Re: starting init on StrongARM (shark)

[un23@rz.uni-karlsruhe.de (Alexander Schulz)]

According to Philip Blundell:
> 
> >So I would like some hints from people who know the exec stuff. Just some
> >intelligent guesses perhaps. Are there any Architecture dependences in the
> >MMU-code (not just processor-specific)? How can I debug this? Where can I
> >find some documentation about this part of the kernel (I only have Rubini
> >here, and starting init ist not exactly a device :-)
> 
> It does sound like your memory mappings are not being set up properly.  
> What exactly do you have in memory.h, and what is the physical layout of the 
> Shark's memory map?

Memory starts at 0x0e00.0000 (but this depends on the amount of memory
installed, I have 32MB) I don't have the source here at the moment, I will
have a look as soon as I am at home. Kernel is mapped at c000.0000 as on the
other arms.

BTW: I forgot to tell this the last time:
The first elf_bss that should be cleared is at 0x4002.1408 the second is at
0x0200.e3d8 each one hangs the machine.

When I comment them out, I get to the debug messages which tell me that the
code also is around 0x0200.0000, the stack at 0xbfff.ff30

So some numbers at least, I'll send you the memory.h stuff.

Thanks
  Alexander