On Tue, 5 May 2020 10:28:50 -0700 "Paul E. McKenney" <paul...@kernel.org> wrote:
> On Tue, May 05, 2020 at 09:37:42AM -0700, Eric Dumazet wrote: > > > > > > On 5/5/20 9:31 AM, Eric Dumazet wrote: > > > > > > > > > On 5/5/20 9:25 AM, Eric Dumazet wrote: > > >> > > >> > > >> On 5/5/20 9:13 AM, SeongJae Park wrote: > > >>> On Tue, 5 May 2020 09:00:44 -0700 Eric Dumazet <eduma...@google.com> > > >>> wrote: > > >>> > > >>>> On Tue, May 5, 2020 at 8:47 AM SeongJae Park <sjp...@amazon.com> wrote: > > >>>>> > > >>>>> On Tue, 5 May 2020 08:20:50 -0700 Eric Dumazet > > >>>>> <eric.duma...@gmail.com> wrote: > > >>>>> > > >>>>>> > > >>>>>> > > >>>>>> On 5/5/20 8:07 AM, SeongJae Park wrote: > > >>>>>>> On Tue, 5 May 2020 07:53:39 -0700 Eric Dumazet > > >>>>>>> <eduma...@google.com> wrote: > > >>>>>>> > > >>>>>> > > >>>>>>>> Why do we have 10,000,000 objects around ? Could this be because of > > >>>>>>>> some RCU problem ? > > >>>>>>> > > >>>>>>> Mainly because of a long RCU grace period, as you guess. I have no > > >>>>>>> idea how > > >>>>>>> the grace period became so long in this case. > > >>>>>>> > > >>>>>>> As my test machine was a virtual machine instance, I guess RCU > > >>>>>>> readers > > >>>>>>> preemption[1] like problem might affected this. > > >>>>>>> > > >>>>>>> [1] > > >>>>>>> https://www.usenix.org/system/files/conference/atc17/atc17-prasad.pdf > > >>>>>>> > > >>>>>>>> > > >>>>>>>> Once Al patches reverted, do you have 10,000,000 sock_alloc around > > >>>>>>>> ? > > >>>>>>> > > >>>>>>> Yes, both the old kernel that prior to Al's patches and the recent > > >>>>>>> kernel > > >>>>>>> reverting the Al's patches didn't reproduce the problem. > > >>>>>>> > > >>>>>> > > >>>>>> I repeat my question : Do you have 10,000,000 (smaller) objects kept > > >>>>>> in slab caches ? > > >>>>>> > > >>>>>> TCP sockets use the (very complex, error prone) > > >>>>>> SLAB_TYPESAFE_BY_RCU, but not the struct socket_wq > > >>>>>> object that was allocated in sock_alloc_inode() before Al patches. > > >>>>>> > > >>>>>> These objects should be visible in kmalloc-64 kmem cache. > > >>>>> > > >>>>> Not exactly the 10,000,000, as it is only the possible highest > > >>>>> number, but I > > >>>>> was able to observe clear exponential increase of the number of the > > >>>>> objects > > >>>>> using slabtop. Before the start of the problematic workload, the > > >>>>> number of > > >>>>> objects of 'kmalloc-64' was 5760, but I was able to observe the > > >>>>> number increase > > >>>>> to 1,136,576. > > >>>>> > > >>>>> OBJS ACTIVE USE OBJ SIZE SLABS OBJ/SLAB CACHE SIZE NAME > > >>>>> before: 5760 5088 88% 0.06K 90 64 360K > > >>>>> kmalloc-64 > > >>>>> after: 1136576 1136576 100% 0.06K 17759 64 71036K > > >>>>> kmalloc-64 > > >>>>> > > >>>> > > >>>> Great, thanks. > > >>>> > > >>>> How recent is the kernel you are running for your experiment ? > > >>> > > >>> It's based on 5.4.35. > > >>> > > >>>> > > >>>> Let's make sure the bug is not in RCU. > > >>> > > >>> One thing I can currently say is that the grace period passes at last. > > >>> I > > >>> modified the benchmark to repeat not 10,000 times but only 5,000 times > > >>> to run > > >>> the test without OOM but easily observable memory pressure. As soon as > > >>> the > > >>> benchmark finishes, the memory were freed. > > >>> > > >>> If you need more tests, please let me know. > > >>> > > >> > > >> I would ask Paul opinion on this issue, because we have many objects > > >> being freed after RCU grace periods. > > >> > > >> If RCU subsystem can not keep-up, I guess other workloads will also > > >> suffer. > > >> > > >> Sure, we can revert patches there and there trying to work around the > > >> issue, > > >> but for objects allocated from process context, we should not have these > > >> problems. > > >> > > > > > > I wonder if simply adjusting rcu_divisor to 6 or 5 would help > > > > > > diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c > > > index > > > d9a49cd6065a20936edbda1b334136ab597cde52..fde833bac0f9f81e8536211b4dad6e7575c1219a > > > 100644 > > > --- a/kernel/rcu/tree.c > > > +++ b/kernel/rcu/tree.c > > > @@ -427,7 +427,7 @@ module_param(qovld, long, 0444); > > > static ulong jiffies_till_first_fqs = ULONG_MAX; > > > static ulong jiffies_till_next_fqs = ULONG_MAX; > > > static bool rcu_kick_kthreads; > > > -static int rcu_divisor = 7; > > > +static int rcu_divisor = 6; > > > module_param(rcu_divisor, int, 0644); > > > > > > /* Force an exit from rcu_do_batch() after 3 milliseconds. */ > > > > > > > To be clear, you can adjust the value without building a new kernel. > > > > echo 6 >/sys/module/rcutree/parameters/rcu_divisor > > Worth a try! If that helps significantly, I have some ideas for updating > that heuristic, such as checking for sudden increases in the number of > pending callbacks. > > But I would really also like to know whether there are long readers and > whether v5.6 fares better. I will share the results as soon as possible :) Thanks, SeongJae Park > > Thanx, Paul
