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;
MIME-Version: 1.0
References: <X97pprdcRXusLGnq@google.com> <DDA15360-D6D4-46A8-95A4-5EE34107A407@gmail.com>
 <20201221172711.GE6640@xz-x1> <76B4F49B-ED61-47EA-9BE4-7F17A26B610D@gmail.com>
 <X+D0hTZCrWS3P5Pi@google.com> <CAHk-=wg_UBuo7ro1fpEGkMyFKA1+PxrE85f9J_AhUfr-nJPpLQ@mail.gmail.com>
 <9E301C7C-882A-4E0F-8D6D-1170E792065A@gmail.com> <CAHk-=wg-Y+svNy3CDkJjj0X_CJkSbpERLg64-Vqwq5u7SC4z0g@mail.gmail.com>
 <X+ESkna2z3WjjniN@google.com> <1FCC8F93-FF29-44D3-A73A-DF943D056680@gmail.com>
 <20201221223041.GL6640@xz-x1> <CAHk-=wh-bG4thjXUekLtrCg8FRrdWjtT40ibXXLSm_hzQG8eOw@mail.gmail.com>
 <CALCETrV=8tY7h=aaudWBEn-MJnNkm2wz5qjH49SYqwkjYTpOaA@mail.gmail.com> <CAHk-=wj=CcOHQpG0cUGfoMCt2=Uaifpqq-p-mMOmW8XmrBn4fQ@mail.gmail.com>
In-Reply-To: <CAHk-=wj=CcOHQpG0cUGfoMCt2=Uaifpqq-p-mMOmW8XmrBn4fQ@mail.gmail.com>
From:   Andy Lutomirski <luto@kernel.org>
Date:   Tue, 22 Dec 2020 12:19:11 -0800
Message-ID: <CALCETrVdT4jSa3UKU8rQ2KO8CE1Jq2r23ZwuqmynAo6kYUwA=Q@mail.gmail.com>
Subject: Re: [PATCH] mm/userfaultfd: fix memory corruption due to writeprotect
To:     Linus Torvalds <torvalds@linux-foundation.org>
Cc:     Andy Lutomirski <luto@kernel.org>, Peter Xu <peterx@redhat.com>,
        Nadav Amit <nadav.amit@gmail.com>, Yu Zhao <yuzhao@google.com>,
        Andrea Arcangeli <aarcange@redhat.com>,
        linux-mm <linux-mm@kvack.org>,
        lkml <linux-kernel@vger.kernel.org>,
        Pavel Emelyanov <xemul@openvz.org>,
        Mike Kravetz <mike.kravetz@oracle.com>,
        Mike Rapoport <rppt@linux.vnet.ibm.com>,
        stable <stable@vger.kernel.org>,
        Minchan Kim <minchan@kernel.org>,
        Will Deacon <will@kernel.org>,
        Peter Zijlstra <peterz@infradead.org>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Mon, Dec 21, 2020 at 8:16 PM Linus Torvalds
<torvalds@linux-foundation.org> wrote:
>
> On Mon, Dec 21, 2020 at 7:19 PM Andy Lutomirski <luto@kernel.org> wrote:
> >
> > Ugh, this is unpleasantly complicated.
>
> What I *should* have said is that *because* userfaultfd is changing
> the VM layout, it should always act as if it had to take the mmap_sem
> for writing, and that the RW->RO change is an example of when
> downgrading that write-lock to a read lock is simply not valid -
> because it basically breaks the rules about what a lookup (ie a read)
> can do.
>
> A lookup can never turn a writable page non-writable. A lookup -
> through COW - _can_ turn a read-only page writable. So that's why
> "RO->RW" can be ok under the read lock, but RW->RO is not.
>
> Does that clarify the situation when I phrase it that way instead?

It's certainly more clear, but I'm still not thrilled with a bunch of
functions in different files maintained by different people that
cooperate using an undocumented lockless protocol.  It makes me
nervous from a maintainability standpoint even if the code is, in
fact, entirely correct.

But I didn't make my question clear either: when I asked about
mark_screen_rdonly(), I wasn't asking about locking or races at all.
mark_screen_rdonly() will happily mark *any* PTE readonly.  I can
easily believe that writable mappings of non-shared mappings all
follow the same CoW rules in the kernel and that, assuming all the
locking is correct, marking them readonly is conceptually okay.  But
shared mappings are a whole different story.  Is it actually the case
that all, or even most, drivers and other sources of shared mappings
will function correctly if one of their PTEs becomes readonly behind
their back?  Or maybe there are less bizarre ways for this to happen
without vm86 shenanigans and this is completely safe after all.

P.S. This whole mess reminds me that I need to take a closer look at
the upcoming SHSTK code.  Shadow stack pages violate all common sense
about how the various PTE bits work.