1 |
Hi Matthias, |
2 |
|
3 |
Matthias Gerstner <mgerstner@××××.de> writes: |
4 |
|
5 |
> I think I found the cause of the trouble: |
6 |
> |
7 |
> The group id used comes a long way ... in |
8 |
> |
9 |
> ${EPREFIX}/usr/lib64/python3.4/site-packages/portage/data.py function _init(settings) |
10 |
> |
11 |
> a global variable '_portage_grpname' is initialized from a settings |
12 |
> variable 'PORTAGE_GRPNAME' which is setup in |
13 |
> |
14 |
> ${EPREFIX}/usr/lib64/python3.4/site-packages/portage/package/ebuild/config.py function config._init(...) |
15 |
> |
16 |
> where basically an os.stat() is made on 'eroot_or_parent' which in this |
17 |
> case is ${EPREFIX}. |
18 |
> |
19 |
> And in my case the ${EPREFIX} was group-owned by root. Doing a |
20 |
> |
21 |
> chown "${EPREFIX}" <my-group> |
22 |
> |
23 |
> fixed the situation. The stage3 went through now. |
24 |
|
25 |
Thank you for your time trouble-shooting. Portage should have been |
26 |
smarter in such cases. |
27 |
|
28 |
By the way, I have set up a openSUSE 42.1 yesterday on lxc and tested |
29 |
out the bootstrap. It also worked well. Appended an openSUSE entry to |
30 |
the table: |
31 |
|
32 |
https://wiki.gentoo.org/wiki/Prefix/libc#Tested_Distributions |
33 |
|
34 |
Yours, |
35 |
Benda |