On Mon, 2014-05-26 at 23:18 -0400, Nicolas Pitre wrote:
On Tue, 27 May 2014, Mike Galbraith wrote:
On Mon, 2014-05-26 at 18:19 -0400, Nicolas Pitre wrote:
@@ -1046,7 +1046,7 @@ static void update_numa_stats(struct numa_stats *ns, int nid) ns->nr_running += rq->nr_running; ns->load += weighted_cpuload(cpu);
ns->power += power_of(cpu);
ns->compute_capacity += power_of(cpu);power_of(cpu) as a capacity input looks odd now..
@@ -1062,9 +1062,10 @@ static void update_numa_stats(struct numa_stats *ns, int nid) if (!cpus) return;
- ns->load = (ns->load * SCHED_POWER_SCALE) / ns->power;
- ns->capacity = DIV_ROUND_CLOSEST(ns->power, SCHED_POWER_SCALE);
- ns->has_capacity = (ns->nr_running < ns->capacity);
- ns->load = (ns->load * SCHED_POWER_SCALE) / ns->compute_capacity;
- ns->task_capacity =
DIV_ROUND_CLOSEST(ns->compute_capacity, SCHED_POWER_SCALE);..as do SCHED_POWER_SCALE, update_cpu_power() etc.
The rest is renamed in a later patch. I wanted to split it into multiple patches to keep those changes manageable.
I don't see a lot of benefit in creating intermediate inconsistencies vs one bulk rename, but it does make smaller patches, so never mind.
-Mike