[ECOS] new_net crashes

[NavEcos <ecos at navosha dot com>]

Has anybody else experienced crashes using the new_net template with
i386-pc target?  When I fell back to the "old" net, it still crashed,
but much worse, so that gdb disconnects.

I am setting up a server that just spews out data as fast as it can when
a client connects.  When I try to trace the stack in gdb, the "server" that
I made cannot be traced, the PC ends up at 0x00004 or something
that's impossible.  This makes sense since the error I get is "illegal
instruction" - so the stack must have been corrupted.

I did the obvious and I checked the stack, and I'm not running out of
space.  You can only trace the stack backward on 386 so tracing
the stack requires coding which I am working on.

Is anybody experiencing similar problems?  I'm attempting to debug it
now, but it's very difficult.  The server will NOT crash as LOW data
rates, only high data rates which makes it very difficult to see what
is corrupting the stack.

Thanks,
-Rich

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