+ //This test is fun, though, as it may happen that
+ //at time t : CPU 0 : process_free
+ //at time t+150ns : CPU 1 : schedule out
+ //Clearly due to time imprecision, we disable it. (Mathieu)
+ //If this weird case happen, we have no choice but to put the
+ //Currently running process on the cpu to 0.
+ //I re-enable it following time precision fixes. (Mathieu)
+ //Well, in the case where an process is freed by a process on another CPU
+ //and still scheduled, it happens that this is the schedchange that will
+ //drop the last reference count. Do not free it here!