Atmel Xplained Pro Evaluation Kit ATSAM4E-XPRO ATSAM4E-XPRO Data Sheet

SAM4E [DATASHEET]

Atmel-11157D-ATARM-SAM4E16-SAM4E8-Datasheet_12-Jun-14

186

Converts a value in a register from floating-point to and from fixed-point.

Syntax

VCVT{cond}.Td.F32 Sd, Sd, #fbits

VCVT{cond}.F32.Td Sd, Sd, #fbits

where:

cond

is an optional condition code, see 

.

Td

is the data type for the fixed-point number. It must be one of:

S16     signed 16-bit value.
U16     unsigned 16-bit value.

S32     signed 32-bit value.
U32     unsigned 32-bit value.

Sd

is the destination register and the operand register.

fbits

is the number of fraction bits in the fixed-point number:

If Td is S16 or U16, fbits must be in the range 0–16.
If Td is S32 or U32, fbits must be in the range 1–32.

Operation

These instructions:

1.

Either

̶

Converts a value in a register from floating-point to fixed-point.

̶

Converts a value in a register from fixed-point to floating-point.

2.

Places the result in a second register.

The floating-point values are single-precision.

The fixed-point value can be 16-bit or 32-bit. Conversions from fixed-point values take their operand from the low-
order bits of the source register and ignore any remaining bits. 

Signed conversions to fixed-point values sign-extend the result value to the destination register width. 

Unsigned conversions to fixed-point values zero-extend the result value to the destination register width.

The floating-point to fixed-point operation uses the Round towards Zero rounding mode. The fixed-point to floating-
point operation uses the Round to Nearest rounding mode.

Restrictions

There are no restrictions.

Condition Flags

These instructions do not change the flags.