Hi, On Sun, Dec 15, 2013 at 11:24:12PM +0100, Andreas Larsson wrote: > >>>Also, the way you're using thread IRQs is quite wrong. I can't let that > >>>pass and get merged upstream, sorry. > >> > >>What is quite wrong? What is it that I need to fix? > > > >Ideally the hardirq handler should be usually to actually check if > >$this_device generated the IRQ, that should involve reading a IRQSTATUS > >register of some sort. > > > >Sure, check that IRQs are actually enabled, but you also need to read > >STATUS register before waking the thread up. > > I agree that that would be preferable. Unfortunately, the hardware > lacks status register bits that indicates whether interrupts has been > generated or not. That is why the only check is if interrupts are > disabled, because that is the only sure way to know that the softirq > handler does not need to go through everything. so it seems, fair enough then. Keep it the way it is. cheers -- balbi
Attachment:
signature.asc
Description: Digital signature
