From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754135AbcANL1l (ORCPT ); Thu, 14 Jan 2016 06:27:41 -0500 Received: from gate.crashing.org ([63.228.1.57]:44570 "EHLO gate.crashing.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753501AbcANL1i (ORCPT ); Thu, 14 Jan 2016 06:27:38 -0500 Message-ID: <1452770755.31558.13.camel@kernel.crashing.org> Subject: Re: [PATCH v1 3/3] ARM64 LPC: update binding doc From: Benjamin Herrenschmidt To: Rongrong Zou , arnd@arndb.de, catalin.marinas@arm.com, will.deacon@arm.com Cc: lijianhua@huawei.com, lixiancai@huawei.com, linuxarm@huawei.com, linux-kernel@vger.kernel.org, minyard@acm.org, gregkh@linuxfoundation.org Date: Thu, 14 Jan 2016 22:25:55 +1100 In-Reply-To: <5697274C.1000606@gmail.com> References: <1451396032-23708-1-git-send-email-zourongrong@gmail.com> <1451396032-23708-4-git-send-email-zourongrong@gmail.com> <1452727756.2403.47.camel@kernel.crashing.org> <569701E3.2090307@gmail.com> <1452742757.31558.8.camel@kernel.crashing.org> <5697274C.1000606@gmail.com> Content-Type: text/plain; charset="UTF-8" X-Mailer: Evolution 3.18.3 (3.18.3-1.fc23) Mime-Version: 1.0 Content-Transfer-Encoding: 8bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Thu, 2016-01-14 at 12:42 +0800, Rongrong Zou wrote: > > Right, and the "compatible" property should be something like the > > specific implementation of the LPC bridge. For example, ibm,power8- > > lpc > > in my case. Not something generic. > > > > Maybe we could allow for a generic one if the LPC is directly MMIO > > mapped via the ranges property. > > It is not directly MMIO mapped actually. I know yours is not. But some are. My point is that we should have a binding that is either completely specific to your ARM64 LPC or we should have a generic LPC binding with provisions for implementation specific stuff such as ARM64 or POWER8 which are both not MMIO mapped. I go for the latter. So "ranges" if you are mapped, otherwise "reg", and in the latter case, the compatible property should be much more specific like it is for P8,   .../... > The big problem is we do not want the "ranges" property, but we can't > get resource if the property is absent, you could see discussion at > https://lkml.org/lkml/2016/1/11/631. That's fixable. I missed the discussion but I'll have a look tomorrow. Cheers Ben.