diff options
| -rw-r--r-- | ChangeLog | 8243 | ||||
| -rw-r--r-- | ChangeLogs/ChangeLog-03a | 8243 |
2 files changed, 8 insertions, 16478 deletions
diff --git a/ChangeLog b/ChangeLog index 415de5b54d7..d0f450a8ab1 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,3 +1,7 @@ +Thu Feb 20 19:18:17 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> + + * ChangeLog: Shortened the changelog for ease of use. + Thu Feb 20 10:35:15 2003 Ossama Othman <ossama@uci.edu> * ace/Strategies_T.h: @@ -316,8243 +320,4 @@ Thu Feb 20 09:23:51 2003 Ossama Othman <ossama@uci.edu> ACE_SSL_SOCK_{Acceptor,Connector} classes. They are unnecessary. -Wed Feb 19 12:50:58 2003 Heather Drury <oci@cs.wustl.edu> - - * ACE version 5.3.1 released. - -Mon Feb 17 15:21:18 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: The Server_Protocol and the ORB_init - test in RTCORBA will not run in ST builds since the operations - ACE_OS::thr_getprio () are not supported. - -Fri Feb 14 09:53:43 2003 Rich Seibel <seibel_r@ociweb.com> - - * ace/config-osf1-4.0.h: Tru64 lacks the clearerr - function when built without threads (thread=0). - Thanks to Abhay Kulkarni <Abhay.Kulkarni@veritas.com> - for reporting the problem. - -Thu Feb 13 17:50:07 2003 Krishnakumar B <kitty@cse.wustl.edu> - - * include/makeinclude/rules.lib.GNU: Build static libraries - using xargs with the list of all the object files instead of - using them directly on the command line. This works around - brokenness of systems that have a small limit on the length of - the command-line arguments. Thanks to Abhay Kulkarni - <Abhay.Kulkarni@veritas.com> for reporting the problem. - -Thu Feb 13 12:43:27 2003 Pradeep Gore <pradeep@oomworks.com> - - * bin/auto_run_tests.lst: - Added TAO/examples/RTCORBA/Activity/run_test.pl to this list. - -Thu Feb 13 07:36:40 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * THANKS: Added Dmitri Belogaj to the hall of fame. - -Wed Feb 12 12:40:21 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: Prevented a few RTCORBA tests to be run - in our daily builds since changing priorities requires root - access. - -Wed Feb 12 12:06:05 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/Strategies.h: Added this file for backward - compatibility. Backward compatibility was broken with the - following checkin - - "Tue Nov 20 12:48:39 2001 Balachandran Natarajan - <bala@cs.wustl.edu>". Thanks to Dmitri Belogaj - <belogaj@nentec.de> for reporting this problem. - -Wed Feb 12 10:18:21 2003 Nanbor Wang <nanbor@cs.wustl.edu> - - * bin/MakeProjectCreator/templates/vc7.mpd: Checked in Chad's - missing ssl_libs fix. - -Wed Feb 12 08:59:12 UTC 2003 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/UPIPE_Addr.h: - Removed ACE_UPIPE_Addr class which isn't used, typedef ACE_UPIPE_Addr - still is there. - -Tue Feb 11 22:19:39 2003 Krishnakumar B <kitty@cse.wustl.edu> - - * examples/DLL/Makefile.Today (SHLIB): - * examples/DLL/Makefile.Newsweek (SHLIB): - - Removed extra space at the end which caused compilation to - break. Thanks to Bill Cassanova <BCassanova@weather.com> for - reporting the problem. - -Wed Feb 12 00:00:12 UTC 2003 Craig Rodrigues <crodrigu@bbn.com> - - * ace/OS.h: - * ace/OS.i (thr_setprio): Add a new default thr_policy parameter which - allows us to explicitly set the policy in pthread_setschedparam(). - Introduced to eliminate a bug discovered in RT-CORBA implementation. - -Tue Feb 11 14:01:21 2003 Nanbor Wang <nanbor@cs.wustl.edu> - - * vc7_nightly.mwc: Added a temporary file specifying core nightly - build targets. Notice this file includes both ACE and TAO. - - * ace/ace.mpc: Added XML_Svc_Conf.cpp. - - * ace/SSL/ssl.mpc: Fixed dllflags setting. - -Mon Feb 10 18:08:12 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * THANKS: Added Otis Nyandoro to the hall of fame. - -Mon Feb 10 18:02:41 2003 Steve Huston <shuston@riverace.com> - - * ace/OS.h: Don't special-case SEH support on ACE_HAS_WINCE. First, - CE can do SEH; second, the switching of this on/off can and should - be done in the config file. - - * ace/config-win32-common.h: Enable ACE_HAS_WIN32_STRUCTURAL_EXCEPTIONS - for WinCE. - - Thanks to Emmanuel Thevenot Beaufort - <emmanuel.thevenot-beaufort@jci.com> for this fix. - - * ace/Handle_Set.{h i}: Remove the ACE_Handle_Set destructor for - ACE_HAS_WINCE. The need for this predates the supported vc 3 and 4 - compilers, and having it in there triggers an SEH warning. - - * ace/SPIPE_Stream.i: Follow-up to: - Mon Feb 10 09:22:04 2003 Phil Mesnier <mesnier_p@ociweb.com> - Added the need for ACE_HAS_WINSOCK2 to enable this code. The calls - and structures used aren't available in Winsock1. - - * ace/config-wince.h: Disable Winsock2 for CE versions less than 4.0. - -Mon Feb 10 12:14:35 2003 Jeffrey Graham <jgraham@titan.com> - - * bin/ace-install: - - Fixed problem where all TAO executables were not installed. - - Fixed installed file permission problem. Files that should have - global access didn't have it. - - Fixed problem where the script did not easily flatten the - install directory, requires the ACE_ARCH subdirectory trees be - built. - - Fixed problem where all binaries were not filtered out. - -Mon Feb 10 19:42:12 UTC 2003 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/ATM_Acceptor.h: - Converted old style documentation to doxygen style - - * ace/Arg_Shifter.h: - Added private declaration of copy constructor and assignment - operator. Instances of these class can not be copied using the - default ones. Also fixes a MSVC6 level 4 warning. - - * ace/Configuration.cpp: - Removed not needed MingW ifdef, the normal code also compiles with - MingW - - * ace/config-win32.h: - Removed cygwin from this file. For cygwin the config-cygwin32.h - must be included directly in config.h - - * ace/Notification_Strategy.h: - * ace/Notification_Strategy.inl: - Made mask() method const - - * include/makeinclude/ace_flags.bor: - If TAO_ROOT isn't defined, define it as $(ACE_ROOT)/TAO - - * bin/MakeProjectCreator/templates/bor.mpd: - * include/makeinclude/build_dll.bor: - * include/makeinclude/build_exe.bor: - Thanks for Christopher Kohlhoff <chris@kohlhoff.com> for providing - a fix for the singleton template with Borland. The problem is that - with a singleton template the code is implicitly generated into - every object file that references it. The linker then strips out - all but a single copy of the code. When linking an exe that refers - to a singleton template in a dll, the linker would choose one - of the template code instances in the exe's .obj files, and - completely ignore the instance that was exported by the dll. - - The key to the solution is that it appears that the borland linker - chooses the first instance of the template symbols that it sees. The - linker will let you list the .lib files in the place normally used - for object files. By listing the library files before any of the - object files, it resolves the template code against the dll exports. - - * examples/Service_Configurator/Misc/Makefile: - * examples/Service_Configurator/Misc/Makefile.Timer: - * examples/Service_Configurator/Misc/Makefile.main: - Splitted the original GNU makefile in multiple files, one for the - shared library and one for the exe and one to trigger the other - two. This fixes the compile errors with this example with Cygwin - and MingW. Thanks for Vince Mounts <vince@mounts.cc> for making - these makefiles. - -Mon Feb 10 09:22:04 2003 Phil Mesnier <mesnier_p@ociweb.com> - - * ace/SPIPE_Stream.i: Added a special case for win32 platforms to - ACE_SPIPE_Stream::send_handle() and recv_handle() to allow handle - passing. The extension is necessary since the receiving side must - open the duplicate handle before the sending side closes it. - -Fri Feb 7 17:22:39 2003 Steve Huston <shuston@riverace.com> - - * ace/OS.cpp (ACE_OS::thr_create): Moved the pthread_setstack() call - (and the check for whether it's available) from inside the - Pthreads draft 4/6 section to the Pthreads standard section - and corrected the condition it's used (defined - ACE_HAS_PTHREAD_SETSTACK, not !defined ACE_HAS_PTHREAD_SETSTACK). - - Also see: - Thu Jan 16 19:06:19 2003 Stephen Torri <storri@cse.wustl.edu> - -Fri Feb 7 16:55:58 2003 Heather Drury <drury_h@ociweb.com> - - * ace/config-aix-4.x.h: Reverted change to aix config file: - - Fri Feb 7 15:27:40 2003 Heather Drury <drury_h@ociweb.com> - -Fri Feb 7 15:37:19 2003 Steve Huston <shuston@riverace.com> - - * include/makeinclude/platform_aix_ibm.GNU: Define DCCFLAGS with the - debugging options for xlC_r; Change DCFLAGS to just -g (for the C - compiler). Thanks to Trevor Fields <fields_t@ociweb.com> for - reporting this problem. - -Fri Feb 7 15:27:40 2003 Heather Drury <drury_h@ociweb.com> - - * ace/config-aix-4.x.h: Restored the defining of - ACE_TEMPLATES_REQUIRE_SOURCE for AIX VisualAge 5, - which is needed for building TAO. Thanks to Trevor - Fields <fields_t@ociweb.com> for this fix. - -Fri Feb 7 13:04:51 2003 Steve Huston <shuston@riverace.com> - - * ace/ace_wchar.h: Added a new macro, ACE_TEXT_ALWAYS_WCHAR(STRING). - It's the inverse of ACE_TEXT_ALWAYS_CHAR - whatever STRING is, - the macro produces a wide-char string. Thanks to - Emmanuel Thevenot Beaufort <emmanuel.thevenot-beaufort@jci.com> - for this macro. - -Thu Feb 6 19:30:18 2003 Steve Huston <shuston@riverace.com> - - * tests/SString_Test.cpp: Added a test for wide-string compares of - equal-length strings that differ in the last character. - - * ace/String_Base.i (compare): When memcmp()-ing to compare, take - the size of CHAR into account when calculating byte length. Thanks - to Emmanuel Thevenot Beaufort <emmanuel.thevenot-beaufort@jci.com> - for this fix. - -Thu Feb 6 16:06:29 2003 Rich Seibel <seibel_r@ociweb.com> - - * ace/config-tru64.h: Added a guard around the - ACE_HAS_SNPRINTF as versions earlier than 5.0 do not seem - have this function. - -Thu Feb 6 15:03:04 2003 Krishnakumar B <kitty@cs.wustl.edu> - - * include/makeinclude/rules.local.GNU: Make the removal of IDL - generated files in realclean target dependent upon the - definition of IDL_FILES. MingW make crashes on empty "rm -f " - commands. - -Thu Feb 6 15:19:28 2003 Heather Drury <drury_h@ociweb.com> - - * ace/Refcounted_Auto_Ptr.h: - * ace/Refcounted_Auto_Ptr.i: Made the other get() method const. See - previous ChangeLog entry. - -Thu Feb 6 07:13:06 2003 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * ace/Refcounted_Auto_Ptr.{h,i} (get): Made the get() method const. - Thanks to Jenny Kowald <jkowald@agile.tv> for suggesting this. - -Wed Feb 5 20:48:47 2003 Krishnakumar B <kitty@cse.wustl.edu> - - * include/makeinclude/rules.local.GNU: Added rules to generate - object files from C++ file extensions .cxx, .C and .cc, in - addition to .cpp. - -Wed Feb 5 20:15:18 2003 Krishnakumar B <kitty@cse.wustl.edu> - - * include/makeinclude/rules.lib.GNU: Removed the conditional - setting of VSHOBJS and VSHOBJS1 based on PRELIB. There is too - much bogosity here. This should fix HP-UX and AIX breakage. - -Sat Feb 1 08:43:53 2003 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * ace/Stream_Modules.cpp (put): Return 0 rather than -1 to be consistent - wrt the put() semantics elsewhere in ACE. Thanks to Jody Hagins - <jody@atdesk.com> for help with this. - - * tests/FIFO_Test.cpp (server): Only compile this on platforms - that aren't Win32 since FIFO's don't work there. Thanks to Carl - Halvorson <cyberdesk@yahoo.com> for reporting this. - -Wed Feb 5 09:45:12 UTC 2003 Johnny Willemsen <jwillemsen@remedy.nl> - - * include/makeinclude/platform_mingw32.GNU: - Removed desciption to MINGW_BASE. - -Tue Feb 4 13:13:08 2003 Krishnakumar B <kitty@cs.wustl.edu> - - * include/makeinclude/rules.lib.GNU: - * include/makeinclude/rules.bin.GNU: - - Generalized the creation of objects from source files by - removing the dependency that the source files be named *.cpp. - Also removed a lot of junk. - - * include/makeinclude/wrapper_macros.GNU: Removed some more - accumulated cruft. - - * include/makeinclude/rules.nested.GNU: Fix recursive make so - that users can name the makefiles in sub-directories as - something other than Makefile by introducing a new variable - SUBDIR_MAKEFILE. - - * ace/Makefile: Fix the Makefile to use the above extension when - recursing into directories. - - * include/makeinclude/rules.local.GNU: Add a new local target - clean_idl_stubs.local which cleans all the IDL generated files. - This target is linked to realclean so the IDL generated files - will only be cleaned with realclean. - - Thanks to James Mansion <james.mansion@uk.nomura.com> for - reporting the bugs and suggesting some fixes. - -Tue Feb 4 11:21:13 MST 2003 Rob Andzik <andzik@rtlogic.com> - - * ACE_wrappers/include/makeinclude/platform_vxworks5.x_g++.GNU - - Added three new variables - GCC_VERSION (defaults to gcc-2.96) - VXWORKS_VERSION_FLAG (defaults to -DACE_VXWORKS=0x542) - VXWORKS_STDINC_FLAG (defaults to <null>) - - - Modified all sets of GCCLIB_DIR to use GCC_VERSION - - Modified all references to ACE_VXWORKS to use VXWORKS_VERSION_FLAG - - replaced -nostdinc with VXWORKS_STDINC_FLAG - - * ACE_wrappers/ace/config-vxworks5.x.h - - Added define ACE_LACKS_CLEARERR (wrapped in ACE_VXWORKS check) - - * ACE_wrappers/ace/OS.i - - Added ACE_UNUSED_ARG where warnings occurred. - -Mon Feb 3 16:33:02 2003 Heather Drury <drury_h@ociweb.com> - - * ace/config-freebsd.h: msgsnd() with const parameter was recently - added to FreeBSD-CURRENT. Thanks to Craig Rodrigues' for this change. - -Sat Feb 1 09:34:45 2003 Ossama Othman <ossama@uci.edu> - - * bin/auto_run_tests.lst: - - Disable TAO's IORInterceptor test for the GIOP 1.0 build - configuration. The test requires IOR tagged components that are - not supported in GIOP 1.0. Fixes a false positive in our - scoreboard. - -Sat Feb 1 07:50 2003 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * Our deepest sympathies to the families and friends of the crew of - the US Space Shuttle Columbia, which was lost over Texas this - morning. - - "We shall never forget them nor the last time we saw them, as - they prepared for their mission and waved good-bye and slipped - the surly bonds of Earth to touch the face of God." - - - President Reagan (addressing NASA employees following the - tragic loss of the Challenger 7 crew on STS-51L.) - - Ad astra per aspera. - -Sat Feb 1 09:00:00 2003 Steve Huston <shuston@riverace.com> - - * Space Shuttle Columbia breaks up at 200,000ft altitude at - 12,500mph. 7 astronauts killed; the US and Israel mourn their loss. - -Fri Jan 31 12:55:26 2003 Frank Hunleth <frank@hunleth.com> - - * tests/Reactor_Exceptions_Test.cpp (worker): Added #if defined - (ACE_HAS_THREADS) guard around worker () function to remove a - compiler warning on single threaded configurations that have - exceptions enabled. - -Thu Jan 30 21:57:58 2003 Frank Hunleth <frank@hunleth.com> - - * examples/Threads/thread_specific.cpp (worker): Added __OpenBSD__ - to the list of OS/compilers that can check printf arguments. - This removes a compilation warning. - -Thu Jan 30 21:50:27 2003 Frank Hunleth <frank@hunleth.com> - - * ace/config-openbsd.h: Added ACE_LACKS_SETSCHED, since it turns - out that sched_setscheduler hasn't been implemented on OpenBSD. - - * bin/auto_run_tests.lst: Disabled the MT_Cubit test for OpenBSD, - since it fails completely when sched_params returns an error. - -Thu Jan 30 10:34:02 2003 Ossama Othman <ossama@uci.edu> - - From Olivier Brunet <o.brunet@free.fr> - * ace/SSL/SSL_Context.cpp (report_error): - - Corrected format specifier for wide string case. - -Wed Jan 29 12:31:51 2003 Frank Hunleth <frank@hunleth.com> - - * ace/config-openbsd.h: - * ace/config-openbsd-pthreads.h: Merged config files together and - removed config-openbsd-pthreads.h like what was done for - FreeBSD. - - * include/makeinclude/platform_openbsd.GNU: - * include/makeinclude/platform_openbsd_pthreads.GNU: Merged files - together and removed platform_openbsd_pthreads.GNU like - FreeBSD. Also removed gratuitous link of -lstdc++. g++ already - does that for us. Thanks to Craig Rodrigues' changelog entry. - -Wed Jan 29 11:47:29 2003 Steve Huston <shuston@riverace.com> - - * ACEXML/tests/ContentHandler_Test.cpp: Fixed compile errors by - adding/correcting use of the ACEXML exception macros. - -Wed Jan 29 08:30:12 UTC 2003 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/Vector_T.cpp: - Fixed bug in resize() method, of the original vector, the last - item was lost during resize(). Thanks to Denis Parnaland - <denis.parnaland@tde.alstom.com> for reporting this. - - * tests/Vector_Test.cpp: - Extended this test to test the resize() method of ACE_Vector. - - * THANKS: Added Denis Parnaland. - -Tue Jan 28 18:46:41 2003 Steve Huston <shuston@riverace.com> - - * ACEXML/parser/parser/Parser.cpp (parse_element): When converting - an escape sequence, add its length, don't lose the length of - already-parsed content. Thanks to Emmanuel Thevenot Beaufort - <emmanuel.thevenot-beaufort@jci.com> for this fix. - - * ACEXML/tests/ContentHandler_Test.{cpp dsp}: New test that - illustrates the problem above and validates the fix. - - * ACEXML/tests/Tests.dsw: - * ACEXML/Makefile: - * ACEXML/Makefile.bor: Added ContentHandler_Test. - - * THANKS: Added Emmanuel Thevenot Beaufort to the Hall of Fame. - -Tue Jan 28 15:47:29 2003 Steve Huston <shuston@riverace.com> - - * include/makeinclude/platform_hpux_aCC.GNU: For distrib builds, - changed the +h option to not specify the whole path; put the path - in the path search option (+b). - -Mon Jan 27 14:28:16 2003 Heather Drury <drury_h@ociweb.com> - - * include/makeinclude/wrapper_macros.GNU: Remove unnecessary use - of GNU make basename command. Supplied by Ganesh Pai <gpai@sonusnet.com>. - -Sun Jan 26 16:42:23 UTC 2003 Johnny Willemsen <jwillemsen@remedy.nl> - - * include/makeinclude/platform_linux_borland.GNU: - When AIO support is found, set the right define but also add the - library liblrt to the lists to link with. Only when building with - threading support link with the pthread library. - -Sun Jan 26 16:42:23 UTC 2003 Johnny Willemsen <jwillemsen@remedy.nl> - - * include/makeinclude/platform_linux_borland.GNU: - Don't set any BUILD_DLL defines in this file, the BUILD_DLL defines - are set in the specific makefiles. - -Sat Jan 25 20:31:56 2003 Frank Hunleth <frank@hunleth.com> - - * ace/config-openbsd.h: - * ace/config-openbsd-pthread.h: Added ACE_HAS_SNPRINTF since - OpenBSD has snprintf() and to fix buffer overflow tested for by - tests/Log_Msg_Test. - -Sat Jan 25 18:15:48 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: Added a new test to the daily builds. - -Sat Jan 25 13:39:23 UTC 2003 Johnny Willemsen <jwillemsen@remedy.nl> - - * ACE-INSTALL.html: - * include/makeinclude/platform_mingw32.GNU: - * include/makeinclude/platform_gnuwin32_common.GNU: - Removed the usage of MINGW_BASE. This is not needed for using MingW. - Thankx to Viktor Ransmayr <viktor.ransmayr@t-online.de> for - confirming that this is not needed. - -Fri Jan 24 07:54:12 UTC 2003 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/OS.h: - VxWorks 5.4 lacks some ANSI prototypes. These are defined by - VxWorks 5.5 so we only define these ANSI prototypes when defined - ACE_VXWORKS and this is equal or smaller 0x540. Else we just - include apra/inet.h. - -Thu Jan 23 17:57:05 2003 Heather Drury <drury_h@ociweb.com> - - * ace/README: Deleted non-ascii characters around ACE_HAS_PTHREAD_SETSTACK. - -Thu Jan 23 17:44:38 UTC 2003 Craig Rodrigues <crodrigu@bbn.com> - - * include/makeinclude/platform_freebsd.GNU: Remove gratuitous link - of -lstdc++. g++ already does that for us. - -Wed Jan 22 21:08:48 2003 Steve Huston <shuston@riverace.com> - - * examples/C++NPv2/WFMO_Reactor_Logging_Server.cpp (Quit_Handler): - Sync with code in the book. - -Wed Jan 22 18:12:41 2003 Steve Huston <shuston@riverace.com> - - * examples/C++NPv2/Service_Reporter.cpp: - * examples/C++NPv2/Configurable_Logging_Server.cpp: Change the - name of the static service descriptor from Reporter to - Reporter_Descriptor. This is more descriptive and matches - the book. - -Wed Jan 22 18:08:27 UTC 2003 Craig Rodrigues <crodrigu@bbn.com> - - * ace/config-freebsd.h: Add ACE_HAS_SNPRINTF. Fixes Log_Msg_Test. - -Wed Jan 22 18:03:35 UTC 2003 Craig Rodrigues <crodrigu@bbn.com> - - * include/makeinclude/platform_linux.GNU: - * include/makeinclude/platform_linux_icc.GNU: - * include/makeinclude/platform_linux_kcc.GNU: - Remove defines of _POSIX_THREADS and _POSIX_THREAD_SAFE_FUNCTIONS. - These POSIX constants are not meant to be defined by - the user. The user must include <unistd.h> and *check* - the value of these constants. On Linux glibc systems, - after including <unistd.h>, the header <bits/posix_opt.h> - will be eventually be included...these constants are - defined by the glibc implementation in that header. - -Wed Jan 22 10:55:39 2003 Steve Huston <shuston@riverace.com> - - * ace/config-tru64.h: Added ACE_HAS_SNPRINTF. Thanks to Vladimir - Chovanec <Vladimir.CHOVANEC@asset.sk> for this fix. - -Wed Jan 22 01:24:51 2003 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * ace/Message_Queue_T.h: Corrected the documentation for - ACE_Message_Block_Ex so it says ACE_MESSAGE_TYPE rather than - ACE_Message_Block. - -Tue Jan 21 21:44:04 2003 Nanbor Wang <nanbor@cs.wustl.edu> - - * ACEXML/examples/SAXPrint/broken.xml: Added a new broken XML file - to make sure we handle this case properly. - - * ACEXML/parser/parser/Parser.i (get): Added a check for - successful <get> operation from the input CharStream to avoid - out of the infinite loop when reading broken XML files. Thanks - to Volodymyr Orlenko <vorlenko@jaalam.com> for reporting this - and submitting the patch. - -Mon Jan 20 20:32:55 2003 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * ace/Log_Msg.cpp (log): Added a '}' at line 1531 to fix a - nasty compiler bug on Digital UNIX. Thanks to Igor Pisarenko - <Igor.Pisarenko@computershare.com.au> for reporting this. - -Tue Jan 21 19:29:04 2003 Heather Drury <drury_h@ociweb.com> - - * bin/auto_run_tests.lst: Disabled orbsvcs/Test/EC_MT_MCast on - static builds. - -Tue Jan 21 08:43:24 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/generate_doxygen.pl: Fixed a cut and paste error. - -Mon Jan 20 16:22:43 2003 Heather Drury <drury_h@ociweb.com> - - * ace/NT_Service.cpp: Modified return statement as the - wait_for_service_state function is declared as void and it was - returning an int. - -Mon Jan 20 15:48:35 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: Added a TAO test to the daily builds. - -Mon Jan 20 13:58:25 2003 Rich Seibel <seibel_r@ociweb.com> - - * ace/config-macosx.h: Added ACE_HAS_SNPRINTF as well as - ACE_LACKS_PERFECT_MULTICAST_FILTERING to make the ACE - Mac OSX 10.2.x build pass the Log_Msg_Test and the - Multicast_Test. Thanks to John Michael Zorko <j.zorko@att.net> - for providing this patch. - -Mon Jan 20 18:59:12 UTC 2003 Johnny Willemsen <jwillemsen@remedy.nl> - - * examples/C++NPv2/Makefile.CLD: - * examples/C++NPv2/Makefile.SLD: - * examples/C++NPv2/Makefile.SLDex: - * examples/C++NPv2/Makefile.TPCLS: - * examples/C++NPv2/Makefile.TPLS: - * ace/RMCast/Makefile: - Set the correct x_BUILD_DLL defines. These are needed for the - Cygwin/MingW and Kylix compiler. Thanks to Vince Mounts - <vince@mounts.cc> for providing the patches. - -Mon Jan 20 11:58:00 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/generate_doxygen.pl: Fixed the spaces in the file names - of generated man pages. Thanks to Olver Kellogg for providing - this patch. - -Sun Jan 19 11:40:10 2003 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * ace/NT_Service.{h,cpp}: Enhanced the wait_For_service_state() so - that it actually uses the wait_time parameter. Thanks to Theo - Landman <tlandman@justcroft.com> for contributing this. - - * ace/Process_Manager.cpp (spawn): Cleanup dynamically allocated - memory if the spawn() fails. Thanks to Kobi Cohen-Arazi - <kobi@mivzak.com> for this fix. - -Sun Jan 19 15:03:13 UTC 2003 Johnny Willemsen <jwillemsen@remedy.nl> - - * websvcs/lib/Makefile: - Set the correct x_BUILD_DLL defines. These are needed for the - Cygwin/MingW and Kylix compiler. - -Sat Jan 18 17:49:20 2003 Steve Huston <shuston@riverace.com> - - * Makefile: Removed PACE references. - -Sat Jan 18 17:11:21 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/make_release: Use doxygen 1.2.18 as opposed to doxygen - 1.2.13.1. - -Fri Jan 17 09:00:17 2003 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * ACE-INSTALL.html (make): Removed all claims that we support - SunC++ 4.2. Thanks to Kirat Singh <kirat.singh@gs.com> for - reporting this. - -Sat Jan 18 09:07:12 UTC 2003 Johnny Willemsen <jwillemsen@remedy.nl> - - * tests/Makefile.DLL_Test: - * tests/Makefile.Framework_Component_DLL: - * tests/Makefile.Service_Config_DLL: - * examples/Service_Configurator/IPC-tests/server/Makefile: - Set the correct x_BUILD_DLL defines. These are needed for the - Cygwin/MingW and Kylix compiler. - -Fri Jan 17 12:34:51 2003 Steve Huston <shuston@riverace.com> - - * ACE-INSTALL.html: Updated platform and compiler information. - -Thu Jan 16 19:06:19 2003 Stephen Torri <storri@cse.wustl.edu> - - * ace/OS.cpp: - * ace/config-linux-common.h: - * ace/config-freebsd.h: - * ace/README: - - Added a new macro called ACE_HAS_PTHREAD_SETSTACK for systems - that use the newer glibc (2.3+). Eliminates linker warning about - not using the pthread function called pthread_setstack. - -Thu Jan 16 16:25:45 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: Added a TAO/orbsvcs/tests/EC_MT_Mcast - test to the daily builds. - -Thu Jan 16 12:27:24 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * COPYING: Extended the validity of the copyright notice from 2002 - to 2003. Should have been done yesterday, anyway, better late - than never. - -Wed Jan 15 18:02:01 2003 Bala <bala@cs.wustl.edu> - - * ACE version 5.3 released. - -Wed Jan 15 10:48:11 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/make_release: Did the following enhancements - - - If it is a minor release do not append the beta version. Looks - like somone broke this and the script started writing out - things as x.3.0, whcih is bad BTW. - - - We now POSIX.1 tar, instead of old tar format with the -H - option in cpio. This helps to tar files deep in the hierarchy - tree to be properly zipped. We were loosing this in many - betas. - - - Fixes for the generated filenames copied into the - previous_versions directory. - - - Fix to get the right Version.h file in TAO. - - -Tue Jan 14 18:51:49 2003 Steve Huston <shuston@riverace.com> - - * ace/Auto_Ptr.h: Added pragma to disable warning 4284 for Microsoft - compilers. The warned-of behavior is exactly what is intended. - - * bin/MakeProjectCreator/templates/em3vcpdll.mpt: - * bin/MakeProjectCreator/templates/em3vcpdllexe.mpt: - * bin/MakeProjectCreator/templates/em3vcplib.mpt: - * bin/MakeProjectCreator/templates/em3vcplibexe.mpt: - Added a "WCE emulator" configuration section. This allows builds - for the WinCE emulator, necessary for working with the WinCE.NET - Standard SDK emulator. - -Tue Jan 14 14:56:12 2003 Steve Huston <shuston@riverace.com> - - * ace/config-win32-common.h: For WinCE, don't include wce.h unless - building with MFC. wce.h is an MFC file. - - * ace/config-WinCE.h: Add pragmas to link _with_ corelibc.lib and - _without_ oldnames.lib. These weren't needed with WinCE 3/PocketPC - (but they don't hurt) and are needed for WinCE .NET. The ATL headers - do these pragmas, but we don't want to add a dependency on ATL. - -Tue Jan 14 11:07:06 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/generate_export_file.pl: Added a patch from - Gonzalo Diethelm <gonzalo.diethelm@aditiva.com> which generates - an empty CVS Id string. - -Mon Jan 13 12:53:12 UTC 2003 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/config-win32-common.h: - The MingW compiler can handle the export/import macros. - - * ace/config-cygwin32.h: - Corrected the guards around this header file. - -Sat Jan 11 09:37:24 2003 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * ace/OS_String.cpp (wcslen_emulation): Fixed a nasty formatting - glitch that made the code hard to read. Thanks to Ruslan - Zasukhin <sunshine@public.kherson.ua> for reporting this. - -Fri Jan 3 21:46:00 2003 John Michael Zorko <j.zorko@att.net> - - * netsvcs/clients/Naming/Client/Makefile: Added $(ACELIB) to - ACE_SHLIBS to fix multple -lACE problem. This wasn't an - issue on Mac OS X 10.2, but could be on other platforms. - - * examples/ASX/CCM_App/Makefile: same as above - * netsvcs/clients/Naming/Dump_Restore/Makefile: same as above - -Wed Jan 8 19:27:49 2003 Steve Huston <shuston@riverace.com> - - * ace/Message_Block.h: Doxygenized comments for release() and copy(). - - * ace/UPIPE_Stream.{h cpp}: Removed the remaining_ member; use the - ACE_Message_BLock::length() member function to find the amount - of remaining data. - -Wed Jan 8 12:17:10 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/ace-install: Fixed the improper usage of "or" in the "find " - command used to find the inlined files with extensions .i and - .inl. Thanks to Oliver Kellog for providing the patch. - -Tue Jan 7 13:08:12 UTC 2003 Johnny Willemsen <jwillemsen@remedy.nl> - - * ACEXML/common/Makefile: - Set ACEXML_BUILD_DLL when building the shared lib. - -Mon Jan 6 18:31:09 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/generate_topinfo_charts.sh: Removed the range for the - Y-AXIS, for the graphs to auto-adjust while plotting. - -Mon Jan 6 16:56:04 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/topinfo_iorsize_stats.sh: Scripts for capturing the size of - the server and client when shipping large IOR's - - * bin/topinfo_stats.sh: Renamed to topinfo_simple_stats.sh. This - is a better to show that we capture the sizes of a simple server - and client. - - * bin/topinfo_simple_stats.sh: New file. We need to do some code - sharing between these two scripts. Just postponing that for the - timebeing. - -Mon Jan 6 11:23:12 UTC 2003 Johnny Willemsen <jwillemsen@remedy.nl> - - * include/makeinclude/platform_gnuwin32_common.GNU: - Removed setting ACE_OS_HAS_DLL to 1, this should be done more - specific in the makefiles - -Sat Jan 4 18:46:19 2003 Steve Huston <shuston@riverace.com> - - * ace/config-aix5.1.h: Add a #define ACE_AIX_VERS 501 if ACE_AIX_VERS - isn't already defined. Helps to build with Visual Age C++ incremental - mode. Support for incremental mode is going away after ACE 5.3, but - this crutch can be used for now. - - * ace/Makefile.ace: Remove SOCK_Dgram_Mcast_T from TEMPLATE_FILES. See - Tue Jun 25 23:22:09 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * ace/ace.icc: Corrected the list of source files to build ACE with. - In addition to updating the source list, if building with Visual - Age C++ 5, add the TEMPLATE_FILES to the sources that get compiled. - Visual Age C++ 5 incremental needs to see them all. - - * examples/C++NPv1/Makefile: Add .NOTPARALLEL since some of the - sub-makes share object files and they can get confused. - - * tests/makeicc.pl: Add additional template-getting stuff for - Visual Age C++ 5. - - * tests/ARGV_Test.icc: - * tests/Aio_Platform_Test.icc: - * tests/Arg_Shifter_Test.icc: - * tests/Atomic_Op_Test.icc: - * tests/Auto_IncDec_Test.icc: - * tests/Barrier_Test.icc: - * tests/Basic_Types_Test.icc: - * tests/Bound_Ptr_Test.icc: - * tests/Buffer_Stream_Test.icc: - * tests/CDR_Array_Test.icc: - * tests/CDR_File_Test.icc: - * tests/CDR_Test.icc: - * tests/Cache_Map_Manager_Test.icc: - * tests/Cached_Accept_Conn_Test.icc: - * tests/Cached_Allocator_Test.icc: - * tests/Cached_Conn_Test.icc: - * tests/Capabilities_Test.icc: - * tests/Codecs_Test.icc: - * tests/Collection_Test.icc: - * tests/Config_Test.icc: - * tests/Conn_Test.icc: - * tests/DLL_Test.icc: - * tests/DLL_Test_Impl.icc: - * tests/DLList_Test.icc: - * tests/Date_Time_Test.icc: - * tests/Dev_Poll_Reactor_Test.icc: - * tests/Dirent_Test.icc: - * tests/Dynamic_Priority_Test.icc: - * tests/Enum_Interfaces_Test.icc: - * tests/Env_Value_Test.icc: - * tests/FIFO_Test.icc: - * tests/FlReactor_Test.icc: - * tests/Framework_Component_DLL.icc: - * tests/Framework_Component_Test.icc: - * tests/Future_Set_Test.icc: - * tests/Future_Test.icc: - * tests/Get_Opt_Test.icc: - * tests/Handle_Set_Test.icc: - * tests/Hash_Map_Bucket_Iterator_Test.icc: - * tests/Hash_Map_Manager_Test.icc: - * tests/High_Res_Timer_Test.icc: - * tests/INET_Addr_Test.icc: - * tests/IOStream_Test.icc: - * tests/Lazy_Map_Manager_Test.icc: - * tests/Log_Msg_Test.icc: - * tests/Logging_Strategy_Test.icc: - * tests/MEM_Stream_Test.icc: - * tests/MM_Shared_Memory_Test.icc: - * tests/MT_Reactor_Timer_Test.icc: - * tests/MT_Reactor_Upcall_Test.icc: - * tests/MT_SOCK_Test.icc: - * tests/Malloc_Test.icc: - * tests/Map_Manager_Test.icc: - * tests/Map_Test.icc: - * tests/Max_Default_Port_Test.icc: - * tests/Mem_Map_Test.icc: - * tests/Message_Block_Test.icc: - * tests/Message_Queue_Notifications_Test.icc: - * tests/Message_Queue_Test.icc: - * tests/Message_Queue_Test_Ex.icc: - * tests/Multicast_Test.icc: - * tests/Naming_Test.icc: - * tests/New_Fail_Test.icc: - * tests/Notify_Performance_Test.icc: - * tests/OS_Test.icc: - * tests/Object_Manager_Test.icc: - * tests/Obstack_Test.icc: - * tests/OrdMultiSet_Test.icc: - * tests/Pipe_Test.icc: - * tests/Priority_Buffer_Test.icc: - * tests/Priority_Reactor_Test.icc: - * tests/Priority_Task_Test.icc: - * tests/Proactor_Scatter_Gather_Test.icc: - * tests/Proactor_Test.icc: - * tests/Proactor_Timer_Test.icc: - * tests/Process_Manager_Test.icc: - * tests/Process_Mutex_Test.icc: - * tests/Process_Strategy_Test.icc: - * tests/RB_Tree_Test.icc: - * tests/Reactor_Dispatch_Order_Test.icc: - * tests/Reactor_Exceptions_Test.icc: - * tests/Reactor_Notify_Test.icc: - * tests/Reactor_Performance_Test.icc: - * tests/Reactor_Timer_Test.icc: - * tests/Reactors_Test.icc: - * tests/Reader_Writer_Test.icc: - * tests/Recursive_Condition_Test.icc: - * tests/Recursive_Mutex_Test.icc: - * tests/Refcounted_Auto_Ptr_Test.icc: - * tests/Reverse_Lock_Test.icc: - * tests/SOCK_Connector_Test.icc: - * tests/SOCK_Send_Recv_Test.icc: - * tests/SOCK_Test.icc: - * tests/SPIPE_Test.icc: - * tests/SString_Test.icc: - * tests/SV_Shared_Memory_Test.icc: - * tests/Semaphore_Test.icc: - * tests/Service_Config_DLL.icc: - * tests/Service_Config_Test.icc: - * tests/Signal_Test.icc: - * tests/Sigset_Ops_Test.icc: - * tests/Simple_Message_Block_Test.icc: - * tests/Svc_Handler_Test.icc: - * tests/TP_Reactor_Test.icc: - * tests/TSS_Test.icc: - * tests/Task_Test.icc: - * tests/Thread_Manager_Test.icc: - * tests/Thread_Mutex_Test.icc: - * tests/Thread_Pool_Reactor_Resume_Test.icc: - * tests/Thread_Pool_Reactor_Test.icc: - * tests/Thread_Pool_Test.icc: - * tests/Time_Service_Test.icc: - * tests/Time_Value_Test.icc: - * tests/Timeprobe_Test.icc: - * tests/Timer_Queue_Test.icc: - * tests/TkReactor_Test.icc: - * tests/Token_Strategy_Test.icc: - * tests/Tokens_Test.icc: - * tests/UPIPE_SAP_Test.icc: - * tests/Upgradable_RW_Test.icc: - * tests/Vector_Test.icc: - * tests/XtReactor_Test.icc: Mostly updated to build with Visual Age - C++ 5 in incremental mode. Most of the tests that use templates - don't build. They require explicitly including the needed template - source files names in the .icc file. I don't have time to add them - all now - if anyone uses this compiler, please contact me. No wonder - IBM ditched this piece of junk. - -Fri Jan 3 17:58:21 2003 Steve Huston <shuston@riverace.com> - - * ace/SSL/SSL_Context.cpp: - * netsvcs/lib/Name_Handler.cpp: - * netsvcs/lib/Server_Logging_Handler.cpp: Add __hpux to condition - where template members are explicitly instantiated. Also see: - Tue Dec 3 20:47:39 2002 Steve Huston <shuston@riverace.com> - -Fri Jan 3 17:45:29 2003 Steve Huston <shuston@riverace.com> - - * include/makeinclude/platform_hpux_gcc.GNU: Use HP's /usr/bin/ar - instead of GNU ar (which may be picked up first in PATH). GNU ar - crashes - missing libfl.sl. - - * performance-tests/Misc/test_singleton.cpp: Add __hpux to conditions - for explicitly instantiating ACE_Singleton::singleton_. - -Fri Jan 3 07:49:09 2003 Chad Elliott <elliott_c@ociweb.com> - - * bin/aix_shr: - - Changed the name of the AIX template instantiation macro and - commented on why it's needed. - -Thu Jan 2 17:44:29 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/config-cygwin32-common.h: Set the custom export - macro's. These will be used when building dynamically with the - cygwin compiler. - - * bin/MakeProjectCreator/templates/gnu.mpd: Generate dllflags and - libflags into the GNU makefiles. These are important for the - Cygwin/MingW and Kylix compiler. - -Thu Jan 2 16:11:38 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/config-freebsd.h: Fix use of macros for FreeBSD 4.6: - ACE_LACKS_RWLOCK_T, ACE_LACKS_READDIR_R, ACE_LACKS_SETSCHED, - ACE_LACKS_PTHREAD_THR_SIGSETMASK. Do not define SCHED_RR, - SCHED_IO, SCHED_FIFO, SCHED_OTHER for FreeBSD >= 4.0, since they - are defined in sched.h. Thanks to Craig Rodrigues for providng - the patch. - -Thu Jan 2 13:33:01 2003 Chad Elliott <elliott_c@ociweb.com> - - * ace/Refcounted_Auto_Ptr.i: - - Added #include of ace/Log_Msg.h for ACE_ASSERT. - - * bin/aix_shr: - - Added -DTAO_INSTANTIATING_TEMPLATES to the link command line. - -Thu Jan 2 12:38:55 2003 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/Makefile.ace: Set the dllflags in Makefile. This makes it - possible to do __declspec (dllexport) for the symbols in the - dll. This is needed for the Cygwin/MingW and Kylix compiler. - - Thid change takes care of the static and dynamic library cases. - -Thu Jan 2 08:57:26 2003 Johnny Willemsen <jwillemsen@remedy.nl> - - * include/makeinclude/platform_cygwin32.GNU: Add - --enable-auto-import to the linker flags. This will result in - the linker doing auto-import without giving warnings about - it. - -Wed Jan 1 14:26:09 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/topinfo_stats.sh: Forgot that we run this scripts in cron - environment. Set the path and other details that would help to - ease things. - -Tue Dec 31 21:21:26 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/generate_topinfo_charts.sh: - * bin/topinfo_stats.sh: Scripts to track the memory usage and plot - them. At present it works only for the simple test in - $TAO_ROOT/performance-tests/Memory/Single_Threaded. - -Tue Dec 31 17:57:19 2002 Steve Huston <shuston@riverace.com> - - * examples/Service_Configurator/IPC-tests/server/Handle_Thr_Stream.h: - Add #include "ace/Svc_Handler.h" - -Tue Dec 31 17:50:39 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.h: For WinCE 4 (.NET) and up, include <stddef.h> rather than - typedef-ing ptrdiff_t. - -Tue Dec 31 17:26:41 2002 Steve Huston <shuston@riverace.com> - - * tests/tests.mpc (DLL Test Lib): Add dllflags = ACE_BUILD_SVC_DLL to - add the correct /D option. The source uses ACE_Svc_Export to flag - dll import/export correctly. - - * tests/CE_fostream.h: #include "ace/config-all.h" before checking for - #if defined (ACE_HAS_WINCE). - -Tue Dec 31 15:05:42 2002 Steve Huston <shuston@riverace.com> - - * ACEXML/apps/svcconf/Svcconf.cpp: Changed the condition for when - a dummy entrypoint is needed from AIX 4.3 to AIX and Visual Age C++ - version 5. Thanks for Jon Wackley for this adjustment. Also see: - Fri Dec 27 19:46:31 2002 Steve Huston <shuston@riverace.com> - - * netsvcs/clients/Naming/Client/Makefile: If building this stuff, - reset OBJDIRS to $(VDIR) and $(VSHDIR) to clean up both. - -Mon Dec 30 21:30:41 2002 Steve Huston <shuston@riverace.com> - - * include/makeinclude/platform_hpux_gcc.GNU: Removed the SOVERSION= - line to allow versions libraries to be set up properly. - - * ace/Based_Pointer_Repository.cpp: - * ace/Dynamic.cpp: - * ace/SString.cpp: - * tests/Framework_Component_DLL.cpp: - * tests/MEM_Stream_Test.cpp: - * tests/Process_Strategy_Test.cpp: - * tests/Timeprobe_Test.cpp: - * apps/drwho/File_Manager.cpp: - * examples/IPC_SAP/SOCK_SAP/CPP-inclient.cpp: - * examples/IPC_SAP/SOCK_SAP/CPP-inserver-fancy.cpp: - * examples/Logger/Acceptor-server/server_loggerd.cpp: - * examples/Logger/simple-server/server_loggerd.cpp: - * examples/Threads/auto_event.cpp: - * examples/Threads/tss2.cpp: - Add __hpux to the conditions needed to instantiate - static template members. Also see - Sat Nov 23 18:12:37 2002 Steve Huston <shuston@riverace.com> - -Mon Dec 30 19:32:18 2002 Steve Huston <shuston@riverace.com> - - * tests/Codecs_Test.cpp: Add explicit template instantiation for - ACE_Auto_Basic_Array_Ptr<ACE_Byte>. - -Mon Dec 30 18:44:20 Steve Huston <shuston@riverace.com> - - * examples/C++NPv2/TP_Logging_Server.cpp: - * examples/IPC_SAP/SOCK_SAP/CPP-inclient.cpp: - * examples/IPC_SAP/SOCK_SAP/CPP-inserver-fancy.cpp: - * examples/Logger/Acceptor-server/server_loggerd.cpp: - * examples/Logger/simple-server/server_loggerd.cpp: - * examples/Threads/auto_event.cpp: - * examples/Threads/tss2.cpp: - * performance-tests/Misc/test_singleton.cpp: - More singleton_ instantiations for g++ on AIX. - -Mon Dec 30 18:11:39 2002 Steve Huston <shuston@riverace.com> - - * apps/drwho/File_Manager.cpp: Add explicit instantiation of - ACE_Singleton<File_Manager, ACE_Null_Mutex>::singleton_. - Similar to those done for: - Tue Dec 3 20:47:39 2002 Steve Huston <shuston@riverace.com> - -Mon Dec 30 17:38:31 2002 Steve Huston <shuston@riverace.com> - - * examples/Service_Configurator/IPC-tests/server/Handle_Thr_Stream.cpp: - For Visual Age C++ 5 and up on AIX, explicitly instantiate the - CLI_Stream<> constructor. The compiler misses this one when - instantiating templates. Thank you very much to Jonathan - Wackley <jonw@mountaincable.net> for diagnosing this problem and - leading to a workaround. - -Mon Dec 30 15:25:26 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ACEXML/common/StrCharStream.cpp (open): - * ACEXML/common/FileCharStream.cpp (open): Changed to allow open - to succeed if automatic determination of encoding fails. This - allows the "Simple parsing" feature of ACEXML to work. Thanks - to Michael F"olsl <michael.foelsl@gmx.net> for reporting the - problem. - - * ACEXML/examples/SAXPrint/main.cpp: Added a new option to test - out ACEXML's simpling parsing feature. - - * ACEXML/examples/SAXPrint/simple.svcconf.xml: Added this new - "simple" XML document to test out the "Simple" parsing feature - in ACEXML. - -Mon Dec 30 12:21:33 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/MakeProjectCreator/config/taoversion.mpb: Fixed the version - information. This should automatically get updated during the - next beta/release. - -Sun Dec 29 18:27:38 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: Prevented BiDirGIOP tests from running - in GIOP 10 builds. - -Sun Dec 29 16:43:03 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ACEXML/common/FileCharStream.cpp (close): Changed to reset the - file pointer, filename_ and encoding_ strings after they are - release. Thanks to Andrew Marlow <apm35@student.open.ac.uk> for - suggesting the change. - -Sun Dec 29 00:47:58 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: Security tests using SSLIOP will not be - run in static builds. - -Sat Dec 28 23:05:51 2002 Steve Huston <shuston@riverace.com> - - * bin/make_release: Use -relative to replace ACE_ROOT with a relative - path in generated makefiles. - -Sat Dec 28 13:16:24 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/MEM_IO.cpp (recv_buf): Reverted my previous hack in: - - Fri Dec 27 08:38:29 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - and make the connection close handling more robust. - - * tests/MEM_Stream_Test.cpp: Reverted my previous change in: - - Sat Dec 28 11:04:38 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - It was not necessary. Added debug information for close - connection. - -Sat Dec 28 11:04:38 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * tests/MEM_Stream_Test.cpp: Made sure this test ignore the EIO - error from the following change when a connection was closed. - - Fri Dec 27 08:38:29 2002 Nanbor Wang <nanbor@cs.wustl.edu> - -Fri Dec 27 20:06:39 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/Malloc_T.h: - * ace/Malloc_T.i: - * ace/Malloc_T.cpp: Added a new ACE_Malloc_T::bad () method to - query whether a Malloc_T object was constructed successfully. - This approach, IMO, is sub-optimal compared to using an explicit - init method approach. The existing constructor signatures, - however, prevent us from doing so without breaking backward - compatibility. - - * ace/MEM_SAP.cpp: Now that there's a way to check whether the - underlying Malloc is functioning or not. Changed MEM_SAP to - check for proper initialization of underlying Malloc object and - clean up OS resources if the Malloc object is defected. This - change makes MEM_Stream transport mechanism more robust. - -Fri Dec 27 19:46:31 2002 Steve Huston <shuston@riverace.com> - - * ACEXML/apps/svcconf/Svcconf.cpp: Add a dummy entrypoint when - ACE_USES_CLASS_SVC_CONF == 0 on AIX 4.3. Else the AIX 4.3 linker - complains there are no csects or exported symbols saved at link - time. Thanks to Jon Wackley <jonw@mountaincable.net> for reporting - this problem. - -Fri Dec 27 17:43:29 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.h: For WinCE, only typedef ptrdiff_t for _WIN32_WCE < 400. - eMbedded Visual C++ 4 (Win CE.NET) defines this type. - - * bin/MakeProjectCreator/templates/em3vcp.mpd: Removed /M$(CECrtMT) - from the CPP options. Neither evc3 nor 4 support any /M options to - set the C run-time like MSVC does. It's controlled via the _DEBUG - macro only. This also allows the generated vcp file to be used - with either evc3 or evc4 (Windows CE .NET). evc3 ignored the option, - but evc4 flags it as unknown. - -Fri Dec 27 08:38:29 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/MEM_IO.cpp (recv_buf): Changed to return error (-1) and set - the errno to EIO when we receive 0 byte off the signaling socket - as this should never happen under normal circumstance. However, - on Win32 and Solaris, when one end of a connection closes the - socket while the other end is still reading the socket, the - active end will continuously receive several 0 bytes messages. - This confuses the hack of the MEM_IO mechanism and causes the - memory management routine to fall into a tight spinning - deadlock. - -Fri Dec 27 07:49:28 2002 Craig Rodrigues <crodrigu@bbn.com> - - * ace/config-macosx.h: Remove temporary hacks for buggy - gcc 3.1 compiler. Thanks to John Zorko for reminding me about - this. - - * examples/Service_Configurator/IPC-tests/server/Makefile: - * examples/Service_Configurator/Misc/Makefile: - * examples/Timer_Queue/Makefile: - Add ACELIBS to ACE_SHLIBS to fix MacOS X linking problem. - -Thu Dec 26 17:11:54 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/make_release: Fixed the following problems in the release - script which caused problems during the beta cutting process: - - - cvs commits of TAO/Changelog TAO/VERSION etc. failed. The - failure was because we were also commiting taoversion.mpb but - from $TAO_ROOT as "cvs commit ChangeLog VERSION - other_files ../bin/Makeyada". Using ../ to indicate files in - higher directories is an unpardonable mistake for cvs - 1.10.x. It used to work 1.9.x. But with 1.10* versions the - protocol changed disallowing the ../ - - - we didnt set the file permissions on the generated files - properly - - - the generated files from mpc were not added to the list of - files for which LF->CRLF conversions are supposed to be - masked during zip! - -Thu Dec 26 16:17:41 2002 Steve Huston <shuston@riverace.com> - - * tests/tests.mpc: For DLL Test Lib, the source is DLL_Test_Impl.cpp, - not DLL_Test.cpp. - -Thu Dec 26 09:14:38 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * Makefile: Removed configure.in from the list of controlled files. - -Thu Dec 26 07:36:10 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * include/makeinclude/platform_freebsd_pthread.GNU: Removed for - sure now. This should have been removed in this "Thu Nov 14 - 22:47:45 UTC 2002 Craig Rodrigues <crodrigu@bbn.com>" - checkin. Again, thanks to Craig for catching this. - -Thu Dec 26 07:31:42 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * apps/Gateway/Gateway/Makefile: Fixed a ttypo. Thanks to Craig - Rodrigues for catching this. - -Wed Dec 25 17:47:12 2002 Bala <bala@cs.wustl.edu> - - * ACE version 5.2.8 released. - -Wed Dec 25 17:39:05 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/make_release: Fixed problems with generation of vcproj etc. - -Wed Dec 25 06:14:17 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * Merry Christmas to all. May we have peace and justice for all in 2003. - -Tue Dec 24 17:29:11 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * docs/ACE-subsets.html: Documented the compiler options that - Steve Vranyes experimented and tested with to get a good - footprint reduction on Solaris. - -Tue Dec 24 07:31:54 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/Argv_Type_Converter.h: - * ace/Argv_Type_Converter.cpp: Reverted the following changes "Mon - Dec 23 17:26:42 2002 Steve Huston <shuston@riverace.com>" and - "Mon Dec 23 22:57:18 2002 Balachandran Natarajan - <bala@isis-server.isis.vanderbilt.edu>". The argument parsing in - TAO gets busted due to these changes and most of tests in our - daily builds fail. - -Mon Dec 23 22:57:18 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/Argv_Type_Converter.h: Changed saved_argc_ to be of type int - as opposed to int &. This seems to be necessitated by the change - "Mon Dec 23 17:26:42 2002 Steve Huston <shuston@riverace.com>" - which gives compile errors on all platforms. - -Mon Dec 23 21:07:20 2002 Steve Huston <shuston@riverace.com> - - * bin/MakeProjectCreator/templates/em3vcpdll.mpt: Changed the Debug - define from DEBUG to _DEBUG, matching all others, and what the - ACE files expect. Fixes mismatch inlines in ACE tests. - - * tests/tests.mpc: Added CE_fostream.cpp to the CDR_File_Test inputs. - - * tests/CE_fostream.{cpp h}: Added #if defined (ACE_HAS_WINCE) to - prevent trying to compile this on non-CE systems. - -Mon Dec 23 17:26:42 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.h (main): For WinCE, eVC can't grok the #include within the - 'main' macro expansion, so #include "ace/Argv_Type_Converter.h" is - moved outside the macro definition. - - * ace/Argv_Type_Converter.{h cpp} (ctor): Change argc from int& to - int, else it can't be used in the way it's already coded into use - in the 'main' definition (above, OS.h). - - * tests/Dev_Poll_Reactor_Test.cpp: - * tests/Proactor_Test.cpp: - * tests/Proactor_Scatter_Gather_Test.cpp: Changed main() to - ACE_TMAIN(). - - * tests/FlReactor_Test.cpp: - * tests/XtReactor_Test.cpp: Add /**/ between #include and filename - to silence MS warnings about missing files during dependency updates. - -Mon Dec 23 16:16:05 2002 Boris S Kolpackov <boris@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: Added - TAO/orbsvcs/tests/Security/Crash_Test to the list of tests. - -Mon Dec 23 17:08:17 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * THANKS: Changed Boris's information in the Hall of Fame. - -Mon Dec 23 12:13:51 2002 Steve Huston <shuston@riverace.com> - - * examples/C++NPv1/Process_Per_Connection_Logging_Server.h: Add - 'public' keyword to Process_Per_Connection_Logging_Server's - inheritance from Logging_Server. - - * examples/C++NPv2/Makefile: Add a $(RM) -rf tempinc after each - mkfile make to clean up template instantiations for Visual Age C++ - on AIX. - - * examples/C++NPv2/Makefile.SLD: Add Logging_Handler to the files - comprising this service. - - * examples/C++NPv2/Makefile.SLDex: Added Logging_Acceptor, - Logging_Event_Handler, and Logging_Handler to the files for this - service. Removed the TEMPINC trick for Visual Age C++ to keep the - templates from each compile around for the link stage where they're - needed. - - * examples/C++NPv2/Makefile.CLD: Added Logging_Handler to files. Remove - TEMPINC destruction from each compile. - - * examples/C++NPv2/Makefile.TPLS: Add Logging_Handler to files. Remove - TEMPINC destruction from each compile. - -Mon Dec 23 14:05:21 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/config-cygwin32-common.h: - When ACE_IOV_MAX is not defined, define it is 64. The Cygwin header - files define IOV_MAX as MAX_INT-1 and ACE_IOV_MAX is defined as - IOV_MAX by default. Allocating an array of this size on the stack - results in compile errors, so we define ACE_IOV_MAX to a smaller and - more realistic number. - -Mon Dec 23 12:16:33 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/RMCast/RMCast_IO_UDP.cpp: - * ace/RMCast/RMCast_Fragment.cpp: - Use ACE_IOV_MAX instead of IOV_MAX. - -Mon Dec 23 12:02:33 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * bin/ace_ld: - Extended this perl script so that the location of the __ctordtor.o - can be passed as -o command line option. The location is dependent - on the $VDIR in the vxworks make files. - - * include/makeinclude/platform_vxworks5.x_diab.GNU: - * include/makeinclude/platform_vxworks5.x_g++.GNU: - When calling the ace_ld perl script, pass the $VDIR as command line option - -o to the script. - -Sun Dec 22 11:41:56 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/Connector.cpp: Fixed fuzz errors. - -Sun Dec 22 11:37:44 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * THANKS: Added Daniel White to the hall of fame! - -Sun Dec 22 11:31:53 2002 Craig Rodrigues <crodrigu@bbn.com> - - * apps/Gateway/Gateway/Makefile: - * apps/Gateway/Peer/Makefile: Add $(ACELIB) to $(ACE_SHLIBS) so - that -lACE is added to the linkline when a shared library is - built. Solves linking problem MacOS X. Thanks to John Zorko and - Daniel white <ygor@comcast.net> for pointing this out. - -Sun Dec 22 10:25:35 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/Connector.cpp: Fixed a warning with g++. - -Sun Dec 22 09:18:00 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/Connector.cpp: - * ace/Connector.h: Added a lock to the ACE_Connector class to - make the following atomic - - . Registration of AST with the handler_map_ - . Registration of the ACE_Connector with the Reactor - . and the registration of timers with the Reactor's timer - queue. - - If they are not atomic, it throws open race conditions such as - the ones documented in BUG 1405. This is not complete fix for - BUG 1405, since it taxes applications using only one thread with a - lock. This is a short-term fix to prevent the race condition - from occuring. - - Other relevant changes include - - . the AST is now refcounted - . the AST cannot be deleted directly. The lifetime of the AST - can only be manipulated using the refcount on the AST. - -Sat Dec 21 18:27:42 2002 Steve Huston <shuston@riverace.com> - - * ace/ace_dll.vcp: - * ace/ace_os_dll.vcp: - * ace/ace.vcw: eMbedded Visual C++ workspace/projects deleted. These - need to be generated from ace.mpc when needed. The make_release - script now generates them for each source kit. When needed on a - CVS workspace build, use mwc/mpc. The autobuild system has commands - generate_workspace and generate_makefile that accomplish this. - See autobuild configs in the riverace directory for examples. - - * tests/tests_WinCE.vcw: - * tests/ARGV_Test_WinCE.vcp: - * tests/Atomic_Op_Test_WinCE.vcp: - * tests/Auto_IncDec_Test_WinCE.vcp: - * tests/Barrier_Test_WinCE.vcp: - * tests/Basic_Types_Test_WinCE.vcp: - * tests/Bound_Ptr_Test_WinCE.vcp: - * tests/Buffer_Stream_Test_WinCE.vcp: - * tests/Cached_Accept_Conn_Test_WinCE.vcp: - * tests/Cached_Conn_Test_WinCE.vcp: - * tests/Cache_Map_Manager_Test_WinCE.vcp: - * tests/Capabilities_Test_WinCE.vcp: - * tests/CDR_Array_Test_WinCE.vcp: - * tests/CDR_File_Test_WinCE.vcp: - * tests/CDR_Test_WinCE.vcp: - * tests/Collection_Test_WinCE.vcp: - * tests/Conn_Test_WinCE.vcp: - * tests/DLList_Test_WinCE.vcp: - * tests/DLL_Test_DLL_WinCE.vcp: - * tests/DLL_Test_WinCE.vcp: - * tests/Dynamic_Priority_Test_WinCE.vcp: - * tests/Enum_Interfaces_Test_WinCE.vcp: - * tests/Future_Set_Test_WinCE.vcp: - * tests/Future_Test_WinCE.vcp: - * tests/Handle_Set_Test_WinCE.vcp: - * tests/Hash_Map_Bucket_Iterator_Test_WinCE.vcp: - * tests/Hash_Map_Manager_Test_WinCE.vcp: - * tests/High_Res_Timer_Test_WinCE.vcp: - * tests/Lazy_Map_Manager_Test_WinCE.vcp: - * tests/Logging_Strategy_Test_WinCE.vcp: - * tests/Log_Msg_Test_WinCE.vcp: - * tests/Malloc_Test_WinCE.vcp: - * tests/Map_Manager_Test_WinCE.vcp: - * tests/Map_Test_WinCE.vcp: - * tests/Max_Default_Port_Test_WinCE.vcp: - * tests/Mem_Map_Test_WinCE.vcp: - * tests/MEM_Stream_Test_WinCE.vcp: - * tests/Message_Block_Test_WinCE.vcp: - * tests/Message_Queue_Notifications_Test_WinCE.vcp: - * tests/Message_Queue_Test_Ex_WinCE.vcp: - * tests/Message_Queue_Test_WinCE.vcp: - * tests/MM_Shared_Memory_Test_WinCE.vcp: - * tests/MT_Reactor_Timer_Test_WinCE.vcp: - * tests/MT_Reactor_Upcall_Test_WinCE.vcp: - * tests/MT_SOCK_Test_WinCE.vcp: - * tests/Naming_Test_WinCE.vcp: - * tests/New_Fail_Test_WinCE.vcp: - * tests/Notify_Performance_Test_WinCE.vcp: - * tests/Object_Manager_Test_WinCE.vcp: - * tests/OrdMultiSet_Test_WinCE.vcp: - * tests/OS_Test_WinCE.vcp: - * tests/Priority_Buffer_Test_WinCE.vcp: - * tests/Priority_Reactor_Test_WinCE.vcp: - * tests/Priority_Task_Test_WinCE.vcp: - * tests/Process_Manager_Test_WinCE.vcp: - * tests/RB_Tree_Test_WinCE.vcp: - * tests/Reactors_Test_WinCE.vcp: - * tests/Reactor_Exceptions_Test_WinCE.vcp: - * tests/Reactor_Notify_Test_WinCE.vcp: - * tests/Reactor_Performance_Test_WinCE.vcp: - * tests/Reactor_Timer_Test_WinCE.vcp: - * tests/Reader_Writer_Test_WinCE.vcp: - * tests/Recursive_Mutex_Test_WinCE.vcp: - * tests/Refcounted_Auto_Ptr_Test_WinCE.vcp: - * tests/Reverse_Lock_Test_WinCE.vcp: - * tests/Semaphore_Test_WinCE.vcp: - * tests/Service_Config_DLL_WinCE.vcp: - * tests/Service_Config_Test_WinCE.vcp: - * tests/Sigset_Ops_Test_WinCE.vcp: - * tests/Simple_Message_Block_Test_WinCE.vcp: - * tests/SOCK_Connector_Test_WinCE.vcp: - * tests/SOCK_Send_Recv_Test_WinCE.vcp: - * tests/SOCK_Test_WinCE.vcp: - * tests/SPIPE_Test_WinCE.vcp: - * tests/SString_Test_WinCE.vcp: - * tests/Svc_Handler_Test_WinCE.vcp: - * tests/Task_Test_WinCE.vcp: - * tests/Thread_Manager_Test_WinCE.vcp: - * tests/Thread_Mutex_Test_WinCE.vcp: - * tests/Thread_Pool_Reactor_Resume_Test_WinCE.vcp: - * tests/Thread_Pool_Reactor_Test_WinCE.vcp: - * tests/Thread_Pool_Test_WinCE.vcp: - * tests/Timeprobe_Test_WinCE.vcp: - * tests/Timer_Queue_Test_WinCE.vcp: - * tests/Time_Service_Test_WinCE.vcp: - * tests/Time_Value_Test_WinCE.vcp: - * tests/Tokens_Test_WinCE.vcp: - * tests/TSS_Test_WinCE.vcp: - * tests/Upgradable_RW_Test_WinCE.vcp: eMbedded Visual C++ - workspace/projects deleted. These need to be generated from - tests.mpc when needed. The make_release script now generates - them for each source kit. When needed on a CVS workspace build, - use mwc/mpc. The autobuild system has commands - generate_workspace and generate_makefile that accomplish this. - See autobuild configs in the riverace directory for examples. - - * tests/tests.mpc: Added New_Fail_Test. - - * ace/config-WinCE.h: Add #define ACE_LACKS_ACE_IOSTREAM. If it has - ACE_LACKS_IOSTREAM_TOTALLY, it can't possibly do ACE_IOStream<>. - - * ace/ATM_Acceptor.{h cpp}: - * ace/ATM_Addr.{h cpp}: - * ace/ATM_Params.h: - * ace/ATM_QoS.h: - * ace/Dev_Poll_Reactor.cpp: - * ace/FlReactor.cpp: - * ace/POSIX_Proactor.cpp: - * ace/QtReactor.h: - * ace/SUN_Proactor.h: - * ace/TkReactor.h: Add /**/ between #include and filename to silence - MS warnings about missing files during dependency updates. - - * ace/OS.i: Fixed typo for ACE_HAS_WINCE. - - * ace/OS.{h cpp}: Enabled ACE_OS::mktime() for ACE_HAS_WINCE. To do - this, define struct tm similarly to the way it is in the standard - C library. When called, converts the struct tm to a SYSTEMTIME, - then to a FILETIME, then inserts that into an ACE_Time_Value and - pulls out the sec() value. - In OS.h, add the ACE_HAS_WINCE section for the ace_timezone() - function. - - * ace/Filecache.cpp (ACE_Filecache_Object::update): For ACE_HAS_WINCE, - return status to indicate every file has changed since cached. - There's probably a way to do this correctly on WinCE, but there's - no time to figure it out right now. - - * ace/Logging_Strategy.cpp: In calls to ACE_OS::fopen(), added - ACE_LIB_TEXT around naked char strings to build clean with Unicode. - - * ace/NT_Service.{h cpp}: Add ACE_HAS_WINCE to the platforms that - don't attempt to build this class. - - * ace/SPIPE_Connector.cpp (connect): Pretty much disabled this for CE. - - * bin/MakeProjectCreator/templates/em3vcp.mpd: When generating - CPU-based /D options, just do /D, not /D<name>=1. If =1 is need - for something, add it to the .mpt file. - - * bin/MakeProjectCreator/templates/em3vcpdll.mpt: - * bin/MakeProjectCreator/templates/em3vcpdllexe.mpt: - * bin/MakeProjectCreator/templates/em3vcplibexe.mpt: Remove /GX, /Gm, - /GR, /Gy, and /Md[d] options - embedded VC++ doesn't do these or the - previous near-working .vcp files didn't use them and no value is - seen in enabling something not previously used. - Insured that the options are consistent, for example, all Release - builds use /Ob2 (high level of inlining). - * bin/MakeProjectCreator/templates/em3vcplib.mpt: In addition to the - above removals, remove /G5 (optimize for Pentium) from Debug build. - -Sat Dec 21 12:38:41 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.cpp (snprintf): Win32 doesn't 0-terminate the string if - the maxlen is exceed, so terminate the string in that case. - -Sat Dec 21 14:36:43 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * include/makeinclude/platform_cygwin32.GNU: - Enabled threads and removed the setting that only static libs are - build. Also set SONAME and SOVERSION to nothing because the - Cygwin linker can't handle a dll name where the version is added - after .dll. - -Fri Dec 20 20:02:41 2002 Steve Huston <shuston@riverace.com> - - * bin/make_release: When updating version files, also update - bin/MakeProjectCreator/config/aceversion.mpb and taoversion.mpb. - Added a new subroutine, generate_makefiles() that's called just - before create_kit. It generates the workspace and project files - for ACE_wrappers/ace and ACE_wrappers/tests. If any of these are - desired for TAO, someone knowledgeable on TAO should add to the - place in generate_makefiles() for the TAO stuff. - -Fri Dec 20 16:12:33 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/ACE.h: - The math.h delived with the Cygwin32 compiler already defines - log2 as macro so added a define acelog2 as log2 to handle this - already defined macro. - -Thu Dec 19 18:47:29 2002 Steve Huston <shuston@riverace.com> - - * tests/tests.mpc: Added Obstack_Test and Recursive_Condition_Test. - -Thu Dec 19 16:47:17 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/SOCK_Dgram.h: - ace/SOCK_CODgram.h: Added metatype information for - ACE_SOCK_Dgram and ACE_SOCK_CODgram. Thanks to Sankaranarayanan - K V <sankar@migv.mot.com> for reporting this. - -Thu Dec 19 10:02:56 2002 Rich Seibel <seibel_r@ociweb.com> - - * tests/MM_Shared_Memory_Test.cpp - - removed extraneous line that made the test appear to fail - -Thu Dec 19 06:47:07 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/config/aceversion.mpb: - * bin/MakeProjectCreator/config/taoversion.mpb: - - Updated the version to reflect the current release version. - - * bin/MakeProjectCreator/modules/Driver.pm: - - Moved the Cwd::getcwd() outside the loop to avoid calling it - multiple times. - - * bin/MakeProjectCreator/templates/vc7.mpd: - - Modified the ProgramDatabaseFile to use the intermediate_dir as - the output location. Thanks to Steve Huston - <shuston@riverace.com> for reporting it. - -Wed Dec 18 20:52:41 2002 Steve Huston <shuston@riverace.com> - - * ace/SOCK_Dgram_Mcast.cpp (open_i): Call shared_open using the - protocol family from the specified address, not PF_INET. Allows - working with IPv6. - - * tests/Multicast_Test.cpp: If join fails, log the error reason. - -Wed Dec 18 19:57:51 2002 Steve Huston <shuston@riverace.com> - - * ace/config-sunos5.7.h: Added ACE_HAS_SNPRINTF. I have verified - that it's available from Solaris 7 on. If someone can determine - it's available earlier, please move the definition to one of the - earlier-version config-sunos5.*.h files. - -Wed Dec 18 18:59:41 2002 Steve Huston <shuston@riverace.com> - - * ace/ace.mpc: - * tests/tests.mpc: Back out the below changes... jumped the gun. - -Wed Dec 18 18:55:31 2002 Steve Huston <shuston@riverace.com> - - * ace/ace.mpc: Added new Hash_Map_Manager_RT[_T].cpp files. - - * tests/tests.mpc: Added new Hash_Map_Manager_RT_Test.cpp project. - -Wed Dec 18 15:03:03 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/Get_Opt.cpp (long_option): For MSVC 7.x compiler, we need to - make sure that the argument passed to "isalnum" is legal, i.e., - between 0 and 0xff, otherwise, it will crash the process. Added - a special case for MSVC 7.x and above to prevent this. - -Wed Dec 18 13:36:49 2002 Steve Huston <shuston@riverace.com> - - * ace/ace.mwc: Workspace definition file for ACE and it's immediate - subordinates, RMCast. - - * tests/tests.mwc: Workspace definition for tests and tests/RMCast. - - * bin/MakeProjectCreator/modules/Driver.pm (run): When processing - specified input files, chdir to the directory containing the file - so the files are read and written properly when referred to using - relative paths. - -Wed Dec 18 11:38:39 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/config-win32-msvc-7.h: Added pragma prefix to disable - warning of using Microsoft extension. - - * ace/Timeprobe.h: Removed MSVC specific disabling and re-enabling - of warning of using Microsoft entension. We already do this - through all the MSVC config files and there's no point of doing - this again here. - -Mon Dec 16 16:34:23 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/OS.i: Applied the following patch from Eugene Alterman - <EugeneA@bremer-inc.com>. - - It appears that Win32 mmap() emulation implements MAP_FIXED - semantic under the wrong assumption about MapViewOfFileEx() - behavior. In fact if a non-null base address is supplied to - MapViewOfFileEx() it behaves like mmap() with MAP_FIXED set - it - maps at that address or fails if it unable to do it. If a null - base address is supplied it behaves like mmap() with MAP_FIXED - not set. - - This means that if MAP_FIXED is not set the emulation should set - the base address argument to 0, and it should fail if MAP_FIXED - is set and <addr> is 0. - - Besides, it makes sense for WinCE emulation to fail whenever - MAP_FIXED is set since WinCE does not support MAP_FIXED - semantic. - -Wed Dec 18 08:29:04 2002 Rich Seibel <seibel_r@ociweb.com> - - * ace/Synch.cpp: Added code to generate a unique name if a - platform lacks unnamed semaphores (MACOSX 10.2). - - * ace/Sock_Connect.cpp: Added a conditional include for - ifaddrs.h (MACOSX 10.2). - - * tests/Priority_Task_Test.cpp: Changed test to notify - if platform does not implement getprio instead of fail. - (MACOSX 10.2) - - * tests/MM_Shared_Memory_Test.cpp: Changed compile time test to - pick up the right implementation on platform that have - POSIX semaphores but not SYSV semaphores (MACOSX 10.2). - - * ace/Lib_Find.cpp - ace/config-macosx.h: - - Gave the builder a choice of where to get the dlcompat package. - - * ace/OS.i - ace/config-macosx.h: - - Added a switch to eliminate an error message since the dlcompat - provided APPLE dynamic linkage does not support unloading thus - complains on the use of dlclose. - -Wed Dec 18 08:29:04 2002 Rich Seibel <seibel_r@ociweb.com> - - * ace/Timer_Wheel_T.cpp: - - Changed types from int to long to match the return - for timer ids. - -Wed Dec 18 05:42:52 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * ace/High_Res_Timer.cpp (global_scale_factor): The INTEL compiler - can also set the global scale factor. Thanks to Chris Kohlhoff - for reporting this. - -Tue Dec 17 11:22:31 2002 Phil Mesnier <mesnier_p@ociweb.com> - - * ace/Codeset_IBM1047.h: - * ace/Codeset_IBM1047.cpp: - - Added implementations of ncs and tcs methods to return the proper - values as defined by the OSF's code and character set registry. - - * ace/CDR_Stream.cpp: - * ace/CDR_Stream.h: - * ace/CDR_Stream.i: - - Stripped out the reference counting on translator objects to avoid - build problems. - -Tue Dec 17 11:30:31 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ACE-INSTALL.html: Fixed fuzz errors. - -Tue Dec 17 08:20:22 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/NMakeWorkspaceCreator.pm: - - Fixed a problem with generated Makefiles where CFG is defaulted. - - * bin/MakeProjectCreator/templates/nmake.mpd: - - Modified the dependencies file name. - -Tue Dec 17 07:03:40 2002 Chad Elliott <elliott_c@ociweb.com> - - * ace/Array_Base.h: - * ace/Atomic_Op_T.cpp: - * ace/Message_Queue.cpp: - * ace/OS.h: - - Added #include's for various files for the Green Hills compiler - for VxWorks. - - * ace/OS.cpp: - - Added a call to ::inet_aton() with a non-const char* for VxWorks. - - * ace/Thread_Adapter.cpp: - - Changed the pointer type from ACE_THR_FUNC_INTERNAL to - ACE_THR_FUNC. - - * ace/config-vxworks5.x.h: - - Added a #define for ACE_LACKS_CLEARERR for the Green Hills - compiler. - -Mon Dec 16 12:28:05 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ACE-INSTALL.html: Fixed some broken links. Thanks to - Jonathan Wackley <jwackley@legato.com> for reporting this. - -Sat Dec 14 13:06:09 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * docs/ACE-categories.html: Removed all mention of the - ACE_Multiplexor class. Thanks to Marc Tardif <marc@sitepak.com> - for reporting this. - -Tue Dec 17 07:37:40 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * tests/README: Fixed typos. - -Mon Dec 16 23:31:25 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * THANKS: Added Guan Joe to the hall of fame. - -Mon Dec 16 18:11:03 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/CDR_Stream.h: Moved the destructor of - ACE_Char_Codeset_Translator to be public. - -Mon Dec 16 13:14:34 2002 Phil Mesnier <mesnier_p@ociweb.com> - - * ace/CDR_Stream.cpp: - * ace/CDR_Stream.h: - * ace/CDR_Stream.i: - Preparing for codeset negotiation. The ACE input/output CDR - streams' support for wide character IO is enhanced by adding a - flag to allow or disallow wchar i/o regardless of whether a - translator is present. There are times when wchar is not - allowed, such as GIOP 1.0, or when a native codeset for wchar - was not specified. In CORBA, a native wchar codeset must be - specified, it may not be defaulted. Within the ACE CDR streams, - wchar i/o is allowed by default, to ensure backwards - compatibility. - - I fixed a few of the reading and writing methods to ensure that - the good_bit_ flag is cleared if an error occurs. This was not - happening in all cases, leading to errors elsewhere. - - I have extended the definition of the - ACE_[W]Char_Codeset_Translator classes to make them reference - counted objects. There are some codesets that support shifting - and may be stateful. By using reference counting, the CDR - streams don't have to worry about ownership of stateful - translators. - - Finally, the translators now have pure virtual methods that - return the native and translated codeset ID values. These values - refer to entries in the DCE Code and Character Set Registry, - which is the reference for CORBA codeset ids. Anyone - implementing a char or wchar codeset will have to supply a - numeric value for the native codeset (ncs) and the translated - codeset (tcs). NCS refers to the codeset of the text on the - application side of the stream, and TCS refers to the codset of - the text on the "wire" side of the stream. - - These changes are not everything that is required to support - codeset negotiation, but it is all the changes required to the - CDR streams. - -Sat Dec 14 08:56:23 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * ace/Timer_Queue_T.cpp (timeout): Commented out the change below - until we figure out how to break the dependencies on the - Reactor. Thanks to Bala for reporting this. - -Fri Dec 13 16:05:05 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/Timer_Queue_T.cpp (timeout): Optimize Steve's patch below - for the case where the upcall is coming from a reactor. Thanks - to Irfan for pointing this out. - -Fri Dec 13 21:12:29 2002 Steve Huston <shuston@riverace.com> - - * ace/test_config.h: Add __hpux to the conditions needed to instantiate - ACE_Singleton<ACE_Test_Output, ACE_Null_Mutex>::singleton_. Also see - Sat Nov 23 18:12:37 2002 Steve Huston <shuston@riverace.com> - -Fri Dec 13 18:25:42 2002 Steve Huston <shuston@riverace.com> - - * ace/config-aix-4.x.h: Set ACE_HAS_THREADS to 1 or 0 based on the - compiler's predefined macros. This lets users build easier without - having to worry about what ACE expects - they just have to invoke - the compiler correctly. Add #include "ace/config-posix.h" to pick - the detectable POSIX-ish settings. - - * ace/config-posix.h: If ACE_HAS_THREADS is defined as 0 on entry, - undef it and don't pick up the POSIX threads settings. Allows - single-threaded builds using this very handy config. - -Fri Dec 13 18:10:31 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.cpp (wchar snprintf, sprintf): Change the selection for - ability to use vswprintf from ACE_HAS_XPG4_MULTIBYTE_CHAR to - _XOPEN_SOURCE >= 500. vswprintf came in at XPG5 (and C99). - - * ace/config-hpux-11.00.h: Enable ACE_HAS_SNPRINTF if - _INCLUDE_HPUX_SOURCE is set after including stdsyms.h. - - * ace/config-posix.h: Set ACE_HAS_SNPRINTF if _XOPEN_SOURCE >= 500. - -Thu Dec 12 20:11:39 2002 Steve Huston <shuston@riverace.com> - - * ace/Timer_Heap_T.cpp (cancel): Release the queue lock before - upcalling to handle_close. Fixed Bugzilla #1387. - -Thu Dec 12 12:50:34 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/OS.cpp: Changed a cast of unique_name to &unique_name to - fix a bug on pSoS. Thanks to Dieter Knueppel - <dknueppel@datus.com> for reporting this. - -Thu Dec 12 12:55:41 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.h: Add missing snprintf() method declaration for ACE_HAS_WCHAR - -Thu Dec 12 08:49:31 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/Log_Msg.cpp: Fixed a warning in LynxOS builds. - -Thu Dec 12 06:18:02 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/Parser.pm: - - Fixed a bug with setting and retrieving the current working - directory. Thanks to Steve Huston <shuston@riverace.com> for - reporting this. - -Thu Dec 12 00:01:37 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * netsvcs/clients/Naming/Client/Makefile: Fixed a small problem in - static builds. Hopefully they should be fine in the next round - of builds. - -Wed Dec 11 23:11:09 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/clean_sems.sh: A shell script to remove - semaphores. Integrating the script that Ossama used to the - daily build system. - -Wed Dec 11 21:05:49 2002 Steve Huston <shuston@riverace.com> - - * ace/Message_Block_T.cpp: - * examples/ASX/Message_Queue/priority_buffer.cpp: - * examples/Misc/test_read_buffer.cpp: - * examples/Threads/TSS_Obj.h: Added missing - #include "ace/Malloc_Base.h" to build clean on AIX, Visual Age C++ 6. - - * examples/C++NPv1/Logging_Client.cpp: - * examples/C++NPv2/Reactor_Logging_Server.cpp: - * examples/C++NPv2/Service_Reporter.cpp: - * examples/C++NPv2/SLD.cpp: - * examples/C++NPv2/SLDex.cpp: Add missing #include - "ace/Log_Msg.h" to pick up ACE_ERROR_RETURN for AIX Visual Age C++ 6. - - * examples/C++NPv2/Makefile.SLD: Remove the tempinc-bashing - redefinition of COMPILE.cc... that only works if there's a single - source file for each program. - -Wed Dec 11 19:33:19 2002 Steve Huston <shuston@riverace.com> - - * ace/config-hpux-10.x-hpc++.h: - * ace/config-hpux-10.x.h: Picked up these changes from the 11.x config: - - Sat Nov 23 17:54:18 2002 Steve Huston <shuston@riverace.com> - - * ace/config-hpux-11.00.h: Control the ACE_HAS_STANDARD_CPP_LIBRARY - and ACE_USES_STD_NAMESPACE_FOR_STDCPP_LIB settings by compiler- - defined macro _HP_NAMESPACE_STD, not passing them through from - platform_hpux_aCC.GNU. Similarly with ACE_HAS_EXCEPTIONS (define - it unless __HPACC_NOEH is set). This breaks a coupling from the - platform_hpux_aCC.GNU file, and will make autoconf easier as well - as making it easier for users to write makefiles without following - the ties through and trying to duplicate them. - Also, include config-posix.h to let it pick up system-supplied - settings. - - * ace/Time_Value.h: For HP-UX 10, need to typedef timespec_t because - the OS doesn't define it. - -Wed Dec 11 17:05:28 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * examples/Reactor/WFMO_Reactor/Talker.cpp: Fixed a warning in - Borland builds. - -Wed Dec 11 17:00:21 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/WIN32_Asynch_IO.cpp: Fixed a warning in Borland builds. - -Tue Dec 10 19:42:29 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.{h cpp}: Added new method, ACE_OS::snprintf() that wraps - native snprintf() if available (ACE_HAS_SNPRINTF for char, - ACE_HAS_XPG4_MULTIBYTE or ACE_WIN32 for wide char); else returns - -1, ENOTSUP. - - * ace/config-linux-common.h: Add detection for ACE_HAS_SNPRINTF. - - * ace/config-win32-common.h: Removed #define ACE_HAS_VSWPRINTF. - vswprintf() on Windows is not the standard way, and this setting - was only used for Win32. See comments in OS.cpp, sprintf(wchar_t...) - for more info. Added ACE_HAS_SNPRINTF in its place. - - * ace/Log_Msg.cpp (log): Don't overrun the msg_ buffer while - building the logged message. Silently stops logging characters - when the limit is reached. - Also, on %a aborts, don't just do a "while (1);" on WinCE. - Forcing the user to cycle power is not a good idea. - - * tests/Log_Msg_Test.cpp: Add a test case for logging a string - that can't possibly fit in the logging buffers. - -Tue Dec 10 18:10:41 2002 Steve Huston <shuston@riverace.com> - - * tests/Proactor_Test.cpp: Changed the default proactor type to - DEFAULT. If a specific proactor implementation type is desired, that - must be specified now. This matches ACE's behavior that users will - get unless they specifically change it. Also, changed loglevel to - be 0 (only highlights and errors), 1 (more progress information), - and 2 (all operation parameters and result information). The default - is 0, allowing for increased load on this test without creating - gigantic log files. All platforms now use the same test parameters - (clients, byte counts, threads, etc.) except that if the SUN - proactor is selected, the test will run with one thread regardless - of user selection. The ACE_SUN_Proactor's completion detector - (aiowait) is not multithread safe. - -Tue Dec 10 17:19:06 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * tests/OS_Test.cpp (ACE_TMAIN): Integrated the latest tests from - Alain for string_strsncpy_test (). The test is more robust and - tests things better. It also has comments on what it is testing - :-). - -Tue Dec 10 16:53:29 2002 Steve Huston <shuston@riverace.com> - - * tests/OS_Test.cpp: Add missing '=' signs to fix compile error on - wide-char builds. - -Tue Dec 10 16:05:37 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * Makefile: Added a target 'reverseclean' - * bin/reverse_clean: Added a small perl script that does a - realclean in the reverse order on the directories passed to - it. - - The motivation for this is that some of our builds dont clean up - properly due to the dependencies. The reason for this is that - ace/ACE_COMPONENTS.list gets wiped out first, preventing things - that examine the components list (like netsvcs) from getting - cleaned. Hence we need to do a 'realclean' in reverse order of - the builds. - -Tue Dec 10 13:23:37 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * bin/PerlACE/Process.pm: - * bin/PerlACE/Process_Win32.pm: - * bin/PerlACE/Process_Unix.pm: Changed WAIT_DELAY_FACTORY from a - data member in Process_Win32 to a global data in - PerlACE::Process. You can now define an environment variable - "ACE_RUNTEST_DELAY" on both Win32 and UNIX to multiply the wait - time with the number. The same trick that I added last time - into Process_Win32 to invoke tests with purify should also work - with valgrind. - - * bin/PerlACE/Run_Test.pm (waitforfile): Changed to multiply the - wait time with the PerlACE::Process::WAIT_DELAY_FACTOR . - -Tue Dec 10 10:20:35 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/Synch.cpp: Commented out the dump () call on ACE_cond_t, - which caused build errors from the checkin "Mon Dec 9 18:54:09 - 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu>". There is - no dump method on ACE_cond_t. - - Fixed the return type of the implementation to match the - declaration. - -Mon Dec 9 18:54:09 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/Synch.{h,cpp}: Added a dump() method to the template - specialization of ACE_Condition<ACE_Recursive_Thread_Mutex>. - Thanks to Mathias Waack <mathias.waack@schlund.de> for reporting - this. - - * ace/Get_Opt.{h,i} Added a new argc() accessor. Thanks to - Volker Boerchers <vboerchers@tecon.de> for contributing this. - -Mon Dec 9 17:58:31 2002 Steve Huston <shuston@riverace.com> - - * tests/Multicast_Test.cpp: Added missing template instantiation for - Sun Workshop 5 compiler. - -Mon Dec 9 17:27:49 2002 Steve Huston <shuston@riverace.com> - - * tests/Recursive_Condition_Test.cpp: In test_4, wait for the - waiter thread to wake up and exit itself before checking the - mutex recursion count. Else, the waiter may have the mutex while - the main thread is checking the count and detect a false fail. - -Mon Dec 9 17:02:41 2002 Steve Huston <shuston@riverace.com> - - * include/makeinclude/platform_aix_ibm.GNU: Change from using - -qtemplateregistry to use -qtempinc for Visual Age C++ 6 template - handling. -qtemplateregistry is a nice idea, but generates too - many errors from missing templates. Using -qtempinc still generates - gobs of multiple-instantiation warnings, but the build is correct - and the tests run. - - * bin/aix_shr: Add the error number patterns for VA 6's multiple - template instantiation warnings to the filter-out pattern. - -Mon Dec 9 15:42:39 UTC 2002 Don Hinton <dhinton@ieee.org> - - * examples/C++NPv1/Reactive_Logging_Server_Ex.cpp: Removed - duplicate template specialization for ACE_Hash<ACE_HANDLE> on - Windows that's no longer needed. - -Mon Dec 9 07:41:39 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * tests/OS_Test.cpp: - * ace/OS_String.cpp (strsncpy): Fixed a cut and paste error. - -Mon Dec 9 00:09:55 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/INET_Addr.cpp (get_host_addr): Changed the usage of strncpy - () to strsncpy (). Thanks to Alain Decamps for donating this patch. - -Sun Dec 8 23:56:09 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/OS_String.cpp: Applied patches from Alain Decamps to make - strsncpy () more robust. If dst == src, the return string will - have utmost 'maxlen'. - - * tests/OS_Test.cpp: Added a test case for strsncpy (). Thanks to - Alain Decamps for donating the patch. - -Sun Dec 08 22:44:07 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * ace/Functor.h (void *): - * ace/Functor.i: Added a template specialization for ACE_Hash <void *>. - -Sun Dec 08 09:45:34 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/Map_Manager.h: - * ace/SString.h: - * ace/Unbounded_Queue.h: - Improved doxygen documentation - -Sun Dec 08 09:37:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * etc/*.doxygen: - Set JAVADOC_AUTOBRIEF to NO in all doxygen config files. When - only javadoc style documentation is used, only detailed documentation - is generated. - -Sun Dec 08 01:17:26 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/Event_Handler.cpp: Fixed a typo. - -Thu Dec 5 20:30:56 2002 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * ace/Event_Handler.cpp: Updated the implementation of - ACE_Event_Handler::read_adapter() so that the code is consistent - with the stated semantics. Thanks to Ron Muck <rlm@sdiusa.com> - for the fix. - - * ace/OS_Dirent.inl (readdir_r): Enhanced the - ACE_OS_Directn::readdir_r() wrapper facade so that it returns 1 - (i.e., "done") when an error occurs. Thanks to Abhay Kulkarni - <Abhay.Kulkarni@veritas.com> for reporting this. - -Sat Dec 7 19:54:12 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.h: For WinCE, define S_IFDIR and S_IFREG file mode masks. - Also, added define for ENXIO if not supplied (like on WinCE). - -Sat Dec 07 14:29:47 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/ACE.cpp (format_hexdump): This change fixed the problem that - a character is missing when the last line has less than 16 bytes - or less to display. Thanks to Terry Lao <terry@ec-tone.com> for - submitting the patch, and Johnny Willemsen - <jwillemsen@remedy.nl> for reviewing the change. - -Fri Dec 6 23:58:47 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/CDR_Stream.cpp: - * ace/CDR_Stream.i: Set good_bit_ to zero whenever the sanity - check for remaining length in the stream fails. Thanks to Jerry - D. De Master <jdemaste@rite-solutions.com> for pointing out the - problem. - - * ace/CDR_Array_Test.cpp: Extended the test to include test for - checking the value of good_bit () by forcing an erroneous read. - -Fri Dec 6 16:26:49 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ace/FILE_Addr.cpp (set): Windows appends ASCII 'a' - 'z' and - 'A' - 'Z' to generate unique filenames and gives up if the - character in the template is a non-alphabetic character. In such - cases, it returns a NULL. Make sure that we supply a template - with a alphabetic character before XXXXXX as well as check for a - NULL return value. - -Thu Dec 5 22:19:41 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.{h i}: Added ACE_OS::sigwaitinfo() wrapper. - - * ace/POSIX_Proactor.{h cpp} (ACE_POSIX_SIG_Proactor): Clean up some - spelling and formatting in documentation comments. Clean up the - signal calls, and substitue ACE_OS calls for naked API calls. - Unless we have specific evidence to the contrary, trust what the - OS says with signal information. Otherwise, if we run through - all the aiocbs on every signal, we could get into a situation - where deferred I/O operations are never restarted. - - * ace/POSIX_CB_Proactor.h: Changed the comments to say this is a - callback-based proactor, not an SGI IRIX one. The callback - paradigm is part of newer POSIXes and available on more than - SGI... HP-UX, for instance. - -Thu Dec 5 20:33:18 2002 Steve Huston <shuston@riverace.com> - - * ace/config-aix-4.x.h: Removed ACE_NEEDS_DEV_IO_CONVERSION. This - seems like a ludicrous workaround to some old, buggy compiler. - -Thu Dec 5 14:35:50 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * netsvcs/clients/Naming/Client/Makefile: Do not build in static - builds. - -Thu Dec 5 09:45:25 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * examples/Synch/proc_sema.cpp: Fixed some typos. Thanks to - Andy King <andyking@vernon.com> for reporting this. - -Thu Dec 5 07:39:45 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: Removed Bug_1020_Regression from the - nightly regression test on Win32. - -Thu Dec 5 09:58:45 UTC 2002 Don Hinton <dhinton@ieee.org> - - * ace/config-openbsd-pthread.h: Added - ACE_LACKS_PERFECT_MULTICAST_FILTERING 1. This should - fix the errors in the test/Multicast_Test in the - nightly autobuilds. - -Wed Dec 4 21:53:28 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.h: Include <sys/pstat.h> for __hpux platform. - * ace/OS.cpp (num_processors_online): Try the syscall for this - on HP-UX if __hpux is defined, not hpux (only set for g++). - -Wed Dec 4 21:18:31 2002 Steve Huston <shuston@riverace.com> - - * ace/Proactor_Timer_Test.cpp: Fix incorrect use of pointer-integer - casting. Log pointers with %@, not %x. - -Wed Dec 4 21:11:41 2002 Steve Huston <shuston@riverace.com> - - * tests/Proactor_Test.cpp: Set all the default options and values - before parsing command line options to make it easy to predictably - vary things by hand. Also, rather than run for a determined number - of seconds, run a specific number of bytes around the loop. This - limits the size of the log, especially nice for fast processors. - Thanks to Johnny Willemsen <jwillemsen@remedy.nl> for this idea. - -Wed Dec 4 18:37:10 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: Added a Server_Connection_Purging to the - nightly regression tests. - -Wed Dec 4 17:47:49 2002 Steve Huston <shuston@riverace.com> - - * tests/Proactor_Test.cpp: On Sender read complete, start another - read if either in full duplex mode OR writing has been stopped. - We want to catch the Receiver closing the socket (read 0 bytes). - -Wed Dec 4 11:41:15 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: Added a Bug_1020_Regression to the - nightly regression test on Win32. - -Tue Dec 3 20:47:39 2002 Steve Huston <shuston@riverace.com> - - * ace/SSL/SSL_Context.cpp: Add explicit instantiation of - ACE_Singleton<ACE_SSL_Context, ACE_SYNCH_MUTEX>::singleton_. - * netsvcs/lib/Name_Handler.cpp: Add explicit instantiation of - ACE_Singleton<Naming_Context, ACE_SYNCH_NULL_MUTEX>::singleton_. - * netsvcs/lib/Server_Logging_Handler.cpp: Add explicit instantiation of - ACE_Server_Logging_Handler_T<LOGGING_PEER_STREAM, u_long, - ACE_NULL_SYNCH, Null_Synch_Static_Receiver>::request_count_; - * tests/Framework_Component_DLL.cpp: Add explicit instantiation of - ACE_DLL_Singleton_T<FWCT_DLL_Singleton_Adapter_T <Simple_Service>, - ACE_SYNCH_MUTEX>::singleton_. - * tests/MEM_Stream_Test.cpp: Add explicit instantiation of - ACE_Singleton<ACE_Atomic_Op<ACE_SYNCH_MUTEX, u_short>, - ACE_SYNCH_RECURSIVE_MUTEX>::singleton_. - * tests/Process_Strategy_Test.cpp: Add explicit instantiation of - ACE_Singleton<Options, ACE_Null_Mutex>::singleton_. - * tests/Timeprobe_Test.cpp: Add explicit instantiation of - ACE_Singleton<int, ACE_SYNCH_RECURSIVE_MUTEX>::singleton_. - - These are similar to those done in: - Sat Nov 23 18:12:37 2002 Steve Huston <shuston@riverace.com> - - * netsvcs/clients/Naming/Client/Makefile: - * netsvcs/clients/Naming/Dump_Restore/Makefile: Add ACELIB to LIBS - to include -lACE for building the shared libs. rules.lib.GNU - doesn't add VLDLIBS when linking a shared lib, but rules.bin.GNU - adds it for binaries. This should probably be changed, but it's a - quagmire to wade into at this point, and autoconf is coming. - This fix is needed now to get this stuff building with g++ on AIX. - -Mon Dec 2 20:09:29 2002 Steve Huston <shuston@riverace.com> - - * examples/C++NPv1/Reactive_Logging_Server_Ex.cpp: - * ACEXML/common/Mem_Map_Stream.cpp: - * ACEXML/common/URL_Addr.cpp: - * ACEXML/common/XML_Codecs.cpp: Added missing explicit template - instantiations to build clean on Sun Workshop 5. - -Mon Dec 2 19:41:27 2002 Steve Huston <shuston@riverace.com> - - * tests/Proactor_Test.cpp: Rather than close connections outright - when beginning to shut down test, just shut down the write side. - This is an effort to make the receiving closes cleaner. - -Mon Dec 2 17:58:06 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/ACE.cpp: Fixed a compile error with g++. - -Sat Nov 30 09:18:51 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/ACE.{h,cpp}: Added a new crc32() method that works on - an iovec. Thanks to Tufan Oruk <toruk@usa.net> for contributing - this. - - * THANKS: Just added our 1,600 contributor to ACE+TAO! - -Mon Dec 02 18:59:33 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/DLL.h: - * ace/Dirent.h: - * ace/Future_Set.h: - * ace/UPIPE_Connector.h: - * ace/UPIPE_Acceptor.h: - * ace/UNIX_Addr.h: - * ace/Configuration_Import_Export.h: - Improved doxygen documentation. - -Mon Dec 02 18:47:32 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * THANKS: Added Tim Smith <timsmith_s@hotmail.com> - -Sun Dec 01 10:16:45 2002 Balachandran <bala@cs.wustl.edu> - - * ACE version 5.2.7 released. - -Sat Nov 30 22:46:00 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * THANKS: Added Christophe Vedel to the hall of fame. - -Fri Nov 29 12:23:14 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/Thread_Manager.h (ACE_Thread_Manager): Changed all the - accessors to the Thread_Descriptor to be protected. - - * ace/Thread_Manager.i: Removed assertion test for null - Thread_Descriptor pointer. This can happen when calling - thread_desc_self() from a thread not spawned by the - Thread_Manager, e.g., the main thread. - - Added checks in all ACE_Thread_Manager::at_exit for null - thread_desc_self to avoid crash. - - Thanks to Mathias Waack <mathias.waack@schlund.de> for reporting - the bug. - -Thu Nov 28 06:36:04 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * apps/soreduce/Makefile: Changed INSBIN from $ACE_ROOT/bin/soreduce/ - to $ACE_ROOT/bin/. Thanks to Jaroslaw Nozderko - <jaroslaw.nozderko@polkomtel.com.pl> for reporting this. - -Thu Nov 28 07:54:28 2002 Craig Rodrigues <crodrigu@bbn.com> - - * docs/tutorials/022/Acceptor_Service.h: Add include of svc_export.h - to fix compilation errors. Thanks to Viktor Ransmayr - <viktor.ransmayr@t-online.de> for reporting the problem. Not - sure whether this example would still compile fine. - -Thu Nov 28 09:04:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/FILE.h: - * ace/Default_Constants.h: - Improved usage of doxygen commands. - -Thu Nov 28 07:23:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/Typed_SV_Message_Queue.h: - * ace/Strategies_T.h: - * ace/Framework_Component.h: - Improved doxygen documentation. - - * ace/TLI_Connector.i: - Made method definition and implementation the same. - -Wed Nov 27 22:38:15 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/Connector.h: - * ace/Connector.cpp: Fix for BUG 1361. Please see - http://deuce.doc.wustl.edu/bugzilla/show_bug.cgi?id=1361 for - more details on the problem. - - The fix is based on the assumption that ACE_Connector need not - be resumed by the TP_Reactor when used for asynch connects. - The supporting reason for this assumption is that during - connection completion or connection closure, the handler - ie. the ACE_Connector in this case is removed from the - Reactor. Hence this fix, which essentially implements the - virtual function resume_handler () and prevents itself from - being resumed. Thaks to Irfan for reviewing the fix. - -Wed Nov 27 20:35:49 2002 Steve Huston <shuston@riverace.com> - - * include/makeinclude/platform_aix_ibm.GNU: Add Visual Age C++ 6 to - the list of compiler versions it knows about. For VA 6, use the - new -qtemplateregistry build option to try to be more efficient - about template compilation. It's still lamer than most though, and - Visual Age C++ 6 needs the aixshr filter to weed out the Duplicate - Symbol warnings. - - * include/makeinclude/rules.local.GNU: Add templateregistry to the - files/directories removed on clean. It's used by Visual Age C++ 6. - -Wed Nov 27 19:42:51 2002 Steve Huston <shuston@riverace.com> - - * tests/MT_Reactor_Upcall_Test.cpp: Send and receive each message - field specifically; don't rely on the type and size fields being - adjacent in memory. On 64-bit Solaris w/ Forte 7 and on 64-bit - Windows XP, they aren't. - -Wed Nov 27 15:47:41 2002 Steve Huston <shuston@riverace.com> - - * ace/config-aix5.1.h: Add #define ACE_HAS_AIO_CALLS, but commented - out. AIX did things differently enough here to be a real problem. - It should be possible to get AIO working on AIX 5, but it will - take some real work. - - * ace/POSIX_Proactor.cpp: Add #include "ace/ACE.h" - - * ace/Sock_Connect.cpp: Visual Age C++ 6 still needs to hack to - get the templates built correctly :-( Changed the compile-time check - to an alleged Visual Age C++ 7 to see if they get it right then. - - * ACEXML/common/Env.i: Add missing #include "ace/Log_Msg.h" to pick - up ACE_ASSERT. - - * ACEXML/common/FileCharStream.cpp: - * ACEXML/common/StrCharStream.cpp: Add missing #include "ace/Log_Msg.h" - - * ACEXML/common/NamespaceSupport.h: - * ACEXML/parser/parser/Entity_Manager.h: Add missing #include - "ace/Synch.h" - -Wed Nov 27 07:23:30 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: Two new tests cannot run in single - threaded builds. - -Wed Nov 27 09:30:33 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/DLL.h: - * ace/Singleton.h: - Improved doxygen documentation. - -Tue Nov 26 20:04:38 2002 Steve Huston <shuston@riverace.com> - - * tests/Proactor_Test.cpp: Added checks for condition where a read or - write fails to initiate because the socket was closed. For Windows, - this is a ERROR_NETNAME_DELETED error on the ReadFile/WriteFile call - under ACE_WIN32_Asynch_Read/Write_Stream. The test now checks for - this and doesn't report an error, but this is a departure from - what one would expect from a socket. It might be better if the - underlying call was changed to WSARecv/WSASend for Winsock2-enabled - systems. - -Tue Nov 26 16:08:22 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: Added two new tests to our daily builds. - -Tue Nov 26 15:10:57 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ACE-INSTALL.html: Updated the file to have the right GNU make - version. Our make's dont work well with 3.78 and lower. - -Mon Nov 25 23:49:38 2002 Craig Rodrigues <crodrigu@bbn.com> - - * ace/config-macosx.h: Updates for pthread configuration, provided - by John Zorko <j.zorko@att.net> and Rich Seibel - <seibel_r@ociweb.com>. - - * ace/OS.i: - * ace/Sock_Connect.cpp: Remove unnecessary __MACOSX__ macros. - See: Thu Oct 17 19:18:02 UTC 2002 Craig Rodrigues - <crodrigu@bbn.com> - -Mon Nov 25 18:20:41 2002 Steve Huston <shuston@riverace.com> - - * ace/config-hpux-11.00.h: Added #define ACE_SIZEOF_WCHAR _WCHAR_T - to get ACEXML to compile. ACEXML's usage of ACE_SIZEOF_WCHAR breaks - the previous allowance for ACE_SIZEOF_WCHAR to be sizeof (wchar_t) - because it uses ACE_SIZEOF_WCHAR in a math expression. - - * examples/C++NPv2/Reactor_Logging_Server_Adapter.cpp: Added - #include "ace/INET_Addr.h" to build clean on HP-UX. - -Mon Nov 25 20:32:02 UTC 2002 Don Hinton <dhinton@ieee.org> - - * ace/README: Added ACE_LACKS_PERFECT_MULTICAST_FILTERING. - -Mon Nov 25 18:58:33 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/Local_Name_Space_T.cpp: - * ace/Process.h: - * ace/Local_Name_Space_T.h: - * ace/ATM_Params.h: - * ace/Timer_Queue_Adapters.h: - Improved documentation for doxygen generation. - - * ace/Svc_Handler.cpp: - Made method definition and implementation arguments match. - -Sat Nov 23 20:01:36 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * apps/*/Makefile: Updated dependecies. Thanks to Erwin Rol for - pointing out the problem. - -Sat Nov 23 18:12:37 2002 Steve Huston <shuston@riverace.com> - - * ace/SString.cpp: Add explicit instantiation of - ACE_String_Base<char>::NULL_String_ and - ACE_String_Base<ACE_WSTRING_TYPE>::NULL_String. - * ace/Based_Pointer_Repository.cpp: Add explicit instantiation of - ACE_Singleton<ACE_Based_Pointer_Repository,ACE_SYNCH_RW_MUTEX>:: - singleton_. - * ace/Dynamic.cpp: Add explicit instantiation of - ACE_TSS_Singleton<ACE_Dynamic, ACE_Null_Mutex>::singleton_. - * tests/test_config.h: Added explicit instantiation of - ACE_Singleton<ACE_Test_Output, ACE_Null_Mutex>::singleton_. - - The above are only done when using automatic template instantiation - on AIX with g++. Without this, the compiler and linker don't get - things done right and static data members in templates don't get - instantiated correctly. Thanks to Max F. Bilyk <mbilyk@europe.com> - for reporting this problem. - - * ace/SSL/Makefile: Set LIBS = $(ACELIB). Else it won't link on - AIX, g++. - - * include/makeinclude/platform_aix4_g++.GNU: Removed. Use - platform_aix_g++.GNU instead. They were similar, though not equal, - and confusing each other. - - * include/makeinclude/platform_aix_g++.GNU: Fixed to work correctly - and is now the one to use for AIX 4 and 5 with g++. - -Sat Nov 23 18:03:41 2002 Steve Huston <shuston@riverace.com> - - * ace/String_Base.i (compare): Fix a size_t/int mismatch. - -Sat Nov 23 17:54:18 2002 Steve Huston <shuston@riverace.com> - - * ace/POSIX_CB_Proactor.{h cpp}: Changes to align with recent - ACE_POSIX_AIOCB_Proactor changes: changed handle_events (ulong) to - handle_events_i() and made nonvirtual. Removed get_result_status(), - cancel_aiocb(), and cancel_aio() since they simply forwarded to - ACE_POSIX_AIOCB_Proactor. Remove most of allocate_aio_slot(); call - up to ACE_POSIX_AIOCB_Proactor::allocate_aio_slot then tweak the - aiocb fields for the callback. - - * ace/config-hpux-11.00.h: Control the ACE_HAS_STANDARD_CPP_LIBRARY - and ACE_USES_STD_NAMESPACE_FOR_STDCPP_LIB settings by compiler- - defined macro _HP_NAMESPACE_STD, not passing them through from - platform_hpux_aCC.GNU. Similarly with ACE_HAS_EXCEPTIONS (define - it unless __HPACC_NOEH is set). This breaks a coupling from the - platform_hpux_aCC.GNU file, and will make autoconf easier as well - as making it easier for users to write makefiles without following - the ties through and trying to duplicate them. - Also, include config-posix.h to let it pick up system-supplied - settings. - - * ace/OS_Memory.h: In the HP aCC section, remove the checks for - aCC version > 32500 and _HP_NAMESPACE_STD... just use the - ACE_HAS_STANDARD_CPP_LIBRARY setting from config-hpux-11.00.h. - - * ace/Proactor.cpp (ctor): When no implementation is supplied, be - sure to make one up, even w/o presence of realtime signals. - - * include/makeinclude/platform_hpux_aCC.GNU: If stdcpplib is not - specified by user, set it to 1 always, unless the compiler is - a .2x version and not .27 (03.27 (01.27 on 10.20) is the first - version that can do -AA); makes it unnecessary to update this every - time HP releases a new compiler version. Also rearranged the - thread-related compiler option settings to be simpler and easier - to match with what the documentation says to set. Removed the - ACE_HAS_STANDARD_CPP_LIBRARY settings; they're defined in - ace/config-hpux-11.00.h when proper (see above). - Also, added +W930 to CCLFAGS when -AA to stop a spurious warning - about placement operator delete on aC++ 03.37, 03.39. - -Sat Nov 23 17:43:29 2002 Steve Huston <shuston@riverace.com> - - * ace/ACE.cpp (send_n): Properly update byte count for each message - block in the chain to send. Thanks to Venkita Subramonian - <venkita@cse.wustl.edu> for reporting this and sending a fix. - -Sat Nov 23 16:55:39 2002 Steve Huston <shuston@riverace.com> - - * tests/MEM_Stream_Test.cpp: - * tests/Notify_Performance_Test.cpp: - * tests/Priority_Reactor_Test.cpp: - * tests/Process_Manager_Test.cpp: - * tests/Reactor_Performance_Test.cpp: Replaced calls to the - deprecated ACE_Reactor::run_event_loop() and end_event_loop() - with run_reactor_event_loop() and end_reactor_event_loop(). - - * tests/Proactor_Test.cpp: Add missing \n in an ACE_DEBUG message. - -Sat Nov 23 00:16:59 2002 Craig Rodrigues <crodrigu@bbn.com> - - * ace/config-posix.h: Define ACE_HAS_PTHREADS_STD - if platform defines _POSIX_THREADS. - -Fri Nov 22 13:48:39 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/Containers_T.cpp (init_head): Broke up the cascading pointer - assignments to avoid a problem when using position independent - pointer. Thanks to Eugene Alterman <eugalt@myrealbox.com> for - reporting this. - - * ace/Svc_Conf.l: - * ace/Svc_Conf_l.cpp: Changed to always remove the matching quote - character (' or ") instead of first search for double quote, - then single quote. Thanks to Johnny Willemsen - <jwillemsen@remedy.nl> for sumitting the patch. - -Fri Nov 22 04:52:16 2002 Bala <bala@cs.wustl.edu> - - * ACE version 5.2.6 released. - -Tue Nov 19 22:36:38 2002 Ossama Othman <ossama@uci.edu> - - * ace/Time_Value.h (sec, usec, msec): - - Documentation updates/clarifications. - - * ace/Time_Value.cpp (operator++, operator--): - - Fixed postfix increment and decrement operators. Previously - they had the same implementation as their prefix increment and - decrement operator counterparts. A copy of the ACE_Time_Value - is now done prior to in/decrementing it via the corresponding - prefix operator. That copy is returned. A copy is only - performed for the postfix case. - -Tue Nov 19 18:14:10 2002 Pradeep Gore <pradeep@oomworks.com> - - * bin/auto_run_tests.lst: - Modified Notify configurations. - - * bin/auto_run_tests.pl: - Checkin on behalf of Irfan - Changed logic so that it checks for - the existence of the executable and not of the parameters passed - to the executable. - -Tue Nov 19 07:09:15 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: Tests not runnable in minimum - configuration were being run. Hence prevented them from - running. - -Mon Nov 18 20:49:33 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * examples/IPC_SAP/FILE_SAP/client.cpp (ACE_TMAIN): Changed to use - predefined ACE_DEFAULT_FILE_PERMS instead of the non-portable 0666. - -Mon Nov 18 12:27:46 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/Dev_Poll_Reactor.cpp (handler): - * ace/WFMO_Reactor.i (handler): - * ace/Select_Reactor_Base_T.cpp (handler_i): The methods that - returns a signal handler associated with a certain type of - signal, does restrictive sanity checks. One of the checks that - it performs is to see whether a pointer to an event_handler is 0 - just before assigning a new event handler pointer. This check - was not required at all and hence it is now removed. Thanks to - Edan Ayal for reporting this. - -Sat Nov 16 09:38:38 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/*/Makefile: - * tests/Makefile: - * examples/*/Makefile: - * performance-tests/*/Makefile: - * ACEXML/*/Makefile: Updated dependencies. - -Fri Nov 15 20:33:02 UTC 2002 Don Hinton <dhinton@ieee.org> - - * ace/SOCK_Dgram_Mcast.h: - * ace/config-all.h: Added new macro for platforms that don't yet - implement the new IGMPv3, perfect multicast filtering, and - replaced the #ifdef linux in the ACE_SOCK_Dgram_Mcast::ctor - with the new macro, since this isn't just a linux thing. - - * ace/config-aix-4.x.h: - * ace/config-freebsd.h: - * ace/config-linux-common.h: - * ace/config-netbsd.h: - * ace/config-openbsd.h: - Defined ACE_LACKS_PERFECT_MULTICAST_FILTERING for these platforms. - - * ace/SOCK_Dgram_Mcast.cpp (open_i): Removed redundant call to - ACE_SOCK::set_option() to set SO_REUSEADDR, since it's set in - the call to ACE_SOCK::open(). This fixes [bug 1111]. - - * tests/Multicast_Test.cpp: Enhanced test by adding new options - control the wait time when running test as a CONSUMER only, - and contol the TTL so the test could be used over routers. - Also cleaned up the code a bit to make it more readable. Thanks - to Bill Fulton for motivating these enhancements. - -Fri Nov 15 14:16:10 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/Thread_Exit.h: - * ace/Thread_Exit.cpp (cleanup): Changed the type of the argument - <cleanup> takes to the generic (void*) and avoid including - <Synch.h> in Thread_Exit.h. This eliminated the circular header - dependencies when ACE_LEGACY_MODE is define. - - * ace/ace_dll.dsp: - * ace/ace_lib.dsp: Removed some template files from the build - lists and make sure all template files are not being built. - -Fri Nov 15 13:22:41 2002 Steve Huston <shuston@riverace.com> - - * tests/Proactor_Test.cpp: On Win32, check for canceled I/O using - ERROR_OPERATION_ABORTED instead of ECANCELED. - -Fri Nov 15 16:51:11 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * ace/POSIX_CB_Proactor.h: - * ace/POSIX_CB_Proactor.cpp: - Do not build on Lynx. - -Fri Nov 15 14:14:39 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * ace/POSIX_CB_Proactor.h: - * ace/POSIX_CB_Proactor.cpp: - Do not build on Sun. - -Fri Nov 15 14:08:11 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * include/makeinclude/platform_linux.GNU: - Linux glibc has aio_ and pthread_ calls in librt.so. - Add fix so that librt.so is linked in if threads=0. - -Fri Nov 15 07:48:37 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * tests/OS_Test.cpp: Added the test cases for itoa from - Joseph Sarbak <joseph.sarbak@csfb.com>. - -Fri Nov 15 07:08:46 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: Fixed a few minor problems that showed - up in the daily builds. - -Thu Nov 14 22:08:49 2002 Steve Huston <shuston@riverace.com> - - * tests/Proactor_Test.cpp: Make this test behave better: - - At the end of running time, close all the Senders. This should - cancel all the Sender outstanding operations and trip all the - Receivers into finishing their receives and closing themselves. - - Change the default Sender/Receiver count from 20 to 10. This - shrinks the log a bit without sacrificing test quality. - - Change the "inferior platforms" default Sender/Receiver count - from 1 to 3. With Proactor improvements made so far, let's give - these platforms some more exercise and see what happens. - -Thu Nov 14 23:10:32 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * ace/POSIX_CB_Proactor.cpp: - * ace/POSIX_CB_Proactor.h: - * ace/POSIX_CB_Proactor.i: - * ace/POSIX_Proactor.cpp: - * ace/POSIX_Proactor.h: - * ace/POSIX_Proactor.i: - * ace/Proactor.cpp: - * tests/Proactor_Test.cpp: - Fixed Proactor so that on platforms that support POSIX IO, - but not POSIX RT signals, it will compile. Use the new - ACE_HAS_POSIX_REALTIME_SIGNALS macro. - -Thu Nov 14 22:47:45 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * include/makeinclude/platform_freebsd.GNU: Merged with the - pthread version of this file. To compile without threads, - set, do make threads=0, or set threads=0 in this file. - * include/makeinclude/platform_freebsd_pthread.GNU: Removed. - * ace/config-freebsd.h: Merge with pthread version of this file. - * ace/config-freebsd-pthread.h: Removed. - * ace/OS.h: Use new ACE_HAS_POSIX_REALTIME_SIGNALS macro for - POSIX RT signals stuff, instead of ACE_HAS_AIO_CALLS. - FreeBSD has POSIX AIO, but not POSIX RT signals. - * ace/config-sunos5.6.h: Define ACE_HAS_POSIX_REALTIME_SIGNALS, - ACE_HAS_POSIX_MESSAGE_PASSING. - * ace/config-linux-common.h: Define ACE_HAS_POSIX_REALTIME_SIGNALS, - ACE_HAS_AIO_CALLS. - -Thu Nov 14 16:19:57 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: Added two tests, Bug_1269_Regression and - Bug_1270_Regression into the daily regression suite. They will - not run on Win32. - -Thu Nov 14 13:35:44 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/Containers_T.cpp: Changed to deallocate internal nodes in - ACE_Double_Linked_List with the allocator. Thanks to Eugene - Alterman <EugeneA@bremer-inc.com> for reporting the problem. - -Thu Nov 14 10:57:47 2002 Carlos O'Ryan <coryan@atdesk.com> - - * ace/OS.h (ace_timezone): - - Fixed broken implementation for Linux. Call tzset() and then - return the global "timezone" variable instead. [Bug 1359] - -Thu Nov 14 12:19:51 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/Creator.pm: - - Modify the way projects are created on windows. If using - cygwin perl then use \n instead of \r\n. - - * bin/MakeProjectCreator/modules/VC7WorkspaceCreator.pm: - - Sort the list of projects to make the generated solution file more - similar to actual solution files. - -Thu Nov 14 06:16:51 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/Parser.pm: - - Take into account relative directories in the cd() method. - The original optimization overlooked this. - -Thu Nov 14 06:59:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/Date_Time.{h,i}: - Added update() method with which it is possible to set the values - in ACE_Date_Time based on a passed ACE_Time_Value. - - * ace/Time_Request_Reply.{h,cpp}: - Made the argument of the timeout method const&. Updated doxygen - documentation and marked init() as deprecated because it is a no-op. - -Wed Nov 13 18:28:39 2002 Steve Huston <shuston@riverace.com> - - * tests/FIFO_Test.cpp: Changed to both clarify the usage of - ACE_FIFO_(Send|Recv)_Msg with and without ACE_HAS_STREAM_PIPES - and test that usage. - -Wed Nov 13 17:39:41 2002 Steve Huston <shuston@riverace.com> - - * ace/POSIX_Proactor.{h cpp}: - * ace/SUN_Proactor.{h cpp}: Leave two handle_events() methods - available; one with a timeout, one without. Each proactor's - implementations calls an internal handle_events_i() method - that does the right thing. - (start_aio): This is a virtual method used to start an asynch I/O - request from outside the proactor (such as from an asynch read - stream). When the actual operation is to start, start_aio calls - start_aio_i(), implemented differently for the different proactors. - Similar arrangement for cancel_aio(). - - * tests/Proactor_Test.cpp: Cancel all outstanding operations before - trying to abruptly stop everything. - -Wed Nov 13 17:29:31 2002 Steve Huston <shuston@riverace.com> - - * ace/WIN32_Asynch_IO.h (cancel_uncompleted): Filled in - description of the 'set' parameter. Thanks to Craig - Rodrigues for noticing its omission. - -Wed Nov 13 16:52:04 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * ace/Lib_Find.cpp (ldname): MacOS X prepends an underscore to - symbols in shared libaries. - - * ace/config-macosx.h: - * include/makeinclude/platform_macosx.GNU: - - - add support for dlcompat, a third party package which provides - dlopen(), dlsym(), dlcose() functions - - - Define ACE_LD_SEARCH_PATH to DYLD_LIBRARY_PATH and - ACE_LD_SUFFIX to .dylib - -Tue Nov 12 14:12:59 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - This is a hackish fix for BUG 1361. Please see - http://deuce.doc.wustl.edu/bugzilla/show_bug.cgi?id=1361 - for details of this problem - - * ace/TP_Reactor.cpp (handle_socket_events): A hack to prevent - BUG 1361 from surfacing for applications like TAO, using the - TP_Reactor. This fix basically ensures that resume_i () is - not called naively. - - * ace/Event_Handler.h: Added a enum type which could be could be - used by the TP_Reactor. - -Tue Nov 12 11:53:29 2002 Steve Huston <shuston@riverace.com> - - * tests/run_test.lst: Change FIFO_Test from !MSVC to !Win32 to - prevent it from running with Borland on Windows. Thanks to - Johnny Willemsen for pointing this out. - -Tue Nov 12 14:11:33 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/Time_Value.h: - Doxygen documentation improvements. - -Tue Nov 12 13:15:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/Message_Queue_T.h: - Doxygen documentation improvements. - -Mon Nov 11 20:37:18 2002 Ossama Othman <ossama@uci.edu> - - * bin/auto_run_tests.lst: - - Added Jai's application-controlled membership load balancing - test to the list of regression tests to be run. - -Mon Nov 11 17:57:12 2002 Ossama Othman <ossama@uci.edu> - - * bin/auto_run_tests.lst: - - Added Carlos' new PortableInterceptor AMI test to the list of - regression tests to be run. - -Mon Nov 11 16:17:59 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * bin/MakeProjectCreator/modules/TemplateParser.pm (relative): - Changed to invoke slash_to_backslash from $self. The original - code only runs on some specific version of perl. - -Mon Nov 11 13:35:53 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/Hash_Map_Manager_T.h: Updated - comments on the class ACE_Hash_Map_Bucket_Iterator. Thanks to - Kitty and Irfan for helping with this. - -Mon Nov 11 12:52:41 2002 Steve Huston <shuston@riverace.com> - - * docs/ACE-subsets.html: Added the Timer component to the list of - buildable pieces. Thanks to Jason Pasion <J.Pasion@motorola.com> - for reporting this problem. - - * THANKS: Added Jason Pasion to the Hall of Fame. - -Mon Nov 11 12:39:18 2002 Steve Huston <shuston@riverace.com> - - * ace/FIFO_Send_Msg.cpp (send): When setting iovec length, - static_cast the int length to size_t; the target is not an - int (as the strbuf length is). When returning the sent byte - count, don't include any additional bytes added internally - for record framing. - - * tests/FIFO_Test.cpp: New test to exercise basic and overflow - message-based FIFO I/O. - - * tests/Makefile: - * tests/run_test.lst: Added FIFO_Test. Disabled on MSVC builds - since Windows doesn't have FIFOs. - -Mon Nov 11 09:58:13 2002 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * ace/CDR_Stream.{i,cpp} Add casts to ACE_CDR::Octet and - ACE_CDR::Boolean to prevent warnings when run with MSVC++ level - 4 warnings. Thanks to Lukas Gruetzmacher - <gruetzmacher@ais-dresden.de> for motivating this fix. - - * ace/OS_String.cpp (itoa_emulation): Updated the code so that it'll - work correctly with negative numbers. Thanks to Joseph Sarbak - <joseph.sarbak@csfb.com> for reporting this and providing a fix. - -Sun Nov 10 19:38:45 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/Message_Block.cpp(copy): The copy () methods were copying - data till the end_ pointer. This is a wrong assumption, - especially in places where the alloted bytes are less than the - allocated bytes. Now we use the space () method to find out the - amount of space available in the data block for copying. Thanks - to Edan Ayal for finding this and providing a patch. - - * ace/Simple_Message_Block_Test.cpp: Added a test which tests - copying in a case where the alloted space is less than the - allocated space. Thanks to Edan Ayal for this test case. - -Sat Nov 9 12:37:54 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * docs/ACE-guidelines.html: Updated the ACE guidelines to - refer to the correct Doxygen links. Thanks to - Johnny Willemsen for reporting this. - -Sat Nov 9 21:48:31 2002 Steve Huston <shuston@riverace.com> - - * tests/Proactor_Test.cpp: When disabling Nagle's, use IPPROTO_TCP, - not SOL_SOCKET, for the setsockopt level. Fixes set_option errors - in this test log. - -Sat Nov 9 18:49:11 2002 Steve Huston <shuston@riverace.com> - - * ace/POSIX_Proactor.{h cpp i} (create_asynch_read_stream, - create_asynch_write_stream, create_asynch_read_dgram, - create_asynch_write_dgram, create_asynch_read_file, - create_asynch_write_file, create_asynch_accept, - create_asynch_connect, create_asynch_transmit_file): - Refactored all the factory creation methods from - ACE_POSIX_AIOCB_Proactor to ACE_POSIX_Proactor. - Moved the pseudo task from ACE_POSIX_AIOCB_Proactor to - ACE_POSIX_Proactor as well. - - (handle_events, cancel_aio): - Added these methods to ACE_POSIX_Proactor as pure virtual, - forcing their definition in the specific implementations. - - (start_aio): This is now a pure virtual function that all - more specific proactors need to implement. The old start_aio() - is now start_aio_i(). What used to be ACE_POSIX_AIOCB_Proactor:: - register_and_start_aio() is now ACE_POSIX_AIOCB_Proactor::start_aio. - Which type of I/O to start is now defined by an enum Opcode - that's passed to start_aio(). - - * ace/POSIX_Asynch_IO.{h cpp} (ACE_POSIX_Asynch_Operation): - Replace references to ACE_POSIX_AIOCB_Proactor with - ACE_POSIX_Proactor. Refactoring things out of specific proactor - types. - (ACE_POSIX_Asynch_Read_Stream, ACE_POSIX_Write_Stream, - ACE_POSIX_Asynch_Read_File, ACE_POSIX_Asynch_Write_File, - ACE_POSIX_Asynch_Accept, ACE_POSIX_Asynch_Connect, - ACE_POSIX_Asynch_Transmit_File, ACE_POSIX_Asynch_Transmit_Handler, - ACE_POSIX_Asynch_Read_Dgram, ACE_POSIX_Asynch_Write_Dgram): - Replace args of type ACE_POSIX_AIOCB_Proactor with - ACE_POSIX_Proactor. Matches POSIX_Proactor.* changes above. - Removed all the methods that said they were only there to kill - dominance warnings. These are not needed. - - These things all simplify the proactor framework code. Thanks to - Alex Libman for these ideas. - -Sat Nov 9 13:16:50 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * THANKS: Added Simon McQueen to the hall of fame. - -Fri Nov 8 13:04:10 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/ProjectCreator.pm: - - Fixed a bug in MPC where the idl file name could cause - an error on Windows if it had a slash in it. It would be - converted to a back slash and then interpreted inside a - regular expression which would cause MPC to abort. - -Fri Nov 8 13:00:56 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/BorlandProjectCreator.pm: - * bin/MakeProjectCreator/modules/BorlandWorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/Creator.pm: - * bin/MakeProjectCreator/modules/Driver.pm: - * bin/MakeProjectCreator/modules/EM3ProjectCreator.pm: - * bin/MakeProjectCreator/modules/EM3WorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/GHSProjectCreator.pm: - * bin/MakeProjectCreator/modules/GHSWorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/GNUProjectCreator.pm: - * bin/MakeProjectCreator/modules/GNUWorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/GUID.pm: - * bin/MakeProjectCreator/modules/NMakeProjectCreator.pm: - * bin/MakeProjectCreator/modules/NMakeWorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/Parser.pm: - * bin/MakeProjectCreator/modules/ProjectCreator.pm: - * bin/MakeProjectCreator/modules/TemplateInputReader.pm: - * bin/MakeProjectCreator/modules/TemplateParser.pm: - * bin/MakeProjectCreator/modules/VA4ProjectCreator.pm: - * bin/MakeProjectCreator/modules/VA4WorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/VC6ProjectCreator.pm: - * bin/MakeProjectCreator/modules/VC6WorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/VC7ProjectCreator.pm: - * bin/MakeProjectCreator/modules/VC7WorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/WorkspaceCreator.pm: - - Performed various optimization which dramatically decreased the - amount of time to generate projects for the entire ACE tree. In - tests, the time went from 2 minutes to 1 minute. - -Fri Nov 8 12:57:57 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/README: - - Added documentation for the new verbatim keyword. - - * bin/MakeProjectCreator/modules/ProjectCreator.pm: - * bin/MakeProjectCreator/modules/TemplateParser.pm: - - Added functionality to place test verbatim into a generated - project file. - - * bin/MakeProjectCreator/templates/gnu.mpd: - - Added verbatim markers top, macros, local and bottom. - -Fri Nov 8 12:52:16 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/README: - - Added documentation about the new idlgendir keyword. - - * bin/MakeProjectCreator/modules/GHSWorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/Parser.pm: - * bin/MakeProjectCreator/modules/ProjectCreator.pm: - * bin/MakeProjectCreator/modules/TemplateParser.pm: - * bin/MakeProjectCreator/modules/VC6WorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/VC7WorkspaceCreator.pm: - - Modified flag_overrides to work with both idlflags and idlgendir. - - * bin/MakeProjectCreator/templates/bor.mpd: - * bin/MakeProjectCreator/templates/em3vcp.mpd: - * bin/MakeProjectCreator/templates/gnu.mpd: - * bin/MakeProjectCreator/templates/nmake.mpd: - * bin/MakeProjectCreator/templates/vc6dsp.mpd: - * bin/MakeProjectCreator/templates/vc7.mpd: - - Modified all templates that deal with idl files to utilize the - idlgendir keyword. - -Fri Nov 8 04:03:29 UTC 2002 Don Hinton <dhinton@ieee.org> - - * tests/Multicast_Test.cpp: Cleaned up error messages and - error handling and insure that errors are returned. Added - 2 second timeout to call to ACE_Thread_Manager::wait() so - that lost messages won't cause test to hang. - -Wed Nov 6 18:55:22 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.{h i} (recursive_mutex_cond_unlock, - recursive_mutex_cond_relock): For Windows, don't try to save/restore - CRITICAL_SECTION members individually - there are undocumented - details at work there. Instead, release and acquire the mutex - one fewer times than the recursion count, letting Windows take - care of its internals. Fixes the hung Recursive_Condition_Test on - Windows. This also neatly removes the WinCE conditional code in - this area. - - * tests/Recursive_Condition_Test.cpp: Fix a const error that gcc - pointed out. - -Wed Nov 6 11:12:28 2002 Steve Huston <shuston@riverace.com> - - * tests/Recursive_Condition_Test.cpp: Replaced ACE_TRACE with - ACE_DEBUG so the log gets all the info. Re-enabled the timer queue - part of the test and added a check to be sure "enough" expirations - are recorded. Added a piece of data protected by the mutex/condition - to be sure threads aren't released too early. - -Tue Nov 5 21:54:40 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * tests/run_tests.lst: - * tests/run_tests_remote.lst: Removed these files, since they - served no purpose other than adding to the noise. - -Tue Nov 5 15:51:29 2002 Steve Huston <shuston@riverace.com> - - * ace/INET_Addr.cpp (set_addr): Add missing ';' in the IPv6 case. - -Tue Nov 5 20:26:33 UTC 2002 Don Hinton <dhinton@ieee.org> - - * tests/tests.mpc: Added Multicast_Test entry. - - * tests/README: Upadate instructions for adding new tests. - -Tue Nov 5 12:12:05 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * tests/Multicast_Test.cpp: Fixed compile errors in single threaded - builds. - -Tue Nov 5 11:24:03 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/Process.h: - * ace/Process.i: Added a new method <return_value> which returns - the actual value returned or <exit>'ed by the child process. - The <exit_code> accessor returns the raw exit status returned by - the system APIs and the value is OS dependent. Changed the - <exit_code> mutator to be a proctected method and declare - ACE_Process_Manager a fried of ACE_Process as <exit_code> - mutator should only be used by ACE_Process_Manager. - - Thanks to Stephen Blake <sblake@speakeasy.net> for reporting the - problem. - -Tue Nov 5 05:12:16 2002 Ossama Othman <ossama@uci.edu> - - * ace/Acceptor.h: - * ace/Connector.h: - * ace/Strategies_T.h: - - Added missing STL-style template traits found in C++NPv2 - including the following: - - addr_type - acceptor_type - connector_type - handler_type - stream_type - base_type - accept_strategy_type - connect_strategy_type - creation_strategy_type - concurrency_strategy_type - recycling_strategy_type - scheduling_strategy_type - - where applicable. The old traits are still available and remain - unchanged. - -Tue Nov 5 05:54:16 UTC 2002 Don Hinton <dhinton@ieee.org> - - * tests/Multicast_Test.cpp: Fixed typo in pragma instantiation, - and replaced an instance of char* with ACE_TCHAR* to fix unicode - builds. Added #ifdef for ACE_HAS_THREADS, since it uses - ACE_Task as an active object. - - * tests/run_test.lst: Turned off Multicast_Test and - Token_Stategy_Test for single threaded builds. - -Mon Nov 4 18:01:25 UTC 2002 Don Hinton <dhinton@ieee.org> - - * tests/Makefile.bor: Added new Multicast_Test. Thanks to - Johnny Willemsen <jwillemsen@remedy.nl> for noticing this. - -Mon Nov 4 16:14:21 UTC 2002 Don Hinton <dhinton@ieee.org> - - * tests/Multicast_Test.cpp (get_handle): Made return type of - definition ACE_HANDLE to match the declaration. Thanks to - Ed Skees <eskees@dctd.saic.com> for reporting this. - - * tests/Multicast_Test.dsp: - * tests/tests.dsw: Added new dsp. Thanks again to Ed Skees - for providing this. - - * THANKS: Added Ed Skees to hall of fame... - -Mon Nov 4 09:28:34 UTC 2002 Don Hinton <dhinton@ieee.org> - - * ace/SOCK_Dgram.cpp (recv): Corrected method signature in .cpp to - match the one in the .h. The .h version specified iovec*, but - the .cpp had iovec[] which confused doxygen. This fixes bug [1353]. - -Sun Nov 3 19:41:30 UTC 2002 Don Hinton <dhinton@ieee.org> - - * tests/Multicast_Test.cpp: - * tests/Makefile: - * tests/run_test.lst: Added a new test to exercise some of the new - functionality of ACE_SOCK_Dgram_Mcast. - -Sun Nov 3 10:36:50 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * tests/run_test.lst: Prevented SV_Shared_Memory test from running - in RH_7.1 configurations. Please see BUGID 1351 for details. - -Sat Nov 2 17:44:00 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/Select_Reactor_Base.h: - * ace/Select_Reactor_Base.cpp: Fixed a subtle bug in the - Select_Reactor. The following were performed during the bind () - operation, which binds a handle with interested mask to an - event_handler, - - - The new masks were added to the existing wait_set - - and the new masks were also removed from the suspend_set. - - Though the above steps were fine, there exists a possibility - when the reactor could land up with a state when the handle - could be both in the wait_set and suspend_set. If suppose a - handle is initially registered for READ and EXCEPTION with the - reactor and then suspended. If the user calls register_handler - () to register the same handle for READ and WRITE, things could - go totally bizzare. - - The present fix does the following - - - If the handle is suspended, the masks are added to the - suspend_set - - else the masks are added to the wait_set and the flag that - indicates state changes within the reactor is set to 1. - - A pure virtual call is_suspended_i () was added to the - ACE_Select_Reactor_Base class to make the check for a suspended - handle possible within the bind () call. Thanks to Irfan - Pyarali<irfan@oomworks.com> who reported this problem. This - should fix the problem reported in [BUG 1305]. - -Fri Nov 1 07:35:10 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: Added the new - MProfile_Connection_Timeout test to the daily regression tests. - -Thu Oct 31 19:15:49 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ACEXML/common/Attributes.h: Added the missing inclusion of - ACEXML_Export.h and the #pragma once directive. Thanks to Paul - Marquis <pmarquis@pobox.com> for reporting this. - -Thu Oct 31 12:57:10 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/OS.i (fstat): Added check for FILE_ATTRIBUTE_DIRECTORY on - Win32 to set the file type in the st_mode so one can determine - if it's a file or a directory. Thanks to Paul Marquis - <pmarquis@pobox.com> for submitting the patch. - -Wed Oct 30 13:47:14 2002 Steve Huston <shuston@riverace.com> - - * ace/Thread_Exit.cpp: Add #include "ace/Managed_Object.h" to get - the definition of ACE_Managed_Object. - -Wed Oct 30 11:38:36 2002 Irfan Pyarali <irfan@oomworks.com> - - * ace/Timeprobe.h: Fixed formatting of comments that somehow got - messed up in: - - http://cvs.doc.wustl.edu/viewcvs.cgi/ace/Timeprobe.h.diff?r1=4.29&r2=4.30 - -Tue Oct 29 22:27:24 2002 Frank Hunleth <frank@hunleth.com> - - * apps/gperf/src/Key_List.cpp: Added cast to int on UCHAR_MAX and - USHRT_MAX to remove warnings of comparing signed and unsigned - values under OpenBSD. - - * ace/config-openbsd.h: - * ace/config-openbsd-pthread.h: Added ACE_HAS_MKSTEMP since - OpenBSD has the mkstemp() function, and it helps clear up - linker warnings. Also added ACE_HAS_ALT_CUSERID to clear - up errors due to _cuserid(). Thanks to Craig Rodrigues - for the latter patch. - -Tue Oct 29 15:45:06 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/Thread_Exit.h: - * ace/Thread_Exit.cpp: - * ace/Thread_Manager.h: - * ace/Thread_Manager.cpp: Changed to use ACE_Thread_Manager to - clean up the global TSS ACE_Thread_Exit object instead of - depending on the more general ACE_Object_Manager::at_exit method - so we can better control the order of destruction. The change - makes sure the global TSS ACE_Thread_Exit object will only get - deleted *after* ACE_Thread_Manager has cleaned up all threads - and destroyed itself. - - Thanks to Patrick McNerthney <pat@mcnerthney.com> and - Daniel Bell <Daniel.Bell@colorbus.com.au> for reporting - the problem. - -Tue Oct 29 16:19:54 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * ace/INET_Addr.cpp: - * ace/INET_Addr.i: - * ace/INET_Addr.h: - - Added a reset() method to initialize the underlying sockaddr_in - or sockaddr_in6 structure, and set the address family. Changed - all constructors to use this reset() method. [Bug 1342] - - (INET_Addr::set_addr) the port number was not being set, - so set it. [Bug 1329] - * tests/INET_Addr_Test.cpp: Add test to verify that address family - is being set. - * THANKS: Added Thomas Devanneaux <tdevanne@akamai.com> who provided - these fixes. - -Mon Oct 28 20:46:33 2002 Steve Huston <shuston@riverace.com> - - * ace/SOCK_Dgram_Mcast.cpp (ACE_SDM_helpers::addr_to_string): Use - the "<?>" string in place rather than try to declare it a static. - gcc reports the original usage won't do what you think (keep one - copy for the method) but will duplicate it in each generated - method. Use it as a literal and let the compiler try to place a - single copy in a read-only section. - -Mon Oct 28 20:38:27 2002 Steve Huston <shuston@riverace.com> - - * ace/Asynch_IO.cpp (ACE_Service_Handler::addresses()): Don't - print addresses from here. It's inappropriate for a framework - to be printing things out without being asked to. - - * ace/POSIX_Asynch_IO.cpp (ACE_POSIX_Asynch_Read_Stream::read): - * ace/WIN32_Asynch_IO.cpp (ACE_WIN32_Asynch_Read_Stream::read): - Don't print a message for a 0-byte/no space read - set errno to - ENOSPC so the caller can figure out what's going on. - - * ace/SUN_Proactor.{h cpp}: - * ace/POSIX_Proactor.{h cpp} (ACE_POSIX_AIOCB_Proactor):Change - 'return_status' arg to get_result_status(), find_completed_aio() - from int to size_t and rename transfer_count; get_result_status() - takes care of sensing -1 count and changing to 0. Passing back - a size_t smooths the path from here through to the result object. - Removed application_specific_code() - reuse the one from - ACE_POSIX_Proactor - this one called it; remove the middle-man. - - * tests/Proactor_Test.cpp: Added addresses() method implementations - to print address with session IDs; helps to match Sender/Receiver - pairs in the log. Also added some logging of basic send/recv info - to help try to track down why this facility doesn't work well. - Added a check for comparable sends/receives when a session ends. - Added a warning if there are outstanding I/O when the session - ends. This probably should be an error, but I haven't thought - through it enough to go that far. For the SUN Proactor, use one - thread by default (not 3) - aiosuspend() is not MT safe. - -Mon Oct 28 12:48:14 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * bin/PerlACE/Process_Unix.pm (Spawn): Return 0 when the function - complete successfully. Thanks to Charles F. Frasch - <cfrasch@spawar.navy.mil> for reporting and suggesting the fix. - -Sun Oct 27 07:44:32 2002 John Aughey <jha@aughey.com> - - * ace/INET_Addr.cpp: Changed the default constructor to only - set the address family rather than doing a full set. This - should exhibit the same behavior without additional overhead. - -Sat Oct 26 21:49:15 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/Process.cpp (wait): - * ace/Process.i (wait): Changed to always update this->exit_code_ - when waiting on a child process to allow <exit_code> method to - function properly. Removed the use of the macro - ACE_OSCALL_RETURN on ACE_OS::waitpid since ACE_OS::waitpid is - already signal safe. - - * ace/Process.h: Changed the type of this->exit_code_ from int to - ACE_exitcode. Updated documentation for mutator method - <exit_code> to better reflect the intention for using this - method. - - Thanks to Artashes Ghazaryan <arty@unicad.am>, Jaroslaw Nozderko - <jaroslaw.nozderko@polkomtel.com.pl>, and Stephen Blake - <sblake@speakeasy.net> for reporting the problem. - -Sat Oct 26 11:42:04 UTC 2002 John Aughey <jha@aughey.com> - - * ace/INET_Addr.cpp: Changed default behavior to set 0 - port and INADDR_ANY address. - -Sat Oct 26 09:45:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/Timer_Heap_T.h: - Doxygen documentation fix - - * ace/OS.h: - The BCB4 header files don't define ENOTEMPTY, so define this when - detecting the BCB4 compiler. Thanks to Mike Curtis <mcurtis@wt.net> - for reporting this. - - * examples/Reactor/Proactor/test_proactor2.cpp: - Removed not needed return. Fixes a BCB4 build error - - * THANKS: Added Mike Curtis - -Wed Oct 23 16:37:15 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * THANKS: Added Sandeep to the Hall of Fame. - -Wed Oct 23 17:20:18 UTC 2002 Don Hinton <dhinton@ieee.org> - - * ace/Service_Object.{h,cpp} (suspend, resume, fini): - Changed return value from void to int so that the - return value of the contained ACE_Service_Type could be - propagated. - - * ace/Service_Repository.cpp (suspend, resume, fini): - Changed to propagate the return value from the - ACE_Service_Type instead of always 0. - -Tue Oct 22 16:26:56 2002 Rich Seibel <seibel_r@ociweb.com> - - * ace/Basic_Types.h: - Added a test for _BYTE_ORDER, same as BYTE_ORDER and - __BYTE_ORDER. VxWorks defines this. Also note, compilers - that previously automatically looked for _BYTE_ORDER when - only BYTE_ORDER was specified will likely not do so in the - future. This will proactively fix them also. - - * ace/config-all.h: - Moved the use of FUNCPTR until after the include that defines - it. Only affects VxWorks. - -Tue Oct 22 13:11:42 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/OS.cpp (DllMain): Added code to invoke - DisableThreadLibraryCalls during PROCESS_ATTACH when - ACE_DISABLES_THREAD_LIBRARY_CALLS is defined to 1 as described - below. This should enhance performance of applications that - spawn a lot of threads. Thanks to Patrick McNerthney - <pat@mcnerthney.com> for suggesting the change. - - We should probably apply the same optimization for the core TAO - libraries. - - * ace/config-win32-common.h: Defined a Win32-only new macros - ACE_DISABLES_THREAD_LIBRARY_CALLS which is default to 0 if not - defined by users. - -Mon Oct 21 17:26:15 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/SPIPE_Acceptor.cpp: Modified - SPIPE_Acceptor::create_new_instance to manually signal the event - used in the ConnectNamedPipe call when an ERROR_PIPE_CONNECTED - status is returned. Also added the handling of the status of - ERROR_NO_DATA to the same logic as ERROR_PIPE_CONNECTED. Thanks - to Patrick J. McNerthney <pat@mcnerthney.com> for contributing - this. - -Tue Oct 22 10:20:41 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * examples/C++NPv2/Reactor_Logging_Server_T.cpp: Fixed a typo that - broke the builds. My bad :( - -Mon Oct 21 20:13:27 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * examples/C++NPv2/Reactor_Logging_Server_T.cpp: - * examples/C++NPv1/Logging_Server.cpp: - * examples/C++NPv1/Process_Per_Connection_Logging_Server.cpp: - * examples/C++NPv1/RT_Thread_Per_Connection_Logging_Server.cpp: - * examples/C++NPv1/Thread_Per_Connection_Logging_Server.cpp: - Fix for compile errors on Lynx platform. Thanks to Craig - Rodrigues for supplying the patches. - -Tue Oct 22 12:36:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/Basic_Stats.h: - * ace/Dirent_Selector.h: - * ace/Service_Config.h: - * ace/Timer_Heap_T.h: - Improved doxygen documentation. - -Mon Oct 21 22:30:51 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * tests/SString_Test.cpp (ACE_TMAIN): Fixed a warning in Win32 builds. - -Mon Oct 21 20:04:12 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/WIN32_Asynch_IO.h: - * ace/WIN32_Asynch_IO.cpp: Integrated the latest patches from Eric - Frias <efrias@syncad.com>. - -Mon Oct 21 06:54:12 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * tests/Recursive_Condition_Test.cpp (test_4): - * tests/RB_Tree_Test.cpp: Fixed compile errors arising out of M$ - quirks. The scoping rules for 'for' loops are busted and hence - these fixes. These fixes basically change the variable names for - the different for loops inside the same scope. - -Sun Oct 20 21:40:13 2002 Krishnakumar B <kitty@cse.wustl.edu> - - * ACEXML/common/HttpCharStream.cpp: Handle all the error paths - gracefully and don't leak memory/sockets. - -Sun Oct 20 14:06:36 2002 Krishnakumar B <kitty@cse.wustl.edu> - - * ACEXML/common/FileCharStream.cpp (peek_i): Fixed more - compilation errors in Borland WCHAR builds. MSVC sucks... - -Sun Oct 20 11:36:34 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ACEXML/common/FileCharStream.cpp: Fixed a compile error in - Borland builds. The compile error is due to strict enforcement - of scoping rules in a for loop by Borland compiler. - -Sun Oct 20 11:31:45 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ACEXML/common/Makefile.bor: Added Encoding.obj to the list of - object files. Thanks to Johnny for pointing this out. - -Sun Oct 20 11:30:04 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * tests/SPIPE_Test.cpp: Fixed compile errors in Borland builds. - -Sun Oct 20 10:16:23 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * examples/C++NPv2/display_logfile.cpp: Fix for single threaded - builds. - -Sun Oct 20 10:12:42 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * tests/Recursive_Condition_Test.cpp: Fix for single threaded - builds. The ACE_HAS_THREADS were improperly placed. - -Sun Oct 20 09:26:58 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * examples/C++NPv1/Makefile: Added changes from "Fri Oct 18 - 21:01:25 UTC 2002 Craig Rodrigues <crodrigu@bbn.com>" back - since the changes are fine. Thanks to Craig for pointing that - out. - -Sun Oct 20 13:34:45 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * examples/C++NPv2/Makefile: Replace with recursive Makefile which - does not use for loops. Add .NOTPARALLEL target to fix - build errors on parallel builds. - -Sun Oct 20 08:10:43 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * tests/Recursive_Condition_Test.cpp: Fixed a unused variable warning. - -Sun Oct 20 00:27:02 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * tests/Recursive_Condition_Test.cpp: Fixed a fuzz error. - -Sun Oct 20 00:14:32 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * examples/C++NPv1/Makefile: Reverted the change "Fri Oct 18 - 21:01:25 UTC 2002 Craig Rodrigues <crodrigu@bbn.com>" since - this opens up some problems. Craig will look into it later and - provide a clean fix. Till then.. - -Sat Oct 19 20:00:00 2002 Stephen Torri <storri@cse.wustl.edu> - - * tests/Recursive_Condition_Test.cpp: - - Added test to check nesting_level after requiring a Recursive - Condition Mutex multiple times. Also after releasing for the - same number of times. - - * tests/SString_Test.cpp: - * ace/String_Base.i: - - Bug 1323 (Reported by Wayne Erchak <wayne@stentor.com>): Fixed. - In the effort to protect from memory problems with non-null - terminated strings the operators '<' and '>' where - rendered useless. Now the comparions use memcmp instead of - strncmp. A series of tests have been added to SString_Test.cpp - to ensure this does not happen again. - -Sat Oct 19 21:58:00 2002 Krishnakumar B <kitty@cse.wustl.edu> - - * ace/Argv_Type_Converter.cpp: - * ace/Argv_Type_Converter.h: - * ace/Argv_Type_Converter.inl: - * ace/Codecs.cpp: - * ace/Codecs.h: - * ace/OS.cpp: - * ace/Vector_T.cpp: - * ace/Vector_T.h: - * ace/WFMO_Reactor.cpp: - * tests/Proactor_Scatter_Gather_Test.cpp: Convert bool to use int. - - * ace/SString.cpp: Removed a redundant initialization to 0. - - * tests/Makefile: Add DIRS_PARALLEL=1 to allow building in - parallel in $ACE_ROOT/tests. - -Sat Oct 19 11:40:22 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.h: Added a ptrdiff_t typedef for ACE_HAS_WINCE. eVC3, - PocketPC2002 don't define this type. - - * ace/OS.i (recursive_mutex_cond_unlock, recursive_mutex_cond_relock): - PocketPC 2002 doesn't have the RecursionCount member in the - CRITICAL_SECTION. - - * ace/OS.cpp: ACE-ified the ACE_CE_ARGV class. - - * ace/OS_Errno.cpp: Add #include "ace/OS_Memory.h" to get ACE_NEW. - - * ace/ace_os_dll.vcp: Added files Time_Value.{h cpp inl}. Changed - preprocessor file locations from $ACE_ROOT to ".." to allow - working with more than one ACE version/build. - - * ace/ace_dll.vcp: Added files DLL_Manager.{h cpp}. - Changed preprocessor file locations from $ACE_ROOT to ".." to - allow working with more than one ACE version/build. - - * ace/config-win32-common.h: Set ACE_HAS_WIN32_TRYLOCK for WinCE - in addition to ACE_HAS_WINNT4. Also ACE_HAS_INTERLOCKED_EXCHANGEADD, - but only for PPC. - - * ace/Atomic_Op.h: Only set ACE_HAS_BUILTIN_ATOMIC_OP without - ACE_HAS_INTERLOCKED_EXCHANGEADD for MSVC on x86. It's not available - for eVC, CE on other than PPC. - - * tests/ARGV_Test_WinCE.vcp - * tests/Atomic_Op_Test_WinCE.vcp - * tests/Auto_IncDec_Test_WinCE.vcp - * tests/Barrier_Test_WinCE.vcp - * tests/Basic_Types_Test_WinCE.vcp - * tests/Bound_Ptr_Test_WinCE.vcp - * tests/Buffer_Stream_Test_WinCE.vcp - * tests/Cached_Accept_Conn_Test_WinCE.vcp - * tests/Cached_Conn_Test_WinCE.vcp - * tests/Cache_Map_Manager_Test_WinCE.vcp - * tests/Capabilities_Test_WinCE.vcp - * tests/CDR_Array_Test_WinCE.vcp - * tests/CDR_File_Test_WinCE.vcp - * tests/CDR_Test_WinCE.vcp - * tests/Collection_Test_WinCE.vcp - * tests/Conn_Test_WinCE.vcp - * tests/DLList_Test_WinCE.vcp - * tests/DLL_Test_DLL_WinCE.vcp - * tests/DLL_Test_WinCE.vcp - * tests/Dynamic_Priority_Test_WinCE.vcp - * tests/Enum_Interfaces_Test_WinCE.vcp - * tests/Future_Set_Test_WinCE.vcp - * tests/Future_Test_WinCE.vcp - * tests/Handle_Set_Test_WinCE.vcp - * tests/Hash_Map_Bucket_Iterator_Test_WinCE.vcp - * tests/Hash_Map_Manager_Test_WinCE.vcp - * tests/High_Res_Timer_Test_WinCE.vcpL - * tests/Lazy_Map_Manager_Test_WinCE.vcp - * tests/Logging_Strategy_Test_WinCE.vcp - * tests/Log_Msg_Test_WinCE.vcp - * tests/Malloc_Test_WinCE.vcp - * tests/Map_Manager_Test_WinCE.vcp - * tests/Map_Test_WinCE.vcp - * tests/Max_Default_Port_Test_WinCE.vcp - * tests/Mem_Map_Test_WinCE.vcp - * tests/MEM_Stream_Test_WinCE.vcp - * tests/Message_Block_Test_WinCE.vcp - * tests/Message_Queue_Notifications_Test_WinCE.vcp - * tests/Message_Queue_Test_Ex_WinCE.vcp - * tests/Message_Queue_Test_WinCE.vcp - * tests/MM_Shared_Memory_Test_WinCE.vcp - * tests/MT_Reactor_Timer_Test_WinCE.vcp - * tests/MT_Reactor_Upcall_Test_WinCE.vcp - * tests/MT_SOCK_Test_WinCE.vcp - * tests/Naming_Test_WinCE.vcp - * tests/New_Fail_Test_WinCE.vcp - * tests/Notify_Performance_Test_WinCE.vcp - * tests/Object_Manager_Test_WinCE.vcp - * tests/OrdMultiSet_Test_WinCE.vcp - * tests/OS_Test_WinCE.vcp - * tests/Priority_Buffer_Test_WinCE.vcp - * tests/Priority_Reactor_Test_WinCE.vcp - * tests/Priority_Task_Test_WinCE.vcp - * tests/Process_Manager_Test_WinCE.vcp - * tests/RB_Tree_Test_WinCE.vcp - * tests/Reactors_Test_WinCE.vcp - * tests/Reactor_Exceptions_Test_WinCE.vcp - * tests/Reactor_Notify_Test_WinCE.vcp - * tests/Reactor_Performance_Test_WinCE.vcp - * tests/Reactor_Timer_Test_WinCE.vcp - * tests/Reader_Writer_Test_WinCE.vcp - * tests/Recursive_Mutex_Test_WinCE.vcp - * tests/Refcounted_Auto_Ptr_Test_WinCE.vcp - * tests/Reverse_Lock_Test_WinCE.vcp - * tests/Semaphore_Test_WinCE.vcp - * tests/Service_Config_DLL_WinCE.vcp - * tests/Service_Config_Test_WinCE.vcp - * tests/Sigset_Ops_Test_WinCE.vcp - * tests/Simple_Message_Block_Test_WinCE.vcp - * tests/SOCK_Connector_Test_WinCE.vcp - * tests/SOCK_Send_Recv_Test_WinCE.vcp - * tests/SOCK_Test_WinCE.vcp - * tests/SPIPE_Test_WinCE.vcp - * tests/SString_Test_WinCE.vcpS - * tests/Svc_Handler_Test_WinCE.vcp - * tests/Task_Test_WinCE.vcp - * tests/Thread_Manager_Test_WinCE.vcp - * tests/Thread_Mutex_Test_WinCE.vcp - * tests/Thread_Pool_Reactor_Resume_Test_WinCE.vcp - * tests/Thread_Pool_Reactor_Test_WinCE.vcp - * tests/Thread_Pool_Test_WinCE.vcp - * tests/Timeprobe_Test_WinCE.vcp - * tests/Timer_Queue_Test_WinCE.vcp - * tests/Time_Service_Test_WinCE.vcp - * tests/Time_Value_Test_WinCE.vcp - * tests/Tokens_Test_WinCE.vcp - * tests/TSS_Test_WinCE.vcp - * tests/Upgradable_RW_Test_WinCE.vcp - Changed include file path from $ACE_ROOT to .. to allow building - multiple versions. Added aced.lib to CE ARM Debug builds. - -Sat Oct 19 16:57:44 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * etc/tao_*.doxygen: - Disabled the doxygen warnings. The warnings about not documented - members are still on. The warnings are not disabled in the ace - libraries because we could fix them there, but we can't fix them - in TAO because of the environment macro's that cause the method - definition arguments to be different than the implementation - arguments. - -Sat Oct 19 16:47:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/ace_wchar.h: - Added ACE_TEXT_WaitNamedPipe to the list of Windows functions we use - and for which there is an ascii and unicode version. - - * ace/SPIPE_Connector.cpp: - Used ACE_TEXT_WaitNamedPipe instead of ::WaitNamedPipe. This fixes - the problems in the BCB Unicode build. - -Sat Oct 19 10:09:38 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * examples/C++NPv2/Makefile: Reverted the change "Fri Oct 18 - 21:01:25 UTC 2002 Craig Rodrigues <crodrigu@bbn.com>" since - this opens up some problems. Craig will look into it later and - provide a clean fix. Till then.. - -Sat Oct 19 09:25:52 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * examples/Reactor/Misc/test_timer_queue.cpp, - * tests/Timer_Queue_Test.cpp: Fixed warnings with G++. - Thanks to Bala for reporting this. - -Fri Oct 18 22:07:31 2002 Steve Huston <shuston@riverace.com> - - * ace/config-aix-4.x.h: Only define ACE_HAS_ALLOCA for IBM - compilers. They do it via "builtin" trickery and g++ doesn't - know that trick. - - * include/makeinclude/platform_aix4_g++.GNU: Set with_ld to aix - to avoid trying to do an soname - AIX's ld doesn't do that. - - * include/makeinclude/platform_g++_common.GNU: If with_ld is aix, - don't try to add to soname option to the linker. - -Fri Oct 18 20:13:29 2002 Steve Huston <shuston@riverace.com> - - * include/makeinclude/platform_hpux_gcc.GNU: Added an optional - build parameter, with_ld, whose default value is hpux. This - says which ld is used with g++. - - * include/makeinclude/platform_g++_common.GNU: When setting the - SOFLAGS for setting SONAME, if with_ld is hpux, use +h, not -h. - -h means "hide symbol" in HP-UX ld. - -Fri Oct 18 17:12:03 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * examples/C++NPv2/display_logfile.cpp: - * examples/C++NPv2/Client_Logging_Daemon.cpp: - * examples/C++NPv2/TP_Logging_Server.h: - * examples/C++NPv2/TP_Logging_Server.cpp: Fixed compile errors in - single threaded builds. All they needed was the usage of the - right macros like ACE_SYNCH instead of ACE_MT_SYNCH and - ACE_SYNCH_MUTEX instead of ACE_Thread_Mutex. - -Fri Oct 18 16:25:09 2002 Yuanfang Zhang <yfzhang@cse.wustl.edu> - - * tests/RB_Tree_Test.h: - * tests/RB_Tree_Test.cpp: - - add a new method to test forward and reverse partial tree traversal. - -Fri Oct 18 21:01:25 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * examples/C++NPv1/Makefile: Fix comment. - * examples/C++NPv1/Makefile: Replace with recursive Makefile which - does not use for loops. - -Fri Oct 18 15:44:19 2002 Priyanka Gontla <pgontla@ociweb.com> - - * THANKS: - Added Christian Barheine <cbarheine@sympatec.com> to the hall - of fame. - -Fri Oct 18 19:37:20 GMT 2002 Craig Rodrigues <crodrigu@bbn.com> - - * ace/config-macosx.h: Define ACE_HAS_NONSTATIC_OBJECT_MANAGER - for gperf. - -Fri Oct 18 14:52:37 2002 Steve Huston <shuston@riverace.com> - - * examples/C++NPv2/Reactor_Logging_Server.cpp: - * examples/C++NPv2/Select_Reactor_Logging_Server.cpp: - * examples/C++NPv2/TP_Reactor_Logging_Server.cpp: - * examples/C++NPv2/display_logfile.cpp: Added necessary explicit - template instantiations. - -Fri Oct 18 13:21:05 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * tests/Timer_Queue_Test.cpp, - * examples/Reactor/Misc/test_timer_queue.cpp: - Fixed these tests so that it passes dynamically allocated memory - rather that values via the "act" parameter to - ACE_Timer_Queue::schedule(). This ensures the test will work - properly on strange OS platforms, such as Win64 where pointers - and longs/ints are different sizes. Thanks to Cristian - Ungureanu <cristian.ungureanu@amdocs.com> for motivating this. - -Fri Oct 18 17:20:02 UTC 2002 Don Hinton <dhinton@ieee.org> - - * ace/SOCK_Dgram_Mcast.{h,cpp}: - - Factored out common code to an open_i() method that open() - can call at this level and in inherited classes. - - * ace/QoS/SOCK_Dgram_Mcast_QoS.{h.cpp}: - - Modified ctor to take same default param as base class and - invoke the base class ctor. - - Removed duplicate code in open() and have it call base::open_i() - after opening the socket with QoS. Changed 1st parameter from - ACE_Addr to ACE_INET_Addr to match base classes. - - Removed duplicate code from subscribe by having it call base - class methods. - -Fri Oct 18 11:51:44 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * bin/PerlACE/Run_Test.pm: Created a new variable - $PerlACE::svcconf_ext which defines the canonical - svc.conf file extention used by various run_test scripts. By - default, this is defined to ".conf". It can be overwritten - using the environment variable ACE_RUNTEST_SVCCONF_EXT. For - example, you can define ACE_RUNTEST_SVCCONF_EXT to ".conf.xml" - if you have built your ACE with XML-based service configurator. - -Fri Oct 18 11:55:33 2002 Steve Huston <shuston@riverace.com> - - * ace/POSIX_CB_Proactor.cpp: Removed unneeded argument to - application_specific_code(). Thanks to Tommy Persson - <tpe@ida.liu.se> for contributing this fix. - - Fixes Bugzilla #1325. - - * THANKS: Added Tommy Persson to the Hall of Fame. - -Fri Oct 18 11:39:29 2002 Steve Huston <shuston@riverace.com> - - * examples/C++NPv1/Process_Per_Connection_Logging_Server.cpp: - * examples/C++NPv1/RT_Thread_Per_Connection_Logging_Server.cpp: - * examples/C++NPv1/Thread_Per_Connection_Logging_Server.cpp: - Put the sigterm_handler function in an unnamed namespace - to avoid issues with g++ not liking 'static' and 'extern "C"' - together. - -Fri Oct 18 03:55:51 UTC 2002 Don Hinton <dhinton@objectsciences.com> - Bill Fulton <Bill_A_Fulton@raytheon.com> - - * ace/SOCK_Dgram_Mcast.{h,i,cpp}: - - Enhanced support for join()'ing multiple multicast groups. Ctor - now offers finer control over a bind()'ing to an address or - interface. Appropriate default behavior has been specified, - but note that due to issues on Linux, portable code should not - join() more than one multicast group per instance. - - An open() method has been added/exposed. If not explicitely - called, it will be implicitely called by the first join(), and - will bind() the address if the OPT_BINDADDR_YES options was - specified in the ctor. - - Error handling has also been enhanced. For instance, if - OPT_BINDADDR_YES has been specified (the default on Linux), - additional join()'s will fail since the socket gets bound on - the first call to open() or join(). Previously, subsequent - calls to join() would silently fail, i.e., the call succeeded - but only datagrams sent to the first, bound address where - actually delivered. - - If ACE_SOCK_DGRAM_MCAST_DUMPABLE is defined, a list of all - subscribed groups is maintained and can be dump()'ed, although - access to this list is not public. - - Added new join()/leave() methods that remove unused, deprecated - protocol_family and protocol arguments. These values can be - determined on the fly. The old subscribe()/unsubscribe() - methods have been deprecated and are now implemented in terms - of join()/leave() respectively. - - The unsubscribe(void) method has also been deprecated. It was - designed to allow the caller to unsubscribe/leave() all - groups at once, but since a list was never kept (and still - isn't unless ACE_SOCK_DGRAM_MCAST_DUMPABLE is defined) only - the last group join()'ed was ever left. A better way would - be to either keep track of you own groups and call leave() - for each, or let the dtor close() the socket which automatically - unsubscribes from all groups. - - Some IPv6 support has been added either directly or by removing - the need for the caller to handle it (see join()/leave() above), - although the internal calls to setsockopt() are still hard wired - for IPv4. - -Fri Oct 18 06:51:07 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * ace/String_Base.{i,cpp}: Fixed the formatting in these files... - -Thu Oct 17 20:02:39 2002 Steve Huston <shuston@riverace.com> - - * examples/C++NPv1/Process_Per_Connection_Logging_Server.cpp: - * examples/C++NPv1/RT_Thread_Per_Connection_Logging_Server.cpp: - * examples/C++NPv1/Thread_Per_Connection_Logging_Server.cpp: - Add extern "C" to the signal handler function declaration to conform - strictly to its definition. - When explicitly instantiating auto_ptr<>, fully specify the - Thread_Args type else modern compilers (Forte 7) can't see it. - -Thu Oct 17 17:32:39 2002 Krishnakumar B <kitty@cse.wustl.edu> - - * ACEXML/examples/svcconf/Svcconf_Handler.cpp: - * ACEXML/examples/svcconf/Svcconf_Handler.h: - * ACEXML/examples/svcconf/main.cpp: - - Fixed compilation errors under Windows. - -Thu Oct 17 19:18:02 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * ace/config-macosx.h: - - Remove define of __MACOSX__, and use constant __APPLE__ - instead defined by Apple's gcc. - - Removed definitions: - ACE_HAS_MACOSX_DYLIB, ACE_LACKS_SEMBUF_T - - Added definitions: - ACE_HAS_SEMUN, ACE_HAS_SIGINFO_T, ACE_LACKS_SIGINFO_H, - ACE_HAS_UCONTEXT_T, ACE_HAS_GETIFADDRS, - ACE_LACKS_STATIC_DATA_MEMBER_TEMPLATES - - Defined ACE_HAS_POSITION_INDEPENDENT_POINTERS to 0 for temporary - Apple gcc bug workaround. - - * ace/Log_Msg.cpp (ACE_Log_Msg::log): Add check for __APPLE__ to catch - right prototype for sys_errlist. - - * ace/OS.h: Remove unnecessary __MACOSX__ check. - - * include/makeinclude/platform_macosx.GNU: Minor cleanups. - -Thu Oct 17 13:47:39 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.h: Added typedef for ACE_recursive_mutex_state for - !defined (ACE_HAS_THREADS) section. - -Thu Oct 17 16:28:05 UTC 2002 Don Hinton <dhinton@ieee.org> - - * THANKS: Added Jon Franklin <jfranklin@dctd.saic.com> - to the hall of fame. - -Thu Oct 17 10:26:59 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * examples/C++NPv2/display_logfile.cpp (ACE_MT_SYNCH>): Fixed a - uninitialized member usage warning. - -Thu Oct 17 00:26:13 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/Malloc_T.i (release): Instead of releasing the memory_pool_ - explicitly, changed to use remove () to make sure all resources - are removed properly. - -Wed Oct 16 20:23:09 2002 Krishnakumar B <kitty@cse.wustl.edu> - - * ACEXML/common/FileCharStream.cpp (getchar_i): - * ACEXML/common/FileCharStream.h: - - Borland defines getchar() as a macro. Rename it as getchar_i to - resolve the clash and fix compilation errors in the build. - - -Wed Oct 16 18:47:44 2002 Stephen Torri <storri@cse.wustl.edu> - - * examples/C++NPv1/Makefile: - - Replaced Makefile with non-for-loop version provided by - Craig Rodrigues. - -Wed Oct 16 16:52:52 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/MEM_SAP.cpp (close_shm_malloc): Fixed a memory leak that I - added last time when I fixed a race condition in removing mmap - file. - -Wed Oct 16 15:35:22 2002 Steve Huston <shuston@riverace.com> - - * tests/SPIPE_Test.cpp: Fixed compile error. - -Wed Oct 16 14:02:27 2002 Yuanfang Zhang <yfzhang@cse.wustl.edu> - - * ace/RB_Tree.cpp: - * ace/RB_Tree.h: - - forward and reverse partial tree traversal. - Changed the find_i method by adding a parameter "find_Exact". - -Wed Oct 16 12:18:03 2002 Stephen Torri <storri@cse.wustl.edu> - - * examples/C++v1/Makefile: - * examples/C++NPv1/Makefile.Iterative_Logging_Server: - * examples/C++NPv1/Makefile.Logging_Client: - * examples/C++NPv1/Makefile.Reactive_Logging_Server: - * examples/C++NPv1/Makefile.Reactive_Logging_Server_Ex: - * examples/C++NPv1/Makefile.RT_Thread_Per_Connection_Logging_Server: - * examples/C++NPv1/Makefile.Thread_Per_Connection_Logging_Server: - * examples/C++NPv1/Makefile.wProcess_Per_Connection_Logging_Server: - - Renamed from *.mak to Makefile.* to avoid looking like an - exported MSVC project NMAKE file. (Following Steve Huston's - change to examples/C++v2). - - Fixed Make error that make realclean was not in Makefile. - - * tests/Recursive_Condition_Test.cpp: - - Changed function call of signal() to broadcast() to awaken - multiple threads waiting on a shared condition variable. - -Oct 16 11:44:27 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/Thread_Manager.cpp (check_state): Changed to return 0 if - the thread id is not managed by the Thread_Manager. - - * ace/Thread_Manager.h: Updated documentation for testcancel(), - testsuspend(), and testresume (). They all return false now if - the you try to test the status of a tid that is not managed by - the Thread_Manager. - - Thanks to Giovanni Zito <g.zito@acsys.it> for reporting it. - -Wed Oct 16 12:41:29 2002 Steve Huston <shuston@riverace.com> - - * examples/Makefile: Add C++NPv1 and C++NPv2 to the DIRS list. - - * examples/C++NPv2/Makefile: Only build the SSL-using examples - (AC_CLD, AIO_CLD, TPCLS) is ssl=1 is set. - - * examples/C++NPv2/Makefile.AC_CLD: - * examples/C++NPv2/Makefile.AIO_CLD: - * examples/C++NPv2/Makefile.TPCLS: - Set up the SSL build params correctly and warn if ssl=1 not used. - -Wed Oct 16 12:08:21 2002 Steve Huston <shuston@riverace.com> - - * examples/C++NPv2/AC_CLD.mak: - * examples/C++NPv2/AIO_CLD.mak - * examples/C++NPv2/CLD.mak: - * examples/C++NPv2/Configurable_Logging_Server.mak: - * examples/C++NPv2/display_logfile.mak: - * examples/C++NPv2/Reactor_Logging_Server.mak: - * examples/C++NPv2/Select_Reactor_Logging_Server.mak: - * examples/C++NPv2/SLDex.mak: - * examples/C++NPv2/SLD.mak: - * examples/C++NPv2/SR_Configurable_Logging_Server.mak: - * examples/C++NPv2/TPCLS.mak: - * examples/C++NPv2/TPLS.mak: - * examples/C++NPv2/TP_Reactor_Logging_Server.mak: - Renamed from *.mak to Makefile.* to avoid looking like an - exported MSVC project NMAKE file. - - * examples/C++NPv2/Makefile: Changed to process new Makefile names. - -Wed Oct 16 11:02:41 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * ace/High_Res_Timer.cpp (get_cpuinfo): Removed the recalibration of - the scale factor every time the application is launched. Thanks - to Bo Balder <bba@post.utfors.se> for reporting this. - -Wed Oct 16 10:55:11 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/config-linux-common.h: Added support for ACE_CC_NAME, - ACE_CC_MAJOR_VERSION and ACE_CC_MINOR_VERSION for linux. Thanks - to Jaroslaw Nozderko <jaroslaw.nozderko@polkomtel.com.pl> for - suggesting this. - -Wed Oct 16 10:55:44 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/config-win32-msvc.h (ACE_HAS_ALLOCA): Enabled alloca support - for MSVC. Thanks to Pat McNertney <pat@mcnerthney.com> for - providing the patch. - -Tue Oct 15 18:49:25 2002 Stephen Torri <storri@cse.wustl.edu> - - * tests/Recursive_Condition_Test.cpp: - - Added two tests. - Test 2: repeat of test 1 but with a 2 second delay - after main thread signals the waiting thread. (Works) - Test 3: repeat of test 1 but with 4 threads waiting - on the condition variable. (Broken - no threads - awaken). - -Tue Oct 15 17:27:34 2002 Krishnakumar B <kitty@cse.wustl.edu> - - * ACEXML/docs/bugs.txt: - * ACEXML/docs/TODO.txt: - - Updated to reflect the current status. - -Tue Oct 15 17:23:43 2002 Krishnakumar B <kitty@cse.wustl.edu> - - * tests/Service_Config_Test.UTF-16.conf.xml: New file added which - is used in Service_Config_Test when ACE_HAS_XML_SVC_CONF and - ACE_USES_WCHAR is enabled. - -Tue Oct 15 17:17:44 2002 Krishnakumar B <kitty@cse.wustl.edu> - - * ACEXML/common/Encoding.cpp: - * ACEXML/common/Encoding.h: - - New files which handle the automatic detection of encoding in - XML files. - - * ACEXML/apps/svcconf/Makefile: - * ACEXML/common/CharStream.h: - * ACEXML/common/Exception.cpp: - * ACEXML/common/FileCharStream.cpp: - * ACEXML/common/FileCharStream.h: - * ACEXML/common/HttpCharStream.cpp: - * ACEXML/common/HttpCharStream.h: - * ACEXML/common/InputSource.cpp: - * ACEXML/common/Makefile: - * ACEXML/common/Mem_Map_Stream.cpp: - * ACEXML/common/Mem_Map_Stream.h: - * ACEXML/common/StrCharStream.cpp: - * ACEXML/common/StrCharStream.h: - * ACEXML/common/XML_Common.dsp: - * ACEXML/examples/SAXPrint/Makefile: - * ACEXML/parser/parser/Makefile: - * ACEXML/parser/parser/Parser.cpp: - - Handle the automatic detection of encoding in XML files. - -Tue Oct 15 17:15:41 2002 Steve Huston <shuston@riverace.com> - - Fixes for Bugzilla #1319: - - * ace/OS.i (ctime_r): For the 2-param ctime_r platforms, restrict - buflen to at least 26, just as 3-param platforms do. If less than - 26, return 0 and errno = ERANGE. - - * ace/OS.cpp (ctime_r): The WinCE implementation of ctime_r already - had the 26-char minimum check, but didn't set errno to ERANGE. - Now it does. - - * tests/OS_Test.cpp (ctime_r_test): Make sure that ctime_r with length - less than 26 fails and doesn't scribble on the buffer. - -Tue Oct 15 11:43:27 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/SPIPE_Connector.cpp (connect): Convert the first argument in - the ::WaitNamedPipe () call to char * always using the macro - defined. This shold fix compilation errors in WCHAR builds. - -Tue Oct 15 16:19:33 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - [Bug 1090] - * ace/config-freebsd-pthread.h: Add ACE_HAS_GETIFADDRS. This fixes - Bugzilla bug 1090. - -Tue Oct 15 10:47:20 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/SPIPE_Connector.cpp: Fixed a unused variable warning. - -Tue Oct 15 06:56:18 2002 Patrick J. McNerthney <ace@mcnerthney.com> - - * tests/SPIPE_Test.cpp: Added support for testing the new Windows - bytestream mode. - -Mon Oct 14 20:47:21 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * ace/SPIPE_Acceptor.cpp (open): Changed "mode" to "pipe_mode", - which is the name of the parameter. Thanks to Don Hinton for - reporting this. - -Mon Oct 14 19:20:35 2002 Frank Hunleth <frank@hunleth.com> - - * ACE-INSTALL.html: Added installation notes for the OpenBSD port. - - * ace/config-openbsd.h: - * ace/config-openbsd-pthread.h: - * include/makeinclude/platform_openbsd.GNU: - * include/makeinclude/platform_openbsd_pthread.GNU: Configuration - files for OpenBSD. - - * ace/ACE.cpp: Added OpenBSD-specific error code to ACE::out_of_handles - - * ace/OS.h: Added #define for OpenBSD-specific path to siginfo.h and - ACE_LACKS_CLEARERR guards around clearerr() for OpenBSD. - - * ace/OS.i: Added support for getifaddrs() with ACE_HAS_GETIFADDRS - and fix for segfault when calling dlsym with a null handle. Added - guards on clearerr using ACE_LACKS_CLEARERR. (All only affect - OpenBSD.) - - * ace/POSIX_Proactor.cpp: - * ace/POSIX_Proactor.h: Added SystemType family for OpenBSD. - - * ace/README: Added new ACE #defines to support OpenBSD. These include: - ACE_HAS_GETIFADDRS - for support of the getifaddrs() system call - ACE_HAS_DLSYM_SEGFAULT_ON_INVALID_HANDLE - add check to dlsym call - to protect against NULL. OpenBSD 3.1 needs this. - ACE_LACKS_CLEARERR - #define to remove clearerr definition due - to bogus syntax error on OpenBSD. - - * ace/Sock_Connect.cpp: Added code to get_ip_interfaces to use - getifaddrs() if it is supported. OpenBSD is the only platform to - use this currently. Under OpenBSD, the code to parse the return - values from the corresponding ioctl calls gets confused when IPv6 - interfaces are present. - - * tests/Malloc_Test.cpp: Added OpenBSD to the list of platforms that - can't handle the remap of fault part of the test. - - * tests/Reactor_Dispatch_Order_Test.cpp: Added code for OpenBSD to - clear the write mask after handle_output is called, since OpenBSD - calls handle_output more than once before the data is received - over the pipe. - -Mon Oct 14 18:21:40 2002 Patrick J. McNerthney <ace@mcnerthney.com> - - * ace/OS.h: Added FILE_FLAG_WRITE_THROUGH defines for - non-supporting operating systems and added various PIPE_* - defines for non-supporting operating systems. - - * ace/SPIPE_Connector.{h,cpp}: Added support to enable the - selection of whether bytestream or message mode is used on - Windows. Also added support for both blocking and timeout - connections on Windows, only non-blocking connections where - support before. - - * ace/SPIPE_Acceptor.{h.cpp}: Added support to enable the - selection of whether bytestream or message mode is used on - Windows. - -Mon Oct 14 20:19:51 2002 Steve Huston <shuston@riverace.com> - - * ChangeLog: Removed a large chunk of duplicated dates from Oct 11 - back to July. - - * ace/OS.{h i}: Removed the owning_thread_ member from - ACE_recursive_mutex_state on Windows. This isn't needed since the - OS mutex restores the owner properly - it was in to do a sanity - check ACE_ASSERT, but can't do that from ACE_OS. - - * ace/WFMO_Reactor.h: Expanded and clarified the class comments - to resolve Bugzilla #772. - -Mon Oct 14 17:23:43 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/CDR_Base.cpp (consolidate): Reverted the optimization - performed in "Tue Mar 12 17:43:32 2002 Balachandran Natarajan - <bala@cs.wustl.edu>". This optimizations created problems for - CORBA::Any implementation which assumes that the data is totally - contained within the Any. Using the optimization created problems - for Any's that are copied between different threads. Thanks to - Jian Chen <jchen@huawei.com> for reporting this problem. - -Mon Oct 14 10:12:31 2002 Steve Huston <shuston@riverace.com> - - Fix for Bugzilla # 1316: - - * ace/Synch.h: - * ace/OS.h: Moved the ACE_recursive_mutex_state from Synch.h - to OS.h. The recursive mutex methods need to handle this stuff - internally to properly synchronize releasing and reacquiring - the mutex around condition variable waits. - - * ace/OS.i: Added ACE_OS::recursive_mutex_cond_unlock() and - ACE_OS::recursive_mutex_cond_relock() methods to support - recursive mutex handling around waiting on a condition variable. - - * ace/Synch.cpp (ACE_Condition<ACE_Recursive_Thread_Mutex>): - Removed the guards from the broadcast() and signal() methods. - Removed the ACE_recursive_mutex_state class methods - this - functionality is now in the ACE_OS space; see above. - Reimplemented the wait() method to make use of the new ACE_OS - methods (above), enabling proper reclaiming of the recursive - mutex after cond_wait, and avoiding deadlock. - -Mon Oct 14 00:12:48 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ACEXML/docs/bugs.txt: Updated to match the current status of - ACEXML parser. - -Sun Oct 13 19:38:35 2002 Krishnakumar B <kitty@cse.wustl.edu> - - * ace/Exception_Macros.h: - - Consolidated the checking of exceptions preprocessor hacking - into this file. - - * ACEXML/common/XML_Macros.h: - - Include the above file and get rid of redundant and broken - checking for exceptions. - - * ace/CORBA_macros.h: - - Moved over the stuff from Exception_Macros.h to where it was - originally. This should fix all the build breakage. Thanks to - Bala for pointing out the errors in the build. - -Sun Oct 13 18:50:53 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * ace/SPIPE_Stream.h: Enhanced the documentation of ACE_SPIPE_Stream. - Thanks to Pat McNerthney <ace@mcnerthney.com> for motivating - this. - -Sun Oct 13 15:57:04 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/Asynch_IO.cpp: - * ace/Asynch_IO.h: Protected ACE_Asynch_Read_File::readv () and - ACE_Asynch_Write_File::writev () from being called when - ACE_HAS_WINNT4 == 0. - * ace/WIN32_Asynch_IO.cpp: - * ace/WIN32_Asynch_IO.h: Protected - ACE_WIN32_Asynch_Read_File::readv () and - ACE_WIN32_Asynch_Write_File::writev () from being called when - ACE_HAS_WINNT4 == 0. - - Thanks to Eric Frias <efrias@syncad.com> for reporting the - problem and Alex Libman for suggesting the fix. - - * THANKS: Added Eric Frias <efrias@syncad.com> to the hall of - fame. - -Sun Oct 13 15:54:59 2002 Krishnakumar B <kitty@cse.wustl.edu> - - * ACEXML/apps/svcconf/Svcconf_Handler.cpp: Fixed a couple of - warnings related to the exception argument handling. - -Sun Oct 13 00:55:27 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ACEXML/common/XML_Common.dsp: - - Added XML_Macros.h to the project file. - - * ACEXML/examples/SAXPrint/svc.conf.xml: - - Reverted a mistaken change used for testing. - - * ACEXML/apps/svcconf/Svcconf_Handler.cpp: - * ACEXML/common/SAXExceptions.cpp: - * ACEXML/parser/debug_validator/Debug_DTD_Manager.cpp: - - Added ACE_TEXT around strings so that WCHAR build is happy. - -Sat Oct 12 23:01:29 2002 Krishnakumar B <kitty@cse.wustl.edu> - - * ACEXML/apps/svcconf/Svcconf.h (ACEXML_Svcconf_Parser): Removed - the unused member env_. - -Sat Oct 12 22:50:02 2002 Krishnakumar B <kitty@cse.wustl.edu> - - * ACEXML/common/XML_Macros.h: - - New file which contains the macros needed to enable support for - Exceptions in ACEXML. - - * ACEXML/apps/svcconf/Makefile: - * ACEXML/common/Makefile: - * ACEXML/examples/SAXPrint/Makefile: - * ACEXML/parser/parser/Makefile: - * ACEXML/tests/Makefile: - - Regenerate dependencies. - - * ACEXML/common/Env.cpp: - * ACEXML/common/Env.h: - * ACEXML/common/Env.i: - * ACEXML/common/Exception.cpp: - * ACEXML/common/Exception.h: - * ACEXML/common/Exception.i: - - Enable support for Native Exceptions. - - * ACEXML/apps/svcconf/Svcconf.cpp: - * ACEXML/apps/svcconf/Svcconf.h: - * ACEXML/apps/svcconf/Svcconf_Handler.cpp: - * ACEXML/apps/svcconf/Svcconf_Handler.h: - * ACEXML/common/Attributes_Def_Builder.h: - * ACEXML/common/ContentHandler.h: - * ACEXML/common/DTDHandler.h: - * ACEXML/common/DTD_Manager.h: - * ACEXML/common/DefaultHandler.cpp: - * ACEXML/common/DefaultHandler.h: - * ACEXML/common/Element_Def_Builder.h: - * ACEXML/common/EntityResolver.h: - * ACEXML/common/ErrorHandler.h: - * ACEXML/common/HttpCharStream.cpp: - * ACEXML/common/SAXExceptions.cpp: - * ACEXML/common/SAXExceptions.h: - * ACEXML/common/SAXExceptions.i: - * ACEXML/common/Validator.h: - * ACEXML/common/XMLFilterImpl.cpp: - * ACEXML/common/XMLFilterImpl.h: - * ACEXML/common/XMLReader.h: - * ACEXML/examples/SAXPrint/Print_Handler.cpp: - * ACEXML/examples/SAXPrint/Print_Handler.h: - * ACEXML/examples/SAXPrint/SAXPrint_Handler.cpp: - * ACEXML/examples/SAXPrint/SAXPrint_Handler.h: - * ACEXML/examples/SAXPrint/main.cpp: - * ACEXML/examples/SAXPrint/svc.conf.xml: - * ACEXML/examples/svcconf/Svcconf_Handler.cpp: - * ACEXML/examples/svcconf/Svcconf_Handler.h: - * ACEXML/examples/svcconf/main.cpp: - * ACEXML/parser/debug_validator/Debug_Attributes_Builder.cpp: - * ACEXML/parser/debug_validator/Debug_Attributes_Builder.h: - * ACEXML/parser/debug_validator/Debug_DTD_Manager.cpp: - * ACEXML/parser/debug_validator/Debug_DTD_Manager.h: - * ACEXML/parser/debug_validator/Debug_Element_Builder.cpp: - * ACEXML/parser/debug_validator/Debug_Element_Builder.h: - - Switch over the files to support Native Exceptions. - - * ACEXML/parser/parser/Parser.cpp: - * ACEXML/parser/parser/Parser.h: - - Use the obstack unwind to unwind the stack and reuse memory. - Currently not done throughout the parser. That will come later. - - * ace/Makefile.ace: - - Added Codecs to ACE_COMPONENTS when building FOR_TAO. Thanks to - Don Hinton for suggesting this. - - * include/makeinclude/rules.local.GNU: - - Clean up the old symlinks under $ACE_ROOT/ace when building a - new version of the libraries. - - * tests/Makefile: - - Cleanup the Codecs_Test when $ACE_ROOT/ace has been realcleaned - previously. - -Fri Oct 11 12:44:58 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * RTJava/*: Removed. Angelo Corsaro <corsaro@ece.uci.edu> - keeps RTJperf related files in another CVS repository. - - * ace/config-freebsd-pthread.h: Improved check of - _POSIX_REALTIME_SIGNALS preprocessor constant. - -Thu Oct 10 16:02:20 2002 Ossama Othman <othman@cs.wustl.edu> - - * ACE version 5.2.5 released. - -Thu Oct 10 13:56:35 2002 Ossama Othman <ossama@uci.edu> - - * bin/make_release (check_workspace): - - Ignore platform_macros.GNU and config.h files. They are - needed when cutting the beta/release. This isn't a problem - since the ACE/TAO kits are created from a "cvs export"ed - workspace that doesn't include those files. - -Thu Oct 10 15:11:07 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/make_release: A pattern matching issue. The cvs server now - gives a message starting with "cvs server:.." instead of "cvs - checkout:..". This created problems for Ossama when he was - trying to cut the beta. - -Wed Oct 9 23:43:51 2002 Steve Huston <shuston@riverace.com> - - * ace/WIN32_Asynch_IO.cpp (ACE_Asynch_Read_Stream::readv, - ACE_Asynch_Write_Stream::writev, ACE_Asynch_Read_Dgram::recv, - ACE_Asynch_Write_Dgram::send): Don't double-increment iovcnt - while checking for the need for multiple iov per message block. - Fixes bug introduced in: - Fri Oct 4 20:22:41 2002 Steve Huston <shuston@riverace.com> - -Wed Oct 9 23:08:19 2002 Steve Huston <shuston@riverace.com> - - * include/makeinclude/rules.lib.GNU: For the funky AIX shr.o/lib*.a - machination, declare shr.o as .INTERMEDIATE to avoid rebuilding it - all the time for nothing. Thanks to Craig L. Ching for motivating - this. - -Wed Oct 09 12:19:40 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * bin/PerlACE/Process_Win32.pm: Added hack to purify a process. - This will allow us to purify all the examples, tests much - easier. It is, however, tightly coupled to Purify 6.0. The - script automaticaly switch to purifying the process if the - ACE_RUN_PURIFY_CMD environment variable is defined. See the - comments in this file for more details. - -Tue Oct 8 21:06:23 2002 Steve Huston <shuston@riverace.com> - - * ace/config-sunos5.6.h: Undef ACE_HAS_TEMPLATE_TYPEDEFS for the - Forte 7 compiler (C++ 5.4). It should work, but Malloc_T doesn't - compile correctly with it enabled. Thanks to Serge Kolgan - <skolgan@objectsciences.com> for this fix. - - * ace/config-sunos5.9.h: New config.h file for Solaris 9. Right now, - it just includes config-sunos5.8.h, but Solaris 9 has some new things - to play with as time goes on. - -Tue Oct 8 09:22:41 2002 Steve Huston <shuston@riverace.com> - - * ace/SOCK_Dgram_Bcast.cpp (send): Use next_ to find next - ACE_Bcast_Node in if_list_ rather than incrementing pointer. - Thanks to <k2@biogate.com> for this fix. - -Mon Oct 7 17:29:14 2002 Ossama Othman <ossama@uci.edu> - - * tests/Proactor_Scatter_Gather_Test.cpp (initiate_read_file): - - Cast one of the operands in a subtraction operation to a size_t - to address a problem exhibited by Borland C++ where the ACE_MIN - template expects two parameters of the same type, but was - getting two parameter of different types. This problem occurred - on platforms where "unsigned int" and "unsigned long" are - different types. - -Mon Oct 07 15:21:44 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ACEXML/apps/svcconf/XML_Svc_Conf_Parser.dsp: - * ACEXML/common/XML_Common.dsp: - * ACEXML/parser/debug_validator/Debug_Validator.dsp: - * ACEXML/parser/parser/Parser.dsp: - * ACEXML/tests/HttpCharStream_Test.dsp: - * ACEXML/tests/NamespaceSupport_Test.dsp: - * ACEXML/tests/Transcoder_Test.dsp: - * apps/JAWS2/server.dsp: - * examples/Export/test.dsp: - * examples/IPC_SAP/FILE_SAP/client.dsp: - * examples/Logger/client/client.dsp: - * examples/Logger/simple-server/server_loggerd.dsp: - * examples/QOS/Simple/receiver.dsp: - * examples/QOS/Simple/sender.dsp: - * examples/Reactor/Multicast/client.dsp: - * examples/Reactor/Multicast/server.dsp: - * examples/Reactor/Proactor/test_udp_proactor.dsp: - * examples/Synch/proc_sema.dsp: - * tests/DLL_Test_Impl.dsp: - * tests/Framework_Component_DLL.dsp: Disabled the generation - of precompiled headers. They were taking up too much disk space - on the nightly build machines. Thanks to Ossama for noticing - this. - -Sun Oct 6 09:03:28 2002 Ossama Othman <ossama@uci.edu> - - * ace/UPIPE_Stream.cpp (send): - - Silenced a signed/unsigned conversion warning using a - static_cast<> until a better solution presents itself. - -Sat Oct 5 17:04:38 2002 Steve Huston <shuston@riverace.com> - - * ace/Memory_Pool.cpp (ACE_Pagefile_Memory_Pool::acquire): Remove - casts on comparisons, allowing size_t values to compare directly. - - * apps/JAWS/server/IO.cpp (JAWS_Synch_IO::transmit_file): Remove - cast from send_n() return value to avoid signed/unsigned compare. - -Sat Oct 5 12:30:49 2002 Steve Huston <shuston@riverace.com> - - * ace/Service_Manager.cpp (list_services): Remove extra right paren. - - * tests/Proactor_Timer_Test.cpp: Fixed signed/unsigned mismatch. - -Fri Oct 04 22:24:23 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ACEXML/common/URL_Addr.cpp (string_to_addr): Added a new wrapper - which matches the signature in ACE_INET_Addr, and delegates the - work to the ACEXML_Char function to work around Borland BCB - warning about hidden virtual function. - -Fri Oct 4 21:32:51 2002 Steve Huston <shuston@riverace.com> - - * ace/Service_Manager.cpp (list_services): Correct compiler - warning about sign/unsigned comparison. Don't bother for checking - transfer count != requested, since send_n always reports - an error (0 or -1) if there was a fail before the whole 'n' sent. - -Fri Oct 4 20:22:41 2002 Steve Huston <shuston@riverace.com> - - * ace/ace_dll64.mak: Removed the /wd4267 to re-enable the - conversion size_t to int warning. - Added Atomic_Op.obj to the files to compile directly. - - * ace/ACE.cpp (recv, send): Extract int, not size_t, from implicit - char */int pairs. - (send_n, recv_n with ACE_Message_Blocks): Message block length - is a size_t but iov_len is u_long, so make as many iovs as needed - to fit the entire requested length. - - * ace/Asynch_Acceptor.cpp: Added casts to fix size_t->DWORD warnings. - - * ace/Asynch_IO.cpp (transmit_buffers): Cast lengths to DWORD for - Win64. - - * ace/Configuration.{h cpp} (set_binary_value, get_binary_value): - Change length arg from u_int to size_t. - - * ace/Configuration_Import_Export.cpp: Align use of u_int, size_t. - - * ace/Containers_T.cpp (ACE_Fixed_Set::insert): Use ssize_t, not int, - to track free slot since max size is a size_t. - - * ace/FIFO_Recv_Msg.i (recv): - * ace/FIFO_Send_Msg.i (send): Squeeze size_t max_len down to an int - because this gets run through the ACE_Str_Buf variant of recv/send. - If this restriction is too much, the code will need to be - refactored. - - * ace/FILE_Addr.cpp (set): Cast size_t back to int for base_set(). - - * ace/FILE_IO.cpp (recv): Count iov tuples as an int to match what's - expected at the ACE_OS::readv level. - - * ace/Filecache.h (ACE_File_Object): Change size_ from size_t to off_t. - It's a better match for what it's used for (checking against - stat.st_size, which is an off_t also). - - * ace/Functor_T.i (ACE_Pointer_Hash): Add #pragmas for Win64 to - shut up the compiler warnings about casting a pointer to u_long. - - * ace/INET_Addr.{h i} (ip_addr_size): Return an int, not size_t. This - value is just passed on to other methods (like host db functions) - that are expecting an int. Also, being the size of a IP addressing - structure, all the values are well within range of an int. - - * ace/Lib_Find.cpp (ldfind, get_temp_dir): Cast path length value to - DWORD for calls to SearchPath and GetTempPath on Windows. - - * ace/Local_Name_Space.{h cpp} (ACE_NS_String::strstr): Cast return - value from size_t to int. (hash): Change return type from size_t - to u_long to match what's expected from ACE_Hash<>, which this is - an adapter for. - - * ace/Log_Msg_NT_Event_Log.cpp (open): Cast msg_file length to - DWORD for use in Windows API. Use size_t for reg key string length - since it need not be passed to a Windows API. - - * ace/Log_Record.cpp (round_up): Do all the rounding math in size_t - terms, since all factors are size_t, but cast it back to ACE_UINT32 - since the ACE_Log_Record length_ field is ACE_UINT32 and should - stay that way due to the need to send it over the network as a - known-length field. - (print): Cast verbose_msg string length to int so it can be compared - to fprintf return value. - - * ace/MEM_Acceptor.cpp (accept): Cast string length to ACE_UINT16. - - * ace/MEM_SAP.h: - * ace/MEM_IO.{h cpp} (ACE_Reactive_MEM_IO, ACE_MT_MEM_IO): Changed - return type of send_buf and recv_buf from int to ssize_t. This - is consistent with other ACE sock-type send methods, and matches - the range of sizes allowed to request sending of and the return - type expected by ACE_MEM_IO::send() and recv(). - - * ace/Mem_Map.i (advise): Get a size_t length as expected by - ACE_OS::madvise(). - - * ace/Mem_Map.cpp (map_it): Use off_t instead of size_t to calculate - null_byte_positition when extending the file. This is the type - that ACE_OS::pwrite() expects. - - * ace/Memory_Pool.h (ACE_Pagefile_Memory_Pool::Shared_Control_Block): - Changed mapped_size_ from int to size_t to match the sizes it can - take (max_size_)and what's expected in the Win32/64 API calls. - Changed free_offset_ from int to ptrdiff_t to represent the full - range of mapped address range. Changed free_size_ from int to - size_t to match max_size_ and the range of mappable sizes. - Changed 'append' arg to map() from int to size_t, also to represent - full range of mappable size. - - * ace/Memory_Pool.i (round_up): Cast nbytes to off_t for call to - ACE::round_to_page_size(). - - * ace/Memory_Pool.cpp: Adjust usage of size_t/off_t, usually with - casts where needed. - (map): For Win64, allow full 64-bit range of file mapping size. - - * ace/Message_Queue.{h i cpp} (ACE_Message_Queue_NT): Changed - max_threads arg (ctor and open()) and max_cthrs_/cur_thrs_ from - size_t to DWORD, as expected by the underlying API (this class - is not portable, or compilable outside of ACE_WIN32). - - * ace/Msg_WFMO_Reactor.{h cpp} (poll_remaining_handles()): Changed - argument from size_t to DWORD - matches similar changes to - WFMO_Reactor, below. - - * ace/Name_Request_Reply.cpp (ctor): Cast length to ACE_UINT32. - - * ace/Naming_Context.cpp (info): Cast info string length to int. - - * ace/OS.{h i cpp}: (sema_post): Changed count from size_t to u_int, - to match both the value range for initializing (sema_init) and - the underlying OS calls on all platforms. - (getipnodebyaddr): For IPv4 case, cast len to int. - (recvfrom): size_t->int casting for Windows. - (read, write, pread, pwrite): Reduce byte count to DWORD range - for Windows. - (getcwd): Windows offers an int len for getcwd(), as opposed to - size_t everywhere else... so special-case the Windows need to - cast the length down. - (thr_create): Cast stack_size to u_int for beginthreadex(). - Other miscellaneous int/size_t casts to get Win64 to build clean. - (iovec): Changed iov_len field from size_t to u_long to match - WSABUF, as comment says to. - - * ace/OS_String.cpp (strtok_r_emulation): Change string lengths from - int to size_t, matching that returned from strlen(). - - * ace/Process.{h cpp} (setenv_i): Change len arg from int to size_t. - Matches string lengths in general, and the expectations of both - callers and callees with this information. - Changed environment_buf_index_ and environment_buf_len_ from int - to size_t as well, matching available range. - - * ace/Process_Manager.cpp: size_t to DWORD casting for current_count_. - - * ace/Registry.cpp (make_name): Use ssize_t instead of int for string - positions - matches interface for ACE_TString, which this method - accepts and uses. - - * ace/Remote_Name_Space.cpp (bind, rebind, resolve, list_names, - list_values, list_types, list_name_entries, list_value_entries, - list_type_entries): Cast length values to ACE_UINT32 as - ACE_Name_Request expects. - - * ace/Select_Reactor_Base.cpp (ACE_Select_Reactor_Handler_Repository): - In open(), cast size to int for call to ACE::set_handle_limit(). - - * ace/Service_Config.cpp (ctor): Cast size value to int for passing to - ACE_Service_Repository::instance(). The ACE_Service_Repository - size should probably be changed to size_t for consistency. - - * ace/Service_Manager.cpp (info): Cast return string length to int. - - * ace/Service_Types.cpp (info): Cast return string length to int. - - * ace/Shared_Memory_MM.i (get_segment_size): - * ace/Shared_Memory_SV.i (get_segment_size): Cast shared_memory_ size - to int. Legit since the open() call's size is an int; otherwise - this should be changed to return a size_t. - - * ace/SOCK_Connector.cpp (shared_connect_start): Align use of - int/size_t. - - * ace/SOCK_Dgram.{h cpp} (recv, send): - * ace/SOCK_Dgram_Bcast.{h cpp} (recv, send): - * ace/SOCK_Sgram_Mcast.{h i} (send, recv): Change iovcnt from - size_t to int to match the args used by other iovcnt-accepting - methods in ACE. - - * ace/SOCK_Dgram_Bcast.i (send): Align use of int/size_t. - - * ace/SOCK_IO.cpp (recv (size_t, ...), send(size_t, ...)): The - comments say the pairs are char*, int, but the code extracted - ssize_t values instead of int. Changed to extract ints as the - comment says. The values are put in iovecs which accept - unsigned long, not ssize_t. - - * ace/SPIPE_Addr.cpp (set): Align use of int/size_t. - - * ace/SPIPE_Stream.cpp (send, recv): Align use of int/size_t. - - * ace/SString.i (ACE_NS_WString): Add missing arg for ACE_WString - ctor initialization. - (ACE_SString::rfind): Cast string len to int for searching... note - this limits the string to 'int' range, though the other size - arguments/parameters are size_t. There are crocodiles lurking - here... hopefully this class is on its way out. - - * ace/SV_Shared_Memory.{h i} (get_segment_size, round_up): Changed - return type from int to size_t, matching it's argument type. - Also changed class's size_ member from int to size_t, matching - the arg from which it's set. - - * ace/Synch.{h i} (ACE_Semaphore::release): Change release count from - size_t to u_int - also see corresponding change to OS.h, above. - - * ace/Thread.{h cpp} (spawn_n): Change return value from int to size_t, - matching the range available via the 'n' argument. - - * ace/Thread_Manager.{h i cpp}: - (count_threads): Change return value from int to size_t. Matches - the requestable number of threads in spawn_n, as well as - underlying container size return. - (find_task): Change slot argument from int to size_t... matches - range of available slots. The default changed from -1 to 0, but - should make no difference since first check is for 0 >= <slot>. - (thread_all_list, task_all_list, task_list, thread_list, - hthread_list, thread_grp_list, hthread_grp_list): Changed - return value from int to ssize_t, to match <n> argument. - - * ace/Timer_Hash_T.{h cpp}: This timer queue generates timer IDs - by casting a pointer to a dynamically-allocated structure to - a long. This, of course, is invalid on Win64. On Win64, the ID - is obtained by masking off the lower 32 bits of the pointer, - and saving the upper 32 bits. This adds an assumption that all - the Hash_Tokens allocated will have the same upper 32 bits in - the pointer value. Note that the original pointer value is used - as the ACT value in the timer scheduled within this class. The - timer ID must only be reconverted to cancel a timer by ID, or - to explicitly reschedule it. Expiration and cancel-by-handler - are all handled by reconverting the ACT to the dynamically-allocated - memory area (Hash_Token). - Also changed hash starter from time.usec() to time.sec(). usec() - isn't usually anything of value. - - * ace/Timer_Wheel_T.cpp: Added #pragma to disable compile warnings. - We know of the issue and deliberately used the fields this way. - - * ace/UPIPE_Stream.{h cpp} (send, recv): Changed return value from - int to ssize_t, consistent with other IPC streams in ACE. - - * ace/WFMO_Reactor.{h cpp} (ACE_WFMO_Reactor_Handler_Repository): - Changed max_handlep1_ and max_handlep1 () from size_t to DWORD. - Matches what they're used for (input to WaitForMultipleObjects). - (ACE_WFMO_Reactor): Changed wait_for_multiple_events() to return - DWORD, not int - matches what WaitForMultipleObjects returns. - Changed wait_status and slot args to dispatch(), safe_dispatch(), - dispatch_handles (), dispatch_handler (), simple_dispatch_handler (), - complex_dispatch_handler (), and poll_remaining_handles () to also - be DWORD, not int, as they're working directly (or close to it) - with the same WFMO return value. - - * ace/WIN32_Asynch_IO.cpp (ACE_WIN32_Asynch_Read_Stream::readv, - ACE_WIN32_Write_Stream::writev, ACE_WIN32_Asynch_Read_Dgram::recv, - ACE_WIN32_Asynch_Write_Dgram::send): - Make multiple iovecs if needed to transfer all data requested. - (ACE_WIN32_Asynch_Read_Stream::shared_read, - ACE_WIN32_Asynch_Write_Stream::shared_write, - ACE_WIN32_Asynch_Read_File::readv, - ACE_WIN32_Asynch_Write_File::writev, - ACE_WIN32_Asynch_Accept::accept, - ACE_WIN32_Asynch_Transmit_File::transmit_file): Limit requested - bytes to MAXDWORD. Return -1, ERANGE if over. - - * ace/WIN32_Proactor.{h cpp}: Changed number_of_threads_ member from - size_t to DWORD, matching the API calls it's passed to. - (post_completion): Cast transfer count from size_t back to DWORD - for posting. The cast is ok since the original requested transfer - counts (in WIN32_Asynch_IO) were limited to DWORD range. - - * tests/Config_Test.cpp: - * tests/Proactor_Scatter_Gather_Test.cpp: - * tests/Proactor_Test.cpp: - * tests/Proactor_Timer_Test.cpp: - * tests/Reactor_Timer_Test.cpp: - * tests/Semaphore_Test.cpp: - * tests/TP_Reactor_Test.cpp: size_t/int fixes. - -Fri Oct 4 18:39:19 2002 Steve Huston <shuston@riverace.com> - - * tests/Recursive_Condition_Test.cpp: Added a deactivate/wait - on the timer queue to make sure the queue shuts down. This - reveals a bug in the recursive condition... it's stuck. - Also added a much simpler wait/signal test that shows the issue - without the additional stuff to trudge through while looking for - the problem. Thanks to Philippe Haussy <p.haussy@xpedite.com> - for reporting this and sending a nice, concise test case. - - * THANKS: Added Philippe Haussy to the Hall of Fame. - -Thu Oct 3 17:29:36 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/OS_Dirent.cpp (readdir_emulation): Don't do a FindNextFile() - right after FindFirstFile() to avoid overwriting the existing - data! Thanks to Dmitry Botcharnikov <dimab@garant.ru> for - reporting this. - -Wed Oct 2 20:44:19 2002 Steve Huston <shuston@riverace.com> - - * tests/Proactor_Scatter_Gather_Test.cpp: Added some more diagnostic - info and removed ACE_ASSERT on file not compared correctly - just - note the error and return. When about to writev a chain of blocks, - update the odd/even chains before writing in case the write fails; - fixes a crash on NT4... test still fails, but doesn't crash. - -Wed Oct 2 10:50:49 2002 Ossama Othman <ossama@uci.edu> - - * ace/Basic_Types.h: - * ace/Basic_Types.cpp: - - Do not rely on the u_int and u_long typedefs. They are not - always defined in system headers. Addresses a compile-time - issue when _XOPEN_SOURCE is explicitly set to 600 on Glibc - platforms, such as Linux. - - * ace/OS.h: - - Explicitly define the u_{char,short,int,long} typedefs on Glibc - platforms when _BSD_SOURCE is not defined. Addresses a - compile-time issue when _BSD_SOURCE is not defined. - - * ace/config-linux.h: - - If ACE_HAS_PTHREADS_UNIX98_EXT is not defined, do not define - ACE_LACKS_RWLOCK_T. Addresses a compile-time issue when - _XOPEN_SOURCE is not set to a value >= 600. - - * ace/config-linux-common.h: - - Reverted change that explicitly defined _XOPEN_SOURCE to 600. - Explicitly defining _XOPEN_SOURCE to 600 causes many - compile-time errors and warnings to occur. - - Moved the check for (_XOPEN_SOURCE >= 600) after the - <features.h> include. That header may define _XOPEN_SOURCE when - other macros, such as _GNU_SOURCE, are defined. - -Wed Oct 2 01:08:51 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ace/SString.cpp (ACE_NS_WString): Fix the silly off-by-one error - causing heap corruption and core dumps in $ACE_ROOT/Naming_Test. - Thanks to Vladislav <vladislav@gtss.spb.ru> for reporting the - problem. - -Tue Oct 1 20:46:35 2002 Christopher Kohlhoff <chris@kohlhoff.com> - - * ace/Atomic_Op.cpp: - Added explicit instantiations for ACE_Atomic_Op<ACE_Thread_Mutex,long> - and ACE_Atomic_Op_Ex<ACE_Thread_Mutex,long>. Programs using ACE should - no longer explicitly instantiate these classes. This change eliminates - the complexity of having to test for the ACE_HAS_BUILTIN_ATOMIC_OP - define when doing explicit instantiations. - - * tests/Atomic_Op_Test.cpp: - * tests/Notify_Performance_Test.cpp: - Removed ACE_Atomic_Op explicit instantiations that are now made in the - ACE library itself. - -Tue Oct 1 19:49:36 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ace/Exception_Macros.h: - - Fixed some documentation for using the Exception macros. - - * ace/OS.i: - * ace/config-aix-4.x.h: - * ace/config-hpux-11.00.h: - * ace/config-linux-common.h: - * ace/config-linux.h: - - Enable support for using native rwlocks on Linux. - -Tue Oct 1 05:06:18 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * ace/WIN32_Asynch_IO.cpp: Minor cosmetic changes. - -Mon Sep 30 12:21:32 2002 Ossama Othman <ossama@uci.edu> - - From Christophe Juniet <cjuniet@entreview.com> - * ace/Sock_Connect.cpp (get_bcast_addr, get_ip_interfaces) - (count_interfaces): - - The ifr_addr.sa_len field was not taken into account when - incrementing the ifreq pointer on FreeBSD. Addresses network - interface enumeration problems in FreeBSD builds. - -Mon Sep 30 09:33:42 2002 Christopher Kohlhoff <chris@kohlhoff.com> - - * bin/MakeProjectCreator/templates/bor.mpd: - * include/makeinclude/build_dll.bor: - * include/makeinclude/build_exe.bor: - * include/makeinclude/build_lib.bor: - * include/makeinclude/clean.bor: - * include/makeinclude/outputdir.bor: - Add support for the 'install', 'dllout' and 'libpaths' - MakeProjectCreator settings in generated Borland makefiles. - -Mon Sep 30 13:25:12 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * include/makeinclude/platform_linux_borland.GNU: - Now it is possible to create shared libraries, but no executables - yet with Kylix3. - -Mon Sep 30 11:33:12 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * include/makeinclude/platform_linux_borland.GNU: - Extended Kylix3 GNU file. - -Mon Sep 30 11:13:33 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/config-borland-common.h: - * ace/config-win32-common.h: - Moved the dll export stuff for the Borland compiler from the win32 - file to the borland file because the Kylix compiler also supports - the dll export stuff for shared objects. - -Mon Sep 30 10:19:33 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * include/makeinclude/rules.bin.GNU: - Added special part for Kylix. The Kylix linker expects that the - object files and the name of the executable are separated using - a comma. - -Sun Sep 29 08:19:09 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ACEXML/examples/SAXPrint/Makefile: Removed inclusion of - rules.lib.GNU to allow mingw compilation to proceed. Thanks to - Vince Mounts <vince@mounts.cc> for reporting this. - - * ace/SString.h: Moved the ACE_INLINE before ACE_Export in - declaration of operator+ to resolved a mingw warning. Thanks to - Vince Mounts <vince@mounts.cc> for reporting this. - -Fri Sep 27 12:49:33 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * include/makeinclude/rules.lib.GNU: - Added special part for Kylix - -Thu Sep 26 21:52:05 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * THANKS: Added Erich Hochmuth" <erich_hochmuth@hotmail.com> to - the hall of fame. - -Thu Sep 26 13:48:26 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/TP_Reactor.cpp (remove_handler): Make a check for the return - value from the remove_handler_i () call before calling - handle_close (). Thanks to Eric Malenfant - <EMalenfant@interstarinc.com> for suggesting this. - - Eric mentions that doesnt show up the BUG 1020 problem. But - local testing here doesnt tell so. I still see some core dumps - with a test, albeit slowly! The patch cannot do any harm and - hence this is checked in. - -Thu Sep 26 07:16:56 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/Local_Name_Space.h (ACE_NS_String): Removed the comment - about ACE_NS_String not having a destructor since this seems to - have been overtaken by events. Thanks to Vladislav Zverev - <vladislav@gtss.spb.ru> for reporting this inconsistency. - -Thu Sep 26 09:18:55 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/Logging_Strategy.h: - Improved doxygen documentation. - -Thu Sep 26 05:42:33 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * include/makeinclude/platform_gnuwin32_common.GNU: - Since mingw can now handle exceptions across dll boundries - removed that restriction. Thanks for Vince Mounts <vince@mounts.cc> - for reporting this. - -Wed Sep 25 17:58:43 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/Basic_Stats.h: - * ace/Connector.h: - * ace/INET_Addr.h: - * ace/OS.h: - * ace/OS.i: - * ace/Service_Config.h: - * ace/Stats.h: - * ace/Timer_Queue_T.cpp: - * ace/TLI_Connector.cpp: - * ace/XTI_ATM_Mcast.i: - * ace/RMCast/RMCast_Reliable_Factory.h: - Small fixes in method argument names, types, comments to fix doxygen - warnings and improve the generated documentation. - -Wed Sep 25 08:44:13 2002 Ossama Othman <ossama@uci.edu> - - * tests/Notify_Performance_Test.cpp: - - Do not explicitly instantiate an ACE_Atomic_Op<ACE_Thread_Mutex, - long> template if ACE_HAS_BUILTIN_ATOMIC_OP is defined. - Defining it causes template specialization with the same - parameters to be declared. Fixes a compile-time problem in - explicit template instantiation builds. - -Wed Sep 25 14:49:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * bin/fuzz.pl: - Extended the types of files that are tested in the doxygen - @file with the files_idl. Some of the idl files are parsed to - generate doxygen documentation and when then a @file tag - exists the filename must be correct. - -Tue Sep 24 20:17:50 UTC 2002 Don Hinton <dhinton@objectsciences.com> - - * tests/Process_Manager_Test.cpp: Removed the ACE_HAS_THREADS - requirement and the unneeded call to - ACE_Reactor::instance ()->owner () so that it would run cleanly - on single threaded builds. Thanks to Steve Huston for this - suggestion. - - * tests/run_test.lst: Added !STATIC to Framework_Component_Test since - it must load a shared library as part of the test. - -Tue Sep 24 16:07:21 2002 Steve Huston <shuston@riverace.com> - - * examples/C++NPv1/Logging_Client.cpp: - * examples/C++NPv2/Select_Reactor_Logging_Server.cpp: - * examples/C++NPv2/Server_Shutdown.cpp: - * examples/C++NPv2/TP_Reactor_Logging_Server.cpp: - For Windows builds without ACE_HAS_STANDARD_CPP_LIBRARY, use - gets() and a char[] instead of the standard string and getline - functions. We don't want to be going this way for ACE in general, - but this fixes the legacy-mode builds. - -Tue Sep 24 06:30:31 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/ACE.h: - * ace/Acceptor.h: - * ace/Activation_Queue.h: - * ace/Asynch_Acceptor.h: - * ace/Asynch_Connector.h: - * ace/Arg_Shifter.h: - * ace/Get_Opt.h: - * ace/Local_Tokens.h: - * ace/Method_Request.h: - * ace/Thread_Manager.h: - * ace/WIN32_Asynch_IO.h: - * ace/INET_Addr.h: - Corrected doxygen tags. - - * ace/CDR_Stream.cpp (write_string): - * ace/Map_T.i (find, unbind): - * ace/Timer_Hash_T.cpp: - Made arguments in method definition and implementation the - same. - - * ace/Future.h: - Improved doxygen documentation. - - * ace/POSIX_CB_Proactor.i: - Added same ifdef as in the header file around the code in this - file, so that doxygen doesn't parse it. - -Mon Sep 23 13:43:11 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/WorkspaceCreator.pm: - - Fixed a bug with the per-project workspaces. - -Mon Sep 23 13:00:13 2002 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * ace/Log_Msg_NT_Event_Log.{h,cpp}: Made the #ifdefs consistent - in both the *.h and *.cpp files. Thanks to Johnny Willemsen for - reporting this. - -Mon Sep 23 13:14:28 2002 Steve Huston <shuston@riverace.com> - - * ace/Task.h (wait()): - * ace/Thread_Manager.h (wait_task()): Improved documentation, also - clarifying that the waited-for threads have been joined upon return. - -Mon Sep 23 12:05:38 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/BorlandWorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/GNUWorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/NMakeWorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/WorkspaceCreator.pm: - - For GNU, NMake and Borland projects, create a workspace per - project file. - - * bin/MakeProjectCreator/modules/BorlandProjectCreator.pm: - * bin/MakeProjectCreator/modules/Creator.pm: - * bin/MakeProjectCreator/modules/GHSWorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/TemplateParser.pm: - - Remove trailing whitespace. - -Sun Sep 22 08:49:41 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/Svc_Handler.cpp (flush_i): Don't flush the queued messages - if the connection fails. Thanks to Andrew Finnell - <andrew@activesol.net> for this fix. - -Sat Sep 21 23:05:56 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * bin/subst_env.pl: Improved conversion to different coding - styles. Thanks to Oliver Kellog for suggesting this. - -Fri Sep 20 13:52:23 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/NMakeProjectCreator.pm: - - Fixed a problem with implicit idl dependencies. - - * bin/MakeProjectCreator/README: - * bin/MakeProjectCreator/modules/Creator.pm: - * bin/MakeProjectCreator/modules/ProjectCreator.pm: - * bin/MakeProjectCreator/modules/TemplateParser.pm: - * bin/MakeProjectCreator/templates/bor.mpd: - * bin/MakeProjectCreator/templates/em3vcp.mpd: - * bin/MakeProjectCreator/templates/gnu.mpd: - * bin/MakeProjectCreator/templates/nmake.mpd: - * bin/MakeProjectCreator/templates/vc6dsp.mpd: - * bin/MakeProjectCreator/templates/vc7.mpd: - - Added the ability to override or modify idlflags on a per file - group basis. - -Thu Sep 19 22:13:13 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ACEXML/common/HttpCharStream.cpp: Added license for the FSM in - HttpCharStream.cpp. It is a shortened BSD license which is - compatible with ACE/TAO licensing. - -Wed Sep 18 12:26:22 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * examples/Misc/test_trace.cpp: The call to activate() was - being invoked with the parameters in the wrong order. Thanks to - Jeff W <jeffw_1234@yahoo.com> for reporting this. - -Tue Sep 17 17:26:40 2002 Ossama Othman <ossama@uci.edu> - - * ace/Basic_Types.cpp: - - Cosmetic changes to clarify some preprocessor blocks. - - * ace/CDR_Base.h: - - Corrected placement of '#' character in preprocessor - directives. It belongs in the first column, not arbitrary - ones. - -Tue Sep 17 01:35:54 UTC 2002 Don Hinton <dhinton@ieee.org> - - * ace/OS.h: Remove redundant MAXPATHLEN definition when - building VxWorks. Thanks to Rob Andzik <andzik@rtlogic.com> - for pointing this out. - -Mon Sep 16 17:36:33 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/Hash_Cache_Map_Manager_T.cpp: - Use typedefs from header file. - - * ace/Process.i (setreugid): - Made method implementation same as definition - -Mon Sep 16 15:26:56 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/config-doxygen.h: - Added define ACE_HAS_TEMPLATE_TYPEDEFS to generate - ACE_Event_Handler_T documentation. - -Mon Sep 16 07:35:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/config-doxygen.h: - Added define ACE_HAS_EVENT_POLL to generate Dev_Poll reactor - documentation - - * ace/Acceptor.h: - * ace/Containers_T.{h,cpp}: - * ace/Dev_Poll_Reactor.h: - * ace/Sample_History.h: - * ace/Service_Config.h: - Improved doxygen documentation. - -Sat Sep 14 10:34:34 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/config-freebsd-pthread.h: Added - #undef ACE_LACKS_UCONTEXT_H if we're compiling for FreeBSD - version > 4. Thanks to Christophe Juniet - <cjuniet@entreview.com> for reporting this. - -Fri Sep 13 08:37:09 2002 Ossama Othman <ossama@uci.edu> - - * ace/Time_Value.h: - * ace/Time_Value.inl: - - Removed copy constructor. The default compiler-generated copy - constructor will do fine since ACE_Time_Value is a POD class. - This may also avoid potential confusion due to a missing - explicit assignment operator. The compiler-generated assignment - operator will suffice in this case, too. - - From Gonzalo Diethelm <gonzalo.diethelm@aditiva.com> - * bin/.cvsignore: - * tests/.cvsignore: - * tests/log/.cvsignore: - - Files to silence annoying CVS messages regarding generated files - it doesn't know about. - -Fri Sep 13 08:57:03 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/USAGE: - * bin/MakeProjectCreator/modules/Creator.pm: - * bin/MakeProjectCreator/modules/Driver.pm: - * bin/MakeProjectCreator/modules/ProjectCreator.pm: - * bin/MakeProjectCreator/modules/TemplateParser.pm: - * bin/MakeProjectCreator/modules/WorkspaceCreator.pm: - * bin/MakeProjectCreator/templates/bor.mpd: - * bin/MakeProjectCreator/templates/em3vcp.mpd: - * bin/MakeProjectCreator/templates/gnu.mpd: - * bin/MakeProjectCreator/templates/nmake.mpd: - * bin/MakeProjectCreator/templates/vc6dsp.mpd: - * bin/MakeProjectCreator/templates/vc7.mpd: - - Added the ability to modify template and project variable - assignments from the command line. Also added rules to process - idl files and place the output in the directory relative to the - specified idl file. - - Modified to not add duplicates to file lists. - -Thu Sep 12 17:18:43 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/Synch.{h,cpp}: Only define the recursive condition support if - we're running on Win32. Hopefully, this will be generalized for - VxWorks at some point. Thanks to Rob Andzik - <andzik@rtlogic.com> for reporting this. - -Thu Sep 12 19:29:41 2002 Steve Huston <shuston@riverace.com> - - * ace/Atomic_Op_T.i (dump): Don't try to return a value from a - 'void' method. - -Thu Sep 12 17:38:10 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/Malloc_T.h: - * ace/Malloc_T.i (release): Added an extra argument (with default - value) to release so that we could decrease the refcount and - release the backing store if needed atomically (while holding the - lock.) This was causing a race condition when removing mmap - files used by MEM_Stream's. - - * ace/Memory_Pool.h: - * ace/Memory_Pool.i: - * ace/Memory_Pool.cpp: Changed the release method to take an extra - argument so that we can control whether we want to perform - cleanup on backing store or not. - - * ace/MEM_SAP.cpp (close_shm_malloc): Changed to use the extended - release method to remove the shared_memory. - - * ace/MEM_IO.cpp: When sending/receiving data reactively, make - sure we don't even try if the socket is not valid. - -Thu Sep 12 08:39:46 2002 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * ace/Synch.h (class ACE_recursive_mutex_state): Changed "HANDLE" to - "ACE_HANDLE" to avoid problems with VxWorks. Thanks to Rob - Andzik <andzik@rtlogic.com> and Don Hinton <dhinton@ieee.org> - for reporting this. - -Thu Sep 12 17:36:36 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * include/makeinclude/wrapper_macros.GNU: Use awk instead of - perl to determine the ACE major, minor, and beta versions. - -Thu Sep 12 12:02:31 2002 Steve Huston <shuston@riverace.com> - - * tests/TSS_Test.cpp: Pass iteration count to worker() as a pointer - to int, not an int cast to a pointer and back. - -Wed Sep 11 08:44:28 2002 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * Our thoughts and prayers go out to the victims of 9/11/01 on the - anniversary of this terrible tragedy. - -Wed Sep 11 13:03:48 2002 Carlos O'Ryan <coryan@atdesk.com> - - * ace/Select_Reactor_Base.cpp: - Check the suspend mask before completely forgetting about a - handle, otherwise things do not work correctly with TP_Reactors - and multiply-registered handlers. This fixes: - http://deuce.doc.wustl.edu/bugzilla/show_bug.cgi?id=1306 - -Tue Sep 10 17:46:51 2002 Steve Huston <shuston@riverace.com> - - * ace/Malloc_T.{h i cpp} (ACE_Malloc_T): Changed the lock_ member from - an ACE_LOCK to an ACE_LOCK *. Also added another constructor that - accepts an ACE_LOCK pointer instead of a ACE_TCHAR *lock_name. - This allows someone to form an ACE_LOCK externally and supply - it to the ACE_Malloc_T object. This is useful if special - permissions need to be set on the lock, such as the mode_t for - ACE_Process_Mutex. - -Tue Sep 10 14:06:02 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ace/Obstack_T.cpp: - * ace/Obstack_T.h: - - Obstack is no longer restricted to fixed-size chunks. Chunk size - is increased by a power of two if we run out of memory in a - chunk. Added implementation of unwind operations. Stack can be - unwound by giving the address of an object on the stack. Fixed a - couple of bugs in the existing implementation when used with - WCHAR. We were incrementing the pointers by sizeof (CHAR) as - well as multiplying by sizeof (CHAR) to get the number of bytes - to be copied. This improvement doesn't deal with the alignment - issues as it requires a change in Obchunk which is postponed - till after ACE 5.3 is released. - - * tests/Obstack_Test.cpp: - - New, improved test which exercises all the code paths in the - implementation unlike the previous test. - -Tue Sep 10 12:32:33 2002 Carlos O'Ryan <coryan@atdesk.com> - - * ace/Connector.cpp: - Be more informative in error messages, at least report in what - library, class and function was the error detected. This fixes: - http://deuce.doc.wustl.edu/bugzilla/show_bug.cgi?id=1304 - -Tue Sep 10 12:28:41 2002 Carlos O'Ryan <coryan@atdesk.com> - - * ace/Connector.cpp: - Be more paranoid about using AST's returned from cleanup_AST() - and always check the return value of that function. This fixes: - http://deuce.doc.wustl.edu/bugzilla/show_bug.cgi?id=1303 - -Mon Sep 9 16:44:00 2002 Balachandran Natarajan <bala@isis-server.vuse.vanderbilt.edu> - - * THANKS: Surprise, surprise, Gautam Thaker'ss name was not there - in the THANKS file. So added it. - -Mon Sep 9 16:57:41 2002 Steve Huston <shuston@riverace.com> - - * ace/POSIX_Asynch_IO.cpp (cancel): Don't sanity-check the - proactor type - just call cancel_aio and let the proactor impl - figure it out. - - * examples/Reactor/Proactor/test_aiocb.cpp: Fixes to not crash. - -Mon Sep 09 11:37:13 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/OS.cpp (cond_timedwait): Changed to pass absolute time when - using emulated semaphore on WinCE. Thanks to Herbert Wang - <herbert@nowsol.com> for reporting this. - - * ace/Thread_Manager.cpp (join): This function was not cropping - the thread return status when a thread had already exited when - we called and its Thread_Descriptor had been moved to - terminated_thr_list_. Thanks to Roger Weeks - <rogerw@firstlogic.com> for reporting the problem. - -Mon Sep 9 00:12:04 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * ace/Makefile.dirs: - * examples/Makefile: - * tests/Makefile: - Remove references to obsolete ACE_HAS_GNUG_PRE_2_8 make - variable. - -Sat Sep 7 23:22:20 2002 Balachandran Natarajan <bala@isis-server.vuse.vanderbilt.edu> - - * bin/make_release: The tar ball that is put out during the - release, is copied now to the the previous_versions directory - with the appopriate version information. This should work right - since this was tested against a local repo. - - Thanks to Preston A. Elder <prez@gentoo.org> for motivating - this change. - -Sat Sep 7 18:03:21 2002 Steve Huston <shuston@riverace.com> - - * ace/Proactor.cpp: Removed some ACE_Proactor:: prefixes on member - variable access, replacing with this-> to clarify what's happening. - - * ace/SUN_Proactor.cpp (handle_events): No need to cast a pointer to - long to check for 0. - Re-enabled some "should never happen" ACE_DEBUG lines until we - find out what's going wrong with this stuff. - -Thu Sep 5 12:13:51 2002 Steve Huston <shuston@riverace.com> - - * ACEXML/common/URL_Addr.inl: Removed ACE_RCSID - it's in URL_Addr.cpp. - -Wed Sep 4 16:31:29 2002 Douglas C. Schmidt <schmidt@cs.wustl.edu> - - * Kelly Clarkson wins "American Idol." Go Kelly!! - -Wed Sep 4 17:50:27 2002 Steve Huston <shuston@riverace.com> - - * tests/run_test.pl: Change run time limit from 300 to 400 seconds. - Some tests are timing out, and when run by hand, they run just over - 5 minutes, so give them a little more time. - -Wed Sep 4 17:31:24 2002 Steve Huston <shuston@riverace.com> - - * tests/Cached_Conn_Test.cpp: ACE_DEBUG %@ for pointers, not %d. - -Wed Sep 4 20:59:40 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * include/makeinclude/platform_irix5.2.GNU: - * include/makeinclude/platform_m88k.GNU: - * include/makeinclude/platform_netbsd.GNU: - * include/makeinclude/platform_qnx_neutrino.GNU: - * include/makeinclude/platform_qnx_rtp_gcc.GNU: - * include/makeinclude/platform_sco-nothread.GNU: - Instead of setting -fno-implicit-templates, - use $(TEMPLATES_FLAG). Include platform_g++.GNU to get the - correct value of $(TEMPLATES_FLAG). - -Wed Sep 4 16:31:29 2002 Steve Huston <shuston@riverace.com> - - * ace/Timer_Wheel_T.cpp: Formatting tweaks for ACE coding style. - -Wed Sep 4 20:08:54 UTC 2002 Don Hinton <dhinton@objectsciences.com> - - * ace/Framework_Component.cpp (close): - - Delete each component explicitely instead of calling - close_singleton () to avoid a memory leak. The dtor calls - close_singleton () anyway. Thanks Vladimir Chovanec - <Vladimir CHOVANEC@asset.sk> for reporting the leak and - suggesting the fix. - -Tue Sep 03 22:59:27 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ACEXML/common/XML_Common.dsp: - - Added URL_Addr.inl to common. - -Tue Sep 03 22:40:22 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ACEXML/common/HttpCharStream.cpp: - - Handle WCHAR properly when getting data from URLs. - - * ACEXML/common/URL_Addr.cpp: - * ACEXML/common/URL_Addr.h: - * ACEXML/common/URL_Addr.inl: - - Fixes for proper parsing of URLs when URLs are WCHARs. - - * ACEXML/examples/SAXPrint/main.cpp: - - Use int in GetOpt instead of char or ACEXML_Char. - -Tue Sep 3 17:12:59 2002 Ossama Othman <ossama@uci.edu> - - * ace/OS.cpp (num_processors_online): - - Fixed ACE_TRACE macro. - -Tue Sep 3 16:38:47 2002 Ossama Othman <ossama@uci.edu> - - * ace/ace_dll.dsp: - * ace/ace_lib.dsp: - - Atomic_Op.cpp is no longer a template source file. Moved it to - the "Source Files" folder. - - * ace/Synch.cpp (ACE_Mutex): - - Fixed "unused 'mode' argument" warning. - -Tue Sep 3 15:54:33 2002 Christopher Kohlhoff <chris@kohlhoff.com> - - * ace/OS.h: - * ace/OS.cpp: - - Added new functions ACE_OS::num_processors() and - ACE_OS::num_processors_online() for determining the number of CPUs - available on a system. - - * ace/Atomic_Op.cpp: - * ace/Atomic_Op.h: - * ace/Atomic_Op.i: - * ace/Atomic_Op_T.cpp: - * ace/Atomic_Op_T.h: - * ace/Atomic_Op_T.i: - * ace/Object_Manager.cpp: - * ace/Makefile.bor: - * ace/Makefile.ace: - * ace/ace.mpc: - * tests/Atomic_Op_Test.cpp: - - Added a new fast, lightweight ACE_Atomic_Op template specialization - for x86 platforms that uses the architecture's atomic integer - primitives. It currently supports Win32 using MSVC6 or BCB, and g++ - for (in theory) all x86 operating systems. It selects the - appropriate implementation for single- or multi-CPU systems at - runtime, based on the return value of the ACE_OS::num_processors - function added above. - - Note that Atomic_Op.cpp is now a real .cpp file and not a template - file, and so project files need to be updated accordingly. The - template definitions have been moved to the new Atomic_Op_T.* files. - - Note also that the ACE_Atomic_Op::mutex() operation has been - deprecated (and is not supported for the new specialization). If - you need this functionality, consider using the ACE_Atomic_Op_Ex - template instead. - -Tue Sep 3 15:40:39 2002 Ossama Othman <ossama@uci.edu> - - * tests/XtReactor_Test.cpp (sock_callback): - - Removed this unused static function. - - * tests/TkReactor_Test.cpp (sock_callback): - - Likewise. - - (ACE_TMAIN): - - Fixed "deprecated conversion from string constant to char *" - warning. - -Tue Sep 3 15:16:03 2002 Balachandran Natarajan <bala@isis-server.vuse.vanderbilt.edu> - - * bin/auto_run_tests.lst: Fixed a typo! - -Mon Sep 2 16:48:29 2002 Steve Huston <shuston@riverace.com> - - * tests/OS_Test.cpp (ctime_r_test): Corrected ACE_Log_Msg to use - %P, not %p, to list process ID. - -Mon Sep 2 13:17:26 2002 Steve Huston <shuston@riverace.com> - - * ace/Memory_Pool.{h cpp}: - (ACE_MMAP_Memory_Pool_Options): Added optional file_mode arg - to set the protection mode on the mapped file, if it is - created; defaults to ACE_DEFAULT_FILE_PERMS. - (ACE_MMAP_Memory_Pool): Added file_mode_ member that can be set - by the ACE_MMAP_Memory_Pool_Options. Defaults to - ACE_DEFAULT_FILE_PERMS. It's used for call to ACE_Mem_Map::open() - and ACE_Mem_Map::map(). - - * ace/Process_Mutex.{h cpp}: ACE_Process_Mutex ctor takes an optional - mode_t arg to set the backing store mode (for ACE_Mutex use) or the - mode for the System V semaphore, when it's used. - - * ace/Synch.{h cpp} (ACE_Mutex): Constructor takes an optional mode - to set backing store protection for Pthreads process mutex case, - where memory must be allocated to hold the mutex. - - These are laying the groundwork for fixing Bugzilla #1219. - -Mon Sep 2 14:07:12 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ace/Exception_Macros.h: - - New file which holds the exception macros previously present in - CORBA_macros.h. This file also makes the macros a bit generic so - that they can be used in other modules than TAO. - - * ace/CORBA_macros.h: - - Include only the definitions needed for TAO/CORBA. This file can - be placed under $TAO_ROOT now. Leaving it for compatibility - reasons. - - * ace/OS.i: - - Changed the definition of ctime_buf_size to const so that gcc - doesn't complain about using variable-sized array which is - non-ISO C++. - -Mon Sep 2 10:00:40 2002 Balachandran Natarajan <bala@isis-server.vuse.vanderbilt.edu> - - * ace/Timer_List_T.cpp (get_first_i): Fixes fuzz errors. - -Sun Sep 01 08:18:37 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/Task.cpp (svc_run): - The latest version of the Borland compiler doesn't like the - static cast, so only do the static cast when the compiler - version is smaller than 0x570 - - * ace/Handle_Set.cpp: - When using Borland on Linux, we don't have fds_bits, but we have - __fds_bits. - -Sun Sep 01 07:57:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/config-win32-borland.h: - * ace/config-borland-common.h: - Improved Borland config files for Kylix support. - -Sun Sep 01 07:25:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * include/makeinclude/platform_linux_borland.GNU: - Added first version of platform .GNU file for the Borland Kylix - compiler. Thanks to Newton Aird <naird@ix.netcom.com> for - providing this first version. - - * THANKS: Added Newton Aird <naird@ix.netcom.com> - -Sat Aug 31 14:57:53 2002 Balachandran Natarajan <bala@isis-server.vuse.vanderbilt.edu> - - * THANKS: Added Cemal Yilmaz <cyilmaz@cs.umd.edu> to the hall of - fame. - -Sat Aug 31 13:55:52 2002 Balachandran Natarajan <bala@isis-server.vuse.vanderbilt.edu> - - * ace/OS.i (ctime_r): This fixes a problem on Linux for the call - to ::ctime_r where the pointer returned (result) is the same as - the pointer (buf) passed in, causing the buffer to get clobbed - if strsncpy is used subsequently to move the result back to the - caller's buffer. - - * tests/OS_Test.cpp: Added a test for ACE_OS::ctime_r (). - - Thanks to Michael Searles <msearles@base16.com> for providing - this patch. - -Sat Aug 31 17:45:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/config-borland-common.h: - Created a new config file for the Borland compiler. This will be - used on the Win32 platform and the Kylix3 compiler on Linux - - * ace/config-linux-common.h: - When the Borland compiler is detected, include the - config-borland-common.h file - - * ace/config-win32-borland.h: - Moved most of the contents to the config-borland-common.h file, - only the lines that are Win32 specific stay here. - -Sat Aug 31 10:32:38 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * ace/Future.cpp (set): Removed a stray reference to "iterator" - and replaced the for loop with a while loop. Thanks to Craig - Rodrigues for reporting this. - -Sat Aug 31 01:20:09 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - Gary Duzan <gduzan@bbn.com> - - * include/makeinclude/rules.nested.GNU: Rewritten by Gary Duzan. - If a list of directories is specified with DIRS= for a recursive - make, make will halt immediately and propagate a non-zero - return code to the shell if an error is encountered in a - subdirectory. The present behavior does not propagate any - errors to make if there is an error in a subdirectory. If make -k - is specified, make will keep on going even if it encounters an - error in a subdirectory, but it will also propagate a non-zero - return code to make when the make finishes. The current behavior - is to keep on going if an error is encountered in a subdirectory, - but NOT propagate a non-zero return code upon termination of make. - - * ace/Makefile.dirs: (Added) Specifies the DIRS= list of directories - to invoke recursive makes on. - - * ace/Makefile.ace: (Added). This builds libACE, like the old Makefile - did. - - * ace/Makefile: Invoke Makefile.ace first, then Makefile,dirs. It - was necessary to split apart the DIRS= rule from the part of the - Makefile which builds libACE in order to have parallel makes (-j 2) - work on recursive makes. - -Fri Aug 30 17:34:00 2002 Justin Michel <michel_j@ociweb.com> - - * ace/Timer_List_T.cpp: - * ace/Timer_List_T.h: - * ace/Timer_Wheel_T.cpp: - * tests/Timer_Queue_Test.cpp: - - Reimplemented timer list to be compatible with Itanium. It no longer - uses a Node* as the timer_id, so the cancel operation is now much - more expensive. On the plus side I was able to make a few optimizations - that improve the speed of the schedule operations. - I also made a small optimization to the timer wheel, and updated - the test slightly. - -Fri Aug 30 15:53:18 2002 Steve Huston <shuston@riverace.com> - - * tests/SSL/Makefile: Move the addition of SSL options/flags up to - before the hackery of the compile command for AIX. - -Fri Aug 30 05:48:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/OS.i (tempnam): - Changed defines because special things only happen on Win32 - -Thu Aug 29 22:01:27 2002 Steve Huston <shuston@riverace.com> - - * include/makeinclude/platform_hpux_gcc.GNU: Make the build options - settable by the command line (and/or platform_macros.GNU), and - fix the threads and soname options, as well we remove the - -fstrict-prototype option from the compile options - it was needed - for an older g++ (nothing prior to g++ 3 works right on HP-UX - anyway) and is not valid for g++ 3.2 and later. Thanks to - Ville Lehtiniemi <guderian@ees2.oulu.fi> for reporting these - issues and sending fixes. - -Thu Aug 29 15:18:31 2002 Steve Huston <shuston@riverace.com> - - * ace/Get_Opt.{h i cpp} (long_option): Only print error messages for - long message addition problems if opterr is set. Also added - ACE_LIB_TEXT to naked strings for ACE_ERROR. - - Added opt_opt() method to return the most recently matched - short option character. Thanks to Rich (Mr. POSIX) Seibel - for the idea and POSIX spec info for this new method. - - * ace/Object_Manager.cpp (init): When finishing up initialization of - the object manager, call ACE_Trace::start_tracing() to reenable - trace in cases where a user does init/fini/init. - - * tests/Get_Opt_Test.cpp: Added usage of ACE_Get_Opt::opt_opt(). - -Thu Aug 29 15:03:23 2002 Ossama Othman <ossama@uci.edu> - - * ace/README: - - Documented ACE_HAS_BROKEN_UNEXPECTED_EXCEPTIONS macro. - - * ace/config-win32-msvc-6.h - (ACE_HAS_BROKEN_UNEXPECTED_EXCEPTIONS): - - MSVC++ 6 does not correctly handle exceptions thrown from - functions/methods that do not list them in the exception - specification. In particular, the unexpected exception handler - is not invoked. - -Thu Aug 29 06:12:56 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * ace/Future.cpp (set): Fixed the loop code so it's possible for - the observer object to be deleted or detached in its update() - operation. Thanks to Tomer Amiaz <Tomer@bandwiz.com> for this - fix. - -Thu Aug 29 09:40:20 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ACEXML/common/Transcode.h: - * ACEXML/parser/parser/Parser.h: Fixed incorrect usage of Doxygen - tag. Thanks to Johnny for pointing this out. - -Thu Aug 29 12:20:17 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/config-win32-borland.h: - Removed the 1 after 2 defines, so that they are like the defines - in other files. - -Thu Aug 29 05:53:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * examples/Reactor/Proactor/post_completions.cpp: - Fixed compile error. - - * ace/OS_Dirent.h: - * ace/OS_Errno.cpp: - * ace/Time_Value.h: - * ace/Time_Value.inl: - Made method signature the same in definition and implementation. - - * ace/TkReactor.cpp: - * ace/TkReactor.h: - Made method argument names the same as in the base so that doxygen - can generate the documentation. - -Wed Aug 28 23:30:31 2002 Steve Huston <shuston@riverace.com> - - * include/makeinclude/wrapper_macros.GNU: Remove addition of SSL - compile/link options so that not all parts of ACE, needing SSL or - not, get these options. - - * ace/SSL/Makefile: - * tests/SSL/Makefile: Added the SSL compile/link options from - wrapper_macros.GNU to these files, where they're needed. - - * ace/POSIX_Proactor.cpp (post_wakeup_completions): Use int - to count completions, not ssize_t - match the passed-in arg type. - Post the completion directly via the proactor, not via the - wakeup completion; the dynamic_cast in ACE_POSIX_Asynch_IO::post_ - completion loses the subclass of ACE_POSIX_AIOCB_Proactor and - when using any other subtype of proactor, the notification is - queued incorrectly. - - (POSIX_SIG_Proactor::notify_completion()) - always queue the signal. - Not doing so leaves a window where MT proactors can lose a signal. - - * tests/Proactor_Test.cpp: Use ACE_Proactor::close_singleton() to - delete proactor instance. Just to be sure we try it. - - * ace/SSL/SSL_Asynch_Stream.{h cpp}: u_long->size_t changes. - -Wed Aug 28 22:42:41 2002 Steve Huston <shuston@riverace.com> - - * examples/Service_Configurator/Misc/main.cpp: - * netsvcs/clients/Naming/Client/main.cpp: - * tests/ARGV_Test.cpp: - * tests/Get_Opt_Test.cpp: Change size_t to int as a side-affect of: - Tue Aug 27 13:28:17 2002 Steve Huston <shuston@riverace.com> - * ace/ARGV.{h i cpp}: Changed argc() value from size_t to int. It - needs to get passed to many places that expect int (argc, argv). - - * tests/Semaphore_Test.cpp: - * tests/TP_Reactor_Test.{cpp h}: - * tests/Upgradable_RW_Test.cpp: Further int/size_t corrections. - -Wed Aug 28 19:56:16 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * docs/ACE-subsets.html: Updated the document to point to the - soreduce tool and to remove the discussion of PACE. Thanks to - Chumsu Kim <chskim@add.re.kr> for motivating this. - -Wed Aug 28 14:24:36 2002 Steve Huston <shuston@riverace.com> - - * ace/Log_Msg.h: Fixed Doxygenation of ACE_Log_Msg. Thanks to - Philip Miller <pwmiller@sarnoff.com> for reporting this. - -Wed Aug 28 11:21:04 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/templates/em3vcpdll.mpt: - * bin/MakeProjectCreator/templates/em3vcpdllexe.mpt: - * bin/MakeProjectCreator/templates/em3vcplib.mpt: - * bin/MakeProjectCreator/templates/em3vcplibexe.mpt: - * bin/MakeProjectCreator/templates/nmakedll.mpt: - * bin/MakeProjectCreator/templates/nmakeexe.mpt: - * bin/MakeProjectCreator/templates/vc6dspdll.mpt: - * bin/MakeProjectCreator/templates/vc6dspdllexe.mpt: - * bin/MakeProjectCreator/templates/vc6dsplib.mpt: - * bin/MakeProjectCreator/templates/vc6dsplibexe.mpt: - - Enabled RTTI for all VC based template input files. - -Wed Aug 28 06:18:35 2002 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * ace: Removed the config-sunos5.8-sunc++-5.4.h file since these - changes should go in the config-sunos5.8.h file rather than - a special file. Thanks to Steve Huston for noticing this. - - * include/makeincludes: Removed the platform_sunos5_sunc++5.4.GNU - file since it's no different from the platform_sunos5_sunc++.GNU - file. Thanks to Steve Huston for noticing this. - -Wed Aug 28 06:12:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * etc/tao.doxygen: - Removed Services.h of the list of not to parse files - - * etc/tao_rtportableserver.doxygen: - Added TAO_NAMESPACE to list of to expanded macros - - * tests/Timer_Queue_Test.cpp: - Fixed unicode build error - - * ace/Timer_Wheel_T.cpp: - Fixed doxygen tags. - -Tue Aug 27 19:09:17 2002 Steve Huston <shuston@riverace.com> - - * ace/Get_Opt.h: Documentation improvements. - -Tue Aug 27 17:00:20 2002 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * ace/Signal.cpp: Fixed the ACE_Sig_Action constructors so that - it doesn't break out of the loop just because a signal handler - failed. Thanks to Marcel Loose <loose@astron.nl> for reporting - this. - -Tue Aug 27 19:03:05 2002 Balachandran Natarajan <bala@isis-server.vuse.vanderbilt.edu> - - * bin/auto_run_tests.lst: Added $TAO_ROOT/tests/ORB_shutdown test - to the daily builds. - -Tue Aug 27 18:52:59 2002 Balachandran Natarajan <bala@isis-server.vuse.vanderbilt.edu> - - * ace/Select_Reactor_T.cpp (wait_for_multiple_events): Normally, - select() will reset the bits in dispatch_set so that only those - filed descriptors that are ready will have bits set. However, - when an error occurs, the bit set remains as it was when the - select call was first made. Thus, we now have a dispatch_set - that has every file descriptor that was originally waited for, - which is not correct. We must clear all the bit sets because we - have no idea if any of the file descriptors is ready. - - Thanks to Jody Hagins <jody@atdesk.com> and Carlos O'Ryan - <coryan@atdesk.com> for reporting this problem and proving the - patches. - -Tue Aug 27 15:56:10 2002 Ossama Othman <ossama@uci.edu> - - * ace/config-sunos5.8-sunc++-5.4.h: - * include/makeinclude/platform_sunos5_sunc++5.4.GNU: - - Solaris 8 + SunC++ SPRO 5.4 platform configuration files. - Thanks to Serge Kolgan <skolgan@objectsciences.com> for - contributing these files. - -Tue Aug 27 17:25:39 2002 Steve Huston <shuston@riverace.com> - - * ace/WIN32_Proactor.{h cpp}: More u_long->size_t corrections. - -Tue Aug 27 16:50:41 2002 Steve Huston <shuston@riverace.com> - - * tests/Proactor_Test.cpp: Fixed mismatched printf formats for - size_t arguments. - -Tue Aug 27 16:32:18 2002 Steve Huston <shuston@riverace.com> - - * ace/Proactor_Impl.h: Fixed some u_long->size_t changes missed - on POSIX platforms. - -Tue Aug 27 14:44:00 2002 Justin Michel <michel_j@ociweb.com> - - * ace/Timer_Wheel_T.cpp: - - Fixed problem with incorrect name in ACE_TRACE macro. - -Tue Aug 27 13:28:17 2002 Steve Huston <shuston@riverace.com> - - * ace/config-win32-common.h: Added #define ACE_MALLOC_ALIGN 8 for - Win64. - - * ace/Malloc_T.cpp (ACE_Cached_Allocator): When dividing up the - allocated space, maintain the platform's alignment requirements. - - * ace/ace_dll64.mak: Removed references to Hash_Map_Manager_rt.*. - - * ace/Stream_Modules.cpp (info): Cast string length back to int. - - * ace/ARGV.{h i cpp}: Changed argc() value from size_t to int. It - needs to get passed to many places that expect int (argc, argv). - - * ace/OS.{h cpp} (ACE_OS::string_to_argv): argc changed from size_t& - to int& to match the change above. - - * ace/FILE_IO.{h i cpp}: - * ace/SPIPE_Stream.{h i}: Follow-up to ACE_SOCK_Stream change in: - Mon Jul 22 16:39:21 2002 Steve Huston <shuston@riverace.com> - (sendv_n, recvv_n, sendv, send, recv) - change iov count from - size_t to int. - - * ace/Functor_T.i (ACE_Pointer_Hash::operator()): Use reinterpret_cast - to go from pointer to u_long. - - * ace/config-win32-msvc-7.h: Define ACE_AUTO_PTR_LACKS_RESET for - Win64 SDK compiler. - - * ace/Strategies_T.{h i cpp} (ACE_Thread_Strategy): Changed n_threads - parameter (and n_threads_ member) from size_t to int - matches - what's expected by ACE_Task::activate(), which is where this - value is passed. - - * ace/Asynch_Acceptor.{h cpp}: - * ace/Asynch_IO.{h cpp}: - * ace/Asynch_IO_Impl.h: - * ace/POSIX_Asynch_IO.{h cpp}: - * ace/POSIX_Proactor.{h cpp}: - * ace/WIN32_Asynch_IO.{h cpp}: - * ace/WIN32_Proactor.{h cpp}: All byte-count arguments and members - in all classes changed from u_long to size_t. Some already were... - - * ace/SOCK_Dgram.i (send): address length is an int, not size_t. - - * tests/Cache_Map_Manager_Test.h (Hash_Key::operator()): Cast the - size_t argument to u_long for return. - - * tests/Cache_Map_Manager_Test.cpp: Cast arg for ACE_OS::srand to - u_int (matches ACE_OS signature), not size_t. - - * tests/CDR_Array_Test.cpp (zero): Change len arg to size_t, and - also the values passed to it. Also, correct length args to - ACE_InputCDR and ACE_OutputCDR - they're size_t, not int. - - * tests/Handle_Set_Test.cpp: Disable the conversion warnings - initializing the array of ACE_HANDLE values on Win64. They're fine. - - * tests/Mem_Map_Test.cpp (reverse_file): Changed size arg from int - to size_t. - - * tests/Buffer_Stream_Test.cpp: - * tests/CDR_Test.cpp: - * tests/Message_Block_Test.cpp: - * tests/Message_Queue_Notifications_Test.cpp: - * tests/Priority_Buffer_Test.cpp: - * tests/Proactor_Scatter_Gather_Test.cpp: - * tests/Proactor_Test.cpp: - * tests/Semaphore_Test.cpp: - * tests/Thread_Pool_Test.cpp: - * tests/TP_Reactor_Test.{h cpp}: - * tests/Upgradable_RW_Test.cpp: Align size_t/int usage. - - * tests/Message_Queue_Notifications_Test.cpp (Watermark_Test::producer) - Change hwm from size_t to ssize_t so it can go less than zero and - break the 'for' loop properly. - - * tests/MT_SOCK_Test.cpp: - * tests/SOCK_Test.cpp: Use an int to pass to select(), always 0 - for Win64. - - * tests/SOCK_Send_Recv_Test.cpp: - * tests/Thread_Pool_Reactor_Resume_Test.cpp: - * tests/Thread_Pool_Reactor_Test.cpp: - * tests/TSS_TEst.cpp: Add missing type cast. - -Tue Aug 27 11:08:54 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * ace/Pair_T.{h,i}: Added support for operator== to the ACE_Pair - can be used in maps, etc. Thanks to Gonzalo Diethelm - <gonzalo.diethelm@aditiva.com> for this addition. - -Tue Aug 27 06:45:58 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/ProjectCreator.pm: - - Fixed a problem where .i and .h files would be added even if they - didn't exist. - -Tue Aug 27 10:09:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * etc/tao_dynamicany.doxygen: - * etc/tao_dynamicinterface.doxygen: - * etc/tao_iormanip.doxygen: - * etc/tao_iortable.doxygen: - * etc/tao_portableserver.doxygen: - * etc/tao_rtcorba.doxygen : - Added TAO_NAMESPACE to the list of defines to expand. - -Mon Aug 26 17:33:54 2002 Ossama Othman <ossama@uci.edu> - - * bin/msvc_auto_compile.pl (Build_Core): - - Build the Messaging, PortableGroup and CosLoadBalancing - libraries in a specific order. CosLoadBalancing depends on the - first two. Fixes automated Win32 static builds. - -Mon Aug 26 17:14:20 2002 Ossama Othman <ossama@uci.edu> - - * tests/TkReactor_Test.cpp: - - Fixed a number of warnings. - -Mon Aug 26 18:21:34 UTC 2002 Justin Michel <michel_j@ociweb.com> - - * ace/Timer_Wheel_T.cpp: - * ace/Timer_Wheel_T.h: - * tests/Timer_Queue_Test.cpp: - - New and improved timer wheel implementation. - -Mon Aug 26 09:51:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/FlReactor.{h,cpp}: - * ace/TkReactor.{h,cpp}: - * ace/QtReactor.{h,cpp}: - Made the names of the arguments of the schedule_timer() method the - same as in the base class. That way doxygen generates the same - documentation for this overruled method as in the base. - - * ace/TLI_Connector.i: - * ace/XTI_ATM_MCast.i: - Made argument names the same as in the header file. - -Mon Aug 26 08:29:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/Connector.cpp: - Replaced all short defines with the original ones. Doxygen can't - generate the documentation because of the short defines. - -Sun Aug 25 14:15:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/config-win32-borland.h: - Certain defines should only be defined without a value. - -Sun Aug 25 12:45:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * etc/tao.doxygen: - Added TAO_NAMESPACE to the list of macro's that must be expanded - when generating doxygen documentation. - -Sat Aug 24 22:47:22 2002 Christopher Kohlhoff <chris@kohlhoff.com> - - * include/makeinclude/ace_flags.bor: - * include/makeinclude/clean.bor: - * include/makeinclude/compiler.bor: - - Updated to support MakeProjectCreator. - - * bin/MakeProjectCreator/config/core.mpb: - * bin/MakeProjectCreator/modules/ProjectCreator.pm: - * bin/MakeProjectCreator/modules/BorlandProjectCreator.pm: - * bin/MakeProjectCreator/modules/BorlandWorkspaceCreator.pm: - * bin/MakeProjectCreator/templates/bor.mpd: - - Fixed support for Borland makefile generation. - - New "core" config for projects that would typically be installed - on production system, i.e. pretty much everything except examples - and tests. - - New "include_dir" project attribute used to specify a directory - path where header files for the project should be installed. - - * ace/ace.mpc: - * ace/QoS/qos.mpc: - * ace/RMCast/rmcast.mpc: - * ace/SSL/ssl.mpc: - - Added "core" as a base project and added "include_dir" value. - - * apps/gperf/src/gperf.mpc: - - Added "core" as a base project. - -Sat Aug 24 16:41:36 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ACEXML/common/FileCharStream.cpp (get): - - Remove :: in front of the macro feof. VC is hopelessly confused. - - * ace/Argv_Type_Converter.h: - - Added #include "ace/OS_Memory.h" to get ACE_NEW. I needed to - touch this to test my changes. - -Sat Aug 24 12:01:54 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * ace/OS.h: #undef the atop macro so that it won't break the new - ACE_OS::atop() method on FreeBSD. Thanks to Russell Carter - <rcarter@pinyon.org> and Christophe Juniet - <cjuniet@entreview.com> for this report. - - * ace/config-lynxos.h: Added some #defines for LynxOS 4.0. We'll - leave these comments out for now until we figure out how to add - them without breaking earlier versions of LynxOS. Thanks to - Umberto Mascia <u.mascia@pxl.it> for reporting this. - - * docs/tutorials/015/Protocol_Stream.h (class Protocol_Stream): - Changed a comment to say "server-side applications" rather than - "client-side applications." Thanks to Mark Xu <ltxd@371.net> - for reporting this. - -Fri Aug 23 18:20:31 2002 Steve Huston <shuston@riverace.com> - - * ace/config-all.h: - * ace/OS_Memory.h: Moved the ACE_NEW macros from config-all.h to - OS_Memory.h - - * ace/Log_Msg.h: Removed #include "ace/Basic_Types.h" (not needed) - and replaced "ace/OS_Errno.h" with "ace/OS.h" since thread defs - are necessary. - - * ace/Basic_Types.h: Don't make thread definitions here - they belong - in ace/OS.h. - - * ace/Cache_Map_Manager_T.h: Added #include "ace/Default_Constants.h" - to get ACE_DEFAULT_MAP_SIZE when instantiating templates on AIX - with Visual Age C++. - - * ace/Caching_Utility_T.cpp: #include "ace/Min_Max.h" and - "ace/OS_Memory.h" so Visual Age C++ can see ACE_MAX and ACE_NEW - when instantiating templates. - - * ace/Arg_Shifter.cpp: Added #include "ace/OS_Memory.h" to get ACE_NEW. - - * netsvcs/clients/Naming/Dump_Restore/Dump_Restore.cpp: Added - #include "ace/Malloc_Base.h" to pick up ACE_Allocator. - - * tests/Cached_Allocator_Test.cpp: Changed default speed tests loops - from 100,000,000 to 10,000,000. run_test.pl times it out on - non-nuclear processors ;-) - -Fri Aug 23 16:59:51 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * bin/svcconf-convert.pl: Fixed regex to handle empty option - strings in static/dynamic entries correctly. - -Fri Aug 23 16:25:47 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ace/OS.i: - * ace/OS.h: - - Removed the wrappers for functions that are known to be - overloaded by macros. #undef will break user code for these - functions. This includes feof(), ferror() and ungetc(). The - other functions seem to only have a function implementation. - - * ACEXML/common/FileCharStream.cpp: - - Removed the use of ACE_OS wrappers for the above functions. - -Fri Aug 23 08:40:38 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * ace/config-qnx-rtp.h: Added some new #defines so ACE will - compile on QNX 6.2. Thanks to Leen Van Kampen - <leen.van.kampen@nl.vanderlande.com> for reporting this. - - * ace/Task.cpp (svc_run): Added a #ifdef for _MSC_VER <= 1200 - since there are complaints about reinterpret_cast from int to - unsigned long. Thanks to Espen Harlinn - <espen.harlinn@seamos.no> for this fix. - -Fri Aug 23 13:04:34 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/Service_Config.cpp: Some minor cosmetic changes. - -Fri Aug 23 07:16:13 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/templates/vc6dsp.mpd: - - Fixed a problem where the pch header files is not in the list of - header files. - -Fri Aug 23 06:07:38 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ace/OS.h (ACE_OS): - - ungetc() is a macro under Borland. Undefine it to make Borland - compiler happy. - -Thu Aug 22 21:27:56 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/OS_String.cpp (strtok_r_emulation): Applied the patch Frank - Kuhlman <kuhlmann@materna.de> submitted in [Bug 1272] which - resolved an overrun error. - -Thu Aug 22 21:20:01 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * include/makeinclude/rules.local.GNU (realclean.local): - - Remove older cruft when doing a realclean. Note this is a - sledgehammer (libfoo.so*). Don't blame me if you shoot yourself - in the feet, which technically shouldn't happen as long as you - have the sources to the libraries that you built. Thanks to - Ossama for pointing out that cruft was left behind. - -Thu Aug 22 19:14:44 2002 Ossama Othman <ossama@uci.edu> - - * ace/OS.i (fgetwc, ungetwc): - - These functions are supposed to return WEOF on error, not -1 as - their ACE_OS wrappers were doing. Corrects a warning about - initialization of a negative value to "wint_t", which is an - unsigned integer on some platforms (e.g. GNU libc). - -Thu Aug 15 10:43:51 2002 Steve Huston <shuston@riverace.com> - - * ace/Message_Queue_T.cpp (enqueue_tail): Moved notify() call outside - lock scope, as Sun May 5 19:14:34 2002 Douglas C. Schmidt - suggested. - -Wed Aug 14 18:14:57 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.cpp (ACE_TSS_Cleanup::remove()): Always mark the key no - longer in use, instead of simply decrementing the reference count. - Shouldn't call remove() until there are no more references to - the key. - (ACE_TSS_Cleanup::detach()): Decrement the reference count when - detaching. Previously, the count could never get to 0 and the - remove() method would never be called until object manager shutdown. - This resolves a problem where a service using - ACE_Unmanaged_TSS_Singleton would do its cleanup(), but there would - still be a reference to the key... if the destructor (in an unloaded - service DLL) was called at process shutdown, it would crash. - Thanks to Craig L. Ching <cching@mqsoftware.com> for pointing this - problem area out. - -Fri Aug 9 16:11:39 2002 Steve Huston <shuston@riverace.com> - - * ace/Message_Block.cpp: - (ACE_Data_Block::ACE_Data_Block): If the data allocation fails, - reset the cur_size_ and max_size_ to 0 to reflect the fact that - there is no legitimate memory allocated. - (ACE_Message_Block::init_i): If the data block size() that results - from allocating a new ACE_Data_Block is not what was requested, - return -1 to indicate the failure. - - * ace/Malloc_T.i (ACE_Cached_Allocator::free): Don't attempt to - return a 0 pointer to the free_list_. Thanks to Laxmikant - Bopalkar <LBopalkar@protrader.com> for this fix. - - * THANKS: Added Laxmikant Bopalkar to the Hall of Fame. - -Thu Aug 22 17:34:18 2002 Steve Huston <shuston@riverace.com> - - * ace/FIFO_Recv_Msg.{h i}: Fixed recv() in the non- - ACE_HAS_STREAM_PIPES case to throw away the part of a - message longer than the max number of bytes requested by - the caller. Filled out the doxygen comments in the header - to explain the behavior, and the differences when (not) - using ACE_HAS_STREAM_PIPES. - -Wed Aug 21 17:48:29 2002 Steve Huston <shuston@riverace.com> - - * examples/C++NPv2/TP_Logging_Server.cpp (TP_Logging_Task): - * examples/C++NPv2/display_logfile.cpp (svc): - Must use reinterpret_cast, not static_cast, to cast between - unrelated types (char *, class *). - - * examples/C++NPv2/Makefile: Added AIO_CLD to top-level build. - -Thu Aug 22 18:10:46 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * bin/nightlybuilds/builds.lst: Move RedHat_Static build - to new scoreboard. - -Thu Aug 22 01:32:46 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ACEXML/common/StreamFactory.cpp: - - Use ACE_TEXT for literals. Add destructor to avoid linker - problems. - - * ACEXML/parser/parser/Parser.cpp: - - Use ACEXML_Char instead of char* to fix compilation with WCHAR. - - * ACEXML/common/XML_Common.dsp: - - Add files StreamFactory.{h,cpp} to the project files. - -Thu Aug 22 01:16:16 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ace/OS.h: - - Fixed compilation error under Windows by undefining the macros. - -Wed Aug 21 22:22:42 2002 Ossama Othman <ossama@uci.edu> - - * bin/msvc_auto_compile.pl (Build_Core): - - Removed "LoadBalancing_Static.dsp" from the list of static - library projects to build. That MSVC++ project was for the old - load balancer, and no longer exists. - -Thu Aug 22 03:50:02 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * bin/performance_stats.sh: - * bin/footprint_stats.sh: Calculate value of DATE - inside these scripts instead of passing it in as the - third parameter. This will make it easier to make - an autobuild script around this. - -Wed Aug 21 20:40:20 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ace/OS.h: - * ace/OS.i: - - Added wrappers for fgetc, fgetwc, ferror, clearerr, feof, - ungetc, ungetwc. All these are guaranteed to not be macros. - - * ACEXML/apps/svcconf/Svcconf_Handler.cpp: - * ACEXML/apps/svcconf/Svcconf_Handler.h: - - Fixed compilation errors. - - * ACEXML/common/FileCharStream.cpp: - * ACEXML/common/HttpCharStream.cpp: - * ACEXML/common/XMLFilterImpl.cpp: - * ACEXML/examples/SAXPrint/Print_Handler.cpp: - * ACEXML/examples/SAXPrint/SAXPrint_Handler.cpp: - * ACEXML/examples/SAXPrint/main.cpp: - * ACEXML/examples/svcconf/Svcconf_Handler.cpp: - * ACEXML/parser/debug_validator/Debug_Attributes_Builder.cpp: - * ACEXML/parser/debug_validator/Debug_DTD_Manager.cpp: - * ACEXML/parser/debug_validator/Debug_Element_Builder.cpp: - * ACEXML/parser/debug_validator/Element_Tree.cpp: - * ACEXML/parser/parser/Parser.cpp: - - Use ACE_TEXT instead of ACE_LIB_TEXT. It results in reducing - vertical space and the wchar.txt document also encourages it. - -Wed Aug 21 17:02:44 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/OS_String.cpp: Changed to enabled wchar version of - strtok_r_emulation when both ACE_HAS_WCHAR and ACE_LACKS_WCSTOK - are both defined (instead of ACE_HAS_REENTRANT_FUNCTIONS.) - Thanks to Brian Appel <bappel@integ.com> for reporting this. - -Wed Aug 21 12:41:51 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/config/notifytest.mpb: - * tests/SSL/tests.mpc: - - Fixed a few problems with some of the mpc files. - - * nightly.mwc: - - Added an initial nightly build workspace file. - -Wed Aug 21 11:57:37 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * bin/msvc_auto_compile.pl: Added ACEXML/apps to the list of - compiled ACE directories. - -Wed Aug 21 11:15:11 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/NMakeWorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/TemplateParser.pm: - * bin/MakeProjectCreator/templates/nmake.mpd: - - Fixed a problem dealing with Windows, NMake and idl files. - -Wed Aug 21 11:04:09 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/Object_Manager.cpp (init): Changed to return failure (-1) if - ACE_Object_Manager fail to acquire the TSS ACE_Log_Msg instance - for whatever reason. Thanks to Jeremy Altavilla - <jeremya@artifact-entertainment.com> for suggesting the fix. - -Wed Aug 21 07:43:12 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/ATM_Addr.{h,i}: - Made the signature definition and implementation of set_selector() - and get_selector() the same. - - * ace/CDR_Stream.cpp: - Made the read_string() signature the same as in the definition - - * ace/config-doxygen.h: - Added ACE_HAS_XTI_ATM to the list of defines that must be set - when generating the doxygen documentation - - * ace/config-win32-mingw.h: - Added extern "C" around the win32api methods ace uses but which - are not defined in the mingw win32api. - - * ace/Lib_Find.cpp: - Made signature of ldfind method the same as in the definition - - * ace/Log_Record.cpp: - Made signature of print method the same as in the definition - - * ace/Multiplexor.i: - Added #ifdef 0/#endif around the contents of this file, because the - header and cpp file also do this. Doxygen parses now this .i file - and gives a lot of warnings about this file because it cannot find - the definition of the methods in this file. - - * ace/Configuration.h: - * ace/Containers_T.cpp: - * ace/Lib_Find.h: - * ace/OS.h: - * ace/RMCast/RMCast_Reliable_Factory.h: - * ace/RMCast/RMCast_UDP_Proxy.h: - Small improvements for doxygen documentation. - - * include/makeinclude/ace_flags.bor: - Added cflags and lib for TAO PSDL library - - * include/makeinclude/platform_mingw32.GNU: - MingW doesn't support the adding of the version number after - the .dll, so set SONAME and SOVERSION to empty. - - * examples/Connection/non_blocking/test_lsock_acceptor.cpp: - * examples/Connection/non_blocking/test_lsock_connector.cpp: - Added include of Log_Msg to fix compile errors with MingW - -Tue Aug 20 14:19:55 2002 Jaiganesh@doc <bugzilla@cs.wustl.edu> - - * ACE version 5.2.4 released. - -Fri Aug 16 10:25:07 2002 Ossama Othman <ossama@uci.edu> - - * include/makeinclude/ace_flags.bor: - - Corrected CosLoadBalancing library related variables. They - were still setup for use with the old load balancer prototype. - -Fri Aug 16 11:15:49 2002 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * examples/Reactor/Proactor/test_aiosig_ace.cpp: Added - #include "ace/Log_Msg.h" so things will compile. Thanks to - Victor Poznyak <VPoznyak@Webley.COM> for reporting this and - Alexander Libman <alexl@aurema.com> for suggesting the fix. - -Thu Aug 15 15:22:49 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * tests/tests.dsw: Re-added Thread_Manager_Test.dsp into this - workspace. It got accidentally removed when someone merged a - branch into the main trunk. Thanks to Ronald Berger - <ronald.berger@wpafb.af.mil> for noticing this. - -Mon Aug 12 10:17:55 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/Thread_Manager.h (ACE_Thread_Manager): Fixed a type where - "manipulated" was spelled "maniputated". Thanks to - Eric Newton <ecn@metaslash.com> for reporting this. - - * PROBLEM-REPORT-FORM: Fixed a typo where "ASCII text" was spelled - "ASCII test". Thanks to Eric Newton <ecn@metaslash.com> for - reporting this. - -Fri Aug 9 12:13:21 2002 Jaiganesh Balasubramanian <jai@doc.ece.uci.edu> - - * bin/nightlybuilds/builds.1st: - Moved the Debian_Core builds to the new scoreboard. - Thanks to CRaig Rodrigues <crodrigu@bbn.com> for the fix. - -Fri Aug 9 09:24:08 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/Time_Value.{inl,cpp}: Make sure to normalize the set() methods... - Thanks to Francois <francois@vibes.net> for reporting this. This - fixes BugID 1275. - -Thu Aug 8 23:11:30 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * include/makeinclude/rules.nested.GNU: Reverted this change "Thu - Aug 8 18:28:34 UTC 2002 Craig Rodrigues <crodrigu@bbn.com>" - since it would break our daily builds. - -Thu Aug 8 18:28:34 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * include/makeinclude/rules.nested.GNU: If we fail during a recursive - make, exit with error code 1. This makes it easier to detect errors - in automated builds. Thanks to Gary Duzan <gduzan@bbn.com> for showing - inspiring the fix. - -Wed Aug 7 12:39:31 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/Driver.pm: - - Speed up execution by changing the include path order. - - * bin/MakeProjectCreator/modules/GNUWorkspaceCreator.pm: - - Don't generate a for loop if there is only one project. - - * bin/MakeProjectCreator/modules/ProjectCreator.pm: - - Fixed a bug with the global assignments overwriting project - assignments. - -Sun Aug 4 16:47:29 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * tests/Framework_Component_DLL.cpp: Added a missing template - instantiation. It was caught only in SunCC53 builds. - -Sun Aug 4 16:01:28 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * bin/auto_run_tests.lst: Do not the Stack_Recursion tests in - single threaded builds. - -Sat Aug 03 17:57:33 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * ace/ace_dll.dsp: Removed the Hash_Map_Manager_rt.cpp from the - project file. - -Sat Aug 03 15:42:49 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * examples/Log_Msg/Log_Msg_MFC/Log_Msg_MFC.dsp: Made sure that it - builds only in MFC debug builds. - -Sat Aug 3 14:56:23 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ACEXML/common/HttpCharStream.cpp: - - Fixed a conversion error from char to wchar_t. Thanks to Johnny - for pointing this out. - - * ACEXML/parser/parser/Parser.cpp: - - Fixed an unreachable statement warning. Thanks to Bala for - pointing this out. - -Sat Aug 3 19:53:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ACEXML/common/XML_Codecs.h: - * ACEXML/parser/parser/Parser.h: - * ace/Registry.h: - * ace/String_Base.h: - * ace/QoS/QoS_Session_Factory.h: - Fixed doxygen commands. - -Sat Aug 3 12:36:55 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * ace/Hash_Map_Manager_rt.h: - * ace/Hash_Map_Manager_rt.cpp: - * ace/Hash_Map_Manager_rt.i: Removed these files from the main - trunk. They shouldnt have been there in the first place. - -Fri Aug 2 22:00:26 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * examples/Log_Msg/Log_Msg_MFC/*: Added example to show how - to send ACE_DEBUG output to the Microsoft Visual Studio debugger console. - Thanks to Don Hinton <dhinton@ieee.org> and - Chris Hafey <chafey@stentor.com> for help with this. - -Fri Aug 2 13:10:05 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ACEXML/examples/SAXPrint/main.cpp (ACE_TMAIN): - * ACEXML/common/InputSource.cpp: - - Reverted the previous check-in with respect to the delete - statement. Memory is getting leaked somewhere down in ACE. - -Fri Aug 2 12:31:34 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ACEXML/common/ContentHandler.h: - * ACEXML/parser/parser/Entity_Manager.cpp: - * ACEXML/common/NamespaceSupport.cpp: - - Fixed minor typos in the comments. - - * ACEXML/common/InputSource.cpp: - - Removed an erroneous delete statement. Not sure if this is a - implementation bug or a thinko. - - * ACEXML/common/HttpCharStream.cpp: - * ACEXML/common/Mem_Map_Stream.cpp: - * ACEXML/common/Mem_Map_Stream.h: - - Use char* when sending data on the network. Technically it - should be byte. - - * ACEXML/examples/SAXPrint/Print_Handler.cpp: - * ACEXML/examples/SAXPrint/main.cpp: - - Fixed a bug with getopt() when WCHAR is enabled. Since EOF is - -1, never use a WCHAR in the while loop. - - * ACEXML/parser/parser/Parser.cpp: - * ACEXML/parser/parser/Parser.h: - - Improved Namespace support. Modularized functions in the parser. - - * ACEXML/docs/TODO.txt: - * ACEXML/docs/bugs.txt: - * ACEXML/docs/guidelines.txt: - * ACEXML/docs/parser_features.txt: - - Updated to reflect the current status. People were reading these - old docs and assuming a lot of the features aren't implemented. - -Fri Aug 2 16:06:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/config-doxygen.h: - Added ACE_HAS_ATM define to also generate the doxygen documentation - of the ACE ATM classes. This is one of the last parts of ACE that - are missing in the doxygen documentation. - -Fri Aug 2 07:54:24 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * ace/Malloc.cpp: Rearranged member initialization list in the - constructor of ACE_Malloc_Stats, to fix warnings with explicit - template builds. - -Thu Aug 01 13:40:59 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * tests/Malloc_Test.cpp: Changed the usage of ASYS_TEXT to - ACE_TEXT. Should fix most of the compile errors with g++. - -Thu Aug 1 12:14:21 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/templates/gnu.mpd: - - Added $(ACELIB) to the list of ACE_SHLIBS. - -Thu Aug 1 11:13:20 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * ace/Malloc_Allocator.h: - * ace/Malloc_Allocator.cpp: Removed the declaration and definition - of ACE_Malloc_Stats since they are not needed in this - translation unit. - - * ace/Malloc.h: - * ace/Malloc.cpp: Retained the definition if ACE_Malloc_Stats here - (yes, there were two declarations) and added the definition and - template instantiation for ACE_Atomic_Op used by - ACE_Alloc_Stats. This should fix the erros showing up on the - scoreboard explicit template builds which has - ACE_HAS_MALLOC_STATS turned on. - -Thu Aug 1 09:08:36 2002 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * ace/Message_Block.h: Made the non-static duplicate() and release() - methods virtual so they can be overridden by subclasses. Thanks - to Brian Raven <brianr@liffe.com> for this suggestion. - -Thu Aug 01 15:24:32 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/OS_Thread_Adapter.h: - Fixed incorrected doxygen @file command. - -Thu Aug 01 14:52:44 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * bin/fuzz.pl: - Improved the check for the @file doxygen command. Incomplete file - names where not detected as error. - -Thu Aug 1 07:03:05 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/README: - - Added a bit of background on the tool before plunging into the - details. - -Thu Aug 01 07:59:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/Asynch_Acceptor.h: - * ace/XtReactor.cpp: - * ace/XtReactor.h: - Corrected argument names to fix doxygen warnings. - -Thu Aug 01 07:34:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/config-win32-mingw.h: - Made the definition of the missing parts of the Win32api of MingW - dependent on the version of the w32api instead of our own defines. - This way we automically define the missing things when needed. Thanks - to Craig for suggesting this. - -Wed Jul 31 23:39:27 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * etc/*.doxygen: Change DOT_PATH from /usr/local/bin/ to - empty, forcing doxygen to look for the dot program in PATH. - This eliminates warnings in the Doxygen nightly build. - -Wed Jul 31 12:23:47 2002 Steve Huston <shuston@riverace.com> - - * ace/Malloc.cpp: - * ace/Malloc_T.cpp: - * tests/Malloc_Test.cpp: Changed logging of pointers to %@ - - * tests/Mem_Map_Test.cpp: Fixed loop bound error. - -Wed Jul 31 11:34:58 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * docs/index.html: Added a link to ACEXML document. - -Wed Jul 31 10:47:54 2002 Chad Elliott <elliott_c@ociweb.com> - - * include/makeinclude/rules.local.GNU: - - Added a gnu macro (DEPEND_CMD_ARGS) to the g++dep line to allow - pasing arguments (such as -R) to g++dep when running "make depend" - -Wed Jul 31 08:41:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/config-win32-mingw.h: - For the compilation of Win32_Asynch_IO we now need to have - FILE_SEGMENT_ELEMENT from winnt.h and ReadFileScatter and - WriteFileScatter from winbase.h. - These are not in the header files of the MingW w32api 1.4/1.5 - release. That's why we do these union/defines also in this header - file. By adding the defines ACE_MINGW_LACKS_READWRITE_FILESCATTER - and ACE_MINGW_LACKS_FILE_SEGMENT_ELEMENT to your config.h file - these defines/union are done by ACE for MingW32 so that everything - will compile. - -Wed Jul 31 08:11:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * etc/Task.cpp: - Also the MinGW32 GCC3.1 compiler has problems with the - reinterpret_cast in the svc_run method. Use the static_cast instead. - -Wed Jul 31 07:40:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * etc/*.doxygen: - Enabled warnings to get them in the doxygen scoreboard. - - * ace/Configuration.cpp: - Fixed method argument names to fix doxygen warnings. - -Tue Jul 30 15:45:23 2002 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * ace/SOCK_Dgram.h: Fixed the comment for send(), which had a - cut and paste error. Thanks to Jeff Wilson - <wilsonj@cs.ecs.baylor.edu> for reporting this. - -Tue Jul 30 14:09:35 2002 Chad Elliott <elliott_c@ociweb.com> - - * tests/tests.mpc: - - Update the mpc file to include the libraries. - -Tue Jul 30 09:07:29 2002 Ossama Othman <ossama@uci.edu> - - * ace/Parse_Node.h: - * ace/Parse_Node.cpp: - - Unexported all classes in this file. They are only meant for - internal use by ACE's Service Configurator. Updated Doxygen - comments accordingly. - - (apply, symbol, open_dll): - - These methods now accept a "yyerrno" reference argument to allow - for modification of the error count in a reentrant manner. - - * ace/Svc_Conf.h: - - Removed "ace_yyerrno" and "ace_yylineno" global variables. - Local ones are now used. - - Altered signature of ace_yyerror() function to accept two - additional parameters, yyerrno and yylineno. Addresses a race - condition. - - * ace/Svc_Conf.y: - - Removed all remnants of the global "yyerrno" and "yylineno" - variables. Only the reentrant versions are used now. - - Pass the "yyerrno" variable as a parameter to all functions that - read or modify the parse error count. Addresses several race - conditions. - - Fixed call to ACE_Location_Node::handle(). This method no - longer exists. The correct call is ACE_Location_Node::dll(). - For some reason, Svc_Conf_y.cpp (a generated file!!!) was - "fixed" but the parser grammar file was never fixed. - - * ace/Svc_Conf.l: - - Removed all remnants of the global "yylineno" variable. Only - the reentrant version is used now. - - * ace/Service_Config.cpp (process_directives_i): - - Removed initialization of "ace_yyerrno" and "ace_yylineno" - global variables. They no longer exist. Poof! - - Reverted my previous change. The above changes obviate the need - to take into account the global parse error count. - - * ace/Makefile: - - Tweaked Svc_Conf_y.cpp and Svc_Conf_l.cpp generation rules so - that yyerrno and yylineno are not prepended with an "ace_". - That is no longer needed. - - * etc/Svc_Conf_y.cpp.diff: - - Regenerated this patch against parsers generated by the latest - GNU Bison (1.35). - - * etc/Svc_Conf_l.cpp.diff: - - Regenerated this patch to pull in new changes to Svc_Conf_l.cpp. - -Tue Jul 30 10:06:04 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/Thread_Manager.cpp: Moved a stray colon until *after* - the #if !defined(ACE_USE_ONE_SHOT_AT_THREAD_EXIT) check. Thanks - to Alan Tanga <stang1@lucent.com> for reporting this. - - * ace/POSIX_Proactor.cpp (delete_result_aiocb_list): Moved - the definition of size_t ai out of the for loops to avoid - problems on older C++ compilers. Thanks to George Varsamis - <gv@prismtechnologies.com> for reporting this. - -Tue Jul 30 08:14:34 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/Creator.pm: - - Added an error check for invalid inheritance syntax. - - * bin/MakeProjectCreator/modules/Driver.pm: - - Changed the progress indicator to allow for cleaner output - when an error occurs. - - * bin/MakeProjectCreator/modules/ProjectCreator.pm: - - Remove duplicate libs, libpaths and includes when addition or - subtraction is involved. - - * bin/MakeProjectCreator/templates/gnu.mpd: - - Cleaned up the template to make generated Makefiles a little more - consistent. - -Tue Jul 30 06:50:26 2002 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * ace/Dynamic_Service.cpp: Moved ACE_Dynamic_Service::instance() - from the .i file into the .cpp file and added a #include of - "ace/Service_Object.h" to the .cpp file so the complete type of - ACE_Service_Object would be available to the HP/UX C++ compiler. - Thanks to Torsten Kuepper <Torsten.Kuepper@nokia.com> for - reporting this. - -Tue Jul 30 07:51:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * etc/ace.doxygen: - Enabled warnings and undocument warnings to get these warnings - into the doxygen scoreboard. When this works I will enable the - warnings in all other .doxygen files later. - -Mon Jul 29 14:47:46 2002 Ossama Othman <ossama@uci.edu> - - * ace/Service_Config.cpp (process_directives_i): - - In addition to the thread-safe yyerrno variable, check the - global ace_yyerrno variable when determining if errors have - occurred when processing Service Configurator directives. - Thanks to Doug for tracking down the problem and suggesting a - fix and to Denginere <denginere@hotmail.com> for reporting this. - -Mon Jul 29 17:30:27 2002 Steve Huston <shuston@riverace.com> - - * ace/Get_Opt.h (ctor): Changed documented name of the skip_argv0 - argument to skip_args to better reflect that it's not just - argv[0] that can be skipped. - -Mon Jul 29 07:45:12 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/OS.cpp (uname): Only check for PROCESSOR_ARCHITECTURE_IA64 - when this is defined. The MingW w32api doesn't define this constant. - -Sun Jul 28 17:17:50 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * docs/ACE-subsets.html: Updated the breakdown of files to - each subset component so it's consistent with the - $ACE_ROOT/Makefile. Thanks to Ruslan Zasukhin - <sunshine@public.kherson.ua> for reporting this. - -Sun Jul 28 10:31:12 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * include/makeinclude/platform_gnuwin32_common.GNU: - Use -mcpu=$(TCPU) instead of -m$(TCPU), because MingW GCC 3.1 - now gives warnings that -m is deprecated. Thanks to Craig Rodrigues - for confirming that -m was already deprecated in GCC 2.95.3 - -Sat Jul 27 17:35:06 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * include/makeinclude/platform_g++_common.GNU (SOFLAGS): - - Allow building of shared libraries with gcc and Solaris ld. - There wasn't any reason that it wouldn't work. - - * Changelog: Remove >>>>> left out by merge conflicts. People - should be more careful about checking for conflicts. - -Sat Jul 27 16:20:19 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * examples/Service_Configurator/Misc/main.cpp (ACE_TMAIN): Changed - the code to be more resilient against errors that occur during - parsing of the svc.conf file. Thanks to Denginere - <denginere@hotmail.com> for motivating this change. - -Sat Jul 27 15:50:12 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ACEXML/examples/svcconf/main.cpp: - Made this file compiling with BCB and Unicode - - * ACEXML/examples/svcconf/Svcconf_Handler.h: - Made includes like the other examples so that we can build it with - BCB - - * ACEXML/examples/svcconf/Svcconf_Handler.cpp: - Fixed unused argument warnings - - * ACEXML/examples/svcconf/Makefile.bor: - Added this BCB makefile - - * ACEXML/examples/Makefile.bor - Added svcconf directory - -Fri Jul 26 21:00:32 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * ace/TP_Reactor.cpp: Fixed a unused variable warning in g++ - builds. - -Fri Jul 26 20:08:41 2002 Steve Huston <shuston@riverace.com> - - * include/makeinclude/platform_aix4_g++.GNU: Corrected option for - using threads from -mthreads to -pthread. This may not work with - pre-gcc 3 compilers, but those compilers are busted on AIX anyway. - Added support for buildbits=32 and buildbits=64. Moved in the new - AIX library-building setup from platform_aix_ibm.GNU. - -Fri Jul 26 15:52:12 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * ace/TP_Reactor.cpp: Fixed a problem with handle_event - (ACE_Time_Value &). It should be calling the TP_Reactor's - handle_event () instead of Select_Reactor's. This should fix - [BUGID 1257]. - -Fri Jul 26 10:11:41 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/config/messaging.mpb: - * bin/MakeProjectCreator/config/messaging_only.mpb: - - Set up a new base project that just contains information about - the Messaging library. - - * bin/MakeProjectCreator/modules/GNUProjectCreator.pm: - * bin/MakeProjectCreator/templates/gnu.mpd: - - Allow the file lists to contain names of files that correspond - to relative directories. - -Fri Jul 26 06:40:12 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/OS_String.inl: Added casts of the input parameter to - (unsigned char) for the ace_isprint() and ace_isspace() methods - on ACE_OS_String to prevent problems with MSVC++ 7.0. Thanks to - Colin Weaver <colin.weaver@invensys.com> for suggesting this - fix. - -Fri Jul 26 10:37:12 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * etc/tao_pss.doxygen: - Added doxygen config file to generate the new TAO PSS. - - * bin/generate_doxygen_pl: - Added tao_pss.doxygen to the list of files to generate. - - * htlm/index.html: - Added link for the new PPS doxygen documentation. - -Thu Jul 25 22:58:24 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * ace/Synch.cpp: Removed explicit instantiation of - ACE_Condition<ACE_Recursive_Thread_Mutex> since a template - specialization exists. This was causing a compile error when - explicit template instantiations were used with - g++-2.95.3. - -Thu Jul 25 11:32:41 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/README: - - Added a little more detail about what the depends keyword refers - to. - - * bin/MakeProjectCreator/USAGE: - * bin/MakeProjectCreator/modules/Driver.pm: - - Changed the code to show the default type in the usage message. - - * bin/MakeProjectCreator/modules/GNUWorkspaceCreator.pm: - - Change the workspace to use a for loop when building on UNIX. - - * tests/RMCast/tests.mpc: - * tests/SSL/tests.mpc: - - Fixed the case on the dependencies lines. - -Wed Jul 24 13:38:40 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/Creator.pm: - * bin/MakeProjectCreator/modules/Driver.pm: - * bin/MakeProjectCreator/modules/ProjectCreator.pm: - * bin/MakeProjectCreator/modules/WorkspaceCreator.pm: - - Added a progress indicator (for those big workspaces). - -Wed Jul 24 13:10:19 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/ProjectCreator.pm: - - Add README and readme to the default documentation - files. - - * bin/MakeProjectCreator/modules/Creator.pm: - * bin/MakeProjectCreator/modules/GUID.pm: - * bin/MakeProjectCreator/modules/VC7ProjectCreator.pm: - - Change the GUID generation code to be deterministic (but still - unique) to allow for multiple solution files to contain the same - project files without having to regenerate the solutions. - -Wed Jul 24 13:42:37 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.cpp (uname): Added case for PROCESSOR_ARCHITECTURE_IA64 - when decoding processor type. - -Wed Jul 24 10:58:22 2002 Steve Huston <shuston@riverace.com> - - * ace/Acceptor.cpp (ACE_Acceptor::handle_input): When calling select(), - use a int value instead of casting a ACE_HANDLE to an int. - - * tests/Cached_Accept_Conn_Test.cpp: Use %@ instead of %d when - logging pointers. - - * tests/MT_Reactor_Upcall_Test.cpp: Added ACE_TEXT to char strings - where needed. Changed the send to send fixed header, then the - variable-size string. Sending all at once doesn't work if the - string doesn't immediately follow the size_ member in memory, - which happened on Win64. - -Wed Jul 24 10:07:24 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/README: - * bin/MakeProjectCreator/USAGE: - - Update the documentation to reflect the current state. - -Wed Jul 24 08:13:48 2002 Christopher Kohlhoff <chris@kohlhoff.com> - - * include/makeinclude/build_dll.bor: - * include/makeinclude/build_exe.bor: - * include/makeinclude/build_lib.bor: - * include/makeinclude/make_flags.bor: - Improve debugging support with Borland C++Builder by invoking the - compiler with a full path to each source file. This can be disabled - by setting a NO_FULL_PATH environment variable to 1 or by passing - -DNO_FULL_PATH when running make, which may be necessary when doing - a debug build on Win9x. - -Wed Jul 24 07:45:43 2002 Chad Elliott <elliott_c@ociweb.com> - - * THANKS: Added Tim Bradley to the hall of fame. - -Tue Jul 23 21:46:54 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/Process_Mutex.h: - * ace/Process_Mutex.cpp: Disallowed anonymous Process_Mutex by - invoking this->unique_name() to create a temporary name if - there's no mutex name passed to the constructor. Thanks John - Michael Zorko <j.zorko@att.net> for reporting this problem on - Mac OSX and submitting the patch. - -Tue Jul 23 13:08:27 2002 Steve Huston <shuston@riverace.com> - - * ace/Log_Msg.{h cpp} (log_hexdump): - * ace/ACE.{h cpp} (format_hexdump): Change size arguments from int - to size_t. Change format_hexdump()'s return from int to size_t. - -Tue Jul 23 09:48:15 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/Driver.pm: - - Always add the default include paths. - - * bin/MakeProjectCreator/modules/ProjectCreator.pm: - - Make sure that idl generated source files go at the front of the - file list in auto-generates source lists. - - * bin/MakeProjectCreator/templates/gnu.mpd: - - Only add LIB and SHLIB assignments if we are generating a Makefile - for libraries. - -Mon Jul 22 22:30:53 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ACEXML/common/XML_Common.dsp: Added the macro definition - ACEXML_HAS_DLL=1 to the release configuration to resolve a - nightly biuld problem. - -Mon Jul 22 18:28:48 2002 Steve Huston <shuston@riverace.com> - - * ace/Configuration.h: Doxygen fixes. - -Mon Jul 22 16:57:19 2002 Chris Cleeland <cleeland_c@ociweb.com> - - * bin/g++dep: Added a -R option that allows you to specify a - variable which should be used to generate relative paths if it's - defined. - -Mon Jul 22 17:45:48 2002 Steve Huston <shuston@riverace.com> - - * ace/Get_Opt.h: Doxygen fixes. - -Mon Jul 22 16:39:21 2002 Steve Huston <shuston@riverace.com> - - * ace/Acceptor.cpp (info): - * ace/Connector.cpp (info): Cast return val to int. - - * ace/Message_Queue.h (message_count): - * ace/Message_Queue_T.{h cpp i} (message_count): Change return value - from size_t to int. Changed cur_count_ member (and accessors that - return it) from size_t to int. This is a compromise to unify the - length of arguments without changing the return types of all - methods that return a message count (all the enqueue/dequeue methods, - as well as a few others). - Changed a number of @arg to @param for doxygen. - - * ace/SOCK_IO.{h i} (recvv, sendv): - * ace/SOCK_Stream.{h i}(recvv_n, sendv_n): Changed iovcnt arg from - size_t to int. int, though sort of odd as a choice, is the - UNIX98 standard and more commonly implemented. Removes an arg - size mismatch between the SOCK_IO and ACE layers, restricting the - iovcnt range at the SOCK_IO layers to match the layer below. - Changed the recv and send iov-enabled, deprecated methods to cast - their size_t iovcnt args to int... these deprecated methods should - be removed after 5.3 is out. - - * tests/Conn_Test.cpp: Replaced %d with %@ when logging pointers. - Added missing string arg for a ACE_ERROR %p directive. Use a - int select_width instead of casting ACE_HANDLE for Win64. - -Mon Jul 22 15:13:38 2002 Carlos O'Ryan <coryan@atdesk.com> - - * ace/Atomic_Op.h: - * ace/Atomic_Op.i: - Fixed bug 1253, add missing copy constructor. - -Mon Jul 22 13:25:29 2002 Stephen Torri <storri@cs.wustl.edu> - - * ace/streams.h: Fixed _MSC_VER compiler warning messages stating - concern about using MSC_VER without assurance that the macro - was defined. - -Mon Jul 22 12:20:21 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * THANKS: Added Sean Ogle to the hall of fame. - -Mon Jul 22 09:19:34 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * THANKS: Aded Kew Whitney and Colin Weaver to the hall of fame. - -Sat Jul 20 08:53:05 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/SOCK_Dgram.cpp (set_nic): Moved the #endif for - ACE_WIN32 *ahead* of the set_options() call. Thanks to Joe - Hayes <Joe.Hayes@lmco.com> for reporting this. - -Fri Jul 19 14:42:37 2002 Steve Huston <shuston@riverace.com> - - * ace/WFMO_Reactor.cpp (make_changes_in_suspension_infos, - make_changes_in_current_infos): - Fixed bug resulting from: Wed Jul 17 11:34:28 2002 Steve Huston - -Thu Jul 18 15:48:29 2002 Steve Huston <shuston@riverace.com> - - * ace/Future.h: Doxygenation improvements. - -Thu Jul 18 08:39:08 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/README: - - Update the documentation to reflect the additions to the grammar. - - * bin/MakeProjectCreator/config/avstreamsexe.mpb: - * bin/MakeProjectCreator/config/corba_messaging.mpb: - * bin/MakeProjectCreator/config/minimum_corba.mpb: - * bin/MakeProjectCreator/config/namingexe.mpb: - * bin/MakeProjectCreator/config/notify.mpb: - * bin/MakeProjectCreator/config/orbsvcsexe.mpb: - * bin/MakeProjectCreator/config/orbsvcslib.mpb: - * bin/MakeProjectCreator/config/rt_client.mpb: - * bin/MakeProjectCreator/config/rt_server.mpb: - * bin/MakeProjectCreator/config/rteventexe.mpb: - - Add the GNU specific info for generating conditional target - Makefiles. - - * bin/MakeProjectCreator/modules/GNUProjectCreator.pm: - - Added code to generate the conditional sections of the GNU - Makefiles. - - * bin/MakeProjectCreator/modules/ProjectCreator.pm: - * bin/MakeProjectCreator/modules/TemplateParser.pm: - - Fix some minor problems with generated project files. - - * bin/MakeProjectCreator/templates/gnu.mpd: - - Added template info for conditional targets. - - * bin/MakeProjectCreator/templates/vc6dsp.mpd: - * bin/MakeProjectCreator/templates/vc7.mpd: - - Removed the tie class idl targets. - - * bin/ace_components: - - Added a new append option that will be used by the generated GNU - Makefiles. - -Thu Jul 18 06:17:19 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * Doug turns 40 today. Fortunately, 40 Octal is only 32 Decimal ;-). - -Wed Jul 17 15:18:31 2002 Steve Huston <shuston@riverace.com> - - * ace/Configuration.h: Doxygen improvements. - - * ace/Configuration.cpp: All error cases where a return - (-2, -3, -4, -5) occurred now return -1 with errno set. - - * ace/OS.{i cpp}: Follow-up to Wed Jul 17 11:34:28 2002 Steve Huston - Only make ACE_WIN64 use intptr_t - must have a new SDK to get - intptr_t, and not all have it.... so use it only for ACE_WIN64 - since the compiler comes in the newer SDK... Win32 does - fine with long, which is what an intrptr_t is on Win32. - - Also fixed the PACE-busted conditional compiles around siginfo_t - to match the declaration in OS.h. Win32 needs this stuff. - -Wed Jul 17 14:25:39 2002 Steve Huston <shuston@riverace.com> - - * include/makeinclude/platform_aix_ibm.GNU: Added -qfuncsect to the - compile options. Allows for smaller executables/libraries, and - gets rid of most of the multiple definition warnings for templates. - Unfortunately, there are still a few stragglers so the aix_shr - post-processing script is still required. - - * ace/Sock_Connect.cpp: Added a hack to get Visual Age C++ on AIX - build running clean. Have to include a template instantiation we - don't need to get the compiler to pull the proper header files - at template compilation time. - -Wed Jul 17 11:34:28 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.{i cpp}: Windows, in calls to _open_osfhandle(), cast - handle to long for C++Builder. Only MSVC has the intptr_t type. - - * ace/WFMO_Reactor.cpp (make_changes_in_current_infos, - make_changes_in_suspension_infos): Run the index from 0 to - last_valid_slot rather than valid slot down to 0 to allow - correct use of size_t index. - - * ace/WIN32_Proactor.cpp (register_handle, post_completion): Use - a reinterpret_cast, not static_cast to cast void* to ULONG for - non-Win64. - -Wed Jul 17 03:50:56 UTC 2002 Don Hinton <dhinton@objectsciences.com> - - * ace/ACE.cpp: - * ace/OS.{h,i,cpp}: - * ace/OS_Dirent.inl: - * ace/OS_Memory.h - * ace/OS_String.inl: - * ace/Thread_Manager.cpp: - * ace/WFMO_Reactor.cpp: - * ace/config-all.h: - * ace/config-win32-borland.h: - * ace/config-win32-msvc-5.h: - * ace/config-win32-msvc-6.h: - * include/makeinclude/wrapper_macros.GNU: - - Removed PACE. - -Tue Jul 16 14:19:37 2002 Steve Huston <shuston@riverace.com> - - * ace/DLL_Manager.cpp: - * ace/Parse_Node.cpp: (also see Fri Jul 12 12:37:42 2002 Steve Huston) - Always use the ptrdiff_t trick to cast a void* to a func ptr. - Nanbor was right ;-) Thanks to Carlos O'Ryan <coryan@atdesk.com> - for clarifying this issue, including C++ chapter and verse. - (It's in ISO C++ section 5.2.10 for those interested). - - * tests/DLL_Test.cpp: Use ptrdiff_t as a mediary to cast a void * - to a func ptr, not a long. - -Mon Jul 15 21:23:47 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * ace/Message_Block.i (base): Fixed a warning in TRU 64 builds. - -Mon Jul 15 18:56:47 2002 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * ace/String_Base.cpp (substring): Changed "nil" to "nill" to - avoid conflicts with a Mac OS X macro. Thanks to Ruslan - Zasukhin <sunshine@public.kherson.ua> for reporting this. - -Mon Jul 15 15:18:33 2002 Steve Huston <shuston@riverace.com> - - * ace/Configuration.{h cpp}: Changed ACE_Configuration_Value_IntId - data_ member to be a union with both pointer and u_int members. - Also, the length_ member is now size_t, not u_int. - - * ace/Memory_Pool.cpp (ACE_Pagefile_Memory_Pool::map): Use - INVALID_HANDLE_VALUE instead of a literal 0xFFFFFFFF for call to - CreateFileMapping. - - * ace/OS.{h i}: New method, void *atop(const char *s) converts a - string to void *. - - * ace/OS.i (fdopen): Don't cast arg to _open_osfhandle - Microsoft - changed the type to something sensible. - (isatty): open a C run-time handle for the ACE_HANDLE. - - * ace/OS.cpp (fopen): Don't cast arg to _open_osfhandle - Microsoft - changed the type to something sensible. - (unique_name): Use sprintf %p for pointer rather than casting - to an int. - - * ace/Naming_Context.cpp (ACE_Name_Options::parse_args): Use - ACE_OS::atop instead of ACE_OS::atoi for converting an argument - to a pointer. - - * ace/SOCK_Dgram.cpp (recv, send): - * ace/SOCK_IO.cpp (recvv): Use an int select_width - to pass width to select() so it can be ignored on ACE_WIN64. - - * ace/SV_Semaphore_Simple.cpp (name_2_key): Disable the type cast - conversion warning casting ACE::crc32() to key_t... it's fine. - - * ace/WIN32_Proactor.cpp (register_handle, post_completion): The - I/O completion port's completionKey argument is a ULONG in - "older" compiler/SDK versions, but was changed to a - ULONG_PTR (new type in newer SDK/compiler) for 64-bit - transition. The new type is used conditionally based on whether or - not we're building a 64-bit version. - -Mon Jul 15 09:26:17 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/SString.cpp (substring): Changed "nil" to "nill" to - avoid conflicts with a Mac OS X macro. Thanks to Ruslan - Zasukhin <sunshine@public.kherson.ua> for reporting this. - -Mon Jul 15 10:14:44 2002 Carlos O'Ryan <coryan@atdesk.com> - - * bin/MakeProjectCreator/README: - Fixed typo - -Sun Jul 14 20:14:16 2002 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * FAQ: Fixed the use of "*.ps.gz" files. Thanks to - Carol Hunsicker <carol.hunsicker@lmco.com> for pointing this - out. - -Sun Jul 14 01:55:03 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * ace/Message_Block.i (reset_allocators): Reset the allocators in - the chain if the message block is chained. - -Sun Jul 14 01:43:32 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * ace/Message_Block.h: - * ace/Message_Block.i: Added two methods viz. reset_allocators () - and access_allocators () to ACE_Message_Block. - -Sat Jul 13 17:29:35 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * bin/fuzz.pl: Removed the word "warning" from a print statement, - since our scoreboard scripts seems to catch that as a warning. - -Sat Jul 13 13:31:54 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * bin/auto_run_tests.lst: Added the Stack_Recursion test to the - daily builds. - -Fri Jul 12 20:28:52 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ACEXML/common/URL_Addr.cpp (string_to_addr): - - Fixed confusion in handling normal and wchar strings. It should - compile now without problems. - - * ACEXML/apps/svcconf/Svcconf_Handler.cpp: - * ACEXML/apps/svcconf/Svcconf_Handler.i: - - Fixed memory leaks, some typos and used the locator to print - out error messages. - -Fri Jul 12 17:55:08 2002 Carlos O'Ryan <coryan@atdesk.com> - - * ace/String_Base.i: - Fixed rfind() to check for pos == npos also, that solved the - regression test failures. - -Fri Jul 12 16:03:22 2002 Steve Huston <shuston@riverace.com> - - * ace/Basic_Types.h: Itanium is little-endian, like x86. Although - the Itanium architecture allows big-endian operation, Windows - doesn't use it. - - * tests/test_config.h (randomize): Fixed some int/size_t mix-ups. - -Fri Jul 12 13:02:51 2002 Steve Huston <shuston@riverace.com> - - * ace/String_Base.{h i} (rfind): Changed pos argument from size_t to - ssize_t to make proper use of npos and avoid compile error for - g++ 2.95.2. - -Fri Jul 12 12:37:42 2002 Steve Huston <shuston@riverace.com> - - * ace/DLL_Manager.cpp: - * ace/Parse_Node.cpp: g++ complains that ISO C++ won't allow cast - from void * to func ptr... so cast void * to ptrdiff_t, then to - func ptr. - -Fri Jul 12 12:26:27 2002 Steve Huston <shuston@riverace.com> - - * ace/Timer_Heap_T.cpp (ctor): Alloc timer_ids_ as ssize_t[], - not long[]. - - * examples/C++NPv2/TPCLS.dsp: Added SSL libraries to link input. - - * examples/C++NPv2/TPLS.dsp: Added Logging_Handler.{h cpp} to project. - -Fri Jul 12 12:15:49 2002 Steve Huston <shuston@riverace.com> - - * ace/Timer_Heap_T.cpp (grow_heap): Fix compile warning. - -Thu Jul 11 21:09:36 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * Makefile.bor: Added a new BUILD target called "nightlybuild" to - simplify nightlybuild script. - -Thu Jul 11 20:42:52 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/OS.i (creat): The Win32 implementation of this function was - broken. Thanks to Marvin Greenberg <mgreenberg@dctd.saic.com> - for submitting the fix. - -Thu Jul 11 18:37:35 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ACEXML/common/URL_Addr.cpp (string_to_addr): - * ACEXML/common/URL_Addr.h: - - Change the argument to match the base class INET_Addr so that - warnings under Borland BCB is eliminated. - - * ACEXML/common/HttpCharStream.cpp: - - Use ACE_TEXT_ALWAYS_CHAR to convert an Unicode URL to char* - before converting it to an address. Thanks to Nanbor for - pointing this out. - - * ACEXML/common/Makefile: - * ACEXML/common/StreamFactory.h: - * ACEXML/common/StreamFactory.cpp: - - New class to create appropriate streams based on the URI. This - is needed for handling the parameter entities. - - * ACEXML/examples/SAXPrint/main.cpp: - - Move common error message to a new usage function. - -Thu Jul 11 16:28:24 2002 Steve Huston <shuston@riverace.com> - - * tests/Proactor_Scatter_Gather_Test.cpp: Changed Writer to derive - from ACE_Handler, not ACE_Service_Handler. It's not the target of - an asynch connection factory, and declares an open() method that - hides the one in ACE_Service_Handler. - -Thu Jul 11 16:15:49 2002 Steve Huston <shuston@riverace.com> - - * examples/C++NPv2/SR_Configurable_Logging_Server.dsp: - * examples/TPCLS.dsp: - * examples/TPLS.dsp: - * examples/C++NPv2/examples.dsw: Added SR_Configurable_Logging_Server, - TPCLS, and TPLS projects. - - * examples/C++NPv2/TPC_Logging_Server.h: Use ACE traits macros to - allow build on Win32. - * examples/C++NPv2/TPC_Logging_Server.cpp: Fix build problems on - Win32. - - * examples/TP_Logging_Server.h (init): Added #include "ace/Auto_Ptr.h" - and corrected use of ACE_NEW_NORETURN; use ACE_Auto_Ptr::get() to get - the argv pointer. - -Thu Jul 11 15:03:47 2002 Steve Huston <shuston@riverace.com> - - * ace/ace_dll64.mak: TEMPORARILY added /wd4267 to disable the - conversion size_t to int warning. - - * ace/Activation_Queue.{h i} (method_count): Change return type - from int to size_t. - - * ace/Arg_Shifter.cpp (cur_arg_strncasecmp): Change flag_length var - to size_t to match string function signatures. If returning a - 'remaining' length, cast back to int. - - * ace/ARGV.cpp (ctor): Build string length using size_t, not int. - - * ace/CDR_Base.cpp (ACE_CDR::consolidate): Use ptrdiff_t, not the - nonstandard ptr_arith_t. - - * ace/CDR_Stream.cpp (grow_and_adjust): Use ptrdiff_t, notptr_arith_t. - (write_string): Cast ACE_SString:length result to ACE_CDR::ULong. - A number of other size_t<->ACE_CDR::ULong casts. - - * ace/CDR_Stream.i (write_string, operator<<): Use a ACE_CDR::ULong - len, can't directly use the return from strlen(), a size_t. - - * ace/Codecs.cpp: size_t conversions. - - * ace/DEV_Addr.i (set): Cast strlen result to int for base_set. - - * ace/DEV_IO.{i cpp} (send, recv): Cast iovec count from size_t to - int for call to ACE_OS level. - - * ace/Hash_Map_Manager_T.{h, i, cpp} (shared_find): Change loc argument - from u_long to size_t. Better match for its use. - - * ace/OS.h (ACE_align_binary, ACE_ptr_align_binary): Change from - using ptr_arith_t to the standard ptrdiff_t. - - * ace/DLL_Manager.cpp: - * ace/Parse_Node.cpp: Avoid casting a void * to a long then back to a - function pointer. Cast using proper calling signature. - - * ace/Reactor_Impl.h (ACE_Reactor_Notify::notify): - * ace/Select_Reactor_Base.{h cpp} (ACE_Select_Reactor_Notify::notify): - Change return type from ssize_t to int to match advertised return - type from ACE_Reactor. - - * ace/Select_Reactor_T.cpp (check_handles): Use an int select_width - to pass width to select() so it can be ignored on ACE_WIN64. - - * ace/String_Base_Const.{h cpp}: Changed npos from int to ssize_t. - - * ace/String_Base.{h i}: (operator[]) changed slot arg from int - to size_t. (strstr(), find(), rfind()) changed return type from - int to ssize_t; change pos arg from int to size_t. - - * ace/Timer_Heap_T.{h cpp}: Change 'slot' arg from int to size_t. - Also changed timer_ids_ to use ssize_t rather than long to - avoid length mismatches between the max size and the timer - queue (a size_t) and the arrays. - - * ace/WFMO_Reactor.h (ACE_WFMO_Reactor_Notify): Change return type - from ssize_to to int to match corresponding change in - ACE_Reactor_Notify, above. The implementation already was returning - int. - - * ace/WFMO_Reactor.cpp (make_changes_in_current_infos, - make_changes_in_suspension_infos): Align use of size_t. - (wait_for_multiple_events)...........Should maybe make the - sizes in this class DWORD????? - - * ace/WIN32_Proactor.cpp (close, handle_events): completion_key - needs to be a different on Win64 than on Win32. - -Thu Jul 11 08:34:22 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/USAGE: - * bin/MakeProjectCreator/modules/Driver.pm: - - Added new options and documentation dealing with generation of - relative paths in projects. - -Wed Jul 10 11:46:02 2002 Mayur Deshpande <mayur@ics.uci.edu> - - * bin/auto_run_tests.lst: - - Added tests/AMH_Exception and examples/AMH/Sink_Server as tests - for AMH - -Wed Jul 10 13:31:49 2002 Steve Huston <shuston@riverace.com> - - * examples/C++NPv2/CLD.dsp: - * examples/C++NPv2/examples.dsw: Added CLD project. - - * examples/Client_Logging_Daemon.cpp: Disable the info, suspend, and - resume methods - these are left as an exercise for the reader. - -Wed Jul 10 11:24:39 2002 Steve Huston <shuston@riverace.com> - - * examples/C++NPv2/AIO_CLD.dsp: - * examples/C++NPv2/examples.dsw: Added AIO_CLD project. - - * examples/C++NPv2/AIO_Client_Logging_Daemon.cpp: Fix Win32 compile - problems. - -Wed Jul 10 07:22:24 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/TemplateParser.pm: - - If the project creator needs the slashes converted, then - do so when generating relative paths. - -Tue Jul 9 23:24:44 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * include/makeinclude/rules.local.GNU: - - Print out more explanative messages. - - * include/makeinclude/platform_osf1_4.x_cxx.GNU: - * include/makeinclude/platform_sunos5_sunc++.GNU: - - Turn on symbol versioning for these platforms with the - proprietary compilers that ships with these systems viz, Sun CC - and cxx. - - -Tue Jul 9 16:24:46 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/Containers_T.cpp: Fixed inconsistent initializer order for - ACE_Fixed_Stack and ACE_Bounded_Stack. Thanks to Michael - Searles <msearles@base16.com> for reporting this. - -Tue Jul 9 11:10:38 2002 Steve Huston <shuston@riverace.com> - - * apps/Gateway/Peer/Peer.{h cpp}: Renamed Peer_Acceptor::open() - to Peer_Acceptor::start() to avoid clashes with virtual - ACE_Acceptor::open. - -Tue Jul 9 09:01:14 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/config/taolib_with_idl.mpb: - - Change the default tao_idl option from -Gs to -Sc. - -Tue Jul 9 08:07:20 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/TemplateParser.pm: - - When using Cygwin Perl, we need to convert the cwd to a windows - path to match up with environment variables that are usable - outside of Cygwin. - -Tue Jul 9 07:46:57 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/TemplateParser.pm: - - Fix up the relative value matching code to work correctly - on Windows. - -Tue Jul 9 07:31:48 2002 Chad Elliott <elliott_c@ociweb.com> - - * ace/ace.mpc: - * ace/QoS/qos.mpc: - * ace/RMCast/rmcast.mpc: - * ace/SSL/ssl.mpc: - * apps/gperf/src/gperf.mpc: - * tests/tests.mpc: - * tests/RMCast/tests.mpc: - * tests/SSL/tests.mpc: - - Input files for the Make Project Creator for just the basic - libraries, exe's and tests. - -Tue Jul 9 07:24:06 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/mpc.pl: - * bin/mwc.pl: - - The starting point for creating workspaces and projects. - - * bin/MakeProjectCreator/README: - * bin/MakeProjectCreator/USAGE: - - A bit of lite documentation on the Make Project Creator. - - * bin/MakeProjectCreator/config/aceexe.mpb: - * bin/MakeProjectCreator/config/acelib.mpb: - * bin/MakeProjectCreator/config/aceversion.mpb: - * bin/MakeProjectCreator/config/avstreamsexe.mpb: - * bin/MakeProjectCreator/config/client.mpb: - * bin/MakeProjectCreator/config/global.mpb: - * bin/MakeProjectCreator/config/messaging.mpb: - * bin/MakeProjectCreator/config/namingexe.mpb: - * bin/MakeProjectCreator/config/notify.mpb: - * bin/MakeProjectCreator/config/notifytest.mpb: - * bin/MakeProjectCreator/config/orbsvcsexe.mpb: - * bin/MakeProjectCreator/config/orbsvcslib.mpb: - * bin/MakeProjectCreator/config/rt_client.mpb: - * bin/MakeProjectCreator/config/rt_server.mpb: - * bin/MakeProjectCreator/config/rteventexe.mpb: - * bin/MakeProjectCreator/config/server.mpb: - * bin/MakeProjectCreator/config/strategies.mpb: - * bin/MakeProjectCreator/config/taoexe.mpb: - * bin/MakeProjectCreator/config/taolib.mpb: - * bin/MakeProjectCreator/config/taolib_with_idl.mpb: - * bin/MakeProjectCreator/config/taoversion.mpb: - - Base projects for most of ACE and TAO. - - - * bin/MakeProjectCreator/modules/Driver.pm: - * bin/MakeProjectCreator/modules/Creator.pm: - * bin/MakeProjectCreator/modules/WorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/GUID.pm: - * bin/MakeProjectCreator/modules/Parser.pm: - * bin/MakeProjectCreator/modules/ProjectCreator.pm: - * bin/MakeProjectCreator/modules/TemplateInputReader.pm: - * bin/MakeProjectCreator/modules/TemplateParser.pm: - - The main modules of the Make Project Creator which provide most of - the functionalty. - - * bin/MakeProjectCreator/modules/BorlandProjectCreator.pm: - * bin/MakeProjectCreator/modules/BorlandWorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/EM3ProjectCreator.pm: - * bin/MakeProjectCreator/modules/EM3WorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/GHSProjectCreator.pm: - * bin/MakeProjectCreator/modules/GHSWorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/GNUProjectCreator.pm: - * bin/MakeProjectCreator/modules/GNUWorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/NMakeProjectCreator.pm: - * bin/MakeProjectCreator/modules/NMakeWorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/VA4ProjectCreator.pm: - * bin/MakeProjectCreator/modules/VA4WorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/VC6ProjectCreator.pm: - * bin/MakeProjectCreator/modules/VC6WorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/VC7ProjectCreator.pm: - * bin/MakeProjectCreator/modules/VC7WorkspaceCreator.pm: - - Subclassed workspace and project creators. - - * bin/MakeProjectCreator/templates/bor.mpd: - * bin/MakeProjectCreator/templates/bordll.mpt: - * bin/MakeProjectCreator/templates/borexe.mpt: - * bin/MakeProjectCreator/templates/em3vcp.mpd: - * bin/MakeProjectCreator/templates/em3vcpdll.mpt: - * bin/MakeProjectCreator/templates/em3vcpdllexe.mpt: - * bin/MakeProjectCreator/templates/em3vcplib.mpt: - * bin/MakeProjectCreator/templates/em3vcplibexe.mpt: - * bin/MakeProjectCreator/templates/ghs.mpd: - * bin/MakeProjectCreator/templates/gnu.mpd: - * bin/MakeProjectCreator/templates/gnudll.mpt: - * bin/MakeProjectCreator/templates/gnuexe.mpt: - * bin/MakeProjectCreator/templates/nmake.mpd: - * bin/MakeProjectCreator/templates/nmakedll.mpt: - * bin/MakeProjectCreator/templates/nmakeexe.mpt: - * bin/MakeProjectCreator/templates/va4icc.mpd: - * bin/MakeProjectCreator/templates/va4iccdll.mpt: - * bin/MakeProjectCreator/templates/va4iccdllexe.mpt: - * bin/MakeProjectCreator/templates/va4icclib.mpt: - * bin/MakeProjectCreator/templates/va4icclibexe.mpt: - * bin/MakeProjectCreator/templates/vc6dsp.mpd: - * bin/MakeProjectCreator/templates/vc6dspdll.mpt: - * bin/MakeProjectCreator/templates/vc6dspdllexe.mpt: - * bin/MakeProjectCreator/templates/vc6dsplib.mpt: - * bin/MakeProjectCreator/templates/vc6dsplibexe.mpt: - * bin/MakeProjectCreator/templates/vc7.mpd: - * bin/MakeProjectCreator/templates/vc7dll.mpt: - * bin/MakeProjectCreator/templates/vc7exe.mpt: - - Template and template input files for all of the different project - creators. - -Tue Jul 9 07:21:29 2002 Chad Elliott <elliott_c@ociweb.com> - - * ace/Map_T.h: - * ace/Map.cpp: - * ace/Pair_T.h: - * ace/Pair.cpp: - - Remove the use of Map.h and Pair.h. - - * ace/gethrtime.cpp: - - Only compile the contents of this file if using GHS and are - building for the x86. - -Mon Jul 8 17:26:51 2002 Steve Huston <shuston@riverace.com> - - * ace/Task.cpp (svc_run): Borland complains about reinterpret_cast from - int to unsigned long, so change the cast to ACE_THR_FUNC_RETURN to - a static_cast for that compiler. Thanks to Johnny Willemsen - <jwillemsen@remedy.nl> for this fix. - -Fri Jul 5 10:30:44 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * ace/Array_Base.cpp: Swapped the order of includes so that - Array_Base.h comes after Malloc_Base.h. Thanks to Al Pariante - <Albert.Pariante@axcelis.com> for reporting this. - -Thu Jul 4 11:27:05 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * Happy 226th Birthday USA! - -Sat Jul 06 19:32:40 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * ace/Time_Value.h (ACE_Time_Value): Fix for [BUGID 1240]. Some of - the operators were missing explicit declarations in the header - file (they were anyway defined in the inlined files) which - created problems in builds where inlining was diabled. Please - see - - http://deuce.doc.wustl.edu/bugzilla/show_bug.cgi?id=1240 for - more details. Thanks to Marina Spivak <marina@atdesk.com> for - reporting this. - -Fri Jul 5 18:44:39 2002 Steve Huston <shuston@riverace.com> - - * ace/ACE.cpp: Corrected undeclared 'select_width' errors and - signed/unsigned mismatch warnings. - -Fri Jul 5 16:14:51 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.{h i} (send, sendto, recv, recvfrom): Change 'len' argument - from int to size_t. - Removed conditional typedef of ssize_t; it's done in Basic_Types.h. - - * ace/ACE.h (recv): Filled in documentation for varargs variant. - * ace/ACE.cpp (send, recv): Properly convert size_t 'count' arg to int - iovec count for call to ACE_OS::sendv/recvv(). - (handle_timed_accept, handle_timed_complete, handle_ready): Use - an int select_width for ACE_OS::select() calls to get the types - correct. This arg is ignored on Windows anyway. - - * ace/Basic_Types.h: If ACE_HAS_SSIZE_T not defined, and Win64, - typedef SSIZE_T ssize_t; else use the existing int typedef. - -Wed Jul 3 11:09:31 2002 Steve Huston <shuston@riverace.com> - - * tests/SSL/Thread_Pool_Reactor_SSL_Test.cpp: - Change thread function to return ACE_THR_FUNC_RETURN, not void *. - -Tue Jul 02 21:43:53 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ACEXML/common/XMLFilterImpl.h: One more pre/post mismatch. - -Tue Jul 2 16:58:55 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ACEXML/common/XML_Types.h: - - Removed unnecessary includes and fixed a typo. - -Tue Jul 02 16:35:19 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ACEXML/common/Attributes.h: - * ACEXML/common/Locator.h: Fixed mismatched pre/post inclusions. - - * ACEXML/parser/parser/Parser.cpp: Fixed compilation errors. - -Tue Jul 2 11:53:29 2002 Steve Huston <shuston@riverace.com> - - * ace/config-all.h: Changed definition of ACE_THR_FUNC to incorporate - the platform's thread function return type (DWORD on Windows, - void * on most others). Uses a new typedef ACE_THR_FUNC_RETURN - to specify the platform's thread function return type. - - * ace/OS.{h i cpp} (thr_exit, thr_join): The status arg is now - a ACE_THR_FUNC_RETURN *. - Also moved remaining thread function definition type things to - config-all.h with the rest of them. - - * ace/Base_Thread_Adapter.{h cpp} (invoke(), ace_thread_adapter()): - * ace/OS_Thread_Adapter.{h cpp} - * ace/Thread_Adapter.{h cpp} (invoke(), invoke_i()): - Return type changed from void * to ACE_THR_FUNC_RETURN. - - * ace/Event_Handler.{cpp h} (read_adapter): Change return value from - void* to ACE_THR_FUNC_RETURN. - - * ace/Thread_Hook.{h cpp}: ACE_Thread_Hook::start () returns - ACE_THR_FUNC_RETURN, not void *. - - * ace/Thread.{h i} (exit, join): 'status' arg is now - ACE_THR_FUNC_RETURN *, not void *. - - * ace/Thread_Control.{h inl cpp} (status, exit): - * ace/Thread_Manager.{h cpp} (exit, join): Thread exit status - type changed from void * to ACE_THR_FUNC_RETURN. - - * ace/Task.{h cpp} (ACE_Task_Base): Return type from svc_run now - ACE_THR_FUNC_RETURN. - - * examples/C++NPv1/RT_Thread_Per_Connection_Logging_Server.cpp: - * examples/C++NPv1/Thread_Per_Connection_Logging_Server.{cpp, h}: - * examples/C++NPv2/Select_Reactor_Logging_Server.cpp: - * examples/C++NPv2/Server_Shutdown.cpp: - * examples/C++NPv2/TP_Reactor_Logging_Server.cpp: - * examples/C++NPv2/WFMO_Reactor_Logging_Server.cpp: - * examples/C++NPv2/Client_Logging_Daemon.cpp: - * examples/Connection/blocking/SPIPE-acceptor.cpp: - * examples/IPC_SAP/SOCK_SAP/CPP-inserver.cpp: - * examples/IPC_SAP/SSL_SAP/SSL-server.cpp: - * examples/Reactor/WFMO_Reactor/Abandoned.cpp: - * performance-tests/TCP/tcp_test.cpp: - * tests/ACE_Init_Test.cpp: - * tests/MEM_Stream_Test.cpp: - * tests/Service_Config_DLL.cpp: - * tests/Task_Test.cpp: - * tests/Thread_Pool_Reactor_Resume_Test.cpp: - * tests/Thread_Pool_Reactor_Test.cpp: - Change thread function to return ACE_THR_FUNC_RETURN, not void *. - - * examples/C++NPv2/TP_Logging_Server.cpp (handle_input, svc): - * examples/C++NPv2/AC_Client_Logging_Daemon.cpp (open): - * examples/C++NPv2/display_logfile.cpp (svc): Change - ACE_reinterpret_cast to ACE_static_cast. Thanks to Doug Schmidt - for finding this. - - * examples/C++NPv2/TP_Logging_Server.h: - * examples/C++NPv2/Reactor_Logging_Server_Adapter.cpp: Corrected use - of ACE_NEW_RETURN. - -Mon Jul 1 23:16:47 2002 Jeff Parsons <parsons@cs.wustl.edu> - - * */Makefile: - - Updated dependencies. - -Mon Jul 1 21:59:44 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ACEXML/common/Attributes.h: - * ACEXML/common/AttributesImpl.h: - * ACEXML/common/Attributes_Def_Builder.h: - * ACEXML/common/CharStream.h: - * ACEXML/common/ContentHandler.h: - * ACEXML/common/DTDHandler.h: - * ACEXML/common/DTD_Manager.h: - * ACEXML/common/DefaultHandler.h: - * ACEXML/common/Element_Def_Builder.h: - * ACEXML/common/EntityResolver.h: - * ACEXML/common/Env.h: - * ACEXML/common/ErrorHandler.h: - * ACEXML/common/Exception.h: - * ACEXML/common/FileCharStream.h: - * ACEXML/common/HttpCharStream.h: - * ACEXML/common/InputSource.h: - * ACEXML/common/Locator.h: - * ACEXML/common/LocatorImpl.h: - * ACEXML/common/Mem_Map_Stream.h: - * ACEXML/common/NamespaceSupport.h: - * ACEXML/common/SAXExceptions.h: - * ACEXML/common/StrCharStream.h: - * ACEXML/common/Transcode.h: - * ACEXML/common/URL_Addr.h: - * ACEXML/common/Validator.h: - * ACEXML/common/XMLFilter.h: - * ACEXML/common/XMLFilterImpl.h: - * ACEXML/common/XMLReader.h: - * ACEXML/common/XML_Codecs.h: - * ACEXML/common/XML_Types.h: - * ACEXML/parser/debug_validator/Debug_Attributes_Builder.cpp: - * ACEXML/parser/debug_validator/Debug_Attributes_Builder.h: - * ACEXML/parser/debug_validator/Debug_DTD_Manager.h: - * ACEXML/parser/debug_validator/Debug_Element_Builder.h: - * ACEXML/parser/debug_validator/Element_Tree.h: - * ACEXML/parser/parser/Entity_Manager.h: - * ACEXML/parser/parser/Parser.cpp: - * ACEXML/parser/parser/Parser.h: - - ACE_LACKS_PRAGMA_ONCE fixes. - -Mon Jul 1 06:21:00 2002 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - * ace/Sock_Connect.cpp (count_interfaces): Added a MacOSX-specific - value for OSIOCGIFCONF. Thanks to John Zorko <j.zorko@att.net> - for this fix. diff --git a/ChangeLogs/ChangeLog-03a b/ChangeLogs/ChangeLog-03a index 415de5b54d7..d0f450a8ab1 100644 --- a/ChangeLogs/ChangeLog-03a +++ b/ChangeLogs/ChangeLog-03a @@ -1,3 +1,7 @@ +Thu Feb 20 19:18:17 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> + + * ChangeLog: Shortened the changelog for ease of use. + Thu Feb 20 10:35:15 2003 Ossama Othman <ossama@uci.edu> * ace/Strategies_T.h: @@ -316,8243 +320,4 @@ Thu Feb 20 09:23:51 2003 Ossama Othman <ossama@uci.edu> ACE_SSL_SOCK_{Acceptor,Connector} classes. They are unnecessary. -Wed Feb 19 12:50:58 2003 Heather Drury <oci@cs.wustl.edu> - - * ACE version 5.3.1 released. - -Mon Feb 17 15:21:18 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: The Server_Protocol and the ORB_init - test in RTCORBA will not run in ST builds since the operations - ACE_OS::thr_getprio () are not supported. - -Fri Feb 14 09:53:43 2003 Rich Seibel <seibel_r@ociweb.com> - - * ace/config-osf1-4.0.h: Tru64 lacks the clearerr - function when built without threads (thread=0). - Thanks to Abhay Kulkarni <Abhay.Kulkarni@veritas.com> - for reporting the problem. - -Thu Feb 13 17:50:07 2003 Krishnakumar B <kitty@cse.wustl.edu> - - * include/makeinclude/rules.lib.GNU: Build static libraries - using xargs with the list of all the object files instead of - using them directly on the command line. This works around - brokenness of systems that have a small limit on the length of - the command-line arguments. Thanks to Abhay Kulkarni - <Abhay.Kulkarni@veritas.com> for reporting the problem. - -Thu Feb 13 12:43:27 2003 Pradeep Gore <pradeep@oomworks.com> - - * bin/auto_run_tests.lst: - Added TAO/examples/RTCORBA/Activity/run_test.pl to this list. - -Thu Feb 13 07:36:40 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * THANKS: Added Dmitri Belogaj to the hall of fame. - -Wed Feb 12 12:40:21 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: Prevented a few RTCORBA tests to be run - in our daily builds since changing priorities requires root - access. - -Wed Feb 12 12:06:05 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/Strategies.h: Added this file for backward - compatibility. Backward compatibility was broken with the - following checkin - - "Tue Nov 20 12:48:39 2001 Balachandran Natarajan - <bala@cs.wustl.edu>". Thanks to Dmitri Belogaj - <belogaj@nentec.de> for reporting this problem. - -Wed Feb 12 10:18:21 2003 Nanbor Wang <nanbor@cs.wustl.edu> - - * bin/MakeProjectCreator/templates/vc7.mpd: Checked in Chad's - missing ssl_libs fix. - -Wed Feb 12 08:59:12 UTC 2003 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/UPIPE_Addr.h: - Removed ACE_UPIPE_Addr class which isn't used, typedef ACE_UPIPE_Addr - still is there. - -Tue Feb 11 22:19:39 2003 Krishnakumar B <kitty@cse.wustl.edu> - - * examples/DLL/Makefile.Today (SHLIB): - * examples/DLL/Makefile.Newsweek (SHLIB): - - Removed extra space at the end which caused compilation to - break. Thanks to Bill Cassanova <BCassanova@weather.com> for - reporting the problem. - -Wed Feb 12 00:00:12 UTC 2003 Craig Rodrigues <crodrigu@bbn.com> - - * ace/OS.h: - * ace/OS.i (thr_setprio): Add a new default thr_policy parameter which - allows us to explicitly set the policy in pthread_setschedparam(). - Introduced to eliminate a bug discovered in RT-CORBA implementation. - -Tue Feb 11 14:01:21 2003 Nanbor Wang <nanbor@cs.wustl.edu> - - * vc7_nightly.mwc: Added a temporary file specifying core nightly - build targets. Notice this file includes both ACE and TAO. - - * ace/ace.mpc: Added XML_Svc_Conf.cpp. - - * ace/SSL/ssl.mpc: Fixed dllflags setting. - -Mon Feb 10 18:08:12 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * THANKS: Added Otis Nyandoro to the hall of fame. - -Mon Feb 10 18:02:41 2003 Steve Huston <shuston@riverace.com> - - * ace/OS.h: Don't special-case SEH support on ACE_HAS_WINCE. First, - CE can do SEH; second, the switching of this on/off can and should - be done in the config file. - - * ace/config-win32-common.h: Enable ACE_HAS_WIN32_STRUCTURAL_EXCEPTIONS - for WinCE. - - Thanks to Emmanuel Thevenot Beaufort - <emmanuel.thevenot-beaufort@jci.com> for this fix. - - * ace/Handle_Set.{h i}: Remove the ACE_Handle_Set destructor for - ACE_HAS_WINCE. The need for this predates the supported vc 3 and 4 - compilers, and having it in there triggers an SEH warning. - - * ace/SPIPE_Stream.i: Follow-up to: - Mon Feb 10 09:22:04 2003 Phil Mesnier <mesnier_p@ociweb.com> - Added the need for ACE_HAS_WINSOCK2 to enable this code. The calls - and structures used aren't available in Winsock1. - - * ace/config-wince.h: Disable Winsock2 for CE versions less than 4.0. - -Mon Feb 10 12:14:35 2003 Jeffrey Graham <jgraham@titan.com> - - * bin/ace-install: - - Fixed problem where all TAO executables were not installed. - - Fixed installed file permission problem. Files that should have - global access didn't have it. - - Fixed problem where the script did not easily flatten the - install directory, requires the ACE_ARCH subdirectory trees be - built. - - Fixed problem where all binaries were not filtered out. - -Mon Feb 10 19:42:12 UTC 2003 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/ATM_Acceptor.h: - Converted old style documentation to doxygen style - - * ace/Arg_Shifter.h: - Added private declaration of copy constructor and assignment - operator. Instances of these class can not be copied using the - default ones. Also fixes a MSVC6 level 4 warning. - - * ace/Configuration.cpp: - Removed not needed MingW ifdef, the normal code also compiles with - MingW - - * ace/config-win32.h: - Removed cygwin from this file. For cygwin the config-cygwin32.h - must be included directly in config.h - - * ace/Notification_Strategy.h: - * ace/Notification_Strategy.inl: - Made mask() method const - - * include/makeinclude/ace_flags.bor: - If TAO_ROOT isn't defined, define it as $(ACE_ROOT)/TAO - - * bin/MakeProjectCreator/templates/bor.mpd: - * include/makeinclude/build_dll.bor: - * include/makeinclude/build_exe.bor: - Thanks for Christopher Kohlhoff <chris@kohlhoff.com> for providing - a fix for the singleton template with Borland. The problem is that - with a singleton template the code is implicitly generated into - every object file that references it. The linker then strips out - all but a single copy of the code. When linking an exe that refers - to a singleton template in a dll, the linker would choose one - of the template code instances in the exe's .obj files, and - completely ignore the instance that was exported by the dll. - - The key to the solution is that it appears that the borland linker - chooses the first instance of the template symbols that it sees. The - linker will let you list the .lib files in the place normally used - for object files. By listing the library files before any of the - object files, it resolves the template code against the dll exports. - - * examples/Service_Configurator/Misc/Makefile: - * examples/Service_Configurator/Misc/Makefile.Timer: - * examples/Service_Configurator/Misc/Makefile.main: - Splitted the original GNU makefile in multiple files, one for the - shared library and one for the exe and one to trigger the other - two. This fixes the compile errors with this example with Cygwin - and MingW. Thanks for Vince Mounts <vince@mounts.cc> for making - these makefiles. - -Mon Feb 10 09:22:04 2003 Phil Mesnier <mesnier_p@ociweb.com> - - * ace/SPIPE_Stream.i: Added a special case for win32 platforms to - ACE_SPIPE_Stream::send_handle() and recv_handle() to allow handle - passing. The extension is necessary since the receiving side must - open the duplicate handle before the sending side closes it. - -Fri Feb 7 17:22:39 2003 Steve Huston <shuston@riverace.com> - - * ace/OS.cpp (ACE_OS::thr_create): Moved the pthread_setstack() call - (and the check for whether it's available) from inside the - Pthreads draft 4/6 section to the Pthreads standard section - and corrected the condition it's used (defined - ACE_HAS_PTHREAD_SETSTACK, not !defined ACE_HAS_PTHREAD_SETSTACK). - - Also see: - Thu Jan 16 19:06:19 2003 Stephen Torri <storri@cse.wustl.edu> - -Fri Feb 7 16:55:58 2003 Heather Drury <drury_h@ociweb.com> - - * ace/config-aix-4.x.h: Reverted change to aix config file: - - Fri Feb 7 15:27:40 2003 Heather Drury <drury_h@ociweb.com> - -Fri Feb 7 15:37:19 2003 Steve Huston <shuston@riverace.com> - - * include/makeinclude/platform_aix_ibm.GNU: Define DCCFLAGS with the - debugging options for xlC_r; Change DCFLAGS to just -g (for the C - compiler). Thanks to Trevor Fields <fields_t@ociweb.com> for - reporting this problem. - -Fri Feb 7 15:27:40 2003 Heather Drury <drury_h@ociweb.com> - - * ace/config-aix-4.x.h: Restored the defining of - ACE_TEMPLATES_REQUIRE_SOURCE for AIX VisualAge 5, - which is needed for building TAO. Thanks to Trevor - Fields <fields_t@ociweb.com> for this fix. - -Fri Feb 7 13:04:51 2003 Steve Huston <shuston@riverace.com> - - * ace/ace_wchar.h: Added a new macro, ACE_TEXT_ALWAYS_WCHAR(STRING). - It's the inverse of ACE_TEXT_ALWAYS_CHAR - whatever STRING is, - the macro produces a wide-char string. Thanks to - Emmanuel Thevenot Beaufort <emmanuel.thevenot-beaufort@jci.com> - for this macro. - -Thu Feb 6 19:30:18 2003 Steve Huston <shuston@riverace.com> - - * tests/SString_Test.cpp: Added a test for wide-string compares of - equal-length strings that differ in the last character. - - * ace/String_Base.i (compare): When memcmp()-ing to compare, take - the size of CHAR into account when calculating byte length. Thanks - to Emmanuel Thevenot Beaufort <emmanuel.thevenot-beaufort@jci.com> - for this fix. - -Thu Feb 6 16:06:29 2003 Rich Seibel <seibel_r@ociweb.com> - - * ace/config-tru64.h: Added a guard around the - ACE_HAS_SNPRINTF as versions earlier than 5.0 do not seem - have this function. - -Thu Feb 6 15:03:04 2003 Krishnakumar B <kitty@cs.wustl.edu> - - * include/makeinclude/rules.local.GNU: Make the removal of IDL - generated files in realclean target dependent upon the - definition of IDL_FILES. MingW make crashes on empty "rm -f " - commands. - -Thu Feb 6 15:19:28 2003 Heather Drury <drury_h@ociweb.com> - - * ace/Refcounted_Auto_Ptr.h: - * ace/Refcounted_Auto_Ptr.i: Made the other get() method const. See - previous ChangeLog entry. - -Thu Feb 6 07:13:06 2003 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * ace/Refcounted_Auto_Ptr.{h,i} (get): Made the get() method const. - Thanks to Jenny Kowald <jkowald@agile.tv> for suggesting this. - -Wed Feb 5 20:48:47 2003 Krishnakumar B <kitty@cse.wustl.edu> - - * include/makeinclude/rules.local.GNU: Added rules to generate - object files from C++ file extensions .cxx, .C and .cc, in - addition to .cpp. - -Wed Feb 5 20:15:18 2003 Krishnakumar B <kitty@cse.wustl.edu> - - * include/makeinclude/rules.lib.GNU: Removed the conditional - setting of VSHOBJS and VSHOBJS1 based on PRELIB. There is too - much bogosity here. This should fix HP-UX and AIX breakage. - -Sat Feb 1 08:43:53 2003 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * ace/Stream_Modules.cpp (put): Return 0 rather than -1 to be consistent - wrt the put() semantics elsewhere in ACE. Thanks to Jody Hagins - <jody@atdesk.com> for help with this. - - * tests/FIFO_Test.cpp (server): Only compile this on platforms - that aren't Win32 since FIFO's don't work there. Thanks to Carl - Halvorson <cyberdesk@yahoo.com> for reporting this. - -Wed Feb 5 09:45:12 UTC 2003 Johnny Willemsen <jwillemsen@remedy.nl> - - * include/makeinclude/platform_mingw32.GNU: - Removed desciption to MINGW_BASE. - -Tue Feb 4 13:13:08 2003 Krishnakumar B <kitty@cs.wustl.edu> - - * include/makeinclude/rules.lib.GNU: - * include/makeinclude/rules.bin.GNU: - - Generalized the creation of objects from source files by - removing the dependency that the source files be named *.cpp. - Also removed a lot of junk. - - * include/makeinclude/wrapper_macros.GNU: Removed some more - accumulated cruft. - - * include/makeinclude/rules.nested.GNU: Fix recursive make so - that users can name the makefiles in sub-directories as - something other than Makefile by introducing a new variable - SUBDIR_MAKEFILE. - - * ace/Makefile: Fix the Makefile to use the above extension when - recursing into directories. - - * include/makeinclude/rules.local.GNU: Add a new local target - clean_idl_stubs.local which cleans all the IDL generated files. - This target is linked to realclean so the IDL generated files - will only be cleaned with realclean. - - Thanks to James Mansion <james.mansion@uk.nomura.com> for - reporting the bugs and suggesting some fixes. - -Tue Feb 4 11:21:13 MST 2003 Rob Andzik <andzik@rtlogic.com> - - * ACE_wrappers/include/makeinclude/platform_vxworks5.x_g++.GNU - - Added three new variables - GCC_VERSION (defaults to gcc-2.96) - VXWORKS_VERSION_FLAG (defaults to -DACE_VXWORKS=0x542) - VXWORKS_STDINC_FLAG (defaults to <null>) - - - Modified all sets of GCCLIB_DIR to use GCC_VERSION - - Modified all references to ACE_VXWORKS to use VXWORKS_VERSION_FLAG - - replaced -nostdinc with VXWORKS_STDINC_FLAG - - * ACE_wrappers/ace/config-vxworks5.x.h - - Added define ACE_LACKS_CLEARERR (wrapped in ACE_VXWORKS check) - - * ACE_wrappers/ace/OS.i - - Added ACE_UNUSED_ARG where warnings occurred. - -Mon Feb 3 16:33:02 2003 Heather Drury <drury_h@ociweb.com> - - * ace/config-freebsd.h: msgsnd() with const parameter was recently - added to FreeBSD-CURRENT. Thanks to Craig Rodrigues' for this change. - -Sat Feb 1 09:34:45 2003 Ossama Othman <ossama@uci.edu> - - * bin/auto_run_tests.lst: - - Disable TAO's IORInterceptor test for the GIOP 1.0 build - configuration. The test requires IOR tagged components that are - not supported in GIOP 1.0. Fixes a false positive in our - scoreboard. - -Sat Feb 1 07:50 2003 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * Our deepest sympathies to the families and friends of the crew of - the US Space Shuttle Columbia, which was lost over Texas this - morning. - - "We shall never forget them nor the last time we saw them, as - they prepared for their mission and waved good-bye and slipped - the surly bonds of Earth to touch the face of God." - - - President Reagan (addressing NASA employees following the - tragic loss of the Challenger 7 crew on STS-51L.) - - Ad astra per aspera. - -Sat Feb 1 09:00:00 2003 Steve Huston <shuston@riverace.com> - - * Space Shuttle Columbia breaks up at 200,000ft altitude at - 12,500mph. 7 astronauts killed; the US and Israel mourn their loss. - -Fri Jan 31 12:55:26 2003 Frank Hunleth <frank@hunleth.com> - - * tests/Reactor_Exceptions_Test.cpp (worker): Added #if defined - (ACE_HAS_THREADS) guard around worker () function to remove a - compiler warning on single threaded configurations that have - exceptions enabled. - -Thu Jan 30 21:57:58 2003 Frank Hunleth <frank@hunleth.com> - - * examples/Threads/thread_specific.cpp (worker): Added __OpenBSD__ - to the list of OS/compilers that can check printf arguments. - This removes a compilation warning. - -Thu Jan 30 21:50:27 2003 Frank Hunleth <frank@hunleth.com> - - * ace/config-openbsd.h: Added ACE_LACKS_SETSCHED, since it turns - out that sched_setscheduler hasn't been implemented on OpenBSD. - - * bin/auto_run_tests.lst: Disabled the MT_Cubit test for OpenBSD, - since it fails completely when sched_params returns an error. - -Thu Jan 30 10:34:02 2003 Ossama Othman <ossama@uci.edu> - - From Olivier Brunet <o.brunet@free.fr> - * ace/SSL/SSL_Context.cpp (report_error): - - Corrected format specifier for wide string case. - -Wed Jan 29 12:31:51 2003 Frank Hunleth <frank@hunleth.com> - - * ace/config-openbsd.h: - * ace/config-openbsd-pthreads.h: Merged config files together and - removed config-openbsd-pthreads.h like what was done for - FreeBSD. - - * include/makeinclude/platform_openbsd.GNU: - * include/makeinclude/platform_openbsd_pthreads.GNU: Merged files - together and removed platform_openbsd_pthreads.GNU like - FreeBSD. Also removed gratuitous link of -lstdc++. g++ already - does that for us. Thanks to Craig Rodrigues' changelog entry. - -Wed Jan 29 11:47:29 2003 Steve Huston <shuston@riverace.com> - - * ACEXML/tests/ContentHandler_Test.cpp: Fixed compile errors by - adding/correcting use of the ACEXML exception macros. - -Wed Jan 29 08:30:12 UTC 2003 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/Vector_T.cpp: - Fixed bug in resize() method, of the original vector, the last - item was lost during resize(). Thanks to Denis Parnaland - <denis.parnaland@tde.alstom.com> for reporting this. - - * tests/Vector_Test.cpp: - Extended this test to test the resize() method of ACE_Vector. - - * THANKS: Added Denis Parnaland. - -Tue Jan 28 18:46:41 2003 Steve Huston <shuston@riverace.com> - - * ACEXML/parser/parser/Parser.cpp (parse_element): When converting - an escape sequence, add its length, don't lose the length of - already-parsed content. Thanks to Emmanuel Thevenot Beaufort - <emmanuel.thevenot-beaufort@jci.com> for this fix. - - * ACEXML/tests/ContentHandler_Test.{cpp dsp}: New test that - illustrates the problem above and validates the fix. - - * ACEXML/tests/Tests.dsw: - * ACEXML/Makefile: - * ACEXML/Makefile.bor: Added ContentHandler_Test. - - * THANKS: Added Emmanuel Thevenot Beaufort to the Hall of Fame. - -Tue Jan 28 15:47:29 2003 Steve Huston <shuston@riverace.com> - - * include/makeinclude/platform_hpux_aCC.GNU: For distrib builds, - changed the +h option to not specify the whole path; put the path - in the path search option (+b). - -Mon Jan 27 14:28:16 2003 Heather Drury <drury_h@ociweb.com> - - * include/makeinclude/wrapper_macros.GNU: Remove unnecessary use - of GNU make basename command. Supplied by Ganesh Pai <gpai@sonusnet.com>. - -Sun Jan 26 16:42:23 UTC 2003 Johnny Willemsen <jwillemsen@remedy.nl> - - * include/makeinclude/platform_linux_borland.GNU: - When AIO support is found, set the right define but also add the - library liblrt to the lists to link with. Only when building with - threading support link with the pthread library. - -Sun Jan 26 16:42:23 UTC 2003 Johnny Willemsen <jwillemsen@remedy.nl> - - * include/makeinclude/platform_linux_borland.GNU: - Don't set any BUILD_DLL defines in this file, the BUILD_DLL defines - are set in the specific makefiles. - -Sat Jan 25 20:31:56 2003 Frank Hunleth <frank@hunleth.com> - - * ace/config-openbsd.h: - * ace/config-openbsd-pthread.h: Added ACE_HAS_SNPRINTF since - OpenBSD has snprintf() and to fix buffer overflow tested for by - tests/Log_Msg_Test. - -Sat Jan 25 18:15:48 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: Added a new test to the daily builds. - -Sat Jan 25 13:39:23 UTC 2003 Johnny Willemsen <jwillemsen@remedy.nl> - - * ACE-INSTALL.html: - * include/makeinclude/platform_mingw32.GNU: - * include/makeinclude/platform_gnuwin32_common.GNU: - Removed the usage of MINGW_BASE. This is not needed for using MingW. - Thankx to Viktor Ransmayr <viktor.ransmayr@t-online.de> for - confirming that this is not needed. - -Fri Jan 24 07:54:12 UTC 2003 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/OS.h: - VxWorks 5.4 lacks some ANSI prototypes. These are defined by - VxWorks 5.5 so we only define these ANSI prototypes when defined - ACE_VXWORKS and this is equal or smaller 0x540. Else we just - include apra/inet.h. - -Thu Jan 23 17:57:05 2003 Heather Drury <drury_h@ociweb.com> - - * ace/README: Deleted non-ascii characters around ACE_HAS_PTHREAD_SETSTACK. - -Thu Jan 23 17:44:38 UTC 2003 Craig Rodrigues <crodrigu@bbn.com> - - * include/makeinclude/platform_freebsd.GNU: Remove gratuitous link - of -lstdc++. g++ already does that for us. - -Wed Jan 22 21:08:48 2003 Steve Huston <shuston@riverace.com> - - * examples/C++NPv2/WFMO_Reactor_Logging_Server.cpp (Quit_Handler): - Sync with code in the book. - -Wed Jan 22 18:12:41 2003 Steve Huston <shuston@riverace.com> - - * examples/C++NPv2/Service_Reporter.cpp: - * examples/C++NPv2/Configurable_Logging_Server.cpp: Change the - name of the static service descriptor from Reporter to - Reporter_Descriptor. This is more descriptive and matches - the book. - -Wed Jan 22 18:08:27 UTC 2003 Craig Rodrigues <crodrigu@bbn.com> - - * ace/config-freebsd.h: Add ACE_HAS_SNPRINTF. Fixes Log_Msg_Test. - -Wed Jan 22 18:03:35 UTC 2003 Craig Rodrigues <crodrigu@bbn.com> - - * include/makeinclude/platform_linux.GNU: - * include/makeinclude/platform_linux_icc.GNU: - * include/makeinclude/platform_linux_kcc.GNU: - Remove defines of _POSIX_THREADS and _POSIX_THREAD_SAFE_FUNCTIONS. - These POSIX constants are not meant to be defined by - the user. The user must include <unistd.h> and *check* - the value of these constants. On Linux glibc systems, - after including <unistd.h>, the header <bits/posix_opt.h> - will be eventually be included...these constants are - defined by the glibc implementation in that header. - -Wed Jan 22 10:55:39 2003 Steve Huston <shuston@riverace.com> - - * ace/config-tru64.h: Added ACE_HAS_SNPRINTF. Thanks to Vladimir - Chovanec <Vladimir.CHOVANEC@asset.sk> for this fix. - -Wed Jan 22 01:24:51 2003 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * ace/Message_Queue_T.h: Corrected the documentation for - ACE_Message_Block_Ex so it says ACE_MESSAGE_TYPE rather than - ACE_Message_Block. - -Tue Jan 21 21:44:04 2003 Nanbor Wang <nanbor@cs.wustl.edu> - - * ACEXML/examples/SAXPrint/broken.xml: Added a new broken XML file - to make sure we handle this case properly. - - * ACEXML/parser/parser/Parser.i (get): Added a check for - successful <get> operation from the input CharStream to avoid - out of the infinite loop when reading broken XML files. Thanks - to Volodymyr Orlenko <vorlenko@jaalam.com> for reporting this - and submitting the patch. - -Mon Jan 20 20:32:55 2003 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * ace/Log_Msg.cpp (log): Added a '}' at line 1531 to fix a - nasty compiler bug on Digital UNIX. Thanks to Igor Pisarenko - <Igor.Pisarenko@computershare.com.au> for reporting this. - -Tue Jan 21 19:29:04 2003 Heather Drury <drury_h@ociweb.com> - - * bin/auto_run_tests.lst: Disabled orbsvcs/Test/EC_MT_MCast on - static builds. - -Tue Jan 21 08:43:24 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/generate_doxygen.pl: Fixed a cut and paste error. - -Mon Jan 20 16:22:43 2003 Heather Drury <drury_h@ociweb.com> - - * ace/NT_Service.cpp: Modified return statement as the - wait_for_service_state function is declared as void and it was - returning an int. - -Mon Jan 20 15:48:35 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: Added a TAO test to the daily builds. - -Mon Jan 20 13:58:25 2003 Rich Seibel <seibel_r@ociweb.com> - - * ace/config-macosx.h: Added ACE_HAS_SNPRINTF as well as - ACE_LACKS_PERFECT_MULTICAST_FILTERING to make the ACE - Mac OSX 10.2.x build pass the Log_Msg_Test and the - Multicast_Test. Thanks to John Michael Zorko <j.zorko@att.net> - for providing this patch. - -Mon Jan 20 18:59:12 UTC 2003 Johnny Willemsen <jwillemsen@remedy.nl> - - * examples/C++NPv2/Makefile.CLD: - * examples/C++NPv2/Makefile.SLD: - * examples/C++NPv2/Makefile.SLDex: - * examples/C++NPv2/Makefile.TPCLS: - * examples/C++NPv2/Makefile.TPLS: - * ace/RMCast/Makefile: - Set the correct x_BUILD_DLL defines. These are needed for the - Cygwin/MingW and Kylix compiler. Thanks to Vince Mounts - <vince@mounts.cc> for providing the patches. - -Mon Jan 20 11:58:00 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/generate_doxygen.pl: Fixed the spaces in the file names - of generated man pages. Thanks to Olver Kellogg for providing - this patch. - -Sun Jan 19 11:40:10 2003 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * ace/NT_Service.{h,cpp}: Enhanced the wait_For_service_state() so - that it actually uses the wait_time parameter. Thanks to Theo - Landman <tlandman@justcroft.com> for contributing this. - - * ace/Process_Manager.cpp (spawn): Cleanup dynamically allocated - memory if the spawn() fails. Thanks to Kobi Cohen-Arazi - <kobi@mivzak.com> for this fix. - -Sun Jan 19 15:03:13 UTC 2003 Johnny Willemsen <jwillemsen@remedy.nl> - - * websvcs/lib/Makefile: - Set the correct x_BUILD_DLL defines. These are needed for the - Cygwin/MingW and Kylix compiler. - -Sat Jan 18 17:49:20 2003 Steve Huston <shuston@riverace.com> - - * Makefile: Removed PACE references. - -Sat Jan 18 17:11:21 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/make_release: Use doxygen 1.2.18 as opposed to doxygen - 1.2.13.1. - -Fri Jan 17 09:00:17 2003 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * ACE-INSTALL.html (make): Removed all claims that we support - SunC++ 4.2. Thanks to Kirat Singh <kirat.singh@gs.com> for - reporting this. - -Sat Jan 18 09:07:12 UTC 2003 Johnny Willemsen <jwillemsen@remedy.nl> - - * tests/Makefile.DLL_Test: - * tests/Makefile.Framework_Component_DLL: - * tests/Makefile.Service_Config_DLL: - * examples/Service_Configurator/IPC-tests/server/Makefile: - Set the correct x_BUILD_DLL defines. These are needed for the - Cygwin/MingW and Kylix compiler. - -Fri Jan 17 12:34:51 2003 Steve Huston <shuston@riverace.com> - - * ACE-INSTALL.html: Updated platform and compiler information. - -Thu Jan 16 19:06:19 2003 Stephen Torri <storri@cse.wustl.edu> - - * ace/OS.cpp: - * ace/config-linux-common.h: - * ace/config-freebsd.h: - * ace/README: - - Added a new macro called ACE_HAS_PTHREAD_SETSTACK for systems - that use the newer glibc (2.3+). Eliminates linker warning about - not using the pthread function called pthread_setstack. - -Thu Jan 16 16:25:45 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: Added a TAO/orbsvcs/tests/EC_MT_Mcast - test to the daily builds. - -Thu Jan 16 12:27:24 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * COPYING: Extended the validity of the copyright notice from 2002 - to 2003. Should have been done yesterday, anyway, better late - than never. - -Wed Jan 15 18:02:01 2003 Bala <bala@cs.wustl.edu> - - * ACE version 5.3 released. - -Wed Jan 15 10:48:11 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/make_release: Did the following enhancements - - - If it is a minor release do not append the beta version. Looks - like somone broke this and the script started writing out - things as x.3.0, whcih is bad BTW. - - - We now POSIX.1 tar, instead of old tar format with the -H - option in cpio. This helps to tar files deep in the hierarchy - tree to be properly zipped. We were loosing this in many - betas. - - - Fixes for the generated filenames copied into the - previous_versions directory. - - - Fix to get the right Version.h file in TAO. - - -Tue Jan 14 18:51:49 2003 Steve Huston <shuston@riverace.com> - - * ace/Auto_Ptr.h: Added pragma to disable warning 4284 for Microsoft - compilers. The warned-of behavior is exactly what is intended. - - * bin/MakeProjectCreator/templates/em3vcpdll.mpt: - * bin/MakeProjectCreator/templates/em3vcpdllexe.mpt: - * bin/MakeProjectCreator/templates/em3vcplib.mpt: - * bin/MakeProjectCreator/templates/em3vcplibexe.mpt: - Added a "WCE emulator" configuration section. This allows builds - for the WinCE emulator, necessary for working with the WinCE.NET - Standard SDK emulator. - -Tue Jan 14 14:56:12 2003 Steve Huston <shuston@riverace.com> - - * ace/config-win32-common.h: For WinCE, don't include wce.h unless - building with MFC. wce.h is an MFC file. - - * ace/config-WinCE.h: Add pragmas to link _with_ corelibc.lib and - _without_ oldnames.lib. These weren't needed with WinCE 3/PocketPC - (but they don't hurt) and are needed for WinCE .NET. The ATL headers - do these pragmas, but we don't want to add a dependency on ATL. - -Tue Jan 14 11:07:06 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/generate_export_file.pl: Added a patch from - Gonzalo Diethelm <gonzalo.diethelm@aditiva.com> which generates - an empty CVS Id string. - -Mon Jan 13 12:53:12 UTC 2003 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/config-win32-common.h: - The MingW compiler can handle the export/import macros. - - * ace/config-cygwin32.h: - Corrected the guards around this header file. - -Sat Jan 11 09:37:24 2003 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * ace/OS_String.cpp (wcslen_emulation): Fixed a nasty formatting - glitch that made the code hard to read. Thanks to Ruslan - Zasukhin <sunshine@public.kherson.ua> for reporting this. - -Fri Jan 3 21:46:00 2003 John Michael Zorko <j.zorko@att.net> - - * netsvcs/clients/Naming/Client/Makefile: Added $(ACELIB) to - ACE_SHLIBS to fix multple -lACE problem. This wasn't an - issue on Mac OS X 10.2, but could be on other platforms. - - * examples/ASX/CCM_App/Makefile: same as above - * netsvcs/clients/Naming/Dump_Restore/Makefile: same as above - -Wed Jan 8 19:27:49 2003 Steve Huston <shuston@riverace.com> - - * ace/Message_Block.h: Doxygenized comments for release() and copy(). - - * ace/UPIPE_Stream.{h cpp}: Removed the remaining_ member; use the - ACE_Message_BLock::length() member function to find the amount - of remaining data. - -Wed Jan 8 12:17:10 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/ace-install: Fixed the improper usage of "or" in the "find " - command used to find the inlined files with extensions .i and - .inl. Thanks to Oliver Kellog for providing the patch. - -Tue Jan 7 13:08:12 UTC 2003 Johnny Willemsen <jwillemsen@remedy.nl> - - * ACEXML/common/Makefile: - Set ACEXML_BUILD_DLL when building the shared lib. - -Mon Jan 6 18:31:09 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/generate_topinfo_charts.sh: Removed the range for the - Y-AXIS, for the graphs to auto-adjust while plotting. - -Mon Jan 6 16:56:04 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/topinfo_iorsize_stats.sh: Scripts for capturing the size of - the server and client when shipping large IOR's - - * bin/topinfo_stats.sh: Renamed to topinfo_simple_stats.sh. This - is a better to show that we capture the sizes of a simple server - and client. - - * bin/topinfo_simple_stats.sh: New file. We need to do some code - sharing between these two scripts. Just postponing that for the - timebeing. - -Mon Jan 6 11:23:12 UTC 2003 Johnny Willemsen <jwillemsen@remedy.nl> - - * include/makeinclude/platform_gnuwin32_common.GNU: - Removed setting ACE_OS_HAS_DLL to 1, this should be done more - specific in the makefiles - -Sat Jan 4 18:46:19 2003 Steve Huston <shuston@riverace.com> - - * ace/config-aix5.1.h: Add a #define ACE_AIX_VERS 501 if ACE_AIX_VERS - isn't already defined. Helps to build with Visual Age C++ incremental - mode. Support for incremental mode is going away after ACE 5.3, but - this crutch can be used for now. - - * ace/Makefile.ace: Remove SOCK_Dgram_Mcast_T from TEMPLATE_FILES. See - Tue Jun 25 23:22:09 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * ace/ace.icc: Corrected the list of source files to build ACE with. - In addition to updating the source list, if building with Visual - Age C++ 5, add the TEMPLATE_FILES to the sources that get compiled. - Visual Age C++ 5 incremental needs to see them all. - - * examples/C++NPv1/Makefile: Add .NOTPARALLEL since some of the - sub-makes share object files and they can get confused. - - * tests/makeicc.pl: Add additional template-getting stuff for - Visual Age C++ 5. - - * tests/ARGV_Test.icc: - * tests/Aio_Platform_Test.icc: - * tests/Arg_Shifter_Test.icc: - * tests/Atomic_Op_Test.icc: - * tests/Auto_IncDec_Test.icc: - * tests/Barrier_Test.icc: - * tests/Basic_Types_Test.icc: - * tests/Bound_Ptr_Test.icc: - * tests/Buffer_Stream_Test.icc: - * tests/CDR_Array_Test.icc: - * tests/CDR_File_Test.icc: - * tests/CDR_Test.icc: - * tests/Cache_Map_Manager_Test.icc: - * tests/Cached_Accept_Conn_Test.icc: - * tests/Cached_Allocator_Test.icc: - * tests/Cached_Conn_Test.icc: - * tests/Capabilities_Test.icc: - * tests/Codecs_Test.icc: - * tests/Collection_Test.icc: - * tests/Config_Test.icc: - * tests/Conn_Test.icc: - * tests/DLL_Test.icc: - * tests/DLL_Test_Impl.icc: - * tests/DLList_Test.icc: - * tests/Date_Time_Test.icc: - * tests/Dev_Poll_Reactor_Test.icc: - * tests/Dirent_Test.icc: - * tests/Dynamic_Priority_Test.icc: - * tests/Enum_Interfaces_Test.icc: - * tests/Env_Value_Test.icc: - * tests/FIFO_Test.icc: - * tests/FlReactor_Test.icc: - * tests/Framework_Component_DLL.icc: - * tests/Framework_Component_Test.icc: - * tests/Future_Set_Test.icc: - * tests/Future_Test.icc: - * tests/Get_Opt_Test.icc: - * tests/Handle_Set_Test.icc: - * tests/Hash_Map_Bucket_Iterator_Test.icc: - * tests/Hash_Map_Manager_Test.icc: - * tests/High_Res_Timer_Test.icc: - * tests/INET_Addr_Test.icc: - * tests/IOStream_Test.icc: - * tests/Lazy_Map_Manager_Test.icc: - * tests/Log_Msg_Test.icc: - * tests/Logging_Strategy_Test.icc: - * tests/MEM_Stream_Test.icc: - * tests/MM_Shared_Memory_Test.icc: - * tests/MT_Reactor_Timer_Test.icc: - * tests/MT_Reactor_Upcall_Test.icc: - * tests/MT_SOCK_Test.icc: - * tests/Malloc_Test.icc: - * tests/Map_Manager_Test.icc: - * tests/Map_Test.icc: - * tests/Max_Default_Port_Test.icc: - * tests/Mem_Map_Test.icc: - * tests/Message_Block_Test.icc: - * tests/Message_Queue_Notifications_Test.icc: - * tests/Message_Queue_Test.icc: - * tests/Message_Queue_Test_Ex.icc: - * tests/Multicast_Test.icc: - * tests/Naming_Test.icc: - * tests/New_Fail_Test.icc: - * tests/Notify_Performance_Test.icc: - * tests/OS_Test.icc: - * tests/Object_Manager_Test.icc: - * tests/Obstack_Test.icc: - * tests/OrdMultiSet_Test.icc: - * tests/Pipe_Test.icc: - * tests/Priority_Buffer_Test.icc: - * tests/Priority_Reactor_Test.icc: - * tests/Priority_Task_Test.icc: - * tests/Proactor_Scatter_Gather_Test.icc: - * tests/Proactor_Test.icc: - * tests/Proactor_Timer_Test.icc: - * tests/Process_Manager_Test.icc: - * tests/Process_Mutex_Test.icc: - * tests/Process_Strategy_Test.icc: - * tests/RB_Tree_Test.icc: - * tests/Reactor_Dispatch_Order_Test.icc: - * tests/Reactor_Exceptions_Test.icc: - * tests/Reactor_Notify_Test.icc: - * tests/Reactor_Performance_Test.icc: - * tests/Reactor_Timer_Test.icc: - * tests/Reactors_Test.icc: - * tests/Reader_Writer_Test.icc: - * tests/Recursive_Condition_Test.icc: - * tests/Recursive_Mutex_Test.icc: - * tests/Refcounted_Auto_Ptr_Test.icc: - * tests/Reverse_Lock_Test.icc: - * tests/SOCK_Connector_Test.icc: - * tests/SOCK_Send_Recv_Test.icc: - * tests/SOCK_Test.icc: - * tests/SPIPE_Test.icc: - * tests/SString_Test.icc: - * tests/SV_Shared_Memory_Test.icc: - * tests/Semaphore_Test.icc: - * tests/Service_Config_DLL.icc: - * tests/Service_Config_Test.icc: - * tests/Signal_Test.icc: - * tests/Sigset_Ops_Test.icc: - * tests/Simple_Message_Block_Test.icc: - * tests/Svc_Handler_Test.icc: - * tests/TP_Reactor_Test.icc: - * tests/TSS_Test.icc: - * tests/Task_Test.icc: - * tests/Thread_Manager_Test.icc: - * tests/Thread_Mutex_Test.icc: - * tests/Thread_Pool_Reactor_Resume_Test.icc: - * tests/Thread_Pool_Reactor_Test.icc: - * tests/Thread_Pool_Test.icc: - * tests/Time_Service_Test.icc: - * tests/Time_Value_Test.icc: - * tests/Timeprobe_Test.icc: - * tests/Timer_Queue_Test.icc: - * tests/TkReactor_Test.icc: - * tests/Token_Strategy_Test.icc: - * tests/Tokens_Test.icc: - * tests/UPIPE_SAP_Test.icc: - * tests/Upgradable_RW_Test.icc: - * tests/Vector_Test.icc: - * tests/XtReactor_Test.icc: Mostly updated to build with Visual Age - C++ 5 in incremental mode. Most of the tests that use templates - don't build. They require explicitly including the needed template - source files names in the .icc file. I don't have time to add them - all now - if anyone uses this compiler, please contact me. No wonder - IBM ditched this piece of junk. - -Fri Jan 3 17:58:21 2003 Steve Huston <shuston@riverace.com> - - * ace/SSL/SSL_Context.cpp: - * netsvcs/lib/Name_Handler.cpp: - * netsvcs/lib/Server_Logging_Handler.cpp: Add __hpux to condition - where template members are explicitly instantiated. Also see: - Tue Dec 3 20:47:39 2002 Steve Huston <shuston@riverace.com> - -Fri Jan 3 17:45:29 2003 Steve Huston <shuston@riverace.com> - - * include/makeinclude/platform_hpux_gcc.GNU: Use HP's /usr/bin/ar - instead of GNU ar (which may be picked up first in PATH). GNU ar - crashes - missing libfl.sl. - - * performance-tests/Misc/test_singleton.cpp: Add __hpux to conditions - for explicitly instantiating ACE_Singleton::singleton_. - -Fri Jan 3 07:49:09 2003 Chad Elliott <elliott_c@ociweb.com> - - * bin/aix_shr: - - Changed the name of the AIX template instantiation macro and - commented on why it's needed. - -Thu Jan 2 17:44:29 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/config-cygwin32-common.h: Set the custom export - macro's. These will be used when building dynamically with the - cygwin compiler. - - * bin/MakeProjectCreator/templates/gnu.mpd: Generate dllflags and - libflags into the GNU makefiles. These are important for the - Cygwin/MingW and Kylix compiler. - -Thu Jan 2 16:11:38 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/config-freebsd.h: Fix use of macros for FreeBSD 4.6: - ACE_LACKS_RWLOCK_T, ACE_LACKS_READDIR_R, ACE_LACKS_SETSCHED, - ACE_LACKS_PTHREAD_THR_SIGSETMASK. Do not define SCHED_RR, - SCHED_IO, SCHED_FIFO, SCHED_OTHER for FreeBSD >= 4.0, since they - are defined in sched.h. Thanks to Craig Rodrigues for providng - the patch. - -Thu Jan 2 13:33:01 2003 Chad Elliott <elliott_c@ociweb.com> - - * ace/Refcounted_Auto_Ptr.i: - - Added #include of ace/Log_Msg.h for ACE_ASSERT. - - * bin/aix_shr: - - Added -DTAO_INSTANTIATING_TEMPLATES to the link command line. - -Thu Jan 2 12:38:55 2003 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/Makefile.ace: Set the dllflags in Makefile. This makes it - possible to do __declspec (dllexport) for the symbols in the - dll. This is needed for the Cygwin/MingW and Kylix compiler. - - Thid change takes care of the static and dynamic library cases. - -Thu Jan 2 08:57:26 2003 Johnny Willemsen <jwillemsen@remedy.nl> - - * include/makeinclude/platform_cygwin32.GNU: Add - --enable-auto-import to the linker flags. This will result in - the linker doing auto-import without giving warnings about - it. - -Wed Jan 1 14:26:09 2003 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/topinfo_stats.sh: Forgot that we run this scripts in cron - environment. Set the path and other details that would help to - ease things. - -Tue Dec 31 21:21:26 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/generate_topinfo_charts.sh: - * bin/topinfo_stats.sh: Scripts to track the memory usage and plot - them. At present it works only for the simple test in - $TAO_ROOT/performance-tests/Memory/Single_Threaded. - -Tue Dec 31 17:57:19 2002 Steve Huston <shuston@riverace.com> - - * examples/Service_Configurator/IPC-tests/server/Handle_Thr_Stream.h: - Add #include "ace/Svc_Handler.h" - -Tue Dec 31 17:50:39 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.h: For WinCE 4 (.NET) and up, include <stddef.h> rather than - typedef-ing ptrdiff_t. - -Tue Dec 31 17:26:41 2002 Steve Huston <shuston@riverace.com> - - * tests/tests.mpc (DLL Test Lib): Add dllflags = ACE_BUILD_SVC_DLL to - add the correct /D option. The source uses ACE_Svc_Export to flag - dll import/export correctly. - - * tests/CE_fostream.h: #include "ace/config-all.h" before checking for - #if defined (ACE_HAS_WINCE). - -Tue Dec 31 15:05:42 2002 Steve Huston <shuston@riverace.com> - - * ACEXML/apps/svcconf/Svcconf.cpp: Changed the condition for when - a dummy entrypoint is needed from AIX 4.3 to AIX and Visual Age C++ - version 5. Thanks for Jon Wackley for this adjustment. Also see: - Fri Dec 27 19:46:31 2002 Steve Huston <shuston@riverace.com> - - * netsvcs/clients/Naming/Client/Makefile: If building this stuff, - reset OBJDIRS to $(VDIR) and $(VSHDIR) to clean up both. - -Mon Dec 30 21:30:41 2002 Steve Huston <shuston@riverace.com> - - * include/makeinclude/platform_hpux_gcc.GNU: Removed the SOVERSION= - line to allow versions libraries to be set up properly. - - * ace/Based_Pointer_Repository.cpp: - * ace/Dynamic.cpp: - * ace/SString.cpp: - * tests/Framework_Component_DLL.cpp: - * tests/MEM_Stream_Test.cpp: - * tests/Process_Strategy_Test.cpp: - * tests/Timeprobe_Test.cpp: - * apps/drwho/File_Manager.cpp: - * examples/IPC_SAP/SOCK_SAP/CPP-inclient.cpp: - * examples/IPC_SAP/SOCK_SAP/CPP-inserver-fancy.cpp: - * examples/Logger/Acceptor-server/server_loggerd.cpp: - * examples/Logger/simple-server/server_loggerd.cpp: - * examples/Threads/auto_event.cpp: - * examples/Threads/tss2.cpp: - Add __hpux to the conditions needed to instantiate - static template members. Also see - Sat Nov 23 18:12:37 2002 Steve Huston <shuston@riverace.com> - -Mon Dec 30 19:32:18 2002 Steve Huston <shuston@riverace.com> - - * tests/Codecs_Test.cpp: Add explicit template instantiation for - ACE_Auto_Basic_Array_Ptr<ACE_Byte>. - -Mon Dec 30 18:44:20 Steve Huston <shuston@riverace.com> - - * examples/C++NPv2/TP_Logging_Server.cpp: - * examples/IPC_SAP/SOCK_SAP/CPP-inclient.cpp: - * examples/IPC_SAP/SOCK_SAP/CPP-inserver-fancy.cpp: - * examples/Logger/Acceptor-server/server_loggerd.cpp: - * examples/Logger/simple-server/server_loggerd.cpp: - * examples/Threads/auto_event.cpp: - * examples/Threads/tss2.cpp: - * performance-tests/Misc/test_singleton.cpp: - More singleton_ instantiations for g++ on AIX. - -Mon Dec 30 18:11:39 2002 Steve Huston <shuston@riverace.com> - - * apps/drwho/File_Manager.cpp: Add explicit instantiation of - ACE_Singleton<File_Manager, ACE_Null_Mutex>::singleton_. - Similar to those done for: - Tue Dec 3 20:47:39 2002 Steve Huston <shuston@riverace.com> - -Mon Dec 30 17:38:31 2002 Steve Huston <shuston@riverace.com> - - * examples/Service_Configurator/IPC-tests/server/Handle_Thr_Stream.cpp: - For Visual Age C++ 5 and up on AIX, explicitly instantiate the - CLI_Stream<> constructor. The compiler misses this one when - instantiating templates. Thank you very much to Jonathan - Wackley <jonw@mountaincable.net> for diagnosing this problem and - leading to a workaround. - -Mon Dec 30 15:25:26 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ACEXML/common/StrCharStream.cpp (open): - * ACEXML/common/FileCharStream.cpp (open): Changed to allow open - to succeed if automatic determination of encoding fails. This - allows the "Simple parsing" feature of ACEXML to work. Thanks - to Michael F"olsl <michael.foelsl@gmx.net> for reporting the - problem. - - * ACEXML/examples/SAXPrint/main.cpp: Added a new option to test - out ACEXML's simpling parsing feature. - - * ACEXML/examples/SAXPrint/simple.svcconf.xml: Added this new - "simple" XML document to test out the "Simple" parsing feature - in ACEXML. - -Mon Dec 30 12:21:33 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/MakeProjectCreator/config/taoversion.mpb: Fixed the version - information. This should automatically get updated during the - next beta/release. - -Sun Dec 29 18:27:38 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: Prevented BiDirGIOP tests from running - in GIOP 10 builds. - -Sun Dec 29 16:43:03 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ACEXML/common/FileCharStream.cpp (close): Changed to reset the - file pointer, filename_ and encoding_ strings after they are - release. Thanks to Andrew Marlow <apm35@student.open.ac.uk> for - suggesting the change. - -Sun Dec 29 00:47:58 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: Security tests using SSLIOP will not be - run in static builds. - -Sat Dec 28 23:05:51 2002 Steve Huston <shuston@riverace.com> - - * bin/make_release: Use -relative to replace ACE_ROOT with a relative - path in generated makefiles. - -Sat Dec 28 13:16:24 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/MEM_IO.cpp (recv_buf): Reverted my previous hack in: - - Fri Dec 27 08:38:29 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - and make the connection close handling more robust. - - * tests/MEM_Stream_Test.cpp: Reverted my previous change in: - - Sat Dec 28 11:04:38 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - It was not necessary. Added debug information for close - connection. - -Sat Dec 28 11:04:38 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * tests/MEM_Stream_Test.cpp: Made sure this test ignore the EIO - error from the following change when a connection was closed. - - Fri Dec 27 08:38:29 2002 Nanbor Wang <nanbor@cs.wustl.edu> - -Fri Dec 27 20:06:39 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/Malloc_T.h: - * ace/Malloc_T.i: - * ace/Malloc_T.cpp: Added a new ACE_Malloc_T::bad () method to - query whether a Malloc_T object was constructed successfully. - This approach, IMO, is sub-optimal compared to using an explicit - init method approach. The existing constructor signatures, - however, prevent us from doing so without breaking backward - compatibility. - - * ace/MEM_SAP.cpp: Now that there's a way to check whether the - underlying Malloc is functioning or not. Changed MEM_SAP to - check for proper initialization of underlying Malloc object and - clean up OS resources if the Malloc object is defected. This - change makes MEM_Stream transport mechanism more robust. - -Fri Dec 27 19:46:31 2002 Steve Huston <shuston@riverace.com> - - * ACEXML/apps/svcconf/Svcconf.cpp: Add a dummy entrypoint when - ACE_USES_CLASS_SVC_CONF == 0 on AIX 4.3. Else the AIX 4.3 linker - complains there are no csects or exported symbols saved at link - time. Thanks to Jon Wackley <jonw@mountaincable.net> for reporting - this problem. - -Fri Dec 27 17:43:29 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.h: For WinCE, only typedef ptrdiff_t for _WIN32_WCE < 400. - eMbedded Visual C++ 4 (Win CE.NET) defines this type. - - * bin/MakeProjectCreator/templates/em3vcp.mpd: Removed /M$(CECrtMT) - from the CPP options. Neither evc3 nor 4 support any /M options to - set the C run-time like MSVC does. It's controlled via the _DEBUG - macro only. This also allows the generated vcp file to be used - with either evc3 or evc4 (Windows CE .NET). evc3 ignored the option, - but evc4 flags it as unknown. - -Fri Dec 27 08:38:29 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/MEM_IO.cpp (recv_buf): Changed to return error (-1) and set - the errno to EIO when we receive 0 byte off the signaling socket - as this should never happen under normal circumstance. However, - on Win32 and Solaris, when one end of a connection closes the - socket while the other end is still reading the socket, the - active end will continuously receive several 0 bytes messages. - This confuses the hack of the MEM_IO mechanism and causes the - memory management routine to fall into a tight spinning - deadlock. - -Fri Dec 27 07:49:28 2002 Craig Rodrigues <crodrigu@bbn.com> - - * ace/config-macosx.h: Remove temporary hacks for buggy - gcc 3.1 compiler. Thanks to John Zorko for reminding me about - this. - - * examples/Service_Configurator/IPC-tests/server/Makefile: - * examples/Service_Configurator/Misc/Makefile: - * examples/Timer_Queue/Makefile: - Add ACELIBS to ACE_SHLIBS to fix MacOS X linking problem. - -Thu Dec 26 17:11:54 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/make_release: Fixed the following problems in the release - script which caused problems during the beta cutting process: - - - cvs commits of TAO/Changelog TAO/VERSION etc. failed. The - failure was because we were also commiting taoversion.mpb but - from $TAO_ROOT as "cvs commit ChangeLog VERSION - other_files ../bin/Makeyada". Using ../ to indicate files in - higher directories is an unpardonable mistake for cvs - 1.10.x. It used to work 1.9.x. But with 1.10* versions the - protocol changed disallowing the ../ - - - we didnt set the file permissions on the generated files - properly - - - the generated files from mpc were not added to the list of - files for which LF->CRLF conversions are supposed to be - masked during zip! - -Thu Dec 26 16:17:41 2002 Steve Huston <shuston@riverace.com> - - * tests/tests.mpc: For DLL Test Lib, the source is DLL_Test_Impl.cpp, - not DLL_Test.cpp. - -Thu Dec 26 09:14:38 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * Makefile: Removed configure.in from the list of controlled files. - -Thu Dec 26 07:36:10 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * include/makeinclude/platform_freebsd_pthread.GNU: Removed for - sure now. This should have been removed in this "Thu Nov 14 - 22:47:45 UTC 2002 Craig Rodrigues <crodrigu@bbn.com>" - checkin. Again, thanks to Craig for catching this. - -Thu Dec 26 07:31:42 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * apps/Gateway/Gateway/Makefile: Fixed a ttypo. Thanks to Craig - Rodrigues for catching this. - -Wed Dec 25 17:47:12 2002 Bala <bala@cs.wustl.edu> - - * ACE version 5.2.8 released. - -Wed Dec 25 17:39:05 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/make_release: Fixed problems with generation of vcproj etc. - -Wed Dec 25 06:14:17 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * Merry Christmas to all. May we have peace and justice for all in 2003. - -Tue Dec 24 17:29:11 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * docs/ACE-subsets.html: Documented the compiler options that - Steve Vranyes experimented and tested with to get a good - footprint reduction on Solaris. - -Tue Dec 24 07:31:54 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/Argv_Type_Converter.h: - * ace/Argv_Type_Converter.cpp: Reverted the following changes "Mon - Dec 23 17:26:42 2002 Steve Huston <shuston@riverace.com>" and - "Mon Dec 23 22:57:18 2002 Balachandran Natarajan - <bala@isis-server.isis.vanderbilt.edu>". The argument parsing in - TAO gets busted due to these changes and most of tests in our - daily builds fail. - -Mon Dec 23 22:57:18 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/Argv_Type_Converter.h: Changed saved_argc_ to be of type int - as opposed to int &. This seems to be necessitated by the change - "Mon Dec 23 17:26:42 2002 Steve Huston <shuston@riverace.com>" - which gives compile errors on all platforms. - -Mon Dec 23 21:07:20 2002 Steve Huston <shuston@riverace.com> - - * bin/MakeProjectCreator/templates/em3vcpdll.mpt: Changed the Debug - define from DEBUG to _DEBUG, matching all others, and what the - ACE files expect. Fixes mismatch inlines in ACE tests. - - * tests/tests.mpc: Added CE_fostream.cpp to the CDR_File_Test inputs. - - * tests/CE_fostream.{cpp h}: Added #if defined (ACE_HAS_WINCE) to - prevent trying to compile this on non-CE systems. - -Mon Dec 23 17:26:42 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.h (main): For WinCE, eVC can't grok the #include within the - 'main' macro expansion, so #include "ace/Argv_Type_Converter.h" is - moved outside the macro definition. - - * ace/Argv_Type_Converter.{h cpp} (ctor): Change argc from int& to - int, else it can't be used in the way it's already coded into use - in the 'main' definition (above, OS.h). - - * tests/Dev_Poll_Reactor_Test.cpp: - * tests/Proactor_Test.cpp: - * tests/Proactor_Scatter_Gather_Test.cpp: Changed main() to - ACE_TMAIN(). - - * tests/FlReactor_Test.cpp: - * tests/XtReactor_Test.cpp: Add /**/ between #include and filename - to silence MS warnings about missing files during dependency updates. - -Mon Dec 23 16:16:05 2002 Boris S Kolpackov <boris@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: Added - TAO/orbsvcs/tests/Security/Crash_Test to the list of tests. - -Mon Dec 23 17:08:17 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * THANKS: Changed Boris's information in the Hall of Fame. - -Mon Dec 23 12:13:51 2002 Steve Huston <shuston@riverace.com> - - * examples/C++NPv1/Process_Per_Connection_Logging_Server.h: Add - 'public' keyword to Process_Per_Connection_Logging_Server's - inheritance from Logging_Server. - - * examples/C++NPv2/Makefile: Add a $(RM) -rf tempinc after each - mkfile make to clean up template instantiations for Visual Age C++ - on AIX. - - * examples/C++NPv2/Makefile.SLD: Add Logging_Handler to the files - comprising this service. - - * examples/C++NPv2/Makefile.SLDex: Added Logging_Acceptor, - Logging_Event_Handler, and Logging_Handler to the files for this - service. Removed the TEMPINC trick for Visual Age C++ to keep the - templates from each compile around for the link stage where they're - needed. - - * examples/C++NPv2/Makefile.CLD: Added Logging_Handler to files. Remove - TEMPINC destruction from each compile. - - * examples/C++NPv2/Makefile.TPLS: Add Logging_Handler to files. Remove - TEMPINC destruction from each compile. - -Mon Dec 23 14:05:21 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/config-cygwin32-common.h: - When ACE_IOV_MAX is not defined, define it is 64. The Cygwin header - files define IOV_MAX as MAX_INT-1 and ACE_IOV_MAX is defined as - IOV_MAX by default. Allocating an array of this size on the stack - results in compile errors, so we define ACE_IOV_MAX to a smaller and - more realistic number. - -Mon Dec 23 12:16:33 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/RMCast/RMCast_IO_UDP.cpp: - * ace/RMCast/RMCast_Fragment.cpp: - Use ACE_IOV_MAX instead of IOV_MAX. - -Mon Dec 23 12:02:33 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * bin/ace_ld: - Extended this perl script so that the location of the __ctordtor.o - can be passed as -o command line option. The location is dependent - on the $VDIR in the vxworks make files. - - * include/makeinclude/platform_vxworks5.x_diab.GNU: - * include/makeinclude/platform_vxworks5.x_g++.GNU: - When calling the ace_ld perl script, pass the $VDIR as command line option - -o to the script. - -Sun Dec 22 11:41:56 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/Connector.cpp: Fixed fuzz errors. - -Sun Dec 22 11:37:44 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * THANKS: Added Daniel White to the hall of fame! - -Sun Dec 22 11:31:53 2002 Craig Rodrigues <crodrigu@bbn.com> - - * apps/Gateway/Gateway/Makefile: - * apps/Gateway/Peer/Makefile: Add $(ACELIB) to $(ACE_SHLIBS) so - that -lACE is added to the linkline when a shared library is - built. Solves linking problem MacOS X. Thanks to John Zorko and - Daniel white <ygor@comcast.net> for pointing this out. - -Sun Dec 22 10:25:35 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/Connector.cpp: Fixed a warning with g++. - -Sun Dec 22 09:18:00 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/Connector.cpp: - * ace/Connector.h: Added a lock to the ACE_Connector class to - make the following atomic - - . Registration of AST with the handler_map_ - . Registration of the ACE_Connector with the Reactor - . and the registration of timers with the Reactor's timer - queue. - - If they are not atomic, it throws open race conditions such as - the ones documented in BUG 1405. This is not complete fix for - BUG 1405, since it taxes applications using only one thread with a - lock. This is a short-term fix to prevent the race condition - from occuring. - - Other relevant changes include - - . the AST is now refcounted - . the AST cannot be deleted directly. The lifetime of the AST - can only be manipulated using the refcount on the AST. - -Sat Dec 21 18:27:42 2002 Steve Huston <shuston@riverace.com> - - * ace/ace_dll.vcp: - * ace/ace_os_dll.vcp: - * ace/ace.vcw: eMbedded Visual C++ workspace/projects deleted. These - need to be generated from ace.mpc when needed. The make_release - script now generates them for each source kit. When needed on a - CVS workspace build, use mwc/mpc. The autobuild system has commands - generate_workspace and generate_makefile that accomplish this. - See autobuild configs in the riverace directory for examples. - - * tests/tests_WinCE.vcw: - * tests/ARGV_Test_WinCE.vcp: - * tests/Atomic_Op_Test_WinCE.vcp: - * tests/Auto_IncDec_Test_WinCE.vcp: - * tests/Barrier_Test_WinCE.vcp: - * tests/Basic_Types_Test_WinCE.vcp: - * tests/Bound_Ptr_Test_WinCE.vcp: - * tests/Buffer_Stream_Test_WinCE.vcp: - * tests/Cached_Accept_Conn_Test_WinCE.vcp: - * tests/Cached_Conn_Test_WinCE.vcp: - * tests/Cache_Map_Manager_Test_WinCE.vcp: - * tests/Capabilities_Test_WinCE.vcp: - * tests/CDR_Array_Test_WinCE.vcp: - * tests/CDR_File_Test_WinCE.vcp: - * tests/CDR_Test_WinCE.vcp: - * tests/Collection_Test_WinCE.vcp: - * tests/Conn_Test_WinCE.vcp: - * tests/DLList_Test_WinCE.vcp: - * tests/DLL_Test_DLL_WinCE.vcp: - * tests/DLL_Test_WinCE.vcp: - * tests/Dynamic_Priority_Test_WinCE.vcp: - * tests/Enum_Interfaces_Test_WinCE.vcp: - * tests/Future_Set_Test_WinCE.vcp: - * tests/Future_Test_WinCE.vcp: - * tests/Handle_Set_Test_WinCE.vcp: - * tests/Hash_Map_Bucket_Iterator_Test_WinCE.vcp: - * tests/Hash_Map_Manager_Test_WinCE.vcp: - * tests/High_Res_Timer_Test_WinCE.vcp: - * tests/Lazy_Map_Manager_Test_WinCE.vcp: - * tests/Logging_Strategy_Test_WinCE.vcp: - * tests/Log_Msg_Test_WinCE.vcp: - * tests/Malloc_Test_WinCE.vcp: - * tests/Map_Manager_Test_WinCE.vcp: - * tests/Map_Test_WinCE.vcp: - * tests/Max_Default_Port_Test_WinCE.vcp: - * tests/Mem_Map_Test_WinCE.vcp: - * tests/MEM_Stream_Test_WinCE.vcp: - * tests/Message_Block_Test_WinCE.vcp: - * tests/Message_Queue_Notifications_Test_WinCE.vcp: - * tests/Message_Queue_Test_Ex_WinCE.vcp: - * tests/Message_Queue_Test_WinCE.vcp: - * tests/MM_Shared_Memory_Test_WinCE.vcp: - * tests/MT_Reactor_Timer_Test_WinCE.vcp: - * tests/MT_Reactor_Upcall_Test_WinCE.vcp: - * tests/MT_SOCK_Test_WinCE.vcp: - * tests/Naming_Test_WinCE.vcp: - * tests/New_Fail_Test_WinCE.vcp: - * tests/Notify_Performance_Test_WinCE.vcp: - * tests/Object_Manager_Test_WinCE.vcp: - * tests/OrdMultiSet_Test_WinCE.vcp: - * tests/OS_Test_WinCE.vcp: - * tests/Priority_Buffer_Test_WinCE.vcp: - * tests/Priority_Reactor_Test_WinCE.vcp: - * tests/Priority_Task_Test_WinCE.vcp: - * tests/Process_Manager_Test_WinCE.vcp: - * tests/RB_Tree_Test_WinCE.vcp: - * tests/Reactors_Test_WinCE.vcp: - * tests/Reactor_Exceptions_Test_WinCE.vcp: - * tests/Reactor_Notify_Test_WinCE.vcp: - * tests/Reactor_Performance_Test_WinCE.vcp: - * tests/Reactor_Timer_Test_WinCE.vcp: - * tests/Reader_Writer_Test_WinCE.vcp: - * tests/Recursive_Mutex_Test_WinCE.vcp: - * tests/Refcounted_Auto_Ptr_Test_WinCE.vcp: - * tests/Reverse_Lock_Test_WinCE.vcp: - * tests/Semaphore_Test_WinCE.vcp: - * tests/Service_Config_DLL_WinCE.vcp: - * tests/Service_Config_Test_WinCE.vcp: - * tests/Sigset_Ops_Test_WinCE.vcp: - * tests/Simple_Message_Block_Test_WinCE.vcp: - * tests/SOCK_Connector_Test_WinCE.vcp: - * tests/SOCK_Send_Recv_Test_WinCE.vcp: - * tests/SOCK_Test_WinCE.vcp: - * tests/SPIPE_Test_WinCE.vcp: - * tests/SString_Test_WinCE.vcp: - * tests/Svc_Handler_Test_WinCE.vcp: - * tests/Task_Test_WinCE.vcp: - * tests/Thread_Manager_Test_WinCE.vcp: - * tests/Thread_Mutex_Test_WinCE.vcp: - * tests/Thread_Pool_Reactor_Resume_Test_WinCE.vcp: - * tests/Thread_Pool_Reactor_Test_WinCE.vcp: - * tests/Thread_Pool_Test_WinCE.vcp: - * tests/Timeprobe_Test_WinCE.vcp: - * tests/Timer_Queue_Test_WinCE.vcp: - * tests/Time_Service_Test_WinCE.vcp: - * tests/Time_Value_Test_WinCE.vcp: - * tests/Tokens_Test_WinCE.vcp: - * tests/TSS_Test_WinCE.vcp: - * tests/Upgradable_RW_Test_WinCE.vcp: eMbedded Visual C++ - workspace/projects deleted. These need to be generated from - tests.mpc when needed. The make_release script now generates - them for each source kit. When needed on a CVS workspace build, - use mwc/mpc. The autobuild system has commands - generate_workspace and generate_makefile that accomplish this. - See autobuild configs in the riverace directory for examples. - - * tests/tests.mpc: Added New_Fail_Test. - - * ace/config-WinCE.h: Add #define ACE_LACKS_ACE_IOSTREAM. If it has - ACE_LACKS_IOSTREAM_TOTALLY, it can't possibly do ACE_IOStream<>. - - * ace/ATM_Acceptor.{h cpp}: - * ace/ATM_Addr.{h cpp}: - * ace/ATM_Params.h: - * ace/ATM_QoS.h: - * ace/Dev_Poll_Reactor.cpp: - * ace/FlReactor.cpp: - * ace/POSIX_Proactor.cpp: - * ace/QtReactor.h: - * ace/SUN_Proactor.h: - * ace/TkReactor.h: Add /**/ between #include and filename to silence - MS warnings about missing files during dependency updates. - - * ace/OS.i: Fixed typo for ACE_HAS_WINCE. - - * ace/OS.{h cpp}: Enabled ACE_OS::mktime() for ACE_HAS_WINCE. To do - this, define struct tm similarly to the way it is in the standard - C library. When called, converts the struct tm to a SYSTEMTIME, - then to a FILETIME, then inserts that into an ACE_Time_Value and - pulls out the sec() value. - In OS.h, add the ACE_HAS_WINCE section for the ace_timezone() - function. - - * ace/Filecache.cpp (ACE_Filecache_Object::update): For ACE_HAS_WINCE, - return status to indicate every file has changed since cached. - There's probably a way to do this correctly on WinCE, but there's - no time to figure it out right now. - - * ace/Logging_Strategy.cpp: In calls to ACE_OS::fopen(), added - ACE_LIB_TEXT around naked char strings to build clean with Unicode. - - * ace/NT_Service.{h cpp}: Add ACE_HAS_WINCE to the platforms that - don't attempt to build this class. - - * ace/SPIPE_Connector.cpp (connect): Pretty much disabled this for CE. - - * bin/MakeProjectCreator/templates/em3vcp.mpd: When generating - CPU-based /D options, just do /D, not /D<name>=1. If =1 is need - for something, add it to the .mpt file. - - * bin/MakeProjectCreator/templates/em3vcpdll.mpt: - * bin/MakeProjectCreator/templates/em3vcpdllexe.mpt: - * bin/MakeProjectCreator/templates/em3vcplibexe.mpt: Remove /GX, /Gm, - /GR, /Gy, and /Md[d] options - embedded VC++ doesn't do these or the - previous near-working .vcp files didn't use them and no value is - seen in enabling something not previously used. - Insured that the options are consistent, for example, all Release - builds use /Ob2 (high level of inlining). - * bin/MakeProjectCreator/templates/em3vcplib.mpt: In addition to the - above removals, remove /G5 (optimize for Pentium) from Debug build. - -Sat Dec 21 12:38:41 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.cpp (snprintf): Win32 doesn't 0-terminate the string if - the maxlen is exceed, so terminate the string in that case. - -Sat Dec 21 14:36:43 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * include/makeinclude/platform_cygwin32.GNU: - Enabled threads and removed the setting that only static libs are - build. Also set SONAME and SOVERSION to nothing because the - Cygwin linker can't handle a dll name where the version is added - after .dll. - -Fri Dec 20 20:02:41 2002 Steve Huston <shuston@riverace.com> - - * bin/make_release: When updating version files, also update - bin/MakeProjectCreator/config/aceversion.mpb and taoversion.mpb. - Added a new subroutine, generate_makefiles() that's called just - before create_kit. It generates the workspace and project files - for ACE_wrappers/ace and ACE_wrappers/tests. If any of these are - desired for TAO, someone knowledgeable on TAO should add to the - place in generate_makefiles() for the TAO stuff. - -Fri Dec 20 16:12:33 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/ACE.h: - The math.h delived with the Cygwin32 compiler already defines - log2 as macro so added a define acelog2 as log2 to handle this - already defined macro. - -Thu Dec 19 18:47:29 2002 Steve Huston <shuston@riverace.com> - - * tests/tests.mpc: Added Obstack_Test and Recursive_Condition_Test. - -Thu Dec 19 16:47:17 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/SOCK_Dgram.h: - ace/SOCK_CODgram.h: Added metatype information for - ACE_SOCK_Dgram and ACE_SOCK_CODgram. Thanks to Sankaranarayanan - K V <sankar@migv.mot.com> for reporting this. - -Thu Dec 19 10:02:56 2002 Rich Seibel <seibel_r@ociweb.com> - - * tests/MM_Shared_Memory_Test.cpp - - removed extraneous line that made the test appear to fail - -Thu Dec 19 06:47:07 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/config/aceversion.mpb: - * bin/MakeProjectCreator/config/taoversion.mpb: - - Updated the version to reflect the current release version. - - * bin/MakeProjectCreator/modules/Driver.pm: - - Moved the Cwd::getcwd() outside the loop to avoid calling it - multiple times. - - * bin/MakeProjectCreator/templates/vc7.mpd: - - Modified the ProgramDatabaseFile to use the intermediate_dir as - the output location. Thanks to Steve Huston - <shuston@riverace.com> for reporting it. - -Wed Dec 18 20:52:41 2002 Steve Huston <shuston@riverace.com> - - * ace/SOCK_Dgram_Mcast.cpp (open_i): Call shared_open using the - protocol family from the specified address, not PF_INET. Allows - working with IPv6. - - * tests/Multicast_Test.cpp: If join fails, log the error reason. - -Wed Dec 18 19:57:51 2002 Steve Huston <shuston@riverace.com> - - * ace/config-sunos5.7.h: Added ACE_HAS_SNPRINTF. I have verified - that it's available from Solaris 7 on. If someone can determine - it's available earlier, please move the definition to one of the - earlier-version config-sunos5.*.h files. - -Wed Dec 18 18:59:41 2002 Steve Huston <shuston@riverace.com> - - * ace/ace.mpc: - * tests/tests.mpc: Back out the below changes... jumped the gun. - -Wed Dec 18 18:55:31 2002 Steve Huston <shuston@riverace.com> - - * ace/ace.mpc: Added new Hash_Map_Manager_RT[_T].cpp files. - - * tests/tests.mpc: Added new Hash_Map_Manager_RT_Test.cpp project. - -Wed Dec 18 15:03:03 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/Get_Opt.cpp (long_option): For MSVC 7.x compiler, we need to - make sure that the argument passed to "isalnum" is legal, i.e., - between 0 and 0xff, otherwise, it will crash the process. Added - a special case for MSVC 7.x and above to prevent this. - -Wed Dec 18 13:36:49 2002 Steve Huston <shuston@riverace.com> - - * ace/ace.mwc: Workspace definition file for ACE and it's immediate - subordinates, RMCast. - - * tests/tests.mwc: Workspace definition for tests and tests/RMCast. - - * bin/MakeProjectCreator/modules/Driver.pm (run): When processing - specified input files, chdir to the directory containing the file - so the files are read and written properly when referred to using - relative paths. - -Wed Dec 18 11:38:39 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/config-win32-msvc-7.h: Added pragma prefix to disable - warning of using Microsoft extension. - - * ace/Timeprobe.h: Removed MSVC specific disabling and re-enabling - of warning of using Microsoft entension. We already do this - through all the MSVC config files and there's no point of doing - this again here. - -Mon Dec 16 16:34:23 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/OS.i: Applied the following patch from Eugene Alterman - <EugeneA@bremer-inc.com>. - - It appears that Win32 mmap() emulation implements MAP_FIXED - semantic under the wrong assumption about MapViewOfFileEx() - behavior. In fact if a non-null base address is supplied to - MapViewOfFileEx() it behaves like mmap() with MAP_FIXED set - it - maps at that address or fails if it unable to do it. If a null - base address is supplied it behaves like mmap() with MAP_FIXED - not set. - - This means that if MAP_FIXED is not set the emulation should set - the base address argument to 0, and it should fail if MAP_FIXED - is set and <addr> is 0. - - Besides, it makes sense for WinCE emulation to fail whenever - MAP_FIXED is set since WinCE does not support MAP_FIXED - semantic. - -Wed Dec 18 08:29:04 2002 Rich Seibel <seibel_r@ociweb.com> - - * ace/Synch.cpp: Added code to generate a unique name if a - platform lacks unnamed semaphores (MACOSX 10.2). - - * ace/Sock_Connect.cpp: Added a conditional include for - ifaddrs.h (MACOSX 10.2). - - * tests/Priority_Task_Test.cpp: Changed test to notify - if platform does not implement getprio instead of fail. - (MACOSX 10.2) - - * tests/MM_Shared_Memory_Test.cpp: Changed compile time test to - pick up the right implementation on platform that have - POSIX semaphores but not SYSV semaphores (MACOSX 10.2). - - * ace/Lib_Find.cpp - ace/config-macosx.h: - - Gave the builder a choice of where to get the dlcompat package. - - * ace/OS.i - ace/config-macosx.h: - - Added a switch to eliminate an error message since the dlcompat - provided APPLE dynamic linkage does not support unloading thus - complains on the use of dlclose. - -Wed Dec 18 08:29:04 2002 Rich Seibel <seibel_r@ociweb.com> - - * ace/Timer_Wheel_T.cpp: - - Changed types from int to long to match the return - for timer ids. - -Wed Dec 18 05:42:52 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * ace/High_Res_Timer.cpp (global_scale_factor): The INTEL compiler - can also set the global scale factor. Thanks to Chris Kohlhoff - for reporting this. - -Tue Dec 17 11:22:31 2002 Phil Mesnier <mesnier_p@ociweb.com> - - * ace/Codeset_IBM1047.h: - * ace/Codeset_IBM1047.cpp: - - Added implementations of ncs and tcs methods to return the proper - values as defined by the OSF's code and character set registry. - - * ace/CDR_Stream.cpp: - * ace/CDR_Stream.h: - * ace/CDR_Stream.i: - - Stripped out the reference counting on translator objects to avoid - build problems. - -Tue Dec 17 11:30:31 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ACE-INSTALL.html: Fixed fuzz errors. - -Tue Dec 17 08:20:22 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/NMakeWorkspaceCreator.pm: - - Fixed a problem with generated Makefiles where CFG is defaulted. - - * bin/MakeProjectCreator/templates/nmake.mpd: - - Modified the dependencies file name. - -Tue Dec 17 07:03:40 2002 Chad Elliott <elliott_c@ociweb.com> - - * ace/Array_Base.h: - * ace/Atomic_Op_T.cpp: - * ace/Message_Queue.cpp: - * ace/OS.h: - - Added #include's for various files for the Green Hills compiler - for VxWorks. - - * ace/OS.cpp: - - Added a call to ::inet_aton() with a non-const char* for VxWorks. - - * ace/Thread_Adapter.cpp: - - Changed the pointer type from ACE_THR_FUNC_INTERNAL to - ACE_THR_FUNC. - - * ace/config-vxworks5.x.h: - - Added a #define for ACE_LACKS_CLEARERR for the Green Hills - compiler. - -Mon Dec 16 12:28:05 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ACE-INSTALL.html: Fixed some broken links. Thanks to - Jonathan Wackley <jwackley@legato.com> for reporting this. - -Sat Dec 14 13:06:09 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * docs/ACE-categories.html: Removed all mention of the - ACE_Multiplexor class. Thanks to Marc Tardif <marc@sitepak.com> - for reporting this. - -Tue Dec 17 07:37:40 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * tests/README: Fixed typos. - -Mon Dec 16 23:31:25 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * THANKS: Added Guan Joe to the hall of fame. - -Mon Dec 16 18:11:03 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/CDR_Stream.h: Moved the destructor of - ACE_Char_Codeset_Translator to be public. - -Mon Dec 16 13:14:34 2002 Phil Mesnier <mesnier_p@ociweb.com> - - * ace/CDR_Stream.cpp: - * ace/CDR_Stream.h: - * ace/CDR_Stream.i: - Preparing for codeset negotiation. The ACE input/output CDR - streams' support for wide character IO is enhanced by adding a - flag to allow or disallow wchar i/o regardless of whether a - translator is present. There are times when wchar is not - allowed, such as GIOP 1.0, or when a native codeset for wchar - was not specified. In CORBA, a native wchar codeset must be - specified, it may not be defaulted. Within the ACE CDR streams, - wchar i/o is allowed by default, to ensure backwards - compatibility. - - I fixed a few of the reading and writing methods to ensure that - the good_bit_ flag is cleared if an error occurs. This was not - happening in all cases, leading to errors elsewhere. - - I have extended the definition of the - ACE_[W]Char_Codeset_Translator classes to make them reference - counted objects. There are some codesets that support shifting - and may be stateful. By using reference counting, the CDR - streams don't have to worry about ownership of stateful - translators. - - Finally, the translators now have pure virtual methods that - return the native and translated codeset ID values. These values - refer to entries in the DCE Code and Character Set Registry, - which is the reference for CORBA codeset ids. Anyone - implementing a char or wchar codeset will have to supply a - numeric value for the native codeset (ncs) and the translated - codeset (tcs). NCS refers to the codeset of the text on the - application side of the stream, and TCS refers to the codset of - the text on the "wire" side of the stream. - - These changes are not everything that is required to support - codeset negotiation, but it is all the changes required to the - CDR streams. - -Sat Dec 14 08:56:23 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * ace/Timer_Queue_T.cpp (timeout): Commented out the change below - until we figure out how to break the dependencies on the - Reactor. Thanks to Bala for reporting this. - -Fri Dec 13 16:05:05 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/Timer_Queue_T.cpp (timeout): Optimize Steve's patch below - for the case where the upcall is coming from a reactor. Thanks - to Irfan for pointing this out. - -Fri Dec 13 21:12:29 2002 Steve Huston <shuston@riverace.com> - - * ace/test_config.h: Add __hpux to the conditions needed to instantiate - ACE_Singleton<ACE_Test_Output, ACE_Null_Mutex>::singleton_. Also see - Sat Nov 23 18:12:37 2002 Steve Huston <shuston@riverace.com> - -Fri Dec 13 18:25:42 2002 Steve Huston <shuston@riverace.com> - - * ace/config-aix-4.x.h: Set ACE_HAS_THREADS to 1 or 0 based on the - compiler's predefined macros. This lets users build easier without - having to worry about what ACE expects - they just have to invoke - the compiler correctly. Add #include "ace/config-posix.h" to pick - the detectable POSIX-ish settings. - - * ace/config-posix.h: If ACE_HAS_THREADS is defined as 0 on entry, - undef it and don't pick up the POSIX threads settings. Allows - single-threaded builds using this very handy config. - -Fri Dec 13 18:10:31 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.cpp (wchar snprintf, sprintf): Change the selection for - ability to use vswprintf from ACE_HAS_XPG4_MULTIBYTE_CHAR to - _XOPEN_SOURCE >= 500. vswprintf came in at XPG5 (and C99). - - * ace/config-hpux-11.00.h: Enable ACE_HAS_SNPRINTF if - _INCLUDE_HPUX_SOURCE is set after including stdsyms.h. - - * ace/config-posix.h: Set ACE_HAS_SNPRINTF if _XOPEN_SOURCE >= 500. - -Thu Dec 12 20:11:39 2002 Steve Huston <shuston@riverace.com> - - * ace/Timer_Heap_T.cpp (cancel): Release the queue lock before - upcalling to handle_close. Fixed Bugzilla #1387. - -Thu Dec 12 12:50:34 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/OS.cpp: Changed a cast of unique_name to &unique_name to - fix a bug on pSoS. Thanks to Dieter Knueppel - <dknueppel@datus.com> for reporting this. - -Thu Dec 12 12:55:41 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.h: Add missing snprintf() method declaration for ACE_HAS_WCHAR - -Thu Dec 12 08:49:31 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/Log_Msg.cpp: Fixed a warning in LynxOS builds. - -Thu Dec 12 06:18:02 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/Parser.pm: - - Fixed a bug with setting and retrieving the current working - directory. Thanks to Steve Huston <shuston@riverace.com> for - reporting this. - -Thu Dec 12 00:01:37 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * netsvcs/clients/Naming/Client/Makefile: Fixed a small problem in - static builds. Hopefully they should be fine in the next round - of builds. - -Wed Dec 11 23:11:09 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/clean_sems.sh: A shell script to remove - semaphores. Integrating the script that Ossama used to the - daily build system. - -Wed Dec 11 21:05:49 2002 Steve Huston <shuston@riverace.com> - - * ace/Message_Block_T.cpp: - * examples/ASX/Message_Queue/priority_buffer.cpp: - * examples/Misc/test_read_buffer.cpp: - * examples/Threads/TSS_Obj.h: Added missing - #include "ace/Malloc_Base.h" to build clean on AIX, Visual Age C++ 6. - - * examples/C++NPv1/Logging_Client.cpp: - * examples/C++NPv2/Reactor_Logging_Server.cpp: - * examples/C++NPv2/Service_Reporter.cpp: - * examples/C++NPv2/SLD.cpp: - * examples/C++NPv2/SLDex.cpp: Add missing #include - "ace/Log_Msg.h" to pick up ACE_ERROR_RETURN for AIX Visual Age C++ 6. - - * examples/C++NPv2/Makefile.SLD: Remove the tempinc-bashing - redefinition of COMPILE.cc... that only works if there's a single - source file for each program. - -Wed Dec 11 19:33:19 2002 Steve Huston <shuston@riverace.com> - - * ace/config-hpux-10.x-hpc++.h: - * ace/config-hpux-10.x.h: Picked up these changes from the 11.x config: - - Sat Nov 23 17:54:18 2002 Steve Huston <shuston@riverace.com> - - * ace/config-hpux-11.00.h: Control the ACE_HAS_STANDARD_CPP_LIBRARY - and ACE_USES_STD_NAMESPACE_FOR_STDCPP_LIB settings by compiler- - defined macro _HP_NAMESPACE_STD, not passing them through from - platform_hpux_aCC.GNU. Similarly with ACE_HAS_EXCEPTIONS (define - it unless __HPACC_NOEH is set). This breaks a coupling from the - platform_hpux_aCC.GNU file, and will make autoconf easier as well - as making it easier for users to write makefiles without following - the ties through and trying to duplicate them. - Also, include config-posix.h to let it pick up system-supplied - settings. - - * ace/Time_Value.h: For HP-UX 10, need to typedef timespec_t because - the OS doesn't define it. - -Wed Dec 11 17:05:28 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * examples/Reactor/WFMO_Reactor/Talker.cpp: Fixed a warning in - Borland builds. - -Wed Dec 11 17:00:21 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/WIN32_Asynch_IO.cpp: Fixed a warning in Borland builds. - -Tue Dec 10 19:42:29 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.{h cpp}: Added new method, ACE_OS::snprintf() that wraps - native snprintf() if available (ACE_HAS_SNPRINTF for char, - ACE_HAS_XPG4_MULTIBYTE or ACE_WIN32 for wide char); else returns - -1, ENOTSUP. - - * ace/config-linux-common.h: Add detection for ACE_HAS_SNPRINTF. - - * ace/config-win32-common.h: Removed #define ACE_HAS_VSWPRINTF. - vswprintf() on Windows is not the standard way, and this setting - was only used for Win32. See comments in OS.cpp, sprintf(wchar_t...) - for more info. Added ACE_HAS_SNPRINTF in its place. - - * ace/Log_Msg.cpp (log): Don't overrun the msg_ buffer while - building the logged message. Silently stops logging characters - when the limit is reached. - Also, on %a aborts, don't just do a "while (1);" on WinCE. - Forcing the user to cycle power is not a good idea. - - * tests/Log_Msg_Test.cpp: Add a test case for logging a string - that can't possibly fit in the logging buffers. - -Tue Dec 10 18:10:41 2002 Steve Huston <shuston@riverace.com> - - * tests/Proactor_Test.cpp: Changed the default proactor type to - DEFAULT. If a specific proactor implementation type is desired, that - must be specified now. This matches ACE's behavior that users will - get unless they specifically change it. Also, changed loglevel to - be 0 (only highlights and errors), 1 (more progress information), - and 2 (all operation parameters and result information). The default - is 0, allowing for increased load on this test without creating - gigantic log files. All platforms now use the same test parameters - (clients, byte counts, threads, etc.) except that if the SUN - proactor is selected, the test will run with one thread regardless - of user selection. The ACE_SUN_Proactor's completion detector - (aiowait) is not multithread safe. - -Tue Dec 10 17:19:06 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * tests/OS_Test.cpp (ACE_TMAIN): Integrated the latest tests from - Alain for string_strsncpy_test (). The test is more robust and - tests things better. It also has comments on what it is testing - :-). - -Tue Dec 10 16:53:29 2002 Steve Huston <shuston@riverace.com> - - * tests/OS_Test.cpp: Add missing '=' signs to fix compile error on - wide-char builds. - -Tue Dec 10 16:05:37 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * Makefile: Added a target 'reverseclean' - * bin/reverse_clean: Added a small perl script that does a - realclean in the reverse order on the directories passed to - it. - - The motivation for this is that some of our builds dont clean up - properly due to the dependencies. The reason for this is that - ace/ACE_COMPONENTS.list gets wiped out first, preventing things - that examine the components list (like netsvcs) from getting - cleaned. Hence we need to do a 'realclean' in reverse order of - the builds. - -Tue Dec 10 13:23:37 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * bin/PerlACE/Process.pm: - * bin/PerlACE/Process_Win32.pm: - * bin/PerlACE/Process_Unix.pm: Changed WAIT_DELAY_FACTORY from a - data member in Process_Win32 to a global data in - PerlACE::Process. You can now define an environment variable - "ACE_RUNTEST_DELAY" on both Win32 and UNIX to multiply the wait - time with the number. The same trick that I added last time - into Process_Win32 to invoke tests with purify should also work - with valgrind. - - * bin/PerlACE/Run_Test.pm (waitforfile): Changed to multiply the - wait time with the PerlACE::Process::WAIT_DELAY_FACTOR . - -Tue Dec 10 10:20:35 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/Synch.cpp: Commented out the dump () call on ACE_cond_t, - which caused build errors from the checkin "Mon Dec 9 18:54:09 - 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu>". There is - no dump method on ACE_cond_t. - - Fixed the return type of the implementation to match the - declaration. - -Mon Dec 9 18:54:09 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/Synch.{h,cpp}: Added a dump() method to the template - specialization of ACE_Condition<ACE_Recursive_Thread_Mutex>. - Thanks to Mathias Waack <mathias.waack@schlund.de> for reporting - this. - - * ace/Get_Opt.{h,i} Added a new argc() accessor. Thanks to - Volker Boerchers <vboerchers@tecon.de> for contributing this. - -Mon Dec 9 17:58:31 2002 Steve Huston <shuston@riverace.com> - - * tests/Multicast_Test.cpp: Added missing template instantiation for - Sun Workshop 5 compiler. - -Mon Dec 9 17:27:49 2002 Steve Huston <shuston@riverace.com> - - * tests/Recursive_Condition_Test.cpp: In test_4, wait for the - waiter thread to wake up and exit itself before checking the - mutex recursion count. Else, the waiter may have the mutex while - the main thread is checking the count and detect a false fail. - -Mon Dec 9 17:02:41 2002 Steve Huston <shuston@riverace.com> - - * include/makeinclude/platform_aix_ibm.GNU: Change from using - -qtemplateregistry to use -qtempinc for Visual Age C++ 6 template - handling. -qtemplateregistry is a nice idea, but generates too - many errors from missing templates. Using -qtempinc still generates - gobs of multiple-instantiation warnings, but the build is correct - and the tests run. - - * bin/aix_shr: Add the error number patterns for VA 6's multiple - template instantiation warnings to the filter-out pattern. - -Mon Dec 9 15:42:39 UTC 2002 Don Hinton <dhinton@ieee.org> - - * examples/C++NPv1/Reactive_Logging_Server_Ex.cpp: Removed - duplicate template specialization for ACE_Hash<ACE_HANDLE> on - Windows that's no longer needed. - -Mon Dec 9 07:41:39 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * tests/OS_Test.cpp: - * ace/OS_String.cpp (strsncpy): Fixed a cut and paste error. - -Mon Dec 9 00:09:55 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/INET_Addr.cpp (get_host_addr): Changed the usage of strncpy - () to strsncpy (). Thanks to Alain Decamps for donating this patch. - -Sun Dec 8 23:56:09 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/OS_String.cpp: Applied patches from Alain Decamps to make - strsncpy () more robust. If dst == src, the return string will - have utmost 'maxlen'. - - * tests/OS_Test.cpp: Added a test case for strsncpy (). Thanks to - Alain Decamps for donating the patch. - -Sun Dec 08 22:44:07 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * ace/Functor.h (void *): - * ace/Functor.i: Added a template specialization for ACE_Hash <void *>. - -Sun Dec 08 09:45:34 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/Map_Manager.h: - * ace/SString.h: - * ace/Unbounded_Queue.h: - Improved doxygen documentation - -Sun Dec 08 09:37:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * etc/*.doxygen: - Set JAVADOC_AUTOBRIEF to NO in all doxygen config files. When - only javadoc style documentation is used, only detailed documentation - is generated. - -Sun Dec 08 01:17:26 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/Event_Handler.cpp: Fixed a typo. - -Thu Dec 5 20:30:56 2002 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * ace/Event_Handler.cpp: Updated the implementation of - ACE_Event_Handler::read_adapter() so that the code is consistent - with the stated semantics. Thanks to Ron Muck <rlm@sdiusa.com> - for the fix. - - * ace/OS_Dirent.inl (readdir_r): Enhanced the - ACE_OS_Directn::readdir_r() wrapper facade so that it returns 1 - (i.e., "done") when an error occurs. Thanks to Abhay Kulkarni - <Abhay.Kulkarni@veritas.com> for reporting this. - -Sat Dec 7 19:54:12 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.h: For WinCE, define S_IFDIR and S_IFREG file mode masks. - Also, added define for ENXIO if not supplied (like on WinCE). - -Sat Dec 07 14:29:47 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/ACE.cpp (format_hexdump): This change fixed the problem that - a character is missing when the last line has less than 16 bytes - or less to display. Thanks to Terry Lao <terry@ec-tone.com> for - submitting the patch, and Johnny Willemsen - <jwillemsen@remedy.nl> for reviewing the change. - -Fri Dec 6 23:58:47 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/CDR_Stream.cpp: - * ace/CDR_Stream.i: Set good_bit_ to zero whenever the sanity - check for remaining length in the stream fails. Thanks to Jerry - D. De Master <jdemaste@rite-solutions.com> for pointing out the - problem. - - * ace/CDR_Array_Test.cpp: Extended the test to include test for - checking the value of good_bit () by forcing an erroneous read. - -Fri Dec 6 16:26:49 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ace/FILE_Addr.cpp (set): Windows appends ASCII 'a' - 'z' and - 'A' - 'Z' to generate unique filenames and gives up if the - character in the template is a non-alphabetic character. In such - cases, it returns a NULL. Make sure that we supply a template - with a alphabetic character before XXXXXX as well as check for a - NULL return value. - -Thu Dec 5 22:19:41 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.{h i}: Added ACE_OS::sigwaitinfo() wrapper. - - * ace/POSIX_Proactor.{h cpp} (ACE_POSIX_SIG_Proactor): Clean up some - spelling and formatting in documentation comments. Clean up the - signal calls, and substitue ACE_OS calls for naked API calls. - Unless we have specific evidence to the contrary, trust what the - OS says with signal information. Otherwise, if we run through - all the aiocbs on every signal, we could get into a situation - where deferred I/O operations are never restarted. - - * ace/POSIX_CB_Proactor.h: Changed the comments to say this is a - callback-based proactor, not an SGI IRIX one. The callback - paradigm is part of newer POSIXes and available on more than - SGI... HP-UX, for instance. - -Thu Dec 5 20:33:18 2002 Steve Huston <shuston@riverace.com> - - * ace/config-aix-4.x.h: Removed ACE_NEEDS_DEV_IO_CONVERSION. This - seems like a ludicrous workaround to some old, buggy compiler. - -Thu Dec 5 14:35:50 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * netsvcs/clients/Naming/Client/Makefile: Do not build in static - builds. - -Thu Dec 5 09:45:25 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * examples/Synch/proc_sema.cpp: Fixed some typos. Thanks to - Andy King <andyking@vernon.com> for reporting this. - -Thu Dec 5 07:39:45 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: Removed Bug_1020_Regression from the - nightly regression test on Win32. - -Thu Dec 5 09:58:45 UTC 2002 Don Hinton <dhinton@ieee.org> - - * ace/config-openbsd-pthread.h: Added - ACE_LACKS_PERFECT_MULTICAST_FILTERING 1. This should - fix the errors in the test/Multicast_Test in the - nightly autobuilds. - -Wed Dec 4 21:53:28 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.h: Include <sys/pstat.h> for __hpux platform. - * ace/OS.cpp (num_processors_online): Try the syscall for this - on HP-UX if __hpux is defined, not hpux (only set for g++). - -Wed Dec 4 21:18:31 2002 Steve Huston <shuston@riverace.com> - - * ace/Proactor_Timer_Test.cpp: Fix incorrect use of pointer-integer - casting. Log pointers with %@, not %x. - -Wed Dec 4 21:11:41 2002 Steve Huston <shuston@riverace.com> - - * tests/Proactor_Test.cpp: Set all the default options and values - before parsing command line options to make it easy to predictably - vary things by hand. Also, rather than run for a determined number - of seconds, run a specific number of bytes around the loop. This - limits the size of the log, especially nice for fast processors. - Thanks to Johnny Willemsen <jwillemsen@remedy.nl> for this idea. - -Wed Dec 4 18:37:10 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: Added a Server_Connection_Purging to the - nightly regression tests. - -Wed Dec 4 17:47:49 2002 Steve Huston <shuston@riverace.com> - - * tests/Proactor_Test.cpp: On Sender read complete, start another - read if either in full duplex mode OR writing has been stopped. - We want to catch the Receiver closing the socket (read 0 bytes). - -Wed Dec 4 11:41:15 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: Added a Bug_1020_Regression to the - nightly regression test on Win32. - -Tue Dec 3 20:47:39 2002 Steve Huston <shuston@riverace.com> - - * ace/SSL/SSL_Context.cpp: Add explicit instantiation of - ACE_Singleton<ACE_SSL_Context, ACE_SYNCH_MUTEX>::singleton_. - * netsvcs/lib/Name_Handler.cpp: Add explicit instantiation of - ACE_Singleton<Naming_Context, ACE_SYNCH_NULL_MUTEX>::singleton_. - * netsvcs/lib/Server_Logging_Handler.cpp: Add explicit instantiation of - ACE_Server_Logging_Handler_T<LOGGING_PEER_STREAM, u_long, - ACE_NULL_SYNCH, Null_Synch_Static_Receiver>::request_count_; - * tests/Framework_Component_DLL.cpp: Add explicit instantiation of - ACE_DLL_Singleton_T<FWCT_DLL_Singleton_Adapter_T <Simple_Service>, - ACE_SYNCH_MUTEX>::singleton_. - * tests/MEM_Stream_Test.cpp: Add explicit instantiation of - ACE_Singleton<ACE_Atomic_Op<ACE_SYNCH_MUTEX, u_short>, - ACE_SYNCH_RECURSIVE_MUTEX>::singleton_. - * tests/Process_Strategy_Test.cpp: Add explicit instantiation of - ACE_Singleton<Options, ACE_Null_Mutex>::singleton_. - * tests/Timeprobe_Test.cpp: Add explicit instantiation of - ACE_Singleton<int, ACE_SYNCH_RECURSIVE_MUTEX>::singleton_. - - These are similar to those done in: - Sat Nov 23 18:12:37 2002 Steve Huston <shuston@riverace.com> - - * netsvcs/clients/Naming/Client/Makefile: - * netsvcs/clients/Naming/Dump_Restore/Makefile: Add ACELIB to LIBS - to include -lACE for building the shared libs. rules.lib.GNU - doesn't add VLDLIBS when linking a shared lib, but rules.bin.GNU - adds it for binaries. This should probably be changed, but it's a - quagmire to wade into at this point, and autoconf is coming. - This fix is needed now to get this stuff building with g++ on AIX. - -Mon Dec 2 20:09:29 2002 Steve Huston <shuston@riverace.com> - - * examples/C++NPv1/Reactive_Logging_Server_Ex.cpp: - * ACEXML/common/Mem_Map_Stream.cpp: - * ACEXML/common/URL_Addr.cpp: - * ACEXML/common/XML_Codecs.cpp: Added missing explicit template - instantiations to build clean on Sun Workshop 5. - -Mon Dec 2 19:41:27 2002 Steve Huston <shuston@riverace.com> - - * tests/Proactor_Test.cpp: Rather than close connections outright - when beginning to shut down test, just shut down the write side. - This is an effort to make the receiving closes cleaner. - -Mon Dec 2 17:58:06 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/ACE.cpp: Fixed a compile error with g++. - -Sat Nov 30 09:18:51 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/ACE.{h,cpp}: Added a new crc32() method that works on - an iovec. Thanks to Tufan Oruk <toruk@usa.net> for contributing - this. - - * THANKS: Just added our 1,600 contributor to ACE+TAO! - -Mon Dec 02 18:59:33 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/DLL.h: - * ace/Dirent.h: - * ace/Future_Set.h: - * ace/UPIPE_Connector.h: - * ace/UPIPE_Acceptor.h: - * ace/UNIX_Addr.h: - * ace/Configuration_Import_Export.h: - Improved doxygen documentation. - -Mon Dec 02 18:47:32 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * THANKS: Added Tim Smith <timsmith_s@hotmail.com> - -Sun Dec 01 10:16:45 2002 Balachandran <bala@cs.wustl.edu> - - * ACE version 5.2.7 released. - -Sat Nov 30 22:46:00 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * THANKS: Added Christophe Vedel to the hall of fame. - -Fri Nov 29 12:23:14 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/Thread_Manager.h (ACE_Thread_Manager): Changed all the - accessors to the Thread_Descriptor to be protected. - - * ace/Thread_Manager.i: Removed assertion test for null - Thread_Descriptor pointer. This can happen when calling - thread_desc_self() from a thread not spawned by the - Thread_Manager, e.g., the main thread. - - Added checks in all ACE_Thread_Manager::at_exit for null - thread_desc_self to avoid crash. - - Thanks to Mathias Waack <mathias.waack@schlund.de> for reporting - the bug. - -Thu Nov 28 06:36:04 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * apps/soreduce/Makefile: Changed INSBIN from $ACE_ROOT/bin/soreduce/ - to $ACE_ROOT/bin/. Thanks to Jaroslaw Nozderko - <jaroslaw.nozderko@polkomtel.com.pl> for reporting this. - -Thu Nov 28 07:54:28 2002 Craig Rodrigues <crodrigu@bbn.com> - - * docs/tutorials/022/Acceptor_Service.h: Add include of svc_export.h - to fix compilation errors. Thanks to Viktor Ransmayr - <viktor.ransmayr@t-online.de> for reporting the problem. Not - sure whether this example would still compile fine. - -Thu Nov 28 09:04:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/FILE.h: - * ace/Default_Constants.h: - Improved usage of doxygen commands. - -Thu Nov 28 07:23:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/Typed_SV_Message_Queue.h: - * ace/Strategies_T.h: - * ace/Framework_Component.h: - Improved doxygen documentation. - - * ace/TLI_Connector.i: - Made method definition and implementation the same. - -Wed Nov 27 22:38:15 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/Connector.h: - * ace/Connector.cpp: Fix for BUG 1361. Please see - http://deuce.doc.wustl.edu/bugzilla/show_bug.cgi?id=1361 for - more details on the problem. - - The fix is based on the assumption that ACE_Connector need not - be resumed by the TP_Reactor when used for asynch connects. - The supporting reason for this assumption is that during - connection completion or connection closure, the handler - ie. the ACE_Connector in this case is removed from the - Reactor. Hence this fix, which essentially implements the - virtual function resume_handler () and prevents itself from - being resumed. Thaks to Irfan for reviewing the fix. - -Wed Nov 27 20:35:49 2002 Steve Huston <shuston@riverace.com> - - * include/makeinclude/platform_aix_ibm.GNU: Add Visual Age C++ 6 to - the list of compiler versions it knows about. For VA 6, use the - new -qtemplateregistry build option to try to be more efficient - about template compilation. It's still lamer than most though, and - Visual Age C++ 6 needs the aixshr filter to weed out the Duplicate - Symbol warnings. - - * include/makeinclude/rules.local.GNU: Add templateregistry to the - files/directories removed on clean. It's used by Visual Age C++ 6. - -Wed Nov 27 19:42:51 2002 Steve Huston <shuston@riverace.com> - - * tests/MT_Reactor_Upcall_Test.cpp: Send and receive each message - field specifically; don't rely on the type and size fields being - adjacent in memory. On 64-bit Solaris w/ Forte 7 and on 64-bit - Windows XP, they aren't. - -Wed Nov 27 15:47:41 2002 Steve Huston <shuston@riverace.com> - - * ace/config-aix5.1.h: Add #define ACE_HAS_AIO_CALLS, but commented - out. AIX did things differently enough here to be a real problem. - It should be possible to get AIO working on AIX 5, but it will - take some real work. - - * ace/POSIX_Proactor.cpp: Add #include "ace/ACE.h" - - * ace/Sock_Connect.cpp: Visual Age C++ 6 still needs to hack to - get the templates built correctly :-( Changed the compile-time check - to an alleged Visual Age C++ 7 to see if they get it right then. - - * ACEXML/common/Env.i: Add missing #include "ace/Log_Msg.h" to pick - up ACE_ASSERT. - - * ACEXML/common/FileCharStream.cpp: - * ACEXML/common/StrCharStream.cpp: Add missing #include "ace/Log_Msg.h" - - * ACEXML/common/NamespaceSupport.h: - * ACEXML/parser/parser/Entity_Manager.h: Add missing #include - "ace/Synch.h" - -Wed Nov 27 07:23:30 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: Two new tests cannot run in single - threaded builds. - -Wed Nov 27 09:30:33 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/DLL.h: - * ace/Singleton.h: - Improved doxygen documentation. - -Tue Nov 26 20:04:38 2002 Steve Huston <shuston@riverace.com> - - * tests/Proactor_Test.cpp: Added checks for condition where a read or - write fails to initiate because the socket was closed. For Windows, - this is a ERROR_NETNAME_DELETED error on the ReadFile/WriteFile call - under ACE_WIN32_Asynch_Read/Write_Stream. The test now checks for - this and doesn't report an error, but this is a departure from - what one would expect from a socket. It might be better if the - underlying call was changed to WSARecv/WSASend for Winsock2-enabled - systems. - -Tue Nov 26 16:08:22 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: Added two new tests to our daily builds. - -Tue Nov 26 15:10:57 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ACE-INSTALL.html: Updated the file to have the right GNU make - version. Our make's dont work well with 3.78 and lower. - -Mon Nov 25 23:49:38 2002 Craig Rodrigues <crodrigu@bbn.com> - - * ace/config-macosx.h: Updates for pthread configuration, provided - by John Zorko <j.zorko@att.net> and Rich Seibel - <seibel_r@ociweb.com>. - - * ace/OS.i: - * ace/Sock_Connect.cpp: Remove unnecessary __MACOSX__ macros. - See: Thu Oct 17 19:18:02 UTC 2002 Craig Rodrigues - <crodrigu@bbn.com> - -Mon Nov 25 18:20:41 2002 Steve Huston <shuston@riverace.com> - - * ace/config-hpux-11.00.h: Added #define ACE_SIZEOF_WCHAR _WCHAR_T - to get ACEXML to compile. ACEXML's usage of ACE_SIZEOF_WCHAR breaks - the previous allowance for ACE_SIZEOF_WCHAR to be sizeof (wchar_t) - because it uses ACE_SIZEOF_WCHAR in a math expression. - - * examples/C++NPv2/Reactor_Logging_Server_Adapter.cpp: Added - #include "ace/INET_Addr.h" to build clean on HP-UX. - -Mon Nov 25 20:32:02 UTC 2002 Don Hinton <dhinton@ieee.org> - - * ace/README: Added ACE_LACKS_PERFECT_MULTICAST_FILTERING. - -Mon Nov 25 18:58:33 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/Local_Name_Space_T.cpp: - * ace/Process.h: - * ace/Local_Name_Space_T.h: - * ace/ATM_Params.h: - * ace/Timer_Queue_Adapters.h: - Improved documentation for doxygen generation. - - * ace/Svc_Handler.cpp: - Made method definition and implementation arguments match. - -Sat Nov 23 20:01:36 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * apps/*/Makefile: Updated dependecies. Thanks to Erwin Rol for - pointing out the problem. - -Sat Nov 23 18:12:37 2002 Steve Huston <shuston@riverace.com> - - * ace/SString.cpp: Add explicit instantiation of - ACE_String_Base<char>::NULL_String_ and - ACE_String_Base<ACE_WSTRING_TYPE>::NULL_String. - * ace/Based_Pointer_Repository.cpp: Add explicit instantiation of - ACE_Singleton<ACE_Based_Pointer_Repository,ACE_SYNCH_RW_MUTEX>:: - singleton_. - * ace/Dynamic.cpp: Add explicit instantiation of - ACE_TSS_Singleton<ACE_Dynamic, ACE_Null_Mutex>::singleton_. - * tests/test_config.h: Added explicit instantiation of - ACE_Singleton<ACE_Test_Output, ACE_Null_Mutex>::singleton_. - - The above are only done when using automatic template instantiation - on AIX with g++. Without this, the compiler and linker don't get - things done right and static data members in templates don't get - instantiated correctly. Thanks to Max F. Bilyk <mbilyk@europe.com> - for reporting this problem. - - * ace/SSL/Makefile: Set LIBS = $(ACELIB). Else it won't link on - AIX, g++. - - * include/makeinclude/platform_aix4_g++.GNU: Removed. Use - platform_aix_g++.GNU instead. They were similar, though not equal, - and confusing each other. - - * include/makeinclude/platform_aix_g++.GNU: Fixed to work correctly - and is now the one to use for AIX 4 and 5 with g++. - -Sat Nov 23 18:03:41 2002 Steve Huston <shuston@riverace.com> - - * ace/String_Base.i (compare): Fix a size_t/int mismatch. - -Sat Nov 23 17:54:18 2002 Steve Huston <shuston@riverace.com> - - * ace/POSIX_CB_Proactor.{h cpp}: Changes to align with recent - ACE_POSIX_AIOCB_Proactor changes: changed handle_events (ulong) to - handle_events_i() and made nonvirtual. Removed get_result_status(), - cancel_aiocb(), and cancel_aio() since they simply forwarded to - ACE_POSIX_AIOCB_Proactor. Remove most of allocate_aio_slot(); call - up to ACE_POSIX_AIOCB_Proactor::allocate_aio_slot then tweak the - aiocb fields for the callback. - - * ace/config-hpux-11.00.h: Control the ACE_HAS_STANDARD_CPP_LIBRARY - and ACE_USES_STD_NAMESPACE_FOR_STDCPP_LIB settings by compiler- - defined macro _HP_NAMESPACE_STD, not passing them through from - platform_hpux_aCC.GNU. Similarly with ACE_HAS_EXCEPTIONS (define - it unless __HPACC_NOEH is set). This breaks a coupling from the - platform_hpux_aCC.GNU file, and will make autoconf easier as well - as making it easier for users to write makefiles without following - the ties through and trying to duplicate them. - Also, include config-posix.h to let it pick up system-supplied - settings. - - * ace/OS_Memory.h: In the HP aCC section, remove the checks for - aCC version > 32500 and _HP_NAMESPACE_STD... just use the - ACE_HAS_STANDARD_CPP_LIBRARY setting from config-hpux-11.00.h. - - * ace/Proactor.cpp (ctor): When no implementation is supplied, be - sure to make one up, even w/o presence of realtime signals. - - * include/makeinclude/platform_hpux_aCC.GNU: If stdcpplib is not - specified by user, set it to 1 always, unless the compiler is - a .2x version and not .27 (03.27 (01.27 on 10.20) is the first - version that can do -AA); makes it unnecessary to update this every - time HP releases a new compiler version. Also rearranged the - thread-related compiler option settings to be simpler and easier - to match with what the documentation says to set. Removed the - ACE_HAS_STANDARD_CPP_LIBRARY settings; they're defined in - ace/config-hpux-11.00.h when proper (see above). - Also, added +W930 to CCLFAGS when -AA to stop a spurious warning - about placement operator delete on aC++ 03.37, 03.39. - -Sat Nov 23 17:43:29 2002 Steve Huston <shuston@riverace.com> - - * ace/ACE.cpp (send_n): Properly update byte count for each message - block in the chain to send. Thanks to Venkita Subramonian - <venkita@cse.wustl.edu> for reporting this and sending a fix. - -Sat Nov 23 16:55:39 2002 Steve Huston <shuston@riverace.com> - - * tests/MEM_Stream_Test.cpp: - * tests/Notify_Performance_Test.cpp: - * tests/Priority_Reactor_Test.cpp: - * tests/Process_Manager_Test.cpp: - * tests/Reactor_Performance_Test.cpp: Replaced calls to the - deprecated ACE_Reactor::run_event_loop() and end_event_loop() - with run_reactor_event_loop() and end_reactor_event_loop(). - - * tests/Proactor_Test.cpp: Add missing \n in an ACE_DEBUG message. - -Sat Nov 23 00:16:59 2002 Craig Rodrigues <crodrigu@bbn.com> - - * ace/config-posix.h: Define ACE_HAS_PTHREADS_STD - if platform defines _POSIX_THREADS. - -Fri Nov 22 13:48:39 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/Containers_T.cpp (init_head): Broke up the cascading pointer - assignments to avoid a problem when using position independent - pointer. Thanks to Eugene Alterman <eugalt@myrealbox.com> for - reporting this. - - * ace/Svc_Conf.l: - * ace/Svc_Conf_l.cpp: Changed to always remove the matching quote - character (' or ") instead of first search for double quote, - then single quote. Thanks to Johnny Willemsen - <jwillemsen@remedy.nl> for sumitting the patch. - -Fri Nov 22 04:52:16 2002 Bala <bala@cs.wustl.edu> - - * ACE version 5.2.6 released. - -Tue Nov 19 22:36:38 2002 Ossama Othman <ossama@uci.edu> - - * ace/Time_Value.h (sec, usec, msec): - - Documentation updates/clarifications. - - * ace/Time_Value.cpp (operator++, operator--): - - Fixed postfix increment and decrement operators. Previously - they had the same implementation as their prefix increment and - decrement operator counterparts. A copy of the ACE_Time_Value - is now done prior to in/decrementing it via the corresponding - prefix operator. That copy is returned. A copy is only - performed for the postfix case. - -Tue Nov 19 18:14:10 2002 Pradeep Gore <pradeep@oomworks.com> - - * bin/auto_run_tests.lst: - Modified Notify configurations. - - * bin/auto_run_tests.pl: - Checkin on behalf of Irfan - Changed logic so that it checks for - the existence of the executable and not of the parameters passed - to the executable. - -Tue Nov 19 07:09:15 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: Tests not runnable in minimum - configuration were being run. Hence prevented them from - running. - -Mon Nov 18 20:49:33 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * examples/IPC_SAP/FILE_SAP/client.cpp (ACE_TMAIN): Changed to use - predefined ACE_DEFAULT_FILE_PERMS instead of the non-portable 0666. - -Mon Nov 18 12:27:46 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/Dev_Poll_Reactor.cpp (handler): - * ace/WFMO_Reactor.i (handler): - * ace/Select_Reactor_Base_T.cpp (handler_i): The methods that - returns a signal handler associated with a certain type of - signal, does restrictive sanity checks. One of the checks that - it performs is to see whether a pointer to an event_handler is 0 - just before assigning a new event handler pointer. This check - was not required at all and hence it is now removed. Thanks to - Edan Ayal for reporting this. - -Sat Nov 16 09:38:38 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/*/Makefile: - * tests/Makefile: - * examples/*/Makefile: - * performance-tests/*/Makefile: - * ACEXML/*/Makefile: Updated dependencies. - -Fri Nov 15 20:33:02 UTC 2002 Don Hinton <dhinton@ieee.org> - - * ace/SOCK_Dgram_Mcast.h: - * ace/config-all.h: Added new macro for platforms that don't yet - implement the new IGMPv3, perfect multicast filtering, and - replaced the #ifdef linux in the ACE_SOCK_Dgram_Mcast::ctor - with the new macro, since this isn't just a linux thing. - - * ace/config-aix-4.x.h: - * ace/config-freebsd.h: - * ace/config-linux-common.h: - * ace/config-netbsd.h: - * ace/config-openbsd.h: - Defined ACE_LACKS_PERFECT_MULTICAST_FILTERING for these platforms. - - * ace/SOCK_Dgram_Mcast.cpp (open_i): Removed redundant call to - ACE_SOCK::set_option() to set SO_REUSEADDR, since it's set in - the call to ACE_SOCK::open(). This fixes [bug 1111]. - - * tests/Multicast_Test.cpp: Enhanced test by adding new options - control the wait time when running test as a CONSUMER only, - and contol the TTL so the test could be used over routers. - Also cleaned up the code a bit to make it more readable. Thanks - to Bill Fulton for motivating these enhancements. - -Fri Nov 15 14:16:10 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/Thread_Exit.h: - * ace/Thread_Exit.cpp (cleanup): Changed the type of the argument - <cleanup> takes to the generic (void*) and avoid including - <Synch.h> in Thread_Exit.h. This eliminated the circular header - dependencies when ACE_LEGACY_MODE is define. - - * ace/ace_dll.dsp: - * ace/ace_lib.dsp: Removed some template files from the build - lists and make sure all template files are not being built. - -Fri Nov 15 13:22:41 2002 Steve Huston <shuston@riverace.com> - - * tests/Proactor_Test.cpp: On Win32, check for canceled I/O using - ERROR_OPERATION_ABORTED instead of ECANCELED. - -Fri Nov 15 16:51:11 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * ace/POSIX_CB_Proactor.h: - * ace/POSIX_CB_Proactor.cpp: - Do not build on Lynx. - -Fri Nov 15 14:14:39 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * ace/POSIX_CB_Proactor.h: - * ace/POSIX_CB_Proactor.cpp: - Do not build on Sun. - -Fri Nov 15 14:08:11 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * include/makeinclude/platform_linux.GNU: - Linux glibc has aio_ and pthread_ calls in librt.so. - Add fix so that librt.so is linked in if threads=0. - -Fri Nov 15 07:48:37 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * tests/OS_Test.cpp: Added the test cases for itoa from - Joseph Sarbak <joseph.sarbak@csfb.com>. - -Fri Nov 15 07:08:46 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: Fixed a few minor problems that showed - up in the daily builds. - -Thu Nov 14 22:08:49 2002 Steve Huston <shuston@riverace.com> - - * tests/Proactor_Test.cpp: Make this test behave better: - - At the end of running time, close all the Senders. This should - cancel all the Sender outstanding operations and trip all the - Receivers into finishing their receives and closing themselves. - - Change the default Sender/Receiver count from 20 to 10. This - shrinks the log a bit without sacrificing test quality. - - Change the "inferior platforms" default Sender/Receiver count - from 1 to 3. With Proactor improvements made so far, let's give - these platforms some more exercise and see what happens. - -Thu Nov 14 23:10:32 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * ace/POSIX_CB_Proactor.cpp: - * ace/POSIX_CB_Proactor.h: - * ace/POSIX_CB_Proactor.i: - * ace/POSIX_Proactor.cpp: - * ace/POSIX_Proactor.h: - * ace/POSIX_Proactor.i: - * ace/Proactor.cpp: - * tests/Proactor_Test.cpp: - Fixed Proactor so that on platforms that support POSIX IO, - but not POSIX RT signals, it will compile. Use the new - ACE_HAS_POSIX_REALTIME_SIGNALS macro. - -Thu Nov 14 22:47:45 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * include/makeinclude/platform_freebsd.GNU: Merged with the - pthread version of this file. To compile without threads, - set, do make threads=0, or set threads=0 in this file. - * include/makeinclude/platform_freebsd_pthread.GNU: Removed. - * ace/config-freebsd.h: Merge with pthread version of this file. - * ace/config-freebsd-pthread.h: Removed. - * ace/OS.h: Use new ACE_HAS_POSIX_REALTIME_SIGNALS macro for - POSIX RT signals stuff, instead of ACE_HAS_AIO_CALLS. - FreeBSD has POSIX AIO, but not POSIX RT signals. - * ace/config-sunos5.6.h: Define ACE_HAS_POSIX_REALTIME_SIGNALS, - ACE_HAS_POSIX_MESSAGE_PASSING. - * ace/config-linux-common.h: Define ACE_HAS_POSIX_REALTIME_SIGNALS, - ACE_HAS_AIO_CALLS. - -Thu Nov 14 16:19:57 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: Added two tests, Bug_1269_Regression and - Bug_1270_Regression into the daily regression suite. They will - not run on Win32. - -Thu Nov 14 13:35:44 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/Containers_T.cpp: Changed to deallocate internal nodes in - ACE_Double_Linked_List with the allocator. Thanks to Eugene - Alterman <EugeneA@bremer-inc.com> for reporting the problem. - -Thu Nov 14 10:57:47 2002 Carlos O'Ryan <coryan@atdesk.com> - - * ace/OS.h (ace_timezone): - - Fixed broken implementation for Linux. Call tzset() and then - return the global "timezone" variable instead. [Bug 1359] - -Thu Nov 14 12:19:51 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/Creator.pm: - - Modify the way projects are created on windows. If using - cygwin perl then use \n instead of \r\n. - - * bin/MakeProjectCreator/modules/VC7WorkspaceCreator.pm: - - Sort the list of projects to make the generated solution file more - similar to actual solution files. - -Thu Nov 14 06:16:51 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/Parser.pm: - - Take into account relative directories in the cd() method. - The original optimization overlooked this. - -Thu Nov 14 06:59:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/Date_Time.{h,i}: - Added update() method with which it is possible to set the values - in ACE_Date_Time based on a passed ACE_Time_Value. - - * ace/Time_Request_Reply.{h,cpp}: - Made the argument of the timeout method const&. Updated doxygen - documentation and marked init() as deprecated because it is a no-op. - -Wed Nov 13 18:28:39 2002 Steve Huston <shuston@riverace.com> - - * tests/FIFO_Test.cpp: Changed to both clarify the usage of - ACE_FIFO_(Send|Recv)_Msg with and without ACE_HAS_STREAM_PIPES - and test that usage. - -Wed Nov 13 17:39:41 2002 Steve Huston <shuston@riverace.com> - - * ace/POSIX_Proactor.{h cpp}: - * ace/SUN_Proactor.{h cpp}: Leave two handle_events() methods - available; one with a timeout, one without. Each proactor's - implementations calls an internal handle_events_i() method - that does the right thing. - (start_aio): This is a virtual method used to start an asynch I/O - request from outside the proactor (such as from an asynch read - stream). When the actual operation is to start, start_aio calls - start_aio_i(), implemented differently for the different proactors. - Similar arrangement for cancel_aio(). - - * tests/Proactor_Test.cpp: Cancel all outstanding operations before - trying to abruptly stop everything. - -Wed Nov 13 17:29:31 2002 Steve Huston <shuston@riverace.com> - - * ace/WIN32_Asynch_IO.h (cancel_uncompleted): Filled in - description of the 'set' parameter. Thanks to Craig - Rodrigues for noticing its omission. - -Wed Nov 13 16:52:04 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * ace/Lib_Find.cpp (ldname): MacOS X prepends an underscore to - symbols in shared libaries. - - * ace/config-macosx.h: - * include/makeinclude/platform_macosx.GNU: - - - add support for dlcompat, a third party package which provides - dlopen(), dlsym(), dlcose() functions - - - Define ACE_LD_SEARCH_PATH to DYLD_LIBRARY_PATH and - ACE_LD_SUFFIX to .dylib - -Tue Nov 12 14:12:59 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - This is a hackish fix for BUG 1361. Please see - http://deuce.doc.wustl.edu/bugzilla/show_bug.cgi?id=1361 - for details of this problem - - * ace/TP_Reactor.cpp (handle_socket_events): A hack to prevent - BUG 1361 from surfacing for applications like TAO, using the - TP_Reactor. This fix basically ensures that resume_i () is - not called naively. - - * ace/Event_Handler.h: Added a enum type which could be could be - used by the TP_Reactor. - -Tue Nov 12 11:53:29 2002 Steve Huston <shuston@riverace.com> - - * tests/run_test.lst: Change FIFO_Test from !MSVC to !Win32 to - prevent it from running with Borland on Windows. Thanks to - Johnny Willemsen for pointing this out. - -Tue Nov 12 14:11:33 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/Time_Value.h: - Doxygen documentation improvements. - -Tue Nov 12 13:15:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/Message_Queue_T.h: - Doxygen documentation improvements. - -Mon Nov 11 20:37:18 2002 Ossama Othman <ossama@uci.edu> - - * bin/auto_run_tests.lst: - - Added Jai's application-controlled membership load balancing - test to the list of regression tests to be run. - -Mon Nov 11 17:57:12 2002 Ossama Othman <ossama@uci.edu> - - * bin/auto_run_tests.lst: - - Added Carlos' new PortableInterceptor AMI test to the list of - regression tests to be run. - -Mon Nov 11 16:17:59 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * bin/MakeProjectCreator/modules/TemplateParser.pm (relative): - Changed to invoke slash_to_backslash from $self. The original - code only runs on some specific version of perl. - -Mon Nov 11 13:35:53 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/Hash_Map_Manager_T.h: Updated - comments on the class ACE_Hash_Map_Bucket_Iterator. Thanks to - Kitty and Irfan for helping with this. - -Mon Nov 11 12:52:41 2002 Steve Huston <shuston@riverace.com> - - * docs/ACE-subsets.html: Added the Timer component to the list of - buildable pieces. Thanks to Jason Pasion <J.Pasion@motorola.com> - for reporting this problem. - - * THANKS: Added Jason Pasion to the Hall of Fame. - -Mon Nov 11 12:39:18 2002 Steve Huston <shuston@riverace.com> - - * ace/FIFO_Send_Msg.cpp (send): When setting iovec length, - static_cast the int length to size_t; the target is not an - int (as the strbuf length is). When returning the sent byte - count, don't include any additional bytes added internally - for record framing. - - * tests/FIFO_Test.cpp: New test to exercise basic and overflow - message-based FIFO I/O. - - * tests/Makefile: - * tests/run_test.lst: Added FIFO_Test. Disabled on MSVC builds - since Windows doesn't have FIFOs. - -Mon Nov 11 09:58:13 2002 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * ace/CDR_Stream.{i,cpp} Add casts to ACE_CDR::Octet and - ACE_CDR::Boolean to prevent warnings when run with MSVC++ level - 4 warnings. Thanks to Lukas Gruetzmacher - <gruetzmacher@ais-dresden.de> for motivating this fix. - - * ace/OS_String.cpp (itoa_emulation): Updated the code so that it'll - work correctly with negative numbers. Thanks to Joseph Sarbak - <joseph.sarbak@csfb.com> for reporting this and providing a fix. - -Sun Nov 10 19:38:45 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/Message_Block.cpp(copy): The copy () methods were copying - data till the end_ pointer. This is a wrong assumption, - especially in places where the alloted bytes are less than the - allocated bytes. Now we use the space () method to find out the - amount of space available in the data block for copying. Thanks - to Edan Ayal for finding this and providing a patch. - - * ace/Simple_Message_Block_Test.cpp: Added a test which tests - copying in a case where the alloted space is less than the - allocated space. Thanks to Edan Ayal for this test case. - -Sat Nov 9 12:37:54 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * docs/ACE-guidelines.html: Updated the ACE guidelines to - refer to the correct Doxygen links. Thanks to - Johnny Willemsen for reporting this. - -Sat Nov 9 21:48:31 2002 Steve Huston <shuston@riverace.com> - - * tests/Proactor_Test.cpp: When disabling Nagle's, use IPPROTO_TCP, - not SOL_SOCKET, for the setsockopt level. Fixes set_option errors - in this test log. - -Sat Nov 9 18:49:11 2002 Steve Huston <shuston@riverace.com> - - * ace/POSIX_Proactor.{h cpp i} (create_asynch_read_stream, - create_asynch_write_stream, create_asynch_read_dgram, - create_asynch_write_dgram, create_asynch_read_file, - create_asynch_write_file, create_asynch_accept, - create_asynch_connect, create_asynch_transmit_file): - Refactored all the factory creation methods from - ACE_POSIX_AIOCB_Proactor to ACE_POSIX_Proactor. - Moved the pseudo task from ACE_POSIX_AIOCB_Proactor to - ACE_POSIX_Proactor as well. - - (handle_events, cancel_aio): - Added these methods to ACE_POSIX_Proactor as pure virtual, - forcing their definition in the specific implementations. - - (start_aio): This is now a pure virtual function that all - more specific proactors need to implement. The old start_aio() - is now start_aio_i(). What used to be ACE_POSIX_AIOCB_Proactor:: - register_and_start_aio() is now ACE_POSIX_AIOCB_Proactor::start_aio. - Which type of I/O to start is now defined by an enum Opcode - that's passed to start_aio(). - - * ace/POSIX_Asynch_IO.{h cpp} (ACE_POSIX_Asynch_Operation): - Replace references to ACE_POSIX_AIOCB_Proactor with - ACE_POSIX_Proactor. Refactoring things out of specific proactor - types. - (ACE_POSIX_Asynch_Read_Stream, ACE_POSIX_Write_Stream, - ACE_POSIX_Asynch_Read_File, ACE_POSIX_Asynch_Write_File, - ACE_POSIX_Asynch_Accept, ACE_POSIX_Asynch_Connect, - ACE_POSIX_Asynch_Transmit_File, ACE_POSIX_Asynch_Transmit_Handler, - ACE_POSIX_Asynch_Read_Dgram, ACE_POSIX_Asynch_Write_Dgram): - Replace args of type ACE_POSIX_AIOCB_Proactor with - ACE_POSIX_Proactor. Matches POSIX_Proactor.* changes above. - Removed all the methods that said they were only there to kill - dominance warnings. These are not needed. - - These things all simplify the proactor framework code. Thanks to - Alex Libman for these ideas. - -Sat Nov 9 13:16:50 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * THANKS: Added Simon McQueen to the hall of fame. - -Fri Nov 8 13:04:10 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/ProjectCreator.pm: - - Fixed a bug in MPC where the idl file name could cause - an error on Windows if it had a slash in it. It would be - converted to a back slash and then interpreted inside a - regular expression which would cause MPC to abort. - -Fri Nov 8 13:00:56 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/BorlandProjectCreator.pm: - * bin/MakeProjectCreator/modules/BorlandWorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/Creator.pm: - * bin/MakeProjectCreator/modules/Driver.pm: - * bin/MakeProjectCreator/modules/EM3ProjectCreator.pm: - * bin/MakeProjectCreator/modules/EM3WorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/GHSProjectCreator.pm: - * bin/MakeProjectCreator/modules/GHSWorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/GNUProjectCreator.pm: - * bin/MakeProjectCreator/modules/GNUWorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/GUID.pm: - * bin/MakeProjectCreator/modules/NMakeProjectCreator.pm: - * bin/MakeProjectCreator/modules/NMakeWorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/Parser.pm: - * bin/MakeProjectCreator/modules/ProjectCreator.pm: - * bin/MakeProjectCreator/modules/TemplateInputReader.pm: - * bin/MakeProjectCreator/modules/TemplateParser.pm: - * bin/MakeProjectCreator/modules/VA4ProjectCreator.pm: - * bin/MakeProjectCreator/modules/VA4WorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/VC6ProjectCreator.pm: - * bin/MakeProjectCreator/modules/VC6WorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/VC7ProjectCreator.pm: - * bin/MakeProjectCreator/modules/VC7WorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/WorkspaceCreator.pm: - - Performed various optimization which dramatically decreased the - amount of time to generate projects for the entire ACE tree. In - tests, the time went from 2 minutes to 1 minute. - -Fri Nov 8 12:57:57 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/README: - - Added documentation for the new verbatim keyword. - - * bin/MakeProjectCreator/modules/ProjectCreator.pm: - * bin/MakeProjectCreator/modules/TemplateParser.pm: - - Added functionality to place test verbatim into a generated - project file. - - * bin/MakeProjectCreator/templates/gnu.mpd: - - Added verbatim markers top, macros, local and bottom. - -Fri Nov 8 12:52:16 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/README: - - Added documentation about the new idlgendir keyword. - - * bin/MakeProjectCreator/modules/GHSWorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/Parser.pm: - * bin/MakeProjectCreator/modules/ProjectCreator.pm: - * bin/MakeProjectCreator/modules/TemplateParser.pm: - * bin/MakeProjectCreator/modules/VC6WorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/VC7WorkspaceCreator.pm: - - Modified flag_overrides to work with both idlflags and idlgendir. - - * bin/MakeProjectCreator/templates/bor.mpd: - * bin/MakeProjectCreator/templates/em3vcp.mpd: - * bin/MakeProjectCreator/templates/gnu.mpd: - * bin/MakeProjectCreator/templates/nmake.mpd: - * bin/MakeProjectCreator/templates/vc6dsp.mpd: - * bin/MakeProjectCreator/templates/vc7.mpd: - - Modified all templates that deal with idl files to utilize the - idlgendir keyword. - -Fri Nov 8 04:03:29 UTC 2002 Don Hinton <dhinton@ieee.org> - - * tests/Multicast_Test.cpp: Cleaned up error messages and - error handling and insure that errors are returned. Added - 2 second timeout to call to ACE_Thread_Manager::wait() so - that lost messages won't cause test to hang. - -Wed Nov 6 18:55:22 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.{h i} (recursive_mutex_cond_unlock, - recursive_mutex_cond_relock): For Windows, don't try to save/restore - CRITICAL_SECTION members individually - there are undocumented - details at work there. Instead, release and acquire the mutex - one fewer times than the recursion count, letting Windows take - care of its internals. Fixes the hung Recursive_Condition_Test on - Windows. This also neatly removes the WinCE conditional code in - this area. - - * tests/Recursive_Condition_Test.cpp: Fix a const error that gcc - pointed out. - -Wed Nov 6 11:12:28 2002 Steve Huston <shuston@riverace.com> - - * tests/Recursive_Condition_Test.cpp: Replaced ACE_TRACE with - ACE_DEBUG so the log gets all the info. Re-enabled the timer queue - part of the test and added a check to be sure "enough" expirations - are recorded. Added a piece of data protected by the mutex/condition - to be sure threads aren't released too early. - -Tue Nov 5 21:54:40 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * tests/run_tests.lst: - * tests/run_tests_remote.lst: Removed these files, since they - served no purpose other than adding to the noise. - -Tue Nov 5 15:51:29 2002 Steve Huston <shuston@riverace.com> - - * ace/INET_Addr.cpp (set_addr): Add missing ';' in the IPv6 case. - -Tue Nov 5 20:26:33 UTC 2002 Don Hinton <dhinton@ieee.org> - - * tests/tests.mpc: Added Multicast_Test entry. - - * tests/README: Upadate instructions for adding new tests. - -Tue Nov 5 12:12:05 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * tests/Multicast_Test.cpp: Fixed compile errors in single threaded - builds. - -Tue Nov 5 11:24:03 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/Process.h: - * ace/Process.i: Added a new method <return_value> which returns - the actual value returned or <exit>'ed by the child process. - The <exit_code> accessor returns the raw exit status returned by - the system APIs and the value is OS dependent. Changed the - <exit_code> mutator to be a proctected method and declare - ACE_Process_Manager a fried of ACE_Process as <exit_code> - mutator should only be used by ACE_Process_Manager. - - Thanks to Stephen Blake <sblake@speakeasy.net> for reporting the - problem. - -Tue Nov 5 05:12:16 2002 Ossama Othman <ossama@uci.edu> - - * ace/Acceptor.h: - * ace/Connector.h: - * ace/Strategies_T.h: - - Added missing STL-style template traits found in C++NPv2 - including the following: - - addr_type - acceptor_type - connector_type - handler_type - stream_type - base_type - accept_strategy_type - connect_strategy_type - creation_strategy_type - concurrency_strategy_type - recycling_strategy_type - scheduling_strategy_type - - where applicable. The old traits are still available and remain - unchanged. - -Tue Nov 5 05:54:16 UTC 2002 Don Hinton <dhinton@ieee.org> - - * tests/Multicast_Test.cpp: Fixed typo in pragma instantiation, - and replaced an instance of char* with ACE_TCHAR* to fix unicode - builds. Added #ifdef for ACE_HAS_THREADS, since it uses - ACE_Task as an active object. - - * tests/run_test.lst: Turned off Multicast_Test and - Token_Stategy_Test for single threaded builds. - -Mon Nov 4 18:01:25 UTC 2002 Don Hinton <dhinton@ieee.org> - - * tests/Makefile.bor: Added new Multicast_Test. Thanks to - Johnny Willemsen <jwillemsen@remedy.nl> for noticing this. - -Mon Nov 4 16:14:21 UTC 2002 Don Hinton <dhinton@ieee.org> - - * tests/Multicast_Test.cpp (get_handle): Made return type of - definition ACE_HANDLE to match the declaration. Thanks to - Ed Skees <eskees@dctd.saic.com> for reporting this. - - * tests/Multicast_Test.dsp: - * tests/tests.dsw: Added new dsp. Thanks again to Ed Skees - for providing this. - - * THANKS: Added Ed Skees to hall of fame... - -Mon Nov 4 09:28:34 UTC 2002 Don Hinton <dhinton@ieee.org> - - * ace/SOCK_Dgram.cpp (recv): Corrected method signature in .cpp to - match the one in the .h. The .h version specified iovec*, but - the .cpp had iovec[] which confused doxygen. This fixes bug [1353]. - -Sun Nov 3 19:41:30 UTC 2002 Don Hinton <dhinton@ieee.org> - - * tests/Multicast_Test.cpp: - * tests/Makefile: - * tests/run_test.lst: Added a new test to exercise some of the new - functionality of ACE_SOCK_Dgram_Mcast. - -Sun Nov 3 10:36:50 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * tests/run_test.lst: Prevented SV_Shared_Memory test from running - in RH_7.1 configurations. Please see BUGID 1351 for details. - -Sat Nov 2 17:44:00 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/Select_Reactor_Base.h: - * ace/Select_Reactor_Base.cpp: Fixed a subtle bug in the - Select_Reactor. The following were performed during the bind () - operation, which binds a handle with interested mask to an - event_handler, - - - The new masks were added to the existing wait_set - - and the new masks were also removed from the suspend_set. - - Though the above steps were fine, there exists a possibility - when the reactor could land up with a state when the handle - could be both in the wait_set and suspend_set. If suppose a - handle is initially registered for READ and EXCEPTION with the - reactor and then suspended. If the user calls register_handler - () to register the same handle for READ and WRITE, things could - go totally bizzare. - - The present fix does the following - - - If the handle is suspended, the masks are added to the - suspend_set - - else the masks are added to the wait_set and the flag that - indicates state changes within the reactor is set to 1. - - A pure virtual call is_suspended_i () was added to the - ACE_Select_Reactor_Base class to make the check for a suspended - handle possible within the bind () call. Thanks to Irfan - Pyarali<irfan@oomworks.com> who reported this problem. This - should fix the problem reported in [BUG 1305]. - -Fri Nov 1 07:35:10 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/auto_run_tests.lst: Added the new - MProfile_Connection_Timeout test to the daily regression tests. - -Thu Oct 31 19:15:49 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ACEXML/common/Attributes.h: Added the missing inclusion of - ACEXML_Export.h and the #pragma once directive. Thanks to Paul - Marquis <pmarquis@pobox.com> for reporting this. - -Thu Oct 31 12:57:10 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/OS.i (fstat): Added check for FILE_ATTRIBUTE_DIRECTORY on - Win32 to set the file type in the st_mode so one can determine - if it's a file or a directory. Thanks to Paul Marquis - <pmarquis@pobox.com> for submitting the patch. - -Wed Oct 30 13:47:14 2002 Steve Huston <shuston@riverace.com> - - * ace/Thread_Exit.cpp: Add #include "ace/Managed_Object.h" to get - the definition of ACE_Managed_Object. - -Wed Oct 30 11:38:36 2002 Irfan Pyarali <irfan@oomworks.com> - - * ace/Timeprobe.h: Fixed formatting of comments that somehow got - messed up in: - - http://cvs.doc.wustl.edu/viewcvs.cgi/ace/Timeprobe.h.diff?r1=4.29&r2=4.30 - -Tue Oct 29 22:27:24 2002 Frank Hunleth <frank@hunleth.com> - - * apps/gperf/src/Key_List.cpp: Added cast to int on UCHAR_MAX and - USHRT_MAX to remove warnings of comparing signed and unsigned - values under OpenBSD. - - * ace/config-openbsd.h: - * ace/config-openbsd-pthread.h: Added ACE_HAS_MKSTEMP since - OpenBSD has the mkstemp() function, and it helps clear up - linker warnings. Also added ACE_HAS_ALT_CUSERID to clear - up errors due to _cuserid(). Thanks to Craig Rodrigues - for the latter patch. - -Tue Oct 29 15:45:06 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/Thread_Exit.h: - * ace/Thread_Exit.cpp: - * ace/Thread_Manager.h: - * ace/Thread_Manager.cpp: Changed to use ACE_Thread_Manager to - clean up the global TSS ACE_Thread_Exit object instead of - depending on the more general ACE_Object_Manager::at_exit method - so we can better control the order of destruction. The change - makes sure the global TSS ACE_Thread_Exit object will only get - deleted *after* ACE_Thread_Manager has cleaned up all threads - and destroyed itself. - - Thanks to Patrick McNerthney <pat@mcnerthney.com> and - Daniel Bell <Daniel.Bell@colorbus.com.au> for reporting - the problem. - -Tue Oct 29 16:19:54 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * ace/INET_Addr.cpp: - * ace/INET_Addr.i: - * ace/INET_Addr.h: - - Added a reset() method to initialize the underlying sockaddr_in - or sockaddr_in6 structure, and set the address family. Changed - all constructors to use this reset() method. [Bug 1342] - - (INET_Addr::set_addr) the port number was not being set, - so set it. [Bug 1329] - * tests/INET_Addr_Test.cpp: Add test to verify that address family - is being set. - * THANKS: Added Thomas Devanneaux <tdevanne@akamai.com> who provided - these fixes. - -Mon Oct 28 20:46:33 2002 Steve Huston <shuston@riverace.com> - - * ace/SOCK_Dgram_Mcast.cpp (ACE_SDM_helpers::addr_to_string): Use - the "<?>" string in place rather than try to declare it a static. - gcc reports the original usage won't do what you think (keep one - copy for the method) but will duplicate it in each generated - method. Use it as a literal and let the compiler try to place a - single copy in a read-only section. - -Mon Oct 28 20:38:27 2002 Steve Huston <shuston@riverace.com> - - * ace/Asynch_IO.cpp (ACE_Service_Handler::addresses()): Don't - print addresses from here. It's inappropriate for a framework - to be printing things out without being asked to. - - * ace/POSIX_Asynch_IO.cpp (ACE_POSIX_Asynch_Read_Stream::read): - * ace/WIN32_Asynch_IO.cpp (ACE_WIN32_Asynch_Read_Stream::read): - Don't print a message for a 0-byte/no space read - set errno to - ENOSPC so the caller can figure out what's going on. - - * ace/SUN_Proactor.{h cpp}: - * ace/POSIX_Proactor.{h cpp} (ACE_POSIX_AIOCB_Proactor):Change - 'return_status' arg to get_result_status(), find_completed_aio() - from int to size_t and rename transfer_count; get_result_status() - takes care of sensing -1 count and changing to 0. Passing back - a size_t smooths the path from here through to the result object. - Removed application_specific_code() - reuse the one from - ACE_POSIX_Proactor - this one called it; remove the middle-man. - - * tests/Proactor_Test.cpp: Added addresses() method implementations - to print address with session IDs; helps to match Sender/Receiver - pairs in the log. Also added some logging of basic send/recv info - to help try to track down why this facility doesn't work well. - Added a check for comparable sends/receives when a session ends. - Added a warning if there are outstanding I/O when the session - ends. This probably should be an error, but I haven't thought - through it enough to go that far. For the SUN Proactor, use one - thread by default (not 3) - aiosuspend() is not MT safe. - -Mon Oct 28 12:48:14 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * bin/PerlACE/Process_Unix.pm (Spawn): Return 0 when the function - complete successfully. Thanks to Charles F. Frasch - <cfrasch@spawar.navy.mil> for reporting and suggesting the fix. - -Sun Oct 27 07:44:32 2002 John Aughey <jha@aughey.com> - - * ace/INET_Addr.cpp: Changed the default constructor to only - set the address family rather than doing a full set. This - should exhibit the same behavior without additional overhead. - -Sat Oct 26 21:49:15 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/Process.cpp (wait): - * ace/Process.i (wait): Changed to always update this->exit_code_ - when waiting on a child process to allow <exit_code> method to - function properly. Removed the use of the macro - ACE_OSCALL_RETURN on ACE_OS::waitpid since ACE_OS::waitpid is - already signal safe. - - * ace/Process.h: Changed the type of this->exit_code_ from int to - ACE_exitcode. Updated documentation for mutator method - <exit_code> to better reflect the intention for using this - method. - - Thanks to Artashes Ghazaryan <arty@unicad.am>, Jaroslaw Nozderko - <jaroslaw.nozderko@polkomtel.com.pl>, and Stephen Blake - <sblake@speakeasy.net> for reporting the problem. - -Sat Oct 26 11:42:04 UTC 2002 John Aughey <jha@aughey.com> - - * ace/INET_Addr.cpp: Changed default behavior to set 0 - port and INADDR_ANY address. - -Sat Oct 26 09:45:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/Timer_Heap_T.h: - Doxygen documentation fix - - * ace/OS.h: - The BCB4 header files don't define ENOTEMPTY, so define this when - detecting the BCB4 compiler. Thanks to Mike Curtis <mcurtis@wt.net> - for reporting this. - - * examples/Reactor/Proactor/test_proactor2.cpp: - Removed not needed return. Fixes a BCB4 build error - - * THANKS: Added Mike Curtis - -Wed Oct 23 16:37:15 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * THANKS: Added Sandeep to the Hall of Fame. - -Wed Oct 23 17:20:18 UTC 2002 Don Hinton <dhinton@ieee.org> - - * ace/Service_Object.{h,cpp} (suspend, resume, fini): - Changed return value from void to int so that the - return value of the contained ACE_Service_Type could be - propagated. - - * ace/Service_Repository.cpp (suspend, resume, fini): - Changed to propagate the return value from the - ACE_Service_Type instead of always 0. - -Tue Oct 22 16:26:56 2002 Rich Seibel <seibel_r@ociweb.com> - - * ace/Basic_Types.h: - Added a test for _BYTE_ORDER, same as BYTE_ORDER and - __BYTE_ORDER. VxWorks defines this. Also note, compilers - that previously automatically looked for _BYTE_ORDER when - only BYTE_ORDER was specified will likely not do so in the - future. This will proactively fix them also. - - * ace/config-all.h: - Moved the use of FUNCPTR until after the include that defines - it. Only affects VxWorks. - -Tue Oct 22 13:11:42 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/OS.cpp (DllMain): Added code to invoke - DisableThreadLibraryCalls during PROCESS_ATTACH when - ACE_DISABLES_THREAD_LIBRARY_CALLS is defined to 1 as described - below. This should enhance performance of applications that - spawn a lot of threads. Thanks to Patrick McNerthney - <pat@mcnerthney.com> for suggesting the change. - - We should probably apply the same optimization for the core TAO - libraries. - - * ace/config-win32-common.h: Defined a Win32-only new macros - ACE_DISABLES_THREAD_LIBRARY_CALLS which is default to 0 if not - defined by users. - -Mon Oct 21 17:26:15 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/SPIPE_Acceptor.cpp: Modified - SPIPE_Acceptor::create_new_instance to manually signal the event - used in the ConnectNamedPipe call when an ERROR_PIPE_CONNECTED - status is returned. Also added the handling of the status of - ERROR_NO_DATA to the same logic as ERROR_PIPE_CONNECTED. Thanks - to Patrick J. McNerthney <pat@mcnerthney.com> for contributing - this. - -Tue Oct 22 10:20:41 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * examples/C++NPv2/Reactor_Logging_Server_T.cpp: Fixed a typo that - broke the builds. My bad :( - -Mon Oct 21 20:13:27 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * examples/C++NPv2/Reactor_Logging_Server_T.cpp: - * examples/C++NPv1/Logging_Server.cpp: - * examples/C++NPv1/Process_Per_Connection_Logging_Server.cpp: - * examples/C++NPv1/RT_Thread_Per_Connection_Logging_Server.cpp: - * examples/C++NPv1/Thread_Per_Connection_Logging_Server.cpp: - Fix for compile errors on Lynx platform. Thanks to Craig - Rodrigues for supplying the patches. - -Tue Oct 22 12:36:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/Basic_Stats.h: - * ace/Dirent_Selector.h: - * ace/Service_Config.h: - * ace/Timer_Heap_T.h: - Improved doxygen documentation. - -Mon Oct 21 22:30:51 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * tests/SString_Test.cpp (ACE_TMAIN): Fixed a warning in Win32 builds. - -Mon Oct 21 20:04:12 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/WIN32_Asynch_IO.h: - * ace/WIN32_Asynch_IO.cpp: Integrated the latest patches from Eric - Frias <efrias@syncad.com>. - -Mon Oct 21 06:54:12 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * tests/Recursive_Condition_Test.cpp (test_4): - * tests/RB_Tree_Test.cpp: Fixed compile errors arising out of M$ - quirks. The scoping rules for 'for' loops are busted and hence - these fixes. These fixes basically change the variable names for - the different for loops inside the same scope. - -Sun Oct 20 21:40:13 2002 Krishnakumar B <kitty@cse.wustl.edu> - - * ACEXML/common/HttpCharStream.cpp: Handle all the error paths - gracefully and don't leak memory/sockets. - -Sun Oct 20 14:06:36 2002 Krishnakumar B <kitty@cse.wustl.edu> - - * ACEXML/common/FileCharStream.cpp (peek_i): Fixed more - compilation errors in Borland WCHAR builds. MSVC sucks... - -Sun Oct 20 11:36:34 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ACEXML/common/FileCharStream.cpp: Fixed a compile error in - Borland builds. The compile error is due to strict enforcement - of scoping rules in a for loop by Borland compiler. - -Sun Oct 20 11:31:45 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ACEXML/common/Makefile.bor: Added Encoding.obj to the list of - object files. Thanks to Johnny for pointing this out. - -Sun Oct 20 11:30:04 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * tests/SPIPE_Test.cpp: Fixed compile errors in Borland builds. - -Sun Oct 20 10:16:23 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * examples/C++NPv2/display_logfile.cpp: Fix for single threaded - builds. - -Sun Oct 20 10:12:42 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * tests/Recursive_Condition_Test.cpp: Fix for single threaded - builds. The ACE_HAS_THREADS were improperly placed. - -Sun Oct 20 09:26:58 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * examples/C++NPv1/Makefile: Added changes from "Fri Oct 18 - 21:01:25 UTC 2002 Craig Rodrigues <crodrigu@bbn.com>" back - since the changes are fine. Thanks to Craig for pointing that - out. - -Sun Oct 20 13:34:45 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * examples/C++NPv2/Makefile: Replace with recursive Makefile which - does not use for loops. Add .NOTPARALLEL target to fix - build errors on parallel builds. - -Sun Oct 20 08:10:43 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * tests/Recursive_Condition_Test.cpp: Fixed a unused variable warning. - -Sun Oct 20 00:27:02 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * tests/Recursive_Condition_Test.cpp: Fixed a fuzz error. - -Sun Oct 20 00:14:32 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * examples/C++NPv1/Makefile: Reverted the change "Fri Oct 18 - 21:01:25 UTC 2002 Craig Rodrigues <crodrigu@bbn.com>" since - this opens up some problems. Craig will look into it later and - provide a clean fix. Till then.. - -Sat Oct 19 20:00:00 2002 Stephen Torri <storri@cse.wustl.edu> - - * tests/Recursive_Condition_Test.cpp: - - Added test to check nesting_level after requiring a Recursive - Condition Mutex multiple times. Also after releasing for the - same number of times. - - * tests/SString_Test.cpp: - * ace/String_Base.i: - - Bug 1323 (Reported by Wayne Erchak <wayne@stentor.com>): Fixed. - In the effort to protect from memory problems with non-null - terminated strings the operators '<' and '>' where - rendered useless. Now the comparions use memcmp instead of - strncmp. A series of tests have been added to SString_Test.cpp - to ensure this does not happen again. - -Sat Oct 19 21:58:00 2002 Krishnakumar B <kitty@cse.wustl.edu> - - * ace/Argv_Type_Converter.cpp: - * ace/Argv_Type_Converter.h: - * ace/Argv_Type_Converter.inl: - * ace/Codecs.cpp: - * ace/Codecs.h: - * ace/OS.cpp: - * ace/Vector_T.cpp: - * ace/Vector_T.h: - * ace/WFMO_Reactor.cpp: - * tests/Proactor_Scatter_Gather_Test.cpp: Convert bool to use int. - - * ace/SString.cpp: Removed a redundant initialization to 0. - - * tests/Makefile: Add DIRS_PARALLEL=1 to allow building in - parallel in $ACE_ROOT/tests. - -Sat Oct 19 11:40:22 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.h: Added a ptrdiff_t typedef for ACE_HAS_WINCE. eVC3, - PocketPC2002 don't define this type. - - * ace/OS.i (recursive_mutex_cond_unlock, recursive_mutex_cond_relock): - PocketPC 2002 doesn't have the RecursionCount member in the - CRITICAL_SECTION. - - * ace/OS.cpp: ACE-ified the ACE_CE_ARGV class. - - * ace/OS_Errno.cpp: Add #include "ace/OS_Memory.h" to get ACE_NEW. - - * ace/ace_os_dll.vcp: Added files Time_Value.{h cpp inl}. Changed - preprocessor file locations from $ACE_ROOT to ".." to allow - working with more than one ACE version/build. - - * ace/ace_dll.vcp: Added files DLL_Manager.{h cpp}. - Changed preprocessor file locations from $ACE_ROOT to ".." to - allow working with more than one ACE version/build. - - * ace/config-win32-common.h: Set ACE_HAS_WIN32_TRYLOCK for WinCE - in addition to ACE_HAS_WINNT4. Also ACE_HAS_INTERLOCKED_EXCHANGEADD, - but only for PPC. - - * ace/Atomic_Op.h: Only set ACE_HAS_BUILTIN_ATOMIC_OP without - ACE_HAS_INTERLOCKED_EXCHANGEADD for MSVC on x86. It's not available - for eVC, CE on other than PPC. - - * tests/ARGV_Test_WinCE.vcp - * tests/Atomic_Op_Test_WinCE.vcp - * tests/Auto_IncDec_Test_WinCE.vcp - * tests/Barrier_Test_WinCE.vcp - * tests/Basic_Types_Test_WinCE.vcp - * tests/Bound_Ptr_Test_WinCE.vcp - * tests/Buffer_Stream_Test_WinCE.vcp - * tests/Cached_Accept_Conn_Test_WinCE.vcp - * tests/Cached_Conn_Test_WinCE.vcp - * tests/Cache_Map_Manager_Test_WinCE.vcp - * tests/Capabilities_Test_WinCE.vcp - * tests/CDR_Array_Test_WinCE.vcp - * tests/CDR_File_Test_WinCE.vcp - * tests/CDR_Test_WinCE.vcp - * tests/Collection_Test_WinCE.vcp - * tests/Conn_Test_WinCE.vcp - * tests/DLList_Test_WinCE.vcp - * tests/DLL_Test_DLL_WinCE.vcp - * tests/DLL_Test_WinCE.vcp - * tests/Dynamic_Priority_Test_WinCE.vcp - * tests/Enum_Interfaces_Test_WinCE.vcp - * tests/Future_Set_Test_WinCE.vcp - * tests/Future_Test_WinCE.vcp - * tests/Handle_Set_Test_WinCE.vcp - * tests/Hash_Map_Bucket_Iterator_Test_WinCE.vcp - * tests/Hash_Map_Manager_Test_WinCE.vcp - * tests/High_Res_Timer_Test_WinCE.vcpL - * tests/Lazy_Map_Manager_Test_WinCE.vcp - * tests/Logging_Strategy_Test_WinCE.vcp - * tests/Log_Msg_Test_WinCE.vcp - * tests/Malloc_Test_WinCE.vcp - * tests/Map_Manager_Test_WinCE.vcp - * tests/Map_Test_WinCE.vcp - * tests/Max_Default_Port_Test_WinCE.vcp - * tests/Mem_Map_Test_WinCE.vcp - * tests/MEM_Stream_Test_WinCE.vcp - * tests/Message_Block_Test_WinCE.vcp - * tests/Message_Queue_Notifications_Test_WinCE.vcp - * tests/Message_Queue_Test_Ex_WinCE.vcp - * tests/Message_Queue_Test_WinCE.vcp - * tests/MM_Shared_Memory_Test_WinCE.vcp - * tests/MT_Reactor_Timer_Test_WinCE.vcp - * tests/MT_Reactor_Upcall_Test_WinCE.vcp - * tests/MT_SOCK_Test_WinCE.vcp - * tests/Naming_Test_WinCE.vcp - * tests/New_Fail_Test_WinCE.vcp - * tests/Notify_Performance_Test_WinCE.vcp - * tests/Object_Manager_Test_WinCE.vcp - * tests/OrdMultiSet_Test_WinCE.vcp - * tests/OS_Test_WinCE.vcp - * tests/Priority_Buffer_Test_WinCE.vcp - * tests/Priority_Reactor_Test_WinCE.vcp - * tests/Priority_Task_Test_WinCE.vcp - * tests/Process_Manager_Test_WinCE.vcp - * tests/RB_Tree_Test_WinCE.vcp - * tests/Reactors_Test_WinCE.vcp - * tests/Reactor_Exceptions_Test_WinCE.vcp - * tests/Reactor_Notify_Test_WinCE.vcp - * tests/Reactor_Performance_Test_WinCE.vcp - * tests/Reactor_Timer_Test_WinCE.vcp - * tests/Reader_Writer_Test_WinCE.vcp - * tests/Recursive_Mutex_Test_WinCE.vcp - * tests/Refcounted_Auto_Ptr_Test_WinCE.vcp - * tests/Reverse_Lock_Test_WinCE.vcp - * tests/Semaphore_Test_WinCE.vcp - * tests/Service_Config_DLL_WinCE.vcp - * tests/Service_Config_Test_WinCE.vcp - * tests/Sigset_Ops_Test_WinCE.vcp - * tests/Simple_Message_Block_Test_WinCE.vcp - * tests/SOCK_Connector_Test_WinCE.vcp - * tests/SOCK_Send_Recv_Test_WinCE.vcp - * tests/SOCK_Test_WinCE.vcp - * tests/SPIPE_Test_WinCE.vcp - * tests/SString_Test_WinCE.vcpS - * tests/Svc_Handler_Test_WinCE.vcp - * tests/Task_Test_WinCE.vcp - * tests/Thread_Manager_Test_WinCE.vcp - * tests/Thread_Mutex_Test_WinCE.vcp - * tests/Thread_Pool_Reactor_Resume_Test_WinCE.vcp - * tests/Thread_Pool_Reactor_Test_WinCE.vcp - * tests/Thread_Pool_Test_WinCE.vcp - * tests/Timeprobe_Test_WinCE.vcp - * tests/Timer_Queue_Test_WinCE.vcp - * tests/Time_Service_Test_WinCE.vcp - * tests/Time_Value_Test_WinCE.vcp - * tests/Tokens_Test_WinCE.vcp - * tests/TSS_Test_WinCE.vcp - * tests/Upgradable_RW_Test_WinCE.vcp - Changed include file path from $ACE_ROOT to .. to allow building - multiple versions. Added aced.lib to CE ARM Debug builds. - -Sat Oct 19 16:57:44 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * etc/tao_*.doxygen: - Disabled the doxygen warnings. The warnings about not documented - members are still on. The warnings are not disabled in the ace - libraries because we could fix them there, but we can't fix them - in TAO because of the environment macro's that cause the method - definition arguments to be different than the implementation - arguments. - -Sat Oct 19 16:47:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/ace_wchar.h: - Added ACE_TEXT_WaitNamedPipe to the list of Windows functions we use - and for which there is an ascii and unicode version. - - * ace/SPIPE_Connector.cpp: - Used ACE_TEXT_WaitNamedPipe instead of ::WaitNamedPipe. This fixes - the problems in the BCB Unicode build. - -Sat Oct 19 10:09:38 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * examples/C++NPv2/Makefile: Reverted the change "Fri Oct 18 - 21:01:25 UTC 2002 Craig Rodrigues <crodrigu@bbn.com>" since - this opens up some problems. Craig will look into it later and - provide a clean fix. Till then.. - -Sat Oct 19 09:25:52 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * examples/Reactor/Misc/test_timer_queue.cpp, - * tests/Timer_Queue_Test.cpp: Fixed warnings with G++. - Thanks to Bala for reporting this. - -Fri Oct 18 22:07:31 2002 Steve Huston <shuston@riverace.com> - - * ace/config-aix-4.x.h: Only define ACE_HAS_ALLOCA for IBM - compilers. They do it via "builtin" trickery and g++ doesn't - know that trick. - - * include/makeinclude/platform_aix4_g++.GNU: Set with_ld to aix - to avoid trying to do an soname - AIX's ld doesn't do that. - - * include/makeinclude/platform_g++_common.GNU: If with_ld is aix, - don't try to add to soname option to the linker. - -Fri Oct 18 20:13:29 2002 Steve Huston <shuston@riverace.com> - - * include/makeinclude/platform_hpux_gcc.GNU: Added an optional - build parameter, with_ld, whose default value is hpux. This - says which ld is used with g++. - - * include/makeinclude/platform_g++_common.GNU: When setting the - SOFLAGS for setting SONAME, if with_ld is hpux, use +h, not -h. - -h means "hide symbol" in HP-UX ld. - -Fri Oct 18 17:12:03 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * examples/C++NPv2/display_logfile.cpp: - * examples/C++NPv2/Client_Logging_Daemon.cpp: - * examples/C++NPv2/TP_Logging_Server.h: - * examples/C++NPv2/TP_Logging_Server.cpp: Fixed compile errors in - single threaded builds. All they needed was the usage of the - right macros like ACE_SYNCH instead of ACE_MT_SYNCH and - ACE_SYNCH_MUTEX instead of ACE_Thread_Mutex. - -Fri Oct 18 16:25:09 2002 Yuanfang Zhang <yfzhang@cse.wustl.edu> - - * tests/RB_Tree_Test.h: - * tests/RB_Tree_Test.cpp: - - add a new method to test forward and reverse partial tree traversal. - -Fri Oct 18 21:01:25 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * examples/C++NPv1/Makefile: Fix comment. - * examples/C++NPv1/Makefile: Replace with recursive Makefile which - does not use for loops. - -Fri Oct 18 15:44:19 2002 Priyanka Gontla <pgontla@ociweb.com> - - * THANKS: - Added Christian Barheine <cbarheine@sympatec.com> to the hall - of fame. - -Fri Oct 18 19:37:20 GMT 2002 Craig Rodrigues <crodrigu@bbn.com> - - * ace/config-macosx.h: Define ACE_HAS_NONSTATIC_OBJECT_MANAGER - for gperf. - -Fri Oct 18 14:52:37 2002 Steve Huston <shuston@riverace.com> - - * examples/C++NPv2/Reactor_Logging_Server.cpp: - * examples/C++NPv2/Select_Reactor_Logging_Server.cpp: - * examples/C++NPv2/TP_Reactor_Logging_Server.cpp: - * examples/C++NPv2/display_logfile.cpp: Added necessary explicit - template instantiations. - -Fri Oct 18 13:21:05 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * tests/Timer_Queue_Test.cpp, - * examples/Reactor/Misc/test_timer_queue.cpp: - Fixed these tests so that it passes dynamically allocated memory - rather that values via the "act" parameter to - ACE_Timer_Queue::schedule(). This ensures the test will work - properly on strange OS platforms, such as Win64 where pointers - and longs/ints are different sizes. Thanks to Cristian - Ungureanu <cristian.ungureanu@amdocs.com> for motivating this. - -Fri Oct 18 17:20:02 UTC 2002 Don Hinton <dhinton@ieee.org> - - * ace/SOCK_Dgram_Mcast.{h,cpp}: - - Factored out common code to an open_i() method that open() - can call at this level and in inherited classes. - - * ace/QoS/SOCK_Dgram_Mcast_QoS.{h.cpp}: - - Modified ctor to take same default param as base class and - invoke the base class ctor. - - Removed duplicate code in open() and have it call base::open_i() - after opening the socket with QoS. Changed 1st parameter from - ACE_Addr to ACE_INET_Addr to match base classes. - - Removed duplicate code from subscribe by having it call base - class methods. - -Fri Oct 18 11:51:44 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * bin/PerlACE/Run_Test.pm: Created a new variable - $PerlACE::svcconf_ext which defines the canonical - svc.conf file extention used by various run_test scripts. By - default, this is defined to ".conf". It can be overwritten - using the environment variable ACE_RUNTEST_SVCCONF_EXT. For - example, you can define ACE_RUNTEST_SVCCONF_EXT to ".conf.xml" - if you have built your ACE with XML-based service configurator. - -Fri Oct 18 11:55:33 2002 Steve Huston <shuston@riverace.com> - - * ace/POSIX_CB_Proactor.cpp: Removed unneeded argument to - application_specific_code(). Thanks to Tommy Persson - <tpe@ida.liu.se> for contributing this fix. - - Fixes Bugzilla #1325. - - * THANKS: Added Tommy Persson to the Hall of Fame. - -Fri Oct 18 11:39:29 2002 Steve Huston <shuston@riverace.com> - - * examples/C++NPv1/Process_Per_Connection_Logging_Server.cpp: - * examples/C++NPv1/RT_Thread_Per_Connection_Logging_Server.cpp: - * examples/C++NPv1/Thread_Per_Connection_Logging_Server.cpp: - Put the sigterm_handler function in an unnamed namespace - to avoid issues with g++ not liking 'static' and 'extern "C"' - together. - -Fri Oct 18 03:55:51 UTC 2002 Don Hinton <dhinton@objectsciences.com> - Bill Fulton <Bill_A_Fulton@raytheon.com> - - * ace/SOCK_Dgram_Mcast.{h,i,cpp}: - - Enhanced support for join()'ing multiple multicast groups. Ctor - now offers finer control over a bind()'ing to an address or - interface. Appropriate default behavior has been specified, - but note that due to issues on Linux, portable code should not - join() more than one multicast group per instance. - - An open() method has been added/exposed. If not explicitely - called, it will be implicitely called by the first join(), and - will bind() the address if the OPT_BINDADDR_YES options was - specified in the ctor. - - Error handling has also been enhanced. For instance, if - OPT_BINDADDR_YES has been specified (the default on Linux), - additional join()'s will fail since the socket gets bound on - the first call to open() or join(). Previously, subsequent - calls to join() would silently fail, i.e., the call succeeded - but only datagrams sent to the first, bound address where - actually delivered. - - If ACE_SOCK_DGRAM_MCAST_DUMPABLE is defined, a list of all - subscribed groups is maintained and can be dump()'ed, although - access to this list is not public. - - Added new join()/leave() methods that remove unused, deprecated - protocol_family and protocol arguments. These values can be - determined on the fly. The old subscribe()/unsubscribe() - methods have been deprecated and are now implemented in terms - of join()/leave() respectively. - - The unsubscribe(void) method has also been deprecated. It was - designed to allow the caller to unsubscribe/leave() all - groups at once, but since a list was never kept (and still - isn't unless ACE_SOCK_DGRAM_MCAST_DUMPABLE is defined) only - the last group join()'ed was ever left. A better way would - be to either keep track of you own groups and call leave() - for each, or let the dtor close() the socket which automatically - unsubscribes from all groups. - - Some IPv6 support has been added either directly or by removing - the need for the caller to handle it (see join()/leave() above), - although the internal calls to setsockopt() are still hard wired - for IPv4. - -Fri Oct 18 06:51:07 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * ace/String_Base.{i,cpp}: Fixed the formatting in these files... - -Thu Oct 17 20:02:39 2002 Steve Huston <shuston@riverace.com> - - * examples/C++NPv1/Process_Per_Connection_Logging_Server.cpp: - * examples/C++NPv1/RT_Thread_Per_Connection_Logging_Server.cpp: - * examples/C++NPv1/Thread_Per_Connection_Logging_Server.cpp: - Add extern "C" to the signal handler function declaration to conform - strictly to its definition. - When explicitly instantiating auto_ptr<>, fully specify the - Thread_Args type else modern compilers (Forte 7) can't see it. - -Thu Oct 17 17:32:39 2002 Krishnakumar B <kitty@cse.wustl.edu> - - * ACEXML/examples/svcconf/Svcconf_Handler.cpp: - * ACEXML/examples/svcconf/Svcconf_Handler.h: - * ACEXML/examples/svcconf/main.cpp: - - Fixed compilation errors under Windows. - -Thu Oct 17 19:18:02 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * ace/config-macosx.h: - - Remove define of __MACOSX__, and use constant __APPLE__ - instead defined by Apple's gcc. - - Removed definitions: - ACE_HAS_MACOSX_DYLIB, ACE_LACKS_SEMBUF_T - - Added definitions: - ACE_HAS_SEMUN, ACE_HAS_SIGINFO_T, ACE_LACKS_SIGINFO_H, - ACE_HAS_UCONTEXT_T, ACE_HAS_GETIFADDRS, - ACE_LACKS_STATIC_DATA_MEMBER_TEMPLATES - - Defined ACE_HAS_POSITION_INDEPENDENT_POINTERS to 0 for temporary - Apple gcc bug workaround. - - * ace/Log_Msg.cpp (ACE_Log_Msg::log): Add check for __APPLE__ to catch - right prototype for sys_errlist. - - * ace/OS.h: Remove unnecessary __MACOSX__ check. - - * include/makeinclude/platform_macosx.GNU: Minor cleanups. - -Thu Oct 17 13:47:39 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.h: Added typedef for ACE_recursive_mutex_state for - !defined (ACE_HAS_THREADS) section. - -Thu Oct 17 16:28:05 UTC 2002 Don Hinton <dhinton@ieee.org> - - * THANKS: Added Jon Franklin <jfranklin@dctd.saic.com> - to the hall of fame. - -Thu Oct 17 10:26:59 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * examples/C++NPv2/display_logfile.cpp (ACE_MT_SYNCH>): Fixed a - uninitialized member usage warning. - -Thu Oct 17 00:26:13 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/Malloc_T.i (release): Instead of releasing the memory_pool_ - explicitly, changed to use remove () to make sure all resources - are removed properly. - -Wed Oct 16 20:23:09 2002 Krishnakumar B <kitty@cse.wustl.edu> - - * ACEXML/common/FileCharStream.cpp (getchar_i): - * ACEXML/common/FileCharStream.h: - - Borland defines getchar() as a macro. Rename it as getchar_i to - resolve the clash and fix compilation errors in the build. - - -Wed Oct 16 18:47:44 2002 Stephen Torri <storri@cse.wustl.edu> - - * examples/C++NPv1/Makefile: - - Replaced Makefile with non-for-loop version provided by - Craig Rodrigues. - -Wed Oct 16 16:52:52 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/MEM_SAP.cpp (close_shm_malloc): Fixed a memory leak that I - added last time when I fixed a race condition in removing mmap - file. - -Wed Oct 16 15:35:22 2002 Steve Huston <shuston@riverace.com> - - * tests/SPIPE_Test.cpp: Fixed compile error. - -Wed Oct 16 14:02:27 2002 Yuanfang Zhang <yfzhang@cse.wustl.edu> - - * ace/RB_Tree.cpp: - * ace/RB_Tree.h: - - forward and reverse partial tree traversal. - Changed the find_i method by adding a parameter "find_Exact". - -Wed Oct 16 12:18:03 2002 Stephen Torri <storri@cse.wustl.edu> - - * examples/C++v1/Makefile: - * examples/C++NPv1/Makefile.Iterative_Logging_Server: - * examples/C++NPv1/Makefile.Logging_Client: - * examples/C++NPv1/Makefile.Reactive_Logging_Server: - * examples/C++NPv1/Makefile.Reactive_Logging_Server_Ex: - * examples/C++NPv1/Makefile.RT_Thread_Per_Connection_Logging_Server: - * examples/C++NPv1/Makefile.Thread_Per_Connection_Logging_Server: - * examples/C++NPv1/Makefile.wProcess_Per_Connection_Logging_Server: - - Renamed from *.mak to Makefile.* to avoid looking like an - exported MSVC project NMAKE file. (Following Steve Huston's - change to examples/C++v2). - - Fixed Make error that make realclean was not in Makefile. - - * tests/Recursive_Condition_Test.cpp: - - Changed function call of signal() to broadcast() to awaken - multiple threads waiting on a shared condition variable. - -Oct 16 11:44:27 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/Thread_Manager.cpp (check_state): Changed to return 0 if - the thread id is not managed by the Thread_Manager. - - * ace/Thread_Manager.h: Updated documentation for testcancel(), - testsuspend(), and testresume (). They all return false now if - the you try to test the status of a tid that is not managed by - the Thread_Manager. - - Thanks to Giovanni Zito <g.zito@acsys.it> for reporting it. - -Wed Oct 16 12:41:29 2002 Steve Huston <shuston@riverace.com> - - * examples/Makefile: Add C++NPv1 and C++NPv2 to the DIRS list. - - * examples/C++NPv2/Makefile: Only build the SSL-using examples - (AC_CLD, AIO_CLD, TPCLS) is ssl=1 is set. - - * examples/C++NPv2/Makefile.AC_CLD: - * examples/C++NPv2/Makefile.AIO_CLD: - * examples/C++NPv2/Makefile.TPCLS: - Set up the SSL build params correctly and warn if ssl=1 not used. - -Wed Oct 16 12:08:21 2002 Steve Huston <shuston@riverace.com> - - * examples/C++NPv2/AC_CLD.mak: - * examples/C++NPv2/AIO_CLD.mak - * examples/C++NPv2/CLD.mak: - * examples/C++NPv2/Configurable_Logging_Server.mak: - * examples/C++NPv2/display_logfile.mak: - * examples/C++NPv2/Reactor_Logging_Server.mak: - * examples/C++NPv2/Select_Reactor_Logging_Server.mak: - * examples/C++NPv2/SLDex.mak: - * examples/C++NPv2/SLD.mak: - * examples/C++NPv2/SR_Configurable_Logging_Server.mak: - * examples/C++NPv2/TPCLS.mak: - * examples/C++NPv2/TPLS.mak: - * examples/C++NPv2/TP_Reactor_Logging_Server.mak: - Renamed from *.mak to Makefile.* to avoid looking like an - exported MSVC project NMAKE file. - - * examples/C++NPv2/Makefile: Changed to process new Makefile names. - -Wed Oct 16 11:02:41 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * ace/High_Res_Timer.cpp (get_cpuinfo): Removed the recalibration of - the scale factor every time the application is launched. Thanks - to Bo Balder <bba@post.utfors.se> for reporting this. - -Wed Oct 16 10:55:11 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/config-linux-common.h: Added support for ACE_CC_NAME, - ACE_CC_MAJOR_VERSION and ACE_CC_MINOR_VERSION for linux. Thanks - to Jaroslaw Nozderko <jaroslaw.nozderko@polkomtel.com.pl> for - suggesting this. - -Wed Oct 16 10:55:44 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/config-win32-msvc.h (ACE_HAS_ALLOCA): Enabled alloca support - for MSVC. Thanks to Pat McNertney <pat@mcnerthney.com> for - providing the patch. - -Tue Oct 15 18:49:25 2002 Stephen Torri <storri@cse.wustl.edu> - - * tests/Recursive_Condition_Test.cpp: - - Added two tests. - Test 2: repeat of test 1 but with a 2 second delay - after main thread signals the waiting thread. (Works) - Test 3: repeat of test 1 but with 4 threads waiting - on the condition variable. (Broken - no threads - awaken). - -Tue Oct 15 17:27:34 2002 Krishnakumar B <kitty@cse.wustl.edu> - - * ACEXML/docs/bugs.txt: - * ACEXML/docs/TODO.txt: - - Updated to reflect the current status. - -Tue Oct 15 17:23:43 2002 Krishnakumar B <kitty@cse.wustl.edu> - - * tests/Service_Config_Test.UTF-16.conf.xml: New file added which - is used in Service_Config_Test when ACE_HAS_XML_SVC_CONF and - ACE_USES_WCHAR is enabled. - -Tue Oct 15 17:17:44 2002 Krishnakumar B <kitty@cse.wustl.edu> - - * ACEXML/common/Encoding.cpp: - * ACEXML/common/Encoding.h: - - New files which handle the automatic detection of encoding in - XML files. - - * ACEXML/apps/svcconf/Makefile: - * ACEXML/common/CharStream.h: - * ACEXML/common/Exception.cpp: - * ACEXML/common/FileCharStream.cpp: - * ACEXML/common/FileCharStream.h: - * ACEXML/common/HttpCharStream.cpp: - * ACEXML/common/HttpCharStream.h: - * ACEXML/common/InputSource.cpp: - * ACEXML/common/Makefile: - * ACEXML/common/Mem_Map_Stream.cpp: - * ACEXML/common/Mem_Map_Stream.h: - * ACEXML/common/StrCharStream.cpp: - * ACEXML/common/StrCharStream.h: - * ACEXML/common/XML_Common.dsp: - * ACEXML/examples/SAXPrint/Makefile: - * ACEXML/parser/parser/Makefile: - * ACEXML/parser/parser/Parser.cpp: - - Handle the automatic detection of encoding in XML files. - -Tue Oct 15 17:15:41 2002 Steve Huston <shuston@riverace.com> - - Fixes for Bugzilla #1319: - - * ace/OS.i (ctime_r): For the 2-param ctime_r platforms, restrict - buflen to at least 26, just as 3-param platforms do. If less than - 26, return 0 and errno = ERANGE. - - * ace/OS.cpp (ctime_r): The WinCE implementation of ctime_r already - had the 26-char minimum check, but didn't set errno to ERANGE. - Now it does. - - * tests/OS_Test.cpp (ctime_r_test): Make sure that ctime_r with length - less than 26 fails and doesn't scribble on the buffer. - -Tue Oct 15 11:43:27 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/SPIPE_Connector.cpp (connect): Convert the first argument in - the ::WaitNamedPipe () call to char * always using the macro - defined. This shold fix compilation errors in WCHAR builds. - -Tue Oct 15 16:19:33 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - [Bug 1090] - * ace/config-freebsd-pthread.h: Add ACE_HAS_GETIFADDRS. This fixes - Bugzilla bug 1090. - -Tue Oct 15 10:47:20 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/SPIPE_Connector.cpp: Fixed a unused variable warning. - -Tue Oct 15 06:56:18 2002 Patrick J. McNerthney <ace@mcnerthney.com> - - * tests/SPIPE_Test.cpp: Added support for testing the new Windows - bytestream mode. - -Mon Oct 14 20:47:21 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * ace/SPIPE_Acceptor.cpp (open): Changed "mode" to "pipe_mode", - which is the name of the parameter. Thanks to Don Hinton for - reporting this. - -Mon Oct 14 19:20:35 2002 Frank Hunleth <frank@hunleth.com> - - * ACE-INSTALL.html: Added installation notes for the OpenBSD port. - - * ace/config-openbsd.h: - * ace/config-openbsd-pthread.h: - * include/makeinclude/platform_openbsd.GNU: - * include/makeinclude/platform_openbsd_pthread.GNU: Configuration - files for OpenBSD. - - * ace/ACE.cpp: Added OpenBSD-specific error code to ACE::out_of_handles - - * ace/OS.h: Added #define for OpenBSD-specific path to siginfo.h and - ACE_LACKS_CLEARERR guards around clearerr() for OpenBSD. - - * ace/OS.i: Added support for getifaddrs() with ACE_HAS_GETIFADDRS - and fix for segfault when calling dlsym with a null handle. Added - guards on clearerr using ACE_LACKS_CLEARERR. (All only affect - OpenBSD.) - - * ace/POSIX_Proactor.cpp: - * ace/POSIX_Proactor.h: Added SystemType family for OpenBSD. - - * ace/README: Added new ACE #defines to support OpenBSD. These include: - ACE_HAS_GETIFADDRS - for support of the getifaddrs() system call - ACE_HAS_DLSYM_SEGFAULT_ON_INVALID_HANDLE - add check to dlsym call - to protect against NULL. OpenBSD 3.1 needs this. - ACE_LACKS_CLEARERR - #define to remove clearerr definition due - to bogus syntax error on OpenBSD. - - * ace/Sock_Connect.cpp: Added code to get_ip_interfaces to use - getifaddrs() if it is supported. OpenBSD is the only platform to - use this currently. Under OpenBSD, the code to parse the return - values from the corresponding ioctl calls gets confused when IPv6 - interfaces are present. - - * tests/Malloc_Test.cpp: Added OpenBSD to the list of platforms that - can't handle the remap of fault part of the test. - - * tests/Reactor_Dispatch_Order_Test.cpp: Added code for OpenBSD to - clear the write mask after handle_output is called, since OpenBSD - calls handle_output more than once before the data is received - over the pipe. - -Mon Oct 14 18:21:40 2002 Patrick J. McNerthney <ace@mcnerthney.com> - - * ace/OS.h: Added FILE_FLAG_WRITE_THROUGH defines for - non-supporting operating systems and added various PIPE_* - defines for non-supporting operating systems. - - * ace/SPIPE_Connector.{h,cpp}: Added support to enable the - selection of whether bytestream or message mode is used on - Windows. Also added support for both blocking and timeout - connections on Windows, only non-blocking connections where - support before. - - * ace/SPIPE_Acceptor.{h.cpp}: Added support to enable the - selection of whether bytestream or message mode is used on - Windows. - -Mon Oct 14 20:19:51 2002 Steve Huston <shuston@riverace.com> - - * ChangeLog: Removed a large chunk of duplicated dates from Oct 11 - back to July. - - * ace/OS.{h i}: Removed the owning_thread_ member from - ACE_recursive_mutex_state on Windows. This isn't needed since the - OS mutex restores the owner properly - it was in to do a sanity - check ACE_ASSERT, but can't do that from ACE_OS. - - * ace/WFMO_Reactor.h: Expanded and clarified the class comments - to resolve Bugzilla #772. - -Mon Oct 14 17:23:43 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/CDR_Base.cpp (consolidate): Reverted the optimization - performed in "Tue Mar 12 17:43:32 2002 Balachandran Natarajan - <bala@cs.wustl.edu>". This optimizations created problems for - CORBA::Any implementation which assumes that the data is totally - contained within the Any. Using the optimization created problems - for Any's that are copied between different threads. Thanks to - Jian Chen <jchen@huawei.com> for reporting this problem. - -Mon Oct 14 10:12:31 2002 Steve Huston <shuston@riverace.com> - - Fix for Bugzilla # 1316: - - * ace/Synch.h: - * ace/OS.h: Moved the ACE_recursive_mutex_state from Synch.h - to OS.h. The recursive mutex methods need to handle this stuff - internally to properly synchronize releasing and reacquiring - the mutex around condition variable waits. - - * ace/OS.i: Added ACE_OS::recursive_mutex_cond_unlock() and - ACE_OS::recursive_mutex_cond_relock() methods to support - recursive mutex handling around waiting on a condition variable. - - * ace/Synch.cpp (ACE_Condition<ACE_Recursive_Thread_Mutex>): - Removed the guards from the broadcast() and signal() methods. - Removed the ACE_recursive_mutex_state class methods - this - functionality is now in the ACE_OS space; see above. - Reimplemented the wait() method to make use of the new ACE_OS - methods (above), enabling proper reclaiming of the recursive - mutex after cond_wait, and avoiding deadlock. - -Mon Oct 14 00:12:48 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ACEXML/docs/bugs.txt: Updated to match the current status of - ACEXML parser. - -Sun Oct 13 19:38:35 2002 Krishnakumar B <kitty@cse.wustl.edu> - - * ace/Exception_Macros.h: - - Consolidated the checking of exceptions preprocessor hacking - into this file. - - * ACEXML/common/XML_Macros.h: - - Include the above file and get rid of redundant and broken - checking for exceptions. - - * ace/CORBA_macros.h: - - Moved over the stuff from Exception_Macros.h to where it was - originally. This should fix all the build breakage. Thanks to - Bala for pointing out the errors in the build. - -Sun Oct 13 18:50:53 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * ace/SPIPE_Stream.h: Enhanced the documentation of ACE_SPIPE_Stream. - Thanks to Pat McNerthney <ace@mcnerthney.com> for motivating - this. - -Sun Oct 13 15:57:04 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/Asynch_IO.cpp: - * ace/Asynch_IO.h: Protected ACE_Asynch_Read_File::readv () and - ACE_Asynch_Write_File::writev () from being called when - ACE_HAS_WINNT4 == 0. - * ace/WIN32_Asynch_IO.cpp: - * ace/WIN32_Asynch_IO.h: Protected - ACE_WIN32_Asynch_Read_File::readv () and - ACE_WIN32_Asynch_Write_File::writev () from being called when - ACE_HAS_WINNT4 == 0. - - Thanks to Eric Frias <efrias@syncad.com> for reporting the - problem and Alex Libman for suggesting the fix. - - * THANKS: Added Eric Frias <efrias@syncad.com> to the hall of - fame. - -Sun Oct 13 15:54:59 2002 Krishnakumar B <kitty@cse.wustl.edu> - - * ACEXML/apps/svcconf/Svcconf_Handler.cpp: Fixed a couple of - warnings related to the exception argument handling. - -Sun Oct 13 00:55:27 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ACEXML/common/XML_Common.dsp: - - Added XML_Macros.h to the project file. - - * ACEXML/examples/SAXPrint/svc.conf.xml: - - Reverted a mistaken change used for testing. - - * ACEXML/apps/svcconf/Svcconf_Handler.cpp: - * ACEXML/common/SAXExceptions.cpp: - * ACEXML/parser/debug_validator/Debug_DTD_Manager.cpp: - - Added ACE_TEXT around strings so that WCHAR build is happy. - -Sat Oct 12 23:01:29 2002 Krishnakumar B <kitty@cse.wustl.edu> - - * ACEXML/apps/svcconf/Svcconf.h (ACEXML_Svcconf_Parser): Removed - the unused member env_. - -Sat Oct 12 22:50:02 2002 Krishnakumar B <kitty@cse.wustl.edu> - - * ACEXML/common/XML_Macros.h: - - New file which contains the macros needed to enable support for - Exceptions in ACEXML. - - * ACEXML/apps/svcconf/Makefile: - * ACEXML/common/Makefile: - * ACEXML/examples/SAXPrint/Makefile: - * ACEXML/parser/parser/Makefile: - * ACEXML/tests/Makefile: - - Regenerate dependencies. - - * ACEXML/common/Env.cpp: - * ACEXML/common/Env.h: - * ACEXML/common/Env.i: - * ACEXML/common/Exception.cpp: - * ACEXML/common/Exception.h: - * ACEXML/common/Exception.i: - - Enable support for Native Exceptions. - - * ACEXML/apps/svcconf/Svcconf.cpp: - * ACEXML/apps/svcconf/Svcconf.h: - * ACEXML/apps/svcconf/Svcconf_Handler.cpp: - * ACEXML/apps/svcconf/Svcconf_Handler.h: - * ACEXML/common/Attributes_Def_Builder.h: - * ACEXML/common/ContentHandler.h: - * ACEXML/common/DTDHandler.h: - * ACEXML/common/DTD_Manager.h: - * ACEXML/common/DefaultHandler.cpp: - * ACEXML/common/DefaultHandler.h: - * ACEXML/common/Element_Def_Builder.h: - * ACEXML/common/EntityResolver.h: - * ACEXML/common/ErrorHandler.h: - * ACEXML/common/HttpCharStream.cpp: - * ACEXML/common/SAXExceptions.cpp: - * ACEXML/common/SAXExceptions.h: - * ACEXML/common/SAXExceptions.i: - * ACEXML/common/Validator.h: - * ACEXML/common/XMLFilterImpl.cpp: - * ACEXML/common/XMLFilterImpl.h: - * ACEXML/common/XMLReader.h: - * ACEXML/examples/SAXPrint/Print_Handler.cpp: - * ACEXML/examples/SAXPrint/Print_Handler.h: - * ACEXML/examples/SAXPrint/SAXPrint_Handler.cpp: - * ACEXML/examples/SAXPrint/SAXPrint_Handler.h: - * ACEXML/examples/SAXPrint/main.cpp: - * ACEXML/examples/SAXPrint/svc.conf.xml: - * ACEXML/examples/svcconf/Svcconf_Handler.cpp: - * ACEXML/examples/svcconf/Svcconf_Handler.h: - * ACEXML/examples/svcconf/main.cpp: - * ACEXML/parser/debug_validator/Debug_Attributes_Builder.cpp: - * ACEXML/parser/debug_validator/Debug_Attributes_Builder.h: - * ACEXML/parser/debug_validator/Debug_DTD_Manager.cpp: - * ACEXML/parser/debug_validator/Debug_DTD_Manager.h: - * ACEXML/parser/debug_validator/Debug_Element_Builder.cpp: - * ACEXML/parser/debug_validator/Debug_Element_Builder.h: - - Switch over the files to support Native Exceptions. - - * ACEXML/parser/parser/Parser.cpp: - * ACEXML/parser/parser/Parser.h: - - Use the obstack unwind to unwind the stack and reuse memory. - Currently not done throughout the parser. That will come later. - - * ace/Makefile.ace: - - Added Codecs to ACE_COMPONENTS when building FOR_TAO. Thanks to - Don Hinton for suggesting this. - - * include/makeinclude/rules.local.GNU: - - Clean up the old symlinks under $ACE_ROOT/ace when building a - new version of the libraries. - - * tests/Makefile: - - Cleanup the Codecs_Test when $ACE_ROOT/ace has been realcleaned - previously. - -Fri Oct 11 12:44:58 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * RTJava/*: Removed. Angelo Corsaro <corsaro@ece.uci.edu> - keeps RTJperf related files in another CVS repository. - - * ace/config-freebsd-pthread.h: Improved check of - _POSIX_REALTIME_SIGNALS preprocessor constant. - -Thu Oct 10 16:02:20 2002 Ossama Othman <othman@cs.wustl.edu> - - * ACE version 5.2.5 released. - -Thu Oct 10 13:56:35 2002 Ossama Othman <ossama@uci.edu> - - * bin/make_release (check_workspace): - - Ignore platform_macros.GNU and config.h files. They are - needed when cutting the beta/release. This isn't a problem - since the ACE/TAO kits are created from a "cvs export"ed - workspace that doesn't include those files. - -Thu Oct 10 15:11:07 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * bin/make_release: A pattern matching issue. The cvs server now - gives a message starting with "cvs server:.." instead of "cvs - checkout:..". This created problems for Ossama when he was - trying to cut the beta. - -Wed Oct 9 23:43:51 2002 Steve Huston <shuston@riverace.com> - - * ace/WIN32_Asynch_IO.cpp (ACE_Asynch_Read_Stream::readv, - ACE_Asynch_Write_Stream::writev, ACE_Asynch_Read_Dgram::recv, - ACE_Asynch_Write_Dgram::send): Don't double-increment iovcnt - while checking for the need for multiple iov per message block. - Fixes bug introduced in: - Fri Oct 4 20:22:41 2002 Steve Huston <shuston@riverace.com> - -Wed Oct 9 23:08:19 2002 Steve Huston <shuston@riverace.com> - - * include/makeinclude/rules.lib.GNU: For the funky AIX shr.o/lib*.a - machination, declare shr.o as .INTERMEDIATE to avoid rebuilding it - all the time for nothing. Thanks to Craig L. Ching for motivating - this. - -Wed Oct 09 12:19:40 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * bin/PerlACE/Process_Win32.pm: Added hack to purify a process. - This will allow us to purify all the examples, tests much - easier. It is, however, tightly coupled to Purify 6.0. The - script automaticaly switch to purifying the process if the - ACE_RUN_PURIFY_CMD environment variable is defined. See the - comments in this file for more details. - -Tue Oct 8 21:06:23 2002 Steve Huston <shuston@riverace.com> - - * ace/config-sunos5.6.h: Undef ACE_HAS_TEMPLATE_TYPEDEFS for the - Forte 7 compiler (C++ 5.4). It should work, but Malloc_T doesn't - compile correctly with it enabled. Thanks to Serge Kolgan - <skolgan@objectsciences.com> for this fix. - - * ace/config-sunos5.9.h: New config.h file for Solaris 9. Right now, - it just includes config-sunos5.8.h, but Solaris 9 has some new things - to play with as time goes on. - -Tue Oct 8 09:22:41 2002 Steve Huston <shuston@riverace.com> - - * ace/SOCK_Dgram_Bcast.cpp (send): Use next_ to find next - ACE_Bcast_Node in if_list_ rather than incrementing pointer. - Thanks to <k2@biogate.com> for this fix. - -Mon Oct 7 17:29:14 2002 Ossama Othman <ossama@uci.edu> - - * tests/Proactor_Scatter_Gather_Test.cpp (initiate_read_file): - - Cast one of the operands in a subtraction operation to a size_t - to address a problem exhibited by Borland C++ where the ACE_MIN - template expects two parameters of the same type, but was - getting two parameter of different types. This problem occurred - on platforms where "unsigned int" and "unsigned long" are - different types. - -Mon Oct 07 15:21:44 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ACEXML/apps/svcconf/XML_Svc_Conf_Parser.dsp: - * ACEXML/common/XML_Common.dsp: - * ACEXML/parser/debug_validator/Debug_Validator.dsp: - * ACEXML/parser/parser/Parser.dsp: - * ACEXML/tests/HttpCharStream_Test.dsp: - * ACEXML/tests/NamespaceSupport_Test.dsp: - * ACEXML/tests/Transcoder_Test.dsp: - * apps/JAWS2/server.dsp: - * examples/Export/test.dsp: - * examples/IPC_SAP/FILE_SAP/client.dsp: - * examples/Logger/client/client.dsp: - * examples/Logger/simple-server/server_loggerd.dsp: - * examples/QOS/Simple/receiver.dsp: - * examples/QOS/Simple/sender.dsp: - * examples/Reactor/Multicast/client.dsp: - * examples/Reactor/Multicast/server.dsp: - * examples/Reactor/Proactor/test_udp_proactor.dsp: - * examples/Synch/proc_sema.dsp: - * tests/DLL_Test_Impl.dsp: - * tests/Framework_Component_DLL.dsp: Disabled the generation - of precompiled headers. They were taking up too much disk space - on the nightly build machines. Thanks to Ossama for noticing - this. - -Sun Oct 6 09:03:28 2002 Ossama Othman <ossama@uci.edu> - - * ace/UPIPE_Stream.cpp (send): - - Silenced a signed/unsigned conversion warning using a - static_cast<> until a better solution presents itself. - -Sat Oct 5 17:04:38 2002 Steve Huston <shuston@riverace.com> - - * ace/Memory_Pool.cpp (ACE_Pagefile_Memory_Pool::acquire): Remove - casts on comparisons, allowing size_t values to compare directly. - - * apps/JAWS/server/IO.cpp (JAWS_Synch_IO::transmit_file): Remove - cast from send_n() return value to avoid signed/unsigned compare. - -Sat Oct 5 12:30:49 2002 Steve Huston <shuston@riverace.com> - - * ace/Service_Manager.cpp (list_services): Remove extra right paren. - - * tests/Proactor_Timer_Test.cpp: Fixed signed/unsigned mismatch. - -Fri Oct 04 22:24:23 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ACEXML/common/URL_Addr.cpp (string_to_addr): Added a new wrapper - which matches the signature in ACE_INET_Addr, and delegates the - work to the ACEXML_Char function to work around Borland BCB - warning about hidden virtual function. - -Fri Oct 4 21:32:51 2002 Steve Huston <shuston@riverace.com> - - * ace/Service_Manager.cpp (list_services): Correct compiler - warning about sign/unsigned comparison. Don't bother for checking - transfer count != requested, since send_n always reports - an error (0 or -1) if there was a fail before the whole 'n' sent. - -Fri Oct 4 20:22:41 2002 Steve Huston <shuston@riverace.com> - - * ace/ace_dll64.mak: Removed the /wd4267 to re-enable the - conversion size_t to int warning. - Added Atomic_Op.obj to the files to compile directly. - - * ace/ACE.cpp (recv, send): Extract int, not size_t, from implicit - char */int pairs. - (send_n, recv_n with ACE_Message_Blocks): Message block length - is a size_t but iov_len is u_long, so make as many iovs as needed - to fit the entire requested length. - - * ace/Asynch_Acceptor.cpp: Added casts to fix size_t->DWORD warnings. - - * ace/Asynch_IO.cpp (transmit_buffers): Cast lengths to DWORD for - Win64. - - * ace/Configuration.{h cpp} (set_binary_value, get_binary_value): - Change length arg from u_int to size_t. - - * ace/Configuration_Import_Export.cpp: Align use of u_int, size_t. - - * ace/Containers_T.cpp (ACE_Fixed_Set::insert): Use ssize_t, not int, - to track free slot since max size is a size_t. - - * ace/FIFO_Recv_Msg.i (recv): - * ace/FIFO_Send_Msg.i (send): Squeeze size_t max_len down to an int - because this gets run through the ACE_Str_Buf variant of recv/send. - If this restriction is too much, the code will need to be - refactored. - - * ace/FILE_Addr.cpp (set): Cast size_t back to int for base_set(). - - * ace/FILE_IO.cpp (recv): Count iov tuples as an int to match what's - expected at the ACE_OS::readv level. - - * ace/Filecache.h (ACE_File_Object): Change size_ from size_t to off_t. - It's a better match for what it's used for (checking against - stat.st_size, which is an off_t also). - - * ace/Functor_T.i (ACE_Pointer_Hash): Add #pragmas for Win64 to - shut up the compiler warnings about casting a pointer to u_long. - - * ace/INET_Addr.{h i} (ip_addr_size): Return an int, not size_t. This - value is just passed on to other methods (like host db functions) - that are expecting an int. Also, being the size of a IP addressing - structure, all the values are well within range of an int. - - * ace/Lib_Find.cpp (ldfind, get_temp_dir): Cast path length value to - DWORD for calls to SearchPath and GetTempPath on Windows. - - * ace/Local_Name_Space.{h cpp} (ACE_NS_String::strstr): Cast return - value from size_t to int. (hash): Change return type from size_t - to u_long to match what's expected from ACE_Hash<>, which this is - an adapter for. - - * ace/Log_Msg_NT_Event_Log.cpp (open): Cast msg_file length to - DWORD for use in Windows API. Use size_t for reg key string length - since it need not be passed to a Windows API. - - * ace/Log_Record.cpp (round_up): Do all the rounding math in size_t - terms, since all factors are size_t, but cast it back to ACE_UINT32 - since the ACE_Log_Record length_ field is ACE_UINT32 and should - stay that way due to the need to send it over the network as a - known-length field. - (print): Cast verbose_msg string length to int so it can be compared - to fprintf return value. - - * ace/MEM_Acceptor.cpp (accept): Cast string length to ACE_UINT16. - - * ace/MEM_SAP.h: - * ace/MEM_IO.{h cpp} (ACE_Reactive_MEM_IO, ACE_MT_MEM_IO): Changed - return type of send_buf and recv_buf from int to ssize_t. This - is consistent with other ACE sock-type send methods, and matches - the range of sizes allowed to request sending of and the return - type expected by ACE_MEM_IO::send() and recv(). - - * ace/Mem_Map.i (advise): Get a size_t length as expected by - ACE_OS::madvise(). - - * ace/Mem_Map.cpp (map_it): Use off_t instead of size_t to calculate - null_byte_positition when extending the file. This is the type - that ACE_OS::pwrite() expects. - - * ace/Memory_Pool.h (ACE_Pagefile_Memory_Pool::Shared_Control_Block): - Changed mapped_size_ from int to size_t to match the sizes it can - take (max_size_)and what's expected in the Win32/64 API calls. - Changed free_offset_ from int to ptrdiff_t to represent the full - range of mapped address range. Changed free_size_ from int to - size_t to match max_size_ and the range of mappable sizes. - Changed 'append' arg to map() from int to size_t, also to represent - full range of mappable size. - - * ace/Memory_Pool.i (round_up): Cast nbytes to off_t for call to - ACE::round_to_page_size(). - - * ace/Memory_Pool.cpp: Adjust usage of size_t/off_t, usually with - casts where needed. - (map): For Win64, allow full 64-bit range of file mapping size. - - * ace/Message_Queue.{h i cpp} (ACE_Message_Queue_NT): Changed - max_threads arg (ctor and open()) and max_cthrs_/cur_thrs_ from - size_t to DWORD, as expected by the underlying API (this class - is not portable, or compilable outside of ACE_WIN32). - - * ace/Msg_WFMO_Reactor.{h cpp} (poll_remaining_handles()): Changed - argument from size_t to DWORD - matches similar changes to - WFMO_Reactor, below. - - * ace/Name_Request_Reply.cpp (ctor): Cast length to ACE_UINT32. - - * ace/Naming_Context.cpp (info): Cast info string length to int. - - * ace/OS.{h i cpp}: (sema_post): Changed count from size_t to u_int, - to match both the value range for initializing (sema_init) and - the underlying OS calls on all platforms. - (getipnodebyaddr): For IPv4 case, cast len to int. - (recvfrom): size_t->int casting for Windows. - (read, write, pread, pwrite): Reduce byte count to DWORD range - for Windows. - (getcwd): Windows offers an int len for getcwd(), as opposed to - size_t everywhere else... so special-case the Windows need to - cast the length down. - (thr_create): Cast stack_size to u_int for beginthreadex(). - Other miscellaneous int/size_t casts to get Win64 to build clean. - (iovec): Changed iov_len field from size_t to u_long to match - WSABUF, as comment says to. - - * ace/OS_String.cpp (strtok_r_emulation): Change string lengths from - int to size_t, matching that returned from strlen(). - - * ace/Process.{h cpp} (setenv_i): Change len arg from int to size_t. - Matches string lengths in general, and the expectations of both - callers and callees with this information. - Changed environment_buf_index_ and environment_buf_len_ from int - to size_t as well, matching available range. - - * ace/Process_Manager.cpp: size_t to DWORD casting for current_count_. - - * ace/Registry.cpp (make_name): Use ssize_t instead of int for string - positions - matches interface for ACE_TString, which this method - accepts and uses. - - * ace/Remote_Name_Space.cpp (bind, rebind, resolve, list_names, - list_values, list_types, list_name_entries, list_value_entries, - list_type_entries): Cast length values to ACE_UINT32 as - ACE_Name_Request expects. - - * ace/Select_Reactor_Base.cpp (ACE_Select_Reactor_Handler_Repository): - In open(), cast size to int for call to ACE::set_handle_limit(). - - * ace/Service_Config.cpp (ctor): Cast size value to int for passing to - ACE_Service_Repository::instance(). The ACE_Service_Repository - size should probably be changed to size_t for consistency. - - * ace/Service_Manager.cpp (info): Cast return string length to int. - - * ace/Service_Types.cpp (info): Cast return string length to int. - - * ace/Shared_Memory_MM.i (get_segment_size): - * ace/Shared_Memory_SV.i (get_segment_size): Cast shared_memory_ size - to int. Legit since the open() call's size is an int; otherwise - this should be changed to return a size_t. - - * ace/SOCK_Connector.cpp (shared_connect_start): Align use of - int/size_t. - - * ace/SOCK_Dgram.{h cpp} (recv, send): - * ace/SOCK_Dgram_Bcast.{h cpp} (recv, send): - * ace/SOCK_Sgram_Mcast.{h i} (send, recv): Change iovcnt from - size_t to int to match the args used by other iovcnt-accepting - methods in ACE. - - * ace/SOCK_Dgram_Bcast.i (send): Align use of int/size_t. - - * ace/SOCK_IO.cpp (recv (size_t, ...), send(size_t, ...)): The - comments say the pairs are char*, int, but the code extracted - ssize_t values instead of int. Changed to extract ints as the - comment says. The values are put in iovecs which accept - unsigned long, not ssize_t. - - * ace/SPIPE_Addr.cpp (set): Align use of int/size_t. - - * ace/SPIPE_Stream.cpp (send, recv): Align use of int/size_t. - - * ace/SString.i (ACE_NS_WString): Add missing arg for ACE_WString - ctor initialization. - (ACE_SString::rfind): Cast string len to int for searching... note - this limits the string to 'int' range, though the other size - arguments/parameters are size_t. There are crocodiles lurking - here... hopefully this class is on its way out. - - * ace/SV_Shared_Memory.{h i} (get_segment_size, round_up): Changed - return type from int to size_t, matching it's argument type. - Also changed class's size_ member from int to size_t, matching - the arg from which it's set. - - * ace/Synch.{h i} (ACE_Semaphore::release): Change release count from - size_t to u_int - also see corresponding change to OS.h, above. - - * ace/Thread.{h cpp} (spawn_n): Change return value from int to size_t, - matching the range available via the 'n' argument. - - * ace/Thread_Manager.{h i cpp}: - (count_threads): Change return value from int to size_t. Matches - the requestable number of threads in spawn_n, as well as - underlying container size return. - (find_task): Change slot argument from int to size_t... matches - range of available slots. The default changed from -1 to 0, but - should make no difference since first check is for 0 >= <slot>. - (thread_all_list, task_all_list, task_list, thread_list, - hthread_list, thread_grp_list, hthread_grp_list): Changed - return value from int to ssize_t, to match <n> argument. - - * ace/Timer_Hash_T.{h cpp}: This timer queue generates timer IDs - by casting a pointer to a dynamically-allocated structure to - a long. This, of course, is invalid on Win64. On Win64, the ID - is obtained by masking off the lower 32 bits of the pointer, - and saving the upper 32 bits. This adds an assumption that all - the Hash_Tokens allocated will have the same upper 32 bits in - the pointer value. Note that the original pointer value is used - as the ACT value in the timer scheduled within this class. The - timer ID must only be reconverted to cancel a timer by ID, or - to explicitly reschedule it. Expiration and cancel-by-handler - are all handled by reconverting the ACT to the dynamically-allocated - memory area (Hash_Token). - Also changed hash starter from time.usec() to time.sec(). usec() - isn't usually anything of value. - - * ace/Timer_Wheel_T.cpp: Added #pragma to disable compile warnings. - We know of the issue and deliberately used the fields this way. - - * ace/UPIPE_Stream.{h cpp} (send, recv): Changed return value from - int to ssize_t, consistent with other IPC streams in ACE. - - * ace/WFMO_Reactor.{h cpp} (ACE_WFMO_Reactor_Handler_Repository): - Changed max_handlep1_ and max_handlep1 () from size_t to DWORD. - Matches what they're used for (input to WaitForMultipleObjects). - (ACE_WFMO_Reactor): Changed wait_for_multiple_events() to return - DWORD, not int - matches what WaitForMultipleObjects returns. - Changed wait_status and slot args to dispatch(), safe_dispatch(), - dispatch_handles (), dispatch_handler (), simple_dispatch_handler (), - complex_dispatch_handler (), and poll_remaining_handles () to also - be DWORD, not int, as they're working directly (or close to it) - with the same WFMO return value. - - * ace/WIN32_Asynch_IO.cpp (ACE_WIN32_Asynch_Read_Stream::readv, - ACE_WIN32_Write_Stream::writev, ACE_WIN32_Asynch_Read_Dgram::recv, - ACE_WIN32_Asynch_Write_Dgram::send): - Make multiple iovecs if needed to transfer all data requested. - (ACE_WIN32_Asynch_Read_Stream::shared_read, - ACE_WIN32_Asynch_Write_Stream::shared_write, - ACE_WIN32_Asynch_Read_File::readv, - ACE_WIN32_Asynch_Write_File::writev, - ACE_WIN32_Asynch_Accept::accept, - ACE_WIN32_Asynch_Transmit_File::transmit_file): Limit requested - bytes to MAXDWORD. Return -1, ERANGE if over. - - * ace/WIN32_Proactor.{h cpp}: Changed number_of_threads_ member from - size_t to DWORD, matching the API calls it's passed to. - (post_completion): Cast transfer count from size_t back to DWORD - for posting. The cast is ok since the original requested transfer - counts (in WIN32_Asynch_IO) were limited to DWORD range. - - * tests/Config_Test.cpp: - * tests/Proactor_Scatter_Gather_Test.cpp: - * tests/Proactor_Test.cpp: - * tests/Proactor_Timer_Test.cpp: - * tests/Reactor_Timer_Test.cpp: - * tests/Semaphore_Test.cpp: - * tests/TP_Reactor_Test.cpp: size_t/int fixes. - -Fri Oct 4 18:39:19 2002 Steve Huston <shuston@riverace.com> - - * tests/Recursive_Condition_Test.cpp: Added a deactivate/wait - on the timer queue to make sure the queue shuts down. This - reveals a bug in the recursive condition... it's stuck. - Also added a much simpler wait/signal test that shows the issue - without the additional stuff to trudge through while looking for - the problem. Thanks to Philippe Haussy <p.haussy@xpedite.com> - for reporting this and sending a nice, concise test case. - - * THANKS: Added Philippe Haussy to the Hall of Fame. - -Thu Oct 3 17:29:36 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/OS_Dirent.cpp (readdir_emulation): Don't do a FindNextFile() - right after FindFirstFile() to avoid overwriting the existing - data! Thanks to Dmitry Botcharnikov <dimab@garant.ru> for - reporting this. - -Wed Oct 2 20:44:19 2002 Steve Huston <shuston@riverace.com> - - * tests/Proactor_Scatter_Gather_Test.cpp: Added some more diagnostic - info and removed ACE_ASSERT on file not compared correctly - just - note the error and return. When about to writev a chain of blocks, - update the odd/even chains before writing in case the write fails; - fixes a crash on NT4... test still fails, but doesn't crash. - -Wed Oct 2 10:50:49 2002 Ossama Othman <ossama@uci.edu> - - * ace/Basic_Types.h: - * ace/Basic_Types.cpp: - - Do not rely on the u_int and u_long typedefs. They are not - always defined in system headers. Addresses a compile-time - issue when _XOPEN_SOURCE is explicitly set to 600 on Glibc - platforms, such as Linux. - - * ace/OS.h: - - Explicitly define the u_{char,short,int,long} typedefs on Glibc - platforms when _BSD_SOURCE is not defined. Addresses a - compile-time issue when _BSD_SOURCE is not defined. - - * ace/config-linux.h: - - If ACE_HAS_PTHREADS_UNIX98_EXT is not defined, do not define - ACE_LACKS_RWLOCK_T. Addresses a compile-time issue when - _XOPEN_SOURCE is not set to a value >= 600. - - * ace/config-linux-common.h: - - Reverted change that explicitly defined _XOPEN_SOURCE to 600. - Explicitly defining _XOPEN_SOURCE to 600 causes many - compile-time errors and warnings to occur. - - Moved the check for (_XOPEN_SOURCE >= 600) after the - <features.h> include. That header may define _XOPEN_SOURCE when - other macros, such as _GNU_SOURCE, are defined. - -Wed Oct 2 01:08:51 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ace/SString.cpp (ACE_NS_WString): Fix the silly off-by-one error - causing heap corruption and core dumps in $ACE_ROOT/Naming_Test. - Thanks to Vladislav <vladislav@gtss.spb.ru> for reporting the - problem. - -Tue Oct 1 20:46:35 2002 Christopher Kohlhoff <chris@kohlhoff.com> - - * ace/Atomic_Op.cpp: - Added explicit instantiations for ACE_Atomic_Op<ACE_Thread_Mutex,long> - and ACE_Atomic_Op_Ex<ACE_Thread_Mutex,long>. Programs using ACE should - no longer explicitly instantiate these classes. This change eliminates - the complexity of having to test for the ACE_HAS_BUILTIN_ATOMIC_OP - define when doing explicit instantiations. - - * tests/Atomic_Op_Test.cpp: - * tests/Notify_Performance_Test.cpp: - Removed ACE_Atomic_Op explicit instantiations that are now made in the - ACE library itself. - -Tue Oct 1 19:49:36 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ace/Exception_Macros.h: - - Fixed some documentation for using the Exception macros. - - * ace/OS.i: - * ace/config-aix-4.x.h: - * ace/config-hpux-11.00.h: - * ace/config-linux-common.h: - * ace/config-linux.h: - - Enable support for using native rwlocks on Linux. - -Tue Oct 1 05:06:18 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * ace/WIN32_Asynch_IO.cpp: Minor cosmetic changes. - -Mon Sep 30 12:21:32 2002 Ossama Othman <ossama@uci.edu> - - From Christophe Juniet <cjuniet@entreview.com> - * ace/Sock_Connect.cpp (get_bcast_addr, get_ip_interfaces) - (count_interfaces): - - The ifr_addr.sa_len field was not taken into account when - incrementing the ifreq pointer on FreeBSD. Addresses network - interface enumeration problems in FreeBSD builds. - -Mon Sep 30 09:33:42 2002 Christopher Kohlhoff <chris@kohlhoff.com> - - * bin/MakeProjectCreator/templates/bor.mpd: - * include/makeinclude/build_dll.bor: - * include/makeinclude/build_exe.bor: - * include/makeinclude/build_lib.bor: - * include/makeinclude/clean.bor: - * include/makeinclude/outputdir.bor: - Add support for the 'install', 'dllout' and 'libpaths' - MakeProjectCreator settings in generated Borland makefiles. - -Mon Sep 30 13:25:12 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * include/makeinclude/platform_linux_borland.GNU: - Now it is possible to create shared libraries, but no executables - yet with Kylix3. - -Mon Sep 30 11:33:12 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * include/makeinclude/platform_linux_borland.GNU: - Extended Kylix3 GNU file. - -Mon Sep 30 11:13:33 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/config-borland-common.h: - * ace/config-win32-common.h: - Moved the dll export stuff for the Borland compiler from the win32 - file to the borland file because the Kylix compiler also supports - the dll export stuff for shared objects. - -Mon Sep 30 10:19:33 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * include/makeinclude/rules.bin.GNU: - Added special part for Kylix. The Kylix linker expects that the - object files and the name of the executable are separated using - a comma. - -Sun Sep 29 08:19:09 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ACEXML/examples/SAXPrint/Makefile: Removed inclusion of - rules.lib.GNU to allow mingw compilation to proceed. Thanks to - Vince Mounts <vince@mounts.cc> for reporting this. - - * ace/SString.h: Moved the ACE_INLINE before ACE_Export in - declaration of operator+ to resolved a mingw warning. Thanks to - Vince Mounts <vince@mounts.cc> for reporting this. - -Fri Sep 27 12:49:33 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * include/makeinclude/rules.lib.GNU: - Added special part for Kylix - -Thu Sep 26 21:52:05 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * THANKS: Added Erich Hochmuth" <erich_hochmuth@hotmail.com> to - the hall of fame. - -Thu Sep 26 13:48:26 2002 Balachandran Natarajan <bala@isis-server.isis.vanderbilt.edu> - - * ace/TP_Reactor.cpp (remove_handler): Make a check for the return - value from the remove_handler_i () call before calling - handle_close (). Thanks to Eric Malenfant - <EMalenfant@interstarinc.com> for suggesting this. - - Eric mentions that doesnt show up the BUG 1020 problem. But - local testing here doesnt tell so. I still see some core dumps - with a test, albeit slowly! The patch cannot do any harm and - hence this is checked in. - -Thu Sep 26 07:16:56 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/Local_Name_Space.h (ACE_NS_String): Removed the comment - about ACE_NS_String not having a destructor since this seems to - have been overtaken by events. Thanks to Vladislav Zverev - <vladislav@gtss.spb.ru> for reporting this inconsistency. - -Thu Sep 26 09:18:55 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/Logging_Strategy.h: - Improved doxygen documentation. - -Thu Sep 26 05:42:33 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * include/makeinclude/platform_gnuwin32_common.GNU: - Since mingw can now handle exceptions across dll boundries - removed that restriction. Thanks for Vince Mounts <vince@mounts.cc> - for reporting this. - -Wed Sep 25 17:58:43 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/Basic_Stats.h: - * ace/Connector.h: - * ace/INET_Addr.h: - * ace/OS.h: - * ace/OS.i: - * ace/Service_Config.h: - * ace/Stats.h: - * ace/Timer_Queue_T.cpp: - * ace/TLI_Connector.cpp: - * ace/XTI_ATM_Mcast.i: - * ace/RMCast/RMCast_Reliable_Factory.h: - Small fixes in method argument names, types, comments to fix doxygen - warnings and improve the generated documentation. - -Wed Sep 25 08:44:13 2002 Ossama Othman <ossama@uci.edu> - - * tests/Notify_Performance_Test.cpp: - - Do not explicitly instantiate an ACE_Atomic_Op<ACE_Thread_Mutex, - long> template if ACE_HAS_BUILTIN_ATOMIC_OP is defined. - Defining it causes template specialization with the same - parameters to be declared. Fixes a compile-time problem in - explicit template instantiation builds. - -Wed Sep 25 14:49:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * bin/fuzz.pl: - Extended the types of files that are tested in the doxygen - @file with the files_idl. Some of the idl files are parsed to - generate doxygen documentation and when then a @file tag - exists the filename must be correct. - -Tue Sep 24 20:17:50 UTC 2002 Don Hinton <dhinton@objectsciences.com> - - * tests/Process_Manager_Test.cpp: Removed the ACE_HAS_THREADS - requirement and the unneeded call to - ACE_Reactor::instance ()->owner () so that it would run cleanly - on single threaded builds. Thanks to Steve Huston for this - suggestion. - - * tests/run_test.lst: Added !STATIC to Framework_Component_Test since - it must load a shared library as part of the test. - -Tue Sep 24 16:07:21 2002 Steve Huston <shuston@riverace.com> - - * examples/C++NPv1/Logging_Client.cpp: - * examples/C++NPv2/Select_Reactor_Logging_Server.cpp: - * examples/C++NPv2/Server_Shutdown.cpp: - * examples/C++NPv2/TP_Reactor_Logging_Server.cpp: - For Windows builds without ACE_HAS_STANDARD_CPP_LIBRARY, use - gets() and a char[] instead of the standard string and getline - functions. We don't want to be going this way for ACE in general, - but this fixes the legacy-mode builds. - -Tue Sep 24 06:30:31 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/ACE.h: - * ace/Acceptor.h: - * ace/Activation_Queue.h: - * ace/Asynch_Acceptor.h: - * ace/Asynch_Connector.h: - * ace/Arg_Shifter.h: - * ace/Get_Opt.h: - * ace/Local_Tokens.h: - * ace/Method_Request.h: - * ace/Thread_Manager.h: - * ace/WIN32_Asynch_IO.h: - * ace/INET_Addr.h: - Corrected doxygen tags. - - * ace/CDR_Stream.cpp (write_string): - * ace/Map_T.i (find, unbind): - * ace/Timer_Hash_T.cpp: - Made arguments in method definition and implementation the - same. - - * ace/Future.h: - Improved doxygen documentation. - - * ace/POSIX_CB_Proactor.i: - Added same ifdef as in the header file around the code in this - file, so that doxygen doesn't parse it. - -Mon Sep 23 13:43:11 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/WorkspaceCreator.pm: - - Fixed a bug with the per-project workspaces. - -Mon Sep 23 13:00:13 2002 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * ace/Log_Msg_NT_Event_Log.{h,cpp}: Made the #ifdefs consistent - in both the *.h and *.cpp files. Thanks to Johnny Willemsen for - reporting this. - -Mon Sep 23 13:14:28 2002 Steve Huston <shuston@riverace.com> - - * ace/Task.h (wait()): - * ace/Thread_Manager.h (wait_task()): Improved documentation, also - clarifying that the waited-for threads have been joined upon return. - -Mon Sep 23 12:05:38 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/BorlandWorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/GNUWorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/NMakeWorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/WorkspaceCreator.pm: - - For GNU, NMake and Borland projects, create a workspace per - project file. - - * bin/MakeProjectCreator/modules/BorlandProjectCreator.pm: - * bin/MakeProjectCreator/modules/Creator.pm: - * bin/MakeProjectCreator/modules/GHSWorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/TemplateParser.pm: - - Remove trailing whitespace. - -Sun Sep 22 08:49:41 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/Svc_Handler.cpp (flush_i): Don't flush the queued messages - if the connection fails. Thanks to Andrew Finnell - <andrew@activesol.net> for this fix. - -Sat Sep 21 23:05:56 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * bin/subst_env.pl: Improved conversion to different coding - styles. Thanks to Oliver Kellog for suggesting this. - -Fri Sep 20 13:52:23 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/NMakeProjectCreator.pm: - - Fixed a problem with implicit idl dependencies. - - * bin/MakeProjectCreator/README: - * bin/MakeProjectCreator/modules/Creator.pm: - * bin/MakeProjectCreator/modules/ProjectCreator.pm: - * bin/MakeProjectCreator/modules/TemplateParser.pm: - * bin/MakeProjectCreator/templates/bor.mpd: - * bin/MakeProjectCreator/templates/em3vcp.mpd: - * bin/MakeProjectCreator/templates/gnu.mpd: - * bin/MakeProjectCreator/templates/nmake.mpd: - * bin/MakeProjectCreator/templates/vc6dsp.mpd: - * bin/MakeProjectCreator/templates/vc7.mpd: - - Added the ability to override or modify idlflags on a per file - group basis. - -Thu Sep 19 22:13:13 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ACEXML/common/HttpCharStream.cpp: Added license for the FSM in - HttpCharStream.cpp. It is a shortened BSD license which is - compatible with ACE/TAO licensing. - -Wed Sep 18 12:26:22 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * examples/Misc/test_trace.cpp: The call to activate() was - being invoked with the parameters in the wrong order. Thanks to - Jeff W <jeffw_1234@yahoo.com> for reporting this. - -Tue Sep 17 17:26:40 2002 Ossama Othman <ossama@uci.edu> - - * ace/Basic_Types.cpp: - - Cosmetic changes to clarify some preprocessor blocks. - - * ace/CDR_Base.h: - - Corrected placement of '#' character in preprocessor - directives. It belongs in the first column, not arbitrary - ones. - -Tue Sep 17 01:35:54 UTC 2002 Don Hinton <dhinton@ieee.org> - - * ace/OS.h: Remove redundant MAXPATHLEN definition when - building VxWorks. Thanks to Rob Andzik <andzik@rtlogic.com> - for pointing this out. - -Mon Sep 16 17:36:33 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/Hash_Cache_Map_Manager_T.cpp: - Use typedefs from header file. - - * ace/Process.i (setreugid): - Made method implementation same as definition - -Mon Sep 16 15:26:56 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/config-doxygen.h: - Added define ACE_HAS_TEMPLATE_TYPEDEFS to generate - ACE_Event_Handler_T documentation. - -Mon Sep 16 07:35:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/config-doxygen.h: - Added define ACE_HAS_EVENT_POLL to generate Dev_Poll reactor - documentation - - * ace/Acceptor.h: - * ace/Containers_T.{h,cpp}: - * ace/Dev_Poll_Reactor.h: - * ace/Sample_History.h: - * ace/Service_Config.h: - Improved doxygen documentation. - -Sat Sep 14 10:34:34 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/config-freebsd-pthread.h: Added - #undef ACE_LACKS_UCONTEXT_H if we're compiling for FreeBSD - version > 4. Thanks to Christophe Juniet - <cjuniet@entreview.com> for reporting this. - -Fri Sep 13 08:37:09 2002 Ossama Othman <ossama@uci.edu> - - * ace/Time_Value.h: - * ace/Time_Value.inl: - - Removed copy constructor. The default compiler-generated copy - constructor will do fine since ACE_Time_Value is a POD class. - This may also avoid potential confusion due to a missing - explicit assignment operator. The compiler-generated assignment - operator will suffice in this case, too. - - From Gonzalo Diethelm <gonzalo.diethelm@aditiva.com> - * bin/.cvsignore: - * tests/.cvsignore: - * tests/log/.cvsignore: - - Files to silence annoying CVS messages regarding generated files - it doesn't know about. - -Fri Sep 13 08:57:03 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/USAGE: - * bin/MakeProjectCreator/modules/Creator.pm: - * bin/MakeProjectCreator/modules/Driver.pm: - * bin/MakeProjectCreator/modules/ProjectCreator.pm: - * bin/MakeProjectCreator/modules/TemplateParser.pm: - * bin/MakeProjectCreator/modules/WorkspaceCreator.pm: - * bin/MakeProjectCreator/templates/bor.mpd: - * bin/MakeProjectCreator/templates/em3vcp.mpd: - * bin/MakeProjectCreator/templates/gnu.mpd: - * bin/MakeProjectCreator/templates/nmake.mpd: - * bin/MakeProjectCreator/templates/vc6dsp.mpd: - * bin/MakeProjectCreator/templates/vc7.mpd: - - Added the ability to modify template and project variable - assignments from the command line. Also added rules to process - idl files and place the output in the directory relative to the - specified idl file. - - Modified to not add duplicates to file lists. - -Thu Sep 12 17:18:43 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/Synch.{h,cpp}: Only define the recursive condition support if - we're running on Win32. Hopefully, this will be generalized for - VxWorks at some point. Thanks to Rob Andzik - <andzik@rtlogic.com> for reporting this. - -Thu Sep 12 19:29:41 2002 Steve Huston <shuston@riverace.com> - - * ace/Atomic_Op_T.i (dump): Don't try to return a value from a - 'void' method. - -Thu Sep 12 17:38:10 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/Malloc_T.h: - * ace/Malloc_T.i (release): Added an extra argument (with default - value) to release so that we could decrease the refcount and - release the backing store if needed atomically (while holding the - lock.) This was causing a race condition when removing mmap - files used by MEM_Stream's. - - * ace/Memory_Pool.h: - * ace/Memory_Pool.i: - * ace/Memory_Pool.cpp: Changed the release method to take an extra - argument so that we can control whether we want to perform - cleanup on backing store or not. - - * ace/MEM_SAP.cpp (close_shm_malloc): Changed to use the extended - release method to remove the shared_memory. - - * ace/MEM_IO.cpp: When sending/receiving data reactively, make - sure we don't even try if the socket is not valid. - -Thu Sep 12 08:39:46 2002 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * ace/Synch.h (class ACE_recursive_mutex_state): Changed "HANDLE" to - "ACE_HANDLE" to avoid problems with VxWorks. Thanks to Rob - Andzik <andzik@rtlogic.com> and Don Hinton <dhinton@ieee.org> - for reporting this. - -Thu Sep 12 17:36:36 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * include/makeinclude/wrapper_macros.GNU: Use awk instead of - perl to determine the ACE major, minor, and beta versions. - -Thu Sep 12 12:02:31 2002 Steve Huston <shuston@riverace.com> - - * tests/TSS_Test.cpp: Pass iteration count to worker() as a pointer - to int, not an int cast to a pointer and back. - -Wed Sep 11 08:44:28 2002 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * Our thoughts and prayers go out to the victims of 9/11/01 on the - anniversary of this terrible tragedy. - -Wed Sep 11 13:03:48 2002 Carlos O'Ryan <coryan@atdesk.com> - - * ace/Select_Reactor_Base.cpp: - Check the suspend mask before completely forgetting about a - handle, otherwise things do not work correctly with TP_Reactors - and multiply-registered handlers. This fixes: - http://deuce.doc.wustl.edu/bugzilla/show_bug.cgi?id=1306 - -Tue Sep 10 17:46:51 2002 Steve Huston <shuston@riverace.com> - - * ace/Malloc_T.{h i cpp} (ACE_Malloc_T): Changed the lock_ member from - an ACE_LOCK to an ACE_LOCK *. Also added another constructor that - accepts an ACE_LOCK pointer instead of a ACE_TCHAR *lock_name. - This allows someone to form an ACE_LOCK externally and supply - it to the ACE_Malloc_T object. This is useful if special - permissions need to be set on the lock, such as the mode_t for - ACE_Process_Mutex. - -Tue Sep 10 14:06:02 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ace/Obstack_T.cpp: - * ace/Obstack_T.h: - - Obstack is no longer restricted to fixed-size chunks. Chunk size - is increased by a power of two if we run out of memory in a - chunk. Added implementation of unwind operations. Stack can be - unwound by giving the address of an object on the stack. Fixed a - couple of bugs in the existing implementation when used with - WCHAR. We were incrementing the pointers by sizeof (CHAR) as - well as multiplying by sizeof (CHAR) to get the number of bytes - to be copied. This improvement doesn't deal with the alignment - issues as it requires a change in Obchunk which is postponed - till after ACE 5.3 is released. - - * tests/Obstack_Test.cpp: - - New, improved test which exercises all the code paths in the - implementation unlike the previous test. - -Tue Sep 10 12:32:33 2002 Carlos O'Ryan <coryan@atdesk.com> - - * ace/Connector.cpp: - Be more informative in error messages, at least report in what - library, class and function was the error detected. This fixes: - http://deuce.doc.wustl.edu/bugzilla/show_bug.cgi?id=1304 - -Tue Sep 10 12:28:41 2002 Carlos O'Ryan <coryan@atdesk.com> - - * ace/Connector.cpp: - Be more paranoid about using AST's returned from cleanup_AST() - and always check the return value of that function. This fixes: - http://deuce.doc.wustl.edu/bugzilla/show_bug.cgi?id=1303 - -Mon Sep 9 16:44:00 2002 Balachandran Natarajan <bala@isis-server.vuse.vanderbilt.edu> - - * THANKS: Surprise, surprise, Gautam Thaker'ss name was not there - in the THANKS file. So added it. - -Mon Sep 9 16:57:41 2002 Steve Huston <shuston@riverace.com> - - * ace/POSIX_Asynch_IO.cpp (cancel): Don't sanity-check the - proactor type - just call cancel_aio and let the proactor impl - figure it out. - - * examples/Reactor/Proactor/test_aiocb.cpp: Fixes to not crash. - -Mon Sep 09 11:37:13 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/OS.cpp (cond_timedwait): Changed to pass absolute time when - using emulated semaphore on WinCE. Thanks to Herbert Wang - <herbert@nowsol.com> for reporting this. - - * ace/Thread_Manager.cpp (join): This function was not cropping - the thread return status when a thread had already exited when - we called and its Thread_Descriptor had been moved to - terminated_thr_list_. Thanks to Roger Weeks - <rogerw@firstlogic.com> for reporting the problem. - -Mon Sep 9 00:12:04 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * ace/Makefile.dirs: - * examples/Makefile: - * tests/Makefile: - Remove references to obsolete ACE_HAS_GNUG_PRE_2_8 make - variable. - -Sat Sep 7 23:22:20 2002 Balachandran Natarajan <bala@isis-server.vuse.vanderbilt.edu> - - * bin/make_release: The tar ball that is put out during the - release, is copied now to the the previous_versions directory - with the appopriate version information. This should work right - since this was tested against a local repo. - - Thanks to Preston A. Elder <prez@gentoo.org> for motivating - this change. - -Sat Sep 7 18:03:21 2002 Steve Huston <shuston@riverace.com> - - * ace/Proactor.cpp: Removed some ACE_Proactor:: prefixes on member - variable access, replacing with this-> to clarify what's happening. - - * ace/SUN_Proactor.cpp (handle_events): No need to cast a pointer to - long to check for 0. - Re-enabled some "should never happen" ACE_DEBUG lines until we - find out what's going wrong with this stuff. - -Thu Sep 5 12:13:51 2002 Steve Huston <shuston@riverace.com> - - * ACEXML/common/URL_Addr.inl: Removed ACE_RCSID - it's in URL_Addr.cpp. - -Wed Sep 4 16:31:29 2002 Douglas C. Schmidt <schmidt@cs.wustl.edu> - - * Kelly Clarkson wins "American Idol." Go Kelly!! - -Wed Sep 4 17:50:27 2002 Steve Huston <shuston@riverace.com> - - * tests/run_test.pl: Change run time limit from 300 to 400 seconds. - Some tests are timing out, and when run by hand, they run just over - 5 minutes, so give them a little more time. - -Wed Sep 4 17:31:24 2002 Steve Huston <shuston@riverace.com> - - * tests/Cached_Conn_Test.cpp: ACE_DEBUG %@ for pointers, not %d. - -Wed Sep 4 20:59:40 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * include/makeinclude/platform_irix5.2.GNU: - * include/makeinclude/platform_m88k.GNU: - * include/makeinclude/platform_netbsd.GNU: - * include/makeinclude/platform_qnx_neutrino.GNU: - * include/makeinclude/platform_qnx_rtp_gcc.GNU: - * include/makeinclude/platform_sco-nothread.GNU: - Instead of setting -fno-implicit-templates, - use $(TEMPLATES_FLAG). Include platform_g++.GNU to get the - correct value of $(TEMPLATES_FLAG). - -Wed Sep 4 16:31:29 2002 Steve Huston <shuston@riverace.com> - - * ace/Timer_Wheel_T.cpp: Formatting tweaks for ACE coding style. - -Wed Sep 4 20:08:54 UTC 2002 Don Hinton <dhinton@objectsciences.com> - - * ace/Framework_Component.cpp (close): - - Delete each component explicitely instead of calling - close_singleton () to avoid a memory leak. The dtor calls - close_singleton () anyway. Thanks Vladimir Chovanec - <Vladimir CHOVANEC@asset.sk> for reporting the leak and - suggesting the fix. - -Tue Sep 03 22:59:27 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ACEXML/common/XML_Common.dsp: - - Added URL_Addr.inl to common. - -Tue Sep 03 22:40:22 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ACEXML/common/HttpCharStream.cpp: - - Handle WCHAR properly when getting data from URLs. - - * ACEXML/common/URL_Addr.cpp: - * ACEXML/common/URL_Addr.h: - * ACEXML/common/URL_Addr.inl: - - Fixes for proper parsing of URLs when URLs are WCHARs. - - * ACEXML/examples/SAXPrint/main.cpp: - - Use int in GetOpt instead of char or ACEXML_Char. - -Tue Sep 3 17:12:59 2002 Ossama Othman <ossama@uci.edu> - - * ace/OS.cpp (num_processors_online): - - Fixed ACE_TRACE macro. - -Tue Sep 3 16:38:47 2002 Ossama Othman <ossama@uci.edu> - - * ace/ace_dll.dsp: - * ace/ace_lib.dsp: - - Atomic_Op.cpp is no longer a template source file. Moved it to - the "Source Files" folder. - - * ace/Synch.cpp (ACE_Mutex): - - Fixed "unused 'mode' argument" warning. - -Tue Sep 3 15:54:33 2002 Christopher Kohlhoff <chris@kohlhoff.com> - - * ace/OS.h: - * ace/OS.cpp: - - Added new functions ACE_OS::num_processors() and - ACE_OS::num_processors_online() for determining the number of CPUs - available on a system. - - * ace/Atomic_Op.cpp: - * ace/Atomic_Op.h: - * ace/Atomic_Op.i: - * ace/Atomic_Op_T.cpp: - * ace/Atomic_Op_T.h: - * ace/Atomic_Op_T.i: - * ace/Object_Manager.cpp: - * ace/Makefile.bor: - * ace/Makefile.ace: - * ace/ace.mpc: - * tests/Atomic_Op_Test.cpp: - - Added a new fast, lightweight ACE_Atomic_Op template specialization - for x86 platforms that uses the architecture's atomic integer - primitives. It currently supports Win32 using MSVC6 or BCB, and g++ - for (in theory) all x86 operating systems. It selects the - appropriate implementation for single- or multi-CPU systems at - runtime, based on the return value of the ACE_OS::num_processors - function added above. - - Note that Atomic_Op.cpp is now a real .cpp file and not a template - file, and so project files need to be updated accordingly. The - template definitions have been moved to the new Atomic_Op_T.* files. - - Note also that the ACE_Atomic_Op::mutex() operation has been - deprecated (and is not supported for the new specialization). If - you need this functionality, consider using the ACE_Atomic_Op_Ex - template instead. - -Tue Sep 3 15:40:39 2002 Ossama Othman <ossama@uci.edu> - - * tests/XtReactor_Test.cpp (sock_callback): - - Removed this unused static function. - - * tests/TkReactor_Test.cpp (sock_callback): - - Likewise. - - (ACE_TMAIN): - - Fixed "deprecated conversion from string constant to char *" - warning. - -Tue Sep 3 15:16:03 2002 Balachandran Natarajan <bala@isis-server.vuse.vanderbilt.edu> - - * bin/auto_run_tests.lst: Fixed a typo! - -Mon Sep 2 16:48:29 2002 Steve Huston <shuston@riverace.com> - - * tests/OS_Test.cpp (ctime_r_test): Corrected ACE_Log_Msg to use - %P, not %p, to list process ID. - -Mon Sep 2 13:17:26 2002 Steve Huston <shuston@riverace.com> - - * ace/Memory_Pool.{h cpp}: - (ACE_MMAP_Memory_Pool_Options): Added optional file_mode arg - to set the protection mode on the mapped file, if it is - created; defaults to ACE_DEFAULT_FILE_PERMS. - (ACE_MMAP_Memory_Pool): Added file_mode_ member that can be set - by the ACE_MMAP_Memory_Pool_Options. Defaults to - ACE_DEFAULT_FILE_PERMS. It's used for call to ACE_Mem_Map::open() - and ACE_Mem_Map::map(). - - * ace/Process_Mutex.{h cpp}: ACE_Process_Mutex ctor takes an optional - mode_t arg to set the backing store mode (for ACE_Mutex use) or the - mode for the System V semaphore, when it's used. - - * ace/Synch.{h cpp} (ACE_Mutex): Constructor takes an optional mode - to set backing store protection for Pthreads process mutex case, - where memory must be allocated to hold the mutex. - - These are laying the groundwork for fixing Bugzilla #1219. - -Mon Sep 2 14:07:12 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ace/Exception_Macros.h: - - New file which holds the exception macros previously present in - CORBA_macros.h. This file also makes the macros a bit generic so - that they can be used in other modules than TAO. - - * ace/CORBA_macros.h: - - Include only the definitions needed for TAO/CORBA. This file can - be placed under $TAO_ROOT now. Leaving it for compatibility - reasons. - - * ace/OS.i: - - Changed the definition of ctime_buf_size to const so that gcc - doesn't complain about using variable-sized array which is - non-ISO C++. - -Mon Sep 2 10:00:40 2002 Balachandran Natarajan <bala@isis-server.vuse.vanderbilt.edu> - - * ace/Timer_List_T.cpp (get_first_i): Fixes fuzz errors. - -Sun Sep 01 08:18:37 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/Task.cpp (svc_run): - The latest version of the Borland compiler doesn't like the - static cast, so only do the static cast when the compiler - version is smaller than 0x570 - - * ace/Handle_Set.cpp: - When using Borland on Linux, we don't have fds_bits, but we have - __fds_bits. - -Sun Sep 01 07:57:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/config-win32-borland.h: - * ace/config-borland-common.h: - Improved Borland config files for Kylix support. - -Sun Sep 01 07:25:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * include/makeinclude/platform_linux_borland.GNU: - Added first version of platform .GNU file for the Borland Kylix - compiler. Thanks to Newton Aird <naird@ix.netcom.com> for - providing this first version. - - * THANKS: Added Newton Aird <naird@ix.netcom.com> - -Sat Aug 31 14:57:53 2002 Balachandran Natarajan <bala@isis-server.vuse.vanderbilt.edu> - - * THANKS: Added Cemal Yilmaz <cyilmaz@cs.umd.edu> to the hall of - fame. - -Sat Aug 31 13:55:52 2002 Balachandran Natarajan <bala@isis-server.vuse.vanderbilt.edu> - - * ace/OS.i (ctime_r): This fixes a problem on Linux for the call - to ::ctime_r where the pointer returned (result) is the same as - the pointer (buf) passed in, causing the buffer to get clobbed - if strsncpy is used subsequently to move the result back to the - caller's buffer. - - * tests/OS_Test.cpp: Added a test for ACE_OS::ctime_r (). - - Thanks to Michael Searles <msearles@base16.com> for providing - this patch. - -Sat Aug 31 17:45:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/config-borland-common.h: - Created a new config file for the Borland compiler. This will be - used on the Win32 platform and the Kylix3 compiler on Linux - - * ace/config-linux-common.h: - When the Borland compiler is detected, include the - config-borland-common.h file - - * ace/config-win32-borland.h: - Moved most of the contents to the config-borland-common.h file, - only the lines that are Win32 specific stay here. - -Sat Aug 31 10:32:38 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * ace/Future.cpp (set): Removed a stray reference to "iterator" - and replaced the for loop with a while loop. Thanks to Craig - Rodrigues for reporting this. - -Sat Aug 31 01:20:09 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - Gary Duzan <gduzan@bbn.com> - - * include/makeinclude/rules.nested.GNU: Rewritten by Gary Duzan. - If a list of directories is specified with DIRS= for a recursive - make, make will halt immediately and propagate a non-zero - return code to the shell if an error is encountered in a - subdirectory. The present behavior does not propagate any - errors to make if there is an error in a subdirectory. If make -k - is specified, make will keep on going even if it encounters an - error in a subdirectory, but it will also propagate a non-zero - return code to make when the make finishes. The current behavior - is to keep on going if an error is encountered in a subdirectory, - but NOT propagate a non-zero return code upon termination of make. - - * ace/Makefile.dirs: (Added) Specifies the DIRS= list of directories - to invoke recursive makes on. - - * ace/Makefile.ace: (Added). This builds libACE, like the old Makefile - did. - - * ace/Makefile: Invoke Makefile.ace first, then Makefile,dirs. It - was necessary to split apart the DIRS= rule from the part of the - Makefile which builds libACE in order to have parallel makes (-j 2) - work on recursive makes. - -Fri Aug 30 17:34:00 2002 Justin Michel <michel_j@ociweb.com> - - * ace/Timer_List_T.cpp: - * ace/Timer_List_T.h: - * ace/Timer_Wheel_T.cpp: - * tests/Timer_Queue_Test.cpp: - - Reimplemented timer list to be compatible with Itanium. It no longer - uses a Node* as the timer_id, so the cancel operation is now much - more expensive. On the plus side I was able to make a few optimizations - that improve the speed of the schedule operations. - I also made a small optimization to the timer wheel, and updated - the test slightly. - -Fri Aug 30 15:53:18 2002 Steve Huston <shuston@riverace.com> - - * tests/SSL/Makefile: Move the addition of SSL options/flags up to - before the hackery of the compile command for AIX. - -Fri Aug 30 05:48:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/OS.i (tempnam): - Changed defines because special things only happen on Win32 - -Thu Aug 29 22:01:27 2002 Steve Huston <shuston@riverace.com> - - * include/makeinclude/platform_hpux_gcc.GNU: Make the build options - settable by the command line (and/or platform_macros.GNU), and - fix the threads and soname options, as well we remove the - -fstrict-prototype option from the compile options - it was needed - for an older g++ (nothing prior to g++ 3 works right on HP-UX - anyway) and is not valid for g++ 3.2 and later. Thanks to - Ville Lehtiniemi <guderian@ees2.oulu.fi> for reporting these - issues and sending fixes. - -Thu Aug 29 15:18:31 2002 Steve Huston <shuston@riverace.com> - - * ace/Get_Opt.{h i cpp} (long_option): Only print error messages for - long message addition problems if opterr is set. Also added - ACE_LIB_TEXT to naked strings for ACE_ERROR. - - Added opt_opt() method to return the most recently matched - short option character. Thanks to Rich (Mr. POSIX) Seibel - for the idea and POSIX spec info for this new method. - - * ace/Object_Manager.cpp (init): When finishing up initialization of - the object manager, call ACE_Trace::start_tracing() to reenable - trace in cases where a user does init/fini/init. - - * tests/Get_Opt_Test.cpp: Added usage of ACE_Get_Opt::opt_opt(). - -Thu Aug 29 15:03:23 2002 Ossama Othman <ossama@uci.edu> - - * ace/README: - - Documented ACE_HAS_BROKEN_UNEXPECTED_EXCEPTIONS macro. - - * ace/config-win32-msvc-6.h - (ACE_HAS_BROKEN_UNEXPECTED_EXCEPTIONS): - - MSVC++ 6 does not correctly handle exceptions thrown from - functions/methods that do not list them in the exception - specification. In particular, the unexpected exception handler - is not invoked. - -Thu Aug 29 06:12:56 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * ace/Future.cpp (set): Fixed the loop code so it's possible for - the observer object to be deleted or detached in its update() - operation. Thanks to Tomer Amiaz <Tomer@bandwiz.com> for this - fix. - -Thu Aug 29 09:40:20 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ACEXML/common/Transcode.h: - * ACEXML/parser/parser/Parser.h: Fixed incorrect usage of Doxygen - tag. Thanks to Johnny for pointing this out. - -Thu Aug 29 12:20:17 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/config-win32-borland.h: - Removed the 1 after 2 defines, so that they are like the defines - in other files. - -Thu Aug 29 05:53:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * examples/Reactor/Proactor/post_completions.cpp: - Fixed compile error. - - * ace/OS_Dirent.h: - * ace/OS_Errno.cpp: - * ace/Time_Value.h: - * ace/Time_Value.inl: - Made method signature the same in definition and implementation. - - * ace/TkReactor.cpp: - * ace/TkReactor.h: - Made method argument names the same as in the base so that doxygen - can generate the documentation. - -Wed Aug 28 23:30:31 2002 Steve Huston <shuston@riverace.com> - - * include/makeinclude/wrapper_macros.GNU: Remove addition of SSL - compile/link options so that not all parts of ACE, needing SSL or - not, get these options. - - * ace/SSL/Makefile: - * tests/SSL/Makefile: Added the SSL compile/link options from - wrapper_macros.GNU to these files, where they're needed. - - * ace/POSIX_Proactor.cpp (post_wakeup_completions): Use int - to count completions, not ssize_t - match the passed-in arg type. - Post the completion directly via the proactor, not via the - wakeup completion; the dynamic_cast in ACE_POSIX_Asynch_IO::post_ - completion loses the subclass of ACE_POSIX_AIOCB_Proactor and - when using any other subtype of proactor, the notification is - queued incorrectly. - - (POSIX_SIG_Proactor::notify_completion()) - always queue the signal. - Not doing so leaves a window where MT proactors can lose a signal. - - * tests/Proactor_Test.cpp: Use ACE_Proactor::close_singleton() to - delete proactor instance. Just to be sure we try it. - - * ace/SSL/SSL_Asynch_Stream.{h cpp}: u_long->size_t changes. - -Wed Aug 28 22:42:41 2002 Steve Huston <shuston@riverace.com> - - * examples/Service_Configurator/Misc/main.cpp: - * netsvcs/clients/Naming/Client/main.cpp: - * tests/ARGV_Test.cpp: - * tests/Get_Opt_Test.cpp: Change size_t to int as a side-affect of: - Tue Aug 27 13:28:17 2002 Steve Huston <shuston@riverace.com> - * ace/ARGV.{h i cpp}: Changed argc() value from size_t to int. It - needs to get passed to many places that expect int (argc, argv). - - * tests/Semaphore_Test.cpp: - * tests/TP_Reactor_Test.{cpp h}: - * tests/Upgradable_RW_Test.cpp: Further int/size_t corrections. - -Wed Aug 28 19:56:16 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * docs/ACE-subsets.html: Updated the document to point to the - soreduce tool and to remove the discussion of PACE. Thanks to - Chumsu Kim <chskim@add.re.kr> for motivating this. - -Wed Aug 28 14:24:36 2002 Steve Huston <shuston@riverace.com> - - * ace/Log_Msg.h: Fixed Doxygenation of ACE_Log_Msg. Thanks to - Philip Miller <pwmiller@sarnoff.com> for reporting this. - -Wed Aug 28 11:21:04 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/templates/em3vcpdll.mpt: - * bin/MakeProjectCreator/templates/em3vcpdllexe.mpt: - * bin/MakeProjectCreator/templates/em3vcplib.mpt: - * bin/MakeProjectCreator/templates/em3vcplibexe.mpt: - * bin/MakeProjectCreator/templates/nmakedll.mpt: - * bin/MakeProjectCreator/templates/nmakeexe.mpt: - * bin/MakeProjectCreator/templates/vc6dspdll.mpt: - * bin/MakeProjectCreator/templates/vc6dspdllexe.mpt: - * bin/MakeProjectCreator/templates/vc6dsplib.mpt: - * bin/MakeProjectCreator/templates/vc6dsplibexe.mpt: - - Enabled RTTI for all VC based template input files. - -Wed Aug 28 06:18:35 2002 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * ace: Removed the config-sunos5.8-sunc++-5.4.h file since these - changes should go in the config-sunos5.8.h file rather than - a special file. Thanks to Steve Huston for noticing this. - - * include/makeincludes: Removed the platform_sunos5_sunc++5.4.GNU - file since it's no different from the platform_sunos5_sunc++.GNU - file. Thanks to Steve Huston for noticing this. - -Wed Aug 28 06:12:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * etc/tao.doxygen: - Removed Services.h of the list of not to parse files - - * etc/tao_rtportableserver.doxygen: - Added TAO_NAMESPACE to list of to expanded macros - - * tests/Timer_Queue_Test.cpp: - Fixed unicode build error - - * ace/Timer_Wheel_T.cpp: - Fixed doxygen tags. - -Tue Aug 27 19:09:17 2002 Steve Huston <shuston@riverace.com> - - * ace/Get_Opt.h: Documentation improvements. - -Tue Aug 27 17:00:20 2002 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * ace/Signal.cpp: Fixed the ACE_Sig_Action constructors so that - it doesn't break out of the loop just because a signal handler - failed. Thanks to Marcel Loose <loose@astron.nl> for reporting - this. - -Tue Aug 27 19:03:05 2002 Balachandran Natarajan <bala@isis-server.vuse.vanderbilt.edu> - - * bin/auto_run_tests.lst: Added $TAO_ROOT/tests/ORB_shutdown test - to the daily builds. - -Tue Aug 27 18:52:59 2002 Balachandran Natarajan <bala@isis-server.vuse.vanderbilt.edu> - - * ace/Select_Reactor_T.cpp (wait_for_multiple_events): Normally, - select() will reset the bits in dispatch_set so that only those - filed descriptors that are ready will have bits set. However, - when an error occurs, the bit set remains as it was when the - select call was first made. Thus, we now have a dispatch_set - that has every file descriptor that was originally waited for, - which is not correct. We must clear all the bit sets because we - have no idea if any of the file descriptors is ready. - - Thanks to Jody Hagins <jody@atdesk.com> and Carlos O'Ryan - <coryan@atdesk.com> for reporting this problem and proving the - patches. - -Tue Aug 27 15:56:10 2002 Ossama Othman <ossama@uci.edu> - - * ace/config-sunos5.8-sunc++-5.4.h: - * include/makeinclude/platform_sunos5_sunc++5.4.GNU: - - Solaris 8 + SunC++ SPRO 5.4 platform configuration files. - Thanks to Serge Kolgan <skolgan@objectsciences.com> for - contributing these files. - -Tue Aug 27 17:25:39 2002 Steve Huston <shuston@riverace.com> - - * ace/WIN32_Proactor.{h cpp}: More u_long->size_t corrections. - -Tue Aug 27 16:50:41 2002 Steve Huston <shuston@riverace.com> - - * tests/Proactor_Test.cpp: Fixed mismatched printf formats for - size_t arguments. - -Tue Aug 27 16:32:18 2002 Steve Huston <shuston@riverace.com> - - * ace/Proactor_Impl.h: Fixed some u_long->size_t changes missed - on POSIX platforms. - -Tue Aug 27 14:44:00 2002 Justin Michel <michel_j@ociweb.com> - - * ace/Timer_Wheel_T.cpp: - - Fixed problem with incorrect name in ACE_TRACE macro. - -Tue Aug 27 13:28:17 2002 Steve Huston <shuston@riverace.com> - - * ace/config-win32-common.h: Added #define ACE_MALLOC_ALIGN 8 for - Win64. - - * ace/Malloc_T.cpp (ACE_Cached_Allocator): When dividing up the - allocated space, maintain the platform's alignment requirements. - - * ace/ace_dll64.mak: Removed references to Hash_Map_Manager_rt.*. - - * ace/Stream_Modules.cpp (info): Cast string length back to int. - - * ace/ARGV.{h i cpp}: Changed argc() value from size_t to int. It - needs to get passed to many places that expect int (argc, argv). - - * ace/OS.{h cpp} (ACE_OS::string_to_argv): argc changed from size_t& - to int& to match the change above. - - * ace/FILE_IO.{h i cpp}: - * ace/SPIPE_Stream.{h i}: Follow-up to ACE_SOCK_Stream change in: - Mon Jul 22 16:39:21 2002 Steve Huston <shuston@riverace.com> - (sendv_n, recvv_n, sendv, send, recv) - change iov count from - size_t to int. - - * ace/Functor_T.i (ACE_Pointer_Hash::operator()): Use reinterpret_cast - to go from pointer to u_long. - - * ace/config-win32-msvc-7.h: Define ACE_AUTO_PTR_LACKS_RESET for - Win64 SDK compiler. - - * ace/Strategies_T.{h i cpp} (ACE_Thread_Strategy): Changed n_threads - parameter (and n_threads_ member) from size_t to int - matches - what's expected by ACE_Task::activate(), which is where this - value is passed. - - * ace/Asynch_Acceptor.{h cpp}: - * ace/Asynch_IO.{h cpp}: - * ace/Asynch_IO_Impl.h: - * ace/POSIX_Asynch_IO.{h cpp}: - * ace/POSIX_Proactor.{h cpp}: - * ace/WIN32_Asynch_IO.{h cpp}: - * ace/WIN32_Proactor.{h cpp}: All byte-count arguments and members - in all classes changed from u_long to size_t. Some already were... - - * ace/SOCK_Dgram.i (send): address length is an int, not size_t. - - * tests/Cache_Map_Manager_Test.h (Hash_Key::operator()): Cast the - size_t argument to u_long for return. - - * tests/Cache_Map_Manager_Test.cpp: Cast arg for ACE_OS::srand to - u_int (matches ACE_OS signature), not size_t. - - * tests/CDR_Array_Test.cpp (zero): Change len arg to size_t, and - also the values passed to it. Also, correct length args to - ACE_InputCDR and ACE_OutputCDR - they're size_t, not int. - - * tests/Handle_Set_Test.cpp: Disable the conversion warnings - initializing the array of ACE_HANDLE values on Win64. They're fine. - - * tests/Mem_Map_Test.cpp (reverse_file): Changed size arg from int - to size_t. - - * tests/Buffer_Stream_Test.cpp: - * tests/CDR_Test.cpp: - * tests/Message_Block_Test.cpp: - * tests/Message_Queue_Notifications_Test.cpp: - * tests/Priority_Buffer_Test.cpp: - * tests/Proactor_Scatter_Gather_Test.cpp: - * tests/Proactor_Test.cpp: - * tests/Semaphore_Test.cpp: - * tests/Thread_Pool_Test.cpp: - * tests/TP_Reactor_Test.{h cpp}: - * tests/Upgradable_RW_Test.cpp: Align size_t/int usage. - - * tests/Message_Queue_Notifications_Test.cpp (Watermark_Test::producer) - Change hwm from size_t to ssize_t so it can go less than zero and - break the 'for' loop properly. - - * tests/MT_SOCK_Test.cpp: - * tests/SOCK_Test.cpp: Use an int to pass to select(), always 0 - for Win64. - - * tests/SOCK_Send_Recv_Test.cpp: - * tests/Thread_Pool_Reactor_Resume_Test.cpp: - * tests/Thread_Pool_Reactor_Test.cpp: - * tests/TSS_TEst.cpp: Add missing type cast. - -Tue Aug 27 11:08:54 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * ace/Pair_T.{h,i}: Added support for operator== to the ACE_Pair - can be used in maps, etc. Thanks to Gonzalo Diethelm - <gonzalo.diethelm@aditiva.com> for this addition. - -Tue Aug 27 06:45:58 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/ProjectCreator.pm: - - Fixed a problem where .i and .h files would be added even if they - didn't exist. - -Tue Aug 27 10:09:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * etc/tao_dynamicany.doxygen: - * etc/tao_dynamicinterface.doxygen: - * etc/tao_iormanip.doxygen: - * etc/tao_iortable.doxygen: - * etc/tao_portableserver.doxygen: - * etc/tao_rtcorba.doxygen : - Added TAO_NAMESPACE to the list of defines to expand. - -Mon Aug 26 17:33:54 2002 Ossama Othman <ossama@uci.edu> - - * bin/msvc_auto_compile.pl (Build_Core): - - Build the Messaging, PortableGroup and CosLoadBalancing - libraries in a specific order. CosLoadBalancing depends on the - first two. Fixes automated Win32 static builds. - -Mon Aug 26 17:14:20 2002 Ossama Othman <ossama@uci.edu> - - * tests/TkReactor_Test.cpp: - - Fixed a number of warnings. - -Mon Aug 26 18:21:34 UTC 2002 Justin Michel <michel_j@ociweb.com> - - * ace/Timer_Wheel_T.cpp: - * ace/Timer_Wheel_T.h: - * tests/Timer_Queue_Test.cpp: - - New and improved timer wheel implementation. - -Mon Aug 26 09:51:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/FlReactor.{h,cpp}: - * ace/TkReactor.{h,cpp}: - * ace/QtReactor.{h,cpp}: - Made the names of the arguments of the schedule_timer() method the - same as in the base class. That way doxygen generates the same - documentation for this overruled method as in the base. - - * ace/TLI_Connector.i: - * ace/XTI_ATM_MCast.i: - Made argument names the same as in the header file. - -Mon Aug 26 08:29:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/Connector.cpp: - Replaced all short defines with the original ones. Doxygen can't - generate the documentation because of the short defines. - -Sun Aug 25 14:15:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/config-win32-borland.h: - Certain defines should only be defined without a value. - -Sun Aug 25 12:45:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * etc/tao.doxygen: - Added TAO_NAMESPACE to the list of macro's that must be expanded - when generating doxygen documentation. - -Sat Aug 24 22:47:22 2002 Christopher Kohlhoff <chris@kohlhoff.com> - - * include/makeinclude/ace_flags.bor: - * include/makeinclude/clean.bor: - * include/makeinclude/compiler.bor: - - Updated to support MakeProjectCreator. - - * bin/MakeProjectCreator/config/core.mpb: - * bin/MakeProjectCreator/modules/ProjectCreator.pm: - * bin/MakeProjectCreator/modules/BorlandProjectCreator.pm: - * bin/MakeProjectCreator/modules/BorlandWorkspaceCreator.pm: - * bin/MakeProjectCreator/templates/bor.mpd: - - Fixed support for Borland makefile generation. - - New "core" config for projects that would typically be installed - on production system, i.e. pretty much everything except examples - and tests. - - New "include_dir" project attribute used to specify a directory - path where header files for the project should be installed. - - * ace/ace.mpc: - * ace/QoS/qos.mpc: - * ace/RMCast/rmcast.mpc: - * ace/SSL/ssl.mpc: - - Added "core" as a base project and added "include_dir" value. - - * apps/gperf/src/gperf.mpc: - - Added "core" as a base project. - -Sat Aug 24 16:41:36 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ACEXML/common/FileCharStream.cpp (get): - - Remove :: in front of the macro feof. VC is hopelessly confused. - - * ace/Argv_Type_Converter.h: - - Added #include "ace/OS_Memory.h" to get ACE_NEW. I needed to - touch this to test my changes. - -Sat Aug 24 12:01:54 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * ace/OS.h: #undef the atop macro so that it won't break the new - ACE_OS::atop() method on FreeBSD. Thanks to Russell Carter - <rcarter@pinyon.org> and Christophe Juniet - <cjuniet@entreview.com> for this report. - - * ace/config-lynxos.h: Added some #defines for LynxOS 4.0. We'll - leave these comments out for now until we figure out how to add - them without breaking earlier versions of LynxOS. Thanks to - Umberto Mascia <u.mascia@pxl.it> for reporting this. - - * docs/tutorials/015/Protocol_Stream.h (class Protocol_Stream): - Changed a comment to say "server-side applications" rather than - "client-side applications." Thanks to Mark Xu <ltxd@371.net> - for reporting this. - -Fri Aug 23 18:20:31 2002 Steve Huston <shuston@riverace.com> - - * ace/config-all.h: - * ace/OS_Memory.h: Moved the ACE_NEW macros from config-all.h to - OS_Memory.h - - * ace/Log_Msg.h: Removed #include "ace/Basic_Types.h" (not needed) - and replaced "ace/OS_Errno.h" with "ace/OS.h" since thread defs - are necessary. - - * ace/Basic_Types.h: Don't make thread definitions here - they belong - in ace/OS.h. - - * ace/Cache_Map_Manager_T.h: Added #include "ace/Default_Constants.h" - to get ACE_DEFAULT_MAP_SIZE when instantiating templates on AIX - with Visual Age C++. - - * ace/Caching_Utility_T.cpp: #include "ace/Min_Max.h" and - "ace/OS_Memory.h" so Visual Age C++ can see ACE_MAX and ACE_NEW - when instantiating templates. - - * ace/Arg_Shifter.cpp: Added #include "ace/OS_Memory.h" to get ACE_NEW. - - * netsvcs/clients/Naming/Dump_Restore/Dump_Restore.cpp: Added - #include "ace/Malloc_Base.h" to pick up ACE_Allocator. - - * tests/Cached_Allocator_Test.cpp: Changed default speed tests loops - from 100,000,000 to 10,000,000. run_test.pl times it out on - non-nuclear processors ;-) - -Fri Aug 23 16:59:51 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * bin/svcconf-convert.pl: Fixed regex to handle empty option - strings in static/dynamic entries correctly. - -Fri Aug 23 16:25:47 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ace/OS.i: - * ace/OS.h: - - Removed the wrappers for functions that are known to be - overloaded by macros. #undef will break user code for these - functions. This includes feof(), ferror() and ungetc(). The - other functions seem to only have a function implementation. - - * ACEXML/common/FileCharStream.cpp: - - Removed the use of ACE_OS wrappers for the above functions. - -Fri Aug 23 08:40:38 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * ace/config-qnx-rtp.h: Added some new #defines so ACE will - compile on QNX 6.2. Thanks to Leen Van Kampen - <leen.van.kampen@nl.vanderlande.com> for reporting this. - - * ace/Task.cpp (svc_run): Added a #ifdef for _MSC_VER <= 1200 - since there are complaints about reinterpret_cast from int to - unsigned long. Thanks to Espen Harlinn - <espen.harlinn@seamos.no> for this fix. - -Fri Aug 23 13:04:34 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/Service_Config.cpp: Some minor cosmetic changes. - -Fri Aug 23 07:16:13 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/templates/vc6dsp.mpd: - - Fixed a problem where the pch header files is not in the list of - header files. - -Fri Aug 23 06:07:38 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ace/OS.h (ACE_OS): - - ungetc() is a macro under Borland. Undefine it to make Borland - compiler happy. - -Thu Aug 22 21:27:56 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/OS_String.cpp (strtok_r_emulation): Applied the patch Frank - Kuhlman <kuhlmann@materna.de> submitted in [Bug 1272] which - resolved an overrun error. - -Thu Aug 22 21:20:01 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * include/makeinclude/rules.local.GNU (realclean.local): - - Remove older cruft when doing a realclean. Note this is a - sledgehammer (libfoo.so*). Don't blame me if you shoot yourself - in the feet, which technically shouldn't happen as long as you - have the sources to the libraries that you built. Thanks to - Ossama for pointing out that cruft was left behind. - -Thu Aug 22 19:14:44 2002 Ossama Othman <ossama@uci.edu> - - * ace/OS.i (fgetwc, ungetwc): - - These functions are supposed to return WEOF on error, not -1 as - their ACE_OS wrappers were doing. Corrects a warning about - initialization of a negative value to "wint_t", which is an - unsigned integer on some platforms (e.g. GNU libc). - -Thu Aug 15 10:43:51 2002 Steve Huston <shuston@riverace.com> - - * ace/Message_Queue_T.cpp (enqueue_tail): Moved notify() call outside - lock scope, as Sun May 5 19:14:34 2002 Douglas C. Schmidt - suggested. - -Wed Aug 14 18:14:57 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.cpp (ACE_TSS_Cleanup::remove()): Always mark the key no - longer in use, instead of simply decrementing the reference count. - Shouldn't call remove() until there are no more references to - the key. - (ACE_TSS_Cleanup::detach()): Decrement the reference count when - detaching. Previously, the count could never get to 0 and the - remove() method would never be called until object manager shutdown. - This resolves a problem where a service using - ACE_Unmanaged_TSS_Singleton would do its cleanup(), but there would - still be a reference to the key... if the destructor (in an unloaded - service DLL) was called at process shutdown, it would crash. - Thanks to Craig L. Ching <cching@mqsoftware.com> for pointing this - problem area out. - -Fri Aug 9 16:11:39 2002 Steve Huston <shuston@riverace.com> - - * ace/Message_Block.cpp: - (ACE_Data_Block::ACE_Data_Block): If the data allocation fails, - reset the cur_size_ and max_size_ to 0 to reflect the fact that - there is no legitimate memory allocated. - (ACE_Message_Block::init_i): If the data block size() that results - from allocating a new ACE_Data_Block is not what was requested, - return -1 to indicate the failure. - - * ace/Malloc_T.i (ACE_Cached_Allocator::free): Don't attempt to - return a 0 pointer to the free_list_. Thanks to Laxmikant - Bopalkar <LBopalkar@protrader.com> for this fix. - - * THANKS: Added Laxmikant Bopalkar to the Hall of Fame. - -Thu Aug 22 17:34:18 2002 Steve Huston <shuston@riverace.com> - - * ace/FIFO_Recv_Msg.{h i}: Fixed recv() in the non- - ACE_HAS_STREAM_PIPES case to throw away the part of a - message longer than the max number of bytes requested by - the caller. Filled out the doxygen comments in the header - to explain the behavior, and the differences when (not) - using ACE_HAS_STREAM_PIPES. - -Wed Aug 21 17:48:29 2002 Steve Huston <shuston@riverace.com> - - * examples/C++NPv2/TP_Logging_Server.cpp (TP_Logging_Task): - * examples/C++NPv2/display_logfile.cpp (svc): - Must use reinterpret_cast, not static_cast, to cast between - unrelated types (char *, class *). - - * examples/C++NPv2/Makefile: Added AIO_CLD to top-level build. - -Thu Aug 22 18:10:46 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * bin/nightlybuilds/builds.lst: Move RedHat_Static build - to new scoreboard. - -Thu Aug 22 01:32:46 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ACEXML/common/StreamFactory.cpp: - - Use ACE_TEXT for literals. Add destructor to avoid linker - problems. - - * ACEXML/parser/parser/Parser.cpp: - - Use ACEXML_Char instead of char* to fix compilation with WCHAR. - - * ACEXML/common/XML_Common.dsp: - - Add files StreamFactory.{h,cpp} to the project files. - -Thu Aug 22 01:16:16 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ace/OS.h: - - Fixed compilation error under Windows by undefining the macros. - -Wed Aug 21 22:22:42 2002 Ossama Othman <ossama@uci.edu> - - * bin/msvc_auto_compile.pl (Build_Core): - - Removed "LoadBalancing_Static.dsp" from the list of static - library projects to build. That MSVC++ project was for the old - load balancer, and no longer exists. - -Thu Aug 22 03:50:02 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * bin/performance_stats.sh: - * bin/footprint_stats.sh: Calculate value of DATE - inside these scripts instead of passing it in as the - third parameter. This will make it easier to make - an autobuild script around this. - -Wed Aug 21 20:40:20 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ace/OS.h: - * ace/OS.i: - - Added wrappers for fgetc, fgetwc, ferror, clearerr, feof, - ungetc, ungetwc. All these are guaranteed to not be macros. - - * ACEXML/apps/svcconf/Svcconf_Handler.cpp: - * ACEXML/apps/svcconf/Svcconf_Handler.h: - - Fixed compilation errors. - - * ACEXML/common/FileCharStream.cpp: - * ACEXML/common/HttpCharStream.cpp: - * ACEXML/common/XMLFilterImpl.cpp: - * ACEXML/examples/SAXPrint/Print_Handler.cpp: - * ACEXML/examples/SAXPrint/SAXPrint_Handler.cpp: - * ACEXML/examples/SAXPrint/main.cpp: - * ACEXML/examples/svcconf/Svcconf_Handler.cpp: - * ACEXML/parser/debug_validator/Debug_Attributes_Builder.cpp: - * ACEXML/parser/debug_validator/Debug_DTD_Manager.cpp: - * ACEXML/parser/debug_validator/Debug_Element_Builder.cpp: - * ACEXML/parser/debug_validator/Element_Tree.cpp: - * ACEXML/parser/parser/Parser.cpp: - - Use ACE_TEXT instead of ACE_LIB_TEXT. It results in reducing - vertical space and the wchar.txt document also encourages it. - -Wed Aug 21 17:02:44 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/OS_String.cpp: Changed to enabled wchar version of - strtok_r_emulation when both ACE_HAS_WCHAR and ACE_LACKS_WCSTOK - are both defined (instead of ACE_HAS_REENTRANT_FUNCTIONS.) - Thanks to Brian Appel <bappel@integ.com> for reporting this. - -Wed Aug 21 12:41:51 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/config/notifytest.mpb: - * tests/SSL/tests.mpc: - - Fixed a few problems with some of the mpc files. - - * nightly.mwc: - - Added an initial nightly build workspace file. - -Wed Aug 21 11:57:37 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * bin/msvc_auto_compile.pl: Added ACEXML/apps to the list of - compiled ACE directories. - -Wed Aug 21 11:15:11 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/NMakeWorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/TemplateParser.pm: - * bin/MakeProjectCreator/templates/nmake.mpd: - - Fixed a problem dealing with Windows, NMake and idl files. - -Wed Aug 21 11:04:09 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/Object_Manager.cpp (init): Changed to return failure (-1) if - ACE_Object_Manager fail to acquire the TSS ACE_Log_Msg instance - for whatever reason. Thanks to Jeremy Altavilla - <jeremya@artifact-entertainment.com> for suggesting the fix. - -Wed Aug 21 07:43:12 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/ATM_Addr.{h,i}: - Made the signature definition and implementation of set_selector() - and get_selector() the same. - - * ace/CDR_Stream.cpp: - Made the read_string() signature the same as in the definition - - * ace/config-doxygen.h: - Added ACE_HAS_XTI_ATM to the list of defines that must be set - when generating the doxygen documentation - - * ace/config-win32-mingw.h: - Added extern "C" around the win32api methods ace uses but which - are not defined in the mingw win32api. - - * ace/Lib_Find.cpp: - Made signature of ldfind method the same as in the definition - - * ace/Log_Record.cpp: - Made signature of print method the same as in the definition - - * ace/Multiplexor.i: - Added #ifdef 0/#endif around the contents of this file, because the - header and cpp file also do this. Doxygen parses now this .i file - and gives a lot of warnings about this file because it cannot find - the definition of the methods in this file. - - * ace/Configuration.h: - * ace/Containers_T.cpp: - * ace/Lib_Find.h: - * ace/OS.h: - * ace/RMCast/RMCast_Reliable_Factory.h: - * ace/RMCast/RMCast_UDP_Proxy.h: - Small improvements for doxygen documentation. - - * include/makeinclude/ace_flags.bor: - Added cflags and lib for TAO PSDL library - - * include/makeinclude/platform_mingw32.GNU: - MingW doesn't support the adding of the version number after - the .dll, so set SONAME and SOVERSION to empty. - - * examples/Connection/non_blocking/test_lsock_acceptor.cpp: - * examples/Connection/non_blocking/test_lsock_connector.cpp: - Added include of Log_Msg to fix compile errors with MingW - -Tue Aug 20 14:19:55 2002 Jaiganesh@doc <bugzilla@cs.wustl.edu> - - * ACE version 5.2.4 released. - -Fri Aug 16 10:25:07 2002 Ossama Othman <ossama@uci.edu> - - * include/makeinclude/ace_flags.bor: - - Corrected CosLoadBalancing library related variables. They - were still setup for use with the old load balancer prototype. - -Fri Aug 16 11:15:49 2002 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * examples/Reactor/Proactor/test_aiosig_ace.cpp: Added - #include "ace/Log_Msg.h" so things will compile. Thanks to - Victor Poznyak <VPoznyak@Webley.COM> for reporting this and - Alexander Libman <alexl@aurema.com> for suggesting the fix. - -Thu Aug 15 15:22:49 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * tests/tests.dsw: Re-added Thread_Manager_Test.dsp into this - workspace. It got accidentally removed when someone merged a - branch into the main trunk. Thanks to Ronald Berger - <ronald.berger@wpafb.af.mil> for noticing this. - -Mon Aug 12 10:17:55 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/Thread_Manager.h (ACE_Thread_Manager): Fixed a type where - "manipulated" was spelled "maniputated". Thanks to - Eric Newton <ecn@metaslash.com> for reporting this. - - * PROBLEM-REPORT-FORM: Fixed a typo where "ASCII text" was spelled - "ASCII test". Thanks to Eric Newton <ecn@metaslash.com> for - reporting this. - -Fri Aug 9 12:13:21 2002 Jaiganesh Balasubramanian <jai@doc.ece.uci.edu> - - * bin/nightlybuilds/builds.1st: - Moved the Debian_Core builds to the new scoreboard. - Thanks to CRaig Rodrigues <crodrigu@bbn.com> for the fix. - -Fri Aug 9 09:24:08 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/Time_Value.{inl,cpp}: Make sure to normalize the set() methods... - Thanks to Francois <francois@vibes.net> for reporting this. This - fixes BugID 1275. - -Thu Aug 8 23:11:30 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * include/makeinclude/rules.nested.GNU: Reverted this change "Thu - Aug 8 18:28:34 UTC 2002 Craig Rodrigues <crodrigu@bbn.com>" - since it would break our daily builds. - -Thu Aug 8 18:28:34 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * include/makeinclude/rules.nested.GNU: If we fail during a recursive - make, exit with error code 1. This makes it easier to detect errors - in automated builds. Thanks to Gary Duzan <gduzan@bbn.com> for showing - inspiring the fix. - -Wed Aug 7 12:39:31 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/Driver.pm: - - Speed up execution by changing the include path order. - - * bin/MakeProjectCreator/modules/GNUWorkspaceCreator.pm: - - Don't generate a for loop if there is only one project. - - * bin/MakeProjectCreator/modules/ProjectCreator.pm: - - Fixed a bug with the global assignments overwriting project - assignments. - -Sun Aug 4 16:47:29 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * tests/Framework_Component_DLL.cpp: Added a missing template - instantiation. It was caught only in SunCC53 builds. - -Sun Aug 4 16:01:28 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * bin/auto_run_tests.lst: Do not the Stack_Recursion tests in - single threaded builds. - -Sat Aug 03 17:57:33 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * ace/ace_dll.dsp: Removed the Hash_Map_Manager_rt.cpp from the - project file. - -Sat Aug 03 15:42:49 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * examples/Log_Msg/Log_Msg_MFC/Log_Msg_MFC.dsp: Made sure that it - builds only in MFC debug builds. - -Sat Aug 3 14:56:23 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ACEXML/common/HttpCharStream.cpp: - - Fixed a conversion error from char to wchar_t. Thanks to Johnny - for pointing this out. - - * ACEXML/parser/parser/Parser.cpp: - - Fixed an unreachable statement warning. Thanks to Bala for - pointing this out. - -Sat Aug 3 19:53:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ACEXML/common/XML_Codecs.h: - * ACEXML/parser/parser/Parser.h: - * ace/Registry.h: - * ace/String_Base.h: - * ace/QoS/QoS_Session_Factory.h: - Fixed doxygen commands. - -Sat Aug 3 12:36:55 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * ace/Hash_Map_Manager_rt.h: - * ace/Hash_Map_Manager_rt.cpp: - * ace/Hash_Map_Manager_rt.i: Removed these files from the main - trunk. They shouldnt have been there in the first place. - -Fri Aug 2 22:00:26 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * examples/Log_Msg/Log_Msg_MFC/*: Added example to show how - to send ACE_DEBUG output to the Microsoft Visual Studio debugger console. - Thanks to Don Hinton <dhinton@ieee.org> and - Chris Hafey <chafey@stentor.com> for help with this. - -Fri Aug 2 13:10:05 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ACEXML/examples/SAXPrint/main.cpp (ACE_TMAIN): - * ACEXML/common/InputSource.cpp: - - Reverted the previous check-in with respect to the delete - statement. Memory is getting leaked somewhere down in ACE. - -Fri Aug 2 12:31:34 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ACEXML/common/ContentHandler.h: - * ACEXML/parser/parser/Entity_Manager.cpp: - * ACEXML/common/NamespaceSupport.cpp: - - Fixed minor typos in the comments. - - * ACEXML/common/InputSource.cpp: - - Removed an erroneous delete statement. Not sure if this is a - implementation bug or a thinko. - - * ACEXML/common/HttpCharStream.cpp: - * ACEXML/common/Mem_Map_Stream.cpp: - * ACEXML/common/Mem_Map_Stream.h: - - Use char* when sending data on the network. Technically it - should be byte. - - * ACEXML/examples/SAXPrint/Print_Handler.cpp: - * ACEXML/examples/SAXPrint/main.cpp: - - Fixed a bug with getopt() when WCHAR is enabled. Since EOF is - -1, never use a WCHAR in the while loop. - - * ACEXML/parser/parser/Parser.cpp: - * ACEXML/parser/parser/Parser.h: - - Improved Namespace support. Modularized functions in the parser. - - * ACEXML/docs/TODO.txt: - * ACEXML/docs/bugs.txt: - * ACEXML/docs/guidelines.txt: - * ACEXML/docs/parser_features.txt: - - Updated to reflect the current status. People were reading these - old docs and assuming a lot of the features aren't implemented. - -Fri Aug 2 16:06:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/config-doxygen.h: - Added ACE_HAS_ATM define to also generate the doxygen documentation - of the ACE ATM classes. This is one of the last parts of ACE that - are missing in the doxygen documentation. - -Fri Aug 2 07:54:24 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * ace/Malloc.cpp: Rearranged member initialization list in the - constructor of ACE_Malloc_Stats, to fix warnings with explicit - template builds. - -Thu Aug 01 13:40:59 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * tests/Malloc_Test.cpp: Changed the usage of ASYS_TEXT to - ACE_TEXT. Should fix most of the compile errors with g++. - -Thu Aug 1 12:14:21 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/templates/gnu.mpd: - - Added $(ACELIB) to the list of ACE_SHLIBS. - -Thu Aug 1 11:13:20 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * ace/Malloc_Allocator.h: - * ace/Malloc_Allocator.cpp: Removed the declaration and definition - of ACE_Malloc_Stats since they are not needed in this - translation unit. - - * ace/Malloc.h: - * ace/Malloc.cpp: Retained the definition if ACE_Malloc_Stats here - (yes, there were two declarations) and added the definition and - template instantiation for ACE_Atomic_Op used by - ACE_Alloc_Stats. This should fix the erros showing up on the - scoreboard explicit template builds which has - ACE_HAS_MALLOC_STATS turned on. - -Thu Aug 1 09:08:36 2002 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * ace/Message_Block.h: Made the non-static duplicate() and release() - methods virtual so they can be overridden by subclasses. Thanks - to Brian Raven <brianr@liffe.com> for this suggestion. - -Thu Aug 01 15:24:32 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/OS_Thread_Adapter.h: - Fixed incorrected doxygen @file command. - -Thu Aug 01 14:52:44 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * bin/fuzz.pl: - Improved the check for the @file doxygen command. Incomplete file - names where not detected as error. - -Thu Aug 1 07:03:05 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/README: - - Added a bit of background on the tool before plunging into the - details. - -Thu Aug 01 07:59:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/Asynch_Acceptor.h: - * ace/XtReactor.cpp: - * ace/XtReactor.h: - Corrected argument names to fix doxygen warnings. - -Thu Aug 01 07:34:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/config-win32-mingw.h: - Made the definition of the missing parts of the Win32api of MingW - dependent on the version of the w32api instead of our own defines. - This way we automically define the missing things when needed. Thanks - to Craig for suggesting this. - -Wed Jul 31 23:39:27 UTC 2002 Craig Rodrigues <crodrigu@bbn.com> - - * etc/*.doxygen: Change DOT_PATH from /usr/local/bin/ to - empty, forcing doxygen to look for the dot program in PATH. - This eliminates warnings in the Doxygen nightly build. - -Wed Jul 31 12:23:47 2002 Steve Huston <shuston@riverace.com> - - * ace/Malloc.cpp: - * ace/Malloc_T.cpp: - * tests/Malloc_Test.cpp: Changed logging of pointers to %@ - - * tests/Mem_Map_Test.cpp: Fixed loop bound error. - -Wed Jul 31 11:34:58 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * docs/index.html: Added a link to ACEXML document. - -Wed Jul 31 10:47:54 2002 Chad Elliott <elliott_c@ociweb.com> - - * include/makeinclude/rules.local.GNU: - - Added a gnu macro (DEPEND_CMD_ARGS) to the g++dep line to allow - pasing arguments (such as -R) to g++dep when running "make depend" - -Wed Jul 31 08:41:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/config-win32-mingw.h: - For the compilation of Win32_Asynch_IO we now need to have - FILE_SEGMENT_ELEMENT from winnt.h and ReadFileScatter and - WriteFileScatter from winbase.h. - These are not in the header files of the MingW w32api 1.4/1.5 - release. That's why we do these union/defines also in this header - file. By adding the defines ACE_MINGW_LACKS_READWRITE_FILESCATTER - and ACE_MINGW_LACKS_FILE_SEGMENT_ELEMENT to your config.h file - these defines/union are done by ACE for MingW32 so that everything - will compile. - -Wed Jul 31 08:11:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * etc/Task.cpp: - Also the MinGW32 GCC3.1 compiler has problems with the - reinterpret_cast in the svc_run method. Use the static_cast instead. - -Wed Jul 31 07:40:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * etc/*.doxygen: - Enabled warnings to get them in the doxygen scoreboard. - - * ace/Configuration.cpp: - Fixed method argument names to fix doxygen warnings. - -Tue Jul 30 15:45:23 2002 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * ace/SOCK_Dgram.h: Fixed the comment for send(), which had a - cut and paste error. Thanks to Jeff Wilson - <wilsonj@cs.ecs.baylor.edu> for reporting this. - -Tue Jul 30 14:09:35 2002 Chad Elliott <elliott_c@ociweb.com> - - * tests/tests.mpc: - - Update the mpc file to include the libraries. - -Tue Jul 30 09:07:29 2002 Ossama Othman <ossama@uci.edu> - - * ace/Parse_Node.h: - * ace/Parse_Node.cpp: - - Unexported all classes in this file. They are only meant for - internal use by ACE's Service Configurator. Updated Doxygen - comments accordingly. - - (apply, symbol, open_dll): - - These methods now accept a "yyerrno" reference argument to allow - for modification of the error count in a reentrant manner. - - * ace/Svc_Conf.h: - - Removed "ace_yyerrno" and "ace_yylineno" global variables. - Local ones are now used. - - Altered signature of ace_yyerror() function to accept two - additional parameters, yyerrno and yylineno. Addresses a race - condition. - - * ace/Svc_Conf.y: - - Removed all remnants of the global "yyerrno" and "yylineno" - variables. Only the reentrant versions are used now. - - Pass the "yyerrno" variable as a parameter to all functions that - read or modify the parse error count. Addresses several race - conditions. - - Fixed call to ACE_Location_Node::handle(). This method no - longer exists. The correct call is ACE_Location_Node::dll(). - For some reason, Svc_Conf_y.cpp (a generated file!!!) was - "fixed" but the parser grammar file was never fixed. - - * ace/Svc_Conf.l: - - Removed all remnants of the global "yylineno" variable. Only - the reentrant version is used now. - - * ace/Service_Config.cpp (process_directives_i): - - Removed initialization of "ace_yyerrno" and "ace_yylineno" - global variables. They no longer exist. Poof! - - Reverted my previous change. The above changes obviate the need - to take into account the global parse error count. - - * ace/Makefile: - - Tweaked Svc_Conf_y.cpp and Svc_Conf_l.cpp generation rules so - that yyerrno and yylineno are not prepended with an "ace_". - That is no longer needed. - - * etc/Svc_Conf_y.cpp.diff: - - Regenerated this patch against parsers generated by the latest - GNU Bison (1.35). - - * etc/Svc_Conf_l.cpp.diff: - - Regenerated this patch to pull in new changes to Svc_Conf_l.cpp. - -Tue Jul 30 10:06:04 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/Thread_Manager.cpp: Moved a stray colon until *after* - the #if !defined(ACE_USE_ONE_SHOT_AT_THREAD_EXIT) check. Thanks - to Alan Tanga <stang1@lucent.com> for reporting this. - - * ace/POSIX_Proactor.cpp (delete_result_aiocb_list): Moved - the definition of size_t ai out of the for loops to avoid - problems on older C++ compilers. Thanks to George Varsamis - <gv@prismtechnologies.com> for reporting this. - -Tue Jul 30 08:14:34 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/Creator.pm: - - Added an error check for invalid inheritance syntax. - - * bin/MakeProjectCreator/modules/Driver.pm: - - Changed the progress indicator to allow for cleaner output - when an error occurs. - - * bin/MakeProjectCreator/modules/ProjectCreator.pm: - - Remove duplicate libs, libpaths and includes when addition or - subtraction is involved. - - * bin/MakeProjectCreator/templates/gnu.mpd: - - Cleaned up the template to make generated Makefiles a little more - consistent. - -Tue Jul 30 06:50:26 2002 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * ace/Dynamic_Service.cpp: Moved ACE_Dynamic_Service::instance() - from the .i file into the .cpp file and added a #include of - "ace/Service_Object.h" to the .cpp file so the complete type of - ACE_Service_Object would be available to the HP/UX C++ compiler. - Thanks to Torsten Kuepper <Torsten.Kuepper@nokia.com> for - reporting this. - -Tue Jul 30 07:51:12 UTC 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * etc/ace.doxygen: - Enabled warnings and undocument warnings to get these warnings - into the doxygen scoreboard. When this works I will enable the - warnings in all other .doxygen files later. - -Mon Jul 29 14:47:46 2002 Ossama Othman <ossama@uci.edu> - - * ace/Service_Config.cpp (process_directives_i): - - In addition to the thread-safe yyerrno variable, check the - global ace_yyerrno variable when determining if errors have - occurred when processing Service Configurator directives. - Thanks to Doug for tracking down the problem and suggesting a - fix and to Denginere <denginere@hotmail.com> for reporting this. - -Mon Jul 29 17:30:27 2002 Steve Huston <shuston@riverace.com> - - * ace/Get_Opt.h (ctor): Changed documented name of the skip_argv0 - argument to skip_args to better reflect that it's not just - argv[0] that can be skipped. - -Mon Jul 29 07:45:12 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ace/OS.cpp (uname): Only check for PROCESSOR_ARCHITECTURE_IA64 - when this is defined. The MingW w32api doesn't define this constant. - -Sun Jul 28 17:17:50 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * docs/ACE-subsets.html: Updated the breakdown of files to - each subset component so it's consistent with the - $ACE_ROOT/Makefile. Thanks to Ruslan Zasukhin - <sunshine@public.kherson.ua> for reporting this. - -Sun Jul 28 10:31:12 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * include/makeinclude/platform_gnuwin32_common.GNU: - Use -mcpu=$(TCPU) instead of -m$(TCPU), because MingW GCC 3.1 - now gives warnings that -m is deprecated. Thanks to Craig Rodrigues - for confirming that -m was already deprecated in GCC 2.95.3 - -Sat Jul 27 17:35:06 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * include/makeinclude/platform_g++_common.GNU (SOFLAGS): - - Allow building of shared libraries with gcc and Solaris ld. - There wasn't any reason that it wouldn't work. - - * Changelog: Remove >>>>> left out by merge conflicts. People - should be more careful about checking for conflicts. - -Sat Jul 27 16:20:19 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * examples/Service_Configurator/Misc/main.cpp (ACE_TMAIN): Changed - the code to be more resilient against errors that occur during - parsing of the svc.conf file. Thanks to Denginere - <denginere@hotmail.com> for motivating this change. - -Sat Jul 27 15:50:12 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * ACEXML/examples/svcconf/main.cpp: - Made this file compiling with BCB and Unicode - - * ACEXML/examples/svcconf/Svcconf_Handler.h: - Made includes like the other examples so that we can build it with - BCB - - * ACEXML/examples/svcconf/Svcconf_Handler.cpp: - Fixed unused argument warnings - - * ACEXML/examples/svcconf/Makefile.bor: - Added this BCB makefile - - * ACEXML/examples/Makefile.bor - Added svcconf directory - -Fri Jul 26 21:00:32 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * ace/TP_Reactor.cpp: Fixed a unused variable warning in g++ - builds. - -Fri Jul 26 20:08:41 2002 Steve Huston <shuston@riverace.com> - - * include/makeinclude/platform_aix4_g++.GNU: Corrected option for - using threads from -mthreads to -pthread. This may not work with - pre-gcc 3 compilers, but those compilers are busted on AIX anyway. - Added support for buildbits=32 and buildbits=64. Moved in the new - AIX library-building setup from platform_aix_ibm.GNU. - -Fri Jul 26 15:52:12 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * ace/TP_Reactor.cpp: Fixed a problem with handle_event - (ACE_Time_Value &). It should be calling the TP_Reactor's - handle_event () instead of Select_Reactor's. This should fix - [BUGID 1257]. - -Fri Jul 26 10:11:41 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/config/messaging.mpb: - * bin/MakeProjectCreator/config/messaging_only.mpb: - - Set up a new base project that just contains information about - the Messaging library. - - * bin/MakeProjectCreator/modules/GNUProjectCreator.pm: - * bin/MakeProjectCreator/templates/gnu.mpd: - - Allow the file lists to contain names of files that correspond - to relative directories. - -Fri Jul 26 06:40:12 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/OS_String.inl: Added casts of the input parameter to - (unsigned char) for the ace_isprint() and ace_isspace() methods - on ACE_OS_String to prevent problems with MSVC++ 7.0. Thanks to - Colin Weaver <colin.weaver@invensys.com> for suggesting this - fix. - -Fri Jul 26 10:37:12 2002 Johnny Willemsen <jwillemsen@remedy.nl> - - * etc/tao_pss.doxygen: - Added doxygen config file to generate the new TAO PSS. - - * bin/generate_doxygen_pl: - Added tao_pss.doxygen to the list of files to generate. - - * htlm/index.html: - Added link for the new PPS doxygen documentation. - -Thu Jul 25 22:58:24 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * ace/Synch.cpp: Removed explicit instantiation of - ACE_Condition<ACE_Recursive_Thread_Mutex> since a template - specialization exists. This was causing a compile error when - explicit template instantiations were used with - g++-2.95.3. - -Thu Jul 25 11:32:41 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/README: - - Added a little more detail about what the depends keyword refers - to. - - * bin/MakeProjectCreator/USAGE: - * bin/MakeProjectCreator/modules/Driver.pm: - - Changed the code to show the default type in the usage message. - - * bin/MakeProjectCreator/modules/GNUWorkspaceCreator.pm: - - Change the workspace to use a for loop when building on UNIX. - - * tests/RMCast/tests.mpc: - * tests/SSL/tests.mpc: - - Fixed the case on the dependencies lines. - -Wed Jul 24 13:38:40 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/Creator.pm: - * bin/MakeProjectCreator/modules/Driver.pm: - * bin/MakeProjectCreator/modules/ProjectCreator.pm: - * bin/MakeProjectCreator/modules/WorkspaceCreator.pm: - - Added a progress indicator (for those big workspaces). - -Wed Jul 24 13:10:19 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/ProjectCreator.pm: - - Add README and readme to the default documentation - files. - - * bin/MakeProjectCreator/modules/Creator.pm: - * bin/MakeProjectCreator/modules/GUID.pm: - * bin/MakeProjectCreator/modules/VC7ProjectCreator.pm: - - Change the GUID generation code to be deterministic (but still - unique) to allow for multiple solution files to contain the same - project files without having to regenerate the solutions. - -Wed Jul 24 13:42:37 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.cpp (uname): Added case for PROCESSOR_ARCHITECTURE_IA64 - when decoding processor type. - -Wed Jul 24 10:58:22 2002 Steve Huston <shuston@riverace.com> - - * ace/Acceptor.cpp (ACE_Acceptor::handle_input): When calling select(), - use a int value instead of casting a ACE_HANDLE to an int. - - * tests/Cached_Accept_Conn_Test.cpp: Use %@ instead of %d when - logging pointers. - - * tests/MT_Reactor_Upcall_Test.cpp: Added ACE_TEXT to char strings - where needed. Changed the send to send fixed header, then the - variable-size string. Sending all at once doesn't work if the - string doesn't immediately follow the size_ member in memory, - which happened on Win64. - -Wed Jul 24 10:07:24 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/README: - * bin/MakeProjectCreator/USAGE: - - Update the documentation to reflect the current state. - -Wed Jul 24 08:13:48 2002 Christopher Kohlhoff <chris@kohlhoff.com> - - * include/makeinclude/build_dll.bor: - * include/makeinclude/build_exe.bor: - * include/makeinclude/build_lib.bor: - * include/makeinclude/make_flags.bor: - Improve debugging support with Borland C++Builder by invoking the - compiler with a full path to each source file. This can be disabled - by setting a NO_FULL_PATH environment variable to 1 or by passing - -DNO_FULL_PATH when running make, which may be necessary when doing - a debug build on Win9x. - -Wed Jul 24 07:45:43 2002 Chad Elliott <elliott_c@ociweb.com> - - * THANKS: Added Tim Bradley to the hall of fame. - -Tue Jul 23 21:46:54 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/Process_Mutex.h: - * ace/Process_Mutex.cpp: Disallowed anonymous Process_Mutex by - invoking this->unique_name() to create a temporary name if - there's no mutex name passed to the constructor. Thanks John - Michael Zorko <j.zorko@att.net> for reporting this problem on - Mac OSX and submitting the patch. - -Tue Jul 23 13:08:27 2002 Steve Huston <shuston@riverace.com> - - * ace/Log_Msg.{h cpp} (log_hexdump): - * ace/ACE.{h cpp} (format_hexdump): Change size arguments from int - to size_t. Change format_hexdump()'s return from int to size_t. - -Tue Jul 23 09:48:15 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/Driver.pm: - - Always add the default include paths. - - * bin/MakeProjectCreator/modules/ProjectCreator.pm: - - Make sure that idl generated source files go at the front of the - file list in auto-generates source lists. - - * bin/MakeProjectCreator/templates/gnu.mpd: - - Only add LIB and SHLIB assignments if we are generating a Makefile - for libraries. - -Mon Jul 22 22:30:53 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ACEXML/common/XML_Common.dsp: Added the macro definition - ACEXML_HAS_DLL=1 to the release configuration to resolve a - nightly biuld problem. - -Mon Jul 22 18:28:48 2002 Steve Huston <shuston@riverace.com> - - * ace/Configuration.h: Doxygen fixes. - -Mon Jul 22 16:57:19 2002 Chris Cleeland <cleeland_c@ociweb.com> - - * bin/g++dep: Added a -R option that allows you to specify a - variable which should be used to generate relative paths if it's - defined. - -Mon Jul 22 17:45:48 2002 Steve Huston <shuston@riverace.com> - - * ace/Get_Opt.h: Doxygen fixes. - -Mon Jul 22 16:39:21 2002 Steve Huston <shuston@riverace.com> - - * ace/Acceptor.cpp (info): - * ace/Connector.cpp (info): Cast return val to int. - - * ace/Message_Queue.h (message_count): - * ace/Message_Queue_T.{h cpp i} (message_count): Change return value - from size_t to int. Changed cur_count_ member (and accessors that - return it) from size_t to int. This is a compromise to unify the - length of arguments without changing the return types of all - methods that return a message count (all the enqueue/dequeue methods, - as well as a few others). - Changed a number of @arg to @param for doxygen. - - * ace/SOCK_IO.{h i} (recvv, sendv): - * ace/SOCK_Stream.{h i}(recvv_n, sendv_n): Changed iovcnt arg from - size_t to int. int, though sort of odd as a choice, is the - UNIX98 standard and more commonly implemented. Removes an arg - size mismatch between the SOCK_IO and ACE layers, restricting the - iovcnt range at the SOCK_IO layers to match the layer below. - Changed the recv and send iov-enabled, deprecated methods to cast - their size_t iovcnt args to int... these deprecated methods should - be removed after 5.3 is out. - - * tests/Conn_Test.cpp: Replaced %d with %@ when logging pointers. - Added missing string arg for a ACE_ERROR %p directive. Use a - int select_width instead of casting ACE_HANDLE for Win64. - -Mon Jul 22 15:13:38 2002 Carlos O'Ryan <coryan@atdesk.com> - - * ace/Atomic_Op.h: - * ace/Atomic_Op.i: - Fixed bug 1253, add missing copy constructor. - -Mon Jul 22 13:25:29 2002 Stephen Torri <storri@cs.wustl.edu> - - * ace/streams.h: Fixed _MSC_VER compiler warning messages stating - concern about using MSC_VER without assurance that the macro - was defined. - -Mon Jul 22 12:20:21 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * THANKS: Added Sean Ogle to the hall of fame. - -Mon Jul 22 09:19:34 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * THANKS: Aded Kew Whitney and Colin Weaver to the hall of fame. - -Sat Jul 20 08:53:05 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/SOCK_Dgram.cpp (set_nic): Moved the #endif for - ACE_WIN32 *ahead* of the set_options() call. Thanks to Joe - Hayes <Joe.Hayes@lmco.com> for reporting this. - -Fri Jul 19 14:42:37 2002 Steve Huston <shuston@riverace.com> - - * ace/WFMO_Reactor.cpp (make_changes_in_suspension_infos, - make_changes_in_current_infos): - Fixed bug resulting from: Wed Jul 17 11:34:28 2002 Steve Huston - -Thu Jul 18 15:48:29 2002 Steve Huston <shuston@riverace.com> - - * ace/Future.h: Doxygenation improvements. - -Thu Jul 18 08:39:08 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/README: - - Update the documentation to reflect the additions to the grammar. - - * bin/MakeProjectCreator/config/avstreamsexe.mpb: - * bin/MakeProjectCreator/config/corba_messaging.mpb: - * bin/MakeProjectCreator/config/minimum_corba.mpb: - * bin/MakeProjectCreator/config/namingexe.mpb: - * bin/MakeProjectCreator/config/notify.mpb: - * bin/MakeProjectCreator/config/orbsvcsexe.mpb: - * bin/MakeProjectCreator/config/orbsvcslib.mpb: - * bin/MakeProjectCreator/config/rt_client.mpb: - * bin/MakeProjectCreator/config/rt_server.mpb: - * bin/MakeProjectCreator/config/rteventexe.mpb: - - Add the GNU specific info for generating conditional target - Makefiles. - - * bin/MakeProjectCreator/modules/GNUProjectCreator.pm: - - Added code to generate the conditional sections of the GNU - Makefiles. - - * bin/MakeProjectCreator/modules/ProjectCreator.pm: - * bin/MakeProjectCreator/modules/TemplateParser.pm: - - Fix some minor problems with generated project files. - - * bin/MakeProjectCreator/templates/gnu.mpd: - - Added template info for conditional targets. - - * bin/MakeProjectCreator/templates/vc6dsp.mpd: - * bin/MakeProjectCreator/templates/vc7.mpd: - - Removed the tie class idl targets. - - * bin/ace_components: - - Added a new append option that will be used by the generated GNU - Makefiles. - -Thu Jul 18 06:17:19 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * Doug turns 40 today. Fortunately, 40 Octal is only 32 Decimal ;-). - -Wed Jul 17 15:18:31 2002 Steve Huston <shuston@riverace.com> - - * ace/Configuration.h: Doxygen improvements. - - * ace/Configuration.cpp: All error cases where a return - (-2, -3, -4, -5) occurred now return -1 with errno set. - - * ace/OS.{i cpp}: Follow-up to Wed Jul 17 11:34:28 2002 Steve Huston - Only make ACE_WIN64 use intptr_t - must have a new SDK to get - intptr_t, and not all have it.... so use it only for ACE_WIN64 - since the compiler comes in the newer SDK... Win32 does - fine with long, which is what an intrptr_t is on Win32. - - Also fixed the PACE-busted conditional compiles around siginfo_t - to match the declaration in OS.h. Win32 needs this stuff. - -Wed Jul 17 14:25:39 2002 Steve Huston <shuston@riverace.com> - - * include/makeinclude/platform_aix_ibm.GNU: Added -qfuncsect to the - compile options. Allows for smaller executables/libraries, and - gets rid of most of the multiple definition warnings for templates. - Unfortunately, there are still a few stragglers so the aix_shr - post-processing script is still required. - - * ace/Sock_Connect.cpp: Added a hack to get Visual Age C++ on AIX - build running clean. Have to include a template instantiation we - don't need to get the compiler to pull the proper header files - at template compilation time. - -Wed Jul 17 11:34:28 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.{i cpp}: Windows, in calls to _open_osfhandle(), cast - handle to long for C++Builder. Only MSVC has the intptr_t type. - - * ace/WFMO_Reactor.cpp (make_changes_in_current_infos, - make_changes_in_suspension_infos): Run the index from 0 to - last_valid_slot rather than valid slot down to 0 to allow - correct use of size_t index. - - * ace/WIN32_Proactor.cpp (register_handle, post_completion): Use - a reinterpret_cast, not static_cast to cast void* to ULONG for - non-Win64. - -Wed Jul 17 03:50:56 UTC 2002 Don Hinton <dhinton@objectsciences.com> - - * ace/ACE.cpp: - * ace/OS.{h,i,cpp}: - * ace/OS_Dirent.inl: - * ace/OS_Memory.h - * ace/OS_String.inl: - * ace/Thread_Manager.cpp: - * ace/WFMO_Reactor.cpp: - * ace/config-all.h: - * ace/config-win32-borland.h: - * ace/config-win32-msvc-5.h: - * ace/config-win32-msvc-6.h: - * include/makeinclude/wrapper_macros.GNU: - - Removed PACE. - -Tue Jul 16 14:19:37 2002 Steve Huston <shuston@riverace.com> - - * ace/DLL_Manager.cpp: - * ace/Parse_Node.cpp: (also see Fri Jul 12 12:37:42 2002 Steve Huston) - Always use the ptrdiff_t trick to cast a void* to a func ptr. - Nanbor was right ;-) Thanks to Carlos O'Ryan <coryan@atdesk.com> - for clarifying this issue, including C++ chapter and verse. - (It's in ISO C++ section 5.2.10 for those interested). - - * tests/DLL_Test.cpp: Use ptrdiff_t as a mediary to cast a void * - to a func ptr, not a long. - -Mon Jul 15 21:23:47 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * ace/Message_Block.i (base): Fixed a warning in TRU 64 builds. - -Mon Jul 15 18:56:47 2002 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * ace/String_Base.cpp (substring): Changed "nil" to "nill" to - avoid conflicts with a Mac OS X macro. Thanks to Ruslan - Zasukhin <sunshine@public.kherson.ua> for reporting this. - -Mon Jul 15 15:18:33 2002 Steve Huston <shuston@riverace.com> - - * ace/Configuration.{h cpp}: Changed ACE_Configuration_Value_IntId - data_ member to be a union with both pointer and u_int members. - Also, the length_ member is now size_t, not u_int. - - * ace/Memory_Pool.cpp (ACE_Pagefile_Memory_Pool::map): Use - INVALID_HANDLE_VALUE instead of a literal 0xFFFFFFFF for call to - CreateFileMapping. - - * ace/OS.{h i}: New method, void *atop(const char *s) converts a - string to void *. - - * ace/OS.i (fdopen): Don't cast arg to _open_osfhandle - Microsoft - changed the type to something sensible. - (isatty): open a C run-time handle for the ACE_HANDLE. - - * ace/OS.cpp (fopen): Don't cast arg to _open_osfhandle - Microsoft - changed the type to something sensible. - (unique_name): Use sprintf %p for pointer rather than casting - to an int. - - * ace/Naming_Context.cpp (ACE_Name_Options::parse_args): Use - ACE_OS::atop instead of ACE_OS::atoi for converting an argument - to a pointer. - - * ace/SOCK_Dgram.cpp (recv, send): - * ace/SOCK_IO.cpp (recvv): Use an int select_width - to pass width to select() so it can be ignored on ACE_WIN64. - - * ace/SV_Semaphore_Simple.cpp (name_2_key): Disable the type cast - conversion warning casting ACE::crc32() to key_t... it's fine. - - * ace/WIN32_Proactor.cpp (register_handle, post_completion): The - I/O completion port's completionKey argument is a ULONG in - "older" compiler/SDK versions, but was changed to a - ULONG_PTR (new type in newer SDK/compiler) for 64-bit - transition. The new type is used conditionally based on whether or - not we're building a 64-bit version. - -Mon Jul 15 09:26:17 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/SString.cpp (substring): Changed "nil" to "nill" to - avoid conflicts with a Mac OS X macro. Thanks to Ruslan - Zasukhin <sunshine@public.kherson.ua> for reporting this. - -Mon Jul 15 10:14:44 2002 Carlos O'Ryan <coryan@atdesk.com> - - * bin/MakeProjectCreator/README: - Fixed typo - -Sun Jul 14 20:14:16 2002 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - - * FAQ: Fixed the use of "*.ps.gz" files. Thanks to - Carol Hunsicker <carol.hunsicker@lmco.com> for pointing this - out. - -Sun Jul 14 01:55:03 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * ace/Message_Block.i (reset_allocators): Reset the allocators in - the chain if the message block is chained. - -Sun Jul 14 01:43:32 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * ace/Message_Block.h: - * ace/Message_Block.i: Added two methods viz. reset_allocators () - and access_allocators () to ACE_Message_Block. - -Sat Jul 13 17:29:35 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * bin/fuzz.pl: Removed the word "warning" from a print statement, - since our scoreboard scripts seems to catch that as a warning. - -Sat Jul 13 13:31:54 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * bin/auto_run_tests.lst: Added the Stack_Recursion test to the - daily builds. - -Fri Jul 12 20:28:52 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ACEXML/common/URL_Addr.cpp (string_to_addr): - - Fixed confusion in handling normal and wchar strings. It should - compile now without problems. - - * ACEXML/apps/svcconf/Svcconf_Handler.cpp: - * ACEXML/apps/svcconf/Svcconf_Handler.i: - - Fixed memory leaks, some typos and used the locator to print - out error messages. - -Fri Jul 12 17:55:08 2002 Carlos O'Ryan <coryan@atdesk.com> - - * ace/String_Base.i: - Fixed rfind() to check for pos == npos also, that solved the - regression test failures. - -Fri Jul 12 16:03:22 2002 Steve Huston <shuston@riverace.com> - - * ace/Basic_Types.h: Itanium is little-endian, like x86. Although - the Itanium architecture allows big-endian operation, Windows - doesn't use it. - - * tests/test_config.h (randomize): Fixed some int/size_t mix-ups. - -Fri Jul 12 13:02:51 2002 Steve Huston <shuston@riverace.com> - - * ace/String_Base.{h i} (rfind): Changed pos argument from size_t to - ssize_t to make proper use of npos and avoid compile error for - g++ 2.95.2. - -Fri Jul 12 12:37:42 2002 Steve Huston <shuston@riverace.com> - - * ace/DLL_Manager.cpp: - * ace/Parse_Node.cpp: g++ complains that ISO C++ won't allow cast - from void * to func ptr... so cast void * to ptrdiff_t, then to - func ptr. - -Fri Jul 12 12:26:27 2002 Steve Huston <shuston@riverace.com> - - * ace/Timer_Heap_T.cpp (ctor): Alloc timer_ids_ as ssize_t[], - not long[]. - - * examples/C++NPv2/TPCLS.dsp: Added SSL libraries to link input. - - * examples/C++NPv2/TPLS.dsp: Added Logging_Handler.{h cpp} to project. - -Fri Jul 12 12:15:49 2002 Steve Huston <shuston@riverace.com> - - * ace/Timer_Heap_T.cpp (grow_heap): Fix compile warning. - -Thu Jul 11 21:09:36 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * Makefile.bor: Added a new BUILD target called "nightlybuild" to - simplify nightlybuild script. - -Thu Jul 11 20:42:52 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ace/OS.i (creat): The Win32 implementation of this function was - broken. Thanks to Marvin Greenberg <mgreenberg@dctd.saic.com> - for submitting the fix. - -Thu Jul 11 18:37:35 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ACEXML/common/URL_Addr.cpp (string_to_addr): - * ACEXML/common/URL_Addr.h: - - Change the argument to match the base class INET_Addr so that - warnings under Borland BCB is eliminated. - - * ACEXML/common/HttpCharStream.cpp: - - Use ACE_TEXT_ALWAYS_CHAR to convert an Unicode URL to char* - before converting it to an address. Thanks to Nanbor for - pointing this out. - - * ACEXML/common/Makefile: - * ACEXML/common/StreamFactory.h: - * ACEXML/common/StreamFactory.cpp: - - New class to create appropriate streams based on the URI. This - is needed for handling the parameter entities. - - * ACEXML/examples/SAXPrint/main.cpp: - - Move common error message to a new usage function. - -Thu Jul 11 16:28:24 2002 Steve Huston <shuston@riverace.com> - - * tests/Proactor_Scatter_Gather_Test.cpp: Changed Writer to derive - from ACE_Handler, not ACE_Service_Handler. It's not the target of - an asynch connection factory, and declares an open() method that - hides the one in ACE_Service_Handler. - -Thu Jul 11 16:15:49 2002 Steve Huston <shuston@riverace.com> - - * examples/C++NPv2/SR_Configurable_Logging_Server.dsp: - * examples/TPCLS.dsp: - * examples/TPLS.dsp: - * examples/C++NPv2/examples.dsw: Added SR_Configurable_Logging_Server, - TPCLS, and TPLS projects. - - * examples/C++NPv2/TPC_Logging_Server.h: Use ACE traits macros to - allow build on Win32. - * examples/C++NPv2/TPC_Logging_Server.cpp: Fix build problems on - Win32. - - * examples/TP_Logging_Server.h (init): Added #include "ace/Auto_Ptr.h" - and corrected use of ACE_NEW_NORETURN; use ACE_Auto_Ptr::get() to get - the argv pointer. - -Thu Jul 11 15:03:47 2002 Steve Huston <shuston@riverace.com> - - * ace/ace_dll64.mak: TEMPORARILY added /wd4267 to disable the - conversion size_t to int warning. - - * ace/Activation_Queue.{h i} (method_count): Change return type - from int to size_t. - - * ace/Arg_Shifter.cpp (cur_arg_strncasecmp): Change flag_length var - to size_t to match string function signatures. If returning a - 'remaining' length, cast back to int. - - * ace/ARGV.cpp (ctor): Build string length using size_t, not int. - - * ace/CDR_Base.cpp (ACE_CDR::consolidate): Use ptrdiff_t, not the - nonstandard ptr_arith_t. - - * ace/CDR_Stream.cpp (grow_and_adjust): Use ptrdiff_t, notptr_arith_t. - (write_string): Cast ACE_SString:length result to ACE_CDR::ULong. - A number of other size_t<->ACE_CDR::ULong casts. - - * ace/CDR_Stream.i (write_string, operator<<): Use a ACE_CDR::ULong - len, can't directly use the return from strlen(), a size_t. - - * ace/Codecs.cpp: size_t conversions. - - * ace/DEV_Addr.i (set): Cast strlen result to int for base_set. - - * ace/DEV_IO.{i cpp} (send, recv): Cast iovec count from size_t to - int for call to ACE_OS level. - - * ace/Hash_Map_Manager_T.{h, i, cpp} (shared_find): Change loc argument - from u_long to size_t. Better match for its use. - - * ace/OS.h (ACE_align_binary, ACE_ptr_align_binary): Change from - using ptr_arith_t to the standard ptrdiff_t. - - * ace/DLL_Manager.cpp: - * ace/Parse_Node.cpp: Avoid casting a void * to a long then back to a - function pointer. Cast using proper calling signature. - - * ace/Reactor_Impl.h (ACE_Reactor_Notify::notify): - * ace/Select_Reactor_Base.{h cpp} (ACE_Select_Reactor_Notify::notify): - Change return type from ssize_t to int to match advertised return - type from ACE_Reactor. - - * ace/Select_Reactor_T.cpp (check_handles): Use an int select_width - to pass width to select() so it can be ignored on ACE_WIN64. - - * ace/String_Base_Const.{h cpp}: Changed npos from int to ssize_t. - - * ace/String_Base.{h i}: (operator[]) changed slot arg from int - to size_t. (strstr(), find(), rfind()) changed return type from - int to ssize_t; change pos arg from int to size_t. - - * ace/Timer_Heap_T.{h cpp}: Change 'slot' arg from int to size_t. - Also changed timer_ids_ to use ssize_t rather than long to - avoid length mismatches between the max size and the timer - queue (a size_t) and the arrays. - - * ace/WFMO_Reactor.h (ACE_WFMO_Reactor_Notify): Change return type - from ssize_to to int to match corresponding change in - ACE_Reactor_Notify, above. The implementation already was returning - int. - - * ace/WFMO_Reactor.cpp (make_changes_in_current_infos, - make_changes_in_suspension_infos): Align use of size_t. - (wait_for_multiple_events)...........Should maybe make the - sizes in this class DWORD????? - - * ace/WIN32_Proactor.cpp (close, handle_events): completion_key - needs to be a different on Win64 than on Win32. - -Thu Jul 11 08:34:22 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/USAGE: - * bin/MakeProjectCreator/modules/Driver.pm: - - Added new options and documentation dealing with generation of - relative paths in projects. - -Wed Jul 10 11:46:02 2002 Mayur Deshpande <mayur@ics.uci.edu> - - * bin/auto_run_tests.lst: - - Added tests/AMH_Exception and examples/AMH/Sink_Server as tests - for AMH - -Wed Jul 10 13:31:49 2002 Steve Huston <shuston@riverace.com> - - * examples/C++NPv2/CLD.dsp: - * examples/C++NPv2/examples.dsw: Added CLD project. - - * examples/Client_Logging_Daemon.cpp: Disable the info, suspend, and - resume methods - these are left as an exercise for the reader. - -Wed Jul 10 11:24:39 2002 Steve Huston <shuston@riverace.com> - - * examples/C++NPv2/AIO_CLD.dsp: - * examples/C++NPv2/examples.dsw: Added AIO_CLD project. - - * examples/C++NPv2/AIO_Client_Logging_Daemon.cpp: Fix Win32 compile - problems. - -Wed Jul 10 07:22:24 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/TemplateParser.pm: - - If the project creator needs the slashes converted, then - do so when generating relative paths. - -Tue Jul 9 23:24:44 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * include/makeinclude/rules.local.GNU: - - Print out more explanative messages. - - * include/makeinclude/platform_osf1_4.x_cxx.GNU: - * include/makeinclude/platform_sunos5_sunc++.GNU: - - Turn on symbol versioning for these platforms with the - proprietary compilers that ships with these systems viz, Sun CC - and cxx. - - -Tue Jul 9 16:24:46 2002 Douglas C. Schmidt <schmidt@ace.cs.wustl.edu> - - * ace/Containers_T.cpp: Fixed inconsistent initializer order for - ACE_Fixed_Stack and ACE_Bounded_Stack. Thanks to Michael - Searles <msearles@base16.com> for reporting this. - -Tue Jul 9 11:10:38 2002 Steve Huston <shuston@riverace.com> - - * apps/Gateway/Peer/Peer.{h cpp}: Renamed Peer_Acceptor::open() - to Peer_Acceptor::start() to avoid clashes with virtual - ACE_Acceptor::open. - -Tue Jul 9 09:01:14 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/config/taolib_with_idl.mpb: - - Change the default tao_idl option from -Gs to -Sc. - -Tue Jul 9 08:07:20 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/TemplateParser.pm: - - When using Cygwin Perl, we need to convert the cwd to a windows - path to match up with environment variables that are usable - outside of Cygwin. - -Tue Jul 9 07:46:57 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/MakeProjectCreator/modules/TemplateParser.pm: - - Fix up the relative value matching code to work correctly - on Windows. - -Tue Jul 9 07:31:48 2002 Chad Elliott <elliott_c@ociweb.com> - - * ace/ace.mpc: - * ace/QoS/qos.mpc: - * ace/RMCast/rmcast.mpc: - * ace/SSL/ssl.mpc: - * apps/gperf/src/gperf.mpc: - * tests/tests.mpc: - * tests/RMCast/tests.mpc: - * tests/SSL/tests.mpc: - - Input files for the Make Project Creator for just the basic - libraries, exe's and tests. - -Tue Jul 9 07:24:06 2002 Chad Elliott <elliott_c@ociweb.com> - - * bin/mpc.pl: - * bin/mwc.pl: - - The starting point for creating workspaces and projects. - - * bin/MakeProjectCreator/README: - * bin/MakeProjectCreator/USAGE: - - A bit of lite documentation on the Make Project Creator. - - * bin/MakeProjectCreator/config/aceexe.mpb: - * bin/MakeProjectCreator/config/acelib.mpb: - * bin/MakeProjectCreator/config/aceversion.mpb: - * bin/MakeProjectCreator/config/avstreamsexe.mpb: - * bin/MakeProjectCreator/config/client.mpb: - * bin/MakeProjectCreator/config/global.mpb: - * bin/MakeProjectCreator/config/messaging.mpb: - * bin/MakeProjectCreator/config/namingexe.mpb: - * bin/MakeProjectCreator/config/notify.mpb: - * bin/MakeProjectCreator/config/notifytest.mpb: - * bin/MakeProjectCreator/config/orbsvcsexe.mpb: - * bin/MakeProjectCreator/config/orbsvcslib.mpb: - * bin/MakeProjectCreator/config/rt_client.mpb: - * bin/MakeProjectCreator/config/rt_server.mpb: - * bin/MakeProjectCreator/config/rteventexe.mpb: - * bin/MakeProjectCreator/config/server.mpb: - * bin/MakeProjectCreator/config/strategies.mpb: - * bin/MakeProjectCreator/config/taoexe.mpb: - * bin/MakeProjectCreator/config/taolib.mpb: - * bin/MakeProjectCreator/config/taolib_with_idl.mpb: - * bin/MakeProjectCreator/config/taoversion.mpb: - - Base projects for most of ACE and TAO. - - - * bin/MakeProjectCreator/modules/Driver.pm: - * bin/MakeProjectCreator/modules/Creator.pm: - * bin/MakeProjectCreator/modules/WorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/GUID.pm: - * bin/MakeProjectCreator/modules/Parser.pm: - * bin/MakeProjectCreator/modules/ProjectCreator.pm: - * bin/MakeProjectCreator/modules/TemplateInputReader.pm: - * bin/MakeProjectCreator/modules/TemplateParser.pm: - - The main modules of the Make Project Creator which provide most of - the functionalty. - - * bin/MakeProjectCreator/modules/BorlandProjectCreator.pm: - * bin/MakeProjectCreator/modules/BorlandWorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/EM3ProjectCreator.pm: - * bin/MakeProjectCreator/modules/EM3WorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/GHSProjectCreator.pm: - * bin/MakeProjectCreator/modules/GHSWorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/GNUProjectCreator.pm: - * bin/MakeProjectCreator/modules/GNUWorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/NMakeProjectCreator.pm: - * bin/MakeProjectCreator/modules/NMakeWorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/VA4ProjectCreator.pm: - * bin/MakeProjectCreator/modules/VA4WorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/VC6ProjectCreator.pm: - * bin/MakeProjectCreator/modules/VC6WorkspaceCreator.pm: - * bin/MakeProjectCreator/modules/VC7ProjectCreator.pm: - * bin/MakeProjectCreator/modules/VC7WorkspaceCreator.pm: - - Subclassed workspace and project creators. - - * bin/MakeProjectCreator/templates/bor.mpd: - * bin/MakeProjectCreator/templates/bordll.mpt: - * bin/MakeProjectCreator/templates/borexe.mpt: - * bin/MakeProjectCreator/templates/em3vcp.mpd: - * bin/MakeProjectCreator/templates/em3vcpdll.mpt: - * bin/MakeProjectCreator/templates/em3vcpdllexe.mpt: - * bin/MakeProjectCreator/templates/em3vcplib.mpt: - * bin/MakeProjectCreator/templates/em3vcplibexe.mpt: - * bin/MakeProjectCreator/templates/ghs.mpd: - * bin/MakeProjectCreator/templates/gnu.mpd: - * bin/MakeProjectCreator/templates/gnudll.mpt: - * bin/MakeProjectCreator/templates/gnuexe.mpt: - * bin/MakeProjectCreator/templates/nmake.mpd: - * bin/MakeProjectCreator/templates/nmakedll.mpt: - * bin/MakeProjectCreator/templates/nmakeexe.mpt: - * bin/MakeProjectCreator/templates/va4icc.mpd: - * bin/MakeProjectCreator/templates/va4iccdll.mpt: - * bin/MakeProjectCreator/templates/va4iccdllexe.mpt: - * bin/MakeProjectCreator/templates/va4icclib.mpt: - * bin/MakeProjectCreator/templates/va4icclibexe.mpt: - * bin/MakeProjectCreator/templates/vc6dsp.mpd: - * bin/MakeProjectCreator/templates/vc6dspdll.mpt: - * bin/MakeProjectCreator/templates/vc6dspdllexe.mpt: - * bin/MakeProjectCreator/templates/vc6dsplib.mpt: - * bin/MakeProjectCreator/templates/vc6dsplibexe.mpt: - * bin/MakeProjectCreator/templates/vc7.mpd: - * bin/MakeProjectCreator/templates/vc7dll.mpt: - * bin/MakeProjectCreator/templates/vc7exe.mpt: - - Template and template input files for all of the different project - creators. - -Tue Jul 9 07:21:29 2002 Chad Elliott <elliott_c@ociweb.com> - - * ace/Map_T.h: - * ace/Map.cpp: - * ace/Pair_T.h: - * ace/Pair.cpp: - - Remove the use of Map.h and Pair.h. - - * ace/gethrtime.cpp: - - Only compile the contents of this file if using GHS and are - building for the x86. - -Mon Jul 8 17:26:51 2002 Steve Huston <shuston@riverace.com> - - * ace/Task.cpp (svc_run): Borland complains about reinterpret_cast from - int to unsigned long, so change the cast to ACE_THR_FUNC_RETURN to - a static_cast for that compiler. Thanks to Johnny Willemsen - <jwillemsen@remedy.nl> for this fix. - -Fri Jul 5 10:30:44 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * ace/Array_Base.cpp: Swapped the order of includes so that - Array_Base.h comes after Malloc_Base.h. Thanks to Al Pariante - <Albert.Pariante@axcelis.com> for reporting this. - -Thu Jul 4 11:27:05 2002 Douglas C. Schmidt <schmidt@tango.doc.wustl.edu> - - * Happy 226th Birthday USA! - -Sat Jul 06 19:32:40 2002 Balachandran Natarajan <bala@cs.wustl.edu> - - * ace/Time_Value.h (ACE_Time_Value): Fix for [BUGID 1240]. Some of - the operators were missing explicit declarations in the header - file (they were anyway defined in the inlined files) which - created problems in builds where inlining was diabled. Please - see - - http://deuce.doc.wustl.edu/bugzilla/show_bug.cgi?id=1240 for - more details. Thanks to Marina Spivak <marina@atdesk.com> for - reporting this. - -Fri Jul 5 18:44:39 2002 Steve Huston <shuston@riverace.com> - - * ace/ACE.cpp: Corrected undeclared 'select_width' errors and - signed/unsigned mismatch warnings. - -Fri Jul 5 16:14:51 2002 Steve Huston <shuston@riverace.com> - - * ace/OS.{h i} (send, sendto, recv, recvfrom): Change 'len' argument - from int to size_t. - Removed conditional typedef of ssize_t; it's done in Basic_Types.h. - - * ace/ACE.h (recv): Filled in documentation for varargs variant. - * ace/ACE.cpp (send, recv): Properly convert size_t 'count' arg to int - iovec count for call to ACE_OS::sendv/recvv(). - (handle_timed_accept, handle_timed_complete, handle_ready): Use - an int select_width for ACE_OS::select() calls to get the types - correct. This arg is ignored on Windows anyway. - - * ace/Basic_Types.h: If ACE_HAS_SSIZE_T not defined, and Win64, - typedef SSIZE_T ssize_t; else use the existing int typedef. - -Wed Jul 3 11:09:31 2002 Steve Huston <shuston@riverace.com> - - * tests/SSL/Thread_Pool_Reactor_SSL_Test.cpp: - Change thread function to return ACE_THR_FUNC_RETURN, not void *. - -Tue Jul 02 21:43:53 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ACEXML/common/XMLFilterImpl.h: One more pre/post mismatch. - -Tue Jul 2 16:58:55 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ACEXML/common/XML_Types.h: - - Removed unnecessary includes and fixed a typo. - -Tue Jul 02 16:35:19 2002 Nanbor Wang <nanbor@cs.wustl.edu> - - * ACEXML/common/Attributes.h: - * ACEXML/common/Locator.h: Fixed mismatched pre/post inclusions. - - * ACEXML/parser/parser/Parser.cpp: Fixed compilation errors. - -Tue Jul 2 11:53:29 2002 Steve Huston <shuston@riverace.com> - - * ace/config-all.h: Changed definition of ACE_THR_FUNC to incorporate - the platform's thread function return type (DWORD on Windows, - void * on most others). Uses a new typedef ACE_THR_FUNC_RETURN - to specify the platform's thread function return type. - - * ace/OS.{h i cpp} (thr_exit, thr_join): The status arg is now - a ACE_THR_FUNC_RETURN *. - Also moved remaining thread function definition type things to - config-all.h with the rest of them. - - * ace/Base_Thread_Adapter.{h cpp} (invoke(), ace_thread_adapter()): - * ace/OS_Thread_Adapter.{h cpp} - * ace/Thread_Adapter.{h cpp} (invoke(), invoke_i()): - Return type changed from void * to ACE_THR_FUNC_RETURN. - - * ace/Event_Handler.{cpp h} (read_adapter): Change return value from - void* to ACE_THR_FUNC_RETURN. - - * ace/Thread_Hook.{h cpp}: ACE_Thread_Hook::start () returns - ACE_THR_FUNC_RETURN, not void *. - - * ace/Thread.{h i} (exit, join): 'status' arg is now - ACE_THR_FUNC_RETURN *, not void *. - - * ace/Thread_Control.{h inl cpp} (status, exit): - * ace/Thread_Manager.{h cpp} (exit, join): Thread exit status - type changed from void * to ACE_THR_FUNC_RETURN. - - * ace/Task.{h cpp} (ACE_Task_Base): Return type from svc_run now - ACE_THR_FUNC_RETURN. - - * examples/C++NPv1/RT_Thread_Per_Connection_Logging_Server.cpp: - * examples/C++NPv1/Thread_Per_Connection_Logging_Server.{cpp, h}: - * examples/C++NPv2/Select_Reactor_Logging_Server.cpp: - * examples/C++NPv2/Server_Shutdown.cpp: - * examples/C++NPv2/TP_Reactor_Logging_Server.cpp: - * examples/C++NPv2/WFMO_Reactor_Logging_Server.cpp: - * examples/C++NPv2/Client_Logging_Daemon.cpp: - * examples/Connection/blocking/SPIPE-acceptor.cpp: - * examples/IPC_SAP/SOCK_SAP/CPP-inserver.cpp: - * examples/IPC_SAP/SSL_SAP/SSL-server.cpp: - * examples/Reactor/WFMO_Reactor/Abandoned.cpp: - * performance-tests/TCP/tcp_test.cpp: - * tests/ACE_Init_Test.cpp: - * tests/MEM_Stream_Test.cpp: - * tests/Service_Config_DLL.cpp: - * tests/Task_Test.cpp: - * tests/Thread_Pool_Reactor_Resume_Test.cpp: - * tests/Thread_Pool_Reactor_Test.cpp: - Change thread function to return ACE_THR_FUNC_RETURN, not void *. - - * examples/C++NPv2/TP_Logging_Server.cpp (handle_input, svc): - * examples/C++NPv2/AC_Client_Logging_Daemon.cpp (open): - * examples/C++NPv2/display_logfile.cpp (svc): Change - ACE_reinterpret_cast to ACE_static_cast. Thanks to Doug Schmidt - for finding this. - - * examples/C++NPv2/TP_Logging_Server.h: - * examples/C++NPv2/Reactor_Logging_Server_Adapter.cpp: Corrected use - of ACE_NEW_RETURN. - -Mon Jul 1 23:16:47 2002 Jeff Parsons <parsons@cs.wustl.edu> - - * */Makefile: - - Updated dependencies. - -Mon Jul 1 21:59:44 2002 Krishnakumar B <kitty@cs.wustl.edu> - - * ACEXML/common/Attributes.h: - * ACEXML/common/AttributesImpl.h: - * ACEXML/common/Attributes_Def_Builder.h: - * ACEXML/common/CharStream.h: - * ACEXML/common/ContentHandler.h: - * ACEXML/common/DTDHandler.h: - * ACEXML/common/DTD_Manager.h: - * ACEXML/common/DefaultHandler.h: - * ACEXML/common/Element_Def_Builder.h: - * ACEXML/common/EntityResolver.h: - * ACEXML/common/Env.h: - * ACEXML/common/ErrorHandler.h: - * ACEXML/common/Exception.h: - * ACEXML/common/FileCharStream.h: - * ACEXML/common/HttpCharStream.h: - * ACEXML/common/InputSource.h: - * ACEXML/common/Locator.h: - * ACEXML/common/LocatorImpl.h: - * ACEXML/common/Mem_Map_Stream.h: - * ACEXML/common/NamespaceSupport.h: - * ACEXML/common/SAXExceptions.h: - * ACEXML/common/StrCharStream.h: - * ACEXML/common/Transcode.h: - * ACEXML/common/URL_Addr.h: - * ACEXML/common/Validator.h: - * ACEXML/common/XMLFilter.h: - * ACEXML/common/XMLFilterImpl.h: - * ACEXML/common/XMLReader.h: - * ACEXML/common/XML_Codecs.h: - * ACEXML/common/XML_Types.h: - * ACEXML/parser/debug_validator/Debug_Attributes_Builder.cpp: - * ACEXML/parser/debug_validator/Debug_Attributes_Builder.h: - * ACEXML/parser/debug_validator/Debug_DTD_Manager.h: - * ACEXML/parser/debug_validator/Debug_Element_Builder.h: - * ACEXML/parser/debug_validator/Element_Tree.h: - * ACEXML/parser/parser/Entity_Manager.h: - * ACEXML/parser/parser/Parser.cpp: - * ACEXML/parser/parser/Parser.h: - - ACE_LACKS_PRAGMA_ONCE fixes. - -Mon Jul 1 06:21:00 2002 Douglas C. Schmidt <schmidt@macarena.cs.wustl.edu> - * ace/Sock_Connect.cpp (count_interfaces): Added a MacOSX-specific - value for OSIOCGIFCONF. Thanks to John Zorko <j.zorko@att.net> - for this fix. |