From: David Daney Subject: Re: [PATCH] net: mdio-octeon: Add PCI driver binding. Date: Thu, 24 Sep 2015 15:04:23 -0700 Message-ID: <56047367.3060101@caviumnetworks.com> References: <1442968896-30776-1-git-send-email-ddaney.cavm@gmail.com> <20150924.145206.897667718297700953.davem@davemloft.net> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-7; format=flowed Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: In-Reply-To: <20150924.145206.897667718297700953.davem@davemloft.net> Sender: linux-kernel-owner@vger.kernel.org To: David Miller Cc: ddaney.cavm@gmail.com, linux-kernel@vger.kernel.org, robh+dt@kernel.org, pawel.moll@arm.com, mark.rutland@arm.com, ijc+devicetree@hellion.org.uk, galak@codeaurora.org, devicetree@vger.kernel.org, f.fainelli@gmail.com, netdev@vger.kernel.org, david.daney@cavium.com List-Id: devicetree@vger.kernel.org On 09/24/2015 02:52 PM, David Miller wrote: > From: David Daney > Date: Tue, 22 Sep 2015 17:41:36 -0700 > >> From: David Daney >> >> When the Cavium mdio-octeon devices appear in the Thunder family of >> arm64 based SoCs, they show up as PCI devices. Add PCI driver >> wrapping so the driver is bound in the standard PCI device scan. >> >> When in this form, a single PCI device may have more than a single >> bus, we call this a "nexus" of buses. The standard firmware >> device_for_each_child_node() iterator is used to find the individual >> buses underneath the "nexus". >> >> Update the device tree binding documentation for the new PCI driver >> binding. >> >> Signed-off-by: David Daney > > This patch breaks the build: =46or which architecture? I tested it on mips and arm64. I will try x86, as I guess that is wher= e=20 you tried your test build. > > CC [M] drivers/net/phy/mdio-octeon.o > In file included from drivers/net/phy/mdio-octeon.c:13:0: > include/linux/module.h:128:27: error: redefinition of =A1__inittest=A2 > static inline initcall_t __inittest(void) \ > ^ [...] > > And frankly I'm not sure I like this change anyways. If the OF nodes > are there, simply add code to match on those OF nodes. > > This is better than assuming what sits underneath a PCI node, without > any checks for the 'name' or 'compatible' properties at all. That's > what they are there for afterall. There is, somewhat of, a method behind the madness here. In order to use MSI-X interrupts, we need a corresponding PCI device.=20 Now, this driver doesn't currently use interrupts, but other devices in= =20 the SoC do, so they must be PCI devices. The idea is to have the type of all drivers uniformly be PCI, rather=20 than a random mix of PCI and platform, and then switch them back and=20 forth as PCI features are/are-not used in the driver. Also we need to consider ACPI firmware in addition to OF device tree. David Daney