blob: c0861e781cdbdfde4b7c42ec1b90032f2059e69a [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>
20#include <linux/list.h>
21#include <linux/spinlock.h>
Chris Masonb51912c2009-02-04 09:23:24 -050022#include <linux/freezer.h>
Chris Mason8b712842008-06-11 16:50:36 -040023#include "async-thread.h"
24
Chris Mason4a69a412008-11-06 22:03:00 -050025#define WORK_QUEUED_BIT 0
26#define WORK_DONE_BIT 1
27#define WORK_ORDER_DONE_BIT 2
Chris Masond313d7a2009-04-20 15:50:09 -040028#define WORK_HIGH_PRIO_BIT 3
Chris Mason4a69a412008-11-06 22:03:00 -050029
Chris Mason8b712842008-06-11 16:50:36 -040030/*
31 * container for the kthread task pointer and the list of pending work
32 * One of these is allocated per thread.
33 */
34struct btrfs_worker_thread {
Chris Mason35d8ba62008-06-11 20:21:24 -040035 /* pool we belong to */
36 struct btrfs_workers *workers;
37
Chris Mason8b712842008-06-11 16:50:36 -040038 /* list of struct btrfs_work that are waiting for service */
39 struct list_head pending;
Chris Masond313d7a2009-04-20 15:50:09 -040040 struct list_head prio_pending;
Chris Mason8b712842008-06-11 16:50:36 -040041
42 /* list of worker threads from struct btrfs_workers */
43 struct list_head worker_list;
44
45 /* kthread */
46 struct task_struct *task;
47
48 /* number of things on the pending list */
49 atomic_t num_pending;
Chris Mason53863232008-08-15 15:34:18 -040050
Chris Mason90428462009-08-04 16:56:34 -040051 /* reference counter for this struct */
52 atomic_t refs;
53
Chris Mason4854ddd2008-08-15 15:34:17 -040054 unsigned long sequence;
Chris Mason8b712842008-06-11 16:50:36 -040055
56 /* protects the pending list. */
57 spinlock_t lock;
58
59 /* set to non-zero when this thread is already awake and kicking */
60 int working;
Chris Mason35d8ba62008-06-11 20:21:24 -040061
62 /* are we currently idle */
63 int idle;
Chris Mason8b712842008-06-11 16:50:36 -040064};
65
66/*
Chris Mason61d92c32009-10-02 19:11:56 -040067 * btrfs_start_workers uses kthread_run, which can block waiting for memory
68 * for a very long time. It will actually throttle on page writeback,
69 * and so it may not make progress until after our btrfs worker threads
70 * process all of the pending work structs in their queue
71 *
72 * This means we can't use btrfs_start_workers from inside a btrfs worker
73 * thread that is used as part of cleaning dirty memory, which pretty much
74 * involves all of the worker threads.
75 *
76 * Instead we have a helper queue who never has more than one thread
77 * where we scheduler thread start operations. This worker_start struct
78 * is used to contain the work and hold a pointer to the queue that needs
79 * another worker.
80 */
81struct worker_start {
82 struct btrfs_work work;
83 struct btrfs_workers *queue;
84};
85
86static void start_new_worker_func(struct btrfs_work *work)
87{
88 struct worker_start *start;
89 start = container_of(work, struct worker_start, work);
90 btrfs_start_workers(start->queue, 1);
91 kfree(start);
92}
93
94static int start_new_worker(struct btrfs_workers *queue)
95{
96 struct worker_start *start;
97 int ret;
98
99 start = kzalloc(sizeof(*start), GFP_NOFS);
100 if (!start)
101 return -ENOMEM;
102
103 start->work.func = start_new_worker_func;
104 start->queue = queue;
105 ret = btrfs_queue_worker(queue->atomic_worker_start, &start->work);
106 if (ret)
107 kfree(start);
108 return ret;
109}
110
111/*
Chris Mason35d8ba62008-06-11 20:21:24 -0400112 * helper function to move a thread onto the idle list after it
113 * has finished some requests.
114 */
115static void check_idle_worker(struct btrfs_worker_thread *worker)
116{
117 if (!worker->idle && atomic_read(&worker->num_pending) <
118 worker->workers->idle_thresh / 2) {
119 unsigned long flags;
120 spin_lock_irqsave(&worker->workers->lock, flags);
121 worker->idle = 1;
Chris Mason3e99d8e2009-09-15 19:57:42 -0400122
123 /* the list may be empty if the worker is just starting */
124 if (!list_empty(&worker->worker_list)) {
125 list_move(&worker->worker_list,
126 &worker->workers->idle_list);
127 }
Chris Mason35d8ba62008-06-11 20:21:24 -0400128 spin_unlock_irqrestore(&worker->workers->lock, flags);
129 }
130}
131
132/*
133 * helper function to move a thread off the idle list after new
134 * pending work is added.
135 */
136static void check_busy_worker(struct btrfs_worker_thread *worker)
137{
138 if (worker->idle && atomic_read(&worker->num_pending) >=
139 worker->workers->idle_thresh) {
140 unsigned long flags;
141 spin_lock_irqsave(&worker->workers->lock, flags);
142 worker->idle = 0;
Chris Mason3e99d8e2009-09-15 19:57:42 -0400143
144 if (!list_empty(&worker->worker_list)) {
145 list_move_tail(&worker->worker_list,
146 &worker->workers->worker_list);
147 }
Chris Mason35d8ba62008-06-11 20:21:24 -0400148 spin_unlock_irqrestore(&worker->workers->lock, flags);
149 }
150}
151
Chris Mason90428462009-08-04 16:56:34 -0400152static void check_pending_worker_creates(struct btrfs_worker_thread *worker)
153{
154 struct btrfs_workers *workers = worker->workers;
155 unsigned long flags;
156
157 rmb();
158 if (!workers->atomic_start_pending)
159 return;
160
161 spin_lock_irqsave(&workers->lock, flags);
162 if (!workers->atomic_start_pending)
163 goto out;
164
165 workers->atomic_start_pending = 0;
Chris Mason61d92c32009-10-02 19:11:56 -0400166 if (workers->num_workers + workers->num_workers_starting >=
167 workers->max_workers)
Chris Mason90428462009-08-04 16:56:34 -0400168 goto out;
169
Chris Mason61d92c32009-10-02 19:11:56 -0400170 workers->num_workers_starting += 1;
Chris Mason90428462009-08-04 16:56:34 -0400171 spin_unlock_irqrestore(&workers->lock, flags);
Chris Mason61d92c32009-10-02 19:11:56 -0400172 start_new_worker(workers);
Chris Mason90428462009-08-04 16:56:34 -0400173 return;
174
175out:
176 spin_unlock_irqrestore(&workers->lock, flags);
177}
178
Chris Mason4a69a412008-11-06 22:03:00 -0500179static noinline int run_ordered_completions(struct btrfs_workers *workers,
180 struct btrfs_work *work)
181{
Chris Mason4a69a412008-11-06 22:03:00 -0500182 if (!workers->ordered)
183 return 0;
184
185 set_bit(WORK_DONE_BIT, &work->flags);
186
Chris Mason4e3f9c52009-08-05 16:36:45 -0400187 spin_lock(&workers->order_lock);
Chris Mason4a69a412008-11-06 22:03:00 -0500188
Chris Masond313d7a2009-04-20 15:50:09 -0400189 while (1) {
190 if (!list_empty(&workers->prio_order_list)) {
191 work = list_entry(workers->prio_order_list.next,
192 struct btrfs_work, order_list);
193 } else if (!list_empty(&workers->order_list)) {
194 work = list_entry(workers->order_list.next,
195 struct btrfs_work, order_list);
196 } else {
197 break;
198 }
Chris Mason4a69a412008-11-06 22:03:00 -0500199 if (!test_bit(WORK_DONE_BIT, &work->flags))
200 break;
201
202 /* we are going to call the ordered done function, but
203 * we leave the work item on the list as a barrier so
204 * that later work items that are done don't have their
205 * functions called before this one returns
206 */
207 if (test_and_set_bit(WORK_ORDER_DONE_BIT, &work->flags))
208 break;
209
Chris Mason4e3f9c52009-08-05 16:36:45 -0400210 spin_unlock(&workers->order_lock);
Chris Mason4a69a412008-11-06 22:03:00 -0500211
212 work->ordered_func(work);
213
214 /* now take the lock again and call the freeing code */
Chris Mason4e3f9c52009-08-05 16:36:45 -0400215 spin_lock(&workers->order_lock);
Chris Mason4a69a412008-11-06 22:03:00 -0500216 list_del(&work->order_list);
217 work->ordered_free(work);
218 }
219
Chris Mason4e3f9c52009-08-05 16:36:45 -0400220 spin_unlock(&workers->order_lock);
Chris Mason4a69a412008-11-06 22:03:00 -0500221 return 0;
222}
223
Chris Mason90428462009-08-04 16:56:34 -0400224static void put_worker(struct btrfs_worker_thread *worker)
225{
226 if (atomic_dec_and_test(&worker->refs))
227 kfree(worker);
228}
229
230static int try_worker_shutdown(struct btrfs_worker_thread *worker)
231{
232 int freeit = 0;
233
234 spin_lock_irq(&worker->lock);
Chris Mason627e4212009-09-15 20:00:36 -0400235 spin_lock(&worker->workers->lock);
Chris Mason90428462009-08-04 16:56:34 -0400236 if (worker->workers->num_workers > 1 &&
237 worker->idle &&
238 !worker->working &&
239 !list_empty(&worker->worker_list) &&
240 list_empty(&worker->prio_pending) &&
Chris Mason6e740572009-09-15 20:02:33 -0400241 list_empty(&worker->pending) &&
242 atomic_read(&worker->num_pending) == 0) {
Chris Mason90428462009-08-04 16:56:34 -0400243 freeit = 1;
244 list_del_init(&worker->worker_list);
245 worker->workers->num_workers--;
246 }
Chris Mason627e4212009-09-15 20:00:36 -0400247 spin_unlock(&worker->workers->lock);
Chris Mason90428462009-08-04 16:56:34 -0400248 spin_unlock_irq(&worker->lock);
249
250 if (freeit)
251 put_worker(worker);
252 return freeit;
253}
254
Chris Mason4f878e82009-08-07 09:27:38 -0400255static struct btrfs_work *get_next_work(struct btrfs_worker_thread *worker,
256 struct list_head *prio_head,
257 struct list_head *head)
258{
259 struct btrfs_work *work = NULL;
260 struct list_head *cur = NULL;
261
262 if(!list_empty(prio_head))
263 cur = prio_head->next;
264
265 smp_mb();
266 if (!list_empty(&worker->prio_pending))
267 goto refill;
268
269 if (!list_empty(head))
270 cur = head->next;
271
272 if (cur)
273 goto out;
274
275refill:
276 spin_lock_irq(&worker->lock);
277 list_splice_tail_init(&worker->prio_pending, prio_head);
278 list_splice_tail_init(&worker->pending, head);
279
280 if (!list_empty(prio_head))
281 cur = prio_head->next;
282 else if (!list_empty(head))
283 cur = head->next;
284 spin_unlock_irq(&worker->lock);
285
286 if (!cur)
287 goto out_fail;
288
289out:
290 work = list_entry(cur, struct btrfs_work, list);
291
292out_fail:
293 return work;
294}
295
Chris Mason35d8ba62008-06-11 20:21:24 -0400296/*
Chris Mason8b712842008-06-11 16:50:36 -0400297 * main loop for servicing work items
298 */
299static int worker_loop(void *arg)
300{
301 struct btrfs_worker_thread *worker = arg;
Chris Mason4f878e82009-08-07 09:27:38 -0400302 struct list_head head;
303 struct list_head prio_head;
Chris Mason8b712842008-06-11 16:50:36 -0400304 struct btrfs_work *work;
Chris Mason4f878e82009-08-07 09:27:38 -0400305
306 INIT_LIST_HEAD(&head);
307 INIT_LIST_HEAD(&prio_head);
308
Chris Mason8b712842008-06-11 16:50:36 -0400309 do {
Chris Mason4f878e82009-08-07 09:27:38 -0400310again:
Chris Masond313d7a2009-04-20 15:50:09 -0400311 while (1) {
Chris Mason4f878e82009-08-07 09:27:38 -0400312
313
314 work = get_next_work(worker, &prio_head, &head);
315 if (!work)
Chris Masond313d7a2009-04-20 15:50:09 -0400316 break;
317
Chris Mason8b712842008-06-11 16:50:36 -0400318 list_del(&work->list);
Chris Mason4a69a412008-11-06 22:03:00 -0500319 clear_bit(WORK_QUEUED_BIT, &work->flags);
Chris Mason8b712842008-06-11 16:50:36 -0400320
321 work->worker = worker;
Chris Mason8b712842008-06-11 16:50:36 -0400322
323 work->func(work);
324
325 atomic_dec(&worker->num_pending);
Chris Mason4a69a412008-11-06 22:03:00 -0500326 /*
327 * unless this is an ordered work queue,
328 * 'work' was probably freed by func above.
329 */
330 run_ordered_completions(worker->workers, work);
331
Chris Mason90428462009-08-04 16:56:34 -0400332 check_pending_worker_creates(worker);
333
Chris Mason8b712842008-06-11 16:50:36 -0400334 }
Chris Mason4f878e82009-08-07 09:27:38 -0400335
336 spin_lock_irq(&worker->lock);
337 check_idle_worker(worker);
338
Chris Mason8b712842008-06-11 16:50:36 -0400339 if (freezing(current)) {
Chris Masonb51912c2009-02-04 09:23:24 -0500340 worker->working = 0;
341 spin_unlock_irq(&worker->lock);
Chris Mason8b712842008-06-11 16:50:36 -0400342 refrigerator();
343 } else {
Chris Mason8b712842008-06-11 16:50:36 -0400344 spin_unlock_irq(&worker->lock);
Chris Masonb51912c2009-02-04 09:23:24 -0500345 if (!kthread_should_stop()) {
346 cpu_relax();
347 /*
348 * we've dropped the lock, did someone else
349 * jump_in?
350 */
351 smp_mb();
Chris Masond313d7a2009-04-20 15:50:09 -0400352 if (!list_empty(&worker->pending) ||
353 !list_empty(&worker->prio_pending))
Chris Masonb51912c2009-02-04 09:23:24 -0500354 continue;
355
356 /*
357 * this short schedule allows more work to
358 * come in without the queue functions
359 * needing to go through wake_up_process()
360 *
361 * worker->working is still 1, so nobody
362 * is going to try and wake us up
363 */
364 schedule_timeout(1);
365 smp_mb();
Chris Masond313d7a2009-04-20 15:50:09 -0400366 if (!list_empty(&worker->pending) ||
367 !list_empty(&worker->prio_pending))
Chris Masonb51912c2009-02-04 09:23:24 -0500368 continue;
369
Amit Gudb5555f72009-04-02 17:01:27 -0400370 if (kthread_should_stop())
371 break;
372
Chris Masonb51912c2009-02-04 09:23:24 -0500373 /* still no more work?, sleep for real */
374 spin_lock_irq(&worker->lock);
375 set_current_state(TASK_INTERRUPTIBLE);
Chris Masond313d7a2009-04-20 15:50:09 -0400376 if (!list_empty(&worker->pending) ||
Chris Mason4f878e82009-08-07 09:27:38 -0400377 !list_empty(&worker->prio_pending)) {
378 spin_unlock_irq(&worker->lock);
379 goto again;
380 }
Chris Masonb51912c2009-02-04 09:23:24 -0500381
382 /*
383 * this makes sure we get a wakeup when someone
384 * adds something new to the queue
385 */
386 worker->working = 0;
387 spin_unlock_irq(&worker->lock);
388
Chris Mason90428462009-08-04 16:56:34 -0400389 if (!kthread_should_stop()) {
390 schedule_timeout(HZ * 120);
391 if (!worker->working &&
392 try_worker_shutdown(worker)) {
393 return 0;
394 }
395 }
Chris Masonb51912c2009-02-04 09:23:24 -0500396 }
Chris Mason8b712842008-06-11 16:50:36 -0400397 __set_current_state(TASK_RUNNING);
398 }
399 } while (!kthread_should_stop());
400 return 0;
401}
402
403/*
404 * this will wait for all the worker threads to shutdown
405 */
406int btrfs_stop_workers(struct btrfs_workers *workers)
407{
408 struct list_head *cur;
409 struct btrfs_worker_thread *worker;
Chris Mason90428462009-08-04 16:56:34 -0400410 int can_stop;
Chris Mason8b712842008-06-11 16:50:36 -0400411
Chris Mason90428462009-08-04 16:56:34 -0400412 spin_lock_irq(&workers->lock);
Chris Mason35d8ba62008-06-11 20:21:24 -0400413 list_splice_init(&workers->idle_list, &workers->worker_list);
Chris Masond3977122009-01-05 21:25:51 -0500414 while (!list_empty(&workers->worker_list)) {
Chris Mason8b712842008-06-11 16:50:36 -0400415 cur = workers->worker_list.next;
416 worker = list_entry(cur, struct btrfs_worker_thread,
417 worker_list);
Chris Mason90428462009-08-04 16:56:34 -0400418
419 atomic_inc(&worker->refs);
420 workers->num_workers -= 1;
421 if (!list_empty(&worker->worker_list)) {
422 list_del_init(&worker->worker_list);
423 put_worker(worker);
424 can_stop = 1;
425 } else
426 can_stop = 0;
427 spin_unlock_irq(&workers->lock);
428 if (can_stop)
429 kthread_stop(worker->task);
430 spin_lock_irq(&workers->lock);
431 put_worker(worker);
Chris Mason8b712842008-06-11 16:50:36 -0400432 }
Chris Mason90428462009-08-04 16:56:34 -0400433 spin_unlock_irq(&workers->lock);
Chris Mason8b712842008-06-11 16:50:36 -0400434 return 0;
435}
436
437/*
438 * simple init on struct btrfs_workers
439 */
Chris Mason61d92c32009-10-02 19:11:56 -0400440void btrfs_init_workers(struct btrfs_workers *workers, char *name, int max,
441 struct btrfs_workers *async_helper)
Chris Mason8b712842008-06-11 16:50:36 -0400442{
443 workers->num_workers = 0;
Chris Mason61d92c32009-10-02 19:11:56 -0400444 workers->num_workers_starting = 0;
Chris Mason8b712842008-06-11 16:50:36 -0400445 INIT_LIST_HEAD(&workers->worker_list);
Chris Mason35d8ba62008-06-11 20:21:24 -0400446 INIT_LIST_HEAD(&workers->idle_list);
Chris Mason4a69a412008-11-06 22:03:00 -0500447 INIT_LIST_HEAD(&workers->order_list);
Chris Masond313d7a2009-04-20 15:50:09 -0400448 INIT_LIST_HEAD(&workers->prio_order_list);
Chris Mason8b712842008-06-11 16:50:36 -0400449 spin_lock_init(&workers->lock);
Chris Mason4e3f9c52009-08-05 16:36:45 -0400450 spin_lock_init(&workers->order_lock);
Chris Mason8b712842008-06-11 16:50:36 -0400451 workers->max_workers = max;
Chris Mason61b49442008-07-31 15:42:53 -0400452 workers->idle_thresh = 32;
Chris Mason5443be42008-08-15 15:34:16 -0400453 workers->name = name;
Chris Mason4a69a412008-11-06 22:03:00 -0500454 workers->ordered = 0;
Chris Mason90428462009-08-04 16:56:34 -0400455 workers->atomic_start_pending = 0;
Chris Mason61d92c32009-10-02 19:11:56 -0400456 workers->atomic_worker_start = async_helper;
Chris Mason8b712842008-06-11 16:50:36 -0400457}
458
459/*
460 * starts new worker threads. This does not enforce the max worker
461 * count in case you need to temporarily go past it.
462 */
Chris Mason61d92c32009-10-02 19:11:56 -0400463static int __btrfs_start_workers(struct btrfs_workers *workers,
464 int num_workers)
Chris Mason8b712842008-06-11 16:50:36 -0400465{
466 struct btrfs_worker_thread *worker;
467 int ret = 0;
468 int i;
469
470 for (i = 0; i < num_workers; i++) {
471 worker = kzalloc(sizeof(*worker), GFP_NOFS);
472 if (!worker) {
473 ret = -ENOMEM;
474 goto fail;
475 }
476
477 INIT_LIST_HEAD(&worker->pending);
Chris Masond313d7a2009-04-20 15:50:09 -0400478 INIT_LIST_HEAD(&worker->prio_pending);
Chris Mason8b712842008-06-11 16:50:36 -0400479 INIT_LIST_HEAD(&worker->worker_list);
480 spin_lock_init(&worker->lock);
Chris Mason4e3f9c52009-08-05 16:36:45 -0400481
Chris Mason8b712842008-06-11 16:50:36 -0400482 atomic_set(&worker->num_pending, 0);
Chris Mason90428462009-08-04 16:56:34 -0400483 atomic_set(&worker->refs, 1);
Shin Hongfd0fb032009-06-10 20:11:29 -0400484 worker->workers = workers;
Chris Mason5443be42008-08-15 15:34:16 -0400485 worker->task = kthread_run(worker_loop, worker,
486 "btrfs-%s-%d", workers->name,
487 workers->num_workers + i);
Chris Mason8b712842008-06-11 16:50:36 -0400488 if (IS_ERR(worker->task)) {
489 ret = PTR_ERR(worker->task);
Jiri Slaby9b627e92009-07-02 13:50:58 -0400490 kfree(worker);
Chris Mason8b712842008-06-11 16:50:36 -0400491 goto fail;
492 }
Chris Mason8b712842008-06-11 16:50:36 -0400493 spin_lock_irq(&workers->lock);
Chris Mason35d8ba62008-06-11 20:21:24 -0400494 list_add_tail(&worker->worker_list, &workers->idle_list);
Chris Mason4854ddd2008-08-15 15:34:17 -0400495 worker->idle = 1;
Chris Mason8b712842008-06-11 16:50:36 -0400496 workers->num_workers++;
Chris Mason61d92c32009-10-02 19:11:56 -0400497 workers->num_workers_starting--;
498 WARN_ON(workers->num_workers_starting < 0);
Chris Mason8b712842008-06-11 16:50:36 -0400499 spin_unlock_irq(&workers->lock);
500 }
501 return 0;
502fail:
503 btrfs_stop_workers(workers);
504 return ret;
505}
506
Chris Mason61d92c32009-10-02 19:11:56 -0400507int btrfs_start_workers(struct btrfs_workers *workers, int num_workers)
508{
509 spin_lock_irq(&workers->lock);
510 workers->num_workers_starting += num_workers;
511 spin_unlock_irq(&workers->lock);
512 return __btrfs_start_workers(workers, num_workers);
513}
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;
Chris Mason8b712842008-06-11 16:50:36 -0400569
570again:
571 spin_lock_irqsave(&workers->lock, flags);
572 worker = next_worker(workers);
Chris Mason8b712842008-06-11 16:50:36 -0400573
574 if (!worker) {
Chris Mason61d92c32009-10-02 19:11:56 -0400575 if (workers->num_workers + workers->num_workers_starting >=
576 workers->max_workers) {
Chris Mason90428462009-08-04 16:56:34 -0400577 goto fallback;
578 } else if (workers->atomic_worker_start) {
579 workers->atomic_start_pending = 1;
580 goto fallback;
Chris Mason8b712842008-06-11 16:50:36 -0400581 } else {
Chris Mason61d92c32009-10-02 19:11:56 -0400582 workers->num_workers_starting++;
Chris Mason8b712842008-06-11 16:50:36 -0400583 spin_unlock_irqrestore(&workers->lock, flags);
584 /* we're below the limit, start another worker */
Chris Mason61d92c32009-10-02 19:11:56 -0400585 __btrfs_start_workers(workers, 1);
Chris Mason8b712842008-06-11 16:50:36 -0400586 goto again;
587 }
588 }
Chris Mason6e740572009-09-15 20:02:33 -0400589 goto found;
Chris Mason90428462009-08-04 16:56:34 -0400590
591fallback:
592 fallback = NULL;
593 /*
594 * we have failed to find any workers, just
595 * return the first one we can find.
596 */
597 if (!list_empty(&workers->worker_list))
598 fallback = workers->worker_list.next;
599 if (!list_empty(&workers->idle_list))
600 fallback = workers->idle_list.next;
601 BUG_ON(!fallback);
602 worker = list_entry(fallback,
603 struct btrfs_worker_thread, worker_list);
Chris Mason6e740572009-09-15 20:02:33 -0400604found:
605 /*
606 * this makes sure the worker doesn't exit before it is placed
607 * onto a busy/idle list
608 */
609 atomic_inc(&worker->num_pending);
Chris Mason90428462009-08-04 16:56:34 -0400610 spin_unlock_irqrestore(&workers->lock, flags);
611 return worker;
Chris Mason8b712842008-06-11 16:50:36 -0400612}
613
614/*
615 * btrfs_requeue_work just puts the work item back on the tail of the list
616 * it was taken from. It is intended for use with long running work functions
617 * that make some progress and want to give the cpu up for others.
618 */
619int btrfs_requeue_work(struct btrfs_work *work)
620{
621 struct btrfs_worker_thread *worker = work->worker;
622 unsigned long flags;
Chris Masona6837052009-02-04 09:19:41 -0500623 int wake = 0;
Chris Mason8b712842008-06-11 16:50:36 -0400624
Chris Mason4a69a412008-11-06 22:03:00 -0500625 if (test_and_set_bit(WORK_QUEUED_BIT, &work->flags))
Chris Mason8b712842008-06-11 16:50:36 -0400626 goto out;
627
628 spin_lock_irqsave(&worker->lock, flags);
Chris Masond313d7a2009-04-20 15:50:09 -0400629 if (test_bit(WORK_HIGH_PRIO_BIT, &work->flags))
630 list_add_tail(&work->list, &worker->prio_pending);
631 else
632 list_add_tail(&work->list, &worker->pending);
Chris Masonb51912c2009-02-04 09:23:24 -0500633 atomic_inc(&worker->num_pending);
Chris Mason75ccf472008-09-30 19:24:06 -0400634
635 /* by definition we're busy, take ourselves off the idle
636 * list
637 */
638 if (worker->idle) {
Julia Lawall29c5e8c2009-07-22 16:49:00 -0400639 spin_lock(&worker->workers->lock);
Chris Mason75ccf472008-09-30 19:24:06 -0400640 worker->idle = 0;
641 list_move_tail(&worker->worker_list,
Chris Mason6e740572009-09-15 20:02:33 -0400642 &worker->workers->worker_list);
Julia Lawall29c5e8c2009-07-22 16:49:00 -0400643 spin_unlock(&worker->workers->lock);
Chris Mason75ccf472008-09-30 19:24:06 -0400644 }
Chris Masona6837052009-02-04 09:19:41 -0500645 if (!worker->working) {
646 wake = 1;
647 worker->working = 1;
648 }
Chris Mason75ccf472008-09-30 19:24:06 -0400649
Chris Masona6837052009-02-04 09:19:41 -0500650 if (wake)
651 wake_up_process(worker->task);
Chris Mason90428462009-08-04 16:56:34 -0400652 spin_unlock_irqrestore(&worker->lock, flags);
Chris Mason8b712842008-06-11 16:50:36 -0400653out:
Chris Masona6837052009-02-04 09:19:41 -0500654
Chris Mason8b712842008-06-11 16:50:36 -0400655 return 0;
656}
657
Chris Masond313d7a2009-04-20 15:50:09 -0400658void btrfs_set_work_high_prio(struct btrfs_work *work)
659{
660 set_bit(WORK_HIGH_PRIO_BIT, &work->flags);
661}
662
Chris Mason8b712842008-06-11 16:50:36 -0400663/*
664 * places a struct btrfs_work into the pending queue of one of the kthreads
665 */
666int btrfs_queue_worker(struct btrfs_workers *workers, struct btrfs_work *work)
667{
668 struct btrfs_worker_thread *worker;
669 unsigned long flags;
670 int wake = 0;
671
672 /* don't requeue something already on a list */
Chris Mason4a69a412008-11-06 22:03:00 -0500673 if (test_and_set_bit(WORK_QUEUED_BIT, &work->flags))
Chris Mason8b712842008-06-11 16:50:36 -0400674 goto out;
675
676 worker = find_worker(workers);
Chris Mason4a69a412008-11-06 22:03:00 -0500677 if (workers->ordered) {
Chris Mason4e3f9c52009-08-05 16:36:45 -0400678 /*
679 * you're not allowed to do ordered queues from an
680 * interrupt handler
681 */
682 spin_lock(&workers->order_lock);
Chris Masond313d7a2009-04-20 15:50:09 -0400683 if (test_bit(WORK_HIGH_PRIO_BIT, &work->flags)) {
684 list_add_tail(&work->order_list,
685 &workers->prio_order_list);
686 } else {
687 list_add_tail(&work->order_list, &workers->order_list);
688 }
Chris Mason4e3f9c52009-08-05 16:36:45 -0400689 spin_unlock(&workers->order_lock);
Chris Mason4a69a412008-11-06 22:03:00 -0500690 } else {
691 INIT_LIST_HEAD(&work->order_list);
692 }
Chris Mason8b712842008-06-11 16:50:36 -0400693
694 spin_lock_irqsave(&worker->lock, flags);
Chris Masona6837052009-02-04 09:19:41 -0500695
Chris Masond313d7a2009-04-20 15:50:09 -0400696 if (test_bit(WORK_HIGH_PRIO_BIT, &work->flags))
697 list_add_tail(&work->list, &worker->prio_pending);
698 else
699 list_add_tail(&work->list, &worker->pending);
Chris Mason35d8ba62008-06-11 20:21:24 -0400700 check_busy_worker(worker);
Chris Mason8b712842008-06-11 16:50:36 -0400701
702 /*
703 * avoid calling into wake_up_process if this thread has already
704 * been kicked
705 */
706 if (!worker->working)
707 wake = 1;
708 worker->working = 1;
709
Chris Mason8b712842008-06-11 16:50:36 -0400710 if (wake)
711 wake_up_process(worker->task);
Chris Mason90428462009-08-04 16:56:34 -0400712 spin_unlock_irqrestore(&worker->lock, flags);
713
Chris Mason8b712842008-06-11 16:50:36 -0400714out:
715 return 0;
716}