crypto-pubkey-types rsa updates

GaloisInc / RSA

Haskell RSA Library

Other

20 stars 7 forks source link

crypto-pubkey-types rsa updates #4

Closed vincenthz closed 12 years ago

vincenthz commented 12 years ago

Hi, i'm planning to update the private key RSA structure with the following patch:

    -    { private_size :: Int     -- ^ size of key in bytes
    -    , private_n    :: Integer -- ^ private p*q
    +    { private_pub  :: PublicKey -- ^ public part of a private key (size, n and e)

This makes the public key a sub field of a private key, instead of having 2 duplicate fields (private_n and private_size). In this case, e is useful for doing crypto-blinding when using chinese remainder optimisations.

is this update acceptable for RSA ? if so i'll do a pull request, otherwise i'm willing to discuss other possible solutions.

acw commented 12 years ago

Are these the only two reasons to combine the structures? It seems a bit confusing to me for some reason. I'd almost rather duplicate e as well, if that's what you're really looking for.

vincenthz commented 12 years ago

yes, i agree it's not necessarily ideal, and i'm not totally sure how to handle it the best way. However practically it seems to make sense:

ASN1 format of a "private key" has all the information about the public key (e&n)
It's useful for signing and decryption computation to get n&e available. e is quite important to have if using CRT optimisations, recomputing the inverse of d might cancel all improvements.
It's easier to compare the compatibility of a public key with a private key (private_pub priv == pub). It's easier than comparing the fields if they were duplicated, or worse recomputing e*d === 1 [phi].
Likewise it's easier to extract the public key from the private key.
it seems cleaner than to duplicate fields, as it's easier to see which parts is which. Maybe also capturing the asymmetric part of RSA.

acw commented 12 years ago

OK, you've convinced me. Send me a pull request and I'll pull it in.