Hi!

Another solution occurs to me: clk_enable() which returns -EAGAIN if
it's called in an atomic context but needs to sleep, and let the driver
deal with the clock not being enabled when it wants it.  The down side
is it requires the driver to have additional code to sort out such a
problem.

I do not think that "atomic context" is possible to reliably detect
inside a function like that.
									Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html