Description of problem: After a few minutes (>5) the system runs into a complete hardlock when opening kmenu, or activating the lockscreen on an i7 sandybridge with nvidia gtx680 and nouveau. Version-Release number of selected component (if applicable): kde 5 on fedora 22 beta (same problem existed in alpha) How reproducible: login into kde, do something or leave the system idle an wait for the screen lock. system freezes completely. Steps to Reproduce: 1. Login and wait until screen lock becomes active or activate manually or use kmenu after a few minutes (5-15) 2. System runs into a hardlock. Actual results: Complete hardlock Expected results: No hardlock :-) Additional info:
If you can change VT's, can you get some log entries? Anything interesting from /var/log/Xorg.0.log and/or ~/.xsession-errors would be nice. Else, getting that information after rebooting too. The evidence here so far is pointing to a video driver issue.
Created attachment 1018108 [details] Xorg.0.log after reboot
Created attachment 1018109 [details] Xorg.0.log before reboot, after hardlock
Created attachment 1018111 [details] xsession-errors after reboot Sorry but there is no file from before
the only thing from nouveau i have in messages from before the crash is this: Apr 23 19:34:47 starscream kernel: nouveau E[ PFIFO][0000:01:00.0] read fault at 0x0000011000 [UNSUPPORTED_KIND] from PBDMA0/HOST on channel 0x007edb2000 [unknown] but there is still another entry after that from 19:34:53 regarding journal activities. this is the last entry. journalctl shows something similiar direct after detecting the screen lock event which caused the hardlock: Apr 23 19:34:47 starscream.cosmonate.net org.kde.kglobalaccel[1339]: Got XKeyPress event Apr 23 19:34:47 starscream.cosmonate.net org.kde.kglobalaccel[1339]: "Ctrl+Alt+L" = "Lock Session" Apr 23 19:34:47 starscream.cosmonate.net kernel: nouveau E[ PFIFO][0000:01:00.0] read fault at 0x0000011000 [UNSUPPORTED_KIND] from PBDMA0/HOST on channel 0x007edb2000 [unknown]
OK, now we may be getting somewhere. Anything in /var/log/messages and/or dmesg? I got it hit on, https://bugs.freedesktop.org/show_bug.cgi?id=76414 on the message about PFIFO read fault...
I have the same problem in 3 computers 1 with i7 3820 chipset x79, MB Asus Rampage IV Formula, nvidia gtx 690 2 with i7 960 chipset x58, MB Intel DX58SO, nvidia 220/560 In others computers the instalation complete perfect I whant help, please, let me know what to do for help with this insue Serching in google i found something in common in all similar bugs is the presence of core i7 both motherboards I tried disabling all extra features making just functional board I I disabled all secondary chips beyond the x79 / x58, audio, network, cpu processes, UEFI, overclocking, and more The bug continued to freeze the screen A note in the core i7 960 freezes less The system also bug using the video minimum installation option This problem also occurs with opensuse 42.1 (kernel4) This problem does not occur with Ubuntu 15 (kernel4)