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
|
--- linux-2.4.18-pre9/fs/reiserfs/stree.c Thu Feb 7 13:04:22 2002
+++ linux-akpm/fs/reiserfs/stree.c Mon Feb 11 15:28:07 2002
@@ -648,9 +648,8 @@ int search_by_key (struct super_block *
stop at leaf level - set to
DISK_LEAF_NODE_LEVEL */
) {
- int n_block_number = SB_ROOT_BLOCK (p_s_sb),
- expected_level = SB_TREE_HEIGHT (p_s_sb),
- n_block_size = p_s_sb->s_blocksize;
+ int n_block_number, expected_level;
+ int n_block_size = p_s_sb->s_blocksize;
struct buffer_head * p_s_bh;
struct path_element * p_s_last_element;
int n_node_level, n_retval;
@@ -662,7 +661,8 @@ int search_by_key (struct super_block *
#endif
PROC_INFO_INC( p_s_sb, search_by_key );
-
+ conditional_schedule();
+
/* As we add each node to a path we increase its count. This means that
we must be careful to release all nodes in a path before we either
discard the path struct or re-use the path struct, as we do here. */
@@ -674,6 +674,8 @@ int search_by_key (struct super_block *
/* With each iteration of this loop we search through the items in the
current node, and calculate the next current node(next path element)
for the next iteration of this loop.. */
+ n_block_number = SB_ROOT_BLOCK (p_s_sb);
+ expected_level = SB_TREE_HEIGHT (p_s_sb);
while ( 1 ) {
#ifdef CONFIG_REISERFS_CHECK
@@ -1099,6 +1101,8 @@ static char prepare_for_delete_or_cut(
for (n_counter = *p_n_removed;
n_counter < n_unfm_number; n_counter++, p_n_unfm_pointer-- ) {
+
+ conditional_schedule();
if (item_moved (&s_ih, p_s_path)) {
need_research = 1 ;
--- linux-2.4.18-pre9/include/linux/low-latency.h Thu Jan 1 00:00:00 1970
+++ linux-akpm/include/linux/low-latency.h Mon Feb 11 15:28:07 2002
@@ -0,0 +1,109 @@
+/*
+ * include/linux/low-latency.h
+ *
+ * Andrew Morton <akpm@zip.com.au>
+ */
+
+#ifndef LOW_LATENCY_H_INCLUDED
+#define LOW_LATENCY_H_INCLUDED
+
+#if defined(CONFIG_LOLAT)
+#define LOWLATENCY_NEEDED 1
+#else
+#define LOWLATENCY_NEEDED 0
+#endif
+
+#if LOWLATENCY_NEEDED
+
+#include <linux/cache.h> /* For ____cacheline_aligned */
+
+#ifdef CONFIG_LOLAT_SYSCTL
+extern struct low_latency_enable_struct {
+ int yep;
+} ____cacheline_aligned __enable_lowlatency;
+#define enable_lowlatency __enable_lowlatency.yep
+
+#else
+#define enable_lowlatency 1
+#endif
+
+/*
+ * Set this non-zero to generate low-latency instrumentation
+ */
+#define LOWLATENCY_DEBUG 0
+
+/*
+ * Set this non-zero for robustness testing
+ */
+#define LOWLATENCY_ALWAYS_SCHEDULE 0
+
+#if LOWLATENCY_DEBUG
+
+#if LOWLATENCY_ALWAYS_SCHEDULE
+#define conditional_schedule_needed() ((enable_lowlatency == 2) || (enable_lowlatency && current->need_resched))
+#else
+#define conditional_schedule_needed() (enable_lowlatency && current->need_resched)
+#endif
+
+struct lolat_stats_t {
+ unsigned long count;
+ int visited;
+ const char *file;
+ int line;
+ struct lolat_stats_t *next;
+};
+
+void set_running_and_schedule(struct lolat_stats_t *stats);
+
+#define unconditional_schedule() \
+ do { \
+ static struct lolat_stats_t stats = { \
+ file: __FILE__, \
+ line: __LINE__, \
+ }; \
+ set_running_and_schedule(&stats); \
+ } while (0)
+
+extern void show_lolat_stats(void);
+
+#else /* LOWLATENCY_DEBUG */
+
+#if LOWLATENCY_ALWAYS_SCHEDULE
+#define conditional_schedule_needed() 1
+#else
+#define conditional_schedule_needed() (current->need_resched)
+#endif
+
+void set_running_and_schedule(void);
+#define unconditional_schedule() set_running_and_schedule()
+
+#endif /* LOWLATENCY_DEBUG */
+
+#define conditional_schedule() \
+ do { \
+ if (conditional_schedule_needed()) \
+ unconditional_schedule(); \
+ } while (0)
+
+#define DEFINE_RESCHED_COUNT int resched_count = 0
+#define TEST_RESCHED_COUNT(n) (enable_lowlatency && (++resched_count > (n)))
+#define RESET_RESCHED_COUNT() resched_count = 0
+extern int ll_copy_to_user(void *to_user, const void *from, unsigned long len);
+extern int ll_copy_from_user(void *to, const void *from_user, unsigned long len);
+
+#else /* LOWLATENCY_NEEDED */
+
+#define conditional_schedule_needed() 0
+#define conditional_schedule()
+#define unconditional_schedule()
+
+#define DEFINE_RESCHED_COUNT
+#define TEST_RESCHED_COUNT(n) 0
+#define RESET_RESCHED_COUNT()
+#define ll_copy_to_user(to_user, from, len) copy_to_user((to_user), (from), (len))
+#define ll_copy_from_user(to, from_user, len) copy_from_user((to), (from_user), (len))
+
+#endif /* LOWLATENCY_NEEDED */
+
+#endif /* LOW_LATENCY_H_INCLUDED */
+
--- linux-2.4.18-pre9/include/linux/mm.h Fri Dec 21 11:19:23 2001
+++ linux-akpm/include/linux/mm.h Mon Feb 11 15:28:07 2002
@@ -121,6 +121,8 @@ extern int vm_max_readahead;
*/
extern pgprot_t protection_map[16];
+/* Actions for zap_page_range() */
+#define ZPR_COND_RESCHED 1 /* Do a conditional_schedule() occasionally */
/*
* These are the virtual MM functions - opening of an area, closing and
@@ -404,7 +406,7 @@ struct file *shmem_file_setup(char * nam
extern void shmem_lock(struct file * file, int lock);
extern int shmem_zero_setup(struct vm_area_struct *);
-extern void zap_page_range(struct mm_struct *mm, unsigned long address, unsigned long size);
+extern void zap_page_range(struct mm_struct *mm, unsigned long address, unsigned long size, int actions);
extern int copy_page_range(struct mm_struct *dst, struct mm_struct *src, struct vm_area_struct *vma);
extern int remap_page_range(unsigned long from, unsigned long to, unsigned long size, pgprot_t prot);
extern int zeromap_page_range(unsigned long from, unsigned long size, pgprot_t prot);
--- linux-2.4.18-pre9/include/linux/reiserfs_fs.h Fri Dec 21 11:19:23 2001
+++ linux-akpm/include/linux/reiserfs_fs.h Mon Feb 11 15:28:07 2002
@@ -1155,8 +1155,8 @@ static inline loff_t max_reiserfs_offset
#define fs_generation(s) ((s)->u.reiserfs_sb.s_generation_counter)
#define get_generation(s) atomic_read (&fs_generation(s))
#define FILESYSTEM_CHANGED_TB(tb) (get_generation((tb)->tb_sb) != (tb)->fs_gen)
-#define fs_changed(gen,s) (gen != get_generation (s))
-
+#define __fs_changed(gen,s) (gen != get_generation (s))
+#define fs_changed(gen,s) ({conditional_schedule(); __fs_changed(gen,s);})
/***************************************************************************/
/* FIXATE NODES */
--- linux-2.4.18-pre9/include/linux/sched.h Fri Dec 21 11:19:23 2001
+++ linux-akpm/include/linux/sched.h Mon Feb 11 15:28:07 2002
@@ -26,6 +26,7 @@ extern unsigned long event;
#include <linux/signal.h>
#include <linux/securebits.h>
#include <linux/fs_struct.h>
+#include <linux/low-latency.h>
struct exec_domain;
--- linux-2.4.18-pre9/include/linux/sysctl.h Mon Nov 26 11:52:07 2001
+++ linux-akpm/include/linux/sysctl.h Mon Feb 11 15:28:07 2002
@@ -124,6 +124,7 @@ enum
KERN_CORE_USES_PID=52, /* int: use core or core.%pid */
KERN_TAINTED=53, /* int: various kernel tainted flags */
KERN_CADPID=54, /* int: PID of the process to notify on CAD */
+ KERN_LOWLATENCY=55, /* int: enable low latency scheduling */
};
--- linux-2.4.18-pre9/arch/i386/config.in Thu Feb 7 13:04:11 2002
+++ linux-akpm/arch/i386/config.in Mon Feb 11 15:28:07 2002
@@ -26,6 +26,9 @@ endmenu
mainmenu_option next_comment
comment 'Processor type and features'
+bool 'Low latency scheduling' CONFIG_LOLAT
+dep_bool 'Control low latency with sysctl' CONFIG_LOLAT_SYSCTL $CONFIG_LOLAT
+
choice 'Processor family' \
"386 CONFIG_M386 \
486 CONFIG_M486 \
--- linux-2.4.18-pre9/drivers/block/ll_rw_blk.c Thu Feb 7 13:04:11 2002
+++ linux-akpm/drivers/block/ll_rw_blk.c Mon Feb 11 15:28:07 2002
@@ -917,6 +917,7 @@ void submit_bh(int rw, struct buffer_hea
kstat.pgpgin += count;
break;
}
+ conditional_schedule();
}
/**
--- linux-2.4.18-pre9/drivers/char/mem.c Fri Dec 21 11:19:13 2001
+++ linux-akpm/drivers/char/mem.c Mon Feb 11 15:28:07 2002
@@ -400,7 +400,7 @@ static inline size_t read_zero_pagealign
if (count > size)
count = size;
- zap_page_range(mm, addr, count);
+ zap_page_range(mm, addr, count, 0);
zeromap_page_range(addr, count, PAGE_COPY);
size -= count;
--- linux-2.4.18-pre9/drivers/char/random.c Thu Nov 22 23:02:57 2001
+++ linux-akpm/drivers/char/random.c Mon Feb 11 15:28:07 2002
@@ -1369,6 +1369,11 @@ static ssize_t extract_entropy(struct en
buf += i;
ret += i;
add_timer_randomness(&extract_timer_state, nbytes);
+#if LOWLATENCY_NEEDED
+ /* This can happen in softirq's, but that's what we want */
+ if (conditional_schedule_needed())
+ break;
+#endif
}
/* Wipe data just returned from memory */
--- linux-2.4.18-pre9/drivers/i2c/i2c-core.c Thu Nov 22 23:02:57 2001
+++ linux-akpm/drivers/i2c/i2c-core.c Mon Feb 11 15:28:17 2002
@@ -762,6 +762,8 @@ int i2c_transfer(struct i2c_adapter * ad
{
int ret;
+ conditional_schedule();
+
if (adap->algo->master_xfer) {
DEB2(printk("i2c-core.o: master_xfer: %s with %d msgs.\n",
adap->name,num));
@@ -784,6 +786,8 @@ int i2c_master_send(struct i2c_client *c
struct i2c_adapter *adap=client->adapter;
struct i2c_msg msg;
+ conditional_schedule();
+
if (client->adapter->algo->master_xfer) {
msg.addr = client->addr;
msg.flags = client->flags & I2C_M_TEN;
@@ -813,6 +817,9 @@ int i2c_master_recv(struct i2c_client *c
struct i2c_adapter *adap=client->adapter;
struct i2c_msg msg;
int ret;
+
+ conditional_schedule();
+
if (client->adapter->algo->master_xfer) {
msg.addr = client->addr;
msg.flags = client->flags & I2C_M_TEN;
--- linux-2.4.18-pre9/fs/buffer.c Thu Feb 7 13:04:21 2002
+++ linux-akpm/fs/buffer.c Mon Feb 11 15:28:07 2002
@@ -204,8 +204,10 @@ static int write_some_buffers(kdev_t dev
if (dev && bh->b_dev != dev)
continue;
- if (test_and_set_bit(BH_Lock, &bh->b_state))
+ if (test_and_set_bit(BH_Lock, &bh->b_state)) {
+ __refile_buffer(bh);
continue;
+ }
if (atomic_set_buffer_clean(bh)) {
__refile_buffer(bh);
get_bh(bh);
@@ -215,6 +217,7 @@ static int write_some_buffers(kdev_t dev
spin_unlock(&lru_list_lock);
write_locked_buffers(array, count);
+ conditional_schedule();
return -EAGAIN;
}
unlock_buffer(bh);
@@ -249,12 +252,19 @@ static int wait_for_buffers(kdev_t dev,
struct buffer_head * next;
int nr;
- next = lru_list[index];
nr = nr_buffers_type[index];
+repeat:
+ next = lru_list[index];
while (next && --nr >= 0) {
struct buffer_head *bh = next;
next = bh->b_next_free;
+ if (conditional_schedule_needed()) {
+ spin_unlock(&lru_list_lock);
+ unconditional_schedule();
+ spin_lock(&lru_list_lock);
+ goto repeat;
+ }
if (!buffer_locked(bh)) {
if (refile)
__refile_buffer(bh);
@@ -262,7 +272,6 @@ static int wait_for_buffers(kdev_t dev,
}
if (dev && bh->b_dev != dev)
continue;
-
get_bh(bh);
spin_unlock(&lru_list_lock);
wait_on_buffer (bh);
@@ -273,7 +282,7 @@ static int wait_for_buffers(kdev_t dev,
return 0;
}
-static inline void wait_for_some_buffers(kdev_t dev)
+static void wait_for_some_buffers(kdev_t dev)
{
spin_lock(&lru_list_lock);
wait_for_buffers(dev, BUF_LOCKED, 1);
@@ -301,6 +310,15 @@ int sync_buffers(kdev_t dev, int wait)
{
int err = 0;
+#if LOWLATENCY_NEEDED
+ /*
+ * syncing devA when there are lots of buffers dirty against
+ * devB is expensive.
+ */
+ if (enable_lowlatency)
+ dev = NODEV;
+#endif
+
/* One pass for no-wait, three for wait:
* 0) write out all dirty, unlocked buffers;
* 1) wait for all dirty locked buffers;
@@ -682,6 +700,16 @@ void invalidate_bdev(struct block_device
/* Not hashed? */
if (!bh->b_pprev)
continue;
+
+ if (conditional_schedule_needed()) {
+ get_bh(bh);
+ spin_unlock(&lru_list_lock);
+ unconditional_schedule();
+ spin_lock(&lru_list_lock);
+ put_bh(bh);
+ slept = 1;
+ }
+
if (buffer_locked(bh)) {
get_bh(bh);
spin_unlock(&lru_list_lock);
@@ -833,7 +861,8 @@ int fsync_inode_buffers(struct inode *in
struct buffer_head *bh;
struct inode tmp;
int err = 0, err2;
-
+ DEFINE_RESCHED_COUNT;
+
INIT_LIST_HEAD(&tmp.i_dirty_buffers);
spin_lock(&lru_list_lock);
@@ -854,8 +883,18 @@ int fsync_inode_buffers(struct inode *in
spin_lock(&lru_list_lock);
}
}
+ if (TEST_RESCHED_COUNT(32)) {
+ RESET_RESCHED_COUNT();
+ if (conditional_schedule_needed()) {
+ spin_unlock(&lru_list_lock);
+ unconditional_schedule(); /* Syncing many dirty buffers */
+ spin_lock(&lru_list_lock);
+ }
+ }
}
+ RESET_RESCHED_COUNT();
+
while (!list_empty(&tmp.i_dirty_buffers)) {
bh = BH_ENTRY(tmp.i_dirty_buffers.prev);
remove_inode_queue(bh);
@@ -882,12 +921,18 @@ int fsync_inode_data_buffers(struct inod
struct buffer_head *bh;
struct inode tmp;
int err = 0, err2;
-
+
INIT_LIST_HEAD(&tmp.i_dirty_data_buffers);
-
+
+repeat:
spin_lock(&lru_list_lock);
while (!list_empty(&inode->i_dirty_data_buffers)) {
+ if (conditional_schedule_needed()) {
+ spin_unlock(&lru_list_lock);
+ unconditional_schedule();
+ goto repeat;
+ }
bh = BH_ENTRY(inode->i_dirty_data_buffers.next);
list_del(&bh->b_inode_buffers);
if (!buffer_dirty(bh) && !buffer_locked(bh))
@@ -914,6 +959,7 @@ int fsync_inode_data_buffers(struct inod
if (!buffer_uptodate(bh))
err = -EIO;
brelse(bh);
+ conditional_schedule();
spin_lock(&lru_list_lock);
}
@@ -942,14 +988,23 @@ int osync_inode_buffers(struct inode *in
struct buffer_head *bh;
struct list_head *list;
int err = 0;
+ DEFINE_RESCHED_COUNT;
+repeat:
+ conditional_schedule();
spin_lock(&lru_list_lock);
- repeat:
-
for (list = inode->i_dirty_buffers.prev;
bh = BH_ENTRY(list), list != &inode->i_dirty_buffers;
list = bh->b_inode_buffers.prev) {
+ if (TEST_RESCHED_COUNT(32)) {
+ RESET_RESCHED_COUNT();
+ if (conditional_schedule_needed()) {
+ spin_unlock(&lru_list_lock);
+ goto repeat;
+ }
+ }
+
if (buffer_locked(bh)) {
get_bh(bh);
spin_unlock(&lru_list_lock);
@@ -957,7 +1012,6 @@ int osync_inode_buffers(struct inode *in
if (!buffer_uptodate(bh))
err = -EIO;
brelse(bh);
- spin_lock(&lru_list_lock);
goto repeat;
}
}
@@ -971,14 +1025,23 @@ int osync_inode_data_buffers(struct inod
struct buffer_head *bh;
struct list_head *list;
int err = 0;
+ DEFINE_RESCHED_COUNT;
+repeat:
+ conditional_schedule();
spin_lock(&lru_list_lock);
-
- repeat:
for (list = inode->i_dirty_data_buffers.prev;
bh = BH_ENTRY(list), list != &inode->i_dirty_data_buffers;
list = bh->b_inode_buffers.prev) {
+ if (TEST_RESCHED_COUNT(32)) {
+ RESET_RESCHED_COUNT();
+ if (conditional_schedule_needed()) {
+ spin_unlock(&lru_list_lock);
+ goto repeat;
+ }
+ }
+
if (buffer_locked(bh)) {
get_bh(bh);
spin_unlock(&lru_list_lock);
@@ -986,7 +1049,6 @@ int osync_inode_data_buffers(struct inod
if (!buffer_uptodate(bh))
err = -EIO;
brelse(bh);
- spin_lock(&lru_list_lock);
goto repeat;
}
}
@@ -1004,12 +1066,24 @@ int osync_inode_data_buffers(struct inod
void invalidate_inode_buffers(struct inode *inode)
{
struct list_head * entry;
-
+
+repeat:
+ conditional_schedule();
spin_lock(&lru_list_lock);
- while ((entry = inode->i_dirty_buffers.next) != &inode->i_dirty_buffers)
+ while ((entry = inode->i_dirty_buffers.next) != &inode->i_dirty_buffers) {
+ if (conditional_schedule_needed()) {
+ spin_unlock(&lru_list_lock);
+ goto repeat;
+ }
remove_inode_queue(BH_ENTRY(entry));
- while ((entry = inode->i_dirty_data_buffers.next) != &inode->i_dirty_data_buffers)
+ }
+ while ((entry = inode->i_dirty_data_buffers.next) != &inode->i_dirty_data_buffers) {
+ if (conditional_schedule_needed()) {
+ spin_unlock(&lru_list_lock);
+ goto repeat;
+ }
remove_inode_queue(BH_ENTRY(entry));
+ }
spin_unlock(&lru_list_lock);
}
@@ -1184,8 +1258,10 @@ struct buffer_head * bread(kdev_t dev, i
bh = getblk(dev, block, size);
touch_buffer(bh);
- if (buffer_uptodate(bh))
+ if (buffer_uptodate(bh)) {
+ conditional_schedule();
return bh;
+ }
ll_rw_block(READ, 1, &bh);
wait_on_buffer(bh);
if (buffer_uptodate(bh))
@@ -2825,7 +2901,7 @@ void __init buffer_init(unsigned long me
DECLARE_WAIT_QUEUE_HEAD(bdflush_wait);
-void wakeup_bdflush(void)
+void wakeup_bdflush(void)
{
wake_up_interruptible(&bdflush_wait);
}
--- linux-2.4.18-pre9/fs/dcache.c Thu Feb 7 13:04:21 2002
+++ linux-akpm/fs/dcache.c Mon Feb 11 15:28:07 2002
@@ -320,11 +320,23 @@ static inline void prune_one_dentry(stru
void prune_dcache(int count)
{
+ DEFINE_RESCHED_COUNT;
+
+redo:
spin_lock(&dcache_lock);
for (;;) {
struct dentry *dentry;
struct list_head *tmp;
+ if (TEST_RESCHED_COUNT(100)) {
+ RESET_RESCHED_COUNT();
+ if (conditional_schedule_needed()) {
+ spin_unlock(&dcache_lock);
+ unconditional_schedule();
+ goto redo;
+ }
+ }
+
tmp = dentry_unused.prev;
if (tmp == &dentry_unused)
@@ -479,6 +491,7 @@ static int select_parent(struct dentry *
struct dentry *this_parent = parent;
struct list_head *next;
int found = 0;
+ DEFINE_RESCHED_COUNT;
spin_lock(&dcache_lock);
repeat:
@@ -493,6 +506,13 @@ resume:
list_add(&dentry->d_lru, dentry_unused.prev);
found++;
}
+
+ if (TEST_RESCHED_COUNT(500) && found > 10) {
+ if (conditional_schedule_needed()) /* Typically sys_rmdir() */
+ goto out;
+ RESET_RESCHED_COUNT();
+ }
+
/*
* Descend a level if the d_subdirs list is non-empty.
*/
@@ -517,6 +537,7 @@ this_parent->d_parent->d_name.name, this
#endif
goto resume;
}
+out:
spin_unlock(&dcache_lock);
return found;
}
@@ -532,8 +553,10 @@ void shrink_dcache_parent(struct dentry
{
int found;
- while ((found = select_parent(parent)) != 0)
+ while ((found = select_parent(parent)) != 0) {
prune_dcache(found);
+ conditional_schedule(); /* Typically sys_rmdir() */
+ }
}
/*
--- linux-2.4.18-pre9/fs/exec.c Fri Dec 21 11:19:14 2001
+++ linux-akpm/fs/exec.c Mon Feb 11 15:28:07 2002
@@ -223,7 +223,7 @@ int copy_strings(int argc,char ** argv,
if (new)
memset(kaddr+offset+len, 0, PAGE_SIZE-offset-len);
}
- err = copy_from_user(kaddr + offset, str, bytes_to_copy);
+ err = ll_copy_from_user(kaddr + offset, str, bytes_to_copy);
kunmap(page);
if (err)
--- linux-2.4.18-pre9/fs/ext2/dir.c Thu Feb 7 13:04:21 2002
+++ linux-akpm/fs/ext2/dir.c Mon Feb 11 15:28:07 2002
@@ -153,6 +153,7 @@ static struct page * ext2_get_page(struc
struct address_space *mapping = dir->i_mapping;
struct page *page = read_cache_page(mapping, n,
(filler_t*)mapping->a_ops->readpage, NULL);
+ conditional_schedule(); /* Scanning large directories */
if (!IS_ERR(page)) {
wait_on_page(page);
kmap(page);
--- linux-2.4.18-pre9/fs/ext2/inode.c Thu Feb 7 13:04:21 2002
+++ linux-akpm/fs/ext2/inode.c Mon Feb 11 15:28:07 2002
@@ -714,8 +714,13 @@ static inline void ext2_free_data(struct
{
unsigned long block_to_free = 0, count = 0;
unsigned long nr;
+ DEFINE_RESCHED_COUNT;
for ( ; p < q ; p++) {
+ if (TEST_RESCHED_COUNT(32)) {
+ RESET_RESCHED_COUNT();
+ conditional_schedule();
+ }
nr = le32_to_cpu(*p);
if (nr) {
*p = 0;
@@ -758,6 +763,7 @@ static void ext2_free_branches(struct in
if (depth--) {
int addr_per_block = EXT2_ADDR_PER_BLOCK(inode->i_sb);
for ( ; p < q ; p++) {
+ conditional_schedule(); /* Deleting large files */
nr = le32_to_cpu(*p);
if (!nr)
continue;
--- linux-2.4.18-pre9/fs/inode.c Fri Dec 21 11:19:14 2001
+++ linux-akpm/fs/inode.c Mon Feb 11 15:28:07 2002
@@ -229,6 +229,8 @@ static inline void __sync_one(struct ino
filemap_fdatawait(inode->i_mapping);
+ conditional_schedule();
+
spin_lock(&inode_lock);
inode->i_state &= ~I_LOCK;
if (!(inode->i_state & I_FREEING)) {
@@ -539,6 +541,7 @@ static void dispose_list(struct list_hea
while ((inode_entry = head->next) != head)
{
+ conditional_schedule();
list_del(inode_entry);
inode = list_entry(inode_entry, struct inode, i_list);
@@ -567,9 +570,22 @@ static int invalidate_list(struct list_h
if (tmp == head)
break;
inode = list_entry(tmp, struct inode, i_list);
+
+ if (conditional_schedule_needed()) {
+ atomic_inc(&inode->i_count);
+ spin_unlock(&inode_lock);
+ unconditional_schedule();
+ spin_lock(&inode_lock);
+ atomic_dec(&inode->i_count);
+ }
+
if (inode->i_sb != sb)
continue;
+ atomic_inc(&inode->i_count);
+ spin_unlock(&inode_lock);
invalidate_inode_buffers(inode);
+ spin_lock(&inode_lock);
+ atomic_dec(&inode->i_count);
if (!atomic_read(&inode->i_count)) {
list_del_init(&inode->i_hash);
list_del(&inode->i_list);
@@ -667,15 +683,28 @@ void prune_icache(int goal)
struct list_head *entry, *freeable = &list;
int count;
struct inode * inode;
+ int nr_to_scan = inodes_stat.nr_unused;
+resume:
spin_lock(&inode_lock);
-
count = 0;
entry = inode_unused.prev;
- while (entry != &inode_unused)
- {
+ while (entry != &inode_unused && nr_to_scan--) {
struct list_head *tmp = entry;
+ if (conditional_schedule_needed()) {
+ /*
+ * Need to drop the lock. Reposition
+ * the list head so we start here next time.
+ * This can corrupt the LRU nature of the
+ * unused list, but this isn't very important.
+ */
+ list_del(&inode_unused);
+ list_add(&inode_unused, entry);
+ spin_unlock(&inode_lock);
+ unconditional_schedule();
+ goto resume;
+ }
entry = entry->prev;
inode = INODE(tmp);
if (inode->i_state & (I_FREEING|I_CLEAR|I_LOCK))
@@ -841,6 +870,8 @@ static struct inode * get_new_inode(stru
inode = alloc_inode();
if (inode) {
struct inode * old;
+
+ conditional_schedule(); /* sync_old_buffers */
spin_lock(&inode_lock);
/* We released the lock, so.. */
--- linux-2.4.18-pre9/fs/proc/array.c Thu Oct 11 09:00:01 2001
+++ linux-akpm/fs/proc/array.c Mon Feb 11 15:28:07 2002
@@ -412,9 +412,11 @@ static inline void statm_pte_range(pmd_t
if (end > PMD_SIZE)
end = PMD_SIZE;
do {
- pte_t page = *pte;
+ pte_t page;
struct page *ptpage;
+ conditional_schedule(); /* For `top' and `ps' */
+ page = *pte;
address += PAGE_SIZE;
pte++;
if (pte_none(page))
--- linux-2.4.18-pre9/fs/proc/generic.c Fri Sep 7 10:53:59 2001
+++ linux-akpm/fs/proc/generic.c Mon Feb 11 15:28:07 2002
@@ -98,6 +98,8 @@ proc_file_read(struct file * file, char
retval = n;
break;
}
+
+ conditional_schedule(); /* Some /proc files are large */
/* This is a hack to allow mangling of file pos independent
* of actual bytes read. Simply place the data at page,
--- linux-2.4.18-pre9/fs/reiserfs/bitmap.c Fri Dec 21 11:19:22 2001
+++ linux-akpm/fs/reiserfs/bitmap.c Mon Feb 11 15:28:07 2002
@@ -412,17 +412,27 @@ free_and_return:
}
- reiserfs_prepare_for_journal(s, SB_AP_BITMAP(s)[i], 1) ;
+/* this check needs to go before preparing the buffer because that can
+** schedule when low-latency patches are in use. It is ok if the buffer
+** is locked, preparing it will wait on it, and we handle the case where
+** this block was allocated while we sleep below.
+*/
+ RFALSE( is_reusable (s, search_start, 0) == 0,
+ "vs-4140: bad block number found");
- RFALSE( buffer_locked (SB_AP_BITMAP (s)[i]) ||
- is_reusable (s, search_start, 0) == 0,
- "vs-4140: bitmap block is locked or bad block number found");
+ reiserfs_prepare_for_journal(s, SB_AP_BITMAP(s)[i], 1) ;
/* if this bit was already set, we've scheduled, and someone else
** has allocated it. loop around and try again
*/
if (reiserfs_test_and_set_le_bit (j, SB_AP_BITMAP (s)[i]->b_data)) {
reiserfs_restore_prepared_buffer(s, SB_AP_BITMAP(s)[i]) ;
+ /* if this block has been allocated while we slept, it is
+ ** impossible to find any more contiguous blocks for ourselves.
+ ** If we are doing preallocation, give up now and return.
+ */
+ if (for_prealloc)
+ goto free_and_return ;
amount_needed++ ;
continue ;
}
--- linux-2.4.18-pre9/fs/reiserfs/buffer2.c Fri Dec 21 11:19:22 2001
+++ linux-akpm/fs/reiserfs/buffer2.c Mon Feb 11 15:28:07 2002
@@ -55,6 +55,7 @@ struct buffer_head * reiserfs_bread (st
PROC_EXP( unsigned int ctx_switches = kstat.context_swtch );
result = bread (super -> s_dev, n_block, n_size);
+ conditional_schedule();
PROC_INFO_INC( super, breads );
PROC_EXP( if( kstat.context_swtch != ctx_switches )
PROC_INFO_INC( super, bread_miss ) );
--- linux-2.4.18-pre9/fs/reiserfs/journal.c Thu Feb 7 13:04:22 2002
+++ linux-akpm/fs/reiserfs/journal.c Mon Feb 11 15:28:07 2002
@@ -574,6 +574,7 @@ inline void insert_journal_hash(struct r
/* lock the current transaction */
inline static void lock_journal(struct super_block *p_s_sb) {
PROC_INFO_INC( p_s_sb, journal.lock_journal );
+ conditional_schedule();
while(atomic_read(&(SB_JOURNAL(p_s_sb)->j_wlock)) > 0) {
PROC_INFO_INC( p_s_sb, journal.lock_journal_wait );
sleep_on(&(SB_JOURNAL(p_s_sb)->j_wait)) ;
@@ -704,6 +705,7 @@ reiserfs_panic(s, "journal-539: flush_co
mark_buffer_dirty(tbh) ;
}
ll_rw_block(WRITE, 1, &tbh) ;
+ conditional_schedule();
count++ ;
put_bh(tbh) ; /* once for our get_hash */
}
@@ -833,6 +835,7 @@ static int _update_journal_header_block(
set_bit(BH_Dirty, &(SB_JOURNAL(p_s_sb)->j_header_bh->b_state)) ;
ll_rw_block(WRITE, 1, &(SB_JOURNAL(p_s_sb)->j_header_bh)) ;
wait_on_buffer((SB_JOURNAL(p_s_sb)->j_header_bh)) ;
+ conditional_schedule();
if (!buffer_uptodate(SB_JOURNAL(p_s_sb)->j_header_bh)) {
printk( "reiserfs: journal-837: IO error during journal replay\n" );
return -EIO ;
@@ -2089,6 +2092,7 @@ static int journal_join(struct reiserfs_
}
int journal_begin(struct reiserfs_transaction_handle *th, struct super_block * p_s_sb, unsigned long nblocks) {
+ conditional_schedule();
return do_journal_begin_r(th, p_s_sb, nblocks, 0) ;
}
@@ -2229,6 +2233,7 @@ int journal_mark_dirty_nolog(struct reis
}
int journal_end(struct reiserfs_transaction_handle *th, struct super_block *p_s_sb, unsigned long nblocks) {
+ conditional_schedule();
return do_journal_end(th, p_s_sb, nblocks, 0) ;
}
@@ -2680,6 +2685,7 @@ void reiserfs_prepare_for_journal(struct
RFALSE( buffer_locked(bh) && cur_tb != NULL,
"waiting while do_balance was running\n") ;
wait_on_buffer(bh) ;
+ conditional_schedule();
}
PROC_INFO_INC( p_s_sb, journal.prepare_retry );
retry_count++ ;
@@ -2852,6 +2858,7 @@ printk("journal-2020: do_journal_end: BA
/* copy all the real blocks into log area. dirty log blocks */
if (test_bit(BH_JDirty, &cn->bh->b_state)) {
struct buffer_head *tmp_bh ;
+ conditional_schedule();
tmp_bh = sb_getblk(p_s_sb, reiserfs_get_journal_block(p_s_sb) +
((cur_write_start + jindex) % JOURNAL_BLOCK_COUNT)) ;
mark_buffer_uptodate(tmp_bh, 1) ;
--- linux-2.4.18-pre9/fs/jbd/checkpoint.c Thu Feb 7 13:04:21 2002
+++ linux-akpm/fs/jbd/checkpoint.c Mon Feb 11 15:28:07 2002
@@ -431,7 +431,11 @@ int __journal_clean_checkpoint_list(jour
{
transaction_t *transaction, *last_transaction, *next_transaction;
int ret = 0;
+ int ll_retries = 4; /* lowlatency addition */
+restart:
+ if (ll_retries-- == 0)
+ goto out;
transaction = journal->j_checkpoint_transactions;
if (transaction == 0)
goto out;
@@ -451,6 +455,12 @@ int __journal_clean_checkpoint_list(jour
jh = next_jh;
next_jh = jh->b_cpnext;
ret += __try_to_free_cp_buf(jh);
+ if (conditional_schedule_needed()) {
+ spin_unlock(&journal_datalist_lock);
+ unconditional_schedule();
+ spin_lock(&journal_datalist_lock);
+ goto restart;
+ }
} while (jh != last_jh);
}
} while (transaction != last_transaction);
--- linux-2.4.18-pre9/fs/jbd/commit.c Thu Feb 7 13:04:21 2002
+++ linux-akpm/fs/jbd/commit.c Mon Feb 11 15:28:07 2002
@@ -212,6 +212,16 @@ write_out_data_locked:
__journal_remove_journal_head(bh);
refile_buffer(bh);
__brelse(bh);
+ if (conditional_schedule_needed()) {
+ if (commit_transaction->t_sync_datalist)
+ commit_transaction->t_sync_datalist =
+ next_jh;
+ if (bufs)
+ break;
+ spin_unlock(&journal_datalist_lock);
+ unconditional_schedule();
+ goto write_out_data;
+ }
}
}
if (bufs == ARRAY_SIZE(wbuf)) {
@@ -235,8 +245,7 @@ write_out_data_locked:
journal_brelse_array(wbuf, bufs);
lock_journal(journal);
spin_lock(&journal_datalist_lock);
- if (bufs)
- goto write_out_data_locked;
+ goto write_out_data_locked;
}
/*
@@ -272,6 +281,14 @@ sync_datalist_empty:
*/
while ((jh = commit_transaction->t_async_datalist)) {
struct buffer_head *bh = jh2bh(jh);
+ if (conditional_schedule_needed()) {
+ spin_unlock(&journal_datalist_lock);
+ unlock_journal(journal);
+ unconditional_schedule();
+ lock_journal(journal);
+ spin_lock(&journal_datalist_lock);
+ continue; /* List may have changed */
+ }
if (buffer_locked(bh)) {
spin_unlock(&journal_datalist_lock);
unlock_journal(journal);
@@ -486,6 +503,8 @@ start_journal_io:
wait_for_iobuf:
while (commit_transaction->t_iobuf_list != NULL) {
struct buffer_head *bh;
+
+ conditional_schedule();
jh = commit_transaction->t_iobuf_list->b_tprev;
bh = jh2bh(jh);
if (buffer_locked(bh)) {
@@ -622,6 +641,8 @@ skip_commit:
while (commit_transaction->t_forget) {
transaction_t *cp_transaction;
struct buffer_head *bh;
+
+ conditional_schedule(); /* journal is locked */
jh = commit_transaction->t_forget;
J_ASSERT_JH(jh, jh->b_transaction == commit_transaction ||
--- linux-2.4.18-pre9/fs/ext3/balloc.c Thu Feb 7 13:04:21 2002
+++ linux-akpm/fs/ext3/balloc.c Mon Feb 11 15:28:07 2002
@@ -363,6 +363,9 @@ do_more:
}
}
#endif
+ /* superblock lock is held, so this is safe */
+ conditional_schedule();
+
BUFFER_TRACE(bitmap_bh, "clear bit");
if (!ext3_clear_bit (bit + i, bitmap_bh->b_data)) {
ext3_error (sb, __FUNCTION__,
--- linux-2.4.18-pre9/fs/ext3/inode.c Thu Feb 7 13:04:21 2002
+++ linux-akpm/fs/ext3/inode.c Mon Feb 11 15:28:07 2002
@@ -903,6 +903,8 @@ struct buffer_head *ext3_bread(handle_t
prev_blocks = inode->i_blocks;
+ conditional_schedule(); /* Reading large directories */
+
bh = ext3_getblk (handle, inode, block, create, err);
if (!bh)
return bh;
@@ -1595,6 +1597,7 @@ ext3_clear_blocks(handle_t *handle, stru
*/
for (p = first; p < last; p++) {
u32 nr = le32_to_cpu(*p);
+ conditional_schedule();
if (nr) {
struct buffer_head *bh;
@@ -1649,6 +1652,7 @@ static void ext3_free_data(handle_t *han
}
for (p = first; p < last; p++) {
+ conditional_schedule();
nr = le32_to_cpu(*p);
if (nr) {
/* accumulate blocks to free if they're contiguous */
--- linux-2.4.18-pre9/fs/ext3/namei.c Thu Nov 22 23:02:58 2001
+++ linux-akpm/fs/ext3/namei.c Mon Feb 11 15:28:07 2002
@@ -157,6 +157,7 @@ restart:
if ((bh = bh_use[ra_ptr++]) == NULL)
goto next;
wait_on_buffer(bh);
+ conditional_schedule();
if (!buffer_uptodate(bh)) {
/* read error, skip block & hope for the best */
brelse(bh);
--- linux-2.4.18-pre9/kernel/exit.c Thu Feb 7 13:04:22 2002
+++ linux-akpm/kernel/exit.c Mon Feb 11 15:28:07 2002
@@ -203,6 +203,7 @@ static inline void close_files(struct fi
}
i++;
set >>= 1;
+ conditional_schedule(); /* sys_exit, many files open */
}
}
}
--- linux-2.4.18-pre9/kernel/ksyms.c Thu Feb 7 13:04:22 2002
+++ linux-akpm/kernel/ksyms.c Mon Feb 11 15:28:07 2002
@@ -443,6 +443,13 @@ EXPORT_SYMBOL(xtime);
EXPORT_SYMBOL(do_gettimeofday);
EXPORT_SYMBOL(do_settimeofday);
+#if LOWLATENCY_NEEDED
+EXPORT_SYMBOL(set_running_and_schedule);
+#ifdef CONFIG_LOLAT_SYSCTL
+EXPORT_SYMBOL(__enable_lowlatency);
+#endif
+#endif
+
#if !defined(__ia64__)
EXPORT_SYMBOL(loops_per_jiffy);
#endif
--- linux-2.4.18-pre9/kernel/module.c Thu Nov 22 23:02:59 2001
+++ linux-akpm/kernel/module.c Mon Feb 11 15:28:07 2002
@@ -1174,6 +1174,11 @@ static void *s_start(struct seq_file *m,
return ERR_PTR(-ENOMEM);
lock_kernel();
for (v = module_list, n = *pos; v; n -= v->nsyms, v = v->next) {
+#if 0
+ /* We can't actually do this, because we'd create a
+ * race against module unload. Need a semaphore. */
+ conditional_schedule();
+#endif
if (n < v->nsyms) {
p->mod = v;
p->index = n;
--- linux-2.4.18-pre9/kernel/sched.c Fri Dec 21 11:19:23 2001
+++ linux-akpm/kernel/sched.c Mon Feb 11 15:28:07 2002
@@ -302,6 +302,17 @@ send_now_idle:
if (tsk->processor != this_cpu)
smp_send_reschedule(tsk->processor);
}
+#if LOWLATENCY_NEEDED
+ if (enable_lowlatency && (p->policy != SCHED_OTHER)) {
+ struct task_struct *t;
+ for (i = 0; i < smp_num_cpus; i++) {
+ cpu = cpu_logical_map(i);
+ t = cpu_curr(cpu);
+ if (t != tsk)
+ t->need_resched = 1;
+ }
+ }
+#endif
return;
@@ -1342,3 +1353,93 @@ void __init sched_init(void)
atomic_inc(&init_mm.mm_count);
enter_lazy_tlb(&init_mm, current, cpu);
}
+
+#if LOWLATENCY_NEEDED
+#if LOWLATENCY_DEBUG
+
+static struct lolat_stats_t *lolat_stats_head;
+static spinlock_t lolat_stats_lock = SPIN_LOCK_UNLOCKED;
+
+void set_running_and_schedule(struct lolat_stats_t *stats)
+{
+ spin_lock(&lolat_stats_lock);
+ if (stats->visited == 0) {
+ stats->visited = 1;
+ stats->next = lolat_stats_head;
+ lolat_stats_head = stats;
+ }
+ stats->count++;
+ spin_unlock(&lolat_stats_lock);
+
+ if (current->state != TASK_RUNNING)
+ set_current_state(TASK_RUNNING);
+ schedule();
+}
+
+void show_lolat_stats(void)
+{
+ struct lolat_stats_t *stats = lolat_stats_head;
+
+ printk("Low latency scheduling stats:\n");
+ while (stats) {
+ printk("%s:%d: %lu\n", stats->file, stats->line, stats->count);
+ stats->count = 0;
+ stats = stats->next;
+ }
+}
+
+#else /* LOWLATENCY_DEBUG */
+
+void set_running_and_schedule()
+{
+ if (current->state != TASK_RUNNING)
+ __set_current_state(TASK_RUNNING);
+ schedule();
+}
+
+#endif /* LOWLATENCY_DEBUG */
+
+int ll_copy_to_user(void *to_user, const void *from, unsigned long len)
+{
+ while (len) {
+ unsigned long n_to_copy = len;
+ unsigned long remainder;
+
+ if (n_to_copy > 4096)
+ n_to_copy = 4096;
+ remainder = copy_to_user(to_user, from, n_to_copy);
+ if (remainder)
+ return remainder + len;
+ to_user = ((char *)to_user) + n_to_copy;
+ from = ((char *)from) + n_to_copy;
+ len -= n_to_copy;
+ conditional_schedule();
+ }
+ return 0;
+}
+
+int ll_copy_from_user(void *to, const void *from_user, unsigned long len)
+{
+ while (len) {
+ unsigned long n_to_copy = len;
+ unsigned long remainder;
+
+ if (n_to_copy > 4096)
+ n_to_copy = 4096;
+ remainder = copy_from_user(to, from_user, n_to_copy);
+ if (remainder)
+ return remainder + len;
+ to = ((char *)to) + n_to_copy;
+ from_user = ((char *)from_user) + n_to_copy;
+ len -= n_to_copy;
+ conditional_schedule();
+ }
+ return 0;
+}
+
+#ifdef CONFIG_LOLAT_SYSCTL
+struct low_latency_enable_struct __enable_lowlatency = { 0, };
+#endif
+
+#endif /* LOWLATENCY_NEEDED */
+
--- linux-2.4.18-pre9/kernel/sysctl.c Fri Dec 21 11:19:23 2001
+++ linux-akpm/kernel/sysctl.c Mon Feb 11 15:28:07 2002
@@ -256,6 +256,10 @@ static ctl_table kern_table[] = {
{KERN_S390_USER_DEBUG_LOGGING,"userprocess_debug",
&sysctl_userprocess_debug,sizeof(int),0644,NULL,&proc_dointvec},
#endif
+#ifdef CONFIG_LOLAT_SYSCTL
+ {KERN_LOWLATENCY, "lowlatency", &enable_lowlatency, sizeof (int),
+ 0644, NULL, &proc_dointvec},
+#endif
{0}
};
--- linux-2.4.18-pre9/mm/filemap.c Thu Feb 7 13:04:22 2002
+++ linux-akpm/mm/filemap.c Mon Feb 11 15:28:07 2002
@@ -187,6 +187,18 @@ void invalidate_inode_pages(struct inode
page = list_entry(curr, struct page, list);
curr = curr->next;
+ if (conditional_schedule_needed() && !TryLockPage(page)) {
+ /*
+ * Page is locked. We can drop the spinlocks
+ */
+ spin_unlock(&pagecache_lock);
+ spin_unlock(&pagemap_lru_lock);
+ unconditional_schedule();
+ spin_lock(&pagemap_lru_lock);
+ spin_lock(&pagecache_lock);
+ UnlockPage(page);
+ }
+
/* We cannot invalidate something in dirty.. */
if (PageDirty(page))
continue;
@@ -256,12 +268,24 @@ static int truncate_list_pages(struct li
struct list_head *curr;
struct page * page;
int unlocked = 0;
+ int restart_count = 4;
restart:
curr = head->prev;
while (curr != head) {
unsigned long offset;
+ if (conditional_schedule_needed() && restart_count) {
+ restart_count--;
+ list_del(head);
+ list_add(head, curr); /* Restart on this page */
+ spin_unlock(&pagecache_lock);
+ unconditional_schedule();
+ spin_lock(&pagecache_lock);
+ unlocked = 1;
+ goto restart;
+ }
+
page = list_entry(curr, struct page, list);
offset = page->index;
@@ -294,13 +318,11 @@ static int truncate_list_pages(struct li
} else
wait_on_page(page);
- page_cache_release(page);
-
- if (current->need_resched) {
- __set_current_state(TASK_RUNNING);
- schedule();
+ if (LOWLATENCY_NEEDED) {
+ restart_count = 4; /* We made progress */
}
+ page_cache_release(page);
spin_lock(&pagecache_lock);
goto restart;
}
@@ -474,6 +496,7 @@ static int do_buffer_fdatasync(struct li
page_cache_get(page);
spin_unlock(&pagecache_lock);
+ conditional_schedule(); /* sys_msync() (only used by minixfs, udf) */
lock_page(page);
/* The buffers could have been free'd while we waited for the page lock */
@@ -560,12 +583,14 @@ int filemap_fdatasync(struct address_spa
list_del(&page->list);
list_add(&page->list, &mapping->locked_pages);
- if (!PageDirty(page))
- continue;
-
page_cache_get(page);
spin_unlock(&pagecache_lock);
+ conditional_schedule(); /* sys_msync() */
+
+ if (!PageDirty(page))
+ goto clean;
+
lock_page(page);
if (PageDirty(page)) {
@@ -576,7 +601,7 @@ int filemap_fdatasync(struct address_spa
ret = err;
} else
UnlockPage(page);
-
+clean:
page_cache_release(page);
spin_lock(&pagecache_lock);
}
@@ -594,7 +619,8 @@ int filemap_fdatasync(struct address_spa
int filemap_fdatawait(struct address_space * mapping)
{
int ret = 0;
-
+ DEFINE_RESCHED_COUNT;
+restart:
spin_lock(&pagecache_lock);
while (!list_empty(&mapping->locked_pages)) {
@@ -603,6 +629,17 @@ int filemap_fdatawait(struct address_spa
list_del(&page->list);
list_add(&page->list, &mapping->clean_pages);
+ if (TEST_RESCHED_COUNT(32)) {
+ RESET_RESCHED_COUNT();
+ if (conditional_schedule_needed()) {
+ page_cache_get(page);
+ spin_unlock(&pagecache_lock);
+ unconditional_schedule();
+ page_cache_release(page);
+ goto restart;
+ }
+ }
+
if (!PageLocked(page))
continue;
@@ -703,8 +740,10 @@ static int page_cache_read(struct file *
spin_lock(&pagecache_lock);
page = __find_page_nolock(mapping, offset, *hash);
spin_unlock(&pagecache_lock);
- if (page)
+ if (page) {
+ conditional_schedule();
return 0;
+ }
page = page_cache_alloc(mapping);
if (!page)
@@ -869,6 +908,11 @@ static struct page * __find_lock_page_he
* the hash-list needs a held write-lock.
*/
repeat:
+ if (conditional_schedule_needed()) {
+ spin_unlock(&pagecache_lock);
+ unconditional_schedule();
+ spin_lock(&pagecache_lock);
+ }
page = __find_page_nolock(mapping, offset, hash);
if (page) {
page_cache_get(page);
@@ -1333,6 +1377,8 @@ found_page:
page_cache_get(page);
spin_unlock(&pagecache_lock);
+ conditional_schedule(); /* sys_read() */
+
if (!Page_Uptodate(page))
goto page_not_up_to_date;
generic_file_readahead(reada_ok, filp, inode, page);
@@ -2031,6 +2077,12 @@ static inline int filemap_sync_pte_range
address += PAGE_SIZE;
pte++;
} while (address && (address < end));
+
+ if (conditional_schedule_needed()) {
+ spin_unlock(&vma->vm_mm->page_table_lock);
+ unconditional_schedule(); /* syncing large mapped files */
+ spin_lock(&vma->vm_mm->page_table_lock);
+ }
return error;
}
@@ -2438,7 +2490,9 @@ static long madvise_dontneed(struct vm_a
if (vma->vm_flags & VM_LOCKED)
return -EINVAL;
- zap_page_range(vma->vm_mm, start, end - start);
+ zap_page_range(vma->vm_mm, start, end - start,
+ ZPR_COND_RESCHED); /* sys_madvise(MADV_DONTNEED) */
+
return 0;
}
@@ -3003,6 +3057,9 @@ generic_file_write(struct file *file,con
goto sync_failure;
page_fault = __copy_from_user(kaddr+offset, buf, bytes);
flush_dcache_page(page);
+
+ conditional_schedule();
+
status = mapping->a_ops->commit_write(file, page, offset, offset+bytes);
if (page_fault)
goto fail_write;
--- linux-2.4.18-pre9/mm/memory.c Fri Dec 21 11:19:23 2001
+++ linux-akpm/mm/memory.c Mon Feb 11 15:28:07 2002
@@ -355,7 +355,7 @@ static inline int zap_pmd_range(mmu_gath
/*
* remove user pages in a given range.
*/
-void zap_page_range(struct mm_struct *mm, unsigned long address, unsigned long size)
+static void do_zap_page_range(struct mm_struct *mm, unsigned long address, unsigned long size)
{
mmu_gather_t *tlb;
pgd_t * dir;
@@ -465,6 +465,10 @@ int get_user_pages(struct task_struct *t
struct page *map;
while (!(map = follow_page(mm, start, write))) {
spin_unlock(&mm->page_table_lock);
+
+ /* Pinning down many physical pages (kiobufs, mlockall) */
+ conditional_schedule();
+
switch (handle_mm_fault(mm, vma, start, write)) {
case 1:
tsk->min_flt++;
@@ -609,6 +613,21 @@ void unmap_kiobuf (struct kiobuf *iobuf)
iobuf->locked = 0;
}
+#define MAX_ZAP_BYTES 256*PAGE_SIZE
+
+void zap_page_range(struct mm_struct *mm, unsigned long address, unsigned long size, int actions)
+{
+ while (size) {
+ unsigned long chunk = size;
+ if (actions & ZPR_COND_RESCHED && chunk > MAX_ZAP_BYTES)
+ chunk = MAX_ZAP_BYTES;
+ do_zap_page_range(mm, address, chunk);
+ if (actions & ZPR_COND_RESCHED)
+ conditional_schedule();
+ address += chunk;
+ size -= chunk;
+ }
+}
/*
* Lock down all of the pages of a kiovec for IO.
@@ -718,11 +737,18 @@ int unlock_kiovec(int nr, struct kiobuf
return 0;
}
-static inline void zeromap_pte_range(pte_t * pte, unsigned long address,
- unsigned long size, pgprot_t prot)
+static inline void zeromap_pte_range(struct mm_struct *mm, pte_t * pte,
+ unsigned long address, unsigned long size,
+ pgprot_t prot)
{
unsigned long end;
+ if (conditional_schedule_needed()) {
+ spin_unlock(&mm->page_table_lock);
+ unconditional_schedule(); /* mmap(/dev/zero) */
+ spin_lock(&mm->page_table_lock);
+ }
+
address &= ~PMD_MASK;
end = address + size;
if (end > PMD_SIZE)
@@ -750,7 +776,7 @@ static inline int zeromap_pmd_range(stru
pte_t * pte = pte_alloc(mm, pmd, address);
if (!pte)
return -ENOMEM;
- zeromap_pte_range(pte, address, end - address, prot);
+ zeromap_pte_range(mm, pte, address, end - address, prot);
address = (address + PMD_SIZE) & PMD_MASK;
pmd++;
} while (address && (address < end));
@@ -984,7 +1010,7 @@ static void vmtruncate_list(struct vm_ar
/* mapping wholly truncated? */
if (mpnt->vm_pgoff >= pgoff) {
- zap_page_range(mm, start, len);
+ zap_page_range(mm, start, len, 0);
continue;
}
@@ -997,7 +1023,7 @@ static void vmtruncate_list(struct vm_ar
/* Ok, partially affected.. */
start += diff << PAGE_SHIFT;
len = (len - diff) << PAGE_SHIFT;
- zap_page_range(mm, start, len);
+ zap_page_range(mm, start, len, 0);
} while ((mpnt = mpnt->vm_next_share) != NULL);
}
--- linux-2.4.18-pre9/mm/mmap.c Mon Nov 5 21:01:12 2001
+++ linux-akpm/mm/mmap.c Mon Feb 11 15:28:07 2002
@@ -569,7 +569,7 @@ unmap_and_free_vma:
fput(file);
/* Undo any partial mapping done by a device driver. */
- zap_page_range(mm, vma->vm_start, vma->vm_end - vma->vm_start);
+ zap_page_range(mm, vma->vm_start, vma->vm_end - vma->vm_start, 0);
free_vma:
kmem_cache_free(vm_area_cachep, vma);
return error;
@@ -967,7 +967,7 @@ int do_munmap(struct mm_struct *mm, unsi
remove_shared_vm_struct(mpnt);
mm->map_count--;
- zap_page_range(mm, st, size);
+ zap_page_range(mm, st, size, ZPR_COND_RESCHED); /* sys_munmap() */
/*
* Fix the mapping, and free the old area if it wasn't reused.
@@ -1127,7 +1127,7 @@ void exit_mmap(struct mm_struct * mm)
}
mm->map_count--;
remove_shared_vm_struct(mpnt);
- zap_page_range(mm, start, size);
+ zap_page_range(mm, start, size, ZPR_COND_RESCHED); /* sys_exit() */
if (mpnt->vm_file)
fput(mpnt->vm_file);
kmem_cache_free(vm_area_cachep, mpnt);
--- linux-2.4.18-pre9/mm/mremap.c Thu Sep 20 20:31:26 2001
+++ linux-akpm/mm/mremap.c Mon Feb 11 15:28:07 2002
@@ -118,7 +118,7 @@ oops_we_failed:
flush_cache_range(mm, new_addr, new_addr + len);
while ((offset += PAGE_SIZE) < len)
move_one_page(mm, new_addr + offset, old_addr + offset);
- zap_page_range(mm, new_addr, len);
+ zap_page_range(mm, new_addr, len, 0);
return -1;
}
--- linux-2.4.18-pre9/mm/slab.c Fri Dec 21 11:19:23 2001
+++ linux-akpm/mm/slab.c Mon Feb 11 15:28:07 2002
@@ -937,6 +937,7 @@ static int __kmem_cache_shrink(kmem_cach
spin_unlock_irq(&cachep->spinlock);
kmem_slab_destroy(cachep, slabp);
+ conditional_schedule(); /* Can take 30 milliseconds */
spin_lock_irq(&cachep->spinlock);
}
ret = !list_empty(&cachep->slabs_full) || !list_empty(&cachep->slabs_partial);
@@ -1821,6 +1822,7 @@ perfect:
*/
spin_unlock_irq(&best_cachep->spinlock);
kmem_slab_destroy(best_cachep, slabp);
+ conditional_schedule(); /* try_to_free_pages() */
spin_lock_irq(&best_cachep->spinlock);
}
spin_unlock_irq(&best_cachep->spinlock);
--- linux-2.4.18-pre9/mm/swapfile.c Thu Feb 7 13:04:22 2002
+++ linux-akpm/mm/swapfile.c Mon Feb 11 15:28:07 2002
@@ -823,7 +823,7 @@ int get_swaparea_info(char *buf)
len += sprintf(buf + len, "partition\t");
usedswap = 0;
- for (j = 0; j < ptr->max; ++j)
+ for (j = 0; j < ptr->max; ++j) {
switch (ptr->swap_map[j]) {
case SWAP_MAP_BAD:
case 0:
@@ -831,6 +831,8 @@ int get_swaparea_info(char *buf)
default:
usedswap++;
}
+ conditional_schedule();
+ }
len += sprintf(buf + len, "%d\t%d\t%d\n", ptr->pages << (PAGE_SHIFT - 10),
usedswap << (PAGE_SHIFT - 10), ptr->prio);
}
@@ -1124,6 +1126,11 @@ void si_swapinfo(struct sysinfo *val)
if (swap_info[i].flags != SWP_USED)
continue;
for (j = 0; j < swap_info[i].max; ++j) {
+ if (conditional_schedule_needed()) {
+ swap_list_unlock();
+ conditional_schedule();
+ swap_list_lock();
+ }
switch (swap_info[i].swap_map[j]) {
case 0:
case SWAP_MAP_BAD:
--- linux-2.4.18-pre9/mm/vmscan.c Thu Feb 7 13:04:22 2002
+++ linux-akpm/mm/vmscan.c Mon Feb 11 15:28:07 2002
@@ -157,6 +157,7 @@ static inline int swap_out_pmd(struct mm
{
pte_t * pte;
unsigned long pmd_end;
+ DEFINE_RESCHED_COUNT;
if (pmd_none(*dir))
return count;
@@ -182,11 +183,17 @@ static inline int swap_out_pmd(struct mm
address += PAGE_SIZE;
break;
}
+ if (TEST_RESCHED_COUNT(4)) {
+ if (conditional_schedule_needed())
+ goto out;
+ RESET_RESCHED_COUNT();
+ }
}
}
address += PAGE_SIZE;
pte++;
} while (address && (address < end));
+out:
mm->swap_address = address;
return count;
}
@@ -215,6 +222,8 @@ static inline int swap_out_pgd(struct mm
count = swap_out_pmd(mm, vma, pmd, address, end, count, classzone);
if (!count)
break;
+ if (conditional_schedule_needed())
+ return count;
address = (address + PMD_SIZE) & PMD_MASK;
pmd++;
} while (address && (address < end));
@@ -240,6 +249,8 @@ static inline int swap_out_vma(struct mm
count = swap_out_pgd(mm, vma, pgdir, address, end, count, classzone);
if (!count)
break;
+ if (conditional_schedule_needed())
+ return count;
address = (address + PGDIR_SIZE) & PGDIR_MASK;
pgdir++;
} while (address && (address < end));
@@ -261,6 +272,7 @@ static inline int swap_out_mm(struct mm_
* Find the proper vm-area after freezing the vma chain
* and ptes.
*/
+continue_scan:
spin_lock(&mm->page_table_lock);
address = mm->swap_address;
if (address == TASK_SIZE || swap_mm != mm) {
@@ -278,6 +290,12 @@ static inline int swap_out_mm(struct mm_
vma = vma->vm_next;
if (!vma)
break;
+ if (conditional_schedule_needed()) { /* Scanning a large vma */
+ spin_unlock(&mm->page_table_lock);
+ unconditional_schedule();
+ /* Continue from where we left off */
+ goto continue_scan;
+ }
if (!count)
goto out_unlock;
address = vma->vm_start;
--- linux-2.4.18-pre9/net/core/iovec.c Mon Sep 10 07:57:00 2001
+++ linux-akpm/net/core/iovec.c Mon Feb 11 15:28:07 2002
@@ -88,7 +88,7 @@ int memcpy_toiovec(struct iovec *iov, un
if(iov->iov_len)
{
int copy = min_t(unsigned int, iov->iov_len, len);
- if (copy_to_user(iov->iov_base, kdata, copy))
+ if (ll_copy_to_user(iov->iov_base, kdata, copy))
goto out;
kdata+=copy;
len-=copy;
--- linux-2.4.18-pre9/net/ipv4/tcp_minisocks.c Mon Oct 1 09:19:57 2001
+++ linux-akpm/net/ipv4/tcp_minisocks.c Mon Feb 11 15:28:07 2002
@@ -434,6 +434,9 @@ static void SMP_TIMER_NAME(tcp_twkill)(u
{
struct tcp_tw_bucket *tw;
int killed = 0;
+#if LOWLATENCY_NEEDED
+ int max_killed = 0;
+#endif
/* NOTE: compare this to previous version where lock
* was released after detaching chain. It was racy,
@@ -447,6 +450,13 @@ static void SMP_TIMER_NAME(tcp_twkill)(u
goto out;
while((tw = tcp_tw_death_row[tcp_tw_death_row_slot]) != NULL) {
+#if LOWLATENCY_NEEDED
+ /* This loop takes ~6 usecs per iteration. */
+ if (killed > 100) {
+ max_killed = 1;
+ break;
+ }
+#endif
tcp_tw_death_row[tcp_tw_death_row_slot] = tw->next_death;
tw->pprev_death = NULL;
spin_unlock(&tw_death_lock);
@@ -457,12 +467,24 @@ static void SMP_TIMER_NAME(tcp_twkill)(u
killed++;
spin_lock(&tw_death_lock);
+
+ }
+
+#if LOWLATENCY_NEEDED
+ if (max_killed) { /* More to do: do it soon */
+ mod_timer(&tcp_tw_timer, jiffies+2);
+ tcp_tw_count -= killed;
+ }
+ else
+#endif
+ {
+ tcp_tw_death_row_slot =
+ ((tcp_tw_death_row_slot + 1) & (TCP_TWKILL_SLOTS - 1));
+
+ if ((tcp_tw_count -= killed) != 0)
+ mod_timer(&tcp_tw_timer, jiffies+TCP_TWKILL_PERIOD);
}
- tcp_tw_death_row_slot =
- ((tcp_tw_death_row_slot + 1) & (TCP_TWKILL_SLOTS - 1));
- if ((tcp_tw_count -= killed) != 0)
- mod_timer(&tcp_tw_timer, jiffies+TCP_TWKILL_PERIOD);
net_statistics[smp_processor_id()*2].TimeWaited += killed;
out:
spin_unlock(&tw_death_lock);
|