From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1761555AbXGLTGX (ORCPT ); Thu, 12 Jul 2007 15:06:23 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1754405AbXGLTGE (ORCPT ); Thu, 12 Jul 2007 15:06:04 -0400 Received: from nz-out-0506.google.com ([64.233.162.224]:64137 "EHLO nz-out-0506.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753119AbXGLTGB (ORCPT ); Thu, 12 Jul 2007 15:06:01 -0400 DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=beta; h=received:message-id:date:from:to:subject:cc:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; b=pi0vRcA4qTyL8SobGCGmiogGZCiqYstSByLclcs4QfO5BVxj1yQWlHbTzpPnVlA3FPCco8Doj/v72mK8h56WVKWVmF3/cXCiN3G6nr+iagu6YMLdE5M036H9V+4YZNk1CJkF4OJjcRucctR4D2gcN/Tmcgpa9ErToS6VmonfPLY= Message-ID: <86802c440707121206i3186fe9chba55fe2e5e5e4c5@mail.gmail.com> Date: Thu, 12 Jul 2007 12:06:00 -0700 From: "Yinghai Lu" To: "Greg KH" Subject: Re: [PATCH] try parent numa_node at first before using default Cc: "Cornelia Huck" , "Stefan Richter" , "Andrew Morton" , "Andi Kleen" , rientjes@google.com, "Christoph Lameter" , "Christoph Hellwig" , "David Miller" , "Linux Kernel Mailing List" , netdev@vger.kernel.org In-Reply-To: <20070712183141.GB22471@kroah.com> MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Content-Disposition: inline References: <200707101641.17672.yinghai.lu@sun.com> <469610D4.70009@s5r6.in-berlin.de> <20070712172310.4d662e1f@gondolin.boeblingen.de.ibm.com> <200707121059.54320.yinghai.lu@sun.com> <20070712183141.GB22471@kroah.com> Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On 7/12/07, Greg KH wrote: > On Thu, Jul 12, 2007 at 10:59:53AM -0700, Yinghai Lu wrote: > Please split this into two separate patches, as they are doing two > different things. One for the driver core, and one for pci devices. they need to applied together, otherwise all pci_dev numa_node will be change to device_add to -1 after pci_device_add. > > Also, what kind of performance issues have you seen with this patch in > place? As your previous patches were incorrect, I am guessing that you > are not able to test this? forcedeth on second node. without this patch in addition to get_mpbus_to_node patch, the numa_node is always 0. LBsuse91AMD64:/sys/devices/pci0000:80/0000:80:08.0 # cat numa_node 1 LBsuse91AMD64:/sys/devices/pci0000:80/0000:80:08.0 # cat net/eth2/numa_node 1 and after initialization LBsuse91AMD64:~ # tail -10 /var/log/messages Jul 13 03:05:16 LBsuse91AMD64 kernel: forcedeth 0000:00:08.0: YHLU gart_map_single - !need_iommu: ffff81011d33d010, 11d33d010, 670 Jul 13 03:05:26 LBsuse91AMD64 kernel: forcedeth 0000:00:08.0: YHLU gart_map_single - !need_iommu: ffff81011d322010, 11d322010, 670 Jul 13 03:05:29 LBsuse91AMD64 kernel: forcedeth 0000:00:08.0: YHLU gart_map_single - !need_iommu: ffff81011d33a810, 11d33a810, 670 Jul 13 03:05:31 LBsuse91AMD64 kernel: forcedeth 0000:00:08.0: YHLU gart_map_single - !need_iommu: ffff81011d33e810, 11d33e810, 670 Jul 13 03:05:31 LBsuse91AMD64 kernel: forcedeth 0000:00:08.0: YHLU gart_map_single - !need_iommu: ffff81011d358810, 11d358810, 670 Jul 13 03:05:31 LBsuse91AMD64 kernel: forcedeth 0000:00:08.0: YHLU gart_map_single - !need_iommu: ffff81011cfe1010, 11cfe1010, 670 Jul 13 03:05:31 LBsuse91AMD64 kernel: forcedeth 0000:00:08.0: YHLU gart_map_single - !need_iommu: ffff81011cfe1810, 11cfe1810, 670 Jul 13 03:05:31 LBsuse91AMD64 kernel: forcedeth 0000:00:08.0: YHLU gart_map_single - !need_iommu: ffff81011c7f9010, 11c7f9010, 670 Jul 13 03:05:31 LBsuse91AMD64 kernel: forcedeth 0000:00:08.0: YHLU gart_map_single - !need_iommu: ffff81011c7f9810, 11c7f9810, 670 Jul 13 03:05:34 LBsuse91AMD64 kernel: forcedeth 0000:00:08.0: YHLU gart_map_single - !need_iommu: ffff81011c2a7010, 11c2a7010, 670 gart_map_single is used all the time via netdev_alloc_skb for forcedth. YH