On Thu, 2010-01-07 at 11:35 +0800, Brice Rebsamen wrote:
> OK, shame on me: after I wrote this bug report I decided to try to
> read the kernel message in details to understand more about it and
> while I was doing so the kernel booted. The explanation is that since
> the realtek firmware are missing, the kernel fails to turn on the
> ethernet link, and hangs there for some time, then gives up and carry
> on with the boot process.
That is not the true explanation. The kernel sends requests for
firmware as device events, which are normally handled by udev. If
firmware is missing, it can immediately cancel the request. The problem
is that udev is not running, so the kernel keeps on waiting for a
response up to the time limit (60 seconds).
This is a known bug in initramfs-tools, which I think will be fixed
soon.
> So this is no longer a critical bug. Sorry about the false alarm.
>
> However, there is still a problem with the packaging: where are the
> firmwares required?
Not clearly distributable. I will contact Realtek again to request
clear licencing.
Ben.
--
Ben Hutchings
The generation of random numbers is too important to be left to chance.
- Robert Coveyou
Attachment:
signature.asc
Description: This is a digitally signed message part