Re: [Simulavr-devel] Re: raising signals from gdb

[Keith Gudger]

Ted:

simulavr was segmentation faulting..  But, of course, it was my *fault*.
I fixed it and now ports E & F work.  I've also allowed port sizes of 2,
4, 7 & 8 bits.  Any reason we shouldn't allow sizes of 1-8 bits?

Keith

On Fri, 10 Oct 2003, Theodore A. Roth wrote:

> 
> 
> On Fri, 10 Oct 2003, Keith Gudger wrote:
> 
> > Ted:
> >
> > I implemented the signal stuff, using "SIG80" as the start interrupt (to
> > make room for all 32 vectors).  It works great.
> >
> > I checked out my changes to devsupp.c and intvects.c for the USB devices,
> > and they look OK.
> >
> > To finish checking it out, I need to add support for Ports E & F.  I did
> > that, and it compiled OK.  My ports are 4 bits wide, so I added "
> > PORT_4_BIT  = 4" to the ports.h file.  At that point, simulavr can finally
> > try out a couple of the new devices ... and it gives a core dump.
> 
> I assume that you are talking about the 'core_avr_dump.core' file and
> not that simulavr is seg faulting. ;-)
> 
> >
> > Can you point me to documentation on what to do with the dump?  Thanks.
> 
> I haven't looked at one of those files in a really long time. It's
> just a dump of the state of the avr at the time of the dump.
> 
> About the only thing you can do with it is look at it with a text
> editor. ;-)
> 
> Way back before I added the gdbserver mode, the dump file was the only
> way to look at the state of the program being simulated. It's really
> been obsoleted by the disp programs and gdb support.
> 
> Ted Roth
>