gen405 bsp state?

[Ed Sutter]

Thanks for the responses; however, I'm still stumped.
I saw the issue regarding the linker command file and I think
it's fixed in the latest release.  Also, keep in mind that I'm
simply trying to run a hello world 'C' application.  No C++ involved.

Joel, I guess if you were able to run this BSP, then I have to
conclude that my problem is configuration based; although I used
the recommendations from "Getting Started".  I'm just building the
"hello world" example.  Do you recall what you used as your
initial test application?

When I build this image, the disassembly of _init is...

0001e070 <__init>:
    1e070:   94 21 ff f0     stwu    r1,-16(r1)
    1e074:   7c 08 02 a6     mflr    r0
    1e078:   90 01 00 14     stw r0,20(r1)

0001e07c <__fini>:
    1e07c:   94 21 ff f0     stwu    r1,-16(r1)
    1e080:   7c 08 02 a6     mflr    r0
    1e084:   90 01 00 14     stw r0,20(r1)

This is the end of text, so instructions after this point are invalid.
This is consistent with the linkcmds file.  The output of 'nm' shows
_init() and __init() at the same location.  So clearly the call to
_init() is going to crash the machine.

I'm running with a Cygwin host, is there any reason to be suspicious of
anything there?

Ed



>> in the gen405 BSP, there are currently some issues concerning the linker
>> command file (I think you hit some of these already). Have a look at the
>> mailing list archive, searching for "gen405".
> 
> 
> The last build I did from CVS on this BSP worked.  I recall the
> issue was that the default linkcmds was for RAM and there wasn't
> enough of it for all the tests.  We swapped RAM and Flash as the default
> linker script.
> 
>> wkr,
>> Thomas.
>>
>> Ed Sutter schrieb:
>>
>>> Anyone aware of the state of the gen405 bsp?
>>> I was hoping to do a quick port of it to a ppc405-based board.
>>> Ed Sutter
>>
>>
>>
>>
> 
>