Re: Hibernate fails
On 27/06/11 05:36, Carl Fink wrote:
> On Mon, Jun 27, 2011 at 12:47:29AM +1000, Scott Ferguson wrote:
> /etc/network/interfaces is:
> # This file describes the network interfaces available on your system
> # and how to activate them. For more information, see interfaces(5).
> # The loopback network interface
> auto lo
> iface lo inet loopback
> # The primary network interface
> allow-hotplug eth0
> #NetworkManager#iface eth0 inet dhcp
Thanks - that seems to confirm what I've read....
> I use WPA. Works great.
> I added your wpa_supplicant.conf. Still can't hibernate. I don't have WiFi
> here to try to connect to. (I'm at work, where I have to connect this
> netbook using T-Mobile via bluetooth tethering.)
No worries. I "suspect" what is happening is that part of the process
for hibernation is to test network states in order to resume them after
hibernation - and that when attempting to test the state of wireless
there is a problem.
My, limited, understanding is that turning off/disabling wifi and/or
rmmod-ing the related modules will have no effect on the hibernate hook
that is failing (I and suspect that's the fail point in hibernation)
"Running hook /usr/lib/pm-utils/sleep.d/60_wpa_supplicant hibernate
Failed to connect to wpa_supplicant - wpa_ctrl_open: No such file or
The error message seems to indicate that:-
pm-utils is looking for /etc/wpa_supplicant/wpa_supplicant.conf (which
you now have) but is not finding a path to wpa_crtl in that file....
(see my question further down about /var/run)
> While testing this, I discovered not one but two new bugs.
> One, if I use this box's "increase brightness" key-combo (FN-rightarrow) in
> a virtual terminal, it works but also draws control sequences in the VT. And
> two, if I run "/etc/init.d/gdm3 stop" from a VT, then press Alt-F7, I can't
> ever enter any more text, ever. The visual for VT1 continues to display, but
> nothing I do on the keyboard except Ctrl-Alt-Del has any effect.
Hmm - could be that hibernate is not working because of a broken bios.
If it's an ACPI issue (hibernation failure) it should be possible to
work around it by either using the hibernate package or tweaking the
pm-utils settings (but I digress). The simplest, partial, solution would
be to remove the wpa_supplicant hook from the hibernate process - which
would mean wireless wouldn't work after hibernation, and wouldn't be
useful for fixing the "bug" in hibernation (for the package, not you).
Having got this far in trying to isolate the problem with pm-utils it
could be worth moving to pm-test (your call).
If you want to start trying pm-test start by confirming that your kernel
is compiled with pm_debug support:-
$ cat /boot/config-`uname -r` | grep -i config_pm_debug
What does lspci identify as your wireless?
*do you have a /var/run/wpa_supplicant??*
"I can't watch TV longer than five minutes without praying for nuclear
~ Bill Hicks