summaryrefslogtreecommitdiffstats
path: root/rubbos/app/httpd-2.0.64/server/mpm/experimental/threadpool/threadpool.c
blob: 50de500e6a8904c7c9ff901ccf3c1875e538a46d (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
2073
2074
2075
2076
2077
2078
2079
2080
2081
2082
2083
2084
2085
2086
2087
2088
2089
2090
2091
2092
2093
2094
2095
2096
2097
2098
2099
2100
2101
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
2128
2129
2130
2131
2132
2133
2134
2135
2136
2137
2138
2139
2140
2141
2142
2143
2144
2145
2146
2147
2148
2149
2150
2151
2152
2153
2154
2155
2156
2157
2158
2159
2160
2161
2162
2163
2164
2165
2166
2167
2168
2169
2170
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
2184
2185
2186
2187
2188
2189
2190
2191
2192
2193
2194
2195
2196
2197
2198
2199
2200
2201
2202
2203
2204
2205
2206
2207
2208
2209
2210
2211
2212
2213
2214
2215
2216
2217
2218
2219
2220
2221
2222
2223
2224
2225
2226
2227
2228
2229
/* Licensed to the Apache Software Foundation (ASF) under one or more
 * contributor license agreements.  See the NOTICE file distributed with
 * this work for additional information regarding copyright ownership.
 * The ASF licenses this file to You under the Apache License, Version 2.0
 * (the "License"); you may not use this file except in compliance with
 * the License.  You may obtain a copy of the License at
 *
 *     http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */

/* The purpose of this MPM is to fix the design flaws in the threaded
 * model.  Because of the way that pthreads and mutex locks interact,
 * it is basically impossible to cleanly gracefully shutdown a child
 * process if multiple threads are all blocked in accept.  This model
 * fixes those problems.
 */

#include "apr.h"
#include "apr_portable.h"
#include "apr_strings.h"
#include "apr_file_io.h"
#include "apr_thread_proc.h"
#include "apr_signal.h"
#include "apr_poll.h"
#include "apr_thread_mutex.h"
#include "apr_thread_cond.h"
#include "apr_proc_mutex.h"
#define APR_WANT_STRFUNC
#include "apr_want.h"

#if APR_HAVE_UNISTD_H
#include <unistd.h>
#endif
#if APR_HAVE_SYS_SOCKET_H
#include <sys/socket.h>
#endif
#if APR_HAVE_SYS_WAIT_H
#include <sys/wait.h> 
#endif
#ifdef HAVE_SYS_PROCESSOR_H
#include <sys/processor.h> /* for bindprocessor() */
#endif

#if !APR_HAS_THREADS
#error The Worker MPM requires APR threads, but they are unavailable.
#endif

#define CORE_PRIVATE 
 
#include "ap_config.h"
#include "httpd.h" 
#include "http_main.h" 
#include "http_log.h" 
#include "http_config.h"        /* for read_config */ 
#include "http_core.h"          /* for get_remote_host */ 
#include "http_connection.h"
#include "ap_mpm.h"
#include "pod.h"
#include "mpm_common.h"
#include "ap_listen.h"
#include "scoreboard.h" 
#include "mpm_default.h"

#include <signal.h>
#include <limits.h>             /* for INT_MAX */

/* Limit on the total --- clients will be locked out if more servers than
 * this are needed.  It is intended solely to keep the server from crashing
 * when things get out of hand.
 *
 * We keep a hard maximum number of servers, for two reasons --- first off,
 * in case something goes seriously wrong, we want to stop the fork bomb
 * short of actually crashing the machine we're running on by filling some
 * kernel table.  Secondly, it keeps the size of the scoreboard file small
 * enough that we can read the whole thing without worrying too much about
 * the overhead.
 */
#ifndef DEFAULT_SERVER_LIMIT
#define DEFAULT_SERVER_LIMIT 16
#endif

/* Admin can't tune ServerLimit beyond MAX_SERVER_LIMIT.  We want
 * some sort of compile-time limit to help catch typos.
 */
#ifndef MAX_SERVER_LIMIT
#define MAX_SERVER_LIMIT 20000
#endif

/* Limit on the threads per process.  Clients will be locked out if more than
 * this  * server_limit are needed.
 *
 * We keep this for one reason it keeps the size of the scoreboard file small
 * enough that we can read the whole thing without worrying too much about
 * the overhead.
 */
#ifndef DEFAULT_THREAD_LIMIT
#define DEFAULT_THREAD_LIMIT 64 
#endif

/* Admin can't tune ThreadLimit beyond MAX_THREAD_LIMIT.  We want
 * some sort of compile-time limit to help catch typos.
 */
#ifndef MAX_THREAD_LIMIT
#define MAX_THREAD_LIMIT 20000
#endif

/*
 * Actual definitions of config globals
 */

int ap_threads_per_child = 0;         /* Worker threads per child */
static int ap_daemons_to_start = 0;
static int min_spare_threads = 0;
static int max_spare_threads = 0;
static int ap_daemons_limit = 0;
static int server_limit = DEFAULT_SERVER_LIMIT;
static int first_server_limit;
static int thread_limit = DEFAULT_THREAD_LIMIT;
static int first_thread_limit;
static int changed_limit_at_restart;
static int dying = 0;
static int workers_may_exit = 0;
static int start_thread_may_exit = 0;
static int listener_may_exit = 0;
static int requests_this_child;
static int num_listensocks = 0;
static int resource_shortage = 0;
static int mpm_state = AP_MPMQ_STARTING;

/* The structure used to pass unique initialization info to each thread */
typedef struct {
    int pid;
    int tid;
    int sd;
} proc_info;

/* Structure used to pass information to the thread responsible for 
 * creating the rest of the threads.
 */
typedef struct {
    apr_thread_t **threads;
    apr_thread_t *listener;
    int child_num_arg;
    apr_threadattr_t *threadattr;
} thread_starter;

#define ID_FROM_CHILD_THREAD(c, t)    ((c * thread_limit) + t)

/*
 * The max child slot ever assigned, preserved across restarts.  Necessary
 * to deal with MaxClients changes across AP_SIG_GRACEFUL restarts.  We 
 * use this value to optimize routines that have to scan the entire 
 * scoreboard.
 */
int ap_max_daemons_limit = -1;

static ap_pod_t *pod;

/* *Non*-shared http_main globals... */

server_rec *ap_server_conf;

/* The worker MPM respects a couple of runtime flags that can aid
 * in debugging. Setting the -DNO_DETACH flag will prevent the root process
 * from detaching from its controlling terminal. Additionally, setting
 * the -DONE_PROCESS flag (which implies -DNO_DETACH) will get you the
 * child_main loop running in the process which originally started up.
 * This gives you a pretty nice debugging environment.  (You'll get a SIGHUP
 * early in standalone_main; just continue through.  This is the server
 * trying to kill off any child processes which it might have lying
 * around --- Apache doesn't keep track of their pids, it just sends
 * SIGHUP to the process group, ignoring it in the root process.
 * Continue through and you'll be fine.).
 */

static int one_process = 0;

#ifdef DEBUG_SIGSTOP
int raise_sigstop_flags;
#endif

static apr_pool_t *pconf;                 /* Pool for config stuff */
static apr_pool_t *pchild;                /* Pool for httpd child stuff */

static pid_t ap_my_pid; /* Linux getpid() doesn't work except in main 
                           thread. Use this instead */
static pid_t parent_pid;
static apr_os_thread_t *listener_os_thread;

/* Locks for accept serialization */
static apr_proc_mutex_t *accept_mutex;

#if APR_O_NONBLOCK_INHERITED
#undef SINGLE_LISTEN_UNSERIALIZED_ACCEPT
#endif /* APR_O_NONBLOCK_INHERITED */

#ifdef SINGLE_LISTEN_UNSERIALIZED_ACCEPT
#define SAFE_ACCEPT(stmt) (ap_listeners->next ? (stmt) : APR_SUCCESS)
#else
#define SAFE_ACCEPT(stmt) (stmt)
#endif

/* The LISTENER_SIGNAL signal will be sent from the main thread to the 
 * listener thread to wake it up for graceful termination (what a child 
 * process from an old generation does when the admin does "apachectl 
 * graceful").  This signal will be blocked in all threads of a child
 * process except for the listener thread.
 */
#define LISTENER_SIGNAL     SIGHUP


/* Possible states of a worker thread. */
typedef enum {
    WORKER_IDLE,
    WORKER_BUSY,
    WORKER_TERMINATED
} worker_state_e;

/* Structure used to wake up an idle worker thread
 */
typedef struct {
    apr_pool_t *pool;
    apr_socket_t *csd;
    worker_state_e state;
    apr_thread_cond_t *cond;
    apr_thread_mutex_t *mutex;
} worker_wakeup_info;

/* Structure used to hold a stack of idle worker threads 
 */
typedef struct {
    apr_thread_mutex_t *mutex;
    apr_thread_cond_t *cond;
    worker_wakeup_info **stack;
    apr_size_t nelts;
    apr_size_t nalloc;
    int terminated;
} worker_stack;

static worker_stack* worker_stack_create(apr_pool_t *pool, apr_size_t max)
{
    apr_status_t rv;
    worker_stack *stack = (worker_stack *)apr_palloc(pool, sizeof(*stack));

    if ((rv = apr_thread_mutex_create(&stack->mutex, APR_THREAD_MUTEX_DEFAULT,
                                      pool)) != APR_SUCCESS) {
        return NULL;
    }
    if ((rv = apr_thread_cond_create(&stack->cond, pool)) != APR_SUCCESS) {
        return NULL;
    }
    stack->nelts = 0;
    stack->nalloc = max;
    stack->stack =
        (worker_wakeup_info **)apr_palloc(pool, stack->nalloc *
                                          sizeof(worker_wakeup_info *));
    stack->terminated = 0;
    return stack;
}

static apr_status_t worker_stack_wait(worker_stack *stack,
                                      worker_wakeup_info *wakeup)
{
    apr_status_t rv;

    wakeup->state = WORKER_IDLE;
    
    if ((rv = apr_thread_mutex_lock(stack->mutex)) != APR_SUCCESS) {
        return rv;
    }
    if (stack->terminated) {
        if ((rv = apr_thread_mutex_unlock(stack->mutex)) != APR_SUCCESS) {
            return rv;
        }
        return APR_EOF;
    }
    if (stack->nelts == stack->nalloc) {
        if ((rv = apr_thread_mutex_unlock(stack->mutex)) != APR_SUCCESS) {
            return rv;
        }
        return APR_ENOSPC;
    }
    stack->stack[stack->nelts] = wakeup;
    /* Signal a blocking listener thread only if we just made the
     * stack non-empty. */
    if (stack->nelts++ == 0) {
        (void)apr_thread_cond_signal(stack->cond);
    }
    if ((rv = apr_thread_mutex_unlock(stack->mutex)) != APR_SUCCESS) {
        return rv;
    }

    /* At this point we've already added this worker to the stack, now
     * we just wait until the listener has accept()ed a connection
     * for us. */
    if ((rv = apr_thread_mutex_lock(wakeup->mutex)) != APR_SUCCESS) {
        return rv;
    }
    while (wakeup->state == WORKER_IDLE) {
        if ((rv = apr_thread_cond_wait(wakeup->cond, wakeup->mutex)) !=
            APR_SUCCESS) {
            return rv;
        }
    }
    if ((rv = apr_thread_mutex_unlock(wakeup->mutex)) != APR_SUCCESS) {
        return rv;
    }
    return APR_SUCCESS;
}

static apr_status_t worker_stack_pop(worker_stack *stack,
                                     worker_wakeup_info **worker)
{
    apr_status_t rv;
    if ((rv = apr_thread_mutex_lock(stack->mutex)) != APR_SUCCESS) {
        return rv;
    }
    AP_DEBUG_ASSERT(stack->nelts >= 0);
    while ((stack->nelts == 0) && (!stack->terminated)) {
        rv = apr_thread_cond_wait(stack->cond, stack->mutex);
        if (rv != APR_SUCCESS) {
            apr_status_t rv2;
            rv2 = apr_thread_mutex_unlock(stack->mutex);
            if (rv2 != APR_SUCCESS) {
                return rv2;
            }
            return rv;
        }
    }
    if (stack->terminated) {
        if ((rv = apr_thread_mutex_unlock(stack->mutex)) != APR_SUCCESS) {
            return rv;
        }
        return APR_EOF;
    }
    *worker = stack->stack[--stack->nelts];
    if ((rv = apr_thread_mutex_unlock(stack->mutex)) != APR_SUCCESS) {
        return rv;
    }
    return APR_SUCCESS;
}

static apr_status_t worker_stack_terminate(worker_stack *stack)
{
    apr_status_t rv;
    worker_wakeup_info *worker;

    if ((rv = apr_thread_mutex_lock(stack->mutex)) != APR_SUCCESS) {
        return rv;
    }
    stack->terminated = 1;
    /* Wake up the listener thread. Although there will never be
     * more than one thread blocking on this condition, broadcast
     * just in case. */
    apr_thread_cond_broadcast(stack->cond);
    while (stack->nelts) {
        worker = stack->stack[--stack->nelts];
        apr_thread_mutex_lock(worker->mutex);
        worker->csd = 0;
        worker->state = WORKER_TERMINATED;
        apr_thread_cond_signal(worker->cond);
        apr_thread_mutex_unlock(worker->mutex);
    }
    if ((rv = apr_thread_mutex_unlock(stack->mutex)) != APR_SUCCESS) {
        return rv;
    }
    return APR_SUCCESS;
}

static worker_stack *idle_worker_stack;

static void wakeup_listener(void)
{
    apr_status_t rv;

    listener_may_exit = 1;
    if (!idle_worker_stack) {
        return;
    }
    if ((rv = apr_thread_mutex_lock(idle_worker_stack->mutex)) != APR_SUCCESS) {
        return;
    }
    if ((rv = apr_thread_cond_signal(idle_worker_stack->cond)) !=
        APR_SUCCESS) {
        return;
    }
    if ((rv = apr_thread_mutex_unlock(idle_worker_stack->mutex)) != APR_SUCCESS) {
        return;
    }
    if (!listener_os_thread) {
        /* XXX there is an obscure path that this doesn't handle perfectly:
         *     right after listener thread is created but before 
         *     listener_os_thread is set, the first worker thread hits an
         *     error and starts graceful termination
         */
        return;
    }
    /*
     * we should just be able to "kill(ap_my_pid, LISTENER_SIGNAL)" on all
     * platforms and wake up the listener thread since it is the only thread 
     * with SIGHUP unblocked, but that doesn't work on Linux
     */
#ifdef HAVE_PTHREAD_KILL
    pthread_kill(*listener_os_thread, LISTENER_SIGNAL);
#else
    kill(ap_my_pid, LISTENER_SIGNAL);
#endif
}

#define ST_INIT              0
#define ST_GRACEFUL          1
#define ST_UNGRACEFUL        2

static int terminate_mode = ST_INIT;

static void signal_threads(int mode)
{
    if (terminate_mode == mode) {
        return;
    }
    terminate_mode = mode;
    mpm_state = AP_MPMQ_STOPPING;

    /* in case we weren't called from the listener thread, wake up the
     * listener thread
     */
    wakeup_listener();

    /* for ungraceful termination, let the workers exit now;
     * for graceful termination, the listener thread will notify the
     * workers to exit once it has stopped accepting new connections
     */
    if (mode == ST_UNGRACEFUL) {
        workers_may_exit = 1;
        worker_stack_terminate(idle_worker_stack);
    }
}

AP_DECLARE(apr_status_t) ap_mpm_query(int query_code, int *result)
{
    switch(query_code){
        case AP_MPMQ_MAX_DAEMON_USED:
            *result = ap_max_daemons_limit;
            return APR_SUCCESS;
        case AP_MPMQ_IS_THREADED:
            *result = AP_MPMQ_STATIC;
            return APR_SUCCESS;
        case AP_MPMQ_IS_FORKED:
            *result = AP_MPMQ_DYNAMIC;
            return APR_SUCCESS;
        case AP_MPMQ_HARD_LIMIT_DAEMONS:
            *result = server_limit;
            return APR_SUCCESS;
        case AP_MPMQ_HARD_LIMIT_THREADS:
            *result = thread_limit;
            return APR_SUCCESS;
        case AP_MPMQ_MAX_THREADS:
            *result = ap_threads_per_child;
            return APR_SUCCESS;
        case AP_MPMQ_MIN_SPARE_DAEMONS:
            *result = 0;
            return APR_SUCCESS;
        case AP_MPMQ_MIN_SPARE_THREADS:    
            *result = min_spare_threads;
            return APR_SUCCESS;
        case AP_MPMQ_MAX_SPARE_DAEMONS:
            *result = 0;
            return APR_SUCCESS;
        case AP_MPMQ_MAX_SPARE_THREADS:
            *result = max_spare_threads;
            return APR_SUCCESS;
        case AP_MPMQ_MAX_REQUESTS_DAEMON:
            *result = ap_max_requests_per_child;
            return APR_SUCCESS;
        case AP_MPMQ_MAX_DAEMONS:
            *result = ap_daemons_limit;
            return APR_SUCCESS;
        case AP_MPMQ_MPM_STATE:
            *result = mpm_state;
            return APR_SUCCESS;
    }
    return APR_ENOTIMPL;
}

/* a clean exit from a child with proper cleanup */ 
static void clean_child_exit(int code) __attribute__ ((noreturn));
static void clean_child_exit(int code)
{
    mpm_state = AP_MPMQ_STOPPING;
    if (pchild) {
        apr_pool_destroy(pchild);
    }
    exit(code);
}

static void just_die(int sig)
{
    clean_child_exit(0);
}

/*****************************************************************
 * Connection structures and accounting...
 */

/* volatile just in case */
static int volatile shutdown_pending;
static int volatile restart_pending;
static int volatile is_graceful;
static volatile int child_fatal;
ap_generation_t volatile ap_my_generation;

/*
 * ap_start_shutdown() and ap_start_restart(), below, are a first stab at
 * functions to initiate shutdown or restart without relying on signals. 
 * Previously this was initiated in sig_term() and restart() signal handlers, 
 * but we want to be able to start a shutdown/restart from other sources --
 * e.g. on Win32, from the service manager. Now the service manager can
 * call ap_start_shutdown() or ap_start_restart() as appropiate.  Note that
 * these functions can also be called by the child processes, since global
 * variables are no longer used to pass on the required action to the parent.
 *
 * These should only be called from the parent process itself, since the
 * parent process will use the shutdown_pending and restart_pending variables
 * to determine whether to shutdown or restart. The child process should
 * call signal_parent() directly to tell the parent to die -- this will
 * cause neither of those variable to be set, which the parent will
 * assume means something serious is wrong (which it will be, for the
 * child to force an exit) and so do an exit anyway.
 */

static void ap_start_shutdown(void)
{
    mpm_state = AP_MPMQ_STOPPING;
    if (shutdown_pending == 1) {
        /* Um, is this _probably_ not an error, if the user has
         * tried to do a shutdown twice quickly, so we won't
         * worry about reporting it.
         */
        return;
    }
    shutdown_pending = 1;
}

/* do a graceful restart if graceful == 1 */
static void ap_start_restart(int graceful)
{
    mpm_state = AP_MPMQ_STOPPING;
    if (restart_pending == 1) {
        /* Probably not an error - don't bother reporting it */
        return;
    }
    restart_pending = 1;
    is_graceful = graceful;
}

static void sig_term(int sig)
{
    ap_start_shutdown();
}

static void restart(int sig)
{
    ap_start_restart(sig == AP_SIG_GRACEFUL);
}

static void set_signals(void)
{
#ifndef NO_USE_SIGACTION
    struct sigaction sa;
#endif

    if (!one_process) {
        ap_fatal_signal_setup(ap_server_conf, pconf);
    }

#ifndef NO_USE_SIGACTION
    sigemptyset(&sa.sa_mask);
    sa.sa_flags = 0;

    sa.sa_handler = sig_term;
    if (sigaction(SIGTERM, &sa, NULL) < 0)
        ap_log_error(APLOG_MARK, APLOG_WARNING, errno, ap_server_conf, 
                     "sigaction(SIGTERM)");
#ifdef SIGINT
    if (sigaction(SIGINT, &sa, NULL) < 0)
        ap_log_error(APLOG_MARK, APLOG_WARNING, errno, ap_server_conf, 
                     "sigaction(SIGINT)");
#endif
#ifdef SIGXCPU
    sa.sa_handler = SIG_DFL;
    if (sigaction(SIGXCPU, &sa, NULL) < 0)
        ap_log_error(APLOG_MARK, APLOG_WARNING, errno, ap_server_conf, 
                     "sigaction(SIGXCPU)");
#endif
#ifdef SIGXFSZ
    sa.sa_handler = SIG_DFL;
    if (sigaction(SIGXFSZ, &sa, NULL) < 0)
        ap_log_error(APLOG_MARK, APLOG_WARNING, errno, ap_server_conf, 
                     "sigaction(SIGXFSZ)");
#endif
#ifdef SIGPIPE
    sa.sa_handler = SIG_IGN;
    if (sigaction(SIGPIPE, &sa, NULL) < 0)
        ap_log_error(APLOG_MARK, APLOG_WARNING, errno, ap_server_conf, 
                     "sigaction(SIGPIPE)");
#endif

    /* we want to ignore HUPs and AP_SIG_GRACEFUL while we're busy 
     * processing one */
    sigaddset(&sa.sa_mask, SIGHUP);
    sigaddset(&sa.sa_mask, AP_SIG_GRACEFUL);
    sa.sa_handler = restart;
    if (sigaction(SIGHUP, &sa, NULL) < 0)
        ap_log_error(APLOG_MARK, APLOG_WARNING, errno, ap_server_conf, 
                     "sigaction(SIGHUP)");
    if (sigaction(AP_SIG_GRACEFUL, &sa, NULL) < 0)
        ap_log_error(APLOG_MARK, APLOG_WARNING, errno, ap_server_conf, 
                     "sigaction(" AP_SIG_GRACEFUL_STRING ")");
#else
    if (!one_process) {
#ifdef SIGXCPU
        apr_signal(SIGXCPU, SIG_DFL);
#endif /* SIGXCPU */
#ifdef SIGXFSZ
        apr_signal(SIGXFSZ, SIG_DFL);
#endif /* SIGXFSZ */
    }

    apr_signal(SIGTERM, sig_term);
#ifdef SIGHUP
    apr_signal(SIGHUP, restart);
#endif /* SIGHUP */
#ifdef AP_SIG_GRACEFUL
    apr_signal(AP_SIG_GRACEFUL, restart);
#endif /* AP_SIG_GRACEFUL */
#ifdef SIGPIPE
    apr_signal(SIGPIPE, SIG_IGN);
#endif /* SIGPIPE */

#endif
}

/*****************************************************************
 * Here follows a long bunch of generic server bookkeeping stuff...
 */

int ap_graceful_stop_signalled(void)
    /* XXX this is really a bad confusing obsolete name
     * maybe it should be ap_mpm_process_exiting?
     */
{
    /* note: for a graceful termination, listener_may_exit will be set before
     *       workers_may_exit, so check listener_may_exit
     */
    return listener_may_exit;
}

/*****************************************************************
 * Child process main loop.
 */

static void process_socket(apr_pool_t *p, apr_socket_t *sock, int my_child_num,
                           int my_thread_num, apr_bucket_alloc_t *bucket_alloc)
{
    conn_rec *current_conn;
    long conn_id = ID_FROM_CHILD_THREAD(my_child_num, my_thread_num);
    int csd;
    ap_sb_handle_t *sbh;

    ap_create_sb_handle(&sbh, p, my_child_num, my_thread_num);
    apr_os_sock_get(&csd, sock);

    current_conn = ap_run_create_connection(p, ap_server_conf, sock,
                                            conn_id, sbh, bucket_alloc);
    if (current_conn) {
        ap_process_connection(current_conn, sock);
        ap_lingering_close(current_conn);
    }
}

/* requests_this_child has gone to zero or below.  See if the admin coded
   "MaxRequestsPerChild 0", and keep going in that case.  Doing it this way
   simplifies the hot path in worker_thread */
static void check_infinite_requests(void)
{
    if (ap_max_requests_per_child) {
        signal_threads(ST_GRACEFUL);
    }
    else {
        /* wow! if you're executing this code, you may have set a record.
         * either this child process has served over 2 billion requests, or
         * you're running a threaded 2.0 on a 16 bit machine.  
         *
         * I'll buy pizza and beers at Apachecon for the first person to do
         * the former without cheating (dorking with INT_MAX, or running with
         * uncommitted performance patches, for example).    
         *
         * for the latter case, you probably deserve a beer too.   Greg Ames
         */
            
        requests_this_child = INT_MAX;      /* keep going */ 
    }
}

static void unblock_signal(int sig)
{
    sigset_t sig_mask;

    sigemptyset(&sig_mask);
    sigaddset(&sig_mask, sig);
#if defined(SIGPROCMASK_SETS_THREAD_MASK)
    sigprocmask(SIG_UNBLOCK, &sig_mask, NULL);
#else
    pthread_sigmask(SIG_UNBLOCK, &sig_mask, NULL);
#endif
}

static void dummy_signal_handler(int sig)
{
    /* XXX If specifying SIG_IGN is guaranteed to unblock a syscall,
     *     then we don't need this goofy function.
     */
}

static void *listener_thread(apr_thread_t *thd, void * dummy)
{
    proc_info * ti = dummy;
    int process_slot = ti->pid;
    apr_pool_t *tpool = apr_thread_pool_get(thd);
    void *csd = NULL;
    apr_pool_t *ptrans;                /* Pool for per-transaction stuff */
    int n;
    apr_pollfd_t *pollset;
    apr_status_t rv;
    ap_listen_rec *lr, *last_lr = ap_listeners;
    worker_wakeup_info *worker = NULL;

    free(ti);

    apr_poll_setup(&pollset, num_listensocks, tpool);
    for(lr = ap_listeners ; lr != NULL ; lr = lr->next)
        apr_poll_socket_add(pollset, lr->sd, APR_POLLIN);

    /* Unblock the signal used to wake this thread up, and set a handler for
     * it.
     */
    unblock_signal(LISTENER_SIGNAL);
    apr_signal(LISTENER_SIGNAL, dummy_signal_handler);

    /* TODO: Switch to a system where threads reuse the results from earlier
       poll calls - manoj */
    while (1) {
        /* TODO: requests_this_child should be synchronized - aaron */
        if (requests_this_child <= 0) {
            check_infinite_requests();
        }
        if (listener_may_exit) break;

        if (worker == NULL) {
            rv = worker_stack_pop(idle_worker_stack, &worker);
            if (APR_STATUS_IS_EOF(rv)) {
                break;
            }
            else if (rv != APR_SUCCESS) {
                ap_log_error(APLOG_MARK, APLOG_CRIT, rv, ap_server_conf,
                             "worker_stack_pop failed");
                break;
            }
            ptrans = worker->pool;
        }
        AP_DEBUG_ASSERT(worker->state == WORKER_IDLE);

        if ((rv = SAFE_ACCEPT(apr_proc_mutex_lock(accept_mutex)))
            != APR_SUCCESS) {
            int level = APLOG_EMERG;

            if (listener_may_exit) {
                break;
            }
            if (ap_scoreboard_image->parent[process_slot].generation != 
                ap_scoreboard_image->global->running_generation) {
                level = APLOG_DEBUG; /* common to get these at restart time */
            }
            ap_log_error(APLOG_MARK, level, rv, ap_server_conf,
                         "apr_proc_mutex_lock failed. Attempting to shutdown "
                         "process gracefully.");
            signal_threads(ST_GRACEFUL);
            break;                    /* skip the lock release */
        }

        if (!APR_O_NONBLOCK_INHERITED && !ap_listeners->next) {
            /* Only one listener, so skip the poll */
            lr = ap_listeners;
        }
        else {
            while (!listener_may_exit) {
                apr_status_t ret;
                apr_int16_t event;

                ret = apr_poll(pollset, num_listensocks, &n, -1);
                if (ret != APR_SUCCESS) {
                    if (APR_STATUS_IS_EINTR(ret)) {
                        continue;
                    }

                    /* apr_poll() will only return errors in catastrophic
                     * circumstances. Let's try exiting gracefully, for now. */
                    ap_log_error(APLOG_MARK, APLOG_ERR, ret, (const server_rec *)
                                 ap_server_conf, "apr_poll: (listen)");
                    signal_threads(ST_GRACEFUL);
                }

                if (listener_may_exit) break;

                /* find a listener */
                lr = last_lr;
                do {
                    lr = lr->next;
                    if (lr == NULL) {
                        lr = ap_listeners;
                    }
                    /* XXX: Should we check for POLLERR? */
                    apr_poll_revents_get(&event, lr->sd, pollset);
                    if (event & APR_POLLIN) {
                        last_lr = lr;
                        goto got_fd;
                    }
                } while (lr != last_lr);
            }
        }
    got_fd:
        if (!listener_may_exit) {
            rv = lr->accept_func(&csd, lr, ptrans);
            /* later we trash rv and rely on csd to indicate success/failure */
            AP_DEBUG_ASSERT(rv == APR_SUCCESS || !csd);

            if (rv == APR_EGENERAL) {
                /* E[NM]FILE, ENOMEM, etc */
                resource_shortage = 1;
                signal_threads(ST_GRACEFUL);
            }
            if ((rv = SAFE_ACCEPT(apr_proc_mutex_unlock(accept_mutex)))
                != APR_SUCCESS) {
                int level = APLOG_EMERG;

                if (listener_may_exit) {
                    break;
                }
                if (ap_scoreboard_image->parent[process_slot].generation != 
                    ap_scoreboard_image->global->running_generation) {
                    level = APLOG_DEBUG; /* common to get these at restart time */
                }
                ap_log_error(APLOG_MARK, level, rv, ap_server_conf,
                             "apr_proc_mutex_unlock failed. Attempting to "
                             "shutdown process gracefully.");
                signal_threads(ST_GRACEFUL);
            }
            if (csd != NULL) {
                /* Wake up the sleeping worker. */
                apr_thread_mutex_lock(worker->mutex);
                worker->csd = (apr_socket_t *)csd;
                worker->state = WORKER_BUSY;
                /* Posix allows us to signal this condition without
                 * owning the associated mutex, but in that case it can
                 * not guarantee predictable scheduling. See
                 * _UNIX Network Programming: Interprocess Communication_
                 * by W. Richard Stevens, Vol 2, 2nd Ed, pp. 170-171. */
                apr_thread_cond_signal(worker->cond);
                apr_thread_mutex_unlock(worker->mutex);
                worker = NULL;
            }
        }
        else {
            if ((rv = SAFE_ACCEPT(apr_proc_mutex_unlock(accept_mutex)))
                != APR_SUCCESS) {
                ap_log_error(APLOG_MARK, APLOG_EMERG, rv, ap_server_conf,
                             "apr_proc_mutex_unlock failed. Attempting to "
                             "shutdown process gracefully.");
                signal_threads(ST_GRACEFUL);
            }
            break;
        }
    }

    workers_may_exit = 1;
    if (worker) {
        apr_thread_mutex_lock(worker->mutex);
        worker->state = WORKER_TERMINATED;
        /* Posix allows us to signal this condition without
         * owning the associated mutex, but in that case it can
         * not guarantee predictable scheduling. See
         * _UNIX Network Programming: Interprocess Communication_
         * by W. Richard Stevens, Vol 2, 2nd Ed, pp. 170-171. */
        apr_thread_cond_signal(worker->cond);
        apr_thread_mutex_unlock(worker->mutex);
    }
    worker_stack_terminate(idle_worker_stack);
    dying = 1;
    ap_scoreboard_image->parent[process_slot].quiescing = 1;

    /* wake up the main thread */
    kill(ap_my_pid, SIGTERM);

    apr_thread_exit(thd, APR_SUCCESS);
    return NULL;
}

/* XXX For ungraceful termination/restart, we definitely don't want to
 *     wait for active connections to finish but we may want to wait
 *     for idle workers to get out of the queue code and release mutexes,
 *     since those mutexes are cleaned up pretty soon and some systems
 *     may not react favorably (i.e., segfault) if operations are attempted
 *     on cleaned-up mutexes.
 */
static void * APR_THREAD_FUNC worker_thread(apr_thread_t *thd, void * dummy)
{
    proc_info * ti = dummy;
    int process_slot = ti->pid;
    int thread_slot = ti->tid;
    apr_bucket_alloc_t *bucket_alloc;
    apr_pool_t *tpool = apr_thread_pool_get(thd);
    apr_pool_t *ptrans;                /* Pool for per-transaction stuff */
    apr_allocator_t *allocator;
    apr_status_t rv;
    worker_wakeup_info *wakeup;

    free(ti);

    ap_update_child_status_from_indexes(process_slot, thread_slot, SERVER_STARTING, NULL);

    apr_allocator_create(&allocator);
    apr_allocator_max_free_set(allocator, ap_max_mem_free);
    /* XXX: why is ptrans's parent not tpool? --jcw 08/2003 */
    apr_pool_create_ex(&ptrans, NULL, NULL, allocator);
    apr_allocator_owner_set(allocator, ptrans);
    bucket_alloc = apr_bucket_alloc_create_ex(allocator);

    wakeup = (worker_wakeup_info *)apr_palloc(tpool, sizeof(*wakeup));
    wakeup->pool = ptrans;
    if ((rv = apr_thread_cond_create(&wakeup->cond, tpool)) != APR_SUCCESS) {
        ap_log_error(APLOG_MARK, APLOG_EMERG, rv, ap_server_conf,
                     "apr_thread_cond_create failed. Attempting to shutdown "
                     "process gracefully.");
        signal_threads(ST_GRACEFUL);
        apr_thread_exit(thd, rv);
    }
    if ((rv = apr_thread_mutex_create(&wakeup->mutex, APR_THREAD_MUTEX_DEFAULT,
                                      tpool)) != APR_SUCCESS) {
        ap_log_error(APLOG_MARK, APLOG_EMERG, rv, ap_server_conf,
                     "apr_thread_mutex_create failed. Attempting to shutdown "
                     "process gracefully.");
        signal_threads(ST_GRACEFUL);
        apr_thread_exit(thd, rv);
    }

    while (!workers_may_exit) {
        ap_update_child_status_from_indexes(process_slot, thread_slot, SERVER_READY, NULL);
        rv = worker_stack_wait(idle_worker_stack, wakeup);
        if (APR_STATUS_IS_EOF(rv)) {
            break; /* The queue has been terminated. */
        }
        else if (rv != APR_SUCCESS) {
            ap_log_error(APLOG_MARK, APLOG_CRIT, rv, ap_server_conf,
                         "worker_stack_wait failed");
            break; /* Treat all other errors as fatal. */
        }
        else if (wakeup->state == WORKER_TERMINATED) {
            break; /* They told us to quit. */
        }
        AP_DEBUG_ASSERT(wakeup->state != WORKER_IDLE);
        process_socket(ptrans, wakeup->csd,
                       process_slot, thread_slot, bucket_alloc);
        requests_this_child--; /* FIXME: should be synchronized - aaron */
        apr_pool_clear(ptrans);
    }

    ap_update_child_status_from_indexes(process_slot, thread_slot,
        (dying) ? SERVER_DEAD : SERVER_GRACEFUL, (request_rec *) NULL);

    apr_bucket_alloc_destroy(bucket_alloc);

    apr_thread_exit(thd, APR_SUCCESS);
    return NULL;
}

static int check_signal(int signum)
{
    switch (signum) {
    case SIGTERM:
    case SIGINT:
        return 1;
    }
    return 0;
}

static void create_listener_thread(thread_starter *ts)
{
    int my_child_num = ts->child_num_arg;
    apr_threadattr_t *thread_attr = ts->threadattr;
    proc_info *my_info;
    apr_status_t rv;

    my_info = (proc_info *)malloc(sizeof(proc_info));
    my_info->pid = my_child_num;
    my_info->tid = -1; /* listener thread doesn't have a thread slot */
    my_info->sd = 0;
    rv = apr_thread_create(&ts->listener, thread_attr, listener_thread,
                           my_info, pchild);
    if (rv != APR_SUCCESS) {
        ap_log_error(APLOG_MARK, APLOG_ALERT, rv, ap_server_conf,
                     "apr_thread_create: unable to create listener thread");
        /* In case system resources are maxxed out, we don't want
         * Apache running away with the CPU trying to fork over and
         * over and over again if we exit.
         * XXX Jeff doesn't see how Apache is going to try to fork again since
         * the exit code is APEXIT_CHILDFATAL
         */
        apr_sleep(10 * APR_USEC_PER_SEC);
        clean_child_exit(APEXIT_CHILDFATAL);
    }
    apr_os_thread_get(&listener_os_thread, ts->listener);
}

/* XXX under some circumstances not understood, children can get stuck
 *     in start_threads forever trying to take over slots which will
 *     never be cleaned up; for now there is an APLOG_DEBUG message issued
 *     every so often when this condition occurs
 */
static void * APR_THREAD_FUNC start_threads(apr_thread_t *thd, void *dummy)
{
    thread_starter *ts = dummy;
    apr_thread_t **threads = ts->threads;
    apr_threadattr_t *thread_attr = ts->threadattr;
    int child_num_arg = ts->child_num_arg;
    int my_child_num = child_num_arg;
    proc_info *my_info;
    apr_status_t rv;
    int i;
    int threads_created = 0;
    int loops;
    int prev_threads_created;

    idle_worker_stack = worker_stack_create(pchild, ap_threads_per_child);
    if (idle_worker_stack == NULL) {
        ap_log_error(APLOG_MARK, APLOG_ALERT, 0, ap_server_conf,
                     "worker_stack_create() failed");
        clean_child_exit(APEXIT_CHILDFATAL);
    }

    loops = prev_threads_created = 0;
    while (1) {
        /* ap_threads_per_child does not include the listener thread */
        for (i = 0; i < ap_threads_per_child; i++) {
            int status = ap_scoreboard_image->servers[child_num_arg][i].status;

            if (status != SERVER_GRACEFUL && status != SERVER_DEAD) {
                continue;
            }

            my_info = (proc_info *)malloc(sizeof(proc_info));
            if (my_info == NULL) {
                ap_log_error(APLOG_MARK, APLOG_ALERT, errno, ap_server_conf,
                             "malloc: out of memory");
                clean_child_exit(APEXIT_CHILDFATAL);
            }
            my_info->pid = my_child_num;
            my_info->tid = i;
            my_info->sd = 0;
        
            /* We are creating threads right now */
            ap_update_child_status_from_indexes(my_child_num, i,
                                                SERVER_STARTING, NULL);
            /* We let each thread update its own scoreboard entry.  This is
             * done because it lets us deal with tid better.
             */
            rv = apr_thread_create(&threads[i], thread_attr, 
                                   worker_thread, my_info, pchild);
            if (rv != APR_SUCCESS) {
                ap_log_error(APLOG_MARK, APLOG_ALERT, rv, ap_server_conf,
                    "apr_thread_create: unable to create worker thread");
                /* In case system resources are maxxed out, we don't want
                   Apache running away with the CPU trying to fork over and
                   over and over again if we exit. */
                apr_sleep(10 * APR_USEC_PER_SEC);
                clean_child_exit(APEXIT_CHILDFATAL);
            }
            threads_created++;
            if (threads_created == 1) {
                /* now that we have a worker thread, it makes sense to create
                 * a listener thread (we don't want a listener without a worker!)
                 */
                create_listener_thread(ts);
            }
        }
        if (start_thread_may_exit || threads_created == ap_threads_per_child) {
            break;
        }
        /* wait for previous generation to clean up an entry */
        apr_sleep(1 * APR_USEC_PER_SEC);
        ++loops;
        if (loops % 120 == 0) { /* every couple of minutes */
            if (prev_threads_created == threads_created) {
                ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, ap_server_conf,
                             "child %" APR_PID_T_FMT " isn't taking over "
                             "slots very quickly (%d of %d)",
                             ap_my_pid, threads_created, ap_threads_per_child);
            }
            prev_threads_created = threads_created;
        }
    }
    
    /* What state should this child_main process be listed as in the 
     * scoreboard...?
     *  ap_update_child_status_from_indexes(my_child_num, i, SERVER_STARTING, 
     *                                      (request_rec *) NULL);
     * 
     *  This state should be listed separately in the scoreboard, in some kind
     *  of process_status, not mixed in with the worker threads' status.   
     *  "life_status" is almost right, but it's in the worker's structure, and 
     *  the name could be clearer.   gla
     */
    apr_thread_exit(thd, APR_SUCCESS);
    return NULL;
}

static void join_workers(apr_thread_t *listener, apr_thread_t **threads)
{
    int i;
    apr_status_t rv, thread_rv;

    if (listener) {
        int iter;

        /* deal with a rare timing window which affects waking up the
         * listener thread...  if the signal sent to the listener thread
         * is delivered between the time it verifies that the
         * listener_may_exit flag is clear and the time it enters a
         * blocking syscall, the signal didn't do any good...  work around
         * that by sleeping briefly and sending it again
         */

        iter = 0;
        while (iter < 10 && 
#ifdef HAVE_PTHREAD_KILL
               pthread_kill(*listener_os_thread, 0)
#else
               kill(ap_my_pid, 0)
#endif
               == 0) {
            /* listener not dead yet */
            apr_sleep(APR_USEC_PER_SEC / 2);
            wakeup_listener();
            ++iter;
        }
        if (iter >= 10) {
            ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, ap_server_conf,
                         "the listener thread didn't exit");
        }
        else {
            rv = apr_thread_join(&thread_rv, listener);
            if (rv != APR_SUCCESS) {
                ap_log_error(APLOG_MARK, APLOG_CRIT, rv, ap_server_conf,
                             "apr_thread_join: unable to join listener thread");
            }
        }
    }
    
    for (i = 0; i < ap_threads_per_child; i++) {
        if (threads[i]) { /* if we ever created this thread */
            rv = apr_thread_join(&thread_rv, threads[i]);
            if (rv != APR_SUCCESS) {
                ap_log_error(APLOG_MARK, APLOG_CRIT, rv, ap_server_conf,
                             "apr_thread_join: unable to join worker "
                             "thread %d",
                             i);
            }
        }
    }
}

static void join_start_thread(apr_thread_t *start_thread_id)
{
    apr_status_t rv, thread_rv;

    start_thread_may_exit = 1; /* tell it to give up in case it is still 
                                * trying to take over slots from a 
                                * previous generation
                                */
    rv = apr_thread_join(&thread_rv, start_thread_id);
    if (rv != APR_SUCCESS) {
        ap_log_error(APLOG_MARK, APLOG_CRIT, rv, ap_server_conf,
                     "apr_thread_join: unable to join the start "
                     "thread");
    }
}

static void child_main(int child_num_arg)
{
    apr_thread_t **threads;
    apr_status_t rv;
    thread_starter *ts;
    apr_threadattr_t *thread_attr;
    apr_thread_t *start_thread_id;

    mpm_state = AP_MPMQ_STARTING; /* for benefit of any hooks that run as this
                                   * child initializes
                                   */
    ap_my_pid = getpid();
    ap_fatal_signal_child_setup(ap_server_conf);
    apr_pool_create(&pchild, pconf);

    /*stuff to do before we switch id's, so we have permissions.*/
    ap_reopen_scoreboard(pchild, NULL, 0);

    rv = SAFE_ACCEPT(apr_proc_mutex_child_init(&accept_mutex, ap_lock_fname,
                                               pchild));
    if (rv != APR_SUCCESS) {
        ap_log_error(APLOG_MARK, APLOG_EMERG, rv, ap_server_conf,
                     "Couldn't initialize cross-process lock in child");
        clean_child_exit(APEXIT_CHILDFATAL);
    }

    if (unixd_setup_child()) {
        clean_child_exit(APEXIT_CHILDFATAL);
    }

    ap_run_child_init(pchild, ap_server_conf);

    /* done with init critical section */

    /* Just use the standard apr_setup_signal_thread to block all signals
     * from being received.  The child processes no longer use signals for
     * any communication with the parent process.
     */
    rv = apr_setup_signal_thread();
    if (rv != APR_SUCCESS) {
        ap_log_error(APLOG_MARK, APLOG_EMERG, rv, ap_server_conf,
                     "Couldn't initialize signal thread");
        clean_child_exit(APEXIT_CHILDFATAL);
    }

    if (ap_max_requests_per_child) {
        requests_this_child = ap_max_requests_per_child;
    }
    else {
        /* coding a value of zero means infinity */
        requests_this_child = INT_MAX;
    }
    
    /* Setup worker threads */

    /* clear the storage; we may not create all our threads immediately, 
     * and we want a 0 entry to indicate a thread which was not created
     */
    threads = (apr_thread_t **)calloc(1, 
                                sizeof(apr_thread_t *) * ap_threads_per_child);
    if (threads == NULL) {
        ap_log_error(APLOG_MARK, APLOG_ALERT, errno, ap_server_conf,
                     "malloc: out of memory");
        clean_child_exit(APEXIT_CHILDFATAL);
    }

    ts = (thread_starter *)apr_palloc(pchild, sizeof(*ts));

    apr_threadattr_create(&thread_attr, pchild);
    /* 0 means PTHREAD_CREATE_JOINABLE */
    apr_threadattr_detach_set(thread_attr, 0);

    ts->threads = threads;
    ts->listener = NULL;
    ts->child_num_arg = child_num_arg;
    ts->threadattr = thread_attr;

    rv = apr_thread_create(&start_thread_id, thread_attr, start_threads,
                           ts, pchild);
    if (rv != APR_SUCCESS) {
        ap_log_error(APLOG_MARK, APLOG_ALERT, rv, ap_server_conf,
                     "apr_thread_create: unable to create worker thread");
        /* In case system resources are maxxed out, we don't want
           Apache running away with the CPU trying to fork over and
           over and over again if we exit. */
        apr_sleep(10 * APR_USEC_PER_SEC);
        clean_child_exit(APEXIT_CHILDFATAL);
    }

    mpm_state = AP_MPMQ_RUNNING;
    
    /* If we are only running in one_process mode, we will want to
     * still handle signals. */
    if (one_process) {
        /* Block until we get a terminating signal. */
        apr_signal_thread(check_signal);
        /* make sure the start thread has finished; signal_threads() 
         * and join_workers() depend on that
         */
        /* XXX join_start_thread() won't be awakened if one of our
         *     threads encounters a critical error and attempts to
         *     shutdown this child
         */
        join_start_thread(start_thread_id);
        signal_threads(ST_UNGRACEFUL); /* helps us terminate a little more
                           * quickly than the dispatch of the signal thread
                           * beats the Pipe of Death and the browsers
                           */
        /* A terminating signal was received. Now join each of the
         * workers to clean them up.
         *   If the worker already exited, then the join frees
         *   their resources and returns.
         *   If the worker hasn't exited, then this blocks until
         *   they have (then cleans up).
         */
        join_workers(ts->listener, threads);
    }
    else { /* !one_process */
        /* remove SIGTERM from the set of blocked signals...  if one of
         * the other threads in the process needs to take us down
         * (e.g., for MaxRequestsPerChild) it will send us SIGTERM
         */
        unblock_signal(SIGTERM);
        apr_signal(SIGTERM, dummy_signal_handler);
        /* Watch for any messages from the parent over the POD */
        while (1) {
            rv = ap_mpm_pod_check(pod);
            if (rv == AP_NORESTART) {
                /* see if termination was triggered while we slept */
                switch(terminate_mode) {
                case ST_GRACEFUL:
                    rv = AP_GRACEFUL;
                    break;
                case ST_UNGRACEFUL:
                    rv = AP_RESTART;
                    break;
                }
            }
            if (rv == AP_GRACEFUL || rv == AP_RESTART) {
                /* make sure the start thread has finished; 
                 * signal_threads() and join_workers depend on that
                 */
                join_start_thread(start_thread_id);
                signal_threads(rv == AP_GRACEFUL ? ST_GRACEFUL : ST_UNGRACEFUL);
                break;
            }
        }

        if (rv == AP_GRACEFUL) {
            /* A terminating signal was received. Now join each of the
             * workers to clean them up.
             *   If the worker already exited, then the join frees
             *   their resources and returns.
             *   If the worker hasn't exited, then this blocks until
             *   they have (then cleans up).
             */
            join_workers(ts->listener, threads);
        }
    }

    free(threads);

    clean_child_exit(resource_shortage ? APEXIT_CHILDSICK : 0);
}

static int make_child(server_rec *s, int slot) 
{
    int pid;

    if (slot + 1 > ap_max_daemons_limit) {
        ap_max_daemons_limit = slot + 1;
    }

    if (one_process) {
        set_signals();
        ap_scoreboard_image->parent[slot].pid = getpid();
        child_main(slot);
    }

    if ((pid = fork()) == -1) {
        ap_log_error(APLOG_MARK, APLOG_ERR, errno, s, 
                     "fork: Unable to fork new process");

        /* fork didn't succeed. Fix the scoreboard or else
         * it will say SERVER_STARTING forever and ever
         */
        ap_update_child_status_from_indexes(slot, 0, SERVER_DEAD, NULL);

        /* In case system resources are maxxed out, we don't want
           Apache running away with the CPU trying to fork over and
           over and over again. */
        apr_sleep(10 * APR_USEC_PER_SEC);

        return -1;
    }

    if (!pid) {
#ifdef HAVE_BINDPROCESSOR
        /* By default, AIX binds to a single processor.  This bit unbinds
         * children which will then bind to another CPU.
         */
        int status = bindprocessor(BINDPROCESS, (int)getpid(),
                               PROCESSOR_CLASS_ANY);
        if (status != OK)
            ap_log_error(APLOG_MARK, APLOG_WARNING, errno, 
                         ap_server_conf,
                         "processor unbind failed %d", status);
#endif
        RAISE_SIGSTOP(MAKE_CHILD);

        apr_signal(SIGTERM, just_die);
        child_main(slot);

        clean_child_exit(0);
    }
    /* else */
    ap_scoreboard_image->parent[slot].quiescing = 0;
    ap_scoreboard_image->parent[slot].pid = pid;
    return 0;
}

/* start up a bunch of children */
static void startup_children(int number_to_start)
{
    int i;

    for (i = 0; number_to_start && i < ap_daemons_limit; ++i) {
        if (ap_scoreboard_image->parent[i].pid != 0) {
            continue;
        }
        if (make_child(ap_server_conf, i) < 0) {
            break;
        }
        --number_to_start;
    }
}


/*
 * idle_spawn_rate is the number of children that will be spawned on the
 * next maintenance cycle if there aren't enough idle servers.  It is
 * doubled up to MAX_SPAWN_RATE, and reset only when a cycle goes by
 * without the need to spawn.
 */
static int idle_spawn_rate = 1;
#ifndef MAX_SPAWN_RATE
#define MAX_SPAWN_RATE        (32)
#endif
static int hold_off_on_exponential_spawning;

static void perform_idle_server_maintenance(void)
{
    int i, j;
    int idle_thread_count;
    worker_score *ws;
    process_score *ps;
    int free_length;
    int totally_free_length = 0;
    int free_slots[MAX_SPAWN_RATE];
    int last_non_dead;
    int total_non_dead;

    /* initialize the free_list */
    free_length = 0;

    idle_thread_count = 0;
    last_non_dead = -1;
    total_non_dead = 0;

    for (i = 0; i < ap_daemons_limit; ++i) {
        /* Initialization to satisfy the compiler. It doesn't know
         * that ap_threads_per_child is always > 0 */
        int status = SERVER_DEAD;
        int any_dying_threads = 0;
        int any_dead_threads = 0;
        int all_dead_threads = 1;

        if (i >= ap_max_daemons_limit && totally_free_length == idle_spawn_rate)
            break;
        ps = &ap_scoreboard_image->parent[i];
        for (j = 0; j < ap_threads_per_child; j++) {
            ws = &ap_scoreboard_image->servers[i][j];
            status = ws->status;

            /* XXX any_dying_threads is probably no longer needed    GLA */
            any_dying_threads = any_dying_threads || 
                                (status == SERVER_GRACEFUL);
            any_dead_threads = any_dead_threads || (status == SERVER_DEAD);
            all_dead_threads = all_dead_threads &&
                                   (status == SERVER_DEAD ||
                                    status == SERVER_GRACEFUL);

            /* We consider a starting server as idle because we started it
             * at least a cycle ago, and if it still hasn't finished starting
             * then we're just going to swamp things worse by forking more.
             * So we hopefully won't need to fork more if we count it.
             * This depends on the ordering of SERVER_READY and SERVER_STARTING.
             */
            if (status <= SERVER_READY && status != SERVER_DEAD &&
                    !ps->quiescing &&
                    ps->generation == ap_my_generation &&
                 /* XXX the following shouldn't be necessary if we clean up 
                  *     properly after seg faults, but we're not yet    GLA 
                  */     
                    ps->pid != 0) {
                ++idle_thread_count;
            }
        }
        if (any_dead_threads && totally_free_length < idle_spawn_rate 
                && (!ps->pid               /* no process in the slot */
                    || ps->quiescing)) {   /* or at least one is going away */
            if (all_dead_threads) {
                /* great! we prefer these, because the new process can
                 * start more threads sooner.  So prioritize this slot 
                 * by putting it ahead of any slots with active threads.
                 *
                 * first, make room by moving a slot that's potentially still
                 * in use to the end of the array
                 */
                free_slots[free_length] = free_slots[totally_free_length];
                free_slots[totally_free_length++] = i;
            }
            else {
                /* slot is still in use - back of the bus
                 */
            free_slots[free_length] = i;
            }
            ++free_length;
        }
        /* XXX if (!ps->quiescing)     is probably more reliable  GLA */
        if (!any_dying_threads) {
            last_non_dead = i;
            ++total_non_dead;
        }
    }
    ap_max_daemons_limit = last_non_dead + 1;

    if (idle_thread_count > max_spare_threads) {
        /* Kill off one child */
        ap_mpm_pod_signal(pod, TRUE);
        idle_spawn_rate = 1;
    }
    else if (idle_thread_count < min_spare_threads) {
        /* terminate the free list */
        if (free_length == 0) {
            /* only report this condition once */
            static int reported = 0;
            
            if (!reported) {
                ap_log_error(APLOG_MARK, APLOG_ERR, 0, 
                             ap_server_conf,
                             "server reached MaxClients setting, consider"
                             " raising the MaxClients setting");
                reported = 1;
            }
            idle_spawn_rate = 1;
        }
        else {
            if (free_length > idle_spawn_rate) {
                free_length = idle_spawn_rate;
            }
            if (idle_spawn_rate >= 8) {
                ap_log_error(APLOG_MARK, APLOG_INFO, 0, 
                             ap_server_conf,
                             "server seems busy, (you may need "
                             "to increase StartServers, ThreadsPerChild "
                             "or Min/MaxSpareThreads), "
                             "spawning %d children, there are around %d idle "
                             "threads, and %d total children", free_length,
                             idle_thread_count, total_non_dead);
            }
            for (i = 0; i < free_length; ++i) {
                make_child(ap_server_conf, free_slots[i]);
            }
            /* the next time around we want to spawn twice as many if this
             * wasn't good enough, but not if we've just done a graceful
             */
            if (hold_off_on_exponential_spawning) {
                --hold_off_on_exponential_spawning;
            }
            else if (idle_spawn_rate < MAX_SPAWN_RATE) {
                idle_spawn_rate *= 2;
            }
        }
    }
    else {
      idle_spawn_rate = 1;
    }
}

static void server_main_loop(int remaining_children_to_start)
{
    int child_slot;
    apr_exit_why_e exitwhy;
    int status, processed_status;
    apr_proc_t pid;
    int i;

    while (!restart_pending && !shutdown_pending) {
        ap_wait_or_timeout(&exitwhy, &status, &pid, pconf);
        
        if (pid.pid != -1) {
            processed_status = ap_process_child_status(&pid, exitwhy, status);
            if (processed_status == APEXIT_CHILDFATAL) {
                shutdown_pending = 1;
                child_fatal = 1;
                return;
            }
            /* non-fatal death... note that it's gone in the scoreboard. */
            child_slot = find_child_by_pid(&pid);
            if (child_slot >= 0) {
                for (i = 0; i < ap_threads_per_child; i++)
                    ap_update_child_status_from_indexes(child_slot, i, SERVER_DEAD, 
                                                        (request_rec *) NULL);
                
                ap_scoreboard_image->parent[child_slot].pid = 0;
                ap_scoreboard_image->parent[child_slot].quiescing = 0;
                if (processed_status == APEXIT_CHILDSICK) {
                    /* resource shortage, minimize the fork rate */
                    idle_spawn_rate = 1;
                }
                else if (remaining_children_to_start
                    && child_slot < ap_daemons_limit) {
                    /* we're still doing a 1-for-1 replacement of dead
                     * children with new children
                     */
                    make_child(ap_server_conf, child_slot);
                    --remaining_children_to_start;
                }
#if APR_HAS_OTHER_CHILD
            }
            else if (apr_proc_other_child_read(&pid, status) == 0) {
                /* handled */
#endif
            }
            else if (is_graceful) {
                /* Great, we've probably just lost a slot in the
                 * scoreboard.  Somehow we don't know about this child.
                 */
                ap_log_error(APLOG_MARK, APLOG_WARNING, 0,
                             ap_server_conf,
                             "long lost child came home! (pid %ld)",
                             (long)pid.pid);
            }
            /* Don't perform idle maintenance when a child dies,
             * only do it when there's a timeout.  Remember only a
             * finite number of children can die, and it's pretty
             * pathological for a lot to die suddenly.
             */
            continue;
        }
        else if (remaining_children_to_start) {
            /* we hit a 1 second timeout in which none of the previous
             * generation of children needed to be reaped... so assume
             * they're all done, and pick up the slack if any is left.
             */
            startup_children(remaining_children_to_start);
            remaining_children_to_start = 0;
            /* In any event we really shouldn't do the code below because
             * few of the servers we just started are in the IDLE state
             * yet, so we'd mistakenly create an extra server.
             */
            continue;
        }

        perform_idle_server_maintenance();
    }
}

int ap_mpm_run(apr_pool_t *_pconf, apr_pool_t *plog, server_rec *s)
{
    int remaining_children_to_start;
    apr_status_t rv;

    ap_log_pid(pconf, ap_pid_fname);

    first_server_limit = server_limit;
    first_thread_limit = thread_limit;
    if (changed_limit_at_restart) {
        ap_log_error(APLOG_MARK, APLOG_WARNING, 0, s,
                     "WARNING: Attempt to change ServerLimit or ThreadLimit "
                     "ignored during restart");
        changed_limit_at_restart = 0;
    }
    
    /* Initialize cross-process accept lock */
    ap_lock_fname = apr_psprintf(_pconf, "%s.%" APR_PID_T_FMT,
                                 ap_server_root_relative(_pconf, ap_lock_fname),
                                 ap_my_pid);

    rv = apr_proc_mutex_create(&accept_mutex, ap_lock_fname, 
                               ap_accept_lock_mech, _pconf);
    if (rv != APR_SUCCESS) {
        ap_log_error(APLOG_MARK, APLOG_EMERG, rv, s,
                     "Couldn't create accept lock");
        mpm_state = AP_MPMQ_STOPPING;
        return 1;
    }

#if APR_USE_SYSVSEM_SERIALIZE
    if (ap_accept_lock_mech == APR_LOCK_DEFAULT || 
        ap_accept_lock_mech == APR_LOCK_SYSVSEM) {
#else
    if (ap_accept_lock_mech == APR_LOCK_SYSVSEM) {
#endif
        rv = unixd_set_proc_mutex_perms(accept_mutex);
        if (rv != APR_SUCCESS) {
            ap_log_error(APLOG_MARK, APLOG_EMERG, rv, s,
                         "Couldn't set permissions on cross-process lock; "
                         "check User and Group directives");
            mpm_state = AP_MPMQ_STOPPING;
            return 1;
        }
    }

    if (!is_graceful) {
        if (ap_run_pre_mpm(s->process->pool, SB_SHARED) != OK) {
            mpm_state = AP_MPMQ_STOPPING;
            return 1;
        }
        /* fix the generation number in the global score; we just got a new,
         * cleared scoreboard
         */
        ap_scoreboard_image->global->running_generation = ap_my_generation;
    }

    set_signals();
    /* Don't thrash... */
    if (max_spare_threads < min_spare_threads + ap_threads_per_child)
        max_spare_threads = min_spare_threads + ap_threads_per_child;

    /* If we're doing a graceful_restart then we're going to see a lot
     * of children exiting immediately when we get into the main loop
     * below (because we just sent them AP_SIG_GRACEFUL).  This happens pretty
     * rapidly... and for each one that exits we'll start a new one until
     * we reach at least daemons_min_free.  But we may be permitted to
     * start more than that, so we'll just keep track of how many we're
     * supposed to start up without the 1 second penalty between each fork.
     */
    remaining_children_to_start = ap_daemons_to_start;
    if (remaining_children_to_start > ap_daemons_limit) {
        remaining_children_to_start = ap_daemons_limit;
    }
    if (!is_graceful) {
        startup_children(remaining_children_to_start);
        remaining_children_to_start = 0;
    }
    else {
        /* give the system some time to recover before kicking into
            * exponential mode */
        hold_off_on_exponential_spawning = 10;
    }

    ap_log_error(APLOG_MARK, APLOG_NOTICE, 0, ap_server_conf,
                "%s configured -- resuming normal operations",
                ap_get_server_version());
    ap_log_error(APLOG_MARK, APLOG_INFO, 0, ap_server_conf,
                "Server built: %s", ap_get_server_built());
#ifdef AP_MPM_WANT_SET_ACCEPT_LOCK_MECH
    ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, ap_server_conf,
		"AcceptMutex: %s (default: %s)",
		apr_proc_mutex_name(accept_mutex),
		apr_proc_mutex_defname());
#endif
    restart_pending = shutdown_pending = 0;
    mpm_state = AP_MPMQ_RUNNING;

    server_main_loop(remaining_children_to_start);
    mpm_state = AP_MPMQ_STOPPING;

    if (shutdown_pending) {
        /* Time to gracefully shut down:
         * Kill child processes, tell them to call child_exit, etc...
         * (By "gracefully" we don't mean graceful in the same sense as 
         * "apachectl graceful" where we allow old connections to finish.)
         */
        ap_mpm_pod_killpg(pod, ap_daemons_limit, FALSE);
        ap_reclaim_child_processes(1);                /* Start with SIGTERM */

        if (!child_fatal) {
            /* cleanup pid file on normal shutdown */
            const char *pidfile = NULL;
            pidfile = ap_server_root_relative (pconf, ap_pid_fname);
            if ( pidfile != NULL && unlink(pidfile) == 0)
                ap_log_error(APLOG_MARK, APLOG_INFO, 0,
                             ap_server_conf,
                             "removed PID file %s (pid=%ld)",
                             pidfile, (long)getpid());
    
            ap_log_error(APLOG_MARK, APLOG_NOTICE, 0,
                         ap_server_conf, "caught SIGTERM, shutting down");
        }
        return 1;
    }

    /* we've been told to restart */
    apr_signal(SIGHUP, SIG_IGN);

    if (one_process) {
        /* not worth thinking about */
        return 1;
    }

    /* advance to the next generation */
    /* XXX: we really need to make sure this new generation number isn't in
     * use by any of the children.
     */
    ++ap_my_generation;
    ap_scoreboard_image->global->running_generation = ap_my_generation;
    
    if (is_graceful) {
        ap_log_error(APLOG_MARK, APLOG_NOTICE, 0, ap_server_conf,
                     AP_SIG_GRACEFUL_STRING " received.  Doing graceful restart");
        /* wake up the children...time to die.  But we'll have more soon */
        ap_mpm_pod_killpg(pod, ap_daemons_limit, TRUE);
    

        /* This is mostly for debugging... so that we know what is still
         * gracefully dealing with existing request.
         */
        
    }
    else {
        /* Kill 'em all.  Since the child acts the same on the parents SIGTERM 
         * and a SIGHUP, we may as well use the same signal, because some user
         * pthreads are stealing signals from us left and right.
         */
        ap_mpm_pod_killpg(pod, ap_daemons_limit, FALSE);

        ap_reclaim_child_processes(1);                /* Start with SIGTERM */
        ap_log_error(APLOG_MARK, APLOG_NOTICE, 0, ap_server_conf,
                    "SIGHUP received.  Attempting to restart");
    }

    return 0;
}

/* This really should be a post_config hook, but the error log is already
 * redirected by that point, so we need to do this in the open_logs phase.
 */
static int worker_open_logs(apr_pool_t *p, apr_pool_t *plog, apr_pool_t *ptemp, server_rec *s)
{
    apr_status_t rv;
    ap_listen_rec *lr;

    pconf = p;
    ap_server_conf = s;

    if ((num_listensocks = ap_setup_listeners(ap_server_conf)) < 1) {
        ap_log_error(APLOG_MARK, APLOG_ALERT|APLOG_STARTUP, 0,
                     NULL, "no listening sockets available, shutting down");
        return DONE;
    }

#if APR_O_NONBLOCK_INHERITED
    for(lr = ap_listeners ; lr != NULL ; lr = lr->next) {
        apr_socket_opt_set(lr->sd, APR_SO_NONBLOCK, 1);
    }
#endif /* APR_O_NONBLOCK_INHERITED */

    if (!one_process) {
        if ((rv = ap_mpm_pod_open(pconf, &pod))) {
            ap_log_error(APLOG_MARK, APLOG_CRIT|APLOG_STARTUP, rv, NULL,
                    "Could not open pipe-of-death.");
            return DONE;
        }
    }
    return OK;
}

static int worker_pre_config(apr_pool_t *pconf, apr_pool_t *plog, 
                             apr_pool_t *ptemp)
{
    static int restart_num = 0;
    int no_detach, debug, foreground;
    ap_directive_t *pdir;
    ap_directive_t *max_clients = NULL;
    apr_status_t rv;

    mpm_state = AP_MPMQ_STARTING;
    
    /* make sure that "ThreadsPerChild" gets set before "MaxClients" */
    for (pdir = ap_conftree; pdir != NULL; pdir = pdir->next) {
        if (strncasecmp(pdir->directive, "ThreadsPerChild", 15) == 0) {
            if (!max_clients) {
                break; /* we're in the clear, got ThreadsPerChild first */
            }
            else {
                /* now to swap the data */
                ap_directive_t temp;

                temp.directive = pdir->directive;
                temp.args = pdir->args;
                /* Make sure you don't change 'next', or you may get loops! */
                /* XXX: first_child, parent, and data can never be set
                 * for these directives, right? -aaron */
                temp.filename = pdir->filename;
                temp.line_num = pdir->line_num;

                pdir->directive = max_clients->directive;
                pdir->args = max_clients->args;
                pdir->filename = max_clients->filename;
                pdir->line_num = max_clients->line_num;
                
                max_clients->directive = temp.directive;
                max_clients->args = temp.args;
                max_clients->filename = temp.filename;
                max_clients->line_num = temp.line_num;
                break;
            }
        }
        else if (!max_clients
                 && strncasecmp(pdir->directive, "MaxClients", 10) == 0) {
            max_clients = pdir;
        }
    }

    debug = ap_exists_config_define("DEBUG");

    if (debug) {
        foreground = one_process = 1;
        no_detach = 0;
    }
    else {
        one_process = ap_exists_config_define("ONE_PROCESS");
        no_detach = ap_exists_config_define("NO_DETACH");
        foreground = ap_exists_config_define("FOREGROUND");
    }

    /* sigh, want this only the second time around */
    if (restart_num++ == 1) {
        is_graceful = 0;

        if (!one_process && !foreground) {
            rv = apr_proc_detach(no_detach ? APR_PROC_DETACH_FOREGROUND
                                           : APR_PROC_DETACH_DAEMONIZE);
            if (rv != APR_SUCCESS) {
                ap_log_error(APLOG_MARK, APLOG_CRIT, rv, NULL,
                             "apr_proc_detach failed");
                return HTTP_INTERNAL_SERVER_ERROR;
            }
        }
        parent_pid = ap_my_pid = getpid();
    }

    unixd_pre_config(ptemp);
    ap_listen_pre_config();
    ap_daemons_to_start = DEFAULT_START_DAEMON;
    min_spare_threads = DEFAULT_MIN_FREE_DAEMON * DEFAULT_THREADS_PER_CHILD;
    max_spare_threads = DEFAULT_MAX_FREE_DAEMON * DEFAULT_THREADS_PER_CHILD;
    ap_daemons_limit = server_limit;
    ap_threads_per_child = DEFAULT_THREADS_PER_CHILD;
    ap_pid_fname = DEFAULT_PIDLOG;
    ap_lock_fname = DEFAULT_LOCKFILE;
    ap_max_requests_per_child = DEFAULT_MAX_REQUESTS_PER_CHILD;
    ap_extended_status = 0;
#ifdef AP_MPM_WANT_SET_MAX_MEM_FREE
	ap_max_mem_free = APR_ALLOCATOR_MAX_FREE_UNLIMITED;
#endif

    apr_cpystrn(ap_coredump_dir, ap_server_root, sizeof(ap_coredump_dir));

    return OK;
}

static void threadpool_hooks(apr_pool_t *p)
{
    /* The worker open_logs phase must run before the core's, or stderr
     * will be redirected to a file, and the messages won't print to the
     * console.
     */
    static const char *const aszSucc[] = {"core.c", NULL};
    one_process = 0;

    ap_hook_open_logs(worker_open_logs, NULL, aszSucc, APR_HOOK_MIDDLE);
    /* we need to set the MPM state before other pre-config hooks use MPM query
     * to retrieve it, so register as REALLY_FIRST
     */
    ap_hook_pre_config(worker_pre_config, NULL, NULL, APR_HOOK_REALLY_FIRST);
}

static const char *set_daemons_to_start(cmd_parms *cmd, void *dummy,
                                        const char *arg) 
{
    const char *err = ap_check_cmd_context(cmd, GLOBAL_ONLY);
    if (err != NULL) {
        return err;
    }

    ap_daemons_to_start = atoi(arg);
    return NULL;
}

static const char *set_min_spare_threads(cmd_parms *cmd, void *dummy,
                                         const char *arg)
{
    const char *err = ap_check_cmd_context(cmd, GLOBAL_ONLY);
    if (err != NULL) {
        return err;
    }

    min_spare_threads = atoi(arg);
    if (min_spare_threads <= 0) {
       ap_log_error(APLOG_MARK, APLOG_STARTUP, 0, NULL, 
                    "WARNING: detected MinSpareThreads set to non-positive.");
       ap_log_error(APLOG_MARK, APLOG_STARTUP, 0, NULL, 
                    "Resetting to 1 to avoid almost certain Apache failure.");
       ap_log_error(APLOG_MARK, APLOG_STARTUP, 0, NULL, 
                    "Please read the documentation.");
       min_spare_threads = 1;
    }
       
    return NULL;
}

static const char *set_max_spare_threads(cmd_parms *cmd, void *dummy,
                                         const char *arg)
{
    const char *err = ap_check_cmd_context(cmd, GLOBAL_ONLY);
    if (err != NULL) {
        return err;
    }

    max_spare_threads = atoi(arg);
    return NULL;
}

static const char *set_max_clients (cmd_parms *cmd, void *dummy,
                                     const char *arg) 
{
    int max_clients;
    const char *err = ap_check_cmd_context(cmd, GLOBAL_ONLY);
    if (err != NULL) {
        return err;
    }

    /* It is ok to use ap_threads_per_child here because we are
     * sure that it gets set before MaxClients in the pre_config stage. */
    max_clients = atoi(arg);
    if (max_clients < ap_threads_per_child) {
       ap_log_error(APLOG_MARK, APLOG_STARTUP, 0, NULL, 
                    "WARNING: MaxClients (%d) must be at least as large",
                    max_clients);
       ap_log_error(APLOG_MARK, APLOG_STARTUP, 0, NULL, 
                    " large as ThreadsPerChild (%d). Automatically",
                    ap_threads_per_child);
       ap_log_error(APLOG_MARK, APLOG_STARTUP, 0, NULL, 
                    " increasing MaxClients to %d.",
                    ap_threads_per_child);
       max_clients = ap_threads_per_child;
    }
    ap_daemons_limit = max_clients / ap_threads_per_child;
    if ((max_clients > 0) && (max_clients % ap_threads_per_child)) {
       ap_log_error(APLOG_MARK, APLOG_STARTUP, 0, NULL, 
                    "WARNING: MaxClients (%d) is not an integer multiple",
                    max_clients);
       ap_log_error(APLOG_MARK, APLOG_STARTUP, 0, NULL, 
                    " of ThreadsPerChild (%d), lowering MaxClients to %d",
                    ap_threads_per_child,
                    ap_daemons_limit * ap_threads_per_child);
       ap_log_error(APLOG_MARK, APLOG_STARTUP, 0, NULL, 
                    " for a maximum of %d child processes,",
                    ap_daemons_limit);
       max_clients = ap_daemons_limit * ap_threads_per_child; 
    }
    if (ap_daemons_limit > server_limit) {
       ap_log_error(APLOG_MARK, APLOG_STARTUP, 0, NULL, 
                    "WARNING: MaxClients of %d would require %d servers,",
                    max_clients, ap_daemons_limit);
       ap_log_error(APLOG_MARK, APLOG_STARTUP, 0, NULL, 
                    " and would exceed the ServerLimit value of %d.",
                    server_limit);
       ap_log_error(APLOG_MARK, APLOG_STARTUP, 0, NULL, 
                    " Automatically lowering MaxClients to %d.  To increase,",
                    server_limit * ap_threads_per_child);
       ap_log_error(APLOG_MARK, APLOG_STARTUP, 0, NULL, 
                    " please see the ServerLimit directive.");
       ap_daemons_limit = server_limit;
    } 
    else if (ap_daemons_limit < 1) {
        ap_log_error(APLOG_MARK, APLOG_STARTUP, 0, NULL, 
                     "WARNING: Require MaxClients > 0, setting to 1");
        ap_daemons_limit = 1;
    }
    return NULL;
}

static const char *set_threads_per_child (cmd_parms *cmd, void *dummy,
                                          const char *arg) 
{
    const char *err = ap_check_cmd_context(cmd, GLOBAL_ONLY);
    if (err != NULL) {
        return err;
    }

    ap_threads_per_child = atoi(arg);
    if (ap_threads_per_child > thread_limit) {
        ap_log_error(APLOG_MARK, APLOG_STARTUP, 0, NULL, 
                     "WARNING: ThreadsPerChild of %d exceeds ThreadLimit "
                     "value of %d", ap_threads_per_child,
                     thread_limit);
        ap_log_error(APLOG_MARK, APLOG_STARTUP, 0, NULL, 
                     "threads, lowering ThreadsPerChild to %d. To increase, please"
                     " see the", thread_limit);
        ap_log_error(APLOG_MARK, APLOG_STARTUP, 0, NULL, 
                     " ThreadLimit directive.");
        ap_threads_per_child = thread_limit;
    }
    else if (ap_threads_per_child < 1) {
        ap_log_error(APLOG_MARK, APLOG_STARTUP, 0, NULL, 
                     "WARNING: Require ThreadsPerChild > 0, setting to 1");
        ap_threads_per_child = 1;
    }
    return NULL;
}

static const char *set_server_limit (cmd_parms *cmd, void *dummy, const char *arg) 
{
    int tmp_server_limit;
    
    const char *err = ap_check_cmd_context(cmd, GLOBAL_ONLY);
    if (err != NULL) {
        return err;
    }

    tmp_server_limit = atoi(arg);
    /* you cannot change ServerLimit across a restart; ignore
     * any such attempts
     */
    if (first_server_limit &&
        tmp_server_limit != server_limit) {
        /* how do we log a message?  the error log is a bit bucket at this
         * point; we'll just have to set a flag so that ap_mpm_run()
         * logs a warning later
         */
        changed_limit_at_restart = 1;
        return NULL;
    }
    server_limit = tmp_server_limit;
    
    if (server_limit > MAX_SERVER_LIMIT) {
       ap_log_error(APLOG_MARK, APLOG_STARTUP, 0, NULL, 
                    "WARNING: ServerLimit of %d exceeds compile time limit "
                    "of %d servers,", server_limit, MAX_SERVER_LIMIT);
       ap_log_error(APLOG_MARK, APLOG_STARTUP, 0, NULL, 
                    " lowering ServerLimit to %d.", MAX_SERVER_LIMIT);
       server_limit = MAX_SERVER_LIMIT;
    } 
    else if (server_limit < 1) {
	ap_log_error(APLOG_MARK, APLOG_STARTUP, 0, NULL, 
                     "WARNING: Require ServerLimit > 0, setting to 1");
	server_limit = 1;
    }
    return NULL;
}

static const char *set_thread_limit (cmd_parms *cmd, void *dummy, const char *arg) 
{
    int tmp_thread_limit;
    
    const char *err = ap_check_cmd_context(cmd, GLOBAL_ONLY);
    if (err != NULL) {
        return err;
    }

    tmp_thread_limit = atoi(arg);
    /* you cannot change ThreadLimit across a restart; ignore
     * any such attempts
     */
    if (first_thread_limit &&
        tmp_thread_limit != thread_limit) {
        /* how do we log a message?  the error log is a bit bucket at this
         * point; we'll just have to set a flag so that ap_mpm_run()
         * logs a warning later
         */
        changed_limit_at_restart = 1;
        return NULL;
    }
    thread_limit = tmp_thread_limit;
    
    if (thread_limit > MAX_THREAD_LIMIT) {
       ap_log_error(APLOG_MARK, APLOG_STARTUP, 0, NULL, 
                    "WARNING: ThreadLimit of %d exceeds compile time limit "
                    "of %d servers,", thread_limit, MAX_THREAD_LIMIT);
       ap_log_error(APLOG_MARK, APLOG_STARTUP, 0, NULL, 
                    " lowering ThreadLimit to %d.", MAX_THREAD_LIMIT);
       thread_limit = MAX_THREAD_LIMIT;
    } 
    else if (thread_limit < 1) {
	ap_log_error(APLOG_MARK, APLOG_STARTUP, 0, NULL, 
                     "WARNING: Require ThreadLimit > 0, setting to 1");
	thread_limit = 1;
    }
    return NULL;
}

static const command_rec threadpool_cmds[] = {
UNIX_DAEMON_COMMANDS,
LISTEN_COMMANDS,
AP_INIT_TAKE1("StartServers", set_daemons_to_start, NULL, RSRC_CONF,
  "Number of child processes launched at server startup"),
AP_INIT_TAKE1("MinSpareThreads", set_min_spare_threads, NULL, RSRC_CONF,
  "Minimum number of idle children, to handle request spikes"),
AP_INIT_TAKE1("MaxSpareThreads", set_max_spare_threads, NULL, RSRC_CONF,
  "Maximum number of idle children"),
AP_INIT_TAKE1("MaxClients", set_max_clients, NULL, RSRC_CONF,
  "Maximum number of children alive at the same time"),
AP_INIT_TAKE1("ThreadsPerChild", set_threads_per_child, NULL, RSRC_CONF,
  "Number of threads each child creates"),
AP_INIT_TAKE1("ServerLimit", set_server_limit, NULL, RSRC_CONF,
  "Maximum value of MaxClients for this run of Apache"),
AP_INIT_TAKE1("ThreadLimit", set_thread_limit, NULL, RSRC_CONF,
  "Maximum worker threads in a server for this run of Apache"),
{ NULL }
};

module AP_MODULE_DECLARE_DATA mpm_threadpool_module = {
    MPM20_MODULE_STUFF,
    ap_mpm_rewrite_args,        /* hook to run before apache parses args */
    NULL,                       /* create per-directory config structure */
    NULL,                       /* merge per-directory config structures */
    NULL,                       /* create per-server config structure */
    NULL,                       /* merge per-server config structures */
    threadpool_cmds,            /* command apr_table_t */
    threadpool_hooks            /* register_hooks */
};