[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: bpo kernel 4.7 broken? Re: Some function keys not working on a ThinkPad W550s



On Tue, 29 Nov 2016, Richard Hector wrote:
> On 26/11/16 14:27, Henrique de Moraes Holschuh wrote:
> > On Fri, 25 Nov 2016, Celejar wrote:
> >> > applications and a few things from backports. I'm currently running
> >> > kernel 4.7.0-0.bpo.1-amd64 from backports.
> > Don't.  That kernel is very broken.  Switch to the latest 4.8 kernel
> > available from either unstable or kernel.org, or compile the latest
> > 4.4-longterm kernel from kernel.org.
> 
> Woah. I was just about to install that. Am I ok, and better off,
> sticking with the 4.6 bpo kernel?

Likely so.  As usual, it is not going to be broken for everyone, and
really broken for a few.

> Will 4.8 hit bpo soon?

I don't know, but I would not be surprised if it never happens and the
next bpo kernel update ends up being a 4.9 kernel.

Linux kernel 4.9 is going to be the next longterm kernel, so I'd expect
the Debian kernel team to ignore 4.8 in unstable in order to focus their
efforts on being able to upload 4.9 (which is about to be released) to
unstable soonish.  Our release kernel target for Stretch is likely to be
4.9-longterm (it would make a *lot* of sense for it to be so).

And it is not like the current kernel in unstable/testing (4.8.7-based)
is up-to-date: it isn't.  And the fixes it is missing when compared to
4.8.11 are non-trivial, and some of them are rather important.

So, 4.8.7 might be even more broken than the 4.7 kernel currently in
backports (wherever 4.7 works reasonably well) for all I know.

> And is there a list I should be on to see this kind of thing?

LKML, mostly.  Also, git commit logs for the upstream kernel trees.

> (I just found the debian-backports list, but a quick scan only shows a
> couple of quite specific bugs in 4.7)

The Debian BTS is not the best place to look for kernel issues, it is
just one more place you have to check.  You also need to check the LKML
and the kernel bugzilla :-(

-- 
  Henrique Holschuh


Reply to: