Benchmarking

Table of contents

Introduction to Benchmarking

Developing a working and stable application, from the scribbles to the final executing binary, is a long and hard task. During this process, developers may come across stability and perfomance issues. In addition to these issues, some specified QoS might be difficult to quantify. Solving those problems without the proper tools might be frustrating, tedious tasks leading to reduce developers efficiency. An adapted benchmarking tool could overcome all those development obstacles and increase development time. There are different KPI (Keep Performance Indicators) that one might be interested into. In the framework of micro-ROS, the KPI can be freely chosen by the developer. In this way, the benchmarking tool will remain flexible and allow the community to constantly add support for a lot of different KPI.

The problems we want to tackle are:

  • Out there, many benchmarking tools exist, each of them targeting different KPIs.
  • Different platforms (Linux/Nuttx/Baremetal et.c.).
  • Too few time/resources to code benchmarking tool for each.
  • Avoid code overhead: Keep code clarity.
  • Avoid execution overhead: Do not want to make execution slower when benchmarking.

Our Benchmarking tool framework

The benchmarking tool under development is providing a framework to allow developers to create their own benchmarking tool. Each part a developer wants to benchmark can be added as a plugin using the provided framework. In this way plugins can be shared and this improves re-usability as much as possible.

Trace Framework Abstraction

The Shadow builder alone only parse comments from the application and pass it along to the Trace Framework Abstraction (TFA) Core. The TFA core is aware of the plugins that are available, all the plugins’ capabilities and platform target. The process goes as explained below:

  • The line containing the functionality Benchmarking::XX::YY will be checked against all the available plugins.
  • Plugins that are capable of handling functionality will respond with a piece of code that will be replaced with a piece of code.
  • Then the output file will be added in a folder corresponding to the platform type and benchmarking type.

Being generic is the key for this benchmarking tool. The plugins will in contrary bring the specific implementation needed to benchmark a specific platform. Every plugin will provide information as requested by the parser:

  • Provide a list of supported platforms.
  • Provide a list of functions that are handled.
  • Provide snippet codes that will be added for benchmarking.
  • Provide a list of patches and/or patch code
  • Optional provide an end script to run and execute the benchmarks

Shadow Builder

This section will introduce some concepts related to the shadow builder (SB).

The Shadow builder is a tool that will transparently instrument the code to benchmark. The tools will be able to output an “instrumented code” that will be later be compiled as a normal code. The following steps describe what the shadow builder process flow:

  • Get configuration file from the user (Benchmarking Configuration File).
  • Get appropriate sources.
  • Execute Trace Framework Abstraction Configuration file.
  • Parse the sources file needed Injecting code.
  • Compile the targeted binary for different platforms.
  • If needed, depending what type benchmark is undertaken, compile another target binary benchmarking.

The SB (Shadow Builder) is meant to be as transparent as possible for the user. And if the benchmarking is not activated, it should be bypassed.

The SB is in charge of getting the path/git repository to the source code that needs to be benchmarking. The benchmarking. The sources are specified by the user in the benchmarking configuration file.

In order to inject code, there are some tools that allow this. CLang AST tool will allow to inject some code.

Binary generation for instrumented code

The binary generation is the process of compiling the source code. In order to benchmark, previously to compile the source code, it is necessary to instrument the code. The code will be instrumented in a transparent way for the programmer/user. Therefore, a configuration file provided by the programmer will be parsed and code injected as described in a configuration file.

Receiving inputs

The binary generation’s pipeline receives two inputs to work with:

  • Configuration Benchmarking file.
  • Source code to benchmark.

In short, the configuration describes:

  • What is benchmarked (sources).
  • Where to benchmark.
  • What type of benchmark.
  • Optionally against what base line to compare (base line source)

Parse and Check

Once the input received the Shadow Builder parses the configuration file. From the configuration file, the Shadow builder gets:

  • The different benchmarking to be achieved.
  • The targeted platforms.

In addition to parsing, the Shadow Builder is in charge of checking capabilities and consistency within the configuration file and the different TFA’s plugins registered in the TFA module.

TFA Execution

Once parsed and checked against the TFA module capabilities, the Shadow Builder will be in charge of translating configuration into source code. The translated sources will also be achieved in cooperation with the TFA module. The detailed steps of the TFA can be found here. At the end of this step, the TFA will generate the new forged source code ready for compilation. In addition to patched source code, the TFA will generate scripts that will the benchmarks.

Compilation

The compilation will happen for every kind of benchmarks and platforms targeted. Depending on the kind of benchmark that is being executed, there will be one or more binaries per benchmarks session. The number of binary generated also depends on what plugins are provided by the user to the shadow builder. The shadow builder will retrieve capabilities of the plugins and request from the developer, match them and generated software according to the matches.

Step to start benchmarking

The shadow Builder will be executed as follow:

  • Software sources are passed to the Shadow Builder.
  • The source are passed and upon comments containing /Benchmarking::XX::YY/ (a tag) the code line is passed to the Trace Framework Abstraction module. Using comments is preferable → No includes needed.
  • All plugins that registered to the TFA the Benchmarking::XX::YY functionality will return a piece of code that will be added to the source.
  • Once all parsed, the shadow builder will compile for all the different platforms requested either by plugins or by user configuration.