On Wed, May 10, 2023 at 5:58 PM Greg Kroah-Hartman gregkh@linuxfoundation.org wrote:
Thanks! Turns out someone put the wrong "Fixes:" tag in that commit which is why I missed it.
Hmm. Presumably the real commit ceeadb83aea2 at some point got rebased, and had had that other mentioned SHA1 before that.
It might be a good idea in general - not just for stable - if we had some automation that said "this refers to a commit ID that doesn't exist".
Of course, sometimes those commits might exist elsewhere (ie the stable tree obviously refers to upstream commits that are *not* directly reachable from the commit that refers to it, and thus relies on another tree not rebasing itself).
But on the whole, I would expect that the normal situation, outside of that "upstream commit" issue in the stable tree, is that you only refer to commits that are actually reachable from the referrer commit.
Or do people refer to other branches' (or even other projects') commit IDs?
It might be interesting to have some automation, particularly if it then might highlight the situation where the same one-line description does exist under a different commit name...
Linus