Re: [Xenomai] Porting xenomai 2.6.3 and ipipe to 3.12 kernel

[Gilles Chanteperdrix <gilles.chanteperdrix@xenomai.org>]

On 03/31/2014 11:25 PM, Lennart Sorensen wrote:
> On Mon, Mar 31, 2014 at 11:06:51PM +0200, Gilles Chanteperdrix wrote:
>> On 03/31/2014 10:57 PM, Lennart Sorensen wrote:
>>> On Mon, Mar 31, 2014 at 04:53:39PM -0400, Lennart Sorensen wrote:
>>>> [341548.649928] =================================
>>>> [341548.649934] [ INFO: inconsistent lock state ]
>>>> [341548.649943] 3.12-1-dra7xx #1 Not tainted
>>>> [341548.649950] ---------------------------------
>>>> [341548.649957] inconsistent {IN-HARDIRQ-W} -> {HARDIRQ-ON-W} usage.
>>
>> This looks fishy. It means that the root domain is not stalled while
>> interrupt handler is called. This would need to be checked.
>>
>>> No change in behaviour.  switchtest works, clocktest works, most other
>>> things have illegal instruction.  Just waiting for the library and
>>> runtime to recompile against 3.12's headers in case it matters.
>>> After that it will be debug time.
>>>
>>
>> Ok. You seem to have many debug options enabled. Please find attached
>> the config I have used during my tests on omap5.
> 
> I will try and look through it and see which options to use.

Please post your .config, I will have a look. Please disable unlocked
context switch as well. Also please note that the context switch has
been reworked in 3.11, so you need Xenomai 2.6.x from the git repository.

>
>  This is
> an omap57xx, not a 54xx, also known as the dra7xx.
> 
a.k.a Vayu ?

I made quite a few other changes for omap5, I never merged them in the
I-pipe branch since the port was not finished. You may want to have a
look at them if you are continuing the work on omap5.

I have pushed the omap5-ipipe branch in ipipe-gch with this work in
progress.

-- 
                                                                Gilles.