1 |
On Wednesday 29 November 2006 17:33, Richard Fish wrote: |
2 |
> On 11/27/06, Mrugesh Karnik <mrugeshkarnik@×××××.com> wrote: |
3 |
> > On Tuesday 28 November 2006 07:31, Richard Fish wrote: |
4 |
> > > > can see a 75-persistent-net-generator.rules file in there.. |
5 |
> > > |
6 |
> > > Hmm, not sure how I got a 70-persistent-net.rules. There is some |
7 |
> > > interaction between that and 75-persistent-net-generator.rules (and |
8 |
> > > the /lib/udev/write_net_rules script), but I'm a bit too tired to |
9 |
> > > figure it out ATM. It looks like 70-... should be created by the |
10 |
> > > write_net_rules script... |
11 |
> > |
12 |
> > RULES_FILE='/etc/udev/rules.d/70-persistent-net.rules' |
13 |
> > |
14 |
> > That's the first line of write_net_rules. |
15 |
> |
16 |
> Right. I just wasn't able to figure out why you didn't already have |
17 |
> this file created, nor why my laptop had it but not my desktop. |
18 |
> |
19 |
> So the story is that 75-persistent-net-generator.rules will call the |
20 |
> script when ethernet devices are added, and it is up to the |
21 |
> write_net_rules script to generate 70-persistent-net.rules. The |
22 |
> problem is that when udev starts very early in the boot process, your |
23 |
> root filesystem may still be mounted read-only, preventing this file |
24 |
> from being created. |
25 |
> |
26 |
> This worked on my laptop, because I added module aliases to prevent |
27 |
> udev from coldplugging the ipw3945 driver, since it requires a daemon |
28 |
> to be running in order to work and that required /var to be mounted. |
29 |
> The module is loaded later in the boot process, after all of the |
30 |
> filesystems are mounted read-write, and that allowed udev to create |
31 |
> the rules file for me, but only for that adapter. |
32 |
> |
33 |
> The upshot of this is this: by far the easiest way to solve the |
34 |
> net-naming problem is to run |
35 |
> |
36 |
> /lib/udev/write_net_rules all_interfaces |
37 |
> |
38 |
> This will generate the rules for all interfaces, and then you can just |
39 |
> edit the file to change the names as you like. So I guess I'll know |
40 |
> that for the next person that asks. :-P |
41 |
|
42 |
Not sure if/how it is related to the OP, but this is what was created in |
43 |
my /etc/udev/rules.d/70-persistent-net.rules: |
44 |
===================================== |
45 |
# USB device 0x050d:0x7050 (rt2500usb) |
46 |
SUBSYSTEM=="net", DRIVERS=="?*", ATTRS{address}=="00:11:50:18:55:3f", |
47 |
ATTRS{type}=="1", NAME="wlan0" |
48 |
===================================== |
49 |
|
50 |
However, if I boot with the USB WiFi adaptor plugged in it, the device is not |
51 |
being detected. If I plug it in after the system has booted then there is no |
52 |
problem. The USB devices are 'udev-plugged' relatively late in the boot |
53 |
process, well after udevd has started. Therefore I cannot understand why |
54 |
this adaptor is not being detected. Any ideas? |
55 |
-- |
56 |
Regards, |
57 |
Mick |