Re: [ECOS] GDB stubs problems ARM/AT91/EB40 -

[Jonathan Larmour <jlarmour at redhat dot com>]

Marc Pignat wrote:
> 
> I think that I've found a (partial) answer to my question ...
> 
> the stack pointer of the main thread is the stack pointer of gdb stubs.

Actually what happens is that if the stubs are to be entered, the state is
stored on the stack of the current thread, whatever that is, and then we
switch to a different stack. So calling into the stub shouldn't make much
difference to the stack usage of any threads.
 
> eCos as reserved a stack for the main thread (as defined in the configuration tool, but the main thread don't use it.
> 
> why the main thread does not take this stack ???

Perhaps it hasn't run yet, or has already exited.

You can change the config to stop there being a main thread at all if you
like.

But from your trace it looks like there's some weird recursion going on
which is what is eating through your stack. It looks like you haven't
finished processing one timeslice interrupt before getting another. 

Jifl
-- 
Red Hat, Rustat House, Clifton Road, Cambridge, UK. Tel: +44 (1223) 271062
Maybe this world is another planet's Hell -Aldous Huxley || Opinions==mine

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