Black screen after suspend with i915

[Tero Karvinen]

Binary package hint: acpi-support

After waking up from suspend-to-ram, screen stays black.

To reproduce,
1) Right click gnome-power-manager panel applet, choose suspend.
- Computer suspends, fan shuts down
2) Click a key to wake up.
What happens
- Screen stays black. Everything else seems to work: X (blindly), sounds, ssh, network, backlight.
What should happen.
- Computer should wake up to the state it was in before suspend. There should be a picture on the screen.

I am using Acer Travelmate 3004WTMi with 915 graphics controller, Dapper, acpi-support 0.82, Linux 2.6.15-22-386. Command outputs, including lscpi and dmidecode:
http://www.iki.fi/karvinen/acer_travelmate_3004wtmi_command_outputs.html

Complete freeze after suspend (bug #33372) seems to have been fixed. i810 bug #28326 was a bit like this one, but in that bug I was asked to open a new one against acpi-support.

[Luka Renko (lure)]

Does closing and reopening the lid helps? I have similar problem on HP nw8240 (see bug 33827) and close/open lid solves it for me...

[Tero Karvinen (karvinen+launchpad)]

Closing and reopening the lid does not help. Thanks for the tip anyway.

[Paul Sladen (sladen)]

Does switching aware to a console and back help:

  Ctrl-Alt-F1
  Alt-F7

do you see anything on the console, or is that black too?

[Tero Karvinen (karvinen+launchpad)]

Switching virtual consoles (alt-ctrl-f1) does not help. It blinks the screen, though. Console is completely black just like X.

I have also tried 'sudo vbetool post', but it did not help.

[Bernhard J. M. Grün (bernhard-gruen)]

I have the same problem with a notebook from MSI called S260. At least for me I am quite sure that the problem comes from the vga bios. The vga bios does not have a native mode for the 1280x800 resolution used on that notebook. Before the release of X.org 6.9.0 I had to use 915resolution to fix that problem. With 6.9/7.0 this problem is solved by a patch that was included into the mainstream.

[Tero Karvinen (karvinen+launchpad)]

The screen stays black after suspend even if 915resolution has not been loaded. Thus, the problem is not related to 915resolution.

Most likely BIOS is buggy. However, I think it's quite difficult to find a laptop BIOS that is not full of bugs, so OS should just work around them. Suspend works on Windows with the same hardware.

[Marcel Saatkamp (marcel-saatkamp)]

I have a similar bug. Acer Aspire 5652 WLMi
https://launchpad.net/distros/ubuntu/+bug/46292

[neurol23 (neurol23)]

I have the same problem on Toshiba SA50-111. Like in Bug #46677 X server appears to restart several times (mouse pointer or "busy wheel" appear briefly multiple times) before finally hanging with a blank screen. This doesn't happen every time my notebook wakes from suspend to ram, only just 1 in 3-4 times. Only thing that helps, is rebooting the computer to "safe mode" and then rebooting back to normal mode.

I looked to the logs for gdm and each has something like this:

X Window System Version 7.0.0
Release Date: 21 December 2005
X Protocol Version 11, Revision 0, Release 7.0
Build Operating System:Linux 2.6.12 i686
Current Operating System: Linux notebook-zbynek 2.6.15-23-386 #1 PREEMPT Tue May 23 13:49:40 UTC 2006 i686
Build Date: 16 March 2006
 Before reporting problems, check http://wiki.x.org
 to make sure that you have the latest version.
Module Loader present
Markers: (--) probed, (**) from config file, (==) default setting,
 (++) from command line, (!!) notice, (II) informational,
 (WW) warning, (EE) error, (NI) not implemented, (??) unknown.
(==) Log file: "/var/log/Xorg.0.log", Time: Wed May 31 00:03:59 2006
(==) Using config file: "/etc/X11/xorg.conf"
ADVANCE_LP_RING: outring (0x9) isn't on a QWord boundaryerror opening security policy file /etc/X11/xserver/SecurityPolicy
Synaptics DeviceInit called
SynapticsCtrl called.
Synaptics DeviceOn called
ADVANCE_LP_RING: outring (0x31) isn't on a QWord boundaryADVANCE_LP_RING: outring (0x49) isn't on a QWord boundaryADVANCE_LP_RING: outring (0x51) isn't on a QWord boundaryError in I830WaitLpRing(), now is -2024591521, start is -2024593522
pgetbl_ctl: 0x1ffe0001 pgetbl_err: 0x0
ipeir: 0 iphdr: c000000
LP ring tail: 50 head: 0 len: 1f001 start 0
eir: 0 esr: 1 emr: ffff
instdone: ffc1 instpm: 40
memmode: 108 instps: 20
hwstam: ffff ier: 82 imr: 9 iir: 20
space: 130983 wanted 131064

Fatal server error:
lockup

ADVANCE_LP_RING: outring (0x59) isn't on a QWord boundaryError in I830WaitLpRing(), now is -2024589520, start is -2024591521
pgetbl_ctl: 0x1ffe0001 pgetbl_err: 0x0
ipeir: 0 iphdr: c000000
LP ring tail: 58 head: 0 len: 1f001 start 0
eir: 0 esr: 1 emr: ffff
instdone: ffc1 instpm: 40
memmode: 108 instps: 20
hwstam: ffff ier: 0 imr: ffff iir: 0
space: 130975 wanted 131064

FatalError re-entered, aborting
lockup

Hope it helps, this error is really annoying.

[neurol23 (neurol23)]

seems solved for me so far

what helped me, was adding the boot parameter "acpi_sleep=s3_bios,s3_mode" in the grub. as far as i know in some computers only s3_bios or s3_mode is neeeded (not both)

[Tero Karvinen (karvinen+launchpad)]

These boot parameters don't work:
acpi_sleep=s3_bios,s3_mode (complete freeze on wakeup)
acpi_sleep=s3_bios (complete freeze on wakeup)
acpi_sleep=s3_mode (no effect)
Tested on Acer Travelmate 3000, suspend-to-ram and Dapper updated 2006-06-01.

[Tero Karvinen (karvinen+launchpad)]

I can get the picture back, but X usually crashes with this:
1) sudo vbetool vbestate save > terosvbestate
User is often logged out and returned to X login screen. Sometimes X stops responding to mouse events.
2) Gnome panel: Power Manager: Suspend
Computer goes to suspend-to-ram, fan shuts down.
3) Press a key to wake up.
Computer wakes up. There is a correct picture on screen!
Typically, user is logged out or X is frozen. Sometimes it is possible to log in after suspend. I commented out all other display modes except one in xorg.conf:
        SubSection "Display"
                Viewport 0 0
                Depth 24
                Modes "1280x800"
        EndSubSection

Tested on Dapper with latest updates on 2006-06-03, using 915resolution 0.5-1ubuntu6, xserver-xorg 7.0.0-0ubuntu45, linux 2.6.15-23-686 on Acer Travelmate 3000 (3004WTMi).

[Tero Karvinen (karvinen+launchpad)]

Corrected version with restoring vbestate:
I can get the picture back, but X usually crashes with this:
1) sudo vbetool vbestate save > terosvbestate
User is often logged out and returned to X login screen. Sometimes X stops responding to mouse events.
2) Gnome panel: Power Manager: Suspend
Computer goes to suspend-to-ram, fan shuts down.
3) Press a key to wake up.
Computer wakes up.
4) $ sudo vbetool vbestate restore < terosvbestate
"Function not supported"
There is a correct picture on screen!

[Tero Karvinen (karvinen+launchpad)]

Bypass for bug: display works after suspend if
Option "VBERestore" "true"
is set and user changes virtual console back and foth. This works with 915resolution too.
http://www.iki.fi/karvinen/acer_travelmate_3004wtmi_with_linux.html#suspend-to-ram_s3

Hope this setting gets included in Dapper updates.

[disposable (disposable)]

the only thing that helped me was using old (ubuntu 5.10) /etc/acpi/ in 6.06. if anyone here's desperate and has no backup of old /etc/acpi, feel free to use mine. http://acpi.logrus.co.uk/acpi.tar.bz2

[robobart (snapp)]

Similar problem on a ThinkPad T23.

Usually I can get the backlight come back on once. Then every suspend after that I need to play around quite a bit to get it to come on.

[Martin Bergner (martin-bergner)]

Does leaving everything like it was and changing SAVE_VBE_STATE to false in /etc/default/acpi-support, solve the issue?

[Tero Karvinen (karvinen+launchpad)]

My Acer Travelmate 3004WTMi machine is allready in use, so it's not possible for me to install from scratch or remove all settings to try that out. Looking at my notes, I could guess it does not help as manually restoring vbe state was once a great step forward towards working suspend. I'll try it out if can get a suitable test machine.

Can someone else with similar symptoms try it out?

[Jorge Juan (jjchico)]

Solution found!

My laptop is a Fujitsu-Siemens S7020 with a i915 chipset with the same problem. Before, I tried playing with options in /etc/default/acpi-support and /etc/X11/xorg.conf with Dapper and pre-release Edgy with no success.

Today I got it working by commenting out the SAVE_VIDEO_PCI_STATE=true option in /etc/default/acpi-support. Everything else have default values as provided by Edgy.

Changing SAVE_VBE_STATE to false, as suggested in a previous post did not helped.

I hope it can help other users.

[GioSico (john-giosico)]

I am running feisty on an HP Pavilion dv6000 and this is already commented out by default and I still get a black screen after suspend resumes. I am still searching for a solution.

[Tero Karvinen (karvinen+launchpad)]

I've been using a working bypass for some time already. Quoting myself on http://www.iki.fi/karvinen/acer_travelmate_3004wtmi_with_linux.html#suspend-to-ram_s3
> Add this to /etc/X11/xorg.conf:
> Section "Device"
> #...
> Option "VBERestore" "true"
> EndSection
>Picture comes up after changing virtual terminals back and forth, but computer can do this automatically. In /etc/default/acpi-support
> DOUBLE_CONSOLE_SWITCH=true
> Suspend by right clicking battery image on panel and choosing suspend. Display works now. Also works with 915resolution.

[Launchpad Janitor (janitor)]

[Expired for acpi-support (Ubuntu) because there has been no activity for 60 days.]

[Jorge Juan (jjchico)]

My laptop (Fujitsu-Siemens described above) now works OK with the default /etc/default/acpi-support in Gutsy.

[Benoit Heurter (benoit-heurter)]

mine not !

Acer Travelmate 8002 LMIB. default acpi-support from gutsy do not work

enable SAVE_VIDEO_PCI_STATE=true neither

[Inigo Montoya (a9601252)]

I had the same problem an a HP nc2400 laptop. With Kubuntu 7.04 i never got supsend to ram work properly. The system always stuck with a black screen after wakeup. Switching to 7.10 with the original 7.10 /etc/default/acpi-support file didn't help. But thanks to this bugreport i tried the DOUBLE_CONSOLE_SWITCH=true option and now it works. Thats so über-cool!! Thanks!!!

[Habutre (habutre)]

For me the ACPI_SLEEP_MODE variable solved the problem. when I leave my laptop for a many time inactive then blank screen came on, then nothing more works, only the power button :(. But changing the value of mem to standby works for me.
My laptop wake-up without problem.

[Aglari (aglari)]

Setting "VBIRestore" to true in xorg.conf and "DOUBLE_CONSOLE_SWITCH" to true in acpi-support worked for me, as Tero Karvinen posted in comment #20. Thanks.