2 * linux/kernel/workqueue.c
4 * Generic mechanism for defining kernel helper threads for running
5 * arbitrary tasks in process context.
7 * Started by Ingo Molnar, Copyright (C) 2002
9 * Derived from the taskqueue/keventd code by:
11 * David Woodhouse <dwmw2@infradead.org>
12 * Andrew Morton <andrewm@uow.edu.au>
13 * Kai Petzke <wpp@marie.physik.tu-berlin.de>
14 * Theodore Ts'o <tytso@mit.edu>
16 * Made to use alloc_percpu by Christoph Lameter <clameter@sgi.com>.
19 #include <linux/module.h>
20 #include <linux/kernel.h>
21 #include <linux/sched.h>
22 #include <linux/init.h>
23 #include <linux/signal.h>
24 #include <linux/completion.h>
25 #include <linux/workqueue.h>
26 #include <linux/slab.h>
27 #include <linux/cpu.h>
28 #include <linux/notifier.h>
29 #include <linux/kthread.h>
30 #include <linux/hardirq.h>
31 #include <linux/mempolicy.h>
32 #include <linux/freezer.h>
33 #include <linux/kallsyms.h>
34 #include <linux/debug_locks.h>
37 * The per-CPU workqueue (if single thread, we always use the first
40 * The sequence counters are for flush_scheduled_work(). It wants to wait
41 * until all currently-scheduled works are completed, but it doesn't
42 * want to be livelocked by new, incoming ones. So it waits until
43 * remove_sequence is >= the insert_sequence which pertained when
44 * flush_scheduled_work() was called.
46 struct cpu_workqueue_struct
{
50 long remove_sequence
; /* Least-recently added (next to run) */
51 long insert_sequence
; /* Next to add */
53 struct list_head worklist
;
54 wait_queue_head_t more_work
;
55 wait_queue_head_t work_done
;
57 struct workqueue_struct
*wq
;
58 struct task_struct
*thread
;
60 int run_depth
; /* Detect run_workqueue() recursion depth */
62 int freezeable
; /* Freeze the thread during suspend */
63 } ____cacheline_aligned
;
66 * The externally visible workqueue abstraction is an array of
69 struct workqueue_struct
{
70 struct cpu_workqueue_struct
*cpu_wq
;
72 struct list_head list
; /* Empty if single thread */
75 /* All the per-cpu workqueues on the system, for hotplug cpu to add/remove
76 threads to each one as cpus come/go. */
77 static DEFINE_MUTEX(workqueue_mutex
);
78 static LIST_HEAD(workqueues
);
80 static int singlethread_cpu
;
82 /* If it's single threaded, it isn't in the list of workqueues. */
83 static inline int is_single_threaded(struct workqueue_struct
*wq
)
85 return list_empty(&wq
->list
);
89 * Set the workqueue on which a work item is to be run
90 * - Must *only* be called if the pending flag is set
92 static inline void set_wq_data(struct work_struct
*work
, void *wq
)
96 BUG_ON(!work_pending(work
));
98 new = (unsigned long) wq
| (1UL << WORK_STRUCT_PENDING
);
99 new |= WORK_STRUCT_FLAG_MASK
& *work_data_bits(work
);
100 atomic_long_set(&work
->data
, new);
103 static inline void *get_wq_data(struct work_struct
*work
)
105 return (void *) (atomic_long_read(&work
->data
) & WORK_STRUCT_WQ_DATA_MASK
);
108 static int __run_work(struct cpu_workqueue_struct
*cwq
, struct work_struct
*work
)
113 spin_lock_irqsave(&cwq
->lock
, flags
);
115 * We need to re-validate the work info after we've gotten
116 * the cpu_workqueue lock. We can run the work now iff:
118 * - the wq_data still matches the cpu_workqueue_struct
119 * - AND the work is still marked pending
120 * - AND the work is still on a list (which will be this
121 * workqueue_struct list)
123 * All these conditions are important, because we
124 * need to protect against the work being run right
125 * now on another CPU (all but the last one might be
126 * true if it's currently running and has not been
127 * released yet, for example).
129 if (get_wq_data(work
) == cwq
130 && work_pending(work
)
131 && !list_empty(&work
->entry
)) {
132 work_func_t f
= work
->func
;
133 list_del_init(&work
->entry
);
134 spin_unlock_irqrestore(&cwq
->lock
, flags
);
136 if (!test_bit(WORK_STRUCT_NOAUTOREL
, work_data_bits(work
)))
140 spin_lock_irqsave(&cwq
->lock
, flags
);
141 cwq
->remove_sequence
++;
142 wake_up(&cwq
->work_done
);
145 spin_unlock_irqrestore(&cwq
->lock
, flags
);
150 * run_scheduled_work - run scheduled work synchronously
153 * This checks if the work was pending, and runs it
154 * synchronously if so. It returns a boolean to indicate
155 * whether it had any scheduled work to run or not.
157 * NOTE! This _only_ works for normal work_structs. You
158 * CANNOT use this for delayed work, because the wq data
159 * for delayed work will not point properly to the per-
160 * CPU workqueue struct, but will change!
162 int fastcall
run_scheduled_work(struct work_struct
*work
)
165 struct cpu_workqueue_struct
*cwq
;
167 if (!work_pending(work
))
169 if (list_empty(&work
->entry
))
171 /* NOTE! This depends intimately on __queue_work! */
172 cwq
= get_wq_data(work
);
175 if (__run_work(cwq
, work
))
179 EXPORT_SYMBOL(run_scheduled_work
);
181 /* Preempt must be disabled. */
182 static void __queue_work(struct cpu_workqueue_struct
*cwq
,
183 struct work_struct
*work
)
187 spin_lock_irqsave(&cwq
->lock
, flags
);
188 set_wq_data(work
, cwq
);
189 list_add_tail(&work
->entry
, &cwq
->worklist
);
190 cwq
->insert_sequence
++;
191 wake_up(&cwq
->more_work
);
192 spin_unlock_irqrestore(&cwq
->lock
, flags
);
196 * queue_work - queue work on a workqueue
197 * @wq: workqueue to use
198 * @work: work to queue
200 * Returns 0 if @work was already on a queue, non-zero otherwise.
202 * We queue the work to the CPU it was submitted, but there is no
203 * guarantee that it will be processed by that CPU.
205 int fastcall
queue_work(struct workqueue_struct
*wq
, struct work_struct
*work
)
207 int ret
= 0, cpu
= get_cpu();
209 if (!test_and_set_bit(WORK_STRUCT_PENDING
, work_data_bits(work
))) {
210 if (unlikely(is_single_threaded(wq
)))
211 cpu
= singlethread_cpu
;
212 BUG_ON(!list_empty(&work
->entry
));
213 __queue_work(per_cpu_ptr(wq
->cpu_wq
, cpu
), work
);
219 EXPORT_SYMBOL_GPL(queue_work
);
221 static void delayed_work_timer_fn(unsigned long __data
)
223 struct delayed_work
*dwork
= (struct delayed_work
*)__data
;
224 struct workqueue_struct
*wq
= get_wq_data(&dwork
->work
);
225 int cpu
= smp_processor_id();
227 if (unlikely(is_single_threaded(wq
)))
228 cpu
= singlethread_cpu
;
230 __queue_work(per_cpu_ptr(wq
->cpu_wq
, cpu
), &dwork
->work
);
234 * queue_delayed_work - queue work on a workqueue after delay
235 * @wq: workqueue to use
236 * @dwork: delayable work to queue
237 * @delay: number of jiffies to wait before queueing
239 * Returns 0 if @work was already on a queue, non-zero otherwise.
241 int fastcall
queue_delayed_work(struct workqueue_struct
*wq
,
242 struct delayed_work
*dwork
, unsigned long delay
)
245 struct timer_list
*timer
= &dwork
->timer
;
246 struct work_struct
*work
= &dwork
->work
;
249 return queue_work(wq
, work
);
251 if (!test_and_set_bit(WORK_STRUCT_PENDING
, work_data_bits(work
))) {
252 BUG_ON(timer_pending(timer
));
253 BUG_ON(!list_empty(&work
->entry
));
255 /* This stores wq for the moment, for the timer_fn */
256 set_wq_data(work
, wq
);
257 timer
->expires
= jiffies
+ delay
;
258 timer
->data
= (unsigned long)dwork
;
259 timer
->function
= delayed_work_timer_fn
;
265 EXPORT_SYMBOL_GPL(queue_delayed_work
);
268 * queue_delayed_work_on - queue work on specific CPU after delay
269 * @cpu: CPU number to execute work on
270 * @wq: workqueue to use
271 * @dwork: work to queue
272 * @delay: number of jiffies to wait before queueing
274 * Returns 0 if @work was already on a queue, non-zero otherwise.
276 int queue_delayed_work_on(int cpu
, struct workqueue_struct
*wq
,
277 struct delayed_work
*dwork
, unsigned long delay
)
280 struct timer_list
*timer
= &dwork
->timer
;
281 struct work_struct
*work
= &dwork
->work
;
283 if (!test_and_set_bit(WORK_STRUCT_PENDING
, work_data_bits(work
))) {
284 BUG_ON(timer_pending(timer
));
285 BUG_ON(!list_empty(&work
->entry
));
287 /* This stores wq for the moment, for the timer_fn */
288 set_wq_data(work
, wq
);
289 timer
->expires
= jiffies
+ delay
;
290 timer
->data
= (unsigned long)dwork
;
291 timer
->function
= delayed_work_timer_fn
;
292 add_timer_on(timer
, cpu
);
297 EXPORT_SYMBOL_GPL(queue_delayed_work_on
);
299 static void run_workqueue(struct cpu_workqueue_struct
*cwq
)
304 * Keep taking off work from the queue until
307 spin_lock_irqsave(&cwq
->lock
, flags
);
309 if (cwq
->run_depth
> 3) {
310 /* morton gets to eat his hat */
311 printk("%s: recursion depth exceeded: %d\n",
312 __FUNCTION__
, cwq
->run_depth
);
315 while (!list_empty(&cwq
->worklist
)) {
316 struct work_struct
*work
= list_entry(cwq
->worklist
.next
,
317 struct work_struct
, entry
);
318 work_func_t f
= work
->func
;
320 list_del_init(cwq
->worklist
.next
);
321 spin_unlock_irqrestore(&cwq
->lock
, flags
);
323 BUG_ON(get_wq_data(work
) != cwq
);
324 if (!test_bit(WORK_STRUCT_NOAUTOREL
, work_data_bits(work
)))
328 if (unlikely(in_atomic() || lockdep_depth(current
) > 0)) {
329 printk(KERN_ERR
"BUG: workqueue leaked lock or atomic: "
331 current
->comm
, preempt_count(),
333 printk(KERN_ERR
" last function: ");
334 print_symbol("%s\n", (unsigned long)f
);
335 debug_show_held_locks(current
);
339 spin_lock_irqsave(&cwq
->lock
, flags
);
340 cwq
->remove_sequence
++;
341 wake_up(&cwq
->work_done
);
344 spin_unlock_irqrestore(&cwq
->lock
, flags
);
347 static int worker_thread(void *__cwq
)
349 struct cpu_workqueue_struct
*cwq
= __cwq
;
350 DECLARE_WAITQUEUE(wait
, current
);
351 struct k_sigaction sa
;
354 if (!cwq
->freezeable
)
355 current
->flags
|= PF_NOFREEZE
;
357 set_user_nice(current
, -5);
359 /* Block and flush all signals */
360 sigfillset(&blocked
);
361 sigprocmask(SIG_BLOCK
, &blocked
, NULL
);
362 flush_signals(current
);
365 * We inherited MPOL_INTERLEAVE from the booting kernel.
366 * Set MPOL_DEFAULT to insure node local allocations.
368 numa_default_policy();
370 /* SIG_IGN makes children autoreap: see do_notify_parent(). */
371 sa
.sa
.sa_handler
= SIG_IGN
;
373 siginitset(&sa
.sa
.sa_mask
, sigmask(SIGCHLD
));
374 do_sigaction(SIGCHLD
, &sa
, (struct k_sigaction
*)0);
376 set_current_state(TASK_INTERRUPTIBLE
);
377 while (!kthread_should_stop()) {
381 add_wait_queue(&cwq
->more_work
, &wait
);
382 if (list_empty(&cwq
->worklist
))
385 __set_current_state(TASK_RUNNING
);
386 remove_wait_queue(&cwq
->more_work
, &wait
);
388 if (!list_empty(&cwq
->worklist
))
390 set_current_state(TASK_INTERRUPTIBLE
);
392 __set_current_state(TASK_RUNNING
);
396 static void flush_cpu_workqueue(struct cpu_workqueue_struct
*cwq
)
398 if (cwq
->thread
== current
) {
400 * Probably keventd trying to flush its own queue. So simply run
401 * it by hand rather than deadlocking.
406 long sequence_needed
;
408 spin_lock_irq(&cwq
->lock
);
409 sequence_needed
= cwq
->insert_sequence
;
411 while (sequence_needed
- cwq
->remove_sequence
> 0) {
412 prepare_to_wait(&cwq
->work_done
, &wait
,
413 TASK_UNINTERRUPTIBLE
);
414 spin_unlock_irq(&cwq
->lock
);
416 spin_lock_irq(&cwq
->lock
);
418 finish_wait(&cwq
->work_done
, &wait
);
419 spin_unlock_irq(&cwq
->lock
);
424 * flush_workqueue - ensure that any scheduled work has run to completion.
425 * @wq: workqueue to flush
427 * Forces execution of the workqueue and blocks until its completion.
428 * This is typically used in driver shutdown handlers.
430 * This function will sample each workqueue's current insert_sequence number and
431 * will sleep until the head sequence is greater than or equal to that. This
432 * means that we sleep until all works which were queued on entry have been
433 * handled, but we are not livelocked by new incoming ones.
435 * This function used to run the workqueues itself. Now we just wait for the
436 * helper threads to do it.
438 void fastcall
flush_workqueue(struct workqueue_struct
*wq
)
442 if (is_single_threaded(wq
)) {
443 /* Always use first cpu's area. */
444 flush_cpu_workqueue(per_cpu_ptr(wq
->cpu_wq
, singlethread_cpu
));
448 mutex_lock(&workqueue_mutex
);
449 for_each_online_cpu(cpu
)
450 flush_cpu_workqueue(per_cpu_ptr(wq
->cpu_wq
, cpu
));
451 mutex_unlock(&workqueue_mutex
);
454 EXPORT_SYMBOL_GPL(flush_workqueue
);
456 static struct task_struct
*create_workqueue_thread(struct workqueue_struct
*wq
,
457 int cpu
, int freezeable
)
459 struct cpu_workqueue_struct
*cwq
= per_cpu_ptr(wq
->cpu_wq
, cpu
);
460 struct task_struct
*p
;
462 spin_lock_init(&cwq
->lock
);
465 cwq
->insert_sequence
= 0;
466 cwq
->remove_sequence
= 0;
467 cwq
->freezeable
= freezeable
;
468 INIT_LIST_HEAD(&cwq
->worklist
);
469 init_waitqueue_head(&cwq
->more_work
);
470 init_waitqueue_head(&cwq
->work_done
);
472 if (is_single_threaded(wq
))
473 p
= kthread_create(worker_thread
, cwq
, "%s", wq
->name
);
475 p
= kthread_create(worker_thread
, cwq
, "%s/%d", wq
->name
, cpu
);
482 struct workqueue_struct
*__create_workqueue(const char *name
,
483 int singlethread
, int freezeable
)
485 int cpu
, destroy
= 0;
486 struct workqueue_struct
*wq
;
487 struct task_struct
*p
;
489 wq
= kzalloc(sizeof(*wq
), GFP_KERNEL
);
493 wq
->cpu_wq
= alloc_percpu(struct cpu_workqueue_struct
);
500 mutex_lock(&workqueue_mutex
);
502 INIT_LIST_HEAD(&wq
->list
);
503 p
= create_workqueue_thread(wq
, singlethread_cpu
, freezeable
);
509 list_add(&wq
->list
, &workqueues
);
510 for_each_online_cpu(cpu
) {
511 p
= create_workqueue_thread(wq
, cpu
, freezeable
);
513 kthread_bind(p
, cpu
);
519 mutex_unlock(&workqueue_mutex
);
522 * Was there any error during startup? If yes then clean up:
525 destroy_workqueue(wq
);
530 EXPORT_SYMBOL_GPL(__create_workqueue
);
532 static void cleanup_workqueue_thread(struct workqueue_struct
*wq
, int cpu
)
534 struct cpu_workqueue_struct
*cwq
;
536 struct task_struct
*p
;
538 cwq
= per_cpu_ptr(wq
->cpu_wq
, cpu
);
539 spin_lock_irqsave(&cwq
->lock
, flags
);
542 spin_unlock_irqrestore(&cwq
->lock
, flags
);
548 * destroy_workqueue - safely terminate a workqueue
549 * @wq: target workqueue
551 * Safely destroy a workqueue. All work currently pending will be done first.
553 void destroy_workqueue(struct workqueue_struct
*wq
)
559 /* We don't need the distraction of CPUs appearing and vanishing. */
560 mutex_lock(&workqueue_mutex
);
561 if (is_single_threaded(wq
))
562 cleanup_workqueue_thread(wq
, singlethread_cpu
);
564 for_each_online_cpu(cpu
)
565 cleanup_workqueue_thread(wq
, cpu
);
568 mutex_unlock(&workqueue_mutex
);
569 free_percpu(wq
->cpu_wq
);
572 EXPORT_SYMBOL_GPL(destroy_workqueue
);
574 static struct workqueue_struct
*keventd_wq
;
577 * schedule_work - put work task in global workqueue
578 * @work: job to be done
580 * This puts a job in the kernel-global workqueue.
582 int fastcall
schedule_work(struct work_struct
*work
)
584 return queue_work(keventd_wq
, work
);
586 EXPORT_SYMBOL(schedule_work
);
589 * schedule_delayed_work - put work task in global workqueue after delay
590 * @dwork: job to be done
591 * @delay: number of jiffies to wait or 0 for immediate execution
593 * After waiting for a given time this puts a job in the kernel-global
596 int fastcall
schedule_delayed_work(struct delayed_work
*dwork
, unsigned long delay
)
598 return queue_delayed_work(keventd_wq
, dwork
, delay
);
600 EXPORT_SYMBOL(schedule_delayed_work
);
603 * schedule_delayed_work_on - queue work in global workqueue on CPU after delay
605 * @dwork: job to be done
606 * @delay: number of jiffies to wait
608 * After waiting for a given time this puts a job in the kernel-global
609 * workqueue on the specified CPU.
611 int schedule_delayed_work_on(int cpu
,
612 struct delayed_work
*dwork
, unsigned long delay
)
614 return queue_delayed_work_on(cpu
, keventd_wq
, dwork
, delay
);
616 EXPORT_SYMBOL(schedule_delayed_work_on
);
619 * schedule_on_each_cpu - call a function on each online CPU from keventd
620 * @func: the function to call
622 * Returns zero on success.
623 * Returns -ve errno on failure.
625 * Appears to be racy against CPU hotplug.
627 * schedule_on_each_cpu() is very slow.
629 int schedule_on_each_cpu(work_func_t func
)
632 struct work_struct
*works
;
634 works
= alloc_percpu(struct work_struct
);
638 mutex_lock(&workqueue_mutex
);
639 for_each_online_cpu(cpu
) {
640 struct work_struct
*work
= per_cpu_ptr(works
, cpu
);
642 INIT_WORK(work
, func
);
643 set_bit(WORK_STRUCT_PENDING
, work_data_bits(work
));
644 __queue_work(per_cpu_ptr(keventd_wq
->cpu_wq
, cpu
), work
);
646 mutex_unlock(&workqueue_mutex
);
647 flush_workqueue(keventd_wq
);
652 void flush_scheduled_work(void)
654 flush_workqueue(keventd_wq
);
656 EXPORT_SYMBOL(flush_scheduled_work
);
659 * cancel_rearming_delayed_workqueue - reliably kill off a delayed
660 * work whose handler rearms the delayed work.
661 * @wq: the controlling workqueue structure
662 * @dwork: the delayed work struct
664 void cancel_rearming_delayed_workqueue(struct workqueue_struct
*wq
,
665 struct delayed_work
*dwork
)
667 while (!cancel_delayed_work(dwork
))
670 EXPORT_SYMBOL(cancel_rearming_delayed_workqueue
);
673 * cancel_rearming_delayed_work - reliably kill off a delayed keventd
674 * work whose handler rearms the delayed work.
675 * @dwork: the delayed work struct
677 void cancel_rearming_delayed_work(struct delayed_work
*dwork
)
679 cancel_rearming_delayed_workqueue(keventd_wq
, dwork
);
681 EXPORT_SYMBOL(cancel_rearming_delayed_work
);
684 * execute_in_process_context - reliably execute the routine with user context
685 * @fn: the function to execute
686 * @ew: guaranteed storage for the execute work structure (must
687 * be available when the work executes)
689 * Executes the function immediately if process context is available,
690 * otherwise schedules the function for delayed execution.
692 * Returns: 0 - function was executed
693 * 1 - function was scheduled for execution
695 int execute_in_process_context(work_func_t fn
, struct execute_work
*ew
)
697 if (!in_interrupt()) {
702 INIT_WORK(&ew
->work
, fn
);
703 schedule_work(&ew
->work
);
707 EXPORT_SYMBOL_GPL(execute_in_process_context
);
711 return keventd_wq
!= NULL
;
714 int current_is_keventd(void)
716 struct cpu_workqueue_struct
*cwq
;
717 int cpu
= smp_processor_id(); /* preempt-safe: keventd is per-cpu */
722 cwq
= per_cpu_ptr(keventd_wq
->cpu_wq
, cpu
);
723 if (current
== cwq
->thread
)
730 /* Take the work from this (downed) CPU. */
731 static void take_over_work(struct workqueue_struct
*wq
, unsigned int cpu
)
733 struct cpu_workqueue_struct
*cwq
= per_cpu_ptr(wq
->cpu_wq
, cpu
);
734 struct list_head list
;
735 struct work_struct
*work
;
737 spin_lock_irq(&cwq
->lock
);
738 list_replace_init(&cwq
->worklist
, &list
);
740 while (!list_empty(&list
)) {
741 printk("Taking work for %s\n", wq
->name
);
742 work
= list_entry(list
.next
,struct work_struct
,entry
);
743 list_del(&work
->entry
);
744 __queue_work(per_cpu_ptr(wq
->cpu_wq
, smp_processor_id()), work
);
746 spin_unlock_irq(&cwq
->lock
);
749 /* We're holding the cpucontrol mutex here */
750 static int __devinit
workqueue_cpu_callback(struct notifier_block
*nfb
,
751 unsigned long action
,
754 unsigned int hotcpu
= (unsigned long)hcpu
;
755 struct workqueue_struct
*wq
;
759 mutex_lock(&workqueue_mutex
);
760 /* Create a new workqueue thread for it. */
761 list_for_each_entry(wq
, &workqueues
, list
) {
762 if (!create_workqueue_thread(wq
, hotcpu
, 0)) {
763 printk("workqueue for %i failed\n", hotcpu
);
770 /* Kick off worker threads. */
771 list_for_each_entry(wq
, &workqueues
, list
) {
772 struct cpu_workqueue_struct
*cwq
;
774 cwq
= per_cpu_ptr(wq
->cpu_wq
, hotcpu
);
775 kthread_bind(cwq
->thread
, hotcpu
);
776 wake_up_process(cwq
->thread
);
778 mutex_unlock(&workqueue_mutex
);
781 case CPU_UP_CANCELED
:
782 list_for_each_entry(wq
, &workqueues
, list
) {
783 if (!per_cpu_ptr(wq
->cpu_wq
, hotcpu
)->thread
)
785 /* Unbind so it can run. */
786 kthread_bind(per_cpu_ptr(wq
->cpu_wq
, hotcpu
)->thread
,
787 any_online_cpu(cpu_online_map
));
788 cleanup_workqueue_thread(wq
, hotcpu
);
790 mutex_unlock(&workqueue_mutex
);
793 case CPU_DOWN_PREPARE
:
794 mutex_lock(&workqueue_mutex
);
797 case CPU_DOWN_FAILED
:
798 mutex_unlock(&workqueue_mutex
);
802 list_for_each_entry(wq
, &workqueues
, list
)
803 cleanup_workqueue_thread(wq
, hotcpu
);
804 list_for_each_entry(wq
, &workqueues
, list
)
805 take_over_work(wq
, hotcpu
);
806 mutex_unlock(&workqueue_mutex
);
813 void init_workqueues(void)
815 singlethread_cpu
= first_cpu(cpu_possible_map
);
816 hotcpu_notifier(workqueue_cpu_callback
, 0);
817 keventd_wq
= create_workqueue("events");