pytest fixture and test both dependent on another fixture

I have a pair of fixtures for testing a function with x-y inputs. One generates the x-values, the other the y-values. The test depends on both. The y-values depend on the x-values as well. Here is a summary of the setup:

import pytest

@pytest.fixture(params=['a', 'b', 'c'])
def x_data(request):
    return request.param

@pytest.fixture(params=['1', '2', '3'])
def y_data(request, x_data):
    return request.param + x_data

def test_mathfunc(x_data, y_data):
    print(f'\nTesting {x_data} vs {y_data}', end='')

When I run this with pytest -s, there are exactly 9 tests run, and the printout in test_func shows that the x_data passed in to the test is always the same as the one used to generate the y_data (which is exactly what I want, of course):

Testing a vs 1a.
Testing a vs 2a.
Testing a vs 3a.
Testing b vs 1b.
Testing b vs 2b.
Testing b vs 3b.
Testing c vs 1c.
Testing c vs 2c.
Testing c vs 3c.

I am aware of the fact that pytest caches fixture results and attempts to evaluate fixtures the minimum possible number of times. For this simple case, that would imply that the behavior is fairly reliable.

The question is, can I really rely on this behavior? Is there a situation that I am missing where the test could end up getting called with, for example, a, 3c?