//#include <bits/stdc++.h>
#include <iostream>
#include <iomanip>
#include <math.h>
#include <cmath>
#include <algorithm>
#include <climits>
#include <functional>
#include <cstring>
#include <string>
#include <cstdlib>
#include <ctime>
#include <cstdio>
#include <vector>
#include <stack>
#include <queue>
#include <deque>
#include <map>
#include <set>
#include <bitset>
#include <complex>
//#include <ext/pb_ds/assoc_container.hpp>
//#include <ext/pb_ds/tree_policy.hpp>
#define itn int
#define nit int
#define ll long long
#define ms multiset
#define F(i,a,b) for(register int i=a,i##end=b;i<=i##end;++i)
#define UF(i,a,b) for(register int i=a,i##end=b;i>=i##end;--i)
#define re register
#define ri re int
#define il inline
#define pii pair<int,int>
#define cp complex<double>
//#pra gma G CC opti mize(3)
using namespace std;
using std::bitset;
//using namespace __gnu_pbds;
const double Pi=acos(-1);
namespace fastIO {
	template<class T>
	inline void read(T &x) {
		x=0;
		bool fu=0;
		char ch=0;
		while(ch>'9'||ch<'0') {
			ch=getchar();
			if(ch=='-')fu=1;
		}
		while(ch<='9'&&ch>='0') x=(x*10-48+ch),ch=getchar();
		if(fu)x=-x;
	}
	inline int read() {
		int x=0;
		bool fu=0;
		char ch=0;
		while(ch>'9'||ch<'0') {
			ch=getchar();
			if(ch=='-')fu=1;
		}
		while(ch<='9'&&ch>='0') x=(x*10-48+ch),ch=getchar();
		return fu?-x:x;
	}
	template<class T,class... Args>
	inline void read(T& t,Args&... args) {
		read(t);
		read(args...);
	}
	char _n_u_m_[40];
	template<class T>
	inline void write(T x ) {
		if(x==0){
			putchar('0');
			return;
		}
		T tmp = x > 0 ? x : -x ;
		if( x < 0 ) putchar('-') ;
		register int cnt = 0 ;
		while( tmp > 0 ) {
			_n_u_m_[ cnt ++ ] = tmp % 10 + '0' ;
			tmp /= 10 ;
		}
		while( cnt > 0 ) putchar(_n_u_m_[ -- cnt ]) ;
	}
	template<class T>
	inline void write(T x ,char ch) {
		write(x);
		putchar(ch);
	}
}
using namespace fastIO;
const int MAXN=1000005,MAXM=4000005;
int head[MAXN],nxt[MAXM],n,m,to[MAXM],x,y,sta[MAXM],cnt,num;
vector<pii >ans;
bool vis[MAXM>>1];
inline void dfs(int pos){
	for(int &e=head[pos];e;e=nxt[e]){
		if(vis[e])continue;
		int temp=e;
		vis[e]=1;
		dfs(to[e]);
		sta[++cnt]=temp;
	}
}
inline void ed(int x,int y){
	to[++num]=y;
	nxt[num]=head[x];
	head[x]=num;
}
vector<int>v[402][402];
vector<int>A;
int main() {
	cin>>n>>m;
	A.push_back(0);
	F(i,1,n){
		F(j,1,m){
			int w=read();A.push_back(w);
			v[i][w].push_back((i-1)*m+j);
		}
		F(j,1,m){
			if(v[i][j].empty())ed(MAXN-j,i);
			if(v[i][j].size()>=2)F(k,1,v[i][j].size()-1)ed(i,MAXN-j);
		}
	}
	A.push_back(0);
	F(i,1,MAXN-1)if(head[i]&&!vis[head[i]]){
		cnt=0;
		dfs(i);
		reverse(sta+1,sta+cnt+1);
		F(j,1,cnt)sta[j]=to[sta[j]];
		sta[0]=sta[cnt],sta[cnt+1]=sta[1];
		vector<int>R;
		F(j,1,cnt){
			if(sta[j]>n){
				R.push_back(v[sta[j-1]][MAXN-sta[j]].back());
				v[sta[j-1]][MAXN-sta[j]].pop_back();
			}
		}
		
		ans.emplace_back(R.back(),n*m+1);
		
		UF(i,R.size()-2,0)ans.emplace_back(R[i],R[i+1]);
		ans.emplace_back(n*m+1,R[0]);
	}
	write(ans.size(),'\n');
	for(auto i:ans){
		write(i.first,' ');write(i.second,'\n');
		A[i.second]=A[i.first];
	}
//	F(i,1,n*m+1)cerr<<A[i]<<" ";
	return 0;
}
/*
5 6
4 1 3 1 6 5 2 3 2 3 5 6 2 1 4 5 6 4 1 3 2 4 5 5 1 2 3 4 6 6
*/