On Wed, 2011-04-06 at 16:29 +0200, Antonio Ospite wrote: > On Wed, 6 Apr 2011 23:11:33 +0900 > Mark Brown <broo...@opensource.wolfsonmicro.com> wrote: > > > On Wed, Apr 06, 2011 at 11:21:19AM +0200, Antonio Ospite wrote: > > > > > + tristate "Generic rfkill regulator driver" > > > + depends on RFKILL || !RFKILL > > > > That looks *odd*. > > Taken from Documentation/rfkill.txt section 3. Kernel API. > I guess I can drop it if we want to be stricter and just require RFKILL > to be enabled. Johannes?
I guess it depends on what you're looking to do. Since all you implement is set_block() you might very well not need to be able to have this if nothing is ever going to invoke set_block(), in which case you can do "depends on RFKILL". The reason for this usually is that a driver, like a wireless driver, should work even if there's no rfkill API available, but it shouldn't need to put #ifdefs into the code itself. johannes
