[hardy] regression: suspend fails to resume (with or without restricted drivers installed)

dmesg is attached. There are no useful messages that I can tell written to the log.

same here, feisty worked fine, gutsy causes same symptoms, with or without composite desktop running. dell xps m1210, gutsy beta installed off cd (reformatted / and not /home partitions), nvidia binary driver.

Setting as confirmed as this bug has been seconded.

Adding lspci -v -v -n -n output.

Followed these instructions: https://wiki.ubuntu.com/DebuggingKernelSuspend

After running

sync; echo 1 > /sys/power/pm_trace; /etc/acpi/sleep.sh force

same symptoms occur (suspend works, resume fails). Did hard reboot after resume failure, then did

dmesg > dmesg_pmtrace.txt

which is attached. The only named hash match after resume.c is "device ptytd." What is that? Some sort of terminal? It's not a named kernel module.

Just tried the update to the -13 kernel. Symptoms persist: suspend successfully, does not resume. I am attaching a dmesg_pmtrace.txt from the -13 kernel now.

I have the exact same problem and also the exact same result from debugging. Tested with Kubuntu 7.10 beta, with kernels 2.6.22-12-generic and 2.6.22-13-generic both with and without restricted graphics drivers. Here is the output of the debug:
[ 13.981806] Magic number: 0:35:872
[ 13.981808] hash matches /build/buildd/linux-source-2.6.22-2.6.22/drivers/base/power/resume.c:57
[ 13.981848] hash matches device ptytd

My laptop is Comptel HEL80 with nvidia geforce go 7600.

Symptoms persist with the -14 kernel.

Same for me, with exact same hardware ...

Hi, changing "POST_VIDEO=true" to "POST_VIDEO=false" in /etc/default/acpi-support (I found the trick somewhere in Ubuntu's forums)

Hi, changing "POST_VIDEO=true" to "POST_VIDEO=false" in /etc/default/acpi-support did the jod for me (I found this somewhere in Ubuntu's forums, many thanks to whoever the author is ..)

Doesn't work for me... what is your exact hardware spec, gerard76?

Never mind... your workaround is good WITH restricted drivers and compiz installed. I just got it working. Yippee!

Still doesn't work without restricted though...

This is the first time I've heard of restricted drivers being *required* for suspend :-D

I guess this bug is not officially low priority.

I have this problem too. Unfortunately there is no fglrx driver for my hardware (ATI Mobility m6 ly) so I have to use vesa driver. Problem seems to be unrelated to the kernel as I've also tried custom 2.6.23.1 with SLAB instead of SLUB. And the Feisty kernel.

This will be retargeted towards the Hardy kernel once it is released. I've tagged this as "hardy-kernel-candidate" so that we make sure to retarget this report once the new release is out. However against the linux-source-2.6.22 package this is being marked as "Won't Fix" as it does not meet the criteria for a stable release update. To learn more about the stable release update process please refer to https://wiki.ubuntu.com/StableReleaseUpdates . Thanks!

This doesn't seem to actually be a kernel problem. I've tried running 2.6.20 feisty kernel, building my own 2.6.23 but with the same result. I only way to get my computer to resume from sleep was to use the VESA driver. Also with Linux Mint sleep is working, but I don't know what the difference is. Hopefully the regression can be reversed in Hardy (for me loosing sleep functionality is a pretty big regression).

Hardy Heron Alpha2 was recently released. It contains an updated version of the kernel. You can download and try the new Hardy Heron Alpha2 release from http://cdimage.ubuntu.com/releases/hardy/alpha-2/ . You should be able to then test the new kernel via the LiveCD. If you can, please verify if this bug still exists or not and report back your results. General information regarding the release can also be found here: http://www.ubuntu.com/testing/hardy/alpha2 . Thanks!

Tried with Hardy Heron Alpha 3. Suspend fails as before.

Setting this to confirmed as it has been independently verified.

Thanks Pausanias,

Care to run the the latest Hardy Alpha 6 release: http://www.ubuntu.com/testing care to attach the following per the kernel team bug policy. Please be sure to attach each file as a separate attachment.

* cat /proc/version_signature > version.log
* dmesg > dmesg.log
* sudo lspci -vvnn > lspci-vvnn.log

Also, it would be helpful to attach the dmesg output after a suspend/resume cycle as outlined here: https://wiki.ubuntu.com/DebuggingKernelSuspend. For more information regarding the kernel team bug policy, please refer to https://wiki.ubuntu.com/KernelTeamBugPolicies . Thanks again and we appreciate your help and feedback.

Tried again Hardy Alpha 6. It's a no go. Exactly the same problem.

Leann Ogasawara, thanks for your comment,
Resume bug also occurs on my laptop .. specs are :
Model : Dell Precision M70
VCard : Nvidia Quadro Fx GO with restricted drivers (nvidia_glx_new)
Ubuntu : Hardy Heron Beta 6, fresh install

I've set POST_VIDEO=false in /etc/default/acpi-support but the resume part of the RAM suspend still fails when called with the FnKey... What is strange is that when using " sync; echo 1 > /sys/power/pm_trace; /etc/acpi/sleep.sh force " as root makes suspend/resume working fine..

In attachement, you may find results of "cat /proc/version_signature", "dmesg", "lspci -vvnn" and a second "dmesg" done after the resume.

Mathieu Petit

Reporting with Hardy Beta. You're right, gerard. POST_VIDEO=false doesn't work in either Hardy Alpha 6 or Hardy Beta. Earlier I reported incorrectly that POST_VIDEO=false works in Hardy Alpha 6, but that's just in Feisty.

Thanks for the logs gerard. Since you've seconded and documented this error, I'm changing the status back to confirmed.

I'll attach my own logs as soon as I have a chance.

s/feisty/gutsy/g in previous post.

Hello,
I solved my problem .. it seems that with Hardy, ACPI-support is now delegated to pm-utils and Nvidia-kernel drivers doesn't mix so well with it. Following the recommendations of Vlowther in https://bugs.launchpad.net/ubuntu/+source/pm-utils/+bug/180378 , I created a file /etc/pm/config.d/ containing

# /bin/sh
# override the vbe-related video options if the nvidia driver is loaded.
# the nvidia kernel module knows how to handle this stuff properly.
if grep -q nvidia /proc/modules; then
 DISPLAY_QUIRK_VBESTATE_RESTORE=false
 DISPLAY_QUIRK_VBEMODE_RESTORE=false
 DISPLAY_QUIRK_VGA_MODE_3=false
 DISPLAY_QUIRK_DPMS_SUSPEND=false
 DISPLAY_QUIRK_VBE_POST=false
fi

and voila .. resume does not lock the display and computer anymore ...

Mathieu Petit

Sorry .. the complete filename to be created is /etc/pm/config.d/99-non-free-nvidia ..

Thank you again, gerard76.. or Mathieu... your sleuthing skills are unparalleled.

I can verify that the above script causes Hardy to suspend/resume correctly!

Just to be absolutely clear, at least on Precision M70, no modification of /etc/default/acpi-support is required one 99-non-free-nvidia is created as described above.

Marking this bug as duplicate of 180378, since that bug has a much clearer description of what's going on.