From mboxrd@z Thu Jan 1 00:00:00 1970 From: arnd@arndb.de (Arnd Bergmann) Date: Wed, 14 Nov 2012 23:26:22 +0000 Subject: arch-vt8500 and wm8850 support In-Reply-To: <1352914966.3238.3.camel@gitbox> References: <1352879641.2728.8.camel@gitbox> <201211141117.33485.arnd@arndb.de> <1352914966.3238.3.camel@gitbox> Message-ID: <201211142326.23023.arnd@arndb.de> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org On Wednesday 14 November 2012, Tony Prisk wrote: > Makes sense - That's basically what I had in mind. My concern is with > the bit that made no sense to me - the non-multiplatform version of the > same. At the moment, we have the default xx_SINGLE for ARCH_VT8500. > Would I need to add an additional xx_SINGLE variant for each different > version? I think either one option or two (v5 and v6/v7), but we don't need three separate ones, as it should be possible to build v6 and v7 together. > We would need ARCH_WM8505 (To cover VT8500/WM8505/WM8650 ARM926), > ARCH_WM8750 (for ARMv6) and ARCH_WM8850 (for ARMv7). Ah, right, I was mixing up the numbers. > Do I need a ARCH_WM8750_SINGLE, ARCH_WM8850_SINGLE etc for each one to > allow use to continue using earlyprintk? You just need to ensure that we always set at least one of them, and that ARCH_WM8505 isn't set at the same time as one of the v6/v7 parts. Maybe something like config ARCH_VT8500 bool select FOO select BAR config ARCH_WM8505 bool "WonderMedia WM8505/WM8650 or VIA VT8500" if !ARCH_VT8500_SINGLE depends on ARCH_MULTI_V5 default ARCH_VT8500_SINGLE select ARCH_VT8500 select CPU_ARM926 config ARCH_WM8750 bool "WonderMedia WM 8750" depends on ARCH_MULTI_V6 || ARCH_WM8750_SINGLE select ARCH_VT8500 select CPU_V6 config ARCH_WM8850 bool "WonderMedia WM 8850/8950" depends on ARCH_MULTI_V7 || ARCH_WM8750_SINGLE select ARCH_VT8500 select CPU_V7 config ARCH_WM8850_AUTO def_bool y depends on ARCH_WM8750_SINGLE && !ARCH_WM8750 select ARCH_WM8850 # ensure that ARCH_WM8850 is on if ARCH_WM8750 is off Arnd