#pragma GCC optimize ("Ofast")
#include<bits/stdc++.h>
using namespace std;
void *wmem;
template<class T> inline void walloc1d(T **arr, int x, void **mem = &wmem){
  static int skip[16]={0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1};
  (*mem) = (void*)( ((char*)(*mem)) + skip[((unsigned long long)(*mem)) & 15] );
  (*arr)=(T*)(*mem);
  (*mem)=((*arr)+x);
}
inline void rd(int &x){
  int k, m=0;
  x=0;
  for(;;){
    k = getchar_unlocked();
    if(k=='-'){
      m=1;
      break;
    }
    if('0'<=k&&k<='9'){
      x=k-'0';
      break;
    }
  }
  for(;;){
    k = getchar_unlocked();
    if(k<'0'||k>'9'){
      break;
    }
    x=x*10+k-'0';
  }
  if(m){
    x=-x;
  }
}
inline void wt_L(char a){
  putchar_unlocked(a);
}
inline void wt_L(int x){
  char f[10];
  int m=0, s=0;
  if(x<0){
    m=1;
    x=-x;
  }
  while(x){
    f[s++]=x%10;
    x/=10;
  }
  if(!s){
    f[s++]=0;
  }
  if(m){
    putchar_unlocked('-');
  }
  while(s--){
    putchar_unlocked(f[s]+'0');
  }
}
template<class S, class T> inline S chmin(S &a, T b){
  if(a>b){
    a=b;
  }
  return a;
}
template<class T> struct segtree{
  T *addval, *fixval, *mn, *sum;
  char *fixed;
  int N, logN, *mnind;
  void malloc(int maxN, int once = 0){
    int i;
    for(i=1;i<maxN;i*=2){
      ;
    }
    sum = (T*)std::malloc(sizeof(T)*2*i);
    mn = (T*)std::malloc(sizeof(T)*2*i);
    mnind = (int*)std::malloc(sizeof(T)*2*i);
    fixval = (T*)std::malloc(sizeof(T)*i);
    addval = (T*)std::malloc(sizeof(T)*i);
    fixed = (char*)std::malloc(sizeof(char)*i);
    if(once){
      setN(maxN);
    }
  }
  void walloc(int maxN, int once = 0, void **mem = &wmem){
    int i;
    for(i=1;i<maxN;i*=2){
      ;
    }
    walloc1d(&sum, 2*i, mem);
    walloc1d(&mn, 2*i, mem);
    walloc1d(&mnind, 2*i, mem);
    walloc1d(&fixval, i, mem);
    walloc1d(&addval, i, mem);
    walloc1d(&fixed, i, mem);
    if(once){
      setN(maxN);
    }
  }
  void free(void){
    std::free(sum);
    std::free(mn);
    std::free(mnind);
    std::free(fixval);
    std::free(addval);
    std::free(fixed);
  }
  T& operator[](int i){
    return sum[N+i];
  }
  void setN(int n, int zerofill = 1, int dobuild = 1){
    int i;
    for(i=1,logN=0;i<n;i*=2,logN++){
      ;
    }
    N = i;
    if(zerofill){
      for(i=0;i<(N);i++){
        sum[N+i] = 0;
      }
    }
    if(dobuild){
      build();
    }
  }
  void build(void){
    int i;
    for(i=0;i<(N);i++){
      mn[N+i] = sum[N+i];
      mnind[N+i] = i;
    }
    for(i=N-1;i;i--){
      sum[i] = sum[2*i] + sum[2*i+1];
      if(mn[2*i] <= mn[2*i+1]){
        mn[i] = mn[2*i];
        mnind[i] = mnind[2*i];
      }
      else{
        mn[i] = mn[2*i+1];
        mnind[i] = mnind[2*i+1];
      }
    }
    for(i=(1);i<(N);i++){
      fixed[i] = 0;
    }
    for(i=(1);i<(N);i++){
      addval[i] = 0;
    }
  }
  inline void push_one(int a, int sz, int st){
    if(fixed[a]){
      if(sz > 1){
        fixed[a*2] = fixed[a*2+1] = 1;
        fixval[a*2] = fixval[a*2+1] = fixval[a];
        sum[a*2] = sum[a*2+1] = sz * fixval[a];
        mn[a*2] = mn[a*2+1] = fixval[a];
        mnind[a*2] = st;
        mnind[a*2+1] = st + sz;
      }
      else{
        sum[a*2] = sum[a*2+1] = sz * fixval[a];
        mn[a*2] = mn[a*2+1] = fixval[a];
        mnind[a*2] = st;
        mnind[a*2+1] = st + sz;
      }
      fixed[a] = 0;
      addval[a] = 0;
      return;
    }
    if(addval[a] != 0){
      if(sz > 1){
        if(fixed[a*2]){
          fixval[a*2] += addval[a];
        }
        else{
          addval[a*2] += addval[a];
        }
        if(fixed[a*2+1]){
          fixval[a*2+1] += addval[a];
        }
        else{
          addval[a*2+1] += addval[a];
        }
        sum[a*2] += sz * addval[a];
        sum[a*2+1] += sz * addval[a];
        mn[a*2] += addval[a];
        mn[a*2+1] += addval[a];
      }
      else{
        sum[a*2] += sz * addval[a];
        sum[a*2+1] += sz * addval[a];
        mn[a*2] += addval[a];
        mn[a*2+1] += addval[a];
      }
      addval[a] = 0;
      return;
    }
  }
  inline void push(int a){
    int aa=a - N, i, nd, st, sz;
    for(i=logN;i;i--){
      nd = a>>i;
      sz = 1<<(i-1);
      st = 2 * sz * (aa>>i);
      push_one(nd, sz, st);
    }
  }
  inline void build(int a){
    int st=a - N, sz=1;
    while(a > 1){
      if(a%2){
        st += sz;
      }
      a /= 2;
      sz *= 2;
      if(fixed[a]){
        sum[a] = sz * fixval[a];
        mn[a] = fixval[a];
      }
      else{
        sum[a] = sum[a*2] + sum[a*2+1];
        if(mn[a*2] <= mn[a*2+1]){
          mn[a] = mn[a*2];
          mnind[a] = mnind[a*2];
        }
        else{
          mn[a] = mn[a*2+1];
          mnind[a] = mnind[a*2+1];
        }
        if(addval[a] != 0){
          mn[a] += addval[a];
          sum[a] += sz * addval[a];
        }
      }
    }
  }
  inline void change(int a, int b, T val){
    int aa, bb, st_a=a, st_b=b, sz=1;
    if(a >= b){
      return;
    }
    aa = (a += N);
    bb = (b += N);
    push(a);
    push(b-1);
    if(a%2){
      sum[a] = mn[a] = val;
      a++;
      st_a += sz;
    }
    if(b%2){
      b--;
      st_b -= sz;
      sum[b] = mn[b] = val;
    }
    a /= 2;
    b /= 2;
    while(a < b){
      sz *= 2;
      if(a%2){
        fixed[a]=1;
        fixval[a]=val;
        sum[a] = sz * val;
        mn[a] = val;
        mnind[a] = st_a;
        a++;
        st_a += sz;
      }
      if(b%2){
        b--;
        st_b -= sz;
        fixed[b]=1;
        fixval[b]=val;
        sum[b] = sz * val;
        mn[b] = val;
        mnind[b] = st_b;
      }
      a /= 2;
      b /= 2;
    }
    build(aa);
    build(bb-1);
  }
  inline void add(int a, int b, T val){
    int aa, bb, sz=1;
    if(a >= b){
      return;
    }
    aa = (a += N);
    bb = (b += N);
    push(a);
    push(b-1);
    if(a%2){
      sum[a] += val;
      mn[a] += val;
      a++;
    }
    if(b%2){
      b--;
      sum[b] += val;
      mn[b] += val;
    }
    a /= 2;
    b /= 2;
    while(a < b){
      sz *= 2;
      if(a%2){
        if(fixed[a]){
          fixval[a] += val;
        }
        else{
          addval[a] += val;
        }
        sum[a] += sz * val;
        mn[a] += val;
        a++;
      }
      if(b%2){
        b--;
        if(fixed[b]){
          fixval[b] += val;
        }
        else{
          addval[b] += val;
        }
        sum[b] += sz * val;
        mn[b] += val;
      }
      a /= 2;
      b /= 2;
    }
    build(aa);
    build(bb-1);
  }
  inline pair<T,int> getMin(int a, int b){
    int sz=1;
    pair<T,int> res;
    a += N;
    b += N;
    push(a);
    push(b-1);
    res.first = numeric_limits<T>::max();
    res.second = -1;
    while(a < b){
      if(a%2){
        chmin(res, make_pair(mn[a], mnind[a]));
        a++;
      }
      if(b%2){
        b--;
        chmin(res, make_pair(mn[b], mnind[b]));
      }
      a /= 2;
      b /= 2;
    }
    return res;
  }
  inline T getMinVal(int a, int b){
    return getMin(a,b).first;
  }
  inline int getMinInd(int a, int b){
    return getMin(a,b).second;
  }
  inline T getSum(int a, int b){
    T res;
    int sz=1;
    a += N;
    b += N;
    push(a);
    push(b-1);
    res = 0;
    while(a < b){
      if(a%2){
        res += sum[a++];
      }
      if(b%2){
        res += sum[--b];
      }
      a /= 2;
      b /= 2;
    }
    return res;
  }
}
;
char memarr[96000000];
int N;
int Q;
int A[100000];
int T;
int L;
int R;
int main(){
  int KL2GvlyY, i, k;
  segtree<int> t;
  wmem = memarr;
  rd(N);
  rd(Q);
  {
    int Lj4PdHRW;
    for(Lj4PdHRW=0;Lj4PdHRW<(N);Lj4PdHRW++){
      rd(A[Lj4PdHRW]);A[Lj4PdHRW] += (-1);
    }
  }
  t.malloc(N);
  t.setN(N);
  for(i=0;i<(N);i++){
    t[i] = A[i];
  }
  t.build();
  for(KL2GvlyY=0;KL2GvlyY<(Q);KL2GvlyY++){
    rd(T);
    rd(L);L += (-1);
    rd(R);R += (-1);
    if(T==1){
      swap(A[L], A[R]);
      t.change(L,L+1,A[L]);
      t.change(R,R+1,A[R]);
    }
    else{
      k = t.getMinInd(L, R+1);
      wt_L(k+1);
      wt_L('\n');
    }
  }
  return 0;
}
// cLay varsion 20190902-1

// --- original code ---
// int N, Q, A[1d5], T, L, R;
// {
//   int k;
//   segtree<int> t;
//   rd(N,Q,(A--)(N));
//   t.malloc(N);
//   t.setN(N);
//   rep(i,N) t[i] = A[i];
//   t.build();
//   rep(Q){
//     rd(T,L--,R--);
//     if(T==1){
//       swap(A[L], A[R]);
//       t.change(L,L+1,A[L]);
//       t.change(R,R+1,A[R]);
//     } else {
//       k = t.getMinInd(L, R+1);
//       wt(k+1);
//     }
//   }
// }