From mboxrd@z Thu Jan 1 00:00:00 1970 From: Tejun Heo Subject: Re: [PATCH] memcg: add per cgroup dirty page accounting Date: Mon, 9 Mar 2015 11:31:16 -0400 Message-ID: <20150309153116.GV13283@htj.duckdns.org> References: <1425876632-6681-1-git-send-email-gthelen@google.com> <20150309135234.GU13283@htj.duckdns.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Andrew Morton , Johannes Weiner , Michal Hocko , Konstantin Khebnikov , Dave Chinner , Sha Zhengju , Kamezawa Hiroyuki , linux-fsdevel@vger.kernel.org, "linux-mm@kvack.org" To: Greg Thelen Return-path: Received: from mail-qc0-f170.google.com ([209.85.216.170]:41809 "EHLO mail-qc0-f170.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750930AbbCIPbU (ORCPT ); Mon, 9 Mar 2015 11:31:20 -0400 Received: by qcrw7 with SMTP id w7so5905520qcr.8 for ; Mon, 09 Mar 2015 08:31:20 -0700 (PDT) Content-Disposition: inline In-Reply-To: Sender: linux-fsdevel-owner@vger.kernel.org List-ID: On Mon, Mar 09, 2015 at 11:29:05AM -0400, Greg Thelen wrote: > On Mon, Mar 9, 2015 at 9:52 AM, Tejun Heo wrote: > > Hello, Greg. > > > > On Sun, Mar 08, 2015 at 09:50:32PM -0700, Greg Thelen wrote: > >> When modifying PG_Dirty on cached file pages, update the new > >> MEM_CGROUP_STAT_DIRTY counter. This is done in the same places where > >> global NR_FILE_DIRTY is managed. The new memcg stat is visible in the > >> per memcg memory.stat cgroupfs file. The most recent past attempt at > >> this was http://thread.gmane.org/gmane.linux.kernel.cgroups/8632 > > > > Awesome. I had a similar but inferior (haven't noticed the irqsave > > problem) patch in my series. Replaced that with this one. I'm > > getting ready to post the v2 of the cgroup writeback patchset. Do you > > mind routing this patch together in the patchset? > > I don't object to routing this patch with the larger writeback series. > But I do have small concern that merging the writeback series might > take a while and this patch has independent value. For now, I'd say: > go for it. If the series gets stalled we might want to split it off. Yeah, sure, either is fine for me. Hmm... I'm gonna move this patch to the head of the series so that it can go either way. Thanks. -- tejun