From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S262399AbVFIXTZ (ORCPT ); Thu, 9 Jun 2005 19:19:25 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S262417AbVFIXTZ (ORCPT ); Thu, 9 Jun 2005 19:19:25 -0400 Received: from smtp.osdl.org ([65.172.181.4]:49811 "EHLO smtp.osdl.org") by vger.kernel.org with ESMTP id S262399AbVFIXTN (ORCPT ); Thu, 9 Jun 2005 19:19:13 -0400 Date: Thu, 9 Jun 2005 16:20:59 -0700 (PDT) From: Linus Torvalds To: Andreas Koch cc: Ivan Kokshaysky , Benjamin Herrenschmidt , linux-pci@atrey.karlin.mff.cuni.cz, linux-kernel@vger.kernel.org, gregkh@suse.de Subject: Re: PROBLEM: Devices behind PCI Express-to-PCI bridge not mapped In-Reply-To: <20050609223835.GB26023@erebor.esa.informatik.tu-darmstadt.de> Message-ID: References: <20050604155742.GA14384@erebor.esa.informatik.tu-darmstadt.de> <20050605204645.A28422@jurassic.park.msu.ru> <20050606002739.GA943@erebor.esa.informatik.tu-darmstadt.de> <20050606184335.A30338@jurassic.park.msu.ru> <20050608173409.GA32004@erebor.esa.informatik.tu-darmstadt.de> <20050609023639.A7067@jurassic.park.msu.ru> <1118289850.6850.164.camel@gaston> <20050609175441.C9187@jurassic.park.msu.ru> <20050609175429.GA26023@erebor.esa.informatik.tu-darmstadt.de> <20050609223835.GB26023@erebor.esa.informatik.tu-darmstadt.de> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Fri, 10 Jun 2005, Andreas Koch wrote: > > I did some more experimentation, and to my great the surprise, the > serial port on the dock _is_ functioning, even when the rest of the > dock is dead. I wonder whether the bridge is effectively a negative decode thing, and the only "real" problem is that because the kernel doesn't know that, it doesn't know that it can allocate just about any resource at all on the other end.. It would be pretty strange, since the PCI spec (afaik, and for pretty obvious reasons) disallows two negative bridges on the same bus (and you already have the other mobile bridge there), but it's technically possible if they just have different priorities for how fast they react to a PCI address cycle and claim it. Ivan, can you cook up some silly patch that just marks that one device transparent, and see if that "just fixes it". Linus