From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Michael S. Tsirkin" Subject: Re: [PATCH v3] virtio_balloon: Convert "vballoon" kthread into a workqueue Date: Thu, 20 Nov 2014 18:55:30 +0200 Message-ID: <20141120165530.GB7495@redhat.com> References: <1416499397-16669-1-git-send-email-pmladek@suse.cz> <20141120160746.GI14877@htj.dyndns.org> <20141120162543.GA7466@redhat.com> <20141120162624.GA7479@redhat.com> <20141120162935.GK14877@htj.dyndns.org> <20141120164711.GA7495@redhat.com> <20141120164933.GL14877@htj.dyndns.org> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: Content-Disposition: inline In-Reply-To: <20141120164933.GL14877@htj.dyndns.org> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: virtualization-bounces@lists.linux-foundation.org Errors-To: virtualization-bounces@lists.linux-foundation.org To: Tejun Heo Cc: Jiri Kosina , linux-kernel@vger.kernel.org, Petr Mladek , virtualization@lists.linux-foundation.org, Jeff Epler List-Id: virtualization@lists.linuxfoundation.org On Thu, Nov 20, 2014 at 11:49:33AM -0500, Tejun Heo wrote: > On Thu, Nov 20, 2014 at 06:47:11PM +0200, Michael S. Tsirkin wrote: > > > There's cancel_work_sync() to stop the self-requeueing ones. > > > > What happens if queue_work runs while cancel_work_sync is in progress? > > Does it fail to queue? > > cancel_work_sync() is guaranteed to take self-requeueing work items no > matter when it's called or what's going on. External (non-self) > queueings of course should be stopped in other ways. Excellent, thanks a lot. > > > > From that POV a dedicated WQ kept it simple. > > > > > > A dedicated wq doesn't do anything for that. You can't shut down a > > > workqueue with a pending work item on it. destroy_workqueue() will > > > try to drain the target wq, warn if it doesn't finish in certain > > > number of iterations and just keep trying indefinitely. > > > > > > Thanks. > > > > Right, so eventually we'll stop requeueuing and it will succeed? > > Yeah, sure, it's a silly reason to use a separate workqueue tho. > Don't do it that way. > > Thanks. Since there's cancel_work_sync I agree. > -- > tejun