summaryrefslogtreecommitdiffstats
path: root/src/core/cgroup.c
blob: 34fd2a250c41e7c4cb2c666259409eb8484086a3 (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
4375
4376
4377
4378
4379
4380
4381
4382
4383
4384
4385
4386
4387
4388
4389
4390
4391
4392
4393
4394
4395
4396
4397
4398
4399
4400
4401
4402
4403
4404
4405
4406
4407
4408
4409
4410
4411
4412
4413
4414
4415
4416
4417
4418
4419
4420
4421
4422
4423
4424
4425
4426
4427
4428
4429
4430
4431
4432
4433
4434
4435
4436
4437
4438
4439
4440
4441
4442
4443
4444
4445
4446
4447
4448
4449
4450
4451
4452
4453
4454
4455
4456
4457
4458
4459
4460
4461
4462
4463
4464
4465
4466
4467
4468
4469
4470
4471
4472
4473
4474
4475
4476
4477
4478
4479
4480
4481
4482
4483
4484
4485
4486
4487
4488
4489
4490
4491
4492
4493
4494
4495
4496
4497
4498
4499
4500
4501
4502
4503
4504
4505
4506
4507
4508
4509
4510
4511
4512
4513
4514
4515
4516
4517
4518
4519
4520
4521
4522
4523
4524
4525
4526
4527
4528
4529
4530
4531
4532
4533
4534
4535
4536
4537
4538
4539
4540
4541
4542
4543
4544
4545
4546
4547
4548
4549
4550
4551
4552
4553
4554
4555
4556
4557
4558
4559
4560
4561
4562
4563
4564
4565
4566
4567
4568
4569
4570
4571
4572
4573
4574
4575
4576
4577
4578
4579
4580
4581
4582
4583
4584
4585
4586
4587
4588
4589
4590
4591
4592
4593
4594
4595
4596
4597
4598
4599
4600
4601
4602
4603
4604
4605
4606
4607
4608
4609
4610
4611
4612
4613
4614
4615
4616
4617
4618
4619
4620
4621
4622
4623
4624
4625
4626
4627
4628
4629
4630
4631
4632
4633
4634
4635
4636
4637
4638
4639
4640
4641
4642
4643
4644
4645
4646
4647
4648
4649
4650
4651
4652
4653
4654
4655
4656
4657
4658
4659
4660
4661
4662
4663
4664
4665
4666
4667
4668
4669
4670
4671
4672
4673
4674
4675
4676
4677
4678
4679
4680
4681
4682
4683
4684
4685
4686
4687
4688
4689
4690
4691
4692
4693
4694
4695
4696
4697
4698
4699
4700
4701
4702
4703
4704
4705
4706
4707
4708
4709
4710
4711
4712
4713
4714
4715
4716
4717
4718
4719
4720
4721
4722
4723
4724
4725
4726
4727
4728
4729
4730
4731
4732
4733
4734
4735
4736
4737
4738
4739
4740
4741
4742
4743
4744
4745
4746
4747
4748
4749
4750
4751
4752
4753
4754
4755
4756
4757
4758
4759
4760
4761
4762
4763
4764
4765
4766
4767
4768
4769
4770
4771
4772
4773
4774
4775
4776
4777
4778
4779
4780
4781
4782
4783
4784
4785
4786
4787
4788
4789
4790
4791
4792
4793
4794
4795
4796
4797
4798
4799
4800
4801
4802
4803
4804
4805
4806
4807
4808
4809
4810
4811
4812
4813
4814
4815
4816
4817
4818
4819
4820
4821
4822
4823
4824
4825
4826
4827
4828
4829
4830
4831
4832
4833
4834
4835
4836
4837
4838
4839
4840
4841
4842
4843
4844
4845
4846
4847
4848
4849
4850
4851
4852
4853
4854
4855
4856
4857
4858
4859
4860
4861
4862
4863
4864
4865
4866
4867
4868
4869
4870
4871
4872
4873
4874
4875
4876
4877
4878
4879
4880
4881
4882
4883
4884
4885
4886
4887
4888
4889
4890
4891
4892
4893
4894
4895
4896
4897
4898
4899
4900
4901
4902
4903
4904
4905
4906
4907
4908
4909
4910
4911
4912
4913
4914
4915
4916
4917
4918
4919
4920
4921
4922
4923
4924
4925
4926
4927
4928
4929
4930
4931
4932
4933
4934
4935
4936
4937
4938
4939
4940
4941
4942
4943
4944
4945
4946
4947
4948
4949
4950
4951
4952
4953
4954
4955
4956
4957
4958
4959
4960
4961
4962
4963
4964
4965
4966
4967
4968
4969
4970
4971
4972
4973
4974
4975
4976
4977
4978
4979
4980
4981
4982
4983
4984
4985
4986
4987
4988
4989
4990
4991
4992
4993
4994
4995
4996
4997
4998
4999
5000
5001
5002
5003
5004
5005
5006
5007
5008
5009
5010
5011
5012
5013
5014
5015
5016
5017
5018
5019
5020
5021
5022
5023
5024
5025
5026
5027
5028
5029
5030
5031
5032
5033
5034
5035
5036
5037
5038
5039
5040
5041
5042
5043
5044
5045
5046
5047
5048
5049
5050
5051
5052
5053
5054
5055
5056
5057
5058
5059
5060
5061
5062
5063
5064
5065
5066
5067
5068
5069
5070
5071
5072
5073
5074
5075
5076
5077
5078
5079
5080
5081
5082
5083
5084
5085
5086
5087
5088
5089
5090
5091
5092
5093
5094
5095
5096
5097
5098
5099
5100
5101
5102
5103
5104
5105
5106
5107
5108
5109
5110
5111
5112
5113
5114
5115
5116
5117
5118
5119
5120
5121
5122
5123
5124
5125
5126
5127
5128
5129
5130
5131
5132
5133
5134
5135
5136
5137
5138
5139
5140
5141
5142
5143
5144
5145
5146
5147
5148
5149
5150
5151
5152
5153
5154
5155
5156
5157
5158
5159
5160
5161
5162
5163
5164
5165
5166
5167
5168
5169
5170
5171
5172
5173
5174
5175
5176
5177
5178
5179
5180
5181
5182
5183
5184
5185
5186
5187
5188
5189
5190
5191
5192
5193
5194
5195
5196
5197
5198
5199
5200
5201
5202
5203
5204
5205
5206
5207
5208
5209
5210
5211
5212
5213
5214
5215
5216
5217
5218
5219
5220
5221
5222
5223
5224
5225
5226
5227
5228
5229
5230
5231
5232
5233
5234
5235
5236
5237
5238
5239
5240
5241
5242
5243
5244
5245
5246
5247
5248
5249
5250
5251
5252
5253
5254
5255
5256
5257
5258
5259
5260
5261
5262
5263
5264
5265
5266
5267
5268
5269
5270
5271
5272
5273
5274
5275
5276
5277
5278
5279
5280
5281
5282
5283
5284
5285
5286
5287
5288
5289
5290
5291
5292
5293
5294
5295
5296
5297
5298
5299
5300
5301
5302
5303
5304
5305
5306
5307
5308
5309
5310
5311
5312
5313
5314
5315
5316
5317
5318
5319
5320
5321
5322
5323
5324
5325
5326
5327
5328
5329
5330
5331
5332
5333
5334
5335
5336
5337
5338
5339
5340
5341
5342
5343
5344
5345
5346
5347
5348
5349
5350
5351
5352
5353
5354
5355
5356
5357
5358
5359
5360
5361
5362
5363
5364
5365
5366
5367
5368
5369
5370
5371
5372
5373
5374
5375
5376
5377
5378
5379
5380
5381
5382
5383
5384
5385
5386
5387
5388
5389
5390
5391
5392
5393
5394
5395
5396
5397
5398
5399
5400
5401
5402
5403
5404
5405
5406
5407
5408
5409
5410
5411
5412
5413
5414
5415
5416
5417
5418
5419
5420
5421
5422
5423
5424
5425
5426
5427
5428
5429
5430
5431
5432
5433
5434
5435
5436
5437
5438
5439
5440
5441
5442
5443
5444
5445
5446
5447
5448
5449
5450
5451
5452
5453
5454
5455
5456
5457
5458
5459
5460
5461
5462
5463
5464
5465
5466
5467
5468
5469
5470
5471
5472
5473
5474
5475
5476
5477
5478
5479
5480
5481
5482
5483
5484
5485
5486
5487
5488
5489
5490
5491
5492
5493
5494
5495
5496
5497
5498
5499
5500
5501
5502
5503
5504
5505
5506
5507
5508
5509
5510
5511
5512
5513
5514
5515
5516
5517
5518
5519
5520
5521
5522
5523
5524
5525
5526
5527
5528
5529
5530
5531
5532
5533
5534
5535
5536
5537
5538
5539
5540
5541
5542
5543
5544
5545
5546
5547
5548
5549
5550
5551
5552
5553
5554
5555
5556
5557
5558
5559
5560
5561
5562
5563
5564
5565
5566
5567
5568
5569
5570
5571
5572
5573
5574
5575
5576
5577
5578
5579
5580
5581
5582
5583
5584
5585
5586
5587
5588
5589
5590
5591
5592
5593
5594
5595
5596
5597
5598
5599
5600
5601
5602
5603
5604
5605
5606
5607
5608
5609
5610
5611
5612
5613
5614
5615
5616
5617
5618
5619
5620
5621
5622
5623
5624
5625
5626
5627
5628
5629
5630
5631
5632
5633
5634
5635
5636
5637
5638
5639
5640
5641
5642
5643
5644
5645
5646
5647
5648
5649
5650
5651
5652
5653
5654
5655
5656
5657
5658
5659
5660
5661
5662
5663
5664
5665
5666
5667
5668
5669
5670
5671
5672
5673
5674
5675
5676
5677
5678
5679
/* SPDX-License-Identifier: LGPL-2.1-or-later */

#include <fcntl.h>

#include "sd-messages.h"

#include "af-list.h"
#include "alloc-util.h"
#include "blockdev-util.h"
#include "bpf-devices.h"
#include "bpf-firewall.h"
#include "bpf-foreign.h"
#include "bpf-restrict-ifaces.h"
#include "bpf-socket-bind.h"
#include "btrfs-util.h"
#include "bus-error.h"
#include "bus-locator.h"
#include "cgroup-setup.h"
#include "cgroup-util.h"
#include "cgroup.h"
#include "devnum-util.h"
#include "fd-util.h"
#include "fileio.h"
#include "firewall-util.h"
#include "in-addr-prefix-util.h"
#include "inotify-util.h"
#include "io-util.h"
#include "ip-protocol-list.h"
#include "limits-util.h"
#include "nulstr-util.h"
#include "parse-util.h"
#include "path-util.h"
#include "percent-util.h"
#include "process-util.h"
#include "procfs-util.h"
#include "set.h"
#include "serialize.h"
#include "special.h"
#include "stdio-util.h"
#include "string-table.h"
#include "string-util.h"
#include "virt.h"

#if BPF_FRAMEWORK
#include "bpf-dlopen.h"
#include "bpf-link.h"
#include "bpf/restrict_fs/restrict-fs-skel.h"
#endif

#define CGROUP_CPU_QUOTA_DEFAULT_PERIOD_USEC ((usec_t) 100 * USEC_PER_MSEC)

/* Returns the log level to use when cgroup attribute writes fail. When an attribute is missing or we have access
 * problems we downgrade to LOG_DEBUG. This is supposed to be nice to container managers and kernels which want to mask
 * out specific attributes from us. */
#define LOG_LEVEL_CGROUP_WRITE(r) (IN_SET(abs(r), ENOENT, EROFS, EACCES, EPERM) ? LOG_DEBUG : LOG_WARNING)

uint64_t cgroup_tasks_max_resolve(const CGroupTasksMax *tasks_max) {
        if (tasks_max->scale == 0)
                return tasks_max->value;

        return system_tasks_max_scale(tasks_max->value, tasks_max->scale);
}

bool manager_owns_host_root_cgroup(Manager *m) {
        assert(m);

        /* Returns true if we are managing the root cgroup. Note that it isn't sufficient to just check whether the
         * group root path equals "/" since that will also be the case if CLONE_NEWCGROUP is in the mix. Since there's
         * appears to be no nice way to detect whether we are in a CLONE_NEWCGROUP namespace we instead just check if
         * we run in any kind of container virtualization. */

        if (MANAGER_IS_USER(m))
                return false;

        if (detect_container() > 0)
                return false;

        return empty_or_root(m->cgroup_root);
}

bool unit_has_startup_cgroup_constraints(Unit *u) {
        assert(u);

        /* Returns true if this unit has any directives which apply during
         * startup/shutdown phases. */

        CGroupContext *c;

        c = unit_get_cgroup_context(u);
        if (!c)
                return false;

        return c->startup_cpu_shares != CGROUP_CPU_SHARES_INVALID ||
               c->startup_io_weight != CGROUP_WEIGHT_INVALID ||
               c->startup_blockio_weight != CGROUP_BLKIO_WEIGHT_INVALID ||
               c->startup_cpuset_cpus.set ||
               c->startup_cpuset_mems.set ||
               c->startup_memory_high_set ||
               c->startup_memory_max_set ||
               c->startup_memory_swap_max_set||
               c->startup_memory_zswap_max_set ||
               c->startup_memory_low_set;
}

bool unit_has_host_root_cgroup(Unit *u) {
        assert(u);

        /* Returns whether this unit manages the root cgroup. This will return true if this unit is the root slice and
         * the manager manages the root cgroup. */

        if (!manager_owns_host_root_cgroup(u->manager))
                return false;

        return unit_has_name(u, SPECIAL_ROOT_SLICE);
}

static int set_attribute_and_warn(Unit *u, const char *controller, const char *attribute, const char *value) {
        int r;

        assert(u);

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return -EOWNERDEAD;

        r = cg_set_attribute(controller, crt->cgroup_path, attribute, value);
        if (r < 0)
                log_unit_full_errno(u, LOG_LEVEL_CGROUP_WRITE(r), r, "Failed to set '%s' attribute on '%s' to '%.*s': %m",
                                    strna(attribute), empty_to_root(crt->cgroup_path), (int) strcspn(value, NEWLINE), value);

        return r;
}

static void cgroup_compat_warn(void) {
        static bool cgroup_compat_warned = false;

        if (cgroup_compat_warned)
                return;

        log_warning("cgroup compatibility translation between legacy and unified hierarchy settings activated. "
                    "See cgroup-compat debug messages for details.");

        cgroup_compat_warned = true;
}

#define log_cgroup_compat(unit, fmt, ...) do {                                  \
                cgroup_compat_warn();                                           \
                log_unit_debug(unit, "cgroup-compat: " fmt, ##__VA_ARGS__);     \
        } while (false)

void cgroup_context_init(CGroupContext *c) {
        assert(c);

        /* Initialize everything to the kernel defaults. When initializing a bool member to 'true', make
         * sure to serialize in execute-serialize.c using serialize_bool() instead of
         * serialize_bool_elide(), as sd-executor will initialize here to 'true', but serialize_bool_elide()
         * skips serialization if the value is 'false' (as that's the common default), so if the value at
         * runtime is zero it would be lost after deserialization. Same when initializing uint64_t and other
         * values, update/add a conditional serialization check. This is to minimize the amount of
         * serialized data that is sent to the sd-executor, so that there is less work to do on the default
         * cases. */

        *c = (CGroupContext) {
                .cpu_weight = CGROUP_WEIGHT_INVALID,
                .startup_cpu_weight = CGROUP_WEIGHT_INVALID,
                .cpu_quota_per_sec_usec = USEC_INFINITY,
                .cpu_quota_period_usec = USEC_INFINITY,

                .cpu_shares = CGROUP_CPU_SHARES_INVALID,
                .startup_cpu_shares = CGROUP_CPU_SHARES_INVALID,

                .memory_high = CGROUP_LIMIT_MAX,
                .startup_memory_high = CGROUP_LIMIT_MAX,
                .memory_max = CGROUP_LIMIT_MAX,
                .startup_memory_max = CGROUP_LIMIT_MAX,
                .memory_swap_max = CGROUP_LIMIT_MAX,
                .startup_memory_swap_max = CGROUP_LIMIT_MAX,
                .memory_zswap_max = CGROUP_LIMIT_MAX,
                .startup_memory_zswap_max = CGROUP_LIMIT_MAX,

                .memory_limit = CGROUP_LIMIT_MAX,

                .memory_zswap_writeback = true,

                .io_weight = CGROUP_WEIGHT_INVALID,
                .startup_io_weight = CGROUP_WEIGHT_INVALID,

                .blockio_weight = CGROUP_BLKIO_WEIGHT_INVALID,
                .startup_blockio_weight = CGROUP_BLKIO_WEIGHT_INVALID,

                .tasks_max = CGROUP_TASKS_MAX_UNSET,

                .moom_swap = MANAGED_OOM_AUTO,
                .moom_mem_pressure = MANAGED_OOM_AUTO,
                .moom_preference = MANAGED_OOM_PREFERENCE_NONE,

                .memory_pressure_watch = _CGROUP_PRESSURE_WATCH_INVALID,
                .memory_pressure_threshold_usec = USEC_INFINITY,
        };
}

int cgroup_context_add_io_device_weight_dup(CGroupContext *c, const CGroupIODeviceWeight *w) {
        _cleanup_free_ CGroupIODeviceWeight *n = NULL;

        assert(c);
        assert(w);

        n = new(CGroupIODeviceWeight, 1);
        if (!n)
                return -ENOMEM;

        *n = (CGroupIODeviceWeight) {
                .path = strdup(w->path),
                .weight = w->weight,
        };
        if (!n->path)
                return -ENOMEM;

        LIST_PREPEND(device_weights, c->io_device_weights, TAKE_PTR(n));
        return 0;
}

int cgroup_context_add_io_device_limit_dup(CGroupContext *c, const CGroupIODeviceLimit *l) {
        _cleanup_free_ CGroupIODeviceLimit *n = NULL;

        assert(c);
        assert(l);

        n = new0(CGroupIODeviceLimit, 1);
        if (!n)
                return -ENOMEM;

        n->path = strdup(l->path);
        if (!n->path)
                return -ENOMEM;

        for (CGroupIOLimitType type = 0; type < _CGROUP_IO_LIMIT_TYPE_MAX; type++)
                n->limits[type] = l->limits[type];

        LIST_PREPEND(device_limits, c->io_device_limits, TAKE_PTR(n));
        return 0;
}

int cgroup_context_add_io_device_latency_dup(CGroupContext *c, const CGroupIODeviceLatency *l) {
        _cleanup_free_ CGroupIODeviceLatency *n = NULL;

        assert(c);
        assert(l);

        n = new(CGroupIODeviceLatency, 1);
        if (!n)
                return -ENOMEM;

        *n = (CGroupIODeviceLatency) {
                .path = strdup(l->path),
                .target_usec = l->target_usec,
        };
        if (!n->path)
                return -ENOMEM;

        LIST_PREPEND(device_latencies, c->io_device_latencies, TAKE_PTR(n));
        return 0;
}

int cgroup_context_add_block_io_device_weight_dup(CGroupContext *c, const CGroupBlockIODeviceWeight *w) {
        _cleanup_free_ CGroupBlockIODeviceWeight *n = NULL;

        assert(c);
        assert(w);

        n = new(CGroupBlockIODeviceWeight, 1);
        if (!n)
                return -ENOMEM;

        *n = (CGroupBlockIODeviceWeight) {
                .path = strdup(w->path),
                .weight = w->weight,
        };
        if (!n->path)
                return -ENOMEM;

        LIST_PREPEND(device_weights, c->blockio_device_weights, TAKE_PTR(n));
        return 0;
}

int cgroup_context_add_block_io_device_bandwidth_dup(CGroupContext *c, const CGroupBlockIODeviceBandwidth *b) {
        _cleanup_free_ CGroupBlockIODeviceBandwidth *n = NULL;

        assert(c);
        assert(b);

        n = new(CGroupBlockIODeviceBandwidth, 1);
        if (!n)
                return -ENOMEM;

        *n = (CGroupBlockIODeviceBandwidth) {
                .rbps = b->rbps,
                .wbps = b->wbps,
        };

        LIST_PREPEND(device_bandwidths, c->blockio_device_bandwidths, TAKE_PTR(n));
        return 0;
}

int cgroup_context_add_device_allow_dup(CGroupContext *c, const CGroupDeviceAllow *a) {
        _cleanup_free_ CGroupDeviceAllow *n = NULL;

        assert(c);
        assert(a);

        n = new(CGroupDeviceAllow, 1);
        if (!n)
                return -ENOMEM;

        *n = (CGroupDeviceAllow) {
                .path = strdup(a->path),
                .permissions = a->permissions,
        };
        if (!n->path)
                return -ENOMEM;

        LIST_PREPEND(device_allow, c->device_allow, TAKE_PTR(n));
        return 0;
}

static int cgroup_context_add_socket_bind_item_dup(CGroupContext *c, const CGroupSocketBindItem *i, CGroupSocketBindItem *h) {
        _cleanup_free_ CGroupSocketBindItem *n = NULL;

        assert(c);
        assert(i);

        n = new(CGroupSocketBindItem, 1);
        if (!n)
                return -ENOMEM;

        *n = (CGroupSocketBindItem) {
                .address_family = i->address_family,
                .ip_protocol    = i->ip_protocol,
                .nr_ports       = i->nr_ports,
                .port_min       = i->port_min,
        };

        LIST_PREPEND(socket_bind_items, h, TAKE_PTR(n));
        return 0;
}

int cgroup_context_add_socket_bind_item_allow_dup(CGroupContext *c, const CGroupSocketBindItem *i) {
        return cgroup_context_add_socket_bind_item_dup(c, i, c->socket_bind_allow);
}

int cgroup_context_add_socket_bind_item_deny_dup(CGroupContext *c, const CGroupSocketBindItem *i) {
        return cgroup_context_add_socket_bind_item_dup(c, i, c->socket_bind_deny);
}

int cgroup_context_copy(CGroupContext *dst, const CGroupContext *src) {
        struct in_addr_prefix *i;
        char *iface;
        int r;

        assert(src);
        assert(dst);

        dst->cpu_accounting = src->cpu_accounting;
        dst->io_accounting = src->io_accounting;
        dst->blockio_accounting = src->blockio_accounting;
        dst->memory_accounting = src->memory_accounting;
        dst->tasks_accounting = src->tasks_accounting;
        dst->ip_accounting = src->ip_accounting;

        dst->memory_oom_group = src->memory_oom_group;

        dst->cpu_weight = src->cpu_weight;
        dst->startup_cpu_weight = src->startup_cpu_weight;
        dst->cpu_quota_per_sec_usec = src->cpu_quota_per_sec_usec;
        dst->cpu_quota_period_usec = src->cpu_quota_period_usec;

        dst->cpuset_cpus = src->cpuset_cpus;
        dst->startup_cpuset_cpus = src->startup_cpuset_cpus;
        dst->cpuset_mems = src->cpuset_mems;
        dst->startup_cpuset_mems = src->startup_cpuset_mems;

        dst->io_weight = src->io_weight;
        dst->startup_io_weight = src->startup_io_weight;

        LIST_FOREACH_BACKWARDS(device_weights, w, LIST_FIND_TAIL(device_weights, src->io_device_weights)) {
                r = cgroup_context_add_io_device_weight_dup(dst, w);
                if (r < 0)
                        return r;
        }

        LIST_FOREACH_BACKWARDS(device_limits, l, LIST_FIND_TAIL(device_limits, src->io_device_limits)) {
                r = cgroup_context_add_io_device_limit_dup(dst, l);
                if (r < 0)
                        return r;
        }

        LIST_FOREACH_BACKWARDS(device_latencies, l, LIST_FIND_TAIL(device_latencies, src->io_device_latencies)) {
                r = cgroup_context_add_io_device_latency_dup(dst, l);
                if (r < 0)
                        return r;
        }

        dst->default_memory_min = src->default_memory_min;
        dst->default_memory_low = src->default_memory_low;
        dst->default_startup_memory_low = src->default_startup_memory_low;
        dst->memory_min = src->memory_min;
        dst->memory_low = src->memory_low;
        dst->startup_memory_low = src->startup_memory_low;
        dst->memory_high = src->memory_high;
        dst->startup_memory_high = src->startup_memory_high;
        dst->memory_max = src->memory_max;
        dst->startup_memory_max = src->startup_memory_max;
        dst->memory_swap_max = src->memory_swap_max;
        dst->startup_memory_swap_max = src->startup_memory_swap_max;
        dst->memory_zswap_max = src->memory_zswap_max;
        dst->startup_memory_zswap_max = src->startup_memory_zswap_max;

        dst->default_memory_min_set = src->default_memory_min_set;
        dst->default_memory_low_set = src->default_memory_low_set;
        dst->default_startup_memory_low_set = src->default_startup_memory_low_set;
        dst->memory_min_set = src->memory_min_set;
        dst->memory_low_set = src->memory_low_set;
        dst->startup_memory_low_set = src->startup_memory_low_set;
        dst->startup_memory_high_set = src->startup_memory_high_set;
        dst->startup_memory_max_set = src->startup_memory_max_set;
        dst->startup_memory_swap_max_set = src->startup_memory_swap_max_set;
        dst->startup_memory_zswap_max_set = src->startup_memory_zswap_max_set;
        dst->memory_zswap_writeback = src->memory_zswap_writeback;

        SET_FOREACH(i, src->ip_address_allow) {
                r = in_addr_prefix_add(&dst->ip_address_allow, i);
                if (r < 0)
                        return r;
        }

        SET_FOREACH(i, src->ip_address_deny) {
                r = in_addr_prefix_add(&dst->ip_address_deny, i);
                if (r < 0)
                        return r;
        }

        dst->ip_address_allow_reduced = src->ip_address_allow_reduced;
        dst->ip_address_deny_reduced = src->ip_address_deny_reduced;

        if (!strv_isempty(src->ip_filters_ingress)) {
                dst->ip_filters_ingress = strv_copy(src->ip_filters_ingress);
                if (!dst->ip_filters_ingress)
                        return -ENOMEM;
        }

        if (!strv_isempty(src->ip_filters_egress)) {
                dst->ip_filters_egress = strv_copy(src->ip_filters_egress);
                if (!dst->ip_filters_egress)
                        return -ENOMEM;
        }

        LIST_FOREACH_BACKWARDS(programs, l, LIST_FIND_TAIL(programs, src->bpf_foreign_programs)) {
                r = cgroup_context_add_bpf_foreign_program_dup(dst, l);
                if (r < 0)
                        return r;
        }

        SET_FOREACH(iface, src->restrict_network_interfaces) {
                r = set_put_strdup(&dst->restrict_network_interfaces, iface);
                if (r < 0)
                        return r;
        }
        dst->restrict_network_interfaces_is_allow_list = src->restrict_network_interfaces_is_allow_list;

        dst->cpu_shares = src->cpu_shares;
        dst->startup_cpu_shares = src->startup_cpu_shares;

        dst->blockio_weight = src->blockio_weight;
        dst->startup_blockio_weight = src->startup_blockio_weight;

        LIST_FOREACH_BACKWARDS(device_weights, l, LIST_FIND_TAIL(device_weights, src->blockio_device_weights)) {
                r = cgroup_context_add_block_io_device_weight_dup(dst, l);
                if (r < 0)
                        return r;
        }

        LIST_FOREACH_BACKWARDS(device_bandwidths, l, LIST_FIND_TAIL(device_bandwidths, src->blockio_device_bandwidths)) {
                r = cgroup_context_add_block_io_device_bandwidth_dup(dst, l);
                if (r < 0)
                        return r;
        }

        dst->memory_limit = src->memory_limit;

        dst->device_policy = src->device_policy;
        LIST_FOREACH_BACKWARDS(device_allow, l, LIST_FIND_TAIL(device_allow, src->device_allow)) {
                r = cgroup_context_add_device_allow_dup(dst, l);
                if (r < 0)
                        return r;
        }

        LIST_FOREACH_BACKWARDS(socket_bind_items, l, LIST_FIND_TAIL(socket_bind_items, src->socket_bind_allow)) {
                r = cgroup_context_add_socket_bind_item_allow_dup(dst, l);
                if (r < 0)
                        return r;

        }

        LIST_FOREACH_BACKWARDS(socket_bind_items, l, LIST_FIND_TAIL(socket_bind_items, src->socket_bind_deny)) {
                r = cgroup_context_add_socket_bind_item_deny_dup(dst, l);
                if (r < 0)
                        return r;
        }

        dst->tasks_max = src->tasks_max;

        return 0;
}

void cgroup_context_free_device_allow(CGroupContext *c, CGroupDeviceAllow *a) {
        assert(c);
        assert(a);

        LIST_REMOVE(device_allow, c->device_allow, a);
        free(a->path);
        free(a);
}

void cgroup_context_free_io_device_weight(CGroupContext *c, CGroupIODeviceWeight *w) {
        assert(c);
        assert(w);

        LIST_REMOVE(device_weights, c->io_device_weights, w);
        free(w->path);
        free(w);
}

void cgroup_context_free_io_device_latency(CGroupContext *c, CGroupIODeviceLatency *l) {
        assert(c);
        assert(l);

        LIST_REMOVE(device_latencies, c->io_device_latencies, l);
        free(l->path);
        free(l);
}

void cgroup_context_free_io_device_limit(CGroupContext *c, CGroupIODeviceLimit *l) {
        assert(c);
        assert(l);

        LIST_REMOVE(device_limits, c->io_device_limits, l);
        free(l->path);
        free(l);
}

void cgroup_context_free_blockio_device_weight(CGroupContext *c, CGroupBlockIODeviceWeight *w) {
        assert(c);
        assert(w);

        LIST_REMOVE(device_weights, c->blockio_device_weights, w);
        free(w->path);
        free(w);
}

void cgroup_context_free_blockio_device_bandwidth(CGroupContext *c, CGroupBlockIODeviceBandwidth *b) {
        assert(c);
        assert(b);

        LIST_REMOVE(device_bandwidths, c->blockio_device_bandwidths, b);
        free(b->path);
        free(b);
}

void cgroup_context_remove_bpf_foreign_program(CGroupContext *c, CGroupBPFForeignProgram *p) {
        assert(c);
        assert(p);

        LIST_REMOVE(programs, c->bpf_foreign_programs, p);
        free(p->bpffs_path);
        free(p);
}

void cgroup_context_remove_socket_bind(CGroupSocketBindItem **head) {
        assert(head);

        LIST_CLEAR(socket_bind_items, *head, free);
}

void cgroup_context_done(CGroupContext *c) {
        assert(c);

        while (c->io_device_weights)
                cgroup_context_free_io_device_weight(c, c->io_device_weights);

        while (c->io_device_latencies)
                cgroup_context_free_io_device_latency(c, c->io_device_latencies);

        while (c->io_device_limits)
                cgroup_context_free_io_device_limit(c, c->io_device_limits);

        while (c->blockio_device_weights)
                cgroup_context_free_blockio_device_weight(c, c->blockio_device_weights);

        while (c->blockio_device_bandwidths)
                cgroup_context_free_blockio_device_bandwidth(c, c->blockio_device_bandwidths);

        while (c->device_allow)
                cgroup_context_free_device_allow(c, c->device_allow);

        cgroup_context_remove_socket_bind(&c->socket_bind_allow);
        cgroup_context_remove_socket_bind(&c->socket_bind_deny);

        c->ip_address_allow = set_free(c->ip_address_allow);
        c->ip_address_deny = set_free(c->ip_address_deny);

        c->ip_filters_ingress = strv_free(c->ip_filters_ingress);
        c->ip_filters_egress = strv_free(c->ip_filters_egress);

        while (c->bpf_foreign_programs)
                cgroup_context_remove_bpf_foreign_program(c, c->bpf_foreign_programs);

        c->restrict_network_interfaces = set_free_free(c->restrict_network_interfaces);

        cpu_set_reset(&c->cpuset_cpus);
        cpu_set_reset(&c->startup_cpuset_cpus);
        cpu_set_reset(&c->cpuset_mems);
        cpu_set_reset(&c->startup_cpuset_mems);

        c->delegate_subgroup = mfree(c->delegate_subgroup);

        nft_set_context_clear(&c->nft_set_context);
}

static int unit_get_kernel_memory_limit(Unit *u, const char *file, uint64_t *ret) {
        assert(u);

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return -EOWNERDEAD;

        return cg_get_attribute_as_uint64("memory", crt->cgroup_path, file, ret);
}

static int unit_compare_memory_limit(Unit *u, const char *property_name, uint64_t *ret_unit_value, uint64_t *ret_kernel_value) {
        CGroupContext *c;
        CGroupMask m;
        const char *file;
        uint64_t unit_value;
        int r;

        /* Compare kernel memcg configuration against our internal systemd state. Unsupported (and will
         * return -ENODATA) on cgroup v1.
         *
         * Returns:
         *
         * <0: On error.
         *  0: If the kernel memory setting doesn't match our configuration.
         * >0: If the kernel memory setting matches our configuration.
         *
         * The following values are only guaranteed to be populated on return >=0:
         *
         * - ret_unit_value will contain our internal expected value for the unit, page-aligned.
         * - ret_kernel_value will contain the actual value presented by the kernel. */

        assert(u);

        r = cg_all_unified();
        if (r < 0)
                return log_debug_errno(r, "Failed to determine cgroup hierarchy version: %m");

        /* Unsupported on v1.
         *
         * We don't return ENOENT, since that could actually mask a genuine problem where somebody else has
         * silently masked the controller. */
        if (r == 0)
                return -ENODATA;

        /* The root slice doesn't have any controller files, so we can't compare anything. */
        if (unit_has_name(u, SPECIAL_ROOT_SLICE))
                return -ENODATA;

        /* It's possible to have MemoryFoo set without systemd wanting to have the memory controller enabled,
         * for example, in the case of DisableControllers= or cgroup_disable on the kernel command line. To
         * avoid specious errors in these scenarios, check that we even expect the memory controller to be
         * enabled at all. */
        m = unit_get_target_mask(u);
        if (!FLAGS_SET(m, CGROUP_MASK_MEMORY))
                return -ENODATA;

        assert_se(c = unit_get_cgroup_context(u));

        bool startup = u->manager && IN_SET(manager_state(u->manager), MANAGER_STARTING, MANAGER_INITIALIZING, MANAGER_STOPPING);

        if (streq(property_name, "MemoryLow")) {
                unit_value = unit_get_ancestor_memory_low(u);
                file = "memory.low";
        } else if (startup && streq(property_name, "StartupMemoryLow")) {
                unit_value = unit_get_ancestor_startup_memory_low(u);
                file = "memory.low";
        } else if (streq(property_name, "MemoryMin")) {
                unit_value = unit_get_ancestor_memory_min(u);
                file = "memory.min";
        } else if (streq(property_name, "MemoryHigh")) {
                unit_value = c->memory_high;
                file = "memory.high";
        } else if (startup && streq(property_name, "StartupMemoryHigh")) {
                unit_value = c->startup_memory_high;
                file = "memory.high";
        } else if (streq(property_name, "MemoryMax")) {
                unit_value = c->memory_max;
                file = "memory.max";
        } else if (startup && streq(property_name, "StartupMemoryMax")) {
                unit_value = c->startup_memory_max;
                file = "memory.max";
        } else if (streq(property_name, "MemorySwapMax")) {
                unit_value = c->memory_swap_max;
                file = "memory.swap.max";
        } else if (startup && streq(property_name, "StartupMemorySwapMax")) {
                unit_value = c->startup_memory_swap_max;
                file = "memory.swap.max";
        } else if (streq(property_name, "MemoryZSwapMax")) {
                unit_value = c->memory_zswap_max;
                file = "memory.zswap.max";
        } else if (startup && streq(property_name, "StartupMemoryZSwapMax")) {
                unit_value = c->startup_memory_zswap_max;
                file = "memory.zswap.max";
        } else
                return -EINVAL;

        r = unit_get_kernel_memory_limit(u, file, ret_kernel_value);
        if (r < 0)
                return log_unit_debug_errno(u, r, "Failed to parse %s: %m", file);

        /* It's intended (soon) in a future kernel to not expose cgroup memory limits rounded to page
         * boundaries, but instead separate the user-exposed limit, which is whatever userspace told us, from
         * our internal page-counting. To support those future kernels, just check the value itself first
         * without any page-alignment. */
        if (*ret_kernel_value == unit_value) {
                *ret_unit_value = unit_value;
                return 1;
        }

        /* The current kernel behaviour, by comparison, is that even if you write a particular number of
         * bytes into a cgroup memory file, it always returns that number page-aligned down (since the kernel
         * internally stores cgroup limits in pages). As such, so long as it aligns properly, everything is
         * cricket. */
        if (unit_value != CGROUP_LIMIT_MAX)
                unit_value = PAGE_ALIGN_DOWN(unit_value);

        *ret_unit_value = unit_value;

        return *ret_kernel_value == *ret_unit_value;
}

#define FORMAT_CGROUP_DIFF_MAX 128

static char *format_cgroup_memory_limit_comparison(Unit *u, const char *property_name, char *buf, size_t l) {
        uint64_t kval, sval;
        int r;

        assert(u);
        assert(property_name);
        assert(buf);
        assert(l > 0);

        r = unit_compare_memory_limit(u, property_name, &sval, &kval);

        /* memory.swap.max is special in that it relies on CONFIG_MEMCG_SWAP (and the default swapaccount=1).
         * In the absence of reliably being able to detect whether memcg swap support is available or not,
         * only complain if the error is not ENOENT. This is similarly the case for memory.zswap.max relying
         * on CONFIG_ZSWAP. */
        if (r > 0 || IN_SET(r, -ENODATA, -EOWNERDEAD) ||
            (r == -ENOENT && STR_IN_SET(property_name,
                                        "MemorySwapMax",
                                        "StartupMemorySwapMax",
                                        "MemoryZSwapMax",
                                        "StartupMemoryZSwapMax")))
                buf[0] = 0;
        else if (r < 0) {
                errno = -r;
                (void) snprintf(buf, l, " (error getting kernel value: %m)");
        } else
                (void) snprintf(buf, l, " (different value in kernel: %" PRIu64 ")", kval);

        return buf;
}

const char *cgroup_device_permissions_to_string(CGroupDevicePermissions p) {
        static const char *table[_CGROUP_DEVICE_PERMISSIONS_MAX] = {
                /* Lets simply define a table with every possible combination. As long as those are just 8 we
                 * can get away with it. If this ever grows to more we need to revisit this logic though. */
                [0]                                                          = "",
                [CGROUP_DEVICE_READ]                                         = "r",
                [CGROUP_DEVICE_WRITE]                                        = "w",
                [CGROUP_DEVICE_MKNOD]                                        = "m",
                [CGROUP_DEVICE_READ|CGROUP_DEVICE_WRITE]                     = "rw",
                [CGROUP_DEVICE_READ|CGROUP_DEVICE_MKNOD]                     = "rm",
                [CGROUP_DEVICE_WRITE|CGROUP_DEVICE_MKNOD]                    = "wm",
                [CGROUP_DEVICE_READ|CGROUP_DEVICE_WRITE|CGROUP_DEVICE_MKNOD] = "rwm",
        };

        if (p < 0 || p >= _CGROUP_DEVICE_PERMISSIONS_MAX)
                return NULL;

        return table[p];
}

CGroupDevicePermissions cgroup_device_permissions_from_string(const char *s) {
        CGroupDevicePermissions p = 0;

        if (!s)
                return _CGROUP_DEVICE_PERMISSIONS_INVALID;

        for (const char *c = s; *c; c++) {
                if (*c == 'r')
                        p |= CGROUP_DEVICE_READ;
                else if (*c == 'w')
                        p |= CGROUP_DEVICE_WRITE;
                else if (*c == 'm')
                        p |= CGROUP_DEVICE_MKNOD;
                else
                        return _CGROUP_DEVICE_PERMISSIONS_INVALID;
        }

        return p;
}

void cgroup_context_dump(Unit *u, FILE* f, const char *prefix) {
        _cleanup_free_ char *disable_controllers_str = NULL, *delegate_controllers_str = NULL, *cpuset_cpus = NULL, *cpuset_mems = NULL, *startup_cpuset_cpus = NULL, *startup_cpuset_mems = NULL;
        CGroupContext *c;
        struct in_addr_prefix *iaai;
        char cda[FORMAT_CGROUP_DIFF_MAX], cdb[FORMAT_CGROUP_DIFF_MAX], cdc[FORMAT_CGROUP_DIFF_MAX], cdd[FORMAT_CGROUP_DIFF_MAX],
                cde[FORMAT_CGROUP_DIFF_MAX], cdf[FORMAT_CGROUP_DIFF_MAX], cdg[FORMAT_CGROUP_DIFF_MAX], cdh[FORMAT_CGROUP_DIFF_MAX],
                cdi[FORMAT_CGROUP_DIFF_MAX], cdj[FORMAT_CGROUP_DIFF_MAX], cdk[FORMAT_CGROUP_DIFF_MAX];

        assert(u);
        assert(f);

        assert_se(c = unit_get_cgroup_context(u));

        prefix = strempty(prefix);

        (void) cg_mask_to_string(c->disable_controllers, &disable_controllers_str);
        (void) cg_mask_to_string(c->delegate_controllers, &delegate_controllers_str);

        /* "Delegate=" means "yes, but no controllers". Show this as "(none)". */
        const char *delegate_str = delegate_controllers_str ?: c->delegate ? "(none)" : "no";

        cpuset_cpus = cpu_set_to_range_string(&c->cpuset_cpus);
        startup_cpuset_cpus = cpu_set_to_range_string(&c->startup_cpuset_cpus);
        cpuset_mems = cpu_set_to_range_string(&c->cpuset_mems);
        startup_cpuset_mems = cpu_set_to_range_string(&c->startup_cpuset_mems);

        fprintf(f,
                "%sCPUAccounting: %s\n"
                "%sIOAccounting: %s\n"
                "%sBlockIOAccounting: %s\n"
                "%sMemoryAccounting: %s\n"
                "%sTasksAccounting: %s\n"
                "%sIPAccounting: %s\n"
                "%sCPUWeight: %" PRIu64 "\n"
                "%sStartupCPUWeight: %" PRIu64 "\n"
                "%sCPUShares: %" PRIu64 "\n"
                "%sStartupCPUShares: %" PRIu64 "\n"
                "%sCPUQuotaPerSecSec: %s\n"
                "%sCPUQuotaPeriodSec: %s\n"
                "%sAllowedCPUs: %s\n"
                "%sStartupAllowedCPUs: %s\n"
                "%sAllowedMemoryNodes: %s\n"
                "%sStartupAllowedMemoryNodes: %s\n"
                "%sIOWeight: %" PRIu64 "\n"
                "%sStartupIOWeight: %" PRIu64 "\n"
                "%sBlockIOWeight: %" PRIu64 "\n"
                "%sStartupBlockIOWeight: %" PRIu64 "\n"
                "%sDefaultMemoryMin: %" PRIu64 "\n"
                "%sDefaultMemoryLow: %" PRIu64 "\n"
                "%sMemoryMin: %" PRIu64 "%s\n"
                "%sMemoryLow: %" PRIu64 "%s\n"
                "%sStartupMemoryLow: %" PRIu64 "%s\n"
                "%sMemoryHigh: %" PRIu64 "%s\n"
                "%sStartupMemoryHigh: %" PRIu64 "%s\n"
                "%sMemoryMax: %" PRIu64 "%s\n"
                "%sStartupMemoryMax: %" PRIu64 "%s\n"
                "%sMemorySwapMax: %" PRIu64 "%s\n"
                "%sStartupMemorySwapMax: %" PRIu64 "%s\n"
                "%sMemoryZSwapMax: %" PRIu64 "%s\n"
                "%sStartupMemoryZSwapMax: %" PRIu64 "%s\n"
                "%sMemoryZSwapWriteback: %s\n"
                "%sMemoryLimit: %" PRIu64 "\n"
                "%sTasksMax: %" PRIu64 "\n"
                "%sDevicePolicy: %s\n"
                "%sDisableControllers: %s\n"
                "%sDelegate: %s\n"
                "%sManagedOOMSwap: %s\n"
                "%sManagedOOMMemoryPressure: %s\n"
                "%sManagedOOMMemoryPressureLimit: " PERMYRIAD_AS_PERCENT_FORMAT_STR "\n"
                "%sManagedOOMPreference: %s\n"
                "%sMemoryPressureWatch: %s\n"
                "%sCoredumpReceive: %s\n",
                prefix, yes_no(c->cpu_accounting),
                prefix, yes_no(c->io_accounting),
                prefix, yes_no(c->blockio_accounting),
                prefix, yes_no(c->memory_accounting),
                prefix, yes_no(c->tasks_accounting),
                prefix, yes_no(c->ip_accounting),
                prefix, c->cpu_weight,
                prefix, c->startup_cpu_weight,
                prefix, c->cpu_shares,
                prefix, c->startup_cpu_shares,
                prefix, FORMAT_TIMESPAN(c->cpu_quota_per_sec_usec, 1),
                prefix, FORMAT_TIMESPAN(c->cpu_quota_period_usec, 1),
                prefix, strempty(cpuset_cpus),
                prefix, strempty(startup_cpuset_cpus),
                prefix, strempty(cpuset_mems),
                prefix, strempty(startup_cpuset_mems),
                prefix, c->io_weight,
                prefix, c->startup_io_weight,
                prefix, c->blockio_weight,
                prefix, c->startup_blockio_weight,
                prefix, c->default_memory_min,
                prefix, c->default_memory_low,
                prefix, c->memory_min, format_cgroup_memory_limit_comparison(u, "MemoryMin", cda, sizeof(cda)),
                prefix, c->memory_low, format_cgroup_memory_limit_comparison(u, "MemoryLow", cdb, sizeof(cdb)),
                prefix, c->startup_memory_low, format_cgroup_memory_limit_comparison(u, "StartupMemoryLow", cdc, sizeof(cdc)),
                prefix, c->memory_high, format_cgroup_memory_limit_comparison(u, "MemoryHigh", cdd, sizeof(cdd)),
                prefix, c->startup_memory_high, format_cgroup_memory_limit_comparison(u, "StartupMemoryHigh", cde, sizeof(cde)),
                prefix, c->memory_max, format_cgroup_memory_limit_comparison(u, "MemoryMax", cdf, sizeof(cdf)),
                prefix, c->startup_memory_max, format_cgroup_memory_limit_comparison(u, "StartupMemoryMax", cdg, sizeof(cdg)),
                prefix, c->memory_swap_max, format_cgroup_memory_limit_comparison(u, "MemorySwapMax", cdh, sizeof(cdh)),
                prefix, c->startup_memory_swap_max, format_cgroup_memory_limit_comparison(u, "StartupMemorySwapMax", cdi, sizeof(cdi)),
                prefix, c->memory_zswap_max, format_cgroup_memory_limit_comparison(u, "MemoryZSwapMax", cdj, sizeof(cdj)),
                prefix, c->startup_memory_zswap_max, format_cgroup_memory_limit_comparison(u, "StartupMemoryZSwapMax", cdk, sizeof(cdk)),
                prefix, yes_no(c->memory_zswap_writeback),
                prefix, c->memory_limit,
                prefix, cgroup_tasks_max_resolve(&c->tasks_max),
                prefix, cgroup_device_policy_to_string(c->device_policy),
                prefix, strempty(disable_controllers_str),
                prefix, delegate_str,
                prefix, managed_oom_mode_to_string(c->moom_swap),
                prefix, managed_oom_mode_to_string(c->moom_mem_pressure),
                prefix, PERMYRIAD_AS_PERCENT_FORMAT_VAL(UINT32_SCALE_TO_PERMYRIAD(c->moom_mem_pressure_limit)),
                prefix, managed_oom_preference_to_string(c->moom_preference),
                prefix, cgroup_pressure_watch_to_string(c->memory_pressure_watch),
                prefix, yes_no(c->coredump_receive));

        if (c->delegate_subgroup)
                fprintf(f, "%sDelegateSubgroup: %s\n",
                        prefix, c->delegate_subgroup);

        if (c->memory_pressure_threshold_usec != USEC_INFINITY)
                fprintf(f, "%sMemoryPressureThresholdSec: %s\n",
                        prefix, FORMAT_TIMESPAN(c->memory_pressure_threshold_usec, 1));

        LIST_FOREACH(device_allow, a, c->device_allow)
                /* strna() below should be redundant, for avoiding -Werror=format-overflow= error. See #30223. */
                fprintf(f,
                        "%sDeviceAllow: %s %s\n",
                        prefix,
                        a->path,
                        strna(cgroup_device_permissions_to_string(a->permissions)));

        LIST_FOREACH(device_weights, iw, c->io_device_weights)
                fprintf(f,
                        "%sIODeviceWeight: %s %" PRIu64 "\n",
                        prefix,
                        iw->path,
                        iw->weight);

        LIST_FOREACH(device_latencies, l, c->io_device_latencies)
                fprintf(f,
                        "%sIODeviceLatencyTargetSec: %s %s\n",
                        prefix,
                        l->path,
                        FORMAT_TIMESPAN(l->target_usec, 1));

        LIST_FOREACH(device_limits, il, c->io_device_limits)
                for (CGroupIOLimitType type = 0; type < _CGROUP_IO_LIMIT_TYPE_MAX; type++)
                        if (il->limits[type] != cgroup_io_limit_defaults[type])
                                fprintf(f,
                                        "%s%s: %s %s\n",
                                        prefix,
                                        cgroup_io_limit_type_to_string(type),
                                        il->path,
                                        FORMAT_BYTES(il->limits[type]));

        LIST_FOREACH(device_weights, w, c->blockio_device_weights)
                fprintf(f,
                        "%sBlockIODeviceWeight: %s %" PRIu64,
                        prefix,
                        w->path,
                        w->weight);

        LIST_FOREACH(device_bandwidths, b, c->blockio_device_bandwidths) {
                if (b->rbps != CGROUP_LIMIT_MAX)
                        fprintf(f,
                                "%sBlockIOReadBandwidth: %s %s\n",
                                prefix,
                                b->path,
                                FORMAT_BYTES(b->rbps));
                if (b->wbps != CGROUP_LIMIT_MAX)
                        fprintf(f,
                                "%sBlockIOWriteBandwidth: %s %s\n",
                                prefix,
                                b->path,
                                FORMAT_BYTES(b->wbps));
        }

        SET_FOREACH(iaai, c->ip_address_allow)
                fprintf(f, "%sIPAddressAllow: %s\n", prefix,
                        IN_ADDR_PREFIX_TO_STRING(iaai->family, &iaai->address, iaai->prefixlen));
        SET_FOREACH(iaai, c->ip_address_deny)
                fprintf(f, "%sIPAddressDeny: %s\n", prefix,
                        IN_ADDR_PREFIX_TO_STRING(iaai->family, &iaai->address, iaai->prefixlen));

        STRV_FOREACH(path, c->ip_filters_ingress)
                fprintf(f, "%sIPIngressFilterPath: %s\n", prefix, *path);
        STRV_FOREACH(path, c->ip_filters_egress)
                fprintf(f, "%sIPEgressFilterPath: %s\n", prefix, *path);

        LIST_FOREACH(programs, p, c->bpf_foreign_programs)
                fprintf(f, "%sBPFProgram: %s:%s",
                        prefix, bpf_cgroup_attach_type_to_string(p->attach_type), p->bpffs_path);

        if (c->socket_bind_allow) {
                fprintf(f, "%sSocketBindAllow: ", prefix);
                cgroup_context_dump_socket_bind_items(c->socket_bind_allow, f);
                fputc('\n', f);
        }

        if (c->socket_bind_deny) {
                fprintf(f, "%sSocketBindDeny: ", prefix);
                cgroup_context_dump_socket_bind_items(c->socket_bind_deny, f);
                fputc('\n', f);
        }

        if (c->restrict_network_interfaces) {
                char *iface;
                SET_FOREACH(iface, c->restrict_network_interfaces)
                        fprintf(f, "%sRestrictNetworkInterfaces: %s\n", prefix, iface);
        }

        FOREACH_ARRAY(nft_set, c->nft_set_context.sets, c->nft_set_context.n_sets)
                fprintf(f, "%sNFTSet: %s:%s:%s:%s\n", prefix, nft_set_source_to_string(nft_set->source),
                        nfproto_to_string(nft_set->nfproto), nft_set->table, nft_set->set);
}

void cgroup_context_dump_socket_bind_item(const CGroupSocketBindItem *item, FILE *f) {
        const char *family, *colon1, *protocol = "", *colon2 = "";

        family = strempty(af_to_ipv4_ipv6(item->address_family));
        colon1 = isempty(family) ? "" : ":";

        if (item->ip_protocol != 0) {
                protocol = ip_protocol_to_tcp_udp(item->ip_protocol);
                colon2 = ":";
        }

        if (item->nr_ports == 0)
                fprintf(f, "%s%s%s%sany", family, colon1, protocol, colon2);
        else if (item->nr_ports == 1)
                fprintf(f, "%s%s%s%s%" PRIu16, family, colon1, protocol, colon2, item->port_min);
        else {
                uint16_t port_max = item->port_min + item->nr_ports - 1;
                fprintf(f, "%s%s%s%s%" PRIu16 "-%" PRIu16, family, colon1, protocol, colon2,
                        item->port_min, port_max);
        }
}

void cgroup_context_dump_socket_bind_items(const CGroupSocketBindItem *items, FILE *f) {
        bool first = true;

        LIST_FOREACH(socket_bind_items, bi, items) {
                if (first)
                        first = false;
                else
                        fputc(' ', f);

                cgroup_context_dump_socket_bind_item(bi, f);
        }
}

int cgroup_context_add_device_allow(CGroupContext *c, const char *dev, CGroupDevicePermissions p) {
        _cleanup_free_ CGroupDeviceAllow *a = NULL;
        _cleanup_free_ char *d = NULL;

        assert(c);
        assert(dev);
        assert(p >= 0 && p < _CGROUP_DEVICE_PERMISSIONS_MAX);

        if (p == 0)
                p = _CGROUP_DEVICE_PERMISSIONS_ALL;

        a = new(CGroupDeviceAllow, 1);
        if (!a)
                return -ENOMEM;

        d = strdup(dev);
        if (!d)
                return -ENOMEM;

        *a = (CGroupDeviceAllow) {
                .path = TAKE_PTR(d),
                .permissions = p,
        };

        LIST_PREPEND(device_allow, c->device_allow, a);
        TAKE_PTR(a);

        return 0;
}

int cgroup_context_add_or_update_device_allow(CGroupContext *c, const char *dev, CGroupDevicePermissions p) {
        assert(c);
        assert(dev);
        assert(p >= 0 && p < _CGROUP_DEVICE_PERMISSIONS_MAX);

        if (p == 0)
                p = _CGROUP_DEVICE_PERMISSIONS_ALL;

        LIST_FOREACH(device_allow, b, c->device_allow)
                if (path_equal(b->path, dev)) {
                        b->permissions = p;
                        return 0;
                }

        return cgroup_context_add_device_allow(c, dev, p);
}

int cgroup_context_add_bpf_foreign_program(CGroupContext *c, uint32_t attach_type, const char *bpffs_path) {
        CGroupBPFForeignProgram *p;
        _cleanup_free_ char *d = NULL;

        assert(c);
        assert(bpffs_path);

        if (!path_is_normalized(bpffs_path) || !path_is_absolute(bpffs_path))
                return log_error_errno(SYNTHETIC_ERRNO(EINVAL), "Path is not normalized.");

        d = strdup(bpffs_path);
        if (!d)
                return log_oom();

        p = new(CGroupBPFForeignProgram, 1);
        if (!p)
                return log_oom();

        *p = (CGroupBPFForeignProgram) {
                .attach_type = attach_type,
                .bpffs_path = TAKE_PTR(d),
        };

        LIST_PREPEND(programs, c->bpf_foreign_programs, TAKE_PTR(p));

        return 0;
}

#define UNIT_DEFINE_ANCESTOR_MEMORY_LOOKUP(entry)                       \
        uint64_t unit_get_ancestor_##entry(Unit *u) {                   \
                CGroupContext *c;                                       \
                                                                        \
                /* 1. Is entry set in this unit? If so, use that.       \
                 * 2. Is the default for this entry set in any          \
                 *    ancestor? If so, use that.                        \
                 * 3. Otherwise, return CGROUP_LIMIT_MIN. */            \
                                                                        \
                assert(u);                                              \
                                                                        \
                c = unit_get_cgroup_context(u);                         \
                if (c && c->entry##_set)                                \
                        return c->entry;                                \
                                                                        \
                while ((u = UNIT_GET_SLICE(u))) {                       \
                        c = unit_get_cgroup_context(u);                 \
                        if (c && c->default_##entry##_set)              \
                                return c->default_##entry;              \
                }                                                       \
                                                                        \
                /* We've reached the root, but nobody had default for   \
                 * this entry set, so set it to the kernel default. */  \
                return CGROUP_LIMIT_MIN;                                \
}

UNIT_DEFINE_ANCESTOR_MEMORY_LOOKUP(memory_low);
UNIT_DEFINE_ANCESTOR_MEMORY_LOOKUP(startup_memory_low);
UNIT_DEFINE_ANCESTOR_MEMORY_LOOKUP(memory_min);

static void unit_set_xattr_graceful(Unit *u, const char *name, const void *data, size_t size) {
        int r;

        assert(u);
        assert(name);

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return;

        r = cg_set_xattr(crt->cgroup_path, name, data, size, 0);
        if (r < 0)
                log_unit_debug_errno(u, r, "Failed to set '%s' xattr on control group %s, ignoring: %m", name, empty_to_root(crt->cgroup_path));
}

static void unit_remove_xattr_graceful(Unit *u, const char *name) {
        int r;

        assert(u);
        assert(name);

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return;

        r = cg_remove_xattr(crt->cgroup_path, name);
        if (r < 0 && !ERRNO_IS_XATTR_ABSENT(r))
                log_unit_debug_errno(u, r, "Failed to remove '%s' xattr flag on control group %s, ignoring: %m", name, empty_to_root(crt->cgroup_path));
}

static void cgroup_oomd_xattr_apply(Unit *u) {
        CGroupContext *c;

        assert(u);

        c = unit_get_cgroup_context(u);
        if (!c)
                return;

        if (c->moom_preference == MANAGED_OOM_PREFERENCE_OMIT)
                unit_set_xattr_graceful(u, "user.oomd_omit", "1", 1);

        if (c->moom_preference == MANAGED_OOM_PREFERENCE_AVOID)
                unit_set_xattr_graceful(u, "user.oomd_avoid", "1", 1);

        if (c->moom_preference != MANAGED_OOM_PREFERENCE_AVOID)
                unit_remove_xattr_graceful(u, "user.oomd_avoid");

        if (c->moom_preference != MANAGED_OOM_PREFERENCE_OMIT)
                unit_remove_xattr_graceful(u, "user.oomd_omit");
}

static int cgroup_log_xattr_apply(Unit *u) {
        ExecContext *c;
        size_t len, allowed_patterns_len, denied_patterns_len;
        _cleanup_free_ char *patterns = NULL, *allowed_patterns = NULL, *denied_patterns = NULL;
        char *last;
        int r;

        assert(u);

        c = unit_get_exec_context(u);
        if (!c)
                /* Some unit types have a cgroup context but no exec context, so we do not log
                 * any error here to avoid confusion. */
                return 0;

        if (set_isempty(c->log_filter_allowed_patterns) && set_isempty(c->log_filter_denied_patterns)) {
                unit_remove_xattr_graceful(u, "user.journald_log_filter_patterns");
                return 0;
        }

        r = set_make_nulstr(c->log_filter_allowed_patterns, &allowed_patterns, &allowed_patterns_len);
        if (r < 0)
                return log_debug_errno(r, "Failed to make nulstr from set: %m");

        r = set_make_nulstr(c->log_filter_denied_patterns, &denied_patterns, &denied_patterns_len);
        if (r < 0)
                return log_debug_errno(r, "Failed to make nulstr from set: %m");

        /* Use nul character separated strings without trailing nul */
        allowed_patterns_len = LESS_BY(allowed_patterns_len, 1u);
        denied_patterns_len = LESS_BY(denied_patterns_len, 1u);

        len = allowed_patterns_len + 1 + denied_patterns_len;
        patterns = new(char, len);
        if (!patterns)
                return log_oom_debug();

        last = mempcpy_safe(patterns, allowed_patterns, allowed_patterns_len);
        *(last++) = '\xff';
        memcpy_safe(last, denied_patterns, denied_patterns_len);

        unit_set_xattr_graceful(u, "user.journald_log_filter_patterns", patterns, len);

        return 0;
}

static void cgroup_invocation_id_xattr_apply(Unit *u) {
        bool b;

        assert(u);

        b = !sd_id128_is_null(u->invocation_id);
        FOREACH_STRING(xn, "trusted.invocation_id", "user.invocation_id") {
                if (b)
                        unit_set_xattr_graceful(u, xn, SD_ID128_TO_STRING(u->invocation_id), 32);
                else
                        unit_remove_xattr_graceful(u, xn);
        }
}

static void cgroup_coredump_xattr_apply(Unit *u) {
        CGroupContext *c;

        assert(u);

        c = unit_get_cgroup_context(u);
        if (!c)
                return;

        if (unit_cgroup_delegate(u) && c->coredump_receive)
                unit_set_xattr_graceful(u, "user.coredump_receive", "1", 1);
        else
                unit_remove_xattr_graceful(u, "user.coredump_receive");
}

static void cgroup_delegate_xattr_apply(Unit *u) {
        bool b;

        assert(u);

        /* Indicate on the cgroup whether delegation is on, via an xattr. This is best-effort, as old kernels
         * didn't support xattrs on cgroups at all. Later they got support for setting 'trusted.*' xattrs,
         * and even later 'user.*' xattrs. We started setting this field when 'trusted.*' was added, and
         * given this is now pretty much API, let's continue to support that. But also set 'user.*' as well,
         * since it is readable by any user, not just CAP_SYS_ADMIN. This hence comes with slightly weaker
         * security (as users who got delegated cgroups could turn it off if they like), but this shouldn't
         * be a big problem given this communicates delegation state to clients, but the manager never reads
         * it. */
        b = unit_cgroup_delegate(u);
        FOREACH_STRING(xn, "trusted.delegate", "user.delegate") {
                if (b)
                        unit_set_xattr_graceful(u, xn, "1", 1);
                else
                        unit_remove_xattr_graceful(u, xn);
        }
}

static void cgroup_survive_xattr_apply(Unit *u) {
        int r;

        assert(u);

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt)
                return;

        if (u->survive_final_kill_signal) {
                r = cg_set_xattr(
                                crt->cgroup_path,
                                "user.survive_final_kill_signal",
                                "1",
                                1,
                                /* flags= */ 0);
                /* user xattr support was added in kernel v5.7 */
                if (ERRNO_IS_NEG_NOT_SUPPORTED(r))
                        r = cg_set_xattr(
                                        crt->cgroup_path,
                                        "trusted.survive_final_kill_signal",
                                        "1",
                                        1,
                                        /* flags= */ 0);
                if (r < 0)
                        log_unit_debug_errno(u,
                                             r,
                                             "Failed to set 'survive_final_kill_signal' xattr on control "
                                             "group %s, ignoring: %m",
                                             empty_to_root(crt->cgroup_path));
        } else {
                unit_remove_xattr_graceful(u, "user.survive_final_kill_signal");
                unit_remove_xattr_graceful(u, "trusted.survive_final_kill_signal");
        }
}

static void cgroup_xattr_apply(Unit *u) {
        assert(u);

        /* The 'user.*' xattrs can be set from a user manager. */
        cgroup_oomd_xattr_apply(u);
        cgroup_log_xattr_apply(u);
        cgroup_coredump_xattr_apply(u);

        if (!MANAGER_IS_SYSTEM(u->manager))
                return;

        cgroup_invocation_id_xattr_apply(u);
        cgroup_delegate_xattr_apply(u);
        cgroup_survive_xattr_apply(u);
}

static int lookup_block_device(const char *p, dev_t *ret) {
        dev_t rdev, dev = 0;
        mode_t mode;
        int r;

        assert(p);
        assert(ret);

        r = device_path_parse_major_minor(p, &mode, &rdev);
        if (r == -ENODEV) { /* not a parsable device node, need to go to disk */
                struct stat st;

                if (stat(p, &st) < 0)
                        return log_warning_errno(errno, "Couldn't stat device '%s': %m", p);

                mode = st.st_mode;
                rdev = st.st_rdev;
                dev = st.st_dev;
        } else if (r < 0)
                return log_warning_errno(r, "Failed to parse major/minor from path '%s': %m", p);

        if (S_ISCHR(mode))
                return log_warning_errno(SYNTHETIC_ERRNO(ENOTBLK),
                                         "Device node '%s' is a character device, but block device needed.", p);
        if (S_ISBLK(mode))
                *ret = rdev;
        else if (major(dev) != 0)
                *ret = dev; /* If this is not a device node then use the block device this file is stored on */
        else {
                /* If this is btrfs, getting the backing block device is a bit harder */
                r = btrfs_get_block_device(p, ret);
                if (r == -ENOTTY)
                        return log_warning_errno(SYNTHETIC_ERRNO(ENODEV),
                                                 "'%s' is not a block device node, and file system block device cannot be determined or is not local.", p);
                if (r < 0)
                        return log_warning_errno(r, "Failed to determine block device backing btrfs file system '%s': %m", p);
        }

        /* If this is a LUKS/DM device, recursively try to get the originating block device */
        while (block_get_originating(*ret, ret) > 0);

        /* If this is a partition, try to get the originating block device */
        (void) block_get_whole_disk(*ret, ret);
        return 0;
}

static bool cgroup_context_has_cpu_weight(CGroupContext *c) {
        return c->cpu_weight != CGROUP_WEIGHT_INVALID ||
                c->startup_cpu_weight != CGROUP_WEIGHT_INVALID;
}

static bool cgroup_context_has_cpu_shares(CGroupContext *c) {
        return c->cpu_shares != CGROUP_CPU_SHARES_INVALID ||
                c->startup_cpu_shares != CGROUP_CPU_SHARES_INVALID;
}

static bool cgroup_context_has_allowed_cpus(CGroupContext *c) {
        return c->cpuset_cpus.set || c->startup_cpuset_cpus.set;
}

static bool cgroup_context_has_allowed_mems(CGroupContext *c) {
        return c->cpuset_mems.set || c->startup_cpuset_mems.set;
}

uint64_t cgroup_context_cpu_weight(CGroupContext *c, ManagerState state) {
        assert(c);

        if (IN_SET(state, MANAGER_STARTING, MANAGER_INITIALIZING, MANAGER_STOPPING) &&
            c->startup_cpu_weight != CGROUP_WEIGHT_INVALID)
                return c->startup_cpu_weight;
        else if (c->cpu_weight != CGROUP_WEIGHT_INVALID)
                return c->cpu_weight;
        else
                return CGROUP_WEIGHT_DEFAULT;
}

static uint64_t cgroup_context_cpu_shares(CGroupContext *c, ManagerState state) {
        if (IN_SET(state, MANAGER_STARTING, MANAGER_INITIALIZING, MANAGER_STOPPING) &&
            c->startup_cpu_shares != CGROUP_CPU_SHARES_INVALID)
                return c->startup_cpu_shares;
        else if (c->cpu_shares != CGROUP_CPU_SHARES_INVALID)
                return c->cpu_shares;
        else
                return CGROUP_CPU_SHARES_DEFAULT;
}

static CPUSet *cgroup_context_allowed_cpus(CGroupContext *c, ManagerState state) {
        if (IN_SET(state, MANAGER_STARTING, MANAGER_INITIALIZING, MANAGER_STOPPING) &&
            c->startup_cpuset_cpus.set)
                return &c->startup_cpuset_cpus;
        else
                return &c->cpuset_cpus;
}

static CPUSet *cgroup_context_allowed_mems(CGroupContext *c, ManagerState state) {
        if (IN_SET(state, MANAGER_STARTING, MANAGER_INITIALIZING, MANAGER_STOPPING) &&
            c->startup_cpuset_mems.set)
                return &c->startup_cpuset_mems;
        else
                return &c->cpuset_mems;
}

usec_t cgroup_cpu_adjust_period(usec_t period, usec_t quota, usec_t resolution, usec_t max_period) {
        /* kernel uses a minimum resolution of 1ms, so both period and (quota * period)
         * need to be higher than that boundary. quota is specified in USecPerSec.
         * Additionally, period must be at most max_period. */
        assert(quota > 0);

        return MIN(MAX3(period, resolution, resolution * USEC_PER_SEC / quota), max_period);
}

static usec_t cgroup_cpu_adjust_period_and_log(Unit *u, usec_t period, usec_t quota) {
        usec_t new_period;

        assert(u);

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt)
                return USEC_INFINITY;

        if (quota == USEC_INFINITY)
                /* Always use default period for infinity quota. */
                return CGROUP_CPU_QUOTA_DEFAULT_PERIOD_USEC;

        if (period == USEC_INFINITY)
                /* Default period was requested. */
                period = CGROUP_CPU_QUOTA_DEFAULT_PERIOD_USEC;

        /* Clamp to interval [1ms, 1s] */
        new_period = cgroup_cpu_adjust_period(period, quota, USEC_PER_MSEC, USEC_PER_SEC);

        if (new_period != period) {
                log_unit_full(u, crt->warned_clamping_cpu_quota_period ? LOG_DEBUG : LOG_WARNING,
                              "Clamping CPU interval for cpu.max: period is now %s",
                              FORMAT_TIMESPAN(new_period, 1));
                crt->warned_clamping_cpu_quota_period = true;
        }

        return new_period;
}

static void cgroup_apply_unified_cpu_weight(Unit *u, uint64_t weight) {
        char buf[DECIMAL_STR_MAX(uint64_t) + 2];

        if (weight == CGROUP_WEIGHT_IDLE)
                return;
        xsprintf(buf, "%" PRIu64 "\n", weight);
        (void) set_attribute_and_warn(u, "cpu", "cpu.weight", buf);
}

static void cgroup_apply_unified_cpu_idle(Unit *u, uint64_t weight) {
        int r;
        bool is_idle;
        const char *idle_val;

        assert(u);

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return;

        is_idle = weight == CGROUP_WEIGHT_IDLE;
        idle_val = one_zero(is_idle);
        r = cg_set_attribute("cpu", crt->cgroup_path, "cpu.idle", idle_val);
        if (r < 0 && (r != -ENOENT || is_idle))
                log_unit_full_errno(u, LOG_LEVEL_CGROUP_WRITE(r), r, "Failed to set '%s' attribute on '%s' to '%s': %m",
                                    "cpu.idle", empty_to_root(crt->cgroup_path), idle_val);
}

static void cgroup_apply_unified_cpu_quota(Unit *u, usec_t quota, usec_t period) {
        char buf[(DECIMAL_STR_MAX(usec_t) + 1) * 2 + 1];

        assert(u);

        period = cgroup_cpu_adjust_period_and_log(u, period, quota);
        if (quota != USEC_INFINITY)
                xsprintf(buf, USEC_FMT " " USEC_FMT "\n",
                         MAX(quota * period / USEC_PER_SEC, USEC_PER_MSEC), period);
        else
                xsprintf(buf, "max " USEC_FMT "\n", period);
        (void) set_attribute_and_warn(u, "cpu", "cpu.max", buf);
}

static void cgroup_apply_legacy_cpu_shares(Unit *u, uint64_t shares) {
        char buf[DECIMAL_STR_MAX(uint64_t) + 2];

        xsprintf(buf, "%" PRIu64 "\n", shares);
        (void) set_attribute_and_warn(u, "cpu", "cpu.shares", buf);
}

static void cgroup_apply_legacy_cpu_quota(Unit *u, usec_t quota, usec_t period) {
        char buf[DECIMAL_STR_MAX(usec_t) + 2];

        period = cgroup_cpu_adjust_period_and_log(u, period, quota);

        xsprintf(buf, USEC_FMT "\n", period);
        (void) set_attribute_and_warn(u, "cpu", "cpu.cfs_period_us", buf);

        if (quota != USEC_INFINITY) {
                xsprintf(buf, USEC_FMT "\n", MAX(quota * period / USEC_PER_SEC, USEC_PER_MSEC));
                (void) set_attribute_and_warn(u, "cpu", "cpu.cfs_quota_us", buf);
        } else
                (void) set_attribute_and_warn(u, "cpu", "cpu.cfs_quota_us", "-1\n");
}

static uint64_t cgroup_cpu_shares_to_weight(uint64_t shares) {
        return CLAMP(shares * CGROUP_WEIGHT_DEFAULT / CGROUP_CPU_SHARES_DEFAULT,
                     CGROUP_WEIGHT_MIN, CGROUP_WEIGHT_MAX);
}

static uint64_t cgroup_cpu_weight_to_shares(uint64_t weight) {
        /* we don't support idle in cgroupv1 */
        if (weight == CGROUP_WEIGHT_IDLE)
                return CGROUP_CPU_SHARES_MIN;

        return CLAMP(weight * CGROUP_CPU_SHARES_DEFAULT / CGROUP_WEIGHT_DEFAULT,
                     CGROUP_CPU_SHARES_MIN, CGROUP_CPU_SHARES_MAX);
}

static void cgroup_apply_unified_cpuset(Unit *u, const CPUSet *cpus, const char *name) {
        _cleanup_free_ char *buf = NULL;

        buf = cpu_set_to_range_string(cpus);
        if (!buf) {
                log_oom();
                return;
        }

        (void) set_attribute_and_warn(u, "cpuset", name, buf);
}

static bool cgroup_context_has_io_config(CGroupContext *c) {
        return c->io_accounting ||
                c->io_weight != CGROUP_WEIGHT_INVALID ||
                c->startup_io_weight != CGROUP_WEIGHT_INVALID ||
                c->io_device_weights ||
                c->io_device_latencies ||
                c->io_device_limits;
}

static bool cgroup_context_has_blockio_config(CGroupContext *c) {
        return c->blockio_accounting ||
                c->blockio_weight != CGROUP_BLKIO_WEIGHT_INVALID ||
                c->startup_blockio_weight != CGROUP_BLKIO_WEIGHT_INVALID ||
                c->blockio_device_weights ||
                c->blockio_device_bandwidths;
}

static uint64_t cgroup_context_io_weight(CGroupContext *c, ManagerState state) {
        if (IN_SET(state, MANAGER_STARTING, MANAGER_INITIALIZING, MANAGER_STOPPING) &&
            c->startup_io_weight != CGROUP_WEIGHT_INVALID)
                return c->startup_io_weight;
        if (c->io_weight != CGROUP_WEIGHT_INVALID)
                return c->io_weight;
        return CGROUP_WEIGHT_DEFAULT;
}

static uint64_t cgroup_context_blkio_weight(CGroupContext *c, ManagerState state) {
        if (IN_SET(state, MANAGER_STARTING, MANAGER_INITIALIZING, MANAGER_STOPPING) &&
            c->startup_blockio_weight != CGROUP_BLKIO_WEIGHT_INVALID)
                return c->startup_blockio_weight;
        if (c->blockio_weight != CGROUP_BLKIO_WEIGHT_INVALID)
                return c->blockio_weight;
        return CGROUP_BLKIO_WEIGHT_DEFAULT;
}

static uint64_t cgroup_weight_blkio_to_io(uint64_t blkio_weight) {
        return CLAMP(blkio_weight * CGROUP_WEIGHT_DEFAULT / CGROUP_BLKIO_WEIGHT_DEFAULT,
                     CGROUP_WEIGHT_MIN, CGROUP_WEIGHT_MAX);
}

static uint64_t cgroup_weight_io_to_blkio(uint64_t io_weight) {
        return CLAMP(io_weight * CGROUP_BLKIO_WEIGHT_DEFAULT / CGROUP_WEIGHT_DEFAULT,
                     CGROUP_BLKIO_WEIGHT_MIN, CGROUP_BLKIO_WEIGHT_MAX);
}

static int set_bfq_weight(Unit *u, const char *controller, dev_t dev, uint64_t io_weight) {
        static const char * const prop_names[] = {
                "IOWeight",
                "BlockIOWeight",
                "IODeviceWeight",
                "BlockIODeviceWeight",
        };
        static bool warned = false;
        char buf[DECIMAL_STR_MAX(dev_t)*2+2+DECIMAL_STR_MAX(uint64_t)+STRLEN("\n")];
        const char *p;
        uint64_t bfq_weight;
        int r;

        assert(u);

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return -EOWNERDEAD;

        /* FIXME: drop this function when distro kernels properly support BFQ through "io.weight"
         * See also: https://github.com/systemd/systemd/pull/13335 and
         * https://github.com/torvalds/linux/commit/65752aef0a407e1ef17ec78a7fc31ba4e0b360f9. */
        p = strjoina(controller, ".bfq.weight");
        /* Adjust to kernel range is 1..1000, the default is 100. */
        bfq_weight = BFQ_WEIGHT(io_weight);

        if (major(dev) > 0)
                xsprintf(buf, DEVNUM_FORMAT_STR " %" PRIu64 "\n", DEVNUM_FORMAT_VAL(dev), bfq_weight);
        else
                xsprintf(buf, "%" PRIu64 "\n", bfq_weight);

        r = cg_set_attribute(controller, crt->cgroup_path, p, buf);

        /* FIXME: drop this when kernels prior
         * 795fe54c2a82 ("bfq: Add per-device weight") v5.4
         * are not interesting anymore. Old kernels will fail with EINVAL, while new kernels won't return
         * EINVAL on properly formatted input by us. Treat EINVAL accordingly. */
        if (r == -EINVAL && major(dev) > 0) {
               if (!warned) {
                        log_unit_warning(u, "Kernel version does not accept per-device setting in %s.", p);
                        warned = true;
               }
               r = -EOPNOTSUPP; /* mask as unconfigured device */
        } else if (r >= 0 && io_weight != bfq_weight)
                log_unit_debug(u, "%s=%" PRIu64 " scaled to %s=%" PRIu64,
                               prop_names[2*(major(dev) > 0) + streq(controller, "blkio")],
                               io_weight, p, bfq_weight);
        return r;
}

static void cgroup_apply_io_device_weight(Unit *u, const char *dev_path, uint64_t io_weight) {
        char buf[DECIMAL_STR_MAX(dev_t)*2+2+DECIMAL_STR_MAX(uint64_t)+1];
        dev_t dev;
        int r, r1, r2;

        assert(u);

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return;

        if (lookup_block_device(dev_path, &dev) < 0)
                return;

        r1 = set_bfq_weight(u, "io", dev, io_weight);

        xsprintf(buf, DEVNUM_FORMAT_STR " %" PRIu64 "\n", DEVNUM_FORMAT_VAL(dev), io_weight);
        r2 = cg_set_attribute("io", crt->cgroup_path, "io.weight", buf);

        /* Look at the configured device, when both fail, prefer io.weight errno. */
        r = r2 == -EOPNOTSUPP ? r1 : r2;

        if (r < 0)
                log_unit_full_errno(u, LOG_LEVEL_CGROUP_WRITE(r),
                                    r, "Failed to set 'io[.bfq].weight' attribute on '%s' to '%.*s': %m",
                                    empty_to_root(crt->cgroup_path), (int) strcspn(buf, NEWLINE), buf);
}

static void cgroup_apply_blkio_device_weight(Unit *u, const char *dev_path, uint64_t blkio_weight) {
        char buf[DECIMAL_STR_MAX(dev_t)*2+2+DECIMAL_STR_MAX(uint64_t)+1];
        dev_t dev;
        int r;

        r = lookup_block_device(dev_path, &dev);
        if (r < 0)
                return;

        xsprintf(buf, DEVNUM_FORMAT_STR " %" PRIu64 "\n", DEVNUM_FORMAT_VAL(dev), blkio_weight);
        (void) set_attribute_and_warn(u, "blkio", "blkio.weight_device", buf);
}

static void cgroup_apply_io_device_latency(Unit *u, const char *dev_path, usec_t target) {
        char buf[DECIMAL_STR_MAX(dev_t)*2+2+7+DECIMAL_STR_MAX(uint64_t)+1];
        dev_t dev;
        int r;

        r = lookup_block_device(dev_path, &dev);
        if (r < 0)
                return;

        if (target != USEC_INFINITY)
                xsprintf(buf, DEVNUM_FORMAT_STR " target=%" PRIu64 "\n", DEVNUM_FORMAT_VAL(dev), target);
        else
                xsprintf(buf, DEVNUM_FORMAT_STR " target=max\n", DEVNUM_FORMAT_VAL(dev));

        (void) set_attribute_and_warn(u, "io", "io.latency", buf);
}

static void cgroup_apply_io_device_limit(Unit *u, const char *dev_path, uint64_t *limits) {
        char limit_bufs[_CGROUP_IO_LIMIT_TYPE_MAX][DECIMAL_STR_MAX(uint64_t)],
             buf[DECIMAL_STR_MAX(dev_t)*2+2+(6+DECIMAL_STR_MAX(uint64_t)+1)*4];
        dev_t dev;

        if (lookup_block_device(dev_path, &dev) < 0)
                return;

        for (CGroupIOLimitType type = 0; type < _CGROUP_IO_LIMIT_TYPE_MAX; type++)
                if (limits[type] != cgroup_io_limit_defaults[type])
                        xsprintf(limit_bufs[type], "%" PRIu64, limits[type]);
                else
                        xsprintf(limit_bufs[type], "%s", limits[type] == CGROUP_LIMIT_MAX ? "max" : "0");

        xsprintf(buf, DEVNUM_FORMAT_STR " rbps=%s wbps=%s riops=%s wiops=%s\n", DEVNUM_FORMAT_VAL(dev),
                 limit_bufs[CGROUP_IO_RBPS_MAX], limit_bufs[CGROUP_IO_WBPS_MAX],
                 limit_bufs[CGROUP_IO_RIOPS_MAX], limit_bufs[CGROUP_IO_WIOPS_MAX]);
        (void) set_attribute_and_warn(u, "io", "io.max", buf);
}

static void cgroup_apply_blkio_device_limit(Unit *u, const char *dev_path, uint64_t rbps, uint64_t wbps) {
        char buf[DECIMAL_STR_MAX(dev_t)*2+2+DECIMAL_STR_MAX(uint64_t)+1];
        dev_t dev;

        if (lookup_block_device(dev_path, &dev) < 0)
                return;

        sprintf(buf, DEVNUM_FORMAT_STR " %" PRIu64 "\n", DEVNUM_FORMAT_VAL(dev), rbps);
        (void) set_attribute_and_warn(u, "blkio", "blkio.throttle.read_bps_device", buf);

        sprintf(buf, DEVNUM_FORMAT_STR " %" PRIu64 "\n", DEVNUM_FORMAT_VAL(dev), wbps);
        (void) set_attribute_and_warn(u, "blkio", "blkio.throttle.write_bps_device", buf);
}

static bool unit_has_unified_memory_config(Unit *u) {
        CGroupContext *c;

        assert(u);

        assert_se(c = unit_get_cgroup_context(u));

        return unit_get_ancestor_memory_min(u) > 0 ||
               unit_get_ancestor_memory_low(u) > 0 || unit_get_ancestor_startup_memory_low(u) > 0 ||
               c->memory_high != CGROUP_LIMIT_MAX || c->startup_memory_high_set ||
               c->memory_max != CGROUP_LIMIT_MAX || c->startup_memory_max_set ||
               c->memory_swap_max != CGROUP_LIMIT_MAX || c->startup_memory_swap_max_set ||
               c->memory_zswap_max != CGROUP_LIMIT_MAX || c->startup_memory_zswap_max_set;
}

static void cgroup_apply_unified_memory_limit(Unit *u, const char *file, uint64_t v) {
        char buf[DECIMAL_STR_MAX(uint64_t) + 1] = "max\n";

        if (v != CGROUP_LIMIT_MAX)
                xsprintf(buf, "%" PRIu64 "\n", v);

        (void) set_attribute_and_warn(u, "memory", file, buf);
}

static void cgroup_apply_firewall(Unit *u) {
        assert(u);

        /* Best-effort: let's apply IP firewalling and/or accounting if that's enabled */

        if (bpf_firewall_compile(u) < 0)
                return;

        (void) bpf_firewall_load_custom(u);
        (void) bpf_firewall_install(u);
}

void unit_modify_nft_set(Unit *u, bool add) {
        int r;

        assert(u);

        if (!MANAGER_IS_SYSTEM(u->manager))
                return;

        if (!UNIT_HAS_CGROUP_CONTEXT(u))
                return;

        if (cg_all_unified() <= 0)
                return;

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || crt->cgroup_id == 0)
                return;

        if (!u->manager->fw_ctx) {
                r = fw_ctx_new_full(&u->manager->fw_ctx, /* init_tables= */ false);
                if (r < 0)
                        return;

                assert(u->manager->fw_ctx);
        }

        CGroupContext *c = ASSERT_PTR(unit_get_cgroup_context(u));

        FOREACH_ARRAY(nft_set, c->nft_set_context.sets, c->nft_set_context.n_sets) {
                if (nft_set->source != NFT_SET_SOURCE_CGROUP)
                        continue;

                uint64_t element = crt->cgroup_id;

                r = nft_set_element_modify_any(u->manager->fw_ctx, add, nft_set->nfproto, nft_set->table, nft_set->set, &element, sizeof(element));
                if (r < 0)
                        log_warning_errno(r, "Failed to %s NFT set: family %s, table %s, set %s, cgroup %" PRIu64 ", ignoring: %m",
                                          add? "add" : "delete", nfproto_to_string(nft_set->nfproto), nft_set->table, nft_set->set, crt->cgroup_id);
                else
                        log_debug("%s NFT set: family %s, table %s, set %s, cgroup %" PRIu64,
                                  add? "Added" : "Deleted", nfproto_to_string(nft_set->nfproto), nft_set->table, nft_set->set, crt->cgroup_id);
        }
}

static void cgroup_apply_socket_bind(Unit *u) {
        assert(u);

        (void) bpf_socket_bind_install(u);
}

static void cgroup_apply_restrict_network_interfaces(Unit *u) {
        assert(u);

        (void) bpf_restrict_ifaces_install(u);
}

static int cgroup_apply_devices(Unit *u) {
        _cleanup_(bpf_program_freep) BPFProgram *prog = NULL;
        CGroupContext *c;
        CGroupDevicePolicy policy;
        int r;

        assert_se(c = unit_get_cgroup_context(u));

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return -EOWNERDEAD;

        policy = c->device_policy;

        if (cg_all_unified() > 0) {
                r = bpf_devices_cgroup_init(&prog, policy, c->device_allow);
                if (r < 0)
                        return log_unit_warning_errno(u, r, "Failed to initialize device control bpf program: %m");

        } else {
                /* Changing the devices list of a populated cgroup might result in EINVAL, hence ignore
                 * EINVAL here. */

                if (c->device_allow || policy != CGROUP_DEVICE_POLICY_AUTO)
                        r = cg_set_attribute("devices", crt->cgroup_path, "devices.deny", "a");
                else
                        r = cg_set_attribute("devices", crt->cgroup_path, "devices.allow", "a");
                if (r < 0)
                        log_unit_full_errno(u, IN_SET(r, -ENOENT, -EROFS, -EINVAL, -EACCES, -EPERM) ? LOG_DEBUG : LOG_WARNING, r,
                                            "Failed to reset devices.allow/devices.deny: %m");
        }

        bool allow_list_static = policy == CGROUP_DEVICE_POLICY_CLOSED ||
                (policy == CGROUP_DEVICE_POLICY_AUTO && c->device_allow);

        bool any = false;
        if (allow_list_static) {
                r = bpf_devices_allow_list_static(prog, crt->cgroup_path);
                if (r > 0)
                        any = true;
        }

        LIST_FOREACH(device_allow, a, c->device_allow) {
                const char *val;

                if (a->permissions == 0)
                        continue;

                if (path_startswith(a->path, "/dev/"))
                        r = bpf_devices_allow_list_device(prog, crt->cgroup_path, a->path, a->permissions);
                else if ((val = startswith(a->path, "block-")))
                        r = bpf_devices_allow_list_major(prog, crt->cgroup_path, val, 'b', a->permissions);
                else if ((val = startswith(a->path, "char-")))
                        r = bpf_devices_allow_list_major(prog, crt->cgroup_path, val, 'c', a->permissions);
                else {
                        log_unit_debug(u, "Ignoring device '%s' while writing cgroup attribute.", a->path);
                        continue;
                }

                if (r > 0)
                        any = true;
        }

        if (prog && !any) {
                log_unit_warning(u, "No devices matched by device filter.");

                /* The kernel verifier would reject a program we would build with the normal intro and outro
                   but no allow-listing rules (outro would contain an unreachable instruction for successful
                   return). */
                policy = CGROUP_DEVICE_POLICY_STRICT;
        }

        r = bpf_devices_apply_policy(&prog, policy, any, crt->cgroup_path, &crt->bpf_device_control_installed);
        if (r < 0) {
                static bool warned = false;

                log_full_errno(warned ? LOG_DEBUG : LOG_WARNING, r,
                               "Unit %s configures device ACL, but the local system doesn't seem to support the BPF-based device controller.\n"
                               "Proceeding WITHOUT applying ACL (all devices will be accessible)!\n"
                               "(This warning is only shown for the first loaded unit using device ACL.)", u->id);

                warned = true;
        }
        return r;
}

static void set_io_weight(Unit *u, uint64_t weight) {
        char buf[STRLEN("default \n")+DECIMAL_STR_MAX(uint64_t)];

        assert(u);

        (void) set_bfq_weight(u, "io", makedev(0, 0), weight);

        xsprintf(buf, "default %" PRIu64 "\n", weight);
        (void) set_attribute_and_warn(u, "io", "io.weight", buf);
}

static void set_blkio_weight(Unit *u, uint64_t weight) {
        char buf[STRLEN("\n")+DECIMAL_STR_MAX(uint64_t)];

        assert(u);

        (void) set_bfq_weight(u, "blkio", makedev(0, 0), weight);

        xsprintf(buf, "%" PRIu64 "\n", weight);
        (void) set_attribute_and_warn(u, "blkio", "blkio.weight", buf);
}

static void cgroup_apply_bpf_foreign_program(Unit *u) {
        assert(u);

        (void) bpf_foreign_install(u);
}

static void cgroup_context_apply(
                Unit *u,
                CGroupMask apply_mask,
                ManagerState state) {

        bool is_host_root, is_local_root;
        const char *path;
        CGroupContext *c;
        int r;

        assert(u);

        /* Nothing to do? Exit early! */
        if (apply_mask == 0)
                return;

        /* Some cgroup attributes are not supported on the host root cgroup, hence silently ignore them here. And other
         * attributes should only be managed for cgroups further down the tree. */
        is_local_root = unit_has_name(u, SPECIAL_ROOT_SLICE);
        is_host_root = unit_has_host_root_cgroup(u);

        assert_se(c = unit_get_cgroup_context(u));

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return;

        path = crt->cgroup_path;

        if (is_local_root) /* Make sure we don't try to display messages with an empty path. */
                path = "/";

        /* We generally ignore errors caused by read-only mounted cgroup trees (assuming we are running in a container
         * then), and missing cgroups, i.e. EROFS and ENOENT. */

        /* In fully unified mode these attributes don't exist on the host cgroup root. On legacy the weights exist, but
         * setting the weight makes very little sense on the host root cgroup, as there are no other cgroups at this
         * level. The quota exists there too, but any attempt to write to it is refused with EINVAL. Inside of
         * containers we want to leave control of these to the container manager (and if cgroup v2 delegation is used
         * we couldn't even write to them if we wanted to). */
        if ((apply_mask & CGROUP_MASK_CPU) && !is_local_root) {

                if (cg_all_unified() > 0) {
                        uint64_t weight;

                        if (cgroup_context_has_cpu_weight(c))
                                weight = cgroup_context_cpu_weight(c, state);
                        else if (cgroup_context_has_cpu_shares(c)) {
                                uint64_t shares;

                                shares = cgroup_context_cpu_shares(c, state);
                                weight = cgroup_cpu_shares_to_weight(shares);

                                log_cgroup_compat(u, "Applying [Startup]CPUShares=%" PRIu64 " as [Startup]CPUWeight=%" PRIu64 " on %s",
                                                  shares, weight, path);
                        } else
                                weight = CGROUP_WEIGHT_DEFAULT;

                        cgroup_apply_unified_cpu_idle(u, weight);
                        cgroup_apply_unified_cpu_weight(u, weight);
                        cgroup_apply_unified_cpu_quota(u, c->cpu_quota_per_sec_usec, c->cpu_quota_period_usec);

                } else {
                        uint64_t shares;

                        if (cgroup_context_has_cpu_weight(c)) {
                                uint64_t weight;

                                weight = cgroup_context_cpu_weight(c, state);
                                shares = cgroup_cpu_weight_to_shares(weight);

                                log_cgroup_compat(u, "Applying [Startup]CPUWeight=%" PRIu64 " as [Startup]CPUShares=%" PRIu64 " on %s",
                                                  weight, shares, path);
                        } else if (cgroup_context_has_cpu_shares(c))
                                shares = cgroup_context_cpu_shares(c, state);
                        else
                                shares = CGROUP_CPU_SHARES_DEFAULT;

                        cgroup_apply_legacy_cpu_shares(u, shares);
                        cgroup_apply_legacy_cpu_quota(u, c->cpu_quota_per_sec_usec, c->cpu_quota_period_usec);
                }
        }

        if ((apply_mask & CGROUP_MASK_CPUSET) && !is_local_root) {
                cgroup_apply_unified_cpuset(u, cgroup_context_allowed_cpus(c, state), "cpuset.cpus");
                cgroup_apply_unified_cpuset(u, cgroup_context_allowed_mems(c, state), "cpuset.mems");
        }

        /* The 'io' controller attributes are not exported on the host's root cgroup (being a pure cgroup v2
         * controller), and in case of containers we want to leave control of these attributes to the container manager
         * (and we couldn't access that stuff anyway, even if we tried if proper delegation is used). */
        if ((apply_mask & CGROUP_MASK_IO) && !is_local_root) {
                bool has_io, has_blockio;
                uint64_t weight;

                has_io = cgroup_context_has_io_config(c);
                has_blockio = cgroup_context_has_blockio_config(c);

                if (has_io)
                        weight = cgroup_context_io_weight(c, state);
                else if (has_blockio) {
                        uint64_t blkio_weight;

                        blkio_weight = cgroup_context_blkio_weight(c, state);
                        weight = cgroup_weight_blkio_to_io(blkio_weight);

                        log_cgroup_compat(u, "Applying [Startup]BlockIOWeight=%" PRIu64 " as [Startup]IOWeight=%" PRIu64,
                                          blkio_weight, weight);
                } else
                        weight = CGROUP_WEIGHT_DEFAULT;

                set_io_weight(u, weight);

                if (has_io) {
                        LIST_FOREACH(device_weights, w, c->io_device_weights)
                                cgroup_apply_io_device_weight(u, w->path, w->weight);

                        LIST_FOREACH(device_limits, limit, c->io_device_limits)
                                cgroup_apply_io_device_limit(u, limit->path, limit->limits);

                        LIST_FOREACH(device_latencies, latency, c->io_device_latencies)
                                cgroup_apply_io_device_latency(u, latency->path, latency->target_usec);

                } else if (has_blockio) {
                        LIST_FOREACH(device_weights, w, c->blockio_device_weights) {
                                weight = cgroup_weight_blkio_to_io(w->weight);

                                log_cgroup_compat(u, "Applying BlockIODeviceWeight=%" PRIu64 " as IODeviceWeight=%" PRIu64 " for %s",
                                                  w->weight, weight, w->path);

                                cgroup_apply_io_device_weight(u, w->path, weight);
                        }

                        LIST_FOREACH(device_bandwidths, b, c->blockio_device_bandwidths) {
                                uint64_t limits[_CGROUP_IO_LIMIT_TYPE_MAX];

                                for (CGroupIOLimitType type = 0; type < _CGROUP_IO_LIMIT_TYPE_MAX; type++)
                                        limits[type] = cgroup_io_limit_defaults[type];

                                limits[CGROUP_IO_RBPS_MAX] = b->rbps;
                                limits[CGROUP_IO_WBPS_MAX] = b->wbps;

                                log_cgroup_compat(u, "Applying BlockIO{Read|Write}Bandwidth=%" PRIu64 " %" PRIu64 " as IO{Read|Write}BandwidthMax= for %s",
                                                  b->rbps, b->wbps, b->path);

                                cgroup_apply_io_device_limit(u, b->path, limits);
                        }
                }
        }

        if (apply_mask & CGROUP_MASK_BLKIO) {
                bool has_io, has_blockio;

                has_io = cgroup_context_has_io_config(c);
                has_blockio = cgroup_context_has_blockio_config(c);

                /* Applying a 'weight' never makes sense for the host root cgroup, and for containers this should be
                 * left to our container manager, too. */
                if (!is_local_root) {
                        uint64_t weight;

                        if (has_io) {
                                uint64_t io_weight;

                                io_weight = cgroup_context_io_weight(c, state);
                                weight = cgroup_weight_io_to_blkio(cgroup_context_io_weight(c, state));

                                log_cgroup_compat(u, "Applying [Startup]IOWeight=%" PRIu64 " as [Startup]BlockIOWeight=%" PRIu64,
                                                  io_weight, weight);
                        } else if (has_blockio)
                                weight = cgroup_context_blkio_weight(c, state);
                        else
                                weight = CGROUP_BLKIO_WEIGHT_DEFAULT;

                        set_blkio_weight(u, weight);

                        if (has_io)
                                LIST_FOREACH(device_weights, w, c->io_device_weights) {
                                        weight = cgroup_weight_io_to_blkio(w->weight);

                                        log_cgroup_compat(u, "Applying IODeviceWeight=%" PRIu64 " as BlockIODeviceWeight=%" PRIu64 " for %s",
                                                          w->weight, weight, w->path);

                                        cgroup_apply_blkio_device_weight(u, w->path, weight);
                                }
                        else if (has_blockio)
                                LIST_FOREACH(device_weights, w, c->blockio_device_weights)
                                        cgroup_apply_blkio_device_weight(u, w->path, w->weight);
                }

                /* The bandwidth limits are something that make sense to be applied to the host's root but not container
                 * roots, as there we want the container manager to handle it */
                if (is_host_root || !is_local_root) {
                        if (has_io)
                                LIST_FOREACH(device_limits, l, c->io_device_limits) {
                                        log_cgroup_compat(u, "Applying IO{Read|Write}Bandwidth=%" PRIu64 " %" PRIu64 " as BlockIO{Read|Write}BandwidthMax= for %s",
                                                          l->limits[CGROUP_IO_RBPS_MAX], l->limits[CGROUP_IO_WBPS_MAX], l->path);

                                        cgroup_apply_blkio_device_limit(u, l->path, l->limits[CGROUP_IO_RBPS_MAX], l->limits[CGROUP_IO_WBPS_MAX]);
                                }
                        else if (has_blockio)
                                LIST_FOREACH(device_bandwidths, b, c->blockio_device_bandwidths)
                                        cgroup_apply_blkio_device_limit(u, b->path, b->rbps, b->wbps);
                }
        }

        /* In unified mode 'memory' attributes do not exist on the root cgroup. In legacy mode 'memory.limit_in_bytes'
         * exists on the root cgroup, but any writes to it are refused with EINVAL. And if we run in a container we
         * want to leave control to the container manager (and if proper cgroup v2 delegation is used we couldn't even
         * write to this if we wanted to.) */
        if ((apply_mask & CGROUP_MASK_MEMORY) && !is_local_root) {

                if (cg_all_unified() > 0) {
                        uint64_t max, swap_max = CGROUP_LIMIT_MAX, zswap_max = CGROUP_LIMIT_MAX, high = CGROUP_LIMIT_MAX;

                        if (unit_has_unified_memory_config(u)) {
                                bool startup = IN_SET(state, MANAGER_STARTING, MANAGER_INITIALIZING, MANAGER_STOPPING);

                                high = startup && c->startup_memory_high_set ? c->startup_memory_high : c->memory_high;
                                max = startup && c->startup_memory_max_set ? c->startup_memory_max : c->memory_max;
                                swap_max = startup && c->startup_memory_swap_max_set ? c->startup_memory_swap_max : c->memory_swap_max;
                                zswap_max = startup && c->startup_memory_zswap_max_set ? c->startup_memory_zswap_max : c->memory_zswap_max;
                        } else {
                                max = c->memory_limit;

                                if (max != CGROUP_LIMIT_MAX)
                                        log_cgroup_compat(u, "Applying MemoryLimit=%" PRIu64 " as MemoryMax=", max);
                        }

                        cgroup_apply_unified_memory_limit(u, "memory.min", unit_get_ancestor_memory_min(u));
                        cgroup_apply_unified_memory_limit(u, "memory.low", unit_get_ancestor_memory_low(u));
                        cgroup_apply_unified_memory_limit(u, "memory.high", high);
                        cgroup_apply_unified_memory_limit(u, "memory.max", max);
                        cgroup_apply_unified_memory_limit(u, "memory.swap.max", swap_max);
                        cgroup_apply_unified_memory_limit(u, "memory.zswap.max", zswap_max);

                        (void) set_attribute_and_warn(u, "memory", "memory.oom.group", one_zero(c->memory_oom_group));
                        (void) set_attribute_and_warn(u, "memory", "memory.zswap.writeback", one_zero(c->memory_zswap_writeback));

                } else {
                        char buf[DECIMAL_STR_MAX(uint64_t) + 1];
                        uint64_t val;

                        if (unit_has_unified_memory_config(u)) {
                                val = c->memory_max;
                                if (val != CGROUP_LIMIT_MAX)
                                        log_cgroup_compat(u, "Applying MemoryMax=%" PRIu64 " as MemoryLimit=", val);
                        } else
                                val = c->memory_limit;

                        if (val == CGROUP_LIMIT_MAX)
                                strncpy(buf, "-1\n", sizeof(buf));
                        else
                                xsprintf(buf, "%" PRIu64 "\n", val);

                        (void) set_attribute_and_warn(u, "memory", "memory.limit_in_bytes", buf);
                }
        }

        /* On cgroup v2 we can apply BPF everywhere. On cgroup v1 we apply it everywhere except for the root of
         * containers, where we leave this to the manager */
        if ((apply_mask & (CGROUP_MASK_DEVICES | CGROUP_MASK_BPF_DEVICES)) &&
            (is_host_root || cg_all_unified() > 0 || !is_local_root))
                (void) cgroup_apply_devices(u);

        if (apply_mask & CGROUP_MASK_PIDS) {

                if (is_host_root) {
                        /* So, the "pids" controller does not expose anything on the root cgroup, in order not to
                         * replicate knobs exposed elsewhere needlessly. We abstract this away here however, and when
                         * the knobs of the root cgroup are modified propagate this to the relevant sysctls. There's a
                         * non-obvious asymmetry however: unlike the cgroup properties we don't really want to take
                         * exclusive ownership of the sysctls, but we still want to honour things if the user sets
                         * limits. Hence we employ sort of a one-way strategy: when the user sets a bounded limit
                         * through us it counts. When the user afterwards unsets it again (i.e. sets it to unbounded)
                         * it also counts. But if the user never set a limit through us (i.e. we are the default of
                         * "unbounded") we leave things unmodified. For this we manage a global boolean that we turn on
                         * the first time we set a limit. Note that this boolean is flushed out on manager reload,
                         * which is desirable so that there's an official way to release control of the sysctl from
                         * systemd: set the limit to unbounded and reload. */

                        if (cgroup_tasks_max_isset(&c->tasks_max)) {
                                u->manager->sysctl_pid_max_changed = true;
                                r = procfs_tasks_set_limit(cgroup_tasks_max_resolve(&c->tasks_max));
                        } else if (u->manager->sysctl_pid_max_changed)
                                r = procfs_tasks_set_limit(TASKS_MAX);
                        else
                                r = 0;
                        if (r < 0)
                                log_unit_full_errno(u, LOG_LEVEL_CGROUP_WRITE(r), r,
                                                    "Failed to write to tasks limit sysctls: %m");
                }

                /* The attribute itself is not available on the host root cgroup, and in the container case we want to
                 * leave it for the container manager. */
                if (!is_local_root) {
                        if (cgroup_tasks_max_isset(&c->tasks_max)) {
                                char buf[DECIMAL_STR_MAX(uint64_t) + 1];

                                xsprintf(buf, "%" PRIu64 "\n", cgroup_tasks_max_resolve(&c->tasks_max));
                                (void) set_attribute_and_warn(u, "pids", "pids.max", buf);
                        } else
                                (void) set_attribute_and_warn(u, "pids", "pids.max", "max\n");
                }
        }

        if (apply_mask & CGROUP_MASK_BPF_FIREWALL)
                cgroup_apply_firewall(u);

        if (apply_mask & CGROUP_MASK_BPF_FOREIGN)
                cgroup_apply_bpf_foreign_program(u);

        if (apply_mask & CGROUP_MASK_BPF_SOCKET_BIND)
                cgroup_apply_socket_bind(u);

        if (apply_mask & CGROUP_MASK_BPF_RESTRICT_NETWORK_INTERFACES)
                cgroup_apply_restrict_network_interfaces(u);

        unit_modify_nft_set(u, /* add = */ true);
}

static bool unit_get_needs_bpf_firewall(Unit *u) {
        CGroupContext *c;
        assert(u);

        c = unit_get_cgroup_context(u);
        if (!c)
                return false;

        if (c->ip_accounting ||
            !set_isempty(c->ip_address_allow) ||
            !set_isempty(c->ip_address_deny) ||
            c->ip_filters_ingress ||
            c->ip_filters_egress)
                return true;

        /* If any parent slice has an IP access list defined, it applies too */
        for (Unit *p = UNIT_GET_SLICE(u); p; p = UNIT_GET_SLICE(p)) {
                c = unit_get_cgroup_context(p);
                if (!c)
                        return false;

                if (!set_isempty(c->ip_address_allow) ||
                    !set_isempty(c->ip_address_deny))
                        return true;
        }

        return false;
}

static bool unit_get_needs_bpf_foreign_program(Unit *u) {
        CGroupContext *c;
        assert(u);

        c = unit_get_cgroup_context(u);
        if (!c)
                return false;

        return !!c->bpf_foreign_programs;
}

static bool unit_get_needs_socket_bind(Unit *u) {
        CGroupContext *c;
        assert(u);

        c = unit_get_cgroup_context(u);
        if (!c)
                return false;

        return c->socket_bind_allow || c->socket_bind_deny;
}

static bool unit_get_needs_restrict_network_interfaces(Unit *u) {
        CGroupContext *c;
        assert(u);

        c = unit_get_cgroup_context(u);
        if (!c)
                return false;

        return !set_isempty(c->restrict_network_interfaces);
}

static CGroupMask unit_get_cgroup_mask(Unit *u) {
        CGroupMask mask = 0;
        CGroupContext *c;

        assert(u);

        assert_se(c = unit_get_cgroup_context(u));

        /* Figure out which controllers we need, based on the cgroup context object */

        if (c->cpu_accounting)
                mask |= get_cpu_accounting_mask();

        if (cgroup_context_has_cpu_weight(c) ||
            cgroup_context_has_cpu_shares(c) ||
            c->cpu_quota_per_sec_usec != USEC_INFINITY)
                mask |= CGROUP_MASK_CPU;

        if (cgroup_context_has_allowed_cpus(c) || cgroup_context_has_allowed_mems(c))
                mask |= CGROUP_MASK_CPUSET;

        if (cgroup_context_has_io_config(c) || cgroup_context_has_blockio_config(c))
                mask |= CGROUP_MASK_IO | CGROUP_MASK_BLKIO;

        if (c->memory_accounting ||
            c->memory_limit != CGROUP_LIMIT_MAX ||
            unit_has_unified_memory_config(u))
                mask |= CGROUP_MASK_MEMORY;

        if (c->device_allow ||
            c->device_policy != CGROUP_DEVICE_POLICY_AUTO)
                mask |= CGROUP_MASK_DEVICES | CGROUP_MASK_BPF_DEVICES;

        if (c->tasks_accounting ||
            cgroup_tasks_max_isset(&c->tasks_max))
                mask |= CGROUP_MASK_PIDS;

        return CGROUP_MASK_EXTEND_JOINED(mask);
}

static CGroupMask unit_get_bpf_mask(Unit *u) {
        CGroupMask mask = 0;

        /* Figure out which controllers we need, based on the cgroup context, possibly taking into account children
         * too. */

        if (unit_get_needs_bpf_firewall(u))
                mask |= CGROUP_MASK_BPF_FIREWALL;

        if (unit_get_needs_bpf_foreign_program(u))
                mask |= CGROUP_MASK_BPF_FOREIGN;

        if (unit_get_needs_socket_bind(u))
                mask |= CGROUP_MASK_BPF_SOCKET_BIND;

        if (unit_get_needs_restrict_network_interfaces(u))
                mask |= CGROUP_MASK_BPF_RESTRICT_NETWORK_INTERFACES;

        return mask;
}

CGroupMask unit_get_own_mask(Unit *u) {
        CGroupContext *c;

        /* Returns the mask of controllers the unit needs for itself. If a unit is not properly loaded, return an empty
         * mask, as we shouldn't reflect it in the cgroup hierarchy then. */

        if (u->load_state != UNIT_LOADED)
                return 0;

        c = unit_get_cgroup_context(u);
        if (!c)
                return 0;

        return unit_get_cgroup_mask(u) | unit_get_bpf_mask(u) | unit_get_delegate_mask(u);
}

CGroupMask unit_get_delegate_mask(Unit *u) {
        CGroupContext *c;

        /* If delegation is turned on, then turn on selected controllers, unless we are on the legacy hierarchy and the
         * process we fork into is known to drop privileges, and hence shouldn't get access to the controllers.
         *
         * Note that on the unified hierarchy it is safe to delegate controllers to unprivileged services. */

        if (!unit_cgroup_delegate(u))
                return 0;

        if (cg_all_unified() <= 0) {
                ExecContext *e;

                e = unit_get_exec_context(u);
                if (e && !exec_context_maintains_privileges(e))
                        return 0;
        }

        assert_se(c = unit_get_cgroup_context(u));
        return CGROUP_MASK_EXTEND_JOINED(c->delegate_controllers);
}

static CGroupMask unit_get_subtree_mask(Unit *u) {

        /* Returns the mask of this subtree, meaning of the group
         * itself and its children. */

        return unit_get_own_mask(u) | unit_get_members_mask(u);
}

CGroupMask unit_get_members_mask(Unit *u) {
        assert(u);

        /* Returns the mask of controllers all of the unit's children require, merged */

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (crt && crt->cgroup_members_mask_valid)
                return crt->cgroup_members_mask; /* Use cached value if possible */

        CGroupMask m = 0;
        if (u->type == UNIT_SLICE) {
                Unit *member;

                UNIT_FOREACH_DEPENDENCY(member, u, UNIT_ATOM_SLICE_OF)
                        m |= unit_get_subtree_mask(member); /* note that this calls ourselves again, for the children */
        }

        if (crt) {
                crt->cgroup_members_mask = m;
                crt->cgroup_members_mask_valid = true;
        }

        return m;
}

CGroupMask unit_get_siblings_mask(Unit *u) {
        Unit *slice;
        assert(u);

        /* Returns the mask of controllers all of the unit's siblings
         * require, i.e. the members mask of the unit's parent slice
         * if there is one. */

        slice = UNIT_GET_SLICE(u);
        if (slice)
                return unit_get_members_mask(slice);

        return unit_get_subtree_mask(u); /* we are the top-level slice */
}

static CGroupMask unit_get_disable_mask(Unit *u) {
        CGroupContext *c;

        c = unit_get_cgroup_context(u);
        if (!c)
                return 0;

        return c->disable_controllers;
}

CGroupMask unit_get_ancestor_disable_mask(Unit *u) {
        CGroupMask mask;
        Unit *slice;

        assert(u);
        mask = unit_get_disable_mask(u);

        /* Returns the mask of controllers which are marked as forcibly
         * disabled in any ancestor unit or the unit in question. */

        slice = UNIT_GET_SLICE(u);
        if (slice)
                mask |= unit_get_ancestor_disable_mask(slice);

        return mask;
}

CGroupMask unit_get_target_mask(Unit *u) {
        CGroupMask own_mask, mask;

        /* This returns the cgroup mask of all controllers to enable for a specific cgroup, i.e. everything
         * it needs itself, plus all that its children need, plus all that its siblings need. This is
         * primarily useful on the legacy cgroup hierarchy, where we need to duplicate each cgroup in each
         * hierarchy that shall be enabled for it. */

        own_mask = unit_get_own_mask(u);

        if (own_mask & CGROUP_MASK_BPF_FIREWALL & ~u->manager->cgroup_supported)
                emit_bpf_firewall_warning(u);

        mask = own_mask | unit_get_members_mask(u) | unit_get_siblings_mask(u);

        mask &= u->manager->cgroup_supported;
        mask &= ~unit_get_ancestor_disable_mask(u);

        return mask;
}

CGroupMask unit_get_enable_mask(Unit *u) {
        CGroupMask mask;

        /* This returns the cgroup mask of all controllers to enable
         * for the children of a specific cgroup. This is primarily
         * useful for the unified cgroup hierarchy, where each cgroup
         * controls which controllers are enabled for its children. */

        mask = unit_get_members_mask(u);
        mask &= u->manager->cgroup_supported;
        mask &= ~unit_get_ancestor_disable_mask(u);

        return mask;
}

void unit_invalidate_cgroup_members_masks(Unit *u) {
        Unit *slice;

        assert(u);

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt)
                return;

        /* Recurse invalidate the member masks cache all the way up the tree */
        crt->cgroup_members_mask_valid = false;

        slice = UNIT_GET_SLICE(u);
        if (slice)
                unit_invalidate_cgroup_members_masks(slice);
}

const char *unit_get_realized_cgroup_path(Unit *u, CGroupMask mask) {

        /* Returns the realized cgroup path of the specified unit where all specified controllers are available. */

        while (u) {
                CGroupRuntime *crt = unit_get_cgroup_runtime(u);
                if (crt &&
                    crt->cgroup_path &&
                    crt->cgroup_realized &&
                    FLAGS_SET(crt->cgroup_realized_mask, mask))
                        return crt->cgroup_path;

                u = UNIT_GET_SLICE(u);
        }

        return NULL;
}

static const char *migrate_callback(CGroupMask mask, void *userdata) {
        /* If not realized at all, migrate to root ("").
         * It may happen if we're upgrading from older version that didn't clean up.
         */
        return strempty(unit_get_realized_cgroup_path(userdata, mask));
}

int unit_default_cgroup_path(const Unit *u, char **ret) {
        _cleanup_free_ char *p = NULL;
        int r;

        assert(u);
        assert(ret);

        if (unit_has_name(u, SPECIAL_ROOT_SLICE))
                p = strdup(u->manager->cgroup_root);
        else {
                _cleanup_free_ char *escaped = NULL, *slice_path = NULL;
                Unit *slice;

                slice = UNIT_GET_SLICE(u);
                if (slice && !unit_has_name(slice, SPECIAL_ROOT_SLICE)) {
                        r = cg_slice_to_path(slice->id, &slice_path);
                        if (r < 0)
                                return r;
                }

                r = cg_escape(u->id, &escaped);
                if (r < 0)
                        return r;

                p = path_join(empty_to_root(u->manager->cgroup_root), slice_path, escaped);
        }
        if (!p)
                return -ENOMEM;

        *ret = TAKE_PTR(p);
        return 0;
}

int unit_set_cgroup_path(Unit *u, const char *path) {
        _cleanup_free_ char *p = NULL;
        CGroupRuntime *crt;
        int r;

        assert(u);

        crt = unit_get_cgroup_runtime(u);

        if (crt && streq_ptr(crt->cgroup_path, path))
                return 0;

        unit_release_cgroup(u);

        crt = unit_setup_cgroup_runtime(u);
        if (!crt)
                return -ENOMEM;

        if (path) {
                p = strdup(path);
                if (!p)
                        return -ENOMEM;

                r = hashmap_put(u->manager->cgroup_unit, p, u);
                if (r < 0)
                        return r;
        }

        assert(!crt->cgroup_path);
        crt->cgroup_path = TAKE_PTR(p);

        return 1;
}

int unit_watch_cgroup(Unit *u) {
        _cleanup_free_ char *events = NULL;
        int r;

        assert(u);

        /* Watches the "cgroups.events" attribute of this unit's cgroup for "empty" events, but only if
         * cgroupv2 is available. */

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return 0;

        if (crt->cgroup_control_inotify_wd >= 0)
                return 0;

        /* Only applies to the unified hierarchy */
        r = cg_unified_controller(SYSTEMD_CGROUP_CONTROLLER);
        if (r < 0)
                return log_error_errno(r, "Failed to determine whether the name=systemd hierarchy is unified: %m");
        if (r == 0)
                return 0;

        /* No point in watch the top-level slice, it's never going to run empty. */
        if (unit_has_name(u, SPECIAL_ROOT_SLICE))
                return 0;

        r = hashmap_ensure_allocated(&u->manager->cgroup_control_inotify_wd_unit, &trivial_hash_ops);
        if (r < 0)
                return log_oom();

        r = cg_get_path(SYSTEMD_CGROUP_CONTROLLER, crt->cgroup_path, "cgroup.events", &events);
        if (r < 0)
                return log_oom();

        crt->cgroup_control_inotify_wd = inotify_add_watch(u->manager->cgroup_inotify_fd, events, IN_MODIFY);
        if (crt->cgroup_control_inotify_wd < 0) {

                if (errno == ENOENT) /* If the directory is already gone we don't need to track it, so this
                                      * is not an error */
                        return 0;

                return log_unit_error_errno(u, errno, "Failed to add control inotify watch descriptor for control group %s: %m", empty_to_root(crt->cgroup_path));
        }

        r = hashmap_put(u->manager->cgroup_control_inotify_wd_unit, INT_TO_PTR(crt->cgroup_control_inotify_wd), u);
        if (r < 0)
                return log_unit_error_errno(u, r, "Failed to add control inotify watch descriptor for control group %s to hash map: %m", empty_to_root(crt->cgroup_path));

        return 0;
}

int unit_watch_cgroup_memory(Unit *u) {
        _cleanup_free_ char *events = NULL;
        int r;

        assert(u);

        /* Watches the "memory.events" attribute of this unit's cgroup for "oom_kill" events, but only if
         * cgroupv2 is available. */

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return 0;

        CGroupContext *c = unit_get_cgroup_context(u);
        if (!c)
                return 0;

        /* The "memory.events" attribute is only available if the memory controller is on. Let's hence tie
         * this to memory accounting, in a way watching for OOM kills is a form of memory accounting after
         * all. */
        if (!c->memory_accounting)
                return 0;

        /* Don't watch inner nodes, as the kernel doesn't report oom_kill events recursively currently, and
         * we also don't want to generate a log message for each parent cgroup of a process. */
        if (u->type == UNIT_SLICE)
                return 0;

        if (crt->cgroup_memory_inotify_wd >= 0)
                return 0;

        /* Only applies to the unified hierarchy */
        r = cg_all_unified();
        if (r < 0)
                return log_error_errno(r, "Failed to determine whether the memory controller is unified: %m");
        if (r == 0)
                return 0;

        r = hashmap_ensure_allocated(&u->manager->cgroup_memory_inotify_wd_unit, &trivial_hash_ops);
        if (r < 0)
                return log_oom();

        r = cg_get_path(SYSTEMD_CGROUP_CONTROLLER, crt->cgroup_path, "memory.events", &events);
        if (r < 0)
                return log_oom();

        crt->cgroup_memory_inotify_wd = inotify_add_watch(u->manager->cgroup_inotify_fd, events, IN_MODIFY);
        if (crt->cgroup_memory_inotify_wd < 0) {

                if (errno == ENOENT) /* If the directory is already gone we don't need to track it, so this
                                      * is not an error */
                        return 0;

                return log_unit_error_errno(u, errno, "Failed to add memory inotify watch descriptor for control group %s: %m", empty_to_root(crt->cgroup_path));
        }

        r = hashmap_put(u->manager->cgroup_memory_inotify_wd_unit, INT_TO_PTR(crt->cgroup_memory_inotify_wd), u);
        if (r < 0)
                return log_unit_error_errno(u, r, "Failed to add memory inotify watch descriptor for control group %s to hash map: %m", empty_to_root(crt->cgroup_path));

        return 0;
}

int unit_pick_cgroup_path(Unit *u) {
        _cleanup_free_ char *path = NULL;
        int r;

        assert(u);

        if (!UNIT_HAS_CGROUP_CONTEXT(u))
                return -EINVAL;

        CGroupRuntime *crt = unit_setup_cgroup_runtime(u);
        if (!crt)
                return -ENOMEM;
        if (crt->cgroup_path)
                return 0;

        r = unit_default_cgroup_path(u, &path);
        if (r < 0)
                return log_unit_error_errno(u, r, "Failed to generate default cgroup path: %m");

        r = unit_set_cgroup_path(u, path);
        if (r == -EEXIST)
                return log_unit_error_errno(u, r, "Control group %s exists already.", empty_to_root(path));
        if (r < 0)
                return log_unit_error_errno(u, r, "Failed to set unit's control group path to %s: %m", empty_to_root(path));

        return 0;
}

static int unit_update_cgroup(
                Unit *u,
                CGroupMask target_mask,
                CGroupMask enable_mask,
                ManagerState state) {

        bool created, is_root_slice;
        CGroupMask migrate_mask = 0;
        _cleanup_free_ char *cgroup_full_path = NULL;
        int r;

        assert(u);

        if (!UNIT_HAS_CGROUP_CONTEXT(u))
                return 0;

        if (u->freezer_state != FREEZER_RUNNING)
                return log_unit_error_errno(u, SYNTHETIC_ERRNO(EBUSY), "Cannot realize cgroup for frozen unit.");

        /* Figure out our cgroup path */
        r = unit_pick_cgroup_path(u);
        if (r < 0)
                return r;

        CGroupRuntime *crt = ASSERT_PTR(unit_get_cgroup_runtime(u));

        /* First, create our own group */
        r = cg_create_everywhere(u->manager->cgroup_supported, target_mask, crt->cgroup_path);
        if (r < 0)
                return log_unit_error_errno(u, r, "Failed to create cgroup %s: %m", empty_to_root(crt->cgroup_path));
        created = r;

        if (cg_unified_controller(SYSTEMD_CGROUP_CONTROLLER) > 0) {
                uint64_t cgroup_id = 0;

                r = cg_get_path(SYSTEMD_CGROUP_CONTROLLER, crt->cgroup_path, NULL, &cgroup_full_path);
                if (r == 0) {
                        r = cg_path_get_cgroupid(cgroup_full_path, &cgroup_id);
                        if (r < 0)
                                log_unit_full_errno(u, ERRNO_IS_NOT_SUPPORTED(r) ? LOG_DEBUG : LOG_WARNING, r,
                                                    "Failed to get cgroup ID of cgroup %s, ignoring: %m", cgroup_full_path);
                } else
                        log_unit_warning_errno(u, r, "Failed to get full cgroup path on cgroup %s, ignoring: %m", empty_to_root(crt->cgroup_path));

                crt->cgroup_id = cgroup_id;
        }

        /* Start watching it */
        (void) unit_watch_cgroup(u);
        (void) unit_watch_cgroup_memory(u);

        /* For v2 we preserve enabled controllers in delegated units, adjust others,
         * for v1 we figure out which controller hierarchies need migration. */
        if (created || !crt->cgroup_realized || !unit_cgroup_delegate(u)) {
                CGroupMask result_mask = 0;

                /* Enable all controllers we need */
                r = cg_enable_everywhere(u->manager->cgroup_supported, enable_mask, crt->cgroup_path, &result_mask);
                if (r < 0)
                        log_unit_warning_errno(u, r, "Failed to enable/disable controllers on cgroup %s, ignoring: %m", empty_to_root(crt->cgroup_path));

                /* Remember what's actually enabled now */
                crt->cgroup_enabled_mask = result_mask;

                migrate_mask = crt->cgroup_realized_mask ^ target_mask;
        }

        /* Keep track that this is now realized */
        crt->cgroup_realized = true;
        crt->cgroup_realized_mask = target_mask;

        /* Migrate processes in controller hierarchies both downwards (enabling) and upwards (disabling).
         *
         * Unnecessary controller cgroups are trimmed (after emptied by upward migration).
         * We perform migration also with whole slices for cases when users don't care about leave
         * granularity. Since delegated_mask is subset of target mask, we won't trim slice subtree containing
         * delegated units.
         */
        if (cg_all_unified() == 0) {
                r = cg_migrate_v1_controllers(u->manager->cgroup_supported, migrate_mask, crt->cgroup_path, migrate_callback, u);
                if (r < 0)
                        log_unit_warning_errno(u, r, "Failed to migrate controller cgroups from %s, ignoring: %m", empty_to_root(crt->cgroup_path));

                is_root_slice = unit_has_name(u, SPECIAL_ROOT_SLICE);
                r = cg_trim_v1_controllers(u->manager->cgroup_supported, ~target_mask, crt->cgroup_path, !is_root_slice);
                if (r < 0)
                        log_unit_warning_errno(u, r, "Failed to delete controller cgroups %s, ignoring: %m", empty_to_root(crt->cgroup_path));
        }

        /* Set attributes */
        cgroup_context_apply(u, target_mask, state);
        cgroup_xattr_apply(u);

        /* For most units we expect that memory monitoring is set up before the unit is started and we won't
         * touch it after. For PID 1 this is different though, because we couldn't possibly do that given
         * that PID 1 runs before init.scope is even set up. Hence, whenever init.scope is realized, let's
         * try to open the memory pressure interface anew. */
        if (unit_has_name(u, SPECIAL_INIT_SCOPE))
                (void) manager_setup_memory_pressure_event_source(u->manager);

        return 0;
}

static int unit_attach_pid_to_cgroup_via_bus(Unit *u, pid_t pid, const char *suffix_path) {
        _cleanup_(sd_bus_error_free) sd_bus_error error = SD_BUS_ERROR_NULL;
        char *pp;
        int r;

        assert(u);

        if (MANAGER_IS_SYSTEM(u->manager))
                return -EINVAL;

        if (!u->manager->system_bus)
                return -EIO;

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return -EOWNERDEAD;

        /* Determine this unit's cgroup path relative to our cgroup root */
        pp = path_startswith(crt->cgroup_path, u->manager->cgroup_root);
        if (!pp)
                return -EINVAL;

        pp = strjoina("/", pp, suffix_path);
        path_simplify(pp);

        r = bus_call_method(u->manager->system_bus,
                            bus_systemd_mgr,
                            "AttachProcessesToUnit",
                            &error, NULL,
                            "ssau",
                            NULL /* empty unit name means client's unit, i.e. us */, pp, 1, (uint32_t) pid);
        if (r < 0)
                return log_unit_debug_errno(u, r, "Failed to attach unit process " PID_FMT " via the bus: %s", pid, bus_error_message(&error, r));

        return 0;
}

int unit_attach_pids_to_cgroup(Unit *u, Set *pids, const char *suffix_path) {
        _cleanup_free_ char *joined = NULL;
        CGroupMask delegated_mask;
        const char *p;
        PidRef *pid;
        int ret, r;

        assert(u);

        if (!UNIT_HAS_CGROUP_CONTEXT(u))
                return -EINVAL;

        if (set_isempty(pids))
                return 0;

        /* Load any custom firewall BPF programs here once to test if they are existing and actually loadable.
         * Fail here early since later errors in the call chain unit_realize_cgroup to cgroup_context_apply are ignored. */
        r = bpf_firewall_load_custom(u);
        if (r < 0)
                return r;

        r = unit_realize_cgroup(u);
        if (r < 0)
                return r;

        CGroupRuntime *crt = ASSERT_PTR(unit_get_cgroup_runtime(u));

        if (isempty(suffix_path))
                p = crt->cgroup_path;
        else {
                joined = path_join(crt->cgroup_path, suffix_path);
                if (!joined)
                        return -ENOMEM;

                p = joined;
        }

        delegated_mask = unit_get_delegate_mask(u);

        ret = 0;
        SET_FOREACH(pid, pids) {

                /* Unfortunately we cannot add pids by pidfd to a cgroup. Hence we have to use PIDs instead,
                 * which of course is racy. Let's shorten the race a bit though, and re-validate the PID
                 * before we use it */
                r = pidref_verify(pid);
                if (r < 0) {
                        log_unit_info_errno(u, r, "PID " PID_FMT " vanished before we could move it to target cgroup '%s', skipping: %m", pid->pid, empty_to_root(p));
                        continue;
                }

                /* First, attach the PID to the main cgroup hierarchy */
                r = cg_attach(SYSTEMD_CGROUP_CONTROLLER, p, pid->pid);
                if (r < 0) {
                        bool again = MANAGER_IS_USER(u->manager) && ERRNO_IS_PRIVILEGE(r);

                        log_unit_full_errno(u, again ? LOG_DEBUG : LOG_INFO,  r,
                                            "Couldn't move process "PID_FMT" to%s requested cgroup '%s': %m",
                                            pid->pid, again ? " directly" : "", empty_to_root(p));

                        if (again) {
                                int z;

                                /* If we are in a user instance, and we can't move the process ourselves due
                                 * to permission problems, let's ask the system instance about it instead.
                                 * Since it's more privileged it might be able to move the process across the
                                 * leaves of a subtree whose top node is not owned by us. */

                                z = unit_attach_pid_to_cgroup_via_bus(u, pid->pid, suffix_path);
                                if (z < 0)
                                        log_unit_info_errno(u, z, "Couldn't move process "PID_FMT" to requested cgroup '%s' (directly or via the system bus): %m", pid->pid, empty_to_root(p));
                                else {
                                        if (ret >= 0)
                                                ret++; /* Count successful additions */
                                        continue; /* When the bus thing worked via the bus we are fully done for this PID. */
                                }
                        }

                        if (ret >= 0)
                                ret = r; /* Remember first error */

                        continue;
                } else if (ret >= 0)
                        ret++; /* Count successful additions */

                r = cg_all_unified();
                if (r < 0)
                        return r;
                if (r > 0)
                        continue;

                /* In the legacy hierarchy, attach the process to the request cgroup if possible, and if not to the
                 * innermost realized one */

                for (CGroupController c = 0; c < _CGROUP_CONTROLLER_MAX; c++) {
                        CGroupMask bit = CGROUP_CONTROLLER_TO_MASK(c);
                        const char *realized;

                        if (!(u->manager->cgroup_supported & bit))
                                continue;

                        /* If this controller is delegated and realized, honour the caller's request for the cgroup suffix. */
                        if (delegated_mask & crt->cgroup_realized_mask & bit) {
                                r = cg_attach(cgroup_controller_to_string(c), p, pid->pid);
                                if (r >= 0)
                                        continue; /* Success! */

                                log_unit_debug_errno(u, r, "Failed to attach PID " PID_FMT " to requested cgroup %s in controller %s, falling back to unit's cgroup: %m",
                                                     pid->pid, empty_to_root(p), cgroup_controller_to_string(c));
                        }

                        /* So this controller is either not delegate or realized, or something else weird happened. In
                         * that case let's attach the PID at least to the closest cgroup up the tree that is
                         * realized. */
                        realized = unit_get_realized_cgroup_path(u, bit);
                        if (!realized)
                                continue; /* Not even realized in the root slice? Then let's not bother */

                        r = cg_attach(cgroup_controller_to_string(c), realized, pid->pid);
                        if (r < 0)
                                log_unit_debug_errno(u, r, "Failed to attach PID " PID_FMT " to realized cgroup %s in controller %s, ignoring: %m",
                                                     pid->pid, realized, cgroup_controller_to_string(c));
                }
        }

        return ret;
}

static bool unit_has_mask_realized(
                Unit *u,
                CGroupMask target_mask,
                CGroupMask enable_mask) {

        assert(u);

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt)
                return false;

        /* Returns true if this unit is fully realized. We check four things:
         *
         * 1. Whether the cgroup was created at all
         * 2. Whether the cgroup was created in all the hierarchies we need it to be created in (in case of cgroup v1)
         * 3. Whether the cgroup has all the right controllers enabled (in case of cgroup v2)
         * 4. Whether the invalidation mask is currently zero
         *
         * If you wonder why we mask the target realization and enable mask with CGROUP_MASK_V1/CGROUP_MASK_V2: note
         * that there are three sets of bitmasks: CGROUP_MASK_V1 (for real cgroup v1 controllers), CGROUP_MASK_V2 (for
         * real cgroup v2 controllers) and CGROUP_MASK_BPF (for BPF-based pseudo-controllers). Now, cgroup_realized_mask
         * is only matters for cgroup v1 controllers, and cgroup_enabled_mask only used for cgroup v2, and if they
         * differ in the others, we don't really care. (After all, the cgroup_enabled_mask tracks with controllers are
         * enabled through cgroup.subtree_control, and since the BPF pseudo-controllers don't show up there, they
         * simply don't matter. */

        return crt->cgroup_realized &&
                ((crt->cgroup_realized_mask ^ target_mask) & CGROUP_MASK_V1) == 0 &&
                ((crt->cgroup_enabled_mask ^ enable_mask) & CGROUP_MASK_V2) == 0 &&
                crt->cgroup_invalidated_mask == 0;
}

static bool unit_has_mask_disables_realized(
                Unit *u,
                CGroupMask target_mask,
                CGroupMask enable_mask) {

        assert(u);

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt)
                return true;

        /* Returns true if all controllers which should be disabled are indeed disabled.
         *
         * Unlike unit_has_mask_realized, we don't care what was enabled, only that anything we want to remove is
         * already removed. */

        return !crt->cgroup_realized ||
                (FLAGS_SET(crt->cgroup_realized_mask, target_mask & CGROUP_MASK_V1) &&
                 FLAGS_SET(crt->cgroup_enabled_mask, enable_mask & CGROUP_MASK_V2));
}

static bool unit_has_mask_enables_realized(
                Unit *u,
                CGroupMask target_mask,
                CGroupMask enable_mask) {

        assert(u);

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt)
                return false;

        /* Returns true if all controllers which should be enabled are indeed enabled.
         *
         * Unlike unit_has_mask_realized, we don't care about the controllers that are not present, only that anything
         * we want to add is already added. */

        return crt->cgroup_realized &&
                ((crt->cgroup_realized_mask | target_mask) & CGROUP_MASK_V1) == (crt->cgroup_realized_mask & CGROUP_MASK_V1) &&
                ((crt->cgroup_enabled_mask | enable_mask) & CGROUP_MASK_V2) == (crt->cgroup_enabled_mask & CGROUP_MASK_V2);
}

void unit_add_to_cgroup_realize_queue(Unit *u) {
        assert(u);

        if (u->in_cgroup_realize_queue)
                return;

        LIST_APPEND(cgroup_realize_queue, u->manager->cgroup_realize_queue, u);
        u->in_cgroup_realize_queue = true;
}

static void unit_remove_from_cgroup_realize_queue(Unit *u) {
        assert(u);

        if (!u->in_cgroup_realize_queue)
                return;

        LIST_REMOVE(cgroup_realize_queue, u->manager->cgroup_realize_queue, u);
        u->in_cgroup_realize_queue = false;
}

/* Controllers can only be enabled breadth-first, from the root of the
 * hierarchy downwards to the unit in question. */
static int unit_realize_cgroup_now_enable(Unit *u, ManagerState state) {
        CGroupMask target_mask, enable_mask, new_target_mask, new_enable_mask;
        Unit *slice;
        int r;

        assert(u);

        /* First go deal with this unit's parent, or we won't be able to enable
         * any new controllers at this layer. */
        slice = UNIT_GET_SLICE(u);
        if (slice) {
                r = unit_realize_cgroup_now_enable(slice, state);
                if (r < 0)
                        return r;
        }

        target_mask = unit_get_target_mask(u);
        enable_mask = unit_get_enable_mask(u);

        /* We can only enable in this direction, don't try to disable anything.
         */
        if (unit_has_mask_enables_realized(u, target_mask, enable_mask))
                return 0;

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);

        new_target_mask = (crt ? crt->cgroup_realized_mask : 0) | target_mask;
        new_enable_mask = (crt ? crt->cgroup_enabled_mask : 0) | enable_mask;

        return unit_update_cgroup(u, new_target_mask, new_enable_mask, state);
}

/* Controllers can only be disabled depth-first, from the leaves of the
 * hierarchy upwards to the unit in question. */
static int unit_realize_cgroup_now_disable(Unit *u, ManagerState state) {
        Unit *m;

        assert(u);

        if (u->type != UNIT_SLICE)
                return 0;

        UNIT_FOREACH_DEPENDENCY(m, u, UNIT_ATOM_SLICE_OF) {
                CGroupMask target_mask, enable_mask, new_target_mask, new_enable_mask;
                int r;

                CGroupRuntime *rt = unit_get_cgroup_runtime(m);
                if (!rt)
                        continue;

                /* The cgroup for this unit might not actually be fully realised yet, in which case it isn't
                 * holding any controllers open anyway. */
                if (!rt->cgroup_realized)
                        continue;

                /* We must disable those below us first in order to release the controller. */
                if (m->type == UNIT_SLICE)
                        (void) unit_realize_cgroup_now_disable(m, state);

                target_mask = unit_get_target_mask(m);
                enable_mask = unit_get_enable_mask(m);

                /* We can only disable in this direction, don't try to enable anything. */
                if (unit_has_mask_disables_realized(m, target_mask, enable_mask))
                        continue;

                new_target_mask = rt->cgroup_realized_mask & target_mask;
                new_enable_mask = rt->cgroup_enabled_mask & enable_mask;

                r = unit_update_cgroup(m, new_target_mask, new_enable_mask, state);
                if (r < 0)
                        return r;
        }

        return 0;
}

/* Check if necessary controllers and attributes for a unit are in place.
 *
 * - If so, do nothing.
 * - If not, create paths, move processes over, and set attributes.
 *
 * Controllers can only be *enabled* in a breadth-first way, and *disabled* in
 * a depth-first way. As such the process looks like this:
 *
 * Suppose we have a cgroup hierarchy which looks like this:
 *
 *             root
 *            /    \
 *           /      \
 *          /        \
 *         a          b
 *        / \        / \
 *       /   \      /   \
 *      c     d    e     f
 *     / \   / \  / \   / \
 *     h i   j k  l m   n o
 *
 * 1. We want to realise cgroup "d" now.
 * 2. cgroup "a" has DisableControllers=cpu in the associated unit.
 * 3. cgroup "k" just started requesting the memory controller.
 *
 * To make this work we must do the following in order:
 *
 * 1. Disable CPU controller in k, j
 * 2. Disable CPU controller in d
 * 3. Enable memory controller in root
 * 4. Enable memory controller in a
 * 5. Enable memory controller in d
 * 6. Enable memory controller in k
 *
 * Notice that we need to touch j in one direction, but not the other. We also
 * don't go beyond d when disabling -- it's up to "a" to get realized if it
 * wants to disable further. The basic rules are therefore:
 *
 * - If you're disabling something, you need to realise all of the cgroups from
 *   your recursive descendants to the root. This starts from the leaves.
 * - If you're enabling something, you need to realise from the root cgroup
 *   downwards, but you don't need to iterate your recursive descendants.
 *
 * Returns 0 on success and < 0 on failure. */
static int unit_realize_cgroup_now(Unit *u, ManagerState state) {
        CGroupMask target_mask, enable_mask;
        Unit *slice;
        int r;

        assert(u);

        unit_remove_from_cgroup_realize_queue(u);

        target_mask = unit_get_target_mask(u);
        enable_mask = unit_get_enable_mask(u);

        if (unit_has_mask_realized(u, target_mask, enable_mask))
                return 0;

        /* Disable controllers below us, if there are any */
        r = unit_realize_cgroup_now_disable(u, state);
        if (r < 0)
                return r;

        /* Enable controllers above us, if there are any */
        slice = UNIT_GET_SLICE(u);
        if (slice) {
                r = unit_realize_cgroup_now_enable(slice, state);
                if (r < 0)
                        return r;
        }

        /* Now actually deal with the cgroup we were trying to realise and set attributes */
        r = unit_update_cgroup(u, target_mask, enable_mask, state);
        if (r < 0)
                return r;

        CGroupRuntime *crt = ASSERT_PTR(unit_get_cgroup_runtime(u));

        /* Now, reset the invalidation mask */
        crt->cgroup_invalidated_mask = 0;
        return 0;
}

unsigned manager_dispatch_cgroup_realize_queue(Manager *m) {
        ManagerState state;
        unsigned n = 0;
        Unit *i;
        int r;

        assert(m);

        state = manager_state(m);

        while ((i = m->cgroup_realize_queue)) {
                assert(i->in_cgroup_realize_queue);

                if (UNIT_IS_INACTIVE_OR_FAILED(unit_active_state(i))) {
                        /* Maybe things changed, and the unit is not actually active anymore? */
                        unit_remove_from_cgroup_realize_queue(i);
                        continue;
                }

                r = unit_realize_cgroup_now(i, state);
                if (r < 0)
                        log_warning_errno(r, "Failed to realize cgroups for queued unit %s, ignoring: %m", i->id);

                n++;
        }

        return n;
}

void unit_add_family_to_cgroup_realize_queue(Unit *u) {
        assert(u);
        assert(u->type == UNIT_SLICE);

        /* Family of a unit for is defined as (immediate) children of the unit and immediate children of all
         * its ancestors.
         *
         * Ideally we would enqueue ancestor path only (bottom up). However, on cgroup-v1 scheduling becomes
         * very weird if two units that own processes reside in the same slice, but one is realized in the
         * "cpu" hierarchy and one is not (for example because one has CPUWeight= set and the other does
         * not), because that means individual processes need to be scheduled against whole cgroups. Let's
         * avoid this asymmetry by always ensuring that siblings of a unit are always realized in their v1
         * controller hierarchies too (if unit requires the controller to be realized).
         *
         * The function must invalidate cgroup_members_mask of all ancestors in order to calculate up to date
         * masks. */

        do {
                CGroupRuntime *crt = unit_get_cgroup_runtime(u);

                /* Children of u likely changed when we're called */
                if (crt)
                        crt->cgroup_members_mask_valid = false;

                Unit *m;
                UNIT_FOREACH_DEPENDENCY(m, u, UNIT_ATOM_SLICE_OF) {

                        /* No point in doing cgroup application for units without active processes. */
                        if (UNIT_IS_INACTIVE_OR_FAILED(unit_active_state(m)))
                                continue;

                        /* We only enqueue siblings if they were realized once at least, in the main
                         * hierarchy. */
                        crt = unit_get_cgroup_runtime(m);
                        if (!crt || !crt->cgroup_realized)
                                continue;

                        /* If the unit doesn't need any new controllers and has current ones
                         * realized, it doesn't need any changes. */
                        if (unit_has_mask_realized(m,
                                                   unit_get_target_mask(m),
                                                   unit_get_enable_mask(m)))
                                continue;

                        unit_add_to_cgroup_realize_queue(m);
                }

                /* Parent comes after children */
                unit_add_to_cgroup_realize_queue(u);

                u = UNIT_GET_SLICE(u);
        } while (u);
}

int unit_realize_cgroup(Unit *u) {
        Unit *slice;

        assert(u);

        if (!UNIT_HAS_CGROUP_CONTEXT(u))
                return 0;

        /* So, here's the deal: when realizing the cgroups for this unit, we need to first create all
         * parents, but there's more actually: for the weight-based controllers we also need to make sure
         * that all our siblings (i.e. units that are in the same slice as we are) have cgroups, too.  On the
         * other hand, when a controller is removed from realized set, it may become unnecessary in siblings
         * and ancestors and they should be (de)realized too.
         *
         * This call will defer work on the siblings and derealized ancestors to the next event loop
         * iteration and synchronously creates the parent cgroups (unit_realize_cgroup_now). */

        slice = UNIT_GET_SLICE(u);
        if (slice)
                unit_add_family_to_cgroup_realize_queue(slice);

        /* And realize this one now (and apply the values) */
        return unit_realize_cgroup_now(u, manager_state(u->manager));
}

void unit_release_cgroup(Unit *u) {
        assert(u);

        /* Forgets all cgroup details for this cgroup — but does *not* destroy the cgroup. This is hence OK to call
         * when we close down everything for reexecution, where we really want to leave the cgroup in place. */

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt)
                return;

        if (crt->cgroup_path) {
                (void) hashmap_remove(u->manager->cgroup_unit, crt->cgroup_path);
                crt->cgroup_path = mfree(crt->cgroup_path);
        }

        if (crt->cgroup_control_inotify_wd >= 0) {
                if (inotify_rm_watch(u->manager->cgroup_inotify_fd, crt->cgroup_control_inotify_wd) < 0)
                        log_unit_debug_errno(u, errno, "Failed to remove cgroup control inotify watch %i for %s, ignoring: %m", crt->cgroup_control_inotify_wd, u->id);

                (void) hashmap_remove(u->manager->cgroup_control_inotify_wd_unit, INT_TO_PTR(crt->cgroup_control_inotify_wd));
                crt->cgroup_control_inotify_wd = -1;
        }

        if (crt->cgroup_memory_inotify_wd >= 0) {
                if (inotify_rm_watch(u->manager->cgroup_inotify_fd, crt->cgroup_memory_inotify_wd) < 0)
                        log_unit_debug_errno(u, errno, "Failed to remove cgroup memory inotify watch %i for %s, ignoring: %m", crt->cgroup_memory_inotify_wd, u->id);

                (void) hashmap_remove(u->manager->cgroup_memory_inotify_wd_unit, INT_TO_PTR(crt->cgroup_memory_inotify_wd));
                crt->cgroup_memory_inotify_wd = -1;
        }

        *(CGroupRuntime**) ((uint8_t*) u + UNIT_VTABLE(u)->cgroup_runtime_offset) = cgroup_runtime_free(crt);
}

int unit_cgroup_is_empty(Unit *u) {
        int r;

        assert(u);

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt)
                return -ENXIO;
        if (!crt->cgroup_path)
                return -EOWNERDEAD;

        r = cg_is_empty_recursive(SYSTEMD_CGROUP_CONTROLLER, crt->cgroup_path);
        if (r < 0)
                return log_unit_debug_errno(u, r, "Failed to determine whether cgroup %s is empty, ignoring: %m", empty_to_root(crt->cgroup_path));

        return r;
}

bool unit_maybe_release_cgroup(Unit *u) {
        int r;

        assert(u);

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return true;

        /* Don't release the cgroup if there are still processes under it. If we get notified later when all
         * the processes exit (e.g. the processes were in D-state and exited after the unit was marked as
         * failed) we need the cgroup paths to continue to be tracked by the manager so they can be looked up
         * and cleaned up later. */
        r = unit_cgroup_is_empty(u);
        if (r == 1) {
                unit_release_cgroup(u);
                return true;
        }

        return false;
}

void unit_prune_cgroup(Unit *u) {
        int r;
        bool is_root_slice;

        assert(u);

        /* Removes the cgroup, if empty and possible, and stops watching it. */
        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return;

        /* Cache the last CPU and memory usage values before we destroy the cgroup */
        (void) unit_get_cpu_usage(u, /* ret = */ NULL);

        for (CGroupMemoryAccountingMetric metric = 0; metric <= _CGROUP_MEMORY_ACCOUNTING_METRIC_CACHED_LAST; metric++)
                (void) unit_get_memory_accounting(u, metric, /* ret = */ NULL);

#if BPF_FRAMEWORK
        (void) bpf_restrict_fs_cleanup(u); /* Remove cgroup from the global LSM BPF map */
#endif

        unit_modify_nft_set(u, /* add = */ false);

        is_root_slice = unit_has_name(u, SPECIAL_ROOT_SLICE);

        r = cg_trim_everywhere(u->manager->cgroup_supported, crt->cgroup_path, !is_root_slice);
        if (r < 0)
                /* One reason we could have failed here is, that the cgroup still contains a process.
                 * However, if the cgroup becomes removable at a later time, it might be removed when
                 * the containing slice is stopped. So even if we failed now, this unit shouldn't assume
                 * that the cgroup is still realized the next time it is started. Do not return early
                 * on error, continue cleanup. */
                log_unit_full_errno(u, r == -EBUSY ? LOG_DEBUG : LOG_WARNING, r, "Failed to destroy cgroup %s, ignoring: %m", empty_to_root(crt->cgroup_path));

        if (is_root_slice)
                return;

        if (!unit_maybe_release_cgroup(u)) /* Returns true if the cgroup was released */
                return;

        crt = unit_get_cgroup_runtime(u); /* The above might have destroyed the runtime object, let's see if it's still there */
        if (!crt)
                return;

        crt->cgroup_realized = false;
        crt->cgroup_realized_mask = 0;
        crt->cgroup_enabled_mask = 0;

        crt->bpf_device_control_installed = bpf_program_free(crt->bpf_device_control_installed);
}

int unit_search_main_pid(Unit *u, PidRef *ret) {
        _cleanup_(pidref_done) PidRef pidref = PIDREF_NULL;
        _cleanup_fclose_ FILE *f = NULL;
        int r;

        assert(u);
        assert(ret);

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return -ENXIO;

        r = cg_enumerate_processes(SYSTEMD_CGROUP_CONTROLLER, crt->cgroup_path, &f);
        if (r < 0)
                return r;

        for (;;) {
                _cleanup_(pidref_done) PidRef npidref = PIDREF_NULL;

                /* cg_read_pidref() will return an error on unmapped PIDs.
                 * We can't reasonably deal with units that contain those. */
                r = cg_read_pidref(f, &npidref, CGROUP_DONT_SKIP_UNMAPPED);
                if (r < 0)
                        return r;
                if (r == 0)
                        break;

                if (pidref_equal(&pidref, &npidref)) /* seen already, cgroupfs reports duplicates! */
                        continue;

                if (pidref_is_my_child(&npidref) <= 0) /* ignore processes further down the tree */
                        continue;

                if (pidref_is_set(&pidref) != 0)
                        /* Dang, there's more than one daemonized PID in this group, so we don't know what
                         * process is the main process. */
                        return -ENODATA;

                pidref = TAKE_PIDREF(npidref);
        }

        if (!pidref_is_set(&pidref))
                return -ENODATA;

        *ret = TAKE_PIDREF(pidref);
        return 0;
}

static int unit_watch_pids_in_path(Unit *u, const char *path) {
        _cleanup_closedir_ DIR *d = NULL;
        _cleanup_fclose_ FILE *f = NULL;
        int ret = 0, r;

        assert(u);
        assert(path);

        r = cg_enumerate_processes(SYSTEMD_CGROUP_CONTROLLER, path, &f);
        if (r < 0)
                RET_GATHER(ret, r);
        else {
                for (;;) {
                        _cleanup_(pidref_done) PidRef pid = PIDREF_NULL;

                        r = cg_read_pidref(f, &pid, /* flags = */ 0);
                        if (r == 0)
                                break;
                        if (r < 0) {
                                RET_GATHER(ret, r);
                                break;
                        }

                        RET_GATHER(ret, unit_watch_pidref(u, &pid, /* exclusive= */ false));
                }
        }

        r = cg_enumerate_subgroups(SYSTEMD_CGROUP_CONTROLLER, path, &d);
        if (r < 0)
                RET_GATHER(ret, r);
        else {
                for (;;) {
                        _cleanup_free_ char *fn = NULL, *p = NULL;

                        r = cg_read_subgroup(d, &fn);
                        if (r == 0)
                                break;
                        if (r < 0) {
                                RET_GATHER(ret, r);
                                break;
                        }

                        p = path_join(empty_to_root(path), fn);
                        if (!p)
                                return -ENOMEM;

                        RET_GATHER(ret, unit_watch_pids_in_path(u, p));
                }
        }

        return ret;
}

int unit_synthesize_cgroup_empty_event(Unit *u) {
        int r;

        assert(u);

        /* Enqueue a synthetic cgroup empty event if this unit doesn't watch any PIDs anymore. This is compatibility
         * support for non-unified systems where notifications aren't reliable, and hence need to take whatever we can
         * get as notification source as soon as we stopped having any useful PIDs to watch for. */

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return -ENOENT;

        r = cg_unified_controller(SYSTEMD_CGROUP_CONTROLLER);
        if (r < 0)
                return r;
        if (r > 0) /* On unified we have reliable notifications, and don't need this */
                return 0;

        if (!set_isempty(u->pids))
                return 0;

        unit_add_to_cgroup_empty_queue(u);
        return 0;
}

int unit_watch_all_pids(Unit *u) {
        int r;

        assert(u);

        /* Adds all PIDs from our cgroup to the set of PIDs we
         * watch. This is a fallback logic for cases where we do not
         * get reliable cgroup empty notifications: we try to use
         * SIGCHLD as replacement. */

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return -ENOENT;

        r = cg_unified_controller(SYSTEMD_CGROUP_CONTROLLER);
        if (r < 0)
                return r;
        if (r > 0) /* On unified we can use proper notifications */
                return 0;

        return unit_watch_pids_in_path(u, crt->cgroup_path);
}

static int on_cgroup_empty_event(sd_event_source *s, void *userdata) {
        Manager *m = ASSERT_PTR(userdata);
        Unit *u;
        int r;

        assert(s);

        u = m->cgroup_empty_queue;
        if (!u)
                return 0;

        assert(u->in_cgroup_empty_queue);
        u->in_cgroup_empty_queue = false;
        LIST_REMOVE(cgroup_empty_queue, m->cgroup_empty_queue, u);

        if (m->cgroup_empty_queue) {
                /* More stuff queued, let's make sure we remain enabled */
                r = sd_event_source_set_enabled(s, SD_EVENT_ONESHOT);
                if (r < 0)
                        log_debug_errno(r, "Failed to reenable cgroup empty event source, ignoring: %m");
        }

        /* Update state based on OOM kills before we notify about cgroup empty event */
        (void) unit_check_oom(u);
        (void) unit_check_oomd_kill(u);

        unit_add_to_gc_queue(u);

        if (IN_SET(unit_active_state(u), UNIT_INACTIVE, UNIT_FAILED))
                unit_prune_cgroup(u);
        else if (UNIT_VTABLE(u)->notify_cgroup_empty)
                UNIT_VTABLE(u)->notify_cgroup_empty(u);

        return 0;
}

void unit_add_to_cgroup_empty_queue(Unit *u) {
        int r;

        assert(u);

        /* Note that there are four different ways how cgroup empty events reach us:
         *
         * 1. On the unified hierarchy we get an inotify event on the cgroup
         *
         * 2. On the legacy hierarchy, when running in system mode, we get a datagram on the cgroup agent socket
         *
         * 3. On the legacy hierarchy, when running in user mode, we get a D-Bus signal on the system bus
         *
         * 4. On the legacy hierarchy, in service units we start watching all processes of the cgroup for SIGCHLD as
         *    soon as we get one SIGCHLD, to deal with unreliable cgroup notifications.
         *
         * Regardless which way we got the notification, we'll verify it here, and then add it to a separate
         * queue. This queue will be dispatched at a lower priority than the SIGCHLD handler, so that we always use
         * SIGCHLD if we can get it first, and only use the cgroup empty notifications if there's no SIGCHLD pending
         * (which might happen if the cgroup doesn't contain processes that are our own child, which is typically the
         * case for scope units). */

        if (u->in_cgroup_empty_queue)
                return;

        /* Let's verify that the cgroup is really empty */
        r = unit_cgroup_is_empty(u);
        if (r <= 0)
                return;

        LIST_PREPEND(cgroup_empty_queue, u->manager->cgroup_empty_queue, u);
        u->in_cgroup_empty_queue = true;

        /* Trigger the defer event */
        r = sd_event_source_set_enabled(u->manager->cgroup_empty_event_source, SD_EVENT_ONESHOT);
        if (r < 0)
                log_debug_errno(r, "Failed to enable cgroup empty event source: %m");
}

static void unit_remove_from_cgroup_empty_queue(Unit *u) {
        assert(u);

        if (!u->in_cgroup_empty_queue)
                return;

        LIST_REMOVE(cgroup_empty_queue, u->manager->cgroup_empty_queue, u);
        u->in_cgroup_empty_queue = false;
}

int unit_check_oomd_kill(Unit *u) {
        _cleanup_free_ char *value = NULL;
        bool increased;
        uint64_t n = 0;
        int r;

        assert(u);

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return 0;

        r = cg_all_unified();
        if (r < 0)
                return log_unit_debug_errno(u, r, "Couldn't determine whether we are in all unified mode: %m");
        else if (r == 0)
                return 0;

        r = cg_get_xattr_malloc(crt->cgroup_path, "user.oomd_ooms", &value);
        if (r < 0 && !ERRNO_IS_XATTR_ABSENT(r))
                return r;

        if (!isempty(value)) {
                 r = safe_atou64(value, &n);
                 if (r < 0)
                         return r;
        }

        increased = n > crt->managed_oom_kill_last;
        crt->managed_oom_kill_last = n;

        if (!increased)
                return 0;

        n = 0;
        value = mfree(value);
        r = cg_get_xattr_malloc(crt->cgroup_path, "user.oomd_kill", &value);
        if (r >= 0 && !isempty(value))
                (void) safe_atou64(value, &n);

        if (n > 0)
                log_unit_struct(u, LOG_NOTICE,
                                "MESSAGE_ID=" SD_MESSAGE_UNIT_OOMD_KILL_STR,
                                LOG_UNIT_INVOCATION_ID(u),
                                LOG_UNIT_MESSAGE(u, "systemd-oomd killed %"PRIu64" process(es) in this unit.", n),
                                "N_PROCESSES=%" PRIu64, n);
        else
                log_unit_struct(u, LOG_NOTICE,
                                "MESSAGE_ID=" SD_MESSAGE_UNIT_OOMD_KILL_STR,
                                LOG_UNIT_INVOCATION_ID(u),
                                LOG_UNIT_MESSAGE(u, "systemd-oomd killed some process(es) in this unit."));

        unit_notify_cgroup_oom(u, /* ManagedOOM= */ true);

        return 1;
}

int unit_check_oom(Unit *u) {
        _cleanup_free_ char *oom_kill = NULL;
        bool increased;
        uint64_t c;
        int r;

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return 0;

        r = cg_get_keyed_attribute(
                        "memory",
                        crt->cgroup_path,
                        "memory.events",
                        STRV_MAKE("oom_kill"),
                        &oom_kill);
        if (IN_SET(r, -ENOENT, -ENXIO)) /* Handle gracefully if cgroup or oom_kill attribute don't exist */
                c = 0;
        else if (r < 0)
                return log_unit_debug_errno(u, r, "Failed to read oom_kill field of memory.events cgroup attribute: %m");
        else {
                r = safe_atou64(oom_kill, &c);
                if (r < 0)
                        return log_unit_debug_errno(u, r, "Failed to parse oom_kill field: %m");
        }

        increased = c > crt->oom_kill_last;
        crt->oom_kill_last = c;

        if (!increased)
                return 0;

        log_unit_struct(u, LOG_NOTICE,
                        "MESSAGE_ID=" SD_MESSAGE_UNIT_OUT_OF_MEMORY_STR,
                        LOG_UNIT_INVOCATION_ID(u),
                        LOG_UNIT_MESSAGE(u, "A process of this unit has been killed by the OOM killer."));

        unit_notify_cgroup_oom(u, /* ManagedOOM= */ false);

        return 1;
}

static int on_cgroup_oom_event(sd_event_source *s, void *userdata) {
        Manager *m = ASSERT_PTR(userdata);
        Unit *u;
        int r;

        assert(s);

        u = m->cgroup_oom_queue;
        if (!u)
                return 0;

        assert(u->in_cgroup_oom_queue);
        u->in_cgroup_oom_queue = false;
        LIST_REMOVE(cgroup_oom_queue, m->cgroup_oom_queue, u);

        if (m->cgroup_oom_queue) {
                /* More stuff queued, let's make sure we remain enabled */
                r = sd_event_source_set_enabled(s, SD_EVENT_ONESHOT);
                if (r < 0)
                        log_debug_errno(r, "Failed to reenable cgroup oom event source, ignoring: %m");
        }

        (void) unit_check_oom(u);
        unit_add_to_gc_queue(u);

        return 0;
}

static void unit_add_to_cgroup_oom_queue(Unit *u) {
        int r;

        assert(u);

        if (u->in_cgroup_oom_queue)
                return;

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return;

        LIST_PREPEND(cgroup_oom_queue, u->manager->cgroup_oom_queue, u);
        u->in_cgroup_oom_queue = true;

        /* Trigger the defer event */
        if (!u->manager->cgroup_oom_event_source) {
                _cleanup_(sd_event_source_unrefp) sd_event_source *s = NULL;

                r = sd_event_add_defer(u->manager->event, &s, on_cgroup_oom_event, u->manager);
                if (r < 0) {
                        log_error_errno(r, "Failed to create cgroup oom event source: %m");
                        return;
                }

                r = sd_event_source_set_priority(s, EVENT_PRIORITY_CGROUP_OOM);
                if (r < 0) {
                        log_error_errno(r, "Failed to set priority of cgroup oom event source: %m");
                        return;
                }

                (void) sd_event_source_set_description(s, "cgroup-oom");
                u->manager->cgroup_oom_event_source = TAKE_PTR(s);
        }

        r = sd_event_source_set_enabled(u->manager->cgroup_oom_event_source, SD_EVENT_ONESHOT);
        if (r < 0)
                log_error_errno(r, "Failed to enable cgroup oom event source: %m");
}

static int unit_check_cgroup_events(Unit *u) {
        char *values[2] = {};
        int r;

        assert(u);

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return 0;

        r = cg_get_keyed_attribute_graceful(
                        SYSTEMD_CGROUP_CONTROLLER,
                        crt->cgroup_path,
                        "cgroup.events",
                        STRV_MAKE("populated", "frozen"),
                        values);
        if (r < 0)
                return r;

        /* The cgroup.events notifications can be merged together so act as we saw the given state for the
         * first time. The functions we call to handle given state are idempotent, which makes them
         * effectively remember the previous state. */
        if (values[0]) {
                if (streq(values[0], "1"))
                        unit_remove_from_cgroup_empty_queue(u);
                else
                        unit_add_to_cgroup_empty_queue(u);
        }

        /* Disregard freezer state changes due to operations not initiated by us.
         * See: https://github.com/systemd/systemd/pull/13512/files#r416469963 and
         *      https://github.com/systemd/systemd/pull/13512#issuecomment-573007207 */
        if (values[1] && IN_SET(u->freezer_state, FREEZER_FREEZING, FREEZER_FREEZING_BY_PARENT, FREEZER_THAWING)) {
                if (streq(values[1], "0"))
                        unit_thawed(u);
                else
                        unit_frozen(u);
        }

        free(values[0]);
        free(values[1]);

        return 0;
}

static int on_cgroup_inotify_event(sd_event_source *s, int fd, uint32_t revents, void *userdata) {
        Manager *m = ASSERT_PTR(userdata);

        assert(s);
        assert(fd >= 0);

        for (;;) {
                union inotify_event_buffer buffer;
                ssize_t l;

                l = read(fd, &buffer, sizeof(buffer));
                if (l < 0) {
                        if (ERRNO_IS_TRANSIENT(errno))
                                return 0;

                        return log_error_errno(errno, "Failed to read control group inotify events: %m");
                }

                FOREACH_INOTIFY_EVENT_WARN(e, buffer, l) {
                        Unit *u;

                        if (e->wd < 0)
                                /* Queue overflow has no watch descriptor */
                                continue;

                        if (e->mask & IN_IGNORED)
                                /* The watch was just removed */
                                continue;

                        /* Note that inotify might deliver events for a watch even after it was removed,
                         * because it was queued before the removal. Let's ignore this here safely. */

                        u = hashmap_get(m->cgroup_control_inotify_wd_unit, INT_TO_PTR(e->wd));
                        if (u)
                                unit_check_cgroup_events(u);

                        u = hashmap_get(m->cgroup_memory_inotify_wd_unit, INT_TO_PTR(e->wd));
                        if (u)
                                unit_add_to_cgroup_oom_queue(u);
                }
        }
}

static int cg_bpf_mask_supported(CGroupMask *ret) {
        CGroupMask mask = 0;
        int r;

        /* BPF-based firewall */
        r = bpf_firewall_supported();
        if (r < 0)
                return r;
        if (r > 0)
                mask |= CGROUP_MASK_BPF_FIREWALL;

        /* BPF-based device access control */
        r = bpf_devices_supported();
        if (r < 0)
                return r;
        if (r > 0)
                mask |= CGROUP_MASK_BPF_DEVICES;

        /* BPF pinned prog */
        r = bpf_foreign_supported();
        if (r < 0)
                return r;
        if (r > 0)
                mask |= CGROUP_MASK_BPF_FOREIGN;

        /* BPF-based bind{4|6} hooks */
        r = bpf_socket_bind_supported();
        if (r < 0)
                return r;
        if (r > 0)
                mask |= CGROUP_MASK_BPF_SOCKET_BIND;

        /* BPF-based cgroup_skb/{egress|ingress} hooks */
        r = bpf_restrict_ifaces_supported();
        if (r < 0)
                return r;
        if (r > 0)
                mask |= CGROUP_MASK_BPF_RESTRICT_NETWORK_INTERFACES;

        *ret = mask;
        return 0;
}

int manager_setup_cgroup(Manager *m) {
        _cleanup_free_ char *path = NULL;
        const char *scope_path;
        int r, all_unified;
        CGroupMask mask;
        char *e;

        assert(m);

        /* 1. Determine hierarchy */
        m->cgroup_root = mfree(m->cgroup_root);
        r = cg_pid_get_path(SYSTEMD_CGROUP_CONTROLLER, 0, &m->cgroup_root);
        if (r < 0)
                return log_error_errno(r, "Cannot determine cgroup we are running in: %m");

        /* Chop off the init scope, if we are already located in it */
        e = endswith(m->cgroup_root, "/" SPECIAL_INIT_SCOPE);

        /* LEGACY: Also chop off the system slice if we are in
         * it. This is to support live upgrades from older systemd
         * versions where PID 1 was moved there. Also see
         * cg_get_root_path(). */
        if (!e && MANAGER_IS_SYSTEM(m)) {
                e = endswith(m->cgroup_root, "/" SPECIAL_SYSTEM_SLICE);
                if (!e)
                        e = endswith(m->cgroup_root, "/system"); /* even more legacy */
        }
        if (e)
                *e = 0;

        /* And make sure to store away the root value without trailing slash, even for the root dir, so that we can
         * easily prepend it everywhere. */
        delete_trailing_chars(m->cgroup_root, "/");

        /* 2. Show data */
        r = cg_get_path(SYSTEMD_CGROUP_CONTROLLER, m->cgroup_root, NULL, &path);
        if (r < 0)
                return log_error_errno(r, "Cannot find cgroup mount point: %m");

        r = cg_unified();
        if (r < 0)
                return log_error_errno(r, "Couldn't determine if we are running in the unified hierarchy: %m");

        all_unified = cg_all_unified();
        if (all_unified < 0)
                return log_error_errno(all_unified, "Couldn't determine whether we are in all unified mode: %m");
        if (all_unified > 0)
                log_debug("Unified cgroup hierarchy is located at %s.", path);
        else {
                r = cg_unified_controller(SYSTEMD_CGROUP_CONTROLLER);
                if (r < 0)
                        return log_error_errno(r, "Failed to determine whether systemd's own controller is in unified mode: %m");
                if (r > 0)
                        log_debug("Unified cgroup hierarchy is located at %s. Controllers are on legacy hierarchies.", path);
                else
                        log_debug("Using cgroup controller " SYSTEMD_CGROUP_CONTROLLER_LEGACY ". File system hierarchy is at %s.", path);
        }

        /* 3. Allocate cgroup empty defer event source */
        m->cgroup_empty_event_source = sd_event_source_disable_unref(m->cgroup_empty_event_source);
        r = sd_event_add_defer(m->event, &m->cgroup_empty_event_source, on_cgroup_empty_event, m);
        if (r < 0)
                return log_error_errno(r, "Failed to create cgroup empty event source: %m");

        /* Schedule cgroup empty checks early, but after having processed service notification messages or
         * SIGCHLD signals, so that a cgroup running empty is always just the last safety net of
         * notification, and we collected the metadata the notification and SIGCHLD stuff offers first. */
        r = sd_event_source_set_priority(m->cgroup_empty_event_source, EVENT_PRIORITY_CGROUP_EMPTY);
        if (r < 0)
                return log_error_errno(r, "Failed to set priority of cgroup empty event source: %m");

        r = sd_event_source_set_enabled(m->cgroup_empty_event_source, SD_EVENT_OFF);
        if (r < 0)
                return log_error_errno(r, "Failed to disable cgroup empty event source: %m");

        (void) sd_event_source_set_description(m->cgroup_empty_event_source, "cgroup-empty");

        /* 4. Install notifier inotify object, or agent */
        if (cg_unified_controller(SYSTEMD_CGROUP_CONTROLLER) > 0) {

                /* In the unified hierarchy we can get cgroup empty notifications via inotify. */

                m->cgroup_inotify_event_source = sd_event_source_disable_unref(m->cgroup_inotify_event_source);
                safe_close(m->cgroup_inotify_fd);

                m->cgroup_inotify_fd = inotify_init1(IN_NONBLOCK|IN_CLOEXEC);
                if (m->cgroup_inotify_fd < 0)
                        return log_error_errno(errno, "Failed to create control group inotify object: %m");

                r = sd_event_add_io(m->event, &m->cgroup_inotify_event_source, m->cgroup_inotify_fd, EPOLLIN, on_cgroup_inotify_event, m);
                if (r < 0)
                        return log_error_errno(r, "Failed to watch control group inotify object: %m");

                /* Process cgroup empty notifications early. Note that when this event is dispatched it'll
                 * just add the unit to a cgroup empty queue, hence let's run earlier than that. Also see
                 * handling of cgroup agent notifications, for the classic cgroup hierarchy support. */
                r = sd_event_source_set_priority(m->cgroup_inotify_event_source, EVENT_PRIORITY_CGROUP_INOTIFY);
                if (r < 0)
                        return log_error_errno(r, "Failed to set priority of inotify event source: %m");

                (void) sd_event_source_set_description(m->cgroup_inotify_event_source, "cgroup-inotify");

        } else if (MANAGER_IS_SYSTEM(m) && manager_owns_host_root_cgroup(m) && !MANAGER_IS_TEST_RUN(m)) {

                /* On the legacy hierarchy we only get notifications via cgroup agents. (Which isn't really reliable,
                 * since it does not generate events when control groups with children run empty. */

                r = cg_install_release_agent(SYSTEMD_CGROUP_CONTROLLER, SYSTEMD_CGROUPS_AGENT_PATH);
                if (r < 0)
                        log_warning_errno(r, "Failed to install release agent, ignoring: %m");
                else if (r > 0)
                        log_debug("Installed release agent.");
                else if (r == 0)
                        log_debug("Release agent already installed.");
        }

        /* 5. Make sure we are in the special "init.scope" unit in the root slice. */
        scope_path = strjoina(m->cgroup_root, "/" SPECIAL_INIT_SCOPE);
        r = cg_create_and_attach(SYSTEMD_CGROUP_CONTROLLER, scope_path, 0);
        if (r >= 0) {
                /* Also, move all other userspace processes remaining in the root cgroup into that scope. */
                r = cg_migrate(SYSTEMD_CGROUP_CONTROLLER, m->cgroup_root, SYSTEMD_CGROUP_CONTROLLER, scope_path, 0);
                if (r < 0)
                        log_warning_errno(r, "Couldn't move remaining userspace processes, ignoring: %m");

                /* 6. And pin it, so that it cannot be unmounted */
                safe_close(m->pin_cgroupfs_fd);
                m->pin_cgroupfs_fd = open(path, O_RDONLY|O_CLOEXEC|O_DIRECTORY|O_NOCTTY|O_NONBLOCK);
                if (m->pin_cgroupfs_fd < 0)
                        return log_error_errno(errno, "Failed to open pin file: %m");

        } else if (!MANAGER_IS_TEST_RUN(m))
                return log_error_errno(r, "Failed to create %s control group: %m", scope_path);

        /* 7. Always enable hierarchical support if it exists... */
        if (!all_unified && !MANAGER_IS_TEST_RUN(m))
                (void) cg_set_attribute("memory", "/", "memory.use_hierarchy", "1");

        /* 8. Figure out which controllers are supported */
        r = cg_mask_supported_subtree(m->cgroup_root, &m->cgroup_supported);
        if (r < 0)
                return log_error_errno(r, "Failed to determine supported controllers: %m");

        /* 9. Figure out which bpf-based pseudo-controllers are supported */
        r = cg_bpf_mask_supported(&mask);
        if (r < 0)
                return log_error_errno(r, "Failed to determine supported bpf-based pseudo-controllers: %m");
        m->cgroup_supported |= mask;

        /* 10. Log which controllers are supported */
        for (CGroupController c = 0; c < _CGROUP_CONTROLLER_MAX; c++)
                log_debug("Controller '%s' supported: %s", cgroup_controller_to_string(c),
                          yes_no(m->cgroup_supported & CGROUP_CONTROLLER_TO_MASK(c)));

        return 0;
}

void manager_shutdown_cgroup(Manager *m, bool delete) {
        assert(m);

        /* We can't really delete the group, since we are in it. But
         * let's trim it. */
        if (delete && m->cgroup_root && !FLAGS_SET(m->test_run_flags, MANAGER_TEST_RUN_MINIMAL))
                (void) cg_trim(SYSTEMD_CGROUP_CONTROLLER, m->cgroup_root, false);

        m->cgroup_empty_event_source = sd_event_source_disable_unref(m->cgroup_empty_event_source);

        m->cgroup_control_inotify_wd_unit = hashmap_free(m->cgroup_control_inotify_wd_unit);
        m->cgroup_memory_inotify_wd_unit = hashmap_free(m->cgroup_memory_inotify_wd_unit);

        m->cgroup_inotify_event_source = sd_event_source_disable_unref(m->cgroup_inotify_event_source);
        m->cgroup_inotify_fd = safe_close(m->cgroup_inotify_fd);

        m->pin_cgroupfs_fd = safe_close(m->pin_cgroupfs_fd);

        m->cgroup_root = mfree(m->cgroup_root);
}

Unit* manager_get_unit_by_cgroup(Manager *m, const char *cgroup) {
        char *p;
        Unit *u;

        assert(m);
        assert(cgroup);

        u = hashmap_get(m->cgroup_unit, cgroup);
        if (u)
                return u;

        p = strdupa_safe(cgroup);
        for (;;) {
                char *e;

                e = strrchr(p, '/');
                if (!e || e == p)
                        return hashmap_get(m->cgroup_unit, SPECIAL_ROOT_SLICE);

                *e = 0;

                u = hashmap_get(m->cgroup_unit, p);
                if (u)
                        return u;
        }
}

Unit *manager_get_unit_by_pidref_cgroup(Manager *m, const PidRef *pid) {
        _cleanup_free_ char *cgroup = NULL;

        assert(m);

        if (cg_pidref_get_path(SYSTEMD_CGROUP_CONTROLLER, pid, &cgroup) < 0)
                return NULL;

        return manager_get_unit_by_cgroup(m, cgroup);
}

Unit *manager_get_unit_by_pidref_watching(Manager *m, const PidRef *pid) {
        Unit *u, **array;

        assert(m);

        if (!pidref_is_set(pid))
                return NULL;

        u = hashmap_get(m->watch_pids, pid);
        if (u)
                return u;

        array = hashmap_get(m->watch_pids_more, pid);
        if (array)
                return array[0];

        return NULL;
}

Unit *manager_get_unit_by_pidref(Manager *m, const PidRef *pid) {
        Unit *u;

        assert(m);

        /* Note that a process might be owned by multiple units, we return only one here, which is good
         * enough for most cases, though not strictly correct. We prefer the one reported by cgroup
         * membership, as that's the most relevant one as children of the process will be assigned to that
         * one, too, before all else. */

        if (!pidref_is_set(pid))
                return NULL;

        if (pidref_is_self(pid))
                return hashmap_get(m->units, SPECIAL_INIT_SCOPE);
        if (pid->pid == 1)
                return NULL;

        u = manager_get_unit_by_pidref_cgroup(m, pid);
        if (u)
                return u;

        u = manager_get_unit_by_pidref_watching(m, pid);
        if (u)
                return u;

        return NULL;
}

Unit *manager_get_unit_by_pid(Manager *m, pid_t pid) {
        assert(m);

        if (!pid_is_valid(pid))
                return NULL;

        return manager_get_unit_by_pidref(m, &PIDREF_MAKE_FROM_PID(pid));
}

int manager_notify_cgroup_empty(Manager *m, const char *cgroup) {
        Unit *u;

        assert(m);
        assert(cgroup);

        /* Called on the legacy hierarchy whenever we get an explicit cgroup notification from the cgroup agent process
         * or from the --system instance */

        log_debug("Got cgroup empty notification for: %s", cgroup);

        u = manager_get_unit_by_cgroup(m, cgroup);
        if (!u)
                return 0;

        unit_add_to_cgroup_empty_queue(u);
        return 1;
}

int unit_get_memory_available(Unit *u, uint64_t *ret) {
        uint64_t available = UINT64_MAX, current = 0;

        assert(u);
        assert(ret);

        /* If data from cgroups can be accessed, try to find out how much more memory a unit can
         * claim before hitting the configured cgroup limits (if any). Consider both MemoryHigh
         * and MemoryMax, and also any slice the unit might be nested below. */

        do {
                uint64_t unit_available, unit_limit = UINT64_MAX;
                CGroupContext *unit_context;

                /* No point in continuing if we can't go any lower */
                if (available == 0)
                        break;

                unit_context = unit_get_cgroup_context(u);
                if (!unit_context)
                        return -ENODATA;

                CGroupRuntime *crt = unit_get_cgroup_runtime(u);
                if (!crt || !crt->cgroup_path)
                        continue;

                (void) unit_get_memory_current(u, &current);
                /* in case of error, previous current propagates as lower bound */

                if (unit_has_name(u, SPECIAL_ROOT_SLICE))
                        unit_limit = physical_memory();
                else if (unit_context->memory_max == UINT64_MAX && unit_context->memory_high == UINT64_MAX)
                        continue;
                unit_limit = MIN3(unit_limit, unit_context->memory_max, unit_context->memory_high);

                unit_available = LESS_BY(unit_limit, current);
                available = MIN(unit_available, available);
        } while ((u = UNIT_GET_SLICE(u)));

        *ret = available;

        return 0;
}

int unit_get_memory_current(Unit *u, uint64_t *ret) {
        int r;

        // FIXME: Merge this into unit_get_memory_accounting after support for cgroup v1 is dropped

        assert(u);
        assert(ret);

        if (!UNIT_CGROUP_BOOL(u, memory_accounting))
                return -ENODATA;

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return -ENODATA;

        /* The root cgroup doesn't expose this information, let's get it from /proc instead */
        if (unit_has_host_root_cgroup(u))
                return procfs_memory_get_used(ret);

        if ((crt->cgroup_realized_mask & CGROUP_MASK_MEMORY) == 0)
                return -ENODATA;

        r = cg_all_unified();
        if (r < 0)
                return r;

        return cg_get_attribute_as_uint64("memory", crt->cgroup_path, r > 0 ? "memory.current" : "memory.usage_in_bytes", ret);
}

int unit_get_memory_accounting(Unit *u, CGroupMemoryAccountingMetric metric, uint64_t *ret) {

        static const char* const attributes_table[_CGROUP_MEMORY_ACCOUNTING_METRIC_MAX] = {
                [CGROUP_MEMORY_PEAK]          = "memory.peak",
                [CGROUP_MEMORY_SWAP_CURRENT]  = "memory.swap.current",
                [CGROUP_MEMORY_SWAP_PEAK]     = "memory.swap.peak",
                [CGROUP_MEMORY_ZSWAP_CURRENT] = "memory.zswap.current",
        };

        uint64_t bytes;
        bool updated = false;
        int r;

        assert(u);
        assert(metric >= 0);
        assert(metric < _CGROUP_MEMORY_ACCOUNTING_METRIC_MAX);

        if (!UNIT_CGROUP_BOOL(u, memory_accounting))
                return -ENODATA;

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt)
                return -ENODATA;
        if (!crt->cgroup_path)
                /* If the cgroup is already gone, we try to find the last cached value. */
                goto finish;

        /* The root cgroup doesn't expose this information. */
        if (unit_has_host_root_cgroup(u))
                return -ENODATA;

        if (!FLAGS_SET(crt->cgroup_realized_mask, CGROUP_MASK_MEMORY))
                return -ENODATA;

        r = cg_all_unified();
        if (r < 0)
                return r;
        if (r == 0)
                return -ENODATA;

        r = cg_get_attribute_as_uint64("memory", crt->cgroup_path, attributes_table[metric], &bytes);
        if (r < 0 && r != -ENODATA)
                return r;
        updated = r >= 0;

finish:
        if (metric <= _CGROUP_MEMORY_ACCOUNTING_METRIC_CACHED_LAST) {
                uint64_t *last = &crt->memory_accounting_last[metric];

                if (updated)
                        *last = bytes;
                else if (*last != UINT64_MAX)
                        bytes = *last;
                else
                        return -ENODATA;

        } else if (!updated)
                return -ENODATA;

        if (ret)
                *ret = bytes;

        return 0;
}

int unit_get_tasks_current(Unit *u, uint64_t *ret) {
        assert(u);
        assert(ret);

        if (!UNIT_CGROUP_BOOL(u, tasks_accounting))
                return -ENODATA;

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return -ENODATA;

        /* The root cgroup doesn't expose this information, let's get it from /proc instead */
        if (unit_has_host_root_cgroup(u))
                return procfs_tasks_get_current(ret);

        if ((crt->cgroup_realized_mask & CGROUP_MASK_PIDS) == 0)
                return -ENODATA;

        return cg_get_attribute_as_uint64("pids", crt->cgroup_path, "pids.current", ret);
}

static int unit_get_cpu_usage_raw(Unit *u, nsec_t *ret) {
        uint64_t ns;
        int r;

        assert(u);
        assert(ret);

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return -ENODATA;

        /* The root cgroup doesn't expose this information, let's get it from /proc instead */
        if (unit_has_host_root_cgroup(u))
                return procfs_cpu_get_usage(ret);

        /* Requisite controllers for CPU accounting are not enabled */
        if ((get_cpu_accounting_mask() & ~crt->cgroup_realized_mask) != 0)
                return -ENODATA;

        r = cg_all_unified();
        if (r < 0)
                return r;
        if (r > 0) {
                _cleanup_free_ char *val = NULL;
                uint64_t us;

                r = cg_get_keyed_attribute("cpu", crt->cgroup_path, "cpu.stat", STRV_MAKE("usage_usec"), &val);
                if (IN_SET(r, -ENOENT, -ENXIO))
                        return -ENODATA;
                if (r < 0)
                        return r;

                r = safe_atou64(val, &us);
                if (r < 0)
                        return r;

                ns = us * NSEC_PER_USEC;
        } else
                return cg_get_attribute_as_uint64("cpuacct", crt->cgroup_path, "cpuacct.usage", ret);

        *ret = ns;
        return 0;
}

int unit_get_cpu_usage(Unit *u, nsec_t *ret) {
        nsec_t ns;
        int r;

        assert(u);

        /* Retrieve the current CPU usage counter. This will subtract the CPU counter taken when the unit was
         * started. If the cgroup has been removed already, returns the last cached value. To cache the value, simply
         * call this function with a NULL return value. */

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return -ENODATA;

        if (!UNIT_CGROUP_BOOL(u, cpu_accounting))
                return -ENODATA;

        r = unit_get_cpu_usage_raw(u, &ns);
        if (r == -ENODATA && crt->cpu_usage_last != NSEC_INFINITY) {
                /* If we can't get the CPU usage anymore (because the cgroup was already removed, for example), use our
                 * cached value. */

                if (ret)
                        *ret = crt->cpu_usage_last;
                return 0;
        }
        if (r < 0)
                return r;

        if (ns > crt->cpu_usage_base)
                ns -= crt->cpu_usage_base;
        else
                ns = 0;

        crt->cpu_usage_last = ns;
        if (ret)
                *ret = ns;

        return 0;
}

int unit_get_ip_accounting(
                Unit *u,
                CGroupIPAccountingMetric metric,
                uint64_t *ret) {

        uint64_t value;
        int fd, r;

        assert(u);
        assert(metric >= 0);
        assert(metric < _CGROUP_IP_ACCOUNTING_METRIC_MAX);
        assert(ret);

        if (!UNIT_CGROUP_BOOL(u, ip_accounting))
                return -ENODATA;

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return -ENODATA;

        fd = IN_SET(metric, CGROUP_IP_INGRESS_BYTES, CGROUP_IP_INGRESS_PACKETS) ?
                crt->ip_accounting_ingress_map_fd :
                crt->ip_accounting_egress_map_fd;
        if (fd < 0)
                return -ENODATA;

        if (IN_SET(metric, CGROUP_IP_INGRESS_BYTES, CGROUP_IP_EGRESS_BYTES))
                r = bpf_firewall_read_accounting(fd, &value, NULL);
        else
                r = bpf_firewall_read_accounting(fd, NULL, &value);
        if (r < 0)
                return r;

        /* Add in additional metrics from a previous runtime. Note that when reexecing/reloading the daemon we compile
         * all BPF programs and maps anew, but serialize the old counters. When deserializing we store them in the
         * ip_accounting_extra[] field, and add them in here transparently. */

        *ret = value + crt->ip_accounting_extra[metric];

        return r;
}

static uint64_t unit_get_effective_limit_one(Unit *u, CGroupLimitType type) {
        CGroupContext *cc;

        assert(u);
        assert(UNIT_HAS_CGROUP_CONTEXT(u));

        if (unit_has_name(u, SPECIAL_ROOT_SLICE))
                switch (type) {
                        case CGROUP_LIMIT_MEMORY_MAX:
                        case CGROUP_LIMIT_MEMORY_HIGH:
                                return physical_memory();
                        case CGROUP_LIMIT_TASKS_MAX:
                                return system_tasks_max();
                        default:
                                assert_not_reached();
                }

        cc = ASSERT_PTR(unit_get_cgroup_context(u));
        switch (type) {
                /* Note: on legacy/hybrid hierarchies memory_max stays CGROUP_LIMIT_MAX unless configured
                 * explicitly. Effective value of MemoryLimit= (cgroup v1) is not implemented. */
                case CGROUP_LIMIT_MEMORY_MAX:
                        return cc->memory_max;
                case CGROUP_LIMIT_MEMORY_HIGH:
                        return cc->memory_high;
                case CGROUP_LIMIT_TASKS_MAX:
                        return cgroup_tasks_max_resolve(&cc->tasks_max);
                default:
                        assert_not_reached();
        }
}

int unit_get_effective_limit(Unit *u, CGroupLimitType type, uint64_t *ret) {
        uint64_t infimum;

        assert(u);
        assert(ret);
        assert(type >= 0);
        assert(type < _CGROUP_LIMIT_TYPE_MAX);

        if (!UNIT_HAS_CGROUP_CONTEXT(u))
                return -EINVAL;

        infimum = unit_get_effective_limit_one(u, type);
        for (Unit *slice = UNIT_GET_SLICE(u); slice; slice = UNIT_GET_SLICE(slice))
                infimum = MIN(infimum, unit_get_effective_limit_one(slice, type));

        *ret = infimum;
        return 0;
}

static int unit_get_io_accounting_raw(Unit *u, uint64_t ret[static _CGROUP_IO_ACCOUNTING_METRIC_MAX]) {
        static const char *const field_names[_CGROUP_IO_ACCOUNTING_METRIC_MAX] = {
                [CGROUP_IO_READ_BYTES]       = "rbytes=",
                [CGROUP_IO_WRITE_BYTES]      = "wbytes=",
                [CGROUP_IO_READ_OPERATIONS]  = "rios=",
                [CGROUP_IO_WRITE_OPERATIONS] = "wios=",
        };
        uint64_t acc[_CGROUP_IO_ACCOUNTING_METRIC_MAX] = {};
        _cleanup_free_ char *path = NULL;
        _cleanup_fclose_ FILE *f = NULL;
        int r;

        assert(u);

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return -ENODATA;

        if (unit_has_host_root_cgroup(u))
                return -ENODATA; /* TODO: return useful data for the top-level cgroup */

        r = cg_all_unified();
        if (r < 0)
                return r;
        if (r == 0)
                return -ENODATA;

        if (!FLAGS_SET(crt->cgroup_realized_mask, CGROUP_MASK_IO))
                return -ENODATA;

        r = cg_get_path("io", crt->cgroup_path, "io.stat", &path);
        if (r < 0)
                return r;

        f = fopen(path, "re");
        if (!f)
                return -errno;

        for (;;) {
                _cleanup_free_ char *line = NULL;
                const char *p;

                r = read_line(f, LONG_LINE_MAX, &line);
                if (r < 0)
                        return r;
                if (r == 0)
                        break;

                p = line;
                p += strcspn(p, WHITESPACE); /* Skip over device major/minor */
                p += strspn(p, WHITESPACE);  /* Skip over following whitespace */

                for (;;) {
                        _cleanup_free_ char *word = NULL;

                        r = extract_first_word(&p, &word, NULL, EXTRACT_RETAIN_ESCAPE);
                        if (r < 0)
                                return r;
                        if (r == 0)
                                break;

                        for (CGroupIOAccountingMetric i = 0; i < _CGROUP_IO_ACCOUNTING_METRIC_MAX; i++) {
                                const char *x;

                                x = startswith(word, field_names[i]);
                                if (x) {
                                        uint64_t w;

                                        r = safe_atou64(x, &w);
                                        if (r < 0)
                                                return r;

                                        /* Sum up the stats of all devices */
                                        acc[i] += w;
                                        break;
                                }
                        }
                }
        }

        memcpy(ret, acc, sizeof(acc));
        return 0;
}

int unit_get_io_accounting(
                Unit *u,
                CGroupIOAccountingMetric metric,
                bool allow_cache,
                uint64_t *ret) {

        uint64_t raw[_CGROUP_IO_ACCOUNTING_METRIC_MAX];
        int r;

        /* Retrieve an IO account parameter. This will subtract the counter when the unit was started. */

        if (!UNIT_CGROUP_BOOL(u, io_accounting))
                return -ENODATA;

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return -ENODATA;

        if (allow_cache && crt->io_accounting_last[metric] != UINT64_MAX)
                goto done;

        r = unit_get_io_accounting_raw(u, raw);
        if (r == -ENODATA && crt->io_accounting_last[metric] != UINT64_MAX)
                goto done;
        if (r < 0)
                return r;

        for (CGroupIOAccountingMetric i = 0; i < _CGROUP_IO_ACCOUNTING_METRIC_MAX; i++) {
                /* Saturated subtraction */
                if (raw[i] > crt->io_accounting_base[i])
                        crt->io_accounting_last[i] = raw[i] - crt->io_accounting_base[i];
                else
                        crt->io_accounting_last[i] = 0;
        }

done:
        if (ret)
                *ret = crt->io_accounting_last[metric];

        return 0;
}

int unit_reset_cpu_accounting(Unit *u) {
        int r;

        assert(u);

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return 0;

        crt->cpu_usage_last = NSEC_INFINITY;

        r = unit_get_cpu_usage_raw(u, &crt->cpu_usage_base);
        if (r < 0) {
                crt->cpu_usage_base = 0;
                return r;
        }

        return 0;
}

void unit_reset_memory_accounting_last(Unit *u) {
        assert(u);

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return;

        FOREACH_ELEMENT(i, crt->memory_accounting_last)
                *i = UINT64_MAX;
}

int unit_reset_ip_accounting(Unit *u) {
        int r = 0;

        assert(u);

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return 0;

        if (crt->ip_accounting_ingress_map_fd >= 0)
                RET_GATHER(r, bpf_firewall_reset_accounting(crt->ip_accounting_ingress_map_fd));

        if (crt->ip_accounting_egress_map_fd >= 0)
                RET_GATHER(r, bpf_firewall_reset_accounting(crt->ip_accounting_egress_map_fd));

        zero(crt->ip_accounting_extra);

        return r;
}

void unit_reset_io_accounting_last(Unit *u) {
        assert(u);

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return;

        FOREACH_ARRAY(i, crt->io_accounting_last, _CGROUP_IO_ACCOUNTING_METRIC_MAX)
                *i = UINT64_MAX;
}

int unit_reset_io_accounting(Unit *u) {
        int r;

        assert(u);

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return 0;

        unit_reset_io_accounting_last(u);

        r = unit_get_io_accounting_raw(u, crt->io_accounting_base);
        if (r < 0) {
                zero(crt->io_accounting_base);
                return r;
        }

        return 0;
}

int unit_reset_accounting(Unit *u) {
        int r = 0;

        assert(u);

        RET_GATHER(r, unit_reset_cpu_accounting(u));
        RET_GATHER(r, unit_reset_io_accounting(u));
        RET_GATHER(r, unit_reset_ip_accounting(u));
        unit_reset_memory_accounting_last(u);

        return r;
}

void unit_invalidate_cgroup(Unit *u, CGroupMask m) {
        assert(u);

        if (!UNIT_HAS_CGROUP_CONTEXT(u))
                return;

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt)
                return;

        if (m == 0)
                return;

        /* always invalidate compat pairs together */
        if (m & (CGROUP_MASK_IO | CGROUP_MASK_BLKIO))
                m |= CGROUP_MASK_IO | CGROUP_MASK_BLKIO;

        if (m & (CGROUP_MASK_CPU | CGROUP_MASK_CPUACCT))
                m |= CGROUP_MASK_CPU | CGROUP_MASK_CPUACCT;

        if (FLAGS_SET(crt->cgroup_invalidated_mask, m)) /* NOP? */
                return;

        crt->cgroup_invalidated_mask |= m;
        unit_add_to_cgroup_realize_queue(u);
}

void unit_invalidate_cgroup_bpf(Unit *u) {
        assert(u);

        if (!UNIT_HAS_CGROUP_CONTEXT(u))
                return;

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt)
                return;

        if (crt->cgroup_invalidated_mask & CGROUP_MASK_BPF_FIREWALL) /* NOP? */
                return;

        crt->cgroup_invalidated_mask |= CGROUP_MASK_BPF_FIREWALL;
        unit_add_to_cgroup_realize_queue(u);

        /* If we are a slice unit, we also need to put compile a new BPF program for all our children, as the IP access
         * list of our children includes our own. */
        if (u->type == UNIT_SLICE) {
                Unit *member;

                UNIT_FOREACH_DEPENDENCY(member, u, UNIT_ATOM_SLICE_OF)
                        unit_invalidate_cgroup_bpf(member);
        }
}

void unit_cgroup_catchup(Unit *u) {
        assert(u);

        if (!UNIT_HAS_CGROUP_CONTEXT(u))
                return;

        /* We dropped the inotify watch during reexec/reload, so we need to
         * check these as they may have changed.
         * Note that (currently) the kernel doesn't actually update cgroup
         * file modification times, so we can't just serialize and then check
         * the mtime for file(s) we are interested in. */
        (void) unit_check_cgroup_events(u);
        unit_add_to_cgroup_oom_queue(u);
}

bool unit_cgroup_delegate(Unit *u) {
        CGroupContext *c;

        assert(u);

        if (!UNIT_VTABLE(u)->can_delegate)
                return false;

        c = unit_get_cgroup_context(u);
        if (!c)
                return false;

        return c->delegate;
}

void manager_invalidate_startup_units(Manager *m) {
        Unit *u;

        assert(m);

        SET_FOREACH(u, m->startup_units)
                unit_invalidate_cgroup(u, CGROUP_MASK_CPU|CGROUP_MASK_IO|CGROUP_MASK_BLKIO|CGROUP_MASK_CPUSET);
}

static int unit_cgroup_freezer_kernel_state(Unit *u, FreezerState *ret) {
        _cleanup_free_ char *val = NULL;
        FreezerState s;
        int r;

        assert(u);
        assert(ret);

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return -EOWNERDEAD;

        r = cg_get_keyed_attribute(
                        SYSTEMD_CGROUP_CONTROLLER,
                        crt->cgroup_path,
                        "cgroup.events",
                        STRV_MAKE("frozen"),
                        &val);
        if (IN_SET(r, -ENOENT, -ENXIO))
                return -ENODATA;
        if (r < 0)
                return r;

        if (streq(val, "0"))
                s = FREEZER_RUNNING;
        else if (streq(val, "1"))
                s = FREEZER_FROZEN;
        else {
                log_unit_debug(u, "Unexpected cgroup frozen state: %s", val);
                s = _FREEZER_STATE_INVALID;
        }

        *ret = s;
        return 0;
}

int unit_cgroup_freezer_action(Unit *u, FreezerAction action) {
        _cleanup_free_ char *path = NULL;
        FreezerState target, current, next;
        int r;

        assert(u);
        assert(IN_SET(action, FREEZER_FREEZE, FREEZER_PARENT_FREEZE,
                              FREEZER_THAW, FREEZER_PARENT_THAW));

        if (!cg_freezer_supported())
                return 0;

        unit_next_freezer_state(u, action, &next, &target);

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_realized) {
                /* No realized cgroup = nothing to freeze */
                u->freezer_state = freezer_state_finish(next);
                return 0;
        }

        r = unit_cgroup_freezer_kernel_state(u, &current);
        if (r < 0)
                return r;

        if (current == target)
                next = freezer_state_finish(next);
        else if (IN_SET(next, FREEZER_FROZEN, FREEZER_FROZEN_BY_PARENT, FREEZER_RUNNING)) {
                /* We're transitioning into a finished state, which implies that the cgroup's
                 * current state already matches the target and thus we'd return 0. But, reality
                 * shows otherwise. This indicates that our freezer_state tracking has diverged
                 * from the real state of the cgroup, which can happen if someone meddles with the
                 * cgroup from underneath us. This really shouldn't happen during normal operation,
                 * though. So, let's warn about it and fix up the state to be valid */

                log_unit_warning(u, "Unit wants to transition to %s freezer state but cgroup is unexpectedly %s, fixing up.",
                                 freezer_state_to_string(next), freezer_state_to_string(current) ?: "(invalid)");

                if (next == FREEZER_FROZEN)
                        next = FREEZER_FREEZING;
                else if (next == FREEZER_FROZEN_BY_PARENT)
                        next = FREEZER_FREEZING_BY_PARENT;
                else if (next == FREEZER_RUNNING)
                        next = FREEZER_THAWING;
        }

        r = cg_get_path(SYSTEMD_CGROUP_CONTROLLER, crt->cgroup_path, "cgroup.freeze", &path);
        if (r < 0)
                return r;

        log_unit_debug(u, "Unit freezer state was %s, now %s.",
                       freezer_state_to_string(u->freezer_state),
                       freezer_state_to_string(next));

        r = write_string_file(path, one_zero(target == FREEZER_FROZEN), WRITE_STRING_FILE_DISABLE_BUFFER);
        if (r < 0)
                return r;

        u->freezer_state = next;
        return target != current;
}

int unit_get_cpuset(Unit *u, CPUSet *cpus, const char *name) {
        _cleanup_free_ char *v = NULL;
        int r;

        assert(u);
        assert(cpus);

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt || !crt->cgroup_path)
                return -ENODATA;

        if ((crt->cgroup_realized_mask & CGROUP_MASK_CPUSET) == 0)
                return -ENODATA;

        r = cg_all_unified();
        if (r < 0)
                return r;
        if (r == 0)
                return -ENODATA;

        r = cg_get_attribute("cpuset", crt->cgroup_path, name, &v);
        if (r == -ENOENT)
                return -ENODATA;
        if (r < 0)
                return r;

        return parse_cpu_set_full(v, cpus, false, NULL, NULL, 0, NULL);
}

CGroupRuntime *cgroup_runtime_new(void) {
        _cleanup_(cgroup_runtime_freep) CGroupRuntime *crt = NULL;

        crt = new(CGroupRuntime, 1);
        if (!crt)
                return NULL;

        *crt = (CGroupRuntime) {
                .cpu_usage_last = NSEC_INFINITY,

                .cgroup_control_inotify_wd = -1,
                .cgroup_memory_inotify_wd = -1,

                .ip_accounting_ingress_map_fd = -EBADF,
                .ip_accounting_egress_map_fd = -EBADF,

                .ipv4_allow_map_fd = -EBADF,
                .ipv6_allow_map_fd = -EBADF,
                .ipv4_deny_map_fd = -EBADF,
                .ipv6_deny_map_fd = -EBADF,

                .cgroup_invalidated_mask = _CGROUP_MASK_ALL,
        };

        FOREACH_ELEMENT(i, crt->memory_accounting_last)
                *i = UINT64_MAX;
        FOREACH_ELEMENT(i, crt->io_accounting_base)
                *i = UINT64_MAX;
        FOREACH_ELEMENT(i, crt->io_accounting_last)
                *i = UINT64_MAX;
        FOREACH_ELEMENT(i, crt->ip_accounting_extra)
                *i = UINT64_MAX;

        return TAKE_PTR(crt);
}

CGroupRuntime *cgroup_runtime_free(CGroupRuntime *crt) {
        if (!crt)
                return NULL;

        fdset_free(crt->initial_socket_bind_link_fds);
#if BPF_FRAMEWORK
        bpf_link_free(crt->ipv4_socket_bind_link);
        bpf_link_free(crt->ipv6_socket_bind_link);
#endif
        hashmap_free(crt->bpf_foreign_by_key);

        bpf_program_free(crt->bpf_device_control_installed);

#if BPF_FRAMEWORK
        bpf_link_free(crt->restrict_ifaces_ingress_bpf_link);
        bpf_link_free(crt->restrict_ifaces_egress_bpf_link);
#endif
        fdset_free(crt->initial_restrict_ifaces_link_fds);

        safe_close(crt->ipv4_allow_map_fd);
        safe_close(crt->ipv6_allow_map_fd);
        safe_close(crt->ipv4_deny_map_fd);
        safe_close(crt->ipv6_deny_map_fd);

        bpf_program_free(crt->ip_bpf_ingress);
        bpf_program_free(crt->ip_bpf_ingress_installed);
        bpf_program_free(crt->ip_bpf_egress);
        bpf_program_free(crt->ip_bpf_egress_installed);

        set_free(crt->ip_bpf_custom_ingress);
        set_free(crt->ip_bpf_custom_ingress_installed);
        set_free(crt->ip_bpf_custom_egress);
        set_free(crt->ip_bpf_custom_egress_installed);

        free(crt->cgroup_path);

        return mfree(crt);
}

static const char* const ip_accounting_metric_field_table[_CGROUP_IP_ACCOUNTING_METRIC_MAX] = {
        [CGROUP_IP_INGRESS_BYTES]   = "ip-accounting-ingress-bytes",
        [CGROUP_IP_INGRESS_PACKETS] = "ip-accounting-ingress-packets",
        [CGROUP_IP_EGRESS_BYTES]    = "ip-accounting-egress-bytes",
        [CGROUP_IP_EGRESS_PACKETS]  = "ip-accounting-egress-packets",
};

DEFINE_PRIVATE_STRING_TABLE_LOOKUP(ip_accounting_metric_field, CGroupIPAccountingMetric);

static const char* const io_accounting_metric_field_base_table[_CGROUP_IO_ACCOUNTING_METRIC_MAX] = {
        [CGROUP_IO_READ_BYTES]       = "io-accounting-read-bytes-base",
        [CGROUP_IO_WRITE_BYTES]      = "io-accounting-write-bytes-base",
        [CGROUP_IO_READ_OPERATIONS]  = "io-accounting-read-operations-base",
        [CGROUP_IO_WRITE_OPERATIONS] = "io-accounting-write-operations-base",
};

DEFINE_PRIVATE_STRING_TABLE_LOOKUP(io_accounting_metric_field_base, CGroupIOAccountingMetric);

static const char* const io_accounting_metric_field_last_table[_CGROUP_IO_ACCOUNTING_METRIC_MAX] = {
        [CGROUP_IO_READ_BYTES]       = "io-accounting-read-bytes-last",
        [CGROUP_IO_WRITE_BYTES]      = "io-accounting-write-bytes-last",
        [CGROUP_IO_READ_OPERATIONS]  = "io-accounting-read-operations-last",
        [CGROUP_IO_WRITE_OPERATIONS] = "io-accounting-write-operations-last",
};

DEFINE_PRIVATE_STRING_TABLE_LOOKUP(io_accounting_metric_field_last, CGroupIOAccountingMetric);

static const char* const memory_accounting_metric_field_last_table[_CGROUP_MEMORY_ACCOUNTING_METRIC_CACHED_LAST + 1] = {
        [CGROUP_MEMORY_PEAK]      = "memory-accounting-peak",
        [CGROUP_MEMORY_SWAP_PEAK] = "memory-accounting-swap-peak",
};

DEFINE_PRIVATE_STRING_TABLE_LOOKUP(memory_accounting_metric_field_last, CGroupMemoryAccountingMetric);

static int serialize_cgroup_mask(FILE *f, const char *key, CGroupMask mask) {
        _cleanup_free_ char *s = NULL;
        int r;

        assert(f);
        assert(key);

        if (mask == 0)
                return 0;

        r = cg_mask_to_string(mask, &s);
        if (r < 0)
                return log_error_errno(r, "Failed to format cgroup mask: %m");

        return serialize_item(f, key, s);
}

int cgroup_runtime_serialize(Unit *u, FILE *f, FDSet *fds) {
        int r;

        assert(u);
        assert(f);
        assert(fds);

        CGroupRuntime *crt = unit_get_cgroup_runtime(u);
        if (!crt)
                return 0;

        (void) serialize_item_format(f, "cpu-usage-base", "%" PRIu64, crt->cpu_usage_base);
        if (crt->cpu_usage_last != NSEC_INFINITY)
                (void) serialize_item_format(f, "cpu-usage-last", "%" PRIu64, crt->cpu_usage_last);

        if (crt->managed_oom_kill_last > 0)
                (void) serialize_item_format(f, "managed-oom-kill-last", "%" PRIu64, crt->managed_oom_kill_last);

        if (crt->oom_kill_last > 0)
                (void) serialize_item_format(f, "oom-kill-last", "%" PRIu64, crt->oom_kill_last);

        for (CGroupMemoryAccountingMetric metric = 0; metric <= _CGROUP_MEMORY_ACCOUNTING_METRIC_CACHED_LAST; metric++) {
                uint64_t v;

                r = unit_get_memory_accounting(u, metric, &v);
                if (r >= 0)
                        (void) serialize_item_format(f, memory_accounting_metric_field_last_to_string(metric), "%" PRIu64, v);
        }

        for (CGroupIPAccountingMetric m = 0; m < _CGROUP_IP_ACCOUNTING_METRIC_MAX; m++) {
                uint64_t v;

                r = unit_get_ip_accounting(u, m, &v);
                if (r >= 0)
                        (void) serialize_item_format(f, ip_accounting_metric_field_to_string(m), "%" PRIu64, v);
        }

        for (CGroupIOAccountingMetric im = 0; im < _CGROUP_IO_ACCOUNTING_METRIC_MAX; im++) {
                (void) serialize_item_format(f, io_accounting_metric_field_base_to_string(im), "%" PRIu64, crt->io_accounting_base[im]);

                if (crt->io_accounting_last[im] != UINT64_MAX)
                        (void) serialize_item_format(f, io_accounting_metric_field_last_to_string(im), "%" PRIu64, crt->io_accounting_last[im]);
        }

        if (crt->cgroup_path)
                (void) serialize_item(f, "cgroup", crt->cgroup_path);
        if (crt->cgroup_id != 0)
                (void) serialize_item_format(f, "cgroup-id", "%" PRIu64, crt->cgroup_id);

        (void) serialize_bool(f, "cgroup-realized", crt->cgroup_realized);
        (void) serialize_cgroup_mask(f, "cgroup-realized-mask", crt->cgroup_realized_mask);
        (void) serialize_cgroup_mask(f, "cgroup-enabled-mask", crt->cgroup_enabled_mask);
        (void) serialize_cgroup_mask(f, "cgroup-invalidated-mask", crt->cgroup_invalidated_mask);

        (void) bpf_socket_bind_serialize(u, f, fds);

        (void) bpf_program_serialize_attachment(f, fds, "ip-bpf-ingress-installed", crt->ip_bpf_ingress_installed);
        (void) bpf_program_serialize_attachment(f, fds, "ip-bpf-egress-installed", crt->ip_bpf_egress_installed);
        (void) bpf_program_serialize_attachment(f, fds, "bpf-device-control-installed", crt->bpf_device_control_installed);
        (void) bpf_program_serialize_attachment_set(f, fds, "ip-bpf-custom-ingress-installed", crt->ip_bpf_custom_ingress_installed);
        (void) bpf_program_serialize_attachment_set(f, fds, "ip-bpf-custom-egress-installed", crt->ip_bpf_custom_egress_installed);

        (void) bpf_restrict_ifaces_serialize(u, f, fds);

        return 0;
}

#define MATCH_DESERIALIZE(u, key, l, v, parse_func, target)             \
        ({                                                              \
                bool _deserialize_matched = streq(l, key);              \
                if (_deserialize_matched) {                             \
                        CGroupRuntime *crt = unit_setup_cgroup_runtime(u); \
                        if (!crt)                                       \
                                log_oom_debug();                        \
                        else {                                          \
                                int _deserialize_r = parse_func(v);     \
                                if (_deserialize_r < 0)                 \
                                        log_unit_debug_errno(u, _deserialize_r, \
                                                             "Failed to parse \"%s=%s\", ignoring.", l, v); \
                                else                                    \
                                        crt->target = _deserialize_r; \
                        }                                               \
                }                                                       \
                _deserialize_matched;                                   \
        })

#define MATCH_DESERIALIZE_IMMEDIATE(u, key, l, v, parse_func, target)   \
        ({                                                              \
                 bool _deserialize_matched = streq(l, key);             \
                 if (_deserialize_matched) {                            \
                         CGroupRuntime *crt = unit_setup_cgroup_runtime(u); \
                         if (!crt)                                      \
                                 log_oom_debug();                       \
                         else {                                         \
                                 int _deserialize_r = parse_func(v, &crt->target); \
                                 if (_deserialize_r < 0)                \
                                         log_unit_debug_errno(u, _deserialize_r, \
                                                              "Failed to parse \"%s=%s\", ignoring", l, v); \
                         }                                              \
                 }                                                      \
                _deserialize_matched;                                   \
        })

#define MATCH_DESERIALIZE_METRIC(u, key, l, v, parse_func, target)             \
        ({                                                              \
                bool _deserialize_matched = streq(l, key);              \
                if (_deserialize_matched) {                             \
                        CGroupRuntime *crt = unit_setup_cgroup_runtime(u); \
                        if (!crt)                                       \
                                log_oom_debug();                        \
                        else {                                          \
                                int _deserialize_r = parse_func(v);     \
                                if (_deserialize_r < 0)                 \
                                        log_unit_debug_errno(u, _deserialize_r, \
                                                             "Failed to parse \"%s=%s\", ignoring.", l, v); \
                                else                                    \
                                        crt->target = _deserialize_r; \
                        }                                               \
                }                                                       \
                _deserialize_matched;                                   \
        })

int cgroup_runtime_deserialize_one(Unit *u, const char *key, const char *value, FDSet *fds) {
        int r;

        assert(u);
        assert(value);

        if (!UNIT_HAS_CGROUP_CONTEXT(u))
                return 0;

        if (MATCH_DESERIALIZE_IMMEDIATE(u, "cpu-usage-base", key, value, safe_atou64, cpu_usage_base) ||
            MATCH_DESERIALIZE_IMMEDIATE(u, "cpuacct-usage-base", key, value, safe_atou64, cpu_usage_base))
                return 1;

        if (MATCH_DESERIALIZE_IMMEDIATE(u, "cpu-usage-last", key, value, safe_atou64, cpu_usage_last))
                return 1;

        if (MATCH_DESERIALIZE_IMMEDIATE(u, "managed-oom-kill-last", key, value, safe_atou64, managed_oom_kill_last))
                return 1;

        if (MATCH_DESERIALIZE_IMMEDIATE(u, "oom-kill-last", key, value, safe_atou64, oom_kill_last))
                return 1;

        if (streq(key, "cgroup")) {
                r = unit_set_cgroup_path(u, value);
                if (r < 0)
                        log_unit_debug_errno(u, r, "Failed to set cgroup path %s, ignoring: %m", value);

                (void) unit_watch_cgroup(u);
                (void) unit_watch_cgroup_memory(u);
                return 1;
        }

        if (MATCH_DESERIALIZE_IMMEDIATE(u, "cgroup-id", key, value, safe_atou64, cgroup_id))
                return 1;

        if (MATCH_DESERIALIZE(u, "cgroup-realized", key, value, parse_boolean, cgroup_realized))
                return 1;

        if (MATCH_DESERIALIZE_IMMEDIATE(u, "cgroup-realized-mask", key, value, cg_mask_from_string, cgroup_realized_mask))
                return 1;

        if (MATCH_DESERIALIZE_IMMEDIATE(u, "cgroup-enabled-mask", key, value, cg_mask_from_string, cgroup_enabled_mask))
                return 1;

        if (MATCH_DESERIALIZE_IMMEDIATE(u, "cgroup-invalidated-mask", key, value, cg_mask_from_string, cgroup_invalidated_mask))
                return 1;

        if (STR_IN_SET(key, "ipv4-socket-bind-bpf-link-fd", "ipv6-socket-bind-bpf-link-fd")) {
                int fd;

                fd = deserialize_fd(fds, value);
                if (fd >= 0)
                        (void) bpf_socket_bind_add_initial_link_fd(u, fd);

                return 1;
        }

        if (STR_IN_SET(key,
                       "ip-bpf-ingress-installed", "ip-bpf-egress-installed",
                       "bpf-device-control-installed",
                       "ip-bpf-custom-ingress-installed", "ip-bpf-custom-egress-installed")) {

                CGroupRuntime *crt = unit_setup_cgroup_runtime(u);
                if (!crt)
                        log_oom_debug();
                else {
                        if (streq(key, "ip-bpf-ingress-installed"))
                                (void) bpf_program_deserialize_attachment(value, fds, &crt->ip_bpf_ingress_installed);

                        if (streq(key, "ip-bpf-egress-installed"))
                                (void) bpf_program_deserialize_attachment(value, fds, &crt->ip_bpf_egress_installed);

                        if (streq(key, "bpf-device-control-installed"))
                                (void) bpf_program_deserialize_attachment(value, fds, &crt->bpf_device_control_installed);

                        if (streq(key, "ip-bpf-custom-ingress-installed"))
                                (void) bpf_program_deserialize_attachment_set(value, fds, &crt->ip_bpf_custom_ingress_installed);

                        if (streq(key, "ip-bpf-custom-egress-installed"))
                                (void) bpf_program_deserialize_attachment_set(value, fds, &crt->ip_bpf_custom_egress_installed);
                }

                return 1;
        }

        if (streq(key, "restrict-ifaces-bpf-fd")) {
                int fd;

                fd = deserialize_fd(fds, value);
                if (fd >= 0)
                        (void) bpf_restrict_ifaces_add_initial_link_fd(u, fd);
                return 1;
        }

        CGroupMemoryAccountingMetric mm = memory_accounting_metric_field_last_from_string(key);
        if (mm >= 0) {
                uint64_t c;

                r = safe_atou64(value, &c);
                if (r < 0)
                        log_unit_debug(u, "Failed to parse memory accounting last value %s, ignoring.", value);
                else {
                        CGroupRuntime *crt = unit_setup_cgroup_runtime(u);
                        if (!crt)
                                log_oom_debug();
                        else
                                crt->memory_accounting_last[mm] = c;
                }

                return 1;
        }

        CGroupIPAccountingMetric ipm = ip_accounting_metric_field_from_string(key);
        if (ipm >= 0) {
                uint64_t c;

                r = safe_atou64(value, &c);
                if (r < 0)
                        log_unit_debug(u, "Failed to parse IP accounting value %s, ignoring.", value);
                else {
                        CGroupRuntime *crt = unit_setup_cgroup_runtime(u);
                        if (!crt)
                                log_oom_debug();
                        else
                                crt->ip_accounting_extra[ipm] = c;
                }

                return 1;
        }

        CGroupIOAccountingMetric iom = io_accounting_metric_field_base_from_string(key);
        if (iom >= 0) {
                uint64_t c;

                r = safe_atou64(value, &c);
                if (r < 0)
                        log_unit_debug(u, "Failed to parse IO accounting base value %s, ignoring.", value);
                else {
                        CGroupRuntime *crt = unit_setup_cgroup_runtime(u);
                        if (!crt)
                                log_oom_debug();
                        else
                                crt->io_accounting_base[iom] = c;
                }

                return 1;
        }

        iom = io_accounting_metric_field_last_from_string(key);
        if (iom >= 0) {
                uint64_t c;

                r = safe_atou64(value, &c);
                if (r < 0)
                        log_unit_debug(u, "Failed to parse IO accounting last value %s, ignoring.", value);
                else {
                        CGroupRuntime *crt = unit_setup_cgroup_runtime(u);
                        if (!crt)
                                log_oom_debug();
                        else
                                crt->io_accounting_last[iom] = c;
                }
                return 1;
        }

        return 0;
}

static const char* const cgroup_device_policy_table[_CGROUP_DEVICE_POLICY_MAX] = {
        [CGROUP_DEVICE_POLICY_AUTO]   = "auto",
        [CGROUP_DEVICE_POLICY_CLOSED] = "closed",
        [CGROUP_DEVICE_POLICY_STRICT] = "strict",
};

DEFINE_STRING_TABLE_LOOKUP(cgroup_device_policy, CGroupDevicePolicy);

static const char* const cgroup_pressure_watch_table[_CGROUP_PRESSURE_WATCH_MAX] = {
        [CGROUP_PRESSURE_WATCH_OFF]  = "off",
        [CGROUP_PRESSURE_WATCH_AUTO] = "auto",
        [CGROUP_PRESSURE_WATCH_ON]   = "on",
        [CGROUP_PRESSURE_WATCH_SKIP] = "skip",
};

DEFINE_STRING_TABLE_LOOKUP_WITH_BOOLEAN(cgroup_pressure_watch, CGroupPressureWatch, CGROUP_PRESSURE_WATCH_ON);

static const char* const cgroup_ip_accounting_metric_table[_CGROUP_IP_ACCOUNTING_METRIC_MAX] = {
        [CGROUP_IP_INGRESS_BYTES]   = "IPIngressBytes",
        [CGROUP_IP_EGRESS_BYTES]    = "IPEgressBytes",
        [CGROUP_IP_INGRESS_PACKETS] = "IPIngressPackets",
        [CGROUP_IP_EGRESS_PACKETS]  = "IPEgressPackets",
};

DEFINE_STRING_TABLE_LOOKUP(cgroup_ip_accounting_metric, CGroupIPAccountingMetric);

static const char* const cgroup_io_accounting_metric_table[_CGROUP_IO_ACCOUNTING_METRIC_MAX] = {
        [CGROUP_IO_READ_BYTES]       = "IOReadBytes",
        [CGROUP_IO_WRITE_BYTES]      = "IOWriteBytes",
        [CGROUP_IO_READ_OPERATIONS]  = "IOReadOperations",
        [CGROUP_IO_WRITE_OPERATIONS] = "IOWriteOperations",
};

DEFINE_STRING_TABLE_LOOKUP(cgroup_io_accounting_metric, CGroupIOAccountingMetric);

static const char* const cgroup_memory_accounting_metric_table[_CGROUP_MEMORY_ACCOUNTING_METRIC_MAX] = {
        [CGROUP_MEMORY_PEAK]          = "MemoryPeak",
        [CGROUP_MEMORY_SWAP_CURRENT]  = "MemorySwapCurrent",
        [CGROUP_MEMORY_SWAP_PEAK]     = "MemorySwapPeak",
        [CGROUP_MEMORY_ZSWAP_CURRENT] = "MemoryZSwapCurrent",
};

DEFINE_STRING_TABLE_LOOKUP(cgroup_memory_accounting_metric, CGroupMemoryAccountingMetric);

static const char *const cgroup_effective_limit_type_table[_CGROUP_LIMIT_TYPE_MAX] = {
        [CGROUP_LIMIT_MEMORY_MAX]  = "EffectiveMemoryMax",
        [CGROUP_LIMIT_MEMORY_HIGH] = "EffectiveMemoryHigh",
        [CGROUP_LIMIT_TASKS_MAX]   = "EffectiveTasksMax",
};

DEFINE_STRING_TABLE_LOOKUP(cgroup_effective_limit_type, CGroupLimitType);