Re: [PATCH] possible scheduler deadlock in 2.6.16

* Anton Blanchard [off-list ref] wrote:

One way to solve this is to always take runqueues in cpu id order. To 
do this we add a cpu variable to the runqueue and check it in the 
double runqueue locking functions.

Thoughts?

it's fine with me - the overhead to double_rq_lock() is minimal, and 
it's not critical code either.

Acked-by: Ingo Molnar <redacted>

	Ingo