From: "Gary Guo" <gary@garyguo.net>
To: "Jarkko Sakkinen" <jarkko@kernel.org>, "Gary Guo" <gary@garyguo.net>
Cc: "David Howells" <dhowells@redhat.com>,
"Paul Moore" <paul@paul-moore.com>,
"James Morris" <jmorris@namei.org>,
"Serge E. Hallyn" <serge@hallyn.com>, <keyrings@vger.kernel.org>,
<linux-security-module@vger.kernel.org>,
<linux-kernel@vger.kernel.org>
Subject: Re: [PATCH] keys: allow request-key path to be configured via Kconfig
Date: Wed, 10 Jun 2026 14:31:40 +0100 [thread overview]
Message-ID: <DJ5ENJM9BACZ.27BGMVKTKHAZD@garyguo.net> (raw)
In-Reply-To: <ailgG5JF6Qu_6v0P@kernel.org>
On Wed Jun 10, 2026 at 2:01 PM BST, Jarkko Sakkinen wrote:
> On Wed, Jun 10, 2026 at 03:57:37PM +0300, Jarkko Sakkinen wrote:
>> On Mon, Jun 08, 2026 at 11:30:06AM +0100, Gary Guo wrote:
>> > On Mon Jun 8, 2026 at 5:59 AM BST, Jarkko Sakkinen wrote:
>> > > On Mon, Jun 08, 2026 at 07:50:03AM +0300, Jarkko Sakkinen wrote:
>> > >> On Sun, Jun 07, 2026 at 02:49:27PM +0100, Gary Guo wrote:
>> > >> > From: Gary Guo <gary@garyguo.net>
>> > >> >
>> > >> > Some Linux distributions (e.g. NixOS) does not have /sbin present, and they
>> > >> > currently carry patches to replace /sbin/request-key to some other path.
>> > >>
>> > >> Sorry but no configuration for introducing API divergence.
>> >
>> > What is the API divergence here? Distros can already patch the kernel or place a
>> > different binary there, so I don't see what's being gained on not allowing to
>> > change this with Kconfig.
>>
>> There's lot of out-of-tree drivers too that distributions. I'm not
>> finding anything usefel in this argument.
Out-of-tree drivers are, well, out of tree. This one requires patching the tree.
Unlike many other distros, so far the only patches needed for NixOS is patching
out /sbin.
>>
>> >
>> > Also to note, the actual binary being called can already be swapped out by
>> > CONFIG_STATIC_USERMODEHELPER_PATH, although for the NixOS this is not the proper
>> > mechanism as it affects coredump too which isn't a fixed path binary in /sbin.
>>
>> I have not seen actual uses of CONFIG_STATIC_USERMODEHELPER_PATH. You
>> could probably use it with busybox?
I think it's used for hardening.
>> >
>> > This is really just for distros to be able to configure where /sbin is located.
>> > Given usr merge and (some distros) bin/sbin merge, the canonical path of
>> > request-key binary is very likely not /sbin/request-key anymore, so it seems to
>> > make sense to me to allow this to be changed rather than always go through
>> > compatibility symlinks.
>>
>> I doubt there's a huge demand other than NixOS. Just basing this on that
>> no other noise have been made so far.
>>
>> >
>> > How about a something like CONFIG_DEFAULT_USERMODEHELPER_PATH which defaults to
>> > /sbin, and then request-key uses that concatenated with "/request-key"?
>> >
>> > [snip]
>>
>> I don't frankly care how NixOS works per se in details. Scope this into
>> message to problem that it addresses.
Well, I reckon that's what's going to happen, so in the commit message I just
included "binary is not in /sbin". But the idea is that there's a good reason
that it's not in /sbin.
>
> Not 100% NAK but this does not have "universal logic" embedded into it"
>
> "Distro's use it" is popularity opinion, which has no place over here.
> Mastodon, Threads etc. work for that so much better.
I disagree. Distro is really just a collection of users. I would rather than
phrase this as "user's using it this way". If something needs to be patched to
be used, I think that's rather a good reason to make the change.
I think "user wants to control where UMH lives" is a pretty good motivation, but
it looks like you disagree. Anyhow, if you don't like the idea, I'll just drop
this patch, as I am not the one maintaining these distro patches anyway. I just
think it's the best if Kconfig can meet user demand and more people can run
unpatched kernels.
Best,
Gary
> Perhaps if the motivation-stimuli-solution type of logics gets carved
> crystal clear we can move forward. I.e. you need to work on this. I've
> given my feedback for this version, and it is not good enough, sorry.
>
> BR, Jarkko
next prev parent reply other threads:[~2026-06-10 13:31 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-06-07 13:49 [PATCH] keys: allow request-key path to be configured via Kconfig Gary Guo
2026-06-07 19:55 ` Serge E. Hallyn
2026-06-08 4:49 ` Jarkko Sakkinen
2026-06-08 4:59 ` Jarkko Sakkinen
2026-06-08 10:30 ` Gary Guo
2026-06-10 12:57 ` Jarkko Sakkinen
2026-06-10 13:01 ` Jarkko Sakkinen
2026-06-10 13:31 ` Gary Guo [this message]
2026-06-10 13:37 ` Gary Guo
2026-06-15 11:49 ` Jarkko Sakkinen
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=DJ5ENJM9BACZ.27BGMVKTKHAZD@garyguo.net \
--to=gary@garyguo.net \
--cc=dhowells@redhat.com \
--cc=jarkko@kernel.org \
--cc=jmorris@namei.org \
--cc=keyrings@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-security-module@vger.kernel.org \
--cc=paul@paul-moore.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