From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id E1D90C433FE for ; Fri, 4 Nov 2022 21:47:46 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S229927AbiKDVrq (ORCPT ); Fri, 4 Nov 2022 17:47:46 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:48948 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S230048AbiKDVrp (ORCPT ); Fri, 4 Nov 2022 17:47:45 -0400 Received: from mail-oa1-f49.google.com (mail-oa1-f49.google.com [209.85.160.49]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 1A8A545EE7; Fri, 4 Nov 2022 14:47:44 -0700 (PDT) Received: by mail-oa1-f49.google.com with SMTP id 586e51a60fabf-13ba86b5ac0so6965694fac.1; Fri, 04 Nov 2022 14:47:44 -0700 (PDT) X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=in-reply-to:content-disposition:mime-version:references:message-id :subject:cc:to:from:date:x-gm-message-state:from:to:cc:subject:date :message-id:reply-to; bh=+4AWz3heTvW9/SjNGOnADQKijVdWZEZo5RmZcyg0wzU=; b=mJQGUpXWC7qfgGqVpQ/nleXiy45mPrMe1tV07AOnoKawzW3FvzOYiSRjXhIqvDSDcL XAV5du5dAEASbPk91y1VuO+5s2UNrVEDjrSlkCaGKPz9AvKjSPqCW0NJl6xIQlaQVQNt 31LMfmMJxCeBJbxRah76bhzvM5M8ExFP4pdN2AftyK9GMsfj21XPQT45qc2yjQVx/4Kx 8fMtJ5RGyyjjimplESaK/eGpNw1qv9x3Go1lHgK4EMFJg5G5R4HFGgHZgzVL/gdG7EJq IDMwzWSFiTgTEQMI6JvPRqntRwBYe2cZq5epbZJXhrKfw7LuGfWrh7JAJ2j375JV0ktL jzvA== X-Gm-Message-State: ACrzQf2LQMIGXNbqH5dSe+imVLYCks9w5pZl5P+Gz+t1TWWnjJYzD7kU c4xMC030491V1jUvnfNXaQ== X-Google-Smtp-Source: AMsMyM75XqDPp8F1ac7KnwNd5teq4r4YMJDMqo39kggUwLf7Jw9pTQVKqM7XDouGLMc4cawPhnXH7Q== X-Received: by 2002:a05:6870:8911:b0:13b:69b2:86b2 with SMTP id i17-20020a056870891100b0013b69b286b2mr23730705oao.253.1667598463241; Fri, 04 Nov 2022 14:47:43 -0700 (PDT) Received: from robh_at_kernel.org (66-90-144-107.dyn.grandenetworks.net. [66.90.144.107]) by smtp.gmail.com with ESMTPSA id ek17-20020a056870f61100b0013c50b812a2sm72744oab.36.2022.11.04.14.47.42 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Fri, 04 Nov 2022 14:47:42 -0700 (PDT) Received: (nullmailer pid 2885635 invoked by uid 1000); Fri, 04 Nov 2022 21:47:44 -0000 Date: Fri, 4 Nov 2022 16:47:44 -0500 From: Rob Herring To: Krzysztof Kozlowski Cc: Krzysztof Kozlowski , Matthias Brugger , devicetree@vger.kernel.org, linux-kernel@vger.kernel.org, linux-arm-kernel@lists.infradead.org, linux-mediatek@lists.infradead.org Subject: Re: [PATCH] dt-bindings: examples: add a list of templates and solutions Message-ID: <20221104214744.GA2878314-robh@kernel.org> References: <20221028233701.572280-1-krzysztof.kozlowski@linaro.org> <20221101130720.GA963805-robh@kernel.org> <34f40c1c-f933-cee6-f022-14c37650dc3a@linaro.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <34f40c1c-f933-cee6-f022-14c37650dc3a@linaro.org> Precedence: bulk List-ID: X-Mailing-List: devicetree@vger.kernel.org On Fri, Nov 04, 2022 at 10:52:22AM -0400, Krzysztof Kozlowski wrote: > On 01/11/2022 09:07, Rob Herring wrote: > > On Fri, Oct 28, 2022 at 07:37:01PM -0400, Krzysztof Kozlowski wrote: > >> It is useful to start from existing bindings when writing new ones, > >> especially when one does not know that much DT schema. However we have > >> several bindings which are not the best examples, so people tend to copy > >> their issues into new bindings. > >> > >> Beginners also might not know how to achieve some more complex solutions > >> in DT schema, e.g. how one of two properties should be required by the > >> bindings. Some of such solutions are already in example-schema.yaml, > >> but several other are missing. Add reference with such re-usable > >> design-patterns. > > > > My main concern here is what's a good example today is not tomorrow... > > Yes, I agree. The problem I want to solve is some folks copy-paste some > existing schema as starting point and then are surprised when receive > basic style feedback. > > Of course the optimal solution would be to make all schemas in same > (proper) style, but this is going take some time... > > How to solve this problem other way? I don't know. shrug > >> Signed-off-by: Krzysztof Kozlowski > >> --- > >> .../devicetree/bindings/examples.rst | 63 +++++++++++++++++++ > >> Documentation/devicetree/bindings/index.rst | 1 + > >> 2 files changed, 64 insertions(+) > >> create mode 100644 Documentation/devicetree/bindings/examples.rst > >> > >> diff --git a/Documentation/devicetree/bindings/examples.rst b/Documentation/devicetree/bindings/examples.rst > >> new file mode 100644 > >> index 000000000000..710eea81d8b7 > >> --- /dev/null > >> +++ b/Documentation/devicetree/bindings/examples.rst > >> @@ -0,0 +1,63 @@ > >> +.. SPDX-License-Identifier: GPL-2.0 > >> + > >> +Examples of Devicetree Bindings to use a base > >> +============================================= > >> + > >> +Following Devicetree Bindings in DT Schema are a known good starting point when > >> +writing new bindings: > >> + > >> +1. Simple SPI device: > >> + Documentation/devicetree/bindings/iio/adc/maxim,max11205.yaml > >> + > >> +2. PMIC (MFD) with several sub-devices: > >> + Documentation/devicetree/bindings/mfd/mediatek,mt6370.yaml > >> + > >> +3. Battery charger (power supply): > >> + Documentation/devicetree/bindings/power/supply/bq256xx.yaml > >> + (but use vendor prefix in filename) > >> + > >> +4. Clock controller for several devices with different clock inputs: > >> + Documentation/devicetree/bindings/clock/qcom,mmcc.yaml > >> + > >> +5. GPIO controller: > >> + Documentation/devicetree/bindings/gpio/qcom,wcd934x-gpio.yaml > >> + > >> + > >> +Re-usable design patterns when writing your own bindings > >> +======================================================== > >> + > >> +Following bindings show how to use common pattern of writing bindings: > >> + > >> +1. Property required and present only for one variant. Property cannot appear > >> + on other variants: > >> + Documentation/devicetree/bindings/example-schema.yaml > >> + Line: 212 > >> + > >> +2. Excluding properties, but none are required: > >> + Documentation/devicetree/bindings/mfd/samsung,s5m8767.yaml > >> + Line: 155 > >> + > >> +3. Excluding required properties, but one is required: > >> + Documentation/devicetree/bindings/reserved-memory/reserved-memory.yaml > >> + Line: 91 > >> + > >> +4. Array with numbers (items) from given range - min/max: > >> + Documentation/devicetree/bindings/arm/l2c2x0.yaml > >> + Line: 74 > >> + > >> +5. Array with numbers (items) from given range - enum: > >> + Documentation/devicetree/bindings/display/msm/dsi-controller-main.yaml > >> + Line: 101 > >> + > >> +6. Uint32 matrix, variable length of two-items: > >> + Documentation/devicetree/bindings/iio/adc/st,stm32-adc.yaml > >> + Line: 278 > >> + > >> +7. Phandle to syscon with offset: > >> + Documentation/devicetree/bindings/soc/samsung/exynos-usi.yaml > >> + Line: 42 > >> + > >> +8. Variable length of array (e.g. clocks and clock-names) but narrowed to > >> + specific variant: > >> + Documentation/devicetree/bindings/clock/qcom,mmcc.yaml > >> + Lines: 33 and 71 > > > > It seems like some of these that are just a single property we could add > > to example-schema.yaml. > > I am afraid the example-schema will grow too big for folks to look into. > It's already quite complicated, with explanations of the dtschema > behavior itself. How about splitting up the top-level descriptions and property examples? > How about then RST file with small code snippets? Then they don't validate. > > > Also, perhaps a reference to this from writing-schema.rst. > > I can do this as well. > > The problem with my approach above (and adding these to writing-schema) > is that examples above can change, lines can be inadequate. I just meant a link in writing-schema.rst to this doc. We already have that for example-schema.yaml. Rob