From mboxrd@z Thu Jan 1 00:00:00 1970 From: James Bottomley Subject: Re: [RFC] pm_qos: reimplement using plists Date: Sat, 05 Jun 2010 13:19:29 -0500 Message-ID: <1275761969.7227.19.camel@mulgrave.site> References: <1275760688.7227.10.camel@mulgrave.site> <201006052017.22856.rjw@sisk.pl> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <201006052017.22856.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, markgross@thegnar.org List-Id: linux-pm@vger.kernel.org On Sat, 2010-06-05 at 20:17 +0200, Rafael J. Wysocki wrote: > On Saturday 05 June 2010, James Bottomley wrote: > > A lot of the pm_qos extremal value handling is really duplicating what a > > priority ordered list does, just in a less efficient fashion. Simply > > redoing the implementation in terms of a plist gets rid of a lot of this > > junk (although there are several other strange things that could do with > > tidying up, like pm_qos_request_list has to carry the pm_qos_class with > > every node, simply because it doesn't get passed in to > > pm_qos_update_request even though every caller knows full well what > > parameter it's updating). > > > > I think this redo is a win independent of android, so we should do > > something like this now. > > I like it. > > > There is one nasty that should probably be fixed in plists not open > > coded here: plist_first gives the highest priority value, but there's no > > corresponding API to give the lowest (even though you can get it from > > the head.nodes_list.prev) ... if the sched people are OK, I'll correct > > this with the final patch set. > > I assume you'll send a signed-off version at one point. Yes ... probably as a two parter with the plist changes. James