From mboxrd@z Thu Jan  1 00:00:00 1970
From: Anatolij Gustschin <agust@denx.de>
Date: Fri, 7 Nov 2014 00:40:43 +0100
Subject: [U-Boot] socfpga: mainline cannot boot linux
In-Reply-To: <545BFE54.8010909@opensource.altera.com>
References: <1414007764-24094-1-git-send-email-marex@denx.de>
	<1414007764-24094-7-git-send-email-marex@denx.de>
	<545BFE54.8010909@opensource.altera.com>
Message-ID: <20141107004043.2f09f721@crub>
List-Id: <u-boot.lists.denx.de>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
To: u-boot@lists.denx.de

Hi Dinh,

On Thu, 6 Nov 2014 17:03:48 -0600
Dinh Nguyen <dinguyen@opensource.altera.com> wrote:

> Hi Marek,
> 
> Have you been able to successfully boot Linux using the mainline uboot?
> I cannot seem to be able to boot linux on the C5 SocDK.
> 
> It starts to boot linux, but then reboots after a bit back into u-boot.
> So I commented out #define CONFIG_HW_WATCHDOG in
> include/configs/socfpga_cyclone5.h. That helped so that I can look that
> __log_buf.
> I was able to trace it down that it was not able to get a tty console,
> because serial8250_console_setup() is failing at:
> 
> if (!port->iobase && !port->membase)
> 
> For reasons I haven't figured out yet, both port->iobase and
> port->membase are NULL.
> 
> So no debug console, but Linux is able to boot all the way to the point
> of tyring to mount a FS.
> 
> Perhaps some kind of memory setup issue? I've seen a thread where
> Chin-Liang was able to boot, but I don't know if he did any other tricks.
> 
> Just wondering if I'm missing a patch or something?

the issue could be the wrong console= argument on the kernel command
line. In the current socfpga_config5.h I see

  #define CONFIG_BOOTARGS         "console=ttyS0" __stringify(CONFIG_BAUDRATE)

A comma is missing after "console=ttyS0"

It should be

  #define CONFIG_BOOTARGS         "console=ttyS0," __stringify(CONFIG_BAUDRATE)

Could you please test it? Thanks.

Best regards,

Anatolij