Permutation test number of tails

Description of the problem

The p-value produced by the function permutation_t_test does not depend on whether a one- or two-tailed test is requested. This is not the expected behaviour.

Steps to reproduce

import numpy as np
import mne.stats as stats

testData = np.atleast_2d(np.arange(5)).T

for thisTail in [0, 1]:
    _, p_values, _ = stats.permutation_t_test(testData, tail=thisTail)
    print('p-value: {}'.format(p_values[0]))

Link to data

No response

Expected results

The p-value for a one-tailed test should be about half the size of the p-value for a two-tailed test.

Actual results

The p-value for a one-tailed test and two-tailed test are the same.

Permuting 255 times (exact test)...
p-value: 0.0078125
Permuting 511 times (exact test)...
p-value: 0.0078125

Additional information

MNE version: 0.22.0
Platform: Windows-10
Python: 3.8.6

numpy: 1.19.5 {blas=NO_ATLAS_INFO, lapack=lapack}
scipy: 1.6.0
matplotlib: 3.3.3 {backend=Qt5Agg}

sklearn: 0.24.0
numba: 0.52.0
nibabel: 3.2.1
nilearn: 0.7.0
dipy: 1.3.0
cupy: Not found
pandas: 1.2.0
mayavi: 4.7.2
pyvista: 0.27.4 {pyvistaqt=0.2.0, OpenGL 4.5.0 - Build 30.0.101.1692 via Intel(R) UHD Graphics 620}
vtk: 9.0.1
PyQt5: 5.12.3