drm/amdgpu: rework how we handle TLB fences

Add a new VM flag to indicate whether or not we need
a TLB fence.  Userqs (KFD or KGD) require a TLB fence.
A TLB fence is not strictly required for kernel queues,
but it shouldn't hurt.  That said, enabling this
unconditionally should be fine, but it seems to tickle
some issues in KIQ/MES.  Only enable them for KFD,
or when KGD userq queues are enabled (currently via module
parameter).

Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/4798
Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/4749
Fixes: f3854e04b7 ("drm/amdgpu: attach tlb fence to the PTs update")
Cc: Christian König <christian.koenig@amd.com>
Cc: Prike Liang <Prike.Liang@amd.com>
Reviewed-by: Prike Liang <Prike.Liang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit 69c5fbd2b93b5ced77c6e79afe83371bca84c788)
Cc: stable@vger.kernel.org
This commit is contained in:
Alex Deucher
2026-03-16 11:04:46 -04:00
parent 3fc4648b53
commit e9f58ff991
2 changed files with 8 additions and 1 deletions

View File

@@ -1069,7 +1069,10 @@ amdgpu_vm_tlb_flush(struct amdgpu_vm_update_params *params,
}
/* Prepare a TLB flush fence to be attached to PTs */
if (!params->unlocked) {
/* The check for need_tlb_fence should be dropped once we
* sort out the issues with KIQ/MES TLB invalidation timeouts.
*/
if (!params->unlocked && vm->need_tlb_fence) {
amdgpu_vm_tlb_fence_create(params->adev, vm, fence);
/* Makes sure no PD/PT is freed before the flush */
@@ -2602,6 +2605,7 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm,
ttm_lru_bulk_move_init(&vm->lru_bulk_move);
vm->is_compute_context = false;
vm->need_tlb_fence = amdgpu_userq_enabled(&adev->ddev);
vm->use_cpu_for_update = !!(adev->vm_manager.vm_update_mode &
AMDGPU_VM_USE_CPU_FOR_GFX);
@@ -2739,6 +2743,7 @@ int amdgpu_vm_make_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm)
dma_fence_put(vm->last_update);
vm->last_update = dma_fence_get_stub();
vm->is_compute_context = true;
vm->need_tlb_fence = true;
unreserve_bo:
amdgpu_bo_unreserve(vm->root.bo);

View File

@@ -441,6 +441,8 @@ struct amdgpu_vm {
struct ttm_lru_bulk_move lru_bulk_move;
/* Flag to indicate if VM is used for compute */
bool is_compute_context;
/* Flag to indicate if VM needs a TLB fence (KFD or KGD) */
bool need_tlb_fence;
/* Memory partition number, -1 means any partition */
int8_t mem_id;