On Mon, Jul 16, 2012 at 10:29:17AM -0700, John Stultz wrote: > Hey Greg,
Hm, you really should send this to [email protected] as I'm not the only one doing stable releases anymore. > Now that I think all the issues are put to bed upstream, I > wanted to check with you on advice for backporting the leapsecond > fixes to all the various stable trees. > > The core problem is that its a non-trivial amount of patches. > > For 3.4 and before kernels there are two issues. And for each of > those issues, there are fixes, then multiple fixups to those fixes. > > Since there have been structural cleanups along the way, most of the > changes don't backport cleanly. > > I realize quite a bit of care and caution is needed here, but I also > don't want to overload any -stable reviewers with layered patches. > > > So: > > Do you think it is best to generate one patch per upstream commit > (including any short-term bugs that they introduce)? Yes, it's easier that way to ensure that we get all of the fixes in. The number of patches don't matter, we can handle quantity easily. > Or is it better to roll a series of upstream commits into one > -stable commit so reviewers can see it is correct, without slogging > through a series? Nope, series is best, if at all possible. But, if things are messier for older kernel releases, then it might make sense to do a smaller number of patches, only if it is still readable and makes sense to do so. > (Or am I just kidding myself that there will be other people > actually reviewing these in detail?) I hope people are reviewing them :) thanks, greg k-h -- To unsubscribe from this list: send the line "unsubscribe stable" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html
