[Ksummit-2012-discuss] [ATTEND] linux-next and process

James Bottomley James.Bottomley at HansenPartnership.com
Tue Jun 19 09:59:19 UTC 2012


On Tue, 2012-06-19 at 10:25 +0100, Mark Brown wrote:
> On Tue, Jun 19, 2012 at 08:37:41AM +0100, James Bottomley wrote:
> 
> > I'd also be interested in the other statistic:  What were the 1643
> > commits that were in next-20120521 but never made it to v3.5-rc1 doing
> > there (that's 17% of the next tree)?  next is supposed to be our merge
> > canary, so in theory, only commits destined for the next merge window
> > should be in there.
> 
> There's a fair chunk of -next commits that don't get merged from the KVM
> userspace tools which don't seem to be going towards mainline at any
> great rate (there is some debate about putting them in the kernel tree
> at all) but on the other hand aren't really impacting anything else
> either.

There surely aren't 1643, are there?  But if they're not on upstream
track, they shouldn't be in linux-next (especially if there's debate
about whether they will go upstream).  If they have no impact (as in
they're sitting in a defined subdirectory and are never built by a
standard kernel build and don't get looked at by any of the kernel tests
run on next) was there actually any point in putting them in next in the
first place?

James




More information about the Ksummit-2012-discuss mailing list