( Nvidia340xx+archlinux64+gnome3.14)the background of desktop and lockscreen mess after resume from

( Nvidia340xx+g310M+archlinux64+gnome3.14)the background of desktop and lockscreen mess after resume from sleep. It’s work well when using the opensource driver, so i think it’s a bug of nvidia driver. the last driver version 340.58 don’t solve this problem.

my kernel is linux 3.17.2.

I’m seeing the same thing on a GTX 680, driver version 343.22, kernel 3.17.2 in Fedora 21 beta. Here is an example of what happens (just the background, I anonimized the window contents afterwards)

Imgur

Does any body meet the same problems?

Please provide nvidia bug report by running nvidia-bug-report.sh script as root user.

Internally tracking this issue under Bug 200055189

i did as you said
nvidia-bug-report.log.gz (74.9 KB)

i did it

Glad to see a bug report concerning this problem !
Same on ubuntu-gnome 14.10 (gnome 3.14).
I’ll provide a bug report too attached to the bug 200055189 when i’m back home.

Here are my debug logs too: nvidia-bug-report.log.gz

“Our current software architecture doesn’t preserve the contents of FBOs across modeswitches or power events. As a result, our implementation relies on applications re-rendering to their FBOs every frame. This is a fundamental limitation of our design that will be fixed in the future.”

Hope to be able to early correction

Could you tell me how to query the bug status?

I am using OpenSuse with Gnome 3.14.1 and an Nvidia GeForce GT 610. This bug appears after every suspend-resume with Nvidia driver versions 340.65 and 343.36. Currently using the nouveau driver to work around it.

Just tried driver version 3.49.16 (with Gnome 3.16.1) and no longer see this issue. Can’t see any direct mention of a fix in the release notes but it does appear fixed. Very cool!

Is this something you are still working on? Any hope of seeing this fixed in a future driver release?

Yes, it’s still in the bug list. But it’s a deep architectural change, not a quick patch fix, so I can’t promise any particular timeframe for when it might be fixed.

Thank you for the reply.

I just noticed that the thread topic refers to the 340.* legacy driver series specifically. Just to make any expectations clear, any improvements made to future drivers for this will not be backported to the 340 legacy series.

That’s understandable and clear. Thank you. I’m using the 381.22 driver.
I posted in this particular thread instead of started a new thread because this one had an accurate description of the situation: