From mboxrd@z Thu Jan 1 00:00:00 1970 From: Zhang Rui Subject: Re: [RFC][PATCH 0/3] PM: Asynchronous suspend and resume Date: Thu, 13 Aug 2009 16:18:29 +0800 Message-ID: <1250151509.2906.61.camel@rzhang-dt> References: <200908122343.20554.rjw@sisk.pl> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: Received: from mga02.intel.com ([134.134.136.20]:36065 "EHLO mga02.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752935AbZHMITh (ORCPT ); Thu, 13 Aug 2009 04:19:37 -0400 In-Reply-To: <200908122343.20554.rjw@sisk.pl> Sender: linux-acpi-owner@vger.kernel.org List-Id: linux-acpi@vger.kernel.org To: "Rafael J. Wysocki" Cc: Alan Stern , linux-pm , linux-acpi , Linux Kernel Mailing List , Len Brown , Arjan van de Ven On Thu, 2009-08-13 at 05:43 +0800, Rafael J. Wysocki wrote: > On Wednesday 12 August 2009, Alan Stern wrote: > > On Wed, 12 Aug 2009, Rafael J. Wysocki wrote: > >=20 > > > Hi, > > >=20 > > > The following patches introduce a mechanism allowing us to execut= e device > > > drivers' suspend and resume callbacks asynchronously during syste= m sleep > > > transitions, such as suspend to RAM. The idea is explained in th= e [1/1] patch > > > message. > > >=20 > > > Comments welcome. > >=20 > > I get the idea. Not bad. >=20 > Thanks! >=20 > > Have you tried it in a serious way? For example, turning on the > > async_suspend flag for every device? >=20 > No, I've only tested it with a few selected drivers. I'm going to tr= y the > "async everyone" scenario, though. >=20 > > In one way it isn't as efficient as it could be. You fire off a bu= nch > > of async threads and then make many of them wait for parent or chil= d > > devices. They could be doing useful work instead. >=20 =EF=BB=BF=EF=BB=BFare you talking about this scenario, or I find anothe= r problem of this approach: there is a part of dpm_list, dev1->dev_aaa->...->dev_bbb->dev2 dev2 is dev1's first child. dev1 resume takes 1s dev_aaa~dev_bbb resume takes 0.1s. if we call =EF=BB=BFdevice_enable_async_suspend(dev1, true) in order to= resume device1 asynchronously, the real asynchronous resume only happens between dev1 and dev_aaa to dev_bbb because dev2 needs to wait until dev1 resume finished. so kernel schedules dev1 resume in an async thread first, and then take= s 0.1s to finish the dev_aaa to dev_bbb resume, and then sleep 0.9s > I kind of agree, but then the patches would be more complicated. >=20 The problem is that we need to invoke device_resume for every device synchronously. I wonder if we can make the child devices inherit the parent's =EF=BB=BFdev->power.async_suspend flag, so that devices that n= eed to wait are resumed asynchronously, i.e. we never wait/sleep when parsing the dpm_list. this doesn't bring too much benefit in suspend case but it can speed up the resume process a lot. Of cause, this is not a problem if we turn on the async_suspend flag fo= r every device. > > It would be interesting to invent a way of representing explicitly = the=20 > > non-tree dependencies -- assuming there aren't too many of them! (= I=20 > > can just hear the TI guys hollering about power and timer domains..= =2E) >=20 > I have an idea. >=20 > Every such dependency involves two devices, one of which is a "master= " > and the second of which is a "slave", meaning that the "slave" have t= o be > suspended before the "master" and cannot be resumed before it. In pr= inciple > we could give each device two lists of "dependency objects", one cont= aining > "dependency objects" where the device is the "master" and the other c= ontaining > "dependency objects" where the device is the "slave". Then, each "de= pendency > object" could be represented as >=20 > struct pm_connection { > struct device *master; > struct list_head master_hook; > struct device *slave; > struct list_head slave_hook; > }; >=20 > Add some locking, helpers for adding / removing "dependency objects" = etc. > and it should work. Instead of checking the parent, walk the list of > "masters", instead of walking the list of children, walk the list of = "slaves". >=20 > The core could create those objects for parent-child relationships > automatically, the other ones would have to be added by platforms / b= us types / > drivers etc. >=20 =EF=BB=BFthis sounds great. :) thanks, rui > This approach has a problem that it's prone to adding circular depend= encies by > mistake, but then I think it would apply to any other approach just a= s well. >=20 > Best, > Rafael -- To unsubscribe from this list: send the line "unsubscribe linux-acpi" i= n the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html