On Wed, Jul 2, 2008 at 8:31 PM, Alan Cox <alan@lxorguk.ukuu.org.uk> wrote:
quoted text
>> Note that the irq is disabled instead of acked.  So this should not be a
>> problem.  The userspace part then is responsible to ack (first) and
>> unmask the irq.
>
> That still doesn't make sense.
>
> If you call disable_irq in the IRQ handler you will deadlock because it
> waits until the IRQ handler has completed.

Right, the uio_userirq_handler() function should use
disable_irq_nosync() instead of disable_irq(). Otherwise we'll end up
waiting forever for IRQ_INPROGRESS to be cleared in synchronize_irq()
in the case of CONFIG_SMP.

Thanks!

/ magnus
--
unsubscribe notice
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/