From mboxrd@z Thu Jan 1 00:00:00 1970 From: Simon Horman Date: Tue, 05 Mar 2013 05:57:00 +0000 Subject: Re: [PATCH 1/3] ARM: shmobile: armadillo800eva: enable all errata for cache on defconfig Message-Id: <20130305055659.GC9711@verge.net.au> List-Id: References: <1346808893-14124-2-git-send-email-horms@verge.net.au> In-Reply-To: <1346808893-14124-2-git-send-email-horms@verge.net.au> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: linux-sh@vger.kernel.org On Sun, Mar 03, 2013 at 04:42:08PM -0800, Kuninori Morimoto wrote: > > Hi Simon > > > > > > > Some errata for cache had not enabled on current > > > > > > Armadillo800eva defconfig. > > > > > > This patch enables these. > > > > > > Special thanks to Ishiyama-san > > > > > > > > > > It seems to me that it would be better to handle these in Kconfig > > > > > to enable them on an SoC or board basis. > > > > > > > > > > The reason is that I would like to move away from per-board defconfigs > > > > > and towards a more consolidated defconfig. And in that case it will > > > > > be more difficult to know which ERRATA are needed for each board and SoC. > > > > > > > > > > Obviously this thinking is open to review and it implies that > > > > > several board defconfigs should be updated. > > > > > > We used defconfig selection for CPU/board settings before. > > > But Kconfig selection for ERRATA make sense. > > > > > > Do you want to me to send v2 patch which use Kconfig ? > > > Or do you do that ? > > > > I'd like you to send v2 if possible. > > OK, but I noticed that it will be complex if we select these ERRATA on Kconfig, > Because, these ERRATAs dependency. Thanks, I see this is indeed getting messy. The reason that I asked is that I would like to consolidate the defconfigs. Ideally one for shmobile. But more realistically one per SoC. And I thought that avoiding adding ERRATAs to defconfigs would make that slightly easier in the long run. But I don't think it justifies dealing with the mess you describe below. So I will apply the patch as is. > Ex) > config PL310_ERRATA_588369 > bool "PL310 errata: Clean & Invalidate maintenance operations do not invalidate clean lines" > depends on CACHE_L2X0 > > It depends on CACHE_L2X0, but this CACHE_L2X0 selection is depend on user, > this means that there is no guarantee that CACHE_L2X0 is selected. > Then, the Kconfig will be... > > select PL310_ERRATA_588369 if CACHE_L2X0 > > And more complex ERRARA dependency is... > > config ARM_ERRATA_742230 > bool "ARM errata: DMB operation may be faulty" > depends on CPU_V7 && SMP > depends on !ARCH_MULTIPLATFORM > > Of course I can use above style on Kconfig, > but my worry is we will have unknown trouble > if someone updated these ERRATA implementation and dependency > in the future... > > This is example, but I tested below. > This means we can select unnecessary settings on Kconfig ... > > ---------------- > config ARM_ERRATA_326103 > bool "ARM errata: FSR write bit incorrect on a SWP to read-only memory" > => depends on CPU_V6 > ... > > config ARCH_R8A7740 > bool "R-Mobile A1 (R8A77400)" > select ARCH_WANT_OPTIONAL_GPIOLIB > => select CPU_V7 > => select ARM_ERRATA_326103 > ... > > > grep ERRATA .config > CONFIG_ARM_ERRATA_326103=y > ---------------------- > > Best regards > --- > Kuninori Morimoto >