From mboxrd@z Thu Jan 1 00:00:00 1970 From: Thomas Petazzoni Subject: Re: [PATCH 2/2] net: mvneta: fix usage as a module on RGMII configurations Date: Thu, 3 Apr 2014 11:36:52 +0200 Message-ID: <20140403113652.33f11413@skate> References: <1395789942-20764-1-git-send-email-thomas.petazzoni@free-electrons.com> <1395789942-20764-3-git-send-email-thomas.petazzoni@free-electrons.com> <533C1B75.2020803@free-electrons.com> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Cc: "David S. Miller" , netdev@vger.kernel.org, Ezequiel Garcia , Lior Amsalem , Steve McIntyre , linux-arm-kernel@lists.infradead.org, stable@vger.kernel.org To: Gregory CLEMENT Return-path: In-Reply-To: <533C1B75.2020803@free-electrons.com> Sender: stable-owner@vger.kernel.org List-Id: netdev.vger.kernel.org Dear Gregory CLEMENT, On Wed, 02 Apr 2014 16:15:17 +0200, Gregory CLEMENT wrote: > > This commit was successfully tested with mvneta compiled as a module, > > on both the OpenBlocks AX3 (SGMII configuration) and the Armada XP GP > > (RGMII configuration). > > Unfortunately with this patch, mvneta doesn't work anymore on the > Mirabox (Armada 370 based board) on 3.14.I didn't managed to do a > simple ping. > > Once I removed this commit then the driver worked again. Problem reproduced. It turns out that some RGMII platforms need the PCS_ENABLE bit to be set (e.g: Armada XP GP), while some other platforms need the PCS_ENABLE bit to be cleared (e.g: Armada 370 Mirabox). I've verified that on both platforms. I've asked for more details about this bit to understand in which situation it should be set or cleared. I'll get back to you with an updated patch once I have enough information to write a fix. Thanks for the report, Thomas -- Thomas Petazzoni, CTO, Free Electrons Embedded Linux, Kernel and Android engineering http://free-electrons.com