Skip site navigation (1)Skip section navigation (2)
Date:      Sat, 2 May 2009 17:30:25 +0200
From:      Rafal Jaworowski <raj@semihalf.com>
To:        Donald T Hayford <don@donhayford.com>
Cc:        freebsd-arm@freebsd.org
Subject:   Re: Help with Marvel kernel for 88F6281
Message-ID:  <72D23F68-EE62-4297-88F4-6CA0132F0293@semihalf.com>
In-Reply-To: <49FB8696.8020907@donhayford.com>
References:  <49FA5B75.9090008@donhayford.com> <D12A7520-0AA8-4A25-8FF7-EE75FE91C89D@semihalf.com> <49FADE9B.1080806@donhayford.com> <494D378B-B243-4D97-8554-AC3E74A30B8C@semihalf.com> <49FB8696.8020907@donhayford.com>

next in thread | previous in thread | raw e-mail | index | archive | help

On 2009-05-02, at 01:32, Donald T Hayford wrote:

> Rafal Jaworowski wrote:
>>
>> On 2009-05-01, at 13:35, Donald T Hayford wrote:
>>
>>> Rafal Jaworowski wrote:
>>>>
>>>> On 2009-05-01, at 04:16, Donald T Hayford wrote:
>>>>
>>>>> I built FreeBSD-8 (current) kernel and world for the Marvel  
>>>>> chipsets using the instructions here:
>>>>> http://wiki.freebsd.org/FreeBSDMarvell
>>>>>
>>>>> When I tried booting a Sheevaplug (Marvel 88F6281), the system  
>>>>> hangs as soon as I start it:
>>>>
>>>> Please show the output of the 'map' command at U-Boot prompt.
>>>>
>>> I apologize for the bad pasting job - I'll try again.  Hopefully  
>>> this one is easier to read.
>>>
>>> Marvell>> map CPU Interface
>>> -------------
>>
>> [...]
>>
>> I don't see any mismatches WRT internal SOC registers location etc.  
>> and need to look a bit closer to the SheevaPlug docs.
>>
>> Just a basic clarification: you are 100% sure the correct  
>> DB-88F6XXX kernel image is used, right?
>>
>> Rafal
>>
> I'm as sure as I can be.
>
> [verify the directory on the FreeBSD machine is the 6XXX directory]

OK, thanks for verfiication.

> Marvell>> go 0x900000
> ## Starting application at 0x00900000 ...
> [sheevaplug hangs up]
>
> Note that the file size that was loaded by UBoot was 2863204 bytes  
> long, the same as the length of the file in the directory listing  
> above.  The 5XXX kernel is slightly longer at 2870196 bytes.

Please do a quick experiment: eliminate (#if 0) contents of the  
platform_mpp_init() in sys/arm/mv/kirkwood/db88f6xxx.c and recompile/ 
rerun. The SP device could have MPP/GPIO layed out differently (note  
you're using DB-88F6281 dev board configuration): we could be  
overwriting UART lines connection settings and hence lose console  
output.

Rafal




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?72D23F68-EE62-4297-88F4-6CA0132F0293>