PerfExec

PerfExec is a tool designed for measuring the performance of executable files in a Windows environment. It evaluates the execution time, providing insights into the average, median, fastest, and slowest execution times. PerfExec is easy to use, offering a straightforward command-line interface for conducting performance tests.

Features

Performance Benchmarking: Assess the performance of executable files by measuring execution times.
Execution Count Customization: Specify the number of times the executable file or command should be run.
Additional Command-Line Arguments: Support for passing extra arguments to the executable file or command.
Result Output to File: Option to save benchmark results to a specified file.

Usage

Execute the program from the command line using the following syntax:

benchmark.exe <executable file path> [options]

Options

-t <count>: Specifies the number of times to execute the test. Defaults to 10 executions if not specified. Valid range is 1 to 100. Note: Cannot be used with the -a option.
-a <count>: Executes the specified command a given number of times in parallel. Specifies the number of parallel executions. Note: Cannot be used with the -t option.
-o <filename>: Specifies the filename where benchmark test results will be saved. Defaults to 'benchmark_result.txt' if not specified.
-x "<args>": Passes additional command-line options to the executable file or command. The options should be enclosed in quotes.
-s <filename>: Specifies that the executable is a system-registered executable. This option is used to run executables located in system paths.
-d: Displays detailed results, providing more comprehensive information about the benchmark results.
-h: Displays the help message.
-v: Displays the version of the tool.

Examples

Sequential Execution

To execute example.exe 10 times in sequential order and save the results to results.txt:

benchmark.exe example.exe -t 10 -o results.txt

Parallel Execution

To execute example.exe 30 times in parallel and save the results to parallel_results.txt:

benchmark.exe example.exe -a 30 -o parallel_results.txt

Using System-Registered Executable

To execute a system-registered Python script (script.py) located in system paths, passing 1000 as a command-line argument, and save the results to python_results.txt:

benchmark.exe python -s script.py -x "1000" -t 10 -o python_results.txt

In this example, the -s option is used to specify an executable registered with the system, in this case Python. The -x option is used to pass an additional argument (-1000) to the script, and the -t option specifies the number of executions.

Building

To build this tool, you need a C compiler such as GCC (MinGW). Compile the source file (benchmark.c) using your compiler to generate the executable file.

gcc benchmark.c -o benchmark

License

This project is released under the MIT License. For more details, see the LICENSE.

rxxuzi/PerfExec