From: Jason Gunthorpe <jgg@nvidia.com>
To: Parav Pandit <parav@nvidia.com>
Cc: "Serge E. Hallyn" <serge@hallyn.com>,
"Eric W. Biederman" <ebiederm@xmission.com>,
"linux-rdma@vger.kernel.org" <linux-rdma@vger.kernel.org>,
"linux-security-module@vger.kernel.org"
<linux-security-module@vger.kernel.org>,
Leon Romanovsky <leonro@nvidia.com>
Subject: Re: [PATCH] RDMA/uverbs: Consider capability of the process that opens the file
Date: Wed, 23 Apr 2025 11:46:49 -0300 [thread overview]
Message-ID: <20250423144649.GA1743270@nvidia.com> (raw)
In-Reply-To: <CY8PR12MB71955B492640B228145DB9CFDCBA2@CY8PR12MB7195.namprd12.prod.outlook.com>
On Wed, Apr 23, 2025 at 12:41:26PM +0000, Parav Pandit wrote:
>
> > From: Serge E. Hallyn <serge@hallyn.com>
> > Sent: Tuesday, April 22, 2025 10:00 PM
> >
> > On Tue, Apr 22, 2025 at 01:11:27PM -0300, Jason Gunthorpe wrote:
> > > On Tue, Apr 22, 2025 at 08:14:33AM -0500, Serge E. Hallyn wrote:
> > > > Hi Jason,
> > > >
> > > > On Tue, Apr 22, 2025 at 09:46:40AM -0300, Jason Gunthorpe wrote:
> > > > > On Mon, Apr 21, 2025 at 12:22:36PM -0500, Serge E. Hallyn wrote:
> > > > > > > > 1. the create should check
> > > > > > > > ns_capable(current->nsproxy->net->user_ns,
> > > > > > > > CAP_NET_RAW)
> > > > > > > I believe this is sufficient as this create call happens through the
> > ioctl().
> > > > > > > But more question on #3.
> > > > >
> > > > > I think this is the right one to use everywhere.
> > > >
> > > > It's the right one to use when creating resources, but when later
> > > > using them, since below you say that the resource should in fact be
> > > > tied to the creator's network namespace, that means that checking
> > > > current->nsproxy->net->user_ns would have nothing to do with the
> > > > resource being used, right?
> > >
> > > Yes, in that case you'd check something stored in the uobject.
> >
> > Perfect, that's exactly the kind of thing I was looking for. Thanks.
> >
> It means uboject create path will refcount and store user_ns,
>
> uobject->user_ns = get_user_ns(current->nsproxy->net->user_ns);
>
> And uobject destroy will do,
> put_user_ns(uobject->user_ns).
>
> This will ensure that in below flow we won't have use_after_free.
> 1. process_A created object in user_ns_A
> 2. process_A shared fd with process_B in user_ns_B
> 3. process_A is killed and
> 4. user_ns_A is free is attempted (free is skipped, until uobject is destroyed by process_B).
We only need to do that if something is legimitately doing capable
from a uobject outside of creation? Did you find that?
And I wonder if using the uobjects affiliated netdev's namespace is
OK?
Jason
next prev parent reply other threads:[~2025-04-23 14:46 UTC|newest]
Thread overview: 58+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-03-13 5:08 [PATCH] RDMA/uverbs: Consider capability of the process that opens the file Parav Pandit
2025-03-17 19:31 ` Jason Gunthorpe
2025-03-18 3:43 ` Parav Pandit
2025-03-18 11:20 ` Jason Gunthorpe
2025-03-18 12:30 ` Parav Pandit
2025-03-18 12:44 ` Jason Gunthorpe
2025-03-18 20:00 ` Eric W. Biederman
2025-03-18 22:57 ` Jason Gunthorpe
2025-04-04 14:53 ` Parav Pandit
2025-04-04 15:13 ` Jason Gunthorpe
2025-04-06 14:15 ` Serge E. Hallyn
2025-04-07 11:16 ` Parav Pandit
2025-04-07 14:46 ` sergeh
2025-04-20 12:30 ` Parav Pandit
2025-04-20 13:41 ` Serge E. Hallyn
2025-04-20 17:31 ` Parav Pandit
2025-04-07 16:12 ` Jason Gunthorpe
2025-04-08 14:44 ` Eric W. Biederman
2025-04-21 3:13 ` Serge E. Hallyn
2025-04-21 11:04 ` Parav Pandit
2025-04-21 13:00 ` Serge E. Hallyn
2025-04-21 13:33 ` Parav Pandit
2025-04-21 17:22 ` Serge E. Hallyn
2025-04-22 12:46 ` Jason Gunthorpe
2025-04-22 13:14 ` Serge E. Hallyn
2025-04-22 16:11 ` Jason Gunthorpe
2025-04-22 16:29 ` Serge E. Hallyn
2025-04-23 12:41 ` Parav Pandit
2025-04-23 14:46 ` Jason Gunthorpe [this message]
2025-04-23 15:43 ` Eric W. Biederman
2025-04-23 15:56 ` Parav Pandit
2025-04-23 16:45 ` Jason Gunthorpe
2025-04-24 9:08 ` Parav Pandit
2025-04-24 14:13 ` Jason Gunthorpe
2025-04-25 13:14 ` Parav Pandit
2025-04-25 13:29 ` Jason Gunthorpe
2025-04-25 13:54 ` Parav Pandit
2025-04-25 14:06 ` Serge E. Hallyn
2025-04-25 15:05 ` Parav Pandit
2025-04-25 15:29 ` Serge E. Hallyn
2025-04-25 13:59 ` Serge E. Hallyn
2025-04-25 14:01 ` Serge E. Hallyn
2025-04-25 14:24 ` Jason Gunthorpe
2025-04-25 15:06 ` Serge E. Hallyn
2025-04-25 15:27 ` Parav Pandit
2025-04-25 15:46 ` Eric W. Biederman
2025-04-25 16:16 ` Parav Pandit
2025-04-25 15:32 ` Eric W. Biederman
2025-04-25 16:21 ` Jason Gunthorpe
2025-04-25 17:34 ` Eric W. Biederman
2025-04-25 18:20 ` Parav Pandit
2025-04-25 18:35 ` Jason Gunthorpe
2025-04-27 14:30 ` Serge E. Hallyn
2025-04-28 17:03 ` Eric W. Biederman
2025-04-29 3:56 ` Eric W. Biederman
2025-04-29 10:39 ` Parav Pandit
2025-04-30 3:34 ` Eric W. Biederman
2025-04-30 12:14 ` Parav Pandit
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=20250423144649.GA1743270@nvidia.com \
--to=jgg@nvidia.com \
--cc=ebiederm@xmission.com \
--cc=leonro@nvidia.com \
--cc=linux-rdma@vger.kernel.org \
--cc=linux-security-module@vger.kernel.org \
--cc=parav@nvidia.com \
--cc=serge@hallyn.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