1 |
Rich Freeman wrote: |
2 |
|
3 |
> On Wed, Apr 11, 2012 at 11:09 AM, Steven J Long wrote: |
4 |
>> That might be true for some Linux-only packages, but I really find it |
5 |
>> hard to believe that any upstream targetting more than one OS (just |
6 |
>> adding a BSD is enough) with software that could be considered critical |
7 |
>> (I for one would include all POSIX utilities as well as basic stuff like |
8 |
>> mount, fsck and lvm2) would want to ignore this kind of thing; if the |
9 |
>> build-system is ignoring configuration, it's a bug. |
10 |
> |
11 |
> The issue is that if udev requires libfoo, then libfoo must not be in |
12 |
> /usr. If libfoo is libx11 or dbus or some other complex library, that |
13 |
> will pull in a bunch of other stuff as well. However, I doubt that |
14 |
> the maintainers of all those libraries would consider them |
15 |
> boot-essential, so they may not be inclined to make the move. |
16 |
> |
17 |
> Obviously we're not there now, but it is a possible consequence of |
18 |
> moving in this direction. |
19 |
> |
20 |
Sure, but I'm curious: how would you set up an initramfs under those |
21 |
conditions? |
22 |
|
23 |
Where I'm going with this, is that in both cases (an initramfs or a |
24 |
traditional rootfs system) we need to be aware of what libs pull in what. |
25 |
Given that, there is actually common work both sides need. |
26 |
|
27 |
If we could focus on that, we'd all actually be cooperating to fix things |
28 |
for both setups, instead of arguing about which is best. :) |
29 |
|
30 |
> Keep in mind that systemd in particular does not aim to be |
31 |
> cross-platform - they advertise this as one of their core features. |
32 |
> Since tight integration is their goal that could slowly bring in a lot |
33 |
> of other stuff. The main platform pushing it along is Fedora, and |
34 |
> they're aiming to move everything into /usr, so this is a non-issue |
35 |
> for them. |
36 |
> |
37 |
I have a feeling that "integration" is being used as an excuse to avoid |
38 |
thinking about coupling and cohesion. |
39 |
|
40 |
>> I read the decision from the Council to be "we will continue to support |
41 |
>> the traditional split /usr eg with lvm, and no initramfs" and a Council |
42 |
>> member put himself forward to maintain patches to udev to ensure that |
43 |
>> going forward, since it is needed in his employment. |
44 |
>> |
45 |
>> Given that we can do it with initscripts, and don't need to fork udev at |
46 |
>> all, what's the problem? |
47 |
> |
48 |
> I can't really comment on what the decision from the Council actually |
49 |
> was. |
50 |
|
51 |
Well, I've stated that several times now, and included the snippet from the |
52 |
log where Betelgeuse specifically asked Chainsaw to clarify that a |
53 |
"universal" minimal initramfs was not good enough, which was confirmed. |
54 |
|
55 |
If Council members disagree with that interpretation, I'd welcome their |
56 |
clarification. |
57 |
|
58 |
Split /usr with lvm was specifically discussed as a use-case, since it has |
59 |
been outlined in Gentoo documentation. |
60 |
|
61 |
> However, maintaining patches to udev is effectively the same |
62 |
> thing as forking it. Right now it probably isn't hard, and over time |
63 |
> that could change. |
64 |
> |
65 |
> The only time patches != fork is if the patches have been submitted |
66 |
> upstream and are likely to be merged. |
67 |
> |
68 |
udev != openrc or any other init system, which is what we are patching[1] so |
69 |
that udev is not started til after localmount, should the user set a |
70 |
currently unsupported udev.conf option (initramfs="no"). |
71 |
|
72 |
We're only making minor shell-script modifications to udev and udev-mount |
73 |
initscripts, not openrc itself. |
74 |
|
75 |
Earlymounts[2] is simply an additional initscript. |
76 |
|
77 |
IOW no patches to udev whatsoever, so no fork. |
78 |
|
79 |
> In any case, it isn't a crisis now and waiting a little to see which |
80 |
> way the wind ends up blowing probably makes sense. |
81 |
> |
82 |
No indeed: those of us who wish to stay with our traditional setup, who have |
83 |
already configured our machines to localmount with built-in modules, can use |
84 |
the patched initscripts/earlymounts. What we'd like is for those, or |
85 |
equivalent functionality, to be maintained within Gentoo so we don't have to |
86 |
tweak patches whenever udev-init-scripts is updated, or in the earlymounts |
87 |
case there appear to be more issues, and those could use input from devs |
88 |
imo. |
89 |
|
90 |
I prefer the patched approach, even if it is a bit more setup and I am |
91 |
biased, since it appears to have less potential for interaction with other |
92 |
stuff: if you never needed an initramfs before, and can localmount with just |
93 |
kernel-created device-nodes (eg: you need to change fstab from using |
94 |
/dev/vg/lv to /dev/mapper/vg-lv for lvm), then you're fine. |
95 |
|
96 |
[1] http://forums.gentoo.org/viewtopic-t-901206.html |
97 |
[2] http://forums.gentoo.org/viewtopic-t-918466.html |
98 |
-- |
99 |
#friendly-coders -- We're friendly, but we're not /that/ friendly ;-) |