From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([140.186.70.92]:51190) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1QBqVv-0003uR-1c for qemu-devel@nongnu.org; Mon, 18 Apr 2011 11:37:23 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1QBqVt-00070l-Ni for qemu-devel@nongnu.org; Mon, 18 Apr 2011 11:37:22 -0400 Received: from e9.ny.us.ibm.com ([32.97.182.139]:58614) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1QBqVt-00070Q-Lc for qemu-devel@nongnu.org; Mon, 18 Apr 2011 11:37:21 -0400 Received: from d01relay03.pok.ibm.com (d01relay03.pok.ibm.com [9.56.227.235]) by e9.ny.us.ibm.com (8.14.4/8.13.1) with ESMTP id p3IF9BPN022233 for ; Mon, 18 Apr 2011 11:09:11 -0400 Received: from d01av04.pok.ibm.com (d01av04.pok.ibm.com [9.56.224.64]) by d01relay03.pok.ibm.com (8.13.8/8.13.8/NCO v10.0) with ESMTP id p3IFbJkL345420 for ; Mon, 18 Apr 2011 11:37:19 -0400 Received: from d01av04.pok.ibm.com (loopback [127.0.0.1]) by d01av04.pok.ibm.com (8.14.4/8.13.1/NCO v10.0 AVout) with ESMTP id p3IFbIQp008297 for ; Mon, 18 Apr 2011 11:37:19 -0400 From: Dave Hansen In-Reply-To: <4DA89A64.7040809@codemonkey.ws> References: <1302878695-10256-1-git-send-email-dave@linux.vnet.ibm.com> <4DA87093.7050008@us.ibm.com> <1302885415.16562.3358.camel@nimitz> <4DA87DBD.3030406@us.ibm.com> <1302894918.16562.3880.camel@nimitz> <4DA89A64.7040809@codemonkey.ws> Content-Type: text/plain; charset="ISO-8859-1" Date: Mon, 18 Apr 2011 08:37:14 -0700 Message-ID: <1303141034.9615.2651.camel@nimitz> Mime-Version: 1.0 Content-Transfer-Encoding: 7bit Subject: Re: [Qemu-devel] [PATCH] set VIRTIO_BALLOON_F_MUST_TELL_HOST unconditionally List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Anthony Liguori Cc: Amit Shah , Anthony Liguori , qemu-devel@nongnu.org On Fri, 2011-04-15 at 14:20 -0500, Anthony Liguori wrote: > On 04/15/2011 02:15 PM, Dave Hansen wrote: > > On Fri, 2011-04-15 at 12:17 -0500, Anthony Liguori wrote: > >> If you're in OOM and you need memory, you can't ask the host for more > >> and wait for a response. You have to reclaim it immediately. > > Why not? The call in to the notifier chain the s390 case is > > synchronous. The OOM only affects one task at a time and won't proceed > > elsewhere while this is going on. > > Because if we tell the host, we have to wait for the host to ack which > means we'd sleep waiting for an interrupt. Can you do this in the OOM path? Sure. One of the s390 handlers sleeps today. > > Why do we even _tell_ qemu, though? The MADV_WILLNEED is nice, but far > > from being necessary. We could just skip the entire notification in OOM > > situations. > > There's really no particular reason to other than symmetry. When we get there, it sounds like we can simply _skip_ the notification in the qemu case. -- Dave