From mboxrd@z Thu Jan 1 00:00:00 1970 From: Sergei Shtylyov Date: Thu, 16 May 2013 20:58:23 +0000 Subject: Re: [PATCH] : fix compilation warnings with DT disabled Message-Id: <5195486F.1040603@cogentembedded.com> List-Id: References: <201302190258.25726.sergei.shtylyov@cogentembedded.com> <20130301070925.GA24369@verge.net.au> <5130B12F.20409@gmail.com> <5130B1A7.9060001@gmail.com> <5130B34F.2090700@cogentembedded.com> <5130B506.5020908@gmail.com> In-Reply-To: <5130B506.5020908@gmail.com> MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 8bit To: Rob Herring Cc: Simon Horman , linux-sh@vger.kernel.org, devicetree-discuss@lists.ozlabs.org, linux-kernel@vger.kernel.org, rob.herring@calxeda.com, vladimir.barinov@cogentembedded.com Hello. On 03/01/2013 05:02 PM, Rob Herring wrote: >>>>>> Fix the following compilation warnings (in Simon Horman's >>>>>> renesas.git repo): >>>>>> In file included from arch/arm/mach-shmobile/setup-r8a7779.c:24:0: >>>>>> include/linux/of_platform.h:107:13: warning: ‘struct of_device_id’ >>>>>> declared >>>>>> inside parameter list [enabled by default] >>>>>> include/linux/of_platform.h:107:13: warning: its scope is only this >>>>>> definition >>>>>> or declaration, which is probably not what you want [enabled by >>>>>> default] >>>>>> include/linux/of_platform.h:107:13: warning: ‘struct device_node’ >>>>>> declared >>>>>> inside parameter list [enabled by default] >>>>>> only #include's headers with definitions of >>>>>> the above >>>>>> mentioned structures if CONFIG_OF_DEVICE=y but uses them even if >>>>>> not. One >>>>>> solution is to move some #include's out of #ifdef CONFIG_OF_DEVICE >>>>>> and use >>>>>> incomplete declarations for the rest of the structures where the >>>>>> #ifdef move >>>>>> doesn't help... >>>>>> Reported-by: Vladimir Barinov >>>>>> Signed-off-by: Sergei Shtylyov >>>>> Reviewed-by: Simon Horman >>>>> Grant, could you consider taking this patch? >>>> Yes, I can, but I don't seem to have the original patch. Can you send it >>>> again. >>> Nevermind. Found it. I'll apply it. >> Will you drop 'struct device_node' declaration then or should I >> resend? In fact, I think I should better resend it with the changelog >> somewhat edited. > No, I plan to leave it as is and not rely on device.h by chance > declaring device_node. I hoped to see this fix in 3.10-rc1. Is there any hope to see it in 3.10-rc's? The code it fixes the warnings in is already in Linus tree. > >>> Rob WBR, Sergei