From mboxrd@z Thu Jan 1 00:00:00 1970 From: Matthew Wilcox Date: Fri, 15 Oct 2004 00:36:33 +0000 Subject: Re: [PATCH] Introduce PCI <-> CPU address conversion [1/2] Message-Id: <20041015003633.GX16153@parcelfarce.linux.theplanet.co.uk> List-Id: References: <20041014124737.GM16153@parcelfarce.linux.theplanet.co.uk> <20041014125348.GA9633@infradead.org> <20041014135323.GO16153@parcelfarce.linux.theplanet.co.uk> <20041014180005.GA11954@infradead.org> <20041014180748.GS16153@parcelfarce.linux.theplanet.co.uk> <416EFFBE.7B8F702@sgi.com> In-Reply-To: <416EFFBE.7B8F702@sgi.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: Colin Ngam Cc: Matthew Wilcox , Christoph Hellwig , Greg KH , linux-pci@atrey.karlin.mff.cuni.cz, linux-kernel@vger.kernel.org, linux-ia64@vger.kernel.org On Thu, Oct 14, 2004 at 05:37:51PM -0500, Colin Ngam wrote: > On SGI's Altix system, the sysdata for the device is very much different than > the sysdata for the bus. That's fascinating, because ia64 is one of the architectures that relies on sysdata being the same in both the bus and the device: #define PCI_CONTROLLER(busdev) ((struct pci_controller *) busdev->sysdata) In various places, we have struct pci_controller *controller = PCI_CONTROLLER(dev); and if (PCI_CONTROLLER(bus)->iommu) So what the hell does Altix do? Which sysdata can be used to get to the pci_controller? This seems like a horrible mistake to me. -- "Next the statesmen will invent cheap lies, putting the blame upon the nation that is attacked, and every man will be glad of those conscience-soothing falsities, and will diligently study them, and refuse to examine any refutations of them; and thus he will by and by convince himself that the war is just, and will thank God for the better sleep he enjoys after this process of grotesque self-deception." -- Mark Twain