Re: [PATCH 4/4] sched/topology: the group balance cpu must be a cpu where the group is installed

[Peter Zijlstra]

On Mon, Apr 24, 2017 at 12:19:47PM -0300, Lauro Venancio wrote:
> On 04/24/2017 11:27 AM, Peter Zijlstra wrote:

> >>> Also, would it not make sense to re-order patch 2 to come after this,
> >>> such that we _do_ have the group_mask available and don't have to jump
> >>> through hoops in order to link up the sgc? Afaict we don't actually use
> >>> the sgc until the above (reverse) loop computing the CPU capacities.

> Yes, it has the same result. I duplicated the build_group_mask magic to
> avoid building the complete mask for all instances of a group.
> Currently, the mask is built just once per group.

OK, but it was subtly different (using sched_domain_span(sibling) vs
sched_group_cpus(sibling->groups)). And if there's anything this code
could do with less of its confusion.

I see your point about wasted computation, but this is very slow path
code, nobody should care too much. And I much prefer simpler code.

I'll see if I can come up with some coherent comments as well. This code
certainly can use some.