diff options
Diffstat (limited to 'ChangeLogs/ChangeLog-02a')
-rw-r--r-- | ChangeLogs/ChangeLog-02a | 478 |
1 files changed, 45 insertions, 433 deletions
diff --git a/ChangeLogs/ChangeLog-02a b/ChangeLogs/ChangeLog-02a index 5a7cf713c8a..afb206866f0 100644 --- a/ChangeLogs/ChangeLog-02a +++ b/ChangeLogs/ChangeLog-02a @@ -1,441 +1,53 @@ -Mon Apr 23 17:18:35 2001 Steve Huston <shuston@riverace.com> - - * ace/ACE.cpp (ACE::handle_timed_complete): Set need_to_check for - any non-success result, or on a system where you can't tell (AIX). - Also, to check, use getsockopt to retrieve the status/error if - SOL_SOCKET and SO_ERROR are defined; else use the old recv method. - This gets you a real error code on a failed connect, if the platform - supports it. - - * tests/SOCK_Connector_Test.cpp: For an expected success, allow test - to pass if socket gets a reset but not if it says ENOTCONN. If there - are platforms that can't do SOL_SOCKET/SO_ERROR (above) then this - other code may need to be re-enabled. - -Mon Apr 23 13:36:27 2001 Steve Huston <shuston@riverace.com> - - * ace/ACE.cpp (ACE::handle_timed_complete): Winsock (ACE_WIN32) sets - the exception bit (not the read bit) for a failed async connect. - So on fail, don't bother to do the recv trick to see if the socket - really did fail. - - * ace/Connector.cpp (ACE_Connector::handle_output): - * ace/SOCK_Connector.cpp (complete): Apparantly, Win2000 has - fixed the problem where a connection succeeds but can't be used - for a time. So, the sleep workaround is restructured for ACE_WIN32 - to try the new socket first, and only do the sleep if it fails. - Don't try to do this check (e.g. building on Win2000) at compile - time, since we want to be able to build on NT and run on Win2000. +Sat Apr 14 10:51:39 2001 Carlos O'Ryan <coryan@uci.edu> - * ace/README: - * ace/config-win32-common.h: Removed the macro, - ACE_HAS_BROKEN_NONBLOCKING_CONNECTS. It was only used in the two - cases above, and is not needed any longer. - -Mon Apr 23 07:08:09 2001 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * ace/Based_Pointer_T.h: Updated the documentation to - explain how ACE_Based_Pointer<T> (NULL) behaves. Thanks to Tom - Moog <tmoog@polhode.com> for motivating this. - -Sun Apr 22 14:26:56 2001 Carlos O'Ryan <coryan@uci.edu> - - * ace/OS.h: - Some warnings are automatically disabled, but the pragmas only - make sense for MSVC. - - * ace/streams.h: - Fixed problems for Sun/CC-4.2, the trick is to include - strstream.h on MINGW32 *and* unix-like platforms, while - strstrea.h is included for all other WinXX-based compilers. - -Sat Apr 21 19:02:27 2001 Nanbor Wang <nanbor@cs.wustl.edu> - - * netsvcs/clients/Naming/Dump_Restore/Makefile: - * netsvcs/clients/Naming/Client/Makefile: Removed duplicated ACE - library. I thought I removed them all. - -Sat Apr 21 15:31:27 2001 Carlos O'Ryan <coryan@uci.edu> - - * ace/ACE.cpp: - * ace/Basic_Types.h: - * ace/Configuration.cpp: - * ace/Handle_Gobbler.i: - * ace/Lib_Find.cpp: - * ace/Local_Name_Space_T.cpp: - * ace/Log_Msg.cpp: - * ace/Makefile: - * ace/Mem_Map.cpp: - * ace/Memory_Pool.cpp: - * ace/Message_Queue.i: - * ace/NT_Service.cpp: - * ace/OS.cpp: - * ace/OS.h: - * ace/OS.i: - * ace/OS_String.cpp: - * ace/Process.cpp: - * ace/Process_Manager.cpp: - * ace/Registry.cpp: - * ace/SOCK_Connector.i: - * ace/Sock_Connect.cpp: - * ace/TTY_IO.cpp: - * ace/WFMO_Reactor.cpp: - * ace/WFMO_Reactor.i: - * ace/WIN32_Asynch_IO.cpp: - * ace/config-cygwin32-common.h: - * ace/config-g++-common.h: - * ace/config-win32-common.h: - * ace/config-win32-mingw.h: - * ace/config-win32.h: - * ace/streams.h: - * examples/NT_Service/Makefile: - * examples/NT_Service/main.cpp: - * examples/NT_Service/ntsvc.cpp: - * include/makeinclude/platform_cygwin32.GNU: - * include/makeinclude/platform_gnuwin32_common.GNU: - * include/makeinclude/platform_mingw32.GNU: - * include/makeinclude/rules.local.GNU: - * include/makeinclude/wrapper_macros.GNU: - * tests/CDR_File_Test.cpp: - * tests/Cached_Accept_Conn_Test.cpp: - * tests/Cached_Conn_Test.cpp: - * tests/Conn_Test.cpp: - * tests/Handle_Set_Test.cpp: - * tests/MEM_Stream_Test.cpp: - * tests/MEM_Stream_Test.h: - * tests/MM_Shared_Memory_Test.cpp: - * tests/MT_SOCK_Test.cpp: - * tests/Malloc_Test.cpp: - * tests/Mem_Map_Test.cpp: - * tests/OS_Test.cpp: - * tests/Priority_Reactor_Test.cpp: - * tests/Process_Strategy_Test.cpp: - * tests/Reactor_Exceptions_Test.cpp: - * tests/Reactor_Notify_Test.cpp: - * tests/SOCK_Send_Recv_Test.cpp: - * tests/SOCK_Test.cpp: - * tests/SPIPE_Test.cpp: - * tests/Svc_Handler_Test.cpp: - * tests/run_tests.bat: - * tests/run_tests.sh: - * tests/RMCast/RMCast_Fragment_Test.cpp: - Add support for mingw (www.mingw.org) and partial support for - cygwin (http://cygwin.com/). Many thanks to Cristian Ferretti - <cristian_ferretti@yahoo.com> - -Fri Apr 20 18:21:41 2001 Carlos O'Ryan <coryan@uci.edu> - - * ace/CDR_Stream.i (align_write_ptr): - Don't use grow_and_adjust() here because it forces an - allocation, using simply adjust() is better. - -Fri Apr 20 19:37:53 2001 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/Get_Opt.h: Updated the documentation to match the constructor - arguments for ACE_Get_Opt. Thanks to Benot Desmeules - <bdesmeules@eurekium.com> for reporting this. - - * ace/TP_Reactor.cpp: Added a workaround for compiling ACE on - Linux for the OS/390 OS. Thanks to James Dorsey - <James.Dorsey@acxiom.com> for reporting this. - -Fri Apr 20 00:57:24 2001 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * ace/Containers_T.cpp (operator=): Added a missing assignment to - this->size_ = s.size_. Thanks to Juan Jose Comellas - <jcomellas@novamens.com> for reporting this. - - * ace/SOCK_Dgram: One of the send() methods was lacking a "const" - in front of the ACE_Addr &. This has been fixed. Thanks to - Victor Poznyak <VictorP@Webley.COM> for reporting this. - -Thu Apr 19 16:18:33 2001 Steve Huston <shuston@riverace.com> - - * ace/Log_Msg.h: Doxygen-ized some function arguments; other - minor doc cleanups. - -Thu Apr 19 11:08:15 2001 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * ACE-INSTALL.html: Point out that ACE supports Solaris 8. - Thanks to Logan Modahala <lmodahal@cisco.com> for motivating - this. - -Wed Apr 18 13:57:40 2001 Chad Elliott <elliott_c@ociweb.com> - - * ace/ACE.cpp: - - Modification to check errno for ENOBUFS (in addition to - EWOULDBLOCK) after writing to a file descriptor. This will happen - on VxWorks when a client sends data faster than the server can - receive it. This modification causes this to be a retry situation - instead of an error. [Bug 866] - -Wed Apr 18 11:33:07 2001 Balachandran Natarajan <bala@cs.wustl.edu> - - * bin/auto_run_tests.lst: Added the BiDirectional tests to the - nightly builds. There is no reason why it did not go in so far. - -Wed Apr 18 10:25:03 2001 Balachandran Natarajan <bala@cs.wustl.edu> - - * bin/auto_run_tests.lst: Changed the name of the -config option - for SUNCC5.1. Removed the dot and replaced it with a '_'. - -Wed Apr 18 05:23:53 2001 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * tests/Config_Test.cpp: Fixed several warnings related to - not using ACE_TEXT for wide-strings. Thanks to Bala for - reporting this. - -Wed Apr 18 02:08:16 2001 Nanbor Wang <nanbor@cs.wustl.edu> - - * examples/Synch/proc_sema.dsp: Fixed release build settings. - -Tue Apr 17 12:33:49 2001 Steve Huston <shuston@riverace.com> - - * ace/Service_Config.cpp (open_i): If the caller supplies a - logger key, enable the ACE_Log_Msg::LOGGER bit when opening - the ACE_Log_Msg instance, else the logger key is ignored. - -Tue Apr 17 06:15:00 2001 Michael Kircher <Michael.Kircher@mchp.siemens.de> - - * bin/auto_run_tests.lst: - - Added the UDP Pluggable Protocol test to the list of executed tests. - -Mon Apr 16 20:42:46 2001 Carlos O'Ryan <coryan@uci.edu> - - * include/makeinclude/platform_qnx_rtp_gcc.GNU: - Remove obsolete comments and raise the optimization level from - -O2 to -O3, the latter disables a huge number of warnings for - PACE. - -Mon Apr 16 22:15:28 2001 Balachandran Natarajan <bala@cs.wustl.edu> - - * bin/auto_run_tests.lst: Commented out some tests that are - failing on SUNCC 5.1. The failure is because of a compiler - bug. The problem is that the compiler seems to have difficulty - in calling the virtual method in a class hierarchy which is - built with multiple inheritance. - -Mon Apr 16 20:24:11 2001 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * tests/Config_Test.cpp (main): Remoted the test_io() function. - Thanks to Jerry Odenwelder <jerry.o@mindspring.com> for - confirming this. - -Mon Apr 16 16:18:57 2001 Carlos O'Ryan <coryan@uci.edu> - - * ace/config-qnx-rtp.h: - Add support for gperf under QNX - - * include/makeinclude/platform_qnx_rtp_gcc.GNU: - The -fexceptions flag has no place in the CPPFLAGS macros, it - should go in CCFLAGS (though I could be convinced to put it in - CFLAGS) - -Mon Apr 16 16:13:29 2001 Steve Huston <shuston@riverace.com> - - * ace/Configuration.h (ACE_Configuration::root_section): Made this - method 'const'; makes today's changes easier to deal with. - - * ace/Configuration_Import_Export.cpp (string_has_white_space): - * ace/Configuration.cpp (operator==): Replaced 'true' with 1 and - 'false' with 0; replace const_cast<> with ACE_const_cast (), - per docs/ACE-guidelines.html, and IBM C++ :-) Also, use two - non-const local variables instead of const_casting 'this' and - 'rhs' on every use. - Made ACE_Configuration::root_section (void) const. - - * ace/ace.icc: This was an early hack attempt to use library or DLL - build for Visual Age C++ 4. It is now the Visual Age C++ 5 - configuration for building ACE on AIX. The ace-dll.icc and - ace-lib.icc are still around for Visual Age C++ 4 users, but may - not work and will be removed in a future ACE release. This - config is for Visual Age C++ 5 only and can/will be extended to - do subset builds as well as debug/optimize/etc variants. - - * bin/make_pretty.pl: Pick out AIX ld errors correctly. - - * include/makeinclude/platform_hpux_aCC.GNU: Note HP problem report - number and disposition for inability to use -AA and +noeh together. - -Mon Apr 16 12:32:48 2001 Nanbor Wang <nanbor@cs.wustl.edu> - - * tests/MEM_Stream_Test.cpp: Limited the concurrent connections - used in the multithreaded test on platforms without POSIX or - Win32 semaphores. In this case, SysV semaphores are used and - a user can only instantiate very limited number of them. The - test will also print out an INFO in this case as it implies - there can only be one "multithreaded" connection in one host at - any instance. This fixed the test on some obsolete UNIX - platforms. - -Mon Apr 16 09:58:23 2001 Balachandran Natarajan <bala@cs.wustl.edu> - - * tests/Config_Test.cpp: Fixed a warning with g++ builds. The - warning was because of an unused function. - -Sun Apr 15 21:27:19 2001 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/ace_dll.dsp: - * ace/ace_lib.dsp: Added new Configuration_Import_Export.{h,cpp} - files. - -Sun Apr 15 19:33:19 2001 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * ace/Configuration.cpp, - * ace/Configuration_Import_Export.cpp: Fixed some warnings - that show up with G++. Thanks to Carlos for reporting this. - -Sun Apr 15 17:13:40 2001 Carlos O'Ryan <coryan@uci.edu> - - * bin/auto_run_tests.lst: - Disabled TAO/tests/Client_Leaks on all configurations, it is - failing systematically because it cannot create processes fast - enough. - -Sat Apr 14 18:08:16 2001 Carlos O'Ryan <coryan@uci.edu> - - * ace/Configuration_Import_Export.h: - Fixed small typo in the @file comment for - Configuration_Import_Export.h - -Sat Apr 14 18:31:15 2001 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * ace/Makefile.bor: Added new Configuration_Import_Export.obj file. - - * tests/Config_Test.cpp: Added tests for the new ACE_Configuration* - stuff. Thanks to Jerry Odenwelder <jerry.o@mindspring.com> for - contributing these. - - * ace/Configuration_Import_Export.{h,cpp}: Added two new files that - do the following: - - 1. Remove the import and export functionality from the - ACE_Configuration object. - 2. Provide two new classes to import/export ACE_Configuration - objects: - - * ACE_Registry_ImpExp import and exports in the Win32 - Registry format - this was the old import/export routine on - ACE_Configuration. - * ACE_Ini_ImpExp imports and exports in an INI format. The - import part of this was called import_config_as_strings in - the old ACE_Configuration object. While this object will - export any ACE_Configuration object, it can only import data as - strings. This is due to the inability to distinguish type - in the INI format. - 3. Added operator== to ACE_Configuration. This is to compare - contents not type. - 4. Test programs were modified to exercise this new functionality. - - Thanks to Jerry Odenwelder <jerry.o@mindspring.com> for - contributing these. - -Sat Apr 14 12:44:17 2001 Carlos O'Ryan <coryan@uci.edu> - - * ace/MEM_IO.i: - Reorder operations to eliminate warnings on Tru64/CXX - -Sat Apr 14 11:12:26 2001 Steve Huston <shuston@riverace.com> - - * tests/CLASSIX_Addr_Test.icc: - * tests/CLASSIX_CLD_Connector_Test.icc: - * tests/CLASSIX_Con_Acc_Test.icc: - * tests/CLASSIX_Group_Test.icc: - * tests/CLASSIX_Notify_Test.icc: - * tests/CLASSIX_OS_Test.icc: - * tests/CLASSIX_Reactor_Test.icc: - * tests/CLASSIX_Select_Reactor_Test.icc: - * tests/CLASSIX_Stream_Test.icc: - Removed these files. They were moved/copied to tests/CLASSIX some - time back and don't belong here. - - * tests/t.icc: Removed. This is not used for anything. - - * tests/Makefile.vac: Corrected 'realclean' actions. - -Sat Apr 14 09:17:32 2001 Steve Huston <shuston@riverace.com> - - * examples/QOS/Change_Sender_TSpec/sender.dsp: Added extra level of - directory for include and lib - missed this one in yesterday's - set of changes. - -Sat Apr 14 09:08:33 2001 Steve Huston <shuston@riverace.com> - - * ace/Array_Base.h: Added forward declaration for ACE_Allocator. - * ace/Array_Base.cpp: Added #include "Malloc_Base.h" so IBM C++ - can find ACE_Allocator when instantiating templates. - - * ace/config-hpux-10.x-hpc++.h: When using standard C++ lib, - set the value of ACE_USES_STD_NAMESPACE_FOR_STDCPP_LIB to 1. - - * netsvcs/clients/Naming/Dump_Restore/Makefile: - * netsvcs/clients/Naming/Client/Makefile: Add ACELIB to LIBS. This - allows IBM C++ compilers to build the library clean. - - * tests/Makefile.vac: Don't try to use vacbld_r on the entire - project. It can take so long that make gives up and kills - off the build, making a real mess of the .ics files and leaving - random numbers of tests unbuilt. Instead, run vacbld_r on - each .icc file in the directory. Similarly for clean, don't - reply on vacbld_r on the project file - run through all the - ics files and executables based on existing .icc files. + * ace/Log_Msg.h: + Document the lock related functions in a separate Doxygen + section. -Fri Apr 13 12:44:21 2001 Steve Huston <shuston@riverace.com> +Fri Apr 13 11:58:58 2001 Carlos O'Ryan <coryan@uci.edu> - * netsvcs/lib/Server_Logging_Handler_T.cpp (handle_logging_record): - If logging to STDERR is not enabled, don't dump the log record - there either. + * ace/OS.i (gethrtime): + Add support for the the clock tick register as a high resolution + timer under KCC/Linux. Fri Apr 13 12:21:01 2001 Nanbor Wang <nanbor@cs.wustl.edu> - * ace/Malloc_T.h: Added an explanation on the requirement about - sizeof (TYPE)link of ACE_Cached_Allocator and a link to - ACE_Cached_Mem_Pool_Node where this stuff was explained. Thanks - to Rachel G. Smith <Rachel_G_Smith@raytheon.com> for motivating - the change. + * ace/Malloc_T.h: Added an explanation on the requirement about + sizeof (TYPE)link of ACE_Cached_Allocator and a link to + ACE_Cached_Mem_Pool_Node where this stuff was explained. Thanks + to Rachel G. Smith <Rachel_G_Smith@raytheon.com> for motivating + the change. Fri Apr 13 12:05:38 2001 Nanbor Wang <nanbor@cs.wustl.edu> - * tests/MEM_Stream_Test.h: - * tests/MEM_Stream_Test.cpp: Borland C++ has problem cleaning up - the static Atomic_Op object in Echo_Handler in the right order. - Made that an ACE_Singleton and re-organized the test a big. - Thanks to Chris Kohlhoff <chris@kohlhoff.com> for reporting and - suggesting the fix. [Bug 862] - -Fri Apr 13 11:58:58 2001 Carlos O'Ryan <coryan@uci.edu> - - * ace/OS.i (gethrtime): - Add support for the the clock tick register as a high resolution - timer under KCC/Linux. Thanks to David Levine for help on this. + * tests/MEM_Stream_Test.h: + * tests/MEM_Stream_Test.cpp: Borland C++ has problem cleaning up + the static Atomic_Op object in Echo_Handler in the right order. + Made that an ACE_Singleton and re-organized the test a big. + Thanks to Chris Kohlhoff <chris@kohlhoff.com> for reporting and + suggesting the fix. [Bug 862] Fri Apr 13 05:30:04 2001 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - * ace/OS.i: Make sure that the PTHREADS version of ACE_OS::thr_yield() - gets top priority over the STHREADS version. Thanks to Munagala - Ramanath <amberarrow@yahoo.com> for reporting this. - - * ace/Malloc_T.{h,i}: Added support for one of the calloc() methods - to ACE_Cached_Allocator. Thanks to Rachel G Smith - <Rachel_G_Smith@raytheon.com> for motivating this change. + * ace/Malloc_T.{h,i}: Added support for one of the calloc() methods + to ACE_Cached_Allocator. Thanks to Rachel G Smith + <Rachel_G_Smith@raytheon.com> for motivating this change. - * ace/Malloc_Allocator.h: Updated the comments for ACE_New_Allocator - to indicate that the calloc() methods are also supported. - Thanks to Rachel G Smith <Rachel_G_Smith@raytheon.com> for - motivating this change. + * ace/Malloc_Allocator.h: Updated the comments for ACE_New_Allocator + to indicate that the calloc() methods are also supported. + Thanks to Rachel G Smith <Rachel_G_Smith@raytheon.com> for + motivating this change. - * docs/ACE-bug-process.html: Updated this document to explain how - bugs pertaining to OCI's TAO releases should be handled. Thanks - to Georg Lohrer <GeorgLohrer@web.de> for motivating this. + * docs/ACE-bug-process.html: Updated this document to explain how + bugs pertaining to OCI's TAO releases should be handled. Thanks + to Georg Lohrer <GeorgLohrer@web.de> for motivating this. Thu Apr 12 20:36:30 2001 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - * ace/OS.i (thr_getspecific): Reorder the way in which we check - the flags so that ACE_HAS_PTHREADS takes precedence over - ACE_HAS_STHREADS. Thanks to Munagala Ramanath - <amberarrow@yahoo.com> for reporting this. + * ace/OS.i (thr_getspecific): Reorder the way in which we check + the flags so that ACE_HAS_PTHREADS takes precedence over + ACE_HAS_STHREADS. Thanks to Munagala Ramanath + <amberarrow@yahoo.com> for reporting this. Thu Apr 12 21:18:44 2001 Steve Huston <shuston@riverace.com> @@ -471,24 +83,24 @@ Thu Apr 12 15:48:22 2001 Joe Hoffert <joeh@cs.wustl.edu> Thu Apr 12 13:06:27 2001 Joel Sherrill <joel@OARcorp.com> - * include/makeinclude/Makefile.rtems: Replaced debug version - with real version that queries RTEMS application Makefile - infrastructure. + * include/makeinclude/Makefile.rtems: Replaced debug version + with real version that queries RTEMS application Makefile + infrastructure. * include/makeinclude/platform_rtems.x_g++.GNU: Added --no-print-directory option to all invocations of recursive - make to determine information from Makefile.rtems. This - avoids getting the "entering/exiting directory" messages - included in the results. + make to determine information from Makefile.rtems. This + avoids getting the "entering/exiting directory" messages + included in the results. * ACE-RTEMS.README: Added requirement to configure C++ support in RTEMS. Updated to reflect also compiling for mc68360 - target. + target. Wed Apr 11 20:36:36 2001 Nanbor Wang <nanbor@cs.wustl.edu> - * examples/Synch/proc_sema.cpp (main): Fixd an assigning string - literal to char* warning. + * examples/Synch/proc_sema.cpp (main): Fixd an assigning string + literal to char* warning. Wed Apr 11 18:04:33 2001 Steve Huston <shuston@riverace.com> @@ -498,9 +110,9 @@ Wed Apr 11 18:04:33 2001 Steve Huston <shuston@riverace.com> Wed Apr 11 11:00:50 2001 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - * ACE-INSTALL.html: Added a section illustrating how to build - ACE+TAO for RTEMS. Thanks to Joel Sherrill <joel@OARcorp.com> - for this build procedure. + * ACE-INSTALL.html: Added a section illustrating how to build + ACE+TAO for RTEMS. Thanks to Joel Sherrill <joel@OARcorp.com> + for this build procedure. Wed Apr 11 11:38:37 2001 Steve Huston <shuston@riverace.com> |