[Ksummit-2012-discuss] [ATTEND] stable kernel stuff and grumpy maintainers [bisection/rebase/-next]
sfr at canb.auug.org.au
Wed Jun 20 22:50:02 UTC 2012
On Wed, 20 Jun 2012 21:42:01 +0000 "Luck, Tony" <tony.luck at intel.com> wrote:
> > > This is why I suggested that "-rc5" should be declared as the rebase
> > > deadline for trees feeding into linux-next ... and that maintainers
> > > should be ENCOURAGED TO REBASE for any issue where it makes sense.
> > >
> > We used to do that in -tip, and Linus ripped us a new **** for it.
> It would be good to poke Linus a bit more on why ...
The problem is that once a tree is published, it may be merged into other
trees. I you then rebase the original, you need to track down the trees
that merged it and rebase them as well. Almost every release I see cases
where one tree merges a branch from another and then that branch is
rebased but the tree that merged it is not fixed for some time (if
ever). It causes conflicts when the trees are merged and (in at least
one case) mismerges. It can also lead to us having multiple
copies/versions of the same patches in linux-next and Linus' tree.
So you need to be clear about the difference between test trees and
stable trees. Once something reaches linux-next it should be stable
except for rare circumstances. If it isn't, then that may indicate that
insufficient (unit) testing is being done.
More care and early testing, less haste ...
Stephen Rothwell sfr at canb.auug.org.au
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Size: 836 bytes
Desc: not available
More information about the Ksummit-2012-discuss