Re: [openrisc] Question about or1ksim with gdb

[Marko Mlinar <markom@opencores.org>]

On Monday 26 August 2002 11:37, SUNGYON@aaww.com wrote:
> Chris Ziomkowski and many forerunner!
> Nowasdays i sended many Questions here and i got trouble-shooting answers.
> I finished the compilation of "mini-mad" program with or32-rtems-.....
> thanks to some of you.
> I want to run this code and see the printf message, but even hello.exe(in
> rtems4.5) failed.
>
> I see the README.or1k which Chris wrote in or1k/rtems/, the link of or1ksim
> & gdb5.0.
> There, it seems that he is successful to run "hello.exe" with or1ksim & gdb
> interface.
> But in my case, gdb does not be connected to the or1ksim (server).
> The jtag port is opened for sure, and gdb find the port, but gdb cannot
> finish the jtag connection.
>
> I edited memory section of sim.cfg according to configuation file used by
> 'or32-rtems-ld'
> I try the standalone simulation of or1ksim. it resulted in segmentation
> fault. If i escape the connection peding, and jump to highly probable start
> address(in gdb), it also fails.
> I think the critical problem is insufficient jtag connection.
>
> If i success in the hello.exe of rtems4.5, i think all preliminary work for
> my minimad analysis was done.
> Sincerly.
gdb-or1ksim connection was not maintained, since Chris stopped working. After 
him nobody really did the effort to test everything again. The reason is that 
everybody I know uses just or1ksim for debugging -- it is much simpler for 
low level debugging and there is possibility of execution logs, which really 
speed up development.

I don't have enough time for another project, but I would be willing to help a 
volounteer to test this connection.

gdb-board connection is well tested and widely used.

I am not sure about RTEMS status. Ivan was working on latest RTEMS version and 
newlib.

Marko

--
To unsubscribe from openrisc mailing list please visit http://www.opencores.org/mailinglists.shtml