GSoC 2016 Interested in Tracing was Re:

Isaac Gutekunst isaac.gutekunst at vecna.com
Thu Mar 3 17:00:48 UTC 2016


Hi Vivek,

Chris is definitely the person most knowledgeable in this area.





On 03/03/2016 11:45 AM, Joel Sherrill wrote:
>
>
> On Thu, Mar 3, 2016 at 10:25 AM, Vivek Kukreja <vivekkukreja at outlook.com
> <mailto:vivekkukreja at outlook.com>> wrote:
>
>     Hi everyone
>
>
>     I am Vivek Kukreja, currently pursuing Masters from India. I'm
>     interested in applying for GSOC this year. I am a skilled C
>     Programmer and have strong understanding of OS concepts. I'm
>     interested in system programming and kernel development.
>
>
>     I am interested to work on topics under Tracing. I am intrigued by
>     the topic. I think this will be a good place to start for me.
>
>
> That's a good area that needs work. Chris Johns is the core person in
> this area. He is in
> Sydney and will likely comment once he is awake. :)
>
> I gave a presentation at the Flight Software Workshop in 2014 on the
> RTEMS ecosystem.
> Tracing was included in that presentation and Chris helped with the
> slides. The slides are
> here:
>
> http://flightsoftware.jhuapl.edu/files/2014/Presentations/Day-2/Session-3/2-RTEMSProjectEcosystem.pdf
>
> There is a video of my presentation if you want to see it. But this
> presentation covers the
> background and status.
>
>     I have gone through Getting Started documentation and setup RTEMS on
>     my machine. I have tried running the provided examples. I tried to
>     run the tracing example, but am facing some problem during
>     compilation (wrapper compilation error). I will continue looking
>     into it, and report if I can't solve the problem on my own.
>
>
> Focus first on doing the GSoC Hello World and adding yourself to the
> list of students.
> Then ask specific questions about those examples.
>
>     Please tell me about the open projects under the Tracing Tool. I
>     request you to please give me some pointers to get started.
>
>
> Chris is the person to make sure the task list is current and in priority.
>
> But I know there is work to be done on generation of method trace wrappers,
> getting trace data off target (tcp/ip, etc), and producing Common Trace
> Format
> and integrating with the Linux Trace Toolkit.
>
> My understanding is that the Linux Trace Toolkit can also dynamically
> interact (bidirectionally) with the target system. I don't know what that
> entails but it is desirable. But it requires having basic integration solved
> first.

I have done some work on the side trying to understand the lttng-live 
trace protocol

The Linux Trace Toolkit has one component that runs on a potentially 
remote system, and captures trace information.

A different system can connect to the system being traced using the 
lttng-live protocol.

See 
https://github.com/lttng/lttng-tools/blob/master/doc/live-reading-protocol.txt

It is a fairly straightforward binary protocol with a handshake, and a 
few interactions before raw CTF is sent back in a request/response manner.


The lttng-live relay daemon is what provides this service.

The code for that can be seen here:

https://github.com/lttng/lttng-tools/tree/master/src/common/relayd


I've done some work to implement the lttng-live protocol as a proof of 
concept as part of some development here.

I'd be interested in providing some help to make RTEMS Trace Capture 
Engine support storing logs in CTF format, and making them available via 
lttng-live.


Hope this helps as well,

Isaac





>
> I hope that helps.
>
> --joel
>
>     Regards,
>
>     Vivek
>
>
>
>     _______________________________________________
>     devel mailing list
>     devel at rtems.org <mailto:devel at rtems.org>
>     http://lists.rtems.org/mailman/listinfo/devel
>
>
>
>
> _______________________________________________
> devel mailing list
> devel at rtems.org
> http://lists.rtems.org/mailman/listinfo/devel
>



More information about the devel mailing list