Issue 132 in keyczar: Fixed Cpp wrong IV size for AES keysizes above 128

[key...@googlecode.com]

Status: Accepted
Owner: jtu...@gmail.com
Labels: Type-Defect Priority-High Implementation-Cpp

New issue 132 by jtu...@gmail.com: Fixed Cpp wrong IV size for AES keysizes
above 128
http://code.google.com/p/keyczar/issues/detail?id=132

Before 2010 the java version generated an IV based on keysize instead of
blocksize even though the aes implementations just ignored the rest. The
c++ version compensated for this by padding the IV's with zeros to match
the size. However this was corrected in the java version in 2010, but never
fixed in C++.

Patch:
https://code.google.com/r/jtuley-keyczar-dev/source/detail?r=479ad523003084e333320b7d36032243991f4011&name=cpp/jtuley_dev

Tests:
https://code.google.com/r/jtuley-keyczar-dev/source/detail?r=07144658e516a151e1d976821deaaf9dc5aa1a07&name=cpp/jtuley_dev

I just to remove the padding code and fixed the reading code so it produces
interoperable ciphertexts with java, python, go & c#. I did not try to
implement any backwards compatibility with the broken c++ implementation.

--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[key...@googlecode.com]

Comment #3 on issue 132 by dlundb...@google.com: Fixed Cpp wrong IV size

for AES keysizes above 128
http://code.google.com/p/keyczar/issues/detail?id=132

Just to get this into master before the next release. Here's a commit with
C++ constant naming.
https://code.google.com/p/keyczar/source/detail?r=3ce82e9699661ebdce6b08475b718dc350ecdb78&name=cpp_big_aes_keys

[key...@googlecode.com]

Updates:
Status: Fixed

Comment #4 on issue 132 by dlundb...@google.com: Fixed Cpp wrong IV size

for AES keysizes above 128
http://code.google.com/p/keyczar/issues/detail?id=132

Merged into master
https://code.google.com/p/keyczar/source/detail?r=3ce82e9699661ebdce6b08475b718dc350ecdb78