From mboxrd@z Thu Jan 1 00:00:00 1970 Date: Tue, 8 Jan 2008 10:40:16 +0900 From: KAMEZAWA Hiroyuki Subject: Re: [BUG] at mm/slab.c:3320 Message-Id: <20080108104016.4fa5a4f3.kamezawa.hiroyu@jp.fujitsu.com> In-Reply-To: References: <20071220100541.GA6953@skywalker> <20071225140519.ef8457ff.akpm@linux-foundation.org> <20071227153235.GA6443@skywalker> <20071228051959.GA6385@skywalker> <20080103155046.GA7092@skywalker> <20080107102301.db52ab64.kamezawa.hiroyu@jp.fujitsu.com> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-linux-mm@kvack.org Return-Path: To: Christoph Lameter Cc: "Aneesh Kumar K.V" , linux-mm@kvack.org, nacc@us.ibm.com, lee.schermerhorn@hp.com, bob.picco@hp.com, mel@skynet.ie List-ID: On Mon, 7 Jan 2008 10:10:16 -0800 (PST) Christoph Lameter wrote: > On Mon, 7 Jan 2008, KAMEZAWA Hiroyuki wrote: > > > Seems Node 1 has no NORMAL memory. > > > > Because the patch changes 'online_node' to N_NORMAL_MEMORY, there is a change. > > I'm not sure but cachep->nodelists[] should be created against all online nodes ? > > Well what is the point of creating a memory structure for a node from > which no memory for the slab allocator can be allocated? I think we need a > special cpu_to_node() that only takes normal memory into consideration. > In usual alloc_pages() allocator, this is done by zonelist fallback. > And we need to use that new function (cpu_to_node_normal_memory or so?) to > find memory for the slab and other stuff in the kernel. > It seems that cache->nodelists[nid] == NULL case should be handled even if nid == cpu_to_node(smp_processor_id()). complicated ? Thanks, -Kame -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: email@kvack.org