2008-06-30 13:45:15

by Dmitry Baryshkov

[permalink] [raw]
Subject: Re: linux-next: Tree for June 27 - build fails at tc6393xb_attach_irq() with randconfig build

Kamalesh Babulal wrote:

> Hi Stephen,
>
> randconfig build fails on next-20080627 kernelk with CONFIG_MFD_TC6393XB
> enabled
>
> CC drivers/mfd/tc6393xb.o
> drivers/mfd/tc6393xb.c: In function ‘tc6393xb_attach_irq’:
> drivers/mfd/tc6393xb.c:269: error: implicit declaration of function
> ‘set_irq_flags’ drivers/mfd/tc6393xb.c:269: error: ‘IRQF_VALID’
> undeclared (first use in this function) drivers/mfd/tc6393xb.c:269:
> error: (Each undeclared identifier is reported only once
> drivers/mfd/tc6393xb.c:269: error: for each function it appears in.)
> drivers/mfd/tc6393xb.c:269: error: ‘IRQF_PROBE’ undeclared (first use in
> this function) drivers/mfd/tc6393xb.c:272: error: ‘IRQT_FALLING’
> undeclared (first use in this function) make[2]: ***
> [drivers/mfd/tc6393xb.o] Error 1 make[1]: *** [drivers/mfd] Error 2
> make: *** [drivers] Error 2


IRQT_FALLING will be replaced with IRQ_TYPE_FALLING_EDGE.
However I've a question. set_irq_flags is arm-only thing to change the
status IRQ_NOREQUEST, IRQ_NOPROBE etc. Most other arches (except ppc)
don't even bother touching IRQ_NOREQUEST. Should I change tc6393xb to
depend on ARM or make call to set_irq_flags ARM-dependant?


--
With best wishes
Dmitry