On Wed, 06 Jun 2012 10:57:21 -0500, Bruce Dubbs <bruce.du...@gmail.com> wrote:

> I suggest this topic be dropped as moot.  The discussion doesn't add
> anything and we are agreed that today 3.4.1 is the latest stable version.

On the contrary, I think it's important that folks here understand a bit
about the kernel development/release process so that they know why we choose
the kernels we do.  The following is my understanding and therefore comes with
the usual caveat that I may be talking complete and utter rubbish.

The various kernel subsytem maintainers look after their own git trees.  This
is where they merge patches/pull requests for their particular subsystems.

Stephen Rothwell maintains a linux-next git tree, which one can think of as an
integration area.  This contains the various subsystem maintainer's intended
code for the next kernel to be released by Linus.  This is not quite bleeding
edge stuff (the real bleeding edge stuff resides in the subsystem maintainer's
own git trees); it's usually been compile tested, or at least had git merge
conflicts resolved.

Linus then pulls code from linux-next into his own tree.  Various release
candidates are made until Linus is suitably happy to cut a release from it.
Both the RCs and the final release are given a 'mainline' label.  They
should be relatively free from show stopper bugs, but some may inevitably
creep in due to relatively limited exposure to various hardware and user space
combinations.

Greg Kroah-Hartman maintains a 'stable' git tree.  This is where ongoing
maintenance of a particular mainline release is carried out.  Bugs that are
reported through various channels, and that have already been fixed in Linus'
tree, are squashed here.

So, LFS uses Linus' non-RC mainline releases when they come out, followed by
Greg K-H's stable releases up until Linus' next non-RC mainline release is
made.

Regards,

Matt.

-- 
http://linuxfromscratch.org/mailman/listinfo/lfs-dev
FAQ: http://www.linuxfromscratch.org/faq/
Unsubscribe: See the above information page

Reply via email to