In PR#1748 we tried to enable the Stockham FFT tests.
Their problem is that the size of an array is given as b**e which the code generator translates to
new double [dace::math::pow(b, e)
currently, pow will generate a float, which leads to a compilation error.
In PR#1748, the solution was to specialize pow to handle this case.
But it was concluded that this was the wrong approach.
This is now the second attempt.
In PR#1748 we tried to enable the Stockham FFT tests. Their problem is that the size of an array is given as
b**ewhich the code generator translates tocurrently,
powwill generate a float, which leads to a compilation error. In PR#1748, the solution was to specializepowto handle this case. But it was concluded that this was the wrong approach. This is now the second attempt.