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;
Subject: Re: [PATCH v2 2/3] xen/privcmd: Mark pages as dirty
To:     =?UTF-8?B?SsO8cmdlbiBHcm/Dnw==?= <jgross@suse.com>,
        Souptick Joarder <jrdr.linux@gmail.com>
CC:     Boris Ostrovsky <boris.ostrovsky@oracle.com>,
        <sstabellini@kernel.org>, <xen-devel@lists.xenproject.org>,
        <linux-kernel@vger.kernel.org>, Paul Durrant <xadimgnik@gmail.com>
References: <1594059372-15563-1-git-send-email-jrdr.linux@gmail.com>
 <1594059372-15563-3-git-send-email-jrdr.linux@gmail.com>
 <8fdd8c77-27dd-2847-7929-b5d3098b1b45@suse.com>
 <CAFqt6zZRx3oDO+p2e6EiDig9fzKirME-t6fanzDRh6e7gWx+nA@mail.gmail.com>
 <4abc0dd2-655c-16fa-dfc3-95904196c81f@suse.com>
From:   John Hubbard <jhubbard@nvidia.com>
Message-ID: <4c6e52e7-1d33-132b-1d7e-e57963966dcc@nvidia.com>
Date:   Tue, 7 Jul 2020 12:30:57 -0700
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:68.0) Gecko/20100101
 Thunderbird/68.10.0
MIME-Version: 1.0
In-Reply-To: <4abc0dd2-655c-16fa-dfc3-95904196c81f@suse.com>
Content-Type: text/plain; charset="utf-8"; format=flowed
Content-Language: en-US
Content-Transfer-Encoding: quoted-printable
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On 2020-07-07 04:43, J=C3=BCrgen Gro=C3=9F wrote:
> On 07.07.20 13:30, Souptick Joarder wrote:
>> On Tue, Jul 7, 2020 at 3:08 PM J=C3=BCrgen Gro=C3=9F <jgross@suse.com> w=
rote:
...
>>>> diff --git a/drivers/xen/privcmd.c b/drivers/xen/privcmd.c
>>>> index 33677ea..f6c1543 100644
>>>> --- a/drivers/xen/privcmd.c
>>>> +++ b/drivers/xen/privcmd.c
>>>> @@ -612,8 +612,11 @@ static void unlock_pages(struct page *pages[], un=
signed int nr_pages)
>>>> =C2=A0=C2=A0 {
>>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 unsigned int i;
>>>>
>>>> -=C2=A0=C2=A0=C2=A0=C2=A0 for (i =3D 0; i < nr_pages; i++)
>>>> +=C2=A0=C2=A0=C2=A0=C2=A0 for (i =3D 0; i < nr_pages; i++) {
>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=
=A0 if (!PageDirty(pages[i]))
>>>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=
=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 set_page_dirty_lock(pag=
es[i]);
>>>
>>> With put_page() directly following I think you should be able to use
>>> set_page_dirty() instead, as there is obviously a reference to the page
>>> existing.
>>
>> Patch [3/3] will convert above codes to use unpin_user_pages_dirty_lock(=
)
>> which internally do the same check. So I thought to keep linux-stable an=
d
>> linux-next code in sync. John had a similar concern [1] and later agreed=
 to keep
>> this check.
>>
>> Shall I keep this check ?=C2=A0 No ?

It doesn't matter *too* much, because patch 3/3 fixes up everything by
changing it all to unpin_user_pages_dirty_lock(). However, there is somethi=
ng
to be said for having correct interim patches, too. :)  Details:

>>
>> [1] https://lore.kernel.org/xen-devel/a750e5e5-fd5d-663b-c5fd-261d7c939b=
a7@nvidia.com/
>=20
> I wasn't referring to checking PageDirty(), but to the use of
> set_page_dirty_lock().
>=20
> Looking at the comment just before the implementation of
> set_page_dirty_lock() suggests that it is fine to use set_page_dirty()
> instead (so not calling lock_page()).


no no, that's a misreading of the comment. Unless this xen/privcmd code has
somehow taken a reference on page->mapping->host (which I do *not* think is
the case), then it is still racy to call set_page_dirty() here. Instead,
set_page_dirty_lock() should be used.


>=20
> Only the transition from get_user_pages_fast() to pin_user_pages_fast()
> requires to use the locked version IMO.
>=20

That's a different misunderstanding. :) pin_user_pages*() APIs are meant to=
 be
functionally drop-in replacements for get_user_pages*(). Internally,
pin_user_pages*() functions do some additional tracking, but from a caller'=
s
perspective, it should look the same. In other words, there is nothing
about pin_user_pages_fast() that requires set_page_dirty_lock() upon releas=
e.
The reason set_page_dirty_lock() was chosen is that there are very few
(none at all?) call sites that need to release and dirty a page, that also =
meet
the requirements to safely call set_page_dirty().

That's why there is a unpin_user_pages_dirty_lock(), but there is not a
corresponding unpin_user_pages_dirty() call: the latter has not been requir=
ed
so far, even though the call site conversions are nearly done.


thanks,
--=20
John Hubbard
NVIDIA