Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932252AbVLHTFc (ORCPT ); Thu, 8 Dec 2005 14:05:32 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S932253AbVLHTFc (ORCPT ); Thu, 8 Dec 2005 14:05:32 -0500 Received: from omx1-ext.sgi.com ([192.48.179.11]:24803 "EHLO omx1.americas.sgi.com") by vger.kernel.org with ESMTP id S932252AbVLHTFb (ORCPT ); Thu, 8 Dec 2005 14:05:31 -0500 Date: Thu, 8 Dec 2005 11:05:24 -0800 (PST) From: Christoph Lameter To: Ravikiran G Thirumalai cc: linux-kernel@vger.kernel.org, discuss@x86-64.org Subject: Re: pcibus_to_node value when no pxm info is present for the pci bus In-Reply-To: <20051207223414.GA4493@localhost.localdomain> Message-ID: References: <20051207223414.GA4493@localhost.localdomain> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1068 Lines: 24 On Wed, 7 Dec 2005, Ravikiran G Thirumalai wrote: > Most of the arches seem to return -1 for pcibus_to_node if there is no pxm > kind of proximity information for the pci busses. Arch specific code on > those arches check if nodeid >= 0 before using the nodeid for kmalloc_node > etc. But some code paths in x86_64/i386 does not doe this -- > x86_64/dma_alloc_pages() and e1000 node local descriptor (I am to blame for > the second one). Also, pcibus_to_node seems to be 0 when there is no pxm > info available. kmalloc_node falls back to kmalloc for node == -1. So there does not need to be a check. > The question is, what should be the default pcibus_to_node if there is no > pxm info? Answer seems like -1 -- in which case dma_alloc_pages and e1000 > driver has to be fixed. Why would they have to be fixed? - 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/