Bug 98318 - [BDW] GPU HANG: ecode 8:0:0x8ed9fff2, in RenderThread 1 [16790], reason: Hang on render ring, action: reset (StellarOverload)
Summary: [BDW] GPU HANG: ecode 8:0:0x8ed9fff2, in RenderThread 1 [16790], reason: Hang...
Status: RESOLVED FIXED
Alias: None
Product: Mesa
Classification: Unclassified
Component: Drivers/DRI/i965 (show other bugs)
Version: unspecified
Hardware: x86-64 (AMD64) Linux (All)
: medium normal
Assignee: Intel 3D Bugs Mailing List
QA Contact: Intel 3D Bugs Mailing List
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2016-10-19 02:52 UTC by Chris O
Modified: 2017-03-16 21:14 UTC (History)
2 users (show)

See Also:
i915 platform:
i915 features:


Attachments
/sys/class/drm/card0/error (37.52 KB, application/x-bzip)
2016-10-19 02:52 UTC, Chris O
Details
syslog find GPU HANG (6.89 KB, application/octet-stream)
2016-10-19 03:24 UTC, Chris O
Details
apitrace (30.00 MB, application/octet-stream)
2016-11-12 16:25 UTC, Chris O
Details
apitrace, 2 of 5 (30.00 MB, application/octet-stream)
2016-11-12 16:28 UTC, Chris O
Details
apitrace, 3 of 5 (30.00 MB, application/octet-stream)
2016-11-12 16:30 UTC, Chris O
Details
apitrace, 4 of 5 (30.00 MB, application/octet-stream)
2016-11-12 16:31 UTC, Chris O
Details
apitrace, 5 of 5 (9.54 MB, application/octet-stream)
2016-11-12 16:34 UTC, Chris O
Details

Description Chris O 2016-10-19 02:52:45 UTC
Created attachment 127397 [details]
/sys/class/drm/card0/error

1- Launch game executable .steam/steam/steamapps/common/Stellar\ Overload/StellarOverloadEA1/Binaries/Linux/StellarOverload
2- Start a new game
3- Select histoy mode
4- Click Play on avatar screen
5- Game display a load progess bar, at 75%, display freeze, mouse also, then in the terminal I see the message : 

intel_do_flush_locked failed: Input/output error
pure virtual method called

In dmesg out I see :
[11040.531420] [drm] GPU HANG: ecode 8:0:0x8ed9fff2, in RenderThread 1 [16790], reason: Hang on render ring, action: reset
[11040.531423] [drm] GPU hangs can indicate a bug anywhere in the entire gfx stack, including userspace.
[11040.531424] [drm] Please file a _new_ bug report on bugs.freedesktop.org against DRI -> DRM/Intel
[11040.531425] [drm] drm/i915 developers can then reassign to the right component if it's not a kernel issue.
[11040.531426] [drm] The gpu crash dump is required to analyze gpu hangs, so please always attach it.
[11040.531428] [drm] GPU crash dump saved to /sys/class/drm/card0/error
[11040.531512] drm/i915: Resetting chip after gpu hang
[11050.544688] drm/i915: Resetting chip after gpu hang
[14222.447331] drm/i915: Resetting chip after gpu hang
[14232.462464] drm/i915: Resetting chip after gpu hang

-- system architecture: x86_64
-- kernel version: 4.8.0-25-generic
-- Linux distribution: Ubuntu 16.10
-- Machine or mother board model: NUC5i5RYB, i5-5250U CPU, HD Graphics 6000
-- Display connector: HDMI
Comment 1 Chris O 2016-10-19 03:24:48 UTC
Created attachment 127398 [details]
syslog find GPU HANG
Comment 2 Chris O 2016-10-20 01:18:39 UTC
Mesa version 
OpenGL version string: 3.0 Mesa 12.1.0-devel

libdrm-intel1 version
2.4.71+git1610170956.a44c9c~gd~y

libgl1-mesa-dri
12.1~git1610181930.21af69~gd~y
Comment 3 Matt Turner 2016-11-02 05:02:44 UTC
Could you please capture and upload an apitrace (https://github.com/apitrace/apitrace) for a developer to use to reproduce the hang? Then please reset the bug's state to REOPENED.
Comment 4 Chris O 2016-11-12 16:25:21 UTC
Created attachment 127933 [details]
apitrace

apitrace, file 1 of 5
Comment 5 Chris O 2016-11-12 16:28:11 UTC
Created attachment 127934 [details]
apitrace,  2 of 5
Comment 6 Chris O 2016-11-12 16:30:09 UTC
Created attachment 127935 [details]
apitrace,  3 of 5
Comment 7 Chris O 2016-11-12 16:31:37 UTC
Created attachment 127936 [details]
apitrace, 4 of 5
Comment 8 Chris O 2016-11-12 16:34:23 UTC
Created attachment 127937 [details]
apitrace, 5 of 5

Join the 5 apitrace files parts first to reconstitute the full apitrace.
Comment 9 Anuj Phogat 2016-11-18 18:58:35 UTC
Reproduced this hang with mesa master commit 3ff9f8c on SKL, kernel 4.7.10-100.fc23.x86_64.
Comment 10 Anuj Phogat 2017-01-03 23:19:16 UTC
I couldn't reproduce the issue on SKL with Mesa master commit abcaba4, kernel  4.8.15-300.fc25.x86_64. So, something has fixed it between commit 3ff9f8c and abcaba4. Can you test it on BDW too?
Comment 11 Anuj Phogat 2017-03-16 21:14:47 UTC
Based on the results of my testing on SKL, I'm closing this bug.


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.