Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Date:   Wed, 19 May 2021 23:04:30 +0530
From:   Srikar Dronamraju <srikar@linux.vnet.ibm.com>
To:     Aubrey Li <aubrey.li@linux.intel.com>
Cc:     Ingo Molnar <mingo@kernel.org>,
        Peter Zijlstra <peterz@infradead.org>,
        LKML <linux-kernel@vger.kernel.org>,
        Mel Gorman <mgorman@techsingularity.net>,
        Rik van Riel <riel@surriel.com>,
        Thomas Gleixner <tglx@linutronix.de>,
        Valentin Schneider <valentin.schneider@arm.com>,
        Vincent Guittot <vincent.guittot@linaro.org>,
        Dietmar Eggemann <dietmar.eggemann@arm.com>,
        Gautham R Shenoy <ego@linux.vnet.ibm.com>,
        Parth Shah <parth@linux.ibm.com>
Subject: Re: [PATCH v2 6/8] sched/idle: Move busy_cpu accounting to idle
 callback
Message-ID: <20210519173430.GB2633526@linux.vnet.ibm.com>
Reply-To: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
References: <20210513073112.GV2633526@linux.vnet.ibm.com>
 <5823f298-6fae-5a73-3ab8-f708d90a7e52@linux.intel.com>
 <20210517104058.GW2633526@linux.vnet.ibm.com>
 <9d493353-7a27-16aa-3e99-c6a07e69de25@linux.intel.com>
 <20210517125727.GX2633526@linux.vnet.ibm.com>
 <27ab234c-b36b-bf7f-52f4-92c1804f8245@linux.intel.com>
 <20210518040024.GY2633526@linux.vnet.ibm.com>
 <ce4eba7e-dc2e-1c67-d04d-579f04c2040b@linux.intel.com>
 <20210518071843.GZ2633526@linux.vnet.ibm.com>
 <32b98350-35e4-7475-2d19-9101f50ecc63@linux.intel.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1
Content-Disposition: inline
In-Reply-To: <32b98350-35e4-7475-2d19-9101f50ecc63@linux.intel.com>
Precedence: bulk

* Aubrey Li <aubrey.li@linux.intel.com> [2021-05-19 17:43:55]:

> On 5/18/21 3:18 PM, Srikar Dronamraju wrote:
> 
> >>>> This is v3. It looks like hackbench gets better. And netperf still has
> >>>> some notable changes under 2 x overcommit cases.
> >>>>
> >>>
> >>> Thanks Aubrey for the results. netperf (2X) case does seem to regress.
> >>> I was actually expecting the results to get better with overcommit.
> >>> Can you confirm if this was just v3 or with v3 + set_next_idle_core
> >>> disabled?
> >>
> >> Do you mean set_idle_cores(not set_next_idle_core) actually? Gautham's patch
> >> changed "this" to "target" in set_idle_cores, and I removed it to apply
> >> v3-2-8-sched-fair-Maintain-the-identity-of-idle-core.patch for tip/sched/core
> >> commit-id 915a2bc3c6b7.
> > 
> > Thats correct,
> > 
> > In the 3rd patch, I had introduced set_next_idle_core
> > which is suppose to set idle_cores in the LLC.
> > What I suspected was is this one is causing issues in your 48 CPU LLC.
> > 
> > I am expecting set_next_idle_core to be spending much time in your scenario.
> > I was planning for something like the below on top of my patch.
> > With this we dont look for an idle-core if we already know that we dont find one.
> > But in the mean while I had asked if you could have dropped the call to
> > set_next_idle_core.
> > 
> 
> +	if (atomic_read(&sd->shared->nr_busy_cpus) * 2 >=  per_cpu(sd_llc_size, target))
> +		goto out;
> 
> Does this has side effect if waker and wakee are coalesced on a portion of cores?
> Also, is 2 a SMT2 assumption?

The above line was just a hack to see if things change by not spending time
searching for an idle-core. Since you were running on Intel, I had
hard-coded it to 2.

> 
> I did a quick testing on this, it looks like the regression of netperf 2x cases are 
> improved indeed, but hackbench two mid-load cases get worse.
> 

In the mid-loaded, case, there was a chance of idle-core being around, Since
we dont set it to an idle-core or -2, i.e idle-core is set to -1, it may or
may not get selected.

> process-sockets 	group-2 	 1.00 (  5.32)	-18.40 (  7.32)
> threads-sockets 	group-2 	 1.00 (  5.44)	-20.44 (  4.60)
> 
> Thanks,
> -Aubrey

-- 
Thanks and Regards
Srikar Dronamraju