kzhong130 / emp-ag2pc-max-example

emp-ag2pc-max-example

What is this about

• I am trying to use ag2pc to run the following simple example

n = 3;
int int_size = 32;
Integer *A = new Integer[n];
Integer *B = new Integer[n];
Integer *C = new Integer[n];
Integer *D = new Integer[n];
Integer *E = new Integer[n];

for (int i = 0; i < n; ++i) {
    A[i] = Integer(int_size, bid_vec[i], ALICE);
    B[i] = Integer(int_size, bid_vec[i], BOB);
}

for (int i = 0; i < n; ++i) {
    C[i] = A[i];
    D[i] = B[i];
    E[i] = A[i] + B[i];
}

for (int i = 0; i < n; i++) {
    C[i].reveal<uint32_t>();
}

for (int i = 0; i < n; i++) {
    D[i].reveal<uint32_t>();
}

for (int i = 0; i < n; i++) {
    E[i].reveal<uint32_t>();
}

• Both ALICE and BOB provides N (in this example, it's 3) numbers, and I try to reveal the original input (N numbers from both party) and the sum of them.

• But it turns out that after transforming the output into nine integers, the results is a bit weird. Suppose ALICE inputs 1 2 3 and BOB inputs 4 5 6. The expected output should be A[0]: 1, A[1]: 2, A[2]: 3, B[0]: 4, B[1]: 5, B[2]: 6, A[0]+B[0]: 5, A[1]+B[1]: 7, A[2]+B[2]: 9.

• However, the output from my side is A[0]: 4, A[1]: 6, A[2]: 2, B[0]: 5, B[1]: 1, B[2]: 3, A[0]+B[0]: 9, A[1]+B[1]: 7, A[2]+B[2]: 5.

Way I generate input for circuit

bool* in = new bool[max(cf.n1, cf.n2)];
for (size_t i = 0; i < bids.size(); i++) {
    int_to_bool(in + i * int_size, bids[i], int_size);
}

• The codes above describe how I generate the boolean array required by circuit, bascially I am using int_to_bool() function provides in utils.hpp and transform one input into the right place in the boolean array.

How to run

• Generate the circuit file by compiling max-cuircuit-gen.cpp, and execute it to generate the circuit file. The circuit file should be named as max.txt. I have also attached a circuit file generated from my machine (max-circuit.txt) for reference.

• Please modify line 14: string curcuit_file = "emp/emp-tool/max.txt"; in max-ag2.cpp and then compile it.

• You can use this command to run the binary, $PATH_TO_max-ag2 1 12345 1 2 3 & $PATH_TO_max-ag2 2 12345 4 5 6 > /dev/null to see the output. The last three arguments are input for that party.

Version

• I am using the latest version of emp-ag2pc.