Oh hi Codeforces. I'm doing problem Longest Flight Route. At first I'm like "hum ... directed graph but with no circle, gotta find the longest path from 1 to n using normal BFS and I'm gonna be find ...". So it turned out that I didn't get AC because somehow I got TLE verdict on 2 test cases ... I checked the runtime of other test and saw that longest runtime for other test that got AC is like 0.09 sec so it must my algorithm's fault. I did some more check and found that something has caused my BFS to run like forever. However, it's not possible I think. My solution is just the reversed version of normal BFS : with current node u and it's child v, I check if (d[v] < d[u]+1) then update d[v], push v in the queue and I'm done (Oh sorry, I should have mention that d[i] is the longest path from 1 to i). This graph is guaranteed to have no circle, so how the heck did my code got 2 TLE

const int inf = 0x3c3c3c3c, N = 1e5+1;
int n, m, d[N], p[N];
vt <int> adj[N]; // vector is often defined as vt in my program

void solve() {
	read(n, m); // read function does exactly what you think it does
	for (int i = 0; i < m; i++) {
		int u, v;
		read(u, v);
		adj[u].pb(v); // push_back
	}
	reset(d, -1); // memset d to -1
	queue <int> q;
	d[1] = 0;
	q.push(1);
	while (sz(q)) { // sz(q) == q.size()
		int u = q.front();
		q.pop();
		for (int &v : adj[u]) {
			if (d[v] < d[u] + 1) {
				d[v] = d[u] + 1;
				p[v] = u;
				q.push(v);
			}
		}
	}
	if (d[n] == -1) {
		write("IMPOSSIBLE");
	} else {
		vt <int> path;
		int x = n;
		while (x != 0) {
			path.pb(x);
			x = p[x];
		}
		print(sz(path)); // print(something) equal to cout <<  something << endl;
		for (int i = sz(path) - 1; i >= 0; i--) write(path[i], " "); // work like cout
	}
}

Pls leave some of your thought about this problem and what should I do to fix my code ... Can't see why BFS run forever.