Re: Gentoo guys starting a fork of udev
On Wed, Nov 14, 2012 at 11:34:27PM +0100, John Paul Adrian Glaubitz wrote:
> On Wed, Nov 14, 2012 at 03:41:55PM -0300, gustavo panizzo <gfa> wrote:
> > >udev isn't broken.
> >
> > really?
> >
> > https://bbs.archlinux.org/viewtopic.php?id=134012&p=1
>
> I actually remember having seen this issue on Fedora Rawhide as well,
> but it vanished after an update a few weeks ago, so it rather seems
> like a "normal" bug to me. That's not really what "broken" means in
> this context.
>
> > but don't trust me
> >
> > https://lkml.org/lkml/2012/10/2/505
>
> Well, yes, it's the same issue. Linus is well known for going on a
> rant very quickly, but that doesn't mean that udev is completely
> broken.
>
> Yes, they obviously made a recent change that broke module loading on
> some machines, but that doesn't mean the whole concept is
> broken. That's just an unfair statement. Also, Kay is admitting that
> there is/was a problem with udev that needs to be addressed and it
> seems that they did because I cannot reproduce it anymore with udev
> 195 anymore.
I believe the regression (removal of support for firmware loading
during module loading) has been fixed. However, the udev developers
*knew in advance* that this would be a problem, reported such uses
of firmware loading as being driver bugs. They then went ahead and
changed udev even though the drivers had not all been updated (and it
was evidently not easy to do so in some cases).
Ben.
--
Ben Hutchings
We get into the habit of living before acquiring the habit of thinking.
- Albert Camus
Reply to: