From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S261666AbVFUJh7 (ORCPT ); Tue, 21 Jun 2005 05:37:59 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S262096AbVFUJhV (ORCPT ); Tue, 21 Jun 2005 05:37:21 -0400 Received: from clock-tower.bc.nu ([81.2.110.250]:63650 "EHLO lxorguk.ukuu.org.uk") by vger.kernel.org with ESMTP id S262090AbVFUJg3 (ORCPT ); Tue, 21 Jun 2005 05:36:29 -0400 Subject: Re: PATCH: Fix crashes with hotplug serverworks From: Alan Cox To: Christoph Hellwig Cc: Linux Kernel Mailing List , akpm@osdl.org In-Reply-To: <20050621065221.GA31420@infradead.org> References: <1119298859.3325.43.camel@localhost.localdomain> <20050621065221.GA31420@infradead.org> Content-Type: text/plain Content-Transfer-Encoding: 7bit Message-Id: <1119346417.3325.82.camel@localhost.localdomain> Mime-Version: 1.0 X-Mailer: Ximian Evolution 1.4.6 (1.4.6-2) Date: Tue, 21 Jun 2005 10:33:40 +0100 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Maw, 2005-06-21 at 07:52, Christoph Hellwig wrote: > Well, because of fake-hotplug we really need to mark every ->probe routine > and what's called from it __devinit. Debian has patch to do that forever > but Bart refused to take it. I'm not surprised from our experience either. Actually marking all the devices __devinit may be overkill. One approach that does also seem to work is passing "hotplug yes/no" information when registering the driver. This is then used to run the ide scan at boot and avoid registering that driver with the PCI core for hotplug. A move to __devinit is certainly simpler.