[okl4-developer] More NSLU2 debugging

[Geoffrey Lee]

On Mon, Jun 23, 2008 at 03:07:21PM +1000, Joshua Root wrote:
> Geoffrey Lee wrote:
>> On Sun, Jun 22, 2008 at 07:42:35PM +1000, Joshua Root wrote:
>>> So you'd expect something like:
>>>
>>> add_mapping_init( kspace_phys, 
>>> (addr_t)(XSCALE_DEV_PHYS+CONSOLE_OFFSET),   
>>> (addr_t)(XSCALE_DEV_PHYS+CONSOLE_OFFSET), uncached );
>>>
>>> to work if I do it at the same time as the other 1:1 mappings? And 
>>> after  VM is turned on here (init.cc:353):
>>>
>>> /* Enable virtual memory, caching etc */
>>>     write_cp15_register(C15_control, C15_CRm_default,
>>>                     C15_OP2_default, C15_CONTROL_KERNEL);
>>>
>>> I should be able to write to the UART, yes? (And indeed I can in the  
>>> simulator...) I've been doing pretty much what you suggest to 
>>> pinpoint  where it's failing, but I can't get any output after that   
>>> write_cp15_register call. Could kspace_phys be constructed 
>>> incorrectly?
>>
>> Yes, I guess something hasn't been constructed properly.  I think if
>> you turn off the MMU initialization you will actually get a bit  
>> further - not saying that is the way to move forward, just that
>> it properly is something that got put into the cpu which it didn't
>> like.
>>
>> It might be a good idea to print out the contents of physbase to double
>> check that it's been patched properly, and also the contents of the
>> page table etc that you write to the hardware.
>
> Good call. Turns out that the physbase pointer that's passed to  
> init_memory is null. I guess this means that the build system is failing  
> to set __phys_addr_ram?

Yes, possibly something has gone wrong there.  Just a side note
we never set the lower bound of the physical range to be zero.
In the machines.py description file can you try changing the 
physical range to be something else (say 0x100000L), clean out
the build and then do a full rebuild?

If the contents of __phys_addr_ram is zero that would probably 
make sense since your lower bound for physical range is set to zero,
but it might have assumed zero to mean something else.

>
> - Josh
>

	-gl