Adds verification of RSA signatures #26
Conversation
| [ (TInt 1, TInt 3) | ||
| , (TInt 3, TInt (-257)) | ||
| , (TInt (-1), TBytes (EC.i2osp n)) | ||
| , (TInt (-2), TBytes (EC.i2osp e)) |
There was a problem hiding this comment.
You probably want to use EC.i2ospOf_ 32 as above, as otherwise I fear it might do the wrong things if there are leading zeros.
There was a problem hiding this comment.
I commented on these lines earlier:
From section 4: https://datatracker.ietf.org/doc/html/rfc8230#section-4
All numeric key parameters are encoded in an unsigned big-endian
representation as an octet sequence using the CBOR byte string
type (major type 2). The octet sequence MUST utilize the minimum
number of octets needed to represent the value.
According to https://hackage.haskell.org/package/cryptonite-0.29/docs/Crypto-PubKey-RSA.html#v:generateWith common choices for e are 0x10001 or 3
which makes me doubt 32bytes "utilizes the minimum number of octets needed to represent the value"
Co-authored-by: Joachim Breitner <[email protected]>
|
@nomeata I think I adressed your feedback. PTAL |
| @@ -72,94 +77,122 @@ genClientDataJson challenge = JSON.encode $ JSON.Object $ | |||
| <> "type" JSON..= ("webauthn.get" :: T.Text) | |||
| <> "origin" JSON..= ("ic-ref-test" :: T.Text) | |||
|
|
|||
| parseCOSEKey :: BS.ByteString -> Either T.Text EC.PublicKey | |||
| parseCOSEKey s = do | |||
| verifyCOSESig :: BS.ByteString -> BS.ByteString -> BS.ByteString -> Either T.Text () | |||
There was a problem hiding this comment.
Any chance we can name these types properly so the signature becomes more meaningful?
There was a problem hiding this comment.
(Sorry for the late response, I was on PTO the last two days)
This matches the already exported signature of verify, and it's a module internal function. If we wanted to go with better names we should probably start at verify.
I think part of the reason we're dealing with lots of ByteStrings in ic-ref's API is that it's primarily a tool used for testing, so it wants to be low-friction for putting together a couple of testcases, rather than having to construct via various newtype wrappers.
No description provided.