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: <CABi2SkWx_BnEHzGqqqbDMJi+vi-5a7XkQUCkyesN5PUtk23SgQ@mail.gmail.com>
 <CABi2SkXw6ZD-M1ZrcXNL7abtM=RzQXv716PPM_k=1Tay=5rUFA@mail.gmail.com> <ZIjOlU5EfVNt6NRU@x1n>
In-Reply-To: <ZIjOlU5EfVNt6NRU@x1n>
From:   Jeff Xu <jeffxu@chromium.org>
Date:   Tue, 13 Jun 2023 14:29:14 -0700
Message-ID: <CABi2SkXE4pUhHucZ_c-_4Ux-VcLKic0+HY_DN2wUEC6DGkDvQQ@mail.gmail.com>
Subject: Re: inconsistence in mprotect_fixup mlock_fixup madvise_update_vma
To:     Peter Xu <peterx@redhat.com>
Cc:     linux-mm@kvack.org, linux-hardening@vger.kernel.org,
        Liam.Howlett@oracle.com, zhangpeng.00@bytedance.com,
        akpm@linux-foundation.org, koct9i@gmail.com, david@redhat.com,
        ak@linux.intel.com, hughd@google.com, emunson@akamai.com,
        rppt@linux.ibm.com, aarcange@redhat.com,
        linux-kernel@vger.kernel.org, Lorenzo Stoakes <lstoakes@gmail.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

Hello Peter,

Thanks for responding.

On Tue, Jun 13, 2023 at 1:16=E2=80=AFPM Peter Xu <peterx@redhat.com> wrote:
>
> Hi, Jeff,
>
> On Tue, Jun 13, 2023 at 08:26:26AM -0700, Jeff Xu wrote:
> > + more ppl to the list.
> >
> > On Mon, Jun 12, 2023 at 6:04=E2=80=AFPM Jeff Xu <jeffxu@chromium.org> w=
rote:
> > >
> > > Hello,
> > >
> > > There seems to be inconsistency in different VMA fixup
> > > implementations, for example:
> > > mlock_fixup will skip VMA that is hugettlb, etc, but those checks do
> > > not exist in mprotect_fixup and madvise_update_vma. Wouldn't this be =
a
> > > problem? the merge/split skipped by mlock_fixup, might get acted on i=
n
> > > the madvice/mprotect case.
> > >
> > > mlock_fixup currently check for
> > > if (newflags =3D=3D oldflags || (oldflags & VM_SPECIAL) ||
> > > is_vm_hugetlb_page(vma) || vma =3D=3D get_gate_vma(current->mm) ||
> > > vma_is_dax(vma) || vma_is_secretmem(vma))
>
> The special handling you mentioned in mlock_fixup mostly makes sense to m=
e.
>
> E.g., I think we can just ignore mlock a hugetlb page if it won't be
> swapped anyway.
>
> Do you encounter any issue with above?
>
> > > Should there be a common function to handle VMA merge/split ?
>
> IMHO vma_merge() and split_vma() are the "common functions".  Copy Lorenz=
o
> as I think he has plan to look into the interface to make it even easier =
to
> use.
>
The mprotect_fixup doesn't have the same check as mlock_fixup. When
userspace calls mlock(), two VMAs might not merge or split because of
vma_is_secretmem check, However, when user space calls mprotect() with
the same address range, it will merge/split.  If mlock() is doing the
right thing to merge/split the VMAs, then mprotect() is not ?

Also skipping merge of VMA might be OK, but skipping split doesn't,
wouldn't that cause inconsistent between vma->vm_flags and what is
provisioned in the page ?

Thanks
-Jeff Xu


> --
> Peter Xu
>