On 24 February 2014 14:17,  [off-list ref] wrote:

Sorry, not sure I understand what you mean.

I agree, wording in my commit text might be unclear. I'll fix it after we
agree on the code fix. In the MSM case, each CPU has it's own policy.

I'm assuming your original complaint was about my confusing wording. Maybe
that's not what you were pointing out?

In your case each CPU has a separate policy structure as they have separately
controllable clocks. But you also said that CPU0 is setting CPU1's governor to
NULL. I don't see that happening. Each CPU sets its own governor to NULL on
init().