#pragma region Macros // #pragma GCC target("avx2") #pragma GCC optimize("O3") #include #define rep(i, n) for (long long i = 0; i < (n); i++) #define rrep(i, n) for (long long i = (n - 1); i >= 0; i--) #define ALL(v) v.begin(), v.end() #define endl "\n" #define fi first #define se second #define popcount(bit) __builtin_popcount(bit) #define popcountll(bit) __builtin_popcountll(bit) #define pb push_back #define eb emplace_back using namespace std; using P = pair; using PL = pair; using Graph = vector>; typedef long long ll; template inline bool chmax(T& a, T b) { if (a < b) { a = b; return 1; } return 0; } template inline bool chmin(T& a, T b) { if (a > b) { a = b; return 1; } return 0; } const int dx[4] = {1, 0, -1, 0}; const int dy[4] = {0, 1, 0, -1}; const int fx[8] = {0, 1, 1, 1, 0, -1, -1, -1}; const int fy[8] = {1, 1, 0, -1, -1, -1, 0, 1}; template const auto INF = numeric_limits::max()/2; namespace PrimeCounting { using i64 = long long; static inline i64 my_div(i64 n, i64 p) { return double(n) / p; }; __attribute__((target("avx2"), optimize("O3", "unroll-loops"))) i64 prime_counting(i64 N) { i64 N2 = sqrt(N); i64 NdN2 = my_div(N, N2); vector hl(NdN2); for(int i = 1; i < NdN2; i++) hl[i] = my_div(N, i) - 1; vector hs(N2 + 1); iota(begin(hs), end(hs), -1); for(int x = 2, pi = 0; x <= N2; ++x) { if(hs[x] == hs[x - 1]) continue; i64 x2 = i64(x) * x; i64 imax = min(NdN2, my_div(N, x2) + 1); i64 ix = x; for(i64 i = 1; i < imax; ++i) { hl[i] -= (ix < NdN2 ? hl[ix] : hs[my_div(N, ix)]) - pi; ix += x; } for(int n = N2; n >= x2; n--) { hs[n] -= hs[my_div(n, x)] - pi; } ++pi; } return hl[1]; } } // namespace PrimeCounting /** * @brief 素数カウント( $\mathrm{O}(\frac{N^{\frac{3}{4}}}{\log N})$・高速化版) * @docs docs/multiplicative-function/prime-counting.md */ int main() { ll l, r; cin >> l >> r; auto f = [&](ll l, ll r) { return PrimeCounting::prime_counting(r) - (l > 1 ? PrimeCounting::prime_counting(l - 1) : 0); }; ll ans = f(l, r); if(r > l)ans += f(2* l + 1, 2 * r-1); cout << ans << endl; }