The GPUPipelineError interface of the WebGPU API describes a pipeline failure. This is the value received when a Promise returned by a GPUDevice.createComputePipelineAsync() or GPUDevice.createRenderPipelineAsync() call rejects.

GPUPipelineError() Experimental

Creates a new GPUPipelineError object instance.

Inherits properties from its parent, DOMException.

reason Experimental Read only

An enumerated value that defines the reason the pipeline creation failed in a machine-readable way.

In the following snippet we are attempting to create a GPUComputePipeline using GPUDevice.createComputePipelineAsync(). However, we have misspelt our compute pipeline entryPoint as "maijn" (it should be "main"), therefore pipeline creation fails, and our catch block prints the resulting reason and error message to the console.

// ...

let computePipeline;

try {
  computePipeline = await device.createComputePipelineAsync({
    layout: device.createPipelineLayout({
      bindGroupLayouts: [bindGroupLayout],
    }),
    compute: {
      module: shaderModule,
      entryPoint: "maijn",
    },
  });
} catch (error) {
  // error is a GPUPipelineError object instance
  console.error(error.reason);
  console.error(`Pipeline creation failed: ${error.message}`);
}

// ...

In this case, the given reason is "Validation", and the message is "Entry point "maijn" doesn't exist in the shader module [ShaderModule]."

• The WebGPU API
• WebGPU Error Handling best practices