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
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
|
<!DOCTYPE html>
<html class="writer-html5" lang="en">
<head>
<meta charset="utf-8" />
<meta name="viewport" content="width=device-width, initial-scale=1.0" />
<title>3.1.1. Open MPI v5.0.x series — Open MPI 5.0.8 documentation</title>
<link rel="stylesheet" type="text/css" href="../../_static/pygments.css" />
<link rel="stylesheet" type="text/css" href="../../_static/css/theme.css" />
<!--[if lt IE 9]>
<script src="../../_static/js/html5shiv.min.js"></script>
<![endif]-->
<script data-url_root="../../" id="documentation_options" src="../../_static/documentation_options.js"></script>
<script src="../../_static/jquery.js"></script>
<script src="../../_static/underscore.js"></script>
<script src="../../_static/_sphinx_javascript_frameworks_compat.js"></script>
<script src="../../_static/doctools.js"></script>
<script src="../../_static/sphinx_highlight.js"></script>
<script src="../../_static/js/theme.js"></script>
<link rel="index" title="Index" href="../../genindex.html" />
<link rel="search" title="Search" href="../../search.html" />
<link rel="next" title="3.1.2. Open MPI v4.1.x series" href="v4.1.x.html" />
<link rel="prev" title="3.1. Changelog" href="index.html" />
</head>
<body class="wy-body-for-nav">
<div class="wy-grid-for-nav">
<nav data-toggle="wy-nav-shift" class="wy-nav-side">
<div class="wy-side-scroll">
<div class="wy-side-nav-search" >
<a href="../../index.html" class="icon icon-home">
Open MPI
</a>
<div role="search">
<form id="rtd-search-form" class="wy-form" action="../../search.html" method="get">
<input type="text" name="q" placeholder="Search docs" aria-label="Search docs" />
<input type="hidden" name="check_keywords" value="yes" />
<input type="hidden" name="area" value="default" />
</form>
</div>
</div><div class="wy-menu wy-menu-vertical" data-spy="affix" role="navigation" aria-label="Navigation menu">
<ul class="current">
<li class="toctree-l1"><a class="reference internal" href="../../quickstart.html">1. Quick start</a></li>
<li class="toctree-l1"><a class="reference internal" href="../../getting-help.html">2. Getting help</a></li>
<li class="toctree-l1 current"><a class="reference internal" href="../index.html">3. Release notes</a><ul class="current">
<li class="toctree-l2 current"><a class="reference internal" href="index.html">3.1. Changelog</a><ul class="current">
<li class="toctree-l3 current"><a class="current reference internal" href="#">3.1.1. Open MPI v5.0.x series</a><ul>
<li class="toctree-l4"><a class="reference internal" href="#open-mpi-version-v5-0-8">3.1.1.1. Open MPI Version v5.0.8</a></li>
<li class="toctree-l4"><a class="reference internal" href="#open-mpi-version-v5-0-7">3.1.1.2. Open MPI Version v5.0.7</a></li>
<li class="toctree-l4"><a class="reference internal" href="#open-mpi-version-v5-0-6">3.1.1.3. Open MPI Version v5.0.6</a></li>
<li class="toctree-l4"><a class="reference internal" href="#open-mpi-version-v5-0-5">3.1.1.4. Open MPI version v5.0.5</a></li>
<li class="toctree-l4"><a class="reference internal" href="#open-mpi-version-v5-0-4">3.1.1.5. Open MPI version v5.0.4</a></li>
<li class="toctree-l4"><a class="reference internal" href="#open-mpi-version-v5-0-3">3.1.1.6. Open MPI version v5.0.3</a></li>
<li class="toctree-l4"><a class="reference internal" href="#open-mpi-version-v5-0-2">3.1.1.7. Open MPI version v5.0.2</a></li>
<li class="toctree-l4"><a class="reference internal" href="#open-mpi-version-v5-0-1">3.1.1.8. Open MPI version v5.0.1</a></li>
<li class="toctree-l4"><a class="reference internal" href="#open-mpi-version-5-0-0">3.1.1.9. Open MPI version 5.0.0</a></li>
</ul>
</li>
<li class="toctree-l3"><a class="reference internal" href="v4.1.x.html">3.1.2. Open MPI v4.1.x series</a></li>
<li class="toctree-l3"><a class="reference internal" href="v4.0.x.html">3.1.3. Open MPI v4.0.x series</a></li>
<li class="toctree-l3"><a class="reference internal" href="v3.1.x.html">3.1.4. Open MPI v3.1.x series</a></li>
<li class="toctree-l3"><a class="reference internal" href="v3.0.x.html">3.1.5. Open MPI v3.0.x series</a></li>
<li class="toctree-l3"><a class="reference internal" href="v2.x.html">3.1.6. Open MPI v2.x series</a></li>
<li class="toctree-l3"><a class="reference internal" href="v1.x.html">3.1.7. Open MPI v1.x series</a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="../general.html">3.2. General notes</a></li>
<li class="toctree-l2"><a class="reference internal" href="../platform.html">3.3. Platform Notes</a></li>
<li class="toctree-l2"><a class="reference internal" href="../compilers.html">3.4. Compiler Notes</a></li>
<li class="toctree-l2"><a class="reference internal" href="../run-time.html">3.5. General Run-Time Support Notes</a></li>
<li class="toctree-l2"><a class="reference internal" href="../mpi.html">3.6. MPI Functionality and Features</a></li>
<li class="toctree-l2"><a class="reference internal" href="../openshmem.html">3.7. OpenSHMEM Functionality and Features</a></li>
<li class="toctree-l2"><a class="reference internal" href="../mpi-collectives.html">3.8. MPI Collectives</a></li>
<li class="toctree-l2"><a class="reference internal" href="../openshmem-collectives.html">3.9. OpenSHMEM Collectives</a></li>
<li class="toctree-l2"><a class="reference internal" href="../networks.html">3.10. Network Support</a></li>
<li class="toctree-l2"><a class="reference internal" href="../extensions.html">3.11. Open MPI Extensions</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../installing-open-mpi/index.html">4. Building and installing Open MPI</a></li>
<li class="toctree-l1"><a class="reference internal" href="../../features/index.html">5. Open MPI-specific features</a></li>
<li class="toctree-l1"><a class="reference internal" href="../../validate.html">6. Validating your installation</a></li>
<li class="toctree-l1"><a class="reference internal" href="../../version-numbering.html">7. Version numbers and compatibility</a></li>
<li class="toctree-l1"><a class="reference internal" href="../../mca.html">8. The Modular Component Architecture (MCA)</a></li>
<li class="toctree-l1"><a class="reference internal" href="../../building-apps/index.html">9. Building MPI applications</a></li>
<li class="toctree-l1"><a class="reference internal" href="../../launching-apps/index.html">10. Launching MPI applications</a></li>
<li class="toctree-l1"><a class="reference internal" href="../../tuning-apps/index.html">11. Run-time operation and tuning MPI applications</a></li>
<li class="toctree-l1"><a class="reference internal" href="../../app-debug/index.html">12. Debugging Open MPI Parallel Applications</a></li>
<li class="toctree-l1"><a class="reference internal" href="../../developers/index.html">13. Developer’s guide</a></li>
<li class="toctree-l1"><a class="reference internal" href="../../contributing.html">14. Contributing to Open MPI</a></li>
<li class="toctree-l1"><a class="reference internal" href="../../license/index.html">15. License</a></li>
<li class="toctree-l1"><a class="reference internal" href="../../history.html">16. History of Open MPI</a></li>
<li class="toctree-l1"><a class="reference internal" href="../../man-openmpi/index.html">17. Open MPI manual pages</a></li>
<li class="toctree-l1"><a class="reference internal" href="../../man-openshmem/index.html">18. OpenSHMEM manual pages</a></li>
</ul>
</div>
</div>
</nav>
<section data-toggle="wy-nav-shift" class="wy-nav-content-wrap"><nav class="wy-nav-top" aria-label="Mobile navigation menu" >
<i data-toggle="wy-nav-top" class="fa fa-bars"></i>
<a href="../../index.html">Open MPI</a>
</nav>
<div class="wy-nav-content">
<div class="rst-content">
<div role="navigation" aria-label="Page navigation">
<ul class="wy-breadcrumbs">
<li><a href="../../index.html" class="icon icon-home" aria-label="Home"></a></li>
<li class="breadcrumb-item"><a href="../index.html"><span class="section-number">3. </span>Release notes</a></li>
<li class="breadcrumb-item"><a href="index.html"><span class="section-number">3.1. </span>Changelog</a></li>
<li class="breadcrumb-item active"><span class="section-number">3.1.1. </span>Open MPI v5.0.x series</li>
<li class="wy-breadcrumbs-aside">
<a href="../../_sources/release-notes/changelog/v5.0.x.rst.txt" rel="nofollow"> View page source</a>
</li>
</ul>
<hr/>
</div>
<div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
<div itemprop="articleBody">
<style>
.wy-table-responsive table td,.wy-table-responsive table th{white-space:normal}
</style><div class="section" id="open-mpi-v5-0-x-series">
<h1><span class="section-number">3.1.1. </span>Open MPI v5.0.x series<a class="headerlink" href="#open-mpi-v5-0-x-series" title="Permalink to this heading"></a></h1>
<p>This file contains all the NEWS updates for the Open MPI v5.0.x
series, in reverse chronological order.</p>
<div class="section" id="open-mpi-version-v5-0-8">
<h2><span class="section-number">3.1.1.1. </span>Open MPI Version v5.0.8<a class="headerlink" href="#open-mpi-version-v5-0-8" title="Permalink to this heading"></a></h2>
<dl class="field-list simple">
<dt class="field-odd">Date</dt>
<dd class="field-odd"><p>30 May 2025</p>
</dd>
</dl>
<ul class="simple">
<li><p>Internal PMIx and PRRTe versions:
- PMIx (v5.0.8). Repo: <code class="docutils literal notranslate"><span class="pre">https://github.com/openpmix/openpmix</span></code>. Commit hash: <code class="docutils literal notranslate"><span class="pre">907b1ccaeec61a1197f0ee5264d4fef20b257b84</span></code>.
- PRRTE (v3.0.11). Repo: <code class="docutils literal notranslate"><span class="pre">https://github.com/openpmix/prrte</span></code>. Commit hash: <code class="docutils literal notranslate"><span class="pre">222f03fbb98b71abd293aa205b38fa9a38e57965</span></code>.</p></li>
<li><p>Functionality Enhancements
- UCC collective operations: Improved handling of MPI_IN_PLACE in allgather, allgatherv, alltoall, alltoallv, gather, gatherv, scatter, and scatterv operations
- UCX OSC: Added support for “no_locks” info key to disable lock table usage
- OFI MTL: Enhanced CXI provider support for better compatibility
- AARCH64: Added SVE detection alongside NEON in aarch64 op component
- Fortran: Fixed common symbol sizes and alignments for better compatibility</p></li>
<li><p>Bug Fixes and Minor Enhancements
- Hugepage Mpool: Fixed sizing of hugepages for better memory management
- Configure: Fixed –with-prrte=internal option handling
- Documentation: Updated shared memory and networking documentation
- Build system: Fixed support for flang on OSX
- Removed unused spread contrib sample scripts
- Various documentation updates and improvements
- CI: Removed Ubuntu 18 and RHEL 7 from test matrix
- Fixed type mismatch errors in shared memory components
- Updated Java configuration for latest Fedora releases
- Fixed memory zone reference counting in UCX dynamic windows</p></li>
</ul>
</div>
<div class="section" id="open-mpi-version-v5-0-7">
<h2><span class="section-number">3.1.1.2. </span>Open MPI Version v5.0.7<a class="headerlink" href="#open-mpi-version-v5-0-7" title="Permalink to this heading"></a></h2>
<dl class="field-list simple">
<dt class="field-odd">Date</dt>
<dd class="field-odd"><p>14 February January 2025</p>
</dd>
</dl>
<ul class="simple">
<li><p>Internal PMIx and PRRTe version:
- PMIx (v5.0.5). Repo: <code class="docutils literal notranslate"><span class="pre">https://github.com/openpmix/openpmix</span></code>. Commit hash: <code class="docutils literal notranslate"><span class="pre">707920c99de946a5c3a1850da457340f38c0caf2</span></code>.
- PRRTE (v3.0.8). repo: <code class="docutils literal notranslate"><span class="pre">https://github.com/openpmix/prrte</span></code>. Commit hash: <code class="docutils literal notranslate"><span class="pre">f6f5c181c1dec317c31f61effd73f960ce2eac25</span></code>.</p></li>
<li><p>Updated UCX version requirement to v1.9</p></li>
<li><p>Documentation Improvements
- Updated <cite>coll tuned</cite> rules file documentation to reflect changes in versioning and features.
- Updated CUDA-related networking documentation for consistency.
- Improved descriptions for <cite>mpirun</cite> options, including <code class="docutils literal notranslate"><span class="pre">--pmixmca</span></code> and <code class="docutils literal notranslate"><span class="pre">--prtemca</span></code>.
- Fixed typos and formatting issues in <cite>mpirun</cite> man pages.</p></li>
<li><p>Functionality Enhancements
- Improved argument checking for <code class="docutils literal notranslate"><span class="pre">MPI_Comm_create_from_group</span></code> and <code class="docutils literal notranslate"><span class="pre">MPI_Intercomm_create_from_groups</span></code> to align with MPI 4.0 standards.
- Addressed potential race conditions in <cite>PMIx_Group_construct</cite> calls by adding discriminators.
- Fixed <code class="docutils literal notranslate"><span class="pre">MPI_Wtime</span></code> to initialize timing relative to <code class="docutils literal notranslate"><span class="pre">MPI_Init</span></code> for accurate session-level measurements.
- Added support for the <cite>alltoall_algorithm_max_requests</cite> parameter in the <cite>coll tuned</cite> rules file, maintaining backward compatibility.
- Introduced a version identifier for the <cite>coll tuned</cite> rules file for better parser handling.
- Adjusted MCA variable scopes across multiple components to allow setting via the <cite>MPI_T</cite> interface: <cite>coll adapt,ucc, han, basic, hcoll</cite></p></li>
<li><p>Bug Fixes and Minor Enhancements
- Ensures that singletons properly create their own sessions directory trees.
- Fixed community Jenkins build pipeline to handle branch commits properly.
- Resolved missing profiling symbols in the <cite>mpi_f08</cite> Fortran library.
- Corrected parameter mismatches in <cite>coll/cuda</cite> functions to avoid type issues.
- Fixed build issues on Hurd.
- Removed unused portions of rcache base structure.
- Fixed oshmem base segment address exchange.
- Fixed UCC collective fallback issue.
- Updated Nvidia/Mellanox platform file</p></li>
</ul>
</div>
<div class="section" id="open-mpi-version-v5-0-6">
<h2><span class="section-number">3.1.1.3. </span>Open MPI Version v5.0.6<a class="headerlink" href="#open-mpi-version-v5-0-6" title="Permalink to this heading"></a></h2>
<dl class="field-list simple">
<dt class="field-odd">Date</dt>
<dd class="field-odd"><p>14 November 2024</p>
</dd>
</dl>
<ul class="simple">
<li><p>Internal PMIx and PRRTe version:</p>
<ul>
<li><p>PMIx (v5.0.4). Repo: <code class="docutils literal notranslate"><span class="pre">https://github.com/openpmix/openpmix</span></code>. Commit hash: <code class="docutils literal notranslate"><span class="pre">20ad9b16fe49ede4a76e7489bdf0e2ea05c3de95</span></code>.</p></li>
<li><p>PRRTE (v3.0.7). repo: <code class="docutils literal notranslate"><span class="pre">https://github.com/openpmix/prrte</span></code>. Commit hash: <code class="docutils literal notranslate"><span class="pre">b2f4163e40a84fe777fa36dfbffa4db23d8d31dd</span></code>.</p></li>
<li><p>Session Management Enhancements:</p></li>
<li><p>Finalization Control: Ensures proper finalization of classes when references to the OPAL utility layer reach zero. A new smoke test has been added to validate this behavior.</p></li>
<li><p>Test and CI Additions: A smoke test was introduced for session management, also integrated into macOS GitHub CI to ensure stability across platforms.</p></li>
<li><p>Recognize requests from the same instance rather than same communicator, fixes Waitall/Waitany/Testall/Testany.</p></li>
<li><p>Multiple Init/Finalize Handling: Improvements to handling session multiple initializations and finalizations, addressing issues with parameter management and session destruction.</p></li>
<li><p>CUDA and NVIDIA HPC Compiler Compatibility:</p></li>
<li><p>Stream-ordered Allocations: Adjustments for handling CUDA stream-ordered allocations and assigning primary device context in <cite>opal/cuda</cite> (Akshay Venkatesh).</p></li>
<li><p>nvfortran Compatibility: Configure script updated to handle NVIDIA’s <cite>nvfortran</cite> compiler, enabling proper compilation with <cite>-fPIC</cite> flags for shared libraries (Gilles Gouaillardet).</p></li>
<li><p>CUDA Memory Address Management: Improvements for handling virtual memory manager (VMM) pointers in CUDA applications to prevent runtime issues.</p></li>
<li><p>MPI Communication and Resource Management:</p></li>
<li><p>Info Duplication: Enhanced <cite>MPI_Info_dup</cite> function to ensure consistent key management across duplicated info objects, supporting better memory and resource handling.</p></li>
<li><p>OFI/Initialization Improvements: Refactored code for OFI common components to support session-based initialization models per MPI 4 standards.</p></li>
<li><p>Resource Leak Fixes: Addressed memory leaks and reference count issues in several components to improve memory management and stability in high-demand environments.</p></li>
<li><p>Detailed Locking Protocol: Modified default file-locking protocols in UFS component to ensure data consistency, especially when using data-sieving operations, which require broader locking.</p></li>
<li><p>MPI_T Interface Enhancements: Modified MCA parameter settings to allow runtime tuning of collectives through the MPI_T interface, adding flexibility for dynamic rules in collective operations.</p></li>
<li><p>Corrected various documentation typos and clarified parameter settings in several API references, improving readability and usability for end users.</p></li>
</ul>
</li>
</ul>
</div>
<div class="section" id="open-mpi-version-v5-0-5">
<h2><span class="section-number">3.1.1.4. </span>Open MPI version v5.0.5<a class="headerlink" href="#open-mpi-version-v5-0-5" title="Permalink to this heading"></a></h2>
<dl class="field-list">
<dt class="field-odd">Date</dt>
<dd class="field-odd"><p>23 July 2024</p>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>This is a quick follow-up release that fixes v5.0.4 which does
not compile on macOS (arm64) with the inbox Apple clang compiler (<code class="docutils literal notranslate"><span class="pre">/usr/bin/gcc</span></code>).
The <em>only</em> change in v5.0.5 compared to v5.0.4 is a fix for this compilation issue.</p>
</div>
</dd>
</dl>
<ul class="simple">
<li><p>Internal PMIx and PRRTe versions:</p>
<ul>
<li><p>PMIx (v5.0.3). Commit hash: <code class="docutils literal notranslate"><span class="pre">8ab6d680b90afd6e61766220a8724065a1b554a7</span></code>.</p></li>
<li><p>PRRTE (v3.0.6). Commit hash: <code class="docutils literal notranslate"><span class="pre">b68a0acb32cfc0d3c19249e5514820555bcf438b</span></code>.</p></li>
</ul>
</li>
<li><p>Bugfixes and changes</p>
<ul>
<li><p>Fix a typo that breaks the build on macOS with the Apple clang compiler on arm64 platforms.
Many thanks to Lisandro Dalcin for reporting the issue.</p></li>
</ul>
</li>
</ul>
</div>
<div class="section" id="open-mpi-version-v5-0-4">
<h2><span class="section-number">3.1.1.5. </span>Open MPI version v5.0.4<a class="headerlink" href="#open-mpi-version-v5-0-4" title="Permalink to this heading"></a></h2>
<dl class="field-list simple">
<dt class="field-odd">Date</dt>
<dd class="field-odd"><p>19 July 2024</p>
</dd>
</dl>
<ul class="simple">
<li><p>Internal PMIx and PRRTe versions:</p>
<ul>
<li><p>PMIx (v5.0.3). Commit hash: <code class="docutils literal notranslate"><span class="pre">8ab6d680b90afd6e61766220a8724065a1b554a7</span></code>.</p></li>
<li><p>PRRTE (v3.0.6). Commit hash: <code class="docutils literal notranslate"><span class="pre">b68a0acb32cfc0d3c19249e5514820555bcf438b</span></code>.</p></li>
</ul>
</li>
<li><p>Bugfixes and changes</p>
<ul>
<li><p>btl/ofi: retry posting receive buffer</p></li>
<li><p>btl/uct: check return status from some uct funcs</p></li>
<li><p>fortran: use -Wl,-ld_classic if supported</p></li>
<li><p>fortran: parameterize the install location of Fortran MPI bindings modulefiles
via the configure –with-mpi-moduledir CLI option</p></li>
<li><p>fortran: various configury improvements</p></li>
<li><p>use-mpi-f08: add missing subroutines in the mpi_f08 module</p></li>
<li><p>pml/base: remove redundant copy of ompi_mpi_thread_multiple</p></li>
<li><p>pml/cm: release datatype object in send completion callback</p></li>
<li><p>pml/ucx: propagate MPI serialized thread mode for all worker creations</p></li>
<li><p>pml/ob1: make no. of events an mca parameter</p></li>
<li><p>oshmem/shmem: fix warnings regarding types and unused variables in shmem</p></li>
<li><p>Add an AARCH64 OP component for for NEON and SVE ISA</p></li>
<li><p>Fix MPI function names in error output</p></li>
<li><p>Fix MPI_Neighbor_alltoall{v,w}_init to handle NULL parameters</p></li>
<li><p>Fix a bug in intercomm creation</p></li>
<li><p>Fix a bug in ompi_comm_lookup function to handle partially ready communicators</p></li>
<li><p>Fix error message in ompi_comm_get_rproc</p></li>
<li><p>Fix opal_basename for single character filenames</p></li>
<li><p>Fix rpm build test</p></li>
<li><p>Fix the info subscriber mechanism and hidden info keys</p></li>
<li><p>Fix ulfm bugs in ishrink/agree/iagree</p></li>
<li><p>Improve libcuda.so find command to follow symlinks</p></li>
<li><p>Improve pkgconfig installation logic in script wrapper</p></li>
<li><p>Change default tar format to tar-pax</p></li>
<li><p>Various document updates</p></li>
</ul>
</li>
<li><p>Many thanks to Open MPI community including:
Alex Margolin, Benjamin Kitor, Burlen Loring, Ethan Townsend, Geoffroy Vallee,
Jeff Hammond, Junchao Zhang, Lisandro Dalcin, Nuno Nobre, Rainer Keller,
Roie Danino, Thomas Vegas</p></li>
</ul>
</div>
<div class="section" id="open-mpi-version-v5-0-3">
<h2><span class="section-number">3.1.1.6. </span>Open MPI version v5.0.3<a class="headerlink" href="#open-mpi-version-v5-0-3" title="Permalink to this heading"></a></h2>
<dl class="field-list simple">
<dt class="field-odd">Date</dt>
<dd class="field-odd"><p>9 April 2024</p>
</dd>
</dl>
<ul class="simple">
<li><p>Internal PMIx and PRRTe versions:</p>
<ul>
<li><p>PMIx (v5.0.2). Commit hash: <code class="docutils literal notranslate"><span class="pre">33e93469e1e1f69904ff3e3827394719aa6b3671</span></code>.</p></li>
<li><p>PRRTE (v3.0.5). Commit hash: <code class="docutils literal notranslate"><span class="pre">3a70fac9a21700b31c4a9f9958afa207a627f0fa</span></code>.</p></li>
</ul>
</li>
<li><p>Bugfixes and changes</p>
<ul>
<li><p>Fix HAN MPI_Scatter intra-node data count and type</p></li>
<li><p>Add missing write memory barrier in btl/smcuda</p></li>
<li><p>Fix MPI_Alltoallv and MPI_Ialltoallv data count and type check</p></li>
<li><p>Enhance cuda-aware build by detecting default –with-cuda-lib value</p></li>
<li><p>Enhance ofi NIC selection logic</p></li>
<li><p>Fix an illegal memory access bug in mtl/ofi</p></li>
<li><p>Multiple bugfixes in rocm accelerator</p></li>
<li><p>Fix a double free bug in oshmem</p></li>
<li><p>Add missing oshmem headers</p></li>
<li><p>Fix uninitialized lock in btl/usnic</p></li>
<li><p>Fix uninitialized mutex in file_open</p></li>
<li><p>Fix compiler warnings about unused cid_base variable</p></li>
<li><p>Enhance autogen submodule checks</p></li>
<li><p>Various document updates</p></li>
</ul>
</li>
<li><p>Many thanks to Open MPI community including:
Christoph Niethammer, Christoph van Wüllen, Jakub Klinkovský, Lisandro Dalcin</p></li>
</ul>
</div>
<div class="section" id="open-mpi-version-v5-0-2">
<h2><span class="section-number">3.1.1.7. </span>Open MPI version v5.0.2<a class="headerlink" href="#open-mpi-version-v5-0-2" title="Permalink to this heading"></a></h2>
<dl class="field-list simple">
<dt class="field-odd">Date</dt>
<dd class="field-odd"><p>6 February 2024</p>
</dd>
</dl>
<ul>
<li><p>Internal PMIx and PRRTe versions:</p>
<ul class="simple">
<li><p>PMIx (v4.2.8). Commit hash: <code class="docutils literal notranslate"><span class="pre">d1c1ed0c2e64f19ad49291241a43630ea7fdce28</span></code>.</p></li>
<li><p>PRRTE (v3.0.3). Commit hash: <code class="docutils literal notranslate"><span class="pre">e383f5ad70c2633420d3425e9fb67e69b6bfd9c4</span></code>.</p></li>
</ul>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>Due to the timing of Open MPI v5.0.2 release, the PMIx and PRRTE versions
that are internally bundled in the Open MPI distribution are the same
versions as in the Open MPI v5.0.1 release: PMIx v4.2.8 and PRRTE v3.0.3.
As a result, changes and bug fixes in newer PMIx and PRRTE releases are
not included in Open MPI v5.0.2, notably:</p>
<p>PRRTE</p>
<ul class="simple">
<li><p>Fix the map-by pe-list option</p></li>
<li><p>Use pmix_path_nfs to detect shared file systems</p></li>
<li><p>psets: fix some problems with PMIX_QUERY_PSET_MEMBERSHIP query.
PMIX_QUERY_PSET_MEMBERSHIP query bugfixes.</p></li>
</ul>
<p>PMIx</p>
<ul class="simple">
<li><p>Restore default to enable-devel-check in Git repos</p></li>
<li><p>Protect against empty envar definition for mca_base_param_files</p></li>
<li><p>PMIx_Query_info: removed duplicated PMIX_RELEASE</p></li>
<li><p>Do not add no-unused-parameter for non-devel-check builds</p></li>
<li><p>Cast a few parameters when translating macros to functions</p></li>
<li><p>Fix one more spot for “get” from rank undefined</p></li>
<li><p>Fix “get” of key with undefined rank</p></li>
<li><p>Resolve problem of stack variables and realloc</p></li>
<li><p>Restore support for detecting shared file systems</p></li>
<li><p>Fix broken link in README</p></li>
</ul>
</div>
</li>
<li><p>Bugfixes and changes (in chronological order)</p>
<ul class="simple">
<li><p>Fix the fs/lustre component build</p></li>
<li><p>Fix a mtl/ofi multi-threaded race condition bug</p></li>
<li><p>Add missing fortran profiling interfaces according to MPI 4.0 standard</p></li>
<li><p>Allow 0-size data copy in opal/accelerator</p></li>
<li><p>Fix a btl/ofi and mtl/ofi provider selection bug</p></li>
<li><p>Properly initialize mca_rcache_base_module_t members</p></li>
<li><p>Fix a singleton launch segfault</p></li>
<li><p>Add GCC13 support with MacOS compatibility</p></li>
</ul>
</li>
<li><p>Many thanks to Open MPI community including:
Erik Schnetter</p></li>
</ul>
</div>
<div class="section" id="open-mpi-version-v5-0-1">
<h2><span class="section-number">3.1.1.8. </span>Open MPI version v5.0.1<a class="headerlink" href="#open-mpi-version-v5-0-1" title="Permalink to this heading"></a></h2>
<dl class="field-list simple">
<dt class="field-odd">Date</dt>
<dd class="field-odd"><p>20 December 2023</p>
</dd>
</dl>
<ul class="simple">
<li><p>Internal PMIx and PRRTe versions:</p>
<ul>
<li><p>PMIx (v4.2.8). Commit hash: <code class="docutils literal notranslate"><span class="pre">d1c1ed0c2e64f19ad49291241a43630ea7fdce28</span></code>.</p></li>
<li><p>PRRTE (v3.0.3). Commit hash: <code class="docutils literal notranslate"><span class="pre">e383f5ad70c2633420d3425e9fb67e69b6bfd9c4</span></code>.</p></li>
</ul>
</li>
<li><p>Bugfixes and changes (in chronological order)</p>
<ul>
<li><p>Various documentation related updates and changes</p></li>
<li><p>Fix a bunch of compiler warnings</p></li>
<li><p>Sessions: fix a problem with fortran MPI_Session_get_nth_pset interface</p></li>
<li><p>Correctly access the communicator name is MSGQ</p></li>
<li><p>accelerator/cuda: fix bug in makefile.am preventing correct linkage
in non-standard location</p></li>
<li><p>btlsmcuda: fix problem with makefile</p></li>
<li><p>Fix rcache/gpusm and rcache/rgpsum</p></li>
<li><p>Correctly handle attributes on MPI_COMM_WORLD.</p></li>
<li><p>Minor memory leak fixes in:
btl/tcp, mca_base_framework, ob1</p></li>
<li><p>Fix static initialization of recursive mutexes</p></li>
<li><p>Spack: fix for dlopen missing symbol problem</p></li>
<li><p>opal/mca/accelerator: ROCm 6.0 incompatibility fix</p></li>
<li><p>opal_var_dump_color_keys: fix an array overflow</p></li>
<li><p>SHMEM_LOCKS: MCS implementation of SHMEM LOCKS</p></li>
<li><p>configury: support flang-new</p></li>
<li><p>Update processing of “display_map” info key</p></li>
<li><p>dpm: update PMIX attribute</p></li>
</ul>
</li>
<li><p>Many thanks to Open MPI community including:
Tony Curtis, David Edwards Linaro, Niv Shpak for their contribution.</p></li>
</ul>
</div>
<div class="section" id="open-mpi-version-5-0-0">
<h2><span class="section-number">3.1.1.9. </span>Open MPI version 5.0.0<a class="headerlink" href="#open-mpi-version-5-0-0" title="Permalink to this heading"></a></h2>
<dl class="field-list simple">
<dt class="field-odd">Date</dt>
<dd class="field-odd"><p>25 October 2023</p>
</dd>
</dl>
<div class="warning admonition">
<p class="admonition-title">The MPIR API has been removed</p>
<p>As was announced in the summer of 2017, Open MPI has removed
support for MPIR-based tools beginning with the release of Open MPI
v5.0.0.</p>
<p>Open MPI now uses the <a class="reference external" href="https://github.com/openpmix/prrte">PRRTE</a>
runtime environment, which supports the <a class="reference external" href="https://pmix.org/">PMIx</a>
tools API — instead of the legacy MPIR API — for
debugging parallel jobs.</p>
<p>Users who still need legacy MPIR support should see
<a class="reference external" href="https://github.com/hpc/mpir-to-pmix-guide">https://github.com/hpc/mpir-to-pmix-guide</a> for more information.</p>
</div>
<div class="note admonition">
<p class="admonition-title">Zlib is suggested for better performance</p>
<p><a class="reference external" href="https://pmix.org/">PMIx</a> will optionally use <a class="reference external" href="https://github.com/madler/zlib">Zlib</a> to compress large data streams.
This may result in faster startup times and smaller memory
footprints (compared to not using compression).</p>
<p>The Open MPI community recommends building PMIx with Zlib support,
regardless of whether you are using an externally-installed PMIx or
the bundled PMIx that is included with Open MPI distribution
tarballs.</p>
<p>Note that while the Zlib library <em>may</em> be present on many systems
by default, the Zlib header files — which are needed to build
PMIx with Zlib support — may need to be installed separately
before building PMIx.</p>
</div>
<div class="admonition caution">
<p class="admonition-title">Caution</p>
<p>Open MPI has changed the default behavior of how it
builds and links against its <a class="reference internal" href="../../installing-open-mpi/required-support-libraries.html#label-install-required-support-libraries"><span class="std std-ref">required 3rd-party
packages</span></a>:
<a class="reference external" href="https://libevent.org/">Libevent</a>, <a class="reference external" href="https://www.open-mpi.org/projects/hwloc/">Hardware Locality</a>, <a class="reference external" href="https://pmix.org/">PMIx</a>, and <a class="reference external" href="https://github.com/openpmix/prrte">PRRTE</a>.</p>
<ol class="arabic">
<li><p>Unlike previous versions of Open MPI, Open MPI 5.0 and
later will prefer an external package that meets our
version requirements, even if it is older than our
internal version.</p></li>
<li><p>To simplify managing dependencies, any required
packages that Open MPI v5.0.x bundles will be
installed in Open MPI’s installation prefix, without
name mangling.</p>
<p>For example, if a valid Libevent installation cannot
be found and Open MPI therefore builds its bundled
version, a <code class="docutils literal notranslate"><span class="pre">libevent.so</span></code> will be installed in Open
MPI’s installation tree. This is different from
previous releases, where Open MPI name-mangled the
Libevent symbols and then statically pulled the
library into <code class="docutils literal notranslate"><span class="pre">libmpi.so</span></code>.</p>
</li>
</ol>
</div>
<ul>
<li><p>Internal PMIx and PRRTe versions:</p>
<ul class="simple">
<li><p>PMIx release tag v4.2.7. Commit hash: <code class="docutils literal notranslate"><span class="pre">57c405c52ad76bab0be9f95e29a6df660673081e</span></code>.</p></li>
<li><p>PRRTE release tag v3.0.2. Commit hash: <code class="docutils literal notranslate"><span class="pre">1552e36f0852bbc6d901ec95983369f0a3c283f6</span></code>.</p></li>
</ul>
</li>
<li><p>All other notable updates for v5.0.0:</p>
<ul>
<li><p>MPI-4.0 updates and additions:</p>
<ul class="simple">
<li><p>Support for MPI Sessions has been added.</p></li>
<li><p>Added partitioned communication using persistent sends
and persistent receives.</p></li>
<li><p>Added persistent collectives to the <code class="docutils literal notranslate"><span class="pre">MPI_</span></code> namespace
(they were previously available via the <code class="docutils literal notranslate"><span class="pre">MPIX_</span></code> prefix).</p></li>
<li><p>Added <code class="docutils literal notranslate"><span class="pre">MPI_Isendrecv()</span></code> and its variants.</p></li>
<li><p>Added support for <code class="docutils literal notranslate"><span class="pre">MPI_Comm_idup_with_info()</span></code>.</p></li>
<li><p>Added support for <code class="docutils literal notranslate"><span class="pre">MPI_Info_get_string()</span></code>.</p></li>
<li><p>Added support for <code class="docutils literal notranslate"><span class="pre">initial_error_handler</span></code> and the
<code class="docutils literal notranslate"><span class="pre">ERRORS_ABORT</span></code> infrastructure.</p></li>
<li><p>Added error handling for unbound errors to <code class="docutils literal notranslate"><span class="pre">MPI_COMM_SELF</span></code>.</p></li>
<li><p>Made <code class="docutils literal notranslate"><span class="pre">MPI_Comm_get_info()</span></code>, <code class="docutils literal notranslate"><span class="pre">MPI_File_get_info()</span></code>, and
<code class="docutils literal notranslate"><span class="pre">MPI_Win_get_info()</span></code> compliant to the standard.</p></li>
<li><p>Droped unknown/ignored info keys on communicators, files,
and windows.</p></li>
<li><p>Initial implementations of <code class="docutils literal notranslate"><span class="pre">MPI_COMM_TYPE_HW_GUIDED</span></code> and
<code class="docutils literal notranslate"><span class="pre">MPI_COMM_TYPE_HW_GUIDED</span></code> added.</p></li>
<li><p><code class="docutils literal notranslate"><span class="pre">MPI_Info_get()</span></code> and <code class="docutils literal notranslate"><span class="pre">MPI_Info_get_valuelen()</span></code> are now
deprecated.</p></li>
<li><p>Issue a deprecation warning when <code class="docutils literal notranslate"><span class="pre">MPI_Cancel()</span></code> is called for
a non-blocking send request.</p></li>
</ul>
</li>
<li><p>New Features:</p>
<ul class="simple">
<li><p>ULFM Fault Tolerance support has been added. See <a class="reference internal" href="../../features/ulfm.html#ulfm-label"><span class="std std-ref">the ULFM
section</span></a>.</p></li>
<li><p>CUDA is now supported in the <code class="docutils literal notranslate"><span class="pre">ofi</span></code> MTL.</p></li>
<li><p>A threading framework has been added to allow building Open MPI
with different threading libraries. It currently supports
<a class="reference external" href="https://www.argobots.org/">Argobots</a>, <a class="reference external" href="https://github.com/Qthreads/qthreads">Qthreads</a>, and Pthreads. See the
<code class="docutils literal notranslate"><span class="pre">--with-threads</span></code> option in the <code class="docutils literal notranslate"><span class="pre">configure</span></code> command. Thanks
to Shintaro Iwasaki and Jan Ciesko for their contributions to
this effort.</p></li>
<li><p>New Thread Local Storage API: Removes global visibility of TLS
structures and allows for dynamic TLS handling.</p></li>
<li><p>Added new <code class="docutils literal notranslate"><span class="pre">Accelerator</span></code> framework. CUDA-specific code
was replaced with a generic framework that standardizes various
device features such as copies or pointer type detection. This
allows for modularized implementation of various devices such as
the newly introduced ROCm Accelerator component. The redesign
also allows for Open MPI builds to be shipped with CUDA
support enabled without requiring CUDA libraries.</p></li>
<li><p>Added load-linked, store-conditional atomics support for
AArch64.</p></li>
<li><p>Added atomicity support to the <code class="docutils literal notranslate"><span class="pre">ompio</span></code> component.</p></li>
<li><p><code class="docutils literal notranslate"><span class="pre">osc/rdma</span></code>: Added support for MPI minimum alignment key.</p></li>
<li><p>Add ability to detect patched memory to
<code class="docutils literal notranslate"><span class="pre">memory_patcher</span></code>. Thanks to Rich Welch for the contribution.</p></li>
<li><p><code class="docutils literal notranslate"><span class="pre">coll/ucc</span></code>: Added support for the <code class="docutils literal notranslate"><span class="pre">MPI_Scatter()</span></code> and
<code class="docutils literal notranslate"><span class="pre">MPI_Iscatter()</span></code> collectives.</p></li>
<li><p>New algorithm for Allgather and Allgatherv has been added, based
on the paper <em>“Sparbit: a new logarithmic-cost and data
locality-aware MPI Allgather algorithm”</em>. Default algorithm
selection rules are unchanged; to use these algorithms add:
<code class="docutils literal notranslate"><span class="pre">--mca</span> <span class="pre">coll_tuned_allgather_algorithm</span> <span class="pre">sparbit</span></code> and/or <code class="docutils literal notranslate"><span class="pre">--mca</span>
<span class="pre">coll_tuned_allgatherv_algorithm</span> <span class="pre">sparbit</span></code> to your <code class="docutils literal notranslate"><span class="pre">mpirun</span></code>
command. Thanks to Wilton Jaciel Loch and Guilherme Koslovski
for their contribution.</p></li>
</ul>
</li>
<li><p>Transport updates and improvements</p>
<ul>
<li><p>One-sided Communication:</p>
<ul>
<li><p>Many MPI one-sided and RDMA emulation fixes for the <code class="docutils literal notranslate"><span class="pre">tcp</span></code> BTL.</p>
<p>This patch series fixs many issues when running with <code class="docutils literal notranslate"><span class="pre">--mca</span>
<span class="pre">osc</span> <span class="pre">rdma</span> <span class="pre">--mca</span> <span class="pre">btl</span> <span class="pre">tcp</span></code>, i.e., TCP support for one sided
MPI calls.</p>
</li>
<li><p>Many MPI one-sided fixes for the <code class="docutils literal notranslate"><span class="pre">uct</span></code> BTL.</p></li>
<li><p>Added support for <code class="docutils literal notranslate"><span class="pre">acc_single_intrinsic</span></code> to the one-sided
<code class="docutils literal notranslate"><span class="pre">ucx</span></code> component.</p></li>
<li><p>Removed the legacy <code class="docutils literal notranslate"><span class="pre">pt2pt</span></code> one-sided component. Users should
now utilize the <code class="docutils literal notranslate"><span class="pre">rdma</span></code> one-sided component instead. The
<code class="docutils literal notranslate"><span class="pre">rdma</span></code> component will use BTL components — such as the
TCP BTL — to effect one-sided communications.</p></li>
</ul>
</li>
<li><p>Updated the <code class="docutils literal notranslate"><span class="pre">tcp</span></code> BTL to use graph solving for global
interface matching between peers in order to improve
<code class="docutils literal notranslate"><span class="pre">MPI_Init()</span></code> wireup performance.</p></li>
<li><p>OFI</p>
<ul class="simple">
<li><p>Improved support for the HPE SS11 network.</p></li>
<li><p>Added cache bypass mechanism. This fixes conflicts with
<a class="reference external" href="https://libfabric.org/">Libfabric</a>, which has its own
registration cache. This adds a bypass flag which can be used
for providers known to have their own registration cache.</p></li>
</ul>
</li>
<li><p>Shared Memory:</p>
<ul class="simple">
<li><p>Update the new <code class="docutils literal notranslate"><span class="pre">sm</span></code> BTL to not use Linux Cross Memory Attach
(CMA) in user namespaces.</p></li>
<li><p>Fixed a crash when using the new <code class="docutils literal notranslate"><span class="pre">sm</span></code> BTL when compiled with
Linux Cross Memory Attach (<code class="docutils literal notranslate"><span class="pre">XPMEM</span></code>). Thanks to George
Katevenis for reporting this issue.</p></li>
</ul>
</li>
<li><p>Updated the <code class="docutils literal notranslate"><span class="pre">-mca</span> <span class="pre">pml</span></code> option to only accept one PML, not a list.</p></li>
</ul>
</li>
<li><p>Deprecations and removals:</p>
<ul class="simple">
<li><p>The legacy <code class="docutils literal notranslate"><span class="pre">sm</span></code> (shared memory) BTL has been removed. The
next-generation shared memory BTL <code class="docutils literal notranslate"><span class="pre">vader</span></code> replaces it, and
has been renamed to be <code class="docutils literal notranslate"><span class="pre">sm</span></code> (<code class="docutils literal notranslate"><span class="pre">vader</span></code> will still work as an
alias).</p></li>
<li><p>ORTE, the underlying Open MPI launcher has been removed, and
replaced with the <a class="reference external" href="https://github.com/openpmix/prrte">PMIx Reference RunTime Environment</a> (<code class="docutils literal notranslate"><span class="pre">PRTE</span></code>).</p></li>
<li><p>PMI support has been removed from Open MPI; now only PMIx is
supported. Thanks to Zach Osman for contributing.</p></li>
<li><p>The following components have been removed, and are replaced by
UCX support: PML <code class="docutils literal notranslate"><span class="pre">yalla</span></code>, PML <code class="docutils literal notranslate"><span class="pre">mxm</span></code>, SPML <code class="docutils literal notranslate"><span class="pre">ikrit</span></code>.</p></li>
<li><p>The MTL <code class="docutils literal notranslate"><span class="pre">psm</span></code> component has been removed and is no longer
supported.</p></li>
<li><p>Removed all vestiges of Checkpoint Restart (C/R) support.</p></li>
<li><p>32 bit atomics are now only supported via C11 compliant compilers.</p></li>
<li><p>Explicitly disable support for GNU gcc < v4.8.1 (note: the
default gcc compiler that is included in RHEL 7 is v4.8.5).</p></li>
<li><p>Various atomics support removed: S390/s390x, Sparc v9, ARMv4 and
ARMv5 with CMA support.</p></li>
<li><p>The MPI C++ bindings have been removed.</p></li>
<li><p>The <code class="docutils literal notranslate"><span class="pre">mpirun</span></code> options <code class="docutils literal notranslate"><span class="pre">--am</span></code> and <code class="docutils literal notranslate"><span class="pre">--amca</span></code> options have been
deprecated.</p></li>
<li><p>The <code class="docutils literal notranslate"><span class="pre">libompitrace</span></code> contributed library has been removed.
This library was incomplete and unmaintained. If needed, it
is available in the v4.x series.</p></li>
<li><p>The rankfile format no longer supports physical processor
locations. Only logical processor locations are supported.</p></li>
<li><p>32-bit builds have been disabled. Building Open MPI in a 32-bit
environment is no longer supported. 32 bit support is still
available in the v4.x series.</p></li>
</ul>
</li>
<li><p>Other updates and bug fixes:</p>
<ul>
<li><p>Updated Open MPI to use <code class="docutils literal notranslate"><span class="pre">ROMIO</span></code> v3.4.1.</p></li>
<li><p>Add missing <code class="docutils literal notranslate"><span class="pre">MPI_Status</span></code> conversion subroutines:
<code class="docutils literal notranslate"><span class="pre">MPI_Status_c2f08()</span></code>, <code class="docutils literal notranslate"><span class="pre">MPI_Status_f082c()</span></code>,
<code class="docutils literal notranslate"><span class="pre">MPI_Status_f082f()</span></code>, <code class="docutils literal notranslate"><span class="pre">MPI_Status_f2f08()</span></code> and the
<code class="docutils literal notranslate"><span class="pre">PMPI_*</span></code> related subroutines.</p></li>
<li><p>MPI module: added the <code class="docutils literal notranslate"><span class="pre">mpi_f08</span></code> <code class="docutils literal notranslate"><span class="pre">TYPE(MPI_*)</span></code> types for
Fortran. Thanks to George Katevenis for the report and their
contribution to the patch.</p></li>
<li><p>The default atomics have been changed to be GCC, with C11 as a
fallback. C11 atomics incurs sequential memory ordering, which
in most cases is not desired.</p></li>
<li><p>The default build mode has changed from building Open MPI’s
components as Dynamic Shared Objects (DSOs) to being statically
included in their respective libraries.</p>
<div class="admonition important">
<p class="admonition-title">Important</p>
<p>This has consequences for packagers. Be sure to
read the <a class="reference internal" href="../../installing-open-mpi/packagers.html#label-install-packagers-gnu-libtool-dependency-flattening"><span class="std std-ref">GNU Libtool dependency flattening</span></a>
subsection.</p>
</div>
</li>
<li><p>Various datatype bugfixes and performance improvements.</p></li>
<li><p>Various pack/unpack bugfixes and performance improvements.</p></li>
<li><p>Various OSHMEM bugfixes and performance improvements.</p></li>
<li><p>Thanks to Jeff Hammond, Pak Lui, Felix Uhl, Naribayashi Akira,
Julien Emmanuel, and Yaz Saito for their invaluable contributions.</p></li>
</ul>
</li>
<li><p>Documentation updates and improvements:</p>
<ul class="simple">
<li><p>Open MPI has consolidated and converted all of its documentation
to use <a class="reference external" href="https://www.sphinx-doc.org/en/master/usage/restructuredtext/basics.html">ReStructured Text.</a>
and <a class="reference external" href="https://www.sphinx-doc.org/">Sphinx</a>.</p>
<ul>
<li><p>The resulting documentation is now hosted on
<a class="reference external" href="https://docs.open-mpi.org">https://docs.open-mpi.org</a> (via <a class="reference external" href="https://ReadTheDocs.io/">ReadTheDocs</a>).</p></li>
<li><p>The documentation is also wholly available offline via Open
MPI distribution tarballs, in the <code class="docutils literal notranslate"><span class="pre">docs/_build/html</span></code>
directory.</p></li>
</ul>
</li>
<li><p>Many, many people from the Open MPI community contributed to the
overall documentation effort — not just those who are
listed in the Git commit logs. Indeed, many Open MPI core
developers contributed their time and effort, as did a fairly
large group of non-core developers (e.g., those who participated
just to help the documentation revamp), including (but not
limited to):</p>
<ul>
<li><p>Lachlan Bell</p></li>
<li><p>Simon Byrne</p></li>
<li><p>Samuel Cho</p></li>
<li><p>Tony Curtis</p></li>
<li><p>Lisandro Dalcin</p></li>
<li><p>Sophia Fang</p></li>
<li><p>Rick Gleitz</p></li>
<li><p>Colton Kammes</p></li>
<li><p>Robert Langfield</p></li>
<li><p>Nick Papior</p></li>
<li><p>Luz Paz</p></li>
<li><p>Alex Ross</p></li>
<li><p>Hao Tong</p></li>
<li><p>Mitchell Topaloglu</p></li>
<li><p>Siyu Wu</p></li>
<li><p>Fangcong Yin</p></li>
<li><p>Seth Zegelstein</p></li>
<li><p>Yixin Zhang</p></li>
</ul>
</li>
</ul>
</li>
</ul>
</li>
</ul>
</div>
</div>
</div>
</div>
<footer><div class="rst-footer-buttons" role="navigation" aria-label="Footer">
<a href="index.html" class="btn btn-neutral float-left" title="3.1. Changelog" accesskey="p" rel="prev"><span class="fa fa-arrow-circle-left" aria-hidden="true"></span> Previous</a>
<a href="v4.1.x.html" class="btn btn-neutral float-right" title="3.1.2. Open MPI v4.1.x series" accesskey="n" rel="next">Next <span class="fa fa-arrow-circle-right" aria-hidden="true"></span></a>
</div>
<hr/>
<div role="contentinfo">
<p>© Copyright 2003-2025, The Open MPI Community.
<span class="lastupdated">Last updated on 2025-05-30 16:41:43 UTC.
</span></p>
</div>
Built with <a href="https://www.sphinx-doc.org/">Sphinx</a> using a
<a href="https://github.com/readthedocs/sphinx_rtd_theme">theme</a>
provided by <a href="https://readthedocs.org">Read the Docs</a>.
</footer>
</div>
</div>
</section>
</div>
<script>
jQuery(function () {
SphinxRtdTheme.Navigation.enable(true);
});
</script>
</body>
</html>
|
