Russell King - ARM Linux wrote:
> On Sun, Apr 14, 2002 at 08:10:55PM +0200, Eric Benard / FREE wrote:
> 
>>I'm investigating on a strange probklem I have on Assabet and a SA1110 
>>board similar to Assabet :
> 
> 
> Gah, there's that expression again...  We don't care that a board is
> "similar to the Assabet".  If it has an SA1110, its similar to an
> assabet.  Tells us absolutely nothing.
> 
sorry about this. I was just talking about the RAM/FLash configuration.
> 
>>My question : in the redboot dump, are the exception handlers properly 
>>mapped and is their content good ?
> 
> 
> Difficult to tell.  If your flash is normally mapped at 0x50000000, then
> they're mostly ok.  The only thing that I'd be concerned about is that
> it seems to want to jump to 0x60, but 'reset_vector' is 0x40.
> 
Exactly this was the problem. This is logic and you are the champion in 
solving problems after only seeing a few lines of assembly ;-))) ... I 
was investigating on the problem since thursday without any success ...
I changed 0x60 to 0x40 in the binary file and this did the trick. now 
the board starts everytime ;-)))

But, how  to explain this problem ?
Can this come from the compiler ? I'm using gcc version 2.95.3 20010315 
(release).

Last question (very stupid one but I want to be sure) :
in the redboot dump, there are a .rom_vectors and a .text section
is the following true ?
- everything in .rom_vectors is address of code to jump to
- everything in .text is assembled code

So a "bootable" file can be done in 2 different ways :
- like bootldr : using branch and only a .text section
- like redboot : using vectors

Many thanks Russell

Eric


_______________________________________________
http://lists.arm.linux.org.uk/mailman/listinfo/linux-arm
http://www.arm.linux.org.uk/armlinux/mailinglists.php
Please visit the above addresses for information on this list.

Reply via email to