Bug 105219 - [snb] GPU HANG: ecode 6:0:0x85fffffc, in gnome-shell
Summary: [snb] GPU HANG: ecode 6:0:0x85fffffc, in gnome-shell
Status: RESOLVED MOVED
Alias: None
Product: Mesa
Classification: Unclassified
Component: Drivers/DRI/i965 (show other bugs)
Version: unspecified
Hardware: Other All
: medium normal
Assignee: Intel 3D Bugs Mailing List
QA Contact: Intel 3D Bugs Mailing List
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2018-02-22 23:52 UTC by mikhail.v.gavrilov
Modified: 2019-09-25 19:09 UTC (History)
1 user (show)

See Also:
i915 platform:
i915 features:


Attachments
dmesg (53.74 KB, text/plain)
2018-02-22 23:52 UTC, mikhail.v.gavrilov
Details
/sys/class/drm/card0/error (28.16 KB, text/plain)
2018-02-22 23:52 UTC, mikhail.v.gavrilov
Details
system log (385.30 KB, text/plain)
2018-03-10 19:45 UTC, mikhail.v.gavrilov
Details
/sys/class/drm/card0/error (21.63 KB, text/plain)
2018-03-10 19:45 UTC, mikhail.v.gavrilov
Details
dmesg (58.38 KB, text/plain)
2018-03-10 19:46 UTC, mikhail.v.gavrilov
Details

Description mikhail.v.gavrilov 2018-02-22 23:52:15 UTC
Created attachment 137547 [details]
dmesg

[ 4264.490623] [drm] GPU HANG: ecode 6:0:0x85fffffc, in gnome-shell [1514], reason: Hang on rcs0, action: reset
[ 4264.490625] [drm] GPU hangs can indicate a bug anywhere in the entire gfx stack, including userspace.
[ 4264.490626] [drm] Please file a _new_ bug report on bugs.freedesktop.org against DRI -> DRM/Intel
[ 4264.490626] [drm] drm/i915 developers can then reassign to the right component if it's not a kernel issue.
[ 4264.490626] [drm] The gpu crash dump is required to analyze gpu hangs, so please always attach it.
[ 4264.490627] [drm] GPU crash dump saved to /sys/class/drm/card0/error
[ 4264.490693] i915 0000:00:02.0: Resetting chip after gpu hang
[ 5202.670808] perf: interrupt took too long (6361 > 6315), lowering kernel.perf_event_max_sample_rate to 31000
[ 9323.092160] perf: interrupt took too long (8025 > 7951), lowering kernel.perf_event_max_sample_rate to 24000
Comment 1 mikhail.v.gavrilov 2018-02-22 23:52:44 UTC
Created attachment 137548 [details]
/sys/class/drm/card0/error
Comment 2 Elizabeth 2018-03-06 18:13:51 UTC
What produced this hang? What mesa version are you using? Thank you.
Comment 3 mikhail.v.gavrilov 2018-03-10 12:46:36 UTC
(In reply to Elizabeth from comment #2)
> What produced this hang?
I was watched video film in Totem player. And while watching the video several times freezed.

> What mesa version are you using? Thank you.
18.1.0-0.2.gita5053ba
Comment 4 mikhail.v.gavrilov 2018-03-10 19:45:03 UTC
Today this issue was repeated with latest LLVM 7.0.0-0.1.r326462 and MESA 18.1.0-0.6.gitd41ee9b

[34830.149115] [drm] GPU HANG: ecode 6:0:0x85fffffc, in gnome-shell [1622], reason: Hang on rcs0, action: reset
[34830.149211] [drm] GPU hangs can indicate a bug anywhere in the entire gfx stack, including userspace.
[34830.149213] [drm] Please file a _new_ bug report on bugs.freedesktop.org against DRI -> DRM/Intel
[34830.149215] [drm] drm/i915 developers can then reassign to the right component if it's not a kernel issue.
[34830.149216] [drm] The gpu crash dump is required to analyze gpu hangs, so please always attach it.
[34830.149218] [drm] GPU crash dump saved to /sys/class/drm/card0/error
[34830.149865] i915 0000:00:02.0: Resetting chip after gpu hang
Comment 5 mikhail.v.gavrilov 2018-03-10 19:45:27 UTC
Created attachment 137973 [details]
system log
Comment 6 mikhail.v.gavrilov 2018-03-10 19:45:52 UTC
Created attachment 137974 [details]
/sys/class/drm/card0/error
Comment 7 mikhail.v.gavrilov 2018-03-10 19:46:06 UTC
Created attachment 137975 [details]
dmesg
Comment 8 Elizabeth 2018-03-12 15:25:16 UTC
Hello again, could you try to identify any good commit for this issue? A setup where this worked properly? Thank you.
Comment 9 Denis 2018-06-07 15:24:01 UTC
hi. I checked this issue on Fedora27 ans Suse (with gnome-shell) and mesa version 18.1.0. MPEG4 video played in a loop for a night.Didn't reproduce hang.

Could be that it depends on video file resolution? If so, could you provide this information also?
Thank you
Comment 10 Vitalii Para 2018-06-23 08:58:12 UTC
try add in .drirc (it fix for me):
<driconf>
    <device screen="0" driver="i965">
        <application name="java" executable="java">
            <option name="mesa_no_error" value="true" />
            <option name="always_flush_batch" value="true" />
        </application>
    </device>
</driconf>

In my case it was an error in Batch Buffer.

Mesa 17.3.9
Linux localhost 4.14.44-desktop-2.mga6 #1 SMP Mon May 28 22:35:45 UTC 2018 x86_64 x86_64 x86_64 GNU/Linux
00:02.0 VGA compatible controller: Intel Corporation 2nd Generation Core Processor Family Integrated Graphics Controller (rev 09)

where java write your application name
Comment 11 GitLab Migration User 2019-09-25 19:09:39 UTC
-- GitLab Migration Automatic Message --

This bug has been migrated to freedesktop.org's GitLab instance and has been closed from further activity.

You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.freedesktop.org/mesa/mesa/issues/1695.


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.