[mythtv-users] MythTV problems with 3.7.x and 3.8.x kernel

[Karl Newman]

On Sun, Feb 24, 2013 at 8:38 AM, Neil Salstrom <salstrom at gmail.com> wrote:

> On Sat, Feb 23, 2013 at 9:56 PM, Karl Newman <newmank1 at asme.org> wrote:
> >
> > Since you're set up to compile and use your own kernel, you might try a
> > bisect to find the offending commit. Here's the documentation:
> > http://www.kernel.org/pub//software/scm/git/docs/git-bisect.html In your
> > case you should start with the first 3.6.0 kernel and the first "bad" 3.7
> > kernel (3.7.0?). Based on my limited knowledge about kernel development
> > practices, I believe that will minimize the number of commits between the
> > good and bad trees. Ideally you would find a bad commit which could be
> > reported upstream to the linux kernel list, but if they don't confirm it
> as
> > a bug, at the least you could create a reverse patch that you could apply
> > which would remove the problematic commit for your custom kernel and then
> > the mythtv devs could look at it on their end. I've done this once and it
> > took about 12 iterations to find the commit which caused my problem (not
> > mythtv related).
> >
> > Karl
> >
>
> Wow...  That's crazy.  I had no idea git-bisect was a thing.  Giving
> it a go now!!!!
>
> Do I need to do a kpkg-clean between compiles?
>
> First iteration was "bad".  On to iteration number two....
>

Yeah, git has all kinds of handy functions. I don't know about kpkg-clean
(I use gentoo) but I doubt it. I think you just need to make sure you
install the modules for each iteration with 'make modules_install' (usually
I'll build the kernel and install modules in one line like: 'make && make
modules_install').

Karl
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.mythtv.org/pipermail/mythtv-users/attachments/20130224/23a0a059/attachment.html>