Request for phase 2 approval for CR 7031830 - bad_record_mac failure on TLSv1.2 enabled connection with SSLEngine

[Brad Wetmore]

Edvard/Dalibor,

I think I know the answer to this, but wanted to ask anyway.

There is a broken JDK test on windows-x64 under the jdk_security3 
target.  It doesn't compile due to a missing library PKCS11, and thus 
fails each/every time.

In JDK8, we added this test to the test/ProblemList.txt file.  I can do 
the same for 7u2, or I can just leave it alone.  (This entry was a small 
part of a much larger test cleanup.)

http://hg.openjdk.java.net/jdk8/tl/jdk/annotate/cb83fe13af98/test/ProblemList.txt

weijun at 4455

     + # Directly references PKCS11 class
     + sun/security/pkcs11/Provider/Absolute.java        windows-x64

The downside is that every time jdk_security3 is run on 7uXX, someone 
has to manually check that failure.

Let me know in the next couple hours, otherwise I'll just go forward 
with just the putback for the bad MAC issue.

Brad



On 10/18/2011 10:36 AM, Edvard Wendelin wrote:
> Approved.
>
> Cheers,
> Edvard
>
> On 17 okt 2011, at 20.37, Brad Wetmore wrote:
>
>>
>> Bug: http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=7031830
>>
>> Webrev: http://cr.openjdk.java.net/~wetmore/7031830/
>> See the *.01 revision.
>>
>> Review:
>> http://mail.openjdk.java.net/pipermail/security-dev/2011-October/003915.html
>>
>> The security-dev mailing list.
>>
>> Rvwrs: Xuelei Fan, Weijun Wang, Abhijit Saha
>> (xuelei, weijun, asaha)
>>
>> Important customer needs this fix for their transition to JDK7
>> (customer info is in the bug report). TLS 1.1/1.2 is broken when using
>> byte-buffer backed Java arrays, and needs to be fixed.
>>
>> Brad
>>
>