Hi Hans, others, On Thu, Sep 26, 2024 at 09:27:20AM +0200, Hans Verkuil wrote: > On 25/09/2024 21:58, Laurent Pinchart wrote: > > Hi Mauro, > > > > On Wed, Sep 18, 2024 at 09:24:54AM +0200, Mauro Carvalho Chehab wrote: > >> Em Tue, 17 Sep 2024 14:52:19 +0200 Hans Verkuil escreveu: > >>> On 9/17/24 11:17 AM, Sebastian Fricke wrote: > >>>> Greetings, > >>>> > >>>> I remember that we wanted to still define a couple of processes for the > >>>> multi-committer model for which we hadn't have the time on the media > >>>> summit. Just would like to gather who would be interested to meet for > >>>> that, where we meet (probably LPC venue) and when (Laurent just told me > >>>> that Friday is probably a good slot for that). > >>> > >>> Can you refresh my memory which processes need more work? > >> > >> I have the same doubt. As discussed during the summit, Hans and I had some > >> discussions yesterday, to address a few details. For both of us the process > >> sounds well defined. > >> > >> From my personal notes, this will be the new process: > >> > >> - committers will merge patches at media-committers.git tree at fdo, > >> provided that they'll follow the rules defined on a committers agreement > >> and (partially?) enforced by media-ci checks; > >> - core committers follow the same rules, with a broader privilege of > >> changing kernel API/ABI; > >> - committers will ensure that patchwork will reflect the review process of > >> the patches; > >> - maintainers will double-check if everything is ok and, if ok, merge the > >> changes at linuxtv.org. We intend to rename the tree there to "media.git", > >> being the main tree to be used for development; > >> - pull requests will keep using the same process as currently. The only > >> change is that the media-stage.git tree will be renamed to "media.git"; > >> - maintainers will periodically send patches upstream. > >> > >> The media-commiters.git tree at fdo might be rebased if needed; the > >> media.git tree at linuxtv.org is stable. A large effort will be taken to > >> avoid rebasing it. > >> > >> We may need some helper scripts and/or use pwclient to keep patchwork > >> updated after committers reviews. > > > > What will happen if we update the status of patches in patchwork when > > merging them to the fdo tree, and the tree is later rebased to drop > > commits ? Will the person rebasing handle updating patchwork to move the > > patches back from accepted to a different status ? > > That should be the responsibility of the person doing the rebase. I think > that's what is done today as well in the rare cases we rebase. Sounds reasonable. I'd also like to avoid rebases as much as possible. Do we have a list of cases where a rebase would be needed? A license issue or a missing Sob line, perhaps? -- Kind regards, Sakari Ailus
