Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <fe46e855-b812-c1c5-07ce-7a3715bcbd2e@redhat.com>
Date:   Thu, 3 Aug 2023 13:43:34 -0400
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101
 Thunderbird/102.7.1
Subject: Re: [PATCH v2] sched/core: Use empty mask to reset cpumasks in
 sched_setaffinity()
Content-Language: en-US
To:     Peter Zijlstra <peterz@infradead.org>
Cc:     Ingo Molnar <mingo@redhat.com>, Juri Lelli <juri.lelli@redhat.com>,
        Vincent Guittot <vincent.guittot@linaro.org>,
        Dietmar Eggemann <dietmar.eggemann@arm.com>,
        Steven Rostedt <rostedt@goodmis.org>,
        Ben Segall <bsegall@google.com>, Mel Gorman <mgorman@suse.de>,
        Daniel Bristot de Oliveira <bristot@redhat.com>,
        Valentin Schneider <vschneid@redhat.com>,
        linux-kernel@vger.kernel.org, Phil Auld <pauld@redhat.com>,
        Brent Rowsell <browsell@redhat.com>,
        Peter Hunt <pehunt@redhat.com>
References: <20230717180243.3607603-1-longman@redhat.com>
 <20230721094233.GF3630545@hirez.programming.kicks-ass.net>
From:   Waiman Long <longman@redhat.com>
In-Reply-To: <20230721094233.GF3630545@hirez.programming.kicks-ass.net>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Precedence: bulk


On 7/21/23 05:42, Peter Zijlstra wrote:
> On Mon, Jul 17, 2023 at 02:02:43PM -0400, Waiman Long wrote:
>> Since commit 8f9ea86fdf99 ("sched: Always preserve the user requested
>> cpumask"), user provided CPU affinity via sched_setaffinity(2) is
>> perserved even if the task is being moved to a different cpuset. However,
>> that affinity is also being inherited by any subsequently created child
>> processes which may not want or be aware of that affinity.
>>
>> One way to solve this problem is to provide a way to back off from
>> that user provided CPU affinity.  This patch implements such a scheme
>> by using an empty cpumask to signal a reset of the cpumasks to the
>> default as allowed by the current cpuset.
>>
>> Before this patch, passing in an empty cpumask to sched_setaffinity(2)
>> will always return an EINVAL error. With this patch, an error will no
>> longer be returned if sched_setaffinity(2) has been called before to
>> set up user_cpus_ptr. Instead, the user_cpus_ptr that stores the user
>> provided affinity will be cleared and the task's CPU affinity will be
>> reset to that of the current cpuset. No error will be returned in this
>> case to signal that a reset has happened.
>>
>> If sched_setaffinity(2) has not been called previously, an EINVAL error
>> will be returned with an empty cpumask just like before. As a result,
>> tests or tools that rely on this behavior will not be affected unless
>> they have somehow called sched_setaffinity(2) before.
>>
>> We will have to update the sched_setaffinity(2) manpage to document
>> this possible side effect of passing in an empty cpumask.
> So a normal task, that hasn't had it's affinity changed will have
> possible_mask.
>
> So why not use in_mask == possible_mask to clear the user state?

It is not straight forward for a user application to figure what exactly 
is possible_mask. Using a empty mask, however, is much easier with the 
CPU_ZERO() macro.

In many cases, tasks may be running under a cpuset with cpu list that is 
a proper subset of possible_mask. Since possible_mask is a valid 
cpumask, we can't use it to reset to the cpuset's default which is what 
this patch is.

Cheers,
Longman