Hi, Thanks all who have contributed to the descussion! Much appreciated.
On Mon, Jan 16, 2023 at 03:08:22PM -0700, Sam Hartman wrote: > >>>>> "Moritz" == Moritz Mühlenhoff <j...@inutil.org> writes: > Moritz> Not moving to 6.1.x (which is most likely the next Linux > Moritz> kernel LTS) is by far a worse regression since it applies to > Moritz> every single Debian system. > > Moritz> As a community distro without paid, full time kernel > Moritz> maintainers we can't just randomly stick to an older kernel > Moritz> tree and decide to assess/backport hundreds of patches sent > Moritz> to stable@ every week. > > I think we're all agreed on that point. > What we can do is delay the release if we have a serious enough bug that > is not fixed yet. > I think what people are saying on this bug is that this issue is serious > enough it should be considered as a release blocker---something that > could actually delay the release. I do agree, this is defintively an issue we would not would want to have in a stable release, so the assessment of marking it RC is defintively right. While 6.0.y is now EOL, and 6.1 was aimed to be the LTS release, and the reason we would like to pick that for bookworm, there was not yet an official announce for it, in part because of reported regressions, cf. https://lore.kernel.org/all/y53bputyk+3dj...@kroah.com/ and https://lore.kernel.org/all/20230114084719.ga6...@1wt.eu/ for some context. > From where I sit, thinking about what I've deployed in the last five > years, I agree with that analysis: this *might* be serious enough to > delay the release until there is a fix. > Given that we can't stick on 6.0, I think we should try to get this > into testing as soon as possible so we can see how big of an impact it > is in practice. > I don't like to see testing broken, but I like to see stable broken in > serious ways even less. > And so I'd recommend we see how badly this is going to hurt us. I will bite the bullet (taking full responsibility for it if necessary, don't blame the other kernel team members) and ask here now the release team: Can we let linux 6.1.4-1 despite the RC bug reported, migrate to testing, so we can move on to 6.1.y? Let's keep the bug as RC severity. I'm currently working on uploading as well 6.1.6 or 6.1.7 (depending on the timing) further after that to unstable. Unfortuantely there is still not a solution to address #1028451 but will contain other important fixes (including security ones). Thank you for considering it, Odyx, I feel sorry, this will knowingly impact your and others! Regards, Salvatore
