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
|
/*
* Copyright (c) 2009-2012, Salvatore Sanfilippo <antirez at gmail dot com>
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions are met:
*
* * Redistributions of source code must retain the above copyright notice,
* this list of conditions and the following disclaimer.
* * Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* * Neither the name of Redis nor the names of its contributors may be used
* to endorse or promote products derived from this software without
* specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
* AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
* LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
* CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
* SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
* INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
* CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
* ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
* POSSIBILITY OF SUCH DAMAGE.
*/
#include "server.h"
#include "atomicvar.h"
#include "cluster.h"
#include "script.h"
#include <sys/socket.h>
#include <sys/uio.h>
#include <math.h>
#include <ctype.h>
static void setProtocolError(const char *errstr, client *c);
int postponeClientRead(client *c);
int ProcessingEventsWhileBlocked = 0; /* See processEventsWhileBlocked(). */
/* Return the size consumed from the allocator, for the specified SDS string,
* including internal fragmentation. This function is used in order to compute
* the client output buffer size. */
size_t sdsZmallocSize(sds s) {
void *sh = sdsAllocPtr(s);
return zmalloc_size(sh);
}
/* Return the amount of memory used by the sds string at object->ptr
* for a string object. This includes internal fragmentation. */
size_t getStringObjectSdsUsedMemory(robj *o) {
serverAssertWithInfo(NULL,o,o->type == OBJ_STRING);
switch(o->encoding) {
case OBJ_ENCODING_RAW: return sdsZmallocSize(o->ptr);
case OBJ_ENCODING_EMBSTR: return zmalloc_size(o)-sizeof(robj);
default: return 0; /* Just integer encoding for now. */
}
}
/* Return the length of a string object.
* This does NOT includes internal fragmentation or sds unused space. */
size_t getStringObjectLen(robj *o) {
serverAssertWithInfo(NULL,o,o->type == OBJ_STRING);
switch(o->encoding) {
case OBJ_ENCODING_RAW: return sdslen(o->ptr);
case OBJ_ENCODING_EMBSTR: return sdslen(o->ptr);
default: return 0; /* Just integer encoding for now. */
}
}
/* Client.reply list dup and free methods. */
void *dupClientReplyValue(void *o) {
clientReplyBlock *old = o;
clientReplyBlock *buf = zmalloc(sizeof(clientReplyBlock) + old->size);
memcpy(buf, o, sizeof(clientReplyBlock) + old->size);
return buf;
}
void freeClientReplyValue(void *o) {
zfree(o);
}
int listMatchObjects(void *a, void *b) {
return equalStringObjects(a,b);
}
/* This function links the client to the global linked list of clients.
* unlinkClient() does the opposite, among other things. */
void linkClient(client *c) {
listAddNodeTail(server.clients,c);
/* Note that we remember the linked list node where the client is stored,
* this way removing the client in unlinkClient() will not require
* a linear scan, but just a constant time operation. */
c->client_list_node = listLast(server.clients);
uint64_t id = htonu64(c->id);
raxInsert(server.clients_index,(unsigned char*)&id,sizeof(id),c,NULL);
}
/* Initialize client authentication state.
*/
static void clientSetDefaultAuth(client *c) {
/* If the default user does not require authentication, the user is
* directly authenticated. */
c->user = DefaultUser;
c->authenticated = (c->user->flags & USER_FLAG_NOPASS) &&
!(c->user->flags & USER_FLAG_DISABLED);
}
int authRequired(client *c) {
/* Check if the user is authenticated. This check is skipped in case
* the default user is flagged as "nopass" and is active. */
int auth_required = (!(DefaultUser->flags & USER_FLAG_NOPASS) ||
(DefaultUser->flags & USER_FLAG_DISABLED)) &&
!c->authenticated;
return auth_required;
}
client *createClient(connection *conn) {
client *c = zmalloc(sizeof(client));
/* passing NULL as conn it is possible to create a non connected client.
* This is useful since all the commands needs to be executed
* in the context of a client. When commands are executed in other
* contexts (for instance a Lua script) we need a non connected client. */
if (conn) {
connEnableTcpNoDelay(conn);
if (server.tcpkeepalive)
connKeepAlive(conn,server.tcpkeepalive);
connSetReadHandler(conn, readQueryFromClient);
connSetPrivateData(conn, c);
}
selectDb(c,0);
uint64_t client_id;
atomicGetIncr(server.next_client_id, client_id, 1);
c->id = client_id;
c->resp = 2;
c->conn = conn;
c->name = NULL;
c->bufpos = 0;
c->buf_usable_size = zmalloc_usable_size(c)-offsetof(client,buf);
c->ref_repl_buf_node = NULL;
c->ref_block_pos = 0;
c->qb_pos = 0;
c->querybuf = sdsempty();
c->pending_querybuf = sdsempty();
c->querybuf_peak = 0;
c->reqtype = 0;
c->argc = 0;
c->argv = NULL;
c->argv_len = 0;
c->argv_len_sum = 0;
c->original_argc = 0;
c->original_argv = NULL;
c->cmd = c->lastcmd = NULL;
c->multibulklen = 0;
c->bulklen = -1;
c->sentlen = 0;
c->flags = 0;
c->ctime = c->lastinteraction = server.unixtime;
clientSetDefaultAuth(c);
c->replstate = REPL_STATE_NONE;
c->repl_start_cmd_stream_on_ack = 0;
c->reploff = 0;
c->read_reploff = 0;
c->repl_ack_off = 0;
c->repl_ack_time = 0;
c->repl_last_partial_write = 0;
c->slave_listening_port = 0;
c->slave_addr = NULL;
c->slave_capa = SLAVE_CAPA_NONE;
c->slave_req = SLAVE_REQ_NONE;
c->reply = listCreate();
c->reply_bytes = 0;
c->obuf_soft_limit_reached_time = 0;
listSetFreeMethod(c->reply,freeClientReplyValue);
listSetDupMethod(c->reply,dupClientReplyValue);
c->btype = BLOCKED_NONE;
c->bpop.timeout = 0;
c->bpop.keys = dictCreate(&objectKeyHeapPointerValueDictType);
c->bpop.target = NULL;
c->bpop.xread_group = NULL;
c->bpop.xread_consumer = NULL;
c->bpop.xread_group_noack = 0;
c->bpop.numreplicas = 0;
c->bpop.reploffset = 0;
c->woff = 0;
c->watched_keys = listCreate();
c->pubsub_channels = dictCreate(&objectKeyPointerValueDictType);
c->pubsub_patterns = listCreate();
c->pubsubshard_channels = dictCreate(&objectKeyPointerValueDictType);
c->peerid = NULL;
c->sockname = NULL;
c->client_list_node = NULL;
c->paused_list_node = NULL;
c->pending_read_list_node = NULL;
c->client_tracking_redirection = 0;
c->client_tracking_prefixes = NULL;
c->last_memory_usage = c->last_memory_usage_on_bucket_update = 0;
c->last_memory_type = CLIENT_TYPE_NORMAL;
c->auth_callback = NULL;
c->auth_callback_privdata = NULL;
c->auth_module = NULL;
listSetFreeMethod(c->pubsub_patterns,decrRefCountVoid);
listSetMatchMethod(c->pubsub_patterns,listMatchObjects);
c->mem_usage_bucket = NULL;
c->mem_usage_bucket_node = NULL;
if (conn) linkClient(c);
initClientMultiState(c);
return c;
}
/* This function puts the client in the queue of clients that should write
* their output buffers to the socket. Note that it does not *yet* install
* the write handler, to start clients are put in a queue of clients that need
* to write, so we try to do that before returning in the event loop (see the
* handleClientsWithPendingWrites() function).
* If we fail and there is more data to write, compared to what the socket
* buffers can hold, then we'll really install the handler. */
void clientInstallWriteHandler(client *c) {
/* Schedule the client to write the output buffers to the socket only
* if not already done and, for slaves, if the slave can actually receive
* writes at this stage. */
if (!(c->flags & CLIENT_PENDING_WRITE) &&
(c->replstate == REPL_STATE_NONE ||
(c->replstate == SLAVE_STATE_ONLINE && !c->repl_start_cmd_stream_on_ack)))
{
/* Here instead of installing the write handler, we just flag the
* client and put it into a list of clients that have something
* to write to the socket. This way before re-entering the event
* loop, we can try to directly write to the client sockets avoiding
* a system call. We'll only really install the write handler if
* we'll not be able to write the whole reply at once. */
c->flags |= CLIENT_PENDING_WRITE;
listAddNodeHead(server.clients_pending_write,c);
}
}
/* This function is called every time we are going to transmit new data
* to the client. The behavior is the following:
*
* If the client should receive new data (normal clients will) the function
* returns C_OK, and make sure to install the write handler in our event
* loop so that when the socket is writable new data gets written.
*
* If the client should not receive new data, because it is a fake client
* (used to load AOF in memory), a master or because the setup of the write
* handler failed, the function returns C_ERR.
*
* The function may return C_OK without actually installing the write
* event handler in the following cases:
*
* 1) The event handler should already be installed since the output buffer
* already contains something.
* 2) The client is a slave but not yet online, so we want to just accumulate
* writes in the buffer but not actually sending them yet.
*
* Typically gets called every time a reply is built, before adding more
* data to the clients output buffers. If the function returns C_ERR no
* data should be appended to the output buffers. */
int prepareClientToWrite(client *c) {
/* If it's the Lua client we always return ok without installing any
* handler since there is no socket at all. */
if (c->flags & (CLIENT_SCRIPT|CLIENT_MODULE)) return C_OK;
/* If CLIENT_CLOSE_ASAP flag is set, we need not write anything. */
if (c->flags & CLIENT_CLOSE_ASAP) return C_ERR;
/* CLIENT REPLY OFF / SKIP handling: don't send replies. */
if (c->flags & (CLIENT_REPLY_OFF|CLIENT_REPLY_SKIP)) return C_ERR;
/* Masters don't receive replies, unless CLIENT_MASTER_FORCE_REPLY flag
* is set. */
if ((c->flags & CLIENT_MASTER) &&
!(c->flags & CLIENT_MASTER_FORCE_REPLY)) return C_ERR;
if (!c->conn) return C_ERR; /* Fake client for AOF loading. */
/* Schedule the client to write the output buffers to the socket, unless
* it should already be setup to do so (it has already pending data).
*
* If CLIENT_PENDING_READ is set, we're in an IO thread and should
* not install a write handler. Instead, it will be done by
* handleClientsWithPendingReadsUsingThreads() upon return.
*/
if (!clientHasPendingReplies(c) && io_threads_op == IO_THREADS_OP_IDLE)
clientInstallWriteHandler(c);
/* Authorize the caller to queue in the output buffer of this client. */
return C_OK;
}
/* -----------------------------------------------------------------------------
* Low level functions to add more data to output buffers.
* -------------------------------------------------------------------------- */
/* Attempts to add the reply to the static buffer in the client struct.
* Returns the length of data that is added to the reply buffer.
*
* Sanitizer suppression: client->buf_usable_size determined by
* zmalloc_usable_size() call. Writing beyond client->buf boundaries confuses
* sanitizer and generates a false positive out-of-bounds error */
REDIS_NO_SANITIZE("bounds")
size_t _addReplyToBuffer(client *c, const char *s, size_t len) {
size_t available = c->buf_usable_size - c->bufpos;
/* If there already are entries in the reply list, we cannot
* add anything more to the static buffer. */
if (listLength(c->reply) > 0) return 0;
size_t reply_len = len > available ? available : len;
memcpy(c->buf+c->bufpos,s,reply_len);
c->bufpos+=reply_len;
return reply_len;
}
/* Adds the reply to the reply linked list.
* Note: some edits to this function need to be relayed to AddReplyFromClient. */
void _addReplyProtoToList(client *c, const char *s, size_t len) {
listNode *ln = listLast(c->reply);
clientReplyBlock *tail = ln? listNodeValue(ln): NULL;
/* Note that 'tail' may be NULL even if we have a tail node, because when
* addReplyDeferredLen() is used, it sets a dummy node to NULL just
* to fill it later, when the size of the bulk length is set. */
/* Append to tail string when possible. */
if (tail) {
/* Copy the part we can fit into the tail, and leave the rest for a
* new node */
size_t avail = tail->size - tail->used;
size_t copy = avail >= len? len: avail;
memcpy(tail->buf + tail->used, s, copy);
tail->used += copy;
s += copy;
len -= copy;
}
if (len) {
/* Create a new node, make sure it is allocated to at
* least PROTO_REPLY_CHUNK_BYTES */
size_t usable_size;
size_t size = len < PROTO_REPLY_CHUNK_BYTES? PROTO_REPLY_CHUNK_BYTES: len;
tail = zmalloc_usable(size + sizeof(clientReplyBlock), &usable_size);
/* take over the allocation's internal fragmentation */
tail->size = usable_size - sizeof(clientReplyBlock);
tail->used = len;
memcpy(tail->buf, s, len);
listAddNodeTail(c->reply, tail);
c->reply_bytes += tail->size;
closeClientOnOutputBufferLimitReached(c, 1);
}
}
void _addReplyToBufferOrList(client *c, const char *s, size_t len) {
if (c->flags & CLIENT_CLOSE_AFTER_REPLY) return;
/* Replicas should normally not cause any writes to the reply buffer. In case a rogue replica sent a command on the
* replication link that caused a reply to be generated we'll simply disconnect it.
* Note this is the simplest way to check a command added a response. Replication links are used to write data but
* not for responses, so we should normally never get here on a replica client. */
if (getClientType(c) == CLIENT_TYPE_SLAVE) {
sds cmdname = c->lastcmd ? getFullCommandName(c->lastcmd) : NULL;
logInvalidUseAndFreeClientAsync(c, "Replica generated a reply to command %s",
cmdname ? cmdname : "<unknown>");
sdsfree(cmdname);
return;
}
size_t reply_len = _addReplyToBuffer(c,s,len);
if (len > reply_len) _addReplyProtoToList(c,s+reply_len,len-reply_len);
}
/* -----------------------------------------------------------------------------
* Higher level functions to queue data on the client output buffer.
* The following functions are the ones that commands implementations will call.
* -------------------------------------------------------------------------- */
/* Add the object 'obj' string representation to the client output buffer. */
void addReply(client *c, robj *obj) {
if (prepareClientToWrite(c) != C_OK) return;
if (sdsEncodedObject(obj)) {
_addReplyToBufferOrList(c,obj->ptr,sdslen(obj->ptr));
} else if (obj->encoding == OBJ_ENCODING_INT) {
/* For integer encoded strings we just convert it into a string
* using our optimized function, and attach the resulting string
* to the output buffer. */
char buf[32];
size_t len = ll2string(buf,sizeof(buf),(long)obj->ptr);
_addReplyToBufferOrList(c,buf,len);
} else {
serverPanic("Wrong obj->encoding in addReply()");
}
}
/* Add the SDS 's' string to the client output buffer, as a side effect
* the SDS string is freed. */
void addReplySds(client *c, sds s) {
if (prepareClientToWrite(c) != C_OK) {
/* The caller expects the sds to be free'd. */
sdsfree(s);
return;
}
_addReplyToBufferOrList(c,s,sdslen(s));
sdsfree(s);
}
/* This low level function just adds whatever protocol you send it to the
* client buffer, trying the static buffer initially, and using the string
* of objects if not possible.
*
* It is efficient because does not create an SDS object nor an Redis object
* if not needed. The object will only be created by calling
* _addReplyProtoToList() if we fail to extend the existing tail object
* in the list of objects. */
void addReplyProto(client *c, const char *s, size_t len) {
if (prepareClientToWrite(c) != C_OK) return;
_addReplyToBufferOrList(c,s,len);
}
/* Low level function called by the addReplyError...() functions.
* It emits the protocol for a Redis error, in the form:
*
* -ERRORCODE Error Message<CR><LF>
*
* If the error code is already passed in the string 's', the error
* code provided is used, otherwise the string "-ERR " for the generic
* error code is automatically added.
* Note that 's' must NOT end with \r\n. */
void addReplyErrorLength(client *c, const char *s, size_t len) {
/* If the string already starts with "-..." then the error code
* is provided by the caller. Otherwise we use "-ERR". */
if (!len || s[0] != '-') addReplyProto(c,"-ERR ",5);
addReplyProto(c,s,len);
addReplyProto(c,"\r\n",2);
}
/* Do some actions after an error reply was sent (Log if needed, updates stats, etc.) */
void afterErrorReply(client *c, const char *s, size_t len) {
/* Increment the global error counter */
server.stat_total_error_replies++;
/* Increment the error stats
* If the string already starts with "-..." then the error prefix
* is provided by the caller ( we limit the search to 32 chars). Otherwise we use "-ERR". */
if (s[0] != '-') {
incrementErrorCount("ERR", 3);
} else {
char *spaceloc = memchr(s, ' ', len < 32 ? len : 32);
if (spaceloc) {
const size_t errEndPos = (size_t)(spaceloc - s);
incrementErrorCount(s+1, errEndPos-1);
} else {
/* Fallback to ERR if we can't retrieve the error prefix */
incrementErrorCount("ERR", 3);
}
}
/* Sometimes it could be normal that a slave replies to a master with
* an error and this function gets called. Actually the error will never
* be sent because addReply*() against master clients has no effect...
* A notable example is:
*
* EVAL 'redis.call("incr",KEYS[1]); redis.call("nonexisting")' 1 x
*
* Where the master must propagate the first change even if the second
* will produce an error. However it is useful to log such events since
* they are rare and may hint at errors in a script or a bug in Redis. */
int ctype = getClientType(c);
if (ctype == CLIENT_TYPE_MASTER || ctype == CLIENT_TYPE_SLAVE || c->id == CLIENT_ID_AOF) {
char *to, *from;
if (c->id == CLIENT_ID_AOF) {
to = "AOF-loading-client";
from = "server";
} else if (ctype == CLIENT_TYPE_MASTER) {
to = "master";
from = "replica";
} else {
to = "replica";
from = "master";
}
if (len > 4096) len = 4096;
const char *cmdname = c->lastcmd ? c->lastcmd->name : "<unknown>";
serverLog(LL_WARNING,"== CRITICAL == This %s is sending an error "
"to its %s: '%.*s' after processing the command "
"'%s'", from, to, (int)len, s, cmdname);
if (ctype == CLIENT_TYPE_MASTER && server.repl_backlog &&
server.repl_backlog->histlen > 0)
{
showLatestBacklog();
}
server.stat_unexpected_error_replies++;
}
}
/* The 'err' object is expected to start with -ERRORCODE and end with \r\n.
* Unlike addReplyErrorSds and others alike which rely on addReplyErrorLength. */
void addReplyErrorObject(client *c, robj *err) {
addReply(c, err);
afterErrorReply(c, err->ptr, sdslen(err->ptr)-2); /* Ignore trailing \r\n */
}
/* Sends either a reply or an error reply by checking the first char.
 * If the first char is '-' the reply is considered an error.
 * In any case the given reply is sent, if the reply is also recognize
* as an error we also perform some post reply operations such as
* logging and stats update. */
void addReplyOrErrorObject(client *c, robj *reply) {
serverAssert(sdsEncodedObject(reply));
sds rep = reply->ptr;
if (sdslen(rep) > 1 && rep[0] == '-') {
addReplyErrorObject(c, reply);
} else {
addReply(c, reply);
}
}
/* See addReplyErrorLength for expectations from the input string. */
void addReplyError(client *c, const char *err) {
addReplyErrorLength(c,err,strlen(err));
afterErrorReply(c,err,strlen(err));
}
/* See addReplyErrorLength for expectations from the input string. */
/* As a side effect the SDS string is freed. */
void addReplyErrorSds(client *c, sds err) {
addReplyErrorLength(c,err,sdslen(err));
afterErrorReply(c,err,sdslen(err));
sdsfree(err);
}
/* See addReplyErrorLength for expectations from the formatted string.
* The formatted string is safe to contain \r and \n anywhere. */
void addReplyErrorFormat(client *c, const char *fmt, ...) {
va_list ap;
va_start(ap,fmt);
sds s = sdscatvprintf(sdsempty(),fmt,ap);
va_end(ap);
/* Trim any newlines at the end (ones will be added by addReplyErrorLength) */
s = sdstrim(s, "\r\n");
/* Make sure there are no newlines in the middle of the string, otherwise
* invalid protocol is emitted. */
s = sdsmapchars(s, "\r\n", " ", 2);
addReplyErrorLength(c,s,sdslen(s));
afterErrorReply(c,s,sdslen(s));
sdsfree(s);
}
void addReplyStatusLength(client *c, const char *s, size_t len) {
addReplyProto(c,"+",1);
addReplyProto(c,s,len);
addReplyProto(c,"\r\n",2);
}
void addReplyStatus(client *c, const char *status) {
addReplyStatusLength(c,status,strlen(status));
}
void addReplyStatusFormat(client *c, const char *fmt, ...) {
va_list ap;
va_start(ap,fmt);
sds s = sdscatvprintf(sdsempty(),fmt,ap);
va_end(ap);
addReplyStatusLength(c,s,sdslen(s));
sdsfree(s);
}
/* Sometimes we are forced to create a new reply node, and we can't append to
* the previous one, when that happens, we wanna try to trim the unused space
* at the end of the last reply node which we won't use anymore. */
void trimReplyUnusedTailSpace(client *c) {
listNode *ln = listLast(c->reply);
clientReplyBlock *tail = ln? listNodeValue(ln): NULL;
/* Note that 'tail' may be NULL even if we have a tail node, because when
* addReplyDeferredLen() is used */
if (!tail) return;
/* We only try to trim the space is relatively high (more than a 1/4 of the
* allocation), otherwise there's a high chance realloc will NOP.
* Also, to avoid large memmove which happens as part of realloc, we only do
* that if the used part is small. */
if (tail->size - tail->used > tail->size / 4 &&
tail->used < PROTO_REPLY_CHUNK_BYTES)
{
size_t old_size = tail->size;
tail = zrealloc(tail, tail->used + sizeof(clientReplyBlock));
/* take over the allocation's internal fragmentation (at least for
* memory usage tracking) */
tail->size = zmalloc_usable_size(tail) - sizeof(clientReplyBlock);
c->reply_bytes = c->reply_bytes + tail->size - old_size;
listNodeValue(ln) = tail;
}
}
/* Adds an empty object to the reply list that will contain the multi bulk
* length, which is not known when this function is called. */
void *addReplyDeferredLen(client *c) {
/* Note that we install the write event here even if the object is not
* ready to be sent, since we are sure that before returning to the
* event loop setDeferredAggregateLen() will be called. */
if (prepareClientToWrite(c) != C_OK) return NULL;
/* Replicas should normally not cause any writes to the reply buffer. In case a rogue replica sent a command on the
* replication link that caused a reply to be generated we'll simply disconnect it.
* Note this is the simplest way to check a command added a response. Replication links are used to write data but
* not for responses, so we should normally never get here on a replica client. */
if (getClientType(c) == CLIENT_TYPE_SLAVE) {
sds cmdname = c->lastcmd ? getFullCommandName(c->lastcmd) : NULL;
logInvalidUseAndFreeClientAsync(c, "Replica generated a reply to command %s",
cmdname ? cmdname : "<unknown>");
sdsfree(cmdname);
return NULL;
}
trimReplyUnusedTailSpace(c);
listAddNodeTail(c->reply,NULL); /* NULL is our placeholder. */
return listLast(c->reply);
}
void setDeferredReply(client *c, void *node, const char *s, size_t length) {
listNode *ln = (listNode*)node;
clientReplyBlock *next, *prev;
/* Abort when *node is NULL: when the client should not accept writes
* we return NULL in addReplyDeferredLen() */
if (node == NULL) return;
serverAssert(!listNodeValue(ln));
/* Normally we fill this dummy NULL node, added by addReplyDeferredLen(),
* with a new buffer structure containing the protocol needed to specify
* the length of the array following. However sometimes there might be room
* in the previous/next node so we can instead remove this NULL node, and
* suffix/prefix our data in the node immediately before/after it, in order
* to save a write(2) syscall later. Conditions needed to do it:
*
* - The prev node is non-NULL and has space in it or
* - The next node is non-NULL,
* - It has enough room already allocated
* - And not too large (avoid large memmove) */
if (ln->prev != NULL && (prev = listNodeValue(ln->prev)) &&
prev->size - prev->used > 0)
{
size_t len_to_copy = prev->size - prev->used;
if (len_to_copy > length)
len_to_copy = length;
memcpy(prev->buf + prev->used, s, len_to_copy);
prev->used += len_to_copy;
length -= len_to_copy;
if (length == 0) {
listDelNode(c->reply, ln);
return;
}
s += len_to_copy;
}
if (ln->next != NULL && (next = listNodeValue(ln->next)) &&
next->size - next->used >= length &&
next->used < PROTO_REPLY_CHUNK_BYTES * 4)
{
memmove(next->buf + length, next->buf, next->used);
memcpy(next->buf, s, length);
next->used += length;
listDelNode(c->reply,ln);
} else {
/* Create a new node */
clientReplyBlock *buf = zmalloc(length + sizeof(clientReplyBlock));
/* Take over the allocation's internal fragmentation */
buf->size = zmalloc_usable_size(buf) - sizeof(clientReplyBlock);
buf->used = length;
memcpy(buf->buf, s, length);
listNodeValue(ln) = buf;
c->reply_bytes += buf->size;
closeClientOnOutputBufferLimitReached(c, 1);
}
}
/* Populate the length object and try gluing it to the next chunk. */
void setDeferredAggregateLen(client *c, void *node, long length, char prefix) {
serverAssert(length >= 0);
/* Abort when *node is NULL: when the client should not accept writes
* we return NULL in addReplyDeferredLen() */
if (node == NULL) return;
char lenstr[128];
size_t lenstr_len = sprintf(lenstr, "%c%ld\r\n", prefix, length);
setDeferredReply(c, node, lenstr, lenstr_len);
}
void setDeferredArrayLen(client *c, void *node, long length) {
setDeferredAggregateLen(c,node,length,'*');
}
void setDeferredMapLen(client *c, void *node, long length) {
int prefix = c->resp == 2 ? '*' : '%';
if (c->resp == 2) length *= 2;
setDeferredAggregateLen(c,node,length,prefix);
}
void setDeferredSetLen(client *c, void *node, long length) {
int prefix = c->resp == 2 ? '*' : '~';
setDeferredAggregateLen(c,node,length,prefix);
}
void setDeferredAttributeLen(client *c, void *node, long length) {
serverAssert(c->resp >= 3);
setDeferredAggregateLen(c,node,length,'|');
}
void setDeferredPushLen(client *c, void *node, long length) {
serverAssert(c->resp >= 3);
setDeferredAggregateLen(c,node,length,'>');
}
/* Add a double as a bulk reply */
void addReplyDouble(client *c, double d) {
if (isinf(d)) {
/* Libc in odd systems (Hi Solaris!) will format infinite in a
* different way, so better to handle it in an explicit way. */
if (c->resp == 2) {
addReplyBulkCString(c, d > 0 ? "inf" : "-inf");
} else {
addReplyProto(c, d > 0 ? ",inf\r\n" : ",-inf\r\n",
d > 0 ? 6 : 7);
}
} else {
char dbuf[MAX_LONG_DOUBLE_CHARS+3],
sbuf[MAX_LONG_DOUBLE_CHARS+32];
int dlen, slen;
if (c->resp == 2) {
dlen = snprintf(dbuf,sizeof(dbuf),"%.17g",d);
slen = snprintf(sbuf,sizeof(sbuf),"$%d\r\n%s\r\n",dlen,dbuf);
addReplyProto(c,sbuf,slen);
} else {
dlen = snprintf(dbuf,sizeof(dbuf),",%.17g\r\n",d);
addReplyProto(c,dbuf,dlen);
}
}
}
void addReplyBigNum(client *c, const char* num, size_t len) {
if (c->resp == 2) {
addReplyBulkCBuffer(c, num, len);
} else {
addReplyProto(c,"(",1);
addReplyProto(c,num,len);
addReply(c,shared.crlf);
}
}
/* Add a long double as a bulk reply, but uses a human readable formatting
* of the double instead of exposing the crude behavior of doubles to the
* dear user. */
void addReplyHumanLongDouble(client *c, long double d) {
if (c->resp == 2) {
robj *o = createStringObjectFromLongDouble(d,1);
addReplyBulk(c,o);
decrRefCount(o);
} else {
char buf[MAX_LONG_DOUBLE_CHARS];
int len = ld2string(buf,sizeof(buf),d,LD_STR_HUMAN);
addReplyProto(c,",",1);
addReplyProto(c,buf,len);
addReplyProto(c,"\r\n",2);
}
}
/* Add a long long as integer reply or bulk len / multi bulk count.
* Basically this is used to output <prefix><long long><crlf>. */
void addReplyLongLongWithPrefix(client *c, long long ll, char prefix) {
char buf[128];
int len;
/* Things like $3\r\n or *2\r\n are emitted very often by the protocol
* so we have a few shared objects to use if the integer is small
* like it is most of the times. */
if (prefix == '*' && ll < OBJ_SHARED_BULKHDR_LEN && ll >= 0) {
addReply(c,shared.mbulkhdr[ll]);
return;
} else if (prefix == '$' && ll < OBJ_SHARED_BULKHDR_LEN && ll >= 0) {
addReply(c,shared.bulkhdr[ll]);
return;
}
buf[0] = prefix;
len = ll2string(buf+1,sizeof(buf)-1,ll);
buf[len+1] = '\r';
buf[len+2] = '\n';
addReplyProto(c,buf,len+3);
}
void addReplyLongLong(client *c, long long ll) {
if (ll == 0)
addReply(c,shared.czero);
else if (ll == 1)
addReply(c,shared.cone);
else
addReplyLongLongWithPrefix(c,ll,':');
}
void addReplyAggregateLen(client *c, long length, int prefix) {
serverAssert(length >= 0);
addReplyLongLongWithPrefix(c,length,prefix);
}
void addReplyArrayLen(client *c, long length) {
addReplyAggregateLen(c,length,'*');
}
void addReplyMapLen(client *c, long length) {
int prefix = c->resp == 2 ? '*' : '%';
if (c->resp == 2) length *= 2;
addReplyAggregateLen(c,length,prefix);
}
void addReplySetLen(client *c, long length) {
int prefix = c->resp == 2 ? '*' : '~';
addReplyAggregateLen(c,length,prefix);
}
void addReplyAttributeLen(client *c, long length) {
serverAssert(c->resp >= 3);
addReplyAggregateLen(c,length,'|');
}
void addReplyPushLen(client *c, long length) {
serverAssert(c->resp >= 3);
addReplyAggregateLen(c,length,'>');
}
void addReplyNull(client *c) {
if (c->resp == 2) {
addReplyProto(c,"$-1\r\n",5);
} else {
addReplyProto(c,"_\r\n",3);
}
}
void addReplyBool(client *c, int b) {
if (c->resp == 2) {
addReply(c, b ? shared.cone : shared.czero);
} else {
addReplyProto(c, b ? "#t\r\n" : "#f\r\n",4);
}
}
/* A null array is a concept that no longer exists in RESP3. However
* RESP2 had it, so API-wise we have this call, that will emit the correct
* RESP2 protocol, however for RESP3 the reply will always be just the
* Null type "_\r\n". */
void addReplyNullArray(client *c) {
if (c->resp == 2) {
addReplyProto(c,"*-1\r\n",5);
} else {
addReplyProto(c,"_\r\n",3);
}
}
/* Create the length prefix of a bulk reply, example: $2234 */
void addReplyBulkLen(client *c, robj *obj) {
size_t len = stringObjectLen(obj);
addReplyLongLongWithPrefix(c,len,'$');
}
/* Add a Redis Object as a bulk reply */
void addReplyBulk(client *c, robj *obj) {
addReplyBulkLen(c,obj);
addReply(c,obj);
addReply(c,shared.crlf);
}
/* Add a C buffer as bulk reply */
void addReplyBulkCBuffer(client *c, const void *p, size_t len) {
addReplyLongLongWithPrefix(c,len,'$');
addReplyProto(c,p,len);
addReply(c,shared.crlf);
}
/* Add sds to reply (takes ownership of sds and frees it) */
void addReplyBulkSds(client *c, sds s) {
addReplyLongLongWithPrefix(c,sdslen(s),'$');
addReplySds(c,s);
addReply(c,shared.crlf);
}
/* Set sds to a deferred reply (for symmetry with addReplyBulkSds it also frees the sds) */
void setDeferredReplyBulkSds(client *c, void *node, sds s) {
sds reply = sdscatprintf(sdsempty(), "$%d\r\n%s\r\n", (unsigned)sdslen(s), s);
setDeferredReply(c, node, reply, sdslen(reply));
sdsfree(reply);
sdsfree(s);
}
/* Add a C null term string as bulk reply */
void addReplyBulkCString(client *c, const char *s) {
if (s == NULL) {
addReplyNull(c);
} else {
addReplyBulkCBuffer(c,s,strlen(s));
}
}
/* Add a long long as a bulk reply */
void addReplyBulkLongLong(client *c, long long ll) {
char buf[64];
int len;
len = ll2string(buf,64,ll);
addReplyBulkCBuffer(c,buf,len);
}
/* Reply with a verbatim type having the specified extension.
*
* The 'ext' is the "extension" of the file, actually just a three
* character type that describes the format of the verbatim string.
* For instance "txt" means it should be interpreted as a text only
* file by the receiver, "md " as markdown, and so forth. Only the
* three first characters of the extension are used, and if the
* provided one is shorter than that, the remaining is filled with
* spaces. */
void addReplyVerbatim(client *c, const char *s, size_t len, const char *ext) {
if (c->resp == 2) {
addReplyBulkCBuffer(c,s,len);
} else {
char buf[32];
size_t preflen = snprintf(buf,sizeof(buf),"=%zu\r\nxxx:",len+4);
char *p = buf+preflen-4;
for (int i = 0; i < 3; i++) {
if (*ext == '\0') {
p[i] = ' ';
} else {
p[i] = *ext++;
}
}
addReplyProto(c,buf,preflen);
addReplyProto(c,s,len);
addReplyProto(c,"\r\n",2);
}
}
/* Add an array of C strings as status replies with a heading.
* This function is typically invoked by from commands that support
* subcommands in response to the 'help' subcommand. The help array
* is terminated by NULL sentinel. */
void addReplyHelp(client *c, const char **help) {
sds cmd = sdsnew((char*) c->argv[0]->ptr);
void *blenp = addReplyDeferredLen(c);
int blen = 0;
sdstoupper(cmd);
addReplyStatusFormat(c,
"%s <subcommand> [<arg> [value] [opt] ...]. Subcommands are:",cmd);
sdsfree(cmd);
while (help[blen]) addReplyStatus(c,help[blen++]);
addReplyStatus(c,"HELP");
addReplyStatus(c," Prints this help.");
blen += 1; /* Account for the header. */
blen += 2; /* Account for the footer. */
setDeferredArrayLen(c,blenp,blen);
}
/* Add a suggestive error reply.
* This function is typically invoked by from commands that support
* subcommands in response to an unknown subcommand or argument error. */
void addReplySubcommandSyntaxError(client *c) {
sds cmd = sdsnew((char*) c->argv[0]->ptr);
sdstoupper(cmd);
addReplyErrorFormat(c,
"Unknown subcommand or wrong number of arguments for '%.128s'. Try %s HELP.",
(char*)c->argv[1]->ptr,cmd);
sdsfree(cmd);
}
/* Append 'src' client output buffers into 'dst' client output buffers.
* This function clears the output buffers of 'src' */
void AddReplyFromClient(client *dst, client *src) {
/* If the source client contains a partial response due to client output
* buffer limits, propagate that to the dest rather than copy a partial
* reply. We don't wanna run the risk of copying partial response in case
* for some reason the output limits don't reach the same decision (maybe
* they changed) */
if (src->flags & CLIENT_CLOSE_ASAP) {
sds client = catClientInfoString(sdsempty(),dst);
freeClientAsync(dst);
serverLog(LL_WARNING,"Client %s scheduled to be closed ASAP for overcoming of output buffer limits.", client);
sdsfree(client);
return;
}
/* First add the static buffer (either into the static buffer or reply list) */
addReplyProto(dst,src->buf, src->bufpos);
/* We need to check with prepareClientToWrite again (after addReplyProto)
* since addReplyProto may have changed something (like CLIENT_CLOSE_ASAP) */
if (prepareClientToWrite(dst) != C_OK)
return;
/* We're bypassing _addReplyProtoToList, so we need to add the pre/post
* checks in it. */
if (dst->flags & CLIENT_CLOSE_AFTER_REPLY) return;
/* Concatenate the reply list into the dest */
if (listLength(src->reply))
listJoin(dst->reply,src->reply);
dst->reply_bytes += src->reply_bytes;
src->reply_bytes = 0;
src->bufpos = 0;
/* Check output buffer limits */
closeClientOnOutputBufferLimitReached(dst, 1);
}
/* Logically copy 'src' replica client buffers info to 'dst' replica.
* Basically increase referenced buffer block node reference count. */
void copyReplicaOutputBuffer(client *dst, client *src) {
serverAssert(src->bufpos == 0 && listLength(src->reply) == 0);
if (src->ref_repl_buf_node == NULL) return;
dst->ref_repl_buf_node = src->ref_repl_buf_node;
dst->ref_block_pos = src->ref_block_pos;
((replBufBlock *)listNodeValue(dst->ref_repl_buf_node))->refcount++;
}
/* Return true if the specified client has pending reply buffers to write to
* the socket. */
int clientHasPendingReplies(client *c) {
if (getClientType(c) == CLIENT_TYPE_SLAVE) {
/* Replicas use global shared replication buffer instead of
* private output buffer. */
serverAssert(c->bufpos == 0 && listLength(c->reply) == 0);
if (c->ref_repl_buf_node == NULL) return 0;
/* If the last replication buffer block content is totally sent,
* we have nothing to send. */
listNode *ln = listLast(server.repl_buffer_blocks);
replBufBlock *tail = listNodeValue(ln);
if (ln == c->ref_repl_buf_node &&
c->ref_block_pos == tail->used) return 0;
return 1;
} else {
return c->bufpos || listLength(c->reply);
}
}
/* Return true if client connected from loopback interface */
int islocalClient(client *c) {
/* unix-socket */
if (c->flags & CLIENT_UNIX_SOCKET) return 1;
/* tcp */
char cip[NET_IP_STR_LEN+1] = { 0 };
connPeerToString(c->conn, cip, sizeof(cip)-1, NULL);
return !strcmp(cip,"127.0.0.1") || !strcmp(cip,"::1");
}
void clientAcceptHandler(connection *conn) {
client *c = connGetPrivateData(conn);
if (connGetState(conn) != CONN_STATE_CONNECTED) {
serverLog(LL_WARNING,
"Error accepting a client connection: %s",
connGetLastError(conn));
freeClientAsync(c);
return;
}
/* If the server is running in protected mode (the default) and there
* is no password set, nor a specific interface is bound, we don't accept
* requests from non loopback interfaces. Instead we try to explain the
* user what to do to fix it if needed. */
if (server.protected_mode &&
DefaultUser->flags & USER_FLAG_NOPASS)
{
if (!islocalClient(c)) {
char *err =
"-DENIED Redis is running in protected mode because protected "
"mode is enabled and no password is set for the default user. "
"In this mode connections are only accepted from the loopback interface. "
"If you want to connect from external computers to Redis you "
"may adopt one of the following solutions: "
"1) Just disable protected mode sending the command "
"'CONFIG SET protected-mode no' from the loopback interface "
"by connecting to Redis from the same host the server is "
"running, however MAKE SURE Redis is not publicly accessible "
"from internet if you do so. Use CONFIG REWRITE to make this "
"change permanent. "
"2) Alternatively you can just disable the protected mode by "
"editing the Redis configuration file, and setting the protected "
"mode option to 'no', and then restarting the server. "
"3) If you started the server manually just for testing, restart "
"it with the '--protected-mode no' option. "
"4) Setup a an authentication password for the default user. "
"NOTE: You only need to do one of the above things in order for "
"the server to start accepting connections from the outside.\r\n";
if (connWrite(c->conn,err,strlen(err)) == -1) {
/* Nothing to do, Just to avoid the warning... */
}
server.stat_rejected_conn++;
freeClientAsync(c);
return;
}
}
server.stat_numconnections++;
moduleFireServerEvent(REDISMODULE_EVENT_CLIENT_CHANGE,
REDISMODULE_SUBEVENT_CLIENT_CHANGE_CONNECTED,
c);
}
#define MAX_ACCEPTS_PER_CALL 1000
static void acceptCommonHandler(connection *conn, int flags, char *ip) {
client *c;
char conninfo[100];
UNUSED(ip);
if (connGetState(conn) != CONN_STATE_ACCEPTING) {
serverLog(LL_VERBOSE,
"Accepted client connection in error state: %s (conn: %s)",
connGetLastError(conn),
connGetInfo(conn, conninfo, sizeof(conninfo)));
connClose(conn);
return;
}
/* Limit the number of connections we take at the same time.
*
* Admission control will happen before a client is created and connAccept()
* called, because we don't want to even start transport-level negotiation
* if rejected. */
if (listLength(server.clients) + getClusterConnectionsCount()
>= server.maxclients)
{
char *err;
if (server.cluster_enabled)
err = "-ERR max number of clients + cluster "
"connections reached\r\n";
else
err = "-ERR max number of clients reached\r\n";
/* That's a best effort error message, don't check write errors.
* Note that for TLS connections, no handshake was done yet so nothing
* is written and the connection will just drop. */
if (connWrite(conn,err,strlen(err)) == -1) {
/* Nothing to do, Just to avoid the warning... */
}
server.stat_rejected_conn++;
connClose(conn);
return;
}
/* Create connection and client */
if ((c = createClient(conn)) == NULL) {
serverLog(LL_WARNING,
"Error registering fd event for the new client: %s (conn: %s)",
connGetLastError(conn),
connGetInfo(conn, conninfo, sizeof(conninfo)));
connClose(conn); /* May be already closed, just ignore errors */
return;
}
/* Last chance to keep flags */
c->flags |= flags;
/* Initiate accept.
*
* Note that connAccept() is free to do two things here:
* 1. Call clientAcceptHandler() immediately;
* 2. Schedule a future call to clientAcceptHandler().
*
* Because of that, we must do nothing else afterwards.
*/
if (connAccept(conn, clientAcceptHandler) == C_ERR) {
char conninfo[100];
if (connGetState(conn) == CONN_STATE_ERROR)
serverLog(LL_WARNING,
"Error accepting a client connection: %s (conn: %s)",
connGetLastError(conn), connGetInfo(conn, conninfo, sizeof(conninfo)));
freeClient(connGetPrivateData(conn));
return;
}
}
void acceptTcpHandler(aeEventLoop *el, int fd, void *privdata, int mask) {
int cport, cfd, max = MAX_ACCEPTS_PER_CALL;
char cip[NET_IP_STR_LEN];
UNUSED(el);
UNUSED(mask);
UNUSED(privdata);
while(max--) {
cfd = anetTcpAccept(server.neterr, fd, cip, sizeof(cip), &cport);
if (cfd == ANET_ERR) {
if (errno != EWOULDBLOCK)
serverLog(LL_WARNING,
"Accepting client connection: %s", server.neterr);
return;
}
serverLog(LL_VERBOSE,"Accepted %s:%d", cip, cport);
acceptCommonHandler(connCreateAcceptedSocket(cfd),0,cip);
}
}
void acceptTLSHandler(aeEventLoop *el, int fd, void *privdata, int mask) {
int cport, cfd, max = MAX_ACCEPTS_PER_CALL;
char cip[NET_IP_STR_LEN];
UNUSED(el);
UNUSED(mask);
UNUSED(privdata);
while(max--) {
cfd = anetTcpAccept(server.neterr, fd, cip, sizeof(cip), &cport);
if (cfd == ANET_ERR) {
if (errno != EWOULDBLOCK)
serverLog(LL_WARNING,
"Accepting client connection: %s", server.neterr);
return;
}
serverLog(LL_VERBOSE,"Accepted %s:%d", cip, cport);
acceptCommonHandler(connCreateAcceptedTLS(cfd, server.tls_auth_clients),0,cip);
}
}
void acceptUnixHandler(aeEventLoop *el, int fd, void *privdata, int mask) {
int cfd, max = MAX_ACCEPTS_PER_CALL;
UNUSED(el);
UNUSED(mask);
UNUSED(privdata);
while(max--) {
cfd = anetUnixAccept(server.neterr, fd);
if (cfd == ANET_ERR) {
if (errno != EWOULDBLOCK)
serverLog(LL_WARNING,
"Accepting client connection: %s", server.neterr);
return;
}
serverLog(LL_VERBOSE,"Accepted connection to %s", server.unixsocket);
acceptCommonHandler(connCreateAcceptedSocket(cfd),CLIENT_UNIX_SOCKET,NULL);
}
}
void freeClientOriginalArgv(client *c) {
/* We didn't rewrite this client */
if (!c->original_argv) return;
for (int j = 0; j < c->original_argc; j++)
decrRefCount(c->original_argv[j]);
zfree(c->original_argv);
c->original_argv = NULL;
c->original_argc = 0;
}
void freeClientArgv(client *c) {
int j;
for (j = 0; j < c->argc; j++)
decrRefCount(c->argv[j]);
c->argc = 0;
c->cmd = NULL;
c->argv_len_sum = 0;
c->argv_len = 0;
zfree(c->argv);
c->argv = NULL;
}
/* Close all the slaves connections. This is useful in chained replication
* when we resync with our own master and want to force all our slaves to
* resync with us as well. */
void disconnectSlaves(void) {
listIter li;
listNode *ln;
listRewind(server.slaves,&li);
while((ln = listNext(&li))) {
freeClient((client*)ln->value);
}
}
/* Check if there is any other slave waiting dumping RDB finished expect me.
* This function is useful to judge current dumping RDB can be used for full
* synchronization or not. */
int anyOtherSlaveWaitRdb(client *except_me) {
listIter li;
listNode *ln;
listRewind(server.slaves, &li);
while((ln = listNext(&li))) {
client *slave = ln->value;
if (slave != except_me &&
slave->replstate == SLAVE_STATE_WAIT_BGSAVE_END)
{
return 1;
}
}
return 0;
}
/* Remove the specified client from global lists where the client could
* be referenced, not including the Pub/Sub channels.
* This is used by freeClient() and replicationCacheMaster(). */
void unlinkClient(client *c) {
listNode *ln;
/* If this is marked as current client unset it. */
if (server.current_client == c) server.current_client = NULL;
/* Certain operations must be done only if the client has an active connection.
* If the client was already unlinked or if it's a "fake client" the
* conn is already set to NULL. */
if (c->conn) {
/* Remove from the list of active clients. */
if (c->client_list_node) {
uint64_t id = htonu64(c->id);
raxRemove(server.clients_index,(unsigned char*)&id,sizeof(id),NULL);
listDelNode(server.clients,c->client_list_node);
c->client_list_node = NULL;
}
/* Check if this is a replica waiting for diskless replication (rdb pipe),
* in which case it needs to be cleaned from that list */
if (c->flags & CLIENT_SLAVE &&
c->replstate == SLAVE_STATE_WAIT_BGSAVE_END &&
server.rdb_pipe_conns)
{
int i;
for (i=0; i < server.rdb_pipe_numconns; i++) {
if (server.rdb_pipe_conns[i] == c->conn) {
rdbPipeWriteHandlerConnRemoved(c->conn);
server.rdb_pipe_conns[i] = NULL;
break;
}
}
}
connClose(c->conn);
c->conn = NULL;
}
/* Remove from the list of pending writes if needed. */
if (c->flags & CLIENT_PENDING_WRITE) {
ln = listSearchKey(server.clients_pending_write,c);
serverAssert(ln != NULL);
listDelNode(server.clients_pending_write,ln);
c->flags &= ~CLIENT_PENDING_WRITE;
}
/* Remove from the list of pending reads if needed. */
serverAssert(io_threads_op == IO_THREADS_OP_IDLE);
if (c->pending_read_list_node != NULL) {
listDelNode(server.clients_pending_read,c->pending_read_list_node);
c->pending_read_list_node = NULL;
}
/* When client was just unblocked because of a blocking operation,
* remove it from the list of unblocked clients. */
if (c->flags & CLIENT_UNBLOCKED) {
ln = listSearchKey(server.unblocked_clients,c);
serverAssert(ln != NULL);
listDelNode(server.unblocked_clients,ln);
c->flags &= ~CLIENT_UNBLOCKED;
}
/* Clear the tracking status. */
if (c->flags & CLIENT_TRACKING) disableTracking(c);
}
/* Clear the client state to resemble a newly connected client. */
void clearClientConnectionState(client *c) {
listNode *ln;
/* MONITOR clients are also marked with CLIENT_SLAVE, we need to
* distinguish between the two.
*/
if (c->flags & CLIENT_MONITOR) {
ln = listSearchKey(server.monitors,c);
serverAssert(ln != NULL);
listDelNode(server.monitors,ln);
c->flags &= ~(CLIENT_MONITOR|CLIENT_SLAVE);
}
serverAssert(!(c->flags &(CLIENT_SLAVE|CLIENT_MASTER)));
if (c->flags & CLIENT_TRACKING) disableTracking(c);
selectDb(c,0);
c->resp = 2;
clientSetDefaultAuth(c);
moduleNotifyUserChanged(c);
discardTransaction(c);
pubsubUnsubscribeAllChannels(c,0);
pubsubUnsubscribeShardAllChannels(c, 0);
pubsubUnsubscribeAllPatterns(c,0);
if (c->name) {
decrRefCount(c->name);
c->name = NULL;
}
/* Selectively clear state flags not covered above */
c->flags &= ~(CLIENT_ASKING|CLIENT_READONLY|CLIENT_PUBSUB|
CLIENT_REPLY_OFF|CLIENT_REPLY_SKIP_NEXT);
}
void freeClient(client *c) {
listNode *ln;
/* If a client is protected, yet we need to free it right now, make sure
* to at least use asynchronous freeing. */
if (c->flags & CLIENT_PROTECTED) {
freeClientAsync(c);
return;
}
/* For connected clients, call the disconnection event of modules hooks. */
if (c->conn) {
moduleFireServerEvent(REDISMODULE_EVENT_CLIENT_CHANGE,
REDISMODULE_SUBEVENT_CLIENT_CHANGE_DISCONNECTED,
c);
}
/* Notify module system that this client auth status changed. */
moduleNotifyUserChanged(c);
/* If this client was scheduled for async freeing we need to remove it
* from the queue. Note that we need to do this here, because later
* we may call replicationCacheMaster() and the client should already
* be removed from the list of clients to free. */
if (c->flags & CLIENT_CLOSE_ASAP) {
ln = listSearchKey(server.clients_to_close,c);
serverAssert(ln != NULL);
listDelNode(server.clients_to_close,ln);
}
/* If it is our master that's being disconnected we should make sure
* to cache the state to try a partial resynchronization later.
*
* Note that before doing this we make sure that the client is not in
* some unexpected state, by checking its flags. */
if (server.master && c->flags & CLIENT_MASTER) {
serverLog(LL_WARNING,"Connection with master lost.");
if (!(c->flags & (CLIENT_PROTOCOL_ERROR|CLIENT_BLOCKED))) {
c->flags &= ~(CLIENT_CLOSE_ASAP|CLIENT_CLOSE_AFTER_REPLY);
replicationCacheMaster(c);
return;
}
}
/* Log link disconnection with slave */
if (getClientType(c) == CLIENT_TYPE_SLAVE) {
serverLog(LL_WARNING,"Connection with replica %s lost.",
replicationGetSlaveName(c));
}
/* Free the query buffer */
sdsfree(c->querybuf);
sdsfree(c->pending_querybuf);
c->querybuf = NULL;
/* Deallocate structures used to block on blocking ops. */
if (c->flags & CLIENT_BLOCKED) unblockClient(c);
dictRelease(c->bpop.keys);
/* UNWATCH all the keys */
unwatchAllKeys(c);
listRelease(c->watched_keys);
/* Unsubscribe from all the pubsub channels */
pubsubUnsubscribeAllChannels(c,0);
pubsubUnsubscribeShardAllChannels(c, 0);
pubsubUnsubscribeAllPatterns(c,0);
dictRelease(c->pubsub_channels);
listRelease(c->pubsub_patterns);
dictRelease(c->pubsubshard_channels);
/* Free data structures. */
listRelease(c->reply);
freeReplicaReferencedReplBuffer(c);
freeClientArgv(c);
freeClientOriginalArgv(c);
/* Unlink the client: this will close the socket, remove the I/O
* handlers, and remove references of the client from different
* places where active clients may be referenced. */
unlinkClient(c);
/* Master/slave cleanup Case 1:
* we lost the connection with a slave. */
if (c->flags & CLIENT_SLAVE) {
/* If there is no any other slave waiting dumping RDB finished, the
* current child process need not continue to dump RDB, then we kill it.
* So child process won't use more memory, and we also can fork a new
* child process asap to dump rdb for next full synchronization or bgsave.
* But we also need to check if users enable 'save' RDB, if enable, we
* should not remove directly since that means RDB is important for users
* to keep data safe and we may delay configured 'save' for full sync. */
if (server.saveparamslen == 0 &&
c->replstate == SLAVE_STATE_WAIT_BGSAVE_END &&
server.child_type == CHILD_TYPE_RDB &&
server.rdb_child_type == RDB_CHILD_TYPE_DISK &&
anyOtherSlaveWaitRdb(c) == 0)
{
killRDBChild();
}
if (c->replstate == SLAVE_STATE_SEND_BULK) {
if (c->repldbfd != -1) close(c->repldbfd);
if (c->replpreamble) sdsfree(c->replpreamble);
}
list *l = (c->flags & CLIENT_MONITOR) ? server.monitors : server.slaves;
ln = listSearchKey(l,c);
serverAssert(ln != NULL);
listDelNode(l,ln);
/* We need to remember the time when we started to have zero
* attached slaves, as after some time we'll free the replication
* backlog. */
if (getClientType(c) == CLIENT_TYPE_SLAVE && listLength(server.slaves) == 0)
server.repl_no_slaves_since = server.unixtime;
refreshGoodSlavesCount();
/* Fire the replica change modules event. */
if (c->replstate == SLAVE_STATE_ONLINE)
moduleFireServerEvent(REDISMODULE_EVENT_REPLICA_CHANGE,
REDISMODULE_SUBEVENT_REPLICA_CHANGE_OFFLINE,
NULL);
}
/* Master/slave cleanup Case 2:
* we lost the connection with the master. */
if (c->flags & CLIENT_MASTER) replicationHandleMasterDisconnection();
/* Remove the contribution that this client gave to our
* incrementally computed memory usage. */
server.stat_clients_type_memory[c->last_memory_type] -=
c->last_memory_usage;
/* Remove client from memory usage buckets */
if (c->mem_usage_bucket) {
c->mem_usage_bucket->mem_usage_sum -= c->last_memory_usage;
listDelNode(c->mem_usage_bucket->clients, c->mem_usage_bucket_node);
}
/* Release other dynamically allocated client structure fields,
* and finally release the client structure itself. */
if (c->name) decrRefCount(c->name);
freeClientMultiState(c);
sdsfree(c->peerid);
sdsfree(c->sockname);
sdsfree(c->slave_addr);
zfree(c);
}
/* Schedule a client to free it at a safe time in the serverCron() function.
* This function is useful when we need to terminate a client but we are in
* a context where calling freeClient() is not possible, because the client
* should be valid for the continuation of the flow of the program. */
void freeClientAsync(client *c) {
/* We need to handle concurrent access to the server.clients_to_close list
* only in the freeClientAsync() function, since it's the only function that
* may access the list while Redis uses I/O threads. All the other accesses
* are in the context of the main thread while the other threads are
* idle. */
if (c->flags & CLIENT_CLOSE_ASAP || c->flags & CLIENT_SCRIPT) return;
c->flags |= CLIENT_CLOSE_ASAP;
if (server.io_threads_num == 1) {
/* no need to bother with locking if there's just one thread (the main thread) */
listAddNodeTail(server.clients_to_close,c);
return;
}
static pthread_mutex_t async_free_queue_mutex = PTHREAD_MUTEX_INITIALIZER;
pthread_mutex_lock(&async_free_queue_mutex);
listAddNodeTail(server.clients_to_close,c);
pthread_mutex_unlock(&async_free_queue_mutex);
}
/* Log errors for invalid use and free the client in async way.
* We will add additional information about the client to the message. */
void logInvalidUseAndFreeClientAsync(client *c, const char *fmt, ...) {
va_list ap;
va_start(ap, fmt);
sds info = sdscatvprintf(sdsempty(), fmt, ap);
va_end(ap);
sds client = catClientInfoString(sdsempty(), c);
serverLog(LL_WARNING, "%s, disconnecting it: %s", info, client);
sdsfree(info);
sdsfree(client);
freeClientAsync(c);
}
/* Perform processing of the client before moving on to processing the next client
* this is useful for performing operations that affect the global state but can't
* wait until we're done with all clients. In other words can't wait until beforeSleep()
* return C_ERR in case client is no longer valid after call.
* The input client argument: c, may be NULL in case the previous client was
* freed before the call. */
int beforeNextClient(client *c) {
/* Skip the client processing if we're in an IO thread, in that case we'll perform
this operation later (this function is called again) in the fan-in stage of the threading mechanism */
if (io_threads_op != IO_THREADS_OP_IDLE)
return C_OK;
/* Handle async frees */
/* Note: this doesn't make the server.clients_to_close list redundant because of
* cases where we want an async free of a client other than myself. For example
* in ACL modifications we disconnect clients authenticated to non-existent
* users (see ACL LOAD). */
if (c && (c->flags & CLIENT_CLOSE_ASAP)) {
freeClient(c);
return C_ERR;
}
return C_OK;
}
/* Free the clients marked as CLOSE_ASAP, return the number of clients
* freed. */
int freeClientsInAsyncFreeQueue(void) {
int freed = 0;
listIter li;
listNode *ln;
listRewind(server.clients_to_close,&li);
while ((ln = listNext(&li)) != NULL) {
client *c = listNodeValue(ln);
if (c->flags & CLIENT_PROTECTED) continue;
c->flags &= ~CLIENT_CLOSE_ASAP;
freeClient(c);
listDelNode(server.clients_to_close,ln);
freed++;
}
return freed;
}
/* Return a client by ID, or NULL if the client ID is not in the set
* of registered clients. Note that "fake clients", created with -1 as FD,
* are not registered clients. */
client *lookupClientByID(uint64_t id) {
id = htonu64(id);
client *c = raxFind(server.clients_index,(unsigned char*)&id,sizeof(id));
return (c == raxNotFound) ? NULL : c;
}
/* This function does actual writing output buffers to different types of
* clients, it is called by writeToClient.
* If we write successfully, it return C_OK, otherwise, C_ERR is returned,
* And 'nwritten' is a output parameter, it means how many bytes server write
* to client. */
int _writeToClient(client *c, ssize_t *nwritten) {
*nwritten = 0;
if (getClientType(c) == CLIENT_TYPE_SLAVE) {
serverAssert(c->bufpos == 0 && listLength(c->reply) == 0);
replBufBlock *o = listNodeValue(c->ref_repl_buf_node);
serverAssert(o->used >= c->ref_block_pos);
/* Send current block if it is not fully sent. */
if (o->used > c->ref_block_pos) {
*nwritten = connWrite(c->conn, o->buf+c->ref_block_pos,
o->used-c->ref_block_pos);
if (*nwritten <= 0) return C_ERR;
c->ref_block_pos += *nwritten;
}
/* If we fully sent the object on head, go to the next one. */
listNode *next = listNextNode(c->ref_repl_buf_node);
if (next && c->ref_block_pos == o->used) {
o->refcount--;
((replBufBlock *)(listNodeValue(next)))->refcount++;
c->ref_repl_buf_node = next;
c->ref_block_pos = 0;
incrementalTrimReplicationBacklog(REPL_BACKLOG_TRIM_BLOCKS_PER_CALL);
}
return C_OK;
}
if (c->bufpos > 0) {
*nwritten = connWrite(c->conn,c->buf+c->sentlen,c->bufpos-c->sentlen);
if (*nwritten <= 0) return C_ERR;
c->sentlen += *nwritten;
/* If the buffer was sent, set bufpos to zero to continue with
* the remainder of the reply. */
if ((int)c->sentlen == c->bufpos) {
c->bufpos = 0;
c->sentlen = 0;
}
} else {
clientReplyBlock *o = listNodeValue(listFirst(c->reply));
size_t objlen = o->used;
if (objlen == 0) {
c->reply_bytes -= o->size;
listDelNode(c->reply,listFirst(c->reply));
return C_OK;
}
*nwritten = connWrite(c->conn, o->buf + c->sentlen, objlen - c->sentlen);
if (*nwritten <= 0) return C_ERR;
c->sentlen += *nwritten;
/* If we fully sent the object on head go to the next one */
if (c->sentlen == objlen) {
c->reply_bytes -= o->size;
listDelNode(c->reply,listFirst(c->reply));
c->sentlen = 0;
/* If there are no longer objects in the list, we expect
* the count of reply bytes to be exactly zero. */
if (listLength(c->reply) == 0)
serverAssert(c->reply_bytes == 0);
}
}
return C_OK;
}
/* Write data in output buffers to client. Return C_OK if the client
* is still valid after the call, C_ERR if it was freed because of some
* error. If handler_installed is set, it will attempt to clear the
* write event.
*
* This function is called by threads, but always with handler_installed
* set to 0. So when handler_installed is set to 0 the function must be
* thread safe. */
int writeToClient(client *c, int handler_installed) {
/* Update total number of writes on server */
atomicIncr(server.stat_total_writes_processed, 1);
ssize_t nwritten = 0, totwritten = 0;
while(clientHasPendingReplies(c)) {
int ret = _writeToClient(c, &nwritten);
if (ret == C_ERR) break;
totwritten += nwritten;
/* Note that we avoid to send more than NET_MAX_WRITES_PER_EVENT
* bytes, in a single threaded server it's a good idea to serve
* other clients as well, even if a very large request comes from
* super fast link that is always able to accept data (in real world
* scenario think about 'KEYS *' against the loopback interface).
*
* However if we are over the maxmemory limit we ignore that and
* just deliver as much data as it is possible to deliver.
*
* Moreover, we also send as much as possible if the client is
* a slave or a monitor (otherwise, on high-speed traffic, the
* replication/output buffer will grow indefinitely) */
if (totwritten > NET_MAX_WRITES_PER_EVENT &&
(server.maxmemory == 0 ||
zmalloc_used_memory() < server.maxmemory) &&
!(c->flags & CLIENT_SLAVE)) break;
}
atomicIncr(server.stat_net_output_bytes, totwritten);
if (nwritten == -1) {
if (connGetState(c->conn) != CONN_STATE_CONNECTED) {
serverLog(LL_VERBOSE,
"Error writing to client: %s", connGetLastError(c->conn));
freeClientAsync(c);
return C_ERR;
}
}
if (totwritten > 0) {
/* For clients representing masters we don't count sending data
* as an interaction, since we always send REPLCONF ACK commands
* that take some time to just fill the socket output buffer.
* We just rely on data / pings received for timeout detection. */
if (!(c->flags & CLIENT_MASTER)) c->lastinteraction = server.unixtime;
}
if (!clientHasPendingReplies(c)) {
c->sentlen = 0;
/* Note that writeToClient() is called in a threaded way, but
* adDeleteFileEvent() is not thread safe: however writeToClient()
* is always called with handler_installed set to 0 from threads
* so we are fine. */
if (handler_installed) {
serverAssert(io_threads_op == IO_THREADS_OP_IDLE);
connSetWriteHandler(c->conn, NULL);
}
/* Close connection after entire reply has been sent. */
if (c->flags & CLIENT_CLOSE_AFTER_REPLY) {
freeClientAsync(c);
return C_ERR;
}
}
updateClientMemUsage(c);
return C_OK;
}
/* Write event handler. Just send data to the client. */
void sendReplyToClient(connection *conn) {
client *c = connGetPrivateData(conn);
writeToClient(c,1);
}
/* This function is called just before entering the event loop, in the hope
* we can just write the replies to the client output buffer without any
* need to use a syscall in order to install the writable event handler,
* get it called, and so forth. */
int handleClientsWithPendingWrites(void) {
listIter li;
listNode *ln;
int processed = listLength(server.clients_pending_write);
listRewind(server.clients_pending_write,&li);
while((ln = listNext(&li))) {
client *c = listNodeValue(ln);
c->flags &= ~CLIENT_PENDING_WRITE;
listDelNode(server.clients_pending_write,ln);
/* If a client is protected, don't do anything,
* that may trigger write error or recreate handler. */
if (c->flags & CLIENT_PROTECTED) continue;
/* Don't write to clients that are going to be closed anyway. */
if (c->flags & CLIENT_CLOSE_ASAP) continue;
/* Try to write buffers to the client socket. */
if (writeToClient(c,0) == C_ERR) continue;
/* If after the synchronous writes above we still have data to
* output to the client, we need to install the writable handler. */
if (clientHasPendingReplies(c)) {
int ae_barrier = 0;
/* For the fsync=always policy, we want that a given FD is never
* served for reading and writing in the same event loop iteration,
* so that in the middle of receiving the query, and serving it
* to the client, we'll call beforeSleep() that will do the
* actual fsync of AOF to disk. the write barrier ensures that. */
if (server.aof_state == AOF_ON &&
server.aof_fsync == AOF_FSYNC_ALWAYS)
{
ae_barrier = 1;
}
if (connSetWriteHandlerWithBarrier(c->conn, sendReplyToClient, ae_barrier) == C_ERR) {
freeClientAsync(c);
}
}
}
return processed;
}
/* resetClient prepare the client to process the next command */
void resetClient(client *c) {
redisCommandProc *prevcmd = c->cmd ? c->cmd->proc : NULL;
freeClientArgv(c);
c->reqtype = 0;
c->multibulklen = 0;
c->bulklen = -1;
/* We clear the ASKING flag as well if we are not inside a MULTI, and
* if what we just executed is not the ASKING command itself. */
if (!(c->flags & CLIENT_MULTI) && prevcmd != askingCommand)
c->flags &= ~CLIENT_ASKING;
/* We do the same for the CACHING command as well. It also affects
* the next command or transaction executed, in a way very similar
* to ASKING. */
if (!(c->flags & CLIENT_MULTI) && prevcmd != clientCommand)
c->flags &= ~CLIENT_TRACKING_CACHING;
/* Remove the CLIENT_REPLY_SKIP flag if any so that the reply
* to the next command will be sent, but set the flag if the command
* we just processed was "CLIENT REPLY SKIP". */
c->flags &= ~CLIENT_REPLY_SKIP;
if (c->flags & CLIENT_REPLY_SKIP_NEXT) {
c->flags |= CLIENT_REPLY_SKIP;
c->flags &= ~CLIENT_REPLY_SKIP_NEXT;
}
}
/* This function is used when we want to re-enter the event loop but there
* is the risk that the client we are dealing with will be freed in some
* way. This happens for instance in:
*
* * DEBUG RELOAD and similar.
* * When a Lua script is in -BUSY state.
*
* So the function will protect the client by doing two things:
*
* 1) It removes the file events. This way it is not possible that an
* error is signaled on the socket, freeing the client.
* 2) Moreover it makes sure that if the client is freed in a different code
* path, it is not really released, but only marked for later release. */
void protectClient(client *c) {
c->flags |= CLIENT_PROTECTED;
if (c->conn) {
connSetReadHandler(c->conn,NULL);
connSetWriteHandler(c->conn,NULL);
}
}
/* This will undo the client protection done by protectClient() */
void unprotectClient(client *c) {
if (c->flags & CLIENT_PROTECTED) {
c->flags &= ~CLIENT_PROTECTED;
if (c->conn) {
connSetReadHandler(c->conn,readQueryFromClient);
if (clientHasPendingReplies(c)) clientInstallWriteHandler(c);
}
}
}
/* Like processMultibulkBuffer(), but for the inline protocol instead of RESP,
* this function consumes the client query buffer and creates a command ready
* to be executed inside the client structure. Returns C_OK if the command
* is ready to be executed, or C_ERR if there is still protocol to read to
* have a well formed command. The function also returns C_ERR when there is
* a protocol error: in such a case the client structure is setup to reply
* with the error and close the connection. */
int processInlineBuffer(client *c) {
char *newline;
int argc, j, linefeed_chars = 1;
sds *argv, aux;
size_t querylen;
/* Search for end of line */
newline = strchr(c->querybuf+c->qb_pos,'\n');
/* Nothing to do without a \r\n */
if (newline == NULL) {
if (sdslen(c->querybuf)-c->qb_pos > PROTO_INLINE_MAX_SIZE) {
addReplyError(c,"Protocol error: too big inline request");
setProtocolError("too big inline request",c);
}
return C_ERR;
}
/* Handle the \r\n case. */
if (newline != c->querybuf+c->qb_pos && *(newline-1) == '\r')
newline--, linefeed_chars++;
/* Split the input buffer up to the \r\n */
querylen = newline-(c->querybuf+c->qb_pos);
aux = sdsnewlen(c->querybuf+c->qb_pos,querylen);
argv = sdssplitargs(aux,&argc);
sdsfree(aux);
if (argv == NULL) {
addReplyError(c,"Protocol error: unbalanced quotes in request");
setProtocolError("unbalanced quotes in inline request",c);
return C_ERR;
}
/* Newline from slaves can be used to refresh the last ACK time.
* This is useful for a slave to ping back while loading a big
* RDB file. */
if (querylen == 0 && getClientType(c) == CLIENT_TYPE_SLAVE)
c->repl_ack_time = server.unixtime;
/* Masters should never send us inline protocol to run actual
* commands. If this happens, it is likely due to a bug in Redis where
* we got some desynchronization in the protocol, for example
* because of a PSYNC gone bad.
*
* However the is an exception: masters may send us just a newline
* to keep the connection active. */
if (querylen != 0 && c->flags & CLIENT_MASTER) {
sdsfreesplitres(argv,argc);
serverLog(LL_WARNING,"WARNING: Receiving inline protocol from master, master stream corruption? Closing the master connection and discarding the cached master.");
setProtocolError("Master using the inline protocol. Desync?",c);
return C_ERR;
}
/* Move querybuffer position to the next query in the buffer. */
c->qb_pos += querylen+linefeed_chars;
/* Setup argv array on client structure */
if (argc) {
if (c->argv) zfree(c->argv);
c->argv_len = argc;
c->argv = zmalloc(sizeof(robj*)*c->argv_len);
c->argv_len_sum = 0;
}
/* Create redis objects for all arguments. */
for (c->argc = 0, j = 0; j < argc; j++) {
c->argv[c->argc] = createObject(OBJ_STRING,argv[j]);
c->argc++;
c->argv_len_sum += sdslen(argv[j]);
}
zfree(argv);
return C_OK;
}
/* Helper function. Record protocol error details in server log,
* and set the client as CLIENT_CLOSE_AFTER_REPLY and
* CLIENT_PROTOCOL_ERROR. */
#define PROTO_DUMP_LEN 128
static void setProtocolError(const char *errstr, client *c) {
if (server.verbosity <= LL_VERBOSE || c->flags & CLIENT_MASTER) {
sds client = catClientInfoString(sdsempty(),c);
/* Sample some protocol to given an idea about what was inside. */
char buf[256];
if (sdslen(c->querybuf)-c->qb_pos < PROTO_DUMP_LEN) {
snprintf(buf,sizeof(buf),"Query buffer during protocol error: '%s'", c->querybuf+c->qb_pos);
} else {
snprintf(buf,sizeof(buf),"Query buffer during protocol error: '%.*s' (... more %zu bytes ...) '%.*s'", PROTO_DUMP_LEN/2, c->querybuf+c->qb_pos, sdslen(c->querybuf)-c->qb_pos-PROTO_DUMP_LEN, PROTO_DUMP_LEN/2, c->querybuf+sdslen(c->querybuf)-PROTO_DUMP_LEN/2);
}
/* Remove non printable chars. */
char *p = buf;
while (*p != '\0') {
if (!isprint(*p)) *p = '.';
p++;
}
/* Log all the client and protocol info. */
int loglevel = (c->flags & CLIENT_MASTER) ? LL_WARNING :
LL_VERBOSE;
serverLog(loglevel,
"Protocol error (%s) from client: %s. %s", errstr, client, buf);
sdsfree(client);
}
c->flags |= (CLIENT_CLOSE_AFTER_REPLY|CLIENT_PROTOCOL_ERROR);
}
/* Process the query buffer for client 'c', setting up the client argument
* vector for command execution. Returns C_OK if after running the function
* the client has a well-formed ready to be processed command, otherwise
* C_ERR if there is still to read more buffer to get the full command.
* The function also returns C_ERR when there is a protocol error: in such a
* case the client structure is setup to reply with the error and close
* the connection.
*
* This function is called if processInputBuffer() detects that the next
* command is in RESP format, so the first byte in the command is found
* to be '*'. Otherwise for inline commands processInlineBuffer() is called. */
int processMultibulkBuffer(client *c) {
char *newline = NULL;
int ok;
long long ll;
if (c->multibulklen == 0) {
/* The client should have been reset */
serverAssertWithInfo(c,NULL,c->argc == 0);
/* Multi bulk length cannot be read without a \r\n */
newline = strchr(c->querybuf+c->qb_pos,'\r');
if (newline == NULL) {
if (sdslen(c->querybuf)-c->qb_pos > PROTO_INLINE_MAX_SIZE) {
addReplyError(c,"Protocol error: too big mbulk count string");
setProtocolError("too big mbulk count string",c);
}
return C_ERR;
}
/* Buffer should also contain \n */
if (newline-(c->querybuf+c->qb_pos) > (ssize_t)(sdslen(c->querybuf)-c->qb_pos-2))
return C_ERR;
/* We know for sure there is a whole line since newline != NULL,
* so go ahead and find out the multi bulk length. */
serverAssertWithInfo(c,NULL,c->querybuf[c->qb_pos] == '*');
ok = string2ll(c->querybuf+1+c->qb_pos,newline-(c->querybuf+1+c->qb_pos),&ll);
if (!ok || ll > INT_MAX) {
addReplyError(c,"Protocol error: invalid multibulk length");
setProtocolError("invalid mbulk count",c);
return C_ERR;
} else if (ll > 10 && authRequired(c)) {
addReplyError(c, "Protocol error: unauthenticated multibulk length");
setProtocolError("unauth mbulk count", c);
return C_ERR;
}
c->qb_pos = (newline-c->querybuf)+2;
if (ll <= 0) return C_OK;
c->multibulklen = ll;
/* Setup argv array on client structure */
if (c->argv) zfree(c->argv);
c->argv_len = min(c->multibulklen, 1024);
c->argv = zmalloc(sizeof(robj*)*c->argv_len);
c->argv_len_sum = 0;
}
serverAssertWithInfo(c,NULL,c->multibulklen > 0);
while(c->multibulklen) {
/* Read bulk length if unknown */
if (c->bulklen == -1) {
newline = strchr(c->querybuf+c->qb_pos,'\r');
if (newline == NULL) {
if (sdslen(c->querybuf)-c->qb_pos > PROTO_INLINE_MAX_SIZE) {
addReplyError(c,
"Protocol error: too big bulk count string");
setProtocolError("too big bulk count string",c);
return C_ERR;
}
break;
}
/* Buffer should also contain \n */
if (newline-(c->querybuf+c->qb_pos) > (ssize_t)(sdslen(c->querybuf)-c->qb_pos-2))
break;
if (c->querybuf[c->qb_pos] != '$') {
addReplyErrorFormat(c,
"Protocol error: expected '$', got '%c'",
c->querybuf[c->qb_pos]);
setProtocolError("expected $ but got something else",c);
return C_ERR;
}
ok = string2ll(c->querybuf+c->qb_pos+1,newline-(c->querybuf+c->qb_pos+1),&ll);
if (!ok || ll < 0 ||
(!(c->flags & CLIENT_MASTER) && ll > server.proto_max_bulk_len)) {
addReplyError(c,"Protocol error: invalid bulk length");
setProtocolError("invalid bulk length",c);
return C_ERR;
} else if (ll > 16384 && authRequired(c)) {
addReplyError(c, "Protocol error: unauthenticated bulk length");
setProtocolError("unauth bulk length", c);
return C_ERR;
}
c->qb_pos = newline-c->querybuf+2;
if (ll >= PROTO_MBULK_BIG_ARG) {
/* If we are going to read a large object from network
* try to make it likely that it will start at c->querybuf
* boundary so that we can optimize object creation
* avoiding a large copy of data.
*
* But only when the data we have not parsed is less than
* or equal to ll+2. If the data length is greater than
* ll+2, trimming querybuf is just a waste of time, because
* at this time the querybuf contains not only our bulk. */
if (sdslen(c->querybuf)-c->qb_pos <= (size_t)ll+2) {
sdsrange(c->querybuf,c->qb_pos,-1);
c->qb_pos = 0;
/* Hint the sds library about the amount of bytes this string is
* going to contain. */
c->querybuf = sdsMakeRoomForNonGreedy(c->querybuf,ll+2-sdslen(c->querybuf));
}
}
c->bulklen = ll;
}
/* Read bulk argument */
if (sdslen(c->querybuf)-c->qb_pos < (size_t)(c->bulklen+2)) {
/* Not enough data (+2 == trailing \r\n) */
break;
} else {
/* Check if we have space in argv, grow if needed */
if (c->argc >= c->argv_len) {
c->argv_len = min(c->argv_len < INT_MAX/2 ? c->argv_len*2 : INT_MAX, c->argc+c->multibulklen);
c->argv = zrealloc(c->argv, sizeof(robj*)*c->argv_len);
}
/* Optimization: if the buffer contains JUST our bulk element
* instead of creating a new object by *copying* the sds we
* just use the current sds string. */
if (c->qb_pos == 0 &&
c->bulklen >= PROTO_MBULK_BIG_ARG &&
sdslen(c->querybuf) == (size_t)(c->bulklen+2))
{
c->argv[c->argc++] = createObject(OBJ_STRING,c->querybuf);
c->argv_len_sum += c->bulklen;
sdsIncrLen(c->querybuf,-2); /* remove CRLF */
/* Assume that if we saw a fat argument we'll see another one
* likely... */
c->querybuf = sdsnewlen(SDS_NOINIT,c->bulklen+2);
sdsclear(c->querybuf);
} else {
c->argv[c->argc++] =
createStringObject(c->querybuf+c->qb_pos,c->bulklen);
c->argv_len_sum += c->bulklen;
c->qb_pos += c->bulklen+2;
}
c->bulklen = -1;
c->multibulklen--;
}
}
/* We're done when c->multibulk == 0 */
if (c->multibulklen == 0) return C_OK;
/* Still not ready to process the command */
return C_ERR;
}
/* Perform necessary tasks after a command was executed:
*
* 1. The client is reset unless there are reasons to avoid doing it.
* 2. In the case of master clients, the replication offset is updated.
* 3. Propagate commands we got from our master to replicas down the line. */
void commandProcessed(client *c) {
/* If client is blocked(including paused), just return avoid reset and replicate.
*
* 1. Don't reset the client structure for blocked clients, so that the reply
* callback will still be able to access the client argv and argc fields.
* The client will be reset in unblockClient().
* 2. Don't update replication offset or propagate commands to replicas,
* since we have not applied the command. */
if (c->flags & CLIENT_BLOCKED) return;
resetClient(c);
long long prev_offset = c->reploff;
if (c->flags & CLIENT_MASTER && !(c->flags & CLIENT_MULTI)) {
/* Update the applied replication offset of our master. */
c->reploff = c->read_reploff - sdslen(c->querybuf) + c->qb_pos;
}
/* If the client is a master we need to compute the difference
* between the applied offset before and after processing the buffer,
* to understand how much of the replication stream was actually
* applied to the master state: this quantity, and its corresponding
* part of the replication stream, will be propagated to the
* sub-replicas and to the replication backlog. */
if (c->flags & CLIENT_MASTER) {
long long applied = c->reploff - prev_offset;
if (applied) {
replicationFeedStreamFromMasterStream(c->pending_querybuf,applied);
sdsrange(c->pending_querybuf,applied,-1);
}
}
}
/* This function calls processCommand(), but also performs a few sub tasks
* for the client that are useful in that context:
*
* 1. It sets the current client to the client 'c'.
* 2. calls commandProcessed() if the command was handled.
*
* The function returns C_ERR in case the client was freed as a side effect
* of processing the command, otherwise C_OK is returned. */
int processCommandAndResetClient(client *c) {
int deadclient = 0;
client *old_client = server.current_client;
server.current_client = c;
if (processCommand(c) == C_OK) {
commandProcessed(c);
/* Update the client's memory to include output buffer growth following the
* processed command. */
updateClientMemUsage(c);
}
if (server.current_client == NULL) deadclient = 1;
/*
* Restore the old client, this is needed because when a script
* times out, we will get into this code from processEventsWhileBlocked.
* Which will cause to set the server.current_client. If not restored
* we will return 1 to our caller which will falsely indicate the client
* is dead and will stop reading from its buffer.
*/
server.current_client = old_client;
/* performEvictions may flush slave output buffers. This may
* result in a slave, that may be the active client, to be
* freed. */
return deadclient ? C_ERR : C_OK;
}
/* This function will execute any fully parsed commands pending on
* the client. Returns C_ERR if the client is no longer valid after executing
* the command, and C_OK for all other cases. */
int processPendingCommandsAndResetClient(client *c) {
if (c->flags & CLIENT_PENDING_COMMAND) {
c->flags &= ~CLIENT_PENDING_COMMAND;
if (processCommandAndResetClient(c) == C_ERR) {
return C_ERR;
}
}
return C_OK;
}
/* This function is called every time, in the client structure 'c', there is
* more query buffer to process, because we read more data from the socket
* or because a client was blocked and later reactivated, so there could be
* pending query buffer, already representing a full command, to process.
* return C_ERR in case the client was freed during the processing */
int processInputBuffer(client *c) {
/* Keep processing while there is something in the input buffer */
while(c->qb_pos < sdslen(c->querybuf)) {
/* Immediately abort if the client is in the middle of something. */
if (c->flags & CLIENT_BLOCKED) break;
/* Don't process more buffers from clients that have already pending
* commands to execute in c->argv. */
if (c->flags & CLIENT_PENDING_COMMAND) break;
/* Don't process input from the master while there is a busy script
* condition on the slave. We want just to accumulate the replication
* stream (instead of replying -BUSY like we do with other clients) and
* later resume the processing. */
if (scriptIsTimedout() && c->flags & CLIENT_MASTER) break;
/* CLIENT_CLOSE_AFTER_REPLY closes the connection once the reply is
* written to the client. Make sure to not let the reply grow after
* this flag has been set (i.e. don't process more commands).
*
* The same applies for clients we want to terminate ASAP. */
if (c->flags & (CLIENT_CLOSE_AFTER_REPLY|CLIENT_CLOSE_ASAP)) break;
/* Determine request type when unknown. */
if (!c->reqtype) {
if (c->querybuf[c->qb_pos] == '*') {
c->reqtype = PROTO_REQ_MULTIBULK;
} else {
c->reqtype = PROTO_REQ_INLINE;
}
}
if (c->reqtype == PROTO_REQ_INLINE) {
if (processInlineBuffer(c) != C_OK) break;
} else if (c->reqtype == PROTO_REQ_MULTIBULK) {
if (processMultibulkBuffer(c) != C_OK) break;
} else {
serverPanic("Unknown request type");
}
/* Multibulk processing could see a <= 0 length. */
if (c->argc == 0) {
resetClient(c);
} else {
/* If we are in the context of an I/O thread, we can't really
* execute the command here. All we can do is to flag the client
* as one that needs to process the command. */
if (io_threads_op != IO_THREADS_OP_IDLE) {
serverAssert(io_threads_op == IO_THREADS_OP_READ);
c->flags |= CLIENT_PENDING_COMMAND;
break;
}
/* We are finally ready to execute the command. */
if (processCommandAndResetClient(c) == C_ERR) {
/* If the client is no longer valid, we avoid exiting this
* loop and trimming the client buffer later. So we return
* ASAP in that case. */
return C_ERR;
}
}
}
/* Trim to pos */
if (c->qb_pos) {
sdsrange(c->querybuf,c->qb_pos,-1);
c->qb_pos = 0;
}
/* Update client memory usage after processing the query buffer, this is
* important in case the query buffer is big and wasn't drained during
* the above loop (because of partially sent big commands). */
updateClientMemUsage(c);
return C_OK;
}
void readQueryFromClient(connection *conn) {
client *c = connGetPrivateData(conn);
int nread, big_arg = 0;
size_t qblen, readlen;
/* Check if we want to read from the client later when exiting from
* the event loop. This is the case if threaded I/O is enabled. */
if (postponeClientRead(c)) return;
/* Update total number of reads on server */
atomicIncr(server.stat_total_reads_processed, 1);
readlen = PROTO_IOBUF_LEN;
/* If this is a multi bulk request, and we are processing a bulk reply
* that is large enough, try to maximize the probability that the query
* buffer contains exactly the SDS string representing the object, even
* at the risk of requiring more read(2) calls. This way the function
* processMultiBulkBuffer() can avoid copying buffers to create the
* Redis Object representing the argument. */
if (c->reqtype == PROTO_REQ_MULTIBULK && c->multibulklen && c->bulklen != -1
&& c->bulklen >= PROTO_MBULK_BIG_ARG)
{
ssize_t remaining = (size_t)(c->bulklen+2)-sdslen(c->querybuf);
big_arg = 1;
/* Note that the 'remaining' variable may be zero in some edge case,
* for example once we resume a blocked client after CLIENT PAUSE. */
if (remaining > 0) readlen = remaining;
}
qblen = sdslen(c->querybuf);
if (big_arg || sdsalloc(c->querybuf) < PROTO_IOBUF_LEN) {
/* When reading a BIG_ARG we won't be reading more than that one arg
* into the query buffer, so we don't need to pre-allocate more than we
* need, so using the non-greedy growing. For an initial allocation of
* the query buffer, we also don't wanna use the greedy growth, in order
* to avoid collision with the RESIZE_THRESHOLD mechanism. */
c->querybuf = sdsMakeRoomForNonGreedy(c->querybuf, readlen);
} else {
c->querybuf = sdsMakeRoomFor(c->querybuf, readlen);
/* Read as much as possible from the socket to save read(2) system calls. */
readlen = sdsavail(c->querybuf);
}
nread = connRead(c->conn, c->querybuf+qblen, readlen);
if (nread == -1) {
if (connGetState(conn) == CONN_STATE_CONNECTED) {
return;
} else {
serverLog(LL_VERBOSE, "Reading from client: %s",connGetLastError(c->conn));
freeClientAsync(c);
goto done;
}
} else if (nread == 0) {
if (server.verbosity <= LL_VERBOSE) {
sds info = catClientInfoString(sdsempty(), c);
serverLog(LL_VERBOSE, "Client closed connection %s", info);
sdsfree(info);
}
freeClientAsync(c);
goto done;
} else if (c->flags & CLIENT_MASTER) {
/* Append the query buffer to the pending (not applied) buffer
* of the master. We'll use this buffer later in order to have a
* copy of the string applied by the last command executed. */
c->pending_querybuf = sdscatlen(c->pending_querybuf,
c->querybuf+qblen,nread);
}
sdsIncrLen(c->querybuf,nread);
qblen = sdslen(c->querybuf);
if (c->querybuf_peak < qblen) c->querybuf_peak = qblen;
c->lastinteraction = server.unixtime;
if (c->flags & CLIENT_MASTER) c->read_reploff += nread;
atomicIncr(server.stat_net_input_bytes, nread);
if (!(c->flags & CLIENT_MASTER) && sdslen(c->querybuf) > server.client_max_querybuf_len) {
sds ci = catClientInfoString(sdsempty(),c), bytes = sdsempty();
bytes = sdscatrepr(bytes,c->querybuf,64);
serverLog(LL_WARNING,"Closing client that reached max query buffer length: %s (qbuf initial bytes: %s)", ci, bytes);
sdsfree(ci);
sdsfree(bytes);
freeClientAsync(c);
goto done;
}
/* There is more data in the client input buffer, continue parsing it
* and check if there is a full command to execute. */
if (processInputBuffer(c) == C_ERR)
c = NULL;
done:
beforeNextClient(c);
}
/* A Redis "Address String" is a colon separated ip:port pair.
* For IPv4 it's in the form x.y.z.k:port, example: "127.0.0.1:1234".
* For IPv6 addresses we use [] around the IP part, like in "[::1]:1234".
* For Unix sockets we use path:0, like in "/tmp/redis:0".
*
* An Address String always fits inside a buffer of NET_ADDR_STR_LEN bytes,
* including the null term.
*
* On failure the function still populates 'addr' with the "?:0" string in case
* you want to relax error checking or need to display something anyway (see
* anetFdToString implementation for more info). */
void genClientAddrString(client *client, char *addr,
size_t addr_len, int fd_to_str_type) {
if (client->flags & CLIENT_UNIX_SOCKET) {
/* Unix socket client. */
snprintf(addr,addr_len,"%s:0",server.unixsocket);
} else {
/* TCP client. */
connFormatFdAddr(client->conn,addr,addr_len,fd_to_str_type);
}
}
/* This function returns the client peer id, by creating and caching it
* if client->peerid is NULL, otherwise returning the cached value.
* The Peer ID never changes during the life of the client, however it
* is expensive to compute. */
char *getClientPeerId(client *c) {
char peerid[NET_ADDR_STR_LEN];
if (c->peerid == NULL) {
genClientAddrString(c,peerid,sizeof(peerid),FD_TO_PEER_NAME);
c->peerid = sdsnew(peerid);
}
return c->peerid;
}
/* This function returns the client bound socket name, by creating and caching
* it if client->sockname is NULL, otherwise returning the cached value.
* The Socket Name never changes during the life of the client, however it
* is expensive to compute. */
char *getClientSockname(client *c) {
char sockname[NET_ADDR_STR_LEN];
if (c->sockname == NULL) {
genClientAddrString(c,sockname,sizeof(sockname),FD_TO_SOCK_NAME);
c->sockname = sdsnew(sockname);
}
return c->sockname;
}
/* Concatenate a string representing the state of a client in a human
* readable format, into the sds string 's'. */
sds catClientInfoString(sds s, client *client) {
char flags[16], events[3], conninfo[CONN_INFO_LEN], *p;
p = flags;
if (client->flags & CLIENT_SLAVE) {
if (client->flags & CLIENT_MONITOR)
*p++ = 'O';
else
*p++ = 'S';
}
if (client->flags & CLIENT_MASTER) *p++ = 'M';
if (client->flags & CLIENT_PUBSUB) *p++ = 'P';
if (client->flags & CLIENT_MULTI) *p++ = 'x';
if (client->flags & CLIENT_BLOCKED) *p++ = 'b';
if (client->flags & CLIENT_TRACKING) *p++ = 't';
if (client->flags & CLIENT_TRACKING_BROKEN_REDIR) *p++ = 'R';
if (client->flags & CLIENT_TRACKING_BCAST) *p++ = 'B';
if (client->flags & CLIENT_DIRTY_CAS) *p++ = 'd';
if (client->flags & CLIENT_CLOSE_AFTER_REPLY) *p++ = 'c';
if (client->flags & CLIENT_UNBLOCKED) *p++ = 'u';
if (client->flags & CLIENT_CLOSE_ASAP) *p++ = 'A';
if (client->flags & CLIENT_UNIX_SOCKET) *p++ = 'U';
if (client->flags & CLIENT_READONLY) *p++ = 'r';
if (client->flags & CLIENT_NO_EVICT) *p++ = 'e';
if (p == flags) *p++ = 'N';
*p++ = '\0';
p = events;
if (client->conn) {
if (connHasReadHandler(client->conn)) *p++ = 'r';
if (connHasWriteHandler(client->conn)) *p++ = 'w';
}
*p = '\0';
/* Compute the total memory consumed by this client. */
size_t obufmem, total_mem = getClientMemoryUsage(client, &obufmem);
size_t used_blocks_of_repl_buf = 0;
if (client->ref_repl_buf_node) {
replBufBlock *last = listNodeValue(listLast(server.repl_buffer_blocks));
replBufBlock *cur = listNodeValue(client->ref_repl_buf_node);
used_blocks_of_repl_buf = last->id - cur->id + 1;
}
sds cmdname = client->lastcmd ? getFullCommandName(client->lastcmd) : NULL;
sds ret = sdscatfmt(s,
"id=%U addr=%s laddr=%s %s name=%s age=%I idle=%I flags=%s db=%i sub=%i psub=%i multi=%i qbuf=%U qbuf-free=%U argv-mem=%U multi-mem=%U obl=%U oll=%U omem=%U tot-mem=%U events=%s cmd=%s user=%s redir=%I resp=%i",
(unsigned long long) client->id,
getClientPeerId(client),
getClientSockname(client),
connGetInfo(client->conn, conninfo, sizeof(conninfo)),
client->name ? (char*)client->name->ptr : "",
(long long)(server.unixtime - client->ctime),
(long long)(server.unixtime - client->lastinteraction),
flags,
client->db->id,
(int) dictSize(client->pubsub_channels),
(int) listLength(client->pubsub_patterns),
(client->flags & CLIENT_MULTI) ? client->mstate.count : -1,
(unsigned long long) sdslen(client->querybuf),
(unsigned long long) sdsavail(client->querybuf),
(unsigned long long) client->argv_len_sum,
(unsigned long long) client->mstate.argv_len_sums,
(unsigned long long) client->bufpos,
(unsigned long long) listLength(client->reply) + used_blocks_of_repl_buf,
(unsigned long long) obufmem, /* should not include client->buf since we want to see 0 for static clients. */
(unsigned long long) total_mem,
events,
cmdname ? cmdname : "NULL",
client->user ? client->user->name : "(superuser)",
(client->flags & CLIENT_TRACKING) ? (long long) client->client_tracking_redirection : -1,
client->resp);
if (cmdname)
sdsfree(cmdname);
return ret;
}
sds getAllClientsInfoString(int type) {
listNode *ln;
listIter li;
client *client;
sds o = sdsnewlen(SDS_NOINIT,200*listLength(server.clients));
sdsclear(o);
listRewind(server.clients,&li);
while ((ln = listNext(&li)) != NULL) {
client = listNodeValue(ln);
if (type != -1 && getClientType(client) != type) continue;
o = catClientInfoString(o,client);
o = sdscatlen(o,"\n",1);
}
return o;
}
/* This function implements CLIENT SETNAME, including replying to the
* user with an error if the charset is wrong (in that case C_ERR is
* returned). If the function succeeded C_OK is returned, and it's up
* to the caller to send a reply if needed.
*
* Setting an empty string as name has the effect of unsetting the
* currently set name: the client will remain unnamed.
*
* This function is also used to implement the HELLO SETNAME option. */
int clientSetNameOrReply(client *c, robj *name) {
int len = sdslen(name->ptr);
char *p = name->ptr;
/* Setting the client name to an empty string actually removes
* the current name. */
if (len == 0) {
if (c->name) decrRefCount(c->name);
c->name = NULL;
return C_OK;
}
/* Otherwise check if the charset is ok. We need to do this otherwise
* CLIENT LIST format will break. You should always be able to
* split by space to get the different fields. */
for (int j = 0; j < len; j++) {
if (p[j] < '!' || p[j] > '~') { /* ASCII is assumed. */
addReplyError(c,
"Client names cannot contain spaces, "
"newlines or special characters.");
return C_ERR;
}
}
if (c->name) decrRefCount(c->name);
c->name = name;
incrRefCount(name);
return C_OK;
}
/* Reset the client state to resemble a newly connected client.
*/
void resetCommand(client *c) {
/* MONITOR clients are also marked with CLIENT_SLAVE, we need to
* distinguish between the two.
*/
uint64_t flags = c->flags;
if (flags & CLIENT_MONITOR) flags &= ~(CLIENT_MONITOR|CLIENT_SLAVE);
if (flags & (CLIENT_SLAVE|CLIENT_MASTER|CLIENT_MODULE)) {
addReplyError(c,"can only reset normal client connections");
return;
}
clearClientConnectionState(c);
addReplyStatus(c,"RESET");
}
/* Disconnect the current client */
void quitCommand(client *c) {
addReply(c,shared.ok);
c->flags |= CLIENT_CLOSE_AFTER_REPLY;
}
void clientCommand(client *c) {
listNode *ln;
listIter li;
if (c->argc == 2 && !strcasecmp(c->argv[1]->ptr,"help")) {
const char *help[] = {
"CACHING (YES|NO)",
" Enable/disable tracking of the keys for next command in OPTIN/OPTOUT modes.",
"GETREDIR",
" Return the client ID we are redirecting to when tracking is enabled.",
"GETNAME",
" Return the name of the current connection.",
"ID",
" Return the ID of the current connection.",
"INFO",
" Return information about the current client connection.",
"KILL <ip:port>",
" Kill connection made from <ip:port>.",
"KILL <option> <value> [<option> <value> [...]]",
" Kill connections. Options are:",
" * ADDR (<ip:port>|<unixsocket>:0)",
" Kill connections made from the specified address",
" * LADDR (<ip:port>|<unixsocket>:0)",
" Kill connections made to specified local address",
" * TYPE (NORMAL|MASTER|REPLICA|PUBSUB)",
" Kill connections by type.",
" * USER <username>",
" Kill connections authenticated by <username>.",
" * SKIPME (YES|NO)",
" Skip killing current connection (default: yes).",
"LIST [options ...]",
" Return information about client connections. Options:",
" * TYPE (NORMAL|MASTER|REPLICA|PUBSUB)",
" Return clients of specified type.",
"UNPAUSE",
" Stop the current client pause, resuming traffic.",
"PAUSE <timeout> [WRITE|ALL]",
" Suspend all, or just write, clients for <timeout> milliseconds.",
"REPLY (ON|OFF|SKIP)",
" Control the replies sent to the current connection.",
"SETNAME <name>",
" Assign the name <name> to the current connection.",
"GETNAME",
" Get the name of the current connection.",
"UNBLOCK <clientid> [TIMEOUT|ERROR]",
" Unblock the specified blocked client.",
"TRACKING (ON|OFF) [REDIRECT <id>] [BCAST] [PREFIX <prefix> [...]]",
" [OPTIN] [OPTOUT] [NOLOOP]",
" Control server assisted client side caching.",
"TRACKINGINFO",
" Report tracking status for the current connection.",
"NO-EVICT (ON|OFF)",
" Protect current client connection from eviction.",
NULL
};
addReplyHelp(c, help);
} else if (!strcasecmp(c->argv[1]->ptr,"id") && c->argc == 2) {
/* CLIENT ID */
addReplyLongLong(c,c->id);
} else if (!strcasecmp(c->argv[1]->ptr,"info") && c->argc == 2) {
/* CLIENT INFO */
sds o = catClientInfoString(sdsempty(), c);
o = sdscatlen(o,"\n",1);
addReplyVerbatim(c,o,sdslen(o),"txt");
sdsfree(o);
} else if (!strcasecmp(c->argv[1]->ptr,"list")) {
/* CLIENT LIST */
int type = -1;
sds o = NULL;
if (c->argc == 4 && !strcasecmp(c->argv[2]->ptr,"type")) {
type = getClientTypeByName(c->argv[3]->ptr);
if (type == -1) {
addReplyErrorFormat(c,"Unknown client type '%s'",
(char*) c->argv[3]->ptr);
return;
}
} else if (c->argc > 3 && !strcasecmp(c->argv[2]->ptr,"id")) {
int j;
o = sdsempty();
for (j = 3; j < c->argc; j++) {
long long cid;
if (getLongLongFromObjectOrReply(c, c->argv[j], &cid,
"Invalid client ID")) {
sdsfree(o);
return;
}
client *cl = lookupClientByID(cid);
if (cl) {
o = catClientInfoString(o, cl);
o = sdscatlen(o, "\n", 1);
}
}
} else if (c->argc != 2) {
addReplyErrorObject(c,shared.syntaxerr);
return;
}
if (!o)
o = getAllClientsInfoString(type);
addReplyVerbatim(c,o,sdslen(o),"txt");
sdsfree(o);
} else if (!strcasecmp(c->argv[1]->ptr,"reply") && c->argc == 3) {
/* CLIENT REPLY ON|OFF|SKIP */
if (!strcasecmp(c->argv[2]->ptr,"on")) {
c->flags &= ~(CLIENT_REPLY_SKIP|CLIENT_REPLY_OFF);
addReply(c,shared.ok);
} else if (!strcasecmp(c->argv[2]->ptr,"off")) {
c->flags |= CLIENT_REPLY_OFF;
} else if (!strcasecmp(c->argv[2]->ptr,"skip")) {
if (!(c->flags & CLIENT_REPLY_OFF))
c->flags |= CLIENT_REPLY_SKIP_NEXT;
} else {
addReplyErrorObject(c,shared.syntaxerr);
return;
}
} else if (!strcasecmp(c->argv[1]->ptr,"no-evict") && c->argc == 3) {
/* CLIENT NO-EVICT ON|OFF */
if (!strcasecmp(c->argv[2]->ptr,"on")) {
c->flags |= CLIENT_NO_EVICT;
addReply(c,shared.ok);
} else if (!strcasecmp(c->argv[2]->ptr,"off")) {
c->flags &= ~CLIENT_NO_EVICT;
addReply(c,shared.ok);
} else {
addReplyErrorObject(c,shared.syntaxerr);
return;
}
} else if (!strcasecmp(c->argv[1]->ptr,"kill")) {
/* CLIENT KILL <ip:port>
* CLIENT KILL <option> [value] ... <option> [value] */
char *addr = NULL;
char *laddr = NULL;
user *user = NULL;
int type = -1;
uint64_t id = 0;
int skipme = 1;
int killed = 0, close_this_client = 0;
if (c->argc == 3) {
/* Old style syntax: CLIENT KILL <addr> */
addr = c->argv[2]->ptr;
skipme = 0; /* With the old form, you can kill yourself. */
} else if (c->argc > 3) {
int i = 2; /* Next option index. */
/* New style syntax: parse options. */
while(i < c->argc) {
int moreargs = c->argc > i+1;
if (!strcasecmp(c->argv[i]->ptr,"id") && moreargs) {
long tmp;
if (getRangeLongFromObjectOrReply(c, c->argv[i+1], 1, LONG_MAX, &tmp,
"client-id should be greater than 0") != C_OK)
return;
id = tmp;
} else if (!strcasecmp(c->argv[i]->ptr,"type") && moreargs) {
type = getClientTypeByName(c->argv[i+1]->ptr);
if (type == -1) {
addReplyErrorFormat(c,"Unknown client type '%s'",
(char*) c->argv[i+1]->ptr);
return;
}
} else if (!strcasecmp(c->argv[i]->ptr,"addr") && moreargs) {
addr = c->argv[i+1]->ptr;
} else if (!strcasecmp(c->argv[i]->ptr,"laddr") && moreargs) {
laddr = c->argv[i+1]->ptr;
} else if (!strcasecmp(c->argv[i]->ptr,"user") && moreargs) {
user = ACLGetUserByName(c->argv[i+1]->ptr,
sdslen(c->argv[i+1]->ptr));
if (user == NULL) {
addReplyErrorFormat(c,"No such user '%s'",
(char*) c->argv[i+1]->ptr);
return;
}
} else if (!strcasecmp(c->argv[i]->ptr,"skipme") && moreargs) {
if (!strcasecmp(c->argv[i+1]->ptr,"yes")) {
skipme = 1;
} else if (!strcasecmp(c->argv[i+1]->ptr,"no")) {
skipme = 0;
} else {
addReplyErrorObject(c,shared.syntaxerr);
return;
}
} else {
addReplyErrorObject(c,shared.syntaxerr);
return;
}
i += 2;
}
} else {
addReplyErrorObject(c,shared.syntaxerr);
return;
}
/* Iterate clients killing all the matching clients. */
listRewind(server.clients,&li);
while ((ln = listNext(&li)) != NULL) {
client *client = listNodeValue(ln);
if (addr && strcmp(getClientPeerId(client),addr) != 0) continue;
if (laddr && strcmp(getClientSockname(client),laddr) != 0) continue;
if (type != -1 && getClientType(client) != type) continue;
if (id != 0 && client->id != id) continue;
if (user && client->user != user) continue;
if (c == client && skipme) continue;
/* Kill it. */
if (c == client) {
close_this_client = 1;
} else {
freeClient(client);
}
killed++;
}
/* Reply according to old/new format. */
if (c->argc == 3) {
if (killed == 0)
addReplyError(c,"No such client");
else
addReply(c,shared.ok);
} else {
addReplyLongLong(c,killed);
}
/* If this client has to be closed, flag it as CLOSE_AFTER_REPLY
* only after we queued the reply to its output buffers. */
if (close_this_client) c->flags |= CLIENT_CLOSE_AFTER_REPLY;
} else if (!strcasecmp(c->argv[1]->ptr,"unblock") && (c->argc == 3 ||
c->argc == 4))
{
/* CLIENT UNBLOCK <id> [timeout|error] */
long long id;
int unblock_error = 0;
if (c->argc == 4) {
if (!strcasecmp(c->argv[3]->ptr,"timeout")) {
unblock_error = 0;
} else if (!strcasecmp(c->argv[3]->ptr,"error")) {
unblock_error = 1;
} else {
addReplyError(c,
"CLIENT UNBLOCK reason should be TIMEOUT or ERROR");
return;
}
}
if (getLongLongFromObjectOrReply(c,c->argv[2],&id,NULL)
!= C_OK) return;
struct client *target = lookupClientByID(id);
if (target && target->flags & CLIENT_BLOCKED && moduleBlockedClientMayTimeout(target)) {
if (unblock_error)
addReplyError(target,
"-UNBLOCKED client unblocked via CLIENT UNBLOCK");
else
replyToBlockedClientTimedOut(target);
unblockClient(target);
addReply(c,shared.cone);
} else {
addReply(c,shared.czero);
}
} else if (!strcasecmp(c->argv[1]->ptr,"setname") && c->argc == 3) {
/* CLIENT SETNAME */
if (clientSetNameOrReply(c,c->argv[2]) == C_OK)
addReply(c,shared.ok);
} else if (!strcasecmp(c->argv[1]->ptr,"getname") && c->argc == 2) {
/* CLIENT GETNAME */
if (c->name)
addReplyBulk(c,c->name);
else
addReplyNull(c);
} else if (!strcasecmp(c->argv[1]->ptr,"unpause") && c->argc == 2) {
/* CLIENT UNPAUSE */
unpauseClients(PAUSE_BY_CLIENT_COMMAND);
addReply(c,shared.ok);
} else if (!strcasecmp(c->argv[1]->ptr,"pause") && (c->argc == 3 ||
c->argc == 4))
{
/* CLIENT PAUSE TIMEOUT [WRITE|ALL] */
mstime_t end;
int type = CLIENT_PAUSE_ALL;
if (c->argc == 4) {
if (!strcasecmp(c->argv[3]->ptr,"write")) {
type = CLIENT_PAUSE_WRITE;
} else if (!strcasecmp(c->argv[3]->ptr,"all")) {
type = CLIENT_PAUSE_ALL;
} else {
addReplyError(c,
"CLIENT PAUSE mode must be WRITE or ALL");
return;
}
}
if (getTimeoutFromObjectOrReply(c,c->argv[2],&end,
UNIT_MILLISECONDS) != C_OK) return;
pauseClients(PAUSE_BY_CLIENT_COMMAND, end, type);
addReply(c,shared.ok);
} else if (!strcasecmp(c->argv[1]->ptr,"tracking") && c->argc >= 3) {
/* CLIENT TRACKING (on|off) [REDIRECT <id>] [BCAST] [PREFIX first]
* [PREFIX second] [OPTIN] [OPTOUT] [NOLOOP]... */
long long redir = 0;
uint64_t options = 0;
robj **prefix = NULL;
size_t numprefix = 0;
/* Parse the options. */
for (int j = 3; j < c->argc; j++) {
int moreargs = (c->argc-1) - j;
if (!strcasecmp(c->argv[j]->ptr,"redirect") && moreargs) {
j++;
if (redir != 0) {
addReplyError(c,"A client can only redirect to a single "
"other client");
zfree(prefix);
return;
}
if (getLongLongFromObjectOrReply(c,c->argv[j],&redir,NULL) !=
C_OK)
{
zfree(prefix);
return;
}
/* We will require the client with the specified ID to exist
* right now, even if it is possible that it gets disconnected
* later. Still a valid sanity check. */
if (lookupClientByID(redir) == NULL) {
addReplyError(c,"The client ID you want redirect to "
"does not exist");
zfree(prefix);
return;
}
} else if (!strcasecmp(c->argv[j]->ptr,"bcast")) {
options |= CLIENT_TRACKING_BCAST;
} else if (!strcasecmp(c->argv[j]->ptr,"optin")) {
options |= CLIENT_TRACKING_OPTIN;
} else if (!strcasecmp(c->argv[j]->ptr,"optout")) {
options |= CLIENT_TRACKING_OPTOUT;
} else if (!strcasecmp(c->argv[j]->ptr,"noloop")) {
options |= CLIENT_TRACKING_NOLOOP;
} else if (!strcasecmp(c->argv[j]->ptr,"prefix") && moreargs) {
j++;
prefix = zrealloc(prefix,sizeof(robj*)*(numprefix+1));
prefix[numprefix++] = c->argv[j];
} else {
zfree(prefix);
addReplyErrorObject(c,shared.syntaxerr);
return;
}
}
/* Options are ok: enable or disable the tracking for this client. */
if (!strcasecmp(c->argv[2]->ptr,"on")) {
/* Before enabling tracking, make sure options are compatible
* among each other and with the current state of the client. */
if (!(options & CLIENT_TRACKING_BCAST) && numprefix) {
addReplyError(c,
"PREFIX option requires BCAST mode to be enabled");
zfree(prefix);
return;
}
if (c->flags & CLIENT_TRACKING) {
int oldbcast = !!(c->flags & CLIENT_TRACKING_BCAST);
int newbcast = !!(options & CLIENT_TRACKING_BCAST);
if (oldbcast != newbcast) {
addReplyError(c,
"You can't switch BCAST mode on/off before disabling "
"tracking for this client, and then re-enabling it with "
"a different mode.");
zfree(prefix);
return;
}
}
if (options & CLIENT_TRACKING_BCAST &&
options & (CLIENT_TRACKING_OPTIN|CLIENT_TRACKING_OPTOUT))
{
addReplyError(c,
"OPTIN and OPTOUT are not compatible with BCAST");
zfree(prefix);
return;
}
if (options & CLIENT_TRACKING_OPTIN && options & CLIENT_TRACKING_OPTOUT)
{
addReplyError(c,
"You can't specify both OPTIN mode and OPTOUT mode");
zfree(prefix);
return;
}
if ((options & CLIENT_TRACKING_OPTIN && c->flags & CLIENT_TRACKING_OPTOUT) ||
(options & CLIENT_TRACKING_OPTOUT && c->flags & CLIENT_TRACKING_OPTIN))
{
addReplyError(c,
"You can't switch OPTIN/OPTOUT mode before disabling "
"tracking for this client, and then re-enabling it with "
"a different mode.");
zfree(prefix);
return;
}
if (options & CLIENT_TRACKING_BCAST) {
if (!checkPrefixCollisionsOrReply(c,prefix,numprefix)) {
zfree(prefix);
return;
}
}
enableTracking(c,redir,options,prefix,numprefix);
} else if (!strcasecmp(c->argv[2]->ptr,"off")) {
disableTracking(c);
} else {
zfree(prefix);
addReplyErrorObject(c,shared.syntaxerr);
return;
}
zfree(prefix);
addReply(c,shared.ok);
} else if (!strcasecmp(c->argv[1]->ptr,"caching") && c->argc >= 3) {
if (!(c->flags & CLIENT_TRACKING)) {
addReplyError(c,"CLIENT CACHING can be called only when the "
"client is in tracking mode with OPTIN or "
"OPTOUT mode enabled");
return;
}
char *opt = c->argv[2]->ptr;
if (!strcasecmp(opt,"yes")) {
if (c->flags & CLIENT_TRACKING_OPTIN) {
c->flags |= CLIENT_TRACKING_CACHING;
} else {
addReplyError(c,"CLIENT CACHING YES is only valid when tracking is enabled in OPTIN mode.");
return;
}
} else if (!strcasecmp(opt,"no")) {
if (c->flags & CLIENT_TRACKING_OPTOUT) {
c->flags |= CLIENT_TRACKING_CACHING;
} else {
addReplyError(c,"CLIENT CACHING NO is only valid when tracking is enabled in OPTOUT mode.");
return;
}
} else {
addReplyErrorObject(c,shared.syntaxerr);
return;
}
/* Common reply for when we succeeded. */
addReply(c,shared.ok);
} else if (!strcasecmp(c->argv[1]->ptr,"getredir") && c->argc == 2) {
/* CLIENT GETREDIR */
if (c->flags & CLIENT_TRACKING) {
addReplyLongLong(c,c->client_tracking_redirection);
} else {
addReplyLongLong(c,-1);
}
} else if (!strcasecmp(c->argv[1]->ptr,"trackinginfo") && c->argc == 2) {
addReplyMapLen(c,3);
/* Flags */
addReplyBulkCString(c,"flags");
void *arraylen_ptr = addReplyDeferredLen(c);
int numflags = 0;
addReplyBulkCString(c,c->flags & CLIENT_TRACKING ? "on" : "off");
numflags++;
if (c->flags & CLIENT_TRACKING_BCAST) {
addReplyBulkCString(c,"bcast");
numflags++;
}
if (c->flags & CLIENT_TRACKING_OPTIN) {
addReplyBulkCString(c,"optin");
numflags++;
if (c->flags & CLIENT_TRACKING_CACHING) {
addReplyBulkCString(c,"caching-yes");
numflags++;
}
}
if (c->flags & CLIENT_TRACKING_OPTOUT) {
addReplyBulkCString(c,"optout");
numflags++;
if (c->flags & CLIENT_TRACKING_CACHING) {
addReplyBulkCString(c,"caching-no");
numflags++;
}
}
if (c->flags & CLIENT_TRACKING_NOLOOP) {
addReplyBulkCString(c,"noloop");
numflags++;
}
if (c->flags & CLIENT_TRACKING_BROKEN_REDIR) {
addReplyBulkCString(c,"broken_redirect");
numflags++;
}
setDeferredSetLen(c,arraylen_ptr,numflags);
/* Redirect */
addReplyBulkCString(c,"redirect");
if (c->flags & CLIENT_TRACKING) {
addReplyLongLong(c,c->client_tracking_redirection);
} else {
addReplyLongLong(c,-1);
}
/* Prefixes */
addReplyBulkCString(c,"prefixes");
if (c->client_tracking_prefixes) {
addReplyArrayLen(c,raxSize(c->client_tracking_prefixes));
raxIterator ri;
raxStart(&ri,c->client_tracking_prefixes);
raxSeek(&ri,"^",NULL,0);
while(raxNext(&ri)) {
addReplyBulkCBuffer(c,ri.key,ri.key_len);
}
raxStop(&ri);
} else {
addReplyArrayLen(c,0);
}
} else {
addReplySubcommandSyntaxError(c);
}
}
/* HELLO [<protocol-version> [AUTH <user> <password>] [SETNAME <name>] ] */
void helloCommand(client *c) {
long long ver = 0;
int next_arg = 1;
if (c->argc >= 2) {
if (getLongLongFromObjectOrReply(c, c->argv[next_arg++], &ver,
"Protocol version is not an integer or out of range") != C_OK) {
return;
}
if (ver < 2 || ver > 3) {
addReplyError(c,"-NOPROTO unsupported protocol version");
return;
}
}
for (int j = next_arg; j < c->argc; j++) {
int moreargs = (c->argc-1) - j;
const char *opt = c->argv[j]->ptr;
if (!strcasecmp(opt,"AUTH") && moreargs >= 2) {
redactClientCommandArgument(c, j+1);
redactClientCommandArgument(c, j+2);
if (ACLAuthenticateUser(c, c->argv[j+1], c->argv[j+2]) == C_ERR) {
addReplyError(c,"-WRONGPASS invalid username-password pair or user is disabled.");
return;
}
j += 2;
} else if (!strcasecmp(opt,"SETNAME") && moreargs) {
if (clientSetNameOrReply(c, c->argv[j+1]) == C_ERR) return;
j++;
} else {
addReplyErrorFormat(c,"Syntax error in HELLO option '%s'",opt);
return;
}
}
/* At this point we need to be authenticated to continue. */
if (!c->authenticated) {
addReplyError(c,"-NOAUTH HELLO must be called with the client already "
"authenticated, otherwise the HELLO AUTH <user> <pass> "
"option can be used to authenticate the client and "
"select the RESP protocol version at the same time");
return;
}
/* Let's switch to the specified RESP mode. */
if (ver) c->resp = ver;
addReplyMapLen(c,6 + !server.sentinel_mode);
addReplyBulkCString(c,"server");
addReplyBulkCString(c,"redis");
addReplyBulkCString(c,"version");
addReplyBulkCString(c,REDIS_VERSION);
addReplyBulkCString(c,"proto");
addReplyLongLong(c,c->resp);
addReplyBulkCString(c,"id");
addReplyLongLong(c,c->id);
addReplyBulkCString(c,"mode");
if (server.sentinel_mode) addReplyBulkCString(c,"sentinel");
else if (server.cluster_enabled) addReplyBulkCString(c,"cluster");
else addReplyBulkCString(c,"standalone");
if (!server.sentinel_mode) {
addReplyBulkCString(c,"role");
addReplyBulkCString(c,server.masterhost ? "replica" : "master");
}
addReplyBulkCString(c,"modules");
addReplyLoadedModules(c);
}
/* This callback is bound to POST and "Host:" command names. Those are not
* really commands, but are used in security attacks in order to talk to
* Redis instances via HTTP, with a technique called "cross protocol scripting"
* which exploits the fact that services like Redis will discard invalid
* HTTP headers and will process what follows.
*
* As a protection against this attack, Redis will terminate the connection
* when a POST or "Host:" header is seen, and will log the event from
* time to time (to avoid creating a DOS as a result of too many logs). */
void securityWarningCommand(client *c) {
static time_t logged_time = 0;
time_t now = time(NULL);
if (llabs(now-logged_time) > 60) {
serverLog(LL_WARNING,"Possible SECURITY ATTACK detected. It looks like somebody is sending POST or Host: commands to Redis. This is likely due to an attacker attempting to use Cross Protocol Scripting to compromise your Redis instance. Connection aborted.");
logged_time = now;
}
freeClientAsync(c);
}
/* Keep track of the original command arguments so that we can generate
* an accurate slowlog entry after the command has been executed. */
static void retainOriginalCommandVector(client *c) {
/* We already rewrote this command, so don't rewrite it again */
if (c->original_argv) return;
c->original_argc = c->argc;
c->original_argv = zmalloc(sizeof(robj*)*(c->argc));
for (int j = 0; j < c->argc; j++) {
c->original_argv[j] = c->argv[j];
incrRefCount(c->argv[j]);
}
}
/* Redact a given argument to prevent it from being shown
* in the slowlog. This information is stored in the
* original_argv array. */
void redactClientCommandArgument(client *c, int argc) {
retainOriginalCommandVector(c);
decrRefCount(c->argv[argc]);
c->original_argv[argc] = shared.redacted;
}
/* Rewrite the command vector of the client. All the new objects ref count
* is incremented. The old command vector is freed, and the old objects
* ref count is decremented. */
void rewriteClientCommandVector(client *c, int argc, ...) {
va_list ap;
int j;
robj **argv; /* The new argument vector */
argv = zmalloc(sizeof(robj*)*argc);
va_start(ap,argc);
for (j = 0; j < argc; j++) {
robj *a;
a = va_arg(ap, robj*);
argv[j] = a;
incrRefCount(a);
}
replaceClientCommandVector(c, argc, argv);
va_end(ap);
}
/* Completely replace the client command vector with the provided one. */
void replaceClientCommandVector(client *c, int argc, robj **argv) {
int j;
retainOriginalCommandVector(c);
freeClientArgv(c);
zfree(c->argv);
c->argv = argv;
c->argc = argc;
c->argv_len_sum = 0;
for (j = 0; j < c->argc; j++)
if (c->argv[j])
c->argv_len_sum += getStringObjectLen(c->argv[j]);
c->cmd = lookupCommandOrOriginal(c->argv,c->argc);
serverAssertWithInfo(c,NULL,c->cmd != NULL);
}
/* Rewrite a single item in the command vector.
* The new val ref count is incremented, and the old decremented.
*
* It is possible to specify an argument over the current size of the
* argument vector: in this case the array of objects gets reallocated
* and c->argc set to the max value. However it's up to the caller to
*
* 1. Make sure there are no "holes" and all the arguments are set.
* 2. If the original argument vector was longer than the one we
* want to end with, it's up to the caller to set c->argc and
* free the no longer used objects on c->argv. */
void rewriteClientCommandArgument(client *c, int i, robj *newval) {
robj *oldval;
retainOriginalCommandVector(c);
/* We need to handle both extending beyond argc (just update it and
* initialize the new element) or beyond argv_len (realloc is needed).
*/
if (i >= c->argc) {
if (i >= c->argv_len) {
c->argv = zrealloc(c->argv,sizeof(robj*)*(i+1));
c->argv_len = i+1;
}
c->argc = i+1;
c->argv[i] = NULL;
}
oldval = c->argv[i];
if (oldval) c->argv_len_sum -= getStringObjectLen(oldval);
if (newval) c->argv_len_sum += getStringObjectLen(newval);
c->argv[i] = newval;
incrRefCount(newval);
if (oldval) decrRefCount(oldval);
/* If this is the command name make sure to fix c->cmd. */
if (i == 0) {
c->cmd = lookupCommandOrOriginal(c->argv,c->argc);
serverAssertWithInfo(c,NULL,c->cmd != NULL);
}
}
/* This function returns the number of bytes that Redis is
* using to store the reply still not read by the client.
*
* Note: this function is very fast so can be called as many time as
* the caller wishes. The main usage of this function currently is
* enforcing the client output length limits. */
size_t getClientOutputBufferMemoryUsage(client *c) {
if (getClientType(c) == CLIENT_TYPE_SLAVE) {
size_t repl_buf_size = 0;
size_t repl_node_num = 0;
size_t repl_node_size = sizeof(listNode) + sizeof(replBufBlock);
if (c->ref_repl_buf_node) {
replBufBlock *last = listNodeValue(listLast(server.repl_buffer_blocks));
replBufBlock *cur = listNodeValue(c->ref_repl_buf_node);
repl_buf_size = last->repl_offset + last->size - cur->repl_offset;
repl_node_num = last->id - cur->id + 1;
}
return repl_buf_size + (repl_node_size*repl_node_num);
} else {
size_t list_item_size = sizeof(listNode) + sizeof(clientReplyBlock);
return c->reply_bytes + (list_item_size*listLength(c->reply));
}
}
/* Returns the total client's memory usage.
* Optionally, if output_buffer_mem_usage is not NULL, it fills it with
* the client output buffer memory usage portion of the total. */
size_t getClientMemoryUsage(client *c, size_t *output_buffer_mem_usage) {
size_t mem = getClientOutputBufferMemoryUsage(c);
if (output_buffer_mem_usage != NULL)
*output_buffer_mem_usage = mem;
mem += sdsZmallocSize(c->querybuf);
mem += zmalloc_size(c);
/* For efficiency (less work keeping track of the argv memory), it doesn't include the used memory
* i.e. unused sds space and internal fragmentation, just the string length. but this is enough to
* spot problematic clients. */
mem += c->argv_len_sum + sizeof(robj*)*c->argc;
mem += multiStateMemOverhead(c);
/* Add memory overhead of pubsub channels and patterns. Note: this is just the overhead of the robj pointers
* to the strings themselves because they aren't stored per client. */
mem += listLength(c->pubsub_patterns) * sizeof(listNode);
mem += dictSize(c->pubsub_channels) * sizeof(dictEntry) +
dictSlots(c->pubsub_channels) * sizeof(dictEntry*);
/* Add memory overhead of the tracking prefixes, this is an underestimation so we don't need to traverse the entire rax */
if (c->client_tracking_prefixes)
mem += c->client_tracking_prefixes->numnodes * (sizeof(raxNode) * sizeof(raxNode*));
return mem;
}
/* Get the class of a client, used in order to enforce limits to different
* classes of clients.
*
* The function will return one of the following:
* CLIENT_TYPE_NORMAL -> Normal client
* CLIENT_TYPE_SLAVE -> Slave
* CLIENT_TYPE_PUBSUB -> Client subscribed to Pub/Sub channels
* CLIENT_TYPE_MASTER -> The client representing our replication master.
*/
int getClientType(client *c) {
if (c->flags & CLIENT_MASTER) return CLIENT_TYPE_MASTER;
/* Even though MONITOR clients are marked as replicas, we
* want the expose them as normal clients. */
if ((c->flags & CLIENT_SLAVE) && !(c->flags & CLIENT_MONITOR))
return CLIENT_TYPE_SLAVE;
if (c->flags & CLIENT_PUBSUB) return CLIENT_TYPE_PUBSUB;
return CLIENT_TYPE_NORMAL;
}
int getClientTypeByName(char *name) {
if (!strcasecmp(name,"normal")) return CLIENT_TYPE_NORMAL;
else if (!strcasecmp(name,"slave")) return CLIENT_TYPE_SLAVE;
else if (!strcasecmp(name,"replica")) return CLIENT_TYPE_SLAVE;
else if (!strcasecmp(name,"pubsub")) return CLIENT_TYPE_PUBSUB;
else if (!strcasecmp(name,"master")) return CLIENT_TYPE_MASTER;
else return -1;
}
char *getClientTypeName(int class) {
switch(class) {
case CLIENT_TYPE_NORMAL: return "normal";
case CLIENT_TYPE_SLAVE: return "slave";
case CLIENT_TYPE_PUBSUB: return "pubsub";
case CLIENT_TYPE_MASTER: return "master";
default: return NULL;
}
}
/* The function checks if the client reached output buffer soft or hard
* limit, and also update the state needed to check the soft limit as
* a side effect.
*
* Return value: non-zero if the client reached the soft or the hard limit.
* Otherwise zero is returned. */
int checkClientOutputBufferLimits(client *c) {
int soft = 0, hard = 0, class;
unsigned long used_mem = getClientOutputBufferMemoryUsage(c);
class = getClientType(c);
/* For the purpose of output buffer limiting, masters are handled
* like normal clients. */
if (class == CLIENT_TYPE_MASTER) class = CLIENT_TYPE_NORMAL;
/* Note that it doesn't make sense to set the replica clients output buffer
* limit lower than the repl-backlog-size config (partial sync will succeed
* and then replica will get disconnected).
* Such a configuration is ignored (the size of repl-backlog-size will be used).
* This doesn't have memory consumption implications since the replica client
* will share the backlog buffers memory. */
size_t hard_limit_bytes = server.client_obuf_limits[class].hard_limit_bytes;
if (class == CLIENT_TYPE_SLAVE && hard_limit_bytes &&
(long long)hard_limit_bytes < server.repl_backlog_size)
hard_limit_bytes = server.repl_backlog_size;
if (server.client_obuf_limits[class].hard_limit_bytes &&
used_mem >= hard_limit_bytes)
hard = 1;
if (server.client_obuf_limits[class].soft_limit_bytes &&
used_mem >= server.client_obuf_limits[class].soft_limit_bytes)
soft = 1;
/* We need to check if the soft limit is reached continuously for the
* specified amount of seconds. */
if (soft) {
if (c->obuf_soft_limit_reached_time == 0) {
c->obuf_soft_limit_reached_time = server.unixtime;
soft = 0; /* First time we see the soft limit reached */
} else {
time_t elapsed = server.unixtime - c->obuf_soft_limit_reached_time;
if (elapsed <=
server.client_obuf_limits[class].soft_limit_seconds) {
soft = 0; /* The client still did not reached the max number of
seconds for the soft limit to be considered
reached. */
}
}
} else {
c->obuf_soft_limit_reached_time = 0;
}
return soft || hard;
}
/* Asynchronously close a client if soft or hard limit is reached on the
* output buffer size. The caller can check if the client will be closed
* checking if the client CLIENT_CLOSE_ASAP flag is set.
*
* Note: we need to close the client asynchronously because this function is
* called from contexts where the client can't be freed safely, i.e. from the
* lower level functions pushing data inside the client output buffers.
* When `async` is set to 0, we close the client immediately, this is
* useful when called from cron.
*
* Returns 1 if client was (flagged) closed. */
int closeClientOnOutputBufferLimitReached(client *c, int async) {
if (!c->conn) return 0; /* It is unsafe to free fake clients. */
serverAssert(c->reply_bytes < SIZE_MAX-(1024*64));
/* Note that c->reply_bytes is irrelevant for replica clients
* (they use the global repl buffers). */
if ((c->reply_bytes == 0 && getClientType(c) != CLIENT_TYPE_SLAVE) ||
c->flags & CLIENT_CLOSE_ASAP) return 0;
if (checkClientOutputBufferLimits(c)) {
sds client = catClientInfoString(sdsempty(),c);
if (async) {
freeClientAsync(c);
serverLog(LL_WARNING,
"Client %s scheduled to be closed ASAP for overcoming of output buffer limits.",
client);
} else {
freeClient(c);
serverLog(LL_WARNING,
"Client %s closed for overcoming of output buffer limits.",
client);
}
sdsfree(client);
return 1;
}
return 0;
}
/* Helper function used by performEvictions() in order to flush slaves
* output buffers without returning control to the event loop.
* This is also called by SHUTDOWN for a best-effort attempt to send
* slaves the latest writes. */
void flushSlavesOutputBuffers(void) {
listIter li;
listNode *ln;
listRewind(server.slaves,&li);
while((ln = listNext(&li))) {
client *slave = listNodeValue(ln);
int can_receive_writes = connHasWriteHandler(slave->conn) ||
(slave->flags & CLIENT_PENDING_WRITE);
/* We don't want to send the pending data to the replica in a few
* cases:
*
* 1. For some reason there is neither the write handler installed
* nor the client is flagged as to have pending writes: for some
* reason this replica may not be set to receive data. This is
* just for the sake of defensive programming.
*
* 2. The put_online_on_ack flag is true. To know why we don't want
* to send data to the replica in this case, please grep for the
* flag for this flag.
*
* 3. Obviously if the slave is not ONLINE.
*/
if (slave->replstate == SLAVE_STATE_ONLINE &&
can_receive_writes &&
!slave->repl_start_cmd_stream_on_ack &&
clientHasPendingReplies(slave))
{
writeToClient(slave,0);
}
}
}
/* Compute current most restictive pause type and its end time, aggregated for
* all pause purposes. */
static void updateClientPauseTypeAndEndTime(void) {
pause_type old_type = server.client_pause_type;
pause_type type = CLIENT_PAUSE_OFF;
mstime_t end = 0;
for (int i = 0; i < NUM_PAUSE_PURPOSES; i++) {
pause_event *p = server.client_pause_per_purpose[i];
if (p == NULL) {
/* Nothing to do. */
} else if (p->end < server.mstime) {
/* This one expired. */
zfree(p);
server.client_pause_per_purpose[i] = NULL;
} else if (p->type > type) {
/* This type is the most restrictive so far. */
type = p->type;
}
}
/* Find the furthest end time among the pause purposes of the most
* restrictive type */
for (int i = 0; i < NUM_PAUSE_PURPOSES; i++) {
pause_event *p = server.client_pause_per_purpose[i];
if (p != NULL && p->type == type && p->end > end) end = p->end;
}
server.client_pause_type = type;
server.client_pause_end_time = end;
/* If the pause type is less restrictive than before, we unblock all clients
* so they are reprocessed (may get re-paused). */
if (type < old_type) {
listNode *ln;
listIter li;
listRewind(server.paused_clients, &li);
while ((ln = listNext(&li)) != NULL) {
client *c = listNodeValue(ln);
unblockClient(c);
}
}
}
/* Pause clients up to the specified unixtime (in ms) for a given type of
* commands.
*
* A main use case of this function is to allow pausing replication traffic
* so that a failover without data loss to occur. Replicas will continue to receive
* traffic to facilitate this functionality.
*
* This function is also internally used by Redis Cluster for the manual
* failover procedure implemented by CLUSTER FAILOVER.
*
* The function always succeed, even if there is already a pause in progress.
* In such a case, the duration is set to the maximum and new end time and the
* type is set to the more restrictive type of pause. */
void pauseClients(pause_purpose purpose, mstime_t end, pause_type type) {
/* Manage pause type and end time per pause purpose. */
if (server.client_pause_per_purpose[purpose] == NULL) {
server.client_pause_per_purpose[purpose] = zmalloc(sizeof(pause_event));
server.client_pause_per_purpose[purpose]->type = type;
server.client_pause_per_purpose[purpose]->end = end;
} else {
pause_event *p = server.client_pause_per_purpose[purpose];
p->type = max(p->type, type);
p->end = max(p->end, end);
}
updateClientPauseTypeAndEndTime();
/* We allow write commands that were queued
* up before and after to execute. We need
* to track this state so that we don't assert
* in propagateNow(). */
if (server.in_exec) {
server.client_pause_in_transaction = 1;
}
}
/* Unpause clients and queue them for reprocessing. */
void unpauseClients(pause_purpose purpose) {
if (server.client_pause_per_purpose[purpose] == NULL) return;
zfree(server.client_pause_per_purpose[purpose]);
server.client_pause_per_purpose[purpose] = NULL;
updateClientPauseTypeAndEndTime();
}
/* Returns true if clients are paused and false otherwise. */
int areClientsPaused(void) {
return server.client_pause_type != CLIENT_PAUSE_OFF;
}
/* Checks if the current client pause has elapsed and unpause clients
* if it has. Also returns true if clients are now paused and false
* otherwise. */
int checkClientPauseTimeoutAndReturnIfPaused(void) {
if (!areClientsPaused())
return 0;
if (server.client_pause_end_time < server.mstime) {
updateClientPauseTypeAndEndTime();
}
return areClientsPaused();
}
/* This function is called by Redis in order to process a few events from
* time to time while blocked into some not interruptible operation.
* This allows to reply to clients with the -LOADING error while loading the
* data set at startup or after a full resynchronization with the master
* and so forth.
*
* It calls the event loop in order to process a few events. Specifically we
* try to call the event loop 4 times as long as we receive acknowledge that
* some event was processed, in order to go forward with the accept, read,
* write, close sequence needed to serve a client.
*
* The function returns the total number of events processed. */
void processEventsWhileBlocked(void) {
int iterations = 4; /* See the function top-comment. */
/* Update our cached time since it is used to create and update the last
* interaction time with clients and for other important things. */
updateCachedTime(0);
/* Note: when we are processing events while blocked (for instance during
* busy Lua scripts), we set a global flag. When such flag is set, we
* avoid handling the read part of clients using threaded I/O.
* See https://github.com/redis/redis/issues/6988 for more info.
* Note that there could be cases of nested calls to this function,
* specifically on a busy script during async_loading rdb, and scripts
* that came from AOF. */
ProcessingEventsWhileBlocked++;
while (iterations--) {
long long startval = server.events_processed_while_blocked;
long long ae_events = aeProcessEvents(server.el,
AE_FILE_EVENTS|AE_DONT_WAIT|
AE_CALL_BEFORE_SLEEP|AE_CALL_AFTER_SLEEP);
/* Note that server.events_processed_while_blocked will also get
* incremented by callbacks called by the event loop handlers. */
server.events_processed_while_blocked += ae_events;
long long events = server.events_processed_while_blocked - startval;
if (!events) break;
}
whileBlockedCron();
ProcessingEventsWhileBlocked--;
serverAssert(ProcessingEventsWhileBlocked >= 0);
}
/* ==========================================================================
* Threaded I/O
* ========================================================================== */
#define IO_THREADS_MAX_NUM 128
pthread_t io_threads[IO_THREADS_MAX_NUM];
pthread_mutex_t io_threads_mutex[IO_THREADS_MAX_NUM];
redisAtomic unsigned long io_threads_pending[IO_THREADS_MAX_NUM];
int io_threads_op; /* IO_THREADS_OP_IDLE, IO_THREADS_OP_READ or IO_THREADS_OP_WRITE. */ // TODO: should access to this be atomic??!
/* This is the list of clients each thread will serve when threaded I/O is
* used. We spawn io_threads_num-1 threads, since one is the main thread
* itself. */
list *io_threads_list[IO_THREADS_MAX_NUM];
static inline unsigned long getIOPendingCount(int i) {
unsigned long count = 0;
atomicGetWithSync(io_threads_pending[i], count);
return count;
}
static inline void setIOPendingCount(int i, unsigned long count) {
atomicSetWithSync(io_threads_pending[i], count);
}
void *IOThreadMain(void *myid) {
/* The ID is the thread number (from 0 to server.iothreads_num-1), and is
* used by the thread to just manipulate a single sub-array of clients. */
long id = (unsigned long)myid;
char thdname[16];
snprintf(thdname, sizeof(thdname), "io_thd_%ld", id);
redis_set_thread_title(thdname);
redisSetCpuAffinity(server.server_cpulist);
makeThreadKillable();
while(1) {
/* Wait for start */
for (int j = 0; j < 1000000; j++) {
if (getIOPendingCount(id) != 0) break;
}
/* Give the main thread a chance to stop this thread. */
if (getIOPendingCount(id) == 0) {
pthread_mutex_lock(&io_threads_mutex[id]);
pthread_mutex_unlock(&io_threads_mutex[id]);
continue;
}
serverAssert(getIOPendingCount(id) != 0);
/* Process: note that the main thread will never touch our list
* before we drop the pending count to 0. */
listIter li;
listNode *ln;
listRewind(io_threads_list[id],&li);
while((ln = listNext(&li))) {
client *c = listNodeValue(ln);
if (io_threads_op == IO_THREADS_OP_WRITE) {
writeToClient(c,0);
} else if (io_threads_op == IO_THREADS_OP_READ) {
readQueryFromClient(c->conn);
} else {
serverPanic("io_threads_op value is unknown");
}
}
listEmpty(io_threads_list[id]);
setIOPendingCount(id, 0);
}
}
/* Initialize the data structures needed for threaded I/O. */
void initThreadedIO(void) {
server.io_threads_active = 0; /* We start with threads not active. */
/* Indicate that io-threads are currently idle */
io_threads_op = IO_THREADS_OP_IDLE;
/* Don't spawn any thread if the user selected a single thread:
* we'll handle I/O directly from the main thread. */
if (server.io_threads_num == 1) return;
if (server.io_threads_num > IO_THREADS_MAX_NUM) {
serverLog(LL_WARNING,"Fatal: too many I/O threads configured. "
"The maximum number is %d.", IO_THREADS_MAX_NUM);
exit(1);
}
/* Spawn and initialize the I/O threads. */
for (int i = 0; i < server.io_threads_num; i++) {
/* Things we do for all the threads including the main thread. */
io_threads_list[i] = listCreate();
if (i == 0) continue; /* Thread 0 is the main thread. */
/* Things we do only for the additional threads. */
pthread_t tid;
pthread_mutex_init(&io_threads_mutex[i],NULL);
setIOPendingCount(i, 0);
pthread_mutex_lock(&io_threads_mutex[i]); /* Thread will be stopped. */
if (pthread_create(&tid,NULL,IOThreadMain,(void*)(long)i) != 0) {
serverLog(LL_WARNING,"Fatal: Can't initialize IO thread.");
exit(1);
}
io_threads[i] = tid;
}
}
void killIOThreads(void) {
int err, j;
for (j = 0; j < server.io_threads_num; j++) {
if (io_threads[j] == pthread_self()) continue;
if (io_threads[j] && pthread_cancel(io_threads[j]) == 0) {
if ((err = pthread_join(io_threads[j],NULL)) != 0) {
serverLog(LL_WARNING,
"IO thread(tid:%lu) can not be joined: %s",
(unsigned long)io_threads[j], strerror(err));
} else {
serverLog(LL_WARNING,
"IO thread(tid:%lu) terminated",(unsigned long)io_threads[j]);
}
}
}
}
void startThreadedIO(void) {
serverAssert(server.io_threads_active == 0);
for (int j = 1; j < server.io_threads_num; j++)
pthread_mutex_unlock(&io_threads_mutex[j]);
server.io_threads_active = 1;
}
void stopThreadedIO(void) {
/* We may have still clients with pending reads when this function
* is called: handle them before stopping the threads. */
handleClientsWithPendingReadsUsingThreads();
serverAssert(server.io_threads_active == 1);
for (int j = 1; j < server.io_threads_num; j++)
pthread_mutex_lock(&io_threads_mutex[j]);
server.io_threads_active = 0;
}
/* This function checks if there are not enough pending clients to justify
* taking the I/O threads active: in that case I/O threads are stopped if
* currently active. We track the pending writes as a measure of clients
* we need to handle in parallel, however the I/O threading is disabled
* globally for reads as well if we have too little pending clients.
*
* The function returns 0 if the I/O threading should be used because there
* are enough active threads, otherwise 1 is returned and the I/O threads
* could be possibly stopped (if already active) as a side effect. */
int stopThreadedIOIfNeeded(void) {
int pending = listLength(server.clients_pending_write);
/* Return ASAP if IO threads are disabled (single threaded mode). */
if (server.io_threads_num == 1) return 1;
if (pending < (server.io_threads_num*2)) {
if (server.io_threads_active) stopThreadedIO();
return 1;
} else {
return 0;
}
}
/* This function achieves thread safety using a fan-out -> fan-in paradigm:
* Fan out: The main thread fans out work to the io-threads which block until
* setIOPendingCount() is called with a value larger than 0 by the main thread.
* Fan in: The main thread waits until getIOPendingCount() returns 0. Then
* it can safely perform post-processing and return to normal synchronous
* work. */
int handleClientsWithPendingWritesUsingThreads(void) {
int processed = listLength(server.clients_pending_write);
if (processed == 0) return 0; /* Return ASAP if there are no clients. */
/* If I/O threads are disabled or we have few clients to serve, don't
* use I/O threads, but the boring synchronous code. */
if (server.io_threads_num == 1 || stopThreadedIOIfNeeded()) {
return handleClientsWithPendingWrites();
}
/* Start threads if needed. */
if (!server.io_threads_active) startThreadedIO();
/* Distribute the clients across N different lists. */
listIter li;
listNode *ln;
listRewind(server.clients_pending_write,&li);
int item_id = 0;
while((ln = listNext(&li))) {
client *c = listNodeValue(ln);
c->flags &= ~CLIENT_PENDING_WRITE;
/* Remove clients from the list of pending writes since
* they are going to be closed ASAP. */
if (c->flags & CLIENT_CLOSE_ASAP) {
listDelNode(server.clients_pending_write, ln);
continue;
}
/* Since all replicas and replication backlog use global replication
* buffer, to guarantee data accessing thread safe, we must put all
* replicas client into io_threads_list[0] i.e. main thread handles
* sending the output buffer of all replicas. */
if (getClientType(c) == CLIENT_TYPE_SLAVE) {
listAddNodeTail(io_threads_list[0],c);
continue;
}
int target_id = item_id % server.io_threads_num;
listAddNodeTail(io_threads_list[target_id],c);
item_id++;
}
/* Give the start condition to the waiting threads, by setting the
* start condition atomic var. */
io_threads_op = IO_THREADS_OP_WRITE;
for (int j = 1; j < server.io_threads_num; j++) {
int count = listLength(io_threads_list[j]);
setIOPendingCount(j, count);
}
/* Also use the main thread to process a slice of clients. */
listRewind(io_threads_list[0],&li);
while((ln = listNext(&li))) {
client *c = listNodeValue(ln);
writeToClient(c,0);
}
listEmpty(io_threads_list[0]);
/* Wait for all the other threads to end their work. */
while(1) {
unsigned long pending = 0;
for (int j = 1; j < server.io_threads_num; j++)
pending += getIOPendingCount(j);
if (pending == 0) break;
}
io_threads_op = IO_THREADS_OP_IDLE;
/* Run the list of clients again to install the write handler where
* needed. */
listRewind(server.clients_pending_write,&li);
while((ln = listNext(&li))) {
client *c = listNodeValue(ln);
/* Update the client in the mem usage buckets after we're done processing it in the io-threads */
updateClientMemUsageBucket(c);
/* Install the write handler if there are pending writes in some
* of the clients. */
if (clientHasPendingReplies(c) &&
connSetWriteHandler(c->conn, sendReplyToClient) == AE_ERR)
{
freeClientAsync(c);
}
}
listEmpty(server.clients_pending_write);
/* Update processed count on server */
server.stat_io_writes_processed += processed;
return processed;
}
/* Return 1 if we want to handle the client read later using threaded I/O.
* This is called by the readable handler of the event loop.
* As a side effect of calling this function the client is put in the
* pending read clients and flagged as such. */
int postponeClientRead(client *c) {
if (server.io_threads_active &&
server.io_threads_do_reads &&
!ProcessingEventsWhileBlocked &&
!(c->flags & (CLIENT_MASTER|CLIENT_SLAVE|CLIENT_BLOCKED)) &&
io_threads_op == IO_THREADS_OP_IDLE)
{
listAddNodeHead(server.clients_pending_read,c);
c->pending_read_list_node = listFirst(server.clients_pending_read);
return 1;
} else {
return 0;
}
}
/* When threaded I/O is also enabled for the reading + parsing side, the
* readable handler will just put normal clients into a queue of clients to
* process (instead of serving them synchronously). This function runs
* the queue using the I/O threads, and process them in order to accumulate
* the reads in the buffers, and also parse the first command available
* rendering it in the client structures.
* This function achieves thread safety using a fan-out -> fan-in paradigm:
* Fan out: The main thread fans out work to the io-threads which block until
* setIOPendingCount() is called with a value larger than 0 by the main thread.
* Fan in: The main thread waits until getIOPendingCount() returns 0. Then
* it can safely perform post-processing and return to normal synchronous
* work. */
int handleClientsWithPendingReadsUsingThreads(void) {
if (!server.io_threads_active || !server.io_threads_do_reads) return 0;
int processed = listLength(server.clients_pending_read);
if (processed == 0) return 0;
/* Distribute the clients across N different lists. */
listIter li;
listNode *ln;
listRewind(server.clients_pending_read,&li);
int item_id = 0;
while((ln = listNext(&li))) {
client *c = listNodeValue(ln);
int target_id = item_id % server.io_threads_num;
listAddNodeTail(io_threads_list[target_id],c);
item_id++;
}
/* Give the start condition to the waiting threads, by setting the
* start condition atomic var. */
io_threads_op = IO_THREADS_OP_READ;
for (int j = 1; j < server.io_threads_num; j++) {
int count = listLength(io_threads_list[j]);
setIOPendingCount(j, count);
}
/* Also use the main thread to process a slice of clients. */
listRewind(io_threads_list[0],&li);
while((ln = listNext(&li))) {
client *c = listNodeValue(ln);
readQueryFromClient(c->conn);
}
listEmpty(io_threads_list[0]);
/* Wait for all the other threads to end their work. */
while(1) {
unsigned long pending = 0;
for (int j = 1; j < server.io_threads_num; j++)
pending += getIOPendingCount(j);
if (pending == 0) break;
}
io_threads_op = IO_THREADS_OP_IDLE;
/* Run the list of clients again to process the new buffers. */
while(listLength(server.clients_pending_read)) {
ln = listFirst(server.clients_pending_read);
client *c = listNodeValue(ln);
listDelNode(server.clients_pending_read,ln);
c->pending_read_list_node = NULL;
serverAssert(!(c->flags & CLIENT_BLOCKED));
if (beforeNextClient(c) == C_ERR) {
/* If the client is no longer valid, we avoid
* processing the client later. So we just go
* to the next. */
continue;
}
/* Once io-threads are idle we can update the client in the mem usage buckets */
updateClientMemUsageBucket(c);
if (processPendingCommandsAndResetClient(c) == C_ERR) {
/* If the client is no longer valid, we avoid
* processing the client later. So we just go
* to the next. */
continue;
}
if (processInputBuffer(c) == C_ERR) {
/* If the client is no longer valid, we avoid
* processing the client later. So we just go
* to the next. */
continue;
}
/* We may have pending replies if a thread readQueryFromClient() produced
* replies and did not install a write handler (it can't).
*/
if (!(c->flags & CLIENT_PENDING_WRITE) && clientHasPendingReplies(c))
clientInstallWriteHandler(c);
}
/* Update processed count on server */
server.stat_io_reads_processed += processed;
return processed;
}
/* Returns the actual client eviction limit based on current configuration or
* 0 if no limit. */
size_t getClientEvictionLimit(void) {
size_t maxmemory_clients_actual = SIZE_MAX;
/* Handle percentage of maxmemory*/
if (server.maxmemory_clients < 0 && server.maxmemory > 0) {
unsigned long long maxmemory_clients_bytes = (unsigned long long)((double)server.maxmemory * -(double) server.maxmemory_clients / 100);
if (maxmemory_clients_bytes <= SIZE_MAX)
maxmemory_clients_actual = maxmemory_clients_bytes;
}
else if (server.maxmemory_clients > 0)
maxmemory_clients_actual = server.maxmemory_clients;
else
return 0;
/* Don't allow a too small maxmemory-clients to avoid cases where we can't communicate
* at all with the server because of bad configuration */
if (maxmemory_clients_actual < 1024*128)
maxmemory_clients_actual = 1024*128;
return maxmemory_clients_actual;
}
void evictClients(void) {
/* Start eviction from topmost bucket (largest clients) */
int curr_bucket = CLIENT_MEM_USAGE_BUCKETS-1;
listIter bucket_iter;
listRewind(server.client_mem_usage_buckets[curr_bucket].clients, &bucket_iter);
size_t client_eviction_limit = getClientEvictionLimit();
if (client_eviction_limit == 0)
return;
while (server.stat_clients_type_memory[CLIENT_TYPE_NORMAL] +
server.stat_clients_type_memory[CLIENT_TYPE_PUBSUB] >= client_eviction_limit) {
listNode *ln = listNext(&bucket_iter);
if (ln) {
client *c = ln->value;
sds ci = catClientInfoString(sdsempty(),c);
serverLog(LL_NOTICE, "Evicting client: %s", ci);
freeClient(c);
sdsfree(ci);
server.stat_evictedclients++;
} else {
curr_bucket--;
if (curr_bucket < 0) {
serverLog(LL_WARNING, "Over client maxmemory after evicting all evictable clients");
break;
}
listRewind(server.client_mem_usage_buckets[curr_bucket].clients, &bucket_iter);
}
}
}
|