GPUCommandEncoder: finish() method

{{APIRef("WebGPU API")}}{{SecureContext_Header}}{{AvailableInWorkers}}

The finish() method of the {{domxref("GPUCommandEncoder")}} interface completes recording of the command sequence encoded on this GPUCommandEncoder, returning a corresponding {{domxref("GPUCommandBuffer")}}.

Syntax

finish()
finish(descriptor)

Parameters

• descriptor {{optional_inline}}
  • : An object that can contain the following properties:
    • label {{optional_inline}}
      • : A string providing a label for the returned {{domxref("GPUCommandBuffer")}} that can be used to identify it, for example in {{domxref("GPUError")}} messages or console warnings.

Return value

A {{domxref("GPUCommandBuffer")}} object instance.

Validation

The following criteria must be met when calling finish(), otherwise a {{domxref("GPUValidationError")}} is generated and the {{domxref("GPUCommandEncoder")}} becomes invalid:

• There are no active debug groups (i.e., started via {{domxref("GPUCommandEncoder.pushDebugGroup", "pushDebugGroup()")}}).
• The {{domxref("GPUCommandEncoder")}} is in an open state — this means that:
  • There are no child {{domxref("GPUComputePassEncoder")}} or {{domxref("GPURenderPassEncoder")}}s active that have not been ended (by calling end()).
  • The {{domxref("GPUCommandEncoder")}} has not already had finish() called on it (in which case it cannot be used to encode any more commands).

Examples

// …

const commandBuffer = commandEncoder.finish();
device.queue.submit([commandBuffer]);

// …

Specifications

{{Specifications}}

Browser compatibility

{{Compat}}

See also

• The WebGPU API