[Intel-gfx] ✗ Fi.CI.BAT: failure for lib: Ask the kernel to quiescent the GPU

[Chris Wilson]

Quoting Patchwork (2017-09-26 19:13:26)
> == Series Details ==
> 
> Series: lib: Ask the kernel to quiescent the GPU
> URL   : https://patchwork.freedesktop.org/series/30890/
> State : failure
> 
> == Summary ==
> 
> IGT patchset tested on top of latest successful build
> 2885b10f99b4beeb046e75af8b8488c229f629d3 igt/gem_exec_schedule: Ignore set-priority failures on old kernels
> 
> with latest DRM-Tip kernel build CI_DRM_3140
> c4c623d58e38 drm-tip: 2017y-09m-26d-16h-37m-12s UTC integration manifest
> 
> Test chamelium:
>         Subgroup dp-crc-fast:
>                 pass       -> FAIL       (fi-kbl-7500u) fdo#102514
> Test gem_sync:
>         Subgroup basic-all:
>                 pass       -> DMESG-FAIL (fi-blb-e6850)
>                 pass       -> DMESG-FAIL (fi-pnv-d510)
> Test gem_workarounds:
>         Subgroup basic-read:
>                 pass       -> FAIL       (fi-bdw-5557u)
>                 pass       -> FAIL       (fi-bdw-gvtdvm)
>                 pass       -> FAIL       (fi-bsw-n3050)
>                 pass       -> FAIL       (fi-skl-6700k)
>                 pass       -> FAIL       (fi-skl-6770hq)
>                 pass       -> FAIL       (fi-skl-gvtdvm)
>                 pass       -> FAIL       (fi-bxt-j4205)
>                 pass       -> FAIL       (fi-kbl-7500u)
>                 pass       -> FAIL       (fi-kbl-7560u)
>                 pass       -> FAIL       (fi-kbl-r)
>                 pass       -> FAIL       (fi-glk-1)
>                 pass       -> FAIL       (fi-cfl-s)

It is using quiescent_gpu to load a context. Abuse of the old
implementation. Furthermore the test is bogus, since it should be
executing those reads from inside the context. But we've known that
since its inception.
-Chris