Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932506AbcJZNnI (ORCPT ); Wed, 26 Oct 2016 09:43:08 -0400 Received: from mail-wm0-f67.google.com ([74.125.82.67]:38655 "EHLO mail-wm0-f67.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754321AbcJZNnF (ORCPT ); Wed, 26 Oct 2016 09:43:05 -0400 Date: Wed, 26 Oct 2016 15:43:01 +0200 From: Robert Richter To: David Daney Cc: linux-kernel@vger.kernel.org, Rob Herring , Frank Rowand , devicetree@vger.kernel.org, linux-arm-kernel@lists.infradead.org, Will Deacon , Catalin Marinas , Hanjun Guo , Ganapatrao Kulkarni , Gilbert Netzer , David Daney Subject: Re: [PATCH 1/2] of, numa: Add function to disable of_node_to_nid(). Message-ID: <20161026134301.GV25086@rric.localdomain> References: <1477431061-7258-1-git-send-email-ddaney.cavm@gmail.com> <1477431061-7258-2-git-send-email-ddaney.cavm@gmail.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1477431061-7258-2-git-send-email-ddaney.cavm@gmail.com> User-Agent: Mutt/1.5.23 (2014-03-12) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1913 Lines: 49 On 25.10.16 14:31:00, David Daney wrote: > From: David Daney > > On arm64 NUMA kernels we can pass "numa=off" on the command line to > disable NUMA. A side effect of this is that kmalloc_node() calls to > non-zero nodes will crash the system with an OOPS: > > [ 0.000000] [] __alloc_pages_nodemask+0xa4/0xe68 > [ 0.000000] [] new_slab+0xd0/0x57c > [ 0.000000] [] ___slab_alloc+0x2e4/0x514 > [ 0.000000] [] __slab_alloc+0x48/0x58 > [ 0.000000] [] __kmalloc_node+0xd0/0x2e0 > [ 0.000000] [] __irq_domain_add+0x7c/0x164 > [ 0.000000] [] its_probe+0x784/0x81c > [ 0.000000] [] its_init+0x48/0x1b0 > . > . > . > > This is caused by code like this in kernel/irq/irqdomain.c > > domain = kzalloc_node(sizeof(*domain) + (sizeof(unsigned int) * size), > GFP_KERNEL, of_node_to_nid(of_node)); > > When NUMA is disabled, the concept of a node is really undefined, so > of_node_to_nid() should unconditionally return NUMA_NO_NODE. > > Add __of_force_no_numa() to allow of_node_to_nid() to be forced to > return NUMA_NO_NODE. > > The follow on patch will call this new function from the arm64 numa > code. Didn't that work before? numa=off just maps all mem to node 0. If mem allocation is requested for another node it should just fall back to a node with mem (node 0 then). I suspect there is something wrong with the page initialization, see: http://www.spinics.net/lists/arm-kernel/msg535191.html https://bugzilla.redhat.com/show_bug.cgi?id=1387793 What is the complete oops? So I think k*alloc_node() must be able to handle requests to non-existing nodes. Otherwise your fix is incomplete, assume a failed of_numa_init() causing a dummy init but still some devices reporting a node. -Robert