On 04/30/2016 03:16 AM, Paul Rogers wrote:
This could be a problem with that ethernet interface not an issue
with udev. Take a look at this:
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=575805
Indeed, slightly different final octets with his board, but otherwise
looks like the same problem! Thanks for finding that, I didn't hit
the right keywords.
Yes, they are to be different. A MAC address is composed of 3 bytes
after the manufacturer and the rest is to be a unique number for each
specific interface they make.
I've just minutes ago cloned the system to another box, Biostar
MoBo/Conroe CPU/P43 Chipset/RTL8169 NIC, yet to be tested.
OK, I've gone through my normal initial testing as I did with the VIA
Mini-ITX boards. None of this udev addressing trouble found, whether
or not I booted the kernel built without networking support or later
with it. The rules file had the same MAC address. At least with some
initial boots that turned it up before.
So it's looking like it's a faulty interaction between the board/BIOS,
kernel and udev somewhere.
Different machine, different results. I think your are dealing with a
very specific issue. Probably the BIOS already gets the bad number from
the physical interface on boot. Udev is your friend here.
Why don't you try that? Yours could read something like this:
ATTR{address}=="??:??:??:??:96:06"
If you confirm that issue in your machine, this could handle the
zeroing of the first four octets.
Now I'm going to get the VIA systems out again and see if your
suggestion will make the problem go away! At the end of the day, that's
what I want.
Not mine, Bruce's one! I wasn't even aware that you can set matching
patterns in udev rules! So I take something new for me too... ;-)
Good luck.
Alz.
--
http://lists.linuxfromscratch.org/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page
