operand syntax, and the operand data format, the CU completes execution of the instruction
in one of the following ways:
• The instructions that have operand data formats B, W, L, and P are listed in Table
5-1. When the CU receives an instruction with the operand data format of B, W, or L,
it requests the BIU to transfer the necessary operand. Then, the CU waits for the APU
to become idle so that it can hand off the instruction to the APU. If the instruction has
an operand with data format P, the CU does not request the prefetch of the operand.
In this case, it waits until the APU is idle to hand off the instruction to the APU.
• If the instruction is an FMOVE.X FPm,FPn instruction, the CU does the following:
a. Releases the main processor.
b. Prefetches the source operand from FPm, unless the instruction currently operating
in the APU uses FPm as a destination. In that case, the CU waits until the APU is
idle before prefetching.
c. If the selected rounding precision is single or double, waits until the APU is idle
and hands off the instruction to the APU.
d. If the operand data type is NAN, denormalized, or unnormalized, waits until the
APU is idle and hands off the instruction to the APU.
e. If the instruction currently in the APU uses FPn, waits until the APU is idle before
proceeding.
f. Writes the source operand into the destination floating-point data register without
involving the APU.
• If the instruction is an FMOVE <ea>, FPn with an operand format of S, D, or X, the
CU does the following:
a. Prefetches the source operand from memory by using the evaluate <ea> and
transfer
data
primitive with CA = 0. This releases the main processor immediately
after the source operand is written to the operand CIR.
b. Converts the memory source operand to internal extended format. If the selected
rounding precision is single or double, waits until the APU is idle and hands off
the instruction to the APU.
c. Creates a tag that represents the
data
type of the converted source operand (nor-
malized, denormalized, zero, infinity, or NAN).
d. If the operand data type is NAN, unnormalized, or denormalized, waits until the
APU is idle and hands off the instruction to the APU.
e. if the instruction currently in the APU uses FPn, waits until the APU is idle before
proceeding.
f. Writes the converted source operand into FPn without involving the APU.
• If the instruction is an FMOVE FPm,<ea> with a data format of S or D, the CU does
the following:
a. Prefetches the source operand from FPm, unless the instruction currently operating
in the APU uses FPm as a destination. In that case, the CU waits until the APU is
idle before prefetching.
bo If the data type of the source operand is unnormalized, denormalized, or NAN,
waits until the APU is idle and hands off the instruction to the APU.
c. Converts the source operand to the destination data format.
d. If the conversion results in an overflow or underflow, or if the INEX2 trap is enabled,
waits until the APU is idle and hands off the instruction to the APU.
e. Creates a tag that represents the data type of the converted source operand (nor-
• malized, denormalized, zero, infinity, or NAN).
MC68881/MC68882 USER'S MANUAL FREESCALE
5-3
To Next Page
Loading...