You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
After further inspection, both generate_rsa and generate_ec_secp256r1 test purposedly against the fixed unseeded RNG, so this is normal behavior for these two tests. I suggest we use two test RNGs where one is deterministic and used for these type of tests.
We need investigation for private_from_rsa_components_sanity.
The current test RNG is deterministic:
and initialized in
fn test_rng()
asUsing a different RNG and running tests repeatedly reveals test failures:
(In my experiments, I replaced
pub struct TestRandom(XorShiftRng);
withpub type TestRandom = CtrDrbg;
and initialize withOsEntropy
)The text was updated successfully, but these errors were encountered: