diff options
Diffstat (limited to 'drivers/gpu/drm/amd/amdkfd')
| -rw-r--r-- | drivers/gpu/drm/amd/amdkfd/kfd_chardev.c | 7 | ||||
| -rw-r--r-- | drivers/gpu/drm/amd/amdkfd/kfd_migrate.c | 16 | ||||
| -rw-r--r-- | drivers/gpu/drm/amd/amdkfd/kfd_process.c | 23 | ||||
| -rw-r--r-- | drivers/gpu/drm/amd/amdkfd/kfd_svm.c | 2 | ||||
| -rw-r--r-- | drivers/gpu/drm/amd/amdkfd/kfd_topology.c | 5 |
5 files changed, 39 insertions, 14 deletions
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c index 55aa74cbc532..1e6cc0bfc432 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c @@ -1139,7 +1139,7 @@ static int kfd_ioctl_alloc_memory_of_gpu(struct file *filep, goto err_unlock; } offset = dev->adev->rmmio_remap.bus_addr; - if (!offset) { + if (!offset || (PAGE_SIZE > 4096)) { err = -ENOMEM; goto err_unlock; } @@ -2307,7 +2307,7 @@ static int criu_restore_memory_of_gpu(struct kfd_process_device *pdd, return -EINVAL; } offset = pdd->dev->adev->rmmio_remap.bus_addr; - if (!offset) { + if (!offset || (PAGE_SIZE > 4096)) { pr_err("amdgpu_amdkfd_get_mmio_remap_phys_addr failed\n"); return -ENOMEM; } @@ -3349,6 +3349,9 @@ static int kfd_mmio_mmap(struct kfd_node *dev, struct kfd_process *process, if (vma->vm_end - vma->vm_start != PAGE_SIZE) return -EINVAL; + if (PAGE_SIZE > 4096) + return -EINVAL; + address = dev->adev->rmmio_remap.bus_addr; vm_flags_set(vma, VM_IO | VM_DONTCOPY | VM_DONTEXPAND | VM_NORESERVE | diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c b/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c index bdc01ca9609a..5c8d81bfce7a 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c @@ -509,10 +509,19 @@ svm_migrate_ram_to_vram(struct svm_range *prange, uint32_t best_loc, start = start_mgr << PAGE_SHIFT; end = (last_mgr + 1) << PAGE_SHIFT; + r = amdgpu_amdkfd_reserve_mem_limit(node->adev, + prange->npages * PAGE_SIZE, + KFD_IOC_ALLOC_MEM_FLAGS_VRAM, + node->xcp ? node->xcp->id : 0); + if (r) { + dev_dbg(node->adev->dev, "failed to reserve VRAM, r: %ld\n", r); + return -ENOSPC; + } + r = svm_range_vram_node_new(node, prange, true); if (r) { dev_dbg(node->adev->dev, "fail %ld to alloc vram\n", r); - return r; + goto out; } ttm_res_offset = (start_mgr - prange->start + prange->offset) << PAGE_SHIFT; @@ -545,6 +554,11 @@ svm_migrate_ram_to_vram(struct svm_range *prange, uint32_t best_loc, svm_range_vram_node_free(prange); } +out: + amdgpu_amdkfd_unreserve_mem_limit(node->adev, + prange->npages * PAGE_SIZE, + KFD_IOC_ALLOC_MEM_FLAGS_VRAM, + node->xcp ? node->xcp->id : 0); return r < 0 ? r : 0; } diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process.c b/drivers/gpu/drm/amd/amdkfd/kfd_process.c index b79986412cd8..451bb058cc62 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_process.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_process.c @@ -829,6 +829,14 @@ struct kfd_process *kfd_create_process(struct task_struct *thread) if (process) { pr_debug("Process already found\n"); } else { + /* If the process just called exec(3), it is possible that the + * cleanup of the kfd_process (following the release of the mm + * of the old process image) is still in the cleanup work queue. + * Make sure to drain any job before trying to recreate any + * resource for this process. + */ + flush_workqueue(kfd_process_wq); + process = create_process(thread); if (IS_ERR(process)) goto out; @@ -1922,6 +1930,8 @@ static int signal_eviction_fence(struct kfd_process *p) rcu_read_lock(); ef = dma_fence_get_rcu_safe(&p->ef); rcu_read_unlock(); + if (!ef) + return -EINVAL; ret = dma_fence_signal(ef); dma_fence_put(ef); @@ -1949,10 +1959,9 @@ static void evict_process_worker(struct work_struct *work) * they are responsible stopping the queues and scheduling * the restore work. */ - if (!signal_eviction_fence(p)) - queue_delayed_work(kfd_restore_wq, &p->restore_work, - msecs_to_jiffies(PROCESS_RESTORE_TIME_MS)); - else + if (signal_eviction_fence(p) || + mod_delayed_work(kfd_restore_wq, &p->restore_work, + msecs_to_jiffies(PROCESS_RESTORE_TIME_MS))) kfd_process_restore_queues(p); pr_debug("Finished evicting pasid 0x%x\n", p->pasid); @@ -2011,9 +2020,9 @@ static void restore_process_worker(struct work_struct *work) if (ret) { pr_debug("Failed to restore BOs of pasid 0x%x, retry after %d ms\n", p->pasid, PROCESS_BACK_OFF_TIME_MS); - ret = queue_delayed_work(kfd_restore_wq, &p->restore_work, - msecs_to_jiffies(PROCESS_BACK_OFF_TIME_MS)); - WARN(!ret, "reschedule restore work failed\n"); + if (mod_delayed_work(kfd_restore_wq, &p->restore_work, + msecs_to_jiffies(PROCESS_RESTORE_TIME_MS))) + kfd_process_restore_queues(p); } } diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c index f0f7f48af413..386875e6eb96 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c @@ -3426,7 +3426,7 @@ svm_range_trigger_migration(struct mm_struct *mm, struct svm_range *prange, mm, KFD_MIGRATE_TRIGGER_PREFETCH); *migrated = !r; - return r; + return 0; } int svm_range_schedule_evict_svm_bo(struct amdgpu_amdkfd_fence *fence) diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c index c51f131eaa2f..bc9eb847ecfe 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c @@ -1997,9 +1997,8 @@ int kfd_topology_add_device(struct kfd_node *gpu) HSA_CAP_ASIC_REVISION_MASK); dev->node_props.location_id = pci_dev_id(gpu->adev->pdev); - /* On multi-partition nodes, node id = location_id[31:28] */ - if (gpu->kfd->num_nodes > 1) - dev->node_props.location_id |= (dev->gpu->node_id << 28); + if (KFD_GC_VERSION(dev->gpu->kfd) == IP_VERSION(9, 4, 3)) + dev->node_props.location_id |= dev->gpu->node_id; dev->node_props.domain = pci_domain_nr(gpu->adev->pdev->bus); dev->node_props.max_engine_clk_fcompute = |