Bug 89308 - [bdw] GPU HANG on resuming
Summary: [bdw] GPU HANG on resuming
Status: CLOSED FIXED
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/Intel (show other bugs)
Version: XOrg git
Hardware: x86-64 (AMD64) Linux (All)
: medium normal
Assignee: Intel GFX Bugs mailing list
QA Contact: Intel GFX Bugs mailing list
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2015-02-25 05:18 UTC by Jason Harvey
Modified: 2016-05-18 17:07 UTC (History)
3 users (show)

See Also:
i915 platform: BDW
i915 features: GPU hang, power/suspend-resume


Attachments
Gzipped copy of /sys/class/drm/card0/error (353.90 KB, text/plain)
2015-02-25 05:18 UTC, Jason Harvey
no flags Details
intel_reg_dumper output after hang (12.11 KB, text/plain)
2015-02-25 05:23 UTC, Jason Harvey
no flags Details

Description Jason Harvey 2015-02-25 05:18:51 UTC
Created attachment 113805 [details]
Gzipped copy of /sys/class/drm/card0/error

HW is Lenovo X1 Carbon Gen3.

00:02.0 VGA compatible controller: Intel Corporation Broadwell-U Integrated Graphics (rev 09)


When resuming from suspend, desktop hangs briefly. After opening chrome, the entire chrome window is black, and the following message shows up in dmesg:


```
[ 1535.764464] [drm] stuck on render ring
[ 1535.765259] [drm] GPU HANG: ecode 0:0xc625c605, in Xorg [1471], reason: Ring hung, action: reset
[ 1535.765260] [drm] GPU hangs can indicate a bug anywhere in the entire gfx stack, including userspace.
[ 1535.765260] [drm] Please file a _new_ bug report on bugs.freedesktop.org against DRI -> DRM/Intel
[ 1535.765261] [drm] drm/i915 developers can then reassign to the right component if it's not a kernel issue.
[ 1535.765261] [drm] The gpu crash dump is required to analyze gpu hangs, so please always attach it.
[ 1535.765262] [drm] GPU crash dump saved to /sys/class/drm/card0/error
[ 1554.764090] [drm] stuck on render ring
[ 1554.764932] [drm] GPU HANG: ecode 0:0x541fa382, in chrome [8525], reason: Ring hung, action: reset
[ 1560.743786] [drm] stuck on render ring
[ 1560.746022] [drm] GPU HANG: ecode 0:0x541fa382, in chrome [8525], reason: Ring hung, action: reset
[ 1612.775561] [drm] stuck on render ring
[ 1612.777810] [drm] GPU HANG: ecode 0:0xa0cf2c0c, reason: Ring hung, action: reset
```


Gzipped error from /sys/class/drm/card0/error attached, as requested.
Comment 1 Jason Harvey 2015-02-25 05:19:19 UTC
$ uname -a
Linux artoo 3.13.0-37-generic #64-Ubuntu SMP Mon Sep 22 21:28:38 UTC 2014 x86_64 x86_64 x86_64 GNU/Linux
Comment 2 Jason Harvey 2015-02-25 05:23:22 UTC
Created attachment 113806 [details]
intel_reg_dumper output after hang
Comment 3 Jason Harvey 2015-02-27 07:29:38 UTC
Updated to new kernel and had identical symptom with a similar log message:

$ uname -a
Linux artoo 3.16.0-31-generic #41~14.04.1-Ubuntu SMP Wed Feb 11 19:30:13 UTC 2015 x86_64 x86_64 x86_64 GNU/Linux

$ dmesg
[ 9545.493792] [drm] stuck on render ring
[ 9545.494497] [drm] GPU HANG: ecode 0:0x00092929, in Xorg [1407], reason: Ring hung, action: reset
[ 9547.495093] [drm] Enabling RC6 states: RC6 on, RC6p off, RC6pp off
Comment 4 Jason Harvey 2015-03-04 15:42:49 UTC
Upgraded kernel again, this time to the intel-drm-next tag on 3.18:

$ uname -a
Linux artoo 3.18.0-997-generic #201412200341 SMP Sat Dec 20 03:42:55 UTC 2014 x86_64 x86_64 x86_64 GNU/Linux


Issue still occurs after resume from suspend:

$ dmesg
[32420.465811] [drm] stuck on render ring
[32420.466721] [drm] GPU HANG: ecode 8:0:0xfffffffe, in Xorg [1537], reason: Rin
g hung, action: reset
[32420.466722] [drm] GPU hangs can indicate a bug anywhere in the entire gfx sta
ck, including userspace.
[32420.466722] [drm] Please file a _new_ bug report on bugs.freedesktop.org agai
nst DRI -> DRM/Intel
[32420.466723] [drm] drm/i915 developers can then reassign to the right componen
t if it's not a kernel issue.
[32420.466724] [drm] The gpu crash dump is required to analyze gpu hangs, so ple
ase always attach it.
[32420.466724] [drm] GPU crash dump saved to /sys/class/drm/card0/error
[32420.473835] drm/i915: Resetting chip after gpu hang
Comment 5 Jason Harvey 2015-11-09 05:42:01 UTC
Issue has not reoccurred since kernel 4.3.0-040300rc5-generic.


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.