From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754776Ab3AQXUk (ORCPT ); Thu, 17 Jan 2013 18:20:40 -0500 Received: from hrndva-omtalb.mail.rr.com ([71.74.56.122]:3887 "EHLO hrndva-omtalb.mail.rr.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752298Ab3AQXUi (ORCPT ); Thu, 17 Jan 2013 18:20:38 -0500 X-Authority-Analysis: v=2.0 cv=UsXhxpMB c=1 sm=0 a=rXTBtCOcEpjy1lPqhTCpEQ==:17 a=mNMOxpOpBa8A:10 a=WIJlFdhv0j0A:10 a=5SG0PmZfjMsA:10 a=Q9fys5e9bTEA:10 a=meVymXHHAAAA:8 a=_FnEdowTgBoA:10 a=e0rM7KXBUBkOxtCy9Q8A:9 a=PUjeQqilurYA:10 a=jeBq3FmKZ4MA:10 a=rXTBtCOcEpjy1lPqhTCpEQ==:117 X-Cloudmark-Score: 0 X-Authenticated-User: X-Originating-IP: 74.67.115.198 Message-ID: <1358464837.23211.66.camel@gandalf.local.home> Subject: [RFC][PATCH] slub: Keep page and object in sync in slab_alloc_node() From: Steven Rostedt To: Christoph Lameter Cc: LKML , linux-mm , Andrew Morton , Pekka Enberg , Matt Mackall , Thomas Gleixner , RT , Clark Williams , John Kacur , "Luis Claudio R. Goncalves" Date: Thu, 17 Jan 2013 18:20:37 -0500 In-Reply-To: <1358464245.23211.62.camel@gandalf.local.home> References: <1358446258.23211.32.camel@gandalf.local.home> <1358447864.23211.34.camel@gandalf.local.home> <0000013c4a69a2cf-1a19a6f6-e6a3-4f06-99a4-10fdd4b9aca2-000000@email.amazonses.com> <1358458996.23211.46.camel@gandalf.local.home> <0000013c4a7e7fbf-c51fd42a-2455-4fec-bb37-915035956f05-000000@email.amazonses.com> <1358462763.23211.57.camel@gandalf.local.home> <1358464245.23211.62.camel@gandalf.local.home> Content-Type: text/plain; charset="ISO-8859-15" X-Mailer: Evolution 3.4.4-1 Mime-Version: 1.0 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org In slab_alloc_node(), after the cpu_slab is assigned, if the task is preempted and moves to another CPU, there's nothing keeping the page and object in sync. The -rt kernel crashed because page was NULL and object was not, and the node_match() dereferences page. Even though the crash happened on -rt, there's nothing that's keeping this from happening on mainline. The easiest fix is to disable interrupts for the entire time from acquiring the current CPU cpu_slab and assigning the object and page. After that, it's fine to allow preemption. Signed-off-by: Steven Rostedt diff --git a/mm/slub.c b/mm/slub.c index 2fdd96f..c00ce80 100644 --- a/mm/slub.c +++ b/mm/slub.c @@ -2317,6 +2317,8 @@ static __always_inline void *slab_alloc(struct kmem_cache *s, struct kmem_cache_cpu *c; struct page *page; unsigned long tid; + unsigned long flags; + int new_object; if (slab_pre_alloc_hook(s, gfpflags)) return NULL; @@ -2328,7 +2330,10 @@ redo: * enabled. We may switch back and forth between cpus while * reading from one cpu area. That does not matter as long * as we end up on the original cpu again when doing the cmpxchg. + * + * But we need to sync the setting of page and object. */ + local_irq_save(flags); c = __this_cpu_ptr(s->cpu_slab); /* @@ -2338,11 +2343,14 @@ redo: * linked list in between. */ tid = c->tid; - barrier(); object = c->freelist; page = c->page; - if (unlikely(!object || !node_match(page, node))) + + new_object = !object || !node_match(page, node); + local_irq_restore(flags); + + if (new_object) object = __slab_alloc(s, gfpflags, node, addr, c); else {