summaryrefslogtreecommitdiff
path: root/ACE/ace/Atomic_Op.inl
Commit message (Collapse)AuthorAgeFilesLines
* Removed redundant voidJohnny Willemsen2021-03-171-2/+2
|
* Removed redundant voidJohnny Willemsen2020-12-091-9/+9
|
* Removed single line Id tagJohnny Willemsen2015-01-071-2/+0
|
* ChangeLogTag:Sat Sep 15 01:11:15 UTC 2012 Steve Huston <shuston@riverace.com>Steve Huston2012-09-151-0/+3
|
* ChangeLogTag:Thu Sep 13 16:46:03 UTC 2012 Steve Huston <shuston@riverace.com>Steve Huston2012-09-131-0/+52
|
* ChangeLogtag:Mon Dec 5 20;16:51 UTC 2011 Steve Huston <shuston@riverace.com>Steve Huston2011-12-051-0/+32
|
* Tue Nov 2 13:26:56 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>Johnny Willemsen2010-11-021-1/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | * ace/ATM_Stream.inl: * ace/Atomic_Op.inl: * ace/Atomic_Op_GCC_T.inl: Layout changes * ace/OS_NS_dlfcn.inl: Removed old code * ace/config-macros.h: When using ACE_HAS_VALGRIND, disable the dlclose functionality, valgrind needs to have all loaded dll's still in memory in order to show the full callstack, if the dll is unloaded before process shutdown valgrind will show ??? * ace/os_include/os_dlfcn.h: Removed old defines * debian/TODO: * debian/debian.rules: Cleanup * debian/cmp-installation.sh: Removed this file.
* Removed trailing whitespacesJohnny Willemsen2010-09-171-1/+1
|
* Fri Apr 16 13:02:04 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>Johnny Willemsen2010-04-161-0/+28
| | | | | | | | | | * ace/Atomic_Op.h: * ace/Atomic_Op.inl: * ace/Atomic_Op.cpp: * ace/config-sunos5.10.h: Added support for the Solaris 10 Atomic support, thanks to Aleksandar Vukajlovic <aleksandar dot vukajlovic at finsoft dot rs> for providing the patches
* Thu Mar 11 09:33:54 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>Johnny Willemsen2010-03-111-0/+4
| | | | | | | | | | | | | | | | | | | * ace/config-g++-common.h: PPU 4.1.1 for PPC doesn't have atomic builtin support for datatypes of size 1 and 2 * ace/Atomic_Op.h: * ace/Atomic_Op.inl: Added defines to just disable short/unsigned short/bool specializations * tests/run_test.lst: Enable 2610 but mark it as not fixed * ace/Process_Manager.cpp: Layout change * ace/WIN32_Asynch_IO.cpp: Const changes
* Mon Mar 8 08:35:54 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>Johnny Willemsen2010-03-081-0/+78
| | | | | | | * ace/Atomic_Op.h: * ace/Atomic_Op.inl: Added specializations for bool/unsigned short/short, is around 7 times faster than the default atomic op
* (no commit message)Johnny Willemsen2010-03-061-31/+32
|
* Fri Mar 5 12:40:54 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>Johnny Willemsen2010-03-051-13/+3
| | | | | | | | | | * ace/Atomic_Op.h: * ace/Atomic_Op.inl: We did a benchmark on a FC10 64bit system with the GCC builtin support for long/unsigned long. In a release/inline build the performance results using the GCC builtin support is around 2 times faster than when using the ACE assembly version. When we have the GCC builtin support, we now also going to use that for long/unsigned long.
* Fri Mar 5 12:20:54 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>Johnny Willemsen2010-03-051-27/+121
| | | | | | | | | | | | | | | | | | | * ace/Atomic_Op.h: * ace/Atomic_Op.inl: * ace/Atomic_Op_GCC_T.cpp: * ace/Atomic_Op_GCC_T.h: * ace/Atomic_Op_GCC_T.inl: Added new Atomic template for the GCC builtin atomic op support. This is used on PPC64/PPC32/IA64 for int/unsigned int/long/unsigned long. On EM64T we do use this for int/unsigned int, the ACE assembly version for long/unsigned long is faster than the GCC. * ace/config-g++-common.h: Enable the new builtin atomic also for EM64T * ace/ace.mpc: * ace/Makefile.am: Added new files
* Wed Feb 10 18:39:30 UTC 2010 Martin Corino <mcorino@remedy.nl>mcorino2010-02-101-17/+12
| | | | | | | | | | | | | | | | | | * ace/config-g++-common.h: Added logic to detect (usable) support for GCC builtin __sync_XXX atomic op functions. * ace/Atomic_Op.h: * ace/Atomic_Op.inl: Changed autodetection for GCC builtin atomic ops so it will only be used when really wanted. Removed include for stdatomic.h because that has nothing to do with the __sync_XXX builtins but rather with a *proposed* atomic op implementation for the C++1x standard. The __sync_XXX builtins are truly intrinsic, i.e. no header files involved.
* Thu Feb 3 09:57:30 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>Johnny Willemsen2010-02-041-0/+29
| | | | | | | | | | | | | | * ace/Atomic_Op.inl: Added support for the GCC Atomic Builtins, if you want to use these, add ACE_HAS_GCC_ATOMIC_BUILTINS defined as 1 in your config file * ace/Reactor.cpp: * ace/Select_Reactor_Base.cpp: Layout changes * ace/Reactor.h: Doxygen change
* Fri Apr 17 15:03:32 UTC 2009 Johnny Willemsen <jwillemsen@remedy.nl>Johnny Willemsen2009-04-171-0/+28
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * ace/Atomic_Op.{h,inl}: Added support for VxWorks vxAtomicLib. If you want to use that library add ACE_HAS_VXATOMICLIB to your config.h file * ace/config-vxworks6.4.h: VxWorks 6.4 has cpuset_t and taskCpuAffinitySet * ace/os_include/os_sched.h: VxWorks has cpuset_t * ace/OS_NS_sys_mman.inl: Updated VxWorks version checks * ace/OS_NS_sys_utsname.cpp: const change * ace/OS_NS_Thread.cpp: Added support for taskCpuAffinitySet * ace/Sock_Connect.cpp: Changed the checks for VxWorks to correctly detect which implementation of get_ip_interfaces we have to use * rpmbuild/ace-tao-orbsvcs-daemon.patch: This patch doesn't work anymore with svn head, if this is a real issue, it has to be addressed in the real code and not with a patch. This should already be possible with the -ORBDeamon flag * rpmbuild/ace-tao-ciao.spec: Don't apply ace-tao-orbsvcs-daemon.patch, fixed a small provides bug
* undoing accidental deletionWilliam R. Otte2008-03-041-0/+335
|
* swap in externals for ACE and TAOWilliam R. Otte2008-03-041-335/+0
|
* Tue Apr 17 10:03:12 UTC 2007 Johnny Willemsen <jwillemsen@remedy.nl>Johnny Willemsen2007-04-171-0/+2
|
* Tue Apr 17 09:13:12 UTC 2007 Johnny Willemsen <jwillemsen@remedy.nl>Johnny Willemsen2007-04-171-13/+44
|
* ChangeLogTag:Mon Feb 19 21:02:55 UTC 2007 Steve Huston <shuston@riverace.com>Steve Huston2007-02-191-1/+2
|
* Repo restructuringWilliam R. Otte2006-07-241-0/+301