1 |
On 26/09/14 11:47, Samuli Suominen wrote: |
2 |
> On 26/09/14 11:22, Canek Peláez Valdés wrote: |
3 |
>> On Fri, Sep 26, 2014 at 3:07 AM, Neil Bothwick <neil@××××××××××.uk> wrote: |
4 |
>>> On Fri, 26 Sep 2014 05:27:15 +0200, Alan McKinnon wrote: |
5 |
>>> |
6 |
>>>>> I buy machines with one ethernet interface. What I find |
7 |
>>>>> particularly annoying is this doublespeak about calling it |
8 |
>>>>> "predictable". Before the change, it was predicatbly "eth0". Now, |
9 |
>>>>> it's different on every different model. |
10 |
>>>> It's not doublespeak, the interfaces are named exactly according to |
11 |
>>>> where they are on the PCI bus. If you had two interfaces, they show up |
12 |
>>>> to the kernel in random order by time and sometimes eth0/eth1 are nto |
13 |
>>>> the same they were before the reboot. |
14 |
>>> That may be true for PCI devices but not for USB ones. If you unplug a |
15 |
>>> USB device and plug it back into the same port, it will get a different |
16 |
>>> device number. The naming is more predictable, but it's not there yet. |
17 |
>> That doesn't sound right. If unplugging a USB net device and plugging |
18 |
>> it again *in the same port* results in a different device *name*, then |
19 |
>> it is a bug and should be reported; the description of the algorithm |
20 |
>> in [1] sounds like it should get always the same name for the same |
21 |
>> port, unless I'm misunderstanding something. |
22 |
>> |
23 |
>> Regards. |
24 |
>> |
25 |
>> [1] http://cgit.freedesktop.org/systemd/systemd/tree/src/udev/udev-builtin-net_id.c#n51 |
26 |
> I've seen this happening once on a cheap laptop with a stripped down |
27 |
> BIOS I can't |
28 |
> even recall brand for, it had a kludge in the BIOS settings for |
29 |
> hotplugging, turning |
30 |
> it off, allowed the port to remain same, turning it on, some machine |
31 |
> specific code |
32 |
> gets executed and the kernel interprets the same port as different port |
33 |
> |
34 |
> Bad hardware, bad hardware settings, maybe missing exception for that |
35 |
> particular |
36 |
> hardware type in the code that determines the name... I'm not sure, I |
37 |
> don't have |
38 |
> the machine anymore |
39 |
> |
40 |
> - Samuli |
41 |
> |
42 |
|
43 |
Later kernels *can mark interfaces predictable in a new form of |
44 |
metadata*, and udev >= 209 can |
45 |
pick that information up, and then it won't do anykind of userspace |
46 |
renaming on it, since kernel |
47 |
has declared the interface name to be steady...predictable...always |
48 |
same, so I hope |
49 |
we are moving towards kernel assigning predictable names for all drivers |
50 |
and we can get rid of |
51 |
the userspace renaming of interfaces all together at some point |
52 |
I really believe this is a task for the kernel to provide predictable |
53 |
names, and all this userspace |
54 |
renaming is just a bandage we can hopefully soon rip off |
55 |
|
56 |
- Samuli |