Depending on the orientation of the quarter wave plate, it will rotate the polarization of the light passing through that slit from linear, to circular, and back to linear as the plate is rotated. At all points, the light passing through that slit will have a significant component of linear polarization along the same direction--- anywhere between 100 percent and 50 percent.
What you will see on the screen as you rotate the 1/4 waveplate, is that it will change from a perfect slit interference pattern to a 50/50 mixture of the slit interference pattern and the diffraction pattern, and back again to perfect slit interference.
Ok, this may make sense... but I still don't get it why a polarization change induces a diffraction pattern. As far as I understand the double slit experiment, as any interference phenomenon in general, it is only the phase difference on the detection screen between the two light beams coming from the two slits which determine the diffraction and interference pattern. Here, of course, the rotation of the polarizer at 100 percent determines also ##\lambda##/4 phase shift. But I think of it as resulting in a spatial shift of the interference fringes, not an overlap between the diffraction envelope and the fringes. What am I missing here?