Secure Programming for Linux and Unix HOWTO

David A. Wheeler

v2.92, 8 January 2002

This book provides a set of design and implementation guidelines for writing secure programs for Linux and Unix systems. Such programs include application programs used as viewers of remote data, web applications (including CGI scripts), network servers, and setuid/setgid programs. Specific guidelines for C, C++, Java, Perl, PHP, Python, TCL, and Ada95 are included.


Table of Contents
1. Introduction
2. Background
2.1. History of Unix, Linux, and Open Source / Free Software
2.2. Security Principles
2.3. Why do Programmers Write Insecure Code?
2.4. Is Open Source Good for Security?
2.5. Types of Secure Programs
2.6. Paranoia is a Virtue
2.7. Why Did I Write This Document?
2.8. Sources of Design and Implementation Guidelines
2.9. Other Sources of Security Information
2.10. Document Conventions
3. Summary of Linux and Unix Security Features
3.1. Processes
3.2. Files
3.3. System V IPC
3.4. Sockets and Network Connections
3.5. Signals
3.6. Quotas and Limits
3.7. Dynamically Linked Libraries
3.8. Audit
3.9. PAM
3.10. Specialized Security Extensions for Unix-like Systems
4. Validate All Input
4.1. Command line
4.2. Environment Variables
4.3. File Descriptors
4.4. File Contents
4.5. Web-Based Application Inputs (Especially CGI Scripts)
4.6. Other Inputs
4.7. Human Language (Locale) Selection
4.8. Character Encoding
4.9. Prevent Cross-site Malicious Content on Input
4.10. Filter HTML/URIs That May Be Re-presented
4.11. Forbid HTTP GET To Perform Non-Queries
4.12. Limit Valid Input Time and Load Level
5. Avoid Buffer Overflow
5.1. Dangers in C/C++
5.2. Library Solutions in C/C++
5.3. Compilation Solutions in C/C++
5.4. Other Languages
6. Structure Program Internals and Approach
6.1. Follow Good Software Engineering Principles for Secure Programs
6.2. Secure the Interface
6.3. Minimize Privileges
6.4. Avoid Creating Setuid/Setgid Scripts
6.5. Configure Safely and Use Safe Defaults
6.6. Load Initialization Values Safely
6.7. Fail Safe
6.8. Avoid Race Conditions
6.9. Trust Only Trustworthy Channels
6.10. Set up a Trusted Path
6.11. Use Internal Consistency-Checking Code
6.12. Self-limit Resources
6.13. Prevent Cross-Site Malicious Content
6.14. Be Careful with Data Types
7. Carefully Call Out to Other Resources
7.1. Call Only Safe Library Routines
7.2. Limit Call-outs to Valid Values
7.3. Call Only Interfaces Intended for Programmers
7.4. Check All System Call Returns
7.5. Avoid Using vfork(2)
7.6. Counter Web Bugs When Retrieving Embedded Content
7.7. Hide Sensitive Information
8. Send Information Back Judiciously
8.1. Minimize Feedback
8.2. Don't Include Comments
8.3. Handle Full/Unresponsive Output
8.4. Control Data Formatting (``Format Strings'')
8.5. Control Character Encoding in Output
8.6. Prevent Include/Configuration File Access
9. Language-Specific Issues
9.1. C/C++
9.2. Perl
9.3. Python
9.4. Shell Scripting Languages (sh and csh Derivatives)
9.5. Ada
9.6. Java
9.7. TCL
9.8. PHP
10. Special Topics
10.1. Passwords
10.2. Authenticating on the Web
10.3. Random Numbers
10.4. Specially Protect Secrets (Passwords and Keys) in User Memory
10.5. Cryptographic Algorithms and Protocols
10.6. Using PAM
10.7. Tools
10.8. Windows CE
10.9. Write Audit Records
10.10. Miscellaneous
11. Conclusion
12. Bibliography
A. History
B. Acknowledgements
C. About the Documentation License
D. GNU Free Documentation License
E. Endorsements
F. About the Author
List of Tables
4-1. Legal UTF-8 Sequences
List of Figures
1-1. Abstract View of a Program