Bug 105248

Summary: [CI] igt@gem_busy@extended-* - fail - Failed assertion: write[0] == ring - and skips on other gem tests.
Product: DRI Reporter: Marta Löfstedt <marta.lofstedt>
Component: IGTAssignee: Default DRI bug account <dri-devel>
Status: CLOSED FIXED QA Contact:
Severity: normal    
Priority: medium CC: intel-gfx-bugs
Version: DRI git   
Hardware: Other   
OS: All   
Whiteboard: ReadyForDev
i915 platform: KBL i915 features: GEM/Other

Description Marta Löfstedt 2018-02-26 07:05:24 UTC 
Starting at:
https://intel-gfx-ci.01.org/tree/drm-tip/IGT_4292/shard-kbl2/igt@gem_busy@extended-bsd1.html
https://intel-gfx-ci.01.org/tree/drm-tip/IGT_4292/shard-kbl7/igt@gem_busy@extended-bsd2.html
https://intel-gfx-ci.01.org/tree/drm-tip/IGT_4292/shard-kbl1/igt@gem_busy@extended-parallel-blt.html
https://intel-gfx-ci.01.org/tree/drm-tip/IGT_4292/shard-kbl6/igt@gem_busy@extended-parallel-bsd1.html
https://intel-gfx-ci.01.org/tree/drm-tip/IGT_4292/shard-kbl6/igt@gem_busy@extended-parallel-bsd2.html
https://intel-gfx-ci.01.org/tree/drm-tip/IGT_4292/shard-kbl2/igt@gem_busy@extended-parallel-render.html
https://intel-gfx-ci.01.org/tree/drm-tip/IGT_4292/shard-kbl7/igt@gem_busy@extended-parallel-vebox.html

(gem_busy:1919) CRITICAL: Test assertion failure function one, file gem_busy.c:270:
(gem_busy:1919) CRITICAL: Failed assertion: write[0] == ring
(gem_busy:1919) CRITICAL: error: 2 != 8194
Subtest extended-bsd1 failed.

Also, a lot of gem_busy and gem_exec subtests started skipping after the same commit.

This issue is repeated on all IGT runs IGT_4292 - IGT_4297. 
Note, for the runs I am referring to; IGT_4292 - IGT_4297, have both the same IGT commit and kernel. 	

IGT-Version: 1.21-g305ebced (x86_64) (Linux: 4.16.0-rc2-CI-CI_DRM_3830+ x86_64)

The test look OK on CI_DRM_3830, this is the IGT commit that was integrated to IGT_4292:

commit 305ebcedc36e98f3118ac27a5bbde0ce7cd71a74
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date:   Wed Feb 21 14:13:16 2018 +0000

    Iterate over physical engines
    
    We current have a single for_each_engine() iterator which we use to
    generate both a set of uABI engines and a set of physical engines.
    Determining what uABI ring-id corresponds to an actual HW engine is
    tricky, so pull that out to a library function and introduce
    for_each_physical_engine() for cases where we want to issue requests
    once on each HW ring (avoiding aliasing issues).
Comment 1 Chris Wilson 2018-02-26 09:37:59 UTC 
commit d7e0cc6ca4174871660ec78a5a063d9b564e120c (HEAD, upstream/master)
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date:   Sat Feb 24 19:10:54 2018 +0000

    igt/gem_busy: Fix extended-bsd aliasing checks
    
    Although we want to specify exactly which physical engine to run on, the
    busy ioctl can only return the I915_EXEC_RING identifier, i.e. the
    aliased I915_EXEC_BSD for vcs0/vcs1. Horrors.
    
    Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=105248
    Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
    Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Comment 2 Marta Löfstedt 2018-02-26 11:53:33 UTC 
Fix integrated to IGT_4299, results are good on the igt@gem_busy@extended-*.

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.