Bug 96681

Summary: GUI hangs after a few minutes of activity
Product: DRI Reporter: oz
Component: DRM/IntelAssignee: Intel GFX Bugs mailing list <intel-gfx-bugs>
Status: CLOSED FIXED QA Contact: Intel GFX Bugs mailing list <intel-gfx-bugs>
Severity: major    
Priority: medium CC: intel-gfx-bugs
Version: unspecified   
Hardware: x86-64 (AMD64)   
OS: Linux (All)   
Whiteboard:
i915 platform: BSW/CHT i915 features: GPU hang
Attachments:
Description Flags
sys-class-drm-card0-error none

Description oz 2016-06-26 12:57:31 UTC 
Created attachment 124725 [details]
sys-class-drm-card0-error

Dear maintainers,

since 4.6-er series kernel the GUI and input devices hang after a few minutes of activity (for example webbrowsing). Usually I have the chance to return  to console with Ctrl-Alt-F1 (which also lags a few minutes) to shutdown the system.

The GUI is very stable with the former Debian Kernel Name: linux-image-4.5.0-2-amd64  Version: 4.5.5-1. 

The problematic kernel is for example Debian kernel Name: linux-image-4.6.0-1-amd64  Version:4.6.2-2.

The kern.log catches events starting with:
[code]
Jun 26 10:30:23 amy kernel: [  253.706204] [drm] stuck on render ring
Jun 26 10:30:23 amy kernel: [  253.712117] [drm] GPU HANG: ecode 8:0:0xfffffffe, in Xorg [683], reason: Ring hung, action: reset
Jun 26 10:30:23 amy kernel: [  253.712121] [drm] GPU hangs can indicate a bug anywhere in the entire gfx stack, including userspace.
Jun 26 10:30:23 amy kernel: [  253.712123] [drm] Please file a _new_ bug report on bugs.freedesktop.org against DRI -> DRM/Intel
Jun 26 10:30:23 amy kernel: [  253.712125] [drm] drm/i915 developers can then reassign to the right component if it's not a kernel issue.
Jun 26 10:30:23 amy kernel: [  253.712127] [drm] The gpu crash dump is required to analyze gpu hangs, so please always attach it.
Jun 26 10:30:23 amy kernel: [  253.712128] [drm] GPU crash dump saved to /sys/class/drm/card0/error
Jun 26 10:30:23 amy kernel: [  253.714221] drm/i915: Resetting chip after gpu hang
Jun 26 10:30:31 amy kernel: [  261.710739] [drm] stuck on render ring
Jun 26 10:30:31 amy kernel: [  261.716770] [drm] GPU HANG: ecode 8:0:0xfffffffe, in Xorg [683], reason: Ring hung, action: reset
Jun 26 10:30:31 amy kernel: [  261.718255] [drm:i915_set_reset_status [i915]] *ERROR* gpu hanging too fast, banning!
Jun 26 10:30:31 amy kernel: [  261.720424] drm/i915: Resetting chip after gpu hang
Jun 26 10:32:03 amy kernel: [  354.150760] ------------[ cut here ]------------
Jun 26 10:32:03 amy kernel: [  354.150960] WARNING: CPU: 0 PID: 683 at /build/linux-FzSZXp/linux-4.6.2/drivers/gpu/drm/i915/intel_display.c:3931 intel_atomic_commit+0x1476/0x14f0 [i915]
Jun 26 10:32:03 amy kernel: [  354.150971] Removing stuck page flip
[...]
[/code]

Attached is cat /sys/class/drm/card0/error.

Thank you
Comment 1 oz 2016-07-08 13:15:40 UTC 
Since the latest kernel and intel xorg driver my problems are gone under Debian/Sid. These were the version changes:

linux-image-4.6.0-1-amd64:amd64 (4.6.2-2, 4.6.3-1)
xserver-xorg-video-intel:amd64 (2:2.99.917+git20160522-1, 2:2.99.917+git20160706-1)

I was the reporter of this bug.

Thank you!

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.