From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1759810AbZE2N3p (ORCPT ); Fri, 29 May 2009 09:29:45 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1757425AbZE2N3i (ORCPT ); Fri, 29 May 2009 09:29:38 -0400 Received: from mx1.emlix.com ([193.175.82.87]:36188 "EHLO mx1.emlix.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756785AbZE2N3h (ORCPT ); Fri, 29 May 2009 09:29:37 -0400 Message-ID: <4A1FE338.5010802@emlix.com> Date: Fri, 29 May 2009 15:29:28 +0200 From: Simon Braunschmidt User-Agent: Thunderbird 2.0.0.21 (X11/20090409) MIME-Version: 1.0 To: wuzhangjin@gmail.com CC: linux-mips@linux-mips.org, linux-kernel@vger.kernel.org Subject: Re: [loongson-PATCH-v1 16/27] Add Siliconmotion 712 framebuffer driver References: In-Reply-To: Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Organization: emlix gmbh, Goettingen, Germany Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Hi Wu Zhangjin I applied your patch cleanly and run it on a i486SX (little endian) embedded computing board. This provides a huge performance benefit compared to vesa framebuffer, so thanks a lot for the contribution. I tested it in both the library-accelerated (cfb...) and hardware accelerated codepaths, with no problems. Tested with qt-embedded 4.5.1. As i have seen lots of functionality are guarded by #ifdef __BIG_ENDIAN Is there a way to setup a different mode than 1024x600 on kernel-commandline or from userspace, e.g. with fbset? What I currently do is picking appropriate values form the table VGAMode in smtcfb.h and hardcoding them. I tried 800x600x16bit@60hz, which worked also, but 1024x768x24bit@60hz left me with a corrupted display output (distored colors, missing regions), i did not check further wheter it also crashed. I dont know wheter i am subscribed and this will make it through to lkml and linux-mips, feel free to cite me if you want. Also if you want you may add my Tested-by: Simon Braunschmidt see also attached a patch for the debug messages From addec9a5a27512621c538a2f799e31e07720cb9c Mon Sep 17 00:00:00 2001 From: Simon Braunschmidt Date: Fri, 29 May 2009 15:03:51 +0200 Subject: [PATCH] [video] fix debug messages in smtcfb when defining #DEBUG --- drivers/video/smi/smtcfb.c | 6 ++++-- 1 files changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/video/smi/smtcfb.c b/drivers/video/smi/smtcfb.c index 33ce878..ebc6371 100644 --- a/drivers/video/smi/smtcfb.c +++ b/drivers/video/smi/smtcfb.c @@ -44,6 +44,8 @@ #include "smtcfb.h" #include "smtc2d.h" +#define DEBUG + #ifdef DEBUG #define smdbg(format, arg...) printk(KERN_DEBUG format , ## arg) #else @@ -1087,7 +1089,7 @@ static int __init sm712be_setup(char *options) sm712be_flag = 0; if (!options || !*options) { retval = 1; - smdbg("\n No sm712be parameter\n", __LINE__); + smdbg("\n No sm712be parameter\n"); } if (!retval && strstr(options, "enable")) sm712be_flag = 1; @@ -1113,7 +1115,7 @@ static int __init sm712vga_setup(char *options) if (!options || !*options) { retval = 1; - smdbg("\n No vga parameter\n", __LINE__); + smdbg("\n No vga parameter\n"); } screen_info.lfb_width = 0; -- 1.6.3.rc0.1.gf800