glm::intersectRayTriangle might be broken

[dgrat]

I am in a sphere at pos 0,0,0 and the sphere radius is 10 with center 0,0,0.
If I intersect along one direction e.g. dir 1,0,0 I expect one intersecting face. Instead it gives me always several intersections (always two I think). What makes me wonder most is that the Distance is one time positive (as expected) and one time negative despite the directions vector is positive. The only case where this should happen is if I use a line intersection (lineTriangleIntersection). This is why I think the test case and the ray function is broken.

glm version 9.9
from: https://github.com/dgrat/voxelizer