linux-kernel - Re: [PATCH] [v3] Documentation: Provide guidelines for tool-generated content

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <d7625ea9-21c4-4644-b61a-0736b040edc2@lucifer.local>
Date: Fri, 9 Jan 2026 16:37:28 +0000
From: Lorenzo Stoakes <lorenzo.stoakes@...cle.com>
To: Miguel Ojeda <miguel.ojeda.sandonis@...il.com>
Cc: Sasha Levin <sashal@...nel.org>, Dave Hansen <dave.hansen@...el.com>,
        Dave Hansen <dave@...1.net>, Dave Hansen <dave.hansen@...ux.intel.com>,
        linux-kernel@...r.kernel.org, Shuah Khan <shuah@...nel.org>,
        Kees Cook <kees@...nel.org>,
        Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        Miguel Ojeda <ojeda@...nel.org>, Luis Chamberlain <mcgrof@...nel.org>,
        SeongJae Park <sj@...nel.org>, Dan Williams <dan.j.williams@...el.com>,
        Steven Rostedt <rostedt@...dmis.org>, NeilBrown <neilb@...mail.net>,
        Theodore Ts'o <tytso@....edu>, Jonathan Corbet <corbet@....net>,
        Vlastimil Babka <vbabka@...e.cz>, workflows@...r.kernel.org,
        ksummit@...ts.linux.dev
Subject: Re: [PATCH] [v3] Documentation: Provide guidelines for
 tool-generated content

On Fri, Jan 09, 2026 at 05:30:17PM +0100, Miguel Ojeda wrote:
> On Thu, Jan 8, 2026 at 8:28 PM Lorenzo Stoakes
> <lorenzo.stoakes@...cle.com> wrote:
> >
> > 'You _can_ be more transparent by adding information like this:...'
>
> I am not a native speaker, but my reading of that "can" was that it is
> suggesting ways to be more transparent that may or may not apply in
> particular cases, but the requirement of being transparent was already
> established by the previous sentence:
>
>     Second, when making a contribution, be transparent about
>     the origin of content in cover letters and changelogs.
>
> Which is reinforced by another imperative in the bullet point about prompts:
>
>     If code was largely generated from a single or short set of
>     prompts, include those prompts.
>
> Similarly, I read those other "might"s you quote like a set of things
> that could happen or not (and is not exhaustive) in particular cases
> and/or depending on the maintainer etc.

Right I mean I'm not disputing the logic of it, and the document _is_ well
written.

>
> At least that is my reading, and as far as I understood the TAB
> discussions, the goal of this patch was to document that non-trivial
> tool usage needs to be disclosed, including LLM use, and to me the
> patch already did that, but perhaps the wording can be more direct.

Yes, exactly. Really for me the whole thing is about emphasis.

The current version of my proposal is (hopefully) reaching towards quorum as it
takes into account feedback from Dave, Jens, Steven, and others probably who I
forget here (apologies) - see [0] - so I'm hoping that this _should_ be
acceptable as a means of establishing that emphasis without disrupting the
overall aims of the document?

It pleasingly is applicable to _all_ tooling and doesn't take a 'position' per
se on LLMs specifically.

[0]: https://lore.kernel.org/all/1273cff8-b114-4381-bbfe-aa228ce0d20d@lucifer.local/
>
> I hope that clarifies a bit...

Yes indeed :) thanks for that!

>
> Cheers,
> Miguel

Cheers, Lorenzo