Re: hoary kernel point release

Hi!

Fabio Massimo Di Nitto [2005-05-17  7:20 +0200]:
> Right now I have been informed by upstream that one of the SATA patches
> that are inside our tree is extremely dangerous and it can cause data 
> corruption.
> There is no update to that patch so it must to be reverted.
> the impact of this fix is pretty high unfortunatly. It seems to introduce an 
> ABI change
> and it will kill 2 modules from our tree. The 2 modules are not used by d-i 
> so that
> makes things a bit better.

The two modules are the SATA ones, I presume?
 
> A few minutes ago we found a bug on how our initrd is builded and due to this 
> bug
> the capability module is not added to the initrd as it should. The module is 
> still
> loaded at a later stage but this is not optimal.

Hm, should it be in /etc/modules? On my system it's not, and it isn't
loaded automatically on demand; is this already fixed in Breezy?

> My suggestion is to get 34.1 out with the 4 security fixes that we have been 
> working
> on (so we don't delay warty) and prepare a 35 to go in -updates together with 
> linux-meta
> and the whole dance of the abi change.

That works for me. We never really agreed upon a versioning schema for
-updates, it should be a different one than the development release
scheme. However, since the kernel already has a special versioning
scheme and we use 2.6.12 for Breezy, I don't have a problem with
calling it -35. I think your approach is fine; we need a separate
security update anyway.

Thanks!

Martin

-- 
Martin Pitt              http://www.piware.de
Ubuntu Developer   http://www.ubuntulinux.org
Debian Developer        http://www.debian.org

signature.asc
Description: Digital signature