1 |
On Thu, 2005-07-07 at 13:52 -0700, Greg KH wrote: |
2 |
> On Thu, Jul 07, 2005 at 03:55:45PM +0200, Martin Schlemmer wrote: |
3 |
> > On Wed, 2005-07-06 at 15:46 -0700, Greg KH wrote: |
4 |
> > > Ok, now that devfs is removed from the 2.6 kernel tree[1], I think it's |
5 |
> > > time to start to revisit some of the /dev naming rules that we currently |
6 |
> > > are living with[2]. |
7 |
> > > |
8 |
> > > To start with, the 061 version of udev offers a big memory savings if |
9 |
> > > you use the "default" kernel name of a device[3]. If you do that, it does |
10 |
> > > not create a file in its database in /dev/.udevdb/ |
11 |
> > > |
12 |
> > > If we can move away from some of our devfs-like names, we stand to |
13 |
> > > reclaim a lot of memory from everyone's machines. As an example, if we |
14 |
> > > drop all of the tty/pts/vc/vcc symlinks, and just go with the default |
15 |
> > > kernel name, we save 2.5Mb of space in tempfs/ramfs. I've done this on |
16 |
> > > my machines and everything seems to work just fine (it looks like |
17 |
> > > everything that was trying to use a tty node was just using the symlink |
18 |
> > > anyway.) |
19 |
> > > |
20 |
> > > So, anyone have any objections to me changing the default udev naming |
21 |
> > > scheme in this manner? |
22 |
> > > |
23 |
> > |
24 |
> > Fine with me. I assume we will need to keep the rcscript support for |
25 |
> > those die-hard 2.4 users still, but hopefully we can eventually drop |
26 |
> > that as well. |
27 |
> |
28 |
> What rcscript support? |
29 |
> |
30 |
|
31 |
Err, sorry, all the crap in /sbin/rc ... |
32 |
|
33 |
> > > Next up, that loony block device naming scheme (more on that later...) |
34 |
> > > |
35 |
> > |
36 |
|
37 |
> > > [3] HAL needs a patch to be able to handle this. It's posted on the |
38 |
> > > hal development mailing lists and will be checked in real-soon-now. |
39 |
> > |
40 |
> > I just think we need to make sure this is in first ... |
41 |
> |
42 |
> The HAL patch? It's now in HAL's cvs tree, don't know when they will do |
43 |
> a new release. |
44 |
> |
45 |
|
46 |
Well, you did provide the patch, so hopefully foser or somebody else |
47 |
will just add it. Foser ping ... |
48 |
|
49 |
> > Lastly on an unrelated note ... I have a rule: |
50 |
> > |
51 |
> > ----- |
52 |
> > # cat /etc/udev/rules.d/40-dm.rules |
53 |
> > KERNEL="dm-[0-9]*", PROGRAM="/sbin/devmap_name %M %m", NAME="mapper/%c", SYMLINK="%c" |
54 |
> > ----- |
55 |
> > |
56 |
> > And in theory it should be the last rule to set the name ... however the |
57 |
> > default one in 50-udev.rules overrides it, and I have to add |
58 |
> > OPTIONS="last_rule" |
59 |
> |
60 |
> Yes. |
61 |
> |
62 |
> Want me to just change the default rule to yours? |
63 |
> |
64 |
|
65 |
I do not think that will work, as that is not distributed with either |
66 |
udev or device-mapper, but multipath-tools (sorda silly if you ask me, |
67 |
as I think it would have been more appropriate with device-mapper, but |
68 |
what the hey). |
69 |
|
70 |
Anyhow, I'll see if I can hack a patch or something up so that NAME="" |
71 |
will also be seen as as a rule that 'set the name' .... |
72 |
|
73 |
|
74 |
Thanks, |
75 |
|
76 |
-- |
77 |
Martin Schlemmer |
78 |
Gentoo Linux Developer, Desktop/System Team Developer |
79 |
Cape Town, South Africa |