Re: Debian kernel maintainter takeover
On Mon, May 17, 2004 at 05:27:37PM +0200, Andreas Barth wrote:
> * Christian T. Steigies (cts@debian.org) [040517 16:10]:
> > On Mon, May 17, 2004 at 03:35:34PM +0200, Andreas Barth wrote:
>
> > > All I wish is that we reduce the number of source packages for the
> > > kernel, to ease the load for the security team.
>
> > I don't see how this reduces the load for the security team.
>
> Well, I'm not part of the security team, so my answers are not
> authoritative. Martin Schulz has said it, see
> http://lists.debian.org/debian-devel/2004/04/msg06282.html ;
> please read that mail for the full reasons.
I read that some time ago.
> As I understood the security team, the number of kernel source
> packages just needs to be reduced. At the moment, we have e.g.
> kernel-source-2.4.19 testing 2.4.19-11 all source
> kernel-source-2.4.20 testing 2.4.20-14 all source
> kernel-source-2.4.21 testing 2.4.21-8 all source
> kernel-source-2.4.22 testing 2.4.22-7 all source
> kernel-source-2.4.24 testing 2.4.24-3 all source
> kernel-source-2.4.25 testing 2.4.25-1 all source
> in testing. If we manage to reduce this to two versions (one fast one,
> and one slow one), that would be a step forward.
No problem with m68k, actually all 2.4 < 2.4.25 have been removed already.
2.4.25 could be removed as well, since 2.4.26 m68k images made it into
testing recently. I hope with two versions you mean the latest 2.2 and 2.4
though ;-)
> Well, we have two issues right now:
> 1. kernel-source and -binary-packages are independend in their walk
> down to sarge, so we have sometimes just too much packages there.
> 2. If different archs depend on different kernel versions, it's
> necessary to patch more kernel-source-packages.
> These two issues are - as far as I can see - the main showstoppers,
> but our current 11 kernel-source-packages and 48 image packages are
> way too much.
>
> Of course, getting m68k to the fast architectures would be fun. ;)
I don't see a problem with that, after all m68k is not the arch that still
has 2.2.10 or 2.4.19 in the archive. You can put us in the "fast" arch, as
long as you let us play with 2.2 and 2.4. Of course fast is relative, since
it takes me a while to build all those m68k images. Plus the time to create
a working m68k kernel-patch, fixing the bugs introduced by the
kernel-source. But I am optimistic that I can reduce that time by using a
cross compiler for testing. If only I had a working cross compiler for
gcc-2.95 and gcc-272...
So what are the plans for this one-source-package-builds-multiple-arch
kernel-images? Which approach for building the images are we going to use?
Any approach everybody could agree on? I assume everybody just needs to
unpack the kernel-source, apply (arch specific?) kernel-patches, copy an
arch and subarch specific config file and run make-kpkg. Or what am I
missing here?
Christian
Reply to: