Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1422728AbWA1AW4 (ORCPT ); Fri, 27 Jan 2006 19:22:56 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1422730AbWA1AWz (ORCPT ); Fri, 27 Jan 2006 19:22:55 -0500 Received: from smtp.osdl.org ([65.172.181.4]:18324 "EHLO smtp.osdl.org") by vger.kernel.org with ESMTP id S1422728AbWA1AWy (ORCPT ); Fri, 27 Jan 2006 19:22:54 -0500 Date: Fri, 27 Jan 2006 19:22:15 -0500 (EST) From: Linus Torvalds To: "Paul E. McKenney" cc: Ingo Molnar , Andrew Morton , linux-kernel@vger.kernel.org, Arjan van de Ven Subject: Re: [patch, lock validator] fix uidhash_lock <-> RCU deadlock In-Reply-To: <20060126111352.GF4953@us.ibm.com> Message-ID: References: <20060125142307.GA5427@elte.hu> <20060126111352.GF4953@us.ibm.com> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 687 Lines: 19 On Thu, 26 Jan 2006, Paul E. McKenney wrote: > > On Wed, Jan 25, 2006 at 03:23:07PM +0100, Ingo Molnar wrote: > > RCU task-struct freeing can call free_uid(), which is taking > > uidhash_lock - while other users of uidhash_lock are softirq-unsafe. > > I guess I get to feel doubly stupid today. Good catch, great tool!!! I wonder if the right fix wouldn't be to free the user struct early, instead of freeing it from RCU. Hmm? Linus - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/