Re: [PATCH v2 0/9] MR4637: bcrypt DH improvements

[Hans Leidekker (＠hans)]

On Fri Dec  8 11:29:40 2023 +0000, Paul Gofman wrote:
> Another observation, also related to already merged part:
> key_asymmetric_derive_key() relies on gnutls_datum_t returned from
> gnutls_privkey_derive_secret() to have correct size. I didn't check with
> gnutls code if it pads the secret to the expected length on output, but
> based on all the other gnutls functions which return not a crypt value
> but generic big number in gnutls_datum_t which may be shorter than
> expected (without leading zeroes) or instead may have zero prepended to
> avoid negative value, would it probably be better to use
> export_gnutls_datum() with specified size like elsewhere?

This is fixed as well, thanks.