20:34 openglfreak: Hmm, creating a userqueue in a process, i.e. successfully doing an DRM_AMDGPU_USERQ ioctl, makes the process take an additional 200 * $RANDOM milliseconds to disappear, any ideas why? My grep-fu is failing me here
22:34 alyssa: is amd's lower_load_barycentric_at_offset perspective-correct?
22:34 alyssa: I see like 3 different lowerings within src/amd alone
23:54 lumag: Hmm, valentine, vignesh_, I'm getting a weird error from ci_run_n_monitor when run against drm/msm ("Error while fetching schema: {'message': 'Internal server error'}"). Any ideas?