File: test_tracking.py

package info (click to toggle)
python-memray 1.17.0%2Bdfsg-1
  • links: PTS, VCS
  • area: main
  • in suites: forky, sid, trixie
  • size: 24,396 kB
  • sloc: python: 28,451; ansic: 16,507; sh: 10,586; cpp: 8,494; javascript: 1,474; makefile: 822; awk: 12
file content (1750 lines) | stat: -rw-r--r-- 56,924 bytes parent folder | download
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
import collections
import datetime
import mmap
import signal
import subprocess
import sys
import textwrap
import threading
import time
from pathlib import Path

import pytest

from memray import AllocatorType
from memray import FileFormat
from memray import FileReader
from memray import Tracker
from memray._memray import compute_statistics
from memray._test import MemoryAllocator
from memray._test import MmapAllocator
from memray._test import PrimeCaches
from memray._test import PymallocDomain
from memray._test import PymallocMemoryAllocator
from memray._test import _cython_allocate_in_two_places
from memray._test import allocate_cpp_vector
from memray._test import fill_cpp_vector
from tests.utils import filter_relevant_allocations
from tests.utils import run_without_tracer

ALLOCATORS = [
    ("malloc", AllocatorType.MALLOC),
    ("valloc", AllocatorType.VALLOC),
    ("pvalloc", AllocatorType.PVALLOC),
    ("calloc", AllocatorType.CALLOC),
    ("memalign", AllocatorType.MEMALIGN),
    ("posix_memalign", AllocatorType.POSIX_MEMALIGN),
    ("aligned_alloc", AllocatorType.ALIGNED_ALLOC),
    ("realloc", AllocatorType.REALLOC),
]

PYMALLOC_ALLOCATORS = [
    ("malloc", AllocatorType.PYMALLOC_MALLOC),
    ("calloc", AllocatorType.PYMALLOC_CALLOC),
    ("realloc", AllocatorType.PYMALLOC_REALLOC),
]

PYMALLOC_DOMAINS = [
    PymallocDomain.PYMALLOC_RAW,
    PymallocDomain.PYMALLOC_MEM,
    PymallocDomain.PYMALLOC_OBJECT,
]


PAGE_SIZE = mmap.PAGESIZE
ALLOC_SIZE = 123 * 8


@pytest.mark.skipif(
    sys.platform == "darwin", reason="Test triggers some extra allocations in macOS"
)
def test_no_allocations_while_tracking(tmp_path):
    output = tmp_path / "test.bin"
    with Tracker(output):
        pass

    records = list(FileReader(output).get_allocation_records())
    assert not records


@pytest.mark.parametrize(["allocator_func", "allocator_type"], ALLOCATORS)
def test_simple_allocation_tracking(allocator_func, allocator_type, tmp_path):
    # GIVEN
    allocator = MemoryAllocator()
    output = tmp_path / "test.bin"

    # WHEN
    with Tracker(output):
        res = getattr(allocator, allocator_func)(ALLOC_SIZE)
        if res:
            allocator.free()

    if not res:
        pytest.skip(f"Allocator {allocator_func} not supported in this platform")

    # THEN
    allocations = list(FileReader(output).get_allocation_records())
    allocs = [
        event
        for event in allocations
        if event.size == ALLOC_SIZE and event.allocator == allocator_type
    ]
    assert len(allocs) == 1
    (alloc,) = allocs

    frees = [
        event
        for event in allocations
        if event.address == alloc.address and event.allocator == AllocatorType.FREE
    ]
    assert len(frees) >= 1


def test_simple_cpp_allocation_tracking(tmp_path):
    # GIVEN
    output = tmp_path / "test.bin"

    # WHEN
    with Tracker(output):
        allocate_cpp_vector(ALLOC_SIZE)

    # THEN
    allocations = list(FileReader(output).get_allocation_records())
    allocs = [event for event in allocations if event.size == ALLOC_SIZE]
    assert len(allocs) == 1
    (alloc,) = allocs

    frees = [
        event
        for event in allocations
        if event.address == alloc.address and event.allocator == AllocatorType.FREE
    ]
    assert len(frees) >= 1


@pytest.mark.parametrize("domain", PYMALLOC_DOMAINS)
@pytest.mark.parametrize(["allocator_func", "allocator_type"], PYMALLOC_ALLOCATORS)
def test_simple_pymalloc_allocation_tracking(
    allocator_func, allocator_type, domain, tmp_path
):
    # GIVEN
    allocator = PymallocMemoryAllocator(domain)
    output = tmp_path / "test.bin"

    # WHEN
    the_allocator = getattr(allocator, allocator_func)
    with Tracker(output, trace_python_allocators=True):
        res = the_allocator(ALLOC_SIZE)
        if res:
            allocator.free()

    if not res:
        pytest.skip(f"Allocator {allocator_func} not supported in this platform")

    # THEN
    allocations = list(FileReader(output).get_allocation_records())
    allocs = [
        event
        for event in allocations
        if event.size == ALLOC_SIZE and event.allocator == allocator_type
    ]
    assert len(allocs) == 1
    (alloc,) = allocs

    frees = [
        event
        for event in allocations
        if event.address == alloc.address
        and event.allocator == AllocatorType.PYMALLOC_FREE
    ]
    assert len(frees) >= 1


@pytest.mark.parametrize("domain", PYMALLOC_DOMAINS)
@pytest.mark.parametrize(["allocator_func", "allocator_type"], PYMALLOC_ALLOCATORS)
def test_pymalloc_allocation_tracking_deactivated(
    allocator_func, allocator_type, domain, tmp_path
):
    # GIVEN
    allocator = PymallocMemoryAllocator(domain)
    output = tmp_path / "test.bin"

    # WHEN
    the_allocator = getattr(allocator, allocator_func)
    with Tracker(output, trace_python_allocators=False):
        res = the_allocator(ALLOC_SIZE)
        if res:
            allocator.free()

    if not res:
        pytest.skip(f"Allocator {allocator_func} not supported in this platform")

    # THEN
    allocations = list(FileReader(output).get_allocation_records())
    allocs = [event for event in allocations if event.allocator == allocator_type]
    assert not allocs
    frees = [
        event for event in allocations if event.allocator == AllocatorType.PYMALLOC_FREE
    ]
    assert not frees


def test_mmap_tracking(tmp_path):
    # GIVEN / WHEN
    output = tmp_path / "test.bin"
    with Tracker(output):
        with mmap.mmap(-1, length=2048, access=mmap.ACCESS_WRITE) as mmap_obj:
            mmap_obj[0:100] = b"a" * 100

    # THEN
    records = list(FileReader(output).get_allocation_records())
    assert len(records) >= 2

    mmap_records = [
        record
        for record in records
        if AllocatorType.MMAP == record.allocator and record.size == 2048
    ]
    assert len(mmap_records) == 1
    mmunmap_record = [
        record for record in records if AllocatorType.MUNMAP == record.allocator
    ]
    assert len(mmunmap_record) == 1


def test_pthread_tracking(tmp_path):
    # GIVEN
    allocator = MemoryAllocator()

    def tracking_function():  # pragma: no cover
        allocator.valloc(ALLOC_SIZE)
        allocator.free()

    # WHEN
    output = tmp_path / "test.bin"
    with Tracker(output):
        allocator.run_in_pthread(tracking_function)

    # THEN
    allocations = list(FileReader(output).get_allocation_records())
    allocs = [
        event
        for event in allocations
        if event.size == ALLOC_SIZE and event.allocator == AllocatorType.VALLOC
    ]
    assert len(allocs) == 1
    (alloc,) = allocs

    frees = [
        event
        for event in allocations
        if event.address == alloc.address and event.allocator == AllocatorType.FREE
    ]
    assert len(frees) >= 1


def test_tracking_with_SIGKILL(tmpdir):
    """Verify that we can successfully retrieve the allocations after SIGKILL."""
    # GIVEN
    output = Path(tmpdir) / "test.bin"
    subprocess_code = textwrap.dedent(
        f"""
        import os
        import signal
        from memray import Tracker
        from memray._test import MemoryAllocator

        allocator = MemoryAllocator()
        output = "{output}"

        with Tracker(output) as tracker:
            num_flushes = 0
            last_size = os.stat(output).st_size

            # Loop until two flushes occur, since the first flush might include
            # a frame record but not an allocation record.
            while num_flushes < 2:
                allocator.valloc(1024)
                new_size = os.stat(output).st_size
                if new_size != last_size:
                    last_size = new_size
                    num_flushes += 1

            # Kill ourselves without letting the tracker clean itself up.
            os.kill(os.getpid(), signal.SIGKILL)
    """
    )

    # WHEN
    process = subprocess.run([sys.executable, "-c", subprocess_code], timeout=5)

    # THEN
    assert process.returncode == -signal.SIGKILL

    records = list(FileReader(output).get_allocation_records())
    vallocs = [
        record
        for record in filter_relevant_allocations(records)
        if record.allocator == AllocatorType.VALLOC
    ]
    (allocation, *rest) = vallocs
    assert allocation.size == 1024


def test_no_allocations(tmpdir):
    """Verify that we can successfully read a file that has no allocations."""
    # GIVEN
    output = Path(tmpdir) / "test.bin"
    subprocess_code = textwrap.dedent(
        f"""
        import os
        from memray import Tracker
        output = "{output}"
        tracker = Tracker(output)
        with tracker:
            os._exit(0)
    """
    )

    # WHEN
    process = subprocess.run([sys.executable, "-c", subprocess_code], timeout=5)

    # THEN
    assert process.returncode == 0

    records = list(FileReader(output).get_allocation_records())
    assert not records


def test_unsupported_operations_on_aggregated_capture(tmpdir):
    """Verify that we can successfully read a file that has no allocations."""
    # GIVEN
    output = Path(tmpdir) / "test.bin"
    subprocess_code = textwrap.dedent(
        f"""
        import os
        from memray import Tracker, FileFormat
        output = "{output}"
        tracker = Tracker(output, file_format=FileFormat.AGGREGATED_ALLOCATIONS)
        with tracker:
            pass
        """
    )
    process = subprocess.run([sys.executable, "-c", subprocess_code], timeout=5)
    assert process.returncode == 0
    reader = FileReader(output)

    # WHEN / THEN
    with pytest.raises(
        NotImplementedError,
        match="Can't find temporary allocations using a pre-aggregated capture file",
    ):
        list(reader.get_temporary_allocation_records())

    with pytest.raises(
        NotImplementedError,
        match="Can't get all allocations from a pre-aggregated capture file",
    ):
        list(reader.get_allocation_records())

    with pytest.raises(
        NotImplementedError,
        match="Can't compute statistics using a pre-aggregated capture file",
    ):
        compute_statistics(str(output))


@pytest.mark.parametrize(
    "file_format",
    [
        pytest.param(FileFormat.ALL_ALLOCATIONS, id="ALL_ALLOCATIONS"),
        pytest.param(FileFormat.AGGREGATED_ALLOCATIONS, id="AGGREGATED_ALLOCATIONS"),
    ],
)
class TestHighWatermark:
    def test_no_allocations_while_tracking(self, tmp_path, file_format):
        # GIVEN
        output = tmp_path / "test.bin"

        def tracking_function():
            pass

        with PrimeCaches():
            tracking_function()

        # WHEN
        with Tracker(output, native_traces=True, file_format=file_format):
            tracking_function()

        # THEN
        assert list(FileReader(output).get_high_watermark_allocation_records()) == []

    @pytest.mark.parametrize(["allocator_func", "allocator_type"], ALLOCATORS)
    def test_simple_allocation_tracking(
        self, tmp_path, allocator_func, allocator_type, file_format
    ):
        # GIVEN
        allocator = MemoryAllocator()
        output = tmp_path / "test.bin"

        # WHEN
        with Tracker(output, file_format=file_format):
            res = getattr(allocator, allocator_func)(ALLOC_SIZE)
            if res:
                allocator.free()

        if not res:
            pytest.skip(f"Allocator {allocator_func} not supported in this platform")

        # THEN
        peak_allocations_unfiltered = FileReader(
            output
        ).get_high_watermark_allocation_records()
        peak_allocations = [
            record
            for record in peak_allocations_unfiltered
            if record.size == ALLOC_SIZE
        ]
        assert len(peak_allocations) == 1, peak_allocations

        record = peak_allocations[0]
        assert record.allocator == allocator_type
        assert record.n_allocations == 1

    def test_multiple_high_watermark(self, tmp_path, file_format):
        # GIVEN
        allocator = MemoryAllocator()
        output = tmp_path / "test.bin"

        def test_function():
            for _ in range(2):
                allocator.valloc(1024)
                allocator.free()

        with PrimeCaches():
            test_function()

        # WHEN
        with Tracker(output, file_format=file_format):
            test_function()

        # THEN
        reader = FileReader(output)

        if file_format == FileFormat.ALL_ALLOCATIONS:
            all_allocations = list(
                filter_relevant_allocations(reader.get_allocation_records())
            )
            assert len(all_allocations) == 4

        peak_allocations = list(
            filter_relevant_allocations(reader.get_high_watermark_allocation_records())
        )
        assert len(peak_allocations) == 1
        record = peak_allocations[0]

        assert record.allocator == AllocatorType.VALLOC
        assert record.size == 1024
        assert record.n_allocations == 1

    def test_freed_before_high_watermark_do_not_appear(self, tmp_path, file_format):
        # GIVEN
        allocator1 = MemoryAllocator()
        allocator2 = MemoryAllocator()
        output = tmp_path / "test.bin"

        def test_function():
            allocator1.valloc(1024)
            allocator1.free()
            allocator2.valloc(2048)
            allocator2.free()

        with PrimeCaches():
            test_function()

        # WHEN
        with Tracker(output, file_format=file_format):
            test_function()

        # THEN
        reader = FileReader(output)

        if file_format == FileFormat.ALL_ALLOCATIONS:
            all_allocations = list(
                filter_relevant_allocations(reader.get_allocation_records())
            )
            assert len(all_allocations) == 4

        peak_allocations = list(
            filter_relevant_allocations(reader.get_high_watermark_allocation_records())
        )
        assert len(peak_allocations) == 1

        record = peak_allocations[0]
        assert record.allocator == AllocatorType.VALLOC
        assert record.size == 2048
        assert record.n_allocations == 1

    def test_freed_after_high_watermark_do_not_appear(self, tmp_path, file_format):
        # GIVEN
        allocator1 = MemoryAllocator()
        allocator2 = MemoryAllocator()
        output = tmp_path / "test.bin"

        def test_function():
            allocator2.valloc(2048)
            allocator2.free()
            allocator1.valloc(1024)
            allocator1.free()

        with PrimeCaches():
            test_function()

        # WHEN
        with Tracker(output, file_format=file_format):
            test_function()

        # THEN
        reader = FileReader(output)

        if file_format == FileFormat.ALL_ALLOCATIONS:
            all_allocations = list(
                filter_relevant_allocations(reader.get_allocation_records())
            )
            assert len(all_allocations) == 4

        peak_allocations = list(
            filter_relevant_allocations(reader.get_high_watermark_allocation_records())
        )
        assert len(peak_allocations) == 1

        record = peak_allocations[0]
        assert record.allocator == AllocatorType.VALLOC
        assert record.size == 2048
        assert record.n_allocations == 1

    def test_allocations_aggregation_on_same_line(self, tmp_path, file_format):
        # GIVEN
        allocators = []
        output = tmp_path / "test.bin"

        # WHEN
        with Tracker(output, file_format=file_format):
            for _ in range(2):
                allocator = MemoryAllocator()
                allocators.append(allocator)

                allocator.valloc(1024)

            for allocator in allocators:
                allocator.free()

        # THEN
        reader = FileReader(output)

        if file_format == FileFormat.ALL_ALLOCATIONS:
            all_allocations = list(
                filter_relevant_allocations(reader.get_allocation_records())
            )
            assert len(all_allocations) == 4

        peak_allocations = list(
            filter_relevant_allocations(reader.get_high_watermark_allocation_records())
        )
        assert len(peak_allocations) == 1

        record = peak_allocations[0]
        assert record.allocator == AllocatorType.VALLOC
        assert record.size == 2048
        assert record.n_allocations == 2

    def test_aggregation_same_python_stack_and_same_native_stack(
        self, tmp_path, file_format
    ):
        # GIVEN
        allocators = []
        output = tmp_path / "test.bin"

        # WHEN
        with Tracker(output, native_traces=True, file_format=file_format):
            for _ in range(2):
                allocator = MemoryAllocator()
                allocators.append(allocator)

                allocator.valloc(1024)

            for allocator in allocators:
                allocator.free()

        # THEN
        reader = FileReader(output)

        if file_format == FileFormat.ALL_ALLOCATIONS:
            all_allocations = list(
                filter_relevant_allocations(reader.get_allocation_records())
            )
            assert len(all_allocations) == 4

        peak_allocations = list(
            filter_relevant_allocations(reader.get_high_watermark_allocation_records())
        )
        assert len(peak_allocations) == 1

        record = peak_allocations[0]
        assert record.allocator == AllocatorType.VALLOC
        assert record.size == 2048
        assert record.n_allocations == 2

    def test_allocations_aggregation_on_different_lines(self, tmp_path, file_format):
        # GIVEN
        allocator1 = MemoryAllocator()
        allocator2 = MemoryAllocator()
        output = tmp_path / "test.bin"

        # WHEN
        with Tracker(output, file_format=file_format):
            allocator1.valloc(1024)
            allocator2.valloc(2048)
            allocator1.free()
            allocator2.free()

        # THEN
        reader = FileReader(output)

        if file_format == FileFormat.ALL_ALLOCATIONS:
            all_allocations = list(
                filter_relevant_allocations(reader.get_allocation_records())
            )
            assert len(all_allocations) == 4

        peak_allocations = list(
            filter_relevant_allocations(reader.get_high_watermark_allocation_records())
        )
        assert len(peak_allocations) == 2
        assert sum(record.size for record in peak_allocations) == 1024 + 2048
        assert all(record.n_allocations == 1 for record in peak_allocations)

    def test_aggregation_same_python_stack_but_different_native_stack(
        self, tmp_path, file_format
    ):
        # GIVEN
        output = tmp_path / "test.bin"

        # WHEN
        with Tracker(output, native_traces=True, file_format=file_format):
            _cython_allocate_in_two_places(ALLOC_SIZE)

        # THEN
        reader = FileReader(output)

        if file_format == FileFormat.ALL_ALLOCATIONS:
            all_allocations = list(
                filter_relevant_allocations(reader.get_allocation_records())
            )
            assert len(all_allocations) == 4

        peak_allocations = list(
            filter_relevant_allocations(reader.get_high_watermark_allocation_records())
        )
        assert len(peak_allocations) == 2
        assert (
            sum(record.size for record in peak_allocations) == ALLOC_SIZE + ALLOC_SIZE
        )
        assert all(record.n_allocations == 1 for record in peak_allocations)

    def test_non_freed_allocations_are_accounted_for(self, tmp_path, file_format):
        # GIVEN
        allocator1 = MemoryAllocator()
        allocator2 = MemoryAllocator()
        output = tmp_path / "test.bin"

        # WHEN
        with Tracker(output, file_format=file_format):
            allocator1.valloc(1024)
            allocator2.valloc(2048)
            allocator1.free()
            allocator2.free()

        # THEN
        reader = FileReader(output)

        if file_format == FileFormat.ALL_ALLOCATIONS:
            all_allocations = list(
                filter_relevant_allocations(reader.get_allocation_records())
            )
            assert len(all_allocations) == 4

        peak_allocations = list(
            filter_relevant_allocations(reader.get_high_watermark_allocation_records())
        )
        assert len(peak_allocations) == 2
        assert sum(record.size for record in peak_allocations) == 1024 + 2048
        assert all(record.n_allocations == 1 for record in peak_allocations)

    def test_final_allocation_is_peak(self, tmp_path, file_format):
        # GIVEN
        allocator1 = MemoryAllocator()
        allocator2 = MemoryAllocator()
        output = tmp_path / "test.bin"

        # WHEN
        with Tracker(output, file_format=file_format):
            allocator1.valloc(1024)
            allocator1.free()
            allocator2.valloc(2048)
        allocator2.free()

        # THEN
        reader = FileReader(output)

        if file_format == FileFormat.ALL_ALLOCATIONS:
            all_allocations = list(
                filter_relevant_allocations(reader.get_allocation_records())
            )
            assert len(all_allocations) == 3

        peak_allocations = list(
            filter_relevant_allocations(reader.get_high_watermark_allocation_records())
        )
        assert len(peak_allocations) == 1

        record = peak_allocations[0]
        assert record.n_allocations == 1
        assert record.allocator == AllocatorType.VALLOC
        assert record.size == 2048

    def test_spiky_generally_increasing_to_final_peak(self, tmp_path, file_format):
        """Checks multiple aspects with an interesting toy function."""

        # GIVEN
        def recursive(n, chunk_size):
            """Mimics generally-increasing but spiky usage"""
            if not n:
                return

            allocator = MemoryAllocator()
            print(f"+{n:>2} kB")
            allocator.valloc(n * chunk_size)

            # Don't keep allocated memory when recursing, ~50% of the calls.
            if n % 2:
                allocator.free()
                print(f"-{n:>2} kB")
                recursive(n - 1, chunk_size)
            else:
                recursive(n - 1, chunk_size)
                allocator.free()
                print(f"-{n:>2} kB")

        # WHEN
        output = tmp_path / "test.bin"
        with Tracker(output, file_format=file_format):
            recursive(10, 1024)

        # THEN
        reader = FileReader(output)

        if file_format == FileFormat.ALL_ALLOCATIONS:
            all_allocations = list(
                filter_relevant_allocations(reader.get_allocation_records())
            )
            assert len(all_allocations) == 20
            assert sum(record.size for record in all_allocations) == 56320

        peak_allocations = list(
            filter_relevant_allocations(reader.get_high_watermark_allocation_records())
        )
        assert all(record.n_allocations == 1 for record in peak_allocations)

        expected = {10, 8, 6, 4, 2, 1}
        assert len(peak_allocations) == len(expected)
        assert {record.size / 1024 for record in peak_allocations} == expected

    def test_allocations_after_high_watermark_is_freed_do_not_appear(
        self, tmp_path, file_format
    ):
        # GIVEN
        allocator = MemoryAllocator()
        output = tmp_path / "test.bin"

        def test_function():
            allocator.valloc(2048)
            allocator.free()
            allocator.valloc(1024)

        with PrimeCaches():
            test_function()
            allocator.free()

        # WHEN
        with Tracker(output, file_format=file_format):
            test_function()
        allocator.free()

        # THEN
        reader = FileReader(output)

        if file_format == FileFormat.ALL_ALLOCATIONS:
            all_allocations = list(
                filter_relevant_allocations(reader.get_allocation_records())
            )
            assert len(all_allocations) == 3

        peak_allocations = list(
            filter_relevant_allocations(reader.get_high_watermark_allocation_records())
        )
        assert len(peak_allocations) == 1

        record = peak_allocations[0]
        assert record.n_allocations == 1
        assert record.allocator == AllocatorType.VALLOC
        assert record.size == 2048

    def test_partial_munmap(self, tmp_path, file_format):
        """Partial munmap operations should be accurately tracked: we should
        only account for the removal of the actually munmap'd chunk and not
        the entire mmap'd region when a partial munmap is performed."""

        # GIVEN
        output = tmp_path / "test.bin"

        def test_function():
            # Mmap some memory and free just the first page. This should not register
            # the deallocation of the entire mmap'd region, only one page.
            alloc = MmapAllocator(2 * PAGE_SIZE)
            alloc.munmap(PAGE_SIZE)

            # Now perform the peak allocation. This should be detected as the peak index.
            MmapAllocator(10 * PAGE_SIZE)
            # At this point we should have 11 * PAGE_SIZE allocated

        with PrimeCaches():
            test_function()

        # WHEN
        with Tracker(output, file_format=file_format):
            test_function()

        # THEN
        reader = FileReader(output)
        peak_allocations = list(reader.get_high_watermark_allocation_records())
        assert len(peak_allocations) == 2
        peak_memory = sum(x.size for x in peak_allocations)
        assert peak_memory == 11 * PAGE_SIZE

    def test_partial_munmap_gap(self, tmp_path, file_format):
        """Validate that removing chunks from a mmap'd region correctly accounts
        for the parts removed. This test allocates 4 pages and removes the first
        and last pages of the mmap'd region."""

        # GIVEN
        output = tmp_path / "test.bin"

        def test_function():
            # Mmap some memory and free two pages: one at the beginning and one at the
            # end of the region.
            alloc = MmapAllocator(4 * PAGE_SIZE)
            alloc.munmap(PAGE_SIZE)
            alloc.munmap(PAGE_SIZE, 3 * PAGE_SIZE)

            # Now perform the peak allocation. This should be detected as the peak index.
            MmapAllocator(10 * PAGE_SIZE)
            # At this point we should have 12 * PAGE_SIZE allocated

        with PrimeCaches():
            test_function()

        # WHEN
        with Tracker(output, file_format=file_format):
            test_function()

        # THEN
        reader = FileReader(output)
        peak_allocations = list(reader.get_high_watermark_allocation_records())
        assert len(peak_allocations) == 2
        peak_memory = sum(x.size for x in peak_allocations)
        assert peak_memory == 12 * PAGE_SIZE

    def test_munmap_multiple_mmaps(self, tmp_path, file_format):
        """Allocate multiple contiguous mmap'd regions and then deallocate all of them
        with munmap in one go."""

        # GIVEN
        output = tmp_path / "test.bin"
        with Tracker(output, file_format=file_format):
            # Ensure we have a long enough free buffer for the contiguous
            # mmap's. We also need to make sure the allocation addresses are
            # page-aligned later, and mmap does that for us. We can then use
            # the pointer from this allocation for the actual test.
            buf = MmapAllocator(8 * PAGE_SIZE)
            buf.munmap(8 * PAGE_SIZE)

            # WHEN
            # Allocate 2 contiguous chunks of 4 pages (8 pages in total) and free them
            # with a single munmap
            alloc1 = MmapAllocator(4 * PAGE_SIZE, buf.address)
            MmapAllocator(4 * PAGE_SIZE, alloc1.address + (4 * PAGE_SIZE))
            alloc1.munmap(8 * PAGE_SIZE)

            # Now perform the peak allocation. This should be detected as the peak index.
            MmapAllocator(10 * PAGE_SIZE)

        # THEN
        reader = FileReader(output)
        peak_allocations = list(
            filter_relevant_allocations(
                reader.get_high_watermark_allocation_records(), ranged=True
            )
        )
        peak_memory = sum(x.size for x in peak_allocations)
        assert peak_memory == 10 * PAGE_SIZE

    def test_munmap_multiple_mmaps_multiple_munmaps(self, tmp_path, file_format):
        """Allocate multiple contiguous mmap'd regions and then with multiple munmap's, each
        deallocating several mmap'd areas in one go."""
        # GIVEN
        output = tmp_path / "test.bin"
        with Tracker(output, file_format=file_format):
            # Ensure we have a long enough free buffer for the contiguous
            # mmap's. We also need to make sure the allocation addresses are
            # page-aligned later, and mmap does that for us. We can then use
            # the pointer from this allocation for the actual test.
            buf = MmapAllocator(8 * PAGE_SIZE)
            buf.munmap(8 * PAGE_SIZE)

            # WHEN
            alloc1 = MmapAllocator(2 * PAGE_SIZE, buf.address)
            MmapAllocator(2 * PAGE_SIZE, buf.address + (2 * PAGE_SIZE))
            MmapAllocator(2 * PAGE_SIZE, buf.address + (4 * PAGE_SIZE))
            MmapAllocator(2 * PAGE_SIZE, buf.address + (6 * PAGE_SIZE))
            alloc1.munmap(4 * PAGE_SIZE)
            alloc1.munmap(4 * PAGE_SIZE, 4 * PAGE_SIZE)

            # Now perform the peak allocation. This should be detected as the peak index.
            MmapAllocator(10 * PAGE_SIZE)

        # THEN
        reader = FileReader(output)
        peak_allocations = list(
            filter_relevant_allocations(
                reader.get_high_watermark_allocation_records(), ranged=True
            )
        )
        peak_memory = sum(x.size for x in peak_allocations)
        assert peak_memory == 10 * PAGE_SIZE

    def test_partial_munmap_multiple_split_in_middle(self, tmp_path, file_format):
        """Deallocate pages in of a larger mmap'd area, splitting it into 3 areas."""
        # GIVEN
        output = tmp_path / "test.bin"

        def test_function():
            alloc = MmapAllocator(5 * PAGE_SIZE)
            alloc.munmap(PAGE_SIZE, 1 * PAGE_SIZE)
            alloc.munmap(PAGE_SIZE, 3 * PAGE_SIZE)
            MmapAllocator(10 * PAGE_SIZE)

        with PrimeCaches():
            test_function()

        # WHEN
        with Tracker(output, file_format=file_format):
            test_function()

        # THEN
        reader = FileReader(output)
        peak_allocations = list(reader.get_high_watermark_allocation_records())
        assert len(peak_allocations) == 2
        peak_memory = sum(x.size for x in peak_allocations)
        assert peak_memory == 13 * PAGE_SIZE

    def test_partial_munmap_split_in_middle(self, tmp_path, file_format):
        """Deallocate a single page in the middle of a larger mmap'd area."""
        # GIVEN
        output = tmp_path / "test.bin"

        def test_function():
            alloc = MmapAllocator(8 * PAGE_SIZE)
            alloc.munmap(PAGE_SIZE, 4 * PAGE_SIZE)

            MmapAllocator(10 * PAGE_SIZE)

        with PrimeCaches():
            test_function()

        # WHEN
        with Tracker(output, file_format=file_format):
            test_function()

        # THEN
        reader = FileReader(output)
        peak_allocations = list(reader.get_high_watermark_allocation_records())
        assert len(peak_allocations) == 2
        peak_memory = sum(x.size for x in peak_allocations)
        assert peak_memory == 17 * PAGE_SIZE


@pytest.mark.parametrize(
    "file_format",
    [
        pytest.param(FileFormat.ALL_ALLOCATIONS, id="ALL_ALLOCATIONS"),
        pytest.param(FileFormat.AGGREGATED_ALLOCATIONS, id="AGGREGATED_ALLOCATIONS"),
    ],
)
class TestLeaks:
    def test_leaks_allocations_are_detected(self, tmp_path, file_format):
        # GIVEN
        allocator = MemoryAllocator()
        output = tmp_path / "test.bin"

        # WHEN
        with Tracker(output, file_format=file_format):
            allocator.valloc(1024)
        allocator.free()

        # THEN
        reader = FileReader(output)

        if file_format == FileFormat.ALL_ALLOCATIONS:
            all_allocations = list(
                filter_relevant_allocations(reader.get_allocation_records())
            )
            assert len(all_allocations) == 1

        leaked_allocations = list(
            filter_relevant_allocations(reader.get_leaked_allocation_records())
        )
        assert len(leaked_allocations) == 1

        record = leaked_allocations[0]
        assert record.n_allocations == 1
        assert record.allocator == AllocatorType.VALLOC
        assert record.size == 1024

    def test_allocations_that_are_freed_do_not_appear_as_leaks(
        self, tmp_path, file_format
    ):
        # GIVEN
        allocator = MemoryAllocator()
        output = tmp_path / "test.bin"

        # WHEN
        with Tracker(output, file_format=file_format):
            allocator.valloc(1024)
            allocator.free()
            allocator.valloc(1024)
            allocator.free()
            allocator.valloc(1024)
            allocator.free()

        # THEN
        reader = FileReader(output)

        if file_format == FileFormat.ALL_ALLOCATIONS:
            all_allocations = list(
                filter_relevant_allocations(reader.get_allocation_records())
            )
            assert len(all_allocations) == 6

        leaked_allocations = list(
            filter_relevant_allocations(reader.get_leaked_allocation_records())
        )
        assert not leaked_allocations

    def test_leak_that_happens_in_the_middle_is_detected(self, tmp_path, file_format):
        # GIVEN
        allocator = MemoryAllocator()
        leak_allocator = MemoryAllocator()
        output = tmp_path / "test.bin"

        # WHEN
        with Tracker(output, file_format=file_format):
            allocator.valloc(1024)
            allocator.free()
            allocator.valloc(1024)
            leak_allocator.valloc(2048)
            allocator.free()
            allocator.valloc(1024)
            allocator.free()
        leak_allocator.free()

        # THEN
        reader = FileReader(output)

        if file_format == FileFormat.ALL_ALLOCATIONS:
            all_allocations = list(
                filter_relevant_allocations(reader.get_allocation_records())
            )
            assert len(all_allocations) == 7

        leaked_allocations = list(
            filter_relevant_allocations(reader.get_leaked_allocation_records())
        )

        assert len(leaked_allocations) == 1

        record = leaked_allocations[0]
        assert record.n_allocations == 1
        assert record.allocator == AllocatorType.VALLOC
        assert record.size == 2048

    def test_leaks_that_happen_in_different_lines(self, tmp_path, file_format):
        # GIVEN
        allocator1 = MemoryAllocator()
        allocator2 = MemoryAllocator()
        output = tmp_path / "test.bin"

        # WHEN
        with Tracker(output, file_format=file_format):
            allocator1.valloc(1024)
            allocator2.valloc(2048)

        allocator1.free()
        allocator2.free()

        # THEN
        leaked_allocations = list(
            filter_relevant_allocations(
                FileReader(output).get_leaked_allocation_records()
            )
        )
        assert len(leaked_allocations) == 2
        assert sum(record.size for record in leaked_allocations) == 1024 + 2048
        assert all(record.n_allocations == 1 for record in leaked_allocations)

    def test_leaks_that_happen_in_the_same_function_are_aggregated(
        self, tmp_path, file_format
    ):
        # GIVEN
        allocators = []
        output = tmp_path / "test.bin"

        def foo():
            allocator = MemoryAllocator()
            allocator.valloc(1024)
            allocators.append(allocator)

        # WHEN
        with Tracker(output, file_format=file_format):
            for _ in range(10):
                foo()

        for allocator in allocators:
            allocator.free()

        # THEN
        reader = FileReader(output)

        if file_format == FileFormat.ALL_ALLOCATIONS:
            all_allocations = list(
                filter_relevant_allocations(reader.get_allocation_records())
            )
            assert len(all_allocations) == 10

        leaked_allocations = list(
            filter_relevant_allocations(reader.get_leaked_allocation_records())
        )
        assert len(leaked_allocations) == 1
        (allocation,) = leaked_allocations
        assert allocation.size == 1024 * 10
        assert allocation.n_allocations == 10

    def test_unmatched_deallocations_are_not_reported(self, tmp_path, file_format):
        # GIVEN
        allocator = MemoryAllocator()
        output = tmp_path / "test.bin"

        # WHEN
        allocator.valloc(ALLOC_SIZE)
        with Tracker(output, file_format=file_format):
            allocator.free()

        # THEN
        reader = FileReader(output)

        if file_format == FileFormat.ALL_ALLOCATIONS:
            all_allocations = list(reader.get_allocation_records())
            assert len(all_allocations) >= 1

        assert not list(
            filter_relevant_allocations(reader.get_leaked_allocation_records())
        )

    def test_thread_allocations_multiple_threads(self, tmpdir, file_format):
        # GIVEN
        def allocating_function(allocator, amount, stop_flag):
            allocator.posix_memalign(amount)
            allocator.posix_memalign(amount)
            # We need a barrier as pthread might reuse the same thread ID if the first thread
            # finishes before the other one starts
            stop_flag.wait()

        # WHEN
        alloc1 = MemoryAllocator()
        stop_flag1 = threading.Event()
        alloc2 = MemoryAllocator()
        stop_flag2 = threading.Event()
        output = Path(tmpdir) / "test.bin"
        with Tracker(output, file_format=file_format):
            t1 = threading.Thread(
                target=allocating_function, args=(alloc1, 2048, stop_flag1)
            )
            t1.start()
            t2 = threading.Thread(
                target=allocating_function, args=(alloc2, 2048, stop_flag2)
            )
            t2.start()

            stop_flag1.set()
            t1.join()
            stop_flag2.set()
            t2.join()

        # THEN
        reader = FileReader(output)

        if file_format == FileFormat.ALL_ALLOCATIONS:
            all_allocations = [
                record
                for record in reader.get_allocation_records()
                if record.allocator == AllocatorType.POSIX_MEMALIGN
            ]
            assert len(all_allocations) == 4

        high_watermark_records = (
            record
            for record in reader.get_high_watermark_allocation_records(
                merge_threads=False
            )
            if record.allocator == AllocatorType.POSIX_MEMALIGN
        )
        # Group the allocations per thread
        records = collections.defaultdict(list)
        for record in high_watermark_records:
            records[record.tid].append(record)
        assert len(records.keys()) == 2
        # Each thread should have 4096 bytes total allocations
        for allocations in records.values():
            assert sum(allocation.size for allocation in allocations) == 4096


class TestTemporaryAllocations:
    def test_temporary_allocations_are_detected(self, tmp_path):
        # GIVEN
        allocator = MemoryAllocator()
        output = tmp_path / "test.bin"

        # WHEN
        with Tracker(output):
            allocator.valloc(1024)
            allocator.free()

        # THEN
        reader = FileReader(output)
        all_allocations = list(
            filter_relevant_allocations(reader.get_allocation_records())
        )
        assert len(all_allocations) == 2

        temporary_allocations = list(
            filter_relevant_allocations(reader.get_temporary_allocation_records())
        )
        assert len(temporary_allocations) == 1

        record = temporary_allocations[0]
        assert record.n_allocations == 1
        assert record.allocator == AllocatorType.VALLOC
        assert record.size == 1024

    def test_temporary_allocations_with_two_allocators_are_detected(self, tmp_path):
        # GIVEN
        allocator1 = MemoryAllocator()
        allocator2 = MemoryAllocator()
        output = tmp_path / "test.bin"

        # WHEN
        with Tracker(output):
            allocator1.valloc(1024)
            allocator1.free()
            allocator2.valloc(1024)
            allocator2.free()
        # THEN
        reader = FileReader(output)
        all_allocations = list(
            filter_relevant_allocations(reader.get_allocation_records())
        )
        assert len(all_allocations) == 4

        temporary_allocations = list(
            filter_relevant_allocations(reader.get_temporary_allocation_records())
        )
        assert len(temporary_allocations) == 2

        for record in temporary_allocations:
            assert record.n_allocations == 1
            assert record.allocator == AllocatorType.VALLOC
            assert record.size == 1024

    @pytest.mark.parametrize(
        "buffer_size",
        [
            1,
            2,
            5,
            10,
        ],
    )
    def test_temporary_allocations_outside_buffer_are_not_detected(
        self, tmp_path, buffer_size
    ):
        # GIVEN
        allocators = [MemoryAllocator() for _ in range(buffer_size + 1)]
        output = tmp_path / "test.bin"

        # WHEN
        with Tracker(output):
            for allocator in allocators:
                allocator.valloc(1024)
            for allocator in reversed(allocators):
                allocator.free()

        # THEN
        reader = FileReader(output)
        all_allocations = list(
            filter_relevant_allocations(reader.get_allocation_records())
        )
        assert len(all_allocations) == 2 * (buffer_size + 1)

        temporary_allocations = list(
            filter_relevant_allocations(
                reader.get_temporary_allocation_records(threshold=buffer_size - 1)
            )
        )
        assert len(temporary_allocations) == 1
        (allocation,) = temporary_allocations
        assert allocation.n_allocations == buffer_size

    def test_temporary_allocations_that_happen_in_different_lines(self, tmp_path):
        # GIVEN
        allocator1 = MemoryAllocator()
        allocator2 = MemoryAllocator()
        output = tmp_path / "test.bin"

        # WHEN
        with Tracker(output):
            allocator1.valloc(1024)
            allocator1.free()
            allocator2.valloc(2048)
            allocator2.free()

        # THEN
        temporary_allocations = list(
            filter_relevant_allocations(
                FileReader(output).get_temporary_allocation_records()
            )
        )
        assert len(temporary_allocations) == 2
        assert sum(record.size for record in temporary_allocations) == 1024 + 2048
        assert all(record.n_allocations == 1 for record in temporary_allocations)

    def test_temporary_allocations_that_happen_in_the_same_function_are_aggregated(
        self, tmp_path
    ):
        # GIVEN
        output = tmp_path / "test.bin"

        def foo():
            allocator = MemoryAllocator()
            allocator.valloc(1024)
            allocator.free()

        # WHEN
        with Tracker(output):
            for _ in range(10):
                foo()

        # THEN
        reader = FileReader(output)
        all_allocations = list(
            filter_relevant_allocations(reader.get_allocation_records())
        )
        assert len(all_allocations) == 10 + 10  # 10 x valloc + 10 x free

        temporary_allocations = list(
            filter_relevant_allocations(reader.get_temporary_allocation_records())
        )
        assert len(temporary_allocations) == 1
        (allocation,) = temporary_allocations
        assert allocation.size == 1024 * 10
        assert allocation.n_allocations == 10

    def test_unmatched_allocations_are_not_reported(self, tmp_path):
        # GIVEN
        allocator = MemoryAllocator()
        output = tmp_path / "test.bin"

        # WHEN
        with Tracker(output):
            allocator.valloc(ALLOC_SIZE)
        allocator.free()

        # THEN
        reader = FileReader(output)
        all_allocations = list(reader.get_allocation_records())
        assert len(all_allocations) >= 1
        assert not list(
            filter_relevant_allocations(reader.get_temporary_allocation_records())
        )

    def test_thread_allocations_multiple_threads(self, tmpdir):
        # GIVEN
        def allocating_function(allocator, amount, stop_flag):
            allocator.posix_memalign(amount)
            allocator.free()
            allocator.posix_memalign(amount)
            allocator.free()
            # We need a barrier as pthread might reuse the same thread ID if the first thread
            # finishes before the other one starts
            stop_flag.wait()

        # WHEN
        alloc1 = MemoryAllocator()
        stop_flag1 = threading.Event()
        alloc2 = MemoryAllocator()
        stop_flag2 = threading.Event()
        output = Path(tmpdir) / "test.bin"
        with Tracker(output):
            t1 = threading.Thread(
                target=allocating_function, args=(alloc1, 2048, stop_flag1)
            )
            t1.start()
            t2 = threading.Thread(
                target=allocating_function, args=(alloc2, 2048, stop_flag2)
            )
            t2.start()

            stop_flag1.set()
            t1.join()
            stop_flag2.set()
            t2.join()

        # THEN
        reader = FileReader(output)
        all_allocations = [
            record
            for record in reader.get_allocation_records()
            if record.allocator == AllocatorType.POSIX_MEMALIGN
        ]
        assert len(all_allocations) == 4

        temporary_allocation_records = (
            record
            for record in reader.get_temporary_allocation_records(merge_threads=False)
            if record.allocator == AllocatorType.POSIX_MEMALIGN
        )
        # Group the allocations per thread
        records = collections.defaultdict(list)
        for record in temporary_allocation_records:
            records[record.tid].append(record)
        assert len(records.keys()) == 2
        # Each thread should have 4096 bytes total allocations
        for allocations in records.values():
            assert sum(allocation.size for allocation in allocations) == 4096

    def test_intertwined_temporary_allocations_in_threads(self, tmpdir):
        """This test checks that temporary allocations are correctly detected
        when they happen in different threads with interleaved calls to malloc
        and free.

        Note that there is not an easy way to synchronize this test because
        using condition variables may allocate, disrupting the test as there
        will be other spurious allocations between our call to posix_memalign
        and the call to free.

        The best effort is that both threads are synchronized by spinlocking
        around a python boolean that uses the GIL as means to ensure the correct
        ordering of allocations.
        """

        # GIVEN
        thread1_allocated = False
        thread1_should_free = False

        def thread1_body(allocator):
            nonlocal thread1_allocated
            allocator.posix_memalign(1234)
            thread1_allocated = True
            while not thread1_should_free:
                pass
            allocator.free()

        def thread2_body(allocator):
            nonlocal thread1_should_free
            while not thread1_allocated:
                pass
            allocator.posix_memalign(1234)
            allocator.free()
            thread1_should_free = True

        # WHEN
        alloc1 = MemoryAllocator()
        alloc2 = MemoryAllocator()
        output = Path(tmpdir) / "test.bin"
        with Tracker(output):
            t1 = threading.Thread(target=thread1_body, args=[alloc1])
            t1.start()
            t2 = threading.Thread(
                target=thread2_body,
                args=[alloc2],
            )
            t2.start()

            t1.join()
            t2.join()

        # THEN
        reader = FileReader(output)
        all_allocations = [
            record
            for record in reader.get_allocation_records()
            if record.allocator == AllocatorType.POSIX_MEMALIGN
        ]
        assert len(all_allocations) == 2

        temporary_allocation_records = (
            record
            for record in reader.get_temporary_allocation_records(merge_threads=False)
            if record.allocator == AllocatorType.POSIX_MEMALIGN
        )
        # Group the allocations per thread
        records = collections.defaultdict(list)
        for record in temporary_allocation_records:
            records[record.tid].append(record)
        assert len(records.keys()) == 2
        # Each thread should have 1234 bytes total allocations
        for allocations in records.values():
            assert sum(allocation.size for allocation in allocations) == 1234


class TestHeader:
    def test_get_header(self, monkeypatch, tmpdir):
        # GIVEN
        allocator = MemoryAllocator()
        output = Path(tmpdir) / "test.bin"

        with PrimeCaches():
            pass

        # WHEN

        monkeypatch.setattr(sys, "argv", ["python", "-m", "pytest"])
        with run_without_tracer():
            start_time = datetime.datetime.now().astimezone()
            with Tracker(output):
                for _ in range(100):
                    allocator.valloc(1024)
            end_time = datetime.datetime.now().astimezone()

        reader = FileReader(output)
        n_records = len(list(reader.get_allocation_records()))
        metadata = reader.metadata

        # THEN
        assert metadata.end_time > metadata.start_time
        assert abs(metadata.start_time - start_time).seconds < 1
        assert abs(metadata.end_time - end_time).seconds < 1
        assert metadata.total_allocations == n_records
        assert metadata.command_line == "python -m pytest"
        assert metadata.peak_memory == 1024 * 100

    def test_get_header_after_snapshot(self, monkeypatch, tmpdir):
        """Verify that we can successfully retrieve the metadata after querying
        the high watermark snapshot."""
        # GIVEN
        allocator = MemoryAllocator()
        output = Path(tmpdir) / "test.bin"

        with PrimeCaches():
            pass

        # WHEN

        monkeypatch.setattr(sys, "argv", ["python", "-m", "pytest"])
        start_time = datetime.datetime.now().astimezone()
        with Tracker(output):
            for _ in range(100):
                allocator.valloc(1024)
        end_time = datetime.datetime.now().astimezone()

        reader = FileReader(output)
        peak, *_ = list(reader.get_high_watermark_allocation_records())
        metadata = reader.metadata

        # THEN
        assert metadata.end_time > metadata.start_time
        assert abs(metadata.start_time - start_time).seconds < 1
        assert abs(metadata.end_time - end_time).seconds < 1
        assert metadata.total_allocations == peak.n_allocations
        assert metadata.command_line == "python -m pytest"
        assert metadata.peak_memory == 1024 * 100

    @pytest.mark.parametrize(
        "allocator, allocator_name",
        [
            ("malloc", "malloc"),
            ("pymalloc", "pymalloc"),
            ("pymalloc_debug", "pymalloc debug"),
        ],
    )
    def test_header_allocator(self, allocator, allocator_name, tmpdir):
        # GIVEN
        output = Path(tmpdir) / "test.bin"

        # WHEN

        subprocess_code = textwrap.dedent(
            f"""
        from memray import Tracker
        from memray._test import MemoryAllocator
        allocator = MemoryAllocator()

        with Tracker('{output}'):
            allocator.valloc(1024)
            allocator.free()
        """
        )

        subprocess.run(
            [sys.executable, "-c", subprocess_code],
            timeout=5,
            env={"PYTHONMALLOC": allocator},
        )

        reader = FileReader(output)
        metadata = reader.metadata

        # THEN
        assert metadata.python_allocator == allocator_name


class TestMemorySnapshots:
    @pytest.mark.valgrind
    def test_memory_snapshots_are_written(self, tmp_path):
        # GIVEN
        allocator = MemoryAllocator()
        output = tmp_path / "test.bin"

        # WHEN
        with Tracker(output):
            allocator.valloc(ALLOC_SIZE)
            time.sleep(0.11)
            allocator.free()

        memory_snapshots = list(FileReader(output).get_memory_snapshots())

        assert memory_snapshots
        assert all(record.rss > 0 for record in memory_snapshots)
        assert any(record.heap >= ALLOC_SIZE for record in memory_snapshots)
        assert sorted(memory_snapshots, key=lambda r: r.time) == memory_snapshots
        assert all(
            _next.time - prev.time >= 10
            for prev, _next in zip(memory_snapshots, memory_snapshots[1:])
        )

    def test_memory_snapshots_tick_interval(self, tmp_path):
        # GIVEN
        allocator = MemoryAllocator()
        output = tmp_path / "test.bin"

        # WHEN
        with Tracker(output, memory_interval_ms=20):
            allocator.valloc(ALLOC_SIZE)
            time.sleep(1)

        memory_snapshots = list(FileReader(output).get_memory_snapshots())

        assert len(memory_snapshots)
        assert all(record.rss > 0 for record in memory_snapshots)
        assert any(record.heap >= ALLOC_SIZE for record in memory_snapshots)
        assert sorted(memory_snapshots, key=lambda r: r.time) == memory_snapshots
        assert all(
            _next.time - prev.time >= 20
            for prev, _next in zip(memory_snapshots, memory_snapshots[1:])
        )

    def test_memory_snapshots_limit_when_reading(self, tmp_path):
        # GIVEN
        allocator = MemoryAllocator()
        output = tmp_path / "test.bin"

        # WHEN
        with Tracker(output):
            for _ in range(2):
                allocator.valloc(ALLOC_SIZE)
                time.sleep(0.11)
                allocator.free()

        reader = FileReader(output)
        memory_snapshots = list(reader.get_memory_snapshots())
        temporal_records = list(reader.get_temporal_allocation_records())

        assert memory_snapshots
        n_snapshots = len(memory_snapshots)
        n_temporal_records = len(temporal_records)

        reader = FileReader(output, max_memory_records=n_snapshots // 2)
        memory_snapshots = list(reader.get_memory_snapshots())
        temporal_records = list(reader.get_temporal_allocation_records())

        assert memory_snapshots
        assert len(memory_snapshots) <= n_snapshots // 2 + 1
        assert len(temporal_records) <= n_temporal_records // 2 + 1

    def test_temporary_allocations_when_filling_vector_without_preallocating(
        self, tmp_path
    ):
        # GIVEN
        output = tmp_path / "test.bin"

        # WHEN
        with Tracker(output):
            elements = fill_cpp_vector(2 << 10)

        # THEN
        reader = FileReader(output)
        temporary_allocations = [
            alloc
            for alloc in reader.get_temporary_allocation_records(threshold=1)
            if __file__ in alloc.stack_trace()[0][1]
        ]
        assert elements == 512
        assert len(temporary_allocations) == 1
        (record,) = temporary_allocations
        assert record.n_allocations == 10
        assert record.allocator == AllocatorType.MALLOC
        assert record.size >= (2 << 10)

    def test_temporary_allocations_when_filling_vector_without_preallocating_small_buffer(
        self, tmp_path
    ):
        # GIVEN
        output = tmp_path / "test.bin"

        # WHEN
        with Tracker(output):
            elements = fill_cpp_vector(2 << 10)

        # THEN
        reader = FileReader(output)
        temporary_allocations = [
            alloc
            for alloc in reader.get_temporary_allocation_records(threshold=0)
            if __file__ in alloc.stack_trace()[0][1]
        ]
        assert elements == 512
        assert len(temporary_allocations) == 1
        (record,) = temporary_allocations
        # With a buffer of 1 we only see the last allocation
        assert record.n_allocations == 1
        assert record.allocator == AllocatorType.MALLOC
        assert record.size == 2 << 10