Cpu which is put into quiescent mode, would set its sd
member as NULL, and want others not disturb its task running.
But current scheduler would not checking whether that cpu is
setting in such mode, and still insist the quiescent
cpu to response the nohz load balance.
Fix it by preventing such cpu set nohz.idle_cpus_mask in the
first place.
Signed-off-by: Lei Wen <lei...@marvell.com>
Cc: Peter Zijlstra <pet...@infradead.org>
Cc: Mike Galbraith <bitbuc...@online.de>
---
Much thanks to Mike Pointing out the root span would be merged when the
last cpu becomes isolated from the crash result checking!
kernel/sched/fair.c | 8 ++++++++
1 file changed, 8 insertions(+)
diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index 235cfa7..af30b6a 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -6883,6 +6883,14 @@ void nohz_balance_enter_idle(int cpu)
if (!cpu_active(cpu))
return;
+ /*
+ * If this cpu is isolated, its rq's sd member would becomes NULL.
+ * Base on this observation, we could exclude this cpu from nohz
+ * idle balance, so that it would not be disturbed.
+ */
+ if (!this_rq()->sd)
+ return;
+
if (test_bit(NOHZ_TICK_STOPPED, nohz_flags(cpu)))
return;
--
1.8.3.2
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
