When will we stop shipping WLAN improvements ahead of upstream in released Fedora version?

[John W. Linville]

On Tue, Aug 05, 2008 at 05:05:33PM -0400, Chuck Ebbert wrote:
> John W. Linville wrote:
> >
> >I still think that for continuity's sake f8 and f9 should continue
> >to get wireless fixes from 2.6.27.  Those should only be specific
> >bug fixes (although I suppose there is still time for a new driver),
> >so hopefully the nattering nabobs won't be opposed to continuing with
> >that part of the original plan.
> >
>
> If you push wireless fixes to -stable Fedora would then get them for "free".
> I don't see many wireless patches in there generally, though the ath5k
> memory corruption fix just went in.

Well I see where you had to revert a few in the F-9 changelogs as
you rebased on later -stable kernels, so there must be a few getting
through. :-)

If you would like to nominate more wireless fixes for -stable, feel
free to do so.  As it is, I mostly rely on my driver/stack maintainers
to identify appropriate patches for stable -- they are aware of the
process, including the Cc: stable at kernel.org trick.

Even so, that only works for those problems which date back to 2.6.26.
Since Fedora kernels already have 2.6.27 code, then the 2.6.27 patches
would seem appropriate for Fedora too.  Of course you could elide the
2.6.27 code from the Fedora kernels as you have suggested elsewhere,
but then you will reintroduce other problems.

> And new drivers would be great. Lots of people seem to want ath9k, for
> example.

We'll see.  Honestly I've been dropping other things in favor of
keeping Fedora up-to-date for the last year or more.  Rather than find
ways to spend just as much time on Fedora while "getting less bang for
my buck", I'll probably find something else to occupy more of my time.
In particular, I think ath9k will make it into the 2.6.27 queue.

John
-- 
John W. Linville
linville at redhat.com