On Tue, Mar 12, 2013 at 10:42:30PM -0600, Stephen Warren wrote:
I believe the point Russell is making is not that the idea behind this patch is wrong, but simply that the function where you put the hooks is wrong. The hooks should at least be in clk_enable/clk_disable and not
Indeed, remembering that clk_enable/clk_disable can be called from atomic contexts.
If the hook needs to be non-atomic (iow, it can schedule) then it can't go into clk_enable/clk_disable, and must go into clk_prepare/clk_unprepare, which is the schedulable half of that API.