Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S262785AbVA1T2m (ORCPT ); Fri, 28 Jan 2005 14:28:42 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S262775AbVA1TY5 (ORCPT ); Fri, 28 Jan 2005 14:24:57 -0500 Received: from fire.osdl.org ([65.172.181.4]:9857 "EHLO smtp.osdl.org") by vger.kernel.org with ESMTP id S262769AbVA1TVr (ORCPT ); Fri, 28 Jan 2005 14:21:47 -0500 Subject: Re: out of memory question From: Mark Haverkamp To: Matt Domsch Cc: linux-kernel In-Reply-To: <20050128191458.GA21601@lists.us.dell.com> References: <1106934186.27858.40.camel@markh1.pdx.osdl.net> <20050128191458.GA21601@lists.us.dell.com> Content-Type: text/plain Date: Fri, 28 Jan 2005 11:21:42 -0800 Message-Id: <1106940102.27858.41.camel@markh1.pdx.osdl.net> Mime-Version: 1.0 X-Mailer: Evolution 2.0.2 (2.0.2-4) Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1609 Lines: 41 On Fri, 2005-01-28 at 13:14 -0600, Matt Domsch wrote: > On Fri, Jan 28, 2005 at 09:43:06AM -0800, Mark Haverkamp wrote: > > > > I have a situation where the out of memory killer kicked in and killed > > off a process. From the information displayed, it looks like there was > > a lot of free memory available. I need some help interpreting the > > output. I have included the console output from the oom killer. > > > > It is running 2.6.11-rc2 and has a patch from Nick Piggin: > > http://marc.theaimsgroup.com/?l=linux-kernel&m=110665524811826&w=2 > > The machine is running as an iscsi target with 4K luns configured. > > What is eating all of your ZONE_DMA? Of the 16MB available > > > DMA: 55*4kB 4*8kB 2*16kB 0*32kB 1*64kB 0*128kB 0*256kB 0*512kB 0*1024kB 0*2048kB 0*4096kB = 348kB > > DMA free:348kB min:68kB low:84kB high:100kB active:4kB inactive:0kB present:16384kB pages_scanned:0 all_unreclaimable? no > > protections[]: 0 0 0 > > only 348kB is available, and something is requesting more... > > > oom-killer: gfp_mask=0xd1 (__GFP_FS|__GFP_IO|__GFP_WAIT|__GFP_DMA) > > > This isn't a 64-bit architecture (you've got some ZONE_HIGHMEM), so > it's not like the x86_64/ia64 iommu. Perhaps a > <32-bit DMA address mask PCI device? > Thanks for the help. I have somewhere to start looking now. Mark. > -- Mark Haverkamp - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/