Bug 82755 - [ivb] GPU Crash causes font corruption
Summary: [ivb] GPU Crash causes font corruption
Status: CLOSED WORKSFORME
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: 2014-08-18 08:45 UTC by hirner
Modified: 2017-07-24 22:52 UTC (History)
3 users (show)

See Also:
i915 platform: IVB
i915 features: GPU hang

Attachments
Screenshot of font corruption (32.50 KB, image/png)
2014-08-18 08:45 UTC, hirner 		no flags Details
GPU crash dump, taken from /sys/class/drm/card0/error (2.04 MB, text/plain)
2014-08-18 08:46 UTC, hirner 		no flags Details
View All

Description hirner 2014-08-18 08:45:49 UTC 
Created attachment 104805 [details]
Screenshot of font corruption

I sometimes notice a short interruption where the system is not responsive. After a few seconds, it's usable again, but graphics in opened windows like the browser is corrupted (most times, only single font characters are corrupted, for instance all "9" are replaced by "e" - this is not a single rendering issue but persists and can even be captured by a screenshot, see attached screenshot). dmesg shows:

[ 3738.273393] [drm] stuck on render ring
[ 3738.273403] [drm] GPU crash dump saved to /sys/class/drm/card0/error
[ 3738.273404] [drm] GPU hangs can indicate a bug anywhere in the entire gfx stack, including userspace.
[ 3738.273405] [drm] Please file a _new_ bug report on bugs.freedesktop.org against DRI -> DRM/Intel
[ 3738.273406] [drm] drm/i915 developers can then reassign to the right component if it's not a kernel issue.
[ 3738.273407] [drm] The gpu crash dump is required to analyze gpu hangs, so please always attach it.

Hardware: Lenovo ThinkPad T340s, "Intel Corporation 3rd Gen Core processor Graphics Controller"
Software: Ubuntu 14.04 LTS
Comment 1 hirner 2014-08-18 08:46:55 UTC 
Created attachment 104806 [details]
GPU crash dump, taken from /sys/class/drm/card0/error
Comment 2 Ben Widawsky 2014-12-19 19:27:39 UTC 
Can you please try to reproduce this with the latest drm-intel-nightly kernel?
Comment 3 Jani Nikula 2015-10-23 09:58:16 UTC 
(In reply to Ben Widawsky from comment #2)
> Can you please try to reproduce this with the latest drm-intel-nightly
> kernel?

I guess not. Timeout, closing. Please reopen if the problem persists with latest kernels.
Comment 4 rh 2015-10-23 09:59:04 UTC 
Yes, sorry, I don't have this laptop anymore.
Comment 5 Chris Wilson 2015-10-23 10:03:11 UTC 
Yes, the problem still persists. We don't have batch level granularity in our reset, therefore when one component hangs the GPU all pending rendering is lost and other clients suffer corruption.
Comment 6 Jani Nikula 2015-10-23 11:30:37 UTC 
(In reply to Chris Wilson from comment #5)
> Yes, the problem still persists. We don't have batch level granularity in
> our reset, therefore when one component hangs the GPU all pending rendering
> is lost and other clients suffer corruption.

Is this an enhancement you want to track in bugzilla?

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.