how many rows does pandas' interpolate consider?

How does pandas' DataFrame.interpolate() work in relation to the amount of rows it considers:

1. is it just the row before the NaNs and the row right after?
2. Or is it the whole DataFrame (how does that work at 1 million rows?)
3. Or another way (please explain)

each of the methods is relevant.

‘linear’: Ignore the index and treat the values as equally spaced. This is the only method supported on MultiIndexes.
                
                    ‘time’: Works on daily and higher resolution data to interpolate given length of interval.
                
                    ‘index’, ‘values’: use the actual numerical values of the index.
                
                    ‘pad’: Fill in NaNs using existing values.
                
                    ‘nearest’, ‘zero’, ‘slinear’, ‘quadratic’, ‘cubic’, ‘spline’, ‘barycentric’, ‘polynomial’: Passed to scipy.interpolate.interp1d. These methods use the numerical values of the index. 
                                                                                                               Both ‘polynomial’ and ‘spline’ require that you also specify an order (int), e.g. df.interpolate(method='polynomial', order=5).
                
                    ‘krogh’, ‘piecewise_polynomial’, ‘spline’, ‘pchip’, ‘akima’: Wrappers around the SciPy interpolation methods of similar names. See Notes.
                    
                    ‘from_derivatives’: Refers to scipy.interpolate.BPoly.from_derivatives which replaces ‘piecewise_polynomial’ interpolation method in scipy 0.18.