On 11/08/2026 15:42, Philipp Stanner wrote:
+Cc a few more people
On Tue, 2026-08-11 at 15:07 +0100, Tvrtko Ursulin wrote:
There was a late minute user regression report[1] so lets revert this so it can
be debugged in peace.
I have not been able to locally reproduce it yet, but at least a partial fix
looks to be this[2]. Further debug will be needed, possibly after the holiday/
vacation season is over.
[1]
https://lore.kernel.org/dri-devel/TfhgV1W0W5LI6RWUO6J35B3R8QIYH_FN3Eihzdo_9PH39hfn1AVByT-QBPHmWiAR-L0Kqi2sppM_EhFPKwZPGmb5pFgpF2MrzeFzkZDmpG8=@proton.me/
[2]
https://lore.kernel.org/dri-devel/[email protected]/
Cc: [email protected]
Cc: Matthew Brost <[email protected]>
Cc: Danilo Krummrich <[email protected]>
Cc: Philipp Stanner <[email protected]>
Cc: Christian König <[email protected]>
v2:
* Forgot the SoBs.
Tvrtko Ursulin (20):
Revert "drm/sched: Remove redundant entity->rq initialization and
checks"
Revert "drm/sched: Remove drm_sched_init_args->num_rqs"
Revert "drm/xe: Remove drm_sched_init_args->num_rqs usage"
Revert "drm/v3d: Remove drm_sched_init_args->num_rqs usage"
Revert "drm/sched: Remove drm_sched_init_args->num_rqs usage"
Revert "drm/panthor: Remove drm_sched_init_args->num_rqs usage"
Revert "drm/panfrost: Remove drm_sched_init_args->num_rqs usage"
Revert "drm/nouveau: Remove drm_sched_init_args->num_rqs usage"
Revert "drm/msm: Remove drm_sched_init_args->num_rqs usage"
Revert "drm/lima: Remove drm_sched_init_args->num_rqs usage"
Revert "drm/imagination: Remove drm_sched_init_args->num_rqs usage"
Revert "drm/etnaviv: Remove drm_sched_init_args->num_rqs usage"
Revert "drm/amdgpu: Remove drm_sched_init_args->num_rqs usage"
Revert "accel/ethosu: Remove drm_sched_init_args->num_rqs usage"
Revert "accel/rocket: Remove drm_sched_init_args->num_rqs usage"
Revert "accel/amdxdna: Remove drm_sched_init_args->num_rqs usage"
Revert "drm/sched: Embed run queue singleton into the scheduler"
Revert "drm/sched: Remove FIFO and RR and simplify to a single run
queue"
Revert "drm/sched: Switch default policy to fair"
drm/sched: Mark fair policy as experimental
Acked-by: Philipp Stanner <[email protected]>
Doesn't yet apply to drm-misc-fixes, I suppose because of the merge
yesterday.
From my POV you can push this to misc-fixes if you can make it apply;
hopefully it's not too late for v7.2
I've rebased against drm-misc-fixes, posted v3 and pushed it. I wanted
to prioritise getting the reverts in quickly to not miss the fixes pull
rq so I did not log the acks against all patches since there is a lot of
them. There were two small conflicts while building drm-tip so fingers
crossed that's all good. I compile re-tested both drm-misc-fixes with
all affected drivers, and drm-tip for x86 only.
Regards,
Tvrtko
