Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Subject: Re: [RESEND RFC PATCH V3] sched: Improve scalability of
 select_idle_sibling using SMT balance
To:     Peter Zijlstra <peterz@infradead.org>
Cc:     subhra mazumdar <subhra.mazumdar@oracle.com>,
        linux-kernel@vger.kernel.org, mingo@redhat.com,
        dhaval.giani@oracle.com
References: <20180129233102.19018-1-subhra.mazumdar@oracle.com>
 <20180201123335.GV2249@hirez.programming.kicks-ass.net>
 <911d42cf-54c7-4776-c13e-7c11f8ebfd31@oracle.com>
 <20180202172153.GO2269@hirez.programming.kicks-ass.net>
 <ed78c502-c1f8-43b3-f3d4-e0256e4085ee@oracle.com>
 <5b9f0828-8d35-885b-3eba-d31ca46da642@oracle.com>
 <20180202200401.GS2269@hirez.programming.kicks-ass.net>
From:   Steven Sistare <steven.sistare@oracle.com>
Organization: Oracle Corporation
Message-ID: <70ff99d5-1240-69ae-63d4-773b201d8c0e@oracle.com>
Date:   Fri, 2 Feb 2018 16:17:52 -0500
User-Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:52.0) Gecko/20100101
 Thunderbird/52.5.2
MIME-Version: 1.0
In-Reply-To: <20180202200401.GS2269@hirez.programming.kicks-ass.net>
Content-Type: text/plain; charset=utf-8
Content-Language: en-US
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On 2/2/2018 3:04 PM, Peter Zijlstra wrote:
> On Fri, Feb 02, 2018 at 01:34:58PM -0500, Steven Sistare wrote:
>> Actually, I take back my take back.  I suspect the primary benefit
>> of random selection is that it breaks up resonance states where
>> CPUs that are busy tend to stay busy, and CPUs that are idle tend
>> to stay idle, which is reinforced by starting the search at target =
>> last cpu ran.
> 
> Which, according to there here patches:
> 
>   https://lkml.kernel.org/r/20180130104555.4125-1-mgorman@techsingularity.net
> 
> is a good thing, because of power management.

Yes, but it's a bad thing if ready to run tasks pile on busy CPUs and idle 
CPUs go unused.  Stating the obvious, when the search for idle fails, the 
thread goes on a busy CPU. The existing logic that checks and uses the
initial target if it is idle reduces unnecessary spreading and is power
friendly (indeed, added in the patch you reference).  Subhra's patch does
not change that.

- Steve