1 |
On 10/27/2019 06:06, James Le Cuirot wrote: |
2 |
> On Sun, 27 Oct 2019 05:38:48 -0400 |
3 |
> Joshua Kinard <kumba@g.o> wrote: |
4 |
> |
5 |
>> Why do I not like an initramfs, though? Well, for one, it complicates the |
6 |
>> kernel compiles (and it makes them bigger, something which is an issue on |
7 |
>> the old SGI systems at times). Two, it's another layer that I have to |
8 |
>> maintain. Three, it violates, in my mind, the simplicity of keeping the |
9 |
>> kernel and userland separated (e.g., kernel does kernel-y things, userland |
10 |
>> does userland-y things). |
11 |
> |
12 |
> You make it sound like the initramfs has to be built into the kernel |
13 |
> image. It can be but it usually isn't. I suspect you know that though? |
14 |
> Admittedly that does depend on support from your bootloader. While GRUB |
15 |
> and U-Boot have supported this for years, I forget what oddball |
16 |
> bootloaders your hardware may be using. |
17 |
|
18 |
As mentioned (and later pointed out in the thread), I use LILO. Considering |
19 |
upstream is dead now, how much longer that will persist is an open question. |
20 |
I am a creature very wedded to habit, and LILO, for whatever its ills and |
21 |
warts are, has always just worked for me, so I have stuck with it. Having |
22 |
to edit a config file before each kernel update is the trivialest of tasks, |
23 |
especially since I tend to only update the kernel once every few months. |
24 |
|
25 |
The SGI systems all boot using netboot because I compile their kernels on my |
26 |
dev box and then host them over TFTP. It's just a lot easier that way, |
27 |
especially since ARCS (the SGI version of a BIOS) has had network booting |
28 |
built in since pretty much 1990 (and earlier, I think). It's really the SGI |
29 |
Indy (IP22) that refuses to netboot if the kernel's size is >11MB. All of |
30 |
those kernels are monolithic, since the hardware in such systems is pretty |
31 |
much fixed for all time. Granted, my Indy gave up the ghost some time ago |
32 |
(probably the infamous RTC battery issue), so I am really just left with an |
33 |
O2 (which refuses to netboot anything >41MB), Octane, and two IP27-class |
34 |
systems, which don't have size limits (that I've found yet) for what they'll |
35 |
boot. |
36 |
|
37 |
For actual disk booting on those systems, well, there's really just the |
38 |
super-old ArcLoad project. Written by the mastermind behind Octane's Linux |
39 |
port, he aimed to make it very GRUB-like, so it has some of GRUB's |
40 |
filesystem code in it for direct booting, or it can boot kernels stored in a |
41 |
special area of the disk called the volume header (on disks w/ SGI's |
42 |
partition layout). Except that imported GRUB code is beyond ancient (11+ |
43 |
years old?), so it probably won't load a kernel off of ext4 or even XFSv5 |
44 |
partitions now. Thus you really just have the volume header approach, and |
45 |
that has its own limits. |
46 |
|
47 |
That all said, it doesn't really matter if the initramfs is built into the |
48 |
kernel or loaded somewhere into memory by the bootloader. It's still a |
49 |
wholly-separate userland layer that needs its own libc and whatever small |
50 |
amount of tools to help mount required partitions and/or flash CPU microcode |
51 |
before the kernel tries invoking /sbin/init. That's just a terrible design. |
52 |
I mean, you can put lipstick on a pig, but it's still a bloody pig. |
53 |
|
54 |
|
55 |
>> Maybe I'm just a old codger who refuses to accept change. I'm fine with |
56 |
>> that description. I like things to remain somewhat simple, and my view of |
57 |
>> Linux, both kernel and userland, over the last few years is one of growing |
58 |
>> dismay due to the constant introduction of subsystem layer atop subsystem |
59 |
>> layer for very little gain. How much longer until we need a kernel to boot |
60 |
>> the kernel to mount the userland that mounts the userland (yo dawg)? |
61 |
> |
62 |
> Isn't that what the BIOS and bootloader do? On the plus side, you can |
63 |
> now boot straight from UEFI to kernel without a bootloader but on the |
64 |
> other hand, UEFI is horrifically over-complex. |
65 |
|
66 |
Yeah, UEFI is one of those Eldritch horrors spoken of in the Necronomicon. |
67 |
It's really just there so gamerz bois can have their uber-fancy gaming |
68 |
system fully specced out. Ironically, SGI's ARCS was capable of doing most |
69 |
of what UEFI does 20 years ago, including pretty GFX and even playing music |
70 |
when the system boots up (e.g., Indy's infamous "ass trumpet"). |
71 |
|
72 |
-- |
73 |
Joshua Kinard |
74 |
Gentoo/MIPS |
75 |
kumba@g.o |
76 |
rsa6144/5C63F4E3F5C6C943 2015-04-27 |
77 |
177C 1972 1FB8 F254 BAD0 3E72 5C63 F4E3 F5C6 C943 |
78 |
|
79 |
"The past tempts us, the present confuses us, the future frightens us. And |
80 |
our lives slip away, moment by moment, lost in that vast, terrible in-between." |
81 |
|
82 |
--Emperor Turhan, Centauri Republic |