From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932117AbWGKUCt (ORCPT ); Tue, 11 Jul 2006 16:02:49 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S932116AbWGKUCt (ORCPT ); Tue, 11 Jul 2006 16:02:49 -0400 Received: from mxsf18.cluster1.charter.net ([209.225.28.218]:55193 "EHLO mxsf18.cluster1.charter.net") by vger.kernel.org with ESMTP id S1751213AbWGKUCs (ORCPT ); Tue, 11 Jul 2006 16:02:48 -0400 X-IronPort-AV: i="4.06,230,1149480000"; d="scan'208"; a="387608855:sNHT20976968" MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <17588.997.688135.786150@stoffel.org> Date: Tue, 11 Jul 2006 16:02:45 -0400 From: "John Stoffel" To: Russell King Cc: John Stoffel , linux-kernel@vger.kernel.org, akpm@osdl.org Subject: Re: 2.6.18-rc1-mm1 - bad serial port count messages In-Reply-To: <20060711185630.GA1240@flint.arm.linux.org.uk> References: <17587.42397.168635.821696@stoffel.org> <20060711185630.GA1240@flint.arm.linux.org.uk> X-Mailer: VM 7.19 under Emacs 21.4.1 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Russell> On Tue, Jul 11, 2006 at 09:20:29AM -0400, John Stoffel wrote: >> I'm getting the following messages in dmesg: >> >> uart_close: bad serial port count; tty->count is 1, state->count is 0 >> uart_close: bad serial port count for ttyS0: -1 >> uart_close: bad serial port count for ttyS0: -1 Russell> I assume that it's 100% reproducable, and doesn't happen with Russell> mainline? Not sure, haven't rebooted yet to make sure it happens again. The hardware is a Dell Dimension 630, Dual Xeon PIII 550mhz, 768mb of RAM. It's been my main machine for a number of years now. Russell> I'm not aware of any serial core patches in -mm which would Russell> produce this type of breakage - maybe there's something funny Russell> with the tty layer in that it's trying to close the port more Russell> times than it's been opened... Hmm. I thought I saw something float by where someone had raised some new locking or count variables? I dunno... I'll work on it tonight and see what happens after another reboot. John