[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <629c5dc5-cde4-76de-4680-9fb5ecfd26cd@oracle.com>
Date: Tue, 23 Oct 2018 09:18:42 -0400
From: Steven Sistare <steven.sistare@...cle.com>
To: mingo@...hat.com, peterz@...radead.org
Cc: subhra.mazumdar@...cle.com, dhaval.giani@...cle.com,
rohit.k.jain@...cle.com, daniel.m.jordan@...cle.com,
pavel.tatashin@...rosoft.com, matt@...eblueprint.co.uk,
umgwanakikbuti@...il.com, riel@...hat.com, jbacik@...com,
juri.lelli@...hat.com, linux-kernel@...r.kernel.org
Subject: Re: [PATCH 09/10] sched/fair: disable stealing if too many NUMA nodes
On 10/22/2018 10:59 AM, Steve Sistare wrote:
[...]
> diff --git a/kernel/sched/topology.c b/kernel/sched/topology.c
> index f18c416..b7d2070 100644
> --- a/kernel/sched/topology.c
> +++ b/kernel/sched/topology.c
> @@ -1337,6 +1337,30 @@ static void init_numa_topology_type(void)
> }
> }
>
> +DEFINE_STATIC_KEY_TRUE(sched_steal_allow);
> +static int sched_steal_node_limit;
> +#define SCHED_STEAL_NODE_LIMIT_DEFAULT 1
This is an oversight; I meant to suggest a value of 2. All my 2-socket
testing was done with stealing enabled.
- Steve
Powered by blists - more mailing lists