3.6.4:
New features
* Build Python 3.7 wheels on Linux, Windows and Mac.
Resolved issues
* Rename _cpuid module to make upgrades more robust.
* More meaningful exceptions in case of mismatch in IV length (CBC/OFB/CFB modes).
* Fix compilation issues on Solaris 10/11.
3.6.3:
Resolved issues
GH175: Fixed incorrect results for CTR encryption/decryption with more than 8 blocks.
3.6.2:
New features
ChaCha20 accepts 96 bit nonces (in addition to 64 bit nonces) as defined in RFC7539.
Accelerate AES-GCM on x86 using PCLMULQDQ instruction.
Accelerate AES-ECB and AES-CTR on x86 by pipelining AESNI instructions.
As result of the two improvements above, on x86 (Broadwell):
AES-ECB and AES-CTR are 3x faster
AES-GCM is 9x faster
Resolved issues
On Windows, MPIR library was stilled pulled in if renamed to gmp.dll.
Breaks in compatibility
In Crypto.Util.number, functions floor_div and exact_div have been removed. Also, ceil_div is limited to non-negative terms only.
3.6.1:
New features
Added Google Wycheproof tests (https://github.com/google/wycheproof) for RSA, DSA, ECDSA, GCM, SIV, EAX, CMAC.
New parameter mac_len (length of MAC tag) for CMAC.
Resolved issues
In certain circumstances (at counter wrapping, which happens on average after 32 GBi) AES GCM produced wrong ciphertexts.
Method encrypt() of AES SIV cipher could be still called, whereas only encrypt_and_digest() should be allowed.
3.6.0:
New features
Introduced export_key and deprecated exportKey for DSA and RSA key objects.
Ciphers and hash functions accept memoryview objects in input.
Added support for SHA-512/224 and SHA-512/256.
Resolved issues
Reintroduced Crypto.__version__ variable as in PyCrypto.
Fixed compilation problem with MinGW.
New features
* Import and export of ECC curves in compressed form.
* The initial counter for a cipher in CTR mode can be a byte string (in addition to an integer).
* Faster PBKDF2 for HMAC-based PRFs (at least 20x for short passwords, more for longer passwords). Thanks to Christian Heimes for pointing out the implementation was under-optimized.
* The salt for PBKDF2 can be either a string or bytes.
Resolved issues
* Without libgmp, modular exponentiation (since v3.4.8) crashed on 32-bit big-endian systems.
Breaks in compatibility
* Removed support for Python < 2.6.
3.4.11:
Resolved issues
* GH-121. the record list was still not correct due to PEP3147
and __pycache__ directories.
3.4.10:
Resolved issues
* When creating ElGamal keys, the generator wasn't a square residue:
ElGamal encryption done with those keys cannot be secure under
the DDH assumption.
3.4.9:
New features
* More meaningful error messages while importing an ECC key.
Resolved issues
* GH-123 and #125. The SSE2 command line switch was not always passed on 32-bit x86 platforms.
* GH-121. The record list (--record) was not always correctly filled for the
pycryptodomex package.
3.4.8:
New features
* Added a native extension in pure C for modular exponentiation, optimized for SSE2 on x86.
In the process, we drop support for the arbitrary arithmetic library MPIR
on Windows, which is painful to compile and deploy.
The custom modular exponentiation is 130% (160%) slower on an Intel CPU in 32-bit (64-bit) mode,
compared to MPIR. Still, that is much faster that CPython's own `pow()` function which
is 900% (855%) slower than MPIR. Support for the GMP library on Unix remains.
* Added support for *manylinux* wheels.
* Support for Python 3.7.
Resolved issues
* The DSA parameter 'p' prime was created with 255 bits cleared
(but still with the correct strength).
* Not all docs were included in the tar ball.
* ECDSA verification failed for DER encoded signatures.
* Human-friendly messages for padding errors with ECB and CBC.
3.4.7:
New features
* API documentation is made with sphinx instead of epydoc.
* Start using ``importlib`` instead of ``imp``` where available.
Resolved issues
* Fixed PEM header for RSA/DSA public keys.
The library can also be compiled using MinGW.
Removed use of alloca().
[Security] Removed implementation of deprecated "quick check" feature of PGP block cipher mode.
Improved the performance of scrypt by converting some Python to C.
