From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Tue, 12 Jun 2001 08:32:26 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Tue, 12 Jun 2001 08:32:15 -0400 Received: from kosmo.edeal.de ([62.40.13.104]:45072 "EHLO kosmo.edeal.de") by vger.kernel.org with ESMTP id ; Tue, 12 Jun 2001 08:32:06 -0400 Date: Tue, 12 Jun 2001 14:31:42 +0200 From: Lukas Schroeder To: Ben Pfaff Cc: Alan Cox , Lukas Schroeder , zab@redhat.com, linux-kernel@vger.kernel.org Subject: Re: [patch] ess maestro, support for hardware volume control Message-ID: <20010612143142.A17450@kosmo.edeal.de> In-Reply-To: <200106091931.f59JVw731673@devserv.devel.redhat.com> <87elst2vr2.fsf@pfaffben.user.msu.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <87elst2vr2.fsf@pfaffben.user.msu.edu> User-Agent: Mutt/1.3.18i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org the little patch below (vs. 2.4.5-ac13) fixes a freeze which happens when the maestro module was unloaded and a HWV button gets pushed generating the interrupt. so this disables all irqs of that chip on remove... regards, lukas --- linux-2.4.5-ac13/drivers/sound/maestro.c Tue Jun 12 13:41:24 2001 +++ linux/drivers/sound/maestro.c Tue Jun 12 14:13:40 2001 @@ -3575,6 +3575,12 @@ struct ess_card *card = pci_get_drvdata(pcidev); int i; + /* turn off all irqs; _especially_ the one for hardware volume + control (bit 6), which locks the machine dead when occurring after + the maestro module was removed. + */ + outw(0x0, card->iobase+0x18); + /* XXX maybe should force stop bob, but should be all stopped by _release by now */ free_irq(card->irq, card);