From: Trond Myklebust <trondmy@hammerspace.com>
To: "neil@brown.name" <neil@brown.name>
Cc: "snitzer@kernel.org" <snitzer@kernel.org>,
"linux-nfs@vger.kernel.org" <linux-nfs@vger.kernel.org>
Subject: Re: [PATCH] NFS/localio: Fix a race in nfs_local_open_fh()
Date: Wed, 23 Apr 2025 00:14:41 +0000 [thread overview]
Message-ID: <e3fedef06b142901d0c8fd962a38671052de9eb4.camel@hammerspace.com> (raw)
In-Reply-To: <174535921354.500591.6488717737987093498@noble.neil.brown.name>
On Wed, 2025-04-23 at 08:00 +1000, NeilBrown wrote:
> On Tue, 22 Apr 2025, trondmy@kernel.org wrote:
> > From: Trond Myklebust <trond.myklebust@hammerspace.com>
> >
> > Once the clp->cl_uuid.lock has been dropped, another CPU could come
> > in
> > and free the struct nfsd_file that was just added. To prevent that
> > from
> > happening, take the RCU read lock before dropping the spin lock.
>
> I think there is a race here but I think the better fix would be to
> use
> nfs_local_file_get() to get an extra reference earlier. That ensures
> we
> won't lose the nfsd_file.
>
Yes. This patch only ensures that the address pointed to by "nf" still
contains a valid object. It doesn't guarantee that the reference count
won't be zero, nor does it fix the issue that the initial dereference
to '*pnf' might be non-zero, yet point to an object that has a zero
refcount.
I'm not sure how much we really care, since those races should be
extremely rare and will only have a performance impact (as opposed to
the existing potential use-after-free).
That said, alternative patches are definitely welcome.
> I'm working on a patch in this area which I hope to post soon. It
> will
> address this.
>
> Thanks,
> NeilBrown
>
>
>
> >
> > Fixes: 86e00412254a ("nfs: cache all open LOCALIO nfsd_file(s) in
> > client")
> > Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
> > ---
> > fs/nfs/localio.c | 2 +-
> > 1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/fs/nfs/localio.c b/fs/nfs/localio.c
> > index 5c21caeae075..4ec952f9f47d 100644
> > --- a/fs/nfs/localio.c
> > +++ b/fs/nfs/localio.c
> > @@ -278,6 +278,7 @@ nfs_local_open_fh(struct nfs_client *clp, const
> > struct cred *cred,
> > new = __nfs_local_open_fh(clp, cred, fh, nfl,
> > mode);
> > if (IS_ERR(new))
> > return NULL;
> > + rcu_read_lock();
> > /* try to swap in the pointer */
> > spin_lock(&clp->cl_uuid.lock);
> > nf = rcu_dereference_protected(*pnf, 1);
> > @@ -287,7 +288,6 @@ nfs_local_open_fh(struct nfs_client *clp, const
> > struct cred *cred,
> > rcu_assign_pointer(*pnf, nf);
> > }
> > spin_unlock(&clp->cl_uuid.lock);
> > - rcu_read_lock();
> > }
> > nf = nfs_local_file_get(nf);
> > rcu_read_unlock();
> > --
> > 2.49.0
> >
> >
> >
--
Trond Myklebust
Linux NFS client maintainer, Hammerspace
trond.myklebust@hammerspace.com
next prev parent reply other threads:[~2025-04-23 0:14 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-04-22 22:00 [PATCH] NFS/localio: Fix a race in nfs_local_open_fh() NeilBrown
2025-04-23 0:14 ` Trond Myklebust [this message]
-- strict thread matches above, loose matches on Subject: below --
2025-04-21 18:52 trondmy
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=e3fedef06b142901d0c8fd962a38671052de9eb4.camel@hammerspace.com \
--to=trondmy@hammerspace.com \
--cc=linux-nfs@vger.kernel.org \
--cc=neil@brown.name \
--cc=snitzer@kernel.org \
/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