gnome-power-manager when removing or inserting power cable

[11011san]

Binary package hint: gnome-power-manager

was moving the computer and puled out the power cable and when i sate down the power manager informed me that it had crashed.
running Ubuntu 9.10
gnome-power-manager:
  Installed: 2.28.0-0ubuntu3
  Candidate: 2.28.0-0ubuntu3
  Version table:
 *** 2.28.0-0ubuntu3 0
        500 http://se.archive.ubuntu.com karmic/main Packages
        100 /var/lib/dpkg/status

ProblemType: Crash
Architecture: i386
Date: Fri Oct 9 17:06:06 2009
DistroRelease: Ubuntu 9.10
ExecutablePath: /usr/bin/gnome-power-manager
NonfreeKernelModules: wl
Package: gnome-power-manager 2.28.0-0ubuntu3
ProcCmdline: gnome-power-manager
ProcEnviron:
 PATH=(custom, user)
 LANG=en_US.UTF-8
 SHELL=/bin/bash
ProcVersionSignature: Ubuntu 2.6.31-12.41-generic
SegvAnalysis:
 Segfault happened at: 0x8052de3: mov 0x8(%eax),%eax
 PC (0x08052de3) ok
 source "0x8(%eax)" (0x00000008) not located in a known VMA region (needed readable region)!
 destination "%eax" ok
SegvReason: reading NULL VMA
Signal: 11
SourcePackage: gnome-power-manager
StacktraceTop:
 ?? ()
 ?? ()
 g_cclosure_marshal_VOID__VOID ()
 g_closure_invoke () from /usr/lib/libgobject-2.0.so.0
 ?? () from /usr/lib/libgobject-2.0.so.0
Title: gnome-power-manager crashed with SIGSEGV in g_cclosure_marshal_VOID__VOID()
Uname: Linux 2.6.31-12-generic i686
UserGroups: adm admin cdrom dialout fuse lpadmin plugdev sambashare

[Apport retracing service (apport)]

StacktraceTop:gpm_disks_set_spindown_timeout (disks=0x80ac260, timeout=60)
gpm_manager_client_changed_cb (client=0x80d0898,
g_cclosure_marshal_VOID__VOID ()
g_closure_invoke () from /usr/lib/libgobject-2.0.so.0
?? () from /usr/lib/libgobject-2.0.so.0

[Scott Howard (showard314)]

Hi, could you please try the version of GPM from this PPA:
https://launchpad.net/~showard314/+archive/ppa

and try to recreate this bug?

[Scott Howard (showard314)]

From upstream:

"[Richard Hughes]: This looks like a problem in dbus-glib:

/* set spindown timeouts */
ret = dbus_g_proxy_call (disks->priv->proxy, "DriveSetAllSpindownTimeouts",
&error,
             G_TYPE_INT, timeout,
             G_TYPE_STRV, options,
             G_TYPE_INVALID,
             G_TYPE_STRING, disks->priv->cookie,
             G_TYPE_INVALID);
if (!ret) {
        egg_warning ("failed to set spindown timeout: %s", error->message);
                                                                ^- GOES BANG
        g_error_free (error);
        goto out;
}

But in the docs:

dbus_g_proxy_call()...
Returns : FALSE if an error is set, TRUE otherwise.

So it's returning FALSE without setting ERROR non-NULL."

"[Scott Howard]: Thanks for looking at this, Richard. I found this, and think we're on to
something:
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=541632
"By passing dbus_g_proxy_call an incorrect signature, we can cause
the function to fail but not provide any error message. This isn't
helpful. This patch causes it to set an error when demarshalling
a type that it's not expecting. Instead of a NULL error, one instead
sees something like the following when the method returns a path
object but the client expected a string:"

I think we're passing "timeout" as a guint when it is expecting a gint
according to its signature."

Testing this is important for getting this fixed.

[Scott Howard (showard314)]

Upstream applied the patch, could someone with the bug confirm that it is fixed in this PPA so we can get this into Karmic before release? Thanks.
https://launchpad.net/~showard314/+archive/ppa

[Scott Howard (showard314)]

Fixed in 2.28.1

gnome-power-manager (2.28.1-0ubuntu1) karmic; urgency=low

  * New upstream bug fix release:
    - Add missing indirection to return value of DriveSetAllSpindownTimeouts
      call.
    - Fix failure error messages and comments.
    - Help the kernel through its sleep key confusion.
    - Improve configure summary wrt HAL.
    - Use separate variables for dbus and dbus-glib.
    - Remove devkit-power tool configure check.
    - Correctly set the focus on the last used device in
      gnome-power-statistics.
    - Do not hide some radio buttons depending on the current machine state.
    - Fix up all the issues reported by clang.
    - Don't try to match and refresh some devices, to stop g-p-s crashing on
      virtual devices.
    - DeviceKit-disks is expecting an int as a value, not a uint. *********************************<---- this one
    - Throttle screensaver before suspend/hibernate.
    - Translation updates.
  * Drop 03-fix-dkdisks-arguments.patch, applied upstream.

[Scott Howard (showard314)]

- DeviceKit-disks is expecting an int as a value, not a uint.