linux-kernel - Re: [PATCH v1 2/6] mm/page_isolation: don't dump_page(NULL) in set_migratetype

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:   Wed, 29 Jul 2020 20:08:56 +0200
From:   David Hildenbrand <david@...hat.com>
To:     Mike Kravetz <mike.kravetz@...cle.com>,
        linux-kernel@...r.kernel.org
Cc:     virtualization@...ts.linux-foundation.org, linux-mm@...ck.org,
        Andrew Morton <akpm@...ux-foundation.org>,
        Michal Hocko <mhocko@...e.com>,
        "Michael S.Tsirkin" <mst@...hat.com>
Subject: Re: [PATCH v1 2/6] mm/page_isolation: don't dump_page(NULL) in
 set_migratetype_isolate()

On 29.07.20 19:31, Mike Kravetz wrote:
> On 6/30/20 7:26 AM, David Hildenbrand wrote:
>> Right now, if we have two isolations racing, we might trigger the
>> WARN_ON_ONCE() and to dump_page(NULL), dereferencing NULL. Let's just
>> return directly.
> 
> Just curious, what call path has the WARN_ON_ONCE()/dump_page(NULL)?

See below, two set_migratetype_isolate() caller racing.

> 
>>
>> In the future, we might want to report -EAGAIN to the caller instead, as
>> this could indicate a temporary isolation failure only.
>>
>> Cc: Andrew Morton <akpm@...ux-foundation.org>
>> Cc: Michal Hocko <mhocko@...e.com>
>> Cc: Michael S. Tsirkin <mst@...hat.com>
>> Signed-off-by: David Hildenbrand <david@...hat.com>
> 
> Hi David,
> 
> That 'return -EAGAIN' was added as a sort of synchronization mechanism.
> See commit message for 2c7452a075d4d.  Before adding the 'return -EAGAIN',
> I could create races which would abandon isolated pageblocks.  Repeating
> those races over and over would result in a good chunk of system memory
> being isolated and unusable.

It's actually -EBUSY, it should maybe later be changed to -EAGAIN (see
comment), so caller can decide to retry immediately. Other discussion.

> 
> Admittedly, these races are rare and I had to work really hard to produce
> them.  I'll try to find my testing mechanism.  My concern is reintroducing
> this abandoning of pageblocks.  I have not looked further in your series
> to see if this potentially addressed later.  If not, then we should not
> remove the return code.
> 

Memory offlining could race with alloc_contig_range(), e.g., called when
allocating gigantic pages, or when virtio-mem tries to unplug memory.
The latter two could also race.

We are getting more alloc_contig_range() users, which is why these races
will become more relevant.

I have no clue what you mean with "reintroducing this abandoning of
pageblocks". All this patch is changing is not doing the dump_page() -
or am I missing something important?


-- 
Thanks,

David / dhildenb