From mboxrd@z Thu Jan 1 00:00:00 1970 From: Benjamin Herrenschmidt Subject: Re: [RFC] Architecture independent pcibios? Date: Wed, 09 Oct 2013 07:31:18 +1100 Message-ID: <1381264278.645.232.camel@pasglop> References: <20131008144211.GA25231@e102652-lin.cambridge.arm.com> <20131008171343.GB25606@e102652-lin.cambridge.arm.com> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20131008171343.GB25606@e102652-lin.cambridge.arm.com> Sender: linux-pci-owner@vger.kernel.org To: Liviu Dudau Cc: Bjorn Helgaas , "linux-pci@vger.kernel.org" , Catalin Marinas , Olof Johansson , Arnd Bergmann , Michal Simek , "grant.likely@linaro.org" , "rob.herring@calxeda.com" , "devicetree@vger.kernel.org" List-Id: devicetree@vger.kernel.org On Tue, 2013-10-08 at 18:13 +0100, Liviu Dudau wrote: > There are at least a handful of architectures that seem to share the same > implementation for most (all?) of the pcibios_*() functions. (PowerPC and > microblaze the most obvious offenders, x86 to a certain extent). That's because microblaze started by copying the powerpc code :-) The other part about x86 is mostly me over the year changing the powerpc code to make it closer to x86 especially in the area of resource management. This leads to less overall surprises and issues when the generic code changes and in a few cases has made it easier to move bits to the core. The main remaining difference is our use of IORESOURCE_UNSET and our flags to force reallocation. Ben.