Re: [RFC v3 1/6] Track the active utilisation

[Peter Zijlstra <peterz@infradead.org>]

On Tue, Oct 25, 2016 at 09:58:11AM -0400, Steven Rostedt wrote:

> I agree with Daniel, especially since I don't usually trust the
> compiler. And the added variable is more of a distraction as it doesn't
> seem to have any real purpose.

I don't think there's anything here to trust the compiler on. Either
it inlines or it doesn't, it should generate 'correct' code either way.

If it doesn't inline, its a dumb compiler and it will make these dumb
decisions throughout the tree and your kernel will be slow, not my
problem really ;-)

That said, I agree that the single line thing is actually easier to
read.

That said; there's something to be said for:

	u64 running_bw;

static void add_running_bw(struct sched_dl_entity *dl_se, struct dl_rq *dl_rq)
{
	u64 old = dl_rq->running_bw;

	dl_rq->running_bw += dl_se->dl_bw;
	SCHED_WARN_ON(dl_rq->running_bw < old); /* overflow */
}

static void sub_running_bw(struct sched_dl_entity *dl_se, struct dl_rq *dl_rq)
{
	u64 old = dl_rq->running_bw;

	dl_rq->running_bw -= dl_se->dl_bw;
	SCHED_WARN_ON(dl_rq->running_bw > old); /* underflow */
}