Bug 500994

Please attach the information requested on the webpage:

http://en.opensuse.org/Bugs:X

Created attachment 290155 [details]
xorg.conf

Created attachment 290156 [details]
Xorg.0.log

Created attachment 290157 [details]
Xorg.0.log.old

Created attachment 290158 [details]
pm-suspend.log

Created attachment 290159 [details]
Xorg.0.log.old

Created attachment 290161 [details]
kdm.log

rpm --changelog -q libdrm | head -n3 
rpm --changelog -q Mesa | head -n3 
rpm --changelog -q xorg-x11-server | head -n3 
rpm --changelog -q xorg-x11-driver-input | head -n3 
rpm --changelog -q xorg-x11-driver-video | head -n3 
rpm --changelog -q xorg-x11-driver-video-radeonhd | head -n3 (if you use radeonhd driver) 
rpm --changelog -q xorg-x11-driver-video-unichrome | head -n3 (if you use unichrome driver)

================================================

pavel@linux-trdb:/etc/X11> rpm --changelog -q libdrm | head -n3
* Fri Oct 31 2008 sndirsch@suse.de
- libdrm 2.4.1
  * New libdrm release to fix the symbol name collision between
pavel@linux-trdb:/etc/X11> rpm --changelog -q Mesa | head -n3
* Tue Feb 10 2009 sndirsch@suse.de
- commit-7d99ddc.diff
  * intel: Fix a number of memory leaks on context destroy.
pavel@linux-trdb:/etc/X11> rpm --changelog -q xorg-x11-server | head -n3
* Tue Feb 24 2009 sndirsch@suse.de
- bug474071-fix1.diff
  * fixes Xserver issue of bnc #474071
pavel@linux-trdb:/etc/X11> rpm --changelog -q xorg-x11-driver-input | head -n3
* Wed Jan 14 2009 mhopf@suse.de
- xf86-input-evdev-sig11-bug437651.diff
  Fix evdev segfault (bnc #437651)
pavel@linux-trdb:/etc/X11> rpm --changelog -q xorg-x11-driver-video | head -n3
* Tue Feb 24 2009 sndirsch@suse.de
- xf86-video-ati-bug474071-fix2.diff
  * fixes driver issue of bnc #474071
pavel@linux-trdb:/etc/X11> rpm --changelog -q xorg-x11-driver-video-radeonhd | head -n3
* Thu Jan 15 2009 sndirsch@suse.de
- Identity.map
  * use device id as device name if empty (bnc #465154)
pavel@linux-trdb:/etc/X11> rpm --changelog -q xorg-x11-driver-video-unichrome | head -n3
* Mon Aug 18 2008 ro@suse.de
- change prereq to coreutils

================================================

linux-trdb:/etc/X11 # hwinfo --gfx
09: PCI 02.0: 0300 VGA compatible controller (VGA)              
  [Created at pci.318]                                          
  UDI: /org/freedesktop/Hal/devices/pci_8086_2a02               
  Unique ID: _Znp.3gR64TvADaC                                   
  SysFS ID: /devices/pci0000:00/0000:00:02.0
  SysFS BusID: 0000:00:02.0
  Hardware Class: graphics card
  Model: "Lenovo T61"
  Vendor: pci 0x8086 "Intel Corporation"
  Device: pci 0x2a02 "965 GM"
  SubVendor: pci 0x17aa "Lenovo"
  SubDevice: pci 0x20b5 "T61"
  Revision: 0x0c
  Memory Range: 0xf8100000-0xf81fffff (rw,non-prefetchable)
  Memory Range: 0xe0000000-0xefffffff (rw,prefetchable)
  I/O Ports: 0x1800-0x1807 (rw)
  IRQ: 16 (10370188 events)
  I/O Ports: 0x3c0-0x3df (rw)
  Module Alias: "pci:v00008086d00002A02sv000017AAsd000020B5bc03sc00i00"
  Driver Info #0:
    XFree86 v4 Server Module: intel
  Driver Info #1:
    XFree86 v4 Server Module: intel
    3D Support: yes
    Extensions: dri
  Config Status: cfg=no, avail=yes, need=no, active=unknown

10: PCI 02.1: 0380 Display controller
  [Created at pci.318]
  UDI: /org/freedesktop/Hal/devices/pci_8086_2a03
  Unique ID: ruGf.a6pkzICrUB2
  SysFS ID: /devices/pci0000:00/0000:00:02.1
  SysFS BusID: 0000:00:02.1
  Hardware Class: graphics card
  Model: "Lenovo T61"
  Vendor: pci 0x8086 "Intel Corporation"
  Device: pci 0x2a03 "Mobile GM965/GL960 Integrated Graphics Controller"
  SubVendor: pci 0x17aa "Lenovo"
  SubDevice: pci 0x20b5 "T61"
  Revision: 0x0c
  Memory Range: 0xf8200000-0xf82fffff (rw,non-prefetchable)
  Module Alias: "pci:v00008086d00002A03sv000017AAsd000020B5bc03sc80i00"
  Config Status: cfg=no, avail=yes, need=no, active=unknown

Primary display adapter: #9

================================================

I would need a Xorg.0.log.old right after such a Xserver crash.

I mean I don't see any Xserver crash in the logfiles you've attached.

(In reply to comment #9)
> I would need a Xorg.0.log.old right after such a Xserver crash.

i'll attach it as soon as possible...

Created attachment 290296 [details]
After crashing X-server... and login again..

Created attachment 290297 [details]
After crashing X-server... and login again.. (kdm.log)

Created attachment 290298 [details]
After crashing X-server... and login again.. (Xorg.0.log)

After crashing X-server... and login again.. (file var/log/messages). Look at string 401... may be it's something showing us ablut crashing?

http://paste.org.ru/?4jyfke

(In reply to comment #13)
> Created an attachment (id=290297) [details]
> After crashing X-server... and login again.. (kdm.log)

That's it.

Xorg: intel_bufmgr_fake.c:769: drm_intel_bufmgr_fake_contended_lock_take: Assertion `_fence_test(bufmgr_fake, block->fence)' failed.

It's a dup.

*** This bug has been marked as a duplicate of bug 470725 ***

(In reply to comment #17)
> It's a dup.
> 
> *** This bug has been marked as a duplicate of bug 470725 ***

To resolve the bug 470725 there recommendations is to update video drivers from http://download.opensuse.org/repositories/X11:/XOrg/openSUSE_11.1/
BTW if I do it... i have another bug - 483424

so... i should reopen this bug again...

Stefan..  i've googled something about bug 483424 see http://forums.opensuse.org/install-boot-login/409510-x11-xorg-repository.html

what can you tell me?

(In reply to comment #19)
> so... i should reopen this bug again...

I don't want this. Probably you can workaround the initial issue for now by adding

  Disable "DRI"

to Modules section of xorg.conf and not using compiz/desktop effects.

(In reply to comment #20)
> Stefan..  i've googled something about bug 483424 see
> http://forums.opensuse.org/install-boot-login/409510-x11-xorg-repository.html
> 
> what can you tell me?

We still can't reproduce this issue inhouse.

(In reply to comment #21)
> (In reply to comment #19)
> > so... i should reopen this bug again...
> 
> I don't want this. Probably you can workaround the initial issue for now by
> adding
> 
>   Disable "DRI"
> 
> to Modules section of xorg.conf and not using compiz/desktop effects.
> 
> (In reply to comment #20)
> > Stefan..  i've googled something about bug 483424 see
> > http://forums.opensuse.org/install-boot-login/409510-x11-xorg-repository.html
> > 
> > what can you tell me?
> 
> We still can't reproduce this issue inhouse.

I put Disable "DRI"... in section "Module" of my xorg.conf but after reading http://www.puppylinux.com/blog/?viewDetailed=00522 i added also  - Option "DRI" "0" in the section "Device"
Then after restarting X i won't use kwin effects.. Stefan correct me if i do not right something...

and id i don't use DRI... i can't play some opengl game.. such as OpenArena? 
ps. i've run it now and it runs very slowly...

so.. i can only disable kwin effects only? And i've not need to change my xorg.conf? (add Disable "DRI" option..), have I?

StefaN?

You're right, you might need 'Option "DRI" "off" as well. With disabled DRI 3D support is no longer available.

(In reply to comment #26)
> You're right, you might need 'Option "DRI" "off" as well. With disabled DRI 3D
> support is no longer available.

Hi, Stefan. I suppose fix it..
I modified my xorg.conf in section "Device" :

Section "Device"
  BoardName    "965 GM"
  Driver       "intel"
  Identifier   "Device[0]"
  Option       "monitor-LVDS" "Monitor[0]"
  VendorName   "Intel"
  Option "AccelMethod" "exa"
  Option "MigrationHeuristic" "smart" #greedy
  Option "ExaNoComposite" "false"
  Option "XaaNoOffscreenPixmaps" "true"
  Option "EXAOptimizeMigration" "true"
  Option "DRI" "on"
EndSection

Now it works fine. I use Intel video drivers from "openSUSE 11.1 : Update" repo.

> Option "AccelMethod" "exa"

default anyway

> Option "MigrationHeuristic" "smart" #greedy

ok.

> Option "ExaNoComposite" "false"

we set this by default

> Option "XaaNoOffscreenPixmaps" "true"

doesn't matter when using EXA

> Option "EXAOptimizeMigration" "true"

ok.

> Option "DRI" "on"

default anyway

So it seems setting  MigrationHeuristic(smart) and or EXAOptimizeMigration(true) workarounds that issue.

I have bad news... today my xserver has crashed again... (after suspend)..
it's my log http://pastie.org/477155

Stefan, i'll try to update my BIOS.

linux-trdb:/home/pavel # dmidecode -s bios-version
7NETB7WW (2.17 )

But here is updates:
http://www-307.ibm.com/pc/support/site.wss/document.do?sitestyle=lenovo&lndocid=MIGR-67983
here 

Version:	2.20
Release Date:	2009/04/24

ps. and also Version 7NETB8WW (2.18) havs already (New) Intel Video BIOS update.

so.. i'll update my bios and then try X11:Xorg packages...

So.. i has sucuessfuly updated my BIOS version to 2.20. But anyway after updating intel video drivers from X11:Xorg repo.. i get black screen.. bug #483424

Stefan, do you have any answers for my trouble?

Unfortunately I don't.

Stefan, can you tell me configuration of the machine where have you test it on?

i want to know - Linux Kernel version also...


ps. Is it true that X11:Xorg packages work only with kernel version >= 2.6.28 ?

(In reply to comment #34)
> Stefan, can you tell me configuration of the machine where have you test it on?

Did I say so?

> ps. Is it true that X11:Xorg packages work only with kernel version >= 2.6.28 ?

Not at all. Who says this?

(In reply to comment #35)
> (In reply to comment #34)
> > Stefan, can you tell me configuration of the machine where have you test it on?
> 
> Did I say so?
> 

Sorry.. I thought that you'd tested

> > ps. Is it true that X11:Xorg packages work only with kernel version >= 2.6.28 ?
> 
> Not at all. Who says this?


One man on the ICR channel.. may be ho isn't right (i also checked with linux kernel 2.6.29 after upgrading my BIOS and all packages from X11:XOrg - the same results.. doesn't work.. only black screen).


PS. I've also noticed that when this screen appears i can see "_" symbol in the top left corner of the screen (there are only two key combination which i can do: Fn+F5 - switch BT and POWER BTN about 3 sec to turn off laptop). Except this i can introduce you all neccesary logs after it again.. may be i'd did something wrong when I placed it on this site..

Yesterday i boot from KDE4 LIVE CD (from factory) which i'd downloded here : 
http://download.opensuse.org/factory/iso/

xserver starts fine...

(i think that there is the same packages that in the X11:Xorg repo.. )


So.. now i want you to read my message before this one again and answer me )

Stefan i'd like to add that it seemd to me that this black screen appears always on my working system (without updating from X11:Xorg) while xserver starting... but it's during is about 1-2 seconds.. and then appears mouse cursor etc.. and xserver starts.. but after updating from X11:Xorg it appears and that's all.. (black screen and xserver doesn't start..)..

(In reply to comment #37)
> Yesterday i boot from KDE4 LIVE CD (from factory) which i'd downloded here : 
> http://download.opensuse.org/factory/iso/
> 
> xserver starts fine...
> 
> (i think that there is the same packages that in the X11:Xorg repo.. )

Isn't this a relaunch of openSUSE 11.1? Then for sure the X11:XOrg packages are not included.

(In reply to comment #39)
> (In reply to comment #37)
> > Yesterday i boot from KDE4 LIVE CD (from factory) which i'd downloded here : 
> > http://download.opensuse.org/factory/iso/
> > 
> > xserver starts fine...
> > 
> > (i think that there is the same packages that in the X11:Xorg repo.. )
> 
> Isn't this a relaunch of openSUSE 11.1? Then for sure the X11:XOrg packages are
> not included.

I don't know... it's factory iso of opensuse 11.1 which will be 11.2 soon

> it's factory iso of opensuse 11.1

What ?!? Factory and OS 11.1 are different !!!

(In reply to comment #41)
> > it's factory iso of opensuse 11.1
> 
> What ?!? Factory and OS 11.1 are different !!!

yes

Stefan. Finally, I resolved this bug by updating ALL (!) packages from X11:Xorg repo.

:)

Cheers!

Stefan, i have bad news(

Today i wake up my laptop from suspend and xserver crashed againg...(got login screen)


here is kdm.log: linux-trdb:/var/log # grep intel_buf kdm.log
Xorg: intel_bufmgr_fake.c:774: drm_intel_bufmgr_fake_contended_lock_take: Assertion `_fence_test(bufmgr_fake, block->fence)' failed.

PS. It's with last update from X11:Xorg..

Is this easily reproducable? Are you still using

  Option "MigrationHeuristic" "smart"
  Option "EXAOptimizeMigration" "true"

(In reply to comment #46)
> Is this easily reproducable? Are you still using
> 
>   Option "MigrationHeuristic" "smart"
>   Option "EXAOptimizeMigration" "true"

No, i don't use this 2 options. I've commented it in xorg.conf

Thanks. Still the first question remains. Is this easily reproducable?

1. Turn my laptop on
2. Enter user login and pass
3. Loading kde4
4. Begin working ... (launch some programms)
5. Suspend to disk
6. Come back by turning laptop on
7. Xserver restart (by not always)....sometime it's normal

*** Bug 470725 has been marked as a duplicate of this bug. ***

(In reply to comment #50)
> *** Bug 470725 has been marked as a duplicate of this bug. ***

As for me after latest updates from X11:Xorg xserver sometimes crashes...

*** Bug 470725 has been marked as a duplicate of this bug. ***

Stefan, any idea how to fix this bug?

Not at all.

(In reply to comment #54)
> Not at all.

Stefan, i was updating yesterday some packages from X11:Xorg, may be they were all not properly updated last time.. and then start testing:

1. After ~10 cycles (suspend to disk <--> wake up) system work good. (xserver didn't crash)

2. But 1 time a have a worse moment: suddenly after suspend there were a message: somethink like this " composite too slow if you want return press alt+shift+f12". After this i press this key combination and the strange bugs began (see paragraph number 3)

3. If i press alt+shift+f12 desktop hides and i get grey background with several lines.. (may be default suse wallpaper?) then if i do alt+shift+f12 again composite still doesn't turn on and i get windows of programms but not a panle at the bottom of the screen...

4. So... to continue normal working with kwin i needed to reboot...

Now, i have  a question may be you can help me How can i turn off this message? I don't want to meet this bug in the future..

I think with all the issues you're seing it would be the best to turn off 3D support or at least kwin effects.

Stefan i don't wan't to use kde 4.2 without 3d effects)

i'll mark this bug as "RESOLVED" for me now

Today i do suspend to disk with a big percent of RAM used.. (using oracle DB and etc.). After suspend to disk screen hangs (but it wasn't black color, but more.. red). Then after booting to init 3 i found in kdm.log: http://paste.org.ru/?8sjwfg

PS. Used last updated from X11:Xorg

Stefan, if last patches really work.. why i have the same errors? (Xorg: intel_bufmgr_fake.c:774: drm_intel_bufmgr_fake_contended_lock_take: Assertion `_fence_test(bufmgr_fake, block->fence)' failed)

Or may be it's because my RAM was almost full?

Well, you closed the bug as resolved fixed, not me.

(In reply to comment #60)
> Well, you closed the bug as resolved fixed, not me.

Yes, but today xserver hangs after suspend to disk... so what should i do now? open the new bug?

Why another one? You reopened the bug already.

And i suppose that you couldn't help me with issue "intel_bufmgr_fake.c:774"?

The bug is set to assigned. We'll have a look at it - when time permits. It's not that easy to reproduce and debug you might imagine ...

(In reply to comment #64)
> The bug is set to assigned. We'll have a look at it - when time permits. It's
> not that easy to reproduce and debug you might imagine ...

Stefan, do you have any news?

As long as there are no new comments by us there are no news. It's that simple.

Created attachment 295375 [details]
Xorg.0.log

Today Xserver crashed again after wake up from suspend to disk. Bottom i'll add attachments logs

Created attachment 295376 [details]
Xorg.0.log.old

Created attachment 295377 [details]
localmessages

Created attachment 295378 [details]
messages

Created attachment 295379 [details]
pm-suspend.log

Created attachment 295380 [details]
warn

I'd like to add that "crashing" was at time ~ 15:28 - 15:32

May be it'll help to solve this issue

ps. May 31 15:28:00 - 15:32:00

applications which were running before:

* konversation
* kmail
* opera
* dolphin
* qutim

err.  that's all..

Suspicious:

May 31 15:29:26 linux-trdb kernel: Xorg:23002 conflicting memory types e0000000-f0000000 write-combining<->uncached-minus
May 31 15:29:26 linux-trdb kernel: reserve_memtype failed 0xe0000000-0xf0000000, track write-combining, req write-combining
May 31 15:29:27 linux-trdb kernel: Xorg:23002 conflicting memory types e0000000-f0000000 write-combining<->uncached-minus
May 31 15:29:27 linux-trdb kernel: reserve_memtype failed 0xe0000000-0xf0000000, track write-combining, req write-combining
May 31 15:29:27 linux-trdb kernel: Xorg:23007 freeing invalid memtype e0000000-f0000000
May 31 15:29:27 linux-trdb kernel: Xorg:23002 conflicting memory types e0000000-f0000000 write-combining<->uncached-minus
May 31 15:29:27 linux-trdb kernel: reserve_memtype failed 0xe0000000-0xf0000000, track write-combining, req write-combining
May 31 15:29:27 linux-trdb kernel: Xorg:23008 freeing invalid memtype e0000000-f0000000
May 31 15:29:27 linux-trdb kernel: Xorg:23002 conflicting memory types e0000000-f0000000 write-combining<->uncached-minus
May 31 15:29:27 linux-trdb kernel: reserve_memtype failed 0xe0000000-0xf0000000, track write-combining, req write-combining
May 31 15:29:27 linux-trdb kernel: Xorg:23009 freeing invalid memtype e0000000-f0000000
May 31 15:29:27 linux-trdb kernel: Xorg:23002 conflicting memory types e0000000-f0000000 write-combining<->uncached-minus
May 31 15:29:27 linux-trdb kernel: reserve_memtype failed 0xe0000000-0xf0000000, track write-combining, req write-combining
May 31 15:29:27 linux-trdb kernel: Xorg:23010 freeing invalid memtype e0000000-f0000000

What does it mean?

Stefan, i've googled for this 
http://bbs.archlinux.org/viewtopic.php?id=60723
may be adding "nopat" to kernel startup string will resolve this issue?

Feel free to give it a try.

(In reply to comment #79)
> Feel free to give it a try.

Ok.. or may be i have to test my laptop with only 2gb RAM (instead of 4gb ).. hm

Today xserver have crashed again... (with added "nopat" option)...

in kdm.log:
Xorg:
intel_bufmgr_fake.c:774: drm_intel_bufmgr_fake_contended_lock_take: Assertion
`_fence_test(bufmgr_fake, block->fence)' failed

Hi Stefan. I have some news..

I've switched back to GNOME. Xserver still crashes after hibernate sometimes... 

grep drm /var/log/messages gives: 
Jun  6 18:24:41 linux-trdb kernel: [drm:i915_getparam] *ERROR* Unknown parameter 5

PS. I've also turned off the hardware signature by adding acpi_sleep=s4_nohwsig as the kernel parameter.

Or may be i need install new Kernel 2.6.28 with GEM support and that might fix this bug?

also find dublicate (ubuntu) - https://bugs.launchpad.net/ubuntu/+source/linux/+bug/379417

another upstream bug like this https://bugs.freedesktop.org/show_bug.cgi?id=19731

I would suggest testing 11.2 MI8 - or the latest X packages from X11:XOrg if you don't want to update your complete system.

There have been quite a number of fixes to the intel driver, so chances are that your issue is fixed as well. I wouldn't hold my breath, though.

Any chance to give 11.2-RC1 or the latest X packages from X11:Xorg buildservice project a try?

Let's assume the issue has been fixed with packages from obs://X11:Xorg and openSUSE 11.2 >= RC2. Otherwise please reopen. Thanks.