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: <20230511182426.1898675-1-axelrasmussen@google.com>
 <CADrL8HXFiTL-RDnETS2BUg_qH8CvcCMZiX-kutsrS1-8Uy25=w@mail.gmail.com>
 <ZGVRUeCWr8209m8d@x1n> <ZGVTMnVKNcQDM0x4@x1n> <CAJHvVcgXynHcuoS6eCfOAB2SgzqYy_zMGrRMR2kFuxOtSdUwvQ@mail.gmail.com>
 <CACw3F52MNOVv6KA5n7wRYDT2ujwYkco=aYngbo-zGA3zW1yq+w@mail.gmail.com>
 <ZGZMtK6PzoTuLZ1b@x1n> <CAJHvVcgcYPu-G3RDVrkrM_J48NUiUY0SH0G1sd+=X9BDgnQEuQ@mail.gmail.com>
 <32fdc2c8-b86b-92f3-1d5e-64db6be29126@redhat.com> <CACw3F50qvf13-fUx4XrL1jkhbo2mQ5sPV=E_i7_Gt2NaWXJfnQ@mail.gmail.com>
 <ZGehyTCtAtTneiE8@x1n>
In-Reply-To: <ZGehyTCtAtTneiE8@x1n>
From:   Axel Rasmussen <axelrasmussen@google.com>
Date:   Fri, 19 May 2023 10:32:13 -0700
Message-ID: <CAJHvVcgVcGigruVdCyOZ2gyrOMboHjuMBwyx9+9LFiv_xH2Q5A@mail.gmail.com>
Subject: Re: [PATCH 1/3] mm: userfaultfd: add new UFFDIO_SIGBUS ioctl
To:     Peter Xu <peterx@redhat.com>
Cc:     Jiaqi Yan <jiaqiyan@google.com>,
        David Hildenbrand <david@redhat.com>,
        James Houghton <jthoughton@google.com>,
        Alexander Viro <viro@zeniv.linux.org.uk>,
        Andrew Morton <akpm@linux-foundation.org>,
        Christian Brauner <brauner@kernel.org>,
        Hongchen Zhang <zhanghongchen@loongson.cn>,
        Huang Ying <ying.huang@intel.com>,
        "Liam R. Howlett" <Liam.Howlett@oracle.com>,
        Miaohe Lin <linmiaohe@huawei.com>,
        "Mike Rapoport (IBM)" <rppt@kernel.org>,
        Nadav Amit <namit@vmware.com>,
        Naoya Horiguchi <naoya.horiguchi@nec.com>,
        Shuah Khan <shuah@kernel.org>,
        ZhangPeng <zhangpeng362@huawei.com>,
        linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org,
        linux-mm@kvack.org, linux-kselftest@vger.kernel.org,
        Anish Moorthy <amoorthy@google.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Fri, May 19, 2023 at 9:20=E2=80=AFAM Peter Xu <peterx@redhat.com> wrote:
>
> Hi, Jiaqi,
>
> On Fri, May 19, 2023 at 08:04:09AM -0700, Jiaqi Yan wrote:
> > I don't think CAP_ADMIN is something we can work around: a VMM must be
> > a good citizen to avoid introducing any vulnerability to the host or
> > guest.
> >
> > On the other hand, "Userfaults allow the implementation of on-demand
> > paging from userland and more generally they allow userland to take
> > control of various memory page faults, something otherwise only the
> > kernel code could do." [3]. I am not familiar with the UFFD internals,
> > but our use case seems to match what UFFD wants to provide: without
> > affecting the whole world, give a specific userspace (without
> > CAP_ADMIN) the ability to handle page faults (indirectly emulate a
> > HWPOISON page (in my mind I treat it as SetHWPOISON(page) +
> > TestHWPOISON(page) operation in kernel's PF code)). So is it fair to
> > say what Axel provided here is "provide !ADMIN somehow"?
> >
> > [3]https://docs.kernel.org/admin-guide/mm/userfaultfd.html
>
> Userfault keywords on "user", IMHO.  We don't strictly need userfault to
> resolve anything regarding CAP_ADMIN problems.  MADV_DONTNEED also dosn't
> need CAP_ADMIN, same to any new madvise() if we want to make it useful fo=
r
> injecting poisoned ptes with !ADMIN and limit it within current->mm.
>
> But I think you're right that userfaultfd always tried to avoid having
> ADMIN and keep everything within its own scope of permissions.
>
> So again, totally no objection on make it uffd specific for now if you gu=
ys
> are all happy with it, but just to be clear that it's (to me) mostly for
> avoiding another WAKE, and afaics that's not really for solving the ADMIN
> issue here.

How about this plan:

Since the concrete use case we have (postcopy live migration) is
UFFD-specific, let's leave it as a UFFDIO_* operation for now.

If in the future we come up with a non-UFFD use case, we can add a new
MADV_* which does this operation at that point. From my perspective
they could even share most of the same implementation code.

I don't think it's a big problem keeping the UFFDIO_* version too at
that point, because it still provides some (perhaps small) value:

- Combines the operation + waking into one syscall
- It allows us to support additional UFFD flags which modify / extend
the operation in UFFD-specific ways, if we want to add those in the
future

Seem reasonable?

If so, I'll send a v2 with documentation updates.

>
> Thanks,
>
> --
> Peter Xu
>