Comment 0 for bug 1723680

Revision history for this message
Aun (aun-sswick) wrote :

When using linux kernel linux-image-4.13.0-12-generic to linux-image-4.13.0-16-generic Xorg fails to start with error kernel error messages below. Kernel boots and displays messages allows for GUI to enter disk encryption key then tries to start X login. Shows blank screen, sometimes part of login dialog. Then hangs and goes blank usually showing the low graphics mode error screen from Xorg and no functionality after a while. Can ctrl+alt+f1 to console. Using linux-image-4.10.0-35-generic everything works as expected. System worked without issue with Ubuntu 17.04

Kernel error messages:
[ 43.836641] [drm] GPU HANG: ecode 8:0:0x00dfffff, in Xorg [1027], reason: Hang on rcs0, action: reset
[ 43.836645] [drm] GPU hangs can indicate a bug anywhere in the entire gfx stack, including userspace.
[ 43.836646] [drm] Please file a _new_ bug report on bugs.freedesktop.org against DRI -> DRM/Intel
[ 43.836647] [drm] drm/i915 developers can then reassign to the right component if it's not a kernel issue.
[ 43.836648] [drm] The gpu crash dump is required to analyze gpu hangs, so please always attach it.
[ 43.836649] [drm] GPU crash dump saved to /sys/class/drm/card0/error
[ 43.836718] drm/i915: Resetting chip after gpu hang
[ 55.812149] drm/i915: Resetting chip after gpu hang
[ 63.840645] drm/i915: Resetting chip after gpu hang
[ 71.836614] drm/i915: Resetting chip after gpu hang
[ 79.836882] drm/i915: Resetting chip after gpu hang
[ 88.828508] drm/i915: Resetting chip after gpu hang

have tried kernel parameter intel_iommu=off, intel_iommu=igfx_off, i915.enable_rc6=0 with no better results.
lshw, dmesg and card0/error file attached.