Time Domain Setup saves incorrect H5 files if Maximum Rabi pulse length is not a multiple of the step size

[thilokru]

As stated in the title, the rabi measurement of the TD3 setup saves x-axis and data with conflicting dimensions. This can be fixed, by forcing the maximum pulse length to be a multiple of the step size:

t_step = 8e-9
t_max = t_step * 32

This is not obvious, however, and should probably be fixed.

[thilokru]

The Rabi Job appears to be inconsistent in the ql.jobs.Rabi.time_range function.

[thilokru]

Not reproducible.