Bug #1395547 “xfsettingsd and xfce4-volumed using a lot of CPU a...” : Bugs : Xfce4 Settings

Revision history for this message

icyrock.com (icyrock-com) wrote on 2014-11-23:

#1

Dependencies.txt Edit (8.3 KiB, text/plain; charset="utf-8")
ProcEnviron.txt Edit (103 bytes, text/plain; charset="utf-8")

Revision history for this message

Thaddaeus Tintenfisch (thad-fisch-deactivatedaccount) wrote on 2014-11-25:

#2

Does xfce4-volumed stop working properly (bug 1384831)?

Revision history for this message

icyrock.com (icyrock-com) wrote on 2014-11-26:

#3

- xfce4-volumed seems not to be working properly after that
- Restarting xfce4-volumed fixes the multimedia keys not working. One CPU core is still used by xfsettingsd
- Restarting both xfsettingsd and xfce4-volumed or just xfce4-volumed fixes the CPU issue
- If I kill xfsettingsd, leave xfce4-volumed running, run xflock4, unlock the screen, xfce4-volumed seems to be working fine and CPU issue does not manifest
- If I kill xfce4-volumed, leave xfsettingsd running, run xflock4, unlock the screen, I run into the CPU issue. Restarting xfsettingsd resolves the issue

This leads me to believe it's not a xfce4-volumed issue. It looks like these two are influencing each other, though, or something along those lines. Maybe xfsettingsd and xflock4 influencing each other?

One more thing - when unlocking with xfsettingsd running, the whole system seems to freeze for about 10 seconds or so, even the CPU monitors or time displayed in the corner are not moving during that period.

Revision history for this message

Launchpad Janitor (janitor) wrote on 2014-11-29:

#4

Status changed to 'Confirmed' because the bug affects multiple users.

Changed in xfce4-settings (Ubuntu):
status:	New → Confirmed
Changed in xfce4-volumed (Ubuntu):
status:	New → Confirmed

Revision history for this message

icyrock.com (icyrock-com) wrote on 2014-12-02:

#6

May be related to https://bugs.launchpad.net/ubuntu/+source/linux/+bug/998310, using Xubuntu 14.10. The behavior doesn't look completely the same, only similar.

Revision history for this message

icyrock.com (icyrock-com) wrote on 2014-12-02:

#7

After reading the bug I mentioned in #6, I tried waiting after logging in. The CPU usage went down in about 3 minutes without restarting xfsettingsd. This leads me to believe these bugs are more related than I thought. Contrary to #10 in that bug (https://bugs.launchpad.net/ubuntu/+source/linux/+bug/998310/comments/10), I could not fix it by saving a session or by running xmodmap ~/.Xmodmap.

Revision history for this message

Steve Dodier-Lazaro (sidi) wrote on 2014-12-03:

#8

If that's any useful, some of the bug reports I've had with xfce4-volumed were related to how (now defunct) libkeybinder handles keyboard layout changes. Are such changes (even from/to the same layout) triggerred by switching to whatever launcher Xubuntu uses?

Revision history for this message

icyrock.com (icyrock-com) wrote on 2014-12-12:

#9

I sometimes see these in a terminal after unlocking the screen:

xmodmap: please release the following keys within 2 seconds:
r (keysym 0x72, keycode 27)
Control_R (keysym 0xffe4, keycode 105)

Not sure if that is related at all to this bug.

Revision history for this message

Thaddaeus Tintenfisch (thad-fisch-deactivatedaccount) wrote on 2014-12-13:

#10

Please forward this bug to the Xfce bug tracker. Thanks in advance.

https://bugzilla.xfce.org/

Revision history for this message

Steve Dodier-Lazaro (sidi) wrote on 2014-12-13:

#11

The bug would be rejected against upstream xfce4-volumed with the information available so far. I cannot fix bugs that I cannot reproduce. It could help to report against xfce4-settings. Note the following though:

So far the following conditions are known to trigger the bug:
OS: Ubuntu 14.10 64bits
GPU: AMD ???
Keyboard: ???
Layout: ???
Concurrent software that binds keys: ???
X11 version: ???

I don't have Ubuntu and AMD and so I cannot reproduce the bug. I strongly suspect that the issue is not just with xfce4 apps though but with key binding or key pressing, e.g., a X11/driver bug that is triggered (by some unknown conditions) and causes apps that bind keys to freeze.

OP, could you please attach a 'lshw' output and bring this bug report to the attention of X11 Ubuntu maintainers? They might have seen something similar for GNOME/KDE/Unity users.

Thanks.

Revision history for this message

icyrock.com (icyrock-com) wrote on 2014-12-14:

#12

lshw.txt Edit (17.8 KiB, text/plain)

Added lshw output.

Answers to other questions:

>> OS: Ubuntu 14.10 64bits
<
Linux aclap 3.13.0-24-generic #47-Ubuntu SMP Fri May 2 23:30:00 UTC 2014 x86_64 x86_64 x86_64 GNU/Linux

>> GPU: AMD ???
<
02:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Madison [Mobility Radeon HD 5650/5750 / 6530M/6550M]

>> Keyboard: ???
<
Bus 001 Device 004: ID 045e:00db Microsoft Corp. Natural Ergonomic Keyboard 4000 V1.0

>> Layout: ???
<
English (US)

>> Concurrent software that binds keys: ???
<
I don't think I have any.

>> X11 version: ???
<
ii xserver-xorg 1:7.7+7ubuntu2 amd64 X.Org X server

X.Org X Server 1.16.0
Release Date: 2014-07-16

Alberto Salvia Novella (es20490446e) on 2014-12-15

Changed in xfce4-volumed (Ubuntu):
importance:	Undecided → High
Changed in xfce4-settings (Ubuntu):
importance:	Undecided → High

Revision history for this message

icyrock.com (icyrock-com) wrote on 2015-02-02:

#13

I ran a strace on the xfsettingsd pid, logged out, logged in and I could see a lot of these repeating:

poll([{fd=3, events=POLLIN|POLLOUT}], 1, 4294967295) = 1 ([{fd=3, revents=POLLOUT}])
writev(3, [{"\20\0\3\0\3\0\0\0AltL", 12}, {NULL, 0}, {"", 0}], 3) = 12
poll([{fd=3, events=POLLIN}], 1, 4294967295) = 1 ([{fd=3, revents=POLLIN}])
recvmsg(3, {msg_name(0)=NULL, msg_iov(1)=[{"\1\0\213\346\0\0\0\0\232\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0", 4096}], msg_controllen=0, msg_flags=0}, 0) = 32
recvmsg(3, 0x7fffc492a350, 0) = -1 EAGAIN (Resource temporarily unavailable)
recvmsg(3, 0x7fffc492a350, 0) = -1 EAGAIN (Resource temporarily unavailable)
poll([{fd=3, events=POLLIN|POLLOUT}], 1, 4294967295) = 1 ([{fd=3, revents=POLLOUT}])
writev(3, [{"\20\0\5\0\n\0\0\0LevelThree\0\0", 20}, {NULL, 0}, {"", 0}], 3) = 20
poll([{fd=3, events=POLLIN}], 1, 4294967295) = 1 ([{fd=3, revents=POLLIN}])
recvmsg(3, {msg_name(0)=NULL, msg_iov(1)=[{"\1\0\214\346\0\0\0\0\233\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0", 4096}], msg_controllen=0, msg_flags=0}, 0) = 32
recvmsg(3, 0x7fffc492a350, 0) = -1 EAGAIN (Resource temporarily unavailable)
recvmsg(3, 0x7fffc492a350, 0) = -1 EAGAIN (Resource temporarily unavailable)
poll([{fd=3, events=POLLIN|POLLOUT}], 1, 4294967295) = 1 ([{fd=3, revents=POLLOUT}])
writev(3, [{"\20\0\3\0\4\0\0\0LAlt", 12}, {NULL, 0}, {"", 0}], 3) = 12
poll([{fd=3, events=POLLIN}], 1, 4294967295) = 1 ([{fd=3, revents=POLLIN}])
recvmsg(3, {msg_name(0)=NULL, msg_iov(1)=[{"\1\0\215\346\0\0\0\0\234\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0", 4096}], msg_controllen=0, msg_flags=0}, 0) = 32
recvmsg(3, 0x7fffc492a350, 0) = -1 EAGAIN (Resource temporarily unavailable)
recvmsg(3, 0x7fffc492a350, 0) = -1 EAGAIN (Resource temporarily unavailable)

The values are different between the calls as you can see (though usually similar), but they seem to be repeating endlessly, till the system settles down several minutes down the road, when it all stops.

Not sure if relevant, let me know if I can provide any information that could help track the issue down.

I ran a strace on the xfsettingsd pid, logged out, logged in and I could see a lot of these repeating:

poll([{fd=3, events=POLLIN|POLLOUT}], 1, 4294967295) = 1 ([{fd=3, revents=POLLOUT}])
writev(3, [{"\20\0\3\0\3\0\0\0AltL", 12}, {NULL, 0}, {"", 0}], 3) = 12
poll([{fd=3, events=POLLIN}], 1, 4294967295) = 1 ([{fd=3, revents=POLLIN}])
recvmsg(3, {msg_name(0)=NULL, msg_iov(1)=[{"\1\0\213\346\0\0\0\0\232\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0", 4096}], msg_controllen=0, msg_flags=0}, 0) = 32
recvmsg(3, 0x7fffc492a350, 0)           = -1 EAGAIN (Resource temporarily unavailable)
recvmsg(3, 0x7fffc492a350, 0)           = -1 EAGAIN (Resource temporarily unavailable)
poll([{fd=3, events=POLLIN|POLLOUT}], 1, 4294967295) = 1 ([{fd=3, revents=POLLOUT}])
writev(3, [{"\20\0\5\0\n\0\0\0LevelThree\0\0", 20}, {NULL, 0}, {"", 0}], 3) = 20
poll([{fd=3, events=POLLIN}], 1, 4294967295) = 1 ([{fd=3, revents=POLLIN}])
recvmsg(3, {msg_name(0)=NULL, msg_iov(1)=[{"\1\0\214\346\0\0\0\0\233\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0", 4096}], msg_controllen=0, msg_flags=0}, 0) = 32
recvmsg(3, 0x7fffc492a350, 0)           = -1 EAGAIN (Resource temporarily unavailable)
recvmsg(3, 0x7fffc492a350, 0)           = -1 EAGAIN (Resource temporarily unavailable)
poll([{fd=3, events=POLLIN|POLLOUT}], 1, 4294967295) = 1 ([{fd=3, revents=POLLOUT}])
writev(3, [{"\20\0\3\0\4\0\0\0LAlt", 12}, {NULL, 0}, {"", 0}], 3) = 12
poll([{fd=3, events=POLLIN}], 1, 4294967295) = 1 ([{fd=3, revents=POLLIN}])
recvmsg(3, {msg_name(0)=NULL, msg_iov(1)=[{"\1\0\215\346\0\0\0\0\234\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0", 4096}], msg_controllen=0, msg_flags=0}, 0) = 32
recvmsg(3, 0x7fffc492a350, 0)           = -1 EAGAIN (Resource temporarily unavailable)
recvmsg(3, 0x7fffc492a350, 0)           = -1 EAGAIN (Resource temporarily unavailable)

The values are different between the calls as you can see (though usually similar), but they seem to be repeating endlessly, till the system settles down several minutes down the road, when it all stops.

Not sure if relevant, let me know if I can provide any information that could help track the issue down.

Revision history for this message

In Xfce Bugzilla #11497, icyrock.com (icyrock-com) wrote on 2015-02-02:

#15

*Note*: the below is copy & paste from https://bugs.launchpad.net/ubuntu/+source/xfce4-settings/+bug/1395547 as per request by Thaddaeus Tintenfisch (thad-fisch) on 2014-12-13. Let me know if I should copy other things from there or if having a reference to that is enough / acceptable.

Just upgraded to Xubuntu 14.10, had issues with fglrx (see http://askubuntu.com/questions/552795/xubuntu-14-10-upgrade-issues-with-fglrx), went back to radeon drivers. Now it looks to be OK, however after locking the screen using xflock4 and unlocking, I get these three processes using a lot of CPU (this is on 4-core Phenom II X4 N950):

PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
1572 root 20 0 262892 56672 26480 R 89.7 1.4 6:48.70 Xorg
5025 usr 20 0 312116 11336 4632 R 36.9 0.3 0:17.42 xfce4-volumed
5036 usr 20 0 327328 10556 5760 S 31.5 0.3 0:13.38 xfsettingsd

Additional information:

- Killing and restarting xfsettingsd and xfce4-volumed resolves the issue until next sleep cycle
- When running only xfce4-volumed, but not xfsettingsd, the issue does not appear
- This can be reproduced always. ATI Radeon Mobility HD 5650 video card
- The CPU usage does not seem to go down after a while
- If I got it correctly, it looks like some other processes are also going to high CPU usage after killing xfce4-volumed. I think I saw xfce4-xkb-plugin going to 95% CPU usage
- The issue happens before unlock, i.e. locking then pressing a key to get a login dialog, then switching to vty1 and doing a top yields the above

Not sure what else I should check / post that can be useful for troubleshooting, let me know.

ProblemType: Bug
DistroRelease: Ubuntu 14.10
Package: xfce4-settings 4.11.3-0ubuntu2
ProcVersionSignature: Ubuntu 3.13.0-24.47-generic 3.13.9
Uname: Linux 3.13.0-24-generic x86_64
ApportVersion: 2.14.7-0ubuntu8
Architecture: amd64
CurrentDesktop: XFCE
Date: Sun Nov 23 18:34:02 2014
InstallationDate: Installed on 2014-04-24 (213 days ago)
InstallationMedia: Xubuntu 14.04 LTS "Trusty Tahr" - Release amd64 (20140416.2)
SourcePackage: xfce4-settings
UpgradeStatus: Upgraded to utopic on 2014-11-22 (0 days ago)

*Note*: the below is copy & paste from https://bugs.launchpad.net/ubuntu/+source/xfce4-settings/+bug/1395547 as per request by Thaddaeus Tintenfisch (thad-fisch) on 2014-12-13. Let me know if I should copy other things from there or if having a reference to that is enough / acceptable.

Just upgraded to Xubuntu 14.10, had issues with fglrx (see http://askubuntu.com/questions/552795/xubuntu-14-10-upgrade-issues-with-fglrx), went back to radeon drivers. Now it looks to be OK, however after locking the screen using xflock4 and unlocking, I get these three processes using a lot of CPU (this is on 4-core Phenom II X4 N950):

PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
 1572 root 20 0 262892 56672 26480 R 89.7 1.4 6:48.70 Xorg
 5025 usr 20 0 312116 11336 4632 R 36.9 0.3 0:17.42 xfce4-volumed
 5036 usr 20 0 327328 10556 5760 S 31.5 0.3 0:13.38 xfsettingsd

Additional information:

- Killing and restarting xfsettingsd and xfce4-volumed resolves the issue until next sleep cycle
- When running only xfce4-volumed, but not xfsettingsd, the issue does not appear
- This can be reproduced always. ATI Radeon Mobility HD 5650 video card
- The CPU usage does not seem to go down after a while
- If I got it correctly, it looks like some other processes are also going to high CPU usage after killing xfce4-volumed. I think I saw xfce4-xkb-plugin going to 95% CPU usage
- The issue happens before unlock, i.e. locking then pressing a key to get a login dialog, then switching to vty1 and doing a top yields the above

Not sure what else I should check / post that can be useful for troubleshooting, let me know.

ProblemType: Bug
DistroRelease: Ubuntu 14.10
Package: xfce4-settings 4.11.3-0ubuntu2
ProcVersionSignature: Ubuntu 3.13.0-24.47-generic 3.13.9
Uname: Linux 3.13.0-24-generic x86_64
ApportVersion: 2.14.7-0ubuntu8
Architecture: amd64
CurrentDesktop: XFCE
Date: Sun Nov 23 18:34:02 2014
InstallationDate: Installed on 2014-04-24 (213 days ago)
InstallationMedia: Xubuntu 14.04 LTS "Trusty Tahr" - Release amd64 (20140416.2)
SourcePackage: xfce4-settings
UpgradeStatus: Upgraded to utopic on 2014-11-22 (0 days ago)

Revision history for this message

icyrock.com (icyrock-com) wrote on 2015-02-02:

#14

As per request by Thaddaeus Tintenfisch (thad-fisch) on 2014-12-13:, I've created a bug report on bugzilla.xfce.org: https://bugzilla.xfce.org/show_bug.cgi?id=11497

Bug Watch Updater (bug-watch-updater) on 2015-02-02

Changed in xfce4-settings:
importance:	Unknown → Medium
status:	Unknown → Confirmed

Revision history for this message

In Xfce Bugzilla #11497, Timo-heister (timo-heister) wrote on 2015-02-27:

#16

I have the same issue but with intel graphics. One difference is that that the high load on Xorg goes away after about a minute and the system behaves normally.

I can also produce this high load using
xmodmap -pke > .Xmodmap
xmodmap ~/.Xmodmap

Revision history for this message

In Xfce Bugzilla #11497, Timo-heister (timo-heister) wrote on 2015-02-27:

#17

*** Bug 11611 has been marked as a duplicate of this bug. ***

Revision history for this message

In Xfce Bugzilla #11497, icyrock.com (icyrock-com) wrote on 2015-02-28:

#18

Thanks Timo. Check the comment #7 I posted on the launchpad bug - (https://bugs.launchpad.net/ubuntu/+source/xfce4-settings/+bug/1395547/comments/7). It actually goes down for me too, though it takes a bit longer, 3 minutes on my particular machine.

I tried the thing you used to reproduce:

xmodmap -pke > .Xmodmap
xmodmap ~/.Xmodmap

and I can confirm it generates high load on my machine, too. In my case, it took 4 minutes for this one also to calm down, which is somewhat consistent with the 3 minutes I measured before when unlocking the screen.

Revision history for this message

In Xfce Bugzilla #11497, Timo-heister (timo-heister) wrote on 2015-02-28:

#19

Thanks for your comment!

Do you have a workaround on your system? Next thing I was going to try is to experiment with different tools to provide screen locking (light-locker, xscreensaver, gnome-screensaver, ...).

Revision history for this message

In Xfce Bugzilla #11497, icyrock.com (icyrock-com) wrote on 2015-02-28:

#20

Unfortunately, nothing better than just:

killall xfsettingsd; killall xfce4-volumed; sleep 6; xfsettingsd & xfce4-volumed &

I don't think it is a problem with the locker itself though, to me it looks like it's just a trigger for an issue in xfce4-volumed or something related to its interaction with the X system.

Check my comment #3 on the referenced launchpad bug (https://bugs.launchpad.net/ubuntu/+source/xfce4-settings/+bug/1395547/comments/3), where I outline that when I kill xfsettingsd, lock, unlock, I don't see the CPU issue.

Given comment #8 on that bug (https://bugs.launchpad.net/ubuntu/+source/xfce4-settings/+bug/1395547/comments/8), it might be that it's related to libkeybinder + layout changes, which might be triggered by the xmodmap commands you can reproduce the bug with.

I put as much information on the other bug report as I knew how to gather / thought would be useful. If somebody can let me know what could help fixing, I can try providing those or doing additional testing.

Revision history for this message

In Xfce Bugzilla #11497, Timo-heister (timo-heister) wrote on 2015-02-28:

#21

Okay, I found a great workaround for me:
1. I made sure I removed light-locker, gnome-screensaver, xscreensaver, kscreensaver, etc..
2. installed xscreensaver
3. reboot

Standby/resume and screen unlocking now works perfectly. While xmodmap is still broken, I don't care about that.

Revision history for this message

In Xfce Bugzilla #11497, icyrock.com (icyrock-com) wrote on 2015-02-28:

#22

Thanks Timo - I only had light-locker installed, so I did this:

1. Uninstall light-locker
2. Install xcreensaver
3. Reboot

Agreed, it seems to be working fine now, so might be something related either to light-locker or how it interacts with xfsettingsd.

Thanks for posting a workaround!

Revision history for this message

Launchpad Janitor (janitor) wrote on 2015-03-03:

#23

Status changed to 'Confirmed' because the bug affects multiple users.

Changed in xserver-xorg-video-ati (Ubuntu):
status:	New → Confirmed

Alberto Salvia Novella (es20490446e) on 2015-03-03

Changed in xserver-xorg-video-ati (Ubuntu):
importance:	Undecided → High

penalvch (penalvch) on 2015-09-26

no longer affects:

xserver-xorg-video-ati (Ubuntu)

Xfce4 Settings

xfsettingsd and xfce4-volumed using a lot of CPU after screen lock / unlock

Bug Description

Duplicates of this bug

Other bug subscribers

Bug attachments

Remote bug watches

	Status	Importance	Assigned to
Xfce4 Settings	Confirmed	Medium	xfce-bugs #11497
xfce4-settings (Ubuntu)	Confirmed	High	Unassigned
xfce4-volumed (Ubuntu)	Confirmed	High	Unassigned