1 |
people actually need an initramfs? |
2 |
|
3 |
my kernel has only what it needs, and nothing it doesn't. |
4 |
|
5 |
On Wed, Mar 21, 2012 at 5:39 AM, Pandu Poluan <pandu@××××××.info> wrote: |
6 |
> |
7 |
> On Mar 21, 2012 4:23 PM, "Halassy Zoltán" <zhalassy@×××××××.hu> wrote: |
8 |
>>> |
9 |
>>> IMO, initramfs adds yet another black box during server boot. |
10 |
>> |
11 |
>> |
12 |
>> The other way around, for me at least. I build my own initramfs, yet I |
13 |
>> don't know anything about mdev, just the fact it's part of busybox. So for |
14 |
>> me, mdev is a black box, while my initramfs definitely isn't. |
15 |
>> |
16 |
>> |
17 |
> |
18 |
> I see. Well, different views for different people, I guess. |
19 |
> |
20 |
> It's easier for me to bypass mdev (if it's b0rken) than to bypass initramfs. |
21 |
> |
22 |
>>> And yet |
23 |
>>> another daemon in memory, something I certainly don't need on my static |
24 |
>>> virtualized servers. |
25 |
>> |
26 |
>> |
27 |
>> I agree with that. But why do you need mdev for a static system? A few |
28 |
>> mknods would suffice. |
29 |
>> |
30 |
> |
31 |
> It allows triggered action when I (for example) attach a (virtual) hard disk |
32 |
> to my VM. |
33 |
> |
34 |
> Rgds, |