Messages in this thread | | | Date | Fri, 23 Feb 2018 08:53:17 -0800 | From | Tejun Heo <> | Subject | Re: [PATCH] percpu: add a schedule point in pcpu_balance_workfn() |
| |
On Fri, Feb 23, 2018 at 08:12:42AM -0800, Eric Dumazet wrote: > From: Eric Dumazet <edumazet@google.com> > > 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 <edumazet@google.com>
Applied to percpu/for-4.16-fixes.
Thank, Eric.
-- tejun
| |