linux-next.git/drivers/gpu/drm/amd/amdgpu/amdgpu_userq.c, branch master

drm/amdgpu: trigger GPU recovery when userq destroy fails to unmap a hung queue

2026-07-06T20:56:15+00:00

Destroying a hung user queue issues a MES REMOVE_QUEUE that times out, The destroy path only logged the error and freed the queue, so the next userq submission failed and forced a GPU reset attributed to an innocent workload. Kick the userq reset work when unmap fails so the GPU is recovered at destroy time. Acked-by: Alex Deucher Signed-off-by: Jesse Zhang Signed-off-by: Alex Deucher

drm/amdgpu: reject mapping a reserved doorbell to a new queue

2026-07-01T15:28:50+00:00

When creating an user-queue, the user space provides a doorbell BO handle and an offset within the bo to obtain a doorbell. However current implementation using xa_store_irq() to store a doorbell, which allows a later queue created with the same BO and offset parameters to overwrite an existing queue and doorbell mapping. This can cause problems like misrouting fence IRQ processing to a wrong queue, and mislead the cleanup process of one queue erasing the mapping of another queue. This commit fixes this issue by replacing xa_store_irq with xa_insert_irq, which rejects mapping a reserved doorbell to a newly created queue Signed-off-by: Zhu Lingshan Reviewed-by: Christian König Signed-off-by: Alex Deucher

drm/amdgpu/userq: add reset helper and identify guilty user queue

2026-07-01T15:15:36+00:00

If we get an interrupt for a bad user queue (bad opcode, etc.), add a helper to handle the reset for user queues. v2: squash in fixes v3: - schedule the reset via amdgpu_userq_start_hang_detect_work() instead of open-coding mod_delayed_work() - drop the per-queue guilty flag; always reset the queue the hang detect work belongs to, matching the non-compute reset path Co-developed-by: Alex Deucher Signed-off-by: Alex Deucher Signed-off-by: Jesse Zhang Reviewed-by: Alex Deucher Signed-off-by: Alex Deucher

drm/amdgpu: do not release the root bo after vm validate

2026-07-01T15:12:11+00:00

Make sure to not release the vm root bo after vm validation and to make that happen we moved the restore function within amdgpu_userq_vm_validate function. Also update the function name to reflect the intent. Suggested-by: Christian König Signed-off-by: Zhu Lingshan Signed-off-by: Sunil Khatri Reviewed-by: Christian König Signed-off-by: Alex Deucher

drm/amdgpu: add userq job and state transition trace events

2026-06-17T20:35:25+00:00

Add ftrace events for tracking the userq fence emit, signal and queue state transition. Signed-off-by: Pierre-Eric Pelloux-Prayer Co-developed-by: Prike Liang Reviewed-by: Christian König Signed-off-by: Alex Deucher

drm/amdgpu: add userq create and destroy tracepoints

2026-06-17T20:35:22+00:00

Add ftrace events around user queue creation and destruction to profile queue setup and teardown latency. Signed-off-by: Prike Liang Reviewed-by: Christian König Signed-off-by: Alex Deucher

drm/amdgpu: set the userq xcp_id

2026-06-17T20:35:18+00:00

Initialize the userq xcp_id. Signed-off-by: Prike Liang Acked-by: Christian König Signed-off-by: Alex Deucher

drm/amdgpu: implement per-process MES context

2026-06-17T20:20:37+00:00

MES process context is a process-level page where process specific context is saved for MES scheduler. However, current user-queue code path assigns fw_obj of a queue to MES process_context_addr when adding the queue to MES. This means every new queue from the same process would replace the previous process context address with that queue's fw_obj address. What's worse is, when user space frees a queue, its fw_obj will be freed as well, causing MES working on a NULL page pointer. This issue leads to inconsistency and crash in the scheduler. This commit allocates a process-level page for MES process contexts for a process other than queue-level Signed-off-by: Zhu Lingshan Reviewed-by: Christian König Signed-off-by: Alex Deucher

drm/amdgpu: correct reservation fence slots for userq per-vm BOs eviction

2026-06-17T20:14:16+00:00

It fixes both the move overflow and the eviction fence add for evicting these per-vm BOs. Signed-off-by: Prike Liang Reviewed-by: Christian König Signed-off-by: Alex Deucher

drm/amdgpu: Remove faulty queue before resume

2026-06-17T19:51:36+00:00

When driver already knows a bad queue but MES suspend_all is successful and MES hung queue detection doesn't detect it, remove this queue refore resume_all. Signed-off-by: Amber Lin Reviewed-by: Alex Deucher Signed-off-by: Alex Deucher