Received-SPF: pass (google.com: domain of linux-kernel+bounces-27035-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:4601:e00::3 as permitted sender) client-ip=2604:1380:4601:e00::3;
Precedence: bulk
MIME-Version: 1.0
References: <20240104162510.72773-1-urezki@gmail.com> <20240104162510.72773-5-urezki@gmail.com>
 <CALm+0cUM=5bg0eKQ4D-mm7ZaAnQbf+2NjetUYnqHOLq5uR0w5g@mail.gmail.com>
 <ZaUNCJyYREsw7O3h@pc636> <ZaUPo97uzZlGkNdY@pc636>
In-Reply-To: <ZaUPo97uzZlGkNdY@pc636>
From: Z qiang <qiang.zhang1211@gmail.com>
Date: Tue, 16 Jan 2024 14:19:34 +0800
Message-ID: <CALm+0cWVJPis4c6VhGviXaHF90+njEVXvOjVZ-COL43dJFgNbw@mail.gmail.com>
Subject: Re: [PATCH v4 4/4] rcu: Support direct wake-up of synchronize_rcu() users
To: Uladzislau Rezki <urezki@gmail.com>
Cc: "Paul E . McKenney" <paulmck@kernel.org>, RCU <rcu@vger.kernel.org>, 
	Neeraj upadhyay <Neeraj.Upadhyay@amd.com>, Boqun Feng <boqun.feng@gmail.com>, 
	Hillf Danton <hdanton@sina.com>, Joel Fernandes <joel@joelfernandes.org>, 
	LKML <linux-kernel@vger.kernel.org>, 
	Oleksiy Avramchenko <oleksiy.avramchenko@sony.com>, Frederic Weisbecker <frederic@kernel.org>
Content-Type: text/plain; charset="UTF-8"

>
> > Hello, Zqiang.
> >
> > > >
> > > >         // concurrent sr_normal_gp_cleanup work might observe this update.
> > > >         smp_store_release(&rcu_state.srs_done_tail, wait_tail);
> > > >         ASSERT_EXCLUSIVE_WRITER(rcu_state.srs_done_tail);
> > > >
> > > > -       if (wait_tail)
> > > > +       if (wait_tail->next)
> > > >                 queue_work(system_highpri_wq, &sr_normal_gp_cleanup);
> > > >
> > >
> > > I'm testing these patches :) , one question is as follows:
> > > Can we use (WQ_MEM_RECLAIM | WQ_HIGHPR)type of workqueue to perform
> > > wake-up actions? avoid kworker creation failure under memory pressure, causing
> > > the wake-up action to be delayed.
> > >
> > I do not have any objections in not doing that, so we can add.
> >
> > Thank for testing this!
> >
> I forgot to ask, is your testing simulates a low memory condition so
> you see the failure you refer to? Or it is just a possible scenario?
>

I'm not currently testing this feature in low memory scenarios,  I thought
of this possible scenario.  I will test it in a low memory scenario later and
let you know if it happens :) .

Thanks
Zqiang

>
> Thanks!

>
> --
> Uladzislau Rezki
>