On Tue, Jun 30, 2015 at 07:08:26PM +0200, Marc Le Douarain wrote:
> Le 29/06/2015 21:45, Gilles Chanteperdrix a écrit :
> >On Mon, Jun 29, 2015 at 09:27:11PM +0200, Marc Le Douarain wrote:
> >>Hello everybody,
> >>
> >>tested Xenomai 3.0-rc5 on the little Arietta (with patch
> >>ipipe-core-3.18.12-arm-2.patch),
> >>and even if it seems to work (I-pipe registered and latenty test running),
> >>theses lines seeems quite suspect :
> >>
> >>NR_IRQS:16 nr_irqs:16 16
> >>------------[ cut here ]------------
> >>WARNING: CPU: 0 PID: 0 at kernel/irq/pm.c:48
> >>irq_pm_install_action+0x94/0xd0()
> >>Modules linked in:
> >>CPU: 0 PID: 0 Comm: swapper Not tainted 3.18.12-xeno-mld #1
> >>[<c000dc84>] (unwind_backtrace) from [<c000b9b4>] (show_stack+0x10/0x14)
> >>[<c000b9b4>] (show_stack) from [<c001671c>] (warn_slowpath_common+0x60/0x80)
> >>[<c001671c>] (warn_slowpath_common) from [<c0016754>]
> >>(warn_slowpath_null+0x18/0x1c)
> >>[<c0016754>] (warn_slowpath_null) from [<c003fc28>]
> >>(irq_pm_install_action+0x94/0xd0)
> >>[<c003fc28>] (irq_pm_install_action) from [<c003bcd8>]
> >>(__setup_irq+0x398/0x4f4)
> >>[<c003bcd8>] (__setup_irq) from [<c003bef8>]
> >>(request_threaded_irq+0xc4/0x124)
> >>[<c003bef8>] (request_threaded_irq) from [<c0526050>]
> >>(at91sam926x_pit_common_init+0xcc/0x1a0)
> >>[<c0526050>] (at91sam926x_pit_common_init) from [<c0525f40>]
> >>(clocksource_of_init+0x44/0x88)
> >>[<c0525f40>] (clocksource_of_init) from [<c050dae8>]
> >>(start_kernel+0x240/0x3cc)
> >>[<c050dae8>] (start_kernel) from [<20008040>] (0x20008040)
> >>---[ end trace cb88537fdc8fa200 ]---
> >Does not the exact same kernel not patched with Xenomai and I-pipe
> >with the exact same configuration, only without CONFIG_IPIPE and
> >CONFIG_XENOMAI, have the same problem?
> >
> Thanks for the idea of this simple test to be done,
> and you are absolutely right, same strange messages with the vanilla kernel
> without I-pipe patch !
You are welcome. You should have a look at the corresponding line of
code to know whether this is an important issue. But the reason is
probably that the configuration you test is not the same as the one
the maintainer of the board uses, probably in the power management
area.
--
Gilles.
https://click-hack.org
_______________________________________________
Xenomai mailing list
[email protected]
http://xenomai.org/mailman/listinfo/xenomai
