Closed zdohnal closed 1 year ago
Fixes #762 .
I've tried to compare results with hashes from GNUTLS:
Hashing "Hello world" by GNUTLS by all supported hashes:
Hash: md5 Final hash size: 16 >%� y�ƛgL��g�, Hash: sha Final hash size: 20 {P,:H�`����c��9g? Hash: sha2-224 Final hash size: 28 �#��w��n���Xd��� Hash: sha2-256 Final hash size: 32 d�� Hash: sha2-384 Final hash size: 48 �Ҙ�6'��Xx�c7��2���&P�諌'�3��ߔ�� Hash: sha2-512 Final hash size: 64 ����킗�ۑtbO���i������& Hash: sha2-512_224 Final hash size: 28 ����킗�ۑtbO���i������& Hash: sha2-512_256 Final hash size: 32 ����킗�ۑtbO���i������&
Hashing by OpenSSL (OpenSSL supports those truncated algorithms by default, but it gives a different output...):
Hash: md5 Final hash size: 16 >%� y�ƛgL��g�, Hash: sha Final hash size: 20 {P,:H�`����c��9g? Hash: sha2-224 Final hash size: 28 �#��w��n���Xd��� Hash: sha2-256 Final hash size: 32 d�� Hash: sha2-384 Final hash size: 48 �Ҙ�6'��Xx�c7��2���&P�諌'�3��ߔ�� Hash: sha2-512 Final hash size: 64 ����킗�ۑtbO���i������& Hash: sha2-512_224 Final hash size: 28 �����g�O�[Y4/QO>F����� Hash: sha2-512_256 Final hash size: 32 ��XrԮ�h;���(�v�6/�Z���K��_
@michaelrsweet WDYT?
@michaelrsweet ok, I'll take the fix from libcups without breaking API (without two block hash feature).
This should be it.
Fixes #762 .
I've tried to compare results with hashes from GNUTLS:
Hashing "Hello world" by GNUTLS by all supported hashes:
Hashing by OpenSSL (OpenSSL supports those truncated algorithms by default, but it gives a different output...):
@michaelrsweet WDYT?