Dear Andreas Bießmann,
please do not hijack a thread!
sorry I didn't want do this. I wanted to give some more information.
Ok, you use 2010.09. This was last release before relocate changes. Therefore it has no connection with the relocation stuff.
yes I. I tried to make a patch for 2010.12 but the uboot don't starts so I tried in the first step the version before. It is a similar board to the at91rm9200dk. It hangs in the low level_init file. But I have only a keil jtag debugger and couldn't debug it. or is there any possibility to use it with the gdb?
I found the problem in the board.c of the arm architecture. But can't see why. Line 303 is reached but line 312 not. Also line 306 is not reached. I don't know why, maybe you can give me a hint?
I think if the problem is found the problem with the at91rm9200dk is also found.
Is it nessesary to set a console parameter in the bootcmd in uboot version 2010.09?
We don't know your configuration. What tells your working setup in kernel log messages about 'Command line'? -> /proc/cmdline
I only give the MAC and the running image in the cmdline. I think the problem must be in the uboot.
What is your board exactly? Is it currently in official u-boot or do you have another tree? at91rm9200dk is most likely broken with current u-boot (2010.12) and will be removed for 2011.03. So if you have a at91rm9200dk at hand please try to get the current u-boot working with that board and provide patches. You may see my patches for at91rm9200ek which at least make at91rm9200ek boot from SDRAM when setup correctly with an jtag (-> booting with at91bootstrab should also work but is untested).
It isn't in official u-boot tree. You can have a look in the patch.
Best regards / Mit freundlichen Grüßen
Maik Hänig