linux-kernel - Re: [PATCH] mm, vmscan: Do not wait for page writeback for GFP

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, 1 Jul 2015 08:17:31 +0200
From:	Michal Hocko <mhocko@...e.cz>
To:	Nikolay Borisov <kernel@...p.com>,
	Johannes Weiner <hannes@...xchg.org>,
	Andrew Morton <akpm@...ux-foundation.org>
Cc:	Dave Chinner <david@...morbit.com>, Theodore Ts'o <tytso@....edu>,
	Marian Marinov <mm@...com>, linux-mm@...ck.org,
	LKML <linux-kernel@...r.kernel.org>, linux-ext4@...r.kernel.org
Subject: Re: [PATCH] mm, vmscan: Do not wait for page writeback for GFP_NOFS
 allocations

On Tue 30-06-15 17:17:17, Michal Hocko wrote:
[...]
> Hi,
> the issue has been reported http://marc.info/?l=linux-kernel&m=143522730927480.
> This obviously requires a patch ot make ext4_ext_grow_indepth call
> sb_getblk with the GFP_NOFS mask but that one makes sense on its own
> and Ted has mentioned he will push it. I haven't marked the patch for
> stable yet. This is the first time the issue has been reported and
> ext4 writeout code has changed considerably in 3.11 and I am not sure
> the issue was present before. e62e384e9da8 which has introduced the
> wait_on_page_writeback has been merged in 3.6 which is quite some time
> ago. If we go with stable I would suggest marking it for 3.11+ and it
> should obviously go with the ext4_ext_grow_indepth fix.

After Dave's additional explanation
(http://marc.info/?l=linux-ext4&m=143570521212215) it is clear that the
lack of __GFP_FS check was wrong from the very beginning. XFS is doing
the similar thing from before the e62e384e9da8 was merged. I guess we
were just lucky not to hit this problem sooner.

That being said I think the patch should be marked for stable and the
changelog updated:

As per David Chinner the xfs is doing similar thing since 2.6.15 already
so ext4 is not the only affected filesystem. Moreover he notes:
: For example: IO completion might require unwritten extent conversion
: which executes filesystem transactions and GFP_NOFS allocations. The
: writeback flag on the pages can not be cleared until unwritten
: extent conversion completes. Hence memory reclaim cannot wait on
: page writeback to complete in GFP_NOFS context because it is not
: safe to do so, memcg reclaim or otherwise.

Cc: stable # 3.6+
Fixes: e62e384e9da8 ("memcg: prevent OOM with too many dirty pages")

Andrew let me know whether I should repost the patch with the updated
changelog or you can take it from here.
-- 
Michal Hocko
SUSE Labs
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/