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 3DAD4C433EF for ; Sat, 16 Jul 2022 16:09:24 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S229544AbiGPQJX (ORCPT ); Sat, 16 Jul 2022 12:09:23 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:33062 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S229457AbiGPQJW (ORCPT ); Sat, 16 Jul 2022 12:09:22 -0400 Received: from dfw.source.kernel.org (dfw.source.kernel.org [IPv6:2604:1380:4641:c500::1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 91C2065F9; Sat, 16 Jul 2022 09:09:21 -0700 (PDT) Received: from smtp.kernel.org (relay.kernel.org [52.25.139.140]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by dfw.source.kernel.org (Postfix) with ESMTPS id 2E462611A0; Sat, 16 Jul 2022 16:09:21 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id 234B0C34114; Sat, 16 Jul 2022 16:09:20 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1657987760; bh=nuK8GK3uKLya8oToFj66vEMC13us5ufvacB/Y1/YHR0=; h=Date:From:To:Cc:Subject:References:In-Reply-To:From; b=sPJSozviix4Hz3h/Ln2GbrVWzORwGVK9GBOFiW/u2egcTXJqlDjjHnvWHyJlfl9jM 5szkTR/EjgPJ6Pyoq0Xl7TPREBwaFx11KtsMQO8WmwwrUehku6DkVn2VLP7sCPTnj2 bOt9FicTOU7ygEthnMiM4u/IhLXZt3nXzsEKFz3bIpqDpSfsUF4qqjf7Jdr9sVOtxT vZUuFYyPHsOE+HSzZrC6uq4I66Wj4avbNFY1PyeI2e4iHxI2gIX85vC3I2+gqJc6xO GW9gxYe3HjKVMPuu7ZtG47Ygv1hEW4vMWi8X76jg2GnmtLiK7NTZe5jBIPXW+0Ok7h GbijlDI3fO5Tw== Received: by pali.im (Postfix) id DC602A52; Sat, 16 Jul 2022 18:09:16 +0200 (CEST) Date: Sat, 16 Jul 2022 18:09:16 +0200 From: Pali =?utf-8?B?Um9ow6Fy?= To: Uwe =?utf-8?Q?Kleine-K=C3=B6nig?= Cc: Thierry Reding , Lee Jones , Linus Walleij , Bartosz Golaszewski , Rob Herring , Krzysztof Kozlowski , Andrew Lunn , Gregory Clement , Sebastian Hesselbarth , Marek =?utf-8?B?QmVow7pu?= , linux-pwm@vger.kernel.org, linux-gpio@vger.kernel.org, devicetree@vger.kernel.org, linux-kernel@vger.kernel.org, linux-arm-kernel@lists.infradead.org Subject: Re: [PATCH v2 3/4] ARM: dts: armada-38x: Fix compatible string for gpios Message-ID: <20220716160916.jp37siznitgzw6qf@pali> References: <20220714115515.5748-1-pali@kernel.org> <20220714183328.4137-1-pali@kernel.org> <20220714183328.4137-3-pali@kernel.org> <20220716144028.rzwcn4wl5uyxepjd@pengutronix.de> <20220716145019.nps3oh4a22fsuzup@pali> <20220716150751.6yaknmo3qwusyy5h@pengutronix.de> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline Content-Transfer-Encoding: 8bit In-Reply-To: <20220716150751.6yaknmo3qwusyy5h@pengutronix.de> User-Agent: NeoMutt/20180716 Precedence: bulk List-ID: X-Mailing-List: linux-gpio@vger.kernel.org On Saturday 16 July 2022 17:07:51 Uwe Kleine-König wrote: > On Sat, Jul 16, 2022 at 04:50:19PM +0200, Pali Rohár wrote: > > On Saturday 16 July 2022 16:40:28 Uwe Kleine-König wrote: > > > On Thu, Jul 14, 2022 at 08:33:27PM +0200, Pali Rohár wrote: > > > > Armada 38x supports per CPU interrupts for gpios, like Armada XP. Pre-XP > > > > variants like Armada 370 do not support per CPU interrupts for gpios. > > > > > > > > So change compatible string for Armada 38x from "marvell,armada-370-gpio" > > > > which indicates pre-XP variant to "marvell,armadaxp-gpio" which indicates > > > > XP variant or new. > > > > > > > > Driver gpio-mvebu.c which handles both pre-XP and XP variants already > > > > provides support for per CPU interrupts on XP and newer variants. > > > > > > > > Signed-off-by: Pali Rohár > > > > Fixes: 7cb2acb3fbae ("ARM: dts: mvebu: Add PWM properties for armada-38x") > > > > --- > > > > arch/arm/boot/dts/armada-38x.dtsi | 4 ++-- > > > > 1 file changed, 2 insertions(+), 2 deletions(-) > > > > > > > > diff --git a/arch/arm/boot/dts/armada-38x.dtsi b/arch/arm/boot/dts/armada-38x.dtsi > > > > index df3c8d1d8f64..9343de6947b3 100644 > > > > --- a/arch/arm/boot/dts/armada-38x.dtsi > > > > +++ b/arch/arm/boot/dts/armada-38x.dtsi > > > > @@ -292,7 +292,7 @@ > > > > }; > > > > > > > > gpio0: gpio@18100 { > > > > - compatible = "marvell,armada-370-gpio", > > > > + compatible = "marvell,armadaxp-gpio", > > > > "marvell,orion-gpio"; > > > > > > If you can treat the XP variant as 370 and everything that is supposed > > > to work on 370 works then, then maybe the right incarnation is: > > > > > > compatible = "marvell,armadaxp-gpio", "marvell,armada-370-gpio", "marvell,orion-gpio"; > > > > > > ? > > > > For pre-XP variants is "marvell,orion-gpio" enough and for XP + post-XP > > is needed "marvell,armadaxp-gpio" (with possible "marvell,orion-gpio" > > for backward compatibility). > > > > So I do not see reason why to add "marvell,armada-370-gpio" nor what > > value it brings. > > If you boot an older kernel (i.e. one that doesn't support > marvell,armadaxp-gpio, but does support marvell,armada-370-gpio), it Is there such kernel version? > will work better as there are relevant differences between > marvell,orion-gpio and marvell,armada-370-gpio. And if yes, do we really need this in DTS files for new kernel versions? I can imagine that such change can be relevant for old LTS kernel version, but not for new versions. > For example some > registers seem to have a different offset ... armada-370-gpio is mapped to MVEBU_GPIO_SOC_VARIANT_ORION, so it will get same offsets as orion-gpio. So no change.