Bug 85828

Summary: [SNB/IVB/HSW]igt gem_caching doesn't exit /sys/kernel/debug/dri/0/clients
Product: DRI Reporter: lu hua <huax.lu>
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: normal    
Priority: medium CC: christophe.prigent, intel-gfx-bugs
Version: unspecified   
Hardware: All   
OS: Linux (All)   
Whiteboard:
i915 platform: HSW, IVB, SNB i915 features: GEM/Other
Attachments:
Description Flags
run log none

Description lu hua 2014-11-04 02:21:55 UTC
Created attachment 108868 [details]
run log

==System Environment==
--------------------------
Regression: not sure 
Non-working platforms: IVB

==kernel==
--------------------------
drm-intel-nightly/203b34bf78a127a85cf89bbd8cf4decd935d780f

==Bug detailed description==
Run automation testing, run gem_basic_create-fd-close then run gem_caching_reads, it doesn't exit /sys/kernel/debug/dri/0/clients. 
Run it 5 cycles manually, I can't reproduce this issue.

/sys/kernel/debug/dri/0/clients   
          command   pid dev master a   uid      magic
         gem_caching  8043   0   y    y     0          0
         gem_caching  8043   0   n    y     0          0
Comment 1 Mika Kuoppala 2014-11-04 09:57:00 UTC
Do the gem_caching clients disappear if you wait long enough?

While the clients are still there, does 'strace -p <gem_caching_pid>' show any output?
Comment 2 Mika Kuoppala 2014-11-04 10:06:29 UTC
According to the log, the gem_caching is showing in the client list too early.
Perhaps ordering/race issue in the testrunner.

Similar symptomps as in
https://bugs.freedesktop.org/show_bug.cgi?id=85831
Comment 3 Mika Kuoppala 2014-11-04 10:11:11 UTC
*** Bug 85831 has been marked as a duplicate of this bug. ***
Comment 4 lu hua 2014-11-05 08:10:55 UTC
(In reply to Mika Kuoppala from comment #1)
> Do the gem_caching clients disappear if you wait long enough?
> 
> While the clients are still there, does 'strace -p <gem_caching_pid>' show
> any output?

It only happens on automation testing via piglit. I can't reproduce it manually.
Comment 5 lu hua 2014-11-10 02:39:03 UTC
igt_gem_multi_bsd_sync_loop also has this issue.
@test: Intel_gpu_tools/igt_gem_multi_bsd_sync_loop
info: @@@Returncode: 0

test case start at: Mon Nov 10 04:57:17 2014
test case end at:   Mon Nov 10 04:57:21 2014

Errors:


Dmesg:


Output:
             command   pid dev master a   uid      magic
     gem_multi_bsd_s  6808   0   y    y     0          0
     gem_multi_bsd_s  6808   0   n    y     0          0
     gem_multi_bsd_s  6808   0   n    y     0          0
     gem_multi_bsd_s  6808   0   n    y     0          0
     gem_multi_bsd_s  6808   0   n    y     0          0
     gem_multi_bsd_s  6808   0   n    y     0          0
     gem_multi_bsd_s  6808   0   n    y     0          0
     gem_multi_bsd_s  6808   0   n    y     0          0
     gem_multi_bsd_s  6808   0   n    y     0          0
     gem_multi_bsd_s  6808   0   n    y     0          0
     gem_multi_bsd_s  6808   0   n    y     0          0
     gem_multi_bsd_s  6808   0   n    y     0          0
     gem_multi_bsd_s  6808   0   n    y     0          0
     gem_multi_bsd_s  6808   0   n    y     0          0
     gem_multi_bsd_s  6808   0   n    y     0          0
     gem_multi_bsd_s  6808   0   n    y     0          0
     gem_multi_bsd_s  6808   0   n    y     0          0
     gem_multi_bsd_s  6808   0   n    y     0          0
     gem_multi_bsd_s  6808   0   n    y     0          0
     gem_multi_bsd_s  6808   0   n    y     0          0
     gem_multi_bsd_s  6808   0   n    y     0          0
     gem_multi_bsd_s  6808   0   n    y     0          0
     gem_multi_bsd_s  6808   0   n    y     0          0
     gem_multi_bsd_s  6808   0   n    y     0          0
     gem_multi_bsd_s  6808   0   n    y     0          0
     gem_multi_bsd_s  6808   0   n    y     0          0
     gem_multi_bsd_s  6808   0   n    y     0          0
     gem_multi_bsd_s  6808   0   n    y     0          0
     gem_multi_bsd_s  6808   0   n    y     0          0
Comment 6 lu hua 2014-11-10 02:42:06 UTC
kms_flip_wf_vblank also has this issue.
@test: Intel_gpu_tools/igt_kms_flip_wf_vblank
info: @@@Returncode: 0

test case start at: Mon Nov 10 07:59:07 2014
test case end at:   Mon Nov 10 07:59:12 2014

Errors:


Dmesg:


Output:
             command   pid dev master a   uid      magic
            kms_flip  5699   0   y    y     0          0
            kms_flip  5699   0   n    y     0          0
Test Environment check: other drm clients running!
[1/1] fail: 1 Running Test(s): 0
[1/1] fail: 1 Running Test(s): 1
Comment 7 lu hua 2014-11-10 03:11:10 UTC
kms_flip_wf_vblank also has this issue on HSW.
@test: Intel_gpu_tools/igt_kms_flip_wf_vblank
info: @@@Returncode: 0

test case start at: Sat Nov  8 12:58:45 2014
test case end at:   Sat Nov  8 12:58:49 2014

Errors:


Dmesg:


Output:
             command   pid dev master a   uid      magic
            kms_flip  6021   0   y    y     0          0
            kms_flip  6021   0   n    y     0          0
Test Environment check: other drm clients running!
[1/1] fail: 1 Running Test(s): 0
[1/1] fail: 1 Running Test(s): 1
Comment 8 lu hua 2014-12-30 05:33:26 UTC
SNB also has this issue.
Comment 9 yann 2016-12-12 16:59:14 UTC
These tests are passing and exit correctly, including when they are called via piglit, on SNB, IVB & HSW

Kernel:
commit 8f02d2103390e48e9e76c9e75d16dbb4396484c8
Author: Jani Nikula <jani.nikula@intel.com>
Date:   Mon Nov 28 14:38:48 2016 +0200

IGT:
	tag       : intel-gpu-tools-1.16-164-g7797517
	commit    : 7797517 

Please fill a new bug if this is occurring on your side.

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.