1. Added a check on the hex2bn function for null strings to be treated like the NULL pointer.
2. Added some more test cases to the big number unit test.
3. Added wolfSSL_EVP_add_cipher() analogous to wolfSSL_EVP_add_digest().
4. Implemented `RSA_public_encrypt()` and `RSA_private_decrypt()` for the OpenSSL compatibility layer.
5. Added `RSA_PKCS1_OAEP_PADDING` as an option to the new functions.
6. Fixed the constness on the `from` parameter in function `RSA_public_decrypt()`. Note: this does not allow OAEP to be used in FIPS mode.
7. Added size constants for EVP_MAX_BLOCK_LENGTH, RSA_PKCS1_PADDING_SIZE, and RSA_PKCS1_OAEP_PADDING_SIZE.
8. Added functions `wolfSSL_EVP_PKEY_new_mac_key()` and `wolfSSL_EVP_PKEY_get0_hmac()` for storing HMAC keys in an EVP_PKEY structure for use later.
9. Added the routines EVP_DigestSignXXX(). They only work w.r.t HMAC.
10. Added the ForceZero() to EVP_MD_CTX_cleanup().
11. Modified the EVP_MD_CTX structure to have a union with both the hasher and a wolfCrypt Hmac structure.
12. Added a mapping of SHA1->SHA to the list of digest name aliases in wolfSSL_EVP_get_digestbyname().
13. Changed the mapping of ssl3-sha1->SHA.
14. When using RSA blinding, the OpenSSL compatibility layer RSA object was attempting to decrypt without an RNG. Added an RNG if needed.
15. Add comment to wolfSSL_BN_hex2bn() to cover usage w.r.t. string length. Remember to keep the length even, front-pad with zeroes if needed.
16. Added tests for the new functions.
wolfSSL_set_secret() was checking the preMasterSz against the sizeof
the preMasterSecret member of the Arrays structure. That member
was changed to a pointer and allocated dynamically for the session
write-duping. The comparison between the passed in size and the size
of the parameter started failing. The check now uses the constant
that is used for allocating the preMasterSecret member.
1. Restore original EmbedRecvFrom callback.
2. Add new EmbedRecvFromMcast callback. (EmbedSendTo still the same.)
3. Set new receive from callback when setting the member ID.
4. Fixed bad rebase change.
1. Adding the prototypes for the sequence number high water callback.
2. Added the accessors to set the highwater callback function,
trigger levels, and application context.
3. Calls the highwater callback at specified sequence number thresholds
per peer.
1. Separated the peer ID from the array index into the peer sequence
list. This allows peer IDs to range from 0..255, and to have an
arbitrary size for the sequence list.
2. Add API to add and remove peer IDs from the sequence number list.
