Document that IRQ_NONE should be returned when IRQ not actually handled
authorDavid Woodhouse <dwmw2@infradead.org>
Wed, 28 Oct 2015 07:14:31 +0000 (16:14 +0900)
committerThomas Gleixner <tglx@linutronix.de>
Fri, 30 Oct 2015 09:13:26 +0000 (10:13 +0100)
commitd9e4ad5badf4ccbfddee208c898fb8fd0c8836b1
treed3e50f3afbdb57b8c3eb84d03188cd357c1bec14
parent3b93baf56dafa2d27e4fc227990dcd3ffeb10510
Document that IRQ_NONE should be returned when IRQ not actually handled

Our IRQ storm detection works when an interrupt handler returns
IRQ_NONE for thousands of consecutive interrupts in a second. It
doesn't hurt to occasionally return IRQ_NONE when the interrupt is
actually genuine.

Drivers should only be returning IRQ_HANDLED if they have actually
*done* something to stop an interrupt from happening — it doesn't just
mean "this really *was* my device".

Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Cc: davem@davemloft.net
Link: http://lkml.kernel.org/r/1446016471.3405.201.camel@infradead.org
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
include/linux/irqreturn.h