From: Amir Goldstein <amir73il@gmail.com>
To: Christian Brauner <christian.brauner@ubuntu.com>
Cc: Jan Kara <jack@suse.cz>,
linux-fsdevel <linux-fsdevel@vger.kernel.org>,
Linux API <linux-api@vger.kernel.org>,
"J. Bruce Fields" <bfields@fieldses.org>
Subject: Re: open_by_handle_at() in userns
Date: Thu, 8 Apr 2021 18:34:01 +0300 [thread overview]
Message-ID: <CAOQ4uxi=jKRxLoRs=fNh96zAzqeoG5OZrOi6i7m7Ooy8b5zxBA@mail.gmail.com> (raw)
In-Reply-To: <20210408125530.gnv5hqcmgewklypn@wittgenstein>
On Thu, Apr 8, 2021 at 3:55 PM Christian Brauner
<christian.brauner@ubuntu.com> wrote:
>
> On Thu, Apr 08, 2021 at 02:44:47PM +0300, Amir Goldstein wrote:
> > > One thing your patch
> > >
> > > commit ea31e84fda83c17b88851de399f76f5d9fc1abf4
> > > Author: Amir Goldstein <amir73il@gmail.com>
> > > Date: Sat Mar 20 12:58:12 2021 +0200
> > >
> > > fs: allow open by file handle inside userns
> > >
> > > open_by_handle_at(2) requires CAP_DAC_READ_SEARCH in init userns,
> > > where most filesystems are mounted.
> > >
> > > Relax the requirement to allow a user with CAP_DAC_READ_SEARCH
> > > inside userns to open by file handle in filesystems that were
> > > mounted inside that userns.
> > >
> > > In addition, also allow open by handle in an idmapped mount, which is
> > > mapped to the userns while verifying that the returned open file path
> > > is under the root of the idmapped mount.
> > >
> > > This is going to be needed for setting an fanotify mark on a filesystem
> > > and watching events inside userns.
> > >
> > > Signed-off-by: Amir Goldstein <amir73il@gmail.com>
> > >
> > > Requires fs/exportfs/expfs.c to be made idmapped mounts aware.
> > > open_by_handle_at() uses exportfs_decode_fh() which e.g. has the
> > > following and other callchains:
> > >
> > > exportfs_decode_fh()
> > > -> exportfs_decode_fh_raw()
> > > -> lookup_one_len()
> > > -> inode_permission(mnt_userns, ...)
> > >
> > > That's not a huge problem though I did all these changes for the
> > > overlayfs support for idmapped mounts I have in a branch from an earlier
> > > version of the idmapped mounts patchset. Basically lookup_one_len(),
> > > lookup_one_len_unlocked(), and lookup_positive_unlocked() need to take
> > > the mnt_userns into account. I can rebase my change and send it for
> > > consideration next cycle. If you can live without the
> > > open_by_handle_at() support for now in this patchset (Which I think you
> > > said you could.) then it's not a blocker either. Sorry for the
> > > inconvenience.
> > >
> >
> > Christian,
> >
> > I think making exportfs_decode_fh() idmapped mount aware is not
> > enough, because when a dentry alias is found in dcache, none of
> > those lookup functions are called.
> >
> > I think we will also need something like this:
> > https://github.com/amir73il/linux/commits/fhandle_userns
> >
> > I factored-out a helper from nfsd_apcceptable() which implements
> > the "subtree_check" nfsd logic and uses it for open_by_handle_at().
> >
> > I've also added a small patch to name_to_handle_at() with a UAPI
> > change that could make these changes usable by userspace nfs
> > server inside userns, but I have no demo nor tests for that and frankly,
> > I have little incentive to try and promote this UAPI change without
> > anybody asking for it...
>
> Ah, at first I was confused about why this would matter but it matters
> because nfsd already implements a check of that sort directly in nfsd
> independent of idmapped mounts:
> https://github.com/amir73il/linux/commit/4bef9ff1718935b7b42afbae71cfaab7770e8436
>
The check is needed for slightly different reasons.
nfsd "subtree_check" feature explicitly meant to forbid access in case
file was moved "out of reach", for example, out of the export path.
Note the nfsd "subtree_check" affects both file handle encoding
(i.e. "connectable") and file handle decoding (i.e. nfsd_acceptable()).
open_by_handle_at() in idmapped mount needs to verify that the ancestry
inode owners can be mapped to the userns, because we already checked
that user has CAP_DAC_READ_SEARCH in userns, but it's nicer to do
the full inode_permission() check IMO.
> Afaict, an nfs server can't be mounted inside of userns right now. That
> is something that folks from Netflix and from Kinvolk have been
> interested in enabling. They also want the ability to use idmapped
> mounts + nfs. Understandable that you don't want to drive this of
> course. I'll sync with them about this.
>
> Independent of that, I thought our last understanding was that you
> wouldn't need to handle open_by_handle_at() for now.
>
I don't need it. But I realized that the fanotify_userns demo branch
I provided you is buggy in terms of security, so I wanted to give you
(or whoever wants to pursue this) a better reference.
It was one of those things that are easier to code than to explain ;-)
Thanks,
Amir.
next prev parent reply other threads:[~2021-04-08 15:34 UTC|newest]
Thread overview: 34+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-03-28 15:56 [RFC][PATCH] fanotify: allow setting FAN_CREATE in mount mark mask Amir Goldstein
2021-03-30 7:31 ` Christian Brauner
2021-03-30 9:31 ` Amir Goldstein
2021-03-30 16:24 ` Amir Goldstein
2021-03-31 10:08 ` Christian Brauner
2021-03-31 10:57 ` Amir Goldstein
2021-04-08 11:44 ` open_by_handle_at() in userns Amir Goldstein
2021-04-08 12:55 ` Christian Brauner
2021-04-08 14:15 ` J. Bruce Fields
2021-04-08 15:54 ` Amir Goldstein
2021-04-08 16:08 ` J. Bruce Fields
2021-04-08 16:48 ` Frank Filz
2021-04-08 15:34 ` Amir Goldstein [this message]
2021-04-08 15:41 ` Christian Brauner
2021-03-30 12:12 ` [RFC][PATCH] fanotify: allow setting FAN_CREATE in mount mark mask Christian Brauner
2021-03-30 12:33 ` Amir Goldstein
2021-03-30 12:53 ` Christian Brauner
2021-03-30 12:55 ` Christian Brauner
2021-03-30 13:54 ` Amir Goldstein
2021-03-30 14:17 ` Christian Brauner
2021-03-30 14:56 ` Amir Goldstein
2021-03-31 9:46 ` Christian Brauner
2021-03-31 11:29 ` Amir Goldstein
2021-03-31 12:17 ` Christian Brauner
2021-03-31 12:59 ` Amir Goldstein
2021-03-31 12:54 ` Jan Kara
2021-03-31 14:06 ` Amir Goldstein
2021-03-31 20:59 ` fsnotify path hooks Amir Goldstein
2021-04-01 10:29 ` Jan Kara
2021-04-01 14:18 ` Amir Goldstein
2021-04-02 8:20 ` Amir Goldstein
2021-04-06 8:35 ` Jan Kara
2021-03-31 13:06 ` [RFC][PATCH] fanotify: allow setting FAN_CREATE in mount mark mask J. Bruce Fields
2021-03-30 12:20 ` Amir Goldstein
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='CAOQ4uxi=jKRxLoRs=fNh96zAzqeoG5OZrOi6i7m7Ooy8b5zxBA@mail.gmail.com' \
--to=amir73il@gmail.com \
--cc=bfields@fieldses.org \
--cc=christian.brauner@ubuntu.com \
--cc=jack@suse.cz \
--cc=linux-api@vger.kernel.org \
--cc=linux-fsdevel@vger.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;
as well as URLs for NNTP newsgroup(s).