1 |
On Sun, 7 Feb 2010 02:20:19 -0800 |
2 |
James Ausmus wrote: |
3 |
|
4 |
> On Sat, Feb 6, 2010 at 8:07 PM, David Relson |
5 |
> <relson@×××××××××××××.com>wrote: |
6 |
> |
7 |
> > On Sat, 6 Feb 2010 19:13:33 -0500 |
8 |
> > Willie Wong wrote: |
9 |
> > |
10 |
> > > On Sat, Feb 06, 2010 at 06:29:27PM -0500, David Relson wrote: |
11 |
> > > > Your replies are much appreciated as we're in an area of Linux |
12 |
> > > > about which I'm poorly informed. |
13 |
> > > > |
14 |
> > > > Output (below) of "rc-status sysinit" indicated devfs stopped, |
15 |
> > > > so I started devfs (which didn't change /dev/pt*), then |
16 |
> > > > restarted udev (which didn't affect /dev/pt*). |
17 |
> > > |
18 |
> > > Right, but can you ssh in to the machine now (or open a terminal |
19 |
> > > emulator in X)? |
20 |
> > > |
21 |
> > > /dev/pts is just the mount point for the devpts pseudo |
22 |
> > > filesystem. In modern versions of linux the pts devices are |
23 |
> > > created on-the-fly when requested (as opposed to other versions |
24 |
> > > and some modern unixes where there will be a fixed number of |
25 |
> > > device nodes under /dev/pts or equivalent). All that just goes to |
26 |
> > > say that if /dev/pts is empty right after you restart the devfs |
27 |
> > > service, it is normal. A device file should be created |
28 |
> > > automatically now when userspace programs demand it. (E.g. if you |
29 |
> > > now ssh in, and if it succeeds, ls /dev/pts should show one |
30 |
> > > entry.) |
31 |
> > > |
32 |
> > > Try it, let me know if the problem is still there. |
33 |
> > |
34 |
> > Nope. Both ssh and X terminal emulators are still broken. No |
35 |
> > change in behavior. |
36 |
> > |
37 |
> > FWIW, most of the entries in /dev are timestamped 02/02 23:34 which |
38 |
> > is when I updated udev earlier this week. Today's upgrade/downgrade |
39 |
> > emerge hasn't affected the timestamps. |
40 |
> > |
41 |
> > A comparison of /etc/udev/rules.d to a saved copy didn't show |
42 |
> > much. The only puzzling difference is: |
43 |
> > --- 90-hal.rules (revision 51) |
44 |
> > +++ 90-hal.rules (working copy) |
45 |
> > @@ -1,2 +1,2 @@ |
46 |
> > # pass all events to the HAL daemon |
47 |
> > -RUN+="socket:/org/freedesktop/hal/udev_event" |
48 |
> > +RUN+="socket:@/org/freedesktop/hal/udev_event" |
49 |
> > |
50 |
> > removing the "@" and restarting udev hasn't helped. Since the rule |
51 |
> > is hal related, I also restarted hald -- which hasn't helped. |
52 |
> > |
53 |
> > |
54 |
> What happens if you do: |
55 |
> |
56 |
> mount -t devpts none /dev/pts |
57 |
> |
58 |
> Does the problem go away? |
59 |
> |
60 |
> -James |
61 |
|
62 |
Eureka! Problem fixed. |
63 |
|
64 |
Looking in /etc/mtab, the last line is: |
65 |
|
66 |
none /dev/pts devpts rw 0 0 |
67 |
|
68 |
Perhaps the mount devpts command should have been issued as part of |
69 |
emerging udev, openrc, or sysinit ??? Should this be reported to |
70 |
b.g.o.?? |
71 |
|
72 |
David |