networking.c 28.6 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13
#include "redis.h"
#include <sys/uio.h>

void *dupClientReplyValue(void *o) {
    incrRefCount((robj*)o);
    return o;
}

int listMatchObjects(void *a, void *b) {
    return equalStringObjects(a,b);
}

redisClient *createClient(int fd) {
14
    redisClient *c = zmalloc(sizeof(redisClient));
15
    c->bufpos = 0;
16 17 18 19

    anetNonBlock(NULL,fd);
    anetTcpNoDelay(NULL,fd);
    if (!c) return NULL;
20 21 22 23 24 25 26 27
    if (aeCreateFileEvent(server.el,fd,AE_READABLE,
        readQueryFromClient, c) == AE_ERR)
    {
        close(fd);
        zfree(c);
        return NULL;
    }

28 29 30
    selectDb(c,0);
    c->fd = fd;
    c->querybuf = sdsempty();
31
    c->reqtype = 0;
32 33
    c->argc = 0;
    c->argv = NULL;
34
    c->multibulklen = 0;
35 36 37 38 39 40 41 42 43
    c->bulklen = -1;
    c->sentlen = 0;
    c->flags = 0;
    c->lastinteraction = time(NULL);
    c->authenticated = 0;
    c->replstate = REDIS_REPL_NONE;
    c->reply = listCreate();
    listSetFreeMethod(c->reply,decrRefCount);
    listSetDupMethod(c->reply,dupClientReplyValue);
44 45 46 47
    c->bpop.keys = NULL;
    c->bpop.count = 0;
    c->bpop.timeout = 0;
    c->bpop.target = NULL;
48 49 50 51 52 53 54 55 56 57 58 59
    c->io_keys = listCreate();
    c->watched_keys = listCreate();
    listSetFreeMethod(c->io_keys,decrRefCount);
    c->pubsub_channels = dictCreate(&setDictType,NULL);
    c->pubsub_patterns = listCreate();
    listSetFreeMethod(c->pubsub_patterns,decrRefCount);
    listSetMatchMethod(c->pubsub_patterns,listMatchObjects);
    listAddNodeTail(server.clients,c);
    initClientMultiState(c);
    return c;
}

60 61
/* Set the event loop to listen for write events on the client's socket.
 * Typically gets called every time a reply is built. */
62
int _installWriteEvent(redisClient *c) {
63 64 65
    /* When CLOSE_AFTER_REPLY is set, no more replies may be added! */
    redisAssert(!(c->flags & REDIS_CLOSE_AFTER_REPLY));

66
    if (c->fd <= 0) return REDIS_ERR;
67
    if (c->bufpos == 0 && listLength(c->reply) == 0 &&
68 69 70
        (c->replstate == REDIS_REPL_NONE ||
         c->replstate == REDIS_REPL_ONLINE) &&
        aeCreateFileEvent(server.el, c->fd, AE_WRITABLE,
71 72 73 74
        sendReplyToClient, c) == AE_ERR) return REDIS_ERR;
    return REDIS_OK;
}

75 76 77 78 79 80 81 82 83 84 85 86 87 88
/* Create a duplicate of the last object in the reply list when
 * it is not exclusively owned by the reply list. */
robj *dupLastObjectIfNeeded(list *reply) {
    robj *new, *cur;
    listNode *ln;
    redisAssert(listLength(reply) > 0);
    ln = listLast(reply);
    cur = listNodeValue(ln);
    if (cur->refcount > 1) {
        new = dupStringObject(cur);
        decrRefCount(cur);
        listNodeValue(ln) = new;
    }
    return listNodeValue(ln);
89 90
}

91
int _addReplyToBuffer(redisClient *c, char *s, size_t len) {
92
    size_t available = sizeof(c->buf)-c->bufpos;
93 94 95 96 97 98 99

    /* If there already are entries in the reply list, we cannot
     * add anything more to the static buffer. */
    if (listLength(c->reply) > 0) return REDIS_ERR;

    /* Check that the buffer has enough space available for this string. */
    if (len > available) return REDIS_ERR;
100

101 102 103
    memcpy(c->buf+c->bufpos,s,len);
    c->bufpos+=len;
    return REDIS_OK;
104 105
}

106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125
void _addReplyObjectToList(redisClient *c, robj *o) {
    robj *tail;
    if (listLength(c->reply) == 0) {
        incrRefCount(o);
        listAddNodeTail(c->reply,o);
    } else {
        tail = listNodeValue(listLast(c->reply));

        /* Append to this object when possible. */
        if (tail->ptr != NULL &&
            sdslen(tail->ptr)+sdslen(o->ptr) <= REDIS_REPLY_CHUNK_BYTES)
        {
            tail = dupLastObjectIfNeeded(c->reply);
            tail->ptr = sdscatlen(tail->ptr,o->ptr,sdslen(o->ptr));
        } else {
            incrRefCount(o);
            listAddNodeTail(c->reply,o);
        }
    }
}
126

127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142
/* This method takes responsibility over the sds. When it is no longer
 * needed it will be free'd, otherwise it ends up in a robj. */
void _addReplySdsToList(redisClient *c, sds s) {
    robj *tail;
    if (listLength(c->reply) == 0) {
        listAddNodeTail(c->reply,createObject(REDIS_STRING,s));
    } else {
        tail = listNodeValue(listLast(c->reply));

        /* Append to this object when possible. */
        if (tail->ptr != NULL &&
            sdslen(tail->ptr)+sdslen(s) <= REDIS_REPLY_CHUNK_BYTES)
        {
            tail = dupLastObjectIfNeeded(c->reply);
            tail->ptr = sdscatlen(tail->ptr,s,sdslen(s));
            sdsfree(s);
143
        } else {
144
            listAddNodeTail(c->reply,createObject(REDIS_STRING,s));
145
        }
146 147 148 149 150 151 152
    }
}

void _addReplyStringToList(redisClient *c, char *s, size_t len) {
    robj *tail;
    if (listLength(c->reply) == 0) {
        listAddNodeTail(c->reply,createStringObject(s,len));
153
    } else {
154 155 156 157 158 159 160 161
        tail = listNodeValue(listLast(c->reply));

        /* Append to this object when possible. */
        if (tail->ptr != NULL &&
            sdslen(tail->ptr)+len <= REDIS_REPLY_CHUNK_BYTES)
        {
            tail = dupLastObjectIfNeeded(c->reply);
            tail->ptr = sdscatlen(tail->ptr,s,len);
162
        } else {
163
            listAddNodeTail(c->reply,createStringObject(s,len));
164 165 166
        }
    }
}
167

168
void addReply(redisClient *c, robj *obj) {
169 170 171 172 173 174 175 176 177 178 179 180 181
    if (_installWriteEvent(c) != REDIS_OK) return;
    redisAssert(!server.vm_enabled || obj->storage == REDIS_VM_MEMORY);

    /* This is an important place where we can avoid copy-on-write
     * when there is a saving child running, avoiding touching the
     * refcount field of the object if it's not needed.
     *
     * If the encoding is RAW and there is room in the static buffer
     * we'll be able to send the object to the client without
     * messing with its page. */
    if (obj->encoding == REDIS_ENCODING_RAW) {
        if (_addReplyToBuffer(c,obj->ptr,sdslen(obj->ptr)) != REDIS_OK)
            _addReplyObjectToList(c,obj);
182
    } else {
183 184 185
        /* FIXME: convert the long into string and use _addReplyToBuffer()
         * instead of calling getDecodedObject. As this place in the
         * code is too performance critical. */
186
        obj = getDecodedObject(obj);
187 188 189
        if (_addReplyToBuffer(c,obj->ptr,sdslen(obj->ptr)) != REDIS_OK)
            _addReplyObjectToList(c,obj);
        decrRefCount(obj);
190 191 192 193
    }
}

void addReplySds(redisClient *c, sds s) {
194
    if (_installWriteEvent(c) != REDIS_OK) {
195 196 197 198
        /* The caller expects the sds to be free'd. */
        sdsfree(s);
        return;
    }
199
    if (_addReplyToBuffer(c,s,sdslen(s)) == REDIS_OK) {
200 201
        sdsfree(s);
    } else {
202 203
        /* This method free's the sds when it is no longer needed. */
        _addReplySdsToList(c,s);
204
    }
205 206
}

207
void addReplyString(redisClient *c, char *s, size_t len) {
208
    if (_installWriteEvent(c) != REDIS_OK) return;
209 210
    if (_addReplyToBuffer(c,s,len) != REDIS_OK)
        _addReplyStringToList(c,s,len);
211
}
212

213 214 215 216
void _addReplyError(redisClient *c, char *s, size_t len) {
    addReplyString(c,"-ERR ",5);
    addReplyString(c,s,len);
    addReplyString(c,"\r\n",2);
217 218
}

219 220 221
void addReplyError(redisClient *c, char *err) {
    _addReplyError(c,err,strlen(err));
}
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
void addReplyErrorFormat(redisClient *c, const char *fmt, ...) {
    va_list ap;
    va_start(ap,fmt);
    sds s = sdscatvprintf(sdsempty(),fmt,ap);
    va_end(ap);
    _addReplyError(c,s,sdslen(s));
    sdsfree(s);
}

void _addReplyStatus(redisClient *c, char *s, size_t len) {
    addReplyString(c,"+",1);
    addReplyString(c,s,len);
    addReplyString(c,"\r\n",2);
}

void addReplyStatus(redisClient *c, char *status) {
    _addReplyStatus(c,status,strlen(status));
}

void addReplyStatusFormat(redisClient *c, const char *fmt, ...) {
    va_list ap;
    va_start(ap,fmt);
    sds s = sdscatvprintf(sdsempty(),fmt,ap);
    va_end(ap);
    _addReplyStatus(c,s,sdslen(s));
    sdsfree(s);
}

251 252 253
/* Adds an empty object to the reply list that will contain the multi bulk
 * length, which is not known when this function is called. */
void *addDeferredMultiBulkLength(redisClient *c) {
254 255 256 257
    /* Note that we install the write event here even if the object is not
     * ready to be sent, since we are sure that before returning to the
     * event loop setDeferredMultiBulkLength() will be called. */
    if (_installWriteEvent(c) != REDIS_OK) return NULL;
258
    listAddNodeTail(c->reply,createObject(REDIS_STRING,NULL));
259 260 261 262 263 264 265 266 267 268 269 270 271 272 273
    return listLast(c->reply);
}

/* Populate the length object and try glueing it to the next chunk. */
void setDeferredMultiBulkLength(redisClient *c, void *node, long length) {
    listNode *ln = (listNode*)node;
    robj *len, *next;

    /* Abort when *node is NULL (see addDeferredMultiBulkLength). */
    if (node == NULL) return;

    len = listNodeValue(ln);
    len->ptr = sdscatprintf(sdsempty(),"*%ld\r\n",length);
    if (ln->next != NULL) {
        next = listNodeValue(ln->next);
274

275
        /* Only glue when the next node is non-NULL (an sds in this case) */
276
        if (next->ptr != NULL) {
277
            len->ptr = sdscatlen(len->ptr,next->ptr,sdslen(next->ptr));
278 279
            listDelNode(c->reply,ln->next);
        }
280
    }
281 282
}

283
/* Add a duble as a bulk reply */
284 285 286 287 288 289
void addReplyDouble(redisClient *c, double d) {
    char dbuf[128], sbuf[128];
    int dlen, slen;
    dlen = snprintf(dbuf,sizeof(dbuf),"%.17g",d);
    slen = snprintf(sbuf,sizeof(sbuf),"$%d\r\n%s\r\n",dlen,dbuf);
    addReplyString(c,sbuf,slen);
290 291
}

292 293
/* Add a long long as integer reply or bulk len / multi bulk count.
 * Basically this is used to output <prefix><long long><crlf>. */
294
void _addReplyLongLong(redisClient *c, long long ll, char prefix) {
295
    char buf[128];
296 297
    int len;
    buf[0] = prefix;
298 299 300
    len = ll2string(buf+1,sizeof(buf)-1,ll);
    buf[len+1] = '\r';
    buf[len+2] = '\n';
301
    addReplyString(c,buf,len+3);
302 303
}

304 305 306
void addReplyLongLong(redisClient *c, long long ll) {
    _addReplyLongLong(c,ll,':');
}
307

308 309
void addReplyMultiBulkLen(redisClient *c, long length) {
    _addReplyLongLong(c,length,'*');
310 311
}

312
/* Create the length prefix of a bulk reply, example: $2234 */
313
void addReplyBulkLen(redisClient *c, robj *obj) {
314
    size_t len;
315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330

    if (obj->encoding == REDIS_ENCODING_RAW) {
        len = sdslen(obj->ptr);
    } else {
        long n = (long)obj->ptr;

        /* Compute how many bytes will take this integer as a radix 10 string */
        len = 1;
        if (n < 0) {
            len++;
            n = -n;
        }
        while((n = n/10) != 0) {
            len++;
        }
    }
331
    _addReplyLongLong(c,len,'$');
332 333
}

334
/* Add a Redis Object as a bulk reply */
335 336 337 338 339 340
void addReplyBulk(redisClient *c, robj *obj) {
    addReplyBulkLen(c,obj);
    addReply(c,obj);
    addReply(c,shared.crlf);
}

341 342 343 344 345 346 347 348
/* Add a C buffer as bulk reply */
void addReplyBulkCBuffer(redisClient *c, void *p, size_t len) {
    _addReplyLongLong(c,len,'$');
    addReplyString(c,p,len);
    addReply(c,shared.crlf);
}

/* Add a C nul term string as bulk reply */
349 350 351 352
void addReplyBulkCString(redisClient *c, char *s) {
    if (s == NULL) {
        addReply(c,shared.nullbulk);
    } else {
353
        addReplyBulkCBuffer(c,s,strlen(s));
354 355 356
    }
}

357 358 359 360 361 362 363 364 365
/* Add a long long as a bulk reply */
void addReplyBulkLongLong(redisClient *c, long long ll) {
    char buf[64];
    int len;

    len = ll2string(buf,64,ll);
    addReplyBulkCBuffer(c,buf,len);
}

366
static void acceptCommonHandler(int fd) {
367
    redisClient *c;
368
    if ((c = createClient(fd)) == NULL) {
369
        redisLog(REDIS_WARNING,"Error allocating resoures for the client");
370
        close(fd); /* May be already closed, just ingore errors */
371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389
        return;
    }
    /* If maxclient directive is set and this is one client more... close the
     * connection. Note that we create the client instead to check before
     * for this condition, since now the socket is already set in nonblocking
     * mode and we can send an error for free using the Kernel I/O */
    if (server.maxclients && listLength(server.clients) > server.maxclients) {
        char *err = "-ERR max number of clients reached\r\n";

        /* That's a best effort error message, don't check write errors */
        if (write(c->fd,err,strlen(err)) == -1) {
            /* Nothing to do, Just to avoid the warning... */
        }
        freeClient(c);
        return;
    }
    server.stat_numconnections++;
}

390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411
void acceptTcpHandler(aeEventLoop *el, int fd, void *privdata, int mask) {
    int cport, cfd;
    char cip[128];
    REDIS_NOTUSED(el);
    REDIS_NOTUSED(mask);
    REDIS_NOTUSED(privdata);

    cfd = anetTcpAccept(server.neterr, fd, cip, &cport);
    if (cfd == AE_ERR) {
        redisLog(REDIS_VERBOSE,"Accepting client connection: %s", server.neterr);
        return;
    }
    redisLog(REDIS_VERBOSE,"Accepted %s:%d", cip, cport);
    acceptCommonHandler(cfd);
}

void acceptUnixHandler(aeEventLoop *el, int fd, void *privdata, int mask) {
    int cfd;
    REDIS_NOTUSED(el);
    REDIS_NOTUSED(mask);
    REDIS_NOTUSED(privdata);

412
    cfd = anetUnixAccept(server.neterr, fd);
413 414 415 416 417 418 419 420 421
    if (cfd == AE_ERR) {
        redisLog(REDIS_VERBOSE,"Accepting client connection: %s", server.neterr);
        return;
    }
    redisLog(REDIS_VERBOSE,"Accepted connection to %s", server.unixsocket);
    acceptCommonHandler(cfd);
}


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
static void freeClientArgv(redisClient *c) {
    int j;
    for (j = 0; j < c->argc; j++)
        decrRefCount(c->argv[j]);
    c->argc = 0;
}

void freeClient(redisClient *c) {
    listNode *ln;

    /* Note that if the client we are freeing is blocked into a blocking
     * call, we have to set querybuf to NULL *before* to call
     * unblockClientWaitingData() to avoid processInputBuffer() will get
     * called. Also it is important to remove the file events after
     * this, because this call adds the READABLE event. */
    sdsfree(c->querybuf);
    c->querybuf = NULL;
    if (c->flags & REDIS_BLOCKED)
        unblockClientWaitingData(c);

    /* UNWATCH all the keys */
    unwatchAllKeys(c);
    listRelease(c->watched_keys);
    /* Unsubscribe from all the pubsub channels */
    pubsubUnsubscribeAllChannels(c,0);
    pubsubUnsubscribeAllPatterns(c,0);
    dictRelease(c->pubsub_channels);
    listRelease(c->pubsub_patterns);
    /* Obvious cleanup */
    aeDeleteFileEvent(server.el,c->fd,AE_READABLE);
    aeDeleteFileEvent(server.el,c->fd,AE_WRITABLE);
    listRelease(c->reply);
    freeClientArgv(c);
    close(c->fd);
    /* Remove from the list of clients */
    ln = listSearchKey(server.clients,c);
    redisAssert(ln != NULL);
    listDelNode(server.clients,ln);
460 461 462 463 464 465 466 467 468 469
    /* Remove from the list of clients waiting for swapped keys, or ready
     * to be restarted, but not yet woken up again. */
    if (c->flags & REDIS_IO_WAIT) {
        redisAssert(server.vm_enabled);
        if (listLength(c->io_keys) == 0) {
            ln = listSearchKey(server.io_ready_clients,c);

            /* When this client is waiting to be woken up (REDIS_IO_WAIT),
             * it should be present in the list io_ready_clients */
            redisAssert(ln != NULL);
470
            listDelNode(server.io_ready_clients,ln);
471 472 473 474 475
        } else {
            while (listLength(c->io_keys)) {
                ln = listFirst(c->io_keys);
                dontWaitForSwappedKey(c,ln->value);
            }
476
        }
477
        server.vm_blocked_clients--;
478 479
    }
    listRelease(c->io_keys);
480 481
    /* Master/slave cleanup.
     * Case 1: we lost the connection with a slave. */
482 483 484 485 486 487 488 489
    if (c->flags & REDIS_SLAVE) {
        if (c->replstate == REDIS_REPL_SEND_BULK && c->repldbfd != -1)
            close(c->repldbfd);
        list *l = (c->flags & REDIS_MONITOR) ? server.monitors : server.slaves;
        ln = listSearchKey(l,c);
        redisAssert(ln != NULL);
        listDelNode(l,ln);
    }
490 491

    /* Case 2: we lost the connection with the master. */
492 493
    if (c->flags & REDIS_MASTER) {
        server.master = NULL;
494
        /* FIXME */
495
        server.replstate = REDIS_REPL_CONNECT;
496 497 498 499 500 501 502 503 504
        /* Since we lost the connection with the master, we should also
         * close the connection with all our slaves if we have any, so
         * when we'll resync with the master the other slaves will sync again
         * with us as well. Note that also when the slave is not connected
         * to the master it will keep refusing connections by other slaves. */
        while (listLength(server.slaves)) {
            ln = listFirst(server.slaves);
            freeClient((redisClient*)ln->value);
        }
505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527
    }
    /* Release memory */
    zfree(c->argv);
    freeClientMultiState(c);
    zfree(c);
}

void sendReplyToClient(aeEventLoop *el, int fd, void *privdata, int mask) {
    redisClient *c = privdata;
    int nwritten = 0, totwritten = 0, objlen;
    robj *o;
    REDIS_NOTUSED(el);
    REDIS_NOTUSED(mask);

    /* Use writev() if we have enough buffers to send */
    if (!server.glueoutputbuf &&
        listLength(c->reply) > REDIS_WRITEV_THRESHOLD &&
        !(c->flags & REDIS_MASTER))
    {
        sendReplyToClientWritev(el, fd, privdata, mask);
        return;
    }

528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548
    while(c->bufpos > 0 || listLength(c->reply)) {
        if (c->bufpos > 0) {
            if (c->flags & REDIS_MASTER) {
                /* Don't reply to a master */
                nwritten = c->bufpos - c->sentlen;
            } else {
                nwritten = write(fd,c->buf+c->sentlen,c->bufpos-c->sentlen);
                if (nwritten <= 0) break;
            }
            c->sentlen += nwritten;
            totwritten += nwritten;

            /* If the buffer was sent, set bufpos to zero to continue with
             * the remainder of the reply. */
            if (c->sentlen == c->bufpos) {
                c->bufpos = 0;
                c->sentlen = 0;
            }
        } else {
            o = listNodeValue(listFirst(c->reply));
            objlen = sdslen(o->ptr);
549

550 551 552 553
            if (objlen == 0) {
                listDelNode(c->reply,listFirst(c->reply));
                continue;
            }
554

555 556 557 558 559 560 561 562 563
            if (c->flags & REDIS_MASTER) {
                /* Don't reply to a master */
                nwritten = objlen - c->sentlen;
            } else {
                nwritten = write(fd, ((char*)o->ptr)+c->sentlen,objlen-c->sentlen);
                if (nwritten <= 0) break;
            }
            c->sentlen += nwritten;
            totwritten += nwritten;
564

565 566 567 568 569
            /* If we fully sent the object on head go to the next one */
            if (c->sentlen == objlen) {
                listDelNode(c->reply,listFirst(c->reply));
                c->sentlen = 0;
            }
570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591
        }
        /* Note that we avoid to send more thank REDIS_MAX_WRITE_PER_EVENT
         * bytes, in a single threaded server it's a good idea to serve
         * other clients as well, even if a very large request comes from
         * super fast link that is always able to accept data (in real world
         * scenario think about 'KEYS *' against the loopback interfae) */
        if (totwritten > REDIS_MAX_WRITE_PER_EVENT) break;
    }
    if (nwritten == -1) {
        if (errno == EAGAIN) {
            nwritten = 0;
        } else {
            redisLog(REDIS_VERBOSE,
                "Error writing to client: %s", strerror(errno));
            freeClient(c);
            return;
        }
    }
    if (totwritten > 0) c->lastinteraction = time(NULL);
    if (listLength(c->reply) == 0) {
        c->sentlen = 0;
        aeDeleteFileEvent(server.el,c->fd,AE_WRITABLE);
P
Pieter Noordhuis 已提交
592 593

        /* Close connection after entire reply has been sent. */
594
        if (c->flags & REDIS_CLOSE_AFTER_REPLY) freeClient(c);
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
    }
}

void sendReplyToClientWritev(aeEventLoop *el, int fd, void *privdata, int mask)
{
    redisClient *c = privdata;
    int nwritten = 0, totwritten = 0, objlen, willwrite;
    robj *o;
    struct iovec iov[REDIS_WRITEV_IOVEC_COUNT];
    int offset, ion = 0;
    REDIS_NOTUSED(el);
    REDIS_NOTUSED(mask);

    listNode *node;
    while (listLength(c->reply)) {
        offset = c->sentlen;
        ion = 0;
        willwrite = 0;

        /* fill-in the iov[] array */
        for(node = listFirst(c->reply); node; node = listNextNode(node)) {
            o = listNodeValue(node);
            objlen = sdslen(o->ptr);

            if (totwritten + objlen - offset > REDIS_MAX_WRITE_PER_EVENT)
                break;

            if(ion == REDIS_WRITEV_IOVEC_COUNT)
                break; /* no more iovecs */

            iov[ion].iov_base = ((char*)o->ptr) + offset;
            iov[ion].iov_len = objlen - offset;
            willwrite += objlen - offset;
            offset = 0; /* just for the first item */
            ion++;
        }

        if(willwrite == 0)
            break;

        /* write all collected blocks at once */
        if((nwritten = writev(fd, iov, ion)) < 0) {
            if (errno != EAGAIN) {
                redisLog(REDIS_VERBOSE,
                         "Error writing to client: %s", strerror(errno));
                freeClient(c);
                return;
            }
            break;
        }

        totwritten += nwritten;
        offset = c->sentlen;

        /* remove written robjs from c->reply */
        while (nwritten && listLength(c->reply)) {
            o = listNodeValue(listFirst(c->reply));
            objlen = sdslen(o->ptr);

            if(nwritten >= objlen - offset) {
                listDelNode(c->reply, listFirst(c->reply));
                nwritten -= objlen - offset;
                c->sentlen = 0;
            } else {
                /* partial write */
                c->sentlen += nwritten;
                break;
            }
            offset = 0;
        }
    }

    if (totwritten > 0)
        c->lastinteraction = time(NULL);

    if (listLength(c->reply) == 0) {
        c->sentlen = 0;
        aeDeleteFileEvent(server.el,c->fd,AE_WRITABLE);
    }
}

/* resetClient prepare the client to process the next command */
void resetClient(redisClient *c) {
    freeClientArgv(c);
679 680
    c->reqtype = 0;
    c->multibulklen = 0;
681 682 683 684 685 686 687 688 689 690 691 692 693 694 695
    c->bulklen = -1;
}

void closeTimedoutClients(void) {
    redisClient *c;
    listNode *ln;
    time_t now = time(NULL);
    listIter li;

    listRewind(server.clients,&li);
    while ((ln = listNext(&li)) != NULL) {
        c = listNodeValue(ln);
        if (server.maxidletime &&
            !(c->flags & REDIS_SLAVE) &&    /* no timeout for slaves */
            !(c->flags & REDIS_MASTER) &&   /* no timeout for masters */
696
            !(c->flags & REDIS_BLOCKED) &&  /* no timeout for BLPOP */
697 698 699 700 701 702 703
            dictSize(c->pubsub_channels) == 0 && /* no timeout for pubsub */
            listLength(c->pubsub_patterns) == 0 &&
            (now - c->lastinteraction > server.maxidletime))
        {
            redisLog(REDIS_VERBOSE,"Closing idle client");
            freeClient(c);
        } else if (c->flags & REDIS_BLOCKED) {
704
            if (c->bpop.timeout != 0 && c->bpop.timeout < now) {
705 706 707 708 709 710 711
                addReply(c,shared.nullmultibulk);
                unblockClientWaitingData(c);
            }
        }
    }
}

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
int processInlineBuffer(redisClient *c) {
    char *newline = strstr(c->querybuf,"\r\n");
    int argc, j;
    sds *argv;
    size_t querylen;

    /* Nothing to do without a \r\n */
    if (newline == NULL)
        return REDIS_ERR;

    /* Split the input buffer up to the \r\n */
    querylen = newline-(c->querybuf);
    argv = sdssplitlen(c->querybuf,querylen," ",1,&argc);

    /* Leave data after the first line of the query in the buffer */
    c->querybuf = sdsrange(c->querybuf,querylen+2,-1);

    /* Setup argv array on client structure */
    if (c->argv) zfree(c->argv);
    c->argv = zmalloc(sizeof(robj*)*argc);

    /* Create redis objects for all arguments. */
    for (c->argc = 0, j = 0; j < argc; j++) {
        if (sdslen(argv[j])) {
            c->argv[c->argc] = createObject(REDIS_STRING,argv[j]);
            c->argc++;
        } else {
            sdsfree(argv[j]);
        }
    }
    zfree(argv);
    return REDIS_OK;
}

/* Helper function. Trims query buffer to make the function that processes
 * multi bulk requests idempotent. */
static void setProtocolError(redisClient *c, int pos) {
    c->flags |= REDIS_CLOSE_AFTER_REPLY;
    c->querybuf = sdsrange(c->querybuf,pos,-1);
}

int processMultibulkBuffer(redisClient *c) {
    char *newline = NULL;
    char *eptr;
    int pos = 0, tolerr;
    long bulklen;

    if (c->multibulklen == 0) {
        /* The client should have been reset */
        redisAssert(c->argc == 0);

        /* Multi bulk length cannot be read without a \r\n */
        newline = strstr(c->querybuf,"\r\n");
        if (newline == NULL)
            return REDIS_ERR;

        /* We know for sure there is a whole line since newline != NULL,
         * so go ahead and find out the multi bulk length. */
        redisAssert(c->querybuf[0] == '*');
        c->multibulklen = strtol(c->querybuf+1,&eptr,10);
        pos = (newline-c->querybuf)+2;
        if (c->multibulklen <= 0) {
            c->querybuf = sdsrange(c->querybuf,pos,-1);
            return REDIS_OK;
776 777 778 779
        } else if (c->multibulklen > 1024*1024) {
            addReplyError(c,"Protocol error: invalid multibulk length");
            setProtocolError(c,pos);
            return REDIS_ERR;
780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801
        }

        /* Setup argv array on client structure */
        if (c->argv) zfree(c->argv);
        c->argv = zmalloc(sizeof(robj*)*c->multibulklen);

        /* Search new newline */
        newline = strstr(c->querybuf+pos,"\r\n");
    }

    redisAssert(c->multibulklen > 0);
    while(c->multibulklen) {
        /* Read bulk length if unknown */
        if (c->bulklen == -1) {
            newline = strstr(c->querybuf+pos,"\r\n");
            if (newline != NULL) {
                if (c->querybuf[pos] != '$') {
                    addReplyErrorFormat(c,
                        "Protocol error: expected '$', got '%c'",
                        c->querybuf[pos]);
                    setProtocolError(c,pos);
                    return REDIS_ERR;
802
                }
803 804 805 806 807 808 809 810 811 812 813 814 815 816 817

                bulklen = strtol(c->querybuf+pos+1,&eptr,10);
                tolerr = (eptr[0] != '\r');
                if (tolerr || bulklen == LONG_MIN || bulklen == LONG_MAX ||
                    bulklen < 0 || bulklen > 1024*1024*1024)
                {
                    addReplyError(c,"Protocol error: invalid bulk length");
                    setProtocolError(c,pos);
                    return REDIS_ERR;
                }
                pos += eptr-(c->querybuf+pos)+2;
                c->bulklen = bulklen;
            } else {
                /* No newline in current buffer, so wait for more data */
                break;
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
        }

        /* Read bulk argument */
        if (sdslen(c->querybuf)-pos < (unsigned)(c->bulklen+2)) {
            /* Not enough data (+2 == trailing \r\n) */
            break;
        } else {
            c->argv[c->argc++] = createStringObject(c->querybuf+pos,c->bulklen);
            pos += c->bulklen+2;
            c->bulklen = -1;
            c->multibulklen--;
        }
    }

    /* Trim to pos */
    c->querybuf = sdsrange(c->querybuf,pos,-1);

    /* We're done when c->multibulk == 0 */
    if (c->multibulklen == 0) {
        return REDIS_OK;
    }
    return REDIS_ERR;
}

void processInputBuffer(redisClient *c) {
    /* Keep processing while there is something in the input buffer */
    while(sdslen(c->querybuf)) {
846
        /* Immediately abort if the client is in the middle of something. */
847 848
        if (c->flags & REDIS_BLOCKED || c->flags & REDIS_IO_WAIT) return;

849 850 851 852
        /* REDIS_CLOSE_AFTER_REPLY closes the connection once the reply is
         * written to the client. Make sure to not let the reply grow after
         * this flag has been set (i.e. don't process more commands). */
        if (c->flags & REDIS_CLOSE_AFTER_REPLY) return;
853 854 855 856 857

        /* Determine request type when unknown. */
        if (!c->reqtype) {
            if (c->querybuf[0] == '*') {
                c->reqtype = REDIS_REQ_MULTIBULK;
858
            } else {
859
                c->reqtype = REDIS_REQ_INLINE;
860 861
            }
        }
862 863 864 865 866 867 868

        if (c->reqtype == REDIS_REQ_INLINE) {
            if (processInlineBuffer(c) != REDIS_OK) break;
        } else if (c->reqtype == REDIS_REQ_MULTIBULK) {
            if (processMultibulkBuffer(c) != REDIS_OK) break;
        } else {
            redisPanic("Unknown request type");
869
        }
870 871

        /* Multibulk processing could see a <= 0 length. */
872 873 874 875 876 877 878
        if (c->argc == 0) {
            resetClient(c);
        } else {
            /* Only reset the client when the command was executed. */
            if (processCommand(c) == REDIS_OK)
                resetClient(c);
        }
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
    }
}

void readQueryFromClient(aeEventLoop *el, int fd, void *privdata, int mask) {
    redisClient *c = (redisClient*) privdata;
    char buf[REDIS_IOBUF_LEN];
    int nread;
    REDIS_NOTUSED(el);
    REDIS_NOTUSED(mask);

    nread = read(fd, buf, REDIS_IOBUF_LEN);
    if (nread == -1) {
        if (errno == EAGAIN) {
            nread = 0;
        } else {
            redisLog(REDIS_VERBOSE, "Reading from client: %s",strerror(errno));
            freeClient(c);
            return;
        }
    } else if (nread == 0) {
        redisLog(REDIS_VERBOSE, "Client closed connection");
        freeClient(c);
        return;
    }
    if (nread) {
904
        c->querybuf = sdscatlen(c->querybuf,buf,nread);
905 906 907 908 909 910
        c->lastinteraction = time(NULL);
    } else {
        return;
    }
    processInputBuffer(c);
}