N,D = list(map(int,input().split()))

dp = [[0 for i in range(2)] for j in range(N+1)]

for i in range(1,N+1):
    T,K = list(map(int,input().split()))
    if i!=1:
        dp[i][0] = max(dp[i-1][1]-D+T ,dp[i-1][0]+T)
        dp[i][1] = max(dp[i-1][0]-D+K ,dp[i-1][1]+K)
    else:
        dp[i][0] = max(dp[i-1][1]-D+T ,dp[i-1][0]+T)
        dp[i][1] = K-D

print(max(dp[N]))