Nearest-even binary arithmetic kernels

The six nearest-even kernels behind the default Model.add, Model.sub, Model.mul, Model.div, Model.sqrt, and Model.fma call the structural dispatchers selected for a validated format descriptor. They sit below the directed and status kernels, which reuse them for nearest-even rounding, and the public operations in Arithmetic.Runtime select among the three.

def FloatLib.Floats.Formats.BinaryInterchange.Model.Internal.Nearest.add {fmt : FloatFormat} (x y : Model fmt) : Model fmt

Automatically dispatched nearest-even addition for any validated format descriptor.

def FloatLib.Floats.Formats.BinaryInterchange.Model.Internal.Nearest.sub {fmt : FloatFormat} (x y : Model fmt) : Model fmt

Automatically dispatched nearest-even subtraction for any validated format descriptor.

def FloatLib.Floats.Formats.BinaryInterchange.Model.Internal.Nearest.mul {fmt : FloatFormat} (x y : Model fmt) : Model fmt

Automatically dispatched nearest-even multiplication for any validated format descriptor.

def FloatLib.Floats.Formats.BinaryInterchange.Model.Internal.Nearest.div {fmt : FloatFormat} (x y : Model fmt) : Model fmt

Automatically dispatched nearest-even division for any validated format descriptor.

def FloatLib.Floats.Formats.BinaryInterchange.Model.Internal.Nearest.sqrt {fmt : FloatFormat} (x : Model fmt) : Model fmt

Automatically dispatched nearest-even square root for any validated format descriptor.

def FloatLib.Floats.Formats.BinaryInterchange.Model.Internal.Nearest.fma {fmt : FloatFormat} (x y z : Model fmt) : Model fmt

Automatically dispatched nearest-even fused multiply-add for any validated format descriptor.