Re: [PATCH] mm, vmscan: add cond_resched into shrink_node_memcg

2016-12-09 Thread Donald Buczek
On 12/02/16 10:58, Michal Hocko wrote: From: Michal Hocko Boris Zhmurov has reported RCU stalls during the kswapd reclaim: 17511.573645] INFO: rcu_sched detected stalls on CPUs/tasks: [17511.573699] 23-...: (22 ticks this GP) idle=92f/140/0

Re: [PATCH] mm, vmscan: add cond_resched into shrink_node_memcg

2016-12-09 Thread Donald Buczek
On 12/02/16 10:58, Michal Hocko wrote: From: Michal Hocko Boris Zhmurov has reported RCU stalls during the kswapd reclaim: 17511.573645] INFO: rcu_sched detected stalls on CPUs/tasks: [17511.573699] 23-...: (22 ticks this GP) idle=92f/140/0 softirq=2638404/2638404 fqs=23

Re: [PATCH] mm, vmscan: add cond_resched into shrink_node_memcg

2016-12-05 Thread Paul E. McKenney
On Mon, Dec 05, 2016 at 01:49:55PM +0100, Michal Hocko wrote: > [CC Paul - sorry I've tried to save you from more emails...] > > On Mon 05-12-16 23:44:27, Balbir Singh wrote: > > > > > > Hi, > > > there were multiple reportes of the similar RCU stalls. Only Boris has > > > confirmed that this

Re: [PATCH] mm, vmscan: add cond_resched into shrink_node_memcg

2016-12-05 Thread Paul E. McKenney
On Mon, Dec 05, 2016 at 01:49:55PM +0100, Michal Hocko wrote: > [CC Paul - sorry I've tried to save you from more emails...] > > On Mon 05-12-16 23:44:27, Balbir Singh wrote: > > > > > > Hi, > > > there were multiple reportes of the similar RCU stalls. Only Boris has > > > confirmed that this

Re: [PATCH] mm, vmscan: add cond_resched into shrink_node_memcg

2016-12-05 Thread Michal Hocko
[CC Paul - sorry I've tried to save you from more emails...] On Mon 05-12-16 23:44:27, Balbir Singh wrote: > > > > Hi, > > there were multiple reportes of the similar RCU stalls. Only Boris has > > confirmed that this patch helps in his workload. Others might see a > > slightly different issue

Re: [PATCH] mm, vmscan: add cond_resched into shrink_node_memcg

2016-12-05 Thread Michal Hocko
[CC Paul - sorry I've tried to save you from more emails...] On Mon 05-12-16 23:44:27, Balbir Singh wrote: > > > > Hi, > > there were multiple reportes of the similar RCU stalls. Only Boris has > > confirmed that this patch helps in his workload. Others might see a > > slightly different issue

Re: [PATCH] mm, vmscan: add cond_resched into shrink_node_memcg

2016-12-05 Thread Balbir Singh
> > Hi, > there were multiple reportes of the similar RCU stalls. Only Boris has > confirmed that this patch helps in his workload. Others might see a > slightly different issue and that should be investigated if it is the > case. As pointed out by Paul [1] cond_resched might be not sufficient >

Re: [PATCH] mm, vmscan: add cond_resched into shrink_node_memcg

2016-12-05 Thread Balbir Singh
> > Hi, > there were multiple reportes of the similar RCU stalls. Only Boris has > confirmed that this patch helps in his workload. Others might see a > slightly different issue and that should be investigated if it is the > case. As pointed out by Paul [1] cond_resched might be not sufficient >

[PATCH] mm, vmscan: add cond_resched into shrink_node_memcg

2016-12-02 Thread Michal Hocko
From: Michal Hocko Boris Zhmurov has reported RCU stalls during the kswapd reclaim: 17511.573645] INFO: rcu_sched detected stalls on CPUs/tasks: [17511.573699] 23-...: (22 ticks this GP) idle=92f/140/0 softirq=2638404/2638404 fqs=23 [17511.573740] (detected by 4,

[PATCH] mm, vmscan: add cond_resched into shrink_node_memcg

2016-12-02 Thread Michal Hocko
From: Michal Hocko Boris Zhmurov has reported RCU stalls during the kswapd reclaim: 17511.573645] INFO: rcu_sched detected stalls on CPUs/tasks: [17511.573699] 23-...: (22 ticks this GP) idle=92f/140/0 softirq=2638404/2638404 fqs=23 [17511.573740] (detected by 4, t=6389 jiffies,