On Mon, Dec 12, 2022 at 06:56:59PM -0800, Linus Torvalds wrote: > On Mon, Dec 12, 2022 at 3:19 AM Christian Brauner <brauner@xxxxxxxxxx> wrote: > > > > For a long and detailed > > explanation for just some of the issues [1] provides a good summary. > > There is no link [1]. > > > A few implementation details: > > > > * The series makes sure to retain exactly the same security and integrity module > > permission checks. See [2] for annotated callchains. > > There is no link [2]. > > This was an extensive changelog for my merge commit, so it's all fine > and I've pulled it, but it does look like some pieces were either > missing, or there was a bit of a cut-and-paste from previous > explanations without the links.. Bah, there was a single stray word "mainline" in the pr after the /* Conflicts */ section because I copied it over a small (relatively uninteresting) paragraph. So the two missing links are: [1]: https://lore.kernel.org/all/20220801145520.1532837-1-brauner@xxxxxxxxxx [2]: https://gist.github.com/brauner/12c795b93a05dc3b3056b1982549a633 they are also listed in the cover letter for the series. I think I might need a script to look for missing links in the pull request. I've had a missing link before. Hopefully I'll get around to this. About the cut-and-paste: What follows is just my personal theory/preference but I think it helps to understand how the pr message come together. If the series is a self-contained topic and not a collection of a pile of commits than I aim for the description given in the cover letter and the description given in the merge/pull request message to be almost indistinguishable. So the cover letter I keep in git edit --branch-description will morph into the merge/pull request message. So I often will try to write the cover letter in a form that is suitable for the merge commit. Of course, the cover letter will often contain additional technical information that might not be suitable for the cover letter. Such sections will then be cut our or rewritten for the pr message. Christian
