Re: [GNU Crypto] EMSA_PSS.MGF Bug? Version 2.0.1

[Casey Marshall]

>>>>> "TAKIGUCHI" == TAKIGUCHI Naruhito <address@hidden> writes:

TAKIGUCHI> Hi, Marshall.
>> I think you are correct, and that indeed using a '<=' is correct
>> here.
>> 
>> Have you tried GNU Crypto's version with that change? Does it work
>> then?

TAKIGUCHI> I have changed version 2.0.1's source, and it works
TAKIGUCHI> successfully.  ( Trailing some bytes of the dbMask are not
TAKIGUCHI> 0x00.)

Ok, I'll check a fix into the stable branch sometime soon.

>> Also, if you have a test case for this (say, an example signature) I
>> would like to add it to the test suite.

TAKIGUCHI> Sorry, I have no test case yet.

TAKIGUCHI> What do you need a set for test caese?  private
TAKIGUCHI> key(privateExponent, modulus, publicExponent), message and
TAKIGUCHI> signature?

Yeah, a private key, message, and known-good signature would be fine,
and the test case would just generate the signature and check it
against the good signature.

-- 
Casey Marshall || address@hidden