Replace the HK_TYPE_MISC isolation mask with HKR_TYPE_SCHED in
find_new_ilb(). This ultimately resolves [1], i.e. selecting
a CPU for the nohz idle balance which cannot pull tasks
through:
sched_balance_trigger()
\-nohz_balancer_kick()
\-kick_ilb()
\-find_new_ilb()
\-smp_call_function_single_async()
\-nohz_csd_cpu()
\-raise_softirq_irqoff()
sched_balance_softirq()
\-_nohz_idle_balance()
\-sched_balance_domains()
Indeed, a following patch will add/remove CPUs with a NULL sched
domain to the HKR_TYPE_SCHED cpumask.
[1] https://lore.kernel.org/all/[email protected]/
Reported-by: Zhang Rui <[email protected]>
Signed-off-by: Pierre Gondois <[email protected]>
---
kernel/sched/fair.c | 5 +----
1 file changed, 1 insertion(+), 4 deletions(-)
diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index e3d707243ca7..0665f5eb4703 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -11842,16 +11842,13 @@ static inline int on_null_domain(struct rq *rq)
* - When one of the busy CPUs notices that there may be an idle rebalancing
* needed, they will kick the idle load balancer, which then does idle
* load balancing for all the idle CPUs.
- *
- * - HK_TYPE_MISC CPUs are used for this task, because HK_TYPE_SCHED is not set
- * anywhere yet.
*/
static inline int find_new_ilb(void)
{
const struct cpumask *hk_mask;
int ilb_cpu;
- hk_mask = housekeeping_cpumask(HK_TYPE_MISC);
+ hk_mask = housekeeping_runtime_cpumask(HKR_TYPE_SCHED);
for_each_cpu_and(ilb_cpu, nohz.idle_cpus_mask, hk_mask) {
--
2.25.1