From mboxrd@z Thu Jan 1 00:00:00 1970 From: James Morse Subject: Re: [PATCH 2/2] edac: add support for Amazon's Annapurna Labs EDAC Date: Wed, 5 Jun 2019 16:13:03 +0100 Message-ID: References: <1559211329-13098-1-git-send-email-hhhawa@amazon.com> <1559211329-13098-3-git-send-email-hhhawa@amazon.com> <20190531051400.GA2275@cz.tnic> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20190531051400.GA2275@cz.tnic> Content-Language: en-GB Sender: linux-kernel-owner@vger.kernel.org To: Borislav Petkov , "Herrenschmidt, Benjamin" Cc: "robh+dt@kernel.org" , "Woodhouse, David" , "paulmck@linux.ibm.com" , "mchehab@kernel.org" , "mark.rutland@arm.com" , "Hawa, Hanna" , "gregkh@linuxfoundation.org" , "davem@davemloft.net" , "nicolas.ferre@microchip.com" , "devicetree@vger.kernel.org" , "Shenhar, Talel" , "linux-kernel@vger.kernel.org" , "Chocron, Jonathan" , "Krupnik, Ronen" , "linux-edac@vger.kernel.org" , "Hanoch, Uri" List-Id: devicetree@vger.kernel.org Hi guys, On 31/05/2019 06:14, Borislav Petkov wrote: > On Fri, May 31, 2019 at 01:15:33AM +0000, Herrenschmidt, Benjamin wrote: >> This isn't terribly helpful, there's nothing telling anybody which of >> those files corresponds to an ARM SoC :-) > > drivers/edac/altera_edac.c is one example. > > Also, James and I have a small writeup on how an arm driver should look > like, we just need to polish it up and post it. > > James? Yes I should get on with that. Its mostly for platforms which end up with multiple piecemeal drivers and some co-ordination is needed. It doesn't look like that will be a problem here. >> That said ... >> >> You really want a single EDAC driver that contains all the stuff for >> the caches, the memory controller, etc... ? This has to be platform specific as it has integration-time dependencies and firmware dependencies. Doing it as a platform driver matched from the machine-compatible may be more straightforward today. The DT will already say "compatible = arm,cortex-a57" for the Alpine-v2, what that 'edac_l1_l2' node is telling us is the integration/firmware stuff has been done, and the imp-def instructions can be used. Thanks, James