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;
Subject: Re: sched: Move SCHED_DEBUG sysctl to debugfs
To:     Steven Rostedt <rostedt@goodmis.org>
Cc:     peterz@infradead.org, bristot@redhat.com, bsegall@google.com,
        dietmar.eggemann@arm.com, greg@kroah.com,
        gregkh@linuxfoundation.org, joshdon@google.com,
        juri.lelli@redhat.com, linux-kernel@vger.kernel.org,
        linux@rasmusvillemoes.dk, mgorman@suse.de, mingo@kernel.org,
        valentin.schneider@arm.com, vincent.guittot@linaro.org,
        linux-s390@vger.kernel.org, kvm@vger.kernel.org
References: <20210412102001.287610138@infradead.org>
 <20210427145925.5246-1-borntraeger@de.ibm.com>
 <20210427110926.24f41fbb@gandalf.local.home>
From:   Christian Borntraeger <borntraeger@de.ibm.com>
Message-ID: <ae04549f-b009-9d90-d312-5c544f5a5e14@de.ibm.com>
Date:   Tue, 27 Apr 2021 17:17:09 +0200
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101
 Thunderbird/78.9.1
MIME-Version: 1.0
In-Reply-To: <20210427110926.24f41fbb@gandalf.local.home>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Language: en-US
Content-Transfer-Encoding: 7bit
Precedence: bulk


On 27.04.21 17:09, Steven Rostedt wrote:
> On Tue, 27 Apr 2021 16:59:25 +0200
> Christian Borntraeger <borntraeger@de.ibm.com> wrote:
> 
>> Peter,
>>
>> I just realized that we moved away sysctl tunabled to debugfs in next.
>> We have seen several cases where it was benefitial to set
>> sched_migration_cost_ns to a lower value. For example with KVM I can
>> easily get 50% more transactions with 50000 instead of 500000.
>> Until now it was possible to use tuned or /etc/sysctl.conf to set
>> these things permanently.
>>
>> Given that some people do not want to have debugfs mounted all the time
>> I would consider this a regression. The sysctl tunable was always
>> available.
>>
>> I am ok with the "informational" things being in debugfs, but not
>> the tunables. So how do we proceed here?
> 
> Should there be a schedfs created?
> 
> This is the reason I created the tracefs file system, was to get the
> tracing code out of debugfs, as debugfs is a catch all for everything and
> can lead to poor and insecure interfaces that people do not want to add on
> systems that they still want tracing on.
> 
> Or perhaps we should add a "tunefs" for tunables that are stable interfaces
> that should not be in /proc but also not in debugfs.

Yes, a tunefs or schedfs could be considered a replacement for sysctl.
It will still break existing setups with kernel.sched* things in /etc/sysctl.conf
but at least there is a stable transition path.