Bug 108614 - [snb] Random system freezes. GPU hang on rcs0
Summary: [snb] Random system freezes. GPU hang on rcs0
Status: NEW
Alias: None
Product: Mesa
Classification: Unclassified
Component: Drivers/DRI/i965 (show other bugs)
Version: 18.2
Hardware: x86-64 (AMD64) Linux (All)
: medium critical
Assignee: Intel 3D Bugs Mailing List
QA Contact: Intel 3D Bugs Mailing List
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2018-10-31 20:18 UTC by Gustavo
Modified: 2018-12-10 09:51 UTC (History)
1 user (show)

See Also:
i915 platform:
i915 features:


Attachments
GPU Crash Dump (53.88 KB, text/plain)
2018-10-31 20:18 UTC, Gustavo
Details

Note You need to log in before you can comment on or make changes to this bug.
Description Gustavo 2018-10-31 20:18:03 UTC
Created attachment 142307 [details]
GPU Crash Dump

Using Kubuntu 18.10 on Kernel version: 4.18.0-10-generic

While using the system, random freezes are occuring, lasting from 5 to 20 seconds, mouse keeps working but everything else is unresponsive. KWin crashes and restarts, and the notification pops: "KWin: Desktop effects were restarted due to a graphics reset". I'm having around 5-6 crashes on a 10 minute period.

Running dmesg I get:

[  920.806022] [drm] GPU HANG: ecode 6:0:0x87e8effd, in chrome [4579], reason: hang on rcs0, action: reset
[  920.806024] [drm] GPU hangs can indicate a bug anywhere in the entire gfx stack, including userspace.
[  920.806024] [drm] Please file a _new_ bug report on bugs.freedesktop.org against DRI -> DRM/Intel
[  920.806025] [drm] drm/i915 developers can then reassign to the right component if it's not a kernel issue.
[  920.806025] [drm] The gpu crash dump is required to analyze gpu hangs, so please always attach it.
[  920.806026] [drm] GPU crash dump saved to /sys/class/drm/card0/error


and a bunch of:

[  920.806092] i915 0000:00:02.0: Resetting chip for hang on rcs0
[  976.832401] i915 0000:00:02.0: Resetting chip for hang on rcs0
[ 1049.792379] i915 0000:00:02.0: Resetting chip for hang on rcs0
[ 1085.792388] i915 0000:00:02.0: Resetting chip for hang on rcs0
[ 1093.792021] i915 0000:00:02.0: Resetting chip for hang on rcs0
[ 1101.792036] i915 0000:00:02.0: Resetting chip for hang on rcs0
[ 1171.840381] i915 0000:00:02.0: Resetting chip for hang on rcs0
[ 1198.784395] i915 0000:00:02.0: Resetting chip for hang on rcs0
[ 1213.824390] i915 0000:00:02.0: Resetting chip for hang on rcs0
[ 1227.808413] i915 0000:00:02.0: Resetting chip for hang on rcs0
[ 1235.808051] i915 0000:00:02.0: Resetting chip for hang on rcs0
[ 1243.808052] i915 0000:00:02.0: Resetting chip for hang on rcs0
[ 1293.792374] i915 0000:00:02.0: Resetting chip for hang on rcs0


Tried to upgrade from mesa 18.2.2 to 18.2.3 but didn't solve the problem.

The GPU Crash dump is attatched.
Comment 1 Chris Wilson 2018-10-31 21:02:17 UTC
One of those fun little snb bugs where the CS keeps on going but the GPU stops responding.
Comment 2 Georg 2018-12-10 09:51:47 UTC
I have a 'gpu hang' issue as well. May be a different or the same root cause. My investigation to my situation shows that this is directly or indirectly a kernel issue. See here: https://bugs.freedesktop.org/show_bug.cgi?id=107420.


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.