Change cancel_work_sync() and cancel_delayed_work_sync() to return a boolean
indicating whether the work was actually cancelled.  A zero return value means
that the work was not pending/queued.

Without that kind of change it is not possible to avoid flush_workqueue()
sometimes, see the next patch as an example.

Also, this patch unifies both functions and kills the (unlikely) busy-wait
loop.
Signed-off-by: NOleg Nesterov <oleg@tv-sign.ru>
Acked-by: NJarek Poplawski <jarkao2@o2.pl>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Imho, the current naming of cancel_xxx workqueue functions is very confusing.

	cancel_delayed_work()
	cancel_rearming_delayed_work()
	cancel_rearming_delayed_workqueue()	// obsolete

	cancel_work_sync()

This looks as if the first 2 functions differ in "type" of their argument
which is not true any longer, nowadays the difference is the behaviour.

The semantics of cancel_rearming_delayed_work(dwork) was changed
significantly, it doesn't require that dwork rearms itself, and cancels dwork
synchronously.

Rename it to cancel_delayed_work_sync().  This matches cancel_delayed_work()
and cancel_work_sync().  Re-create cancel_rearming_delayed_work() as a simple
inline obsolete wrapper, like cancel_rearming_delayed_workqueue().
Signed-off-by: NOleg Nesterov <oleg@tv-sign.ru>
Acked-by: NJarek Poplawski <jarkao2@o2.pl>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

As pointed out by Jarek Poplawski, the patch

	[WORKQUEUE]: cancel_delayed_work: use del_timer() instead of del_timer_sync()
	commit: 071b6386

was wrong, it was merged by mistake after that.

From the changelog:

	after this patch:
		...
		delayed_work_timer_fn->__queue_work() in progress.

		The latter doesn't differ from the caller's POV,

it does make a difference if the caller calls flush_workqueue() after
cancel_delayed_work(), in that case flush_workqueue() can miss this
work_struct.
Signed-off-by: NOleg Nesterov <oleg@tv-sign.ru>
Cc: Jarek Poplawski <jarkao2@o2.pl>
Cc: David Howells <dhowells@redhat.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

It is a known fact that freezeable multithreaded workqueues doesn't like
CPU_DEAD. We keep them only for the incoming CPU-hotplug rework.

Sadly, we can't just kill create_freezeable_workqueue() right now, make
them singlethread.
Signed-off-by: NOleg Nesterov <oleg@tv-sign.ru>
Cc: "Rafael J. Wysocki" <rjw@sisk.pl>
Cc: Gautham R Shenoy <ego@in.ibm.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

flush_work(wq, work) doesn't need the first parameter, we can use cwq->wq
(this was possible from the very beginnig, I missed this).  So we can unify
flush_work_keventd and flush_work.

Also, rename flush_work() to cancel_work_sync() and fix all callers.
Perhaps this is not the best name, but "flush_work" is really bad.

(akpm: this is why the earlier patches bypassed maintainers)
Signed-off-by: NOleg Nesterov <oleg@tv-sign.ru>
Cc: Jeff Garzik <jeff@garzik.org>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Jens Axboe <jens.axboe@oracle.com>
Cc: Tejun Heo <htejun@gmail.com>
Cc: Auke Kok <auke-jan.h.kok@intel.com>,
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

We don't have any users, and it is not so trivial to use NOAUTOREL works
correctly.  It is better to simplify API.

Delete NOAUTOREL support and rename work_release to work_clear_pending to
avoid a confusion.
Signed-off-by: NOleg Nesterov <oleg@tv-sign.ru>
Acked-by: NDavid Howells <dhowells@redhat.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

cancel_rearming_delayed_workqueue(wq, dwork) doesn't need the first
parameter.  We don't hang on un-queued dwork any longer, and work->data
doesn't change its type.  This means we can always figure out "wq" from
dwork when it is needed.

Remove this parameter, and rename the function to
cancel_rearming_delayed_work().  Re-create an inline "obsolete"
cancel_rearming_delayed_workqueue(wq) which just calls
cancel_rearming_delayed_work().
Signed-off-by: NOleg Nesterov <oleg@tv-sign.ru>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Because it has no callers.

Actually, I think the whole idea of run_scheduled_work() was not right, not
good to mix "unqueue this work and execute its ->func()" in one function.
Signed-off-by: NOleg Nesterov <oleg@tv-sign.ru>
Cc: Ingo Molnar <mingo@elte.hu>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

A basic problem with flush_scheduled_work() is that it blocks behind _all_
presently-queued works, rather than just the work whcih the caller wants to
flush.  If the caller holds some lock, and if one of the queued work happens
to want that lock as well then accidental deadlocks can occur.

One example of this is the phy layer: it wants to flush work while holding
rtnl_lock().  But if a linkwatch event happens to be queued, the phy code will
deadlock because the linkwatch callback function takes rtnl_lock.

So we implement a new function which will flush a *single* work - just the one
which the caller wants to free up.  Thus we avoid the accidental deadlocks
which can arise from unrelated subsystems' callbacks taking shared locks.

flush_work() non-blockingly dequeues the work_struct which we want to kill,
then it waits for its handler to complete on all CPUs.

Add ->current_work to the "struct cpu_workqueue_struct", it points to
currently running "struct work_struct". When flush_work(work) detects
->current_work == work, it inserts a barrier at the _head_ of ->worklist
(and thus right _after_ that work) and waits for completition. This means
that the next work fired on that CPU will be this barrier, or another
barrier queued by concurrent flush_work(), so the caller of flush_work()
will be woken before any "regular" work has a chance to run.

When wait_on_work() unlocks workqueue_mutex (or whatever we choose to protect
against CPU hotplug), CPU may go away. But in that case take_over_work() will
move a barrier we queued to another CPU, it will be fired sometime, and
wait_on_work() will be woken.

Actually, we are doing cleanup_workqueue_thread()->kthread_stop() before
take_over_work(), so cwq->thread should complete its ->worklist (and thus
the barrier), because currently we don't check kthread_should_stop() in
run_workqueue(). But even if we did, everything should be ok.

[akpm@osdl.org: cleanup]
[akpm@osdl.org: add flush_work_keventd() wrapper]
Signed-off-by: NOleg Nesterov <oleg@tv-sign.ru>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Add a new deferrable delayed work init.  This can be used to schedule work
that are 'unimportant' when CPU is idle and can be called later, when CPU
eventually comes out of idle.

Use this init in cpufreq ondemand governor.
Signed-off-by: NVenkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Cc: Dave Jones <davej@codemonkey.org.uk>
Cc: Oleg Nesterov <oleg@tv-sign.ru>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

del_timer_sync() buys nothing for cancel_delayed_work(), but it is less
efficient since it locks the timer unconditionally, and may wait for the
completion of the delayed_work_timer_fn().

cancel_delayed_work() == 0 means:

	before this patch:
		work->func may still be running or queued

	after this patch:
		work->func may still be running or queued, or
		delayed_work_timer_fn->__queue_work() in progress.

		The latter doesn't differ from the caller's POV,
		delayed_work_timer_fn() is called with _PENDING
		bit set.

cancel_delayed_work() == 1 with this patch adds a new possibility:

	delayed_work->work was cancelled, but delayed_work_timer_fn
	is still running (this is only possible for the re-arming
	works on single-threaded workqueue).

	In this case the timer was re-started by work->func(), nobody
	else can do this. This in turn means that delayed_work_timer_fn
	has already passed __queue_work() (and wont't touch delayed_work)
	because nobody else can queue delayed_work->work.
Signed-off-by: NOleg Nesterov <oleg@tv-sign.ru>
Signed-Off-By: NDavid Howells <dhowells@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

On architectures where the atomicity of the bit operations is handled by
external means (ie a separate spinlock to protect concurrent accesses),
just doing a direct assignment on the workqueue data field (as done by
commit 4594bf15) can cause the
assignment to be lost due to lack of serialization with the bitops on
the same word.

So we need to serialize the assignment with the locks on those
architectures (notably older ARM chips, PA-RISC and sparc32).

So rather than using an "unsigned long", let's use "atomic_long_t",
which already has a safe assignment operation (atomic_long_set()) on
such architectures.

This requires that the atomic operations use the same atomicity locks as
the bit operations do, but that is largely the case anyway.  Sparc32
will probably need fixing.

Architectures (including modern ARM with LL/SC) that implement sane
atomic operations for SMP won't see any of this matter.

Cc: Russell King <rmk+lkml@arm.linux.org.uk>
Cc: David Howells <dhowells@redhat.com>
Cc: David Miller <davem@davemloft.com>
Cc: Matthew Wilcox <matthew@wil.cx>
Cc: Linux Arch Maintainers <linux-arch@vger.kernel.org>
Cc: Andrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Nobody uses it, but it was still wrong.  Using the macro argument name
'work' meant that when we used 'work' as a member name, that would also
get replaced by the macro argument.
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

This allows workqueue users to run just their own pending work, rather
than wait for the whole workqueue to finish running.  This solves the
deadlock with networking libphy that was due to other workqueue entries
possibly needing a lock that was held by the routine that wanted to
flush its own work.

It's not wonderful: if you absolutely need to synchronize with the work
function having been executed, any user strictly speaking should have
its own completion tracking logic, since when we run things explicitly
by hand, the generic workqueue layer can no longer help us synchronize.

Also, this is strictly only usable for work that has been scheduled
without any delayed timers.  You can not mix the new interface with
schedule_delayed_work().

But it's better than what we had currently.
Acked-by: NMaciej W. Rozycki <macro@linux-mips.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Make it possible to create a workqueue the worker thread of which will be
frozen during suspend, along with other kernel threads.
Signed-off-by: NRafael J. Wysocki <rjw@sisk.pl>
Acked-by: NPavel Machek <pavel@ucw.cz>
Cc: Nigel Cunningham <nigel@suspend2.net>
Cc: David Chinner <dgc@sgi.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Pass the work_struct pointer to the work function rather than context data.
The work function can use container_of() to work out the data.

For the cases where the container of the work_struct may go away the moment the
pending bit is cleared, it is made possible to defer the release of the
structure by deferring the clearing of the pending bit.

To make this work, an extra flag is introduced into the management side of the
work_struct.  This governs auto-release of the structure upon execution.

Ordinarily, the work queue executor would release the work_struct for further
scheduling or deallocation by clearing the pending bit prior to jumping to the
work function.  This means that, unless the driver makes some guarantee itself
that the work_struct won't go away, the work function may not access anything
else in the work_struct or its container lest they be deallocated..  This is a
problem if the auxiliary data is taken away (as done by the last patch).

However, if the pending bit is *not* cleared before jumping to the work
function, then the work function *may* access the work_struct and its container
with no problems.  But then the work function must itself release the
work_struct by calling work_release().

In most cases, automatic release is fine, so this is the default.  Special
initiators exist for the non-auto-release case (ending in _NAR).
Signed-Off-By: NDavid Howells <dhowells@redhat.com>

Reclaim a word from the size of the work_struct by folding the pending bit and
the wq_data pointer together.  This shouldn't cause misalignment problems as
all pointers should be at least 4-byte aligned.
Signed-Off-By: NDavid Howells <dhowells@redhat.com>

Define a type for the work function prototype.  It's not only kept in the
work_struct struct, it's also passed as an argument to several functions.

This makes it easier to change it.
Signed-Off-By: NDavid Howells <dhowells@redhat.com>

Separate delayable work items from non-delayable work items be splitting them
into a separate structure (delayed_work), which incorporates a work_struct and
the timer_list removed from work_struct.

The work_struct struct is huge, and this limits it's usefulness.  On a 64-bit
architecture it's nearly 100 bytes in size.  This reduces that by half for the
non-delayable type of event.
Signed-Off-By: NDavid Howells <dhowells@redhat.com>

Add queue_delayed_work_on() interface for workqueues.
Signed-off-by: NAlexey Starikovskiy <alexey.y.starikovskiy@intel.com>
Signed-off-by: NVenkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Signed-off-by: NDave Jones <davej@redhat.com>

We have several points in the SCSI stack (primarily for our device
functions) where we need to guarantee process context, but (given the
place where the last reference was released) we cannot guarantee this.

This API gets around the issue by executing the function directly if
the caller has process context, but scheduling a workqueue to execute
in process context if the caller doesn't have it.
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

swap migration's isolate_lru_page() currently uses an IPI to notify other
processors that the lru caches need to be drained if the page cannot be
found on the LRU.  The IPI interrupt may interrupt a processor that is just
processing lru requests and cause a race condition.

This patch introduces a new function run_on_each_cpu() that uses the
keventd() to run the LRU draining on each processor.  Processors disable
preemption when dealing the LRU caches (these are per processor) and thus
executing LRU draining from another process is safe.

Thanks to Lee Schermerhorn <lee.schermerhorn@hp.com> for finding this race
condition.
Signed-off-by: NChristoph Lameter <clameter@sgi.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

This was unexported by Arjan because we have no current users.

However, during a conversion from tasklets to workqueues of the parisc led
functions, we ran across a case where this was needed.  In particular, the
open coded equivalent of cancel_rearming_delayed_workqueue was implemented
incorrectly, which is, I think, all the evidence necessary that this is a
useful API.
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Initial git repository build. I'm not bothering with the full history,
even though we have it. We can create a separate "historical" git
archive of that later if we want to, and in the meantime it's about
3.2GB when imported into git - space that would just make the early
git days unnecessarily complicated, when we don't have a lot of good
infrastructure for it.

Let it rip!