vpc.c 23.0 KB
Newer Older
1
/*
S
Stefan Weil 已提交
2
 * Block driver for Connectix / Microsoft Virtual PC images
3
 *
4
 * Copyright (c) 2005 Alex Beregszaszi
A
aliguori 已提交
5
 * Copyright (c) 2009 Kevin Wolf <kwolf@suse.de>
6
 *
7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
 * Permission is hereby granted, free of charge, to any person obtaining a copy
 * of this software and associated documentation files (the "Software"), to deal
 * in the Software without restriction, including without limitation the rights
 * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
 * copies of the Software, and to permit persons to whom the Software is
 * furnished to do so, subject to the following conditions:
 *
 * The above copyright notice and this permission notice shall be included in
 * all copies or substantial portions of the Software.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
 * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
 * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
 * THE SOFTWARE.
 */
P
pbrook 已提交
25
#include "qemu-common.h"
26
#include "block/block_int.h"
27
#include "qemu/module.h"
28
#include "migration/migration.h"
29 30 31
#if defined(CONFIG_UUID)
#include <uuid/uuid.h>
#endif
32 33 34 35 36 37 38

/**************************************************************/

#define HEADER_SIZE 512

//#define CACHE

39 40 41 42 43 44
enum vhd_type {
    VHD_FIXED           = 2,
    VHD_DYNAMIC         = 3,
    VHD_DIFFERENCING    = 4,
};

A
aliguori 已提交
45 46 47
// Seconds since Jan 1, 2000 0:00:00 (UTC)
#define VHD_TIMESTAMP_BASE 946684800

48
// always big-endian
49
struct vhd_footer {
50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81
    char        creator[8]; // "conectix"
    uint32_t    features;
    uint32_t    version;

    // Offset of next header structure, 0xFFFFFFFF if none
    uint64_t    data_offset;

    // Seconds since Jan 1, 2000 0:00:00 (UTC)
    uint32_t    timestamp;

    char        creator_app[4]; // "vpc "
    uint16_t    major;
    uint16_t    minor;
    char        creator_os[4]; // "Wi2k"

    uint64_t    orig_size;
    uint64_t    size;

    uint16_t    cyls;
    uint8_t     heads;
    uint8_t     secs_per_cyl;

    uint32_t    type;

    // Checksum of the Hard Disk Footer ("one's complement of the sum of all
    // the bytes in the footer without the checksum field")
    uint32_t    checksum;

    // UUID used to identify a parent hard disk (backing file)
    uint8_t     uuid[16];

    uint8_t     in_saved_state;
82 83 84
};

struct vhd_dyndisk_header {
85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113
    char        magic[8]; // "cxsparse"

    // Offset of next header structure, 0xFFFFFFFF if none
    uint64_t    data_offset;

    // Offset of the Block Allocation Table (BAT)
    uint64_t    table_offset;

    uint32_t    version;
    uint32_t    max_table_entries; // 32bit/entry

    // 2 MB by default, must be a power of two
    uint32_t    block_size;

    uint32_t    checksum;
    uint8_t     parent_uuid[16];
    uint32_t    parent_timestamp;
    uint32_t    reserved;

    // Backing file name (in UTF-16)
    uint8_t     parent_name[512];

    struct {
        uint32_t    platform;
        uint32_t    data_space;
        uint32_t    data_length;
        uint32_t    reserved;
        uint64_t    data_offset;
    } parent_locator[8];
114 115 116
};

typedef struct BDRVVPCState {
117
    CoMutex lock;
A
aliguori 已提交
118 119
    uint8_t footer_buf[HEADER_SIZE];
    uint64_t free_data_block_offset;
120
    int max_table_entries;
121
    uint32_t *pagetable;
A
aliguori 已提交
122 123
    uint64_t bat_offset;
    uint64_t last_bitmap_offset;
124

125
    uint32_t block_size;
A
aliguori 已提交
126 127
    uint32_t bitmap_size;

128 129 130 131
#ifdef CACHE
    uint8_t *pageentry_u8;
    uint32_t *pageentry_u32;
    uint16_t *pageentry_u16;
132

133 134
    uint64_t last_bitmap;
#endif
K
Kevin Wolf 已提交
135 136

    Error *migration_blocker;
137 138
} BDRVVPCState;

A
aliguori 已提交
139 140 141 142 143 144 145 146 147 148 149 150
static uint32_t vpc_checksum(uint8_t* buf, size_t size)
{
    uint32_t res = 0;
    int i;

    for (i = 0; i < size; i++)
        res += buf[i];

    return ~res;
}


151 152
static int vpc_probe(const uint8_t *buf, int buf_size, const char *filename)
{
T
ths 已提交
153
    if (buf_size >= 8 && !strncmp((char *)buf, "conectix", 8))
154 155 156 157
	return 100;
    return 0;
}

158
static int vpc_open(BlockDriverState *bs, int flags)
159 160
{
    BDRVVPCState *s = bs->opaque;
161
    int i;
162 163 164
    struct vhd_footer* footer;
    struct vhd_dyndisk_header* dyndisk_header;
    uint8_t buf[HEADER_SIZE];
A
aliguori 已提交
165
    uint32_t checksum;
166
    int err = -1;
167
    int disk_type = VHD_DYNAMIC;
168

169
    if (bdrv_pread(bs->file, 0, s->footer_buf, HEADER_SIZE) != HEADER_SIZE)
170 171
        goto fail;

A
aliguori 已提交
172
    footer = (struct vhd_footer*) s->footer_buf;
173 174 175 176 177 178 179 180 181 182 183 184 185 186 187
    if (strncmp(footer->creator, "conectix", 8)) {
        int64_t offset = bdrv_getlength(bs->file);
        if (offset < HEADER_SIZE) {
            goto fail;
        }
        /* If a fixed disk, the footer is found only at the end of the file */
        if (bdrv_pread(bs->file, offset-HEADER_SIZE, s->footer_buf, HEADER_SIZE)
                != HEADER_SIZE) {
            goto fail;
        }
        if (strncmp(footer->creator, "conectix", 8)) {
            goto fail;
        }
        disk_type = VHD_FIXED;
    }
188

A
aliguori 已提交
189 190 191 192
    checksum = be32_to_cpu(footer->checksum);
    footer->checksum = 0;
    if (vpc_checksum(s->footer_buf, HEADER_SIZE) != checksum)
        fprintf(stderr, "block-vpc: The header checksum of '%s' is "
193
            "incorrect.\n", bs->filename);
A
aliguori 已提交
194

195 196 197
    /* Write 'checksum' back to footer, or else will leave it with zero. */
    footer->checksum = be32_to_cpu(checksum);

A
aliguori 已提交
198 199 200 201 202 203
    // The visible size of a image in Virtual PC depends on the geometry
    // rather than on the size stored in the footer (the size in the footer
    // is too large usually)
    bs->total_sectors = (int64_t)
        be16_to_cpu(footer->cyls) * footer->heads * footer->secs_per_cyl;

204 205
    /* Allow a maximum disk size of approximately 2 TB */
    if (bs->total_sectors >= 65535LL * 255 * 255) {
206 207 208 209
        err = -EFBIG;
        goto fail;
    }

210 211 212 213 214
    if (disk_type == VHD_DYNAMIC) {
        if (bdrv_pread(bs->file, be64_to_cpu(footer->data_offset), buf,
                HEADER_SIZE) != HEADER_SIZE) {
            goto fail;
        }
215

216
        dyndisk_header = (struct vhd_dyndisk_header *) buf;
217

218 219 220
        if (strncmp(dyndisk_header->magic, "cxsparse", 8)) {
            goto fail;
        }
221

222 223
        s->block_size = be32_to_cpu(dyndisk_header->block_size);
        s->bitmap_size = ((s->block_size / (8 * 512)) + 511) & ~511;
A
aliguori 已提交
224

225 226
        s->max_table_entries = be32_to_cpu(dyndisk_header->max_table_entries);
        s->pagetable = g_malloc(s->max_table_entries * 4);
227

228 229 230 231 232
        s->bat_offset = be64_to_cpu(dyndisk_header->table_offset);
        if (bdrv_pread(bs->file, s->bat_offset, s->pagetable,
                s->max_table_entries * 4) != s->max_table_entries * 4) {
            goto fail;
        }
233

234 235
        s->free_data_block_offset =
            (s->bat_offset + (s->max_table_entries * 4) + 511) & ~511;
A
aliguori 已提交
236

237 238 239 240 241
        for (i = 0; i < s->max_table_entries; i++) {
            be32_to_cpus(&s->pagetable[i]);
            if (s->pagetable[i] != 0xFFFFFFFF) {
                int64_t next = (512 * (int64_t) s->pagetable[i]) +
                    s->bitmap_size + s->block_size;
A
aliguori 已提交
242

243 244 245 246
                if (next > s->free_data_block_offset) {
                    s->free_data_block_offset = next;
                }
            }
A
aliguori 已提交
247 248
        }

249
        s->last_bitmap_offset = (int64_t) -1;
250 251

#ifdef CACHE
252 253 254 255
        s->pageentry_u8 = g_malloc(512);
        s->pageentry_u32 = s->pageentry_u8;
        s->pageentry_u16 = s->pageentry_u8;
        s->last_pagetable = -1;
256
#endif
257
    }
258

259
    qemu_co_mutex_init(&s->lock);
K
Kevin Wolf 已提交
260 261 262 263 264 265 266

    /* Disable migration when VHD images are used */
    error_set(&s->migration_blocker,
              QERR_BLOCK_FORMAT_FEATURE_NOT_SUPPORTED,
              "vpc", bs->device_name, "live migration");
    migrate_add_blocker(s->migration_blocker);

267 268
    return 0;
 fail:
269
    return err;
270 271
}

J
Jeff Cody 已提交
272 273 274 275 276 277
static int vpc_reopen_prepare(BDRVReopenState *state,
                              BlockReopenQueue *queue, Error **errp)
{
    return 0;
}

278 279 280
/*
 * Returns the absolute byte offset of the given sector in the image file.
 * If the sector is not allocated, -1 is returned instead.
A
aliguori 已提交
281 282 283
 *
 * The parameter write must be 1 if the offset will be used for a write
 * operation (the block bitmaps is updated then), 0 otherwise.
284
 */
A
aliguori 已提交
285 286
static inline int64_t get_sector_offset(BlockDriverState *bs,
    int64_t sector_num, int write)
287 288 289 290 291 292
{
    BDRVVPCState *s = bs->opaque;
    uint64_t offset = sector_num * 512;
    uint64_t bitmap_offset, block_offset;
    uint32_t pagetable_index, pageentry_index;

293 294
    pagetable_index = offset / s->block_size;
    pageentry_index = (offset % s->block_size) / 512;
295

A
aliguori 已提交
296 297
    if (pagetable_index >= s->max_table_entries || s->pagetable[pagetable_index] == 0xffffffff)
        return -1; // not allocated
298

299
    bitmap_offset = 512 * (uint64_t) s->pagetable[pagetable_index];
A
aliguori 已提交
300 301 302 303 304 305 306 307 308 309 310 311
    block_offset = bitmap_offset + s->bitmap_size + (512 * pageentry_index);

    // We must ensure that we don't write to any sectors which are marked as
    // unused in the bitmap. We get away with setting all bits in the block
    // bitmap each time we write to a new block. This might cause Virtual PC to
    // miss sparse read optimization, but it's not a problem in terms of
    // correctness.
    if (write && (s->last_bitmap_offset != bitmap_offset)) {
        uint8_t bitmap[s->bitmap_size];

        s->last_bitmap_offset = bitmap_offset;
        memset(bitmap, 0xff, s->bitmap_size);
312
        bdrv_pwrite_sync(bs->file, bitmap_offset, bitmap, s->bitmap_size);
A
aliguori 已提交
313
    }
314

B
bellard 已提交
315
//    printf("sector: %" PRIx64 ", index: %x, offset: %x, bioff: %" PRIx64 ", bloff: %" PRIx64 "\n",
316 317 318 319 320 321 322 323 324 325 326
//	sector_num, pagetable_index, pageentry_index,
//	bitmap_offset, block_offset);

// disabled by reason
#if 0
#ifdef CACHE
    if (bitmap_offset != s->last_bitmap)
    {
	lseek(s->fd, bitmap_offset, SEEK_SET);

	s->last_bitmap = bitmap_offset;
327

328 329 330 331 332 333 334 335 336 337 338
	// Scary! Bitmap is stored as big endian 32bit entries,
	// while we used to look it up byte by byte
	read(s->fd, s->pageentry_u8, 512);
	for (i = 0; i < 128; i++)
	    be32_to_cpus(&s->pageentry_u32[i]);
    }

    if ((s->pageentry_u8[pageentry_index / 8] >> (pageentry_index % 8)) & 1)
	return -1;
#else
    lseek(s->fd, bitmap_offset + (pageentry_index / 8), SEEK_SET);
339

340 341 342 343 344 345 346
    read(s->fd, &bitmap_entry, 1);

    if ((bitmap_entry >> (pageentry_index % 8)) & 1)
	return -1; // not allocated
#endif
#endif

347
    return block_offset;
348 349
}

A
aliguori 已提交
350 351 352 353 354 355 356 357 358 359 360 361
/*
 * Writes the footer to the end of the image file. This is needed when the
 * file grows as it overwrites the old footer
 *
 * Returns 0 on success and < 0 on error
 */
static int rewrite_footer(BlockDriverState* bs)
{
    int ret;
    BDRVVPCState *s = bs->opaque;
    int64_t offset = s->free_data_block_offset;

362
    ret = bdrv_pwrite_sync(bs->file, offset, s->footer_buf, HEADER_SIZE);
A
aliguori 已提交
363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396
    if (ret < 0)
        return ret;

    return 0;
}

/*
 * Allocates a new block. This involves writing a new footer and updating
 * the Block Allocation Table to use the space at the old end of the image
 * file (overwriting the old footer)
 *
 * Returns the sectors' offset in the image file on success and < 0 on error
 */
static int64_t alloc_block(BlockDriverState* bs, int64_t sector_num)
{
    BDRVVPCState *s = bs->opaque;
    int64_t bat_offset;
    uint32_t index, bat_value;
    int ret;
    uint8_t bitmap[s->bitmap_size];

    // Check if sector_num is valid
    if ((sector_num < 0) || (sector_num > bs->total_sectors))
        return -1;

    // Write entry into in-memory BAT
    index = (sector_num * 512) / s->block_size;
    if (s->pagetable[index] != 0xFFFFFFFF)
        return -1;

    s->pagetable[index] = s->free_data_block_offset / 512;

    // Initialize the block's bitmap
    memset(bitmap, 0xff, s->bitmap_size);
397
    ret = bdrv_pwrite_sync(bs->file, s->free_data_block_offset, bitmap,
398
        s->bitmap_size);
399 400 401
    if (ret < 0) {
        return ret;
    }
A
aliguori 已提交
402 403 404 405 406 407 408 409 410 411

    // Write new footer (the old one will be overwritten)
    s->free_data_block_offset += s->block_size + s->bitmap_size;
    ret = rewrite_footer(bs);
    if (ret < 0)
        goto fail;

    // Write BAT entry to disk
    bat_offset = s->bat_offset + (4 * index);
    bat_value = be32_to_cpu(s->pagetable[index]);
412
    ret = bdrv_pwrite_sync(bs->file, bat_offset, &bat_value, 4);
A
aliguori 已提交
413 414 415 416 417 418 419 420 421 422
    if (ret < 0)
        goto fail;

    return get_sector_offset(bs, sector_num, 0);

fail:
    s->free_data_block_offset -= (s->block_size + s->bitmap_size);
    return -1;
}

423
static int vpc_read(BlockDriverState *bs, int64_t sector_num,
424 425
                    uint8_t *buf, int nb_sectors)
{
426
    BDRVVPCState *s = bs->opaque;
427
    int ret;
428
    int64_t offset;
429
    int64_t sectors, sectors_per_block;
430
    struct vhd_footer *footer = (struct vhd_footer *) s->footer_buf;
431

432 433 434
    if (cpu_to_be32(footer->type) == VHD_FIXED) {
        return bdrv_read(bs->file, sector_num, buf, nb_sectors);
    }
435
    while (nb_sectors > 0) {
A
aliguori 已提交
436
        offset = get_sector_offset(bs, sector_num, 0);
437

438 439 440 441 442 443
        sectors_per_block = s->block_size >> BDRV_SECTOR_BITS;
        sectors = sectors_per_block - (sector_num % sectors_per_block);
        if (sectors > nb_sectors) {
            sectors = nb_sectors;
        }

444
        if (offset == -1) {
445
            memset(buf, 0, sectors * BDRV_SECTOR_SIZE);
446
        } else {
447 448 449
            ret = bdrv_pread(bs->file, offset, buf,
                sectors * BDRV_SECTOR_SIZE);
            if (ret != sectors * BDRV_SECTOR_SIZE) {
450
                return -1;
451
            }
452 453
        }

454 455 456
        nb_sectors -= sectors;
        sector_num += sectors;
        buf += sectors * BDRV_SECTOR_SIZE;
457 458 459 460
    }
    return 0;
}

461 462 463 464 465 466 467 468 469 470 471
static coroutine_fn int vpc_co_read(BlockDriverState *bs, int64_t sector_num,
                                    uint8_t *buf, int nb_sectors)
{
    int ret;
    BDRVVPCState *s = bs->opaque;
    qemu_co_mutex_lock(&s->lock);
    ret = vpc_read(bs, sector_num, buf, nb_sectors);
    qemu_co_mutex_unlock(&s->lock);
    return ret;
}

A
aliguori 已提交
472 473 474
static int vpc_write(BlockDriverState *bs, int64_t sector_num,
    const uint8_t *buf, int nb_sectors)
{
475
    BDRVVPCState *s = bs->opaque;
A
aliguori 已提交
476
    int64_t offset;
477
    int64_t sectors, sectors_per_block;
A
aliguori 已提交
478
    int ret;
479
    struct vhd_footer *footer =  (struct vhd_footer *) s->footer_buf;
A
aliguori 已提交
480

481 482 483
    if (cpu_to_be32(footer->type) == VHD_FIXED) {
        return bdrv_write(bs->file, sector_num, buf, nb_sectors);
    }
A
aliguori 已提交
484 485 486
    while (nb_sectors > 0) {
        offset = get_sector_offset(bs, sector_num, 1);

487 488 489 490 491 492
        sectors_per_block = s->block_size >> BDRV_SECTOR_BITS;
        sectors = sectors_per_block - (sector_num % sectors_per_block);
        if (sectors > nb_sectors) {
            sectors = nb_sectors;
        }

A
aliguori 已提交
493 494 495 496 497 498
        if (offset == -1) {
            offset = alloc_block(bs, sector_num);
            if (offset < 0)
                return -1;
        }

499 500
        ret = bdrv_pwrite(bs->file, offset, buf, sectors * BDRV_SECTOR_SIZE);
        if (ret != sectors * BDRV_SECTOR_SIZE) {
A
aliguori 已提交
501
            return -1;
502
        }
A
aliguori 已提交
503

504 505 506
        nb_sectors -= sectors;
        sector_num += sectors;
        buf += sectors * BDRV_SECTOR_SIZE;
A
aliguori 已提交
507 508 509 510 511
    }

    return 0;
}

512 513 514 515 516 517 518 519 520 521 522
static coroutine_fn int vpc_co_write(BlockDriverState *bs, int64_t sector_num,
                                     const uint8_t *buf, int nb_sectors)
{
    int ret;
    BDRVVPCState *s = bs->opaque;
    qemu_co_mutex_lock(&s->lock);
    ret = vpc_write(bs, sector_num, buf, nb_sectors);
    qemu_co_mutex_unlock(&s->lock);
    return ret;
}

A
aliguori 已提交
523 524 525 526 527 528 529
/*
 * Calculates the number of cylinders, heads and sectors per cylinder
 * based on a given number of sectors. This is the algorithm described
 * in the VHD specification.
 *
 * Note that the geometry doesn't always exactly match total_sectors but
 * may round it down.
530
 *
531 532 533
 * Returns 0 on success, -EFBIG if the size is larger than ~2 TB. Override
 * the hardware EIDE and ATA-2 limit of 16 heads (max disk size of 127 GB)
 * and instead allow up to 255 heads.
A
aliguori 已提交
534
 */
535
static int calculate_geometry(int64_t total_sectors, uint16_t* cyls,
A
aliguori 已提交
536 537 538 539
    uint8_t* heads, uint8_t* secs_per_cyl)
{
    uint32_t cyls_times_heads;

540 541
    /* Allow a maximum disk size of approximately 2 TB */
    if (total_sectors > 65535LL * 255 * 255) {
542
        return -EFBIG;
543
    }
A
aliguori 已提交
544 545 546

    if (total_sectors > 65535 * 16 * 63) {
        *secs_per_cyl = 255;
547 548 549 550 551
        if (total_sectors > 65535 * 16 * 255) {
            *heads = 255;
        } else {
            *heads = 16;
        }
A
aliguori 已提交
552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573
        cyls_times_heads = total_sectors / *secs_per_cyl;
    } else {
        *secs_per_cyl = 17;
        cyls_times_heads = total_sectors / *secs_per_cyl;
        *heads = (cyls_times_heads + 1023) / 1024;

        if (*heads < 4)
            *heads = 4;

        if (cyls_times_heads >= (*heads * 1024) || *heads > 16) {
            *secs_per_cyl = 31;
            *heads = 16;
            cyls_times_heads = total_sectors / *secs_per_cyl;
        }

        if (cyls_times_heads >= (*heads * 1024)) {
            *secs_per_cyl = 63;
            *heads = 16;
            cyls_times_heads = total_sectors / *secs_per_cyl;
        }
    }

574
    *cyls = cyls_times_heads / *heads;
575 576

    return 0;
A
aliguori 已提交
577 578
}

579
static int create_dynamic_disk(int fd, uint8_t *buf, int64_t total_sectors)
A
aliguori 已提交
580 581 582 583
{
    struct vhd_dyndisk_header* dyndisk_header =
        (struct vhd_dyndisk_header*) buf;
    size_t block_size, num_bat_entries;
584
    int i;
B
Blue Swirl 已提交
585
    int ret = -EIO;
A
aliguori 已提交
586 587 588 589 590

    // Write the footer (twice: at the beginning and at the end)
    block_size = 0x200000;
    num_bat_entries = (total_sectors + block_size / 512) / (block_size / 512);

B
Blue Swirl 已提交
591 592 593
    if (write(fd, buf, HEADER_SIZE) != HEADER_SIZE) {
        goto fail;
    }
A
aliguori 已提交
594

B
Blue Swirl 已提交
595 596 597 598 599 600
    if (lseek(fd, 1536 + ((num_bat_entries * 4 + 511) & ~511), SEEK_SET) < 0) {
        goto fail;
    }
    if (write(fd, buf, HEADER_SIZE) != HEADER_SIZE) {
        goto fail;
    }
A
aliguori 已提交
601 602

    // Write the initial BAT
B
Blue Swirl 已提交
603 604 605
    if (lseek(fd, 3 * 512, SEEK_SET) < 0) {
        goto fail;
    }
A
aliguori 已提交
606 607

    memset(buf, 0xFF, 512);
B
Blue Swirl 已提交
608 609 610 611 612
    for (i = 0; i < (num_bat_entries * 4 + 511) / 512; i++) {
        if (write(fd, buf, 512) != 512) {
            goto fail;
        }
    }
A
aliguori 已提交
613 614 615 616

    // Prepare the Dynamic Disk Header
    memset(buf, 0, 1024);

617
    memcpy(dyndisk_header->magic, "cxsparse", 8);
A
aliguori 已提交
618

619 620 621 622 623
    /*
     * Note: The spec is actually wrong here for data_offset, it says
     * 0xFFFFFFFF, but MS tools expect all 64 bits to be set.
     */
    dyndisk_header->data_offset = be64_to_cpu(0xFFFFFFFFFFFFFFFFULL);
A
aliguori 已提交
624 625 626 627 628 629 630 631
    dyndisk_header->table_offset = be64_to_cpu(3 * 512);
    dyndisk_header->version = be32_to_cpu(0x00010000);
    dyndisk_header->block_size = be32_to_cpu(block_size);
    dyndisk_header->max_table_entries = be32_to_cpu(num_bat_entries);

    dyndisk_header->checksum = be32_to_cpu(vpc_checksum(buf, 1024));

    // Write the header
B
Blue Swirl 已提交
632 633 634
    if (lseek(fd, 512, SEEK_SET) < 0) {
        goto fail;
    }
A
aliguori 已提交
635

B
Blue Swirl 已提交
636 637 638 639 640
    if (write(fd, buf, 1024) != 1024) {
        goto fail;
    }
    ret = 0;

641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698
 fail:
    return ret;
}

static int create_fixed_disk(int fd, uint8_t *buf, int64_t total_size)
{
    int ret = -EIO;

    /* Add footer to total size */
    total_size += 512;
    if (ftruncate(fd, total_size) != 0) {
        ret = -errno;
        goto fail;
    }
    if (lseek(fd, -512, SEEK_END) < 0) {
        goto fail;
    }
    if (write(fd, buf, HEADER_SIZE) != HEADER_SIZE) {
        goto fail;
    }

    ret = 0;

 fail:
    return ret;
}

static int vpc_create(const char *filename, QEMUOptionParameter *options)
{
    uint8_t buf[1024];
    struct vhd_footer *footer = (struct vhd_footer *) buf;
    QEMUOptionParameter *disk_type_param;
    int fd, i;
    uint16_t cyls = 0;
    uint8_t heads = 0;
    uint8_t secs_per_cyl = 0;
    int64_t total_sectors;
    int64_t total_size;
    int disk_type;
    int ret = -EIO;

    /* Read out options */
    total_size = get_option_parameter(options, BLOCK_OPT_SIZE)->value.n;

    disk_type_param = get_option_parameter(options, BLOCK_OPT_SUBFMT);
    if (disk_type_param && disk_type_param->value.s) {
        if (!strcmp(disk_type_param->value.s, "dynamic")) {
            disk_type = VHD_DYNAMIC;
        } else if (!strcmp(disk_type_param->value.s, "fixed")) {
            disk_type = VHD_FIXED;
        } else {
            return -EINVAL;
        }
    } else {
        disk_type = VHD_DYNAMIC;
    }

    /* Create the file */
699
    fd = qemu_open(filename, O_WRONLY | O_CREAT | O_TRUNC | O_BINARY, 0644);
700 701 702 703
    if (fd < 0) {
        return -EIO;
    }

704 705 706 707 708
    /*
     * Calculate matching total_size and geometry. Increase the number of
     * sectors requested until we get enough (or fail). This ensures that
     * qemu-img convert doesn't truncate images, but rather rounds up.
     */
709
    total_sectors = total_size / BDRV_SECTOR_SIZE;
710 711 712 713
    for (i = 0; total_sectors > (int64_t)cyls * heads * secs_per_cyl; i++) {
        if (calculate_geometry(total_sectors + i, &cyls, &heads,
                               &secs_per_cyl))
        {
714 715 716 717
            ret = -EFBIG;
            goto fail;
        }
    }
718

719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753
    total_sectors = (int64_t) cyls * heads * secs_per_cyl;

    /* Prepare the Hard Disk Footer */
    memset(buf, 0, 1024);

    memcpy(footer->creator, "conectix", 8);
    /* TODO Check if "qemu" creator_app is ok for VPC */
    memcpy(footer->creator_app, "qemu", 4);
    memcpy(footer->creator_os, "Wi2k", 4);

    footer->features = be32_to_cpu(0x02);
    footer->version = be32_to_cpu(0x00010000);
    if (disk_type == VHD_DYNAMIC) {
        footer->data_offset = be64_to_cpu(HEADER_SIZE);
    } else {
        footer->data_offset = be64_to_cpu(0xFFFFFFFFFFFFFFFFULL);
    }
    footer->timestamp = be32_to_cpu(time(NULL) - VHD_TIMESTAMP_BASE);

    /* Version of Virtual PC 2007 */
    footer->major = be16_to_cpu(0x0005);
    footer->minor = be16_to_cpu(0x0003);
    if (disk_type == VHD_DYNAMIC) {
        footer->orig_size = be64_to_cpu(total_sectors * 512);
        footer->size = be64_to_cpu(total_sectors * 512);
    } else {
        footer->orig_size = be64_to_cpu(total_size);
        footer->size = be64_to_cpu(total_size);
    }
    footer->cyls = be16_to_cpu(cyls);
    footer->heads = heads;
    footer->secs_per_cyl = secs_per_cyl;

    footer->type = be32_to_cpu(disk_type);

754 755 756
#if defined(CONFIG_UUID)
    uuid_generate(footer->uuid);
#endif
757 758 759 760 761 762 763 764 765

    footer->checksum = be32_to_cpu(vpc_checksum(buf, HEADER_SIZE));

    if (disk_type == VHD_DYNAMIC) {
        ret = create_dynamic_disk(fd, buf, total_sectors);
    } else {
        ret = create_fixed_disk(fd, buf, total_size);
    }

B
Blue Swirl 已提交
766
 fail:
767
    qemu_close(fd);
B
Blue Swirl 已提交
768
    return ret;
A
aliguori 已提交
769 770
}

771 772 773
static void vpc_close(BlockDriverState *bs)
{
    BDRVVPCState *s = bs->opaque;
774
    g_free(s->pagetable);
775
#ifdef CACHE
776
    g_free(s->pageentry_u8);
777
#endif
K
Kevin Wolf 已提交
778 779 780

    migrate_del_blocker(s->migration_blocker);
    error_free(s->migration_blocker);
781 782
}

783
static QEMUOptionParameter vpc_create_options[] = {
784 785 786 787 788
    {
        .name = BLOCK_OPT_SIZE,
        .type = OPT_SIZE,
        .help = "Virtual disk size"
    },
789 790 791 792 793 794 795
    {
        .name = BLOCK_OPT_SUBFMT,
        .type = OPT_STRING,
        .help =
            "Type of virtual hard disk format. Supported formats are "
            "{dynamic (default) | fixed} "
    },
796 797 798
    { NULL }
};

799
static BlockDriver bdrv_vpc = {
K
Kevin Wolf 已提交
800 801
    .format_name    = "vpc",
    .instance_size  = sizeof(BDRVVPCState),
802

K
Kevin Wolf 已提交
803 804 805
    .bdrv_probe     = vpc_probe,
    .bdrv_open      = vpc_open,
    .bdrv_close     = vpc_close,
J
Jeff Cody 已提交
806
    .bdrv_reopen_prepare = vpc_reopen_prepare,
K
Kevin Wolf 已提交
807
    .bdrv_create    = vpc_create,
808

809 810 811
    .bdrv_read              = vpc_co_read,
    .bdrv_write             = vpc_co_write,

812
    .create_options = vpc_create_options,
813
};
814 815 816 817 818 819 820

static void bdrv_vpc_init(void)
{
    bdrv_register(&bdrv_vpc);
}

block_init(bdrv_vpc_init);