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;
MIME-Version: 1.0
References: <20210419085027.761150-1-elver@google.com> <20210419085027.761150-2-elver@google.com>
 <20210419094044.311-1-hdanton@sina.com> <CANpmjNMR-DPj=0mQMevyEQ7k3RJh0eq_nkt9M6kLvwC-abr_SQ@mail.gmail.com>
In-Reply-To: <CANpmjNMR-DPj=0mQMevyEQ7k3RJh0eq_nkt9M6kLvwC-abr_SQ@mail.gmail.com>
From:   Marco Elver <elver@google.com>
Date:   Mon, 19 Apr 2021 11:49:04 +0200
Message-ID: <CANpmjNNO3AgK3Fr07KXQhGpqt6-z7xNJFP=UoODg-Ft=u9cGfA@mail.gmail.com>
Subject: Re: [PATCH 1/3] kfence: await for allocation using wait_event
To:     Hillf Danton <hdanton@sina.com>
Cc:     Andrew Morton <akpm@linux-foundation.org>,
        Alexander Potapenko <glider@google.com>,
        Dmitry Vyukov <dvyukov@google.com>,
        Jann Horn <jannh@google.com>,
        Mark Rutland <mark.rutland@arm.com>,
        LKML <linux-kernel@vger.kernel.org>,
        Linux Memory Management List <linux-mm@kvack.org>,
        kasan-dev <kasan-dev@googlegroups.com>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Mon, 19 Apr 2021 at 11:44, Marco Elver <elver@google.com> wrote:
>
> On Mon, 19 Apr 2021 at 11:41, Hillf Danton <hdanton@sina.com> wrote:
> >
> > On Mon, 19 Apr 2021 10:50:25 Marco Elver wrote:
> > > +
> > > +     WRITE_ONCE(kfence_timer_waiting, true);
> > > +     smp_mb(); /* See comment in __kfence_alloc(). */
> >
> > This is not needed given task state change in wait_event().
>
> Yes it is. We want to avoid the unconditional irq_work in
> __kfence_alloc(). When the system is under load doing frequent
> allocations, at least in my tests this avoids the irq_work almost
> always. Without the irq_work you'd be correct of course.

And in case this is about the smp_mb() here, yes it definitely is
required. We *must* order the write of kfence_timer_waiting *before*
the check of kfence_allocation_gate, which wait_event() does before
anything else (including changing the state). Otherwise the write may
be reordered after the read, and we could potentially never wake up
because __kfence_alloc() not waking us.

This is documented in __kfence_alloc().

> > > +     wait_event_timeout(allocation_wait, atomic_read(&kfence_allocation_gate), HZ);
> > > +     smp_store_release(&kfence_timer_waiting, false); /* Order after wait_event(). */
> > > +