From mboxrd@z Thu Jan 1 00:00:00 1970 From: Stephen Boyd Subject: Re: [PATCH v6] soc: qcom: Add SoC info driver Date: Wed, 21 Dec 2016 16:31:22 -0800 Message-ID: <20161222003122.GY8288@codeaurora.org> References: <1481555889-29380-1-git-send-email-kimran@codeaurora.org> <20161214002617.GS5423@codeaurora.org> <20161217012615.GV5423@codeaurora.org> <05cdb699-6406-cff8-cce6-fcedf6ac6c4e@codeaurora.org> <20161220225007.GD5423@codeaurora.org> <1305f89c-2f9b-aa0b-4e22-951bf8af9344@codeaurora.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from smtp.codeaurora.org ([198.145.29.96]:46032 "EHLO smtp.codeaurora.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1760107AbcLVAeG (ORCPT ); Wed, 21 Dec 2016 19:34:06 -0500 Content-Disposition: inline In-Reply-To: <1305f89c-2f9b-aa0b-4e22-951bf8af9344@codeaurora.org> Sender: linux-arm-msm-owner@vger.kernel.org List-Id: linux-arm-msm@vger.kernel.org To: Imran Khan Cc: andy.gross@linaro.org, lee.jones@linaro.org, David Brown , linux-kernel@vger.kernel.org, linux-arm-msm@vger.kernel.org, linux-soc@vger.kernel.org On 12/21, Imran Khan wrote: > On 12/21/2016 4:20 AM, Stephen Boyd wrote: > > > > I'll wait to see what the next patch version has. We will > > probably need to have some way to know which ODM the kernel is > > running on, so we can interpret the platform type/subtype fields > > properly. That part seems to be lacking from this patch right > > now. We assume it's always qcom as the ODM, which isn't true. > > > Now I get this point. So far we don't have any mechanism in the driver that > gives ODM information. As far as generic soc_device_attribute's vendor field > is concerned we use Qualcomm since this will be true for SoC. > For hardware type and sub-types the various relevant values in SMEM are numeric > values and indeed it would be very difficult to estimate how some other ODM > will use the same number. > So for the h/w types and sub-types can we keep the numeric values rather than > showing strings as attribute values. We can leave the interpretation of these > values to ODM specific code. Raw numbers sounds fine, but how do we know what ODM it is to understand how to parse the numbers appropriately? Perhaps the smem DT entry needs to have a property indicating the ODM that has configured these numbers, and then we can have an ODM sysfs node that we use to expose that string property to userspace? -- Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, a Linux Foundation Collaborative Project