From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756128Ab2HGTcl (ORCPT ); Tue, 7 Aug 2012 15:32:41 -0400 Received: from mail-bk0-f46.google.com ([209.85.214.46]:46062 "EHLO mail-bk0-f46.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751639Ab2HGTck (ORCPT ); Tue, 7 Aug 2012 15:32:40 -0400 Message-ID: <50216D52.3030402@gmail.com> Date: Tue, 07 Aug 2012 21:32:34 +0200 From: Sebastian Hesselbarth User-Agent: Mozilla/5.0 (X11; Linux i686; rv:10.0.5) Gecko/20120624 Icedove/10.0.5 MIME-Version: 1.0 To: Russell King - ARM Linux CC: Arnd Bergmann , Jason Cooper , Andrew Lunn , linux-arm-kernel@lists.infradead.org, linux-kernel@vger.kernel.org, Rabeeh Khoury , Ian Molton , Maen Suleiman , Olof Johansson , Thomas Petazzoni Subject: Re: [PATCH 0/6] ARM: dove: DT support for Marvell Dove References: <1344255815-4457-1-git-send-email-sebastian.hesselbarth@gmail.com> <201208061402.22815.arnd@arndb.de> <20120807165350.GG18957@n2100.arm.linux.org.uk> In-Reply-To: <20120807165350.GG18957@n2100.arm.linux.org.uk> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 08/07/2012 06:53 PM, Russell King - ARM Linux wrote: > I've recently pushed a fix into mainline (and stable) for VFP support > with Thumb instructions. I noticed that the Cubox kernel had a different > fix, this "different" fix will need to be removed when Cubox folk update > to the next kernel from mainline. Russell, I read about that fix but honestly, that is way to deep into ARM architecture for me. All I realized while having my Cubox is that the supplied kernel tree is very messy and on Debian wheezy/armhf almost all applications started over ssh -X fail with segfault. I was hoping that bringing the cubox into mainline kernel tree will help here - and luckily the (hacking) community is growing! > Video support, not great. The supplied Ubuntu Lucid stuff install > (including the latest dev kernel from the solid-run folk) doesn't work > with any of the tree HDMI devices I've tried it with: HDMI is another topic that I want to work on as soon as the main SoC stuff is ported over to mainline. HDMI output actually is done by some nxp HDMI transmitter that translates the rgb 888 video from lcd controller on dove. AFAIR the HDMI transmitter code is "open", i.e. free to clean up and port. I haven't looked closely into lcd yet but on Cubox the lcd pixclk comes from an external clk chip (si5351a) that I have a quite ready driver for. The existing driver does a table based setup of the plls and does not hit 148.5 MHz precisely. The driver I wrote is based on an AN from SiLabs and is also already based on common clk framework. > . A LCD PC monitor with HDMI input. If the monitor is disconnected, the > Cubox booted, and the monitor connected, you get a picture, but you > lose the top, left, bottom and right of the display off the side of the > monitor. No amount of adjusting on the monitor fixes this. There have been many reports of people no being able to get proper resolutions with the original kernel setup. I know there are people working on gfx support and xbmc. Maybe you hit the solid-run.com forums and have a look there. > What is obvious is that there is no HDMI hotplug support and reconfiguring > of the display. So, I've been investigating the Linux DRM/DRI support - > both the kernel and X11 side, and I have something that is getting close > to being ready. Last night I finally disabled the kernel dovefb driver, > and have X11 using solely my new DRM driver. Great, keep me posted and I ll be happy to test it. > However, this won't be using the on-chip mpeg decoding (yet). CPU > usage sits at around 30% while vlc playback is running. The bad thing with GLES and video decoding is that it all relies on some dirty closed source user space libraries. But again, as I am not that familiar with gfx and decoding drivers, I didn't look into that. Sebastian