Re: [PATCH?] Livelock in pick_next_task_fair() / idle_balance()

[Mike Galbraith]

On Wed, 2015-07-01 at 16:55 +0200, Rabin Vincent wrote:

> So, we will not hit the "if (env->src_rq->load.weight <=
> env->dst_rq->load.weight + d_load)" condition to break out of the loop until we
> actualy move all tasks.  So the patch will not have any effect on this case.
> Or am I missing something?

Probably not.  I did have it breaking if dst_rq would pass
src_rq->nr_running, which would certainly stop it, but thought I try to
let it watch weights.

Either way, task_h_load(p) returning 0 is not very wonderful.

	-Mike

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/