diff options
Diffstat (limited to 'drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c')
| -rw-r--r-- | drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 91 |
1 files changed, 59 insertions, 32 deletions
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index f2beb980e3c3..73abac6be5b3 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -139,6 +139,20 @@ static void amdgpu_vm_assert_locked(struct amdgpu_vm *vm) } /** + * amdgpu_vm_is_bo_always_valid - check if the BO is VM always valid + * + * @vm: VM to test against. + * @bo: BO to be tested. + * + * Returns true if the BO shares the dma_resv object with the root PD and is + * always guaranteed to be valid inside the VM. + */ +bool amdgpu_vm_is_bo_always_valid(struct amdgpu_vm *vm, struct amdgpu_bo *bo) +{ + return bo && bo->tbo.base.resv == vm->root.bo->tbo.base.resv; +} + +/** * amdgpu_vm_bo_evicted - vm_bo is evicted * * @vm_bo: vm_bo which is evicted @@ -764,12 +778,9 @@ bool amdgpu_vm_need_pipeline_sync(struct amdgpu_ring *ring, * @need_pipe_sync: is pipe sync needed * * Emit a VM flush when it is necessary. - * - * Returns: - * 0 on success, errno otherwise. */ -int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job, - bool need_pipe_sync) +void amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job, + bool need_pipe_sync) { struct amdgpu_device *adev = ring->adev; struct amdgpu_isolation *isolation = &adev->isolation[ring->xcp_id]; @@ -783,8 +794,7 @@ int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job, bool cleaner_shader_needed = false; bool pasid_mapping_needed = false; struct dma_fence *fence = NULL; - unsigned int patch; - int r; + unsigned int patch = 0; if (amdgpu_vmid_had_gpu_reset(adev, id)) { gds_switch_needed = true; @@ -812,9 +822,20 @@ int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job, if (!vm_flush_needed && !gds_switch_needed && !need_pipe_sync && !cleaner_shader_needed) - return 0; + return; amdgpu_ring_ib_begin(ring); + + /* There is no matching insert_end for this on purpose for the vm flush. + * The IB portion of the submission has both. Having multiple + * insert_start sequences is ok, but you can only have one insert_end + * per submission based on the way VCN FW works. For JPEG + * you can as many insert_start and insert_end sequences as you like as + * long as the rest of the packets come between start and end sequences. + */ + if (ring->funcs->insert_start) + ring->funcs->insert_start(ring); + if (ring->funcs->init_cond_exec) patch = amdgpu_ring_init_cond_exec(ring, ring->cond_exe_gpu_addr); @@ -845,9 +866,7 @@ int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job, } if (vm_flush_needed || pasid_mapping_needed || cleaner_shader_needed) { - r = amdgpu_fence_emit(ring, job->hw_vm_fence, 0); - if (r) - return r; + amdgpu_fence_emit(ring, job->hw_vm_fence, 0); fence = &job->hw_vm_fence->base; /* get a ref for the job */ dma_fence_get(fence); @@ -892,7 +911,6 @@ int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job, } amdgpu_ring_ib_end(ring); - return 0; } /** @@ -1069,7 +1087,10 @@ 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 +2623,7 @@ 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 +2761,7 @@ 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); @@ -2785,8 +2808,8 @@ void amdgpu_vm_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm) dma_fence_put(vm->last_unlocked); dma_fence_wait(vm->last_tlb_flush, false); /* Make sure that all fence callbacks have completed */ - spin_lock_irqsave(vm->last_tlb_flush->lock, flags); - spin_unlock_irqrestore(vm->last_tlb_flush->lock, flags); + dma_fence_lock_irqsave(vm->last_tlb_flush, flags); + dma_fence_unlock_irqrestore(vm->last_tlb_flush, flags); dma_fence_put(vm->last_tlb_flush); list_for_each_entry_safe(mapping, tmp, &vm->freed, list) { @@ -2893,6 +2916,7 @@ void amdgpu_vm_manager_fini(struct amdgpu_device *adev) xa_destroy(&adev->vm_manager.pasids); amdgpu_vmid_mgr_fini(adev); + amdgpu_pasid_mgr_cleanup(); } /** @@ -2968,14 +2992,14 @@ bool amdgpu_vm_handle_fault(struct amdgpu_device *adev, u32 pasid, if (!root) return false; - addr /= AMDGPU_GPU_PAGE_SIZE; - if (is_compute_context && !svm_range_restore_pages(adev, pasid, vmid, - node_id, addr, ts, write_fault)) { + node_id, addr >> PAGE_SHIFT, ts, write_fault)) { amdgpu_bo_unref(&root); return true; } + addr /= AMDGPU_GPU_PAGE_SIZE; + r = amdgpu_bo_reserve(root, true); if (r) goto error_unref; @@ -3188,20 +3212,6 @@ void amdgpu_vm_update_fault_cache(struct amdgpu_device *adev, xa_unlock_irqrestore(&adev->vm_manager.pasids, flags); } -/** - * amdgpu_vm_is_bo_always_valid - check if the BO is VM always valid - * - * @vm: VM to test against. - * @bo: BO to be tested. - * - * Returns true if the BO shares the dma_resv object with the root PD and is - * always guaranteed to be valid inside the VM. - */ -bool amdgpu_vm_is_bo_always_valid(struct amdgpu_vm *vm, struct amdgpu_bo *bo) -{ - return bo && bo->tbo.base.resv == vm->root.bo->tbo.base.resv; -} - void amdgpu_vm_print_task_info(struct amdgpu_device *adev, struct amdgpu_task_info *task_info) { @@ -3210,3 +3220,20 @@ void amdgpu_vm_print_task_info(struct amdgpu_device *adev, task_info->process_name, task_info->tgid, task_info->task.comm, task_info->task.pid); } + +void amdgpu_sdma_set_vm_pte_scheds(struct amdgpu_device *adev, + const struct amdgpu_vm_pte_funcs *vm_pte_funcs) +{ + struct drm_gpu_scheduler *sched; + int i; + + for (i = 0; i < adev->sdma.num_instances; i++) { + if (adev->sdma.has_page_queue) + sched = &adev->sdma.instance[i].page.sched; + else + sched = &adev->sdma.instance[i].ring.sched; + adev->vm_manager.vm_pte_scheds[i] = sched; + } + adev->vm_manager.vm_pte_num_scheds = adev->sdma.num_instances; + adev->vm_manager.vm_pte_funcs = vm_pte_funcs; +} |