Bug 106024

Summary:	[CI]igt@kms_* - incomplete - last line i dmesg :[drm:i915_fifo_underrun_reset_write [i915]] Re-arming FIFO underruns on pipe A
Product:	DRI	Reporter:	Marta Löfstedt <marta.lofstedt>
Component:	DRM/Intel	Assignee:	Intel GFX Bugs mailing list <intel-gfx-bugs>
Status:	CLOSED WORKSFORME	QA Contact:	Intel GFX Bugs mailing list <intel-gfx-bugs>
Severity:	normal
Priority:	high	CC:	intel-gfx-bugs
Version:	DRI git
Hardware:	Other
OS:	All
Whiteboard:	ReadyForDev
i915 platform:	KBL	i915 features:	display/Other
Bug Depends on:
Bug Blocks:	105980

Description Marta Löfstedt 2018-04-13 07:34:40 UTC

https://intel-gfx-ci.01.org/tree/drm-tip/drmtip_18/fi-kbl-7567u/igt@kms_cursor_crc@cursor-128x128-suspend.html

run.log
running: igt/gem_exec_suspend/basic-s4-devices

[31/75] skip: 9, pass: 22 \                   
FATAL: command execution failed
...
Completed CI_IGT_test drmtip_18/fi-kbl-7567u/14 : FAILURE
CI_IGT_test runtime 543 seconds
Rebooting fi-kbl-7567u

from dmesg:
<7>[  228.912937] [IGT] gem_exec_suspend: starting subtest basic-S4-devices
...
<7>[  236.220968] [IGT] kms_cursor_crc: executing
<7>[  236.271155] [drm:drm_helper_probe_single_connector_modes] [CONNECTOR:79:DP-2]
<7>[  236.271194] [drm:intel_dp_detect [i915]] [CONNECTOR:79:DP-2]
<7>[  236.271207] [drm:drm_helper_probe_single_connector_modes] [CONNECTOR:79:DP-2] disconnected
<7>[  236.271349] [drm:drm_helper_probe_single_connector_modes] [CONNECTOR:84:HDMI-A-1]
<7>[  236.271377] [drm:intel_hdmi_detect [i915]] [CONNECTOR:84:HDMI-A-1]
<7>[  236.324829] [drm:do_gmbus_xfer [i915]] GMBUS [i915 gmbus dpc] timed out, falling back to bit banging on pin 4
<7>[  236.327677] [drm:drm_do_probe_ddc_edid] drm: skipping non-existent adapter i915 gmbus dpc
<7>[  236.327750] [drm:intel_hdmi_set_edid [i915]] HDMI GMBUS EDID read failed, retry using GPIO bit-banging
<7>[  236.327822] [drm:intel_gmbus_force_bit [i915]] enabling bit-banging on i915 gmbus dpc. force bit now 1
<7>[  236.330638] [drm:drm_do_probe_ddc_edid] drm: skipping non-existent adapter i915 gmbus dpc
<7>[  236.330710] [drm:intel_gmbus_force_bit [i915]] disabling bit-banging on i915 gmbus dpc. force bit now 0
<7>[  236.384690] [drm:do_gmbus_xfer [i915]] GMBUS [i915 gmbus dpc] timed out, falling back to bit banging on pin 4
<7>[  236.387545] [drm:drm_dp_dual_mode_detect] DP dual mode HDMI ID:  (err -6)
<7>[  236.387555] [drm:drm_helper_probe_single_connector_modes] [CONNECTOR:84:HDMI-A-1] disconnected
<7>[  236.387910] [drm:drm_mode_addfb2] [FB:90]
<7>[  236.388864] [drm:drm_mode_addfb2] [FB:90]
<7>[  236.389309] [drm:drm_mode_addfb2] [FB:90]
<7>[  236.389697] [IGT] kms_cursor_crc: starting subtest cursor-128x128-suspend
<7>[  236.390307] [drm:drm_mode_addfb2] [FB:117]
<7>[  236.416267] [drm:drm_mode_setcrtc] [CRTC:41:pipe A]
<7>[  236.416331] [drm:drm_mode_setcrtc] [CONNECTOR:71:DP-1]
<7>[  236.433833] [drm:drm_mode_setcrtc] [CRTC:55:pipe B]
<7>[  236.434994] [drm:drm_mode_setcrtc] [CRTC:69:pipe C]
<7>[  236.450171] [drm:i915_fifo_underrun_reset_write [i915]] Re-arming FIFO underruns on pipe A

Comment 1 Marta Löfstedt 2018-04-13 07:37:48 UTC

I put this on a specific bug due to last dmesg: Re-arming FIFO underruns on pipe A

if from new feature to re-set the FIFO underrun to be able to re-enable FBC, that was added with:

 commit d52ad9cb9d6d3b696d6b7ad20a381a8f5520ea03
Author: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Date:   Wed Mar 28 12:05:26 2018 +0200

    drm/i915: Add debugfs file to clear FIFO underruns.

Comment 2 Marta Löfstedt 2018-04-17 05:52:44 UTC

https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_4058/shard-glkb4/igt@kms_atomic_transition@plane-all-transition.html

<7>[  530.159294] [drm:i915_fifo_underrun_reset_write [i915]] Re-arming FIFO underruns on pipe A

then last dmesg:
<6>[  532.692727] Console: switching to colour frame buffer device 400x112
<6>[  532.904364] Console: switching to colour dummy device 80x25
<7>[  532.904460] [IGT] kms_atomic_transition: executing

Comment 3 Marta Löfstedt 2018-04-18 10:31:30 UTC

https://intel-gfx-ci.01.org/tree/drm-tip/drmtip_22/fi-skl-6770hq/igt@kms_frontbuffer_tracking@fbc-1p-rte.html

Last dmesg:
<7>[  901.458499] [drm:i915_fifo_underrun_reset_write [i915]] Re-arming FIFO underruns on pipe B

Comment 4 Francesco Balestrieri 2018-06-01 07:25:25 UTC

Not seen for 4 weeks. Resolving.

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.