From mboxrd@z Thu Jan 1 00:00:00 1970 From: Pavel Machek Subject: Re: syncing the disks when entering sleep Date: Tue, 26 Jan 2010 14:59:43 +0100 Message-ID: <20100126135942.GD1764@ucw.cz> References: <76FA3B279DD9DA48896E2B404944957204E0B241@USA7061MS02.na.xerox.net> <201001222149.46839.rjw@sisk.pl> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: Content-Disposition: inline In-Reply-To: <201001222149.46839.rjw@sisk.pl> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: linux-pm-bounces@lists.linux-foundation.org Errors-To: linux-pm-bounces@lists.linux-foundation.org To: "Rafael J. Wysocki" Cc: linux-pm@lists.linux-foundation.org, "Leisner, Martin" List-Id: linux-pm@vger.kernel.org Hi! > > I noticed when we enter sleep states, it seems there's always a > > sys_sync given. > > > > When having an embedded system doing WOL (already in laptop mode to > > cache > > all the writes), it reasonable to wake up dozens of times of day and > > quickly go back to sleep. If the disk is not spun up, > > and there's some work to be done, won't the sys_sync spin up the disk > > before going to sleep? > > > > It would be good to have a knob (in /sys) you can tweak from usespace to > > > > configure the behavior of whether you want to do a sync when entering > > sleep. > > > > The ideal behavior would be: > > > > if(disk is spun up) > > then let the sync happen > > I'm not against that. Patch welcome. :-) I'd say such knob would be ugly. But maybe acceptable way would be echo mem-nosync > power, or maybe it can already be done using s2disk ioctl interface...? Pavel -- (english) http://www.livejournal.com/~pavelmachek (cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html