302.07 no resume from suspend

[artem]

Originally Posted by ledoc

Seems to affect only nv50 family chips then?

Most likely, except there's a person in the thread who mentions GTX285.

[Dragoran]

Originally Posted by artem

Most likely, except there's a person in the thread who mentions GTX285.

Which falls into "nv50 class GPU" which basically covers

GeForce 8, GeForce 9, GeForce 100, GeForce 200, GeForce 300

[1]

[mdriftmeyer]

I'm not seeing any outputs from X or any information on which versions of X each person expressing concern about failure has installed.

I'm sure it will help Nvidia if you folks post your known configuration beyond, the GPU you have installed and which Kernel you are running it on.

After all, wouldn't this be an intercommunication issue between X and Nvidia's driver?/

[ltmon]

Getting this problem myself, with drivers installed from Ubuntu Xorg edge PPA (i.e. very recent Xorg). I was assuming using this was at least part of the issue. You can see all the packages it updates from standard Ubuntu 12.04 here:

https://launchpad.net/~xorg-edgers/+...filter=precise

Quadro FX 880M
X.Org X Server 1.12.1 (Release Date: 2012-04-13)
Kernel 3.2.0-24-generic #38-Ubuntu SMP Tue May 1 16:18:50 UTC 2012 x86_64 x86_64 x86_64 GNU/Linux

The Xorg log after this problem shows nothing post suspend -- just the messages it gives when suspending. None of the other system logs show anything at all either.

Nvidia bug report attached if it's useful.

Luke.

[artem]

There is the only nvidia-bug-report attached to this thread.

Please, if you are affected by this bug, post your nvidia-bug-report as well.

[sandipt]

NVIDIA internal Bug ID: 983043 to track this issue.

[ngoonee]

Attaching my own, should I have reinstalled 302.07 to run it? As I understand it nvidia-bug-report.sh is mainly to provide system-specific data?

[artem]

Originally Posted by ngoonee

Attaching my own, should I have reinstalled 302.07 to run it? As I understand it nvidia-bug-report.sh is mainly to provide system-specific data?

It's not necessary to reinstall 302.07 to run it. You're right.

[NvFuchs]

Originally Posted by sandipt

Please provide nvidia bug report

http://nvnews.net/vbulletin/showthread.php?t=180018

For another report if you need it. Thanks for working on it, I hope that there will be a fix soon.

Fuchs

[MsTiFtS]

Same here, on a Lenovo T410 with nVidia Quadro NVS 3100m running Ubuntu 12.04 LTS with 302.07 driver from xorg-edgers.

Suspend to disk/RAM fails to wake up properly, displaying only a black screen, and I have to resort to sysrq to reboot it. If, however, I switch to tty1-6 before suspending, it wakes up just fine, and will only lock up as soon as I switch back to tty7 (X server). If, after suspending/resuming, I restart the X server before switching back to it, things continue to work fine. Those various pm-suspend quirks don't seem to help at all.

As soon as the problem occurs I get a couple of these messages in dmesg:

NVRM: os_schedule: Attempted to yield the CPU while in atomic or interrupt context.

Feel free to contact me for further details, I'm very willing to help nail this down or test beta drivers. This bug is really driving me nuts.

[valentine]

Affects me as well.

Lenovo Thinkpad W510
01:00.0 VGA compatible controller: NVIDIA Corporation GT216 [Quadro FX 880M] (rev a2)
Ubuntu 12.04
nvidia-current 302.07-0ubuntu1~xedgers~precise1

[MadCatX]

I'm seeing a similar problem, except that my system doesn't lock up while resuming, but when it's just about to power down for sleeping. All I get is a black screen with a text cursor blinking in the top left corner. Switching to other ttys has no effect.

9800M GTS, kernel 3.4-rc7, Arch Linux 64bit