Re: CFB Bug?

Jan Luehe (luehe@laguna.eng.sun.com)
Thu, 12 Aug 1999 16:37:44 -0700 (PDT)

Message-Id: <199908122337.QAA12935@laguna.eng.sun.com>
Date: Thu, 12 Aug 1999 16:37:44 -0700 (PDT)
From: Jan Luehe <luehe@laguna.eng.sun.com>
Subject: Re: CFB Bug?
To: java-security@java.sun.com, jepolito@lucent.com

Jonathan:

The SunJCE provider yields the exact same results as
the Crypto-J provider from RSA Data Security, Inc, when
using the algorithm/mode/padding combination specified
in your test program.

Jan

> Date: Thu, 12 Aug 1999 15:56:09 -0400
> From: Jonathan Polito - Lucent ASCC <jep@ascc.lucent.com>
> X-Accept-Language: en
> MIME-Version: 1.0
> To: java-security@java.sun.com
> Subject: CFB Bug?
>
>
> Using jce1_2-do.jar (latest JCE) CFB/NoPadding is not giving the same
> results as other implementations, namely:
> Cryptix 3.0.3 (java 1.1)
> SSLeay/OpenSSL
> Entrust/IAIK (java 1.2)
>
> CBC mode gives the same results. I have run tests on Solaris/sparc and
> Cygwin32/x86. Simply, if you encrypt some plaintext with the same key
> and iv, you do not get the same results with Sun's JCE as you do with
> the other packages (which all agree). Attached is the sample program for
> Java2/JCE 1.2 (Solaris VM (build Solaris_JDK_1.2.1_03, native threads,
> sunwjit) with JCE jce1_2-doc.jar).
>
> Regards,
> Jonathan Polito.
>
>