Re: [PATCH 4/4] anonvma: when setting up page->mapping, we need to pick the _oldest_ anonvma

[Minchan Kim <minchan.kim@gmail.com>]

On Tue, Apr 13, 2010 at 10:08 AM, Linus Torvalds
<torvalds@linux-foundation.org> wrote:
>
>
> On Tue, 13 Apr 2010, Johannes Weiner wrote:
>>
>> Would you mind pasting that nice description of the error case from your
>> other email into that changelog?  I skimmed over the description but when
>> I read this patch several hours later, I had to go back to that previous
>> email to fully make sense of it.
>
> It now looks like this..
>
>                Linus
> ---
> From: Linus Torvalds <torvalds@linux-foundation.org>
> Date: Mon, 12 Apr 2010 12:44:29 -0700
> Subject: [PATCH 4/4] anonvma: when setting up page->mapping, we need to pick the _oldest_ anonvma
>
> Otherwise we might be mapping in a page in a new mapping, but that page
> (through the swapcache) would later be mapped into an old mapping too.
> The page->mapping must be the case that works for everybody, not just
> the mapping that happened to page it in first.
>
> Here's the scenario:
>
>  - page gets allocated/mapped by process A. Let's call the anon_vma we
>   associate the page with 'A' to keep it easy to track.
>
>  - Process A forks, creating process B. The anon_vma in B is 'B', and has
>   a chain that looks like 'B' -> 'A'. Everything is fine.
>
>  - Swapping happens. The page (with mapping pointing to 'A') gets swapped
>   out (perhaps not to disk - it's enough to assume that it's just not
>   mapped any more, and lives entirely in the swap-cache)
>
>  - Process B pages it in, which goes like this:
>
>        do_swap_page ->
>          page = lookup_swap_cache(entry);
>         ...
>          set_pte_at(mm, address, page_table, pte);
>          page_add_anon_rmap(page, vma, address);
>
>   And think about what happens here!
>
>   In particular, what happens is that this will now be the "first"
>   mapping of that page, so page_add_anon_rmap() used to do
>
>        if (first)
>                __page_set_anon_rmap(page, vma, address);
>
>   and notice what anon_vma it will use? It will use the anon_vma for
>   process B!
>
>   What happens then? Trivial: process 'A' also pages it in (nothing
>   happens, it's not the first mapping), and then process 'B' execve's
>   or exits or unmaps, making anon_vma B go away.
>
>   End result: process A has a page that points to anon_vma B, but
>   anon_vma B does not exist any more.  This can go on forever.  Forget
>   about RCU grace periods, forget about locking, forget anything like
>   that.  The bug is simply that page->mapping points to an anon_vma
>   that was correct at one point, but was _not_ the one that was shared
>   by all users of that possible mapping.
>
> Changing it to always use the deepest anon_vma in the anonvma chain gets
> us to the safest model.
>
> This can be improved in certain cases: if we know the page is private to
> just this particular mapping (for example, it's a new page, or it is the
> only swapcache entry), we could pick the top (most specific) anon_vma.
>
> But that's a future optimization. Make it _work_ reliably first.
>
> Reviewed-by: Rik van Riel <riel@redhat.com>
> Acked-by: Johannes Weiner <hannes@cmpxchg.org>
> Tested-by: Borislav Petkov <bp@alien8.de> [ "What do you know, I think you fixed it!" ]
> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Reviewed-by: Minchan Kim <minchan.kim>

It was great hunting and was a chance to learn many things
from LKML smart guys.
I feel again about OSS's power and great procedure of linux evolution

Thanks for everybody.

-- 
Kind regards,
Minchan Kim