Re: Suspend to disk fails after Squeeze upgrade
On Fri, 25 Feb 2011 10:56:26 -0800, Richard Lawrence wrote:
> Thanks, Camaleón, for these suggestions.
>
>> Make a quick and easy test: create a new user and try to
>> hibernate/resume from there and see how it goes.
>
> Ok, here's where I'm at:
>
> 1) A new user (pmtest) can indeed hibernate and resume from the console
> without issue
>
> 2) My regular user account (rwl) can hibernate and resume from the
> console without issue
Humm... the fact that you can hibernate/resume fine when you trigger the
command from console can be "relevant" but I cannot see the reason.
Look, if we read this doc:
http://wiki.debian.org/Suspend
It seems that when using (when it's installed) "pm-utils", and you
trigger the hibernation command from any of the menus, pm-utils is
called, the same way as you do when running from command line. So I don't
know what can make the difference here :-?
When you run the pm-suspend command, were you in a tty or X console?
> 3) pmtest can hibernate the system from within Gnome in X, but a kernel
> error results on resume -- though the system recovers. Details below.
(...)
> Kernel failure message 1:
> BUG: unable to handle kernel paging request at a63b404f
> IP: [<08a5bc0a>] 0x8a5bc0a
> *pde = 00000000
> Oops: 0002 [#1] SMP
(...)
> Pid: 2925, comm: pm-hibernate Not tainted (2.6.32-5-686 #1)
(...)
> <0> eede2800 00007bbf 20200a00 206c6e20 Call Trace:
> Code: Bad EIP value.
> EIP: [<08a5bc0a>] 0x8a5bc0a SS:ESP 0068:de02ffcf CR2: 00000000a63b404f
> ---[ end trace 0e93aec76eafb3ca ]---
Thanks for the detailed logs :-)
I'd say that's good.
Well, no kernel oops is "good" but if you can reproduce this kernel bug
every time you trigger the hibernation from GNOME when resuming the
system, at least you can open a bug for this in Debian BTS. You are
clearly having some sort of problem with the kernel.
OTOH, because of the weird behaviour you are experiencing with the delay
on booting, I would also look for any BIOS update.
Greetings,
--
Camaleón
Reply to: