n,d=map(int,input().split())
t_i=[]
k_i=[]
for i in range(n):
    t,k=list(map(int,input().split()))
    t_i.append(t)
    k_i.append(k)

for i in range(1,n):
    t_i[i]=max(t_i[i-1],k_i[i-1]-d)+t_i[i]
    k_i[i]=max(t_i[i-1]-d,k_i[i-1])+k_i[i]
print(max(max(t_i,k_i))-d)