You are not logged in.

#1 2012-10-17 20:46:15

ViPa
Member
From: Germany
Registered: 2012-10-17
Posts: 2

Flickering display on intel graphics after resume from hibernation

After latest kernel upgrade (to 3.6.2), when resuming from hibernation, the display is flickering.
Additionally, the /var/log/messages.log say (at time of resume):

  Oct 17 06:29:09 localhost kernel: [ 5642.562850] ------------[ cut here ]------------
  Oct 17 06:29:09 localhost kernel: [ 5642.562867] WARNING: at drivers/gpu/drm/i915/intel_display.c:1225 assert_pipe+0x8e/0x90 [i915]()
  Oct 17 06:29:09 localhost kernel: [ 5642.562868] Hardware name: K50IJ
  Oct 17 06:29:09 localhost kernel: [ 5642.562870] pipe B assertion failure (expected off, current on)
  Oct 17 06:29:09 localhost kernel: [ 5642.562914] Modules linked in: <list removed>
  Oct 17 06:29:09 localhost kernel: [ 5642.562916] Pid: 9770, comm: kworker/u:5 Not tainted 3.6.2-1-ARCH #1
  Oct 17 06:29:09 localhost kernel: [ 5642.562917] Call Trace:
  Oct 17 06:29:09 localhost kernel: [ 5642.562924]  [<ffffffff81055c2f>] warn_slowpath_common+0x7f/0xc0
  Oct 17 06:29:09 localhost kernel: [ 5642.562927]  [<ffffffff81055d26>] warn_slowpath_fmt+0x46/0x50
  Oct 17 06:29:09 localhost kernel: [ 5642.562935]  [<ffffffffa0409cfe>] assert_pipe+0x8e/0x90 [i915]
  Oct 17 06:29:09 localhost kernel: [ 5642.562943]  [<ffffffffa040f319>] intel_crtc_disable+0x69/0xb0 [i915]
  Oct 17 06:29:09 localhost kernel: [ 5642.562948]  [<ffffffffa01d8885>] drm_helper_disable_unused_functions+0x115/0x170 [drm_kms_helper]
  Oct 17 06:29:09 localhost kernel: [ 5642.562951]  [<ffffffffa01d9370>] drm_helper_resume_force_mode+0x130/0x170 [drm_kms_helper]
  Oct 17 06:29:09 localhost kernel: [ 5642.562957]  [<ffffffffa03e3200>] i915_drm_thaw+0xd0/0x150 [i915]
  Oct 17 06:29:09 localhost kernel: [ 5642.562963]  [<ffffffffa03e35ce>] i915_resume+0x4e/0x80 [i915]
  Oct 17 06:29:09 localhost kernel: [ 5642.562969]  [<ffffffffa03e3616>] i915_pm_resume+0x16/0x20 [i915]
  Oct 17 06:29:09 localhost kernel: [ 5642.562972]  [<ffffffff8127ba53>] pci_pm_restore+0x73/0xd0
  Oct 17 06:29:09 localhost kernel: [ 5642.562974]  [<ffffffff8127b9e0>] ? pci_pm_suspend_noirq+0x1a0/0x1a0
  Oct 17 06:29:09 localhost kernel: [ 5642.562977]  [<ffffffff8132e91b>] dpm_run_callback.isra.4+0x3b/0x70
  Oct 17 06:29:09 localhost kernel: [ 5642.562979]  [<ffffffff8132f638>] device_resume+0xa8/0x160
  Oct 17 06:29:09 localhost kernel: [ 5642.562981]  [<ffffffff8132f711>] async_resume+0x21/0x50
  Oct 17 06:29:09 localhost kernel: [ 5642.562984]  [<ffffffff8108149e>] async_run_entry_fn+0x8e/0x1b0
  Oct 17 06:29:09 localhost kernel: [ 5642.562987]  [<ffffffff810735c9>] process_one_work+0x139/0x500
  Oct 17 06:29:09 localhost kernel: [ 5642.562990]  [<ffffffff81485fee>] ? mutex_unlock+0xe/0x10
  Oct 17 06:29:09 localhost kernel: [ 5642.562993]  [<ffffffff81081410>] ? async_schedule+0x20/0x20
  Oct 17 06:29:09 localhost kernel: [ 5642.562995]  [<ffffffff81073d5e>] worker_thread+0x15e/0x460
  Oct 17 06:29:09 localhost kernel: [ 5642.562997]  [<ffffffff81487789>] ? preempt_schedule+0x49/0x70
  Oct 17 06:29:09 localhost kernel: [ 5642.562999]  [<ffffffff81073c00>] ? rescuer_thread+0x230/0x230
  Oct 17 06:29:09 localhost kernel: [ 5642.563002]  [<ffffffff81079693>] kthread+0x93/0xa0
  Oct 17 06:29:09 localhost kernel: [ 5642.563005]  [<ffffffff81490dc4>] kernel_thread_helper+0x4/0x10
  Oct 17 06:29:09 localhost kernel: [ 5642.563007]  [<ffffffff81079600>] ? kthread_freezable_should_stop+0x70/0x70
  Oct 17 06:29:09 localhost kernel: [ 5642.563009]  [<ffffffff81490dc0>] ? gs_change+0x13/0x13
  Oct 17 06:29:09 localhost kernel: [ 5642.563010] ---[ end trace 41226cd292a0cffc ]---
  Oct 17 06:29:09 localhost kernel: [ 5642.566182] PM: restore of devices complete after 791.242 msecs
  Oct 17 06:29:09 localhost kernel: [ 5642.566579] Restarting tasks ... done.
  Oct 17 06:29:09 localhost kernel: [ 5642.575410] video LNXVIDEO:00: Restoring backlight state
  Oct 17 06:29:09 localhost kernel: [ 5642.586217] ------------[ cut here ]------------


This warning is repeated several times.

Offline

#2 2012-10-20 07:13:04

cmcc
Member
Registered: 2012-10-20
Posts: 2

Re: Flickering display on intel graphics after resume from hibernation

I had the same issue, it seems to disappear if using the new SNA acceleration method.
I created the file /etc/X11/xorg.conf.d/20-intel.conf with the following lines

Section "Device"
   Identifier  "Intel Graphics"
   Driver      "intel"
   Option      "AccelMethod"  "sna"
EndSection

rebooted and it's resuming correctly from hibernation again.
Hope this helps!

Offline

#3 2012-10-20 07:44:50

WonderWoofy
Member
From: Los Gatos, CA
Registered: 2012-05-19
Posts: 8,412

Re: Flickering display on intel graphics after resume from hibernation

I think using sna is only wise if you don't have old hardware though.  Not saying it is not worth a try, but you didn't even mention what specific hardware you were using.

Offline

#4 2012-10-20 22:07:02

ViPa
Member
From: Germany
Registered: 2012-10-17
Posts: 2

Re: Flickering display on intel graphics after resume from hibernation

I also thougt  that SNA is not relevant for my approx. 4 years old hardware, however this Gentoo page claims that SNA is also useful for chipsets other than SandyBridge.

I have plain old GM45, so I'm not sure. In the meantime I've also filed a bug at freedesktop.org.

Strangely enough, I currently do not see anymore these artefacts, but the WARNINGs in dmesg still exist after resume from hibernate.

Offline

#5 2012-10-21 09:56:12

cmcc
Member
Registered: 2012-10-20
Posts: 2

Re: Flickering display on intel graphics after resume from hibernation

Yes, I didn't specify my hardware, sorry: it's an Intel GM45 integrated chipset, using the i915 kernel module.
Anyway, yesterday, after the latest upgrade to extra/xf86-video-intel-2.20.10-1 and extra/xorg-server-1.13.0-3 and reverting to the old UXA acceleration method, the issue for my card is fixed.

Offline

#6 2012-12-28 04:42:28

girzel
Member
Registered: 2010-06-27
Posts: 42

Re: Flickering display on intel graphics after resume from hibernation

I've had the exact same screen flickering and dmesg error as OP for quite a while now. I have the file /etc/X11/xorg.conf.d/20-intel.conf, but previously there was no AccelMethod specified in there at all. After finding this thread I tried specifying both "uxa" and "sna", but had the same problem each time (though the pattern of the flickering did change). It's still doing it now, though I noticed that after I resume from hibernation and get the flickering, I can then suspend to RAM and resume again, and the problem goes away.

All my packages are up to date. lspci reports a "Intel Corporation Mobile 4 Series Chipset Integrated Graphics Controller (rev 07)", using the i915 driver. Is there anything else I can try?

Thanks!

Offline

Board footer

Powered by FluxBB