diff --git a/libs/vkd3d/command.c b/libs/vkd3d/command.c index 6c9b145f..0d30f430 100644 --- a/libs/vkd3d/command.c +++ b/libs/vkd3d/command.c @@ -80,6 +80,29 @@ void vkd3d_queue_release(struct vkd3d_queue *queue) pthread_mutex_unlock(&queue->mutex); } +static VkResult vkd3d_queue_wait_idle(struct vkd3d_queue *queue, + const struct vkd3d_vk_device_procs *vk_procs) +{ + VkQueue vk_queue; + VkResult vr; + + if ((vk_queue = vkd3d_queue_acquire(queue))) + { + vr = VK_CALL(vkQueueWaitIdle(vk_queue)); + vkd3d_queue_release(queue); + + if (vr < 0) + WARN("Failed to wait for queue, vr %d.\n", vr); + } + else + { + ERR("Failed to acquire queue %p.\n", queue); + vr = VK_ERROR_OUT_OF_HOST_MEMORY; + } + + return vr; +} + /* Fence worker thread */ static HRESULT vkd3d_enqueue_gpu_fence(struct vkd3d_fence_worker *worker, VkFence vk_fence, struct d3d12_fence *fence, uint64_t value) @@ -377,6 +400,104 @@ static struct d3d12_fence *impl_from_ID3D12Fence(ID3D12Fence *iface) return CONTAINING_RECORD(iface, struct d3d12_fence, ID3D12Fence_iface); } +#define VKD3D_MAX_VK_SEMAPHORE_COUNT_PER_D3D12_FENCE 4 + +static void d3d12_fence_destroy_vk_semaphores_locked(struct d3d12_fence *fence, bool destroy_all) +{ + struct d3d12_device *device = fence->device; + const struct vkd3d_vk_device_procs *vk_procs = &device->vk_procs; + VkSemaphore vk_semaphore; + size_t semaphore_count; + unsigned int i, j; + + semaphore_count = fence->semaphore_count; + if (!destroy_all && fence->semaphore_count < VKD3D_MAX_VK_SEMAPHORE_COUNT_PER_D3D12_FENCE) + return; + + for (i = 0, j = 0; i < fence->semaphore_count; ++i) + { + if (!destroy_all && semaphore_count < VKD3D_MAX_VK_SEMAPHORE_COUNT_PER_D3D12_FENCE) + break; + + /* The semaphore doesn't have a pending signal operation if the fence + * was signaled. */ + if (!fence->semaphores[i].vk_fence || destroy_all) + { + vk_semaphore = fence->semaphores[i].vk_semaphore; + if (fence->semaphores[i].vk_fence) + WARN("Destroying potentially pending semaphore.\n"); + VK_CALL(vkDestroySemaphore(device->vk_device, vk_semaphore, NULL)); + --semaphore_count; + } + else + { + if (i != j) + fence->semaphores[j] = fence->semaphores[i]; + ++j; + } + } + for (; i < fence->semaphore_count; ++i) + { + fence->semaphores[j++] = fence->semaphores[i]; + } + + if (j != fence->semaphore_count) + TRACE("Destroyed %zu Vulkan semaphores.\n", fence->semaphore_count - j); + + fence->semaphore_count = j; +} + +static void d3d12_fence_destroy_vk_semaphores(struct d3d12_fence *fence) +{ + int rc; + + if ((rc = pthread_mutex_lock(&fence->mutex))) + { + ERR("Failed to lock mutex, error %d.\n", rc); + return; + } + + d3d12_fence_destroy_vk_semaphores_locked(fence, true); + + pthread_mutex_unlock(&fence->mutex); +} + +static HRESULT d3d12_fence_add_vk_semaphore(struct d3d12_fence *fence, + VkSemaphore vk_semaphore, VkFence vk_fence, uint64_t value) +{ + HRESULT hr = S_OK; + int rc; + + TRACE("fence %p, value %#"PRIx64".\n", fence, value); + + if ((rc = pthread_mutex_lock(&fence->mutex))) + { + ERR("Failed to lock mutex, error %d.\n", rc); + return E_FAIL; + } + + /* Destroy old Vulkan semaphores. */ + d3d12_fence_destroy_vk_semaphores_locked(fence, false); + + if (vkd3d_array_reserve((void **)&fence->semaphores, &fence->semaphores_size, + fence->semaphore_count + 1, sizeof(*fence->semaphores))) + { + fence->semaphores[fence->semaphore_count].value = value; + fence->semaphores[fence->semaphore_count].vk_semaphore = vk_semaphore; + fence->semaphores[fence->semaphore_count].vk_fence = vk_fence; + ++fence->semaphore_count; + } + else + { + ERR("Failed to add semaphore.\n"); + hr = E_OUTOFMEMORY; + } + + pthread_mutex_unlock(&fence->mutex); + + return hr; +} + static HRESULT d3d12_fence_signal(struct d3d12_fence *fence, uint64_t value, VkFence vk_fence) { struct d3d12_device *device = fence->device; @@ -414,6 +535,17 @@ static HRESULT d3d12_fence_signal(struct d3d12_fence *fence, uint64_t value, VkF } fence->event_count = j; + if (vk_fence) + { + for (i = 0; i < fence->semaphore_count; ++i) + { + struct vkd3d_signaled_semaphore *current = &fence->semaphores[i]; + + if (current->vk_fence == vk_fence) + current->vk_fence = VK_NULL_HANDLE; + } + } + pthread_mutex_unlock(&fence->mutex); return S_OK; @@ -467,6 +599,10 @@ static ULONG STDMETHODCALLTYPE d3d12_fence_Release(ID3D12Fence *iface) vkd3d_fence_worker_remove_fence(&device->fence_worker, fence); + /* Destroy all Vulkan semaphores associated with this fence. */ + d3d12_fence_destroy_vk_semaphores(fence); + vkd3d_free(fence->semaphores); + vkd3d_free(fence->events); if ((rc = pthread_mutex_destroy(&fence->mutex))) ERR("Failed to destroy mutex, error %d.\n", rc); @@ -656,6 +792,9 @@ static HRESULT d3d12_fence_init(struct d3d12_fence *fence, struct d3d12_device * fence->events = NULL; fence->events_size = 0; fence->event_count = 0; + fence->semaphores = NULL; + fence->semaphores_size = 0; + fence->semaphore_count = 0; if (FAILED(hr = vkd3d_private_store_init(&fence->private_store))) { @@ -4915,58 +5054,106 @@ static HRESULT STDMETHODCALLTYPE d3d12_command_queue_Signal(ID3D12CommandQueue * { struct d3d12_command_queue *command_queue = impl_from_ID3D12CommandQueue(iface); const struct vkd3d_vk_device_procs *vk_procs; + VkSemaphore vk_semaphore = VK_NULL_HANDLE; + VkSemaphoreCreateInfo semaphore_info; + VkFence vk_fence = VK_NULL_HANDLE; VkFenceCreateInfo fence_info; struct d3d12_device *device; struct d3d12_fence *fence; - VkFence vk_fence; + VkSubmitInfo submit_info; VkQueue vk_queue; VkResult vr; + HRESULT hr; TRACE("iface %p, fence %p, value %#"PRIx64".\n", iface, fence_iface, value); device = command_queue->device; vk_procs = &device->vk_procs; + /* XXX: It may be a good idea to re-use Vulkan fences/semaphores. */ + fence = unsafe_impl_from_ID3D12Fence(fence_iface); fence_info.sType = VK_STRUCTURE_TYPE_FENCE_CREATE_INFO; fence_info.pNext = NULL; fence_info.flags = 0; - - /* XXX: It may be a good idea to re-use Vulkan fences. */ if ((vr = VK_CALL(vkCreateFence(device->vk_device, &fence_info, NULL, &vk_fence))) < 0) { WARN("Failed to create Vulkan fence, vr %d.\n", vr); - return hresult_from_vk_result(vr); + vk_fence = VK_NULL_HANDLE; + goto fail_vkresult; } + semaphore_info.sType = VK_STRUCTURE_TYPE_SEMAPHORE_CREATE_INFO; + semaphore_info.pNext = NULL; + semaphore_info.flags = 0; + if ((vr = VK_CALL(vkCreateSemaphore(device->vk_device, &semaphore_info, NULL, &vk_semaphore))) < 0) + { + WARN("Failed to create Vulkan semaphore, vr %d.\n", vr); + vk_semaphore = VK_NULL_HANDLE; + goto fail_vkresult; + } + + submit_info.sType = VK_STRUCTURE_TYPE_SUBMIT_INFO; + submit_info.pNext = NULL; + submit_info.waitSemaphoreCount = 0; + submit_info.pWaitSemaphores = NULL; + submit_info.pWaitDstStageMask = NULL; + submit_info.commandBufferCount = 0; + submit_info.pCommandBuffers = NULL; + submit_info.signalSemaphoreCount = 1; + submit_info.pSignalSemaphores = &vk_semaphore; + if (!(vk_queue = vkd3d_queue_acquire(command_queue->vkd3d_queue))) { ERR("Failed to acquire queue %p.\n", command_queue->vkd3d_queue); - return E_FAIL; + hr = E_FAIL; + goto fail; } - vr = VK_CALL(vkQueueSubmit(vk_queue, 0, NULL, vk_fence)); + vr = VK_CALL(vkQueueSubmit(vk_queue, 1, &submit_info, vk_fence)); vkd3d_queue_release(command_queue->vkd3d_queue); if (vr < 0) { - WARN("Failed to submit fence, vr %d.\n", vr); - VK_CALL(vkDestroyFence(device->vk_device, vk_fence, NULL)); - return hresult_from_vk_result(vr); + WARN("Failed to submit, vr %d.\n", vr); + goto fail_vkresult; } + if (SUCCEEDED(hr = d3d12_fence_add_vk_semaphore(fence, vk_semaphore, vk_fence, value))) + vk_semaphore = VK_NULL_HANDLE; + vr = VK_CALL(vkGetFenceStatus(device->vk_device, vk_fence)); - if (vr == VK_SUCCESS) + if (vr == VK_NOT_READY) { - return d3d12_fence_signal(fence, value, vk_fence); + if (SUCCEEDED(hr = vkd3d_enqueue_gpu_fence(&device->fence_worker, vk_fence, fence, value))) + vk_fence = VK_NULL_HANDLE; } - if (vr != VK_NOT_READY) + else if (vr == VK_SUCCESS) { - WARN("Failed to get fence status, vr %d.\n", vr); - VK_CALL(vkDestroyFence(device->vk_device, vk_fence, NULL)); - return hresult_from_vk_result(vr); + TRACE("Already signaled %p, value %#"PRIx64".\n", fence, value); + hr = d3d12_fence_signal(fence, value, vk_fence); + vk_fence = VK_NULL_HANDLE; + } + else + { + FIXME("Failed to get fence status, vr %d.\n", vr); + hr = hresult_from_vk_result(vr); } - return vkd3d_enqueue_gpu_fence(&device->fence_worker, vk_fence, fence, value); + if (vk_fence || vk_semaphore) + { + /* In case of an unexpected failure, try to safely destroy Vulkan objects. */ + vkd3d_queue_wait_idle(command_queue->vkd3d_queue, vk_procs); + goto fail; + } + + return hr; + +fail_vkresult: + hr = hresult_from_vk_result(vr); +fail: + VK_CALL(vkDestroyFence(device->vk_device, vk_fence, NULL)); + VK_CALL(vkDestroySemaphore(device->vk_device, vk_semaphore, NULL)); + return hr; } static HRESULT STDMETHODCALLTYPE d3d12_command_queue_Wait(ID3D12CommandQueue *iface, diff --git a/libs/vkd3d/vkd3d_private.h b/libs/vkd3d/vkd3d_private.h index 68a51c0d..3ca80c0c 100644 --- a/libs/vkd3d/vkd3d_private.h +++ b/libs/vkd3d/vkd3d_private.h @@ -260,6 +260,15 @@ struct d3d12_fence size_t events_size; size_t event_count; + struct vkd3d_signaled_semaphore + { + uint64_t value; + VkSemaphore vk_semaphore; + VkFence vk_fence; + } *semaphores; + size_t semaphores_size; + size_t semaphore_count; + struct d3d12_device *device; struct vkd3d_private_store private_store;