I am currently testing out the grid leveling routine on my Delta Kossel, and ran into what seems to be a bug. When the probe is physically wired to be non-inverting, and is also not inverted in the smoothieware firmware (1.28^), the grid calibration appears to start and run. However when I changed my probe to be physically inverting, and the firmware to be inverting (1.28!^), the grid calibration refuses to detect that the probe has been pressed at all, and the firmware gets stuck forever at the "press endstop to continue" check.
The workaround for this currently is to wire the probe so that it connects the + and S pins, and run the pin as a pull down (1.28v) so that it is no longer inverting, however this is a little cumbersome. Is anyone else able to confirm this on their own printers? I have confirmed that the probe is reporting it's correct state via M119, and it also functions correctly with the standard delta calibration.