Introducing CHAP - A Program to Clarify Dynamic Memory Usage in Un-Instrumented Cores

Explore a powerful tool for analyzing memory usage and leaks in C and C++ programs without code instrumentation. Learn about CHAP (formerly ah64), an open-source program developed at VMware that examines un-instrumented core files to detect leaks, understand memory growth issues, and identify memory corruption. Discover how CHAP avoids false leak reports and timing distortions associated with traditional instrumentation methods. Gain insights into the tool's functionality, usage scenarios, and its ability to analyze memory issues in live cores or crash dumps. Understand CHAP's current capabilities, including support for Linux ELF cores using glibc malloc, and learn about potential future enhancements such as support for additional allocators and dump formats. Benefit from the presenter's extensive experience in C and C++ programming and debugging tool development at VMware.