On Thu, Mar 28, 2013 at 02:42:50AM +0200, Andrei Epure wrote: > Patch found using coccinelle. > > Signed-off-by: Andrei Epure <[email protected]>
Some of these may be false positives especially for PMICs - often the interrupts are guaranteed to be generated from the interrupt controller on the device which runs in thread context. More generally though why not fix this in request_threaded_irq() itself, if there's no hard IRQ handler then add the flag transparently? It'd save boilerplate. Indeed I could've sworn someone did that...
signature.asc
Description: Digital signature
