From: Peter Zijlstra <peterz@infradead.org>
To: Mel Gorman <mgorman@techsingularity.net>
Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com>,
Ingo Molnar <mingo@kernel.org>, Rik van Riel <riel@surriel.com>,
LKML <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH 4/4] sched/numa: Do not move imbalanced load purely on the basis of an idle CPU
Date: Fri, 7 Sep 2018 13:33:09 +0200 [thread overview]
Message-ID: <20180907113309.GU24106@hirez.programming.kicks-ass.net> (raw)
In-Reply-To: <20180907101139.20760-5-mgorman@techsingularity.net>
On Fri, Sep 07, 2018 at 11:11:39AM +0100, Mel Gorman wrote:
> Commit 305c1fac3225 ("sched/numa: Evaluate move once per node")
> restructured how task_numa_compare evaluates load but there is an anomaly.
> task_numa_find_cpu() checks if the load balance between too nodes is too
> imbalanced with the intent of only swapping tasks if it would improve
> the balance overall. However, if an idle CPU is encountered, the task is
> still moved if it's the best improvement but an idle cpu is always going
> to appear to be the best improvement.
>
> If a machine is lightly loaded such that all tasks can fit on one node then
> the idle CPUs are found and the tasks migrate to one socket. From a NUMA
> perspective, this seems intuitively great because memory accesses are all
> local but there are two counter-intuitive effects.
>
> First, the load balancer may move tasks so the machine is more evenly
> utilised and conflict with automatic NUMA balancing which may respond by
> scanning more frequently and increasing overhead. Second, sockets usually
> have their own memory channels so using one socket means that fewer
> channels are available yielding less memory bandwidth overall. For
> memory-bound tasks, it can be beneficial to migrate to another socket and
> migrate the data to increase bandwidth even though the accesses are remote
> in the short term.
> ---
> kernel/sched/fair.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
> index d59d3e00a480..d4c289c11012 100644
> --- a/kernel/sched/fair.c
> +++ b/kernel/sched/fair.c
> @@ -1560,7 +1560,7 @@ static bool task_numa_compare(struct task_numa_env *env,
> goto unlock;
>
> if (!cur) {
> - if (maymove || imp > env->best_imp)
> + if (maymove)
> goto assign;
> else
> goto unlock;
Srikar's patch here:
http://lkml.kernel.org/r/1533276841-16341-4-git-send-email-srikar@linux.vnet.ibm.com
Also frobs this condition, but in a less radical way. Does that yield
similar results?
next prev parent reply other threads:[~2018-09-07 11:33 UTC|newest]
Thread overview: 22+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-09-07 10:11 [PATCH 0/4] Follow-up fixes for v4.19-rc1 NUMA balancing Mel Gorman
2018-09-07 10:11 ` [PATCH 1/4] sched/numa: Remove redundant numa_stats nr_running field Mel Gorman
2018-09-07 10:11 ` [PATCH 2/4] sched/numa: Remove unused calculations in update_numa_stats Mel Gorman
2018-09-07 10:11 ` [PATCH 3/4] sched/numa: Stop comparing tasks for NUMA placement after selecting an idle core Mel Gorman
2018-09-07 13:05 ` Srikar Dronamraju
2018-09-07 14:20 ` Mel Gorman
2018-09-07 10:11 ` [PATCH 4/4] sched/numa: Do not move imbalanced load purely on the basis of an idle CPU Mel Gorman
2018-09-07 11:33 ` Peter Zijlstra [this message]
2018-09-07 12:37 ` Mel Gorman
2018-09-07 12:44 ` Peter Zijlstra
2018-09-07 13:42 ` Srikar Dronamraju
2018-09-07 14:28 ` Mel Gorman
2018-09-10 9:41 ` Mel Gorman
2018-09-12 6:54 ` Srikar Dronamraju
2018-09-12 9:36 ` Peter Zijlstra
2018-09-12 10:45 ` Srikar Dronamraju
2018-09-12 9:57 ` Ingo Molnar
2018-09-12 10:27 ` Srikar Dronamraju
2018-09-12 10:57 ` Mel Gorman
2018-09-12 10:52 ` Mel Gorman
2018-09-07 11:24 ` [PATCH 0/4] Follow-up fixes for v4.19-rc1 NUMA balancing Peter Zijlstra
2018-09-07 12:29 ` Mel Gorman
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20180907113309.GU24106@hirez.programming.kicks-ass.net \
--to=peterz@infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mgorman@techsingularity.net \
--cc=mingo@kernel.org \
--cc=riel@surriel.com \
--cc=srikar@linux.vnet.ibm.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).