Compartilhar via


Interface ID3D12CommandList (d3d12.h)

Uma interface da qual ID3D12GraphicsCommandList herda. Ele representa um conjunto ordenado de comandos que a GPU executa, permitindo que a extensão dê suporte a outras listas de comandos do que apenas aquelas para gráficos (como computação e cópia).

Herança

A interface ID3D12CommandList herda de ID3D12DeviceChild. ID3D12CommandList também tem estes tipos de membros:

Métodos

A interface ID3D12CommandList tem esses métodos.

 
ID3D12CommandList::GetType

Obtém o tipo da lista de comandos, como direto, pacote, computação ou cópia.

Comentários

Use ID3D12Device::CreateCommandList para criar um objeto de lista de comandos.

Consulte também ID3D12GraphicsCommandList, que deriva de ID3D12CommandList.

Uma lista de comandos corresponde a um conjunto de comandos que a GPU (unidade de processamento gráfico) executa. Os comandos definem o estado, desenham, limpam, copiam e assim por diante.

As listas de comandos do Direct3D 12 dão suporte apenas a estes 2 níveis de indireção:

  • Uma lista de comandos diretos corresponde a um buffer de comando que a GPU pode executar.
  • Um pacote só pode ser executado diretamente por meio de uma lista de comandos diretos.

Exemplos

O exemplo D3D12nBodyGravity usa ID3D12CommandList da seguinte maneira:

DWORD D3D12nBodyGravity::AsyncComputeThreadProc(int threadIndex)
{
    ID3D12CommandQueue* pCommandQueue = m_computeCommandQueue[threadIndex].Get();
    ID3D12CommandAllocator* pCommandAllocator = m_computeAllocator[threadIndex].Get();
    ID3D12GraphicsCommandList* pCommandList = m_computeCommandList[threadIndex].Get();
    ID3D12Fence* pFence = m_threadFences[threadIndex].Get();

    while (0 == InterlockedGetValue(&m_terminating))
    {
        // Run the particle simulation.
        Simulate(threadIndex);

        // Close and execute the command list.
        ThrowIfFailed(pCommandList->Close());
        ID3D12CommandList* ppCommandLists[] = { pCommandList };

        pCommandQueue->ExecuteCommandLists(1, ppCommandLists);

        // Wait for the compute shader to complete the simulation.
        UINT64 threadFenceValue = InterlockedIncrement(&m_threadFenceValues[threadIndex]);
        ThrowIfFailed(pCommandQueue->Signal(pFence, threadFenceValue));
        ThrowIfFailed(pFence->SetEventOnCompletion(threadFenceValue, m_threadFenceEvents[threadIndex]));
        WaitForSingleObject(m_threadFenceEvents[threadIndex], INFINITE);

        // Wait for the render thread to be done with the SRV so that
        // the next frame in the simulation can run.
        UINT64 renderContextFenceValue = InterlockedGetValue(&m_renderContextFenceValues[threadIndex]);
        if (m_renderContextFence->GetCompletedValue() < renderContextFenceValue)
        {
            ThrowIfFailed(pCommandQueue->Wait(m_renderContextFence.Get(), renderContextFenceValue));
            InterlockedExchange(&m_renderContextFenceValues[threadIndex], 0);
        }

        // Swap the indices to the SRV and UAV.
        m_srvIndex[threadIndex] = 1 - m_srvIndex[threadIndex];

        // Prepare for the next frame.
        ThrowIfFailed(pCommandAllocator->Reset());
        ThrowIfFailed(pCommandList->Reset(pCommandAllocator, m_computeState.Get()));
    }

    return 0;
}

Consulte o Código de Exemplo na Referência D3D12.

Requisitos

   
Plataforma de Destino Windows
Cabeçalho d3d12.h

Confira também

Interfaces principais

ID3D12DeviceChild

ID3D12GraphicsCommandList

Definindo heaps de descritor