[Xenomai] Kernel choice for Xenomai on the Beagle Bone

[Xenomai] Kernel choice for Xenomai on the Beagle Bone

[Jonathan Wallace]

First I just want to say that I've been doing a lot of reading in the
archives of this mailing list and I'm really impressed with the community,
I'm excited to get my project going with this software!

I'm really just trying to get my bearings before I start mucking around and
one question has been lingering during all my research that I've yet to see
a clear answer for. I'ts one that seems to come up every now and then here,
and I hope I'm not beating a dead horse with it.

Essentially I want to know what kernel I should apply the Xenomai patches
against. From what I've read, the naming conventions on the adeos patches
in .../xenomai-2.6.2.1/ksrc/arch/arm/patches/ should indicate the latest
suitable kernel version with which the patch will work. As of yesterday,
that appears to be ipipe-core-3.5.7-arm-3.patch. I have two hesitations
though. The first is the arm-3 suffix, does this imply some sort of
processor restriction e.g. no compatibility with ARMv7 isa or
something?(forgive me if that's a dumb question) And my second hesitation
is that the 3.5.7 kernel is currently EOL. Would it be a wiser choice to
use 3.4.30 which is still supported?

Thanks,
Jon RW

Re: [Xenomai] Kernel choice for Xenomai on the Beagle Bone

[Gilles Chanteperdrix]

On 02/14/2013 01:09 AM, Jonathan Wallace wrote:

> First I just want to say that I've been doing a lot of reading in the
> archives of this mailing list and I'm really impressed with the community,
> I'm excited to get my project going with this software!
> 
> I'm really just trying to get my bearings before I start mucking around and
> one question has been lingering during all my research that I've yet to see
> a clear answer for. I'ts one that seems to come up every now and then here,
> and I hope I'm not beating a dead horse with it.
> 
> Essentially I want to know what kernel I should apply the Xenomai patches
> against. From what I've read, the naming conventions on the adeos patches
> in .../xenomai-2.6.2.1/ksrc/arch/arm/patches/ should indicate the latest
> suitable kernel version with which the patch will work. As of yesterday,
> that appears to be ipipe-core-3.5.7-arm-3.patch. I have two hesitations
> though. The first is the arm-3 suffix, does this imply some sort of
> processor restriction e.g. no compatibility with ARMv7 isa or
> something?


No, it is a version number of the I-pipe patch.


> And my second hesitation
> is that the 3.5.7 kernel is currently EOL. Would it be a wiser choice to
> use 3.4.30 which is still supported?


There is also an I-pipe patch for Linux 3.4, you may use that one if you
prefer.

-- 
                                                                Gilles.

[Xenomai] Kernel choice for Xenomai on the Beagle Bone

[Jonathan Wallace]

On Fri, Feb 15, 2013 at 4:17 AM, Gilles Chanteperdrix <
gilles.chanteperdrix@xenomai.org> wrote:

> On 02/15/2013 12:12 PM, Jonathan Wallace wrote:
>
> > Thanks for the info Gilles, I went with 3.2.21. There's a lot
> > more information on getting that version patched for the Beagle Bone
> > floating around. So I've gotten as far as a successful compile, but the
> > board hangs during boot without any useful information. Here's the
> > console log:
> >
>
> > (...)
> >
>
> > Uncompressing Linux... done, booting the kernel.
> > [    0.230661] tps65910 1-002d: could not be detected
>
>
> See:
>
> http://xenomai.org/index.php/I-pipe-core:ArmPorting#The_kernel_stops_after_the_message_.22Uncompressing_Linux..._done.2C_booting_the_kernel..22
>
> --
>                                                                 Gilles.
>
You rock Gilles, that's exactly what I needed. But I ended up side-stepping
the problem by using a known-working .config (part of git://
github.com/RobertCNelson/linux-dev.git for the record) and merging in the
xenomai options. The true problem was trying to configure the kernel for my
board from scratch. I'm booted up now and I think that's enough for now
(its 4:30am, I'll save getting the latency test to run for tomorrow.)
Thanks again for the help, I'm doing my best to figure this out as I go.

-Jon RW