Non-Portable Types - Oracle® Solaris Studio 12.4: Numerical Computation Guide

Oracle® Solaris Studio 12.4: Numerical Computation Guide

Exit Print View

Documentation Home » Oracle Solaris Studio 12.4 Information Library » ... » Reproducible Results » Non-Portable Types

Updated: January 2015

Oracle^® Solaris Studio 12.4: Numerical Computation Guide

Document Information

Using This Documentation

Chapter 1 Introduction

Chapter 2 IEEE Arithmetic

Chapter 3 The Math Libraries

Chapter 4 Exceptions and Exception Handling

Chapter 5 Compiler Code Generation

5.1 Supported Operation Systems, Hardware, and Memory Model

5.2 Code Generation Options

5.3 Default Address Model and Code Generation

5.4 Compilation Options

5.5 Reproducible Results

5.5.1 Transcendental Functions

5.5.2 Associative Operations

5.5.3 Indeterminate Evaluation

5.5.4 Non-Portable Types

5.5.5 Implicit Higher Precision

5.6 Independent Confirmation

Appendix A Examples

Appendix B SPARC Behavior and Implementation

Appendix C x86 Behavior and Implementation

Appendix D Addendum to What Every Computer Scientist Should Know About Floating-Point Arithmetic

Appendix E Standards Compliance

Appendix F References

Please tell us how to improve our documentation:

Terms of Use | Your Privacy Rights

Thank you for your feedback! If you want to participate in content improvement and share additional information, please click Feedback button.

Rate this document:

Language:

5.5.4 Non-Portable Types

long double in C/C++ is implemented differently on SPARC and x86 in Studio, with 113 significant bits and 64 significant bits respectively. Programs with explicit long double variables are thus bound to behave differently on SPARC and x86.

Copyright © 2015, Oracle and/or its affiliates. All rights reserved. Legal Notices