On Mon, 2017-01-30 at 08:56 -0800, Eric Anholt wrote: > Michael Zoran <mzoran@xxxxxxxxxxxx> writes: > > > This driver makes no sense outside of ARM or ARM64. > > Add an explicit build dependency on: > > (ARM || ARM64 || COMPILE_TEST) > > > > Also set the default build to n > > > > Signed-off-by: Michael Zoran <mzoran@xxxxxxxxxxxx> > > --- > > drivers/staging/bcm2835-audio/Kconfig | 2 ++ > > 1 file changed, 2 insertions(+) > > > > diff --git a/drivers/staging/bcm2835-audio/Kconfig > > b/drivers/staging/bcm2835-audio/Kconfig > > index 32a2ff9ef9b2..840faa21f665 100644 > > --- a/drivers/staging/bcm2835-audio/Kconfig > > +++ b/drivers/staging/bcm2835-audio/Kconfig > > @@ -1,6 +1,8 @@ > > config SND_BCM2835 > > tristate "BCM2835 ALSA driver" > > depends on ARCH_BCM2835 && BCM2835_VCHIQ && SND > > + depends on (ARM || ARM64 || COMPILE_TEST) > > + default n > > select SND_PCM > > help > > Say Y or M if you want to support BCM2835 Alsa pcm card > > driver > > We've already got a DEPENDS on ARCH_BCM2835, how would that be > present > without one of those three being set? > I can't think of a reason why ARCH_BCM2835 would be set without ARM || ARM64. I'm just trying to be cautious. I still would like the "default n" added just as a precaution. I thought the default was n unless provided but I see other drivers doing that and build config has some complex options for upgrading old configurations that I don't completely understand. _______________________________________________ devel mailing list devel@xxxxxxxxxxxxxxxxxxxxxx http://driverdev.linuxdriverproject.org/mailman/listinfo/driverdev-devel
