Re: [avr-gcc-list] gdb-5.0 patch + simulator + hardware

[torhr]

Karsten Schmidt wrote:

> > gdb4.18 won't work with binutils 2.11.2, you'll need an earlier
> > version and that will mean an earlier gcc as well as there were
> > some register name convention changes at some stage.

> I'm disagreeing on this. I successful compiled gdb4.18 with binutils
> 2.11.2 and tested the stuff with avrmon-stk200-0.6.0. (See
> http://www.pvv.ntnu.no/~torhr/avrmon-stk200 for this).

Please don't take my word for a fact when it comes to claiming
portability between gdb-4.18 and binutils-2.11.2 / gcc-3.0. I simply
used the latest version available for each tool at a certain time.

When that's said, I can confirm that the configuration seems to work
pretty well most of the time. I have seen one problem, however, that I
can't remember were there before; trying to dump a call stack trace
seem to get the debugger stuck in some endless loop.

> ok it's slow, 

Yup. The communication protocol isn't exactly optimised for
speed. It's very robust, though. I don't think I've ever managed to
make it break down, even when I use the same pins both for debugger
communication and doing real work at the same time.

> Using avrmon-stk200-0.6.0 together with the newly and cleanly
> compiled gdb5.0 and setting the target to avr-mon results in the
> following:
>
> [...]

I guess I'll have to look into that as soon as get gdb-5.0 up and
running.


- Tor Ringstad -