[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <bae72e9f-c88c-43ec-a91a-40f217ea2adc@e43.eu>
Date: Thu, 14 Nov 2024 13:56:45 +0100
From: Erin Shepherd <erin.shepherd@....eu>
To: Christoph Hellwig <hch@...radead.org>
Cc: Christian Brauner <brauner@...nel.org>,
Alexander Viro <viro@...iv.linux.org.uk>, Jan Kara <jack@...e.cz>,
Chuck Lever <chuck.lever@...cle.com>, linux-fsdevel@...r.kernel.org,
linux-kernel@...r.kernel.org, Jeff Layton <jlayton@...nel.org>,
Amir Goldstein <amir73il@...il.com>, linux-nfs@...r.kernel.org
Subject: Re: [PATCH v2 2/3] exportfs: allow fs to disable CAP_DAC_READ_SEARCH
check
On 14/11/2024 05:37, Christoph Hellwig wrote:
> On Wed, Nov 13, 2024 at 05:55:24PM +0000, Erin Shepherd wrote:
>> For pidfs, there is no reason to restrict file handle decoding by
>> CAP_DAC_READ_SEARCH.
> Why is there no reason, i.e. why do you think it is safe.
A process can use both open_by_handle_at to open the exact same set of
pidfds as they can by pidfd_open. i.e. there is no reason to additionally
restrict access to the former API.
>> Introduce an export_ops flag that can indicate
>> this
> Also why is is desirable?
>
> To be this looks more than sketchy with the actual exporting hat on,
> but I guess that's now how the cool kids use open by handle these days.
Right - we have a bunch of API file systems where userspace wants stable
non-reused file references for the same reasons network filesystems do.
The first example of this was cgroupfs, but the same rationale exists for
pidfs and process tracking.
Powered by blists - more mailing lists