https://bugs.kde.org/show_bug.cgi?id=422572
--- Comment #1 from Leyenda <neopy...@gmail.com> --- SUMMARY STEPS TO REPRODUCE 1. Use the gpu from an amd Ryzen 5 2400g cpu, with Radeon Vega Graphics 2. Use a 5.6.x kernel and kwin-wayland from plasma 5.19 or master 3. Browse the web with firefox flatpak OBSERVED RESULT When a graphic error from amdgpu driver occurs, plasma crashes and logs out to sddm, ending your session. That doesn't happen on x11, on x11 when an error occurs x11 restars itself withouth any problem. EXPECTED RESULT Graphic server restarts itself without problems. SOFTWARE/OS VERSIONS Windows: macOS: Linux/KDE Plasma: KDE Neon Developer edition (available in About System) KDE Plasma Version: 5.19.80 KDE Frameworks Version: 5.71 Qt Version: 5.14.2 ADDITIONAL INFORMATION Kernel 5.6.15 info on dmesg: [ 2218.863065] amdgpu 0000:06:00.0: [gfxhub0] retry page fault (src_id:0 ring:0 vmid:4 pasid:32772, for process Xwayland pid 2022 thread Xwayland:cs0 pid 2024) [ 2218.863066] amdgpu 0000:06:00.0: in page starting at address 0x0000800100aac000 from client 27 [ 2218.863067] amdgpu 0000:06:00.0: VM_L2_PROTECTION_FAULT_STATUS:0x00401031 [ 2218.863068] amdgpu 0000:06:00.0: MORE_FAULTS: 0x1 [ 2218.863070] amdgpu 0000:06:00.0: WALKER_ERROR: 0x0 [ 2218.863071] amdgpu 0000:06:00.0: PERMISSION_FAULTS: 0x3 [ 2218.863072] amdgpu 0000:06:00.0: MAPPING_ERROR: 0x0 [ 2218.863073] amdgpu 0000:06:00.0: RW: 0x0 [ 2219.018238] [drm:amdgpu_job_timedout [amdgpu]] *ERROR* ring gfx timeout, but soft recovered -- You are receiving this mail because: You are watching all bug changes.
