Bug 476743 - KDE 4 freezes completely after repeated <alt>+<tab> window switches
KDE 4 freezes completely after repeated <alt>+<tab> window switches
Status: CLOSED WORKSFORME
Product: Fedora
Classification: Fedora
Component: kernel (Show other bugs)
10
i686 Linux
low Severity urgent
: ---
: ---
Assigned To: Adam Jackson
Fedora Extras Quality Assurance
: Patch
: 508635 (view as bug list)
Depends On:
Blocks:
  Show dependency treegraph
 
Reported: 2008-12-16 16:16 EST by Matthias Scheutz
Modified: 2009-11-18 16:07 EST (History)
16 users (show)

See Also:
Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of:
Environment:
Last Closed: 2009-11-18 16:07:52 EST
Type: ---
Regression: ---
Mount Type: ---
Documentation: ---
CRM:
Verified Versions:
Category: ---
oVirt Team: ---
RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: ---


Attachments (Terms of Use)
Xorg.0.log after kde freeze (55.39 KB, text/plain)
2009-02-12 08:11 EST, iarly selbir
no flags Details
xsession-errors after kde freeze with me (272.49 KB, text/plain)
2009-02-12 08:14 EST, iarly selbir
no flags Details

  None (edit)
Description Matthias Scheutz 2008-12-16 16:16:53 EST
Description of problem:

KDE 4 freezes completely after repeated and quick presses of <alt>+<tab> to switch between windows... mouse still works, but nothing else; without network access, a hard-reset is needed


Version-Release number of selected component (if applicable):


How reproducible:
Cannot be reliably produced, but will happen occasionally


Steps to Reproduce:
1.
2.
3.
  
Actual results:


Expected results:


Additional info:
Comment 1 Rex Dieter 2008-12-16 16:22:17 EST
"KDE freezes completely" = even the mouse?
What video hw/driver are you using?
Have you enabled desktop effects/compositing?
Comment 2 Matthias Scheutz 2008-12-16 16:48:06 EST
no, the mouse is still working, everything else is frozen, including the keyboard

the computer is a dell D830 latitude, with a mobile GM965/GL960 integrated graphics controller

desktop effects are turned on (improved window management, shadows, various animations, including box switch)
Comment 3 Rex Dieter 2008-12-16 16:53:12 EST
Mind trying to reproduce with effects turned off (or perhaps change DE effects prefs from opengl -> xrender)?
Comment 4 Steven M. Parrish 2009-01-10 10:20:01 EST
Ping?
Comment 5 Matthias Scheutz 2009-01-10 10:35:39 EST
Thanks for the reminder -- yes, I think the problem is caused by desktop effects because after I turned them off, I have not had any of the problems since.
Comment 6 Rex Dieter 2009-01-10 11:19:26 EST
alright, almost certainly an X/driver bug.

Would you rather we close this, or re-assign to X?
Comment 7 Matthias Scheutz 2009-01-10 11:24:43 EST
Re-assigning is fine.  I would make sure that the KDE folks are aware of it (I've never had anything like that happen before in Linux)
Comment 8 Rex Dieter 2009-01-10 11:44:46 EST
ok, there's already a multitude of intel driver-related bugs, not sure which one to dup.  I'll go spelunking.
Comment 9 Rex Dieter 2009-01-10 11:45:52 EST
oh, did you have a chance to try the suggestion in comment #3?
Comment 10 Matthias Scheutz 2009-01-10 11:49:12 EST
You mean "change DE effects prefs from opengl -> xrender", no, I haven't -- right now I simply need a working system... (I'll try to remember when I have more time)
Comment 11 Steven M. Parrish 2009-02-04 12:34:34 EST
Rex, how did the spelunking go.  Did you determine where you wanted to reassign this one?
Comment 12 Rex Dieter 2009-02-04 12:41:25 EST
didn't find anything that was an obvious fit.

Matthias, I'm sure the X folks will want a copy of /var/log/Xorg.0.log + /etc/X11/xorg.conf (if it exists), at least.
Comment 13 Matěj Cepl 2009-02-05 04:53:06 EST
Thanks for the bug report.  We have reviewed the information you have provided above, and there is some additional information we require that will be helpful in our diagnosis of this issue.

Please attach your X server config file (/etc/X11/xorg.conf, if available), /var/log/messages, /var/log/dmesg, and X server log file (/var/log/Xorg.*.log) to the bug report as individual uncompressed file attachments using the bugzilla file attachment link below.

We will review this issue again once you've had a chance to attach this information.

Thanks in advance.
Comment 14 iarly selbir 2009-02-10 12:40:00 EST
(In reply to comment #13)
> Thanks for the bug report.  We have reviewed the information you have provided
> above, and there is some additional information we require that will be helpful
> in our diagnosis of this issue.
> 
> Please attach your X server config file (/etc/X11/xorg.conf, if available),
> /var/log/messages, /var/log/dmesg, and X server log file (/var/log/Xorg.*.log)
> to the bug report as individual uncompressed file attachments using the
> bugzilla file attachment link below.
> 
> We will review this issue again once you've had a chance to attach this
> information.
> 
> Thanks in advance.

I'm having the same problem...

Only kde freezes, if I try ssh access I access normally.

I'm right now erasing all logs, enabling the desktop effects and tracking until freeze again.
Comment 15 Matěj Cepl 2009-02-11 13:46:20 EST
Matthias, could we get required information as well, please?
Comment 16 Matthias Scheutz 2009-02-11 21:44:37 EST
I'm afraid that I don't have time right now to collect the data you need (I'll try to send it when I have more time)
Comment 17 iarly selbir 2009-02-12 06:27:21 EST
I'm waiting for a new freeze... when this happen, I sent all logs


Note:

My Graphic card: 00:02.0 VGA compatible controller: Intel Corporation Mobile GM965/GL960 Integrated Graphics Controller (rev 0c)

Kde: 4.2 ( from updates-testing )
Comment 18 iarly selbir 2009-02-12 08:11:41 EST
Created attachment 331685 [details]
Xorg.0.log after kde freeze

the Xorg.0.log ( /var/log/Xorg.0.log ) after kde freeze.
Comment 19 iarly selbir 2009-02-12 08:14:08 EST
Created attachment 331686 [details]
xsession-errors after kde freeze with me

Following the xsession-errors from my home (/home/user/.xsession-errors ) after kde freeze
Comment 20 iarly selbir 2009-02-12 08:15:11 EST
The /var/log/messages don't had anything related to the x errors just alsa errors:



Feb 12 09:08:06 ski0s kernel: bridge-eth0: enabled promiscuous mode
Feb 12 09:08:37 ski0s kernel: [3876]: /dev/rtc enable interrupt failed: -1
Feb 12 09:08:37 ski0s kernel: [3878]: /dev/rtc enable interrupt failed: -1
Feb 12 09:09:06 ski0s pulseaudio[3095]: module-alsa-sink.c: ALSA woke us up to write new data to the device, but there was actually nothing to write! Most likely this is an ALSA driver bug. Please report this issue to the PulseAudio developers.
Feb 12 09:09:21 ski0s kernel: [3878]: /dev/rtc enable interrupt failed: -1
Feb 12 09:09:31 ski0s kernel: [3878]: /dev/rtc enable interrupt failed: -1
Feb 12 09:49:24 ski0s pulseaudio[3095]: module-alsa-sink.c: ALSA woke us up to write new data to the device, but there was actually nothing to write! Most likely this is an ALSA driver bug. Please report this issue to the PulseAudio developers.
Feb 12 09:50:39 ski0s pulseaudio[3095]: module-alsa-sink.c: ALSA woke us up to write new data to the device, but there was actually nothing to write! Most likely this is an ALSA driver bug. Please report this issue to the PulseAudio developers.
Feb 12 09:50:39 ski0s pulseaudio[3095]: module-alsa-sink.c: ALSA woke us up to write new data to the device, but there was actually nothing to write! Most likely this is an ALSA driver bug. Please report this issue to the PulseAudio developers.
Comment 21 Edney Matias 2009-04-06 17:12:16 EDT
Maybe this 

https://bugzilla.redhat.com/show_bug.cgi?id=487106#c5

has something to do with this bug?
Comment 22 Andrew Meakovski 2009-07-02 18:20:56 EDT
After recent update (2 jun 2009) got similar bug. 

Fc11/KDE4.2/Video intel GM965

When trying to use alt+tab or "show windows" kde composited effects — system freezes with no reaction to any input(no caps-lock blinking, no mouse moving).

Other effects (transparency, expo, etc) works well. When switching to Xrender, system also works normal.

Some possibly usefull output from /var/log/messages

pasted here:
http://dpaste.com/hold/62678/
Comment 23 Matěj Cepl 2009-07-03 12:42:36 EDT
*** Bug 508635 has been marked as a duplicate of this bug. ***
Comment 24 Philippe Troin 2009-07-04 17:15:59 EDT
This patch:

http://gitorious.org/linux-omap/mainline/commit/0e7ddf7eeeef5aea85412120539ab5369577faeb

fixes the problem for me.

Phil.
Comment 25 Kevin Kofler 2009-07-04 17:26:28 EDT
That's a patch to the DRM (direct rendering, not digital restrictions ;-) ) code in the kernel, so reassigning to kernel.
Comment 26 Philippe Troin 2009-07-05 11:23:18 EDT
Note that my issue was slightly different, as X was hanging when playing quake3.

However the Oops from comment #21 was exactly the same, which makes me think the problem is similar.

I rolled my own kernel with this patch, if anybody experiencing the KDE problem is willing to give it a try, I'll post some URLs here.

Phil.
Comment 27 Edney Matias 2009-07-06 16:26:34 EDT
I'm seeing this also but running kde 4.2.95 from redhat-kde repository. System algo hangs when i try to paly Warcraft using Wine, but not when playing Urban Terror. 

This started to happen after an update of mesa libs, but maybe it pushed some update on DRM also i didn't notice. Don't know, anyway here are my versions:

mesa-libGL-7.6-0.1.fc11.i586
mesa-libGL-devel-7.6-0.1.fc11.i586
mesa-libGLU-devel-7.6-0.1.fc11.i586
mesa-libGLU-7.6-0.1.fc11.i586
mesa-dri-drivers-7.6-0.1.fc11.i586
libdrm-2.4.6-7.fc11.i586
libdrm-devel-2.4.6-7.fc11.i586
xorg-x11-server-Xorg-1.6.1.901-2.fc11.i586

Going to try libdrm from updates-testing.
Comment 28 Edney Matias 2009-07-06 16:50:48 EDT
It didn't work. :/
Comment 29 Edney Matias 2009-07-06 21:16:51 EDT
I think this is my oops
http://www.kerneloops.org/submitresult.php?number=517690
Comment 30 Mathew Topper 2009-07-08 04:06:26 EDT
I have the same alt-tab bug with KDE4. This occurs when the OpenGL desktop effects are turned on. It only seems to be the window management effects that cause the machine to freeze along with a last beep out of KDE.

The machine is a dell vostro 1310 with an X3100 graphics card.

Possibly related is that the blue marble plasma widget says it needs opengl shader support. I'm pretty sure the x3100 has this. Could be wrong. 

The bug is 'very' repeatable. Is there any output that would be useful to diagnose the problem?
Comment 31 Mathew Topper 2009-07-08 16:40:57 EDT
My var/log/messages: (I'm running the 2.6.29.5-191.fc11.x86_64 kernel)

Jul  8 20:06:24 hopalong kernel: ------------[ cut here ]------------
Jul  8 20:06:24 hopalong kernel: kernel BUG at drivers/gpu/drm/i915/i915_gem.c:2136!
Jul  8 20:06:24 hopalong kernel: invalid opcode: 0000 [#1] SMP 
Jul  8 20:06:24 hopalong kernel: last sysfs file: /sys/devices/pci0000:00/0000:00:1b.0/sound/card0/uevent
Jul  8 20:06:24 hopalong kernel: CPU 1 
Jul  8 20:06:24 hopalong kernel: Modules linked in: cisco_ipsec(P) sco bridge stp llc bnep l2cap bluetooth coretemp hwmon sunrpc ip6t_REJECT nf_conntrack_ipv6 ip6table_filter ip6_tables ipv6 cpufreq_ondemand acpi_cpufreq freq_table fuse dm_multipath uinput snd_hda_codec_realtek lib80211_crypt_tkip snd_hda_intel wl(P) snd_hda_codec snd_hwdep snd_pcm uvcvideo r8169 snd_timer joydev snd videodev v4l1_compat v4l2_compat_ioctl32 firewire_ohci firewire_core mii wmi dell_laptop iTCO_wdt sdhci_pci sdhci dcdbas mmc_core soundcore pcspkr i2c_i801 snd_page_alloc crc_itu_t iTCO_vendor_support lib80211 i915 drm i2c_algo_bit i2c_core video output [last unloaded: microcode]
Jul  8 20:06:24 hopalong kernel: Pid: 1875, comm: X Tainted: P           2.6.29.5-191.fc11.x86_64 #1 Vostro1310
Jul  8 20:06:24 hopalong kernel: RIP: 0010:[<ffffffffa005f3d2>]  [<ffffffffa005f3d2>] i915_gem_object_get_fence_reg+0x221/0x61e [i915]
Jul  8 20:06:24 hopalong kernel: RSP: 0000:ffff880131189be8  EFLAGS: 00013202
Jul  8 20:06:24 hopalong kernel: RAX: 000000000000d815 RBX: ffff880118d1be40 RCX: 0000000000000010
Jul  8 20:06:24 hopalong kernel: RDX: 0000000000002a02 RSI: 000000000000d815 RDI: ffff88011a89cc00
Jul  8 20:06:24 hopalong kernel: RBP: ffff880131189c28 R08: 0000000000000003 R09: ffff88013d60a1e8
Jul  8 20:06:24 hopalong kernel: R10: 0000000000000200 R11: 0000000000000040 R12: ffff88011a8ad900
Jul  8 20:06:24 hopalong kernel: R13: ffff88013d60a000 R14: ffff880118d1bd80 R15: ffff88013d609000
Jul  8 20:06:24 hopalong kernel: FS:  00007f6f21ad27b0(0000) GS:ffff88013ecdd280(0000) knlGS:0000000000000000
Jul  8 20:06:24 hopalong kernel: CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
Jul  8 20:06:24 hopalong kernel: CR2: 00007f6f1ca0e000 CR3: 000000013110b000 CR4: 00000000000006e0
Jul  8 20:06:24 hopalong kernel: DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
Jul  8 20:06:24 hopalong kernel: DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
Jul  8 20:06:24 hopalong kernel: Process X (pid: 1875, threadinfo ffff880131188000, task ffff8801310d5c00)
Jul  8 20:06:24 hopalong kernel: Stack:
Jul  8 20:06:24 hopalong kernel: ffff880131189c38 ffff88013d60a1e8 ffff88013d609020 ffff880118d1bd80
Jul  8 20:06:24 hopalong kernel: ffff88013d609020 ffff880118d1be40 ffff880118d879a0 ffff880131189ce8
Jul  8 20:06:24 hopalong kernel: ffff880131189c98 ffffffffa00610d3 ffff880131189c68 00007f6f1ca0e000
Jul  8 20:06:24 hopalong kernel: Call Trace:
Jul  8 20:06:24 hopalong kernel: [<ffffffffa00610d3>] i915_gem_fault+0xc1/0x136 [i915]
Jul  8 20:06:24 hopalong kernel: [<ffffffff810b213f>] __do_fault+0x55/0x3d5
Jul  8 20:06:24 hopalong kernel: [<ffffffff812429bb>] ? agp_flush_chipset+0x1b/0x1d
Jul  8 20:06:24 hopalong kernel: [<ffffffffa005d523>] ? i915_gem_object_flush_cpu_write_domain+0x26/0x32 [i915]
Jul  8 20:06:24 hopalong kernel: [<ffffffff810b4475>] handle_mm_fault+0x349/0x7c5
Jul  8 20:06:24 hopalong kernel: [<ffffffff813ae615>] do_page_fault+0x5b5/0x9e9
Jul  8 20:06:24 hopalong kernel: [<ffffffff813ac01a>] ? unlock_kernel+0x2f/0x32
Jul  8 20:06:24 hopalong kernel: [<ffffffff810e0e27>] ? vfs_ioctl+0x76/0x87
Jul  8 20:06:24 hopalong kernel: [<ffffffff810e12bb>] ? do_vfs_ioctl+0x462/0x4a3
Jul  8 20:06:24 hopalong kernel: [<ffffffff813abab5>] ? trace_hardirqs_off_thunk+0x3a/0x6c
Jul  8 20:06:24 hopalong kernel: [<ffffffff813ac175>] page_fault+0x25/0x30
Jul  8 20:06:24 hopalong kernel: Code: ff e8 35 e9 ff ff 85 c0 0f 84 b3 fe ff ff e9 06 04 00 00 41 83 7c 24 20 00 75 10 48 8b 55 c8 48 8b 02 f7 40 70 be ff ff ff 74 04 <0f> 0b eb fe 49 8b bf 38 01 00 00 48 8b 70 38 48 85 ff 74 1a 48 
Jul  8 20:06:24 hopalong kernel: RIP  [<ffffffffa005f3d2>] i915_gem_object_get_fence_reg+0x221/0x61e [i915]
Jul  8 20:06:24 hopalong kernel: RSP <ffff880131189be8>
Jul  8 20:06:24 hopalong kernel: ---[ end trace 8e0b2086a8f4e619 ]---
Jul  8 20:06:31 hopalong kerneloops: Submitted 1 kernel oopses to www.kerneloops.org
Jul  8 20:06:31 hopalong kerneloops: URL of last kerneloops report: http://www.kerneloops.org/submitresult.php?number=522937
Comment 33 Edney Matias 2009-07-22 17:19:53 EDT
I applied some updates to my box from updates-testing but still seeing this box. Here are my new versions:

kernel 2.6.29.6-213.fc11.i686.PAE
xorg-x11-server-Xorg-1.6.2-2.fc11.i586
xorg-x11-drv-intel-2.7.0-7.fc11.i586
xorg-x11-server-common-1.6.2-2.fc11.i586
mesa-libGL-7.6-0.1.fc11.i586
mesa-libGLU-7.6-0.1.fc11.i586
mesa-dri-drivers-7.6-0.1.fc11.i586
libdrm-2.4.11-2.fc11.i586

Is there any test I can do? Maybe try out any new package.
Comment 34 Edney Matias 2009-07-31 23:34:38 EDT
Is the patch supposed to correct this problem already on kernel 2.6.29.6-213.fc11.i686.PAE? How could i apply it and use it if not?
Comment 35 Philippe Troin 2009-08-01 10:53:55 EDT
(In reply to comment #34)
> Is the patch supposed to correct this problem already on kernel
> 2.6.29.6-213.fc11.i686.PAE? How could i apply it and use it if not?  

Well, actually you need a few more patches if you want rock-solid i915 functionality.  Rock-solid is defined by:  XV, hardware GL, and XRANDR.

In addition to the above patch for the kernel, you also want xorg-x11-drv-intel 2.8.0, which is not packaged yet.

And if you want XRANDR to work with TV-out and VGA-out and GLX, turn off kms (nomodeset on the kernel command line).

I have a patched kernel and xorg-x11-drv-intel available at http://rpm.fifi.org/
Feel free to give it a shot.  Be aware that the kernel available for download there has other patches (fixed ALSA timing, l7filter, IMQ, kqemu) that you may or might not want.

Phil.
Comment 36 Edney Matias 2009-08-03 08:42:39 EDT
I updated as you said but i need to disable KMS on boot to be able to play DoTA with wine, KDE works with some glitches but no way to play without disabling KMS. 

I used to play with KMS enabled before the update that started all this but I can play again now.
Comment 37 Edney Matias 2009-08-05 07:52:48 EDT
I must correct an statement here, i do can play DoTA with the updates you suggested. It worked as soon as i ajusted the game display resolution to match the one of my desktop. Now it's fine.

But now i have these new updates coming from official updates that want to update my kernel to version 2.6.29.6-217.2.3.f11. What will happen? Is it safe to update?
Comment 38 iarly selbir 2009-08-27 10:03:34 EDT
Hi all, it seems fixed since kernel 2.6.29, some days ago it back ( look bug https://bugzilla.redhat.com/show_bug.cgi?id=509519 ) and now is fixed again

I'm using f11 fully updated + kde 4.3

I guess it fixed to everybody as well, I hope =)

--
Fedora Bug Triage Team Member
https://fedoraproject.org/wiki/BugZappers
Comment 39 Bug Zapper 2009-11-18 04:27:01 EST
This message is a reminder that Fedora 10 is nearing its end of life.
Approximately 30 (thirty) days from now Fedora will stop maintaining
and issuing updates for Fedora 10.  It is Fedora's policy to close all
bug reports from releases that are no longer maintained.  At that time
this bug will be closed as WONTFIX if it remains open with a Fedora 
'version' of '10'.

Package Maintainer: If you wish for this bug to remain open because you
plan to fix it in a currently maintained version, simply change the 'version' 
to a later Fedora version prior to Fedora 10's end of life.

Bug Reporter: Thank you for reporting this issue and we are sorry that 
we may not be able to fix it before Fedora 10 is end of life.  If you 
would still like to see this bug fixed and are able to reproduce it 
against a later version of Fedora please change the 'version' of this 
bug to the applicable version.  If you are unable to change the version, 
please add a comment here and someone will do it for you.

Although we aim to fix as many bugs as possible during every release's 
lifetime, sometimes those efforts are overtaken by events.  Often a 
more recent Fedora release includes newer upstream software that fixes 
bugs or makes them obsolete.

The process we are following is described here: 
http://fedoraproject.org/wiki/BugZappers/HouseKeeping
Comment 40 Steven M. Parrish 2009-11-18 16:07:52 EST
Based on the recent comments it appears this has been resolved.  Going to close it as WORKSFORME.  Please feel free to reopen if the problem returns.

Note You need to log in before you can comment on or make changes to this bug.