On Thu, 3 Dec 2020 13:18:13 -0800 Linus Torvalds wrote: > On Thu, Dec 3, 2020 at 12:45 PM Jakub Kicinski <k...@kernel.org> wrote: > > Networking fixes for 5.10-rc7, including fixes from bpf, netfilter, > > wireless drivers, wireless mesh and can. > > Thanks, pulled. > > And btw - maybe I've already talked about this, but since next week is > (hopefully) going to be the last week of rc release: since the > networking pulls tend to be some of the bigger ones, one thing I've > asked David to do in the past is to (a) not send a big networking pull > request right before the final release and (b) let me know whether > there is anything worrisome going on in networking. > > So if you send it on a Thursday (like this one), then that's all good > - it's the "Oh, it's Sunday noon, I was planning on a final release in > the afternoon, and I have a big networking fix pull request in my > mailbox" that I'd prefer to not see.
Make sense. I'm not anticipating that the last PR will be much smaller, given we get a constant stream of fixes for older releases and the review coverage is pretty good so we can apply stuff with confidence. Sounds like a comparable PR size will not be a major concern to you as long as the PR comes in early on Thu and we are reporting any sign of trouble. Sounds good! > A heads up on the "Uhhuh - we have something bad going in the > networking tree" kind of situation you can obviously send at any time. > If there are known issues, I'll just make an rc8 - I prefer not to > _have_ to, of course, but I'd always much rather be safe than release > the final kernel just because I didn't know of some pending issue. Will do! > (And the reverse - just a note saying "everything looks fine, none of > this is scary and there's nothing pending that looks at all worrisome > either" - for the last rc pull is obviously also always appreciated, > but generally I'll assume that unless something else is said, we're in > good shape). Ack, it's been smooth sailing so far in this release. No big scares, knock on wood. This time around (other than the large-ish ibmvnic set which was in the works for a while) the PR was smaller, but I think that's only due to Turkey lethargy. Thanks for this note, I was shy to ask about the endgame :)
