path: root/TAO/ChangeLogs/ChangeLog-02a

Thu Aug 23 15:54:44 2001  Nanbor Wang  <nanbor@cs.wustl.edu>

	* tao/TAO.dsp: Flipped on "generate browse information" flag.  We
	  need to remove this setting when we merge the branch into the
	  main trunk.

	* tao/Abstract_Servant_Base.h:
	* tao/Object.cpp:
	* tao/Object.h:
	* tao/Object_Proxy_Impl.h:
	* tao/Remote_Object_Proxy_Impl.cpp:
	* tao/Remote_Object_Proxy_Impl.h:
	* tao/PortableServer/Direct_Object_Proxy_Impl.cpp:
	* tao/PortableServer/Direct_Object_Proxy_Impl.h:
	* tao/PortableServer/Servant_Base.cpp:
	* tao/PortableServer/Servant_Base.h:
	* tao/PortableServer/ThruPOA_Object_Proxy_Impl.cpp:
	* tao/PortableServer/ThruPOA_Object_Proxy_Impl.h: Added
	  _get_component() as one of the default operations an CORBA
	  Object supports.  The default implementation is to return a nil
	  object reference.

Fri Aug 17 14:04:20 2001  Nanbor Wang  <nanbor@cs.wustl.edu>

	* TAO_IDL/be/be_sequence.cpp:
	* TAO_IDL/be/be_visitor_sequence/gen_bounded_obj_sequence_ch.cpp:
	* TAO_IDL/be/be_visitor_sequence/gen_bounded_obj_sequence_ci.cpp:
	* TAO_IDL/be/be_visitor_sequence/gen_bounded_obj_sequence_cs.cpp:
	* TAO_IDL/be/be_visitor_sequence/gen_unbounded_obj_sequence_ch.cpp:
	* TAO_IDL/be/be_visitor_sequence/gen_unbounded_obj_sequence_ci.cpp:
	* TAO_IDL/be/be_visitor_sequence/gen_unbounded_obj_sequence_cs.cpp:
	  More fixes to handle the forward declared valuetypes correctly.4

Thu Aug 16 21:05:19 2001  Nanbor Wang  <nanbor@cs.wustl.edu>

	* TAO_IDL/be/be_sequence.cpp:
	* TAO_IDL/be/be_visitor_sequence/gen_bounded_obj_sequence_ch.cpp:
	* TAO_IDL/be/be_visitor_sequence/gen_bounded_obj_sequence_ci.cpp:
	* TAO_IDL/be/be_visitor_sequence/gen_bounded_obj_sequence_cs.cpp:
	* TAO_IDL/be/be_visitor_sequence/gen_unbounded_obj_sequence_ch.cpp:
	* TAO_IDL/be/be_visitor_sequence/gen_unbounded_obj_sequence_ci.cpp:
	* TAO_IDL/be/be_visitor_sequence/gen_unbounded_obj_sequence_cs.cpp:
	  Fixed IDL_Test.

Thu Aug 16 15:16:18 2001  Nanbor Wang  <nanbor@cs.wustl.edu>

	* TAO_IDL/be/be_sequence.cpp:
	* TAO_IDL/be/be_visitor_sequence/buffer_type.cpp:
	* TAO_IDL/be/be_visitor_sequence/cdr_op_cs.cpp:
	* TAO_IDL/be/be_visitor_sequence/elemtype.cpp:
	* TAO_IDL/be/be_visitor_sequence/gen_bounded_obj_sequence_ch.cpp:
	* TAO_IDL/be/be_visitor_sequence/gen_bounded_obj_sequence_ci.cpp:
	* TAO_IDL/be/be_visitor_sequence/gen_bounded_obj_sequence_cs.cpp:
	* TAO_IDL/be/be_visitor_sequence/gen_unbounded_obj_sequence_ch.cpp:
	* TAO_IDL/be/be_visitor_sequence/gen_unbounded_obj_sequence_ci.cpp:
	* TAO_IDL/be/be_visitor_sequence/gen_unbounded_obj_sequence_cs.cpp:
	* TAO_IDL/be/be_visitor_sequence/sequence_base.cpp:
	* TAO_IDL/be/be_visitor_sequence/sequence_ch.cpp:
	* TAO_IDL/be/be_visitor_sequence/sequence_ci.cpp:
	* TAO_IDL/be/be_visitor_sequence/sequence_cs.cpp:
	* TAO_IDL/be_include/be_sequence.h:
	* TAO_IDL/be_include/be_visitor_sequence/buffer_type.h:
	* TAO_IDL/be_include/be_visitor_sequence/cdr_op_cs.h:
	* TAO_IDL/be_include/be_visitor_sequence/elemtype.h:
	* TAO_IDL/be_include/be_visitor_sequence/sequence_base.h: Allowed
	  using valuetypes in sequences.  This is required by the CCM IDL.

Mon Aug 13 16:48:05 2001  Nanbor Wang  <nanbor@cs.wustl.edu>

	* tao/Sequence_T.h:
	* tao/Sequence_T.i:
	* tao/Sequence_T.cpp: Added TAO_Valuetype_Manager,
	  TAO_Unbounded_Valuetype_Sequence, and
	  TAO_Bounded_Valuetype_Sequence.  Valuetype's memory is managed
	  thru _add_ref and _remove_ref (as oppose to _duplicate and
	  release.)

	* examples/OBV/Sequence/README:
	* examples/OBV/Sequence/client.cpp:
	* examples/OBV/Sequence/client.dsp:
	* examples/OBV/Sequence/registry_i.cpp:
	* examples/OBV/Sequence/registry_i.h:
	* examples/OBV/Sequence/sequence.dsw:
	* examples/OBV/Sequence/series.idl:
	* examples/OBV/Sequence/series_i.cpp:
	* examples/OBV/Sequence/series_i.h:
	* examples/OBV/Sequence/server.cpp:
	* examples/OBV/Sequence/server.dsp: New example to show how to use
	  valuetypes in structured data type.  Currently only sequence is
	  tested and we'll need to hack the IDL compiler to get this to
	  work out of the box.

	* examples/OBV/Sequence/seriesC.cpp.bak:
	* examples/OBV/Sequence/seriesC.h.bak:
	* examples/OBV/Sequence/seriesC.i.bak: These files showed what
	  needs to be done for the test to run.

Thu Aug  9 16:15:03 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

	* tests/RTCORBA/ORB_init/ORB_init.cpp: Fixed a warning in KCC
	  builds on Linux.

Thu Aug  9 11:10:59 2001  Ossama Othman  <ossama@uci.edu>

	* tao/default_resource.cpp:

	  Temporarily reverted previous change.  The default resource
	  factory is still destroyed before the ORB core is destroyed.
	  Another solution is needed.

Thu Aug  9 11:46:27 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

	* tao/CDR.h: Fixed the ambigous constructor problem. The InputCDR
	  class had two constructors which were  similar if the default
	  values for the arguments were exercised. Thanks to Torbjorn
	  Backstrom <torbjorn.k.backstrom@volvo.com>  for reporting this.

Wed Aug  8 22:53:37 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

	* tao/DLL_ORB.cpp: Fixed a unused variable warning.

Wed Aug  8 16:15:59 2001  Carlos O'Ryan  <coryan@uci.edu>

        * tests/AMI_Buffering/client.cpp:
        * tests/Oneway_Buffering/client.cpp:
          The liveness test was blowing the stack on the server.  On each
          request the server calls the 'admin' interface to report the
          number of bytes received.  While waiting for a response on that
          call the server receives another call (the client is pushing
          oneways or AMI calls like crazy).
          This results in an segfault when the stack is overflowed, now we
          sync the server every so many calls.

        * tests/AMI_Buffering/svc.conf:
          Removed, the default configuration works now.

        * tests/Big_Oneways/Session.cpp:
          Made the test more silent.

Wed Aug 08 16:14:01 2001  Ossama Othman  <ossama@uci.edu>

        * tao/default_resource.h (TAO_Default_Resource_Factory):

          The default resource factory now also inherits from ACE_Cleanup
          so that it may be registered for destruction with the
          TAO_Singleton_Manager.  This generally only necessary for the
          default resource factory since it may be created during ORB
          initialization, and it must exist long enough for the ORB
          Core to make some calls on it during ORB Core finalization.
          Resource factories that are statically or dynamically loaded
          (i.e. registered with the Service Repository) before the ORB is
          initialized need not inherit from ACE_Cleanup, nor is it
          necessary for them to register with the TAO_Singleton_Manager.

        * tao/default_resource.cpp (_make_TAO_Default_Resource_Factory):

          A custom factory definition is used instead of the commonly used
          one created by the ACE_FACTORY_DEFINE macro.  This is necessary
          to get around dynamic unloading issues.  The default resource
          factory must exist long enough for the ORB to make the a
          resource factory reclaim the reactor, for example.  The easiest
          (and probably best) way to do that is register the
          TAO_Default_Resource_Factory for destruction with the
          TAO_Singleton_Manager.  This fixes a segmentation fault (memory
          access violation) that occurred when an ORB that was initialized
          within a dynamically loaded object (e.g. within an
          ACE_Service_Object instance) was finalized/destroyed.

          (ACE_STATIC_SVC_DEFINE):

          Do not set the ACE_Service_Type::DELETE_OBJ flag within the
          TAO_Default_Resource_Factory-specific declaration/expansion of
          this macro.  The "payload" object, i.e. the default resource
          factory, is now managed by the TAO_Singleton manager instead of
          the Service Repository.

        * tao/DLL_ORB.h (TAO_DLL_ORB):

          Updated class documentation to mention the fact that this class
          is deprecated, and why it should not be used.

        * tao/DLL_ORB.cpp (init):

          Added debugging message that mentions that the TAO_DLL_ORB class
          is deprecated.

        * tao/TAO_Internal.cpp (open_i):
        * tao/TAO_Singleton_Manager.h (TAO_Singleton_Manager):

          Updated and corrected class documentation.

        * tests/DLL_ORB/Test_Client_Module.cpp (fini):

          Explicitly release the reference to the target object.  This is
          a bit of a hack.  The ORB Core's lifetime is tied to the
          lifetime of an object reference.  We need to wipe out all object
          references before we call fini() on the TAO_Singleton_Manager.
          This is a tough problem to correct.  Fortunately, this is not a
          problem when a resource factory is loaded into the Service
          Repository before the ORB is initialized, i.e. when the default
          resource factory is not used.

Wed Aug  8 15:03:58 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Strategies/DIOP_Transport.cpp (send_request): There was no
          reason to idle the transport after a send. In other protocols,
          where the handles are cached the call idle_after_send () does
          certain things based on the muxed strategy. But for DIOP, we
          dont cache the handles and the call started doing something
          really bad like closing the handles. We now just return a 0
          after sending the request. This should fix the problem in the
          full builds with DIOP.

Wed Aug  8 12:16:22 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * performance-tests/Thread_Per_Connection_Latency/svc.conf:
        * performance-tests/Thread_Pool_Latency/svc.conf: Made the client
          to use Wait_On_Read_Write as we see a big performance drop if we
          use the Wait_On_LF. To make the test work without any problems,
          we now use the exclusive transport mux strategy. Hopefully this
          will get our performance back to normal.

        * tao/Muxed_TMS.cpp: Fixed a subtle problem in connection_closed
          (). While iterating through the reply_disptacher table to close
          the  reply dispatchers, a wrong comparision was made in the for
          loop. This would not allow any of the reply dispatchers to be
          closed and in turn prevent the transport cleaning up the
          resources. This lead to the client hanging on select () even
          after the connection is closed by the server. I am surprised
          that this has not been found out so far. God save people who
          have been using the Muxed strategy :-).

          BTW, the above fix should fix problems seen in the Faults and
          Crashed_Callback tests in the daily builds.

Tue Aug 07 21:54:31 2001  Ossama Othman  <ossama@uci.edu>

        * tests/IORManipulation/IORTest.cpp (main):

          Updated URL IORs used in test to be of the standard "corbaloc"
          form.  Fixes a CORBA::INV_OBJREF exception.

Tue Aug  7 23:29:34 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * performance-tests/Thread_Per_Connection_Latency/svc.conf:  Did
          the same change as the last checkin, ie. removed the RW
          connection handler. But beware, we *could* have a performance
          drop in our stats page. If wehave performance drop, then we
          change the Transport_Mux_Strategy for the following test. Even
          if we do change the Mux_Strategy and the connection handler to
          RW, it seems that this would be in odd with what we preached a
          few days back at the TAO workshop.

Tue Aug  7 23:02:09 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * performance-tests/Thread_Pool_Latency/svc.conf: Looks like our
          past seems to haunt us. Once upon a time this svc.conf file was
          used in Latency measurements. When the tests were changed to a
          new directory the svc.conf file was copied. The svc.conf file
          had a RW connection handler. That was enough to cause the test
          to hang. Further removed the -ORBReactor type as it was trying
          to load a TP reactor. This should fix the problem in the daily
          builds with this test.

Tue Aug  7 14:37:10 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Transport.cpp (handle_input_i): Added the transport
          information  to a debugging output.

        * tao/orbconf.h: Fixed a simple typo.

Mon Aug  6 09:49:39 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/RTPortableServer/TAO_RTPortableServer.dsp: Fixed a build
          problem in Release builds. Should have nailed the problem by
          now.

Sun Aug  5 16:58:39 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tests/DLL_ORB/Makefile: Thanks to Ossama for helping to fix the
          the warnings in the daily builds.

Sun Aug  5 16:00:28 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tests/Oneway_Buffering/Test.idl: Changed the oneway in the
          Test.idl to a twoway call.

Sun Aug 05 15:12:31 2001  Balachandran Natarajan <bala@cs.wustl.edu>

        * tao/RTPortableServer/TAO_RTPortableServer.dsp: Fixed a build
          problem in Release builds.

Sun Aug 05 15:03:31 2001  Balachandran Natarajan <bala@cs.wustl.edu>

        * tao/TAO_Static.dsp: Added the new flushing strategies to the
          static builds.

Fri Aug 04 3:33:31 2001  Yamuna Krishnamurthy  <yamuna@cs.wustl.edu>

        * TAO/orbsvcs/orbsvcs/AV/QoS_UDP.h:
        * TAO/orbsvcs/orbsvcs/AV/QoS_UDP.cpp:
        * TAO/orbsvcs/orbsvcs/AV/QoS_UDP.i:

          Added a helper class to create qos sessions, activate qos
          handlers and set qos on a session. Fixed a bug in translation
          from ACE_Flow_Spec to AVStreams::streamQoS. Fixed addressing
          problems.

Fri Aug 03 20:50:42 2001  Ossama Othman  <ossama@uci.edu>

        * tests/README:

          Added summary about the new DLL_ORB test.

Fri Aug 03 20:33:51 2001  Ossama Othman  <ossama@uci.edu>

        * tests/DLL_ORB/Makefile (MAKEFLAGS):

          Force non-parallel build of this test since the IDL file rules
          in both `Makefile.Test_Client_Module' and
          `Makefile.Test_Server_Module' cause corruption of the generated
          stubs and skeletons when these Makefiles are run in parallel.
          Fixes build problems in some of our parallel build enabled
          nightly builds.

Fri Aug 03 10:16:37 2001  Carlos O'Ryan  <coryan@uci.edu>

        * docs/Options.html:
          Add documentation for the new Leader/Followers flushing
          strategy.

Fri Aug  3 10:37:06 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * TAO_IDL/be/be_visitor_operation/operation_cs.cpp (visit_operation):

          Added ACE_UNUSED_ARG generation to satisfy strict compilers,
          for the recently modified cases where the operation returns
          a long long or a long double.

Fri Aug  3 07:59:24 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tests/DLL_ORB/Test_Client_Module.cpp: Changed the order of
          #includes.

Fri Aug  3 07:50:49 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/TAO_Singleton_Manager.h: Changed the #ifndef to
          TAO_SINGLETON_MANAGER_H instead of TAO_OBJECT_MANAGER_H.

Fri Aug  3 07:30:02 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tests/DLL_ORB/Test_Client_Module.cpp:
        * tests/DLL_ORB/Test_Server_Module.cpp: Added missing
          #includes. This should take care of the problem with g++.

Thu Aug 02 20:39:19 2001  Ossama Othman  <ossama@uci.edu>

        * tao/Stub.h (TAO_Exception_Data):

          The "id" member should be "const char *" not "char *".  Fixes a
          warning exhibited by some strict compilers (e.g. g++ with
          "-pedantic" flag enabled).  Conversion from a string constant to
          a "char *" is deprecated.

Thu Aug  2 17:21:27 2001  Ossama Othman  <ossama@uci.edu>

        * tests/Makefile:

          Added DLL_ORB test directory to the list of directories to
          build.

Thu Aug 02 17:09:53 2001  Ossama Othman  <ossama@uci.edu>

        * tests/DLL_ORB/*:

          New test that dynamically loads a shared object that initializes
          an ORB upon initialization of that shared object, and destroys
          that ORB upon finalization of the shared object.  [Bug 832]

Thu Aug  2 11:44:51 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Leader_Follower.cpp: Added a #include. This should take care
          of the No AMI builds.

Thu Aug  2 10:17:12 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * TAO_IDL/be/be_visitor_operation/operation_cs.cpp (visit_operation):

          Recent changes to this method made it necessary now to unalias
          the return type before using it.

Thu Aug  2 09:09:26 2001  Douglas C. Schmidt  <schmidt@tango.doc.wustl.edu>

        * tests/Smart_Proxies/On_Demand/client.cpp (parse_args): Fixed a
          but where there was a missing ":" in the get_opts string.
          Thanks to Goran Lowkrantz <goran.lowkrantz@ismobile.com> for
          reporting this.  Fixes bugid 995.

Wed Aug  1 18:17:38 2001  Carlos O'Ryan  <coryan@uci.edu>

        * tao/PortableServer/ImplRepoC.h:
        * tao/PortableServer/ImplRepoC.i:
        * tao/PortableServer/ImplRepoC.cpp:
        * tao/PortableServer/ImplRepoS.h:
        * tao/PortableServer/ImplRepoS.i:
        * tao/PortableServer/ImplRepoS.cpp:
        * tao/PortableServer/ImplRepoS_T.h:
        * tao/PortableServer/ImplRepoS_T.i:
        * tao/PortableServer/ImplRepoS_T.cpp:
        * tao/PortableServer/diffs/ImplRepo.diff:
          After Jeff's changes it was necessary to regenerate the ImplRepo
          pre-compiled IDL.

        * tao/diffs/ImplRepoC.cpp.diff:
        * tao/diffs/ImplRepoC.h.diff:
        * tao/diffs/ImplRepoC.i.diff:
        * tao/diffs/ImplRepoS.cpp.diff:
        * tao/diffs/ImplRepoS.h.diff:
        * tao/diffs/ImplRepoS.i.diff:
        * tao/diffs/ImplRepoS_T.cpp.diff:
        * tao/diffs/ImplRepoS_T.h.diff:
        * tao/diffs/ImplRepoS_T.i.diff:
          Move diffs to $TAO_ROOT/tao/PortableServer/ImplRepo.diff

Wed Aug  1 16:05:36 2001  Carlos O'Ryan  <coryan@uci.edu>

        * Merged changes from the fix_886 branch

        * tao/Transport.cpp:
          There was a subtle race condition in the handle_output() method:
          the state of the queue was checked in drain_queue(), while
          holding the mutex, if the queue was empty the decision was made
          to call cancel_output().
          However, that was performed *outside* the context of the mutex,
          so another thread could attempt to send data, queue it,
          schedule_output() only to find out that is was cancelled right
          after it did...
          Before the patches below it was not possible to move the
          cancel_output() to the context of the mutex, the ORB would
          deadlock for other reasons. I took us (Bala and myself) a couple
          of days to track this one down, obviously I still don't know how
          to write MT-safe code :-)

        Tue Jul 31 12:55:07 2001  Carlos O'Ryan  <coryan@uci.edu>

        * tao/Follower.h:
        * tao/Follower.inl:
        * tao/Follower.cpp:
        * tao/Follower_Auto_Ptr.h:
        * tao/Follower_Auto_Ptr.inl:
        * tao/Follower_Auto_Ptr.cpp:
        * tao/LF_Follower.h:
        * tao/LF_Follower.inl:
        * tao/LF_Follower.cpp:
        * tao/LF_Follower_Auto_Ptr.h:
        * tao/LF_Follower_Auto_Ptr.inl:
        * tao/LF_Follower_Auto_Ptr.cpp:
        * tao/Makefile:
        * tao/Makefile.am:
        * tao/Makefile.bor:
        * tao/TAO.dsp:
        * tao/TAO_Static.dsp:
          Renamed TAO_Follower to TAO_LF_Follower and
          TAO_Follower_Auto_Ptr to TAO_LF_Follower_Auto_Ptr, I think this
          is more consistent with the other files in the project.

        * tao/LF_Follower_Auto_Adder.h:
        * tao/LF_Follower_Auto_Adder.inl:
        * tao/LF_Follower_Auto_Adder.cpp:
          Remove unused code from the Auto_Adder files.

        * tao/LF_Event.h:
        * tao/LF_Event.inl:
        * tao/LF_Event.cpp:
        * tao/LF_Event_Binder.h:
        * tao/LF_Event_Binder.inl:
        * tao/Leader_Follower.h:
        * tao/Leader_Follower.i:
        * tao/Leader_Follower.cpp:
          Use the new class names and #includes.

        Mon Jul 30 14:41:43 2001  Carlos O'Ryan  <coryan@uci.edu>

        * tao/LF_Follower_Auto_Adder.h:
        * tao/LF_Follower_Auto_Adder.inl:
        * tao/LF_Follower_Auto_Adder.cpp:
          Automatically manipulate the L/F follower set: its constructor
          inserts a follower into the set and the destructor removes it.

        * tao/Leader_Follower.cpp:
          Use the new LF_Follower_Auto_Adder class.

        * tao/Makefile:
        * tao/Makefile.am:
        * tao/Makefile.bor:
        * tao/TAO.dsp:
        * tao/TAO_Static.dsp:
          Add the new files to Makefiles and projects.

        Sun Jul 29 11:58:51 2001  Carlos O'Ryan  <coryan@uci.edu>

        * tao/Leader_Follower.cpp:
          If an error is detected while waiting as a follower the loop
          should return -1.

        * tao/Transport.cpp:
          Merged in some bug fixes from the main trunk.

        Sun Jul 29 09:00:57 2001  Carlos O'Ryan  <coryan@uci.edu>

        * tao/LF_Event.cpp:
          Fixed the error_detected() function

        Fri Jul 27 17:34:40 2001  Carlos O'Ryan  <coryan@cs.wustl.edu>

        * tao/Leader_Follower.h:
        * tao/default_resource.cpp:
          Fixed warnings and compilation errors for gcc-2.7.2

        Fri Jul 27 10:59:50 2001  Carlos O'Ryan  <coryan@uci.edu>

        * tao/LF_Event.h:
        * tao/LF_Event.inl:
          Add new method to unbind a LF_Event and its Follower:
          Reply_Dispatchers can be used multiple times to wait for several
          replys, mostly when a LOCATION_FORWARD message is received.

        * tao/LF_Event.cpp:
          Modify the state machine: the state can go back to ACTIVE from
          CONNECTION_CLOSED or SUCCESSFUL.  This represents the location
          forward scenario described above.

        * tao/LF_Event_Binder.h:
        * tao/LF_Event_Binder.inl:
        * tao/LF_Event_Binder.cpp:
          Helper class to automate the bind/unbind calls to a LF_Event.

        * tao/Leader_Follower.cpp:
          Use LF_Event_Binder to handle the bind/unbind calls into the
          LF_Event.

        * tao/Makefile:
        * tao/Makefile.bor:
        * tao/TAO.dsp:
        * tao/TAO_Static.dsp:
          Add new files to the projects and Makefiles.

        * tao/Invocation.cpp:
        * tao/Wait_On_Read.cpp:
        * tao/Wait_On_Reactor.cpp:
        * tao/Synch_Reply_Dispatcher.h:
        * tao/Synch_Reply_Dispatcher.cpp:
          Remove the reply_received() flag from Synch_Reply_Dispatcher,
          the LF_Event state is enough to know what happens.

        Thu Jul 26 18:00:12 2001  Carlos O'Ryan  <coryan@uci.edu>

        * tao/Strategies/advanced_resource.cpp:
          Fixed typo in last commit.

        Thu Jul 26 16:50:46 2001  Carlos O'Ryan  <coryan@uci.edu>

        * Part of the fixes for
          http://ace.cs.wustl.edu/bugzilla/show_bug.cgi?id=886

          the changes also close the following bug:

          http://ace.cs.wustl.edu/bugzilla/show_bug.cgi?id=296

        * tao/Leader_Follower_Flushing_Strategy.h:
        * tao/Leader_Follower_Flushing_Strategy.cpp:
          New flushing strategy that participates in the Leader/Followers
          protocol.
          To support this several changes to the Leader/Followers
          implementation were required.  The most important involved using
          some abstract representation for the events that the
          Leader/Followers wait for, in the old days there were only reply
          events, so there was no need to abstract anything, but now the
          Leader/Followers set can wait for both 'message flushed' events,
          as well as 'reply received'.
          With this explicit representation for events at hand it was
          easier to encapsulate the Leader/Followers wait loop in
          TAO_Leader_Follower class, instead of hidden in
          Wait_On_Leader_Follower.
          To match the events that L/F waits for and the threads waiting
          for them we addd a class that represents a Follower thread.
          These TAO_Follower objects had to implement an intrusive list
          for fast addition into the follower set, once that intrusive
          list was implemented adding a free list was trivial, and thus we
          could solve bug 296 easily too.

        * tao/Asynch_Queued_Message.cpp:
        * tao/Synch_Queued_Message.cpp:
          Use the TAO_LF_Event methods to signal any waiters when the
          state changes.

        * tao/Follower.h:
        * tao/Follower.inl:
        * tao/Follower.cpp:
          This class represents a thread playing the Follower role.  It
          contains the condition variable used by the thread.
          The class provides the necessary hooks to implement an intrusive
          linked list.

        * tao/Invocation.cpp:
          The waiting strategy wants the complete Synch_Reply_Dispatcher,
          not just the reply_received flag.

        * tao/LF_Event.h:
        * tao/LF_Event.inl:
        * tao/LF_Event.cpp:
          New class to represent events that the Leader/Followers loop
          waits for.  Used as a base class for both TAO_Queued_Message and
          for TAO_Synch_Reply.

        * tao/LF_Event_Loop_Thread_Helper.h:
        * tao/LF_Event_Loop_Thread_Helper.inl:
        * tao/LF_Event_Loop_Thread_Helper.cpp:
          Move helper class to its own file, no sense in exposing it to
          everybody through the Leader_Follower.h file.

        * tao/Leader_Follower.h:
        * tao/Leader_Follower.i:
        * tao/Leader_Follower.cpp:
          Add free list for TAO_Follower, as well as allocation and
          deallocation methods.
          Move Leader/Followers main loop to this class.
          Move LF_Strategy and friends to their own files.

        * tao/ORB_Core.h:
        * tao/ORB_Core.i:
        * tao/ORB_Core.cpp:
          Removed the TSS Leader/Followers condition variable, the
          Leader/Followers free list implements the same optimization with
          less problems (i.e. without bug 296).

        * tao/Queued_Message.h:
        * tao/Queued_Message.cpp:
        * tao/Synch_Reply_Dispatcher.h:
        * tao/Synch_Reply_Dispatcher.cpp:
          This class derives from TAO_LF_Event now.  Any state or methods
          required to detect timeouts, closed connections or transmition
          errors are in the base class.

        * tao/Reply_Dispatcher.h:
        * tao/Asynch_Reply_Dispatcher.h:
        * tao/Asynch_Reply_Dispatcher.cpp:
          Remove the dispatcher_bound() calls, they are no longer required
          to match follower threads and their reply dispatchers, this is
          now done in the TAO_LF_Event::bind() method, called from
          TAO_Leader_Follower::wait_for_event()

        * tao/Transport.h:
        * tao/Transport.cpp:

        * tao/Transport_Mux_Strategy.h:
        * tao/Transport_Mux_Strategy.cpp:
        * tao/Muxed_TMS.cpp:
        * tao/Exclusive_TMS.cpp:
          Since there is no need to call dispatcher_bound() anymore the
          bind_dispatcher() methods were simplified.

        * tao/Wait_On_Leader_Follower.h:
        * tao/Wait_On_Leader_Follower.cpp:
        * tao/Wait_On_Reactor.h:
        * tao/Wait_On_Reactor.cpp:
        * tao/Wait_On_Read.h:
        * tao/Wait_On_Read.cpp:
        * tao/Wait_Strategy.h:
        * tao/Wait_Strategy.cpp:
          Use a TAO_Synch_Reply_Dispatcher to wait for a reply.  The hack
          using a reply_received flag + a cond.var. was too ugly, plus it
          was tightly coupling the Leader/Followers loop to the reply
          dispatching logic.

        * tao/default_resource.h:
        * tao/default_resource.cpp:
          Made Leader_Follower_Flushing_Strategy the default.

        * tao/orbconf.h:
        * tao/default_client.cpp:
          Made Muxed_TMS the default

        * tao/LF_Strategy.h:
        * tao/LF_Strategy.inl:
        * tao/LF_Strategy.cpp:
        * tao/LF_Strategy_Complete.h:
        * tao/LF_Strategy_Complete.inl:
        * tao/LF_Strategy_Complete.cpp:
          Move the LF_Strategy classes to their own files, no sense in
          exposing them to everybody through the Leader_Follower.h file.

        * tao/Follower_Auto_Ptr.h:
        * tao/Follower_Auto_Ptr.inl:
        * tao/Follower_Auto_Ptr.cpp:
          Helper class to automatically allocate and deallocate
          TAO_Follower objects from the Leader/Followers set.

        * tao/GIOP_Message_Base.cpp:
        * tao/GIOP_Message_Lite.cpp:
        * tao/Reactor_Registry.cpp:
          Must #include the "LF_Strategy.h" file explicitly.

        * tao/TAO.dsp:
        * tao/TAO_Static.dsp:
        * tao/Makefile:
        * tao/Makefile.bor:
        * tao/Strategies/TAO_Strategies.dsp:
        * tao/Strategies/TAO_Strategies_Static.dsp:
        * tao/Strategies/Makefile:
        * tao/Strategies/Makefile.bor:
          Add new files to the projects and Makefile


        * tao/Strategies/advanced_resource.cpp:
        * tao/Strategies/LF_Strategy_Null.h:
        * tao/Strategies/LF_Strategy_Null.inl:
        * tao/Strategies/LF_Strategy_Null.cpp:
          Move the Null Leader/Follower Strategy to the TAO_Strategies
          library, it was in TAO, but was only used here.

        * tao/RTPortableServer/TAO_RTPortableServer.dsp:
          Fixed missing libraries in link line.

        * tao/TAO.dsw:
          Add missing dependencies for RTPortableServer and RTCORBA


Wed Aug  1 13:15:10 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * TAO_IDL/be/be_visitor_operation/ami_exception_holder_operation_cs.cpp:
        * TAO_IDL/be/be_visitor_operation/exceptlist_cs.cpp:
        * TAO_IDL/be/be_visitor_operation/interceptors_cs.cpp:
        * TAO_IDL/be/be_visitor_operation/interceptors_exceptlist.cpp:
        * TAO_IDL/be/be_visitor_operation/interceptors_ss.cpp:
        * tao/Stub.h:
        * tao/Invocation.cpp:

          Modified the struct TAO_Exception_Data to contain the repository id
          instead of the type code, since the type code was used only to
          furnish the repo id string. The twoway invoke() method and generated
          code have been modified accordingly. Also, the interceptor method
          exceptions() now creates a simple array of type codes instead of
          an array of TAO_Exception_Data structs. If type codes are suppressed
          in generated code, the interceptor methods result() (which returns an
          Any) and exception() (from which the spec requires a list of type
          codes as the return value) will throw CORBA::NO_IMPLEMENT(). If only
          Anys are suppressed in generated code, only result() will throw the
          exception. Thanks to Russell Mora <rd.mora@econz.co.nz> for reporting
          the clash of type code suppression and exceptions, and to Ossama Othman
          <othman@cs.wustl.edu> and Carlos O'Ryan <coryan@ece.uci.edu> for
          their suggestions.

Wed Aug  1 11:32:26 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Strategies/UIOP_Connection_Handler.cpp:Fixed a compile error
          with g++.

Wed Aug  1 09:36:37 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * TAO_IDL/be/be_visitor_operation/operation_cs.cpp:

          Fixed an ACE_CHECK_RETURN in generated operation body where
          the return type is CORBA::LongLong or CORBA::LongDouble.

Wed Aug 1 8:38:06 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/orbconf.h: Removed the #define TAO_DOESNT_YADA_YADA as there
          is no use for it.

        * tao/Connection_Handler.h: Removed the fetch_handle () as there
          is no use for it.

        * tao/IIOP_Connection_Handler.h:
        * tao/IIOP_Connection_Handler.cpp:
        * tao/Strategies/DIOP_Connection_Handler.cpp:
        * tao/Strategies/DIOP_Connection_Handler.h:
        * tao/Strategies/SHMIOP_Connection_Handler.h:
        * tao/Strategies/SHMIOP_Connection_Handler.cpp:
        * tao/Strategies/UIOP_Connection_Handler.h:
        * tao/Strategies/UIOP_Connection_Handler.cpp:
        * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Connection_Handler.h:
        * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Connection_Handler.cpp: Removed
          the implementation of fetch_handle (). Thanks to Carlos, the
          foot-print police!!, for motivating this change.

Wed Aug 1 8:02:06 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/IIOP_Connection_Handler.h:
        * tao/IIOP_Connection_Handler.cpp: Removed the flag <resume_flag_>
          that was  introduced so that the input and output datapath would
          behave differently. The output data path would allow the reactor
          to resume the handler whereas the input data path would resume
          the handle. This was creating more confusion than
          necessary. Hence zapped the the flag and made the input and
          output data path consistent. Thanks to Carlos for motivating
          this stuff.

        * tao/Strategies/DIOP_Connection_Handler.cpp:
        * tao/Strategies/DIOP_Connection_Handler.h:
        * tao/Strategies/SHMIOP_Connection_Handler.h:
        * tao/Strategies/SHMIOP_Connection_Handler.cpp:
        * tao/Strategies/UIOP_Connection_Handler.h:
        * tao/Strategies/UIOP_Connection_Handler.cpp:
        * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Connection_Handler.h:
        * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Connection_Handler.cpp: Appled the
          same changes to the above mentioned protocols.

Tue Jul 31 12:53:06 2001  Carlos O'Ryan  <coryan@uci.edu>

        * tao/Transport.cpp (register_handler):
          Fixed race condition, the register_handler_i() method may use
          the connection_handler_, but we do not check if it is nil before
          calling.

Tue Jul 31 13:40:59 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tests/Oneways_Invoking_Twoways/client.dsp: Fixed a compile error
          in release builds.

Tue Jul 31 10:39:18 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * tests/IDL_Test/gperf.idl:

          New file in IDL_Test containing examples sent in by
          Karl Proese <karl.proese@mchp.siemens.de> and Vsevolod Novikov
          <novikov@df.nnov.rfnet.ru> that uncovered bugs in gperf.

        * tests/IDL_Test/idl_test.dsp:
        * tests/IDL_Test/Makefile:
        * tests/IDL_Test/Makefile.bor:

          Updated project and makefiles to include the new generated files.

Tue Jul 31 08:25:46 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * TAO_IDL/fe/idl.ll:
        * TAO_IDL/fe/idl.yy:
        * TAO_IDL/fe/lex.yy.cpp:
        * TAO_IDL/fe/lex.yy.cpp.diff:
        * TAO_IDL/fe/y.tab.cpp:
        * TAO_IDL/fe/y.tab.cpp.diff:
        * TAO_IDL/fe/y.tab.h:

          Removed OBV-related token IDL_INIT and replaced it with
          IDL_FACTORY, which is returned when the string 'factory' is seen
          in an IDL file. This string signals the declaration of an value
          type's initializing member function. Also changed the production
          rule 'init_decl' to be closer to the correct grammar for this
          type of function, although this feature is still completely
          unimplemented. And finally, modified two production rules to
          eliminate long-standing shift/reduce error messages when
          generating y.tab.* from idl.yy. Thanks to Russ Noseworthy
          <j.russell.noseworthy@objectsciences.com> for reporting that an
          'init' identifier string in an IDL file caused an error when
          compiled with the -Gv option (value types) enabled.

Tue Jul 31 06:58:59 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        This checkin is for fixing a race condition while trying to
        manipulate the number of upcalls. This was not a problem before
        575 fix, as the manipulation was done when there was an implicit
        synchronisation in the TP Reactor. As the implicit synchronisation
        has been broken, we had a race condition. The surpsising element
        was the fact that it took sometime to figure out this race
        condition. We have now added a lock that will be held by the
        thread  before the variable is manipulated.

        * tao/Connection_Handler.cpp:
        * tao/Connection_Handler.h:
        * tao/ Connection_Handler.inl: Added a lock to the class. Also
          added three methods, incr_pending_upcalls (),
          decr_pending_upcalls () and pending_upcalls (). The first two
          does the manipulation of the pending_upcalls_ variable after
          holding the lock.

        * tao/IIOP_Connection_Handler.cpp:
        * tao/IIOP_Connection_Handler.h: Removed he peding_upcalls_
          variable and called the incr_pending_upcalls () and
          decr_pending_upcalls () to achieve what needs to be done.

        * tao/Strategies/DIOP_Connection_Handler.cpp
        * tao/Strategies/DIOP_Connection_Handler.h
        * tao/Strategies/SHMIOP_Connection_Handler.cpp
        * tao/Strategies/SHMIOP_Connection_Handler.h
        * tao/Strategies/UIOP_Connection_Handler.cpp
        * tao/Strategies/UIOP_Connection_Handler.h:
        * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Connection_Handler.h:
        * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Connection_Handler.cpp:
          Replicated the changes from IIOP to the above protocols.

Sun Jul 29 19:31:34 2001  Douglas C. Schmidt  <schmidt@tango.doc.wustl.edu>

        * tao/RTCORBA/RT_Mutex.cpp (try_lock): Since we fixed the timed
          ACE_OS::mutex_lock() to set errno to ETIME we can cleanup the
          code here to remove the checks for errno == EBUSY and errno ==
          ETIMEDOUT.

Sun Jul 29 20:00:00 2001  Craig Rodrigues  <crodrigu@bbn.com>

        * tao/PortableServer/ORB_Manager.h:

          Improved comments and converted to doxygen format.

Sun Jul 29 10:15:13 2001  Ossama Othman  <ossama@uci.edu>

        * tao/TAO_Internal.cpp (open_services_i):

          Reverted my change that prevented the default resource factory
          from being inserted into the Service Repository.  The resource
          factory must be inserted into the Service Repository so that
          Service Configurator directives such as 'static Resource_Factory
          "-ORBResources global"' actually work properly.  An alternative
          solution for the dynamic loading problem related the default
          resource factory is necessary.

Sun Jul 29 00:59:28 2001  Ossama Othman  <ossama@uci.edu>

        * docs/Options.html:

          Updated documentation for the "-ORBSkipServiceConfigOpen" ORB
          option.  It is no longer necessary the Service Configurator is
          now reentrant and thread-safe.  This option is deprecated, and
          will be removed in releases of TAO in the near future.

Sun Jul 29 00:22:30 2001  Ossama Othman  <ossama@uci.edu>

        * examples/Simple/time-date/svc.conf:

          The support for nested Service Configurator directive processing
          that was just added to ACE made it possible to greatly simply
          this `svc.conf' file.  It is no longer necessary to pre-load
          services that are by default loaded by the ORB.  It is also no
          longer necessary to use the "-ORBSkipServiceConfigOpen" ORB
          option since the Service Configurator is now
          reentrant/thread-safe.

          Instead of the nine Service Configurator directives that were
          listed in this file, only three of them are now necessary.

Sat Jul 28 23:29:21 2001  Ossama Othman  <ossama@uci.edu>

        * tao/TAO_Internal.cpp (open_services_i):

          We cannot insert the default resource factory into the Service
          Repository before the ORB is created since it will be finalized
          before the ORB is finalized.  The ORB requires that a resource
          factory exist in order to reclaim the reactor during
          finalization.

          This was only a problem when the ORB was dynamically
          loaded/unloaded.

Wed Jul 25 23:48:58 2001  Douglas C. Schmidt  <schmidt@ace.cs.wustl.edu>

        * tests/Blocking_Sync_None/client.cpp (main): Changed an
          error messages so that the "Right Thing[TM]" will happen when
          this test fails, i.e., the build system will automagically
          detect it.  Thanks to Johnny Willemsen for reportng this.

Fri Jul 27 21:48:12 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * performance-tests/Cubit/TAO/IDL_Cubit/run_test.pl: Removed all
          the references to the GIOP lite protocol from this script.

        * performance-tests/Cubit/TAO/IDL_Cubit/run_test_lite.pl: We dont
          seem to be having any sort of tests that are run for GIOP
          Lite. Now we will run this in our nightly builds. This test uses
          IIOP_Lite & UIOP_Lite to run the IDL_Cubit test. Having a
          protocol that we are not sure whether it works may be a bad
          idea.

Fri Jul 27 21:38:58 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/GIOP_Message_Base.cpp: Added a TAO_debug_level guard around
          the places where dump_msg () is called. Though dump_msg ()
          prints out information only if the TAO_debug_level is set, this
          extra guard will prevent us from calling this method when we are
          trying to do performance measurements.

        * tao/GIOP_Message_Lite.cpp: For some reason this class never had
          a hexdump. Added the hexdump in dump_msg ().

Fri Jul 27 16:33:24 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Strategies/DIOP_Connection_Handler.cpp:
        * tao/Strategies/UIOP_Connection_Handler.cpp:
        * tao/Strategies/SHMIOP_Connection_Handler.cpp: Enabled GIOP lite
          flag to pass through to their transport object.

Fri Jul 27 16:32:46 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/IIOP_Connection_Handler.cpp: Enabled GIOP lite flag to pass
          through to their transport object.

        * tao/GIOP_Message_Lite.cpp: Transfered the reply that has been
          received to the another CDR which is used to dispatch the
          reply. This was actually fixed in my branch. Looks like this got
          missed when the branch was merged to the main trunk.  Added some
          cosmetic fixes to the debugging output. Thanks to Paul Calabrese
          for alerting me about this miss.

Fri Jul 27 14:43:46 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tests/BiDirectional_NestedUpcall/svc.conf: Removed from the
          repository.  It is no longer needed as this test will work with
          a TP Reactor (the default one).

Fri Jul 27 12:58:51 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Transport.cpp: Removed a debug statement.

Fri Jul 27 12:25:49 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Transport.cpp: Reverted the change made in this Wed Jul 25
          22:10:21 2001  Balachandran Natarajan  <bala@cs.wustl.edu>. We
          need to resume the handle as soon the reply is ready for
          dispatching.  The situation where it can create problems can be
          easily seen in $TAO_ROOT/tests/LongUpcalls/run_ami_test.pl.

          Further, the fix that was done earlier was to resume the handle
          after dispatching the reply. It was thought that it would fix a
          race condition. I had a long discussion with Irfan yesterday,
          and looks like the possibility of a race condition is not
          there at all.

Fri Jul 27 11:09:25 2001  Chad Elliott  <elliott_c@ociweb.com>

        * tests/RTCORBA/Banded_Connections/bands.hpux:

          Added to fix the runtime problem on HP-UX.

        * tests/RTCORBA/Banded_Connections/run_test.pl:
        * tests/RTCORBA/Client_Propagated/client.cpp:
        * tests/RTCORBA/Client_Propagated/server.cpp:
        * tests/RTCORBA/MT_Client_Protocol_Priority/run_test.pl:
        * tests/RTCORBA/Server_Declared/run_test.pl:

          Corrections to fix some of the RTCORBA runtime test problems for
          HP-UX.  These tests still have some problems (especially with
          SHMIOP).

Fri Jul 27 10:50:42 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * tests/Smart_Proxies/Benchmark/Makefile:
        * tests/Smart_Proxies/Policy/Makefile:

          Replaced VLDLIBS with TAO_SRVR_LIBS for the server
          target in each Makefile, in order to pull in
          TAO_PortableServer. Thanks to Ekkehard Hoffmann
          <ehoffman@fzi.de> for pointing out the link errors.

Thu Jul 26 22:41:00 2001  Craig Rodrigues  <crodrigu@bbn.com>

       * orbsvcs/orbsvcs/AV/AVStreams_i.h:

         Make TAO_FlowConnection and TAO_MMDevice virtually
         inherit TAO_PropertySet.  Thanks to Rob Ruff <rruff@scires.com>
         for pointing this out.

Thu Jul 26 21:48:00 2001  Craig Rodrigues  <crodrigu@bbn.com>

        * orbsvcs/tests/AVStreams/mpeg/*: REMOVED

          Example uses archaic version of AVStreams and does not
          work well.

        * orbsvcs/tests/AVStreams/mpeg/README.uav: (added)

Thu Jul 26 17:12:00 2001  Ossama Othman  <ossama@uci.edu>

        * tao/default_resource.cpp (get_parser_names):

          Fixed problem where the FILE and DLL parser Service Objects were
          not inserted into the parser registry.  This problem surfaced
          when attempting to dynamically load the ORB, and manifested
          itself as a CORBA::INV_OBJREF exception when attempting to
          destringify an IOR such as "file://foo.ior".

Thu Jul 26 09:44:00 2001  Craig Rodrigues  <crodrigu@bbn.com>

        * orbsvcs/orbsvcs/AV/QoS_UDP.cpp:

          Hide more debugging messages behind: if( TAO_debug_level > 0)

Thu Jul 26 07:37:29 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tests/RTCORBA/Server_Protocol/server.cpp:
        * tests/RTCORBA/Client_Propagated/server.cpp:
        * tests/RTCORBA/Thread_Pool/server.cpp: Fixed warnings in g++.

Wed Jul 25 23:37:00 2001  Craig Rodrigues  <crodrigu@bbn.com>

        * orbsvcs/tests/AVStreams/Full_Profile/Makefile:
        * orbsvcs/tests/AVStreams/Latency/Makefile:
        * orbsvcs/tests/AVStreams/Modify_QoS/Makefile:
        * orbsvcs/tests/AVStreams/Multicast/Makefile:
        * orbsvcs/tests/AVStreams/Multicast_Full_Profile/Makefile:
        * orbsvcs/tests/AVStreams/Pluggable/Makefile:
        * orbsvcs/tests/AVStreams/Simple_Three_Stage/Makefile:
        * orbsvcs/tests/AVStreams/Simple_Two_Stage/Makefile:
        * orbsvcs/orbsvcs/Makefile.av:

          Correct link flags so that ACE_QoS is linked in when
          rapi=1 is specified in platform_macros.GNU.

Wed Jul 25 22:45:10 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Strategies/SHMIOP_Transport.cpp:
        * tao/Strategies/DIOP_Transport.cpp: Fixed a compile error that
          came up from my previous change.

Wed Jul 25 22:39:32 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tests/RTCORBA/Server_Protocol/server.cpp:
        * tests/RTCORBA/Client_Propagated/server.cpp:
        * tests/RTCORBA/Thread_Pool/server.cpp: Added checks & debugging
          statments for a null RootPOA. thanks to Johnny Willemsen for
          pointing this out. Did some minor cosmetic fixes.

Wed Jul 25 22:10:21 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Transport.cpp:
        * tao/Transport.h: Fixed a subtle problem that seems to have lead
          to the  Muxing tests failing randomly. The problem is something
          like this
          - multiple client threads can try to share a connection
          - because of the above, more than one message are sent on the
            same connection
          - if the server is multi-threaded, the messages can be processed
            concurrently
          - there may be a possibility of more than two replies coming on
            the same connection.
          - one of the client threads can pick up both the replies
          - one of the replies would be queued up and the first one can be
            its own
          - after queueing up the second it would wake up another thread
          - if the woken up thread does not own the reply, it could just
            take the reply and try to transfer ownership to the right
            thread.
          - before the second thread transfers the reply, teh second
            thread would have resumed the handler and because of which one
            of the threads would have gone into the reactor from the LF.
          - at exactly the same instant the seccond thread will have
            difficulty in waking up the thread on select () is it is the
            owner.
          Fixed this problem by not resuming the handle till we dispatch
          the reply. We dont buy anything by resuming the handle before
          dispatching the reply because, the dispatching will not be
          unbounded. The forces that apply to the server thread, which
          resumes the handle before making an upcall does not apply to the
          client threads that reads and processes replies. This fix should
          ideally fix the Muxing test failure on different paltforms. If
          it doesnt, it will atleast prevent the race condition outlined
          above :-)

Wed Jul 25 20:33:21 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * examples/Simple/time-date/Makefile.bor:
        * examples/Simple/time-date/server.bor:
        * examples/Simple/time-date/time_date.bor: Fixed Borland builds
          for this example. This commit is for Johnny Willemsen who is
          away from his work.

Wed Jul 25 12:50:00 2001  Michael Kircher <Michael.Kircher@mchp.siemens.de>

        * tao/Strategies/DIOP_Factory.cpp:

          Changed the return value of requires_explicit_endpoint () to 0
          and documented that this return code is not reflecting that
          the endpoints are not cleaned-up but that we disable it by default
          because DIOP is only suitable for certain use cases, e.g. it only
          supports one-ways.

Wed Jul 25 08:41:39 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * TAO_IDL/be/be_visitor_interface/interface_cs.cpp:

          Fixed formatting in generation of _unchecked_narrow().

Tue Jul 25 01:00:00 2001  Craig Rodrigues <crodrigu@bbn.com>

        * orbsvcs/tests/AVStreams/Multicast/run_test.pl:
        * orbsvcs/tests/AVVStreams/Asynch_Three_Stage/run_test.pl:
        * orbsvcs/tests/AVStreams/Asynch_Three_Stage/input:

         Increase the times that the perl scripts expecting the
         the CORBA processes to run for.  Decrease the size
         of the Asynch_Three_Stage input file by 4000 lines.

Tue Jul 24 18:16:13 2001  Krishnakumar B  <kitty@cs.wustl.edu>

        * tests/Exposed_Policies/Policy_Verifier.cpp (init):
        * tests/Exposed_Policies/RT_Properties.cpp:

          Changed handling of the default argument values. Added options
          "-BaseObjectIOR" and "-OverriddenIOR" to enable waiting on the IOR
          file instead of sleeping in the run_test.pl

        * tests/Exposed_Policies/run_test.pl:

          Fixed priorities to handle Tru64. Added the new options
          mentioned above to specify the IOR.

        * tests/Exposed_Policies/POA.cfg:
        * tests/Exposed_Policies/Object.cfg:

          Since the IOR files are passed from the command-line, remove
          them from here.

        * tests/Exposed_Policies/POA.cfg.tru64: *
          tests/Exposed_Policies/Object.cfg.tru64:

          Added new files with proper priority values for Tru64.

          OCI folks, you want to change similarly for HP-UX.


Tue Jul 24 11:12:25 2001  Ossama Othman  <ossama@uci.edu>

        * tao/IIOP_Acceptor.cpp (create_shared_profile,
          create_new_profile):
        * tao/Strategies/DIOP_Acceptor.cpp
          (create_shared_profile, create_new_profile):
        * tao/Strategies/SHMIOP_Acceptor.cpp (create_profile):
        * tao/Strategies/UIOP_Acceptor.cpp (create_profile):

          Do not add any tagged components to the profile if an IIOP 1.0
          endpoint/profile is being created.  Tagged components were
          introduced in IIOP 1.1.  The same convention is adopted for the
          other pluggable protocols distributed with TAO (UIOP, SHMIOP,
          and DIOP).  SSLIOP already does this check since it requires
          tagged components in order to convey security association
          information to the client.  These changes address
          interoperability issues.

Tue Jul 24 12:33:17 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tests/Big_Reply/Client_Task.cpp (validate_connection): Some
          cosmetic changes.

Tue Jul 24 08:38:14 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * orbsvcs/examples/Security/SecurityLevel1/SLevel1_Test.idl:

          Removed 'void' parameter from two operations. Thanks to
          Ugendreshwar Kudupudi <ugenderk@rediffmail.com>
          for reporting the bug.

Mon Jul 23 22:31:18 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tests/Big_Reply/Client_Task.cpp: Added some debufg
          statments. Further fixed a small goof up. A method to
          validate connections was added but never called :(. Fixed that
          by calling validate_connection () before an actual call to the
          server object.

Mon Jul 23 22:24:05 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * examples/Simple/time-date/svc.conf: Removed the entry for
          TAO_RT_Protocol_Hooks. Looks like they have taken some other
          form after the RTCORBA subsetting effort. The test were failing
          in the daily builds but never came up on the scoreboard.

Mon Jul 23 21:40:38 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/GIOP_Message_Lite.cpp: Fixed a warning in Win32 builds.

Mon Jul 23 1:44:23 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/TAO_Static.dsp: Added GIOP_Lite files to the builds.

Mon Jul 23 11:44:30 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        Support for GIOP Lite and GIOP Fragments are being added in this
        checkin. Merged from the branch giop_lite_fragment.

        * tao/Makefile:
        * tao/Makefile.bor: Added GIOP Lite  files to the list of files.

Sat Jul 14 16:42:07 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Strategies/DIOP_Transport.cpp:
        * tao/Strategies/SHMIOP_Transport.cpp:
        * tao/Strategies/UIOP_Transport.cpp: Added support for GIOP lite.

Fri Jul 13 16:54:07 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Pluggable_Messaging.h:
        * tao/GIOP_Message_Base.h:
        * tao/GIOP_Message_Base.cpp:
        * tao/Transport.cpp:
        * tao/Transport.h:
        * tao/Transport.inl: Added support for GIOP fragments. The GIOP
          fragmentation supportseems to have some copying and allocation
          overhead. Need to look into this when it gets
          important. Further, the fragmentation support hasnt been
          tested at all as we have no way of testing it.

        * tao/IIOP_Transport.cpp: Enabled GIOP lite support
        * tao/TAO.dsp: Added GIOP lite files back to the builds.

        * tao/Connection_Handler.h:
        * tao/orbconf.h: Moved some of the #defines from
          Connection_Handler.h to orbconf.h.

Mon Jul  9 10:23:07 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/GIOP_Message_Generator_Parser_12.cpp: The long talked about
          alignment for the LocateReply messages have been removed. We
          dont align the messages on an 8 byte boundary as described by
          the CORBA 2.4 spec. This is one of the urgent resolutions
          adopted by the OMG.

Mon Jul  9 09:40:07 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/GIOP_Message_Lite.h:
        * tao/GIOP_Message_Lite.cpp: Added support for GIOP lite. The
          implementation now  is similar to the regular GIOP.

        * tao/GIOP_Message_Base.cpp: Used the payload_size () in
          GIOP_Message_State  for claculating the payload. We were using
          the message_size () and then subtracting the length of the GIOP
          header in consolidate_node (). This looked cumbersome.



Mon Jul 23 07:46:30 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/orbconf.h: Enabled DIOP for minimum CORBA.

Sun Jul 22 21:53:23 2001  Irfan Pyarali  <irfan@cs.wustl.edu>

        * $ACE_ROOT/auto_run_tests.lst: Disabled Exposed_Policies test for
          single threaded configurations and Linux.  This test requires
          multiple threads and support for multiple native priorities.

        * tests/Exposed_Policies/Policy_Tester.cpp (Policy_Tester): Fixed
          several things:

          - No need for the Policy_Tester::~Policy_Tester to call
          shutdown().  This is already done by the servant.

          - No need for TRY/CATCH blocks in each function.  This
          unnecessarily stops any exceptions from propagating to higher
          layers.

          - Fixed return values and added error checking.

          - No need to call destroy() on the POAs. ORB::shutdown is
          enough.

          - Removed unnecessary default values from the function
          prototypes.

        * tests/Exposed_Policies/run_test.pl:
        * tests/Exposed_Policies/POA.cfg:
        * tests/Exposed_Policies/Object.cfg:

          Updated config files so that the priorities would be ok for
          Win32.

        * tests/Exposed_Policies/server.cpp: Improved error checking.

        * tests/Exposed_Policies/server.conf: Removed Thread Pool Reactor
          directive from the service config file.

Sun Jul 22 16:44:16 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tests/Big_Reply/Client_Task.cpp:
        * tests/Big_Reply/Client_Task.h: Fixed a compile error in no
          exception builds.

Sun Jul 22 13:36:21 2001  Craig Rodrigues  <crodrigu@bbn.com>

        * orbsvcs/orbsvcs/AV/RTP.cpp:
          If a TAO_AV_frame_info is not specified in the
          TAO_AV_RTP_Object::sendframe() method, insert a
          timestamp into the RTP header based on the current system
          time.  The timestamp will be in milliseconds.
          Thanks to Rob Ruff <rruff@scires.com> for pointing this out.

Sun Jul 20 12:30:00 2001  Michael Kircher <Michael.Kircher@mchp.siemens.de>

        * tao/Asynch_Invocation.cpp:
        * tao/Asynch_Invocation.i:

          Changed the code which sets up a reply dispatcher for
          AMI calls to only set up one, if a non-nil reply handler
          got registerd. This change is related to my change on
          Fri Jul 20 08:10:00 2001. It assumes that the ORB
          will drop replies to which no reply dispatcher is registered.

Sun Jul 22 09:43:09 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tests/Big_Reply/Test.idl: Added a no-op ping () operation which
          is used to setup a connection properly.

        * tests/Big_Reply/Client_Task.cpp:
        * tests/Big_Reply/Client_Task.h: Called the ping () method before
          getting the replies.

        * tests/Big_Reply/Big_Reply_i.h:
        * tests/Big_Reply/Big_Reply_i.cpp: Implementation for the ping ()
          method.

Sun Jul 22 09:27:30 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tests/Oneway_Buffering/Test.idl: Changed the operation
          request_received () in Oneway_Buffering_Admin interface to a
          oneway. The reason for doing this has been well documented in
          bug #982.

Fri Jul 20 23:58:12 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Transport.cpp: Fixed a subtle bug. The problem was when
          reading  a big message. We read part of the message into a stack
          allocated buffer. We then grow the buffer to read the rest of
          the message. This was working fine. But in the read after
          grwoing the  buffer, if we get an error we were just returning
          to the reactor. This was right. But we have to queue up the
          message before returning to the reactor incase we get a
          EWOULDBLOCK.

Fri Jul 20 12:58:12 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/IIOP_Connection_Handler.h:
        * tao/IIOP_Connection_Handler.cpp: Removed the resumption of the
          handler in the  handle_output () call. Looks like we dont win
          anything by resuming the handler at this level. We need to get
          back to this after 1.2 if it is required.

        * tao/Connection_Handler.h: Removed some of the #defines
        * tao/orbconf.h: Moved the #defines here from
          Connection_handler.h.

        * tao/Strategies/DIOP_Connection_Handler.h:
        * tao/Strategies/DIOP_Connection_Handler.cpp:
        * tao/Strategies/UIOP_Connection_Handler.h:
        * tao/Strategies/UIOP_Connection_Handler.cpp:
        * tao/Strategies/SHMIOP_Connection_Handler.h:
        * tao/Strategies/SHMIOP_Connection_Handler.cpp:
        * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Connection_Handler.h:
        * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Connection_Handler.cpp: Replicated
          IIOP changes to these protocols.


Fri Jul 20 09:25:00 2001  Craig Rodrigues  <crodrigu@bbn.com>

        * tests/ior_corbaloc/run_test.pl:
        * tests/InterOp-Naming/run_test.pl:

          Do not include ACEutils Perl module in scripts which use
          PerlACE::Run_Test module.  ACEutils prevents
          PerlACE::Run_Test from seeing the -ExeSubDir argument.
          This causes test failures on certain platforms.
          Thanks to Darrell Brunsch for explaining this and pointing
          this out.

Tue Jul 20 08:27:31 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * orbsvcs/tests/AVStreams/Component_Switching/receiver.dsp: Fixed
          a typo in the  library names.

Fri Jul 20 08:10:00 2001  Michael Kircher <Michael.Kircher@mchp.siemens.de>

        * tao/ORB_Core.cpp:
        * tao/Leader_Follower.i:
        * tao/Strategies/DIOP_Acceptor.cpp:
        * tao/Strategies/DIOP_Connection_Handler.cpp:

          Removed old, meanwhile unrelevant comments from me.

        * tao/Asynch_Reply_Dispatcher.cpp:

          Added a check for a nil reply handler. In the case of a nil reply
          handler no response is dispatched. Thanks to Andreas Geisler
          <Andreas.Geisler@erl9.siemens.de> for pointing this out.

        * tests/AMI/simple_client.cpp:

          Added a test case for the above change.

        * tests/AMI/client.cpp:

          Did a cosmetic change.

Thu Jul 19 18:48:09 2001  Irfan Pyarali  <irfan@cs.wustl.edu>

        * tao/ORB.h (run/work_pending/perform_work): Updated documentation
          wrt timeouts.  Thanks to Jean-Christophe Dubois <jcd@one.com>
          for suggesting this.

Thu Jul 19 10:37:55 2001  Paul Calabrese <calabrese_p@ociweb.com>

  * docs/Options.html:

          Document the differences between the default and
          advanced resource factories.

Tue Jul 19 09:30:31 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Transport.cpp (make_queued_data): Created a data block for
          the size that is required instead of the size of the incoming
          data block. This should fix the lingering problems with
          Single_Read tests in our daily builds.

        * tao/GIOP_Message_Base.cpp: Added some comments.

Tue Jul 19 09:34:00 2001  Craig Rodrigues  <crodrigu@bbn.com>

        * orbsvcs/tests/AVStreams/Component_Switching/sender.bor:
        * orbsvcs/tests/AVStreams/Component_Switching/receiver.bor:
        * orbsvcs/tests/AVStreams/Component_Switching/distributer.bor:
          Fixed the Borland makefiles files to link in the strategies library.

Tue Jul 19 07:56:31 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * orbsvcs/tests/AVStreams/Component_Switching/sender.dsp:
        * orbsvcs/tests/AVStreams/Component_Switching/receiver.dsp:
        * orbsvcs/tests/AVStreams/Component_Switching/distributer.dsp:
          Fixed the dsp files to link in the strategies library.

Tue Jul 19 12:49:31 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Sequence.i (replace):
        * tao/Sequence.cpp: Set the read and write pointers properly for
          copied message blocks at the positions.

Tue Jul 19 12:27:31 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Transport.cpp (make_queued_data): Fixed a typo and added a
          new line for a DEBUG statement.

Tue Jul 18 11:25:31 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Transport.h:
        * tao/Transport.cpp: Refactored some code in to a new method
          called  get_queued_data ().

Wed Jul 18 23:58:00 2001  Craig Rodrigues <crodrigu@bbn.com>

        * orbsvcs/tests/AVStreams/Component_Switching/sender.cpp:
        * orbsvcs/tests/AVStreams/Component_Switching/receiver.cpp:
        * orbsvcs/tests/AVStreams/Component_Switching/distributer.cpp:

          Add #include "tao/Strategies/advanced_resource.h"
          so that the ace_static_svc_TAO_Advanced_Resource_Factory
          symbol is defined in each binary, and the TAO_Strategies
          library is actually linked in during static builds.
          Fixes failures of this test with static builds.

        * orbsvcs/tests/AVStreams/Component_Switching/components_svc.conf:
          Fix syntax so Service Configurator file can be used
          for static and dynamic builds.

Wed Jul 18 19:22:48 2001  Irfan Pyarali  <irfan@cs.wustl.edu>

        * tao/Leader_Follower.i (set/reset_client_leader_thread): Fixed a
          subtle bug in the Leader Follower code: <client_leader_thread_>
          was marked 1 in set_client_leader_thread() and 0 in
          reset_client_leader_thread().  This was a problem when the
          client thread was made leader thread multiple times in nested
          upcall situations.  Changing the code to increment
          <client_leader_thread_> in set_client_leader_thread() and
          decrement in reset_client_leader_thread() gave us an accurate
          count of how many times the client was made the leader.

        * tao/Asynch_Reply_Dispatcher.cpp: Removed unnecessary include
          file "Leader_Followers.h".

        * tests/NestedUpcall/Simple/run_test.pl: Changed MT configuration
          to use parameters with which the test was previously failing.

Wed Jul 18 17:44:56 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * TAO_IDL/driver/drv_preproc.cpp (DRV_pre_proc):
        * orbsvcs/IFR_Service/drv_preproc_ifr.cpp (DRV_pre_proc):

          Added one more flag set in the ACE_Log_Msg instance to
          ensure that the output of the preprocessor dump (-E command
          line option) goes to stdout.

Tue Jul 18 09:11:13 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Transport.cpp: Made sure that the timeout values are passed
          to  the recv  () calls in the handle_input_i (). The
          thread-per-connection uses the timeout value. Thanks to Mike
          Pyle <mike.pyle@burning-glass.com> for influencing this change.

Wed Jul 18 08:55:29 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * orbsvcs/IFR_Service/drv_preproc_ifr.cpp (DRV_pre_proc):

          Dump of preprocessed IDL file (from -E command line option)
          now goes to stdout instead of stderr. Similar to change in
          TAO IDL compiler in yesterday's entry.

Tue Jul 18 07:58:30 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Transport.cpp (consolidate_message): Fixed a subtle that
          hardly showed up.  The problem stemmed from the fact we were
          trying to queue up messages if we had a partial read. The
          Message blocks were duplicated but not the underlying data
          blocks. Fixed them by doing a proper copying onto the message
          queue.

        * tao/Sequence.i:
        * tao/Sequence.cpp: Used the self_flags () instead of flags () to
          check on the origin of the data block.

Tue Jul 17 21:13:30 2001  Irfan Pyarali  <irfan@cs.wustl.edu>

        * tests/RTCORBA/Server_Protocol/run_test.pl: Updated and fixed
          run_test.pl and add new file server_uiop.conf.

Tue Jul 17 16:24:00 2001  Craig Rodrigues  <crodrigu@bbn.com>

        * orbsvcs/tests/AVStreams/Component_Switching/run_test.pl:
        * orbsvcs/tests/AVStreams/Component_Switching/svc.conf: (removed)
        * orbsvcs/tests/AVStreams/Component_Switching/components_svc.conf: (added)
          Renamed svc.conf to components_svc.conf to fix problems
          in statically compiled builds.

Tue Jul 17 14:09:16 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tests/Single_Read/client.cpp:
        * tests/Single_Read/test_i.cpp: Added some debug statements. This
          test seems to be failing only in one of the Full builds. Not
          sure what could be the problem. The test runs to perfection but
          the srever doesn't shutdown. I havent been able to reproduce the
          error. These debug statements should give me sufficient clue to
          what is happening.

Tue Jul 17 13:48:00 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Strategies/UIOP_Transport.cpp: Fixed a warning.

Tue Jul 17 13:44:49 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * performance-tests/POA/Demux/Makefile: Updated dependencies. This
          should fix the compile error with FORTE.

Tue Jul 17 11:54:07 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * TAO_IDL/driver/drv_preproc.cpp:

          Modified preprocessor dump to go to stdout instead of stderr. This
          feature was requested by Alex Hornby <alex@anvil.co.uk>.

Mon Jul 16 22:24:17 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * examples/PluggableUDP/tests/SimplePerformance/run_test.pl:
        * examples/PluggableUDP/tests/Basic/run_test.pl: Fixed the run
          test scripts to make sure that a DIOP endpoint is specified when
          the server is invoked. This would fix the errors seen in our
          builds.

        * examples/PluggableUDP/tests/Basicsvc.conf: Just removed the file
          from the repository. Looks like it was not needed and confused
          the client too much leading to SEGV's.

Mon Jul 16 22:03:13 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/IIOP_Connection_Handler.cpp:
        * tao/Strategies/DIOP_Connection_Handler.cpp:
        * tao/Strategies/UIOP_Connection_Handler.cpp:
        * tao/Strategies/SHMIOP_Connection_Handler.cpp:
        * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Connection_Handler.cpp: Fixed,
          what looked like a subtle race condition. If a thread that calls
          handle_input () gets a -1 on read, we return -1 to the
          reactor. This makes the reactor close the handler and remove
          that from its internal map. If we resume the handler for such
          cases, before returning from handle_input (), looked like one
          more thread was woken up by the reactor to read from the
          handle. When this occurs on another thread, the first thread
          went about doing its task of closing the handle and removing the
          handler from its internal map. Bad things started happening.

Mon Jul 16 17:23:57 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * tao/IORManipulation/TAO_IORManip.dsp:

          Further fixes to the MFC versions of this project.

Mon Jul 16 14:25:23 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/GIOP_Message_State.cpp (read_ulong): Looks like some of the
          compilers  have problems when they try to dereference a pointer
          got out of reinterpret_cast'ing another pointer.  The problem
          showed up on solaris builds. The fix that has been applied is to
          make a local copy on the stack of the data that is needed before
          calling a reinterpret_cast on it.

Mon Jul 16 11:44:30 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tests/Big_Reply/server.cpp: The reply sent was too huge. The bug
          #957 started showing up. Reduced the size of the reply to
          approximately 4 MB.

Mon Jul 16 11:36:00 2001  Frank Hunleth  <fhunleth@cs.wustl.edu>

        * tao/Strategies/DIOP_Factory.cpp (requires_explicit_endpoint):

          Make DIOP require an explicit endpoint to prevent a DIOP
          endpoint from being automatically created when the Strategies
          library is linked.

Mon Jul 16 11:20:40 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * tao/DyanmicInterface/TAO_DynamicInterface.dsp:
        * tao/IORTable/TAO_IORTable.dsp:
        * tao/IORManipulation/TAO_IORManip.dsp:
        * orbsvcs/orbsvcs/LoadBalancing.dsp:
        * orbsvcs/orbsvcs/CosNotification.dsp:

          Fixed MFC settings for these projects. Thanks to
          truename <shi_bing@netease.com> for reporting the
          problems.

Mon Jul 16 9:17:43 2001  Christopher Kohlhoff <chris@kohlhoff.com>

        * performance-tests/Throughput/client.bor:
        * performance-tests/Throughput/server.bor:
          Corrected definitions of CFLAGS and LIBFILES to fix Borland
          compile error.

Mon Jul 16 07:49:22 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * performance-tests/Cubit/TAO/IDL_Cubit/collocation_test.dsp:
        * performance-tests/RTCorba/Multiple_Endpoints/Orb_Per_Priority/Server.dsp:
        * performance-tests/RTCorba/Multiple_Endpoints/Orb_Per_Priority/Client.dsp:
          These dsp files needed the strategies library in their release
          configuration.  Added them to fix Win32 builds. I thought I had
          made the changes during my last checkin.  Apparently I had not :(.

Mon Jul 16 07:34:22 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * docs/tutorials/Quoter/Event_Service/Makefile: Updated
          dependencies.

Sun Jul 15 9:59:24 2001  Balachandran Natarajan <bala@cs.wustl.edu>

        * tests/Param_Test/anyop.dsp (InputPath): Set the dependecy for
          the generated code to the IDL compiler. Missed out when
          generated this file.

Sun Jul 15 9:33:34 2001  Balachandran Natarajan <bala@cs.wustl.edu>

        * tao/Sequence.cpp:
        * tao/Sequence.i: A fix at the places where
          ACE_Message_Block::duplicate () is called.  The duplicate ()
          call just allocated memory for a new Message block and then
          duplicated the exisiting data block -- which just increments the
          reference count of the data block. This could be bad for cases
          where the incoming message is on a data block on stack. By
          incrementing the reference count we get nothing for such
          cases. The fix that has been put in does the following

          - checks whether the data block is on stack, if so does a deep
            copy before calling duplicate on the message block.
          - if the data block is already on the heap just calls duplicate
            () on the message block.

        * orbsvcs/orbsvcs/Event/EC_Gateway_UDP.cpp: Added a comment where
          duplicate () is used. But the code that uses duplicate () has
          been commented out .

Sat Jul 14 20:18:36 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * docs/tutorials/Quoter/Event_Service/Makefile: Updated
          dependencies.

Sat Jul 14 18:59:48 2001  Balachandran Natarajan <bala@cs.wustl.edu>

        * performance-tests/Cubit/TAO/IDL_Cubit/collocation_test.dsp:
        * performance-tests/RTCorba/Multiple_Endpoints/Orb_Per_Priority/Server.dsp:
        * performance-tests/RTCorba/Multiple_Endpoints/Orb_Per_Priority/Client.dsp:
          These dsp files needed the strategies library. Added them to fix
          Win32 builds.

Sat Jul 14 17:44:28 2001  Balachandran Natarajan <bala@cs.wustl.edu>

        * tests/Param_Test/anyop.dsp (RSC): Looks like this has been
          broken for  a long time now. Fixed the dsp file so that we dont
          any errors during loading.

Sat Jul 14 00:33:28 2001  Paul Calabrese <calabrese_p@ociweb.com>

        * examples/PluggableUDP/tests/SimplePerformance/svc.conf:

          Switch this configuration file to use the
          Advanced_Resource_Factory.  This way the options
          actually do something.

Fri Jul 13 16:07:04 2001  Douglas C. Schmidt  <schmidt@tango.doc.wustl.edu>

        * orbsvcs/ImplRepo_Service/ImplRepo_i.cpp: Make sure
          that this->ior_multicast is non-NULL in ~ImplRepo_i() before
          using it to lookup in the reactor.  Thanks to Victor Chernenko
          <v_chernenko@hotmail.com> for reporting this.

Thu Jul 13 14:30:16 2001  Paul Calabrese <calabrese_p@ociweb.com>

        * examples/AMI/FL_Callback/AMI_Peer.dsp:
        * examples/AMI/FL_Callback/AMI_Progress.dsp:
        * orbsvcs/tests/AVStreams/Latency/control.dsp:
        * orbsvcs/tests/AVStreams/Latency/ping.dsp:
        * orbsvcs/tests/AVStreams/Latency/pong.dsp:
        * performance-tests/Callback/client.dsp:
        * performance-tests/Callback/server.dsp:
        * performance-tests/Cubit/TAO/MT_Cubit/client.dsp:
        * performance-tests/Cubit/TAO/MT_Cubit/server.dsp:
        * tests/Collocation/Collocation.dsp:
        * tests/Exposed_Policies/Client.dsp:
        * tests/Exposed_Policies/Server.dsp:

          Add in missing path for TAO_Strategies library.

Fri Jul 13 10:48:03 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * TAO_IDL/be/be_visitor_union/union_cs.cpp:
        * TAO_IDL/be/be_visitor_union_branch/pulbic_assign_cs.cpp:

          Fixed code generation for copy constructor and assignment
          operator for object reference union members. These operations
          were failing if the rhs union was uninitialized. Thanks to
          Matt Cheers <matt.cheers@boeing.com> for sending in the example
          IDL file and application code.

Thu Jul 12 22:22:31 2001  Ossama Othman  <ossama@uci.edu>

        * tao/Connector_Registry.cpp (create_profile):

          Print the hexadecimal value of the unknown profile tag in the
          debugging output rather than the decimal value.  The former is
          generally more useful.

Thu Jul 12 21:36:53 2001  Ossama Othman  <ossama@uci.edu>

        * orbsvcs/orbsvcs/Makefile (MKLIST):

          The AV Service needs the CosNaming and CosProperty libraries.
          Add their Makefiles to the build list if they aren't already
          there.

Thu Jul 12 20:40:51 2001  Ossama Othman  <ossama@uci.edu>

        * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Connector.cpp (connect):

          Corrected placement of an ACE_CHECK_RETURN statement.

Thu Jul 12 20:29:42 2001  Ossama Othman  <ossama@uci.edu>

        * tao/ORB_Core.i (resolve_rt_orb, resolve_rt_current):

          Fixed broken emulated exception code.  ACE_CHECK_RETURN should
          have been used instead of ACE_TRY_FLAG.

Thu Jul 12 20:21:36 2001  Ossama Othman  <ossama@uci.edu>

        * docs/tutorials/Quoter/Makefile (DIRS):
        * docs/tutorials/Quoter/Simple/Makefile (DIRS):

          Do not build certain directories if AMI is disabled or if
          Minimum CORBA is enabled.  Some tutorials require AMI, or
          features that are not available in the minimum CORBA
          configuration.

Thu Jul 12 19:45:06 2001  Ossama Othman  <ossama@uci.edu>

        * orbsvcs/Naming_Service/NT_Naming_Server.cpp (ConsoleHandler):

          Fixed unused argument warning.

Thu Jul 12 17:30:42 2001  Frank Hunleth  <fhunleth@cs.wustl.edu>

        * tao/orbconf.h:
        * tao/ORB.cpp:
        * tao/ORB_Core.cpp:
        * tao/ORB_Core.h:
        * tao/ORB_Core.i:

          Removed the RTORB and RTCurrent from the hardcoded list of
          initial references.  Now, both get added dynamically when the
          RTCORBA library is loaded using the PortableInterceptors.

          This should fix a segfault that was reported by Max
          Voronoy <M.Voronoy@telesenskscl.com.ua> and investigated by
          Ossama.

        * tao/RTCORBA/RTCORBA.cpp:
        * tao/RTCORBA/RT_ORB.cpp:
        * tao/RTCORBA/RT_ORB.h:
        * tao/RTCORBA/RT_ORBInitializer.cpp:
        * tao/RTCORBA/RT_ORBInitializer.h:
        * tao/RTCORBA/RT_ORB_Loader.cpp:
        * tao/RTCORBA/RT_ORB_Loader.h:
        * tao/RTCORBA/Thread_Pool.cpp:
        * tao/RTCORBA/Thread_Pool.h:

          Code changes necessary to support creating the RTORB and
          RTCurrent in the pre_init method rather than on demand.
          This is necessary to add both objects to the initial
          references list using the PortableInterceptors.

        * tests/RTCORBA/Makefile:
        * tests/RTCORBA/Makefile.bor:
        * tests/RTCORBA/README:
        * tests/RTCORBA/ORB_init/Makefile:
        * tests/RTCORBA/ORB_init/Makefile.bor:
        * tests/RTCORBA/ORB_init/ORB_init.cpp:
        * tests/RTCORBA/ORB_init/ORB_init.dsp:
        * tests/RTCORBA/ORB_init/README:
        * tests/RTCORBA/ORB_init/run_test.pl:

          Added unit test to check for correct processing when
          instantiating multiple RT enabled ORBs.  Looking at some
          of the RT code indicated that there might be a problem,
          and there actually was.  Thanks to Irfan for writing
          the test.

        * tao/TAO_Internal.cpp:

          Added check to automatically initialize RTCORBA if it
          has been linked in.  This should fix a common source
          of RTCORBA errors.  Thanks to Irfan for the suggestion.

        * tests/RTCORBA/Private_Connection/svc.conf:
        * tests/RTCORBA/Server_Declared/server.conf:
        * tests/RTCORBA/Server_Declared/svc.conf:
        * tests/RTCORBA/Server_Protocol/server_iiop.conf:
        * tests/RTCORBA/Server_Protocol/server_reverse.conf:
        * tests/RTCORBA/Server_Protocol/server_reverse_nt.conf:
        * tests/RTCORBA/Server_Protocol/server_shmiop.conf:
        * tests/RTCORBA/Server_Protocol/svc.conf:
        * tests/RTCORBA/Thread_Pool/svc.conf:
        * tests/Exposed_Policies/server.conf:
        * tests/Exposed_Policies/svc.conf:
        * tests/RTCORBA/Banded_Connections/server.conf:
        * tests/RTCORBA/Banded_Connections/svc.conf:
        * tests/RTCORBA/Client_Propagated/svc.conf:
        * tests/RTCORBA/Client_Protocol/svc.conf:
        * tests/RTCORBA/Explicit_Binding/svc.conf:
        * tests/RTCORBA/MT_Client_Protocol_Priority/server.conf:
        * tests/RTCORBA/MT_Client_Protocol_Priority/svc.conf:

          Updated configuration files to remove RT_ORB initialization
          to test out the above change.


Thu Jul 12 15:10:52 2001  Ossama Othman  <ossama@uci.edu>

        * TAO_IDL/driver/drv_preproc.cpp (DRV_pre_proc):

          Open the temporary file with the O_EXCL flag to close a symbolic
          link attack vulnerability.

Thu Jul 12 13:31:25 2001  Ossama Othman  <ossama@uci.edu>

        * orbsvcs/orbsvcs/Security/EstablishTrustPolicy.h:

          Corrected constructor signature to match the implementation.

        * orbsvcs/orbsvcs/Security/Security_PolicyFactory.cpp
          (create_policy):

          Support creation of the SecurityLevel2::EstablishTrustPolicy.

        * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Connector.h:
        * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Connector.cpp (connect):

          Refactored IIOP-specific and SSLIOP-specific code into new
          iiop_connect() and ssliop_connect() methods, respectively.

          (ssliop_connect):

          Added support for the SecurityLevel2::EstablishTrustPolicy.

        * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Endpoint.h:
        * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Endpoint.i (object_addr):

          Cache the SSLIOP-specific ACE_INET_Addr in the
          TAO_SSLIOP_Endpoint rather than initializing one each time an
          invocation is made.  This should improve SSLIOP performance.

        * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Endpoint.cpp:

          Removed left over Security::Detect{Replay,Misordering} security
          association bits from the default security association value.
          They were actually unused, but removed them anwyay since SSLIOP
          doesn't support either.  No visible run-time change will occur.

        * orbsvcs/orbsvcs/Makefile.Security (FILES):
        * orbsvcs/orbsvcs/Security.bor (OBJFILES):
        * orbsvcs/orbsvcs/Security.dsp:

          Added EstablishTrustPolicy files to the list of sources.

Thu Jul 12 15:10:46 2001  Paul Calabrese <calabrese_p@ociweb.com>

        * examples/PluggableUDP/tests/Basic/svc.conf:
        * examples/PluggableUDP/tests/Performance/svc.conf:
        * tests/RTCORBA/Server_Protocol/server_iiop.conf:
        * tests/RTCORBA/Server_Protocol/server_reverse.conf:
        * tests/RTCORBA/Server_Protocol/server_reverse_nt.conf:
        * tests/RTCORBA/Server_Protocol/server_shmiop.conf:

          Switch these configuration files to use the
          Advanced_Resource_Factory.  This way their options
          actually do something.

Thu Jul 12 13:07:20 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * tests/IDL_Test/interface.idl:
        * tests/IDL_Test/reopened_modules.idl:

          Moved some example code having nested modules from
          interface.idl to reopened_modules.idl. This move hides
          the generated code from both configurations of VxWorks,
          which can't compile C++ code generated from nested IDL
          modules.

Thu Jul 12 07:11:39 2001  Ossama Othman  <ossama@uci.edu>

        * tao/Strategies/advanced_resource.cpp (init):

          Fixed broken code that initialized a stack allocated string
          array with a non-const size.  Use a CORBA::StringSeq instead.

Wed Jul 11 18:00:13 2001  Ossama Othman  <ossama@uci.edu>

        * tao/Invocation_Endpoint_Selectors.cpp (select_endpoint):

          Applied patch from Wayne Erchak <werchak@stentor.com> that fixes
          problem where the endpoint list in a given profile was not
          iterated through during connection failures.  "Fail-over"
          semantics for a non-RTCORBA configured ORB once again work.
          [Bug 927]

        * orbsvcs/orbsvcs/Security/EstablishTrustPolicy.h:
        * orbsvcs/orbsvcs/Security/QOPPolicy.h:

          Added Doxygen comments for the TAO_QOPPolicy and
          TAO_EstablishTrustPolicy classes.

Wed Jul 11 16:03:49 2001  Ossama Othman  <ossama@uci.edu>

        * orbsvcs/orbsvcs/Security/EstablishTrustPolicy.h:
        * orbsvcs/orbsvcs/Security/EstablishTrustPolicy.cpp:

          Implementation of the SecurityLevel2::EstablishTrustPolicy.  It
          makes it possible to control whether or not client/target
          authentication is performed.

Wed Jul 11 16:03:58 2001  Paul Calabrese <calabrese_p@ociweb.com>

        * tao/Object_Loader.h:
        * tao/Resource_Factory.cpp:
        * tao/Resource_Factory.h:
        * tao/Server_Strategy_Factory.cpp:
        * tao/Server_Strategy_Factory.h:
        * tao/default_client.cpp:
        * tao/default_client.h:
        * tao/default_resource.cpp:
        * tao/default_resource.h:
        * tao/default_server.cpp:
        * tao/default_server.h:
        * tao/Strategies/advanced_resource.cpp:
        * tao/Strategies/advanced_resource.h:

          Improved processing of service configurator options.  Many
          failures were not being reported to the user.  These changes
          cause many failure to print warning messages but should not
          affect execution of the code.  The following situations now
          result in warning messages:
            - Passing unknown -ORB* options to a factory
              (This includes passing advanced resource factory
              options to the default resource factory)
            - Passing an unknown value to a defined option
            - Passing options to the default resource factory
              (Resource_Factory) when the advanced resource factory
              (Advanced_Resource_Factory) is being used

          I also removed several deprecated options: -ORBEventLoopLock,
          -ORBDemuxStrategy, and -ORBConnectorLock.

        * examples/AMI/FL_Callback/AMI_Peer.dsp:
        * examples/AMI/FL_Callback/AMI_Progress.dsp:
        * examples/AMI/FL_Callback/Makefile:
        * examples/AMI/FL_Callback/peer.conf:
        * examples/AMI/FL_Callback/peer.cpp:
        * examples/AMI/FL_Callback/svc.conf:
        * orbsvcs/tests/AVStreams/Latency/Makefile:
        * orbsvcs/tests/AVStreams/Latency/control.cpp:
        * orbsvcs/tests/AVStreams/Latency/control.dsp:
        * orbsvcs/tests/AVStreams/Latency/ping.cpp:
        * orbsvcs/tests/AVStreams/Latency/ping.dsp:
        * orbsvcs/tests/AVStreams/Latency/pong.cpp:
        * orbsvcs/tests/AVStreams/Latency/pong.dsp:
        * orbsvcs/tests/AVStreams/Latency/svc.conf:
        * orbsvcs/tests/Event/Performance/latency.conf:
        * orbsvcs/tests/Notify/performance-tests/RedGreen/svc.conf:
        * performance-tests/Callback/Makefile:
        * performance-tests/Callback/client.bor:
        * performance-tests/Callback/client.cpp:
        * performance-tests/Callback/client.dsp:
        * performance-tests/Callback/server.bor:
        * performance-tests/Callback/server.cpp:
        * performance-tests/Callback/server.dsp:
        * performance-tests/Callback/svc.conf:
        * performance-tests/Cubit/TAO/IDL_Cubit/collocation_test.cpp:
        * performance-tests/Cubit/TAO/IDL_Cubit/iiop_lite.conf:
        * performance-tests/Cubit/TAO/IDL_Cubit/run_test.pl:
        * performance-tests/Cubit/TAO/IDL_Cubit/svc.mt_server.conf:
        * performance-tests/Cubit/TAO/IDL_Cubit/svc.st_client.conf:
        * performance-tests/Cubit/TAO/IDL_Cubit/svc.st_server.conf:
        * performance-tests/Cubit/TAO/IDL_Cubit/svc.zero_lock.conf:
        * performance-tests/Cubit/TAO/IDL_Cubit/uiop_lite.conf:
        * performance-tests/Cubit/TAO/IDL_Cubit/collocation/svc.conf:
        * performance-tests/Cubit/TAO/MT_Cubit/Makefile:
        * performance-tests/Cubit/TAO/MT_Cubit/client.bor:
        * performance-tests/Cubit/TAO/MT_Cubit/client.cpp:
        * performance-tests/Cubit/TAO/MT_Cubit/client.dsp:
        * performance-tests/Cubit/TAO/MT_Cubit/server.cpp:
        * performance-tests/Cubit/TAO/MT_Cubit/server.dsp:
        * performance-tests/Cubit/TAO/MT_Cubit/svc.conf:
        * performance-tests/RTCorba/Multiple_Endpoints/Orb_Per_Priority/client.conf:
        * performance-tests/RTCorba/Multiple_Endpoints/Orb_Per_Priority/server.conf:
        * performance-tests/RTCorba/Multiple_Endpoints/Orb_Per_Priority/client.cpp:
        * performance-tests/RTCorba/Multiple_Endpoints/Orb_Per_Priority/server.cpp:
        * performance-tests/RTCorba/Multiple_Endpoints/Orb_Per_Priority/Makefile:
        * performance-tests/RTCorba/Multiple_Endpoints/Single_Endpoint/client.conf:
        * performance-tests/Throughput/Makefile:
        * performance-tests/Throughput/client.bor:
        * performance-tests/Throughput/client.cpp:
        * performance-tests/Throughput/server.bor:
        * performance-tests/Throughput/server.cpp:
        * performance-tests/Throughput/svc.conf:
        * tests/AMI_Timeouts/svc.conf:
        * tests/Collocation/Collocation.bor:
        * tests/Collocation/Collocation.cpp:
        * tests/Collocation/Collocation.dsp:
        * tests/Collocation/Makefile.test:
        * tests/Collocation/svc.conf:
        * tests/Exposed_Policies/Client.dsp:
        * tests/Exposed_Policies/Makefile:
        * tests/Exposed_Policies/Server.dsp:
        * tests/Exposed_Policies/client.bor:
        * tests/Exposed_Policies/client.cpp:
        * tests/Exposed_Policies/server.bor:
        * tests/Exposed_Policies/server.conf:
        * tests/Exposed_Policies/server.cpp:
        * tests/Exposed_Policies/svc.conf:
        * tests/FL_Cube/Makefile:
        * tests/FL_Cube/client.bor:
        * tests/FL_Cube/client.cpp:
        * tests/FL_Cube/server.bor:
        * tests/FL_Cube/server.cpp:
        * tests/FL_Cube/svc.conf:
        * tests/LongUpcalls/svc.conf:
        * tests/MT_Server/server.conf:
        * tests/Strategies/Makefile:
        * tests/Strategies/README:
        * tests/Strategies/client.bor:
        * tests/Strategies/server.bor:
        * tests/Strategies/svc.conf

          These are changes to TAO tests and examples with
          broken service config files.  The above changes
          caused these files to generate warnings.  The vast
          majority of the problems were the passing of advanced
          resource factory options to the default resource
          factory.  Most were changed to simply use the advanced
          resource factory.  Some were modified to use
          the defaults of the default resource factory.

Wed Jul 11 14:41:20 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * TAO_IDL/ast/ast_decl.cpp (compute_prefix):

          Fixed function to work if the #pragma prefix string looks like
          '#pragma prefix      "foo.bar"'. The preprocessor takes care
          of the whitespace, if any, between '#pragma' and 'prefix'.

Wed Jul 11 14:38:00 2001  Craig Rodrigues <crodrigu@bbn.com>

       * tao/Strategies/DIOP_Connection_Handler.cpp:
         Only display debugging message if TAO_debug_level is set > 5.

Wed Jul 11 13:49:00 2001  Craig Rodrigues <crodrigu@bbn.com>

        * utils/catior/catior.cpp:
          Add support for printing DIOP (GIOP over UDP) profiles.

Wed Jul 11 08:39:14 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * tao/RTCORBA/RT_Protocols_Hooks.cpp (get_thread_CORBA_priority):

          Replace ACE_CHECK with ACE_CHECK_RETURN (-1) - function must
          return an int.

Wed Jul 11 09:25:00 2001  Craig Rodrigues  <crodrigu@bbn.com>

        * orbsvcs/tests/Makefile.bor:
          Add Simple_Naming test

Wed Jul 11 02:19:59 2001  Irfan Pyarali  <irfan@cs.wustl.edu>

        * tao/RTPortableServer/RT_Servant_Dispatcher.cpp: Remember the
          native priority of the invoking thread.  Once the invocation
          completes, restore the thread to the original native thread
          priority.

          Previously, when the priority was restored, it was done it terms
          of CORBA priority.  The problem with this approach was that in
          the conversion from native to CORBA and then back to native, we
          can lose precision and hence the native priority of the native
          thread may not be the same as its original native priority.
          Therefore, remembering the original native priority will
          alleviate this problem.

        * tao/Protocols_Hooks:
        * tao/Default_Protocols_Hooks:
        * tao/RT_Protocols_Hooks.cpp:

          Expanded the priority interfaces on the Protocols_Hooks such
          that the user can obtain both the native and CORBA priorities.
          Also, the user can set both.

        * tao/RTCORBA/RT_Invocation_Endpoint_Selectors.cpp (select_endpoint):
        * tao/RTCORBA/RT_Current.cpp (the_priority):
        * tao/Strategies/Reactor_Per_Priority.cpp (reactor):

          Changed to use new priority interfaces.

Tue Jul 10 22:37:48 2001  Ossama Othman  <ossama@uci.edu>

        * orbsvcs/orbsvcs/Security/QOPPolicy.cpp (copy):

          Perform a deep copy, not a shallow copy.  The copy is supposed
          to be independent of the original.

Tue Jul 10 15:59:31 2001  Ossama Othman  <ossama@uci.edu>

        * tao/PortableInterceptorC.h:

          Include "PolicyC.h" to pull in some policy related exception
          definitions.  This fixed a problem with some minimum CORBA
          builds with native exception support enabled.  Thanks to
          Sangeetha Ramadurai <Sangeetha.Ramadurai@geind.ge.com> for
          reporting the problem and providing a fix.

Tue Jul 10 15:56:13 2001  Krishnakumar B  <kitty@cs.wustl.edu>

        * tao/GIOP_Message_Lite.h:
        * tao/GIOP_Message_Lite.cpp:
        * tao/Pluggable_Messaging.h:
        * tao/Pluggable_Messaging.cpp:

          Fixed warning about virtual function override. Was caught by the
          Tru64 compiler.

Tue Jul 10 13:09:18 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * TAO_IDL/be/be_visitor_valuetype/valuetype.cpp:

          Corrected a generation of skel_export_macro to
          stub_export_macro. Thanks to Alexander Rieger
          <Alexander.Rieger@inka.de> for tracking this down.

Tue Jul 10 11:51:05 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * TAO_IDL/be/be_visitor_interface/interface_ch.cpp:
        * TAO_IDL/be/be_visitor_interface/interface_sh.cpp:

          Changed code so that the inheritance list of a class
          declaration consists of fully scoped names. The stub
          generation was using ACE_NESTED_CLASS, and the skeleton
          generation was using a computed relative name. Both
          generated uncompilable code in some cases. Thanks to
          Richard L. Johnson <rich@huey.jpl.nasa.gov> for reporting
          this bug and for sending in the example IDL file.

        * tests/IDL_Test/interface.idl:

          Added example IDL to this file in the IDL test suite.

Tue Jul 10 05:52:09 2001  Douglas C. Schmidt  <schmidt@ace.cs.wustl.edu>

        * tao/Profile.h,
        * tao/Object.h: Fixed typos in comments.  Thanks to Johnny
          Willemsen for reporting this.

        * tao/Connector_Registry.h (TAO_Connector_Registry): Fixed a typo
          where svc.comf should have been svc.conf.  Thanks to Johnny
          Willemsen for reporting this.

        * tao/ORB.cpp: Added an addition expression to the #ifdef for the
          using std::set_unexpected declaration.  Thanks to Scott Plant
          <splant@softhome.net> for reporting this.

Tue Jul 10 00:45:00 2001  Ossama Othman  <ossama@uci.edu>

        * tao/RTCORBA/RT_ORB.cpp (TAO_RT_CORBA_Priority_Normalizer):

          Fixed busted code that did not check if the resolved
          PriorityMappingManager object reference was nil.  Since this
          check was missing, a seg fault would occur when attempting to
          invoke a method on that mapping manager.  Throw a
          CORBA::INTERNAL() exception if the reference is nil.  This at
          leasts lets us identify a problem without seg faulting.

          Fixed broken code that did not use emulated exceptions properly.
          The ACE_TRY/CATCH block was missing.

Tue Jul 10 09:13:12 2001  Johnny Willemsen <jwillemsen@remedy.nl>

        * tao/orbsvcs/orbsvcs/security.bor

          Added the missing files Security_PolicyFactory and QOPPolicy

Mon Jul 09 23:43:29 2001  Ossama Othman  <ossama@uci.edu>

        * tao/params.i (default_init_ref):

          Use CORBA::string_dup() to return a duplicate string instead of
          relying on ACE_CString::rep().  This allows us to portably take
          advantage of CORBA::String_var in TAO_ORB_Core::resolve_rir().

        * tao/ORB_Core.cpp (list_initial_references):

          Corrected long standing bug where the stringified object
          reference was placed in the ObjectIdList instead of the ObjectId
          that represents it.

          Include the ObjectIds stored in the underlying table for
          the ORB::register_initial_reference() mechanism in the returned
          ObjectIdList.  This was a bug.

          (TAO_ORB_Core, fini):

          The rt_priority_mapping_manager_ cached object reference member
          was unused.  Removed its initialization and finalization code.

          (resolve_rir):

          Improved exception-safety of this method by using a
          CORBA::String_var instead of relying solely on delete().

        * tao/ORB_Core.h (rt_priority_mapping_manager_):

          Removed this unused attribute.

        * tao/Object_Ref_Table.h (begin, end):

          Made iterator accessors public so that the ORB_Core can use
          them.

        * tao/Object_Ref_Table.cpp:

          Removed three unnecessary Hash Map related template
          instantiations.

          (current_size):

          Return the current size of the underlying table.

        * tests/InterOp-Naming/README:

          Corrected format for corbaloc IORs.

Mon Jul  9 17:39:21 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * orbsvcs/tests/Security/Secure_Invocation/client.cpp:

          Added .in() to a CORBA::ORB_var passed to a function.

Mon Jul  9 14:56:26 2001  Douglas C. Schmidt  <schmidt@tango.doc.wustl.edu>

        * docs/components.html (bgcolor): Fixed a typo.  Thanks to
          John Ashmun for reporting this.

Mon Jul 09 10:56:30 2001  Ossama Othman  <ossama@uci.edu>

        * tao/DynamicInterface/Server_Request.h:
        * tao/DynamicInterface/Server_Request.inl (_tao_server_request):

          Added this accessor that returns a reference to the underlying
          TAO_ServerRequest object.  Thanks to James Megquier
          <jmegq@bbn.com> and Mouna Seri <seri@crhc.uiuc.edu> for
          suggesting addition of this accessor.

Mon Jul  9 11:16:36 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * tao/IFR_Client/TAO_IFR_Client.dsp:
        * tao/TypeCodeFactory/TypeCodeFactory.dsp:

          Added MFC debug and release versions to these projects.
          Thanks to Francois Bernier <fbernier@gel.ulaval.ca> for
          sending them in.

Mon Jul  9 10:13:54 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * TAO_IDL/be/be_interface.cpp:
        * TAO_IDL/be_visitor_interface/interface_ch.cpp:
        * TAO_IDL/be_visitor_interface/interface_cs.cpp:
        * tao/Object.cpp:
        * tao/Object.h:
        * tao/Object.i:

          Added static int _tao_class_id to CORBA::Object and
          to all classes generated from IDL interfaces. The
          address of this variable is used in all
          _unchecked_narrow() and _tao_QueryInterface() methods.
          Formerly, the address of the _narrow() method was used,
          but on BCB, this address is not usable across DLLs.
          Thanks to Christopher Kohlhoff <chris@kohlhoff.com> for
          suggesting the fix.

Mon Jul  9 15:27:12 2001  Johnny Willemsen <jwillemsen@remedy.nl>

        * tao/orbsvcs/tests/AVStreams/Makefile.bor
        * tao/orbsvcs/tests/AVStreams/Multicast_Full_Profile/*.bor
        * tao/orbsvcs/tests/AVStreams/Full_Profile/*.bor

          Added BCB makefiles for the mentioned AVStreams test

Mon Jul 09 04:11:45 2001  Ossama Othman  <ossama@uci.edu>

        * orbsvcs/orbsvcs/Security/QOPPolicy.h:

          Added missing "ace/post.h" include.

Mon Jul 09 02:43:38 2001  Ossama Othman  <ossama@uci.edu>

        * orbsvcs/tests/Security/Secure_Invocation/Makefile:
        * orbsvcs/tests/Security/Secure_Invocation/client.cpp:
        * orbsvcs/tests/Security/Secure_Invocation/client.dsp:
        * orbsvcs/tests/Security/Secure_Invocation/run_test.pl:

          Updated this test to take advantage of the newly added
          Security::QOPPolicy support.  The test is more self-contained
          now, i.e. it no longer depends so much on the "run_test.pl" Perl
          script to test all features.

Mon Jul 09 02:21:32 2001  Ossama Othman  <ossama@uci.edu>

        * docs/releasenotes/index.html:

          Updated Security Service release notes.

Mon Jul 09 01:51:27 2001  Ossama Othman  <ossama@uci.edu>

        * tao/Invocation.h:
        * tao/Invocation.i:

          Added accessors to the transport and max_wait_time members.

        * tao/Invocation.cpp (perform_call):

          Updated invocation of the connect() method to match the
          signature change described below.

        * tao/Connector_Registry.cpp:
        * tao/Connector_Registry.h:
        * tao/IIOP_Connector.cpp:
        * tao/IIOP_Connector.h:
        * tao/Pluggable.h:
        * tao/Strategies/DIOP_Connector.cpp:
        * tao/Strategies/DIOP_Connector.h:
        * tao/Strategies/SHMIOP_Connector.cpp:
        * tao/Strategies/SHMIOP_Connector.h:
        * tao/Strategies/UIOP_Connector.cpp:
        * tao/Strategies/UIOP_Connector.h:
        * orbsvcs/orbsvcs/SSLIOP/IIOP_SSL_Connector.h:
        * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Connector.h:

          Changed signature of connect() method to accept a pointer to a
          TAO_GIOP_Invocation object.  Some pluggable protocols
          (e.g. SSLIOP) may need access to some of the information in that
          object.

        * orbsvcs/orbsvcs/Security.idl:

          Corrected spelling of Security::SecFeaturePolicy.  "Feature" not
          "Features".

        * orbsvcs/orbsvcs/Security/QOPPolicy.cpp:
        * orbsvcs/orbsvcs/Security/QOPPolicy.h:

          Implementation of the Security::QOPPolicy policy.  This policy
          is used to set the desired invocation Quality-of-Protection
          (QoP).  It can be created using ORB::create_policy(), and used
          in conjunction with the standard policy manipulation CORBA
          features (e.g. PolicyManager, PolicyCurrent), meaning that this
          policy can be set on a per-ORB, per-thread or per-object basis.

          This policy makes it possible to, for example, make both secure
          and insecure invocations within the same client process.

        * orbsvcs/orbsvcs/Security/Security_PolicyFactory.cpp:
        * orbsvcs/orbsvcs/Security/Security_PolicyFactory.h:

          Security policy factory implementation that is registered with
          the ORB's policy factory registry.

        * orbsvcs/orbsvcs/Security/Security_ORBInitializer.cpp (post_init,
          register_policy_factories):

          Register the supported security policy factories with the ORB.

        * orbsvcs/orbsvcs/Security/Security_ORBInitializer.h:

          Added a shared security policy factory member to this
          ORBInitializer.  The factory is reentrant so there is no need
          create one for each ORB.  Just share one between all ORBs.

        * orbsvcs/orbsvcs/SSLIOP/IIOP_SSL_Connector.cpp (connect):

          Moved code that obtained the TAO_IIOP_Endpoint from the
          TAO_SSLIOP_Endpoint from this method to the corresponding one in
          TAO_SSLIOP_Connector.  This cleans up the code a bit.  It
          shouldn't differ all that much from the
          IIOP_Connector::connect() implementation.

        * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Connector.cpp (connect):

          Now that both secure and insecure invocations are supported
          within the same client process, an IIOP-only transport
          descriptor must be used instead of the one passed to this method
          since the latter is used for SSLIOP connections.  Doing so
          prevents an IIOP-only cached transport from being associated
          with an SSLIOP connection.

          This fixes a problem that was revealed when support for secure
          and insecure invocations within the same client process was
          added (i.e. via the Security::QOPPolicy support).  An insecure
          cached transport was matched against an SSLIOP connection.

        * tao/IIOP_Endpoint.i:
        * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Endpoint.cpp:

          Cosmetic updates.

        * tao/Exception.cpp:

          Corrected "unknown description" message.

        * tao/Strategies/SHMIOP_Endpoint.h:
        * tao/Strategies/UIOP_Endpoint.h:

          Doxygen-ated these headers.

        * orbsvcs/orbsvcs/Makefile.Security:
        * orbsvcs/orbsvcs/Security.dsp:
        * orbsvcs/orbsvcs/Security/Security.bor:

          Added new QOPPolicy and Security_PolicyFactory filenames to
          these Makefiles and project files.

Mon Jul  9 08:41:12 2001  Johnny Willemsen <jwillemsen@remedy.nl>

        * tao/orbsvcs/tests/AVStreams/Makefile.bor
        * tao/orbsvcs/tests/AVStreams/Pluggable/*.bor
        * tao/orbsvcs/tests/AVStreams/Multicast/*.bor

          Added BCB makefiles for the mentioned AVStreams test

Sat Jul  7 19:44:55 2001  Krishnakumar B  <kitty@cs.wustl.edu>

        * tao/Incoming_Message_Queue.inl (get_queued_data):

          Moved the definition to the beginning to fix warning. Was caught
          by Tru64 cxx compiler.

Fri Jul  6 22:04:51 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tests/LongUpcalls/run_ami_test.pl: Made the perl script
          runnable.  It was giving weird errors in the builds.

Fri Jul 06 21:53:26 2001  Christopher Kohlhoff  <chris@kohlhoff.com>

        * tao/Makefile.bor:
        * tao/Strategies/Makefile.bor:
          Removed GIOP Lite support and added new source files.

        * orbsvcs/orbsvcs/AV/AV_Core.h:
          Fixed borland #pragmas to correctly reset previous options.

Fri Jul  6 16:58:52 2001  Ossama Othman  <ossama@uci.edu>

        * orbsvcs/Notify_Service/Notify_Service.cpp (shutdown):

          Applied the same fix from Jody's patch described below to this
          method.

Fri Jul  6 15:59:17 2001  Ossama Othman  <ossama@uci.edu>

        * orbsvcs/Notify_Service/Notify_Service.h (naming_):
        * orbsvcs/Notify_Service/Notify_Service.cpp (init,
          resolve_naming_service):

          Integrated patch from Jody Hagins <jody@atdesk.com>.  Take
          advantage of the methods provided by the
          CosNaming::NamingContextExt interface so that the Notification
          Service does the right thing with an option such as
          "-Factory Foo.Bar/My_Event_Channel".

Fri Jul  6 18:02:07 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/TAO_Static.dsp: Removed support for GIOP Lite for the time
          being. Just want to make sure that the interface and other stuff
          stabilises before we can squeeze this in.

Fri Jul  6 17:55:25 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * TAO_IDL/be/be_visitor_argument/pre_invoke_cs.cpp (void_return_type):

          Change a call to base_node_type() on the operation's return type
          node to node_type(), which is the same but does no unaliasing.
          We are checking for a return type of void, which can't be aliased
          anyway. The return from base_node_type() checks for equality to
          the enum value NT_pre_defined, which now returns true for some
          aliased return types, and that causes problems later in the
          function. Thanks to Walter Wehrli <walter.wehrli@ubs.com> for
          sending in the example IDL file that uncovered this bug.

Fri Jul  6 17:13:20 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * tao/ORB_Core.cpp:

          Changed the string passed to
          ACE_Dynamic_Service<TAO_Object_Loader>::instance from
          "TypeCodeFactory" to "TypeCodeFactor_Loader" to match the string
          in the ACE_STATIC_SVC_DEFINE macro in
          TypeCodeFactory_Loader.cpp. Thanks to Francois Bernier
          <fbernier@gel.ulaval.ca> for reporting the bug.

Fri Jul  6 17:10:07 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/GIOP_Message_Base.cpp:
        * tao/Transport.cpp: Added support for LocateRequest & LocateReply
          that. It should have been added before the code from my branch
          came to the main trunk, but somehow got missed.

Fri Jul  6 16:01:45 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Transport.cpp (process_queue_head):
        * tao/GIOP_Message_State.cpp: Fixed a problem that has long been
          there. We have been unlucky that we did not get to this before.

Fri Jul 06 14:58:09 2001  Nanbor Wang  <nanbor@cs.wustl.edu>

        To build "CosEvent - Win32 MFC Release":

        * TAOACE.dsw: Make RTEvent a dependent of CosEvent.

        * orbsvcs/orbsvcs/RTEvent.dsp: Set up IDL custom build rules for
          MFC configs.

        * orbsvcs/orbsvcs/CosEvent.dsp: Changed the name for release
          version libarary to XXXmfc.lib.

          Thanks to Kristopher Johnson <kristopher.johnson@transcore.com>
          for providing the fix.  [Bug 898]

Fri Jul  6 13:33:35 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * TAO_IDL/Makefile.BE:
        * TAO_IDL/Makefile.FE:

          Added lines to force static lib compiliation for mingw. Thanks
          to Cristian Ferretti <cristian_ferretti@yahoo.com> for the
          patches.

Fri Jul  6 13:23:46 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Synch_Reply_Dispatcher.cpp:
        * tao/Asynch_Reply_Dispatcher.cpp: Removed the initialization of a
          buffer  from the constructor. Looks like I used to do it in some
          life of mine and I dont which one it was. :( SunCC 4.2 doesnt
          like that.

        * tao/Resume_Handle.cpp: Added a #include of the reactor to fix
          compile errors in Sun CC4.2.

Fri Jul  6 13:19:25 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * TAO_IDL/be/be_visitor_sequence/gen_bounded_obj_sequence_ch.cpp:
        * TAO_IDL/be/be_visitor_sequence/gen_bounded_sequence_ch.cpp:
        * TAO_IDL/be/be_visitor_sequence/gen_bounded_str_sequence_ch.cpp:
        * TAO_IDL/be/be_visitor_sequence/gen_bounded_wstr_sequence_ch.cpp:
          Added generation of 'TAO_EXPORT_MACRO' to class declaration.
          Thanks to Craig Rodrigues and Christopher Kohlhoff for helping to
          track this stuff down.

         * TAO_IDL/be/be_visitor_sequence/gen_unbounded_sequence_ch.cpp:
          Changed generation of 'TAO_EXPORT_NESTED_MACRO' to 'TAO_EXPORT_MACRO'.

Fri Jul  6 12:38:06 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Transport.cpp: Fixed a bug with thread-per-connection
          model. We dont need to send notify () to the reactor as there is
          no reactor.

Fri Jul  6 11:27:00 2001  Craig Rodrigues  <crodrigu@bbn.com>

        * tao/IFR_Client/IFR_BaseC.h:
        * tao/DynamicAny/DynamicAnyC.h:
        * tao/IORManipulation/IORC.h:
          Export more classes to appease Borland C++.

Fri Jul  6 08:27:06 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Incoming_Message_Queue.h:
        * tao/Resume_Handle.h:
        * orbsvcs/orbsvcs/SSLIOP/IIOP_SSL_Transport.h: Fixed fuzz errors.

Fri Jul  6 08:14:48 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * performance-tests/Cubit/TAO/MT_Cubit/Makefile:
        * performance-tests/Cubit/TAO/IDL_Cubit/Makefile: Generated
          dependencies again.

Fri Jul  6 07:43:15 2001  Chad Elliott  <elliott_c@ociweb.com>

        * orbsvcs/IFR_Service/Makefile:

          Disable building this for Visual Age C++ due to the fact that the
          using keyword is not properly supported.

Fri Jul  6 07:38:55 2001  Chad Elliott  <elliott_c@ociweb.com>

        * performance-tests/Cubit/TAO/MT_Cubit/Task_Client.h:

          Fix warning when using gcc on Tru64.

Fri Jul  6 07:38:44 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * orbsvcs/orbsvcs/SSLIOP.bor:
        * orbsvcs/orbsvcs/SSLIOP.dsp: Added the new file
          IIOP_SSL_Transport.cpp  to the above files.

Fri Jul  6 07:12:44 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * orbsvcs/orbsvcs/Makefile.av: Generated dependencies again. Looks
          like the last dependency generation did not do a good job. This
          should fix all the compile errors seen in the builds.

Fri Jul  6 06:56:59 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Makefile.bor:
        * tao/Strategies/Makefile.bor: Fixed the makefiles. Thanks to
          Johnny Willemsen for alerting me on this.

Thu Jul  6 01:06:55 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Strategies/DIOP_Connection_Handler.h: Fixed a link error
          on Win32 builds.

Fri Jul  6 00:39:38 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * orbsvcs/*/Makefile: Updated dependencies.

Fri Jul  6 00:38:05 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * examples/*/Makefile: Updated dependencies.

Fri Jul  6 00:34:13 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * performance-tests/*/Makefile: Updated dependencies.

Fri Jul  6 00:27:54 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tests/*/Makefile: Updated dependencies.

Thu Jul  6 00:16:55 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/TAO_Static.dsp:
        * tao/Strategies/TAO_Strategies.dsp:
        * tao/Strategies/TAO_Strategies_Static.dsp: Added the new files
          and removed old ones.

Thu Jul  5 23:57:55 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Makefile:
        * tao/*/Makefile: Updated dependencies.

Thu Jul  5 23:49:34 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * orbsvcs/orbsvcs/Makefile.SSLIOP: Added IIOP_SSL_Transport to it.

Thu Jul  5 23:44:16 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Strategies/GIOP_Message_NonReactive_Base.h:
        * tao/Strategies/GIOP_Message_NonReactive_Base.cpp:
        * tao/Strategies/GIOP_Message_NonReactive_Handler.h:
        * tao/Strategies/GIOP_Message_NonReactive_Handler.cpp:
        * tao/Strategies/GIOP_Message_NonReactive_Handler.inl:
        * tao/GIOP_Message_Reactive_Handler.h:
        * tao/GIOP_Message_Reactive_Handler.cpp:
        * tao/GIOP_Message_Reactive_Handler.inl: Removed them from the
          main trunk. They are no longer needed.

Thu Jul  5 23:42:17 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/GIOP_Message_State.i: Removed from the main trunk. Replaced
          that with GIOP_Message_State.inl.

Thu Jul  5 23:30:07 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

          The long awaited fix for bug 575 is finally in!! This has been
          merged  from bug_575_stage_2. The ChangeLog entries start here,

Thu Jul  5 23:00:43 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Connection_Handler.cpp
        * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Connection_Handler.h
        * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Transport.cpp
        * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Transport.h: Fixed some more
          compilation errors in g++.

Thu Jul  5 22:35:07 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * orbsvcs/orbsvcs/SSLIOP/IIOP_SSL_Transport.h:
        * orbsvcs/orbsvcs/SSLIOP/IIOP_SSL_Transport.cpp: Fixed compile
          errors in g++. The new IIOP_SSL_Transport classes now does the
          functionality of the IIOP_SSL_Connection_Handler classes. The
          connection handler classes exist now only to create the
          transport.

Thu Jul  5 21:44:59 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * orbsvcs/orbsvcs/SSLIOP/IIOP_SSL_Transport.h:
        * orbsvcs/orbsvcs/SSLIOP/IIOP_SSL_Transport.cpp: New files for the
          SSLIOP.

        * orbsvcs/orbsvcs/SSLIOP/IIOP_SSL_Connection_Handler.h:
        * orbsvcs/orbsvcs/SSLIOP/IIOP_SSL_Connection_Handler.cpp: Moved
          the handle_input () from the connection handler to transport.

        * tao/Strategies/DIOP_*:  Fixed it for the new setup. We now make
          a buffer of dgram size and use that to read messages.

Thu Jul  5 14:44:59 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        Merged with the main trunk and moved it to a new branch by name
        bug_575_stage_2.

Wed Jul  4 18:53:22 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Transport.cpp (recv_i): Added the
          error checking.

        * tao/Pluggable_Messaging.h: Made reset () as a pure virtual
          function.

        * tao/Strategies/UIOP_Transport.cpp: Removed the close_connection
          () call on TMS.

        * tao/LIST_OF_TODO: Updated..

Wed Jul  4 18:45:22 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/GIOP_Message_State.inl (reset):
        * tao/GIOP_Message_State.h:
        * tao/GIOP_Message_State.cpp:
        * tao/GIOP_Message_Base.h:
        * tao/GIOP_Message_Base.cpp:
        * tao/GIOP_Message_Base.inl:
        * tao/Incoming_Message_Queue.cpp:
        * tao/Pluggable_Messaging.h:
        * tao/Pluggable_Messaging.cpp: Added documentation.

        * tao/Transport.h:
        * tao/Transport.cpp: Fixed a minor bug. In consolidate_message ()
          we were not checking the queue before processing the message on
          hand. Now we check the queue and process the head of the queue
          if it is not empty. In the process we add the message on hand in
          the queue.

Wed Jul  4 16:21:22 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/IIOP_Transport.cpp: Removed the connection_closed () from
          the recv_i () call.

        * tao/Transport.h:
        * tao/Transport.cpp: Added a number of comments and cleaned up the
          code for readability.

        * tao/LIST_OF_TODO: Updated the list.

Wed Jul  4 09:20:22 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Transport.h (TAO_Transport): Added some documentation for
          the incoming_data_path.

        * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Transport.cpp:
        * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Transport.h:
        * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Connection_Handler.cpp:
        * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Connection_Handler.h:  Fixed the
          SSLIOP for the changes. Havent compiled this yet.

Tue Jul  3 17:09:22 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Strategies/SHMIOP_Transport.cpp: Fixed a bug in
          consolidate_message (). The IDL_Cubit tests work fine with the
          SHMIOP & UIOP transports.

Tue Jul  3 16:29:06 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Strategies/SHMIOP_Connection_Handler.cpp:
        * tao/Strategies/SHMIOP_Connection_Handler.h:
        * tao/Strategies/SHMIOP_Transport.cpp: Fixed warnings and errors
          in g++.

        * tao/Strategies/UIOP_Transport.cpp:
        * tao/Strategies/UIOP_Transport.h:
        * tao/Strategies/UIOP_Connection_Handler.h:
        * tao/Strategies/UIOP_Connection_Handler.cpp: Fixed warnungs and
          errors in g++.

        * tao/IIOP_Connection_Handler.cpp: Used fetch_handle () to get the
          handle instead of the handle passed as an argument of the
          handle_input () call.

        * tao/Connection_Handler.h: Added a new #define and changed
          TAO_CONNECTION_HANDLER_BUF_SIZE as
          TAO_CONNECTION_HANDLER_STACK_BUF_SIZE.


Tue Jul 03 15:45:03 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Strategies/SHMIOP_Transport.h:
        * tao/Strategies/SHMIOP_Transport.cpp:
        * tao/Strategies/SHMIOP_Connection_Handler.h:
        * tao/Strategies/SHMIOP_Connection_Handler.cpp: Fixed the protocol
          to work  with the latest changes.

        * tao/Strategies/TAO_Strategies.dsp: Removed the
          GIOP_NonReactive_* files.

        * tao/Resume_Handle.cpp: Fixed a bug in resume_handle (). If we
          have a null ORB_Core we just dont resume the handle.

Mon Jul 02 23:16:03 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/IIOP_Transport.cpp:
        * tao/IIOP_Transport.h: Removed some functions which would no
          longer be used.

        * tao/Transport.h: Added some documentation for the incoming data
          path.

        * tao/Strategies/UIOP_Transport.h:
        * tao/Strategies/UIOP_Transport.cpp:
        * tao/Strategies/UIOP_Connection_Handler.cpp: Brought them up in
          sync with IIOP.

Mon Jul 02 20:28:03 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        This checkin fixes many problems with multiple calls being read in
        a single read.

        * tao/GIOP_Message_Base.cpp: Changed the way we calculate the
          remaining length that needs to be copied if the last read had
          left with a very small piece of message ie. less than 12 bytes
          in the queue.

        * tao/GIOP_Message_State.cpp: Fixed a check condition before we go
          ahead to parse the header.

        * tao/Incoming_Message_Queue.h:
        * tao/Incoming_Message_Queue.cpp:
        * tao/Incoming_Message_Queue.inl: Added a method set_flags
          (). Further the condition for resuming the handler has been
          relaxed. Clients of this class can decide not to resume handlers
          at all. This comes in handy at times. Also added a operator=
          method.

        * tao/Transport.cpp:
        * tao/Transport.h: One of the bigger problems have been
          solved. The way we now multiple oneways is like this

          - The leader thread uses the transport and reads more messages
          - It splits up the messages in to pieces and queues them up.
          - It just takes one message and processes (the head of the
            queue). The changes above makes sure that only one of the
            message is processed.
          - Before processing if it finds one more message it sends a
            notify () to the reactor. An important point is that it does
            this without resuming the handler.
          - The notify call processes one more message. Before processing
            if it finds one more complete message it just sends another
            notify to the reactor.
          - The thread that reads the last complete message from the queue
            resumes the handler before processing the message.

          By the above process we dont starve any thread and at the
          same time ensures concurrency within the ORB.

        * tao/LIST_OF_TODO: Updated..

Sun Jul 01 18:35:03 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Asynch_Reply_Dispatcher.h:
        * tao/Asynch_Reply_Dispatcher.cpp:
        * tao/DynamicInterface/DII_Reply_Dispatcher.h:
        * tao/DynamicInterface/DII_Reply_Dispatcher.cpp:  Applied the same
          optimization that was done in the last checkin. The last checkin
          works fine. It is no more a suspect.

Sun Jul 01 18:05:03 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/GIOP_Message_Base.cpp: Fixed small problems in getting
          message types.

        * tao/Synch_Reply_Dispatcher.h:
        * tao/Synch_Reply_Dispatcher.cpp:
        * tao/Pluggable_Messaging_Utils.h:
        * tao/Pluggable_Messaging_Utils.cpp: Suspect optimisations added
          to create datablocks on stack.

        * tao/LIST_OF_TODO: Updated list..

Sun Jul 01 09:05:03 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Transport_Mux_Strategy.h:
        * tao/Muxed_TMS.h:
        * tao/Muxed_TMS.cpp:
        * tao/Exclusive_TMS.cpp:
        * tao/Exclusive_TMS.h: Removed commented out code. They have been
          commented out for sometime that we dont need them anymore.

        * tao/Incoming_Message_Queue.cpp:
        * tao/Incoming_Message_Queue.inl:
        * tao/Incoming_Message_Queue.h: Added lots of comments. Changed
          the name of the method copy_message () to copy_tail () and
          missing_data () to missing_data_tail (). Added a new static
          method TAO_Queued_Data::release () that releases a node.

        * tao/Transport.cpp: Accomodated the changes to the
          Incoming_Message_Queue to the Transport class.

        * tao/LIST_OF_TODO: Updated ..

Sat Jun 30 13:00:03 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/GIOP_Message_Base.cpp:
        * tao/GIOP_Message_Base.h: Fixed a subtle bug while allocating
          memory for the datablock in the queue. Memory was allocated only
          for the exact number of bytes without thinking about the
          alignment. Increased the number of bytes allocated.

          Created the outgoing CDR streams on the stack with memory drawn
          from the TSS. We cannot share the buffers like the way it has
          been done so far.

        * tao/IIOP_Transport.cpp: Increased the debug_level for a debug
          statement.

        * This checkin fixes the AMI problems that have been seen so far.

        * tao/LIST_OF_TODO: Updated the list.

Fri Jun 29 18:30:03 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/GIOP_Message_Base.cpp: Fixed a bug in calculating the number
          of bytes that  needs to be copied into queue.

        * tao/Incoming_Message_Queue.cpp:
        * tao/Incoming_Message_Queue.h: Changed the signature of copy
          (). It returns the number of bytes copied instead of a void.

        * tao/Transport.cpp (parse_consolidate_messages):
        * tao/Transport.h: Added a new method parse_consolidate_messages
          () to the class. It does some common functaionalities like
          parsing the messages & consolidating the messages.

Thu Jun 28 18:30:03 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Resume_Handle.cpp: Added a check for resumable_handlers
          before actually resuming the handle.

Thu Jun 28 18:25:03 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Transport.cpp (process_queue_head):
        * tao/Transport.h: Added a new method that would take the message
          from the head of the incoming queue and send it for processing.

        * tao/GIOP_Message_Base.cpp: Minor formatting.

Thu Jun 28 17:16:50 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/GIOP_Message_Base.cpp:
        * tao/GIOP_Message_Base.h:
        * tao/Pluggable_Messaging.h: Removed all the arguments that had an
          TAO_ORB_Core as one of the arguments. GIOP classes hold a copy
          of the pointer to the ORB_Core and so it is no
          necessary. Removed the following methods

          - is_message_complete ()
          - message_type ()

          The message_type () method is now local to the GIOP classes and
          it returns the Pluggable Message type from the
          GIOP_Message_State. This method is now used to fill the node
          with the right information type about the message.

        * tao/Transport.cpp: Used the information from the node of the
          queue to determine the type of message before processing it.

Thu Jun 28 15:12:32 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/GIOP_Message_Base.cpp:
        * tao/GIOP_Message_Base.i: Moved the get_queued_data () from the
          .i  file to the .cpp file.

        * tao/Incoming_Message_Queue.h: #include'd a file.
        * tao/Makefile: Added new files.
        * tao/Resume_Handle.inl: Added the ACE_INLINE macros to the
          methods defined in this file.

Thu Jun 28 09:30:43 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Transport.cpp:
        * tao/GIOP_Message_Base.h:
        * tao/GIOP_Message_Base.i:
        * tao/GIOP_Message_Base.cpp: Made lots of changes to deal with the
          following

          - to parse & queue up messages if we have read multiple messages.
          - to consolidate messages if we had read half of a message
          - to process a consolidated message.

         We now use the node of the Incoming Message Queue to share data
         between the Transport layer and the GIOP layer.

        * tao/Incoming_Message_Queue.h:
        * tao/Incoming_Message_Queue.cpp:
        * tao/Incoming_Message_Queue.inl: Made the TAO_Queued_Data as a
          seperate class. This helped using the node of the queue to be
          shared between the GIOP layer. Removed many of the useless
          methods like wr_ptr (), is_message_complete (), add_message ()
          etc. We have the following methods that are more meaningful

          - is_tail_complete ()
          - is_message_complete ()
          - dequeue_head ()
          - dequeue_tail ()
          - enqueue_tail ()

         Added a static method get_queued_data () to create a node in the
         Message Queue.

         Added the protocol version information to the Queued Data.  We
         also need the message type and that has also been added.

        * tao/Resume_Handle.h:
        * tao/Resume_Handle.cpp:
        * tao/Resume_Handle.inl: This is a utility class that is used to
          resume handlers. This works more or less similar to our
          ACE_GUARD macros, but uses a flag to keep track whether the
          handle has been resumed.

        * tao/IIOP_Connection_Handler.cpp: Installed the Resume_Handle in
          handle_input () methods.

        * tao/Pluggable_Messaging.h: Removed the method  byte_order () and
          added the methods consolidate_node (), get_message_data () and
          extract_next_message ().

        * tao/Wait_On_Read.cpp:
        * tao/Connection_Handler.cpp: Changes to keep in sync with the
          changes to the signature of handle_input_i () in Transport
          class.

        * tao/LIST_OF_TODO: Updated the list of TODO's.

Mon Jun 25 19:21:43 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/GIOP_Message_Base.cpp:
        * tao/GIOP_Message_Base.h:
        * tao/GIOP_Message_Base.i:
        * tao/GIOP_Message_State.inl:
        * tao/IIOP_Transport.cpp:
        * tao/Incoming_Message_Queue.cpp:
        * tao/Incoming_Message_Queue.h:
        * tao/Incoming_Message_Queue.inl:
        * tao/Pluggable_Messaging.h:
        * tao/Transport.cpp:
        * tao/Transport.h:  An inconsistent checkin only to transfer the
          files to the box at home.

Mon Jun 25 12:10:15 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Incoming_Message_Queue.{h,cpp,inl}:
        * tao/LIST_OF_TODO:
        * tao/GIOP_Message_State.inl: Added these to my branch.

Mon Jun 25 07:54:31 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/* : Merged the files from the main branch.

Mon Jun 25 07:45:38 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * Created this file.

Fri Jun 22 17:00:38 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/LIST_OF_TODO: Updated the list of TODO's.

Mon Jun 18 13:31:38 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Asynch_Reply_Dispatcher.cpp (dispatch_reply):
        * tao/Synch_Reply_Dispatcher.cpp:
        * tao/GIOP_Message_Base.cpp:
        * tao/Transport.cpp:
        * tao/DynamicInterface/DII_Reply_Dispatcher.cpp: Fixed warnings in
          g++ builds.

Sat Jun 17 17:46:23 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        This set of changes comes with complete revamping of the previous
        design. The flaws with the previous design were as follows
        (1) We were unnecessarily penalising large data blocks. We were
            trying to read a particular size of data till the data was
            completely removed from the socket. This was totally
            ridiculous because we were doing more reads than required.

        (2) The message block that was constructed on the stack with a
            buffer from stack never did what we wanted. It was allocating
            a data block on the heap and was thus spoiling whatever
            optimization that we had tried putting in.

        (3) The incoming message Queue is now managed by the TAO_Transport
            object instead of the GIOP classes.

        * tao/GIOP_Message_Base.cpp:
        * tao/GIOP_Message_Base.h: Removed the references to the incoming
          message queue. Implemention for two methods missing_data () and
          byte_order (). Added an extra argument to the methods
          process_request_message () and process_reply_message (). Used
          the incoming message block to create a input CDR with the
          DONT_DELETE flag so that the data block is not deleted after
          request processing.

        * tao/GIOP_Message_State.cpp: Removed the inclusion of
          Transport.h.

        * tao/Incoming_Message_Queue.h:
        * tao/Incoming_Message_Queue.cpp:
        * tao/Incoming_Message_Queue.inl: Added an argument to the
          add_message (). Further the implementation of add_message () has
          changed a bit. It now adds only a new message to the queue. It
          doesn't modify a half filled queue. The TAO_Transport object
          does that job. So declared the TAO_Transport as the friend class
          of the Incoming_Message_Queue.

          Changed the name of the methods complete_message () as
          is_complete_message (). Removed the methods current_message ()&
          current_byte_order ().

          Added a new method copy_message () which copies messages into
          the half empty nodes.

          Added a method wr_ptr () to access the write pointer of the tail
          node that has halfempty message.

        * tao/Transport.cpp:
        * tao/Transport.h: The Incoming Message Queue is now managed by
          this class. Added the following methods
          - missing_data ()
          - parse_incoming_messages ()
          - check_message_integrity ()
          - consolidate_message ()
          - conslodate_message_queue ()

        * tao/Pluggable_Messaging.h: Added two new virtual functions
          missing_data () and byte_order ().

        * tao/ORB_Core.h:
        * tao/ORB_Core.cpp:
        * tao/ORB_Core.i: Added an accessor for the locking_strategy used
          for the CDR blocks.

        * tao/Synch_Reply_Dispatcher.cpp:
        * tao/Asynch_Reply_Dispatcher.cpp:
        * tao/DynamicInterface/DII_Reply_Dispatcher.cpp: Changed the
          exchange_data_block () to clone_from () which is a new method in
          ACE_InputCDR.

        * tao/LIST_OF_TODO: Updated the list

Sat Jun 16 15:49:23 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Any.cpp:
        * tao/Asynch_Reply_Dispatcher.cpp:
        * tao/CDR.cpp:
        * tao/CDR.h:
        * tao/CDR.i:
        * tao/GIOP_Message_Lite.cpp:
        * tao/IIOP_Profile.cpp:
        * tao/Invocation.cpp:
        * tao/ORB.cpp:
        * tao/Pluggable_Messaging_Utils.cpp:
        * tao/Synch_Reply_Dispatcher.cpp:
        * tao/TAO_Server_Request.cpp:
        * tao/DynamicInterface/DII_Reply_Dispatcher.cpp: Integrated some
          of the changes from the main trunk in this branch.

Wed Jun 13 17:55:24 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Incoming_Message_Queue.cpp:
        * tao/Incoming_Message_Queue.inl: Fixed a bad way to use the tail
          of the circular linked list. This is now replaced by the size of
          the linked list. This fixes quite a few errors.

        * tao/Transport.cpp: On read () if we get errno == EWOULDBLOCK we
          were closing the connection prematurely. This is not right. We
          should only return a 0 to the reactor, so that it can call us
          back when there is data in the socket.

Wed Jun 13 10:45:24 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/GIOP_Message_Base.cpp (process_reply_message): Fixed the
          reply parsing and generation of the CDR stream that is passed on
          to the higher layers of the ORB.

Wed Jun 13 07:25:24 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Makefile: Added new files.
        * tao/Exclusive_TMS.cpp: Removed some commented code.

Tue Jun  12 18:42:55 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/GIOP_Message_State.h:
        * tao/GIOP_Message_State.inl: The message_size () now returns the
          payload size + the GIOP header length. Added a new method
          payload_size () that returns the payload size alone.

        * tao/GIOP_Message_Base.cpp: Fixed a warning with g++ builds.

        * tao/Connection_Handler.h:
        * tao/Connection_Handler.cpp: The svc_i () method now calls
          handle_input_i () on the transport instead of the method in the
          same class. The handle_input_i () in the TAO_Connection_Handler
          class has been removed as it is no longer used.

        * tao/IIOP_Connection_Handler.h:
        * tao/IIOP_Connection_Handler.cpp: Removed the implementation of
          handle_input_i ().

        * tao/Incoming_Message_Queue.cpp: When trying to make a new data
          block we dont add the size of GIOP header. The message_size ()
          now returns with that value.

        * tao/Incoming_Message_Queue.inl: Fixed a link error in g++
          builds.

Tue Jun  12 17:42:55 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        First set of checkins for big two ways.

        * tao/Connection_Handler.h: Added #define for the default
          buffer size . Not sure yet whether this is the right place for
          it.

        * tao/Incoming_Message_Queue.h:
        * tao/Incoming_Message_Queue.cpp:
        * tao/Incoming_Message_Queue.inl: Queue up the incoming
          messages. We form a circular linked list of messages that are
          bigger than a particular buffer size. During processing we take
          messages of the queue to pass it onto the higher layers of the
          ORB.

        * tao/GIOP_Message_Base.cpp:
        * tao/GIOP_Message_Base.h: Added support for big two way
          requests. If the message has only been partially read we add the
          message in the queue and then go for the next read. During
          processing we check if the queue has messages before processing
          the message on hand.

        * tao/GIOP_Message_State.h:
        * tao/GIOP_Message_State.cpp:
        * tao/GIOP_Message_State.inl: Added accessor methods.

        * tao/Transport.cpp: Added support for bigger two ways. If we
          receive a two way bigger than a particular size we read and
          queue the message and do a further read to retrieve the rest of
          the message.

        * tao/TAO.dsp: Added new files.
        * tao/default_resource.cpp: For single threaded configuration we
          have been creating a locked data block and that seems to defy
          logic. We now should create a lock free datablock for single
          threaded configurations.

Sat Jun  2 12:02:55 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/* : Merged with the main trunk.

Fri Jun  1 17:22:29 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/GIOP_Message_State.cpp (parse_message_header_i):
        * tao/Connection_Handler.cpp:
        * tao/GIOP_Message_Base.cpp:
        * tao/Makefile:
        * tao/Transport.cpp: Fixed warnings and errors in Linux g++.

Thu Jun 01 13:39:02 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Transport.cpp:
        * tao/Transport.h:  Added the following methods -- handle_input_i
          (), process_parsed_message (). The handle_input_i () creates a
          buffer on the stack. It uses the buffer to read the
          messages. Any errors in read () or dealt here directly. The read
          message is then parsed and processed by the GIOP classes. Moved
          most of the code for processing here as they seem to be common
          between different transport protocols. Once we have received a
          full message, we just resume the handler and go ahead with
          processing the message.

        * tao/IIOP_Connection_Handler.cpp (handle_input ()): Calls
          handle_input_i () on the transport.

        * tao/Connection_Handler.h:
        * tao/Connection_Handler.cpp:  The svc_i () calls the
          handle_input_i () on the transport instead of the same call on
          the connection handler. This way we should be able to share the
          same code among different protocol objects.

        * tao/GIOP_Message_State.h:
        * tao/GIOP_Message_State.cpp: Much of the message parsing and
          state information is stored here.

        * tao/GIOP_Message_Base.h:
        * tao/GIOP_Message_Base.cpp:
        * tao/Pluggable_Messaging.h: made some changes to suit the
          above. But more changes could come in, to suit different
          protocols. We may want to change the interfaces in
          Pluggable_Messaging to give a much cleaner interface to the
          world.

        * tao/Wait_On_Read.cpp: Used the handle_input_i () on the
          transport for processing the incoming data.

        * tao/TAO.dsp:
        * tao/PortableServer/TAO_PortableServer.dsp: Added new files and
          removed files from GIOP_Lite in the first round.

        * tao/GIOP_Message_Reactive_Handler.cpp:
        * tao/GIOP_Message_Reactive_Handler.h: Looks like these files
          would be removed during the final merge.

        The above checkins works for simple two way calls. The
        BiDirectional_NestedUpcall test works with the TP_Reactor. This
        works on Win32.


Thu May 24 12:19:02 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/Transport.h: Added some design forces for the input data
          path.

        End of ChangeLog entries from branch bug_575_stage_2.

Thu Jul  5 22:04:24 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tests/LongUpcalls/run_ami_test.pl: Adding a new perl script to
          this test. This is supposed to fail in the daily builds as this
          needs bug fix 575.

Thu Jul  5 21:16:00 2001  Craig Rodrigues <crodrigu@bbn.com>

         * orbsvcs/orbsvcs/AV/AV_Core.h

           Added some Borland compiler pragma options to prevent
           singleton problems when AV_Core is in a DLL.

Thu Jul  5 20:41:00 2001  Craig Rodrigues <crodrigu@bbn.com>

         * tao/CONV_FRAMEC.h
         * tao/DynamicC.h
         * tao/PolicyC.h
            Added export macros for some Unbounded Sequence types.

         * TAO_IDL/be/be_visitor_sequence/gen_unbounded_obj_sequence_ch.cpp
            Add export macro for unbounded object sequence types

Thu Jul  5 09:22:12 2001  Johnny Willemsen <jwillemsen@remedy.nl>

        * orbsvcs/tests/AVStreams/Makefile.bor
          Added  Asynch_Three_Stage
        * orbsvcs/tests/AVStreams/Asynch_Three_Stage/*.bor
          Added BCB makefiles

Thu Jul  5 04:53:00 2001  Craig Rodrigues <crodrigu@bbn.com>

        * orbsvcs/tests/AVStreams/Component_Switching/Makefile
          Add missing link flag -lTAO_Strategies, needed
          for Advanced_Resource_Factory in svc.conf.

        * orbsvcs/tests/AVStreams/Component_Switching/receiver.cpp
          Fix typo

        * orbsvcs/tests/AVStreams/Component_Switching/run_test.pl
          Add -ORBSkipServiceConfigOpen to startup of Naming_Service,
          it did not like the svc.conf file in this directory.

Wed Jul  4 09:53:01 2001  Frank Hunleth  <fhunleth@cs.wustl.edu>

        * tao/Strategies/TAO_Strategies_Static.dsp:

          Updated static build with DIOP files.

Wed Jul  4 09:52:20 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * tao/DynamicAny/DynCommon.cpp (insert_reference):

          Added a call to _is_a() as a final check to see if there
          is a type mismatch between the dynamic any and the
          object reference argument. Thanks to Jonathan Biggar
          <jon@floorboard.com> for clarifying the spec and to
          Philippe Merle <Philippe.Merle@lifl.fr> for suggesting
          optimizations.

Wed Jul  4 09:10:24 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * TAO_IDL/be/be_visitor_interface/tie_sh.cpp:
        * TAO_IDL/be/be_visitor_root/root_sth.cpp:

          Removed ACE_HAS_USING_KEYWORD guard from each TIE class
          declaration, and added it to the global module reopening
          where all TIE class declarations will occur. This
          includes the TAO_NAMESPACE macro inside the guard, which
          is what we want if the platform does not support
          namespaces.

Wed Jul  4 14:05:12 2001  Johnny Willemsen <jwillemsen@remedy.nl>

        * orbsvcs/tests/Makefile.bor
          Updated to include AVStreams

        * orbsvcs/tests/AVStreams/Component_Switching/*.bor
        * orbsvcs/tests/AVStreams/Simple_Three_Stage/*.bor
        * orbsvcs/tests/AVStreams/Simgle_Two_Stage/*.bor
          Updated these BCB makefiles so that these tests are now
          build without errors

Tue Jul  3 20:09:58 2001  Frank Hunleth  <fhunleth@cs.wustl.edu>

        * tao/Strategies/TAO_Strategies.dsp:

          Added DIOP source files to project.

Tue Jul  3 20:09:58 2001  Frank Hunleth  <fhunleth@cs.wustl.edu>

        * examples/PluggableUDP/tests/Basic/Makefile:
        * examples/PluggableUDP/tests/Performance/Makefile:
        * examples/PluggableUDP/tests/SimplePerformance/Makefile:
        * tao/Strategies/Makefile:

          Updated dependencies.

Tue Jul  3 18:25:41 2001  Frank Hunleth  <fhunleth@cs.wustl.edu>

        * examples/PluggableUDP/DIOP/*:

          Moved DIOP source files from PluggableUDP to Strategies.

        * examples/PluggableUDP/README:
        * examples/PluggableUDP/DIOP/README:

          Updated README files to note file location change.

        * examples/PluggableUDP/Makefile:
        * examples/PluggableUDP/Makefile.bor:

          Removed DIOP directory from build.

        * examples/PluggableUDP/tests/Basic/Makefile:
        * examples/PluggableUDP/tests/Basic/client.bor:
        * examples/PluggableUDP/tests/Basic/client.cpp:
        * examples/PluggableUDP/tests/Basic/client.dsp:
        * examples/PluggableUDP/tests/Basic/server.bor:
        * examples/PluggableUDP/tests/Basic/server.cpp:
        * examples/PluggableUDP/tests/Basic/server.dsp:
        * examples/PluggableUDP/tests/Basic/svc.conf:
        * examples/PluggableUDP/tests/Performance/Makefile:
        * examples/PluggableUDP/tests/Performance/client.bor:
        * examples/PluggableUDP/tests/Performance/client.cpp:
        * examples/PluggableUDP/tests/Performance/client.dsp:
        * examples/PluggableUDP/tests/Performance/server.bor:
        * examples/PluggableUDP/tests/Performance/server.cpp:
        * examples/PluggableUDP/tests/Performance/server.dsp:
        * examples/PluggableUDP/tests/Performance/svc.conf:
        * examples/PluggableUDP/tests/SimplePerformance/Makefile:
        * examples/PluggableUDP/tests/SimplePerformance/client.bor:
        * examples/PluggableUDP/tests/SimplePerformance/client.cpp:
        * examples/PluggableUDP/tests/SimplePerformance/client.dsp:
        * examples/PluggableUDP/tests/SimplePerformance/server.bor:
        * examples/PluggableUDP/tests/SimplePerformance/server.cpp:
        * examples/PluggableUDP/tests/SimplePerformance/server.dsp:
        * examples/PluggableUDP/tests/SimplePerformance/svc.conf:

          Updated source files and Makefiles to look for the DIOP
          code in the Strategies library rather than in the DIOP library.

        * tao/orbconf.h:

          Added #define to control whether DIOP is compiled in or not.

        * tao/Strategies/DIOP_Acceptor.cpp:
        * tao/Strategies/DIOP_Acceptor.h:
        * tao/Strategies/DIOP_Acceptor.i:
        * tao/Strategies/DIOP_Connection_Handler.cpp:
        * tao/Strategies/DIOP_Connection_Handler.h:
        * tao/Strategies/DIOP_Connection_Handler.i:
        * tao/Strategies/DIOP_Connector.cpp:
        * tao/Strategies/DIOP_Connector.h:
        * tao/Strategies/DIOP_Endpoint.cpp:
        * tao/Strategies/DIOP_Endpoint.h:
        * tao/Strategies/DIOP_Endpoint.i:
        * tao/Strategies/DIOP_Factory.cpp:
        * tao/Strategies/DIOP_Factory.h:
        * tao/Strategies/DIOP_Profile.cpp:
        * tao/Strategies/DIOP_Profile.h:
        * tao/Strategies/DIOP_Profile.i:
        * tao/Strategies/DIOP_Transport.cpp:
        * tao/Strategies/DIOP_Transport.h:
        * tao/Strategies/DIOP_Transport.i:

          Moved DIOP files to Strategies.  Updated linking specifiers and
          include files to reflect change.

        * tao/Strategies/Makefile:
        * tao/Strategies/Makefile.bor:

          Added DIOP files.

        * tao/Strategies/TAO_Strategies_Internal.cpp:
        * tao/Strategies/advanced_resource.cpp:

          Added service configurator hooks for DIOP.

Tue Jul  3 14:53:00 2001  Craig Rodrigues <crodrigu@bbn.com>

        * orbsvcs/orbsvcs/AV/AV_Core.h
        * orbsvcs/orbsvcs/AV/RTCP.h
        * orbsvcs/orbsvcs/AV/RTP.h
        * orbsvcs/orbsvcs/AV/Transport.h
        * orbsvcs/orbsvcs/AV/UDP.h

          Fix some includes, add TAO_AV_Export in more class
          declarations.

Tue Jul  3 11:27:03 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * TAO_IDL/be/be_codegen.cpp:
        * TAO_IDL/be/be_visitor_interface/tie_sh.cpp:
        * TAO_IDL/be/be_visitor_interface/tie_si.cpp:

          Removed the #if defined (ACE_HAS_USING_KEYWORD)
          guards from the top and bottom of the *S_T.{h,i}
          files and added their generation on a TIE class by
          TIE class basis, conditionally, if the original
          interface is defined inside a module. TIE class
          code will compile for interfaces declared at global
          scope, even on platforms that do not support
          namespaces. This enhancement was requested by
          Marco Kranawetter <Marco.Kranawetter@icn.siemens.de>.

Tue Jul  3 09:55:22 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * tao/DynamicAny/DynUnion_i.cpp:

          Added missing .in() to a CORBA::Any_var.

        * tao/DynamicAny/DynCommon.cpp:

          Removed unnecessary break statement.

Tue Jul  3 08:45:13 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * tests/Smart_Proxies/Benchmark/client.dsp:
        * tests/Smart_Proxies/Benchmark/server.dsp:

          Fixed IDL compiler settings in the release version.

Mon Jul  2 08:59:16 2001  Johnny Willemsen <jwillemsen@remedy.nl>

        * orbsvcs/tests/AVStreams/Makefile.bor
        Added BCB makefile

Mon Jul  2 23:13:00 2001  Craig Rodrigues <crodrigu@bbn.com>

        * orbsvcs/orbsvcs/AV/AVStreams_i.h
        * orbsvcs/orbsvcs/AV/Protocol_Factory.h
        * orbsvcs/orbsvcs/AV/Transport.h
        * orbsvcs/orbsvcs/AV/UDP.h

        Added TAO_AV_Export to more classes to fix Borland DLL builds.

        * orbsvcs/tests/AVStreams/Simple_Two_Stage/receiver.bor
        * orbsvcs/tests/AVStreams/Simple_Two_Stage/sender.bor

        Added more linker flags to fix Borland DLL builds.

Mon Jul  2 19:32:09 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * tests/Queued_Message_Test/Queued_Message_Test.cpp:

          Changed use of ACE_OS::rand_r(seed) to ACE_OS::srand(seed)
          and ACE_OS::rand(). ACE_OS::rand_r(seed) is not supported
          on Win32 platforms, and was causing the test to hang.

Mon Jul  2 17:06:22 2001  Balachandran Natarajan  <bala@cs.wustl.edu>

        * tao/ORB_Core.cpp: Initialization of the transport_cache ()
          should preceed the preconnects (). Thanks to Irfan for pointing
          this out.

Mon Jul  2 15:53:29 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * orbsvcs/tests/InterfaceRepo/IFR_Test/Admin_Client.cpp:
        * tao/Typecode.cpp:
        * tao/append.cpp:
        * tao/skip.cpp:
        * tao/DynamicAny/DynUnion_i.cpp:

          Modified CORBA::TypeCode::member_label() to return a duplicate
          Any pointer (to be comsumed by the caller), making it consistent
          with other ORB functions that return pseudo-object types. Also
          modified all uses of member_label() in TAO to prevent memory
          leaks. Thanks to Philippe Merle <Philippe.Merle@lifl.fr> for
          pointing out this inconsistency, and to Jonanthan Biggar
          <jon@floorboard.com> for clarifying the spec with regard to
          memory management of pseudo-objects.

Mon Jul  2 14:29:49 2001  Frank Hunleth  <fhunleth@cs.wustl.edu>

        * performance-tests/RTCorba/Multiple_Endpoints/Single_Endpoint/Makefile:
        * performance-tests/RTCorba/Multiple_Endpoints/Orb_Per_Priority/Makefile:

          Removed duplicate ACE and TAO library includes.  Should fix KCC
          linker warnings.

Mon Jul 2  14:50:00 2001  Craig Rodrigues <crodrigu@bbn.com>

        * orbsvcs/AV/AVStreams_i.cpp

          Only display certain debug messages when higher ORB debug level
          is specified.

Mon Jul  2 13:37:45 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * TAO_IDL/fe/idl.ll:
        * TAO_IDL/fe/lex.yy.cpp:

          Added '\x' to the list of escaped characters in the
          regular expression for strings in the TAO IDL lexer,
          and regenerated the C++ file. Legal IDL strings such
          as "\xA" were producing syntax errors. Thanks to
          Jules Colding <dsl11814@vip.cybercity.dk> for reporting
          this bug.

Mon Jul  2 13:18:14 2001  Douglas C. Schmidt  <schmidt@tango.doc.wustl.edu>

        * Moved all the ChangeLog-* files into the new ChangeLogs
          directory.  Thanks to Holger P . Krekel <krekel@merlinux.de> for
          this suggestion.

Mon Jul  2 10:57:26 2001  Jeff Parsons <parsons@cs.wustl.edu>

        * tests/Smart_Proxies/Benchmark/Benchmark.dsw:
        * tests/Smart_Proxies/Benchmark/client.dsp:
        * tests/Smart_Proxies/Benchmark/server.dsp:

          New files to build the smart proxies benchmark test on
          Win32 platforms.

        * tests/Smart_Proxies/Benchmark/client.cpp:
        * tests/Smart_Proxies/Benchmark/Smart_Proxy_Impl.cpp:
        * tests/Smart_Proxies/Benchmark/Smart_Proxy_Impl.h:

          Removed unnecessary TAO_HAS_SMART_PROXIES macro from
          these files.

        * docs/Smart_Proxies.html:

          Updated documentation.