From: Daniel Jordan <daniel.m.jordan-QHcLZuEGTsvQT0dZR+AlfA@public.gmane.org>
To: Alexey Klimov <aklimov-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>,
linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
cgroups-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
Cc: peterz-wEGCiKHe2LqWVfeAwA7xHQ@public.gmane.org,
yury.norov-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org,
tglx-hfZtesqFncYOwBW4kG4KsQ@public.gmane.org,
jobaker-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org,
audralmitchel-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org,
arnd-r2nGTMty4D4@public.gmane.org,
gregkh-hQyY1W1yCW8ekmWlsbkhG0B+6BGkLq7r@public.gmane.org,
rafael-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org,
tj-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org,
qais.yousef-5wv7dgnIgG8@public.gmane.org,
hannes-druUgvl0LCNAfugRpC6u6w@public.gmane.org,
klimov.linux-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org
Subject: Re: [PATCH v3] cpu/hotplug: wait for cpuset_hotplug_work to finish on cpu onlining
Date: Thu, 18 Mar 2021 15:28:09 -0400 [thread overview]
Message-ID: <877dm4uura.fsf@oracle.com> (raw)
In-Reply-To: <20210317003616.2817418-1-aklimov-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
Alexey Klimov <aklimov-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org> writes:
> When a CPU offlined and onlined via device_offline() and device_online()
> the userspace gets uevent notification. If, after receiving "online" uevent,
> userspace executes sched_setaffinity() on some task trying to move it
> to a recently onlined CPU, then it sometimes fails with -EINVAL. Userspace
> needs to wait around 5..30 ms before sched_setaffinity() will succeed for
> recently onlined CPU after receiving uevent.
>
> If in_mask argument for sched_setaffinity() has only recently onlined CPU,
> it could fail with such flow:
>
> sched_setaffinity()
> cpuset_cpus_allowed()
> guarantee_online_cpus() <-- cs->effective_cpus mask does not
> contain recently onlined cpu
> cpumask_and() <-- final new_mask is empty
> __set_cpus_allowed_ptr()
> cpumask_any_and_distribute() <-- returns dest_cpu equal to nr_cpu_ids
> returns -EINVAL
>
> Cpusets used in guarantee_online_cpus() are updated using workqueue from
> cpuset_update_active_cpus() which in its turn is called from cpu hotplug callback
> sched_cpu_activate() hence it may not be observable by sched_setaffinity() if
> it is called immediately after uevent.
>
> Out of line uevent can be avoided if we will ensure that cpuset_hotplug_work
> has run to completion using cpuset_wait_for_hotplug() after onlining the
> cpu in cpu_device_up() and in cpuhp_smt_enable().
>
> Cc: Daniel Jordan <daniel.m.jordan-QHcLZuEGTsvQT0dZR+AlfA@public.gmane.org>
> Reviewed-by: Qais Yousef <qais.yousef-5wv7dgnIgG8@public.gmane.org>
> Co-analyzed-by: Joshua Baker <jobaker-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
> Signed-off-by: Alexey Klimov <aklimov-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
Looks good to me.
Reviewed-by: Daniel Jordan <daniel.m.jordan-QHcLZuEGTsvQT0dZR+AlfA@public.gmane.org>
WARNING: multiple messages have this Message-ID (diff)
From: Daniel Jordan <daniel.m.jordan@oracle.com>
To: Alexey Klimov <aklimov@redhat.com>,
linux-kernel@vger.kernel.org, cgroups@vger.kernel.org
Cc: peterz@infradead.org, yury.norov@gmail.com, tglx@linutronix.de,
jobaker@redhat.com, audralmitchel@gmail.com, arnd@arndb.de,
gregkh@linuxfoundation.org, rafael@kernel.org, tj@kernel.org,
qais.yousef@arm.com, hannes@cmpxchg.org, klimov.linux@gmail.com
Subject: Re: [PATCH v3] cpu/hotplug: wait for cpuset_hotplug_work to finish on cpu onlining
Date: Thu, 18 Mar 2021 15:28:09 -0400 [thread overview]
Message-ID: <877dm4uura.fsf@oracle.com> (raw)
In-Reply-To: <20210317003616.2817418-1-aklimov@redhat.com>
Alexey Klimov <aklimov@redhat.com> writes:
> When a CPU offlined and onlined via device_offline() and device_online()
> the userspace gets uevent notification. If, after receiving "online" uevent,
> userspace executes sched_setaffinity() on some task trying to move it
> to a recently onlined CPU, then it sometimes fails with -EINVAL. Userspace
> needs to wait around 5..30 ms before sched_setaffinity() will succeed for
> recently onlined CPU after receiving uevent.
>
> If in_mask argument for sched_setaffinity() has only recently onlined CPU,
> it could fail with such flow:
>
> sched_setaffinity()
> cpuset_cpus_allowed()
> guarantee_online_cpus() <-- cs->effective_cpus mask does not
> contain recently onlined cpu
> cpumask_and() <-- final new_mask is empty
> __set_cpus_allowed_ptr()
> cpumask_any_and_distribute() <-- returns dest_cpu equal to nr_cpu_ids
> returns -EINVAL
>
> Cpusets used in guarantee_online_cpus() are updated using workqueue from
> cpuset_update_active_cpus() which in its turn is called from cpu hotplug callback
> sched_cpu_activate() hence it may not be observable by sched_setaffinity() if
> it is called immediately after uevent.
>
> Out of line uevent can be avoided if we will ensure that cpuset_hotplug_work
> has run to completion using cpuset_wait_for_hotplug() after onlining the
> cpu in cpu_device_up() and in cpuhp_smt_enable().
>
> Cc: Daniel Jordan <daniel.m.jordan@oracle.com>
> Reviewed-by: Qais Yousef <qais.yousef@arm.com>
> Co-analyzed-by: Joshua Baker <jobaker@redhat.com>
> Signed-off-by: Alexey Klimov <aklimov@redhat.com>
Looks good to me.
Reviewed-by: Daniel Jordan <daniel.m.jordan@oracle.com>
next prev parent reply other threads:[~2021-03-18 19:28 UTC|newest]
Thread overview: 18+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-03-17 0:36 [PATCH v3] cpu/hotplug: wait for cpuset_hotplug_work to finish on cpu onlining Alexey Klimov
2021-03-17 0:36 ` Alexey Klimov
[not found] ` <20210317003616.2817418-1-aklimov-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
2021-03-18 19:28 ` Daniel Jordan [this message]
2021-03-18 19:28 ` Daniel Jordan
2021-03-27 21:01 ` Thomas Gleixner
2021-03-27 21:01 ` Thomas Gleixner
[not found] ` <87tuowcnv3.ffs-ecDvlHI5BZPZikZi3RtOZ1XZhhPuCNm+@public.gmane.org>
2021-04-01 14:06 ` Qais Yousef
2021-04-01 14:06 ` Qais Yousef
2021-04-04 2:32 ` Alexey Klimov
2021-04-04 2:32 ` Alexey Klimov
[not found] ` <CALW4P+L9_tYgfOPv0riWWnv54HPhKPDJ4EK4yYaWsz0MdDGqfw-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2021-04-15 1:30 ` Alexey Klimov
2021-04-15 1:30 ` Alexey Klimov
[not found] ` <CAFBcO+8NBZxNdXtVuTXt9_m9gWTq7kxrcDcdFntvVjR_0rM13A-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2021-06-14 18:14 ` Alexey Klimov
2021-06-14 18:14 ` Alexey Klimov
[not found] ` <CAFBcO+9wLjDW6n-ZSean_UQHSJ44Tpw9XBz-3UMoVCeUridj4Q-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2021-06-19 20:18 ` Thomas Gleixner
2021-06-19 20:18 ` Thomas Gleixner
2021-06-19 20:27 ` [tip: smp/urgent] cpu/hotplug: Cure the cpusets trainwreck tip-bot2 for Thomas Gleixner
2021-06-21 8:36 ` tip-bot2 for Thomas Gleixner
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=877dm4uura.fsf@oracle.com \
--to=daniel.m.jordan-qhclzuegtsvqt0dzr+alfa@public.gmane.org \
--cc=aklimov-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org \
--cc=arnd-r2nGTMty4D4@public.gmane.org \
--cc=audralmitchel-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org \
--cc=cgroups-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
--cc=gregkh-hQyY1W1yCW8ekmWlsbkhG0B+6BGkLq7r@public.gmane.org \
--cc=hannes-druUgvl0LCNAfugRpC6u6w@public.gmane.org \
--cc=jobaker-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org \
--cc=klimov.linux-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org \
--cc=linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
--cc=peterz-wEGCiKHe2LqWVfeAwA7xHQ@public.gmane.org \
--cc=qais.yousef-5wv7dgnIgG8@public.gmane.org \
--cc=rafael-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org \
--cc=tglx-hfZtesqFncYOwBW4kG4KsQ@public.gmane.org \
--cc=tj-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org \
--cc=yury.norov-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.