[Xenomai] RaspberryPi kernel 3.8 issue
gregd72002 at gmail.com
Sat Mar 1 13:27:43 CET 2014
Uploaded a new trace in here with them being reduced to 10 and 9
On 28 February 2014 11:41, Gilles Chanteperdrix <
gilles.chanteperdrix at xenomai.org> wrote:
> On 02/27/2014 09:33 PM, Gregory Dymarek wrote:
> > Not sure if I understand this. This is the first call to
> __report_bad_irq (
> > I added the trace on line 192.
> > AFAIK this is the first call to the function. the trace freeze i added is
> > not conditional.
> I think the issue you observe is that the irqs_unhandled counter goes
> over 99000, meaning that the driver returns IRQ_NONE and does not
> acknowledge the interrupt at device level, so that the interrupt
> triggers again. So, having enough trace points to see what happens
> before this irq storm will be hard. Could you try reducing the 100000
> and 99000 thresholds in kernel/irq/spurious.c, function note_interrupt?
More information about the Xenomai