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: <20230728161356.1784568-1-fengwei.yin@intel.com>
 <3bbfde16-ced1-dca8-6a3f-da893e045bc5@arm.com> <56c8f4f9-b54b-b0bb-250c-ec8643accfc7@intel.com>
 <3541d2de-5cf8-2f84-8153-277e2bfc0101@arm.com> <5f98748a-97ca-6426-1e24-a5675da75381@intel.com>
 <a590da86-0c42-7d46-d320-c661a59a46c1@arm.com> <837ba176-c97f-f81b-c044-eb6aa3d88bb7@intel.com>
In-Reply-To: <837ba176-c97f-f81b-c044-eb6aa3d88bb7@intel.com>
From:   Yu Zhao <yuzhao@google.com>
Date:   Thu, 3 Aug 2023 14:46:23 -0600
Message-ID: <CAOUHufY9EQ70Pn-n2zVa9=Gm3-WHxxphp7VHia4qv9x2domdbg@mail.gmail.com>
Subject: Re: [PATCH 0/2] don't use mapcount() to check large folio sharing
To:     "Yin, Fengwei" <fengwei.yin@intel.com>
Cc:     Ryan Roberts <ryan.roberts@arm.com>, linux-mm@kvack.org,
        linux-kernel@vger.kernel.org, stable@vger.kernel.org,
        akpm@linux-foundation.org, willy@infradead.org,
        vishal.moola@gmail.com, wangkefeng.wang@huawei.com,
        minchan@kernel.org, david@redhat.com, shy828301@gmail.com
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Wed, Aug 2, 2023 at 6:56=E2=80=AFAM Yin, Fengwei <fengwei.yin@intel.com>=
 wrote:
>
>
>
> On 8/2/2023 8:49 PM, Ryan Roberts wrote:
> > On 02/08/2023 13:42, Yin, Fengwei wrote:
> >>
> >>
> >> On 8/2/2023 8:40 PM, Ryan Roberts wrote:
> >>> On 02/08/2023 13:35, Yin, Fengwei wrote:
> >>>>
> >>>>
> >>>> On 8/2/2023 6:27 PM, Ryan Roberts wrote:
> >>>>> On 28/07/2023 17:13, Yin Fengwei wrote:
> >>>>>> In madvise_cold_or_pageout_pte_range() and madvise_free_pte_range(=
),
> >>>>>> folio_mapcount() is used to check whether the folio is shared. But=
 it's
> >>>>>> not correct as folio_mapcount() returns total mapcount of large fo=
lio.
> >>>>>>
> >>>>>> Use folio_estimated_sharers() here as the estimated number is enou=
gh.
> >>>>>>
> >>>>>> Yin Fengwei (2):
> >>>>>>   madvise: don't use mapcount() against large folio for sharing ch=
eck
> >>>>>>   madvise: don't use mapcount() against large folio for sharing ch=
eck
> >>>>>>
> >>>>>>  mm/huge_memory.c | 2 +-
> >>>>>>  mm/madvise.c     | 6 +++---
> >>>>>>  2 files changed, 4 insertions(+), 4 deletions(-)
> >>>>>>
> >>>>>
> >>>>> As a set of fixes, I agree this is definitely an improvement, so:
> >>>>>
> >>>>> Reviewed-By: Ryan Roberts
> >>>> Thanks.
> >>>>
> >>>>>
> >>>>>
> >>>>> But I have a couple of comments around further improvements;
> >>>>>
> >>>>> Once we have the scheme that David is working on to be able to prov=
ide precise
> >>>>> exclusive vs shared info, we will probably want to move to that. Al=
though that
> >>>>> scheme will need access to the mm_struct of a process known to be m=
apping the
> >>>>> folio. We have that info, but its not passed to folio_estimated_sha=
rers() so we
> >>>>> can't just reimplement folio_estimated_sharers() - we will need to =
rework these
> >>>>> call sites again.
> >>>> Yes. This could be extra work. Maybe should delay till David's work =
is done.
> >>>
> >>> What you have is definitely an improvement over what was there before=
. And is
> >>> probably the best we can do without David's scheme. So I wouldn't del=
ay this.
> >>> Just pointing out that we will be able to make it even better later o=
n (if
> >>> David's stuff goes in).
> >> Yes. I agree that we should wait for David's work ready and do fix bas=
ed on that.
> >
> > I was suggesting the opposite - not waiting. Then we can do separate im=
provement
> > later.
> Let's wait for David's work ready.

Waiting is fine as long as we don't miss the next merge window -- we
don't want these two bugs to get into another release. Also I think we
should cc stable, since as David mentioned, they have been causing
selftest failures.