From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Miller Subject: Re: [PATCH (net.git) 0/2] stmmac: review and fix the dwmac-sti glue-logic Date: Tue, 14 Oct 2014 16:41:17 -0400 (EDT) Message-ID: <20141014.164117.553206792335727136.davem@davemloft.net> References: <1413267176-325-1-git-send-email-peppe.cavallaro@st.com> Mime-Version: 1.0 Content-Type: Text/Plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: netdev@vger.kernel.org, maxime.coquelin@st.com To: peppe.cavallaro@st.com Return-path: Received: from shards.monkeyblade.net ([149.20.54.216]:44499 "EHLO shards.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755270AbaJNUlT (ORCPT ); Tue, 14 Oct 2014 16:41:19 -0400 In-Reply-To: <1413267176-325-1-git-send-email-peppe.cavallaro@st.com> Sender: netdev-owner@vger.kernel.org List-ID: From: Giuseppe Cavallaro Date: Tue, 14 Oct 2014 08:12:54 +0200 > This patch is to review the whole glue logic adopted on STi SoCs that > was bugged. > In the old glue-logic there was a lot of confusion when setup the > retiming especially for STiD127 where, for example, the bits 6 and 7 > (in the GMAC control register) have a different meaning of what is > used for STiH4xx SoCs. So we cannot adopt the same glue for all these > SoCs. > Moreover, GiGa on STiD127 didn't work and, for all the SoCs, the RGMII > couldn't run when the speed was 10Mbps (because the clock was not properly > managed). > Note that the phy clock needs to be provided by the platform as well as > documented in the related binding file (updated as consequence). > > The old code supported too many configurations never adopted and validated. > This made the code very complex to maintain and debug in case of issues. > > The patch simplifies all the configurations as commented in the tables > inside the file and obviously it has been tested on all the boards > based on the SoCs mentioned. > > With this patch, the dwmac-sti is also ready to support new configurations that > will be available on next SoC generations. Series applied.