* Sorry, I forgot to mention that this problem started with 167-1.
* I was surprised that my kernel wasn't using devtmpfs. This must have
got deselected at some point. Recompiling with that option fixed it.
* I also forgot to mention I was getting the ".udev/ already exists on the
static
$udev_root!" message which was kind of misleading because . . .
* It turned out I had a static /run with junk in it which is what the
above message was referring to as /dev/.udev. Deleting this erroneous
/run also solved the problem independent of using devtmpfs.
So I'm not sure what to conclude from all this in terms of smoothing the
upgrade path.
At the very least I think the ".udev/ already exists . . ." message
should only be displayed if that's actually the case. A separate "/run/udev
already exists
. . ." should be reserved for the case I had. This would have led to a
far quicker diagnosis.x
Kevin
On Tue 11.04.12 04:51, Marco d'Itri wrote:
On Apr 12, Kevin Mitchell <kevmi...@math.sfu.ca> wrote:
It would seem that the "udevadm settle" isn't doing what's expected?
Also, do you have a /run directory? Is it a tmpfs?
Does /run/udev/ exist? What about /dev/.udev/?
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
