GUI slows down significantly after random period of time.

Bug #996008 reported by Nick Tuckett
62
This bug affects 14 people
Affects Status Importance Assigned to Milestone
compiz (Ubuntu)
Confirmed
Undecided
Unassigned

Bug Description

Like a number of other users, I'm getting some variant of graphical slowdown after using my desktop PC for a random amount of time - between minutes and hours. I suspect the problem lies with compiz, OpenGL or graphics card drivers (NVidia in my case) as I have seen this behaviour with both Linux Mint 11 and 12 in the past, on the same PC. I don't believe there to be a hardware problem as this machine works perfectly fine running high-end games under Windows.

Restarting or logging out and in often temporarily resolves the issue... but it eventually returns.

ProblemType: Bug
DistroRelease: Ubuntu 12.04
Package: compiz 1:0.9.7.6-0ubuntu1
ProcVersionSignature: Ubuntu 3.2.0-23.36-generic 3.2.14
Uname: Linux 3.2.0-23-generic x86_64
NonfreeKernelModules: nvidia
.proc.driver.nvidia.gpus.0: Error: [Errno 21] Is a directory: '/proc/driver/nvidia/gpus/0'
.proc.driver.nvidia.registry: Binary: ""
.proc.driver.nvidia.version:
 NVRM version: NVIDIA UNIX x86_64 Kernel Module 295.40 Thu Apr 5 21:37:00 PDT 2012
 GCC version: gcc version 4.6.3 (Ubuntu/Linaro 4.6.3-1ubuntu5)
.tmp.unity.support.test.0:

ApportVersion: 2.0.1-0ubuntu7
Architecture: amd64
CompizPlugins: [core,composite,opengl,decor,grid,vpswitch,compiztoolbox,snap,place,regex,mousepoll,gnomecompat,imgpng,wall,move,resize,session,unitymtgrabhandles,animation,fade,workarounds,expo,scale,ezoom,unityshell]
CompositorRunning: compiz
Date: Mon May 7 16:38:55 2012
DistUpgraded: Fresh install
DistroCodename: precise
DistroVariant: ubuntu
DkmsStatus:
 nvidia-current, 295.40, 3.2.0-23-generic, x86_64: installed
 nvidia-current, 295.40, 3.2.0-24-generic, x86_64: installed
GraphicsCard:
 NVIDIA Corporation GF110 [GeForce GTX 580] [10de:1080] (rev a1) (prog-if 00 [VGA controller])
   Subsystem: Micro-Star International Co., Ltd. Device [1462:2550]
InstallationMedia: Ubuntu 12.04 LTS "Precise Pangolin" - Release amd64 (20120425)
JockeyStatus:
 xorg:nvidia_current - NVIDIA accelerated graphics driver (Proprietary, Enabled, In use)
 xorg:nvidia_current_updates - NVIDIA accelerated graphics driver (post-release updates) (Proprietary, Disabled, Not in use)
MachineType: System manufacturer System Product Name
PackageArchitecture: all
ProcEnviron:
 LANGUAGE=en_GB:en
 TERM=xterm
 PATH=(custom, no user)
 LANG=en_GB.UTF-8
 SHELL=/bin/bash
ProcKernelCmdLine: BOOT_IMAGE=/boot/vmlinuz-3.2.0-23-generic root=UUID=87d49759-39b7-4671-bae2-41815b3a4f34 ro quiet splash
SourcePackage: compiz
UpgradeStatus: No upgrade log present (probably fresh install)
dmi.bios.date: 03/02/2011
dmi.bios.vendor: American Megatrends Inc.
dmi.bios.version: 1502
dmi.board.asset.tag: To be filled by O.E.M.
dmi.board.name: P8P67 PRO
dmi.board.vendor: ASUSTeK Computer INC.
dmi.board.version: Rev 1.xx
dmi.chassis.asset.tag: Asset-1234567890
dmi.chassis.type: 3
dmi.chassis.vendor: Chassis Manufacture
dmi.chassis.version: Chassis Version
dmi.modalias: dmi:bvnAmericanMegatrendsInc.:bvr1502:bd03/02/2011:svnSystemmanufacturer:pnSystemProductName:pvrSystemVersion:rvnASUSTeKComputerINC.:rnP8P67PRO:rvrRev1.xx:cvnChassisManufacture:ct3:cvrChassisVersion:
dmi.product.name: System Product Name
dmi.product.version: System Version
dmi.sys.vendor: System manufacturer
version.compiz: compiz 1:0.9.7.6-0ubuntu1
version.ia32-libs: ia32-libs N/A
version.libdrm2: libdrm2 2.4.32-1ubuntu1
version.libgl1-mesa-dri: libgl1-mesa-dri 8.0.2-0ubuntu3
version.libgl1-mesa-dri-experimental: libgl1-mesa-dri-experimental N/A
version.libgl1-mesa-glx: libgl1-mesa-glx 8.0.2-0ubuntu3
version.nvidia-graphics-drivers: nvidia-graphics-drivers N/A
version.xserver-xorg-core: xserver-xorg-core 2:1.11.4-0ubuntu10.1
version.xserver-xorg-input-evdev: xserver-xorg-input-evdev 1:2.7.0-0ubuntu1
version.xserver-xorg-video-ati: xserver-xorg-video-ati 1:6.14.99~git20111219.aacbd629-0ubuntu2
version.xserver-xorg-video-intel: xserver-xorg-video-intel 2:2.17.0-1ubuntu4
version.xserver-xorg-video-nouveau: xserver-xorg-video-nouveau 1:0.0.16+git20111201+b5534a1-1build2

Revision history for this message
Nick Tuckett (nick.tuckett) wrote :
affects: ubuntu → compiz (Ubuntu)
Revision history for this message
Launchpad Janitor (janitor) wrote :

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

Changed in compiz (Ubuntu):
status: New → Confirmed
Revision history for this message
Christoph Buchner (bilderbuchi) wrote :

I think I'm affectd by the same bug. I foud ou that when the bug is triggered, I get a message in the syslog about
irq 11: nobody cared. tis only ever happns whnthe bug is triggered

You gota similar thing in curentDmesg, I sspect this could be related to the bug trigger:
irq 16: nobody cared (try booting with the "irqpoll" option)
...
Disabling IRQ #16

p.s. sorry for typing like a retard, the lag is etig my keystrokes...

Revision history for this message
Christoph Buchner (bilderbuchi) wrote :

Also, I just discovered that instlling thenew nvidia driver from precise-proposed (295.49) does NOT fix this issue... :-(
Please bump the severity, this consistently makes the system unusable after a time!

Revision history for this message
Christoph Buchner (bilderbuchi) wrote :

Another data point: After a workday of using Unity2D instead of Unity, I haven't experienced any slowdown.

Revision history for this message
Nick Tuckett (nick.tuckett) wrote :

I updated my machine with the latest 12.04 updates in the past week, which included some compiz updates (to 0.9.7.8), and the problem is still occurring.

As mentioned by Christoph, around the time of the latest occurrence I got an "irq 16: nobody cared" in syslog:

May 16 19:40:51 ntuckett-ubuntu kernel: [20742.749768] irq 16: nobody cared (try booting with the "irqpoll" option)
May 16 19:40:51 ntuckett-ubuntu kernel: [20742.749779] <IRQ> [<ffffffff810db37d>] __report_bad_irq+0x3d/0xe0
May 16 19:40:51 ntuckett-ubuntu kernel: [20742.749795] [<ffffffff810d8e69>] handle_irq_event_percpu+0xa9/0x220
May 16 19:40:51 ntuckett-ubuntu kernel: [20742.749799] [<ffffffff810d9031>] handle_irq_event+0x51/0x80
May 16 19:40:51 ntuckett-ubuntu kernel: [20742.749803] [<ffffffff810dc05a>] handle_fasteoi_irq+0x6a/0x110
May 16 19:40:51 ntuckett-ubuntu kernel: [20742.749808] [<ffffffff81015282>] handle_irq+0x22/0x40
May 16 19:40:51 ntuckett-ubuntu kernel: [20742.749883] [<ffffffffa0064370>] irq_handler

Unfortunately I was away from my machine at the time, so can't yet confirm if this was a coinciding event.

Revision history for this message
Christoph Buchner (bilderbuchi) wrote :

Today, I got that error even while using Unity2D, so comment #5 has to be disregarded. :-(

Can someone bump the importance? - Now that avoiding this by switching to unity2d is impossible, this makes working basically impossible after an undetermined time in the session.

Revision history for this message
Nick Tuckett (nick.tuckett) wrote :

Did some more research today and came across this post: http://www.linuxquestions.org/questions/slackware-14/disabling-irq-16-a-879964/

Checking my /proc/interrupts, I have a firewire module on IRQ16 alongside the Nvidia driver; I believe this is from my PCI soundcard. So I'm trying to run for a while with this module removed - manually atm via rmmod, as I couldn't get the blacklist to work.

Will post results after a few days use.

Revision history for this message
Christoph Buchner (bilderbuchi) wrote :

my /proc/interrupts shows ohci_hcd:usb2, snd_hda_intel, nvidia on the affected IRQ11. Does that mean I can choose to have either usb, sound, or a graphics card? I can hardly disable two of those three...

Does anybody know something to do (a command to run or whatever) to alleviate this error if it occurs? For me, only a reboot helps.

Revision history for this message
Nick Tuckett (nick.tuckett) wrote :

I've been running with my firewire module removed for a couple of weeks now, and have not experienced the slowdown.

Christoph, have you tried the first option mentioned in this post? http://www.linuxquestions.org/questions/slackware-14/disabling-irq-16-a-879964/#post4354587

There's some more on enabling MSI for Nvidia cards here: http://www.nvnews.net/vbulletin/showthread.php?t=149171.
Seems like it isn't always a piece of cake, though...

Revision history for this message
Christoph Buchner (bilderbuchi) wrote :

For the record, I have tried booting with the irqpoll option, does not help.

Doing echo 'options nvidia NVreg_EnableMSI=1' > /etc/modprobe.d/nvidia.conf as recommended in that post you linked does apparently not activate MSI for nvidia as intended. There was no file with that name in that directory, but several otehr nvidia* files (nvidia-current-updates_hybrid.conf, nvidia-current_hybrid.conf, nvidia-graphics-drivers.conf). I did not know which one to add that option to, if necessary.

However, using dmesg | grep MSI to check for enabled MSI, the following drew my attention:
hda_intel: Disabling MSI
hda_intel is one of the modules which conflicts on irq11, I think (/proc/interrupts says on 11: ohci_hcd:usb2, snd_hda_intel, nvidia), so I'll further investigate. I have a theory that on precise, the MSI for the sound driver has been disabled for some reason, creating this conflict in the first place.

I have to be a bit careful about wildly changing system settings, as this is a production machine and I can't currently afford to lose it for an extended period of time if I have to nuke the system and reinstall or whatever.

Revision history for this message
Rahul Jain (equites-vero) wrote :

I'm having a similar issue in Saucy Salamander. After using my computer for a while, GUI slows down somewhat, not unusably slow, but animations are signficantly slower. Even the Window side by side snap animation shows a little lag, I begin to see the fade effect in icon right click menu in launcher bar where I did not notice it before, dash becomes slightly stiffer, etc. Dmesg | tail shows no irq messages for me, but hitting unity --replace solves the problem, until it occurs again. I like Unity but the desktop seems to be a mess with lots of users experiencing bugs either with the dash or with performance, etc.. I hope this issue is fixed soon.

Revision history for this message
William Söderberg (william-soderberg) wrote :
Download full text (4.3 KiB)

I also experience this problem. I've been having it for a while. Logging out and logging in again temporarily fixes the issue. Affects both KDE and Gnome. Got this in /var/log/syslog when the problem happened just now:

Mar 31 21:57:54 dreadnought kernel: [ 2390.058249] irq 16: nobody cared (try booting with the "irqpoll" option)
Mar 31 21:57:54 dreadnought kernel: [ 2390.058253] CPU: 0 PID: 0 Comm: swapper/0 Tainted: PF O 3.11.0-18-generic #32-Ubuntu
Mar 31 21:57:54 dreadnought kernel: [ 2390.058254] Hardware name: System manufacturer System Product Name/P8P67-M, BIOS 1501 08/30/2011
Mar 31 21:57:54 dreadnought kernel: [ 2390.058255] ffff880400974c9c ffff880417403e68 ffffffff816e8571 ffff880400974c00
Mar 31 21:57:54 dreadnought kernel: [ 2390.058257] ffff880417403e90 ffffffff810f88e2 ffff880400974c00 0000000000000010
Mar 31 21:57:54 dreadnought kernel: [ 2390.058259] 0000000000000000 ffff880417403ed0 ffffffff810f8d11 ffffffff815a158f
Mar 31 21:57:54 dreadnought kernel: [ 2390.058261] Call Trace:
Mar 31 21:57:54 dreadnought kernel: [ 2390.058262] <IRQ> [<ffffffff816e8571>] dump_stack+0x45/0x56
Mar 31 21:57:54 dreadnought kernel: [ 2390.058270] [<ffffffff810f88e2>] __report_bad_irq+0x32/0xd0
Mar 31 21:57:54 dreadnought kernel: [ 2390.058271] [<ffffffff810f8d11>] note_interrupt+0x141/0x200
Mar 31 21:57:54 dreadnought kernel: [ 2390.058274] [<ffffffff815a158f>] ? cpuidle_enter_state+0x4f/0xc0
Mar 31 21:57:54 dreadnought kernel: [ 2390.058276] [<ffffffff810f6681>] handle_irq_event_percpu+0xe1/0x1e0
Mar 31 21:57:54 dreadnought kernel: [ 2390.058277] [<ffffffff810f67bf>] handle_irq_event+0x3f/0x60
Mar 31 21:57:54 dreadnought kernel: [ 2390.058279] [<ffffffff810f988a>] handle_fasteoi_irq+0x5a/0x100
Mar 31 21:57:54 dreadnought kernel: [ 2390.058283] [<ffffffff81014b8e>] handle_irq+0x1e/0x30
Mar 31 21:57:54 dreadnought kernel: [ 2390.058285] [<ffffffff816fa70d>] do_IRQ+0x4d/0xc0
Mar 31 21:57:54 dreadnought kernel: [ 2390.058287] [<ffffffff816efd6d>] common_interrupt+0x6d/0x6d
Mar 31 21:57:54 dreadnought kernel: [ 2390.058288] <EOI> [<ffffffff815a158f>] ? cpuidle_enter_state+0x4f/0xc0
Mar 31 21:57:54 dreadnought kernel: [ 2390.058291] [<ffffffff815a1588>] ? cpuidle_enter_state+0x48/0xc0
Mar 31 21:57:54 dreadnought kernel: [ 2390.058293] [<ffffffff815a16c9>] cpuidle_idle_call+0xc9/0x210
Mar 31 21:57:54 dreadnought kernel: [ 2390.058295] [<ffffffff8101bafe>] arch_cpu_idle+0xe/0x30
Mar 31 21:57:54 dreadnought kernel: [ 2390.058298] [<ffffffff810b5825>] cpu_startup_entry+0xe5/0x280
Mar 31 21:57:54 dreadnought kernel: [ 2390.058300] [<ffffffff816d7767>] rest_init+0x77/0x80
Mar 31 21:57:54 dreadnought kernel: [ 2390.058303] [<ffffffff81d26ef7>] start_kernel+0x40a/0x416
Mar 31 21:57:54 dreadnought kernel: [ 2390.058305] [<ffffffff81d268f6>] ? repair_env_string+0x5c/0x5c
Mar 31 21:57:54 dreadnought kernel: [ 2390.058307] [<ffffffff81d26120>] ? early_idt_handlers+0x120/0x120
Mar 31 21:57:54 dreadnought kernel: [ 2390.058309] [<ffffffff81d265de>] x86_64_start_reservations+0x2a/0x2c
Mar 31 21:57:54 dreadnought kernel: [ 2390.058310] [<ffffffff81d266e8>] x86_64_start_kernel+0x108/0x117
Mar 31 21:57:54 dreadnought kernel: [ 2390.0...

Read more...

Revision history for this message
William Söderberg (william-soderberg) wrote :

I can confirm that the problem stopped when I removed my NIC. Now IRQ 16 is only used by my video card.

Revision history for this message
fcole90 (fcole90) wrote :

As part of the big bug review for 16.04 LTS I have tested this on 15.10 and the bug is still there.

tags: added: desktop-bugscrub-triaged
Revision history for this message
fcole90 (fcole90) wrote :

Importance Medium

Revision history for this message
TheGateKeeperX (thegatekeeperx) wrote :

This worked for me. I had a unused Fiber NIC still installed and removed it.
FYI: NIC Had LSI Chip (L2A1157) and I have a Nvidia Quadro FX 4800/PCIe/SSE2 graphics card installed.

Seems to be some type of Hardware & IRQ issue.
Problem has been gone for last 2 days.
Previously this issue would appear within half hour or so after logging in.

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

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.