On 9/1/2026 08:04, Russell, Kent wrote:

+     .hqd_gfx_cleaner = kgd_gfx_v9_cleaner

Prefer a name like fault_cleaner. This is not related to a HQD.

will update

+void kgd_gfx_v9_cleaner(struct amdgpu_device *adev)
+{
+     gfx_v9_4_2_shader_cleaner_workarounds(adev);

Similarly, fault_cleaner.

+static void gfx942_workaround_cleaner(struct kfd_node *dev, uint16_t pasid)

Similarly, fault_cleaner

+     /* Run the GFX cleaner workaround on the faulting device.
+      * Only run once per page fault (evicted == 1 after first evict).
+      */
+     pdd = kfd_get_process_device_data(dev, p);
+     if (pdd) {
+             /* Only call cleaner once per page fault. Run the workaround
+              * only on the first eviction (evicted == 1 after evict above).
+              */
+             dqm = pdd->dev->dqm;
+             dqm_lock(dqm);
+             if (pdd->qpd.evicted == 1 && dqm->dev->kfd2kgd-

hqd_gfx_cleaner)

+                     dqm->dev->kfd2kgd->hqd_gfx_cleaner(dqm->dev-

adev);

+             dqm_unlock(dqm);
+     }
+
+     /* Restore all process devices after workaround completes */

I'm not sure that we want to do this. Once a process faults there is no recovery path for its queues. If we restore them then other waves may cause further faults, leading us back here.

Similarly, if the workqueue contains pipelined fault interrupts we don't want to do this multiple times. Perhaps check pdd->qpd.evicted before evicting queues and only run the workaround if (evicted == 0).