Bug in sklearn WhiteKernel: always returns zero

White Kernel, which should behave like this

k(x1, x2) = 0 if x1 != x2 else noies_level

is always returning zero:

from sklearn.gaussian_process import kernels
import numpy as np
import sklearn
print(kernels.WhiteKernel(noise_level=2.0)(np.array([[1.0]]), np.array([[1.0]])))
print(sklearn.__version__)

Result:

[[0.]]
1.1.1

I'm expecting nonzero result. What am I missing here?