[SRU][Q][PATCH 3/3] drm/amdgpu: rework how we handle TLB fences

[Aaron Ma]

From: Alex Deucher <alexander.deucher at amd.com>

BugLink: https://bugs.launchpad.net/bugs/2148538

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: f3854e04b708 ("drm/amdgpu: attach tlb fence to the PTs update")
Cc: Christian König <christian.koenig at amd.com>
Cc: Prike Liang <Prike.Liang at amd.com>
Reviewed-by: Prike Liang <Prike.Liang at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
(cherry picked from commit 69c5fbd2b93b5ced77c6e79afe83371bca84c788)
Cc: stable at vger.kernel.org
(cherry picked from commit e9f58ff991dd4be13fd7a651bbf64329c090af09)
Signed-off-by: Aaron Ma <aaron.ma at canonical.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 7 ++++++-
 drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h | 2 ++
 2 files changed, 8 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
index 0d610091c7316..cccb8bf23f003 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
@@ -1056,7 +1056,10 @@ 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 */
@@ -2571,6 +2574,7 @@ 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);
@@ -2694,6 +2698,7 @@ 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);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
index fd086efd8457e..e0cb7197facae 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
@@ -421,6 +421,8 @@ 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;
-- 
2.43.0