Re: [PATCH 16/19] sched/numa: Detect if node actively handling migration
From: Srikar Dronamraju
Date: Wed Jun 06 2018 - 08:56:08 EST
> >
> > I thought about this. Lets say we evaluated that destination node can
> > allow movement. While we iterate through the list of cpus trying to
> > find
> > the best cpu node, we find a idle cpu towards the end of the list.
> > However if another task as already raced with us to move a task to
> > this
> > node, then we should bail out. Keeping the check in task_numa_compare
> > will allow us to do this.
>
> Your check is called once for every invocation
> of task_numa_compare. It does not matter whether
> it is inside or outside, except on the outside
> the variable manipulation will be easier to read.
>
Okay I mistook your comment; Basically you want the check to be moved
within the for-loop in task_numa_find_cpu.
I will do the needful.
> >
> > While we can't complete avoid this, the second check will try to make
> > sure we don't hop on/hop off just for small incremental numa
> > improvement.
>
> However, all those racing tasks start searching
> the CPUs on a node from the same start position.
>
> That means they may all get stuck on the same
> task/cpu A, and not select the better task/cpu B.
>
> What am I missing?
All tasks will not be stuck at task/cpu A.
"[PATCH 10/19] sched/numa: Stop multiple tasks from moving to the
cpu..." the first task to set cpu A as swap target will ensure
subsequent tasks wont be allowed to set cpu A as target for swap till it
finds a better task/cpu. Because of this there a very very small chance
of a second task unable to find a task to swap.