Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752949AbcCKVf1 (ORCPT ); Fri, 11 Mar 2016 16:35:27 -0500 Received: from vps0.lunn.ch ([178.209.37.122]:55126 "EHLO vps0.lunn.ch" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752581AbcCKVfZ (ORCPT ); Fri, 11 Mar 2016 16:35:25 -0500 Date: Fri, 11 Mar 2016 22:35:16 +0100 From: Andrew Lunn To: David Daney Cc: Florian Fainelli , Mark Rutland , Robert Richter , Pawel Moll , Ian Campbell , netdev@vger.kernel.org, David Daney , David Daney , linux-kernel@vger.kernel.org, Rob Herring , Kumar Gala , Sunil Goutham , Radha Mohan Chintakuntla , "David S. Miller" , linux-arm-kernel@lists.infradead.org Subject: Re: [PATCH 1/3] net: thunderx: Cleanup PHY probing code. Message-ID: <20160311213516.GE19277@lunn.ch> References: <1457714822-5754-1-git-send-email-ddaney.cavm@gmail.com> <1457714822-5754-2-git-send-email-ddaney.cavm@gmail.com> <20160311173125.GI3153@lunn.ch> <56E30332.7060003@caviumnetworks.com> <20160311180030.GB19277@lunn.ch> <56E30DDF.5040506@caviumnetworks.com> <20160311190627.GC19277@lunn.ch> <56E31E7A.6080905@gmail.com> <56E330E4.9090800@gmail.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <56E330E4.9090800@gmail.com> User-Agent: Mutt/1.5.21 (2010-09-15) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 725 Lines: 19 > For this phy, we have: > > compatible = "cortina,cs4223-slice"; That actually means something else is happening, i think. of_mdiobus_register() looks at the children, and decides if each child is a phy or an mdio device, by calling of_mdiobus_child_is_phy(). Since this compatible string is not in whitelist_phys[], it will return false. of_mdiobus_register() will then do a of_mdiobus_register_device(). This compatible means it is an MDIO device, not a PHY. So when you later call of_phy_find_device() it is always going to return NULL, because there is no PHY there, only an MDIO device. How usable is the hardware without a PHY driver? Is a better solution that your write a very minimal PHY driver? Andrew