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;
MIME-Version: 1.0
References: <20230712060144.3006358-1-fengwei.yin@intel.com>
 <20230712060144.3006358-4-fengwei.yin@intel.com> <CAOUHufYef--8MxFettL6fOGjVx2vyZHZQU6EEaTCoW0XBvuC8Q@mail.gmail.com>
 <CAOUHufZ6=9P_=CAOQyw0xw-3q707q-1FVV09dBNDC-hpcpj2Pg@mail.gmail.com>
 <40cbc39e-5179-c2f4-3cea-0a98395aaff1@intel.com> <CAOUHufZHyEvU-c2O6B6stM_QVMxc22zV4Szn52myYqjdZvptUA@mail.gmail.com>
 <d93030e9-938e-f394-8067-a7307f6acd30@intel.com> <045fe36e-ed5b-2b02-a40d-c185796464b2@intel.com>
In-Reply-To: <045fe36e-ed5b-2b02-a40d-c185796464b2@intel.com>
From:   Yu Zhao <yuzhao@google.com>
Date:   Mon, 17 Jul 2023 21:59:11 -0600
Message-ID: <CAOUHufa63EfXu_5pi8Rs2BQSdfv-x56foH1p8ZtBm5YJTVNBVQ@mail.gmail.com>
Subject: Re: [RFC PATCH v2 3/3] mm: mlock: update mlock_pte_range to handle
 large folio
To:     Yin Fengwei <fengwei.yin@intel.com>
Cc:     linux-mm@kvack.org, linux-kernel@vger.kernel.org,
        akpm@linux-foundation.org, willy@infradead.org, david@redhat.com,
        ryan.roberts@arm.com, shy828301@gmail.com
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Mon, Jul 17, 2023 at 8:07=E2=80=AFPM Yin Fengwei <fengwei.yin@intel.com>=
 wrote:
>
> On 7/17/23 16:12, Yin Fengwei wrote:
> >
> > On 7/17/23 08:35, Yu Zhao wrote:
> >> On Sun, Jul 16, 2023 at 6:00=E2=80=AFPM Yin, Fengwei <fengwei.yin@inte=
l.com> wrote:
> >>>
> >>> On 7/15/2023 2:06 PM, Yu Zhao wrote:
> >>>> There is a problem here that I didn't have the time to elaborate: we
> >>>> can't mlock() a folio that is within the range but not fully mapped
> >>>> because this folio can be on the deferred split queue. When the spli=
t
> >>>> happens, those unmapped folios (not mapped by this vma but are mappe=
d
> >>>> into other vmas) will be stranded on the unevictable lru.
> >>>
> >>> This should be fine unless I missed something. During large folio spl=
it,
> >>> the unmap_folio() will be migrate(anon)/unmap(file) folio. Folio will=
 be
> >>> munlocked in unmap_folio(). So the head/tail pages will be evictable =
always.
> >>
> >> It's close but not entirely accurate: munlock can fail on isolated fol=
ios.
> >
> > I suppose normal 4K page can hit this problem also and following patch =
could
> > fix it:
> No. This patch is not necessary as unevictable folio will not be picked u=
p by
> page reclaim. It's not possible to munlock the isolated folio from lru li=
st.
>
> The possible cases I am ware are: page_migrate, madvise and damon_pa_page=
out and
> lru_gen_look_around. The first three already handle this case correctly b=
y call
>   folio_putback_lru().
>
> If folio is isolated, the split_folio() will just fail. So looks we are f=
ine
> for this corner case. Let me know if I miss something here.

The race is between isolation and munlock -- split_folio() only fails
if a folio is still isolated when it tries to freeze its refcnt, e.g.,
cpu 1      cpu 2
           split_folio()
isolation  unmap_folio()
putback
           freeze refcnt