Described the problem here, similar to what others have reported on this forum: http://askubuntu.com/questions/828853/screen-and-keyboard-sometimes-dont-work-after-suspend
I did not start the X server with startx -- -logverbose 6. I might be able to do it if necessary. Could also have someone ssh in maybe, I’ve never used it.
I’m attaching the crash report from before the problem (old.gz) as well, though it’s probably not useful, because it successfully ran /usr/bin/nvidia-debugdump -D. The report after the problem says this:
ERROR: internal_dumpGpuComponent() failed, return code: 0xf
My general experience has been that sound is still working after a malfunctioning suspend. I was testing this, and while it usually worked, it seemed like once it didn’t work, in that an ffplay instance that had been running before the suspend was no longer outputting sound.
In other testing, it seemed like ffplay would silently fail normally (after printing out version information). The first time successful test this seemed to happen within two seconds. The second time (having avoided -hide_banner option and correctly redirected standard error), ‘date’ command outputs before and after it differ by 16 seconds, though I think I only put ‘sleep 5’. Only just noticed this and probably not relevant. I don’t think any sound played from this instance of ffplay, and it had the -t 3 option to restrict playback length.
But with the -nodisp option that turns off a visual display of frequencies for an mp3 file, it gave the error ‘xcb_connection_has_error() returned true’ but correctly played, with lots of playback position information in the output.
Even if there isn’t a solution, I would like to verify whether this is an Nvidia problem! I was suspending numerous times while testing ffplay without ever encountering a malfunctioning suspend, and noticed that I had removed my USB wireless device, possibly to reduce spam in the log files; once I plugged it back in and finally got it to work again (taking it out during a suspend or something might have bugged something), the next suspend did malfunction.
Attached is the ‘old.gz’ when the system was functioning well, and the nvidia-bug-report.log.gz from after the system had incorrectly resumed after a suspend.
nvidia-bug-report.log.old.gz (116 KB)
nvidia-bug-report.log.gz (103 KB)