From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932345AbWGEDYk (ORCPT ); Tue, 4 Jul 2006 23:24:40 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S932349AbWGEDYk (ORCPT ); Tue, 4 Jul 2006 23:24:40 -0400 Received: from mxl145v69.mxlogic.net ([208.65.145.69]:3237 "EHLO p02c11o146.mxlogic.net") by vger.kernel.org with ESMTP id S932345AbWGEDYj (ORCPT ); Tue, 4 Jul 2006 23:24:39 -0400 Date: Wed, 5 Jul 2006 06:07:45 +0300 From: "Michael S. Tsirkin" To: Roland Dreier Cc: Zach Brown , Andrew Morton , Ingo Molnar , linux-kernel@vger.kernel.org, openib-general@openib.org, Arjan van de Ven Subject: Re: [openib-general] [PATCH] mthca: initialize send and receive queue locks separately Message-ID: <20060705030745.GA20709@mellanox.co.il> Reply-To: "Michael S. Tsirkin" References: Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.4.2.1i X-OriginalArrivalTime: 05 Jul 2006 03:12:27.0453 (UTC) FILETIME=[D8267ED0:01C69FE0] X-Spam: [F=0.0100000000; S=0.010(2006062901)] X-MAIL-FROM: X-SOURCE-IP: [63.251.237.3] Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Quoting r. Roland Dreier : > I think that is actually a very minor bug you've found. If someone > were posting a work request at the same time as they transitioned a QP > to reset (which is a legitimate if not sensible thing to do), then the > spinlock could get reinitialized while it was held. Which would be > bad. I think you can't post work requests to a QP in reset state, since IB spec forbids this. If you do, it seems bad things will happen anyway, for example head/tail pointers may get out of sync as mthca_wq_init clears these. -- MST