All of lore.kernel.org
 help / color / mirror / Atom feed
From: Andrew Morton <akpm@linux-foundation.org>
To: Ingo Molnar <mingo@elte.hu>
Cc: oleg@redhat.com, a.p.zijlstra@chello.nl, rusty@rustcorp.com.au,
	travis@sgi.com, mingo@redhat.com, davej@redhat.com,
	cpufreq@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH 2/3] work_on_cpu: Use our own workqueue.
Date: Mon, 26 Jan 2009 14:01:16 -0800	[thread overview]
Message-ID: <20090126140116.35f9c173.akpm@linux-foundation.org> (raw)
In-Reply-To: <20090126214516.GA22142@elte.hu>

On Mon, 26 Jan 2009 22:45:16 +0100
Ingo Molnar <mingo@elte.hu> wrote:

> 
> * Andrew Morton <akpm@linux-foundation.org> wrote:
> 
> > On Mon, 26 Jan 2009 22:27:27 +0100
> > Ingo Molnar <mingo@elte.hu> wrote:
> > 
> > > 
> > > * Andrew Morton <akpm@linux-foundation.org> wrote:
> > > 
> > > > > So if it's generic it ought to be implemented in a generic way - not a 
> > > > > "dont use from any codepath that has a lock held that might 
> > > > > occasionally also be held in a keventd worklet". (which is a totally 
> > > > > unmaintainable proposition and which would just cause repeat bugs 
> > > > > again and again.)
> > > > 
> > > > That's different.  The core fault here lies in the keventd workqueue 
> > > > handling code.  If we're flushing work A then we shouldn't go and 
> > > > block behind unrelated work B.
> > > 
> > > the blocking is inherent in the concept of "a queue of worklets 
> > > handled by a single thread".
> > > 
> > > If a worklet is blocked then all other work performed by that thread 
> > > is blocked as well. So by waiting on a piece of work in the queue, we 
> > > wait for all prior work queued up there as well.
> > > 
> > > The only way to decouple that and to make them independent (and hence 
> > > independently flushable) is to create more parallel flows of 
> > > execution: i.e. by creating another thread (another workqueue).
> > > 
> > 
> > Nope.  As I said, the caller of flush_work() can detach the work item 
> > and run it directly.
> 
> that would change the concept of execution but indeed it would be 
> interesting to try. It's outside the scope of late -rcs i guess, but 
> worthwile nevertheless.
> 

Well it turns out that I was having a less-than-usually-senile moment:

: commit b89deed32ccc96098bd6bc953c64bba6b847774f
: Author:     Oleg Nesterov <oleg@tv-sign.ru>
: AuthorDate: Wed May 9 02:33:52 2007 -0700
: Commit:     Linus Torvalds <torvalds@woody.linux-foundation.org>
: CommitDate: Wed May 9 12:30:50 2007 -0700
: 
:     implement flush_work()
:     
:     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.


Why isn't that working in this case??


  reply	other threads:[~2009-01-26 22:02 UTC|newest]

Thread overview: 70+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-01-16 19:11 [PATCH 0/3] cpu freq: fix problems with work_on_cpu usage in acpi-cpufreq Mike Travis
2009-01-16 19:11 ` [PATCH 1/3] work_on_cpu: dont try to get_online_cpus() in work_on_cpu Mike Travis
2009-01-16 19:11 ` [PATCH 2/3] work_on_cpu: Use our own workqueue Mike Travis
2009-01-24  8:15   ` Andrew Morton
     [not found]     ` <200901261711.43943.rusty@rustcorp.com.au>
2009-01-26  7:01       ` Andrew Morton
2009-01-26 17:16         ` Ingo Molnar
2009-01-26 18:35           ` Andrew Morton
2009-01-26 20:20             ` Ingo Molnar
2009-01-26 20:43               ` Mike Travis
2009-01-26 21:00               ` Andrew Morton
2009-01-26 21:27                 ` Ingo Molnar
2009-01-26 21:35                   ` Andrew Morton
2009-01-26 21:45                     ` Ingo Molnar
2009-01-26 22:01                       ` Andrew Morton [this message]
2009-01-26 22:05                         ` Ingo Molnar
2009-01-26 22:16                           ` Andrew Morton
2009-01-26 22:20                             ` Ingo Molnar
2009-01-26 22:50                               ` Andrew Morton
2009-01-26 22:59                                 ` Ingo Molnar
2009-01-26 23:42                                   ` Andrew Morton
2009-01-26 23:53                                     ` Ingo Molnar
2009-01-27  0:42                                       ` Andrew Morton
2009-01-26 22:31                             ` Oleg Nesterov
2009-01-26 22:15                         ` Oleg Nesterov
2009-01-26 22:24                           ` Ingo Molnar
2009-01-26 22:37                             ` Oleg Nesterov
2009-01-26 22:42                               ` Ingo Molnar
2009-01-26 21:50                     ` Oleg Nesterov
2009-01-26 22:17                       ` Ingo Molnar
2009-01-26 23:01                         ` Mike Travis
2009-01-27  0:09                           ` Oleg Nesterov
2009-01-27  7:15                         ` Rusty Russell
2009-01-27 17:55                           ` Oleg Nesterov
2009-01-27  7:05         ` Rusty Russell
2009-01-27  7:25           ` Andrew Morton
2009-01-27 15:28             ` Ingo Molnar
2009-01-27 16:51               ` Andrew Morton
2009-01-28 13:02             ` Rusty Russell
2009-01-28 17:19               ` Mike Travis
2009-01-28 17:32                 ` Mike Travis
2009-01-29 10:39                   ` Rusty Russell
2009-01-28 19:44               ` Andrew Morton
2009-01-29  1:43                 ` Rusty Russell
2009-01-29  2:12                   ` Andrew Morton
2009-01-30  6:03                     ` Rusty Russell
2009-01-30  6:30                       ` Andrew Morton
2009-01-30 13:49                         ` Ingo Molnar
2009-01-30 17:08                           ` Andrew Morton
2009-01-30 21:59                         ` Rusty Russell
2009-01-30 22:17                           ` Andrew Morton
2009-02-02 12:35                             ` Rusty Russell
2009-02-03  4:06                               ` Andrew Morton
2009-02-04  2:44                                 ` Rusty Russell
2009-02-04  3:01                                   ` Andrew Morton
2009-02-04 10:41                                     ` Rusty Russell
2009-02-04 15:36                                       ` Andrew Morton
2009-02-04 21:35                                         ` Ingo Molnar
2009-02-04 21:48                                           ` Andrew Morton
2009-02-04 21:54                                             ` Ingo Molnar
2009-02-04 23:45                                             ` Rusty Russell
2009-02-05 12:19                                             ` Pavel Machek
2009-02-05 17:44                                             ` Dmitry Adamushko
2009-02-10  8:54                                         ` Rusty Russell
2009-02-10  9:35                                           ` Andrew Morton
2009-02-11  0:32                                             ` Rusty Russell
2009-01-16 19:11 ` [PATCH 3/3] cpufreq: use work_on_cpu in acpi-cpufreq.c for drv_read and drv_write Mike Travis
2009-01-16 23:38 ` [PATCH 0/3] cpu freq: fix problems with work_on_cpu usage in acpi-cpufreq [PULL request] Mike Travis
2009-01-17 22:08   ` Ingo Molnar
2009-01-19 17:11     ` Mike Travis
2009-01-19 17:26       ` Ingo Molnar

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20090126140116.35f9c173.akpm@linux-foundation.org \
    --to=akpm@linux-foundation.org \
    --cc=a.p.zijlstra@chello.nl \
    --cc=cpufreq@vger.kernel.org \
    --cc=davej@redhat.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@elte.hu \
    --cc=mingo@redhat.com \
    --cc=oleg@redhat.com \
    --cc=rusty@rustcorp.com.au \
    --cc=travis@sgi.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.