Bug 78723

Summary:

[BDW]igt some gem_reset_stats subcases cause GPU hang

Product:

DRI

Reporter:

lu hua <huax.lu>

Component:

DRM/Intel

Assignee:

Mika Kuoppala <mika.kuoppala>

Status:

CLOSED FIXED

QA Contact:

Intel GFX Bugs mailing list <intel-gfx-bugs>

Severity:

major

Priority:

high

CC:

intel-gfx-bugs

Version:

unspecified

Hardware:

All

OS:

Linux (All)

Whiteboard:

i915 platform:

i915 features:

Attachments:

Description	Flags
dmesg	none
/sys/class/drm/card0/error	none

Description lu hua 2014-05-15 06:06:58 UTC

Created attachment 99056 [details]
dmesg

system Environment:
--------------------------
Platform: Broadwell
Kernel:(drm-intel-nightly)c74cad3c2599b47438b168ca5629fbb00ab63f95

Bug detailed description:
-----------------------------
Run some gem_reset_stats subcases, GPU HANG reports in dmesg. It happens on  Broadwell with -nightly, queued and -fixes kernel.
Run earlier kernel, this issue still exists.

run  ./gem_reset_stats --run-subtest ban-blt
output:
IGT-Version: 1.6-g351e7d3 (x86_64) (Linux: 3.15.0-rc3_drm-intel-nightly_c74cad_20140515+ x86_64)
Subtest ban-blt: SUCCESS

dmesg -r | egrep "<[1-6]>" |grep drm
<6>[  106.713869] [drm] stuck on blitter ring
<6>[  106.715357] [drm] GPU HANG: ecode 2:0xe77ffff2, in gem_reset_stats [4470], reason: Ring hung, action: reset
<6>[  106.715360] [drm] GPU hangs can indicate a bug anywhere in the entire gfx stack, including userspace.
<6>[  106.715362] [drm] Please file a _new_ bug report on bugs.freedesktop.org against DRI -> DRM/Intel
<6>[  106.715366] [drm] drm/i915 developers can then reassign to the right component if it's not a kernel issue.
<6>[  106.715370] [drm] The gpu crash dump is required to analyze gpu hangs, so please always attach it.
<6>[  106.715371] [drm] GPU crash dump saved to /sys/class/drm/card0/error
<6>[  106.716771] [drm] Simulated gpu hang, resetting stop_rings
<6>[  112.711645] [drm] stuck on blitter ring
<6>[  112.712839] [drm] GPU HANG: ecode 2:0xe77ffff2, in gem_reset_stats [4470], reason: Ring hung, action: reset
<6>[  112.714606] [drm] Simulated gpu hang, resetting stop_rings

Reproduce steps:
----------------------------
1. ./gem_reset_stats --run-subtest ban-blt

Comment 1 lu hua 2014-05-15 06:08:24 UTC

Created attachment 99057 [details]
/sys/class/drm/card0/error

Comment 2 Mika Kuoppala 2014-05-15 09:48:06 UTC

<6>[  112.714606] [drm] Simulated gpu hang, resetting stop_rings

This test will produce two hangs in a row, to test the context banning feature of the driver. Both hangs are marked as simulated gpu hangs in the log.

I don't see what is wrong in here, could you elaborate?

Comment 3 Daniel Vetter 2014-05-15 14:12:26 UTC

Your dmesg filtering is b0rked, please see what piglit does or check out latest bkms.

Comment 4 Elizabeth 2017-10-06 14:38:11 UTC

Closing old verified.

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.