From mboxrd@z Thu Jan 1 00:00:00 1970 From: Phil Sutter Subject: Re: [PATCH] irq handler must disable the handled irq Date: Mon, 27 Oct 2008 16:41:06 +0100 Message-ID: <20081027154106.GA27227@nuty> References: <1225116107-9753-1-git-send-email-n0-1@freewrt.org> <20081027141509.2d0be2c7@lxorguk.ukuu.org.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from orbit.nwl.cc ([81.169.176.177]:56838 "EHLO mail.ifyouseekate.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752134AbYJ0Prn (ORCPT ); Mon, 27 Oct 2008 11:47:43 -0400 Received: from nuty (localhost [127.0.0.1]) by mail.ifyouseekate.net (Postfix) with ESMTP id D239B38C2BF5 for ; Mon, 27 Oct 2008 16:40:11 +0100 (CET) Content-Disposition: inline In-Reply-To: <20081027141509.2d0be2c7@lxorguk.ukuu.org.uk> Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: linux-ide@vger.kernel.org On Mon, Oct 27, 2008 at 02:15:09PM +0000, Alan Cox wrote: > You can't call disable_irq in an IRQ handler. What are you actually > trying to do ? Well, what I'm trying to do is getting the driver to a working state. But it seems like I don't know what I'm actually doing. :/ The symptom I discovered was rb532_pata_irq_handler() being called over and over again, endlessly. Calling disable_irq() from within stopped that, so I assumed it to be right (obviously the wrong direction for a proof). As the working driver for 2.6.19 I have here is not based on libata, I'm not quite sure how much code is still required and where to hook it into the libata subsystem. Greetings, Phil