From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932133AbcARNET (ORCPT ); Mon, 18 Jan 2016 08:04:19 -0500 Received: from unicorn.mansr.com ([81.2.72.234]:44736 "EHLO unicorn.mansr.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755073AbcARNEQ convert rfc822-to-8bit (ORCPT ); Mon, 18 Jan 2016 08:04:16 -0500 From: =?iso-8859-1?Q?M=E5ns_Rullg=E5rd?= To: Marc Gonzalez Cc: Thomas Gleixner , Jason Cooper , Marc Zyngier , LKML , Linux ARM , Sebastian Frias Subject: Re: [RFC PATCH v2] irqchip: Add support for Tango interrupt controller References: <569CE0F2.1060507@sigmadesigns.com> Date: Mon, 18 Jan 2016 13:04:08 +0000 In-Reply-To: <569CE0F2.1060507@sigmadesigns.com> (Marc Gonzalez's message of "Mon, 18 Jan 2016 13:56:18 +0100") Message-ID: User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/24.5 (gnu/linux) MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8BIT Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Marc Gonzalez writes: > + if (of_property_read_u32(node, "reg", &ctl)) > + panic("%s: failed to get reg base", node->name); > + > + chip = kzalloc(sizeof(*chip), GFP_KERNEL); > + chip->ctl = ctl; > + chip->base = base; I think it might be better to let the OF core convert the "reg" property to a struct resource, then get the offset by subtracting from the base (physical) address. Otherwise it will break badly if someone thinks it's safe to drop "ranges" from the outer node and use absolute addresses in the child nodes (that's how the "reg" property typically behaves). Anyway, thanks for revisiting this. I never got a straight answer (or any answer for that matter) when I asked which approach would be preferred, and then I had better things to do than keep trying patches until people got happy. -- Måns Rullgård