Update Se_status.f90

[rdemaria]

Fix an off bound error that show using openmp. @piotrskowronski could you verify?

[ldeniau]

The fix looks ok to me as the polygon is closed with POLYGN points when allocated in Sn_mad_like.f90.
Apart from small details, it's the same code (after fixing the bound) as mine in MAD-NG (written 5 years ago):
https://github.com/MethodicalAcceleratorDesign/MAD-NG/blob/0d1749f893ef52f932cb24379fb7800743607831/src/mad_poly.c#L37
I found a nice ref with a C implementation at http://geomalgorithms.com/a03-_inclusion.html