https://bugzilla.kernel.org/show_bug.cgi?id=209253 Mair O'Nire (maironire@xxxxxxxxx) changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |maironire@xxxxxxxxx --- Comment #23 from Mair O'Nire (maironire@xxxxxxxxx) --- Hi all. I'm new here, registered exactly to track this bug. I've got a new laptop recently, HP Omen 2020, with AMD Ryzen 4800H and Nvidia 1660Ti, and whilst was trying to setup kvm virtualized system I ran into the same issue. During investigation I've found Ian's report, and then this one, too. Backtraces were very similar to other posted above, all at fs/eventfd.c:74, but sometimes with different origin. So I decided to give Alex's patch a try, and built mainline kernel with this patch applied. Mine host kernel was 5.8.[6,17,18], Fedora 33, and they all seems to be affected. Guest is Windows 10. So, I do confirm - patch works, system became stable, and there was no crashes for about a week, nor under regular work load, neither while running benchmarks. Before the fix guest uptime varied from 2 to 10 minutes and then crash. Moreover, I had to reboot host to get Nvidia back to work in vm again. Thanks Alex, Ian, and others, you guys doin' a great work if you don't know this already :) -- You are receiving this mail because: You are watching the assignee of the bug.
