Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751678AbcCHUlE (ORCPT ); Tue, 8 Mar 2016 15:41:04 -0500 Received: from resqmta-ch2-05v.sys.comcast.net ([69.252.207.37]:45150 "EHLO resqmta-ch2-05v.sys.comcast.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751582AbcCHUkz (ORCPT ); Tue, 8 Mar 2016 15:40:55 -0500 Date: Tue, 8 Mar 2016 14:40:52 -0600 (CST) From: Christoph Lameter X-X-Sender: cl@east.gentwo.org To: Vineet Gupta cc: linux-mm@kvack.org, Pekka Enberg , David Rientjes , Joonsoo Kim , Andrew Morton , Noam Camus , stable@vger.kernel.org, linux-kernel@vger.kernel.org, linux-snps-arc@lists.infradead.org Subject: Re: [PATCH] mm: slub: Ensure that slab_unlock() is atomic In-Reply-To: <56DEF3D3.6080008@synopsys.com> Message-ID: References: <1457447457-25878-1-git-send-email-vgupta@synopsys.com> <56DEF3D3.6080008@synopsys.com> Content-Type: text/plain; charset=US-ASCII Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 696 Lines: 18 On Tue, 8 Mar 2016, Vineet Gupta wrote: > # set the bit > 80543b8e: ld_s r2,[r13,0] <--- (A) Finds PG_locked is set > 80543b90: or r3,r2,1 <--- (B) other core unlocks right here > 80543b94: st_s r3,[r13,0] <--- (C) sets PG_locked (overwrites unlock) Duh. Guess you need to take the spinlock also in the arch specific implementation of __bit_spin_unlock(). This is certainly not the only case in which we use the __ op to unlock. You need a true atomic op or you need to take the "spinlock" in all cases where you modify the bit. If you take the lock in __bit_spin_unlock then the race cannot happen. > Are you convinced now ! Yes, please fix your arch specific code.