On Wednesday 27 November 2013 15:50:18 Lars-Peter Clausen wrote: > On 11/27/2013 01:14 PM, Hans Verkuil wrote: > [...] > > >>> For our systems the adv7604 interrupts is not always hooked up to a gpio > >>> irq, instead a register has to be read to figure out which device > >>> actually produced the irq. > >> > >> Where is that register located ? Shouldn't it be modeled as an interrupt > >> controller ? > > > > It's a PCIe interrupt whose handler needs to read several FPGA registers > > in order to figure out which interrupt was actually triggered. I don't > > know enough about interrupt controller to understand whether it can be > > modeled as a 'standard' interrupt. > > This sounds as if it should be implemented as a irq_chip driver. There are a > couple of examples in drivers/irqchip/ Exactly, that was my point. A piece of hardware that takes several interrupt inputs, includes mask and flag registers and generate a single interrupt towards the system is an interrupt controller and should have be handled by the Linux irqchip infrastructure. -- Regards, Laurent Pinchart -- To unsubscribe from this list: send the line "unsubscribe linux-media" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
