From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S265696AbTFNRsM (ORCPT ); Sat, 14 Jun 2003 13:48:12 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S265697AbTFNRsM (ORCPT ); Sat, 14 Jun 2003 13:48:12 -0400 Received: from pc2-cwma1-4-cust86.swan.cable.ntl.com ([213.105.254.86]:51645 "EHLO lxorguk.ukuu.org.uk") by vger.kernel.org with ESMTP id S265696AbTFNRsL (ORCPT ); Sat, 14 Jun 2003 13:48:11 -0400 Subject: Re: Linux 2.4.21-ac1 From: Alan Cox To: walt Cc: Linux Kernel Mailing List In-Reply-To: <3EEAF192.4050102@myrealbox.com> References: <200306141430.h5EEUuV31162@devserv.devel.redhat.com> <3EEAF192.4050102@myrealbox.com> Content-Type: text/plain Content-Transfer-Encoding: 7bit Organization: Message-Id: <1055613583.8281.23.camel@dhcp22.swansea.linux.org.uk> Mime-Version: 1.0 X-Mailer: Ximian Evolution 1.2.2 (1.2.2-5) Date: 14 Jun 2003 18:59:44 +0100 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Sad, 2003-06-14 at 10:57, walt wrote: > This is the offending line in serialmidi.c: > > if (tty->count > 1) { > snd_printk(KERN_ERR "tty %s is already used", serial->sdev); > > The compilation will finish if I make this change: > > if (atomic_read(&(tty->count)) > 1) { > > Since I don't have a clue what I'm doing this is probably not the right fix :0) > I suspect that changing the alsa code is starting at the wrong end of the > problem anyway. The tty code changed to atomic counters to fix a race. Your fix is right.