Commit | Line | Data |
---|---|---|
8b712842 CM |
1 | /* |
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> | |
5a0e3ad6 | 20 | #include <linux/slab.h> |
8b712842 CM |
21 | #include <linux/list.h> |
22 | #include <linux/spinlock.h> | |
b51912c9 | 23 | #include <linux/freezer.h> |
8b712842 CM |
24 | #include "async-thread.h" |
25 | ||
4a69a410 CM |
26 | #define WORK_QUEUED_BIT 0 |
27 | #define WORK_DONE_BIT 1 | |
28 | #define WORK_ORDER_DONE_BIT 2 | |
d313d7a3 | 29 | #define WORK_HIGH_PRIO_BIT 3 |
4a69a410 | 30 | |
8b712842 CM |
31 | /* |
32 | * container for the kthread task pointer and the list of pending work | |
33 | * One of these is allocated per thread. | |
34 | */ | |
35 | struct btrfs_worker_thread { | |
35d8ba66 CM |
36 | /* pool we belong to */ |
37 | struct btrfs_workers *workers; | |
38 | ||
8b712842 CM |
39 | /* list of struct btrfs_work that are waiting for service */ |
40 | struct list_head pending; | |
d313d7a3 | 41 | struct list_head prio_pending; |
8b712842 CM |
42 | |
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; | |
53863232 | 51 | |
9042846b CM |
52 | /* reference counter for this struct */ |
53 | atomic_t refs; | |
54 | ||
4854ddd0 | 55 | unsigned long sequence; |
8b712842 CM |
56 | |
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; | |
35d8ba66 CM |
62 | |
63 | /* are we currently idle */ | |
64 | int idle; | |
8b712842 CM |
65 | }; |
66 | ||
0dc3b84a JB |
67 | static int __btrfs_start_workers(struct btrfs_workers *workers); |
68 | ||
61d92c32 CM |
69 | /* |
70 | * 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 | */ | |
84 | struct worker_start { | |
85 | struct btrfs_work work; | |
86 | struct btrfs_workers *queue; | |
87 | }; | |
88 | ||
89 | static void start_new_worker_func(struct btrfs_work *work) | |
90 | { | |
91 | struct worker_start *start; | |
92 | start = container_of(work, struct worker_start, work); | |
0dc3b84a | 93 | __btrfs_start_workers(start->queue); |
61d92c32 CM |
94 | kfree(start); |
95 | } | |
96 | ||
35d8ba66 CM |
97 | /* |
98 | * helper function to move a thread onto the idle list after it | |
99 | * has finished some requests. | |
100 | */ | |
101 | static 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; | |
3e99d8eb CM |
108 | |
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 | } | |
35d8ba66 CM |
114 | 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 | */ | |
122 | static 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; | |
3e99d8eb CM |
129 | |
130 | if (!list_empty(&worker->worker_list)) { | |
131 | list_move_tail(&worker->worker_list, | |
132 | &worker->workers->worker_list); | |
133 | } | |
35d8ba66 CM |
134 | spin_unlock_irqrestore(&worker->workers->lock, flags); |
135 | } | |
136 | } | |
137 | ||
9042846b CM |
138 | static void check_pending_worker_creates(struct btrfs_worker_thread *worker) |
139 | { | |
140 | struct btrfs_workers *workers = worker->workers; | |
0dc3b84a | 141 | struct worker_start *start; |
9042846b CM |
142 | unsigned long flags; |
143 | ||
144 | rmb(); | |
145 | if (!workers->atomic_start_pending) | |
146 | return; | |
147 | ||
0dc3b84a JB |
148 | 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 | ||
9042846b CM |
155 | spin_lock_irqsave(&workers->lock, flags); |
156 | if (!workers->atomic_start_pending) | |
157 | goto out; | |
158 | ||
159 | workers->atomic_start_pending = 0; | |
61d92c32 CM |
160 | if (workers->num_workers + workers->num_workers_starting >= |
161 | workers->max_workers) | |
9042846b CM |
162 | goto out; |
163 | ||
61d92c32 | 164 | workers->num_workers_starting += 1; |
9042846b | 165 | spin_unlock_irqrestore(&workers->lock, flags); |
0dc3b84a | 166 | btrfs_queue_worker(workers->atomic_worker_start, &start->work); |
9042846b CM |
167 | return; |
168 | ||
169 | out: | |
0dc3b84a | 170 | kfree(start); |
9042846b CM |
171 | spin_unlock_irqrestore(&workers->lock, flags); |
172 | } | |
173 | ||
143bede5 | 174 | static noinline void run_ordered_completions(struct btrfs_workers *workers, |
4a69a410 CM |
175 | struct btrfs_work *work) |
176 | { | |
4a69a410 | 177 | if (!workers->ordered) |
143bede5 | 178 | return; |
4a69a410 CM |
179 | |
180 | set_bit(WORK_DONE_BIT, &work->flags); | |
181 | ||
4e3f9c50 | 182 | spin_lock(&workers->order_lock); |
4a69a410 | 183 | |
d313d7a3 CM |
184 | 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 | } | |
4a69a410 CM |
194 | 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 | ||
4e3f9c50 | 205 | spin_unlock(&workers->order_lock); |
4a69a410 CM |
206 | |
207 | work->ordered_func(work); | |
208 | ||
209 | /* now take the lock again and call the freeing code */ | |
4e3f9c50 | 210 | spin_lock(&workers->order_lock); |
4a69a410 CM |
211 | list_del(&work->order_list); |
212 | work->ordered_free(work); | |
213 | } | |
214 | ||
4e3f9c50 | 215 | spin_unlock(&workers->order_lock); |
4a69a410 CM |
216 | } |
217 | ||
9042846b CM |
218 | static void put_worker(struct btrfs_worker_thread *worker) |
219 | { | |
220 | if (atomic_dec_and_test(&worker->refs)) | |
221 | kfree(worker); | |
222 | } | |
223 | ||
224 | static int try_worker_shutdown(struct btrfs_worker_thread *worker) | |
225 | { | |
226 | int freeit = 0; | |
227 | ||
228 | spin_lock_irq(&worker->lock); | |
627e421a | 229 | spin_lock(&worker->workers->lock); |
9042846b CM |
230 | if (worker->workers->num_workers > 1 && |
231 | worker->idle && | |
232 | !worker->working && | |
233 | !list_empty(&worker->worker_list) && | |
234 | list_empty(&worker->prio_pending) && | |
6e74057c CM |
235 | list_empty(&worker->pending) && |
236 | atomic_read(&worker->num_pending) == 0) { | |
9042846b CM |
237 | freeit = 1; |
238 | list_del_init(&worker->worker_list); | |
239 | worker->workers->num_workers--; | |
240 | } | |
627e421a | 241 | spin_unlock(&worker->workers->lock); |
9042846b CM |
242 | spin_unlock_irq(&worker->lock); |
243 | ||
244 | if (freeit) | |
245 | put_worker(worker); | |
246 | return freeit; | |
247 | } | |
248 | ||
4f878e84 CM |
249 | static struct btrfs_work *get_next_work(struct btrfs_worker_thread *worker, |
250 | struct list_head *prio_head, | |
251 | struct list_head *head) | |
252 | { | |
253 | struct btrfs_work *work = NULL; | |
254 | struct list_head *cur = NULL; | |
255 | ||
256 | if(!list_empty(prio_head)) | |
257 | cur = prio_head->next; | |
258 | ||
259 | smp_mb(); | |
260 | if (!list_empty(&worker->prio_pending)) | |
261 | goto refill; | |
262 | ||
263 | if (!list_empty(head)) | |
264 | cur = head->next; | |
265 | ||
266 | if (cur) | |
267 | goto out; | |
268 | ||
269 | refill: | |
270 | spin_lock_irq(&worker->lock); | |
271 | list_splice_tail_init(&worker->prio_pending, prio_head); | |
272 | list_splice_tail_init(&worker->pending, head); | |
273 | ||
274 | if (!list_empty(prio_head)) | |
275 | cur = prio_head->next; | |
276 | else if (!list_empty(head)) | |
277 | cur = head->next; | |
278 | spin_unlock_irq(&worker->lock); | |
279 | ||
280 | if (!cur) | |
281 | goto out_fail; | |
282 | ||
283 | out: | |
284 | work = list_entry(cur, struct btrfs_work, list); | |
285 | ||
286 | out_fail: | |
287 | return work; | |
288 | } | |
289 | ||
8b712842 CM |
290 | /* |
291 | * main loop for servicing work items | |
292 | */ | |
293 | static int worker_loop(void *arg) | |
294 | { | |
295 | struct btrfs_worker_thread *worker = arg; | |
4f878e84 CM |
296 | struct list_head head; |
297 | struct list_head prio_head; | |
8b712842 | 298 | struct btrfs_work *work; |
4f878e84 CM |
299 | |
300 | INIT_LIST_HEAD(&head); | |
301 | INIT_LIST_HEAD(&prio_head); | |
302 | ||
8b712842 | 303 | do { |
4f878e84 | 304 | again: |
d313d7a3 | 305 | while (1) { |
4f878e84 CM |
306 | |
307 | ||
308 | work = get_next_work(worker, &prio_head, &head); | |
309 | if (!work) | |
d313d7a3 CM |
310 | break; |
311 | ||
8b712842 | 312 | list_del(&work->list); |
4a69a410 | 313 | clear_bit(WORK_QUEUED_BIT, &work->flags); |
8b712842 CM |
314 | |
315 | work->worker = worker; | |
8b712842 CM |
316 | |
317 | work->func(work); | |
318 | ||
319 | atomic_dec(&worker->num_pending); | |
4a69a410 CM |
320 | /* |
321 | * unless this is an ordered work queue, | |
322 | * 'work' was probably freed by func above. | |
323 | */ | |
324 | run_ordered_completions(worker->workers, work); | |
325 | ||
9042846b | 326 | check_pending_worker_creates(worker); |
8f3b65a3 | 327 | cond_resched(); |
8b712842 | 328 | } |
4f878e84 CM |
329 | |
330 | spin_lock_irq(&worker->lock); | |
331 | check_idle_worker(worker); | |
332 | ||
8b712842 | 333 | if (freezing(current)) { |
b51912c9 CM |
334 | worker->working = 0; |
335 | spin_unlock_irq(&worker->lock); | |
a0acae0e | 336 | try_to_freeze(); |
8b712842 | 337 | } else { |
8b712842 | 338 | spin_unlock_irq(&worker->lock); |
b51912c9 CM |
339 | if (!kthread_should_stop()) { |
340 | cpu_relax(); | |
341 | /* | |
342 | * we've dropped the lock, did someone else | |
343 | * jump_in? | |
344 | */ | |
345 | smp_mb(); | |
d313d7a3 CM |
346 | if (!list_empty(&worker->pending) || |
347 | !list_empty(&worker->prio_pending)) | |
b51912c9 CM |
348 | continue; |
349 | ||
350 | /* | |
351 | * this short schedule allows more work to | |
352 | * come in without the queue functions | |
353 | * needing to go through wake_up_process() | |
354 | * | |
355 | * worker->working is still 1, so nobody | |
356 | * is going to try and wake us up | |
357 | */ | |
358 | schedule_timeout(1); | |
359 | smp_mb(); | |
d313d7a3 CM |
360 | if (!list_empty(&worker->pending) || |
361 | !list_empty(&worker->prio_pending)) | |
b51912c9 CM |
362 | continue; |
363 | ||
b5555f77 AG |
364 | if (kthread_should_stop()) |
365 | break; | |
366 | ||
b51912c9 CM |
367 | /* still no more work?, sleep for real */ |
368 | spin_lock_irq(&worker->lock); | |
369 | set_current_state(TASK_INTERRUPTIBLE); | |
d313d7a3 | 370 | if (!list_empty(&worker->pending) || |
4f878e84 CM |
371 | !list_empty(&worker->prio_pending)) { |
372 | spin_unlock_irq(&worker->lock); | |
ed3b3d31 | 373 | set_current_state(TASK_RUNNING); |
4f878e84 CM |
374 | goto again; |
375 | } | |
b51912c9 CM |
376 | |
377 | /* | |
378 | * this makes sure we get a wakeup when someone | |
379 | * adds something new to the queue | |
380 | */ | |
381 | worker->working = 0; | |
382 | spin_unlock_irq(&worker->lock); | |
383 | ||
9042846b CM |
384 | if (!kthread_should_stop()) { |
385 | schedule_timeout(HZ * 120); | |
386 | if (!worker->working && | |
387 | try_worker_shutdown(worker)) { | |
388 | return 0; | |
389 | } | |
390 | } | |
b51912c9 | 391 | } |
8b712842 CM |
392 | __set_current_state(TASK_RUNNING); |
393 | } | |
394 | } while (!kthread_should_stop()); | |
395 | return 0; | |
396 | } | |
397 | ||
398 | /* | |
399 | * this will wait for all the worker threads to shutdown | |
400 | */ | |
143bede5 | 401 | void btrfs_stop_workers(struct btrfs_workers *workers) |
8b712842 CM |
402 | { |
403 | struct list_head *cur; | |
404 | struct btrfs_worker_thread *worker; | |
9042846b | 405 | int can_stop; |
8b712842 | 406 | |
9042846b | 407 | spin_lock_irq(&workers->lock); |
35d8ba66 | 408 | list_splice_init(&workers->idle_list, &workers->worker_list); |
d397712b | 409 | while (!list_empty(&workers->worker_list)) { |
8b712842 CM |
410 | cur = workers->worker_list.next; |
411 | worker = list_entry(cur, struct btrfs_worker_thread, | |
412 | worker_list); | |
9042846b CM |
413 | |
414 | atomic_inc(&worker->refs); | |
415 | workers->num_workers -= 1; | |
416 | if (!list_empty(&worker->worker_list)) { | |
417 | list_del_init(&worker->worker_list); | |
418 | put_worker(worker); | |
419 | can_stop = 1; | |
420 | } else | |
421 | can_stop = 0; | |
422 | spin_unlock_irq(&workers->lock); | |
423 | if (can_stop) | |
424 | kthread_stop(worker->task); | |
425 | spin_lock_irq(&workers->lock); | |
426 | put_worker(worker); | |
8b712842 | 427 | } |
9042846b | 428 | spin_unlock_irq(&workers->lock); |
8b712842 CM |
429 | } |
430 | ||
431 | /* | |
432 | * simple init on struct btrfs_workers | |
433 | */ | |
61d92c32 CM |
434 | void btrfs_init_workers(struct btrfs_workers *workers, char *name, int max, |
435 | struct btrfs_workers *async_helper) | |
8b712842 CM |
436 | { |
437 | workers->num_workers = 0; | |
61d92c32 | 438 | workers->num_workers_starting = 0; |
8b712842 | 439 | INIT_LIST_HEAD(&workers->worker_list); |
35d8ba66 | 440 | INIT_LIST_HEAD(&workers->idle_list); |
4a69a410 | 441 | INIT_LIST_HEAD(&workers->order_list); |
d313d7a3 | 442 | INIT_LIST_HEAD(&workers->prio_order_list); |
8b712842 | 443 | spin_lock_init(&workers->lock); |
4e3f9c50 | 444 | spin_lock_init(&workers->order_lock); |
8b712842 | 445 | workers->max_workers = max; |
61b49440 | 446 | workers->idle_thresh = 32; |
5443be45 | 447 | workers->name = name; |
4a69a410 | 448 | workers->ordered = 0; |
9042846b | 449 | workers->atomic_start_pending = 0; |
61d92c32 | 450 | workers->atomic_worker_start = async_helper; |
8b712842 CM |
451 | } |
452 | ||
453 | /* | |
454 | * starts new worker threads. This does not enforce the max worker | |
455 | * count in case you need to temporarily go past it. | |
456 | */ | |
0dc3b84a | 457 | static int __btrfs_start_workers(struct btrfs_workers *workers) |
8b712842 CM |
458 | { |
459 | struct btrfs_worker_thread *worker; | |
460 | int ret = 0; | |
8b712842 | 461 | |
0dc3b84a JB |
462 | worker = kzalloc(sizeof(*worker), GFP_NOFS); |
463 | if (!worker) { | |
464 | ret = -ENOMEM; | |
465 | goto fail; | |
466 | } | |
8b712842 | 467 | |
0dc3b84a JB |
468 | INIT_LIST_HEAD(&worker->pending); |
469 | INIT_LIST_HEAD(&worker->prio_pending); | |
470 | INIT_LIST_HEAD(&worker->worker_list); | |
471 | spin_lock_init(&worker->lock); | |
472 | ||
473 | atomic_set(&worker->num_pending, 0); | |
474 | atomic_set(&worker->refs, 1); | |
475 | worker->workers = workers; | |
476 | worker->task = kthread_run(worker_loop, worker, | |
477 | "btrfs-%s-%d", workers->name, | |
478 | workers->num_workers + 1); | |
479 | if (IS_ERR(worker->task)) { | |
480 | ret = PTR_ERR(worker->task); | |
481 | kfree(worker); | |
482 | goto fail; | |
8b712842 | 483 | } |
0dc3b84a JB |
484 | spin_lock_irq(&workers->lock); |
485 | list_add_tail(&worker->worker_list, &workers->idle_list); | |
486 | worker->idle = 1; | |
487 | workers->num_workers++; | |
488 | workers->num_workers_starting--; | |
489 | WARN_ON(workers->num_workers_starting < 0); | |
490 | spin_unlock_irq(&workers->lock); | |
491 | ||
8b712842 CM |
492 | return 0; |
493 | fail: | |
0dc3b84a JB |
494 | spin_lock_irq(&workers->lock); |
495 | workers->num_workers_starting--; | |
496 | spin_unlock_irq(&workers->lock); | |
8b712842 CM |
497 | return ret; |
498 | } | |
499 | ||
0dc3b84a | 500 | int btrfs_start_workers(struct btrfs_workers *workers) |
61d92c32 CM |
501 | { |
502 | spin_lock_irq(&workers->lock); | |
0dc3b84a | 503 | workers->num_workers_starting++; |
61d92c32 | 504 | spin_unlock_irq(&workers->lock); |
0dc3b84a | 505 | return __btrfs_start_workers(workers); |
61d92c32 CM |
506 | } |
507 | ||
8b712842 CM |
508 | /* |
509 | * run through the list and find a worker thread that doesn't have a lot | |
510 | * to do right now. This can return null if we aren't yet at the thread | |
511 | * count limit and all of the threads are busy. | |
512 | */ | |
513 | static struct btrfs_worker_thread *next_worker(struct btrfs_workers *workers) | |
514 | { | |
515 | struct btrfs_worker_thread *worker; | |
516 | struct list_head *next; | |
61d92c32 CM |
517 | int enforce_min; |
518 | ||
519 | enforce_min = (workers->num_workers + workers->num_workers_starting) < | |
520 | workers->max_workers; | |
8b712842 | 521 | |
8b712842 | 522 | /* |
35d8ba66 CM |
523 | * if we find an idle thread, don't move it to the end of the |
524 | * idle list. This improves the chance that the next submission | |
525 | * will reuse the same thread, and maybe catch it while it is still | |
526 | * working | |
8b712842 | 527 | */ |
35d8ba66 CM |
528 | if (!list_empty(&workers->idle_list)) { |
529 | next = workers->idle_list.next; | |
8b712842 CM |
530 | worker = list_entry(next, struct btrfs_worker_thread, |
531 | worker_list); | |
35d8ba66 | 532 | return worker; |
8b712842 | 533 | } |
35d8ba66 CM |
534 | if (enforce_min || list_empty(&workers->worker_list)) |
535 | return NULL; | |
536 | ||
8b712842 | 537 | /* |
35d8ba66 | 538 | * if we pick a busy task, move the task to the end of the list. |
d352ac68 CM |
539 | * hopefully this will keep things somewhat evenly balanced. |
540 | * Do the move in batches based on the sequence number. This groups | |
541 | * requests submitted at roughly the same time onto the same worker. | |
8b712842 | 542 | */ |
35d8ba66 CM |
543 | next = workers->worker_list.next; |
544 | worker = list_entry(next, struct btrfs_worker_thread, worker_list); | |
4854ddd0 | 545 | worker->sequence++; |
d352ac68 | 546 | |
53863232 | 547 | if (worker->sequence % workers->idle_thresh == 0) |
4854ddd0 | 548 | list_move_tail(next, &workers->worker_list); |
8b712842 CM |
549 | return worker; |
550 | } | |
551 | ||
d352ac68 CM |
552 | /* |
553 | * selects a worker thread to take the next job. This will either find | |
554 | * an idle worker, start a new worker up to the max count, or just return | |
555 | * one of the existing busy workers. | |
556 | */ | |
8b712842 CM |
557 | static struct btrfs_worker_thread *find_worker(struct btrfs_workers *workers) |
558 | { | |
559 | struct btrfs_worker_thread *worker; | |
560 | unsigned long flags; | |
9042846b | 561 | struct list_head *fallback; |
0dc3b84a | 562 | int ret; |
8b712842 | 563 | |
8b712842 | 564 | spin_lock_irqsave(&workers->lock, flags); |
8d532b2a | 565 | again: |
8b712842 | 566 | worker = next_worker(workers); |
8b712842 CM |
567 | |
568 | if (!worker) { | |
61d92c32 CM |
569 | if (workers->num_workers + workers->num_workers_starting >= |
570 | workers->max_workers) { | |
9042846b CM |
571 | goto fallback; |
572 | } else if (workers->atomic_worker_start) { | |
573 | workers->atomic_start_pending = 1; | |
574 | goto fallback; | |
8b712842 | 575 | } else { |
61d92c32 | 576 | workers->num_workers_starting++; |
8b712842 CM |
577 | spin_unlock_irqrestore(&workers->lock, flags); |
578 | /* we're below the limit, start another worker */ | |
0dc3b84a | 579 | ret = __btrfs_start_workers(workers); |
8d532b2a | 580 | spin_lock_irqsave(&workers->lock, flags); |
0dc3b84a JB |
581 | if (ret) |
582 | goto fallback; | |
8b712842 CM |
583 | goto again; |
584 | } | |
585 | } | |
6e74057c | 586 | goto found; |
9042846b CM |
587 | |
588 | fallback: | |
589 | fallback = NULL; | |
590 | /* | |
591 | * we have failed to find any workers, just | |
592 | * return the first one we can find. | |
593 | */ | |
594 | if (!list_empty(&workers->worker_list)) | |
595 | fallback = workers->worker_list.next; | |
596 | if (!list_empty(&workers->idle_list)) | |
597 | fallback = workers->idle_list.next; | |
598 | BUG_ON(!fallback); | |
599 | worker = list_entry(fallback, | |
600 | struct btrfs_worker_thread, worker_list); | |
6e74057c CM |
601 | found: |
602 | /* | |
603 | * this makes sure the worker doesn't exit before it is placed | |
604 | * onto a busy/idle list | |
605 | */ | |
606 | atomic_inc(&worker->num_pending); | |
9042846b CM |
607 | spin_unlock_irqrestore(&workers->lock, flags); |
608 | return worker; | |
8b712842 CM |
609 | } |
610 | ||
611 | /* | |
612 | * btrfs_requeue_work just puts the work item back on the tail of the list | |
613 | * it was taken from. It is intended for use with long running work functions | |
614 | * that make some progress and want to give the cpu up for others. | |
615 | */ | |
143bede5 | 616 | void btrfs_requeue_work(struct btrfs_work *work) |
8b712842 CM |
617 | { |
618 | struct btrfs_worker_thread *worker = work->worker; | |
619 | unsigned long flags; | |
a6837051 | 620 | int wake = 0; |
8b712842 | 621 | |
4a69a410 | 622 | if (test_and_set_bit(WORK_QUEUED_BIT, &work->flags)) |
143bede5 | 623 | return; |
8b712842 CM |
624 | |
625 | spin_lock_irqsave(&worker->lock, flags); | |
d313d7a3 CM |
626 | if (test_bit(WORK_HIGH_PRIO_BIT, &work->flags)) |
627 | list_add_tail(&work->list, &worker->prio_pending); | |
628 | else | |
629 | list_add_tail(&work->list, &worker->pending); | |
b51912c9 | 630 | atomic_inc(&worker->num_pending); |
75ccf47d CM |
631 | |
632 | /* by definition we're busy, take ourselves off the idle | |
633 | * list | |
634 | */ | |
635 | if (worker->idle) { | |
29c5e8ce | 636 | spin_lock(&worker->workers->lock); |
75ccf47d CM |
637 | worker->idle = 0; |
638 | list_move_tail(&worker->worker_list, | |
6e74057c | 639 | &worker->workers->worker_list); |
29c5e8ce | 640 | spin_unlock(&worker->workers->lock); |
75ccf47d | 641 | } |
a6837051 CM |
642 | if (!worker->working) { |
643 | wake = 1; | |
644 | worker->working = 1; | |
645 | } | |
75ccf47d | 646 | |
a6837051 CM |
647 | if (wake) |
648 | wake_up_process(worker->task); | |
9042846b | 649 | spin_unlock_irqrestore(&worker->lock, flags); |
8b712842 CM |
650 | } |
651 | ||
d313d7a3 CM |
652 | void btrfs_set_work_high_prio(struct btrfs_work *work) |
653 | { | |
654 | set_bit(WORK_HIGH_PRIO_BIT, &work->flags); | |
655 | } | |
656 | ||
8b712842 CM |
657 | /* |
658 | * places a struct btrfs_work into the pending queue of one of the kthreads | |
659 | */ | |
0dc3b84a | 660 | void btrfs_queue_worker(struct btrfs_workers *workers, struct btrfs_work *work) |
8b712842 CM |
661 | { |
662 | struct btrfs_worker_thread *worker; | |
663 | unsigned long flags; | |
664 | int wake = 0; | |
665 | ||
666 | /* don't requeue something already on a list */ | |
4a69a410 | 667 | if (test_and_set_bit(WORK_QUEUED_BIT, &work->flags)) |
0dc3b84a | 668 | return; |
8b712842 CM |
669 | |
670 | worker = find_worker(workers); | |
4a69a410 | 671 | if (workers->ordered) { |
4e3f9c50 CM |
672 | /* |
673 | * you're not allowed to do ordered queues from an | |
674 | * interrupt handler | |
675 | */ | |
676 | spin_lock(&workers->order_lock); | |
d313d7a3 CM |
677 | if (test_bit(WORK_HIGH_PRIO_BIT, &work->flags)) { |
678 | list_add_tail(&work->order_list, | |
679 | &workers->prio_order_list); | |
680 | } else { | |
681 | list_add_tail(&work->order_list, &workers->order_list); | |
682 | } | |
4e3f9c50 | 683 | spin_unlock(&workers->order_lock); |
4a69a410 CM |
684 | } else { |
685 | INIT_LIST_HEAD(&work->order_list); | |
686 | } | |
8b712842 CM |
687 | |
688 | spin_lock_irqsave(&worker->lock, flags); | |
a6837051 | 689 | |
d313d7a3 CM |
690 | if (test_bit(WORK_HIGH_PRIO_BIT, &work->flags)) |
691 | list_add_tail(&work->list, &worker->prio_pending); | |
692 | else | |
693 | list_add_tail(&work->list, &worker->pending); | |
35d8ba66 | 694 | check_busy_worker(worker); |
8b712842 CM |
695 | |
696 | /* | |
697 | * avoid calling into wake_up_process if this thread has already | |
698 | * been kicked | |
699 | */ | |
700 | if (!worker->working) | |
701 | wake = 1; | |
702 | worker->working = 1; | |
703 | ||
8b712842 CM |
704 | if (wake) |
705 | wake_up_process(worker->task); | |
9042846b | 706 | spin_unlock_irqrestore(&worker->lock, flags); |
8b712842 | 707 | } |