[arch-dev-public] Module breakage (WAS: Re: [signoff] kernel 184.108.40.206-1)
t.powa at gmx.de
Sat Jan 9 10:57:17 EST 2010
Am Samstag 09 Januar 2010 schrieb Dan McGee:
> On Sat, Jan 9, 2010 at 9:03 AM, Thomas Bächler <thomas at archlinux.org> wrote:
> > Am 08.01.2010 09:06, schrieb Tobias Powalowski:
> > > Hi guys,
> > > bump to latest bugfix version.
> > >
> > > Arch Linux bugfixes/feature requests:
> > > http://bugs.archlinux.org/task/17538 # added blktrace
> > > http://bugs.archlinux.org/task/17106 # finally added CONFIG_MMIOTRACE
> > One of these changes broke aufs, a rebuild is necessary. At least
> > virtualbox is also affected, maybe more. I am rebuilding aufs, please
> > post if any other modules are affected.
> This is where I find our "don't rebuild the modules" policy lacking.
> We never see these breakages until they have already affected a fair
> number of people. You also have very little transparency into what
> kernel minor version you are actually running on your system unless I
> am missing something, so doing something like verifying you have an
> unaffected kernel to a security issue or something without doing a
> bunch of looking at reboot dates, build dates, etc.
> Do we do it to save ourselves time, save users extra downloads, or
> what is the primary reason? If it is one of these two things, then we
> can address both in different ways- 1. make an auto-build system for
> things like this 2. xdelta packages.
We normally only add such changes on big upgrades,
nvidia worked here fine that's the reason i didn't discovered this.
For safety i will rebuild every binary module now.
In some hours this all mirrors will be fixed.
Archlinux Developer & Package Maintainer (tpowa)
tpowa at archlinux.org
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Size: 198 bytes
Desc: This is a digitally signed message part.
More information about the arch-dev-public