This issue still happens. Motes stay at S_RXDATAOFFSET at the beginning of the slot, this means that for some reason the previous slot did not finish correctly and the state machine has been broken.
the S_RXDATAOFFSET state is set at the beginning of the slot:
and a timer is set to transit to the new state.
If this state remains this can be attributed to the timer not triggering well (which is odd as timers work well). I'm gonna check that.
when the timer triggers correctly the following function is executed: