[Ace-users] [tao-users] [TAO]: AMI+DSI+AMH facility usage ...a few problems, and their fixes

[Douglas C. Schmidt]

Hi Venkat,

   Thanks for using the PRF.

>    TAO VERSION: 1.6.1
>    ACE VERSION: 5.6.1
>
>    HOST MACHINE and OPERATING SYSTEM:
>        amd64, NetBSD 3.1
>
>    TARGET MACHINE and OPERATING SYSTEM, if different from HOST:
>             same
>
>   THE $ACE_ROOT/ace/config.h FILE [if you use a link to a platform-
>specific file, simply state which one]:
>                 config-netbsd.h
>
>    THE $ACE_ROOT/include/makeinclude/platform_macros.GNU FILE :
>      platform_netbsd.GNU
>
>    CONTENTS OF $ACE_ROOT/bin/MakeProjectCreator/config/
>default.features
>              ssl=1
>
>    AREA/CLASS/EXAMPLE AFFECTED:
>            tao/DynamicInterface
>            TAO/tests/DSI_AMH
>            TAO/tests/DSI_AMI_Gateway
>
>    DOES THE PROBLEM AFFECT:
>       COMPILATION? No
>       LINKING? No
>       EXECUTION? Yes
>       TAO and application are effected
>
>    SYNOPSIS:
>       Following problems are seen:
>(1) Incorrect implementation of TAO_AMH_DSI_Response_Handler_var, and
>other miscellaneous
>(2)  Mostly Memory related problems - Several types of memory leaks
>(3)  Incorrect usage of amh_response_handler_allocator() at
>application (tests)
>
>   DESCRIPTION:
>
>        While verifying the DIS/AMI/AMH features via TAO/tests/
>DSI_AMH, TAO/tests/DSI_AMI_Gateway, I came across several types of
>problems related to memory management. I have not done exhaustive
>tests, though - various exceptions generations. I notice several TODO
>comments, so there probably is intent to evolve the code in future.
>
>        With some changes I made, I am successful in running and
>removing several memory leaks. Most of my testing at this time is
>limited to "normal" behavior; that is, no exceptions are generated in
>the gateway or passed through gateway. I would pursue this later.
>
>        Unfortunately, I'm still seeing some memory leaks. I'm very
>sure that there are no leaks - in the tests I've performed - in test
>programs nor in the response and reply handlers. How could I say
>that?
>
>        I have taken 5.5.7/1.5.7 release, and ran the same tests with
>the same changes I attached to this PRF. I ran the client for a few
>thousands of iterations, and there is no process memory growth.
>Without my fixes, I see lots of process growth.
>
>        I notice changes in the guts of tao (/orb) between the
>releases. It's too big a space to cover by me. I don't have any memory
>leak too at this time for my platform!
>
>     Any help in knowing/fixing the leaks is very appreciated.
>
>    REPEAT BY:
>
>    SAMPLE FIX/WORKAROUND:
>        I've fixes, but I dont' see an option to attach to this post.
>(somehow my direct post to tao-bugs is reflected. I subscribed to it.)

It would be great if you could please send us patches that fix the
problems.  You might also want to add the patches to bugzilla 

http://deuce.doc.wustl.edu/bugzilla/

so we don't lose track of them!

Thanks,

        Doug
-- 
Dr. Douglas C. Schmidt                       Professor and Associate Chair
Electrical Engineering and Computer Science  TEL: (615) 343-8197
Vanderbilt University                        WEB: www.dre.vanderbilt.edu/~schmidt
Nashville, TN 37203                          NET: d.schmidt at vanderbilt.edu