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

Re: 2.6.4 kernel



On Mon, Mar 22, 2004 at 06:10:58PM +0100, Jens Schmalzing wrote:
> Hi,
> 
> Sven Luther writes:
> 
> > Why shouldn't they ? Mmm, i see why, they install into : 
> > 
> >   /usr/src/kernel-patches/powerpc/debian-powerpc.diff.gz
> > 
> > Do you think kernel-package would be ready to have it moved to :
> > 
> >   /usr/src/kernel-patches/powerpc/2.4.25/debian-powerpc.diff.gz
> 
> Of course.  The location of the patch is entirely up to the apply and
> unapply scripts, which you should generate with dh_installkpatches

And what about the apply/unapply patches ? I only cited the patch, but
the same problem happens with the rest of the patch stuff.

> from the dh-kpatches package anyway.  Take a look at kernel-patch-xfs
> or the late kernel-patch-benh, both by Yours truly.

I will have a look at these packages, thanks for the hint. I didn't
touch those stuff since i took over the package from Dan.

> > Somehow i doubt it.
> 
> > Why do you need to have them both installed ? because of the mol
> > modules you want to build ?
> 
> Yep.  I didn't really mind dropping support for 2.4.24 when 2.4.25

2.4.22 you mean, there was never an official debian 2.4.24 powerpc
kernel package.

> came out.  I would hate to drop support for 2.4.25 when 2.6.4 comes
> out, though, and building is a lot easier if I can have both patches
> installed simultaneously.

Yeah, i understand.

> > I still think it would be nicer to build the mol modules directly from
> > the powerpc kernel package, but i understand why you prefer it the other
> > way around.
> 
> You can have the mol-modules package when you pry it from my cold dead
> fingers.

Did we not decide you would co-maintain the powerpc kernels ? 

> I *do* hope you come up with the correct answer to the above phrase.
> 
> That said, the reason for keeping the pre-built mol modules in a
> source package separate from both the remainder of mol and the kernels
> is simple.  The modules have to be rebuilt both when mol changes and
> when the kernels change.  Both changes happened several times since we
> last talked about this.

Yeah, i know.

Still, the fact that you need to unpack the kernel again and configure
it is no nice, so the solution i propose would have been nicer.

Friendly,

Sven Luther



Reply to: