The correctness of the garbler result

[marvellous987]

Hi,

I am using the semihonest_2pc.rs in examples for testing some garbled circuit txts. I found that the sha256 circuit comes from https://homes.esat.kuleuven.be/~nsmart/MPC/old-circuits.html where it provides some test vectors. However, I found that the result is not compatible with the given ones...

There are only one correct with all 1 (0xffff.ffff) as the first (garbler) input, but the remaining aren't... I did think of the endian problem, but the case with all 1 is correct and all 0 is not correct...

Could anyone give me some help or any approaches?

Thank you very much!

[themighty1]

May or may not be similar to what we encountered here #14

[marvellous987]

Hi, thanks for the reply.

I think not really. The case now is about the circuit provided in the repo (circuits/sha-256.txt), I checked that it is the same as the one in https://homes.esat.kuleuven.be/~nsmart/MPC/old-circuits.html but the semihonest_2pc.rs outputs a different result, except the one with 0xfffff...fff input.

[marvellous987]

I did not modify any in the code, apart from some println!.