summaryrefslogtreecommitdiffstats
path: root/testpar/t_shapesame.c
blob: 98e307772a92117af9268eaac2e5c52ba6c63c82 (plain)
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
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
2073
2074
2075
2076
2077
2078
2079
2080
2081
2082
2083
2084
2085
2086
2087
2088
2089
2090
2091
2092
2093
2094
2095
2096
2097
2098
2099
2100
2101
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
2128
2129
2130
2131
2132
2133
2134
2135
2136
2137
2138
2139
2140
2141
2142
2143
2144
2145
2146
2147
2148
2149
2150
2151
2152
2153
2154
2155
2156
2157
2158
2159
2160
2161
2162
2163
2164
2165
2166
2167
2168
2169
2170
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
2184
2185
2186
2187
2188
2189
2190
2191
2192
2193
2194
2195
2196
2197
2198
2199
2200
2201
2202
2203
2204
2205
2206
2207
2208
2209
2210
2211
2212
2213
2214
2215
2216
2217
2218
2219
2220
2221
2222
2223
2224
2225
2226
2227
2228
2229
2230
2231
2232
2233
2234
2235
2236
2237
2238
2239
2240
2241
2242
2243
2244
2245
2246
2247
2248
2249
2250
2251
2252
2253
2254
2255
2256
2257
2258
2259
2260
2261
2262
2263
2264
2265
2266
2267
2268
2269
2270
2271
2272
2273
2274
2275
2276
2277
2278
2279
2280
2281
2282
2283
2284
2285
2286
2287
2288
2289
2290
2291
2292
2293
2294
2295
2296
2297
2298
2299
2300
2301
2302
2303
2304
2305
2306
2307
2308
2309
2310
2311
2312
2313
2314
2315
2316
2317
2318
2319
2320
2321
2322
2323
2324
2325
2326
2327
2328
2329
2330
2331
2332
2333
2334
2335
2336
2337
2338
2339
2340
2341
2342
2343
2344
2345
2346
2347
2348
2349
2350
2351
2352
2353
2354
2355
2356
2357
2358
2359
2360
2361
2362
2363
2364
2365
2366
2367
2368
2369
2370
2371
2372
2373
2374
2375
2376
2377
2378
2379
2380
2381
2382
2383
2384
2385
2386
2387
2388
2389
2390
2391
2392
2393
2394
2395
2396
2397
2398
2399
2400
2401
2402
2403
2404
2405
2406
2407
2408
2409
2410
2411
2412
2413
2414
2415
2416
2417
2418
2419
2420
2421
2422
2423
2424
2425
2426
2427
2428
2429
2430
2431
2432
2433
2434
2435
2436
2437
2438
2439
2440
2441
2442
2443
2444
2445
2446
2447
2448
2449
2450
2451
2452
2453
2454
2455
2456
2457
2458
2459
2460
2461
2462
2463
2464
2465
2466
2467
2468
2469
2470
2471
2472
2473
2474
2475
2476
2477
2478
2479
2480
2481
2482
2483
2484
2485
2486
2487
2488
2489
2490
2491
2492
2493
2494
2495
2496
2497
2498
2499
2500
2501
2502
2503
2504
2505
2506
2507
2508
2509
2510
2511
2512
2513
2514
2515
2516
2517
2518
2519
2520
2521
2522
2523
2524
2525
2526
2527
2528
2529
2530
2531
2532
2533
2534
2535
2536
2537
2538
2539
2540
2541
2542
2543
2544
2545
2546
2547
2548
2549
2550
2551
2552
2553
2554
2555
2556
2557
2558
2559
2560
2561
2562
2563
2564
2565
2566
2567
2568
2569
2570
2571
2572
2573
2574
2575
2576
2577
2578
2579
2580
2581
2582
2583
2584
2585
2586
2587
2588
2589
2590
2591
2592
2593
2594
2595
2596
2597
2598
2599
2600
2601
2602
2603
2604
2605
2606
2607
2608
2609
2610
2611
2612
2613
2614
2615
2616
2617
2618
2619
2620
2621
2622
2623
2624
2625
2626
2627
2628
2629
2630
2631
2632
2633
2634
2635
2636
2637
2638
2639
2640
2641
2642
2643
2644
2645
2646
2647
2648
2649
2650
2651
2652
2653
2654
2655
2656
2657
2658
2659
2660
2661
2662
2663
2664
2665
2666
2667
2668
2669
2670
2671
2672
2673
2674
2675
2676
2677
2678
2679
2680
2681
2682
2683
2684
2685
2686
2687
2688
2689
2690
2691
2692
2693
2694
2695
2696
2697
2698
2699
2700
2701
2702
2703
2704
2705
2706
2707
2708
2709
2710
2711
2712
2713
2714
2715
2716
2717
2718
2719
2720
2721
2722
2723
2724
2725
2726
2727
2728
2729
2730
2731
2732
2733
2734
2735
2736
2737
2738
2739
2740
2741
2742
2743
2744
2745
2746
2747
2748
2749
2750
2751
2752
2753
2754
2755
2756
2757
2758
2759
2760
2761
2762
2763
2764
2765
2766
2767
2768
2769
2770
2771
2772
2773
2774
2775
2776
2777
2778
2779
2780
2781
2782
2783
2784
2785
2786
2787
2788
2789
2790
2791
2792
2793
2794
2795
2796
2797
2798
2799
2800
2801
2802
2803
2804
2805
2806
2807
2808
2809
2810
2811
2812
2813
2814
2815
2816
2817
2818
2819
2820
2821
2822
2823
2824
2825
2826
2827
2828
2829
2830
2831
2832
2833
2834
2835
2836
2837
2838
2839
2840
2841
2842
2843
2844
2845
2846
2847
2848
2849
2850
2851
2852
2853
2854
2855
2856
2857
2858
2859
2860
2861
2862
2863
2864
2865
2866
2867
2868
2869
2870
2871
2872
2873
2874
2875
2876
2877
2878
2879
2880
2881
2882
2883
2884
2885
2886
2887
2888
2889
2890
2891
2892
2893
2894
2895
2896
2897
2898
2899
2900
2901
2902
2903
2904
2905
2906
2907
2908
2909
2910
2911
2912
2913
2914
2915
2916
2917
2918
2919
2920
2921
2922
2923
2924
2925
2926
2927
2928
2929
2930
2931
2932
2933
2934
2935
2936
2937
2938
2939
2940
2941
2942
2943
2944
2945
2946
2947
2948
2949
2950
2951
2952
2953
2954
2955
2956
2957
2958
2959
2960
2961
2962
2963
2964
2965
2966
2967
2968
2969
2970
2971
2972
2973
2974
2975
2976
2977
2978
2979
2980
2981
2982
2983
2984
2985
2986
2987
2988
2989
2990
2991
2992
2993
2994
2995
2996
2997
2998
2999
3000
3001
3002
3003
3004
3005
3006
3007
3008
3009
3010
3011
3012
3013
3014
3015
3016
3017
3018
3019
3020
3021
3022
3023
3024
3025
3026
3027
3028
3029
3030
3031
3032
3033
3034
3035
3036
3037
3038
3039
3040
3041
3042
3043
3044
3045
3046
3047
3048
3049
3050
3051
3052
3053
3054
3055
3056
3057
3058
3059
3060
3061
3062
3063
3064
3065
3066
3067
3068
3069
3070
3071
3072
3073
3074
3075
3076
3077
3078
3079
3080
3081
3082
3083
3084
3085
3086
3087
3088
3089
3090
3091
3092
3093
3094
3095
3096
3097
3098
3099
3100
3101
3102
3103
3104
3105
3106
3107
3108
3109
3110
3111
3112
3113
3114
3115
3116
3117
3118
3119
3120
3121
3122
3123
3124
3125
3126
3127
3128
3129
3130
3131
3132
3133
3134
3135
3136
3137
3138
3139
3140
3141
3142
3143
3144
3145
3146
3147
3148
3149
3150
3151
3152
3153
3154
3155
3156
3157
3158
3159
3160
3161
3162
3163
3164
3165
3166
3167
3168
3169
3170
3171
3172
3173
3174
3175
3176
3177
3178
3179
3180
3181
3182
3183
3184
3185
3186
3187
3188
3189
3190
3191
3192
3193
3194
3195
3196
3197
3198
3199
3200
3201
3202
3203
3204
3205
3206
3207
3208
3209
3210
3211
3212
3213
3214
3215
3216
3217
3218
3219
3220
3221
3222
3223
3224
3225
3226
3227
3228
3229
3230
3231
3232
3233
3234
3235
3236
3237
3238
3239
3240
3241
3242
3243
3244
3245
3246
3247
3248
3249
3250
3251
3252
3253
3254
3255
3256
3257
3258
3259
3260
3261
3262
3263
3264
3265
3266
3267
3268
3269
3270
3271
3272
3273
3274
3275
3276
3277
3278
3279
3280
3281
3282
3283
3284
3285
3286
3287
3288
3289
3290
3291
3292
3293
3294
3295
3296
3297
3298
3299
3300
3301
3302
3303
3304
3305
3306
3307
3308
3309
3310
3311
3312
3313
3314
3315
3316
3317
3318
3319
3320
3321
3322
3323
3324
3325
3326
3327
3328
3329
3330
3331
3332
3333
3334
3335
3336
3337
3338
3339
3340
3341
3342
3343
3344
3345
3346
3347
3348
3349
3350
3351
3352
3353
3354
3355
3356
3357
3358
3359
3360
3361
3362
3363
3364
3365
3366
3367
3368
3369
3370
3371
3372
3373
3374
3375
3376
3377
3378
3379
3380
3381
3382
3383
3384
3385
3386
3387
3388
3389
3390
3391
3392
3393
3394
3395
3396
3397
3398
3399
3400
3401
3402
3403
3404
3405
3406
3407
3408
3409
3410
3411
3412
3413
3414
3415
3416
3417
3418
3419
3420
3421
3422
3423
3424
3425
3426
3427
3428
3429
3430
3431
3432
3433
3434
3435
3436
3437
3438
3439
3440
3441
3442
3443
3444
3445
3446
3447
3448
3449
3450
3451
3452
3453
3454
3455
3456
3457
3458
3459
3460
3461
3462
3463
3464
3465
3466
3467
3468
3469
3470
3471
3472
3473
3474
3475
3476
3477
3478
3479
3480
3481
3482
3483
3484
3485
3486
3487
3488
3489
3490
3491
3492
3493
3494
3495
3496
3497
3498
3499
3500
3501
3502
3503
3504
3505
3506
3507
3508
3509
3510
3511
3512
3513
3514
3515
3516
3517
3518
3519
3520
3521
3522
3523
3524
3525
3526
3527
3528
3529
3530
3531
3532
3533
3534
3535
3536
3537
3538
3539
3540
3541
3542
3543
3544
3545
3546
3547
3548
3549
3550
3551
3552
3553
3554
3555
3556
3557
3558
3559
3560
3561
3562
3563
3564
3565
3566
3567
3568
3569
3570
3571
3572
3573
3574
3575
3576
3577
3578
3579
3580
3581
3582
3583
3584
3585
3586
3587
3588
3589
3590
3591
3592
3593
3594
3595
3596
3597
3598
3599
3600
3601
3602
3603
3604
3605
3606
3607
3608
3609
3610
3611
3612
3613
3614
3615
3616
3617
3618
3619
3620
3621
3622
3623
3624
3625
3626
3627
3628
3629
3630
3631
3632
3633
3634
3635
3636
3637
3638
3639
3640
3641
3642
3643
3644
3645
3646
3647
3648
3649
3650
3651
3652
3653
3654
3655
3656
3657
3658
3659
3660
3661
3662
3663
3664
3665
3666
3667
3668
3669
3670
3671
3672
3673
3674
3675
3676
3677
3678
3679
3680
3681
3682
3683
3684
3685
3686
3687
3688
3689
3690
3691
3692
3693
3694
3695
3696
3697
3698
3699
3700
3701
3702
3703
3704
3705
3706
3707
3708
3709
3710
3711
3712
3713
3714
3715
3716
3717
3718
3719
3720
3721
3722
3723
3724
3725
3726
3727
3728
3729
3730
3731
3732
3733
3734
3735
3736
3737
3738
3739
3740
3741
3742
3743
3744
3745
3746
3747
3748
3749
3750
3751
3752
3753
3754
3755
3756
3757
3758
3759
3760
3761
3762
3763
3764
3765
3766
3767
3768
3769
3770
3771
3772
3773
3774
3775
3776
3777
3778
3779
3780
3781
3782
3783
3784
3785
3786
3787
3788
3789
3790
3791
3792
3793
3794
3795
3796
3797
3798
3799
3800
3801
3802
3803
3804
3805
3806
3807
3808
3809
3810
3811
3812
3813
3814
3815
3816
3817
3818
3819
3820
3821
3822
3823
3824
3825
3826
3827
3828
3829
3830
3831
3832
3833
3834
3835
3836
3837
3838
3839
3840
3841
3842
3843
3844
3845
3846
3847
3848
3849
3850
3851
3852
3853
3854
3855
3856
3857
3858
3859
3860
3861
3862
3863
3864
3865
3866
3867
3868
3869
3870
3871
3872
3873
3874
3875
3876
3877
3878
3879
3880
3881
3882
3883
3884
3885
3886
3887
3888
3889
3890
3891
3892
3893
3894
3895
3896
3897
3898
3899
3900
3901
3902
3903
3904
3905
3906
3907
3908
3909
3910
3911
3912
3913
3914
3915
3916
3917
3918
3919
3920
3921
3922
3923
3924
3925
3926
3927
3928
3929
3930
3931
3932
3933
3934
3935
3936
3937
3938
3939
3940
3941
3942
3943
3944
3945
3946
3947
3948
3949
3950
3951
3952
3953
3954
3955
3956
3957
3958
3959
3960
3961
3962
3963
3964
3965
3966
3967
3968
3969
3970
3971
3972
3973
3974
3975
3976
3977
3978
3979
3980
3981
3982
3983
3984
3985
3986
3987
3988
3989
3990
3991
3992
3993
3994
3995
3996
3997
3998
3999
4000
4001
4002
4003
4004
4005
4006
4007
4008
4009
4010
4011
4012
4013
4014
4015
4016
4017
4018
4019
4020
4021
4022
4023
4024
4025
4026
4027
4028
4029
4030
4031
4032
4033
4034
4035
4036
4037
4038
4039
4040
4041
4042
4043
4044
4045
4046
4047
4048
4049
4050
4051
4052
4053
4054
4055
4056
4057
4058
4059
4060
4061
4062
4063
4064
4065
4066
4067
4068
4069
4070
4071
4072
4073
4074
4075
4076
4077
4078
4079
4080
4081
4082
4083
4084
4085
4086
4087
4088
4089
4090
4091
4092
4093
4094
4095
4096
4097
4098
4099
4100
4101
4102
4103
4104
4105
4106
4107
4108
4109
4110
4111
4112
4113
4114
4115
4116
4117
4118
4119
4120
4121
4122
4123
4124
4125
4126
4127
4128
4129
4130
4131
4132
4133
4134
4135
4136
4137
4138
4139
4140
4141
4142
4143
4144
4145
4146
4147
4148
4149
4150
4151
4152
4153
4154
4155
4156
4157
4158
4159
4160
4161
4162
4163
4164
4165
4166
4167
4168
4169
4170
4171
4172
4173
4174
4175
4176
4177
4178
4179
4180
4181
4182
4183
4184
4185
4186
4187
4188
4189
4190
4191
4192
4193
4194
4195
4196
4197
4198
4199
4200
4201
4202
4203
4204
4205
4206
4207
4208
4209
4210
4211
4212
4213
4214
4215
4216
4217
4218
4219
4220
4221
4222
4223
4224
4225
4226
4227
4228
4229
4230
4231
4232
4233
4234
4235
4236
4237
4238
4239
4240
4241
4242
4243
4244
4245
4246
4247
4248
4249
4250
4251
4252
4253
4254
4255
4256
4257
4258
4259
4260
4261
4262
4263
4264
4265
4266
4267
4268
4269
4270
4271
4272
4273
4274
4275
4276
4277
4278
4279
4280
4281
4282
4283
4284
4285
4286
4287
4288
4289
4290
4291
4292
4293
4294
4295
4296
4297
4298
4299
4300
4301
4302
4303
4304
4305
4306
4307
4308
4309
4310
4311
4312
4313
4314
4315
4316
4317
4318
4319
4320
4321
4322
4323
4324
4325
4326
4327
4328
4329
4330
4331
4332
4333
4334
4335
4336
4337
4338
4339
4340
4341
4342
4343
4344
4345
4346
4347
4348
4349
4350
4351
4352
4353
4354
4355
4356
4357
4358
4359
4360
4361
4362
4363
4364
4365
4366
4367
4368
4369
4370
4371
4372
4373
4374
/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
 * Copyright by The HDF Group.                                               *
 * All rights reserved.                                                      *
 *                                                                           *
 * This file is part of HDF5.  The full HDF5 copyright notice, including     *
 * terms governing use, modification, and redistribution, is contained in    *
 * the COPYING file, which can be found at the root of the source code       *
 * distribution tree, or in https://www.hdfgroup.org/licenses.               *
 * If you do not have access to either file, you may request a copy from     *
 * help@hdfgroup.org.                                                        *
 * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */

/*
   This program will test independent and collective reads and writes between
   selections of different rank that non-the-less are deemed as having the
   same shape by H5Sselect_shape_same().
 */

#define H5S_FRIEND /*suppress error about including H5Spkg   */

/* Define this macro to indicate that the testing APIs should be available */
#define H5S_TESTING

#include "H5Spkg.h" /* Dataspaces                           */
#include "testphdf5.h"

/* On Lustre (and perhaps other parallel file systems?), we have severe
 * slow downs if two or more processes attempt to access the same file system
 * block.  To minimize this problem, we set alignment in the shape same tests
 * to the default Lustre block size -- which greatly reduces contention in
 * the chunked dataset case.
 */

#define SHAPE_SAME_TEST_ALIGNMENT ((hsize_t)(4 * 1024 * 1024))

#define PAR_SS_DR_MAX_RANK 5 /* must update code if this changes */

struct hs_dr_pio_test_vars_t {
    int       mpi_size;
    int       mpi_rank;
    MPI_Comm  mpi_comm;
    MPI_Info  mpi_info;
    int       test_num;
    int       edge_size;
    int       checker_edge_size;
    int       chunk_edge_size;
    int       small_rank;
    int       large_rank;
    hid_t     dset_type;
    uint32_t *small_ds_buf_0;
    uint32_t *small_ds_buf_1;
    uint32_t *small_ds_buf_2;
    uint32_t *small_ds_slice_buf;
    uint32_t *large_ds_buf_0;
    uint32_t *large_ds_buf_1;
    uint32_t *large_ds_buf_2;
    uint32_t *large_ds_slice_buf;
    int       small_ds_offset;
    int       large_ds_offset;
    hid_t     fid; /* HDF5 file ID */
    hid_t     xfer_plist;
    hid_t     full_mem_small_ds_sid;
    hid_t     full_file_small_ds_sid;
    hid_t     mem_small_ds_sid;
    hid_t     file_small_ds_sid_0;
    hid_t     file_small_ds_sid_1;
    hid_t     small_ds_slice_sid;
    hid_t     full_mem_large_ds_sid;
    hid_t     full_file_large_ds_sid;
    hid_t     mem_large_ds_sid;
    hid_t     file_large_ds_sid_0;
    hid_t     file_large_ds_sid_1;
    hid_t     file_large_ds_process_slice_sid;
    hid_t     mem_large_ds_process_slice_sid;
    hid_t     large_ds_slice_sid;
    hid_t     small_dataset; /* Dataset ID */
    hid_t     large_dataset; /* Dataset ID */
    size_t    small_ds_size;
    size_t    small_ds_slice_size;
    size_t    large_ds_size;
    size_t    large_ds_slice_size;
    hsize_t   dims[PAR_SS_DR_MAX_RANK];
    hsize_t   chunk_dims[PAR_SS_DR_MAX_RANK];
    hsize_t   start[PAR_SS_DR_MAX_RANK];
    hsize_t   stride[PAR_SS_DR_MAX_RANK];
    hsize_t   count[PAR_SS_DR_MAX_RANK];
    hsize_t   block[PAR_SS_DR_MAX_RANK];
    hsize_t  *start_ptr;
    hsize_t  *stride_ptr;
    hsize_t  *count_ptr;
    hsize_t  *block_ptr;
    int       skips;
    int       max_skips;
    int64_t   total_tests;
    int64_t   tests_run;
    int64_t   tests_skipped;
};

/*-------------------------------------------------------------------------
 * Function:    hs_dr_pio_test__setup()
 *
 * Purpose:    Do setup for tests of I/O to/from hyperslab selections of
 *         different rank in the parallel case.
 *
 * Return:    void
 *
 *-------------------------------------------------------------------------
 */

#define CONTIG_HS_DR_PIO_TEST__SETUP__DEBUG 0

static void
hs_dr_pio_test__setup(const int test_num, const int edge_size, const int checker_edge_size,
                      const int chunk_edge_size, const int small_rank, const int large_rank,
                      const bool use_collective_io, const hid_t dset_type, const int express_test,
                      struct hs_dr_pio_test_vars_t *tv_ptr)
{
#if CONTIG_HS_DR_PIO_TEST__SETUP__DEBUG
    const char *fcnName = "hs_dr_pio_test__setup()";
#endif /* CONTIG_HS_DR_PIO_TEST__SETUP__DEBUG */
    const char *filename;
    bool        mis_match = false;
    int         i;
    int         mrc;
    int         mpi_rank; /* needed by the VRFY macro */
    uint32_t    expected_value;
    uint32_t   *ptr_0;
    uint32_t   *ptr_1;
    hid_t       acc_tpl; /* File access templates */
    hid_t       small_ds_dcpl_id = H5P_DEFAULT;
    hid_t       large_ds_dcpl_id = H5P_DEFAULT;
    herr_t      ret; /* Generic return value */

    assert(edge_size >= 6);
    assert(edge_size >= chunk_edge_size);
    assert((chunk_edge_size == 0) || (chunk_edge_size >= 3));
    assert(1 < small_rank);
    assert(small_rank < large_rank);
    assert(large_rank <= PAR_SS_DR_MAX_RANK);

    tv_ptr->test_num          = test_num;
    tv_ptr->edge_size         = edge_size;
    tv_ptr->checker_edge_size = checker_edge_size;
    tv_ptr->chunk_edge_size   = chunk_edge_size;
    tv_ptr->small_rank        = small_rank;
    tv_ptr->large_rank        = large_rank;
    tv_ptr->dset_type         = dset_type;

    MPI_Comm_size(MPI_COMM_WORLD, &(tv_ptr->mpi_size));
    MPI_Comm_rank(MPI_COMM_WORLD, &(tv_ptr->mpi_rank));
    /* the VRFY() macro needs the local variable mpi_rank -- set it up now */
    mpi_rank = tv_ptr->mpi_rank;

    assert(tv_ptr->mpi_size >= 1);

    tv_ptr->mpi_comm = MPI_COMM_WORLD;
    tv_ptr->mpi_info = MPI_INFO_NULL;

    for (i = 0; i < tv_ptr->small_rank - 1; i++) {
        tv_ptr->small_ds_size *= (size_t)(tv_ptr->edge_size);
        tv_ptr->small_ds_slice_size *= (size_t)(tv_ptr->edge_size);
    }
    tv_ptr->small_ds_size *= (size_t)(tv_ptr->mpi_size + 1);

    /* used by checker board tests only */
    tv_ptr->small_ds_offset = PAR_SS_DR_MAX_RANK - tv_ptr->small_rank;

    assert(0 < tv_ptr->small_ds_offset);
    assert(tv_ptr->small_ds_offset < PAR_SS_DR_MAX_RANK);

    for (i = 0; i < tv_ptr->large_rank - 1; i++) {

        tv_ptr->large_ds_size *= (size_t)(tv_ptr->edge_size);
        tv_ptr->large_ds_slice_size *= (size_t)(tv_ptr->edge_size);
    }
    tv_ptr->large_ds_size *= (size_t)(tv_ptr->mpi_size + 1);

    /* used by checker board tests only */
    tv_ptr->large_ds_offset = PAR_SS_DR_MAX_RANK - tv_ptr->large_rank;

    assert(0 <= tv_ptr->large_ds_offset);
    assert(tv_ptr->large_ds_offset < PAR_SS_DR_MAX_RANK);

    /* set up the start, stride, count, and block pointers */
    /* used by contiguous tests only */
    tv_ptr->start_ptr  = &(tv_ptr->start[PAR_SS_DR_MAX_RANK - tv_ptr->large_rank]);
    tv_ptr->stride_ptr = &(tv_ptr->stride[PAR_SS_DR_MAX_RANK - tv_ptr->large_rank]);
    tv_ptr->count_ptr  = &(tv_ptr->count[PAR_SS_DR_MAX_RANK - tv_ptr->large_rank]);
    tv_ptr->block_ptr  = &(tv_ptr->block[PAR_SS_DR_MAX_RANK - tv_ptr->large_rank]);

    /* Allocate buffers */
    tv_ptr->small_ds_buf_0 = (uint32_t *)malloc(sizeof(uint32_t) * tv_ptr->small_ds_size);
    VRFY((tv_ptr->small_ds_buf_0 != NULL), "malloc of small_ds_buf_0 succeeded");

    tv_ptr->small_ds_buf_1 = (uint32_t *)malloc(sizeof(uint32_t) * tv_ptr->small_ds_size);
    VRFY((tv_ptr->small_ds_buf_1 != NULL), "malloc of small_ds_buf_1 succeeded");

    tv_ptr->small_ds_buf_2 = (uint32_t *)malloc(sizeof(uint32_t) * tv_ptr->small_ds_size);
    VRFY((tv_ptr->small_ds_buf_2 != NULL), "malloc of small_ds_buf_2 succeeded");

    tv_ptr->small_ds_slice_buf = (uint32_t *)malloc(sizeof(uint32_t) * tv_ptr->small_ds_slice_size);
    VRFY((tv_ptr->small_ds_slice_buf != NULL), "malloc of small_ds_slice_buf succeeded");

    tv_ptr->large_ds_buf_0 = (uint32_t *)malloc(sizeof(uint32_t) * tv_ptr->large_ds_size);
    VRFY((tv_ptr->large_ds_buf_0 != NULL), "malloc of large_ds_buf_0 succeeded");

    tv_ptr->large_ds_buf_1 = (uint32_t *)malloc(sizeof(uint32_t) * tv_ptr->large_ds_size);
    VRFY((tv_ptr->large_ds_buf_1 != NULL), "malloc of large_ds_buf_1 succeeded");

    tv_ptr->large_ds_buf_2 = (uint32_t *)malloc(sizeof(uint32_t) * tv_ptr->large_ds_size);
    VRFY((tv_ptr->large_ds_buf_2 != NULL), "malloc of large_ds_buf_2 succeeded");

    tv_ptr->large_ds_slice_buf = (uint32_t *)malloc(sizeof(uint32_t) * tv_ptr->large_ds_slice_size);
    VRFY((tv_ptr->large_ds_slice_buf != NULL), "malloc of large_ds_slice_buf succeeded");

    /* initialize the buffers */

    ptr_0 = tv_ptr->small_ds_buf_0;
    for (i = 0; i < (int)(tv_ptr->small_ds_size); i++)
        *ptr_0++ = (uint32_t)i;
    memset(tv_ptr->small_ds_buf_1, 0, sizeof(uint32_t) * tv_ptr->small_ds_size);
    memset(tv_ptr->small_ds_buf_2, 0, sizeof(uint32_t) * tv_ptr->small_ds_size);

    memset(tv_ptr->small_ds_slice_buf, 0, sizeof(uint32_t) * tv_ptr->small_ds_slice_size);

    ptr_0 = tv_ptr->large_ds_buf_0;
    for (i = 0; i < (int)(tv_ptr->large_ds_size); i++)
        *ptr_0++ = (uint32_t)i;
    memset(tv_ptr->large_ds_buf_1, 0, sizeof(uint32_t) * tv_ptr->large_ds_size);
    memset(tv_ptr->large_ds_buf_2, 0, sizeof(uint32_t) * tv_ptr->large_ds_size);

    memset(tv_ptr->large_ds_slice_buf, 0, sizeof(uint32_t) * tv_ptr->large_ds_slice_size);

    filename = (const char *)GetTestParameters();
    assert(filename != NULL);
#if CONTIG_HS_DR_PIO_TEST__SETUP__DEBUG
    if (MAINPROCESS) {

        fprintf(stdout, "%d: test num = %d.\n", tv_ptr->mpi_rank, tv_ptr->test_num);
        fprintf(stdout, "%d: mpi_size = %d.\n", tv_ptr->mpi_rank, tv_ptr->mpi_size);
        fprintf(stdout, "%d: small/large rank = %d/%d, use_collective_io = %d.\n", tv_ptr->mpi_rank,
                tv_ptr->small_rank, tv_ptr->large_rank, (int)use_collective_io);
        fprintf(stdout, "%d: edge_size = %d, chunk_edge_size = %d.\n", tv_ptr->mpi_rank, tv_ptr->edge_size,
                tv_ptr->chunk_edge_size);
        fprintf(stdout, "%d: checker_edge_size = %d.\n", tv_ptr->mpi_rank, tv_ptr->checker_edge_size);
        fprintf(stdout, "%d: small_ds_size = %d, large_ds_size = %d.\n", tv_ptr->mpi_rank,
                (int)(tv_ptr->small_ds_size), (int)(tv_ptr->large_ds_size));
        fprintf(stdout, "%d: filename = %s.\n", tv_ptr->mpi_rank, filename);
    }
#endif /* CONTIG_HS_DR_PIO_TEST__SETUP__DEBUG */
    /* ----------------------------------------
     * CREATE AN HDF5 FILE WITH PARALLEL ACCESS
     * ---------------------------------------*/
    /* setup file access template */
    acc_tpl = create_faccess_plist(tv_ptr->mpi_comm, tv_ptr->mpi_info, facc_type);
    VRFY((acc_tpl >= 0), "create_faccess_plist() succeeded");

    /* set the alignment -- need it large so that we aren't always hitting the
     * the same file system block.  Do this only if express_test is greater
     * than zero.
     */
    if (express_test > 0) {

        ret = H5Pset_alignment(acc_tpl, (hsize_t)0, SHAPE_SAME_TEST_ALIGNMENT);
        VRFY((ret != FAIL), "H5Pset_alignment() succeeded");
    }

    /* create the file collectively */
    tv_ptr->fid = H5Fcreate(filename, H5F_ACC_TRUNC, H5P_DEFAULT, acc_tpl);
    VRFY((tv_ptr->fid >= 0), "H5Fcreate succeeded");

    MESG("File opened.");

    /* Release file-access template */
    ret = H5Pclose(acc_tpl);
    VRFY((ret >= 0), "H5Pclose(acc_tpl) succeeded");

    /* setup dims: */
    tv_ptr->dims[0] = (hsize_t)(tv_ptr->mpi_size + 1);
    tv_ptr->dims[1] = tv_ptr->dims[2] = tv_ptr->dims[3] = tv_ptr->dims[4] = (hsize_t)(tv_ptr->edge_size);

    /* Create small ds dataspaces */
    tv_ptr->full_mem_small_ds_sid = H5Screate_simple(tv_ptr->small_rank, tv_ptr->dims, NULL);
    VRFY((tv_ptr->full_mem_small_ds_sid != 0), "H5Screate_simple() full_mem_small_ds_sid succeeded");

    tv_ptr->full_file_small_ds_sid = H5Screate_simple(tv_ptr->small_rank, tv_ptr->dims, NULL);
    VRFY((tv_ptr->full_file_small_ds_sid != 0), "H5Screate_simple() full_file_small_ds_sid succeeded");

    tv_ptr->mem_small_ds_sid = H5Screate_simple(tv_ptr->small_rank, tv_ptr->dims, NULL);
    VRFY((tv_ptr->mem_small_ds_sid != 0), "H5Screate_simple() mem_small_ds_sid succeeded");

    tv_ptr->file_small_ds_sid_0 = H5Screate_simple(tv_ptr->small_rank, tv_ptr->dims, NULL);
    VRFY((tv_ptr->file_small_ds_sid_0 != 0), "H5Screate_simple() file_small_ds_sid_0 succeeded");

    /* used by checker board tests only */
    tv_ptr->file_small_ds_sid_1 = H5Screate_simple(tv_ptr->small_rank, tv_ptr->dims, NULL);
    VRFY((tv_ptr->file_small_ds_sid_1 != 0), "H5Screate_simple() file_small_ds_sid_1 succeeded");

    tv_ptr->small_ds_slice_sid = H5Screate_simple(tv_ptr->small_rank - 1, &(tv_ptr->dims[1]), NULL);
    VRFY((tv_ptr->small_ds_slice_sid != 0), "H5Screate_simple() small_ds_slice_sid succeeded");

    /* Create large ds dataspaces */
    tv_ptr->full_mem_large_ds_sid = H5Screate_simple(tv_ptr->large_rank, tv_ptr->dims, NULL);
    VRFY((tv_ptr->full_mem_large_ds_sid != 0), "H5Screate_simple() full_mem_large_ds_sid succeeded");

    tv_ptr->full_file_large_ds_sid = H5Screate_simple(tv_ptr->large_rank, tv_ptr->dims, NULL);
    VRFY((tv_ptr->full_file_large_ds_sid != FAIL), "H5Screate_simple() full_file_large_ds_sid succeeded");

    tv_ptr->mem_large_ds_sid = H5Screate_simple(tv_ptr->large_rank, tv_ptr->dims, NULL);
    VRFY((tv_ptr->mem_large_ds_sid != FAIL), "H5Screate_simple() mem_large_ds_sid succeeded");

    tv_ptr->file_large_ds_sid_0 = H5Screate_simple(tv_ptr->large_rank, tv_ptr->dims, NULL);
    VRFY((tv_ptr->file_large_ds_sid_0 != FAIL), "H5Screate_simple() file_large_ds_sid_0 succeeded");

    /* used by checker board tests only */
    tv_ptr->file_large_ds_sid_1 = H5Screate_simple(tv_ptr->large_rank, tv_ptr->dims, NULL);
    VRFY((tv_ptr->file_large_ds_sid_1 != FAIL), "H5Screate_simple() file_large_ds_sid_1 succeeded");

    tv_ptr->mem_large_ds_process_slice_sid = H5Screate_simple(tv_ptr->large_rank, tv_ptr->dims, NULL);
    VRFY((tv_ptr->mem_large_ds_process_slice_sid != FAIL),
         "H5Screate_simple() mem_large_ds_process_slice_sid succeeded");

    tv_ptr->file_large_ds_process_slice_sid = H5Screate_simple(tv_ptr->large_rank, tv_ptr->dims, NULL);
    VRFY((tv_ptr->file_large_ds_process_slice_sid != FAIL),
         "H5Screate_simple() file_large_ds_process_slice_sid succeeded");

    tv_ptr->large_ds_slice_sid = H5Screate_simple(tv_ptr->large_rank - 1, &(tv_ptr->dims[1]), NULL);
    VRFY((tv_ptr->large_ds_slice_sid != 0), "H5Screate_simple() large_ds_slice_sid succeeded");

    /* if chunk edge size is greater than zero, set up the small and
     * large data set creation property lists to specify chunked
     * datasets.
     */
    if (tv_ptr->chunk_edge_size > 0) {

        /* Under Lustre (and perhaps other parallel file systems?) we get
         * locking delays when two or more processes attempt to access the
         * same file system block.
         *
         * To minimize this problem, I have changed chunk_dims[0]
         * from (mpi_size + 1) to just when any sort of express test is
         * selected.  Given the structure of the test, and assuming we
         * set the alignment large enough, this avoids the contention
         * issue by seeing to it that each chunk is only accessed by one
         * process.
         *
         * One can argue as to whether this is a good thing to do in our
         * tests, but for now it is necessary if we want the test to complete
         * in a reasonable amount of time.
         *
         *                                         JRM -- 9/16/10
         */

        tv_ptr->chunk_dims[0] = 1;

        tv_ptr->chunk_dims[1] = tv_ptr->chunk_dims[2] = tv_ptr->chunk_dims[3] = tv_ptr->chunk_dims[4] =
            (hsize_t)(tv_ptr->chunk_edge_size);

        small_ds_dcpl_id = H5Pcreate(H5P_DATASET_CREATE);
        VRFY((ret != FAIL), "H5Pcreate() small_ds_dcpl_id succeeded");

        ret = H5Pset_layout(small_ds_dcpl_id, H5D_CHUNKED);
        VRFY((ret != FAIL), "H5Pset_layout() small_ds_dcpl_id succeeded");

        ret = H5Pset_chunk(small_ds_dcpl_id, tv_ptr->small_rank, tv_ptr->chunk_dims);
        VRFY((ret != FAIL), "H5Pset_chunk() small_ds_dcpl_id succeeded");

        large_ds_dcpl_id = H5Pcreate(H5P_DATASET_CREATE);
        VRFY((ret != FAIL), "H5Pcreate() large_ds_dcpl_id succeeded");

        ret = H5Pset_layout(large_ds_dcpl_id, H5D_CHUNKED);
        VRFY((ret != FAIL), "H5Pset_layout() large_ds_dcpl_id succeeded");

        ret = H5Pset_chunk(large_ds_dcpl_id, tv_ptr->large_rank, tv_ptr->chunk_dims);
        VRFY((ret != FAIL), "H5Pset_chunk() large_ds_dcpl_id succeeded");
    }

    /* create the small dataset */
    tv_ptr->small_dataset =
        H5Dcreate2(tv_ptr->fid, "small_dataset", tv_ptr->dset_type, tv_ptr->file_small_ds_sid_0, H5P_DEFAULT,
                   small_ds_dcpl_id, H5P_DEFAULT);
    VRFY((ret != FAIL), "H5Dcreate2() small_dataset succeeded");

    /* create the large dataset */
    tv_ptr->large_dataset =
        H5Dcreate2(tv_ptr->fid, "large_dataset", tv_ptr->dset_type, tv_ptr->file_large_ds_sid_0, H5P_DEFAULT,
                   large_ds_dcpl_id, H5P_DEFAULT);
    VRFY((ret != FAIL), "H5Dcreate2() large_dataset succeeded");

    /* setup xfer property list */
    tv_ptr->xfer_plist = H5Pcreate(H5P_DATASET_XFER);
    VRFY((tv_ptr->xfer_plist >= 0), "H5Pcreate(H5P_DATASET_XFER) succeeded");

    if (use_collective_io) {
        ret = H5Pset_dxpl_mpio(tv_ptr->xfer_plist, H5FD_MPIO_COLLECTIVE);
        VRFY((ret >= 0), "H5Pset_dxpl_mpio succeeded");
    }

    /* setup selection to write initial data to the small and large data sets */
    tv_ptr->start[0]  = (hsize_t)(tv_ptr->mpi_rank);
    tv_ptr->stride[0] = (hsize_t)(2 * (tv_ptr->mpi_size + 1));
    tv_ptr->count[0]  = 1;
    tv_ptr->block[0]  = 1;

    for (i = 1; i < tv_ptr->large_rank; i++) {

        tv_ptr->start[i]  = 0;
        tv_ptr->stride[i] = (hsize_t)(2 * tv_ptr->edge_size);
        tv_ptr->count[i]  = 1;
        tv_ptr->block[i]  = (hsize_t)(tv_ptr->edge_size);
    }

    /* setup selections for writing initial data to the small data set */
    ret = H5Sselect_hyperslab(tv_ptr->mem_small_ds_sid, H5S_SELECT_SET, tv_ptr->start, tv_ptr->stride,
                              tv_ptr->count, tv_ptr->block);
    VRFY((ret >= 0), "H5Sselect_hyperslab(mem_small_ds_sid, set) succeeded");

    ret = H5Sselect_hyperslab(tv_ptr->file_small_ds_sid_0, H5S_SELECT_SET, tv_ptr->start, tv_ptr->stride,
                              tv_ptr->count, tv_ptr->block);
    VRFY((ret >= 0), "H5Sselect_hyperslab(file_small_ds_sid_0, set) succeeded");

    if (MAINPROCESS) { /* add an additional slice to the selections */

        tv_ptr->start[0] = (hsize_t)(tv_ptr->mpi_size);

        ret = H5Sselect_hyperslab(tv_ptr->mem_small_ds_sid, H5S_SELECT_OR, tv_ptr->start, tv_ptr->stride,
                                  tv_ptr->count, tv_ptr->block);
        VRFY((ret >= 0), "H5Sselect_hyperslab(mem_small_ds_sid, or) succeeded");

        ret = H5Sselect_hyperslab(tv_ptr->file_small_ds_sid_0, H5S_SELECT_OR, tv_ptr->start, tv_ptr->stride,
                                  tv_ptr->count, tv_ptr->block);
        VRFY((ret >= 0), "H5Sselect_hyperslab(file_small_ds_sid_0, or) succeeded");
    }

    /* write the initial value of the small data set to file */
    ret = H5Dwrite(tv_ptr->small_dataset, tv_ptr->dset_type, tv_ptr->mem_small_ds_sid,
                   tv_ptr->file_small_ds_sid_0, tv_ptr->xfer_plist, tv_ptr->small_ds_buf_0);

    VRFY((ret >= 0), "H5Dwrite() small_dataset initial write succeeded");

    /* sync with the other processes before checking data */
    mrc = MPI_Barrier(MPI_COMM_WORLD);
    VRFY((mrc == MPI_SUCCESS), "Sync after small dataset writes");

    /* read the small data set back to verify that it contains the
     * expected data.  Note that each process reads in the entire
     * data set and verifies it.
     */
    ret = H5Dread(tv_ptr->small_dataset, H5T_NATIVE_UINT32, tv_ptr->full_mem_small_ds_sid,
                  tv_ptr->full_file_small_ds_sid, tv_ptr->xfer_plist, tv_ptr->small_ds_buf_1);
    VRFY((ret >= 0), "H5Dread() small_dataset initial read succeeded");

    /* verify that the correct data was written to the small data set */
    expected_value = 0;
    mis_match      = false;
    ptr_1          = tv_ptr->small_ds_buf_1;

    i = 0;
    for (i = 0; i < (int)(tv_ptr->small_ds_size); i++) {

        if (*ptr_1 != expected_value) {

            mis_match = true;
        }
        ptr_1++;
        expected_value++;
    }
    VRFY((mis_match == false), "small ds init data good.");

    /* setup selections for writing initial data to the large data set */

    tv_ptr->start[0] = (hsize_t)(tv_ptr->mpi_rank);

    ret = H5Sselect_hyperslab(tv_ptr->mem_large_ds_sid, H5S_SELECT_SET, tv_ptr->start, tv_ptr->stride,
                              tv_ptr->count, tv_ptr->block);
    VRFY((ret >= 0), "H5Sselect_hyperslab(mem_large_ds_sid, set) succeeded");

    ret = H5Sselect_hyperslab(tv_ptr->file_large_ds_sid_0, H5S_SELECT_SET, tv_ptr->start, tv_ptr->stride,
                              tv_ptr->count, tv_ptr->block);
    VRFY((ret >= 0), "H5Sselect_hyperslab(file_large_ds_sid_0, set) succeeded");

    /* In passing, setup the process slice dataspaces as well */

    ret = H5Sselect_hyperslab(tv_ptr->mem_large_ds_process_slice_sid, H5S_SELECT_SET, tv_ptr->start,
                              tv_ptr->stride, tv_ptr->count, tv_ptr->block);
    VRFY((ret >= 0), "H5Sselect_hyperslab(mem_large_ds_process_slice_sid, set) succeeded");

    ret = H5Sselect_hyperslab(tv_ptr->file_large_ds_process_slice_sid, H5S_SELECT_SET, tv_ptr->start,
                              tv_ptr->stride, tv_ptr->count, tv_ptr->block);
    VRFY((ret >= 0), "H5Sselect_hyperslab(file_large_ds_process_slice_sid, set) succeeded");

    if (MAINPROCESS) { /* add an additional slice to the selections */

        tv_ptr->start[0] = (hsize_t)(tv_ptr->mpi_size);

        ret = H5Sselect_hyperslab(tv_ptr->mem_large_ds_sid, H5S_SELECT_OR, tv_ptr->start, tv_ptr->stride,
                                  tv_ptr->count, tv_ptr->block);
        VRFY((ret >= 0), "H5Sselect_hyperslab(mem_large_ds_sid, or) succeeded");

        ret = H5Sselect_hyperslab(tv_ptr->file_large_ds_sid_0, H5S_SELECT_OR, tv_ptr->start, tv_ptr->stride,
                                  tv_ptr->count, tv_ptr->block);
        VRFY((ret >= 0), "H5Sselect_hyperslab(file_large_ds_sid_0, or) succeeded");
    }

    /* write the initial value of the large data set to file */
    ret = H5Dwrite(tv_ptr->large_dataset, tv_ptr->dset_type, tv_ptr->mem_large_ds_sid,
                   tv_ptr->file_large_ds_sid_0, tv_ptr->xfer_plist, tv_ptr->large_ds_buf_0);
    if (ret < 0)
        H5Eprint2(H5E_DEFAULT, stderr);
    VRFY((ret >= 0), "H5Dwrite() large_dataset initial write succeeded");

    /* sync with the other processes before checking data */
    mrc = MPI_Barrier(MPI_COMM_WORLD);
    VRFY((mrc == MPI_SUCCESS), "Sync after large dataset writes");

    /* read the large data set back to verify that it contains the
     * expected data.  Note that each process reads in the entire
     * data set.
     */
    ret = H5Dread(tv_ptr->large_dataset, H5T_NATIVE_UINT32, tv_ptr->full_mem_large_ds_sid,
                  tv_ptr->full_file_large_ds_sid, tv_ptr->xfer_plist, tv_ptr->large_ds_buf_1);
    VRFY((ret >= 0), "H5Dread() large_dataset initial read succeeded");

    /* verify that the correct data was written to the large data set */
    expected_value = 0;
    mis_match      = false;
    ptr_1          = tv_ptr->large_ds_buf_1;

    i = 0;
    for (i = 0; i < (int)(tv_ptr->large_ds_size); i++) {

        if (*ptr_1 != expected_value) {

            mis_match = true;
        }
        ptr_1++;
        expected_value++;
    }
    VRFY((mis_match == false), "large ds init data good.");

    /* sync with the other processes before changing data */
    mrc = MPI_Barrier(MPI_COMM_WORLD);
    VRFY((mrc == MPI_SUCCESS), "Sync initial values check");

    return;

} /* hs_dr_pio_test__setup() */

/*-------------------------------------------------------------------------
 * Function:    hs_dr_pio_test__takedown()
 *
 * Purpose:    Do takedown after tests of I/O to/from hyperslab selections
 *        of different rank in the parallel case.
 *
 * Return:    void
 *
 *-------------------------------------------------------------------------
 */

#define HS_DR_PIO_TEST__TAKEDOWN__DEBUG 0

static void
hs_dr_pio_test__takedown(struct hs_dr_pio_test_vars_t *tv_ptr)
{
#if HS_DR_PIO_TEST__TAKEDOWN__DEBUG
    const char *fcnName = "hs_dr_pio_test__takedown()";
#endif               /* HS_DR_PIO_TEST__TAKEDOWN__DEBUG */
    int    mpi_rank; /* needed by the VRFY macro */
    herr_t ret;      /* Generic return value */

    /* initialize the local copy of mpi_rank */
    mpi_rank = tv_ptr->mpi_rank;

    /* Close property lists */
    if (tv_ptr->xfer_plist != H5P_DEFAULT) {
        ret = H5Pclose(tv_ptr->xfer_plist);
        VRFY((ret != FAIL), "H5Pclose(xfer_plist) succeeded");
    }

    /* Close dataspaces */
    ret = H5Sclose(tv_ptr->full_mem_small_ds_sid);
    VRFY((ret != FAIL), "H5Sclose(full_mem_small_ds_sid) succeeded");

    ret = H5Sclose(tv_ptr->full_file_small_ds_sid);
    VRFY((ret != FAIL), "H5Sclose(full_file_small_ds_sid) succeeded");

    ret = H5Sclose(tv_ptr->mem_small_ds_sid);
    VRFY((ret != FAIL), "H5Sclose(mem_small_ds_sid) succeeded");

    ret = H5Sclose(tv_ptr->file_small_ds_sid_0);
    VRFY((ret != FAIL), "H5Sclose(file_small_ds_sid_0) succeeded");

    ret = H5Sclose(tv_ptr->file_small_ds_sid_1);
    VRFY((ret != FAIL), "H5Sclose(file_small_ds_sid_1) succeeded");

    ret = H5Sclose(tv_ptr->small_ds_slice_sid);
    VRFY((ret != FAIL), "H5Sclose(small_ds_slice_sid) succeeded");

    ret = H5Sclose(tv_ptr->full_mem_large_ds_sid);
    VRFY((ret != FAIL), "H5Sclose(full_mem_large_ds_sid) succeeded");

    ret = H5Sclose(tv_ptr->full_file_large_ds_sid);
    VRFY((ret != FAIL), "H5Sclose(full_file_large_ds_sid) succeeded");

    ret = H5Sclose(tv_ptr->mem_large_ds_sid);
    VRFY((ret != FAIL), "H5Sclose(mem_large_ds_sid) succeeded");

    ret = H5Sclose(tv_ptr->file_large_ds_sid_0);
    VRFY((ret != FAIL), "H5Sclose(file_large_ds_sid_0) succeeded");

    ret = H5Sclose(tv_ptr->file_large_ds_sid_1);
    VRFY((ret != FAIL), "H5Sclose(file_large_ds_sid_1) succeeded");

    ret = H5Sclose(tv_ptr->mem_large_ds_process_slice_sid);
    VRFY((ret != FAIL), "H5Sclose(mem_large_ds_process_slice_sid) succeeded");

    ret = H5Sclose(tv_ptr->file_large_ds_process_slice_sid);
    VRFY((ret != FAIL), "H5Sclose(file_large_ds_process_slice_sid) succeeded");

    ret = H5Sclose(tv_ptr->large_ds_slice_sid);
    VRFY((ret != FAIL), "H5Sclose(large_ds_slice_sid) succeeded");

    /* Close Datasets */
    ret = H5Dclose(tv_ptr->small_dataset);
    VRFY((ret != FAIL), "H5Dclose(small_dataset) succeeded");

    ret = H5Dclose(tv_ptr->large_dataset);
    VRFY((ret != FAIL), "H5Dclose(large_dataset) succeeded");

    /* close the file collectively */
    MESG("about to close file.");
    ret = H5Fclose(tv_ptr->fid);
    VRFY((ret != FAIL), "file close succeeded");

    /* Free memory buffers */

    if (tv_ptr->small_ds_buf_0 != NULL)
        free(tv_ptr->small_ds_buf_0);
    if (tv_ptr->small_ds_buf_1 != NULL)
        free(tv_ptr->small_ds_buf_1);
    if (tv_ptr->small_ds_buf_2 != NULL)
        free(tv_ptr->small_ds_buf_2);
    if (tv_ptr->small_ds_slice_buf != NULL)
        free(tv_ptr->small_ds_slice_buf);

    if (tv_ptr->large_ds_buf_0 != NULL)
        free(tv_ptr->large_ds_buf_0);
    if (tv_ptr->large_ds_buf_1 != NULL)
        free(tv_ptr->large_ds_buf_1);
    if (tv_ptr->large_ds_buf_2 != NULL)
        free(tv_ptr->large_ds_buf_2);
    if (tv_ptr->large_ds_slice_buf != NULL)
        free(tv_ptr->large_ds_slice_buf);

    return;

} /* hs_dr_pio_test__takedown() */

/*-------------------------------------------------------------------------
 * Function:    contig_hs_dr_pio_test__d2m_l2s()
 *
 * Purpose:    Part one of a series of tests of I/O to/from hyperslab
 *        selections of different rank in the parallel.
 *
 *        Verify that we can read from disk correctly using
 *        selections of different rank that H5Sselect_shape_same()
 *        views as being of the same shape.
 *
 *              In this function, we test this by reading small_rank - 1
 *        slices from the on disk large cube, and verifying that the
 *        data read is correct.  Verify that H5Sselect_shape_same()
 *        returns true on the memory and file selections.
 *
 * Return:    void
 *
 *-------------------------------------------------------------------------
 */

#define CONTIG_HS_DR_PIO_TEST__D2M_L2S__DEBUG 0

static void
contig_hs_dr_pio_test__d2m_l2s(struct hs_dr_pio_test_vars_t *tv_ptr)
{
#if CONTIG_HS_DR_PIO_TEST__D2M_L2S__DEBUG
    const char *fcnName = "contig_hs_dr_pio_test__run_test()";
#endif /* CONTIG_HS_DR_PIO_TEST__D2M_L2S__DEBUG */
    bool      mis_match = false;
    int       i, j, k, l;
    size_t    n;
    int       mpi_rank; /* needed by the VRFY macro */
    uint32_t  expected_value;
    uint32_t *ptr_1;
    htri_t    check; /* Shape comparison return value */
    herr_t    ret;   /* Generic return value */

    /* initialize the local copy of mpi_rank */
    mpi_rank = tv_ptr->mpi_rank;

    /* We have already done a H5Sselect_all() on the dataspace
     * small_ds_slice_sid in the initialization phase, so no need to
     * call H5Sselect_all() again.
     */

    /* set up start, stride, count, and block -- note that we will
     * change start[] so as to read slices of the large cube.
     */
    for (i = 0; i < PAR_SS_DR_MAX_RANK; i++) {

        tv_ptr->start[i]  = 0;
        tv_ptr->stride[i] = (hsize_t)(2 * tv_ptr->edge_size);
        tv_ptr->count[i]  = 1;
        if ((PAR_SS_DR_MAX_RANK - i) > (tv_ptr->small_rank - 1)) {

            tv_ptr->block[i] = 1;
        }
        else {

            tv_ptr->block[i] = (hsize_t)(tv_ptr->edge_size);
        }
    }

    /* zero out the buffer we will be reading into */
    memset(tv_ptr->small_ds_slice_buf, 0, sizeof(uint32_t) * tv_ptr->small_ds_slice_size);

#if CONTIG_HS_DR_PIO_TEST__D2M_L2S__DEBUG
    fprintf(stdout, "%s reading slices from big cube on disk into small cube slice.\n", fcnName);
#endif /* CONTIG_HS_DR_PIO_TEST__D2M_L2S__DEBUG */

    /* in serial versions of this test, we loop through all the dimensions
     * of the large data set.  However, in the parallel version, each
     * process only works with that slice of the large cube indicated
     * by its rank -- hence we set the most slowly changing index to
     * mpi_rank, and don't iterate over it.
     */

    if (PAR_SS_DR_MAX_RANK - tv_ptr->large_rank == 0) {

        i = tv_ptr->mpi_rank;
    }
    else {

        i = 0;
    }

    /* since large_rank is at most PAR_SS_DR_MAX_RANK, no need to
     * loop over it -- either we are setting i to mpi_rank, or
     * we are setting it to zero.  It will not change during the
     * test.
     */

    if (PAR_SS_DR_MAX_RANK - tv_ptr->large_rank == 1) {

        j = tv_ptr->mpi_rank;
    }
    else {

        j = 0;
    }

    do {
        if (PAR_SS_DR_MAX_RANK - tv_ptr->large_rank == 2) {

            k = tv_ptr->mpi_rank;
        }
        else {

            k = 0;
        }

        do {
            /* since small rank >= 2 and large_rank > small_rank, we
             * have large_rank >= 3.  Since PAR_SS_DR_MAX_RANK == 5
             * (baring major re-orgaization), this gives us:
             *
             *     (PAR_SS_DR_MAX_RANK - large_rank) <= 2
             *
             * so no need to repeat the test in the outer loops --
             * just set l = 0.
             */

            l = 0;
            do {
                if ((tv_ptr->skips)++ < tv_ptr->max_skips) { /* skip the test */

                    (tv_ptr->tests_skipped)++;
                }
                else { /* run the test */

                    tv_ptr->skips = 0; /* reset the skips counter */

                    /* we know that small_rank - 1 >= 1 and that
                     * large_rank > small_rank by the assertions at the head
                     * of this function.  Thus no need for another inner loop.
                     */
                    tv_ptr->start[0] = (hsize_t)i;
                    tv_ptr->start[1] = (hsize_t)j;
                    tv_ptr->start[2] = (hsize_t)k;
                    tv_ptr->start[3] = (hsize_t)l;
                    tv_ptr->start[4] = 0;

                    ret = H5Sselect_hyperslab(tv_ptr->file_large_ds_sid_0, H5S_SELECT_SET, tv_ptr->start_ptr,
                                              tv_ptr->stride_ptr, tv_ptr->count_ptr, tv_ptr->block_ptr);
                    VRFY((ret != FAIL), "H5Sselect_hyperslab(file_large_cube_sid) succeeded");

                    /* verify that H5Sselect_shape_same() reports the two
                     * selections as having the same shape.
                     */
                    check = H5Sselect_shape_same(tv_ptr->small_ds_slice_sid, tv_ptr->file_large_ds_sid_0);
                    VRFY((check == true), "H5Sselect_shape_same passed");

                    /* Read selection from disk */
#if CONTIG_HS_DR_PIO_TEST__D2M_L2S__DEBUG
                    fprintf(stdout, "%s:%d: start = %d %d %d %d %d.\n", fcnName, (int)(tv_ptr->mpi_rank),
                            (int)(tv_ptr->start[0]), (int)(tv_ptr->start[1]), (int)(tv_ptr->start[2]),
                            (int)(tv_ptr->start[3]), (int)(tv_ptr->start[4]));
                    fprintf(stdout, "%s slice/file extent dims = %d/%d.\n", fcnName,
                            H5Sget_simple_extent_ndims(tv_ptr->small_ds_slice_sid),
                            H5Sget_simple_extent_ndims(tv_ptr->file_large_ds_sid_0));
#endif /* CONTIG_HS_DR_PIO_TEST__D2M_L2S__DEBUG */
                    ret =
                        H5Dread(tv_ptr->large_dataset, H5T_NATIVE_UINT32, tv_ptr->small_ds_slice_sid,
                                tv_ptr->file_large_ds_sid_0, tv_ptr->xfer_plist, tv_ptr->small_ds_slice_buf);
                    VRFY((ret >= 0), "H5Dread() slice from large ds succeeded.");

                    /* verify that expected data is retrieved */

                    mis_match = false;
                    ptr_1     = tv_ptr->small_ds_slice_buf;
                    expected_value =
                        (uint32_t)((i * tv_ptr->edge_size * tv_ptr->edge_size * tv_ptr->edge_size *
                                    tv_ptr->edge_size) +
                                   (j * tv_ptr->edge_size * tv_ptr->edge_size * tv_ptr->edge_size) +
                                   (k * tv_ptr->edge_size * tv_ptr->edge_size) + (l * tv_ptr->edge_size));

                    for (n = 0; n < tv_ptr->small_ds_slice_size; n++) {

                        if (*ptr_1 != expected_value) {

                            mis_match = true;
                        }

                        *ptr_1 = 0; /* zero data for next use */

                        ptr_1++;
                        expected_value++;
                    }

                    VRFY((mis_match == false), "small slice read from large ds data good.");

                    (tv_ptr->tests_run)++;
                }

                l++;

                (tv_ptr->total_tests)++;

            } while ((tv_ptr->large_rank > 2) && ((tv_ptr->small_rank - 1) <= 1) && (l < tv_ptr->edge_size));
            k++;
        } while ((tv_ptr->large_rank > 3) && ((tv_ptr->small_rank - 1) <= 2) && (k < tv_ptr->edge_size));
        j++;
    } while ((tv_ptr->large_rank > 4) && ((tv_ptr->small_rank - 1) <= 3) && (j < tv_ptr->edge_size));

    return;

} /* contig_hs_dr_pio_test__d2m_l2s() */

/*-------------------------------------------------------------------------
 * Function:    contig_hs_dr_pio_test__d2m_s2l()
 *
 * Purpose:    Part two of a series of tests of I/O to/from hyperslab
 *        selections of different rank in the parallel.
 *
 *        Verify that we can read from disk correctly using
 *        selections of different rank that H5Sselect_shape_same()
 *        views as being of the same shape.
 *
 *        In this function, we test this by reading slices of the
 *        on disk small data set into slices through the in memory
 *        large data set, and verify that the correct data (and
 *        only the correct data) is read.
 *
 * Return:    void
 *
 *-------------------------------------------------------------------------
 */

#define CONTIG_HS_DR_PIO_TEST__D2M_S2L__DEBUG 0

static void
contig_hs_dr_pio_test__d2m_s2l(struct hs_dr_pio_test_vars_t *tv_ptr)
{
#if CONTIG_HS_DR_PIO_TEST__D2M_S2L__DEBUG
    const char *fcnName = "contig_hs_dr_pio_test__d2m_s2l()";
#endif /* CONTIG_HS_DR_PIO_TEST__D2M_S2L__DEBUG */
    bool      mis_match = false;
    int       i, j, k, l;
    size_t    n;
    int       mpi_rank; /* needed by the VRFY macro */
    size_t    start_index;
    size_t    stop_index;
    uint32_t  expected_value;
    uint32_t *ptr_1;
    htri_t    check; /* Shape comparison return value */
    herr_t    ret;   /* Generic return value */

    /* initialize the local copy of mpi_rank */
    mpi_rank = tv_ptr->mpi_rank;

    /* Read slices of the on disk small data set into slices
     * through the in memory large data set, and verify that the correct
     * data (and only the correct data) is read.
     */

    tv_ptr->start[0]  = (hsize_t)(tv_ptr->mpi_rank);
    tv_ptr->stride[0] = (hsize_t)(2 * (tv_ptr->mpi_size + 1));
    tv_ptr->count[0]  = 1;
    tv_ptr->block[0]  = 1;

    for (i = 1; i < tv_ptr->large_rank; i++) {

        tv_ptr->start[i]  = 0;
        tv_ptr->stride[i] = (hsize_t)(2 * tv_ptr->edge_size);
        tv_ptr->count[i]  = 1;
        tv_ptr->block[i]  = (hsize_t)(tv_ptr->edge_size);
    }

    ret = H5Sselect_hyperslab(tv_ptr->file_small_ds_sid_0, H5S_SELECT_SET, tv_ptr->start, tv_ptr->stride,
                              tv_ptr->count, tv_ptr->block);
    VRFY((ret >= 0), "H5Sselect_hyperslab(file_small_ds_sid_0, set) succeeded");

#if CONTIG_HS_DR_PIO_TEST__D2M_S2L__DEBUG
    fprintf(stdout, "%s reading slices of on disk small data set into slices of big data set.\n", fcnName);
#endif /* CONTIG_HS_DR_PIO_TEST__D2M_S2L__DEBUG */

    /* zero out the in memory large ds */
    memset(tv_ptr->large_ds_buf_1, 0, sizeof(uint32_t) * tv_ptr->large_ds_size);

    /* set up start, stride, count, and block -- note that we will
     * change start[] so as to read slices of the large cube.
     */
    for (i = 0; i < PAR_SS_DR_MAX_RANK; i++) {

        tv_ptr->start[i]  = 0;
        tv_ptr->stride[i] = (hsize_t)(2 * tv_ptr->edge_size);
        tv_ptr->count[i]  = 1;
        if ((PAR_SS_DR_MAX_RANK - i) > (tv_ptr->small_rank - 1)) {

            tv_ptr->block[i] = 1;
        }
        else {

            tv_ptr->block[i] = (hsize_t)(tv_ptr->edge_size);
        }
    }

    /* in serial versions of this test, we loop through all the dimensions
     * of the large data set that don't appear in the small data set.
     *
     * However, in the parallel version, each process only works with that
     * slice of the large (and small) data set indicated by its rank -- hence
     * we set the most slowly changing index to mpi_rank, and don't iterate
     * over it.
     */

    if (PAR_SS_DR_MAX_RANK - tv_ptr->large_rank == 0) {

        i = tv_ptr->mpi_rank;
    }
    else {

        i = 0;
    }

    /* since large_rank is at most PAR_SS_DR_MAX_RANK, no need to
     * loop over it -- either we are setting i to mpi_rank, or
     * we are setting it to zero.  It will not change during the
     * test.
     */

    if (PAR_SS_DR_MAX_RANK - tv_ptr->large_rank == 1) {

        j = tv_ptr->mpi_rank;
    }
    else {

        j = 0;
    }

    do {
        if (PAR_SS_DR_MAX_RANK - tv_ptr->large_rank == 2) {

            k = tv_ptr->mpi_rank;
        }
        else {

            k = 0;
        }

        do {
            /* since small rank >= 2 and large_rank > small_rank, we
             * have large_rank >= 3.  Since PAR_SS_DR_MAX_RANK == 5
             * (baring major re-orgaization), this gives us:
             *
             *     (PAR_SS_DR_MAX_RANK - large_rank) <= 2
             *
             * so no need to repeat the test in the outer loops --
             * just set l = 0.
             */

            l = 0;
            do {
                if ((tv_ptr->skips)++ < tv_ptr->max_skips) { /* skip the test */

                    (tv_ptr->tests_skipped)++;
                }
                else { /* run the test */

                    tv_ptr->skips = 0; /* reset the skips counter */

                    /* we know that small_rank >= 1 and that large_rank > small_rank
                     * by the assertions at the head of this function.  Thus no
                     * need for another inner loop.
                     */
                    tv_ptr->start[0] = (hsize_t)i;
                    tv_ptr->start[1] = (hsize_t)j;
                    tv_ptr->start[2] = (hsize_t)k;
                    tv_ptr->start[3] = (hsize_t)l;
                    tv_ptr->start[4] = 0;

                    ret = H5Sselect_hyperslab(tv_ptr->mem_large_ds_sid, H5S_SELECT_SET, tv_ptr->start_ptr,
                                              tv_ptr->stride_ptr, tv_ptr->count_ptr, tv_ptr->block_ptr);
                    VRFY((ret != FAIL), "H5Sselect_hyperslab(mem_large_ds_sid) succeeded");

                    /* verify that H5Sselect_shape_same() reports the two
                     * selections as having the same shape.
                     */
                    check = H5Sselect_shape_same(tv_ptr->file_small_ds_sid_0, tv_ptr->mem_large_ds_sid);
                    VRFY((check == true), "H5Sselect_shape_same passed");

                    /* Read selection from disk */
#if CONTIG_HS_DR_PIO_TEST__D2M_S2L__DEBUG
                    fprintf(stdout, "%s:%d: start = %d %d %d %d %d.\n", fcnName, (int)(tv_ptr->mpi_rank),
                            (int)(tv_ptr->start[0]), (int)(tv_ptr->start[1]), (int)(tv_ptr->start[2]),
                            (int)(tv_ptr->start[3]), (int)(tv_ptr->start[4]));
                    fprintf(stdout, "%s:%d: mem/file extent dims = %d/%d.\n", fcnName, tv_ptr->mpi_rank,
                            H5Sget_simple_extent_ndims(tv_ptr->mem_large_ds_sid),
                            H5Sget_simple_extent_ndims(tv_ptr->file_small_ds_sid_0));
#endif /* CONTIG_HS_DR_PIO_TEST__D2M_S2L__DEBUG */
                    ret = H5Dread(tv_ptr->small_dataset, H5T_NATIVE_UINT32, tv_ptr->mem_large_ds_sid,
                                  tv_ptr->file_small_ds_sid_0, tv_ptr->xfer_plist, tv_ptr->large_ds_buf_1);
                    VRFY((ret >= 0), "H5Dread() slice from small ds succeeded.");

                    /* verify that the expected data and only the
                     * expected data was read.
                     */
                    ptr_1          = tv_ptr->large_ds_buf_1;
                    expected_value = (uint32_t)((size_t)(tv_ptr->mpi_rank) * tv_ptr->small_ds_slice_size);
                    start_index =
                        (size_t)((i * tv_ptr->edge_size * tv_ptr->edge_size * tv_ptr->edge_size *
                                  tv_ptr->edge_size) +
                                 (j * tv_ptr->edge_size * tv_ptr->edge_size * tv_ptr->edge_size) +
                                 (k * tv_ptr->edge_size * tv_ptr->edge_size) + (l * tv_ptr->edge_size));
                    stop_index = start_index + tv_ptr->small_ds_slice_size - 1;

                    assert(start_index < stop_index);
                    assert(stop_index <= tv_ptr->large_ds_size);

                    for (n = 0; n < tv_ptr->large_ds_size; n++) {

                        if ((n >= start_index) && (n <= stop_index)) {

                            if (*ptr_1 != expected_value) {

                                mis_match = true;
                            }
                            expected_value++;
                        }
                        else {

                            if (*ptr_1 != 0) {

                                mis_match = true;
                            }
                        }
                        /* zero out the value for the next pass */
                        *ptr_1 = 0;

                        ptr_1++;
                    }

                    VRFY((mis_match == false), "small slice read from large ds data good.");

                    (tv_ptr->tests_run)++;
                }

                l++;

                (tv_ptr->total_tests)++;

            } while ((tv_ptr->large_rank > 2) && ((tv_ptr->small_rank - 1) <= 1) && (l < tv_ptr->edge_size));
            k++;
        } while ((tv_ptr->large_rank > 3) && ((tv_ptr->small_rank - 1) <= 2) && (k < tv_ptr->edge_size));
        j++;
    } while ((tv_ptr->large_rank > 4) && ((tv_ptr->small_rank - 1) <= 3) && (j < tv_ptr->edge_size));

    return;

} /* contig_hs_dr_pio_test__d2m_s2l() */

/*-------------------------------------------------------------------------
 * Function:    contig_hs_dr_pio_test__m2d_l2s()
 *
 * Purpose:    Part three of a series of tests of I/O to/from hyperslab
 *        selections of different rank in the parallel.
 *
 *        Verify that we can write from memory to file using
 *        selections of different rank that H5Sselect_shape_same()
 *        views as being of the same shape.
 *
 *        Do this by writing small_rank - 1 dimensional slices from
 *        the in memory large data set to the on disk small cube
 *        dataset.  After each write, read the slice of the small
 *        dataset back from disk, and verify that it contains
 *        the expected data. Verify that H5Sselect_shape_same()
 *        returns true on the memory and file selections.
 *
 * Return:    void
 *
 *-------------------------------------------------------------------------
 */

#define CONTIG_HS_DR_PIO_TEST__M2D_L2S__DEBUG 0

static void
contig_hs_dr_pio_test__m2d_l2s(struct hs_dr_pio_test_vars_t *tv_ptr)
{
#if CONTIG_HS_DR_PIO_TEST__M2D_L2S__DEBUG
    const char *fcnName = "contig_hs_dr_pio_test__m2d_l2s()";
#endif /* CONTIG_HS_DR_PIO_TEST__M2D_L2S__DEBUG */
    bool      mis_match = false;
    int       i, j, k, l;
    size_t    n;
    int       mpi_rank; /* needed by the VRFY macro */
    size_t    start_index;
    size_t    stop_index;
    uint32_t  expected_value;
    uint32_t *ptr_1;
    htri_t    check; /* Shape comparison return value */
    herr_t    ret;   /* Generic return value */

    /* initialize the local copy of mpi_rank */
    mpi_rank = tv_ptr->mpi_rank;

    /* now we go in the opposite direction, verifying that we can write
     * from memory to file using selections of different rank that
     * H5Sselect_shape_same() views as being of the same shape.
     *
     * Start by writing small_rank - 1 dimensional slices from the in memory large
     * data set to the on disk small cube dataset.  After each write, read the
     * slice of the small dataset back from disk, and verify that it contains
     * the expected data. Verify that H5Sselect_shape_same() returns true on
     * the memory and file selections.
     */

    tv_ptr->start[0]  = (hsize_t)(tv_ptr->mpi_rank);
    tv_ptr->stride[0] = (hsize_t)(2 * (tv_ptr->mpi_size + 1));
    tv_ptr->count[0]  = 1;
    tv_ptr->block[0]  = 1;

    for (i = 1; i < tv_ptr->large_rank; i++) {

        tv_ptr->start[i]  = 0;
        tv_ptr->stride[i] = (hsize_t)(2 * tv_ptr->edge_size);
        tv_ptr->count[i]  = 1;
        tv_ptr->block[i]  = (hsize_t)(tv_ptr->edge_size);
    }

    ret = H5Sselect_hyperslab(tv_ptr->file_small_ds_sid_0, H5S_SELECT_SET, tv_ptr->start, tv_ptr->stride,
                              tv_ptr->count, tv_ptr->block);
    VRFY((ret >= 0), "H5Sselect_hyperslab(file_small_ds_sid_0, set) succeeded");

    ret = H5Sselect_hyperslab(tv_ptr->mem_small_ds_sid, H5S_SELECT_SET, tv_ptr->start, tv_ptr->stride,
                              tv_ptr->count, tv_ptr->block);
    VRFY((ret >= 0), "H5Sselect_hyperslab(mem_small_ds_sid, set) succeeded");

    /* set up start, stride, count, and block -- note that we will
     * change start[] so as to read slices of the large cube.
     */
    for (i = 0; i < PAR_SS_DR_MAX_RANK; i++) {

        tv_ptr->start[i]  = 0;
        tv_ptr->stride[i] = (hsize_t)(2 * tv_ptr->edge_size);
        tv_ptr->count[i]  = 1;
        if ((PAR_SS_DR_MAX_RANK - i) > (tv_ptr->small_rank - 1)) {

            tv_ptr->block[i] = 1;
        }
        else {

            tv_ptr->block[i] = (hsize_t)(tv_ptr->edge_size);
        }
    }

    /* zero out the in memory small ds */
    memset(tv_ptr->small_ds_buf_1, 0, sizeof(uint32_t) * tv_ptr->small_ds_size);

#if CONTIG_HS_DR_PIO_TEST__M2D_L2S__DEBUG
    fprintf(stdout, "%s writing slices from big ds to slices of small ds on disk.\n", fcnName);
#endif /* CONTIG_HS_DR_PIO_TEST__M2D_L2S__DEBUG */

    /* in serial versions of this test, we loop through all the dimensions
     * of the large data set that don't appear in the small data set.
     *
     * However, in the parallel version, each process only works with that
     * slice of the large (and small) data set indicated by its rank -- hence
     * we set the most slowly changing index to mpi_rank, and don't iterate
     * over it.
     */

    if (PAR_SS_DR_MAX_RANK - tv_ptr->large_rank == 0) {

        i = tv_ptr->mpi_rank;
    }
    else {

        i = 0;
    }

    /* since large_rank is at most PAR_SS_DR_MAX_RANK, no need to
     * loop over it -- either we are setting i to mpi_rank, or
     * we are setting it to zero.  It will not change during the
     * test.
     */

    if (PAR_SS_DR_MAX_RANK - tv_ptr->large_rank == 1) {

        j = tv_ptr->mpi_rank;
    }
    else {

        j = 0;
    }

    j = 0;
    do {
        if (PAR_SS_DR_MAX_RANK - tv_ptr->large_rank == 2) {

            k = tv_ptr->mpi_rank;
        }
        else {

            k = 0;
        }

        do {
            /* since small rank >= 2 and large_rank > small_rank, we
             * have large_rank >= 3.  Since PAR_SS_DR_MAX_RANK == 5
             * (baring major re-orgaization), this gives us:
             *
             *     (PAR_SS_DR_MAX_RANK - large_rank) <= 2
             *
             * so no need to repeat the test in the outer loops --
             * just set l = 0.
             */

            l = 0;
            do {
                if ((tv_ptr->skips)++ < tv_ptr->max_skips) { /* skip the test */

                    (tv_ptr->tests_skipped)++;
                }
                else { /* run the test */

                    tv_ptr->skips = 0; /* reset the skips counter */

                    /* we know that small_rank >= 1 and that large_rank > small_rank
                     * by the assertions at the head of this function.  Thus no
                     * need for another inner loop.
                     */

                    /* zero out this rank's slice of the on disk small data set */
                    ret = H5Dwrite(tv_ptr->small_dataset, H5T_NATIVE_UINT32, tv_ptr->mem_small_ds_sid,
                                   tv_ptr->file_small_ds_sid_0, tv_ptr->xfer_plist, tv_ptr->small_ds_buf_2);
                    VRFY((ret >= 0), "H5Dwrite() zero slice to small ds succeeded.");

                    /* select the portion of the in memory large cube from which we
                     * are going to write data.
                     */
                    tv_ptr->start[0] = (hsize_t)i;
                    tv_ptr->start[1] = (hsize_t)j;
                    tv_ptr->start[2] = (hsize_t)k;
                    tv_ptr->start[3] = (hsize_t)l;
                    tv_ptr->start[4] = 0;

                    ret = H5Sselect_hyperslab(tv_ptr->mem_large_ds_sid, H5S_SELECT_SET, tv_ptr->start_ptr,
                                              tv_ptr->stride_ptr, tv_ptr->count_ptr, tv_ptr->block_ptr);
                    VRFY((ret >= 0), "H5Sselect_hyperslab() mem_large_ds_sid succeeded.");

                    /* verify that H5Sselect_shape_same() reports the in
                     * memory slice through the cube selection and the
                     * on disk full square selections as having the same shape.
                     */
                    check = H5Sselect_shape_same(tv_ptr->file_small_ds_sid_0, tv_ptr->mem_large_ds_sid);
                    VRFY((check == true), "H5Sselect_shape_same passed.");

                    /* write the slice from the in memory large data set to the
                     * slice of the on disk small dataset. */
#if CONTIG_HS_DR_PIO_TEST__M2D_L2S__DEBUG
                    fprintf(stdout, "%s:%d: start = %d %d %d %d %d.\n", fcnName, (int)(tv_ptr->mpi_rank),
                            (int)(tv_ptr->start[0]), (int)(tv_ptr->start[1]), (int)(tv_ptr->start[2]),
                            (int)(tv_ptr->start[3]), (int)(tv_ptr->start[4]));
                    fprintf(stdout, "%s:%d: mem/file extent dims = %d/%d.\n", fcnName, tv_ptr->mpi_rank,
                            H5Sget_simple_extent_ndims(tv_ptr->mem_large_ds_sid),
                            H5Sget_simple_extent_ndims(tv_ptr->file_small_ds_sid_0));
#endif /* CONTIG_HS_DR_PIO_TEST__M2D_L2S__DEBUG */
                    ret = H5Dwrite(tv_ptr->small_dataset, H5T_NATIVE_UINT32, tv_ptr->mem_large_ds_sid,
                                   tv_ptr->file_small_ds_sid_0, tv_ptr->xfer_plist, tv_ptr->large_ds_buf_0);
                    VRFY((ret >= 0), "H5Dwrite() slice to large ds succeeded.");

                    /* read the on disk square into memory */
                    ret = H5Dread(tv_ptr->small_dataset, H5T_NATIVE_UINT32, tv_ptr->mem_small_ds_sid,
                                  tv_ptr->file_small_ds_sid_0, tv_ptr->xfer_plist, tv_ptr->small_ds_buf_1);
                    VRFY((ret >= 0), "H5Dread() slice from small ds succeeded.");

                    /* verify that expected data is retrieved */

                    mis_match = false;
                    ptr_1     = tv_ptr->small_ds_buf_1;

                    expected_value =
                        (uint32_t)((i * tv_ptr->edge_size * tv_ptr->edge_size * tv_ptr->edge_size *
                                    tv_ptr->edge_size) +
                                   (j * tv_ptr->edge_size * tv_ptr->edge_size * tv_ptr->edge_size) +
                                   (k * tv_ptr->edge_size * tv_ptr->edge_size) + (l * tv_ptr->edge_size));

                    start_index = (size_t)(tv_ptr->mpi_rank) * tv_ptr->small_ds_slice_size;
                    stop_index  = start_index + tv_ptr->small_ds_slice_size - 1;

                    assert(start_index < stop_index);
                    assert(stop_index <= tv_ptr->small_ds_size);

                    for (n = 0; n < tv_ptr->small_ds_size; n++) {

                        if ((n >= start_index) && (n <= stop_index)) {

                            if (*ptr_1 != expected_value) {

                                mis_match = true;
                            }
                            expected_value++;
                        }
                        else {

                            if (*ptr_1 != 0) {

                                mis_match = true;
                            }
                        }
                        /* zero out the value for the next pass */
                        *ptr_1 = 0;

                        ptr_1++;
                    }

                    VRFY((mis_match == false), "small slice write from large ds data good.");

                    (tv_ptr->tests_run)++;
                }

                l++;

                (tv_ptr->total_tests)++;

            } while ((tv_ptr->large_rank > 2) && ((tv_ptr->small_rank - 1) <= 1) && (l < tv_ptr->edge_size));
            k++;
        } while ((tv_ptr->large_rank > 3) && ((tv_ptr->small_rank - 1) <= 2) && (k < tv_ptr->edge_size));
        j++;
    } while ((tv_ptr->large_rank > 4) && ((tv_ptr->small_rank - 1) <= 3) && (j < tv_ptr->edge_size));

    return;

} /* contig_hs_dr_pio_test__m2d_l2s() */

/*-------------------------------------------------------------------------
 * Function:    contig_hs_dr_pio_test__m2d_s2l()
 *
 * Purpose:    Part four of a series of tests of I/O to/from hyperslab
 *        selections of different rank in the parallel.
 *
 *        Verify that we can write from memory to file using
 *        selections of different rank that H5Sselect_shape_same()
 *        views as being of the same shape.
 *
 *        Do this by writing the contents of the process's slice of
 *        the in memory small data set to slices of the on disk
 *        large data set.  After each write, read the process's
 *        slice of the large data set back into memory, and verify
 *        that it contains the expected data.
 *
 *        Verify that H5Sselect_shape_same() returns true on the
 *        memory and file selections.
 *
 * Return:    void
 *
 *-------------------------------------------------------------------------
 */

#define CONTIG_HS_DR_PIO_TEST__M2D_S2L__DEBUG 0

static void
contig_hs_dr_pio_test__m2d_s2l(struct hs_dr_pio_test_vars_t *tv_ptr)
{
#if CONTIG_HS_DR_PIO_TEST__M2D_S2L__DEBUG
    const char *fcnName = "contig_hs_dr_pio_test__m2d_s2l()";
#endif /* CONTIG_HS_DR_PIO_TEST__M2D_S2L__DEBUG */
    bool      mis_match = false;
    int       i, j, k, l;
    size_t    n;
    int       mpi_rank; /* needed by the VRFY macro */
    size_t    start_index;
    size_t    stop_index;
    uint32_t  expected_value;
    uint32_t *ptr_1;
    htri_t    check; /* Shape comparison return value */
    herr_t    ret;   /* Generic return value */

    /* initialize the local copy of mpi_rank */
    mpi_rank = tv_ptr->mpi_rank;

    /* Now write the contents of the process's slice of the in memory
     * small data set to slices of the on disk large data set.  After
     * each write, read the process's slice of the large data set back
     * into memory, and verify that it contains the expected data.
     * Verify that H5Sselect_shape_same() returns true on the memory
     * and file selections.
     */

    /* select the slice of the in memory small data set associated with
     * the process's mpi rank.
     */
    tv_ptr->start[0]  = (hsize_t)(tv_ptr->mpi_rank);
    tv_ptr->stride[0] = (hsize_t)(2 * (tv_ptr->mpi_size + 1));
    tv_ptr->count[0]  = 1;
    tv_ptr->block[0]  = 1;

    for (i = 1; i < tv_ptr->large_rank; i++) {

        tv_ptr->start[i]  = 0;
        tv_ptr->stride[i] = (hsize_t)(2 * tv_ptr->edge_size);
        tv_ptr->count[i]  = 1;
        tv_ptr->block[i]  = (hsize_t)(tv_ptr->edge_size);
    }

    ret = H5Sselect_hyperslab(tv_ptr->mem_small_ds_sid, H5S_SELECT_SET, tv_ptr->start, tv_ptr->stride,
                              tv_ptr->count, tv_ptr->block);
    VRFY((ret >= 0), "H5Sselect_hyperslab(mem_small_ds_sid, set) succeeded");

    /* set up start, stride, count, and block -- note that we will
     * change start[] so as to write slices of the small data set to
     * slices of the large data set.
     */
    for (i = 0; i < PAR_SS_DR_MAX_RANK; i++) {

        tv_ptr->start[i]  = 0;
        tv_ptr->stride[i] = (hsize_t)(2 * tv_ptr->edge_size);
        tv_ptr->count[i]  = 1;
        if ((PAR_SS_DR_MAX_RANK - i) > (tv_ptr->small_rank - 1)) {

            tv_ptr->block[i] = 1;
        }
        else {

            tv_ptr->block[i] = (hsize_t)(tv_ptr->edge_size);
        }
    }

    /* zero out the in memory large ds */
    memset(tv_ptr->large_ds_buf_1, 0, sizeof(uint32_t) * tv_ptr->large_ds_size);

#if CONTIG_HS_DR_PIO_TEST__M2D_S2L__DEBUG
    fprintf(stdout, "%s writing process slices of small ds to slices of large ds on disk.\n", fcnName);
#endif /* CONTIG_HS_DR_PIO_TEST__M2D_S2L__DEBUG */

    if (PAR_SS_DR_MAX_RANK - tv_ptr->large_rank == 0) {

        i = tv_ptr->mpi_rank;
    }
    else {

        i = 0;
    }

    /* since large_rank is at most PAR_SS_DR_MAX_RANK, no need to
     * loop over it -- either we are setting i to mpi_rank, or
     * we are setting it to zero.  It will not change during the
     * test.
     */

    if (PAR_SS_DR_MAX_RANK - tv_ptr->large_rank == 1) {

        j = tv_ptr->mpi_rank;
    }
    else {

        j = 0;
    }

    do {
        if (PAR_SS_DR_MAX_RANK - tv_ptr->large_rank == 2) {

            k = tv_ptr->mpi_rank;
        }
        else {

            k = 0;
        }

        do {
            /* since small rank >= 2 and large_rank > small_rank, we
             * have large_rank >= 3.  Since PAR_SS_DR_MAX_RANK == 5
             * (baring major re-orgaization), this gives us:
             *
             *     (PAR_SS_DR_MAX_RANK - large_rank) <= 2
             *
             * so no need to repeat the test in the outer loops --
             * just set l = 0.
             */

            l = 0;
            do {
                if ((tv_ptr->skips)++ < tv_ptr->max_skips) { /* skip the test */

                    (tv_ptr->tests_skipped)++;

#if CONTIG_HS_DR_PIO_TEST__M2D_S2L__DEBUG
                    tv_ptr->start[0] = (hsize_t)i;
                    tv_ptr->start[1] = (hsize_t)j;
                    tv_ptr->start[2] = (hsize_t)k;
                    tv_ptr->start[3] = (hsize_t)l;
                    tv_ptr->start[4] = 0;

                    fprintf(stdout, "%s:%d: skipping test with start = %d %d %d %d %d.\n", fcnName,
                            (int)(tv_ptr->mpi_rank), (int)(tv_ptr->start[0]), (int)(tv_ptr->start[1]),
                            (int)(tv_ptr->start[2]), (int)(tv_ptr->start[3]), (int)(tv_ptr->start[4]));
                    fprintf(stdout, "%s:%d: mem/file extent dims = %d/%d.\n", fcnName, tv_ptr->mpi_rank,
                            H5Sget_simple_extent_ndims(tv_ptr->mem_small_ds_sid),
                            H5Sget_simple_extent_ndims(tv_ptr->file_large_ds_sid_0));
#endif /* CONTIG_HS_DR_PIO_TEST__M2D_S2L__DEBUG */
                }
                else { /* run the test */

                    tv_ptr->skips = 0; /* reset the skips counter */

                    /* we know that small_rank >= 1 and that large_rank > small_rank
                     * by the assertions at the head of this function.  Thus no
                     * need for another inner loop.
                     */

                    /* Zero out this processes slice of the on disk large data set.
                     * Note that this will leave one slice with its original data
                     * as there is one more slice than processes.
                     */
                    ret = H5Dwrite(tv_ptr->large_dataset, H5T_NATIVE_UINT32, tv_ptr->large_ds_slice_sid,
                                   tv_ptr->file_large_ds_process_slice_sid, tv_ptr->xfer_plist,
                                   tv_ptr->large_ds_buf_2);
                    VRFY((ret != FAIL), "H5Dwrite() to zero large ds succeeded");

                    /* select the portion of the in memory large cube to which we
                     * are going to write data.
                     */
                    tv_ptr->start[0] = (hsize_t)i;
                    tv_ptr->start[1] = (hsize_t)j;
                    tv_ptr->start[2] = (hsize_t)k;
                    tv_ptr->start[3] = (hsize_t)l;
                    tv_ptr->start[4] = 0;

                    ret = H5Sselect_hyperslab(tv_ptr->file_large_ds_sid_0, H5S_SELECT_SET, tv_ptr->start_ptr,
                                              tv_ptr->stride_ptr, tv_ptr->count_ptr, tv_ptr->block_ptr);
                    VRFY((ret != FAIL), "H5Sselect_hyperslab() target large ds slice succeeded");

                    /* verify that H5Sselect_shape_same() reports the in
                     * memory small data set slice selection and the
                     * on disk slice through the large data set selection
                     * as having the same shape.
                     */
                    check = H5Sselect_shape_same(tv_ptr->mem_small_ds_sid, tv_ptr->file_large_ds_sid_0);
                    VRFY((check == true), "H5Sselect_shape_same passed");

                    /* write the small data set slice from memory to the
                     * target slice of the disk data set
                     */
#if CONTIG_HS_DR_PIO_TEST__M2D_S2L__DEBUG
                    fprintf(stdout, "%s:%d: start = %d %d %d %d %d.\n", fcnName, (int)(tv_ptr->mpi_rank),
                            (int)(tv_ptr->start[0]), (int)(tv_ptr->start[1]), (int)(tv_ptr->start[2]),
                            (int)(tv_ptr->start[3]), (int)(tv_ptr->start[4]));
                    fprintf(stdout, "%s:%d: mem/file extent dims = %d/%d.\n", fcnName, tv_ptr->mpi_rank,
                            H5Sget_simple_extent_ndims(tv_ptr->mem_small_ds_sid),
                            H5Sget_simple_extent_ndims(tv_ptr->file_large_ds_sid_0));
#endif /* CONTIG_HS_DR_PIO_TEST__M2D_S2L__DEBUG */
                    ret = H5Dwrite(tv_ptr->large_dataset, H5T_NATIVE_UINT32, tv_ptr->mem_small_ds_sid,
                                   tv_ptr->file_large_ds_sid_0, tv_ptr->xfer_plist, tv_ptr->small_ds_buf_0);
                    VRFY((ret != FAIL), "H5Dwrite of small ds slice to large ds succeeded");

                    /* read this processes slice on the on disk large
                     * data set into memory.
                     */

                    ret = H5Dread(
                        tv_ptr->large_dataset, H5T_NATIVE_UINT32, tv_ptr->mem_large_ds_process_slice_sid,
                        tv_ptr->file_large_ds_process_slice_sid, tv_ptr->xfer_plist, tv_ptr->large_ds_buf_1);
                    VRFY((ret != FAIL), "H5Dread() of process slice of large ds succeeded");

                    /* verify that the expected data and only the
                     * expected data was read.
                     */
                    ptr_1          = tv_ptr->large_ds_buf_1;
                    expected_value = (uint32_t)((size_t)(tv_ptr->mpi_rank) * tv_ptr->small_ds_slice_size);

                    start_index =
                        (size_t)((i * tv_ptr->edge_size * tv_ptr->edge_size * tv_ptr->edge_size *
                                  tv_ptr->edge_size) +
                                 (j * tv_ptr->edge_size * tv_ptr->edge_size * tv_ptr->edge_size) +
                                 (k * tv_ptr->edge_size * tv_ptr->edge_size) + (l * tv_ptr->edge_size));
                    stop_index = start_index + tv_ptr->small_ds_slice_size - 1;

                    assert(start_index < stop_index);
                    assert(stop_index < tv_ptr->large_ds_size);

                    for (n = 0; n < tv_ptr->large_ds_size; n++) {

                        if ((n >= start_index) && (n <= stop_index)) {

                            if (*ptr_1 != expected_value) {

                                mis_match = true;
                            }

                            expected_value++;
                        }
                        else {

                            if (*ptr_1 != 0) {

                                mis_match = true;
                            }
                        }
                        /* zero out buffer for next test */
                        *ptr_1 = 0;
                        ptr_1++;
                    }

                    VRFY((mis_match == false), "small ds slice write to large ds slice data good.");

                    (tv_ptr->tests_run)++;
                }

                l++;

                (tv_ptr->total_tests)++;

            } while ((tv_ptr->large_rank > 2) && ((tv_ptr->small_rank - 1) <= 1) && (l < tv_ptr->edge_size));
            k++;
        } while ((tv_ptr->large_rank > 3) && ((tv_ptr->small_rank - 1) <= 2) && (k < tv_ptr->edge_size));
        j++;
    } while ((tv_ptr->large_rank > 4) && ((tv_ptr->small_rank - 1) <= 3) && (j < tv_ptr->edge_size));

    return;

} /* contig_hs_dr_pio_test__m2d_s2l() */

/*-------------------------------------------------------------------------
 * Function:    contig_hs_dr_pio_test__run_test()
 *
 * Purpose:    Test I/O to/from hyperslab selections of different rank in
 *        the parallel.
 *
 * Return:    void
 *
 *-------------------------------------------------------------------------
 */

#define CONTIG_HS_DR_PIO_TEST__RUN_TEST__DEBUG 0

static void
contig_hs_dr_pio_test__run_test(const int test_num, const int edge_size, const int chunk_edge_size,
                                const int small_rank, const int large_rank, const bool use_collective_io,
                                const hid_t dset_type, int express_test, int *skips_ptr, int max_skips,
                                int64_t *total_tests_ptr, int64_t *tests_run_ptr, int64_t *tests_skipped_ptr)
{
#if CONTIG_HS_DR_PIO_TEST__RUN_TEST__DEBUG
    const char *fcnName = "contig_hs_dr_pio_test__run_test()";
#endif /* CONTIG_HS_DR_PIO_TEST__RUN_TEST__DEBUG */
    struct hs_dr_pio_test_vars_t test_vars = {
        /* int           mpi_size                        = */ -1,
        /* int         mpi_rank                        = */ -1,
        /* MPI_Comm    mpi_comm                        = */ MPI_COMM_NULL,
        /* MPI_Inf     mpi_info                        = */ MPI_INFO_NULL,
        /* int         test_num                        = */ -1,
        /* int         edge_size                       = */ -1,
        /* int         checker_edge_size               = */ -1,
        /* int         chunk_edge_size                 = */ -1,
        /* int         small_rank                      = */ -1,
        /* int         large_rank                      = */ -1,
        /* hid_t       dset_type                       = */ H5I_INVALID_HID,
        /* uint32_t  * small_ds_buf_0                  = */ NULL,
        /* uint32_t  * small_ds_buf_1                  = */ NULL,
        /* uint32_t  * small_ds_buf_2                  = */ NULL,
        /* uint32_t  * small_ds_slice_buf              = */ NULL,
        /* uint32_t  * large_ds_buf_0                  = */ NULL,
        /* uint32_t  * large_ds_buf_1                  = */ NULL,
        /* uint32_t  * large_ds_buf_2                  = */ NULL,
        /* uint32_t  * large_ds_slice_buf              = */ NULL,
        /* int         small_ds_offset                 = */ -1,
        /* int         large_ds_offset                 = */ -1,
        /* hid_t       fid                             = */ H5I_INVALID_HID, /* HDF5 file ID */
        /* hid_t       xfer_plist                      = */ H5P_DEFAULT,
        /* hid_t       full_mem_small_ds_sid           = */ H5I_INVALID_HID,
        /* hid_t       full_file_small_ds_sid          = */ H5I_INVALID_HID,
        /* hid_t       mem_small_ds_sid                = */ H5I_INVALID_HID,
        /* hid_t       file_small_ds_sid_0             = */ H5I_INVALID_HID,
        /* hid_t       file_small_ds_sid_1             = */ H5I_INVALID_HID,
        /* hid_t       small_ds_slice_sid              = */ H5I_INVALID_HID,
        /* hid_t       full_mem_large_ds_sid           = */ H5I_INVALID_HID,
        /* hid_t       full_file_large_ds_sid          = */ H5I_INVALID_HID,
        /* hid_t       mem_large_ds_sid                = */ H5I_INVALID_HID,
        /* hid_t       file_large_ds_sid_0             = */ H5I_INVALID_HID,
        /* hid_t       file_large_ds_sid_1             = */ H5I_INVALID_HID,
        /* hid_t       file_large_ds_process_slice_sid = */ H5I_INVALID_HID,
        /* hid_t       mem_large_ds_process_slice_sid  = */ H5I_INVALID_HID,
        /* hid_t       large_ds_slice_sid              = */ H5I_INVALID_HID,
        /* hid_t       small_dataset                   = */ H5I_INVALID_HID, /* Dataset ID */
        /* hid_t       large_dataset                   = */ H5I_INVALID_HID, /* Dataset ID */
        /* size_t      small_ds_size                   = */ 1,
        /* size_t      small_ds_slice_size             = */ 1,
        /* size_t      large_ds_size                   = */ 1,
        /* size_t      large_ds_slice_size             = */ 1,
        /* hsize_t     dims[PAR_SS_DR_MAX_RANK]        = */ {0, 0, 0, 0, 0},
        /* hsize_t     chunk_dims[PAR_SS_DR_MAX_RANK]  = */ {0, 0, 0, 0, 0},
        /* hsize_t     start[PAR_SS_DR_MAX_RANK]       = */ {0, 0, 0, 0, 0},
        /* hsize_t     stride[PAR_SS_DR_MAX_RANK]      = */ {0, 0, 0, 0, 0},
        /* hsize_t     count[PAR_SS_DR_MAX_RANK]       = */ {0, 0, 0, 0, 0},
        /* hsize_t     block[PAR_SS_DR_MAX_RANK]       = */ {0, 0, 0, 0, 0},
        /* hsize_t   * start_ptr                       = */ NULL,
        /* hsize_t   * stride_ptr                      = */ NULL,
        /* hsize_t   * count_ptr                       = */ NULL,
        /* hsize_t   * block_ptr                       = */ NULL,
        /* int            skips                           = */ 0,
        /* int            max_skips                       = */ 0,
        /* int64_t     total_tests                     = */ 0,
        /* int64_t     tests_run                       = */ 0,
        /* int64_t     tests_skipped                   = */ 0};
    struct hs_dr_pio_test_vars_t *tv_ptr = &test_vars;

    hs_dr_pio_test__setup(test_num, edge_size, -1, chunk_edge_size, small_rank, large_rank, use_collective_io,
                          dset_type, express_test, tv_ptr);

    /* initialize skips & max_skips */
    tv_ptr->skips     = *skips_ptr;
    tv_ptr->max_skips = max_skips;

#if CONTIG_HS_DR_PIO_TEST__RUN_TEST__DEBUG
    if (MAINPROCESS) {
        fprintf(stdout, "test %d: small rank = %d, large rank = %d.\n", test_num, small_rank, large_rank);
        fprintf(stdout, "test %d: Initialization complete.\n", test_num);
    }
#endif /* CONTIG_HS_DR_PIO_TEST__RUN_TEST__DEBUG */

    /* first, verify that we can read from disk correctly using selections
     * of different rank that H5Sselect_shape_same() views as being of the
     * same shape.
     *
     * Start by reading small_rank - 1 dimensional slice from the on disk
     * large cube, and verifying that the data read is correct.  Verify that
     * H5Sselect_shape_same() returns true on the memory and file selections.
     */

#if CONTIG_HS_DR_PIO_TEST__RUN_TEST__DEBUG
    if (MAINPROCESS) {
        fprintf(stdout, "test %d: running contig_hs_dr_pio_test__d2m_l2s.\n", test_num);
    }
#endif /* CONTIG_HS_DR_PIO_TEST__RUN_TEST__DEBUG */
    contig_hs_dr_pio_test__d2m_l2s(tv_ptr);

    /* Second, read slices of the on disk small data set into slices
     * through the in memory large data set, and verify that the correct
     * data (and only the correct data) is read.
     */

#if CONTIG_HS_DR_PIO_TEST__RUN_TEST__DEBUG
    if (MAINPROCESS) {
        fprintf(stdout, "test %d: running contig_hs_dr_pio_test__d2m_s2l.\n", test_num);
    }
#endif /* CONTIG_HS_DR_PIO_TEST__RUN_TEST__DEBUG */
    contig_hs_dr_pio_test__d2m_s2l(tv_ptr);

    /* now we go in the opposite direction, verifying that we can write
     * from memory to file using selections of different rank that
     * H5Sselect_shape_same() views as being of the same shape.
     *
     * Start by writing small_rank - 1 D slices from the in memory large data
     * set to the on disk small cube dataset.  After each write, read the
     * slice of the small dataset back from disk, and verify that it contains
     * the expected data. Verify that H5Sselect_shape_same() returns true on
     * the memory and file selections.
     */

#if CONTIG_HS_DR_PIO_TEST__RUN_TEST__DEBUG
    if (MAINPROCESS) {
        fprintf(stdout, "test %d: running contig_hs_dr_pio_test__m2d_l2s.\n", test_num);
    }
#endif /* CONTIG_HS_DR_PIO_TEST__RUN_TEST__DEBUG */
    contig_hs_dr_pio_test__m2d_l2s(tv_ptr);

    /* Now write the contents of the process's slice of the in memory
     * small data set to slices of the on disk large data set.  After
     * each write, read the process's slice of the large data set back
     * into memory, and verify that it contains the expected data.
     * Verify that H5Sselect_shape_same() returns true on the memory
     * and file selections.
     */

#if CONTIG_HS_DR_PIO_TEST__RUN_TEST__DEBUG
    if (MAINPROCESS) {
        fprintf(stdout, "test %d: running contig_hs_dr_pio_test__m2d_s2l.\n", test_num);
    }
#endif /* CONTIG_HS_DR_PIO_TEST__RUN_TEST__DEBUG */
    contig_hs_dr_pio_test__m2d_s2l(tv_ptr);

#if CONTIG_HS_DR_PIO_TEST__RUN_TEST__DEBUG
    if (MAINPROCESS) {
        fprintf(stdout, "test %d: Subtests complete -- tests run/skipped/total = %lld/%lld/%lld.\n", test_num,
                (long long)(tv_ptr->tests_run), (long long)(tv_ptr->tests_skipped),
                (long long)(tv_ptr->total_tests));
    }
#endif /* CONTIG_HS_DR_PIO_TEST__RUN_TEST__DEBUG */

    hs_dr_pio_test__takedown(tv_ptr);

#if CONTIG_HS_DR_PIO_TEST__RUN_TEST__DEBUG
    if (MAINPROCESS) {
        fprintf(stdout, "test %d: Takedown complete.\n", test_num);
    }
#endif /* CONTIG_HS_DR_PIO_TEST__RUN_TEST__DEBUG */

    *skips_ptr = tv_ptr->skips;
    *total_tests_ptr += tv_ptr->total_tests;
    *tests_run_ptr += tv_ptr->tests_run;
    *tests_skipped_ptr += tv_ptr->tests_skipped;

    return;

} /* contig_hs_dr_pio_test__run_test() */

/*-------------------------------------------------------------------------
 * Function:    contig_hs_dr_pio_test(ShapeSameTestMethods sstest_type)
 *
 * Purpose:    Test I/O to/from hyperslab selections of different rank in
 *        the parallel case.
 *
 * Return:    void
 *
 *-------------------------------------------------------------------------
 */

#define CONTIG_HS_DR_PIO_TEST__DEBUG 0

static void
contig_hs_dr_pio_test(ShapeSameTestMethods sstest_type)
{
    int express_test;
    int local_express_test;
    int mpi_rank = -1;
    int mpi_size;
    int test_num = 0;
    int edge_size;
    int chunk_edge_size = 0;
    int small_rank;
    int large_rank;
    int mpi_result;
    int skips     = 0;
    int max_skips = 0;
    /* The following table list the number of sub-tests skipped between
     * each test that is actually executed as a function of the express
     * test level.  Note that any value in excess of 4880 will cause all
     * sub tests to be skipped.
     */
    int     max_skips_tbl[4] = {0, 4, 64, 1024};
    hid_t   dset_type        = H5T_NATIVE_UINT;
    int64_t total_tests      = 0;
    int64_t tests_run        = 0;
    int64_t tests_skipped    = 0;

    HDcompile_assert(sizeof(uint32_t) == sizeof(unsigned));

    MPI_Comm_size(MPI_COMM_WORLD, &mpi_size);
    MPI_Comm_rank(MPI_COMM_WORLD, &mpi_rank);

    edge_size = (mpi_size > 6 ? mpi_size : 6);

    local_express_test = GetTestExpress();

    mpi_result = MPI_Allreduce((void *)&local_express_test, (void *)&express_test, 1, MPI_INT, MPI_MAX,
                               MPI_COMM_WORLD);

    VRFY((mpi_result == MPI_SUCCESS), "MPI_Allreduce(0) succeeded");

    if (local_express_test < 0) {
        max_skips = max_skips_tbl[0];
    }
    else if (local_express_test > 3) {
        max_skips = max_skips_tbl[3];
    }
    else {
        max_skips = max_skips_tbl[local_express_test];
    }

    for (large_rank = 3; large_rank <= PAR_SS_DR_MAX_RANK; large_rank++) {

        for (small_rank = 2; small_rank < large_rank; small_rank++) {

            switch (sstest_type) {
                case IND_CONTIG:
                    /* contiguous data set, independent I/O */
                    chunk_edge_size = 0;

                    contig_hs_dr_pio_test__run_test(test_num, edge_size, chunk_edge_size, small_rank,
                                                    large_rank, false, dset_type, express_test, &skips,
                                                    max_skips, &total_tests, &tests_run, &tests_skipped);
                    test_num++;
                    break;
                    /* end of case IND_CONTIG */

                case COL_CONTIG:
                    /* contiguous data set, collective I/O */
                    chunk_edge_size = 0;

                    contig_hs_dr_pio_test__run_test(test_num, edge_size, chunk_edge_size, small_rank,
                                                    large_rank, true, dset_type, express_test, &skips,
                                                    max_skips, &total_tests, &tests_run, &tests_skipped);
                    test_num++;
                    break;
                    /* end of case COL_CONTIG */

                case IND_CHUNKED:
                    /* chunked data set, independent I/O */
                    chunk_edge_size = 5;

                    contig_hs_dr_pio_test__run_test(test_num, edge_size, chunk_edge_size, small_rank,
                                                    large_rank, false, dset_type, express_test, &skips,
                                                    max_skips, &total_tests, &tests_run, &tests_skipped);
                    test_num++;
                    break;
                    /* end of case IND_CHUNKED */

                case COL_CHUNKED:
                    /* chunked data set, collective I/O */
                    chunk_edge_size = 5;

                    contig_hs_dr_pio_test__run_test(test_num, edge_size, chunk_edge_size, small_rank,
                                                    large_rank, true, dset_type, express_test, &skips,
                                                    max_skips, &total_tests, &tests_run, &tests_skipped);
                    test_num++;
                    break;
                    /* end of case COL_CHUNKED */

                default:
                    VRFY((false), "unknown test type");
                    break;

            } /* end of switch(sstest_type) */
#if CONTIG_HS_DR_PIO_TEST__DEBUG
            if ((MAINPROCESS) && (tests_skipped > 0)) {
                fprintf(stdout, "    run/skipped/total = %lld/%lld/%lld.\n", tests_run, tests_skipped,
                        total_tests);
            }
#endif /* CONTIG_HS_DR_PIO_TEST__DEBUG */
        }
    }

    if ((MAINPROCESS) && (tests_skipped > 0)) {
        fprintf(stdout, "    %" PRId64 " of %" PRId64 " subtests skipped to expedite testing.\n",
                tests_skipped, total_tests);
    }

    return;

} /* contig_hs_dr_pio_test() */

/****************************************************************
**
**  ckrbrd_hs_dr_pio_test__slct_ckrbrd():
**    Given a dataspace of tgt_rank, and dimensions:
**
**        (mpi_size + 1), edge_size, ... , edge_size
**
**    edge_size, and a checker_edge_size, select a checker
**    board selection of a sel_rank (sel_rank < tgt_rank)
**    dimensional slice through the dataspace parallel to the
**      sel_rank fastest changing indices, with origin (in the
**    higher indices) as indicated by the start array.
**
**    Note that this function, like all its relatives, is
**    hard coded to presume a maximum dataspace rank of 5.
**    While this maximum is declared as a constant, increasing
**    it will require extensive coding in addition to changing
**      the value of the constant.
**
**                    JRM -- 10/8/09
**
****************************************************************/

#define CKRBRD_HS_DR_PIO_TEST__SELECT_CHECKER_BOARD__DEBUG 0

static void
ckrbrd_hs_dr_pio_test__slct_ckrbrd(const int mpi_rank, const hid_t tgt_sid, const int tgt_rank,
                                   const int edge_size, const int checker_edge_size, const int sel_rank,
                                   hsize_t sel_start[])
{
#if CKRBRD_HS_DR_PIO_TEST__SELECT_CHECKER_BOARD__DEBUG
    const char *fcnName = "ckrbrd_hs_dr_pio_test__slct_ckrbrd():";
#endif
    bool      first_selection = true;
    int       i, j, k, l, m;
    int       n_cube_offset;
    int       sel_offset;
    const int test_max_rank = PAR_SS_DR_MAX_RANK; /* must update code if */
                                                  /* this changes        */
    hsize_t base_count;
    hsize_t offset_count;
    hsize_t start[PAR_SS_DR_MAX_RANK];
    hsize_t stride[PAR_SS_DR_MAX_RANK];
    hsize_t count[PAR_SS_DR_MAX_RANK];
    hsize_t block[PAR_SS_DR_MAX_RANK];
    herr_t  ret; /* Generic return value */

    assert(edge_size >= 6);
    assert(0 < checker_edge_size);
    assert(checker_edge_size <= edge_size);
    assert(0 < sel_rank);
    assert(sel_rank <= tgt_rank);
    assert(tgt_rank <= test_max_rank);
    assert(test_max_rank <= PAR_SS_DR_MAX_RANK);

    sel_offset = test_max_rank - sel_rank;
    assert(sel_offset >= 0);

    n_cube_offset = test_max_rank - tgt_rank;
    assert(n_cube_offset >= 0);
    assert(n_cube_offset <= sel_offset);

#if CKRBRD_HS_DR_PIO_TEST__SELECT_CHECKER_BOARD__DEBUG
    fprintf(stdout, "%s:%d: edge_size/checker_edge_size = %d/%d\n", fcnName, mpi_rank, edge_size,
            checker_edge_size);
    fprintf(stdout, "%s:%d: sel_rank/sel_offset = %d/%d.\n", fcnName, mpi_rank, sel_rank, sel_offset);
    fprintf(stdout, "%s:%d: tgt_rank/n_cube_offset = %d/%d.\n", fcnName, mpi_rank, tgt_rank, n_cube_offset);
#endif /* CKRBRD_HS_DR_PIO_TEST__SELECT_CHECKER_BOARD__DEBUG */

    /* First, compute the base count (which assumes start == 0
     * for the associated offset) and offset_count (which
     * assumes start == checker_edge_size for the associated
     * offset).
     *
     * Note that the following computation depends on the C99
     * requirement that integer division discard any fraction
     * (truncation towards zero) to function correctly. As we
     * now require C99, this shouldn't be a problem, but noting
     * it may save us some pain if we are ever obliged to support
     * pre-C99 compilers again.
     */

    base_count = (hsize_t)(edge_size / (checker_edge_size * 2));

    if ((edge_size % (checker_edge_size * 2)) > 0) {

        base_count++;
    }

    offset_count = (hsize_t)((edge_size - checker_edge_size) / (checker_edge_size * 2));

    if (((edge_size - checker_edge_size) % (checker_edge_size * 2)) > 0) {

        offset_count++;
    }

    /* Now set up the stride and block arrays, and portions of the start
     * and count arrays that will not be altered during the selection of
     * the checker board.
     */
    i = 0;
    while (i < n_cube_offset) {

        /* these values should never be used */
        start[i]  = 0;
        stride[i] = 0;
        count[i]  = 0;
        block[i]  = 0;

        i++;
    }

    while (i < sel_offset) {

        start[i]  = sel_start[i];
        stride[i] = (hsize_t)(2 * edge_size);
        count[i]  = 1;
        block[i]  = 1;

        i++;
    }

    while (i < test_max_rank) {

        stride[i] = (hsize_t)(2 * checker_edge_size);
        block[i]  = (hsize_t)checker_edge_size;

        i++;
    }

    i = 0;
    do {
        if (0 >= sel_offset) {

            if (i == 0) {

                start[0] = 0;
                count[0] = base_count;
            }
            else {

                start[0] = (hsize_t)checker_edge_size;
                count[0] = offset_count;
            }
        }

        j = 0;
        do {
            if (1 >= sel_offset) {

                if (j == 0) {

                    start[1] = 0;
                    count[1] = base_count;
                }
                else {

                    start[1] = (hsize_t)checker_edge_size;
                    count[1] = offset_count;
                }
            }

            k = 0;
            do {
                if (2 >= sel_offset) {

                    if (k == 0) {

                        start[2] = 0;
                        count[2] = base_count;
                    }
                    else {

                        start[2] = (hsize_t)checker_edge_size;
                        count[2] = offset_count;
                    }
                }

                l = 0;
                do {
                    if (3 >= sel_offset) {

                        if (l == 0) {

                            start[3] = 0;
                            count[3] = base_count;
                        }
                        else {

                            start[3] = (hsize_t)checker_edge_size;
                            count[3] = offset_count;
                        }
                    }

                    m = 0;
                    do {
                        if (4 >= sel_offset) {

                            if (m == 0) {

                                start[4] = 0;
                                count[4] = base_count;
                            }
                            else {

                                start[4] = (hsize_t)checker_edge_size;
                                count[4] = offset_count;
                            }
                        }

                        if (((i + j + k + l + m) % 2) == 0) {

#if CKRBRD_HS_DR_PIO_TEST__SELECT_CHECKER_BOARD__DEBUG
                            fprintf(stdout, "%s%d: *** first_selection = %d ***\n", fcnName, mpi_rank,
                                    (int)first_selection);
                            fprintf(stdout, "%s:%d: i/j/k/l/m = %d/%d/%d/%d/%d\n", fcnName, mpi_rank, i, j, k,
                                    l, m);
                            fprintf(stdout, "%s:%d: start = %d %d %d %d %d.\n", fcnName, mpi_rank,
                                    (int)start[0], (int)start[1], (int)start[2], (int)start[3],
                                    (int)start[4]);
                            fprintf(stdout, "%s:%d: stride = %d %d %d %d %d.\n", fcnName, mpi_rank,
                                    (int)stride[0], (int)stride[1], (int)stride[2], (int)stride[3],
                                    (int)stride[4]);
                            fprintf(stdout, "%s:%d: count = %d %d %d %d %d.\n", fcnName, mpi_rank,
                                    (int)count[0], (int)count[1], (int)count[2], (int)count[3],
                                    (int)count[4]);
                            fprintf(stdout, "%s:%d: block = %d %d %d %d %d.\n", fcnName, mpi_rank,
                                    (int)block[0], (int)block[1], (int)block[2], (int)block[3],
                                    (int)block[4]);
                            fprintf(stdout, "%s:%d: n-cube extent dims = %d.\n", fcnName, mpi_rank,
                                    H5Sget_simple_extent_ndims(tgt_sid));
                            fprintf(stdout, "%s:%d: selection rank = %d.\n", fcnName, mpi_rank, sel_rank);
#endif

                            if (first_selection) {

                                first_selection = false;

                                ret = H5Sselect_hyperslab(tgt_sid, H5S_SELECT_SET, &(start[n_cube_offset]),
                                                          &(stride[n_cube_offset]), &(count[n_cube_offset]),
                                                          &(block[n_cube_offset]));

                                VRFY((ret != FAIL), "H5Sselect_hyperslab(SET) succeeded");
                            }
                            else {

                                ret = H5Sselect_hyperslab(tgt_sid, H5S_SELECT_OR, &(start[n_cube_offset]),
                                                          &(stride[n_cube_offset]), &(count[n_cube_offset]),
                                                          &(block[n_cube_offset]));

                                VRFY((ret != FAIL), "H5Sselect_hyperslab(OR) succeeded");
                            }
                        }

                        m++;

                    } while ((m <= 1) && (4 >= sel_offset));

                    l++;

                } while ((l <= 1) && (3 >= sel_offset));

                k++;

            } while ((k <= 1) && (2 >= sel_offset));

            j++;

        } while ((j <= 1) && (1 >= sel_offset));

        i++;

    } while ((i <= 1) && (0 >= sel_offset));

#if CKRBRD_HS_DR_PIO_TEST__SELECT_CHECKER_BOARD__DEBUG
    fprintf(stdout, "%s%d: H5Sget_select_npoints(tgt_sid) = %d.\n", fcnName, mpi_rank,
            (int)H5Sget_select_npoints(tgt_sid));
#endif /* CKRBRD_HS_DR_PIO_TEST__SELECT_CHECKER_BOARD__DEBUG */

    /* Clip the selection back to the dataspace proper. */

    for (i = 0; i < test_max_rank; i++) {

        start[i]  = 0;
        stride[i] = (hsize_t)edge_size;
        count[i]  = 1;
        block[i]  = (hsize_t)edge_size;
    }

    ret = H5Sselect_hyperslab(tgt_sid, H5S_SELECT_AND, start, stride, count, block);

    VRFY((ret != FAIL), "H5Sselect_hyperslab(AND) succeeded");

#if CKRBRD_HS_DR_PIO_TEST__SELECT_CHECKER_BOARD__DEBUG
    fprintf(stdout, "%s%d: H5Sget_select_npoints(tgt_sid) = %d.\n", fcnName, mpi_rank,
            (int)H5Sget_select_npoints(tgt_sid));
    fprintf(stdout, "%s%d: done.\n", fcnName, mpi_rank);
#endif /* CKRBRD_HS_DR_PIO_TEST__SELECT_CHECKER_BOARD__DEBUG */

    return;

} /* ckrbrd_hs_dr_pio_test__slct_ckrbrd() */

/****************************************************************
**
**  ckrbrd_hs_dr_pio_test__verify_data():
**
**    Examine the supplied buffer to see if it contains the
**    expected data.  Return true if it does, and false
**      otherwise.
**
**    The supplied buffer is presumed to this process's slice
**    of the target data set.  Each such slice will be an
**    n-cube of rank (rank -1) and the supplied edge_size with
**    origin (mpi_rank, 0, ... , 0) in the target data set.
**
**    Further, the buffer is presumed to be the result of reading
**    or writing a checker board selection of an m (1 <= m <
**      rank) dimensional slice through this processes slice
**    of the target data set.  Also, this slice must be parallel
**    to the fastest changing indices.
**
**    It is further presumed that the buffer was zeroed before
**    the read/write, and that the full target data set (i.e.
**    the buffer/data set for all processes) was initialized
**      with the natural numbers listed in order from the origin
**    along the fastest changing axis.
**
**      Thus for a 20x10x10 dataset, the value stored in location
**    (x, y, z) (assuming that z is the fastest changing index
**    and x the slowest) is assumed to be:
**
**        (10 * 10 * x) + (10 * y) + z
**
**    Further, supposing that this is process 10, this process's
**    slice of the dataset would be a 10 x 10 2-cube with origin
**    (10, 0, 0) in the data set, and would be initialize (prior
**    to the checkerboard selection) as follows:
**
**        1000, 1001, 1002, ... 1008, 1009
**        1010, 1011, 1012, ... 1018, 1019
**          .     .     .         .     .
**          .     .     .         .     .
**          .     .     .         .     .
**        1090, 1091, 1092, ... 1098, 1099
**
**    In the case of a read from the processors slice of another
**    data set of different rank, the values expected will have
**    to be adjusted accordingly.  This is done via the
**    first_expected_val parameter.
**
**    Finally, the function presumes that the first element
**    of the buffer resides either at the origin of either
**    a selected or an unselected checker.  (Translation:
**    if partial checkers appear in the buffer, they will
**    intersect the edges of the n-cube opposite the origin.)
**
****************************************************************/

#define CKRBRD_HS_DR_PIO_TEST__VERIFY_DATA__DEBUG 0

static bool
ckrbrd_hs_dr_pio_test__verify_data(uint32_t *buf_ptr, const int rank, const int edge_size,
                                   const int checker_edge_size, uint32_t first_expected_val,
                                   bool buf_starts_in_checker)
{
#if CKRBRD_HS_DR_PIO_TEST__VERIFY_DATA__DEBUG
    const char *fcnName = "ckrbrd_hs_dr_pio_test__verify_data():";
#endif
    bool      good_data = true;
    bool      in_checker;
    bool      start_in_checker[5];
    uint32_t  expected_value;
    uint32_t *val_ptr;
    int       i, j, k, l, m;     /* to track position in n-cube */
    int       v, w, x, y, z;     /* to track position in checker */
    const int test_max_rank = 5; /* code changes needed if this is increased */

    assert(buf_ptr != NULL);
    assert(0 < rank);
    assert(rank <= test_max_rank);
    assert(edge_size >= 6);
    assert(0 < checker_edge_size);
    assert(checker_edge_size <= edge_size);
    assert(test_max_rank <= PAR_SS_DR_MAX_RANK);

#if CKRBRD_HS_DR_PIO_TEST__VERIFY_DATA__DEBUG

    int mpi_rank;

    MPI_Comm_rank(MPI_COMM_WORLD, &mpi_rank);
    fprintf(stdout, "%s mpi_rank = %d.\n", fcnName, mpi_rank);
    fprintf(stdout, "%s rank = %d.\n", fcnName, rank);
    fprintf(stdout, "%s edge_size = %d.\n", fcnName, edge_size);
    fprintf(stdout, "%s checker_edge_size = %d.\n", fcnName, checker_edge_size);
    fprintf(stdout, "%s first_expected_val = %d.\n", fcnName, (int)first_expected_val);
    fprintf(stdout, "%s starts_in_checker = %d.\n", fcnName, (int)buf_starts_in_checker);
}
#endif

val_ptr        = buf_ptr;
expected_value = first_expected_val;

i                   = 0;
v                   = 0;
start_in_checker[0] = buf_starts_in_checker;
do {
    if (v >= checker_edge_size) {

        start_in_checker[0] = !start_in_checker[0];
        v                   = 0;
    }

    j                   = 0;
    w                   = 0;
    start_in_checker[1] = start_in_checker[0];
    do {
        if (w >= checker_edge_size) {

            start_in_checker[1] = !start_in_checker[1];
            w                   = 0;
        }

        k                   = 0;
        x                   = 0;
        start_in_checker[2] = start_in_checker[1];
        do {
            if (x >= checker_edge_size) {

                start_in_checker[2] = !start_in_checker[2];
                x                   = 0;
            }

            l                   = 0;
            y                   = 0;
            start_in_checker[3] = start_in_checker[2];
            do {
                if (y >= checker_edge_size) {

                    start_in_checker[3] = !start_in_checker[3];
                    y                   = 0;
                }

                m = 0;
                z = 0;
#if CKRBRD_HS_DR_PIO_TEST__VERIFY_DATA__DEBUG
                fprintf(stdout, "%d, %d, %d, %d, %d:", i, j, k, l, m);
#endif
                in_checker = start_in_checker[3];
                do {
#if CKRBRD_HS_DR_PIO_TEST__VERIFY_DATA__DEBUG
                    fprintf(stdout, " %d", (int)(*val_ptr));
#endif
                    if (z >= checker_edge_size) {

                        in_checker = !in_checker;
                        z          = 0;
                    }

                    if (in_checker) {

                        if (*val_ptr != expected_value) {

                            good_data = false;
                        }

                        /* zero out buffer for reuse */
                        *val_ptr = 0;
                    }
                    else if (*val_ptr != 0) {

                        good_data = false;

                        /* zero out buffer for reuse */
                        *val_ptr = 0;
                    }

                    val_ptr++;
                    expected_value++;
                    m++;
                    z++;

                } while ((rank >= (test_max_rank - 4)) && (m < edge_size));
#if CKRBRD_HS_DR_PIO_TEST__VERIFY_DATA__DEBUG
                fprintf(stdout, "\n");
#endif
                l++;
                y++;
            } while ((rank >= (test_max_rank - 3)) && (l < edge_size));
            k++;
            x++;
        } while ((rank >= (test_max_rank - 2)) && (k < edge_size));
        j++;
        w++;
    } while ((rank >= (test_max_rank - 1)) && (j < edge_size));
    i++;
    v++;
} while ((rank >= test_max_rank) && (i < edge_size));

return (good_data);

} /* ckrbrd_hs_dr_pio_test__verify_data() */

/*-------------------------------------------------------------------------
 * Function:    ckrbrd_hs_dr_pio_test__d2m_l2s()
 *
 * Purpose:    Part one of a series of tests of I/O to/from hyperslab
 *        selections of different rank in the parallel.
 *
 *        Verify that we can read from disk correctly using checker
 *        board selections of different rank that
 *        H5Sselect_shape_same() views as being of the same shape.
 *
 *        In this function, we test this by reading small_rank - 1
 *        checker board slices from the on disk large cube, and
 *        verifying that the data read is correct.  Verify that
 *        H5Sselect_shape_same() returns true on the memory and
 *        file selections.
 *
 * Return:    void
 *
 *-------------------------------------------------------------------------
 */

#define CHECKER_BOARD_HS_DR_PIO_TEST__D2M_L2S__DEBUG 0

static void
ckrbrd_hs_dr_pio_test__d2m_l2s(struct hs_dr_pio_test_vars_t *tv_ptr)
{
#if CHECKER_BOARD_HS_DR_PIO_TEST__D2M_L2S__DEBUG
    const char *fcnName = "ckrbrd_hs_dr_pio_test__d2m_l2s()";
    uint32_t   *ptr_0;
#endif /* CHECKER_BOARD_HS_DR_PIO_TEST__D2M_L2S__DEBUG */
    bool     data_ok = false;
    int      i, j, k, l;
    uint32_t expected_value;
    int      mpi_rank; /* needed by VRFY */
    hsize_t  sel_start[PAR_SS_DR_MAX_RANK];
    htri_t   check; /* Shape comparison return value */
    herr_t   ret;   /* Generic return value */

    /* initialize the local copy of mpi_rank */
    mpi_rank = tv_ptr->mpi_rank;

    /* first, verify that we can read from disk correctly using selections
     * of different rank that H5Sselect_shape_same() views as being of the
     * same shape.
     *
     * Start by reading a (small_rank - 1)-D checker board slice from this
     * processes slice of the on disk large data set, and verifying that the
     * data read is correct.  Verify that H5Sselect_shape_same() returns
     * true on the memory and file selections.
     *
     * The first step is to set up the needed checker board selection in the
     * in memory small small cube
     */

    sel_start[0] = sel_start[1] = sel_start[2] = sel_start[3] = sel_start[4] = 0;
    sel_start[tv_ptr->small_ds_offset]                                       = (hsize_t)(tv_ptr->mpi_rank);

    ckrbrd_hs_dr_pio_test__slct_ckrbrd(tv_ptr->mpi_rank, tv_ptr->small_ds_slice_sid, tv_ptr->small_rank - 1,
                                       tv_ptr->edge_size, tv_ptr->checker_edge_size, tv_ptr->small_rank - 1,
                                       sel_start);

    /* zero out the buffer we will be reading into */
    memset(tv_ptr->small_ds_slice_buf, 0, sizeof(uint32_t) * tv_ptr->small_ds_slice_size);

#if CHECKER_BOARD_HS_DR_PIO_TEST__D2M_L2S__DEBUG
    fprintf(stdout, "%s:%d: initial small_ds_slice_buf = ", fcnName, tv_ptr->mpi_rank);
    ptr_0 = tv_ptr->small_ds_slice_buf;
    for (i = 0; i < (int)(tv_ptr->small_ds_slice_size); i++) {
        fprintf(stdout, "%d ", (int)(*ptr_0));
        ptr_0++;
    }
    fprintf(stdout, "\n");
#endif /* CHECKER_BOARD_HS_DR_PIO_TEST__D2M_L2S__DEBUG */

    /* set up start, stride, count, and block -- note that we will
     * change start[] so as to read slices of the large cube.
     */
    for (i = 0; i < PAR_SS_DR_MAX_RANK; i++) {

        tv_ptr->start[i]  = 0;
        tv_ptr->stride[i] = (hsize_t)(2 * tv_ptr->edge_size);
        tv_ptr->count[i]  = 1;
        if ((PAR_SS_DR_MAX_RANK - i) > (tv_ptr->small_rank - 1)) {

            tv_ptr->block[i] = 1;
        }
        else {

            tv_ptr->block[i] = (hsize_t)(tv_ptr->edge_size);
        }
    }

#if CHECKER_BOARD_HS_DR_PIO_TEST__D2M_L2S__DEBUG
    fprintf(stdout, "%s:%d: reading slice from big ds on disk into small ds slice.\n", fcnName,
            tv_ptr->mpi_rank);
#endif /* CHECKER_BOARD_HS_DR_PIO_TEST__D2M_L2S__DEBUG */
    /* in serial versions of this test, we loop through all the dimensions
     * of the large data set.  However, in the parallel version, each
     * process only works with that slice of the large cube indicated
     * by its rank -- hence we set the most slowly changing index to
     * mpi_rank, and don't iterate over it.
     */

    if (PAR_SS_DR_MAX_RANK - tv_ptr->large_rank == 0) {

        i = tv_ptr->mpi_rank;
    }
    else {

        i = 0;
    }

    /* since large_rank is at most PAR_SS_DR_MAX_RANK, no need to
     * loop over it -- either we are setting i to mpi_rank, or
     * we are setting it to zero.  It will not change during the
     * test.
     */

    if (PAR_SS_DR_MAX_RANK - tv_ptr->large_rank == 1) {

        j = tv_ptr->mpi_rank;
    }
    else {

        j = 0;
    }

    do {
        if (PAR_SS_DR_MAX_RANK - tv_ptr->large_rank == 2) {

            k = tv_ptr->mpi_rank;
        }
        else {

            k = 0;
        }

        do {
            /* since small rank >= 2 and large_rank > small_rank, we
             * have large_rank >= 3.  Since PAR_SS_DR_MAX_RANK == 5
             * (baring major re-orgaization), this gives us:
             *
             *     (PAR_SS_DR_MAX_RANK - large_rank) <= 2
             *
             * so no need to repeat the test in the outer loops --
             * just set l = 0.
             */

            l = 0;
            do {
                if ((tv_ptr->skips)++ < tv_ptr->max_skips) { /* skip the test */

                    (tv_ptr->tests_skipped)++;
                }
                else { /* run the test */

                    tv_ptr->skips = 0; /* reset the skips counter */

                    /* we know that small_rank - 1 >= 1 and that
                     * large_rank > small_rank by the assertions at the head
                     * of this function.  Thus no need for another inner loop.
                     */
                    tv_ptr->start[0] = (hsize_t)i;
                    tv_ptr->start[1] = (hsize_t)j;
                    tv_ptr->start[2] = (hsize_t)k;
                    tv_ptr->start[3] = (hsize_t)l;
                    tv_ptr->start[4] = 0;

                    assert((tv_ptr->start[0] == 0) || (0 < tv_ptr->small_ds_offset + 1));
                    assert((tv_ptr->start[1] == 0) || (1 < tv_ptr->small_ds_offset + 1));
                    assert((tv_ptr->start[2] == 0) || (2 < tv_ptr->small_ds_offset + 1));
                    assert((tv_ptr->start[3] == 0) || (3 < tv_ptr->small_ds_offset + 1));
                    assert((tv_ptr->start[4] == 0) || (4 < tv_ptr->small_ds_offset + 1));

                    ckrbrd_hs_dr_pio_test__slct_ckrbrd(
                        tv_ptr->mpi_rank, tv_ptr->file_large_ds_sid_0, tv_ptr->large_rank, tv_ptr->edge_size,
                        tv_ptr->checker_edge_size, tv_ptr->small_rank - 1, tv_ptr->start);

                    /* verify that H5Sselect_shape_same() reports the two
                     * selections as having the same shape.
                     */
                    check = H5Sselect_shape_same(tv_ptr->small_ds_slice_sid, tv_ptr->file_large_ds_sid_0);
                    VRFY((check == true), "H5Sselect_shape_same passed");

                    /* Read selection from disk */
#if CHECKER_BOARD_HS_DR_PIO_TEST__D2M_L2S__DEBUG
                    fprintf(stdout, "%s:%d: start = %d %d %d %d %d.\n", fcnName, tv_ptr->mpi_rank,
                            tv_ptr->start[0], tv_ptr->start[1], tv_ptr->start[2], tv_ptr->start[3],
                            tv_ptr->start[4]);
                    fprintf(stdout, "%s slice/file extent dims = %d/%d.\n", fcnName,
                            H5Sget_simple_extent_ndims(tv_ptr->small_ds_slice_sid),
                            H5Sget_simple_extent_ndims(tv_ptr->file_large_ds_sid_0));
#endif /* CHECKER_BOARD_HS_DR_PIO_TEST__D2M_L2S__DEBUG */

                    ret =
                        H5Dread(tv_ptr->large_dataset, H5T_NATIVE_UINT32, tv_ptr->small_ds_slice_sid,
                                tv_ptr->file_large_ds_sid_0, tv_ptr->xfer_plist, tv_ptr->small_ds_slice_buf);
                    VRFY((ret >= 0), "H5Dread() slice from large ds succeeded.");

#if CHECKER_BOARD_HS_DR_PIO_TEST__D2M_L2S__DEBUG
                    fprintf(stdout, "%s:%d: H5Dread() returns.\n", fcnName, tv_ptr->mpi_rank);
#endif /* CHECKER_BOARD_HS_DR_PIO_TEST__D2M_L2S__DEBUG */

                    /* verify that expected data is retrieved */

                    expected_value =
                        (uint32_t)((i * tv_ptr->edge_size * tv_ptr->edge_size * tv_ptr->edge_size *
                                    tv_ptr->edge_size) +
                                   (j * tv_ptr->edge_size * tv_ptr->edge_size * tv_ptr->edge_size) +
                                   (k * tv_ptr->edge_size * tv_ptr->edge_size) + (l * tv_ptr->edge_size));

                    data_ok = ckrbrd_hs_dr_pio_test__verify_data(
                        tv_ptr->small_ds_slice_buf, tv_ptr->small_rank - 1, tv_ptr->edge_size,
                        tv_ptr->checker_edge_size, expected_value, (bool)true);

                    VRFY((data_ok == true), "small slice read from large ds data good.");

                    (tv_ptr->tests_run)++;
                }

                l++;

                (tv_ptr->total_tests)++;

            } while ((tv_ptr->large_rank > 2) && ((tv_ptr->small_rank - 1) <= 1) && (l < tv_ptr->edge_size));
            k++;
        } while ((tv_ptr->large_rank > 3) && ((tv_ptr->small_rank - 1) <= 2) && (k < tv_ptr->edge_size));
        j++;
    } while ((tv_ptr->large_rank > 4) && ((tv_ptr->small_rank - 1) <= 3) && (j < tv_ptr->edge_size));

    return;

} /* ckrbrd_hs_dr_pio_test__d2m_l2s() */

/*-------------------------------------------------------------------------
 * Function:    ckrbrd_hs_dr_pio_test__d2m_s2l()
 *
 * Purpose:    Part two of a series of tests of I/O to/from hyperslab
 *        selections of different rank in the parallel.
 *
 *        Verify that we can read from disk correctly using
 *        selections of different rank that H5Sselect_shape_same()
 *        views as being of the same shape.
 *
 *        In this function, we test this by reading checker board
 *        slices of the on disk small data set into slices through
 *        the in memory large data set, and verify that the correct
 *        data (and only the correct data) is read.
 *
 * Return:    void
 *
 *-------------------------------------------------------------------------
 */

#define CHECKER_BOARD_HS_DR_PIO_TEST__D2M_S2L__DEBUG 0

static void
ckrbrd_hs_dr_pio_test__d2m_s2l(struct hs_dr_pio_test_vars_t *tv_ptr)
{
#if CHECKER_BOARD_HS_DR_PIO_TEST__D2M_S2L__DEBUG
    const char *fcnName = "ckrbrd_hs_dr_pio_test__d2m_s2l()";
#endif /* CHECKER_BOARD_HS_DR_PIO_TEST__D2M_S2L__DEBUG */
    bool      data_ok = false;
    int       i, j, k, l;
    size_t    u;
    size_t    start_index;
    size_t    stop_index;
    uint32_t  expected_value;
    uint32_t *ptr_1;
    int       mpi_rank; /* needed by VRFY */
    hsize_t   sel_start[PAR_SS_DR_MAX_RANK];
    htri_t    check; /* Shape comparison return value */
    herr_t    ret;   /* Generic return value */

    /* initialize the local copy of mpi_rank */
    mpi_rank = tv_ptr->mpi_rank;

    /* similarly, read slices of the on disk small data set into slices
     * through the in memory large data set, and verify that the correct
     * data (and only the correct data) is read.
     */

    sel_start[0] = sel_start[1] = sel_start[2] = sel_start[3] = sel_start[4] = 0;
    sel_start[tv_ptr->small_ds_offset]                                       = (hsize_t)(tv_ptr->mpi_rank);

    ckrbrd_hs_dr_pio_test__slct_ckrbrd(tv_ptr->mpi_rank, tv_ptr->file_small_ds_sid_0, tv_ptr->small_rank,
                                       tv_ptr->edge_size, tv_ptr->checker_edge_size, tv_ptr->small_rank - 1,
                                       sel_start);

#if CHECKER_BOARD_HS_DR_PIO_TEST__D2M_S2L__DEBUG
    fprintf(stdout, "%s reading slices of on disk small data set into slices of big data set.\n", fcnName);
#endif /* CHECKER_BOARD_HS_DR_PIO_TEST__D2M_S2L__DEBUG */

    /* zero out the buffer we will be reading into */
    memset(tv_ptr->large_ds_buf_1, 0, sizeof(uint32_t) * tv_ptr->large_ds_size);

    /* set up start, stride, count, and block -- note that we will
     * change start[] so as to read the slice of the small data set
     * into different slices of the process slice of the large data
     * set.
     */
    for (i = 0; i < PAR_SS_DR_MAX_RANK; i++) {

        tv_ptr->start[i]  = 0;
        tv_ptr->stride[i] = (hsize_t)(2 * tv_ptr->edge_size);
        tv_ptr->count[i]  = 1;
        if ((PAR_SS_DR_MAX_RANK - i) > (tv_ptr->small_rank - 1)) {

            tv_ptr->block[i] = 1;
        }
        else {

            tv_ptr->block[i] = (hsize_t)(tv_ptr->edge_size);
        }
    }

    /* in serial versions of this test, we loop through all the dimensions
     * of the large data set that don't appear in the small data set.
     *
     * However, in the parallel version, each process only works with that
     * slice of the large (and small) data set indicated by its rank -- hence
     * we set the most slowly changing index to mpi_rank, and don't iterate
     * over it.
     */

    if (PAR_SS_DR_MAX_RANK - tv_ptr->large_rank == 0) {

        i = tv_ptr->mpi_rank;
    }
    else {

        i = 0;
    }

    /* since large_rank is at most PAR_SS_DR_MAX_RANK, no need to
     * loop over it -- either we are setting i to mpi_rank, or
     * we are setting it to zero.  It will not change during the
     * test.
     */

    if (PAR_SS_DR_MAX_RANK - tv_ptr->large_rank == 1) {

        j = tv_ptr->mpi_rank;
    }
    else {

        j = 0;
    }

    do {
        if (PAR_SS_DR_MAX_RANK - tv_ptr->large_rank == 2) {

            k = tv_ptr->mpi_rank;
        }
        else {

            k = 0;
        }

        do {
            /* since small rank >= 2 and large_rank > small_rank, we
             * have large_rank >= 3.  Since PAR_SS_DR_MAX_RANK == 5
             * (baring major re-orgaization), this gives us:
             *
             *     (PAR_SS_DR_MAX_RANK - large_rank) <= 2
             *
             * so no need to repeat the test in the outer loops --
             * just set l = 0.
             */

            l = 0;
            do {
                if ((tv_ptr->skips)++ < tv_ptr->max_skips) { /* skip the test */

                    (tv_ptr->tests_skipped)++;
                }
                else { /* run the test */

                    tv_ptr->skips = 0; /* reset the skips counter */

                    /* we know that small_rank >= 1 and that large_rank > small_rank
                     * by the assertions at the head of this function.  Thus no
                     * need for another inner loop.
                     */
                    tv_ptr->start[0] = (hsize_t)i;
                    tv_ptr->start[1] = (hsize_t)j;
                    tv_ptr->start[2] = (hsize_t)k;
                    tv_ptr->start[3] = (hsize_t)l;
                    tv_ptr->start[4] = 0;

                    assert((tv_ptr->start[0] == 0) || (0 < tv_ptr->small_ds_offset + 1));
                    assert((tv_ptr->start[1] == 0) || (1 < tv_ptr->small_ds_offset + 1));
                    assert((tv_ptr->start[2] == 0) || (2 < tv_ptr->small_ds_offset + 1));
                    assert((tv_ptr->start[3] == 0) || (3 < tv_ptr->small_ds_offset + 1));
                    assert((tv_ptr->start[4] == 0) || (4 < tv_ptr->small_ds_offset + 1));

                    ckrbrd_hs_dr_pio_test__slct_ckrbrd(
                        tv_ptr->mpi_rank, tv_ptr->mem_large_ds_sid, tv_ptr->large_rank, tv_ptr->edge_size,
                        tv_ptr->checker_edge_size, tv_ptr->small_rank - 1, tv_ptr->start);

                    /* verify that H5Sselect_shape_same() reports the two
                     * selections as having the same shape.
                     */
                    check = H5Sselect_shape_same(tv_ptr->file_small_ds_sid_0, tv_ptr->mem_large_ds_sid);
                    VRFY((check == true), "H5Sselect_shape_same passed");

                    /* Read selection from disk */
#if CHECKER_BOARD_HS_DR_PIO_TEST__D2M_S2L__DEBUG
                    fprintf(stdout, "%s:%d: start = %d %d %d %d %d.\n", fcnName, tv_ptr->mpi_rank,
                            tv_ptr->start[0], tv_ptr->start[1], tv_ptr->start[2], tv_ptr->start[3],
                            tv_ptr->start[4]);
                    fprintf(stdout, "%s:%d: mem/file extent dims = %d/%d.\n", fcnName, tv_ptr->mpi_rank,
                            H5Sget_simple_extent_ndims(tv_ptr->large_ds_slice_sid),
                            H5Sget_simple_extent_ndims(tv_ptr->file_small_ds_sid_0));
#endif /* CHECKER_BOARD_HS_DR_PIO_TEST__D2M_S2L__DEBUG */
                    ret = H5Dread(tv_ptr->small_dataset, H5T_NATIVE_UINT32, tv_ptr->mem_large_ds_sid,
                                  tv_ptr->file_small_ds_sid_0, tv_ptr->xfer_plist, tv_ptr->large_ds_buf_1);
                    VRFY((ret >= 0), "H5Dread() slice from small ds succeeded.");

                    /* verify that the expected data and only the
                     * expected data was read.
                     */
                    data_ok        = true;
                    ptr_1          = tv_ptr->large_ds_buf_1;
                    expected_value = (uint32_t)((size_t)(tv_ptr->mpi_rank) * tv_ptr->small_ds_slice_size);
                    start_index =
                        (size_t)((i * tv_ptr->edge_size * tv_ptr->edge_size * tv_ptr->edge_size *
                                  tv_ptr->edge_size) +
                                 (j * tv_ptr->edge_size * tv_ptr->edge_size * tv_ptr->edge_size) +
                                 (k * tv_ptr->edge_size * tv_ptr->edge_size) + (l * tv_ptr->edge_size));
                    stop_index = start_index + tv_ptr->small_ds_slice_size - 1;

#if CHECKER_BOARD_HS_DR_PIO_TEST__D2M_S2L__DEBUG
                    {
                        int m, n;

                        fprintf(stdout, "%s:%d: expected_value = %d.\n", fcnName, tv_ptr->mpi_rank,
                                expected_value);
                        fprintf(stdout, "%s:%d: start/stop index = %d/%d.\n", fcnName, tv_ptr->mpi_rank,
                                start_index, stop_index);
                        n = 0;
                        for (m = 0; (unsigned)m < tv_ptr->large_ds_size; m++) {
                            fprintf(stdout, "%d ", (int)(*ptr_1));
                            ptr_1++;
                            n++;
                            if (n >= tv_ptr->edge_size) {
                                fprintf(stdout, "\n");
                                n = 0;
                            }
                        }
                        fprintf(stdout, "\n");
                        ptr_1 = tv_ptr->large_ds_buf_1;
                    }
#endif /* CHECKER_BOARD_HS_DR_PIO_TEST__D2M_S2L__DEBUG */

                    assert(start_index < stop_index);
                    assert(stop_index <= tv_ptr->large_ds_size);

                    for (u = 0; u < start_index; u++) {

                        if (*ptr_1 != 0) {

                            data_ok = false;
                        }

                        /* zero out the value for the next pass */
                        *ptr_1 = 0;

                        ptr_1++;
                    }

                    VRFY((data_ok == true), "slice read from small to large ds data good(1).");

                    data_ok = ckrbrd_hs_dr_pio_test__verify_data(ptr_1, tv_ptr->small_rank - 1,
                                                                 tv_ptr->edge_size, tv_ptr->checker_edge_size,
                                                                 expected_value, (bool)true);

                    VRFY((data_ok == true), "slice read from small to large ds data good(2).");

                    ptr_1 = tv_ptr->large_ds_buf_1 + stop_index + 1;

                    for (u = stop_index + 1; u < tv_ptr->large_ds_size; u++) {

                        if (*ptr_1 != 0) {

                            data_ok = false;
                        }

                        /* zero out the value for the next pass */
                        *ptr_1 = 0;

                        ptr_1++;
                    }

                    VRFY((data_ok == true), "slice read from small to large ds data good(3).");

                    (tv_ptr->tests_run)++;
                }

                l++;

                (tv_ptr->total_tests)++;

            } while ((tv_ptr->large_rank > 2) && ((tv_ptr->small_rank - 1) <= 1) && (l < tv_ptr->edge_size));
            k++;
        } while ((tv_ptr->large_rank > 3) && ((tv_ptr->small_rank - 1) <= 2) && (k < tv_ptr->edge_size));
        j++;
    } while ((tv_ptr->large_rank > 4) && ((tv_ptr->small_rank - 1) <= 3) && (j < tv_ptr->edge_size));

    return;

} /* ckrbrd_hs_dr_pio_test__d2m_s2l() */

/*-------------------------------------------------------------------------
 * Function:    ckrbrd_hs_dr_pio_test__m2d_l2s()
 *
 * Purpose:    Part three of a series of tests of I/O to/from checker
 *        board hyperslab selections of different rank in the
 *        parallel.
 *
 *        Verify that we can write from memory to file using checker
 *        board selections of different rank that
 *        H5Sselect_shape_same() views as being of the same shape.
 *
 *        Do this by writing small_rank - 1 dimensional checker
 *        board slices from the in memory large data set to the on
 *        disk small cube dataset.  After each write, read the
 *        slice of the small dataset back from disk, and verify
 *        that it contains the expected data. Verify that
 *        H5Sselect_shape_same() returns true on the memory and
 *        file selections.
 *
 * Return:    void
 *
 *-------------------------------------------------------------------------
 */

#define CHECKER_BOARD_HS_DR_PIO_TEST__M2D_L2S__DEBUG 0

static void
ckrbrd_hs_dr_pio_test__m2d_l2s(struct hs_dr_pio_test_vars_t *tv_ptr)
{
#if CHECKER_BOARD_HS_DR_PIO_TEST__M2D_L2S__DEBUG
    const char *fcnName = "ckrbrd_hs_dr_pio_test__m2d_l2s()";
#endif /* CHECKER_BOARD_HS_DR_PIO_TEST__M2D_L2S__DEBUG */
    bool      data_ok = false;
    int       i, j, k, l;
    size_t    u;
    size_t    start_index;
    size_t    stop_index;
    uint32_t  expected_value;
    uint32_t *ptr_1;
    int       mpi_rank; /* needed by VRFY */
    hsize_t   sel_start[PAR_SS_DR_MAX_RANK];
    htri_t    check; /* Shape comparison return value */
    herr_t    ret;   /* Generic return value */

    /* initialize the local copy of mpi_rank */
    mpi_rank = tv_ptr->mpi_rank;

    /* now we go in the opposite direction, verifying that we can write
     * from memory to file using selections of different rank that
     * H5Sselect_shape_same() views as being of the same shape.
     *
     * Start by writing small_rank - 1 D slices from the in memory large data
     * set to the on disk small dataset.  After each write, read the slice of
     * the small dataset back from disk, and verify that it contains the
     * expected data. Verify that H5Sselect_shape_same() returns true on
     * the memory and file selections.
     */

    tv_ptr->start[0]  = (hsize_t)(tv_ptr->mpi_rank);
    tv_ptr->stride[0] = (hsize_t)(2 * (tv_ptr->mpi_size + 1));
    tv_ptr->count[0]  = 1;
    tv_ptr->block[0]  = 1;

    for (i = 1; i < tv_ptr->large_rank; i++) {

        tv_ptr->start[i]  = 0;
        tv_ptr->stride[i] = (hsize_t)(2 * tv_ptr->edge_size);
        tv_ptr->count[i]  = 1;
        tv_ptr->block[i]  = (hsize_t)(tv_ptr->edge_size);
    }

    ret = H5Sselect_hyperslab(tv_ptr->file_small_ds_sid_0, H5S_SELECT_SET, tv_ptr->start, tv_ptr->stride,
                              tv_ptr->count, tv_ptr->block);
    VRFY((ret >= 0), "H5Sselect_hyperslab(file_small_ds_sid_0, set) succeeded");

    ret = H5Sselect_hyperslab(tv_ptr->mem_small_ds_sid, H5S_SELECT_SET, tv_ptr->start, tv_ptr->stride,
                              tv_ptr->count, tv_ptr->block);
    VRFY((ret >= 0), "H5Sselect_hyperslab(mem_small_ds_sid, set) succeeded");

    sel_start[0] = sel_start[1] = sel_start[2] = sel_start[3] = sel_start[4] = 0;
    sel_start[tv_ptr->small_ds_offset]                                       = (hsize_t)(tv_ptr->mpi_rank);

    ckrbrd_hs_dr_pio_test__slct_ckrbrd(tv_ptr->mpi_rank, tv_ptr->file_small_ds_sid_1, tv_ptr->small_rank,
                                       tv_ptr->edge_size, tv_ptr->checker_edge_size, tv_ptr->small_rank - 1,
                                       sel_start);

    /* set up start, stride, count, and block -- note that we will
     * change start[] so as to read slices of the large cube.
     */
    for (i = 0; i < PAR_SS_DR_MAX_RANK; i++) {

        tv_ptr->start[i]  = 0;
        tv_ptr->stride[i] = (hsize_t)(2 * tv_ptr->edge_size);
        tv_ptr->count[i]  = 1;
        if ((PAR_SS_DR_MAX_RANK - i) > (tv_ptr->small_rank - 1)) {

            tv_ptr->block[i] = 1;
        }
        else {

            tv_ptr->block[i] = (hsize_t)(tv_ptr->edge_size);
        }
    }

    /* zero out the in memory small ds */
    memset(tv_ptr->small_ds_buf_1, 0, sizeof(uint32_t) * tv_ptr->small_ds_size);

#if CHECKER_BOARD_HS_DR_PIO_TEST__M2D_L2S__DEBUG
    fprintf(stdout,
            "%s writing checker boards selections of slices from big ds to slices of small ds on disk.\n",
            fcnName);
#endif /* CHECKER_BOARD_HS_DR_PIO_TEST__M2D_L2S__DEBUG */

    /* in serial versions of this test, we loop through all the dimensions
     * of the large data set that don't appear in the small data set.
     *
     * However, in the parallel version, each process only works with that
     * slice of the large (and small) data set indicated by its rank -- hence
     * we set the most slowly changing index to mpi_rank, and don't iterate
     * over it.
     */

    if (PAR_SS_DR_MAX_RANK - tv_ptr->large_rank == 0) {

        i = tv_ptr->mpi_rank;
    }
    else {

        i = 0;
    }

    /* since large_rank is at most PAR_SS_DR_MAX_RANK, no need to
     * loop over it -- either we are setting i to mpi_rank, or
     * we are setting it to zero.  It will not change during the
     * test.
     */

    if (PAR_SS_DR_MAX_RANK - tv_ptr->large_rank == 1) {

        j = tv_ptr->mpi_rank;
    }
    else {

        j = 0;
    }

    j = 0;
    do {
        if (PAR_SS_DR_MAX_RANK - tv_ptr->large_rank == 2) {

            k = tv_ptr->mpi_rank;
        }
        else {

            k = 0;
        }

        do {
            /* since small rank >= 2 and large_rank > small_rank, we
             * have large_rank >= 3.  Since PAR_SS_DR_MAX_RANK == 5
             * (baring major re-orgaization), this gives us:
             *
             *     (PAR_SS_DR_MAX_RANK - large_rank) <= 2
             *
             * so no need to repeat the test in the outer loops --
             * just set l = 0.
             */

            l = 0;
            do {
                if ((tv_ptr->skips)++ < tv_ptr->max_skips) { /* skip the test */

                    (tv_ptr->tests_skipped)++;
                }
                else { /* run the test */

                    tv_ptr->skips = 0; /* reset the skips counter */

                    /* we know that small_rank >= 1 and that large_rank > small_rank
                     * by the assertions at the head of this function.  Thus no
                     * need for another inner loop.
                     */

                    /* zero out this rank's slice of the on disk small data set */
                    ret = H5Dwrite(tv_ptr->small_dataset, H5T_NATIVE_UINT32, tv_ptr->mem_small_ds_sid,
                                   tv_ptr->file_small_ds_sid_0, tv_ptr->xfer_plist, tv_ptr->small_ds_buf_2);
                    VRFY((ret >= 0), "H5Dwrite() zero slice to small ds succeeded.");

                    /* select the portion of the in memory large cube from which we
                     * are going to write data.
                     */
                    tv_ptr->start[0] = (hsize_t)i;
                    tv_ptr->start[1] = (hsize_t)j;
                    tv_ptr->start[2] = (hsize_t)k;
                    tv_ptr->start[3] = (hsize_t)l;
                    tv_ptr->start[4] = 0;

                    assert((tv_ptr->start[0] == 0) || (0 < tv_ptr->small_ds_offset + 1));
                    assert((tv_ptr->start[1] == 0) || (1 < tv_ptr->small_ds_offset + 1));
                    assert((tv_ptr->start[2] == 0) || (2 < tv_ptr->small_ds_offset + 1));
                    assert((tv_ptr->start[3] == 0) || (3 < tv_ptr->small_ds_offset + 1));
                    assert((tv_ptr->start[4] == 0) || (4 < tv_ptr->small_ds_offset + 1));

                    ckrbrd_hs_dr_pio_test__slct_ckrbrd(
                        tv_ptr->mpi_rank, tv_ptr->mem_large_ds_sid, tv_ptr->large_rank, tv_ptr->edge_size,
                        tv_ptr->checker_edge_size, tv_ptr->small_rank - 1, tv_ptr->start);

                    /* verify that H5Sselect_shape_same() reports the in
                     * memory checkerboard selection of the slice through the
                     * large dataset and the checkerboard selection of the process
                     * slice of the small data set as having the same shape.
                     */
                    check = H5Sselect_shape_same(tv_ptr->file_small_ds_sid_1, tv_ptr->mem_large_ds_sid);
                    VRFY((check == true), "H5Sselect_shape_same passed.");

                    /* write the checker board selection of the slice from the in
                     * memory large data set to the slice of the on disk small
                     * dataset.
                     */
#if CHECKER_BOARD_HS_DR_PIO_TEST__M2D_L2S__DEBUG
                    fprintf(stdout, "%s:%d: start = %d %d %d %d %d.\n", fcnName, tv_ptr->mpi_rank,
                            tv_ptr->start[0], tv_ptr->start[1], tv_ptr->start[2], tv_ptr->start[3],
                            tv_ptr->start[4]);
                    fprintf(stdout, "%s:%d: mem/file extent dims = %d/%d.\n", fcnName, tv_ptr->mpi_rank,
                            H5Sget_simple_extent_ndims(tv_ptr->mem_large_ds_sid),
                            H5Sget_simple_extent_ndims(tv_ptr->file_small_ds_sid_1));
#endif /* CHECKER_BOARD_HS_DR_PIO_TEST__M2D_L2S__DEBUG */
                    ret = H5Dwrite(tv_ptr->small_dataset, H5T_NATIVE_UINT32, tv_ptr->mem_large_ds_sid,
                                   tv_ptr->file_small_ds_sid_1, tv_ptr->xfer_plist, tv_ptr->large_ds_buf_0);
                    VRFY((ret >= 0), "H5Dwrite() slice to large ds succeeded.");

                    /* read the on disk process slice of the small dataset into memory */
                    ret = H5Dread(tv_ptr->small_dataset, H5T_NATIVE_UINT32, tv_ptr->mem_small_ds_sid,
                                  tv_ptr->file_small_ds_sid_0, tv_ptr->xfer_plist, tv_ptr->small_ds_buf_1);
                    VRFY((ret >= 0), "H5Dread() slice from small ds succeeded.");

                    /* verify that expected data is retrieved */

                    expected_value =
                        (uint32_t)((i * tv_ptr->edge_size * tv_ptr->edge_size * tv_ptr->edge_size *
                                    tv_ptr->edge_size) +
                                   (j * tv_ptr->edge_size * tv_ptr->edge_size * tv_ptr->edge_size) +
                                   (k * tv_ptr->edge_size * tv_ptr->edge_size) + (l * tv_ptr->edge_size));

                    start_index = (size_t)(tv_ptr->mpi_rank) * tv_ptr->small_ds_slice_size;
                    stop_index  = start_index + tv_ptr->small_ds_slice_size - 1;

                    assert(start_index < stop_index);
                    assert(stop_index <= tv_ptr->small_ds_size);

                    data_ok = true;

                    ptr_1 = tv_ptr->small_ds_buf_1;
                    for (u = 0; u < start_index; u++, ptr_1++) {

                        if (*ptr_1 != 0) {

                            data_ok = false;
                            *ptr_1  = 0;
                        }
                    }

                    data_ok &= ckrbrd_hs_dr_pio_test__verify_data(
                        tv_ptr->small_ds_buf_1 + start_index, tv_ptr->small_rank - 1, tv_ptr->edge_size,
                        tv_ptr->checker_edge_size, expected_value, (bool)true);

                    ptr_1 = tv_ptr->small_ds_buf_1;
                    for (u = stop_index; u < tv_ptr->small_ds_size; u++, ptr_1++) {

                        if (*ptr_1 != 0) {

                            data_ok = false;
                            *ptr_1  = 0;
                        }
                    }

                    VRFY((data_ok == true), "large slice write slice to small slice data good.");

                    (tv_ptr->tests_run)++;
                }

                l++;

                (tv_ptr->total_tests)++;

            } while ((tv_ptr->large_rank > 2) && ((tv_ptr->small_rank - 1) <= 1) && (l < tv_ptr->edge_size));
            k++;
        } while ((tv_ptr->large_rank > 3) && ((tv_ptr->small_rank - 1) <= 2) && (k < tv_ptr->edge_size));
        j++;
    } while ((tv_ptr->large_rank > 4) && ((tv_ptr->small_rank - 1) <= 3) && (j < tv_ptr->edge_size));

    return;

} /* ckrbrd_hs_dr_pio_test__m2d_l2s() */

/*-------------------------------------------------------------------------
 * Function:    ckrbrd_hs_dr_pio_test__m2d_s2l()
 *
 * Purpose:    Part four of a series of tests of I/O to/from checker
 *        board hyperslab selections of different rank in the parallel.
 *
 *        Verify that we can write from memory to file using
 *        selections of different rank that H5Sselect_shape_same()
 *        views as being of the same shape.
 *
 *        Do this by writing checker board selections of the contents
 *        of the process's slice of the in memory small data set to
 *        slices of the on disk large data set.  After each write,
 *        read the process's slice of the large data set back into
 *        memory, and verify that it contains the expected data.
 *
 *        Verify that H5Sselect_shape_same() returns true on the
 *        memory and file selections.
 *
 * Return:    void
 *
 *-------------------------------------------------------------------------
 */

#define CHECKER_BOARD_HS_DR_PIO_TEST__M2D_S2L__DEBUG 0

static void
ckrbrd_hs_dr_pio_test__m2d_s2l(struct hs_dr_pio_test_vars_t *tv_ptr)
{
#if CHECKER_BOARD_HS_DR_PIO_TEST__M2D_S2L__DEBUG
    const char *fcnName = "ckrbrd_hs_dr_pio_test__m2d_s2l()";
#endif /* CONTIG_HS_DR_PIO_TEST__M2D_S2L__DEBUG */
    bool      data_ok = false;
    int       i, j, k, l;
    size_t    u;
    size_t    start_index;
    size_t    stop_index;
    uint32_t  expected_value;
    uint32_t *ptr_1;
    int       mpi_rank; /* needed by VRFY */
    hsize_t   sel_start[PAR_SS_DR_MAX_RANK];
    htri_t    check; /* Shape comparison return value */
    herr_t    ret;   /* Generic return value */

    /* initialize the local copy of mpi_rank */
    mpi_rank = tv_ptr->mpi_rank;

    /* Now write the contents of the process's slice of the in memory
     * small data set to slices of the on disk large data set.  After
     * each write, read the process's slice of the large data set back
     * into memory, and verify that it contains the expected data.
     * Verify that H5Sselect_shape_same() returns true on the memory
     * and file selections.
     */

    tv_ptr->start[0]  = (hsize_t)(tv_ptr->mpi_rank);
    tv_ptr->stride[0] = (hsize_t)(2 * (tv_ptr->mpi_size + 1));
    tv_ptr->count[0]  = 1;
    tv_ptr->block[0]  = 1;

    for (i = 1; i < tv_ptr->large_rank; i++) {

        tv_ptr->start[i]  = 0;
        tv_ptr->stride[i] = (hsize_t)(2 * tv_ptr->edge_size);
        tv_ptr->count[i]  = 1;
        tv_ptr->block[i]  = (hsize_t)(tv_ptr->edge_size);
    }

    ret = H5Sselect_hyperslab(tv_ptr->file_large_ds_sid_0, H5S_SELECT_SET, tv_ptr->start, tv_ptr->stride,
                              tv_ptr->count, tv_ptr->block);
    VRFY((ret >= 0), "H5Sselect_hyperslab(file_large_ds_sid_0, set) succeeded");

    ret = H5Sselect_hyperslab(tv_ptr->mem_large_ds_sid, H5S_SELECT_SET, tv_ptr->start, tv_ptr->stride,
                              tv_ptr->count, tv_ptr->block);
    VRFY((ret >= 0), "H5Sselect_hyperslab(tv_ptr->mem_large_ds_sid, set) succeeded");

    /* setup a checkerboard selection of the slice of the in memory small
     * data set associated with the process's mpi rank.
     */

    sel_start[0] = sel_start[1] = sel_start[2] = sel_start[3] = sel_start[4] = 0;
    sel_start[tv_ptr->small_ds_offset]                                       = (hsize_t)(tv_ptr->mpi_rank);

    ckrbrd_hs_dr_pio_test__slct_ckrbrd(tv_ptr->mpi_rank, tv_ptr->mem_small_ds_sid, tv_ptr->small_rank,
                                       tv_ptr->edge_size, tv_ptr->checker_edge_size, tv_ptr->small_rank - 1,
                                       sel_start);

    /* set up start, stride, count, and block -- note that we will
     * change start[] so as to write checkerboard selections of slices
     * of the small data set to slices of the large data set.
     */
    for (i = 0; i < PAR_SS_DR_MAX_RANK; i++) {

        tv_ptr->start[i]  = 0;
        tv_ptr->stride[i] = (hsize_t)(2 * tv_ptr->edge_size);
        tv_ptr->count[i]  = 1;
        if ((PAR_SS_DR_MAX_RANK - i) > (tv_ptr->small_rank - 1)) {

            tv_ptr->block[i] = 1;
        }
        else {

            tv_ptr->block[i] = (hsize_t)(tv_ptr->edge_size);
        }
    }

    /* zero out the in memory large ds */
    memset(tv_ptr->large_ds_buf_1, 0, sizeof(uint32_t) * tv_ptr->large_ds_size);

#if CHECKER_BOARD_HS_DR_PIO_TEST__M2D_S2L__DEBUG
    fprintf(stdout,
            "%s writing process checkerboard selections of slices of small ds to process slices of large "
            "ds on disk.\n",
            fcnName);
#endif /* CHECKER_BOARD_HS_DR_PIO_TEST__M2D_S2L__DEBUG */

    if (PAR_SS_DR_MAX_RANK - tv_ptr->large_rank == 0) {

        i = tv_ptr->mpi_rank;
    }
    else {

        i = 0;
    }

    /* since large_rank is at most PAR_SS_DR_MAX_RANK, no need to
     * loop over it -- either we are setting i to mpi_rank, or
     * we are setting it to zero.  It will not change during the
     * test.
     */

    if (PAR_SS_DR_MAX_RANK - tv_ptr->large_rank == 1) {

        j = tv_ptr->mpi_rank;
    }
    else {

        j = 0;
    }

    do {
        if (PAR_SS_DR_MAX_RANK - tv_ptr->large_rank == 2) {

            k = tv_ptr->mpi_rank;
        }
        else {

            k = 0;
        }

        do {
            /* since small rank >= 2 and large_rank > small_rank, we
             * have large_rank >= 3.  Since PAR_SS_DR_MAX_RANK == 5
             * (baring major re-orgaization), this gives us:
             *
             *     (PAR_SS_DR_MAX_RANK - large_rank) <= 2
             *
             * so no need to repeat the test in the outer loops --
             * just set l = 0.
             */

            l = 0;
            do {
                if ((tv_ptr->skips)++ < tv_ptr->max_skips) { /* skip the test */

                    (tv_ptr->tests_skipped)++;
                }
                else { /* run the test */

                    tv_ptr->skips = 0; /* reset the skips counter */

                    /* we know that small_rank >= 1 and that large_rank > small_rank
                     * by the assertions at the head of this function.  Thus no
                     * need for another inner loop.
                     */

                    /* Zero out this processes slice of the on disk large data set.
                     * Note that this will leave one slice with its original data
                     * as there is one more slice than processes.
                     */
                    ret = H5Dwrite(tv_ptr->large_dataset, H5T_NATIVE_UINT32, tv_ptr->mem_large_ds_sid,
                                   tv_ptr->file_large_ds_sid_0, tv_ptr->xfer_plist, tv_ptr->large_ds_buf_2);
                    VRFY((ret != FAIL), "H5Dwrite() to zero large ds succeeded");

                    /* select the portion of the in memory large cube to which we
                     * are going to write data.
                     */
                    tv_ptr->start[0] = (hsize_t)i;
                    tv_ptr->start[1] = (hsize_t)j;
                    tv_ptr->start[2] = (hsize_t)k;
                    tv_ptr->start[3] = (hsize_t)l;
                    tv_ptr->start[4] = 0;

                    assert((tv_ptr->start[0] == 0) || (0 < tv_ptr->small_ds_offset + 1));
                    assert((tv_ptr->start[1] == 0) || (1 < tv_ptr->small_ds_offset + 1));
                    assert((tv_ptr->start[2] == 0) || (2 < tv_ptr->small_ds_offset + 1));
                    assert((tv_ptr->start[3] == 0) || (3 < tv_ptr->small_ds_offset + 1));
                    assert((tv_ptr->start[4] == 0) || (4 < tv_ptr->small_ds_offset + 1));

                    ckrbrd_hs_dr_pio_test__slct_ckrbrd(
                        tv_ptr->mpi_rank, tv_ptr->file_large_ds_sid_1, tv_ptr->large_rank, tv_ptr->edge_size,
                        tv_ptr->checker_edge_size, tv_ptr->small_rank - 1, tv_ptr->start);

                    /* verify that H5Sselect_shape_same() reports the in
                     * memory small data set slice selection and the
                     * on disk slice through the large data set selection
                     * as having the same shape.
                     */
                    check = H5Sselect_shape_same(tv_ptr->mem_small_ds_sid, tv_ptr->file_large_ds_sid_1);
                    VRFY((check == true), "H5Sselect_shape_same passed");

                    /* write the small data set slice from memory to the
                     * target slice of the disk data set
                     */
#if CHECKER_BOARD_HS_DR_PIO_TEST__M2D_S2L__DEBUG
                    fprintf(stdout, "%s:%d: start = %d %d %d %d %d.\n", fcnName, tv_ptr->mpi_rank,
                            tv_ptr->start[0], tv_ptr->start[1], tv_ptr->start[2], tv_ptr->start[3],
                            tv_ptr->start[4]);
                    fprintf(stdout, "%s:%d: mem/file extent dims = %d/%d.\n", fcnName, tv_ptr->mpi_rank,
                            H5Sget_simple_extent_ndims(tv_ptr->mem_small_ds_sid),
                            H5Sget_simple_extent_ndims(tv_ptr->file_large_ds_sid_1));
#endif /* CHECKER_BOARD_HS_DR_PIO_TEST__M2D_S2L__DEBUG */
                    ret = H5Dwrite(tv_ptr->large_dataset, H5T_NATIVE_UINT32, tv_ptr->mem_small_ds_sid,
                                   tv_ptr->file_large_ds_sid_1, tv_ptr->xfer_plist, tv_ptr->small_ds_buf_0);
                    VRFY((ret != FAIL), "H5Dwrite of small ds slice to large ds succeeded");

                    /* read this processes slice on the on disk large
                     * data set into memory.
                     */

                    ret = H5Dread(tv_ptr->large_dataset, H5T_NATIVE_UINT32, tv_ptr->mem_large_ds_sid,
                                  tv_ptr->file_large_ds_sid_0, tv_ptr->xfer_plist, tv_ptr->large_ds_buf_1);
                    VRFY((ret != FAIL), "H5Dread() of process slice of large ds succeeded");

                    /* verify that the expected data and only the
                     * expected data was read.
                     */
                    expected_value = (uint32_t)((size_t)(tv_ptr->mpi_rank) * tv_ptr->small_ds_slice_size);

                    start_index =
                        (size_t)((i * tv_ptr->edge_size * tv_ptr->edge_size * tv_ptr->edge_size *
                                  tv_ptr->edge_size) +
                                 (j * tv_ptr->edge_size * tv_ptr->edge_size * tv_ptr->edge_size) +
                                 (k * tv_ptr->edge_size * tv_ptr->edge_size) + (l * tv_ptr->edge_size));
                    stop_index = start_index + tv_ptr->small_ds_slice_size - 1;

                    assert(start_index < stop_index);
                    assert(stop_index < tv_ptr->large_ds_size);

                    data_ok = true;

                    ptr_1 = tv_ptr->large_ds_buf_1;
                    for (u = 0; u < start_index; u++, ptr_1++) {

                        if (*ptr_1 != 0) {

                            data_ok = false;
                            *ptr_1  = 0;
                        }
                    }

                    data_ok &= ckrbrd_hs_dr_pio_test__verify_data(
                        tv_ptr->large_ds_buf_1 + start_index, tv_ptr->small_rank - 1, tv_ptr->edge_size,
                        tv_ptr->checker_edge_size, expected_value, (bool)true);

                    ptr_1 = tv_ptr->large_ds_buf_1;
                    for (u = stop_index; u < tv_ptr->small_ds_size; u++, ptr_1++) {

                        if (*ptr_1 != 0) {

                            data_ok = false;
                            *ptr_1  = 0;
                        }
                    }

                    VRFY((data_ok == true), "small ds cb slice write to large ds slice data good.");

                    (tv_ptr->tests_run)++;
                }

                l++;

                (tv_ptr->total_tests)++;

            } while ((tv_ptr->large_rank > 2) && ((tv_ptr->small_rank - 1) <= 1) && (l < tv_ptr->edge_size));
            k++;
        } while ((tv_ptr->large_rank > 3) && ((tv_ptr->small_rank - 1) <= 2) && (k < tv_ptr->edge_size));
        j++;
    } while ((tv_ptr->large_rank > 4) && ((tv_ptr->small_rank - 1) <= 3) && (j < tv_ptr->edge_size));

    return;

} /* ckrbrd_hs_dr_pio_test__m2d_s2l() */

/*-------------------------------------------------------------------------
 * Function:    ckrbrd_hs_dr_pio_test__run_test()
 *
 * Purpose:    Test I/O to/from checkerboard selections of hyperslabs of
 *        different rank in the parallel.
 *
 * Return:    void
 *
 *-------------------------------------------------------------------------
 */

#define CKRBRD_HS_DR_PIO_TEST__RUN_TEST__DEBUG 0

static void
ckrbrd_hs_dr_pio_test__run_test(const int test_num, const int edge_size, const int checker_edge_size,
                                const int chunk_edge_size, const int small_rank, const int large_rank,
                                const bool use_collective_io, const hid_t dset_type, const int express_test,
                                int *skips_ptr, int max_skips, int64_t *total_tests_ptr,
                                int64_t *tests_run_ptr, int64_t *tests_skipped_ptr)

{
#if CKRBRD_HS_DR_PIO_TEST__RUN_TEST__DEBUG
    const char *fcnName = "ckrbrd_hs_dr_pio_test__run_test()";
#endif /* CKRBRD_HS_DR_PIO_TEST__RUN_TEST__DEBUG */
    struct hs_dr_pio_test_vars_t test_vars = {
        /* int           mpi_size                        = */ -1,
        /* int         mpi_rank                        = */ -1,
        /* MPI_Comm    mpi_comm                        = */ MPI_COMM_NULL,
        /* MPI_Inf     mpi_info                        = */ MPI_INFO_NULL,
        /* int         test_num                        = */ -1,
        /* int         edge_size                       = */ -1,
        /* int         checker_edge_size               = */ -1,
        /* int         chunk_edge_size                 = */ -1,
        /* int         small_rank                      = */ -1,
        /* int         large_rank                      = */ -1,
        /* hid_t       dset_type                       = */ H5I_INVALID_HID,
        /* uint32_t  * small_ds_buf_0                  = */ NULL,
        /* uint32_t  * small_ds_buf_1                  = */ NULL,
        /* uint32_t  * small_ds_buf_2                  = */ NULL,
        /* uint32_t  * small_ds_slice_buf              = */ NULL,
        /* uint32_t  * large_ds_buf_0                  = */ NULL,
        /* uint32_t  * large_ds_buf_1                  = */ NULL,
        /* uint32_t  * large_ds_buf_2                  = */ NULL,
        /* uint32_t  * large_ds_slice_buf              = */ NULL,
        /* int         small_ds_offset                 = */ -1,
        /* int         large_ds_offset                 = */ -1,
        /* hid_t       fid                             = */ H5I_INVALID_HID, /* HDF5 file ID */
        /* hid_t       xfer_plist                      = */ H5P_DEFAULT,
        /* hid_t       full_mem_small_ds_sid           = */ H5I_INVALID_HID,
        /* hid_t       full_file_small_ds_sid          = */ H5I_INVALID_HID,
        /* hid_t       mem_small_ds_sid                = */ H5I_INVALID_HID,
        /* hid_t       file_small_ds_sid_0             = */ H5I_INVALID_HID,
        /* hid_t       file_small_ds_sid_1             = */ H5I_INVALID_HID,
        /* hid_t       small_ds_slice_sid              = */ H5I_INVALID_HID,
        /* hid_t       full_mem_large_ds_sid           = */ H5I_INVALID_HID,
        /* hid_t       full_file_large_ds_sid          = */ H5I_INVALID_HID,
        /* hid_t       mem_large_ds_sid                = */ H5I_INVALID_HID,
        /* hid_t       file_large_ds_sid_0             = */ H5I_INVALID_HID,
        /* hid_t       file_large_ds_sid_1             = */ H5I_INVALID_HID,
        /* hid_t       file_large_ds_process_slice_sid = */ H5I_INVALID_HID,
        /* hid_t       mem_large_ds_process_slice_sid  = */ H5I_INVALID_HID,
        /* hid_t       large_ds_slice_sid              = */ H5I_INVALID_HID,
        /* hid_t       small_dataset                   = */ H5I_INVALID_HID, /* Dataset ID */
        /* hid_t       large_dataset                   = */ H5I_INVALID_HID, /* Dataset ID */
        /* size_t      small_ds_size                   = */ 1,
        /* size_t      small_ds_slice_size             = */ 1,
        /* size_t      large_ds_size                   = */ 1,
        /* size_t      large_ds_slice_size             = */ 1,
        /* hsize_t     dims[PAR_SS_DR_MAX_RANK]        = */ {0, 0, 0, 0, 0},
        /* hsize_t     chunk_dims[PAR_SS_DR_MAX_RANK]  = */ {0, 0, 0, 0, 0},
        /* hsize_t     start[PAR_SS_DR_MAX_RANK]       = */ {0, 0, 0, 0, 0},
        /* hsize_t     stride[PAR_SS_DR_MAX_RANK]      = */ {0, 0, 0, 0, 0},
        /* hsize_t     count[PAR_SS_DR_MAX_RANK]       = */ {0, 0, 0, 0, 0},
        /* hsize_t     block[PAR_SS_DR_MAX_RANK]       = */ {0, 0, 0, 0, 0},
        /* hsize_t   * start_ptr                       = */ NULL,
        /* hsize_t   * stride_ptr                      = */ NULL,
        /* hsize_t   * count_ptr                       = */ NULL,
        /* hsize_t   * block_ptr                       = */ NULL,
        /* int            skips                           = */ 0,
        /* int            max_skips                       = */ 0,
        /* int64_t     total_tests                     = */ 0,
        /* int64_t     tests_run                       = */ 0,
        /* int64_t     tests_skipped                   = */ 0};
    struct hs_dr_pio_test_vars_t *tv_ptr = &test_vars;

    hs_dr_pio_test__setup(test_num, edge_size, checker_edge_size, chunk_edge_size, small_rank, large_rank,
                          use_collective_io, dset_type, express_test, tv_ptr);

    /* initialize skips & max_skips */
    tv_ptr->skips     = *skips_ptr;
    tv_ptr->max_skips = max_skips;

#if CKRBRD_HS_DR_PIO_TEST__RUN_TEST__DEBUG
    if (MAINPROCESS) {
        fprintf(stdout, "test %d: small rank = %d, large rank = %d.\n", test_num, small_rank, large_rank);
        fprintf(stdout, "test %d: Initialization complete.\n", test_num);
    }
#endif /* CKRBRD_HS_DR_PIO_TEST__RUN_TEST__DEBUG */

    /* first, verify that we can read from disk correctly using selections
     * of different rank that H5Sselect_shape_same() views as being of the
     * same shape.
     *
     * Start by reading a (small_rank - 1)-D slice from this processes slice
     * of the on disk large data set, and verifying that the data read is
     * correct.  Verify that H5Sselect_shape_same() returns true on the
     * memory and file selections.
     *
     * The first step is to set up the needed checker board selection in the
     * in memory small small cube
     */

    ckrbrd_hs_dr_pio_test__d2m_l2s(tv_ptr);

    /* similarly, read slices of the on disk small data set into slices
     * through the in memory large data set, and verify that the correct
     * data (and only the correct data) is read.
     */

    ckrbrd_hs_dr_pio_test__d2m_s2l(tv_ptr);

    /* now we go in the opposite direction, verifying that we can write
     * from memory to file using selections of different rank that
     * H5Sselect_shape_same() views as being of the same shape.
     *
     * Start by writing small_rank - 1 D slices from the in memory large data
     * set to the on disk small dataset.  After each write, read the slice of
     * the small dataset back from disk, and verify that it contains the
     * expected data. Verify that H5Sselect_shape_same() returns true on
     * the memory and file selections.
     */

    ckrbrd_hs_dr_pio_test__m2d_l2s(tv_ptr);

    /* Now write the contents of the process's slice of the in memory
     * small data set to slices of the on disk large data set.  After
     * each write, read the process's slice of the large data set back
     * into memory, and verify that it contains the expected data.
     * Verify that H5Sselect_shape_same() returns true on the memory
     * and file selections.
     */

    ckrbrd_hs_dr_pio_test__m2d_s2l(tv_ptr);

#if CKRBRD_HS_DR_PIO_TEST__RUN_TEST__DEBUG
    if (MAINPROCESS) {
        fprintf(stdout, "test %d: Subtests complete -- tests run/skipped/total = %lld/%lld/%lld.\n", test_num,
                (long long)(tv_ptr->tests_run), (long long)(tv_ptr->tests_skipped),
                (long long)(tv_ptr->total_tests));
    }
#endif /* CKRBRD_HS_DR_PIO_TEST__RUN_TEST__DEBUG */

    hs_dr_pio_test__takedown(tv_ptr);

#if CKRBRD_HS_DR_PIO_TEST__RUN_TEST__DEBUG
    if (MAINPROCESS) {
        fprintf(stdout, "test %d: Takedown complete.\n", test_num);
    }
#endif /* CKRBRD_HS_DR_PIO_TEST__RUN_TEST__DEBUG */

    *skips_ptr = tv_ptr->skips;
    *total_tests_ptr += tv_ptr->total_tests;
    *tests_run_ptr += tv_ptr->tests_run;
    *tests_skipped_ptr += tv_ptr->tests_skipped;

    return;

} /* ckrbrd_hs_dr_pio_test__run_test() */

/*-------------------------------------------------------------------------
 * Function:    ckrbrd_hs_dr_pio_test()
 *
 * Purpose:    Test I/O to/from hyperslab selections of different rank in
 *        the parallel case.
 *
 * Return:    void
 *
 *-------------------------------------------------------------------------
 */

static void
ckrbrd_hs_dr_pio_test(ShapeSameTestMethods sstest_type)
{
    int   express_test;
    int   local_express_test;
    int   mpi_size = -1;
    int   mpi_rank = -1;
    int   test_num = 0;
    int   edge_size;
    int   checker_edge_size = 3;
    int   chunk_edge_size   = 0;
    int   small_rank        = 3;
    int   large_rank        = 4;
    int   mpi_result;
    hid_t dset_type = H5T_NATIVE_UINT;
    int   skips     = 0;
    int   max_skips = 0;
    /* The following table list the number of sub-tests skipped between
     * each test that is actually executed as a function of the express
     * test level.  Note that any value in excess of 4880 will cause all
     * sub tests to be skipped.
     */
    int     max_skips_tbl[4] = {0, 4, 64, 1024};
    int64_t total_tests      = 0;
    int64_t tests_run        = 0;
    int64_t tests_skipped    = 0;

    MPI_Comm_size(MPI_COMM_WORLD, &mpi_size);
    MPI_Comm_rank(MPI_COMM_WORLD, &mpi_rank);

    edge_size = (mpi_size > 6 ? mpi_size : 6);

    local_express_test = GetTestExpress();

    HDcompile_assert(sizeof(uint32_t) == sizeof(unsigned));

    mpi_result = MPI_Allreduce((void *)&local_express_test, (void *)&express_test, 1, MPI_INT, MPI_MAX,
                               MPI_COMM_WORLD);

    VRFY((mpi_result == MPI_SUCCESS), "MPI_Allreduce(0) succeeded");

    if (local_express_test < 0) {
        max_skips = max_skips_tbl[0];
    }
    else if (local_express_test > 3) {
        max_skips = max_skips_tbl[3];
    }
    else {
        max_skips = max_skips_tbl[local_express_test];
    }

#if 0
    {
        int DebugWait = 1;

        while (DebugWait) ;
    }
#endif

    for (large_rank = 3; large_rank <= PAR_SS_DR_MAX_RANK; large_rank++) {

        for (small_rank = 2; small_rank < large_rank; small_rank++) {
            switch (sstest_type) {
                case IND_CONTIG:
                    /* contiguous data set, independent I/O */
                    chunk_edge_size = 0;
                    ckrbrd_hs_dr_pio_test__run_test(test_num, edge_size, checker_edge_size, chunk_edge_size,
                                                    small_rank, large_rank, false, dset_type, express_test,
                                                    &skips, max_skips, &total_tests, &tests_run,
                                                    &tests_skipped);
                    test_num++;
                    break;
                    /* end of case IND_CONTIG */

                case COL_CONTIG:
                    /* contiguous data set, collective I/O */
                    chunk_edge_size = 0;
                    ckrbrd_hs_dr_pio_test__run_test(
                        test_num, edge_size, checker_edge_size, chunk_edge_size, small_rank, large_rank, true,
                        dset_type, express_test, &skips, max_skips, &total_tests, &tests_run, &tests_skipped);
                    test_num++;
                    break;
                    /* end of case COL_CONTIG */

                case IND_CHUNKED:
                    /* chunked data set, independent I/O */
                    chunk_edge_size = 5;
                    ckrbrd_hs_dr_pio_test__run_test(test_num, edge_size, checker_edge_size, chunk_edge_size,
                                                    small_rank, large_rank, false, dset_type, express_test,
                                                    &skips, max_skips, &total_tests, &tests_run,
                                                    &tests_skipped);
                    test_num++;
                    break;
                    /* end of case IND_CHUNKED */

                case COL_CHUNKED:
                    /* chunked data set, collective I/O */
                    chunk_edge_size = 5;
                    ckrbrd_hs_dr_pio_test__run_test(
                        test_num, edge_size, checker_edge_size, chunk_edge_size, small_rank, large_rank, true,
                        dset_type, express_test, &skips, max_skips, &total_tests, &tests_run, &tests_skipped);
                    test_num++;
                    break;
                    /* end of case COL_CHUNKED */

                default:
                    VRFY((false), "unknown test type");
                    break;

            } /* end of switch(sstest_type) */
#if CONTIG_HS_DR_PIO_TEST__DEBUG
            if ((MAINPROCESS) && (tests_skipped > 0)) {
                fprintf(stdout, "     run/skipped/total = %" PRId64 "/%" PRId64 "/%" PRId64 ".\n", tests_run,
                        tests_skipped, total_tests);
            }
#endif /* CONTIG_HS_DR_PIO_TEST__DEBUG */
        }
    }

    if ((MAINPROCESS) && (tests_skipped > 0)) {
        fprintf(stdout, "     %" PRId64 " of %" PRId64 " subtests skipped to expedite testing.\n",
                tests_skipped, total_tests);
    }

    return;

} /* ckrbrd_hs_dr_pio_test() */

/* Main Body. Here for now, may have to move them to a separated file later. */

/*
 * Main driver of the Parallel HDF5 tests
 */

#include "testphdf5.h"

#ifndef PATH_MAX
#define PATH_MAX 512
#endif /* !PATH_MAX */

/* global variables */
int dim0;
int dim1;
int chunkdim0;
int chunkdim1;
int nerrors   = 0;               /* errors count */
int ndatasets = 300;             /* number of datasets to create*/
int ngroups   = 512;             /* number of groups to create in root
                                  * group. */
int facc_type       = FACC_MPIO; /*Test file access type */
int dxfer_coll_type = DXFER_COLLECTIVE_IO;

H5E_auto2_t old_func;        /* previous error handler */
void       *old_client_data; /* previous error handler arg.*/

/* other option flags */

/* FILENAME and filenames must have the same number of names.
 * Use PARATESTFILE in general and use a separated filename only if the file
 * created in one test is accessed by a different test.
 * filenames[0] is reserved as the file name for PARATESTFILE.
 */
#define NFILENAME    2
#define PARATESTFILE filenames[0]
const char *FILENAME[NFILENAME] = {"ShapeSameTest", NULL};
char       *filenames[NFILENAME];
hid_t       fapl; /* file access property list */

#ifdef USE_PAUSE
/* pause the process for a moment to allow debugger to attach if desired. */
/* Will pause more if greenlight file is not present but will eventually */
/* continue. */
#include <sys/types.h>
#include <sys/stat.h>

void
pause_proc(void)
{

    int       pid;
    h5_stat_t statbuf;
    char      greenlight[] = "go";
    int       maxloop      = 10;
    int       loops        = 0;
    int       time_int     = 10;

    /* mpi variables */
    int  mpi_size, mpi_rank;
    int  mpi_namelen;
    char mpi_name[MPI_MAX_PROCESSOR_NAME];

    pid = getpid();
    MPI_Comm_size(MPI_COMM_WORLD, &mpi_size);
    MPI_Comm_rank(MPI_COMM_WORLD, &mpi_rank);
    MPI_Get_processor_name(mpi_name, &mpi_namelen);

    if (MAINPROCESS)
        while ((HDstat(greenlight, &statbuf) == -1) && loops < maxloop) {
            if (!loops++) {
                printf("Proc %d (%*s, %d): to debug, attach %d\n", mpi_rank, mpi_namelen, mpi_name, pid, pid);
            }
            printf("waiting(%ds) for file %s ...\n", time_int, greenlight);
            fflush(stdout);
            HDsleep(time_int);
        }
    MPI_Barrier(MPI_COMM_WORLD);
}

/* Use the Profile feature of MPI to call the pause_proc() */
int
MPI_Init(int *argc, char ***argv)
{
    int ret_code;
    ret_code = PMPI_Init(argc, argv);
    pause_proc();
    return (ret_code);
}
#endif /* USE_PAUSE */

/*
 * Show command usage
 */
static void
usage(void)
{
    printf("    [-r] [-w] [-m<n_datasets>] [-n<n_groups>] "
           "[-o] [-f <prefix>] [-d <dim0> <dim1>]\n");
    printf("\t-m<n_datasets>"
           "\tset number of datasets for the multiple dataset test\n");
    printf("\t-n<n_groups>"
           "\tset number of groups for the multiple group test\n");
    printf("\t-f <prefix>\tfilename prefix\n");
    printf("\t-2\t\tuse Split-file together with MPIO\n");
    printf("\t-d <factor0> <factor1>\tdataset dimensions factors. Defaults (%d,%d)\n", ROW_FACTOR,
           COL_FACTOR);
    printf("\t-c <dim0> <dim1>\tdataset chunk dimensions. Defaults (dim0/10,dim1/10)\n");
    printf("\n");
}

/*
 * parse the command line options
 */
static int
parse_options(int argc, char **argv)
{
    int mpi_size, mpi_rank; /* mpi variables */

    MPI_Comm_size(MPI_COMM_WORLD, &mpi_size);
    MPI_Comm_rank(MPI_COMM_WORLD, &mpi_rank);

    /* setup default chunk-size. Make sure sizes are > 0 */

    chunkdim0 = (dim0 + 9) / 10;
    chunkdim1 = (dim1 + 9) / 10;

    while (--argc) {
        if (**(++argv) != '-') {
            break;
        }
        else {
            switch (*(*argv + 1)) {
                case 'm':
                    ndatasets = atoi((*argv + 1) + 1);
                    if (ndatasets < 0) {
                        nerrors++;
                        return (1);
                    }
                    break;
                case 'n':
                    ngroups = atoi((*argv + 1) + 1);
                    if (ngroups < 0) {
                        nerrors++;
                        return (1);
                    }
                    break;
                case 'f':
                    if (--argc < 1) {
                        nerrors++;
                        return (1);
                    }
                    if (**(++argv) == '-') {
                        nerrors++;
                        return (1);
                    }
                    paraprefix = *argv;
                    break;
                case 'i': /* Collective MPI-IO access with independent IO  */
                    dxfer_coll_type = DXFER_INDEPENDENT_IO;
                    break;
                case '2': /* Use the split-file driver with MPIO access */
                    /* Can use $HDF5_METAPREFIX to define the */
                    /* meta-file-prefix. */
                    facc_type = FACC_MPIO | FACC_SPLIT;
                    break;
                case 'd': /* dimensizes */
                    if (--argc < 2) {
                        nerrors++;
                        return (1);
                    }
                    dim0 = atoi(*(++argv)) * mpi_size;
                    argc--;
                    dim1 = atoi(*(++argv)) * mpi_size;
                    /* set default chunkdim sizes too */
                    chunkdim0 = (dim0 + 9) / 10;
                    chunkdim1 = (dim1 + 9) / 10;
                    break;
                case 'c': /* chunk dimensions */
                    if (--argc < 2) {
                        nerrors++;
                        return (1);
                    }
                    chunkdim0 = atoi(*(++argv));
                    argc--;
                    chunkdim1 = atoi(*(++argv));
                    break;
                case 'h': /* print help message--return with nerrors set */
                    return (1);
                default:
                    printf("Illegal option(%s)\n", *argv);
                    nerrors++;
                    return (1);
            }
        }
    } /*while*/

    /* check validity of dimension and chunk sizes */
    if (dim0 <= 0 || dim1 <= 0) {
        printf("Illegal dim sizes (%d, %d)\n", dim0, dim1);
        nerrors++;
        return (1);
    }
    if (chunkdim0 <= 0 || chunkdim1 <= 0) {
        printf("Illegal chunkdim sizes (%d, %d)\n", chunkdim0, chunkdim1);
        nerrors++;
        return (1);
    }

    /* Make sure datasets can be divided into equal portions by the processes */
    if ((dim0 % mpi_size) || (dim1 % mpi_size)) {
        if (MAINPROCESS)
            printf("dim0(%d) and dim1(%d) must be multiples of processes(%d)\n", dim0, dim1, mpi_size);
        nerrors++;
        return (1);
    }

    /* compose the test filenames */
    {
        int i, n;

        n = sizeof(FILENAME) / sizeof(FILENAME[0]) - 1; /* exclude the NULL */

        for (i = 0; i < n; i++)
            if (h5_fixname(FILENAME[i], fapl, filenames[i], PATH_MAX) == NULL) {
                printf("h5_fixname failed\n");
                nerrors++;
                return (1);
            }
        printf("Test filenames are:\n");
        for (i = 0; i < n; i++)
            printf("    %s\n", filenames[i]);
    }

    return (0);
}

/*
 * Create the appropriate File access property list
 */
hid_t
create_faccess_plist(MPI_Comm comm, MPI_Info info, int l_facc_type)
{
    hid_t  ret_pl = H5I_INVALID_HID;
    herr_t ret;      /* generic return value */
    int    mpi_rank; /* mpi variables */

    /* need the rank for error checking macros */
    MPI_Comm_rank(MPI_COMM_WORLD, &mpi_rank);

    ret_pl = H5Pcreate(H5P_FILE_ACCESS);
    VRFY((ret_pl >= 0), "H5P_FILE_ACCESS");

    if (l_facc_type == FACC_DEFAULT)
        return (ret_pl);

    if (l_facc_type == FACC_MPIO) {
        /* set Parallel access with communicator */
        ret = H5Pset_fapl_mpio(ret_pl, comm, info);
        VRFY((ret >= 0), "");
        ret = H5Pset_all_coll_metadata_ops(ret_pl, true);
        VRFY((ret >= 0), "");
        ret = H5Pset_coll_metadata_write(ret_pl, true);
        VRFY((ret >= 0), "");
        return (ret_pl);
    }

    if (l_facc_type == (FACC_MPIO | FACC_SPLIT)) {
        hid_t mpio_pl;

        mpio_pl = H5Pcreate(H5P_FILE_ACCESS);
        VRFY((mpio_pl >= 0), "");
        /* set Parallel access with communicator */
        ret = H5Pset_fapl_mpio(mpio_pl, comm, info);
        VRFY((ret >= 0), "");

        /* setup file access template */
        ret_pl = H5Pcreate(H5P_FILE_ACCESS);
        VRFY((ret_pl >= 0), "");
        /* set Parallel access with communicator */
        ret = H5Pset_fapl_split(ret_pl, ".meta", mpio_pl, ".raw", mpio_pl);
        VRFY((ret >= 0), "H5Pset_fapl_split succeeded");
        H5Pclose(mpio_pl);
        return (ret_pl);
    }

    /* unknown file access types */
    return (ret_pl);
}

/* Shape Same test using contiguous hyperslab using independent IO on contiguous datasets */
static void
sscontig1(void)
{
    contig_hs_dr_pio_test(IND_CONTIG);
}

/* Shape Same test using contiguous hyperslab using collective IO on contiguous datasets */
static void
sscontig2(void)
{
    contig_hs_dr_pio_test(COL_CONTIG);
}

/* Shape Same test using contiguous hyperslab using independent IO on chunked datasets */
static void
sscontig3(void)
{
    contig_hs_dr_pio_test(IND_CHUNKED);
}

/* Shape Same test using contiguous hyperslab using collective IO on chunked datasets */
static void
sscontig4(void)
{
    contig_hs_dr_pio_test(COL_CHUNKED);
}

/* Shape Same test using checker hyperslab using independent IO on contiguous datasets */
static void
sschecker1(void)
{
    ckrbrd_hs_dr_pio_test(IND_CONTIG);
}

/* Shape Same test using checker hyperslab using collective IO on contiguous datasets */
static void
sschecker2(void)
{
    ckrbrd_hs_dr_pio_test(COL_CONTIG);
}

/* Shape Same test using checker hyperslab using independent IO on chunked datasets */
static void
sschecker3(void)
{
    ckrbrd_hs_dr_pio_test(IND_CHUNKED);
}

/* Shape Same test using checker hyperslab using collective IO on chunked datasets */
static void
sschecker4(void)
{
    ckrbrd_hs_dr_pio_test(COL_CHUNKED);
}

int
main(int argc, char **argv)
{
    int mpi_size, mpi_rank; /* mpi variables */

#ifndef H5_HAVE_WIN32_API
    /* Un-buffer the stdout and stderr */
    HDsetbuf(stderr, NULL);
    HDsetbuf(stdout, NULL);
#endif

    MPI_Init(&argc, &argv);
    MPI_Comm_size(MPI_COMM_WORLD, &mpi_size);
    MPI_Comm_rank(MPI_COMM_WORLD, &mpi_rank);

    mpi_rank_framework_g = mpi_rank;

    dim0 = ROW_FACTOR * mpi_size;
    dim1 = COL_FACTOR * mpi_size;

    if (MAINPROCESS) {
        printf("===================================\n");
        printf("Shape Same Tests Start\n");
        printf("    express_test = %d.\n", GetTestExpress());
        printf("===================================\n");
    }

    /* Attempt to turn off atexit post processing so that in case errors
     * happen during the test and the process is aborted, it will not get
     * hung in the atexit post processing in which it may try to make MPI
     * calls.  By then, MPI calls may not work.
     */
    if (H5dont_atexit() < 0) {
        if (MAINPROCESS)
            printf("%d: Failed to turn off atexit processing. Continue.\n", mpi_rank);
    };
    H5open();
    h5_show_hostname();

    memset(filenames, 0, sizeof(filenames));
    for (int i = 0; i < NFILENAME; i++) {
        if (NULL == (filenames[i] = malloc(PATH_MAX))) {
            printf("couldn't allocate filename array\n");
            MPI_Abort(MPI_COMM_WORLD, -1);
        }
    }

    /* Initialize testing framework */
    TestInit(argv[0], usage, parse_options);

    /* Shape Same tests using contiguous hyperslab */
    AddTest("sscontig1", sscontig1, NULL, "Cntg hslab, ind IO, cntg dsets", PARATESTFILE);
    AddTest("sscontig2", sscontig2, NULL, "Cntg hslab, col IO, cntg dsets", PARATESTFILE);
    AddTest("sscontig3", sscontig3, NULL, "Cntg hslab, ind IO, chnk dsets", PARATESTFILE);
    AddTest("sscontig4", sscontig4, NULL, "Cntg hslab, col IO, chnk dsets", PARATESTFILE);

    /* Shape Same tests using checker board hyperslab */
    AddTest("sschecker1", sschecker1, NULL, "Check hslab, ind IO, cntg dsets", PARATESTFILE);
    AddTest("sschecker2", sschecker2, NULL, "Check hslab, col IO, cntg dsets", PARATESTFILE);
    AddTest("sschecker3", sschecker3, NULL, "Check hslab, ind IO, chnk dsets", PARATESTFILE);
    AddTest("sschecker4", sschecker4, NULL, "Check hslab, col IO, chnk dsets", PARATESTFILE);

    /* Display testing information */
    TestInfo(argv[0]);

    /* setup file access property list */
    fapl = H5Pcreate(H5P_FILE_ACCESS);
    H5Pset_fapl_mpio(fapl, MPI_COMM_WORLD, MPI_INFO_NULL);

    /* Parse command line arguments */
    TestParseCmdLine(argc, argv);

    if (dxfer_coll_type == DXFER_INDEPENDENT_IO && MAINPROCESS) {
        printf("===================================\n"
               "   Using Independent I/O with file set view to replace collective I/O \n"
               "===================================\n");
    }

    /* Perform requested testing */
    PerformTests();

    /* make sure all processes are finished before final report, cleanup
     * and exit.
     */
    MPI_Barrier(MPI_COMM_WORLD);

    /* Display test summary, if requested */
    if (MAINPROCESS && GetTestSummary())
        TestSummary();

    /* Clean up test files */
    h5_clean_files(FILENAME, fapl);

    nerrors += GetTestNumErrs();

    /* Gather errors from all processes */
    {
        int temp;
        MPI_Allreduce(&nerrors, &temp, 1, MPI_INT, MPI_MAX, MPI_COMM_WORLD);
        nerrors = temp;
    }

    if (MAINPROCESS) { /* only process 0 reports */
        printf("===================================\n");
        if (nerrors)
            printf("***Shape Same tests detected %d errors***\n", nerrors);
        else
            printf("Shape Same tests finished with no errors\n");
        printf("===================================\n");
    }

    for (int i = 0; i < NFILENAME; i++) {
        free(filenames[i]);
        filenames[i] = NULL;
    }

    /* close HDF5 library */
    H5close();

    /* Release test infrastructure */
    TestShutdown();

    MPI_Finalize();

    /* cannot just return (nerrors) because exit code is limited to 1byte */
    return (nerrors != 0);
}