Add utilities for benchmarking different code versions #4

	name: Run benchmark and profiling

	# The workflow gets triggered by pushes and pull requests
	on:
	push:
	branches: [ "main" ]
	pull_request:
	branches: [ "main" ]

	jobs:
	build:

	runs-on: ubuntu-latest
	strategy:
	fail-fast: false

	steps:
	# checks out the code in the repository
	- uses: actions/checkout@v3
	with:
	submodules: true

	- name: Install dependencies
	run: \|
	sudo apt-get update && apt upgrade -y
	sudo apt-get install -y libboost-all-dev
	python -m pip install --upgrade pip
	pip install matplotlib

	- name: Compile and run benchmark
	working-directory: ${{ github.workspace }}/benchmark/dataset_size
	run: \|
	cmake -S . -B build -DCMAKE_BUILD_TYPE=Release
	cmake --build build -- -j 2
	./build/serial.out 10 18

	- name: Compile and run profiling
	working-directory: ${{ github.workspace }}/benchmark/profiling
	run: \|
	cmake -S . -B build/Debug -DCMAKE_BUILD_TYPE=Debug
	cmake --build build/Debug -- -j 2
	cmake -S . -B build/Release -DCMAKE_BUILD_TYPE=Release
	cmake --build build/Release -- -j 2
	./build/Debug/serial.out ../../data/data_32768.csv
	gprof ./build/Debug/serial.out ../../data/data_32768.csv
	./build/Release/serial.out ../../data/data_32768.csv
	gprof ./build/Release/serial.out ../../data/data_32768.csv

	- name: Check cache misses with perf
	working-directory: ${{ github.workspace }}/benchmark/profiling
	run: \|
	perf stat -B -e cache-misses,cycles,instructions,branches ./build/Debug/serial.out ../../data/data_32768.csv
	perf stat -B -e cache-misses,cycles,instructions,branches ./build/Release/serial.out ../../data/data_32768.csv

Provide feedback