[LITMUS^RT] ARM compilation patches.

[Manohar Vanga]

Hey Chris!

Thanks for the prompt reply!

So I'm a new student here at MPI-SWS in Bjorn's group so I'm still relatively unfamiliar with Litmus-RT. In the coming weeks I'll get more familiar and I'll be able to talk more intelligently about it! :-)

We recently got an ODROID-X board and I've been playing with it for a few days now. As for what I've done so far, I basically just used the hardkernel 3.6-rc6 tree to boot the ODROID-X into a debootstrap rootfs. I was now considering porting the Litmus code to the new kernel but it doesn't look very trivial (Björn informs me that the scheduler architecture has apparently changed quite a bit).

Thus I am now looking to support the older 3.0 kernel although it looks like you've done most of the work already (and I'm definitely not going to pick up bugs like issues with JRCU at this point. Maybe in a few weeks when I'm more familiar with litmus...).

Regarding the enabling of performance counters, it looks like a patch was submitted for enabling and using them on ARMv7 processors (both A8 and A9). I didn't check exactly when this was introduced: (http://lxr.free-electrons.com/source/arch/arm/kernel/perf_event_v7.c) I think you could just try back porting that file into the tree.

I suppose the 3.0 kernel with the ARM patches are in the arm-fixes branch? I'll give it a try and let you know my results!

Thanks and best regards!
/manohar

On Oct 1, 2012, at 5:42 PM, Christopher Kenna wrote:

> On Mon, Oct 1, 2012 at 11:36 AM, Christopher Kenna <cjk at cs.unc.edu> wrote:
>> 
>> JRCU seems
>> particularly bad, as it requires a JRCU daemon to track context
>> switches, which I expect LITMUS^RT would starve.
> 
> Just to be clear, I meant, "if we were to use JRCU", but I disabled it
> in favor of the CONFIG_TREE_PREEMPT_RCU like we have been using for a
> while now. I'd be interested in having someone take a look at a
> configuration and/or code to see if there is something I've missed,
> though.
> 
> _______________________________________________
> litmus-dev mailing list
> litmus-dev at lists.litmus-rt.org
> https://lists.litmus-rt.org/listinfo/litmus-dev