Skip Menu |

Preferred bug tracker

Please visit the preferred bug tracker to report your issue.

This queue is for tickets about the Crypt-Rijndael CPAN distribution.

Report information
The Basics
Id: 27632
Status: resolved
Priority: 0/
Queue: Crypt-Rijndael

People
Owner: Nobody in particular
Requestors: JGG [...] dolby.com
Cc:
AdminCc:

Bug Information
Severity: (no value)
Broken in: (no value)
Fixed in: (no value)



Subject: Bug in Crypt::Rijndael, demo script included.
Date: Mon, 18 Jun 2007 18:11:16 -0700
To: <bug-Crypt-Rijndael [...] rt.cpan.org>
From: "Gilbert, John" <JGG [...] dolby.com>
Hello, If you run the following code (from Lincoln Stein) on Linux it returns 6ba7dc3661305b3479ac06df40d053a5, on Solaris it outputs 9ed879854c4cb2266265045315886072. Solaris returns the same value on AMD64 and UltraSparc. Thanks. John G. #!/usr/bin/perl use strict; use Crypt::Rijndael; my $cipher = new Crypt::Rijndael "a" x 32, Crypt::Rijndael::MODE_CBC; $cipher->set_iv('a' x 16); open MD5,"|md5sum -" or die; print MD5 $cipher->encrypt('testing testing testing testing testing testing '); close MD5; ----------------------------------------- This message (including any attachments) may contain confidential information intended for a specific individual and purpose. If you are not the intended recipient, delete this message. If you are not the intended recipient, disclosing, copying, distributing, or taking any action based on this message is strictly prohibited.
Hi John, Can you tell me which version of Linux and Solaris you are using, which versions of Perl you are using, and which C compilers you're using? Thanks,
Hi John, Can you tell me which version of Linux and Solaris you are using, which versions of Perl you are using, and which C compilers you're using? Thanks,
Subject: RE: [rt.cpan.org #27632] Bug in Crypt::Rijndael, demo script included.
Date: Tue, 19 Jun 2007 10:22:18 -0700
To: <bug-Crypt-Rijndael [...] rt.cpan.org>
From: "Gilbert, John" <JGG [...] dolby.com>
Hi Brian, Yes. Solaris 8, 9, 10 and 11 (OpenSolaris) using Perl built with Sun Studio 10 and 11 on V20s (AMD64), V240s (UltraSPARC III), and a SunBlade 100 (UltraSPARC IIi), with both 64 bit and 32 bit Perl builds, Homebuilt and Sun provided, all return one value. Slackware Linux 9, 10, and 11, the latest Ubuntu, fedora and SUSE, with perl built with gcc 3.6 and 4.1.2, all 32 bit as far as I know all return the other value. Perl is 5.8.8. Crypt::Rijndael is 0.04 and 1.04 on various systems, upgrading doesn't seem to effect outcome. It looks like a gcc vs Sun compiler split as the same AMD64 box running both will get both results. If you do manage to fix this, hopefully you can create a flag so that a translation program is possible. We have about 2.2 million files that are in Solaris only format. I have not yet tested Cygwin/Windows, MacOSX, IRIX, etc. Thanks. John G. Show quoted text
-----Original Message----- From: brian_d_foy via RT [mailto:bug-Crypt-Rijndael@rt.cpan.org] Sent: Tuesday, June 19, 2007 8:17 AM To: Gilbert, John Subject: [rt.cpan.org #27632] Bug in Crypt::Rijndael, demo script included. <URL: http://rt.cpan.org/Ticket/Display.html?id=27632 > Hi John, Can you tell me which version of Linux and Solaris you are using, which versions of Perl you are using, and which C compilers you're using? Thanks, ----------------------------------------- This message (including any attachments) may contain confidential information intended for a specific individual and purpose. If you are not the intended recipient, delete this message. If you are not the intended recipient, disclosing, copying, distributing, or taking any action based on this message is strictly prohibited.
Crypt::Rijndael-1.04_01 is on it's way to CPAN. It creates a test file for this bug, but does not include a fix. This release is to see if other systems experience the problem.
I've just uploaded Crypt-Rijndael-1.04_03. This version takes more care with the size of integers. I think some of the stuff that you might have been seeing was signed ints where there should have been unsigned ints. I tested this with Solaris 10 / gcc 3.4.6 under my VMWare image on my MacBook Pro. It compiled cleanly and passed all the tests.
Subject: RE: [rt.cpan.org #27632] Bug in Crypt::Rijndael, demo script included.
Date: Mon, 15 Oct 2007 14:50:22 -0700
To: <bug-Crypt-Rijndael [...] rt.cpan.org>
From: "Gilbert, John" <JGG [...] dolby.com>
I just tested this on an UltraSPARC 3i, Solaris 10, Sun Studio 12, everything seems to match a reference AES implementation now. I'll keep testing (need to try an x86_64 with a Sun Studio compiled perl), and I'll now need to code up a conversion program for our old data. Thanks! John G. Show quoted text
-----Original Message----- From: brian_d_foy via RT [mailto:bug-Crypt-Rijndael@rt.cpan.org] Sent: Monday, October 15, 2007 12:43 PM To: Gilbert, John Subject: [rt.cpan.org #27632] Bug in Crypt::Rijndael, demo script included. <URL: http://rt.cpan.org/Ticket/Display.html?id=27632 > I've just uploaded Crypt-Rijndael-1.04_03. This version takes more care with the size of integers. I think some of the stuff that you might have been seeing was signed ints where there should have been unsigned ints. I tested this with Solaris 10 / gcc 3.4.6 under my VMWare image on my MacBook Pro. It compiled cleanly and passed all the tests. ----------------------------------------- This message (including any attachments) may contain confidential information intended for a specific individual and purpose. If you are not the intended recipient, delete this message. If you are not the intended recipient, disclosing, copying, distributing, or taking any action based on this message is strictly prohibited.