Fwd: Better GDB support for rtems.

[Dhananjay Balan]

Forgot to CC.

---------- Forwarded message ----------
From: Dhananjay Balan <mb.dhananjay at gmail.com>
Date: 8 April 2013 07:44
Subject: Re: Better GDB support for rtems.
To: Joel Sherrill <joel.sherrill at oarcorp.com>


I have already followed the GDB python tutorials and documentation. The
tutorial and helpers wanted was to get the right consciousness around
debugging an RTEMS application with GDB.

sim-scripts seems a good start. Thanks.


On 6 April 2013 21:26, Joel Sherrill <joel.sherrill at oarcorp.com> wrote:

> On 04/04/2013 03:13 PM, Dhananjay Balan wrote:
>
>> Hi,
>>
>> I was fooling around rtems (Not much, was taking tests till yesterday)
>> and the general debugging stuff (Structures and stuff..), but I couldn't
>> find much documentation about it.
>>
> If your goal is the GSOC project to provide Python plugins to provide more
> insight
> into RTEMS data structures, then you are in a no-man's land between an
> internal
> capability of gdb which is new and likely not well documented and the
> internals
> of RTEMS which require some insight to print in a useful way.
>
> The key point to gdb support for displaying more about an RTEMS application
> is that it doesn't require any code to run on the target hardware.
>
>
>> I was stuck at following points,
>>
>> * Is there any specific tutorials or anything about GDB and rtems?
>>
> Not particularly. Once you connect to the target, any gdb tutorial
> should be sufficient. gdb is always gdb -- the target changes and the
> connection method to the target changes -- but the gdb interface is
> the same.
>
> So you could experiment with Python pretty-printing on a native gdb
> and no RTEMS.
>
>  * I could find from the wiki that rtems provides a primitive shell to
>> debug applications to a limit, I found the document on the shell - but I am
>> confused how to run it in a simulator. Any help?
>>
> See the fileio.exe sample. The "monitor" class of commands and the "rtems"
> class
> of commands give some
>
> Another option to look at is the gdb macros in rtems-testing/sim-scripts.
> This is how
> you pretty printed things before the Python support in gdb. I wrote most
> of those
> and getting Python pretty printing on par with those capabilities would be
> a big
> improvement plus it is fairly easy to see where to go from there. It is
> far from
> complete.
>
>
>  * Afiu the simulator is a sparc sis (arch-run command) I am using, is
>> there any advantage if I run a qemu-sparc instance instead?
>>
>>  Not for adding python pretty printing support for rtems structures to
> gdb.
> It will just make connecting to the sparc target more complicated.
>
>  --
>> Regards,
>> Dhananjay M Balan,
>>
>>
>
> --
> Joel Sherrill, Ph.D.             Director of Research& Development
> joel.sherrill at OARcorp.com        On-Line Applications Research
> Ask me about RTEMS: a free RTOS  Huntsville AL 35806
> Support Available               (256) 722-9985
>
>


-- 
Regards,
Dhananjay M Balan,
Department Of Computer Science,
College Of Engineering, Trivandrum.
http://cet.ac.in



-- 
Regards,
Dhananjay M Balan,
Department Of Computer Science,
College Of Engineering, Trivandrum.
http://cet.ac.in
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.rtems.org/pipermail/devel/attachments/20130408/abfae448/attachment-0001.html>