From mboxrd@z Thu Jan 1 00:00:00 1970 From: Fabrice Gasnier Subject: Re: [RFC PATCH v1 1/3] PCI: designware: Add ARM64 support Date: Wed, 27 May 2015 15:52:51 +0200 Message-ID: <5565CC33.3070301@st.com> References: <000001d096a9$27bf43f0$773dcbd0$@com> <5562F0C9.4090007@hisilicon.com> <000001d096f1$93dca4e0$bb95eea0$@com> <5563451B.4050609@st.com> <5563DF4F.4080601@hisilicon.com> <556428C9.9080202@st.com> <5565C598.7020809@hisilicon.com> Mime-Version: 1.0 Content-Type: text/plain; charset="windows-1252"; format=flowed Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <5565C598.7020809@hisilicon.com> Sender: linux-pci-owner@vger.kernel.org To: Zhou Wang Cc: Jingoo Han , zhudacai@hisilicon.com, devicetree@vger.kernel.org, 'Gabriele Paoloni' , 'Arnd Bergmann' , 'Richard Zhu' , linux-pci@vger.kernel.org, 'Pratyush Anand' , 'Liviu Dudau' , 'Zhichang Yuan' , 'Kishon Vijay Abraham I' , qiuzhenfa@hisilicon.com, 'Bjorn Helgaas' , 'Zhang Jukuo' , 'Liguozhu' , linux-arm-kernel@lists.infradead.org, 'Lucas Stach' List-Id: devicetree@vger.kernel.org On 05/27/2015 03:24 PM, Zhou Wang wrote: >> there is still code in bios32 that requires pci_sys_data. > Right. > >> >No matters pci_common_init_dev() is being called or not : >> > >> >You can see pcibios_msi_controller() has a weak definition (in drivers/pci/msi.c). >> >But bios32 defines it for arm. Then it's being used. > But here in dw_pcie_host_init, I directly set bus->msi = &dw_pcie_msi_chip. So in > drivers/pci/msi.c, we will get the msi controller by dev->bus->msi in pci_msi_controller. > It will not run into pcibios_msi_controller(). > > Maybe there is still something I miss understood, if so, please let me know. Hi Zhou, Sorry, my mistake, you're right. I double checked, issue I'm seeing is in pcibios_align_resource... Please refer to Arnd's answer. Best Regards, Fabrice > > Best Regards, > Zhou >