Re: [ECOS] redboot load command on ipaq

[Gary Thomas <gary at chez-thomas dot org>]

On Mon, 2002-07-29 at 15:39, Kevin Moule wrote:
> 
> The redboot load command on my ipaq seems to be behaving very strangely.
> Initially the load command would simple stall with no output, regardless
> of command line options, and I would be forced to reset the ipaq.
> 
> I tried to add some diag_printfs to the do_load function to see how far it
> was getting but I never did see anything. Even with a diag_printf as the
> first statement.
> 
> Next, I tore out the do_load function and all the related static functions
> in load.c and wrote a simple do_load with a single diag_printf. This
> works, I get the one line statement and i can continue with redboot
> commands.
> 
> I put all the proper do_load stuff back in, added a single diag_printf
> followed by a return and now I'm getting either gdb-stub output (from a
> crash of some sort, i assume) or what looks like the output from running
> the ping command (even though i typed load). Its completely unpredictable.
> 
> I'm building redboot from the CVS repository code. I last updated
> everything on friday. The tool chain i'm using was built on friday aswell
> using the following:

I assume that you are trying to use TFTP.  

I tried this and, indeed, the TFTP tranfer failed.  In fact, all network
stuff seems to be broken. It seems to be related to the network gateway
patches I incorporated from Grant Edwards last month.

Try turning off CYGSEM_REDBOOT_NETWORKING_USE_GATEWAY and see if it helps.



-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss