1 |
Adrian Middleton <lists@×××××××××××.net> posted |
2 |
20061021073901.3e81782e@×××××××××××××××××××××.net, excerpted below, on |
3 |
Sat, 21 Oct 2006 07:39:01 -0400: |
4 |
|
5 |
> So I have been using this same setup perfectly for like a year until |
6 |
> yesterday when I updated my system and rebooted to find that init was |
7 |
> hanging when trying to mount my swap. It appears the problem occured |
8 |
> because of something to do with a cryptsetup upgrade. I forgot to run |
9 |
> dispatch-conf after the system update and found that my system would hang |
10 |
> very early in init with 'Warning: exhausting read requested, but key file |
11 |
> is not a regular file, function might never return' and an invalid |
12 |
> variable 'type=luks' error. So I boot the live cd and mount the root fs |
13 |
> and run dispatch-conf and find that indeed there is no longer a 'type' |
14 |
> variable. |
15 |
|
16 |
You don't mention what version of that or baselayout you are using, and |
17 |
I'm not running encrypted swap tho I've always thought it'd be nice to |
18 |
setup /someday/, so I can't help directly. However, running ~amd64, I've |
19 |
discovered that it's wise to run emerge --pretend --changelog <package> |
20 |
for packages such as baselayout and portage, every time they want to |
21 |
update, just to get an idea what's changing and how it might affect me. |
22 |
It appears you didn't do this and were taken by surprise, particularly so |
23 |
having forgotten to dispatch-conf or etc-update. |
24 |
|
25 |
Anyway, I /did/ see some changes related to that in the last baselayout |
26 |
updates IIRC. I'd suggest you check bugzilla and see what's up. If |
27 |
there's not bugs filed on the issue already, perhaps you'll want to file |
28 |
one, as the number of folks using that feature is going to be relatively |
29 |
small, and maybe you're the first to run into the issue. I know I use a |
30 |
couple corner-case modules (mainly macchanger) and a couple local |
31 |
modifications (to the checkroot and checkfs scripts), and have been the |
32 |
first to come across issues in the updates and file bugs on them, a few |
33 |
times. |
34 |
|
35 |
It's worth keeping in mind the distinction between ordinary packages where |
36 |
Gentoo just packages and distributes the work of upstream, and packages |
37 |
such as portage and baselayout where Gentoo /is/ upstream. On the former, |
38 |
by the time something gets into ~arch, it has normally been reasonably |
39 |
well tested upstream and is a stable candidate. On the latter, Gentoo's |
40 |
own development, ~arch /is/ the testing altho they still aren't going to |
41 |
release stuff that isn't working for them. Thus, an ~arch user anyway can |
42 |
expect to come across a few issues the devs didn't have the config to |
43 |
test, from time to time. Because these packages are by definition core |
44 |
Gentoo packages as well, it's worth taking a bit of extra time to find out |
45 |
just what is changing and be prepared for issues in those areas as one |
46 |
upgrades, from time to time. |
47 |
|
48 |
-- |
49 |
Duncan - List replies preferred. No HTML msgs. |
50 |
"Every nonfree program has a lord, a master -- |
51 |
and if you use the program, he is your master." Richard Stallman |
52 |
|
53 |
-- |
54 |
gentoo-amd64@g.o mailing list |