1 |
On 03/20/2012 04:23 AM, Tony "Chainsaw" Vroon wrote: |
2 |
> On 20/03/12 06:48, Ulrich Mueller wrote: |
3 |
>> In two weeks from now, the council will meet again. This is the time |
4 |
>> to raise and prepare items that the council should put on the agenda |
5 |
>> to discuss or vote on. |
6 |
> |
7 |
> Newer udev (180+) is attempting to force /usr to be mounted *very* early |
8 |
> in the boot sequence. No other software has this requirement, and it is |
9 |
> going to break LVM2 users that have a separate /usr. Our official |
10 |
> documentation recommends this, even today: |
11 |
> http://www.gentoo.org/doc/en/lvm2.xml |
12 |
> |
13 |
> Now that it is unmasked and unleashed on ~arch users, this might result |
14 |
> in an attempt to stable it. I would like for the council to decide on |
15 |
> whether a separate /usr is still a supported configuration. |
16 |
|
17 |
To clarify, the question is whether or not we support a separate /usr |
18 |
_without_ mounting it early via an initramfs. |
19 |
|
20 |
> If it is, newer udev can not be stabled and alternatives should be |
21 |
> investigated. |
22 |
|
23 |
A possible compromise would be to use pkg_pretend to check if /usr is a |
24 |
mount point, and die if the user hasn't set a variable or a USE flag to |
25 |
indicate awareness that /usr must be mounted early. |
26 |
|
27 |
> If it isn't, a lot of documentation will have to be |
28 |
> updated. (And an alternative should likely still be provided) |
29 |
-- |
30 |
Thanks, |
31 |
Zac |