udev-110 ->111 changed behaviour for naming net-devices?

Jens Stroebel drifter at bcsoft.de
Wed May 23 03:21:02 PDT 2007

Bryan Kadzban wrote:
> On Tue, May 22, 2007 at 06:03:39PM +0200, Jens Stroebel wrote:
>> On Mon, May 21, 2007 at 06:22:07PM -0400, Bryan Kadzban wrote:
>>> (Why do you want it, by the way?)
>> That's because we install a laptop with a pcmcia-plugged network card.
>> Should someone at a later time change this card (e.g. it is defect),
>> the MAC would change, too.
> Ah.
> Would it work to use by-path persistence?  That's not supported in the
> book or by the udev developers because (e.g.) USB devices change their
> paths way too often.  But if these laptops only have one PCMCIA slot,
> then that might work.

Having tried the below-mentioned patch, I'd say this should pose no
problems. The location of the PCMCIA slot (id-wise) shouldn't change as
far as I understand; and yes, it only has one of them.
As long as nothing changes with regard to the sys filesystem, we should
be OK with these rules even in case of a NIC replacement.

The patch applied cleanly on udev-110 + udev-111 (although I did only
test it running on udev-110 - works fine there).

> I had a preliminary by-path patch for udev somewhere, let me see if I
> can find it...  Ah:
> http://marc.info/?l=linux-hotplug-devel&m=117141213806593&w=2
> however, reading it again, I said that I didn't think it'd work with
> PCMCIA devices.

I cannot second this. By observation, it seems that the NICs we use in
the PCMCIA slot appear to the kernel as a PCI device.

lspci for the card in PCMCIA slot:
04:00.0 Ethernet controller: Abocom Systems Inc RTL8139 [FE2000VX]
CardBus Fast Ethernet Attached Port Adapter (rev 10)

I don't know about "all NICs plugged to PCMCIA slots", but the ones we
use(d) all behave(d) that way.

> If I knew how those devices showed up in sysfs, I could
> probably add it (otherwise you could just treat this patch as a starting
> point).

Regarding the extension towards PCMCIA-slotted NICs, I can give you no
interesting information, as it works right without any changes in our
case. There isn't even an entry under /sys/bus/pcmcia on the laptop.

PS: thx for the patch, fine thing, that
drifter at bcsoft.de
By caffeine alone I set my mind in motion, By the beans of Java
do thoughts acquire speed,  hands acquire shaking,  the shaking
becomes a warning, By caffeine alone do I set my mind in motion

More information about the lfs-dev mailing list