1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
|
Thu Apr 29 03:34:56 UTC 2010 Chris Cleeland <cleeland@ociweb.com>
* ace/Stack_Trace.cpp: Clarified origins of platform-specific
code in the leading comments for the file.
Wed Apr 28 09:23:47 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>
* netsvcs/lib/Client_Logging_Handler.cpp:
* netsvcs/lib/TS_Server_Handler.cpp:
Use %C for logging the hostname instead of %s. This fixes bugzilla
3846, thanks to Christian Freund <freund at wrz dot de> for reporting
this
Mon Apr 26 08:29:58 UTC 2010 Vladimir Zykov <vladimir.zykov@prismtech.com>
* tests/SSL/SSL_Asynch_Stream_Test.cpp:
* tests/SSL/Bug_2912_Regression_Test.cpp:
* tests/Proactor_UDP_Test.cpp:
* tests/MT_Reference_Counted_Event_Handler_Test.cpp:
* tests/Proactor_Test_IPV6.cpp:
* tests/Bug_2740_Regression_Test.cpp:
* tests/Dev_Poll_Reactor_Test.cpp:
* tests/Proactor_Test.cpp:
* tests/TP_Reactor_Test.cpp:
All these tests need to block SIGPIPE signal. POSIX defines that
newly created threads must inherit signal mask from the creating
thread. The later doesn't happen at least on Solaris 10. With this
change each new thread in the above tests will block SIGPIPE signal
individually.
Thu Apr 22 07:51:42 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>
* Release:
* bin/make_release.py:
Updated because of rename
* debian:
* debianbuild:
Renamed debianbuild to debianbuild
Wed Apr 21 13:36:30 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>
* examples/Connection/blocking/SPIPE-connector.cpp:
Use Reactor instead of Proactor, thanks to Dave <dwh0403 at 163 dot com>
for reporting this
Tue Apr 20 13:02:46 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>
* ace/Atomic_Op.h:
Fix redefinition warning
Mon Apr 19 18:17:30 UTC 2010 UTC 2010 Chad Beaulac <chad@objectivesolutions.com
* ace/Service_Repository.cpp
Modified ASR::fini to print debug info for empty
service entries and handle empty service entries.
This fixes the ASR remove functionality that was
one of the reasons SC_Server was crashing at shutdown.
* examples/ASX/CCM_App/SC_Server.cpp
Removing signal handler before main exit to present SEGV
* tests/run_test.lst
Removed !FIXED_BUGS_ONLY from Bug_3334_Regression_Test
Apparently it was missed in the last commit.
This fixes Bugzilla #2916 and #3205.
I'll create another patch for #3334 after this commit.
Mon Apr 19 12:44:56 UTC 2010 Chad Beaulac <chad@objectivesolutions.com>
* ace/Service_Types.h
* ace/Service_Types.inl
* ace/Service_Types.cpp
Added service_type_ attr to expose the type of service being
managed to the ACE_Service_Repository. This allows the ASR to
manage the lifecycle of the ACE_Module and ACE_Stream in order to
avoid a double delete of ACE_Module at shutdown.
ACE_Stream_Type::fini was modified to not call ACE_Module_Type::fini
as this will results in a double delete when ACE_Service_Repository::fini
is called. ACE_Stream_Type::remove(module_name) was modified to not call
ACE_Module_Type::fini as this will results in a double delete
when ACE_Service_Repository::fini is called also.
* ace/Service_Repository.cpp
Modified ASR::fini to iterate over the service_array_ twice.
ACE_Service_Type::fini is called on all ACE_Stream_Type and
ACE_Service_Object_Type instances first. Then, fini is called on
all ACE_Modules_Type instances. All calls to fini are done in the
order the services appear in the ASR::service_array_ except for
the grouping described here. The calls to ACE_Module_Type::fini
must be done last because ACE_Stream_Type::fini accesses the Modules
so they must not be deleted by a call to ACE_Module_Type::fini
before that.
* tests/Bug_3334_Regresssion_Test.cpp
Added call to ACE_Service_Repository::fini_svcs() to capture all
debug output before the application exits.
* tests/run_tests.lst
Removed !FIXED_BUGS_ONLY from Bug_3334_Regression_Test
This fixes Bugzilla #3334
Mon Apr 19 06:54:01 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>
* bin/MakeProjectCreator/config/global.features:
Enabled LightWeigth CCM as default, if you want to use the
full ccm features set ccm_lw to 0 in your default.features file
Fri Apr 16 13:02:04 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>
* 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
Wed Apr 14 11:22:27 UTC 2010 Steve Huston <shuston@riverace.com>
* ace/Dev_Poll_Reactor.cpp: Fixed compile error on some g++ versions.
Tue Apr 13 20:34:40 UTC 2010 Steve Huston <shuston@riverace.com>
* bin/PerlACE/Process_Unix.pm: Fix typo that broke all tests.
Tue Apr 13 19:34:40 UTC 2010 Steve Huston <shuston@riverace.com>
* ace/Dev_Poll_Reactor.{h cpp}: Reorder the enqueue/pipe operations
when adding a notification to close the race where notifies can get
lost in the ACE_HAS_NOTIFICATION_QUEUE case. Also fixed a misplaced
#endif in read_notification_pipe() and cleaned it up some.
Mon Apr 12 09:31:54 UTC 2010 Martin Corino <mcorino@remedy.nl>
* bin/PerlACE/TestTarget_WinCE.pm:
* bin/PerlACE/ProcessVX_Win32.pm:
* bin/PerlACE/Process_Unix.pm:
* bin/PerlACE/ProcessLVRT.pm:
* bin/PerlACE/TestTarget.pm:
* bin/PerlACE/TestTarget_VxWorks.pm:
* bin/PerlACE/ProcessVX.pm:
* bin/PerlACE/TestTarget_LVRT.pm:
* bin/PerlACE/ProcessVX_Unix.pm:
* bin/PerlACE/Process_Win32.pm:
* bin/PerlACE/Run_Test.pm:
* bin/PerlACE/Process_VMS.pm:
Added support for running multiplatform tests using remote shell
connections.
Currently only supported for *nix-like platforms
(like Linux PPC64 <-> Linux x86_64).
Thu Apr 8 21:25:54 UTC 2010 Abdul Sowayan <sowayan@gmail.com>
* ace/OS_NS_netdb.cpp:
Fixed compile time error encountered in MAC and iPhone platforms.
Thu Apr 8 14:51:45 UTC 2010 Douglas C. Schmidt <schmidt@dre.vanderbilt.edu>
* docs/ACE-guidelines.html: Clarified the style recommendations
for loop and if statement bodies.
Wed Apr 7 16:27:46 UTC 2010 Adam Mitz <mitza@ociweb.com>
* bin/cle.pl:
Added the ability to use an environment variable (CL_CHANGELOG_FILE)
to specify the name of the ChangeLog file to use.
Tue Apr 6 14:37:38 UTC 2010 Douglas C. Schmidt <schmidt@dre.vanderbilt.edu>
* ACE-INSTALL.html (http): Updated the link to the OCI MPC
website. Thanks to Tim <tim at burmair dot com> for suggesting
this.
Wed Apr 7 07:06:31 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>
* bin/MakeProjectCreator/config/global.features:
ccm_lw is disabled by default
Tue Apr 06 06:08:48 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>
* ace/os_include/os_stddef.h:
Fixed incorrect semicolon
* ace/OS_NS_netdb.cpp:
Initialise pointers to 0
* configure.ac:
Fix for cross compilation, thanks to Ed Blackmond
<ed dot blackmond at themis dot com> for reporting this
Fri Apr 02 13:57:48 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>
* bin/diff-builds-and-group-fixed-tests-only.sh:
Updated latest release date
* bin/generate_rel_manpages:
Don't automatically copy the files to ISIS, added the copy
step to the release instructions
* docs/bczar/bczar.html:
Small improvements and manually copy the html files using
scp. Makes it easier to run this from another host
* docs/Download.html:
Updated links to point to x.7.8
* html/index.html:
Added link for x.7.8
* NEWS:
Added 5.7.8 section
Fri Apr 02 10:50:48 CEST 2010 Johnny Willemsen <jwillemsen@remedy.nl>
* ACE version 5.7.8 released.
Tue Mar 30 09:38:08 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>
* debianbuild/*:
Pulled in latest files from pkg-ace
Tue Mar 30 09:34:38 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>
* tests/Thread_Timer_Queue_Adapter_Test.cpp:
Fixed compilation problems with single threaded builds
Tue Mar 30 07:50:02 UTC 2010 Vladimir Zykov <vladimir.zykov@prismtech.com>
* tests/MT_NonBlocking_Connect_Test.cpp:
Fixed compilation problem for single threaded builds.
Mon Mar 29 12:25:38 UTC 2010 Marcel Smit <msmit@remedy.nl>
* tests/NDDS_Timer_Test.cpp:
Resolved test issues.
Mon Mar 29 11:05:23 UTC 2010 Marcel Smit <msmit@remedy.nl>
* tests/NDDS_Timer_Test.cpp:
* tests/run_test.lst:
* tests/tests.mpc:
Added timer test which uses the ndds timers.
Mon Mar 29 06:59:08 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>
* bin/MakeProjectCreator/config/ace_qt4reactor.mpb:
New base project, thanks to Marek Brudka <mbrudka at aster dot pl>
for providing this
Mon Mar 29 06:54:08 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>
* debianbuild/*:
Pulled in latest files from pkg-ace
Fri Mar 26 14:07:55 UTC 2010 Steve Huston <shuston@riverace.com>
* ace/Dev_Poll_Reactor.{h inl cpp}: Refinements to the way notifies
are handled:
1. The notify handler was being dispatched multiple times per
notify; this has been fixed, which increases performance.
2. Notifications are now dispatched in only one thread at a time.
This brings more parity to the way notifies are handled in
other reactor implementations.
Fri Mar 26 13:43:08 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>
* debianbuild/tao-utils.install:
Corrected filenames
* debianbuild/libtao-orbsvcs-dev.install:
Added missing libraries
Fri Mar 26 09:02:50 UTC 2010 Vladimir Zykov <vladimir.zykov@prismtech.com>
* tests/run_test.pl:
Added -l option which allows to run tests selectively.
Wed Mar 24 07:17:22 UTC 2010 Marcel Smit <msmit@remedy.nl>
* bin/fuzz.pl:
Added dds4ccm_trace checks.
Tue Mar 23 10:59:08 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>
* debianbuild/patches/34-bts386713.dpatch:
Updated from debian
Fri Mar 19 16:52:08 UTC 2010 Vladimir Zykov <vladimir.zykov@prismtech.com>
* tests/MT_NonBlocking_Connect_Test.cpp:
And another attempt to workaround WFMO reactor. The previous
changes could break the test on other platforms.
Fri Mar 19 16:44:37 UTC 2010 Steve Huston <shuston@riverace.com>
* ace/Dev_Poll_Reactor.cpp (find_handler): If the handle isn't
registered, don't try to dereference a null pointer; return 0.
Fri Mar 19 16:16:30 UTC 2010 Vladimir Zykov <vladimir.zykov@prismtech.com>
* tests/MT_NonBlocking_Connect_Test.cpp:
Fixed a typo in previous commit.
Fri Mar 19 15:58:52 UTC 2010 Vladimir Zykov <vladimir.zykov@prismtech.com>
* tests/MT_NonBlocking_Connect_Test.cpp:
Fixed a test on Windows for WFMO reactor which doesn't support
work_pending() function.
Fri Mar 19 13:23:18 UTC 2010 Adam Mitz <mitza@ociweb.com>
* ace/OS_NS_Thread.cpp:
Added a helper function for starting tasks on VxWorks.
* bin/PerlACE/ProcessVX_Unix.pm:
* bin/PerlACE/ProcessVX_Win32.pm:
* bin/PerlACE/TestTarget_VxWorks.pm:
Fixed a bug that was triggered with CreateProcess was called with
only one argument.
Also, set the correct relative path to the executable when it's not
in the current directory.
Fri Mar 19 08:56:29 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>
* include/makeinclude/platform_linux.GNU:
Added support for archmodelflag
Thu Mar 18 10:40:29 UTC 2010 Vladimir Zykov <vladimir.zykov@prismtech.com>
* tests/NonBlocking_Conn_Test.cpp:
Made logging a bit more verbose.
Wed Mar 17 20:00:45 UTC 2010 Steve Huston <shuston@riverace.com>
* include/makeinclude/platform_linux_common.GNU: Make the ARFLAGS
setting overridable.
Wed Mar 17 14:24:20 UTC 2010 Vladimir Zykov <vladimir.zykov@prismtech.com>
* ace/Logging_Strategy.cpp:
* ace/Logging_Strategy.h:
Extended ACE_Logging_Strategy to allow safely change reactor
used by this logging strategy.
Wed Mar 17 11:58:10 UTC 2010 Vladimir Zykov <vladimir.zykov@prismtech.com>
Committed the changes that I had to revert before x.7.7.
* ace/Strategies_T.cpp:
Changed the code so that close() is called in case of failure
instead of destroy().
* ace/Connector.cpp:
* ace/Connector.h:
Changed the fix for bug#3731. Now NBCH adds a reference in
constructor and removes it in destructor if the SVC_HANDLER that
it owns is reference counted. This is a cleaner solution than the
one used before.
* tests/NonBlocking_Conn_Test.h:
* tests/Process_Strategy_Test.cpp:
* tests/NonBlocking_Conn_Test.cpp:
Extended NonBlocking_Conn_Test and changed Process_Strategy_Test
because of the change in Strategies_T.cpp.
* tests/MT_NonBlocking_Connect_Test.cpp:
* tests/tests.mpc:
* tests/run_test.lst:
Added a new test. This can be a reproducer for the bug#3731 which
currently doesn't have its own test.
Wed Mar 17 08:57:30 UTC 2010 Vladimir Zykov <vladimir.zykov@prismtech.com>
* ace/Service_Gestalt.h:
* ace/Service_Config.h:
Extended doxygen documentaion related to the order of processing
of command-line directives and service configuration files.
* ace/Service_Gestalt.cpp:
Changed the order of processing. First, service configuration files
are processed then command-line directive. This way default svc.conf
can not override results of user provided command-line directives.
* NEWS:
Added a news entry for this change.
Mon Mar 15 20:08:32 UTC 2010 Olli Savia <ops@iki.fi>
* ace/config-WinCE.h:
* ace/config-hpux-11.00.h:
* ace/config-lynxos.h:
* ace/config-netbsd.h:
* ace/config-openbsd.h:
* ace/config-sunos5.10.h:
* ace/config-sunos5.5.h:
* ace/config-vxworks6.4.h:
* configure.ac:
Removed floorl and ceill related settings which are no longer
needed.
Mon Mar 15 12:50:54 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>
* bin/MakeProjectCreator/templates/gnu.mpd:
When generating with -static set static_libs_only to 1. This fixes
bugzilla 3815. Thanks to Chad Elliot for assisting with this
Mon Mar 15 09:41:54 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>
Reverted change below, it breaks unix builds
Mon Mar 15 07:58:54 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>
* ace/High_Res_Timer.cpp
* ace/High_Res_Timer.h
* ace/High_Res_Timer.inl
Use 64bit calculations to increase our precision. If you want to have the 32bit
calculations, add ACE_USE_WINDOWS_32BIT_HIGH_RES_TIMER_CALCULATIONS as define. If
the scoreboard doesn't show any platforms requiring 32bit, we will remove that code
before the next micro release goes out
Thanks to Alon Diamant <diamant dot alon at gmail dot com> for supplying the patches.
This fixes bugzilla 3703.
Mon Mar 15 07:58:54 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>
* ace/High_Res_Timer.cpp
* ace/High_Res_Timer.h
* ace/High_Res_Timer.inl
Use 64bit calculations to increase our precision. If you want to have the 32bit
calculations, add ACE_USE_WINDOWS_32BIT_HIGH_RES_TIMER_CALCULATIONS as define. If
the scoreboard doesn't show any platforms requiring 32bit, we will remove that code
before the next micro release goes out
Thanks to Alon Diamant <diamant dot alon at gmail dot com> for supplying the patches.
This fixes bugzilla 3703.
* ace/Timer_Queue_Adapters.cpp
* ace/Timer_Queue_Adapters.h
* ace/Timer_Queue_Adapters.inl
Make it possible to use a customer event handler in ACE_Thread_Timer_Queue_Adapter.
Thanks to Alon Diamant <diamant dot alon at gmail dot com> for supplying the patches.
This fixes bugzilla 3614
* tests/tests.mpc:
* tests/Thread_Timer_Queue_Adapter_Test.cpp:
New test for testing custom event handlers in ACE_Thread_Timer_Queue_Adapter.
Thanks to Alon Diamant <diamant dot alon at gmail dot com> for creating
this new test
Fri Mar 12 21:53:30 UTC 2010 Adam Mitz <mitza@ociweb.com>
* bin/PerlACE/ProcessVX.pm:
* bin/PerlACE/ProcessVX_Unix.pm:
Added an option to use the 'expect' program to interact with
the remote target (via telnet). This is only enabled when the
environment variable ACE_RUN_VX_USE_EXPECT is defined.
* tests/Log_Msg_Test.cpp:
Updated the ACE_VXWORKS version check to include version 6.8.
* tests/run_test.lst:
Added !VxWorks to exclude Pipe_Test which can't be run
on VxWorks (kernel mode) because it depends on ACE_Process::spawn().
Fri Mar 12 12:09:02 UTC 2010 Olli Savia <ops@iki.fi>
* ace/OS_NS_math.h:
* ace/OS_NS_math.inl:
Rewrote ceil and floor functions using templates and overloaded
ceil/floor functions provided by standard C++.
* tests/OS_Test.cpp:
Added tests for ceil(float) and floor(float).
Thu Mar 11 09:33:54 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>
* 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
Wed Mar 10 10:31:16 UTC 2010 Vladimir Zykov <vladimir.zykov@prismtech.com>
* ace/Svc_Handler.cpp:
* tests/run_test.lst:
Fixed bug#2609 and enabled the test for it. Thanks to Milan
Cvetkovic <milan dot cvetkovic at mpathix dot com> for
contributing this.
Tue Mar 9 12:20:54 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>
* include/makeinclude/platform_g++_common.GNU:
Added CXX_MACHINE
Mon Mar 8 15:36:55 UTC 2010 Adam Mitz <mitza@ociweb.com>
* bin/mpc.pl:
* bin/mwc.pl:
Some platforms have been reported to return undef from
$FindBin::RealBin. In this case, fall back on dirname($0).
Mon Mar 8 13:53:54 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>
* ace/Atomic_Op_GCC_T.inl:
Fixed gcc warning
Mon Mar 8 08:35:54 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>
* 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
Mon Mar 8 08:24:54 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>
* tests/Atomic_Op_Test.cpp:
Added tests for short/unsigned short/bool
Fri Mar 5 14:02:54 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>
* configure.ac:
Added test for GCC builtin support
Fri Mar 5 12:40:54 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>
* 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>
* 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.
On FC10 64bit the int/unsigned int Atomic_Op is now around 6 to 7 times
faster than before this change
* ace/config-g++-common.h:
Enable the new builtin atomic also for EM64T
* ace/ace.mpc:
* ace/Makefile.am:
Added new files
Fri Mar 5 11:30:54 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>
* tests/Atomic_Op_Test.cpp:
Print the number of iterations
Fri Mar 5 11:28:54 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>
* tests/Atomic_Op_Test.cpp:
Also measure postfix increment/decrement
Fri Mar 5 11:21:54 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>
* tests/Atomic_Op_Test.cpp:
Added more checks to make sure Atomic_Op does work
Tue Mar 2 21:02:51 UTC 2010 Steve Huston <shuston@riverace.com>
* ace/Dev_Poll_Reactor.cpp (mask_ops_i): If the specified handle is
not registered, return -1 instead of crashing.
Thu Mar 4 09:15:54 UTC 2010 Martin Corino <mcorino@remedy.nl>
* bin/make_release.py:
Fix for problems with creating MPC release tag.
Thu Mar 4 07:50:54 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>
* tests/Atomic_Op_Test.cpp:
Rewritten this test to use a template method so that we can
test atomic op easily for various data types
* ace/Time_Value.cpp:
Layout changes and fixed compile problem with CB2010. This fixes
bugzilla 3835, thanks to Jan Kalin <jan dot kalin at zag dot si>
for reporting this.
* ace/config-win32-borland.h:
CB also has long timevalue mismatch
Thu Mar 4 07:50:54 UTC 2010 Johnny Willemsen <jwillemsen@remedy.nl>
* NEWS:
Updated for next micro
* rpmbuild/ace-tao.spec:
Fixed problem with installation of nslist readme
* etc/index.html:
Added 5.7.7
* docs/Download.html:
Updated for 5.7.7
* docs/bczar/bczar.html:
Several updates because we now package on a FC12 system
* debianbuild/rules:
Removed commented out part
* bin/make_release.py:
Next micro release will not ship msvc71 solutions, it is still
supported but the end user should generate the solutions him self.
Also WinCE solutions are not generated anymore for CIAO and the
autoconf bootstrapping is done much earlier in the release process
so when it fails we know early. The hostname check has been removed
* bin/diff-builds-and-group-fixed-tests-only.sh:
Updated because of micro release
Local Variables:
mode: change-log
add-log-time-format: (lambda () (progn (setq tz (getenv "TZ")) (set-time-zone-rule "UTC") (setq time (format-time-string "%a %b %e %H:%M:%S %Z %Y" (current-time))) (set-time-zone-rule tz) time))
indent-tabs-mode: nil
End:
|