System Monitor ignores close request and uses 100% of one core

Bug #441828 reported by paniq on 2009-10-03
60
This bug affects 11 people
Affects Status Importance Assigned to Milestone
gnome-system-monitor (Ubuntu)
Medium
Unassigned

Bug Description

Binary package hint: gnome-system-monitor

I'm unable the close the system monitor by clicking on the X or using the menu. It uses one full core and never closes, just keeps running.

Unlike other bugs filed, this is persistent and 100% reproductible. There is no related output when gnome-system-monitor is started from the console, just this:

** (gnome-system-monitor:493): WARNING **: SELinux was found but is not enabled.

no idea if this is in any way related to the problem.

ProblemType: Bug
Architecture: i386
Date: Sun Oct 4 01:36:59 2009
DistroRelease: Ubuntu 9.10
ExecutablePath: /usr/bin/gnome-system-monitor
NonfreeKernelModules: nvidia
Package: gnome-system-monitor 2.28.0-0ubuntu1 [modified: usr/bin/gnome-system-monitor]
ProcEnviron:
 LANG=en_US.UTF-8
 PATH=(custom, user)
 SHELL=/bin/bash
ProcVersionSignature: Ubuntu 2.6.31-11.38-generic
SourcePackage: gnome-system-monitor
Uname: Linux 2.6.31-11-generic i686

Revision history for this message
paniq (paniq) wrote :
Revision history for this message
Chris Coulson (chrisccoulson) wrote :

Thank you for your bug report. I suspect this is the same issue I'm experiencing, which might be gvfs related. If you run "dbus-monitor --session", do you see a lot of activity? Do you have a CD in your drive by any chance (and does the issue go away if you remove it)?

Changed in gnome-system-monitor (Ubuntu):
importance: Undecided → Medium
status: New → Incomplete
Revision history for this message
paniq (paniq) wrote :

Chris, you are a magician! :)

That's it. There has indeed been a lot of activity with the session monitor, which seems to be drive related - there has been an audio CD in the drive.

When I removed the CD (and put it back in!), the issue went away. I couldn't reproduce the bug again. The CD has to be in the drive when the system boots, and the bug occurs only once.

Revision history for this message
Chris Coulson (chrisccoulson) wrote :

Confirming the issue then. I need to do a little debugging before I work out what component to assign this too and upstream it.

Changed in gnome-system-monitor (Ubuntu):
status: Incomplete → Confirmed
Revision history for this message
Scott Nicholls (koli) wrote :

I'd like to add that this also happens when I plug in my portable Samsung music player, which uses MTP. This also causes dbus-daemon to use an entire core when gnome-system-monitor is running too.

Revision history for this message
John Stewart (cane-cubo) wrote :

Same happens when I connect an iphone to a system running 9.10.

Revision history for this message
Don Cristóbal (doncristobal) wrote :

I can basically confirm this, but the behaviour on my system is slightly different:
* The hyperactivity of the cpu is shared between gnome-system-monitor (~50%) and dbus-daemon (~40%).
* The hyperactivity starts as many times as I start gnome-system-monitor when a CD is mounted, and it goes away when I eject the CD.
* However, the hyperactivity does not restart when I reinsert the disc while gnome-system-monitor is running.
* Sometimes gnome-system-manager refuses to close (and I xkill it), sometimes it goes away voluntarily. I don't see the logic.

"dbus-monitor --session" gives enormous amounts of output (some kind of data structure definitions or so). If you want, I can record some of it, just tell me to.

(Xubuntu 9.10, daily updates)

Revision history for this message
Chris Coulson (chrisccoulson) wrote :

Removing duplicate link there, as the bugs are totally unrelated. The other bug is about general excessive CPU usage, and this one is a specific issue due to a loop somewhere in GVFS

Revision history for this message
John Huffaker (jhuffaker-endeca) wrote :

I see this issue when connecting my iphone in 9.10. dbus-monitor --session produces voluminous output

Other info:
Running vmware player 3.0.1.

Does anyone know of a workaround that allows me to keep my iphone attached and the system stable? It is pretty annoying because my machine heats up pretty rapidly.

Revision history for this message
cuby (cuby) wrote :

this is also affecting Maverick RC amd64...

Revision history for this message
nerval (nerval-mail) wrote :

Confirming this Bug in Maverick amd64 with audiocd inserted. It only occurs if gnome-system-monitor displayed the Processes tab, otherwise it closes normally.

Revision history for this message
Jaime Rave (jaimerave) wrote :

I?m having this issue on Ubuntu natty.

Revision history for this message
David Smith (dave-kts) wrote :

I am having simalar problem in 10.10 32 bit superOS (super ubuntu)
intel DP43BF board
Nvidia quadro FX550 using proprietary driver version 260.19.06
Q8300 quad core2

cpu monitor widget and panel app show idle load at normal 5% or so, as soon as I open Sys monitor, cpu load jumps up to 32-38%, showing usage on all 4 cores. all 4 cores. tried increasing refresh rates and other options, no difference.
System monitor also hangs and will not close sometimes, have to use xkill, sometimes a couple times on it to get it to close.

I posted this elsewhere and then found this page, and can confirm it. Ejected my CD drive, Sys monitor closed itself, I re opened it and it reported normal usage. Upon re-inserting the audio CD, it refuses to open the CD and I get this:
 Could not open location 'cdda://sr1/'
Failed to execute child process "sound-juicer" (No such file or directory)
and Brasero will not open now.

Revision history for this message
Max Barry (max-maxbarry) wrote :

Same here when running gnome-system-monitor with iPhone 3 connected via USB.

Some observations:

1. As noted by other posters, I get tons of output from "dbus-monitor --session"

2. As mentioned above, when starting gnome-system-monitor, at first the only output on the command-line is: "** (gnome-system-monitor:6765): WARNING **: SELinux was found but is not enabled."

3. While experiencing the problem, the command-line that I started gnome-system-monitor from received these two messages thousands of times:

(gnome-system-monitor:6582): GVFS-RemoteVolumeMonitor-WARNING **: cannot add match rule 'type='signal',interface='org.gtk.Private.RemoteVolumeMonitor',sender='org.gtk.Private.GPhoto2VolumeMonitor',': org.freedesktop.DBus.Error.LimitsExceeded: Connection ":1.236" is not allowed to add more match rules (increase limits in configuration file if required)

(gnome-system-monitor:6582): GVFS-RemoteVolumeMonitor-WARNING **: cannot add match rule 'type='signal',interface='org.freedesktop.DBus',member='NameOwnerChanged',arg0='org.gtk.Private.GPhoto2VolumeMonitor'': org.freedesktop.DBus.Error.LimitsExceeded: Connection ":1.236" is not allowed to add more match rules (increase limits in configuration file if required)

4. Usually, after I plug in my iPhone, my system automatically mounts both "Max's iPhone" and "Documents on Max's iPhone". However, when I unplugged the iPhone, restart gnome-system-monitor, and plug the phone back in, only the main device is auto-mounted, not "Documents".

5. After removing and replugging the phone, I was not able to reproduce the problem during that login session. It only happened if the phone was already plugged in before I logged in.

I'm using Ubuntu 11.04.

To post a comment you must log in.
This report contains Public information  Edit
Everyone can see this information.

Duplicates of this bug

Other bug subscribers