From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752091AbbCUASf (ORCPT ); Fri, 20 Mar 2015 20:18:35 -0400 Received: from userp1040.oracle.com ([156.151.31.81]:27551 "EHLO userp1040.oracle.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752000AbbCUASe (ORCPT ); Fri, 20 Mar 2015 20:18:34 -0400 Message-ID: <550CB8D1.9030608@oracle.com> Date: Fri, 20 Mar 2015 18:18:25 -0600 From: David Ahern User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.10; rv:31.0) Gecko/20100101 Thunderbird/31.5.0 MIME-Version: 1.0 To: Linus Torvalds CC: linux-mm , LKML Subject: Re: 4.0.0-rc4: panic in free_block References: <550C37C9.2060200@oracle.com> <550CA3F9.9040201@oracle.com> In-Reply-To: <550CA3F9.9040201@oracle.com> Content-Type: text/plain; charset=utf-8; format=flowed Content-Transfer-Encoding: 7bit X-Source-IP: aserv0022.oracle.com [141.146.126.234] Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 3/20/15 4:49 PM, David Ahern wrote: > On 3/20/15 3:17 PM, Linus Torvalds wrote: >> In other words, if I read that sparc asm right (and it is very likely >> that I do *not*), then "objp" is NULL, and that's why you crash. > > That does appear to be why. I put a WARN_ON before > clear_obj_pfmemalloc() if objpp[i] is NULL. I got 2 splats during an > 'allyesconfig' build and the system stayed up. > >> >> That's odd, because we know that objp cannot be NULL in >> kmem_slab_free() (even if we allowed it, like with kfree(), >> remove_vma() cannot possibly have a NULL vma, since ti dereferences it >> multiple times). >> >> So I must be misreading this completely. Somebody with better sparc >> debugging mojo should double-check my logic. How would objp be NULL? > > I'll add checks to higher layers and see if it reveals anything. > > I did ask around and apparently this bug is hit only with the new M7 > processors. DaveM: that's why you are not hitting this. Here's another data point: If I disable NUMA I don't see the problem. Performance drops, but no NULL pointer splats which would have been panics. The 128 cpu ldom with NUMA enabled shows the problem every single time I do a kernel compile (-j 128). With NUMA disabled I have done 3 allyesconfig compiles without hitting the problem. I'll put the compiles into a loop while I head out for dinner.