diff options
Diffstat (limited to 'trunk/TAO/ChangeLogs/ChangeLog-06b')
| -rw-r--r-- | trunk/TAO/ChangeLogs/ChangeLog-06b | 12577 |
1 files changed, 12577 insertions, 0 deletions
diff --git a/trunk/TAO/ChangeLogs/ChangeLog-06b b/trunk/TAO/ChangeLogs/ChangeLog-06b new file mode 100644 index 00000000000..bbe4135e861 --- /dev/null +++ b/trunk/TAO/ChangeLogs/ChangeLog-06b @@ -0,0 +1,12577 @@ +Mon Nov 20 10:10:00 UTC 2006 Simon Massey <sma@prismtech.com> + + * NEWS: + + Slight tidy-up, to keep release notes / NEWS in step. + +Fri Nov 17 09:08:00 UTC 2006 Simon Massey <sma@prismtech.com> + + * NEWS: + * orbsvcs/tests/ImplRepo/Bug_2604_Regression/MessengerServer.cpp + + Slight tidy-up, to keep release notes / NEWS in step. + Wallace Zhang's fix for shutdown run-time windows test problem. + +Thu Nov 16 10:20:00 UTC 2006 Simon Massey <sma@prismtech.com> + + * tao/IIOP_Profile.cpp: + + Committed for Phil Mesnier. Patch so that Bug_2604_Regression + test passes on windows. The problem was that an IIOP_Profile + that is decoded from a CDR buffer does not correctly set the + is_ipv6_decimal_ flag for the first endpoint only. After the + beta is out more changes have to be made to remove the friend + relationships and cleanup more code. Need to also update + tests/IPV6 so that we have a regression in the repo for this + problem. Thanks to Johnny Willemsen and Martin Corino for + quickly testing this fix. + +Thu Nov 16 00:07:30 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * tao/Makefile.am: + + Regenerated. + + * tao/DynamicAny.mpc: + + Changed to inherit from core.mpb, so that library and headers + will be installed for automake build. + +Wed Nov 15 16:38:29 UTC 2006 Chris Cleeland <cleeland_c@ociweb.com> + + * orbsvcs/orbsvcs/Event/EC_Thread_Flags.h: + * orbsvcs/orbsvcs/Event/EC_Thread_Flags.cpp: + + Removed code that had been in #if 0/#endif, which had been part + of an earlier, more complete design but more difficult to + implement. + + Eliminated #include of non-existent inline file that gave the + fuzz build heartburn. + +Wed Nov 15 10:35:00 UTC 2006 Simon Massey <sma@prismtech.com> + + * NEWS: + + Added entries. + +Wed Nov 15 10:05:00 UTC 2006 Simon Massey <sma@prismtech.com> + + * tests/Oneway_Timeouts/client.cpp: + + Windows warnings removed (for Phil Mesnier). + +Wed Nov 15 08:04:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * docs/Options.html: + + Added ORBServiceConfigLoggerKey, thanks to Doug and Abdul for + finding out this was lacking. + +Wed Nov 15 04:00:22 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * examples/RTScheduling/Fixed_Priority_Scheduler/Makefile.am: + * examples/RTScheduling/MIF_Scheduler/Makefile.am: + * examples/RTScheduling/Makefile.am: + + Regenerated. + +Wed Nov 15 02:04:44 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * tao/Makefile.am: + + Regenerated. + + * tao/tao.mpc: + + Add Header_Files section containing all the fooA.h headers + generated from *.pidl files. These aren't automagically + added (possibly because of the use of MPC's >> construct), + which results in these headers not being installed in + automake builds. + +Tue Nov 14 16:17:04 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * examples/RTScheduling/RTSchedTestLib.mpc: + * examples/RTScheduling/rtschedtestlib_export.h: + + New files, representing a new build created to hold common + code factored out from the other builds. Motivated not only + by good design, but by necessity on our scoreboard + platforms with parallel builds, where strict dependency + ordering in required. + + * examples/RTScheduling/Thread_Task.h: + * examples/RTScheduling/Synch_i.h: + * examples/RTScheduling/Job.mpc: + * examples/RTScheduling/POA_Holder.cpp: + * examples/RTScheduling/Fixed_Priority_Scheduler\FP_DT_Creator.cpp: + * examples/RTScheduling/Fixed_Priority_Scheduler\FP_DT_Creator.h: + * examples/RTScheduling/Fixed_Priority_Scheduler\Fixed_Priority_Scheduler.mpc: + * examples/RTScheduling/POA_Holder.h: + * examples/RTScheduling/Synch.mpc: + * examples/RTScheduling/Job_i.cpp: + * examples/RTScheduling/Starter.cpp: + * examples/RTScheduling/Task_Stats.h: + * examples/RTScheduling/DT_Creator.cpp: + * examples/RTScheduling/Job_i.h: + * examples/RTScheduling/MIF_Scheduler\MIF_Scheduler.mpc: + * examples/RTScheduling/MIF_Scheduler\MIF_Scheduler.h: + * examples/RTScheduling/MIF_Scheduler\test.cpp: + * examples/RTScheduling/MIF_Scheduler\test.h: + * examples/RTScheduling/MIF_Scheduler\MIF_Task.cpp: + * examples/RTScheduling/Starter.h: + * examples/RTScheduling/Thread_Task.cpp: + * examples/RTScheduling/DT_Creator.h: + + Factored files and dependencies out of existing MPC files to + be consistent with the new MPC file above. Also, in C++ + files, re-ordered includes, added export macros, and made + other cosmetic changes. + +Tue Nov 14 15:23:43 UTC 2006 Chad Elliott <elliott_c@ociweb.com> + + * NEWS: + + Updated this file with information about my two bug fixes. + +Tue Nov 14 13:49:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Wait_Strategy.cpp: + * tao/Wait_On_Leader_Follower.cpp: + * tao/default_client.cpp: + Added some missing markers for FOCUS to work + +Tue Nov 14 09:00:00 UTC 2006 Simon Massey <sma@prismtech.com> + + * NEWS: + + Submitted by Jeff Parsons. + +Tue Nov 14 04:59:41 UTC 2006 Chris Cleeland <cleeland_c@ociweb.com> + + * orbsvcs/orbsvcs/Event/EC_Thread_Flags.cpp: + + Correct problems related to missing string functions on + non-inline builds. + +Mon Nov 13 22:40:11 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * orbsvcs/tests/Event/ThreadFlags/Makefile.am: + + Regenerated. + +Mon Nov 13 22:36:51 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * examples/RTScheduling/Fixed_Priority_Scheduler/Makefile.am: + * examples/RTScheduling/MIF_Scheduler/Makefile.am: + * examples/RTScheduling/Makefile.am: + + Regenerated. + +Mon Nov 13 20:01:51 UTC 2006 Chris Cleeland <cleeland_c@ociweb.com> + + * orbsvcs/orbsvcs/Event/EC_Thread_Flags.h: + * orbsvcs/orbsvcs/Event/EC_Thread_Flags.cpp: + * orbsvcs/tests/Event/ThreadFlags/testflags.cpp: + + Various fixes to enable Versioned Namespace builds to complete. + +Mon Nov 13 19:22:06 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * examples/RTScheduling/Job.mpc: + * examples/RTScheduling/job_export.h: + * examples/RTScheduling/Synch.mpc: + * examples/RTScheduling/synch_export.h: + + New files, factored out IDL files common to some or all + of the examples in this directory into separate builds, + to eliminate problems with parallel builds when IDL + files are repeated in builds in the same directory. + + * examples/RTScheduling/Fixed_Priority_Scheduler/Fixed_Priority_Scheduler.mpc: + * examples/RTScheduling/MIF_Scheduler/MIF_Scheduler.mpc: + + Modified these files to make them consistent with the + changes above. + +Mon Nov 13 15:23:05 UTC 2006 Chris Cleeland <cleeland_c@ociweb.com> + + * orbsvcs/orbsvcs/Event/EC_TPC_Dispatching.cpp (TAO_EC_TPC_debug_level): + + Moved declaration to fix versioned namespace builds. + + * orbsvcs/orbsvcs/Event/EC_Thread_Flags.h: + * orbsvcs/orbsvcs/Event/EC_Thread_Flags.cpp: + + Add in versioned namespace boundary macros to fix versioned + namespace builds. + +Mon Nov 13 15:21:58 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tao/default_resource.h: + + It turns out, exporting of the TAO_Codeset_Parameters is indeed + necessary (the advanced resource factory needs it, + too). Partially reverting "Fri Nov 10 21:09:53 UTC 2006". + +Sat Nov 11 17:19:31 UTC 2006 Chris Cleeland <cleeland_c@ociweb.com> + + * orbsvcs/tests/Event/ThreadFlags/testflags.cpp: Remove extraneous + semicolon after ACE_RCSID. There should be a fuzz check for + this. + +Sat Nov 11 09:31:30 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * orbsvcs/tests/Event/ThreadFlags/tftest.mpc: + + Fix broken dependency by inheriting from rtevent_serv.mpb + instead of appending TAO_RTEvent_Serv to libs. + +Fri Nov 10 21:09:53 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tao/default_resource.h: + + Cleaning up the code that fixes bug #2651 to fix typos and + eliminate unnecessary exports. (Thanks Johnny!) + +Fri Nov 10 15:51:29 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tao/ORB_Core.cpp (fini, ~TAO_ORB_Core): + + Calling TAO::ORB::close_services() prior to the dtor may cause + non-yet-destroyed-heap-allocated objects to end up in unmapped + memory (like the policy_factory_registry_, for example). This + can happen if the dynamic service object, which allocated them + gets unloaded before the core dtor deletes them. The change + ensures the core sevice destruction (close_services) is + postponed until everything else the core owns has been deleted. + +Fri Nov 10 09:08:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Wait_On_Read.cpp: + * tao/SystemException.cpp: + * tao/Synch_Invocation.cpp: + * tao/Base_Transport_Property.cpp: + Const improvements + + * tao/Stub.cpp: + Removed TAO_Synchronous_Cancellation_Required, not used and + const improvements + +Fri Nov 10 08:53:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Unbounded_Octet_Sequence_T.h: + Export this template specialization to resolve application link + errors with Intel 9.0 + +Fri Nov 10 01:29:39 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * orbsvcs/tests/CosEvent/Makefile.am: + * orbsvcs/tests/Event/Makefile.am: + * orbsvcs/tests/Event/Mcast/Makefile.am: + + Regenerated to recurse into new test directories. + + * configure.ac: + + Configure Makefiles in new test directories. + + * orbsvcs/tests/CosEvent/Timeout/Makefile.am: + * orbsvcs/tests/Event/Mcast/RTEC_MCast_Federated/Makefile.am: + * orbsvcs/tests/Event/ThreadFlags/Makefile.am: + + New files. Makefile.am's for new test directories. + + * orbsvcs/tests/Event/Mcast/RTEC_MCast_Federated/RTEC_MCast_Federated.mpc: + Add automake specific section so that includes is set relative + to $(srcdir). Fix includes so it is correct on case-sensitive + filesystems. + +Thu Nov 9 22:34:28 UTC 2006 Chris Cleeland <cleeland_c@ociweb.com> + + * orbsvcs/orbsvcs/Event/EC_Thread_Flags.cpp: + + Removed semicolon after ACE_RCSID that caused problems with some + compilers, but not others, and caused a domino effect of no + RTEC-related stuff to build. + +Thu Nov 9 19:27:16 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * orbsvcs/orbsvcs/Makefile.am: + + Regenerated (again, to catch changes made while I was preparing + the last set of changes). + +Thu Nov 9 18:52:15 UTC 2006 Chris Cleeland <cleeland_c@ociweb.com> + + * NEWS: Updated to reflect RTEC changes. + +Thu Nov 9 15:42:32 UTC 2006 Chris Cleeland <cleeland_c@ociweb.com> + + * docs/ec_options.html: Add documentation for TPC and + -ECDispatchingThreadFlags options. + + * orbsvcs/orbsvcs/Event/EC_Default_Factory.h: + * orbsvcs/orbsvcs/Event/EC_Default_Factory.cpp: + + Added new option, -ECDispatchingThreadsFlags, that allows the + user to pass in a list of thread creation flags and priority for + dispatching threads. These can be used for either the MT + dispatching strategy or the TPC dispatching strategy. + + Also added -ECDebug option to enable debugging output from the + RTEC. Only the option and variable was added, but no messages. + Therefore, at the moment, this does not generate much output. + + * orbsvcs/orbsvcs/Event/EC_Factory.h: + * orbsvcs/orbsvcs/Event/EC_Factory.cpp: + + Added support for RTEC's debug level. + + * orbsvcs/orbsvcs/Event/EC_TPC_Dispatching.h: + * orbsvcs/orbsvcs/Event/EC_TPC_Dispatching.cpp: + + (CTOR) Changed so that thread flags can be passed in just like + with the MT dispatching strategy. + + (add_consumer) Modified so that it passes the thread flags set + in the CTOR to the dispatching task's activate call. + + * orbsvcs/orbsvcs/Event/EC_TPC_Factory.cpp (create_dispatching): + + Update to reflect new CTOR args for EC_TPC_Dispatching. + + * orbsvcs/orbsvcs/Event/EC_TPC_Factory.h: + * orbsvcs/orbsvcs/Event/EC_TPC_Factory.cpp: + * orbsvcs/orbsvcs/Event/EC_TPC_ProxyConsumer.cpp: + * orbsvcs/orbsvcs/Event/EC_TPC_ProxySupplier.inl: + * orbsvcs/orbsvcs/Event/EC_TPC_ProxySupplier.cpp: + + Change references to EC_TPC_debug_level to + TAO_EC_TPC_debug_level for proper scoping. + + * orbsvcs/orbsvcs/Event/EC_Thread_Flags.h: + * orbsvcs/orbsvcs/Event/EC_Thread_Flags.cpp: + + Added new abstraction around thread flags. This also handles + parsing of text representation of THR_* options. Ideally this + is part of ACE, but such a low-level change is difficult to push + through in general, much less in a patch. + + * orbsvcs/orbsvcs/RTEvent_Serv.mpc: + + Updated with new files. + + * orbsvcs/tests/Event/ThreadFlags/testflags.cpp: + * orbsvcs/tests/Event/ThreadFlags/tftest.mpc: + + Added unit test for TAO_EC_Thread_Flags abstraction. + Unfortunately, due to the way that some of the thread flags + are encoded on some platforms, e.g., Linux, the test can + give false failing indications. I've left this as-is for + now because I haven't decided if this is a flaw in the test, + a flaw in the TAO_EC_Thread_Flags abstraction, or a flaw in + the way ACE assigns values when a THR_ symbol is not provided + by the platform itself. + +Thu Nov 9 12:17:10 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * docs/tutorials/Quoter/AMI/Makefile.am: + * docs/tutorials/Quoter/Event_Service/Makefile.am: + * docs/tutorials/Quoter/Makefile.am: + * docs/tutorials/Quoter/On_Demand_Activation/Makefile.am: + * docs/tutorials/Quoter/RT_Event_Service/Makefile.am: + * docs/tutorials/Quoter/Simple/ImprovedServer/Makefile.am: + * docs/tutorials/Quoter/idl/Makefile.am: + * examples/AMH/Sink_Server/Makefile.am: + * examples/AMI/FL_Callback/Makefile.am: + * examples/Buffered_AMI/Makefile.am: + * examples/Buffered_Oneways/Makefile.am: + * examples/CSD_Strategy/ThreadPool/Makefile.am: + * examples/CSD_Strategy/ThreadPool2/Makefile.am: + * examples/CSD_Strategy/ThreadPool3/Makefile.am: + * examples/CSD_Strategy/ThreadPool4/Makefile.am: + * examples/CSD_Strategy/ThreadPool5/Makefile.am: + * examples/CSD_Strategy/ThreadPool6/Makefile.am: + * examples/Callback_Quoter/Makefile.am: + * examples/Content_Server/AMI_Iterator/Makefile.am: + * examples/Content_Server/AMI_Observer/Makefile.am: + * examples/Content_Server/SMI_Iterator/Makefile.am: + * examples/Event_Comm/Makefile.am: + * examples/Kokyu_dsrt_schedulers/Makefile.am: + * examples/Kokyu_dsrt_schedulers/fp_example/Makefile.am: + * examples/Kokyu_dsrt_schedulers/mif_example/Makefile.am: + * examples/Kokyu_dsrt_schedulers/muf_example/Makefile.am: + * examples/Load_Balancing/Makefile.am: + * examples/Load_Balancing_persistent/Makefile.am: + * examples/Logging/Makefile.am: + * examples/Makefile.am: + * examples/OBV/Typed_Events/Makefile.am: + * examples/POA/Adapter_Activator/Makefile.am: + * examples/POA/DSI/Makefile.am: + * examples/POA/Default_Servant/Makefile.am: + * examples/POA/Explicit_Activation/Alt_Resources/Makefile.am: + * examples/POA/Explicit_Activation/Makefile.am: + * examples/POA/FindPOA/Makefile.am: + * examples/POA/Forwarding/Makefile.am: + * examples/POA/Generic_Servant/Makefile.am: + * examples/POA/Loader/Makefile.am: + * examples/POA/NewPOA/Makefile.am: + * examples/POA/On_Demand_Activation/Makefile.am: + * examples/POA/On_Demand_Loading/Makefile.am: + * examples/POA/POA_BiDir/Makefile.am: + * examples/POA/Reference_Counted_Servant/Makefile.am: + * examples/POA/RootPOA/Makefile.am: + * examples/POA/TIE/Makefile.am: + * examples/Persistent_Grid/Makefile.am: + * examples/PluggableUDP/tests/Basic/Makefile.am: + * examples/PluggableUDP/tests/Performance/Makefile.am: + * examples/PluggableUDP/tests/SimplePerformance/Makefile.am: + * examples/Quoter/Makefile.am: + * examples/RTCORBA/Activity/Makefile.am: + * examples/RTScheduling/Fixed_Priority_Scheduler/Makefile.am: + * examples/RTScheduling/MIF_Scheduler/Makefile.am: + * examples/Simple/bank/Makefile.am: + * examples/Simple/chat/Makefile.am: + * examples/Simple/echo/Makefile.am: + * examples/Simple/grid/Makefile.am: + * examples/Simple/time-date/Makefile.am: + * examples/Simple/time/Makefile.am: + * examples/Simulator/Event_Supplier/Makefile.am: + * examples/TypeCode_Creation/Makefile.am: + * examples/ior_corbaloc/Makefile.am: + * examples/mfc/Makefile.am: + * interop-tests/wchar/Makefile.am: + * orbsvcs/Concurrency_Service/Makefile.am: + * orbsvcs/CosEvent_Service/Makefile.am: + * orbsvcs/Dump_Schedule/Makefile.am: + * orbsvcs/Event_Service/Makefile.am: + * orbsvcs/FTRT_Event_Service/Event_Service/Makefile.am: + * orbsvcs/FTRT_Event_Service/Factory_Service/Makefile.am: + * orbsvcs/FTRT_Event_Service/Gateway_Service/Makefile.am: + * orbsvcs/FT_ReplicationManager/Makefile.am: + * orbsvcs/Fault_Detector/Makefile.am: + * orbsvcs/Fault_Notifier/Makefile.am: + * orbsvcs/IFR_Service/Makefile.am: + * orbsvcs/ImplRepo_Service/Makefile.am: + * orbsvcs/LifeCycle_Service/Makefile.am: + * orbsvcs/LoadBalancer/Makefile.am: + * orbsvcs/Logging_Service/Basic_Logging_Service/Makefile.am: + * orbsvcs/Logging_Service/Event_Logging_Service/Makefile.am: + * orbsvcs/Logging_Service/Notify_Logging_Service/Makefile.am: + * orbsvcs/Logging_Service/RTEvent_Logging_Service/Makefile.am: + * orbsvcs/Naming_Service/Makefile.am: + * orbsvcs/Notify_Service/Makefile.am: + * orbsvcs/Scheduling_Service/Makefile.am: + * orbsvcs/TAO_Service/Makefile.am: + * orbsvcs/Time_Service/Makefile.am: + * orbsvcs/Trading_Service/Makefile.am: + * orbsvcs/examples/CosEC/Factory/Makefile.am: + * orbsvcs/examples/CosEC/RtEC_Based/bin/Makefile.am: + * orbsvcs/examples/CosEC/RtEC_Based/lib/Makefile.am: + * orbsvcs/examples/CosEC/RtEC_Based/tests/Basic/Makefile.am: + * orbsvcs/examples/CosEC/RtEC_Based/tests/Multiple/Makefile.am: + * orbsvcs/examples/CosEC/Simple/Makefile.am: + * orbsvcs/examples/CosEC/TypedSimple/Makefile.am: + * orbsvcs/examples/FaultTolerance/RolyPoly/Makefile.am: + * orbsvcs/examples/LoadBalancing/Makefile.am: + * orbsvcs/examples/Log/Basic/Makefile.am: + * orbsvcs/examples/Log/Event/Makefile.am: + * orbsvcs/examples/Log/Notify/Makefile.am: + * orbsvcs/examples/Log/RTEvent/Makefile.am: + * orbsvcs/examples/Notify/Federation/Agent/Makefile.am: + * orbsvcs/examples/Notify/Federation/Gate/Makefile.am: + * orbsvcs/examples/Notify/Federation/SpaceCraft/Makefile.am: + * orbsvcs/examples/Notify/Filter/Makefile.am: + * orbsvcs/examples/Notify/Lanes/Makefile.am: + * orbsvcs/examples/Notify/Subscribe/Makefile.am: + * orbsvcs/examples/Notify/ThreadPool/Makefile.am: + * orbsvcs/examples/ORT/Makefile.am: + * orbsvcs/examples/RtEC/IIOPGateway/Makefile.am: + * orbsvcs/examples/RtEC/Kokyu/Makefile.am: + * orbsvcs/examples/RtEC/MCast/Makefile.am: + * orbsvcs/examples/RtEC/Schedule/Makefile.am: + * orbsvcs/examples/RtEC/Simple/Makefile.am: + * orbsvcs/examples/Security/Send_File/Makefile.am: + * orbsvcs/orbsvcs/Makefile.am: + * orbsvcs/performance-tests/LoadBalancing/LBPerf/RPS/Makefile.am: + * orbsvcs/performance-tests/RTEvent/Colocated_Roundtrip/Makefile.am: + * orbsvcs/performance-tests/RTEvent/Federated_Roundtrip/Makefile.am: + * orbsvcs/performance-tests/RTEvent/RTCORBA_Baseline/Makefile.am: + * orbsvcs/performance-tests/RTEvent/RTCORBA_Callback/Makefile.am: + * orbsvcs/performance-tests/RTEvent/Roundtrip/Makefile.am: + * orbsvcs/performance-tests/RTEvent/TCP_Baseline/Makefile.am: + * orbsvcs/performance-tests/RTEvent/lib/Makefile.am: + * orbsvcs/tests/AVStreams/Asynch_Three_Stage/Makefile.am: + * orbsvcs/tests/AVStreams/Bidirectional_Flows/Makefile.am: + * orbsvcs/tests/AVStreams/Component_Switching/Makefile.am: + * orbsvcs/tests/AVStreams/Full_Profile/Makefile.am: + * orbsvcs/tests/AVStreams/Latency/Makefile.am: + * orbsvcs/tests/AVStreams/Modify_QoS/Makefile.am: + * orbsvcs/tests/AVStreams/Multicast/Makefile.am: + * orbsvcs/tests/AVStreams/Multicast_Full_Profile/Makefile.am: + * orbsvcs/tests/AVStreams/Multiple_Flows/Makefile.am: + * orbsvcs/tests/AVStreams/Pluggable/Makefile.am: + * orbsvcs/tests/AVStreams/Pluggable_Flow_Protocol/Makefile.am: + * orbsvcs/tests/AVStreams/Simple_Three_Stage/Makefile.am: + * orbsvcs/tests/AVStreams/Simple_Two_Stage/Makefile.am: + * orbsvcs/tests/AVStreams/Simple_Two_Stage_With_QoS/Makefile.am: + * orbsvcs/tests/BiDir_CORBALOC/Makefile.am: + * orbsvcs/tests/Bug_1334_Regression/Makefile.am: + * orbsvcs/tests/Bug_1393_Regression/Makefile.am: + * orbsvcs/tests/Bug_1395_Regression/Makefile.am: + * orbsvcs/tests/Bug_1630_Regression/Makefile.am: + * orbsvcs/tests/Bug_2074_Regression/Makefile.am: + * orbsvcs/tests/Bug_2137_Regression/Makefile.am: + * orbsvcs/tests/Bug_2247_Regression/Makefile.am: + * orbsvcs/tests/Bug_2248_Regression/Makefile.am: + * orbsvcs/tests/Bug_2285_Regression/Makefile.am: + * orbsvcs/tests/Bug_2287_Regression/Makefile.am: + * orbsvcs/tests/Bug_2316_Regression/Makefile.am: + * orbsvcs/tests/Bug_2377_Regression/Makefile.am: + * orbsvcs/tests/Concurrency/Makefile.am: + * orbsvcs/tests/CosEvent/Basic/Makefile.am: + * orbsvcs/tests/CosEvent/Makefile.am: + * orbsvcs/tests/CosEvent/lib/Makefile.am: + * orbsvcs/tests/EC_Custom_Marshal/Makefile.am: + * orbsvcs/tests/EC_MT_Mcast/Makefile.am: + * orbsvcs/tests/EC_Mcast/Makefile.am: + * orbsvcs/tests/EC_Multiple/Makefile.am: + * orbsvcs/tests/EC_Throughput/Makefile.am: + * orbsvcs/tests/Event/Basic/Makefile.am: + * orbsvcs/tests/Event/Makefile.am: + * orbsvcs/tests/Event/Mcast/Common/Makefile.am: + * orbsvcs/tests/Event/Mcast/Complex/Makefile.am: + * orbsvcs/tests/Event/Mcast/Makefile.am: + * orbsvcs/tests/Event/Mcast/Simple/Makefile.am: + * orbsvcs/tests/Event/Mcast/Two_Way/Makefile.am: + * orbsvcs/tests/Event/Performance/Makefile.am: + * orbsvcs/tests/Event/lib/Makefile.am: + * orbsvcs/tests/FT_App/Makefile.am: + * orbsvcs/tests/FaultTolerance/GroupRef_Manipulation/Makefile.am: + * orbsvcs/tests/FaultTolerance/IOGR/Makefile.am: + * orbsvcs/tests/FaultTolerance/IOGRManipulation/Makefile.am: + * orbsvcs/tests/FtRtEvent/Makefile.am: + * orbsvcs/tests/HTIOP/AMI/Makefile.am: + * orbsvcs/tests/HTIOP/BiDirectional/Makefile.am: + * orbsvcs/tests/HTIOP/Hello/Makefile.am: + * orbsvcs/tests/IOR_MCast/Makefile.am: + * orbsvcs/tests/ImplRepo/Makefile.am: + * orbsvcs/tests/ImplRepo/NameService/Makefile.am: + * orbsvcs/tests/ImplRepo/scale/Makefile.am: + * orbsvcs/tests/InterfaceRepo/Application_Test/Makefile.am: + * orbsvcs/tests/InterfaceRepo/IDL3_Test/Makefile.am: + * orbsvcs/tests/InterfaceRepo/IFR_Inheritance_Test/Makefile.am: + * orbsvcs/tests/InterfaceRepo/IFR_Test/Makefile.am: + * orbsvcs/tests/InterfaceRepo/Latency_Test/Makefile.am: + * orbsvcs/tests/InterfaceRepo/Persistence_Test/Makefile.am: + * orbsvcs/tests/Interoperable_Naming/Makefile.am: + * orbsvcs/tests/LoadBalancing/GenericFactory/Application_Controlled/Makefile.am: + * orbsvcs/tests/LoadBalancing/GenericFactory/Infrastructure_Controlled/Makefile.am: + * orbsvcs/tests/LoadBalancing/GenericFactory/Manage_Object_Group/Makefile.am: + * orbsvcs/tests/LoadBalancing/LoadMonitor/CPU/Makefile.am: + * orbsvcs/tests/Log/Basic_Log_Test/Makefile.am: + * orbsvcs/tests/Makefile.am: + * orbsvcs/tests/Miop/McastHello/Makefile.am: + * orbsvcs/tests/Notify/Basic/Makefile.am: + * orbsvcs/tests/Notify/Blocking/Makefile.am: + * orbsvcs/tests/Notify/Destroy/Makefile.am: + * orbsvcs/tests/Notify/Discarding/Makefile.am: + * orbsvcs/tests/Notify/Driver/Makefile.am: + * orbsvcs/tests/Notify/MT_Dispatching/Makefile.am: + * orbsvcs/tests/Notify/Ordering/Makefile.am: + * orbsvcs/tests/Notify/PluggableTopology/Makefile.am: + * orbsvcs/tests/Notify/RT_lib/Makefile.am: + * orbsvcs/tests/Notify/Reconnecting/Makefile.am: + * orbsvcs/tests/Notify/Sequence_Multi_ETCL_Filter/Makefile.am: + * orbsvcs/tests/Notify/Sequence_Multi_Filter/Makefile.am: + * orbsvcs/tests/Notify/Structured_Filter/Makefile.am: + * orbsvcs/tests/Notify/Structured_Multi_Filter/Makefile.am: + * orbsvcs/tests/Notify/Test_Filter/Makefile.am: + * orbsvcs/tests/Notify/XML_Persistence/Makefile.am: + * orbsvcs/tests/Notify/lib/Makefile.am: + * orbsvcs/tests/Notify/performance-tests/Filter/Makefile.am: + * orbsvcs/tests/Notify/performance-tests/RedGreen/Makefile.am: + * orbsvcs/tests/Notify/performance-tests/Throughput/Makefile.am: + * orbsvcs/tests/Property/Makefile.am: + * orbsvcs/tests/Redundant_Naming/Makefile.am: + * orbsvcs/tests/Sched/Makefile.am: + * orbsvcs/tests/Sched_Conf/Makefile.am: + * orbsvcs/tests/Security/BiDirectional/Makefile.am: + * orbsvcs/tests/Security/Big_Request/Makefile.am: + * orbsvcs/tests/Security/Callback/Makefile.am: + * orbsvcs/tests/Security/Crash_Test/Makefile.am: + * orbsvcs/tests/Security/MT_IIOP_SSL/Makefile.am: + * orbsvcs/tests/Security/MT_SSLIOP/Makefile.am: + * orbsvcs/tests/Security/Secure_Invocation/Makefile.am: + * orbsvcs/tests/Security/ssliop_corbaloc/Makefile.am: + * orbsvcs/tests/Simple_Naming/Makefile.am: + * orbsvcs/tests/Time/Makefile.am: + * orbsvcs/tests/Trading/Makefile.am: + * orbsvcs/tests/ior_corbaname/Makefile.am: + * orbsvcs/tests/tests_svc_loader/Makefile.am: + * performance-tests/Anyop/Makefile.am: + * performance-tests/CSD_Strategy/TestApps/Makefile.am: + * performance-tests/CSD_Strategy/TestInf/Makefile.am: + * performance-tests/CSD_Strategy/TestServant/Makefile.am: + * performance-tests/Callback/Makefile.am: + * performance-tests/Cubit/TAO/DII_Cubit/Makefile.am: + * performance-tests/Cubit/TAO/IDL_Cubit/Makefile.am: + * performance-tests/Cubit/TAO/MT_Cubit/Makefile.am: + * performance-tests/Latency/AMH_Single_Threaded/Makefile.am: + * performance-tests/Latency/AMI/Makefile.am: + * performance-tests/Latency/Collocation/Makefile.am: + * performance-tests/Latency/DII/Makefile.am: + * performance-tests/Latency/DSI/Makefile.am: + * performance-tests/Latency/Deferred/Makefile.am: + * performance-tests/Latency/Single_Threaded/Makefile.am: + * performance-tests/Latency/Thread_Per_Connection/Makefile.am: + * performance-tests/Latency/Thread_Pool/Makefile.am: + * performance-tests/Memory/IORsize/Makefile.am: + * performance-tests/Memory/Single_Threaded/Makefile.am: + * performance-tests/POA/Create_Reference/Makefile.am: + * performance-tests/POA/Demux/Makefile.am: + * performance-tests/POA/Implicit_Activation/Makefile.am: + * performance-tests/POA/Object_Creation_And_Registration/Makefile.am: + * performance-tests/Pluggable/Makefile.am: + * performance-tests/Protocols/Makefile.am: + * performance-tests/RTCorba/Multiple_Endpoints/Common/Makefile.am: + * performance-tests/RTCorba/Multiple_Endpoints/Orb_Per_Priority/Makefile.am: + * performance-tests/RTCorba/Multiple_Endpoints/Single_Endpoint/Makefile.am: + * performance-tests/RTCorba/Oneways/Reliable/Makefile.am: + * performance-tests/RTCorba/Thread_Pool/Makefile.am: + * performance-tests/Sequence_Latency/AMH_Single_Threaded/Makefile.am: + * performance-tests/Sequence_Latency/AMI/Makefile.am: + * performance-tests/Sequence_Latency/DII/Makefile.am: + * performance-tests/Sequence_Latency/DSI/Makefile.am: + * performance-tests/Sequence_Latency/Deferred/Makefile.am: + * performance-tests/Sequence_Latency/Single_Threaded/Makefile.am: + * performance-tests/Sequence_Latency/Thread_Per_Connection/Makefile.am: + * performance-tests/Sequence_Latency/Thread_Pool/Makefile.am: + * performance-tests/Throughput/Makefile.am: + * tao/Makefile.am: + * utils/catior/Makefile.am: + * utils/nslist/Makefile.am: + * utils/wxNamingViewer/Makefile.am: + + Regenerate. + +Thu Nov 9 09:54:12 UTC 2006 Martin Corino <mcorino@remedy.nl> + + * tests/Bug_2654_Regression/Hello.cpp: + + Fixed the last fix to get rid of compile warning;-). + +Thu Nov 9 06:46:21 UTC 2006 Ossama Othman <ossama_othman at symantec dot com> + + * docs/Options.html: + + Added description of new "dev_poll" Advanced_Resource_Factory + reactor type. + +Thu Nov 9 01:19:02 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * TAO_IDL/be/be_visitor_interface/interface_ss.cpp: + * TAO_IDL/be/be_decl.cpp: + * TAO_IDL/be_include/be_decl.h: + + Fixed bug where TIE class skeleton code is generated twice + when AMH code generation is enabled. + +Wed Nov 8 16:16:37 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * docs/releasenotes/index.html: + * docs/compiler.html: + + Update IDL compiler documentation to reflect the changes in + + Tue Nov 7 23:13:51 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + +Wed Nov 8 14:12:12 UTC 2006 Martin Corino <mcorino@remedy.nl> + + * tests/AMH_Exceptions/Test.idl: + * tests/AMH_Exceptions/server.cpp: + * tests/AMH_Exceptions/client.cpp: + * tests/AMH_Exceptions/run_test.pl: + + Added clean shutdown option. + +Wed Nov 8 13:22:12 UTC 2006 Martin Corino <mcorino@remedy.nl> + + * tests/Muxing/Test.idl: + * tests/Muxing/Receiver{.h .cpp}: + * tests/Muxing/server.cpp: + * tests/Muxing/client.cpp: + * tests/Muxing/run_test.pl: + + Added clean shutdown option. + +Wed Nov 8 13:14:12 UTC 2006 Martin Corino <mcorino@remedy.nl> + + * tests/Bug_2654_Regression/Hello.cpp: + + Fixed typo causing compile error. + +Wed Nov 8 07:31:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/DSI_Gateway/client.cpp: + * tests/DSI_Gateway/test_i.{h,cpp,inl}: + * tests/DSI_Gateway/test.idl: + Extended this test to also test an attribute, this extension + exhibits a few bugs in the DII code. Thanks to Venkat + <swara101 at yahoo dot com> for extending this test. + + * tao/DynamicInterface/Context.cpp: + * tao/DynamicInterface/DII_Arguments.cpp: + * tao/DynamicInterface/DII_Arguments_Converter_Impl.cpp: + Const changes and small performance improvements + + * tao/DynamicInterface/Request.cpp: + Fixed a few bugs related to the number of arguments, also + applied a few const and performance improvements. + +Tue Nov 7 23:13:51 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * TAO_IDL/be/be_visitor_operation/tie_si.cpp: + * TAO_IDL/be/be_visitor_interface/tie_si.cpp: + * TAO_IDL/be_include/be_visitor_interface/tie_si.h: + * TAO_IDL/be_include/be_visitor_operation/tie_si.h: + + Removed these files. + + * TAO_IDL/be/be_visitor_operation/tie_ss.cpp: + * TAO_IDL/be/be_visitor_interface/tie_ss.cpp: + * TAO_IDL/be_include/be_visitor_interface/tie_ss.h: + * TAO_IDL/be_include/be_visitor_operation/tie_ss.h: + + Added these files, in effect renaming the ones listed + as removed above. + + * TAO_IDL/be/be_visitor_operation.cpp: + * TAO_IDL/be/be_visitor_interface.cpp: + * TAO_IDL/be/be_visitor_interface/interface_si.cpp: + * TAO_IDL/be/be_visitor_interface/interface.cpp: + * TAO_IDL/be/be_visitor_interface/interface_ss.cpp: + * TAO_IDL/be/be_visitor_interface/tie_sh.cpp: + * TAO_IDL/be/be_codegen.cpp: + * TAO_IDL/be/be_visitor_root/root_si.cpp: + * TAO_IDL/be/be_visitor_root/root_sth.cpp: + * TAO_IDL/be/be_visitor_root/root.cpp: + * TAO_IDL/be/be_visitor_root/root_ss.cpp: + * TAO_IDL/be/be_global.cpp: + * TAO_IDL/be/be_visitor_attribute/attribute.cpp: + * TAO_IDL/be_include/be_codegen.h: + * TAO_IDL/be_include/be_global.h: + * TAO_IDL/be_include/be_visitor_interface.h: + * TAO_IDL/be_include/be_visitor_operation.h: + + Uninlined the generated code for TIE classes, since much + of this code consists of virtual functions, and can thus + cause problems related to RTTI on some compilers. Thanks + to Ossama Othman <ossama@dre.vanderbilt.edu> for + reporting the problem. This fix closes [BUGID:2699]. + +Tue Nov 7 19:45:02 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tests/Bug_2654_Regression/Hello.cpp: + ACE_OS::exit() doesn't seem to work well when called from + non-main threads on windows. Switching to ORB::shutdown() may + improve the performance a bit. + + * tests/Oneway_Timeouts/client.cpp: + Replaced cout/cerr with ACE_DEBUG/ACE_ERROR. Also reset the + timeout test value before issuing the shutdown invocation. I + believe this invocation was the major cause of failures because + it would cause a timeout after a delay in a scenaro that is + otherwise intended to be asynchronous and otherwise have no + delay at all. I also increased the TIME_THRESHOLD value to + better accomodate preemptively multitasking test hosts. + +Tue Nov 7 20:00:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * be/be_visitor_exception/exception_cs.cpp: + Generate the constructor taking all members as first method. The + gcc 3.2.3 based rtems compiler seems to get confused in the + generated tests/IDL_Test/anonymousC.cpp file when it is the last + method. For other compilers it doesn't make a difference but at + least with this change we have an error free compilation for + rtems + + * be/be_visitor_valuetype/valuetype_cs.cpp: + Generate const for retval + +Tue Nov 7 19:06:12 UTC 2006 Ossama Othman <ossama_othman at symantec dot com> + + * tao/ORB_Core.cpp (orbinitializer_registry_i): + + Temporarily reverted + "Thu Nov 2 23:23:16 UTC 2006 Ossama Othman <ossama_othman at + symantec.com>" change. It's not completely correct since it + prevents dynamic ORB unloading from working properly when + ORBInitializers have been registered. The existing code isn't + correct either but go with it for now since it has been tested + more. + +Tue Nov 7 12:44:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Two_Objects/worker.{h,cpp}: + * tests/Two_Objects/server.cpp: + Renamed timeout to orb_timeout, resolves the warnings in the + rtems build where timeout is also something in the rtems + glue code. + +Tue Nov 7 09:26:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/orbsvcs/IFRService/IFR_Service_Utils.cpp: + Fixed warning abour variable not used in rtems build + +Tue Nov 7 06:30:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Bug_2654_Regression/client.cpp: + Fixed warning in vc7 builds + +Mon Nov 6 20:04:53 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tests/Bug_2654_Regression/Hello.cpp: + Replace ::exit() with ACE_OS::exit() for portability. + +Mon Nov 6 19:33:15 UTC 2006 Chad Elliott <elliott_c@ociweb.com> + + * tao/Object.cpp (_set_policy_overrides): + + When setting the policy overrides on an Object, we need to ensure + that when it is collocated it has a collocated servant. + Otherwise, narrowing the Object afterward will cause an exception + to be thrown. Thanks to Artur DeEsperanto <DrEsperanto at gmx + dot net> for reporting this. + + * tests/Bug_2134_Regression/Bug_2134_Regression.mpc: + * tests/Bug_2134_Regression/Hello.cpp: + + Modified this test to utilize _set_policy_overrides() on the + collocated object to exercise the change above. + +Mon Nov 6 09:30:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/CollocationLockup/CollocationLockup.cpp: + * tests/CollocationLockup/AB.idl: + Renamed struct B to struct C to resolve compile problems with + rtems which seems to have a struct B. + + * tests/Multiple_Inheritance/client.cpp: + * tests/Multiple_Inheritance/Multiple_Inheritance.idl: + Renamed B to intB to resolve compile problems with + rtems which seems to have a struct B. + +Fri Nov 3 22:16:28 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * TAO_IDL/ast/ast_sequence.cpp (in_recursion): + + Fixed a possible bug in the check for recursive IDL type. + The change will in any case speed up the performance of + this method somewhat, since it will short-circuit the + checking if the sequence's element type is neither a struct, + a union, nor a sequence. + +Fri Nov 3 18:05:20 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * TAO_IDL/be/be_visitor_field/field_ch.cpp (visit_predefined_type): + + Simplified the code in this method, owing to the changes in + + Thu Nov 2 20:53:16 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + to AST_Type::nested_name(). + +Fri Nov 3 16:10:31 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * TAO_IDL/be/be_visitor_union_branch/private_ch.cpp: + * TAO_IDL/be/be_visitor_union_branch/public_ch.cpp: + + More cases of ad hoc generation of the global "::" that can be + removed because such generation is now handled systematically + in AST_Type::nested_name(), see + + Thu Nov 2 20:53:16 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + Thanks to Simon Massey <simon.massey@prismtechnologies.com> for + pointing out the code generation errors in the Security Service + that uncovered the omissions. + +Fri Nov 3 15:54:35 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/Strategies/SHMIOP_Connector.cpp (make_connection): + scoreboard cleanup. + +Fri Nov 3 12:35:41 UTC 2006 Chad Elliott <elliott_c@ociweb.com> + + * tao/Connection_Handler.inl: + + Added a missing #include "tao/Transport.h". + +Fri Nov 3 00:22:27 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * orbsvcs/orbsvcs/HTIOP/HTIOP_Connection_Handler.cpp: + * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Connection_Handler.cpp: + * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Connector.cpp: + * tao/Connection_Handler.h: + * tao/Connection_Handler.inl: + * tao/Connection_Handler.cpp: + * tao/IIOP_Connection_Handler.h: + * tao/IIOP_Connection_Handler.cpp: + * tao/IIOP_Connector.cpp: + * tao/Strategies/SCIOP_Connection_Handler.cpp: + * tao/Strategies/SCIOP_Connector.cpp: + * tao/Strategies/SHMIOP_Connection_Handler.cpp: + * tao/Strategies/SHMIOP_Connector.cpp: + * tao/Strategies/UIOP_Connection_Handler.cpp: + * tao/Strategies/UIOP_Connector.cpp: + + Propagating the changes applied just to IIOP for bug 2654 to the + other DOC managed ESIOPs. + +Thu Nov 2 23:23:16 UTC 2006 Ossama Othman <ossama_othman at symantec.com> + + * tao/ORB_Core.cpp (orbinitializer_registry_i): + + The ORBInitializer registry is supposed to be (and originally + was) a singleton. Store it in the global service configuration, + not the ORB-specific one. Addresses disjoint ORBInitializer + lists, and redundant ORBInitializer registrations. + + * tests/Portable_Interceptors/Register_Initial_References/Server_ORBInitializer.cpp (pre_init): + + Corrected CORBA::BAD_PARAM minor code check to use + CORBA::OMGVMCID instead. Easier to read and understand. + +Thu Nov 2 20:28:04 UTC 2006 Douglas C. Schmidt <schmidt@dre.vanderbilt.edu> + + * orbsvcs/examples/FaultTolerance/RolyPoly/ReplicaController.cpp + (send_reply): + + Fixed exit(1) to be ACE_OS::exit(1). + +Thu Nov 2 20:53:16 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * TAO_IDL/be/be_visitor_exception/exception_ctor.cpp: + * TAO_IDL/be/be_visitor_union/discriminant_ch.cpp: + * TAO_IDL/be/be_visitor_field/field_ch.cpp: + * TAO_IDL/ast/ast_type.cpp: + + Fixed generation of nested type name to add a double + colon if the name includes the canonical CORBA + namespace, to disambiguate it from any scoped + name which may include a different, nested namespace + named CORBA. Thanks to N Johnson <nj@artesys.info> + for reporting the bug. + + * tests/IDL_Test/nested_scope.idl: + + Added use case from the bug report above to the IDL file. + +Wed Nov 1 18:48:26 UTC 2006 Simon McQueen <sm@prismtech.com> + + * tests/Bug_2174_Regression/run_test.pl: + + Add a sleep to prevent spurious fails on win32. + +Mon Oct 30 19:49:12 UTC 2006 Martin Corino <mcorino@remedy.nl> + + * tests/CodeSets/simple/simple.idl: + * tests/CodeSets/simple/client.cpp: + * tests/CodeSets/simple/server.cpp: + * tests/CodeSets/simple/run_test.pl: + + Added proper shutdown operation to have test behave better + under (at least) OpenVMS. + +Mon Oct 30 19:08:12 UTC 2006 Martin Corino <mcorino@remedy.nl> + + * tests/Muxing/client.cpp: + * tests/Muxing/Server_Task.cpp: + * tests/Muxing/run_test.pl: + + Some adjustments of testparameters for when test running on + OpenVMS where the original test takes far too long for the + the timeouts set. + +Mon Oct 30 18:56:03 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * tao/AnyTypeCode/Any_Unknown_IDL_Type.cpp (to_object): + + Modified the decoding to be from a copy of the class + member CDR stream, rather than from the stream itself, + so the Any will retain ownership of the value as + mandated by the CORBA spec. Thanks to Phil Mesnier + <mesnier_p@ociweb.com> for reporting the bug and + suggesting the fix. Also changed several instances of + '0' to 'false' where CORBA::Boolean is referenced. + + * tao/AnyTypeCode/Marshal.cpp: + + Added CORBA::tk_event case values to the switch + statements in perform_append() and perform_skip(), + so that IDL eventtypes will be handled correctly. + +Mon Oct 30 15:09:24 UTC 2006 Simon McQueen <sm@prismtech.com> + + * TAO_IDL/be/be_visitor_ami_pre_proc.cpp: + + Set the correct pragma prefix value on the AMI reply handlers + for inherited interfaces. + + * tests/Bug_2669_Regression/Bug_2669_Regression.mpc: + * tests/Bug_2669_Regression/Child.idl: + * tests/Bug_2669_Regression/Parent.idl: + * tests/Bug_2669_Regression/README: + * tests/Bug_2669_Regression/client.cpp: + * tests/Bug_2669_Regression/run_test.pl: + + Regression test for the above fix. + +Mon Oct 30 14:35:24 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tests/Bug_2654_Regression/Bug_2654_Regression.mpc: + Do not build on minimum_corba platforms. + +Mon Oct 30 13:48:48 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tests/Bug_2654_Regression/client.cpp (init_callback): + Scoreboard cleanup. + +Sun Oct 29 17:08:21 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tests/Bug_2654_Regression/Hello.cpp: + Took out temporary, unix-only, header. + +Sun Oct 29 16:51:47 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/Connection_Handler.cpp: + * tao/IIOP_Connector.cpp: + * tao/Transport.h: + * tao/Transport.cpp: + + The fix for bug 2654. Thanks to Chris Reed <cr at progress dot + com> an Scott Mitchell <rsm at progress dot com> for help + tracking this down. + + * tests/Bug_2654_Regression/Bug_2654_Regression.mpc: + * tests/Bug_2654_Regression/Hello.h: + * tests/Bug_2654_Regression/Hello.idl: + * tests/Bug_2654_Regression/Hello.cpp: + * tests/Bug_2654_Regression/README: + * tests/Bug_2654_Regression/client.cpp: + * tests/Bug_2654_Regression/run_test.pl: + * tests/Bug_2654_Regression/server.cpp: + + The new test for ensuring bug 2654 is fixed. + +Sun Oct 29 10:58:12 UTC 2006 Martin Corino <mcorino@remedy.nl> + + * tao/Strategies/advanced_resource.cpp: + + Changed size argument for TP_Reactor construction from + ACE_TP_Reactor::DEFAULT_SIZE to ACE::max_handles () (as it is in + default_resource.cpp) to solve CTOR failures on OpenVMS (and + most likely also LynxOS) in OctetSeq and Leader_Followers + tests. + +Sat Oct 28 16:11:12 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/IIOP_Acceptor.cpp (open): + + When using an explicit, default address for the endpoint such as + iiop://:12345 this change is necessary to ensure the encoded + address is correct for either IPv4 or IPv6 as configured. + +Fri Oct 28 03:23:18 UTC 2006 Ossama Othman <ossama_othman at symantec dot com> + + * examples/POA/On_Demand_Loading/Servant_Manager.cpp: + * examples/POA/Loader/Servant_Activator.cpp: + * examples/POA/Loader/Servant_Locator.cpp: + + Addressed 64-bit conversion warnings. + +Thu Oct 27 02:53:53 UTC 2006 Ossama Othman <ossama_othman at symantec dot com> + + * tao/GIOP_Message_Base.cpp (parse_request_id): + + Corrected GIOP version check so that the GIOP 1.{0,1} case + doesn't apply to all major versions greater than or equal to + one, i.e. {1,2,3,...,N}.{0,1}. The case in question is only + meant for GIOP 1.0 or 1.1. + +Thu Oct 27 01:59:19 UTC 2006 Ossama Othman <ossama_othman at symantec dot com> + + * tao/GIOP_Message_Base.cpp (parse_request_id): + + Simplified error return logic. Addresses "code not reached" + warning. + + * tao/ORB.cpp (ORB_init): + + Do not pass down the default CORBA::Environment from TSS in the + native C++ exception case. Simply pass a locally instantiated + CORBA::Environment. Removes an unnecessary TSS access. + + * tao/EndpointPolicy/Endpoint_Acceptor_Filter.h: + + Moved TAO_POA_Manager forward declaration into the versioned + namespace. + + * tao/PortableServer/Servant_var.h: + + Removed suggestion of adding an exception specification. + They've fallen out of favor by most C++ gurus. + + * tao/Strategies/advanced_resource.cpp: + * tao/Strategies/advanced_resource.h: + + Added support ACE_Dev_Poll_Reactor. + + * tao/Valuetype/ValueBase.cpp (_tao_write_value_header): + + Fixed Coverity DEADCODE error. + + From Russell Mora <russell_mora at symantec dot com> + * examples/Content_Server/AMI_Iterator/Content_Iterator_i.cpp: + * examples/Content_Server/SMI_Iterator/Content_Iterator_i.cpp: + * tao/Strategies/SHMIOP_Acceptor.cpp: + * tao/Strategies/SHMIOP_Acceptor.h: + * tao/Strategies/SHMIOP_Factory.h: + + Added support for 64-bit file offsets. + + From Duane Binder <duane_binder at symantec dot com> + * TAO_IDL/be/be_helper.cpp: + + Buffer I/O by removing fflush() calls. Greatly improves tao_idl + performance, particularly over networked filesystems. + + * TAO_IDL/be/be_interface.cpp (gen_gperf_lookup_methods): + + Flush the output stream. Gperf also uses it as output. Ensure + current contents are written before gperf writes. + +Thu Oct 26 14:20:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/tests/Event/UDP/receiver.cpp: + Register the value factory + + * orbsvcs/tests/Event/UDP/run_test.pl: + Also run the test with a valuetype in the event. + +Thu Oct 26 13:30:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/orbsvcs/Event/ECG_CDR_Message_Receiver.cpp: + Reverted the change below. Thew new UDP RTEC test showed + that the consumer was receiving duplicate events without being + able to detect that. So a single send resulted in two events + being received. The original problem needs to be resolved in + a different way because this breaks any assumptions about events + + Mon Aug 21 15:37:23 UTC 2006 Douglas C. Schmidt <schmidt@dre.vanderbilt.edu> + * orbsvcs/orbsvcs/Event/ECG_UDP_Out_Endpoint.i: + * orbsvcs/orbsvcs/Event/ECG_CDR_Message_Receiver.cpp: Made the + simplest fix to the problem where if a UDP RTEC supplier + restarts while the consumer is still running, it gets duplicate + fragment errors. The cost of this fix is that duplicate + completed requests are no longer detected. Thanks to Keith + Nicewarner for these fixes. + +Thu Oct 26 12:54:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/orbsvcs/Event/ECG_CDR_Message_Sender.cpp: + * orbsvcs/orbsvcs/Event/ECG_Mcast_EH.cpp: + Const improvements + +Thu Oct 26 10:53:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/tests/Event/UDP/*: + New regression test for the RTEC using UDP federations. By default + we just send the string ACE/TAO/CIAO in the Any and this works, but + when -v is passed to the run_test.pl we send a valuetype which + contains the string. With the current svn head version of today + this fails because on the receiver side we can't extract the valuetype + out of the Any + +Wed Oct 25 15:33:27 UTC 2006 Chad Elliott <elliott_c@ociweb.com> + + * tao/PortableServer/Root_POA.cpp: + + My change from Mon Oct 23 12:24:57 UTC 2006 uncovered yet another + reference counting problem. The TAO_Root_POA was not manipulating + the reference count of the TAO_POA_Manager that it held. + +Mon Oct 23 12:24:57 UTC 2006 Chad Elliott <elliott_c@ociweb.com> + + * tao/PortableServer/Object_Adapter.h: + * tao/PortableServer/Object_Adapter.cpp: + + Added a helper method to correctly release the + TAO_POAManager_Factory which is used in both the destructor and + the close method. + + * tao/PortableServer/POAManagerFactory.h: + * tao/PortableServer/POAManagerFactory.cpp: + + When CORBA::release() is called on the POAManagerFactory, the + reference count does not necessarily go to zero because it has a + set of objects that are still holding on to a reference to the + factory. + + I added a method to the POAManagerFactory to remove all the + registered poa managers. This is called by the helper method in + the TAO_Object_Adapter class. + + Thanks to Artur DeEsperanto <DrEsperanto at gmx dot net> for + reporting this. This fixes bugid 2686. + +Fri Oct 20 16:54:32 UTC 2006 Chad Elliott <elliott_c@ociweb.com> + + * tao/ORB_Core.cpp: + + Fixed a memory leak where the TAO_Adapter would be leaked in the + event that an exception occurred before the adapter was added to + the adapter registry. Thanks to Artur DeEsperanto <DrEsperanto + at gmx dot net> for reporting this. + + * tao/PortableServer/Object_Adapter.cpp: + + Clean up the TAO_Root_POA and TAO_POAManager_Factory pointers in + the destructor in the event that the close() method was never + called. + +Fri Oct 20 07:44:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Transport_Connector.cpp: + Updated formatting of a debug message to match the others + in TAO + + * orbsvcs/orbsvcs/Event/ECG_CDR_Message_Receiver.cpp: + * orbsvcs/orbsvcs/Event/ECG_UDP_Receiver.cpp: + Use LM_ERROR together with ACE_ERROR, const improvements + +Tue Oct 17 22:04:33 UTC 2006 Ciju John <johnc@ociweb.com> + + * tao/Strategies/advanced_resource.cpp: + Believe the default reactor for advanced resources and default + resources should be the same. No reason for the advanced + resources recator not to automatically restart upon signal + interruption. + +Tue Oct 17 18:02:25 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/IORTable/Table_Adapter.cpp: + Fix to improve compiling on borland. + +Tue Oct 17 15:41:50 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * TAO_IDL/be/be_visitor_typedef/any_op_cs.cpp: + + Cosmetic changes. + + * TAO_IDL/be/be_visitor_sequence/any_op_cs.cpp: + * TAO_IDL/be/be_visitor_array/any_op_cs.cpp: + + Changed generated implementation of Any operators to + pass the alias typecode for typedefed sequences and + arrays. Formerly the raw sequence or array typecode + was used in all cases. + +Tue Oct 17 14:40:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/orbsvcs/Log/Hash_LogRecordStore.cpp: + Fixed warning in VxWorks 5.5.2PPC build, some layout changes + and use prefix increment instead of postfix + +Tue Oct 17 14:24:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Optimized_Connection/run_test.pl: + Removed commented out test code + +Tue Oct 17 11:58:02 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/IORTable/IOR_Table_Impl.h: + * tests/Bug_2683_Regression/server.cpp: + + Clean up scoreboard related problems. + +Mon Oct 16 21:16:14 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/IORTable/Table_Adapter.cpp: + * tao/IORTable/Table_Adapter.h: + * tao/IORTable/IOR_Table_Impl.cpp: + * tao/IORTable/IOR_Table_Impl.h: + Added var type for the impl, to avoid casting. + +Mon Oct 16 10:59:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Nested_Upcall_Crash/run_test.pl: + Check whether we could spawn the server, if not, give an error + and exit directly. + +Mon Oct 16 08:43:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * TAO_IDL/be/be_codegen.cpp: + Const changes + +Sun Oct 15 23:23:53 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/IORTable/Table_Adapter.cpp: + * tao/IORTable/Table_Adapter.h: + Fix for bug 2683. The problem was that if ORB::shutdown() was + called by another thread, the Table_Adapter::root_ reference + could go to nil while it was needed for processing in the + find_object() method. The fix is to use the reference counting + already supplied for the root_ reference to ensure the pointer + can never go nil while it is needed. Further, a synchronized + flag is used to restrict access to find_object() to ensure it is + not called once ORB::shutdown() was invoked. + + * tests/Bug_2683_Regression/*: + A new regression test for this bug. As this bug is the result of + a race condition, it is hard to detect anyway. The best way to + observe the bug is to revert the above change to Table_Adapter* + and add a small sleep to the find_object() method. This simply + forces the race condition. Adding a small sleep along with the + above fix demonstrates that the problem no longer occurs. + + Thanks to Chris Reed, cr at progress dot com, for reporting this + bug. + +Fri Oct 13 10:28:12 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/COIOP/*: + Extended this test so that we check whether a call should + succeed or not. This test requires that COIOP is the only + pluggable protocol available. + +Thu Oct 12 14:17:11 UTC 2006 Chad Elliott <elliott_c@ociweb.com> + + * orbsvcs/ImplRepo_Service/ImR_Locator_i.cpp: + + Split a debug message into two messages to work around a strange + bug in Visual Studio 2005. + +Thu Oct 12 14:13:44 UTC 2006 Chad Elliott <elliott_c@ociweb.com> + + * orbsvcs/tests/Event/Mcast/RTEC_MCast_Federated/run_test.pl: + + Use a random port number for the Naming_Service instead of a + hard-coded port number. + +Tue Oct 10 12:11:12 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/examples/ORT/README: + Layout improvement + + * orbsvcs/orbsvcs/CosConcurrencyControl.idl: + * orbsvcs/orbsvcs/CosLoadBalancing.idl: + * orbsvcs/orbsvcs/CosNaming.idl: + Doxygen improvements + + * orbsvcs/LoadBalancing/LB_ObjectReferenceFactory.cpp: + Const/bool improvements + +Tue Oct 10 09:24:12 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * docs/ior_parsing.html: + Updated that match_prefix returns bool + +Tue Oct 10 09:19:12 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Connection_Handler.cpp: + Layout change + + * tao/CORBALoc_Parser.{h,cpp}: + * tao/CORBAName_Parser.{h,cpp}: + * tao/DLL_Parser.{h,cpp}: + * tao/FILE_Parser.{h,cpp}: + * tao/IOR_Parser.{h,cpp}: + * tao/MCAST_Parser.{h,cpp}: + Changed match_prefix to return a bool instead of an int. + + * tao/CORBALoc_Parser.cpp: + * tao/CSD_ThreadPool/CSD_TP_Strategy_Factory.cpp: + * tao/Environment.cpp: + * tao/Exception.cpp: + * tao/IIOP_Connector.cpp: + * tao/IIOP_Profile.cpp: + * tao/ORB_Core.cpp: + * tao/Strategies/SCIOP_Connector.cpp: + * tao/Strategies/UIOP_Connector.cpp: + * tao/SystemException.cpp: + When using LM_ERROR, use ACE_ERROR as macro + + * tao/IIOP_Connection_Handler.cpp: + Const improvment + + * tao/ObjRefTemplate/ObjectReferenceTemplate.pidl: + Cleanup of comment + + * tao/Strategies/DIOP_Profile.cpp: + Small code improvement assuming the fact that operator>> returns + bool + +Tue Oct 10 06:08:12 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/COIOP/*: + New test for COIOP + +Thu Oct 05 00:35:37 2006 William R. Otte wotte@dre.vanderbilt.edu + + * TAO version 1.5.3 released. + +Tue Oct 3 17:24:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * docs/tutorials/Quoter/Simple/Client/index.html: + Corrected filenames generated by tao_idl + +Tue Oct 3 14:13:55 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * NEWS + + Updated NEWS file. + +Mon Oct 2 12:59:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/POA/POAManagerFactory/POAManagerFactory.cpp: + Renamed VERIFY macro to fix warning in mfc build and use + ACE_ERROR instead of ACE_DEBUG + +Thu Sep 28 15:13:49 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * orbsvcs/orbsvcs/PortableGroup/UIPMC_Connection_Handler.cpp: + * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Connection_Handler.cpp: + * tao/Connection_Handler.h: + * tao/IIOP_Connection_Handler.cpp: + * tao/Strategies/SCIOP_Connection_Handler.cpp: + * tao/Strategies/SHMIOP_Connection_Handler.cpp: + * tao/Strategies/UIOP_Connection_Handler.cpp: + + Fixing a regression on bug 2074 caused by CH::handle_timeout() + trying to update its instance, when the latter has already been + deallocated by CH::close(). This case only occurs when the + reference count on the Connection handler drops down to one. The + solution is to use a kind of auto pointer (TAO_Auto_Reference) + which magically increments the reference count in its + constructor and decrements it in its destructor. + +Thu Sep 28 14:55:06 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tests/Bug_2188_Regression/broken_i.cpp: + Silence the Borland warning. + +Wed Sep 27 07:24:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + Merged in the patch from bugzilla bug 2425, this fixes the crashing + of the notification service with vc8. This matches earlier investigations, + calling the set_qos seems to crash with vc8 because it is implemented + in a strange multiple inheritance setup. + + * orbsvcs/orbsvcs/Notify/ProxyConsumer.cpp: + * orbsvcs/orbsvcs/Notify/ProxySupplier.cpp: + Changed to call TAO_Notify_Object::set_qos() + directly within the init() method (within the scope + of a guard) to work around an apparent memory + corruption error that is seen only in Visual C++ 8 + builds. Thanks to zhamak dehghani <zhamakd at gmail + dot com> for initially reporting the problem and + providing a proposed fix, and to Charles Calkins + <calkinsc at ociweb dot com> for testing and + tweaking the proposed fix and further analyzing the + problem. (Not yet found, however, is the actual + cause of the memory corruption that existed before + this work around.) + +Tue Sep 26 11:59:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Strategies/COIOP_*.*: + New pluggable protocol COIOP, this is Collocated Only IOP. It is + almost finished, some small improvements are pending which will + be added soon. This protocol only allows collocated invocations and + must be explicitly enabled by defined TAO_HAS_COIOP to 1. At the + moment all other protocols are disabled with their defines it is + then possible to run a collocated TAO application without using + any network API's which safes footprint for an application. + + * tao/DynamicAny/DynCommon.h: + No need for a pragma push/pop + + * tao/ORB_Constants.h: + Added COIOP profile constant + + * tao/Strategies.mpc: + Extended for COIOP + + * tao/Strategies/advanced_resource.cpp: + Added COIOP factory + + * utils/catior/catior.cpp: + Added support for COIOP + + * docs/releasenotes/index.html: + Added COIOP + +Tue Sep 26 06:21:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Timeout/run_test.pl: + Don't use a hardcoded timeout and when there is no executable, exit + directly + +Mon Sep 25 22:27:24 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * orbsvcs/orbsvcs/HTIOP/HTIOP_Connection_Handler.cpp: + * orbsvcs/orbsvcs/PortableGroup/UIPMC_Connection_Handler.cpp: + * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Connection_Handler.cpp: + * tao/Connection_Handler.cpp: + * tao/IIOP_Connection_Handler.cpp: + * tao/Strategies/DIOP_Connection_Handler.cpp: + * tao/Strategies/SCIOP_Connection_Handler.cpp: + * tao/Strategies/SHMIOP_Connection_Handler.cpp: + * tao/Strategies/UIOP_Connection_Handler.cpp: + + Removed the call to release_os_resources call from the base + Connection_Handler destructor, and added code to each of the + protocol-specific CHs to report an error if release_os_resources + returns -1. + +Mon Sep 25 17:09:58 UTC 2006 Edward R. Mulholland <emulholl@atl.lmco.com> + + * tests/Timeout/client.cpp: + * tests/Timeout/run_test.pl: + * tests/Timeout/test_i.cpp: + * tests/Timeout/README.expected_behavior: + * tests/Timeout/README: + + Made relatively modest changes to track statistics on per type + of timeout. Also added a file README.expected_behavior that + carefully describes the timeout behavior that we should observe + on a proper real-time platform. Please address comments and + questions to Gautam Thaker <gthaker@atl.lmco.com>. + +Fri Sep 22 21:29:26 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * tao/DynamicAny/DynAnyFactory.h: + * tao/DynamicAny/DynamicAny.pidl: + * tao/DynamicAny/DynCommon.cpp: + * tao/DynamicAny/DynCommon.h: + * tao/DynamicAny/DynAnyFactory.cpp: + * tao/DynamicAny.mpc: + + Updated spec IDL to conform to CORBA + 3.0.3 (formal/2004-03-01) and added + dummy implementations of new operations + in existing interfaces. Also modified + the MPC file to inherit from valuetype.mpb, + to accommodate the new DynValue and + DynValueBox interfaces. + + * tao/DynamicAny/DynValue_i.cpp: + * tao/DynamicAny/DynValue_i.h: + * tao/DynamicAny/DynValueCommon_i.cpp: + * tao/DynamicAny/DynValueBox_i.cpp: + * tao/DynamicAny/DynValueBox_i.h: + * tao/DynamicAny/DynValueCommon_i.h: + + New files containing implementation classes + for new DynAny types related to valuetypes. + All of the operations in these classes are + dummy impls at present. + +Fri Sep 22 15:31:54 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/IIOP_Acceptor.cpp: + Fix for remaining IPV6 test failures. + + * tests/Bug_2188_Regression/broken.mpc: + * tests/Bug_2188_Regression/client.cpp: + Updated the mpc file to be consistent with current builds. Added + an explicit type case to the client to silence some build warnings. + + +Thu Sep 21 14:59:37 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tests/Bug_2188_Regression: + * tests/Bug_2188_Regression/broken.idl: + * tests/Bug_2188_Regression/broken.mpc: + * tests/Bug_2188_Regression/broken_i.h: + * tests/Bug_2188_Regression/broken_i.cpp: + * tests/Bug_2188_Regression/client.cpp: + * tests/Bug_2188_Regression/run_test.pl: + * tests/Bug_2188_Regression/server.cpp: + + A new test for an old problem. The actual problem is now + resolved, but I never got around to adding this test when I + should have. + +Wed Sep 20 21:37:58 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/IIOP_Acceptor.cpp: + + Fix for non-IPV6 builds. This relates back to the test for an + explicitly supplied any addr. + +Wed Sep 20 21:02:14 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/IIOP_Acceptor.cpp: + + Make sure the port_separator_loc value is used to compute the + string length only when it is not 0. + +Wed Sep 20 01:10:58 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tao/default_resource.cpp: + + Fixing the extra semicolon, causing hickups on some + platforms. Thanks to Phil Mesnier for catching it. + +Tue Sep 19 19:31:15 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/IIOP_Acceptor.h: + * tao/IIOP_Acceptor.cpp: + + Fix the treatment of explicitly defaulted endpoints on + potentially dual-stack hosts. Prior to this fix, expressing + certain forms of defaulted endpoints are inappropriately encoded + in the IOR using the literal string provided, rather than using + the probed interface addresses. When an endpoint is specified as + iiop://[::]:, iiop://[]:, or iiop://0.0.0.0:, this patch now + treats the first two as indicating that all IPv6 interfaces + should be included, but no IPv4. The third indicates that only + IPv4 interface addresses should be listed in the IOR. Using + iiop://: means that both IPv4 and IPv6 interface addresses are + to be included. + +Mon Sep 18 23:53:29 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tao/TAO_Internal.cpp: + + Replacing the usage of the global ACE lock with a separate + mutex, because the DLL Manager is also using the global + lock. This can cause a race condition in applications, which + re-enter the Service Configurator to load another DLL as part of + their own initialization. + +Mon Sep 18 23:11:31 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + This change fixes bug#2651, related to the order of destruction + of Codeset Manager instances, owned by the Default Resource + Factory. It is a prerequisite for correctly fixing bug#2612 and + also fixes a design artifact left over from the splitting of the + codesets in their own library. See the bugzilla entry for more + details. + + * tao/ORB_Core.h: + * tao/ORB_Core.cpp: + + The Core is now responsible for managing the life-cycle of the + Codeset Manager instance, instead of the Resource Factory. + + * tao/default_resource.h: + * tao/default_resource.cpp: + + Introducing a TAO_Codeset_Parameters class, which encapsulates + codeset configuration information. Keeping the configuration + information instead of the configured instance, allows the + Resource Factory to give up ownership of objects it creates (as + it should). Removed the ACE_Dynamic_Service_Dependency instance + member, which was a crude attempt at solving the issue, + described in bug#2651. + +Mon Sep 18 19:54:49 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * TAO_IDL/be/be_visitor_union_branch/private_ch.cpp: + * TAO_IDL/be/be_visitor_union_branch/public_ch.cpp: + * TAO_IDL/be/be_visitor_union_branch/public_ci.cpp: + * TAO_IDL/be/be_visitor_union_branch/cdr_op_cs.cpp: + * TAO_IDL/be/be_visitor_union_branch/serializer_op_cs.cpp: + * TAO_IDL/be/be_visitor_argument/arglist.cpp: + * TAO_IDL/be/be_visitor_operation/rettype.cpp: + * TAO_IDL/be/be_visitor_sequence/buffer_type.cpp: + * TAO_IDL/be/be_visitor_sequence/serializer_op_cs.cpp: + * TAO_IDL/be/be_predefined_type.cpp: + * TAO_IDL/be/be_visitor_array/cdr_op_cs.cpp: + * TAO_IDL/be/be_visitor_array/serializer_op_cs.cpp: + * TAO_IDL/be/be_sequence.cpp: + * TAO_IDL/be/be_visitor_field/field_ch.cpp: + * TAO_IDL/be/be_decl.cpp: + + Fixed cases of code generation involving + CORBA::AbstractBase that were overlooked in + + Fri Sep 15 19:02:58 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + +Mon Sep 18 16:44:10 UTC 2006 Ciju John <johnc@ociweb.com> + + * tao/Transport_Connector.cpp: + Removed an extra semi-colon. + +Fri Sep 15 19:02:58 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * TAO_IDL/include/idl_global.h: + * TAO_IDL/include/ast_predefined_type.h: + * TAO_IDL/be/be_codegen.cpp: + * TAO_IDL/ast/ast_predefined_type.cpp: + * TAO_IDL/fe/fe_init.cpp: + * TAO_IDL/util/utl_global.cpp: + * TAO_IDL/util/utl_scope.cpp: + + Added recognition of CORBA::AbstractBase as a pseudo + object, similarly to the existing support for Object, + ValueBase, and TypeCode. Also added automatic inclusion + of tao/Valuetype/AbstractBase.h in generated code + if AbstractBase is seen in IDL. + +Wed Sep 13 14:58:24 UTC 2006 Ciju John <johnc@ociweb.com> + + * tao/Transport_Connector.h: + * tao/Transport_Connector.inl: + * tao/Transport_Connector.cpp: + The TransportCleanupGuard class is for local usage, so am moving + it into an anonymous namespace. + +Fri Sep 8 20:03:26 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * orbsvcs/orbsvcs/HTIOP/HTIOP_Connection_Handler.cpp: + * orbsvcs/orbsvcs/PortableGroup/UIPMC_Connection_Handler.cpp: + * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Connection_Handler.cpp: + * tao/Strategies/DIOP_Connection_Handler.cpp: + * tao/Strategies/SCIOP_Connection_Handler.cpp: + * tao/Strategies/UIOP_Connection_Handler.cpp: + * NEWS: + Applied fix from Tue Aug 29 17:40:01 UTC 2006 Phil Mesnier + <mesnier_p@ociweb.com> to other protocols. This explicitly calls + release_os_resources from the derived destructor, as was + originally intended. This resolves bug 2649. + +Thu Sep 7 14:19:58 UTC 2006 Ciju John <johnc@ociweb.com> + + * tao/Transport_Connector.inl: + Removed the extra semicolon. This was creating havoc on the + windows builds. + +Tue Sep 5 20:39:15 UTC 2006 Ciju John <johnc@ociweb.com> + + * tao/PortableServer/Servant_var.h: + * tao/PortableServer/Servant_var.inl: + Commented out the member templates code. See: + Tue Aug 15 21:25:50 UTC 2006 Ciju John <johnc@ociweb.com> + +Tue Sep 5 13:56:58 UTC 2006 Ciju John <johnc@ociweb.com> + + * tests/Bug_1476_Test/run_test.pl: + Took out a debug print statement. + +Thu Aug 31 23:05:46 UTC 2006 Ciju John <johnc@ociweb.com> + + * tao/ORB_Core.cpp: + Access to the TAO_ORB_Core_Static_Resources was through the + static instance() API. This created a bug when the object which + earlier was a true singleton was converted to a service object + under the control of the service config. The service config + depending upon the access point can either return the global + object or the object local to the particular orb. In this case + (during an invocation) we need to access the local object and + the service config was returning the global object. This led to + incorrect synch-scope policies being set in invocations via a + non-default orb (named orb). Since at this point the ORB_Core + has a pointer to the local service config, its safe to query it + to request the local service object. + + * tao/Transport_Connector.h: + * tao/Transport_Connector.inl: + * tao/Transport_Connector.cpp: + + In case of a asynch connection establishment failure or + registration failure we need to clean up teh transport, else we + have a file descriptor leak. The 'TransportCleanupGuard' takes + care of that. + + * tests/Bug_1476_Test: + * tests/Bug_1476_Test/Bug_1476_Test.mpc: + * tests/Bug_1476_Test/Hello.idl: + * tests/Bug_1476_Test/README: + * tests/Bug_1476_Test/client.cpp: + * tests/Bug_1476_Test/run_test.pl: + Add new regression test for buzilla 1476 + +Fri Sep 1 14:05:12 UTC 2006 Martin Corino <mcorino@remedy.nl> + + * tests/Bug_2183_Regression/run_test.pl: + Fixed spawning of PERL process on OpenVMS. + + * tests/File_IO/run_test.pl: + Replaced literal timeout value ('10') with platformdependent + variable defined in PerlACE/Run_Test.pm for that purpose. + + * tests/Faults/test_i.cpp: + * tests/Crashed_Callback/Crashed_Callback.cpp: + * tests/Nested_Upcall_Crash/Client_Peer.cpp: + Fixed these tests for OpenVMS replacing ACE_OS::abort() as deliberate + crash trigger by ACE_OS::_exit() as using abort() causes these tests + to 'hang' instead of crash. + +Thu Aug 31 18:28:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/PortableServer/PortableServer.cpp: + With CORBA/e micro don't use single thread strategy and non + retain servant retention strategy + +Thu Aug 31 11:22:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Messaging/Messaging_Loader.cpp: + Changed called_once to a real bool + + * tao/Base_Transport_Property.{cpp,inl}: + * tao/Cache_Entries.inl: + * tao/Transport_Descriptor_Interface.{h,cpp}: + Use true/false instead of 1/0 + + * tao/Transport_Cache_Manager.cpp: + Const improvements + +Thu Aug 31 03:35:14 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/Asynch_Queued_Message.cpp: + * tao/Invocation_Adapter.cpp: + Fix for timed, asynch messages that happen to expire during + transport resolution. This fixes part of the Oneway_Timeout test. + + * tests/Big_Oneways/run_test.pl: + This test still has an undefined intermittent deadlock, but I + lowered the timeout thresholds so it doesn't wait 15 minutes in + a deadlock condition. If any client fails, they all deadlock, so + there is no reason to wait 5 minutes before killing clients 2 + and 3. + + * tests/Oneway_Timeouts/client.cpp: + * tests/Oneway_Timeouts/run_test.pl: + + Fixed a math error in run_test.pl that falsely triggered a + timeout failure every time. Rearranged the semantics of forced + connection establishment to ensure it always works without + regard to messaging timeouts. + +Wed Aug 30 13:20:38 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/IIOP_Connection_Handler.h: + * tao/IIOP_Connection_Handler.cpp: + * tao/IIOP_Connector.cpp: + + Revised the fix for bug 2417. That fix added a reference count + to IIOP Connection Handlers associated with pending asynch + connections to accomodate a reference held by an ACE component + that does not participate in reference counting. However, + blindly adding a reference caused the Connection_Handler to not + get cleaned up as part of ORB destruction. This then caused + trouble on windows platforms where any sockets that were open at + program termination were aborted with a RST and undelivered + stream content purged. + + This fix gives control of this additional reference to the + connection handler itself, so that upon successful completion of + connection establishment, that reference is removed and no + resource is leaked. + +Wed Aug 30 13:29:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/PortableServer/PortableServer.cpp: + + Don't do the process_directives for the policies that aren't + supported with CORBA/e micro, this will reduce the footprint for + a static build. + +Wed Aug 30 06:05:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Bug_2560_Regression/server.cpp: + + No need to explicitly set the defaul thread policy, makes it possible + to use this test when minimum corb is enabled + +Wed Aug 30 04:20:00 UTC 2006 Frank Rehberger <frehberger@prismtech.com> + + * NEWS: explains user-visible implications of patch #2301 + +Tue Aug 29 21:55:05 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/Profile_Transport_Resolver.cpp: + Fixing syntax to comply with ACE coding guidelines. + +Tue Aug 29 20:02:01 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * examples/POA/Forwarding/Servant_Activator.h: + + * examples/POA/Loader/Servant_Activator.h: + * examples/POA/Loader/Servant_Locator.h: + + * examples/POA/On_Demand_Activation/Servant_Activator.h: + * examples/POA/On_Demand_Activation/Servant_Locator.h: + + * examples/POA/On_Demand_Loading/Servant_Activator.h: + * examples/POA/On_Demand_Loading/Servant_Locator.h: + + These all use Servant Managers which are local objects, but need + to be reference counted so that they may be shutdown properly + and thereby release their reference to the ORB_Core so it can be + finalized and clean up resources it owns. + +Tue Aug 29 17:40:01 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/IIOP_Connection_Handler.cpp: + Explicitly call release_os_resources from the dtor. + Release_os_resources is a virtual method, but of course calling + that from the base class's destructor won't work. This fix + ensures that sockets are really closed when the connection + handler is destroyed. + +Tue Aug 29 17:26:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Bug_2560_Regression/Bug_2560_Regression.mpc: + Set needed idlflags and fixed build order, should fix the gcc4 linker + errors + +Tue Aug 29 13:46:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Transport.cpp (schedule_output_i): + Check whether we have a reactor, if not, return -1, this fixes + bugzilla bug 2560. Thanks to Andy Salnikov <a_salnikov at yahoo dot com> + for reporting this bug and making a test program + +Tue Aug 29 12:53:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Bug_2560_Regression/*: + New regression for bug 2560. Thanks to Andy Salnikov + <a_salnikov at yahoo dot com> for creating this test. + +Tue Aug 29 11:48:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/orbsvcs/Makefile.am: + Fixed small bug + + * orbsvcs/orbsvcs/RTEvent.mpc: + Replaced .i with .inl + +Tue Aug 29 07:54:00 UTC 2006 Frank Rehberger <frehberger@prismtech.com> + + * tao/PortableServer/Object_Adapter.cpp: + + Handle location forward requests raised by colocal + ServerInterceptor. This fixes bug #2301. + + * orbsvcs/tests/FaultTolerance/GroupRef_Manipulation/server.cpp: + + Extend test to verify the forward request raised by colocal + ServerInterceptor is handled correctly + +Mon Aug 28 14:30:11 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * orbsvcs/tests/Security/BiDirectional/run_test.pl: + + Updated the test to run in SSL mode by default. + + * orbsvcs/tests/Security/BiDirectional/ssl/ca.pem: + * orbsvcs/tests/Security/BiDirectional/ssl/client-cert.pem: + * orbsvcs/tests/Security/BiDirectional/ssl/client-key.pem: + * orbsvcs/tests/Security/BiDirectional/ssl/server-cert.pem: + * orbsvcs/tests/Security/BiDirectional/ssl/server-key.pem: + + Updated expired certificates. + +Mon Aug 28 12:05:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/DII_Collocation_Tests/twoway/Hello.cpp: + Added needed include to fix test failures in static builds + +Mon Aug 28 09:39:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/OBV/Simple/run_test.pl: + Updated for VxWorks + +Mon Aug 28 09:12:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Makefile.am: + Regenerated + +Mon Aug 28 07:18:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Bounded_Sequence_CDR_T.h: + * tao/Unbounded_Sequence_CDR_T.h: + Only deliver the CORBA::WChar specialization when ACE_LACKS_NATIVE_WCHAR_T + is not defined. Thanks to Birgit Platt <bhe at newbase dot de> and + Jules Colding <colding at omesc dot com> for reporting this problem + +Sun Aug 27 11:10:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/DII_Collocation_Tests/twoway/DII_Collocation_Tests.mpc: + New file to build this test + + * tests/DII_Collocation_Tests/twoway/run_test.pl: + Also cleanup the simple test ior file + + * tests/DII_Collocation_Tests/twoway/Server_Task.cpp: + Layout changes and initialiser pointer to 0 + +Sat Aug 26 11:44:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Bug_2429_Regression/run_test.pl: + Updated work with vxworks + +Fri Aug 25 21:53:55 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * TAO_IDL/tao_idl.cpp: + + Added an output message with the name of the IDL file + being processed. + +Fri Aug 25 15:02:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/ORB_Core.cpp: + Use true/false for opt_for_collocation and use_global_collocation + + * tao/GIOP_Message_Lite.cpp: + * tao/GIOP_Message_Base.cpp: + Add include of ace/Min_Max.h to resolve compile errors with VxWorks + 6.3 and the diab 5.4.0 compiler + +Thu Aug 24 13:37:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/orbsvcs/HTIOP/HTIOP_Factory.{h,cpp}: + Removed minor_ and major_ members, where not used at all + +Thu Aug 24 10:41:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/tests/Security/BiDirectional/run_test.pl: + * orbsvcs/tests/Security/ssliop_corbaloc/run_test.pl: + Check the result of spawn so that when there is no executable we + directly return an error + +Wed Aug 23 18:41:15 UTC 2006 Douglas C. Schmidt <schmidt@dre.vanderbilt.edu> + + * orbsvcs/orbsvcs/Event/ECG_UDP_Out_Endpoint.inl: Reverted changes + to the copy constructor so that it initializes the data members + properly. + +Wed Aug 23 17:43:29 UTC 2006 Yan Dai <dai_y@ociweb.com> + + * tao\PortableServer\DII_Arguments_Converter.h: + * tao\PortableServer\DII_Arguments_Converter.cpp: + * tao\DynamicInterface\DII_Arguments_Converter_Impl.h: + * tao\DynamicInterface\DII_Arguments_Converter_Impl.cpp: + + Reverted back the changes + "Tue Aug 22 20:28:26 UTC 2006 Yan Dai <dai_y@ociweb.com>". + Added the virtual destructor to TAO_DII_Arguments_Converter + class to let the rtti work with gcc4 to resolve the oneway + test failure on gcc4 builds. + +Wed Aug 23 12:49:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/IIOP_Factory.{h,cpp}: + * tao/IIOP_Lite_Factory.{h,cpp}: + * tao/Strategies/DIOP_Factory.{h,cpp}: + * tao/Strategies/SCIOP_Factory.{h,cpp}: + * tao/Strategies/SCIOP_Lite_Factory.{h,cpp}: + * tao/Strategies/SHMIOP_Factory.{h,cpp}: + * tao/Strategies/UIOP_Factory.{h,cpp}: + * tao/Strategies/UIOP_Lite_Factory.{h,cpp}: + Removed the minor_ and major_ members, they where only initialized + to a value and then never used at all. Saves a few bytes of footprint + +Tue Aug 22 20:28:26 UTC 2006 Yan Dai <dai_y@ociweb.com> + + * tao\PortableServer\DII_Arguments_Converter.cpp: + * tao\PortableServer\DII_Arguments_Converter.h: + * tao\DynamicInterface\DII_Arguments_Converter_Impl.h: + * tao\DynamicInterface\DII_Arguments_Converter_Impl.cpp: + + Made the TAO_DII_Arguments_Converter service object registered + when PortableServer lib is loaded and the TAO_DII_Arguments_Converter_Impl + object will overwrite the registered "DII_Arguments_Converter" service + object. This would fix dynamic_cast failure when accessing the + "DII_Arguments_Converter" dynamic service object in Upcall_Wrapper::upcall() + on FC4 with gcc4.x builds. This should fix the + $TAO_ROOT/tests/DII_Collocation_Tests/oneway test failure on FC4 with + gcc4.x builds. + +Tue Aug 22 19:39:07 UTC 2006 Ciju John <john@ociweb.com> + + * orbsvcs/tests/Event/Mcast/RTEC_MCast_Federated/EchoEventConsumerMain.cpp: + * orbsvcs/tests/Event/Mcast/RTEC_MCast_Federated/EchoEventSupplierMain.cpp: + For increased portability replaced iostream API with ACE IO + macros. + + * tao/Wait_On_Read.cpp: + If reactor registration fails, the 'is_registered_' flag + shouldn't be set. + + * tests/Oneway_Timeouts/client.cpp: + Std header <cmath> is contains the signature for std::abs (). + +Tue Aug 22 14:57:18 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tests/Oneway_Timeouts/client.cpp: + Added a test case for transient exceptions. The TAO-specific + connection timout causes a connection attempt to give up, which + in turn causes the invocation endpoint selector to try + connecting to a new endpoint. If no more endpoints are available + for connecting, a transient exception is raised. Thus the ORB + code is working correctly, but the test was not handing the + transient exception as it should. + +Tue Aug 22 11:27:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/orbsvcs/Trader/Constraint_Interpreter.cpp: + Initialise pointer with 0 to fix gcc4 warning + +Tue Aug 22 09:58:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Transport_Conector.cpp: + Made some debug statements consistent with the regular + way we format things in TAO + +Tue Aug 22 09:49:48 UTC 2006 Simon McQueen <sm@prismtech.com> + + * tao/Client_Strategy_Factory.h: + * tao/Client_Strategy_Factory.cpp: + * tao/Fault_Tolerance_Service.h: + * tao/Fault_Tolerance_Service.inl: + * tao/Fault_Tolerance_Service.cpp: + * tao/ORB.h: + * tao/ORB.cpp: + * tao/default_client.h: + * tao/default_client.cpp: + + Remove spurious FT methods. This fixes bug #2644. + +Tue Aug 22 09:26:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/PI_Server.mpc: + Removed invalid -Sc flag, thanks to Simon McQueen for reporting + this. + +Tue Aug 22 07:45:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/orbsvcs/Event/ECG_UDP_Out_Endpoint.inl: + Added missing include + +Mon Aug 21 21:38:15 UTC 2006 William R. Otte <wotte@dre.vanderbilt.edu> + + * docs/tutorials/Quoter/RTCORBA/Stock_Database.h + * docs/tutorials/Quoter/RTCORBA/Distributor_i.cpp + * docs/tutorials/Quoter/RTCORBA/Stock_Database.tpp + + Removed a dependancy that the Stock_Database had on IDL. + +Mon Aug 21 18:42:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Seq_Var_T.{cpp,inl}: + Moved include of OS_Memory.h from cpp to inl file to resolve + compile problems when building static and TAO_NO_COPY_OCTET_SEQUENCES + is defined to 0. This file is included to get ACE_NEW which is + only used in the inl file + +Mon Aug 21 18:04:51 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * orbsvcs/tests/InterfaceRepo/IDL3_Test/idl3_client.h: + * orbsvcs/tests/InterfaceRepo/IDL3_Test/idl3_client.cpp: + + Changed a CORBA::Boolean class member to native boolean, + since it isn't used in a CORBA call. + + * orbsvcs/tests/InterfaceRepo/IDL3_Test/test.idl: + + Added a workaround IDL module called 'Components', + containing just enough for the IFR loader to look up + necessary stuff, eliminating the need to include + Components.idl and the corresponding dependency on CIAO. + + * orbsvcs/tests/InterfaceRepo/IDL3_Test/run_test.pl: + + Removed CIAO_ROOT/ciao from the list of include paths. + +Mon Aug 21 15:37:23 UTC 2006 Douglas C. Schmidt <schmidt@dre.vanderbilt.edu> + + * orbsvcs/orbsvcs/Event/ECG_UDP_Out_Endpoint.i: + * orbsvcs/orbsvcs/Event/ECG_CDR_Message_Receiver.cpp: Made the + simplest fix to the problem where if a UDP RTEC supplier + restarts while the consumer is still running, it gets duplicate + fragment errors. The cost of this fix is that duplicate + completed requests are no longer detected. Thanks to Keith + Nicewarner for these fixes. + + * orbsvcs/orbsvcs/FtRtEvent/EventChannel/FT_ProxyAdmin_T.cpp + (set_state): Fixed a problem with type punning. Thanks to + Thomas Girard <thomas.g.girard at free dot fr> for reporting + this. + + * examples/Advanced/ch_8_and_10/server.cpp: Fixed a compile error + by adding a _duplicate() call. Thanks to Christoph Schmalhofer + <christoph dot schmalhofer at gmx dot net> for contributing this + fix. This fixes bugid 2611. + +Mon Aug 21 16:28:51 UTC 2006 William R. Otte <wotte@dre.vanderbilt.edu> + + * docs/tutorials/Quoter/RTCORBA/Stock_Database.cpp + * docs/tutorials/Quoter/RTCORBA/StockNameConsumer_i.cpp + * docs/tutorials/Quoter/RTCORBA/Stock_Database.h + * docs/tutorials/Quoter/RTCORBA/StockNameConsumer_i.h + * docs/tutorials/Quoter/RTCORBA/Common.idl + * docs/tutorials/Quoter/RTCORBA/StockQuoter.mpc + * docs/tutorials/Quoter/RTCORBA/Distributor_i.cpp + * docs/tutorials/Quoter/RTCORBA/Distributor_i.h + * docs/tutorials/Quoter/RTCORBA/Broker_i.cpp + + More updates to the Quoter tutorial, the most significant + being a redesign on the Stock_Database to make is more generic + and easily plugged into other Quoter examples. + +Mon Aug 21 10:07:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Portable_Interceptors/Bug_2510_Regression/*: + Integrated more advanced test, thanks to Martin Cornelius + <Martin dot Cornelius at smiths-heimann dot com> for creating + this version. This test doesn't crash anymore, probably due + to the recent PI refactoring + +Mon Aug 21 05:58:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Unbounded_Octet_Sequence_T.h: + When TAO_NO_COPY_OCTET_SEQUENCES is defined to 0 use the global + operator == and != to resolve compile errors with TAO. Some compilers + had problems with these operators so they where fully commented out + but this made it impossible to define TAO_NO_COPY_OCTET_SEQUENCES to + 0 completely + +Sat Aug 19 22:03:03 UTC 2006 Yan Dai <dai_y@ociweb.com> + + * tests/DII_Collocation_Tests/oneway/Collocated_Test.cpp: + + Added the DII_Arguments_Converter_Impl.h include to force static builds + to load the DII_Arguments_Converter service. + +Fri Aug 18 17:00:15 UTC 2006 Yan Dai <dai_y@ociweb.com> + + Merged OCI changes that fix the problem that CORBA::Request::send_oneway() + crashes on get_in_arg() or gives incorrect arguments when the request is + collocated oneway request. These fixes are combination of + "Tue Dec 27 13:20:58 USMST 2005 Yan Dai <dai_y@ociweb.com>" and + "Tue May 2 16:52:43 UTC 2006 Chad Elliott <elliott_c@ociweb.com>" + (See BugZilla #2545 for details). + Note this only fixed the oneway DII collocation requests, the twoway + collocation request was not tested yet. + + * tao/Invocation_Adapter.cpp: + * tao/Invocation_Adapter.h: + * tao/Invocation_Adapter.inl: + + Added is_dii_request_ data member. + Added an extra parameter is_dii_request(defaults to false - not a dii request) + to the invoke() function so it can be passed to the Operation_Details + to mark the request is a dii request. + + * tao/operation_details.h: + * tao/operation_details.i: + + Added is_dii_request_ data member and the accessor and added the is_dii_request + parameter to the constructor. + + * tao/DynamicInterface/DII_Arguments.h: + * tao/DynamicInterface/DII_Arguments.inl: + + Added accessor to the NVList. + + * tao/PortableServer/Upcall_Wrapper.cpp: + + Updated upcall () to use the DII_Argument_Convert to expand the + DII request parameter from NVList to list of *SArgument. + This would make DII request parameters in NVList from the client + side to be changed to the list of arguments so the server side + can correctly retrieve the arguments. + + * tao/DynamicInterface/DII_Arguments_Converter_Impl.cpp: + * tao/DynamicInterface/DII_Arguments_Converter_Impl.h: + * tao/PortableServer/DII_Arguments_Converter.h: + + An abstract class DII_Arguments_Converter is added for conversion + of the NVList to list of *SArgument. + These new files are added to resolve the library circuit dependency + problem. This makes the conversion of NVList to list of *SArgument + can be done in DynamicInterface instead of in PortableServer. The + expanded skel args are used in get_in_arg() to give the correct + "IN" arguments. + + * tao/DynamicInterface/DII_Invocation_Adapter.cpp: + + Passed is_dii_request true to Invocation_Adapter constructor to + indicate it's a dii request invocation. + + * tao/DynamicInterface/Request.cpp: + + Passed is_dii_request true to construct Invocation_Adapter object to + indicate it's a dii request invocation. + + * tests/DII_Collocation_Tests/oneway/Collocated_Test.cpp: + + Removed the commented include. + +Fri Aug 18 16:18:35 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tests/Timed_Buffered_Oneways/client.cpp: + * tests/Timed_Buffered_Oneways/test_i.cpp: + + I am unable to reproduce the failure in this test locally. The + output in the scoreboard files is inconclusive, but the failure + seems to be consistent on the affected platforms. This is just + some additional output to help make sense of what might be + causing the failure. + +Fri Aug 18 13:34:43 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/Profile_Transport_Resolver.cpp: + Fix for connection timeout errors introduced by Adam's commit of + Justin's changes in Tue Aug 15 14:56:35 UTC 2006 Adam Mitz + <mitza@ociweb.com> + +Fri Aug 18 13:27:02 UTC 2006 Adam Mitz <mitza@ociweb.com> + + * tests/Oneway_Timeouts/client.cpp: + Eliminated warnings due to using var's without ".in()". + +Fri Aug 18 09:09:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/GIOP_Message_Generator_Parser*.*: + Changed all write_* methods to return a bool instead of int, the + return value was used as bool, so made this explicit. Also added + more const usage when possible. + + * tao/GIOP_Message_Base.cpp: + Use true for a bool + + * tao/GIOP_Mesage_Locate_Header.{h,inl}: + Const improvements + + * tao/GIOP_Utils.h: + Removed stuff that was not used + + * tao/GIOP_Utils.cpp: + Removed this file, not used at all + + * tao/tao.mpc: + Removed GIOP_Utils.cpp + +Fri Aug 18 07:17:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/GIOP_Message_Base.cpp: + Fixed incorrect method name in debug macro. When the parsing of the + request header has failed, throw the marshal exception directly, don't + try to process the service context and assign translators, we just + don't know at this stage if that is safe to do, so don't do that. + Also in parse_request_id do the demarshaling of the service contexts + and the request id in two seperate steps, if the demarshaling of the + service context fails, then we can't demarshal the request id + safely. + + * tao/GIOP_Message_Generator_Parser.cpp (parse_reply): + When the demarshaling of the request id failed then return with -1 + instead of trying to demarshal anymore data. Also when we have + received an unknow reply status return -1 instead of returning 0 + + * tao/GIOP_Message_Generator_Parser_10.cpp : + In parse_request_header check whether the demarshaling of the + service contexts didn't fail. Also in parse_reply log an error when + demarshaling fails. + + * tao/IIOP_Connection_Handler.cpp: + * tao/Transport_Cache_Manager.cpp: + Const improvement + +Thu Aug 17 20:46:29 UTC 2006 Adam Mitz <mitza@ociweb.com> + + * orbsvcs/tests/ImplRepo/Bug_2604_Regression/Bug_2604.mpc: + Set the exename explicitly, the default wasn't behaving as expected. + + * tests/Oneway_Timeouts/client.cpp: + Added a cast to int to eliminate Borland compiler error due to + ambiguous call to std::abs(). + + * tests/Timed_Buffered_Oneways/client.cpp: + Eliminated warnings due to using var's without ".in()". + +Thu Aug 17 12:54:58 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * TAO_IDL/fe/lex.yy.cpp: + + Touched this file, to eliminate warnings about it being possible + outdated, since minor changes were made in + + Wed Aug 16 17:29:44 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + to idl.ll, which generates it. + +Wed Aug 16 21:48:22 UTC 2006 Shanshan Jiang <shanshan.jiang@vanderbilt.edu> + + * docs/tutorials/Quoter/RTCORBA/docs/broker.html + * docs/tutorials/Quoter/RTCORBA/docs/database.html + * docs/tutorials/Quoter/RTCORBA/docs/distributor.html + * docs/tutorials/Quoter/RTCORBA/docs/idl.html + * docs/tutorials/Quoter/RTCORBA/docs/index.html + * docs/tutorials/Quoter/RTCORBA/docs/mpc.html + * docs/tutorials/Quoter/RTCORBA/docs/nameConsumer.html + * docs/tutorials/Quoter/RTCORBA/docs/prioritymapping.html + * docs/tutorials/Quoter/RTCORBA/docs/run.html + Update these files according to William R. Otte's modification to the + corresponding source code. + +Wed Aug 16 18:09:44 UTC 2006 William R. Otte <wotte@dre.vanderbilt.edu> + + * docs/tutorials/Quoter/RTCORBA/StockNameConsumer_i.cpp + * docs/tutorials/Quoter/RTCORBA/StockNameConsumer_i.h + + Improved valuetype memory management. + +Wed Aug 16 17:29:44 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * TAO_IDL/fe/lex.yy.cpp: + + Cosmetic changes. + + * TAO_IDL/fe/idl.ll: + + Replace overlooked instances of idl_bool with native bool, + and corresponding values I_TRUE and I_FALSE with the + native versions. + + * TAO_IDL/util/utl_identifier.cpp: + + In the constructor from char*, if the input arg begins + with '_cxx_', the logic is changed. If the tail of the + arg matches a C++ keyword, the arg is left unchanged. + Otherwise, the leading underscore is stripped as with + all other 'escaped' input args. This fix closes [BUGID:2634], + thanks to Boris Kolpackov <boris@dre.vanderbilt.edu> for + reporting it. + +Wed Aug 16 17:17:30 UTC 2006 William R. Otte <wotte@dre.vanderbilt.edu> + + * docs/tutorials/Quoter/RTCORBA/Common_Export.h + * docs/tutorials/Quoter/RTCORBA/Stock_PriorityMapping.cpp + * docs/tutorials/Quoter/RTCORBA/Distributor.idl + * docs/tutorials/Quoter/RTCORBA/Common_i.cpp + * docs/tutorials/Quoter/RTCORBA/Broker_i.h + * docs/tutorials/Quoter/RTCORBA/Common_i.h + * docs/tutorials/Quoter/RTCORBA/Stock_PriorityMapping.h + * docs/tutorials/Quoter/RTCORBA/Stock_Database.cpp + * docs/tutorials/Quoter/RTCORBA/StockNameConsumer_i.cpp + * docs/tutorials/Quoter/RTCORBA/Distributor.cpp + * docs/tutorials/Quoter/RTCORBA/Broker.idl + * docs/tutorials/Quoter/RTCORBA/Stock_Database.h + * docs/tutorials/Quoter/RTCORBA/StockNameConsumer_i.h + * docs/tutorials/Quoter/RTCORBA/Common.idl + * docs/tutorials/Quoter/RTCORBA/StockQuoter.mpc + * docs/tutorials/Quoter/RTCORBA/Distributor_i.cpp + * docs/tutorials/Quoter/RTCORBA/Broker.cpp + * docs/tutorials/Quoter/RTCORBA/Distributor_i.h + * docs/tutorials/Quoter/RTCORBA/Admin.cpp + * docs/tutorials/Quoter/RTCORBA/Broker_i.cpp + + This is a major overhaul of the RTCORBA tutorial to fix bugs, + correct a broken priority model, and add new features. + + Thanks to Doug Schmidt for providing code, feedback, and suggestions. + +Wed Aug 16 15:35:29 UTC 2006 Adam Mitz <mitza@ociweb.com> + + * orbsvcs/orbsvcs/HTIOP/HTIOP_Connector.cpp: fixed warning (unused arg) + * tao/Asynch_Queued_Message.h: fixed error in VERSIONED_NAMESPACE build + +Tue Aug 15 21:25:50 UTC 2006 Ciju John <johnc@ociweb.com> + + * tao/PortableServer/Servant_var.h: + * tao/PortableServer/Servant_var.inl: + + Removed the member templates as + - They dodn't seem to be serving any purpose. It appears that the + only reason these member templates were defined was so that the + hand-crafted constructor could be declared to be + explicit. However since the member template constructor is + implicit this just adds complexity without any benefits. + - They were causing build problems in SunStudio 11 builds and + VxWorks builds using gcc 2.96. + + Made the Servant_var constructor implicit to comply with the c++ + language mapping. + +Wed Aug 16 08:18:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/ObjRefTemplate/ObjectReferenceTemplate_include.pidl + * tao/PortableServer/PortableServer_include.pidl + * tao/RTCORBA/RTCORBA_include.pidl + Added empty pragma prefix to be able to compile these pidl files + correctly when the visual age preprocessor is used on AIX + +Wed Aug 16 04:47:24 UTC 2006 Adam Mitz <mitza@ociweb.com> + + * tests/Timed_Buffered_Oneways/client.cpp: + Added #include for ACE_OS::gettimeofday(). + +Tue Aug 15 19:36:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Policies/Policies.mpc: + Simplified this file + +Tue Aug 15 17:57:20 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * TAO_IDL/driver/drv_preproc.cpp (drv_post_init): + + Added check for environment variable CIAO_ROOT when adding + default include paths for CIAO_ROOT and CIAO_ROOT/ciao. This + check is necessary if the middleware SubVersion modules have + been checked out in non-nested mode, since in that case, + CIAO_ROOT cannot be derived from TAO_ROOT or ACE_ROOT. + +Tue Aug 15 17:24:05 UTC 2006 Ossama Othman <ossama_othman at symantec dot com> + + * tao/Strategies/SCIOP_Endpoint.cpp (preferred_interfaces): + + Use new ACE_CString::size_type type to hold string + position/index value, instead of "int". Addresses "comparison + between signed and unsigned" warnings. + +Tue Aug 15 14:56:35 UTC 2006 Adam Mitz <mitza@ociweb.com> + + Checking in on behalf of Justin Michel <michel_j at ociweb dot com>... + + Fixed bugs related to OCI RT8769. There were several problems + identified in this ticket. + + All of these changes are on the branch "OCI_Oneway_Buffering". + + 1. The SyncScopes SYNC_NONE, SYNC_DELAYED_BUFFERING, and + SYNC_EAGER_BUFFERING use non-blocking connects, but the + non-blocking connect mechanism allows the caller to + effectively block, because it enters the reactor. (Even + though it used a timeout of zero, this still makes it + possible for the caller to be blocked.) The solution is to + queue any messages until the connection is complete, and to + rely on the reactor to complete the connection. This was + already implemented, and the fix just involved NOT polling + the reactor. A side-effect of this change, is that + applications that make connections within oneway requests + with the above SyncScope policies must call orb->run() to + complete the connection. Alternatively an invocation that + doesn't use the above SyncScopes can be made to complete + the connection. + + 2. Connection and request timeouts did not work correctly for + every scenario. (e.g. For non-blocking connects, the + request timeout should not be used even if it's less than + the connect timeout, because the connect does not happen in + the context of the request.) The fix required numerous + changes detailed below. + + 3. Queued messages would never be expired even if a roundtrip + request timeout policy was in place. The solution is to + calculate an absolute expiration time for queued + asynchronous messages that have timeouts. The messages are + then discarded in Transport::drain_queue_i() if expired. + + 4. Connection timeouts result in TRANSIENT instead of TIMEOUT. + The fix required numerous changes detailed below. + + 5. Queued messages were flushed at connection startup, which + effectively ignored the buffering constraint policies. + Removed the code that flushed the messages. + + * NEWS: explains user-visible implications of these changes + + * orbsvcs/orbsvcs/FaultTolerance/FT_ClientPolicy_i.cpp: + * tao/Messaging/Connection_Timeout_Policy_i.cpp: + * tao/Messaging/Messaging_Policy_i.cpp: + + Enhanced display of millisecond timeouts in log + messages. Log messages now say "Connect timeout <50ms>" or + "Request timeout <100ms>" instead of "timeout <100>". + + * examples/Buffered_Oneways/client.cpp: + + Add a check to make sure that the test is run with a valid + buffering constraint policy. + + * tao/Connection_Handler.h: + * tao/Connection_Handler.inl: + + Added an is_timeout() method for Connection_Handler() so that + we can tell the difference between a timeout and a close. + + * orbsvcs/orbsvcs/HTIOP/HTIOP_Connector.cpp: + * orbsvcs/orbsvcs/PortableGroup/UIPMC_Connection_Handler.cpp: + * orbsvcs/orbsvcs/SSLIOP/IIOP_SSL_Connector.cpp: + * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Connection_Handler.cpp: + * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Connector.cpp: + * tao/IIOP_Connection_Handler.cpp: + * tao/IIOP_Connector.cpp: + * tao/Strategies/SCIOP_Connection_Handler.cpp: + * tao/Strategies/SCIOP_Connector.cpp: + * tao/Strategies/SHMIOP_Connection_Handler.cpp: + * tao/Strategies/UIOP_Connection_Handler.cpp: + * tao/Strategies/UIOP_Connector.cpp: + + Update log message during IIOP_Connection_Handler::open() to + show ip address and port instead of just ip address. + + Update all connection handlers that closed a connection due + to timeout to set the state to TAO_LF_Event::LFS_TIMEOUT + instead of closed. + + Changed xxx_Connector to NOT set synch_options.timeout() to + ACE_Time_Value::zero when the connect is + non-blocking. Setting it here is unnecessary and prevents + non-blocking timeouts from working. + + Update the IIOP_Connector logged error message to show + errno in addition to the errno description. + + Increment the reference count of the xxx_Connection_Handler + when the xxx_Connector is unable to complete the connection + synchronously. (Same as BugZilla 2417, but for all protocols + instead of just IIOP.) + + * tao/Invocation_Adapter.cpp: + + Move the max_wait_time variable to an inner scope in + Invocation_Adapter::invoke_i(), because it wasn't needed in + most code paths. + + * tao/Profile_Transport_Resolver.cpp: + + Fixed the logic in profile_transport_resolver::resolve() to + correctly detect TIMEOUT exceptions, and to ensure that the + connection timeout uses the minimum of the connection + timeout policy and request timeout for blocking + connects. Also ensure that request timeout is never used as + the connection timeout for non-blocking connects. + + * tao/Asynch_Queued_Message.h: + * tao/Asynch_Queued_Message.cpp: + * tao/Queued_Message.h: + * tao/Queued_Message.cpp: + * tao/Synch_Queued_Message.cpp: + + Add expiration time to Asynch_Queued_Messages. The + constructor now takes an ACE_Time_Value for the relative + timeout. If the timeout is nonzero and nonnull then it is + used to calculate an absolute expiration time using the high + res timer. + + Update Queued_Message constructor to allow specifying an + absolute timeout, and a new is_expired() method to check + whether the queued message is expired. + + + * tao/Synch_Invocation.cpp: + * tao/Transport.h: + * tao/Transport.cpp: + + Update various places in Transport and Synch_Invocation to + create queued messages with the correct timeout value. This + includes ensuring that partially delivered messages can + never timeout. + + Update Transport to discard expired messages. This happens + in drain_queue_i(). + + Add more descriptive logging output to + Transport::cleanup_queue_i. + + Update Transport post_open to NOT flush queued + messages. This effectively ignored the specified buffering + constraint policies. + + * tao/Transport_Connector.cpp: + + Update Transport_Connector to correctly detect timed out + connections. Also fixed the code that checked the return + value of + base_transport->wait_strategy()->register_handler(). This + was recently changed from "if != 0" to "if == 0", but the + correct check for failure should be "if == -1". + + Updated Transport_Connector to not enter the wait strategy + for non-blocking connects. Instead we rely on the + application to eventually call orb->run() which will + complete the connection, and send any queued messages. + + * tao/Transport_Queueing_Strategies.cpp: + + Update Eager_Transport_Queuing_Strategy to not use a pointer + to something that looks like it could go out of scope. I + think the code was actually safe, but it was confusing. + + * tao/Block_Flushing_Strategy.cpp: + * tao/Flushing_Strategy.h: + + Change block_flushing strategy to NOT synchronously flush in + the schedule_output() method. Instead we added a new special + return value indicating that the caller should do the flush + immediately. This allows the calling code to hold a lock + during the call to schedule_output() that can be released + before calling flush_transport(). + + * tests/AMI_Buffering/client.cpp: + * tests/AMI_Buffering/run_buffer_size.pl: + * tests/AMI_Buffering/run_message_count.pl: + * tests/AMI_Buffering/run_test.pl: + * tests/AMI_Buffering/run_timeout.pl: + * tests/AMI_Buffering/run_timeout_reactive.pl: + * tests/Oneway_Buffering/client.cpp: + * tests/Oneway_Buffering/run_buffer_size.pl: + * tests/Oneway_Buffering/run_message_count.pl: + * tests/Oneway_Buffering/run_test.pl: + * tests/Oneway_Buffering/run_timeout.pl: + * tests/Oneway_Buffering/run_timeout_reactive.pl: + * tests/Queued_Message_Test/Queued_Message_Test.cpp: + * tests/Timed_Buffered_Oneways/client.cpp: + * tests/Timed_Buffered_Oneways/run_test.pl: + * tests/Timed_Buffered_Oneways/server.cpp: + * tests/Timed_Buffered_Oneways/test.idl: + * tests/Timed_Buffered_Oneways/test_i.h: + * tests/Timed_Buffered_Oneways/test_i.cpp: + + Updated various tests to comply with the above changes. + + Fixed tests that used sync_none or sync_delayed_buffering + without calling orb->run(). + + * tests/Oneway_Timeouts: + * tests/Oneway_Timeouts/Test.idl: + * tests/Oneway_Timeouts/client.cpp: + * tests/Oneway_Timeouts/run_test.pl: + * tests/Oneway_Timeouts/server.cpp: + * tests/Oneway_Timeouts/test.mpc: + + Added new test for various combinations of timeouts with oneways: + connection timeouts and relative round-trip timeouts. + +Tue Aug 15 14:02:23 UTC 2006 Chad Elliott <elliott_c@ociweb.com> + + * tao/CSD_Framework/CSD_FW_Server_Request_Wrapper.cpp: + + Before cloning the arguments, check to make sure that we have at + least one argument/return value. + +Tue Aug 15 11:37:06 UTC 2006 Chad Elliott <elliott_c@ociweb.com> + + * tao/AnyTypeCode/TypeCode.h: + + Changed the export on In_Object_Argument_Cloner_T from TAO_Export + to TAO_AnyTypeCode_Export. Thanks to Johnny Willemsen for + pointing out the build warnings. + +Tue Aug 15 11:19:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Big_Reply/Client_Task.cpp: + Added some more debug output to better see why this test fails on + OpenVMS. + +Tue Aug 15 10:50:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/RTCORBA/Server_Protocol/run_test.pl: + OpenVMS doesn't have uiop + +Tue Aug 15 10:45:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Bug_2503_Regression/run_test.pl: + Extended wait time for server to let this test succeed on OpenVMS + + * tests/RTCORBA/Client_Protocol/run_test.pl: + Small update for OpenVMS + +Tue Aug 15 08:40:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/tests/Bug_1393_Regression/run_test.pl: + * tests/CollocationLockup/CollocationLockup.mpc: + * tests/Bug_2494_Regression/Bug_2494_Regression.mpc: + * tests/Abstract_Interface/Abstract_Interface.mpc: + * orbsvcs/tests/Trading/Trading.mpc: + * orbsvcs/tests/ImplRepo/Bug_2604_Regression/Bug_2604.mpc: + * orbsvcs/tests/Bug_2248_Regression/Bug_2248_Regression.mpc: + * orbsvcs/orbsvcs/SSLIOP.mpc: + * orbsvcs/orbsvcs/HTIOP.mpc: + * tao/ImR_Client.mpc: + * tao/PI.mpc: + Don't use -Sc as tao_idl option, this doesn't exist anymore + +Tue Aug 15 08:02:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * TAO_IDL/be/be_visitor_operation/ami_handler_reply_stub_operation_cs.cpp: + Const improvement + + * tao/RTPortableServer.mpc: + * tao/RTScheduler.mpc: + Fixed regeneration instructions + + * TAO_IDL/be/be_visitor_ami_pre_proc.cpp: + Cleanup of old stuff + +Mon Aug 14 19:05:38 UTC 2006 Yan Dai <dai_y@ociweb.com> + + * tao/PortableServer/Object_Adapter.h: + * tao/PortableServer/Object_Adapter.cpp: + + Made the TAO_Object_Adapter keep a reference to POAManager + object on minimum corba builds to fix leaking of POAManager object. + +Mon Aug 14 17:21:00 UTC 2006 Yan Dai <dai_y@ociweb.com> + + * tao/PortableServer/Object_Adapter.cpp: + + Fixed a bug in open() on minimum corba builds. The problems is + the new poa_manager passed to create_Root_POA() was actually + destructed when exit from open() function. This would cause + access violation when application access the poa manager in + the Root_POA. + +Mon Aug 14 13:45:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/*.h: + Added /**/ to the include of the export and versioned + namespace files to get them out of the doxygen include + pictures + +Mon Aug 14 13:11:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/PortableServer.mpc: + Enabed regeneration of PortableServer.pidl + + * tao/PortableServer/PortableServer{A,C,S}.{h,cpp}: + * tao/PortableServer/diffs/PortableServer.diff: + Removed from the repo + +Mon Aug 14 12:59:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/CodecFactory.mpc: + * tao/ObjRefTemplate.mpc: + * tao/PortableServer.mpc: + * tao/RTCORBA.mpc: + * tao/RTPortableServer.mpc: + * tao/RTScheduler.mpc: + * tao/PI_Server.mpc: + Enabed regeneration of *_include.pidl files, this will result + in an almost empty C.cpp file but that is needed to retrigger + the regeneration of the header files. + + * tao/CodecFactory/IOP_Codec_include{A,C,S}.h: + * tao/ObjRefTemplate/ObjectReferenceTemplate_include{A,C,S}.h: + * tao/PI_Server/PI_Server_include{A,C,S}.h: + * tao/PortableServer/PortableServer_include{A,C,S}.h: + * tao/RTCORBA/RTCORBA_include{A,C,S}.h: + * tao/RTPortableServer/RTPortableServer_include{A,C,S}.h: + * tao/RTScheduling/RTScheduler_include{A,C,S}.h: + * tao/RTScheduling/diffs/RTScheduler.diff: + * tao/RTPortableServer/RTPortableServerA.h: + Removed from the archive + +Mon Aug 14 12:49:11 UTC 2006 Chad Elliott <elliott_c@ociweb.com> + + * TAO_IDL/be/be_global.cpp: + * TAO_IDL/be/be_visitor_operation/ami_cs.cpp: + * TAO_IDL/be/be_visitor_operation/operation.cpp: + * TAO_IDL/be_include/be_global.h: + + Added a new option (-b) which controls the use of clonable in + arguments on oneway calls. + + * tao/AnyTypeCode/TypeCode.h: + * tao/AnyTypeCode/TypeCode.cpp: + * tao/Argument.h: + * tao/Argument.cpp: + * tao/BD_String_Argument_T.h: + * tao/BD_String_Argument_T.inl: + * tao/BD_String_Argument_T.cpp: + * tao/Basic_Argument_T.h: + * tao/Basic_Argument_T.inl: + * tao/Basic_Argument_T.cpp: + * tao/Basic_Arguments.h: + * tao/CSD_Framework/CSD_FW_Server_Request_Wrapper.cpp: + * tao/Fixed_Array_Argument_T.h: + * tao/Fixed_Array_Argument_T.inl: + * tao/Fixed_Array_Argument_T.cpp: + * tao/Fixed_Size_Argument_T.h: + * tao/Fixed_Size_Argument_T.inl: + * tao/Fixed_Size_Argument_T.cpp: + * tao/Object.h: + * tao/Object.cpp: + * tao/Object_Argument_T.h: + * tao/Object_Argument_T.inl: + * tao/Object_Argument_T.cpp: + * tao/Special_Basic_Argument_T.h: + * tao/Special_Basic_Argument_T.inl: + * tao/Special_Basic_Argument_T.cpp: + * tao/UB_String_Argument_T.h: + * tao/UB_String_Argument_T.inl: + * tao/UB_String_Argument_T.cpp: + * tao/Var_Array_Argument_T.h: + * tao/Var_Array_Argument_T.inl: + * tao/Var_Array_Argument_T.cpp: + * tao/Var_Size_Argument_T.h: + * tao/Var_Size_Argument_T.inl: + * tao/Var_Size_Argument_T.cpp: + + Created "clonable" template classes based on all In_*_Argument_T. + Through the use of traits, the clonable templates are not + instantiated unless the idl generated code uses them (through the + -b option). + + Code in the CSD_FW_Server_Request is able to determine if the + arguments (for collocated oneway's) can be cloned or not. If they + can be, they will be cloned during the copy of the server request. + If they can not be cloned, it will use the original method of + marshaling the pararameters to copy them. [RT 8372] + + * tests/CSD_Strategy_Tests/TP_Foo_B/csd_tp_foo_b_lib.mpc: + * tests/CSD_Strategy_Tests/TP_Foo_C/csd_tp_foo_c_lib.mpc: + + Use the -b option for the idl files in these projects. I left + some CSD tests to not use the -b option to exercise both methods + of argument copying. + +Mon Aug 14 11:39:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/AnyTypeCode/TypeCode_Case_Enum_T.{h,cpp,inl}: + Fixed gcc warnings + +Mon Aug 14 10:43:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * TAO_IDL/be/be_codegen.cpp: + * TAO_IDL/be_include/be_codegen.h: + Part of fixes for bug 2605, when generating the include of + config-all.h, Versioned_Namespace.h and the export include + do generate /**/ between the include statement and the filename + so that doxygen doesn't add these standard includes to the + include picture to make them more readable. + +Mon Aug 14 09:50:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * docs/compiler.html: + * TAO_IDL/be/be_codegen.cpp: + * TAO_IDL/be/be_global.cpp: + * TAO_IDL/be_include/be_global.h: + Added -Wb,unique_include as option, when this is specifified + instead of the normal includes only the specified include is + done. This is needed for regeneration the *_include.pidl files + in the archive as part of the regular build + + * TAO_IDL/be/be_visitor_component/component_ch.cpp: + * TAO_IDL/be/be_visitor_interface/interface_ch.cpp: + Generate false instead of 0 + +Mon Aug 14 09:07:12 UTC 2006 Martin Corino <mcorino@remedy.nl> + + * examples/Simulator/Event_Supplier/Logging_Sup.cpp: + Small OpenVMS compatibility changes. + +Mon Aug 14 08:35:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Messaging/Messaging.h: + * tao/Messaging/messaging_export.h: + Moved the undef of TRANSPART to the export file to fix + compile errors on solaris. This way we don't need to handcraft + generated files for the messaging lib + +Mon Aug 14 07:44:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/AnyTypeCode/TypeCode_Case_Enum_T.{h,cpp,inl}: + * tao/AnyTypeCode/TypeCode_CDR_Extraction.cpp: + Fixed handling of union with enums. This fixes bugzilla bug + 2446, thanks to Tommy Persson <tpe at ida dot liu dot se> for + reporting this bug. Also it fixes bugzilla bug 2539 which + has been reported by sdjiangwei@sina.com. The real fix has + been made by Dale Boan <boan at sedsystems dot ca>, many thanks + to him. + +Sun Aug 13 23:30:02 UTC 2006 Yan Dai <dai_y@ociweb.com> + + These changes excluded the POAManagerFactory and EndpointPolicy + from minimum corba, minimum poa and CORBA/e builds. + + * tao/PortableServer/POA.pidl: + + Excluded the_POAManagerFactory attribute in POA. + + * tao/PortableServer/POAManagerFactory.pidl: + + Excluded POAManagerFactory interface. + + * tao/PortableServer/POAManager.cpp: + * tao/PortableServer/POAManager.inl: + * tao/PortableServer/POAManager.h: + * tao/PortableServer/Object_Adapter.cpp: + * tao/PortableServer/Object_Adapter.h: + * tao/PortableServer/Acceptor_Filter_Factory.cpp: + * tao/PortableServer/Acceptor_Filter_Factory.h: + * tao/PortableServer/POAManagerFactory.cpp: + * tao/PortableServer/POAManagerFactory.h: + * tao/PortableServer/Root_POA.cpp: + * tao/PortableServer/Root_POA.h: + + Excluded the use of POAManagerFactory in the code. + + * tao/EndpointPolicy.mpc: + * tests/POA/EndpointPolicy/EndpointPolicy.mpc: + * tests/POA/POAManagerFactory/POAManagerFactory.mpc: + + Made these projects inherit from minimum_corba. + +Fri Aug 11 12:49:55 UTC 2006 Ciju John <johnc@ociweb.com> + + * tests/Optimized_Connection/Optimized_Connection.mpc: + * tests/Optimized_Connection/README: + * tests/Optimized_Connection/Test.idl: + * tests/Optimized_Connection/oc.ior: + + This test was failing on OpenVMS due to the way the linker + handles template instantiations. Copied the Hello test IDL here, + so the test is now self-contained. + +Fri Aug 11 11:55:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/RTPortableServer/RT_POA.{h,cpp}: + CORBA/e changes + + * tao/Utils/Auto_Function.h: + Marked this header as deprecated + +Fri Aug 11 09:10:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/IFR_Client.mpc: + Added -GX to the generation of the pidl files to get empty + A.h files + + * tao/IFR_Client/*A.h: + Removed from the repo, do get regenerated now automatically + +Fri Aug 11 07:35:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * TAO_IDL/be/be_global.cpp: + * TAO_IDL/be/be_produce.cpp: + * TAO_IDL/be_include/be_global.h: + * docs/compiler.html: + Added -Scc to suppress generation of client stub and -Ssc of + generation of server skeleton file. These options don't check + if something is generated in these files or not, they just + suppress the generation. Usefull for some pidl files in the + core libraries. This fixes bugzilla 2631 + +Fri Aug 11 07:35:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/orbsvcs/Sched/Reconfig_Scheduler_T.cpp: + Initialise pointer with 0 to address gcc 4.1 warning + +Fri Aug 11 07:09:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Messaging/Messaging_Types.pidl: + Moved all policy type constants to this new pidl file that is + included by the other pidl files. This way the catior utility + only needs to include the generated client file for this pidl + file. + + * tao/Messaging/Messaging_No_Impl.pidl: + * tao/Messaging/Messaging_RT_Policy.pidl: + * tao/Messaging/Messaging_SyncScope_Policy.pidl: + Moved the policy constants out of these file and included the + new file above + + * tao/Messaging.mpc: + Added new pidl file + + * tao/utils/catior/catior.cpp: + Include Messaging_TypesC.h instead of all seperate messaging + files. + +Thu Aug 10 12:20:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Messaging.mpc: + Enabled regeneration of Messaging_No_Impl.pidl. This had a diff + to undef TRANSPARENT if it was defined, we moved that to + Messaging.h + + * tao/Messaging/diffs/*: + Removed, not needed anymore + + * tao/Messaging/Messaging.h: + Undef TRANSPARENT when it is defined + + * tao/Messaging/Messaging.pidl: + Documentation cleanup + + * tao/Messaging/Messaging_No_Impl{A,C,S}.{h,cpp,inl}: + Removed, will get regenerated as part of the build + +Thu Aug 10 12:03:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * TAO_IDL/be/be_codegen.cpp: + In the server header include the safe include when this is defined + + * tao/Messaging.mpc: + Enabled regeneration of Messaging.pidl and added some -GX flags to + get the empty A.h files + + * tao/Messaging/Messaging_No_ImplS.h: + Added empty file, needed because Messaging.pidl gets regenerated + + * tao/Messaging/Messaging.pidl: + Removed include of OldExceptionHolder.pidl which got removed some + time ago. + + * tao/Messaging/Messaging{A,C,S}.{h,cpp,inl}: + Removed because they get regenerated as part of the build + +Thu Aug 10 10:29:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * TAO_IDL/be/be_global.cpp: + * TAO_IDL/be_include/be_global.h: + * TAO_IDL/be/be_codegen.cpp: + * docs/compiler.html + Added -GX option, generates an empty A.h file. This can be used + for the pidl files in the archive which can't be build with -GA + but we do need an A.h file to not get compile errors when end + users do include the pidl files directly. + + * tao/CodecFactory.mpc: + * tao/ImR_Client.mpc: + * tao/ObjRefTemplate.mpc: + * tao/PI.mpc: + * tao/RTScheduler.mpc: + * tao/IORManipulation.mpc: + Use the new -GX option + + * tao/CodecFactory/IOP_CodecA.h: + * tao/ImR_Client/ImplRepoA.h: + * tao/IORManipulation/IORA.h: + * tao/ObjRefTemplate/ObjectReferenceTemplateA.h: + * tao/PI/ORBInitInfoA.h: + * tao/RTScheduling/RTSchedulerA.h: + Removed from the repo, do get generated now with -GX. + +Thu Aug 10 09:43:56 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/CodecFactory/IOP_Codec.pidl: + * tao/CSD_Framework/CSD_Framework.pidl: + Documentation improvements + + * tao/Connector_Registry.cpp: + Improved some debug statements to mention the class and method + where the debug message is coming from + + * tao/CSD_Framework/CSD_FW_Server_Request_Wrapper.cpp: + Initialize some pointers with 0 + + * tao/CSD_Framework/CSD_FW_Server_Request_Wrapper.cpp: + * tao/Invocation_Adapter.cpp: + * tao/LocateRequest_Invocation_Adapter.cpp: + * tao/operation_details.{h,inl}: + Removed the argument_flag as member, we just determine whether + we have arguments or not on the number of arguments. Changed the + accessor to just check it when requested. Introduced a flag + whether the stub arguments should be used or not, this can + then be changed is some part of the code knows the stub + arguments are not valid, for example with AMI. + + * tao/operation_details.h: + Guarded the ft_* methods with TAO_HAS_INTERCEPTORS, the + implementation did use the macro, just not the declaration + + * tao/extra_core.mpb: + Removed a generated file from the list of files + + * tao/GIOP_Message_Generator_Parser_12.cpp: + * tao/Service_Context.{h,inl}: + is_service_id is now returning a real bool + + * tao/Messaging/Async_Invocation_Adapter.{h,cpp}: + Overruled invoke_collocated_i to set the use_stub_args + flag in the operations details to false, when using + collocation we should use the skeleton arguments. This fixes + bugzilla bug 2503, thanks to Charles Frasch <cfrash at atdesk dot com> + for reporting this bug. + + * tao/PortableServer/get_arg.h: + Use the operation_details stub_args method to determine + which args should be used + + * TAO_IDL/be/be_visitor_operation/ami_cs.cpp: + Small optimization for AMI operations with just out arguments as + supplied by Ossama Othman as partly fix for bug 2503. + + * tao/PortableServer/Root_POA.cpp: + Don't use POAManagerFactory with CORBA/e compact. + +Wed Aug 9 15:00:55 UTC 2006 Adam Mitz <mitza@ociweb.com> + + * orbsvcs/tests/ImplRepo/Bug_2604_Regression/Bug_2604.mpc: + * orbsvcs/tests/ImplRepo/Bug_2604_Regression/MessengerServer.cpp: + Added dependency on ImR_Client library so that static builds can make + use of ImR features without dynamic library loading. + +Tue Aug 8 22:39:56 UTC 2006 Ciju John <johnc@ociweb.com> + + * tao/CSD_ThreadPool.mpc: + Replaced the requires for threads with dependency upon + threads.mpb which does essentially the same thing. Only looks + cleaner. + +Wed Aug 9 13:18:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Bug_2503_Regression/run_test.pl: + No need to run the client with orb debug level 10 + +Wed Aug 9 08:17:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/RTCORBA/RT_Protocol_Hooks.cpp: + Use true instead of 1 + + * tao/RTCORBA/RTCORBA.pidl: + Doxygen changes + +Tue Aug 8 14:55:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/BiDIR_GIOP/BiDir_PolicyFactory.cpp: + No need to extract the value twice out of the any + +Tue Aug 8 14:13:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/tests/Codesets/libs/IBM1047_ISO8859/Char_IBM1047_ISO8859_Factory.h: + * tao/tests/Codesets/libs/IBM1047_ISO8859/Wchar_IBM1047_ISO8859_Factory.h: + * tao/tests/Codesets/libs/UCS4_UTF16/WUCS4_UTF16_Factory.h: + Updated includes because of the removal of Codeset_Translator_Factory.h + from the core TAO directory + +Tue Aug 8 09:38:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * TAO_IDL/be/be_visitor_typecode/typecode_defn.cpp: + Initailize some pointers explicitly with 0 to fix gcc41 warnings + +Tue Aug 8 09:33:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * TAO_IDL/ast/ast_interface.cpp: + * TAO_IDL/ast/ast_type.cpp: + Initailize some pointers explicitly with 0 to fix gcc41 warnings + +Tue Aug 8 08:30:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Codeset_Translator_Factory.h: + Removed, it was deprecated for more then a year + + * tao/tao.mpc: + Removed remark about Codeset_Translator_Factory.h + +Tue Aug 8 07:53:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * TAO-INSTALL.html: + Removed vc6 info, added VxWorks 6.3 as supported platform + +Tue Aug 8 07:49:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * TAO_IDL/be/be_visitor_operation/operation.cpp: + Fixed typo in comment + +Mon Aug 7 19:29:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/EndpointPolicy.mpc: + Enabed regeneration of EndpointPolicyType.pidl: + + * tao/EndpointPolicy/EndpointPolicy{A,C}.{h,cpp}: + Removed, not needed anymore + + * tao/EndpointPolicy/EndpointPolicyType.pidl: + Include Policy_Forward.pidl, Policy.pidl pulls in too + much. + + * tao/BiDIR_GIOP/BiDIR_GIOP.{h,cpp}: + Changed is_activated to a bool + + * tao/Profile.cpp (decode): + Only get encap_len when ACE_NLOGGING is not defined, if it is + defined then we don't use this variable, so no need to get it. + +Mon Aug 7 17:06:52 UTC 2006 Yan Dai <dai_y@ociweb.com> + + * tao/PortableServer/POAManagerFactory.cpp: + + Removed hard coded policy checking in create_POAManager(). + Instead, the TAO_Policy_Validator is used to validate the + policies. + + * tao/orbconf.h: + + Fixed a bug that the TAO_CACHED_POLICY_ENDPOINT was set + after TAO_CACHED_POLICY_MAX_CACHED. This would fix the + access beyond cached_policies_ array bound problem. + + * tao/EndpointPolicy/EndpointPolicy_Factory.cpp: + + Fixed a bug in TAO_EndpointPolicy_Factory::create_policy() + that initializes the local variable "found_one" to true which + makes the supplied endpoints are never verified against + the ORB listening endpoints. + + * tao/EndpointPolicy/Endpoint_Acceptor_Filter.cpp: + * tao/EndpointPolicy/Endpoint_Acceptor_Filter.h: + * tao/EndpointPolicy/Endpoint_Acceptor_Filter_Factory.cpp: + + Fixed a bug that does not count all endpoints if multiple + endpoint policies are supplied. Previously, it just counted + the endpoints in the first endpoint policy. + + * tao/EndpointPolicy/EndpointPolicy_i.cpp: + * tao/EndpointPolicy/EndpointPolicy_i.h: + + Added virtual destructor and reset endpoint_list to length 0 + when the policy is destroyed. + + * tao/EndpointPolicy/IIOPEndpointValue_i.cpp: + + Removed the duplicate of the const char* ("") since the + CORBA::String_var constructor duplicates the string internally. + +Mon Aug 7 11:41:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/IORManipulation/IOR.pidl: + Converted documentation to doxygen style + +Fri Aug 4 09:06:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/default_resource.{h,cpp}: + * tao/IIOP_Transport.{h,cpp}: + * tao/MMAP_Allocator.{h,cpp}: + * tao/orbconf.h: + * tao/Transport.{h,cpp}: + * docs/Options.html: + Added TAO_HAS_SENDFILE, this is set to 1 when ACE_HAS_SENDFILE + is enabled, else it is set to 0. Users can explicitly define + this to 0 also to build TAO without sendfile support, this + reduces then the footprint for applications running on a + platform that supports sendfile but where the app doesn't + use it. This fixes bugzilla bug 2587. + +Wed Aug 2 22:47:53 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * NEWS: + + Added a note about the valuetype fix. + +Wed Aug 2 21:51:51 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/Valuetype/ValueBase.cpp: + * tao/Valuetype/ValueBase.h: + + Restore the ability to properly interpret indirect repository + IDs for valuetypes to improve interoperability with other ORBs, + notably JacORB. Secondarily, I've renamed some internal + (private) methods to include the prefix _tao_ so as to avoid + potential namespace collisions with names generated by the IDL + compiler. + + Thanks to Eric Danielou <eric dot danielou at eads dot com> + for identifying this flaw. + +Wed Aug 2 21:32:27 UTC 2006 Adam Mitz <mitza@ociweb.com> + + * orbsvcs/examples/ImR/Advanced/run_test.pl: + * orbsvcs/tests/CosEvent/Timeout/run_test.pl: + * orbsvcs/tests/ImplRepo/scale/run_test.pl: + * orbsvcs/tests/Notify/Reconnecting/run_test.pl: + * orbsvcs/tests/Notify/XML_Persistence/run_test.pl: + + In each case, the Perl scripts were depending on the fact that + $TAO_ROOT and $ACE_ROOT/TAO were the same value. These changes + make use of the new $PerlACE::TAO_ROOT variable so that the tests + will run when $TAO_ROOT is not $ACE_ROOT/TAO but still assume that + $TAO_ROOT is $ACE_ROOT/TAO if $TAO_ROOT is not set. + +Wed Aug 2 20:15:37 UTC 2006 Adam Mitz <mitza@ociweb.com> + + * orbsvcs/ImplRepo_Service/ImR_Activator_i.cpp: + Updated comment to reflect what's actually in the code. + + * orbsvcs/ImplRepo_Service/ImR_Locator_i.h: + * orbsvcs/ImplRepo_Service/ImR_Locator_i.cpp: + + This is a fix for bugzilla bug #2604. This only applies to + per_client activation mode. The ImR may receive + "server_is_running()" invocations from servers that it did + not launch (they may have been launched by a prior instance + of the ImR or manually). When this happens the ImR records + the info about the running server and places it on a queue. + Later, when a client request comes in a new server is + launched (as it should be), but the old queued server info is + used to ping it. This can fail with various exceptions + depending on if the old process is still alive or if some + other process is now running on its port. If it fails enough + times (greater than the retry count) the client invocation + will fail. The fix is simply to keep track of whether or not + the ImR expects a new server (that it launched) to be coming + online. If not, it can ignore the server_is_running(). This + mechanism is already in place for normal activation, but due + to differences in per_client it doesn't work there. + + * orbsvcs/tests/ImplRepo/Bug_2604_Regression: + * orbsvcs/tests/ImplRepo/Bug_2604_Regression/Bug_2604.mpc: + * orbsvcs/tests/ImplRepo/Bug_2604_Regression/Messenger.idl: + * orbsvcs/tests/ImplRepo/Bug_2604_Regression/MessengerClient.cpp: + * orbsvcs/tests/ImplRepo/Bug_2604_Regression/MessengerServer.cpp: + * orbsvcs/tests/ImplRepo/Bug_2604_Regression/Messenger_i.h: + * orbsvcs/tests/ImplRepo/Bug_2604_Regression/Messenger_i.cpp: + * orbsvcs/tests/ImplRepo/Bug_2604_Regression/README: + * orbsvcs/tests/ImplRepo/Bug_2604_Regression/run_test.pl: + + New regeression test for the bugfix described above. + +Wed Aug 2 15:11:08 UTC 2006 Simon McQueen <sm@prismtech.com> + + * orbsvcs/orbsvcs/FaultTolerance/FT_ClientRequest_Interceptor.h: + * orbsvcs/orbsvcs/FaultTolerance/FT_ClientRequest_Interceptor.cpp: + + Do not transmit a FT_REQUEST service context if the destination is + not an IOGR (i.e. does not contain a IOP::TAG_FT_GROUP component). + + * orbsvcs/tests/Bug_2615_Regression/Bug_2615_Regression.mpc: + * orbsvcs/tests/Bug_2615_Regression/Hello.h: + * orbsvcs/tests/Bug_2615_Regression/Hello.cpp: + * orbsvcs/tests/Bug_2615_Regression/README: + * orbsvcs/tests/Bug_2615_Regression/ServerORBInitializer2.h: + * orbsvcs/tests/Bug_2615_Regression/ServerORBInitializer2.cpp: + * orbsvcs/tests/Bug_2615_Regression/ServerRequest_Interceptor2.h: + * orbsvcs/tests/Bug_2615_Regression/ServerRequest_Interceptor2.cpp: + * orbsvcs/tests/Bug_2615_Regression/Test.idl: + * orbsvcs/tests/Bug_2615_Regression/client.cpp: + * orbsvcs/tests/Bug_2615_Regression/run_test.pl: + * orbsvcs/tests/Bug_2615_Regression/server.cpp: + + Regression test for the above. + +Wed Aug 2 13:03:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/DynamicInterface/DII_Invocation_Adapter.{h,cpp}: + Doxygen improvements + +Wed Aug 2 12:58:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * performance-tests/Cubit/TAO/IDL_Cubit/run_test.pl: + Don't run the UIOP test on OpenVMS + +Tue Aug 1 07:06:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/IDL_Test/Bug_2619_Regression.idl: + * tests/IDL_Test/IDL_Test.mpc: + New regression for bug 2619, confirms that this is fixed. Thanks + to Graeme Bell <graemeb at econz dot com> for reporting this. + +Fri Jul 28 18:54:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/IDL_Test/Bug_2616_Regression.idl: + Made the types unique within the IDL_Test project + +Fri Jul 28 15:38:11 UTC 2006 William R. Otte <wotte@dre.vanderbilt.edu> + + * orbsvcs/orbsvcs/Makefile.am + * orbsvcs/tests/Event/lib/Makefile.am + * orbsvcs/tests/Security/Crash_Test/Makefile.am + * orbsvcs/tests/Security/MT_SSLIOP/Makefile.am + * orbsvcs/tests/EC_Mcast/Makefile.am + * orbsvcs/tests/HTIOP/BiDirectional/Makefile.am + * orbsvcs/examples/Security/Send_File/Makefile.am + * tao/Makefile.am + * performance-tests/POA/Demux/Makefile.am + * performance-tests/RTCorba/Oneways/Reliable/Makefile.am + * performance-tests/RTCorba/Multiple_Endpoints/Common/Makefile.am + * examples/Kokyu_dsrt_schedulers/fp_example/Makefile.am + * examples/Kokyu_dsrt_schedulers/mif_example/Makefile.am + * examples/Kokyu_dsrt_schedulers/muf_example/Makefile.am + + Fixed Makefile.ams to reflect changes from + Thu Jul 27 20:50:17 UTC 2006 William R. Otte <wotte@dre.vanderbilt.edu> + +Fri Jul 28 14:29:53 UTC 2006 William R. Otte <wotte@dre.vanderbilt.edu> + + * Most remaining run_test.pl and run_all.pl scripts + + These scripts were missing from + Fri Jul 28 00:44:26 UTC 2006 William R. Otte <wotte@dre.vanderbilt.edu>, + as they used " instead of ' in their use lib line. + +Fri Jul 28 14:08:35 UTC 2006 William R. Otte <wotte@dre.vanderbilt.edu> + + * TAO_IDL/fe/fe_lookup.cpp + + Rolled this file back a version, accidentally got pulled in by + Thu Jul 27 20:50:17 UTC 2006 William R. Otte <wotte@dre.vanderbilt.edu> + + * TAO_IDL/fe/lex.yy.cpp + + Re committing this file to update the timestamp. + +Fri Jul 28 12:17:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/orbsvcs/RTCosScheduling/RTCosScheduling_ServerScheduler_i.inl: + * orbsvcs/orbsvcs/RTCosScheduling/RTCosScheduling_PCP_Manager.inl: + Fixed file tag + +Fri Jul 28 10:17:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/IDL_Test/Bug_2616_Regression.idl: + * tests/IDL_Test/IDL_Test.mpc: + Added regression for bug 2616. Thanks to Keith Nicewarner + <Keith dot Nicewarner at SpaceDev dot com> for delivering this + test case, it failed with 1.5.2 but works on the head version, + just storing it to prevent the bug to appear in the future. + +Fri Jul 28 00:44:26 UTC 2006 William R. Otte <wotte@dre.vanderbilt.edu> + + * Most run_test.pl scripts in the TAO repository + + Updated to use the ACE_ROOT environment variable to find + PerlACE. This is necessary for the flattened ACE/TAO/CIAO + Thanks to Adam Mitz for reporting the problem and + suggesting a solution. + +Thu Jul 27 20:50:17 UTC 2006 William R. Otte <wotte@dre.vanderbilt.edu> + + * orbsvcs/orbsvcs/Scheduler_Factory.inl + * orbsvcs/orbsvcs/Event_Utilities.inl + * orbsvcs/orbsvcs/Scheduler_Utilities.h + * orbsvcs/orbsvcs/Scheduler_Utilities.i + * orbsvcs/orbsvcs/Metrics/Metrics_FrameManager_T.inl + * orbsvcs/orbsvcs/Metrics/Metrics_FrameManager_T.cpp + * orbsvcs/orbsvcs/Metrics/Metrics_UpcallMonitor_T.cpp + * orbsvcs/orbsvcs/Metrics/Metrics_UpcallMonitor_T.inl + * orbsvcs/orbsvcs/Metrics/Metrics_FrameManager_T.h + * orbsvcs/orbsvcs/Metrics/Metrics_FrameManager_T.i + * orbsvcs/orbsvcs/Metrics/Metrics_LocalCache.cpp + * orbsvcs/orbsvcs/Metrics/Metrics_LocalCache.inl + * orbsvcs/orbsvcs/Metrics/Metrics_UpcallMonitor_T.h + * orbsvcs/orbsvcs/Metrics/Metrics_UpcallMonitor_T.i + * orbsvcs/orbsvcs/Metrics/Metrics_LocalCache.h + * orbsvcs/orbsvcs/Metrics/Metrics_LocalCache.i + * orbsvcs/orbsvcs/Metrics/Metrics_FrameManager.cpp + * orbsvcs/orbsvcs/Metrics/Metrics_FrameManager.inl + * orbsvcs/orbsvcs/Metrics/Metrics_UpcallMonitor.cpp + * orbsvcs/orbsvcs/Metrics/Metrics_UpcallMonitor.inl + * orbsvcs/orbsvcs/Metrics/Metrics_FrameManager.h + * orbsvcs/orbsvcs/Metrics/Metrics_FrameManager.i + * orbsvcs/orbsvcs/Metrics/Metrics_UpcallMonitor.h + * orbsvcs/orbsvcs/Metrics/Metrics_UpcallMonitor.i + * orbsvcs/orbsvcs/Metrics/Metrics_Utils.cpp + * orbsvcs/orbsvcs/Metrics/Metrics_Utils.inl + * orbsvcs/orbsvcs/Metrics/Metrics_Utils.h + * orbsvcs/orbsvcs/Metrics/Metrics_Utils.i + * orbsvcs/orbsvcs/Time_Utilities.h + * orbsvcs/orbsvcs/Time_Utilities.i + * orbsvcs/orbsvcs/HTIOP/HTIOP_Acceptor.cpp + * orbsvcs/orbsvcs/HTIOP/HTIOP_Acceptor.inl + * orbsvcs/orbsvcs/HTIOP/HTIOP_Acceptor.h + * orbsvcs/orbsvcs/HTIOP/HTIOP_Acceptor.i + * orbsvcs/orbsvcs/HTIOP/HTIOP_Endpoint.cpp + * orbsvcs/orbsvcs/HTIOP/HTIOP_Endpoint.inl + * orbsvcs/orbsvcs/HTIOP/HTIOP_Endpoint.h + * orbsvcs/orbsvcs/HTIOP/HTIOP_Endpoint.i + * orbsvcs/orbsvcs/Sched/DynSched.cpp + * orbsvcs/orbsvcs/Sched/DynSched.inl + * orbsvcs/orbsvcs/Sched/DynSched.h + * orbsvcs/orbsvcs/Sched/DynSched.i + * orbsvcs/orbsvcs/Sched/SchedEntry.inl + * orbsvcs/orbsvcs/Sched/SchedEntry.cpp + * orbsvcs/orbsvcs/Sched/SchedEntry.h + * orbsvcs/orbsvcs/Sched/SchedEntry.i + * orbsvcs/orbsvcs/Scheduler_Factory.h + * orbsvcs/orbsvcs/Scheduler_Factory.i + * orbsvcs/orbsvcs/Channel_Clients_T.cpp + * orbsvcs/orbsvcs/RTCosScheduling/RTCosScheduling_PCP_Manager.h + * orbsvcs/orbsvcs/RTCosScheduling/RTCosScheduling_PCP_Manager.i + * orbsvcs/orbsvcs/RTCosScheduling/RTCosScheduling_ServerScheduler_i.h + * orbsvcs/orbsvcs/RTCosScheduling/RTCosScheduling_ServerScheduler_i.i + * orbsvcs/orbsvcs/RTCosScheduling/RTCosScheduling_PCP_Manager.cpp + * orbsvcs/orbsvcs/RTCosScheduling/RTCosScheduling_PCP_Manager.inl + * orbsvcs/orbsvcs/RTCosScheduling/RTCosScheduling_ServerScheduler_i.cpp + * orbsvcs/orbsvcs/RTCosScheduling/RTCosScheduling_ServerScheduler_i.inl + * orbsvcs/orbsvcs/PortableGroup/UIPMC_Acceptor.h + * orbsvcs/orbsvcs/PortableGroup/UIPMC_Acceptor.i + * orbsvcs/orbsvcs/PortableGroup/UIPMC_Endpoint.inl + * orbsvcs/orbsvcs/PortableGroup/UIPMC_Acceptor.cpp + * orbsvcs/orbsvcs/PortableGroup/UIPMC_Endpoint.h + * orbsvcs/orbsvcs/PortableGroup/UIPMC_Endpoint.i + * orbsvcs/orbsvcs/PortableGroup/UIPMC_Endpoint.cpp + * orbsvcs/orbsvcs/PortableGroup/UIPMC_Acceptor.inl + * orbsvcs/orbsvcs/Event_Utilities.h + * orbsvcs/orbsvcs/Event_Utilities.i + * orbsvcs/orbsvcs/ESF/ESF_Delayed_Changes.h + * orbsvcs/orbsvcs/ESF/ESF_Delayed_Changes.i + * orbsvcs/orbsvcs/ESF/ESF_Delayed_Command.cpp + * orbsvcs/orbsvcs/ESF/ESF_Delayed_Command.inl + * orbsvcs/orbsvcs/ESF/ESF_Proxy_List.cpp + * orbsvcs/orbsvcs/ESF/ESF_Peer_Workers.h + * orbsvcs/orbsvcs/ESF/ESF_RefCount_Guard.h + * orbsvcs/orbsvcs/ESF/ESF_Proxy_List.inl + * orbsvcs/orbsvcs/ESF/ESF_Peer_Workers.i + * orbsvcs/orbsvcs/ESF/ESF_Proxy_RefCount_Guard.h + * orbsvcs/orbsvcs/ESF/ESF_RefCount_Guard.i + * orbsvcs/orbsvcs/ESF/ESF_Proxy_RefCount_Guard.i + * orbsvcs/orbsvcs/ESF/ESF_Copy_On_Write.inl + * orbsvcs/orbsvcs/ESF/ESF_Copy_On_Write.cpp + * orbsvcs/orbsvcs/ESF/ESF_Shutdown_Proxy.h + * orbsvcs/orbsvcs/ESF/ESF_Shutdown_Proxy.i + * orbsvcs/orbsvcs/ESF/ESF_Proxy_RB_Tree.h + * orbsvcs/orbsvcs/ESF/ESF_Proxy_RB_Tree.i + * orbsvcs/orbsvcs/ESF/ESF_Proxy_Admin.cpp + * orbsvcs/orbsvcs/ESF/ESF_Proxy_Admin.inl + * orbsvcs/orbsvcs/ESF/ESF_Busy_Lock.cpp + * orbsvcs/orbsvcs/ESF/ESF_Busy_Lock.inl + * orbsvcs/orbsvcs/ESF/ESF_Immediate_Changes.h + * orbsvcs/orbsvcs/ESF/ESF_Immediate_Changes.i + * orbsvcs/orbsvcs/ESF/ESF_Delayed_Changes.cpp + * orbsvcs/orbsvcs/ESF/ESF_Delayed_Changes.inl + * orbsvcs/orbsvcs/ESF/ESF_RefCount_Guard.cpp + * orbsvcs/orbsvcs/ESF/ESF_Peer_Workers.cpp + * orbsvcs/orbsvcs/ESF/ESF_Peer_Workers.inl + * orbsvcs/orbsvcs/ESF/ESF_RefCount_Guard.inl + * orbsvcs/orbsvcs/ESF/ESF_Proxy_RefCount_Guard.cpp + * orbsvcs/orbsvcs/ESF/ESF_Proxy_RefCount_Guard.inl + * orbsvcs/orbsvcs/ESF/ESF_Delayed_Command.h + * orbsvcs/orbsvcs/ESF/ESF_Delayed_Command.i + * orbsvcs/orbsvcs/ESF/ESF_Shutdown_Proxy.cpp + * orbsvcs/orbsvcs/ESF/ESF_Shutdown_Proxy.inl + * orbsvcs/orbsvcs/ESF/ESF_Proxy_List.h + * orbsvcs/orbsvcs/ESF/ESF_Proxy_List.i + * orbsvcs/orbsvcs/ESF/ESF_Proxy_RB_Tree.cpp + * orbsvcs/orbsvcs/ESF/ESF_Proxy_RB_Tree.inl + * orbsvcs/orbsvcs/ESF/ESF_Copy_On_Write.h + * orbsvcs/orbsvcs/ESF/ESF_Copy_On_Write.i + * orbsvcs/orbsvcs/ESF/ESF_Immediate_Changes.cpp + * orbsvcs/orbsvcs/ESF/ESF_Immediate_Changes.inl + * orbsvcs/orbsvcs/ESF/ESF_Proxy_Admin.h + * orbsvcs/orbsvcs/ESF/ESF_Proxy_Admin.i + * orbsvcs/orbsvcs/ESF/ESF_Busy_Lock.h + * orbsvcs/orbsvcs/ESF/ESF_Busy_Lock.i + * orbsvcs/orbsvcs/Channel_Clients_T.inl + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPullConsumer.inl + * orbsvcs/orbsvcs/CosEvent/CEC_DynamicImplementation.cpp + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPushConsumer.h + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPushConsumer.i + * orbsvcs/orbsvcs/CosEvent/CEC_ConsumerAdmin.cpp + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPushSupplier.cpp + * orbsvcs/orbsvcs/CosEvent/CEC_TypedProxyPushConsumer.h + * orbsvcs/orbsvcs/CosEvent/CEC_TypedProxyPushConsumer.i + * orbsvcs/orbsvcs/CosEvent/CEC_TypedConsumerAdmin.cpp + * orbsvcs/orbsvcs/CosEvent/CEC_Default_Factory.cpp + * orbsvcs/orbsvcs/CosEvent/CEC_SupplierAdmin.h + * orbsvcs/orbsvcs/CosEvent/CEC_SupplierAdmin.i + * orbsvcs/orbsvcs/CosEvent/CEC_TypedEventChannel.h + * orbsvcs/orbsvcs/CosEvent/CEC_TypedEventChannel.i + * orbsvcs/orbsvcs/CosEvent/CEC_Reactive_SupplierControl.cpp + * orbsvcs/orbsvcs/CosEvent/CEC_TypedSupplierAdmin.h + * orbsvcs/orbsvcs/CosEvent/CEC_TypedEvent.inl + * orbsvcs/orbsvcs/CosEvent/CEC_TypedSupplierAdmin.i + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPushConsumer.cpp + * orbsvcs/orbsvcs/CosEvent/CEC_DynamicImplementation.h + * orbsvcs/orbsvcs/CosEvent/CEC_DynamicImplementation.i + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPullSupplier.cpp + * orbsvcs/orbsvcs/CosEvent/CEC_TypedProxyPushConsumer.cpp + * orbsvcs/orbsvcs/CosEvent/CEC_ConsumerAdmin.h + * orbsvcs/orbsvcs/CosEvent/CEC_EventChannel.cpp + * orbsvcs/orbsvcs/CosEvent/CEC_ConsumerAdmin.i + * orbsvcs/orbsvcs/CosEvent/CEC_SupplierAdmin.inl + * orbsvcs/orbsvcs/CosEvent/CEC_Reactive_ConsumerControl.cpp + * orbsvcs/orbsvcs/CosEvent/CEC_Reactive_Pulling_Strategy.cpp + * orbsvcs/orbsvcs/CosEvent/CEC_TypedConsumerAdmin.h + * orbsvcs/orbsvcs/CosEvent/CEC_Dispatching_Task.inl + * orbsvcs/orbsvcs/CosEvent/CEC_TypedEventChannel.inl + * orbsvcs/orbsvcs/CosEvent/CEC_TypedConsumerAdmin.i + * orbsvcs/orbsvcs/CosEvent/CEC_TypedSupplierAdmin.inl + * orbsvcs/orbsvcs/CosEvent/CEC_DynamicImplementation.inl + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPullConsumer.cpp + * orbsvcs/orbsvcs/CosEvent/CEC_Reactive_SupplierControl.h + * orbsvcs/orbsvcs/CosEvent/CEC_Reactive_SupplierControl.i + * orbsvcs/orbsvcs/CosEvent/CEC_TypedEvent.h + * orbsvcs/orbsvcs/CosEvent/CEC_TypedEvent.i + * orbsvcs/orbsvcs/CosEvent/CEC_ConsumerAdmin.inl + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPushSupplier.inl + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPullSupplier.h + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPullSupplier.i + * orbsvcs/orbsvcs/CosEvent/CEC_TypedConsumerAdmin.inl + * orbsvcs/orbsvcs/CosEvent/CEC_Default_Factory.inl + * orbsvcs/orbsvcs/CosEvent/CEC_EventChannel.h + * orbsvcs/orbsvcs/CosEvent/CEC_EventChannel.i + * orbsvcs/orbsvcs/CosEvent/CEC_Reactive_ConsumerControl.h + * orbsvcs/orbsvcs/CosEvent/CEC_Reactive_Pulling_Strategy.h + * orbsvcs/orbsvcs/CosEvent/CEC_Reactive_ConsumerControl.i + * orbsvcs/orbsvcs/CosEvent/CEC_Dispatching_Task.h + * orbsvcs/orbsvcs/CosEvent/CEC_Reactive_Pulling_Strategy.i + * orbsvcs/orbsvcs/CosEvent/CEC_Dispatching_Task.i + * orbsvcs/orbsvcs/CosEvent/CEC_Reactive_SupplierControl.inl + * orbsvcs/orbsvcs/CosEvent/CEC_TypedEvent.cpp + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPushConsumer.inl + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPullConsumer.h + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPullConsumer.i + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPullSupplier.inl + * orbsvcs/orbsvcs/CosEvent/CEC_TypedProxyPushConsumer.inl + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPushSupplier.h + * orbsvcs/orbsvcs/CosEvent/CEC_EventChannel.inl + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPushSupplier.i + * orbsvcs/orbsvcs/CosEvent/CEC_SupplierAdmin.cpp + * orbsvcs/orbsvcs/CosEvent/CEC_Reactive_ConsumerControl.inl + * orbsvcs/orbsvcs/CosEvent/CEC_Reactive_Pulling_Strategy.inl + * orbsvcs/orbsvcs/CosEvent/CEC_Dispatching_Task.cpp + * orbsvcs/orbsvcs/CosEvent/CEC_TypedEventChannel.cpp + * orbsvcs/orbsvcs/CosEvent/CEC_Default_Factory.h + * orbsvcs/orbsvcs/CosEvent/CEC_Default_Factory.i + * orbsvcs/orbsvcs/CosEvent/CEC_TypedSupplierAdmin.cpp + * orbsvcs/orbsvcs/Scheduler_Utilities.cpp + * orbsvcs/orbsvcs/AV/Policy.cpp + * orbsvcs/orbsvcs/AV/FlowSpec_Entry.cpp + * orbsvcs/orbsvcs/AV/MCast.cpp + * orbsvcs/orbsvcs/AV/QoS_UDP.inl + * orbsvcs/orbsvcs/AV/AVStreams_i.inl + * orbsvcs/orbsvcs/AV/UDP.cpp + * orbsvcs/orbsvcs/AV/Transport.h + * orbsvcs/orbsvcs/AV/Transport.i + * orbsvcs/orbsvcs/AV/AVStreams_i.h + * orbsvcs/orbsvcs/AV/AVStreams_i.i + * orbsvcs/orbsvcs/AV/Transport.inl + * orbsvcs/orbsvcs/AV/UDP.h + * orbsvcs/orbsvcs/AV/UDP.i + * orbsvcs/orbsvcs/AV/Policy.inl + * orbsvcs/orbsvcs/AV/FlowSpec_Entry.inl + * orbsvcs/orbsvcs/AV/MCast.inl + * orbsvcs/orbsvcs/AV/QoS_UDP.cpp + * orbsvcs/orbsvcs/AV/AVStreams_i.cpp + * orbsvcs/orbsvcs/AV/UDP.inl + * orbsvcs/orbsvcs/AV/Policy.h + * orbsvcs/orbsvcs/AV/Policy.i + * orbsvcs/orbsvcs/AV/FlowSpec_Entry.h + * orbsvcs/orbsvcs/AV/FlowSpec_Entry.i + * orbsvcs/orbsvcs/AV/MCast.h + * orbsvcs/orbsvcs/AV/MCast.i + * orbsvcs/orbsvcs/AV/QoS_UDP.h + * orbsvcs/orbsvcs/AV/QoS_UDP.i + * orbsvcs/orbsvcs/AV/Transport.cpp + * orbsvcs/orbsvcs/Time_Utilities.cpp + * orbsvcs/orbsvcs/Event/EC_Busy_Lock.inl + * orbsvcs/orbsvcs/Event/EC_Filter_Builder.cpp + * orbsvcs/orbsvcs/Event/EC_Timeout_Generator.cpp + * orbsvcs/orbsvcs/Event/ECG_Mcast_EH.h + * orbsvcs/orbsvcs/Event/ECG_Simple_Address_Server.h + * orbsvcs/orbsvcs/Event/ECG_Mcast_EH.i + * orbsvcs/orbsvcs/Event/ECG_Simple_Address_Server.i + * orbsvcs/orbsvcs/Event/EC_Supplier_Filter.cpp + * orbsvcs/orbsvcs/Event/EC_Reactive_ConsumerControl.inl + * orbsvcs/orbsvcs/Event/EC_Per_Supplier_Filter.inl + * orbsvcs/orbsvcs/Event/EC_Lifetime_Utils.cpp + * orbsvcs/orbsvcs/Event/EC_Dispatching_Task.cpp + * orbsvcs/orbsvcs/Event/EC_TPC_ProxySupplier.h + * orbsvcs/orbsvcs/Event/EC_TPC_ProxySupplier.i + * orbsvcs/orbsvcs/Event/EC_Sched_Factory.h + * orbsvcs/orbsvcs/Event/EC_Sched_Factory.i + * orbsvcs/orbsvcs/Event/EC_Filter.h + * orbsvcs/orbsvcs/Event/EC_Filter.i + * orbsvcs/orbsvcs/Event/EC_Filter_Builder.h + * orbsvcs/orbsvcs/Event/EC_Filter_Builder.i + * orbsvcs/orbsvcs/Event/EC_Gateway_IIOP_Factory.cpp + * orbsvcs/orbsvcs/Event/EC_ProxySupplier.h + * orbsvcs/orbsvcs/Event/EC_ProxySupplier.i + * orbsvcs/orbsvcs/Event/EC_Reactive_ConsumerControl.h + * orbsvcs/orbsvcs/Event/EC_Per_Supplier_Filter.h + * orbsvcs/orbsvcs/Event/EC_Reactive_ConsumerControl.i + * orbsvcs/orbsvcs/Event/EC_Per_Supplier_Filter.i + * orbsvcs/orbsvcs/Event/EC_Null_Scheduling.h + * orbsvcs/orbsvcs/Event/EC_Null_Scheduling.i + * orbsvcs/orbsvcs/Event/EC_ProxyConsumer.h + * orbsvcs/orbsvcs/Event/EC_ProxyConsumer.i + * orbsvcs/orbsvcs/Event/EC_Lifetime_Utils_T.h + * orbsvcs/orbsvcs/Event/EC_Lifetime_Utils_T.i + * orbsvcs/orbsvcs/Event/ECG_Complex_Address_Server.cpp + * orbsvcs/orbsvcs/Event/EC_Default_Factory.inl + * orbsvcs/orbsvcs/Event/ECG_CDR_Message_Receiver.h + * orbsvcs/orbsvcs/Event/ECG_CDR_Message_Receiver.i + * orbsvcs/orbsvcs/Event/EC_Prefix_Filter_Builder.h + * orbsvcs/orbsvcs/Event/EC_Prefix_Filter_Builder.i + * orbsvcs/orbsvcs/Event/EC_Null_Scheduling.inl + * orbsvcs/orbsvcs/Event/EC_Kokyu_Dispatching.h + * orbsvcs/orbsvcs/Event/EC_Kokyu_Dispatching.i + * orbsvcs/orbsvcs/Event/EC_Kokyu_Factory.cpp + * orbsvcs/orbsvcs/Event/EC_Sched_Filter_Builder.inl + * orbsvcs/orbsvcs/Event/ECG_CDR_Message_Receiver.inl + * orbsvcs/orbsvcs/Event/EC_Kokyu_Scheduling.cpp + * orbsvcs/orbsvcs/Event/EC_ObserverStrategy.cpp + * orbsvcs/orbsvcs/Event/EC_Sched_Factory.cpp + * orbsvcs/orbsvcs/Event/EC_Kokyu_Dispatching.inl + * orbsvcs/orbsvcs/Event/EC_Filter.cpp + * orbsvcs/orbsvcs/Event/EC_Timeout_Filter.cpp + * orbsvcs/orbsvcs/Event/EC_Basic_Filter_Builder.inl + * orbsvcs/orbsvcs/Event/ECG_Mcast_Gateway.cpp + * orbsvcs/orbsvcs/Event/EC_Event_Channel_Base.inl + * orbsvcs/orbsvcs/Event/EC_ProxySupplier.cpp + * orbsvcs/orbsvcs/Event/ECG_UDP_Receiver.inl + * orbsvcs/orbsvcs/Event/EC_ProxyConsumer.cpp + * orbsvcs/orbsvcs/Event/EC_Timeout_Filter.h + * orbsvcs/orbsvcs/Event/EC_Timeout_Filter.i + * orbsvcs/orbsvcs/Event/EC_Prefix_Filter_Builder.inl + * orbsvcs/orbsvcs/Event/ECG_UDP_EH.inl + * orbsvcs/orbsvcs/Event/EC_QOS_Info.inl + * orbsvcs/orbsvcs/Event/ECG_UDP_Out_Endpoint.inl + * orbsvcs/orbsvcs/Event/EC_Busy_Lock.cpp + * orbsvcs/orbsvcs/Event/EC_Kokyu_Filter_Builder.inl + * orbsvcs/orbsvcs/Event/EC_Reactive_ConsumerControl.cpp + * orbsvcs/orbsvcs/Event/EC_Per_Supplier_Filter.cpp + * orbsvcs/orbsvcs/Event/EC_ObserverStrategy.h + * orbsvcs/orbsvcs/Event/EC_ObserverStrategy.i + * orbsvcs/orbsvcs/Event/EC_Priority_Scheduling.inl + * orbsvcs/orbsvcs/Event/ECG_Mcast_EH.inl + * orbsvcs/orbsvcs/Event/ECG_UDP_Out_Endpoint.h + * orbsvcs/orbsvcs/Event/ECG_UDP_Out_Endpoint.i + * orbsvcs/orbsvcs/Event/EC_Lifetime_Utils_T.inl + * orbsvcs/orbsvcs/Event/EC_Supplier_Filter.h + * orbsvcs/orbsvcs/Event/ECG_Complex_Address_Server.h + * orbsvcs/orbsvcs/Event/EC_Supplier_Filter.i + * orbsvcs/orbsvcs/Event/ECG_Complex_Address_Server.i + * orbsvcs/orbsvcs/Event/ECG_CDR_Message_Sender.inl + * orbsvcs/orbsvcs/Event/EC_Reactive_SupplierControl.inl + * orbsvcs/orbsvcs/Event/EC_Priority_Scheduling.h + * orbsvcs/orbsvcs/Event/EC_Priority_Scheduling.i + * orbsvcs/orbsvcs/Event/EC_Kokyu_Factory.h + * orbsvcs/orbsvcs/Event/EC_Kokyu_Factory.i + * orbsvcs/orbsvcs/Event/EC_Default_Factory.cpp + * orbsvcs/orbsvcs/Event/EC_Gateway_IIOP_Factory.h + * orbsvcs/orbsvcs/Event/EC_Gateway_IIOP_Factory.i + * orbsvcs/orbsvcs/Event/EC_Null_Scheduling.cpp + * orbsvcs/orbsvcs/Event/EC_QOS_Info.h + * orbsvcs/orbsvcs/Event/EC_QOS_Info.i + * orbsvcs/orbsvcs/Event/EC_Reactive_SupplierControl.h + * orbsvcs/orbsvcs/Event/ECG_Simple_Address_Server.inl + * orbsvcs/orbsvcs/Event/EC_Reactive_SupplierControl.i + * orbsvcs/orbsvcs/Event/EC_Busy_Lock.h + * orbsvcs/orbsvcs/Event/EC_Busy_Lock.i + * orbsvcs/orbsvcs/Event/EC_Basic_Filter_Builder.h + * orbsvcs/orbsvcs/Event/EC_Sched_Filter_Builder.cpp + * orbsvcs/orbsvcs/Event/EC_Basic_Filter_Builder.i + * orbsvcs/orbsvcs/Event/ECG_CDR_Message_Receiver.cpp + * orbsvcs/orbsvcs/Event/EC_TPC_ProxySupplier.inl + * orbsvcs/orbsvcs/Event/EC_Kokyu_Dispatching.cpp + * orbsvcs/orbsvcs/Event/EC_Default_Factory.h + * orbsvcs/orbsvcs/Event/EC_Default_Factory.i + * orbsvcs/orbsvcs/Event/EC_Kokyu_Filter_Builder.h + * orbsvcs/orbsvcs/Event/EC_Kokyu_Filter_Builder.i + * orbsvcs/orbsvcs/Event/EC_Dispatching_Task.h + * orbsvcs/orbsvcs/Event/EC_Dispatching_Task.i + * orbsvcs/orbsvcs/Event/EC_Filter_Builder.inl + * orbsvcs/orbsvcs/Event/EC_Basic_Filter_Builder.cpp + * orbsvcs/orbsvcs/Event/EC_Timeout_Generator.inl + * orbsvcs/orbsvcs/Event/EC_Event_Channel_Base.cpp + * orbsvcs/orbsvcs/Event/EC_Supplier_Filter.inl + * orbsvcs/orbsvcs/Event/EC_Lifetime_Utils.inl + * orbsvcs/orbsvcs/Event/EC_Dispatching_Task.inl + * orbsvcs/orbsvcs/Event/ECG_UDP_Receiver.cpp + * orbsvcs/orbsvcs/Event/EC_Prefix_Filter_Builder.cpp + * orbsvcs/orbsvcs/Event/EC_Event_Channel_Base.h + * orbsvcs/orbsvcs/Event/EC_Gateway_IIOP_Factory.inl + * orbsvcs/orbsvcs/Event/EC_Event_Channel_Base.i + * orbsvcs/orbsvcs/Event/ECG_UDP_EH.cpp + * orbsvcs/orbsvcs/Event/EC_QOS_Info.cpp + * orbsvcs/orbsvcs/Event/EC_Lifetime_Utils.h + * orbsvcs/orbsvcs/Event/EC_Lifetime_Utils.i + * orbsvcs/orbsvcs/Event/ECG_UDP_Out_Endpoint.cpp + * orbsvcs/orbsvcs/Event/ECG_UDP_Receiver.h + * orbsvcs/orbsvcs/Event/ECG_UDP_Receiver.i + * orbsvcs/orbsvcs/Event/ECG_Complex_Address_Server.inl + * orbsvcs/orbsvcs/Event/EC_Kokyu_Scheduling.h + * orbsvcs/orbsvcs/Event/EC_Kokyu_Scheduling.i + * orbsvcs/orbsvcs/Event/EC_Kokyu_Filter_Builder.cpp + * orbsvcs/orbsvcs/Event/ECG_UDP_EH.h + * orbsvcs/orbsvcs/Event/ECG_UDP_EH.i + * orbsvcs/orbsvcs/Event/EC_Priority_Scheduling.cpp + * orbsvcs/orbsvcs/Event/EC_Kokyu_Factory.inl + * orbsvcs/orbsvcs/Event/ECG_Mcast_EH.cpp + * orbsvcs/orbsvcs/Event/EC_Lifetime_Utils_T.cpp + * orbsvcs/orbsvcs/Event/ECG_Mcast_Gateway.h + * orbsvcs/orbsvcs/Event/ECG_Mcast_Gateway.i + * orbsvcs/orbsvcs/Event/EC_Kokyu_Scheduling.inl + * orbsvcs/orbsvcs/Event/EC_ObserverStrategy.inl + * orbsvcs/orbsvcs/Event/EC_Sched_Factory.inl + * orbsvcs/orbsvcs/Event/ECG_CDR_Message_Sender.cpp + * orbsvcs/orbsvcs/Event/EC_Filter.inl + * orbsvcs/orbsvcs/Event/EC_Reactive_SupplierControl.cpp + * orbsvcs/orbsvcs/Event/EC_Timeout_Filter.inl + * orbsvcs/orbsvcs/Event/ECG_Mcast_Gateway.inl + * orbsvcs/orbsvcs/Event/EC_ProxySupplier.inl + * orbsvcs/orbsvcs/Event/EC_Sched_Filter_Builder.h + * orbsvcs/orbsvcs/Event/EC_Sched_Filter_Builder.i + * orbsvcs/orbsvcs/Event/ECG_CDR_Message_Sender.h + * orbsvcs/orbsvcs/Event/ECG_CDR_Message_Sender.i + * orbsvcs/orbsvcs/Event/ECG_Simple_Address_Server.cpp + * orbsvcs/orbsvcs/Event/EC_ProxyConsumer.inl + * orbsvcs/orbsvcs/Event/EC_Timeout_Generator.h + * orbsvcs/orbsvcs/Event/EC_Timeout_Generator.i + * orbsvcs/orbsvcs/Event/EC_TPC_ProxySupplier.cpp + * orbsvcs/orbsvcs/Scheduler_Factory.cpp + * orbsvcs/orbsvcs/Event_Utilities.cpp + * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Endpoint.h + * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Endpoint.i + * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Connection_Handler.cpp + * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Endpoint.cpp + * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Acceptor.inl + * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Connection_Handler.inl + * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Acceptor.h + * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Acceptor.i + * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Endpoint.inl + * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Acceptor.cpp + * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Connection_Handler.h + * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Connection_Handler.i + * orbsvcs/orbsvcs/Scheduler_Utilities.inl + * orbsvcs/orbsvcs/Channel_Clients_T.h + * orbsvcs/orbsvcs/Channel_Clients_T.i + * orbsvcs/orbsvcs/Time_Utilities.inl + * orbsvcs/tests/Event/lib/Driver.h + * orbsvcs/tests/Event/lib/Driver.i + * orbsvcs/tests/Event/lib/Driver.inl + * orbsvcs/tests/Event/lib/Driver.cpp + * orbsvcs/tests/Security/Crash_Test/test_i.inl + * orbsvcs/tests/Security/Crash_Test/test_i.cpp + * orbsvcs/tests/Security/Crash_Test/test_i.h + * orbsvcs/tests/Security/Crash_Test/test_i.i + * orbsvcs/tests/Security/MT_SSLIOP/test_i.cpp + * orbsvcs/tests/Security/MT_SSLIOP/test_i.inl + * orbsvcs/tests/Security/MT_SSLIOP/test_i.h + * orbsvcs/tests/Security/MT_SSLIOP/test_i.i + * orbsvcs/tests/EC_Mcast/EC_Mcast.cpp + * orbsvcs/tests/EC_Mcast/EC_Mcast.inl + * orbsvcs/tests/EC_Mcast/EC_Mcast.h + * orbsvcs/tests/EC_Mcast/EC_Mcast.i + * orbsvcs/tests/HTIOP/BiDirectional/test_i.cpp + * orbsvcs/tests/HTIOP/BiDirectional/test_i.inl + * orbsvcs/tests/HTIOP/BiDirectional/test_i.h + * orbsvcs/tests/HTIOP/BiDirectional/test_i.i + * orbsvcs/examples/Security/Send_File/test_i.cpp + * orbsvcs/examples/Security/Send_File/test_i.inl + * orbsvcs/examples/Security/Send_File/test_i.h + * orbsvcs/examples/Security/Send_File/test_i.i + * tao/Tagged_Components.cpp + * tao/IIOP_Endpoint.h + * tao/operation_details.cpp + * tao/IIOP_Endpoint.i + * tao/CDR.cpp + * tao/Tagged_Profile.inl + * tao/Endpoint.cpp + * tao/ORB_Core.cpp + * tao/target_specification.h + * tao/target_specification.i + * tao/GIOP_Message_Locate_Header.h + * tao/Object.inl + * tao/GIOP_Message_Locate_Header.i + * tao/Fault_Tolerance_Service.h + * tao/Fault_Tolerance_Service.i + * tao/LocalObject.inl + * tao/Asynch_Reply_Dispatcher_Base.cpp + * tao/Policy_Set.cpp + * tao/PortableServer/Servant_Base.cpp + * tao/PortableServer/POA_Cached_Policies.inl + * tao/PortableServer/POA_Cached_Policies.h + * tao/PortableServer/POA_Cached_Policies.i + * tao/PortableServer/POA_Policy_Set.h + * tao/PortableServer/POA_Policy_Set.i + * tao/PortableServer/POAManager.cpp + * tao/PortableServer/POAManager.h + * tao/PortableServer/POAManager.i + * tao/PortableServer/Active_Object_Map.h + * tao/PortableServer/Active_Object_Map.i + * tao/PortableServer/Object_Adapter.h + * tao/PortableServer/Object_Adapter.i + * tao/PortableServer/POA_Cached_Policies.cpp + * tao/PortableServer/Active_Object_Map.inl + * tao/PortableServer/Object_Adapter.inl + * tao/PortableServer/POA_Policy_Set.inl + * tao/PortableServer/Servant_Base.inl + * tao/PortableServer/Active_Object_Map.cpp + * tao/PortableServer/Object_Adapter.cpp + * tao/PortableServer/Servant_Base.h + * tao/PortableServer/Servant_Base.i + * tao/PortableServer/POAManager.inl + * tao/PortableServer/POA_Policy_Set.cpp + * tao/CORBALOC_Parser.cpp + * tao/IIOP_Endpoint.inl + * tao/Tagged_Components.h + * tao/Tagged_Components.i + * tao/Tagged_Profile.h + * tao/IIOP_Acceptor.cpp + * tao/Tagged_Profile.i + * tao/MCAST_Parser.h + * tao/Parser_Registry.h + * tao/MCAST_Parser.i + * tao/Parser_Registry.i + * tao/Acceptor_Registry.h + * tao/Acceptor_Registry.i + * tao/Buffering_Constraint_Policy.h + * tao/Buffering_Constraint_Policy.i + * tao/Exception.h + * tao/Exception.i + * tao/Bind_Dispatcher_Guard.inl + * tao/Stub.cpp + * tao/MCAST_Parser.cpp + * tao/Endpoint.h + * tao/Endpoint.i + * tao/Exception.inl + * tao/MProfile.inl + * tao/params.h + * tao/params.i + * tao/Reply_Dispatcher.cpp + * tao/MProfile.h + * tao/MProfile.i + * tao/params.cpp + * tao/Parser_Registry.cpp + * tao/target_specification.cpp + * tao/Policy_Manager.inl + * tao/GIOP_Message_Locate_Header.inl + * tao/Asynch_Reply_Dispatcher_Base.h + * tao/Asynch_Reply_Dispatcher_Base.i + * tao/Tagged_Components.inl + * tao/CORBALOC_Parser.h + * tao/operation_details.inl + * tao/CORBALOC_Parser.i + * tao/CDR.inl + * tao/IIOP_Acceptor.h + * tao/IIOP_Acceptor.i + * tao/Reply_Dispatcher.h + * tao/Reply_Dispatcher.i + * tao/Leader_Follower.h + * tao/Leader_Follower.i + * tao/Endpoint.inl + * tao/Fault_Tolerance_Service.cpp + * tao/ORB_Core.inl + * tao/Profile.cpp + * tao/Asynch_Reply_Dispatcher_Base.inl + * tao/Environment.h + * tao/Policy_Set.inl + * tao/Environment.i + * tao/ORB.cpp + * tao/CORBALOC_Parser.inl + * tao/IIOP_Acceptor.inl + * tao/RTPortableServer/RT_POA.cpp + * tao/RTPortableServer/RT_POA.inl + * tao/RTPortableServer/RT_POA.h + * tao/RTPortableServer/RT_POA.i + * tao/ORB.h + * tao/ORB.i + * tao/Leader_Follower.cpp + * tao/Stub.inl + * tao/MCAST_Parser.inl + * tao/Acceptor_Registry.cpp + * tao/Environment.cpp + * tao/Profile.h + * tao/Profile.i + * tao/Object.h + * tao/Object.i + * tao/Stub.h + * tao/Stub.i + * tao/Messaging/Messaging_Policy_i.cpp + * tao/Messaging/Messaging_Policy_i.inl + * tao/Messaging/Messaging_Policy_i.h + * tao/Messaging/Messaging_Policy_i.i + * tao/Policy_Manager.h + * tao/Policy_Manager.i + * tao/Reply_Dispatcher.inl + * tao/Pluggable_Messaging_Utils.cpp + * tao/Principal.cpp + * tao/TAO_Server_Request.cpp + * tao/Connector_Registry.cpp + * tao/params.inl + * tao/Parser_Registry.inl + * tao/target_specification.inl + * tao/ORB_Core.h + * tao/ORB_Core.i + * tao/Connector_Registry.h + * tao/Connector_Registry.i + * tao/Strategies/SHMIOP_Endpoint.inl + * tao/Strategies/SCIOP_Endpoint.cpp + * tao/Strategies/DIOP_Acceptor.h + * tao/Strategies/UIOP_Endpoint.h + * tao/Strategies/DIOP_Acceptor.i + * tao/Strategies/UIOP_Endpoint.i + * tao/Strategies/DIOP_Endpoint.inl + * tao/Strategies/SCIOP_Acceptor.inl + * tao/Strategies/SHMIOP_Endpoint.h + * tao/Strategies/SHMIOP_Endpoint.i + * tao/Strategies/DIOP_Acceptor.cpp + * tao/Strategies/UIOP_Endpoint.inl + * tao/Strategies/SHMIOP_Endpoint.cpp + * tao/Strategies/DIOP_Endpoint.h + * tao/Strategies/DIOP_Endpoint.i + * tao/Strategies/SCIOP_Acceptor.h + * tao/Strategies/SCIOP_Acceptor.i + * tao/Strategies/SCIOP_Endpoint.inl + * tao/Strategies/DIOP_Endpoint.cpp + * tao/Strategies/SCIOP_Acceptor.cpp + * tao/Strategies/DIOP_Acceptor.inl + * tao/Strategies/UIOP_Endpoint.cpp + * tao/Strategies/SCIOP_Endpoint.h + * tao/Strategies/SCIOP_Endpoint.i + * tao/Buffering_Constraint_Policy.cpp + * tao/CDR.h + * tao/CDR.i + * tao/Policy_Set.h + * tao/Policy_Set.i + * tao/TAO_Server_Request.h + * tao/Principal.h + * tao/TAO_Server_Request.i + * tao/Principal.i + * tao/Fault_Tolerance_Service.inl + * tao/Profile.inl + * tao/Tagged_Profile.cpp + * tao/Object.cpp + * tao/RTScheduling/RTScheduler_Manager.inl + * tao/RTScheduling/RTScheduler_Manager.cpp + * tao/RTScheduling/RTScheduler_Manager.h + * tao/RTScheduling/RTScheduler_Manager.i + * tao/ORB.inl + * tao/Bind_Dispatcher_Guard.h + * tao/Bind_Dispatcher_Guard.i + * tao/LocalObject.cpp + * tao/Pluggable_Messaging_Utils.h + * tao/Pluggable_Messaging_Utils.i + * tao/Leader_Follower.inl + * tao/Acceptor_Registry.inl + * tao/IIOP_Endpoint.cpp + * tao/Environment.inl + * tao/operation_details.h + * tao/operation_details.i + * tao/Bind_Dispatcher_Guard.cpp + * tao/Pluggable_Messaging_Utils.inl + * tao/Principal.inl + * tao/TAO_Server_Request.inl + * tao/Connector_Registry.inl + * tao/LocalObject.h + * tao/LocalObject.i + * tao/Exception.cpp + * tao/MProfile.cpp + * tao/Policy_Manager.cpp + * tao/Buffering_Constraint_Policy.inl + * tao/GIOP_Message_Locate_Header.cpp + * tao/RTCORBA/Network_Priority_Mapping_Manager.inl + * tao/RTCORBA/Priority_Mapping_Manager.h + * tao/RTCORBA/Priority_Mapping_Manager.i + * tao/RTCORBA/Network_Priority_Mapping_Manager.h + * tao/RTCORBA/Network_Priority_Mapping_Manager.i + * tao/RTCORBA/Priority_Mapping_Manager.inl + * tao/RTCORBA/Network_Priority_Mapping_Manager.cpp + * tao/RTCORBA/Priority_Mapping_Manager.cpp + * performance-tests/POA/Demux/demux_test_client.cpp + * performance-tests/POA/Demux/demux_test_client.inl + * performance-tests/POA/Demux/demux_test_client.i + * performance-tests/RTCorba/Oneways/Reliable/test_i.cpp + * performance-tests/RTCorba/Oneways/Reliable/test_i.inl + * performance-tests/RTCorba/Oneways/Reliable/test_i.h + * performance-tests/RTCorba/Oneways/Reliable/test_i.i + * performance-tests/RTCorba/Multiple_Endpoints/Common/test_i.cpp + * performance-tests/RTCorba/Multiple_Endpoints/Common/test_i.inl + * performance-tests/RTCorba/Multiple_Endpoints/Common/test_i.h + * performance-tests/RTCorba/Multiple_Endpoints/Common/test_i.i + * tests/Faults/ping_i.cpp + * tests/Faults/ping_i.inl + * tests/Faults/ping_i.h + * tests/Faults/ping_i.i + * tests/Faults/test_i.inl + * tests/Faults/test_i.cpp + * tests/Faults/test_i.h + * tests/Faults/test_i.i + * tests/Muxed_GIOP_Versions/test_i.cpp + * tests/Muxed_GIOP_Versions/test_i.inl + * tests/Muxed_GIOP_Versions/test_i.h + * tests/Muxed_GIOP_Versions/test_i.i + * tests/DSI_Gateway/test_dsi.cpp + * tests/DSI_Gateway/test_dsi.inl + * tests/DSI_Gateway/test_dsi.h + * tests/DSI_Gateway/test_dsi.i + * tests/DSI_Gateway/test_i.inl + * tests/DSI_Gateway/test_i.cpp + * tests/DSI_Gateway/test_i.h + * tests/DSI_Gateway/test_i.i + * tests/MT_Client/test_i.inl + * tests/MT_Client/test_i.cpp + * tests/MT_Client/test_i.h + * tests/MT_Client/test_i.i + * tests/OctetSeq/test_i.inl + * tests/OctetSeq/test_i.cpp + * tests/OctetSeq/test_i.h + * tests/OctetSeq/test_i.i + * tests/BiDirectional_DelayedUpcall/test_i.cpp + * tests/BiDirectional_DelayedUpcall/test_i.inl + * tests/BiDirectional_DelayedUpcall/test_i.h + * tests/BiDirectional_DelayedUpcall/test_i.i + * tests/MT_Server/test_i.cpp + * tests/MT_Server/test_i.inl + * tests/MT_Server/test_i.h + * tests/MT_Server/test_i.i + * tests/BiDirectional/test_i.cpp + * tests/BiDirectional/test_i.inl + * tests/BiDirectional/test_i.h + * tests/BiDirectional/test_i.i + * tests/Strategies/simple_test_i.cpp + * tests/Strategies/simple_test_i.inl + * tests/Strategies/simple_test_i.h + * tests/Strategies/simple_test_i.i + * tests/MProfile_Forwarding/test_i.cpp + * tests/MProfile_Forwarding/test_i.inl + * tests/MProfile_Forwarding/test_i.h + * tests/MProfile_Forwarding/test_i.i + * tests/Native_Exceptions/test_i.cpp + * tests/Native_Exceptions/test_i.inl + * tests/Native_Exceptions/test_i.h + * tests/Native_Exceptions/test_i.i + * tests/Forwarding/test_i.cpp + * tests/Forwarding/test_i.inl + * tests/Forwarding/test_i.h + * tests/Forwarding/test_i.i + * tests/BiDirectional_NestedUpcall/test_i.cpp + * tests/BiDirectional_NestedUpcall/test_i.inl + * tests/BiDirectional_NestedUpcall/test_i.h + * tests/BiDirectional_NestedUpcall/test_i.i + * TAO_IDL/fe/fe_lookup.cpp + * examples/Kokyu_dsrt_schedulers/fp_example/test_i.cpp + * examples/Kokyu_dsrt_schedulers/fp_example/test_i.inl + * examples/Kokyu_dsrt_schedulers/fp_example/test_i.h + * examples/Kokyu_dsrt_schedulers/fp_example/test_i.i + * examples/Kokyu_dsrt_schedulers/mif_example/test_i.cpp + * examples/Kokyu_dsrt_schedulers/mif_example/test_i.inl + * examples/Kokyu_dsrt_schedulers/mif_example/test_i.h + * examples/Kokyu_dsrt_schedulers/mif_example/test_i.i + * examples/Kokyu_dsrt_schedulers/muf_example/test_i.cpp + * examples/Kokyu_dsrt_schedulers/muf_example/test_i.inl + * examples/Kokyu_dsrt_schedulers/muf_example/test_i.h + * examples/Kokyu_dsrt_schedulers/muf_example/test_i.i + + Moved .i to .inl, and updated the associated header and source files. + +Thu Jul 27 19:21:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * TAO-INSTALL.html: + Updated to use subversion and updated minimum corba documentation + +Thu Jul 27 18:59:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * docs/minimumTAO.html: + Fixed some incorrect hyperlinks + +Thu Jul 27 18:07:44 UTC 2006 William R. Otte <wotte@dre.vanderbilt.edu> + + * TAO_ACE.mwc + + Replaced relative reference to ACE with an environment variable. + This is necessary to allow TAO to build in flattened repo layout. + +Wed Jul 26 07:59:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * TAO-INSTALL.html: + * tao/Transport.h: + * docs/releasenotes/ftcorba_services.html: + * docs/Security/SSLIOP-Usage.html: + Updated documentation to point to the new svn repository + +Mon Jul 24 21:54:22 UTC 2006 William R. Otte <wotte@dre.vanderbilt.edu> + + * orbsvcs/orbsvcs/HTIOP/htiop_endpointsC.cpp + * orbsvcs/orbsvcs/HTIOP/htiop_endpointsC.h + * orbsvcs/orbsvcs/SSLIOP/ssl_endpointsC.h + * orbsvcs/orbsvcs/SSLIOP/ssl_endpointsC.cpp + * tao/RTScheduling/RTSchedulerC.cpp + * tao/RTScheduling/RTSchedulerC.h + + Removed these, they were kept around erroneously. + + * ./TAO/performance-tests/AMI_Latency + * ./TAO/performance-tests/Deferred_Latency + * ./TAO/performance-tests/DII_Latency + * ./TAO/performance-tests/DSI_Latency + * ./TAO/performance-tests/Single_Threaded_Latency + * ./TAO/performance-tests/Thread_Per_Connection_Latency + * ./TAO/performance-tests/Thread_Pool_Latency + + Removed these directories, which had been previously merged into + performance-tests/Latency + +Mon Jul 24 19:02:33 UTC 2006 William R. Otte <wotte@dre.vanderbilt.edu> + + * TAO/RTScheduling/RTSchedulerA.h + + Should not have been removed. + +Mon Jul 24 18:52:33 UTC 2006 William R. Otte <wotte@dre.vanderbilt.edu> + + * TAO/RTScheduling/RTSchedulerA.cpp + + Removed, these were kept around erroneously by the conversion script. + +Mon Jul 24 17:39:38 UTC 2006 William R. Otte <wotte@dre.vanderbilt.edu> + + * TAO_IDL/fe/fe_lookup.cpp + + Re committing this file, as the subversion script probably backdated + the timestamp on the file, which was causing it to be regenerated as + part of the build process. + +Mon Jul 24 15:45:34 UTC 2006 William R. Otte <wotte@dre.vanderbilt.edu> + + * Subversion conversion completed at revision 73730 + +Sun Jul 23 15:44:43 UTC 2006 William R. Otte <wotte@dre.vanderbilt.edu> + + * Repository frozen for repository conversion + +Sat Jul 22 11:16:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/QtTests/client.cpp: + Updated include because of QtResource move + +Fri Jul 21 14:17:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Reliable_Oneways/server.cpp: + Initialise pointer with 0 + +Fri Jul 21 14:13:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/RTScheduling/Current/Thread_Task.cpp: + * tests/RTScheduling/Thread_Cancel/Thread_Task.cpp: + Simplified printing the elements of a string sequence + +Fri Jul 21 14:09:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Portable_Interceptors/PICurrent/client.cpp: + * tests/Portable_Interceptors/PICurrent/test.idl: + * tests/Portable_Interceptors/PICurrent/test_i.{h,cpp}: + Extended this test + +Fri Jul 21 13:27:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Connection_Timeout/Connection_Timeout.mpc: + Simplified this file + +Fri Jul 21 13:24:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Bug_2503_Regression/client.cpp: + * tests/Bug_2503_Regression/server.cpp: + * tests/Bug_2503_Regression/Test.idl: + * tests/Bug_2503_Regression/test_i.{h,cpp}: + Extended this test in such a way that the client does a shutdown + invocation on the server when it is ready. + +Fri Jul 21 12:05:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Bug_1551_Regression/Reply_Handler.cpp: + At the moment there is a bug in the AMI code it could happen that + the exception holder pointer is nill, in that case print an error + message instead of trying to invoke a method on it. This shouldn't + normally happen, but makes it a lot easier to analyze the problem + because we now just get an error message instead of an access + violation. + +Fri Jul 21 11:48:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/tests/Notify/lib/Peer.cpp: + Removed include of Arg_Shifter.h, it is already done in the header + file + +Fri Jul 21 08:40:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * utils/catior/catior.cpp: + Added needed include to build catior when IIOP has been disabled + in the orb + + * tao/Object_T.cpp: + Const improvements, use true/false for bool + +Fri Jul 21 07:40:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Strategies/UIOP_Profile.cpp: + * tao/Strategies/DIOP_Profile.cpp: + * tao/Strategies/SCIOP_Profile.cpp: + Use true/false instead of 1/0 + +Fri Jul 21 07:31:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Load_Protocol_Factory_T.h: + Added missing return statement + +Thu Jul 20 20:38:08 UTC 2006 Shanshan Jiang <shanshan.jiang@vanderbilt.edu> + + * docs/tutorials/Quoter/RTCORBA/Broker.cpp + * docs/tutorials/Quoter/RTCORBA/Broker_i.cpp + * docs/tutorials/Quoter/RTCORBA/Broker_i.h + * docs/tutorials/Quoter/RTCORBA/Common_i.h + * docs/tutorials/Quoter/RTCORBA/Distributor_i.cpp + * docs/tutorials/Quoter/RTCORBA/Distributor_i.h + * docs/tutorials/Quoter/RTCORBA/Stock_Database.cpp + * docs/tutorials/Quoter/RTCORBA/Stock_Database.h + * docs/tutorials/Quoter/RTCORBA/Stock_PriorityMapping.cpp + * docs/tutorials/Quoter/RTCORBA/Stock_PriorityMapping.h + + Fixed some errors. Addressed the comments from Will. + +Thu Jul 20 13:16:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/IIOP_Profile.cpp (do_is_equivalent): + Use true/false + +Thu Jul 20 12:36:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Strategies/DIOP_Endpoint.{h,cpp}: + Made object_addr_set_ a bool instead of int + +Thu Jul 20 12:34:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Strategies/DIOP_Endpoint.h: + Removed commented out method + +Thu Jul 20 12:26:59 UTC 2006 Simon McQueen <sm@prismtech.com> + + * tests/Bug_2549_Regression/test.idl: + + Replace <> with "" to fix fuxx error. + +Thu Jul 20 11:38:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Load_Protocol_Factory_T.h: + New template method that can be used to load a specific pluggable + protocol. + + * tao/tao.mpc: + Added new file + + * tao/default_resource.cpp: + * tao/Strategies/advanced_resource.cpp: + Use the new template method to load the pluggable protocols. This + way we can remove a lot of duplicated code that was copied just for + each specific protocol. + +Thu Jul 20 10:24:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Strategies/DIOP_Transport.h: + Removed commented out methods + +Thu Jul 20 10:00:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/tao.mpc: + * tao/EndpointPolicy.mpc: + Add tao_no_iiop as base project, when we enable this feature these + projects are affected + + * tao/IIOP.pidl: + * tao/EndpointPolicy/IIOPEndpointValue.pidl: + Guard this with !TAO_LACKS_IIOP, when tao_no_iiop is enabled as MPC + feature we don't generate code for these files. + + * tao/Endpoint_Acceptor_Filter.cpp: + No need to include IIOP specific files, base classes are all we need + here. + + * tao/EndpointPolicy/IIOPEndpointValue_i.{h,cpp}: + Don't compile these files when IIOP is disabled. + +Thu Jul 20 08:17:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Bug_1639_Regression/struct_client.cpp: + * tests/DynAny_Test/test_dynany.cpp: + * tests/DynAny_Test/analyzer.h: + Reordered includes to fix problems with gcc4 + +Thu Jul 20 07:08:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/IDL_Test/IDL_Test.mpc: + Added regression for bug 2582 again + +Thu Jul 20 03:43:30 UTC 2006 William R. Otte <wotte@dre.vanderbilt.edu> + + * docs/tutorials/Quoter/RTCORBA/Broker_i.cpp + * docs/tutorials/Quoter/RTCORBA/Distributor_i.cpp + * docs/tutorials/Quoter/RTCORBA/Stock_Database.cpp + + Added some comments for Shanshan. + +Wed Jul 19 22:44:16 UTC 2006 Ciju John <johnc@ociweb.com> + + * orbsvcs/tests/Event/Mcast/RTEC_MCast_Federated/SimpleAddressServer.h: + + Add version macros. + +Wed Jul 19 22:28:50 UTC 2006 Ciju John <johnc@ociweb.com> + + * orbsvcs/orbsvcs/Event/ECG_Mcast_EH.cpp: + + The open() initializes the underlying socket that is later + enabled to be non-blocking. Without open() the enable call will + fail as it should. This fixes the scoreboard failure introduced + in: + + Mon Jul 17 14:26:08 UTC 2006 Ciju John <john_c@ociweb.com> + +Wed Jul 19 19:20:28 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * TAO_IDL/be/be_global.cpp: + * TAO_IDL/be_include/be_global.h: + * docs/compiler.html: + * docs/releasenotes/index.html: + + - Added IDL compiler command line option -oS, which sets an + output directory for all skeleton (*S*) files, similar to what + the -oA option does for *A.* files. Overrides the -o option + value, if any. Also updated IDL compiler and related docs. + This feature will make it easier to automate the generation + from IDL files for services like the Interface Repository, + where the stub and skeleton files go to different directories. + This fix closes [BUGID:2603]. + + - Changed all occurrences of 'base_name_only' parameters from + int to bool. + +Wed Jul 19 17:49:47 UTC 2006 Nanbor Wang <nanbor (a) txcorp (d) com> + + * orbsvcs/examples/FaultTolerance/RolyPoly/client.cpp: Enhanced + the client to use more than 2 replicas. The limitation of using + 2 replicas was causing the client to exit when one of the server + crashes. This is due to the way ACE_TMCase handling a single + server as a fault. But allowing more than 2 replicas, the 2 + remaining servers can continue to work and thus demonstrate how + FT work. + + * orbsvcs/examples/FaultTolerance/RolyPoly/RolyPoly_i.cpp: + * orbsvcs/examples/FaultTolerance/RolyPoly/RolyPoly_i.h: + Minor fixup to throw spec. + +Wed Jul 19 17:07:12 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * TAO_IDL/be/be_visitor_arg_traits.cpp: + + Fixed bug in the logic of the decisioin whether or not to generate + the arg traits template specialization for a typedef. This fix + addresses the reopening of [BUGID:2582], which will now be set + to FIXED unless and until another bug is found. + +Wed Jul 19 17:21:38 UTC 2006 Ossama Othman <ossama_othman at symantec dot com> + + * orbsvcs/tests/AVStreams/Asynch_Three_Stage/Connection_Manager.cpp + (load_ep_addr): + + Use new ACE_CString::size_type type to hold string + position/index value, instead of "int". Addresses "comparison + between signed and unsigned" warnings. + + * tao/Strategies/DIOP_Acceptor.cpp (parse_options): + * tao/Strategies/SCIOP_Acceptor.cpp (parse_options): + * tao/Strategies/SHMIOP_Acceptor.cpp (parse_options): + * tao/Strategies/UIOP_Acceptor.cpp (parse_options): + + Applied Ciju's substring fix. See "Tue Jul 18 21:38:55 UTC 2006 + Ciju John <johnc@ociweb.com>" below for details. + +Wed Jul 19 15:52:11 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * tests/IDL_Test/anonymous.idl: + * tests/IDL_Test/typedef.idl: + * tests/IDL_Test/valuetype.idl: + + Entries to test sequence of valuetype as valuetype member + and typedefs of string with the same bound repeated in + different compilation units that are linked in the same + build. + +Wed Jul 19 15:03:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbscvs/orbsvcs/IFRService/*S.h: + Replace ace\pre.h with ace/pre.h, seems generation with MPC is + dependent on windows or linux + +Wed Jul 19 13:18:41 UTC 2006 Chad Elliott <elliott_c@ociweb.com> + + * tao/Codeset/Codeset_Descriptor.cpp: + * tao/Codeset/UTF16_BOM_Translator.cpp: + + Added static casts to remove build warnings from some 64-bit + compilers where unsigned long and ACE_CDR::ULong were mixed. + It was necessary to use static_cast<> as the size of the data in + most cases needed to be size_t/unsigned long. + +Wed Jul 19 12:31:50 UTC 2006 Chad Elliott <elliott_c@ociweb.com> + + * tao/LocalObject.cpp: + * tao/ORB.h: + * tao/ORB.i: + * tao/ORB_Core.h: + * tao/ORB_Core.i: + * tao/Principal.h: + * tao/Principal.i: + * tao/Profile.h: + * tao/Profile.cpp: + + In my change from Mon Jul 17 14:40:43 UTC 2006, I modified these + files to use an ACE_Atomic_Op<> with an unsigned long. They all + had instances where unsigned long and CORBA::ULong were being + mixed which causes warnings with some 64-bit compilers. I have + modified these to all use unsigned long in places where + CORBA::ULong was previously used. + +Wed Jul 19 12:17:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbscvs/orbsvcs/IFRService/*.S*.*: + Regenerated these files + + * tao/IFR_Client.mpc: + Only use the stub export tao_idl options + +Wed Jul 19 10:23:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/RTScheduler.mpc: + Enabled regeneration of the pidl file + + * tao/RTScheduling/RTSchedulerA.cpp: + * tao/RTScheduling/RTSchedulerC.cpp: + * tao/RTScheduling/RTSchedulerC.h: + * tao/RTScheduling/RTSchedulerS.h: + Removed from the repo + +Wed Jul 19 10:03:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * TAO_IDL/be/be_visitor_module.cpp: + * TAO_IDL/be/be_visitor_module/module.cpp: + * TAO_IDL/be_include/be_visitor_module/module.h: + Support generating native within a module + +Wed Jul 19 08:45:00 UTC 2006 Simon Massey <sma@prismtech.com> + + * utils/nslist/nslist.cpp: + + Bug fix for over optimizing Borland compiler. + +Wed Jul 19 08:03:36 UTC 2006 Simon McQueen <sm@prismtech.com> + + * tao/CodecFactory/IOP_CodecA.h: + * tao/ObjRefTemplate/ObjectReferenceTemplateA.h: + * tao/PortableServer/PortableServerA.h: + * tao/RTCORBA/RTCORBAA.h: + * tao/RTCORBA/RTCORBAS.h: + * tao/RTPortableServer/RTPortableServerA.h: + * tao/RTPortableServer/RTPortableServer_includeA.h: + * tao/RTScheduling/RTSchedulerA.h: + * tao/RTScheduling/RTSchedulerS.h: + + Added these 'empty' files. Works around problems with the IDL + preprocessing. This fixes bug #2549. See ticket for more. + + * tests/Bug_2549_Regression/Bug_2549_Regression.mpc: + * tests/Bug_2549_Regression/README: + * tests/Bug_2549_Regression/test.cpp: + * tests/Bug_2549_Regression/test.idl: + + Compile only regression test for the above fix. + +Wed Jul 19 08:08:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * TAO_IDL/be/be_visitor_native/native_ch.cpp: + Add generation for CORBA::VoidData which is a native that is used + in the RTScheduling library. + +Tue Jul 18 21:38:55 UTC 2006 Ciju John <johnc@ociweb.com> + + * tao/IIOP_Acceptor.cpp: + + substring() 2nd parameter is supposed to be the relative + length, not the absolute position as before. + +Tue Jul 18 14:21:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/orbsvcs/HTIOP.mpc: + Simplified this file + + * orbsvcs/orbsvcs/SSLIOP.mpc: + Regenerate the pidl file + + * orbsvcs/orbsvcs/SSLIOP/ssl_endpointsC.{h,cpp}: + Removed from the repo + +Tue Jul 18 13:51:09 UTC 2006 Adam Mitz <mitza@ociweb.com> + + * orbsvcs/tests/CosEvent/Timeout/Timeout.mpc: + Added the "threads" base project to exclude this test from single- + threaded builds. + +Tue Jul 18 11:19:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/VarOut_T.inl: + Fixed bug 2595, when the _var is nil and the .out() is invoked + we have to allocate a new instance instead of deference a + nill pointer. This matches the latest CORBA C++ mapping, thanks + to Abdul Sowayan <abdullah dot sowayan at lmco dot com> for + reporting this bug + +Tue Jul 18 10:29:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Bug_2595_Regression/* + New regression for bug 2595, not fixed yet, makes it possible + to retest things on linux before we enable this test in + the regression suite + +Tue Jul 18 10:29:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Bug_1639_Regression/struct_client.cpp: + * tests/DynAny_Test/test_dynany.h: + Fixed compile errors with gcc4 + +Tue Jul 18 10:01:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/orbsvcs/Event/ECG_UDP_Out_Endpoint.cpp: + Use true/false + + * orbsvcs/orbsvcs/SSLIOP/ssl_endpoints.pidl: + Updated comments + +Tue Jul 18 09:55:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/IIOP_Acceptor.cpp: + Const improvement + +Tue Jul 18 09:37:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/orbsvcs/HTIOP.mpc: + Enabled regeneration of the pidl file + + * orbsvcs/orbsvcs/HTIOP/htiop_endpoints.pidl: + Updated comments + + * orbsvcs/orbsvcs/HTIOP/htiop_endpointsC.{h,cpp}: + Removed from the repo + +Tue Jul 18 09:29:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/QtTests/server.cpp: + * tests/FL_Cube/client.cpp: + Fixed include paths because resource factories got moved + +Tue Jul 18 08:15:00 UTC 2006 Simon Massey <sma@prismtech.com> + + * orbsvcs/orbsvcs/Naming/Storable.h: + * orbsvcs/orbsvcs/Naming/Storable_Naming_Context.cpp: + + Fix bugzilla 2600 - Naming Service Flat File Persistance with + Federated naming contexts. + +Mon Jul 17 18:35:32 UTC 2006 Shanshan Jiang <shanshan.jiang@vanderbilt.edu> + + * docs/tutorials/Quoter/RTCORBA/Admin.cpp + * docs/tutorials/Quoter/RTCORBA/Broker.cpp + * docs/tutorials/Quoter/RTCORBA/Broker_i.cpp + * docs/tutorials/Quoter/RTCORBA/Broker_i.h + * docs/tutorials/Quoter/RTCORBA/Common_i.cpp + * docs/tutorials/Quoter/RTCORBA/Common_i.h + * docs/tutorials/Quoter/RTCORBA/Distributor.cpp + * docs/tutorials/Quoter/RTCORBA/Distributor_i.cpp + * docs/tutorials/Quoter/RTCORBA/Distributor_i.h + * docs/tutorials/Quoter/RTCORBA/Stock_Database.cpp + * docs/tutorials/Quoter/RTCORBA/Stock_Database.h + + * docs/tutorials/Quoter/RTCORBA/docs/broker.html + * docs/tutorials/Quoter/RTCORBA/docs/common.html + * docs/tutorials/Quoter/RTCORBA/docs/database.html + * docs/tutorials/Quoter/RTCORBA/docs/distributor.html + * docs/tutorials/Quoter/RTCORBA/docs/idl.html + * docs/tutorials/Quoter/RTCORBA/docs/run.html + + Make some minor changes. + +Mon Jul 17 17:49:41 UTC 2006 Ossama Othman <ossama_othman at symantec dot com> + + * examples/POA/On_Demand_Loading/Servant_Manager.cpp: + * orbsvcs/orbsvcs/AV/AVStreams_i.cpp: + * orbsvcs/orbsvcs/AV/AV_Core.cpp: + * orbsvcs/orbsvcs/HTIOP/HTIOP_Acceptor.cpp: + * orbsvcs/orbsvcs/IFRService/Container_i.cpp: + * orbsvcs/ImplRepo_Service/ImR_Locator_i.cpp: + * orbsvcs/ImplRepo_Service/utils.h: + * orbsvcs/orbsvcs/PortableGroup/UIPMC_Acceptor.cpp: + * orbsvcs/orbsvcs/PortableGroup/UIPMC_Profile.cpp: + * orbsvcs/tests/AVStreams/Asynch_Three_Stage/Connection_Manager.cpp: + * orbsvcs/tests/InterfaceRepo/Application_Test/test_i.cpp: + * tao/Strategies/DIOP_Acceptor.cpp: + * tao/Strategies/SHMIOP_Acceptor.cpp: + * tao/Strategies/UIOP_Acceptor.cpp: + + Use new ACE_CString::size_type type to hold string + position/index value, instead of "int". Addresses "comparison + between signed and unsigned" warnings. + +Mon Jul 17 14:46:26 UTC 2006 Ciju John <johnc@ociweb.com> + + * orbsvcs/orbsvcs/Event/ECG_Mcast_EH.cpp: + + Replaced deprecated call to ACE_SOCK_Dgram_Mcast::subscribe() + with ACE_SOCK_Dgram_Mcast::join(). Thanks to Johnny for pointing + this out. + +Mon Jul 17 14:40:43 UTC 2006 Chad Elliott <elliott_c@ociweb.com> + + * TAO_IDL/be/be_visitor_union/union_cs.cpp: + + The memset of the discriminator was redundant as it is assigned + directly after the memset of the union data. + + * tao/CDR.h: + * tao/CDR.cpp: + + Added a constructor that takes an ACE_Data_Block* which can be + used to avoid an allocation while constructing the + ACE_Message_Block held by the ACE_OutputCDR. [RT 8372] + + * tao/CSD_Framework/CSD_FW_Server_Request_Wrapper.h: + * tao/CSD_Framework/CSD_FW_Server_Request_Wrapper.cpp: + + Use the new inline TAO_ServerRequest constructor. This avoids + having to allocate a costly dummy input and output cdr. + + Avoid using ACE_CString for the operation name which will avoid + memory allocations when it isn't necessary. Most of the time, the + operation name can be shared from the operation_details_ object. + + When the type of TAO_Tagged_Profile::type_id_ changed in the + previous commit, the cloning needed to take that into account. + [RT 8372] + + When cloning the operation details, allocate an input cdr with a + "large" message block that will be shared by the output cdr as + the marshaling of the arguments is done. This allows us to + avoid the copying of the data block, in most situations, from the + output cdr into the input cdr as was done in the previous version. + [RT 8372] + + * tao/CSD_ThreadPool.mpc: + + Added a requires for threads as this project shouldn't be built if + thread support is not going to be built into TAO. + + * tao/Client_Strategy_Factory.h: + * tao/Client_Strategy_Factory.cpp: + * tao/Configurable_Refcount.h: + * tao/Configurable_Refcount.inl: + * tao/Configurable_Refcount.cpp: + * tao/Object.h: + * tao/Object.i: + * tao/Object.cpp: + * tao/Profile.h: + * tao/Profile.cpp: + * tao/Resource_Factory.h: + * tao/Resource_Factory.cpp: + * tao/default_client.h: + * tao/default_client.cpp: + * tao/default_resource.h: + * tao/tao.mpc: + + The TAO_Configurable_Refcount can be used as an ACE_Atomic_Op with + either a null mutex or a thread mutex. Those classes that use it + used to use either a null mutex or a thread mutex and an integer + based reference count. + + * tao/GIOP_Message_Base.cpp: + + When processing a request, a character buffer is allocated on the + stack. Previously, it was passed as a parameter to the + TAO_OutputCDR and an ACE_Data_Block was heap allocated during the + construction of the ACE_Message_Block held by the ACE_OutputCDR. + + Now, the ACE_Data_Block is constructed on the stack as well and + passed in during the construction of the ACE_Message_Block to + avoid heap allocations altogether. [RT 8372] + + * tao/GIOP_Message_Lite.cpp: + * tao/Incoming_Message_Queue.h: + * tao/Incoming_Message_Queue.cpp: + + Factored various forms of creating a TAO_Queued_Data (with and + without an ACE_Message_Block) into a single method. Doing this + allows us to avoid extra manipulations of the reference count in + the ACE_Data_Block involved. + + * tao/Environment.i: + * tao/Leader_Follower.cpp: + * tao/TAO_Singleton_Manager.cpp: + * tao/Var_Size_Argument_T.cpp: + * tao/default_resource.cpp: + * tao/IORTable/Table_Adapter.cpp: + * tao/CSD_ThreadPool/CSD_TP_Collocated_Synch_Request.cpp: + * tao/RTCORBA/Thread_Pool.cpp: + + Replace many instances of the use of a bare 'new' with the + equivalent ACE_NEW or ACE_NEW_RETURN macro. [RT 8372] + + * tao/LocalObject.h: + * tao/Principal.h: + * tao/Stub.h: + * tao/Stub.cpp: + * tao/AnyTypeCode/Any_Impl.h: + * tao/AnyTypeCode/NVList.h: + * tao/Messaging/AMH_Response_Handler.cpp: + * tao/Valuetype/ValueBase.h: + + Use unsigned long instead of CORBA::ULong to be able to take + advantage of the ACE_Atomic_Op<> template specialization. + CORBA::ULong doesn't necessarily equate to unsigned long. + + * tao/ORB.h: + * tao/ORB.i: + * tao/ORB.cpp: + * tao/ORB_Core.h: + * tao/ORB_Core.i: + + Switch from using a thread mutex and CORBA::ULong to an + ACE_Atomic_Op<> with a thread mutex and unsigned long. When the + assembler based specialization of ACE_Atomic_Op<> is used, there + is a performance benefit. + + * tao/PortableServer/POA_Current_Impl.h: + * tao/PortableServer/POA_Current_Impl.inl: + * tao/PortableServer/POA_Current_Impl.cpp: + * tao/PortableServer/Servant_Upcall.h: + * tao/PortableServer/Servant_Upcall.cpp: + + Use an octet sequence array of a specific size (512 bytes by + default) when constructing the ObjectId. This allows us to avoid + allocations during an upcall in most situations. [RT 8372] + + * tao/PortableServer/Root_POA.cpp: + + Begin the starting_at index at the TAO_OBJECTKEY_PREFIX_SIZE + instead of starting it at zero and then adding the value directly + after that. This is probably immeasurable performance-wise. + + * tao/TAO_Server_Request.h: + * tao/TAO_Server_Request.i: + * tao/TAO_Server_Request.cpp: + + Avoid using ACE_CString for the operation name which will avoid + memory allocations when it isn't necessary. Most of the time, the + operation name can be shared from the operation_details_ object. + + * tao/Tagged_Profile.h: + * tao/Tagged_Profile.i: + * tao/Tagged_Profile.cpp: + + Avoid using ACE_CString as the type_id_. The type_id_ was never + used for anything more than a holder of a char* that it didn't + own, but incurred allocation penalties during construction. + +Mon Jul 17 14:26:08 UTC 2006 Ciju John <john_c@ociweb.com> + + * orbsvcs/orbsvcs/Event/ECG_Mcast_EH.cpp: + + Addendum to + Fri Jul 14 18:34:41 UTC 2006 john_c <johnc@ociweb.com> + + Previously the code created a connection, then made the socket + non-blocking. This change first make the socket + non-blocking then initiates the connection. Previous code left a + small window, which I think was causing a hangup in certain + conditions. + +Mon Jul 17 13:15:58 UTC 2006 Chad Elliott <elliott_c@ociweb.com> + + * orbsvcs/examples/ORT/client.cpp: + + Modified this code to correctly utilize the exception macros. It + was using ACE_TRY_ENV as a parameter to CORBA::ORB_init(). + +Mon Jul 17 12:55:48 UTC 2006 Chad Elliott <elliott_c@ociweb.com> + + * orbsvcs/tests/Event/Mcast/Complex/consumer.cpp: + + Modified this code to correctly utilize the exception macros. It + was using ACE_TRY_ENV as a parameter to CORBA::ORB_init(). + +Mon Jul 18 11:37:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Strategies.mpc: + Enabled regeneration of the pidl files + + * tao/Strategies/sciop_endpoints.pidl: + * tao/Strategies/uiop_endpoints.pidl: + Updated comments block to resolve warnings during recompilation + + * tao/Strategies/uiop_endpointsC.{h,cpp}: + * tao/Strategies/sciop_endpointsC.{h,cpp}: + Removed from the repo + +Mon Jul 18 11:00:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Valuetype/ValueBase.{h,cpp}: + Removed typecode for ValueModifier, it is also in the AnyTypeCode + library but it seems that in the handcrafted code the export + of this typecode was gone, now we regenate things automatically + some linkers complain about the duplicate symbols + +Mon Jul 18 10:10:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/DynamicAny/*.cpp: + Reordered includes and added missing includes to fix compile + errors with gcc + +Mon Jul 17 09:50:45 UTC 2006 Adam Mitz <mitza@ociweb.com> + + * orbsvcs/tests/CosEvent/Timeout/TimeoutTestMain.cpp: + Eliminate warning with String_var usage on GCC 2.96. + +Mon Jul 18 07:58:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/IFR_Client.mpc: + Removed -Sorb, gives problems for users of this lib + +Sun Jul 17 06:49:21 UTC 2006 Ossama Othman <ossama_othman at symantec dot com> + + * tao/PortableServer/Servant_var.inl (_retn): + + Removed extraneous "this" pointer usage on local variable. + +Mon Jul 17 00:20:29 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * NEWS: + Added news entry re Bug 2593. + +Sun Jul 16 23:28:27 UTC 2006 Adam Mitz <mitza@ociweb.com> + + * orbsvcs/orbsvcs/CosEvent/CEC_Default_Factory.h: + * orbsvcs/orbsvcs/CosEvent/CEC_Default_Factory.cpp: + * orbsvcs/orbsvcs/CosEvent/CEC_EventChannel.h: + * orbsvcs/orbsvcs/CosEvent/CEC_EventChannel.cpp: + * orbsvcs/orbsvcs/CosEvent/CEC_Factory.h: + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPullConsumer.h: + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPullConsumer.cpp: + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPullSupplier.h: + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPullSupplier.cpp: + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPushConsumer.h: + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPushConsumer.cpp: + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPushSupplier.h: + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPushSupplier.cpp: + * orbsvcs/orbsvcs/CosEvent/CEC_TypedEventChannel.h: + * orbsvcs/orbsvcs/CosEvent/CEC_TypedEventChannel.cpp: + * orbsvcs/orbsvcs/CosEvent/CEC_TypedProxyPushConsumer.h: + * orbsvcs/orbsvcs/CosEvent/CEC_TypedProxyPushConsumer.cpp: + + Change method signatures to pass ACE_Time_Value by const-reference. + +Sat Jul 15 22:08:58 UTC 2006 Shanshan Jiang <shanshan.jiang@vanderbilt.edu> + + * docs/tutorials/Quoter/RTCORBA/Admin.cpp + * docs/tutorials/Quoter/RTCORBA/Broker.cpp + * docs/tutorials/Quoter/RTCORBA/Broker_i.cpp + * docs/tutorials/Quoter/RTCORBA/Broker_i.h + * docs/tutorials/Quoter/RTCORBA/Common_i.cpp + * docs/tutorials/Quoter/RTCORBA/Common_i.h + * docs/tutorials/Quoter/RTCORBA/Distributor.cpp + * docs/tutorials/Quoter/RTCORBA/Distributor_i.cpp + * docs/tutorials/Quoter/RTCORBA/Distributor_i.h + * docs/tutorials/Quoter/RTCORBA/Stock_Database.cpp + * docs/tutorials/Quoter/RTCORBA/Stock_Database.h + * docs/tutorials/Quoter/RTCORBA/Stock_PriorityMapping.h + + * docs/tutorials/Quoter/RTCORBA/docs/broker.html + * docs/tutorials/Quoter/RTCORBA/docs/common.html + * docs/tutorials/Quoter/RTCORBA/docs/database.html + * docs/tutorials/Quoter/RTCORBA/docs/distributor.html + * docs/tutorials/Quoter/RTCORBA/docs/index.html + * docs/tutorials/Quoter/RTCORBA/docs/mpc.html + + Modify these files based on the suggestions of William R. Otte. + +Sat Jul 15 12:25:11 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/tests/Bug_1393_Regression/client.cpp: + * orbsvcs/tests/Bug_1630_Regression/testclient.cpp: + * orbsvcs/tests/InterfaceRepo/IDL3_Test/idl3_client.cpp: + * orbsvcs/tests/InterfaceRepo/IFR_Inheritance_Test/main.cpp: + * orbsvcs/tests/InterfaceRepo/IFR_Test/Admin_Client.h: + * orbsvcs/tests/InterfaceRepo/Persistence_Test/Ptest.h: + Added needed include + +Sat Jul 15 05:07:11 UTC 2006 William R. Otte <wotte@dre.vanderbilt.edu> + + * docs/tutorials/Quoter/index.html + + Added a link to the RTCORBA tutorial. + + * docs/tutorials/Quoter/RTCORBA/Broker.idl + * docs/tutorials/Quoter/RTCORBA/Common.idl + * docs/tutorials/Quoter/RTCORBA/Distributor.idl + + Corrected IDL files which had been corrupted by overzealous + removal of DOS style line endings. + +Sat Jul 15 00:47:30 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/IDL_Test/IDL_Test.mpc: + Disabled the regression for bug 2582, I tested it weeks ago and + it worked, but now it fails when I added the test + +Sat Jul 15 00:42:30 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/tao.mpc: + Fixed casing of include + +Fri Jul 14 23:44:45 UTC 2006 Ossama Othman <ossama_othman at symantec dot com> + + * tao/Utils/Servant_Var.h: + + Removed "#warning" preprocess directive. It's not portable. + Thanks to Jeff for pointing this out. + +Fri Jul 14 23:42:30 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * TAO_IDL/be/be_codegen.cpp: + When generating a pidl file change the guard we generate in the + header file. Fixes problems with TimeBase.pidl and TimeBase.idl + which both exist. + +Fri Jul 14 21:43:55 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/IIOP_Acceptor.cpp (parse_address): + Reverted change at Johnny's request. See bug 2599 for a + discussion of the reason why. + +Fri Jul 14 21:43:58 UTC 2006 john_c <johnc@ociweb.com> + + * orbsvcs/tests/Event/Mcast/RTEC_MCast_Federated/EchoEventConsumer_i.cpp: + * orbsvcs/tests/Event/Mcast/RTEC_MCast_Federated/EchoEventSupplierMain.cpp: + * orbsvcs/tests/Event/Mcast/RTEC_MCast_Federated/EchoEventSupplier_i.cpp: + * orbsvcs/tests/Event/Mcast/RTEC_MCast_Federated/RTEC_MCast_Federated.mpc: + * tests/Optimized_Connection/oc_client.cpp: + + Minor fixes to make Fuzz builds happy. + +Fri Jul 14 21:42:30 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/AnyTypeCode/BooleanSeqA.* + * tao/AnyTypeCode/CharSeqA.* + * tao/AnyTypeCode/CONV_FRAMEA.* + * tao/AnyTypeCode/CurrentA.* + * tao/AnyTypeCode/DoubleSeqA.* + * tao/AnyTypeCode/FloatSeqA.* + * tao/AnyTypeCode/GIOPA.* + * tao/AnyTypeCode/IIOP_EndpointsA.* + * tao/AnyTypeCode/IIOPA.* + * tao/AnyTypeCode/IOP_IORA.* + * tao/AnyTypeCode/IOPA.* + * tao/AnyTypeCode/LongDoubleSeqA.* + * tao/AnyTypeCode/LongLongSeqA.* + * tao/AnyTypeCode/LongSeqA.* + * tao/AnyTypeCode/Messaging_PolicyValueA.* + * tao/AnyTypeCode/Messaging_SyncScopeA.* + * tao/AnyTypeCode/ObjectIdListA.* + * tao/AnyTypeCode/orb_typesA.* + * tao/AnyTypeCode/ParameterModeA.* + * tao/AnyTypeCode/PI_ForwardA.* + * tao/AnyTypeCode/Policy_ForwardA.* + * tao/AnyTypeCode/PortableInterceptorA.* + * tao/AnyTypeCode/ServicesA.* + * tao/AnyTypeCode/ShortSeqA.* + * tao/AnyTypeCode/TAOA.* + * tao/AnyTypeCode/TimeBaseA.* + * tao/AnyTypeCode/ULongLongSeqA.* + * tao/AnyTypeCode/ULongSeqA.* + * tao/AnyTypeCode/WCharSeqA.* + * tao/AnyTypeCode/WStringSeqA.* + * tao/BooleanSeqC.* + * tao/CharSeqC.* + * tao/CONV_FRAMEC.* + * tao/CurrentC.* + * tao/DoubleSeqC.* + * tao/FloatSeqC.* + * tao/GIOPC.* + * tao/IIOP_EndpointsC.* + * tao/IIOPC.* + * tao/IOP_IORC.* + * tao/IOPC.* + * tao/LongDoubleSeqC.* + * tao/LongLongSeqC.* + * tao/LongSeqC.* + * tao/Messaging_PolicyValueC.* + * tao/Messaging_SyncScopeC.* + * tao/ObjectIdListC.* + * tao/orb_typesC.* + * tao/ParameterModeC.* + * tao/PI_ForwardC.* + * tao/Policy_ForwardC.* + * tao/PortableInterceptorC.* + * tao/ServicesC.* + * tao/ShortSeqC.* + * tao/StringSeqC.* + * tao/TAOC.* + * tao/TimeBaseC.* + * tao/ULongLongSeqC.* + * tao/ULongSeqC.* + * tao/WCharSeqC.* + * tao/WStringSeqC.* + Removed from the repo, all regenerated as part of the build + + * tao/tao.mpc: + Enabled regeneration for a lot of pidl files + + * tao/DynamicAny/diffs/DynamicAny.diff: + Removed, not needed anymore + +Fri Jul 14 21:10:30 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/DynamicAny.mpc: + Enabled regeneration of the pidl files + + * tao/DynamicAny/DynamicAnyC.{h,cpp}: + Removed from the repo, will get regenerated automatically + + * tao/*Resource*.*: + Moved the different resources files (Qt,Fl,Tk, and Xt) to + subdirectories so that in TAO/tao we only have the tao lib + core + +Fri Jul 14 21:15:32 UTC 2006 Ossama Othman <ossama_othman at symantec dot com> + + * TAO_IDL/ast/ast_decl.cpp (version): + * TAO_IDL/be/be_codegen.cpp: + * TAO_IDL/be/be_visitor_operation/amh_rh_ss.cpp: + * tao/Acceptor_Registry.cpp: + * tao/CORBANAME_Parser.cpp: + * tao/IIOP_Acceptor.cpp: + * tao/IIOP_Endpoint.cpp: + * tao/IIOP_Profile.cpp: + * tao/Transport_Connector.cpp: + * tao/params.cpp: + * tao/CSD_ThreadPool/CSD_TP_Strategy_Factory.cpp: + * tao/TypeCodeFactory/TypeCodeFactory_i.cpp: + + ACE string class operations now consistently return and accept an + unsigned index/position value instead of a mix of signed and + unsigned return index types. Use the "size_type" typedef + instead of int or ssize_t, and the "npos" constant instead of + "-1". + + * TAO_IDL/fe/fe_init.cpp: + + Use ACE_CString::npos instead of ACE_String_Base_Const::npos. + The latter is not meant for use outside of ACE. + + * orbsvcs/Trading_Service/Trading_Service.h: + * orbsvcs/orbsvcs/Shutdown_Utilities.h: + + Include "ace/Sig_Handler.h" to pull in ACE_Sig_Handler class + declaration. + + * orbsvcs/orbsvcs/Notify/ConsumerAdmin.h: + + Moved TAO_Notify_ProxySupplier forward declaration within the + TAO versioned namespace. + + * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Accept_Strategy.cpp + (accept_svc_handler): + + Ensure that errno is preserved in case the svc_handler close() + method resets it. + + * orbsvcs/tests/CosEvent/lib/CosEvent_lib.mpc: + + Removed "naming" MPC base project from "CosEvent Test Lib" + project. That test library does not depending on the Naming + Service. + + * tao/GIOP_Message_Base.cpp: + + Cosmetic changes. + + * tao/ORB_Core.cpp: + + Access ORB table lock using the new lock accessor rather than + directly using the lock member because of a (now removed) friend + declaration. Folks, stop being lazy. :) + + * tao/ORB_Table.h: + * tao/ORB_Table.inl: + + The ORB table now has no friends, nor should it! HP aCC 3.63 + was confused by the global scope qualifier in the friend + declaration, due to broken namespace scoping support. + + Added a public lock accessor method. + + * tao/Policy_Manager.h (TAO_Policy_Manager, operator=): + + Added private and unimplemented copy constructor and assignment + operator. TAO_Policy_Manager should not be copied. + + * tao/PortableServer.mpc (Template_Files): + + Explicitly list template files since the Servant_var.* template + sources will otherwise not be considered template sources. + + * tao/AnyTypeCode/Any_Array_Impl_T.cpp (marshal_value, value): + * tao/AnyTypeCode/Any_Array_Impl_T.inl: + * tao/AnyTypeCode/Any_Basic_Impl_T.cpp: + * tao/AnyTypeCode/Any_Basic_Impl_T.inl: + * tao/AnyTypeCode/Any_Dual_Impl_T.cpp: + * tao/AnyTypeCode/Any_Dual_Impl_T.inl: + * tao/AnyTypeCode/Any_Impl_T.cpp: + * tao/AnyTypeCode/Any_Impl_T.inl: + * tao/AnyTypeCode/Any_Special_Impl_T.cpp: + * tao/AnyTypeCode/Any_Special_Impl_T.inl: + + Moved inline virtual methods out of line. Virtual methods + generally should not be inlined. Addresses potential RTTI + problems with g++ 4.x. + + * tao/PortableServer/Object_Adapter.cpp + (wait_for_non_servant_upcalls_to_complete): + + Removed explicit use of CORBA::Environment in native C++ + exception builds. It isn't needed. + + (dispatch, get_collocated_servant): + + Prevent potential buffer overread. (From Andrew Schnable) + + * tao/PortableServer/Object_Adapter.h + (wait_for_non_servant_upcalls_to_complete): + + Removed explicit use of CORBA::Environment in native C++ + exception builds. It isn't needed. + + * tao/PortableServer/PS_ForwardC.h: + + PortableServer::ServantBase_var is now a specialization of the + new (standard) PortableServer::Servant_var<> class template. + + * tao/PortableServer/Servant_Base.cpp: + * tao/PortableServer/Servant_Base.h: + * tao/PortableServer/Servant_Base.i: + + Removed TAO_ServantBase_var class. It's functionality is now + provided by the PortableServer::Servant_var<> class template. + + * tao/PortableServer/Servant_Upcall.cpp (upcall_cleanup): + + Renamed wait_for_non_servant_upcall_to_complete() method + call to the name (..._to_complete_no_throw()). + + * tao/PortableServer/Servant_var.cpp: + * tao/PortableServer/Servant_var.h: + * tao/PortableServer/Servant_var.inl: + + Implementation of the standard PortableServer::Servant_var<> + class template. + + * tao/Utils/Servant_Var.h: + + Marked TAO::Utils::Servant_var<> class template as deprecated. + Use standard PortableServer::Servant_var class template + instead. + + * utils/catior/catior.cpp (catior): + + Fixed crash due to buffer overrun. The message block write + pointer should be set to the end of the message block, not twice + the length of message block. Thanks to Sriram Chandramouli + <sriram_chandramouli at symantec dot com> for pointing out this + problem. + +Fri Jul 14 21:10:30 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/DynamicAny.mpc: + Enabled regeneration of the pidl files + + * tao/DynamicAny/DynamicAnyC.{h,cpp}: + Removed from the repo, will get regenerated automatically + + * tao/*Resource*.*: + Moved the different resources files (Qt,Fl,Tk, and Xt) to + subdirectories so that in TAO/tao we only have the tao lib + core + +Fri Jul 14 21:03:30 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/IDL_Test/IDL_Test.mpc: + This fix closes [BUGID:2582]. Thanks to Heesuk Shin + <heesuks@econz.com> for reporting the bug. + Also fixes bug 2583. + +Fri Jul 14 20:50:30 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * TAO_IDL/ast/ast_decl.cpp: + * TAO_IDL/ast/ast_typedef.cpp: + * TAO_IDL/include/ast_decl.h: + * TAO_IDL/include/ast_typedef.h: + + Fixed bug in the recursive check for local type + containment that was causing CDR operators to + get generated for types that contain a local + interface at some level of nesting. This fix + closes [BUGID:2583]. + +Fri Jul 14 20:38:07 UTC 2006 john_c <johnc@ociweb.com> + + * tests/AlternateIIOP/run_test.pl: + + Split a long string. + + * tests/Optimized_Connection/Optimized_Connection.mpc: + * tests/Optimized_Connection/README: + * tests/Optimized_Connection/oc.ior: + * tests/Optimized_Connection/oc_client.cpp: + * tests/Optimized_Connection/oc_svc.conf: + * tests/Optimized_Connection/run_test.pl: + + Created new test to test the Optimized_Connection Selector. + + * tests/AlternateIIOP/svc.conf: + + Removed this file. This test shouldn't be testing the Optimized + Connection Selector. + +Fri Jul 14 19:20:34 UTC 2006 William R. Otte <wotte@dre.vanderbilt.edu> + + * docs/tutorials/Quoter/RTCORBA/Admin.cpp + * docs/tutorials/Quoter/RTCORBA/Broker.cpp + * docs/tutorials/Quoter/RTCORBA/Broker.idl + * docs/tutorials/Quoter/RTCORBA/Broker_i.cpp + * docs/tutorials/Quoter/RTCORBA/Broker_i.h + * docs/tutorials/Quoter/RTCORBA/Common.idl + * docs/tutorials/Quoter/RTCORBA/Common_i.cpp + * docs/tutorials/Quoter/RTCORBA/Common_i.h + * docs/tutorials/Quoter/RTCORBA/Distributor.cpp + * docs/tutorials/Quoter/RTCORBA/Distributor.idl + * docs/tutorials/Quoter/RTCORBA/Distributor_i.cpp + * docs/tutorials/Quoter/RTCORBA/Distributor_i.h + * docs/tutorials/Quoter/RTCORBA/StockQuoter.mpc + * docs/tutorials/Quoter/RTCORBA/Stock_Database.cpp + * docs/tutorials/Quoter/RTCORBA/Stock_Database.h + * docs/tutorials/Quoter/RTCORBA/Stock_PriorityMapping.cpp + * docs/tutorials/Quoter/RTCORBA/Stock_PriorityMapping.h + * docs/tutorials/Quoter/RTCORBA/docs/broker.html + * docs/tutorials/Quoter/RTCORBA/docs/common.html + * docs/tutorials/Quoter/RTCORBA/docs/database.html + * docs/tutorials/Quoter/RTCORBA/docs/distributor.html + * docs/tutorials/Quoter/RTCORBA/docs/idl.html + * docs/tutorials/Quoter/RTCORBA/docs/index.html + * docs/tutorials/Quoter/RTCORBA/docs/mpc.html + * docs/tutorials/Quoter/RTCORBA/docs/prioritymapping.html + * docs/tutorials/Quoter/RTCORBA/docs/run.html + + A RTCORBA tutorial based on the Quoter by Shanshan Jiang. + +Fri Jul 14 18:34:41 UTC 2006 john_c <johnc@ociweb.com> + + * orbsvcs/orbsvcs/Event/ECG_Mcast_EH.cpp: + + Set handle options after a proper handle has been assigned. This + was causing a race condition at a client location. We couldn't + reproduce the problem, but the change is valid. + + * orbsvcs/tests/Event/Mcast/RTEC_MCast_Federated/EchoEventConsumerMain.cpp: + * orbsvcs/tests/Event/Mcast/RTEC_MCast_Federated/EchoEventConsumer_i.h: + * orbsvcs/tests/Event/Mcast/RTEC_MCast_Federated/EchoEventConsumer_i.cpp: + * orbsvcs/tests/Event/Mcast/RTEC_MCast_Federated/EchoEventSupplierMain.cpp: + * orbsvcs/tests/Event/Mcast/RTEC_MCast_Federated/EchoEventSupplier_i.h: + * orbsvcs/tests/Event/Mcast/RTEC_MCast_Federated/EchoEventSupplier_i.cpp: + * orbsvcs/tests/Event/Mcast/RTEC_MCast_Federated/README: + * orbsvcs/tests/Event/Mcast/RTEC_MCast_Federated/RTEC_MCast_Federated.mpc: + * orbsvcs/tests/Event/Mcast/RTEC_MCast_Federated/SimpleAddressServer.h: + * orbsvcs/tests/Event/Mcast/RTEC_MCast_Federated/SimpleAddressServer.cpp: + * orbsvcs/tests/Event/Mcast/RTEC_MCast_Federated/run_test.pl: + * orbsvcs/tests/Event/Mcast/RTEC_MCast_Federated/supplier.conf: + + Adding a test to traverse the changed code path. + +Fri Jul 14 17:00:49 UTC 2006 Yan Dai <dai_y@ociweb.com> + + * tao/EndpointPolicy/Endpoint_Acceptor_Filter_Factory.cpp: + + Fixed a bug in create_object() that uses an incorrect index + to access the policy in the policy list. + +Fri Jul 14 16:01:38 UTC 2006 Adam Mitz <mitza@ociweb.com> + + * tests/Bug_2494_Regression/client.cpp: + * tests/Bug_2494_Regression/run_test.pl: + + Eliminate false failures of this test due to the client process + printing the exception or exiting with a non-zero status. + +Fri Jul 14 15:11:11 UTC 2006 Adam Mitz <mitza@ociweb.com> + + * orbsvcs/orbsvcs/CosEvent/CEC_Default_Factory.h: + * orbsvcs/orbsvcs/CosEvent/CEC_Default_Factory.cpp: + * orbsvcs/orbsvcs/CosEvent/CEC_EventChannel.h: + * orbsvcs/orbsvcs/CosEvent/CEC_EventChannel.cpp: + * orbsvcs/orbsvcs/CosEvent/CEC_Event_Loader.cpp: + * orbsvcs/orbsvcs/CosEvent/CEC_Factory.h: + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPullConsumer.h: + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPullConsumer.cpp: + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPullSupplier.h: + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPullSupplier.cpp: + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPushConsumer.h: + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPushConsumer.cpp: + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPushSupplier.h: + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPushSupplier.cpp: + * orbsvcs/orbsvcs/CosEvent/CEC_TypedEventChannel.h: + * orbsvcs/orbsvcs/CosEvent/CEC_TypedEventChannel.cpp: + * orbsvcs/orbsvcs/CosEvent/CEC_TypedProxyPushConsumer.h: + * orbsvcs/orbsvcs/CosEvent/CEC_TypedProxyPushConsumer.cpp: + + Added two new options to the CEC_Default_Factory, + -CECConsumerOperationTimeout and -CECSupplierOperationTimeout. + These options apply relative roundtrip timeout policies for consumer + and supplier objects. This allows users to minimize the ill effects of + unresponsive consumer and supplier operations (i.e. push() or pull()). + See bugzilla #2594 and docs/cec_options.html. + Also, exposed the "disconnect_callbacks" event channel attribute + as the "-b" option to the CosEvent_Service executable. + + * NEWS + * docs/cec_options.html: + + Documented the Timeout feature described above. + + * orbsvcs/tests/CosEvent/Timeout/README: + * orbsvcs/tests/CosEvent/Timeout/TestEventConsumer_i.h: + * orbsvcs/tests/CosEvent/Timeout/TestEventConsumer_i.cpp: + * orbsvcs/tests/CosEvent/Timeout/Timeout.mpc: + * orbsvcs/tests/CosEvent/Timeout/TimeoutTestMain.cpp: + * orbsvcs/tests/CosEvent/Timeout/cosevent.conf: + * orbsvcs/tests/CosEvent/Timeout/run_test.pl: + + Functional test for the Timeout feature described above. + +Fri Jul 14 02:02:55 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/IIOP_Acceptor.cpp: + Fix for parsing IPv6 endpoints not using the enclosing []. An + address such as fe80::a0f2:2510:7549:12345 is a legitemate IPv6 + address but would be misinterpreted. + + * tao/Profile.cpp: + * tests/Bug_2593_Regression/Bug_2593_Regression.mpc: + * tests/Bug_2593_Regression/Hello.h: + * tests/Bug_2593_Regression/Hello.cpp: + * tests/Bug_2593_Regression/README: + * tests/Bug_2593_Regression/Test.idl: + * tests/Bug_2593_Regression/client.cpp: + * tests/Bug_2593_Regression/run_test.pl: + * tests/Bug_2593_Regression/server.conf: + * tests/Bug_2593_Regression/server.cpp: + A fix and a new regression test for bug 2593. Thanks to Jules + Colding, <colding at omesc dot com> for finding this. + +Thu Jul 13 10:26:42 UTC 2006 Ossama Othman <ossama_othman at symantec dot com> + + * tao/EndpointPolicy/EndpointPolicyTypeA.h: + * tao/EndpointPolicy/Endpoint_Value_Impl.cpp: + + Added missing versioned namespace macros. + +Thu Jul 12 18:45:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/ObjRefTemplate.mpc: + Added versioning base project + +Wed Jul 12 18:01:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/IORManipulation.mpc: + Enabled regeneration of pidl files + + * tao/IORManipulation/IORA.h: + Temporarily new file to keep other A.h files happy + + * tao/IORManipulation/IORC.{h,cpp}: + * tao/IORManipulation/IORA.cpp: + Removed from the repo + +Wed Jul 12 10:25:12 UTC 2006 Martin Corino <mcorino@remedy.nl> + + * tests/CollocationLockup/CollocationLockup.cpp: + * tests/CollocationLockup/run_test.pl: + Added some customization to have this test run correctly on OpenVMS. + + * tests/Object_Loader/Loader.cpp: + * tests/Smart_Proxies/On_Demand/Smart_Test_Factory.cpp: + Removed deprecated OpenVMS specific include. + +Tue Jul 11 12:12:01 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/BiDir_GIOP/diffs/BiDirPolicy.diff: + * tao/CodecFactory/diff/IOP_Codec.diff: + * tao/IORInterceptor/diff/IORInterceptor.diff: + * tao/IORTable/diff/IORTable.diff: + * tao/ObjRefTemplate/diffs/ObjectReferenceTemplate.diff: + * tao/RTPortableServer/diffs/RTPortableServer.diff: + Removed, not needed anymore + + * tao/RTPortableServer.mpc: + Enabed regeneration of the pidl files + + * tao/RTPortableServer/RTPortableServerA.cpp: + * tao/RTPortableServer/RTPortableServerC.cpp: + * tao/RTPortableServer/RTPortableServerC.h: + * tao/RTPortableServer/RTPortableServerS.h: + Removed from the repo + +Tue Jul 11 12:07:01 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/AnyTypeCode.mpc: + Enabled regeneration of the pidl files + + * tao/AnyTypeCode/TypeCode.h: + Added include that was first pulled in indirectly + + * tao/AnyTypeCode/AnySeqA.cpp + * tao/AnyTypeCode/AnySeqA.h + * tao/AnyTypeCode/AnySeqC.cpp + * tao/AnyTypeCode/AnySeqC.h + * tao/AnyTypeCode/AnySeqS.h + * tao/AnyTypeCode/BoundsA.cpp + * tao/AnyTypeCode/BoundsA.h + * tao/AnyTypeCode/BoundsC.cpp + * tao/AnyTypeCode/BoundsC.h + * tao/AnyTypeCode/Dynamic_ParameterA.cpp + * tao/AnyTypeCode/Dynamic_ParameterA.h + * tao/AnyTypeCode/Dynamic_ParameterC.cpp + * tao/AnyTypeCode/Dynamic_ParameterC.h + * tao/AnyTypeCode/ValueModifierA.cpp + * tao/AnyTypeCode/ValueModifierA.h + * tao/AnyTypeCode/ValueModifierC.cpp + * tao/AnyTypeCode/ValueModifierC.h + * tao/AnyTypeCode/ValueModifierS.h + * tao/AnyTypeCode/VisibilityA.cpp + * tao/AnyTypeCode/VisibilityA.h + * tao/AnyTypeCode/VisibilityC.cpp + * tao/AnyTypeCode/VisibilityC.h + * tao/AnyTypeCode/VisibilityS.h + Removed from the repository, regenerated now automatically + +Tue Jul 11 11:59:01 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/PortableServer/Object_Adapter.cpp: + Fix for CORBA/e + +Tue Jul 11 11:08:01 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/BiDir_GIOP.mpc: + Fixed casing of include + +Tue Jul 11 08:30:00 UTC 2006 Simon Massey <sma@prismtech.com> + + * utils/nslist/nslist.cpp: + * utils/nslist/run_test.pl: + Fix bugzilla 2592. + +Tue Jul 11 08:10:00 UTC 2006 Simon Massey <sma@prismtech.com> + + * orbsvcs/orbsvcs/Naming/Flat_File_Persistence.cpp: + Fix bugzilla 2591. + +Tue Jul 11 01:43:01 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/IDL_Test/Bug_2582_Regression.idl: + * tests/IDL_Test/Bug_2583_Regression.idl: + New tests for TAO_IDL, mpc file will be committed when the latest + change from Jeff is in the archive + +Tue Jul 11 01:40:01 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/TAO_Singleton_Manager.cpp: + Initialise pointer with 0 + +Tue Jul 11 01:20:01 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * TAO_IDL/be_include/be_global.h: + * TAO_IDL/be/be_global.cpp: + * docs/compiler.html: + * TAO_IDL/be/be_codegen.cpp: + Added support for -Wb,include_guard= and -Wb,safe_include=. + The include guard is a define that is generated in the C.h file + that prevents the user from including this generated file, the + safe_include is the file that the user should include instead + of the C.h and which is also used in the generated A.cpp and + C.cpp. With these new options a lot of generated files can be + removed also from the repository + + * tao/RTCORBA/RT_Transport_Descriptor.cpp: + * tao/RTCORBA/RT_Invocation_Endpoint_Selectors.cpp: + * tao/Valuetype/ValueBase.cpp: + Bool and const changes + + * tao/PI_Server/PortableServer_ORBInitializer.cpp: + * tao/PI_Server/PortableServer_PolicyFactory.cpp: + * tao/PortableServer/Root_POA.cpp: + * tao/PortableServer/Object_Adapter.cpp: + Fixes for CORBA/e + + * tao/CodecFactory.mpc: + * tao/BiDir_GIOP.mpc: + * tao/IORInterceptor.mpc: + * tao/IORTable.mpc: + * tao/ObjRefTemplate.mpc: + Enabled regeneration of the pidl files + + * tao/CodecFactory/IOP_CodecC.{h,cpp}: + * tao/CodecFactory/IOP_CodecS.h: + * tao/BiDir_GIOP/BiDirPolicyC.{h,cpp}: + * tao/IORInterceptor/IORInfoC.{h,cpp}: + * tao/IORInterceptor/IORInterceptorC.{h,cpp}: + * tao/IORTable/IORTableC.{h,cpp}: + * tao/ObjRefTemplate/Default_ORTC.{h,cpp,inl}: + * tao/ObjRefTemplate/ObjectReferenceTemplateC.{h,cpp,inl}: + Removed from the repo, will get regenerated as part of the + building of the core lib. + + * tao/BiDir_GIOP/BiDirPolicy.pidl: + * tao/IIOP_Endpoints.pidl: + Updated comments to resolve warnings during recompilation + +Sun Jul 09 05:56:55 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * TAO version 1.5.2 released. + +Tue Jul 3 07:51:12 UTC 2006 Martin Corino <mcorino@remedy.nl> + + * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Endpoint.h: + Fixed missing 'const' in argument declaration. + +Mon Jul 3 15:05:55 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * TAO_IDL/be/be_typedef.cpp: + * TAO_IDL/be/be_visitor_arg_traits.cpp: + * TAO_IDL/be_include/be_typedef.h: + + Changed logic of how the seen_in_operation_ flag is handled + and how the arg traits visitor works for typedefs. There + was an interaction between a typedef valuetype member + and the use of the base type in an interface operation. + This fix closes [BUGID:2577]. Thanks to Heesuk Shin + <heesuks@econz.com> for reporting the bug. + + * tests/IDL_Test/Bug_2577_Regression.idl: + + Removed the preprocessor condition around the line that + reproduces the bug, and added a comment. + +Mon Jul 3 14:29:07 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tests/OBV/Any/client.cpp: + Guard against running the test using Any::to_value() with a + streamed any when compiled with optimized valuetype marshaling + as this test case still fails. See the verbose comment in the + source for details. + +Mon Jul 3 14:17:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/EndpointPolicy/EndpointPolicyTypeC.h: + Fixed versioning macros + +Mon Jul 3 14:07:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * TAO_IDL/be/be_visitor_valuebox/valuebox_ci.cpp: + Fixed versioning macros + +Mon Jul 3 13:06:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * TAO_IDL/ast/ast_decl.cpp: + Const improvements + + * TAO_IDL/fe/fe_init.cpp: + * TAO_IDL/util/utl_stack.cpp: + Initialise pointer with 0 + +Mon Jul 3 11:41:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/PI/ORBInitInfoA.h: + Added this file, needed because other files can include the + pidl file and this file isn't generated by the idl compiler + because we don't use -GA when regenerating the pidl file + +Mon Jul 3 11:15:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Sequence_Unit_Tests/Sequence_Unit_Tests.mpc: + Temporarily disabled the array sequence ut, the mock_array + isn't updated after all array refactoring and this needs + more time then we have for the upcoming beta + +Mon Jul 3 10:59:59 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tests/OBV/Any/OBV_Any.mpc: + * tests/OBV/Any/client.cpp: + + * tests/OBV/Truncatable/OBV_Truncatable.mpc: + * tests/OBV/Truncatable/TruncatableS_impl.cpp: + * tests/OBV/Truncatable/client.cpp: + + Bug 2576 turned out to be invalid. There is already a mechanism + present to address the problem I raised. I've updated the tests + to work with the correct value extraction methods. + +Mon Jul 3 12:00:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/IDL_Test/Bug_2577_Regression.idl: + * tests/IDL_Test/IDL_Test.mpc: + Added regression for bug 2577, this is not fixed yet, but I had + this in my workspace, so commit the test for others. The + file should be changed to reproduce the bug. Thanks to + Heesuk Shin <heesuks at econz dot com> for delivering this test + case. + +Mon Jul 3 10:21:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/PI/ORBInitializer.pidl: + Added workaround for bug 2569 to get the solaris/icc builds + running again. + + * tests/IDL_Test/typeprefix.idl: + Comment part of this test case to get the builds running again + +Mon Jul 3 09:27:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * TAO_IDL/be/be_visitor_exception/exception_cs.cpp: + * TAO_IDL/be/be_visitor_typecode/alias_typecode.cpp + * TAO_IDL/be/be_visitor_typecode/enum_typecode.cpp + * TAO_IDL/be/be_visitor_typecode/objref_typecode.cpp + * TAO_IDL/be/be_visitor_typecode/struct_typecode.cpp + * TAO_IDL/be/be_visitor_typecode/union_typecode.cpp + * TAO_IDL/be/be_visitor_typecode/value_typecode.cpp + Fixed versioning macros generation + +Mon Jul 3 00:54:37 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/AnyTypeCode/skip.cpp: + * tao/Valuetype/Valuetype_Adapter_Impl.h: + * tao/Valuetype/Valuetype_Adapter_Impl.cpp: + * tao/Valuetype_Adapter.h: + Reworked the new interfaces added to the valuetype adapter to be + a little more logical, and added one for detecting chunking. + Skipping the chunks turns out to be really easy, the chunk start + is the length in bytes to skip. Skip that length, then get the + chunk end tag, which is the negated length. If the chunk start + value is negative, that indicates the end of a nesting level, so + this process is repeated until a nesting level of -1 is reached, + indicating the end of the value. + + This change enables truncatable values to be encapsulated in + anys. + + * tao/Valuetype/ValueBase.cpp: + Reformatted a long comment + + * tests/OBV/Any/Any.idl: + * tests/OBV/Any/AnyS_impl.h: + * tests/OBV/Any/AnyS_impl.cpp: + * tests/OBV/Any/OBV_Any.mpc: + * tests/OBV/Any/client.cpp: + * tests/OBV/Truncatable/Extra.idl: + * tests/OBV/Truncatable/OBV_Truncatable.mpc: + * tests/OBV/Truncatable/Truncatable.idl: + * tests/OBV/Truncatable/TruncatableS_impl.h: + * tests/OBV/Truncatable/TruncatableS_impl.cpp: + * tests/OBV/Truncatable/client.cpp: + + Adding the change for encapsulating truncatables gave rise to a + new bug, 2576. This bug illustrates that TAO does not properly + extract a valuetype to a base value pointer, even though the + base value type is carried as part of the Any's typeID + information. So as not to increase the number of failures on the + scoreboard at this time the test code illustrating this bug is + guarded by a #if defined (TAO_TEST_BUG_2576). + +Sat Jul 1 19:16:11 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * NEWS: + Documented the TAO_HAS_OPTIMIZED_VALUETYPE_MARSHALING switch. + +Sat Jul 1 15:32:43 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/AnyTypeCode/skip.cpp: + * tao/Valuetype/Valuetype_Adapter_Impl.h: + * tao/Valuetype/Valuetype_Adapter_Impl.cpp: + * tao/Valuetype_Adapter.h: + + Added an exclusion for Valuetypes marshaled with implied types. + Implicitly typed values are those for which the actual valuetype + matches the formal valuetype. For values inserted into an Any + this is always true as the Any separately marshals the typecode + making it redundant to do this again by the value itself. + + TAO only marshals values this way when compiled to use optimized + valuetype marshaling which is not the default case since it + breaks backwards compatibility. + + This resolves bug 2575. + + * tao/TAO_Server_Request.h: + + Fixed a typo in a comment. + +Sat Jul 1 12:16:20 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/Valuetype/ValueBase.cpp: + Fixed spelling of TAO_HAS_OPTIMIZED_VALUETYPE_MARSHALING + +Fri Jun 30 13:14:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Profile.cpp (is_equivalent): + Use true/false for bool + +Fri Jun 30 10:36:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * TAO_IDL/be/be_visitor_module/module_ch.cpp: + Removed some versioning macro calls + +Thu Jun 29 14:26:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/tests/Notify/performance-tests/RedGreen/RedGreen_Test.{h,cpp}: + * orbsvcs/tests/Notify/performance-tests/RedGreen/main.cpp: + Moved destroy_ec out of the destructor to a seperate method + +Thu Jun 29 10:35:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * TAO_IDL/include/utl_string.h: + * TAO_IDL/util/utl_string.cpp: + Changed compare_quiet to return a real bool. + + * TAO_IDL/util/utl_identifier.cpp: + Const improvement + +Thu Jun 29 09:54:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Bounded_Array_Sequence_T.h: + * tao/Bounded_Sequence_CDR_T.h: + * tao/Unbounded_Array_Sequence_T.h: + * tao/Unbounded_Sequence_CDR_T.h: + Moved (de)marshal_sequence methods from CDR file to the Array + file. + + * TAO_IDL/be/be_codegen.cpp: + * TAO_IDL/be_include/be_codegen.h: + Slightly reworked the way the any includes are generated. This + way the versioning macros are generated better in the A.cpp file. + +Thu Jun 29 08:53:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * rules.tao.GNU: + Removed check for templates=explicit + +Thu Jun 29 01:45:29 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * tao/Makefile.am: + + Regenerated. + +Wed Jun 28 15:23:37 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * docs/libraries.html: + Added text describing the new TAO_EndpointPolicy library. + +Wed Jun 28 13:34:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/IDL_Test/typeprefix.idl: + New file with small example of IDL construct with typeprefix + that fails to compile when TAO_IDL is build with Intel C++ + and Sun C++ at this moment. Changing the forward declared + interface B to a real interface makes it compiling, no idea + what happens and why this only fails with the mentioned + compilers + + * tests/IDL_Test/IDL_Test.mpc: + Added new file + +Wed Jun 28 12:45:00 UTC 2006 Simon Massey <sma@prismtech.com> + + * utils/nslist/nslist.cpp: + Fix overzellous VuWorks renaming of variables from last fix. + +Wed Jun 28 11:02:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/LoadBalancer/LoadManager.cpp: + When we can't register the signal handler, use ACE_ERROR_RETURN + so that it is visible from the test output why things fail. + +Wed Jun 28 09:55:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Valuetype/AbstractBase.cpp: + Use false instead of 0 for a bool + +Wed Jun 28 09:51:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Profile.h: + Removed commented out method + + * tao/Profile.cpp: + Compilation fix for CORBA/e micro + +Wed Jun 28 09:00:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/tao.mpc: + Added base projects for corba/e and minimum corba + +Wed Jun 28 08:00:00 UTC 2006 Simon Massey <sma@prismtech.com> + + * utils/nslist/nsadd.cpp: + * utils/nslist/nslist.cpp: + Fix minor compiler warnings. Anoying cut and paste + ; subsitution for a , in nsadd.cpp. + +Wed Jun 28 06:42:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * utils/catior/catior.mpc: + Removed not needed libpaths and added messaging as base project + so that we are sure we build this exe after messaging + +Tue Jun 27 19:30:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/PortableServer/POAManager.pidl: + Added missing checks for minimum poa + +Tue Jun 27 19:24:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * examples/Content_Server/SMI_Iterator/Content_Iterator_i.{h,cpp}: + * examples/Content_Server/SMI_Iterator/Iterator_Factory_i.cpp: + * examples/Content_Server/SMI_Iterator/Web_Server.idl: + * examples/Content_Server/AMI_Iterator/Content_Iterator_i.{h,cpp}: + * examples/Content_Server/AMI_Iterator/Web_Server.idl: + Made the filesize a CORBA::ULongLong so that we could handle files + bigger then 2Gb, this also resolves some warnings we have on + windows + +Tue Jun 27 19:14:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/PI_Server.mpc: + * tao/PI_Server/ServerRequestInterceptor{A,C,S}.{h,cpp}: + Readded these generated files and removed the regeneration of these + files from the core libs. The generated files had a diff that I + missed this morning, need to resolve that handcrafting first + +Tue Jun 27 19:06:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/AnyTypeCode/WCharSeqA.cpp: + Removed some versioning macros that don't seem correct, if this + solves the errors in the versioning builds I have to change the + IDL compiler for this. + +Tue Jun 27 18:59:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/XtResource.mpc: + And the last mpc file to get an empty IDL_Files section. Thanks to + Chad Elliot for noticing this last missing file + +Tue Jun 27 18:50:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Domain/DomainS.cpp: + Fixed compile error + +Tue Jun 27 18:38:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/FlResource.mpc: + * tao/QtResource.mpc: + * tao/TkResource.mpc: + Added empty IDL_Files section to prevent orb.idl to appear in these + projects when building for corba/e or minimum corba. Thanks to Chad + Elliot for bringing these files to my attention. + +Tue Jun 27 17:35:02 UTC 2006 Chad Elliott <elliott_c@ociweb.com> + + * tao/CSD_Framework.mpc: + * tao/EndpointPolicy.mpc: + * tao/ImR_Client.mpc: + * tao/Messaging.mpc: + * tao/PI.mpc: + * tao/PI_Server.mpc: + * tao/PortableServer.mpc: + + Reverted the work-around for MPC where generated A.cpp files + needed to be added. The bug in MPC has been fixed. + +Tue Jun 27 14:48:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/PortableServer/Fixed_Array_SArgument_T.{h,cpp,inl}: + Fixed crashing of multdim array part of the param test + +Tue Jun 27 14:40:00 UTC 2006 Simon Massey <sma@prismtech.com> + + * utils/nslist/nsadd.cpp: + Fix a resource leak. + +Tue Jun 27 12:07:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * TAO-INSTALL.html: + Updated supported platforms + +Tue Jun 27 12:03:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/CSD_Framework.mpc: + Enabled recompilation for the pidl files of CSD_Framework + + * tao/CSD_Framework/CSD_Framework.pidl: + Updated instructions, removed comments about the patch, not needed + + * tao/CSD_Framework/CSD_Framework*.*: + Removed from the repo + + * tao/CSD_Framework/diffs/CSD_Framework.diff: + Removed, not needed anymore + +Tue Jun 27 11:50:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/PI_Server.mpc: + Enabled recompilation for the pidl files of PI_Server + + * tao/PI_Server/ServerRequestInterceptor*.*: + * tao/PI_Server/ServerRequestInfo*.*: + Removed from the repo + +Tue Jun 27 10:46:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/PI.mpc: + Enabled recompilation for the pidl files of PI + + * tao/PI/ClientRequestInfo*.*: + * tao/PI/ClientRequestInterceptor*.*: + * tao/PI/Interceptor*.*: + * tao/PI/InvalidSlot*.*: + * tao/PI/ORBInitializer*.*: + * tao/PI/ORBInitInfo*.*: + * tao/PI/PICurrent*.*: + * tao/PI/PIForwardRequest*.*: + * tao/PI/ProcessingModePolicy*.*: + * tao/PI/RequestInfo*.*: + Removed from the repo + +Tue Jun 27 10:25:00 UTC 2006 Simon Massey <sma@prismtech.com> + + * NEWS: + * utils/nslist/nsadd.cpp: + * utils/nslist/nsdel.cpp: + * utils/nslist/nslist.cpp: + Enhanced the nslist utilities. + +Tue Jun 27 10:24:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Messaging.mpc: + Enabled recompilation for the pidl files of Messaging + + * tao/Messaging/TAO_ExtC.*: + * tao/Messaging/TAO_ExtA.*: + * tao/Messaging/PollableC.*: + * tao/Messaging/PollableA.*: + * tao/Messaging/Messaging_SyncScope_PolicyC.*: + * tao/Messaging/Messaging_SyncScope_PolicyA.*: + * tao/Messaging/Messaging_RT_PolicyC.*: + * tao/Messaging/Messaging_RT_PolicyA.*: + * tao/Messaging/ExceptionHolderC.*: + * tao/Messaging/ExceptionHolderA.*: + * tao/Messaging/ExceptionHolderS.*: + Removed from the repo + +Tue Jun 27 09:52:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/ImR_Client.mpc: + Enabled recompilation for the pidl files of ImR_Client + + * tao/ImR_Client/ImplRepoS.*: + * tao/ImR_Client/ImplRepoA.cpp: + * tao/ImR_Client/ImplRepoC.*: + * tao/ImR_Client/ServerObjectA.*: + * tao/ImR_Client/ServerObjectC.*: + * tao/ImR_Client/ServerObjectS.*: + Removed from the repo, do get regenerated as part of the building + of the core libs + + * tao/ImR_Client/ImplRepoA.h: + Dummy file that just includes ImplRepoC.h, using -GA for this file + doesn't work + +Tue Jun 27 09:22:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/BiDirectional/run_test.pl: + Use wait_interval_for_process_creation instead of hardcoded value + + * tests/Hello/server.cpp: + Added newline after debug message + + * tests/Stack_Recursion/server.cpp: + Initialise pointer with 0 + + * tests/Stack_Recursion/Sender.{h,cpp}: + Changed is_done to a real bool + +Tue Jun 27 08:50:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/PortableServer/PolicyS.cpp: + Forgot to patch this file + +Tue Jun 27 06:35:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/*.mpc: + Removed tao_versioning_idl_defaults as explicit base, this is coming + from taolib. Added empty IDL_Files section to all files that didn't + had one to prevent the compilation of orb.idl in the corba/e builds + + * tao/IFR_Client/IFR_BaseC.* + * tao/IFR_Client/IFR_BasicC.* + * tao/IFR_Client/IFR_ComponentsC.* + * tao/IFR_Client/IFR_ExtendedC.* + Removed these files from the repo + + * tao/IFR_Client.mpc: + Regenerate the pidl files as part of the build + + * tao/Basic_Arguments.h: + * tao/Basic_Argument_T.{h,cpp.inl}: + * tao/BD_String_Argument_T.{h,cpp,inl}: + * tao/Fixed_Array_Argument_T.{h,cpp,inl}: + * tao/Fixed_Size_Argument_T.{h,cpp,inl}: + * tao/Object_Argument_T.{h,cpp,inl}: + * tao/Object.h: + * tao/Special_Basic_Argument_T.{h,cpp,inl}: + * tao/Special_Basic_Arguments.h: + * tao/UB_String_Argument_T.{h,cpp,inl}: + * tao/UB_String_Argument.h: + * tao/Var_Array_Argument_T.{h,cpp,inl}: + * tao/Var_Size_Argument_T.{h,cpp,inl}: + * tao/AnyTypeCode/Any_Arg_Traits.h: + * tao/AnyTypeCode/TypeCode.h: + * tao/IFR_Client/IFR_Client_Adapter_Impl.cpp: + * tao/PortableServer/Any_SArg_Traits.h: + * tao/PortableServer/Basic_SArgument_T.{h,cpp,inl}: + * tao/PortableServer/Basic_SArguments.h: + * tao/PortableServer/BD_String_SArgument_T.{h,cpp,inl}: + * tao/PortableServer/Fixed_Array_SArgument_T.{h,cpp,inl}: + * tao/PortableServer/Fixed_Size_SArgument_T.{h,cpp,inl}: + * tao/PortableServer/Object_SArgument_T.{h,cpp,inl}: + * tao/PortableServer/Object_SArg_Traits.h: + * tao/PortableServer/Var_Size_SArgument_T.{h,cpp,inl}: + * tao/Valuetype/ValueBase.h: + * tao/Valuetype/AbstractBase.h: + * tao/PortableServer/Special_Basic_SArgument_T.{h,cpp,inl}: + * tao/PortableServer/Special_Basic_SArgument.h: + * tao/PortableServer/TypeCode_SArg_Traits.h: + * tao/PortableServer/Var_Array_SArgument_T.{h,cpp,inl}: + * TAO_IDL/be/be_visitor_arg_traits.cpp: + Removed the introduction of template template arguments for any + insertion strategy. The HP aCC3 compiler doesn't support this at + all so it broke that support fully. + +Mon Jun 26 13:27:00 UTC 2006 Simon Massey <sma@prismtech.com> + + * NEWS: + Added PICurrent refactoring note. + +Sat Jun 24 22:13:38 UTC 2006 Douglas C. Schmidt <schmidt@dre.vanderbilt.edu> + + * docs/tutorials/Quoter/Simple/Impl-Repo/index.html: Enhanced this + file so it explains how to run the Implementation Repository. + Thanks to Ben Creech <bpcreech at alumni dot ncsu dot edu> for + contributing the improvement. + +Mon Jun 26 09:15:00 UTC 2006 Simon Massey <sma@prismtech.com> + + * tao/ServerRequestInterceptor_Adapter.h: + * tao/TAO_Server_Request.cpp: + * tao/TAO_Server_Request.h: + * tao/CSD_Framework/CSD_FW_Server_Request_Wrapper.cpp: + * tao/PI/ClientRequestInfo.cpp: + * tao/PI/ClientRequestInfo.h: + * tao/PI/PICurrent.h: + * tao/PI/PICurrent_Copy_Callback.cpp: (Deleted) + * tao/PI/PICurrent_Copy_Callback.h: (Deleted) + * tao/PI/PICurrent_Impl.cpp: + * tao/PI/PICurrent_Impl.h: + * tao/PI/PICurrent_Impl.inl: + * tao/PI_Server/PICurrent_Guard.cpp: + * tao/PI_Server/PICurrent_Guard.h: + * tao/PI_Server/ServerInterceptorAdapter.cpp: + * tao/PI_Server/ServerInterceptorAdapter.h: + + This refactors the PICurrent implimentaion and fixes issue 2552 + (TAO358). + +Mon Jun 26 08:17:12 UTC 2006 Martin Corino <mcorino@remedy.nl> + + * tao/RTScheduling/RTScheduler_includeA.h: + * tao/RTScheduling/RTScheduler_includeC.h: + Changed include for RTSheduler,h to make it compatible for OpenVMS. + + * tests/Bug_1020_Regression/server.cpp: + * tests/Bug_1361_Regression/shutdown.cpp: + * tests/Nested_Upcall_Crash/scavenger.cpp: + * tests/Object_Loader/Loader.cpp: + * tests/Smart_Proxies/On_Demand/Smart_Test_Factory.cpp: + Small OpenVMS compatibility changes. + +Fri Jun 23 19:03:21 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Valuetype.mpc: + * tao/EndpointPolicy.mpc: + * tao/PortableServer.mpc: + Explicitly list the generated A.cpp in the source section because + MPC doesn't add them at this moment + +Fri Jun 23 18:39:21 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Valuetype.mpc: + * tao/EndpointPolicy.mpc: + Don't substract -Sa from the idlflags, it can be used for the + pidl files + + * tao/PortableServer.mpc: + Handle ForwardRequest.pidl differently because of special + idlflags + + * orbsvcs/IFR_Service/IFR_Service.mpc: + Fixed error in this file + +Fri Jun 23 13:42:54 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * orbsvcs/IFR_Service/IFR_Service.mpc: + + Added $(TAO_ROOT) to the 'includes' line of the IFR loader + executable project, because of a change in the way + TAO/Version.h is included. + +Fri Jun 23 07:36:21 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/PortableServer/Fixed_Array_SArgument_T.h: + Fixed compile problem + +Thu Jun 22 15:27:04 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tests/ORB_Local_Config/Bug_1459/a.conf: + Fix the test by making it use UIOP rather than SSLIOP as the + alternative protocol. The original problem results from first + using SSLIOP in the non-default ORB which loads the SSL library, + which in turn registers with the ACE Object Manager. At shutdown + time, the SSL library is unloaded in order, but then the ACE + Object Manager tries to clean it up, but can't, because it is + already gone. + +Thu Jun 22 16:13:29 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * tao/PortableServer/Fixed_Array_SArgument_T.cpp: + * tao/PortableServer/Fixed_Array_SArgument_T.h: + * tao/PortableServer/Fixed_Array_SArgument_T.inl: + + Changed the member type of all classes except the out arg + class from array type to forany type. Almost every operation + in these classes was first converting the member to forany + type first, and there is no forany constructor directly from + array type. This new design is similar to the stub argument + classes for fixed arrays in TAO. + +Thu Jun 22 15:01:21 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Credentials.h: + + Added missing SSLIOP namespace qualifier to typedef inside + the SSLIOP_Credentials class. + +Thu Jun 22 14:39:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/PortableServer/BD_String_SArgument_T.cpp: + Fixed compile problems in Debian Core build + +Thu Jun 22 07:14:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/PortableServer/Var_Array_SArgument_T.cpp: + Fixed compile problem + +Thu Jun 22 06:59:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/AnyTypeCode/Any.{h,cpp,inl}: + Moved include of Object.h to the header file to resolve compile + problems with Borland in release mode when inlining is enabled + +Thu Jun 22 06:47:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/BD_String_Argument_T.cpp: + Fixed bug in this file + +Wed Jun 21 20:46:09 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * tao/Array_VarOut_T.h: + + Added a typedef of FORANY in the TAO_FixedArray_Var_T<> and + TAO_VarArray_Var_T<> classes that just refer to the existing + typedef for FORANY in the common base class. Some compilers, + when parsing template code, will not look into a base class + unless they are forced to, so the base class typedef was not + seen when referenced in a derived class. Thanks to Will Otte + <wotte@dre.vanderbilt.edu> for figuring this out. + +Wed Jun 21 14:10:38 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * tao/Array_VarOut_T.h: + + Removed 'typename' qualifier in typedef of FORANY inside the + TAO_Array_Var_Base_T template class, since it was causing + problems on the FC4_CIAO_Core build. + +Wed Jun 21 13:50:57 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Credentials.h: + * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Current.h: + * orbsvcs/orbsvcs/SSLIOP/SSLIOP_OwnCredentials.h: + * orbsvcs/orbsvcs/Security/SL3_CredentialsCurator.h: + + Added typedefs _ptr_type, _var_type, and _out_type to + these "interface" classes, since the typedefs are now + required by the *_var and *_out template classes. + +Wed Jun 21 12:13:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * TAO_IDL/be/be_sequence.cpp: + * TAO_IDL/be/be_visitor_traits.cpp: + * TAO_IDL/be/be_visitor_array/array_ch.cpp: + * TAO_IDL/be/be_visitor_array/array_ci.cpp: + * TAO_IDL/be/be_visitor_typedef/typedef_ch.cpp: + * tao/Array_Traits_T.h + * tao/Array_VarOut_T.cpp + * tao/Array_VarOut_T.h + * tao/Array_VarOut_T.inl + * tao/BD_String_Argument_T.cpp + * tao/Bounded_Array_Sequence_T.h + * tao/Bounded_Sequence_CDR_T.h + * tao/Fixed_Array_Argument_T.cpp + * tao/Unbounded_Array_Sequence_T.h + * tao/Unbounded_Sequence_CDR_T.h + * tao/Var_Array_Argument_T.cpp + Reworked array support, remove the traits struct I added and use the + forany class in most templates, this is how Jeff refactoring things + earlier this week, my approach just gave to much problems. + +Wed Jun 21 07:23:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/BD_String_Argument_T.cpp: + Fixed compile problems by added a seperate typedef and use + this in the methods. + +Wed Jun 21 07:15:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/EndpointPolicy.mpc: + * tao/PortableServer.mpc: + * tao/TypeCodeFactory.mpc: + * tao/Valuetype.mpc: + Removed the <%quote%>, it did break on several platforms, just + don't add the quotes in the mpc file, the IDL compiler will + add quotes around the include string + +Wed Jun 21 02:02:59 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * TAO_IDL/Makefile.am: + + Regenerated. + + * TAO_IDL/tao_idl.mpc: + + Changed automake specific include "$(srcdir)/.." to $(TAO_ROOT). + This is enough to trigger the MPC heuristic that this is a TAO + rather than ACE project. No more hand edits! + +Tue Jun 20 20:58:35 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * tao/Makefile.am: + + Regenerated with updated MPC automake.mpd. + +Tue Jun 20 18:17:04 UTC 2006 Martin Corino <mcorino@remedy.nl> + + * TAO_IDL/tao_idl.cpp: + * TAO_IDL/tao_idl.mpc: + Changed include directive for "../tao/Version.h" to "tao/Version.h" + and added ".." and "$(srcdir)/.." to the list of include dirs in + tao_idl.mpc. + OpenVMS doesn't grok the "../tao/Version.h" when toa_idl.cpp is + compiled for the IFR project. + + * TAO_IDL/driver/drv_preproc.cpp: + * TAO_IDL/util/utl_global.cpp: + * TAO_IDL/be/be_interface.cpp: + Changes to circumvent OpenVMS differences regarding subprocesses, + redirection and file naming. + + * TAO_IDL/be/be_helper.cpp: + Disabled ACE_OS::fflush () calls for OpenVMS as these slow down + things dramatically. + +Tue Jun 20 18:12:04 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Credentials.h: + * orbsvcs/orbsvcs/SSLIOP/SSLIOP_OwnCredentials.h: + Fixed compile errors + +Tue Jun 20 17:21:42 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * tao/EndpointPolicy.mpc: + * tao/PortableServer.mpc: + * tao/TypeCodeFactory.mpc: + * tao/Valuetype.mpc: + + Replaced double quotes in these files with <%quote%>. Thanks + to Chad Elliott <elliott_c@ociweb.com> for providing + the fix that allowed the generated *.vcproj files to + load on MSVC 7.1. + +Tue Jun 20 15:28:18 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * TAO_IDL/Makefile.am: + * tao/Makefile.am: + + Regenerated. + + * tao/TypeCodeFactory.mpc: + + Remove TypeCodeFactory subdirectory from PIDL_Files section, + as it's now listed in IDL_Files. + +Tue Jun 20 13:43:05 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * orbsvcs/orbsvcs/Log/Hash_LogRecordStore.cpp: + + Remove setting of servant retention policy when creating the new + POA because this policy is not available with Minimum CORBA. We + were just being pedantic and setting it to the default value + (RETAIN) anyway. + +Tue Jun 20 13:35:04 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/orbsvcs/HTIOP/htiop_endpointsC.h: + Updated for all changes + +Tue Jun 20 12:41:04 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/PortableServer/BD_String_SArgument_T.cpp: + Fixed compilation problems with Borland + +Tue Jun 20 12:29:04 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * docs/Options.html: + Document ORBOutputCDRAllocator and don't refer to cvs.doc.wustl.edu + for files, just get them locally + + * docs/compiler.html: + Document -Sorb + +Tue Jun 20 10:21:04 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/orbsvcs/Security/SL3_CredentialsCurator.h: + Fixed compile error + +Tue Jun 20 08:12:04 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Any_Insert_Policy_T.h: + Reworded the policies to just contain a static method so that we + don't need to create an instance + + * tao/Argument.h: + When we don't have interceptors, just don't declare the In/Out/InOut + classes, just typedef them to Argument + + * tao/Basic_Arguments.h: + * tao/Basic_Argument_T.{h,cpp.inl}: + * tao/BD_String_Argument_T.{h,cpp,inl}: + * tao/Fixed_Array_Argument_T.{h,cpp,inl}: + * tao/Fixed_Size_Argument_T.{h,cpp,inl}: + * tao/Object_Argument_T.{h,cpp,inl}: + * tao/Object.h: + * tao/Special_Basic_Argument_T.{h,cpp,inl}: + * tao/Special_Basic_Arguments.h: + * tao/UB_String_Argument_T.{h,cpp,inl}: + * tao/UB_String_Argument.h: + * tao/Var_Array_Argument_T.{h,cpp,inl}: + * tao/Var_Size_Argument_T.{h,cpp,inl}: + * tao/AnyTypeCode/Any_Arg_Traits.h: + * tao/AnyTypeCode/TypeCode.h: + * tao/IFR_Client/IFR_Client_Adapter_Impl.cpp: + * tao/PortableServer/Any_SArg_Traits.h: + * tao/PortableServer/Basic_SArgument_T.{h,cpp,inl}: + * tao/PortableServer/Basic_SArguments.h: + * tao/PortableServer/BD_String_SArgument_T.{h,cpp,inl}: + * tao/PortableServer/Fixed_Array_SArgument_T.{h,cpp,inl}: + * tao/PortableServer/Fixed_Size_SArgument_T.{h,cpp,inl}: + * tao/PortableServer/Object_SArgument_T.{h,cpp,inl}: + * tao/PortableServer/Object_SArg_Traits.h: + * tao/PortableServer/Var_Size_SArgument_T.{h,cpp,inl}: + * tao/Valuetype/ValueBase.h: + * tao/Valuetype/AbstractBase.h: + * tao/PortableServer/Special_Basic_SArgument_T.{h,cpp,inl}: + * tao/PortableServer/Special_Basic_SArgument.h: + * tao/PortableServer/TypeCode_SArg_Traits.h: + * tao/PortableServer/Var_Array_SArgument_T.{h,cpp,inl}: + * TAO_IDL/be/be_visitor_arg_traits.cpp: + The any insert policy has been changed to a template template + argument. + + * tao/PortableServer/PolicyS.cpp: + * tao/WCharSeqS.h: + * tao/WCharSeqC.h: + * tao/CONV_FRAMEC.h: + * tao/DomainC.cpp: + * tao/IIOP_EndpointsC.h: + * tao/IIOPC.h: + * tao/Messaging_PolicyValueC.h: + * tao/IOP_IORC.h: + * tao/OctetSeqC.h: + * tao/WCharSeqC.cpp: + * tao/Policy_ForwardC.h: + * tao/PolicyC.cpp: + * tao/StringSeqC.h: + * tao/ServicesC.h: + * tao/GIOPC.h: + * tao/AnyTypeCode/Dynamic_ParameterC.h: + * tao/AnyTypeCode/WCharSeqA.{h,cpp}: + * tao/Domain/DomainS.cpp: + * tao/DynamicAny/DynamicAnyC.h: + * tao/IFR_Client/IFR_ExtendedC.{h,cpp}: + * tao/IFR_Client/IFR_ComponentsC.{h,cpp}: + * tao/IFR_Client/IFR_BasicC.{h,cpp}: + * tao/IFR_Client/IFR_BaseC.{h,cpp}: + * tao/ImR_Client/ImplRepoC.{h,cpp}: + * tao/ImR_Client/ImplRepoS.cpp: + * tao/ImR_Client/ServerObjectS.cpp: + * tao/IORInterceptor/IORInfoC.cpp: + * tao/ObjRefTemplate/ObjectReferenceTemplateC.cpp: + * tao/RTCORBA/RTCORBAC.h: + * tao/RTScheduling/RTSchedulerC.cpp: + * tao/Strategies/sciop_endpointsC.h: + * tao/Strategies/uiop_endpointsC.h: + * tao/PortableServer/PortableServerC.{h,cpp}: + * tao/PortableServer/PortableServerS.h: + Updated for all template changes + + * tao/Sequence_T.h: + Reordered includes + + * tao/Blocked_Connect_Strategy.{h,cpp}: + Removed not needed virtual destructor + + * tao/Connection_Handler.cpp: + * tao/Wait_On_Leader_Follower.cpp: + * tao/Wait_On_Reactor.cpp: + * tao/Wait_On_Read.cpp: + * tao/Wait_Strategy.{h,cpp,inl}: + * tao/Transport_Cache_Manager.cpp: + Changed is_registered flag of the wait stratetegy to a real bool + + * tao/Object_Reference_Sequence_Element_T.h: + * tao/String_Sequence_Element_T.h: + Added in accessor + + * tao/Object.cpp: + * tao/PortableServer/IdAssignmentPolicy.{h,cpp}: + * tao/PortableServer/IdUniquenessPolicy.{h,cpp}: + * tao/PortableServer/ImplicitActivationPolicy.{h,cpp}: + * tao/PortableServer/LifespanPolicy.{h,cpp}: + * tao/PortableServer/POA_Cached_Policies.cpp: + * tao/PortableServer/POA_Policy_Set.cpp: + * tao/PortableServer/Object_Adapter.cpp: + * tao/PortableServer/POAManager.{h,cpp}: + * tao/PortableServer/POAManagerFactory.{h,cpp}: + * tao/PortableServer/RequestProcessingPolicy.{h,cpp}: + * tao/ORB.{h,cpp}: + * tao/PortableServer/PortableServer.cpp: + * tao/PortableServer/RequestProcessingStrategyDefaultServant.{h,cpp}: + * tao/PortableServer/RequestProcessingStrategyDefaultServantFI.{h,cpp}: + * tao/PortableServer/RequestProcessingStrategyServantActivator.{h,cpp}: + * tao/PortableServer/RequestProcessingStrategyServantActivatorFI.{h,cpp}: + * tao/PortableServer/RequestProcessingStrategyServantLocator.{h,cpp}: + * tao/PortableServer/RequestProcessingStrategyServantLocatorFI.{h,cpp}: + * tao/PortableServer/RequestProcessingStrategyServantManager.{h,cpp}: + * tao/PortableServer/Servant_Upcall.cpp: + * tao/PortableServer/Root_POA.{h,cpp}: + * tao/PortableServer/ThreadPolicy.{h,cpp}: + * tao/PortableServer/ServantRetentionStrategyRetain.{h,cpp}: + * tao/PortableServer/ServantRetentionStrategyNonRetain.{h,cpp}: + * tao/PortableServer/ServantRetentionStrategy.h: + * tao/PortableServer/ServantRetentionPolicy.{h,cpp}: + Added support for CORBA/e + + * tao/PortableServer/IdAssignmentPolicy.pidl: + * tao/PortableServer/PS_Current.pidl: + * tao/TypeCodeFactory/TypeCodeFactory.pidl: + * tao/Valuetype/StringValue.pidl: + * tao/PortableServer/RequestProcessingPolicy.pidl: + * tao/PortableServer/ServantActivator.pidl: + * tao/PortableServer/ServantLocator.pidl: + * tao/PortableServer/ServantManager.pidl: + * tao/PortableServer/ServantRetentionPolicy.pidl: + * tao/PortableServer/ThreadPolicy.pidl: + Updated regeneration instructions and support CORBA/e + + * tao/OctetSeq.pidl: + Added -Sorb to the regeneration instructions + + * tao/Policy_Forward.pidl: + Updated regeneration instructions, handcrafting is not needed anymore + + * tao/diffs/Policy_Forward.diff: + Removed from the repo + + * tao/VarOut_T.{h,inl}: + Zapped one template argument + + * tao/IIOP_Transport.{h,cpp}: + * tao/default_resource.{h,cpp}: + * tao/Transport.{h,cpp}: + Added support for sendfile + + * tao/orbconf.h: + Added TAO_USE_OUTPUT_CDR_MMAP_MEMORY_POOL with which sendfile + can be forced during compilation + + * tao/Transport_Cache_Manager.cpp: + Use ACE_TEXT with debug macros + + * tao/Messaging/Asynch_Reply_Dispatcher.cpp: + * tao/RTCORBA/RT_Protocols_Hooks.cpp: + * tao/Valuetype/AbstractBase_T.cpp: + Const improvements + + * tao/PortableServer/get_arg.h: + Check if the details do have args before doing a cast, part of + bug 2503 + + * tao/PortableServer/POA.pidl: + * tao/PortableServer/PortableServer.pidl: + Moved most contents to POA.pidl to make regeneration easier + + * tao/Bounded_Sequence_CDR_T.h: + Moved array code to this file + + * tao/Array_VarOut_T.{h,cpp,inl}: + * tao/Array_Traits_T.h: + * tao/Bounded_Array_Sequence_T.h: + * tao/Unbounded_Array_Sequence_T.h: + * tao/Unbounded_Sequence_CDR_T.h: + Refactored array support, mix of the changes from Jeff from + yesterday and my work. Need to cleanup a little when things + compile again. + + * tao/TypeCodeFactory.mpc: + * tao/Valuetype.mpc: + * tao/PortableServer.mpc: + * tao/ImR_Client.mpc: + * tao/EndpointPolicy.mpc: + Enabled regeneration of pidl files as part of building the core + TAO libs. More pidl files will get regenerated as part of the + build, this is the first step. + + * tao/tao.mpc: + Added new sendfile files + + * tao/RTCORBA/RT_Transport_Descriptor.h: + Use false for a boolean + + * tao/Strategies/FIFO_Connection_Purging_Strategy.{h,cpp}: + * tao/Strategies/LFU_Connection_Purging_Strategy.{h,cpp}: + * tao/Strategies/NULL_Connection_Purging_Strategy.{h,cpp}: + Removed not needed virtual destructors + + * tao/PortableServer/diffs/AdapterActivator.diff + * tao/PortableServer/diffs/ForwardRequest.diff + * tao/PortableServer/diffs/ImplicitActivationPolicy.diff + * tao/PortableServer/diffs/POAManager.diff + * tao/PortableServer/diffs/RequestProcessingPolicy.diff + * tao/PortableServer/diffs/ServantActivator.diff + * tao/PortableServer/diffs/ServantLocator.diff + * tao/PortableServer/diffs/ServantManager.diff + * tao/PortableServer/diffs/ServantRetentionPolicy.diff + * tao/PortableServer/diffs/ThreadPolicy.diff + Removed, not needed anymore + + * TAO_IDL/be/be_codegen.cpp: + * TAO_IDL/be/be_global.cpp: + * TAO_IDL/be_include/be_decl.h: + Made generation of ORB.h optional, can be suppressed using -Sorb + + * TAO_IDL/be/be_decl.cpp: + * TAO_IDL/be/be_interface_fwd.cpp: + * TAO_IDL/be/be_type.cpp: + * TAO_IDL/be/be_typedef.cpp: + * TAO_IDL/be_include/be_decl.h: + * TAO_IDL/be_include/be_interface_fwd.h: + * TAO_IDL/be_include/be_type.h: + * TAO_IDL/be_include/be_typedef.h: + Removed old code related to explicit template instantiations + + * TAO_IDL/be/be_sequence.cpp: + Updated generation of array sequences + + * TAO_IDL/be/be_visitor_array/array_ch.cpp: + * TAO_IDL/be/be_visitor_array/array_ci.cpp: + * TAO_IDL/be/be_visitor_array/array_cs.cpp: + * TAO_IDL/be/be_visitor_array/cdr_op_ch.cpp: + Updated generation for arrays + + * TAO_IDL/be/be_visitor_interface.cpp: + Added support for native keyword, needed to generate typedefs from + the pidl files + + * TAO_IDL/be/be_visitor_enum/any_op_cs.cpp: + * TAO_IDL/be/be_visitor_enum/cdr_op_ch.cpp: + * TAO_IDL/be/be_visitor_enum/cdr_op_cs.cpp: + * TAO_IDL/be/be_visitor_sequence/buffer_type.cpp: + * TAO_IDL/be/be_visitor_sequence/cdr_op_ch.cpp: + * TAO_IDL/be/be_visitor_sequence/sequence_ch.cpp: + Small layout improvements + + * TAO_IDL/be/be_visitor_exception/exception_cs.cpp: + * TAO_IDL/be/be_visitor_interface/interface_cs.cpp: + * TAO_IDL/be/be_visitor_module/module_ch.cpp: + * TAO_IDL/be/be_visitor_typecode/alias_typecode.cpp + * TAO_IDL/be/be_visitor_typecode/enum_typecode.cpp + * TAO_IDL/be/be_visitor_typecode/objref_typecode.cpp + * TAO_IDL/be/be_visitor_typecode/struct_typecode.cpp + * TAO_IDL/be/be_visitor_typecode/union_typecode.cpp + * TAO_IDL/be/be_visitor_typecode/value_typecode.cpp + * TAO_IDL/be/be_visitor_valuebox/any_op_cs.cpp: + * TAO_IDL/be/be_visitor_valuebox/valuebox_ci.cpp: + * TAO_IDL/be/be_visitor_valuebox/valuebox_cs.cpp: + Added missing generation of versioning macros. + + * TAO_IDL/be/be_visitor_interface/interface.cpp: + * TAO_IDL/be_include/be_visitor_interface/interface.h: + Handle native + + * TAO_IDL/be/be_visitor_native/native_ch.cpp: + * TAO_IDL/be_include/be_visitor_native.h: + * TAO_IDL/be_include/be_visitor_native/native_ch.h: + New file for native generation + + * TAO_IDL/be/be_visitor_typedef/any_op_ch.cpp: + * TAO_IDL/be/be_visitor_typedef/any_op_ch.cpp: + * TAO_IDL/be/be_visitor_typedef/serializer_op_ch.cpp: + * TAO_IDL/be/be_visitor_typedef/typedef_ch.cpp: + * TAO_IDL/be/be_visitor_typedef/typedef_ci.cpp: + Updated for array changes + + * TAO_IDL/fe/fe_extern.cpp: + Const improvement + +Mon Jun 19 22:55:04 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * tao/Array_VarOut_T.h: + + Added 'typename' to the typedef of FORANY in class + TAO_Array_Var_Base_T. + +Mon Jun 19 19:39:25 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/EndpointPolicy/EndpointPolicyA.*: + * tao/EndpointPolicy/EndpointPolicyC.*: + * tao/EndpointPolicy/EndpointPolicyTypeA.*: + * tao/EndpointPolicy/EndpointPolicyTypeC.*: + * tao/EndpointPolicy/IIOPEndpointValueA.*: + * tao/EndpointPolicy/IIOPEndpointValueC.*: + * tao/PortableServer/AdapterActivatorC.*: + * tao/PortableServer/AdapterActivatorA.*: + * tao/PortableServer/ForwardRequestC.*: + * tao/PortableServer/ForwardRequestA.*: + * tao/PortableServer/IdAssignmentPolicyC.*: + * tao/PortableServer/IdAssignmentPolicyA.*: + * tao/PortableServer/IdUniquenessPolicyC.*: + * tao/PortableServer/IdUniquenessPolicyA.*: + * tao/PortableServer/ThreadPolicyC.*: + * tao/PortableServer/ThreadPolicyA.*: + * tao/PortableServer/ServantRetentionPolicyC.*: + * tao/PortableServer/ServantRetentionPolicyA.*: + * tao/PortableServer/ServantManagerC.*: + * tao/PortableServer/ServantManagerA.*: + * tao/PortableServer/ServantLocatorC.*: + * tao/PortableServer/ServantLocatorA.*: + * tao/PortableServer/ServantActivatorC.*: + * tao/PortableServer/ServantActivatorA.*: + * tao/PortableServer/RequestProcessingPolicyC.*: + * tao/PortableServer/RequestProcessingPolicyA.*: + * tao/PortableServer/PS_CurrentC.*: + * tao/PortableServer/PortableServerA.cpp: + * tao/PortableServer/POAManagerFactoryC.*: + * tao/PortableServer/POAManagerC.*: + * tao/PortableServer/LifespanPolicyC.*: + * tao/PortableServer/LifespanPolicyA.*: + * tao/PortableServer/ImplicitActivationPolicyC.*: + * tao/PortableServer/ImplicitActivationPolicyA.*: + * tao/TypeCodeFactory/TypeCodeFactoryA.*: + * tao/TypeCodeFactory/TypeCodeFactoryC.*: + * tao/Valuetype/StringValueC.*: + Removed from the repo, will be generated with tao_idl during the build + of the core libs in the future. + + * tao/EndpointPolicy/Endpoint_Acceptor_Filter.cpp: + Const improvement + + * tao/EndpointPolicy/EndpointPolicy.pidl: + * tao/EndpointPolicy/EndpointPolicyType.pidl: + * tao/EndpointPolicy/IIOPEndpointValue.pidl: + Removed regeneration instructions, stored in mpc file in the future + + * tao/PortableServer/AdapterActivator.pidl: + * tao/PortableServer/ForwardRequest.pidl: + * tao/PortableServer/IdUniquenessPolicy.pidl: + * tao/PortableServer/ImplicitActivationPolicy.pidl: + * tao/PortableServer/POAManagerFactory.pidl: + * tao/PortableServer/POAManager.pidl: + * tao/PortableServer/LifespanPolicy.pidl: + Removed regeneration instructions and updated for CORBA/e + + * tao/PortableServer/Creation_Time.h: + Made single argument constructor explicit + + * tao/Connector_Registry.{h,cpp,i}: + * tao/Incoming_Message_Queue.{h,inl}: + Const improvements + + * tao/MMAP_Allocator.{h,cpp}: + New files part of the sendfile support + + * tao/Stub.h: + Doxygen improvements + + * tao/PortableServer/Direct_Collocation_Upcall_Wrapper.cpp: + * tao/PortableServer/Default_Policy_Validator.cpp: + * tao/PortableServer/Adapter_Activator.{h,cpp}: + Added support for CORBA/e + +Mon Jun 19 14:49:25 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * tao/Abstract_Servant_Base.h: + * tao/Adapter.h: + * tao/Adapter_Registry.h: + * tao/AnyTypeCode.mpc + * tao/Array_Traits_T.h: + * tao/Array_VarOut_T.cpp: + * tao/Array_VarOut_T.h: + * tao/Array_VarOut_T.inl: + * tao/BD_String_Argument_T.cpp: + * tao/BD_String_Argument_T.h: + * tao/BD_String_Argument_T.inl: + * tao/Basic_Argument_T.h: + * tao/Basic_Arguments.h: + * tao/BooleanSeqC.h: + * tao/Bounded_Array_Sequence_T.h: + * tao/CharSeqC.h: + * tao/Collocation_Proxy_Broker.h: + * tao/DoubleSeqC.h: + * tao/Environment.h: + * tao/Fixed_Array_Argument_T.cpp: + * tao/Fixed_Array_Argument_T.h: + * tao/Fixed_Array_Argument_T.inl: + * tao/Fixed_Size_Argument_T.cpp: + * tao/Fixed_Size_Argument_T.h: + * tao/Fixed_Size_Argument_T.inl: + * tao/FloatSeqC.h: + * tao/LocalObject.h: + * tao/LongDoubleSeqC.h: + * tao/LongLongSeqC.h: + * tao/LongSeqC.h: + * tao/ORB.h: + * tao/Object.h: + * tao/Object_Argument_T.h: + * tao/OctetSeqC.h: + * tao/PolicyC.h: + * tao/Policy_ForwardC.h: + * tao/Principal.h: + * tao/Pseudo_VarOut_T.h: + * tao/Pseudo_VarOut_T.inl: + * tao/Request_Dispatcher.h: + * tao/ShortSeqC.h: + * tao/StringSeqC.h: + * tao/String_Traits_Base_T.h: + * tao/TypeCodeFactory_Adapter.h: + * tao/UB_String_Argument_T.cpp: + * tao/UB_String_Argument_T.h: + * tao/UB_String_Argument_T.inl: + * tao/UB_String_Arguments.h: + * tao/ULongLongSeqC.h: + * tao/ULongSeqC.h: + * tao/UShortSeqC.h: + * tao/Unbounded_Array_Sequence_T.h: + * tao/Var_Array_Argument_T.cpp: + * tao/Var_Array_Argument_T.h: + * tao/Var_Array_Argument_T.inl: + * tao/Var_Size_Argument_T.cpp: + * tao/Var_Size_Argument_T.h: + * tao/Var_Size_Argument_T.inl: + * tao/WCharSeqC.h: + * tao/WStringSeqC.h: + * tao/AnyTypeCode/Any.h: + * tao/AnyTypeCode/AnySeqC.h: + * tao/AnyTypeCode/Any_Arg_Traits.h: + * tao/AnyTypeCode/NVList.h: + * tao/AnyTypeCode/TypeCode.h: + * tao/Domain/DomainS.cpp: + * tao/DynamicInterface/Context.h: + * tao/DynamicInterface/ExceptionList.h: + * tao/DynamicInterface/Request.h: + * tao/DynamicInterface/Server_Request.h: + * tao/IFR_Client/IFR_BaseC.cpp: + * tao/IFR_Client/IFR_BaseC.h: + * tao/IFR_Client/IFR_BaseC.inl: + * tao/IFR_Client/IFR_BasicC.cpp: + * tao/IFR_Client/IFR_BasicC.h: + * tao/IFR_Client/IFR_BasicC.inl: + * tao/IFR_Client/IFR_ComponentsC.cpp: + * tao/IFR_Client/IFR_ComponentsC.h: + * tao/IFR_Client/IFR_ComponentsC.inl: + * tao/IFR_Client/IFR_ExtendedC.cpp: + * tao/IFR_Client/IFR_ExtendedC.h: + * tao/IFR_Client/IFR_ExtendedC.inl: + * tao/ImR_Client/ImplRepoC.cpp: + * tao/ImR_Client/ImplRepoC.h: + * tao/ImR_Client/ImplRepoS.cpp: + * tao/ImR_Client/ServerObjectS.cpp: + * tao/Messaging/MessagingS.cpp: + * tao/ObjRefTemplate/ObjectReferenceTemplateC.cpp: + * tao/PortableServer/Any_SArg_Traits.h: + * tao/PortableServer/BD_String_SArgument_T.cpp: + * tao/PortableServer/BD_String_SArgument_T.h: + * tao/PortableServer/BD_String_SArgument_T.inl: + * tao/PortableServer/Collocated_Object_Proxy_Broker.cpp: + * tao/PortableServer/Direct_Collocation_Upcall_Wrapper.h: + * tao/PortableServer/Fixed_Array_SArgument_T.cpp: + * tao/PortableServer/Fixed_Array_SArgument_T.h: + * tao/PortableServer/Fixed_Array_SArgument_T.inl: + * tao/PortableServer/Fixed_Size_SArgument_T.h: + * tao/PortableServer/Object_SArg_Traits.h: + * tao/PortableServer/Object_SArgument_T.h: + * tao/PortableServer/PolicyS.cpp: + * tao/PortableServer/Servant_Upcall.h: + * tao/PortableServer/TypeCode_SArg_Traits.h: + * tao/PortableServer/Var_Array_SArgument_T.cpp: + * tao/PortableServer/Var_Array_SArgument_T.h: + * tao/PortableServer/Var_Array_SArgument_T.inl: + * tao/PortableServer/Var_Size_SArgument_T.cpp: + * tao/PortableServer/Var_Size_SArgument_T.h: + * tao/PortableServer/Var_Size_SArgument_T.inl: + * tao/PortableServer/get_arg.h: + * tao/RTScheduling/RTSchedulerC.cpp: + * tao/Valuetype/AbstractBase.h: + * tao/Valuetype/ValueBase.h: + + Reduced the number of template parameters in _var, + _out, _forany and arg template classes. + + * TAO_IDL/be/be_codegen.cpp: + * TAO_IDL/be/be_visitor_arg_traits.cpp: + * TAO_IDL/be/be_visitor_traits.cpp: + * TAO_IDL/be/be_visitor_array/array_ci.cpp: + * TAO_IDL/be/be_visitor_component/component_ch.cpp: + * TAO_IDL/be/be_visitor_interface/amh_ch.cpp: + * TAO_IDL/be/be_visitor_interface/interface_ch.cpp: + * TAO_IDL/be/be_visitor_operation/operation.cpp: + * TAO_IDL/be/be_visitor_operation/upcall_command_ss.cpp: + * TAO_IDL/be/be_visitor_sequence/sequence_ch.cpp: + * TAO_IDL/be/be_visitor_structure/structure_ch.cpp: + * TAO_IDL/be/be_visitor_union/union_ch.cpp: + * TAO_IDL/be/be_visitor_valuebox/valuebox_ch.cpp: + * TAO_IDL/be/be_visitor_valuetype/valuetype_ch.cpp: + + Changes to generated code corresponding with the + changes to middleware code described above. + +Mon Jun 19 12:11:09 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * TAO_IDL/driver/drv_preproc.cpp: + Fixed copy past error which resulted in a non working tao_idl + with BCB + +Sat Jun 17 20:47:09 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * TAO_IDL/tao_idl.cpp: + * TAO_IDL/driver/drv_preproc.cpp: + + Changed the type of the DRV_arglist variable from char*[] + to const char*[], since Borland compilers are strict about + array types and won't implicitly qualify char*[] to be + passed as const char*[]. This change also required + casting away constness when the array elements are + deleted during cleanup. Thanks to Johnny Willemsen + <jwillemsen@rememdy.nl> for his expertise with Borland + compilers. + +Fri Jun 16 20:25:45 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * orbsvcs/IFR_Service/ifr_adding_visitor.cpp: + * orbsvcs/IFR_Service/ifr_adding_visitor_operation.cpp: + + Added CORBA::string_dup() around the identifier string of + an argument, when it is inserted into the sequence of + parameter descriptions that is passed to the IFR to make + an entry for an operation or a factory. Since the + sequence in question manages its own memory, we were + getting double deletions of this identifier string before. + Thanks to Johnny Willemsen <jwillemsen@remedy.nl> for + pointing out the Interface Repository tests failures + causes by this problem. + +Fri Jun 16 14:29:23 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * TAO_IDL/fe/fe_interface_header.cpp: + * TAO_IDL/be/be_visitor_ami_pre_proc.cpp: + + Removed unused variables and identifiers for unused args. + +Thu Jun 15 20:35:41 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * TAO_IDL/be/be_visitor_valuebox/valuebox_ci.cpp: + * TAO_IDL/be/be_visitor_valuebox/valuebox_cs.cpp: + + Partially reverted this mornings fix. The + _tao_obv_static_repository_id function is not virtual and may + safely stay inlined. + +Thu Jun 15 15:05:43 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * TAO_IDL/be_include/be_visitor_interface/remote_proxy_impl_cs.h: + * TAO_IDL/be/be_visitor_interface/remote_proxy_impl_cs.cpp: + + Removed these files from the repository, they are no + longer used. + + * TAO_IDL/tao_idl.cpp: + * TAO_IDL/ast/ast_argument.cpp: + * TAO_IDL/ast/ast_array.cpp: + * TAO_IDL/ast/ast_attribute.cpp: + * TAO_IDL/ast/ast_check.cpp: + * TAO_IDL/ast/ast_concrete_type.cpp: + * TAO_IDL/ast/ast_constant.cpp: + * TAO_IDL/ast/ast_decl.cpp: + * TAO_IDL/ast/ast_enum.cpp: + * TAO_IDL/ast/ast_exception.cpp: + * TAO_IDL/ast/ast_expression.cpp: + * TAO_IDL/ast/ast_factory.cpp: + * TAO_IDL/ast/ast_field.cpp: + * TAO_IDL/ast/ast_generator.cpp: + * TAO_IDL/ast/ast_home.cpp: + * TAO_IDL/ast/ast_interface.cpp: + * TAO_IDL/ast/ast_interface_fwd.cpp: + * TAO_IDL/ast/ast_module.cpp: + * TAO_IDL/ast/ast_native.cpp: + * TAO_IDL/ast/ast_operation.cpp: + * TAO_IDL/ast/ast_predefined_type.cpp: + * TAO_IDL/ast/ast_root.cpp: + * TAO_IDL/ast/ast_sequence.cpp: + * TAO_IDL/ast/ast_string.cpp: + * TAO_IDL/ast/ast_structure.cpp: + * TAO_IDL/ast/ast_structure_fwd.cpp: + * TAO_IDL/ast/ast_type.cpp: + * TAO_IDL/ast/ast_typedef.cpp: + * TAO_IDL/ast/ast_union_branch.cpp: + * TAO_IDL/ast/ast_union_label.cpp: + * TAO_IDL/ast/ast_valuebox.cpp: + * TAO_IDL/ast/ast_valuetype.cpp: + * TAO_IDL/be/be_argument.cpp: + * TAO_IDL/be/be_array.cpp: + * TAO_IDL/be/be_attribute.cpp: + * TAO_IDL/be/be_codegen.cpp: + * TAO_IDL/be/be_constant.cpp: + * TAO_IDL/be/be_decl.cpp: + * TAO_IDL/be/be_enum.cpp: + * TAO_IDL/be/be_exception.cpp: + * TAO_IDL/be/be_factory.cpp: + * TAO_IDL/be/be_field.cpp: + * TAO_IDL/be/be_generator.cpp: + * TAO_IDL/be/be_global.cpp: + * TAO_IDL/be/be_home.cpp: + * TAO_IDL/be/be_init.cpp: + * TAO_IDL/be/be_interface.cpp: + * TAO_IDL/be/be_interface_fwd.cpp: + * TAO_IDL/be/be_interface_strategy.cpp: + * TAO_IDL/be/be_module.cpp: + * TAO_IDL/be/be_native.cpp: + * TAO_IDL/be/be_operation.cpp: + * TAO_IDL/be/be_operation_strategy.cpp: + * TAO_IDL/be/be_predefined_type.cpp: + * TAO_IDL/be/be_produce.cpp: + * TAO_IDL/be/be_root.cpp: + * TAO_IDL/be/be_scope.cpp: + * TAO_IDL/be/be_sequence.cpp: + * TAO_IDL/be/be_string.cpp: + * TAO_IDL/be/be_structure.cpp: + * TAO_IDL/be/be_structure_fwd.cpp: + * TAO_IDL/be/be_type.cpp: + * TAO_IDL/be/be_union.cpp: + * TAO_IDL/be/be_union_branch.cpp: + * TAO_IDL/be/be_union_fwd.cpp: + * TAO_IDL/be/be_union_label.cpp: + * TAO_IDL/be/be_valuetype.cpp: + * TAO_IDL/be/be_visitor_amh_pre_proc.cpp: + * TAO_IDL/be/be_visitor_ami_pre_proc.cpp: + * TAO_IDL/be/be_visitor_arg_traits.cpp: + * TAO_IDL/be/be_visitor_attribute.cpp: + * TAO_IDL/be/be_visitor_ccm_pre_proc.cpp: + * TAO_IDL/be/be_visitor_context.cpp: + * TAO_IDL/be/be_visitor_decl.cpp: + * TAO_IDL/be/be_visitor_interface.cpp: + * TAO_IDL/be/be_visitor_valuetype.cpp: + * TAO_IDL/be/be_visitor_attribute/attribute.cpp: + * TAO_IDL/be/be_visitor_interface/any_op_ch.cpp: + * TAO_IDL/be/be_visitor_interface/any_op_cs.cpp: + * TAO_IDL/be/be_visitor_interface/direct_proxy_impl_sh.cpp: + * TAO_IDL/be/be_visitor_interface/direct_proxy_impl_ss.cpp: + * TAO_IDL/be/be_visitor_interface/interface.cpp: + * TAO_IDL/be/be_visitor_interface/interface_ch.cpp: + * TAO_IDL/be/be_visitor_interface/interface_cs.cpp: + * TAO_IDL/be/be_visitor_interface/interface_sh.cpp: + * TAO_IDL/be/be_visitor_interface/interface_si.cpp: + * TAO_IDL/be/be_visitor_interface/interface_ss.cpp: + * TAO_IDL/be/be_visitor_interface_fwd/any_op_ch.cpp: + * TAO_IDL/be/be_visitor_interface_fwd/cdr_op_ch.cpp: + * TAO_IDL/be/be_visitor_operation/amh_rh_ss.cpp: + * TAO_IDL/be/be_visitor_operation/amh_sh.cpp: + * TAO_IDL/be/be_visitor_operation/amh_ss.cpp: + * TAO_IDL/be/be_visitor_operation/ami_ch.cpp: + * TAO_IDL/be/be_visitor_typecode/enum_typecode.cpp: + * TAO_IDL/be/be_visitor_typecode/struct_typecode.cpp: + * TAO_IDL/be/be_visitor_typecode/union_typecode.cpp: + * TAO_IDL/be/be_visitor_valuetype/valuetype.cpp: + * TAO_IDL/be_include/be_argument.h: + * TAO_IDL/be_include/be_array.h: + * TAO_IDL/be_include/be_attribute.h: + * TAO_IDL/be_include/be_codegen.h: + * TAO_IDL/be_include/be_extern.h: + * TAO_IDL/be_include/be_field.h: + * TAO_IDL/be_include/be_generator.h: + * TAO_IDL/be_include/be_global.h: + * TAO_IDL/be_include/be_interface.h: + * TAO_IDL/be_include/be_interface_strategy.h: + * TAO_IDL/be_include/be_native.h: + * TAO_IDL/be_include/be_operation_strategy.h: + * TAO_IDL/be_include/be_root.h: + * TAO_IDL/be_include/be_union_branch.h: + * TAO_IDL/be_include/be_union_label.h: + * TAO_IDL/be_include/be_valuebox.h: + * TAO_IDL/be_include/be_visitor.h: + * TAO_IDL/be_include/be_visitor_ami_pre_proc.h: + * TAO_IDL/be_include/be_visitor_ccm_pre_proc.h: + * TAO_IDL/be_include/be_visitor_decl.h: + * TAO_IDL/driver/drv_args.cpp: + * TAO_IDL/driver/drv_preproc.cpp: + * TAO_IDL/fe/fe_declarator.cpp: + * TAO_IDL/fe/fe_extern.cpp: + * TAO_IDL/fe/fe_init.cpp: + * TAO_IDL/fe/fe_interface_header.cpp: + * TAO_IDL/fe/idl.ll: + * TAO_IDL/fe/idl.yy: + * TAO_IDL/fe/lex.yy.cpp: + * TAO_IDL/fe/y.tab.cpp: + * TAO_IDL/fe/y.tab.cpp:.diff: + * TAO_IDL/fe/y.tab.h: + * TAO_IDL/include/ast_argument.h: + * TAO_IDL/include/ast_array.h: + * TAO_IDL/include/ast_attribute.h: + * TAO_IDL/include/ast_concrete_type.h: + * TAO_IDL/include/ast_decl.h: + * TAO_IDL/include/ast_exception.h: + * TAO_IDL/include/ast_expression.h: + * TAO_IDL/include/ast_field.h: + * TAO_IDL/include/ast_generator.h: + * TAO_IDL/include/ast_interface.h: + * TAO_IDL/include/ast_interface_fwd.h: + * TAO_IDL/include/ast_native.h: + * TAO_IDL/include/ast_sequence.h: + * TAO_IDL/include/ast_structure.h: + * TAO_IDL/include/ast_structure_fwd.h: + * TAO_IDL/include/ast_typedef.h: + * TAO_IDL/include/ast_union_branch.h: + * TAO_IDL/include/ast_union_label.h: + * TAO_IDL/include/ast_valuebox.h: + * TAO_IDL/include/drv_extern.h: + * TAO_IDL/include/fe_extern.h: + * TAO_IDL/include/idl_global.h: + * TAO_IDL/include/utl_exceptlist.h: + * TAO_IDL/include/utl_exprlist.h: + * TAO_IDL/include/utl_labellist.h: + * TAO_IDL/include/utl_list.h: + * TAO_IDL/include/utl_namelist.h: + * TAO_IDL/include/utl_scope.h: + * TAO_IDL/util/utl_exceptlist.cpp: + * TAO_IDL/util/utl_exprlist.cpp: + * TAO_IDL/util/utl_global.cpp: + * TAO_IDL/util/utl_idlist.cpp: + * TAO_IDL/util/utl_labellist.cpp: + * TAO_IDL/util/utl_list.cpp: + * TAO_IDL/util/utl_namelist.cpp: + * TAO_IDL/util/utl_scope.cpp: + * TAO_IDL/util/utl_string.cpp: + + - Plugged all memory leaks occuring during successful + execution, and most of the leaks occuring when there + is some kind of error. For some syntax errors + reported by the parser, it's not known how far + the parser has looked ahead (and thus what memory + it may have already allocated) when the error + is detected. Creating the AST out of types that + manage their own memory, or creating the AST on + the stack, would solve all memory leak problems, + but would require substantial redesign of the IDL + compiler front end. + + - Added generation of sendc_* operations inherited + by local interfaces, if AMI code is being generated. + + - Fixed required check for parallel inheritance of + valuetypes and their corresponding supported + interfaces, if any. + + - Added workarounds to skip problem code generation + when IDL3 constructs and AMI code generation + occur together. Integration of AMI with components + is a future development item. + + - Fixed the logic of the decision whether or not to + generate inheritance of a reference counting + parent class for valuetype OBV_* classes. + + - Added default generation of Any insertion/extraction + operators for local interfaces, and a command line + option - '-Sal' - to suppress it. + + * orbsvcs/IFR_Service/be_extern.h: + * orbsvcs/IFR_Service/be_init.cpp: + * orbsvcs/IFR_Service/be_produce.cpp: + + Changes to the IFR loader backend corresponding to the + memory leak fixes above. + + * docs/compiler.html: + * docs/releasenotes/index.html: + + Documented new IDL compiler command line option + described above. + +Thu Jun 15 12:32:58 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * TAO_IDL/be/be_visitor_valuebox/valuebox_ci.cpp: + * TAO_IDL/be/be_visitor_valuebox/valuebox_cs.cpp: + + Moved virtual methods from inline to out-of-line. + +Wed Jun 14 14:48:07 UTC 2006 Wallace Zhang <zhangw@ociweb.com> + + * tests/ORB_Local_Config/ORB_Local_Config.mwc: + + Removed the nonexistent Limits directory from + the workspace. This should correct the setup + error shown on Win2003_VC71_Debug hosted by OCI. + +Tue Jun 13 15:24:52 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * orbsvcs/orbsvcs/Log/Hash_LogRecordStore.cpp + * orbsvcs/orbsvcs/Log/Hash_LogRecordStore.h + + Removed orb, added LogMgr argument to constructor. + + Changed constructor to create a new log-channel specific poa + for iterators. Activate iterators in the new poa. + + Changed destructor to destroy the new poa, which ensures any + interators that may exist will be destroyed at the same time + as the log channel. This fixes bugzilla bug #2551. + + * orbsvcs/orbsvcs/Log/Hash_LogStore.cpp + * orbsvcs/orbsvcs/Log/Hash_LogStore.h + + Removed orb argument from constructor. + Changed to pass LogMgr to Hash_LogRecordStore constructor. + + * orbsvcs/orbsvcs/Log/Hash_Persistence_Strategy.cpp: + + Changed to pass only LogMgr to Hash_LogStore constructor. + + * orbsvcs/orbsvcs/Log/Hash_Persistence_Strategy.cpp: + * orbsvcs/orbsvcs/Log/Hash_Persistence_Strategy.h: + * orbsvcs/orbsvcs/Log/Log_Persistence_Strategy.h: + + Removed orb argument from create_log_store(). + + * orbsvcs/orbsvcs/Log/LogMgr_i.cpp + * orbsvcs/orbsvcs/Log/LogMgr_i.h + + Added orb(), factory_poa(), and log_poa() methods. We're + changing most objects to have a reference to the LogMgr, so we + don't have pass orbs and poas around, especially for the cases + where we're already passing the LogMgr. + + Changed to pass only "this" to strategy's create_log_store() + method. + +Tue Jun 13 14:56:01 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * orbsvcs/orbsvcs/Log/Log_i.cpp: + + Don't close the LogRecordStore in the TAO_Log_i destructor, if + the log channel was destroyed, it will be a dangling reference. + The LogStore should be responsible for opening and closing its + LogRecordStores. + + * orbsvcs/orbsvcs/Log/Hash_LogStore.cpp (remove): + + Obtain and delete LogRecordStore when unbinding from map. + Avoids leak. + + * orbsvcs/orbsvcs/Log/Hash_LogRecordStore.cpp (query_i): + + Re-indent to match coding conventions. + +Tue Jun 13 13:51:11 UTC 2006 Adam Mitz <mitza@ociweb.com> + + * tests/Bug_2494_Regression/client.cpp: + + The scoreboard was marking this test as "failed" because it printed + "EXCEPTION" (from the client process). This isn't a failure as long + as the server process still exits cleanly. The test is now modified + so that it doesn't print "EXCEPTION". + +Sun Jun 11 16:46:44 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * orbsvcs/orbsvcs/EventLog_i.cpp: + * orbsvcs/orbsvcs/NotifyLog_i.cpp: + * orbsvcs/orbsvcs/RTEventLog_i.cpp: + + Changed destructor to invoke destroy() on the event channel. + This notifies suppliers and consumers, and avoids a resource + leak. Fixes bugzilla bug #2559. + +Sun Jun 11 00:54:57 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * orbsvcs/orbsvcs/Log/BasicLog_i.cpp: + * orbsvcs/orbsvcs/Log/BasicLog_i.h: + * orbsvcs/orbsvcs/Log/EventLog_i.cpp: + * orbsvcs/orbsvcs/Log/EventLog_i.h: + * orbsvcs/orbsvcs/Log/NotifyLog_i.cpp: + * orbsvcs/orbsvcs/Log/NotifyLog_i.h: + * orbsvcs/orbsvcs/Log/RTEventLog_i.cpp: + * orbsvcs/orbsvcs/Log/RTEventLog_i.h: + + Remove member variable logmgr_i_, as it's also a protected + member in the base class. + +Sat Jun 10 21:54:00 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * NEWS: + + Mention that NT_Naming_Service and NT_Notify_Service must + be enabled by new MPC feature variable "winnt". + +Sat Jun 10 18:30:42 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * performance-tests/Makefile.am: + + Disable building CSD_Strategy tests. + +Sat Jun 10 10:40:02 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * examples/RTScheduling/Job_i.h: + * examples/RTScheduling/POA_Holder.h: + * orbsvcs/tests/Notify/lib/Periodic_Consumer.h: + * orbsvcs/tests/Notify/lib/Periodic_Supplier.h: + + More files where ace/Arg_Shifter.h needed to be directly + included. + +Sat Jun 10 05:03:58 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * configure.ac: + + Add AM_CONDITIONAL for BUILD_WINNT and BUILD_WXWINDOWS. + +Sat Jun 10 04:34:32 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * docs/tutorials/Quoter/AMI/Makefile.am: + * docs/tutorials/Quoter/Event_Service/Makefile.am: + * docs/tutorials/Quoter/On_Demand_Activation/Makefile.am: + * docs/tutorials/Quoter/RT_Event_Service/Makefile.am: + * docs/tutorials/Quoter/idl/Makefile.am: + * examples/AMH/Sink_Server/Makefile.am: + * examples/AMI/FL_Callback/Makefile.am: + * examples/Advanced/ch_3/Makefile.am: + * examples/Buffered_AMI/Makefile.am: + * examples/Buffered_Oneways/Makefile.am: + * examples/CSD_Strategy/ThreadPool/Makefile.am: + * examples/CSD_Strategy/ThreadPool2/Makefile.am: + * examples/CSD_Strategy/ThreadPool3/Makefile.am: + * examples/CSD_Strategy/ThreadPool4/Makefile.am: + * examples/CSD_Strategy/ThreadPool5/Makefile.am: + * examples/CSD_Strategy/ThreadPool6/Makefile.am: + * examples/Callback_Quoter/Makefile.am: + * examples/Content_Server/AMI_Iterator/Makefile.am: + * examples/Content_Server/AMI_Observer/Makefile.am: + * examples/Content_Server/SMI_Iterator/Makefile.am: + * examples/Event_Comm/Makefile.am: + * examples/Kokyu_dsrt_schedulers/Makefile.am: + * examples/Kokyu_dsrt_schedulers/fp_example/Makefile.am: + * examples/Kokyu_dsrt_schedulers/mif_example/Makefile.am: + * examples/Kokyu_dsrt_schedulers/muf_example/Makefile.am: + * examples/Load_Balancing/Makefile.am: + * examples/Load_Balancing_persistent/Makefile.am: + * examples/Logging/Makefile.am: + * examples/OBV/Typed_Events/Makefile.am: + * examples/POA/DSI/Makefile.am: + * examples/POA/Default_Servant/Makefile.am: + * examples/POA/Forwarding/Makefile.am: + * examples/POA/Generic_Servant/Makefile.am: + * examples/POA/TIE/Makefile.am: + * examples/Persistent_Grid/Makefile.am: + * examples/PluggableUDP/tests/Basic/Makefile.am: + * examples/PluggableUDP/tests/Performance/Makefile.am: + * examples/PluggableUDP/tests/SimplePerformance/Makefile.am: + * examples/Quoter/Makefile.am: + * examples/RTCORBA/Activity/Makefile.am: + * examples/RTScheduling/Fixed_Priority_Scheduler/Makefile.am: + * examples/RTScheduling/MIF_Scheduler/Makefile.am: + * examples/Simple/bank/Makefile.am: + * examples/Simple/chat/Makefile.am: + * examples/Simple/echo/Makefile.am: + * examples/Simple/grid/Makefile.am: + * examples/Simple/time/Makefile.am: + * examples/Simple/time-date/Makefile.am: + * examples/Simulator/Event_Supplier/Makefile.am: + * examples/TypeCode_Creation/Makefile.am: + * examples/ior_corbaloc/Makefile.am: + * examples/mfc/Makefile.am: + * interop-tests/wchar/Makefile.am: + * orbsvcs/FTRT_Event_Service/Factory_Service/Makefile.am: + * orbsvcs/FTRT_Event_Service/Gateway_Service/Makefile.am: + * orbsvcs/ImplRepo_Service/Makefile.am: + * orbsvcs/Naming_Service/Makefile.am: + * orbsvcs/Notify_Service/Makefile.am: + * orbsvcs/examples/CosEC/Factory/Makefile.am: + * orbsvcs/examples/CosEC/TypedSimple/Makefile.am: + * orbsvcs/examples/FaultTolerance/RolyPoly/Makefile.am: + * orbsvcs/examples/ImR/Advanced/Makefile.am: + * orbsvcs/examples/ImR/Combined_Service/Makefile.am: + * orbsvcs/examples/LoadBalancing/Makefile.am: + * orbsvcs/examples/ORT/Makefile.am: + * orbsvcs/examples/Security/Send_File/Makefile.am: + * orbsvcs/orbsvcs/Makefile.am: + * orbsvcs/performance-tests/LoadBalancing/LBPerf/RPS/Makefile.am: + * orbsvcs/performance-tests/RTEvent/RTCORBA_Baseline/Makefile.am: + * orbsvcs/performance-tests/RTEvent/RTCORBA_Callback/Makefile.am: + * orbsvcs/performance-tests/RTEvent/lib/Makefile.am: + * orbsvcs/tests/BiDir_CORBALOC/Makefile.am: + * orbsvcs/tests/Bug_1395_Regression/Makefile.am: + * orbsvcs/tests/Bug_2247_Regression/Makefile.am: + * orbsvcs/tests/Bug_2285_Regression/Makefile.am: + * orbsvcs/tests/Bug_2287_Regression/Makefile.am: + * orbsvcs/tests/Bug_2316_Regression/Makefile.am: + * orbsvcs/tests/Bug_2377_Regression/Makefile.am: + * orbsvcs/tests/EC_Custom_Marshal/Makefile.am: + * orbsvcs/tests/FT_App/Makefile.am: + * orbsvcs/tests/FaultTolerance/GroupRef_Manipulation/Makefile.am: + * orbsvcs/tests/FaultTolerance/IOGR/Makefile.am: + * orbsvcs/tests/FtRtEvent/Makefile.am: + * orbsvcs/tests/HTIOP/AMI/Makefile.am: + * orbsvcs/tests/HTIOP/BiDirectional/Makefile.am: + * orbsvcs/tests/HTIOP/Hello/Makefile.am: + * orbsvcs/tests/IOR_MCast/Makefile.am: + * orbsvcs/tests/ImplRepo/Makefile.am: + * orbsvcs/tests/ImplRepo/scale/Makefile.am: + * orbsvcs/tests/InterfaceRepo/Application_Test/Makefile.am: + * orbsvcs/tests/Interoperable_Naming/Makefile.am: + * orbsvcs/tests/LoadBalancing/GenericFactory/Application_Controlled/Makefile.am: + * orbsvcs/tests/LoadBalancing/GenericFactory/Infrastructure_Controlled/Makefile.am: + * orbsvcs/tests/LoadBalancing/GenericFactory/Manage_Object_Group/Makefile.am: + * orbsvcs/tests/Miop/McastHello/Makefile.am: + * orbsvcs/tests/Notify/Blocking/Makefile.am: + * orbsvcs/tests/Notify/Discarding/Makefile.am: + * orbsvcs/tests/Notify/MT_Dispatching/Makefile.am: + * orbsvcs/tests/Notify/Ordering/Makefile.am: + * orbsvcs/tests/Notify/Sequence_Multi_ETCL_Filter/Makefile.am: + * orbsvcs/tests/Notify/Sequence_Multi_Filter/Makefile.am: + * orbsvcs/tests/Notify/Structured_Filter/Makefile.am: + * orbsvcs/tests/Notify/Structured_Multi_Filter/Makefile.am: + * orbsvcs/tests/Notify/lib/Makefile.am: + * orbsvcs/tests/Notify/performance-tests/Filter/Makefile.am: + * orbsvcs/tests/Redundant_Naming/Makefile.am: + * orbsvcs/tests/Security/BiDirectional/Makefile.am: + * orbsvcs/tests/Security/Big_Request/Makefile.am: + * orbsvcs/tests/Security/Callback/Makefile.am: + * orbsvcs/tests/Security/Crash_Test/Makefile.am: + * orbsvcs/tests/Security/MT_IIOP_SSL/Makefile.am: + * orbsvcs/tests/Security/MT_SSLIOP/Makefile.am: + * orbsvcs/tests/Security/Secure_Invocation/Makefile.am: + * orbsvcs/tests/Security/ssliop_corbaloc/Makefile.am: + * orbsvcs/tests/Simple_Naming/Makefile.am: + * orbsvcs/tests/Trading/Makefile.am: + * orbsvcs/tests/ior_corbaname/Makefile.am: + * performance-tests/Makefile.am: + * performance-tests/Anyop/Makefile.am: + * performance-tests/CSD_Strategy/TestInf/Makefile.am: + * performance-tests/CSD_Strategy/TestServant/Makefile.am: + * performance-tests/Callback/Makefile.am: + * performance-tests/Cubit/TAO/DII_Cubit/Makefile.am: + * performance-tests/Cubit/TAO/IDL_Cubit/Makefile.am: + * performance-tests/Cubit/TAO/MT_Cubit/Makefile.am: + * performance-tests/Latency/AMH_Single_Threaded/Makefile.am: + * performance-tests/Latency/AMI/Makefile.am: + * performance-tests/Latency/Collocation/Makefile.am: + * performance-tests/Latency/DII/Makefile.am: + * performance-tests/Latency/DSI/Makefile.am: + * performance-tests/Latency/Deferred/Makefile.am: + * performance-tests/Latency/Single_Threaded/Makefile.am: + * performance-tests/Latency/Thread_Per_Connection/Makefile.am: + * performance-tests/Latency/Thread_Pool/Makefile.am: + * performance-tests/Memory/IORsize/Makefile.am: + * performance-tests/Memory/Single_Threaded/Makefile.am: + * performance-tests/POA/Demux/Makefile.am: + * performance-tests/POA/Implicit_Activation/Makefile.am: + * performance-tests/POA/Object_Creation_And_Registration/Makefile.am: + * performance-tests/Pluggable/Makefile.am: + * performance-tests/Protocols/Makefile.am: + * performance-tests/RTCorba/Multiple_Endpoints/Common/Makefile.am: + * performance-tests/RTCorba/Oneways/Reliable/Makefile.am: + * performance-tests/RTCorba/Thread_Pool/Makefile.am: + * performance-tests/Sequence_Latency/AMH_Single_Threaded/Makefile.am: + * performance-tests/Sequence_Latency/AMI/Makefile.am: + * performance-tests/Sequence_Latency/DII/Makefile.am: + * performance-tests/Sequence_Latency/DSI/Makefile.am: + * performance-tests/Sequence_Latency/Deferred/Makefile.am: + * performance-tests/Sequence_Latency/Single_Threaded/Makefile.am: + * performance-tests/Sequence_Latency/Thread_Per_Connection/Makefile.am: + * performance-tests/Sequence_Latency/Thread_Pool/Makefile.am: + * performance-tests/Throughput/Makefile.am: + * tao/Makefile.am: + * utils/Makefile.am: + + Regenerated. + + * configure.ac: + + Configure utils/wxNamingViewer/Makefile.am. + + * utils/wxNamingViewer/Makefile.am: + + New file. + +Sat Jun 10 03:52:25 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * orbsvcs/Naming_Service/Naming_Service.mpc: + + Changed NT_Naming_Service project to require new MPC feature + variable "winnt". Avoids building and installing this on non + Windows systems. Fixes bugzilla bug #2412. + + * orbsvcs/Notify_Service/Notify_Service.mpc: + + Changed NT_Notify_Service project to require new MPC feature + variable "winnt". Avoids building and installing this on non + Windows systems. Fixes bugzilla bug #2411. + +Fri Jun 9 18:33:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/performance-tests/RTEvent/lib/Client_Options.{h,cpp}: + * examples/RTCORBA/Activity/POA_Holder.{h,cpp}: + * examples/RTCORBA/Activity/Periodic_Task.{h,cpp}: + * examples/RTCORBA/Activity/Job_i.{h,cpp}: + * examples/RTCORBA/Activity/Builder.{h,cpp}: + * orbsvcs/tests/Notify/lib/Peer.h: + Just include ACE_Arg_Shifter in the header file instead of + using a forward declaration + +Fri Jun 9 14:49:06 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tests/Stack_Recursion/run_test.pl: + + Doubled the timeout to accomodate really slow machines. This was + necessitated by new sequence initialization code. + +Fri Jun 9 14:48:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/tests/Notify/lib/Options_Parser.{h,cpp}: + Moved the include of Arg_Shifter.h from cpp to header file + +Fri Jun 9 14:35:56 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tests/Big_Request_Muxing/Client_Task.h: + * tests/Big_Request_Muxing/Client_Task.cpp: + * tests/Big_Request_Muxing/Payload_Receiver.h: + * tests/Big_Request_Muxing/Payload_Receiver.cpp: + * tests/Big_Request_Muxing/Test.idl: + * tests/Big_Request_Muxing/client.cpp: + * tests/Big_Request_Muxing/run_test.pl: + * tests/Big_Request_Muxing/server.cpp: + + This test frequently shows false errors due dropped SYNC_NONE + messages. These errors are false because dropping such messages + is not an error. The test now separately accounts for SYNC_NONE + invocations separately by using a different operation. + + I also modified the timeout loop in an earlier attempt to + resolve the periodic dropped messages, thinking the timeout + occured due to heavy loads. Now the timeout less about the whole + event and more about the duration of individual requests. + +Thu Jun 8 02:16:30 UTC 2006 Douglas C. Schmidt <schmidt@dre.vanderbilt.edu> + + * orbsvcs/Event_Service/Event_Service.cpp (parse_args): Added a + missing break statement. Thanks to Sunil Rottoo <sunil dot + rottoo at idilia dot com> for reporting this. + +Wed Jun 7 14:24:55 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * orbsvcs/orbsvcs/Log/Hash_LogRecordStore.h: + + Now with all with the cleanup, we can switch from using a + ACE_Hash_Map to a ACE_RB_Tree by changing one typedef. + + Now that LogRecords are ordered by RecordId, for all practical + purposes this resolves bugzilla bugs #1980 and #1981. While it + doesn't handle the case where the RecordId's wrap, with 64 bits, + that's not worth losing too much sleep over. Even if we logged + 1,000,000 records per second, it would take nearly 600,000 years + to wrap. + +Wed Jun 7 09:04:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Invocation_Adapter.cpp: + + Corrected the check whether a request has arguments or not. This + fixes interoperability issues with Orbix. Thanks to Phil + Billingham <phil_billingham at ml dot com> for reporting + this. This fixes bugzilla bug 2548 + +Tue Jun 6 21:05:19 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tests/Sequence_Unit_Tests/unbounded_octet_sequence_nocopy_ut.cpp: + There was a buffer overrun inside this test case. + +Tue Jun 6 17:25:15 UTC 2006 Chad Elliott <elliott_c@ociweb.com> + + * TAO_IDL/tao_idl.mpc: + + Combined two automake specific sections into one. + + * orbsvcs/FTRT_Event_Service/Factory_Service/FTRTEC_Factory_Service.mpc: + * orbsvcs/FTRT_Event_Service/Gateway_Service/FTRTEC_Gateway_Service.mpc: + * orbsvcs/tests/FtRtEvent/FtRtEvent.mpc: + + Replaced a gnuace specific section to link in tje TAO_Strategies + library with the inheritance of the strategies base project. + +Tue Jun 6 17:02:57 UTC 2006 Yan Dai <dai_y@ociweb.com> + + * tests/DII_Collocation_Tests/Client_Task.cpp: + * tests/DII_Collocation_Tests/Client_Task.h: + * tests/DII_Collocation_Tests/Collocated_Test.cpp: + * tests/DII_Collocation_Tests/Hello.cpp: + * tests/DII_Collocation_Tests/Hello.h: + * tests/DII_Collocation_Tests/README: + * tests/DII_Collocation_Tests/run_test.pl: + * tests/DII_Collocation_Tests/Server_Task.cpp: + * tests/DII_Collocation_Tests/Server_Task.h: + * tests/DII_Collocation_Tests/Test.idl: + + Added incomplete twoway test. More test cases (OUT, INOUT and + RETURN) need be added. + +Tue Jun 6 14:48:33 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * orbsvcs/orbsvcs/Log/Hash_Iterator_i.cpp: + + Use iter->item() instead of (*iter).int_id_ to dereference + iterators. + + Changed get() to set the length of the output sequence to the + maximum number of log records (this will be shrunk to the real + value once we find how many records match the constraint). We + must have got lucky with the old sequence implementation... + +Tue Jun 6 13:56:49 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * orbsvcs/orbsvcs/Log/Hash_Iterator_i.h: + * orbsvcs/orbsvcs/Log/Hash_LogRecordStore.h: + + Consistantly use the LOG_RECORD_STORE typedef instead of + LOG_RECORD_HASH_MAP. The former is supposed to abstract the + latter. + +Tue Jun 6 13:38:39 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * orbsvcs/orbsvcs/Log/Hash_LogRecordStore.cpp: + * orbsvcs/orbsvcs/Log/Hash_LogRecordStore.h: + + Rename rec_hash_ member variable to rec_map_, as we plan to + change the type. + +Tue Jun 6 13:23:06 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * orbsvcs/orbsvcs/Log/Hash_LogRecordStore.cpp: + + Use iter->item() instead of (*iter).int_id_ to dereference + iterators. + +Tue Jun 6 12:29:03 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tests/ORB_Local_Config/Two_DLL_ORB/ORB_DLL.cpp: + Explicitly duplicate the command-line arguments passed to the + client and server ORBs. On some platforms, it wasn't enough to + merely readjust the length by setting last argv to 0. + +Mon Jun 5 03:19:58 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/ORB_Core.cpp: + * tao/ORB_Core.h: + + There is a class for containing certain initialization values + used by the ORB Core that are set during Dynamic loading of + service configuration objects where there is no way to get a + pointer to an ORB Core instance. These values were stored in + a static instance of this class, TAO_ORB_Core_Static_Resources. + However, using a static instance of this class is incongruent + with the notion of per-ORB configuration, so this change + addresses the problem by making the resources container a + service object which is initialized in the local configuration + context for each ORB, as well as a copy in the global + configuration context. + + This fix specifically addresses the problem of having a later + ORB initialization, such as for a second ORB, affecting the + configuration of earlier ORBs. If this happens in separate + threads, a race can occur leading to unpredictable results. + + However, it might be possible for some ORB-related configuration + to be supplied via service configuration after ORB_init is + called. If that is the case, then the appropriate configuration + context must be set with ACE_Service_Config_Guard during the + duration of the configuration. This will ensure the newly loaded + configuration object goes into the correct context. + + * tao/CSD_Framework/CSD_Framework_Loader.cpp: + + Reverted the temporary fix, it is no longer needed. + +Sun Jun 4 16:04:36 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tests/DII_Collocation_Tests/oneway/Server_Task.h: + + Add versioned namespace wrappers for the forward declaration of + an ACE class. + +Sun Jun 4 14:39:56 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/CSD_Framework/CSD_Default_Servant_Dispatcher.cpp: + * tao/CSD_Framework/CSD_Framework_Loader.cpp: + + This is a temporary (duration unknown) resolution to a problem + highlighted by the intermitant failure in the Two_DLL_ORB test. + The problem is that the POA factory name and POA factory + directive are held in a static instance of the + TAO_ORB_Core_Static_Resoures. The problem is that with multiple + configuration contexts, the ORBs are supposed to be configured + separately, but this static resource violates that principle. + +Sat Jun 3 19:38:44 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * orbsvcs/orbsvcs/Log/Hash_LogRecordStore.cpp: + * orbsvcs/orbsvcs/Log/Hash_LogRecordStore.h: + + Added new varient of remove_i() that takes an iterator instead + of a record id as its argument. In almost all cases, we have a + current iterator when deleting removing a log record; if we use + it when unbinding from the map, we avoid an extraneous lookup. + Update callers. + +Sat Jun 3 19:03:53 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * orbsvcs/orbsvcs/Log/Hash_LogRecordStore.cpp: + + Remove TODO comments about storing the id on a list when a log + record is deleted, and reusing those ids when a new log record + is inserted. If we did this, ids would be practically useless + for use by client applications. It's much more useful to have + montonically increasing ids. + +Sat Jun 3 17:37:11 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * orbsvcs/orbsvcs/Log/Hash_LogRecordStore.cpp: + + Changed last instance of use of old iterator API to the new + STL-like API. + + Changed purge_old_records(), delete_records(), and + remove_old_records() to increment iterator before removing log + record. Removing the record invalidates the current iterator, + and incrementing afterwards it resulted in undefined behavior. + +Sat Jun 3 16:10:47 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * orbsvcs/orbsvcs/Log/Hash_LogRecordStore.cpp: + * orbsvcs/orbsvcs/Log/Hash_LogRecordStore.h: + + Changed the match() and delete_records() methods to iterate + through the map by themselves instead of invoking match_i(). + While match_i() factored out common code, it did so at the + expense of an extra conditional in the hot path. Removed + match_i(). + +Sat Jun 3 15:46:45 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * orbsvcs/orbsvcs/Log/Hash_LogRecordStore.cpp: + * orbsvcs/orbsvcs/Log/Hash_LogRecordStore.h: + + Remove private remove() method. Update callers to invoke + remove_i(). + + Rename update() to update_i(). + Rename retrieve() to retrieve_i(). + +Sat Jun 3 05:25:09 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * orbsvcs/orbsvcs/Log/PersistStore.cpp: + * orbsvcs/orbsvcs/Log/PersistStore.h: + * orbsvcs/orbsvcs/Log/LogRecordStore_persist.cpp: + * orbsvcs/orbsvcs/Log/LogRecordStore_persist.h: + + Remove files. This was an old attempt at log record persistence + that predated the plug-in strategy. + +Sat Jun 3 05:20:50 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * orbsvcs/orbsvcs/Log/Hash_LogRecordStore.h: + + Declare LOG_RECORD_HASH_MAP_ENTRY and LOG_RECORD_HASH_MAP_ITER + in terms of LOG_RECORD_HASH_MAP::ENTRY, ...::ITERATOR, instead + of duplicating all the template stuff. Will make it easier to + migrate to an ordered ma types, for bugzilla bugs #1980, #1981, + etc. + +Sat Jun 3 05:19:59 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * orbsvcs/examples/Log/Event/run_test.pl: + * orbsvcs/examples/Log/Notify/run_test.pl: + * orbsvcs/examples/Log/RTEvent/run_test.pl: + + Kill the Consumer process instead of considering the test to be + failed if it did not exit itself. The Consumer connects to the + Log Factory's event channel, not the event/notification channel + for the log, so even though its disconnect_push_consumer method + invokes shutdown on the orb, the it doesn't matter since it's + not going to be called when the log is destroyed. + +Sat Jun 3 05:03:15 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/TAO_Internal.cpp: + Added ACE_MT guards around new synchronization code to make it + single-thread-build safe. + +Sat Jun 3 04:26:08 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tests/ORB_Local_Config/Two_DLL_ORB/README: + + Added a bit of a documentation on the test + + * tests/ORB_Local_Config/Two_DLL_ORB/Test.cpp: + + Modified to conditionaly excersise one of the two test + scenarios. Originally, the test used SSLIOP, however that + service is not built by default. So the change makes use of + another service, which is normally available in "default" + builds. + + * tests/ORB_Local_Config/Two_DLL_ORB/primary-csd.conf: + * tests/ORB_Local_Config/Two_DLL_ORB/primary-ssl.conf: + * tests/ORB_Local_Config/Two_DLL_ORB/secondary-csd.conf: + * tests/ORB_Local_Config/Two_DLL_ORB/secondary-empty.conf: + * tests/ORB_Local_Config/Two_DLL_ORB/secondary-ssl.conf: + + Added a simple names for the the configuration files. + + * tests/ORB_Local_Config/Two_DLL_ORB/Service_Config_ORB_Test.conf: + * tests/ORB_Local_Config/Two_DLL_ORB/Service_Config_ORB_Test2.conf: + * tests/ORB_Local_Config/Two_DLL_ORB/Service_Config_ORB_Test3.conf: + + Removed these files. + +Fri Jun 2 21:02:46 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tests/CodeSets/libs/UCS4_UTF16/WUCS4_UTF16.cpp: + Fixed a typo from the earlier commit. + +Fri Jun 2 19:52:19 UTC 2006 Yan Dai <dai_y@ociweb.com> + + * tests/DII_Collocation_Tests/twoway/Client_Task.cpp: + * tests/DII_Collocation_Tests/twoway/Client_Task.h: + * tests/DII_Collocation_Tests/twoway/Collocated_Test.cpp: + * tests/DII_Collocation_Tests/twoway/DII_Collocation_Tests.mpc: + * tests/DII_Collocation_Tests/twoway/Hello.cpp: + * tests/DII_Collocation_Tests/twoway/Hello.h: + * tests/DII_Collocation_Tests/twoway/README: + * tests/DII_Collocation_Tests/twoway/run_test.pl: + * tests/DII_Collocation_Tests/twoway/Server_Task.cpp: + * tests/DII_Collocation_Tests/twoway/Server_Task.h: + * tests/DII_Collocation_Tests/twoway/Test.idl: + + Removed twoway test. + +Fri Jun 2 12:05:13 USMST 2006 Yan Dai <dai_y@ociweb.com> + + * tests/DII_Collocation_Tests/oneway/Collocated_Test.cpp: + + Fixed compilation errors due to reference to a new file. + +Fri Jun 2 13:33:19 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tests/Big_Twoways/server.cpp: + Fixed a minor memory leak caused by not cleaning up servants. + + * tests/CodeSets/libs/UCS4_UTF16/WUCS4_UTF16.cpp: + Addressed compiler warnings generated by some platforms with + 2-byte wchars. This codeset translator should only be used on + hosts with 4-byte wchars, since UCS4 is a 32-bit codeset, but + not all of the test platforms meet that criterium. + +Fri Jun 2 12:36:58 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tao/TAO_Internal.cpp: + + Adding synchronization, in the form of a condition variable to + fix a race condition occurring when a non-default ORB enters + open_services, before the default ORB has completed + initializing the global service objects. According to the spec, + the default ORB is the one that gets to call ORB_init + first. The ORB-specific configuration implementation designates + the default ORB to initialize globally available service + objects, like the Resource Factory. If a non-default ORB beats + it to the resource initialization, it would cause a SEGV. + + * tests/ORB_Local_Config/Two_DLL_ORB/Service_Config_ORB_Test.conf: + + As a consequence of implementing the ORB-specific + configuration, we discovered many implicit assumptions about + the lifetime of objects. For example, process-global Singletons + are created by service objects, loaded by ORBs, which are + themselves initialized from a DLL-loaded code. If that DLL, + which also contains the code for tearing down the Singleton + gets unloaded (for instance, as a consequence of calling + orb->destroy()), its TEXT segment will no longer be mapped in + memory when the Object Manager tries to destroy the Singleton. + + Ossama has a solution for a similar problem, involving the TAO + Singleton Manager, for DLL-loaded ORBs (see the DLL_ORB test), + however it doesn't cover the ACE Object Manager. + + This configuration change is a workaround the fact that SSLIOP, + loaded as a private service object by a DLL-based ORB, + registers a process-global Singleton - ACE_SLL_Context, with + the Object Manager. In an use-case with multiple dynamically + loaded components containing ORBs, SSLIOP should be loaded + prior to loading any of the dynamic components using it. In + general, any SO that uses ACE Singletons must be treated + similarly, or rewritten. + +Fri Jun 2 04:44:17 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * orbsvcs/orbsvcs/Log/EventLogFactory_i.cpp: + * orbsvcs/orbsvcs/Log/EventLog_i.cpp: + * orbsvcs/orbsvcs/Log/EventLog_i.h: + * orbsvcs/orbsvcs/Log/RTEventLogFactory_i.cpp: + * orbsvcs/orbsvcs/Log/RTEventLog_i.cpp: + * orbsvcs/orbsvcs/Log/RTEventLog_i.h: + + We must pass both the RootPOA and the Log POA to the log servant + ctor, destroy() needs to deactivate the servant on the Log POA. + +Thu Jun 1 23:13:58 UTC 2006 Yan Dai <dai_y@ociweb.com> + + * tests/DII_Collocation_Tests/Client_Task.cpp: + * tests/DII_Collocation_Tests/Client_Task.h: + * tests/DII_Collocation_Tests/Collocated_Test.cpp: + * tests/DII_Collocation_Tests/DII_Collocation_Tests.mpc: + * tests/DII_Collocation_Tests/Hello.cpp: + * tests/DII_Collocation_Tests/Hello.h: + * tests/DII_Collocation_Tests/README: + * tests/DII_Collocation_Tests/run_test.pl: + * tests/DII_Collocation_Tests/Server_Task.cpp: + * tests/DII_Collocation_Tests/Server_Task.h: + * tests/DII_Collocation_Tests/Test.idl: + + These files are moved to the tests/DII_Collocation_Tests/oneway + directory since a new test for twoway collocated DII request + is added. + + * tests/DII_Collocation_Tests/oneway/Client_Task.cpp: + * tests/DII_Collocation_Tests/oneway/Client_Task.h: + * tests/DII_Collocation_Tests/oneway/Collocated_Test.cpp: + * tests/DII_Collocation_Tests/oneway/DII_Collocation_Tests.mpc: + * tests/DII_Collocation_Tests/oneway/Hello.cpp: + * tests/DII_Collocation_Tests/oneway/Hello.h: + * tests/DII_Collocation_Tests/oneway/README: + * tests/DII_Collocation_Tests/oneway/run_test.pl: + * tests/DII_Collocation_Tests/oneway/Server_Task.cpp: + * tests/DII_Collocation_Tests/oneway/Server_Task.h: + * tests/DII_Collocation_Tests/oneway/Test.idl: + + These files are moved from tests/DII_Collocation_Tests directory. + + * tests/DII_Collocation_Tests/twoway/Client_Task.cpp: + * tests/DII_Collocation_Tests/twoway/Client_Task.h: + * tests/DII_Collocation_Tests/twoway/Collocated_Test.cpp: + * tests/DII_Collocation_Tests/twoway/DII_Collocation_Tests.mpc: + * tests/DII_Collocation_Tests/twoway/Hello.cpp: + * tests/DII_Collocation_Tests/twoway/Hello.h: + * tests/DII_Collocation_Tests/twoway/README: + * tests/DII_Collocation_Tests/twoway/run_test.pl: + * tests/DII_Collocation_Tests/twoway/Server_Task.cpp: + * tests/DII_Collocation_Tests/twoway/Server_Task.h: + * tests/DII_Collocation_Tests/twoway/Test.idl: + + Added a test for twoway collocation DII request via invoke (). + This test should fail now for the same reason as the + oneway collocation DII request. See bugzilla bug #2545 + for details. + +Thu Jun 1 17:53:33 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * orbsvcs/orbsvcs/Log/EventLogFactory_i.cpp: + * orbsvcs/orbsvcs/Log/RTEventLogFactory_i.cpp: + + Pass the RootPOA instead of the Log POA to the log servant ctor. + The log servant creates and attempts to activate a event channel + on the Log POA, which was failed because TAO's log event channel + implementation uses implicit activation (_this), and the Log POA + ImplicitActivationPolicy is NO_IMPLICIT_ACTIVATION. In time, we + should consider whether we should create another POA just for + the event channels. + + * orbsvcs/examples/Log/Event/run_test.pl: + * orbsvcs/examples/Log/Notify/run_test.pl: + * orbsvcs/examples/Log/RTEvent/run_test.pl: + + New files. + +Thu Jun 1 16:54:56 UTC 2006 Yan Dai <dai_y@ociweb.com> + + * tests/DII_Collocation_Tests/Client_Task.cpp: + * tests/DII_Collocation_Tests/Client_Task.h: + * tests/DII_Collocation_Tests/Collocated_Test.cpp: + * tests/DII_Collocation_Tests/DII_Collocation_Tests.mpc: + * tests/DII_Collocation_Tests/Hello.cpp: + * tests/DII_Collocation_Tests/Hello.h: + * tests/DII_Collocation_Tests/README: + * tests/DII_Collocation_Tests/run_test.pl: + * tests/DII_Collocation_Tests/Server_Task.cpp: + * tests/DII_Collocation_Tests/Server_Task.h: + * tests/DII_Collocation_Tests/Test.idl: + + Added new DII_Collocation_Tests test to show bug #2545. + The test crashes on get_in_arg() or gives incorrect arguments + when the request is collocated oneway request and has "IN" + arguments. See bugzilla bug #2545. + +Thu Jun 1 14:16:06 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * orbsvcs/orbsvcs/Log/Log_i.cpp: + + Changed reset_capacity_alarm_thresholds() to do nothing if + LogFullActionType is wrap. + +Thu Jun 1 14:06:20 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * orbsvcs/examples/Log/Basic/run_test.pl: + + New file. + +Thu Jun 1 14:02:30 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * orbsvcs/orbsvcs/Log/Hash_LogRecordStore.cpp: + + Changed set_records_attribute() to validate the query language + grammar. + + Changed query_i() to set the length of the output sequence to + the maximum number of log records (this will be shrunk to the + real value once we find how many records match the constraint). + We must have got lucky with the old sequence implementation... + +Thu Jun 1 11:38:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Valuetype/ValueBase.cpp: + Const improvements + +Thu Jun 1 10:26:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/LocateRequest_Invocation_Adapter.cpp: + Use false instead of 0 + + * tao/operation_details.h: + Improved documentation + +Thu Jun 1 08:59:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/AnyTypeCode/Any_Dual_Impl_T.cpp: + Return false instead 0 + + * tao/AnyTypeCode/*.cpp: + Fixed rcsid tags + +Thu Jun 1 02:26:51 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/RTCORBA/RT_PolicyFactory.h: + * tests/AMH_Oneway/client.cpp: + * tests/AMH_Oneway/server.cpp: + Memory leak fixes. + +Wed May 31 17:47:55 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/AnyTypeCode/Marshal.cpp: + Fuzz fix. + +Wed May 31 13:35:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/AnyTypeCode/Marshal.cpp: + Handle a valuebox in the same way as a regular valuetype. Fixes + bugzilla bug 2542. Thanks to Jiang Wei + <jiangwei_1976 at yahoo dot com dot cn> for reporting this. + +Wed May 31 10:26:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/AnyTypeCode/skip.cpp: + Const improvements + +Wed May 31 03:54:20 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/TAO_Internal.cpp: + + Revert Jeff's May 25 change, it prevented the calling of a + necessary function if the debuglevel wasn't set high + enough. Reworked the logic enough to ensure the value Jeff's + change was protecting didn't start causing problems again. + + * tests/ORB_Local_Config/Two_DLL_ORB/ORB_DLL.cpp: + * tests/ORB_Local_Config/Two_DLL_ORB/Service_Config_ORB_Test.conf: + * tests/ORB_Local_Config/Two_DLL_ORB/Service_Config_ORB_Test2.conf: + * tests/ORB_Local_Config/Two_DLL_ORB/Service_Config_ORB_Test3.conf: + * tests/ORB_Local_Config/Two_DLL_ORB/client.cpp: + * tests/ORB_Local_Config/Two_DLL_ORB/server.cpp: + + Update this test to resolve some runtime issues and to ensure + the proper loading of subsequent service configuration files. + +Tue May 30 19:17:28 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * orbsvcs/orbsvcs/Security/Security_PolicyFactory.h: + + Fix a memory leak resulting from the failure to use a reference + counted local object for the policy factory. + + * tao/ORBInitializer_Registry.cpp: + * tao/ORB_Core.cpp: + + Make use of changes in the ACE_Service_Gestalt and + ACE_Dynamic_Service<> classes to address separation of + configuration contexts. + + * tests/ORB_Local_Config/Bug_1459/Test.cpp: + + Fix for memory leaks induced through improper ORB destruction. + +Tue May 30 18:33:12 UTC 2006 Douglas C. Schmidt <schmidt@dre.vanderbilt.edu> + + * tao/DLL_Parser.h: Updated the documentation to explain the + lookup scheme when a filename is given. Thanks to Phlip + <phlip2005 at gmail dot com> for motivating this. + +Tue May 30 16:31:14 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * tests/Bug_2543_Regression/bug_2542_regression.cpp: + + Added .in() to an OctetSeq_var passed as an argument. + +Tue May 30 16:25:24 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * tests/Bug_2542_Regression/bug_2542_regression.cpp: + + Added .in() to an OctetSeq_var passed as an argument. + +Tue May 30 15:53:50 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * orbsvcs/orbsvcs/Log/Hash_LogRecordStore.cpp (set_records_attribute): + + Commented out unused argument. + +Tue May 30 14:34:07 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/IDL_Test/array.idl: + Added FourDArray as testcase + +Tue May 30 14:24:07 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * TAO_IDL/be/be_visitor_valuetype/valuetype_cs.cpp: + Generate false instead of 0 + + * TAO_IDL/be/be_visitor_valuetype/marshal_cs.cpp: + Const improvement to generated code + + * TAO_IDL/be/be_visitor_operation/amh_rh_ss.cpp: + Fixed 64bit conversion warning + + * TAO_IDL/be/be_visitor_interface/interface_ss.cpp: + Const improvement to generated code and use false instead of 0 + +Tue May 30 14:10:07 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/AnyTypeCode/Marshal.cpp: + * tao/AnyTypeCode/Marshal.inl: + * tao/AnyTypeCode/skip.cpp: + * tao/AnyTypeCode/Any_Unknown_IDL_Type.cpp: + * tao/AnyTypeCode/Any.cpp: + * tao/CodecFactory/CDR_Encaps_Codec.cpp: + * tao/PI/PI.cpp: + Use true/false and const improvements + +Tue May 30 13:42:07 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/GIOP_Message_Base.{h,cpp}: + Removed the buffer as member, let the cdr stream get the buffer + from the allocator. This allocation is just done once at the + creation so this shouldn't impact performance a lot. Also use + the size argument passed to the constructor, specific protocol + implementations to pass this down to its base to set a + specific initial buffer size. + +Tue May 30 10:04:07 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Bug_2543_Regression/bug_2543_regression.cpp: + * tests/Bug_2542_Regression/bug_2542_regression.cpp: + Fixed conversion warnings + +Tue May 30 01:59:07 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * orbsvcs/orbsvcs/Log/Hash_LogRecordStore.cpp: + + Changed set_records_attribute() to iterate through all the + records in the hash map instead of calling query() and (if there + are enough matching records) fiddling with iterators, etc. This + is not only faster, it also avoids a deadlock that shows up when + the iterator takes the already held rwlock. + +Tue May 30 01:01:39 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * NEWS: + + Document changes. + + * orbsvcs/orbsvcs/Log/Hash_LogRecordStore.cpp: + * orbsvcs/orbsvcs/Log/Hash_LogRecordStore.h: + * orbsvcs/orbsvcs/Log/LogRecordStore.h: + * orbsvcs/orbsvcs/Log/Log_i.cpp: + + When a log channel's LogFullActionType is wrap, the capacity + threshold alarms "are triggered as if coupled to a gauge that + counts from zero to the highest capacity threshold value and + then resets to zero". + + The log service didn't implement such a gauge and compared the + log channel's current size with the maximum size (as if the + LogFullActionType was halt). This could result in an alarm + being sent for each log record, as the log channel will almost + always be "full". + + Changed plug-in Strategy to maintain gauge. Added get_gauge() + and reset_gauge(). + + Changed log channel to compare the value of the gauge with the + maximum size when the LogFullActionType is wrap. + + Fixes bugzilla #2420. + +Mon May 29 14:28:17 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * NEWS: + + Document changes. + + * orbsvcs/orbsvcs/Log/Hash_LogRecordStore.cpp: + * orbsvcs/orbsvcs/Log/Hash_LogRecordStore.h: + * orbsvcs/orbsvcs/Log/LogRecordStore.h: + * orbsvcs/orbsvcs/Log/Log_i.cpp: + + Changed plug-in Strategy interface. Added get_record_attribute(), + set_record_attribute(), and set_records_attribute(); removed + retrieve(), update(), and remove(). + + This will allow plug-in Strategies to handle these high-level + operations more efficiently. + +Mon May 29 08:05:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Bug_2542_Regression/*: + New regression for Bug_2542. Thanks to Jiang Wei + <jiangwei_1976 at yahoo dot com dot cn> for creating this regression. + + * tests/Bug_2543_Regression/*: + New regression for Bug_2543. Thanks to Jiang Wei + <jiangwei_1976 at yahoo dot com dot cn> for creating this regression. + +Sun May 28 23:15:51 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * orbsvcs/orbsvcs/Log/Log_i.cpp: + + Reworked checks for the log being full. Just log the record, + the LogRecordStore will return an error if it is full. This + avoids extra conditions in the write hot path. + +Sun May 28 23:10:03 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * docs/releasenotes/index.html: + + Update for changes to the telecom logging service that have been + made over the last few months. + +Sun May 28 16:24:24 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/Valuetype/ValueBase.cpp: + Add explicit dereferences where the ValueFactory_var is used in + equality tests. + +Sun May 28 15:18:19 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * orbsvcs/orbsvcs/Log/Log_i.cpp: + + Removed unused LogRecord variable in write_recordlist(). This + avoids its construction and destruction in the write hot path. + +Sat May 27 22:00:08 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * orbsvcs/orbsvcs/Log/Log_i.cpp: + + Don't invoke reset_capacity_thresholds() if no log records were + actually removed. + +Fri May 26 22:09:09 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/Valuetype/ValueBase.cpp: + Fixed a memory leak introduced by the refactoring of + _tao_unmarshal_pre. The ValueFactory reference was moved into + the unmarshal_pre method directly, and in doing so inadvertently + got its _var status stripped. + +Fri May 26 11:42:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/orbsvcs/Runtime_Scheduler.cpp: + * orbsvcs/orbsvcs/Sched/Reconfig_Scheduler_T.cpp: + Applied workarounds for BCB2006 and BCB6 in release mode, the + code results in the original form in an internal backend + error, reported this to Borland as QC27961. + +Thu May 25 23:12:56 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * tao/TAO_Internal.cpp: + + Fixed logic in open_services() to eliminate an unused local + variable warning (which happened when the TAO debug level + was less than 3) and made many cosmetic changes to the + whole file to bring the code into line with the ACE + style guidelines. + +Thu May 25 03:27:44 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/OctetSeqC.h: + * tao/OctetSeqC.cpp: + * tao/diffs/OctetSeq.diff: + Reverted change from Tue May 16 19:08:49 UTC 2006 Phil Mesnier + <mesnier_p@ociweb.com> + + * tao/Unbounded_Octet_Sequence_T.h: + Moved the equality operations into the octet sequence class, + making them member functions. This seems to resolve the + namespace related problems. + + +Wed May 24 19:43:16 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tests/Portable_Interceptors/Collocated/Service_Context_Manipulation/Server_Task.cpp: + My fix for the servant reference counting used a ServantBase_var + as the owner of the pointer to the servant instance, but that + cannot be used as a target within ACE_NEW_RETURN macros on + windows because the MSVC compiler doesn't appropriately map the + post-assignment equality test. This minor change is to use a + servant type pointer for initialization with ACE_NEW, then + assign the result to a ServantBase_var for reference management. + +Wed May 24 18:27:13 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * orbsvcs/ImplRepo_Service/ImplRepo_Service.mpc: + + Along with the /FORCE:MULTIPLE link option that is added by + MPC specifically for em3, nmake, vc6, and vc71 builds, added + the /INCREMENTAL:NO link option since the above option is + incompatible with incremental linking, and produces warnings + on the platforms where it is in force. + +Wed May 24 16:00:03 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/TAO_Internal.cpp: + Refactored recent change to skip over global parameters after + the first initialization pass. + +Wed May 24 12:40:57 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/EndpointPolicy/IIOPEndpointValue_i.h: + Fixed memory leak. Unlike servants, local objects are not + intrinsically reference counted. Therefore it is still necessary + to explicitly inherit from TAO_RefCount_LocalObject rather than + from CORBA::LocalObject. + +Wed May 24 10:36:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/String_Manager_T.h (out, _retn): + Just set ptr_ to 0 instead of initializing it with an default + string. Now we get the same behaviour as with the old string + manager, fixes runtime memory leaks when the string managers + are used + +Wed May 24 09:12:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Strategies/DIOP_Connector.cpp: + Const improvements + +Wed May 24 09:09:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * docs/performance.html: + Added ACE_NLOGGING=1 as one of the options that can be used to + reduce footprint. Adding this to the footprint build resulted in + a footprint drop of about 10% for the Hello client. + +Wed May 24 08:53:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Bounded_Sequence_CDR_T.h: + * tao/Unbounded_Sequence_CDR_T.h: + Fixed memory leak in the demarshaling of (w)string sequences. + +Wed May 24 03:01:14 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/PortableServer/POAManager.cpp: + * tao/PortableServer/POAManagerFactory.cpp: + + Cleaned up redundant debug messages. + + * tests/Portable_Interceptors/Collocated/Service_Context_Manipulation/Client_Task.cpp: + * tests/Portable_Interceptors/Collocated/Service_Context_Manipulation/Server_Task.cpp: + * tests/Portable_Interceptors/Collocated/Service_Context_Manipulation/interceptors.cpp: + * tests/Portable_Interceptors/Collocated/Service_Context_Manipulation/test.idl: + * tests/Portable_Interceptors/Collocated/Service_Context_Manipulation/test_i.cpp: + + Add (%P|%t) to many debug lines to improve ability to track + thread behavior during collocated calls. Also switched the + servant to using the modern reference counting method of using a + ServantBase_var to manage the local reference to the servant + rather than using an explicit call to _remove_ref() from within + the shutdown operation. Finally, added a slight delay after + orb->run() and before orb->destroy() to avoid thread races. See + bugzilla bug #2538 for more information about this race. + +Tue May 23 16:09:05 UTC 2006 Adam Mitz <mitza@ociweb.com> + + * orbsvcs/orbsvcs/PortableGroup/UIPMC_Acceptor.cpp: + * orbsvcs/orbsvcs/PortableGroup/UIPMC_Connector.cpp: + * orbsvcs/orbsvcs/PortableGroup/UIPMC_Endpoint.cpp: + * orbsvcs/orbsvcs/PortableGroup/UIPMC_Factory.cpp: + * orbsvcs/orbsvcs/PortableGroup/UIPMC_Profile.cpp: + * orbsvcs/orbsvcs/PortableGroup/UIPMC_Transport.cpp: + * tao/ORB_Constants.h: + + See bugzilla #2500. Around the time of TAO 1.4.7 the tags for UIPMC + (multicast) changed from TAO-assigned to OMG-assigned. The + ComponentID and the ProfileID were added to tao/IOP_IORC.h but the + change to the new ProfileID was never completed, since code still + referred to the value in tao/ORB_Constants.h. This change eliminates + the old ProfileID and changes all uses to the new one. This causes a + break in multicast interoperability between applications using TAO + before this change and TAO after this change, but should enable + multicast interoperability between TAO and a different ORB. + +Tue May 23 13:05:43 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/ORB_Core.cpp: + Fixed a typo in a service name. + + * tao/TAO_Internal.cpp: + + Added a function to address the situation where a second ORB is + initialized with arguments that are global and expected to be + removed from argv before ORB_Core::init starts parsing args. + + Since the first initialized ORB is the default ORB, it is also + the one that sets global options, such as debug level, and + daemonization. In situations where multiple ORBs are initialized + indeterminately, such as via dynamically loaded service objects, + an application wanting a concrete set of global options should + explicitly initialize a default ORB. + + * tao/IIOP_Acceptor.cpp: + * tao/PortableServer/Root_POA.cpp: + * tests/CollocationLockup/CollocationLockup.cpp: + * tests/InterOp-Naming/INS_test_client.cpp: + * tests/POA/EndpointPolicy/server.cpp: + Fixed memory leaks. + + +Thu May 18 17:16:30 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * orbsvcs/Logging_Service/ + * orbsvcs/Logging_Service/Basic_Logging_Service/Makefile.am: + * orbsvcs/Logging_Service/Notify_Logging_Service/Makefile.am: + + Regenerated. + + NB: The reason the Makefile.am's for the Event and RTEvent + Logging Services were not changed is because currently the + automake config are generated with the typed event channel + support enabled, which also requires valuetype. + + * orbsvcs/Logging_Service/Basic_Logging_Service/Basic_Logging_Service.mpc: + * orbsvcs/Logging_Service/Event_Logging_Service/Event_Logging_Service.mpc: + * orbsvcs/Logging_Service/Notify_Logging_Service/Notify_Logging_Service.mpc: + * orbsvcs/Logging_Service/RTEvent_Logging_Service/RTEvent_Logging_Service.mpc: + + Changed to inherit from the valuetype base project. + + This is required in the case the event / log record embeds a + valuetype. A event or log record contains one or more CORBA + anys. The current implementation demarshals the any before + touching any app code. Thus if the any were to contain a + valuetype, the valuetype library will be called upon to + demarshal the data. + + The same change was made for the Notification Service in: + Mon Jul 18 13:12:15 2005 Ciju John <john_c@ociweb.com> + + This fixes bugzilla issue #2524. + +Thu May 18 15:10:50 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tests/ORB_Local_Config/run_tests_all.pl: + + Updated to remove the Limits test from the list. The + functionality has been moved to the Service_Config_Test under + ACE. + + * tests/ORB_Local_Config/Limits/Limits.mpc: + * tests/ORB_Local_Config/Limits/Test.cpp: + * tests/ORB_Local_Config/Limits/run_test.pl: + + Removed these files. + +Thu May 18 13:52:58 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tests/ORB_Local_Config/Two_DLL_ORB/server.cpp: + + Replaced servant activation using _this() with one explicitly + referencing the desired POA, using activate_object(). As usual, + using _this() outside the scope of an upcall yields surprising + results, because it is using the first ORB in the orb table, no + matter which one that is. Within a single process, it is + possible for the server's servant to get activated in the client + ORB's POA and then the client will fail to communicate with it, + because the client POA is not active. + +Wed May 17 23:07:10 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/OctetSeqC.h: + * tao/OctetSeqC.cpp: + * tao/diffs/OctetSeq.diff: + + Fuzz removal. + +Wed May 17 19:35:09 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * TAO_IDL/be/be_visitor_valuebox/valuebox_ci.cpp: + + In the generated code for setting the member value, + removed the parentheses from the constructor call in + ACE_NEW, if the member type is a fixed-size IDL struct. + These parens were producing a warning on one of the + scoreboard's VC 7.1 builds, due to a behavior change + in the compiler. This behavior change is to initialize + PODs (for which an IDL fixed struct qualifies) to the + default value without requiring the parens denoting + a default constructor call. + + * tests/Bug_2234_Regression/server.cpp: + * tests/OBV/ValueBox/client.cpp: + + Made changes in hand-written client and server code in TAO/tests + similar to the changes in IDL compiler generated code above. + +Wed May 17 19:09:36 UTC 2006 Yan Dai <dai_y@ociweb.com> + + Merged OCI's changes + "Fri May 12 21:59:41 UTC 2006 Yan Dai <dai_y@ociweb.com>" + + * TAO/tao/Intrusive_Ref_Count_Handle_T.inl: + + Fixed a potential memory leaks in operator==(T*) function. + The memory leak could happen when this assignment operator + is used to assign the same instance. + + * TAO/tao/CSD_ThreadPool/CSD_TP_Task.cpp: + + Made the TP_Dispatchable_Visitor object reset() called after + the request is dispatched. This would avoid the delay deletion + of the request and its referenced objects. + + * tao/CSD_Framework/CSD_FW_Server_Request_Wrapper.cpp: + + Made the transport object in TAO_ServerRequest be reference + counted by the CSD. Increment the reference counter when the + TAO_ServerRequest is cloned, and decrement the reference counter + when the server request is destroyed. This would avoid crash + when the transport object is destroyed but CSD has not finished + dispatching the request. + + Merged OCI's changes + "Thu Apr 20 13:29:44 2006 Ciju John <john_c@ociweb.com>" + + Made an SSLIOP endpoint value of 'iiop://:/ssl_port=xyz' listen + on all available network interfaces instead of listening on a + specific IP address. These changes make the 'iiop://:/ssl_port=xyz' + and 'iiop:///ssl_port=xyz' have same semantics. + + * tao/IIOP_Acceptor.h : + * tao/IIOP_Acceptor.cpp : + + Refactored the address parsing code into a new method + 'parse_address'. + + * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Acceptor.cpp (open): + + Use TAO_IIOP_Acceptor::parse_address() to initialize + ACE_INET_Addr. + +Wed May 17 18:47:22 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tests/AMH_Oneway/server.cpp: + + Fix for an error in the construction of the test. This error is + a result of using a stack based servant, as well as a stack + based helper ORB-running helper class. The main for this test + instantiated the helper on the stack first, then the servant. It + was done in this order so that the servant could obtain a + reference to the helper's ORB. However, stack based objects are + destroyed in reverse order, meaning that the servant instance + was destroyed before the helper's instance. Thus the POA in the + helper ended up with a stale pointer to a prematurely deleted + servant when it came time to do an orderly shutdown. + + There were several options for cleaning this up, the servant + could have been allocated on the stack and have all reference + but the POA's removed, the servant's destructor could have + deactivated itself from the POA, or an explicit cleanup method + could be added to the helper class so the main could force the + proper order of desctruction. I chose this last option since the + main was in charge of determining the order of creation. + +Wed May 17 16:34:44 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tests/AMI_Buffering/AMI_Buffering.h: + + Fix scoping for new nested class used to avoid spurious comm + fail exceptions. This fix resolves build fails that only affect + the BCB compiler. + +Wed May 17 14:11:35 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * TAO_IDL/be/be_visitor_typecode/typecode_defn.cpp: + * TAO_IDL/be_include/be_visitor_typecode/typecode_defn.h: + + Added generation of an anonymous namespace inside the + TAO::TypeCode namespaces already generated, for + typecodes ggenerated for anonymous types (sequences, + arrays, and bounded (w)stringts). This change prevents + a multiple definition link error with a typecode + generated for an identical type in another translation + unit. The exisiting generated ifdef guards prevent + the same error within the same translation unit. + Thanks to Ossmama Othman <ossama.othman@symantec.com> + for suggesting the fix. This fix closes [BUGID:2521]. + + Also removed many lines of commented out code from + this file. + +Wed May 17 12:10:58 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/Bounded_Sequence_CDR_T.h: + + Reverting this change, it didn't work. + Mon May 15 22:25:23 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/Bounded_Sequence_CDR_T.h: + Added explicit include to satisfy the HP compiler. + + +Wed May 17 11:34:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * TAO_IDL/be/be_visitor_traits.cpp: + Added extra newline so that the zero method is on its own line + +Wed May 17 09:48:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/IOR_Endpoint_Hostnames/IOR_Endpoint_Hostnames.mpc: + Simplified this mpc file + +Tue May 16 19:08:49 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * NEWS: + Add decription of new features/fixes. + + * tao/BiDir_GIOP/BiDirGIOP.h: + * tao/CSD_Framework/CSD_Object_Adapter_Factory.h: + * tao/EndpointPolicy/EndpointPolicy.h: + * tao/EndpointPolicy/EndpointPolicy.cpp: + * tao/Messaging/Messaging_Loader.h: + * tao/PI_Server/PI_Server_Loader.h: + * tao/RTScheduling/RTScheduler_Loader.h: + + Fix for certain static builds, notably VC71 on WinXP, for + libraries that depend on the Portable Interceptors library for + registering ORB Initializers. + + * tests/Bug_2417_Regression/publisher_impl.cpp: + + This test runs very long for what it is demonstrating. With the + underlying reference counting problem resolved, the server now + runs to completion. Since the test is using asynch connection + establishment and SYNCH_NONE oneways, there is no way for the + publisher to detect that the subscriber is gone until it makes a + synchronizing twoway call. The period between synch tests was so + long that on a slow machine the test would time out. + + * orbsvcs/orbsvcs/Sched/Reconfig_Scheduler_T.cpp: + * orbsvcs/orbsvcs/Trader/Interpreter_Utils.h: + * tao/OctetSeqC.cpp: + * tao/OctetSeqC.h: + * tao/diffs/OctetSeq.diff: + + Applying fixes to the problems still remaining in the versioned + namespace builds. Not all compilers found the problem in + Reconfig_Scheduler_T.cpp, but gcc 3.3.1 did. The OctetSeq change + might represent a candidate for a change to the IDL compiler, + but for now I think not, as octet sequences are treated as a + special case, having an explicit template instanciation provided + along with explicit equality operators. + +Tue May 16 14:15:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Strategies/Strategies.mpc: + * tests/Bug_2134_Regression/Bug_2134_Regression.mpc: + * tests/Bug_2494_Regression/Bug_2494_Regression.mpc: + Simplified these mpc files + +Tue May 16 14:09:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/orbsvcs/AV/TCP.cpp: + Const improvments and initialise some pointers with 0 + + * orbsvcs/orbsvcs/AV/RTP.cpp: + * orbsvcs/orbsvcs/AV/sfp.cpp: + Use a CORBA::ULong to iterate through the TAO_AV_PolicyList + +Tue May 16 14:03:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/orbsvcs/CosProperty.idl: + Corrected incorrect filename in the header of this file + +Tue May 16 12:36:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Asynch_Queued_Message.cpp: + * tao/Synch_Queued_Message.cpp: + * tao/GIOP_Message_Generator_Parser_10.cpp: + * tao/Object.cpp: + * tao/Profile.cpp: + * tao/IIOP_Profile.cpp: + * tao/Strategies/DIOP_Acceptor.cpp: + * tao/Strategies/DIOP_Endpoint.cpp: + Const improvements + + * tao/GIOP_Message_State.cpp: + Improved error message when the GIOP header can't be parsed + + * tao/ORB.cpp: + Initialise pointer with 0. + + * tao/PortableServer/PolicyS_T.h: + Use true instead of 1 + +Tue May 16 05:22:15 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/IIOP_Connector.cpp: + + This is a potential fix for the Bug 2417 flaw. The problem is + that when using nonblocking connects, in conjunction with asynch + invocations, it is possible for a transport to be returned by + the connector even though the network connection has not + completed. For asynchronous invocations using the SYNCH_NONE + policy, this is appropriate, as request messages may be queued + for delivery if/when the connection completes. + + Bug 2417 describes a scenario where such a nonblocking + connection attempt fails, but the actual failure happens after + the transport has already been returned to the caller. This + causes a problem because the underlying ACE connector framework + relies on "borrowing" the reference to the connection handler + during the time it is waiting for connections to complete or + fail. For blocked connects this is fine because either the + transport will be returned to the caller associated with a + completely established connection, or a failure will occur. + + The issue for nonblocking connects is that when a transport is + returned associated with a pending connection, the existing + transport connector and protocol-specific connector end up + associating to referrers to the same connection handler, without + incrementing the reference count. The two are the transport + being returned and the ACE_NonBlock_Connection_Handler that is + actually registered with the reactor waiting for success or + failure on the pending connection. + + When a connection completes OK, the NBCH surrenders its + reference to the connection handler, thus restoring parity, as + the transport and/or cache entry will still hold the remaining + references, and the count is OK. But when the connection fails, + the base connector ends up calling close() on the connection + handler which in turn decrements the reference count. This then + sets the stage for a later crash from an apparent double delete. + + * tao/IIOP_Connection_Handler.cpp: + * tao/Transport_Connector.cpp: + + Added some comments and cleaned up some whitespace. + +Mon May 15 22:25:23 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/Bounded_Sequence_CDR_T.h: + Added explicit include to satisfy the HP compiler. + +Mon May 15 18:17:23 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/Codeset_Manager_Factory_Base.h: + * tao/Codeset_Manager_Factory_Base.cpp: + * tao/PortableServer/Acceptor_Filter_Factory.h: + * tao/PortableServer/Acceptor_Filter_Factory.cpp: + + Moved the static initializer from inside the .cpp to the header + file. This change was necesitated by the VC7.1 static build that + was apparently skipping over the static initializer if it wasn't + in the .h. + +Mon May 15 13:28:01 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tests/AMI_Buffering/AMI_Buffering.h: + * tests/AMI_Buffering/AMI_Buffering.cpp: + * tests/AMI_Buffering/client.cpp: + + The tests still fail on a lot of machines with a series of + COMM_FAIL exceptions being reported. These are due to a nesting + problem within the server, since it is receiving requests from + the client and also making requests to the admin. What happens + is that with the asynchronous calls, a whole bunch of requests + are sent to the server, and before it gets a chance to receive + all the replies from the admin, the client sends a shutdown. + This immediately closes the server's client-side connection to + the admin, and if any replies were pending, those are lost, + causing the comm fails. I've also added a log message reporting + the maximum nesting level attained by the server. + + The client now delegates responsibility of shutting down the + admin to the server as another way of eliminating any races that + might cause spurious error reports. + +Mon May 15 12:58:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Unbounded_Octet_Sequence_T.h: + Fixed compile problem when TAO_NO_COPY_OCTET_SEQUENCES is defined + to 0 + +Mon May 15 09:59:56 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/Unbounded_Octet_Sequence_T.h: + + A potential solution to the namespace problem. This at least + works for the gcc/linux build. + +Mon May 15 02:50:56 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tests/AMI_Buffering/client.cpp: + * tests/Oneway_Buffering/client.cpp: + + Add the header for defining sleep for the platforms that don't + happen to get it indirectly. + +Sun May 14 13:32:33 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tests/AMI_Buffering/client.cpp: + * tests/Oneway_Buffering/client.cpp: + + Minor change to insert a pause after killing the server but + before killing the admin. This gives the server a moment to + flush out any messages it wants to send to the admin. Without + this pause, the test occasionally reports spurious comm failures + that cause the scoreboard to count the test as failed. + + * tests/Connection_Timeout/client.cpp: + + Added more information to failure output. + +Sat May 13 22:07:04 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/LF_CH_Event.h: + * tao/LF_CH_Event.cpp: + * tao/Transport_Connector.cpp: + + This is a fix for intermittent timeout problems when using + asynch connections. Some tests, most notably the AMI_Buffering + timeout test, randomly fail in that the client seems to just + stop dead. I found that this failure occured when using + nonblocking connects with at least 2 threads, one of which is + invoking ORB::run while the other is trying to connect. + + The problem is that an asynch connection might cause the + transport to cached in an unconnected state, relying on a + subsequent connection request to enter the connection wait + strategy and complete the connection. When using the leader + follower wait strategy, a non-blocking connect will "poll" by + setting the timeout value to be ACE_Time_Value::zero. A race + could occur when the other thread actually handles the + connection completion, but after the interested thread starts to + enter the leader-follower. In this case the timeout of zero + causes the LF to change the connection handler's state to + TIMEOUT, but this was not detected as an error condition. Thus + the LF_Event relating to the connection completion was never + successful or an error and the waiting thread became the leader + and was then stuck. + + This fix works by first treating the TIMEOUT state as an error, + to break out of the LF loop, then the connector will reset the + timeout state if that is appropriate. Finally, a second error in + the connector is fixed where the transport's register_handler + method return value was incorrectly tested. + + * tests/AMI_Buffering/client.cpp: + + Adjusted the timeout values a bit. On a sufficiently fast + computer, it is possible for the sender to overflow the TCP + buffers thus taking too long to flush and spuriously reporting + errors. + +Sat May 13 14:34:23 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tests/ORB_Local_Config/Two_DLL_ORB/server.cpp: + + Removing the use of ACE_OS::unlink() since it is causing + unresolved link errors on windoze. Using alternative IOR file + truncation method to get rid of "stale" IOR files from previous + executions. + +Fri May 12 14:00:17 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tests/ORB_Local_Config/Two_DLL_ORB/client.cpp: + * tests/ORB_Local_Config/Two_DLL_ORB/run_test.pl: + * tests/ORB_Local_Config/Two_DLL_ORB/server.cpp: + + Updated the test to account for the case where the server may + not have completed writing out its IOR, by the time the client + tries to use it to get an object reference. Added ACE_TEXT where + appropriate. + +Thu May 11 21:13:22 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tao/AnyTypeCode/Any_Unknown_IDL_Type.h: + * tao/AnyTypeCode/Any_Unknown_IDL_Type.cpp: + + Having a lock_ as a global static makes this code subject to the + whim of the specific compiler implementation and library + ordering. It is up to the compiler to decide the order in which + our instance is initialized and destroyed. Typically, this + becomes a problem when a code that depends on that instance + finds that the runtime has already destroyed it. The scenario + plays almost always in the process shutdown code, after main() + exits (which is a lot of fun to debug :). The change replaces + the static class member with static-local variable, defined + within a static member function. C++ guarantees that the local + static variable will be initialized at the first method + invocation. + + * tao/ORB_Core.cpp: + + Reversing a changes, introduced by this: + Wed Apr 26 20:21:49 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + I had incorrectly assumed that the same pattern of dynamic + loading of factories applies to all factories. In fact many of + them are optional and their usage is predicated on having a svc + conf file, having their library statically linked. This change + removes the overly aggressive attempts to load such services and + the incorrect usage of TAO_AS_STATIC_LIBS macro.. + +Thu May 11 19:21:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + Reverted the change below, it breaks some gcc builds, have to figure + out another change. + + Thu May 11 13:23:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + * tao/Unbounded_Octet_Sequence_T.h: + Moved equal and not equal operators for the octet sequence to + the TAO namespace, this fixes the build error in the vc71 + namespace build. + +Thu May 11 09:22:10 2006 Douglas C. Schmidt <schmidt@cse.wustl.edu> + + * performance-tests/Throughput/Receiver.cpp (done), + * performance-tests/Throughput/client.cpp (main): Changed + division by 10000000 to division by 1000000. Thanks to + Jason Zhao <jason.zhao at lmco dot com>. + +Thu May 11 13:49:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Bug_1254_Regression/BlobServer.h: + * tests/Bug_1254_Regression/client.cpp: + Fixed casing of includes + +Thu May 11 13:23:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Unbounded_Octet_Sequence_T.h: + Moved equal and not equal operators for the octet sequence to + the TAO namespace, this fixes the build error in the vc71 + namespace build. + + * tests/OBV/Truncatable/client.cpp: + Added missing string_dup calls, fixes crashing of this test + with Borland C++ + +Thu May 11 10:35:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Generic_Sequence_T.h: + Small layout change + +Thu May 11 07:19:12 UTC 2006 Kees van Marle <kvmarle@remedy.nl> + + * tests/Bug_1254_Regression/* + New regression for bug 1254 + + * tao/Unbounded_Octet_Sequence_T.h: + Fixed a bug in the octet sequence when used as inout argument and + when shrinking the lenght while a message block was being used + the full message block was send back, not the smaller length. + We now do a copy of the data to really make sure we don't + modify the mb incorrectly. Thanks to Peter van Merkerk + <Peter dot van dot Merkerk at meco dot nl> and + Marc Walrave <marc dot walrave at meco dot nl> for reporting this. + This fixes bugzilla bug 1254. + +Wed May 10 20:13:45 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/PI/ORBInitializer_Registry_Impl.h: + * tao/PI/ORBInitializer_Registry_Impl.cpp: + * tao/PI/PI.cpp: + + My change from last night (04:26:14 UTC) swung the pendulum too + far back in the other direction. I discovered that once again + some dynamic PI tests were breaking. I've rectified this by + putting Iliyan's code back in ORBInitialiser_Registry_Impl but + with a !TAO_AS_STATIC_LIBS guard around the code to avoid the + redundant processing that might occur with static libs. + +Wed May 10 17:21:26 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/EndpointPolicy/Endpoint_Value_Impl.h: + + I missed committing this with the other versioned namespace + changes. + +Wed May 10 11:49:10 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/PortableServer/Acceptor_Filter_Factory.h: + * tao/PortableServer/POAManagerFactoryC.h: + * tao/PortableServer/POAManagerFactoryC.cpp: + + Cleaning up Versioned namespace related issues. + +Wed May 10 04:26:14 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/ORB_Core.cpp: + + Removed the agressive loading of the IORInterceptor adaptor + factory. It truely is optional, thus its failure to load is not + an error. At least not an error in the ORB core. + + * tao/PI/ORBInitializer_Registry_Impl.h: + * tao/PI/ORBInitializer_Registry_Impl.cpp: + * tao/PI/PI.h: + * tao/PI/PI.cpp: + * tao/PI/PolicyFactory_Loader.h: + * tao/PI/PolicyFactory_Loader.cpp: + + Reverted most of the changes used to resolve a circular + dependency problem with the PI initialization. The problem is + that the initial fix then broke the static builds. The solution + of using TAO_AS_STATIC_BUILDS is a reasonable compromise as it + allows dynamic builds to not get caught in a circular + initialization situation, but lets static builds get the + initialization they need. + +Tue May 9 19:05:30 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * examples/CSD_Strategy/ThreadPool6/CSD_Test_ThreadPool6.mpc: + * examples/CSD_Strategy/ThreadPool6/Makefile.am: + Added -GT to the idlflags + +Tue May 9 16:32:01 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tao/ORB_Core.cpp: + + Added a clarifying comment. + + * tao/TAO_Internal.cpp: + + Fixing a problem in processing -ORBSvcConf command-line + options. Both the process-wide and the ORB-local service + gestalts were processing configuration files. This was causing + the loading of the default svc.conf file, even when another one + had been specified by -ORBSvcConf. The fix is to give the + process-wide gestalt a chance to load the svc conf file, if it + is being initialized for the first time. + + * tests/ORB_Local_Config/Bug_1459/Test.cpp: + + Minor updates. + + * tests/ORB_Local_Config/Two_DLL_ORB/client.cpp: + + Updated the tests to eliminate some possibilities for TRANSIENT + exceptions. + +Tue May 9 11:37:30 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * docs/tutorials/Quoter/idl/Quoter_idl.mpc: + * orbsvcs/orbsvcs/Makefile.am: + * examples/POA/TIE/Makefile.am: + * docs/tutorials/Quoter/idl/Makefile.am: + Add -GT to the idlflags + +Tue May 9 07:46:30 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/String_Alloc.cpp: + Small const fix, simplified CORBA::string_dup a little bit + +Tue May 9 07:36:30 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * TAO_IDL/be/be_global.cpp: + Don't generate the TIE files (_S.*) by default anymore. The + commandline option -Sc has been removed, this suppressed the + generation of these files. The new option -GT has been added, + if you need the TIE files, use the -GT option when compiling + your idl files. This safes disk space on all systems and on + slower machines the builds run then faster. Fixes bug 2525 + + * docs/compiler.html: + Removed -Sc, added -GT + + * NEWS: + Mention the TIE behaviour change + + * examples/POA/TIE/POA_TIE.mpc: + Added -GT to the idlflags + + * tao/*.pidl: + Removed the -Sc flag in the regeneration instructions + +Mon May 8 15:52:31 UTC 2006 Jeff Parsons <j.parsons@vanderbilt.edu> + + * TAO_IDL/be_visitor_union_branch/cdr_op_cs.cpp (visit_enum): + + Added code generation to initialize an enum member of a + union when it is declared for demarshaling. This change + eliminates warnings in gcc 4.0.2 and possibly other + compilers. + +Mon May 8 10:06:12 UTC 2006 Martin Corino <mcorino@remedy.nl> + + * docs/Options.html: + * docs/ORBEndpoint.html: + * docs/INS.html: + Added IPv6 specific info regarding endpoint and corbaloc + definitions. + Added some links for easier crossreferencing. + +Mon May 8 09:22:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Hello/run_test.pl: + Added support for a -debug commandline arugment, if this is + passed when starting this script then -ORBDebugLevel 10 is + passed to client and server. + +Sun May 7 21:39:30 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tests/ORB_Local_Config/Bunch/Test.cpp: + + Removed a misplaced semicolon. + +Sun May 7 14:56:20 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tests/ORB_Local_Config/Bunch/Test.cpp: + + Removed semicolons after if() - cleaning the residue from + replacing the ACE_ASSERTs. + +Sun May 7 06:59:54 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tao/ORB_Core.cpp: + + Replaced 'new' with ACE_NEW. + + * tests/ORB_Local_Config/Service_Dependency/Test.cpp: + * tests/ORB_Local_Config/Shared/Test.cpp: + * tests/ORB_Local_Config/Two_DLL_ORB/run_test.pl: + * tests/ORB_Local_Config/lib/Service_Configuration_Per_ORB.h: + + Minor edits and cleanup. + +Fri Apr 28 10:47:26 2006 Douglas C. Schmidt <schmidt@cse.wustl.edu> + + * tao/Stub.cpp: Added a #include of "tao/CDR.h" to support SunC++. + Thanks to Vladimir Panov <gbr at voidland dot org> for reporting + this. + +Fri Apr 14 17:47:18 2006 Douglas C. Schmidt <schmidt@cse.wustl.edu> + + * docs/releasenotes/index.html: Updated the documentation to + include more pluggable protocols. Thanks to Willie Chen + <wchen12 at ucla dot edu> for motivating this. + +Fri May 5 18:48:45 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/TAO_Internal.cpp: + * tao/default_resource.cpp: + In addtion to the reorganization Iliyan did, it was discovered + that the codeset library was not getting properly + initialized. When the codeset negotiation feature was made + optional for footprint considerations, the loading of the + codeset library was put into the default Resource + Factory. However now that we have the potential for multiple + service repositories, it was no longer sufficient do such late + initialization of the codeset library. Doing so put it codeset + manager and default translators in the configuration context of + the first ORB, not in the global configuration context. Moving + the bootstrap loading of the Codeset library to TAO internals + resolves that. The -ORBNegotiateCodeset flag is still evaluated + making codeset loading optional. It will be loaded to the global + configuration context by the first ORB that needs it. + +Fri May 5 18:05:04 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tests/ORB_Local_Config/Bug_1459/Test.cpp: + * tests/ORB_Local_Config/Bunch/Test.cpp: + * tests/ORB_Local_Config/Limits/Test.cpp: + * tests/ORB_Local_Config/Separation/Test.cpp: + * tests/ORB_Local_Config/Service_Dependency/Test.cpp: + * tests/ORB_Local_Config/Shared/Test.cpp: + * tests/ORB_Local_Config/Simple/Test.cpp: + * tests/ORB_Local_Config/Two_DLL_ORB/Test.cpp: + + Eliminating the use of ACE_ASSERT and replacing with if's and + ACE_ERROR. The ACE_ASSERT may cause an abort(), which will cause + resources to not be cleaned correctly on embedded targets like + VxWorks. Thanks to Johnny Willemsen <jwillemsen@remedy.nl> for + clarifying this. + + * tests/ORB_Local_Config/lib/Service_Configuration_Per_ORB.h: + + A little cleanup. + +Fri May 5 16:12:17 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tao/ORB_Core.cpp: + * tao/TAO_Internal.cpp: + + The initialization of the additional services, which runs right + after the global repository initialization (open()) was supposed + to execute in the context of that same global repository. This + is necessary in case the initialization causes additional + services to be registered. Fixes a problem with + $TAO_ROOT/tests/RTCORBA/ORB_init. + +Fri May 5 11:26:42 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Asynch_Queued_Message.{h,cpp}: + * tao/Synch_Queued_Message.{h,cpp}: + Changed is_heap_allocated to a real bool so that it matches + the base class. + + * tao/Unbounded_Octet_Sequence_T.h: + Small const improvment + + * tao/GIOP_Message_Base.cpp: + Fixed typo in comment + + * tao/DynamicInterface/Request.h: + Removed commented out method + +Thu May 4 16:22:42 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tests/ORB_Local_Config/Two_DLL_ORB/ORB_DLL_Export.h: + + Regenerated the file to fix a problem with building the test in + static builds. + +Thu May 4 13:36:00 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tests/ORB_Local_Config/Service_Dependency/Test.cpp: + + Eliminated a "magic" constant, used for the number of expected + services, because it can vary dependent on the particular TAO + configuration, like mincorba, static, etc. + +Thu May 4 08:53:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/PI/ClientRequestInfo.cpp: + Fixed bug 2510 in such a way that TAO doesn't crash but it seems + we then miss some functionalify. The reporter will extend the test + to detect the missing functionality. Thanks to Martin Cornelius + <Martin dot Cornelius at smiths-heimann dot com> + for reporting this bug and providing a regression test + +Thu May 4 07:54:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/EndpointPolicy/Endpoint_Acceptor_Filter.cpp: + Fixed warning in VxWorks 5.5.1 builds + +Thu May 4 00:45:45 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * tao/Makefile.am: + + Removed EndpointPolicy/EndpointPolicyC.inl. + +Wed May 3 21:26:07 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/TAO_Internal.cpp: + The order of initialization of static services was modified as + multi-orb configuration feature implementation. Somehow during + that refactoring some services that are required to be initialized + ended up having that done before the svc.conf file is processed. + For instance RT_ORB_Loader. This change breaks up the loading of + and initializing of these services to ensure that svc.conf always + gets processed first before doing any default initialization. + +Wed May 3 19:14:12 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tests/ORB_Local_Config/Two_DLL_ORB/Two_DLL_ORB.mpc: + + Updated to resolve buld failures (vc71+Windows) - the two DLLs + produced, now contain identical code. It determines at runtime, + whether to act as a client or as a server - a decision + influenced by a command-line option, in the service + configuration file. + +Wed May 3 18:32:29 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tests/POA/EndpointPolicy/run_test.pl: + The test was failing on Windows because TAO apparently gets + built with -ORBDottedDecimalAddresses defaulted to 1. The + current version of the endpoint policy is very simple in that it + uses literal string comparisons for determining the suitability + of a given candidate profile endpoint. The test is specificly + trying to match "localhost" but fails when the server + substitutes "127.0.0.1" Perhaps an alternative solution would be + to test both literal strings. + +Wed May 3 16:54:41 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * TAO_IDL/be/be_visitor_sequence/serializer_op_cs.cpp: + + Fix an apparent typo introduced the previous commit. + +Wed May 3 16:06:56 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tests/ORB_Local_Config/Bug_1459/Bug_1459.mpc: + * tests/ORB_Local_Config/Bunch/Bunch.mpc: + * tests/ORB_Local_Config/Limits/Limits.mpc: + * tests/ORB_Local_Config/Separation/Separation.mpc: + * tests/ORB_Local_Config/Service_Dependency/Service_Dependency.mpc: + * tests/ORB_Local_Config/Shared/Shared.mpc: + * tests/ORB_Local_Config/Simple/Simple.mpc: + * tests/ORB_Local_Config/Two_DLL_ORB/Two_DLL_ORB.mpc: + + Removed the dependency on ACE test_output library to make it + possible to build the TAO tests without having to build ACE + tests. This is often the case on embedded platforms (VxWorks), + where part of the tests run outside the host platform. Thanks to + Johnny Willemsen for pointing that out. + + * tests/ORB_Local_Config/Bug_1459/Test.cpp: + * tests/ORB_Local_Config/Bunch/Test.cpp: + * tests/ORB_Local_Config/Limits/Test.cpp: + * tests/ORB_Local_Config/Separation/Test.cpp: + * tests/ORB_Local_Config/Service_Dependency/Test.cpp: + * tests/ORB_Local_Config/Shared/Test.cpp: + * tests/ORB_Local_Config/Simple/Test.cpp: + * tests/ORB_Local_Config/Two_DLL_ORB/ORB_DLL.cpp: + * tests/ORB_Local_Config/Two_DLL_ORB/Test.cpp: + + Updated the code to use its own ACE_MAIN and fixed some typos. + +Wed May 3 15:42:14 UTC 2006 Yan Dai <dai_y@ociweb.com> + + * TAO_IDL/be/be_visitor_sequence/serializer_op_cs.cpp: + + Fixed the DDS compilation errors due to the recent + unbounded string sequence implementation changes. + +Wed May 3 13:03:03 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * tao/Makefile.am: + + Regenerated. + + * tao/tao.mpc: + + Fix yet another typo. + +Wed May 3 09:56:10 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Portable_Interceptors/Bug_2510_Regression/client.cpp: + At the end of the test shutdown the server + +Wed May 3 03:35:10 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/LF_Multi_Event.h: + * tao/LF_Multi_Event.cpp: + * tao/PortableServer/POAManagerFactory.h: + * tao/PortableServer/POAManagerFactory.cpp: + * tao/PortableServer/POAManagerFactoryC.h: + Adding in Versioned namespace macros. + +Tue May 2 22:01:51 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tests/ORB_Local_Config/Two_DLL_ORB/Two_DLL_ORB.mpc: + Added the portableserver base project to both the client lib and + server lib projects, since both require client and server + behavior. This is required for windows builds where apparently + DLLs must have all symbols fully resolved at link time, unlike + .so's where they only need to be satisfied at runtime. + +Tue May 2 19:08:18 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tests/ORB_Local_Config/Two_DLL_ORB/ORB_DLL.h: + + Changing auto_ptr to ACE_Auto_Ptr, because some platforms + (WinXP64_Intel90_64bit) provide an auto_ptr<> without the reset + method. + +Tue May 2 15:30:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Sequence_Unit_Tests/bounded_sequence_cdr_ut.cpp: + * tests/Sequence_Unit_Tests/unbounded_sequence_cdr_ut.cpp: + Added missing includes. Thanks to Carlos O'Ryan for + reporting this + +Tue May 2 13:24:26 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/tao.mpc: + Reverted J.T.'s change from Mon May 1 20:03:59 UTC 2006 + to flush out the cvs conflict markers and reapplied just his + typo fix. + +Tue May 2 11:47:55 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tests/ORB_Local_Config/Separation/Test.cpp: + * tests/ORB_Local_Config/Shared/Test.cpp: + * tests/ORB_Local_Config/Simple/Test.cpp: + + The test don't need more specialized type than the base + ACE_Service_Object, in order to demonstrate the intent. + +Tue May 2 08:05:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * TAO_IDL/driver/drv_preproc.cpp: + When we can't remove the input or output file use %p to print + the error so that the info from the OS why this couldn't be + done is also reported to the user. + +Tue May 2 03:59:17 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/IIOP_Connector.h: + * tao/IIOP_Connector.cpp: + * tao/Transport_Descriptor_Interface.h: + * tao/Transport_Descriptor_Interface.cpp: + + This fix resolves a problem exposed by the + performance-tests/RTCorba/Oneway/Reliable test. That failure was + showing up as a pure virtual function call, but the root cause + was related to the changes to the IIOP_Connector in how it + cached new transports. I was creating a new transport descriptor + out of the successful endpoint, but using creating a new + Base_Transport_Descriptor rather than reusing the supplied + transport descriptor. This broke RT tests in that subsequent + invocations would fail to find the cached endpoint since the + type was different. I've resolved that by adding the new + reset_endpoint method on the Transport_Descriptor_Interface + which allows the transport connector to set a new endpoint to + the existing transport descriptor before caching the value. + +Mon May 1 21:54:05 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tests/ORB_Local_Config/Bunch/Test.cpp: + + Modified the test not to require the ability to fully + instantiate TAO_CORBANAME_Parser, etc. which is a class in + another library and it is not declared so that it is "visible" + outside of it. The test does not require that in order to be + functional - using the base class ACE_Service_Object. + + * tests/ORB_Local_Config/Two_DLL_ORB/client.cpp: + + Added some more logging. + +Mon May 1 20:03:59 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * tao/Makefile.am: + + Regenerated. + + * tao/tao.mpc: + + Fix typo. + +Mon May 1 19:39:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/OBV/Simple/OBV_Simple.mpc: + This test isn't dependent on minimum_corba + +Mon May 1 18:55:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/AnyTypeCode/TypeCode.{h,cpp}: + Made type TypeCode_ptr argument of operator << const so that + the signature is the same as declared in DynamicA.h + +Mon May 1 15:33:00 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tests/OBV/Truncatable/client.cpp: + This is an improvement to what I committed earlier. Since the + valuetypes are passed by value, it is perfectly reasonable to + allocate them on the stack locally. That way I could get away + from the use of the var to store a reference to the local value + instance, and also avoid the awkward initialization code used to + avoid the ambiguity BCB6 was complaining about. + + * tests/POA/POAManagerFactory/POAManagerFactory.cpp: + Changed the name of an internal catch value. I'm not sure, but I + suspect this is what was causing the BCB compiler to complain at + the point of ACE_CATCHANY. + +Mon May 1 12:04:17 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tests/OBV/Truncatable/client.cpp: + * tests/POA/POAManagerFactory/POAManagerFactory.cpp: + The Borland BCB6 compiler has trouble with var types and const + vs. non-const assignment or comparisons. These changes are an + attempt to address this trouble. I don't like the nature of + these changes as they are moving away from the built-in type + safety C++ is supposed to provide. + +Mon May 1 03:43:57 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tests/Parallel_Connect_Strategy/run_test.pl: + The fix for this test was to ensure the CORBALOC parser put all + the listed endpoints into a single Profile. + +Sun Apr 30 22:53:37 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/ORB_Core.cpp: + Somehow when I applied my patch for the alternate connection + timeout hook, used allow either or both the AMI connection + timeout policy or the optimized connection endpoint selector to + set the connection timeout hook, I managed to apply my change to + the relative round trip timeout hook. This change addresses that + and fixes the AMI_Timeout test. + +Sun Apr 30 20:24:39 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tests/ORB_Local_Config/Bug_1459/Bug_1459.mpc: + * tests/ORB_Local_Config/Bunch/Bunch.mpc: + * tests/ORB_Local_Config/Limits/Limits.mpc: + * tests/ORB_Local_Config/Separation/Separation.mpc: + * tests/ORB_Local_Config/Service_Dependency/Service_Dependency.mpc: + * tests/ORB_Local_Config/Service_Dependency/Test.cpp: + * tests/ORB_Local_Config/Shared/Shared.mpc: + * tests/ORB_Local_Config/Simple/Simple.mpc: + * tests/ORB_Local_Config/Two_DLL_ORB/Test.cpp: + * tests/ORB_Local_Config/Two_DLL_ORB/Two_DLL_ORB.mpc: + * tests/ORB_Local_Config/lib/Service_Configuration_Per_ORB.h: + + Updated to simplify the tests and the build process + + * tests/ORB_Local_Config/lib/Service_Configuration_Per_ORB.cpp: + * tests/ORB_Local_Config/lib/lib.mpc: + + Removed these files. + +Sun Apr 30 15:36:21 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * examples/CSD_Strategy/ThreadPool4/ClientTask.cpp: + * examples/CSD_Strategy/ThreadPool5/ClientTask.cpp: + Fixed the assignment to the sequence to hand a buffer the + sequence can properly release. + +Sun Apr 30 14:24:20 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * TAO/tao/Profile_Transport_Resolver.cpp: + Removed invalid semicolons. + + * TAO/tests/ORB_Local_Config/Two_DLL_ORB/Two_DLL_ORB.mpc: + Fixed the dynamic flag for windows builds. + +Sun Apr 30 04:54:25 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tests/ORB_Local_Config/Bug_1459/Test.cpp: + + Cleaned up the test a bit to clarify the intent. + +Sun Apr 30 00:47:10 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tests/ORB_Local_Config/Bug_1459/Test.cpp: + Undoing the damage caused by my too-hasty previous checkin. + + * examples/CSD_Strategy/ThreadPool4/ClientTask.cpp: + * examples/CSD_Strategy/ThreadPool5/ClientTask.cpp: + Addressing some more compiler warnings + +Sat Apr 29 15:34:15 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/PortableServer/POAManagerFactory.cpp: + + * tests/ORB_Local_Config/Bug_1459/Test.cpp: + + * tests/ORB_Local_Config/Service_Dependency/Service_Config_Test.conf: + * tests/ORB_Local_Config/Service_Dependency/Test.cpp: + + * tests/ORB_Local_Config/Two_DLL_ORB/ORB_DLL.h: + * tests/ORB_Local_Config/Two_DLL_ORB/ORB_DLL.cpp: + * tests/ORB_Local_Config/Two_DLL_ORB/client.cpp: + * tests/ORB_Local_Config/Two_DLL_ORB/server.cpp: + + * tests/ORT/ORT_test_IORInterceptor.cpp: + + * tests/POA/EndpointPolicy/server.cpp: + + Fixes for various build-specific errors/warnings. + +Sat Apr 29 14:17:32 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tests/ORB_Local_Config/Bug_1459/Test.cpp: + + Removed references to <iostream> and unreferenced variables. + +Sat Apr 29 13:13:20 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tests/ORB_Local_Config/Bug_1459/Test.cpp: + * tests/ORB_Local_Config/Service_Dependency/Test.cpp: + Fuzz cleanup. + +Sat Apr 29 02:26:49 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/EndpointPolicy/EndpointPolicy_i.cpp: + Removed the apparently redundant ACE_NESTED_CLASS macros. + +Fri Apr 28 22:25:57 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tests/ORB_Local_Config/Bunch/Test.cpp: + * tests/ORB_Local_Config/Limits/Test.cpp: + * tests/ORB_Local_Config/Separation/Test.cpp: + * tests/ORB_Local_Config/Service_Dependency/Test.cpp: + * tests/ORB_Local_Config/Shared/Test.cpp: + * tests/ORB_Local_Config/Simple/Test.cpp: + * tests/ORB_Local_Config/Two_DLL_ORB/Test.cpp: + + Modified to eliminate warnings about unused variables on + ACE_NDEBUG builds. + +Fri Apr 28 22:04:50 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tests/ORB_Local_Config/Bug_1459/Bug_1459.mpc: + * tests/ORB_Local_Config/Bug_1459/README: + * tests/ORB_Local_Config/Bug_1459/Test.cpp: + * tests/ORB_Local_Config/Bug_1459/a.conf: + * tests/ORB_Local_Config/Bug_1459/b.conf: + * tests/ORB_Local_Config/Bug_1459/m.conf: + * tests/ORB_Local_Config/Bug_1459/m1.conf: + * tests/ORB_Local_Config/Bug_1459/run_test.pl: + * tests/ORB_Local_Config/Bug_1459/server_cert.pem: + * tests/ORB_Local_Config/Bug_1459/server_key.pem: + + Added a test I borrowed from bugzilla 1459. It tests the ability + to have two differently configured ORBs in the same process. + + * examples/Simple/time-date/Time_Date.cpp: + + Provided an ID for the ORB, initialized in a code that was + loaded from the DLL. The ORB-specific Service Repo changes + eliminated the need to make TAO_Singleton_Manager not register + with the Object Manager, when initialized from a DLL. + +Fri Apr 28 15:14:04 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tests/ORB_Local_Config/Bunch/Service_Config_Test.UTF-16.conf: + * tests/ORB_Local_Config/Bunch/Service_Config_Test.UTF-16.conf.xml: + * tests/ORB_Local_Config/Bunch/Service_Config_Test.WCHAR_T.conf: + * tests/ORB_Local_Config/Bunch/Service_Config_Test.WCHAR_T.conf.xml: + * tests/ORB_Local_Config/Bunch/Service_Config_Test.conf: + * tests/ORB_Local_Config/Bunch/Service_Config_Test.conf.xml: + + Moved these files to Service_Dependency, which also received the + code that uses these. + + * tests/ORB_Local_Config/Service_Dependency/Service_Config_Test.UTF-16.conf: + * tests/ORB_Local_Config/Service_Dependency/Service_Config_Test.UTF-16.conf.xml: + * tests/ORB_Local_Config/Service_Dependency/Service_Config_Test.WCHAR_T.conf: + * tests/ORB_Local_Config/Service_Dependency/Service_Config_Test.WCHAR_T.conf.xml: + * tests/ORB_Local_Config/Service_Dependency/Service_Config_Test.conf: + * tests/ORB_Local_Config/Service_Dependency/Service_Config_Test.conf.xml: + * tests/ORB_Local_Config/Service_Dependency/Test.cpp: + + Moved here the config files from the Bunch test as it may be run + in single threaded builds and the code that uses these config + files requires multiple threads. + +Fri Apr 28 14:51:20 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tao/TAO_Internal.cpp: + + Fixed an error, that prevented the initialization of static + services during Service_Config::open, The override for the + default argument (ignore_static_svcs = 1) was missed during the + refactoring. + +Fri Apr 28 13:46:43 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tests/ORB_Local_Config/Bunch/Test.cpp: + * tests/ORB_Local_Config/Limits/Test.cpp: + * tests/ORB_Local_Config/Service_Dependency/Test.cpp: + * tests/ORB_Local_Config/Shared/Test.cpp: + * tests/ORB_Local_Config/Simple/Test.cpp: + * tests/ORB_Local_Config/Two_DLL_ORB/Test.cpp: + + Updated to shut up unused variable warnings on builds where + ACE_NDEBUG has been defined. + +Thu Apr 27 21:09:51 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/Profile.cpp: + Fix a problem that caused the endpoint selector to spin when + using shared profiles and none of the endpoints were valid. + +Thu Apr 27 15:03:08 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * tests/ORB_Local_Config/Bunch/Test.cpp: + * tests/ORB_Local_Config/Limits/Test.cpp: + * tests/ORB_Local_Config/Simple/Test.cpp: + * tests/ORB_Local_Config/Two_DLL_ORB/Test.cpp: + + Fixed fuzz issues (unmatched ACE_TRACE) + + * tests/ORB_Local_Config/Service_Dependency/Service_Config_DLL.cpp: + * tests/ORB_Local_Config/Service_Dependency/Service_Dependency.mpc: + + Modified to change the DLL name to avoid name conflict with the + one in ACE_ROOT/tests. + +Thu Apr 27 14:20:02 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/IIOP_Endpoint.cpp: + Fixed a recursion problem with the IPv6 endpoint selection. + + * tao/PortableServer/POAManagerFactory.cpp: + Cleaned up compiler warnings. + +Thu Apr 27 06:55:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/ORB_Local_Config/lib/lib.mpc: + Added base project, this would at least get our builds running again + +Thu Apr 27 03:46:34 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/EndpointPolicy/Endpoint_Acceptor_Filter.cpp: + + Found the trouble that was crashing the endpoint test. It was + nothing wierd with the new sequence code, it was merely an + incorrect index variable - i should have been j. + +Thu Apr 27 03:05:31 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/params.cpp: + + One of my patches inadvertently flipped the sense of the + share_profiles default from 0 to 1. However, I think that 1 + should be the default, since it generates more compact IORs. It + has been that way for years in the OCI version of TAO, but I am + keeping the default 0 here for tradition. This value is + overridden by using -ORBUseSharedProfile [0|1] ORB_init option. + +Thu Apr 27 02:58:23 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/EndpointPolicy/Endpoint_Value_Impl.cpp: + * tao/EndpointPolicy/Endpoint_Value_Impl.h: + + Added a virtual destructor to this otherwise abstract base + class. + +Wed Apr 26 21:24:54 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + * NEWS: + + Added an entry to the NEWS file. + +Wed Apr 26 20:21:49 UTC 2006 Iliyan Jeliazkov <iliyan@ociweb.com> + + The motivation for these changes was to enable support for + ORB-local Service Objects. This for instance, makes it possible + for differently configured ORBs to coexist within the same + proces. + + In order to accomplish this, each orb (core) owns a "Gestalt", + i.e. a service object repository instance. There is also a + process-wide, or "global" gestalt, which is the default + repository where service objects are registered. The latter + retains the interface and behavioral compatibility with any + existing code. As a consequence of this design choice, any + un-named orb(s) will default to using the ubergestalt, which is + consistent with the prior behavior. + + * tao/DLL_Parser.cpp: + + Fixed the parse_string() method to use the correct ORB gestalt + when looking up a dynamic service object. That resolves a + failure in TAO/tests/Object_Loader test. + + * tao/ORB.cpp: + + Relocated some aging comments about having first to instantiate + the singleton manager to the correct place (ORB_init). Added a + gestalt parameter to the call to open_services. Edited a few + lines to fit within the standard length requirement. + + * tao/ORB_Core.h: + * tao/ORB_Core.i: + * tao/ORB_Core.cpp: + + Added a member and an accessors for the private service gestalt, + owned by the core. Replaced the call to methods that use the + implicit ubergestalt with ones that specify the gestalt to use + for service objects. + + Added an ACE_Service_Config_Guard to make sure the ORB + initialization uses the correct repository. + + Updated more references to process_directive() and instance() to + use ACE_TCHAR. Updated references to process_directive() to use + ACE_TCHAR for consistency. + + Added #if !defined(TAO_AS_STATIC_LIBS)/#endif around code, which + is only meaningful when TAO is _not_ statically compiled; + Updated the service and DLL symbol names used to load the + CodecFactory_Loader, PolicyFactory_Loader and + TypeCodeFactory_Loader dynamic services, in the cases where + those services are not statically linked and TAO supports + dynamic linking. (Thanks Ossama, for pointing that out.) Added + code to try and explicitly load an IORInterceptor adapter and + Concrete_IORInterceptor_Adapter_Factory, if TAO supports (is + built with) dynamic linking; + + Updated the code that loads the IORTable adapter to be exception + safe; Edited a few lines to fit within the standard length + requirement. + + * tao/Parser_Registry.cpp: + + Updated to explicitly specify the correct gestalt for the ORB. + + * tao/TAO_Internal.h: + * tao/TAO_Internal.cpp: + + Refactored the initialization code to separate process-wide + aspects of initialization from those having to do with the ORB + instance. It is necessary to deal with global initialization + because of the large number of use cases, where the first thing + a process does is to call ORB_init, and consequently - + open_services. There are also cases where a process calls + Service_Config::open, initializing the process-wide + configuration and only then proceeds to call ORB_init - for + example when using Service Configurator to load a DLL that uses + an ORB. The close_service is now only responsible for calling + close in the ORB's own gestalt, the ACE Object Manager is the + one that is clobering the process-wide Service Configuration. + Updated to explicitly specify the correct gestalt to be used. + + * tao/default_resource.h: + * tao/default_resource.cpp: + + Added the ACE_Dynamic_Service_Dependency member to the default + resource factory to expressly maintain the factory's dependance + on TAO_Codeset library, because the order of destruction may be + reversed in some cases. The member help us keep our access to + TAO_Codeset_Manager instances by upping the ref count on + TAO_Codeset's DLL. This is far from elegant, but a complete + reference counting scheme for the ORB services is a more complex + undertaking than what the available resources currently permit. + + * tao/CSD_ThreadPool/CSD_TP_Strategy_Factory.cpp: + + Fixed an (unrelated) issue arising from a call to strcmp() with + two different character types - only visible when ACE_USES_WCHAR + is in effect. + + * tao/Codeset/Codeset_Manager_i.h: + * tao/Codeset/Codeset_Manager_i.cpp: + + (minor) Added void as argument to the ctor and dtor. + + * tao/PI/ORBInitializer_Registry_Impl.h: + * tao/PI/ORBInitializer_Registry_Impl.cpp: + + Implemented an init() method, which registers all the static + services, usually taken for granted with the loading of + TAO_PI. Previously, static initializers were used, however the + dependent static services were being registered only globally, + which broke the ORBs that needed ORB-local services. + + * tao/PI/PI.h: + * tao/PI/PI.cpp: + * tao/PI/PolicyFactory_Loader.h: + * tao/PI/PolicyFactory_Loader.cpp: + + Removed the static initializers code and made it part of the + dynamic service's init method. See the comment above. + + * tao/PortableServer/Root_POA.cpp: + + Explicitly specified the gestalt to be used for registering + dynamic services. + + * tests/DLL_ORB/Test_Client_Module.cpp: + * tests/DLL_ORB/Test_Server_Module.cpp: + + Provided an ID for the client and server's ORB. In the future, + an option may be devised so that the user can specify if they + want any ORB to use its own gestalt, even if it does not have an + ID. The reverse would be to force all ORBs to use the global SR, + even if they have an ID. Fixed a a SEGV upon process + termination. The first thing a client process does in its main() + is to load a dynamic service - Test_Client_Module, using a call + to ACE_Service_Config::process_directive(). The service does + call ORB_init(), which causes the population of the SR with a + number of static and dynamic SOs. At process termination now + however, any services registered following the ORB_init () call + are destroyed first and will be unavailable when the + Test_Client_Module is finalized. Like the Resource Factory, for + example. + + The solution is to provide and ORB id for any ORB, which will + loaded as part of a dynamic service. Since the service gestalt + is tied to the ORB id, this will cause the new ORBs to create + and manage the lifetime of their own Service Repositories. The + ORB_init() will be invoked in the context of each distinct SR + and any SO an ORB needs will go there. At process termination, + the Test_Client_Module will be finalized, which will clobber the + ORB's SR and any SO registered there. + + * tests/ORB_Local_Config/ORB_Local_Config.mwc: + * tests/ORB_Local_Config/README: + * tests/ORB_Local_Config/run_tests_all.pl: + + Added tests and examples of the functionality affected by the + introduction of the multiple private (per-ORB) service + configuration repositories. + + * tests/ORB_Local_Config/lib/Service_Configuration_Per_ORB.h: + * tests/ORB_Local_Config/lib/Service_Configuration_Per_ORB.cpp: + * tests/ORB_Local_Config/lib/lib.mpc: + + Common test code. + + * tests/ORB_Local_Config/Bunch/Bunch.mpc: + * tests/ORB_Local_Config/Bunch/Service_Config_Test.UTF-16.conf: + * tests/ORB_Local_Config/Bunch/Service_Config_Test.UTF-16.conf.xml: + * tests/ORB_Local_Config/Bunch/Service_Config_Test.WCHAR_T.conf: + * tests/ORB_Local_Config/Bunch/Service_Config_Test.WCHAR_T.conf.xml: + * tests/ORB_Local_Config/Bunch/Service_Config_Test.conf: + * tests/ORB_Local_Config/Bunch/Service_Config_Test.conf.xml: + * tests/ORB_Local_Config/Bunch/Test.cpp: + * tests/ORB_Local_Config/Bunch/run_test.pl: + + A collection of miscellaneous tests for compatibility of the new + interfaces with the old; Processing of the command-line + directives; Loading dynamic services in a local repository; + Loading the ORBInitializer_Registry locally; Test the helper + components used to implement the temporary substitution of the + repository currently used as "global" for the sake of + registering static services, which are dependent on a dynamic + service; + + * tests/ORB_Local_Config/Limits/Limits.mpc: + * tests/ORB_Local_Config/Limits/Test.cpp: + * tests/ORB_Local_Config/Limits/run_test.pl: + + Testing the size limits of a gestalt. + + * tests/ORB_Local_Config/Separation/Separation.mpc: + * tests/ORB_Local_Config/Separation/Test.cpp: + * tests/ORB_Local_Config/Separation/run_test.pl: + + Services registered with separate repositories must remain + separate and inaccessible through anyone but the gestalt they + were registered with. + + * tests/ORB_Local_Config/Service_Dependency/Service_Config_DLL.h: + * tests/ORB_Local_Config/Service_Dependency/Service_Config_DLL.cpp: + * tests/ORB_Local_Config/Service_Dependency/Service_Config_DLL_Export.h: + * tests/ORB_Local_Config/Service_Dependency/Service_Dependency.mpc: + * tests/ORB_Local_Config/Service_Dependency/Test.cpp: + * tests/ORB_Local_Config/Service_Dependency/run_test.pl: + + Tests the working of the ACE_Dynamic_Service_Dependency class + + * tests/ORB_Local_Config/Shared/Shared.mpc: + * tests/ORB_Local_Config/Shared/Test.cpp: + * tests/ORB_Local_Config/Shared/run_test.pl: + + Test that the default repository is available through any + Service Gestalt, created with its default ctor. + + * tests/ORB_Local_Config/Simple/Simple.mpc: + * tests/ORB_Local_Config/Simple/Test.cpp: + * tests/ORB_Local_Config/Simple/run_test.pl: + + * tests/ORB_Local_Config/Two_DLL_ORB/ORB_DLL.h: + * tests/ORB_Local_Config/Two_DLL_ORB/ORB_DLL.cpp: + * tests/ORB_Local_Config/Two_DLL_ORB/ORB_DLL_Export.h: + * tests/ORB_Local_Config/Two_DLL_ORB/Service_Config_ORB_Test.conf: + * tests/ORB_Local_Config/Two_DLL_ORB/Service_Config_ORB_Test2.conf: + * tests/ORB_Local_Config/Two_DLL_ORB/Test.idl: + * tests/ORB_Local_Config/Two_DLL_ORB/Test.cpp: + * tests/ORB_Local_Config/Two_DLL_ORB/Test_i.h: + * tests/ORB_Local_Config/Two_DLL_ORB/Test_i.cpp: + * tests/ORB_Local_Config/Two_DLL_ORB/Two_DLL_ORB.mpc: + * tests/ORB_Local_Config/Two_DLL_ORB/client.cpp: + * tests/ORB_Local_Config/Two_DLL_ORB/run_test.pl: + * tests/ORB_Local_Config/Two_DLL_ORB/server.cpp: + + Testing the loading a dynamic service, which initializes its own + ORB. The test is a variant of the Hello test with the twist that + both the client and the server are service objects, loaded by + the Service Configuration mechanism. + +Wed Apr 26 20:09:33 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/EndpointPolicy/EndpointPolicyC.h: + * tao/EndpointPolicy/EndpointPolicyC.cpp: + + Fixed fuzz errors. + + * tao/EndpointPolicy/EndpointPolicyC.inl: + + Removed this file. + +Wed Apr 26 19:44:36 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * tao/Makefile.am: + + Regenerated. + + * tao/EndpointPolicy.mpc: + + Added Pkgconfig_Files definition. + + * tao/EndpointPolicy/TAO_EndpointPolicy.pc.in: + + New file, pkg-config *.pc template for TAO_EndpointPolicy + library. + +Wed Apr 26 19:08:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/OBV/Simple/Client_i.h: + Fixed casing of include + +Wed Apr 26 19:13:02 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/IIOP_Connection_Handler.cpp: + * tao/ORB_Core.cpp: + * tao/params.h: + * tao/params.i: + * tao/params.cpp: + + This change was of OCI by David.Gibbs@igindex.co.uk. He had + previously requested the addition of support for SO_KEEPALIVE + and as a result we found that the framework existed for + communicating a value of SO_DONTROUTE, so it was decided that we + go ahead and add the feature. I consider this change provisional + in that it isn't strictly required and so if someone strongly + objects to its existence it can be pulled. Otherwise it is + simply completing what someone else had started a while ago by + adding configuration values for socket options to IIOP protocol + properties definition. + +Wed Apr 26 18:47:23 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/ORB_Core.cpp: + * tao/ORB_Core.h: + + Add an alternate connection timeout hook. This is necessary for + users of the Optimized Connect Endpoint Selector with its + timeout while also using CORBA messaging and the Connection + Timeout policy. Both of these connection timeouts are + dynamically loaded and one would override the other. This change + allows both to be loaded, and if both are initialized to nonzero + values, the lesser of the two timeouts is used. This results + from a bug originally reported to OCI by friedhelm.wolf@homag.de. + + * tao/Strategies/OC_Endpoint_Selector_Loader.cpp: + * tao/Strategies/OC_Endpoint_Selector_Loader.h: + + Cleaned up the initializer to be more consistent with others. + + * tao/Strategies/Optimized_Connection_Endpoint_Selector.cpp: + + Fixed wihtespace in debug output. + +Wed Apr 26 16:42:45 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * docs/Options.html: + + Adding text for the new commandline options used to control the + parallel connect feature. + + * tao/Blocked_Connect_Strategy.h: + * tao/Blocked_Connect_Strategy.cpp: + * tao/CORBALOC_Parser.cpp: + * tao/Client_Strategy_Factory.h: + * tao/Client_Strategy_Factory.cpp: + * tao/Connect_Strategy.h: + * tao/Connect_Strategy.cpp: + * tao/Endpoint.h: + * tao/Endpoint.cpp: + * tao/IIOP_Connection_Handler.h: + * tao/IIOP_Connection_Handler.cpp: + * tao/IIOP_Connector.h: + * tao/IIOP_Connector.cpp: + * tao/IIOP_Endpoint.h: + * tao/IIOP_Endpoint.cpp: + * tao/IIOP_Profile.h: + * tao/IIOP_Profile.cpp: + * tao/Invocation_Endpoint_Selectors.h: + * tao/Invocation_Endpoint_Selectors.cpp: + * tao/LF_CH_Event.h: + * tao/LF_Connect_Strategy.h: + * tao/LF_Connect_Strategy.cpp: + * tao/LF_Event.h: + * tao/LF_Multi_Event.h: + * tao/LF_Multi_Event.cpp: + * tao/MProfile.h: + * tao/MProfile.i: + * tao/MProfile.cpp: + * tao/ORB_Core.cpp: + * tao/Profile.h: + * tao/Profile.cpp: + * tao/Profile_Transport_Resolver.h: + * tao/Profile_Transport_Resolver.cpp: + * tao/Reactive_Connect_Strategy.h: + * tao/Reactive_Connect_Strategy.cpp: + * tao/Transport.cpp: + * tao/Transport_Connector.h: + * tao/Transport_Connector.cpp: + * tao/Transport_Descriptor_Interface.h: + * tao/Transport_Descriptor_Interface.inl: + * tao/Transport_Descriptor_Interface.cpp: + * tao/default_client.h: + * tao/default_client.cpp: + * tao/params.h: + * tao/params.i: + * tao/params.cpp: + * tao/tao.mpc: + + These changes support a new technique for active connection + establishment when presented with a profile containing multiple + possible endpoints. This commit resolves bugzilla bug #2485. + + The technique in question is "parallel connects" meaning + attempting to connect to many endpoints simultaniously. It was + conceived as a way to deal with timeouts when the Invocation + Endpoint Selector would first try to connect to one or more + unreachable endpoints. If those endpoints were defined as IP + addresses (not hostnames) or as resolvable hostnames that + pointed to unreachable IP addresses, the connection + establishment would take potentially several minutes to time out + and eventually encounter a reachable endpoint. In the case of + shared profiles (those using TAG_ALTERNATE_IIOP_ENDPOINT) this + delay impacts every single invocation. + + This parallel connect feature (also referred to somewhat + inacurately as a strategy) avoids this by supplying all the + endpoints in a profile to the connector and letting it first + test to see if any are already cached and available, and if not, + to open connections to each and wait for a winner. When the + first connection completes, any pending connections are + terminated. + + In order to minimize the use of pending connections, an iterator + traverses the list of endpoints creating new connections and + also checking any existing connections for completion. If the + first endpoint happens to be reachable and the server responds + quickly enough, the client may not open any more connections. + + If the server does not respond immediately, a wait strategy is + entered. This wait strategy may be Reactive or Leader/Follower + based. In either case, a specal "multi event" type is used to + allow a single thread to wait on one of many connectors, and + then to clean up those that didn't finish in time. The parallel + connect feature is also available using blocking connects, but + the only advantage there is in checking the cache for all + endpoints in the profile, there is no performance gain during + actual connection establishment. + + The parallel connect strategy differs from another endpoint + selection optimization, available in + tao/Strategies/Optimized_Connection_Endpoint_Selector.*. That + strategy works by examining all profiles simultaniously, this + feature still treats separate profiles separately. This profile + separation is necessary to support Load Balancing and Fault + Tolerence. Also, this feature requires additional support to be + built into protocol specific connectors (IIOP is currently the + only protocol supporting parallel connects) whereas the other + feature works regardless of the protocol. + + As this is a new feature, it is disabled by default. Use the + -ORBUseParallelConnects option to enable its use. A second + option, -ORBParallelConnectDelay, is used to introduce a small + delay between the opening of new potential connections if the + server is particularly busy. This is useful to minimize the + impact on a busy server if more than one of the available + endpoints is reachable. Also, because this feature only focuses + on one profile at a time, the server must be run with + -ORBUseSharedProfiles enabled (it is disabled by default). + + * tests/Parallel_Connect_Strategy/Parallel_Connect_Strategy.mpc: + * tests/Parallel_Connect_Strategy/README: + * tests/Parallel_Connect_Strategy/Test.idl: + * tests/Parallel_Connect_Strategy/Test_i.h: + * tests/Parallel_Connect_Strategy/Test_i.cpp: + * tests/Parallel_Connect_Strategy/blocked.conf: + * tests/Parallel_Connect_Strategy/client.cpp: + * tests/Parallel_Connect_Strategy/reactive.conf: + * tests/Parallel_Connect_Strategy/run_test.pl: + * tests/Parallel_Connect_Strategy/server.cpp: + + This is a new test for the parallel connect feature. It works by + having the server open two endpoints, one aliased to something + unreachable. The client then uses different wait strategies to + make invocations on the server and records the time for + each. These tests also include counter-examples in which + parallel connects are not used, and these take several minutes + to run. On my Linux machine the timeout period is about 3 + minutes which causes the overall test to take about 9 minutes to + run. + +Wed Apr 26 16:30:56 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/PortableServer/POAManagerFactory.cpp: + + Correct a bug found by the Borland compiler. + +Wed Apr 26 13:47:28 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/EndpointPolicy.mpc: + * tao/EndpointPolicy/EndpointPolicy.h: + * tao/EndpointPolicy/EndpointPolicy.pidl: + * tao/EndpointPolicy/EndpointPolicy.cpp: + * tao/EndpointPolicy/EndpointPolicyA.h: + * tao/EndpointPolicy/EndpointPolicyA.cpp: + * tao/EndpointPolicy/EndpointPolicyC.h: + * tao/EndpointPolicy/EndpointPolicyC.inl: + * tao/EndpointPolicy/EndpointPolicyC.cpp: + * tao/EndpointPolicy/EndpointPolicyType.pidl: + * tao/EndpointPolicy/EndpointPolicyTypeA.h: + * tao/EndpointPolicy/EndpointPolicyTypeA.cpp: + * tao/EndpointPolicy/EndpointPolicyTypeC.h: + * tao/EndpointPolicy/EndpointPolicyTypeC.cpp: + * tao/EndpointPolicy/EndpointPolicy_Export.h: + * tao/EndpointPolicy/EndpointPolicy_Factory.h: + * tao/EndpointPolicy/EndpointPolicy_Factory.cpp: + * tao/EndpointPolicy/EndpointPolicy_ORBInitializer.h: + * tao/EndpointPolicy/EndpointPolicy_ORBInitializer.cpp: + * tao/EndpointPolicy/EndpointPolicy_i.h: + * tao/EndpointPolicy/EndpointPolicy_i.cpp: + * tao/EndpointPolicy/Endpoint_Acceptor_Filter.h: + * tao/EndpointPolicy/Endpoint_Acceptor_Filter.cpp: + * tao/EndpointPolicy/Endpoint_Acceptor_Filter_Factory.h: + * tao/EndpointPolicy/Endpoint_Acceptor_Filter_Factory.cpp: + * tao/EndpointPolicy/Endpoint_Value_Impl.h: + * tao/EndpointPolicy/IIOPEndpointValue.pidl: + * tao/EndpointPolicy/IIOPEndpointValueA.h: + * tao/EndpointPolicy/IIOPEndpointValueA.cpp: + * tao/EndpointPolicy/IIOPEndpointValueC.h: + * tao/EndpointPolicy/IIOPEndpointValueC.cpp: + * tao/EndpointPolicy/IIOPEndpointValue_i.h: + * tao/EndpointPolicy/IIOPEndpointValue_i.cpp: + + The EndpointPolicy is a new, TAO-specific policy that is applied + to POAManagers via the POAManagerFactory. This commit resolves + Bugzilla bug #2484. + + The Endpoint policy acts as a filter for constraining the final + endpoints or profiles listed in an IOR when it is created by a + POA associated with the POAManager containing the policy. The + EndpointPolicy value is a sequence, allow multiple endpoints to + be published. + + The way this works is that the ORB is initialized with all the + -ORBEndpoint options it needs to provide access to all the + objects it will serve. Then POAManagers are created with + Endpoint policies that contain only the endpoints that are to be + used for its subset of objects. For instance, the ORB could + define one endpoint for insecure, internal-use-only objects, and + another for secure internet-facing objects. Using the Endpoint + Policy these different objects would only get one or the other + endpoint. Mechanically what happens is that first all Acceptors + are queried to construct an MProfile, then the resulting + profiles/endpoints are compared to entries in the policy, those + not matching are eliminated. It is possible that an endpoint + policy will exclude all the profiles, which would result in an + exception being raised at object reference construction time. + + Endpoints are matched in their final form. This means that if an + IIOP Endpoint makes use of the hostname_in_ior attribute, that + is the name the policy will use to match. + + Endpoint values are protocol specific. A value for IIOP is + provided, but new values must be defined to support other + protocols. The Endpoint value is a local object. New protocol + specific values do not have to be added to the + TAO_EndpointPolicy library, but they must specialize + EndpointPolicy::ValueBase and the implementation must derive + from TAO_Endpoint_Value_Impl. + + * tao/IIOP_Endpoint.h: + * tao/IIOP_Endpoint.cpp: + * tao/IIOP_Profile.h: + * tao/IIOP_Profile.cpp: + * tao/Profile.h: + * tao/Profile.cpp: + + The profile contains the base endpoint as an attribute. This + caused a problem for the endpoint removal scheme mentioned + above. If a profile contains two or more endpoints, and the base + happens to be the one to be eliminated as a result of the + endpoint comparison, the only thing that could be done is to + copy the contents of the first alternate into the base, then + eliminate the duplicate. + + * tao/orbconf.h: + + Added a tag for the new policy. + + * tests/POA/EndpointPolicy/EndpointPolicy.mpc: + * tests/POA/EndpointPolicy/Hello.h: + * tests/POA/EndpointPolicy/Hello.cpp: + * tests/POA/EndpointPolicy/README: + * tests/POA/EndpointPolicy/Test.idl: + * tests/POA/EndpointPolicy/client.cpp: + * tests/POA/EndpointPolicy/run_test.pl: + * tests/POA/EndpointPolicy/server.cpp: + * tests/POA/README: + + A new test case for the endpoint policy. This test currently + fails due to an unresolved interaction with the sequence + code. The error appears to be related to memory corruption, but + the cause has not yet been determined. The error only manifests + when the multiple profiles portion of the test is being + run. This is where an unmodified IOR would contain two profiles, + each with one endpoint. The problem does not occur when a single + profile has two endpoints. + +Wed Apr 26 14:04:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Smart_Proxies/client.cpp: + * tests/Smart_Proxies/Benchmark/client.cpp: + * tests/Smart_Proxies/dtor/client.cpp: + * tests/Smart_Proxies/Policy/client.cpp: + Removed remarks about the KAI compiler, more compilers do + give these warnings and support for the KAI compilers + has been removed + +Wed Apr 26 13:29:44 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * orbsvcs/examples/ORT/Server_IORInterceptor.h: + * orbsvcs/examples/ORT/Server_IORInterceptor.cpp: + + * orbsvcs/orbsvcs/LoadBalancing/LB_IORInterceptor.h: + * orbsvcs/orbsvcs/LoadBalancing/LB_IORInterceptor.cpp: + + * orbsvcs/orbsvcs/PortableGroup/GOA.h: + * orbsvcs/orbsvcs/PortableGroup/GOA.cpp: + * orbsvcs/orbsvcs/PortableGroup/PG_Servant_Dispatcher.h: + * orbsvcs/orbsvcs/PortableGroup/PG_Servant_Dispatcher.cpp: + + * tao/AnyTypeCode/PI_ForwardA.h: + + * tao/CSD_Framework/CSD_Default_Servant_Dispatcher.h: + * tao/CSD_Framework/CSD_Default_Servant_Dispatcher.cpp: + * tao/CSD_Framework/CSD_POA.h: + * tao/CSD_Framework/CSD_POA.cpp: + + * tao/IORInterceptor/IORInfo.h: + * tao/IORInterceptor/IORInfo.cpp: + * tao/IORInterceptor/IORInfoC.h: + * tao/IORInterceptor/IORInterceptorC.h: + * tao/IORInterceptor/IORInterceptor_Adapter_Impl.h: + * tao/IORInterceptor/IORInterceptor_Adapter_Impl.cpp: + + * tao/IORInterceptor_Adapter.h: + * tao/PI_Forward.pidl: + * tao/PI_ForwardC.h: + + * tao/RTPortableServer/RT_POA.h: + * tao/RTPortableServer/RT_POA.cpp: + * tao/RTPortableServer/RT_Servant_Dispatcher.h: + * tao/RTPortableServer/RT_Servant_Dispatcher.cpp: + + * tests/ORT/ORT_test_IORInterceptor.h: + * tests/ORT/ORT_test_IORInterceptor.cpp: + + * tests/Portable_Interceptors/IORInterceptor/FOO_IORInterceptor.h: + * tests/Portable_Interceptors/IORInterceptor/FOO_IORInterceptor.cpp: + + These are more changes related to the POAManagerFactory. They + are coupled with my 13:10:59 utc checkin. + +Wed Apr 26 13:10:59 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/PortableServer/Acceptor_Filter_Factory.h: + * tao/PortableServer/Acceptor_Filter_Factory.cpp: + * tao/PortableServer/Default_Acceptor_Filter.h: + * tao/PortableServer/Default_Servant_Dispatcher.h: + * tao/PortableServer/Default_Servant_Dispatcher.cpp: + * tao/PortableServer/Object_Adapter.h: + * tao/PortableServer/Object_Adapter.cpp: + * tao/PortableServer/POAManager.h: + * tao/PortableServer/POAManager.i: + * tao/PortableServer/POAManager.pidl: + * tao/PortableServer/POAManager.cpp: + * tao/PortableServer/POAManagerC.h: + * tao/PortableServer/POAManagerFactory.h: + * tao/PortableServer/POAManagerFactory.cpp: + * tao/PortableServer/POAManagerFactory.pidl: + * tao/PortableServer/POAManagerFactoryC.h: + * tao/PortableServer/POAManagerFactoryC.cpp: + * tao/PortableServer/PortableServer.h: + * tao/PortableServer/PortableServer.pidl: + * tao/PortableServer/PortableServerC.h: + * tao/PortableServer/Regular_POA.h: + * tao/PortableServer/Regular_POA.cpp: + * tao/PortableServer/Root_POA.h: + * tao/PortableServer/Root_POA.cpp: + * tao/PortableServer/Servant_Dispatcher.h: + + These files are new/updated to support the POAManagerFactory, + which was added to the CORBA 3.0.2 specification. The PMF is + used to allow for the explicit creation of POA Managers which + can then be supplied to POAs during POA creation. POA Managers + may now also carry policies which will influence all POAs + associated with it. This work builds on the effort originally + started by Johnny Willemsen back in the pre-1.4.8 era. This + commit resolves Bugzilla bug #1785. + + * tests/POA/POAManagerFactory/POAManagerFactory.cpp: + * tests/POA/POAManagerFactory/POAManagerFactory.mpc: + * tests/POA/POAManagerFactory/run_test.pl: + + This is a new test for the POAManagerFactory. + +Wed Apr 26 13:01:48 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/Valuetype/AbstractBase.cpp: + * tao/Valuetype/AbstractBase.h: + The _tao_marshal_v method is supposed to be const. + +Wed Apr 26 13:01:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/POA/Bug_2511_Regression/*: + Added new regression for bug 2511. Thanks to Martin Cornelius + <Martin at Cornelius at smiths-heimann dot com> for creating + this regression + +Wed Apr 26 12:20:51 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tests/AMI/ami_test.idl: + * tests/AMI/ami_test_i.cpp: + * tests/AMI/simple_client.cpp: + Reverting earlier test changes. They were causing problems on + some platforms and I don't have a clear enough memory of the + original motivation for the change. + +Wed Apr 26 11:46:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/tests/AVStreams/Pluggable/ftp.cpp: + * orbsvcs/tests/AVStreams/Multicast/ftp.cpp: + Fixed a bug in these tests, in the test code for element 0 of + a string sequence was set and after that the length was set to 1. + The setting of 0 is possible because the OMG doesn't define + exceptions for this so it is allowed, then setting the length to + 1 does reinitialize element 0 so that we don't get old values. + This was not done with the old sequences, then just the old + value was returned and things worked then. + +Wed Apr 26 11:21:57 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/Valuetype/ValueBase.cpp: + + Fix for errant removal of throw, this corrects the OBV/Factory + test. + +Wed Apr 26 10:41:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Unbounded_Sequence_CDR_T.h: + Added missing include of SystemException + +Wed Apr 26 10:03:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/OBV/Simple/*: + Added very simple OBV test + +Wed Apr 26 09:53:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Portable_Interceptors/Bug_2510_Regression: + New regression test, thanks to Martin Cornelius + <Martin at Cornelius at smiths-heimann dot com> for creating + this test. This bug is not fixed yet. + +Wed Apr 26 09:44:12 UTC 2006 Kees van Marle <kvmarle@remedy.nl> + + * tao/Bounded_Sequence_CDR_T.h: + * tao/Unbounded_Sequence_CDR_T.h: + Check in all marshal_sequence methods if we aren't trying to + marshal a nill sequence, this can happen when the user doesn't + initialize an out argument. In that case we throw a BAD_PARAM + exception as described in the C++ spec. This fixes bugzilla bug + 1676. + +Wed Apr 26 08:42:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/ObjRefTemplate/ObjectReferenceTemplateC.h: + * tao/ObjRefTemplate/Default_ORTC.h: + * tao/Messaging/ExceptionHolderC.h: + Include the Valuetype_Adapter_Factory_Impl.h so that the + Valuetype library gets linked in a static build, this is + already updated earlier in the IDL compiler + +Wed Apr 26 08:17:12 UTC 2006 Kees van Marle <kvmarle@remedy.nl> + + * tests/Bug_1676_Regression/client.cpp: + Extended this test to explicitly test for BAD_PARAM exception + when the server not initializes an out argument + +Wed Apr 26 07:24:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/orbsvcs/Trader/Interpreter_Utils.h: + Removed invalid template export + +Wed Apr 26 03:46:16 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * tao/Valuetype/AbstractBase.cpp: + + Fixed scoreboard detected warning. + +Tue Apr 25 19:24:48 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * TAO_IDL/be/be_visitor_valuetype/valuetype_cs.cpp: + + Missed a patch from the earlier commit. + +Tue Apr 25 19:09:08 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * orbsvcs/examples/ORT/Server_IORInterceptor.h: + * orbsvcs/examples/ORT/Server_IORInterceptor.cpp: + + * orbsvcs/orbsvcs/LoadBalancing/LB_IORInterceptor.h: + * orbsvcs/orbsvcs/LoadBalancing/LB_IORInterceptor.cpp: + + * orbsvcs/orbsvcs/PortableGroup/GOA.h: + * orbsvcs/orbsvcs/PortableGroup/GOA.cpp: + * orbsvcs/orbsvcs/PortableGroup/PG_Servant_Dispatcher.h: + * orbsvcs/orbsvcs/PortableGroup/PG_Servant_Dispatcher.cpp: + + * tao/AnyTypeCode/PI_ForwardA.h: + + * tao/CSD_Framework/CSD_Default_Servant_Dispatcher.h: + * tao/CSD_Framework/CSD_Default_Servant_Dispatcher.cpp: + * tao/CSD_Framework/CSD_POA.h: + * tao/CSD_Framework/CSD_POA.cpp: + + * tao/IIOP_Endpoint.h: + * tao/IIOP_Endpoint.cpp: + * tao/IIOP_Profile.h: + * tao/IIOP_Profile.cpp: + + * tao/IORInterceptor/IORInfo.h: + * tao/IORInterceptor/IORInfo.cpp: + * tao/IORInterceptor/IORInfoC.h: + * tao/IORInterceptor/IORInterceptorC.h: + * tao/IORInterceptor/IORInterceptor_Adapter_Impl.h: + * tao/IORInterceptor/IORInterceptor_Adapter_Impl.cpp: + + * tao/IORInterceptor_Adapter.h: + * tao/PI_Forward.pidl: + * tao/PI_ForwardC.h: + + * tao/PortableServer/Default_Acceptor_Filter.h: + * tao/PortableServer/Default_Servant_Dispatcher.h: + * tao/PortableServer/Default_Servant_Dispatcher.cpp: + * tao/PortableServer/Object_Adapter.h: + * tao/PortableServer/Object_Adapter.cpp: + * tao/PortableServer/POAManager.h: + * tao/PortableServer/POAManager.i: + * tao/PortableServer/POAManager.cpp: + * tao/PortableServer/POAManager.pidl: + * tao/PortableServer/POAManagerC.h: + * tao/PortableServer/PortableServer.h: + * tao/PortableServer/PortableServer.pidl: + * tao/PortableServer/PortableServerC.h: + * tao/PortableServer/Regular_POA.h: + * tao/PortableServer/Regular_POA.cpp: + * tao/PortableServer/Root_POA.h: + * tao/PortableServer/Root_POA.cpp: + * tao/PortableServer/Servant_Dispatcher.h: + + * tao/Profile.h: + * tao/Profile.cpp: + + * tao/RTPortableServer/RT_POA.h: + * tao/RTPortableServer/RT_POA.cpp: + * tao/RTPortableServer/RT_Servant_Dispatcher.h: + * tao/RTPortableServer/RT_Servant_Dispatcher.cpp: + + * tao/orbconf.h: + * tao/params.cpp: + + * tests/ORT/ORT_test_IORInterceptor.h: + * tests/ORT/ORT_test_IORInterceptor.cpp: + + * tests/POA/EndpointPolicy/EndpointPolicy.mpc: + * tests/POA/EndpointPolicy/Hello.h: + * tests/POA/EndpointPolicy/Hello.cpp: + * tests/POA/EndpointPolicy/README: + * tests/POA/EndpointPolicy/Test.idl: + * tests/POA/EndpointPolicy/client.cpp: + * tests/POA/EndpointPolicy/run_test.pl: + * tests/POA/EndpointPolicy/server.cpp: + + * tests/POA/POAManagerFactory/POAManagerFactory.cpp: + * tests/POA/POAManagerFactory/POAManagerFactory.mpc: + * tests/POA/POAManagerFactory/run_test.pl: + + * tests/POA/README: + + * tests/Portable_Interceptors/IORInterceptor/FOO_IORInterceptor.h: + * tests/Portable_Interceptors/IORInterceptor/FOO_IORInterceptor.cpp: + +Tue Apr 25 17:38:34 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + * TAO_IDL/be/be_visitor_valuebox/valuebox_ch.cpp: + * TAO_IDL/be/be_visitor_valuebox/valuebox_ci.cpp: + * TAO_IDL/be/be_visitor_valuetype/valuetype_ch.cpp: + * TAO_IDL/be/be_visitor_valuetype/valuetype_cs.cpp: + * tao/Messaging/ExceptionHolderC.h: + * tao/Messaging/ExceptionHolderC.cpp: + * tao/ObjRefTemplate/Default_ORTC.h: + * tao/ObjRefTemplate/Default_ORTC.cpp: + * tao/ObjRefTemplate/ObjectReferenceTemplateC.h: + * tao/ObjRefTemplate/ObjectReferenceTemplateC.cpp: + * tao/Valuetype/AbstractBase.h: + * tao/Valuetype/AbstractBase.cpp: + * tao/Valuetype/StringValueC.h: + * tao/Valuetype/StringValueC.inl: + * tao/Valuetype/ValueBase.h: + * tao/Valuetype/ValueBase.cpp: + + These are further valuetype changes that are related to Bugzilla + #2162. The problem is that TAO assumes that a valuetype's + typecode is always encoded as 0x7FFFFF02 <repID> but that is not + compliant with the spec. It is also valid to encode a valuetype + typecode as 0x7FFFFF00 which indicates that the actual type of + the value matches the formal type for the argument for which the + value is a parameter. TAO already had most of the hooks in + place to support this, but was missing a key bit of + functionality. This patch adds that functionality, the ability + to test that the formal type matches the actual type when + marshaling values. + + Valuetypes encoded this way are substantially more efficient, if + the type can be implied then there is no need to carry the + actual type id. Unfortunately to do so blindly would break + backwards compatibility with all previous versions of TAO. + For the time being, the effective code to cause TAO to marshal + values using the more efficient typecode is disabled using a new + compile-time flag, TAO_HAS_OPTIMIMIZED_VALUETYPE_MARSHALING, + which must be defined to give TAO the opportunity to use this + new technique. THIS FLAG BREAKS BACKWARDS COMPATIBILITY. It is + not a violation of the spec to always encode the valuetype's + type ID when marshaling, so TAO can continue being backwards + compatible and not be in violation of the spec. The only + violation comes when failing to unmarshal a value which is using + an implied type ID. + + I would prefer to not have a compiler flag to guard the use of + optimized marshaling, but I don't know any other way to do it. + At the point where the decision is made, there is no reference + to an ORB Core so there is no easy way to set a dynamic option + that could be used to selectively control this optimization. + +Tue Apr 25 15:14:13 UTC 2006 Phil Mesnier <mesnier_p@ociweb.com> + + Merging in truncatable valuetype support. This work was done on + an OCI controlled patch then brought in via patch. The premise + is to support the "truncatable" keyword for valuetypes as + defined in sections 3.9.1.3, 5.2.5.3 and 15.3.4.1 of the CORBA + 3.0.3 specification. Practically, this means supporting the + marshaling of typecode lists and chunked values. Chunked values + requires the retention of state, namely nesting level, during + the marshaling. This was handled by creating a new ChunkInfo + type that is created on the stack during the marshaling of a + valuetype and is passed through all the intermediate marshal + methods. This commit resolves Bugzilla #2483 + + * TAO_IDL/be/be_visitor_valuebox/cdr_op_ch.cpp: + * TAO_IDL/be/be_visitor_valuebox/cdr_op_cs.cpp: + * TAO_IDL/be/be_visitor_valuebox/valuebox_ch.cpp: + * TAO_IDL/be/be_visitor_valuebox/valuebox_cs.cpp: + + Valueboxes by definition cannot be made truncatable, but they + derive from the same valuebase, thus they must support the same + signature for creating a list of repository ids. Otherwise these + are whitespace only changes. + + * TAO_IDL/be/be_visitor_valuetype/cdr_op_ch.cpp: + * TAO_IDL/be/be_visitor_valuetype/cdr_op_cs.cpp: + * TAO_IDL/be/be_visitor_valuetype/marshal_cs.cpp: + * TAO_IDL/be/be_visitor_valuetype/valuetype_ch.cpp: + * TAO_IDL/be/be_visitor_valuetype/valuetype_ci.cpp: + * TAO_IDL/be/be_visitor_valuetype/valuetype_cs.cpp: + * TAO_IDL/be/be_visitor_valuetype/valuetype_obv_ch.cpp: + * TAO_IDL/be/be_visitor_valuetype/valuetype_obv_cs.cpp: + + These are changes for generating code that manages the chunkinfo + data as it passes through the value members. + + * TAO_IDL/be/be_visitor_valuetype_fwd/cdr_op_ch.cpp: + + Whitespace changes. + + * TAO_IDL/fe/y.tab.cpp: + * TAO_IDL/fe/idl.yy: + + Removed the warning about support for truncatables. + + * tao/AnyTypeCode/skip.cpp: + + Fixed the type for the valuetag. + + * tao/Messaging/ExceptionHolderC.h: + * tao/Messaging/ExceptionHolderC.cpp: + * tao/ObjRefTemplate/Default_ORTC.h: + * tao/ObjRefTemplate/Default_ORTC.cpp: + * tao/ObjRefTemplate/ObjectReferenceTemplateC.h: + * tao/ObjRefTemplate/ObjectReferenceTemplateC.cpp: + + These are the changes required by the truncatable support that + would ordinarily be generated by the IDL compiler. + + * tao/Valuetype/AbstractBase.h: + * tao/Valuetype/AbstractBase.cpp: + * tao/Valuetype/StringValueC.h: + * tao/Valuetype/StringValueC.cpp: + * tao/Valuetype/ValueBase.h: + * tao/Valuetype/ValueBase.inl: + * tao/Valuetype/ValueBase.cpp: + * tao/Valuetype/Value_CORBA_methods.h: + + The changes for StringValue and AbstractBase are the same as the + generated code. The changes in ValueBase are those common to all + value types, used to determine how to marshal values if the + truncatable keyword was defined or not. This also includes the + definintion of the new ChunkInfo type. + + * tests/AMI/ami_test.idl: + * tests/AMI/ami_test_i.cpp: + * tests/AMI/simple_client.cpp: + + Changes test misc. fixes for support of wchar data in exceptions. + + * tests/OBV/Truncatable/OBV_Truncatable.mpc: + * tests/OBV/Truncatable/README: + * tests/OBV/Truncatable/Truncatable.idl: + * tests/OBV/Truncatable/TruncatableS_impl.h: + * tests/OBV/Truncatable/TruncatableS_impl.cpp: + * tests/OBV/Truncatable/client.cpp: + * tests/OBV/Truncatable/run_test.pl: + * tests/OBV/Truncatable/server.cpp: + + A new test specific to validating the truncatable valuetypes. + Note this test is also added to the ace/bin/tao_orb_tests.lst. + + * tests/Param_Test/svc.conf: + + Remove the explicit override of the wchar codeset for the + +Tue Apr 25 14:25:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + Reverted this change below, this change broke this test + + Fri Apr 21 08:11:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + * orbsvcs/tests/Trading/TTest.idl: + Use the CORBA predefined sequence types + +Tue Apr 25 12:20:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Sequence_Unit_Tests/Sequence_Unit_Tests.mpc: + Added missing unbouded array unit test + +Tue Apr 25 11:40:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/orbsvcs/Trader/Constraint_Visitors.cpp + * orbsvcs/orbsvcs/Trader/Interpreter_Utils.{h,cpp}: + * orbsvcs/orbsvcs/Trader/Offer_Database.{h,cpp}: + * orbsvcs/orbsvcs/Trader/Service_Type_Repository.{h,cpp}: + * orbsvcs/orbsvcs/Trader/Trader.h + * orbsvcs/orbsvcs/Trader/Trader_Constraint_Visitors.cpp + * orbsvcs/orbsvcs/Trader/Trader_Interfaces.{h,cpp}: + * orbsvcs/orbsvcs/Trader/Trader_Utils.{h,cpp}: + Fixed duplicate symbols when linking with vc7/vc8. The trading + service used the TAO_String_Hash_Key class to store strings + in hash maps, this class is derived from CORBA::String_var which + has been refactored to a template. This causes problems with + vc7/vc8 because the base template is exported from multiple + libraries. This has been resolved by usign CORBA::String_var + in the hash map and deliver an ACE_Hash, ACE_Equal_To and + ACE_Less_Then template specialization. This solves now the + link problems, the runtime issue that appeared after the + sequence merge has not been fixed yet. Also see bugzilla bug + 2520 for more info. + +Tue Apr 25 08:46:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/ORB.cpp: + Removed runtime check of the sizes of the basic data types. We + now always use bool for CORBA::Boolean independent of the size + of bool, see also bugzilla 2515 + +Tue Apr 25 06:37:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Object.{h,cpp}: + Made the signature of the static marshal method the same in the + definition and the implementation. + +Mon Apr 24 19:05:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/orbsvcs/Trader/Interpreter_Utils.h: + Explicitly export the base template to fix duplicate symbol + errors with vc71/vc8 + +Mon Apr 24 14:16:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/orbsvcs/AV/FlowSpec_Entry.cpp: + Improved debug messages so that we can see if a string is empty + or not. + + * orbsvcs/orbsvcs/AV/FlowSpec_Entry.h: + Doxygen cleanup + + * orbsvcs/orbsvcs/AV/AVStreams_i.cpp: + Improved debugging output to resolve bug that seems to be introduced + by the sequence changes, the flowspec sequence has length of 1 but + just an empty string as value + +Mon Apr 24 13:12:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/params.cpp: + Small const improvement + + * tao/ORB.{h,cpp}: + Updated CORBA::ORB::RequestSeq to CORBA::RequestSeq as the spec + describes. Fixes bugzilla bug 2512. + +Mon Apr 24 12:02:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/ORB.cpp: + Added include of ObjectIdListC.h to resolve compile errors related + to this type with Sun Studio 10 + +Mon Apr 24 11:56:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + Reverted change below, set method is IDL generated. + + Mon Apr 24 09:39:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + * orbsvcs/orbsvcs/Runtime_Scheduler.{h,cpp}: + Made the arguments of the set method a const reference, this fixes + the internal backend errors in the BCB2006 release builds. + +Mon Apr 24 11:31:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/ORB_Core.cpp: + When we can't get a valuetype adapter throw an internal corba + exception + + * TAO_IDL/be/be_codegen.cpp: + Updated include generated for the valuetype library, makes sure + that the valuetype library gets linked into the executable when + building static. + +Mon Apr 24 10:16:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Bug_1676_Regression/*: + Added regression for bug 1676 written by Kees van Marle. This bug + seems not to be fixed yet. + +Mon Apr 24 09:39:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/orbsvcs/Runtime_Scheduler.{h,cpp}: + Made the arguments of the set method a const reference, this fixes + the internal backend errors in the BCB2006 release builds. + +Mon Apr 24 06:59:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/orbsvcs/FaultTolerance/FT_ServerPolicy_i.inl: + Removed some left over ACE_NESTED_CLASS usage + +Sun Apr 23 11:26:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/orbsvcs/IFRService/IFR_ComponentsS.cpp: + Removed some left over ACE_NESTED_CLASS usage + +Fri Apr 21 20:43:24 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * examples/Makefile.am: + * performance-tests/Makefile.am: + + Remove handful of subdirectories that have not yet been updated + to work with autoconf builds. + +Fri Apr 21 19:32:27 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * configure.ac: + + Added orbsvcs/tests/Bug_2377_Regression/Makefile to + AC_CONFIG_FILES. + + * orbsvcs/tests/Bug_2377_Regression/Makefile.am: + + New file. + +Fri Apr 21 14:07:51 UTC 2006 J.T. Conklin <jtc@acorntoolworks.com> + + * TAO_IDL/Makefile.am: + * docs/Makefile.am: + * docs/tutorials/Makefile.am: + * docs/tutorials/Quoter/Makefile.am: + * docs/tutorials/Quoter/AMI/Makefile.am: + * docs/tutorials/Quoter/Event_Service/Makefile.am: + * docs/tutorials/Quoter/Naming_Service/Makefile.am: + * docs/tutorials/Quoter/On_Demand_Activation/Makefile.am: + * docs/tutorials/Quoter/RT_Event_Service/Makefile.am: + * docs/tutorials/Quoter/Simple/Makefile.am: + * docs/tutorials/Quoter/Simple/Client/Makefile.am: + * docs/tutorials/Quoter/Simple/Impl-Repo/Makefile.am: + * docs/tutorials/Quoter/Simple/ImprovedServer/Makefile.am: + * docs/tutorials/Quoter/Simple/Persistent/Makefile.am: + * docs/tutorials/Quoter/Simple/Server/Makefile.am: + * docs/tutorials/Quoter/idl/Makefile.am: + * examples/Makefile.am: + * examples/AMH/Makefile.am: + * examples/AMH/Sink_Server/Makefile.am: + * examples/AMI/Makefile.am: + * examples/AMI/FL_Callback/Makefile.am: + * examples/Advanced/Makefile.am: + * examples/Advanced/ch_3/Makefile.am: + * examples/Buffered_AMI/Makefile.am: + * examples/Buffered_Oneways/Makefile.am: + * examples/CSD_Strategy/Makefile.am: + * examples/CSD_Strategy/ThreadPool/Makefile.am: + * examples/CSD_Strategy/ThreadPool2/Makefile.am: + * examples/CSD_Strategy/ThreadPool3/Makefile.am: + * examples/CSD_Strategy/ThreadPool4/Makefile.am: + * examples/CSD_Strategy/ThreadPool5/Makefile.am: + * examples/CSD_Strategy/ThreadPool6/Makefile.am: + * examples/Callback_Quoter/Makefile.am: + * examples/Content_Server/Makefile.am: + * examples/Content_Server/AMI_Iterator/Makefile.am: + * examples/Content_Server/AMI_Observer/Makefile.am: + * examples/Content_Server/SMI_Iterator/Makefile.am: + * examples/Event_Comm/Makefile.am: + * examples/Kokyu_dsrt_schedulers/Makefile.am: + * examples/Kokyu_dsrt_schedulers/fp_example/Makefile.am: + * examples/Kokyu_dsrt_schedulers/mif_example/Makefile.am: + * examples/Kokyu_dsrt_schedulers/muf_example/Makefile.am: + * examples/Load_Balancing/Makefile.am: + * examples/Load_Balancing_persistent/Makefile.am: + * examples/Logging/Makefile.am: + * examples/OBV/Makefile.am: + * examples/OBV/Typed_Events/Makefile.am: + * examples/POA/Makefile.am: + * examples/POA/Adapter_Activator/Makefile.am: + * examples/POA/DSI/Makefile.am: + * examples/POA/Default_Servant/Makefile.am: + * examples/POA/Explicit_Activation/Makefile.am: + * examples/POA/Explicit_Activation/Alt_Resources/Makefile.am: + * examples/POA/FindPOA/Makefile.am: + * examples/POA/Forwarding/Makefile.am: + * examples/POA/Generic_Servant/Makefile.am: + * examples/POA/Loader/Makefile.am: + * examples/POA/NewPOA/Makefile.am: + * examples/POA/On_Demand_Activation/Makefile.am: + * examples/POA/On_Demand_Loading/Makefile.am: + * examples/POA/POA_BiDir/Makefile.am: + * examples/POA/Reference_Counted_Servant/Makefile.am: + * examples/POA/RootPOA/Makefile.am: + * examples/POA/TIE/Makefile.am: + * examples/Persistent_Grid/Makefile.am: + * examples/PluggableUDP/Makefile.am: + * examples/PluggableUDP/tests/Makefile.am: + * examples/PluggableUDP/tests/Basic/Makefile.am: + * examples/PluggableUDP/tests/Performance/Makefile.am: + * examples/PluggableUDP/tests/SimplePerformance/Makefile.am: + * examples/Quoter/Makefile.am: + * examples/RTCORBA/Makefile.am: + * examples/RTCORBA/Activity/Makefile.am: + * examples/RTScheduling/Makefile.am: + * examples/RTScheduling/Fixed_Priority_Scheduler/Makefile.am: + * examples/RTScheduling/MIF_Scheduler/Makefile.am: + * examples/Simple/Makefile.am: + * examples/Simple/bank/Makefile.am: + * examples/Simple/chat/Makefile.am: + * examples/Simple/echo/Makefile.am: + * examples/Simple/grid/Makefile.am: + * examples/Simple/time/Makefile.am: + * examples/Simple/time-date/Makefile.am: + * examples/Simulator/Makefile.am: + * examples/Simulator/Event_Supplier/Makefile.am: + * examples/TypeCode_Creation/Makefile.am: + * examples/ior_corbaloc/Makefile.am: + * examples/mfc/Makefile.am: + * interop-tests/Makefile.am: + * interop-tests/wchar/Makefile.am: + * orbsvcs/Makefile.am: + * orbsvcs/Concurrency_Service/Makefile.am: + * orbsvcs/CosEvent_Service/Makefile.am: + * orbsvcs/Dump_Schedule/Makefile.am: + * orbsvcs/Event_Service/Makefile.am: + * orbsvcs/FTRT_Event_Service/Makefile.am: + * orbsvcs/FTRT_Event_Service/Event_Service/Makefile.am: + * orbsvcs/FTRT_Event_Service/Factory_Service/Makefile.am: + * orbsvcs/FTRT_Event_Service/Gateway_Service/Makefile.am: + * orbsvcs/FT_ReplicationManager/Makefile.am: + * orbsvcs/Fault_Detector/Makefile.am: + * orbsvcs/Fault_Notifier/Makefile.am: + * orbsvcs/IFR_Service/Makefile.am: + * orbsvcs/ImplRepo_Service/Makefile.am: + * orbsvcs/LoadBalancer/Makefile.am: + * orbsvcs/Logging_Service/Makefile.am: + * orbsvcs/Logging_Service/Basic_Logging_Service/Makefile.am: + * orbsvcs/Logging_Service/Event_Logging_Service/Makefile.am: + * orbsvcs/Logging_Service/Notify_Logging_Service/Makefile.am: + * orbsvcs/Logging_Service/RTEvent_Logging_Service/Makefile.am: + * orbsvcs/Naming_Service/Makefile.am: + * orbsvcs/Notify_Service/Makefile.am: + * orbsvcs/Scheduling_Service/Makefile.am: + * orbsvcs/TAO_Service/Makefile.am: + * orbsvcs/Time_Service/Makefile.am: + * orbsvcs/Trading_Service/Makefile.am: + * orbsvcs/examples/Makefile.am: + * orbsvcs/examples/CosEC/Makefile.am: + * orbsvcs/examples/CosEC/Factory/Makefile.am: + * orbsvcs/examples/CosEC/RtEC_Based/Makefile.am: + * orbsvcs/examples/CosEC/RtEC_Based/bin/Makefile.am: + * orbsvcs/examples/CosEC/RtEC_Based/lib/Makefile.am: + * orbsvcs/examples/CosEC/RtEC_Based/tests/Makefile.am: + * orbsvcs/examples/CosEC/RtEC_Based/tests/Basic/Makefile.am: + * orbsvcs/examples/CosEC/RtEC_Based/tests/Multiple/Makefile.am: + * orbsvcs/examples/CosEC/Simple/Makefile.am: + * orbsvcs/examples/CosEC/TypedSimple/Makefile.am: + * orbsvcs/examples/FaultTolerance/Makefile.am: + * orbsvcs/examples/FaultTolerance/RolyPoly/Makefile.am: + * orbsvcs/examples/ImR/Makefile.am: + * orbsvcs/examples/ImR/Advanced/Makefile.am: + * orbsvcs/examples/ImR/Combined_Service/Makefile.am: + * orbsvcs/examples/LoadBalancing/Makefile.am: + * orbsvcs/examples/Log/Makefile.am: + * orbsvcs/examples/Log/Basic/Makefile.am: + * orbsvcs/examples/Log/Event/Makefile.am: + * orbsvcs/examples/Log/Notify/Makefile.am: + * orbsvcs/examples/Log/RTEvent/Makefile.am: + * orbsvcs/examples/Notify/Makefile.am: + * orbsvcs/examples/Notify/Federation/Makefile.am: + * orbsvcs/examples/Notify/Federation/Agent/Makefile.am: + * orbsvcs/examples/Notify/Federation/Gate/Makefile.am: + * orbsvcs/examples/Notify/Federation/SpaceCraft/Makefile.am: + * orbsvcs/examples/Notify/Filter/Makefile.am: + * orbsvcs/examples/Notify/Lanes/Makefile.am: + * orbsvcs/examples/Notify/Subscribe/Makefile.am: + * orbsvcs/examples/Notify/ThreadPool/Makefile.am: + * orbsvcs/examples/ORT/Makefile.am: + * orbsvcs/examples/RtEC/Makefile.am: + * orbsvcs/examples/RtEC/IIOPGateway/Makefile.am: + * orbsvcs/examples/RtEC/Kokyu/Makefile.am: + * orbsvcs/examples/RtEC/MCast/Makefile.am: + * orbsvcs/examples/RtEC/Schedule/Makefile.am: + * orbsvcs/examples/RtEC/Simple/Makefile.am: + * orbsvcs/examples/Security/Makefile.am: + * orbsvcs/examples/Security/Send_File/Makefile.am: + * orbsvcs/orbsvcs/Makefile.am: + * orbsvcs/performance-tests/Makefile.am: + * orbsvcs/performance-tests/LoadBalancing/Makefile.am: + * orbsvcs/performance-tests/LoadBalancing/LBPerf/Makefile.am: + * orbsvcs/performance-tests/LoadBalancing/LBPerf/RPS/Makefile.am: + * orbsvcs/performance-tests/RTEvent/Makefile.am: + * orbsvcs/performance-tests/RTEvent/Colocated_Roundtrip/Makefile.am: + * orbsvcs/performance-tests/RTEvent/Federated_Roundtrip/Makefile.am: + * orbsvcs/performance-tests/RTEvent/RTCORBA_Baseline/Makefile.am: + * orbsvcs/performance-tests/RTEvent/RTCORBA_Callback/Makefile.am: + * orbsvcs/performance-tests/RTEvent/Roundtrip/Makefile.am: + * orbsvcs/performance-tests/RTEvent/TCP_Baseline/Makefile.am: + * orbsvcs/performance-tests/RTEvent/lib/Makefile.am: + * orbsvcs/tests/Makefile.am: + * orbsvcs/tests/AVStreams/Makefile.am: + * orbsvcs/tests/AVStreams/Asynch_Three_Stage/Makefile.am: + * orbsvcs/tests/AVStreams/Bidirectional_Flows/Makefile.am: + * orbsvcs/tests/AVStreams/Component_Switching/Makefile.am: + * orbsvcs/tests/AVStreams/Full_Profile/Makefile.am: + * orbsvcs/tests/AVStreams/Latency/Makefile.am: + * orbsvcs/tests/AVStreams/Modify_QoS/Makefile.am: + * orbsvcs/tests/AVStreams/Multicast/Makefile.am: + * orbsvcs/tests/AVStreams/Multicast_Full_Profile/Makefile.am: + * orbsvcs/tests/AVStreams/Multiple_Flows/Makefile.am: + * orbsvcs/tests/AVStreams/Pluggable/Makefile.am: + * orbsvcs/tests/AVStreams/Pluggable_Flow_Protocol/Makefile.am: + * orbsvcs/tests/AVStreams/Simple_Three_Stage/Makefile.am: + * orbsvcs/tests/AVStreams/Simple_Two_Stage/Makefile.am: + * orbsvcs/tests/AVStreams/Simple_Two_Stage_With_QoS/Makefile.am: + * orbsvcs/tests/BiDir_CORBALOC/Makefile.am: + * orbsvcs/tests/Bug_1334_Regression/Makefile.am: + * orbsvcs/tests/Bug_1393_Regression/Makefile.am: + * orbsvcs/tests/Bug_1395_Regression/Makefile.am: + * orbsvcs/tests/Bug_1630_Regression/Makefile.am: + * orbsvcs/tests/Bug_2074_Regression/Makefile.am: + * orbsvcs/tests/Bug_2137_Regression/Makefile.am: + * orbsvcs/tests/Bug_2247_Regression/Makefile.am: + * orbsvcs/tests/Bug_2248_Regression/Makefile.am: + * orbsvcs/tests/Bug_2285_Regression/Makefile.am: + * orbsvcs/tests/Bug_2287_Regression/Makefile.am: + * orbsvcs/tests/Bug_2316_Regression/Makefile.am: + * orbsvcs/tests/Concurrency/Makefile.am: + * orbsvcs/tests/CosEvent/Makefile.am: + * orbsvcs/tests/CosEvent/Basic/Makefile.am: + * orbsvcs/tests/CosEvent/lib/Makefile.am: + * orbsvcs/tests/EC_Custom_Marshal/Makefile.am: + * orbsvcs/tests/EC_MT_Mcast/Makefile.am: + * orbsvcs/tests/EC_Mcast/Makefile.am: + * orbsvcs/tests/EC_Multiple/Makefile.am: + * orbsvcs/tests/EC_Throughput/Makefile.am: + * orbsvcs/tests/Event/Makefile.am: + * orbsvcs/tests/Event/Basic/Makefile.am: + * orbsvcs/tests/Event/Mcast/Makefile.am: + * orbsvcs/tests/Event/Mcast/Common/Makefile.am: + * orbsvcs/tests/Event/Mcast/Complex/Makefile.am: + * orbsvcs/tests/Event/Mcast/Simple/Makefile.am: + * orbsvcs/tests/Event/Mcast/Two_Way/Makefile.am: + * orbsvcs/tests/Event/Performance/Makefile.am: + * orbsvcs/tests/Event/lib/Makefile.am: + * orbsvcs/tests/FT_App/Makefile.am: + * orbsvcs/tests/FaultTolerance/Makefile.am: + * orbsvcs/tests/FaultTolerance/GroupRef_Manipulation/Makefile.am: + * orbsvcs/tests/FaultTolerance/IOGR/Makefile.am: + * orbsvcs/tests/FaultTolerance/IOGRManipulation/Makefile.am: + * orbsvcs/tests/FtRtEvent/Makefile.am: + * orbsvcs/tests/HTIOP/Makefile.am: + * orbsvcs/tests/HTIOP/AMI/Makefile.am: + * orbsvcs/tests/HTIOP/BiDirectional/Makefile.am: + * orbsvcs/tests/HTIOP/Hello/Makefile.am: + * orbsvcs/tests/IOR_MCast/Makefile.am: + * orbsvcs/tests/ImplRepo/Makefile.am: + * orbsvcs/tests/ImplRepo/NameService/Makefile.am: + * orbsvcs/tests/ImplRepo/scale/Makefile.am: + * orbsvcs/tests/InterfaceRepo/Makefile.am: + * orbsvcs/tests/InterfaceRepo/Application_Test/Makefile.am: + * orbsvcs/tests/InterfaceRepo/IDL3_Test/Makefile.am: + * orbsvcs/tests/InterfaceRepo/IFR_Inheritance_Test/Makefile.am: + * orbsvcs/tests/InterfaceRepo/IFR_Test/Makefile.am: + * orbsvcs/tests/InterfaceRepo/Latency_Test/Makefile.am: + * orbsvcs/tests/InterfaceRepo/Persistence_Test/Makefile.am: + * orbsvcs/tests/Interoperable_Naming/Makefile.am: + * orbsvcs/tests/LoadBalancing/Makefile.am: + * orbsvcs/tests/LoadBalancing/GenericFactory/Makefile.am: + * orbsvcs/tests/LoadBalancing/GenericFactory/Application_Controlled/Makefile.am: + * orbsvcs/tests/LoadBalancing/GenericFactory/Infrastructure_Controlled/Makefile.am: + * orbsvcs/tests/LoadBalancing/GenericFactory/Manage_Object_Group/Makefile.am: + * orbsvcs/tests/LoadBalancing/LoadMonitor/Makefile.am: + * orbsvcs/tests/LoadBalancing/LoadMonitor/CPU/Makefile.am: + * orbsvcs/tests/Log/Makefile.am: + * orbsvcs/tests/Log/Basic_Log_Test/Makefile.am: + * orbsvcs/tests/Miop/Makefile.am: + * orbsvcs/tests/Miop/McastHello/Makefile.am: + * orbsvcs/tests/Notify/Makefile.am: + * orbsvcs/tests/Notify/Basic/Makefile.am: + * orbsvcs/tests/Notify/Blocking/Makefile.am: + * orbsvcs/tests/Notify/Destroy/Makefile.am: + * orbsvcs/tests/Notify/Discarding/Makefile.am: + * orbsvcs/tests/Notify/Driver/Makefile.am: + * orbsvcs/tests/Notify/MT_Dispatching/Makefile.am: + * orbsvcs/tests/Notify/Ordering/Makefile.am: + * orbsvcs/tests/Notify/PluggableTopology/Makefile.am: + * orbsvcs/tests/Notify/RT_lib/Makefile.am: + * orbsvcs/tests/Notify/Reconnecting/Makefile.am: + * orbsvcs/tests/Notify/Sequence_Multi_ETCL_Filter/Makefile.am: + * orbsvcs/tests/Notify/Sequence_Multi_Filter/Makefile.am: + * orbsvcs/tests/Notify/Structured_Filter/Makefile.am: + * orbsvcs/tests/Notify/Structured_Multi_Filter/Makefile.am: + * orbsvcs/tests/Notify/Test_Filter/Makefile.am: + * orbsvcs/tests/Notify/XML_Persistence/Makefile.am: + * orbsvcs/tests/Notify/lib/Makefile.am: + * orbsvcs/tests/Notify/performance-tests/Makefile.am: + * orbsvcs/tests/Notify/performance-tests/Filter/Makefile.am: + * orbsvcs/tests/Notify/performance-tests/RedGreen/Makefile.am: + * orbsvcs/tests/Notify/performance-tests/Throughput/Makefile.am: + * orbsvcs/tests/Property/Makefile.am: + * orbsvcs/tests/Redundant_Naming/Makefile.am: + * orbsvcs/tests/Sched/Makefile.am: + * orbsvcs/tests/Sched_Conf/Makefile.am: + * orbsvcs/tests/Security/Makefile.am: + * orbsvcs/tests/Security/BiDirectional/Makefile.am: + * orbsvcs/tests/Security/Big_Request/Makefile.am: + * orbsvcs/tests/Security/Callback/Makefile.am: + * orbsvcs/tests/Security/Crash_Test/Makefile.am: + * orbsvcs/tests/Security/MT_IIOP_SSL/Makefile.am: + * orbsvcs/tests/Security/MT_SSLIOP/Makefile.am: + * orbsvcs/tests/Security/Secure_Invocation/Makefile.am: + * orbsvcs/tests/Security/ssliop_corbaloc/Makefile.am: + * orbsvcs/tests/Simple_Naming/Makefile.am: + * orbsvcs/tests/Time/Makefile.am: + * orbsvcs/tests/Trading/Makefile.am: + * orbsvcs/tests/ior_corbaname/Makefile.am: + * orbsvcs/tests/tests_svc_loader/Makefile.am: + * performance-tests/Makefile.am: + * performance-tests/Anyop/Makefile.am: + * performance-tests/CSD_Strategy/Makefile.am: + * performance-tests/CSD_Strategy/TestApps/Makefile.am: + * performance-tests/CSD_Strategy/TestInf/Makefile.am: + * performance-tests/CSD_Strategy/TestServant/Makefile.am: + * performance-tests/Callback/Makefile.am: + * performance-tests/Cubit/Makefile.am: + * performance-tests/Cubit/TAO/Makefile.am: + * performance-tests/Cubit/TAO/DII_Cubit/Makefile.am: + * performance-tests/Cubit/TAO/IDL_Cubit/Makefile.am: + * performance-tests/Cubit/TAO/MT_Cubit/Makefile.am: + * performance-tests/Latency/Makefile.am: + * performance-tests/Latency/AMH_Single_Threaded/Makefile.am: + * performance-tests/Latency/AMI/Makefile.am: + * performance-tests/Latency/Collocation/Makefile.am: + * performance-tests/Latency/DII/Makefile.am: + * performance-tests/Latency/DSI/Makefile.am: + * performance-tests/Latency/Deferred/Makefile.am: + * performance-tests/Latency/Single_Threaded/Makefile.am: + * performance-tests/Latency/Thread_Per_Connection/Makefile.am: + * performance-tests/Latency/Thread_Pool/Makefile.am: + * performance-tests/Memory/Makefile.am: + * performance-tests/Memory/IORsize/Makefile.am: + * performance-tests/Memory/Single_Threaded/Makefile.am: + * performance-tests/POA/Makefile.am: + * performance-tests/POA/Create_Reference/Makefile.am: + * performance-tests/POA/Demux/Makefile.am: + * performance-tests/POA/Implicit_Activation/Makefile.am: + * performance-tests/POA/Object_Creation_And_Registration/Makefile.am: + * performance-tests/Pluggable/Makefile.am: + * performance-tests/Protocols/Makefile.am: + * performance-tests/RTCorba/Makefile.am: + * performance-tests/RTCorba/Multiple_Endpoints/Makefile.am: + * performance-tests/RTCorba/Multiple_Endpoints/Common/Makefile.am: + * performance-tests/RTCorba/Multiple_Endpoints/Orb_Per_Priority/Makefile.am: + * performance-tests/RTCorba/Multiple_Endpoints/Single_Endpoint/Makefile.am: + * performance-tests/RTCorba/Oneways/Makefile.am: + * performance-tests/RTCorba/Oneways/Reliable/Makefile.am: + * performance-tests/RTCorba/Thread_Pool/Makefile.am: + * performance-tests/Sequence_Latency/Makefile.am: + * performance-tests/Sequence_Latency/AMH_Single_Threaded/Makefile.am: + * performance-tests/Sequence_Latency/AMI/Makefile.am: + * performance-tests/Sequence_Latency/DII/Makefile.am: + * performance-tests/Sequence_Latency/DSI/Makefile.am: + * performance-tests/Sequence_Latency/Deferred/Makefile.am: + * performance-tests/Sequence_Latency/Single_Threaded/Makefile.am: + * performance-tests/Sequence_Latency/Thread_Per_Connection/Makefile.am: + * performance-tests/Sequence_Latency/Thread_Pool/Makefile.am: + * performance-tests/Throughput/Makefile.am: + * tao/Makefile.am: + * utils/Makefile.am: + * utils/catior/Makefile.am: + * utils/nslist/Makefile.am: + + Regenerate with latest MPC and *.mpc/*.mpb changes. + +Fri Apr 21 09:25:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Transport_Connector.cpp: + Removed the fix for bug 2417, according to the test stats things + didn't got fixed. + +Fri Apr 21 08:18:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/String_Traits_Base_T.h: + Disabled the warning when wchar_t is not a native type. This + makes the vxworks logs unreadable. + +Fri Apr 21 08:11:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/tests/Trading/TTest.idl: + Use the CORBA predefined sequence types + +Fri Apr 21 07:38:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/orbsvcs/IFRService/IFR_BaseS.cpp: + * orbsvcs/orbsvcs/IFRService/IFR_BasicS.cpp: + * orbsvcs/orbsvcs/IFRService/IFR_ComponentsS.cpp: + * orbsvcs/orbsvcs/IFRService/IFR_ExtendedS.cpp: + * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Connector.cpp: + * orbsvcs/orbsvcs/FaultTolerance/FT_ClientPolicy_i.inl: + * orbsvcs/orbsvcs/Notify/ETCL_Filter.h: + * orbsvcs/orbsvcs/Notify/EventTypeSeq.cpp: + * examples/Kokyu_dsrt_schedulers/FP_Scheduler.cpp: + * examples/Kokyu_dsrt_schedulers/MIF_Scheduler.cpp: + * examples/Kokyu_dsrt_schedulers/MUF_Scheduler.cpp: + * tests/Smart_Proxies/Collocation/Smart_Proxy_Impl.cpp: + Removed usage of ACE_NESTED_CLASS + +Thu Apr 20 14:41:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/extra_core.mpb: + * tao/tao.mpc: + Moved ServicesC.cpp to tao.mpc + +Thu Apr 20 14:24:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Bounded_Sequence_CDR_T.h: + * tao/Unbounded_Sequence_CDR_T.h: + Include orbconf.h instead of one of the sequence header files + +Thu Apr 20 13:37:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Asynch_Reply_Dispatcher_Base.{h,cpp}: + * tao/ObjectKey_Table.cpp: + * tao/Refcounted_ObjectKey.{h,cpp,inl}: + Made the refcounts CORBA::ULong and only return the refcount + from the incr/decr methods when really needed, using the refcount + form external is always tricky. Fixes bugzilla bug 2505. + +Thu Apr 20 12:28:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + Integrated ondemand branch to cvs head. This makes an ondemand write + possible. The user specified maximum is at this moment not a hard + maximum, it is more an indication how large the GIOP fragments + should become, we can send out smaller and larger fragments if needed. + + * tao/tests/Ondemand_Write/*: + New test + + Mon Apr 3 12:30:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/On_Demand_Fragmentation_Strategy.cpp: + Print the debug message after we padded it so that sizes do match + in the logs + + * tao/GIOP_Message_Base.cpp: + For fragments also retrieve the request/reply id + + * tests/Ondemand_Write: + Simple test for ondemand write, needs now inspection of output to + check if things work ok + + Mon Apr 3 07:19:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/CDR.h: + Commented out write_octet_array decleration, there is no + implementation yet. + + Thu Mar 30 13:02:18 UTC 2006 Chad Elliott <elliott_c@ociweb.com> + + * orbsvcs/orbsvcs/PortableGroup/UIPMC_Transport.cpp: + Added the missing transport parameter to the TAO_GIOP_Message_Base + constructor. + + Thu Mar 30 12:41:17 UTC 2006 Chad Elliott <elliott_c@ociweb.com> + + * tao/CDR.cpp: + * tao/On_Demand_Fragmentation_Strategy.cpp: + + Fixed checks for return values. + + Wed Mar 22 20:13:22 UTC 2006 Ossama Othman <ossama@dre.vanderbilt.edu> + + * tao/Messaging/Asynch_Invocation.cpp (remote_invocation): + + Added missing GIOP fragmentation support. + + Wed Mar 22 13:53:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/ORB_Core.h: + Fixed warning of the Intel compiler + + Wed Mar 22 01:33:47 UTC 2006 Ossama Othman <ossama@dre.vanderbilt.edu> + + * tao/CDR.cpp: + * tao/CDR.h: + * tao/CDR.i: + * tao/GIOP_Message_Base.cpp: + * tao/GIOP_Message_Base.h: + * tao/GIOP_Message_Generator_Parser.h: + * tao/GIOP_Message_Generator_Parser_10.cpp: + * tao/GIOP_Message_Generator_Parser_10.h: + * tao/GIOP_Message_Generator_Parser_12.cpp: + * tao/GIOP_Message_Generator_Parser_12.h: + * tao/GIOP_Message_Lite.cpp: + * tao/GIOP_Message_Lite.h: + * tao/ORB_Core.cpp: + * tao/On_Demand_Fragmentation_Strategy.cpp: + * tao/Pluggable_Messaging.h: + * tao/Remote_Invocation.cpp: + * tao/Synch_Invocation.cpp: + * tao/TAO_Server_Request.cpp: + * tao/default_resource.cpp: + * tao/PortableServer/Upcall_Wrapper.cpp: + + Added remaining code necessary to send fragments through the + underlying transport. + + Added missing outgoing GIOP reply fragment support. + + Tue Mar 21 22:18:45 UTC 2006 Ossama Othman <ossama@dre.vanderbilt.edu> + + * docs/Options.html: + + Document new "-ORBMaxMessageSize" ORB option. + + Tue Mar 21 15:16:43 UTC 2006 Ossama Othman <ossama@dre.vanderbilt.edu> + + * tao/GIOP_Message_Base.cpp (set_giop_flags): + + Cast CDR stream buffer to an array of octets. + + * tao/GIOP_Message_Base.h (set_giop_flags): + + Added missing method declaration. + + * tao/Resource_Factory.h (fragmentation_strategy): + * tao/default_resource.cpp: + * tao/default_resource.h: + + Made factory method name consistent with existing naming + convention, i.e. create_fragmentation_strategy(). + + Corrected return value. It should have been + auto_ptr<TAO_GIOP_Fragmentation_Strategy>, not + TAO_GIOP_Fragmentation_Strategy *. + + * tao/IIOP_Transport.cpp: + * tao/Strategies/DIOP_Transport.cpp: + * tao/Strategies/SCIOP_Transport.cpp: + * tao/Strategies/SHMIOP_Transport.cpp: + * tao/Strategies/UIOP_Transport.cpp: + * orbsvcs/orbsvcs/HTIOP/HTIOP_Transport.cpp: + * orbsvcs/orbsvcs/SSLIOP/SSLIOP_Transport.cpp: + + The TAO_GIOP_Message_Base constructor now accepts a pointer to + TAO_Transport parameter. Updated constructor call accordingly. + + * tao/CDR.h: + * tao/CDR.i: + * tao/CDR.cpp: + + Added missing fragmentation-enabling constructor and + fragmentation flag accessors. + + Removed legacy initial implementa code. Addresses build + errors. + + * tao/GIOP_Fragmentation_Strategy.h (TAO_GIOP_Fragmentation_Strategy): + + Export to allow users to provide their own implementation + through the resource factory. + + (fragment): + + Return an "int" instead of "void". Allows the error status of + the underlying transport send to be propagated up the stack. + + * tao/Null_Fragmentation_Strategy.h (fragment): + * tao/Null_Fragmentation_Strategy.cpp (fragment): + * tao/On_Demand_Fragmentation_Strategy.h (fragment): + * tao/On_Demand_Fragmentation_Strategy.cpp (fragment): + + Likewise. + + * tao/operation_details.cpp (marshal_args): + + Mark the CDR as having no other fragments to send after all + arguments have been marshaled, not before the last one is + marshaled. + + * ORB_Core.h (fragmentation_Strategy): + + Added missing transport parameter. + + Removed const qualifier. The resource_factory() accessor isn't + a const method. + + * ORB_Core.cpp (fragmentation_strategy): + + Likewise. + + * params.h: + * params.i: + * params.cpp: + + Added missing max_message_size attribute. + + Fri Mar 17 10:59:02 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Resource_Factory.h: + Added missing include of Basic_Types and added forward declarations + + * tao/On_Demand_Fragmentation_Strategy.h: + Fixed copy constructor/assignment operators + + * tao/CDR.h: + Removed do_fragmentation method, there is no implementation, added + fragment_stream + + * tao/CDR.i: + Fixed typo + + * tao/default_resource.cpp: + Added missing includes and updated signature of + create_fragmentation_strategy to match header file + + Fri Mar 17 07:48:02 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + Updated code after update from Ossama Othman + + * tao/On_Demand_Fragmentation_Strategy.{h,cpp}: + New files + + * tao/default_resource.{h,cpp}: + Added create_fragmentation_strategy + + * tao/GIOP_Fragmentation_Strategy.h: + Doxygen improvements + + * tao/Resource_Factory.h: + Added pure virtual fragmentation_strategy method + + * tao/ORB_Core.{h,cpp}: + Added fragmentation_strategy accessor method + + * tao/GIOP_Message_Base.{h,cpp}: + Added TAO_Transport to the constructor arguments + + * tao/GIOP_Message_Base.cpp: + Added come comments + + * tao/True_Fragmentation_Strategy.{h,cpp}: + Removed again, replaced by On_Demand so far as I can tell + + * tao/tao.mpc: + Added new files + + Thu Mar 16 07:48:02 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + Integrated first set of code from Ossama Othman + + * tao/GIOP_Fragmentation_Strategy.{h,cpp}: + * tao/Null_Fragmentation_Strategy.{h,cpp}: + * tao/True_Fragmentation_Strategy.{h,cpp}: + New files + + * tao/operation_details.cpp: + When marshaling the last argument put this information on the + cdr_stream + + * tao/GIOP_Message_Base.cpp: + Some refactoring + + * tao/CDR.{h,cpp,i}: + Call fragment_stream as part of the streaming calls + +Thu Apr 20 11:50:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/orbsvcs/sfp.idl: + Use CORBA::OctetSeq and CORBA::ULongSeq + + * tao/Strategies/SCIOP_Profile.cpp: + Fixed compile error + +Thu Apr 20 08:32:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/examples/FaultTolerance/RolyPoly/ReplicaController.cpp: + Fixed compile error + +Thu Apr 20 07:15:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/orbsvcs/SSLIOP/ssl_endpointsC.h: + Fixed template instantiation + +Wed Apr 19 18:24:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/UShortSeqC.h: + * tao/OctetSeqC.h: + Do an explicit export of the base template + +Wed Apr 19 16:01:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Messaging/ExceptionHolderC.{h,cpp}: + Added constructor that accepts all values as generated now by the + IDL compiler + + * tao/Messaging/ExceptionHolder_i.cpp: + Use the new constructor + +Wed Apr 19 14:19:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Exception.h: + Made the copy constructor public again, vc7.1 complains when + it is protected. Made a todo in this file again, have to retest + this later. + +Wed Apr 19 13:28:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/orbsvcs/SSLIOP/SSLIOP_CredentialsAcquirer.cpp: + * examples/Advanced/ch_12/icp.cpp: + * examples/Advanced/ch_21/icp.cpp: + * examples/Advanced/ch_18/icp.cpp: + * examples/Advanced/ch_8_and_10/icp.cpp: + * orbsvcs/orbsvcs/SSLIOP/params_dup.h: + Removed workarounds for vc6 + +Wed Apr 19 13:16:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Valuetype/AbstractBase.cpp: + Use true/false, const improvements + + * tao/Valuetype/AbstractBase.cpp: + Use C++ cast instead of C cast + +Wed Apr 19 13:03:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/AnyTypeCode/Any_Unknown_IDL_Type.h: + Removed not needed forward declarations + + * tao/AnyTypeCode/Any.cpp: + * tao/AnyTypeCode/Any_Impl.cpp: + * tao/BiDir_GIOP/BiDirPolicy_Validator.cpp: + Use false/true instead of 0/1 for bool + + * tao/PortableServer/Object_Adapter.h: + Don't export poa_name_iterator and iteratable_poa_name + + * tao/PortableServer/Object_Adapter.cpp: + Use true/false and when the object adapter can't be found + throw a OBJECT_NOT_EXIST with minor code 2 + +Wed Apr 19 12:56:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Transport_Timer.h: + No need to export this class from the TAO lib + + * tao/ORB.h: + Use false for the default of the shutdown method + +Wed Apr 19 12:51:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * examples/RTScheduling/MIF_Scheduler/MIF_Scheduler.mpc: + * examples/RTScheduling/Fixed_Priority_Scheduler/Fixed_Priority_Scheduler.mpc: + Made these projects dependent on each other to make sure + that they don't build in parallel and generate the same idl + file twice at the same moment. Thanks to Chad Elliot for the + info how to do this the easiest + +Wed Apr 19 12:42:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * TAO_IDL/be/be_visitor_interface/tie_sh.cpp: + Generate doxygen documentation style and use true instead of 1 + + * TAO_IDL/be/be_visitor_valuetype/any_op_cs.cpp: + Generate true for boolean instead of 1 + + * TAO_IDL/be/be_visitor_valuetype/field_ch.cpp: + Generate also argument names in the header file so that doxygen + can parse IDL generated code + +Wed Apr 19 12:38:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPushConsumer.{h,cpp,i}: + * orbsvcs/orbsvcs/CosEvent/CEC_ProxyPushSupplier.{h,i}: + Use bool and prefix increment/decrement + +Wed Apr 19 11:58:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Big_Oneways/run_test.pl: + * tests/Hello/run_test.pl: + * performance-tests/Throughput/run_test.pl: + Check the return value of spawn, speedsup the builds when no + executable is build + + * performance-tests/Throughput/Receiver.cpp: + Prefix increment + + * performance-tests/Throughput/Receiver_Factory.cpp: + Initialise pointer with 0 + + * performance-tests/Throughput/Throughput.mpc: + Simplified + +Wed Apr 19 11:39:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + Integrated iioptbranch. This fixes bugzilla 2467 + + * tao/RTPortableServer/RT_Servant_Dispatcher.cpp: + No need to include IIOP files, just use the base classes + + Wed Mar 29 08:01:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/default_resource.cpp + * tao/IIOP_Acceptor.cpp + * tao/IIOP_Acceptor.h + * tao/IIOP_Acceptor.i + * tao/IIOP_Connection_Handler.cpp + * tao/IIOP_Connection_Handler.h + * tao/IIOP_Connector.cpp + * tao/IIOP_Connector.h + * tao/IIOP_Endpoint.cpp + * tao/IIOP_Endpoint.h + * tao/IIOP_Endpoint.i + * tao/IIOP_Factory.cpp + * tao/IIOP_Factory.h + * tao/IIOP_Lite_Factory.cpp + * tao/IIOP_Lite_Factory.h + * tao/IIOP_Profile.cpp + * tao/IIOP_Profile.h + * tao/IIOP_Transport.cpp + * tao/IIOP_Transport.h + * tao/orbconf.h + * tao/TAO_Internal.cpp + Added TAO_HAS_IIOP. This is default set to 1 but can be overridden + in the config.h file to 0 meaning we don't support IIOP. This is + usefull for embedded systems that support one of the other + pluggable protocols and don't need IIOP support at all. With + TAO_HAS_IIOP set to 0 not everything will compile, just the + core libs itself. + +Wed Apr 19 07:48:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + Integrated sequpdate3 branch. Thanks to Carlos O'Ryan for the initial + work for this new sequence implementation which I merged to cvs head + and finished + + This fixes the following bugzilla entries: + 2492 - Simplify TAO_Seq_Out_T + 2493 - Simplify _reset method on union generated code + 2352 - Valuefactory operations not safe + 2353 - Valuefactories stored per process instead of per orb + 2349 - ORB::destroy() should throw BAD_INV_ORDER if called during a + 2315 - Reimplement (w)string_var/_out as templates + 1989 - Footprint reduction issue, split Sequence files + 2273 - Deprecate old AMI support + 2300 - Simplify serialize/deserialize of sequences + 2299 - Reimplement string/wstring managers as templates + 1936 - Unnecessary usage of virtual functions in sequence implementation + 1673 - operator[] of sequence<string> returns TAO_SeqElem_String_Manager + instead of TAO_String_Manager. Thanks to Mark Paulus <mark dot paulus + at mci dot com> for reporting this one. + 1930 - Assignment operator for sequences is not exception safe. + 1931 - The length() member function for sequences is not exception-safe. + 1933 - Incomplete implementation of freebuf() for reference types. + 1934 - const version of operator[] for string sequences allows assignment + 1938 - Possible incorrect duplication in sequences of references + 1928 - Assignment from T_mgr to sequence elements does not duplicate + 2417 - Double delete on Transport when using oneways with sync_none + Thanks to Jan Ohlenburg <jan dot ohlenburg at fit dot fraunhofer dot de> + for reporting this. + 2355 - oneway op. with timeout crashes client due to server termination + Thans to Jan Zima <jan dot zima at sofis dot cz> for reporting this. + + Also did several const changes throughout the code + + Fri Apr 7 08:03:12 UTC 2006 Kees van Marle <kvmarle@remedy.nl> + + * tao/Valuetype_Adapter_Factory.{h,cpp}: + New files, value type adapter factory + + * tao/tao.mpc: + Added new files + + * tao/Valuetype/Valuetype_Adapter_Factory_Impl.h: + Value type adapter factory implementation + + * tao/Valuetype/ValueFactory_Map.{h,cpp}: + Map isn't a singleton anymore and guard access with a mutex + + * tao/Valuetype/Valuetype_Adapter_Impl.{h,cpp}: + The value type adapter isn't loaded with service configurator + anymore, the value type factory is now the one we load on demand + + * tao/ORB.cpp: + Updated the value type methods to use the new ORB_Core method + to get the valuetype adapter, is the ORB_Core can't get the + adapter it will throw already the internal exception + + * tao/ORB_Core.{h,cpp,i}: + Get the value type adapter factory with svc conf instead of the + adapter itself. Create a unique instance per orb. + + * tao/AnyTypeCode/Any_Unknown_IDL_Type.cpp: + * tao/AnyTypeCode/append.cpp: + * tao/AnyTypeCode/skip.cpp: + Changed the way we get the valuetype adapter + + Store the value type factories per orb and made things thread safe. + This fixes bugzilla bugs 2352 and 23253. + + Thu Apr 6 09:17:25 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tao/Transport_Connector.cpp: + Applied fix of bug 2417, let us see what the results are in the + branch build + + Thu Apr 6 09:08:25 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * examples/POA/FindPOA/FindPOA.cpp: + * tests/Bug_2349_Regression/client.cpp: + * tests/Bug_2349_Regression/foo.idl: + * tests/Bug_2349_Regression/server.cpp: + Improved tests + + * tao/Adapter_Registry.{h,cpp}: + Remove the empty throw spec for close and check_close. This way + exceptions from lower layers are propagated up. + + * tao/ORB_Core.{h,cpp}: + Removed empty throw spec from shutdown, if there are exceptions from + a lower layer and as a result we can't shutdown, let the user be + aware of it. This fixes bugzilla bug 2349 + + Mon Apr 3 07:59:25 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * TAO/IDL/be/be_visitor_union/discriminant_ci.cpp + * TAO/IDL/be/be_visitor_union/union_ch.cpp + * TAO/IDL/be/be_visitor_union/union_cs.cpp + * TAO/IDL/be/be_visitor_union_branch/public_ci.cpp + * tao/GIOPC.{h,cpp,inl}: + Removed arguments from the _reset method on the union generated, + these are not used. + + Sun Apr 2 18:56:25 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * TAO_IDL/be/be_visitor_exception/exception_ch.cpp + * TAO_IDL/be/be_visitor_sequence/sequence_ch.cpp + * TAO_IDL/be/be_visitor_valuebox/valuebox_ch.cpp + * tao/BooleanSeqC.h + * tao/CONV_FRAMEC.h + * tao/CharSeqC.h + * tao/DomainC.h + * tao/DoubleSeqC.h + * tao/FloatSeqC.h + * tao/IIOPC.h + * tao/IIOP_EndpointsC.h + * tao/IOP_IORC.h + * tao/LongDoubleSeqC.h + * tao/LongLongSeqC.h + * tao/LongSeqC.h + * tao/Messaging_PolicyValueC.h + * tao/ORB.h + * tao/ObjectIdListC.h + * tao/Object_KeyC.h + * tao/OctetSeqC.h + * tao/Policy_ForwardC.h + * tao/Seq_Out_T.h + * tao/Seq_Out_T.inl + * tao/ServicesC.h + * tao/ShortSeqC.h + * tao/StringSeqC.h + * tao/ULongLongSeqC.h + * tao/ULongSeqC.h + * tao/UShortSeqC.h + * tao/WCharSeqC.h + * tao/WStringSeqC.h + * tao/AnyTypeCode/AnySeqC.h + * tao/AnyTypeCode/DynamicC.h + * tao/DynamicAny/DynamicAnyC.h + * tao/IFR_Client/IFR_BaseC.h + * tao/IFR_Client/IFR_BasicC.h + * tao/IFR_Client/IFR_ComponentsC.h + * tao/IFR_Client/IFR_ExtendedC.h + * tao/IORManipulation/IORC.h + * tao/ImR_Client/ImplRepoC.h + * tao/ObjRefTemplate/ObjectReferenceTemplateC.h + * tao/PortableServer/PortableServerC.h + * tao/RTCORBA/RTCORBAC.h + * tao/RTScheduling/RTSchedulerC.h + * tao/Strategies/sciop_endpointsC.h + * tao/Strategies/uiop_endpointsC.h + Simplified TAO_Seq_Out_T to just use one template argument. Updated + the IDL compiler for this. Also use false for the release argument + of generated sequence constructors and add an argument to + _tao_encode/_tao_decode when generated in a header file to help + doxygen. + + * TAO_Objref_Out_T.{h,inl}: + Assinging _var to _out is not allowed according to the spec + + * tao/Messaging/*: + Removed old AMI mapping. Fixes partly bugzilla bug 2273, need to + update the IDL compiler yet + + * tao/Exception.h: + Moved constructors, assignment operator and copy constructor to + protected now vc6 has been dropped. Also moved + create_system_exception from TAO_Exceptions class to TAO namespace + + * tao/Messaging/Messaging.cpp: + * tao/Sync_Invocation.cpp: + * tao/SystemException.cpp: + Updated because of the create_system_exception move + + * tao/String_Manager_T.h: + Renamed String_Manager to String_Manager_T, this way we can have + TAO::String_Manager and TAO::WString_Manager. These replace + TAO_String_Manager and TAO_WString_Manager + + * CIAO/tools/Config_Handlers/DnC_Dump.cpp + * CIAO/tools/Config_Handlers/DnC_Dump.h + * TAO_IDL/be/be_visitor_array/array.cpp + * TAO_IDL/be/be_visitor_field/field_ch.cpp + * orbsvcs/PSS/PSDL_Exception_Visitor.cpp + * orbsvcs/PSS/PSDL_Node.h + * orbsvcs/PSS/PSDL_Struct_Visitor.cpp + * orbsvcs/orbsvcs/HTIOP/htiop_endpointsC.h + * orbsvcs/orbsvcs/Metrics/Metrics_UpcallMonitor_T.h + * orbsvcs/orbsvcs/Metrics/Metrics_UpcallMonitor_T.i + * orbsvcs/orbsvcs/Notify/ETCL_Filter.cpp + * orbsvcs/orbsvcs/PortableGroup/PG_Object_Group.h + * orbsvcs/orbsvcs/Trader/Constraint_Nodes.cpp + * orbsvcs/orbsvcs/Trader/Constraint_Nodes.h + * tao/CORBA_String.h + * tao/IIOPC.h + * tao/IIOP_EndpointsC.h + * tao/IOP_IORC.h + * tao/String_Manager_T.h + * tao/String_Traits_Base_T.h + * tao/Tagged_Profile.h + * tao/DynamicAny/DynamicAnyC.h + * tao/IFR_Client/IFR_BaseC.h + * tao/IFR_Client/IFR_BasicC.h + * tao/IFR_Client/IFR_ComponentsC.h + * tao/IFR_Client/IFR_ExtendedC.h + * tao/ImR_Client/ImplRepoC.h + * tao/PI/ORBInitInfoC.h + * tao/Strategies/sciop_endpointsC.h + * tao/Strategies/uiop_endpointsC.h + Updated all these files because of TAO::String_Manager and + TAO::WString_Manager introduction + + * tests/Sequence_Unit_Tests/mock_reference.cpp: + * tests/Sequence_Unit_Tests/mock_reference.hpp: + Added counter to count number of marshal calls + + * tests/Sequence_Unit_Tests/bounded_sequence_cdr_ut.cpp: + * tests/Sequence_Unit_Tests/unbounded_sequence_cdr_ut.cpp: + Check the right counter, should be marshal. + + * tao/PI/ClientRequestDetails.cpp: + * tao/PI_Server/ServerRequestDetails.cpp: + Added missing duplicate call + + * TAO_IDL/be/be_interface.cpp + * TAO_IDL/be/be_visitor_component/component_cs.cpp + * TAO_IDL/be/be_visitor_interface/interface_cs.cpp + * TAO_IDL/be/be_visitor_interface/tie_si.cpp + * TAO_IDL/be/be_visitor_root/root.cpp + * tao/CurrentC.cpp + * tao/DomainC.cpp + * tao/ORB_Core.cpp + * tao/Object.cpp + * tao/Object_Ref_Table.cpp + * tao/PolicyC.cpp + * tao/Pseudo_VarOut_T.cpp + * tao/Pseudo_VarOut_T.inl + * tao/TAOC.cpp + * tao/AnyTypeCode/Any_Array_Impl_T.cpp + * tao/AnyTypeCode/Any_Basic_Impl.cpp + * tao/AnyTypeCode/Any_Basic_Impl_T.cpp + * tao/AnyTypeCode/Any_Dual_Impl_T.cpp + * tao/AnyTypeCode/Any_Impl.cpp + * tao/AnyTypeCode/Any_Impl_T.cpp + * tao/AnyTypeCode/Any_Special_Impl_T.cpp + * tao/AnyTypeCode/Any_Unknown_IDL_Type.cpp + * tao/AnyTypeCode/TypeCode.cpp + * tao/AnyTypeCode/TypeCode.inl + * tao/BiDir_GIOP/BiDirPolicyC.cpp + * tao/CSD_Framework/CSD_Default_Servant_Dispatcher.cpp + * tao/CSD_Framework/CSD_FrameworkC.cpp + * tao/CSD_Framework/CSD_POA.cpp + * tao/CodecFactory/IOP_CodecC.cpp + * tao/DynamicAny/DynamicAnyC.cpp + * tao/DynamicInterface/Dynamic_Implementation.cpp + * tao/DynamicInterface/ExceptionList.cpp + * tao/DynamicInterface/Request.cpp + * tao/DynamicInterface/Server_Request.cpp + * tao/IFR_Client/IFR_BaseC.cpp + * tao/IFR_Client/IFR_BasicC.cpp + * tao/IFR_Client/IFR_Client_Adapter_Impl.cpp + * tao/IFR_Client/IFR_ComponentsC.cpp + * tao/IFR_Client/IFR_ExtendedC.cpp + * tao/IORInterceptor/IORInfoC.cpp + * tao/IORInterceptor/IORInterceptorC.cpp + * tao/IORManipulation/IORC.cpp + * tao/IORManipulation/IORManipulation.cpp + * tao/IORTable/IORTableC.cpp + * tao/IORTable/Table_Adapter.cpp + * tao/ImR_Client/ImplRepoC.cpp + * tao/ImR_Client/ServerObjectC.cpp + * tao/Messaging/MessagingC.cpp + * tao/Messaging/Messaging_No_ImplC.cpp + * tao/Messaging/Messaging_RT_PolicyC.cpp + * tao/Messaging/Messaging_SyncScope_PolicyC.cpp + * tao/Messaging/PollableC.cpp + * tao/Messaging/TAO_ExtC.cpp + * tao/PI/ClientRequestInfoC.cpp + * tao/PI/ClientRequestInterceptorC.cpp + * tao/PI/InterceptorC.cpp + * tao/PI/ORBInitInfo.cpp + * tao/PI/ORBInitInfoC.cpp + * tao/PI/ORBInitializerC.cpp + * tao/PI/PICurrentC.cpp + * tao/PI/PolicyFactoryC.cpp + * tao/PI/PolicyFactory_Registry.cpp + * tao/PI/ProcessingModePolicyC.cpp + * tao/PI/RequestInfoC.cpp + * tao/PI_Server/ServerRequestInfoC.cpp + * tao/PI_Server/ServerRequestInterceptorC.cpp + * tao/PortableServer/AdapterActivatorC.cpp + * tao/PortableServer/IdAssignmentPolicyC.cpp + * tao/PortableServer/IdUniquenessPolicyC.cpp + * tao/PortableServer/ImplicitActivationPolicyC.cpp + * tao/PortableServer/LifespanPolicyC.cpp + * tao/PortableServer/Object_Adapter.cpp + * tao/PortableServer/POAManager.cpp + * tao/PortableServer/POAManagerC.cpp + * tao/PortableServer/PS_CurrentC.cpp + * tao/PortableServer/PortableServerC.cpp + * tao/PortableServer/RequestProcessingPolicyC.cpp + * tao/PortableServer/Root_POA.cpp + * tao/PortableServer/ServantActivatorC.cpp + * tao/PortableServer/ServantLocatorC.cpp + * tao/PortableServer/ServantManagerC.cpp + * tao/PortableServer/ServantRetentionPolicyC.cpp + * tao/PortableServer/ThreadPolicyC.cpp + * tao/RTCORBA/RTCORBAC.cpp + * tao/RTPortableServer/RTPortableServerC.cpp + * tao/RTScheduling/Current.cpp + * tao/RTScheduling/RTSchedulerC.cpp + * tao/TypeCodeFactory/TypeCodeFactoryC.cpp + * tao/Utils/Server_Main.cpp + * tao/Valuetype/AbstractBase.cpp: + Use :: before ::CORBA + + * TAO_IDL/be/be_interface.cpp + * tao/DomainC.cpp + * tao/DomainC.inl + * tao/GIOPC.h + * tao/PolicyC.cpp + * tao/PolicyC.h + * tao/PolicyC.inl + * tao/WrongTransactionC.cpp + * tao/AnyTypeCode/Alias_TypeCode.inl + * tao/AnyTypeCode/Alias_TypeCode_Static.inl + * tao/AnyTypeCode/BoundsC.cpp + * tao/AnyTypeCode/Empty_Param_TypeCode.inl + * tao/AnyTypeCode/Enum_TypeCode.inl + * tao/AnyTypeCode/Enum_TypeCode_Static.inl + * tao/AnyTypeCode/Fixed_TypeCode.inl + * tao/AnyTypeCode/Objref_TypeCode.inl + * tao/AnyTypeCode/Objref_TypeCode_Static.inl + * tao/AnyTypeCode/Recursive_Type_TypeCode.cpp + * tao/AnyTypeCode/Sequence_TypeCode.inl + * tao/AnyTypeCode/Sequence_TypeCode_Static.inl + * tao/AnyTypeCode/String_TypeCode.inl + * tao/AnyTypeCode/String_TypeCode_Static.inl + * tao/AnyTypeCode/Struct_TypeCode.inl + * tao/AnyTypeCode/Struct_TypeCode_Static.inl + * tao/AnyTypeCode/Union_TypeCode.inl + * tao/AnyTypeCode/Union_TypeCode_Static.inl + * tao/AnyTypeCode/Value_TypeCode.inl + * tao/AnyTypeCode/Value_TypeCode_Static.inl + * tao/BiDir_GIOP/BiDir_Policy_i.cpp + * tao/CodecFactory/IOP_CodecC.cpp + * tao/Domain/DomainS.cpp + * tao/DynamicAny/DynamicAnyC.cpp + * tao/IFR_Client/IFR_BaseC.cpp + * tao/IFR_Client/IFR_BaseC.h + * tao/IFR_Client/IFR_BaseC.inl + * tao/IFR_Client/IFR_BasicC.cpp + * tao/IFR_Client/IFR_BasicC.h + * tao/IFR_Client/IFR_BasicC.inl + * tao/IFR_Client/IFR_ComponentsC.cpp + * tao/IFR_Client/IFR_ComponentsC.inl + * tao/IFR_Client/IFR_ExtendedC.cpp + * tao/IFR_Client/IFR_ExtendedC.h + * tao/IFR_Client/IFR_ExtendedC.inl + * tao/IORManipulation/IORC.cpp + * tao/IORTable/IORTableC.cpp + * tao/ImR_Client/ImplRepoC.cpp + * tao/ImR_Client/ImplRepoC.inl + * tao/ImR_Client/ServerObjectC.cpp + * tao/ImR_Client/ServerObjectC.inl + * tao/Messaging/Connection_Timeout_Policy_i.cpp + * tao/Messaging/MessagingC.cpp + * tao/Messaging/MessagingC.inl + * tao/Messaging/Messaging_Policy_i.cpp + * tao/Messaging/PollableC.cpp + * tao/PI/InvalidSlotC.cpp + * tao/PI/ORBInitInfoC.cpp + * tao/PI/PIForwardRequestC.cpp + * tao/PortableServer/ForwardRequestC.cpp + * tao/PortableServer/POAManagerC.cpp + * tao/PortableServer/PS_CurrentC.cpp + * tao/PortableServer/PortableServerC.cpp + * tao/RTCORBA/RTCORBAC.cpp + * tao/RTCORBA/RT_Policy_i.cpp + * tao/RTScheduling/RTSchedulerC.cpp + * tao/TypeCodeFactory/Recursive_TypeCode.inl + * tao/Valuetype/StringValueC.inl + Removed ACE_NESTED_CLASS + + * TAO_IDL/ast/ast_type.cpp + * TAO_IDL/be/be_interface.cpp + * TAO_IDL/be/be_visitor_component/component_ci.cpp + * TAO_IDL/be/be_visitor_component/component_cs.cpp + * TAO_IDL/be/be_visitor_exception/exception_cs.cpp + * TAO_IDL/be/be_visitor_interface/amh_ss.cpp + * TAO_IDL/be/be_visitor_interface/interface_ci.cpp + * TAO_IDL/be/be_visitor_interface/interface_cs.cpp + * TAO_IDL/be/be_visitor_interface/interface_is.cpp + * TAO_IDL/be/be_visitor_operation/ami_cs.cpp + * TAO_IDL/be/be_visitor_operation/operation.cpp + * TAO_IDL/be/be_visitor_valuebox/valuebox_ci.cpp + * TAO_IDL/be/be_visitor_valuetype/marshal_cs.cpp + * TAO_IDL/be/be_visitor_valuetype/valuetype_cs.cpp + * TAO_IDL/be/be_visitor_valuetype/valuetype_obv_cs.cpp + * TAO_IDL/be/be_visitor_valuetype/valuetype_ss.cpp + Removed generation of ACE_NESTED_CLASS, just use A::B instead + + All changes below is the merge of the branch sequenceupdate to + sequpdate2. This is the new sequence implementation for TAO. + + * tao/Array_VarOut_T.h: + Only do an empty forward declaration of Array_Traits. This will + make sure that we have to do each specialization explicitly, if + we lack one, we get a compile error instead of this empty default + one. + + * tao/Objref_VarOut_T.h: + Only do an empty forward declaration of Objref_Traits. This will + make sure that we have to do each specialization explicitly, if + we lack one, we get a compile error instead of this empty default + one. + + * tao/Basic_Types.h: + Updated the string types, these are implemented by a template now. + + * tao/BooleanSeqC.{h,cpp}: + * tao/CharSeqC.{h,cpp}: + * tao/CONV_FRAMEC.{h,cpp}: + * tao/CurrentC.{h,cpp}: + * tao/DomainC.{h,cpp}: + * tao/DoubleSeqC.{h,cpp}: + * tao/FloatSeqC.{h,cpp}: + * tao/WStringSeqC.{h,cpp}: + * tao/LongSeqC.{h,cpp}: + * tao/WCharSeqC.{h,cpp}: + * tao/Object_KeyC.{h,cpp}: + * tao/ObjectIdListC.{h,cpp}: + * tao/IIOP_EndpointsC.{h,cpp}: + * tao/LongLongSeqC.{h,cpp}: + * tao/IIOPC.{h,cpp}: + * tao/IOP_IORC.{h,cpp}: + * tao/LongDoubleSeqC.{h,cpp}: + * tao/Messaging_PolicyValueC.{h,cpp}: + * tao/OctetSeqC.{h,cpp}: + * tao/Policy_ForwardC.{h,cpp}: + * tao/PolicyC.{h,cpp}: + * tao/ServicesC.{h,cpp}: + * tao/ShortSeqC.{h,cpp}: + * tao/StringSeqC.{h,cpp}: + * tao/TAOC.{h,cpp}: + * tao/ULongLongSeqC.{h,cpp}: + * tao/ULongSeqC.{h,cpp}: + * tao/UShortSeqC.{h,cpp}: + * tao/AnyTypeCode/AnySeqC.{h,cpp}: + * tao/AnyTypeCode/DynamicC.{h,cpp}: + * tao/DynamicAny/DynamicAnyC.{h,cpp}: + * tao/IFR_Client/IFR_ExtendedC.{h,cpp}: + * tao/IFR_Client/IFR_ComponentsC.{h,cpp}: + * tao/IFR_Client/IFR_BasicC.{h,cpp}: + * tao/IFR_Client/IFR_BaseC.{h,cpp}: + * tao/ImR_Client/ImplRepoC.{h,cpp}: + * tao/IORManipulation/IORC.{h,cpp}: + * tao/ObjRefTemplate/ObjectReferenceTemplateC.{h,cpp}: + * tao/PortableServer/PortableServerC.{h,cpp}: + * tao/RTCORBA/RTCORBAC.{h,cpp}: + * tao/RTScheduling/RTSchedulerC.{h,cpp}: + * tao/Strategies/sciop_endpointsC.{h,cpp}: + * tao/Strategies/uiop_endpointsC.{h,cpp}: + * orbsvcs/orbsvcs/HTIOP/htiop_endpointsC.{h,cpp}: + Updated all these files because of the changes to the sequence + implementation. The base classes are changed including the way we + marshal and demarshal sequences. The argument to the marshal method + is also const. + + * tao/Bounded_Array_Allocation_Traits.h: + * tao/Bounded_Reference_Allocation_Traits_T.h: + * tao/Bounded_Value_Allocation_Traits_T.h: + New allocation traits for bounded sequences + + * tao/Value_Traits_T.h: + New value traits. + + * tao/Bounded_Array_Sequence_T.h: + New template for Bounded Array Sequences + + * tao/Bounded_Basic_String_Sequence_T.h: + * tao/Bounded_String_Sequence_T.h: + * tao/Bounded_Wstring_Sequence_T.h: + New template for bounded strings, derived are string and wstring + bounded sequences + + * tao/Bounded_Object_Reference_Sequence_T.h: + New template for bounded object reference sequences + + New allocation traits for bounded reference + + * tao/Bounded_Sequence_CDR_T.h: + Template method for sequence marshal/demarshal + + * tao/corba.h: + Updated includes, Managed_Types.h is replaced with + String_Manager_T.h + + * tao/CORBA_String.{h,cpp,inl}: + The CORBA::String_var/_out and CORBA::WString_var/_out are now + implemented with the new TAO::String_var/_out template + + * tao/Generic_Sequence_T.h: + New generic sequence template + + * tao/Managed_Types.{h,cpp,i}: + Removed these files + + * tao/MProfile.cpp: + Initialise pointers with 0 and fixed retrieval of a policy + + * tao/Object.{h,cpp}: + Made the argument of the marshal method const + + * tao/operation_details.i: + Changed the way we reset the service info + + * tao/ORB.h: + Updated all typedefs in this file + + * tao/Policy_Set.{h,cpp.i}: + Made the get_policy_by_index const and fixed the + set_policy_overrides to work with the new sequences, as a result + the workarounds could be removed + + * tao/Sequence_T.{cpp,i}: + Removed these files + + * tao/Sequence_T.h: + Include all new sequence template files, makes it easy for old apps + to keep compiling + + * tao/String_Alloc.{h,cpp}: + All string allocation methods + + * tao/VarOut_T.h: + Removed THIS_OUT_TYPE typedef + + * tao/DynamicInterface/Request.h: + Removed include of Sequence.h, not needed + + * tao/Profile.h: + Updated typedef for TAO_opaque + + * tao/Object_Reference_Sequence_Element_T.h: + * tao/Object_Reference_Traits_Base_T.h: + * tao/Object_Reference_Traits_T.h: + * tao/Range_Checking_T.h: + New files + + * tao/Seq_Out_T.{h,inl}: + Removed TAO_MngSeq_Out_T, not needed anymore + + * tao/Sequence.{h,cpp,i}: + Removed + + * tao/Seq_Var_T.{h,cpp.inl}: + Removed TAO_MngSeq_Var_T, not needed anymore + + * tao/String_Manager_T.h: + TAO string manager as template, new file + + * tao/String_Sequence_Element_T.h: + Element in a string sequence + + * tao/String_Traits_Base_T.h: + * tao/String_Traits_T.h: + String traits + + * tao/Unbounded_Array_Allocation_Traits_T.h + * tao/Unbounded_Array_Sequence_T.h + * tao/Unbounded_Basic_String_Sequence_T.h + * tao/Unbounded_Object_Reference_Sequence_T.h + * tao/Unbounded_Octet_Sequence_T.h + * tao/Unbounded_Reference_Allocation_Traits_T.h + * tao/Unbounded_Sequence_CDR_T.h + * tao/Unbounded_String_Sequence_T.h + * tao/Unbounded_Value_Allocation_Traits_T.h + * tao/Unbounded_Value_Sequence_T.h + * tao/Unbounded_Wstring_Sequence_T.h + Unbounded sequence files + + * tao/diffs/Object_Key.diff: + Updated + + * tao/PI/ClientRequestInfo.cpp: + * tao/PI_Server/ServerRequestInfo.cpp: + Removed temporary object usage + + * tao/RTCORBA/RT_Stub.cpp: + * tao/RTScheduling/Request_Interceptor.cpp: + * tao/TypeCodeFactory/TypeCodeFactory_i.cpp: + Updated for the fact that an object sequence now returns a _ptr + on the subscript operators instead of the _var which wasn't + confirming to the CORBA C++ mapping + + * tao/RTScheduling/Current.h: + Updated IdType typedef + + * tao/Valuetype/Bounded_Valuetype_Allocation_Traits_T.h + * tao/Valuetype/Bounded_Valuetype_Sequence_T.h + * tao/Valuetype/Unbounded_Valuetype_Allocation_Traits_T.h + * tao/Valuetype/Unbounded_Valuetype_Sequence_T.h + * tao/Valuetype/Valuetype_Sequence_Element_T.h + * tao/Valuetype/Valuetype_Traits_Base_T.h + * tao/Valuetype/Valuetype_Traits_T.h + New sequence implementated for valuetypes + + * tao/Valuetype/Sequence_T.{cpp,inl}: + Removed + + * tao/Valuetype/Sequence_T.h: + Just include the new files, easier for backward compatibility + + * tao/Valuetype/Value_VarOut_T.{h,cpp}: + Just define an empty Value_Traits, make sure we get all + specializations + + * TAO_IDL/be/be_codegen.cpp: + * TAO_IDL/be/be_sequence.cpp: + * TAO_IDL/be/be_visitor_traits.cpp: + * TAO_IDL/be/be_visitor_array/array_ch.cpp: + * TAO_IDL/be/be_visitor_array/serializer_op_cs.cpp: + * TAO_IDL/be/be_visitor_array/cdr_op_cs.cpp: + * TAO_IDL/be/be_visitor_array/array_cs.cpp: + * TAO_IDL/be/be_visitor_array/array_ci.cpp: + * TAO_IDL/be/be_visitor_sequence/sequence_ch.cpp: + * TAO_IDL/be/be_visitor_sequence/serializer_op_cs.cpp: + * TAO_IDL/be/be_visitor_typedef/typedef_ch.cpp: + * TAO_IDL/be/be_visitor_typedef/typedef_ci.cpp: + * TAO_IDL/be/be_visitor_valuebox/valuebox_cs.cpp: + * TAO_IDL/be/be_visitor_valuetype/valuetype_cs.cpp: + * TAO_IDL/be_include/be_visitor_traits.h: + Updated for new sequence implementation + + * examples/CSD_Strategy/ThreadPool4/ClientTask.cpp: + * examples/CSD_Strategy/ThreadPool5/ClientTask.cpp: + * examples/Load_Balancing/Identity_Client.cpp: + * examples/Load_Balancing_persistent/Identity_Client.cpp: + * examples/POA/NewPOA/NewPOA.cpp: + * examples/POA/POA_BiDir/POA_BiDir.cpp: + * tests/CSD_Strategy_Tests/TP_Foo_B/Foo_B_ClientEngine.cpp: + * tests/DynAny_Test/test_dynsequence.cpp: + * tests/Sequence_Unit_Tests/*: + * tests/ORT/ServerRequestInterceptor.cpp: + * tests/Param_Test/big_union.cpp: + * + tests/Portable_Interceptors/ForwardRequest/Client_ORBInitializer.cpp: + * orbsvcs/IFR_Service/ifr_adding_visitor.cpp: + * orbsvcs/orbsvcs/DsLogAdmin.idl: + * orbsvcs/orbsvcs/AV/AVStreams_i.cpp: + * orbsvcs/orbsvcs/CosEvent/CEC_TypedEventChannel.{cpp,i}: + * orbsvcs/orbsvcs/ETCL/ETCL_Constraint.{h,cpp,i} + * orbsvcs/orbsvcs/IFRService/ComponentContainer_i.cpp + * orbsvcs/orbsvcs/IFRService/ComponentDef_i.cpp + * orbsvcs/orbsvcs/IFRService/Container_i.cpp + * orbsvcs/orbsvcs/IFRService/EnumDef_i.cpp + * orbsvcs/orbsvcs/IFRService/ExtValueDef_i.cpp + * orbsvcs/orbsvcs/IFRService/HomeDef_i.cpp + * orbsvcs/orbsvcs/IFRService/IFR_Service_Utils.cpp + * orbsvcs/orbsvcs/IFRService/InterfaceDef_i.cpp + * orbsvcs/orbsvcs/IFRService/OperationDef_i.cpp + * orbsvcs/orbsvcs/IFRService/ValueDef_i.cpp + * orbsvcs/orbsvcs/LoadBalancing/LB_LoadManager.cpp: + * orbsvcs/orbsvcs/LoadBalancing/LB_ObjectReferenceFactory.cpp: + * orbsvcs/orbsvcs/Log/Log_i.h: + * orbsvcs/orbsvcs/Property/CosPropertyService_i.h: + * orbsvcs/orbsvcs/Trader/Constraint_Nodes.{h,cpp} + * orbsvcs/tests/AVStreams/Component_Switching/distributer.cpp: + * orbsvcs/tests/AVStreams/Component_Switching/receiver.cpp: + * orbsvcs/tests/AVStreams/Component_Switching/sender.cpp: + * orbsvcs/tests/Bug_1393_Regression/client.cpp: + * orbsvcs/tests/FT_App/FT_Client.cpp: + * orbsvcs/tests/InterfaceRepo/IDL3_Test/idl3_client.cpp: + * orbsvcs/tests/InterfaceRepo/IFR_Test/Admin_Client.cpp: + * + orbsvcs/orbsvcs/FtRtEvent/EventChannel/AMI_Primary_Replication_Strategy.cpp: + * orbsvcs/orbsvcs/FtRtEvent/EventChannel/IOGR_Maker.cpp: + * tests/RTScheduling/Current/Thread_Task.cpp: + * tests/RTScheduling/Thread_Cancel/Thread_Task.cpp: + Updated for sequence implementation + + * tests/Sequence_Unit_Tests/*: + Removed files that are now in the core TAO lib + +Wed Apr 19 07:48:12 UTC 2006 Johnny Willemsen <jwillemsen@remedy.nl> + + * tests/Bug_2503_Regression/*: + New regression for bug 2503. Thanks to Carlos O'Ryan for creating + this test + +Tue Apr 18 20:51:48 2006 Wallace Zhang <zhangw@ociweb.com> + + * TAO version 1.5.1 released. + +Local Variables: +mode: change-log +add-log-time-format: (lambda () (progn (setq tz (getenv "TZ")) (set-time-zone-rule "UTC") (setq time (format-time-string "%a %b %e %H:%M:%S %Z %Y" (current-time))) (set-time-zone-rule tz) time)) +indent-tabs-mode: nil +End: |