From mboxrd@z Thu Jan 1 00:00:00 1970 From: Vivek Goyal Subject: Re: [PATCH 06/12] cfq-iosched: implement cfq_group->nr_active and ->level_weight Date: Mon, 17 Dec 2012 16:18:43 -0500 Message-ID: <20121217211843.GA13691@redhat.com> References: <1355524885-22719-1-git-send-email-tj@kernel.org> <1355524885-22719-7-git-send-email-tj@kernel.org> <20121217204609.GH7235@redhat.com> <20121217211517.GC1844@htj.dyndns.org> Mime-Version: 1.0 Return-path: Content-Disposition: inline In-Reply-To: <20121217211517.GC1844-Gd/HAXX7CRxy/B6EtB590w@public.gmane.org> Sender: cgroups-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org List-ID: Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: Tejun Heo Cc: lizefan-hv44wF8Li93QT0dZR+AlfA@public.gmane.org, axboe-tSWWG44O7X1aa/9Udqfwiw@public.gmane.org, containers-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org, cgroups-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, ctalbott-hpIqsD4AKlfQT0dZR+AlfA@public.gmane.org, rni-hpIqsD4AKlfQT0dZR+AlfA@public.gmane.org On Mon, Dec 17, 2012 at 01:15:17PM -0800, Tejun Heo wrote: > Hello, Vivek. > > On Mon, Dec 17, 2012 at 03:46:09PM -0500, Vivek Goyal wrote: > > On Fri, Dec 14, 2012 at 02:41:19PM -0800, Tejun Heo wrote: > > > To prepare for blkcg hierarchy support, add cfqg->nr_active and > > > ->level_weight. cfqg->nr_active counts the number of active cfqgs at > > > the cfqg's level and ->level_weight is sum of weights of those cfqgs. > > > The level covers itself (cfqg->leaf_weight) and immediate children. > > > > This notion of level is really confusing. If one says "at cfqg's level" > > I immediately associate with cfqg's siblings and not with cfqg's children. > > We can explicitly say at children's level but I think it should be > enough to explain it clearly in the comment where the field is > defined. > > > I think confusion happens because we are overloading the definition of > > cfqg. It is competing with its siblings at the same time it is competing > > against its child groups (on behalf of its children tasks). > > While I agree that part is a bit tricky, I can't think of a much > better way to describe it. Any better ideas? Can we call it cfqg->children_weight insted of cfqg->level_weight. Thanks Vivek