Bug 109049 - [CI][BAT] igt@amdgpu/amd_prime@amd-to-i915 - fail - Failed assertion: r == 0, Last errno: 28, No space left on device
Summary: [CI][BAT] igt@amdgpu/amd_prime@amd-to-i915 - fail - Failed assertion: r == 0,...
Status: CLOSED FIXED
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/AMDgpu (show other bugs)
Version: XOrg git
Hardware: Other All
: highest normal
Assignee: Default DRI bug account
QA Contact:
URL:
Whiteboard: ReadyForDev
Keywords:
Depends on:
Blocks:
 
Reported: 2018-12-13 11:37 UTC by Martin Peres
Modified: 2019-01-04 14:05 UTC (History)
1 user (show)

See Also:
i915 platform: SKL
i915 features: display/Other

Attachments

Description Martin Peres 2018-12-13 11:37:47 UTC 
https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_5311/fi-kbl-8809g/igt@amdgpu_amd_prime@amd-to-i915.html

Starting subtest: amd-to-i915
(amd_prime:4058) CRITICAL: Test assertion failure function amd_to_i915, file ../tests/amdgpu/amd_prime.c:358:
(amd_prime:4058) CRITICAL: Failed assertion: r == 0
(amd_prime:4058) CRITICAL: Last errno: 28, No space left on device
(amd_prime:4058) CRITICAL: error: -28 != 0
Subtest amd-to-i915 failed.
Comment 1 Chris Wilson 2018-12-13 11:58:32 UTC 
Previously we created 4147 contexts & submissions within a 2s setup period.
Comment 2 Martin Peres 2018-12-13 12:21:59 UTC 
Moving the bug back to i915 because of the following issue:

https://intel-gfx-ci.01.org/tree/drm-tip/drmtip_168/fi-skl-6700k2/igt@kms_atomic_transition@plane-all-modeset-transition.html

Starting subtest: plane-all-modeset-transition
(kms_atomic_transition:1188) igt_kms-CRITICAL: Test assertion failure function do_display_commit, file ../lib/igt_kms.c:3320:
(kms_atomic_transition:1188) igt_kms-CRITICAL: Failed assertion: ret == 0
(kms_atomic_transition:1188) igt_kms-CRITICAL: Last errno: 28, No space left on device
(kms_atomic_transition:1188) igt_kms-CRITICAL: error: -28 != 0
Subtest plane-all-modeset-transition failed.
Comment 3 Chris Wilson 2018-12-13 13:09:11 UTC 
(In reply to Martin Peres from comment #2)
> Moving the bug back to i915 because of the following issue:
> 
> https://intel-gfx-ci.01.org/tree/drm-tip/drmtip_168/fi-skl-6700k2/
> igt@kms_atomic_transition@plane-all-modeset-transition.html
> 
> Starting subtest: plane-all-modeset-transition
> (kms_atomic_transition:1188) igt_kms-CRITICAL: Test assertion failure
> function do_display_commit, file ../lib/igt_kms.c:3320:
> (kms_atomic_transition:1188) igt_kms-CRITICAL: Failed assertion: ret == 0
> (kms_atomic_transition:1188) igt_kms-CRITICAL: Last errno: 28, No space left
> on device
> (kms_atomic_transition:1188) igt_kms-CRITICAL: error: -28 != 0
> Subtest plane-all-modeset-transition failed.

Nothing to do with the amdgpu ENOSPC.
Comment 4 Chris Wilson 2018-12-13 13:09:36 UTC 
Now limited to 4095 contexts. Isn't that a coincidence!
Comment 5 Chris Wilson 2018-12-13 13:12:03 UTC 
Regression from

commit 08d1bdd4cc57e89d037205687a61d2b6ff9dddd4
Author: Rex Zhu <Rex.Zhu@amd.com>
Date:   Wed Oct 24 16:10:33 2018 +0800

    drm/amdgpu: Limit vm max ctx number to 4096
    
    driver need to reserve resource for each ctx for
    some hw features. so add this limitation.
    
    Reviewed-by: Christian König <christian.koenig@amd.com>
    Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
    Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Comment 6 Chris Wilson 2018-12-17 15:25:36 UTC 
Stopped being so stringent about allocating so many contexts, and just focus on testing fence import:

commit 270da20849db4d170db09673c6b67712c90ec9fe (HEAD, upstream/master)
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date:   Thu Dec 13 11:54:36 2018 +0000

    igt/amdgpu_amd_prime: Bail if we fail to create more contexts
    
    amdgpu has started to report out of space after creating a few contexts.
    This is not the scope of this test as here we are just verifying that
    fences created in amd can be imported and used for synchronisation by
    i915 and for that we just need at least one context created!
    
    References: https://bugs.freedesktop.org/show_bug.cgi?id=109049
    Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
    Reviewed-by: Antonio Argenziano <antonio.argenziano@intel.com>
    Acked-by: Alex Deucher <alexander.deucher@amd.com>
Comment 7 Martin Peres 2019-01-04 14:05:36 UTC 
(In reply to Chris Wilson from comment #6)
> Stopped being so stringent about allocating so many contexts, and just focus
> on testing fence import:
> 
> commit 270da20849db4d170db09673c6b67712c90ec9fe (HEAD, upstream/master)
> Author: Chris Wilson <chris@chris-wilson.co.uk>
> Date:   Thu Dec 13 11:54:36 2018 +0000
> 
>     igt/amdgpu_amd_prime: Bail if we fail to create more contexts
>     
>     amdgpu has started to report out of space after creating a few contexts.
>     This is not the scope of this test as here we are just verifying that
>     fences created in amd can be imported and used for synchronisation by
>     i915 and for that we just need at least one context created!
>     
>     References: https://bugs.freedesktop.org/show_bug.cgi?id=109049
>     Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
>     Reviewed-by: Antonio Argenziano <antonio.argenziano@intel.com>
>     Acked-by: Alex Deucher <alexander.deucher@amd.com>

Thanks, it did the trick!

Archiving :)

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.