Intel® C++ Compiler Classic Developer Guide and Reference

ID 767249
Date 3/31/2023
Public

A newer version of this document is available. Customers should click here to go to the newest version.

Document Table of Contents

fp-port, Qfp-port

Rounds floating-point results after floating-point operations.

Syntax

Linux:

-fp-port

-no-fp-port

macOS:

-fp-port

-no-fp-port

Windows:

/Qfp-port

/Qfp-port-

Arguments

None

Default

-no-fp-port
or /Qfp-port-

The default rounding behavior depends on the compiler's code generation decisions and the precision parameters of the operating system.

Description

This option rounds floating-point results after floating-point operations.

This option is designed to be used with the -mia32 (Linux*) or /arch:IA32 (Windows*) option on a 32-bit compiler. Under those conditions, the compiler implements floating-point calculations using the x87 instruction set, which uses an internal precision that may be higher than the precision specified in the program.

By default, the compiler may keep results of floating-point operations in this higher internal precision. Rounding to program precision occurs at unspecified points. This provides better performance, but the floating-point results are less deterministic. The [Q]fp-port option rounds floating-point results to user-specified precision at assignments and type conversions. This has some impact on speed.

When compiling for newer architectures, the compiler implements floating-point calculations with different instructions, such as Intel® SSE and SSE2. These Intel® Streaming SIMD Extensions round directly to single precision or double precision at every instruction. In these cases, option [Q]fp-port has no effect.

IDE Equivalent

Visual Studio: Optimization > Floating-point Precision Improvements

Eclipse: Floating Point > Round Floating-Point Results

Xcode: Floating Point > Round Floating-Point Results

Alternate Options

None