linux-kernel - Re: [PATCH] mm/userfaultfd: fix memory corruption due to writeprotect

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Tue, 22 Dec 2020 11:20:21 -0800
From:   Nadav Amit <nadav.amit@...il.com>
To:     Yu Zhao <yuzhao@...gle.com>
Cc:     Peter Zijlstra <peterz@...radead.org>,
        Minchan Kim <minchan@...nel.org>,
        Linus Torvalds <torvalds@...ux-foundation.org>,
        Peter Xu <peterx@...hat.com>,
        Andrea Arcangeli <aarcange@...hat.com>,
        linux-mm <linux-mm@...ck.org>,
        lkml <linux-kernel@...r.kernel.org>,
        Pavel Emelyanov <xemul@...nvz.org>,
        Mike Kravetz <mike.kravetz@...cle.com>,
        Mike Rapoport <rppt@...ux.vnet.ibm.com>,
        stable <stable@...r.kernel.org>,
        Andy Lutomirski <luto@...nel.org>,
        Will Deacon <will@...nel.org>
Subject: Re: [PATCH] mm/userfaultfd: fix memory corruption due to writeprotect

> On Dec 22, 2020, at 10:30 AM, Yu Zhao <yuzhao@...gle.com> wrote:
> 
> On Tue, Dec 22, 2020 at 04:40:32AM -0800, Nadav Amit wrote:
>>> On Dec 21, 2020, at 1:24 PM, Yu Zhao <yuzhao@...gle.com> wrote:
>>> 
>>> On Mon, Dec 21, 2020 at 12:26:22PM -0800, Linus Torvalds wrote:
>>>> On Mon, Dec 21, 2020 at 12:23 PM Nadav Amit <nadav.amit@...il.com> wrote:
>>>>> Using mmap_write_lock() was my initial fix and there was a strong pushback
>>>>> on this approach due to its potential impact on performance.
>>>> 
>>>> From whom?
>>>> 
>>>> Somebody who doesn't understand that correctness is more important
>>>> than performance? And that userfaultfd is not the most important part
>>>> of the system?
>>>> 
>>>> The fact is, userfaultfd is CLEARLY BUGGY.
>>>> 
>>>>         Linus
>>> 
>>> Fair enough.
>>> 
>>> Nadav, for your patch (you might want to update the commit message).
>>> 
>>> Reviewed-by: Yu Zhao <yuzhao@...gle.com>
>>> 
>>> While we are all here, there is also clear_soft_dirty() that could
>>> use a similar fix…
>> 
>> Just an update as for why I have still not sent v2: I fixed
>> clear_soft_dirty(), created a reproducer, and the reproducer kept failing.
>> 
>> So after some debugging, it appears that clear_refs_write() does not flush
>> the TLB. It indeed calls tlb_finish_mmu() but since 0758cd830494
>> ("asm-generic/tlb: avoid potential double flush”), tlb_finish_mmu() does not
>> flush the TLB since there is clear_refs_write() does not call to
>> __tlb_adjust_range() (unless there are nested TLBs are pending).
> 
> Sorry Nadav, I assumed you knew this existing problem fixed by:
> https://patchwork.kernel.org/project/linux-mm/cover/20201210121110.10094-1-will@kernel.org/
> 

Thanks, Yu! For some reason I assumed it was already upstreamed and did not
look back (yet if I was cc’d on v2…)

Yet, something still goes bad. Debugging.