Skip to content

FETOXM1A

Operation

Approximates e raised to the source minus one without an intermediate rounded exponential.

Assembler Syntax

FETOXM1A.{S|D}(z) Fn(s), Fn(d)

Privilege

Unprivileged

Required CPUID flags

FP

FPTRANSA

Repeat eligibility

REP eligible

Computes the \(e^x - 1\) reference without first rounding \(e^x\) over all finite values and signed infinity after DAZ and writes an implementation-bounded S- or D-format approximation under the selected FPTRANSA accuracy contract.

The required FPTRANSA_ACCURACY contract is 0x0032.

Special values

  • Positive infinity: the result is +infinity.

  • Negative infinity: the result is -1.

  • Zero source: the result is the source.

Exact anchors

  • \(\operatorname{expm1}(+0) = +0\)

  • \(\operatorname{expm1}(-0) = -0\)

  • \(\operatorname{expm1}(-\infty) = -1\)

Required properties

  • strictly increasing for finite inputs

  • result is at least -1

2.75in

Encodings

FETOXM1A.{S|D}(z) Fn(s), Fn(d)

Format: Instruction format for FETOXM1A.{S|D}(z) Fn(s), Fn(d)

Format: Instruction format for FETOXM1A.{S|D}(z) Fn(s), Fn(d)