From mboxrd@z Thu Jan 1 00:00:00 1970 From: Pavel Machek Subject: Re: swsusp status report Date: Thu, 27 Jul 2006 20:40:46 +0000 Message-ID: <20060727204046.GA4623@ucw.cz> References: <200607251325.14747.rjw@sisk.pl> Mime-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Return-path: Content-Disposition: inline In-Reply-To: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: linux-pm-bounces@lists.osdl.org Errors-To: linux-pm-bounces@lists.osdl.org To: Alan Stern Cc: Linux PM List-Id: linux-pm@vger.kernel.org Hi! > > VIII. Saving and loading the suspend image > > = > > The suspend image is saved to a swap partition in step (7) of the > > suspend-resume cycle and loaded from it in step (9) with the help of st= andard > > block IO callbacks and/or functions designed for accessing swap devices= and/or > > swap files. This code has not changed for a long time, but recently An= drew > > Morton has made it use asynchronous IO (the patches are waiting in the = -mm > > tree now). > > = > > There are almost no problems with this part of swsusp. There have been= only > > a couple of minor bugs found in it, and fixed, for the last 6 months. = Yet, > > it has one major limitation which is that it can only use swap partitio= ns > > for saving the suspend image and only one swap partition can be used at= a > > time. To overcome this limitation I am considering the addition of sup= port > > for swap files to this part of swsusp. > = > The disk I/O while reading or writing the image is very slow. In fact, i= t = > sounds like it's going one sector at a time. Can it be sped up? Do the = > AIO changes accomplish that? Actually yes, it was sector-at-a-time, but it still was ~half of possible speed, so it should not be _that_ bad. -- = Thanks for all the (sleeping) penguins.