Struct ring::signature::RsaKeyPair
source · pub struct RsaKeyPair { /* private fields */ }
Expand description
An RSA key pair, used for signing.
Implementations§
source§impl RsaKeyPair
impl RsaKeyPair
sourcepub fn from_pkcs8(pkcs8: &[u8]) -> Result<Self, KeyRejected>
pub fn from_pkcs8(pkcs8: &[u8]) -> Result<Self, KeyRejected>
Parses an unencrypted PKCS#8-encoded RSA private key.
Only two-prime (not multi-prime) keys are supported. The public modulus (n) must be at least 2047 bits. The public modulus must be no larger than 4096 bits. It is recommended that the public modulus be exactly 2048 or 3072 bits. The public exponent must be at least 65537.
This will generate a 2048-bit RSA private key of the correct form using OpenSSL’s command line tool:
openssl genpkey -algorithm RSA \
-pkeyopt rsa_keygen_bits:2048 \
-pkeyopt rsa_keygen_pubexp:65537 | \
openssl pkcs8 -topk8 -nocrypt -outform der > rsa-2048-private-key.pk8
This will generate a 3072-bit RSA private key of the correct form:
openssl genpkey -algorithm RSA \
-pkeyopt rsa_keygen_bits:3072 \
-pkeyopt rsa_keygen_pubexp:65537 | \
openssl pkcs8 -topk8 -nocrypt -outform der > rsa-3072-private-key.pk8
Often, keys generated for use in OpenSSL-based software are stored in the Base64 “PEM” format without the PKCS#8 wrapper. Such keys can be converted to binary PKCS#8 form using the OpenSSL command line tool like this:
openssl pkcs8 -topk8 -nocrypt -outform der \
-in rsa-2048-private-key.pem > rsa-2048-private-key.pk8
Base64 (“PEM”) PKCS#8-encoded keys can be converted to the binary PKCS#8 form like this:
openssl pkcs8 -nocrypt -outform der \
-in rsa-2048-private-key.pem > rsa-2048-private-key.pk8
The private key is validated according to NIST SP-800-56B rev. 1 section 6.4.1.4.3, crt_pkv (Intended Exponent-Creation Method Unknown), with the following exceptions:
-
Section 6.4.1.2.1, Step 1: Neither a target security level nor an expected modulus length is provided as a parameter, so checks regarding these expectations are not done.
-
Section 6.4.1.2.1, Step 3: Since neither the public key nor the expected modulus length is provided as a parameter, the consistency check between these values and the private key’s value of n isn’t done.
-
Section 6.4.1.2.1, Step 5: No primality tests are done, both for performance reasons and to avoid any side channels that such tests would provide.
-
Section 6.4.1.2.1, Step 6, and 6.4.1.4.3, Step 7:
- ring has a slightly looser lower bound for the values of
p
andq
than what the NIST document specifies. This looser lower bound matches what most other crypto libraries do. The check might be tightened to meet NIST’s requirements in the future. Similarly, the check thatp
andq
are not too close together is skipped currently, but may be added in the future.
- The validity of the mathematical relationship of
dP
,dQ
,e
andn
is verified only during signing. Some size checks ofd
,dP
anddQ
are performed at construction, but some NIST checks are skipped because they would be expensive and/or they would leak information through side channels. If a preemptive check of the consistency ofdP
,dQ
,e
andn
with each other is necessary, that can be done by signing any message with the key pair.
d
is not fully validated, neither at construction nor during signing. This is OK as far as ring’s usage of the key is concerned because ring never uses the value ofd
(ring always usesp
,q
,dP
anddQ
via the Chinese Remainder Theorem, instead). However, ring’s checks would not be sufficient for validating a key pair for use by some other system; that other system must check the value ofd
itself ifd
is to be used.
- ring has a slightly looser lower bound for the values of
In addition to the NIST requirements, ring requires that p > q
and
that e
must be no more than 33 bits.
See RFC 5958 and RFC 3447 Appendix A.1.2 for more details of the encoding of the key.
sourcepub fn from_der(input: &[u8]) -> Result<Self, KeyRejected>
pub fn from_der(input: &[u8]) -> Result<Self, KeyRejected>
Parses an RSA private key that is not inside a PKCS#8 wrapper.
The private key must be encoded as a binary DER-encoded ASN.1
RSAPrivateKey
as described in RFC 3447 Appendix A.1.2). In all other
respects, this is just like from_pkcs8()
. See the documentation for
from_pkcs8()
for more details.
It is recommended to use from_pkcs8()
(with a PKCS#8-encoded key)
instead.
sourcepub fn public_modulus_len(&self) -> usize
pub fn public_modulus_len(&self) -> usize
Returns the length in bytes of the key pair’s public modulus.
A signature has the same length as the public modulus.
source§impl RsaKeyPair
impl RsaKeyPair
sourcepub fn sign(
&self,
padding_alg: &'static dyn RsaEncoding,
rng: &dyn SecureRandom,
msg: &[u8],
signature: &mut [u8]
) -> Result<(), Unspecified>
pub fn sign( &self, padding_alg: &'static dyn RsaEncoding, rng: &dyn SecureRandom, msg: &[u8], signature: &mut [u8] ) -> Result<(), Unspecified>
Sign msg
. msg
is digested using the digest algorithm from
padding_alg
and the digest is then padded using the padding algorithm
from padding_alg
. The signature it written into signature
;
signature
’s length must be exactly the length returned by
public_modulus_len()
. rng
may be used to randomize the padding
(e.g. for PSS).
Many other crypto libraries have signing functions that takes a
precomputed digest as input, instead of the message to digest. This
function does not take a precomputed digest; instead, sign
calculates the digest itself.
Lots of effort has been made to make the signing operations close to constant time to protect the private key from side channel attacks. On x86-64, this is done pretty well, but not perfectly. On other platforms, it is done less perfectly.