Xorg 1.20.8 and 1.20.9 with Ryzen Raven GPU aborting on power saving event
by Yury_T from LinuxQuestions.org on (#57T9A)
After upgrading the system (Slackware-current), I've ran into an issue of Xorg server aborting on power saving event (10 minutes of inactivity, which is the default value I guess).
Screen switches off, but after reactivation I don't get the screen as I left it, but the display manager's login screen is presented instead. In xdm.log there are stack traces pointing to Xorg and its libraries, in Xorg.log there are messages on GPU sort of resetting (?) several times, once per 5 seconds.
The Xorg packages are at 1.20.9. I've tried downgrading back to May 2020 snapshot (Xorg 1.20.8), no luck. The diagnostic is somewhat different, and the issue doesn't come up every time I leave for 10+ minutes, but happens pretty often all the same.
The issue definitely WASN'T there with January 2020 snapshot of Slackware-current, which had Xorg at either 1.20.6 or 1.20.7 versions. I don't have snapshots for dates before May 2020, and so I can't pinpoint the offending modification more precisely.
I'm thinking about submitting a ticket at Xorg bugzilla, is it worth it?
That's what FlyGoat's RyzenAdj tells me:
Detected Raven
CPU Family: Raven
SMU BIOS Interface Version: 7
Screen switches off, but after reactivation I don't get the screen as I left it, but the display manager's login screen is presented instead. In xdm.log there are stack traces pointing to Xorg and its libraries, in Xorg.log there are messages on GPU sort of resetting (?) several times, once per 5 seconds.
The Xorg packages are at 1.20.9. I've tried downgrading back to May 2020 snapshot (Xorg 1.20.8), no luck. The diagnostic is somewhat different, and the issue doesn't come up every time I leave for 10+ minutes, but happens pretty often all the same.
The issue definitely WASN'T there with January 2020 snapshot of Slackware-current, which had Xorg at either 1.20.6 or 1.20.7 versions. I don't have snapshots for dates before May 2020, and so I can't pinpoint the offending modification more precisely.
I'm thinking about submitting a ticket at Xorg bugzilla, is it worth it?
That's what FlyGoat's RyzenAdj tells me:
Detected Raven
CPU Family: Raven
SMU BIOS Interface Version: 7