Created attachment 129752 [details] GPU erro dump Hello, I faced a GPU hung when running libreoffice at Fedora 25, the GPU performance started to downgrade right after I've started the application leading to a total screen freeze and a GPU reset, the following message was shown at dmesg: [ 4877.844879] [drm] GPU HANG: ecode 9:0:0x85dffffb, in Xorg [1315], reason: Hang on render ring, action: reset [ 4877.844883] [drm] GPU hangs can indicate a bug anywhere in the entire gfx stack, including userspace. [ 4877.844885] [drm] Please file a _new_ bug report on bugs.freedesktop.org against DRI -> DRM/Intel [ 4877.844887] [drm] drm/i915 developers can then reassign to the right component if it's not a kernel issue. [ 4877.844888] [drm] The gpu crash dump is required to analyze gpu hangs, so please always attach it. [ 4877.844891] [drm] GPU crash dump saved to /sys/class/drm/card0/error [ 4877.845070] drm/i915: Resetting chip after gpu hang [ 4877.845305] [drm] GuC firmware load skipped [ 4879.828210] [drm] RC6 on No more errors or issues was found previously. I've attached the GPU crash dump and Xorg log. These are the software versions: Kernel: 4.8.6-300.fc25.x86_64 Xorg: X.Org X Server 1.18.99.901 (1.19.0 RC 1) VGA Controller: Intel Corporation HD Graphics 520 [8086:1916] (rev 07) libdrm:2.4.71 mesa-dri-drivers: 12.0.3
Created attachment 129753 [details] Xorg log
Created attachment 129754 [details] dmesg output
Hello, Same for me: LibreOffice 5.0.6.2 Scientific Linux release 7.4 Kernel 4.15.0-1.el7.elrepo.x86_64 xorg-x11-server-Xorg 1.19.3 00:02.0 VGA compatible controller: Intel Corporation HD Graphics 520 (rev 07) mesa-dri-drivers 17.0.1 My dmesg output reads: [93418.593258] perf: interrupt took too long (3979 > 3963), lowering kernel.perf_event_max_sample_rate to 50000 [95872.516352] [drm] GPU HANG: ecode 9:0:0x85dffffb, in X [2038], reason: Hang on rcs0, action: reset [95872.516356] [drm] GPU hangs can indicate a bug anywhere in the entire gfx stack, including userspace. [95872.516357] [drm] Please file a _new_ bug report on bugs.freedesktop.org against DRI -> DRM/Intel [95872.516358] [drm] drm/i915 developers can then reassign to the right component if it's not a kernel issue. [95872.516359] [drm] The gpu crash dump is required to analyze gpu hangs, so please always attach it. [95872.516360] [drm] GPU crash dump saved to /sys/class/drm/card0/error [95872.516367] i915 0000:00:02.0: Resetting rcs0 after gpu hang [95880.509718] i915 0000:00:02.0: Resetting rcs0 after gpu hang [95888.510039] i915 0000:00:02.0: Resetting rcs0 after gpu hang [95896.510253] i915 0000:00:02.0: Resetting rcs0 after gpu hang [95904.510498] i915 0000:00:02.0: Resetting rcs0 after gpu hang
Created attachment 138146 [details] gpu crash dump /sys/class/drm/card0/error
These hangs have been fixed in mesa 17.3.6 *** This bug has been marked as a duplicate of bug 104411 ***
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.