GPU HANG: ecode 9:0:00000000 - coffee lake, 5.6-rc5,rc6,rc7,5.6.2

Mar 09 12:24:33 thinkpad kernel: Asynchronous wait on fence 0000:00:02.0:kwin_x11[1756]:7c6d8 timed out (hint:intel_atomic_commit_ready+0x0/0x54)
Mar 09 12:24:37 thinkpad kernel: i915 0000:00:02.0: GPU HANG: ecode 9:0:00000000
Mar 09 12:24:37 thinkpad kernel: GPU hangs can indicate a bug anywhere in the entire gfx stack, including userspace.
Mar 09 12:24:37 thinkpad kernel: Please file a _new_ bug report at https://gitlab.freedesktop.org/drm/intel/issues/new.
Mar 09 12:24:37 thinkpad kernel: Please see https://gitlab.freedesktop.org/drm/intel/-/wikis/How-to-file-i915-bugs for details.
Mar 09 12:24:37 thinkpad kernel: drm/i915 developers can then reassign to the right component if it's not a kernel issue.
Mar 09 12:24:37 thinkpad kernel: The GPU crash dump is required to analyze GPU hangs, so please always attach it.
Mar 09 12:24:37 thinkpad kernel: GPU crash dump saved to /sys/class/drm/card0/error
Mar 09 12:24:37 thinkpad kernel: i915 0000:00:02.0: Resetting rcs0 for stopped heartbeat on rcs0
Mar 09 12:24:52 thinkpad kernel: i915 0000:00:02.0: GPU HANG: ecode 9:0:00000000
Mar 09 12:24:52 thinkpad kernel: i915 0000:00:02.0: Resetting rcs0 for stopped heartbeat on rcs0
Mar 09 12:25:07 thinkpad kernel: i915 0000:00:02.0: GPU HANG: ecode 9:0:00000000
Mar 09 12:25:07 thinkpad kernel: i915 0000:00:02.0: Resetting rcs0 for stopped heartbeat on rcs0
Mar 09 12:25:22 thinkpad kernel: i915 0000:00:02.0: GPU HANG: ecode 9:0:00000000
Mar 09 12:25:22 thinkpad kernel: i915 0000:00:02.0: Resetting rcs0 for stopped heartbeat on rcs0
Mar 09 12:25:33 thinkpad kernel: Asynchronous wait on fence 0000:00:02.0:Xorg[1550]:431f16 timed out (hint:intel_atomic_commit_ready+0x0/0x54)

See below for /sys/class/drm/card0/error from several different crashes that all go the same way.