[arch-general] intel video & suspend
Hi folks, Regularly I put my ThinkPad R61 into suspend twice a day, and everything worked perfectly until 2.6.32 and 2.6.33 kernels. Now from time to time (≈ once a week) when I wake it up, the screen remains black in both X and text console. Nothing can bring it back, but everything else works just fine. No errors in logs either. The suspend is done with: # echo -n mem >/sys/power/state (just the stock /etc/acpi/handler.sh of acpid package) I have Intel GM965 card and run xorg-server-1.7.6-3. Driver: xf86-video-intel-2.10.0-1 I found that there are a couple of patches [1,2] for i915 and intel bios were made in Nov-Dec 2009 to fix such problems, and they were supposed to be in 2.6.33, but my problem persists. Does anybody have ideas/info? Thanks, Sergey [1] http://bbs.archlinux.org/viewtopic.php?id=87828 [2] https://patchwork.kernel.org/patch/62734/
On 05/06/10 11:49, Sergey Manucharian wrote:
Hi folks,
Regularly I put my ThinkPad R61 into suspend twice a day, and everything worked perfectly until 2.6.32 and 2.6.33 kernels.
Now from time to time (≈ once a week) when I wake it up, the screen remains black in both X and text console. Nothing can bring it back,
What about suspending again, perhaps waiting a few seconds, and resuming again? Does that have a chance to help? (It helps me on my Intel card's graphic glitches sometimes. I can do it easily because closing the laptop lid causes suspend, and opening it causes wake-up...)
Excerpts from Isaac Dupree's message of Thu, 06 May 2010 14:32 -0400:
On 05/06/10 11:49, Sergey Manucharian wrote:
Hi folks,
Regularly I put my ThinkPad R61 into suspend twice a day, and everything worked perfectly until 2.6.32 and 2.6.33 kernels.
Now from time to time (≈ once a week) when I wake it up, the screen remains black in both X and text console. Nothing can bring it back,
What about suspending again, perhaps waiting a few seconds, and resuming again? Does that have a chance to help? (It helps me on my Intel card's graphic glitches sometimes. I can do it easily because closing the laptop lid causes suspend, and opening it causes wake-up...)
I tried that too with no success, next time will try to suspend again for longer time... The bad thing is that it may work properly next 50 times - it's not easily reproducible. Sergey
On Thu, May 06, 2010 at 01:15:53PM -0600, Sergey Manucharian wrote:
Excerpts from Isaac Dupree's message of Thu, 06 May 2010 14:32 -0400:
On 05/06/10 11:49, Sergey Manucharian wrote:
Hi folks,
Regularly I put my ThinkPad R61 into suspend twice a day, and everything worked perfectly until 2.6.32 and 2.6.33 kernels.
Now from time to time (≈ once a week) when I wake it up, the screen remains black in both X and text console. Nothing can bring it back,
What about suspending again, perhaps waiting a few seconds, and resuming again? Does that have a chance to help? (It helps me on my Intel card's graphic glitches sometimes. I can do it easily because closing the laptop lid causes suspend, and opening it causes wake-up...)
I tried that too with no success, next time will try to suspend again for longer time... The bad thing is that it may work properly next 50 times - it's not easily reproducible.
Sergey
I use pm-suspend and never have some problems.
Excerpts from Nick Stepa's message of Fri, 07 May 2010 08:40 +0300:
On Thu, May 06, 2010 at 01:15:53PM -0600, Sergey Manucharian wrote:
Excerpts from Isaac Dupree's message of Thu, 06 May 2010 14:32 -0400:
On 05/06/10 11:49, Sergey Manucharian wrote:
Hi folks,
Regularly I put my ThinkPad R61 into suspend twice a day, and everything worked perfectly until 2.6.32 and 2.6.33 kernels.
Now from time to time (≈ once a week) when I wake it up, the screen remains black in both X and text console. Nothing can bring it back,
What about suspending again, perhaps waiting a few seconds, and resuming again? Does that have a chance to help? (It helps me on my Intel card's graphic glitches sometimes. I can do it easily because closing the laptop lid causes suspend, and opening it causes wake-up...)
I tried that too with no success, next time will try to suspend again for longer time... The bad thing is that it may work properly next 50 times - it's not easily reproducible.
Sergey
I use pm-suspend and never have some problems.
I started using pm-suspend after this advice, and indeed it haven't happened until today (2 weeks). I tried to suspend it again (and bring back) with various pm-suspend quirks, but nothing helped - I just rebooted to get my screen back. I understand that the using of quirks may not help if it already failed and fell into a bad mode. But anyway, does anybody use those quirks with Intel GM965 card? Cheers, Sergey
oh by the way, I started using 2.6.34 kernel and haven't personally had any graphical issues since (it's in Testing -- make sure to get both kernel and firmware, but AFAICT it works fine to download/install those two packages without using any other part of Testing, at least this time around) -Isaac
Now from time to time (≈ once a week) when I wake it up, the screen remains black in both X and text console. Nothing can bring it back,
What about suspending again, perhaps waiting a few seconds, and resuming again? Does that have a chance to help? (It helps me on my Intel card's graphic glitches sometimes. I can do it easily because closing the laptop lid causes suspend, and opening it causes wake-up...)
I have a Pineview Integreted Controller (i915) in my eeepc and this happened to me twice (in the last month) after closing the lid, but I just issue "xset dpms force off" without suspending. No issues with coming up from suspend so far. kernel26 2.6.33.2-1 xorg-server 1.7.6-3 xf86-video-intel 2.10.0-1 Bye, Christoph
On 05/07/10 04:13, Christoph Rissner wrote:
Now from time to time (≈ once a week) when I wake it up, the screen remains black in both X and text console. Nothing can bring it back,
What about suspending again, perhaps waiting a few seconds, and resuming again? Does that have a chance to help? (It helps me on my Intel card's graphic glitches sometimes. I can do it easily because closing the laptop lid causes suspend, and opening it causes wake-up...)
I have a Pineview Integreted Controller (i915) in my eeepc and this happened to me twice (in the last month) after closing the lid, but I just issue "xset dpms force off" without suspending.
please inform us, what does "xset dpms force off" do? And for what circumstances/purpose do you use it? I looked in the man-page of 'xset' and found it, but I wasn't sure enough to risk using it in case it did something I didn't know how to get out of* *like blanking my screen (see "off" + mentioning suspending), or like turning on "DPMS Energy Star Features" (see man-page, "Setting these values implicitly enables the DPMS features."), or something. Please tell us :-) and did you mean -it's a preventative measure to screen-glitches? -it's a remedy for screen glitches after they happen? -it's an alternative to suspending for the purpose of saving power? -... sorry :0) -Isaac
On 05/07/2010 07:16 PM, Isaac Dupree wrote:
please inform us, what does "xset dpms force off" do? And for what circumstances/purpose do you use it?
Sorry it took so long, I missed your reply. "xset dpms force off" is supposed to turn off the screen immediately using DPMS, at least thats what it seems to do. I just don't want to suspend (suspend the whole system to RAM) everytime I close the lid, rather I like it to continue to work on whatever I told it to do :-) Bye Christoph
On 18 June 2010 15:36, Christoph Rissner <c.r@visotech.at> wrote:
On 05/07/2010 07:16 PM, Isaac Dupree wrote:
please inform us, what does "xset dpms force off" do? And for what circumstances/purpose do you use it?
Sorry it took so long, I missed your reply.
"xset dpms force off" is supposed to turn off the screen immediately using DPMS, at least thats what it seems to do. I just don't want to suspend (suspend the whole system to RAM) everytime I close the lid, rather I like it to continue to work on whatever I told it to do :-)
That _is_ the default behaviour - the laptop should do nothing at all when the lid is closed. I know of no particular hardware/BIOS that sets a sleep action upon lid closure. It is only altered by userspace tools/daemons like for eg. when using a DE. You might want to check what else is enforcing a rule to the lid-closing state. -- GPG/PGP ID: B42DDCAD
Excerpts from Ray Rashif's message of 2010-06-18 09:54:55 +0200:
On 18 June 2010 15:36, Christoph Rissner <c.r@visotech.at> wrote:
On 05/07/2010 07:16 PM, Isaac Dupree wrote:
please inform us, what does "xset dpms force off" do? And for what circumstances/purpose do you use it?
Sorry it took so long, I missed your reply.
"xset dpms force off" is supposed to turn off the screen immediately using DPMS, at least thats what it seems to do. I just don't want to suspend (suspend the whole system to RAM) everytime I close the lid, rather I like it to continue to work on whatever I told it to do :-)
That _is_ the default behaviour - the laptop should do nothing at all when the lid is closed. I know of no particular hardware/BIOS that sets a sleep action upon lid closure. It is only altered by userspace tools/daemons like for eg. when using a DE. You might want to check what else is enforcing a rule to the lid-closing state.
Mine does turn off the backlight, and I'm quite sure there's no userspace involved, just plain hardware/bios. No idea whether others do more than that. Turning off the backlight seems like reasonably safe thing to do. -- Regards, Philipp -- "Wir stehen selbst enttäuscht und sehn betroffen / Den Vorhang zu und alle Fragen offen." Bertolt Brecht, Der gute Mensch von Sezuan
On 18 June 2010 16:07, Philipp Überbacher <hollunder@lavabit.com> wrote:
Excerpts from Ray Rashif's message of 2010-06-18 09:54:55 +0200:
On 18 June 2010 15:36, Christoph Rissner <c.r@visotech.at> wrote:
On 05/07/2010 07:16 PM, Isaac Dupree wrote:
please inform us, what does "xset dpms force off" do? And for what circumstances/purpose do you use it?
Sorry it took so long, I missed your reply.
"xset dpms force off" is supposed to turn off the screen immediately using DPMS, at least thats what it seems to do. I just don't want to suspend (suspend the whole system to RAM) everytime I close the lid, rather I like it to continue to work on whatever I told it to do :-)
That _is_ the default behaviour - the laptop should do nothing at all when the lid is closed. I know of no particular hardware/BIOS that sets a sleep action upon lid closure. It is only altered by userspace tools/daemons like for eg. when using a DE. You might want to check what else is enforcing a rule to the lid-closing state.
Mine does turn off the backlight, and I'm quite sure there's no userspace involved, just plain hardware/bios. No idea whether others do more than that. Turning off the backlight seems like reasonably safe thing to do.
Yes, what I meant to say was that most laptop BIOS's switch off the monitor by default, so you shouldn't need to do any fiddling around other than when a possible power management tool has interrupted this behaviour. Except for very old models that do not recognise lid closure as an ACPI event. -- GPG/PGP ID: B42DDCAD
On 06/18/2010 09:54 AM, Ray Rashif wrote:
That_is_ the default behaviour - the laptop should do nothing at all when the lid is closed. I know of no particular hardware/BIOS that sets a sleep action upon lid closure. It is only altered by userspace tools/daemons like for eg. when using a DE.
I know, in my case its acpi-eeepc-generic that, by default, suspends the system when closing the lid. But thats not the point. What I tried to say is, I do have an Intel video card and experienced problems as mentioned in the original post. I didn't have to suspend the system but just close the screen lid (turning it off), then after opening the lid again the screen would remain blank/off occasionally, no matter what I tried. However, it didn't happen to me in the last month, IIRC there was a kernel update, I think this has resolved my issues. Bye, Christoph
On 18 June 2010 16:12, Christoph Rissner <c.r@visotech.at> wrote:
On 06/18/2010 09:54 AM, Ray Rashif wrote:
That_is_ the default behaviour - the laptop should do nothing at all when the lid is closed. I know of no particular hardware/BIOS that sets a sleep action upon lid closure. It is only altered by userspace tools/daemons like for eg. when using a DE.
I know, in my case its acpi-eeepc-generic that, by default, suspends the system when closing the lid.
But thats not the point. What I tried to say is, I do have an Intel video card and experienced problems as mentioned in the original post. I didn't have to suspend the system but just close the screen lid (turning it off), then after opening the lid again the screen would remain blank/off occasionally, no matter what I tried.
However, it didn't happen to me in the last month, IIRC there was a kernel update, I think this has resolved my issues.
Ahh, alright. I myself have last faced it last year, and I did find it very strange that such a hardware-level feature could have anything to do with the OS. I have never bothered to figure it out, it just works now. -- GPG/PGP ID: B42DDCAD
Excerpts from Christoph Rissner's message of 2010-06-18 09:36:27 +0200:
On 05/07/2010 07:16 PM, Isaac Dupree wrote:
please inform us, what does "xset dpms force off" do? And for what circumstances/purpose do you use it?
Sorry it took so long, I missed your reply.
"xset dpms force off" is supposed to turn off the screen immediately using DPMS, at least thats what it seems to do. I just don't want to suspend (suspend the whole system to RAM) everytime I close the lid, rather I like it to continue to work on whatever I told it to do :-)
Bye Christoph
The syntax of that stuff is weird, but here's what I have, seems to work: # turn off screen powersaving setterm -blank 0 -powersave off -powerdown 0 #turn powersave off xset dpms 0 0 0 xset -dpms #turn screensaver off xset s 0 0 xset s noblank xset s noexpose xset s off -- Regards, Philipp -- "Wir stehen selbst enttäuscht und sehn betroffen / Den Vorhang zu und alle Fragen offen." Bertolt Brecht, Der gute Mensch von Sezuan
On Thu, May 6, 2010 at 21:32, Isaac Dupree <ml@isaac.cedarswampstudios.org> wrote:
On 05/06/10 11:49, Sergey Manucharian wrote:
Hi folks,
Regularly I put my ThinkPad R61 into suspend twice a day, and everything worked perfectly until 2.6.32 and 2.6.33 kernels.
Now from time to time (≈ once a week) when I wake it up, the screen remains black in both X and text console. Nothing can bring it back,
What about suspending again, perhaps waiting a few seconds, and resuming again? Does that have a chance to help? (It helps me on my Intel card's graphic glitches sometimes. I can do it easily because closing the laptop lid causes suspend, and opening it causes wake-up...)
Hello all! I had (and still have) the same problem (almost) with my Benq Joybook S32. It has an Intel GM965. But my problem is a little bit more different: * sometimes after a resume the X remains pitch black (I guess it doesn't exit from the DPMS mode), and to "help" it I switch to VT1 and then back to the VT12 (where I have my X configured); * but on the other hand if before I suspend my laptop I have connected a monitor (on VGA), and on resume the monitor is not connected any more (or the other way around, no monitor at suspend, but monitor present at resume), my monitor goes again pitch black, but this time nothing helps except another suspend and resume. Till this day I haven't figured a way to solve the problem, but I didn't investigated it more in depth. Did I mention that I use the mod_setting? Maybe it happens only in this case? I hope that at the end of this thread we solve the problem. :) Ciprian.
Now from time to time (≈ once a week) when I wake it up, the screen remains black in both X and text console. Nothing can bring it back, but everything else works just fine. No errors in logs either.
I am having the same problem on my Fujitsu T900, 64-bit ArchLinux. Tried both pm-suspend and s2ram, and while s2ram *appeared* to not cause this problem as much, it still happened. At first I thought my screen was entirely blank, but on closer inspection realized my screen was simply very dim--flashing a flashlight at the screen helped a lot. Was able to navigate to my brightness settings, and the brightness was on max, as it was before suspend. I have the testing repo enabled, and everything is up-to-date. I don't know if this problem occurred with older kernel versions on the Fujitsu, as this is a recent install. Sara
On Wed, May 26, 2010 at 4:55 AM, Sara Fauzia <sara@archlinux.us> wrote:
Now from time to time (≈ once a week) when I wake it up, the screen remains black in both X and text console. Nothing can bring it back, but everything else works just fine. No errors in logs either.
I am having the same problem on my Fujitsu T900, 64-bit ArchLinux. Tried both pm-suspend and s2ram, and while s2ram *appeared* to not cause this problem as much, it still happened. At first I thought my screen was entirely blank, but on closer inspection realized my screen was simply very dim--flashing a flashlight at the screen helped a lot. Was able to navigate to my brightness settings, and the brightness was on max, as it was before suspend. I have the testing repo enabled, and everything is up-to-date. I don't know if this problem occurred with older kernel versions on the Fujitsu, as this is a recent install.
Sara
You should try looking into vbetool. It can turn the backlight on if necessary. -- Alexander Lam
participants (9)
-
Alexander Lam
-
Christoph Rissner
-
Ciprian Dorin, Craciun
-
Isaac Dupree
-
Nick Stepa
-
Philipp Überbacher
-
Ray Rashif
-
Sara Fauzia
-
Sergey Manucharian