I try to use heavy path decomposition to ask LCA

but I found the code run slowly particularly in the condition of chain

What going on? What wrong with my heavy path decomposition?

#include<bits/stdc++.h>
using namespace std;
#define int long long
#define debug cout<<-1<<endl
const int maxn=2e5+5;
int n,m;
int a,b;
int uu,vv;
int siz[maxn];
int dep[maxn];
int top[maxn];
int son[maxn];
int vis[maxn];
int fa[maxn];
vector<int>q[maxn];
int maxx=0;
struct node{
	int l,r,lca;
}nod[maxn];
bool cmp(node a,node b){
	return dep[a.lca]>dep[b.lca];
}
void dfs1(int u,int last){
	dep[u]=dep[fa[u]]+1;
	siz[u]=1;
	for(int i=0;i<q[u].size();i++){
		int v=q[u][i];
		if(v!=fa[u]){
			fa[v]=u;
			dfs1(v,u);
			siz[u]+=siz[v];
				
			if(siz[v]>maxx)	{
				son[u]=v;
				maxx=siz[v];
			}
		}
	}
}
void dfs2(int u,int tp){
	top[u]=tp;
	if(son[u]){
		dfs2(son[u],tp);	
	}
	for(int i=0;i<q[u].size();i++){
		int v=q[u][i];
		if(v==son[u]||v==fa[u]){
			continue;
		}
		dfs2(v,v);
	}
}
void dfs3(int u,int last){
	vis[u]=1;
	for(int i=0;i<q[u].size();i++){
		int v=q[u][i];
		if(v!=fa[u]){
			dfs3(v,u);
		}
	}
}
int lca(int a,int b){
	while(top[b]!=top[a]){
		if(dep[a]>dep[b]){
			swap(a,b);//depa <depb
		}
		b=fa[top[b]];
	}
	if(dep[a]>dep[b]){
		swap(a,b);//depa <depb
	}
	return a;
}
signed main(){
	freopen("paths.in","r",stdin);
	freopen("paths.out","w",stdout);
	scanf("%lld%lld",&n,&m);
	int opt=0;
	for(int i=1;i<=n-1;i++){
		scanf("%lld%lld",&a,&b);
		q[a].push_back(b);
		q[b].push_back(a);
	}
	dfs1(1,0);
	dfs2(1,1);
	fa[1]=1;
	for(int i=1;i<=m;i++){
		scanf("%lld%lld",&uu,&vv);
		nod[i].l=uu;
		nod[i].r=vv;
		nod[i].lca=lca(uu,vv);	
	}
	sort(nod+1,nod+m+1,cmp);
	int ans=0;
	for(int i=1;i<=m;i++){
		if(!vis[nod[i].l]&&!vis[nod[i].r]&&!vis[nod[i].lca]){
			dfs3(nod[i].lca,fa[nod[i].lca]);
			ans++;
		}
	}
	cout<<ans<<endl;
}