Re: [PATCH] [v2] Documentation: Provide guidelines for tool-generated content
Linus Torvalds <[email protected]>
| Newsgroups | dev.linux.lists.ksummit,org.kernel.vger.linux-kernel,org.kernel.vger.workflows |
|---|---|
| Message-ID | <CAHk-=wgEPve=BO=SOmgEOd4kv76bSbm0jWFzRzcs4Y7EedpgfA@mail.gmail.com> |
On Mon, 10 Nov 2025 at 09:25, Laurent Pinchart <[email protected]> wrote: > > Mechanical transformations are often performed with Coccinelle. Given > how you mention that tool below, I wouldn't frame it as out of scope > here. Honestly, I think the documented rule should not aim to treat AI as anything special at all, and literally just talk about tooling. Exactly because we've used things like coccinelle (and much simpler tools like 'sed', for that matter) for ages. IOW, this should all be about "tool-assisted patches should be described as such, and should explain how the tool was used". If people send in patches that have been generated by tools, we already ask people to just include the script in the commit message. I mean, we already have commit messages that say things like This is a completely mechanical patch (done with a simple "sed -i" statement). when people do mindless conversions that are so straightforward that the actual sed patch isn't even documented (in that case is was something like just sed -i 's/__ASSEMBLY__/__ASSEMBLER__/' or whatever), and in other cases people include the actual script (random example being commit 96b451d53ae9: "drm/{i915,xe}: convert i915 and xe display members into pointers"). I think we should treat any AI generated patches similarly: people should mention the tool it was done with, and the script (ok, the "scripts" are called "prompts", because AI is so "special") used. Sure, AI ends up making the result potentially much more subtle, but I don't think the *issue* is new, and I don't think it should need to be treated as such. Linus