qemu_process.c 74.9 KB
Newer Older
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
/*
 * qemu_process.h: QEMU process management
 *
 * Copyright (C) 2006-2011 Red Hat, Inc.
 *
 * This library is free software; you can redistribute it and/or
 * modify it under the terms of the GNU Lesser General Public
 * License as published by the Free Software Foundation; either
 * version 2.1 of the License, or (at your option) any later version.
 *
 * This library is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * Lesser General Public License for more details.
 *
 * You should have received a copy of the GNU Lesser General Public
 * License along with this library; if not, write to the Free Software
 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307  USA
 *
 */

#include <config.h>

#include <fcntl.h>
#include <unistd.h>
#include <signal.h>
#include <sys/stat.h>

#include "qemu_process.h"
#include "qemu_domain.h"
#include "qemu_cgroup.h"
#include "qemu_capabilities.h"
#include "qemu_monitor.h"
#include "qemu_command.h"
#include "qemu_audit.h"
#include "qemu_hostdev.h"
#include "qemu_hotplug.h"
#include "qemu_bridge_filter.h"

#include "datatypes.h"
#include "logging.h"
#include "virterror_internal.h"
#include "memory.h"
#include "hooks.h"
#include "files.h"
#include "util.h"
#include "c-ctype.h"
#include "nodeinfo.h"
#include "processinfo.h"
#include "domain_nwfilter.h"

#define VIR_FROM_THIS VIR_FROM_QEMU

#define START_POSTFIX ": starting up\n"
#define SHUTDOWN_POSTFIX ": shutting down\n"

/**
 * qemudRemoveDomainStatus
 *
 * remove all state files of a domain from statedir
 *
 * Returns 0 on success
 */
static int
qemuProcessRemoveDomainStatus(struct qemud_driver *driver,
                              virDomainObjPtr vm)
{
    char ebuf[1024];
    char *file = NULL;

    if (virAsprintf(&file, "%s/%s.xml", driver->stateDir, vm->def->name) < 0) {
        virReportOOMError();
        return(-1);
    }

    if (unlink(file) < 0 && errno != ENOENT && errno != ENOTDIR)
        VIR_WARN("Failed to remove domain XML for %s: %s",
                 vm->def->name, virStrerror(errno, ebuf, sizeof(ebuf)));
    VIR_FREE(file);

    if (virFileDeletePid(driver->stateDir, vm->def->name) != 0)
        VIR_WARN("Failed to remove PID file for %s: %s",
                 vm->def->name, virStrerror(errno, ebuf, sizeof(ebuf)));


    return 0;
}


/* XXX figure out how to remove this */
extern struct qemud_driver *qemu_driver;

/*
 * This is a callback registered with a qemuMonitorPtr  instance,
 * and to be invoked when the monitor console hits an end of file
 * condition, or error, thus indicating VM shutdown should be
 * performed
 */
static void
qemuProcessHandleMonitorEOF(qemuMonitorPtr mon ATTRIBUTE_UNUSED,
                            virDomainObjPtr vm,
                            int hasError)
{
    struct qemud_driver *driver = qemu_driver;
    virDomainEventPtr event = NULL;
    qemuDomainObjPrivatePtr priv;

    VIR_DEBUG("Received EOF on %p '%s'", vm, vm->def->name);

110
    qemuDriverLock(driver);
111 112 113 114 115
    virDomainObjLock(vm);

    if (!virDomainObjIsActive(vm)) {
        VIR_DEBUG("Domain %p is not active, ignoring EOF", vm);
        virDomainObjUnlock(vm);
116
        qemuDriverUnlock(driver);
117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133
        return;
    }

    priv = vm->privateData;
    if (!hasError && priv->monJSON && !priv->gotShutdown) {
        VIR_DEBUG("Monitor connection to '%s' closed without SHUTDOWN event; "
                  "assuming the domain crashed", vm->def->name);
        hasError = 1;
    }

    event = virDomainEventNewFromObj(vm,
                                     VIR_DOMAIN_EVENT_STOPPED,
                                     hasError ?
                                     VIR_DOMAIN_EVENT_STOPPED_FAILED :
                                     VIR_DOMAIN_EVENT_STOPPED_SHUTDOWN);

    qemuProcessStop(driver, vm, 0);
134
    qemuAuditDomainStop(vm, hasError ? "failed" : "shutdown");
135 136 137 138 139 140 141 142 143

    if (!vm->persistent)
        virDomainRemoveInactive(&driver->domains, vm);
    else
        virDomainObjUnlock(vm);

    if (event) {
        qemuDomainEventQueue(driver, event);
    }
144
    qemuDriverUnlock(driver);
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
}


static virDomainDiskDefPtr
qemuProcessFindDomainDiskByPath(virDomainObjPtr vm,
                                const char *path)
{
    int i;

    for (i = 0; i < vm->def->ndisks; i++) {
        virDomainDiskDefPtr disk;

        disk = vm->def->disks[i];
        if (disk->src != NULL && STREQ(disk->src, path))
            return disk;
    }

    qemuReportError(VIR_ERR_INTERNAL_ERROR,
                    _("no disk found with path %s"),
                    path);
    return NULL;
}

static virDomainDiskDefPtr
qemuProcessFindDomainDiskByAlias(virDomainObjPtr vm,
                                 const char *alias)
{
    int i;

    if (STRPREFIX(alias, QEMU_DRIVE_HOST_PREFIX))
        alias += strlen(QEMU_DRIVE_HOST_PREFIX);

    for (i = 0; i < vm->def->ndisks; i++) {
        virDomainDiskDefPtr disk;

        disk = vm->def->disks[i];
        if (disk->info.alias != NULL && STREQ(disk->info.alias, alias))
            return disk;
    }

    qemuReportError(VIR_ERR_INTERNAL_ERROR,
                    _("no disk found with alias %s"),
                    alias);
    return NULL;
}

static int
qemuProcessGetVolumeQcowPassphrase(virConnectPtr conn,
                                   virDomainDiskDefPtr disk,
                                   char **secretRet,
                                   size_t *secretLen)
{
    virSecretPtr secret;
    char *passphrase;
    unsigned char *data;
    size_t size;
    int ret = -1;
    virStorageEncryptionPtr enc;

    if (!disk->encryption) {
        qemuReportError(VIR_ERR_INTERNAL_ERROR,
                        _("disk %s does not have any encryption information"),
                        disk->src);
        return -1;
    }
    enc = disk->encryption;

    if (!conn) {
        qemuReportError(VIR_ERR_INTERNAL_ERROR,
                        "%s", _("cannot find secrets without a connection"));
        goto cleanup;
    }

    if (conn->secretDriver == NULL ||
        conn->secretDriver->lookupByUUID == NULL ||
        conn->secretDriver->getValue == NULL) {
        qemuReportError(VIR_ERR_NO_SUPPORT, "%s",
                        _("secret storage not supported"));
        goto cleanup;
    }

    if (enc->format != VIR_STORAGE_ENCRYPTION_FORMAT_QCOW ||
        enc->nsecrets != 1 ||
        enc->secrets[0]->type !=
        VIR_STORAGE_ENCRYPTION_SECRET_TYPE_PASSPHRASE) {
        qemuReportError(VIR_ERR_XML_ERROR,
                        _("invalid <encryption> for volume %s"), disk->src);
        goto cleanup;
    }

    secret = conn->secretDriver->lookupByUUID(conn,
                                              enc->secrets[0]->uuid);
    if (secret == NULL)
        goto cleanup;
    data = conn->secretDriver->getValue(secret, &size,
                                        VIR_SECRET_GET_VALUE_INTERNAL_CALL);
    virUnrefSecret(secret);
    if (data == NULL)
        goto cleanup;

    if (memchr(data, '\0', size) != NULL) {
        memset(data, 0, size);
        VIR_FREE(data);
        qemuReportError(VIR_ERR_XML_ERROR,
                        _("format='qcow' passphrase for %s must not contain a "
                          "'\\0'"), disk->src);
        goto cleanup;
    }

    if (VIR_ALLOC_N(passphrase, size + 1) < 0) {
        memset(data, 0, size);
        VIR_FREE(data);
        virReportOOMError();
        goto cleanup;
    }
    memcpy(passphrase, data, size);
    passphrase[size] = '\0';

    memset(data, 0, size);
    VIR_FREE(data);

    *secretRet = passphrase;
    *secretLen = size;

    ret = 0;

cleanup:
    return ret;
}

static int
qemuProcessFindVolumeQcowPassphrase(qemuMonitorPtr mon ATTRIBUTE_UNUSED,
                                    virConnectPtr conn,
                                    virDomainObjPtr vm,
                                    const char *path,
                                    char **secretRet,
                                    size_t *secretLen)
{
    virDomainDiskDefPtr disk;
    int ret = -1;

    virDomainObjLock(vm);
    disk = qemuProcessFindDomainDiskByPath(vm, path);

    if (!disk)
        goto cleanup;

    ret = qemuProcessGetVolumeQcowPassphrase(conn, disk, secretRet, secretLen);

cleanup:
    virDomainObjUnlock(vm);
    return ret;
}


static int
qemuProcessHandleReset(qemuMonitorPtr mon ATTRIBUTE_UNUSED,
                       virDomainObjPtr vm)
{
    struct qemud_driver *driver = qemu_driver;
    virDomainEventPtr event;

    virDomainObjLock(vm);
    event = virDomainEventRebootNewFromObj(vm);
    virDomainObjUnlock(vm);

    if (event) {
        qemuDriverLock(driver);
        qemuDomainEventQueue(driver, event);
        qemuDriverUnlock(driver);
    }

    return 0;
}


static int
qemuProcessHandleShutdown(qemuMonitorPtr mon ATTRIBUTE_UNUSED,
                          virDomainObjPtr vm)
{
    virDomainObjLock(vm);
    ((qemuDomainObjPrivatePtr) vm->privateData)->gotShutdown = true;
    virDomainObjUnlock(vm);

    return 0;
}


static int
qemuProcessHandleStop(qemuMonitorPtr mon ATTRIBUTE_UNUSED,
                      virDomainObjPtr vm)
{
    struct qemud_driver *driver = qemu_driver;
    virDomainEventPtr event = NULL;

    virDomainObjLock(vm);
    if (vm->state == VIR_DOMAIN_RUNNING) {
342 343
        VIR_DEBUG("Transitioned guest %s to paused state due to unknown event",
                  vm->def->name);
344 345 346 347 348 349

        vm->state = VIR_DOMAIN_PAUSED;
        event = virDomainEventNewFromObj(vm,
                                         VIR_DOMAIN_EVENT_SUSPENDED,
                                         VIR_DOMAIN_EVENT_SUSPENDED_PAUSED);

350 351 352 353
        if (virDomainSaveStatus(driver->caps, driver->stateDir, vm) < 0) {
            VIR_WARN("Unable to save status on vm %s after state change",
                     vm->def->name);
        }
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
    }
    virDomainObjUnlock(vm);

    if (event) {
        qemuDriverLock(driver);
        if (event)
            qemuDomainEventQueue(driver, event);
        qemuDriverUnlock(driver);
    }

    return 0;
}


static int
qemuProcessHandleRTCChange(qemuMonitorPtr mon ATTRIBUTE_UNUSED,
                           virDomainObjPtr vm,
                           long long offset)
{
    struct qemud_driver *driver = qemu_driver;
    virDomainEventPtr event;

    virDomainObjLock(vm);
    event = virDomainEventRTCChangeNewFromObj(vm, offset);

    if (vm->def->clock.offset == VIR_DOMAIN_CLOCK_OFFSET_VARIABLE)
        vm->def->clock.data.adjustment = offset;

    if (virDomainSaveStatus(driver->caps, driver->stateDir, vm) < 0)
        VIR_WARN0("unable to save domain status with RTC change");

    virDomainObjUnlock(vm);

    if (event) {
        qemuDriverLock(driver);
        qemuDomainEventQueue(driver, event);
        qemuDriverUnlock(driver);
    }

    return 0;
}


static int
qemuProcessHandleWatchdog(qemuMonitorPtr mon ATTRIBUTE_UNUSED,
                          virDomainObjPtr vm,
                          int action)
{
    struct qemud_driver *driver = qemu_driver;
    virDomainEventPtr watchdogEvent = NULL;
    virDomainEventPtr lifecycleEvent = NULL;

    virDomainObjLock(vm);
    watchdogEvent = virDomainEventWatchdogNewFromObj(vm, action);

    if (action == VIR_DOMAIN_EVENT_WATCHDOG_PAUSE &&
        vm->state == VIR_DOMAIN_RUNNING) {
        VIR_DEBUG("Transitioned guest %s to paused state due to watchdog", vm->def->name);

        vm->state = VIR_DOMAIN_PAUSED;
        lifecycleEvent = virDomainEventNewFromObj(vm,
                                                  VIR_DOMAIN_EVENT_SUSPENDED,
                                                  VIR_DOMAIN_EVENT_SUSPENDED_WATCHDOG);

418 419 420 421
        if (virDomainSaveStatus(driver->caps, driver->stateDir, vm) < 0) {
            VIR_WARN("Unable to save status on vm %s after watchdog event",
                     vm->def->name);
        }
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
    }

    if (vm->def->watchdog->action == VIR_DOMAIN_WATCHDOG_ACTION_DUMP) {
        struct qemuDomainWatchdogEvent *wdEvent;
        if (VIR_ALLOC(wdEvent) == 0) {
            wdEvent->action = VIR_DOMAIN_WATCHDOG_ACTION_DUMP;
            wdEvent->vm = vm;
            ignore_value(virThreadPoolSendJob(driver->workerPool, wdEvent));
        } else
            virReportOOMError();
    }

    virDomainObjUnlock(vm);

    if (watchdogEvent || lifecycleEvent) {
        qemuDriverLock(driver);
        if (watchdogEvent)
            qemuDomainEventQueue(driver, watchdogEvent);
        if (lifecycleEvent)
            qemuDomainEventQueue(driver, lifecycleEvent);
        qemuDriverUnlock(driver);
    }

    return 0;
}


static int
qemuProcessHandleIOError(qemuMonitorPtr mon ATTRIBUTE_UNUSED,
                         virDomainObjPtr vm,
                         const char *diskAlias,
                         int action,
                         const char *reason)
{
    struct qemud_driver *driver = qemu_driver;
    virDomainEventPtr ioErrorEvent = NULL;
    virDomainEventPtr ioErrorEvent2 = NULL;
    virDomainEventPtr lifecycleEvent = NULL;
    const char *srcPath;
    const char *devAlias;
    virDomainDiskDefPtr disk;

    virDomainObjLock(vm);
    disk = qemuProcessFindDomainDiskByAlias(vm, diskAlias);

    if (disk) {
        srcPath = disk->src;
        devAlias = disk->info.alias;
    } else {
        srcPath = "";
        devAlias = "";
    }

    ioErrorEvent = virDomainEventIOErrorNewFromObj(vm, srcPath, devAlias, action);
    ioErrorEvent2 = virDomainEventIOErrorReasonNewFromObj(vm, srcPath, devAlias, action, reason);

    if (action == VIR_DOMAIN_EVENT_IO_ERROR_PAUSE &&
        vm->state == VIR_DOMAIN_RUNNING) {
        VIR_DEBUG("Transitioned guest %s to paused state due to IO error", vm->def->name);

        vm->state = VIR_DOMAIN_PAUSED;
        lifecycleEvent = virDomainEventNewFromObj(vm,
                                                  VIR_DOMAIN_EVENT_SUSPENDED,
                                                  VIR_DOMAIN_EVENT_SUSPENDED_IOERROR);

        if (virDomainSaveStatus(driver->caps, driver->stateDir, vm) < 0)
            VIR_WARN("Unable to save status on vm %s after IO error", vm->def->name);
    }
    virDomainObjUnlock(vm);

    if (ioErrorEvent || ioErrorEvent2 || lifecycleEvent) {
        qemuDriverLock(driver);
        if (ioErrorEvent)
            qemuDomainEventQueue(driver, ioErrorEvent);
        if (ioErrorEvent2)
            qemuDomainEventQueue(driver, ioErrorEvent2);
        if (lifecycleEvent)
            qemuDomainEventQueue(driver, lifecycleEvent);
        qemuDriverUnlock(driver);
    }

    return 0;
}


static int
qemuProcessHandleGraphics(qemuMonitorPtr mon ATTRIBUTE_UNUSED,
                          virDomainObjPtr vm,
                          int phase,
                          int localFamily,
                          const char *localNode,
                          const char *localService,
                          int remoteFamily,
                          const char *remoteNode,
                          const char *remoteService,
                          const char *authScheme,
                          const char *x509dname,
                          const char *saslUsername)
{
    struct qemud_driver *driver = qemu_driver;
    virDomainEventPtr event;
    virDomainEventGraphicsAddressPtr localAddr = NULL;
    virDomainEventGraphicsAddressPtr remoteAddr = NULL;
    virDomainEventGraphicsSubjectPtr subject = NULL;
    int i;

    virDomainObjLock(vm);

    if (VIR_ALLOC(localAddr) < 0)
        goto no_memory;
    localAddr->family = localFamily;
    if (!(localAddr->service = strdup(localService)) ||
        !(localAddr->node = strdup(localNode)))
        goto no_memory;

    if (VIR_ALLOC(remoteAddr) < 0)
        goto no_memory;
    remoteAddr->family = remoteFamily;
    if (!(remoteAddr->service = strdup(remoteService)) ||
        !(remoteAddr->node = strdup(remoteNode)))
        goto no_memory;

    if (VIR_ALLOC(subject) < 0)
        goto no_memory;
    if (x509dname) {
        if (VIR_REALLOC_N(subject->identities, subject->nidentity+1) < 0)
            goto no_memory;
        if (!(subject->identities[subject->nidentity].type = strdup("x509dname")) ||
            !(subject->identities[subject->nidentity].name = strdup(x509dname)))
            goto no_memory;
        subject->nidentity++;
    }
    if (saslUsername) {
        if (VIR_REALLOC_N(subject->identities, subject->nidentity+1) < 0)
            goto no_memory;
        if (!(subject->identities[subject->nidentity].type = strdup("saslUsername")) ||
            !(subject->identities[subject->nidentity].name = strdup(saslUsername)))
            goto no_memory;
        subject->nidentity++;
    }

    event = virDomainEventGraphicsNewFromObj(vm, phase, localAddr, remoteAddr, authScheme, subject);
    virDomainObjUnlock(vm);

    if (event) {
        qemuDriverLock(driver);
        qemuDomainEventQueue(driver, event);
        qemuDriverUnlock(driver);
    }

    return 0;

no_memory:
    virReportOOMError();
    if (localAddr) {
        VIR_FREE(localAddr->service);
        VIR_FREE(localAddr->node);
        VIR_FREE(localAddr);
    }
    if (remoteAddr) {
        VIR_FREE(remoteAddr->service);
        VIR_FREE(remoteAddr->node);
        VIR_FREE(remoteAddr);
    }
    if (subject) {
        for (i = 0 ; i < subject->nidentity ; i++) {
            VIR_FREE(subject->identities[i].type);
            VIR_FREE(subject->identities[i].name);
        }
        VIR_FREE(subject->identities);
        VIR_FREE(subject);
    }

    return -1;
}


static void qemuProcessHandleMonitorDestroy(qemuMonitorPtr mon,
                                            virDomainObjPtr vm)
{
602 603 604 605
    qemuDomainObjPrivatePtr priv;

    virDomainObjLock(vm);
    priv = vm->privateData;
606 607
    if (priv->mon == mon)
        priv->mon = NULL;
608 609
    if (virDomainObjUnref(vm) > 0)
        virDomainObjUnlock(vm);
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
}

static qemuMonitorCallbacks monitorCallbacks = {
    .destroy = qemuProcessHandleMonitorDestroy,
    .eofNotify = qemuProcessHandleMonitorEOF,
    .diskSecretLookup = qemuProcessFindVolumeQcowPassphrase,
    .domainShutdown = qemuProcessHandleShutdown,
    .domainStop = qemuProcessHandleStop,
    .domainReset = qemuProcessHandleReset,
    .domainRTCChange = qemuProcessHandleRTCChange,
    .domainWatchdog = qemuProcessHandleWatchdog,
    .domainIOError = qemuProcessHandleIOError,
    .domainGraphics = qemuProcessHandleGraphics,
};

static int
qemuConnectMonitor(struct qemud_driver *driver, virDomainObjPtr vm)
{
    qemuDomainObjPrivatePtr priv = vm->privateData;
    int ret = -1;

    if (virSecurityManagerSetSocketLabel(driver->securityManager, vm) < 0) {
        VIR_ERROR(_("Failed to set security context for monitor for %s"),
                  vm->def->name);
        goto error;
    }

    /* Hold an extra reference because we can't allow 'vm' to be
     * deleted while the monitor is active */
    virDomainObjRef(vm);

    priv->mon = qemuMonitorOpen(vm,
                                priv->monConfig,
                                priv->monJSON,
                                &monitorCallbacks);

    if (priv->mon == NULL)
        virDomainObjUnref(vm);

    if (virSecurityManagerClearSocketLabel(driver->securityManager, vm) < 0) {
        VIR_ERROR(_("Failed to clear security context for monitor for %s"),
                  vm->def->name);
        goto error;
    }

    if (priv->mon == NULL) {
        VIR_INFO("Failed to connect monitor for %s", vm->def->name);
        goto error;
    }


    qemuDomainObjEnterMonitorWithDriver(driver, vm);
    ret = qemuMonitorSetCapabilities(priv->mon);
    qemuDomainObjExitMonitorWithDriver(driver, vm);

error:

    return ret;
}

static int
qemuProcessLogFD(struct qemud_driver *driver, const char* name, bool append)
{
    char *logfile;
    mode_t logmode;
    int fd = -1;

    if (virAsprintf(&logfile, "%s/%s.log", driver->logDir, name) < 0) {
        virReportOOMError();
        return -1;
    }

    logmode = O_CREAT | O_WRONLY;
    /* Only logrotate files in /var/log, so only append if running privileged */
    if (driver->privileged || append)
        logmode |= O_APPEND;
    else
        logmode |= O_TRUNC;

    if ((fd = open(logfile, logmode, S_IRUSR | S_IWUSR)) < 0) {
        virReportSystemError(errno,
                             _("failed to create logfile %s"),
                             logfile);
        VIR_FREE(logfile);
        return -1;
    }
    VIR_FREE(logfile);
    if (virSetCloseExec(fd) < 0) {
        virReportSystemError(errno, "%s",
                             _("Unable to set VM logfile close-on-exec flag"));
        VIR_FORCE_CLOSE(fd);
        return -1;
    }
    return fd;
}


static int
qemuProcessLogReadFD(const char* logDir, const char* name, off_t pos)
{
    char *logfile;
    mode_t logmode = O_RDONLY;
    int fd = -1;

    if (virAsprintf(&logfile, "%s/%s.log", logDir, name) < 0) {
        qemuReportError(VIR_ERR_INTERNAL_ERROR,
                        _("failed to build logfile name %s/%s.log"),
                        logDir, name);
        return -1;
    }

    if ((fd = open(logfile, logmode)) < 0) {
        virReportSystemError(errno,
                             _("failed to create logfile %s"),
                             logfile);
        VIR_FREE(logfile);
        return -1;
    }
    if (virSetCloseExec(fd) < 0) {
        virReportSystemError(errno, "%s",
                             _("Unable to set VM logfile close-on-exec flag"));
        VIR_FORCE_CLOSE(fd);
        VIR_FREE(logfile);
        return -1;
    }
    if (pos < 0 || lseek(fd, pos, SEEK_SET) < 0) {
        virReportSystemError(pos < 0 ? 0 : errno,
                             _("Unable to seek to %lld in %s"),
                             (long long) pos, logfile);
        VIR_FORCE_CLOSE(fd);
    }
    VIR_FREE(logfile);
    return fd;
}


typedef int qemuProcessLogHandleOutput(virDomainObjPtr vm,
                                       const char *output,
                                       int fd);

/*
 * Returns -1 for error, 0 on success
 */
static int
qemuProcessReadLogOutput(virDomainObjPtr vm,
                         int fd,
                         char *buf,
                         size_t buflen,
                         qemuProcessLogHandleOutput func,
                         const char *what,
                         int timeout)
{
    int retries = (timeout*10);
    int got = 0;
    buf[0] = '\0';

    while (retries) {
        ssize_t func_ret, ret;
        int isdead = 0;

        func_ret = func(vm, buf, fd);

        if (kill(vm->pid, 0) == -1 && errno == ESRCH)
            isdead = 1;

        /* Any failures should be detected before we read the log, so we
         * always have something useful to report on failure. */
        ret = saferead(fd, buf+got, buflen-got-1);
        if (ret < 0) {
            virReportSystemError(errno,
                                 _("Failure while reading %s log output"),
                                 what);
            return -1;
        }

        got += ret;
        buf[got] = '\0';
        if (got == buflen-1) {
            qemuReportError(VIR_ERR_INTERNAL_ERROR,
                            _("Out of space while reading %s log output: %s"),
                            what, buf);
            return -1;
        }

        if (isdead) {
            qemuReportError(VIR_ERR_INTERNAL_ERROR,
                            _("Process exited while reading %s log output: %s"),
                            what, buf);
            return -1;
        }

        if (func_ret <= 0)
            return func_ret;

        usleep(100*1000);
        retries--;
    }

    qemuReportError(VIR_ERR_INTERNAL_ERROR,
                    _("Timed out while reading %s log output: %s"),
                    what, buf);
    return -1;
}


/*
 * Look at a chunk of data from the QEMU stdout logs and try to
 * find a TTY device, as indicated by a line like
 *
 * char device redirected to /dev/pts/3
 *
 * Returns -1 for error, 0 success, 1 continue reading
 */
static int
qemuProcessExtractTTYPath(const char *haystack,
                          size_t *offset,
                          char **path)
{
    static const char needle[] = "char device redirected to";
    char *tmp, *dev;

    VIR_FREE(*path);
    /* First look for our magic string */
    if (!(tmp = strstr(haystack + *offset, needle))) {
        return 1;
    }
    tmp += sizeof(needle);
    dev = tmp;

    /*
     * And look for first whitespace character and nul terminate
     * to mark end of the pty path
     */
    while (*tmp) {
        if (c_isspace(*tmp)) {
            *path = strndup(dev, tmp-dev);
            if (*path == NULL) {
                virReportOOMError();
                return -1;
            }

            /* ... now further update offset till we get EOL */
            *offset = tmp - haystack;
            return 0;
        }
        tmp++;
    }

    /*
     * We found a path, but didn't find any whitespace,
     * so it must be still incomplete - we should at
     * least see a \n - indicate that we want to carry
     * on trying again
     */
    return 1;
}

static int
qemuProcessFindCharDevicePTYsMonitor(virDomainObjPtr vm,
                                     virHashTablePtr paths)
{
    int i;

#define LOOKUP_PTYS(array, arraylen, idprefix)                            \
    for (i = 0 ; i < (arraylen) ; i++) {                                  \
        virDomainChrDefPtr chr = (array)[i];                              \
        if (chr->source.type == VIR_DOMAIN_CHR_TYPE_PTY) {                \
            char id[16];                                                  \
                                                                          \
            if (snprintf(id, sizeof(id), idprefix "%i", i) >= sizeof(id)) \
                return -1;                                                \
                                                                          \
            const char *path = (const char *) virHashLookup(paths, id);   \
            if (path == NULL) {                                           \
                if (chr->source.data.file.path == NULL) {                 \
                    /* neither the log output nor 'info chardev' had a */ \
                    /* pty path for this chardev, report an error */      \
                    qemuReportError(VIR_ERR_INTERNAL_ERROR,               \
                                    _("no assigned pty for device %s"), id); \
                    return -1;                                            \
                } else {                                                  \
                    /* 'info chardev' had no pty path for this chardev, */\
                    /* but the log output had, so we're fine */           \
                    continue;                                             \
                }                                                         \
            }                                                             \
                                                                          \
            VIR_FREE(chr->source.data.file.path);                         \
            chr->source.data.file.path = strdup(path);                    \
                                                                          \
            if (chr->source.data.file.path == NULL) {                     \
                virReportOOMError();                                      \
                return -1;                                                \
            }                                                             \
        }                                                                 \
    }

    LOOKUP_PTYS(vm->def->serials,   vm->def->nserials,   "serial");
    LOOKUP_PTYS(vm->def->parallels, vm->def->nparallels, "parallel");
    LOOKUP_PTYS(vm->def->channels,  vm->def->nchannels,  "channel");
    if (vm->def->console)
        LOOKUP_PTYS(&vm->def->console, 1,  "console");
#undef LOOKUP_PTYS

    return 0;
}

static int
qemuProcessFindCharDevicePTYs(virDomainObjPtr vm,
                              const char *output,
                              int fd ATTRIBUTE_UNUSED)
{
    size_t offset = 0;
    int ret, i;

    /* The order in which QEMU prints out the PTY paths is
       the order in which it procsses its serial and parallel
       device args. This code must match that ordering.... */

    /* first comes the serial devices */
    for (i = 0 ; i < vm->def->nserials ; i++) {
        virDomainChrDefPtr chr = vm->def->serials[i];
        if (chr->source.type == VIR_DOMAIN_CHR_TYPE_PTY) {
            if ((ret = qemuProcessExtractTTYPath(output, &offset,
                                                 &chr->source.data.file.path)) != 0)
                return ret;
        }
    }

    /* then the parallel devices */
    for (i = 0 ; i < vm->def->nparallels ; i++) {
        virDomainChrDefPtr chr = vm->def->parallels[i];
        if (chr->source.type == VIR_DOMAIN_CHR_TYPE_PTY) {
            if ((ret = qemuProcessExtractTTYPath(output, &offset,
                                                 &chr->source.data.file.path)) != 0)
                return ret;
        }
    }

    /* then the channel devices */
    for (i = 0 ; i < vm->def->nchannels ; i++) {
        virDomainChrDefPtr chr = vm->def->channels[i];
        if (chr->source.type == VIR_DOMAIN_CHR_TYPE_PTY) {
            if ((ret = qemuProcessExtractTTYPath(output, &offset,
                                                 &chr->source.data.file.path)) != 0)
                return ret;
        }
    }

    return 0;
}

962
static void qemuProcessFreePtyPath(void *payload, const void *name ATTRIBUTE_UNUSED)
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
{
    VIR_FREE(payload);
}

static void
qemuProcessReadLogFD(int logfd, char *buf, int maxlen, int off)
{
    int ret;
    char *tmpbuf = buf + off;

    ret = saferead(logfd, tmpbuf, maxlen - off - 1);
    if (ret < 0) {
        ret = 0;
    }

    tmpbuf[ret] = '\0';
}

static int
qemuProcessWaitForMonitor(struct qemud_driver* driver,
                          virDomainObjPtr vm, off_t pos)
{
    char buf[4096] = ""; /* Plenty of space to get startup greeting */
    int logfd;
    int ret = -1;
    virHashTablePtr paths = NULL;

    if ((logfd = qemuProcessLogReadFD(driver->logDir, vm->def->name, pos)) < 0)
        return -1;

    if (qemuProcessReadLogOutput(vm, logfd, buf, sizeof(buf),
                                 qemuProcessFindCharDevicePTYs,
                                 "console", 30) < 0)
        goto closelog;

    VIR_DEBUG("Connect monitor to %p '%s'", vm, vm->def->name);
    if (qemuConnectMonitor(driver, vm) < 0) {
        goto cleanup;
    }

    /* Try to get the pty path mappings again via the monitor. This is much more
     * reliable if it's available.
     * Note that the monitor itself can be on a pty, so we still need to try the
     * log output method. */
1007
    paths = virHashCreate(0, qemuProcessFreePtyPath);
1008
    if (paths == NULL)
1009 1010 1011
        goto cleanup;

    qemuDomainObjPrivatePtr priv = vm->privateData;
E
Eric Blake 已提交
1012
    qemuDomainObjEnterMonitorWithDriver(driver, vm);
1013 1014 1015 1016 1017 1018 1019 1020
    ret = qemuMonitorGetPtyPaths(priv->mon, paths);
    qemuDomainObjExitMonitorWithDriver(driver, vm);

    VIR_DEBUG("qemuMonitorGetPtyPaths returned %i", ret);
    if (ret == 0)
        ret = qemuProcessFindCharDevicePTYsMonitor(vm, paths);

cleanup:
1021
    virHashFree(paths);
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

    if (kill(vm->pid, 0) == -1 && errno == ESRCH) {
        /* VM is dead, any other error raised in the interim is probably
         * not as important as the qemu cmdline output */
        qemuProcessReadLogFD(logfd, buf, sizeof(buf), strlen(buf));
        qemuReportError(VIR_ERR_INTERNAL_ERROR,
                        _("process exited while connecting to monitor: %s"),
                        buf);
        ret = -1;
    }

closelog:
    if (VIR_CLOSE(logfd) < 0) {
        char ebuf[4096];
        VIR_WARN("Unable to close logfile: %s",
                 virStrerror(errno, ebuf, sizeof ebuf));
    }

    return ret;
}

static int
qemuProcessDetectVcpuPIDs(struct qemud_driver *driver,
                          virDomainObjPtr vm)
{
    pid_t *cpupids = NULL;
    int ncpupids;
    qemuDomainObjPrivatePtr priv = vm->privateData;

    if (vm->def->virtType != VIR_DOMAIN_VIRT_KVM) {
        priv->nvcpupids = 1;
        if (VIR_ALLOC_N(priv->vcpupids, priv->nvcpupids) < 0) {
            virReportOOMError();
            return -1;
        }
        priv->vcpupids[0] = vm->pid;
        return 0;
    }

    /* What follows is now all KVM specific */

    qemuDomainObjEnterMonitorWithDriver(driver, vm);
    if ((ncpupids = qemuMonitorGetCPUInfo(priv->mon, &cpupids)) < 0) {
        qemuDomainObjExitMonitorWithDriver(driver, vm);
        return -1;
    }
    qemuDomainObjExitMonitorWithDriver(driver, vm);

    /* Treat failure to get VCPU<->PID mapping as non-fatal */
    if (ncpupids == 0)
        return 0;

    if (ncpupids != vm->def->vcpus) {
        qemuReportError(VIR_ERR_INTERNAL_ERROR,
                        _("got wrong number of vCPU pids from QEMU monitor. "
                          "got %d, wanted %d"),
                        ncpupids, vm->def->vcpus);
        VIR_FREE(cpupids);
        return -1;
    }

    priv->nvcpupids = ncpupids;
    priv->vcpupids = cpupids;
    return 0;
}

/*
 * To be run between fork/exec of QEMU only
 */
static int
qemuProcessInitCpuAffinity(virDomainObjPtr vm)
{
    int i, hostcpus, maxcpu = QEMUD_CPUMASK_LEN;
    virNodeInfo nodeinfo;
    unsigned char *cpumap;
    int cpumaplen;

1099
    VIR_DEBUG0("Setting CPU affinity");
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

    if (nodeGetInfo(NULL, &nodeinfo) < 0)
        return -1;

    /* setaffinity fails if you set bits for CPUs which
     * aren't present, so we have to limit ourselves */
    hostcpus = VIR_NODEINFO_MAXCPUS(nodeinfo);
    if (maxcpu > hostcpus)
        maxcpu = hostcpus;

    cpumaplen = VIR_CPU_MAPLEN(maxcpu);
    if (VIR_ALLOC_N(cpumap, cpumaplen) < 0) {
        virReportOOMError();
        return -1;
    }

    if (vm->def->cpumask) {
        /* XXX why don't we keep 'cpumask' in the libvirt cpumap
         * format to start with ?!?! */
        for (i = 0 ; i < maxcpu && i < vm->def->cpumasklen ; i++)
            if (vm->def->cpumask[i])
                VIR_USE_CPU(cpumap, i);
    } else {
        /* You may think this is redundant, but we can't assume libvirtd
         * itself is running on all pCPUs, so we need to explicitly set
         * the spawned QEMU instance to all pCPUs if no map is given in
         * its config file */
        for (i = 0 ; i < maxcpu ; i++)
            VIR_USE_CPU(cpumap, i);
    }

    /* We are pressuming we are running between fork/exec of QEMU
     * so use '0' to indicate our own process ID. No threads are
     * running at this point
     */
    if (virProcessInfoSetAffinity(0, /* Self */
                                  cpumap, cpumaplen, maxcpu) < 0) {
        VIR_FREE(cpumap);
        return -1;
    }
    VIR_FREE(cpumap);

    return 0;
}


static int
qemuProcessInitPasswords(virConnectPtr conn,
                         struct qemud_driver *driver,
                         virDomainObjPtr vm,
1150
                         virBitmapPtr qemuCaps)
1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171
{
    int ret = 0;
    qemuDomainObjPrivatePtr priv = vm->privateData;

    if (vm->def->ngraphics == 1) {
        if (vm->def->graphics[0]->type == VIR_DOMAIN_GRAPHICS_TYPE_VNC) {
            ret = qemuDomainChangeGraphicsPasswords(driver, vm,
                                                    VIR_DOMAIN_GRAPHICS_TYPE_VNC,
                                                    &vm->def->graphics[0]->data.vnc.auth,
                                                    driver->vncPassword);
        } else if (vm->def->graphics[0]->type == VIR_DOMAIN_GRAPHICS_TYPE_SPICE) {
            ret = qemuDomainChangeGraphicsPasswords(driver, vm,
                                                    VIR_DOMAIN_GRAPHICS_TYPE_SPICE,
                                                    &vm->def->graphics[0]->data.spice.auth,
                                                    driver->spicePassword);
        }
    }

    if (ret < 0)
        goto cleanup;

1172
    if (qemuCapsGet(qemuCaps, QEMU_CAPS_DEVICE)) {
1173 1174 1175 1176 1177
        int i;

        for (i = 0 ; i < vm->def->ndisks ; i++) {
            char *secret;
            size_t secretLen;
E
Eric Blake 已提交
1178
            const char *alias;
1179 1180 1181 1182 1183 1184 1185 1186 1187 1188

            if (!vm->def->disks[i]->encryption ||
                !vm->def->disks[i]->src)
                continue;

            if (qemuProcessGetVolumeQcowPassphrase(conn,
                                                   vm->def->disks[i],
                                                   &secret, &secretLen) < 0)
                goto cleanup;

E
Eric Blake 已提交
1189
            alias = vm->def->disks[i]->info.alias;
1190
            qemuDomainObjEnterMonitorWithDriver(driver, vm);
E
Eric Blake 已提交
1191
            ret = qemuMonitorSetDrivePassphrase(priv->mon, alias, secret);
1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729
            VIR_FREE(secret);
            qemuDomainObjExitMonitorWithDriver(driver, vm);
            if (ret < 0)
                goto cleanup;
        }
    }

cleanup:
    return ret;
}


#define QEMU_PCI_VENDOR_INTEL     0x8086
#define QEMU_PCI_VENDOR_LSI_LOGIC 0x1000
#define QEMU_PCI_VENDOR_REDHAT    0x1af4
#define QEMU_PCI_VENDOR_CIRRUS    0x1013
#define QEMU_PCI_VENDOR_REALTEK   0x10ec
#define QEMU_PCI_VENDOR_AMD       0x1022
#define QEMU_PCI_VENDOR_ENSONIQ   0x1274
#define QEMU_PCI_VENDOR_VMWARE    0x15ad
#define QEMU_PCI_VENDOR_QEMU      0x1234

#define QEMU_PCI_PRODUCT_DISK_VIRTIO 0x1001

#define QEMU_PCI_PRODUCT_BALLOON_VIRTIO 0x1002

#define QEMU_PCI_PRODUCT_NIC_NE2K     0x8029
#define QEMU_PCI_PRODUCT_NIC_PCNET    0x2000
#define QEMU_PCI_PRODUCT_NIC_RTL8139  0x8139
#define QEMU_PCI_PRODUCT_NIC_E1000    0x100E
#define QEMU_PCI_PRODUCT_NIC_VIRTIO   0x1000

#define QEMU_PCI_PRODUCT_VGA_CIRRUS 0x00b8
#define QEMU_PCI_PRODUCT_VGA_VMWARE 0x0405
#define QEMU_PCI_PRODUCT_VGA_STDVGA 0x1111

#define QEMU_PCI_PRODUCT_AUDIO_AC97    0x2415
#define QEMU_PCI_PRODUCT_AUDIO_ES1370  0x5000

#define QEMU_PCI_PRODUCT_CONTROLLER_PIIX 0x7010
#define QEMU_PCI_PRODUCT_CONTROLLER_LSI  0x0012

#define QEMU_PCI_PRODUCT_WATCHDOG_I63000ESB 0x25ab

static int
qemuProcessAssignNextPCIAddress(virDomainDeviceInfo *info,
                                int vendor,
                                int product,
                                qemuMonitorPCIAddress *addrs,
                                int naddrs)
{
    int found = 0;
    int i;

    VIR_DEBUG("Look for %x:%x out of %d", vendor, product, naddrs);

    for (i = 0 ; (i < naddrs) && !found; i++) {
        VIR_DEBUG("Maybe %x:%x", addrs[i].vendor, addrs[i].product);
        if (addrs[i].vendor == vendor &&
            addrs[i].product == product) {
            VIR_DEBUG("Match %d", i);
            found = 1;
            break;
        }
    }
    if (!found) {
        return -1;
    }

    /* Blank it out so this device isn't matched again */
    addrs[i].vendor = 0;
    addrs[i].product = 0;

    if (info->type == VIR_DOMAIN_DEVICE_ADDRESS_TYPE_NONE)
        info->type = VIR_DOMAIN_DEVICE_ADDRESS_TYPE_PCI;

    if (info->type == VIR_DOMAIN_DEVICE_ADDRESS_TYPE_PCI) {
        info->addr.pci.domain = addrs[i].addr.domain;
        info->addr.pci.bus = addrs[i].addr.bus;
        info->addr.pci.slot = addrs[i].addr.slot;
        info->addr.pci.function = addrs[i].addr.function;
    }

    return 0;
}

static int
qemuProcessGetPCIDiskVendorProduct(virDomainDiskDefPtr def,
                                   unsigned *vendor,
                                   unsigned *product)
{
    switch (def->bus) {
    case VIR_DOMAIN_DISK_BUS_VIRTIO:
        *vendor = QEMU_PCI_VENDOR_REDHAT;
        *product = QEMU_PCI_PRODUCT_DISK_VIRTIO;
        break;

    default:
        return -1;
    }

    return 0;
}

static int
qemuProcessGetPCINetVendorProduct(virDomainNetDefPtr def,
                                  unsigned *vendor,
                                  unsigned *product)
{
    if (!def->model)
        return -1;

    if (STREQ(def->model, "ne2k_pci")) {
        *vendor = QEMU_PCI_VENDOR_REALTEK;
        *product = QEMU_PCI_PRODUCT_NIC_NE2K;
    } else if (STREQ(def->model, "pcnet")) {
        *vendor = QEMU_PCI_VENDOR_AMD;
        *product = QEMU_PCI_PRODUCT_NIC_PCNET;
    } else if (STREQ(def->model, "rtl8139")) {
        *vendor = QEMU_PCI_VENDOR_REALTEK;
        *product = QEMU_PCI_PRODUCT_NIC_RTL8139;
    } else if (STREQ(def->model, "e1000")) {
        *vendor = QEMU_PCI_VENDOR_INTEL;
        *product = QEMU_PCI_PRODUCT_NIC_E1000;
    } else if (STREQ(def->model, "virtio")) {
        *vendor = QEMU_PCI_VENDOR_REDHAT;
        *product = QEMU_PCI_PRODUCT_NIC_VIRTIO;
    } else {
        VIR_INFO("Unexpected NIC model %s, cannot get PCI address",
                 def->model);
        return -1;
    }
    return 0;
}

static int
qemuProcessGetPCIControllerVendorProduct(virDomainControllerDefPtr def,
                                         unsigned *vendor,
                                         unsigned *product)
{
    switch (def->type) {
    case VIR_DOMAIN_CONTROLLER_TYPE_SCSI:
        *vendor = QEMU_PCI_VENDOR_LSI_LOGIC;
        *product = QEMU_PCI_PRODUCT_CONTROLLER_LSI;
        break;

    case VIR_DOMAIN_CONTROLLER_TYPE_FDC:
        /* XXX we could put in the ISA bridge address, but
           that's not technically the FDC's address */
        return -1;

    case VIR_DOMAIN_CONTROLLER_TYPE_IDE:
        *vendor = QEMU_PCI_VENDOR_INTEL;
        *product = QEMU_PCI_PRODUCT_CONTROLLER_PIIX;
        break;

    default:
        VIR_INFO("Unexpected controller type %s, cannot get PCI address",
                 virDomainControllerTypeToString(def->type));
        return -1;
    }

    return 0;
}

static int
qemuProcessGetPCIVideoVendorProduct(virDomainVideoDefPtr def,
                                    unsigned *vendor,
                                    unsigned *product)
{
    switch (def->type) {
    case VIR_DOMAIN_VIDEO_TYPE_CIRRUS:
        *vendor = QEMU_PCI_VENDOR_CIRRUS;
        *product = QEMU_PCI_PRODUCT_VGA_CIRRUS;
        break;

    case VIR_DOMAIN_VIDEO_TYPE_VGA:
        *vendor = QEMU_PCI_VENDOR_QEMU;
        *product = QEMU_PCI_PRODUCT_VGA_STDVGA;
        break;

    case VIR_DOMAIN_VIDEO_TYPE_VMVGA:
        *vendor = QEMU_PCI_VENDOR_VMWARE;
        *product = QEMU_PCI_PRODUCT_VGA_VMWARE;
        break;

    default:
        return -1;
    }
    return 0;
}

static int
qemuProcessGetPCISoundVendorProduct(virDomainSoundDefPtr def,
                                    unsigned *vendor,
                                    unsigned *product)
{
    switch (def->model) {
    case VIR_DOMAIN_SOUND_MODEL_ES1370:
        *vendor = QEMU_PCI_VENDOR_ENSONIQ;
        *product = QEMU_PCI_PRODUCT_AUDIO_ES1370;
        break;

    case VIR_DOMAIN_SOUND_MODEL_AC97:
        *vendor = QEMU_PCI_VENDOR_INTEL;
        *product = QEMU_PCI_PRODUCT_AUDIO_AC97;
        break;

    default:
        return -1;
    }

    return 0;
}

static int
qemuProcessGetPCIWatchdogVendorProduct(virDomainWatchdogDefPtr def,
                                       unsigned *vendor,
                                       unsigned *product)
{
    switch (def->model) {
    case VIR_DOMAIN_WATCHDOG_MODEL_I6300ESB:
        *vendor = QEMU_PCI_VENDOR_INTEL;
        *product = QEMU_PCI_PRODUCT_WATCHDOG_I63000ESB;
        break;

    default:
        return -1;
    }

    return 0;
}


static int
qemuProcessGetPCIMemballoonVendorProduct(virDomainMemballoonDefPtr def,
                                         unsigned *vendor,
                                         unsigned *product)
{
    switch (def->model) {
    case VIR_DOMAIN_MEMBALLOON_MODEL_VIRTIO:
        *vendor = QEMU_PCI_VENDOR_REDHAT;
        *product = QEMU_PCI_PRODUCT_BALLOON_VIRTIO;
        break;

    default:
        return -1;
    }

    return 0;
}


/*
 * This entire method assumes that PCI devices in 'info pci'
 * match ordering of devices specified on the command line
 * wrt to devices of matching vendor+product
 *
 * XXXX this might not be a valid assumption if we assign
 * some static addrs on CLI. Have to check that...
 */
static int
qemuProcessDetectPCIAddresses(virDomainObjPtr vm,
                              qemuMonitorPCIAddress *addrs,
                              int naddrs)
{
    unsigned int vendor = 0, product = 0;
    int i;

    /* XXX should all these vendor/product IDs be kept in the
     * actual device data structure instead ?
     */

    for (i = 0 ; i < vm->def->ndisks ; i++) {
        if (qemuProcessGetPCIDiskVendorProduct(vm->def->disks[i], &vendor, &product) < 0)
            continue;

        if (qemuProcessAssignNextPCIAddress(&(vm->def->disks[i]->info),
                                            vendor, product,
                                            addrs, naddrs) < 0) {
            qemuReportError(VIR_ERR_INTERNAL_ERROR,
                            _("cannot find PCI address for VirtIO disk %s"),
                            vm->def->disks[i]->dst);
            return -1;
        }
    }

    for (i = 0 ; i < vm->def->nnets ; i++) {
        if (qemuProcessGetPCINetVendorProduct(vm->def->nets[i], &vendor, &product) < 0)
            continue;

        if (qemuProcessAssignNextPCIAddress(&(vm->def->nets[i]->info),
                                            vendor, product,
                                            addrs,  naddrs) < 0) {
            qemuReportError(VIR_ERR_INTERNAL_ERROR,
                            _("cannot find PCI address for %s NIC"),
                            vm->def->nets[i]->model);
            return -1;
        }
    }

    for (i = 0 ; i < vm->def->ncontrollers ; i++) {
        if (qemuProcessGetPCIControllerVendorProduct(vm->def->controllers[i], &vendor, &product) < 0)
            continue;

        if (qemuProcessAssignNextPCIAddress(&(vm->def->controllers[i]->info),
                                            vendor, product,
                                            addrs,  naddrs) < 0) {
            qemuReportError(VIR_ERR_INTERNAL_ERROR,
                            _("cannot find PCI address for controller %s"),
                            virDomainControllerTypeToString(vm->def->controllers[i]->type));
            return -1;
        }
    }

    for (i = 0 ; i < vm->def->nvideos ; i++) {
        if (qemuProcessGetPCIVideoVendorProduct(vm->def->videos[i], &vendor, &product) < 0)
            continue;

        if (qemuProcessAssignNextPCIAddress(&(vm->def->videos[i]->info),
                                            vendor, product,
                                            addrs,  naddrs) < 0) {
            qemuReportError(VIR_ERR_INTERNAL_ERROR,
                            _("cannot find PCI address for video adapter %s"),
                            virDomainVideoTypeToString(vm->def->videos[i]->type));
            return -1;
        }
    }

    for (i = 0 ; i < vm->def->nsounds ; i++) {
        if (qemuProcessGetPCISoundVendorProduct(vm->def->sounds[i], &vendor, &product) < 0)
            continue;

        if (qemuProcessAssignNextPCIAddress(&(vm->def->sounds[i]->info),
                                    vendor, product,
                                     addrs,  naddrs) < 0) {
            qemuReportError(VIR_ERR_INTERNAL_ERROR,
                            _("cannot find PCI address for sound adapter %s"),
                            virDomainSoundModelTypeToString(vm->def->sounds[i]->model));
            return -1;
        }
    }


    if (vm->def->watchdog &&
        qemuProcessGetPCIWatchdogVendorProduct(vm->def->watchdog, &vendor, &product) == 0) {
        if (qemuProcessAssignNextPCIAddress(&(vm->def->watchdog->info),
                                            vendor, product,
                                            addrs,  naddrs) < 0) {
            qemuReportError(VIR_ERR_INTERNAL_ERROR,
                            _("cannot find PCI address for watchdog %s"),
                            virDomainWatchdogModelTypeToString(vm->def->watchdog->model));
            return -1;
        }
    }

    if (vm->def->memballoon &&
        qemuProcessGetPCIMemballoonVendorProduct(vm->def->memballoon, &vendor, &product) == 0) {
        if (qemuProcessAssignNextPCIAddress(&(vm->def->memballoon->info),
                                            vendor, product,
                                            addrs, naddrs) < 0) {
            qemuReportError(VIR_ERR_INTERNAL_ERROR,
                            _("cannot find PCI address for balloon %s"),
                            virDomainMemballoonModelTypeToString(vm->def->memballoon->model));
            return -1;
        }
    }

    /* XXX console (virtio) */


    /* ... and now things we don't have in our xml */

    /* XXX USB controller ? */

    /* XXX what about other PCI devices (ie bridges) */

    return 0;
}

static int
qemuProcessInitPCIAddresses(struct qemud_driver *driver,
                            virDomainObjPtr vm)
{
    qemuDomainObjPrivatePtr priv = vm->privateData;
    int naddrs;
    int ret;
    qemuMonitorPCIAddress *addrs = NULL;

    qemuDomainObjEnterMonitorWithDriver(driver, vm);
    naddrs = qemuMonitorGetAllPCIAddresses(priv->mon,
                                           &addrs);
    qemuDomainObjExitMonitorWithDriver(driver, vm);

    ret = qemuProcessDetectPCIAddresses(vm, addrs, naddrs);

    VIR_FREE(addrs);

    return ret;
}


static int qemuProcessNextFreePort(struct qemud_driver *driver,
                                   int startPort)
{
    int i;

    for (i = startPort ; i < QEMU_VNC_PORT_MAX; i++) {
        int fd;
        int reuse = 1;
        struct sockaddr_in addr;
        bool used = false;

        if (virBitmapGetBit(driver->reservedVNCPorts,
                            i - QEMU_VNC_PORT_MIN, &used) < 0)
            VIR_DEBUG("virBitmapGetBit failed on bit %d", i - QEMU_VNC_PORT_MIN);

        if (used)
            continue;

        addr.sin_family = AF_INET;
        addr.sin_port = htons(i);
        addr.sin_addr.s_addr = htonl(INADDR_ANY);
        fd = socket(PF_INET, SOCK_STREAM, 0);
        if (fd < 0)
            return -1;

        if (setsockopt(fd, SOL_SOCKET, SO_REUSEADDR, (void*)&reuse, sizeof(reuse)) < 0) {
            VIR_FORCE_CLOSE(fd);
            break;
        }

        if (bind(fd, (struct sockaddr*)&addr, sizeof(addr)) == 0) {
            /* Not in use, lets grab it */
            VIR_FORCE_CLOSE(fd);
            /* Add port to bitmap of reserved ports */
            if (virBitmapSetBit(driver->reservedVNCPorts,
                                i - QEMU_VNC_PORT_MIN) < 0) {
                VIR_DEBUG("virBitmapSetBit failed on bit %d",
                          i - QEMU_VNC_PORT_MIN);
            }
            return i;
        }
        VIR_FORCE_CLOSE(fd);

        if (errno == EADDRINUSE) {
            /* In use, try next */
            continue;
        }
        /* Some other bad failure, get out.. */
        break;
    }
    return -1;
}


static void
qemuProcessReturnPort(struct qemud_driver *driver,
                      int port)
{
    if (port < QEMU_VNC_PORT_MIN)
        return;

    if (virBitmapClearBit(driver->reservedVNCPorts,
                          port - QEMU_VNC_PORT_MIN) < 0)
        VIR_DEBUG("Could not mark port %d as unused", port);
}


static int
qemuProcessPrepareChardevDevice(virDomainDefPtr def ATTRIBUTE_UNUSED,
                                virDomainChrDefPtr dev,
                                void *opaque ATTRIBUTE_UNUSED)
{
    int fd;
    if (dev->source.type != VIR_DOMAIN_CHR_TYPE_FILE)
        return 0;

    if ((fd = open(dev->source.data.file.path,
                   O_CREAT | O_APPEND, S_IRUSR|S_IWUSR)) < 0) {
        virReportSystemError(errno,
                             _("Unable to pre-create chardev file '%s'"),
                             dev->source.data.file.path);
        return -1;
    }

    VIR_FORCE_CLOSE(fd);

    return 0;
}


struct qemuProcessHookData {
    virConnectPtr conn;
    virDomainObjPtr vm;
    struct qemud_driver *driver;
};

static int qemuProcessHook(void *data)
{
    struct qemuProcessHookData *h = data;

    /* This must take place before exec(), so that all QEMU
     * memory allocation is on the correct NUMA node
     */
    if (qemuAddToCgroup(h->driver, h->vm->def) < 0)
        return -1;

    /* This must be done after cgroup placement to avoid resetting CPU
     * affinity */
    if (qemuProcessInitCpuAffinity(h->vm) < 0)
        return -1;

    if (virSecurityManagerSetProcessLabel(h->driver->securityManager, h->vm) < 0)
        return -1;

    return 0;
}


int
qemuProcessPrepareMonitorChr(struct qemud_driver *driver,
                             virDomainChrSourceDefPtr monConfig,
                             const char *vm)
{
    monConfig->type = VIR_DOMAIN_CHR_TYPE_UNIX;
    monConfig->data.nix.listen = true;

    if (virAsprintf(&monConfig->data.nix.path, "%s/%s.monitor",
                    driver->libDir, vm) < 0) {
        virReportOOMError();
        return -1;
    }

    return 0;
}


E
Eric Blake 已提交
1730 1731 1732
int
qemuProcessStartCPUs(struct qemud_driver *driver, virDomainObjPtr vm,
                     virConnectPtr conn)
1733 1734 1735 1736 1737 1738
{
    int ret;
    qemuDomainObjPrivatePtr priv = vm->privateData;

    qemuDomainObjEnterMonitorWithDriver(driver, vm);
    ret = qemuMonitorStartCPUs(priv->mon, conn);
E
Eric Blake 已提交
1739
    qemuDomainObjExitMonitorWithDriver(driver, vm);
1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797
    if (ret == 0) {
        vm->state = VIR_DOMAIN_RUNNING;
    }

    return ret;
}


int qemuProcessStopCPUs(struct qemud_driver *driver, virDomainObjPtr vm)
{
    int ret;
    int oldState = vm->state;
    qemuDomainObjPrivatePtr priv = vm->privateData;

    vm->state = VIR_DOMAIN_PAUSED;
    qemuDomainObjEnterMonitorWithDriver(driver, vm);
    ret = qemuMonitorStopCPUs(priv->mon);
    qemuDomainObjExitMonitorWithDriver(driver, vm);
    if (ret < 0) {
        vm->state = oldState;
    }
    return ret;
}



static int
qemuProcessFiltersInstantiate(virConnectPtr conn,
                              virDomainDefPtr def)
{
    int err = 0;
    int i;

    if (!conn)
        return 1;

    for (i = 0 ; i < def->nnets ; i++) {
        virDomainNetDefPtr net = def->nets[i];
        if ((net->filter) && (net->ifname)) {
           if (virDomainConfNWFilterInstantiate(conn, net)) {
                err = 1;
                break;
            }
        }
    }

    return err;
}

struct qemuProcessReconnectData {
    virConnectPtr conn;
    struct qemud_driver *driver;
};
/*
 * Open an existing VM's monitor, re-detect VCPU threads
 * and re-reserve the security labels in use
 */
static void
1798
qemuProcessReconnect(void *payload, const void *name ATTRIBUTE_UNUSED, void *opaque)
1799 1800 1801 1802 1803
{
    virDomainObjPtr obj = payload;
    struct qemuProcessReconnectData *data = opaque;
    struct qemud_driver *driver = data->driver;
    qemuDomainObjPrivatePtr priv;
1804
    virBitmapPtr qemuCaps = NULL;
1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829
    virConnectPtr conn = data->conn;

    virDomainObjLock(obj);

    VIR_DEBUG("Reconnect monitor to %p '%s'", obj, obj->def->name);

    priv = obj->privateData;

    /* Hold an extra reference because we can't allow 'vm' to be
     * deleted if qemuConnectMonitor() failed */
    virDomainObjRef(obj);

    /* XXX check PID liveliness & EXE path */
    if (qemuConnectMonitor(driver, obj) < 0)
        goto error;

    if (qemuUpdateActivePciHostdevs(driver, obj->def) < 0) {
        goto error;
    }

    /* XXX we should be persisting the original flags in the XML
     * not re-detecting them, since the binary may have changed
     * since launch time */
    if (qemuCapsExtractVersionInfo(obj->def->emulator, obj->def->os.arch,
                                   NULL,
1830 1831
                                   &qemuCaps) >= 0 &&
        qemuCapsGet(qemuCaps, QEMU_CAPS_DEVICE)) {
1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849
        priv->persistentAddrs = 1;

        if (!(priv->pciaddrs = qemuDomainPCIAddressSetCreate(obj->def)) ||
            qemuAssignDevicePCISlots(obj->def, priv->pciaddrs) < 0)
            goto error;
    }

    if (virSecurityManagerReserveLabel(driver->securityManager, obj) < 0)
        goto error;

    if (qemuProcessFiltersInstantiate(conn, obj->def))
        goto error;

    if (obj->def->id >= driver->nextvmid)
        driver->nextvmid = obj->def->id + 1;

    if (virDomainObjUnref(obj) > 0)
        virDomainObjUnlock(obj);
1850 1851

    qemuCapsFree(qemuCaps);
1852 1853 1854
    return;

error:
1855
    qemuCapsFree(qemuCaps);
1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896
    if (!virDomainObjIsActive(obj)) {
        if (virDomainObjUnref(obj) > 0)
            virDomainObjUnlock(obj);
        return;
    }

    if (virDomainObjUnref(obj) > 0) {
        /* We can't get the monitor back, so must kill the VM
         * to remove danger of it ending up running twice if
         * user tries to start it again later */
        qemuProcessStop(driver, obj, 0);
        if (!obj->persistent)
            virDomainRemoveInactive(&driver->domains, obj);
        else
            virDomainObjUnlock(obj);
    }
}

/**
 * qemuProcessReconnectAll
 *
 * Try to re-open the resources for live VMs that we care
 * about.
 */
void
qemuProcessReconnectAll(virConnectPtr conn, struct qemud_driver *driver)
{
    struct qemuProcessReconnectData data = {conn, driver};
    virHashForEach(driver->domains.objs, qemuProcessReconnect, &data);
}

int qemuProcessStart(virConnectPtr conn,
                     struct qemud_driver *driver,
                     virDomainObjPtr vm,
                     const char *migrateFrom,
                     bool start_paused,
                     int stdin_fd,
                     const char *stdin_path,
                     enum virVMOperationType vmop)
{
    int ret;
1897
    virBitmapPtr qemuCaps = NULL;
1898 1899 1900 1901 1902 1903 1904 1905
    off_t pos = -1;
    char ebuf[1024];
    char *pidfile = NULL;
    int logfile = -1;
    char *timestamp;
    qemuDomainObjPrivatePtr priv = vm->privateData;
    virCommandPtr cmd = NULL;
    struct qemuProcessHookData hookData;
E
Eric Blake 已提交
1906
    unsigned long cur_balloon;
1907 1908 1909 1910 1911

    hookData.conn = conn;
    hookData.vm = vm;
    hookData.driver = driver;

1912
    VIR_DEBUG0("Beginning VM startup process");
1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923

    if (virDomainObjIsActive(vm)) {
        qemuReportError(VIR_ERR_OPERATION_INVALID,
                        "%s", _("VM is already active"));
        return -1;
    }

    /* Do this upfront, so any part of the startup process can add
     * runtime state to vm->def that won't be persisted. This let's us
     * report implicit runtime defaults in the XML, like vnc listen/socket
     */
1924
    VIR_DEBUG0("Setting current domain def as transient");
1925 1926 1927 1928 1929
    if (virDomainObjSetDefTransient(driver->caps, vm, true) < 0)
        goto cleanup;

    vm->def->id = driver->nextvmid++;

1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945
    /* Run a early hook to set-up missing devices */
    if (virHookPresent(VIR_HOOK_DRIVER_QEMU)) {
        char *xml = virDomainDefFormat(vm->def, 0);
        int hookret;

        hookret = virHookCall(VIR_HOOK_DRIVER_QEMU, vm->def->name,
                    VIR_HOOK_QEMU_OP_PREPARE, VIR_HOOK_SUBOP_BEGIN, NULL, xml);
        VIR_FREE(xml);

        /*
         * If the script raised an error abort the launch
         */
        if (hookret < 0)
            goto cleanup;
    }

1946
    /* Must be run before security labelling */
1947
    VIR_DEBUG0("Preparing host devices");
1948 1949 1950
    if (qemuPrepareHostDevices(driver, vm->def) < 0)
        goto cleanup;

1951
    VIR_DEBUG0("Preparing chr devices");
1952 1953 1954 1955 1956 1957 1958 1959
    if (virDomainChrDefForeach(vm->def,
                               true,
                               qemuProcessPrepareChardevDevice,
                               NULL) < 0)
        goto cleanup;

    /* If you are using a SecurityDriver with dynamic labelling,
       then generate a security label for isolation */
1960
    VIR_DEBUG0("Generating domain security label (if required)");
1961
    if (virSecurityManagerGenLabel(driver->securityManager, vm) < 0) {
1962
        qemuAuditSecurityLabel(vm, false);
1963 1964
        goto cleanup;
    }
1965
    qemuAuditSecurityLabel(vm, true);
1966

1967
    VIR_DEBUG0("Generating setting domain security labels (if required)");
1968 1969 1970 1971 1972 1973 1974 1975 1976 1977
    if (virSecurityManagerSetAllLabel(driver->securityManager,
                                      vm, stdin_path) < 0)
        goto cleanup;

    if (stdin_fd != -1) {
        /* if there's an fd to migrate from, and it's a pipe, put the
         * proper security label on it
         */
        struct stat stdin_sb;

1978
        VIR_DEBUG0("setting security label on pipe used for migration");
1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991

        if (fstat(stdin_fd, &stdin_sb) < 0) {
            virReportSystemError(errno,
                                 _("cannot stat fd %d"), stdin_fd);
            goto cleanup;
        }
        if (S_ISFIFO(stdin_sb.st_mode) &&
            virSecurityManagerSetFDLabel(driver->securityManager, vm, stdin_fd) < 0)
            goto cleanup;
    }

    /* Ensure no historical cgroup for this VM is lying around bogus
     * settings */
1992
    VIR_DEBUG0("Ensuring no historical cgroup is lying around");
1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037
    qemuRemoveCgroup(driver, vm, 1);

    if (vm->def->ngraphics == 1) {
        if (vm->def->graphics[0]->type == VIR_DOMAIN_GRAPHICS_TYPE_VNC &&
            !vm->def->graphics[0]->data.vnc.socket &&
            vm->def->graphics[0]->data.vnc.autoport) {
            int port = qemuProcessNextFreePort(driver, QEMU_VNC_PORT_MIN);
            if (port < 0) {
                qemuReportError(VIR_ERR_INTERNAL_ERROR,
                                "%s", _("Unable to find an unused VNC port"));
                goto cleanup;
            }
            vm->def->graphics[0]->data.vnc.port = port;
        } else if (vm->def->graphics[0]->type == VIR_DOMAIN_GRAPHICS_TYPE_SPICE &&
                   vm->def->graphics[0]->data.spice.autoport) {
            int port = qemuProcessNextFreePort(driver, QEMU_VNC_PORT_MIN);
            int tlsPort = -1;
            if (port < 0) {
                qemuReportError(VIR_ERR_INTERNAL_ERROR,
                                "%s", _("Unable to find an unused SPICE port"));
                goto cleanup;
            }

            if (driver->spiceTLS) {
                tlsPort = qemuProcessNextFreePort(driver, port + 1);
                if (tlsPort < 0) {
                    qemuReportError(VIR_ERR_INTERNAL_ERROR,
                                    "%s", _("Unable to find an unused SPICE TLS port"));
                    qemuProcessReturnPort(driver, port);
                    goto cleanup;
                }
            }

            vm->def->graphics[0]->data.spice.port = port;
            vm->def->graphics[0]->data.spice.tlsPort = tlsPort;
        }
    }

    if (virFileMakePath(driver->logDir) != 0) {
        virReportSystemError(errno,
                             _("cannot create log directory %s"),
                             driver->logDir);
        goto cleanup;
    }

2038
    VIR_DEBUG0("Creating domain log file");
2039 2040 2041
    if ((logfile = qemuProcessLogFD(driver, vm->def->name, false)) < 0)
        goto cleanup;

2042
    VIR_DEBUG0("Determining emulator version");
2043 2044
    if (qemuCapsExtractVersionInfo(vm->def->emulator, vm->def->os.arch,
                                   NULL,
2045
                                   &qemuCaps) < 0)
2046 2047
        goto cleanup;

2048
    VIR_DEBUG0("Setting up domain cgroup (if required)");
2049 2050 2051 2052 2053 2054 2055 2056
    if (qemuSetupCgroup(driver, vm) < 0)
        goto cleanup;

    if (VIR_ALLOC(priv->monConfig) < 0) {
        virReportOOMError();
        goto cleanup;
    }

2057
    VIR_DEBUG0("Preparing monitor state");
2058 2059 2060 2061
    if (qemuProcessPrepareMonitorChr(driver, priv->monConfig, vm->def->name) < 0)
        goto cleanup;

#if HAVE_YAJL
2062
    if (qemuCapsGet(qemuCaps, QEMU_CAPS_MONITOR_JSON))
2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090
        priv->monJSON = 1;
    else
#endif
        priv->monJSON = 0;

    priv->monitor_warned = 0;
    priv->gotShutdown = false;

    if ((ret = virFileDeletePid(driver->stateDir, vm->def->name)) != 0) {
        virReportSystemError(ret,
                             _("Cannot remove stale PID file for %s"),
                             vm->def->name);
        goto cleanup;
    }

    if (!(pidfile = virFilePid(driver->stateDir, vm->def->name))) {
        virReportSystemError(errno,
                             "%s", _("Failed to build pidfile path."));
        goto cleanup;
    }

    /*
     * Normally PCI addresses are assigned in the virDomainCreate
     * or virDomainDefine methods. We might still need to assign
     * some here to cope with the question of upgrades. Regardless
     * we also need to populate the PCi address set cache for later
     * use in hotplug
     */
2091
    if (qemuCapsGet(qemuCaps, QEMU_CAPS_DEVICE)) {
2092
        VIR_DEBUG0("Assigning domain PCI addresses");
2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110
        /* Populate cache with current addresses */
        if (priv->pciaddrs) {
            qemuDomainPCIAddressSetFree(priv->pciaddrs);
            priv->pciaddrs = NULL;
        }
        if (!(priv->pciaddrs = qemuDomainPCIAddressSetCreate(vm->def)))
            goto cleanup;


        /* Assign any remaining addresses */
        if (qemuAssignDevicePCISlots(vm->def, priv->pciaddrs) < 0)
            goto cleanup;

        priv->persistentAddrs = 1;
    } else {
        priv->persistentAddrs = 0;
    }

2111
    VIR_DEBUG0("Building emulator command line");
2112
    if (!(cmd = qemuBuildCommandLine(conn, driver, vm->def, priv->monConfig,
2113
                                     priv->monJSON != 0, qemuCaps,
2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210
                                     migrateFrom, stdin_fd,
                                     vm->current_snapshot, vmop)))
        goto cleanup;

#if 0
    /* XXX */
    if (qemuDomainSnapshotSetCurrentInactive(vm, driver->snapshotDir) < 0)
        goto cleanup;
#endif

    /* now that we know it is about to start call the hook if present */
    if (virHookPresent(VIR_HOOK_DRIVER_QEMU)) {
        char *xml = virDomainDefFormat(vm->def, 0);
        int hookret;

        hookret = virHookCall(VIR_HOOK_DRIVER_QEMU, vm->def->name,
                    VIR_HOOK_QEMU_OP_START, VIR_HOOK_SUBOP_BEGIN, NULL, xml);
        VIR_FREE(xml);

        /*
         * If the script raised an error abort the launch
         */
        if (hookret < 0)
            goto cleanup;
    }

    if ((timestamp = virTimestamp()) == NULL) {
        virReportOOMError();
        goto cleanup;
    } else {
        if (safewrite(logfile, timestamp, strlen(timestamp)) < 0 ||
            safewrite(logfile, START_POSTFIX, strlen(START_POSTFIX)) < 0) {
            VIR_WARN("Unable to write timestamp to logfile: %s",
                     virStrerror(errno, ebuf, sizeof ebuf));
        }

        VIR_FREE(timestamp);
    }

    virCommandWriteArgLog(cmd, logfile);

    if ((pos = lseek(logfile, 0, SEEK_END)) < 0)
        VIR_WARN("Unable to seek to end of logfile: %s",
                 virStrerror(errno, ebuf, sizeof ebuf));

    VIR_DEBUG("Clear emulator capabilities: %d",
              driver->clearEmulatorCapabilities);
    if (driver->clearEmulatorCapabilities)
        virCommandClearCaps(cmd);

    virCommandSetPreExecHook(cmd, qemuProcessHook, &hookData);

    virCommandSetOutputFD(cmd, &logfile);
    virCommandSetErrorFD(cmd, &logfile);
    virCommandNonblockingFDs(cmd);
    virCommandSetPidFile(cmd, pidfile);
    virCommandDaemonize(cmd);

    ret = virCommandRun(cmd, NULL);
    VIR_FREE(pidfile);

    /* wait for qemu process to to show up */
    if (ret == 0) {
        if (virFileReadPid(driver->stateDir, vm->def->name, &vm->pid)) {
            qemuReportError(VIR_ERR_INTERNAL_ERROR,
                            _("Domain %s didn't show up"), vm->def->name);
            ret = -1;
        }
#if 0
    } else if (ret == -2) {
        /*
         * XXX this is bogus. It isn't safe to set vm->pid = child
         * because the child no longer exists.
         */

        /* The virExec process that launches the daemon failed. Pending on
         * when it failed (we can't determine for sure), there may be
         * extra info in the domain log (if the hook failed for example).
         *
         * Pretend like things succeeded, and let 'WaitForMonitor' report
         * the log contents for us.
         */
        vm->pid = child;
        ret = 0;
#endif
    }

    if (migrateFrom)
        start_paused = true;
    vm->state = start_paused ? VIR_DOMAIN_PAUSED : VIR_DOMAIN_RUNNING;

    if (ret == -1) /* The VM failed to start; tear filters before taps */
        virDomainConfVMNWFilterTeardown(vm);

    if (ret == -1) /* The VM failed to start */
        goto cleanup;

2211
    VIR_DEBUG0("Waiting for monitor to show up");
2212 2213 2214
    if (qemuProcessWaitForMonitor(driver, vm, pos) < 0)
        goto cleanup;

2215
    VIR_DEBUG0("Detecting VCPU PIDs");
2216 2217 2218
    if (qemuProcessDetectVcpuPIDs(driver, vm) < 0)
        goto cleanup;

2219
    VIR_DEBUG0("Setting any required VM passwords");
2220
    if (qemuProcessInitPasswords(conn, driver, vm, qemuCaps) < 0)
2221 2222 2223 2224
        goto cleanup;

    /* If we have -device, then addresses are assigned explicitly.
     * If not, then we have to detect dynamic ones here */
2225
    if (!qemuCapsGet(qemuCaps, QEMU_CAPS_DEVICE)) {
2226
        VIR_DEBUG0("Determining domain device PCI addresses");
2227 2228 2229 2230
        if (qemuProcessInitPCIAddresses(driver, vm) < 0)
            goto cleanup;
    }

2231
    VIR_DEBUG0("Setting initial memory amount");
E
Eric Blake 已提交
2232
    cur_balloon = vm->def->mem.cur_balloon;
2233
    qemuDomainObjEnterMonitorWithDriver(driver, vm);
E
Eric Blake 已提交
2234
    if (qemuMonitorSetBalloon(priv->mon, cur_balloon) < 0) {
2235 2236 2237 2238 2239 2240
        qemuDomainObjExitMonitorWithDriver(driver, vm);
        goto cleanup;
    }
    qemuDomainObjExitMonitorWithDriver(driver, vm);

    if (!start_paused) {
2241
        VIR_DEBUG0("Starting domain CPUs");
2242 2243 2244 2245 2246 2247 2248 2249 2250 2251
        /* Allow the CPUS to start executing */
        if (qemuProcessStartCPUs(driver, vm, conn) < 0) {
            if (virGetLastError() == NULL)
                qemuReportError(VIR_ERR_INTERNAL_ERROR,
                                "%s", _("resume operation failed"));
            goto cleanup;
        }
    }


2252
    VIR_DEBUG0("Writing domain status to disk");
2253 2254 2255
    if (virDomainSaveStatus(driver->caps, driver->stateDir, vm) < 0)
        goto cleanup;

2256
    qemuCapsFree(qemuCaps);
2257 2258 2259 2260 2261 2262 2263 2264 2265
    virCommandFree(cmd);
    VIR_FORCE_CLOSE(logfile);

    return 0;

cleanup:
    /* We jump here if we failed to start the VM for any reason, or
     * if we failed to initialize the now running VM. kill it off and
     * pretend we never started it */
2266
    qemuCapsFree(qemuCaps);
2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437
    virCommandFree(cmd);
    VIR_FORCE_CLOSE(logfile);
    qemuProcessStop(driver, vm, 0);

    return -1;
}


void qemuProcessStop(struct qemud_driver *driver,
                     virDomainObjPtr vm,
                     int migrated)
{
    int ret;
    int retries = 0;
    qemuDomainObjPrivatePtr priv = vm->privateData;
    virErrorPtr orig_err;
    virDomainDefPtr def;
    int i;
    int logfile = -1;
    char *timestamp;
    char ebuf[1024];

    VIR_DEBUG("Shutting down VM '%s' pid=%d migrated=%d",
              vm->def->name, vm->pid, migrated);

    if (!virDomainObjIsActive(vm)) {
        VIR_DEBUG("VM '%s' not active", vm->def->name);
        return;
    }

    if ((logfile = qemuProcessLogFD(driver, vm->def->name, true)) < 0) {
        /* To not break the normal domain shutdown process, skip the
         * timestamp log writing if failed on opening log file. */
        VIR_WARN("Unable to open logfile: %s",
                  virStrerror(errno, ebuf, sizeof ebuf));
    } else {
        if ((timestamp = virTimestamp()) == NULL) {
            virReportOOMError();
        } else {
            if (safewrite(logfile, timestamp, strlen(timestamp)) < 0 ||
                safewrite(logfile, SHUTDOWN_POSTFIX,
                          strlen(SHUTDOWN_POSTFIX)) < 0) {
                VIR_WARN("Unable to write timestamp to logfile: %s",
                         virStrerror(errno, ebuf, sizeof ebuf));
            }

            VIR_FREE(timestamp);
        }

        if (VIR_CLOSE(logfile) < 0)
             VIR_WARN("Unable to close logfile: %s",
                      virStrerror(errno, ebuf, sizeof ebuf));
    }

    /* This method is routinely used in clean up paths. Disable error
     * reporting so we don't squash a legit error. */
    orig_err = virSaveLastError();

    virDomainConfVMNWFilterTeardown(vm);

    if (driver->macFilter) {
        def = vm->def;
        for (i = 0 ; i < def->nnets ; i++) {
            virDomainNetDefPtr net = def->nets[i];
            if (net->ifname == NULL)
                continue;
            if ((errno = networkDisallowMacOnPort(driver, net->ifname,
                                                  net->mac))) {
                virReportSystemError(errno,
             _("failed to remove ebtables rule to allow MAC address on  '%s'"),
                                     net->ifname);
            }
        }
    }

    /* This will safely handle a non-running guest with pid=0 or pid=-1*/
    if (virKillProcess(vm->pid, 0) == 0 &&
        virKillProcess(vm->pid, SIGTERM) < 0)
        virReportSystemError(errno,
                             _("Failed to send SIGTERM to %s (%d)"),
                             vm->def->name, vm->pid);

    if (priv->mon)
        qemuMonitorClose(priv->mon);

    if (priv->monConfig) {
        if (priv->monConfig->type == VIR_DOMAIN_CHR_TYPE_UNIX)
            unlink(priv->monConfig->data.nix.path);
        virDomainChrSourceDefFree(priv->monConfig);
        priv->monConfig = NULL;
    }

    /* shut it off for sure */
    virKillProcess(vm->pid, SIGKILL);

    /* now that we know it's stopped call the hook if present */
    if (virHookPresent(VIR_HOOK_DRIVER_QEMU)) {
        char *xml = virDomainDefFormat(vm->def, 0);

        /* we can't stop the operation even if the script raised an error */
        virHookCall(VIR_HOOK_DRIVER_QEMU, vm->def->name,
                    VIR_HOOK_QEMU_OP_STOPPED, VIR_HOOK_SUBOP_END, NULL, xml);
        VIR_FREE(xml);
    }

    /* Reset Security Labels */
    virSecurityManagerRestoreAllLabel(driver->securityManager,
                                      vm, migrated);
    virSecurityManagerReleaseLabel(driver->securityManager, vm);

    /* Clear out dynamically assigned labels */
    if (vm->def->seclabel.type == VIR_DOMAIN_SECLABEL_DYNAMIC) {
        VIR_FREE(vm->def->seclabel.model);
        VIR_FREE(vm->def->seclabel.label);
        VIR_FREE(vm->def->seclabel.imagelabel);
    }

    virDomainDefClearDeviceAliases(vm->def);
    if (!priv->persistentAddrs) {
        virDomainDefClearPCIAddresses(vm->def);
        qemuDomainPCIAddressSetFree(priv->pciaddrs);
        priv->pciaddrs = NULL;
    }

    qemuDomainReAttachHostDevices(driver, vm->def);

#if WITH_MACVTAP
    def = vm->def;
    for (i = 0; i < def->nnets; i++) {
        virDomainNetDefPtr net = def->nets[i];
        if (net->type == VIR_DOMAIN_NET_TYPE_DIRECT) {
            delMacvtap(net->ifname, net->mac, net->data.direct.linkdev,
                       &net->data.direct.virtPortProfile);
            VIR_FREE(net->ifname);
        }
    }
#endif

retry:
    if ((ret = qemuRemoveCgroup(driver, vm, 0)) < 0) {
        if (ret == -EBUSY && (retries++ < 5)) {
            usleep(200*1000);
            goto retry;
        }
        VIR_WARN("Failed to remove cgroup for %s",
                 vm->def->name);
    }

    qemuProcessRemoveDomainStatus(driver, vm);

    /* Remove VNC port from port reservation bitmap, but only if it was
       reserved by the driver (autoport=yes)
    */
    if ((vm->def->ngraphics == 1) &&
        vm->def->graphics[0]->type == VIR_DOMAIN_GRAPHICS_TYPE_VNC &&
        vm->def->graphics[0]->data.vnc.autoport) {
        qemuProcessReturnPort(driver, vm->def->graphics[0]->data.vnc.port);
    }
    if ((vm->def->ngraphics == 1) &&
        vm->def->graphics[0]->type == VIR_DOMAIN_GRAPHICS_TYPE_SPICE &&
        vm->def->graphics[0]->data.spice.autoport) {
        qemuProcessReturnPort(driver, vm->def->graphics[0]->data.spice.port);
        qemuProcessReturnPort(driver, vm->def->graphics[0]->data.spice.tlsPort);
    }

    vm->pid = -1;
    vm->def->id = -1;
    vm->state = VIR_DOMAIN_SHUTOFF;
    VIR_FREE(priv->vcpupids);
    priv->nvcpupids = 0;

2438 2439 2440 2441 2442 2443 2444 2445 2446 2447
    /* The "release" hook cleans up additional ressources */
    if (virHookPresent(VIR_HOOK_DRIVER_QEMU)) {
        char *xml = virDomainDefFormat(vm->def, 0);

        /* we can't stop the operation even if the script raised an error */
        virHookCall(VIR_HOOK_DRIVER_QEMU, vm->def->name,
                    VIR_HOOK_QEMU_OP_RELEASE, VIR_HOOK_SUBOP_END, NULL, xml);
        VIR_FREE(xml);
    }

2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459
    if (vm->newDef) {
        virDomainDefFree(vm->def);
        vm->def = vm->newDef;
        vm->def->id = -1;
        vm->newDef = NULL;
    }

    if (orig_err) {
        virSetError(orig_err);
        virFreeError(orig_err);
    }
}