[
https://issues.apache.org/jira/browse/QPID-3522?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13123913#comment-13123913
]
[email protected] commented on QPID-3522:
-----------------------------------------------------
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/2271/
-----------------------------------------------------------
(Updated 2011-10-10 06:32:28.713798)
Review request for Alan Conway, michael goulish and Chug Rolke.
Changes
-------
Made changes pointed out by Chuck. Thanks, Chuck!
Summary
-------
The stubs and skeletons used to invoke and handle AMQP 0-10 commands/controls
at present does not distinguish between null and an empty string. It turns out
this distinction is critical for Cyrus SASL integration.
Rather than altering the general approach - which I fear would mean a large
patch and the potential for lots of irritating bugs to creep in - I've
restricted the change to the specific control of relevance here. By operating
on the command body directly rather than using the parameter list to- and from-
which it is converted, the appropriate check can be made.
I have also had to alter the SASL interfaces to make the same distinction with
regard to the initial response.
This addresses bug QPID-3522.
https://issues.apache.org/jira/browse/QPID-3522
Diffs (updated)
-----
/trunk/qpid/cpp/src/qpid/Sasl.h 1179157
/trunk/qpid/cpp/src/qpid/SaslFactory.cpp 1179157
/trunk/qpid/cpp/src/qpid/broker/ConnectionHandler.h 1179157
/trunk/qpid/cpp/src/qpid/broker/ConnectionHandler.cpp 1179157
/trunk/qpid/cpp/src/qpid/broker/SaslAuthenticator.h 1179157
/trunk/qpid/cpp/src/qpid/broker/SaslAuthenticator.cpp 1179157
/trunk/qpid/cpp/src/qpid/broker/windows/SaslAuthenticator.cpp 1179157
/trunk/qpid/cpp/src/qpid/client/ConnectionHandler.cpp 1179157
/trunk/qpid/cpp/src/qpid/client/windows/SaslFactory.cpp 1179157
/trunk/qpid/cpp/src/tests/ssl_test 1179157
Diff: https://reviews.apache.org/r/2271/diff
Testing
-------
Tested both CRAM-MD5 and EXTERNAL work. Added automated test for EXTERNAL which
was previously missing. I haven't tested at all on windows however.
Thanks,
Gordon
> SASL EXTERNAL mechanism no longer works
> ---------------------------------------
>
> Key: QPID-3522
> URL: https://issues.apache.org/jira/browse/QPID-3522
> Project: Qpid
> Issue Type: Bug
> Components: C++ Broker
> Affects Versions: 0.13
> Reporter: Gordon Sim
> Assignee: Gordon Sim
> Priority: Blocker
> Fix For: 0.13
>
>
> Seems to be as a result of QPID-3393 (i.e. regression since 0.12) which was a
> fix for CRAM-MD5. From a simplistic point of view it seems like the CRAM-MD5
> mechanism requires an empty response string to be treated as null, whereas
> for the EXTERNAL mechanism an empty response should be treated as a zero
> length string. It may be though that there is more to this than that.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
---------------------------------------------------------------------
Apache Qpid - AMQP Messaging Implementation
Project: http://qpid.apache.org
Use/Interact: mailto:[email protected]
