]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/blob - sys/kern/subr_taskqueue.c
Upgrade Unbound to 1.8.0. More to follow.
[FreeBSD/FreeBSD.git] / sys / kern / subr_taskqueue.c
1 /*-
2  * SPDX-License-Identifier: BSD-2-Clause-FreeBSD
3  *
4  * Copyright (c) 2000 Doug Rabson
5  * All rights reserved.
6  *
7  * Redistribution and use in source and binary forms, with or without
8  * modification, are permitted provided that the following conditions
9  * are met:
10  * 1. Redistributions of source code must retain the above copyright
11  *    notice, this list of conditions and the following disclaimer.
12  * 2. Redistributions in binary form must reproduce the above copyright
13  *    notice, this list of conditions and the following disclaimer in the
14  *    documentation and/or other materials provided with the distribution.
15  *
16  * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
17  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
18  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
19  * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
20  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
21  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
22  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
23  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
24  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
25  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
26  * SUCH DAMAGE.
27  */
28
29 #include <sys/cdefs.h>
30 __FBSDID("$FreeBSD$");
31
32 #include <sys/param.h>
33 #include <sys/systm.h>
34 #include <sys/bus.h>
35 #include <sys/cpuset.h>
36 #include <sys/interrupt.h>
37 #include <sys/kernel.h>
38 #include <sys/kthread.h>
39 #include <sys/libkern.h>
40 #include <sys/limits.h>
41 #include <sys/lock.h>
42 #include <sys/malloc.h>
43 #include <sys/mutex.h>
44 #include <sys/proc.h>
45 #include <sys/sched.h>
46 #include <sys/smp.h>
47 #include <sys/taskqueue.h>
48 #include <sys/unistd.h>
49 #include <machine/stdarg.h>
50
51 static MALLOC_DEFINE(M_TASKQUEUE, "taskqueue", "Task Queues");
52 static void     *taskqueue_giant_ih;
53 static void     *taskqueue_ih;
54 static void      taskqueue_fast_enqueue(void *);
55 static void      taskqueue_swi_enqueue(void *);
56 static void      taskqueue_swi_giant_enqueue(void *);
57
58 struct taskqueue_busy {
59         struct task     *tb_running;
60         TAILQ_ENTRY(taskqueue_busy) tb_link;
61 };
62
63 struct task * const TB_DRAIN_WAITER = (struct task *)0x1;
64
65 struct taskqueue {
66         STAILQ_HEAD(, task)     tq_queue;
67         taskqueue_enqueue_fn    tq_enqueue;
68         void                    *tq_context;
69         char                    *tq_name;
70         TAILQ_HEAD(, taskqueue_busy) tq_active;
71         struct mtx              tq_mutex;
72         struct thread           **tq_threads;
73         int                     tq_tcount;
74         int                     tq_spin;
75         int                     tq_flags;
76         int                     tq_callouts;
77         taskqueue_callback_fn   tq_callbacks[TASKQUEUE_NUM_CALLBACKS];
78         void                    *tq_cb_contexts[TASKQUEUE_NUM_CALLBACKS];
79 };
80
81 #define TQ_FLAGS_ACTIVE         (1 << 0)
82 #define TQ_FLAGS_BLOCKED        (1 << 1)
83 #define TQ_FLAGS_UNLOCKED_ENQUEUE       (1 << 2)
84
85 #define DT_CALLOUT_ARMED        (1 << 0)
86 #define DT_DRAIN_IN_PROGRESS    (1 << 1)
87
88 #define TQ_LOCK(tq)                                                     \
89         do {                                                            \
90                 if ((tq)->tq_spin)                                      \
91                         mtx_lock_spin(&(tq)->tq_mutex);                 \
92                 else                                                    \
93                         mtx_lock(&(tq)->tq_mutex);                      \
94         } while (0)
95 #define TQ_ASSERT_LOCKED(tq)    mtx_assert(&(tq)->tq_mutex, MA_OWNED)
96
97 #define TQ_UNLOCK(tq)                                                   \
98         do {                                                            \
99                 if ((tq)->tq_spin)                                      \
100                         mtx_unlock_spin(&(tq)->tq_mutex);               \
101                 else                                                    \
102                         mtx_unlock(&(tq)->tq_mutex);                    \
103         } while (0)
104 #define TQ_ASSERT_UNLOCKED(tq)  mtx_assert(&(tq)->tq_mutex, MA_NOTOWNED)
105
106 void
107 _timeout_task_init(struct taskqueue *queue, struct timeout_task *timeout_task,
108     int priority, task_fn_t func, void *context)
109 {
110
111         TASK_INIT(&timeout_task->t, priority, func, context);
112         callout_init_mtx(&timeout_task->c, &queue->tq_mutex,
113             CALLOUT_RETURNUNLOCKED);
114         timeout_task->q = queue;
115         timeout_task->f = 0;
116 }
117
118 static __inline int
119 TQ_SLEEP(struct taskqueue *tq, void *p, struct mtx *m, int pri, const char *wm,
120     int t)
121 {
122         if (tq->tq_spin)
123                 return (msleep_spin(p, m, wm, t));
124         return (msleep(p, m, pri, wm, t));
125 }
126
127 static struct taskqueue *
128 _taskqueue_create(const char *name, int mflags,
129                  taskqueue_enqueue_fn enqueue, void *context,
130                  int mtxflags, const char *mtxname __unused)
131 {
132         struct taskqueue *queue;
133         char *tq_name;
134
135         tq_name = malloc(TASKQUEUE_NAMELEN, M_TASKQUEUE, mflags | M_ZERO);
136         if (tq_name == NULL)
137                 return (NULL);
138
139         queue = malloc(sizeof(struct taskqueue), M_TASKQUEUE, mflags | M_ZERO);
140         if (queue == NULL) {
141                 free(tq_name, M_TASKQUEUE);
142                 return (NULL);
143         }
144
145         snprintf(tq_name, TASKQUEUE_NAMELEN, "%s", (name) ? name : "taskqueue");
146
147         STAILQ_INIT(&queue->tq_queue);
148         TAILQ_INIT(&queue->tq_active);
149         queue->tq_enqueue = enqueue;
150         queue->tq_context = context;
151         queue->tq_name = tq_name;
152         queue->tq_spin = (mtxflags & MTX_SPIN) != 0;
153         queue->tq_flags |= TQ_FLAGS_ACTIVE;
154         if (enqueue == taskqueue_fast_enqueue ||
155             enqueue == taskqueue_swi_enqueue ||
156             enqueue == taskqueue_swi_giant_enqueue ||
157             enqueue == taskqueue_thread_enqueue)
158                 queue->tq_flags |= TQ_FLAGS_UNLOCKED_ENQUEUE;
159         mtx_init(&queue->tq_mutex, tq_name, NULL, mtxflags);
160
161         return (queue);
162 }
163
164 struct taskqueue *
165 taskqueue_create(const char *name, int mflags,
166                  taskqueue_enqueue_fn enqueue, void *context)
167 {
168
169         return _taskqueue_create(name, mflags, enqueue, context,
170                         MTX_DEF, name);
171 }
172
173 void
174 taskqueue_set_callback(struct taskqueue *queue,
175     enum taskqueue_callback_type cb_type, taskqueue_callback_fn callback,
176     void *context)
177 {
178
179         KASSERT(((cb_type >= TASKQUEUE_CALLBACK_TYPE_MIN) &&
180             (cb_type <= TASKQUEUE_CALLBACK_TYPE_MAX)),
181             ("Callback type %d not valid, must be %d-%d", cb_type,
182             TASKQUEUE_CALLBACK_TYPE_MIN, TASKQUEUE_CALLBACK_TYPE_MAX));
183         KASSERT((queue->tq_callbacks[cb_type] == NULL),
184             ("Re-initialization of taskqueue callback?"));
185
186         queue->tq_callbacks[cb_type] = callback;
187         queue->tq_cb_contexts[cb_type] = context;
188 }
189
190 /*
191  * Signal a taskqueue thread to terminate.
192  */
193 static void
194 taskqueue_terminate(struct thread **pp, struct taskqueue *tq)
195 {
196
197         while (tq->tq_tcount > 0 || tq->tq_callouts > 0) {
198                 wakeup(tq);
199                 TQ_SLEEP(tq, pp, &tq->tq_mutex, PWAIT, "taskqueue_destroy", 0);
200         }
201 }
202
203 void
204 taskqueue_free(struct taskqueue *queue)
205 {
206
207         TQ_LOCK(queue);
208         queue->tq_flags &= ~TQ_FLAGS_ACTIVE;
209         taskqueue_terminate(queue->tq_threads, queue);
210         KASSERT(TAILQ_EMPTY(&queue->tq_active), ("Tasks still running?"));
211         KASSERT(queue->tq_callouts == 0, ("Armed timeout tasks"));
212         mtx_destroy(&queue->tq_mutex);
213         free(queue->tq_threads, M_TASKQUEUE);
214         free(queue->tq_name, M_TASKQUEUE);
215         free(queue, M_TASKQUEUE);
216 }
217
218 static int
219 taskqueue_enqueue_locked(struct taskqueue *queue, struct task *task)
220 {
221         struct task *ins;
222         struct task *prev;
223
224         KASSERT(task->ta_func != NULL, ("enqueueing task with NULL func"));
225         /*
226          * Count multiple enqueues.
227          */
228         if (task->ta_pending) {
229                 if (task->ta_pending < USHRT_MAX)
230                         task->ta_pending++;
231                 TQ_UNLOCK(queue);
232                 return (0);
233         }
234
235         /*
236          * Optimise the case when all tasks have the same priority.
237          */
238         prev = STAILQ_LAST(&queue->tq_queue, task, ta_link);
239         if (!prev || prev->ta_priority >= task->ta_priority) {
240                 STAILQ_INSERT_TAIL(&queue->tq_queue, task, ta_link);
241         } else {
242                 prev = NULL;
243                 for (ins = STAILQ_FIRST(&queue->tq_queue); ins;
244                      prev = ins, ins = STAILQ_NEXT(ins, ta_link))
245                         if (ins->ta_priority < task->ta_priority)
246                                 break;
247
248                 if (prev)
249                         STAILQ_INSERT_AFTER(&queue->tq_queue, prev, task, ta_link);
250                 else
251                         STAILQ_INSERT_HEAD(&queue->tq_queue, task, ta_link);
252         }
253
254         task->ta_pending = 1;
255         if ((queue->tq_flags & TQ_FLAGS_UNLOCKED_ENQUEUE) != 0)
256                 TQ_UNLOCK(queue);
257         if ((queue->tq_flags & TQ_FLAGS_BLOCKED) == 0)
258                 queue->tq_enqueue(queue->tq_context);
259         if ((queue->tq_flags & TQ_FLAGS_UNLOCKED_ENQUEUE) == 0)
260                 TQ_UNLOCK(queue);
261
262         /* Return with lock released. */
263         return (0);
264 }
265
266 int
267 taskqueue_enqueue(struct taskqueue *queue, struct task *task)
268 {
269         int res;
270
271         TQ_LOCK(queue);
272         res = taskqueue_enqueue_locked(queue, task);
273         /* The lock is released inside. */
274
275         return (res);
276 }
277
278 static void
279 taskqueue_timeout_func(void *arg)
280 {
281         struct taskqueue *queue;
282         struct timeout_task *timeout_task;
283
284         timeout_task = arg;
285         queue = timeout_task->q;
286         KASSERT((timeout_task->f & DT_CALLOUT_ARMED) != 0, ("Stray timeout"));
287         timeout_task->f &= ~DT_CALLOUT_ARMED;
288         queue->tq_callouts--;
289         taskqueue_enqueue_locked(timeout_task->q, &timeout_task->t);
290         /* The lock is released inside. */
291 }
292
293 int
294 taskqueue_enqueue_timeout_sbt(struct taskqueue *queue,
295     struct timeout_task *timeout_task, sbintime_t sbt, sbintime_t pr, int flags)
296 {
297         int res;
298
299         TQ_LOCK(queue);
300         KASSERT(timeout_task->q == NULL || timeout_task->q == queue,
301             ("Migrated queue"));
302         KASSERT(!queue->tq_spin, ("Timeout for spin-queue"));
303         timeout_task->q = queue;
304         res = timeout_task->t.ta_pending;
305         if (timeout_task->f & DT_DRAIN_IN_PROGRESS) {
306                 /* Do nothing */
307                 TQ_UNLOCK(queue);
308                 res = -1;
309         } else if (sbt == 0) {
310                 taskqueue_enqueue_locked(queue, &timeout_task->t);
311                 /* The lock is released inside. */
312         } else {
313                 if ((timeout_task->f & DT_CALLOUT_ARMED) != 0) {
314                         res++;
315                 } else {
316                         queue->tq_callouts++;
317                         timeout_task->f |= DT_CALLOUT_ARMED;
318                         if (sbt < 0)
319                                 sbt = -sbt; /* Ignore overflow. */
320                 }
321                 if (sbt > 0) {
322                         callout_reset_sbt(&timeout_task->c, sbt, pr,
323                             taskqueue_timeout_func, timeout_task, flags);
324                 }
325                 TQ_UNLOCK(queue);
326         }
327         return (res);
328 }
329
330 int
331 taskqueue_enqueue_timeout(struct taskqueue *queue,
332     struct timeout_task *ttask, int ticks)
333 {
334
335         return (taskqueue_enqueue_timeout_sbt(queue, ttask, ticks * tick_sbt,
336             0, 0));
337 }
338
339 static void
340 taskqueue_task_nop_fn(void *context, int pending)
341 {
342 }
343
344 /*
345  * Block until all currently queued tasks in this taskqueue
346  * have begun execution.  Tasks queued during execution of
347  * this function are ignored.
348  */
349 static void
350 taskqueue_drain_tq_queue(struct taskqueue *queue)
351 {
352         struct task t_barrier;
353
354         if (STAILQ_EMPTY(&queue->tq_queue))
355                 return;
356
357         /*
358          * Enqueue our barrier after all current tasks, but with
359          * the highest priority so that newly queued tasks cannot
360          * pass it.  Because of the high priority, we can not use
361          * taskqueue_enqueue_locked directly (which drops the lock
362          * anyway) so just insert it at tail while we have the
363          * queue lock.
364          */
365         TASK_INIT(&t_barrier, USHRT_MAX, taskqueue_task_nop_fn, &t_barrier);
366         STAILQ_INSERT_TAIL(&queue->tq_queue, &t_barrier, ta_link);
367         t_barrier.ta_pending = 1;
368
369         /*
370          * Once the barrier has executed, all previously queued tasks
371          * have completed or are currently executing.
372          */
373         while (t_barrier.ta_pending != 0)
374                 TQ_SLEEP(queue, &t_barrier, &queue->tq_mutex, PWAIT, "-", 0);
375 }
376
377 /*
378  * Block until all currently executing tasks for this taskqueue
379  * complete.  Tasks that begin execution during the execution
380  * of this function are ignored.
381  */
382 static void
383 taskqueue_drain_tq_active(struct taskqueue *queue)
384 {
385         struct taskqueue_busy tb_marker, *tb_first;
386
387         if (TAILQ_EMPTY(&queue->tq_active))
388                 return;
389
390         /* Block taskq_terminate().*/
391         queue->tq_callouts++;
392
393         /*
394          * Wait for all currently executing taskqueue threads
395          * to go idle.
396          */
397         tb_marker.tb_running = TB_DRAIN_WAITER;
398         TAILQ_INSERT_TAIL(&queue->tq_active, &tb_marker, tb_link);
399         while (TAILQ_FIRST(&queue->tq_active) != &tb_marker)
400                 TQ_SLEEP(queue, &tb_marker, &queue->tq_mutex, PWAIT, "-", 0);
401         TAILQ_REMOVE(&queue->tq_active, &tb_marker, tb_link);
402
403         /*
404          * Wakeup any other drain waiter that happened to queue up
405          * without any intervening active thread.
406          */
407         tb_first = TAILQ_FIRST(&queue->tq_active);
408         if (tb_first != NULL && tb_first->tb_running == TB_DRAIN_WAITER)
409                 wakeup(tb_first);
410
411         /* Release taskqueue_terminate(). */
412         queue->tq_callouts--;
413         if ((queue->tq_flags & TQ_FLAGS_ACTIVE) == 0)
414                 wakeup_one(queue->tq_threads);
415 }
416
417 void
418 taskqueue_block(struct taskqueue *queue)
419 {
420
421         TQ_LOCK(queue);
422         queue->tq_flags |= TQ_FLAGS_BLOCKED;
423         TQ_UNLOCK(queue);
424 }
425
426 void
427 taskqueue_unblock(struct taskqueue *queue)
428 {
429
430         TQ_LOCK(queue);
431         queue->tq_flags &= ~TQ_FLAGS_BLOCKED;
432         if (!STAILQ_EMPTY(&queue->tq_queue))
433                 queue->tq_enqueue(queue->tq_context);
434         TQ_UNLOCK(queue);
435 }
436
437 static void
438 taskqueue_run_locked(struct taskqueue *queue)
439 {
440         struct taskqueue_busy tb;
441         struct taskqueue_busy *tb_first;
442         struct task *task;
443         int pending;
444
445         KASSERT(queue != NULL, ("tq is NULL"));
446         TQ_ASSERT_LOCKED(queue);
447         tb.tb_running = NULL;
448
449         while (STAILQ_FIRST(&queue->tq_queue)) {
450                 TAILQ_INSERT_TAIL(&queue->tq_active, &tb, tb_link);
451
452                 /*
453                  * Carefully remove the first task from the queue and
454                  * zero its pending count.
455                  */
456                 task = STAILQ_FIRST(&queue->tq_queue);
457                 KASSERT(task != NULL, ("task is NULL"));
458                 STAILQ_REMOVE_HEAD(&queue->tq_queue, ta_link);
459                 pending = task->ta_pending;
460                 task->ta_pending = 0;
461                 tb.tb_running = task;
462                 TQ_UNLOCK(queue);
463
464                 KASSERT(task->ta_func != NULL, ("task->ta_func is NULL"));
465                 task->ta_func(task->ta_context, pending);
466
467                 TQ_LOCK(queue);
468                 tb.tb_running = NULL;
469                 wakeup(task);
470
471                 TAILQ_REMOVE(&queue->tq_active, &tb, tb_link);
472                 tb_first = TAILQ_FIRST(&queue->tq_active);
473                 if (tb_first != NULL &&
474                     tb_first->tb_running == TB_DRAIN_WAITER)
475                         wakeup(tb_first);
476         }
477 }
478
479 void
480 taskqueue_run(struct taskqueue *queue)
481 {
482
483         TQ_LOCK(queue);
484         taskqueue_run_locked(queue);
485         TQ_UNLOCK(queue);
486 }
487
488 static int
489 task_is_running(struct taskqueue *queue, struct task *task)
490 {
491         struct taskqueue_busy *tb;
492
493         TQ_ASSERT_LOCKED(queue);
494         TAILQ_FOREACH(tb, &queue->tq_active, tb_link) {
495                 if (tb->tb_running == task)
496                         return (1);
497         }
498         return (0);
499 }
500
501 /*
502  * Only use this function in single threaded contexts. It returns
503  * non-zero if the given task is either pending or running. Else the
504  * task is idle and can be queued again or freed.
505  */
506 int
507 taskqueue_poll_is_busy(struct taskqueue *queue, struct task *task)
508 {
509         int retval;
510
511         TQ_LOCK(queue);
512         retval = task->ta_pending > 0 || task_is_running(queue, task);
513         TQ_UNLOCK(queue);
514
515         return (retval);
516 }
517
518 static int
519 taskqueue_cancel_locked(struct taskqueue *queue, struct task *task,
520     u_int *pendp)
521 {
522
523         if (task->ta_pending > 0)
524                 STAILQ_REMOVE(&queue->tq_queue, task, task, ta_link);
525         if (pendp != NULL)
526                 *pendp = task->ta_pending;
527         task->ta_pending = 0;
528         return (task_is_running(queue, task) ? EBUSY : 0);
529 }
530
531 int
532 taskqueue_cancel(struct taskqueue *queue, struct task *task, u_int *pendp)
533 {
534         int error;
535
536         TQ_LOCK(queue);
537         error = taskqueue_cancel_locked(queue, task, pendp);
538         TQ_UNLOCK(queue);
539
540         return (error);
541 }
542
543 int
544 taskqueue_cancel_timeout(struct taskqueue *queue,
545     struct timeout_task *timeout_task, u_int *pendp)
546 {
547         u_int pending, pending1;
548         int error;
549
550         TQ_LOCK(queue);
551         pending = !!(callout_stop(&timeout_task->c) > 0);
552         error = taskqueue_cancel_locked(queue, &timeout_task->t, &pending1);
553         if ((timeout_task->f & DT_CALLOUT_ARMED) != 0) {
554                 timeout_task->f &= ~DT_CALLOUT_ARMED;
555                 queue->tq_callouts--;
556         }
557         TQ_UNLOCK(queue);
558
559         if (pendp != NULL)
560                 *pendp = pending + pending1;
561         return (error);
562 }
563
564 void
565 taskqueue_drain(struct taskqueue *queue, struct task *task)
566 {
567
568         if (!queue->tq_spin)
569                 WITNESS_WARN(WARN_GIANTOK | WARN_SLEEPOK, NULL, __func__);
570
571         TQ_LOCK(queue);
572         while (task->ta_pending != 0 || task_is_running(queue, task))
573                 TQ_SLEEP(queue, task, &queue->tq_mutex, PWAIT, "-", 0);
574         TQ_UNLOCK(queue);
575 }
576
577 void
578 taskqueue_drain_all(struct taskqueue *queue)
579 {
580
581         if (!queue->tq_spin)
582                 WITNESS_WARN(WARN_GIANTOK | WARN_SLEEPOK, NULL, __func__);
583
584         TQ_LOCK(queue);
585         taskqueue_drain_tq_queue(queue);
586         taskqueue_drain_tq_active(queue);
587         TQ_UNLOCK(queue);
588 }
589
590 void
591 taskqueue_drain_timeout(struct taskqueue *queue,
592     struct timeout_task *timeout_task)
593 {
594
595         /*
596          * Set flag to prevent timer from re-starting during drain:
597          */
598         TQ_LOCK(queue);
599         KASSERT((timeout_task->f & DT_DRAIN_IN_PROGRESS) == 0,
600             ("Drain already in progress"));
601         timeout_task->f |= DT_DRAIN_IN_PROGRESS;
602         TQ_UNLOCK(queue);
603
604         callout_drain(&timeout_task->c);
605         taskqueue_drain(queue, &timeout_task->t);
606
607         /*
608          * Clear flag to allow timer to re-start:
609          */
610         TQ_LOCK(queue);
611         timeout_task->f &= ~DT_DRAIN_IN_PROGRESS;
612         TQ_UNLOCK(queue);
613 }
614
615 static void
616 taskqueue_swi_enqueue(void *context)
617 {
618         swi_sched(taskqueue_ih, 0);
619 }
620
621 static void
622 taskqueue_swi_run(void *dummy)
623 {
624         taskqueue_run(taskqueue_swi);
625 }
626
627 static void
628 taskqueue_swi_giant_enqueue(void *context)
629 {
630         swi_sched(taskqueue_giant_ih, 0);
631 }
632
633 static void
634 taskqueue_swi_giant_run(void *dummy)
635 {
636         taskqueue_run(taskqueue_swi_giant);
637 }
638
639 static int
640 _taskqueue_start_threads(struct taskqueue **tqp, int count, int pri,
641     cpuset_t *mask, const char *name, va_list ap)
642 {
643         char ktname[MAXCOMLEN + 1];
644         struct thread *td;
645         struct taskqueue *tq;
646         int i, error;
647
648         if (count <= 0)
649                 return (EINVAL);
650
651         vsnprintf(ktname, sizeof(ktname), name, ap);
652         tq = *tqp;
653
654         tq->tq_threads = malloc(sizeof(struct thread *) * count, M_TASKQUEUE,
655             M_NOWAIT | M_ZERO);
656         if (tq->tq_threads == NULL) {
657                 printf("%s: no memory for %s threads\n", __func__, ktname);
658                 return (ENOMEM);
659         }
660
661         for (i = 0; i < count; i++) {
662                 if (count == 1)
663                         error = kthread_add(taskqueue_thread_loop, tqp, NULL,
664                             &tq->tq_threads[i], RFSTOPPED, 0, "%s", ktname);
665                 else
666                         error = kthread_add(taskqueue_thread_loop, tqp, NULL,
667                             &tq->tq_threads[i], RFSTOPPED, 0,
668                             "%s_%d", ktname, i);
669                 if (error) {
670                         /* should be ok to continue, taskqueue_free will dtrt */
671                         printf("%s: kthread_add(%s): error %d", __func__,
672                             ktname, error);
673                         tq->tq_threads[i] = NULL;               /* paranoid */
674                 } else
675                         tq->tq_tcount++;
676         }
677         if (tq->tq_tcount == 0) {
678                 free(tq->tq_threads, M_TASKQUEUE);
679                 tq->tq_threads = NULL;
680                 return (ENOMEM);
681         }
682         for (i = 0; i < count; i++) {
683                 if (tq->tq_threads[i] == NULL)
684                         continue;
685                 td = tq->tq_threads[i];
686                 if (mask) {
687                         error = cpuset_setthread(td->td_tid, mask);
688                         /*
689                          * Failing to pin is rarely an actual fatal error;
690                          * it'll just affect performance.
691                          */
692                         if (error)
693                                 printf("%s: curthread=%llu: can't pin; "
694                                     "error=%d\n",
695                                     __func__,
696                                     (unsigned long long) td->td_tid,
697                                     error);
698                 }
699                 thread_lock(td);
700                 sched_prio(td, pri);
701                 sched_add(td, SRQ_BORING);
702                 thread_unlock(td);
703         }
704
705         return (0);
706 }
707
708 int
709 taskqueue_start_threads(struct taskqueue **tqp, int count, int pri,
710     const char *name, ...)
711 {
712         va_list ap;
713         int error;
714
715         va_start(ap, name);
716         error = _taskqueue_start_threads(tqp, count, pri, NULL, name, ap);
717         va_end(ap);
718         return (error);
719 }
720
721 int
722 taskqueue_start_threads_cpuset(struct taskqueue **tqp, int count, int pri,
723     cpuset_t *mask, const char *name, ...)
724 {
725         va_list ap;
726         int error;
727
728         va_start(ap, name);
729         error = _taskqueue_start_threads(tqp, count, pri, mask, name, ap);
730         va_end(ap);
731         return (error);
732 }
733
734 static inline void
735 taskqueue_run_callback(struct taskqueue *tq,
736     enum taskqueue_callback_type cb_type)
737 {
738         taskqueue_callback_fn tq_callback;
739
740         TQ_ASSERT_UNLOCKED(tq);
741         tq_callback = tq->tq_callbacks[cb_type];
742         if (tq_callback != NULL)
743                 tq_callback(tq->tq_cb_contexts[cb_type]);
744 }
745
746 void
747 taskqueue_thread_loop(void *arg)
748 {
749         struct taskqueue **tqp, *tq;
750
751         tqp = arg;
752         tq = *tqp;
753         taskqueue_run_callback(tq, TASKQUEUE_CALLBACK_TYPE_INIT);
754         TQ_LOCK(tq);
755         while ((tq->tq_flags & TQ_FLAGS_ACTIVE) != 0) {
756                 /* XXX ? */
757                 taskqueue_run_locked(tq);
758                 /*
759                  * Because taskqueue_run() can drop tq_mutex, we need to
760                  * check if the TQ_FLAGS_ACTIVE flag wasn't removed in the
761                  * meantime, which means we missed a wakeup.
762                  */
763                 if ((tq->tq_flags & TQ_FLAGS_ACTIVE) == 0)
764                         break;
765                 TQ_SLEEP(tq, tq, &tq->tq_mutex, 0, "-", 0);
766         }
767         taskqueue_run_locked(tq);
768         /*
769          * This thread is on its way out, so just drop the lock temporarily
770          * in order to call the shutdown callback.  This allows the callback
771          * to look at the taskqueue, even just before it dies.
772          */
773         TQ_UNLOCK(tq);
774         taskqueue_run_callback(tq, TASKQUEUE_CALLBACK_TYPE_SHUTDOWN);
775         TQ_LOCK(tq);
776
777         /* rendezvous with thread that asked us to terminate */
778         tq->tq_tcount--;
779         wakeup_one(tq->tq_threads);
780         TQ_UNLOCK(tq);
781         kthread_exit();
782 }
783
784 void
785 taskqueue_thread_enqueue(void *context)
786 {
787         struct taskqueue **tqp, *tq;
788
789         tqp = context;
790         tq = *tqp;
791         wakeup_one(tq);
792 }
793
794 TASKQUEUE_DEFINE(swi, taskqueue_swi_enqueue, NULL,
795                  swi_add(NULL, "task queue", taskqueue_swi_run, NULL, SWI_TQ,
796                      INTR_MPSAFE, &taskqueue_ih));
797
798 TASKQUEUE_DEFINE(swi_giant, taskqueue_swi_giant_enqueue, NULL,
799                  swi_add(NULL, "Giant taskq", taskqueue_swi_giant_run,
800                      NULL, SWI_TQ_GIANT, 0, &taskqueue_giant_ih));
801
802 TASKQUEUE_DEFINE_THREAD(thread);
803
804 struct taskqueue *
805 taskqueue_create_fast(const char *name, int mflags,
806                  taskqueue_enqueue_fn enqueue, void *context)
807 {
808         return _taskqueue_create(name, mflags, enqueue, context,
809                         MTX_SPIN, "fast_taskqueue");
810 }
811
812 static void     *taskqueue_fast_ih;
813
814 static void
815 taskqueue_fast_enqueue(void *context)
816 {
817         swi_sched(taskqueue_fast_ih, 0);
818 }
819
820 static void
821 taskqueue_fast_run(void *dummy)
822 {
823         taskqueue_run(taskqueue_fast);
824 }
825
826 TASKQUEUE_FAST_DEFINE(fast, taskqueue_fast_enqueue, NULL,
827         swi_add(NULL, "fast taskq", taskqueue_fast_run, NULL,
828         SWI_TQ_FAST, INTR_MPSAFE, &taskqueue_fast_ih));
829
830 int
831 taskqueue_member(struct taskqueue *queue, struct thread *td)
832 {
833         int i, j, ret = 0;
834
835         for (i = 0, j = 0; ; i++) {
836                 if (queue->tq_threads[i] == NULL)
837                         continue;
838                 if (queue->tq_threads[i] == td) {
839                         ret = 1;
840                         break;
841                 }
842                 if (++j >= queue->tq_tcount)
843                         break;
844         }
845         return (ret);
846 }