Bug 101912

Summary: [SKL] GPU HANG: ecode 9:0:0x85dffffb, in Xorg [5189], reason: Hang on render ring, action: reset
Product: Mesa Reporter: Hendrik Tews <hendrik>
Component: Drivers/DRI/i965Assignee: Intel 3D Bugs Mailing List <intel-3d-bugs>
Status: RESOLVED MOVED QA Contact: Intel 3D Bugs Mailing List <intel-3d-bugs>
Severity: normal    
Priority: medium CC: intel-gfx-bugs
Version: 13.0   
Hardware: x86-64 (AMD64)   
OS: Linux (All)   
Whiteboard:
i915 platform: SKL i915 features: GPU hang
Attachments: /sys/class/drm/card0/error

Description Hendrik Tews 2017-07-25 09:33:33 UTC
I am experiencing kernel crashes and X11 crashes/logouts when my wacom intuos pen tablet CTH-680 is connected. My syslog says

Jul 25 11:14:40 cert /usr/lib/gdm3/gdm-x-session[5187]: (EE) BUG: triggered 'if (in_input_thread())'
Jul 25 11:14:40 cert /usr/lib/gdm3/gdm-x-session[5187]: (EE) BUG: ../../../../os/io.c:656 in WriteToClient()
Jul 25 11:14:40 cert /usr/lib/gdm3/gdm-x-session[5187]: (EE) ******** WriteToClient called from input thread *********
Jul 25 11:14:40 cert /usr/lib/gdm3/gdm-x-session[5187]: (EE)
Jul 25 11:14:40 cert /usr/lib/gdm3/gdm-x-session[5187]: (EE) Backtrace:
Jul 25 11:14:40 cert /usr/lib/gdm3/gdm-x-session[5187]: (EE) 0: /usr/lib/xorg/Xorg (xorg_backtrace+0x4a) [0x55c2b1f7fcba]
Jul 25 11:14:40 cert /usr/lib/gdm3/gdm-x-session[5187]: (EE) 1: /usr/lib/xorg/Xorg (WriteToClient+0x162) [0x55c2b1f832f2]
Jul 25 11:14:40 cert /usr/lib/gdm3/gdm-x-session[5187]: (EE) 2: /usr/lib/xorg/Xorg (WriteEventsToClient+0x252) [0x55c2b1e25a72]
Jul 25 11:14:40 cert /usr/lib/gdm3/gdm-x-session[5187]: (EE) 3: /usr/lib/xorg/Xorg (TryClientEvents+0x120) [0x55c2b1e25c10]
Jul 25 11:14:40 cert /usr/lib/gdm3/gdm-x-session[5187]: (EE) 4: /usr/lib/xorg/Xorg (0x55c2b1dc6000+0x6337a) [0x55c2b1e2937a]
Jul 25 11:14:40 cert /usr/lib/gdm3/gdm-x-session[5187]: (EE) 5: /usr/lib/xorg/Xorg (DeliverEventsToWindow+0x1d7) [0x55c2b1e29667]
Jul 25 11:14:40 cert /usr/lib/gdm3/gdm-x-session[5187]: (EE) 6: /usr/lib/xorg/Xorg (0x55c2b1dc6000+0x146d99) [0x55c2b1f0cd99]
Jul 25 11:14:40 cert /usr/lib/gdm3/gdm-x-session[5187]: (EE) 7: /usr/lib/xorg/Xorg (0x55c2b1dc6000+0x15279f) [0x55c2b1f1879f]
Jul 25 11:14:40 cert /usr/lib/gdm3/gdm-x-session[5187]: (EE) 8: /usr/lib/xorg/Xorg (XIChangeDeviceProperty+0x358) [0x55c2b1f19158]
Jul 25 11:14:40 cert /usr/lib/gdm3/gdm-x-session[5187]: (EE) 9: /usr/lib/xorg/modules/input/wacom_drv.so (0x7f2b64110000+0x10df9) [0x7f2b64120df9]
Jul 25 11:14:40 cert /usr/lib/gdm3/gdm-x-session[5187]: (EE) 10: /usr/lib/xorg/modules/input/wacom_drv.so (0x7f2b64110000+0x76a7) [0x7f2b641176a7]
Jul 25 11:14:40 cert /usr/lib/gdm3/gdm-x-session[5187]: (EE) 11: /usr/lib/xorg/modules/input/wacom_drv.so (0x7f2b64110000+0x8864) [0x7f2b64118864]
Jul 25 11:14:40 cert /usr/lib/gdm3/gdm-x-session[5187]: (EE) 12: /usr/lib/xorg/modules/input/wacom_drv.so (0x7f2b64110000+0xe38f) [0x7f2b6411e38f]
Jul 25 11:14:40 cert /usr/lib/gdm3/gdm-x-session[5187]: (EE) 13: /usr/lib/xorg/modules/input/wacom_drv.so (0x7f2b64110000+0x616b) [0x7f2b6411616b]
Jul 25 11:14:40 cert /usr/lib/gdm3/gdm-x-session[5187]: (EE) 14: /usr/lib/xorg/modules/input/wacom_drv.so (0x7f2b64110000+0x63a6) [0x7f2b641163a6]
Jul 25 11:14:40 cert /usr/lib/gdm3/gdm-x-session[5187]: (EE) 15: /usr/lib/xorg/Xorg (0x55c2b1dc6000+0x1bbe2c) [0x55c2b1f81e2c]
Jul 25 11:14:40 cert /usr/lib/gdm3/gdm-x-session[5187]: (EE) 16: /usr/lib/xorg/Xorg (0x55c2b1dc6000+0x1be531) [0x55c2b1f84531]
Jul 25 11:14:40 cert /usr/lib/gdm3/gdm-x-session[5187]: (EE) 17: /usr/lib/xorg/Xorg (0x55c2b1dc6000+0x1bbc86) [0x55c2b1f81c86]
Jul 25 11:14:40 cert /usr/lib/gdm3/gdm-x-session[5187]: (EE) 18: /lib/x86_64-linux-gnu/libpthread.so.0 (0x7f2b6d044000+0x7494) [0x7f2b6d04b494]
Jul 25 11:14:40 cert /usr/lib/gdm3/gdm-x-session[5187]: (EE) 19: /lib/x86_64-linux-gnu/libc.so.6 (clone+0x3f) [0x7f2b6cd8daff]
Jul 25 11:14:40 cert /usr/lib/gdm3/gdm-x-session[5187]: (EE)
Jul 25 11:14:41 cert kernel: [ 3449.035862] [drm] GPU HANG: ecode 9:0:0x85dffffb, in Xorg [5189], reason: Hang on render ring, action: reset
Jul 25 11:14:41 cert kernel: [ 3449.035863] [drm] GPU hangs can indicate a bug anywhere in the entire gfx stack, including userspace.
Jul 25 11:14:41 cert kernel: [ 3449.035864] [drm] Please file a _new_ bug report on bugs.freedesktop.org against DRI -> DRM/Intel
Jul 25 11:14:41 cert kernel: [ 3449.035865] [drm] drm/i915 developers can then reassign to the right component if it's not a kernel issue.
Jul 25 11:14:41 cert kernel: [ 3449.035866] [drm] The gpu crash dump is required to analyze gpu hangs, so please always attach it.
Jul 25 11:14:41 cert kernel: [ 3449.035868] [drm] GPU crash dump saved to /sys/class/drm/card0/error
Jul 25 11:14:41 cert kernel: [ 3449.035921] drm/i915: Resetting chip after gpu hang
Jul 25 11:14:41 cert kernel: [ 3449.035998] [drm] RC6 on
Jul 25 11:14:41 cert kernel: [ 3449.047772] [drm] GuC firmware load skipped

/sys/class/drm/card0/error is empty, probably because I can only access it after rebooting/logging in again.

I use xorg 1:7.7+19 from Debian Stretch.
Comment 1 Chris Wilson 2017-07-25 10:08:26 UTC
The wacom input bug is supposed to be fixed in a new Xserver / wacom_drv.so. Without that error-state, we cannot triage the cause of the GPU hang.
Comment 2 Hendrik Tews 2017-07-25 12:22:29 UTC
Thanks for the quick reply! I updated xserver-xorg-input-wacom from 0.34.0 to 0.34.99. Let's see if this helps.
Comment 3 Hendrik Tews 2017-07-25 12:45:02 UTC
Version 0.34.99 has the same problem.

Meanwhile I found out that /sys/class/drm/card0/error is not empty despite ls reporting size 0. I attached its content, does this contain the information you need?

Thanks,

Hendrik
Comment 4 Hendrik Tews 2017-07-25 12:46:15 UTC
Created attachment 132951 [details]
/sys/class/drm/card0/error
Comment 5 Chris Wilson 2017-07-28 15:38:20 UTC
The GPU hang is from mesa, so double check you have the latest stable 17.1.x, 17.2 imminent.
Comment 6 Elizabeth 2017-08-24 16:59:25 UTC
(In reply to Chris Wilson from comment #5)
> The GPU hang is from mesa, so double check you have the latest stable
> 17.1.x, 17.2 imminent.
Hello Hendrik, any update with this information. Thanks.
Comment 7 Hendrik Tews 2017-08-25 08:17:38 UTC
Sorry, no news. I have not yet found the time to upgrade the mesa library (I have 13.0.6 from Debian stable). I will do when 17.2 enters testing. If you don't like that this report stays open for so long, then please close it with "fixed-in-version-17" or something similar.

Thanks again for the quick analysis of the problem.
Comment 8 GitLab Migration User 2019-09-25 19:03:27 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/1613.

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.