1 |
On Thursday 22 February 2007 01:45, Alan McKinnon wrote: |
2 |
> On Thursday 22 February 2007, Boyd Stephen Smith Jr. wrote: |
3 |
> > On Wednesday 21 February 2007, "pat" <pat@××××××××.org> wrote about |
4 |
> > 'Re: |
5 |
> > |
6 |
> > [gentoo-user] ramfs - is it necessary ???': |
7 |
> |
8 |
> First, I think the OP is confused between ramfs and initramfs.Not quite |
9 |
> the same thing... |
10 |
|
11 |
Yeah, I hoped I cleared that up with my first reply. |
12 |
|
13 |
> But the thread has become about initramfs so we'll stick with that |
14 |
|
15 |
I think this is more what the OP was concerned about. |
16 |
|
17 |
> And the OP should keep in mind that the initrd format was dumped many |
18 |
> many kernel versions ago and these days we use initramfs, |
19 |
|
20 |
I am fairly certain I was still using my custom initrd (not an initramfs) |
21 |
until 2.6.17 -- I'm fairly sure 2.6.20 still *supports* initrd format, even |
22 |
if initramfs is preferred now. |
23 |
|
24 |
For the life of me, I always found it easier to get an initrd working |
25 |
rather |
26 |
than an initramfs -- the whole chroot/exec vs. pivot_root vs. switch_root |
27 |
step always failed for me when using an initramfs (and the very same shell |
28 |
script worked as an initrd). Also, a script-made initrd is still just a |
29 |
compressed filesystem, easy to deal with, but a script-made initramfs |
30 |
(particularly one made by genkernel) is not just a cpio archive, it's a |
31 |
series of them separated by some !!!!KERNEL_MAGIC!!!! strings in the middle |
32 |
of binary data -- nearly impossible to work with using standard tools. |
33 |
|
34 |
At least, that's been my experience, others may have found the process |
35 |
easier. |
36 |
|
37 |
-- |
38 |
Boyd Stephen Smith Jr. ,= ,-_-. =. |
39 |
bss03@××××××××××.net ((_/)o o(\_)) |
40 |
ICQ: 514984 YM/AIM: DaTwinkDaddy `-'(. .)`-' |
41 |
http://iguanasuicide.org/ \_/ |
42 |
New GPG Key! Old key expires 2007-03-25. Upgrade NOW! |