This particular type of test that we’re running is attempting to shake out timing bugs when the probe triggers right at (or very close to) the end of the move.
The commanded stop is something we do before “disarming” the probe (a form of cleanup). We put the multi axis in a STOPPED/IDLE state (and wait for it before proceeding) so that we can be certain nothing is moving while we’re changing internal states and sending commands to the drives. That procedure is also inhibiting further calls to MovePT() until cleanup is finished.
Here’s the anomaly again, but only looking at the Axes’ STATUS.AT_TARGET, MultiAxis.MOTION_STATUS.TriggeredModify, and Probe input, armed/success bits.
It looks like what you’ve described (slightly augmented by me) is what’s happening here.
Supposing that the probe triggers just after the end of the commanded decel, but before we’ve disabled the user limit.
What would you expect to happen then?