Re: [Xenomai] issues with debugging enabled

[Jan Kiszka <jan.kiszka@siemens.com>]

On 2013-11-14 21:23, Gilles Chanteperdrix wrote:
> On 11/14/2013 08:36 PM, Jan Kiszka wrote:
>> On 2013-11-13 14:53, Jan Kiszka wrote:
>>> On 2013-11-13 14:41, Gilles Chanteperdrix wrote:
>>>> On 11/13/2013 02:25 PM, Jan Kiszka wrote:
>>>>> On 2013-11-12 21:34, Gilles Chanteperdrix wrote:
>>>>>>
>>>>>> Hi,
>>>>>>
>>>>>> for a change, I ran the xeno-regression-test with a lot of debugging and
>>>>>> options known to having caused problems in the past and found two
>>>>>> issues:
>>>>>>
>>>>>> on x86 SMP, with full dynticks and debugging enabled (preemptible kernel
>>>>>> debugging, mutex, spinlocks, and sleep inside spinlocks), I get the
>>>>>> series of warnings at the end of the mail.
>>>>>
>>>>> We are currently facing crashes inside __switch_to, FPU related
>>>>> (non-existing save area).
>>>>
>>>> Probably the issue solved by this patch:
>>>> http://git.xenomai.org/?p=ipipe-gch.git;a=commit;h=9b3320eef67e3f118f96065c8c2584b715710d94
>>>
>>> Thought I had everything, but I lost track, and it's missing here. Will
>>> merge and give it a try.
>>
>> OK, this fixes the crash in __switch_to, triggerable by terminating
>> xeno-test while running clocktest. But we still face a crash when
>> terminating switchtest early. Are you aware of this?
> 
> No, I was talking about Xenomai 2.6.3, not forge.

Fix for forge with this commit:

http://git.xenomai.org/?p=xenomai-jki.git;a=commitdiff;h=3e6d8ff9a99262e78655329dc043aacc607eb158

I suppose that error scenario does not occur with 2.6 due to the
different kernel thread model. Still, backporting may be worth considering.

Jan

-- 
Siemens AG, Corporate Technology, CT RTC ITP SES-DE
Corporate Competence Center Embedded Linux