lxc_container.c 62.5 KB
Newer Older
1
/*
2
 * Copyright (C) 2008-2012 Red Hat, Inc.
3
 * Copyright (C) 2008 IBM Corp.
4 5 6 7 8
 *
 * lxc_container.c: file description
 *
 * Authors:
 *  David L. Leskovec <dlesko at linux.vnet.ibm.com>
9
 *  Daniel P. Berrange <berrange@redhat.com>
10 11 12 13 14 15 16 17 18 19 20 21
 *
 * 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
22
 * License along with this library.  If not, see
O
Osier Yang 已提交
23
 * <http://www.gnu.org/licenses/>.
24 25 26 27 28 29 30
 */

#include <config.h>

#include <fcntl.h>
#include <limits.h>
#include <stdlib.h>
31
#include <stdio.h>
32 33
#include <sys/ioctl.h>
#include <sys/mount.h>
34
#include <sys/wait.h>
35
#include <sys/stat.h>
36
#include <unistd.h>
37
#include <mntent.h>
38 39
#include <sys/reboot.h>
#include <linux/reboot.h>
40 41 42 43 44 45

/* Yes, we want linux private one, for _syscall2() macro */
#include <linux/unistd.h>

/* For MS_MOVE */
#include <linux/fs.h>
46

47
#if WITH_CAPNG
48
# include <cap-ng.h>
D
Daniel P. Berrange 已提交
49
#endif
50

51
#if WITH_BLKID
52 53 54
# include <blkid/blkid.h>
#endif

55 56 57 58
#if WITH_SELINUX
# include <selinux/selinux.h>
#endif

59
#include "virerror.h"
60
#include "virlog.h"
61
#include "lxc_container.h"
62
#include "virutil.h"
63
#include "viralloc.h"
64
#include "virnetdevveth.h"
65
#include "viruuid.h"
E
Eric Blake 已提交
66
#include "virfile.h"
67
#include "virusb.h"
68
#include "vircommand.h"
69
#include "virnetdev.h"
70
#include "virprocess.h"
71

72 73
#define VIR_FROM_THIS VIR_FROM_LXC

74 75 76 77 78 79
/*
 * GLibc headers are behind the kernel, so we define these
 * constants if they're not present already.
 */

#ifndef CLONE_NEWPID
80
# define CLONE_NEWPID  0x20000000
81 82
#endif
#ifndef CLONE_NEWUTS
83
# define CLONE_NEWUTS  0x04000000
84 85
#endif
#ifndef CLONE_NEWUSER
86
# define CLONE_NEWUSER 0x10000000
87 88
#endif
#ifndef CLONE_NEWIPC
89
# define CLONE_NEWIPC  0x08000000
90 91
#endif
#ifndef CLONE_NEWNET
92
# define CLONE_NEWNET  0x40000000 /* New network namespace */
93 94 95 96 97 98 99 100
#endif

/* messages between parent and container */
typedef char lxc_message_t;
#define LXC_CONTINUE_MSG 'c'

typedef struct __lxc_child_argv lxc_child_argv_t;
struct __lxc_child_argv {
101
    virDomainDefPtr config;
102
    virSecurityManagerPtr securityDriver;
103
    size_t nveths;
104
    char **veths;
105
    int monitor;
106 107
    char **ttyPaths;
    size_t nttyPaths;
108
    int handshakefd;
109 110
};

111 112 113
static int lxcContainerMountFSBlock(virDomainFSDefPtr fs,
                                    const char *srcprefix);

114

115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173
/*
 * reboot(LINUX_REBOOT_CMD_CAD_ON) will return -EINVAL
 * in a child pid namespace if container reboot support exists.
 * Otherwise, it will either succeed or return -EPERM.
 */
ATTRIBUTE_NORETURN static int
lxcContainerRebootChild(void *argv)
{
    int *cmd = argv;
    int ret;

    ret = reboot(*cmd);
    if (ret == -1 && errno == EINVAL)
        _exit(1);
    _exit(0);
}


static
int lxcContainerHasReboot(void)
{
    int flags = CLONE_NEWPID|CLONE_NEWNS|CLONE_NEWUTS|
        CLONE_NEWIPC|SIGCHLD;
    int cpid;
    char *childStack;
    char *stack;
    char *buf;
    int cmd, v;
    int status;
    char *tmp;

    if (virFileReadAll("/proc/sys/kernel/ctrl-alt-del", 10, &buf) < 0)
        return -1;

    if ((tmp = strchr(buf, '\n')))
        *tmp = '\0';

    if (virStrToLong_i(buf, NULL, 10, &v) < 0) {
        virReportError(VIR_ERR_INTERNAL_ERROR,
                       _("Malformed ctrl-alt-del setting '%s'"), buf);
        VIR_FREE(buf);
        return -1;
    }
    VIR_FREE(buf);
    cmd = v ? LINUX_REBOOT_CMD_CAD_ON : LINUX_REBOOT_CMD_CAD_OFF;

    if (VIR_ALLOC_N(stack, getpagesize() * 4) < 0) {
        virReportOOMError();
        return -1;
    }

    childStack = stack + (getpagesize() * 4);

    cpid = clone(lxcContainerRebootChild, childStack, flags, &cmd);
    VIR_FREE(stack);
    if (cpid < 0) {
        virReportSystemError(errno, "%s",
                             _("Unable to clone to check reboot support"));
        return -1;
174
    } else if (virProcessWait(cpid, &status) < 0) {
175 176 177 178 179 180 181 182 183 184 185 186 187 188
        return -1;
    }

    if (WEXITSTATUS(status) != 1) {
        VIR_DEBUG("Containerized reboot support is missing "
                  "(kernel probably too old < 3.4)");
        return 0;
    }

    VIR_DEBUG("Containerized reboot support is available");
    return 1;
}


189
/**
190
 * lxcContainerBuildInitCmd:
191
 * @vmDef: pointer to vm definition structure
192
 *
193
 * Build a virCommandPtr for launching the container 'init' process
194
 *
195
 * Returns a virCommandPtr
196
 */
197
static virCommandPtr lxcContainerBuildInitCmd(virDomainDefPtr vmDef)
198
{
199
    char uuidstr[VIR_UUID_STRING_BUFLEN];
200
    virCommandPtr cmd;
201 202 203

    virUUIDFormat(vmDef->uuid, uuidstr);

204 205
    cmd = virCommandNew(vmDef->os.init);

206 207 208
    if (vmDef->os.initargv && vmDef->os.initargv[0])
        virCommandAddArgSet(cmd, (const char **)vmDef->os.initargv);

209 210
    virCommandAddEnvString(cmd, "PATH=/bin:/sbin");
    virCommandAddEnvString(cmd, "TERM=linux");
211
    virCommandAddEnvString(cmd, "container=lxc-libvirt");
212
    virCommandAddEnvPair(cmd, "container_uuid", uuidstr);
213 214
    virCommandAddEnvPair(cmd, "LIBVIRT_LXC_UUID", uuidstr);
    virCommandAddEnvPair(cmd, "LIBVIRT_LXC_NAME", vmDef->name);
215 216
    if (vmDef->os.cmdline)
        virCommandAddEnvPair(cmd, "LIBVIRT_LXC_CMDLINE", vmDef->os.cmdline);
217

218
    return cmd;
219 220 221
}

/**
222
 * lxcContainerSetStdio:
223 224
 * @control: control FD from parent
 * @ttyfd: FD of tty to set as the container console
225 226 227 228 229 230
 *
 * Sets the given tty as the primary conosole for the container as well as
 * stdout, stdin and stderr.
 *
 * Returns 0 on success or -1 in case of error
 */
231
static int lxcContainerSetStdio(int control, int ttyfd, int handshakefd)
232 233
{
    int rc = -1;
234
    int open_max, i;
235 236

    if (setsid() < 0) {
237
        virReportSystemError(errno, "%s",
238
                             _("setsid failed"));
239
        goto cleanup;
240 241 242
    }

    if (ioctl(ttyfd, TIOCSCTTY, NULL) < 0) {
243
        virReportSystemError(errno, "%s",
244
                             _("ioctl(TIOCSTTY) failed"));
245 246 247
        goto cleanup;
    }

248 249
    /* Just in case someone forget to set FD_CLOEXEC, explicitly
     * close all FDs before executing the container */
250
    open_max = sysconf(_SC_OPEN_MAX);
251
    for (i = 0; i < open_max; i++)
252
        if (i != ttyfd && i != control && i != handshakefd) {
253
            int tmpfd = i;
254
            VIR_MASS_CLOSE(tmpfd);
255
        }
256 257

    if (dup2(ttyfd, 0) < 0) {
258
        virReportSystemError(errno, "%s",
259
                             _("dup2(stdin) failed"));
260 261 262 263
        goto cleanup;
    }

    if (dup2(ttyfd, 1) < 0) {
264
        virReportSystemError(errno, "%s",
265
                             _("dup2(stdout) failed"));
266 267 268 269
        goto cleanup;
    }

    if (dup2(ttyfd, 2) < 0) {
270
        virReportSystemError(errno, "%s",
271
                             _("dup2(stderr) failed"));
272 273 274 275 276 277
        goto cleanup;
    }

    rc = 0;

cleanup:
278
    VIR_DEBUG("rc=%d", rc);
279 280 281 282
    return rc;
}

/**
283
 * lxcContainerSendContinue:
284
 * @control: control FD to child
285
 *
286 287
 * Sends the continue message via the socket pair stored in the vm
 * structure.
288 289 290
 *
 * Returns 0 on success or -1 in case of error
 */
291
int lxcContainerSendContinue(int control)
292 293
{
    int rc = -1;
294 295
    lxc_message_t msg = LXC_CONTINUE_MSG;
    int writeCount = 0;
296

297 298 299
    writeCount = safewrite(control, &msg, sizeof(msg));
    if (writeCount != sizeof(msg)) {
        goto error_out;
300 301
    }

302 303 304
    rc = 0;
error_out:
    return rc;
305 306
}

307
/**
308
 * lxcContainerWaitForContinue:
309
 * @control: Control FD from parent
310 311 312 313 314 315 316
 *
 * This function will wait for the container continue message from the
 * parent process.  It will send this message on the socket pair stored in
 * the vm structure once it has completed the post clone container setup.
 *
 * Returns 0 on success or -1 in case of error
 */
317
int lxcContainerWaitForContinue(int control)
318 319 320 321
{
    lxc_message_t msg;
    int readLen;

322
    readLen = saferead(control, &msg, sizeof(msg));
323 324 325 326 327 328 329
    if (readLen != sizeof(msg)) {
        if (readLen >= 0)
            errno = EIO;
        return -1;
    }
    if (msg != LXC_CONTINUE_MSG) {
        errno = EINVAL;
330
        return -1;
331 332
    }

333
    return 0;
334 335
}

336

337
/**
338
 * lxcContainerRenameAndEnableInterfaces:
339 340
 * @nveths: number of interfaces
 * @veths: interface names
341
 *
342 343 344
 * This function will rename the interfaces to ethN
 * with id ascending order from zero and enable the
 * renamed interfaces for this container.
345 346 347
 *
 * Returns 0 on success or nonzero in case of error
 */
348
static int lxcContainerRenameAndEnableInterfaces(bool privNet,
349
                                                 size_t nveths,
350
                                                 char **veths)
351 352
{
    int rc = 0;
353
    size_t i;
354
    char *newname = NULL;
355

356
    for (i = 0 ; i < nveths ; i++) {
357
        if (virAsprintf(&newname, "eth%zu", i) < 0) {
358 359
            virReportOOMError();
            rc = -1;
360
            goto error_out;
361
        }
362

363
        VIR_DEBUG("Renaming %s to %s", veths[i], newname);
364
        rc = virNetDevSetName(veths[i], newname);
365
        if (rc < 0)
366 367
            goto error_out;

368
        VIR_DEBUG("Enabling %s", newname);
369
        rc = virNetDevSetOnline(newname, true);
370
        if (rc < 0)
371
            goto error_out;
372

373
        VIR_FREE(newname);
374 375 376
    }

    /* enable lo device only if there were other net devices */
377
    if (veths || privNet)
378
        rc = virNetDevSetOnline("lo", true);
379 380

error_out:
381
    VIR_FREE(newname);
382 383 384
    return rc;
}

385

386
/*_syscall2(int, pivot_root, char *, newroot, const char *, oldroot)*/
387 388 389 390 391 392 393
extern int pivot_root(const char * new_root,const char * put_old);

static int lxcContainerChildMountSort(const void *a, const void *b)
{
  const char **sa = (const char**)a;
  const char **sb = (const char**)b;

E
Eric Blake 已提交
394
  /* Deliberately reversed args - we need to unmount deepest
395 396 397 398
     children first */
  return strcmp(*sb, *sa);
}

399
#ifndef MS_REC
400
# define MS_REC          16384
401 402 403
#endif

#ifndef MNT_DETACH
404
# define MNT_DETACH      0x00000002
405 406 407
#endif

#ifndef MS_PRIVATE
408
# define MS_PRIVATE              (1<<18)
409 410 411
#endif

#ifndef MS_SLAVE
412
# define MS_SLAVE                (1<<19)
413 414
#endif

415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521
static int lxcContainerGetSubtree(const char *prefix,
                                  char ***mountsret,
                                  size_t *nmountsret)
{
    FILE *procmnt;
    struct mntent mntent;
    char mntbuf[1024];
    int ret = -1;
    char **mounts = NULL;
    size_t nmounts = 0;

    VIR_DEBUG("prefix=%s", prefix);

    *mountsret = NULL;
    *nmountsret = 0;

    if (!(procmnt = setmntent("/proc/mounts", "r"))) {
        virReportSystemError(errno, "%s",
                             _("Failed to read /proc/mounts"));
        return -1;
    }

    while (getmntent_r(procmnt, &mntent, mntbuf, sizeof(mntbuf)) != NULL) {
        VIR_DEBUG("Got %s", mntent.mnt_dir);
        if (!STRPREFIX(mntent.mnt_dir, prefix))
            continue;

        if (VIR_REALLOC_N(mounts, nmounts+1) < 0) {
            virReportOOMError();
            goto cleanup;
        }
        if (!(mounts[nmounts] = strdup(mntent.mnt_dir))) {
            virReportOOMError();
            goto cleanup;
        }
        nmounts++;
        VIR_DEBUG("Grabbed %s", mntent.mnt_dir);
    }

    if (mounts)
        qsort(mounts, nmounts, sizeof(mounts[0]),
              lxcContainerChildMountSort);

    ret = 0;
cleanup:
    *mountsret = mounts;
    *nmountsret = nmounts;
    endmntent(procmnt);
    return ret;
}

static int lxcContainerUnmountSubtree(const char *prefix,
                                      bool isOldRootFS)
{
    char **mounts = NULL;
    size_t nmounts = 0;
    size_t i;
    int saveErrno;
    const char *failedUmount = NULL;
    int ret = -1;

    VIR_DEBUG("Unmount subtreee from %s", prefix);

    if (lxcContainerGetSubtree(prefix, &mounts, &nmounts) < 0)
        goto cleanup;
    for (i = 0 ; i < nmounts ; i++) {
        VIR_DEBUG("Umount %s", mounts[i]);
        if (umount(mounts[i]) < 0) {
            char ebuf[1024];
            failedUmount = mounts[i];
            saveErrno = errno;
            VIR_WARN("Failed to unmount '%s', trying to detach subtree '%s': %s",
                     failedUmount, mounts[nmounts-1],
                     virStrerror(errno, ebuf, sizeof(ebuf)));
            break;
        }
    }

    if (failedUmount) {
        /* This detaches the subtree */
        if (umount2(mounts[nmounts-1], MNT_DETACH) < 0) {
            virReportSystemError(saveErrno,
                                 _("Failed to unmount '%s' and could not detach subtree '%s'"),
                                 failedUmount, mounts[nmounts-1]);
            goto cleanup;
        }
        /* This unmounts the tmpfs on which the old root filesystem was hosted */
        if (isOldRootFS &&
            umount(mounts[nmounts-1]) < 0) {
            virReportSystemError(saveErrno,
                                 _("Failed to unmount '%s' and could not unmount old root '%s'"),
                                 failedUmount, mounts[nmounts-1]);
            goto cleanup;
        }
    }

    ret = 0;

cleanup:
    for (i = 0 ; i < nmounts ; i++)
        VIR_FREE(mounts[i]);
    VIR_FREE(mounts);

    return ret;
}


522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566
static int lxcContainerPrepareRoot(virDomainDefPtr def,
                                   virDomainFSDefPtr root)
{
    char *dst;
    char *tmp;

    if (root->type == VIR_DOMAIN_FS_TYPE_MOUNT)
        return 0;

    if (root->type == VIR_DOMAIN_FS_TYPE_FILE) {
        virReportError(VIR_ERR_INTERNAL_ERROR, "%s",
                       _("Unexpected root filesystem without loop device"));
        return -1;
    }

    if (root->type != VIR_DOMAIN_FS_TYPE_BLOCK) {
        virReportError(VIR_ERR_CONFIG_UNSUPPORTED,
                       _("Unsupported root filesystem type %s"),
                       virDomainFSTypeToString(root->type));
        return -1;
    }

    if (virAsprintf(&dst, "%s/%s.root",
                    LXC_STATE_DIR, def->name) < 0) {
        virReportOOMError();
        return -1;
    }

    tmp = root->dst;
    root->dst = dst;

    if (lxcContainerMountFSBlock(root, "") < 0) {
        root->dst = tmp;
        VIR_FREE(dst);
        return -1;
    }

    root->dst = tmp;
    root->type = VIR_DOMAIN_FS_TYPE_MOUNT;
    VIR_FREE(root->src);
    root->src = dst;

    return 0;
}

567 568
static int lxcContainerPivotRoot(virDomainFSDefPtr root)
{
569
    int ret;
570
    char *oldroot = NULL, *newroot = NULL;
571

M
Mark McLoughlin 已提交
572 573
    ret = -1;

574 575
    VIR_DEBUG("Pivot via %s", root->src);

576 577
    /* root->parent must be private, so make / private. */
    if (mount("", "/", NULL, MS_PRIVATE|MS_REC, NULL) < 0) {
578
        virReportSystemError(errno, "%s",
579
                             _("Failed to make root private"));
580
        goto err;
581 582
    }

583
    if (virAsprintf(&oldroot, "%s/.oldroot", root->src) < 0) {
584
        virReportOOMError();
585
        goto err;
586 587
    }

588 589
    if (virFileMakePath(oldroot) < 0) {
        virReportSystemError(errno,
590
                             _("Failed to create %s"),
591
                             oldroot);
592 593 594 595 596
        goto err;
    }

    /* Create a tmpfs root since old and new roots must be
     * on separate filesystems */
597
    if (mount("tmprootfs", oldroot, "tmpfs", 0, NULL) < 0) {
598
        virReportSystemError(errno,
599
                             _("Failed to mount empty tmpfs at %s"),
600 601 602
                             oldroot);
        goto err;
    }
M
Mark McLoughlin 已提交
603

604 605
    /* Create a directory called 'new' in tmpfs */
    if (virAsprintf(&newroot, "%s/new", oldroot) < 0) {
606
        virReportOOMError();
607 608 609
        goto err;
    }

610 611
    if (virFileMakePath(newroot) < 0) {
        virReportSystemError(errno,
612
                             _("Failed to create %s"),
613 614 615 616 617 618
                             newroot);
        goto err;
    }

    /* ... and mount our root onto it */
    if (mount(root->src, newroot, NULL, MS_BIND|MS_REC, NULL) < 0) {
619
        virReportSystemError(errno,
620
                             _("Failed to bind new root %s into tmpfs"),
621 622 623 624
                             root->src);
        goto err;
    }

625 626 627 628 629 630 631 632 633
    if (root->readonly) {
        if (mount(root->src, newroot, NULL, MS_BIND|MS_REC|MS_RDONLY|MS_REMOUNT, NULL) < 0) {
            virReportSystemError(errno,
                                 _("Failed to make new root %s readonly"),
                                 root->src);
            goto err;
        }
    }

634 635
    /* Now we chroot into the tmpfs, then pivot into the
     * root->src bind-mounted onto '/new' */
636
    if (chdir(newroot) < 0) {
637
        virReportSystemError(errno,
638
                             _("Failed to chroot into %s"), newroot);
639
        goto err;
640 641 642 643
    }

    /* The old root directory will live at /.oldroot after
     * this and will soon be unmounted completely */
644
    if (pivot_root(".", ".oldroot") < 0) {
645
        virReportSystemError(errno, "%s",
646
                             _("Failed to pivot root"));
647
        goto err;
648 649 650
    }

    /* CWD is undefined after pivot_root, so go to / */
651 652 653
    if (chdir("/") < 0)
        goto err;

M
Mark McLoughlin 已提交
654 655 656
    ret = 0;

err:
657 658 659
    VIR_FREE(oldroot);
    VIR_FREE(newroot);

M
Mark McLoughlin 已提交
660
    return ret;
661 662
}

663

664 665
static int lxcContainerMountBasicFS(bool pivotRoot,
                                    char *sec_mount_options)
666 667
{
    const struct {
668 669 670
        const char *src;
        const char *dst;
        const char *type;
671 672
        const char *opts;
        int mflags;
673
    } mnts[] = {
674
        /* When we want to make a bind mount readonly, for unknown reasons,
675
         * it is currently necessary to bind it once, and then remount the
676
         * bind with the readonly flag. If this is not done, then the original
677
         * mount point in the main OS becomes readonly too which is not what
678 679
         * we want. Hence some things have two entries here.
         */
680 681 682 683 684
        { "proc", "/proc", "proc", NULL, MS_NOSUID|MS_NOEXEC|MS_NODEV },
        { "/proc/sys", "/proc/sys", NULL, NULL, MS_BIND },
        { "/proc/sys", "/proc/sys", NULL, NULL, MS_BIND|MS_REMOUNT|MS_RDONLY },
        { "sysfs", "/sys", "sysfs", NULL, MS_NOSUID|MS_NOEXEC|MS_NODEV },
        { "sysfs", "/sys", "sysfs", NULL, MS_BIND|MS_REMOUNT|MS_RDONLY },
685
#if WITH_SELINUX
686 687
        { SELINUX_MOUNT, SELINUX_MOUNT, "selinuxfs", NULL, MS_NOSUID|MS_NOEXEC|MS_NODEV },
        { SELINUX_MOUNT, SELINUX_MOUNT, NULL, NULL, MS_BIND|MS_REMOUNT|MS_RDONLY },
688
#endif
689
    };
690
    int i, rc = -1;
691
    char *opts = NULL;
692

693
    VIR_DEBUG("Mounting basic filesystems pivotRoot=%d", pivotRoot);
694

695
    for (i = 0 ; i < ARRAY_CARDINALITY(mnts) ; i++) {
696
        const char *srcpath = NULL;
697

698 699
        VIR_DEBUG("Processing %s -> %s",
                  mnts[i].src, mnts[i].dst);
700

701
        srcpath = mnts[i].src;
702 703 704

        /* Skip if mount doesn't exist in source */
        if ((srcpath[0] == '/') &&
705
            (access(srcpath, R_OK) < 0))
706 707
            continue;

708 709 710 711 712 713 714
        if (virFileMakePath(mnts[i].dst) < 0) {
            virReportSystemError(errno,
                                 _("Failed to mkdir %s"),
                                 mnts[i].src);
            goto cleanup;
        }

715 716
        VIR_DEBUG("Mount %s on %s type=%s flags=%x, opts=%s",
                  srcpath, mnts[i].dst, mnts[i].type, mnts[i].mflags, mnts[i].opts);
717
        if (mount(srcpath, mnts[i].dst, mnts[i].type, mnts[i].mflags, mnts[i].opts) < 0) {
718
            virReportSystemError(errno,
719 720 721
                                 _("Failed to mount %s on %s type %s flags=%x opts=%s"),
                                 srcpath, mnts[i].dst, NULLSTR(mnts[i].type),
                                 mnts[i].mflags, NULLSTR(mnts[i].opts));
722
            goto cleanup;
723
        }
724 725
    }

726 727 728 729 730
    if (pivotRoot) {
        /*
         * tmpfs is limited to 64kb, since we only have device nodes in there
         * and don't want to DOS the entire OS RAM usage
         */
731

732
        ignore_value(virAsprintf(&opts,
733
                                 "mode=755,size=65536%s", sec_mount_options));
734
        if (!opts) {
735 736 737 738 739 740 741 742
            virReportOOMError();
            goto cleanup;
        }

        VIR_DEBUG("Mount devfs on /dev type=tmpfs flags=%x, opts=%s",
                  MS_NOSUID, opts);
        if (mount("devfs", "/dev", "tmpfs", MS_NOSUID, opts) < 0) {
            virReportSystemError(errno,
743 744
                                 _("Failed to mount %s on %s type %s (%s)"),
                                 "devfs", "/dev", "tmpfs", opts);
745 746 747 748
            goto cleanup;
        }
    }

749 750 751 752
    rc = 0;

cleanup:
    VIR_DEBUG("rc=%d", rc);
753
    VIR_FREE(opts);
754 755 756
    return rc;
}

757
#if WITH_FUSE
758 759 760 761 762 763 764
static int lxcContainerMountProcFuse(virDomainDefPtr def,
                                     const char *srcprefix)
{
    int ret;
    char *meminfo_path = NULL;

    if ((ret = virAsprintf(&meminfo_path,
765
                           "%s/%s/%s.fuse/meminfo",
766
                           srcprefix ? srcprefix : "", LXC_STATE_DIR,
767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786
                           def->name)) < 0)
        return ret;

    if ((ret = mount(meminfo_path, "/proc/meminfo",
                     NULL, MS_BIND, NULL)) < 0) {
        virReportSystemError(errno,
                             _("Failed to mount %s on /proc/meminfo"),
                             meminfo_path);
    }

    VIR_FREE(meminfo_path);
    return ret;
}
#else
static int lxcContainerMountProcFuse(virDomainDefPtr def ATTRIBUTE_UNUSED,
                                     const char *srcprefix ATTRIBUTE_UNUSED)
{
    return 0;
}
#endif
787

788 789
static int lxcContainerMountFSDevPTS(virDomainDefPtr def,
                                     const char *srcprefix)
790
{
791 792
    int ret;
    char *path = NULL;
793

794 795 796 797 798
    if ((ret = virAsprintf(&path,
                           "%s/%s/%s.devpts",
                           srcprefix ? srcprefix : "", LXC_STATE_DIR,
                           def->name)) < 0)
        return ret;
799

800 801
    if (virFileMakePath("/dev/pts") < 0) {
        virReportSystemError(errno, "%s",
802
                             _("Cannot create /dev/pts"));
803
        goto cleanup;
804
    }
805

806 807 808 809 810 811 812
    VIR_DEBUG("Trying to move %s to /dev/pts", path);

    if ((ret = mount(path, "/dev/pts",
                     NULL, MS_MOVE, NULL)) < 0) {
        virReportSystemError(errno,
                             _("Failed to mount %s on /dev/pts"),
                             path);
813
        goto cleanup;
814
    }
815

816 817
cleanup:
    VIR_FREE(path);
818

819
    return ret;
820 821
}

822
static int lxcContainerPopulateDevices(char **ttyPaths, size_t nttyPaths)
823
{
824
    size_t i;
825 826 827 828 829 830 831 832 833 834 835 836
    const struct {
        int maj;
        int min;
        mode_t mode;
        const char *path;
    } devs[] = {
        { LXC_DEV_MAJ_MEMORY, LXC_DEV_MIN_NULL, 0666, "/dev/null" },
        { LXC_DEV_MAJ_MEMORY, LXC_DEV_MIN_ZERO, 0666, "/dev/zero" },
        { LXC_DEV_MAJ_MEMORY, LXC_DEV_MIN_FULL, 0666, "/dev/full" },
        { LXC_DEV_MAJ_MEMORY, LXC_DEV_MIN_RANDOM, 0666, "/dev/random" },
        { LXC_DEV_MAJ_MEMORY, LXC_DEV_MIN_URANDOM, 0666, "/dev/urandom" },
    };
837 838 839 840 841 842 843 844 845
    const struct {
        const char *src;
        const char *dst;
    } links[] = {
        { "/proc/self/fd/0", "/dev/stdin" },
        { "/proc/self/fd/1", "/dev/stdout" },
        { "/proc/self/fd/2", "/dev/stderr" },
        { "/proc/self/fd", "/dev/fd" },
    };
846 847 848 849

    /* Populate /dev/ with a few important bits */
    for (i = 0 ; i < ARRAY_CARDINALITY(devs) ; i++) {
        dev_t dev = makedev(devs[i].maj, devs[i].min);
850
        if (mknod(devs[i].path, S_IFCHR, dev) < 0 ||
851
            chmod(devs[i].path, devs[i].mode)) {
852
            virReportSystemError(errno,
853
                                 _("Failed to make device %s"),
854
                                 devs[i].path);
855 856 857 858
            return -1;
        }
    }

859 860 861 862 863 864 865 866 867
    for (i = 0 ; i < ARRAY_CARDINALITY(links) ; i++) {
        if (symlink(links[i].src, links[i].dst) < 0) {
            virReportSystemError(errno,
                                 _("Failed to symlink device %s to %s"),
                                 links[i].dst, links[i].src);
            return -1;
        }
    }

868
    if (access("/dev/pts/ptmx", W_OK) == 0) {
869 870 871 872
        /* We have private devpts capability, so bind that */
        if (virFileTouch("/dev/ptmx", 0666) < 0)
            return -1;

S
Serge Hallyn 已提交
873
        if (mount("/dev/pts/ptmx", "/dev/ptmx", "ptmx", MS_BIND, NULL) < 0) {
874
            virReportSystemError(errno, "%s",
875 876 877 878 879 880 881 882 883 884
                                 _("Failed to bind /dev/pts/ptmx on to /dev/ptmx"));
            return -1;
        }
    } else {
        /* Legacy devpts, so we need to just use shared one */
        dev_t dev = makedev(LXC_DEV_MAJ_TTY, LXC_DEV_MIN_PTMX);
        if (mknod("/dev/ptmx", S_IFCHR, dev) < 0 ||
            chmod("/dev/ptmx", 0666)) {
            virReportSystemError(errno, "%s",
                                 _("Failed to make device /dev/ptmx"));
885 886 887 888
            return -1;
        }
    }

889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909
    for (i = 0 ; i < nttyPaths ; i++) {
        char *tty;
        if (virAsprintf(&tty, "/dev/tty%zu", i+1) < 0) {
            virReportOOMError();
            return -1;
        }
        if (symlink(ttyPaths[i], tty) < 0) {
            VIR_FREE(tty);
            virReportSystemError(errno,
                                 _("Failed to symlink %s to %s"),
                                 ttyPaths[i], tty);
            return -1;
        }
        VIR_FREE(tty);
        if (i == 0 &&
            symlink(ttyPaths[i], "/dev/console") < 0) {
            virReportSystemError(errno,
                                 _("Failed to symlink %s to /dev/console"),
                                 ttyPaths[i]);
            return -1;
        }
910
    }
911 912 913 914
    return 0;
}


915 916
static int lxcContainerMountFSBind(virDomainFSDefPtr fs,
                                   const char *srcprefix)
917
{
918 919
    char *src = NULL;
    int ret = -1;
920
    struct stat st;
921 922 923 924 925 926

    if (virAsprintf(&src, "%s%s", srcprefix, fs->src) < 0) {
        virReportOOMError();
        goto cleanup;
    }

927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963
    if (stat(fs->dst, &st) < 0) {
        if (errno != ENOENT) {
            virReportSystemError(errno, _("Unable to stat bind target %s"),
                                 fs->dst);
            goto cleanup;
        }
        /* ENOENT => create the target dir or file */
        if (stat(src, &st) < 0) {
            virReportSystemError(errno, _("Unable to stat bind source %s"),
                                 src);
            goto cleanup;
        }
        if (S_ISDIR(st.st_mode)) {
            if (virFileMakePath(fs->dst) < 0) {
                virReportSystemError(errno,
                                     _("Failed to create %s"),
                                     fs->dst);
                goto cleanup;
            }
        } else {
            /* Create Empty file for target mount point */
            int fd = open(fs->dst, O_WRONLY|O_CREAT|O_NOCTTY|O_NONBLOCK, 0666);
            if (fd < 0) {
                if (errno != EEXIST) {
                    virReportSystemError(errno,
                                         _("Failed to create bind target %s"),
                                         fs->dst);
                    goto cleanup;
                }
            }
            if (VIR_CLOSE(fd) < 0) {
                virReportSystemError(errno,
                                     _("Failed to close bind target %s"),
                                     fs->dst);
                goto cleanup;
            }
        }
964 965 966 967 968 969 970 971 972
    }

    if (mount(src, fs->dst, NULL, MS_BIND, NULL) < 0) {
        virReportSystemError(errno,
                             _("Failed to bind mount directory %s to %s"),
                             src, fs->dst);
        goto cleanup;
    }

973 974
    if (fs->readonly) {
        VIR_DEBUG("Binding %s readonly", fs->dst);
975
        if (mount(src, fs->dst, NULL, MS_BIND|MS_REMOUNT|MS_RDONLY, NULL) < 0) {
976 977 978
            virReportSystemError(errno,
                                 _("Failed to make directory %s readonly"),
                                 fs->dst);
979 980 981 982 983 984 985 986 987 988 989
        }
    }

    ret = 0;

cleanup:
    VIR_FREE(src);
    return ret;
}


990
#ifdef WITH_BLKID
991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060
static int
lxcContainerMountDetectFilesystem(const char *src, char **type)
{
    int fd;
    int ret = -1;
    int rc;
    const char *data = NULL;
    blkid_probe blkid = NULL;

    *type = NULL;

    if ((fd = open(src, O_RDONLY)) < 0) {
        virReportSystemError(errno,
                             _("Unable to open filesystem %s"), src);
        return -1;
    }

    if (!(blkid = blkid_new_probe())) {
        virReportSystemError(errno, "%s",
                             _("Unable to create blkid library handle"));
        goto cleanup;
    }
    if (blkid_probe_set_device(blkid, fd, 0, 0) < 0) {
        virReportSystemError(EINVAL,
                             _("Unable to associate device %s with blkid library"),
                             src);
        goto cleanup;
    }

    blkid_probe_enable_superblocks(blkid, 1);

    blkid_probe_set_superblocks_flags(blkid, BLKID_SUBLKS_TYPE);

    rc = blkid_do_safeprobe(blkid);
    if (rc != 0) {
        if (rc == 1) /* Nothing found, return success with *type == NULL */
            goto done;

        if (rc == -2) {
            virReportSystemError(EINVAL,
                                 _("Too many filesystems detected for %s"),
                                 src);
        } else {
            virReportSystemError(errno,
                                 _("Unable to detect filesystem for %s"),
                                 src);
        }
        goto cleanup;
    }

    if (blkid_probe_lookup_value(blkid, "TYPE", &data, NULL) < 0) {
        virReportSystemError(ENOENT,
                             _("Unable to find filesystem type for %s"),
                             src);
        goto cleanup;
    }

    if (!(*type = strdup(data))) {
        virReportOOMError();
        goto cleanup;
    }

done:
    ret = 0;
cleanup:
    VIR_FORCE_CLOSE(fd);
    if (blkid)
        blkid_free_probe(blkid);
    return ret;
}
1061
#else /* ! WITH_BLKID */
1062 1063 1064 1065 1066 1067 1068 1069
static int
lxcContainerMountDetectFilesystem(const char *src ATTRIBUTE_UNUSED,
                                  char **type)
{
    /* No libblkid, so just return success with no detected type */
    *type = NULL;
    return 0;
}
1070
#endif /* ! WITH_BLKID */
1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127

/*
 * This functions attempts to do automatic detection of filesystem
 * type following the same rules as the util-linux 'mount' binary.
 *
 * The main difference is that we don't (currently) try to use
 * libblkid to detect the format first. We go straight to using
 * /etc/filesystems, and then /proc/filesystems
 */
static int lxcContainerMountFSBlockAuto(virDomainFSDefPtr fs,
                                        int fsflags,
                                        const char *src,
                                        const char *srcprefix)
{
    FILE *fp = NULL;
    int ret = -1;
    bool tryProc = false;
    bool gotStar = false;
    char *fslist = NULL;
    char *line = NULL;
    const char *type;

    VIR_DEBUG("src=%s srcprefix=%s dst=%s", src, srcprefix, fs->dst);

    /* First time around we use /etc/filesystems */
retry:
    if (virAsprintf(&fslist, "%s%s",
                    srcprefix, tryProc ? "/proc/filesystems" : "/etc/filesystems") < 0) {
        virReportOOMError();
        goto cleanup;
    }

    VIR_DEBUG("Open fslist %s", fslist);
    if (!(fp = fopen(fslist, "r"))) {
        /* If /etc/filesystems does not exist, then we need to retry
         * with /proc/filesystems next
         */
        if (errno == ENOENT &&
            !tryProc) {
            tryProc = true;
            VIR_FREE(fslist);
            goto retry;
        }

        virReportSystemError(errno,
                             _("Unable to read %s"),
                             fslist);
        goto cleanup;
    }

    while (!feof(fp)) {
        size_t n;
        VIR_FREE(line);
        if (getline(&line, &n, fp) <= 0) {
            if (feof(fp))
                break;

1128 1129 1130
            goto cleanup;
        }

1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143
        if (strstr(line, "nodev"))
            continue;

        type = strchr(line, '\n');
        if (type)
            line[type-line] = '\0';

        type = line;
        virSkipSpaces(&type);

        /*
         * /etc/filesystems is only allowed to contain '*' on the last line
         */
1144
        if (gotStar && !tryProc) {
1145 1146 1147
            virReportError(VIR_ERR_INTERNAL_ERROR,
                           _("%s has unexpected '*' before last line"),
                           fslist);
1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168
            goto cleanup;
        }

        /* An '*' on the last line in /etc/filesystems
         * means try /proc/filesystems next. We don't
         * jump immediately though, since we need to see
         * if any more lines follow
         */
        if (!tryProc &&
            STREQ(type, "*"))
            gotStar = true;

        VIR_DEBUG("Trying mount %s with %s", src, type);
        if (mount(src, fs->dst, type, fsflags, NULL) < 0) {
            /* These errnos indicate a bogus filesystem type for
             * the image we have, so skip to the next type
             */
            if (errno == EINVAL || errno == ENODEV)
                continue;

            virReportSystemError(errno,
1169
                                 _("Failed to mount device %s to %s"),
1170 1171 1172 1173 1174 1175
                                 src, fs->dst);
            goto cleanup;
        }

        ret = 0;
        break;
1176 1177
    }

1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189
    /* We've got to the end of /etc/filesystems and saw
     * a '*', so we must try /proc/filesystems next
     */
    if (ret != 0 &&
        !tryProc &&
        gotStar) {
        tryProc = true;
        VIR_FREE(fslist);
        VIR_FORCE_FCLOSE(fp);
        goto retry;
    }

1190 1191 1192 1193 1194 1195
    if (ret != 0) {
        virReportSystemError(ENODEV,
                             _("Failed to mount device %s to %s, unable to detect filesystem"),
                             src, fs->dst);
    }

1196 1197 1198 1199
    VIR_DEBUG("Done mounting filesystem ret=%d tryProc=%d", ret, tryProc);

cleanup:
    VIR_FREE(line);
1200
    VIR_FREE(fslist);
1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215
    VIR_FORCE_FCLOSE(fp);
    return ret;
}


/*
 * Mount a block device 'src' on fs->dst, automatically
 * probing for filesystem type
 */
static int lxcContainerMountFSBlockHelper(virDomainFSDefPtr fs,
                                          const char *src,
                                          const char *srcprefix)
{
    int fsflags = 0;
    int ret = -1;
1216 1217
    char *format = NULL;

1218 1219 1220 1221 1222 1223 1224 1225 1226 1227
    if (fs->readonly)
        fsflags |= MS_RDONLY;

    if (virFileMakePath(fs->dst) < 0) {
        virReportSystemError(errno,
                             _("Failed to create %s"),
                             fs->dst);
        goto cleanup;
    }

1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242
    if (lxcContainerMountDetectFilesystem(src, &format) < 0)
        goto cleanup;

    if (format) {
        VIR_DEBUG("Mount %s with detected format %s", src, format);
        if (mount(src, fs->dst, format, fsflags, NULL) < 0) {
            virReportSystemError(errno,
                                 _("Failed to mount device %s to %s as %s"),
                                 src, fs->dst, format);
            goto cleanup;
        }
        ret = 0;
    } else {
        ret = lxcContainerMountFSBlockAuto(fs, fsflags, src, srcprefix);
    }
1243 1244

cleanup:
1245
    VIR_FREE(format);
1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261
    return ret;
}


static int lxcContainerMountFSBlock(virDomainFSDefPtr fs,
                                    const char *srcprefix)
{
    char *src = NULL;
    int ret = -1;

    if (virAsprintf(&src, "%s%s", srcprefix, fs->src) < 0) {
        virReportOOMError();
        goto cleanup;
    }

    ret = lxcContainerMountFSBlockHelper(fs, src, srcprefix);
1262 1263 1264 1265 1266 1267 1268 1269 1270

    VIR_DEBUG("Done mounting filesystem ret=%d", ret);

cleanup:
    VIR_FREE(src);
    return ret;
}


1271 1272
static int lxcContainerMountFSTmpfs(virDomainFSDefPtr fs,
                                    char *sec_mount_options)
1273 1274 1275 1276
{
    int ret = -1;
    char *data = NULL;

1277
    if (virAsprintf(&data,
1278
                    "size=%lldk%s", fs->usage, sec_mount_options) < 0) {
1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313
        virReportOOMError();
        goto cleanup;
    }

    if (virFileMakePath(fs->dst) < 0) {
        virReportSystemError(errno,
                             _("Failed to create %s"),
                             fs->dst);
        goto cleanup;
    }

    if (mount("tmpfs", fs->dst, "tmpfs", MS_NOSUID|MS_NODEV, data) < 0) {
        virReportSystemError(errno,
                             _("Failed to mount directory %s as tmpfs"),
                             fs->dst);
        goto cleanup;
    }

    if (fs->readonly) {
        VIR_DEBUG("Binding %s readonly", fs->dst);
        if (mount(fs->dst, fs->dst, NULL, MS_BIND|MS_REMOUNT|MS_RDONLY, NULL) < 0) {
            virReportSystemError(errno,
                                 _("Failed to make directory %s readonly"),
                                 fs->dst);
        }
    }

    ret = 0;

cleanup:
    VIR_FREE(data);
    return ret;
}


1314
static int lxcContainerMountFS(virDomainFSDefPtr fs,
1315 1316
                               const char *srcprefix,
                               char *sec_mount_options)
1317 1318 1319 1320 1321 1322
{
    switch (fs->type) {
    case VIR_DOMAIN_FS_TYPE_MOUNT:
        if (lxcContainerMountFSBind(fs, srcprefix) < 0)
            return -1;
        break;
1323 1324 1325 1326
    case VIR_DOMAIN_FS_TYPE_BLOCK:
        if (lxcContainerMountFSBlock(fs, srcprefix) < 0)
            return -1;
        break;
1327
    case VIR_DOMAIN_FS_TYPE_RAM:
1328
        if (lxcContainerMountFSTmpfs(fs, sec_mount_options) < 0)
1329 1330
            return -1;
        break;
1331 1332 1333 1334
    case VIR_DOMAIN_FS_TYPE_BIND:
        if (lxcContainerMountFSBind(fs, "") < 0)
            return -1;
        break;
1335
    case VIR_DOMAIN_FS_TYPE_FILE:
1336 1337 1338
        /* We do actually support this, but the lxc controller
         * should have associated the file with a loopback
         * device and changed this to TYPE_BLOCK for us */
1339 1340 1341
        virReportError(VIR_ERR_INTERNAL_ERROR,
                       _("Unexpected filesystem type %s"),
                       virDomainFSTypeToString(fs->type));
1342
        break;
1343
    default:
1344 1345 1346
        virReportError(VIR_ERR_CONFIG_UNSUPPORTED,
                       _("Cannot mount filesystem type %s"),
                       virDomainFSTypeToString(fs->type));
1347 1348 1349 1350 1351 1352 1353 1354
        break;
    }
    return 0;
}


static int lxcContainerMountAllFS(virDomainDefPtr vmDef,
                                  const char *dstprefix,
1355 1356
                                  bool skipRoot,
                                  char *sec_mount_options)
1357 1358 1359
{
    size_t i;
    VIR_DEBUG("Mounting %s %d", dstprefix, skipRoot);
1360 1361

    /* Pull in rest of container's mounts */
1362
    for (i = 0 ; i < vmDef->nfss ; i++) {
1363 1364
        if (skipRoot &&
            STREQ(vmDef->fss[i]->dst, "/"))
1365 1366
            continue;

1367 1368 1369 1370
        if (lxcContainerUnmountSubtree(vmDef->fss[i]->dst,
                                       false) < 0)
            return -1;

1371
        if (lxcContainerMountFS(vmDef->fss[i], dstprefix, sec_mount_options) < 0)
1372 1373 1374
            return -1;
    }

1375
    VIR_DEBUG("Mounted all filesystems");
1376 1377 1378 1379
    return 0;
}


1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479
static int lxcContainerSetupDisk(virDomainDefPtr vmDef,
                                 virDomainDiskDefPtr def,
                                 const char *dstprefix,
                                 virSecurityManagerPtr securityDriver)
{
    char *src = NULL;
    char *dst = NULL;
    int ret = -1;
    struct stat sb;
    mode_t mode;
    char *tmpsrc = def->src;

    if (def->type != VIR_DOMAIN_DISK_TYPE_BLOCK) {
        virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s",
                       _("Can't setup disk for non-block device"));
        goto cleanup;
    }
    if (def->src == NULL) {
        virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s",
                       _("Can't setup disk without media"));
        goto cleanup;
    }

    if (virAsprintf(&src, "%s/%s", dstprefix, def->src) < 0) {
        virReportOOMError();
        goto cleanup;
    }

    if (virAsprintf(&dst, "/dev/%s", def->dst) < 0) {
        virReportOOMError();
        goto cleanup;
    }

    if (stat(src, &sb) < 0) {
        virReportSystemError(errno,
                             _("Unable to access %s"), def->src);
        goto cleanup;
    }

    if (!S_ISCHR(sb.st_mode) && !S_ISBLK(sb.st_mode)) {
        virReportError(VIR_ERR_CONFIG_UNSUPPORTED,
                       _("Disk source %s must be a character/block device"),
                       def->src);
        goto cleanup;
    }

    mode = 0700;
    if (S_ISCHR(sb.st_mode))
        mode |= S_IFCHR;
    else
        mode |= S_IFBLK;

    /* Yes, the device name we're creating may not
     * actually correspond to the major:minor number
     * we're using, but we've no other option at this
     * time. Just have to hope that containerized apps
     * don't get upset that the major:minor is different
     * to that normally implied by the device name
     */
    VIR_DEBUG("Creating dev %s (%d,%d) from %s",
              dst, major(sb.st_rdev), minor(sb.st_rdev), src);
    if (mknod(dst, mode, sb.st_rdev) < 0) {
        virReportSystemError(errno,
                             _("Unable to create device %s"),
                             dst);
        goto cleanup;
    }
    /* Labelling normally operates on src, but we need
     * to actally label the dst here, so hack the config */
    def->src = dst;
    if (virSecurityManagerSetImageLabel(securityDriver, vmDef, def) < 0)
        goto cleanup;

    ret = 0;

cleanup:
    def->src = tmpsrc;
    VIR_FREE(src);
    VIR_FREE(dst);
    return ret;
}

static int lxcContainerSetupAllDisks(virDomainDefPtr vmDef,
                                     const char *dstprefix,
                                     virSecurityManagerPtr securityDriver)
{
    size_t i;
    VIR_DEBUG("Setting up disks %s", dstprefix);

    for (i = 0 ; i < vmDef->ndisks ; i++) {
        if (lxcContainerSetupDisk(vmDef, vmDef->disks[i],
                                  dstprefix, securityDriver) < 0)
            return -1;
    }

    VIR_DEBUG("Setup all disks");
    return 0;
}


1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552
static int lxcContainerSetupHostdevSubsysUSB(virDomainDefPtr vmDef ATTRIBUTE_UNUSED,
                                             virDomainHostdevDefPtr def ATTRIBUTE_UNUSED,
                                             const char *dstprefix ATTRIBUTE_UNUSED,
                                             virSecurityManagerPtr securityDriver ATTRIBUTE_UNUSED)
{
    int ret = -1;
    char *src = NULL;
    char *dstdir = NULL;
    char *dstfile = NULL;
    struct stat sb;
    mode_t mode;

    if (virAsprintf(&dstdir, USB_DEVFS "/%03d",
                    def->source.subsys.u.usb.bus) < 0) {
        virReportOOMError();
        goto cleanup;
    }

    if (virAsprintf(&dstfile, "%s/%03d",
                    dstdir,
                    def->source.subsys.u.usb.device) < 0) {
        virReportOOMError();
        goto cleanup;
    }

    if (virAsprintf(&src, "%s/%s", dstprefix, dstfile) < 0) {
        virReportOOMError();
        goto cleanup;
    }

    if (stat(src, &sb) < 0) {
        virReportSystemError(errno,
                             _("Unable to access %s"), src);
        goto cleanup;
    }

    if (!S_ISCHR(sb.st_mode)) {
        virReportError(VIR_ERR_CONFIG_UNSUPPORTED,
                       _("USB source %s was not a character device"),
                       src);
        goto cleanup;
    }

    mode = 0700 | S_IFCHR;

    if (virFileMakePath(dstdir) < 0) {
        virReportSystemError(errno,
                             _("Unable to create %s"), dstdir);
        goto cleanup;
    }

    VIR_DEBUG("Creating dev %s (%d,%d)",
              dstfile, major(sb.st_rdev), minor(sb.st_rdev));
    if (mknod(dstfile, mode, sb.st_rdev) < 0) {
        virReportSystemError(errno,
                             _("Unable to create device %s"),
                             dstfile);
        goto cleanup;
    }

    if (virSecurityManagerSetHostdevLabel(securityDriver, vmDef, def, NULL) < 0)
        goto cleanup;

    ret = 0;

cleanup:
    VIR_FREE(src);
    VIR_FREE(dstfile);
    VIR_FREE(dstdir);
    return ret;
}


1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610
static int lxcContainerSetupHostdevCapsStorage(virDomainDefPtr vmDef ATTRIBUTE_UNUSED,
                                               virDomainHostdevDefPtr def ATTRIBUTE_UNUSED,
                                               const char *dstprefix ATTRIBUTE_UNUSED,
                                               virSecurityManagerPtr securityDriver ATTRIBUTE_UNUSED)
{
    char *src = NULL;
    int ret = -1;
    struct stat sb;
    mode_t mode;

    if (def->source.caps.u.storage.block == NULL) {
        virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s",
                       _("Missing storage host block path"));
        goto cleanup;
    }

    if (virAsprintf(&src, "%s/%s", dstprefix, def->source.caps.u.storage.block) < 0) {
        virReportOOMError();
        goto cleanup;
    }

    if (stat(src, &sb) < 0) {
        virReportSystemError(errno,
                             _("Unable to access %s"),
                             src);
        goto cleanup;
    }

    if (!S_ISBLK(sb.st_mode)) {
        virReportError(VIR_ERR_CONFIG_UNSUPPORTED,
                       _("Storage source %s must be a block device"),
                       def->source.caps.u.storage.block);
        goto cleanup;
    }

    mode = 0700 | S_IFBLK;

    VIR_DEBUG("Creating dev %s (%d,%d)",
              def->source.caps.u.storage.block,
              major(sb.st_rdev), minor(sb.st_rdev));
    if (mknod(def->source.caps.u.storage.block, mode, sb.st_rdev) < 0) {
        virReportSystemError(errno,
                             _("Unable to create device %s"),
                             def->source.caps.u.storage.block);
        goto cleanup;
    }

    if (virSecurityManagerSetHostdevLabel(securityDriver, vmDef, def, NULL) < 0)
        goto cleanup;

    ret = 0;

cleanup:
    VIR_FREE(src);
    return ret;
}


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
static int lxcContainerSetupHostdevCapsMisc(virDomainDefPtr vmDef ATTRIBUTE_UNUSED,
                                            virDomainHostdevDefPtr def ATTRIBUTE_UNUSED,
                                            const char *dstprefix ATTRIBUTE_UNUSED,
                                            virSecurityManagerPtr securityDriver ATTRIBUTE_UNUSED)
{
    char *src = NULL;
    int ret = -1;
    struct stat sb;
    mode_t mode;

    if (def->source.caps.u.misc.chardev == NULL) {
        virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s",
                       _("Missing storage host block path"));
        goto cleanup;
    }

    if (virAsprintf(&src, "%s/%s", dstprefix, def->source.caps.u.misc.chardev) < 0) {
        virReportOOMError();
        goto cleanup;
    }

    if (stat(src, &sb) < 0) {
        virReportSystemError(errno,
                             _("Unable to access %s"),
                             src);
        goto cleanup;
    }

    if (!S_ISCHR(sb.st_mode)) {
        virReportError(VIR_ERR_CONFIG_UNSUPPORTED,
                       _("Storage source %s must be a character device"),
                       def->source.caps.u.misc.chardev);
        goto cleanup;
    }

    mode = 0700 | S_IFCHR;

    VIR_DEBUG("Creating dev %s (%d,%d)",
              def->source.caps.u.misc.chardev,
              major(sb.st_rdev), minor(sb.st_rdev));
    if (mknod(def->source.caps.u.misc.chardev, mode, sb.st_rdev) < 0) {
        virReportSystemError(errno,
                             _("Unable to create device %s"),
                             def->source.caps.u.misc.chardev);
        goto cleanup;
    }

    if (virSecurityManagerSetHostdevLabel(securityDriver, vmDef, def, NULL) < 0)
        goto cleanup;

    ret = 0;

cleanup:
    VIR_FREE(src);
    return ret;
}

1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685
static int lxcContainerSetupHostdevSubsys(virDomainDefPtr vmDef,
                                          virDomainHostdevDefPtr def,
                                          const char *dstprefix,
                                          virSecurityManagerPtr securityDriver)
{
    switch (def->source.subsys.type) {
    case VIR_DOMAIN_HOSTDEV_SUBSYS_TYPE_USB:
        return lxcContainerSetupHostdevSubsysUSB(vmDef, def, dstprefix, securityDriver);

    default:
        virReportError(VIR_ERR_CONFIG_UNSUPPORTED,
                       _("Unsupported host device mode %s"),
                       virDomainHostdevSubsysTypeToString(def->source.subsys.type));
        return -1;
    }
}


1686 1687 1688 1689 1690 1691 1692 1693 1694
static int lxcContainerSetupHostdevCaps(virDomainDefPtr vmDef,
                                        virDomainHostdevDefPtr def,
                                        const char *dstprefix,
                                        virSecurityManagerPtr securityDriver)
{
    switch (def->source.subsys.type) {
    case VIR_DOMAIN_HOSTDEV_CAPS_TYPE_STORAGE:
        return lxcContainerSetupHostdevCapsStorage(vmDef, def, dstprefix, securityDriver);

1695 1696 1697
    case VIR_DOMAIN_HOSTDEV_CAPS_TYPE_MISC:
        return lxcContainerSetupHostdevCapsMisc(vmDef, def, dstprefix, securityDriver);

1698 1699 1700
    case VIR_DOMAIN_HOSTDEV_CAPS_TYPE_NET:
        return 0; // case is handled in virLXCControllerMoveInterfaces

1701 1702 1703 1704 1705 1706 1707 1708 1709
    default:
        virReportError(VIR_ERR_CONFIG_UNSUPPORTED,
                       _("Unsupported host device mode %s"),
                       virDomainHostdevCapsTypeToString(def->source.subsys.type));
        return -1;
    }
}


1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723
static int lxcContainerSetupAllHostdevs(virDomainDefPtr vmDef,
                                        const char *dstprefix,
                                        virSecurityManagerPtr securityDriver)
{
    size_t i;
    VIR_DEBUG("Setting up hostdevs %s", dstprefix);

    for (i = 0 ; i < vmDef->nhostdevs ; i++) {
        virDomainHostdevDefPtr def = vmDef->hostdevs[i];
        switch (def->mode) {
        case VIR_DOMAIN_HOSTDEV_MODE_SUBSYS:
            if (lxcContainerSetupHostdevSubsys(vmDef, def, dstprefix, securityDriver) < 0)
                return -1;
            break;
1724 1725 1726 1727
        case VIR_DOMAIN_HOSTDEV_MODE_CAPABILITIES:
            if (lxcContainerSetupHostdevCaps(vmDef, def, dstprefix, securityDriver) < 0)
                return -1;
            break;
1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740
        default:
            virReportError(VIR_ERR_CONFIG_UNSUPPORTED,
                           _("Unsupported host device mode %s"),
                           virDomainHostdevModeTypeToString(def->mode));
            return -1;
        }
    }

    VIR_DEBUG("Setup all hostdevs");
    return 0;
}


1741 1742 1743 1744 1745
/* Got a FS mapped to /, we're going the pivot_root
 * approach to do a better-chroot-than-chroot
 * this is based on this thread http://lkml.org/lkml/2008/3/5/29
 */
static int lxcContainerSetupPivotRoot(virDomainDefPtr vmDef,
1746 1747
                                      virDomainFSDefPtr root,
                                      char **ttyPaths,
1748
                                      size_t nttyPaths,
1749
                                      virSecurityManagerPtr securityDriver)
1750
{
1751 1752
    virCgroupPtr cgroup = NULL;
    int rc;
1753
    int ret = -1;
1754 1755 1756 1757
    char *sec_mount_options;

    if (!(sec_mount_options = virSecurityManagerGetMountOptions(securityDriver, vmDef)))
        return -1;
1758 1759 1760

    /* Before pivoting we need to identify any
     * cgroups controllers that are mounted */
1761 1762 1763
    if ((rc = virCgroupNewSelf(&cgroup)) != 0) {
        virReportSystemError(-rc, "%s",
                             _("Cannot identify cgroup placement"));
1764
        goto cleanup;
1765
    }
1766

1767 1768 1769 1770
    /* Ensure the root filesystem is mounted */
    if (lxcContainerPrepareRoot(vmDef, root) < 0)
        goto cleanup;

1771
    /* Gives us a private root, leaving all parent OS mounts on /.oldroot */
1772
    if (lxcContainerPivotRoot(root) < 0)
1773
        goto cleanup;
1774

1775
#if WITH_SELINUX
1776 1777 1778
    /* Some versions of Linux kernel don't let you overmount
     * the selinux filesystem, so make sure we kill it first
     */
1779 1780
    if (STREQ(root->src, "/") &&
        lxcContainerUnmountSubtree(SELINUX_MOUNT, false) < 0)
1781 1782 1783
        goto cleanup;
#endif

1784 1785
    /* If we have the root source being '/', then we need to
     * get rid of any existing stuff under /proc, /sys & /tmp.
1786 1787
     * We need new namespace aware versions of those. We must
     * do /proc last otherwise we won't find /proc/mounts :-) */
1788 1789 1790 1791
    if (STREQ(root->src, "/") &&
        (lxcContainerUnmountSubtree("/sys", false) < 0 ||
         lxcContainerUnmountSubtree("/dev", false) < 0 ||
         lxcContainerUnmountSubtree("/proc", false) < 0))
1792 1793
        goto cleanup;

1794
    /* Mounts the core /proc, /sys, etc filesystems */
1795
    if (lxcContainerMountBasicFS(true, sec_mount_options) < 0)
1796 1797
        goto cleanup;

1798 1799 1800 1801
    /* Mounts /proc/meminfo etc sysinfo */
    if (lxcContainerMountProcFuse(vmDef, "/.oldroot") < 0)
        goto cleanup;

1802 1803
    /* Now we can re-mount the cgroups controllers in the
     * same configuration as before */
1804
    if (virCgroupIsolateMount(cgroup, "/.oldroot/", sec_mount_options) < 0)
1805
        goto cleanup;
1806

1807
    /* Mounts /dev/pts */
1808
    if (lxcContainerMountFSDevPTS(vmDef, "/.oldroot") < 0)
1809
        goto cleanup;
1810

1811
    /* Populates device nodes in /dev/ */
1812
    if (lxcContainerPopulateDevices(ttyPaths, nttyPaths) < 0)
1813
        goto cleanup;
1814

1815
    /* Sets up any non-root mounts from guest config */
1816
    if (lxcContainerMountAllFS(vmDef, "/.oldroot", true, sec_mount_options) < 0)
1817
        goto cleanup;
1818

1819 1820 1821 1822
    /* Sets up any extra disks from guest config */
    if (lxcContainerSetupAllDisks(vmDef, "/.oldroot", securityDriver) < 0)
        goto cleanup;

1823 1824 1825 1826 1827
    /* Sets up any extra host devices from guest config */
    if (lxcContainerSetupAllHostdevs(vmDef, "/.oldroot", securityDriver) < 0)
        goto cleanup;

   /* Gets rid of all remaining mounts from host OS, including /.oldroot itself */
1828
    if (lxcContainerUnmountSubtree("/.oldroot", true) < 0)
1829
        goto cleanup;
1830

1831 1832 1833
    ret = 0;

cleanup:
1834
    virCgroupFree(&cgroup);
1835
    VIR_FREE(sec_mount_options);
1836
    return ret;
1837 1838
}

1839

1840 1841 1842 1843 1844 1845 1846
static int lxcContainerResolveSymlinks(virDomainDefPtr vmDef)
{
    char *newroot;
    size_t i;

    for (i = 0 ; i < vmDef->nfss ; i++) {
        virDomainFSDefPtr fs = vmDef->fss[i];
1847 1848
        if (!fs->src)
            continue;
1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860
        if (virFileResolveAllLinks(fs->src, &newroot) < 0)
            return -1;

        VIR_DEBUG("Resolved '%s' to %s", fs->src, newroot);

        VIR_FREE(fs->src);
        fs->src = newroot;
    }

    return 0;
}

D
Daniel P. Berrange 已提交
1861 1862 1863 1864 1865
/*
 * This is running as the 'init' process insid the container.
 * It removes some capabilities that could be dangerous to
 * host system, since they are not currently "containerized"
 */
E
Eric Blake 已提交
1866
static int lxcContainerDropCapabilities(bool keepReboot ATTRIBUTE_UNUSED)
1867
{
1868
#if WITH_CAPNG
D
Daniel P. Berrange 已提交
1869 1870 1871 1872 1873 1874 1875 1876 1877
    int ret;

    capng_get_caps_process();

    if ((ret = capng_updatev(CAPNG_DROP,
                             CAPNG_EFFECTIVE | CAPNG_PERMITTED |
                             CAPNG_INHERITABLE | CAPNG_BOUNDING_SET,
                             CAP_SYS_MODULE, /* No kernel module loading */
                             CAP_SYS_TIME, /* No changing the clock */
1878
                             CAP_MKNOD, /* No creating device nodes */
D
Daniel P. Berrange 已提交
1879 1880
                             CAP_AUDIT_CONTROL, /* No messing with auditing status */
                             CAP_MAC_ADMIN, /* No messing with LSM config */
1881
                             keepReboot ? -1 : CAP_SYS_BOOT, /* No use of reboot */
1882
                             -1)) < 0) {
1883 1884
        virReportError(VIR_ERR_INTERNAL_ERROR,
                       _("Failed to remove capabilities: %d"), ret);
D
Daniel P. Berrange 已提交
1885 1886
        return -1;
    }
1887

D
Daniel P. Berrange 已提交
1888
    if ((ret = capng_apply(CAPNG_SELECT_BOTH)) < 0) {
1889 1890
        virReportError(VIR_ERR_INTERNAL_ERROR,
                       _("Failed to apply capabilities: %d"), ret);
D
Daniel P. Berrange 已提交
1891
        return -1;
1892
    }
D
Daniel P. Berrange 已提交
1893

1894 1895 1896 1897 1898
    /* We do not need to call capng_lock() in this case. The bounding
     * set restriction will prevent them reacquiring sys_boot/module/time,
     * etc which is all that matters for the container. Once inside the
     * container it is fine for SECURE_NOROOT / SECURE_NO_SETUID_FIXUP to
     * be unmasked  - they can never escape the bounding set. */
D
Daniel P. Berrange 已提交
1899 1900

#else
1901
    VIR_WARN("libcap-ng support not compiled in, unable to clear capabilities");
D
Daniel Veillard 已提交
1902
#endif
1903 1904 1905 1906
    return 0;
}


1907
/**
1908 1909
 * lxcContainerChild:
 * @data: pointer to container arguments
1910 1911 1912 1913 1914 1915 1916 1917 1918
 *
 * This function is run in the process clone()'d in lxcStartContainer.
 * Perform a number of container setup tasks:
 *     Setup container file system
 *     mount container /proca
 * Then exec's the container init
 *
 * Returns 0 on success or -1 in case of error
 */
1919
static int lxcContainerChild(void *data)
1920
{
1921
    lxc_child_argv_t *argv = data;
1922
    virDomainDefPtr vmDef = argv->config;
1923
    int ttyfd = -1;
1924
    int ret = -1;
1925
    char *ttyPath = NULL;
1926
    virDomainFSDefPtr root;
1927
    virCommandPtr cmd = NULL;
1928
    int hasReboot;
1929 1930

    if (NULL == vmDef) {
1931 1932
        virReportError(VIR_ERR_INTERNAL_ERROR,
                       "%s", _("lxcChild() passed invalid vm definition"));
1933
        goto cleanup;
1934 1935
    }

1936 1937 1938
    if ((hasReboot = lxcContainerHasReboot()) < 0)
        goto cleanup;

1939 1940 1941
    cmd = lxcContainerBuildInitCmd(vmDef);
    virCommandWriteArgLog(cmd, 1);

1942
    root = virDomainGetRootFilesystem(vmDef);
1943

1944 1945
    if (argv->nttyPaths) {
        if (root) {
1946 1947 1948 1949 1950
            const char *tty = argv->ttyPaths[0];
            if (STRPREFIX(tty, "/dev/pts/"))
                tty += strlen("/dev/pts/");
            if (virAsprintf(&ttyPath, "%s/%s.devpts/%s",
                            LXC_STATE_DIR, vmDef->name, tty) < 0) {
1951 1952 1953 1954 1955 1956 1957 1958
                virReportOOMError();
                goto cleanup;
            }
        } else {
            if (!(ttyPath = strdup(argv->ttyPaths[0]))) {
                virReportOOMError();
                goto cleanup;
            }
1959 1960
        }
    } else {
1961
        if (!(ttyPath = strdup("/dev/null"))) {
1962
            virReportOOMError();
1963
            goto cleanup;
1964 1965
        }
    }
1966

1967
    VIR_DEBUG("Container TTY path: %s", ttyPath);
1968 1969

    ttyfd = open(ttyPath, O_RDWR|O_NOCTTY);
1970
    if (ttyfd < 0) {
1971
        virReportSystemError(errno,
1972
                             _("Failed to open tty %s"),
1973
                             ttyPath);
1974
        goto cleanup;
1975
    }
1976

1977 1978 1979 1980 1981 1982
    if (lxcContainerResolveSymlinks(vmDef) < 0)
        goto cleanup;

    if (lxcContainerSetupPivotRoot(vmDef, root,
                                   argv->ttyPaths, argv->nttyPaths,
                                   argv->securityDriver) < 0)
1983
        goto cleanup;
1984

1985 1986 1987 1988 1989 1990 1991
    if (!virFileExists(vmDef->os.init)) {
        virReportSystemError(errno,
                    _("cannot find init path '%s' relative to container root"),
                    vmDef->os.init);
        goto cleanup;
    }

1992
    /* Wait for interface devices to show up */
1993 1994 1995
    if (lxcContainerWaitForContinue(argv->monitor) < 0) {
        virReportSystemError(errno, "%s",
                             _("Failed to read the container continue message"));
1996
        goto cleanup;
1997 1998
    }
    VIR_DEBUG("Received container continue message");
1999

2000
    /* rename and enable interfaces */
2001 2002 2003
    if (lxcContainerRenameAndEnableInterfaces(!!(vmDef->features &
                                                 (1 << VIR_DOMAIN_FEATURE_PRIVNET)),
                                              argv->nveths,
2004
                                              argv->veths) < 0) {
2005
        goto cleanup;
2006
    }
2007

2008
    /* drop a set of root capabilities */
2009
    if (lxcContainerDropCapabilities(!!hasReboot) < 0)
2010
        goto cleanup;
2011

2012 2013 2014 2015 2016 2017
    if (lxcContainerSendContinue(argv->handshakefd) < 0) {
        virReportSystemError(errno, "%s",
                            _("failed to send continue signal to controller"));
        goto cleanup;
    }

2018 2019 2020 2021
    VIR_DEBUG("Setting up security labeling");
    if (virSecurityManagerSetProcessLabel(argv->securityDriver, vmDef) < 0)
        goto cleanup;

2022 2023 2024 2025
    if (lxcContainerSetStdio(argv->monitor, ttyfd, argv->handshakefd) < 0) {
        goto cleanup;
    }

2026
    ret = 0;
2027
cleanup:
2028 2029
    VIR_FREE(ttyPath);
    VIR_FORCE_CLOSE(ttyfd);
2030
    VIR_FORCE_CLOSE(argv->monitor);
2031
    VIR_FORCE_CLOSE(argv->handshakefd);
2032 2033

    if (ret == 0) {
E
Eric Blake 已提交
2034
        /* this function will only return if an error occurred */
2035 2036 2037
        ret = virCommandExec(cmd);
    }

2038 2039
    virCommandFree(cmd);
    return ret;
2040
}
2041

2042 2043
static int userns_supported(void)
{
2044 2045 2046 2047 2048 2049
#if 1
    /*
     * put off using userns until uid mapping is implemented
     */
    return 0;
#else
2050
    return lxcContainerAvailable(LXC_CONTAINER_FEATURE_USER) == 0;
2051
#endif
2052 2053
}

2054
virArch lxcContainerGetAlt32bitArch(virArch arch)
2055 2056 2057
{
    /* Any Linux 64bit arch which has a 32bit
     * personality available should be listed here */
2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073
    if (arch == VIR_ARCH_X86_64)
        return VIR_ARCH_I686;
    if (arch == VIR_ARCH_S390X)
        return VIR_ARCH_S390;
    if (arch == VIR_ARCH_PPC64)
        return VIR_ARCH_PPC;
    if (arch == VIR_ARCH_PARISC64)
        return VIR_ARCH_PARISC;
    if (arch == VIR_ARCH_SPARC64)
        return VIR_ARCH_SPARC;
    if (arch == VIR_ARCH_MIPS64)
        return VIR_ARCH_MIPS;
    if (arch == VIR_ARCH_MIPS64EL)
        return VIR_ARCH_MIPSEL;

    return VIR_ARCH_NONE;
2074 2075 2076
}


2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092
static bool
lxcNeedNetworkNamespace(virDomainDefPtr def)
{
    size_t i;
    if (def->nets != NULL)
        return true;
    if (def->features & (1 << VIR_DOMAIN_FEATURE_PRIVNET))
        return true;
    for (i = 0 ; i < def->nhostdevs ; i++) {
        if (def->hostdevs[i]->mode == VIR_DOMAIN_HOSTDEV_MODE_CAPABILITIES &&
            def->hostdevs[i]->source.caps.type == VIR_DOMAIN_HOSTDEV_CAPS_TYPE_NET)
            return true;
    }
    return false;
}

2093 2094
/**
 * lxcContainerStart:
2095 2096 2097 2098 2099
 * @def: pointer to virtual machine structure
 * @nveths: number of interfaces
 * @veths: interface names
 * @control: control FD to the container
 * @ttyPath: path of tty to set as the container console
2100 2101 2102 2103 2104
 *
 * Starts a container process by calling clone() with the namespace flags
 *
 * Returns PID of container on success or -1 in case of error
 */
2105
int lxcContainerStart(virDomainDefPtr def,
2106
                      virSecurityManagerPtr securityDriver,
2107
                      size_t nveths,
2108
                      char **veths,
2109
                      int control,
2110
                      int handshakefd,
2111 2112
                      char **ttyPaths,
                      size_t nttyPaths)
2113 2114
{
    pid_t pid;
E
Eric Blake 已提交
2115
    int cflags;
2116 2117
    int stacksize = getpagesize() * 4;
    char *stack, *stacktop;
2118 2119
    lxc_child_argv_t args = { def, securityDriver,
                              nveths, veths, control,
2120
                              ttyPaths, nttyPaths, handshakefd};
2121 2122 2123

    /* allocate a stack for the container */
    if (VIR_ALLOC_N(stack, stacksize) < 0) {
2124
        virReportOOMError();
2125 2126 2127 2128
        return -1;
    }
    stacktop = stack + stacksize;

E
Eric Blake 已提交
2129
    cflags = CLONE_NEWPID|CLONE_NEWNS|CLONE_NEWUTS|CLONE_NEWIPC|SIGCHLD;
2130

2131
    if (userns_supported()) {
2132
        VIR_DEBUG("Enable user namespaces");
E
Eric Blake 已提交
2133
        cflags |= CLONE_NEWUSER;
2134
    }
2135

2136
    if (lxcNeedNetworkNamespace(def)) {
2137
        VIR_DEBUG("Enable network namespaces");
E
Eric Blake 已提交
2138
        cflags |= CLONE_NEWNET;
2139
    }
2140

E
Eric Blake 已提交
2141
    pid = clone(lxcContainerChild, stacktop, cflags, &args);
2142
    VIR_FREE(stack);
2143
    VIR_DEBUG("clone() completed, new container PID is %d", pid);
2144 2145

    if (pid < 0) {
2146
        virReportSystemError(errno, "%s",
2147
                             _("Failed to run clone container"));
2148 2149 2150 2151 2152 2153
        return -1;
    }

    return pid;
}

2154 2155
ATTRIBUTE_NORETURN static int
lxcContainerDummyChild(void *argv ATTRIBUTE_UNUSED)
2156 2157 2158 2159 2160 2161
{
    _exit(0);
}

int lxcContainerAvailable(int features)
{
2162
    int flags = CLONE_NEWPID|CLONE_NEWNS|CLONE_NEWUTS|
2163 2164 2165 2166 2167
        CLONE_NEWIPC|SIGCHLD;
    int cpid;
    char *childStack;
    char *stack;

2168 2169 2170
    if (features & LXC_CONTAINER_FEATURE_USER)
        flags |= CLONE_NEWUSER;

2171 2172 2173 2174
    if (features & LXC_CONTAINER_FEATURE_NET)
        flags |= CLONE_NEWNET;

    if (VIR_ALLOC_N(stack, getpagesize() * 4) < 0) {
2175
        VIR_DEBUG("Unable to allocate stack");
2176 2177 2178 2179 2180 2181 2182 2183
        return -1;
    }

    childStack = stack + (getpagesize() * 4);

    cpid = clone(lxcContainerDummyChild, childStack, flags, NULL);
    VIR_FREE(stack);
    if (cpid < 0) {
2184
        char ebuf[1024] ATTRIBUTE_UNUSED;
2185
        VIR_DEBUG("clone call returned %s, container support is not enabled",
2186
                  virStrerror(errno, ebuf, sizeof(ebuf)));
2187
        return -1;
2188
    } else if (virProcessWait(cpid, NULL) < 0) {
E
Eric Blake 已提交
2189
        return -1;
2190 2191
    }

G
Gao feng 已提交
2192
    VIR_DEBUG("container support is enabled");
2193
    return 0;
2194
}