From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755732AbZEKToc (ORCPT ); Mon, 11 May 2009 15:44:32 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1753030AbZEKToU (ORCPT ); Mon, 11 May 2009 15:44:20 -0400 Received: from smtp-out.google.com ([216.239.45.13]:32724 "EHLO smtp-out.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752714AbZEKToT (ORCPT ); Mon, 11 May 2009 15:44:19 -0400 DomainKey-Signature: a=rsa-sha1; s=beta; d=google.com; c=nofws; q=dns; h=date:from:x-x-sender:to:cc:subject:in-reply-to:message-id: references:user-agent:mime-version:content-type:x-system-of-record; b=Cm3deO46i0qPGoy/4RFtWI9YoQR6MQ+f265nb46fRqcpvBLJGLhhIExyt6/HhsMA0 ENGclN57HVnYu8LwidlOg== Date: Mon, 11 May 2009 12:44:16 -0700 (PDT) From: David Rientjes X-X-Sender: rientjes@chino.kir.corp.google.com To: Rik van Riel cc: Greg KH , =?UTF-8?Q?Arve_Hj=C3=B8nnev=C3=A5g?= , Andrew Morton , Nick Piggin , San Mehat , KOSAKI Motohiro , linux-kernel@vger.kernel.org, Greg KH Subject: Re: [patch 1/5] oom: cleanup android low memory killer In-Reply-To: <49FF7B02.9050608@redhat.com> Message-ID: References: <20090504220905.GA29402@kroah.com> <20090504225936.GA9876@kroah.com> <49FF7B02.9050608@redhat.com> User-Agent: Alpine 2.00 (DEB 1167 2008-08-23) MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-System-Of-Record: true Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Mon, 4 May 2009, Rik van Riel wrote: > > There was an interest in a low mem userspace notifier that applications can > > poll() on at configurable low mem levels to react accordingly. This would > > probably address the problem that the Android team is trying to fix. > > This would also make it easier for virtual machines to be > ballooned down to size when the host experiences memory > pressure. > > The low memory notification could also be useful for JVMs, > databases that do (some of) their own caching and any garbage > collected program in general. > I agree, and I'd suggest that it be implemented as a cgroup so that individual sets of tasks can be responsible for their own lowmem notifiers. KOSAKI, you were pushing the /dev/mem_notify patchset last year from Marcelo. Do you have any plans of recreating that in a cgroup or have you dropped it?