1 |
On 11/08/2020 15.38, Joonas Niilola wrote: |
2 |
> |
3 |
> On 8/11/20 11:36 AM, Jaco Kroon wrote: |
4 |
>> And I've already provided you one use case where udev doesn't work well |
5 |
>> but eudev does. I've also mentioned some historic issues I believe |
6 |
>> should already be fixed but which did bit me in systemd-udev which was |
7 |
>> never a problem in eudev. |
8 |
>> |
9 |
> Your systems seem to diverge a lot from what I'd consider as 'default'. |
10 |
> You must already make many changes to them. |
11 |
> |
12 |
> Before this thread I didn't even know/remember I was using eudev. It |
13 |
> works and there doesn't seem to be any global issues related to it. |
14 |
> However after reading this thread I'm a bit considered about the |
15 |
> maintenance state of it. |
16 |
> |
17 |
> Switched to udev. Simple as 'emerge -1 sys-fs/udev'. It works, didn't |
18 |
> notice any difference. |
19 |
> |
20 |
> If musl is a problem, to my knowledge musl has its own stage images, so |
21 |
> why can't those stages use eudev while other ones defaulting to udev? |
22 |
|
23 |
For the sake of science, I've also moved one of my systems to |
24 |
sys-fs/udev and noticed a single incompatibility. There are few ways how |
25 |
to disable the systemd/udev predictable NIC names, one of them is to |
26 |
boot with net.ifnames=0, another is to mask rule file that trigger the |
27 |
rename, as described on wiki[1] |
28 |
|
29 |
Long story short, on eudev it's 80-net-name-slot.rules, on udev it's |
30 |
80-net-setup-link.rules |
31 |
|
32 |
The result was that my system booted without working network, as connman |
33 |
started to poke around Ethernet interfaces (this is ok, I had |
34 |
blacklisted eth*, not enp*), which then resulted in switching routing to |
35 |
Ethernet that failed to get IP from DHCP, even that WiFi had a fully |
36 |
working Internet access, so more like connman bug. (And yes, I am aware |
37 |
that net.ifnames=0 will work on both) |
38 |
|
39 |
This however shows two things: eudev is (no longer) drop-in replacement, |
40 |
as some interfaces changed in upstream udev, which leads to second |
41 |
thing, we need to have migration path, because even if(!) Gentoo change |
42 |
default (I am not a fan of this idea really), then it might lead to |
43 |
people doing fresh installation or reinstallation, migrating their |
44 |
configs resulting in not working systems/working in other way that |
45 |
previous one. |
46 |
|
47 |
[1] https://wiki.gentoo.org/wiki/Udev#Keep_classic_.27eth0.27_naming |
48 |
|
49 |
-- Piotr. |