[i965] Resume from suspend leaves me with black screen or a screen of the desktop before it suspended. Compiz hung in intel_update_renderbuffers() from intel_prepare_render() from brw_draw_prims()

Bug #966744 reported by Anmar Oueja
This bug affects 463 people
Affects Status Importance Assigned to Milestone
Compiz
Invalid
Critical
Unassigned
Linux
Fix Released
High
xf86-video-intel
Fix Released
High
xserver-xorg-video-intel (Ubuntu)
Fix Released
Critical
Ivander Loe-A-Foe
Precise
Fix Released
Critical
Timo Aaltonen
Quantal
Fix Released
Critical
Timo Aaltonen

Bug Description

  ****************ATTENTION*****************
  THIS BUG HAS BEEN MARKED AS FIXED AND THE ISSUE HAS BEEN RESOLVED.
  ANYONE EXPERIENCING THESE SYMPTOMS ON A COMPLETELY UP-TO-DATE SYSTEM SHOULD LOG A NEW BUG:

  'ubuntu-bug xserver-xorg-video-intel'

  PLEASE REFRAIN FROM UNNECESSARY COMMENTS. THANKS!
  ****************ATTENTION*****************

[Impact]:

severe, compiz hangs frequently after resuming from suspend or the screensaver

[Testcase]:

- grab xserver-xorg-video-intel from quantal
- build it from the extracted directory
- after the build run 'cd test && make dri2-swap'
- log in X
- start two terminals
- get dpms.sh from this bug, run it in the other terminal
- on the other run ./dri2-swap

compiz should hang during the first couple of runs of dri2-swap

with the fixed package, it doesn't

[Regression potential]

To ease the backporting, I did replace patch 101 with the commit that was pulled upstream. This patch is needed for smooth transition from plymouth to X. There _could_ be some issues with it (meaning additional patches needed from upstream), but it seems to be working just fine for me.

The precise upload has another upstream fix for an issue referenced in bug 992391, it should be safe.

---

This bug is about problems after resume, with Intel video hardware and i965_dri.so. If you're having these symptoms with other hardware, Nvidia or ATI/Radeon, this isn't the place for reports or questions.

[Those comments are building up, so brief links to direct other people to appropriate bugs might be good here, if anyone has them. Thanks.]

----------------------------------------------------------------------------------------

This happens intermittently. Sometimes upon resume, all I see is a black screen and the cursor. the mouse and keyboard respond (the mouse moves) but nothing changes. Switching to console and back doesn't fix it. Killing Compiz also does not fix the issue. Further, if one has password lock disabled on resume the user will be stuck with a frozen display of whatever was last shown on the screen before suspend. The mouse cursor still changes as you hover over various elements but the display is frozen solid. One can switch to another TTY via ctrl+alt+F1 but killing the X session is the only way to get back to the desktop.

Attempts to replace compiz have been fruitless. Trying the 3.4 kernel also did not fix the issue.

BTW, I haven nothing connected to the laptop when I resume and before suspend. This is a very annoying issue because the only way to recover is to reboot, which means all the open files will be lost.

[UPDATE]
It was discovered that switching to a tty, killing compiz, then invoking 'metacity --replace' would successfully switch over to metacity. Attempting to invoke 'compiz --replace' afterwards would put the state right back where it was (frozen screen, etc.) and output error messages saying something similar to "invalid configuration, this should never happen, you should probably raise a bug report about it". After killing Compiz again and invoking compiz --replace yet again I can get basic window management back (though unity --replace still does not work as it whines about unclean exit states). So one still needs to start over again to get it all functioning properly.

ProblemType: Bug
DistroRelease: Ubuntu 12.04
Package: unity 5.8.0-0ubuntu2
ProcVersionSignature: Ubuntu 3.2.0-20.33-generic 3.2.12
Uname: Linux 3.2.0-20-generic x86_64
.tmp.unity.support.test.0:

ApportVersion: 1.95-0ubuntu1
Architecture: amd64
CompizPlugins: [core,composite,opengl,compiztoolbox,decor,vpswitch,snap,mousepoll,resize,place,move,wall,grid,regex,imgpng,session,gnomecompat,animation,fade,unitymtgrabhandles,workarounds,scale,expo,ezoom,unityshell]
CompositorRunning: compiz
Date: Tue Mar 27 21:47:25 2012
DistUpgraded: Fresh install
DistroCodename: precise
DistroVariant: ubuntu
InstallationMedia: Ubuntu 12.04 LTS "Precise Pangolin" - Alpha amd64 (20120124)
MachineType: LENOVO 32492HU
ProcKernelCmdLine: BOOT_IMAGE=/boot/vmlinuz-3.2.0-20-generic root=UUID=8afc43b9-4675-46b0-8bf8-49bfd422c2c7 ro quiet splash i915.i915_enable_rc6=1 vt.handoff=7
SourcePackage: unity
UpgradeStatus: No upgrade log present (probably fresh install)
dmi.bios.date: 06/29/2011
dmi.bios.vendor: LENOVO
dmi.bios.version: 6QET67WW (1.37 )
dmi.board.name: 32492HU
dmi.board.vendor: LENOVO
dmi.board.version: Not Available
dmi.chassis.asset.tag: No Asset Information
dmi.chassis.type: 10
dmi.chassis.vendor: LENOVO
dmi.chassis.version: Not Available
dmi.modalias: dmi:bvnLENOVO:bvr6QET67WW(1.37):bd06/29/2011:svnLENOVO:pn32492HU:pvrThinkPadX201:rvnLENOVO:rn32492HU:rvrNotAvailable:cvnLENOVO:ct10:cvrNotAvailable:
dmi.product.name: 32492HU
dmi.product.version: ThinkPad X201
dmi.sys.vendor: LENOVO
version.compiz: compiz 1:0.9.7.2-0ubuntu4
version.ia32-libs: ia32-libs 20090808ubuntu35
version.libdrm2: libdrm2 2.4.32-1ubuntu1
version.libgl1-mesa-dri: libgl1-mesa-dri 8.0.2-0ubuntu2
version.libgl1-mesa-dri-experimental: libgl1-mesa-dri-experimental N/A
version.libgl1-mesa-glx: libgl1-mesa-glx 8.0.2-0ubuntu2
version.xserver-xorg-core: xserver-xorg-core 2:1.11.4-0ubuntu7
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

X-HWE-Bug: Bug #1004771

Revision history for this message
Anmar Oueja (anmar) wrote :
Revision history for this message
Launchpad Janitor (janitor) wrote :

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

Changed in unity (Ubuntu):
status: New → Confirmed
Revision history for this message
Alessandro Paolino (mark2dp) wrote :

Same for me, it's like the login prompt is under a a black screen (passing the cursor on the center of the screen it changes like when you are for write a text)

Revision history for this message
George Swann (g-swann) wrote :

I've been having the same problem with the various versions of unity 5.10. I've noticed that that when you resume the 1st time after a reboot things work as expected. From the 2/3rd resume from suspend there's a strong chance of this bug occuring.

Revision history for this message
Eugene Romanenko (eros2) wrote :

Same for me, Fujitsu M2010 notebook.

Revision history for this message
Alessandro Paolino (mark2dp) wrote :

It occurred another time today (Ubuntu is up to date), I had to reboot my pc

Revision history for this message
Kevin Hausmann (salema) wrote :

I am seeing this on a XPS 1330 (intel graphics) on each suspend/resume cycle.

Revision history for this message
Anmar Oueja (anmar) wrote : Re: [Bug 966744] Re: resume from suspend leaves me with black screen and a cursor

I have this every once in a while.  Not sure what to do when it happens other than lose all my work and force reboot.

Revision history for this message
Sebastien Bacher (seb128) wrote : Re: resume from suspend leaves me with black screen and a cursor

Thank you for your bug report, is that specific to unity sessions? Could you add your /var/log/Xorg.0.log and Xorg.0.log.old when that happens? What videocard and driver do you use?

Changed in unity (Ubuntu):
importance: Undecided → High
status: Confirmed → Incomplete
Revision history for this message
Éric Piel (pieleric) wrote :

I think I'm affected by a similar bug. Also bugs #975848 and #986600 are probable duplicates. Bug #914557 could also be a duplicate but it's quite older and the computer affected is using a nvidia graphic driver instead of an intel one.

A few remarks from my side. When this happens, I can still move the cursor (which changes shape depending on what type of widget is under it), and change to a console (Ctrl + Alt + F1).

If I kill compiz ("killall -KILL compiz" twice in the console), and restart it ("compiz -replace -display :0"), I get the screen back. However, I'm probably restarting compiz incorectly, and I miss a window-manager (but at least I can see everything and quit each application).

Revision history for this message
Anmar Oueja (anmar) wrote :

ii xorg 1:7.6+12ubuntu1 X.Org X Window System
ii xorg-docs-core 1:1.6-1ubuntu2 Core documentation for the X.org X Window System
ii xserver-xorg 1:7.6+12ubuntu1 X.Org X server
ii xserver-xorg-core 2:1.11.4-0ubuntu10 Xorg X server - core server
ii xserver-xorg-input-all 1:7.6+12ubuntu1 X.Org X server -- input driver metapackage
ii xserver-xorg-input-evdev 1:2.7.0-0ubuntu1 X.Org X server -- evdev input driver
ii xserver-xorg-input-mouse 1:1.7.1-1build3 X.Org X server -- mouse input driver
ii xserver-xorg-input-synaptics 1.5.99.902-0ubuntu5 Synaptics TouchPad driver for X.Org server
ii xserver-xorg-input-vmmouse 1:12.8.0-1 X.Org X server -- VMMouse input driver to use with VMWare
ii xserver-xorg-input-wacom 1:0.14.0-0ubuntu2 X.Org X server -- Wacom input driver
ii xserver-xorg-video-all 1:7.6+12ubuntu1 X.Org X server -- output driver metapackage
ii xserver-xorg-video-fbdev 1:0.4.2-4ubuntu2 X.Org X server -- fbdev display driver
ii xserver-xorg-video-intel 2:2.17.0-1ubuntu4 X.Org X server -- Intel i8xx, i9xx display driver

and

ii unity 5.10.0-0ubuntu6 Interface designed for efficiency of space and interaction.

I am using the thinkpad X201.

Revision history for this message
jhfhlkjlj (fdsuufijjejejejej-deactivatedaccount) wrote :

12.04 final, happened to me today. Didn't get to try to kill/restart compiz as I accidentally rebooted the computer. Attaching my xorg.log and xorg.log.old

Revision history for this message
jhfhlkjlj (fdsuufijjejejejej-deactivatedaccount) wrote :
Revision history for this message
jhfhlkjlj (fdsuufijjejejejej-deactivatedaccount) wrote :

As it's likely Compiz rather than unity itself, I'm reassigning for now.

affects: unity (Ubuntu) → compiz (Ubuntu)
Changed in compiz (Ubuntu):
status: Incomplete → Confirmed
Changed in compiz-core:
status: New → Confirmed
importance: Undecided → High
milestone: none → 0.9.8.0
Revision history for this message
Amos Blanton (lightnin9) wrote :

Same thing, Precise final, xorg log attached.
Thinkpad X220

Note - if I hit ctrl - alt F2 (or 3) to get to a terminal, login and then sudo killall Xorg, I can then get back to the login prompt and login. Not a nice workaround, but a workaround nonetheless.

Revision history for this message
jhfhlkjlj (fdsuufijjejejejej-deactivatedaccount) wrote :

Lightnin, next time you have this issue, could you see if killing/replacing compiz fixes the issue?

Revision history for this message
Andy Clayton (q3aiml) wrote :

I have a x220 as well and have run into this. I've tried restarting compiz via unity --replace but it does not fix it. Nothing stuck out to me with --debug while trying to restart it, but unfortunately I didn't keep a copy that I can attach.

Revision history for this message
dominik (schmidt-dominik) wrote :

I might add that I am experiencing the here reported bug but it did also happen that I saw my wallpaper but the login window didn't show up. Not sure if it is for the same reason. I have reported a duplicate bug here https://bugs.launchpad.net/bugs/986600.

Revision history for this message
jhfhlkjlj (fdsuufijjejejejej-deactivatedaccount) wrote :

Alright, this seems to be a problem with intel's drivers and not Compiz or Unity.

Changed in compiz (Ubuntu):
status: Confirmed → Triaged
status: Triaged → Invalid
affects: compiz-core → xserver-xorg-video-intel (Ubuntu)
Changed in xserver-xorg-video-intel (Ubuntu):
milestone: 0.9.8.0 → none
Revision history for this message
jhfhlkjlj (fdsuufijjejejejej-deactivatedaccount) wrote :

Okay, actually, perhaps I shouldn't have gone and done all that without confirming. Is everyone here on intel drivers?

Also, since this had a milestone assigned to it was this indeed known to be a defect in compiz?

affects: xserver-xorg-video-intel (Ubuntu) → compiz-core
Changed in compiz (Ubuntu):
status: Invalid → Confirmed
Revision history for this message
Anmar Oueja (anmar) wrote : Re: [Bug 966744] Re: resume from suspend leaves me with black screen and a cursor

I am on Lenovo Thinkpad X201 using intel drivers -
xserver-xorg-video-intel - 2:2.17.0-1ubuntu4

Revision history for this message
dominik (schmidt-dominik) wrote : Re: resume from suspend leaves me with black screen and a cursor

i'm on a thinkpad x200s running intel drivers as well.

Revision history for this message
Paul Arnold (athodeca) wrote :

Same problem here. Asus Eeepc 1005p and 1015p. Intel drivers. I have password on resume disabled and instead of a black screen, I see the desktop. Mouse cursor responds but no icons or the launcher is active. Restarting X with altgr-print-k unfreezes display and returns to login screen. Display is then functional again.

affects: compiz-core → xserver-xorg-video-intel (Ubuntu)
Changed in compiz (Ubuntu):
status: Confirmed → Invalid
Changed in xserver-xorg-video-intel (Ubuntu):
status: Confirmed → Triaged
description: updated
description: updated
Revision history for this message
Stéphane Guillou (stephane-guillou) wrote :

How do you check for the drivers? In "Details", I can see that my processor is an "Intel® Atom™ CPU N270 @ 1.60GHz × 2" but when I go in "graphics", the "driver" shows "unknown".
I'm on a Samsung N310 (netbook).

Revision history for this message
jhfhlkjlj (fdsuufijjejejejej-deactivatedaccount) wrote :

Stéphane, that machine uses intel integrated. It should have told you that in the 'driver' section but I just did a quick search.

Revision history for this message
betzi (s-betzinger-t) wrote :

Same here, ran into the bug after installing 12.04 on my ThinkPad X200s wiht xserver-xorg-video-intel driver...
Seems to be related to the intel graphicsdriverproblems under windows... (after suspend no change of brightness possible)

Overall it's all intel...

Revision history for this message
John Larsen (dmessiahpchan) wrote :

Having this problem using Intel driver on a Toshiba Satellite Pro S300L

Revision history for this message
Lazy (ubuntu-bugs-oittaa) wrote :

https://bugs.launchpad.net/ubuntu/+source/linux/+bug/977135

Seems like I have opened a ticket regarding the same problem.

affects: compiz (Ubuntu) → linux (Ubuntu)
Changed in linux (Ubuntu):
status: Invalid → Triaged
status: Triaged → Incomplete
Changed in linux:
importance: Unknown → High
status: Unknown → Incomplete
Revision history for this message
jhfhlkjlj (fdsuufijjejejejej-deactivatedaccount) wrote :

As relayed through the kernel bug report, both 3.2 and 3.4rc upstream are affected. It doesn't appear to be a kernel bug, so unless someone can confirm it to be something other that xorg, I'm incompleting it.

bugbot (bugbot)
tags: added: resume
tags: added: black-screen
Revision history for this message
david.barbion (david-barbion) wrote :

I just updated to latest stable kernel release (3.2.0-24.38) and I still have the problem.
I then started back on oneiric kernel (3.0.0-19.33) and I did not have the problem anymore.

On recent kernel, every time I close/open the laptop lid, X is frozen. I need to reboot or restart lightdm from console.

I'm running precise on an HP 6730b with an "Intel Corporation Mobile 4 Series Chipset Integrated Graphics Controller (rev 07)" which is a GM45.

In the Xorg.0.log I can find these errors every time I have the problem:
[ 145.274] (WW) intel(0): flip queue failed: Invalid argument
[ 145.274] (WW) intel(0): Page flip failed: Invalid argument

Revision history for this message
Alessandro Paolino (mark2dp) wrote :

Intel integrated HD3000 here on samsung rf 511 s02 (disabled nvidia with bumblebee), I tried to reproduce this problem but seems random.
Ubuntu 12.04 final, stable kernel

Revision history for this message
Ryan McClure (mcc-mcc3d) wrote :

Seeing this and worse on a Lenovo ThinkPad T61 with nVidia Quadro 140M.

Anytime the lid is closed (whether this prompts suspend or not, whether this locks the screen or not), re-opening the lid presents the desktop view as it existed prior to closing the lid (if screen lock is not enabled), with a cursor that moves, but with no click response whatsoever.

Revision history for this message
Ryan McClure (mcc-mcc3d) wrote :

To add further clarification:

- I experience the effects described in the initial report *any time* the lid is closed, with 100% reproducible results. Suspend, screen lock, etc. have no extra effect. Close the lid, X is completely locked up other than the mouse cursor.
- I only started experiencing these effects after upgrading from 11.10 to 12.04.
- I can switch to terminal mode (Ctrl Alt F1) and interact there.
- Switching back to X (Ctrl Alt F7) returns me to the locked-up screen. Mouse still moves, but no interaction is possible.
- If the setting for automatically locking the desktop when the screen turns off is enabled, I experience the black screen with mouse cursor result. If this setting is not enabled, I see a "frozen" version of whatever was last on screen.

Revision history for this message
Tomasz Melcer (liori) wrote :

Same effect on Thinkpad X61s (Intel Graphics Media Accelerator X3100). Blank black screen with just mouse pointer visible after closing and opening the lid back (no suspend necessary), happens maybe once in 20 times. Input works—i can move mouse pointer, also cursor changes when over a text field (or I guess it's a text field, it's in the middle of screen, where a password dialog should be).

Revision history for this message
Ryan McClure (mcc-mcc3d) wrote :

Managed to fix my issue by switching to the nVidia proprietary driver. I had a look at my Xorg.0.log and noticed that *none* of the drivers seemed to be activating successfully. It looks like Nouveau eventually kicked in, even after showing a failed to load message, but I'm not sure what the implications of that are.

Went to Additional Drivers and saw two proprietary nVidia drivers there, neither of which were activated. Selected the top driver, listed as recommended, and now am no longer locking up when I close/open the screen.

Revision history for this message
sanmiguel9 (againsttcpa84) wrote :

I am wondering if the new Intel drivers (2.19; in Ubuntu 12.04 are 2.17) would improve the situation?

The PPA from the Ubuntu-X team ( https://launchpad.net/~ubuntu-x-swat/+archive/x-updates ) contains xserver-xorg-video-intel 2:2.19.0-0ubuntu1~xup1

Maybe someone wants to give it a shot... (but please be sure you know how to purge a PPA from the terminal in case sth goes wrong ;-) )

Revision history for this message
Andy Clayton (q3aiml) wrote : Re: [Bug 966744] Re: resume from suspend leaves me with black screen and a cursor
Download full text (4.1 KiB)

Be warned that PPA purge didn't work with multiarch last I checked

-----Original message-----
From: sanmiguel9 <email address hidden>
To: <email address hidden>
Sent: Tue, May 8, 2012 11:16:12 CDT
Subject: [Bug 966744] Re: resume from suspend leaves me with black screen and a cursor

I am wondering if the new Intel drivers (2.19; in Ubuntu 12.04 are 2.17)
would improve the situation?

The PPA from the Ubuntu-X team (
https://launchpad.net/~ubuntu-x-swat/+archive/x-updates ) contains
xserver-xorg-video-intel 2:2.19.0-0ubuntu1~xup1

Maybe someone wants to give it a shot... (but please be sure you know
how to purge a PPA from the terminal in case sth goes wrong ;-) )

--
You received this bug notification because you are subscribed to the bug
report.
https://bugs.launchpad.net/bugs/966744

Title:
  resume from suspend leaves me with black screen and a cursor

Status in The Linux Kernel:
  Incomplete
Status in “linux” package in Ubuntu:
  Incomplete
Status in “xserver-xorg-video-intel” package in Ubuntu:
  Triaged

Bug description:
  This happens intermittently. Sometimes upon resume, all I see is a
  black screen and the cursor. the mouse and keyboard respond (the mouse
  moves) but nothing changes. Switching to console and back doesn't fix
  it. Killing Compiz also does not fix the issue. Further, if one has
  password lock disabled on resume the user will be stuck with a frozen
  display of whatever was last shown on the screen before suspend. The
  mouse cursor still changes as you hover over various elements but the
  display is frozen solid. Killing the X session is the only workaround.

  BTW, I haven nothing connected to the laptop when I resume and before
  suspend. This is a very annoying issue because the only way to
  recover is to reboot, which means all the open files will be lost.

  ProblemType: Bug
  DistroRelease: Ubuntu 12.04
  Package: unity 5.8.0-0ubuntu2
  ProcVersionSignature: Ubuntu 3.2.0-20.33-generic 3.2.12
  Uname: Linux 3.2.0-20-generic x86_64
  .tmp.unity.support.test.0:

  ApportVersion: 1.95-0ubuntu1
  Architecture: amd64
  CompizPlugins: [core,composite,opengl,compiztoolbox,decor,vpswitch,snap,mousepoll,resize,place,move,wall,grid,regex,imgpng,session,gnomecompat,animation,fade,unitymtgrabhandles,workarounds,scale,expo,ezoom,unityshell]
  CompositorRunning: compiz
  Date: Tue Mar 27 21:47:25 2012
  DistUpgraded: Fresh install
  DistroCodename: precise
  DistroVariant: ubuntu
  InstallationMedia: Ubuntu 12.04 LTS "Precise Pangolin" - Alpha amd64 (20120124)
  MachineType: LENOVO 32492HU
  ProcKernelCmdLine: BOOT_IMAGE=/boot/vmlinuz-3.2.0-20-generic root=UUID=8afc43b9-4675-46b0-8bf8-49bfd422c2c7 ro quiet splash i915.i915_enable_rc6=1 vt.handoff=7
  SourcePackage: unity
  UpgradeStatus: No upgrade log present (probably fresh install)
  dmi.bios.date: 06/29/2011
  dmi.bios.vendor: LENOVO
  dmi.bios.version: 6QET67WW (1.37 )
  dmi.board.name: 32492HU
  dmi.board.vendor: LENOVO
  dmi.board.version: Not Available
  dmi.chassis.asset.tag: No Asset Information
  dmi.chassis.type: 10
  dmi.chassis.vendor: LENOVO
  dmi.chassis.version: Not Available
  dmi.modalias: dmi:bvnLENOVO:bvr6QET67WW(1.37):bd06/29/2011:svnLENOVO:pn3249...

Read more...

Revision history for this message
Marc Tommasi (marc-tommasi) wrote : Re: resume from suspend leaves me with black screen and a cursor

I've exactly the same comments than #30. I've tried with kernel 3.4 :
uname -a
Linux fitis 3.4.0-999-generic #201204260409 SMP Thu Apr 26 08:17:40 UTC 2012 i686 i686 i386 GNU/Linux. The only warning I can see in Xorg log is
[ 18337.968] (WW) intel(0): flip queue failed: Invalid argument
[ 18337.968] (WW) intel(0): Page flip failed: Invalid argument

BTW it seems to me that the restore of backlight state fails. Is it related to that ?
http://us.generation-nt.com/answer/3-4-rc1-regression-intel-resume-suspend-disk-hangs-restoring-backlight-state-help-206990502.html

I've not tried such a patch.

Revision history for this message
david.barbion (david-barbion) wrote :

As asked in comment #36, I've installed the package xserver-xorg-video-intel_2.19.0-0ubuntu1~xup1_i386.deb and did *not* solve this issue: got a black screen lock still with working mouse after 4 lid close/open.
However, the error message in X log disappeared.

Revision history for this message
Alessandro Paolino (mark2dp) wrote :

I've noticed that on my pc (uses Intel HD3000) it occurs more often if using dual monitor, so if I close the lid with the VGA cable inserted is more likely to happen than without.

Revision history for this message
Sparhawk (sparhawkthesecond) wrote :

I see the same thing on my Dell XSP 17 L702X. Also, comment #10 works for me to fix it temporarily. However, there's a typo in the command. Killing compiz is correct, with "killall -KILL compiz" (I only needed to do it once), but you should restart it with "compiz --replace --display :0" (N.B. the double dashes).

Revision history for this message
Stéphane Guillou (stephane-guillou) wrote :

I had experienced this problem before without anything attached to my netbook Samsung N310 (no charger, no screen): a black screen with the cursor.
I just had a problem related I guess, but with different conditions: external monitor, charger plugged in. When I reopened my netbook and turned my monitor back on, I was left with the last visual, pointer responding to links (changing from pointer to hand for example) but image completely frozen.
I could go back to the login screen and log back in thanks to Alt Gr + Syst + K.

Revision history for this message
siziwi (siziwi) wrote :

This bug also affects me, at a random basis. I'm using a Thinkpad W500 with the Intel 4500MHD graphic card.

Revision history for this message
jbus (jbus) wrote :

This bug also affects my Lenovo laptop with Intel 4500MHD chipset. This occurs randomly.

Revision history for this message
jhfhlkjlj (fdsuufijjejejejej-deactivatedaccount) wrote :

I'm bumping this up to critical as this is affecting a) a large portion of users and b) "Crashes the entire operating system"/"renders the entire desktop . . . unusable"

Changed in xserver-xorg-video-intel (Ubuntu):
importance: High → Critical
Revision history for this message
neopegasus (s-martes3) wrote :

Hi i am having like the same problem i have a Hp pavilion g6 whit ati display and AMD prosessor, the thing is i dont get the cursor on my screen my screen is totaal black.

i am puthing my pc info attach to the comment

i hope this more info will help

regards Neopegasus

description: updated
description: updated
Revision history for this message
jhfhlkjlj (fdsuufijjejejejej-deactivatedaccount) wrote :

Hi, all!

To see if this issue remains in later versions of xserver-xorg-video-intel, I've updated the driver via the X updates PPA, which is found here:

https://launchpad.net/~ubuntu-x-swat/+archive/x-updates

It would be nice to see if this driver fixes the issue. I'm going to test in the next few days to see if it does fix the issue.

Please note that I urge those that are not familiar with Xorg and the breakage that can occur to please stay away from the above link. Only those that can dig themselves out of a possible hole should test if they want.

summary: - resume from suspend leaves me with black screen and a cursor
+ Resume from suspend leaves me with black screen or a screen of the
+ desktop before it suspended (though the mouse still moves/changes
+ cursor)
Revision history for this message
jhfhlkjlj (fdsuufijjejejejej-deactivatedaccount) wrote : Re: Resume from suspend leaves me with black screen or a screen of the desktop before it suspended (though the mouse still moves/changes cursor)

Well, forget that. Got the freeze with the updated X drivers.

Revision history for this message
jbus (jbus) wrote :

I installed the later version of xserver-xorg-video-intel that is in the PPA yesterday and I have not had a freeze as of yet. I've been making sure to close my laptop lid because that's when it usually happens... When the lid is closed and it's left for a while. I'll update if I get another freeze, but for now it looks like the freezing has stopped with the updated driver.

This is the output from lspci | grep VGA:

00:02.0 VGA compatible controller: Intel Corporation Mobile 4 Series Chipset Integrated Graphics Controller (rev 07)

Revision history for this message
Nick Rhodes (ngrhodes) wrote :

added the x-swat ppa and I am still getting xorg freezes.

Revision history for this message
cento (cento-autistici) wrote :

but the bus is about xorg, gnome-screensaver, unity, intel drivers... what else?

This only happens with ubuntu. Debian and Arch (tested in these days) works fine.

:(

Revision history for this message
Roman Friesen (krokosjablik) wrote :

This problem (initially reported as 995013) occurs also while using Unity 2D.
xserver-xorg-video-intel 2:2.17.0-1ubuntu4

Revision history for this message
cento (cento-autistici) wrote :

also with gnome-shell instead of unity?

Revision history for this message
jbus (jbus) wrote :

A quick update on post #49

I finally did get a freeze with the more up to date Intel driver in the PPA. Though, it did take a lot longer to freeze than had been occurring with the driver in the 12.04 repository. Another difference is that instead of just a black screen with a cursor, I got a screen with my wallpaper and the cursor. Hopefully this will help in determining what the problem is.

Revision history for this message
In , Brettcornwall-1 (brettcornwall-1) wrote :

This problem occured after upgrading from Ubuntu 11.10 (2:2.15.901-1ubuntu2) to Ubuntu 12.04 (2:2.17.0-1ubuntu4). We have attempted to update to 2.19 but it still occurs.

Sometimes upon resume, all I see is a black screen and the cursor. the mouse and keyboard respond (the mouse moves) but nothing changes. Switching to console and back doesn't fix it. Killing Compiz also does not fix the issue. Further, if one has password lock disabled on resume the user will be stuck with a frozen display of whatever was last shown on the desktop before suspend. The mouse cursor still changes as you hover over various elements but the display is frozen solid. One can switch to another TTY via ctrl+alt+F1 but killing the X session is the only way to get back to the desktop.

This is reproducible in metacity, not just compiz. Trying the 3.4 kernel also did not fix the issue. All of those affected are Intel-based machines.

BTW, I haven nothing connected to the laptop when I resume and before suspend. This is a very annoying issue because the only way to recover is to kill the X session, which means all the open files will be lost.

The launchpad page in which this is documented can be found here: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/966744

Thank you for your time. Please let me know how it can be debugged further (I understand that this is likely not enough information)

Revision history for this message
In , Chris Wilson (ickle) wrote :

This sounds like the Xserver damage bug. First install all the driver packages from ppa:xorg-edgers. Then try and reproduce and attach a drm.debug=6 dmesg and Xorg.log from across the suspend and resume.

Revision history for this message
Roman Friesen (krokosjablik) wrote :

@cento (cento-autistici): I have just changed to "Unity 2D" in the login screen and have used it for a while. I don't know, "Unity2D" uses gnome-shell or not.

Changed in xserver-xorg-video-intel:
importance: Unknown → High
status: Unknown → Confirmed
Revision history for this message
In , Brettcornwall-1 (brettcornwall-1) wrote :

Created attachment 61993
Xorg.0.log

I got the bug to occur after upgrading and enabling debugging. Here is my dmesg output and xorg log.

Revision history for this message
In , Brettcornwall-1 (brettcornwall-1) wrote :

Created attachment 61994
dmesg with drm.debug=6

Revision history for this message
In , Chris Wilson (ickle) wrote :

We encounter one warning there that is suspicious:

[ 304.920] (WW) intel(0): flip queue failed: Invalid argument
[ 304.920] (WW) intel(0): Page flip failed: Invalid argument

and nothing that corresponds with it in the dmesg. To hit EINVAL suggests that the display was disabled, which given that it is an LVDS panel is quite, quite bizarre. There is a patch in xf86-video-intel.git (with --enable-sna) that should behave better in such circumstances. Can you try building from git with SNA enabled and see if the bug reoccurs?

Changed in linux (Ubuntu):
status: Incomplete → Triaged
Revision history for this message
In , Chris Wilson (ickle) wrote :

ppa:xorg-edgers has an updated version of UXA that should help with one of the issues you encountered across resume. How does it fare under your testing?

Revision history for this message
In , Brettcornwall-1 (brettcornwall-1) wrote :

Apologies with my reticence. I have been away from a machine in which I can test on. I shall test with the PPA once more.

Revision history for this message
jhfhlkjlj (fdsuufijjejejejej-deactivatedaccount) wrote : Re: Resume from suspend leaves me with black screen or a screen of the desktop before it suspended (though the mouse still moves/changes cursor)

Hi, all. Launchpad hasn't imported some of the other comments from freedesktop: A few days ago there was a commit to git that may have fixed the issue:

"ppa:xorg-edgers has an updated version of UXA that should help with one of the
issues you encountered across resume. How does it fare under your testing?"

I've been testing and it seems to work. Again, it would be helpful to confirm that it does fix it. Would any brave souls be willing to install xorg edgers? Beware, this time around I ended up with some seriously messed-up graphics making real production on that laptop to be pretty impossible. But I've been able to suspend/resume.

Revision history for this message
cento (cento-autistici) wrote :

i'm trying it

Changed in xserver-xorg-video-intel:
status: Confirmed → New
Revision history for this message
cento (cento-autistici) wrote : Re: Resume from suspend leaves me with black screen or a screen of the desktop before it suspended (though the mouse still moves/changes cursor)

tested. don't work for me.

Revision history for this message
Murz (murznn) wrote :

also confirm that updating drivers from ppa:xorg-edgers didn't help.

Revision history for this message
jhfhlkjlj (fdsuufijjejejejej-deactivatedaccount) wrote :

What a bummer, I thought it had helped for me.

Revision history for this message
In , Brettcornwall-1 (brettcornwall-1) wrote :

While I seemed to have success, there are two other people on the bug report that have reported the issue still exists.

Revision history for this message
In , Brettcornwall-1 (brettcornwall-1) wrote :

Annoyingly, after those few days of testing, I, too had the session freeze up on me again. So three people confirmed that it didn't work :)

Revision history for this message
In , Chris Wilson (ickle) wrote :

Don't forget to attach debugging info from the recent freeze so that we can be sure that you are still hitting the same issue every time.

Revision history for this message
David MacBride (drm-redmenace) wrote :
Download full text (3.4 KiB)

I think I've reproduced this same error, though the scenario is slightly different:

System Info
- Lenovo thinkpad X200 notebook
- Intel GM45 Express graphics
- Ubuntu 12.04 32bit (x86-32). up-to-date as of June 2 / 2012

Steps to reproduce
- set setting "when the lid is closed"="do nothing" in Power settings
- set "Lock" to "On" in Brightness and Lock settings
- check "Require my password when waking from suspend" in Brightness and Lock settings
- close the lid
- open the lid

expected results: I see the log in dialog

actual result: backlit but dark screen, mouse cursor still moves as normal, mouse cursor changes as I move it around the screen as if it is responding to diff't UI elements. Occasionally the display is different: instead of blank/black screen other windows are visible, such as the dock and the top panel... but when that happens I cannot interact with them.

The issue is intermittent. Cannot be reproduced all the time.

Snippet from Xorg.0.log:
[204557.220] (II) intel(0): EDID vendor "LEN", prod id 16401
[204557.220] (II) intel(0): Printing DDC gathered Modelines:
[204557.220] (II) intel(0): Modeline "1280x800"x0.0 69.30 1280 1304 1336 1408 800 804 808 820 +hsync +vsync (49.2 kHz)
[204557.220] (II) intel(0): Modeline "1280x800"x0.0 57.79 1280 1304 1336 1408 800 804 808 820 +hsync +vsync (41.0 kHz)
[204558.022] (WW) intel(0): flip queue failed: Invalid argument
[204558.022] (WW) intel(0): Page flip failed: Invalid argument
[204568.948] (II) intel(0): EDID vendor "LEN", prod id 16401
[204568.948] (II) intel(0): Printing DDC gathered Modelines:
[204568.948] (II) intel(0): Modeline "1280x800"x0.0 69.30 1280 1304 1336 1408 800 804 808 820 +hsync +vsync (49.2 kHz)
[204568.948] (II) intel(0): Modeline "1280x800"x0.0 57.79 1280 1304 1336 1408 800 804 808 820 +hsync +vsync (41.0 kHz)
[204569.789] (II) intel(0): Allocated new frame buffer 1920x1080 stride 7680, tiled
[204599.362] (II) PM Event received: Capability Changed
[204599.362] I830PMEvent: Capability change
[232228.461] (II) intel(0): EDID vendor "LEN", prod id 16401
[232228.461] (II) intel(0): Printing DDC gathered Modelines:
[232228.461] (II) intel(0): Modeline "1280x800"x0.0 69.30 1280 1304 1336 1408 800 804 808 820 +hsync +vsync (49.2 kHz)
[232228.461] (II) intel(0): Modeline "1280x800"x0.0 57.79 1280 1304 1336 1408 800 804 808 820 +hsync +vsync (41.0 kHz)
[232229.278] (II) intel(0): Allocated new frame buffer 1280x800 stride 5120, tiled
[232235.012] (II) PM Event received: Capability Changed
[232235.012] I830PMEvent: Capability change
[232240.008] (II) PM Event received: Capability Changed
[232240.009] I830PMEvent: Capability change
[232240.092] (II) intel(0): EDID vendor "LEN", prod id 16401
[232240.093] (II) intel(0): Printing DDC gathered Modelines:
[232240.093] (II) intel(0): Modeline "1280x800"x0.0 69.30 1280 1304 1336 1408 800 804 808 820 +hsync +vsync (49.2 kHz)
[232240.093] (II) intel(0): Modeline "1280x800"x0.0 57.79 1280 1304 1336 1408 800 804 808 820 +hsync +vsync (41.0 kHz)
[232267.256] (II) AIGLX: Suspending AIGLX clients for VT switch

Gut feeling: the windows that are supposed to be drawn when the computer is locked eg:...

Read more...

Revision history for this message
jtdintulsa (james-dickson) wrote :

Same problem here .. I have a Dell Vostro V13 ..

00:02.0 VGA compatible controller: Intel Corporation Mobile 4 Series Chipset Integrated Graphics Controller (rev 07)
00:02.1 Display controller: Intel Corporation Mobile 4 Series Chipset Integrated Graphics Controller (rev 07)

Closing the lid and letting it sit for 5 minutes resumes fine. Closing the lid and letting it sit for longer I get the black screen with the cursor.

Revision history for this message
Marcelo Bello (marcelo-bello) wrote :

Every time I get this error (about once a week or less) dmesg has the following message: "*ERROR* invalid framebuffer id".
I have been careful to check if the framabuffer error happens at the same instant I see the bug manifesting itself and it does.

Lenovo X220 with Intel graphics.

I have the laptop to do nothing when lid closes on AC power, sleep when lid closes on battery. But it was already happening when the computer was set to always sleep when lid is closed (I changed exactly to avoid hitting this bug too often).

My guess is that this bug is being triggered by some sort of sequence of lid opening and closing and AC power going on and off. Also, I have the bug happening twice by closing and instantly reopening the lid (but still, I can't find a way to reproduce the problem).

I am not seeing any correlation between time computer has been left with lid closed (whether it is sleeping of not) and chance of this bug manifesting itself (except for the instant close-then-open case I mentioned above).

Of course these are anecdotal observations.

Revision history for this message
In , Brettcornwall-1 (brettcornwall-1) wrote :

Finally, Finally got it to reproduce again under the PPA:

test@test-Aspire-5734Z:~$ apt-cache policy xserver-xorg-video-intel
xserver-xorg-video-intel:
  Installed: 2:2.19.0+git20120530.cf5b3e2e-0ubuntu0sarvatt~precise
  Candidate: 2:2.19.0+git20120530.cf5b3e2e-0ubuntu0sarvatt~precise
  Version table:
 *** 2:2.19.0+git20120530.cf5b3e2e-0ubuntu0sarvatt~precise 0
        500 http://ppa.launchpad.net/xorg-edgers/ppa/ubuntu/ precise/main amd64

I'll attach dmesg and xorg.0.log again.

Revision history for this message
In , Brettcornwall-1 (brettcornwall-1) wrote :

Created attachment 62482
dmesg-drm.debug=6 with git version

Revision history for this message
In , Brettcornwall-1 (brettcornwall-1) wrote :

Created attachment 62483
xorg.0.log with git version

Revision history for this message
Alexander Neumann (alex-neumann) wrote :

I also do not observe any correlation between the time the lid is closed and occurrence of the bug. What I find suspicious is that most users experiencing the bug are using a Thinkpad of the X series (X200, X200s, X220), in my case it's a Lenovo x200s.

Revision history for this message
Sparhawk (sparhawkthesecond) wrote :

There are two of us using Dells, but it's plausibly an unrelated bug.

Revision history for this message
Alexander Neumann (alex-neumann) wrote :

I don't think it's exclusive to Thinkpads but it seems to appear most frequently there. E.g. in this askubuntu thread we also have mostly Thinkpads from the X series http://askubuntu.com/questions/124399/laptop-screen-goes-blank-after-waking-from-sleep

Revision history for this message
neopegasus (s-martes3) wrote :

I am usin a H.P. pavilion G6 and a dell isperion. And have the same problem H.P. whit AMD prossesor and ati grafick and the dell intel

Revision history for this message
drplix (pjr-1060) wrote :

I also have this bug - same set up and symptoms as #70. I'm on Toshiba Satellite Pro L630 - Intel chipset/video.

Revision history for this message
Andrés Aguila (afaguila) wrote :

I have a samsung with the intel graphic driver and the same problem.

In a side note, for a week now I've been going into suspend mode not by closing the lid, but with the "start" menu and then click on suspend. I've not had this problem ever since. I have not updated anything or changed any "conf" files. I have no idea if that helps others, but it might give some ideas on what the problem is (before, going with the lid, it happened 1 out of 3 times or so).

Revision history for this message
Fernando Carolo (carolo) wrote :

I am also getting these symptoms on a Dell XPS 14z with Intel graphics, so this specific bug or a similar one does affect other brands, not only the Thinkpad X series.

Revision history for this message
jhfhlkjlj (fdsuufijjejejejej-deactivatedaccount) wrote :

Hi, all. This has been confirmed as an intel graphics bug, not specific to certain manufactures. If you are experiencing this bug, please mark "me, too" at the top of this page rather than commenting (as lots of people are subscribed to this report).

Thanks!

Revision history for this message
neopegasus (s-martes3) wrote :

Hi, chauncellor i have a question , if this bug is confirmed as an Intel graphics bug , is a bug for AMD graphics? Because as I say before I do not have a Intel graphics but still I have this same problem.

Thanks in advance

Revision history for this message
neopegasus (s-martes3) wrote :

Hi, chauncellor i have a question , if this bug is confirmed as an Intel graphics bug , is a bug for AMD graphics? So I can report it ther? Because as I say before I do not have a Intel graphics but still I have this same problem.

Thanks in advance

Revision history for this message
jhfhlkjlj (fdsuufijjejejejej-deactivatedaccount) wrote :

neopegasus, it may be the same bug, merely that intel chips are much more affected. If this bug report gets fixed and your issue is not resolved, you'll need to report a new bug. For the meantime, it would be good to follow this bug and see if the problem is fixed with this bug.

For now we're waiting for the reply of an X developer over on the development report.

Revision history for this message
Michael Zehrer (zehrer) wrote :

It's solved on my TP X220 with the intel drivers from the PPA.

Revision history for this message
jhfhlkjlj (fdsuufijjejejejej-deactivatedaccount) wrote :

Michael, I had deemed it solved, too, but when two other people said it was still present I decided to keep trying and eventually got it. It certainly is harder for me to get it but it wasn't gone for me. Oddly enough, the logs I did capture with the PPA did not include the previously harmful-sounding:

[ 304.920] (WW) intel(0): flip queue failed: Invalid argument
[ 304.920] (WW) intel(0): Page flip failed: Invalid argument

though the issue remained the same.

Revision history for this message
Tekkaman_Slade (tekkamanslade) wrote :

I'm having the same issue running Precise on my X61s with gma965. Like afaguila, the problem doesn't occur when I spend via the power menu. Fun thing is that if I set power management to do nothing when the lid is closed, the problem occurs when closing and opening the lid without sleep. This leads me to believe it my be an issue related to the lid-close event.

Thoughts?

Revision history for this message
Sparhawk (sparhawkthesecond) wrote :

FWIW I never close the lid to activate suspend, but I still see this bug regularly. I suspend either through timing out, or through a keyboard shortcut bound to:
dbus-send --print-reply --system --dest=org.freedesktop.UPower /org/freedesktop/UPower org.freedesktop.UPower.Suspend

Revision history for this message
joe williams (joetify) wrote :

FWIW I have this issue and use a X200s.

Revision history for this message
loke (developer-loke) wrote : Re: Resume from suspend leaves me with black screen or a screen of the desktop before it suspended (though the mouse still moves/changes cursor)

The problem seems to be related to compiz. After I switch to a terminal, and killall -9 compiz, and switch back to X, I can see a partially functional desktop, because I can't switch between applications, and window frames are missing. It's not recoverable even after restarting compiz with DISPLAY=:0 compiz &.

Restarting lightdm definitely fixes the issue.

Revision history for this message
Tekkaman_Slade (tekkamanslade) wrote :

I have the same issue using a Ubuntu 2D session, so I don't think the problem is related to compiz.

Has anyone confirmed that kernel 3.4 fixes the issue as mentioned in post #19 of the thread linked below.

https://bugs.launchpad.net/ubuntu/+source/linux/+bug/989674?comments=all

Revision history for this message
Philipp Dreimann (philipp-dreimann-deactivatedaccount) wrote : Re: [Bug 966744] Re: Resume from suspend leaves me with black screen or a screen of the desktop before it suspended (though the mouse still moves/changes cursor)

On 12 June 2012 14:37, Josephus Beale <email address hidden> wrote:
> Has anyone confirmed that kernel 3.4 fixes the issue as mentioned in
> post #19 of the thread linked below.
>
> https://bugs.launchpad.net/ubuntu/+source/linux/+bug/989674?comments=all

It does not fix the issue for me.

Revision history for this message
Sparhawk (sparhawkthesecond) wrote : Re: Resume from suspend leaves me with black screen or a screen of the desktop before it suspended (though the mouse still moves/changes cursor)

@loke: I'm also not sure that this implies that compiz is to blame. If my system is functioning normally, restarting compiz never works for me. I have to restart unity instead (i.e. $ unity --replace ), which presumably restarts compiz itself.

Revision history for this message
Tekkaman_Slade (tekkamanslade) wrote :

I've been running kernel ppa 3.4 from mainline for about 2 days now and it seems to have fixed the problem. I have yet to encounter a black screen or any other issue with resume since installing this kernel. Hopefully, this kernel version will make its way to release via an update.

Revision history for this message
Richard Kemp (richardtkemp) wrote :

To add some info here: it seems that a common theme for those suffering from this bug is that they use laptops with Intel graphics chips. I am also experiencing this bug with symptoms exactly as described here, except I am using a desktop PC with an nVidia graphics card. (Dell Vostro 410, nVidia 9800GT)

I am using nVidia's beta 302.07 drivers for proper function of my 1-rotated 1-normal dual-screen setup, I can't comment on other drivers as I only recently put Ubuntu on this machine and didn't suspend before switching to these drivers. Next time it happens I will grab the logs.

TLDR; nVidia 9800GT, Ubuntu 12.04, Unity 3D, Compiz, experiencing the bug only intermittently.

Revision history for this message
Pablo Huerta Jr. (gpdpob) wrote :

I am also experiencing this but I never use suspend. I have had this problem with both lock on and off in Brightness settings. I am also on intel.

I hope this helps others. I have found a way to restore my session to save any open programs or just to close session cleanly. Once the problem happens I go to a tty(Crtl+Alt+F1), log in and kill compiz and compiz-decorator. Then I run metacity, return to X(Crtl+Alt+F7) and start closing programs. I then opened a terminal (Crlt+Alt+T) and launch gnome-panel which gives me a menu to log out.

I do not know if metacity and gnome-panel ship stock or you have to install them. I never installed gnome 3 but it appears as a session option on my login screen.

Revision history for this message
Sparhawk (sparhawkthesecond) wrote :

@Pablo: Does running metacity have any advantage over just restarting compiz, as per comment #10? (N.B. the typos as per comment #41.)

Revision history for this message
Pablo Huerta Jr. (gpdpob) wrote :

@Sparhawk: I'll need to try the command in comment #41 when it breaks again. I tried just compiz(no arguments) after killing it but all it did was give a wallpaper but nothing else.

Revision history for this message
Tekkaman_Slade (tekkamanslade) wrote :

@philipp: After 3 days of testing, looks like kernel 3.4 didn't fix it for me after all.

I'm using a power management script to switch VTs on resume as a workaround until this is official fixed.

Revision history for this message
siziwi (siziwi) wrote :

As stated above, this bug affected me as well. However, since I've moved to Gnome Shell (instead of Unity), I haven't experienced a single problem (and I've used the suspend/resume quite often). Even more interesting is the fact that I experienced the bug again when using Unity (just tried out a few moments ago, logged in using Unity and tried suspend/resume).
I basically don't know anything about the mechanism behind it and therefore I don't know why it's thought to be a kernel or X problem. From my experience I would say it's related to Unity. Hopefully, this information helps!

Revision history for this message
In , Brettcornwall-1 (brettcornwall-1) wrote :

Has this been sufficient information?

Revision history for this message
Sparhawk (sparhawkthesecond) wrote :

@Pablo: The commands in comment #41 allow you to access applications, but you are left without a window manager, so it's a temporary fix to save data before rebooting. Does metacity leave a workable environment?

@Josephus: Could you please elaborate regarding your workaround? It sounds like it might be useful? Cheers.

Revision history for this message
Tekkaman_Slade (tekkamanslade) wrote :

@Sparkhawk: I used the script mentioned in post #20 in the thread linked below. Unfortunely, it only reduced how often the problem occurs for me and doesn't completely eliminate it.

https://bugs.launchpad.net/ubuntu/+source/linux/+bug/989674?comments=all

@siziwi: You are on to something. I made the switch to gnome 3 shell and it completely eliminated the issue. I am now convinced that the problem is related to unity in some way. I like unity, but gnome shell isn't a bad alternative at all.

"sudo apt-get install gnome-shell" and selecting a gnome session on login is all you need.

I can now use my notebook like it was intended to be used. I hope this issue is fixed soon.

Revision history for this message
Sparhawk (sparhawkthesecond) wrote :

@tekkamanslade (did you just change usernames? :-)
Thanks for that. For others following, that script essentially emulates switching to tty1 and back again. i.e. Ctrl + Alt + F1 and Ctrl + Alt + F7.

Revision history for this message
Bruno Henrique (brunofx7) wrote :

Exactly the same issue here, but video is ATI opensource drivers.

Revision history for this message
Jens (jujehe) wrote :

Hi, I have the same problem on Thinkpad X61s with Intel GM965/GL960.
It also happens under gnome 3 and plugging in an external screen or unplugging caused the same problem.

Revision history for this message
Jens (jujehe) wrote :

Just to clarify, I get the black screen from time to time when the computer is waking up from suspend, but after playing around with a second monitor a bit I was able to cause a black screen with only the mouse pointer visible quite often, when unplugging and plugging back in the second screen and changing the resolution of the display.

dmesg gives me:

[26638.780723] Xorg[18591]: segfault at c0 ip 00007f35bd063558 sp 00007ffffa4f4e30 error 4 in Xorg[7f35bd00e000+1f0000]
[29666.992265] Xorg[25256]: segfault at c0 ip 00007ff44c185558 sp 00007fffdffe66b0 error 4 in Xorg[7ff44c130000+1f0000]
[29712.217686] Xorg[26589]: segfault at c0 ip 00007f43b60c9558 sp 00007fffb0e16ef0 error 4 in Xorg[7f43b6074000+1f0000]
[29860.445754] Xorg[27260]: segfault at c0 ip 00007fa487687558 sp 00007fffa9070d30 error 4 in Xorg[7fa487632000+1f0000]
[29884.673410] Xorg[27981]: segfault at c0 ip 00007f18d473d558 sp 00007fff45a09670 error 4 in Xorg[7f18d46e8000+1f0000]

it might be a different issue, but maybe it helps here too.

Revision history for this message
Anmar Oueja (anmar) wrote :

Now I can reproduce the matter 100% of the time. I have a Thinkpad X201 with Intel graphics of course.

I am running the latest and greatest packages as of June 19 1:00 UTC including the latest 3.2-26 kernel . I get this in the log file.

*ERROR* Invalid framebuffer id.

This message is from the kernel.

At 109 messages, I assume there is enought information for somebody to attempt a fix. As you can imagine, this has a major impact on this issue. I hear GNOME Shell doesn't have that problem. I might end up switching to it just so I can use my machine.

Revision history for this message
Timo Aaltonen (tjaalton) wrote :

Please try with a mainline kernel:

http://kernel.ubuntu.com/~kernel-ppa/mainline/

You can start from 3.4.3, then if that one works 'bisect' backwards to see which kernel fixed it. If 3.4.3 is broken, then try 3.5-rc3. Let us know how it goes.

Bruno, this bug is intel specific.

Changed in xserver-xorg-video-intel (Ubuntu):
status: Triaged → Incomplete
Revision history for this message
Anmar Oueja (anmar) wrote : Re: [Bug 966744] Re: Resume from suspend leaves me with black screen or a screen of the desktop before it suspended (though the mouse still moves/changes cursor)

On 19 June 2012 02:01, Timo Aaltonen <email address hidden> wrote:
> Please try with a mainline kernel:
>
> http://kernel.ubuntu.com/~kernel-ppa/mainline/

3.4 is the last one tagged for precise. Are you saying I can use the
Quantal packages for that because I did and graphics acceleration was
broken when I installed 3.5-rc3 on Precise.

Revision history for this message
Timo Aaltonen (tjaalton) wrote : Re: Resume from suspend leaves me with black screen or a screen of the desktop before it suspended (though the mouse still moves/changes cursor)

quantal images are fine, as long as you remember to install the -extra package as well, which includes the drm drivers..

Revision history for this message
Symbol (asarpeshkar) wrote :

I see this bug with stock 12.04 and Unity on a Thinkpad T400 (Intel graphics), but I don't see it in GNOME 3.

Revision history for this message
Minosone (minosone) wrote :

I'm experiencing this bug too. Intel Graphics here too. However i'm running an Acer 1810TZ laptop, so no Thinkpad.

I have attached the output of the following commands:

sudo lshw -C display > xbuginfo.txt;
lsb_release -a >> xbuginfo.txt;
uname -a >> xbuginfo.txt

Let me know if there are any logs I can attach that might help too. Hope this bug gets fixed soon, because it's rather annoying to loose my work each time my laptop suspends.

Thanks!

Revision history for this message
Anmar Oueja (anmar) wrote : Re: [Bug 966744] Re: Resume from suspend leaves me with black screen or a screen of the desktop before it suspended (though the mouse still moves/changes cursor)

I tried using 3.5-rc3 and so far resume issue seem OK. Will report back in
a weeks time.

Revision history for this message
thehansman (thehansman) wrote : Re: Resume from suspend leaves me with black screen or a screen of the desktop before it suspended (though the mouse still moves/changes cursor)

I didn't experience the bug anymore, after switching to Kubuntu.
Before I was using Ubuntu with Unity and had the same problems with resuming quite frequently. I didn't change anything else in my system, except from moving to KDE.

Revision history for this message
Maarten Lankhorst (mlankhorst) wrote :

For those on precise (not quantal), could you test the xorg-server package in ppa:ubuntu-x-swat/ppa ? X dying on resume seems to look similar and at least 1 person had a dmesg captured of a Xorg segfault here.

Revision history for this message
Minosone (minosone) wrote :

I installed the package 'xserver-xorg-video-intel' in the ppa:ubuntu-x-swat/ppa .
It's now at version 2.19.0-0ubuntu1~xup1 .
Will leave it running for a couple of days and keep you posted.

Revision history for this message
Maarten Lankhorst (mlankhorst) wrote :

Erm that's not from that repository, only the xorg-server package there is useful. 2.19.0-0ubuntu1~xup1 is not in it.

Revision history for this message
Adam Logghe (alogghe) wrote :

This does not appear to be Intel graphics only.

I experience this problem with a 2011 imac and a thinkpad x120e both have AMD graphics, both proprietary drivers (I will test the opensource stack though).

Currently both desktops use Gnome fallback.

Revision history for this message
joe williams (joetify) wrote :

The packages from ppa:ubuntu-x-swat dont seem to correct the problem for me on my x200s.

Revision history for this message
Carlos Moffat (carlos-eldiabloenlosdetalles) wrote :

I've been using the 3.5 rc3 kernel and the intel drivers from x-swat, and I had another "dark resume" today.

Revision history for this message
Carlos Moffat (carlos-eldiabloenlosdetalles) wrote :

Just to add, #123 refers to a Thinkpad X220 (Intel graphics)

Revision history for this message
[mrs_sheep] (mrs-sheep) wrote :

Seems totally arbitrary to me, but at least for me it worked:

I read at AskUbuntu http://askubuntu.com/a/133537/71437 about this being related to the USC. In fact since removing it I did not have a single "dark resume" and I used to have it everytime. As said, this seems completely arbitrary.

[Ubuntu 12.04, MSI FX700 with Integrated Intel Graphics (i5) and disabled (acpi_call) GeForce GT 425M]

Revision history for this message
Heiko (heiko-dudzus) wrote :

I have something similar as C.Schöll said, (#125). I have this problem when update-manager prompted for updates before and was waiting for user interaction

Revision history for this message
bagl0312 (bagl0312) wrote :

Hello,
I was affected by the a similar bug on a dell XPS M1330 since the upgrade to 12.04 (with nvidia graphics and nouveau drivers).
In particular closing the lid with "Do nothing" "when lid is closed" in the power settings was always leaving me with a unusable desktop session after opening the lid (i.e. only the mouse pointer was moving).

I did a sequence of things that solved the problem for me:
first I resumed my session a couple of times using the commands described on post #41,

then in the power settings I first selected "Suspend" as action when lid is closed, then I put back "Do nothing".

After this the problem disappeared "magically"

Sorry for not beiing more precise, but probably it can help somebody else with a similar problem

Revision history for this message
Timo Aaltonen (tjaalton) wrote :

Hmm, I'm actually having this same bug on both of my laptops, and I don't think it's a driver bug but compiz or the xserver. For me, when I kill compiz remotely, the screen flashes with the desktop contents and then when compiz is respawned it leaves me a screen with the background and functional mouse (the cursor changes as it passes windows that are "invisible"). This of course after blindly opening the screensaver :)

Revision history for this message
Timo Aaltonen (tjaalton) wrote :

oh and what's weird is that I get the desktop back (without unity / window-manager) when I kill indicator-printers-service after killing compiz... Running unity --replace restores the broken state.

Revision history for this message
Timo Aaltonen (tjaalton) wrote :

Running metacity once compiz is killed works fine.

Changed in compiz (Ubuntu):
importance: Undecided → High
status: New → Triaged
Revision history for this message
jhfhlkjlj (fdsuufijjejejejej-deactivatedaccount) wrote :

Indeed, it does appear that compiz can be killed and metacity replaced after some wiggling around a TTY once it hits. I am currently running metacity's window decorator. And, indeed, running compiz --replace once I can get the desktop usable again will bring it back to a 'frozen' state with the cursor changing as necessary for window management again. It also spews out error messages that says something like "invalid configuration, this should never happen, you should probably raise a bug report about it".

After an additional metacity --replace, however, I was able to smudge compiz back in there, although I could not get Unity back up and running (was complaining about unknown exit status of the other program).

As some people also had this issue with gnome shell, I feel that it may be bad interaction between xorg and drivers for compositing (but what do I know). I was getting invalid page flips from my dmesg when debugging in the xorg bug report (the developer just kind of disappeared and hasn't replied).

I don't mean to be a prude to the compiz devs (I know they're very busy) But for the meantime I am going to bump up importance to 'critical' as the bug has 'a severe impact on a large portion of Ubuntu users' and 'crashes the entire operating system'. They'll decide better than I once they look at it, though (for all I know, it still isn't in compiz).

Changed in compiz (Ubuntu):
importance: High → Critical
Changed in compiz-core:
status: New → Confirmed
description: updated
Revision history for this message
ethanay (ethan-y-us) wrote :

the "invalid frame buffer" dmesg report got me here...

same symptoms on intel graphics hardware (3 laptops), but *unable* to get a workable desktop without reboot. TTY 1-6 work OK, but TTY7 is stuck on blank screen, mouse cursor only. compiz --replace, metacity --replace etc don't work and all report various error messages when attempted.

i would call this a critical bug -- it renders the system completely unoperable. my parents know enough to "hold down the power button to force the computer to turn off."

Revision history for this message
Marcelo Ruiz (marcelo-javier-ruiz-yahoo) wrote :

I don't know if this is useful, but I am experiencing the blank screen + moving cursor + high CPU ussage + locked keyboard problem.
I am running 12.04 AMD64 with propietary Nvidia graphic drivers.
Suspend and Resume work FINE if I use Unity 2D or Gnome Classic (no effects).
It does show the problem with Unity, Gnome Shell, Gnome Classic, and Cinnamon. In all of these the var/usr/pm-suspend.log shows the resume from suspend worked. I think the problem is in Compiz (at least in my case).
My computer used to resume and suspend with no problem with Maverick.

Revision history for this message
Bruno Santos (bsantos) wrote :

I have this with a 9800 GT and driver 302.17 from x-updates. The last thing I have on the Xorg log is this:

[ 51230.115] (WW) NVIDIA(0): WAIT (2, 6, 0x8000, 0x0000ec68, 0x0000ecf0)
[ 51237.115] (WW) NVIDIA(0): WAIT (1, 6, 0x8000, 0x0000ec68, 0x0000ecf0)
[ 51719.354] [mi] EQ overflowing. Additional events will be discarded until existing events are processed.
[ 51719.360]
Backtrace:
[ 51719.540] 0: /usr/bin/X (xorg_backtrace+0x26) [0x7f7cb38319c6]
[ 51719.540] 1: /usr/bin/X (mieqEnqueue+0x263) [0x7f7cb38120a3]
[ 51719.540] 2: /usr/bin/X (0x7f7cb36a9000+0x62ab4) [0x7f7cb370bab4]
[ 51719.540] 3: /usr/lib/xorg/modules/input/evdev_drv.so (0x7f7cab06b000+0x5d88) [0x7f7cab070d88]
[ 51719.540] 4: /usr/bin/X (0x7f7cb36a9000+0x8afb7) [0x7f7cb3733fb7]
[ 51719.540] 5: /usr/bin/X (0x7f7cb36a9000+0xb0efa) [0x7f7cb3759efa]
[ 51719.540] 6: /lib/x86_64-linux-gnu/libpthread.so.0 (0x7f7cb29cf000+0xfcb0) [0x7f7cb29decb0]
[ 51719.540] 7: /usr/lib/x86_64-linux-gnu/xorg/extra-modules/nvidia_drv.so (0x7f7cac5aa000+0x87900) [0x7f7cac631900]
[ 51719.540] 8: /usr/lib/x86_64-linux-gnu/xorg/extra-modules/nvidia_drv.so (0x7f7cac5aa000+0x89235) [0x7f7cac633235]
[ 51719.540] 9: /usr/lib/x86_64-linux-gnu/xorg/extra-modules/nvidia_drv.so (0x7f7cac5aa000+0x10052b) [0x7f7cac6aa52b]
[ 51719.540] 10: /usr/lib/x86_64-linux-gnu/xorg/extra-modules/nvidia_drv.so (0x7f7cac5aa000+0x4d9130) [0x7f7caca83130]
[ 51719.540] 11: /usr/lib/x86_64-linux-gnu/xorg/extra-modules/nvidia_drv.so (0x7f7cac5aa000+0x4d7bb5) [0x7f7caca81bb5]
[ 51719.540] 12: /usr/lib/x86_64-linux-gnu/xorg/extra-modules/nvidia_drv.so (0x7f7cac5aa000+0x4d936d) [0x7f7caca8336d]
[ 51719.540] 13: /usr/bin/X (0x7f7cb36a9000+0x119b1c) [0x7f7cb37c2b1c]
[ 51719.540] 14: /usr/bin/X (0x7f7cb36a9000+0x4b54e) [0x7f7cb36f454e]
[ 51719.540] 15: /usr/bin/X (0x7f7cb36a9000+0x4e8b1) [0x7f7cb36f78b1]
[ 51719.540] 16: /usr/bin/X (0x7f7cb36a9000+0x3d6da) [0x7f7cb36e66da]
[ 51719.540] 17: /lib/x86_64-linux-gnu/libc.so.6 (__libc_start_main+0xed) [0x7f7cb186476d]
[ 51719.540] 18: /usr/bin/X (0x7f7cb36a9000+0x3d9d1) [0x7f7cb36e69d1]
[ 51719.540] [mi] These backtraces from mieqEnqueue may point to a culprit higher up the stack.
[ 51719.540] [mi] mieq is *NOT* the cause. It is a victim.

Revision history for this message
Nils Herde (hernil) wrote :

I have experienced this far less after I read something about this occuring when Software center or Update manager was requiering user input and thus made sure none of them were open when suspending.

But this is clearly not resolved and I think it's about time someone told us that this is being worked on, and exactly what logs and other things we can provide to help fix this critical bug! And even though it seems that it's a combination of packages that triggers this bug someone should take responsability and not just leave it up to "one of the other teams".

It's something that not only should be fixed in 12.10 but also backported so that the LTS work as it should.

Revision history for this message
cuc (cuc+) wrote :

happens on 2 sandy bridge laptops 8 times out of 10.
never happend on intel gma950 netbook

Revision history for this message
Joonas Saarinen (jza) wrote :

To me it happened also with a GMA950 netbook (LG X110).

Revision history for this message
idsfa (idsfa) wrote :

Both my IdeaPad U310 and my wife's U260 (two different intel chipsets) experience this randomly upon awakening from suspend.

I can confirm that switching to a VT and restarting lightdm gets me back to the login. I will look into the compiz/metacity shuffle next time it happens.

My kernel logs contain the "[drm:drm_mode_getfb] *ERROR* invalid framebuffer id" error message on resume. There are no errors logged in pm-suspend.

I also see the lightdm/x-0.log error message abou mieq, but cannot confirm that is contemporaneous.

Revision history for this message
Marcelo Ruiz (marcelo-javier-ruiz-yahoo) wrote :

For those of you with Nvidia cards having this problem: if you feel like giving it a try I solved it by doing two things, but I am not sure if the combination of both gave me the solution or just the last one:

I first installed Michael Kuhn ppa and upgraded the system:
sudo apt-add-repository -y ppa:suraia/ppa
sudo apt-get update && sudo apt-get -y dist-upgrade

Finally I installed xorg-edgers fresh X crack ppa to use the nvidia beta drivers:
sudo apt-add-repository - y ppa:xorg-edgers/ppa
sudo apt-get update && sudo apt-get install -y nvidia-current

I then disabled this last ppa from software sources to prevent updates to the driver that is finally working as it should: the random red screen bug and freeze at shutdown/reboot and the black screen with the moving cursor problem are gone (I tried suspending the computer with tons of programs opened for hours and it worked as it used to with Maverick).

Note that this will leave your computer running the 3.5.0-4-generic kernel.
YMMV

Toshiba Qosmio X500 Core i7 + Nvidia 360GTM

Revision history for this message
Melanie Roha (melanie-roha) wrote :

 I worked around this issue in 12.04 switching to Gnome. Hardly ideal. Please tell me if I can provide more information to help find a fix.

Sony Vaio SR390 with "Intel Corporation Mobile 4 Series Chipset"

Marc D. (marc.d)
security vulnerability: no → yes
security vulnerability: yes → no
Revision history for this message
Janne Moren (jan-moren-gmail) wrote :

This has started to happen to me from about a week agi. I did not have this problem once before, now I can't ever suspend and resume properly. Likely an update that caused it.

Ubuntu 12.40 64 bit, Lenovo T430, Intel graphics. Kernel 3.4.0 (to get around a separate random freeze issue).

Revision history for this message
Pellaeon Lin (pellaeon) wrote :

Hi, I am also having similiar problem on Acer TimelineX 4830tg, which has:

Intel Sandy Bridge graphics (HD3000)
Nvidia Geforce GT540M

But for me not only resuming from suspend will trigger this issue, simply lock the screen would.

Locking the screen would leave me my original desktop but a darker color (like a black area with 0.5 opacity over the screen).

Resuming from suspend would sometimes leave me a totally black screen with cursor functioning just as others described, but sometimes also a original desktop with darker color.

I use Unity3D on 3.2.0-26-generic kernel (the official repo version)

xserver-xorg-video-intel version is 2:2.19.0-0ubuntu1~xup1 (from x-swat PPA)

Here is a similiar bug: https://bugs.launchpad.net/ubuntu/+source/gnome-screensaver/+bug/996770

Revision history for this message
hypercity (graphiccityflorin) wrote :

started about a week ago after updates , never hapened before except that for me it happens 100% of the times on lid close and when i try to return to the desktop after a screen blank for a period of inactivity

Display controller: Intel Corporation Mobile 945GM/GMS/GME, 943/940GML Express Integrated Graphics Controller (rev 03)
PackardBell Easynote-MX45

Revision history for this message
Ubuntu QA Website (ubuntuqa) wrote :

This bug has been reported on the Ubuntu Package testing tracker.

A list of all reports related to this bug can be found here:
http://packages.qa.ubuntu.com/qatracker/reports/bugs/966744

tags: added: package-qa-testing
Revision history for this message
gbz (andyrising) wrote :

I had the same problem after some X server updates in Ubuntu 12.04. The screen went black ( in fact in white color) after I reopened the screen lid.

This is how I solved it. In synaptic, I can see the version of xserver-xorg-core is 2:1.11.4-0ubuntu10.6. I choose "Force version" and downgrade it to 2:1.11.4-0ubuntu10. Then there is no problem anymore. (BTW, I use the latest verion of compiz and there is no problem).

I am using Thinkpad W520 with latest nvidia-current driver installed.

Revision history for this message
jhfhlkjlj (fdsuufijjejejejej-deactivatedaccount) wrote :

gbz, I've had no such luck in that area. It's very likely that you're having a different (but similar) issue. White screen freezes are not uncommon on those damnedable nvidia binary drivers.

Revision history for this message
gbz (andyrising) wrote :

Update: After I downgraded xserver-xorg-core from 2:1.11.4-0ubuntu10.6 to 2:1.11.4-0ubuntu10, with latest compiz (1:0.9.7.8-0ubuntu1.2) and nvidia-current (302.17-0ubuntu1~precise~xup1), at first this problem disappeared (as I reported in #145), but after about 1 hour this problem comes again (and I didn't fiugre out why).

So I also downgrade compiz to 1:0.9.7.8-0ubuntu1, and nvidia-current to 295.40, both of which are the versions in a fresh installed Ubuntu 12.04. In the last few hours, I tested for serveral times, and there is no such black screen problem after resume from suspend.

Revision history for this message
gbz (andyrising) wrote :

Chauncellor, yes I just realized my problem may be different from that in this bug :P
Maybe someone could use the downgrading method to test whether it could solve this problem.

Revision history for this message
idsfa (idsfa) wrote :

Looking to improve my 3D frame rate, I installed the xorg-edgers PPA. The kernel in that PPA causes a complete hang on resume from suspend. However, when I boot from the stock "precise" kernel, resume works properly and I have not been able to recreate this bug with that package set (precise + xorg-edgers - 3.5.0 kernel)

Unfortunate side effect: A known bug in xorg-edgers cause me to have no icons in the Unity sidebar or window switcher. Since the hover text still works, I consider this an acceptable loss.

Revision history for this message
Richard (w0h0i6a95-2ic6ajw-p7k1qaion) wrote :

Tried upgrading to 3.2.22-030222-generic, but bug is still present.

[699311.573] (WW) intel(0): flip queue failed: Invalid argument
[699311.600] (WW) intel(0): Page flip failed: Invalid argument

I have a cron job that runs, twice a day, switching CCTV monitor on/off.

/usr/bin/xset -display :0.0 dpms force off
/usr/bin/xset -display :0.0 dpms force on

This bug occurs after a few days. Rapid issuing of dpms on/off does not appear to trigger it faster.

Ubuntu 12.04 LTS
Hardware P8H61-M LE/USB3

00:02.0 VGA compatible controller: Intel Corporation 2nd Generation Core Processor Family Integrated Graphics Controller (rev 09) (prog-if 00 [VGA controller])
        Subsystem: ASUSTeK Computer Inc. Device 844d
        Flags: bus master, fast devsel, latency 0, IRQ 52
        Memory at fe000000 (64-bit, non-prefetchable) [size=4M]
        Memory at c0000000 (64-bit, prefetchable) [size=256M]
        I/O ports at f000 [size=64]
        Expansion ROM at <unassigned> [disabled]
        Capabilities: [90] MSI: Enable+ Count=1/1 Maskable- 64bit-
        Capabilities: [d0] Power Management version 2
        Capabilities: [a4] PCI Advanced Features
        Kernel driver in use: i915
        Kernel modules: i915

xserver-xorg-video-intel 2:2.17.0-1ubuntu4 X.Org X server -- Intel i8xx, i9xx display driver

Any useful tests I can do, or upgrades I should try?

Revision history for this message
Richard (w0h0i6a95-2ic6ajw-p7k1qaion) wrote :

Addition information, I just check log and times, the report happened on the /usr/bin/xset -display :0.0 dpms force off not the force on. Hope that helps in fixing bug.

Revision history for this message
idsfa (idsfa) wrote :

Sadly, just experienced this again, even with the xorg-edgers driver. Neither the "mieq" nor the "invalid framebuffer id" errors were in my log files this time, so those appear to be red herrings.

I was unsuccessful in doing the metacity/compiz shuffle to recover and had to restart the lightdm service to regain functionality.

affects: compiz-core → compiz
Revision history for this message
jhfhlkjlj (fdsuufijjejejejej-deactivatedaccount) wrote :

Hi all,

it would be handy to switch to a TTY and analyse the compiz (and other) process carefully. Check to see if it's using much more memory than before hanging, or CPU.

Most important would be to attach a debugger and see where compiz is hung (if at all):

sudo gdb /usr/bin/compiz `pidof compiz`
(gdb) bt

I'll give that a go if I get the time soon enough.

Revision history for this message
Daniel van Vugt (vanvugt) wrote :

Hi all. If compiz is hung then a stack trace of the compiz process would be very helpful. To do this...

0. Download and keep a copy of the attached dstack script handy in your home directory.
Then when the hang occurs;
1. Switch to a VT (Ctrl+Alt+F1) and log in.
2. Note how much memory the compiz process is using. Is it more than usual?
2. sudo ./dstack compiz > compizhang.txt
3. Kill/restart/recover any way you like.
4. Attach compizhang.txt to this bug.

Changed in compiz:
importance: Undecided → High
milestone: none → 0.9.8.0
Revision history for this message
Daniel van Vugt (vanvugt) wrote :

To see the aforementioned "dstack" script you will have to click on the "#154" link.

Revision history for this message
jhfhlkjlj (fdsuufijjejejejej-deactivatedaccount) wrote :

Here you go, Daniel. Got it on the first try :).

Anyway, Compiz report 1.4 Gigs of virtual memory - nothing out of the ordinary for me. I got slightly distracted and forgot to check real memory usage. I can check up on that again. Anyway, here's a stacktrace - it's looking pretty unhelpful....

affects: xserver-xorg-video-intel (Ubuntu) → mesa (Ubuntu)
Changed in mesa (Ubuntu):
status: Incomplete → New
Revision history for this message
Daniel van Vugt (vanvugt) wrote :

Thanks Chauncellor, the stack trace is very good actually. It shows compiz is blocked in the intel driver:
intel_update_renderbuffers () from /usr/lib/x86_64-linux-gnu/dri/i965_dri.so

This is probably the fault of the intel graphics driver, unless...

1. compiz is passing it invalid data. Looks highly unlikely; or
2. compiz is spinning at 100% CPU overwhelming the driver. Is it?

Revision history for this message
Daniel van Vugt (vanvugt) wrote :

Chauncellor - If you run dstack multiple times while compiz is unresponsive, does the resulting output change much?

Revision history for this message
Éric Piel (pieleric) wrote :

I also got a hang. It's very rare on this laptop (every couple of months). It's 32 bits. Attached is the backtrace. As you can see, it's very similar to the 64 bits one.

Compiz is not looping. It's just blocked there. The process doesn't use the cpu and I took a second backtrace a few second later, and it was exactly identical.

Revision history for this message
Daniel van Vugt (vanvugt) wrote :

Thanks Éric. That matches Chauncellor's results.

I wonder if the intel driver gets stuck and confused if you suspend while it's in the middle of a system call...

Revision history for this message
Éric Piel (pieleric) wrote :

Just a note, this backtrace is not taken after suspend, but just closing the lid, monitors go to powersave mode, and opening the lid.

Indeed, it could be due to having this dri call and a call to switch off the monitors at the same time.

Revision history for this message
jhfhlkjlj (fdsuufijjejejejej-deactivatedaccount) wrote :

Daniel, I haven't gotten it to freeze again (strangley, it's like the chances of it happening are greater if you leave the laptop sitting for long periods of time) but have you seen the dmesg reports over on the bugzilla Intel report? Would they be of any help?

Revision history for this message
Launchpad Janitor (janitor) wrote :

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

Changed in mesa (Ubuntu):
status: New → Confirmed
Revision history for this message
Daniel van Vugt (vanvugt) wrote :

Chauncellor, I know of the logs on bugzilla, but they're only helpful to Intel driver/kernel developers.

Revision history for this message
jhfhlkjlj (fdsuufijjejejejej-deactivatedaccount) wrote :

Daniel, I got multiple freeze/stacktraces - they're all exactly the same as the one I uploaded. Compiz also wasn't using any notable cpu consumption and was taking up 100 Mb of memory, which is not any different than normal, really. So it just *dies*.

Also, I'm not sure if you ever saw this in the slew of comments on this report, but if I kill compiz/compiz-decorator and replace with metacity, then shove a 'compiz --replace' onto tty7 via tty1 then the same exact frozen screen will return. So would you indeed say this is still intel's fault? Some sort of triple buffering issue? (ahem, not very knowledgable about graphics stacks...)

Revision history for this message
franglais.125 (franglais.125-deactivatedaccount) wrote :

Hello everyone,

I have also been experiencing this bugs for quite some time now.

I just wanted to add to Eric Piel's comment, as I too experienced this bug in a different situation, i.e. not after suspending my laptop.

I have a Dell Vostro, with intel graphics.

I sometimes use "xset dpms force off" to turn off my screen without locking it nor suspending. Although I cannot reproduce the bug at will, this is what I did:

1) Run "xset dpms force off"
2) Wait 10 minutes until the screen locks automatically
3) Type something on the keyboard to wake up the computer
4) Result: Screen turns on (the laptop was NOT suspended), but it is not blank, I can see my desktop, but the mouse/keyboard can't interact with it. The mouse changes to a cursor in the center of the screen, where the space to input the password should be.

This is, hence, very similar in the result, as the mouse changes to a cursor, but the screen cannot be unlocked. (I did try to put my password to no avail).

Hope this helps to tackle this down. Sorry if it isn't too clear, let me know if I should explain again anything.
Cheers

Revision history for this message
dominik (schmidt-dominik) wrote :
Revision history for this message
Timo Aaltonen (tjaalton) wrote :

there is a new mesa release in precise-proposed, could you try if it fixes the bug?

Changed in mesa (Ubuntu):
status: Confirmed → Incomplete
Revision history for this message
jhfhlkjlj (fdsuufijjejejejej-deactivatedaccount) wrote :

Nope, still got the freeze.

Changed in mesa (Ubuntu):
status: Incomplete → Confirmed
Revision history for this message
ethanay (ethan-y-us) wrote :

the only method of those listed above to help me recover is to reboot or switch to another tty and

sudo killall lightdm

then

sudo lightdm

Revision history for this message
In , Brettcornwall-1 (brettcornwall-1) wrote :

There are some stacktraces of compiz in the launchpad report, which appears to be the package that fuses the powder keg, that may detail why the driver isn't being particularly kind to compiz.

Revision history for this message
kenorb (kenorb) wrote :

Some duplicates of this issue:
https://bugs.launchpad.net/ubuntu/+source/nvidia-graphics-drivers/+bug/982710
https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/966631
https://bugs.launchpad.net/ubuntu/+source/nvidia-graphics-drivers/+bug/921450
https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/1007904
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/988008
https://bugs.launchpad.net/ubuntu/+source/fglrx-installer/+bug/987153
because of the same backtrace.

It affect me as well.

Every 10-15 seconds it freezes, I can't use the computer in normal way.
My machine:
HP Compaq 8100 Elite
Linux kenorb-HP-Compaq-8100-Elite-SFF-PC 3.2.0-27-generic #43-Ubuntu SMP Fri Jul 6 14:25:57 UTC 2012 x86_64 x86_64 x86_64 GNU/Linux
Ubuntu 12.04 LTS

Revision history for this message
Lee (qq510371827) wrote :

Same for me, Lenovo ZHAOYANG E46A with Nvidia nvs 3100M(The laptop has an intel i3 cpu with a gpu integrated on it .But the intel integrated gpu has already been blocked by the laptop manufacturer).
I used arch,ubuntu 11.10,12.04 in the past and am currently using ubuntu 12.10 alpha3. I have used gnome3,xfce and unity as my desktop enviroment. All of the above DEs and OS distros have the same bug.

Revision history for this message
Eddie (system-w) wrote : Re: Resume from suspend leaves me with black screen or a screen of the desktop before it suspended (though the mouse still moves/changes cursor)

Guys, just found an admissible solution which is nice for me.

Watching over the processes I've found a lid.sh running on lid opening and I removed the file and did a reboot then closed the lid again and it started working like "Do Nothing" option.

Will dig around to examine the script and see what goes wrong, but deleting it is an acceptable walkthrough by now.

Location: /etc/acpi/lid.sh

Revision history for this message
GeorgeD (geodosch) wrote :

I tried disabling lid.sh and though it worked for me too, at least after a single test.

Rather than delete the file, I did a chmod -x /etc/acpi/lid.sh, which prevents it from being executed.

In looking at the script, it's checking for the files 'local/lid.sh.pre' and 'local/lid.sh.post' under the /etc/acpi directory. The script executes them if they exist, and skips them if they don't, so that may be normal. However, I wonder if it's a problem that the 'local' directory is missing.

Revision history for this message
Sparhawk (sparhawkthesecond) wrote :

For me, I sometimes hit this bug when not opening the lid to resume. That is, the laptop is open when suspended, and remains open on resume. So, in my case at least, I assume it's unrelated to lid.sh.

Revision history for this message
Henry Gomersall (hgomersall) wrote :

@Sparhawk, or it's entering the same fail state through a different mechanism (it's unlikely to be lid.sh per se), or it's a different bug.

Revision history for this message
NoBugs! (luke32j) wrote :

I tried logging in to "Ubuntu 2d", and for a long time it seemed this bug was gone. A bit after I switched back, it froze at resume again. This is on a Intel GM965/GL960 Integrated Graphics.

Revision history for this message
Éric Piel (pieleric) wrote :

lid.sh does quite a few things, but with Unity, without xscreensaver, and with an intel card, it basically does this when closing the screen:
xset dpms force off

and when opening the screen:
xset dpms force on

Revision history for this message
Sparhawk (sparhawkthesecond) wrote :

@Henry, could be. I figured that lid.sh gets triggered by opening/closing the lid, but perhaps I'm wrong?

Revision history for this message
Henry Gomersall (hgomersall) wrote :

@sparhawk, I'm sure that's the case. What I mean is that whatever it is that lid.sh does to trigger the bug might well be done by other operations as well.

Revision history for this message
Sparhawk (sparhawkthesecond) wrote :

@Henry, ah, I see. That is plausible too.

Also, FWIW, I often use "xset dpms force off", and have never seen this bug from that. (I do have xscreensaver though, and I've never tried "xset dpms force on").

Revision history for this message
David Perkinson (davidp-reed) wrote :

I believe http://ubuntuforums.org/showthread.php?p=11926504 solved the problem for me. (In fact, under 11.04, I fixed a similar problem by creating a file /etc/pm/sleep.d/20_chvt_mjf containing the following:

#!bin/sh

case "$1" in
  suspend) chvt 1; ;;
  resume) chvt 7; ;;
esac

return 0

but I have not tried that under 12.04.) I am running 12.04 on a thinkpad x200t.

Revision history for this message
Pablo Huerta Jr. (gpdpob) wrote :

I just wanted to let you all know that I have not had a problem with this since adding:

 acpi_osi=Linux acpi_backlight=vendor

to

GRUB_CMDLINE_LINUX_DEFAULT="quiet splash"

in /etc/default/grub

and running

update-grub

Revision history for this message
Eddie (system-w) wrote :

If run lid.sh manualy the script works well with its job, I see that there is something else involved with the lid interactions and causes the intel express graphics to fail.

Revision history for this message
idsfa (idsfa) wrote :

The most recent updates from the xorg-edgers PPA work with the kernel from the PPA and has now been a couple of weeks without displaying this bug.

Of course, the known Unity bug with this PPA (no icons on dock) is still unresolved, so this may not be a solution for everyone.

Revision history for this message
In , Brettcornwall-1 (brettcornwall-1) wrote :

One user is reporting that the recent updates to the xorg-edgers PPA has been fine with his machine for the last few weeks. Are there any special commits to pay attention to that may have resolved this issue?

Revision history for this message
Dražen Lučanin (kermit666) wrote :

Started happening to me recently, but only sometimes when I close the lid and suspend that way. Reminds me of the bug that still occurs on my old HP laptop that's caused by the GNOME screensaver.

Here is the report I managed to collect with apport when I entered this black-screen void state by switching to another tty and running ubuntu-bug.

Revision history for this message
Daniel van Vugt (vanvugt) wrote :

Duplicate bug 1035183 shows the same hang and call stack when opening the laptop lid... even when the system never suspended/resumed.

Revision history for this message
Benjamin Bach (benjaoming) wrote :

This is a hard nut to crack :D If someone needs a quick run-down of the most important findings, I think this will do....

OP as well as many others are talking about suspend/resume, and many point to intel drivers and compiz... BUT:

#33, #34 says it's related to closing the lid.

#52 and #94 says the problem occurs also in Unity 2D.

#76 says problem doesn't occur when using suspend button, rather than closing the lid.

#78 says it's an Intel bug (but many nvidia users in here also)

#84 says it's definately related to open/close of laptop lid.

#85 says the bug occurs without using the lid.

#99 never uses suspend and experiences it with BOTH lock on and off.

#108 says plugging and unplugging an external screen causes the same issue.

#127 says that "do nothing" upon closing the lid would always reproduce only until selecting different option and then back to "do nothing".

#142 says that simply locking the screen triggers the problem.

#143 says it happens simply when returning to a blank screen

Revision history for this message
Stefan Skotte (screemo) wrote :

I have an Asus UX32VD with nvidia/intel hd4000 graphics, and I have tried several kernels from 3.5.x -> 3.6-rc1 (both on Ubuntu 12.04 and 12.10), and everytime I get a black wallpaper after I open the lid after suspend to ram.

Just setting the wallpaper manually again will make it appear as usual, and no errors following that.

Is this some compiz redraw missing on the "wallpaper" section of the desktop ? - reason is that using a "Unity 2d" session I cannot replicate the problem (use case: close lid -> suspend to ram (wait a few minutes) -> open lid -> resumes).

I dont see any errors/warning in the logs (syslog, lightdm, Xorg.0.log etc.) .

Omer Akram (om26er)
Changed in compiz:
importance: High → Critical
Revision history for this message
Dac Chartrand (conner-bw) wrote :

This bug happens to me a dozen times a week, seemingly random.

I "fix it" (if by "fix it" I mean driving around in a junker with a piece of wood as the steering and two flat tires instead of, you know, having a decent working car) by doing the following:

CTRL+ALT+F2
[Login as myself]
sudo service lightdm stop
sudo service lightdm start

Revision history for this message
Dac Chartrand (conner-bw) wrote :

PS: I'm on an X220, (Intel® Sandybridge Mobile Graphics Card)

Revision history for this message
benmhall (ubuntu-linuxgruven) wrote :

Dac, I have a Lenovo ThinkPad X200s with the same problem. My workaround was a bit less drastic: I've moved to Unity 2D. It isn't quite as slick to use, but it looks and behaves almost the same and I haven't experienced the problem in four days of heavy use. (Suspend and resume ~20 times a day.)

The worst thing to me about Unity 2D is that metacity won't alt+tab all windows across all desktops. The best solution I've found is to install superswitcher, which is far less than ideal.

I keep trying everyone's suggested workarounds, newer Intel drivers, etc. Nothing works so far other than to avoid using standard Unity and switch to Unity 2D.

Ben

Revision history for this message
Corey O. (7pde) wrote :

I am having the same problem on both my new asus zenbook prime UX32VD (intel 4000) and my old custom-built desktop (well-supported nvidia chipset). The problem exists in both gnome-classic and gnome-shell, so it is not gnome/unity/compiz specific. This is a pretty major bug as it sometimes causes me to lose my work. I should mention that both setups usually use dual monitors. I'm not sure if that makes any difference.

btw, I think that bug #996770 might be a duplicate of this problem

Revision history for this message
Daniel van Vugt (vanvugt) wrote :

Hi all.

We need to be focussed and specific or else bugs like this never get resolved.

Since all the stacks I have seen so far have been identical hangs in i965_dri.so, I have now clarified that this is an Intel graphics driver bug. Although a workaround in compiz might still be possible.

If you're not using Intel graphics, please log a new bug. Or if you have a bug that was made a duplicate of this one and you're not using intel graphics, please change the duplicate status of your bug to not point to this one any more.

summary: - Resume from suspend leaves me with black screen or a screen of the
- desktop before it suspended (though the mouse still moves/changes
- cursor)
+ [i965] Resume from suspend leaves me with black screen or a screen of
+ the desktop before it suspended. Compiz hung in
+ intel_update_renderbuffers() from intel_prepare_render() from
+ brw_draw_prims()
Changed in xserver-xorg-video-intel:
status: New → Confirmed
tags: added: blocks-hwcert-enablement
Revision history for this message
Gabriel (gabriel-nobody) wrote :

same problem on Lenovo laptop with a i915 driver.

Revision history for this message
Zdenek Dlauhy (e-tast-k) wrote :

also same problem here, intel, Asus laptop U36SD, managed to get to console...

Revision history for this message
Pedro Nunes (nunes-p89) wrote :

Same Problem with an Asus 1000H

When I come back from suspend state, all i see is a black screen and the cursor, that i can move the cursor, but cant do anything.
I am currently using ubuntu 12.04LTS kernel: 3.2.0.29

This bug happens random

Revision history for this message
In , idsfa (idsfa) wrote :

Latest xorg-edgers update: The newest batch fixes the Unity issue that kept icons from being displayed. I have now not seen the black screen on resume bug in at least three rounds of updates.

Bad news: the latest kernel in -edgers (3.5.0-10-generic) doesn't work with encrypted home directories, had to boot with the previous one (3.5.0-9-generic).

Changed in compiz:
milestone: 0.9.8.0 → 0.9.8.1
Revision history for this message
In , Chris Wilson (ickle) wrote :

The cause of the EINVAL is an attempt to pageflip with the pipe disabled due to DPMS off. This should be fixed by:

commit c4eb5528a456b65c673f7c984d14a622ac67cdca
Author: Chris Wilson <email address hidden>
Date: Tue Jun 5 16:04:16 2012 +0100

    uxa: Check for DPMS off before scheduling a WAIT_ON_EVENT

    Regression from commit 3f3bde4f0c72f6f31aae322bcdc20b95eade6631
    Author: Chris Wilson <email address hidden>
    Date: Thu May 24 11:58:46 2012 +0100

        uxa: Only consider an output valid if the kernel reports it attached

    When backporting from SNA, a key difference that UXA does not track DPMS
    state in its enabled flag and that a DPMS off CRTC is still bound to the
    fb. So we do need to rescan the outputs and check that we have a
    connector enabled *and* the pipe is running prior to emitting a scanline
    wait.

    References: https://bugs.freedesktop.org/show_bug.cgi?id=50668
    Signed-off-by: Chris Wilson <email address hidden>

Changed in xserver-xorg-video-intel:
status: Confirmed → Fix Released
Revision history for this message
benmhall (ubuntu-linuxgruven) wrote :

Just to clarify, is this bug fixed in a current/scheduled-for-update xserver-xorg-video-intel package, or is it only fixed upstream? I'm currently using xorg-edgers packages to try to avoid the problem but would like to revert to standard packages if a fix has been released to stock Ubuntu 12.04.

My apologies in advance if the answer to this question should be obvious to me somewhere on the launchpad page.

Timo Aaltonen (tjaalton)
no longer affects: linux (Ubuntu)
Revision history for this message
Edward Donovan (edward.donovan) wrote :

Hi Ben, fair question. It's just upstream, and it's hard to tell.

I think this automated message:

   Bug Watch Updater (bug-watch-updater) 3 hours ago
      Changed in xserver-xorg-video-intel:
      status: Confirmed → Fix Released

might be aided by having the word upstream in there somewhere. Or some other way for people to see, without having to puzzle, that 'Fix Released' isn't referring to what's available from their distro. 'Cause those two words sure grab your attention. :)

And, Bugzilla just provides a status of 'Fixed', while Launchpad has 'Fix Committed' and 'Fix Released'. And the bugzilla installations I've seen, like Gnome, fd.o, and the kernel, mark things Fixed upon commit. LP translates Fixed to Fix Released, and I've been confused by this in the past. Maybe I can look into that more, some time. (And I sure don't mean to knock LP. Upstream tracking, and distinguishing committed from released, are both great features.)

While I'm here I'd like to thank Chauncellor for all his work here, and in filing the upstream report. Thanks!

Revision history for this message
Kevin Shenk (batonac) wrote :

I'd also love to know whether we can expect to see this fixed in stock 12.04. I've been holding off several precise installs and upgrades until this is fixed.

Revision history for this message
Timo Aaltonen (tjaalton) wrote :

launchpad refuses me to change the status of the bug..

anyway, I'll add the two commits to -intel and put it on a PPA for testing.

Revision history for this message
Timo Aaltonen (tjaalton) wrote :

you can grab the test package from my ppa:

https://launchpad.net/~tjaalton/+archive/ppa

Changed in mesa (Ubuntu Precise):
assignee: nobody → Timo Aaltonen (tjaalton)
Timo Aaltonen (tjaalton)
Changed in mesa (Ubuntu Precise):
importance: Undecided → Critical
status: New → Triaged
Changed in mesa (Ubuntu):
status: Confirmed → Fix Released
affects: mesa (Ubuntu) → xserver-xorg-video-intel (Ubuntu)
Changed in xserver-xorg-video-intel (Ubuntu Precise):
status: Triaged → Incomplete
Revision history for this message
Timo Aaltonen (tjaalton) wrote :

the ppa builders seem backlogged, so amd64 binary and source also here:

http://kernel.ubuntu.com/~tjaalton/lp966744/

Revision history for this message
Timo Aaltonen (tjaalton) wrote :

-1ubuntu4.1.1 was not enough, so I added another backported patch (from fdo bug 45413), so test -1ubuntu4.1.2 instead. Feels better here.. 20 suspend/resume cycles without issues where the previous one hung after 13.

Revision history for this message
benmhall (ubuntu-linuxgruven) wrote :

Thank you so much, Timo. This has certainly been an annoying bug.

I have installed your -1ubuntu4.1.2 package and will report back if I have any further problems.

Thank you also to Edward for providing a good explanation as to what "Fix Released" means in this context.

It's great to have backed away from the xorg-edgers packages, as I was seeing all sorts of display artifacts in Unity 3D on my ThinkPad X200s. Using Timo's packages, all is well again, and so far so good WRT hanging.

Revision history for this message
jhfhlkjlj (fdsuufijjejejejej-deactivatedaccount) wrote :

I'm sad to say that I just hit the bug with the above PPA's -intel package.

Revision history for this message
Lawrence Patrick Calulo (lugkhast) wrote :

I can confirm, I've also hit the bug even after installing the -1ubuntu4.1.2 version.

Revision history for this message
Andy Clayton (q3aiml) wrote : Re: [Bug 966744] Re: [i965] Resume from suspend leaves me with black screen or a screen of the desktop before it suspended. Compiz hung in intel_update_renderbuffers() from intel_prepare_render() from brw_draw_prims()

Has anyone confirmed that the backtrace after the patches is unchanged?

Revision history for this message
Launchpad Janitor (janitor) wrote :

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

Changed in compiz (Ubuntu Precise):
status: New → Confirmed
Revision history for this message
Lawrence Patrick Calulo (lugkhast) wrote :

The stack trace given by the dstack script is identical, with only the PIDs and parts of the addresses differing.

Revision history for this message
jhfhlkjlj (fdsuufijjejejejej-deactivatedaccount) wrote :

so if the compiz stack is identical, what of the intel report? Should we get another drm.debug=6 output, compare, and bother upstream again?

Revision history for this message
In , Brettcornwall-1 (brettcornwall-1) wrote :

Hi, Chris - I'm afraid we've had a few confirmations that applying these fixes didn't fix the issue. The same problem remains unabated.

Would you like yet another dmesg output with the patches applied?

Changed in xserver-xorg-video-intel:
status: Fix Released → Confirmed
Changed in xserver-xorg-video-intel (Ubuntu Precise):
status: Incomplete → Confirmed
Revision history for this message
In , Chris Wilson (ickle) wrote :

Yes, I think we need a fresh set of debug logs with all the known fixes applied.

Revision history for this message
In , Brettcornwall-1 (brettcornwall-1) wrote :

Created attachment 66517
drm.debug=6 After patches applied

Here you go.

Revision history for this message
In , Brettcornwall-1 (brettcornwall-1) wrote :

Created attachment 66518
Xorg.0.log After patches applied

Changed in oem-priority:
importance: Undecided → Medium
Revision history for this message
Simon Greenwood (sfgreenwood-gmail) wrote :

Happening on System 76 Lemur (lemu3, early 2012 version). It has only started happening since the last kernel update (3.2.0-29).

Revision history for this message
Chris Halse Rogers (raof) wrote :

Invalidating the Compiz tasks; it's not compiz' fault that the Intel driver never completes glSwapBuffers.

Changed in compiz (Ubuntu Precise):
status: Confirmed → Invalid
Changed in compiz (Ubuntu Quantal):
status: Triaged → Invalid
Revision history for this message
Timo Aaltonen (tjaalton) wrote :

reopening for quantal, not fixed yet :/

Changed in xserver-xorg-video-intel (Ubuntu Quantal):
status: Fix Released → Confirmed
tags: added: rls-q-incoming
tags: removed: rls-q-incoming
Changed in xserver-xorg-video-intel (Ubuntu Quantal):
milestone: none → ubuntu-12.10-beta-2
Revision history for this message
Timo Aaltonen (tjaalton) wrote :

if you can reproduce this (hard for me it seems), please install xtrace and run

xtrace compiz --replace > trace.txt

and suspend/resume with anger, then after a hang change to a VT and run

tail -n1000 trace.txt > trace-tail.txt

and attach it here.

Revision history for this message
Timo Aaltonen (tjaalton) wrote :

note that the trace file will grow _very_ quickly, so unless you have reproduced it after ~10 cycles, start over.

Revision history for this message
Lawrence Patrick Calulo (lugkhast) wrote :

I got a trace with a hang - this is with -1ubuntu4.1.2 from the PPA.

Revision history for this message
Lawrence Patrick Calulo (lugkhast) wrote :

After looking at it a bit I'm not 100% sure that the previous log I posted is correct - the modification time is 20 minutes before the time of the hang. Here's another one, this time checking that xtrace was indeed running and that the modification time was only seconds before the hang.

Revision history for this message
Mark Edgington (edgimar) wrote :

Not sure if it helps, but I get the same issue using Ubuntu 12.10 with kernel 3.5.0-14-generic -- I am using this on netbook with the gma500_gfx driver (Dell Mini 12). When resuming from suspend, I also get the black screen with mouse-pointer visible.

What I have found out is that upon resuming, if I wait long enough for the display to be turned off due to inactivity (i.e. the /usr/share/acpi-support/screenblank script is run), then when I move the mouse or press a key to wake the display, everything is working normally again, and the black-screen is no more.

This is reproducible. In fact, I can just type in my password when the black screen is there (knowing that there is a password dialog hiding behind that black screen), blank the screen by running "sudo /usr/share/acpi-support/screenblank" and typing my password for sudo, and after the display is turned off (i.e. backlight turns off), when I move the mouse or hit a key, everything is visible again (i.e. no black screen).

Does this work for anyone else?

By the way, I am not running compiz at all -- just a gnome-classic session with xmonad. So at least in my case, this isn't a compiz issue.

Changed in compiz:
milestone: 0.9.8.2 → 0.9.8.4
Revision history for this message
Jani Monoses (jani) wrote :

This also affects a friend's 14.04 install with Gnome Classic not Unity/Compiz.

00:02.0 VGA compatible controller: Intel Corporation 2nd Generation Core Processor Family Integrated Graphics Controller (rev 09)

Revision history for this message
OC (olivier-c) wrote :

I cannot reproduce this bug anymore on my machine (ThinkPad X201, up-to-date 12.04 with no PPAs). I used to hit the "black screen with mouse" approx. 1 every 5 suspends. Today with 30 suspends in a row (I wanted to investigate Mark Edington’s workaround at #230) I cannot hit it.

Revision history for this message
Matt Price (matt-price) wrote : Re: [Bug 966744] Re: [i965] Resume from suspend leaves me with black screen or a screen of the desktop before it suspended. Compiz hung in intel_update_renderbuffers() from intel_prepare_render() from brw_draw_prims()

On Mon 10 Sep 2012 08:16:57 AM EDT, OC wrote:
> I cannot reproduce this bug anymore on my machine (ThinkPad X201, up-to-
> date 12.04 with no PPAs). I used to hit the "black screen with mouse"
> approx. 1 every 5 suspends. Today with 30 suspends in a row (I wanted to
> investigate Mark Edington’s workaround at #230) I cannot hit it.
>

I am also having trouble reproducing, but I think for me the key was
turning off gnome-screensaver & screen-lock. This indicates to me
(naivey) that the probem isn't in ocmpiz at all. in a crunch time at
work so can't try again right now, but when I have a chance I will see
if turning screensaver/lock back on again can reproduce the problem.

Revision history for this message
Sparhawk (sparhawkthesecond) wrote :

@edgimar: That command doesn't work on my system (command not found), but "xset dpms force off" does. I'll give that a go next time I hit the bug. FWIW, to run this from tty1–6, use "sleep 5 ; xset dpms force off -display :0" and then switch back with ctrl-alt-F7.

@matt-price: I still get this bug (more rarely now), but I had uninstalled gnome-screensaver (and installed xscreensaver instead), and never used screen-lock. So perhaps it's not related to either (in my case at least).

Changed in xserver-xorg-video-intel (Ubuntu Quantal):
assignee: nobody → Timo Aaltonen (tjaalton)
Revision history for this message
Julien-Charles Lévesque (jclevesque) wrote :

Doesn't seem to restricted to i965 since it happens on my laptop running i915 (ASUS UX31, IntelHD 4000)

Revision history for this message
Vladimir Dobriakov (vladimir-geekq) wrote :

I am currently experiencing this problem with my Lenovo X220 Notebook under Ubuntu 12.04, `lsmod|grep video` shows `i915`. But I've seen the same problem for years(!), on e.g. HP Z400 with Nvidia graphics and Ubuntu 11.04 and 11.10. The black-screen-lock happens either on sleep-resume and even after just locking the keyboard (lock, wait till display goes to sleep, move the mouse, display wakes up, mouse cursor visible, but nothing works).

So for me it looks like it is not related to particular hardware or driver - happens in compiz/unity with different hardware.

Changed in oem-priority:
assignee: nobody → James M. Leddy (jm-leddy)
bugbot (bugbot)
tags: added: compiz
Revision history for this message
Alex G (agutfraind2) wrote :

Resolution:
I had the problem on a Thinkpad x201. The underlying cause was actually a file corruption in the packages.

I ran
debsums -s

One of the packages which was corrupted was
gnome-screensaver

Reinstalling gnome-screensaver (and several other packages) fixed the problem.

Revision history for this message
Anmar Oueja (anmar) wrote :

Hello Alex,

If this is indeed the case, can't ubuntu up the number so people get
it when they upgrade. By the number of people we have complaining, I
think it warrants this small bump. What do you think?

Revision history for this message
bmomjian (bruce-momjian) wrote :

I don't think corrupt binaries is the general cause of the hang problem. I ran debsums -s and got no problems reported, but still see the hang. Also, I uninstalled gnome-screensaver a while ago, but still get the hang. I do have an Intel video card, but saw the hangs with an nVidia card as well.

Revision history for this message
Sparhawk (sparhawkthesecond) wrote :

@Alex, nice idea, but I also ran debsums -s with nothing corrupted. Also (as I mentioned above) I uninstalled gnome-screensaver ages ago too.

Revision history for this message
idsfa (idsfa) wrote :

Still haven't seen this bug reappear with xorg-edgers. Unfortunately the kernel there doesn't handle my encrypted home directory properly, so I'm two revisions back on that (3.5.0-9-generic), but I have not seen this problem in a LONG time, and I suspend several times every day.

Revision history for this message
bmomjian (bruce-momjian) wrote :

I am seeing it on a T60 running Ubuntu 12.04 with an Intel 945 video card. I have my lid set to "no action" so I am not even going into suspend, and can get it to fail 20% of the time. There really is no workaround except not closing the lid --- hope this is fixed soon. I have 8 other T60s with the same behavior.

Revision history for this message
bmomjian (bruce-momjian) wrote :

Oh, I should also mention I have seen the hang on a T60 with a Radeon X1300 video card as well.

Revision history for this message
JaniH (janzu78) wrote :

This affects me too. I'm running a clean install of Ubuntu 12.04.1 with Dell D430 w/ intel driver.

Revision history for this message
Matt Price (matt-price) wrote :

In a previous post i asserted that turning off gnome-screensaver stops
these crashes from happening. This turns out not to be the case.
Unfortunately, I am having trouble capturing an xtrace of the crash --
it almost seems as if xtrace itself somehow protects the driver from the
crash! -- anyway, I'm finding that I can sometimes recover from the
crash by switching to VT1 and running

DISPLAY=:0 unity
DISPLAY=:0 xtrace compiz --replace > ~/trace.txt

But usually I'll end up with some problems, like all window management
being switched off so that it seems as though compiz isn't running at
all. And sometimes this attempt at recovery will fail almost
completely, in that I just end up with an unchanging image of my desktop
wallpaper (sometimes the active windows will flash briefly when I switch
to VT7, then disappear behind the wallpaper. I'm really not sure what
the next steps in diagnosis are, htough I'll keep trying to capture an
xtrace. This is on a Lenovo Thikpad T410, wiht intel graphics & (of
course) the i965 driver.

Thanks all!

Revision history for this message
Jason Apol (13l0yp6viz-jason) wrote :
Download full text (6.5 KiB)

Matt - How does one "recover from the crash by switching to VT1"?

I'd be happy if I can get to some kind of desktop and save my work.
--
Jason Apol, Sr. Engineer - Logicraft Industrial Systems, Inc
 www.lcrft.com

On Mon, Sep 17, 2012 at 5:28 PM, Matt Price <email address hidden> wrote:

> In a previous post i asserted that turning off gnome-screensaver stops
> these crashes from happening. This turns out not to be the case.
> Unfortunately, I am having trouble capturing an xtrace of the crash --
> it almost seems as if xtrace itself somehow protects the driver from the
> crash! -- anyway, I'm finding that I can sometimes recover from the
> crash by switching to VT1 and running
>
> DISPLAY=:0 unity
> DISPLAY=:0 xtrace compiz --replace > ~/trace.txt
>
> But usually I'll end up with some problems, like all window management
> being switched off so that it seems as though compiz isn't running at
> all. And sometimes this attempt at recovery will fail almost
> completely, in that I just end up with an unchanging image of my desktop
> wallpaper (sometimes the active windows will flash briefly when I switch
> to VT7, then disappear behind the wallpaper. I'm really not sure what
> the next steps in diagnosis are, htough I'll keep trying to capture an
> xtrace. This is on a Lenovo Thikpad T410, wiht intel graphics & (of
> course) the i965 driver.
>
> Thanks all!
>
> --
> You received this bug notification because you are subscribed to the bug
> report.
> https://bugs.launchpad.net/bugs/966744
>
> Title:
> [i965] Resume from suspend leaves me with black screen or a screen of
> the desktop before it suspended. Compiz hung in
> intel_update_renderbuffers() from intel_prepare_render() from
> brw_draw_prims()
>
> Status in Compiz:
> Confirmed
> Status in The Linux Kernel:
> Incomplete
> Status in OEM Priority Project:
> New
> Status in OEM Priority Project precise series:
> New
> Status in X.org xf86-video-intel:
> Confirmed
> Status in “compiz” package in Ubuntu:
> Invalid
> Status in “xserver-xorg-video-intel” package in Ubuntu:
> Confirmed
> Status in “compiz” source package in Precise:
> Invalid
> Status in “xserver-xorg-video-intel” source package in Precise:
> Confirmed
> Status in “compiz” source package in Quantal:
> Invalid
> Status in “xserver-xorg-video-intel” source package in Quantal:
> Confirmed
>
> Bug description:
> This happens intermittently. Sometimes upon resume, all I see is a
> black screen and the cursor. the mouse and keyboard respond (the mouse
> moves) but nothing changes. Switching to console and back doesn't fix
> it. Killing Compiz also does not fix the issue. Further, if one has
> password lock disabled on resume the user will be stuck with a frozen
> display of whatever was last shown on the screen before suspend. The
> mouse cursor still changes as you hover over various elements but the
> display is frozen solid. One can switch to another TTY via ctrl+alt+F1
> but killing the X session is the only way to get back to the desktop.
>
> Attempts to replace compiz have been fruitless. Trying the 3.4 kernel
> also did not fix the issue.
>
> BTW, I haven nothing c...

Read more...

Revision history for this message
Matt Price (matt-price) wrote :

On 12-09-17 05:38 PM, Jason Apol wrote:
> Matt - How does one "recover from the crash by switching to VT1"?
>
> I'd be happy if I can get to some kind of desktop and save my work.
Ctrl-Alt-F1 will take you to a "Virtual Terminal" -- it's a console
screen where you can log in just as you would at the gnome-terminal
prompt. Then you issue those commands, and switch back to the running X
session, which is running on VT7, with Ctrl-Alt-F7. you cna switch
between the 6 built-in VT's with F1-F6; f7 is the default X session; F8
seems to echo system log messages unless you log in a second user, with
'Swtitch User', at which point a new X session will start on that screen.

Does that help?

ps -- My success rate is not that high -- at best I usually end up with
a barely-functional x system, try to save as much work as I can, then
restart X with 'sudo restart lightdm', which destroys my X session but
also resets whatever that display bug is. At worst the resultant
session is just as nonresponsive as the original crashed one.

Revision history for this message
Lawrence Patrick Calulo (lugkhast) wrote :

Matt, was the battery on your ThinkPad not attached when you weren't able to reproduce the hang?

When I was trying to get the hang to happen on my X200, I got about two days of hang-free Unity. When I attached the battery and continued trying, I got two hangs within three hours of each other.

It doesn't make sense to me that the two are related, but it may be worth a try.

Revision history for this message
Leon Marcus (treesurf) wrote :

Using a Lenovo Ideapad Y550 with integrated Intel everything. I get the hang whether or not the battery is installed.

Revision history for this message
Jason Apol (13l0yp6viz-jason) wrote :
Download full text (6.4 KiB)

I will try that next time my system gets out in the weeds. Thanks!
--
Jason Apol, Sr. Engineer - Logicraft Industrial Systems, Inc
Cell: 616-299-7558 www.lcrft.com

On Mon, Sep 17, 2012 at 7:30 PM, Matt Price <email address hidden> wrote:

> On 12-09-17 05:38 PM, Jason Apol wrote:
> > Matt - How does one "recover from the crash by switching to VT1"?
> >
> > I'd be happy if I can get to some kind of desktop and save my work.
> Ctrl-Alt-F1 will take you to a "Virtual Terminal" -- it's a console
> screen where you can log in just as you would at the gnome-terminal
> prompt. Then you issue those commands, and switch back to the running X
> session, which is running on VT7, with Ctrl-Alt-F7. you cna switch
> between the 6 built-in VT's with F1-F6; f7 is the default X session; F8
> seems to echo system log messages unless you log in a second user, with
> 'Swtitch User', at which point a new X session will start on that screen.
>
> Does that help?
>
> ps -- My success rate is not that high -- at best I usually end up with
> a barely-functional x system, try to save as much work as I can, then
> restart X with 'sudo restart lightdm', which destroys my X session but
> also resets whatever that display bug is. At worst the resultant
> session is just as nonresponsive as the original crashed one.
>
> --
> You received this bug notification because you are subscribed to the bug
> report.
> https://bugs.launchpad.net/bugs/966744
>
> Title:
> [i965] Resume from suspend leaves me with black screen or a screen of
> the desktop before it suspended. Compiz hung in
> intel_update_renderbuffers() from intel_prepare_render() from
> brw_draw_prims()
>
> Status in Compiz:
> Confirmed
> Status in The Linux Kernel:
> Incomplete
> Status in OEM Priority Project:
> New
> Status in OEM Priority Project precise series:
> New
> Status in X.org xf86-video-intel:
> Confirmed
> Status in “compiz” package in Ubuntu:
> Invalid
> Status in “xserver-xorg-video-intel” package in Ubuntu:
> Confirmed
> Status in “compiz” source package in Precise:
> Invalid
> Status in “xserver-xorg-video-intel” source package in Precise:
> Confirmed
> Status in “compiz” source package in Quantal:
> Invalid
> Status in “xserver-xorg-video-intel” source package in Quantal:
> Confirmed
>
> Bug description:
> This happens intermittently. Sometimes upon resume, all I see is a
> black screen and the cursor. the mouse and keyboard respond (the mouse
> moves) but nothing changes. Switching to console and back doesn't fix
> it. Killing Compiz also does not fix the issue. Further, if one has
> password lock disabled on resume the user will be stuck with a frozen
> display of whatever was last shown on the screen before suspend. The
> mouse cursor still changes as you hover over various elements but the
> display is frozen solid. One can switch to another TTY via ctrl+alt+F1
> but killing the X session is the only way to get back to the desktop.
>
> Attempts to replace compiz have been fruitless. Trying the 3.4 kernel
> also did not fix the issue.
>
> BTW, I haven nothing connected to the laptop when I resume and before
> suspend. This is a ...

Read more...

Revision history for this message
Matt Price (matt-price) wrote :

On Mon 17 Sep 2012 07:59:27 PM EDT, Lawrence Patrick Calulo wrote:
> Matt, was the battery on your ThinkPad not attached when you weren't
> able to reproduce the hang?
>
> When I was trying to get the hang to happen on my X200, I got about two
> days of hang-free Unity. When I attached the battery and continued
> trying, I got two hangs within three hours of each other.
>
> It doesn't make sense to me that the two are related, but it may be
> worth a try.
>

The battery is always in my laptop. I sometimes have the AC power cord
attached and sometimes not, but I haven't been keeping very careful
track of that. I've certainly had the error appear when the AC power
is unplugged, because I've had embarrassing moments showing up to
meetings with a crashed computer (arrggh!).

I'll experiment when I don't have 2 lectures to write for the morning!
Thanks Lawrence.

Revision history for this message
Anmar Oueja (anmar) wrote :

I switched to 3.6-rc5 from the kernel PPA
(http://kernel.ubuntu.com/~kernel-ppa/mainline/) and so far, I haven't
had any blank screens upon resume. Please have a try.

BTW, although the kernel says quantal, you can install it on 12.04.
You need 4 packages / per architecture.

Please test and provide feedback in the hope of helping engineers fix it.

Revision history for this message
Matt Price (matt-price) wrote :

On Tue 18 Sep 2012 06:00:08 AM EDT, Anmar Oueja wrote:
> I switched to 3.6-rc5 from the kernel PPA
> (http://kernel.ubuntu.com/~kernel-ppa/mainline/) and so far, I haven't
> had any blank screens upon resume. Please have a try.
>
> BTW, although the kernel says quantal, you can install it on 12.04.
> You need 4 packages / per architecture.
>
does anyone know what exactly is missing from the mainline kernel, or
if using it is really advised? I use a realtek wireless card that used
to be pretty buggy, I would hate to have to give up wireless to get
suspend back...

Changed in compiz:
status: Confirmed → Invalid
no longer affects: oem-priority/precise
no longer affects: oem-priority
description: updated
Revision history for this message
Nazim Aghabayov (nazim-aga) wrote :

On Tue 18 Sep 2012 06:00:08 AM EDT, Anmar Oueja wrote:
> I switched to 3.6-rc5 from the kernel PPA
> (http://kernel.ubuntu.com/~kernel-ppa/mainline/) and so far, I haven't
> had any blank screens upon resume. Please have a try.
>

Switched to v3.6-rc6-quantal and have no suspend issues with Samsung np900xc3. Generic kernels for precise 12.04, 3.2.0-30 and 3.2.0-29 are not resuming correctly giving a black screen with cursor instead of the unity session.

Revision history for this message
Anmar Oueja (anmar) wrote :

Matt,

> does anyone know what exactly is missing from the mainline kernel, or
> if using it is really advised? I use a realtek wireless card that used
> to be pretty buggy, I would hate to have to give up wireless to get
> suspend back...

That is a fairly complicated question to answer but I doubt you will
have a problem with wireless with 3.6 simply because the realtek
drivers are usually updated with every kernel release so if you card
use to work ok, in 3.2, it should work much better in 3.6.

Testing it should be really easy. Just download the package , I have
the address a couple of comments above, and install them, using dpkg
-i, and boot. If doesn't work, remove the packages and you are back to
your original kernel.

Revision history for this message
bmomjian (bruce-momjian) wrote :

Matt Price (matt-price) asked about the underlying cause of the hang, and I think it is most clearly stated in the link to freedesktop-bugs #50227 at the top of this page in the status table, which finally has this bug marked as critical by almost every subsystem.

The fact this bug report goes back to March indicates this bug fix is not open source's finest hour. :-( I wonder what the timeframe is for a fix to be pushed as an Ubuntu 12.04 update.

Revision history for this message
kenneth hsu (kennethhsu) wrote :

Not sure if this information is useful, but I used to encounter this bug at least per day with Ubuntu 12.04 on a Lenovo Thinkpad x200. After installing Xubuntu 12.04 (fresh install, not via the xubuntu-desktop package), I have yet to encounter this bug after almost a month of use.

gcc (chris+ubuntu-qwirx)
Changed in compiz (Ubuntu Precise):
status: Invalid → Confirmed
Omer Akram (om26er)
Changed in compiz (Ubuntu Precise):
status: Confirmed → Invalid
tags: added: kernel-da-key
Revision history for this message
In , Sean Channel (sean-channel) wrote :

I got no problems. Try this: add "i915.semaphores=1" to your kernel options in /etc/default/grub (AND run update-grub AND reboot). E.g:

  GRUB_CMDLINE_LINUX_DEFAULT="quiet splash elevator=noop i915.semaphores=1"

I also do created /usr/share/X11/xorg.conf.d/20-intel.conf with this in it:

  Section "Device"
      Identifier "Intel"
      Driver "intel"
      Option "AccelMethod" "exa"
      Option "MigrationHeuristic" "greedy"
      Option "DebugWait" "true"
      Option "SwapbuffersWait" "false"
  EndSection

^ ^ again, reboot if you do that and/or play with those options.

YMMV.

Revision history for this message
Sean Channel (sean-channel) wrote :

almost forgot: I also use the xserver-xorg-video-intel package from https://launchpad.net/~ubuntu-x-swat

Revision history for this message
paulandmildred (paulandmildred) wrote :

for me, I get this everytime when I'm logged in with Gnome (new), but not when using Gnome Classic. haven't tried in Unity or others.

12.04, 64bit, most recent updates (yesterday), Kernel Linux 3.2.0-30-generic
Intel® Xeon(R) CPU W3505 @ 2.53GHz × 2
Quadro NVS 295 (GPU 0), Driver "nvidia"

Revision history for this message
nfsd (in4mer+launchpad) wrote :

This is happening to me on an xserver-xorg-radeon box, intel processor. Happens after about five days of daily suspend/resumes. I run compiz.

Revision history for this message
paulandmildred (paulandmildred) wrote :

of course, as soon as I posted my comment above, I got the same thing in Gnome Classic as I did in new Gnome.

upon restart (sudo lightdm restart didn't work), the system detected a failure with /usr/lib/x64_86-linux-gnu/colord/colord. not sure if it is related.

Revision history for this message
Bob Merhebi (bobmerhebi) wrote :

Happens on Dell Studio XPS 1640 with [AMD] nee ATI RV730 XT [Mobility Radeon HD 4670]

$ lsb_release -rd
Description: Ubuntu 12.04.1 LTS
Release: 12.04

$ uname -rmpi
3.2.0-30-generic x86_64 x86_64 x86_64

Revision history for this message
Edward Donovan (edward.donovan) wrote :

For the folks in comments 260-263: this bug is just about these symptoms under Intel video hardware. You all mention nvidia or radeon,,so that's different . (It doesn't matter if it's an Intel CPU.) With hundreds of posts here, you may not have seen what Daniel wrote in #196 :

"Hi all.

We need to be focussed and specific or else bugs like this never get resolved.

Since all the stacks I have seen so far have been identical hangs in i965_dri.so, I have now clarified that this is an Intel graphics driver bug. Although a workaround in compiz might still be possible.

If you're not using Intel graphics, please log a new bug. [...]"

Or there may be similar bugs filed about those drivers. I'm sorry I can't point you to them right now. And if there is a compiz workaround, that might help you, so subscribe to this bug if you want. But reporting information about non-Intel video will unfortunately just clog this bug, not help it.

I guess I'll put some form of that information into the main description, since seeing [i965] in the title isn't going to clarify that for everybody. And if we know other bugs to subscribe to, for other hardware, we might mention them there, too.

Thanks!

description: updated
Revision history for this message
justfortherec (dasdarfdochnichtwahrsein) wrote :

Hi,

to me it seems that this got worse since about a week ago. For me this also happened before. But now it happens 9 times out of 10 times. Is it possible that an update screwed this even more?

On September 10th my apt log shows installation of kernel 3.2.0-30.48

I use Ubuntu 12.04 on a Thinkpad x201.

Best Regards

Revision history for this message
Andreas Hohenegger (hohenegger) wrote :

Same on my Lenovo Thinkpad T61 with *intel graphics* (above it was reported with nvidia graphics).

Revision history for this message
Matthew Petty (matthewpettyuk) wrote :

My System76 Lemur just started doing this in the last week. Tech support advised me to use the suspend key before closing thelid, rather than just closing the lid. I will see if that has any affect.

Timo Aaltonen (tjaalton)
no longer affects: compiz (Ubuntu)
no longer affects: compiz (Ubuntu Precise)
no longer affects: compiz (Ubuntu Quantal)
Revision history for this message
In , Chris Wilson (ickle) wrote :

Hmm, the X.log indicates 2.17, there were a few related fixes as well, can you please install a 2.20.8 from your distrobution updates?

Revision history for this message
In , Timo Aaltonen (tjaalton) wrote :

I have a couple of users on ubuntu 12.10 (2.20.8) that also tried 3.6rc7, and the hang happens every time the screen is closed or screensaver kicks in. I can't seem to be able to reproduce it myself though..

Revision history for this message
In , Chris Wilson (ickle) wrote :

Given the batch submit immediately after the vsync'ed copy, I can't see what else userspace can do to prevent the WAIT_FOR_EVENT hang...

Revision history for this message
In , Chris Wilson (ickle) wrote :

Hmm, there is some similarity here between this and bug 51616 if in both cases the kernel reports the pipe as active, but in reality it is disabled.

Revision history for this message
Ayan Nath (mymail-ayaan) wrote :

how to solve the problem. plz be quick and help
thnx in advance

Revision history for this message
Christian Dysthe (christian-dysthe) wrote :

I have a Thinkpad L520 which had this problem in 12.10. I have also had this problem on an Asus machine a while back. I tried the solution I found for the Asus on my Thinkpad and it worked. This is what I did:

Create a file /etc/pm/sleep.d/20_custom_ehci_hcd containing:

#!/bin/sh
# File: "/etc/pm/sleep.d/20_custom-ehci_hcd".
case "${1}" in
 hibernate|suspend)
# Unbind ehci_hcd for first device 0000:00:1a.0:
echo -n "0000:00:1a.0" | tee /sys/bus/pci/drivers/ehci_hcd/unbind
# Unbind ehci_hcd for second device 0000:00:1d.0:
echo -n "0000:00:1d.0" | tee /sys/bus/pci/drivers/ehci_hcd/unbind
;;
 resume|thaw)
# Bind ehci_hcd for first device 0000:00:1a.0:
echo -n "0000:00:1a.0" | tee /sys/bus/pci/drivers/ehci_hcd/bind
# Bind ehci_hcd for second device 0000:00:1d.0:
echo -n "0000:00:1d.0" | tee /sys/bus/pci/drivers/ehci_hcd/bind
;;
esac

Set it 755 and owned by root. Now the display returns after sleep without problems.

tags: removed: kernel-da-key
Revision history for this message
benmhall (ubuntu-linuxgruven) wrote :

Hello.

For me, Christian Dysthe's suggestion (#269) only prevented my ThinkPad X200s from going to sleep.

This is a frustrating bug that greatly affects my use of the system. Interestingly, I took an image of my ThinkPad X200s and put the drive in a ThinkPad T500 (Same vintage but with an ATI 3650) and the system suspends and resumes perfectly.

An ATI card working better at suspend and resume in Linux than an Intel card. Who'd have thought.

Ursula Junque (ursinha)
tags: added: quantal
Revision history for this message
In , Shu Latif (shulatif) wrote :

So far this has helped me out.

Revision history for this message
Timo Aaltonen (tjaalton) wrote :

hey folks, I've pushed a snapshot from current -intel git master (will hopefully become 2.20.9 soon..) to my ppa *for quantal*:

https://launchpad.net/~tjaalton/+archive/ppa
(amd64 built, i386 building soon)

I have confirmations from two subscribers that it has fixed the hang. Further testing is welcome.

Changed in xserver-xorg-video-intel (Ubuntu Quantal):
status: Confirmed → Incomplete
Revision history for this message
Anmar Oueja (anmar) wrote :

Can you make it work for Precise as well?

Revision history for this message
Carlos Moffat (carlos-eldiabloenlosdetalles) wrote :

Timo,

A build for Precise would be great!

Thanks

Revision history for this message
Timo Aaltonen (tjaalton) wrote :

We're working on a proper backport for precise that will be SRU'd, but I've uploaded the snapshot to the ppa so you can try it on precise.. once it's built, that is.

Changed in compiz:
milestone: 0.9.8.4 → none
Revision history for this message
gcc (chris+ubuntu-qwirx) wrote :

Hi @tjaalton,

Unfortunately the i386 build in your PPA seems to have failed:

Processing triggers for libc-bin ...
ldconfig deferred processing now taking place
Checking correctness of source dependencies...
After installing, the following source dependencies are still unsatisfied:
xserver-xorg-dev(inst 2:1.11.4-0ubuntu10.8 ! >= wanted 2:1.11.99.901)
Source-dependencies not satisfied; skipping xserver-xorg-video-intel
******************************************************************************
Finished at 20120927-2317
Build needed 00:00:00, 0k disk space
RUN: /usr/share/launchpad-buildd/slavebin/scan-for-processes ['/usr/share/launchpad-buildd/slavebin/scan-for-processes', 'eca7c87f4ca61935bc5165de1001c7f657a5b738']
Scanning for processes to kill in build /home/buildd/build-eca7c87f4ca61935bc5165de1001c7f657a5b738/chroot-autobuild...
RUN: /usr/share/launchpad-buildd/slavebin/umount-chroot ['umount-chroot', 'eca7c87f4ca61935bc5165de1001c7f657a5b738']
Unmounting chroot for build eca7c87f4ca61935bc5165de1001c7f657a5b738...
RUN: /usr/share/launchpad-buildd/slavebin/remove-build ['remove-build', 'eca7c87f4ca61935bc5165de1001c7f657a5b738']
Removing build eca7c87f4ca61935bc5165de1001c7f657a5b738

https://launchpadlibrarian.net/117574890/buildlog_ubuntu-precise-i386.xserver-xorg-video-intel_2%3A2.20.8-0ubuntu2.1~precise1_MANUALDEPWAIT.txt.gz

Please can you see if you can fix it?

Cheers, Chris.

Revision history for this message
Ursula Junque (ursinha) wrote :

Hi Timo,

I've installed the packages in your PPA and when I rebooted I ran into bug 1058125 (private -- contains sensitive information --, let me know if you can't see it). I'm not sure it's related, but as it was the first issue apport caught after installing your package and rebooting, I thought it might be worth looking.

I haven't checked yet if it fixes this very bug.

Thanks!

Revision history for this message
macstevejb (macstevejb) wrote :

Installing the xserver-xorg-video-intel package from this PPA solved this problem for me on my Asus laptop using Ubuntu 12.10 beta 2 (i386)

Thanks :)

Revision history for this message
david.barbion (david-barbion) wrote :

Can anyone point me (us) on the right package to install? Can't find version 2.20.9 on Timo's PPA.

Revision history for this message
Edward Donovan (edward.donovan) wrote :

David -

It's xserver-xorg-video-intel 2.20.8-0ubuntu2.1 . Timo's saying that this code will soon be 2.20.9, but it isn't yet -- I believe. Thanks.

Revision history for this message
Aatish (aatish) wrote :

Could anyone help point a noob in the right direction? I'm trying to install Timo's PPA, and am running Precise on a Lenovo X220.

I ran sudo add-apt-repository ppa:tjaalton/ppa and sudo apt-get update

But when I run sudo apt-get install xserver-xorg-video-intel it tells me that I already have the latest version. I am running version 2:2.17.0-1ubuntu4.1 but the version in Timo's PPA is 2:2.20.8-0ubuntu2.1~precise1. What am I doing wrong?

Revision history for this message
Carlos Moffat (carlos-eldiabloenlosdetalles) wrote :

Hi Aathish,

You're not doing anything wrong. The driver is not on the PPA yet. Timo set it to build for Precise, but something went wrong, so it's only available for Quantal. Hopefully he'll be able to take a look and get a version for Precise ready soon.

All the best,
Carlos

Revision history for this message
Stefan Tegtmeyer (st-tegtmeyer) wrote :

installing xserver-xorg-video-intel from https://launchpad.net/~tjaalton/+archive/ppa for 12.10 "development branch, Quantal Quetzal"on AMD64 solves this issue for me

Thanks to tjaalton

Revision history for this message
Timo Aaltonen (tjaalton) wrote :

Thanks for testing.. I've pushed a new version for precise that fixes the build-depends. Proper fix for quantal will be uploaded early next week.

Ursula, that bug shouldn't be related to my ppa, since it has happened with the version in the archive.

Changed in xserver-xorg-video-intel (Ubuntu Quantal):
milestone: ubuntu-12.10-beta-2 → none
status: Incomplete → In Progress
Revision history for this message
bmomjian (bruce-momjian) wrote :
Revision history for this message
Timo Aaltonen (tjaalton) wrote :

2.20.9 is out already, pushed to quantal

Changed in xserver-xorg-video-intel (Ubuntu Quantal):
status: In Progress → Fix Committed
Revision history for this message
Launchpad Janitor (janitor) wrote :

This bug was fixed in the package xserver-xorg-video-intel - 2:2.20.9-0ubuntu1

---------------
xserver-xorg-video-intel (2:2.20.9-0ubuntu1) quantal; urgency=low

  * Merge from unrelease debian git.
    - fixes racy UXA pageflip code (LP: #966744)
  * Drop dont-run-intel-mode-fini-before-preinit.diff, upstream.

xserver-xorg-video-intel (2:2.20.9-1) UNRELEASED; urgency=low

  * New upstream bugfix release.
    - Fixes a long-standing bug in UXA pageflip code that was the
      source of many GPU and application crashes.
 -- Timo Aaltonen <email address hidden> Sat, 29 Sep 2012 16:45:35 +0300

Changed in xserver-xorg-video-intel (Ubuntu Quantal):
status: Fix Committed → Fix Released
Revision history for this message
Nick Rhodes (ngrhodes) wrote :

Fixed my suspend issues with Quantal using Timo's PPA version 2.20.8-0ubuntu2.1.

Thanks Timo for the quick fix.
I realised how important this is/was to get fixed for Quantal due to the lack of Unity 2D to fall back on.

Timo, can you confirm that 2.20.8-0ubuntu2.1 has become 2:2.20.9-1 ?

Revision history for this message
Timo Aaltonen (tjaalton) wrote :

Nick: essentially, yes. It has some additional fixes too.

Revision history for this message
Carlos Molina (camolin3) wrote :

I have the same problem. I'm using a Intel integrated video card, with the i915 driver.

Revision history for this message
jhfhlkjlj (fdsuufijjejejejej-deactivatedaccount) wrote :

so far so good with the ppa package for me.

Revision history for this message
Nekhelesh Ramananthan (nik90) wrote :

Timo: Does 2.20.8-0ubuntu2.1 (precise) also provide the fix to this bug? I am running precise and running into this bug. When can expect an update on precise?

Revision history for this message
Éric Piel (pieleric) wrote :

I can also report that on Quantal with 2.20.9-0ubuntu1, I haven't been able to reproduce this bug.

That said, even if this week I've extensivily tried to trigger the bug, it's far from a complete proof the bug is gone. On my hardware it sometimes can take several weeks before the bug triggers again.

Revision history for this message
Anmar Oueja (anmar) wrote : Re: [Bug 966744] Re: [i965] Resume from suspend leaves me with black screen or a screen of the desktop before it suspended. Compiz hung in intel_update_renderbuffers() from intel_prepare_render() from brw_draw_prims()

I tried things on 12.10 Beta 2 and it works every time. I have been
using my TinkPad x201 constantly with no issues whatsoever. I am using
2.20.9 version

Revision history for this message
Matt Price (matt-price) wrote :

On 12-10-04 08:51 AM, Éric Piel wrote:
> I can also report that on Quantal with 2.20.9-0ubuntu1, I haven't been
> able to reproduce this bug.
>
> That said, even if this week I've extensivily tried to trigger the bug,
> it's far from a complete proof the bug is gone. On my hardware it
> sometimes can take several weeks before the bug triggers again.
>
I've also had a pretty long run without the bug now. I really hope it's
fixed! but like Eric, I can't say for sure.

Revision history for this message
Timo Aaltonen (tjaalton) wrote :

I've uploaded a new driver to precise-proposed, and tested it with the dri2-swap testcase in the latest -intel driver, and couldn't reproduce the hang (was trivial with the old version).

Changed in xserver-xorg-video-intel (Ubuntu Precise):
status: Confirmed → In Progress
Revision history for this message
Timo Aaltonen (tjaalton) wrote :
Timo Aaltonen (tjaalton)
description: updated
Timo Aaltonen (tjaalton)
description: updated
Timo Aaltonen (tjaalton)
description: updated
Revision history for this message
Adam Conrad (adconrad) wrote : Please test proposed package

Hello Anmar, or anyone else affected,

Accepted xserver-xorg-video-intel into precise-proposed. The package will build now and be available at http://launchpad.net/ubuntu/+source/xserver-xorg-video-intel/2:2.17.0-1ubuntu4.2 in a few hours, and then in the -proposed repository.

Please help us by testing this new package. See https://wiki.ubuntu.com/Testing/EnableProposed for documentation how to enable and use -proposed. Your feedback will aid us getting this update out to other Ubuntu users.

If this package fixes the bug for you, please change the bug tag from verification-needed to verification-done. If it does not, change the tag to verification-failed. In either case, details of your testing will help us make a better decision.

Further information regarding the verification process can be found at https://wiki.ubuntu.com/QATeam/PerformingSRUVerification . Thank you in advance!

Changed in xserver-xorg-video-intel (Ubuntu Precise):
status: In Progress → Fix Committed
tags: added: verification-needed
Revision history for this message
Terence (helloterence) wrote :

Pretty new to Linux. Can someone provide me steps as to how to build this fix?

Revision history for this message
Eugene Romanenko (eros2) wrote :

Terence, enable "proposed" repository in Update manager (see there https://wiki.ubuntu.com/Testing/EnableProposed ) and update xserver-xorg-video-intel package.

Revision history for this message
Nekhelesh Ramananthan (nik90) wrote :

Seems to works for me.

tags: added: verification-done
removed: verification-needed
Revision history for this message
Rakitha S Beminiwattha (rakithab) wrote :

I have updated xserver-xorg-video-intel package in the precise-proposed. Let me know if I have to do anything to test the fix. I will use my notebook as usual.

Revision history for this message
In , Brettcornwall-1 (brettcornwall-1) wrote :

Actually, someone built a package including some other recent fixes and it looks like the problem has been resolved for a number of people. You can check out the LP report for specifics (sorry, in a bit of a rush).

Regardless, I've installed the package from his PPA and have been using normal lid-closing suspend (had been switching to TTY1, logging in as root, then using pm-suspend for the past few months).

I haven't had a single issue since. So unless there's any claiming otherwise, this bug is likely fixed.

Revision history for this message
In , Chris Wilson (ickle) wrote :

Closing as the UXA DPMS off vs pageflip race.

Changed in xserver-xorg-video-intel:
status: Confirmed → Fix Released
Revision history for this message
Archish (archish) wrote :

Issue is resolved for me after enabling the new intel update from the proposed repository in 12.04. Thanks

Revision history for this message
bmomjian (bruce-momjian) wrote :

You can see the Ubuntu pending update queue here, particularly for Precise (12.04): http://people.canonical.com/~ubuntu-archive/pending-sru.html. It shows the pending update fix as four days old.

Revision history for this message
jhfhlkjlj (fdsuufijjejejejej-deactivatedaccount) wrote :

Just following up: This bug has been obliterated since installing those packages. Push that update! :)

Revision history for this message
Scott Kitterman (kitterman) wrote :

There's a second bug this upload fixes that also has to be verified. Bug #992391

Revision history for this message
bmomjian (bruce-momjian) wrote :

I found an easy way to get this update. This might be obvioius to experienced Ubuntu users, but it wasn't obvious to me.

In System Tools/Administration/Update Manager, under Settings/Updates, check Pre-release Updates. Then close the window and click Check. You will see a lot of proposed updates. Go to the top and uncheck the heading Propsed Updates, then scroll to the bottom and just check the Intel driver update. Choose Install Updates, then go back to Settings/Update and uncheck Pre-release Updates, close the window, and reboot. This will get you just the Intel driver update.

Revision history for this message
Colin Watson (cjwatson) wrote : Update Released

The verification of this Stable Release Update has completed successfully and the package has now been released to -updates. Subsequently, the Ubuntu Stable Release Updates Team is being unsubscribed and will not receive messages about this bug report. In the event that you encounter a regression using the package from -updates please report a new bug using ubuntu-bug and tag the bug report regression-update so we can easily find any regresssions.

Revision history for this message
Launchpad Janitor (janitor) wrote :

This bug was fixed in the package xserver-xorg-video-intel - 2:2.17.0-1ubuntu4.2

---------------
xserver-xorg-video-intel (2:2.17.0-1ubuntu4.2) precise-proposed; urgency=low

  * Add 106-uxa-only-consider-attached-outputs-valid.patch, two upstream
    commits merged together that fix GPU hangs on resume, due to outputs
    getting marked valid before the kernel reports them attached.
    (LP: #992391)
  * Replace 101_copy-fb.patch with the patch that went upstream, backported
    to build against the older xserver. Needed for other backports.
    - refreshed patch 104 to apply.
  * Add 107-uxa-fix-pageflip-race.patch, fixes a race with pageflip and
    dpms, resulting in for instance compiz hanging. (LP: #966744)
    - consists of five upstream commits merged as one and backported to
      build against the xserver in precise
 -- Timo Aaltonen <email address hidden> Fri, 05 Oct 2012 22:40:38 +0300

Changed in xserver-xorg-video-intel (Ubuntu Precise):
status: Fix Committed → Fix Released
Changed in linux:
status: Incomplete → Fix Released
Revision history for this message
Koopee (koopee1234) wrote :

I thought this bug was fixed, but today got the blank screen with mouse moving again.
ASUS UX32VD & Ubuntu 12.04.1
Linux 3.2.0-32-generic #51-Ubuntu SMP Wed Sep 26 21:33:09 UTC 2012

Please ask if you need any more details.

Revision history for this message
jhfhlkjlj (fdsuufijjejejejej-deactivatedaccount) wrote :

@Koopee, please do file a new report as this one has been fixed.

Revision history for this message
Michael (michaeljt) wrote :

@Chauncellor: I am also still getting a black screen on resume from time to time (Dell Latitude E6410 with Intel Arrandale). X.Org is unresponsive, but I can change to a virtual text terminal and blindly enter commands. Can you suggest an "apport-cli" command or other I should enter to get good data for a submission, presumably for a new bug report?

Revision history for this message
jhfhlkjlj (fdsuufijjejejejej-deactivatedaccount) wrote :

@Michael, I would report this bug against intel's driver:

ubuntu-bug xserver-xorg-video-intel

Mention this bug and the fact that this did not fix it for you. Take a look at the freedesktop.org bugzilla report linked at the top of this page - it will provide steps on how to get the proper logs to make sure that this is indeed intel's fault. Attach the same logs in the same manner that I did in the bugzilla report so that Mr. Wilson will be able to get a better look at the problem.

Best of luck.

Revision history for this message
Maakuth (markus-vuorio) wrote :

I did get the black screen after resume today, but this was the first time after the xserver-xorg-video-intel update. It used to do this way more often.

Revision history for this message
roman shaikh (romaan7) wrote :

I am using ASUS K55VM laptop..its annoying man since i need to lock my screen very often ..
using tty to restart lightDM for time being..

Revision history for this message
Andreas Hohenegger (hohenegger) wrote :

I seem to encounter this bug as I just commented accidentally in https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/995013 . Above I read that it is fixed in package xserver-xorg-video-intel - 2:2.20.9-0ubuntu1. According to my system I have xserver-xorg-video-intel - 2:2.20.9-0ubuntu2 (quantal ppa's). Is it fixed there?

Revision history for this message
jhfhlkjlj (fdsuufijjejejejej-deactivatedaccount) wrote :

Anyone experiencing these symptoms on a completely up-to-date system should log a new bug:

'ubuntu-bug xserver-xorg-video-intel'

If anyone opens the bug, please comment here with a link to it so any people still experiencing issues can hop over and 'me, too' it.

Revision history for this message
Andreas Hohenegger (hohenegger) wrote :

My system is up to date but I can report now that it is fixed for me with graphics drivers from xorg-edgers ppa (inluding kernel 3.7.0-4-generic 32 bit). But not in the present packages quoted above.

Revision history for this message
joshyg6 (josheg) wrote :

Still experiencing issue..

Revision history for this message
joshyg6 (josheg) wrote :

PS I try to run the above 'ubuntu-bug' command, and I keep getting an error that the package isn't an official Ubuntu package.. What??

Revision history for this message
joshyg6 (josheg) wrote :

Fix of xserver-xorg-video-intel updates doesn't work in Quantal

Revision history for this message
jhfhlkjlj (fdsuufijjejejejej-deactivatedaccount) wrote :

Please stop commenting on this bug report. Hundreds of people are subscribed and will be spammed with emails. If any support is needed please do go on askubuntu.com or the official forums/IRC channels. If you are experiencing this bug with all the updates, please follow my previous post (post # 325)

@joshyg6, you may be using unofficial packages (PPA?).

description: updated
Revision history for this message
joshyg6 (josheg) wrote : Re: [Bug 966744] Re: [i965] Resume from suspend leaves me with black screen or a screen of the desktop before it suspended. Compiz hung in intel_update_renderbuffers() from intel_prepare_render() from brw_draw_prims()
Download full text (13.2 KiB)

Tried the ppa to see if it fixed problem per recommendation of your
previous comments and the referenced similar bug.. Did not fix it.. And
upgrading to the 3.7 kernel also breaks all the wireless modules.. which
aren't seemingly updated yet..

On Wed, Jan 2, 2013 at 6:00 PM, Chauncellor <email address hidden>wrote:

> Please stop commenting on this bug report. Hundreds of people are
> subscribed and will be spammed with emails. If any support is needed
> please do go on askubuntu.com or the official forums/IRC channels. If
> you are experiencing this bug with all the updates, please follow my
> previous post (post # 325)
>
> @joshyg6, you may be using unofficial packages (PPA?).
>
> ** Description changed:
>
> + ****************ATTENTION*****************
> + THIS BUG HAS BEEN MARKED AS FIXED AND THE ISSUE HAS BEEN RESOLVED.
> + ANYONE EXPERIENCING THESE SYMPTOMS ON A COMPLETELY UP-TO-DATE SYSTEM
> SHOULD LOG A NEW BUG:
> +
> + 'ubuntu-bug xserver-xorg-video-intel'
> +
> + PLEASE REFRAIN FROM UNNECESSARY COMMENTS. THANKS!
> + ****************ATTENTION*****************
> +
> [Impact]:
>
> severe, compiz hangs frequently after resuming from suspend or the
> screensaver
>
> [Testcase]:
>
> - grab xserver-xorg-video-intel from quantal
> - build it from the extracted directory
> - after the build run 'cd test && make dri2-swap'
> - log in X
> - start two terminals
> - get dpms.sh from this bug, run it in the other terminal
> - on the other run ./dri2-swap
>
> compiz should hang during the first couple of runs of dri2-swap
>
> with the fixed package, it doesn't
>
> [Regression potential]
>
> To ease the backporting, I did replace patch 101 with the commit that
> was pulled upstream. This patch is needed for smooth transition from
> plymouth to X. There _could_ be some issues with it (meaning additional
> patches needed from upstream), but it seems to be working just fine for
> me.
>
> The precise upload has another upstream fix for an issue referenced in
> bug 992391, it should be safe.
>
> ---
>
> This bug is about problems after resume, with Intel video hardware and
> i965_dri.so. If you're having these symptoms with other hardware,
> Nvidia or ATI/Radeon, this isn't the place for reports or questions.
>
> [Those comments are building up, so brief links to direct other people
> to appropriate bugs might be good here, if anyone has them. Thanks.]
>
>
> ----------------------------------------------------------------------------------------
>
> This happens intermittently. Sometimes upon resume, all I see is a black
> screen and the cursor. the mouse and keyboard respond (the mouse moves)
> but nothing changes. Switching to console and back doesn't fix it.
> Killing Compiz also does not fix the issue. Further, if one has password
> lock disabled on resume the user will be stuck with a frozen display of
> whatever was last shown on the screen before suspend. The mouse cursor
> still changes as you hover over various elements but the display is
> frozen solid. One can switch to another TTY via ctrl+alt+F1 but killing
> the X session is the only way to get back...

Changed in xserver-xorg-video-intel (Ubuntu):
assignee: Timo Aaltonen (tjaalton) → Ivander Loe-A-Foe (pdsboss)
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Related questions

Remote bug watches

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