This is the documentation of ovni, the Obtuse (but Versatile) Nanoscale Instrumentation project.

The ovni project implements a fast instrumentation library that records small events (starting at 12 bytes) during the execution of programs to later investigate how the execution happened.

The instrumentation process is split in two stages: runtime tracing and emulation.

During runtime, very short binary events are stored on disk which describe what is happening. Once the execution finishes, the events are read and processed to reproduce the execution during the emulation process, and the final execution trace is generated.

By splitting the runtime and emulation processes we can perform expensive computations during the trace generation without disturbing the runtime process.

Each event belongs to a model, which has a direct mapping to a target library or program. Each model is independent of other models, and they can be instrumented concurrently.

Quick start#

To start a trace follow these steps:

• First enable the instrumentation of those libraries or programs that you are interested in (see their documentation).
• Then simply run the program normally. You will see a new ovni directory containing the runtime trace.
• Finally run the ovniemu ovni command to generate the Paraver traces.
• Use the command wxparaver ovni/cpu.prv to load the CPU trace.
• Load the configurations from the ovni/cfg/ directory that you are interested in, to open a timeline view.