n,m = map(int,input().split())
cum = []
for i in range(n):
    cum.append(sum(map(int,input().split())))
dp = [[0]*2 for i in range(n+1)]
for i in range(n):
    dp[i+1][0] = max(dp[i][0],dp[i][1]+cum[i])
    dp[i+1][1] = max(dp[i][1],dp[i][0]-cum[i])
print(max(dp[-1]))