From mboxrd@z Thu Jan 1 00:00:00 1970 From: Bjorn Helgaas Date: Wed, 08 Mar 2006 21:56:52 +0000 Subject: Re: git pull on ia64 linux tree Message-Id: <200603081456.53031.bjorn.helgaas@hp.com> List-Id: References: <200504222203.j3MM3fV17003@unix-os.sc.intel.com> In-Reply-To: <200504222203.j3MM3fV17003@unix-os.sc.intel.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: linux-ia64@vger.kernel.org On Wednesday 08 March 2006 13:47, Luck, Tony wrote: > >On Wed, Mar 08, 2006 at 11:54:26AM -0800, Luck, Tony wrote: > >> Bjorn Helgaas: > >> [IA64] don't report !sn2 or !summit hardware as an error > >> [IA64] SGI SN drivers: don't report !sn2 hardware as an error > > > >I thought hch solved this by fixing the initcall code to not > >treat -ENODEV as an error worthy of reporting? > > That's what happens when I get far, far behind on LKML. I've found > traces of the threads ... though none by hch, and no clear agreement > about whether 0 or -ENODEV is right. Some of this code is linked > into the ia64 generic kernel by default, so I'm leaning a little > towards Bjorn's position that we shouldn't be returning an error. There are a ton of drivers that just return the result of pci_register_driver(). This is typically zero, even when no devices are present. There are also some drivers that return -ENODEV when no devices are present. Personally, I think the first style is best, and we should deal with the problem of loading the right drivers a different way, e.g., with udev. But others disagree: http://lkml.org/lkml/2006/3/7/159