qemu_migration.c 93.8 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
/*
 * qemu_migration.c: QEMU migration handling
 *
 * 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 <sys/time.h>
25 26
#include <gnutls/gnutls.h>
#include <gnutls/x509.h>
27 28 29 30 31 32

#include "qemu_migration.h"
#include "qemu_monitor.h"
#include "qemu_domain.h"
#include "qemu_process.h"
#include "qemu_capabilities.h"
33
#include "qemu_cgroup.h"
34

35
#include "domain_audit.h"
36 37 38 39
#include "logging.h"
#include "virterror_internal.h"
#include "memory.h"
#include "util.h"
E
Eric Blake 已提交
40
#include "virfile.h"
41 42
#include "datatypes.h"
#include "fdstream.h"
43
#include "uuid.h"
44
#include "locking/domain_lock.h"
45

46 47 48

#define VIR_FROM_THIS VIR_FROM_QEMU

49 50 51 52 53 54 55 56 57 58 59 60 61
VIR_ENUM_IMPL(qemuMigrationJobPhase, QEMU_MIGRATION_PHASE_LAST,
              "none",
              "perform2",
              "begin3",
              "perform3",
              "perform3_done",
              "confirm3_cancelled",
              "confirm3",
              "prepare",
              "finish2",
              "finish3",
);

62
enum qemuMigrationCookieFlags {
63
    QEMU_MIGRATION_COOKIE_FLAG_GRAPHICS,
64
    QEMU_MIGRATION_COOKIE_FLAG_LOCKSTATE,
65 66 67 68 69 70 71

    QEMU_MIGRATION_COOKIE_FLAG_LAST
};

VIR_ENUM_DECL(qemuMigrationCookieFlag);
VIR_ENUM_IMPL(qemuMigrationCookieFlag,
              QEMU_MIGRATION_COOKIE_FLAG_LAST,
72
              "graphics", "lockstate");
73 74 75

enum qemuMigrationCookieFeatures {
    QEMU_MIGRATION_COOKIE_GRAPHICS  = (1 << QEMU_MIGRATION_COOKIE_FLAG_GRAPHICS),
76
    QEMU_MIGRATION_COOKIE_LOCKSTATE = (1 << QEMU_MIGRATION_COOKIE_FLAG_LOCKSTATE),
77 78 79 80 81 82 83 84 85 86 87 88
};

typedef struct _qemuMigrationCookieGraphics qemuMigrationCookieGraphics;
typedef qemuMigrationCookieGraphics *qemuMigrationCookieGraphicsPtr;
struct _qemuMigrationCookieGraphics {
    int type;
    int port;
    int tlsPort;
    char *listen;
    char *tlsSubject;
};

89 90 91
typedef struct _qemuMigrationCookie qemuMigrationCookie;
typedef qemuMigrationCookie *qemuMigrationCookiePtr;
struct _qemuMigrationCookie {
E
Eric Blake 已提交
92 93
    unsigned int flags;
    unsigned int flagsMandatory;
94 95

    /* Host properties */
96 97 98 99
    unsigned char localHostuuid[VIR_UUID_BUFLEN];
    unsigned char remoteHostuuid[VIR_UUID_BUFLEN];
    char *localHostname;
    char *remoteHostname;
100 101 102 103

    /* Guest properties */
    unsigned char uuid[VIR_UUID_BUFLEN];
    char *name;
104

105 106 107 108
    /* If (flags & QEMU_MIGRATION_COOKIE_LOCKSTATE) */
    char *lockState;
    char *lockDriver;

109 110
    /* If (flags & QEMU_MIGRATION_COOKIE_GRAPHICS) */
    qemuMigrationCookieGraphicsPtr graphics;
111 112
};

113 114 115 116 117 118 119 120 121
static void qemuMigrationCookieGraphicsFree(qemuMigrationCookieGraphicsPtr grap)
{
    if (!grap)
        return;
    VIR_FREE(grap->listen);
    VIR_FREE(grap->tlsSubject);
    VIR_FREE(grap);
}

122 123 124 125 126 127

static void qemuMigrationCookieFree(qemuMigrationCookiePtr mig)
{
    if (!mig)
        return;

128 129 130
    if (mig->flags & QEMU_MIGRATION_COOKIE_GRAPHICS)
        qemuMigrationCookieGraphicsFree(mig->graphics);

131 132
    VIR_FREE(mig->localHostname);
    VIR_FREE(mig->remoteHostname);
133
    VIR_FREE(mig->name);
134 135
    VIR_FREE(mig->lockState);
    VIR_FREE(mig->lockDriver);
136 137 138 139
    VIR_FREE(mig);
}


140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212
static char *
qemuDomainExtractTLSSubject(const char *certdir)
{
    char *certfile = NULL;
    char *subject = NULL;
    char *pemdata = NULL;
    gnutls_datum_t pemdatum;
    gnutls_x509_crt_t cert;
    int ret;
    size_t subjectlen;

    if (virAsprintf(&certfile, "%s/server-cert.pem", certdir) < 0)
        goto no_memory;

    if (virFileReadAll(certfile, 8192, &pemdata) < 0) {
        qemuReportError(VIR_ERR_INTERNAL_ERROR,
                        _("unable to read server cert %s"), certfile);
        goto error;
    }

    ret = gnutls_x509_crt_init(&cert);
    if (ret < 0) {
        qemuReportError(VIR_ERR_INTERNAL_ERROR,
                        _("cannot initialize cert object: %s"),
                        gnutls_strerror(ret));
        goto error;
    }

    pemdatum.data = (unsigned char *)pemdata;
    pemdatum.size = strlen(pemdata);

    ret = gnutls_x509_crt_import(cert, &pemdatum, GNUTLS_X509_FMT_PEM);
    if (ret < 0) {
        qemuReportError(VIR_ERR_INTERNAL_ERROR,
                        _("cannot load cert data from %s: %s"),
                        certfile, gnutls_strerror(ret));
        goto error;
    }

    subjectlen = 1024;
    if (VIR_ALLOC_N(subject, subjectlen+1) < 0)
        goto no_memory;

    gnutls_x509_crt_get_dn(cert, subject, &subjectlen);
    subject[subjectlen] = '\0';

    VIR_FREE(certfile);
    VIR_FREE(pemdata);

    return subject;

no_memory:
    virReportOOMError();
error:
    VIR_FREE(certfile);
    VIR_FREE(pemdata);
    return NULL;
}


static qemuMigrationCookieGraphicsPtr
qemuMigrationCookieGraphicsAlloc(struct qemud_driver *driver,
                                 virDomainGraphicsDefPtr def)
{
    qemuMigrationCookieGraphicsPtr mig = NULL;
    const char *listenAddr;

    if (VIR_ALLOC(mig) < 0)
        goto no_memory;

    mig->type = def->type;
    if (mig->type == VIR_DOMAIN_GRAPHICS_TYPE_VNC) {
        mig->port = def->data.vnc.port;
213
        listenAddr = virDomainGraphicsListenGetAddress(def, 0);
214 215 216 217 218 219 220 221 222 223 224 225
        if (!listenAddr)
            listenAddr = driver->vncListen;

        if (driver->vncTLS &&
            !(mig->tlsSubject = qemuDomainExtractTLSSubject(driver->vncTLSx509certdir)))
            goto error;
    } else {
        mig->port = def->data.spice.port;
        if (driver->spiceTLS)
            mig->tlsPort = def->data.spice.tlsPort;
        else
            mig->tlsPort = -1;
226
        listenAddr = virDomainGraphicsListenGetAddress(def, 0);
227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246
        if (!listenAddr)
            listenAddr = driver->spiceListen;

        if (driver->spiceTLS &&
            !(mig->tlsSubject = qemuDomainExtractTLSSubject(driver->spiceTLSx509certdir)))
            goto error;
    }
    if (!(mig->listen = strdup(listenAddr)))
        goto no_memory;

    return mig;

no_memory:
    virReportOOMError();
error:
    qemuMigrationCookieGraphicsFree(mig);
    return NULL;
}


247 248 249 250 251 252 253 254 255 256 257 258
static qemuMigrationCookiePtr
qemuMigrationCookieNew(virDomainObjPtr dom)
{
    qemuMigrationCookiePtr mig = NULL;

    if (VIR_ALLOC(mig) < 0)
        goto no_memory;

    if (!(mig->name = strdup(dom->def->name)))
        goto no_memory;
    memcpy(mig->uuid, dom->def->uuid, VIR_UUID_BUFLEN);

259
    if (!(mig->localHostname = virGetHostname(NULL)))
260
        goto error;
261
    if (virGetHostUUID(mig->localHostuuid) < 0) {
262 263 264 265 266 267 268 269 270 271 272 273 274 275 276
        qemuReportError(VIR_ERR_INTERNAL_ERROR, "%s",
                        _("Unable to obtain host UUID"));
        goto error;
    }

    return mig;

no_memory:
    virReportOOMError();
error:
    qemuMigrationCookieFree(mig);
    return NULL;
}


277 278 279 280 281 282 283 284 285 286 287 288 289
static int
qemuMigrationCookieAddGraphics(qemuMigrationCookiePtr mig,
                               struct qemud_driver *driver,
                               virDomainObjPtr dom)
{
    if (mig->flags & QEMU_MIGRATION_COOKIE_GRAPHICS) {
        qemuReportError(VIR_ERR_INTERNAL_ERROR, "%s",
                        _("Migration graphics data already present"));
        return -1;
    }

    if (dom->def->ngraphics == 1 &&
        (dom->def->graphics[0]->type == VIR_DOMAIN_GRAPHICS_TYPE_VNC ||
290 291 292 293 294 295
         dom->def->graphics[0]->type == VIR_DOMAIN_GRAPHICS_TYPE_SPICE)) {
        if (!(mig->graphics =
              qemuMigrationCookieGraphicsAlloc(driver, dom->def->graphics[0])))
            return -1;
        mig->flags |= QEMU_MIGRATION_COOKIE_GRAPHICS;
    }
296 297 298 299 300

    return 0;
}


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
static int
qemuMigrationCookieAddLockstate(qemuMigrationCookiePtr mig,
                                struct qemud_driver *driver,
                                virDomainObjPtr dom)
{
    qemuDomainObjPrivatePtr priv = dom->privateData;

    if (mig->flags & QEMU_MIGRATION_COOKIE_LOCKSTATE) {
        qemuReportError(VIR_ERR_INTERNAL_ERROR, "%s",
                        _("Migration lockstate data already present"));
        return -1;
    }

    if (virDomainObjGetState(dom, NULL) == VIR_DOMAIN_PAUSED) {
        if (priv->lockState &&
            !(mig->lockState = strdup(priv->lockState)))
            return -1;
    } else {
        if (virDomainLockProcessInquire(driver->lockManager, dom, &mig->lockState) < 0)
            return -1;
    }

    if (!(mig->lockDriver = strdup(virLockManagerPluginGetName(driver->lockManager)))) {
        VIR_FREE(mig->lockState);
        return -1;
    }

    mig->flags |= QEMU_MIGRATION_COOKIE_LOCKSTATE;
    mig->flagsMandatory |= QEMU_MIGRATION_COOKIE_LOCKSTATE;

    return 0;
}



336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353
static void qemuMigrationCookieGraphicsXMLFormat(virBufferPtr buf,
                                                 qemuMigrationCookieGraphicsPtr grap)
{
    virBufferAsprintf(buf, "  <graphics type='%s' port='%d' listen='%s'",
                      virDomainGraphicsTypeToString(grap->type),
                      grap->port, grap->listen);
    if (grap->type == VIR_DOMAIN_GRAPHICS_TYPE_SPICE)
        virBufferAsprintf(buf, " tlsPort='%d'", grap->tlsPort);
    if (grap->tlsSubject) {
        virBufferAddLit(buf, ">\n");
        virBufferEscapeString(buf, "    <cert info='subject' value='%s'/>\n", grap->tlsSubject);
        virBufferAddLit(buf, "  </graphics>\n");
    } else {
        virBufferAddLit(buf, "/>\n");
    }
}


354 355 356 357 358
static void qemuMigrationCookieXMLFormat(virBufferPtr buf,
                                         qemuMigrationCookiePtr mig)
{
    char uuidstr[VIR_UUID_STRING_BUFLEN];
    char hostuuidstr[VIR_UUID_STRING_BUFLEN];
359
    int i;
360 361

    virUUIDFormat(mig->uuid, uuidstr);
362
    virUUIDFormat(mig->localHostuuid, hostuuidstr);
363 364 365 366

    virBufferAsprintf(buf, "<qemu-migration>\n");
    virBufferEscapeString(buf, "  <name>%s</name>\n", mig->name);
    virBufferAsprintf(buf, "  <uuid>%s</uuid>\n", uuidstr);
367
    virBufferEscapeString(buf, "  <hostname>%s</hostname>\n", mig->localHostname);
368
    virBufferAsprintf(buf, "  <hostuuid>%s</hostuuid>\n", hostuuidstr);
369

370 371 372 373 374 375
    for (i = 0 ; i < QEMU_MIGRATION_COOKIE_FLAG_LAST ; i++) {
        if (mig->flagsMandatory & (1 << i))
            virBufferAsprintf(buf, "  <feature name='%s'/>\n",
                              qemuMigrationCookieFlagTypeToString(i));
    }

376 377
    if ((mig->flags & QEMU_MIGRATION_COOKIE_GRAPHICS) &&
        mig->graphics)
378 379
        qemuMigrationCookieGraphicsXMLFormat(buf, mig->graphics);

380 381 382 383 384 385 386 387 388
    if ((mig->flags & QEMU_MIGRATION_COOKIE_LOCKSTATE) &&
        mig->lockState) {
        virBufferAsprintf(buf, "  <lockstate driver='%s'>\n",
                          mig->lockDriver);
        virBufferAsprintf(buf, "    <leases>%s</leases>\n",
                          mig->lockState);
        virBufferAddLit(buf, "  </lockstate>\n");
    }

389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407
    virBufferAddLit(buf, "</qemu-migration>\n");
}


static char *qemuMigrationCookieXMLFormatStr(qemuMigrationCookiePtr mig)
{
    virBuffer buf = VIR_BUFFER_INITIALIZER;

    qemuMigrationCookieXMLFormat(&buf, mig);

    if (virBufferError(&buf)) {
        virReportOOMError();
        return NULL;
    }

    return virBufferContentAndReset(&buf);
}


408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445
static qemuMigrationCookieGraphicsPtr
qemuMigrationCookieGraphicsXMLParse(xmlXPathContextPtr ctxt)
{
    qemuMigrationCookieGraphicsPtr grap;
    char *tmp;

    if (VIR_ALLOC(grap) < 0)
        goto no_memory;

    if (!(tmp = virXPathString("string(./graphics/@type)", ctxt))) {
        qemuReportError(VIR_ERR_INTERNAL_ERROR,
                        "%s", _("missing type attribute in migration data"));
        goto error;
    }
    if ((grap->type = virDomainGraphicsTypeFromString(tmp)) < 0) {
        qemuReportError(VIR_ERR_INTERNAL_ERROR,
                        _("unknown graphics type %s"), tmp);
        VIR_FREE(tmp);
        goto error;
    }
    if (virXPathInt("string(./graphics/@port)", ctxt, &grap->port) < 0) {
        qemuReportError(VIR_ERR_INTERNAL_ERROR,
                        "%s", _("missing port attribute in migration data"));
        goto error;
    }
    if (grap->type == VIR_DOMAIN_GRAPHICS_TYPE_SPICE) {
        if (virXPathInt("string(./graphics/@tlsPort)", ctxt, &grap->tlsPort) < 0) {
            qemuReportError(VIR_ERR_INTERNAL_ERROR,
                            "%s", _("missing tlsPort attribute in migration data"));
            goto error;
        }
    }
    if (!(grap->listen = virXPathString("string(./graphics/@listen)", ctxt))) {
        qemuReportError(VIR_ERR_INTERNAL_ERROR,
                        "%s", _("missing listen attribute in migration data"));
        goto error;
    }
    /* Optional */
446
    grap->tlsSubject = virXPathString("string(./graphics/cert[@info='subject']/@value)", ctxt);
447 448 449 450 451 452 453 454 455 456 457 458


    return grap;

no_memory:
    virReportOOMError();
error:
    qemuMigrationCookieGraphicsFree(grap);
    return NULL;
}


459 460 461
static int
qemuMigrationCookieXMLParse(qemuMigrationCookiePtr mig,
                            xmlXPathContextPtr ctxt,
E
Eric Blake 已提交
462
                            unsigned int flags)
463 464 465
{
    char uuidstr[VIR_UUID_STRING_BUFLEN];
    char *tmp;
466 467
    xmlNodePtr *nodes = NULL;
    int i, n;
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

    /* We don't store the uuid, name, hostname, or hostuuid
     * values. We just compare them to local data to do some
     * sanity checking on migration operation
     */

    /* Extract domain name */
    if (!(tmp = virXPathString("string(./name[1])", ctxt))) {
        qemuReportError(VIR_ERR_INTERNAL_ERROR,
                        "%s", _("missing name element in migration data"));
        goto error;
    }
    if (STRNEQ(tmp, mig->name)) {
        qemuReportError(VIR_ERR_INTERNAL_ERROR,
                        _("Incoming cookie data had unexpected name %s vs %s"),
                        tmp, mig->name);
        goto error;
    }
    VIR_FREE(tmp);

    /* Extract domain uuid */
    tmp = virXPathString("string(./uuid[1])", ctxt);
    if (!tmp) {
        qemuReportError(VIR_ERR_INTERNAL_ERROR,
                        "%s", _("missing uuid element in migration data"));
        goto error;
    }
    virUUIDFormat(mig->uuid, uuidstr);
    if (STRNEQ(tmp, uuidstr)) {
        qemuReportError(VIR_ERR_INTERNAL_ERROR,
                        _("Incoming cookie data had unexpected UUID %s vs %s"),
                        tmp, uuidstr);
    }
    VIR_FREE(tmp);

    /* Check & forbid "localhost" migration */
504
    if (!(mig->remoteHostname = virXPathString("string(./hostname[1])", ctxt))) {
505 506 507 508
        qemuReportError(VIR_ERR_INTERNAL_ERROR,
                        "%s", _("missing hostname element in migration data"));
        goto error;
    }
509
    if (STREQ(mig->remoteHostname, mig->localHostname)) {
510 511
        qemuReportError(VIR_ERR_INTERNAL_ERROR,
                        _("Attempt to migrate guest to the same host %s"),
512
                        mig->remoteHostname);
513 514 515 516 517 518 519 520
        goto error;
    }

    if (!(tmp = virXPathString("string(./hostuuid[1])", ctxt))) {
        qemuReportError(VIR_ERR_INTERNAL_ERROR,
                        "%s", _("missing hostuuid element in migration data"));
        goto error;
    }
521 522 523 524 525 526
    if (virUUIDParse(tmp, mig->remoteHostuuid) < 0) {
        qemuReportError(VIR_ERR_INTERNAL_ERROR,
                        "%s", _("malformed hostuuid element in migration data"));
        goto error;
    }
    if (memcmp(mig->remoteHostuuid, mig->localHostuuid, VIR_UUID_BUFLEN) == 0) {
527 528 529 530 531 532 533
        qemuReportError(VIR_ERR_INTERNAL_ERROR,
                        _("Attempt to migrate guest to the same host %s"),
                        tmp);
        goto error;
    }
    VIR_FREE(tmp);

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
    /* Check to ensure all mandatory features from XML are also
     * present in 'flags' */
    if ((n = virXPathNodeSet("./features", ctxt, &nodes)) < 0)
        goto error;

    for (i = 0 ; i < n ; i++) {
        int val;
        char *str = virXMLPropString(nodes[i], "name");
        if (!str) {
            qemuReportError(VIR_ERR_INTERNAL_ERROR,
                            "%s", _("missing feature name"));
            goto error;
        }

        if ((val = qemuMigrationCookieFlagTypeFromString(str)) < 0) {
            qemuReportError(VIR_ERR_INTERNAL_ERROR,
                            _("Unknown migration cookie feature %s"),
                            str);
            VIR_FREE(str);
            goto error;
        }

        if ((flags & (1 << val)) == 0) {
            qemuReportError(VIR_ERR_INTERNAL_ERROR,
                            _("Unsupported migration cookie feature %s"),
                            str);
            VIR_FREE(str);
        }
        VIR_FREE(str);
    }
    VIR_FREE(nodes);

566 567 568 569 570
    if ((flags & QEMU_MIGRATION_COOKIE_GRAPHICS) &&
        virXPathBoolean("count(./graphics) > 0", ctxt) &&
        (!(mig->graphics = qemuMigrationCookieGraphicsXMLParse(ctxt))))
        goto error;

571 572 573 574 575 576 577 578 579 580 581 582 583
    if ((flags & QEMU_MIGRATION_COOKIE_LOCKSTATE) &&
        virXPathBoolean("count(./lockstate) > 0", ctxt)) {
        mig->lockDriver = virXPathString("string(./lockstate[1]/@driver)", ctxt);
        if (!mig->lockDriver) {
            qemuReportError(VIR_ERR_INTERNAL_ERROR, "%s",
                            _("Missing lock driver name in migration cookie"));
            goto error;
        }
        mig->lockState = virXPathString("string(./lockstate[1]/leases[1])", ctxt);
        if (mig->lockState && STREQ(mig->lockState, ""))
            VIR_FREE(mig->lockState);
    }

584 585 586 587
    return 0;

error:
    VIR_FREE(tmp);
588
    VIR_FREE(nodes);
589 590 591 592 593 594 595
    return -1;
}


static int
qemuMigrationCookieXMLParseStr(qemuMigrationCookiePtr mig,
                               const char *xml,
E
Eric Blake 已提交
596
                               unsigned int flags)
597 598 599
{
    xmlDocPtr doc = NULL;
    xmlXPathContextPtr ctxt = NULL;
600
    int ret = -1;
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

    VIR_DEBUG("xml=%s", NULLSTR(xml));

    if (!(doc = virXMLParseString(xml, "qemumigration.xml")))
        goto cleanup;

    if ((ctxt = xmlXPathNewContext(doc)) == NULL) {
        virReportOOMError();
        goto cleanup;
    }

    ctxt->node = xmlDocGetRootElement(doc);

    ret = qemuMigrationCookieXMLParse(mig, ctxt, flags);

cleanup:
    xmlXPathFreeContext(ctxt);
    xmlFreeDoc(doc);

    return ret;
}


static int
qemuMigrationBakeCookie(qemuMigrationCookiePtr mig,
626 627
                        struct qemud_driver *driver,
                        virDomainObjPtr dom,
628 629
                        char **cookieout,
                        int *cookieoutlen,
E
Eric Blake 已提交
630
                        unsigned int flags)
631
{
632 633
    if (!cookieout || !cookieoutlen)
        return 0;
634 635 636

    *cookieoutlen = 0;

637 638 639 640
    if (flags & QEMU_MIGRATION_COOKIE_GRAPHICS &&
        qemuMigrationCookieAddGraphics(mig, driver, dom) < 0)
        return -1;

641 642 643 644
    if (flags & QEMU_MIGRATION_COOKIE_LOCKSTATE &&
        qemuMigrationCookieAddLockstate(mig, driver, dom) < 0)
        return -1;

645 646 647 648 649 650 651 652 653 654 655 656
    if (!(*cookieout = qemuMigrationCookieXMLFormatStr(mig)))
        return -1;

    *cookieoutlen = strlen(*cookieout) + 1;

    VIR_DEBUG("cookielen=%d cookie=%s", *cookieoutlen, *cookieout);

    return 0;
}


static qemuMigrationCookiePtr
657 658
qemuMigrationEatCookie(struct qemud_driver *driver,
                       virDomainObjPtr dom,
659 660
                       const char *cookiein,
                       int cookieinlen,
E
Eric Blake 已提交
661
                       unsigned int flags)
662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683
{
    qemuMigrationCookiePtr mig = NULL;

    /* Parse & validate incoming cookie (if any) */
    if (cookiein && cookieinlen &&
        cookiein[cookieinlen-1] != '\0') {
        qemuReportError(VIR_ERR_INTERNAL_ERROR, "%s",
                        _("Migration cookie was not NULL terminated"));
        goto error;
    }

    VIR_DEBUG("cookielen=%d cookie='%s'", cookieinlen, NULLSTR(cookiein));

    if (!(mig = qemuMigrationCookieNew(dom)))
        return NULL;

    if (cookiein && cookieinlen &&
        qemuMigrationCookieXMLParseStr(mig,
                                       cookiein,
                                       flags) < 0)
        goto error;

684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701
    if (mig->flags & QEMU_MIGRATION_COOKIE_LOCKSTATE) {
        if (!mig->lockDriver) {
            if (virLockManagerPluginUsesState(driver->lockManager)) {
                qemuReportError(VIR_ERR_INTERNAL_ERROR,
                                _("Missing %s lock state for migration cookie"),
                                virLockManagerPluginGetName(driver->lockManager));
                goto error;
            }
        } else if (STRNEQ(mig->lockDriver,
                          virLockManagerPluginGetName(driver->lockManager))) {
            qemuReportError(VIR_ERR_INTERNAL_ERROR,
                            _("Source host lock driver %s different from target %s"),
                            mig->lockDriver,
                            virLockManagerPluginGetName(driver->lockManager));
            goto error;
        }
    }

702 703 704 705 706 707
    return mig;

error:
    qemuMigrationCookieFree(mig);
    return NULL;
}
708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728

bool
qemuMigrationIsAllowed(virDomainDefPtr def)
{
    if (def->nhostdevs > 0) {
        qemuReportError(VIR_ERR_OPERATION_INVALID,
            "%s", _("Domain with assigned host devices cannot be migrated"));
        return false;
    }

    return true;
}

/** qemuMigrationSetOffline
 * Pause domain for non-live migration.
 */
int
qemuMigrationSetOffline(struct qemud_driver *driver,
                        virDomainObjPtr vm)
{
    int ret;
729
    VIR_DEBUG("driver=%p vm=%p", driver, vm);
730 731
    ret = qemuProcessStopCPUs(driver, vm, VIR_DOMAIN_PAUSED_MIGRATION,
                              QEMU_ASYNC_JOB_MIGRATION_OUT);
732 733 734 735 736 737 738 739 740 741 742 743 744 745
    if (ret == 0) {
        virDomainEventPtr event;

        event = virDomainEventNewFromObj(vm,
                                         VIR_DOMAIN_EVENT_SUSPENDED,
                                         VIR_DOMAIN_EVENT_SUSPENDED_MIGRATED);
        if (event)
            qemuDomainEventQueue(driver, event);
    }

    return ret;
}


746 747 748
static int
qemuMigrationUpdateJobStatus(struct qemud_driver *driver,
                             virDomainObjPtr vm,
749 750
                             const char *job,
                             enum qemuDomainAsyncJob asyncJob)
751
{
752
    qemuDomainObjPrivatePtr priv = vm->privateData;
753 754 755 756 757
    int ret = -1;
    int status;
    unsigned long long memProcessed;
    unsigned long long memRemaining;
    unsigned long long memTotal;
758

759 760 761
    ret = qemuDomainObjEnterMonitorAsync(driver, vm, asyncJob);
    if (ret < 0) {
        /* Guest already exited; nothing further to update.  */
762 763
        return -1;
    }
764 765 766 767 768 769
    ret = qemuMonitorGetMigrationStatus(priv->mon,
                                        &status,
                                        &memProcessed,
                                        &memRemaining,
                                        &memTotal);
    qemuDomainObjExitMonitorWithDriver(driver, vm);
770

771 772
    if (ret < 0 || virTimeMs(&priv->job.info.timeElapsed) < 0) {
        priv->job.info.type = VIR_DOMAIN_JOB_FAILED;
773 774
        return -1;
    }
775
    priv->job.info.timeElapsed -= priv->job.start;
776 777 778

    switch (status) {
    case QEMU_MONITOR_MIGRATION_STATUS_INACTIVE:
779
        priv->job.info.type = VIR_DOMAIN_JOB_NONE;
780 781 782 783 784
        qemuReportError(VIR_ERR_OPERATION_FAILED,
                        _("%s: %s"), job, _("is not active"));
        break;

    case QEMU_MONITOR_MIGRATION_STATUS_ACTIVE:
785 786 787
        priv->job.info.dataTotal = memTotal;
        priv->job.info.dataRemaining = memRemaining;
        priv->job.info.dataProcessed = memProcessed;
788

789 790 791
        priv->job.info.memTotal = memTotal;
        priv->job.info.memRemaining = memRemaining;
        priv->job.info.memProcessed = memProcessed;
792 793 794 795 796

        ret = 0;
        break;

    case QEMU_MONITOR_MIGRATION_STATUS_COMPLETED:
797
        priv->job.info.type = VIR_DOMAIN_JOB_COMPLETED;
798 799 800 801
        ret = 0;
        break;

    case QEMU_MONITOR_MIGRATION_STATUS_ERROR:
802
        priv->job.info.type = VIR_DOMAIN_JOB_FAILED;
803 804 805 806 807
        qemuReportError(VIR_ERR_OPERATION_FAILED,
                        _("%s: %s"), job, _("unexpectedly failed"));
        break;

    case QEMU_MONITOR_MIGRATION_STATUS_CANCELLED:
808
        priv->job.info.type = VIR_DOMAIN_JOB_CANCELLED;
809 810 811 812 813 814 815 816 817
        qemuReportError(VIR_ERR_OPERATION_FAILED,
                        _("%s: %s"), job, _("canceled by client"));
        break;
    }

    return ret;
}


818 819 820
static int
qemuMigrationWaitForCompletion(struct qemud_driver *driver, virDomainObjPtr vm,
                               enum qemuDomainAsyncJob asyncJob)
821
{
822
    qemuDomainObjPrivatePtr priv = vm->privateData;
823 824
    const char *job;

825 826
    switch (priv->job.asyncJob) {
    case QEMU_ASYNC_JOB_MIGRATION_OUT:
827 828
        job = _("migration job");
        break;
829
    case QEMU_ASYNC_JOB_SAVE:
830 831
        job = _("domain save job");
        break;
832
    case QEMU_ASYNC_JOB_DUMP:
833 834 835 836 837
        job = _("domain core dump job");
        break;
    default:
        job = _("job");
    }
838

839
    priv->job.info.type = VIR_DOMAIN_JOB_UNBOUNDED;
840

841
    while (priv->job.info.type == VIR_DOMAIN_JOB_UNBOUNDED) {
842 843 844
        /* Poll every 50ms for progress & to allow cancellation */
        struct timespec ts = { .tv_sec = 0, .tv_nsec = 50 * 1000 * 1000ull };

845
        if (qemuMigrationUpdateJobStatus(driver, vm, job, asyncJob) < 0)
846 847 848 849 850 851 852 853 854 855 856 857
            goto cleanup;

        virDomainObjUnlock(vm);
        qemuDriverUnlock(driver);

        nanosleep(&ts, NULL);

        qemuDriverLock(driver);
        virDomainObjLock(vm);
    }

cleanup:
858
    if (priv->job.info.type == VIR_DOMAIN_JOB_COMPLETED)
859 860 861
        return 0;
    else
        return -1;
862 863 864
}


865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884
static int
qemuDomainMigrateGraphicsRelocate(struct qemud_driver *driver,
                                  virDomainObjPtr vm,
                                  qemuMigrationCookiePtr cookie)
{
    qemuDomainObjPrivatePtr priv = vm->privateData;
    int ret;

    if (!cookie)
        return 0;

    if (!cookie->graphics)
        return 0;

    /* QEMU doesn't support VNC relocation yet, so
     * skip it to avoid generating an error
     */
    if (cookie->graphics->type != VIR_DOMAIN_GRAPHICS_TYPE_SPICE)
        return 0;

885 886
    ret = qemuDomainObjEnterMonitorAsync(driver, vm,
                                         QEMU_ASYNC_JOB_MIGRATION_OUT);
887 888 889 890 891 892 893 894 895
    if (ret == 0) {
        ret = qemuMonitorGraphicsRelocate(priv->mon,
                                          cookie->graphics->type,
                                          cookie->remoteHostname,
                                          cookie->graphics->port,
                                          cookie->graphics->tlsPort,
                                          cookie->graphics->tlsSubject);
        qemuDomainObjExitMonitorWithDriver(driver, vm);
    }
896 897 898 899 900

    return ret;
}


901
/* The caller is supposed to lock the vm and start a migration job. */
902 903
char *qemuMigrationBegin(struct qemud_driver *driver,
                         virDomainObjPtr vm,
904
                         const char *xmlin,
905 906 907 908 909
                         char **cookieout,
                         int *cookieoutlen)
{
    char *rv = NULL;
    qemuMigrationCookiePtr mig = NULL;
910
    virDomainDefPtr def = NULL;
911 912
    qemuDomainObjPrivatePtr priv = vm->privateData;

913 914
    VIR_DEBUG("driver=%p, vm=%p, xmlin=%s, cookieout=%p, cookieoutlen=%p",
              driver, vm, NULLSTR(xmlin), cookieout, cookieoutlen);
915

916 917 918 919 920 921
    /* Only set the phase if we are inside QEMU_ASYNC_JOB_MIGRATION_OUT.
     * Otherwise we will start the async job later in the perform phase losing
     * change protection.
     */
    if (priv->job.asyncJob == QEMU_ASYNC_JOB_MIGRATION_OUT)
        qemuMigrationJobSetPhase(driver, vm, QEMU_MIGRATION_PHASE_BEGIN3);
922

923 924 925 926 927 928
    if (qemuProcessAutoDestroyActive(driver, vm)) {
        qemuReportError(VIR_ERR_OPERATION_INVALID,
                        "%s", _("domain is marked for auto destroy"));
        goto cleanup;
    }

929 930 931
    if (!qemuMigrationIsAllowed(vm->def))
        goto cleanup;

932
    if (!(mig = qemuMigrationEatCookie(driver, vm, NULL, 0, 0)))
933 934 935 936
        goto cleanup;

    if (qemuMigrationBakeCookie(mig, driver, vm,
                                cookieout, cookieoutlen,
937
                                QEMU_MIGRATION_COOKIE_LOCKSTATE) < 0)
938 939
        goto cleanup;

940 941
    if (xmlin) {
        if (!(def = virDomainDefParseString(driver->caps, xmlin,
M
Matthias Bolte 已提交
942
                                            QEMU_EXPECTED_VIRT_TYPES,
943 944 945
                                            VIR_DOMAIN_XML_INACTIVE)))
            goto cleanup;

946
        if (!virDomainDefCheckABIStability(vm->def, def))
947 948 949 950 951 952 953 954 955 956
            goto cleanup;

        rv = qemuDomainDefFormatXML(driver, def,
                                    VIR_DOMAIN_XML_SECURE |
                                    VIR_DOMAIN_XML_UPDATE_CPU);
    } else {
        rv = qemuDomainFormatXML(driver, vm,
                                 VIR_DOMAIN_XML_SECURE |
                                 VIR_DOMAIN_XML_UPDATE_CPU);
    }
957 958 959

cleanup:
    qemuMigrationCookieFree(mig);
960
    virDomainDefFree(def);
961 962 963 964
    return rv;
}


965 966
/* Prepare is the first step, and it runs on the destination host.
 */
967 968 969 970 971 972 973 974 975 976 977 978

static int
qemuMigrationPrepareAny(struct qemud_driver *driver,
                        virConnectPtr dconn,
                        const char *cookiein,
                        int cookieinlen,
                        char **cookieout,
                        int *cookieoutlen,
                        const char *dname,
                        const char *dom_xml,
                        const char *migrateFrom,
                        virStreamPtr st)
979 980 981 982 983
{
    virDomainDefPtr def = NULL;
    virDomainObjPtr vm = NULL;
    virDomainEventPtr event = NULL;
    int ret = -1;
984
    int dataFD[2] = { -1, -1 };
985
    qemuDomainObjPrivatePtr priv = NULL;
J
Jiri Denemark 已提交
986
    unsigned long long now;
987
    qemuMigrationCookiePtr mig = NULL;
988
    bool tunnel = !!st;
989

J
Jiri Denemark 已提交
990
    if (virTimeMs(&now) < 0)
991 992 993
        return -1;

    if (!(def = virDomainDefParseString(driver->caps, dom_xml,
M
Matthias Bolte 已提交
994
                                        QEMU_EXPECTED_VIRT_TYPES,
995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020
                                        VIR_DOMAIN_XML_INACTIVE)))
        goto cleanup;

    if (!qemuMigrationIsAllowed(def))
        goto cleanup;

    /* Target domain name, maybe renamed. */
    if (dname) {
        VIR_FREE(def->name);
        def->name = strdup(dname);
        if (def->name == NULL)
            goto cleanup;
    }

    if (virDomainObjIsDuplicate(&driver->domains, def, 1) < 0)
        goto cleanup;

    if (!(vm = virDomainAssignDef(driver->caps,
                                  &driver->domains,
                                  def, true))) {
        /* virDomainAssignDef already set the error */
        goto cleanup;
    }
    def = NULL;
    priv = vm->privateData;

1021 1022
    if (!(mig = qemuMigrationEatCookie(driver, vm, cookiein, cookieinlen,
                                       QEMU_MIGRATION_COOKIE_LOCKSTATE)))
1023 1024
        goto cleanup;

1025
    if (qemuMigrationJobStart(driver, vm, QEMU_ASYNC_JOB_MIGRATION_IN) < 0)
1026
        goto cleanup;
1027
    qemuMigrationJobSetPhase(driver, vm, QEMU_MIGRATION_PHASE_PREPARE);
1028 1029 1030 1031

    /* Domain starts inactive, even if the domain XML had an id field. */
    vm->def->id = -1;

1032 1033
    if (tunnel &&
        (pipe(dataFD) < 0 || virSetCloseExec(dataFD[1]) < 0)) {
1034 1035
        virReportSystemError(errno, "%s",
                             _("cannot create pipe for tunnelled migration"));
1036 1037 1038 1039
        goto endjob;
    }

    /* Start the QEMU daemon, with the same command-line arguments plus
1040
     * -incoming $migrateFrom
1041
     */
1042 1043 1044
    if (qemuProcessStart(dconn, driver, vm, migrateFrom, true,
                         true, dataFD[0], NULL,
                         VIR_VM_OP_MIGRATE_IN_START) < 0) {
1045
        virDomainAuditStart(vm, "migrated", false);
1046 1047 1048 1049 1050 1051
        /* Note that we don't set an error here because qemuProcessStart
         * should have already done that.
         */
        goto endjob;
    }

1052 1053 1054 1055 1056 1057 1058
    if (tunnel) {
        if (virFDStreamOpen(st, dataFD[1]) < 0) {
            virReportSystemError(errno, "%s",
                                 _("cannot pass pipe for tunnelled migration"));
            virDomainAuditStart(vm, "migrated", false);
            qemuProcessStop(driver, vm, 0, VIR_DOMAIN_SHUTOFF_FAILED);
            goto endjob;
1059
        }
1060
        dataFD[1] = -1; /* 'st' owns the FD now & will close it */
1061 1062
    }

1063 1064 1065 1066 1067 1068 1069 1070
    if (mig->lockState) {
        VIR_DEBUG("Received lockstate %s", mig->lockState);
        VIR_FREE(priv->lockState);
        priv->lockState = mig->lockState;
        mig->lockState = NULL;
    } else {
        VIR_DEBUG("Received no lockstate");
    }
1071

1072 1073
    if (qemuMigrationBakeCookie(mig, driver, vm, cookieout, cookieoutlen,
                                QEMU_MIGRATION_COOKIE_GRAPHICS) < 0) {
1074 1075 1076 1077 1078 1079 1080
        /* We could tear down the whole guest here, but
         * cookie data is (so far) non-critical, so that
         * seems a little harsh. We'll just warn for now.
         */
        VIR_WARN("Unable to encode migration cookie");
    }

1081 1082 1083 1084
    virDomainAuditStart(vm, "migrated", true);
    event = virDomainEventNewFromObj(vm,
                                     VIR_DOMAIN_EVENT_STARTED,
                                     VIR_DOMAIN_EVENT_STARTED_MIGRATED);
1085

1086 1087 1088 1089 1090
    /* We keep the job active across API calls until the finish() call.
     * This prevents any other APIs being invoked while incoming
     * migration is taking place.
     */
    if (qemuMigrationJobContinue(vm) == 0) {
1091
        vm = NULL;
1092 1093 1094
        qemuReportError(VIR_ERR_OPERATION_FAILED,
                        "%s", _("domain disappeared"));
        goto cleanup;
1095
    }
1096

1097
    ret = 0;
1098 1099 1100

cleanup:
    virDomainDefFree(def);
1101 1102
    VIR_FORCE_CLOSE(dataFD[0]);
    VIR_FORCE_CLOSE(dataFD[1]);
1103 1104 1105 1106
    if (vm)
        virDomainObjUnlock(vm);
    if (event)
        qemuDomainEventQueue(driver, event);
1107
    qemuMigrationCookieFree(mig);
1108
    return ret;
1109 1110 1111 1112 1113 1114 1115 1116 1117

endjob:
    if (qemuMigrationJobFinish(driver, vm) == 0) {
        vm = NULL;
    } else if (!vm->persistent) {
        virDomainRemoveInactive(&driver->domains, vm);
        vm = NULL;
    }
    goto cleanup;
1118 1119 1120
}


1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152
/*
 * This version starts an empty VM listening on a localhost TCP port, and
 * sets up the corresponding virStream to handle the incoming data.
 */
int
qemuMigrationPrepareTunnel(struct qemud_driver *driver,
                           virConnectPtr dconn,
                           const char *cookiein,
                           int cookieinlen,
                           char **cookieout,
                           int *cookieoutlen,
                           virStreamPtr st,
                           const char *dname,
                           const char *dom_xml)
{
    int ret;

    VIR_DEBUG("driver=%p, dconn=%p, cookiein=%s, cookieinlen=%d, "
              "cookieout=%p, cookieoutlen=%p, st=%p, dname=%s, dom_xml=%s",
              driver, dconn, NULLSTR(cookiein), cookieinlen,
              cookieout, cookieoutlen, st, NULLSTR(dname), dom_xml);

    /* QEMU will be started with -incoming stdio (which qemu_command might
     * convert to exec:cat or fd:n)
     */
    ret = qemuMigrationPrepareAny(driver, dconn, cookiein, cookieinlen,
                                  cookieout, cookieoutlen, dname, dom_xml,
                                  "stdio", st);
    return ret;
}


1153 1154 1155
int
qemuMigrationPrepareDirect(struct qemud_driver *driver,
                           virConnectPtr dconn,
1156 1157 1158 1159
                           const char *cookiein,
                           int cookieinlen,
                           char **cookieout,
                           int *cookieoutlen,
1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170
                           const char *uri_in,
                           char **uri_out,
                           const char *dname,
                           const char *dom_xml)
{
    static int port = 0;
    int this_port;
    char *hostname = NULL;
    char migrateFrom [64];
    const char *p;
    int ret = -1;
J
Jiri Denemark 已提交
1171

1172 1173 1174 1175 1176 1177
    VIR_DEBUG("driver=%p, dconn=%p, cookiein=%s, cookieinlen=%d, "
              "cookieout=%p, cookieoutlen=%p, uri_in=%s, uri_out=%p, "
              "dname=%s, dom_xml=%s",
              driver, dconn, NULLSTR(cookiein), cookieinlen,
              cookieout, cookieoutlen, NULLSTR(uri_in), uri_out,
              NULLSTR(dname), dom_xml);
1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198

    /* The URI passed in may be NULL or a string "tcp://somehostname:port".
     *
     * If the URI passed in is NULL then we allocate a port number
     * from our pool of port numbers and return a URI of
     * "tcp://ourhostname:port".
     *
     * If the URI passed in is not NULL then we try to parse out the
     * port number and use that (note that the hostname is assumed
     * to be a correct hostname which refers to the target machine).
     */
    if (uri_in == NULL) {
        this_port = QEMUD_MIGRATION_FIRST_PORT + port++;
        if (port == QEMUD_MIGRATION_NUM_PORTS) port = 0;

        /* Get hostname */
        if ((hostname = virGetHostname(NULL)) == NULL)
            goto cleanup;

        if (STRPREFIX(hostname, "localhost")) {
            qemuReportError(VIR_ERR_INTERNAL_ERROR, "%s",
1199 1200
                            _("hostname on destination resolved to localhost,"
                              " but migration requires an FQDN"));
1201 1202 1203 1204 1205 1206 1207 1208 1209
            goto cleanup;
        }

        /* XXX this really should have been a properly well-formed
         * URI, but we can't add in tcp:// now without breaking
         * compatability with old targets. We at least make the
         * new targets accept both syntaxes though.
         */
        /* Caller frees */
1210
        if (virAsprintf(uri_out, "tcp:%s:%d", hostname, this_port) < 0) {
1211 1212 1213 1214 1215 1216 1217 1218 1219
            virReportOOMError();
            goto cleanup;
        }
    } else {
        /* Check the URI starts with "tcp:".  We will escape the
         * URI when passing it to the qemu monitor, so bad
         * characters in hostname part don't matter.
         */
        if (!STRPREFIX (uri_in, "tcp:")) {
1220 1221 1222
            qemuReportError(VIR_ERR_INVALID_ARG, "%s",
                            _("only tcp URIs are supported for KVM/QEMU"
                              " migrations"));
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
            goto cleanup;
        }

        /* Get the port number. */
        p = strrchr (uri_in, ':');
        if (p == strchr(uri_in, ':')) {
            /* Generate a port */
            this_port = QEMUD_MIGRATION_FIRST_PORT + port++;
            if (port == QEMUD_MIGRATION_NUM_PORTS)
                port = 0;

            /* Caller frees */
            if (virAsprintf(uri_out, "%s:%d", uri_in, this_port) < 0) {
                virReportOOMError();
                goto cleanup;
            }

        } else {
            p++; /* definitely has a ':' in it, see above */
            this_port = virParseNumber (&p);
            if (this_port == -1 || p-uri_in != strlen (uri_in)) {
                qemuReportError(VIR_ERR_INVALID_ARG,
                                "%s", _("URI ended with incorrect ':port'"));
                goto cleanup;
            }
        }
    }

    if (*uri_out)
        VIR_DEBUG("Generated uri_out=%s", *uri_out);

1254 1255
    /* QEMU will be started with -incoming tcp:0.0.0.0:port */
    snprintf(migrateFrom, sizeof (migrateFrom), "tcp:0.0.0.0:%d", this_port);
1256

1257 1258 1259
    ret = qemuMigrationPrepareAny(driver, dconn, cookiein, cookieinlen,
                                  cookieout, cookieoutlen, dname, dom_xml,
                                  migrateFrom, NULL);
1260 1261 1262 1263 1264 1265 1266 1267
cleanup:
    VIR_FREE(hostname);
    if (ret != 0)
        VIR_FREE(*uri_out);
    return ret;
}


1268 1269 1270 1271
enum qemuMigrationDestinationType {
    MIGRATION_DEST_HOST,
    MIGRATION_DEST_UNIX,
};
1272

1273 1274 1275 1276
enum qemuMigrationForwardType {
    MIGRATION_FWD_DIRECT,
    MIGRATION_FWD_STREAM,
};
1277

1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298
typedef struct _qemuMigrationSpec qemuMigrationSpec;
typedef qemuMigrationSpec *qemuMigrationSpecPtr;
struct _qemuMigrationSpec {
    enum qemuMigrationDestinationType destType;
    union {
        struct {
            const char *name;
            int port;
        } host;

        struct {
            const char *file;
            int sock;
        } unix_socket;
    } dest;

    enum qemuMigrationForwardType fwdType;
    union {
        virStreamPtr stream;
    } fwd;
};
1299 1300 1301

#define TUNNEL_SEND_BUF_SIZE 65536

1302 1303 1304 1305 1306 1307 1308 1309 1310 1311
typedef struct _qemuMigrationIOThread qemuMigrationIOThread;
typedef qemuMigrationIOThread *qemuMigrationIOThreadPtr;
struct _qemuMigrationIOThread {
    virThread thread;
    virStreamPtr st;
    int sock;
    virError err;
};

static void qemuMigrationIOFunc(void *arg)
1312
{
1313
    qemuMigrationIOThreadPtr data = arg;
1314 1315 1316 1317 1318
    char *buffer;
    int nbytes = TUNNEL_SEND_BUF_SIZE;

    if (VIR_ALLOC_N(buffer, TUNNEL_SEND_BUF_SIZE) < 0) {
        virReportOOMError();
1319 1320
        virStreamAbort(data->st);
        goto error;
1321 1322 1323
    }

    for (;;) {
1324
        nbytes = saferead(data->sock, buffer, TUNNEL_SEND_BUF_SIZE);
1325 1326 1327
        if (nbytes < 0) {
            virReportSystemError(errno, "%s",
                                 _("tunnelled migration failed to read from qemu"));
1328
            virStreamAbort(data->st);
1329
            VIR_FREE(buffer);
1330
            goto error;
1331 1332 1333 1334 1335
        }
        else if (nbytes == 0)
            /* EOF; get out of here */
            break;

1336
        if (virStreamSend(data->st, buffer, nbytes) < 0) {
1337
            VIR_FREE(buffer);
1338
            goto error;
1339 1340 1341 1342 1343
        }
    }

    VIR_FREE(buffer);

1344 1345
    if (virStreamFinish(data->st) < 0)
        goto error;
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
    return;

error:
    virCopyLastError(&data->err);
    virResetLastError();
}


static qemuMigrationIOThreadPtr
qemuMigrationStartTunnel(virStreamPtr st,
                         int sock)
{
    qemuMigrationIOThreadPtr io;

    if (VIR_ALLOC(io) < 0) {
        virReportOOMError();
        return NULL;
    }

    io->st = st;
    io->sock = sock;

    if (virThreadCreate(&io->thread, true,
                        qemuMigrationIOFunc,
                        io) < 0) {
        virReportSystemError(errno, "%s",
                             _("Unable to create migration thread"));
        VIR_FREE(io);
        return NULL;
    }

    return io;
}

static int
qemuMigrationStopTunnel(qemuMigrationIOThreadPtr io)
{
    int rv = -1;
    virThreadJoin(&io->thread);

    /* Forward error from the IO thread, to this thread */
    if (io->err.code != VIR_ERR_OK) {
        virSetError(&io->err);
        virResetError(&io->err);
        goto cleanup;
    }

    rv = 0;

cleanup:
    VIR_FREE(io);
    return rv;
1399 1400
}

1401 1402 1403 1404 1405 1406 1407 1408 1409 1410
static int
qemuMigrationRun(struct qemud_driver *driver,
                 virDomainObjPtr vm,
                 const char *cookiein,
                 int cookieinlen,
                 char **cookieout,
                 int *cookieoutlen,
                 unsigned long flags,
                 unsigned long resource,
                 qemuMigrationSpecPtr spec)
1411
{
1412
    int ret = -1;
1413 1414
    unsigned int migrate_flags = QEMU_MONITOR_MIGRATE_BACKGROUND;
    qemuDomainObjPrivatePtr priv = vm->privateData;
1415
    qemuMigrationCookiePtr mig = NULL;
1416
    qemuMigrationIOThreadPtr iothread = NULL;
1417 1418 1419 1420 1421 1422 1423 1424
    int fd = -1;

    VIR_DEBUG("driver=%p, vm=%p, cookiein=%s, cookieinlen=%d, "
              "cookieout=%p, cookieoutlen=%p, flags=%lx, resource=%lu, "
              "spec=%p (dest=%d, fwd=%d)",
              driver, vm, NULLSTR(cookiein), cookieinlen,
              cookieout, cookieoutlen, flags, resource,
              spec, spec->destType, spec->fwdType);
1425

1426 1427 1428
    if (virLockManagerPluginUsesState(driver->lockManager) &&
        !cookieout) {
        qemuReportError(VIR_ERR_INTERNAL_ERROR,
1429 1430
                        _("Migration with lock driver %s requires"
                          " cookie support"),
1431 1432 1433 1434 1435
                        virLockManagerPluginGetName(driver->lockManager));
        return -1;
    }

    if (!(mig = qemuMigrationEatCookie(driver, vm, cookiein, cookieinlen,
1436 1437 1438 1439 1440 1441
                                       QEMU_MIGRATION_COOKIE_GRAPHICS)))
        goto cleanup;

    if (qemuDomainMigrateGraphicsRelocate(driver, vm, mig) < 0)
        VIR_WARN("unable to provide data for graphics client relocation");

1442
    /* Before EnterMonitor, since qemuMigrationSetOffline already does that */
1443 1444 1445 1446 1447 1448
    if (!(flags & VIR_MIGRATE_LIVE) &&
        virDomainObjGetState(vm, NULL) == VIR_DOMAIN_RUNNING) {
        if (qemuMigrationSetOffline(driver, vm) < 0)
            goto cleanup;
    }

1449 1450
    if (qemuDomainObjEnterMonitorAsync(driver, vm,
                                       QEMU_ASYNC_JOB_MIGRATION_OUT) < 0)
1451 1452
        goto cleanup;

1453 1454 1455 1456 1457
    if (resource > 0 &&
        qemuMonitorSetMigrationSpeed(priv->mon, resource) < 0) {
        qemuDomainObjExitMonitorWithDriver(driver, vm);
        goto cleanup;
    }
1458

1459
    if (flags & VIR_MIGRATE_NON_SHARED_DISK)
1460 1461
        migrate_flags |= QEMU_MONITOR_MIGRATE_NON_SHARED_DISK;

1462
    if (flags & VIR_MIGRATE_NON_SHARED_INC)
1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482
        migrate_flags |= QEMU_MONITOR_MIGRATE_NON_SHARED_INC;

    switch (spec->destType) {
    case MIGRATION_DEST_HOST:
        ret = qemuMonitorMigrateToHost(priv->mon, migrate_flags,
                                       spec->dest.host.name,
                                       spec->dest.host.port);
        break;

    case MIGRATION_DEST_UNIX:
        if (qemuCapsGet(priv->qemuCaps, QEMU_CAPS_MIGRATE_QEMU_UNIX)) {
            ret = qemuMonitorMigrateToUnix(priv->mon, migrate_flags,
                                           spec->dest.unix_socket.file);
        } else {
            const char *args[] = {
                "nc", "-U", spec->dest.unix_socket.file, NULL
            };
            ret = qemuMonitorMigrateToCommand(priv->mon, migrate_flags, args);
        }
        break;
1483 1484
    }
    qemuDomainObjExitMonitorWithDriver(driver, vm);
1485
    if (ret < 0)
1486 1487
        goto cleanup;
    ret = -1;
1488 1489 1490 1491 1492 1493 1494 1495 1496 1497

    if (!virDomainObjIsActive(vm)) {
        qemuReportError(VIR_ERR_INTERNAL_ERROR, "%s",
                        _("guest unexpectedly quit"));
        goto cleanup;
    }

    /* From this point onwards we *must* call cancel to abort the
     * migration on source if anything goes wrong */

1498 1499 1500 1501 1502 1503 1504 1505
    if (spec->destType == MIGRATION_DEST_UNIX) {
        /* It is also possible that the migrate didn't fail initially, but
         * rather failed later on.  Check its status before waiting for a
         * connection from qemu which may never be initiated.
         */
        if (qemuMigrationUpdateJobStatus(driver, vm, _("migration job"),
                                         QEMU_ASYNC_JOB_MIGRATION_OUT) < 0)
            goto cancel;
1506

1507 1508 1509 1510 1511 1512 1513
        while ((fd = accept(spec->dest.unix_socket.sock, NULL, NULL)) < 0) {
            if (errno == EAGAIN || errno == EINTR)
                continue;
            virReportSystemError(errno, "%s",
                                 _("failed to accept connection from qemu"));
            goto cancel;
        }
1514 1515
    }

1516 1517
    if (spec->fwdType != MIGRATION_FWD_DIRECT &&
        !(iothread = qemuMigrationStartTunnel(spec->fwd.stream, fd)))
1518 1519
        goto cancel;

1520 1521 1522
    if (qemuMigrationWaitForCompletion(driver, vm,
                                       QEMU_ASYNC_JOB_MIGRATION_OUT) < 0)
        goto cleanup;
1523

1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536
    /* When migration completed, QEMU will have paused the
     * CPUs for us, but unless we're using the JSON monitor
     * we won't have been notified of this, so might still
     * think we're running. For v2 protocol this doesn't
     * matter because we'll kill the VM soon, but for v3
     * this is important because we stay paused until the
     * confirm3 step, but need to release the lock state
     */
    if (virDomainObjGetState(vm, NULL) == VIR_DOMAIN_RUNNING) {
        if (qemuMigrationSetOffline(driver, vm) < 0)
            goto cleanup;
    }

1537
    ret = 0;
1538

1539 1540 1541 1542 1543 1544 1545
cleanup:
    if (spec->fwdType != MIGRATION_FWD_DIRECT) {
        /* Close now to ensure the IO thread quits & is joinable */
        VIR_FORCE_CLOSE(fd);
        if (iothread && qemuMigrationStopTunnel(iothread) < 0)
            ret = -1;
    }
1546

1547 1548 1549 1550
    if (ret == 0 &&
        qemuMigrationBakeCookie(mig, driver, vm, cookieout, cookieoutlen, 0) < 0)
        VIR_WARN("Unable to encode migration cookie");

1551 1552 1553 1554
    qemuMigrationCookieFree(mig);

    return ret;

1555
cancel:
1556
    if (virDomainObjIsActive(vm)) {
1557 1558
        if (qemuDomainObjEnterMonitorAsync(driver, vm,
                                           QEMU_ASYNC_JOB_MIGRATION_OUT) == 0) {
1559 1560 1561
            qemuMonitorMigrateCancel(priv->mon);
            qemuDomainObjExitMonitorWithDriver(driver, vm);
        }
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
    goto cleanup;
}

/* Perform migration using QEMU's native TCP migrate support,
 * not encrypted obviously
 */
static int doNativeMigrate(struct qemud_driver *driver,
                           virDomainObjPtr vm,
                           const char *uri,
                           const char *cookiein,
                           int cookieinlen,
                           char **cookieout,
                           int *cookieoutlen,
                           unsigned long flags,
                           unsigned long resource)
{
    xmlURIPtr uribits = NULL;
    int ret;
    qemuMigrationSpec spec;

    VIR_DEBUG("driver=%p, vm=%p, uri=%s, cookiein=%s, cookieinlen=%d, "
              "cookieout=%p, cookieoutlen=%p, flags=%lx, resource=%lu",
              driver, vm, uri, NULLSTR(cookiein), cookieinlen,
              cookieout, cookieoutlen, flags, resource);

    if (STRPREFIX(uri, "tcp:") && !STRPREFIX(uri, "tcp://")) {
        /* HACK: source host generates bogus URIs, so fix them up */
        char *tmpuri;
        if (virAsprintf(&tmpuri, "tcp://%s", uri + strlen("tcp:")) < 0) {
            virReportOOMError();
            return -1;
        }
        uribits = xmlParseURI(tmpuri);
        VIR_FREE(tmpuri);
    } else {
        uribits = xmlParseURI(uri);
    }
    if (!uribits) {
        qemuReportError(VIR_ERR_INTERNAL_ERROR,
                        _("cannot parse URI %s"), uri);
        return -1;
    }

    spec.destType = MIGRATION_DEST_HOST;
    spec.dest.host.name = uribits->server;
    spec.dest.host.port = uribits->port;
    spec.fwdType = MIGRATION_FWD_DIRECT;

    ret = qemuMigrationRun(driver, vm, cookiein, cookieinlen, cookieout,
                           cookieoutlen, flags, resource, &spec);
    xmlFreeURI(uribits);

    return ret;
}


static int doTunnelMigrate(struct qemud_driver *driver,
                           virDomainObjPtr vm,
                           virStreamPtr st,
                           const char *cookiein,
                           int cookieinlen,
                           char **cookieout,
                           int *cookieoutlen,
                           unsigned long flags,
                           unsigned long resource)
{
    qemuDomainObjPrivatePtr priv = vm->privateData;
    int qemu_sock = -1;
    struct sockaddr_un sa_qemu;
    char *unixfile = NULL;
    int ret = -1;
    qemuMigrationSpec spec;

    VIR_DEBUG("driver=%p, vm=%p, st=%p, cookiein=%s, cookieinlen=%d, "
              "cookieout=%p, cookieoutlen=%p, flags=%lx, resource=%lu",
              driver, vm, st, NULLSTR(cookiein), cookieinlen,
              cookieout, cookieoutlen, flags, resource);

    if (!qemuCapsGet(priv->qemuCaps, QEMU_CAPS_MIGRATE_QEMU_UNIX) &&
        !qemuCapsGet(priv->qemuCaps, QEMU_CAPS_MIGRATE_QEMU_EXEC)) {
        qemuReportError(VIR_ERR_OPERATION_FAILED,
                        "%s", _("Source qemu is too old to support tunnelled migration"));
        goto cleanup;
    }

    if (virAsprintf(&unixfile, "%s/qemu.tunnelmigrate.src.%s",
                    driver->libDir, vm->def->name) < 0) {
        virReportOOMError();
        goto cleanup;
    }

    qemu_sock = socket(AF_UNIX, SOCK_STREAM, 0);
    if (qemu_sock < 0) {
        virReportSystemError(errno, "%s",
                             _("cannot open tunnelled migration socket"));
        goto cleanup;
    }
    memset(&sa_qemu, 0, sizeof(sa_qemu));
    sa_qemu.sun_family = AF_UNIX;
    if (virStrcpy(sa_qemu.sun_path, unixfile,
                  sizeof(sa_qemu.sun_path)) == NULL) {
        qemuReportError(VIR_ERR_INTERNAL_ERROR,
                        _("Unix socket '%s' too big for destination"),
                        unixfile);
        goto cleanup;
    }
    unlink(unixfile);
    if (bind(qemu_sock, (struct sockaddr *)&sa_qemu, sizeof(sa_qemu)) < 0) {
        virReportSystemError(errno,
                             _("Cannot bind to unix socket '%s' for tunnelled migration"),
                             unixfile);
        goto cleanup;
    }
    if (listen(qemu_sock, 1) < 0) {
        virReportSystemError(errno,
                             _("Cannot listen on unix socket '%s' for tunnelled migration"),
                             unixfile);
        goto cleanup;
    }

    if (chown(unixfile, driver->user, driver->group) < 0) {
        virReportSystemError(errno,
                             _("Cannot change unix socket '%s' owner"),
                             unixfile);
        goto cleanup;
    }

    spec.destType = MIGRATION_DEST_UNIX;
    spec.dest.unix_socket.file = unixfile;
    spec.dest.unix_socket.sock = qemu_sock;
    spec.fwdType = MIGRATION_FWD_STREAM;
    spec.fwd.stream = st;

    ret = qemuMigrationRun(driver, vm, cookiein, cookieinlen, cookieout,
                           cookieoutlen, flags, resource, &spec);
1698

1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709
cleanup:
    VIR_FORCE_CLOSE(qemu_sock);
    if (unixfile) {
        unlink(unixfile);
        VIR_FREE(unixfile);
    }

    return ret;
}


1710 1711 1712 1713 1714 1715
/* This is essentially a re-impl of virDomainMigrateVersion2
 * from libvirt.c, but running in source libvirtd context,
 * instead of client app context & also adding in tunnel
 * handling */
static int doPeer2PeerMigrate2(struct qemud_driver *driver,
                               virConnectPtr sconn,
1716 1717
                               virConnectPtr dconn,
                               virDomainObjPtr vm,
1718
                               const char *dconnuri,
1719 1720 1721
                               unsigned long flags,
                               const char *dname,
                               unsigned long resource)
1722 1723 1724
{
    virDomainPtr ddomain = NULL;
    char *uri_out = NULL;
1725
    char *cookie = NULL;
1726 1727 1728 1729 1730
    char *dom_xml = NULL;
    int cookielen = 0, ret;
    virErrorPtr orig_err = NULL;
    int cancelled;
    virStreamPtr st = NULL;
1731
    VIR_DEBUG("driver=%p, sconn=%p, dconn=%p, vm=%p, dconnuri=%s, "
1732
              "flags=%lx, dname=%s, resource=%lu",
1733 1734
              driver, sconn, dconn, vm, NULLSTR(dconnuri),
              flags, NULLSTR(dname), resource);
1735

1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770
    /* In version 2 of the protocol, the prepare step is slightly
     * different.  We fetch the domain XML of the source domain
     * and pass it to Prepare2.
     */
    if (!(dom_xml = qemuDomainFormatXML(driver, vm,
                                        VIR_DOMAIN_XML_SECURE |
                                        VIR_DOMAIN_XML_UPDATE_CPU)))
        return -1;

    if (virDomainObjGetState(vm, NULL) == VIR_DOMAIN_PAUSED)
        flags |= VIR_MIGRATE_PAUSED;

    VIR_DEBUG("Prepare2 %p", dconn);
    if (flags & VIR_MIGRATE_TUNNELLED) {
        /*
         * Tunnelled Migrate Version 2 does not support cookies
         * due to missing parameters in the prepareTunnel() API.
         */

        if (!(st = virStreamNew(dconn, 0)))
            goto cleanup;

        qemuDomainObjEnterRemoteWithDriver(driver, vm);
        ret = dconn->driver->domainMigratePrepareTunnel
            (dconn, st, flags, dname, resource, dom_xml);
        qemuDomainObjExitRemoteWithDriver(driver, vm);
    } else {
        qemuDomainObjEnterRemoteWithDriver(driver, vm);
        ret = dconn->driver->domainMigratePrepare2
            (dconn, &cookie, &cookielen, NULL, &uri_out,
             flags, dname, resource, dom_xml);
        qemuDomainObjExitRemoteWithDriver(driver, vm);
    }
    VIR_FREE(dom_xml);
    if (ret == -1)
1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781
        goto cleanup;

    /* the domain may have shutdown or crashed while we had the locks dropped
     * in qemuDomainObjEnterRemoteWithDriver, so check again
     */
    if (!virDomainObjIsActive(vm)) {
        qemuReportError(VIR_ERR_INTERNAL_ERROR, "%s",
                        _("guest unexpectedly quit"));
        goto cleanup;
    }

1782 1783 1784
    if (!(flags & VIR_MIGRATE_TUNNELLED) &&
        (uri_out == NULL)) {
        qemuReportError(VIR_ERR_INTERNAL_ERROR,
1785
                        _("domainMigratePrepare2 did not set uri"));
1786 1787
        cancelled = 1;
        goto finish;
1788 1789
    }

1790 1791 1792 1793
    /* Perform the migration.  The driver isn't supposed to return
     * until the migration is complete.
     */
    VIR_DEBUG("Perform %p", sconn);
1794
    qemuMigrationJobSetPhase(driver, vm, QEMU_MIGRATION_PHASE_PERFORM2);
1795
    if (flags & VIR_MIGRATE_TUNNELLED)
1796 1797 1798
        ret = doTunnelMigrate(driver, vm, st,
                              NULL, 0, NULL, NULL,
                              flags, resource);
1799 1800 1801 1802
    else
        ret = doNativeMigrate(driver, vm, uri_out,
                              cookie, cookielen,
                              NULL, NULL, /* No out cookie with v2 migration */
1803
                              flags, resource);
1804 1805 1806 1807

    /* Perform failed. Make sure Finish doesn't overwrite the error */
    if (ret < 0)
        orig_err = virSaveLastError();
1808

1809 1810 1811 1812
    /* If Perform returns < 0, then we need to cancel the VM
     * startup on the destination
     */
    cancelled = ret < 0 ? 1 : 0;
1813 1814

finish:
1815 1816 1817 1818
    /* In version 2 of the migration protocol, we pass the
     * status code from the sender to the destination host,
     * so it can do any cleanup if the migration failed.
     */
1819
    dname = dname ? dname : vm->def->name;
1820
    VIR_DEBUG("Finish2 %p ret=%d", dconn, ret);
1821 1822
    qemuDomainObjEnterRemoteWithDriver(driver, vm);
    ddomain = dconn->driver->domainMigrateFinish2
1823
        (dconn, dname, cookie, cookielen,
1824
         uri_out ? uri_out : dconnuri, flags, cancelled);
1825 1826
    qemuDomainObjExitRemoteWithDriver(driver, vm);

1827 1828
cleanup:
    if (ddomain) {
1829
        virUnrefDomain(ddomain);
1830 1831 1832 1833
        ret = 0;
    } else {
        ret = -1;
    }
1834

1835 1836 1837 1838 1839 1840 1841 1842
    if (st)
        virUnrefStream(st);

    if (orig_err) {
        virSetError(orig_err);
        virFreeError(orig_err);
    }
    VIR_FREE(uri_out);
1843
    VIR_FREE(cookie);
1844 1845

    return ret;
1846 1847 1848
}


1849 1850 1851 1852 1853 1854 1855 1856
/* This is essentially a re-impl of virDomainMigrateVersion3
 * from libvirt.c, but running in source libvirtd context,
 * instead of client app context & also adding in tunnel
 * handling */
static int doPeer2PeerMigrate3(struct qemud_driver *driver,
                               virConnectPtr sconn,
                               virConnectPtr dconn,
                               virDomainObjPtr vm,
1857
                               const char *xmlin,
1858
                               const char *dconnuri,
1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874
                               const char *uri,
                               unsigned long flags,
                               const char *dname,
                               unsigned long resource)
{
    virDomainPtr ddomain = NULL;
    char *uri_out = NULL;
    char *cookiein = NULL;
    char *cookieout = NULL;
    char *dom_xml = NULL;
    int cookieinlen = 0;
    int cookieoutlen = 0;
    int ret = -1;
    virErrorPtr orig_err = NULL;
    int cancelled;
    virStreamPtr st = NULL;
1875
    VIR_DEBUG("driver=%p, sconn=%p, dconn=%p, vm=%p, xmlin=%s, "
1876
              "dconnuri=%s, uri=%s, flags=%lx, dname=%s, resource=%lu",
1877 1878 1879
              driver, sconn, dconn, vm, NULLSTR(xmlin),
              NULLSTR(dconnuri), NULLSTR(uri), flags,
              NULLSTR(dname), resource);
1880

1881 1882 1883 1884 1885
    /* Unlike the virDomainMigrateVersion3 counterpart, we don't need
     * to worry about auto-setting the VIR_MIGRATE_CHANGE_PROTECTION
     * bit here, because we are already running inside the context of
     * a single job.  */

1886
    dom_xml = qemuMigrationBegin(driver, vm, xmlin,
1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912
                                 &cookieout, &cookieoutlen);
    if (!dom_xml)
        goto cleanup;

    if (virDomainObjGetState(vm, NULL) == VIR_DOMAIN_PAUSED)
        flags |= VIR_MIGRATE_PAUSED;

    VIR_DEBUG("Prepare3 %p", dconn);
    cookiein = cookieout;
    cookieinlen = cookieoutlen;
    cookieout = NULL;
    cookieoutlen = 0;
    if (flags & VIR_MIGRATE_TUNNELLED) {
        if (!(st = virStreamNew(dconn, 0)))
            goto cleanup;

        qemuDomainObjEnterRemoteWithDriver(driver, vm);
        ret = dconn->driver->domainMigratePrepareTunnel3
            (dconn, st, cookiein, cookieinlen,
             &cookieout, &cookieoutlen,
             flags, dname, resource, dom_xml);
        qemuDomainObjExitRemoteWithDriver(driver, vm);
    } else {
        qemuDomainObjEnterRemoteWithDriver(driver, vm);
        ret = dconn->driver->domainMigratePrepare3
            (dconn, cookiein, cookieinlen, &cookieout, &cookieoutlen,
1913
             uri, &uri_out, flags, dname, resource, dom_xml);
1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932
        qemuDomainObjExitRemoteWithDriver(driver, vm);
    }
    VIR_FREE(dom_xml);
    if (ret == -1)
        goto cleanup;

    if (!(flags & VIR_MIGRATE_TUNNELLED) &&
        (uri_out == NULL)) {
        qemuReportError(VIR_ERR_INTERNAL_ERROR,
                        _("domainMigratePrepare3 did not set uri"));
        cancelled = 1;
        goto finish;
    }

    /* Perform the migration.  The driver isn't supposed to return
     * until the migration is complete. The src VM should remain
     * running, but in paused state until the destination can
     * confirm migration completion.
     */
1933
    VIR_DEBUG("Perform3 %p uri=%s uri_out=%s", sconn, uri, uri_out);
1934
    qemuMigrationJobSetPhase(driver, vm, QEMU_MIGRATION_PHASE_PERFORM3);
1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948
    VIR_FREE(cookiein);
    cookiein = cookieout;
    cookieinlen = cookieoutlen;
    cookieout = NULL;
    cookieoutlen = 0;
    if (flags & VIR_MIGRATE_TUNNELLED)
        ret = doTunnelMigrate(driver, vm, st,
                              cookiein, cookieinlen,
                              &cookieout, &cookieoutlen,
                              flags, resource);
    else
        ret = doNativeMigrate(driver, vm, uri_out,
                              cookiein, cookieinlen,
                              &cookieout, &cookieoutlen,
1949
                              flags, resource);
1950 1951

    /* Perform failed. Make sure Finish doesn't overwrite the error */
1952
    if (ret < 0) {
1953
        orig_err = virSaveLastError();
1954 1955 1956 1957
    } else {
        qemuMigrationJobSetPhase(driver, vm,
                                 QEMU_MIGRATION_PHASE_PERFORM3_DONE);
    }
1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978

    /* If Perform returns < 0, then we need to cancel the VM
     * startup on the destination
     */
    cancelled = ret < 0 ? 1 : 0;

finish:
    /*
     * The status code from the source is passed to the destination.
     * The dest can cleanup in the source indicated it failed to
     * send all migration data. Returns NULL for ddomain if
     * the dest was unable to complete migration.
     */
    VIR_DEBUG("Finish3 %p ret=%d", dconn, ret);
    VIR_FREE(cookiein);
    cookiein = cookieout;
    cookieinlen = cookieoutlen;
    cookieout = NULL;
    cookieoutlen = 0;
    dname = dname ? dname : vm->def->name;
    qemuDomainObjEnterRemoteWithDriver(driver, vm);
1979
    ddomain = dconn->driver->domainMigrateFinish3
1980
        (dconn, dname, cookiein, cookieinlen, &cookieout, &cookieoutlen,
1981
         dconnuri, uri_out ? uri_out : uri, flags, cancelled);
1982 1983
    qemuDomainObjExitRemoteWithDriver(driver, vm);

1984 1985 1986 1987 1988 1989 1990
    /* If ddomain is NULL, then we were unable to start
     * the guest on the target, and must restart on the
     * source. There is a small chance that the ddomain
     * is NULL due to an RPC failure, in which case
     * ddomain could in fact be running on the dest.
     * The lock manager plugins should take care of
     * safety in this scenario.
1991
     */
1992
    cancelled = ddomain == NULL ? 1 : 0;
1993

1994 1995 1996 1997 1998 1999
    /* If finish3 set an error, and we don't have an earlier
     * one we need to preserve it in case confirm3 overwrites
     */
    if (!orig_err)
        orig_err = virSaveLastError();

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
    /*
     * If cancelled, then src VM will be restarted, else
     * it will be killed
     */
    VIR_DEBUG("Confirm3 %p ret=%d vm=%p", sconn, ret, vm);
    VIR_FREE(cookiein);
    cookiein = cookieout;
    cookieinlen = cookieoutlen;
    cookieout = NULL;
    cookieoutlen = 0;
    ret = qemuMigrationConfirm(driver, sconn, vm,
                               cookiein, cookieinlen,
2012
                               flags, cancelled);
2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040
    /* If Confirm3 returns -1, there's nothing more we can
     * do, but fortunately worst case is that there is a
     * domain left in 'paused' state on source.
     */

 cleanup:
    if (ddomain) {
        virUnrefDomain(ddomain);
        ret = 0;
    } else {
        ret = -1;
    }

    if (st)
        virUnrefStream(st);

    if (orig_err) {
        virSetError(orig_err);
        virFreeError(orig_err);
    }
    VIR_FREE(uri_out);
    VIR_FREE(cookiein);
    VIR_FREE(cookieout);

    return ret;
}


2041
static int doPeer2PeerMigrate(struct qemud_driver *driver,
2042
                              virConnectPtr sconn,
2043
                              virDomainObjPtr vm,
2044
                              const char *xmlin,
2045
                              const char *dconnuri,
2046 2047 2048
                              const char *uri,
                              unsigned long flags,
                              const char *dname,
2049 2050
                              unsigned long resource,
                              bool *v3proto)
2051 2052 2053 2054
{
    int ret = -1;
    virConnectPtr dconn = NULL;
    bool p2p;
2055 2056
    virErrorPtr orig_err = NULL;

2057
    VIR_DEBUG("driver=%p, sconn=%p, vm=%p, xmlin=%s, dconnuri=%s, "
2058
              "uri=%s, flags=%lx, dname=%s, resource=%lu",
2059 2060
              driver, sconn, vm, NULLSTR(xmlin), NULLSTR(dconnuri),
              NULLSTR(uri), flags, NULLSTR(dname), resource);
2061 2062 2063 2064 2065 2066

    /* the order of operations is important here; we make sure the
     * destination side is completely setup before we touch the source
     */

    qemuDomainObjEnterRemoteWithDriver(driver, vm);
2067
    dconn = virConnectOpen(dconnuri);
2068 2069 2070 2071 2072 2073 2074 2075 2076 2077
    qemuDomainObjExitRemoteWithDriver(driver, vm);
    if (dconn == NULL) {
        qemuReportError(VIR_ERR_OPERATION_FAILED,
                        _("Failed to connect to remote libvirt URI %s"), uri);
        return -1;
    }

    qemuDomainObjEnterRemoteWithDriver(driver, vm);
    p2p = VIR_DRV_SUPPORTS_FEATURE(dconn->driver, dconn,
                                   VIR_DRV_FEATURE_MIGRATION_P2P);
2078
        /* v3proto reflects whether the caller used Perform3, but with
2079
         * p2p migrate, regardless of whether Perform2 or Perform3
2080 2081 2082 2083
         * were used, we decide protocol based on what target supports
         */
    *v3proto = VIR_DRV_SUPPORTS_FEATURE(dconn->driver, dconn,
                                        VIR_DRV_FEATURE_MIGRATION_V3);
2084
    qemuDomainObjExitRemoteWithDriver(driver, vm);
2085

2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098
    if (!p2p) {
        qemuReportError(VIR_ERR_OPERATION_FAILED, "%s",
                        _("Destination libvirt does not support peer-to-peer migration protocol"));
        goto cleanup;
    }

    /* domain may have been stopped while we were talking to remote daemon */
    if (!virDomainObjIsActive(vm)) {
        qemuReportError(VIR_ERR_INTERNAL_ERROR, "%s",
                        _("guest unexpectedly quit"));
        goto cleanup;
    }

2099 2100 2101 2102 2103 2104 2105
    /* Change protection is only required on the source side (us), and
     * only for v3 migration when begin and perform are separate jobs.
     * But peer-2-peer is already a single job, and we still want to
     * talk to older destinations that would reject the flag.
     * Therefore it is safe to clear the bit here.  */
    flags &= ~VIR_MIGRATE_CHANGE_PROTECTION;

2106
    if (*v3proto)
2107
        ret = doPeer2PeerMigrate3(driver, sconn, dconn, vm, xmlin,
2108
                                  dconnuri, uri, flags, dname, resource);
2109 2110
    else
        ret = doPeer2PeerMigrate2(driver, sconn, dconn, vm,
2111
                                  dconnuri, flags, dname, resource);
2112 2113

cleanup:
2114
    orig_err = virSaveLastError();
2115
    qemuDomainObjEnterRemoteWithDriver(driver, vm);
2116
    virConnectClose(dconn);
2117
    qemuDomainObjExitRemoteWithDriver(driver, vm);
2118 2119 2120 2121
    if (orig_err) {
        virSetError(orig_err);
        virFreeError(orig_err);
    }
2122 2123 2124 2125 2126

    return ret;
}


2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146
/*
 * This implements perform part of the migration protocol when migration job
 * does not need to be active across several APIs, i.e., peer2peer migration or
 * perform phase of v2 non-peer2peer migration.
 */
static int
qemuMigrationPerformJob(struct qemud_driver *driver,
                        virConnectPtr conn,
                        virDomainObjPtr vm,
                        const char *xmlin,
                        const char *dconnuri,
                        const char *uri,
                        const char *cookiein,
                        int cookieinlen,
                        char **cookieout,
                        int *cookieoutlen,
                        unsigned long flags,
                        const char *dname,
                        unsigned long resource,
                        bool v3proto)
2147 2148 2149 2150 2151
{
    virDomainEventPtr event = NULL;
    int ret = -1;
    int resume = 0;

2152
    if (qemuMigrationJobStart(driver, vm, QEMU_ASYNC_JOB_MIGRATION_OUT) < 0)
2153 2154 2155 2156 2157 2158 2159 2160
        goto cleanup;

    if (!virDomainObjIsActive(vm)) {
        qemuReportError(VIR_ERR_OPERATION_INVALID,
                        "%s", _("domain is not running"));
        goto endjob;
    }

2161 2162 2163 2164 2165 2166
    if (qemuProcessAutoDestroyActive(driver, vm)) {
        qemuReportError(VIR_ERR_OPERATION_INVALID,
                        "%s", _("domain is marked for auto destroy"));
        goto endjob;
    }

J
Jiri Denemark 已提交
2167
    resume = virDomainObjGetState(vm, NULL) == VIR_DOMAIN_RUNNING;
2168 2169

    if ((flags & (VIR_MIGRATE_TUNNELLED | VIR_MIGRATE_PEER2PEER))) {
2170 2171 2172
        ret = doPeer2PeerMigrate(driver, conn, vm, xmlin,
                                 dconnuri, uri, flags, dname,
                                 resource, &v3proto);
2173
    } else {
2174 2175 2176
        qemuMigrationJobSetPhase(driver, vm, QEMU_MIGRATION_PHASE_PERFORM2);
        ret = doNativeMigrate(driver, vm, uri, cookiein, cookieinlen,
                              cookieout, cookieoutlen,
2177
                              flags, resource);
2178
    }
2179 2180
    if (ret < 0)
        goto endjob;
2181

2182 2183 2184 2185
    /*
     * In v3 protocol, the source VM is not killed off until the
     * confirm step.
     */
2186
    if (!v3proto) {
2187
        qemuProcessStop(driver, vm, 1, VIR_DOMAIN_SHUTOFF_MIGRATED);
2188
        virDomainAuditStop(vm, "migrated");
2189 2190 2191
        event = virDomainEventNewFromObj(vm,
                                         VIR_DOMAIN_EVENT_STOPPED,
                                         VIR_DOMAIN_EVENT_STOPPED_MIGRATED);
2192
    }
2193
    resume = 0;
2194 2195

endjob:
J
Jiri Denemark 已提交
2196
    if (resume && virDomainObjGetState(vm, NULL) == VIR_DOMAIN_PAUSED) {
2197
        /* we got here through some sort of failure; start the domain again */
J
Jiri Denemark 已提交
2198
        if (qemuProcessStartCPUs(driver, vm, conn,
2199 2200
                                 VIR_DOMAIN_RUNNING_MIGRATION_CANCELED,
                                 QEMU_ASYNC_JOB_MIGRATION_OUT) < 0) {
2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212
            /* Hm, we already know we are in error here.  We don't want to
             * overwrite the previous error, though, so we just throw something
             * to the logs and hope for the best
             */
            VIR_ERROR(_("Failed to resume guest %s after failure"),
                      vm->def->name);
        }

        event = virDomainEventNewFromObj(vm,
                                         VIR_DOMAIN_EVENT_RESUMED,
                                         VIR_DOMAIN_EVENT_RESUMED_MIGRATED);
    }
2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265

    if (qemuMigrationJobFinish(driver, vm) == 0) {
        vm = NULL;
    } else if (!virDomainObjIsActive(vm) &&
               (!vm->persistent ||
                (ret == 0 && (flags & VIR_MIGRATE_UNDEFINE_SOURCE)))) {
        if (flags & VIR_MIGRATE_UNDEFINE_SOURCE)
            virDomainDeleteConfig(driver->configDir, driver->autostartDir, vm);
        virDomainRemoveInactive(&driver->domains, vm);
        vm = NULL;
    }

cleanup:
    if (vm)
        virDomainObjUnlock(vm);
    if (event)
        qemuDomainEventQueue(driver, event);
    return ret;
}

/*
 * This implements perform phase of v3 migration protocol.
 */
static int
qemuMigrationPerformPhase(struct qemud_driver *driver,
                          virConnectPtr conn,
                          virDomainObjPtr vm,
                          const char *uri,
                          const char *cookiein,
                          int cookieinlen,
                          char **cookieout,
                          int *cookieoutlen,
                          unsigned long flags,
                          unsigned long resource)
{
    virDomainEventPtr event = NULL;
    int ret = -1;
    bool resume;
    int refs;

    /* If we didn't start the job in the begin phase, start it now. */
    if (!(flags & VIR_MIGRATE_CHANGE_PROTECTION)) {
        if (qemuMigrationJobStart(driver, vm, QEMU_ASYNC_JOB_MIGRATION_OUT) < 0)
            goto cleanup;
    } else if (!qemuMigrationJobIsActive(vm, QEMU_ASYNC_JOB_MIGRATION_OUT)) {
        goto cleanup;
    }

    qemuMigrationJobStartPhase(driver, vm, QEMU_MIGRATION_PHASE_PERFORM3);

    resume = virDomainObjGetState(vm, NULL) == VIR_DOMAIN_RUNNING;
    ret = doNativeMigrate(driver, vm, uri, cookiein, cookieinlen,
                          cookieout, cookieoutlen,
2266
                          flags, resource);
2267 2268 2269 2270 2271

    if (ret < 0 && resume &&
        virDomainObjGetState(vm, NULL) == VIR_DOMAIN_PAUSED) {
        /* we got here through some sort of failure; start the domain again */
        if (qemuProcessStartCPUs(driver, vm, conn,
2272 2273
                                 VIR_DOMAIN_RUNNING_MIGRATION_CANCELED,
                                 QEMU_ASYNC_JOB_MIGRATION_OUT) < 0) {
2274 2275 2276 2277 2278 2279
            /* Hm, we already know we are in error here.  We don't want to
             * overwrite the previous error, though, so we just throw something
             * to the logs and hope for the best
             */
            VIR_ERROR(_("Failed to resume guest %s after failure"),
                      vm->def->name);
2280
        }
2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301

        event = virDomainEventNewFromObj(vm,
                                         VIR_DOMAIN_EVENT_RESUMED,
                                         VIR_DOMAIN_EVENT_RESUMED_MIGRATED);
    }

    if (ret < 0)
        goto endjob;

    qemuMigrationJobSetPhase(driver, vm, QEMU_MIGRATION_PHASE_PERFORM3_DONE);

endjob:
    if (ret < 0)
        refs = qemuMigrationJobFinish(driver, vm);
    else
        refs = qemuMigrationJobContinue(vm);
    if (refs == 0) {
        vm = NULL;
    } else if (!virDomainObjIsActive(vm) && !vm->persistent) {
        virDomainRemoveInactive(&driver->domains, vm);
        vm = NULL;
2302
    }
2303 2304 2305 2306 2307 2308 2309 2310 2311

cleanup:
    if (vm)
        virDomainObjUnlock(vm);
    if (event)
        qemuDomainEventQueue(driver, event);
    return ret;
}

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
int
qemuMigrationPerform(struct qemud_driver *driver,
                     virConnectPtr conn,
                     virDomainObjPtr vm,
                     const char *xmlin,
                     const char *dconnuri,
                     const char *uri,
                     const char *cookiein,
                     int cookieinlen,
                     char **cookieout,
                     int *cookieoutlen,
                     unsigned long flags,
                     const char *dname,
                     unsigned long resource,
                     bool v3proto)
{
    VIR_DEBUG("driver=%p, conn=%p, vm=%p, xmlin=%s, dconnuri=%s, "
              "uri=%s, cookiein=%s, cookieinlen=%d, cookieout=%p, "
              "cookieoutlen=%p, flags=%lx, dname=%s, resource=%lu, v3proto=%d",
              driver, conn, vm, NULLSTR(xmlin), NULLSTR(dconnuri),
              NULLSTR(uri), NULLSTR(cookiein), cookieinlen,
              cookieout, cookieoutlen, flags, NULLSTR(dname),
              resource, v3proto);

    if ((flags & (VIR_MIGRATE_TUNNELLED | VIR_MIGRATE_PEER2PEER))) {
        if (cookieinlen) {
            qemuReportError(VIR_ERR_OPERATION_INVALID,
                            "%s", _("received unexpected cookie with P2P migration"));
            return -1;
        }

        return qemuMigrationPerformJob(driver, conn, vm, xmlin, dconnuri, uri,
                                       cookiein, cookieinlen, cookieout,
                                       cookieoutlen, flags, dname, resource,
                                       v3proto);
    } else {
        if (dconnuri) {
            qemuReportError(VIR_ERR_INTERNAL_ERROR,
                            "%s", _("Unexpected dconnuri parameter with non-peer2peer migration"));
            return -1;
        }

        if (v3proto) {
            return qemuMigrationPerformPhase(driver, conn, vm, uri,
                                             cookiein, cookieinlen,
                                             cookieout, cookieoutlen,
2358
                                             flags, resource);
2359 2360 2361 2362 2363 2364 2365 2366
        } else {
            return qemuMigrationPerformJob(driver, conn, vm, xmlin, dconnuri,
                                           uri, cookiein, cookieinlen,
                                           cookieout, cookieoutlen, flags,
                                           dname, resource, v3proto);
        }
    }
}
2367 2368 2369 2370 2371 2372 2373 2374 2375 2376

#if WITH_MACVTAP
static void
qemuMigrationVPAssociatePortProfiles(virDomainDefPtr def) {
    int i;
    int last_good_net = -1;
    virDomainNetDefPtr net;

    for (i = 0; i < def->nnets; i++) {
        net = def->nets[i];
2377
        if (virDomainNetGetActualType(net) == VIR_DOMAIN_NET_TYPE_DIRECT) {
2378 2379
            if (vpAssociatePortProfileId(net->ifname,
                                         net->mac,
2380 2381
                                         virDomainNetGetActualDirectDev(net),
                                         virDomainNetGetActualDirectVirtPortProfile(net),
2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393
                                         def->uuid,
                                         VIR_VM_OP_MIGRATE_IN_FINISH) != 0)
                goto err_exit;
        }
        last_good_net = i;
    }

    return;

err_exit:
    for (i = 0; i < last_good_net; i++) {
        net = def->nets[i];
2394
        if (virDomainNetGetActualType(net) == VIR_DOMAIN_NET_TYPE_DIRECT) {
2395 2396
            vpDisassociatePortProfileId(net->ifname,
                                        net->mac,
2397 2398
                                        virDomainNetGetActualDirectDev(net),
                                        virDomainNetGetActualDirectVirtPortProfile(net),
2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412
                                        VIR_VM_OP_MIGRATE_IN_FINISH);
        }
    }
}
#else /* !WITH_MACVTAP */
static void
qemuMigrationVPAssociatePortProfiles(virDomainDefPtr def ATTRIBUTE_UNUSED) { }
#endif /* WITH_MACVTAP */


virDomainPtr
qemuMigrationFinish(struct qemud_driver *driver,
                    virConnectPtr dconn,
                    virDomainObjPtr vm,
2413 2414 2415 2416
                    const char *cookiein,
                    int cookieinlen,
                    char **cookieout,
                    int *cookieoutlen,
2417
                    unsigned long flags,
2418 2419
                    int retcode,
                    bool v3proto)
2420 2421 2422 2423
{
    virDomainPtr dom = NULL;
    virDomainEventPtr event = NULL;
    int newVM = 1;
2424
    qemuMigrationCookiePtr mig = NULL;
2425 2426
    virErrorPtr orig_err = NULL;

2427
    VIR_DEBUG("driver=%p, dconn=%p, vm=%p, cookiein=%s, cookieinlen=%d, "
2428
              "cookieout=%p, cookieoutlen=%p, flags=%lx, retcode=%d",
2429 2430
              driver, dconn, vm, NULLSTR(cookiein), cookieinlen,
              cookieout, cookieoutlen, flags, retcode);
2431

2432
    if (!qemuMigrationJobIsActive(vm, QEMU_ASYNC_JOB_MIGRATION_IN))
2433 2434
        goto cleanup;

2435 2436 2437
    qemuMigrationJobStartPhase(driver, vm,
                               v3proto ? QEMU_MIGRATION_PHASE_FINISH3
                                       : QEMU_MIGRATION_PHASE_FINISH2);
2438

2439 2440
    if (!(mig = qemuMigrationEatCookie(driver, vm, cookiein, cookieinlen, 0)))
        goto endjob;
2441 2442 2443 2444 2445 2446 2447 2448

    /* Did the migration go as planned?  If yes, return the domain
     * object, but if no, clean up the empty qemu process.
     */
    if (retcode == 0) {
        if (!virDomainObjIsActive(vm)) {
            qemuReportError(VIR_ERR_INTERNAL_ERROR, "%s",
                            _("guest unexpectedly quit"));
2449
            goto endjob;
2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468
        }

        qemuMigrationVPAssociatePortProfiles(vm->def);

        if (flags & VIR_MIGRATE_PERSIST_DEST) {
            if (vm->persistent)
                newVM = 0;
            vm->persistent = 1;

            if (virDomainSaveConfig(driver->configDir, vm->def) < 0) {
                /* Hmpf.  Migration was successful, but making it persistent
                 * was not.  If we report successful, then when this domain
                 * shuts down, management tools are in for a surprise.  On the
                 * other hand, if we report failure, then the management tools
                 * might try to restart the domain on the source side, even
                 * though the domain is actually running on the destination.
                 * Return a NULL dom pointer, and hope that this is a rare
                 * situation and management tools are smart.
                 */
2469 2470

                /*
2471 2472
                 * However, in v3 protocol, the source VM is still available
                 * to restart during confirm() step, so we kill it off now.
2473 2474 2475
                 */
                if (v3proto) {
                    qemuProcessStop(driver, vm, 1, VIR_DOMAIN_SHUTOFF_FAILED);
2476
                    virDomainAuditStop(vm, "failed");
2477 2478
                    if (newVM)
                        vm->persistent = 0;
2479
                }
2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497
                goto endjob;
            }

            event = virDomainEventNewFromObj(vm,
                                             VIR_DOMAIN_EVENT_DEFINED,
                                             newVM ?
                                             VIR_DOMAIN_EVENT_DEFINED_ADDED :
                                             VIR_DOMAIN_EVENT_DEFINED_UPDATED);
            if (event)
                qemuDomainEventQueue(driver, event);
            event = NULL;
        }

        if (!(flags & VIR_MIGRATE_PAUSED)) {
            /* run 'cont' on the destination, which allows migration on qemu
             * >= 0.10.6 to work properly.  This isn't strictly necessary on
             * older qemu's, but it also doesn't hurt anything there
             */
J
Jiri Denemark 已提交
2498
            if (qemuProcessStartCPUs(driver, vm, dconn,
2499 2500
                                     VIR_DOMAIN_RUNNING_MIGRATED,
                                     QEMU_ASYNC_JOB_MIGRATION_IN) < 0) {
2501 2502 2503 2504 2505 2506 2507 2508
                if (virGetLastError() == NULL)
                    qemuReportError(VIR_ERR_INTERNAL_ERROR,
                                    "%s", _("resume operation failed"));
                /* Need to save the current error, in case shutting
                 * down the process overwrites it
                 */
                orig_err = virSaveLastError();

2509 2510 2511 2512 2513 2514 2515 2516 2517 2518
                /*
                 * In v3 protocol, the source VM is still available to
                 * restart during confirm() step, so we kill it off
                 * now.
                 * In v2 protocol, the source is dead, so we leave
                 * target in paused state, in case admin can fix
                 * things up
                 */
                if (v3proto) {
                    qemuProcessStop(driver, vm, 1, VIR_DOMAIN_SHUTOFF_FAILED);
2519
                    virDomainAuditStop(vm, "failed");
2520 2521 2522 2523
                    event = virDomainEventNewFromObj(vm,
                                                     VIR_DOMAIN_EVENT_STOPPED,
                                                     VIR_DOMAIN_EVENT_STOPPED_FAILED);
                }
2524 2525 2526 2527
                goto endjob;
            }
        }

2528 2529
        dom = virGetDomain (dconn, vm->def->name, vm->def->uuid);

2530 2531 2532
        event = virDomainEventNewFromObj(vm,
                                         VIR_DOMAIN_EVENT_RESUMED,
                                         VIR_DOMAIN_EVENT_RESUMED_MIGRATED);
J
Jiri Denemark 已提交
2533 2534
        if (virDomainObjGetState(vm, NULL) == VIR_DOMAIN_PAUSED) {
            virDomainObjSetState(vm, VIR_DOMAIN_PAUSED, VIR_DOMAIN_PAUSED_USER);
2535 2536
            if (event)
                qemuDomainEventQueue(driver, event);
2537 2538 2539 2540 2541 2542 2543 2544
            event = virDomainEventNewFromObj(vm,
                                             VIR_DOMAIN_EVENT_SUSPENDED,
                                             VIR_DOMAIN_EVENT_SUSPENDED_PAUSED);
        }
        if (virDomainSaveStatus(driver->caps, driver->stateDir, vm) < 0) {
            VIR_WARN("Failed to save status on vm %s", vm->def->name);
            goto endjob;
        }
2545 2546 2547

        /* Guest is successfully running, so cancel previous auto destroy */
        qemuProcessAutoDestroyRemove(driver, vm);
2548
    } else {
J
Jiri Denemark 已提交
2549
        qemuProcessStop(driver, vm, 1, VIR_DOMAIN_SHUTOFF_FAILED);
2550
        virDomainAuditStop(vm, "failed");
2551 2552 2553 2554 2555
        event = virDomainEventNewFromObj(vm,
                                         VIR_DOMAIN_EVENT_STOPPED,
                                         VIR_DOMAIN_EVENT_STOPPED_FAILED);
    }

2556 2557 2558
    if (qemuMigrationBakeCookie(mig, driver, vm, cookieout, cookieoutlen, 0) < 0)
        VIR_WARN("Unable to encode migration cookie");

2559
endjob:
E
Eric Blake 已提交
2560 2561 2562 2563 2564
    if (qemuMigrationJobFinish(driver, vm) == 0) {
        vm = NULL;
    } else if (!vm->persistent && !virDomainObjIsActive(vm)) {
        virDomainRemoveInactive(&driver->domains, vm);
        vm = NULL;
2565
    }
2566 2567 2568 2569 2570 2571

cleanup:
    if (vm)
        virDomainObjUnlock(vm);
    if (event)
        qemuDomainEventQueue(driver, event);
2572
    qemuMigrationCookieFree(mig);
2573 2574 2575 2576
    if (orig_err) {
        virSetError(orig_err);
        virFreeError(orig_err);
    }
2577 2578
    return dom;
}
2579

2580 2581 2582 2583 2584 2585

int qemuMigrationConfirm(struct qemud_driver *driver,
                         virConnectPtr conn,
                         virDomainObjPtr vm,
                         const char *cookiein,
                         int cookieinlen,
E
Eric Blake 已提交
2586
                         unsigned int flags,
2587
                         int retcode)
2588 2589 2590 2591
{
    qemuMigrationCookiePtr mig;
    virDomainEventPtr event = NULL;
    int rv = -1;
2592
    VIR_DEBUG("driver=%p, conn=%p, vm=%p, cookiein=%s, cookieinlen=%d, "
2593
              "flags=%x, retcode=%d",
2594 2595
              driver, conn, vm, NULLSTR(cookiein), cookieinlen,
              flags, retcode);
2596

2597
    virCheckFlags(QEMU_MIGRATION_FLAGS, -1);
E
Eric Blake 已提交
2598

2599 2600 2601 2602 2603
    qemuMigrationJobSetPhase(driver, vm,
                             retcode == 0
                             ? QEMU_MIGRATION_PHASE_CONFIRM3
                             : QEMU_MIGRATION_PHASE_CONFIRM3_CANCELLED);

2604
    if (!(mig = qemuMigrationEatCookie(driver, vm, cookiein, cookieinlen, 0)))
2605 2606 2607 2608 2609 2610 2611
        return -1;

    /* Did the migration go as planned?  If yes, kill off the
     * domain object, but if no, resume CPUs
     */
    if (retcode == 0) {
        qemuProcessStop(driver, vm, 1, VIR_DOMAIN_SHUTOFF_MIGRATED);
2612
        virDomainAuditStop(vm, "migrated");
2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623

        event = virDomainEventNewFromObj(vm,
                                         VIR_DOMAIN_EVENT_STOPPED,
                                         VIR_DOMAIN_EVENT_STOPPED_MIGRATED);
    } else {

        /* run 'cont' on the destination, which allows migration on qemu
         * >= 0.10.6 to work properly.  This isn't strictly necessary on
         * older qemu's, but it also doesn't hurt anything there
         */
        if (qemuProcessStartCPUs(driver, vm, conn,
2624 2625
                                 VIR_DOMAIN_RUNNING_MIGRATED,
                                 QEMU_ASYNC_JOB_MIGRATION_OUT) < 0) {
2626 2627 2628
            if (virGetLastError() == NULL)
                qemuReportError(VIR_ERR_INTERNAL_ERROR,
                                "%s", _("resume operation failed"));
2629
            goto cleanup;
2630 2631 2632 2633 2634 2635 2636
        }

        event = virDomainEventNewFromObj(vm,
                                         VIR_DOMAIN_EVENT_RESUMED,
                                         VIR_DOMAIN_EVENT_RESUMED_MIGRATED);
        if (virDomainSaveStatus(driver->caps, driver->stateDir, vm) < 0) {
            VIR_WARN("Failed to save status on vm %s", vm->def->name);
2637
            goto cleanup;
2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650
        }
    }

    qemuMigrationCookieFree(mig);
    rv = 0;

cleanup:
    if (event)
        qemuDomainEventQueue(driver, event);
    return rv;
}


2651 2652 2653 2654 2655
/* Helper function called while driver lock is held and vm is active.  */
int
qemuMigrationToFile(struct qemud_driver *driver, virDomainObjPtr vm,
                    int fd, off_t offset, const char *path,
                    const char *compressor,
2656 2657
                    bool is_reg, bool bypassSecurityDriver,
                    enum qemuDomainAsyncJob asyncJob)
2658 2659 2660 2661 2662 2663
{
    qemuDomainObjPrivatePtr priv = vm->privateData;
    virCgroupPtr cgroup = NULL;
    int ret = -1;
    int rc;
    bool restoreLabel = false;
2664 2665
    virCommandPtr cmd = NULL;
    int pipeFD[2] = { -1, -1 };
2666

2667
    if (qemuCapsGet(priv->qemuCaps, QEMU_CAPS_MIGRATE_QEMU_FD) &&
2668
        (!compressor || pipe(pipeFD) == 0)) {
2669
        /* All right! We can use fd migration, which means that qemu
2670 2671 2672
         * doesn't have to open() the file, so while we still have to
         * grant SELinux access, we can do it on fd and avoid cleanup
         * later, as well as skip futzing with cgroup.  */
2673 2674
        if (virSecurityManagerSetImageFDLabel(driver->securityManager, vm,
                                              compressor ? pipeFD[1] : fd) < 0)
2675
            goto cleanup;
2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693
        bypassSecurityDriver = true;
    } else {
        /* Phooey - we have to fall back on exec migration, where qemu
         * has to popen() the file by name.  We might also stumble on
         * a race present in some qemu versions where it does a wait()
         * that botches pclose.  */
        if (!is_reg &&
            qemuCgroupControllerActive(driver,
                                       VIR_CGROUP_CONTROLLER_DEVICES)) {
            if (virCgroupForDomain(driver->cgroup, vm->def->name,
                                   &cgroup, 0) != 0) {
                qemuReportError(VIR_ERR_INTERNAL_ERROR,
                                _("Unable to find cgroup for %s"),
                                vm->def->name);
                goto cleanup;
            }
            rc = virCgroupAllowDevicePath(cgroup, path,
                                          VIR_CGROUP_DEVICE_RW);
2694
            virDomainAuditCgroupPath(vm, cgroup, "allow", path, "rw", rc);
2695 2696 2697 2698 2699 2700
            if (rc < 0) {
                virReportSystemError(-rc,
                                     _("Unable to allow device %s for %s"),
                                     path, vm->def->name);
                goto cleanup;
            }
2701
        }
2702 2703 2704
        if ((!bypassSecurityDriver) &&
            virSecurityManagerSetSavedStateLabel(driver->securityManager,
                                                 vm, path) < 0)
2705
            goto cleanup;
2706
        restoreLabel = true;
2707 2708
    }

2709
    if (qemuDomainObjEnterMonitorAsync(driver, vm, asyncJob) < 0)
2710 2711
        goto cleanup;

2712 2713 2714
    if (!compressor) {
        const char *args[] = { "cat", NULL };

2715
        if (qemuCapsGet(priv->qemuCaps, QEMU_CAPS_MIGRATE_QEMU_FD) &&
2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731
            priv->monConfig->type == VIR_DOMAIN_CHR_TYPE_UNIX) {
            rc = qemuMonitorMigrateToFd(priv->mon,
                                        QEMU_MONITOR_MIGRATE_BACKGROUND,
                                        fd);
        } else {
            rc = qemuMonitorMigrateToFile(priv->mon,
                                          QEMU_MONITOR_MIGRATE_BACKGROUND,
                                          args, path, offset);
        }
    } else {
        const char *prog = compressor;
        const char *args[] = {
            prog,
            "-c",
            NULL
        };
2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750
        if (pipeFD[0] != -1) {
            cmd = virCommandNewArgs(args);
            virCommandSetInputFD(cmd, pipeFD[0]);
            virCommandSetOutputFD(cmd, &fd);
            if (virSetCloseExec(pipeFD[1]) < 0) {
                virReportSystemError(errno, "%s",
                                     _("Unable to set cloexec flag"));
                qemuDomainObjExitMonitorWithDriver(driver, vm);
                goto cleanup;
            }
            if (virCommandRunAsync(cmd, NULL) < 0) {
                qemuDomainObjExitMonitorWithDriver(driver, vm);
                goto cleanup;
            }
            rc = qemuMonitorMigrateToFd(priv->mon,
                                        QEMU_MONITOR_MIGRATE_BACKGROUND,
                                        pipeFD[1]);
            if (VIR_CLOSE(pipeFD[0]) < 0 ||
                VIR_CLOSE(pipeFD[1]) < 0)
2751
                VIR_WARN("failed to close intermediate pipe");
2752 2753 2754 2755 2756
        } else {
            rc = qemuMonitorMigrateToFile(priv->mon,
                                          QEMU_MONITOR_MIGRATE_BACKGROUND,
                                          args, path, offset);
        }
2757
    }
2758
    qemuDomainObjExitMonitorWithDriver(driver, vm);
2759 2760 2761 2762

    if (rc < 0)
        goto cleanup;

2763
    rc = qemuMigrationWaitForCompletion(driver, vm, asyncJob);
2764 2765 2766 2767

    if (rc < 0)
        goto cleanup;

2768 2769 2770
    if (cmd && virCommandWait(cmd, NULL) < 0)
        goto cleanup;

2771 2772 2773
    ret = 0;

cleanup:
2774 2775 2776
    VIR_FORCE_CLOSE(pipeFD[0]);
    VIR_FORCE_CLOSE(pipeFD[1]);
    virCommandFree(cmd);
2777 2778 2779 2780 2781 2782 2783 2784
    if (restoreLabel && (!bypassSecurityDriver) &&
        virSecurityManagerRestoreSavedStateLabel(driver->securityManager,
                                                 vm, path) < 0)
        VIR_WARN("failed to restore save state label on %s", path);

    if (cgroup != NULL) {
        rc = virCgroupDenyDevicePath(cgroup, path,
                                     VIR_CGROUP_DEVICE_RWM);
2785
        virDomainAuditCgroupPath(vm, cgroup, "deny", path, "rwm", rc);
2786 2787 2788 2789 2790 2791 2792
        if (rc < 0)
            VIR_WARN("Unable to deny device %s for %s %d",
                     path, vm->def->name, rc);
        virCgroupFree(&cgroup);
    }
    return ret;
}
2793 2794 2795 2796 2797 2798 2799 2800 2801 2802 2803

int
qemuMigrationJobStart(struct qemud_driver *driver,
                      virDomainObjPtr vm,
                      enum qemuDomainAsyncJob job)
{
    qemuDomainObjPrivatePtr priv = vm->privateData;

    if (qemuDomainObjBeginAsyncJobWithDriver(driver, vm, job) < 0)
        return -1;

2804
    if (job == QEMU_ASYNC_JOB_MIGRATION_IN) {
2805
        qemuDomainObjSetAsyncJobMask(vm, QEMU_JOB_NONE);
2806 2807
    } else {
        qemuDomainObjSetAsyncJobMask(vm, DEFAULT_JOB_MASK |
2808
                                     JOB_MASK(QEMU_JOB_SUSPEND) |
2809 2810
                                     JOB_MASK(QEMU_JOB_MIGRATION_OP));
    }
2811 2812 2813 2814 2815 2816 2817 2818 2819 2820 2821 2822 2823 2824 2825 2826 2827 2828 2829 2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846 2847 2848 2849 2850 2851 2852 2853 2854 2855 2856 2857 2858 2859 2860 2861 2862 2863 2864 2865 2866 2867 2868 2869 2870 2871 2872 2873

    priv->job.info.type = VIR_DOMAIN_JOB_UNBOUNDED;

    return 0;
}

void
qemuMigrationJobSetPhase(struct qemud_driver *driver,
                         virDomainObjPtr vm,
                         enum qemuMigrationJobPhase phase)
{
    qemuDomainObjPrivatePtr priv = vm->privateData;

    if (phase < priv->job.phase) {
        VIR_ERROR(_("migration protocol going backwards %s => %s"),
                  qemuMigrationJobPhaseTypeToString(priv->job.phase),
                  qemuMigrationJobPhaseTypeToString(phase));
        return;
    }

    qemuDomainObjSetJobPhase(driver, vm, phase);
}

void
qemuMigrationJobStartPhase(struct qemud_driver *driver,
                           virDomainObjPtr vm,
                           enum qemuMigrationJobPhase phase)
{
    virDomainObjRef(vm);
    qemuMigrationJobSetPhase(driver, vm, phase);
}

int
qemuMigrationJobContinue(virDomainObjPtr vm)
{
    return virDomainObjUnref(vm);
}

bool
qemuMigrationJobIsActive(virDomainObjPtr vm,
                         enum qemuDomainAsyncJob job)
{
    qemuDomainObjPrivatePtr priv = vm->privateData;

    if (priv->job.asyncJob != job) {
        const char *msg;

        if (job == QEMU_ASYNC_JOB_MIGRATION_IN)
            msg = _("domain '%s' is not processing incoming migration");
        else
            msg = _("domain '%s' is not being migrated");

        qemuReportError(VIR_ERR_OPERATION_INVALID, msg, vm->def->name);
        return false;
    }
    return true;
}

int
qemuMigrationJobFinish(struct qemud_driver *driver, virDomainObjPtr vm)
{
    return qemuDomainObjEndAsyncJob(driver, vm);
}