Received-SPF: pass (google.com: best guess record for domain of linux-ext4-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
MIME-Version: 1.0
References: <20200217193314.GA12604@mit.edu> <20200218170857.GA28774@pc636>
 <20200220045233.GC476845@mit.edu> <20200221003035.GC2935@paulmck-ThinkPad-P72>
 <20200221131455.GA4904@pc636> <20200221202250.GK2935@paulmck-ThinkPad-P72>
 <20200222222415.GC191380@google.com> <20200223011018.GB2935@paulmck-ThinkPad-P72>
 <20200224174030.GA22138@pc636> <20200225020705.GA253171@google.com> <20200225035549.GU2935@paulmck-ThinkPad-P72>
In-Reply-To: <20200225035549.GU2935@paulmck-ThinkPad-P72>
From:   Joel Fernandes <joel@joelfernandes.org>
Date:   Tue, 25 Feb 2020 09:17:11 -0500
Message-ID: <CAEXW_YQbiHW=yKiYAs0=8Mp84W6UunM6OOkHE66yXT6cSchm7A@mail.gmail.com>
Subject: Re: [PATCH RFC] ext4: fix potential race between online resizing and
 write operations
To:     "Paul E. McKenney" <paulmck@kernel.org>
Cc:     Uladzislau Rezki <urezki@gmail.com>,
        "Theodore Y. Ts'o" <tytso@mit.edu>,
        Ext4 Developers List <linux-ext4@vger.kernel.org>,
        Suraj Jitindar Singh <surajjs@amazon.com>,
        LKML <linux-kernel@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Sender: linux-ext4-owner@vger.kernel.org
Precedence: bulk

On Mon, Feb 24, 2020 at 10:55 PM Paul E. McKenney <paulmck@kernel.org> wrote:
[...]
> > > As for "task_struct's rcu_read_lock_nesting". Will it be enough just
> > > have a look at preempt_count of current process? If we have for example
> > > nested rcu_read_locks:
> > >
> > > <snip>
> > > rcu_read_lock()
> > >     rcu_read_lock()
> > >         rcu_read_lock()
> > > <snip>
> > >
> > > the counter would be 3.
> >
> > No, because preempt_count is not incremented during rcu_read_lock(). RCU
> > reader sections can be preempted, they just cannot goto sleep in a reader
> > section (unless the kernel is RT).
>
> You are both right.
>
> Vlad is correct for CONFIG_PREEMPT=n and CONFIG_DEBUG_ATOMIC_SLEEP=y
> and Joel is correct otherwise, give or take the possibility of other
> late-breaking corner cases.  ;-)

Oh yes, but even for PREEMPT=n, rcu_read_lock() is just a NOOP for
that configuration and doesn't really mess around with preempt_count
if I recall :-D. (doesn't need to mess with preempt_count because
being in kernel mode is non-preemptible for PREEMPT=n anyway).

thanks,

- Joel