google benchmark vector type

vector_bench_add

#include <algorithm>
#include <vector>

#include <benchmark/benchmark.h>

template<class T>
static void vector_add(benchmark::State& state)
{
  size_t size = state.range(0);

  std::vector<T> vec1(size);
  std::vector<T> vec2(size);
  std::vector<T> vec3(size);
  std::fill(vec1.begin(), vec1.end(), static_cast<T>(12));
  std::fill(vec2.begin(), vec2.end(), static_cast<T>(2));
  std::fill(vec2.begin(), vec2.end(), static_cast<T>(0));

  for (auto _ : state) {
    for (size_t i = 0; i < size; ++i) {
      vec3[i] = vec1[i] + vec2[i];
    }
    // Make sure the variable is not optimized away by compiler
    benchmark::DoNotOptimize(vec1);
    benchmark::DoNotOptimize(vec2);
    benchmark::DoNotOptimize(vec3);
    benchmark::ClobberMemory();
  }
  state.SetBytesProcessed(state.iterations() * state.range(0) * sizeof(T));
  state.SetItemsProcessed(state.iterations() * state.range(0));
}

/*
BENCHMARK_TEMPLATE(vector_add, uint8_t)
    ->Name("uint8_t")
    ->RangeMultiplier(2)
    ->Range(1, 1048576);
*/

BENCHMARK_TEMPLATE(vector_add, uint8_t)
        ->Arg(1048576);
BENCHMARK_TEMPLATE(vector_add, uint_fast8_t)
        ->Arg(1048576);
BENCHMARK_TEMPLATE(vector_add, uint16_t)
        ->Arg(1048576);
BENCHMARK_TEMPLATE(vector_add, uint_fast16_t)
        ->Arg(1048576);
BENCHMARK_TEMPLATE(vector_add, uint32_t)
        ->Arg(1048576);
BENCHMARK_TEMPLATE(vector_add, uint_fast32_t)
        ->Arg(1048576);
BENCHMARK_TEMPLATE(vector_add, uint64_t)
        ->Arg(1048576);
BENCHMARK_TEMPLATE(vector_add, uint_fast64_t)
        ->Arg(1048576);