RQ_decomposition is there a problem with Y 90 /270 degrees?

edit

Hi I was wondering if anyone has come across a problem with the Euler angles returned by RQDecomp3X3 method when rotating by 90 or 270 degrees in the Y axis. I have tried decomposing a rotation matrix which is composed of 3 rotations in the following order starting with X (rotation matrix) multiplied by Y(rotation matrix) multiplied by Z(rotation matrix) (ZYX) with any arbitrary combination of angles for the X and Z where Y = 90 and always get the following vector of Euler angles(90,90,90) or (90,-90,90) where Y=270. So regardless of my input angles for X and Z the result is always the same. Has anyone come across this, or do you know how to explain why this is occurring?

Any help with this will be greatly appreciated, Suzana