From: NeilBrown <neilb@suse.de>
To: Dave Chinner <david@fromorbit.com>
Cc: Jeff Layton <jlayton@redhat.com>,
linux-nfs@vger.kernel.org, Peter Zijlstra <peterz@infradead.org>,
netdev@vger.kernel.org, Ming Lei <ming.lei@canonical.com>,
linux-kernel@vger.kernel.org, xfs@oss.sgi.com,
linux-mm@kvack.org, Ingo Molnar <mingo@redhat.com>
Subject: Re: [PATCH/RFC 00/19] Support loop-back NFS mounts
Date: Thu, 17 Apr 2014 11:50:18 +1000 [thread overview]
Message-ID: <20140417115018.460345d0@notabene.brown> (raw)
In-Reply-To: <20140417012739.GU15995@dastard>
[-- Attachment #1: Type: text/plain, Size: 2570 bytes --]
On Thu, 17 Apr 2014 11:27:39 +1000 Dave Chinner <david@fromorbit.com> wrote:
> On Thu, Apr 17, 2014 at 10:20:48AM +1000, NeilBrown wrote:
> > A good example is the deadlock with the flush-* threads.
> > flush-* will lock a page, and then call ->writepage. If ->writepage
> > allocates memory it can enter reclaim, call ->releasepage on NFS, and block
> > waiting for a COMMIT to complete.
> > The COMMIT might already be running, performing fsync on that same file that
> > flush-* is flushing. It locks each page in turn. When it gets to the page
> > that flush-* has locked, it will deadlock.
>
> It's nfs_release_page() again....
>
> > In general, if nfsd is allowed to block on local filesystem, and local
> > filesystem is allowed to block on NFS, then a deadlock can happen.
> > We would need a clear hierarchy
> >
> > __GFP_NETFS > __GFP_FS > __GFP_IO
> >
> > for it to work. I'm not sure the extra level really helps a lot and it would
> > be a lot of churn.
>
> I think you are looking at this the wrong way - it's not the other
> filesystems that have to avoid memory reclaim recursion, it's the
> NFS client mount that is on loopback that needs to avoid recursion.
>
> IMO, the fix should be that the NFS client cannot block on messages sent to the NFSD
> on the same host during memory reclaim. That is, nfs_release_page()
> cannot send commit messages to the server if the server is on
> localhost. Instead, it just tells memory reclaim that it can't
> reclaim that page.
>
> If nfs_release_page() no longer blocks in memory reclaim, and all
> these nfsd-gets-blocked-in-GFP_KERNEL-memory-allocation recursion
> problems go away. Do the same for all the other memory reclaim
> operations in the NFS client, and you've got a solution that should
> work without needing to walk all over the rest of the kernel....
Maybe.
It is nfs_release_page() today. I wonder if it could be other things another
day. I want to be sure I have a solution that really makes sense.
However ... the thing that nfs_release_page is doing it sending a COMMIT to
tell the server to flush to stable storage. It does that so that if the
server crashes, then the client can re-send.
Of course when it is a loop-back mount the client is the server so the COMMIT
is completely pointless. If the client notices that it is sending a COMMIT
to itself, it can simply assume a positive reply.
You are right, that would make the patch set a lot less intrusive. I'll give
it some serious thought - thanks.
NeilBrown
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 828 bytes --]
next prev parent reply other threads:[~2014-04-17 1:50 UTC|newest]
Thread overview: 52+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-04-16 4:03 [PATCH/RFC 00/19] Support loop-back NFS mounts NeilBrown
2014-04-16 4:03 ` [PATCH 12/19] NET: set PF_FSTRANS while holding rtnl_lock NeilBrown
2014-04-16 4:03 ` [PATCH 07/19] nfsd and VM: use PF_LESS_THROTTLE to avoid throttle in shrink_inactive_list NeilBrown
2014-04-16 4:03 ` [PATCH 02/19] lockdep: lockdep_set_current_reclaim_state should save old value NeilBrown
2014-04-16 4:03 ` [PATCH 09/19] XFS: ensure xfs_file_*_read cannot deadlock in memory allocation NeilBrown
2014-04-16 6:04 ` Dave Chinner
2014-04-16 6:27 ` NeilBrown
2014-04-16 6:31 ` Dave Chinner
2014-04-16 4:03 ` [PATCH 10/19] NET: set PF_FSTRANS while holding sk_lock NeilBrown
2014-04-16 5:13 ` Eric Dumazet
2014-04-16 5:47 ` NeilBrown
2014-04-16 13:00 ` David Miller
2014-04-17 2:38 ` NeilBrown
2014-04-16 4:03 ` [PATCH 08/19] Set PF_FSTRANS while write_cache_pages calls ->writepage NeilBrown
2014-04-16 4:03 ` [PATCH 14/19] driver core: set PF_FSTRANS while holding gdp_mutex NeilBrown
2014-04-16 4:03 ` [PATCH 04/19] Make effect of PF_FSTRANS to disable __GFP_FS universal NeilBrown
2014-04-16 5:37 ` Dave Chinner
2014-04-16 6:17 ` NeilBrown
2014-04-17 1:03 ` NeilBrown
2014-04-17 4:41 ` Dave Chinner
2014-04-16 4:03 ` [PATCH 13/19] MM: set PF_FSTRANS while allocating per-cpu memory to avoid deadlock NeilBrown
2014-04-16 5:49 ` Dave Chinner
2014-04-16 6:22 ` NeilBrown
2014-04-16 6:30 ` Dave Chinner
2014-04-16 4:03 ` [PATCH 03/19] lockdep: improve scenario messages for RECLAIM_FS errors NeilBrown
2014-04-16 7:22 ` Peter Zijlstra
2014-04-16 4:03 ` [PATCH 01/19] Promote current_{set, restore}_flags_nested from xfs to global NeilBrown
2014-04-16 4:03 ` [PATCH 11/19] FS: set PF_FSTRANS while holding mmap_sem in exec.c NeilBrown
2014-04-16 4:03 ` [PATCH 06/19] nfsd: set PF_FSTRANS for nfsd threads NeilBrown
2014-04-16 7:28 ` Peter Zijlstra
2014-04-16 4:03 ` [PATCH 05/19] SUNRPC: track whether a request is coming from a loop-back interface NeilBrown
2014-04-16 14:47 ` Jeff Layton
2014-04-16 23:25 ` NeilBrown
2014-04-16 4:03 ` [PATCH 19/19] XFS: set PF_FSTRANS while ilock is held in xfs_free_eofblocks NeilBrown
2014-04-16 6:18 ` Dave Chinner
2014-04-16 4:03 ` [PATCH 15/19] nfsd: set PF_FSTRANS when client_mutex is held NeilBrown
2014-04-16 4:03 ` [PATCH 16/19] VFS: use GFP_NOFS rather than GFP_KERNEL in __d_alloc NeilBrown
2014-04-16 6:25 ` Dave Chinner
2014-04-16 6:49 ` NeilBrown
2014-04-16 9:00 ` Dave Chinner
2014-04-17 0:51 ` NeilBrown
2014-04-17 5:58 ` Dave Chinner
2014-04-16 4:03 ` [PATCH 18/19] nfsd: set PF_FSTRANS during nfsd4_do_callback_rpc NeilBrown
2014-04-16 4:03 ` [PATCH 17/19] VFS: set PF_FSTRANS while namespace_sem is held NeilBrown
2014-04-16 4:46 ` Al Viro
2014-04-16 5:52 ` NeilBrown
[not found] ` <20140416155230.4d02e4b9-wvvUuzkyo1EYVZTmpyfIwg@public.gmane.org>
2014-04-16 16:37 ` Al Viro
2014-04-16 14:42 ` [PATCH/RFC 00/19] Support loop-back NFS mounts Jeff Layton
2014-04-17 0:20 ` NeilBrown
2014-04-17 1:27 ` Dave Chinner
2014-04-17 1:50 ` NeilBrown [this message]
2014-04-17 4:23 ` Dave Chinner
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20140417115018.460345d0@notabene.brown \
--to=neilb@suse.de \
--cc=david@fromorbit.com \
--cc=jlayton@redhat.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=linux-nfs@vger.kernel.org \
--cc=ming.lei@canonical.com \
--cc=mingo@redhat.com \
--cc=netdev@vger.kernel.org \
--cc=peterz@infradead.org \
--cc=xfs@oss.sgi.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).