[Qi] bad magic when booting from NAND
Michael Trimarchi
trimarchi at gandalf.sssup.it
Wed Apr 8 11:12:28 CEST 2009
Hi,
Dmitry Kurochkin wrote:
> On Wed, Apr 8, 2009 at 12:18 PM, Michael Trimarchi
> <trimarchi at gandalf.sssup.it> wrote:
>
>> Hi,
>>
>> Michael Trimarchi wrote:
>>
>>> Dmitry Kurochkin wrote:
>>>
>>>> I did some more tests.
>>>>
>>>> When I do build unmodified qi, flash the image, it runs fine.
>>>>
>>>> But if I insert a single puts() line in src/cpu/s3c2442/start_qi.c :
>>>> start_qi, just after qi info is printed and before
>>>> bootloader_second_phase is called, it hangs and prints Ts.
>>>>
>>>>
>>> Two things. You are not in stepping stone size or just you don't have
>>> register
>>> what function must use puts. set_putc_func.
>>> Michael
>>>
>>>
>>>
>> Can you post your diff?
>>
>
> This is the only difference except for config.mk:
>
> diff --git a/src/cpu/s3c2442/start_qi.c b/src/cpu/s3c2442/start_qi.c
> index d7136fd..4dc4804 100644
> --- a/src/cpu/s3c2442/start_qi.c
> +++ b/src/cpu/s3c2442/start_qi.c
> @@ -113,6 +113,7 @@ void start_qi(void)
> puts(", ");
> puts((this_board->get_board_variant)()->name);
> puts("\n");
> + puts("HAHAHA\n");
>
>
.text __steppingstone_always : AT (0x0)
{
src/cpu/s3c2442/start.o (.text .rodata* .data .bss)
src/cpu/s3c2442/lowlevel_init.o (.text .rodata* .data .bss)
src/cpu/s3c2442/start_qi.o (.text .rodata* .data .bss)
src/cpu/s3c2442/nand_read.o (.text .rodata* .data .bss)
src/cpu/s3c2442/serial-s3c24xx.o (.text .rodata* .data .bss)
src/memory-test.o (.text .rodata* .data .bss)
src/utils.o (.text .rodata* .data .bss)
src/ctype.o (.text .rodata* .data .bss)
* (.steppingstone)
}
here is the stepping stone,do you have change any of these file? and add
code,
You have only 4kb :)
Michael
> /*
> * jump to bootloader_second_phase() running from DRAM copy
>
> Regards,
> Dmitry
>
>
>> Michael
>>
>>
>
>
More information about the openmoko-kernel
mailing list