On Thu, Feb 22, 2024 at 07:49:20PM +0000, Richard Sandiford wrote: > Thanks for the update. This is still quite hard to review though. > Sorry to ask for another round, but could you split it up further? > The ideal thing would be if patches that move code do nothing other > than move code, and if patches that change code do those changes > in-place.
In general, if there is a (big) part to the patch that does not change behaviour at all, it should be a separate patch. Such a patch is then easy to review (write down in the commit message that it does not change behaviour though, it helps reviewers). It also makes the remaining tiny patches much easier to review then. Very generally, any patch that makes interesting changes should not have more than a few lines semantic content. That can be repeated of course, and have fall-out mechanical follow-up changes, but that's the essence of good patchsets: one change per patch. And then the commit message can be simple as well, and the chanegelog will be easy to write. That is the litmus test for good patch series :-) Segher
