From mboxrd@z Thu Jan 1 00:00:00 1970 From: Arnd Bergmann Subject: Re: [PATCH v3 1/6] PCI: designware: remove wrong io_base assignment Date: Mon, 23 Nov 2015 17:40:56 +0100 Message-ID: <6899049.45pKpnsxmL@wuerfel> References: <56533D93.5000604@linaro.org> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7Bit Return-path: In-Reply-To: <56533D93.5000604@linaro.org> Sender: linux-arm-msm-owner@vger.kernel.org To: Stanimir Varbanov Cc: Gabriele Paoloni , "linux-arm-msm@vger.kernel.org" , "linux-kernel@vger.kernel.org" , "linux-arm-kernel@lists.infradead.org" , "devicetree@vger.kernel.org" , "linux-pci@vger.kernel.org" , Bjorn Helgaas , Srinivas Kandagatla , Rob Herring , Rob Herring , Mark Rutland , Pawel Moll , Ian Campbell , Jingoo Han , Pratyush Anand , Bjorn Andersson List-Id: devicetree@vger.kernel.org On Monday 23 November 2015 18:23:47 Stanimir Varbanov wrote: > >> > >> Fixes: 0021d22b73d6 ("PCI: designware: Use of_pci_get_host_bridge_resources() > >> to parse DT") > >> Reviewed-by: Arnd Bergmann > > I think the bug is introduced in: > > cbce7900598c ("PCI: designware: Make driver arch-agnostic") > > cause the io_base is correctly calculated in 0021d22b73d6, do you agree? I think cbce7900598c just slightly changes the io_base value, but it's still referring to a bus address, not a cpu address, both before and after the patch. 0021d22b73d6 hower seems to remove the correct 'pp->io_base = range.cpu_addr;' and replaces it with 'pp->io_base = pp->io->start;', so it's now in the wrong address space. Arnd