Paravirtualization on Xen

[Gedare Bloom]

POK acts like a hypervisor, although you may be correct that its
interfaces are not quite as "clean" as a true VMM would be. Doing some
work with Xen might be worthwhile, in the sense of fleshing out what
the paravirtualization interface could look like, but only if you have
pretty good experience already with Xen--otherwise, I think there is
too high of a learning curve to figure out both Xen and RTEMS. Plus,
I'm not sure how large the user base would be for something like RTEMS
on Xen. Whereas RTEMS on POK has a pretty clear use-case for running
RTEMS in an ARINC-653 partitioned system. If you are set on
paravirtualization, I think it would be better to continue the work
with POK, unless others think this direction is fruitless.

-Gedare

On Tue, Feb 25, 2014 at 9:28 PM, Youren Shen <shenyouren at gmail.com> wrote:
> Hello,everyone:
> I am a 20 years old student in embedded system from University of Electronic
> Science and Technology of China, which is located in the Chengdu.
>
> I'm interesting in RTEMS and paravirtualization. Thanks for Gedare's help,
> I'm reading the mail list and related documents, and trying to understand
> what has been done before me.Then I can decide what's my purpose in GSoC.
>
> However, I can continue the work of  Philipp Eppelt [3].But it seems
> difficult because POK is not a hypervisor. So, could I start my work to run
> RTEMS in Xen?
>
> Here is my proof that I build RTEMS:
> My patch: [1]
> The snapshot: [2]
> I also have transplanted RTEMS to my mini2440 board.
>
> Your respectfully.
> Sched.
> [1].https://raw.github.com/HuaiYuSched/rtems_GSoC/master/my.patch
> [2].https://raw.github.com/HuaiYuSched/rtems_GSoC/master/snapshot1.png
> [3].http://rtems.org/gsoc2013_final_report
>
> _______________________________________________
> rtems-devel mailing list
> rtems-devel at rtems.org
> http://www.rtems.org/mailman/listinfo/rtems-devel
>