The pushErrorScope() method of the GPUDevice interface pushes a new GPU error scope onto the device's error scope stack, allowing you to capture errors of a particular type.

Once you are done capturing errors, you can end capture by invoking GPUDevice.popErrorScope(). This pops the scope from the stack and returns a Promise that resolves to an object describing the first error captured in the scope, or null if no errors were captured.

filter

An enumerated value that specifies what type of error will be caught in this particular error scope. Possible values are:

"internal"

The error scope will catch a GPUInternalError.

"out-of-memory"

The error scope will catch a GPUOutOfMemoryError.

"validation"

The error scope will catch a GPUValidationError.

None (Undefined).

The following example uses an error scope to capture a suspected validation error, logging it to the console.

device.pushErrorScope("validation");

let sampler = device.createSampler({
  maxAnisotropy: 0, // Invalid, maxAnisotropy must be at least 1.
});

device.popErrorScope().then((error) => {
  if (error) {
    sampler = null;
    console.error(`An error occurred while creating sampler: ${error.message}`);
  }
});

See WebGPU Error Handling best practices for a lot more examples and information.

• The WebGPU API