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;
Date:   Mon, 19 Oct 2020 11:00:28 +1000
From:   Nicholas Piggin <npiggin@gmail.com>
Subject: Re: [PATCH v2] page_alloc: Fix freeing non-compound pages
To:     Andrew Morton <akpm@linux-foundation.org>,
        Matthew Wilcox <willy@infradead.org>
Cc:     Hugh Dickins <hughd@google.com>, linux-kernel@vger.kernel.org,
        linux-mm@kvack.org,
        =?iso-8859-1?q?Peter=0A?= Zijlstra <peterz@infradead.org>
References: <20200926213919.26642-1-willy@infradead.org>
        <20200928180307.7573f3b6128b5e3007dfc9f0@linux-foundation.org>
        <20200929011719.GD30994@casper.infradead.org>
        <20200928214656.be4a0f29961589c074e518fa@linux-foundation.org>
In-Reply-To: <20200928214656.be4a0f29961589c074e518fa@linux-foundation.org>
MIME-Version: 1.0
Message-Id: <1603066445.g384du2cm9.astroid@bobo.none>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

Excerpts from Andrew Morton's message of September 29, 2020 2:46 pm:
> On Tue, 29 Sep 2020 02:17:19 +0100 Matthew Wilcox <willy@infradead.org> w=
rote:
>=20
>> On Mon, Sep 28, 2020 at 06:03:07PM -0700, Andrew Morton wrote:
>> > On Sat, 26 Sep 2020 22:39:19 +0100 "Matthew Wilcox (Oracle)" <willy@in=
fradead.org> wrote:
>> >=20
>> > > Here is a very rare race which leaks memory:

Great catch! [sorry, a bit behind with emails]

>> >=20
>> > Not worth a cc:stable?
>>=20
>> Yes, it probably should have been.
>=20
> Have you a feeling for how often this occurs?
>=20
>>  I just assume the stablebot will
>> pick up anything that has a Fixes: tag.
>=20
> We asked them not to do that for mm/ patches.  Crazy stuff was getting
> backported.
>=20
>> > >
>> > > --- a/mm/page_alloc.c
>> > > +++ b/mm/page_alloc.c
>> > > @@ -4947,6 +4947,9 @@ void __free_pages(struct page *page, unsigned =
int order)
>> > >  {
>> > >  	if (put_page_testzero(page))
>> > >  		free_the_page(page, order);
>> > > +	else if (!PageHead(page))
>> > > +		while (order-- > 0)
>> > > +			free_the_page(page + (1 << order), order);
>> >=20
>> > Well that's weird and scary looking.  `page' has non-zero refcount yet
>> > we go and free random followon pages.  Methinks it merits an
>> > explanatory comment?
>>=20
>> Well, poot.  I lost that comment in the shuffling of patches.  In a
>> different tree, I have:
>>=20
>> @@ -4943,10 +4943,19 @@ static inline void free_the_page(struct page *pa=
ge, unsi
>> gned int order)
>>                 __free_pages_ok(page, order);
>>  }
>> =20
>> +/*
>> + * If we free a non-compound allocation, another thread may have a
>=20
> "non-compound, higher-order", I suggest?
>=20
>> + * speculative reference to the first page.  It has no way of knowing
>> + * about the rest of the allocation, so we have to free all but the
>> + * first page here.
>> + */
>>  void __free_pages(struct page *page, unsigned int order)
>>  {
>>         if (put_page_testzero(page))
>>                 free_the_page(page, order);
>> +       else if (!PageHead(page))
>> +               while (order-- > 0)
>> +                       free_the_page(page + (1 << order), order);
>>  }
>>  EXPORT_SYMBOL(__free_pages);
>> =20
>>=20
>> Although I'm now thinking of making that comment into kernel-doc and
>> turning it into advice to the caller rather than an internal note to
>> other mm developers.
>=20
> hm.  But what action could the caller take?  The explanatory comment
> seems OK to me.

The version of this without the comment got merged. I didn't mind the
comment...

Thanks,
Nick