Skip to content
This repository
Fetching contributors…

Octocat-spinner-32-eaf2f5

Cannot retrieve contributors at this time

file 1150 lines (1046 sloc) 42.208 kb
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
#include "redis.h"

#include <fcntl.h>
#include <pthread.h>
#include <math.h>
#include <signal.h>

/* Virtual Memory is composed mainly of two subsystems:
* - Blocking Virutal Memory
* - Threaded Virtual Memory I/O
* The two parts are not fully decoupled, but functions are split among two
* different sections of the source code (delimited by comments) in order to
* make more clear what functionality is about the blocking VM and what about
* the threaded (not blocking) VM.
*
* Redis VM design:
*
* Redis VM is a blocking VM (one that blocks reading swapped values from
* disk into memory when a value swapped out is needed in memory) that is made
* unblocking by trying to examine the command argument vector in order to
* load in background values that will likely be needed in order to exec
* the command. The command is executed only once all the relevant keys
* are loaded into memory.
*
* This basically is almost as simple of a blocking VM, but almost as parallel
* as a fully non-blocking VM.
*/

/* =================== Virtual Memory - Blocking Side ====================== */

/* Create a VM pointer object. This kind of objects are used in place of
* values in the key -> value hash table, for swapped out objects. */
vmpointer *createVmPointer(int vtype) {
    vmpointer *vp = zmalloc(sizeof(vmpointer));

    vp->type = REDIS_VMPOINTER;
    vp->storage = REDIS_VM_SWAPPED;
    vp->vtype = vtype;
    return vp;
}

void vmInit(void) {
    off_t totsize;
    int pipefds[2];
    size_t stacksize;
    struct flock fl;

    if (server.vm_max_threads != 0)
        zmalloc_enable_thread_safeness(); /* we need thread safe zmalloc() */

    redisLog(REDIS_NOTICE,"Using '%s' as swap file",server.vm_swap_file);
    /* Try to open the old swap file, otherwise create it */
    if ((server.vm_fp = fopen(server.vm_swap_file,"r+b")) == NULL) {
        server.vm_fp = fopen(server.vm_swap_file,"w+b");
    }
    if (server.vm_fp == NULL) {
        redisLog(REDIS_WARNING,
            "Can't open the swap file: %s. Exiting.",
            strerror(errno));
        exit(1);
    }
    server.vm_fd = fileno(server.vm_fp);
    /* Lock the swap file for writing, this is useful in order to avoid
* another instance to use the same swap file for a config error. */
    fl.l_type = F_WRLCK;
    fl.l_whence = SEEK_SET;
    fl.l_start = fl.l_len = 0;
    if (fcntl(server.vm_fd,F_SETLK,&fl) == -1) {
        redisLog(REDIS_WARNING,
            "Can't lock the swap file at '%s': %s. Make sure it is not used by another Redis instance.", server.vm_swap_file, strerror(errno));
        exit(1);
    }
    /* Initialize */
    server.vm_next_page = 0;
    server.vm_near_pages = 0;
    server.vm_stats_used_pages = 0;
    server.vm_stats_swapped_objects = 0;
    server.vm_stats_swapouts = 0;
    server.vm_stats_swapins = 0;
    totsize = server.vm_pages*server.vm_page_size;
    redisLog(REDIS_NOTICE,"Allocating %lld bytes of swap file",totsize);
    if (ftruncate(server.vm_fd,totsize) == -1) {
        redisLog(REDIS_WARNING,"Can't ftruncate swap file: %s. Exiting.",
            strerror(errno));
        exit(1);
    } else {
        redisLog(REDIS_NOTICE,"Swap file allocated with success");
    }
    server.vm_bitmap = zcalloc((server.vm_pages+7)/8);
    redisLog(REDIS_VERBOSE,"Allocated %lld bytes page table for %lld pages",
        (long long) (server.vm_pages+7)/8, server.vm_pages);

    /* Initialize threaded I/O (used by Virtual Memory) */
    server.io_newjobs = listCreate();
    server.io_processing = listCreate();
    server.io_processed = listCreate();
    server.io_ready_clients = listCreate();
    pthread_mutex_init(&server.io_mutex,NULL);
    pthread_mutex_init(&server.io_swapfile_mutex,NULL);
    server.io_active_threads = 0;
    if (pipe(pipefds) == -1) {
        redisLog(REDIS_WARNING,"Unable to intialized VM: pipe(2): %s. Exiting."
            ,strerror(errno));
        exit(1);
    }
    server.io_ready_pipe_read = pipefds[0];
    server.io_ready_pipe_write = pipefds[1];
    redisAssert(anetNonBlock(NULL,server.io_ready_pipe_read) != ANET_ERR);
    /* LZF requires a lot of stack */
    pthread_attr_init(&server.io_threads_attr);
    pthread_attr_getstacksize(&server.io_threads_attr, &stacksize);

    /* Solaris may report a stacksize of 0, let's set it to 1 otherwise
* multiplying it by 2 in the while loop later will not really help ;) */
    if (!stacksize) stacksize = 1;

    while (stacksize < REDIS_THREAD_STACK_SIZE) stacksize *= 2;
    pthread_attr_setstacksize(&server.io_threads_attr, stacksize);
    /* Listen for events in the threaded I/O pipe */
    if (aeCreateFileEvent(server.el, server.io_ready_pipe_read, AE_READABLE,
        vmThreadedIOCompletedJob, NULL) == AE_ERR)
        oom("creating file event");
}

/* Mark the page as used */
void vmMarkPageUsed(off_t page) {
    off_t byte = page/8;
    int bit = page&7;
    redisAssert(vmFreePage(page) == 1);
    server.vm_bitmap[byte] |= 1<<bit;
}

/* Mark N contiguous pages as used, with 'page' being the first. */
void vmMarkPagesUsed(off_t page, off_t count) {
    off_t j;

    for (j = 0; j < count; j++)
        vmMarkPageUsed(page+j);
    server.vm_stats_used_pages += count;
    redisLog(REDIS_DEBUG,"Mark USED pages: %lld pages at %lld\n",
        (long long)count, (long long)page);
}

/* Mark the page as free */
void vmMarkPageFree(off_t page) {
    off_t byte = page/8;
    int bit = page&7;
    redisAssert(vmFreePage(page) == 0);
    server.vm_bitmap[byte] &= ~(1<<bit);
}

/* Mark N contiguous pages as free, with 'page' being the first. */
void vmMarkPagesFree(off_t page, off_t count) {
    off_t j;

    for (j = 0; j < count; j++)
        vmMarkPageFree(page+j);
    server.vm_stats_used_pages -= count;
    redisLog(REDIS_DEBUG,"Mark FREE pages: %lld pages at %lld\n",
        (long long)count, (long long)page);
}

/* Test if the page is free */
int vmFreePage(off_t page) {
    off_t byte = page/8;
    int bit = page&7;
    return (server.vm_bitmap[byte] & (1<<bit)) == 0;
}

/* Find N contiguous free pages storing the first page of the cluster in *first.
* Returns REDIS_OK if it was able to find N contiguous pages, otherwise
* REDIS_ERR is returned.
*
* This function uses a simple algorithm: we try to allocate
* REDIS_VM_MAX_NEAR_PAGES sequentially, when we reach this limit we start
* again from the start of the swap file searching for free spaces.
*
* If it looks pretty clear that there are no free pages near our offset
* we try to find less populated places doing a forward jump of
* REDIS_VM_MAX_RANDOM_JUMP, then we start scanning again a few pages
* without hurry, and then we jump again and so forth...
*
* This function can be improved using a free list to avoid to guess
* too much, since we could collect data about freed pages.
*
* note: I implemented this function just after watching an episode of
* Battlestar Galactica, where the hybrid was continuing to say "JUMP!"
*/
int vmFindContiguousPages(off_t *first, off_t n) {
    off_t base, offset = 0, since_jump = 0, numfree = 0;

    if (server.vm_near_pages == REDIS_VM_MAX_NEAR_PAGES) {
        server.vm_near_pages = 0;
        server.vm_next_page = 0;
    }
    server.vm_near_pages++; /* Yet another try for pages near to the old ones */
    base = server.vm_next_page;

    while(offset < server.vm_pages) {
        off_t this = base+offset;

        /* If we overflow, restart from page zero */
        if (this >= server.vm_pages) {
            this -= server.vm_pages;
            if (this == 0) {
                /* Just overflowed, what we found on tail is no longer
* interesting, as it's no longer contiguous. */
                numfree = 0;
            }
        }
        if (vmFreePage(this)) {
            /* This is a free page */
            numfree++;
            /* Already got N free pages? Return to the caller, with success */
            if (numfree == n) {
                *first = this-(n-1);
                server.vm_next_page = this+1;
                redisLog(REDIS_DEBUG, "FOUND CONTIGUOUS PAGES: %lld pages at %lld\n", (long long) n, (long long) *first);
                return REDIS_OK;
            }
        } else {
            /* The current one is not a free page */
            numfree = 0;
        }

        /* Fast-forward if the current page is not free and we already
* searched enough near this place. */
        since_jump++;
        if (!numfree && since_jump >= REDIS_VM_MAX_RANDOM_JUMP/4) {
            offset += random() % REDIS_VM_MAX_RANDOM_JUMP;
            since_jump = 0;
            /* Note that even if we rewind after the jump, we are don't need
* to make sure numfree is set to zero as we only jump *if* it
* is set to zero. */
        } else {
            /* Otherwise just check the next page */
            offset++;
        }
    }
    return REDIS_ERR;
}

/* Write the specified object at the specified page of the swap file */
int vmWriteObjectOnSwap(robj *o, off_t page) {
    if (server.vm_enabled) pthread_mutex_lock(&server.io_swapfile_mutex);
    if (fseeko(server.vm_fp,page*server.vm_page_size,SEEK_SET) == -1) {
        if (server.vm_enabled) pthread_mutex_unlock(&server.io_swapfile_mutex);
        redisLog(REDIS_WARNING,
            "Critical VM problem in vmWriteObjectOnSwap(): can't seek: %s",
            strerror(errno));
        return REDIS_ERR;
    }
    rdbSaveObject(server.vm_fp,o);
    fflush(server.vm_fp);
    if (server.vm_enabled) pthread_mutex_unlock(&server.io_swapfile_mutex);
    return REDIS_OK;
}

/* Transfers the 'val' object to disk. Store all the information
* a 'vmpointer' object containing all the information needed to load the
* object back later is returned.
*
* If we can't find enough contiguous empty pages to swap the object on disk
* NULL is returned. */
vmpointer *vmSwapObjectBlocking(robj *val) {
    off_t pages = rdbSavedObjectPages(val);
    off_t page;
    vmpointer *vp;

    redisAssert(val->storage == REDIS_VM_MEMORY);
    redisAssert(val->refcount == 1);
    if (vmFindContiguousPages(&page,pages) == REDIS_ERR) return NULL;
    if (vmWriteObjectOnSwap(val,page) == REDIS_ERR) return NULL;

    vp = createVmPointer(val->type);
    vp->page = page;
    vp->usedpages = pages;
    decrRefCount(val); /* Deallocate the object from memory. */
    vmMarkPagesUsed(page,pages);
    redisLog(REDIS_DEBUG,"VM: object %p swapped out at %lld (%lld pages)",
        (void*) val,
        (unsigned long long) page, (unsigned long long) pages);
    server.vm_stats_swapped_objects++;
    server.vm_stats_swapouts++;
    return vp;
}

robj *vmReadObjectFromSwap(off_t page, int type) {
    robj *o;

    if (server.vm_enabled) pthread_mutex_lock(&server.io_swapfile_mutex);
    if (fseeko(server.vm_fp,page*server.vm_page_size,SEEK_SET) == -1) {
        redisLog(REDIS_WARNING,
            "Unrecoverable VM problem in vmReadObjectFromSwap(): can't seek: %s",
            strerror(errno));
        _exit(1);
    }
    o = rdbLoadObject(type,server.vm_fp);
    if (o == NULL) {
        redisLog(REDIS_WARNING, "Unrecoverable VM problem in vmReadObjectFromSwap(): can't load object from swap file: %s", strerror(errno));
        _exit(1);
    }
    if (server.vm_enabled) pthread_mutex_unlock(&server.io_swapfile_mutex);
    return o;
}

/* Load the specified object from swap to memory.
* The newly allocated object is returned.
*
* If preview is true the unserialized object is returned to the caller but
* the pages are not marked as freed, nor the vp object is freed. */
robj *vmGenericLoadObject(vmpointer *vp, int preview) {
    robj *val;

    redisAssert(vp->type == REDIS_VMPOINTER &&
        (vp->storage == REDIS_VM_SWAPPED || vp->storage == REDIS_VM_LOADING));
    val = vmReadObjectFromSwap(vp->page,vp->vtype);
    if (!preview) {
        redisLog(REDIS_DEBUG, "VM: object %p loaded from disk", (void*)vp);
        vmMarkPagesFree(vp->page,vp->usedpages);
        zfree(vp);
        server.vm_stats_swapped_objects--;
    } else {
        redisLog(REDIS_DEBUG, "VM: object %p previewed from disk", (void*)vp);
    }
    server.vm_stats_swapins++;
    return val;
}

/* Plain object loading, from swap to memory.
*
* 'o' is actually a redisVmPointer structure that will be freed by the call.
* The return value is the loaded object. */
robj *vmLoadObject(robj *o) {
    /* If we are loading the object in background, stop it, we
* need to load this object synchronously ASAP. */
    if (o->storage == REDIS_VM_LOADING)
        vmCancelThreadedIOJob(o);
    return vmGenericLoadObject((vmpointer*)o,0);
}

/* Just load the value on disk, without to modify the key.
* This is useful when we want to perform some operation on the value
* without to really bring it from swap to memory, like while saving the
* dataset or rewriting the append only log. */
robj *vmPreviewObject(robj *o) {
    return vmGenericLoadObject((vmpointer*)o,1);
}

/* How a good candidate is this object for swapping?
* The better candidate it is, the greater the returned value.
*
* Currently we try to perform a fast estimation of the object size in
* memory, and combine it with aging informations.
*
* Basically swappability = idle-time * log(estimated size)
*
* Bigger objects are preferred over smaller objects, but not
* proportionally, this is why we use the logarithm. This algorithm is
* just a first try and will probably be tuned later. */
double computeObjectSwappability(robj *o) {
    /* actual age can be >= minage, but not < minage. As we use wrapping
* 21 bit clocks with minutes resolution for the LRU. */
    time_t minage = estimateObjectIdleTime(o);
    long asize = 0, elesize;
    robj *ele;
    list *l;
    listNode *ln;
    dict *d;
    struct dictEntry *de;
    int z;

    if (minage <= 0) return 0;
    switch(o->type) {
    case REDIS_STRING:
        if (o->encoding != REDIS_ENCODING_RAW) {
            asize = sizeof(*o);
        } else {
            asize = sdslen(o->ptr)+sizeof(*o)+sizeof(long)*2;
        }
        break;
    case REDIS_LIST:
        if (o->encoding == REDIS_ENCODING_ZIPLIST) {
            asize = sizeof(*o)+ziplistSize(o->ptr);
        } else {
            l = o->ptr;
            ln = listFirst(l);
            asize = sizeof(list);
            if (ln) {
                ele = ln->value;
                elesize = (ele->encoding == REDIS_ENCODING_RAW) ?
                                (sizeof(*o)+sdslen(ele->ptr)) : sizeof(*o);
                asize += (sizeof(listNode)+elesize)*listLength(l);
            }
        }
        break;
    case REDIS_SET:
    case REDIS_ZSET:
        z = (o->type == REDIS_ZSET);
        d = z ? ((zset*)o->ptr)->dict : o->ptr;

        if (!z && o->encoding == REDIS_ENCODING_INTSET) {
            intset *is = o->ptr;
            asize = sizeof(*is)+is->encoding*is->length;
        } else {
            asize = sizeof(dict)+(sizeof(struct dictEntry*)*dictSlots(d));
            if (z) asize += sizeof(zset)-sizeof(dict);
            if (dictSize(d)) {
                de = dictGetRandomKey(d);
                ele = dictGetEntryKey(de);
                elesize = (ele->encoding == REDIS_ENCODING_RAW) ?
                                (sizeof(*o)+sdslen(ele->ptr)) : sizeof(*o);
                asize += (sizeof(struct dictEntry)+elesize)*dictSize(d);
                if (z) asize += sizeof(zskiplistNode)*dictSize(d);
            }
        }
        break;
    case REDIS_HASH:
        if (o->encoding == REDIS_ENCODING_ZIPMAP) {
            unsigned char *p = zipmapRewind((unsigned char*)o->ptr);
            unsigned int len = zipmapLen((unsigned char*)o->ptr);
            unsigned int klen, vlen;
            unsigned char *key, *val;

            if ((p = zipmapNext(p,&key,&klen,&val,&vlen)) == NULL) {
                klen = 0;
                vlen = 0;
            }
            asize = len*(klen+vlen+3);
        } else if (o->encoding == REDIS_ENCODING_HT) {
            d = o->ptr;
            asize = sizeof(dict)+(sizeof(struct dictEntry*)*dictSlots(d));
            if (dictSize(d)) {
                de = dictGetRandomKey(d);
                ele = dictGetEntryKey(de);
                elesize = (ele->encoding == REDIS_ENCODING_RAW) ?
                                (sizeof(*o)+sdslen(ele->ptr)) : sizeof(*o);
                ele = dictGetEntryVal(de);
                elesize = (ele->encoding == REDIS_ENCODING_RAW) ?
                                (sizeof(*o)+sdslen(ele->ptr)) : sizeof(*o);
                asize += (sizeof(struct dictEntry)+elesize)*dictSize(d);
            }
        }
        break;
    }
    return (double)minage*log(1+asize);
}

/* Try to swap an object that's a good candidate for swapping.
* Returns REDIS_OK if the object was swapped, REDIS_ERR if it's not possible
* to swap any object at all.
*
* If 'usethreaded' is true, Redis will try to swap the object in background
* using I/O threads. */
int vmSwapOneObject(int usethreads) {
    int j, i;
    struct dictEntry *best = NULL;
    double best_swappability = 0;
    redisDb *best_db = NULL;
    robj *val;
    sds key;

    for (j = 0; j < server.dbnum; j++) {
        redisDb *db = server.db+j;
        /* Why maxtries is set to 100?
* Because this way (usually) we'll find 1 object even if just 1% - 2%
* are swappable objects */
        int maxtries = 100;

        if (dictSize(db->dict) == 0) continue;
        for (i = 0; i < 5; i++) {
            dictEntry *de;
            double swappability;

            if (maxtries) maxtries--;
            de = dictGetRandomKey(db->dict);
            val = dictGetEntryVal(de);
            /* Only swap objects that are currently in memory.
*
* Also don't swap shared objects: not a good idea in general and
* we need to ensure that the main thread does not touch the
* object while the I/O thread is using it, but we can't
* control other keys without adding additional mutex. */
            if (val->storage != REDIS_VM_MEMORY || val->refcount != 1) {
                if (maxtries) i--; /* don't count this try */
                continue;
            }
            swappability = computeObjectSwappability(val);
            if (!best || swappability > best_swappability) {
                best = de;
                best_swappability = swappability;
                best_db = db;
            }
        }
    }
    if (best == NULL) return REDIS_ERR;
    key = dictGetEntryKey(best);
    val = dictGetEntryVal(best);

    redisLog(REDIS_DEBUG,"Key with best swappability: %s, %f",
        key, best_swappability);

    /* Swap it */
    if (usethreads) {
        robj *keyobj = createStringObject(key,sdslen(key));
        vmSwapObjectThreaded(keyobj,val,best_db);
        decrRefCount(keyobj);
        return REDIS_OK;
    } else {
        vmpointer *vp;

        if ((vp = vmSwapObjectBlocking(val)) != NULL) {
            dictGetEntryVal(best) = vp;
            return REDIS_OK;
        } else {
            return REDIS_ERR;
        }
    }
}

int vmSwapOneObjectBlocking() {
    return vmSwapOneObject(0);
}

int vmSwapOneObjectThreaded() {
    return vmSwapOneObject(1);
}

/* Return true if it's safe to swap out objects in a given moment.
* Basically we don't want to swap objects out while there is a BGSAVE
* or a BGAEOREWRITE running in backgroud. */
int vmCanSwapOut(void) {
    return (server.bgsavechildpid == -1 && server.bgrewritechildpid == -1);
}

/* =================== Virtual Memory - Threaded I/O ======================= */

void freeIOJob(iojob *j) {
    if ((j->type == REDIS_IOJOB_PREPARE_SWAP ||
        j->type == REDIS_IOJOB_DO_SWAP ||
        j->type == REDIS_IOJOB_LOAD) && j->val != NULL)
    {
         /* we fix the storage type, otherwise decrRefCount() will try to
* kill the I/O thread Job (that does no longer exists). */
        if (j->val->storage == REDIS_VM_SWAPPING)
            j->val->storage = REDIS_VM_MEMORY;
        decrRefCount(j->val);
    }
    decrRefCount(j->key);
    zfree(j);
}

/* Every time a thread finished a Job, it writes a byte into the write side
* of an unix pipe in order to "awake" the main thread, and this function
* is called.
*
* Note that this is called both by the event loop, when a I/O thread
* sends a byte in the notification pipe, and is also directly called from
* waitEmptyIOJobsQueue().
*
* In the latter case we don't want to swap more, so we use the
* "privdata" argument setting it to a not NULL value to signal this
* condition. */
void vmThreadedIOCompletedJob(aeEventLoop *el, int fd, void *privdata,
            int mask)
{
    char buf[1];
    int retval, processed = 0, toprocess = -1, trytoswap = 1;
    REDIS_NOTUSED(el);
    REDIS_NOTUSED(mask);
    REDIS_NOTUSED(privdata);

    if (privdata != NULL) trytoswap = 0; /* check the comments above... */

    /* For every byte we read in the read side of the pipe, there is one
* I/O job completed to process. */
    while((retval = read(fd,buf,1)) == 1) {
        iojob *j;
        listNode *ln;
        struct dictEntry *de;

        redisLog(REDIS_DEBUG,"Processing I/O completed job");

        /* Get the processed element (the oldest one) */
        lockThreadedIO();
        redisAssert(listLength(server.io_processed) != 0);
        if (toprocess == -1) {
            toprocess = (listLength(server.io_processed)*REDIS_MAX_COMPLETED_JOBS_PROCESSED)/100;
            if (toprocess <= 0) toprocess = 1;
        }
        ln = listFirst(server.io_processed);
        j = ln->value;
        listDelNode(server.io_processed,ln);
        unlockThreadedIO();
        /* If this job is marked as canceled, just ignore it */
        if (j->canceled) {
            freeIOJob(j);
            continue;
        }
        /* Post process it in the main thread, as there are things we
* can do just here to avoid race conditions and/or invasive locks */
        redisLog(REDIS_DEBUG,"COMPLETED Job type: %d, ID %p, key: %s", j->type, (void*)j->id, (unsigned char*)j->key->ptr);
        de = dictFind(j->db->dict,j->key->ptr);
        redisAssert(de != NULL);
        if (j->type == REDIS_IOJOB_LOAD) {
            redisDb *db;
            vmpointer *vp = dictGetEntryVal(de);

            /* Key loaded, bring it at home */
            vmMarkPagesFree(vp->page,vp->usedpages);
            redisLog(REDIS_DEBUG, "VM: object %s loaded from disk (threaded)",
                (unsigned char*) j->key->ptr);
            server.vm_stats_swapped_objects--;
            server.vm_stats_swapins++;
            dictGetEntryVal(de) = j->val;
            incrRefCount(j->val);
            db = j->db;
            /* Handle clients waiting for this key to be loaded. */
            handleClientsBlockedOnSwappedKey(db,j->key);
            freeIOJob(j);
            zfree(vp);
        } else if (j->type == REDIS_IOJOB_PREPARE_SWAP) {
            /* Now we know the amount of pages required to swap this object.
* Let's find some space for it, and queue this task again
* rebranded as REDIS_IOJOB_DO_SWAP. */
            if (!vmCanSwapOut() ||
                vmFindContiguousPages(&j->page,j->pages) == REDIS_ERR)
            {
                /* Ooops... no space or we can't swap as there is
* a fork()ed Redis trying to save stuff on disk. */
                j->val->storage = REDIS_VM_MEMORY; /* undo operation */
                freeIOJob(j);
            } else {
                /* Note that we need to mark this pages as used now,
* if the job will be canceled, we'll mark them as freed
* again. */
                vmMarkPagesUsed(j->page,j->pages);
                j->type = REDIS_IOJOB_DO_SWAP;
                lockThreadedIO();
                queueIOJob(j);
                unlockThreadedIO();
            }
        } else if (j->type == REDIS_IOJOB_DO_SWAP) {
            vmpointer *vp;

            /* Key swapped. We can finally free some memory. */
            if (j->val->storage != REDIS_VM_SWAPPING) {
                vmpointer *vp = (vmpointer*) j->id;
                printf("storage: %d\n",vp->storage);
                printf("key->name: %s\n",(char*)j->key->ptr);
                printf("val: %p\n",(void*)j->val);
                printf("val->type: %d\n",j->val->type);
                printf("val->ptr: %s\n",(char*)j->val->ptr);
            }
            redisAssert(j->val->storage == REDIS_VM_SWAPPING);
            vp = createVmPointer(j->val->type);
            vp->page = j->page;
            vp->usedpages = j->pages;
            dictGetEntryVal(de) = vp;
            /* Fix the storage otherwise decrRefCount will attempt to
* remove the associated I/O job */
            j->val->storage = REDIS_VM_MEMORY;
            decrRefCount(j->val);
            redisLog(REDIS_DEBUG,
                "VM: object %s swapped out at %lld (%lld pages) (threaded)",
                (unsigned char*) j->key->ptr,
                (unsigned long long) j->page, (unsigned long long) j->pages);
            server.vm_stats_swapped_objects++;
            server.vm_stats_swapouts++;
            freeIOJob(j);
            /* Put a few more swap requests in queue if we are still
* out of memory */
            if (trytoswap && vmCanSwapOut() &&
                zmalloc_used_memory() > server.vm_max_memory)
            {
                int more = 1;
                while(more) {
                    lockThreadedIO();
                    more = listLength(server.io_newjobs) <
                            (unsigned) server.vm_max_threads;
                    unlockThreadedIO();
                    /* Don't waste CPU time if swappable objects are rare. */
                    if (vmSwapOneObjectThreaded() == REDIS_ERR) {
                        trytoswap = 0;
                        break;
                    }
                }
            }
        }
        processed++;
        if (processed == toprocess) return;
    }
    if (retval < 0 && errno != EAGAIN) {
        redisLog(REDIS_WARNING,
            "WARNING: read(2) error in vmThreadedIOCompletedJob() %s",
            strerror(errno));
    }
}

void lockThreadedIO(void) {
    pthread_mutex_lock(&server.io_mutex);
}

void unlockThreadedIO(void) {
    pthread_mutex_unlock(&server.io_mutex);
}

/* Remove the specified object from the threaded I/O queue if still not
* processed, otherwise make sure to flag it as canceled. */
void vmCancelThreadedIOJob(robj *o) {
    list *lists[3] = {
        server.io_newjobs, /* 0 */
        server.io_processing, /* 1 */
        server.io_processed /* 2 */
    };
    int i;

    redisAssert(o->storage == REDIS_VM_LOADING || o->storage == REDIS_VM_SWAPPING);
again:
    lockThreadedIO();
    /* Search for a matching object in one of the queues */
    for (i = 0; i < 3; i++) {
        listNode *ln;
        listIter li;

        listRewind(lists[i],&li);
        while ((ln = listNext(&li)) != NULL) {
            iojob *job = ln->value;

            if (job->canceled) continue; /* Skip this, already canceled. */
            if (job->id == o) {
                redisLog(REDIS_DEBUG,"*** CANCELED %p (key %s) (type %d) (LIST ID %d)\n",
                    (void*)job, (char*)job->key->ptr, job->type, i);
                /* Mark the pages as free since the swap didn't happened
* or happened but is now discarded. */
                if (i != 1 && job->type == REDIS_IOJOB_DO_SWAP)
                    vmMarkPagesFree(job->page,job->pages);
                /* Cancel the job. It depends on the list the job is
* living in. */
                switch(i) {
                case 0: /* io_newjobs */
                    /* If the job was yet not processed the best thing to do
* is to remove it from the queue at all */
                    freeIOJob(job);
                    listDelNode(lists[i],ln);
                    break;
                case 1: /* io_processing */
                    /* Oh Shi- the thread is messing with the Job:
*
* Probably it's accessing the object if this is a
* PREPARE_SWAP or DO_SWAP job.
* If it's a LOAD job it may be reading from disk and
* if we don't wait for the job to terminate before to
* cancel it, maybe in a few microseconds data can be
* corrupted in this pages. So the short story is:
*
* Better to wait for the job to move into the
* next queue (processed)... */

                    /* We try again and again until the job is completed. */
                    unlockThreadedIO();
                    /* But let's wait some time for the I/O thread
* to finish with this job. After all this condition
* should be very rare. */
                    usleep(1);
                    goto again;
                case 2: /* io_processed */
                    /* The job was already processed, that's easy...
* just mark it as canceled so that we'll ignore it
* when processing completed jobs. */
                    job->canceled = 1;
                    break;
                }
                /* Finally we have to adjust the storage type of the object
* in order to "UNDO" the operaiton. */
                if (o->storage == REDIS_VM_LOADING)
                    o->storage = REDIS_VM_SWAPPED;
                else if (o->storage == REDIS_VM_SWAPPING)
                    o->storage = REDIS_VM_MEMORY;
                unlockThreadedIO();
                redisLog(REDIS_DEBUG,"*** DONE");
                return;
            }
        }
    }
    unlockThreadedIO();
    printf("Not found: %p\n", (void*)o);
    redisAssert(1 != 1); /* We should never reach this */
}

void *IOThreadEntryPoint(void *arg) {
    iojob *j;
    listNode *ln;
    REDIS_NOTUSED(arg);

    pthread_detach(pthread_self());
    while(1) {
        /* Get a new job to process */
        lockThreadedIO();
        if (listLength(server.io_newjobs) == 0) {
            /* No new jobs in queue, exit. */
            redisLog(REDIS_DEBUG,"Thread %ld exiting, nothing to do",
                (long) pthread_self());
            server.io_active_threads--;
            unlockThreadedIO();
            return NULL;
        }
        ln = listFirst(server.io_newjobs);
        j = ln->value;
        listDelNode(server.io_newjobs,ln);
        /* Add the job in the processing queue */
        j->thread = pthread_self();
        listAddNodeTail(server.io_processing,j);
        ln = listLast(server.io_processing); /* We use ln later to remove it */
        unlockThreadedIO();
        redisLog(REDIS_DEBUG,"Thread %ld got a new job (type %d): %p about key '%s'",
            (long) pthread_self(), j->type, (void*)j, (char*)j->key->ptr);

        /* Process the Job */
        if (j->type == REDIS_IOJOB_LOAD) {
            vmpointer *vp = (vmpointer*)j->id;
            j->val = vmReadObjectFromSwap(j->page,vp->vtype);
        } else if (j->type == REDIS_IOJOB_PREPARE_SWAP) {
            j->pages = rdbSavedObjectPages(j->val);
        } else if (j->type == REDIS_IOJOB_DO_SWAP) {
            if (vmWriteObjectOnSwap(j->val,j->page) == REDIS_ERR)
                j->canceled = 1;
        }

        /* Done: insert the job into the processed queue */
        redisLog(REDIS_DEBUG,"Thread %ld completed the job: %p (key %s)",
            (long) pthread_self(), (void*)j, (char*)j->key->ptr);
        lockThreadedIO();
        listDelNode(server.io_processing,ln);
        listAddNodeTail(server.io_processed,j);
        unlockThreadedIO();

        /* Signal the main thread there is new stuff to process */
        redisAssert(write(server.io_ready_pipe_write,"x",1) == 1);
    }
    return NULL; /* never reached */
}

void spawnIOThread(void) {
    pthread_t thread;
    sigset_t mask, omask;
    int err;

    sigemptyset(&mask);
    sigaddset(&mask,SIGCHLD);
    sigaddset(&mask,SIGHUP);
    sigaddset(&mask,SIGPIPE);
    pthread_sigmask(SIG_SETMASK, &mask, &omask);
    while ((err = pthread_create(&thread,&server.io_threads_attr,IOThreadEntryPoint,NULL)) != 0) {
        redisLog(REDIS_WARNING,"Unable to spawn an I/O thread: %s",
            strerror(err));
        usleep(1000000);
    }
    pthread_sigmask(SIG_SETMASK, &omask, NULL);
    server.io_active_threads++;
}

/* We need to wait for the last thread to exit before we are able to
* fork() in order to BGSAVE or BGREWRITEAOF. */
void waitEmptyIOJobsQueue(void) {
    while(1) {
        int io_processed_len;

        lockThreadedIO();
        if (listLength(server.io_newjobs) == 0 &&
            listLength(server.io_processing) == 0 &&
            server.io_active_threads == 0)
        {
            unlockThreadedIO();
            return;
        }
        /* While waiting for empty jobs queue condition we post-process some
* finshed job, as I/O threads may be hanging trying to write against
* the io_ready_pipe_write FD but there are so much pending jobs that
* it's blocking. */
        io_processed_len = listLength(server.io_processed);
        unlockThreadedIO();
        if (io_processed_len) {
            vmThreadedIOCompletedJob(NULL,server.io_ready_pipe_read,
                                                        (void*)0xdeadbeef,0);
            usleep(1000); /* 1 millisecond */
        } else {
            usleep(10000); /* 10 milliseconds */
        }
    }
}

void vmReopenSwapFile(void) {
    /* Note: we don't close the old one as we are in the child process
* and don't want to mess at all with the original file object. */
    server.vm_fp = fopen(server.vm_swap_file,"r+b");
    if (server.vm_fp == NULL) {
        redisLog(REDIS_WARNING,"Can't re-open the VM swap file: %s. Exiting.",
            server.vm_swap_file);
        _exit(1);
    }
    server.vm_fd = fileno(server.vm_fp);
}

/* This function must be called while with threaded IO locked */
void queueIOJob(iojob *j) {
    redisLog(REDIS_DEBUG,"Queued IO Job %p type %d about key '%s'\n",
        (void*)j, j->type, (char*)j->key->ptr);
    listAddNodeTail(server.io_newjobs,j);
    if (server.io_active_threads < server.vm_max_threads)
        spawnIOThread();
}

int vmSwapObjectThreaded(robj *key, robj *val, redisDb *db) {
    iojob *j;

    j = zmalloc(sizeof(*j));
    j->type = REDIS_IOJOB_PREPARE_SWAP;
    j->db = db;
    j->key = key;
    incrRefCount(key);
    j->id = j->val = val;
    incrRefCount(val);
    j->canceled = 0;
    j->thread = (pthread_t) -1;
    val->storage = REDIS_VM_SWAPPING;

    lockThreadedIO();
    queueIOJob(j);
    unlockThreadedIO();
    return REDIS_OK;
}

/* ============ Virtual Memory - Blocking clients on missing keys =========== */

/* This function makes the clinet 'c' waiting for the key 'key' to be loaded.
* If there is not already a job loading the key, it is craeted.
* The key is added to the io_keys list in the client structure, and also
* in the hash table mapping swapped keys to waiting clients, that is,
* server.io_waited_keys. */
int waitForSwappedKey(redisClient *c, robj *key) {
    struct dictEntry *de;
    robj *o;
    list *l;

    /* If the key does not exist or is already in RAM we don't need to
* block the client at all. */
    de = dictFind(c->db->dict,key->ptr);
    if (de == NULL) return 0;
    o = dictGetEntryVal(de);
    if (o->storage == REDIS_VM_MEMORY) {
        return 0;
    } else if (o->storage == REDIS_VM_SWAPPING) {
        /* We were swapping the key, undo it! */
        vmCancelThreadedIOJob(o);
        return 0;
    }

    /* OK: the key is either swapped, or being loaded just now. */

    /* Add the key to the list of keys this client is waiting for.
* This maps clients to keys they are waiting for. */
    listAddNodeTail(c->io_keys,key);
    incrRefCount(key);

    /* Add the client to the swapped keys => clients waiting map. */
    de = dictFind(c->db->io_keys,key);
    if (de == NULL) {
        int retval;

        /* For every key we take a list of clients blocked for it */
        l = listCreate();
        retval = dictAdd(c->db->io_keys,key,l);
        incrRefCount(key);
        redisAssert(retval == DICT_OK);
    } else {
        l = dictGetEntryVal(de);
    }
    listAddNodeTail(l,c);

    /* Are we already loading the key from disk? If not create a job */
    if (o->storage == REDIS_VM_SWAPPED) {
        iojob *j;
        vmpointer *vp = (vmpointer*)o;

        o->storage = REDIS_VM_LOADING;
        j = zmalloc(sizeof(*j));
        j->type = REDIS_IOJOB_LOAD;
        j->db = c->db;
        j->id = (robj*)vp;
        j->key = key;
        incrRefCount(key);
        j->page = vp->page;
        j->val = NULL;
        j->canceled = 0;
        j->thread = (pthread_t) -1;
        lockThreadedIO();
        queueIOJob(j);
        unlockThreadedIO();
    }
    return 1;
}

/* Preload keys for any command with first, last and step values for
* the command keys prototype, as defined in the command table. */
void waitForMultipleSwappedKeys(redisClient *c, struct redisCommand *cmd, int argc, robj **argv) {
    int j, last;
    if (cmd->vm_firstkey == 0) return;
    last = cmd->vm_lastkey;
    if (last < 0) last = argc+last;
    for (j = cmd->vm_firstkey; j <= last; j += cmd->vm_keystep) {
        redisAssert(j < argc);
        waitForSwappedKey(c,argv[j]);
    }
}

/* Preload keys needed for the ZUNIONSTORE and ZINTERSTORE commands.
* Note that the number of keys to preload is user-defined, so we need to
* apply a sanity check against argc. */
void zunionInterBlockClientOnSwappedKeys(redisClient *c, struct redisCommand *cmd, int argc, robj **argv) {
    int i, num;
    REDIS_NOTUSED(cmd);

    num = atoi(argv[2]->ptr);
    if (num > (argc-3)) return;
    for (i = 0; i < num; i++) {
        waitForSwappedKey(c,argv[3+i]);
    }
}

/* Preload keys needed to execute the entire MULTI/EXEC block.
*
* This function is called by blockClientOnSwappedKeys when EXEC is issued,
* and will block the client when any command requires a swapped out value. */
void execBlockClientOnSwappedKeys(redisClient *c, struct redisCommand *cmd, int argc, robj **argv) {
    int i, margc;
    struct redisCommand *mcmd;
    robj **margv;
    REDIS_NOTUSED(cmd);
    REDIS_NOTUSED(argc);
    REDIS_NOTUSED(argv);

    if (!(c->flags & REDIS_MULTI)) return;
    for (i = 0; i < c->mstate.count; i++) {
        mcmd = c->mstate.commands[i].cmd;
        margc = c->mstate.commands[i].argc;
        margv = c->mstate.commands[i].argv;

        if (mcmd->vm_preload_proc != NULL) {
            mcmd->vm_preload_proc(c,mcmd,margc,margv);
        } else {
            waitForMultipleSwappedKeys(c,mcmd,margc,margv);
        }
    }
}

/* Is this client attempting to run a command against swapped keys?
* If so, block it ASAP, load the keys in background, then resume it.
*
* The important idea about this function is that it can fail! If keys will
* still be swapped when the client is resumed, this key lookups will
* just block loading keys from disk. In practical terms this should only
* happen with SORT BY command or if there is a bug in this function.
*
* Return 1 if the client is marked as blocked, 0 if the client can
* continue as the keys it is going to access appear to be in memory. */
int blockClientOnSwappedKeys(redisClient *c, struct redisCommand *cmd) {
    if (cmd->vm_preload_proc != NULL) {
        cmd->vm_preload_proc(c,cmd,c->argc,c->argv);
    } else {
        waitForMultipleSwappedKeys(c,cmd,c->argc,c->argv);
    }

    /* If the client was blocked for at least one key, mark it as blocked. */
    if (listLength(c->io_keys)) {
        c->flags |= REDIS_IO_WAIT;
        aeDeleteFileEvent(server.el,c->fd,AE_READABLE);
        server.vm_blocked_clients++;
        return 1;
    } else {
        return 0;
    }
}

/* Remove the 'key' from the list of blocked keys for a given client.
*
* The function returns 1 when there are no longer blocking keys after
* the current one was removed (and the client can be unblocked). */
int dontWaitForSwappedKey(redisClient *c, robj *key) {
    list *l;
    listNode *ln;
    listIter li;
    struct dictEntry *de;

    /* The key object might be destroyed when deleted from the c->io_keys
* list (and the "key" argument is physically the same object as the
* object inside the list), so we need to protect it. */
    incrRefCount(key);

    /* Remove the key from the list of keys this client is waiting for. */
    listRewind(c->io_keys,&li);
    while ((ln = listNext(&li)) != NULL) {
        if (equalStringObjects(ln->value,key)) {
            listDelNode(c->io_keys,ln);
            break;
        }
    }
    redisAssert(ln != NULL);

    /* Remove the client form the key => waiting clients map. */
    de = dictFind(c->db->io_keys,key);
    redisAssert(de != NULL);
    l = dictGetEntryVal(de);
    ln = listSearchKey(l,c);
    redisAssert(ln != NULL);
    listDelNode(l,ln);
    if (listLength(l) == 0)
        dictDelete(c->db->io_keys,key);

    decrRefCount(key);
    return listLength(c->io_keys) == 0;
}

/* Every time we now a key was loaded back in memory, we handle clients
* waiting for this key if any. */
void handleClientsBlockedOnSwappedKey(redisDb *db, robj *key) {
    struct dictEntry *de;
    list *l;
    listNode *ln;
    int len;

    de = dictFind(db->io_keys,key);
    if (!de) return;

    l = dictGetEntryVal(de);
    len = listLength(l);
    /* Note: we can't use something like while(listLength(l)) as the list
* can be freed by the calling function when we remove the last element. */
    while (len--) {
        ln = listFirst(l);
        redisClient *c = ln->value;

        if (dontWaitForSwappedKey(c,key)) {
            /* Put the client in the list of clients ready to go as we
* loaded all the keys about it. */
            listAddNodeTail(server.io_ready_clients,c);
        }
    }
}
Something went wrong with that request. Please try again.