On Tue, 4 Mar 2014, Alexander Shiyan wrote:
> +сс Arnd Bergmann
> 
> Вторник,  4 марта 2014, 16:46 +01:00 от Thomas Gleixner <tglx@linutronix.de>:
> > On Tue, 4 Mar 2014, Alexander Shiyan wrote:
> > 
> > > Вторник,  4 марта 2014, 12:05 +01:00 от Thomas Gleixner <tglx@linutronix.de>:
> > > > On Sat, 1 Mar 2014, Alexander Shiyan wrote:
> > > > 
> > > > > This will allow to use dummy IRQ handler no_action() from
> > > > > drivers compiled as module. For example, dummy handler is could
> > > > > be used for drivers that use ARM FIQ interrupts.
> > > > 
> > > > And why exactly requires a driver which uses ARM FIQ interrupts the
> > > > no_action implementation?
> > >
> >  
> > > FIQ Interrupt handler is not used in this function. In FIQ case this
> > > is just a dummy declaration. Real handler is assigned by using the
> > > set_fiq_handler().
> > 
> > Why do you need a dummy declaration at all?
> > 
> > set_fiq_handler() is completely detached from the generic interrupt
> > subsystem.
> 
> Some limitations of hardware, such as bit interleaving for normal and
> FIQ interrupts for mask/status registers, led to the implement single
> driver for interrupt handling.
> As a result, all interrupts can be described equally, and single mechanism
> is used for the request/free and enable/disable.
> Correct me if I'm wrong. Driver that implements it, resent several times
> and expect comments within 3 months [1], you are the only one person,
> Thomas, specified as maintainer for this subsystem.
> I repeat, if I'm wrong in the implementation, let's fix this.
> 
> [1]: http://marc.info/?l=linux-arm-kernel&m=139132855024699

Lemme find this and review it.

Thanks,

	tglx