From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754648AbaJBP6G (ORCPT ); Thu, 2 Oct 2014 11:58:06 -0400 Received: from gum.cmpxchg.org ([85.214.110.215]:36284 "EHLO gum.cmpxchg.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754416AbaJBP6E (ORCPT ); Thu, 2 Oct 2014 11:58:04 -0400 Date: Thu, 2 Oct 2014 11:57:50 -0400 From: Johannes Weiner To: Michal Hocko Cc: Andrew Morton , Greg Thelen , Vladimir Davydov , Dave Hansen , linux-mm@kvack.org, cgroups@vger.kernel.org, linux-kernel@vger.kernel.org Subject: Re: [patch 1/3] mm: memcontrol: do not kill uncharge batching in free_pages_and_swap_cache Message-ID: <20141002155750.GB2035@cmpxchg.org> References: <1411571338-8178-1-git-send-email-hannes@cmpxchg.org> <1411571338-8178-2-git-send-email-hannes@cmpxchg.org> <20140924124234.3fdb59d6cdf7e9c4d6260adb@linux-foundation.org> <20140924210322.GA11017@cmpxchg.org> <20140925134403.GA11080@dhcp22.suse.cz> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20140925134403.GA11080@dhcp22.suse.cz> Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Thu, Sep 25, 2014 at 03:44:03PM +0200, Michal Hocko wrote: > On Wed 24-09-14 17:03:22, Johannes Weiner wrote: > [...] > > In release_pages, break the lock at least every SWAP_CLUSTER_MAX (32) > > pages, then remove the batching from free_pages_and_swap_cache. > > Actually I had something like that originally but then decided to > not change the break out logic to prevent from strange and subtle > regressions. I have focused only on the memcg batching POV and led the > rest untouched. > > I do agree that lru_lock batching can be improved as well. Your change > looks almost correct but you should count all the pages while the lock > is held otherwise you might happen to hold the lock for too long just > because most pages are off the LRU already for some reason. At least > that is what my original attempt was doing. Something like the following > on top of the current patch: Yep, that makes sense. Would you care to send it in such that Andrew can pick it up? Thanks!