Bent Laue crystal ML gives bad result for asymmetry angles in [90,180]

[srio]

Look to this example with PP (which is OK):

Switching to ML:

(note that it seems to work for asymmetry angle -52.2)

PS: Thanks to Rober Farla at DESY for pointing out the problem.