Received-SPF: pass (google.com: domain of linux-nfs+bounces-837-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:45d1:ec00::1 as permitted sender) client-ip=2604:1380:45d1:ec00::1;
Precedence: bulk
MIME-Version: 1.0
References: <20231228201510.985235-1-trondmy@kernel.org> <CAOQ4uxiCf=FWtZWw2uLRmfPvgSxsnmqZC6A+FQgQs=MBQwA30w@mail.gmail.com>
 <a68a47d136decfeb6c1cc7959353ae51aca47ae7.camel@hammerspace.com>
In-Reply-To: <a68a47d136decfeb6c1cc7959353ae51aca47ae7.camel@hammerspace.com>
From: Amir Goldstein <amir73il@gmail.com>
Date: Fri, 29 Dec 2023 19:54:16 +0200
Message-ID: <CAOQ4uxiOZobN76OKB-VBNXWeFKVwLW_eK5QtthGyYzWU9mjb7Q@mail.gmail.com>
Subject: Re: [PATCH] knfsd: fix the fallback implementation of the get_name
 export operation
To: Trond Myklebust <trondmy@hammerspace.com>
Cc: "linux-nfs@vger.kernel.org" <linux-nfs@vger.kernel.org>, 
	"linux-fsdevel@vger.kernel.org" <linux-fsdevel@vger.kernel.org>, 
	"viro@zeniv.linux.org.uk" <viro@zeniv.linux.org.uk>, "chuck.lever@oracle.com" <chuck.lever@oracle.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

On Fri, Dec 29, 2023 at 5:22=E2=80=AFPM Trond Myklebust <trondmy@hammerspac=
e.com> wrote:
>
> On Fri, 2023-12-29 at 07:46 +0200, Amir Goldstein wrote:
> > [CC: fsdevel, viro]
> >
> > On Thu, Dec 28, 2023 at 10:22=E2=80=AFPM <trondmy@kernel.org> wrote:
> > >
> > > From: Trond Myklebust <trond.myklebust@hammerspace.com>
> > >
> > > The fallback implementation for the get_name export operation uses
> > > readdir() to try to match the inode number to a filename. That
> > > filename
> > > is then used together with lookup_one() to produce a dentry.
> > > A problem arises when we match the '.' or '..' entries, since that
> > > causes lookup_one() to fail. This has sometimes been seen to occur
> > > for
> > > filesystems that violate POSIX requirements around uniqueness of
> > > inode
> > > numbers, something that is common for snapshot directories.
> >
> > Ouch. Nasty.
> >
> > Looks to me like the root cause is "filesystems that violate POSIX
> > requirements around uniqueness of inode numbers".
> > This violation can cause any of the parent's children to wrongly
> > match
> > get_name() not only '.' and '..' and fail the d_inode sanity check
> > after
> > lookup_one().
> >
> > I understand why this would be common with parent of snapshot dir,
> > but the only fs that support snapshots that I know of (btrfs,
> > bcachefs)
> > do implement ->get_name(), so which filesystem did you encounter
> > this behavior with? can it be fixed by implementing a snapshot
> > aware ->get_name()?
>
> NFS (i.e. re-exporting NFS).
>

Ah.

> Why do you not want a fix in the generic code?
>

I do not object to your fix at all.
I only objected to the Fixes tag.
I was just pointing out that this is not a complete solution.
A decode of an NFS (re-exported) file handle could fail if
get_name() iterates the parent of a snapshot root dir
and finds a false match (which is not "." nor "..") before it
finds the snapshot subdir name.

It may be solved by nfs_get_name() which does not stop when
if finds an ino match but checks further, but I don't know nfs re-export
to know what else could be checked.

Anyway, for this patch, without the Fixes tag, feel free to add:
Acked-by: Amir Goldstein <amir73il@gmail.com>

I'd prefer the use of is_dot_dotdot(), but I do not insist.

Thanks and a happy new year!
Amir.