Teensy 3.x and SMLAL cycles (32×32+64 signed multiply-add)

[Nominal Animal]

As far as I understand, from the Cortex-M4 and Freescale Kinetis documentation, the SMLAL instruction, which computes the 64-bit product of two 32-bit signed integers, and adds the result to a 64-bit signed integer (in two registers), should only take one cycle (if FASTRUN) on all Teensy 3.x (3.0, 3.1, 3.2, 3.5, and 3.6).

Is this correct?