#include using namespace std; #if __has_include() #include using namespace atcoder; templateistream &operator>>(istream &is,static_modint &a){long long b;is>>b;a=b;return is;} istream &operator>>(istream &is,modint &a){long long b;cin>>b;a=b;return is;} #endif #ifdef LOCAL #include "debug.h" #else #define debug(...) static_cast(0) #define debugg(...) static_cast(0) templateostream &operator<<(ostream &os,const pair&p){os<; templateusing minque=priority_queue,greater>; templatebool chmax(T &a,const T &b){return (abool chmin(T &a,const T &b){return (a>b?(a=b,true):false);} templateistream &operator>>(istream &is,pair&p){is>>p.first>>p.second;return is;} templateistream &operator>>(istream &is,vector &a){for(auto &i:a)is>>i;return is;} templatevoid operator++(pair&a,int n){a.first++,a.second++;} templatevoid operator--(pair&a,int n){a.first--,a.second--;} templatevoid operator++(vector&a,int n){for(auto &i:a)i++;} templatevoid operator--(vector&a,int n){for(auto &i:a)i--;} #define reps(i,a,n) for(int i=(a);i<(n);i++) #define rep(i,n) reps(i,0,n) #define all(x) x.begin(),x.end() #define pcnt(x) __builtin_popcount(x) ll myceil(ll a,ll b){return (a+b-1)/b;} template auto vec(const int (&d)[n],const T &init=T()){ if constexpr (id(d,init)); else return init; } void SOLVE(); int main(){ ios::sync_with_stdio(false); cin.tie(nullptr); #ifdef LOCAL clock_t start=clock(); #endif int testcase=1; cin>>testcase; for(int i=0;i>n>>m; swap(n,m); vectordp(3); dp[0]=1,dp[1]=m,dp[2]=mint::raw(m)*(m-1)/2-m; rep(i,n-1){ vectorep(3); ep[0]=dp[0]+dp[1]+dp[2]; ep[1]=dp[0]*m+dp[1]*(m-1)+dp[2]*(m-2); ep[2]=dp[0]*(mint::raw(m)*(m-1)/2-m)+dp[1]*(mint::raw(m)*(m-1)/2-m-(m-3))+dp[2]*(mint::raw(m-2)*max(0,m-3)/2-(m-4)); swap(dp,ep); } cout<<(dp[0]+dp[1]+dp[2]).val()<