The instructions are case insensitive and ignore spaces between operands. The left side of the arithmetic expression, P=, is implicitly declared and should not be specified.
Valid Operands | D, A, ACIN, B, BCIN, CONCAT, C, PCIN, CARRYIN, CARRYCASCIN, 0 | Note that the CONCAT port is the concatenation of the A and B ports, used when the DSP is used as an adder. This is sometimes referred to as A:B in DSP opmodes. |
Valid Operators | +, -, *, >>17 | >>23 | The >>17 operator targets the DSP48 17-bit wire shift; valid only for P and PCIN operators and not valid for VersalĀ® devices. In Versal devices, this becomes a 23-bit wire shift and the operator changes to >>23, targeting the DSP58 23-bit wire shift. |
Valid Functions | rndsimple, rndsym, rndmacc | Rounding functions require that the P output width is less than full precision. See Supported Functions for further details. |