[Xenomai] Spurious relax test?

[Xenomai] Spurious relax test?

[Yogi A. Patel]

Hi -

I’ve implemented the code provided on the Xenomai website for finding spurious relaxes in Xenomai 2 - however I’m a bit skeptical about whether or not I’ve implemented it correctly since I don’t ever see any output to the terminal.

Is there any way to “induce” a relax to test the implementation and know that it is in fact catching any drops to secondary mode?

Yogi

Re: [Xenomai] Spurious relax test?

[Gilles Chanteperdrix]

On Fri, Dec 26, 2014 at 12:46:58AM -0500, Yogi A. Patel wrote:
> Hi -
> 
> I’ve implemented the code provided on the Xenomai website for finding spurious relaxes in Xenomai 2 - however I’m a bit skeptical about whether or not I’ve implemented it correctly since I don’t ever see any output to the terminal.
> 
> Is there any way to “induce” a relax to test the implementation and know that it is in fact catching any drops to secondary mode?

See examples/native/sigdebug.c

-- 
					    Gilles.

Re: [Xenomai] Spurious relax test?

[Yogi A. Patel]

> On Dec 26, 2014, at 10:58, Gilles Chanteperdrix <gilles.chanteperdrix@xenomai.org> wrote:
> 
> On Fri, Dec 26, 2014 at 12:46:58AM -0500, Yogi A. Patel wrote:
>> Hi -
>> 
>> I’ve implemented the code provided on the Xenomai website for finding spurious relaxes in Xenomai 2 - however I’m a bit skeptical about whether or not I’ve implemented it correctly since I don’t ever see any output to the terminal.
>> 
>> Is there any way to “induce” a relax to test the implementation and know that it is in fact catching any drops to secondary mode?
> 
> See examples/native/sigdebug.c

Thanks, Gilles. I implemented the warning method using the example as a guide.

It appears to be working - however I’m a bit puzzled by the output here.

The only error I am getting is: SIGDEBUG received, reason 2: invoked syscall
rtxi(_Z17rt_switch_warningiP9siginfo_tPv+0x3d)[0x43883d]
/lib/x86_64-linux-gnu/libpthread.so.0(+0x10340)[0x7f71e54a9340]
/lib/x86_64-linux-gnu/libpthread.so.0(sem_post+0x2d)[0x7f71e54a790d]

The application I am testing is a Qt based application - I have a feeling these are all coming from the posting of events to the Qt event stack but can’t tell for sure.

Is there a way/how can I get more information from the trace?

Thanks!

> 
> -- 
> 					    Gilles.

Yogi A. Patel
@yapatel525 <https://twitter.com/yapatel525>

Re: [Xenomai] Spurious relax test?

[Gilles Chanteperdrix]

On Fri, Dec 26, 2014 at 11:55:57AM -0500, Yogi A. Patel wrote:
> 
> > On Dec 26, 2014, at 10:58, Gilles Chanteperdrix <gilles.chanteperdrix@xenomai.org> wrote:
> > 
> > On Fri, Dec 26, 2014 at 12:46:58AM -0500, Yogi A. Patel wrote:
> >> Hi -
> >> 
> >> I’ve implemented the code provided on the Xenomai website for finding spurious relaxes in Xenomai 2 - however I’m a bit skeptical about whether or not I’ve implemented it correctly since I don’t ever see any output to the terminal.
> >> 
> >> Is there any way to “induce” a relax to test the implementation and know that it is in fact catching any drops to secondary mode?
> > 
> > See examples/native/sigdebug.c
> 
> Thanks, Gilles. I implemented the warning method using the example as a guide.
> 
> It appears to be working - however I’m a bit puzzled by the output here.
> 
> The only error I am getting is: SIGDEBUG received, reason 2: invoked syscall
> rtxi(_Z17rt_switch_warningiP9siginfo_tPv+0x3d)[0x43883d]
> /lib/x86_64-linux-gnu/libpthread.so.0(+0x10340)[0x7f71e54a9340]
> /lib/x86_64-linux-gnu/libpthread.so.0(sem_post+0x2d)[0x7f71e54a790d]
> 
> The application I am testing is a Qt based application - I have a feeling these are all coming from the posting of events to the Qt event stack but can’t tell for sure.
> 
> Is there a way/how can I get more information from the trace?

gdb

-- 
					    Gilles.

Re: [Xenomai] Spurious relax test?

[Gilles Chanteperdrix]

On Fri, Dec 26, 2014 at 12:19:02PM -0500, Yogi A. Patel wrote:
> 
> > On Dec 26, 2014, at 12:15, Gilles Chanteperdrix <gilles.chanteperdrix@xenomai.org> wrote:
> > 
> > On Fri, Dec 26, 2014 at 11:55:57AM -0500, Yogi A. Patel wrote:
> >> 
> >>> On Dec 26, 2014, at 10:58, Gilles Chanteperdrix <gilles.chanteperdrix@xenomai.org> wrote:
> >>> 
> >>> On Fri, Dec 26, 2014 at 12:46:58AM -0500, Yogi A. Patel wrote:
> >>>> Hi -
> >>>> 
> >>>> I’ve implemented the code provided on the Xenomai website for finding spurious relaxes in Xenomai 2 - however I’m a bit skeptical about whether or not I’ve implemented it correctly since I don’t ever see any output to the terminal.
> >>>> 
> >>>> Is there any way to “induce” a relax to test the implementation and know that it is in fact catching any drops to secondary mode?
> >>> 
> >>> See examples/native/sigdebug.c
> >> 
> >> Thanks, Gilles. I implemented the warning method using the example as a guide.
> >> 
> >> It appears to be working - however I’m a bit puzzled by the output here.
> >> 
> >> The only error I am getting is: SIGDEBUG received, reason 2: invoked syscall
> >> rtxi(_Z17rt_switch_warningiP9siginfo_tPv+0x3d)[0x43883d]
> >> /lib/x86_64-linux-gnu/libpthread.so.0(+0x10340)[0x7f71e54a9340]
> >> /lib/x86_64-linux-gnu/libpthread.so.0(sem_post+0x2d)[0x7f71e54a790d]
> >> 
> >> The application I am testing is a Qt based application - I have a feeling these are all coming from the posting of events to the Qt event stack but can’t tell for sure.
> >> 
> >> Is there a way/how can I get more information from the trace?
> > 
> > gdb
> 
> How exactly does gdb interface with the output above?..

See:
https://sourceware.org/gdb/current/onlinedocs/gdb/Signals.html#Signals

-- 
					    Gilles.