1 |
Michael Scherer wrote: |
2 |
> On Mon, 23 Apr 2012 18:01:09 -0500 |
3 |
> Dale <rdalek1967@×××××.com> wrote: |
4 |
> |
5 |
>> Michael Scherer wrote: |
6 |
>>> On Mon, 23 Apr 2012 13:26:57 -0500 |
7 |
>>> Dale <rdalek1967@×××××.com> wrote: |
8 |
>>> |
9 |
>>>> Michael Mol wrote: |
10 |
>>>>> So, my habit is to have /usr sit on top of LVM on top of mdraid. I |
11 |
>>>>> really don't want to get into the business of manually managing my |
12 |
>>>>> own initramfs, and udev >= 181 will eventually hit stable. I want |
13 |
>>>>> my systems ready before that. |
14 |
>>>>> |
15 |
>>>>> Now, having looked at the pertinent documentation, it looks like |
16 |
>>>>> the command I need to run is: |
17 |
>>>>> |
18 |
>>>>> genkernel --lvm --mdadm --disklabel initramfs |
19 |
>>>>> |
20 |
>>>>> and then add the initramfs to my grub setup. |
21 |
>>>>> |
22 |
>>>>> The trouble is, genkernel dies on me. I tried for some feedback in |
23 |
>>>>> #gentoo-chat, and DrEeevil gave me two clues: |
24 |
>>>>> |
25 |
>>>>> 13:34 <+DrEeevil> that shouldn't even be possible |
26 |
>>>>> 13:36 <+DrEeevil> mikemol: looks like static linking didn't on |
27 |
>>>>> your system |
28 |
>>>>> |
29 |
>>>>> I don't know where to take things from here. I'm hesitant to file |
30 |
>>>>> a bug on b.g.o, because the tail end of genkernel.log specifically |
31 |
>>>>> says not to file build errors as bug reports. |
32 |
>>>>> |
33 |
>>>>> So...any ideas? This is an amd64 mostly-stable system, and dracut |
34 |
>>>>> is still masked on amd64, which is why I'm trying to get genkernel |
35 |
>>>>> working. |
36 |
>>>>> |
37 |
>>>>> I've attached genkernel.log |
38 |
>>>>> |
39 |
>>>> |
40 |
>>>> |
41 |
>>>> If I read the -dev mailing list correctly, they plan to still |
42 |
>>>> support /usr without a init thingy. After all the mess I went |
43 |
>>>> through, we may not need the init thingy after all. |
44 |
>>>> |
45 |
>>>> Go figure. |
46 |
>>>> |
47 |
>>>> Dale |
48 |
>>>> |
49 |
>>>> :-) :-) |
50 |
>>>> |
51 |
>>> |
52 |
>>> I wouldn't bet on that. |
53 |
>>> |
54 |
>>> But there is a detailed gentoo howto to create an iniramfs |
55 |
>>> that does just that: load /usr. |
56 |
>>> http://wiki.gentoo.org/wiki/Early_Userspace_Mounting |
57 |
>>> It works. You only need tell your bootloader that now you |
58 |
>>> have an initramfs. |
59 |
>>> There's also a more general introduction |
60 |
>>> http://en.gentoo-wiki.com/wiki/Initramfs |
61 |
>>> for a lot of other options |
62 |
>>> |
63 |
>>> michael |
64 |
>>> |
65 |
>>> |
66 |
>> |
67 |
>> |
68 |
>> I'm not betting on it but that is what the Gentoo council voted on and |
69 |
>> it said that /usr on a separate partition was going to be supported. |
70 |
>> |
71 |
>> You must have missed the HUGE thread where I was trying to get a init |
72 |
>> thingy to work huh? |
73 |
>> |
74 |
>> Dale |
75 |
>> |
76 |
>> :-) :-) |
77 |
>> |
78 |
> |
79 |
> Indeed, I missed all but one email, but there was no hint |
80 |
> at all that anything had been said before I came in. |
81 |
> Sorry for my unsolicited comments, and I hope you have |
82 |
> solved that problem in one or another way. |
83 |
> |
84 |
> michael |
85 |
> |
86 |
> |
87 |
> |
88 |
|
89 |
|
90 |
I meant to put a "LOL" after the comment about the huge thread. I'm not |
91 |
sure how I missed that. :/ I did get it to work finally. It took |
92 |
several peoples help and a hammer on my part. ;-) |
93 |
|
94 |
I'm just curious as to how they are going to support this in the future. |
95 |
I have read where some say it can be done but others say it can't, or |
96 |
not easily. Old saying, this is where the rubber meets the road. ^_^ |
97 |
|
98 |
|
99 |
To the OP. I would use dracut. I ran into trouble but I found out |
100 |
later that a earlier attempt at a init thingy was causing the dracut |
101 |
init thingy to mess up. I strongly suspect that if I had known to get |
102 |
rid of the previous attempt, it would have worked the first time. My |
103 |
previous attempt was the one that was built into the kernel itself. |
104 |
Lets just say there was a huge fight and I missed it. Grub was telling |
105 |
one thing to load and the kernel was trying to load something else. I'm |
106 |
sure it was a nice fight. |
107 |
|
108 |
Dale |
109 |
|
110 |
:-) :-) |
111 |
|
112 |
-- |
113 |
I am only responsible for what I said ... Not for what you understood or |
114 |
how you interpreted my words! |
115 |
|
116 |
Miss the compile output? Hint: |
117 |
EMERGE_DEFAULT_OPTS="--quiet-build=n" |