[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Tue, 02 Feb 2021 16:48:23 +0000
From: Valentin Schneider <valentin.schneider@....com>
To: Barry Song <song.bao.hua@...ilicon.com>,
vincent.guittot@...aro.org, mgorman@...e.de, mingo@...nel.org,
peterz@...radead.org, dietmar.eggemann@....com,
morten.rasmussen@....com, linux-kernel@...r.kernel.org
Cc: linuxarm@...neuler.org, xuwei5@...wei.com, liguozhu@...ilicon.com,
tiantao6@...ilicon.com, wanghuiqiang@...wei.com,
prime.zeng@...ilicon.com, jonathan.cameron@...wei.com,
guodong.xu@...aro.org, Barry Song <song.bao.hua@...ilicon.com>,
Meelis Roos <mroos@...ux.ee>
Subject: Re: [PATCH] sched/topology: fix the issue groups don't span domain->span for NUMA diameter > 2
On 02/02/21 15:17, Valentin Schneider wrote:
> That is, rather than building overlapping groups and fixing them whenever
> that breaks (distance > 2), we could have:
> - the local group being the child domain's span (as always)
> - all non-local NUMA groups spanning a single node each, with the right sgc
> cross-referencing.
>
> Thoughts?
>
Hmph I'm thinking this can be broken by domain degeneration, as nothing
will come fix up the ->sgc link and we'd be back to having a reference to a
group that doesn't get updated...
Powered by blists - more mailing lists