vdi.c 24.7 KB
Newer Older
1 2 3
/*
 * Block driver for the Virtual Disk Image (VDI) format
 *
4
 * Copyright (c) 2009, 2012 Stefan Weil
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52
 *
 * This program is free software: you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation, either version 2 of the License, or
 * (at your option) version 3 or any later version.
 *
 * This program 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 General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program.  If not, see <http://www.gnu.org/licenses/>.
 *
 * Reference:
 * http://forums.virtualbox.org/viewtopic.php?t=8046
 *
 * This driver supports create / read / write operations on VDI images.
 *
 * Todo (see also TODO in code):
 *
 * Some features like snapshots are still missing.
 *
 * Deallocation of zero-filled blocks and shrinking images are missing, too
 * (might be added to common block layer).
 *
 * Allocation of blocks could be optimized (less writes to block map and
 * header).
 *
 * Read and write of adjacents blocks could be done in one operation
 * (current code uses one operation per block (1 MiB).
 *
 * The code is not thread safe (missing locks for changes in header and
 * block table, no problem with current QEMU).
 *
 * Hints:
 *
 * Blocks (VDI documentation) correspond to clusters (QEMU).
 * QEMU's backing files could be implemented using VDI snapshot files (TODO).
 * VDI snapshot files may also contain the complete machine state.
 * Maybe this machine state can be converted to QEMU PC machine snapshot data.
 *
 * The driver keeps a block cache (little endian entries) in memory.
 * For the standard block size (1 MiB), a 1 TiB disk will use 4 MiB RAM,
 * so this seems to be reasonable.
 */

#include "qemu-common.h"
53
#include "block/block_int.h"
54
#include "qemu/module.h"
55
#include "migration/migration.h"
56

57
#if defined(CONFIG_UUID)
58 59 60
#include <uuid/uuid.h>
#else
/* TODO: move uuid emulation to some central place in QEMU. */
61
#include "sysemu/sysemu.h"     /* UUID_FMT */
62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87
typedef unsigned char uuid_t[16];
#endif

/* Code configuration options. */

/* Enable debug messages. */
//~ #define CONFIG_VDI_DEBUG

/* Support write operations on VDI images. */
#define CONFIG_VDI_WRITE

/* Support non-standard block (cluster) size. This is untested.
 * Maybe it will be needed for very large images.
 */
//~ #define CONFIG_VDI_BLOCK_SIZE

/* Support static (fixed, pre-allocated) images. */
#define CONFIG_VDI_STATIC_IMAGE

/* Command line option for static images. */
#define BLOCK_OPT_STATIC "static"

#define KiB     1024
#define MiB     (KiB * KiB)

#define SECTOR_SIZE 512
88
#define DEFAULT_CLUSTER_SIZE (1 * MiB)
89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114

#if defined(CONFIG_VDI_DEBUG)
#define logout(fmt, ...) \
                fprintf(stderr, "vdi\t%-24s" fmt, __func__, ##__VA_ARGS__)
#else
#define logout(fmt, ...) ((void)0)
#endif

/* Image signature. */
#define VDI_SIGNATURE 0xbeda107f

/* Image version. */
#define VDI_VERSION_1_1 0x00010001

/* Image type. */
#define VDI_TYPE_DYNAMIC 1
#define VDI_TYPE_STATIC  2

/* Innotek / SUN images use these strings in header.text:
 * "<<< innotek VirtualBox Disk Image >>>\n"
 * "<<< Sun xVM VirtualBox Disk Image >>>\n"
 * "<<< Sun VirtualBox Disk Image >>>\n"
 * The value does not matter, so QEMU created images use a different text.
 */
#define VDI_TEXT "<<< QEMU VM Virtual Disk Image >>>\n"

115 116 117 118 119 120 121
/* A never-allocated block; semantically arbitrary content. */
#define VDI_UNALLOCATED 0xffffffffU

/* A discarded (no longer allocated) block; semantically zero-filled. */
#define VDI_DISCARDED   0xfffffffeU

#define VDI_IS_ALLOCATED(X) ((X) < VDI_DISCARDED)
122

123
#if !defined(CONFIG_UUID)
124
static inline void uuid_generate(uuid_t out)
125
{
126
    memset(out, 0, sizeof(uuid_t));
127 128
}

129
static inline int uuid_is_null(const uuid_t uu)
130 131
{
    uuid_t null_uuid = { 0 };
132
    return memcmp(uu, null_uuid, sizeof(uuid_t)) == 0;
133 134
}

135
static inline void uuid_unparse(const uuid_t uu, char *out)
136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178
{
    snprintf(out, 37, UUID_FMT,
            uu[0], uu[1], uu[2], uu[3], uu[4], uu[5], uu[6], uu[7],
            uu[8], uu[9], uu[10], uu[11], uu[12], uu[13], uu[14], uu[15]);
}
#endif

typedef struct {
    char text[0x40];
    uint32_t signature;
    uint32_t version;
    uint32_t header_size;
    uint32_t image_type;
    uint32_t image_flags;
    char description[256];
    uint32_t offset_bmap;
    uint32_t offset_data;
    uint32_t cylinders;         /* disk geometry, unused here */
    uint32_t heads;             /* disk geometry, unused here */
    uint32_t sectors;           /* disk geometry, unused here */
    uint32_t sector_size;
    uint32_t unused1;
    uint64_t disk_size;
    uint32_t block_size;
    uint32_t block_extra;       /* unused here */
    uint32_t blocks_in_image;
    uint32_t blocks_allocated;
    uuid_t uuid_image;
    uuid_t uuid_last_snap;
    uuid_t uuid_link;
    uuid_t uuid_parent;
    uint64_t unused2[7];
} VdiHeader;

typedef struct {
    /* The block map entries are little endian (even in memory). */
    uint32_t *bmap;
    /* Size of block (bytes). */
    uint32_t block_size;
    /* Size of block (sectors). */
    uint32_t block_sectors;
    /* First sector of block map. */
    uint32_t bmap_sector;
S
Stefan Weil 已提交
179
    /* VDI header (converted to host endianness). */
180
    VdiHeader header;
K
Kevin Wolf 已提交
181 182

    Error *migration_blocker;
183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248
} BDRVVdiState;

/* Change UUID from little endian (IPRT = VirtualBox format) to big endian
 * format (network byte order, standard, see RFC 4122) and vice versa.
 */
static void uuid_convert(uuid_t uuid)
{
    bswap32s((uint32_t *)&uuid[0]);
    bswap16s((uint16_t *)&uuid[4]);
    bswap16s((uint16_t *)&uuid[6]);
}

static void vdi_header_to_cpu(VdiHeader *header)
{
    le32_to_cpus(&header->signature);
    le32_to_cpus(&header->version);
    le32_to_cpus(&header->header_size);
    le32_to_cpus(&header->image_type);
    le32_to_cpus(&header->image_flags);
    le32_to_cpus(&header->offset_bmap);
    le32_to_cpus(&header->offset_data);
    le32_to_cpus(&header->cylinders);
    le32_to_cpus(&header->heads);
    le32_to_cpus(&header->sectors);
    le32_to_cpus(&header->sector_size);
    le64_to_cpus(&header->disk_size);
    le32_to_cpus(&header->block_size);
    le32_to_cpus(&header->block_extra);
    le32_to_cpus(&header->blocks_in_image);
    le32_to_cpus(&header->blocks_allocated);
    uuid_convert(header->uuid_image);
    uuid_convert(header->uuid_last_snap);
    uuid_convert(header->uuid_link);
    uuid_convert(header->uuid_parent);
}

static void vdi_header_to_le(VdiHeader *header)
{
    cpu_to_le32s(&header->signature);
    cpu_to_le32s(&header->version);
    cpu_to_le32s(&header->header_size);
    cpu_to_le32s(&header->image_type);
    cpu_to_le32s(&header->image_flags);
    cpu_to_le32s(&header->offset_bmap);
    cpu_to_le32s(&header->offset_data);
    cpu_to_le32s(&header->cylinders);
    cpu_to_le32s(&header->heads);
    cpu_to_le32s(&header->sectors);
    cpu_to_le32s(&header->sector_size);
    cpu_to_le64s(&header->disk_size);
    cpu_to_le32s(&header->block_size);
    cpu_to_le32s(&header->block_extra);
    cpu_to_le32s(&header->blocks_in_image);
    cpu_to_le32s(&header->blocks_allocated);
    cpu_to_le32s(&header->blocks_allocated);
    uuid_convert(header->uuid_image);
    uuid_convert(header->uuid_last_snap);
    uuid_convert(header->uuid_link);
    uuid_convert(header->uuid_parent);
}

#if defined(CONFIG_VDI_DEBUG)
static void vdi_header_print(VdiHeader *header)
{
    char uuid[37];
    logout("text        %s", header->text);
249
    logout("signature   0x%08x\n", header->signature);
250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276
    logout("header size 0x%04x\n", header->header_size);
    logout("image type  0x%04x\n", header->image_type);
    logout("image flags 0x%04x\n", header->image_flags);
    logout("description %s\n", header->description);
    logout("offset bmap 0x%04x\n", header->offset_bmap);
    logout("offset data 0x%04x\n", header->offset_data);
    logout("cylinders   0x%04x\n", header->cylinders);
    logout("heads       0x%04x\n", header->heads);
    logout("sectors     0x%04x\n", header->sectors);
    logout("sector size 0x%04x\n", header->sector_size);
    logout("image size  0x%" PRIx64 " B (%" PRIu64 " MiB)\n",
           header->disk_size, header->disk_size / MiB);
    logout("block size  0x%04x\n", header->block_size);
    logout("block extra 0x%04x\n", header->block_extra);
    logout("blocks tot. 0x%04x\n", header->blocks_in_image);
    logout("blocks all. 0x%04x\n", header->blocks_allocated);
    uuid_unparse(header->uuid_image, uuid);
    logout("uuid image  %s\n", uuid);
    uuid_unparse(header->uuid_last_snap, uuid);
    logout("uuid snap   %s\n", uuid);
    uuid_unparse(header->uuid_link, uuid);
    logout("uuid link   %s\n", uuid);
    uuid_unparse(header->uuid_parent, uuid);
    logout("uuid parent %s\n", uuid);
}
#endif

277 278
static int vdi_check(BlockDriverState *bs, BdrvCheckResult *res,
                     BdrvCheckMode fix)
279 280 281 282 283 284 285 286
{
    /* TODO: additional checks possible. */
    BDRVVdiState *s = (BDRVVdiState *)bs->opaque;
    uint32_t blocks_allocated = 0;
    uint32_t block;
    uint32_t *bmap;
    logout("\n");

287 288 289 290
    if (fix) {
        return -ENOTSUP;
    }

291
    bmap = g_malloc(s->header.blocks_in_image * sizeof(uint32_t));
292 293 294 295 296
    memset(bmap, 0xff, s->header.blocks_in_image * sizeof(uint32_t));

    /* Check block map and value of blocks_allocated. */
    for (block = 0; block < s->header.blocks_in_image; block++) {
        uint32_t bmap_entry = le32_to_cpu(s->bmap[block]);
297
        if (VDI_IS_ALLOCATED(bmap_entry)) {
298 299
            if (bmap_entry < s->header.blocks_in_image) {
                blocks_allocated++;
300
                if (!VDI_IS_ALLOCATED(bmap[bmap_entry])) {
301 302 303 304
                    bmap[bmap_entry] = bmap_entry;
                } else {
                    fprintf(stderr, "ERROR: block index %" PRIu32
                            " also used by %" PRIu32 "\n", bmap[bmap_entry], bmap_entry);
305
                    res->corruptions++;
306 307 308 309
                }
            } else {
                fprintf(stderr, "ERROR: block index %" PRIu32
                        " too large, is %" PRIu32 "\n", block, bmap_entry);
310
                res->corruptions++;
311 312 313 314 315 316 317
            }
        }
    }
    if (blocks_allocated != s->header.blocks_allocated) {
        fprintf(stderr, "ERROR: allocated blocks mismatch, is %" PRIu32
               ", should be %" PRIu32 "\n",
               blocks_allocated, s->header.blocks_allocated);
318
        res->corruptions++;
319 320
    }

321
    g_free(bmap);
322

323
    return 0;
324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366
}

static int vdi_get_info(BlockDriverState *bs, BlockDriverInfo *bdi)
{
    /* TODO: vdi_get_info would be needed for machine snapshots.
       vm_state_offset is still missing. */
    BDRVVdiState *s = (BDRVVdiState *)bs->opaque;
    logout("\n");
    bdi->cluster_size = s->block_size;
    bdi->vm_state_offset = 0;
    return 0;
}

static int vdi_make_empty(BlockDriverState *bs)
{
    /* TODO: missing code. */
    logout("\n");
    /* The return value for missing code must be 0, see block.c. */
    return 0;
}

static int vdi_probe(const uint8_t *buf, int buf_size, const char *filename)
{
    const VdiHeader *header = (const VdiHeader *)buf;
    int result = 0;

    logout("\n");

    if (buf_size < sizeof(*header)) {
        /* Header too small, no VDI. */
    } else if (le32_to_cpu(header->signature) == VDI_SIGNATURE) {
        result = 100;
    }

    if (result == 0) {
        logout("no vdi image\n");
    } else {
        logout("%s", header->text);
    }

    return result;
}

367
static int vdi_open(BlockDriverState *bs, int flags)
368 369 370 371
{
    BDRVVdiState *s = bs->opaque;
    VdiHeader header;
    size_t bmap_size;
372
    int ret;
373 374 375

    logout("\n");

376 377
    ret = bdrv_read(bs->file, 0, (uint8_t *)&header, 1);
    if (ret < 0) {
378 379 380 381 382 383 384 385
        goto fail;
    }

    vdi_header_to_cpu(&header);
#if defined(CONFIG_VDI_DEBUG)
    vdi_header_print(&header);
#endif

386 387 388 389 390 391 392 393 394
    if (header.disk_size % SECTOR_SIZE != 0) {
        /* 'VBoxManage convertfromraw' can create images with odd disk sizes.
           We accept them but round the disk size to the next multiple of
           SECTOR_SIZE. */
        logout("odd disk size %" PRIu64 " B, round up\n", header.disk_size);
        header.disk_size += SECTOR_SIZE - 1;
        header.disk_size &= ~(SECTOR_SIZE - 1);
    }

S
Stefan Weil 已提交
395 396 397 398 399
    if (header.signature != VDI_SIGNATURE) {
        logout("bad vdi signature %08x\n", header.signature);
        ret = -EMEDIUMTYPE;
        goto fail;
    } else if (header.version != VDI_VERSION_1_1) {
400 401
        logout("unsupported version %u.%u\n",
               header.version >> 16, header.version & 0xffff);
402
        ret = -ENOTSUP;
403 404 405 406
        goto fail;
    } else if (header.offset_bmap % SECTOR_SIZE != 0) {
        /* We only support block maps which start on a sector boundary. */
        logout("unsupported block map offset 0x%x B\n", header.offset_bmap);
407
        ret = -ENOTSUP;
408 409 410 411
        goto fail;
    } else if (header.offset_data % SECTOR_SIZE != 0) {
        /* We only support data blocks which start on a sector boundary. */
        logout("unsupported data offset 0x%x B\n", header.offset_data);
412
        ret = -ENOTSUP;
413 414 415
        goto fail;
    } else if (header.sector_size != SECTOR_SIZE) {
        logout("unsupported sector size %u B\n", header.sector_size);
416
        ret = -ENOTSUP;
417 418 419
        goto fail;
    } else if (header.block_size != 1 * MiB) {
        logout("unsupported block size %u B\n", header.block_size);
420
        ret = -ENOTSUP;
421
        goto fail;
422 423 424
    } else if (header.disk_size >
               (uint64_t)header.blocks_in_image * header.block_size) {
        logout("unsupported disk size %" PRIu64 " B\n", header.disk_size);
425
        ret = -ENOTSUP;
426 427 428
        goto fail;
    } else if (!uuid_is_null(header.uuid_link)) {
        logout("link uuid != 0, unsupported\n");
429
        ret = -ENOTSUP;
430 431 432
        goto fail;
    } else if (!uuid_is_null(header.uuid_parent)) {
        logout("parent uuid != 0, unsupported\n");
433
        ret = -ENOTSUP;
434 435 436 437 438 439 440 441 442 443 444
        goto fail;
    }

    bs->total_sectors = header.disk_size / SECTOR_SIZE;

    s->block_size = header.block_size;
    s->block_sectors = header.block_size / SECTOR_SIZE;
    s->bmap_sector = header.offset_bmap / SECTOR_SIZE;
    s->header = header;

    bmap_size = header.blocks_in_image * sizeof(uint32_t);
S
Stefan Weil 已提交
445
    bmap_size = (bmap_size + SECTOR_SIZE - 1) / SECTOR_SIZE;
446
    s->bmap = g_malloc(bmap_size * SECTOR_SIZE);
447 448
    ret = bdrv_read(bs->file, s->bmap_sector, (uint8_t *)s->bmap, bmap_size);
    if (ret < 0) {
449 450 451
        goto fail_free_bmap;
    }

K
Kevin Wolf 已提交
452 453 454 455 456 457
    /* Disable migration when vdi images are used */
    error_set(&s->migration_blocker,
              QERR_BLOCK_FORMAT_FEATURE_NOT_SUPPORTED,
              "vdi", bs->device_name, "live migration");
    migrate_add_blocker(s->migration_blocker);

458 459 460
    return 0;

 fail_free_bmap:
461
    g_free(s->bmap);
462 463

 fail:
464
    return ret;
465 466
}

J
Jeff Cody 已提交
467 468 469 470 471 472
static int vdi_reopen_prepare(BDRVReopenState *state,
                              BlockReopenQueue *queue, Error **errp)
{
    return 0;
}

473 474
static int coroutine_fn vdi_co_is_allocated(BlockDriverState *bs,
        int64_t sector_num, int nb_sectors, int *pnum)
475 476 477 478 479 480 481 482 483 484 485 486
{
    /* TODO: Check for too large sector_num (in bdrv_is_allocated or here). */
    BDRVVdiState *s = (BDRVVdiState *)bs->opaque;
    size_t bmap_index = sector_num / s->block_sectors;
    size_t sector_in_block = sector_num % s->block_sectors;
    int n_sectors = s->block_sectors - sector_in_block;
    uint32_t bmap_entry = le32_to_cpu(s->bmap[bmap_index]);
    logout("%p, %" PRId64 ", %d, %p\n", bs, sector_num, nb_sectors, pnum);
    if (n_sectors > nb_sectors) {
        n_sectors = nb_sectors;
    }
    *pnum = n_sectors;
487
    return VDI_IS_ALLOCATED(bmap_entry);
488 489
}

490 491
static int vdi_co_read(BlockDriverState *bs,
        int64_t sector_num, uint8_t *buf, int nb_sectors)
492 493 494 495 496 497
{
    BDRVVdiState *s = bs->opaque;
    uint32_t bmap_entry;
    uint32_t block_index;
    uint32_t sector_in_block;
    uint32_t n_sectors;
P
Paolo Bonzini 已提交
498
    int ret = 0;
499 500

    logout("\n");
501

P
Paolo Bonzini 已提交
502 503 504 505 506 507 508
    while (ret >= 0 && nb_sectors > 0) {
        block_index = sector_num / s->block_sectors;
        sector_in_block = sector_num % s->block_sectors;
        n_sectors = s->block_sectors - sector_in_block;
        if (n_sectors > nb_sectors) {
            n_sectors = nb_sectors;
        }
509

P
Paolo Bonzini 已提交
510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525
        logout("will read %u sectors starting at sector %" PRIu64 "\n",
               n_sectors, sector_num);

        /* prepare next AIO request */
        bmap_entry = le32_to_cpu(s->bmap[block_index]);
        if (!VDI_IS_ALLOCATED(bmap_entry)) {
            /* Block not allocated, return zeros, no need to wait. */
            memset(buf, 0, n_sectors * SECTOR_SIZE);
            ret = 0;
        } else {
            uint64_t offset = s->header.offset_data / SECTOR_SIZE +
                              (uint64_t)bmap_entry * s->block_sectors +
                              sector_in_block;
            ret = bdrv_read(bs->file, offset, buf, n_sectors);
        }
        logout("%u sectors read\n", n_sectors);
526

P
Paolo Bonzini 已提交
527 528 529
        nb_sectors -= n_sectors;
        sector_num += n_sectors;
        buf += n_sectors * SECTOR_SIZE;
530
    }
531 532

    return ret;
533 534
}

535 536
static int vdi_co_write(BlockDriverState *bs,
        int64_t sector_num, const uint8_t *buf, int nb_sectors)
537 538 539 540 541 542
{
    BDRVVdiState *s = bs->opaque;
    uint32_t bmap_entry;
    uint32_t block_index;
    uint32_t sector_in_block;
    uint32_t n_sectors;
P
Paolo Bonzini 已提交
543 544 545
    uint32_t bmap_first = VDI_UNALLOCATED;
    uint32_t bmap_last = VDI_UNALLOCATED;
    uint8_t *block = NULL;
P
Paolo Bonzini 已提交
546
    int ret = 0;
547 548

    logout("\n");
549

P
Paolo Bonzini 已提交
550 551 552 553 554 555 556
    while (ret >= 0 && nb_sectors > 0) {
        block_index = sector_num / s->block_sectors;
        sector_in_block = sector_num % s->block_sectors;
        n_sectors = s->block_sectors - sector_in_block;
        if (n_sectors > nb_sectors) {
            n_sectors = nb_sectors;
        }
557

P
Paolo Bonzini 已提交
558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587
        logout("will write %u sectors starting at sector %" PRIu64 "\n",
               n_sectors, sector_num);

        /* prepare next AIO request */
        bmap_entry = le32_to_cpu(s->bmap[block_index]);
        if (!VDI_IS_ALLOCATED(bmap_entry)) {
            /* Allocate new block and write to it. */
            uint64_t offset;
            bmap_entry = s->header.blocks_allocated;
            s->bmap[block_index] = cpu_to_le32(bmap_entry);
            s->header.blocks_allocated++;
            offset = s->header.offset_data / SECTOR_SIZE +
                     (uint64_t)bmap_entry * s->block_sectors;
            if (block == NULL) {
                block = g_malloc(s->block_size);
                bmap_first = block_index;
            }
            bmap_last = block_index;
            /* Copy data to be written to new block and zero unused parts. */
            memset(block, 0, sector_in_block * SECTOR_SIZE);
            memcpy(block + sector_in_block * SECTOR_SIZE,
                   buf, n_sectors * SECTOR_SIZE);
            memset(block + (sector_in_block + n_sectors) * SECTOR_SIZE, 0,
                   (s->block_sectors - n_sectors - sector_in_block) * SECTOR_SIZE);
            ret = bdrv_write(bs->file, offset, block, s->block_sectors);
        } else {
            uint64_t offset = s->header.offset_data / SECTOR_SIZE +
                              (uint64_t)bmap_entry * s->block_sectors +
                              sector_in_block;
            ret = bdrv_write(bs->file, offset, buf, n_sectors);
588
        }
589

P
Paolo Bonzini 已提交
590 591 592
        nb_sectors -= n_sectors;
        sector_num += n_sectors;
        buf += n_sectors * SECTOR_SIZE;
593

P
Paolo Bonzini 已提交
594
        logout("%u sectors written\n", n_sectors);
595 596
    }

597
    logout("finished data write\n");
598 599 600 601 602 603 604 605 606 607 608 609 610 611 612
    if (ret < 0) {
        return ret;
    }

    if (block) {
        /* One or more new blocks were allocated. */
        VdiHeader *header = (VdiHeader *) block;
        uint8_t *base;
        uint64_t offset;

        logout("now writing modified header\n");
        assert(VDI_IS_ALLOCATED(bmap_first));
        *header = s->header;
        vdi_header_to_le(header);
        ret = bdrv_write(bs->file, 0, block, 1);
P
Paolo Bonzini 已提交
613 614
        g_free(block);
        block = NULL;
615 616 617

        if (ret < 0) {
            return ret;
618
        }
619 620 621 622 623 624 625 626 627 628 629 630

        logout("now writing modified block map entry %u...%u\n",
               bmap_first, bmap_last);
        /* Write modified sectors from block map. */
        bmap_first /= (SECTOR_SIZE / sizeof(uint32_t));
        bmap_last /= (SECTOR_SIZE / sizeof(uint32_t));
        n_sectors = bmap_last - bmap_first + 1;
        offset = s->bmap_sector + bmap_first;
        base = ((uint8_t *)&s->bmap[0]) + bmap_first * SECTOR_SIZE;
        logout("will write %u block map sectors starting from entry %u\n",
               n_sectors, bmap_first);
        ret = bdrv_write(bs->file, offset, base, n_sectors);
631 632
    }

633
    return ret;
634 635 636 637 638 639 640 641
}

static int vdi_create(const char *filename, QEMUOptionParameter *options)
{
    int fd;
    int result = 0;
    uint64_t bytes = 0;
    uint32_t blocks;
642
    size_t block_size = DEFAULT_CLUSTER_SIZE;
643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662
    uint32_t image_type = VDI_TYPE_DYNAMIC;
    VdiHeader header;
    size_t i;
    size_t bmap_size;

    logout("\n");

    /* Read out options. */
    while (options && options->name) {
        if (!strcmp(options->name, BLOCK_OPT_SIZE)) {
            bytes = options->value.n;
#if defined(CONFIG_VDI_BLOCK_SIZE)
        } else if (!strcmp(options->name, BLOCK_OPT_CLUSTER_SIZE)) {
            if (options->value.n) {
                /* TODO: Additional checks (SECTOR_SIZE * 2^n, ...). */
                block_size = options->value.n;
            }
#endif
#if defined(CONFIG_VDI_STATIC_IMAGE)
        } else if (!strcmp(options->name, BLOCK_OPT_STATIC)) {
S
Stefan Weil 已提交
663 664 665
            if (options->value.n) {
                image_type = VDI_TYPE_STATIC;
            }
666 667 668 669 670
#endif
        }
        options++;
    }

671 672 673
    fd = qemu_open(filename,
                   O_WRONLY | O_CREAT | O_TRUNC | O_BINARY | O_LARGEFILE,
                   0644);
674 675 676 677
    if (fd < 0) {
        return -errno;
    }

678 679 680 681
    /* We need enough blocks to store the given disk size,
       so always round up. */
    blocks = (bytes + block_size - 1) / block_size;

682 683 684 685
    bmap_size = blocks * sizeof(uint32_t);
    bmap_size = ((bmap_size + SECTOR_SIZE - 1) & ~(SECTOR_SIZE -1));

    memset(&header, 0, sizeof(header));
686
    pstrcpy(header.text, sizeof(header.text), VDI_TEXT);
687 688 689 690 691 692 693 694 695 696
    header.signature = VDI_SIGNATURE;
    header.version = VDI_VERSION_1_1;
    header.header_size = 0x180;
    header.image_type = image_type;
    header.offset_bmap = 0x200;
    header.offset_data = 0x200 + bmap_size;
    header.sector_size = SECTOR_SIZE;
    header.disk_size = bytes;
    header.block_size = block_size;
    header.blocks_in_image = blocks;
S
Stefan Weil 已提交
697 698 699
    if (image_type == VDI_TYPE_STATIC) {
        header.blocks_allocated = blocks;
    }
700 701 702 703 704 705 706 707 708 709 710
    uuid_generate(header.uuid_image);
    uuid_generate(header.uuid_last_snap);
    /* There is no need to set header.uuid_link or header.uuid_parent here. */
#if defined(CONFIG_VDI_DEBUG)
    vdi_header_print(&header);
#endif
    vdi_header_to_le(&header);
    if (write(fd, &header, sizeof(header)) < 0) {
        result = -errno;
    }

711
    if (bmap_size > 0) {
S
Stefan Weil 已提交
712 713 714 715 716 717 718
        uint32_t *bmap = g_malloc0(bmap_size);
        for (i = 0; i < blocks; i++) {
            if (image_type == VDI_TYPE_STATIC) {
                bmap[i] = i;
            } else {
                bmap[i] = VDI_UNALLOCATED;
            }
719
        }
S
Stefan Weil 已提交
720 721 722 723
        if (write(fd, bmap, bmap_size) < 0) {
            result = -errno;
        }
        g_free(bmap);
724
    }
S
Stefan Weil 已提交
725

726 727 728 729 730 731 732 733 734 735 736 737 738 739 740
    if (image_type == VDI_TYPE_STATIC) {
        if (ftruncate(fd, sizeof(header) + bmap_size + blocks * block_size)) {
            result = -errno;
        }
    }

    if (close(fd) < 0) {
        result = -errno;
    }

    return result;
}

static void vdi_close(BlockDriverState *bs)
{
K
Kevin Wolf 已提交
741
    BDRVVdiState *s = bs->opaque;
K
Kevin Wolf 已提交
742 743 744

    g_free(s->bmap);

K
Kevin Wolf 已提交
745 746
    migrate_del_blocker(s->migration_blocker);
    error_free(s->migration_blocker);
747 748 749 750 751 752 753 754 755 756 757 758
}

static QEMUOptionParameter vdi_create_options[] = {
    {
        .name = BLOCK_OPT_SIZE,
        .type = OPT_SIZE,
        .help = "Virtual disk size"
    },
#if defined(CONFIG_VDI_BLOCK_SIZE)
    {
        .name = BLOCK_OPT_CLUSTER_SIZE,
        .type = OPT_SIZE,
759 760
        .help = "VDI cluster (block) size",
        .value = { .n = DEFAULT_CLUSTER_SIZE },
761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779
    },
#endif
#if defined(CONFIG_VDI_STATIC_IMAGE)
    {
        .name = BLOCK_OPT_STATIC,
        .type = OPT_FLAG,
        .help = "VDI static (pre-allocated) image"
    },
#endif
    /* TODO: An additional option to set UUID values might be useful. */
    { NULL }
};

static BlockDriver bdrv_vdi = {
    .format_name = "vdi",
    .instance_size = sizeof(BDRVVdiState),
    .bdrv_probe = vdi_probe,
    .bdrv_open = vdi_open,
    .bdrv_close = vdi_close,
J
Jeff Cody 已提交
780
    .bdrv_reopen_prepare = vdi_reopen_prepare,
781
    .bdrv_create = vdi_create,
782
    .bdrv_co_is_allocated = vdi_co_is_allocated,
783 784
    .bdrv_make_empty = vdi_make_empty,

785
    .bdrv_read = vdi_co_read,
786
#if defined(CONFIG_VDI_WRITE)
787
    .bdrv_write = vdi_co_write,
788 789 790 791 792 793 794 795 796 797 798 799 800 801 802
#endif

    .bdrv_get_info = vdi_get_info,

    .create_options = vdi_create_options,
    .bdrv_check = vdi_check,
};

static void bdrv_vdi_init(void)
{
    logout("\n");
    bdrv_register(&bdrv_vdi);
}

block_init(bdrv_vdi_init);