Omnitrace documentation

Omnitrace is designed for the high-level profiling and comprehensive tracing of applications running on the CPU or the CPU and GPU. It supports dynamic binary instrumentation, call-stack sampling, and various other features for determining which function and line number are currently executing. To learn more, see What is Omnitrace?

The code is open and hosted at ROCm/omnitrace.

Install

• Quick start

• Omnitrace installation

The documentation is structured as follows:

Tutorials

• GitHub examples

• Video tutorials

How to

• Configuring and validating the Omnitrace environment

• Configuring runtime options

• Sampling the call stack

• Instrumenting and rewriting a binary application

• Performing causal profiling

• Understanding the Omnitrace output

• Profiling Python scripts

• Using the Omnitrace API

• General tips for using Omnitrace

Conceptual

• Data collection modes

• The Omnitrace feature set

Reference

• Development guide

• Omnitrace glossary

• API library

• Class member functions

• Globals

• Classes, structures, and interfaces

To contribute to the documentation, refer to Contributing to ROCm.

You can find licensing information on the Licensing page.