OJ Board

Hi, I tried this problem today but got WA.

I thought I implemented the rules of concurrency simulation described well, and I've already searched this forum for help / tricky parts as well as test datas generated by others. But I still can't find my problem is. I think it works well.

So would you ...

i didn't find anything wrong in the base case of solve(k)

Matrix solve(int k)
{
if(k==1) return m;
Matrix ret=solve(k/2);
Matrix pow=power(m,k/2);
Matrix tmp=add(ret,mul(pow,ret));
if(k%2==1)
{
Matrix tt=mul(pow,pow);
Matrix newpow=mul(tt,m);
Matrix ans=add(tmp,newpow);
return ans ...

Can you give me some further hint on this problem?
I first thought to use the dynamic programming method that is similar to that in the Hamilton Path which runs in O(2^n*n^2) time but then found one can past a node several times...

my searching + optimization got TLE....

wondering what is the ...

Problem solved...so careless in the base case of solve(k).

code removed because of getting accepted