From mboxrd@z Thu Jan 1 00:00:00 1970 From: eric@eukrea.com (Eric =?ISO-8859-1?B?QuluYXJk?=) Date: Thu, 24 Oct 2013 10:36:16 +0200 Subject: [PATCHv6][ 3/7] ARM: imx{25,27} DT: Permit the selection of the imxfb framebuffer driver. In-Reply-To: <20131024083143.GB13614@S2101-09.ap.freescale.net> References: <1382601665-6830-1-git-send-email-denis@eukrea.com> <1382601665-6830-3-git-send-email-denis@eukrea.com> <20131024081531.GA13614@S2101-09.ap.freescale.net> <20131024102011.610e5d8b@e6520eb> <20131024083143.GB13614@S2101-09.ap.freescale.net> Message-ID: <20131024103616.0b28b922@e6520eb> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org Hi Shawn, Le Thu, 24 Oct 2013 16:31:45 +0800, Shawn Guo a ?crit : > On Thu, Oct 24, 2013 at 10:20:11AM +0200, Eric B?nard wrote: > > Hi Shawn, > > > > Le Thu, 24 Oct 2013 16:15:34 +0800, > > Shawn Guo a ?crit : > > > > config FB_IMX > > > > tristate "Freescale i.MX1/21/25/27 LCD support" > > > > - depends on FB && IMX_HAVE_PLATFORM_IMX_FB > > > > + depends on FB && HAVE_IMX_FB > > > > > > Or simply have it depend on ARCH_MXC? > > > > > > depends on FB && ARCH_MXC > > > > > but not all MXC have IMX_FB : isn't that a problem ? > > Is it really a problem that you see this option when you configure an > imx51 kernel? The prompt clearly tells you do not need this option for > an imx51 kernel. Or will it cause any real problem even if you select > it for an imx51 kernel, except you build in a driver which will never be > instantiated? I've seen drivers even start drop ARCH_* dependency to > get a better randconfig build test coverage. > not really a problem for me. If the policy is not to narrow the choices to only the SOC which have the peripheral then we follow it and take your ARCH_MXC approach which is far simpler. Thanks, Eric