050030
Chris Mason8b712842008-06-11 16:50:36 -040031/*
32 * container for the kthread task pointer and the list of pending work
33 * One of these is allocated per thread.
34 */
35struct btrfs_worker_thread {
Chris Mason35d8ba62008-06-11 20:21:24 -040036 /* pool we belong to */
37 struct btrfs_workers *workers;
38
Chris Mason8b712842008-06-11 16:50:36 -040039 /* list of struct btrfs_work that are waiting for service */
40 struct list_head pending;
Chris Masond313d7a2009-04-20 15:50:09 -040041 struct list_head prio_pending;
Chris Mason8b712842008-06-11 16:50:36 -040042
43 /* list of worker threads from struct btrfs_workers */
44 struct list_head worker_list;
45
46 /* kthread */
47 struct task_struct *task;
48
49 /* number of things on the pending list */
50 atomic_t num_pending;
Chris Mason53863232008-08-15 15:34:18 -040051
Chris Mason90428462009-08-04 16:56:34 -040052 /* reference counter for this struct */
53 atomic_t refs;
54
Chris Mason4854ddd2008-08-15 15:34:17 -040055 unsigned long sequence;
Chris Mason8b712842008-06-11 16:50:36 -040056
57 /* protects the pending list. */
58 spinlock_t lock;
59
60 /* set to non-zero when this thread is already awake and kicking */
61 int working;
Chris Mason35d8ba62008-06-11 20:21:24 -040062
63 /* are we currently idle */
64 int idle;
Chris Mason8b712842008-06-11 16:50:36 -040065};
66
Josef Bacik0dc3b842011-11-18 14:37:27 -050067static int __btrfs_start_workers(struct btrfs_workers *workers);
68
Chris Mason8b712842008-06-11 16:50:36 -040069/*
Chris Mason61d92c32009-10-02 19:11:56 -040070 * btrfs_start_workers uses kthread_run, which can block waiting for memory
71 * for a very long time. It will actually throttle on page writeback,
72 * and so it may not make progress until after our btrfs worker threads
73 * process all of the pending work structs in their queue
74 *
75 * This means we can't use btrfs_start_workers from inside a btrfs worker
76 * thread that is used as part of cleaning dirty memory, which pretty much
77 * involves all of the worker threads.
78 *
79 * Instead we have a helper queue who never has more than one thread
80 * where we scheduler thread start operations. This worker_start struct
81 * is used to contain the work and hold a pointer to the queue that needs
82 * another worker.
83 */
84struct worker_start {
85 struct btrfs_work work;
86 struct btrfs_workers *queue;
87};
88
89static void start_new_worker_func(struct btrfs_work *work)
90{
91 struct worker_start *start;
92 start = container_of(work, struct worker_start, work);
Josef Bacik0dc3b842011-11-18 14:37:27 -050093 __btrfs_start_workers(start->queue);
Chris Mason61d92c32009-10-02 19:11:56 -040094 kfree(start);
95}
96
Chris Mason61d92c32009-10-02 19:11:56 -040097/*
Chris Mason35d8ba62008-06-11 20:21:24 -040098 * helper function to move a thread onto the idle list after it
99 * has finished some requests.
100 */
101static void check_idle_worker(struct btrfs_worker_thread *worker)
102{
103 if (!worker->idle && atomic_read(&worker->num_pending) <
104 worker->workers->idle_thresh / 2) {
105 unsigned long flags;
106 spin_lock_irqsave(&worker->workers->lock, flags);
107 worker->idle = 1;
Chris Mason3e99d8e2009-09-15 19:57:42 -0400108
109 /* the list may be empty if the worker is just starting */
110 if (!list_empty(&worker->worker_list)) {
111 list_move(&worker->worker_list,
112 &worker->workers->idle_list);
113 }
Chris Mason35d8ba62008-06-11 20:21:24 -0400114 spin_unlock_irqrestore(&worker->workers->lock, flags);
115 }
116}
117
118/*
119 * helper function to move a thread off the idle list after new
120 * pending work is added.
121 */
122static void check_busy_worker(struct btrfs_worker_thread *worker)
123{
124 if (worker->idle && atomic_read(&worker->num_pending) >=
125 worker->workers->idle_thresh) {
126 unsigned long flags;
127 spin_lock_irqsave(&worker->workers->lock, flags);
128 worker->idle = 0;
Chris Mason3e99d8e2009-09-15 19:57:42 -0400129
130 if (!list_empty(&worker->worker_list)) {
131 list_move_tail(&worker->worker_list,
132 &worker->workers->worker_list);
133 }
Chris Mason35d8ba62008-06-11 20:21:24 -0400134 spin_unlock_irqrestore(&worker->workers->lock, flags);
135 }
136}
137
Chris Mason90428462009-08-04 16:56:34 -0400138static void check_pending_worker_creates(struct btrfs_worker_thread *worker)
139{
140 struct btrfs_workers *workers = worker->workers;
Josef Bacik0dc3b842011-11-18 14:37:27 -0500141 struct worker_start *start;
Chris Mason90428462009-08-04 16:56:34 -0400142 unsigned long flags;
143
144 rmb();
145 if (!workers->atomic_start_pending)
146 return;
147
Josef Bacik0dc3b842011-11-18 14:37:27 -0500148 start = kzalloc(sizeof(*start), GFP_NOFS);
149 if (!start)
150 return;
151
152 start->work.func = start_new_worker_func;
153 start->queue = workers;
154
Chris Mason90428462009-08-04 16:56:34 -0400155 spin_lock_irqsave(&workers->lock, flags);
156 if (!workers->atomic_start_pending)
157 goto out;
158
159 workers->atomic_start_pending = 0;
Chris Mason61d92c32009-10-02 19:11:56 -0400160 if (workers->num_workers + workers->num_workers_starting >=
161 workers->max_workers)
Chris Mason90428462009-08-04 16:56:34 -0400162 goto out;
163
Chris Mason61d92c32009-10-02 19:11:56 -0400164 workers->num_workers_starting += 1;
Chris Mason90428462009-08-04 16:56:34 -0400165 spin_unlock_irqrestore(&workers->lock, flags);
Josef Bacik0dc3b842011-11-18 14:37:27 -0500166 btrfs_queue_worker(workers->atomic_worker_start, &start->work);
Chris Mason90428462009-08-04 16:56:34 -0400167 return;
168
169out:
Josef Bacik0dc3b842011-11-18 14:37:27 -0500170 kfree(start);
Chris Mason90428462009-08-04 16:56:34 -0400171 spin_unlock_irqrestore(&workers->lock, flags);
172}
173
Jeff Mahoney143bede2012-03-01 14:56:26 +0100174static noinline void run_ordered_completions(struct btrfs_workers *workers,
Chris Mason4a69a412008-11-06 22:03:00 -0500175 struct btrfs_work *work)
176{
Chris Mason4a69a412008-11-06 22:03:00 -0500177 if (!workers->ordered)
Jeff Mahoney143bede2012-03-01 14:56:26 +0100178 return;
Chris Mason4a69a412008-11-06 22:03:00 -0500179
180 set_bit(WORK_DONE_BIT, &work->flags);
181
Chris Mason4e3f9c52009-08-05 16:36:45 -0400182 spin_lock(&workers->order_lock);
Chris Mason4a69a412008-11-06 22:03:00 -0500183
Chris Masond313d7a2009-04-20 15:50:09 -0400184 while (1) {
185 if (!list_empty(&workers->prio_order_list)) {
186 work = list_entry(workers->prio_order_list.next,
187 struct btrfs_work, order_list);
188 } else if (!list_empty(&workers->order_list)) {
189 work = list_entry(workers->order_list.next,
190 struct btrfs_work, order_list);
191 } else {
192 break;
193 }
Chris Mason4a69a412008-11-06 22:03:00 -0500194 if (!test_bit(WORK_DONE_BIT, &work->flags))
195 break;
196
197 /* we are going to call the ordered done function, but
198 * we leave the work item on the list as a barrier so
199 * that later work items that are done don't have their
200 * functions called before this one returns
201 */
202 if (test_and_set_bit(WORK_ORDER_DONE_BIT, &work->flags))
203 break;
204
Chris Mason4e3f9c52009-08-05 16:36:45 -0400205 spin_unlock(&workers->order_lock);
Chris Mason4a69a412008-11-06 22:03:00 -0500206
207 work->ordered_func(work);
208
Chris Masone9fbcb42012-07-25 15:57:13 -0400209 /* now take the lock again and drop our item from the list */
Chris Mason4e3f9c52009-08-05 16:36:45 -0400210 spin_lock(&workers->order_lock);
Chris Mason4a69a412008-11-06 22:03:00 -0500211 list_del(&work->order_list);
Chris Masone9fbcb42012-07-25 15:57:13 -0400212 spin_unlock(&workers->order_lock);
213
214 /*
215 * we don't want to call the ordered free functions
216 * with the lock held though
217 */
Chris Mason4a69a412008-11-06 22:03:00 -0500218 work->ordered_free(work);
Chris Masone9fbcb42012-07-25 15:57:13 -0400219 spin_lock(&workers->order_lock);
Chris Mason4a69a412008-11-06 22:03:00 -0500220 }
221
Chris Mason4e3f9c52009-08-05 16:36:45 -0400222 spin_unlock(&workers->order_lock);
Chris Mason4a69a412008-11-06 22:03:00 -0500223}
224
Chris Mason90428462009-08-04 16:56:34 -0400225static void put_worker(struct btrfs_worker_thread *worker)
226{
227 if (atomic_dec_and_test(&worker->refs))
228 kfree(worker);
229}
230
231static int try_worker_shutdown(struct btrfs_worker_thread *worker)
232{
233 int freeit = 0;
234
235 spin_lock_irq(&worker->lock);
Chris Mason627e4212009-09-15 20:00:36 -0400236 spin_lock(&worker->workers->lock);
Chris Mason90428462009-08-04 16:56:34 -0400237 if (worker->workers->num_workers > 1 &&
238 worker->idle &&
239 !worker->working &&
240 !list_empty(&worker->worker_list) &&
241 list_empty(&worker->prio_pending) &&
Chris Mason6e740572009-09-15 20:02:33 -0400242 list_empty(&worker->pending) &&
243 atomic_read(&worker->num_pending) == 0) {
Chris Mason90428462009-08-04 16:56:34 -0400244 freeit = 1;
245 list_del_init(&worker->worker_list);
246 worker->workers->num_workers--;
247 }
Chris Mason627e4212009-09-15 20:00:36 -0400248 spin_unlock(&worker->workers->lock);
Chris Mason90428462009-08-04 16:56:34 -0400249 spin_unlock_irq(&worker->lock);
250
251 if (freeit)
252 put_worker(worker);
253 return freeit;
254}
255
Chris Mason4f878e82009-08-07 09:27:38 -0400256static struct btrfs_work *get_next_work(struct btrfs_worker_thread *worker,
257 struct list_head *prio_head,
258 struct list_head *head)
259{
260 struct btrfs_work *work = NULL;
261 struct list_head *cur = NULL;
262
263 if(!list_empty(prio_head))
264 cur = prio_head->next;
265
266 smp_mb();
267 if (!list_empty(&worker->prio_pending))
268 goto refill;
269
270 if (!list_empty(head))
271 cur = head->next;
272
273 if (cur)
274 goto out;
275
276refill:
277 spin_lock_irq(&worker->lock);
278 list_splice_tail_init(&worker->prio_pending, prio_head);
279 list_splice_tail_init(&worker->pending, head);
280
281 if (!list_empty(prio_head))
282 cur = prio_head->next;
283 else if (!list_empty(head))
284 cur = head->next;
285 spin_unlock_irq(&worker->lock);
286
287 if (!cur)
288 goto out_fail;
289
290out:
291 work = list_entry(cur, struct btrfs_work, list);
292
293out_fail:
294 return work;
295}
296
Chris Mason35d8ba62008-06-11 20:21:24 -0400297/*
Chris Mason8b712842008-06-11 16:50:36 -0400298 * main loop for servicing work items
299 */
300static int worker_loop(void *arg)
301{
302 struct btrfs_worker_thread *worker = arg;
Chris Mason4f878e82009-08-07 09:27:38 -0400303 struct list_head head;
304 struct list_head prio_head;
Chris Mason8b712842008-06-11 16:50:36 -0400305 struct btrfs_work *work;
Chris Mason4f878e82009-08-07 09:27:38 -0400306
307 INIT_LIST_HEAD(&head);
308 INIT_LIST_HEAD(&prio_head);
309
Chris Mason8b712842008-06-11 16:50:36 -0400310 do {
Chris Mason4f878e82009-08-07 09:27:38 -0400311again:
Chris Masond313d7a2009-04-20 15:50:09 -0400312 while (1) {
Chris Mason4f878e82009-08-07 09:27:38 -0400313
314
315 work = get_next_work(worker, &prio_head, &head);
316 if (!work)
Chris Masond313d7a2009-04-20 15:50:09 -0400317 break;
318
Chris Mason8b712842008-06-11 16:50:36 -0400319 list_del(&work->list);
Chris Mason4a69a412008-11-06 22:03:00 -0500320 clear_bit(WORK_QUEUED_BIT, &work->flags);
Chris Mason8b712842008-06-11 16:50:36 -0400321
322 work->worker = worker;
Chris Mason8b712842008-06-11 16:50:36 -0400323
324 work->func(work);
325
326 atomic_dec(&worker->num_pending);
Chris Mason4a69a412008-11-06 22:03:00 -0500327 /*
328 * unless this is an ordered work queue,
329 * 'work' was probably freed by func above.
330 */
331 run_ordered_completions(worker->workers, work);
332
Chris Mason90428462009-08-04 16:56:34 -0400333 check_pending_worker_creates(worker);
Chris Mason8f3b65a2011-12-15 09:29:43 -0500334 cond_resched();
Chris Mason8b712842008-06-11 16:50:36 -0400335 }
Chris Mason4f878e82009-08-07 09:27:38 -0400336
337 spin_lock_irq(&worker->lock);
338 check_idle_worker(worker);
339
Chris Mason8b712842008-06-11 16:50:36 -0400340 if (freezing(current)) {
Chris Masonb51912c2009-02-04 09:23:24 -0500341 worker->working = 0;
342 spin_unlock_irq(&worker->lock);
Tejun Heoa0acae02011-11-21 12:32:22 -0800343 try_to_freeze();
Chris Mason8b712842008-06-11 16:50:36 -0400344 } else {
Chris Mason8b712842008-06-11 16:50:36 -0400345 spin_unlock_irq(&worker->lock);
Chris Masonb51912c2009-02-04 09:23:24 -0500346 if (!kthread_should_stop()) {
347 cpu_relax();
348 /*
349 * we've dropped the lock, did someone else
350 * jump_in?
351 */
352 smp_mb();
Chris Masond313d7a2009-04-20 15:50:09 -0400353 if (!list_empty(&worker->pending) ||
354 !list_empty(&worker->prio_pending))
Chris Masonb51912c2009-02-04 09:23:24 -0500355 continue;
356
357 /*
358 * this short schedule allows more work to
359 * come in without the queue functions
360 * needing to go through wake_up_process()
361 *
362 * worker->working is still 1, so nobody
363 * is going to try and wake us up
364 */
365 schedule_timeout(1);
366 smp_mb();
Chris Masond313d7a2009-04-20 15:50:09 -0400367 if (!list_empty(&worker->pending) ||
368 !list_empty(&worker->prio_pending))
Chris Masonb51912c2009-02-04 09:23:24 -0500369 continue;
370
Amit Gudb5555f72009-04-02 17:01:27 -0400371 if (kthread_should_stop())
372 break;
373
Chris Masonb51912c2009-02-04 09:23:24 -0500374 /* still no more work?, sleep for real */
375 spin_lock_irq(&worker->lock);
376 set_current_state(TASK_INTERRUPTIBLE);
Chris Masond313d7a2009-04-20 15:50:09 -0400377 if (!list_empty(&worker->pending) ||
Chris Mason4f878e82009-08-07 09:27:38 -0400378 !list_empty(&worker->prio_pending)) {
379 spin_unlock_irq(&worker->lock);
Chris Masoned3b3d3142010-05-25 10:12:41 -0400380 set_current_state(TASK_RUNNING);
Chris Mason4f878e82009-08-07 09:27:38 -0400381 goto again;
382 }
Chris Masonb51912c2009-02-04 09:23:24 -0500383
384 /*
385 * this makes sure we get a wakeup when someone
386 * adds something new to the queue
387 */
388 worker->working = 0;
389 spin_unlock_irq(&worker->lock);
390
Chris Mason90428462009-08-04 16:56:34 -0400391 if (!kthread_should_stop()) {
392 schedule_timeout(HZ * 120);
393 if (!worker->working &&
394 try_worker_shutdown(worker)) {
395 return 0;
396 }
397 }
Chris Masonb51912c2009-02-04 09:23:24 -0500398 }
Chris Mason8b712842008-06-11 16:50:36 -0400399 __set_current_state(TASK_RUNNING);
400 }
401 } while (!kthread_should_stop());
402 return 0;
403}
404
405/*
406 * this will wait for all the worker threads to shutdown
407 */
Jeff Mahoney143bede2012-03-01 14:56:26 +0100408void btrfs_stop_workers(struct btrfs_workers *workers)
Chris Mason8b712842008-06-11 16:50:36 -0400409{
410 struct list_head *cur;
411 struct btrfs_worker_thread *worker;
Chris Mason90428462009-08-04 16:56:34 -0400412 int can_stop;
Chris Mason8b712842008-06-11 16:50:36 -0400413
Chris Mason90428462009-08-04 16:56:34 -0400414 spin_lock_irq(&workers->lock);
Chris Mason35d8ba62008-06-11 20:21:24 -0400415 list_splice_init(&workers->idle_list, &workers->worker_list);
Chris Masond3977122009-01-05 21:25:51 -0500416 while (!list_empty(&workers->worker_list)) {
Chris Mason8b712842008-06-11 16:50:36 -0400417 cur = workers->worker_list.next;
418 worker = list_entry(cur, struct btrfs_worker_thread,
419 worker_list);
Chris Mason90428462009-08-04 16:56:34 -0400420
421 atomic_inc(&worker->refs);
422 workers->num_workers -= 1;
423 if (!list_empty(&worker->worker_list)) {
424 list_del_init(&worker->worker_list);
425 put_worker(worker);
426 can_stop = 1;
427 } else
428 can_stop = 0;
429 spin_unlock_irq(&workers->lock);
430 if (can_stop)
431 kthread_stop(worker->task);
432 spin_lock_irq(&workers->lock);
433 put_worker(worker);
Chris Mason8b712842008-06-11 16:50:36 -0400434 }
Chris Mason90428462009-08-04 16:56:34 -0400435 spin_unlock_irq(&workers->lock);
Chris Mason8b712842008-06-11 16:50:36 -0400436}
437
438/*
439 * simple init on struct btrfs_workers
440 */
Chris Mason61d92c32009-10-02 19:11:56 -0400441void btrfs_init_workers(struct btrfs_workers *workers, char *name, int max,
442 struct btrfs_workers *async_helper)
Chris Mason8b712842008-06-11 16:50:36 -0400443{
444 workers->num_workers = 0;
Chris Mason61d92c32009-10-02 19:11:56 -0400445 workers->num_workers_starting = 0;
Chris Mason8b712842008-06-11 16:50:36 -0400446 INIT_LIST_HEAD(&workers->worker_list);
Chris Mason35d8ba62008-06-11 20:21:24 -0400447 INIT_LIST_HEAD(&workers->idle_list);
Chris Mason4a69a412008-11-06 22:03:00 -0500448 INIT_LIST_HEAD(&workers->order_list);
Chris Masond313d7a2009-04-20 15:50:09 -0400449 INIT_LIST_HEAD(&workers->prio_order_list);
Chris Mason8b712842008-06-11 16:50:36 -0400450 spin_lock_init(&workers->lock);
Chris Mason4e3f9c52009-08-05 16:36:45 -0400451 spin_lock_init(&workers->order_lock);
Chris Mason8b712842008-06-11 16:50:36 -0400452 workers->max_workers = max;
Chris Mason61b49442008-07-31 15:42:53 -0400453 workers->idle_thresh = 32;
Chris Mason5443be42008-08-15 15:34:16 -0400454 workers->name = name;
Chris Mason4a69a412008-11-06 22:03:00 -0500455 workers->ordered = 0;
Chris Mason90428462009-08-04 16:56:34 -0400456 workers->atomic_start_pending = 0;
Chris Mason61d92c32009-10-02 19:11:56 -0400457 workers->atomic_worker_start = async_helper;
Chris Mason8b712842008-06-11 16:50:36 -0400458}
459
460/*
461 * starts new worker threads. This does not enforce the max worker
462 * count in case you need to temporarily go past it.
463 */
Josef Bacik0dc3b842011-11-18 14:37:27 -0500464static int __btrfs_start_workers(struct btrfs_workers *workers)
Chris Mason8b712842008-06-11 16:50:36 -0400465{
466 struct btrfs_worker_thread *worker;
467 int ret = 0;
Chris Mason8b712842008-06-11 16:50:36 -0400468
Josef Bacik0dc3b842011-11-18 14:37:27 -0500469 worker = kzalloc(sizeof(*worker), GFP_NOFS);
470 if (!worker) {
471 ret = -ENOMEM;
472 goto fail;
Chris Mason8b712842008-06-11 16:50:36 -0400473 }
Josef Bacik0dc3b842011-11-18 14:37:27 -0500474
475 INIT_LIST_HEAD(&worker->pending);
476 INIT_LIST_HEAD(&worker->prio_pending);
477 INIT_LIST_HEAD(&worker->worker_list);
478 spin_lock_init(&worker->lock);
479
480 atomic_set(&worker->num_pending, 0);
481 atomic_set(&worker->refs, 1);
482 worker->workers = workers;
483 worker->task = kthread_run(worker_loop, worker,
484 "btrfs-%s-%d", workers->name,
485 workers->num_workers + 1);
486 if (IS_ERR(worker->task)) {
487 ret = PTR_ERR(worker->task);
488 kfree(worker);
489 goto fail;
490 }
491 spin_lock_irq(&workers->lock);
492 list_add_tail(&worker->worker_list, &workers->idle_list);
493 worker->idle = 1;
494 workers->num_workers++;
495 workers->num_workers_starting--;
496 WARN_ON(workers->num_workers_starting < 0);
497 spin_unlock_irq(&workers->lock);
498
Chris Mason8b712842008-06-11 16:50:36 -0400499 return 0;
500fail:
Josef Bacik0dc3b842011-11-18 14:37:27 -0500501 spin_lock_irq(&workers->lock);
502 workers->num_workers_starting--;
503 spin_unlock_irq(&workers->lock);
Chris Mason8b712842008-06-11 16:50:36 -0400504 return ret;
505}
506
Josef Bacik0dc3b842011-11-18 14:37:27 -0500507int btrfs_start_workers(struct btrfs_workers *workers)
Chris Mason61d92c32009-10-02 19:11:56 -0400508{
509 spin_lock_irq(&workers->lock);
Josef Bacik0dc3b842011-11-18 14:37:27 -0500510 workers->num_workers_starting++;
Chris Mason61d92c32009-10-02 19:11:56 -0400511 spin_unlock_irq(&workers->lock);
Josef Bacik0dc3b842011-11-18 14:37:27 -0500512 return __btrfs_start_workers(workers);
Chris Mason61d92c32009-10-02 19:11:56 -0400513}
514
Chris Mason8b712842008-06-11 16:50:36 -0400515/*
516 * run through the list and find a worker thread that doesn't have a lot
517 * to do right now. This can return null if we aren't yet at the thread
518 * count limit and all of the threads are busy.
519 */
520static struct btrfs_worker_thread *next_worker(struct btrfs_workers *workers)
521{
522 struct btrfs_worker_thread *worker;
523 struct list_head *next;
Chris Mason61d92c32009-10-02 19:11:56 -0400524 int enforce_min;
525
526 enforce_min = (workers->num_workers + workers->num_workers_starting) <
527 workers->max_workers;
Chris Mason8b712842008-06-11 16:50:36 -0400528
Chris Mason8b712842008-06-11 16:50:36 -0400529 /*
Chris Mason35d8ba62008-06-11 20:21:24 -0400530 * if we find an idle thread, don't move it to the end of the
531 * idle list. This improves the chance that the next submission
532 * will reuse the same thread, and maybe catch it while it is still
533 * working
Chris Mason8b712842008-06-11 16:50:36 -0400534 */
Chris Mason35d8ba62008-06-11 20:21:24 -0400535 if (!list_empty(&workers->idle_list)) {
536 next = workers->idle_list.next;
Chris Mason8b712842008-06-11 16:50:36 -0400537 worker = list_entry(next, struct btrfs_worker_thread,
538 worker_list);
Chris Mason35d8ba62008-06-11 20:21:24 -0400539 return worker;
Chris Mason8b712842008-06-11 16:50:36 -0400540 }
Chris Mason35d8ba62008-06-11 20:21:24 -0400541 if (enforce_min || list_empty(&workers->worker_list))
542 return NULL;
543
Chris Mason8b712842008-06-11 16:50:36 -0400544 /*
Chris Mason35d8ba62008-06-11 20:21:24 -0400545 * if we pick a busy task, move the task to the end of the list.
Chris Masond352ac62008-09-29 15:18:18 -0400546 * hopefully this will keep things somewhat evenly balanced.
547 * Do the move in batches based on the sequence number. This groups
548 * requests submitted at roughly the same time onto the same worker.
Chris Mason8b712842008-06-11 16:50:36 -0400549 */
Chris Mason35d8ba62008-06-11 20:21:24 -0400550 next = workers->worker_list.next;
551 worker = list_entry(next, struct btrfs_worker_thread, worker_list);
Chris Mason4854ddd2008-08-15 15:34:17 -0400552 worker->sequence++;
Chris Masond352ac62008-09-29 15:18:18 -0400553
Chris Mason53863232008-08-15 15:34:18 -0400554 if (worker->sequence % workers->idle_thresh == 0)
Chris Mason4854ddd2008-08-15 15:34:17 -0400555 list_move_tail(next, &workers->worker_list);
Chris Mason8b712842008-06-11 16:50:36 -0400556 return worker;
557}
558
Chris Masond352ac62008-09-29 15:18:18 -0400559/*
560 * selects a worker thread to take the next job. This will either find
561 * an idle worker, start a new worker up to the max count, or just return
562 * one of the existing busy workers.
563 */
Chris Mason8b712842008-06-11 16:50:36 -0400564static struct btrfs_worker_thread *find_worker(struct btrfs_workers *workers)
565{
566 struct btrfs_worker_thread *worker;
567 unsigned long flags;
Chris Mason90428462009-08-04 16:56:34 -0400568 struct list_head *fallback;
Josef Bacik0dc3b842011-11-18 14:37:27 -0500569 int ret;
Chris Mason8b712842008-06-11 16:50:36 -0400570
Chris Mason8b712842008-06-11 16:50:36 -0400571 spin_lock_irqsave(&workers->lock, flags);
Chris Mason8d532b22011-12-23 07:53:00 -0500572again:
Chris Mason8b712842008-06-11 16:50:36 -0400573 worker = next_worker(workers);
Chris Mason8b712842008-06-11 16:50:36 -0400574
575 if (!worker) {
Chris Mason61d92c32009-10-02 19:11:56 -0400576 if (workers->num_workers + workers->num_workers_starting >=
577 workers->max_workers) {
Chris Mason90428462009-08-04 16:56:34 -0400578 goto fallback;
579 } else if (workers->atomic_worker_start) {
580 workers->atomic_start_pending = 1;
581 goto fallback;
Chris Mason8b712842008-06-11 16:50:36 -0400582 } else {
Chris Mason61d92c32009-10-02 19:11:56 -0400583 workers->num_workers_starting++;
Chris Mason8b712842008-06-11 16:50:36 -0400584 spin_unlock_irqrestore(&workers->lock, flags);
585 /* we're below the limit, start another worker */
Josef Bacik0dc3b842011-11-18 14:37:27 -0500586 ret = __btrfs_start_workers(workers);
Chris Mason8d532b22011-12-23 07:53:00 -0500587 spin_lock_irqsave(&workers->lock, flags);
Josef Bacik0dc3b842011-11-18 14:37:27 -0500588 if (ret)
589 goto fallback;
Chris Mason8b712842008-06-11 16:50:36 -0400590 goto again;
591 }
592 }
Chris Mason6e740572009-09-15 20:02:33 -0400593 goto found;
Chris Mason90428462009-08-04 16:56:34 -0400594
595fallback:
596 fallback = NULL;
597 /*
598 * we have failed to find any workers, just
599 * return the first one we can find.
600 */
601 if (!list_empty(&workers->worker_list))
602 fallback = workers->worker_list.next;
603 if (!list_empty(&workers->idle_list))
604 fallback = workers->idle_list.next;
605 BUG_ON(!fallback);
606 worker = list_entry(fallback,
607 struct btrfs_worker_thread, worker_list);
Chris Mason6e740572009-09-15 20:02:33 -0400608found:
609 /*
610 * this makes sure the worker doesn't exit before it is placed
611 * onto a busy/idle list
612 */
613 atomic_inc(&worker->num_pending);
Chris Mason90428462009-08-04 16:56:34 -0400614 spin_unlock_irqrestore(&workers->lock, flags);
615 return worker;
Chris Mason8b712842008-06-11 16:50:36 -0400616}
617
618/*
619 * btrfs_requeue_work just puts the work item back on the tail of the list
620 * it was taken from. It is intended for use with long running work functions
621 * that make some progress and want to give the cpu up for others.
622 */
Jeff Mahoney143bede2012-03-01 14:56:26 +0100623void btrfs_requeue_work(struct btrfs_work *work)
Chris Mason8b712842008-06-11 16:50:36 -0400624{
625 struct btrfs_worker_thread *worker = work->worker;
626 unsigned long flags;
Chris Masona6837052009-02-04 09:19:41 -0500627 int wake = 0;
Chris Mason8b712842008-06-11 16:50:36 -0400628
Chris Mason4a69a412008-11-06 22:03:00 -0500629 if (test_and_set_bit(WORK_QUEUED_BIT, &work->flags))
Jeff Mahoney143bede2012-03-01 14:56:26 +0100630 return;
Chris Mason8b712842008-06-11 16:50:36 -0400631
632 spin_lock_irqsave(&worker->lock, flags);
Chris Masond313d7a2009-04-20 15:50:09 -0400633 if (test_bit(WORK_HIGH_PRIO_BIT, &work->flags))
634 list_add_tail(&work->list, &worker->prio_pending);
635 else
636 list_add_tail(&work->list, &worker->pending);
Chris Masonb51912c2009-02-04 09:23:24 -0500637 atomic_inc(&worker->num_pending);
Chris Mason75ccf472008-09-30 19:24:06 -0400638
639 /* by definition we're busy, take ourselves off the idle
640 * list
641 */
642 if (worker->idle) {
Julia Lawall29c5e8c2009-07-22 16:49:00 -0400643 spin_lock(&worker->workers->lock);
Chris Mason75ccf472008-09-30 19:24:06 -0400644 worker->idle = 0;
645 list_move_tail(&worker->worker_list,
Chris Mason6e740572009-09-15 20:02:33 -0400646 &worker->workers->worker_list);
Julia Lawall29c5e8c2009-07-22 16:49:00 -0400647 spin_unlock(&worker->workers->lock);
Chris Mason75ccf472008-09-30 19:24:06 -0400648 }
Chris Masona6837052009-02-04 09:19:41 -0500649 if (!worker->working) {
650 wake = 1;
651 worker->working = 1;
652 }
Chris Mason75ccf472008-09-30 19:24:06 -0400653
Chris Masona6837052009-02-04 09:19:41 -0500654 if (wake)
655 wake_up_process(worker->task);
Chris Mason90428462009-08-04 16:56:34 -0400656 spin_unlock_irqrestore(&worker->lock, flags);
Chris Mason8b712842008-06-11 16:50:36 -0400657}
658
Chris Masond313d7a2009-04-20 15:50:09 -0400659void btrfs_set_work_high_prio(struct btrfs_work *work)
660{
661 set_bit(WORK_HIGH_PRIO_BIT, &work->flags);
662}
663
Chris Mason8b712842008-06-11 16:50:36 -0400664/*
665 * places a struct btrfs_work into the pending queue of one of the kthreads
666 */
Josef Bacik0dc3b842011-11-18 14:37:27 -0500667void btrfs_queue_worker(struct btrfs_workers *workers, struct btrfs_work *work)
Chris Mason8b712842008-06-11 16:50:36 -0400668{
669 struct btrfs_worker_thread *worker;
670 unsigned long flags;
671 int wake = 0;
672
673 /* don't requeue something already on a list */
Chris Mason4a69a412008-11-06 22:03:00 -0500674 if (test_and_set_bit(WORK_QUEUED_BIT, &work->flags))
Josef Bacik0dc3b842011-11-18 14:37:27 -0500675 return;
Chris Mason8b712842008-06-11 16:50:36 -0400676
677 worker = find_worker(workers);
Chris Mason4a69a412008-11-06 22:03:00 -0500678 if (workers->ordered) {
Chris Mason4e3f9c52009-08-05 16:36:45 -0400679 /*
680 * you're not allowed to do ordered queues from an
681 * interrupt handler
682 */
683 spin_lock(&workers->order_lock);
Chris Masond313d7a2009-04-20 15:50:09 -0400684 if (test_bit(WORK_HIGH_PRIO_BIT, &work->flags)) {
685 list_add_tail(&work->order_list,
686 &workers->prio_order_list);
687 } else {
688 list_add_tail(&work->order_list, &workers->order_list);
689 }
Chris Mason4e3f9c52009-08-05 16:36:45 -0400690 spin_unlock(&workers->order_lock);
Chris Mason4a69a412008-11-06 22:03:00 -0500691 } else {
692 INIT_LIST_HEAD(&work->order_list);
693 }
Chris Mason8b712842008-06-11 16:50:36 -0400694
695 spin_lock_irqsave(&worker->lock, flags);
Chris Masona6837052009-02-04 09:19:41 -0500696
Chris Masond313d7a2009-04-20 15:50:09 -0400697 if (test_bit(WORK_HIGH_PRIO_BIT, &work->flags))
698 list_add_tail(&work->list, &worker->prio_pending);
699 else
700 list_add_tail(&work->list, &worker->pending);
Chris Mason35d8ba62008-06-11 20:21:24 -0400701 check_busy_worker(worker);
Chris Mason8b712842008-06-11 16:50:36 -0400702
703 /*
704 * avoid calling into wake_up_process if this thread has already
705 * been kicked
706 */
707 if (!worker->working)
708 wake = 1;
709 worker->working = 1;
710
Chris Mason8b712842008-06-11 16:50:36 -0400711 if (wake)
712 wake_up_process(worker->task);
Chris Mason90428462009-08-04 16:56:34 -0400713 spin_unlock_irqrestore(&worker->lock, flags);
Chris Mason8b712842008-06-11 16:50:36 -0400714}