From mboxrd@z Thu Jan 1 00:00:00 1970 From: gregkh@suse.de (Greg KH) Date: Fri, 4 Mar 2011 08:52:04 -0800 Subject: [PATCH v10 1/1] PRUSS UIO driver support In-Reply-To: <201103041658.05655.arnd@arndb.de> References: <1299213772-21931-1-git-send-email-pratheesh@ti.com> <20110304142028.GD28175@local> <201103041658.05655.arnd@arndb.de> Message-ID: <20110304165204.GA21244@suse.de> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org On Fri, Mar 04, 2011 at 04:58:05PM +0100, Arnd Bergmann wrote: > On Friday 04 March 2011, Nori, Sekhar wrote: > > On Fri, Mar 04, 2011 at 19:50:28, Hans J. Koch wrote: > > > On Fri, Mar 04, 2011 at 10:12:52AM +0530, Pratheesh Gangadhar wrote: > > > > diff --git a/include/linux/uio_pruss.h b/include/linux/uio_pruss.h > > > > new file mode 100644 > > > > index 0000000..8c9b2c9 > > > > --- /dev/null > > > > +++ b/include/linux/uio_pruss.h > > > > > > That should go to arch/arm/mach-davinci/include/mach/ > > > and not to the top level include/linux/ directory. > > > > It was put in here because this driver > > will be used across two ARM architectures > > (DaVinci and OMAP). Although it won't be used > > outside of ARM at least in near future. > > I think include/linux/ is fine for this No, please use include/linux/platform_data/ instead, that is where platform data should be going into, not into the "global" space if include/linux/ > , and arch/arm/include/asm would > be ok as well. You might want to have only a single pruss header > file that defines all the platform data for the child drivers, though. > > Ideally, you would of course not need the header at all. This will > be possible with the move to a flat device tree, where you can store the > offset as a property of the device > > Another alternative would be to move it to a resource of the platform > device. That might be a nice idea for the future. thanks, greg k-h