wakeup problem from suspend

Bug #242344 reported by Oguz Yarimtepe
2
Affects Status Importance Assigned to Milestone
acpi (Ubuntu)
Invalid
Undecided
Unassigned

Bug Description

I know this will be a general bug that was mentioned before but till now i haven't found any solutions from the reported bugs. So i will give detail information about the problem by hoping a solution or guidance.

I am using Ubuntu 8.04 on a laptop. When i choose suspend menu after clicking System->Quit, it goes to sleep mode. When i see the power led flashes, i press the power button and the laptop wakes up, i see my desktop and everything is fine. But when i try o suspend it again, it goes to sleep mode but never wakes up. The power led lights up continuously like it is waken up, but all i see is a black screen. So i have to hold the power button and power down the laptop and reopen it again to get my desktop. So briefly i have a problem from waking up from the second sleep mode.

To find a solution i used the sleep2.sh script to see the console outputs and simulate sleep steps. I also compiled the kernel with CONFIG_ACPI_DEBUG=y parameter to see detail acpi messages. So i increased the verbosity of kernel messages after i installed new compiled kernel:

# cd /proc/acpi
# echo -n "0xA0200097" > debul_level

I will add 3 kernel log files. One is the messages at the first boot of the laptop with new compiled kernel. Second one is after the run of sleep2.sh and 3rd one is after running the sleep2.sh again and powering down and up the laptop. I will also add lspci, dmidecode. acpi-support and xorg.conf files. Hope they will help to find a solution.

Revision history for this message
Oguz Yarimtepe (oguzy) wrote :
Revision history for this message
Oguz Yarimtepe (oguzy) wrote :

Kernel log at the first login.

Revision history for this message
Oguz Yarimtepe (oguzy) wrote :

Kernel log after running sleep2.sh

Revision history for this message
Oguz Yarimtepe (oguzy) wrote :

Kernel log after powering down

Revision history for this message
Oguz Yarimtepe (oguzy) wrote :

cat /proc/version_signature > version.log

Revision history for this message
Oguz Yarimtepe (oguzy) wrote :

cat /etc/lsb-release > lsb-release.log

Revision history for this message
Oguz Yarimtepe (oguzy) wrote :

lspci -vvnn > lspci-vvnn.log

Revision history for this message
Oguz Yarimtepe (oguzy) wrote :

dmidecode > dmidecode.txt

Revision history for this message
Oguz Yarimtepe (oguzy) wrote :

uname -a > uname-a.txt

Revision history for this message
Oguz Yarimtepe (oguzy) wrote :

/etc/default/acpi-support

Revision history for this message
Oguz Yarimtepe (oguzy) wrote :

xorg.conf

Revision history for this message
Oguz Yarimtepe (oguzy) wrote :

/boot/grub/menu.lst

As you can see i tried witl SLAB also but didnt get any result.

Revision history for this message
Oliver Grawert (ogra) wrote :

you should note that acpi-support is deprecated, you need to use /etc/pm/config.d/ ... try with adding a file there with the following line:

ADD_PARAMETERS="--quirk-vbemode-restore --quirk-s3-bios"

that helped on some intel cards i have seen in the past ... also if the screen comes up black, did you try hitting brightness up or down ? it could be that the backlight just didnt get switched on...

Revision history for this message
Oguz Yarimtepe (oguzy) wrote :

I tried pm-suspend from console with the --quirk-s3s -bios and it worked at the first run. After waking up from the sleep i run the pm-suspend command with the same option and this time it didn't wake again. And there was not any keyboard activity again, just a lighting led with a running fun. So there is something i am missing. When i tried the command at my second case i see the output saying the activation failure of i8042. After searching the google i decided to unbind the i8042 and bind it while waking up. I attached my /usr/lib/pm-utils/sleep.d/01unbound8042 script. This didn't solve the problem either.

I even tried to compile a new kernel with the i8042 module and unload it while sleeping and loading back while waking it up, but i lost my keyboard activity and problem still continued.

As a last case o tried the script here: http://launchpadlibrarian.net/7411308/suspend2ram

I was trying the suspend part and after the command echo -n "i8042" > /sys/bus/platform/drivers/i8042/unbindci lost my keyboard and touchpad activity. Running echo -n "i8042" > /sys/bus/platform/drivers/i8042/bind made both of them activated again.

Maybe this can give an idea to someone to solve my problem.

Revision history for this message
Oguz Yarimtepe (oguzy) wrote :

dmesg > dmesg.txt after running pm-suspend --quirk-s3-bios and coming back to sleep succesfully

Revision history for this message
Oguz Yarimtepe (oguzy) wrote :

dmesg > dmesg2.txt after running pm-suspend --quirk-s3-bios at a second time and after powering down and up

Revision history for this message
Oguz Yarimtepe (oguzy) wrote :

kern.log after running pm-suspend --quirk-s3-bios and coming back to sleep succesfully

Revision history for this message
Oguz Yarimtepe (oguzy) wrote :

kern2.log after running pm-suspend --quirk-s3-bios at a second time and after powering down and up

Revision history for this message
Oguz Yarimtepe (oguzy) wrote :

pm-suspend.log after running pm-suspend --quirk-s3-bios and coming back to sleep succesfully

Revision history for this message
Oguz Yarimtepe (oguzy) wrote :

pm-suspend2.log after running pm-suspend --quirk-s3-bios at a second time and after powering down and up

Revision history for this message
Oliver Grawert (ogra) wrote :

could you try with both parameters mentioned above, not only with half of the command ?
(pm-suspend --quirk-s3-bios --quirk-vbemode-restore)
also if there was any problem with the i8042 your first suspend wouldnt work, lets try to stay focused on the actual problem with this bug i.e. the video not coming back properly. if there are actual problems with i8042 please use a separate bugreport.

Revision history for this message
Phillip Susi (psusi) wrote :

It looks like this should have been marked as needs info but wasn't, and thus, should have expired long ago since it has not been updated. Thus, I will now close the report. If you can provide the requested information, feel free to reopen it.

Changed in acpi (Ubuntu):
status: New → Invalid
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.