From accd4f36a7d11c2d54544007eb65e10604dcf2f5 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Fri, 23 Feb 2018 08:12:42 -0800 Subject: percpu: add a schedule point in pcpu_balance_workfn() When a large BPF percpu map is destroyed, I have seen pcpu_balance_workfn() holding cpu for hundreds of milliseconds. On KASAN config and 112 hyperthreads, average time to destroy a chunk is ~4 ms. [ 2489.841376] destroy chunk 1 in 4148689 ns ... [ 2490.093428] destroy chunk 32 in 4072718 ns Signed-off-by: Eric Dumazet Signed-off-by: Tejun Heo --- mm/percpu.c | 1 + 1 file changed, 1 insertion(+) (limited to 'mm') diff --git a/mm/percpu.c b/mm/percpu.c index fa3f854634a1..36e7b65ba6cf 100644 --- a/mm/percpu.c +++ b/mm/percpu.c @@ -1610,6 +1610,7 @@ static void pcpu_balance_workfn(struct work_struct *work) spin_unlock_irq(&pcpu_lock); } pcpu_destroy_chunk(chunk); + cond_resched(); } /* -- cgit v1.2.3