Re: [PATCH 4/4] sched: Limit the amount of NUMA imbalance that can exist at fork time

2020-11-20 Thread Vincent Guittot
On Fri, 20 Nov 2020 at 10:06, Mel Gorman wrote: > > At fork time currently, a local node can be allowed to fill completely > and allow the periodic load balancer to fix the problem. This can be > problematic in cases where a task creates lots of threads that idle until > woken as part of a worker

[PATCH 4/4] sched: Limit the amount of NUMA imbalance that can exist at fork time

2020-11-20 Thread Mel Gorman
At fork time currently, a local node can be allowed to fill completely and allow the periodic load balancer to fix the problem. This can be problematic in cases where a task creates lots of threads that idle until woken as part of a worker poll causing a memory bandwidth problem. However, a

[PATCH 4/4] sched: Limit the amount of NUMA imbalance that can exist at fork time

2020-11-19 Thread Mel Gorman
At fork time currently, a local node can be allowed to fill completely and allow the periodic load balancer to fix the problem. This can be problematic in cases where a task creates lots of threads that idle until woken as part of a worker poll causing a memory bandwidth problem. However, a