Intel GPU hangs with chromeos-4.14 |
|||
Issue descriptionThis is seen with chromeos-4.14, v4.14.13-3558-g295b17d7562b. [ 3304.736813] [drm] GPU HANG: ecode 9:0:0xfffffffe, reason: Hang on rcs0, action: reset [ 3304.745583] [drm] GPU hangs can indicate a bug anywhere in the entire gfx stack, including userspace. [ 3304.755904] [drm] Please file a _new_ bug report on bugs.freedesktop.org against DRI -> DRM/Intel [ 3304.765829] [drm] drm/i915 developers can then reassign to the right component if it's not a kernel issue. [ 3304.776636] [drm] The gpu crash dump is required to analyze gpu hangs, so please always attach it. [ 3304.786779] [drm] GPU crash dump saved to /sys/class/drm/card0/error [ 3304.794081] i915 0000:00:02.0: Resetting rcs0 after gpu hang [ 3320.735337] i915 0000:00:02.0: Resetting rcs0 after gpu hang [ 3322.847652] [drm:i915_gem_idle_work_handler] *ERROR* Timeout waiting for engines to idle [ 3338.719301] i915 0000:00:02.0: Resetting rcs0 after gpu hang [ 3340.832946] [drm:i915_gem_idle_work_handler] *ERROR* Timeout waiting for engines to idle [ 3366.751959] i915 0000:00:02.0: Resetting rcs0 after gpu hang It happens quite reliably when booting a 4.14 kernel on eve, without doing anything after boot.
,
Jan 18 2018
Culprit(s) may be: 5d8c5beecdd9 drm/i915: Apply Display WA #1183 on skl, kbl, and cfl 5d8c39173bd7 drm/i915: Disable DC states around GMBUS on GLK inherited from -stable. Currently testing reverts.
,
Jan 19 2018
2h30 runtime with patches in #2 reverted and no error. Re-applying patches and trying again. Assigning to myself for possible revert.
,
Jan 19 2018
Overnight with patches _not_ reverted and no error. Will keep tracking.
,
Jan 23 2018
Ever since I try to reliably reproduce the problem I don't see it anymore. Closing. Will reopen if/when I see it again. |
|||
►
Sign in to add a comment |
|||
Comment 1 by groeck@chromium.org
, Jan 18 2018