Interactive matplotlib plot backend freezes when using ipkernel 7.x.x with visual code #1466

Configuration:
Visual code with Jupyter extension v2025.9.1, ipykernel 7.0.1, python 3.13.8

Issue:
I get stuck when trying to plot something using an interactive window.
There are 2 strange behaviors:

import numpy as np
import matplotlib
from matplotlib import pyplot as plt
%matplotlib tk

plt.figure()
plt.plot(np.arange(10))
plt.show()

print("hello")

=> an empty interactive window opens and freezes, no "hello" printed and the interactive window can be closed using plt.close('all').

import numpy as np
import matplotlib
from matplotlib import pyplot as plt
matplotlib.use('TkAgg')

plt.figure()
plt.plot(np.arange(10))
plt.show()

print("hello")

=> As expected an interactive window opens properly with the expected plot, no "hello" printed until we close the plot window.

import numpy as np
import matplotlib
from matplotlib import pyplot as plt
matplotlib.use('TkAgg')

plt.figure()
plt.plot(np.arange(10))
plt.show(block=False)

print("hello")

=> same results as 1), an empty interactive window opens and freezes, no "hello" printed and the interactive window can be closed using plt.close('all').

After replacing ipykernel 7.0.1 by 6.31.0, the short code in 3) works properly, i.e. interactive plot window which is not blocking the code execution if it is not closed.