From:   Valentin Schneider <vschneid@redhat.com>
To:     Peter Zijlstra <peterz@infradead.org>,
        Ricardo Neri <ricardo.neri-calderon@linux.intel.com>
Cc:     Juri Lelli <juri.lelli@redhat.com>,
        Vincent Guittot <vincent.guittot@linaro.org>,
        Ricardo Neri <ricardo.neri@intel.com>,
        "Ravi V. Shankar" <ravi.v.shankar@intel.com>,
        Ben Segall <bsegall@google.com>,
        Daniel Bristot de Oliveira <bristot@redhat.com>,
        Dietmar Eggemann <dietmar.eggemann@arm.com>,
        Len Brown <len.brown@intel.com>, Mel Gorman <mgorman@suse.de>,
        "Rafael J. Wysocki" <rafael.j.wysocki@intel.com>,
        Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>,
        Steven Rostedt <rostedt@goodmis.org>,
        Tim Chen <tim.c.chen@linux.intel.com>,
        Ionela Voinescu <ionela.voinescu@arm.com>, x86@kernel.org,
        linux-kernel@vger.kernel.org, "Tim C . Chen" <tim.c.chen@intel.com>
Subject: Re: [PATCH v3 06/10] sched/fair: Use the prefer_sibling flag of the
 current sched domain
In-Reply-To: <Y+YXrk5NRuWaSOGR@hirez.programming.kicks-ass.net>
References: <20230207045838.11243-1-ricardo.neri-calderon@linux.intel.com>
 <20230207045838.11243-7-ricardo.neri-calderon@linux.intel.com>
 <Y+YXrk5NRuWaSOGR@hirez.programming.kicks-ass.net>
Date:   Fri, 10 Feb 2023 14:54:56 +0000
Message-ID: <xhsmhmt5lr2nz.mognet@vschneid.remote.csb>
MIME-Version: 1.0
Content-Type: text/plain
Precedence: bulk

On 10/02/23 11:08, Peter Zijlstra wrote:
> On Mon, Feb 06, 2023 at 08:58:34PM -0800, Ricardo Neri wrote:
>> SD_PREFER_SIBLING is set from the SMT scheduling domain up to the first
>> non-NUMA domain (the exception is systems with SD_ASYM_CPUCAPACITY).
>>
>> Above the SMT sched domain, all domains have a child. The SD_PREFER_
>> SIBLING is honored always regardless of the scheduling domain at which the
>> load balance takes place.
>>
>> There are cases, however, in which the busiest CPU's sched domain has
>> child but the destination CPU's does not. Consider, for instance a non-SMT
>> core (or an SMT core with only one online sibling) doing load balance with
>> an SMT core at the MC level. SD_PREFER_SIBLING will not be honored. We are
>> left with a fully busy SMT core and an idle non-SMT core.
>>
>> Avoid inconsistent behavior. Use the prefer_sibling behavior at the current
>> scheduling domain, not its child.
>>
>> The NUMA sched domain does not have the SD_PREFER_SIBLING flag. Thus, we
>> will not spread load among NUMA sched groups, as desired.
>>
>
> Like many of the others; I don't much like this.
>
> Why not simply detect this asymmetric having of SMT and kill the
> PREFER_SIBLING flag on the SMT leafs in that case?
>
> Specifically, I'm thinking something in the degenerate area where it
> looks if a given domain has equal depth children or so.
>
> Note that this should not be tied to having special hardware, you can
> create the very same weirdness by just offlining a few SMT siblings and
> leaving a few on.

So something like have SD_PREFER_SIBLING affect the SD it's on (and not
its parent), but remove it from the lowest non-degenerated topology level?
(+ add it to the first NUMA level to keep things as they are, even if TBF I
find relying on it for NUMA balancing a bit odd).