block.c 78.0 KB
Newer Older
B
bellard 已提交
1 2
/*
 * QEMU System Emulator block driver
3
 *
B
bellard 已提交
4
 * Copyright (c) 2003 Fabrice Bellard
5
 *
B
bellard 已提交
6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
 * 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.
 */
24
#include "config-host.h"
P
pbrook 已提交
25
#include "qemu-common.h"
26
#include "trace.h"
A
aliguori 已提交
27
#include "monitor.h"
B
bellard 已提交
28
#include "block_int.h"
29
#include "module.h"
30
#include "qemu-objects.h"
B
bellard 已提交
31

J
Juan Quintela 已提交
32
#ifdef CONFIG_BSD
B
bellard 已提交
33 34 35
#include <sys/types.h>
#include <sys/stat.h>
#include <sys/ioctl.h>
B
Blue Swirl 已提交
36
#include <sys/queue.h>
37
#ifndef __DragonFly__
B
bellard 已提交
38 39
#include <sys/disk.h>
#endif
40
#endif
B
bellard 已提交
41

42 43 44 45
#ifdef _WIN32
#include <windows.h>
#endif

46 47
static BlockDriverAIOCB *bdrv_aio_readv_em(BlockDriverState *bs,
        int64_t sector_num, QEMUIOVector *qiov, int nb_sectors,
48
        BlockDriverCompletionFunc *cb, void *opaque);
49 50
static BlockDriverAIOCB *bdrv_aio_writev_em(BlockDriverState *bs,
        int64_t sector_num, QEMUIOVector *qiov, int nb_sectors,
51
        BlockDriverCompletionFunc *cb, void *opaque);
52 53
static BlockDriverAIOCB *bdrv_aio_flush_em(BlockDriverState *bs,
        BlockDriverCompletionFunc *cb, void *opaque);
54 55
static BlockDriverAIOCB *bdrv_aio_noop_em(BlockDriverState *bs,
        BlockDriverCompletionFunc *cb, void *opaque);
56
static int bdrv_read_em(BlockDriverState *bs, int64_t sector_num,
B
bellard 已提交
57 58 59
                        uint8_t *buf, int nb_sectors);
static int bdrv_write_em(BlockDriverState *bs, int64_t sector_num,
                         const uint8_t *buf, int nb_sectors);
B
bellard 已提交
60

61 62
static QTAILQ_HEAD(, BlockDriverState) bdrv_states =
    QTAILQ_HEAD_INITIALIZER(bdrv_states);
63

64 65
static QLIST_HEAD(, BlockDriver) bdrv_drivers =
    QLIST_HEAD_INITIALIZER(bdrv_drivers);
B
bellard 已提交
66

67 68 69
/* The device to use for VM snapshots */
static BlockDriverState *bs_snapshots;

70 71 72
/* If non-zero, use only whitelisted block drivers */
static int use_bdrv_whitelist;

73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105
#ifdef _WIN32
static int is_windows_drive_prefix(const char *filename)
{
    return (((filename[0] >= 'a' && filename[0] <= 'z') ||
             (filename[0] >= 'A' && filename[0] <= 'Z')) &&
            filename[1] == ':');
}

int is_windows_drive(const char *filename)
{
    if (is_windows_drive_prefix(filename) &&
        filename[2] == '\0')
        return 1;
    if (strstart(filename, "\\\\.\\", NULL) ||
        strstart(filename, "//./", NULL))
        return 1;
    return 0;
}
#endif

/* check if the path starts with "<protocol>:" */
static int path_has_protocol(const char *path)
{
#ifdef _WIN32
    if (is_windows_drive(path) ||
        is_windows_drive_prefix(path)) {
        return 0;
    }
#endif

    return strchr(path, ':') != NULL;
}

B
bellard 已提交
106
int path_is_absolute(const char *path)
107
{
B
bellard 已提交
108
    const char *p;
B
bellard 已提交
109 110 111 112 113
#ifdef _WIN32
    /* specific case for names like: "\\.\d:" */
    if (*path == '/' || *path == '\\')
        return 1;
#endif
B
bellard 已提交
114 115 116 117 118
    p = strchr(path, ':');
    if (p)
        p++;
    else
        p = path;
119 120 121 122 123
#ifdef _WIN32
    return (*p == '/' || *p == '\\');
#else
    return (*p == '/');
#endif
124 125
}

B
bellard 已提交
126 127 128 129 130 131
/* if filename is absolute, just copy it to dest. Otherwise, build a
   path to it by considering it is relative to base_path. URL are
   supported. */
void path_combine(char *dest, int dest_size,
                  const char *base_path,
                  const char *filename)
132
{
B
bellard 已提交
133 134 135 136 137 138 139 140 141 142 143 144 145
    const char *p, *p1;
    int len;

    if (dest_size <= 0)
        return;
    if (path_is_absolute(filename)) {
        pstrcpy(dest, dest_size, filename);
    } else {
        p = strchr(base_path, ':');
        if (p)
            p++;
        else
            p = base_path;
146 147 148 149 150 151 152 153 154
        p1 = strrchr(base_path, '/');
#ifdef _WIN32
        {
            const char *p2;
            p2 = strrchr(base_path, '\\');
            if (!p1 || p2 > p1)
                p1 = p2;
        }
#endif
B
bellard 已提交
155 156 157 158 159 160 161 162 163 164 165 166
        if (p1)
            p1++;
        else
            p1 = base_path;
        if (p1 > p)
            p = p1;
        len = p - base_path;
        if (len > dest_size - 1)
            len = dest_size - 1;
        memcpy(dest, base_path, len);
        dest[len] = '\0';
        pstrcat(dest, dest_size, filename);
167 168 169
    }
}

170
void bdrv_register(BlockDriver *bdrv)
B
bellard 已提交
171
{
172
    if (!bdrv->bdrv_aio_readv) {
B
bellard 已提交
173
        /* add AIO emulation layer */
174 175
        bdrv->bdrv_aio_readv = bdrv_aio_readv_em;
        bdrv->bdrv_aio_writev = bdrv_aio_writev_em;
176
    } else if (!bdrv->bdrv_read) {
B
bellard 已提交
177 178 179 180
        /* add synchronous IO emulation layer */
        bdrv->bdrv_read = bdrv_read_em;
        bdrv->bdrv_write = bdrv_write_em;
    }
181 182 183 184

    if (!bdrv->bdrv_aio_flush)
        bdrv->bdrv_aio_flush = bdrv_aio_flush_em;

185
    QLIST_INSERT_HEAD(&bdrv_drivers, bdrv, list);
B
bellard 已提交
186
}
B
bellard 已提交
187 188 189 190

/* create a new block device (by default it is empty) */
BlockDriverState *bdrv_new(const char *device_name)
{
191
    BlockDriverState *bs;
B
bellard 已提交
192 193 194

    bs = qemu_mallocz(sizeof(BlockDriverState));
    pstrcpy(bs->device_name, sizeof(bs->device_name), device_name);
B
bellard 已提交
195
    if (device_name[0] != '\0') {
196
        QTAILQ_INSERT_TAIL(&bdrv_states, bs, list);
B
bellard 已提交
197
    }
B
bellard 已提交
198 199 200
    return bs;
}

B
bellard 已提交
201 202 203
BlockDriver *bdrv_find_format(const char *format_name)
{
    BlockDriver *drv1;
204 205
    QLIST_FOREACH(drv1, &bdrv_drivers, list) {
        if (!strcmp(drv1->format_name, format_name)) {
B
bellard 已提交
206
            return drv1;
207
        }
B
bellard 已提交
208 209 210 211
    }
    return NULL;
}

212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235
static int bdrv_is_whitelisted(BlockDriver *drv)
{
    static const char *whitelist[] = {
        CONFIG_BDRV_WHITELIST
    };
    const char **p;

    if (!whitelist[0])
        return 1;               /* no whitelist, anything goes */

    for (p = whitelist; *p; p++) {
        if (!strcmp(drv->format_name, *p)) {
            return 1;
        }
    }
    return 0;
}

BlockDriver *bdrv_find_whitelisted_format(const char *format_name)
{
    BlockDriver *drv = bdrv_find_format(format_name);
    return drv && bdrv_is_whitelisted(drv) ? drv : NULL;
}

236 237
int bdrv_create(BlockDriver *drv, const char* filename,
    QEMUOptionParameter *options)
B
bellard 已提交
238 239 240
{
    if (!drv->bdrv_create)
        return -ENOTSUP;
241 242

    return drv->bdrv_create(filename, options);
B
bellard 已提交
243 244
}

245 246 247 248
int bdrv_create_file(const char* filename, QEMUOptionParameter *options)
{
    BlockDriver *drv;

249
    drv = bdrv_find_protocol(filename);
250
    if (drv == NULL) {
251
        return -ENOENT;
252 253 254 255 256
    }

    return bdrv_create(drv, filename, options);
}

B
bellard 已提交
257
#ifdef _WIN32
258
void get_tmp_filename(char *filename, int size)
B
bellard 已提交
259
{
260
    char temp_dir[MAX_PATH];
261

262 263
    GetTempPath(MAX_PATH, temp_dir);
    GetTempFileName(temp_dir, "qem", 0, filename);
B
bellard 已提交
264 265
}
#else
266
void get_tmp_filename(char *filename, int size)
B
bellard 已提交
267
{
B
bellard 已提交
268
    int fd;
269
    const char *tmpdir;
B
bellard 已提交
270
    /* XXX: race condition possible */
A
aurel32 已提交
271 272 273 274
    tmpdir = getenv("TMPDIR");
    if (!tmpdir)
        tmpdir = "/tmp";
    snprintf(filename, size, "%s/vl.XXXXXX", tmpdir);
B
bellard 已提交
275 276 277
    fd = mkstemp(filename);
    close(fd);
}
B
bellard 已提交
278
#endif
B
bellard 已提交
279

280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301
/*
 * Detect host devices. By convention, /dev/cdrom[N] is always
 * recognized as a host CDROM.
 */
static BlockDriver *find_hdev_driver(const char *filename)
{
    int score_max = 0, score;
    BlockDriver *drv = NULL, *d;

    QLIST_FOREACH(d, &bdrv_drivers, list) {
        if (d->bdrv_probe_device) {
            score = d->bdrv_probe_device(filename);
            if (score > score_max) {
                score_max = score;
                drv = d;
            }
        }
    }

    return drv;
}

302
BlockDriver *bdrv_find_protocol(const char *filename)
B
bellard 已提交
303 304 305
{
    BlockDriver *drv1;
    char protocol[128];
306
    int len;
B
bellard 已提交
307
    const char *p;
B
bellard 已提交
308

309 310
    /* TODO Drivers without bdrv_file_open must be specified explicitly */

311 312 313 314 315 316 317 318 319 320 321 322
    /*
     * XXX(hch): we really should not let host device detection
     * override an explicit protocol specification, but moving this
     * later breaks access to device names with colons in them.
     * Thanks to the brain-dead persistent naming schemes on udev-
     * based Linux systems those actually are quite common.
     */
    drv1 = find_hdev_driver(filename);
    if (drv1) {
        return drv1;
    }

323
    if (!path_has_protocol(filename)) {
324
        return bdrv_find_format("file");
325
    }
326 327
    p = strchr(filename, ':');
    assert(p != NULL);
328 329 330 331 332
    len = p - filename;
    if (len > sizeof(protocol) - 1)
        len = sizeof(protocol) - 1;
    memcpy(protocol, filename, len);
    protocol[len] = '\0';
333
    QLIST_FOREACH(drv1, &bdrv_drivers, list) {
334
        if (drv1->protocol_name &&
335
            !strcmp(drv1->protocol_name, protocol)) {
B
bellard 已提交
336
            return drv1;
337
        }
B
bellard 已提交
338 339 340 341
    }
    return NULL;
}

342
static int find_image_format(const char *filename, BlockDriver **pdrv)
343 344 345 346 347 348
{
    int ret, score, score_max;
    BlockDriver *drv1, *drv;
    uint8_t buf[2048];
    BlockDriverState *bs;

349
    ret = bdrv_file_open(&bs, filename, 0);
350 351 352 353
    if (ret < 0) {
        *pdrv = NULL;
        return ret;
    }
354

355 356
    /* Return the raw BlockDriver * to scsi-generic devices or empty drives */
    if (bs->sg || !bdrv_is_inserted(bs)) {
357
        bdrv_delete(bs);
358 359 360 361 362 363
        drv = bdrv_find_format("raw");
        if (!drv) {
            ret = -ENOENT;
        }
        *pdrv = drv;
        return ret;
364
    }
365

B
bellard 已提交
366 367 368
    ret = bdrv_pread(bs, 0, buf, sizeof(buf));
    bdrv_delete(bs);
    if (ret < 0) {
369 370
        *pdrv = NULL;
        return ret;
B
bellard 已提交
371 372
    }

B
bellard 已提交
373
    score_max = 0;
374
    drv = NULL;
375
    QLIST_FOREACH(drv1, &bdrv_drivers, list) {
B
bellard 已提交
376 377 378 379 380 381
        if (drv1->bdrv_probe) {
            score = drv1->bdrv_probe(buf, ret, filename);
            if (score > score_max) {
                score_max = score;
                drv = drv1;
            }
B
bellard 已提交
382
        }
B
bellard 已提交
383
    }
384 385 386 387 388
    if (!drv) {
        ret = -ENOENT;
    }
    *pdrv = drv;
    return ret;
B
bellard 已提交
389 390
}

391 392 393 394 395 396 397
/**
 * Set the current 'total_sectors' value
 */
static int refresh_total_sectors(BlockDriverState *bs, int64_t hint)
{
    BlockDriver *drv = bs->drv;

398 399 400 401
    /* Do not attempt drv->bdrv_getlength() on scsi-generic devices */
    if (bs->sg)
        return 0;

402 403 404 405 406 407 408 409 410 411 412 413 414
    /* query actual device if possible, otherwise just trust the hint */
    if (drv->bdrv_getlength) {
        int64_t length = drv->bdrv_getlength(bs);
        if (length < 0) {
            return length;
        }
        hint = length >> BDRV_SECTOR_BITS;
    }

    bs->total_sectors = hint;
    return 0;
}

415 416 417 418 419 420 421 422 423 424
/*
 * Common part for opening disk images and files
 */
static int bdrv_open_common(BlockDriverState *bs, const char *filename,
    int flags, BlockDriver *drv)
{
    int ret, open_flags;

    assert(drv != NULL);

425
    bs->file = NULL;
426
    bs->total_sectors = 0;
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
    bs->encrypted = 0;
    bs->valid_key = 0;
    bs->open_flags = flags;
    /* buffer_alignment defaulted to 512, drivers can change this value */
    bs->buffer_alignment = 512;

    pstrcpy(bs->filename, sizeof(bs->filename), filename);

    if (use_bdrv_whitelist && !bdrv_is_whitelisted(drv)) {
        return -ENOTSUP;
    }

    bs->drv = drv;
    bs->opaque = qemu_mallocz(drv->instance_size);

    /*
     * Yes, BDRV_O_NOCACHE aka O_DIRECT means we have to present a
     * write cache to the guest.  We do need the fdatasync to flush
     * out transactions for block allocations, and we maybe have a
     * volatile write cache in our backing device to deal with.
     */
    if (flags & (BDRV_O_CACHE_WB|BDRV_O_NOCACHE))
        bs->enable_write_cache = 1;

    /*
     * Clear flags that are internal to the block layer before opening the
     * image.
     */
    open_flags = flags & ~(BDRV_O_SNAPSHOT | BDRV_O_NO_BACKING);

    /*
     * Snapshots should be writeable.
     */
    if (bs->is_temporary) {
        open_flags |= BDRV_O_RDWR;
    }

464 465 466 467 468 469 470 471 472 473
    /* Open the image, either directly or using a protocol */
    if (drv->bdrv_file_open) {
        ret = drv->bdrv_file_open(bs, filename, open_flags);
    } else {
        ret = bdrv_file_open(&bs->file, filename, open_flags);
        if (ret >= 0) {
            ret = drv->bdrv_open(bs, open_flags);
        }
    }

474 475 476 477 478
    if (ret < 0) {
        goto free_and_fail;
    }

    bs->keep_read_only = bs->read_only = !(open_flags & BDRV_O_RDWR);
479 480 481 482

    ret = refresh_total_sectors(bs, bs->total_sectors);
    if (ret < 0) {
        goto free_and_fail;
483
    }
484

485 486 487 488 489 490 491 492
#ifndef _WIN32
    if (bs->is_temporary) {
        unlink(filename);
    }
#endif
    return 0;

free_and_fail:
493 494 495 496
    if (bs->file) {
        bdrv_delete(bs->file);
        bs->file = NULL;
    }
497 498 499 500 501 502
    qemu_free(bs->opaque);
    bs->opaque = NULL;
    bs->drv = NULL;
    return ret;
}

K
Kevin Wolf 已提交
503 504 505
/*
 * Opens a file using a protocol (file, host_device, nbd, ...)
 */
B
bellard 已提交
506
int bdrv_file_open(BlockDriverState **pbs, const char *filename, int flags)
B
bellard 已提交
507
{
B
bellard 已提交
508
    BlockDriverState *bs;
509
    BlockDriver *drv;
B
bellard 已提交
510 511
    int ret;

512
    drv = bdrv_find_protocol(filename);
513 514 515 516
    if (!drv) {
        return -ENOENT;
    }

B
bellard 已提交
517
    bs = bdrv_new("");
K
Kevin Wolf 已提交
518
    ret = bdrv_open_common(bs, filename, flags, drv);
B
bellard 已提交
519 520 521
    if (ret < 0) {
        bdrv_delete(bs);
        return ret;
522
    }
523
    bs->growable = 1;
B
bellard 已提交
524 525 526 527
    *pbs = bs;
    return 0;
}

K
Kevin Wolf 已提交
528 529 530
/*
 * Opens a disk image (raw, qcow2, vmdk, ...)
 */
K
Kevin Wolf 已提交
531 532
int bdrv_open(BlockDriverState *bs, const char *filename, int flags,
              BlockDriver *drv)
B
bellard 已提交
533
{
K
Kevin Wolf 已提交
534
    int ret;
B
bellard 已提交
535

B
bellard 已提交
536
    if (flags & BDRV_O_SNAPSHOT) {
B
bellard 已提交
537 538
        BlockDriverState *bs1;
        int64_t total_size;
A
aliguori 已提交
539
        int is_protocol = 0;
K
Kevin Wolf 已提交
540 541
        BlockDriver *bdrv_qcow2;
        QEMUOptionParameter *options;
K
Kevin Wolf 已提交
542 543
        char tmp_filename[PATH_MAX];
        char backing_filename[PATH_MAX];
544

B
bellard 已提交
545 546
        /* if snapshot, we create a temporary backing file and open it
           instead of opening 'filename' directly */
547

B
bellard 已提交
548 549
        /* if there is a backing file, use it */
        bs1 = bdrv_new("");
K
Kevin Wolf 已提交
550
        ret = bdrv_open(bs1, filename, 0, drv);
551
        if (ret < 0) {
B
bellard 已提交
552
            bdrv_delete(bs1);
553
            return ret;
B
bellard 已提交
554
        }
555
        total_size = bdrv_getlength(bs1) & BDRV_SECTOR_MASK;
A
aliguori 已提交
556 557 558 559

        if (bs1->drv && bs1->drv->protocol_name)
            is_protocol = 1;

B
bellard 已提交
560
        bdrv_delete(bs1);
561

B
bellard 已提交
562
        get_tmp_filename(tmp_filename, sizeof(tmp_filename));
A
aliguori 已提交
563 564 565 566 567

        /* Real path is meaningless for protocols */
        if (is_protocol)
            snprintf(backing_filename, sizeof(backing_filename),
                     "%s", filename);
568 569
        else if (!realpath(filename, backing_filename))
            return -errno;
A
aliguori 已提交
570

K
Kevin Wolf 已提交
571 572 573
        bdrv_qcow2 = bdrv_find_format("qcow2");
        options = parse_option_parameters("", bdrv_qcow2->create_options, NULL);

574
        set_option_parameter_int(options, BLOCK_OPT_SIZE, total_size);
K
Kevin Wolf 已提交
575 576 577 578 579 580 581
        set_option_parameter(options, BLOCK_OPT_BACKING_FILE, backing_filename);
        if (drv) {
            set_option_parameter(options, BLOCK_OPT_BACKING_FMT,
                drv->format_name);
        }

        ret = bdrv_create(bdrv_qcow2, tmp_filename, options);
582
        free_option_parameters(options);
583 584
        if (ret < 0) {
            return ret;
B
bellard 已提交
585
        }
K
Kevin Wolf 已提交
586

B
bellard 已提交
587
        filename = tmp_filename;
K
Kevin Wolf 已提交
588
        drv = bdrv_qcow2;
B
bellard 已提交
589 590
        bs->is_temporary = 1;
    }
B
bellard 已提交
591

K
Kevin Wolf 已提交
592
    /* Find the right image format driver */
593
    if (!drv) {
594
        ret = find_image_format(filename, &drv);
595
    }
596

597 598
    if (!drv) {
        goto unlink_and_fail;
B
bellard 已提交
599
    }
K
Kevin Wolf 已提交
600 601 602 603

    /* Open the image */
    ret = bdrv_open_common(bs, filename, flags, drv);
    if (ret < 0) {
604 605 606
        goto unlink_and_fail;
    }

K
Kevin Wolf 已提交
607 608 609 610 611 612 613
    /* If there is a backing file, use it */
    if ((flags & BDRV_O_NO_BACKING) == 0 && bs->backing_file[0] != '\0') {
        char backing_filename[PATH_MAX];
        int back_flags;
        BlockDriver *back_drv = NULL;

        bs->backing_hd = bdrv_new("");
614 615 616 617 618 619 620 621 622 623

        if (path_has_protocol(bs->backing_file)) {
            pstrcpy(backing_filename, sizeof(backing_filename),
                    bs->backing_file);
        } else {
            path_combine(backing_filename, sizeof(backing_filename),
                         filename, bs->backing_file);
        }

        if (bs->backing_format[0] != '\0') {
K
Kevin Wolf 已提交
624
            back_drv = bdrv_find_format(bs->backing_format);
625
        }
K
Kevin Wolf 已提交
626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647

        /* backing files always opened read-only */
        back_flags =
            flags & ~(BDRV_O_RDWR | BDRV_O_SNAPSHOT | BDRV_O_NO_BACKING);

        ret = bdrv_open(bs->backing_hd, backing_filename, back_flags, back_drv);
        if (ret < 0) {
            bdrv_close(bs);
            return ret;
        }
        if (bs->is_temporary) {
            bs->backing_hd->keep_read_only = !(flags & BDRV_O_RDWR);
        } else {
            /* base image inherits from "parent" */
            bs->backing_hd->keep_read_only = bs->keep_read_only;
        }
    }

    if (!bdrv_key_required(bs)) {
        /* call the change callback */
        bs->media_changed = 1;
        if (bs->change_cb)
648
            bs->change_cb(bs->change_opaque, CHANGE_MEDIA);
K
Kevin Wolf 已提交
649 650 651 652 653 654 655 656 657 658 659
    }

    return 0;

unlink_and_fail:
    if (bs->is_temporary) {
        unlink(filename);
    }
    return ret;
}

B
bellard 已提交
660 661
void bdrv_close(BlockDriverState *bs)
{
B
bellard 已提交
662
    if (bs->drv) {
663 664 665
        if (bs == bs_snapshots) {
            bs_snapshots = NULL;
        }
666
        if (bs->backing_hd) {
B
bellard 已提交
667
            bdrv_delete(bs->backing_hd);
668 669
            bs->backing_hd = NULL;
        }
B
bellard 已提交
670 671 672 673 674 675
        bs->drv->bdrv_close(bs);
        qemu_free(bs->opaque);
#ifdef _WIN32
        if (bs->is_temporary) {
            unlink(bs->filename);
        }
B
bellard 已提交
676
#endif
B
bellard 已提交
677 678
        bs->opaque = NULL;
        bs->drv = NULL;
B
bellard 已提交
679

680 681 682 683
        if (bs->file != NULL) {
            bdrv_close(bs->file);
        }

B
bellard 已提交
684
        /* call the change callback */
B
bellard 已提交
685
        bs->media_changed = 1;
B
bellard 已提交
686
        if (bs->change_cb)
687
            bs->change_cb(bs->change_opaque, CHANGE_MEDIA);
B
bellard 已提交
688 689 690
    }
}

691 692 693 694 695 696 697 698 699
void bdrv_close_all(void)
{
    BlockDriverState *bs;

    QTAILQ_FOREACH(bs, &bdrv_states, list) {
        bdrv_close(bs);
    }
}

B
bellard 已提交
700 701
void bdrv_delete(BlockDriverState *bs)
{
702 703
    assert(!bs->peer);

704 705 706 707
    /* remove from list, if necessary */
    if (bs->device_name[0] != '\0') {
        QTAILQ_REMOVE(&bdrv_states, bs, list);
    }
708

B
bellard 已提交
709
    bdrv_close(bs);
710 711 712 713
    if (bs->file != NULL) {
        bdrv_delete(bs->file);
    }

714
    assert(bs != bs_snapshots);
B
bellard 已提交
715
    qemu_free(bs);
B
bellard 已提交
716 717
}

718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737
int bdrv_attach(BlockDriverState *bs, DeviceState *qdev)
{
    if (bs->peer) {
        return -EBUSY;
    }
    bs->peer = qdev;
    return 0;
}

void bdrv_detach(BlockDriverState *bs, DeviceState *qdev)
{
    assert(bs->peer == qdev);
    bs->peer = NULL;
}

DeviceState *bdrv_get_attached(BlockDriverState *bs)
{
    return bs->peer;
}

A
aliguori 已提交
738 739 740
/*
 * Run consistency checks on an image
 *
741 742 743
 * Returns 0 if the check could be completed (it doesn't mean that the image is
 * free of errors) or -errno when an internal error occured. The results of the
 * check are stored in res.
A
aliguori 已提交
744
 */
745
int bdrv_check(BlockDriverState *bs, BdrvCheckResult *res)
A
aliguori 已提交
746 747 748 749 750
{
    if (bs->drv->bdrv_check == NULL) {
        return -ENOTSUP;
    }

751
    memset(res, 0, sizeof(*res));
752
    return bs->drv->bdrv_check(bs, res);
A
aliguori 已提交
753 754
}

755 756
#define COMMIT_BUF_SECTORS 2048

757 758 759
/* commit COW file into the raw image */
int bdrv_commit(BlockDriverState *bs)
{
B
bellard 已提交
760
    BlockDriver *drv = bs->drv;
761
    BlockDriver *backing_drv;
762 763
    int64_t sector, total_sectors;
    int n, ro, open_flags;
764
    int ret = 0, rw_ret = 0;
765
    uint8_t *buf;
766 767
    char filename[1024];
    BlockDriverState *bs_rw, *bs_ro;
768

B
bellard 已提交
769 770
    if (!drv)
        return -ENOMEDIUM;
771 772 773
    
    if (!bs->backing_hd) {
        return -ENOTSUP;
774 775
    }

776 777 778
    if (bs->backing_hd->keep_read_only) {
        return -EACCES;
    }
779 780

    backing_drv = bs->backing_hd->drv;
781 782 783 784 785 786 787 788 789
    ro = bs->backing_hd->read_only;
    strncpy(filename, bs->backing_hd->filename, sizeof(filename));
    open_flags =  bs->backing_hd->open_flags;

    if (ro) {
        /* re-open as RW */
        bdrv_delete(bs->backing_hd);
        bs->backing_hd = NULL;
        bs_rw = bdrv_new("");
790 791
        rw_ret = bdrv_open(bs_rw, filename, open_flags | BDRV_O_RDWR,
            backing_drv);
792 793 794 795
        if (rw_ret < 0) {
            bdrv_delete(bs_rw);
            /* try to re-open read-only */
            bs_ro = bdrv_new("");
796 797
            ret = bdrv_open(bs_ro, filename, open_flags & ~BDRV_O_RDWR,
                backing_drv);
798 799 800 801 802 803 804 805 806 807
            if (ret < 0) {
                bdrv_delete(bs_ro);
                /* drive not functional anymore */
                bs->drv = NULL;
                return ret;
            }
            bs->backing_hd = bs_ro;
            return rw_ret;
        }
        bs->backing_hd = bs_rw;
B
bellard 已提交
808
    }
809

J
Jan Kiszka 已提交
810
    total_sectors = bdrv_getlength(bs) >> BDRV_SECTOR_BITS;
811 812 813 814 815 816 817 818 819 820 821 822 823 824
    buf = qemu_malloc(COMMIT_BUF_SECTORS * BDRV_SECTOR_SIZE);

    for (sector = 0; sector < total_sectors; sector += n) {
        if (drv->bdrv_is_allocated(bs, sector, COMMIT_BUF_SECTORS, &n)) {

            if (bdrv_read(bs, sector, buf, n) != 0) {
                ret = -EIO;
                goto ro_cleanup;
            }

            if (bdrv_write(bs->backing_hd, sector, buf, n) != 0) {
                ret = -EIO;
                goto ro_cleanup;
            }
B
bellard 已提交
825
        }
826
    }
827

828 829 830 831
    if (drv->bdrv_make_empty) {
        ret = drv->bdrv_make_empty(bs);
        bdrv_flush(bs);
    }
832

833 834 835 836 837 838
    /*
     * Make sure all data we wrote to the backing device is actually
     * stable on disk.
     */
    if (bs->backing_hd)
        bdrv_flush(bs->backing_hd);
839 840

ro_cleanup:
841
    qemu_free(buf);
842 843 844 845 846 847

    if (ro) {
        /* re-open as RO */
        bdrv_delete(bs->backing_hd);
        bs->backing_hd = NULL;
        bs_ro = bdrv_new("");
848 849
        ret = bdrv_open(bs_ro, filename, open_flags & ~BDRV_O_RDWR,
            backing_drv);
850 851 852 853 854 855 856 857 858 859
        if (ret < 0) {
            bdrv_delete(bs_ro);
            /* drive not functional anymore */
            bs->drv = NULL;
            return ret;
        }
        bs->backing_hd = bs_ro;
        bs->backing_hd->keep_read_only = 0;
    }

860
    return ret;
861 862
}

863 864 865 866 867 868 869 870 871
void bdrv_commit_all(void)
{
    BlockDriverState *bs;

    QTAILQ_FOREACH(bs, &bdrv_states, list) {
        bdrv_commit(bs);
    }
}

K
Kevin Wolf 已提交
872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891
/*
 * Return values:
 * 0        - success
 * -EINVAL  - backing format specified, but no file
 * -ENOSPC  - can't update the backing file because no space is left in the
 *            image file header
 * -ENOTSUP - format driver doesn't support changing the backing file
 */
int bdrv_change_backing_file(BlockDriverState *bs,
    const char *backing_file, const char *backing_fmt)
{
    BlockDriver *drv = bs->drv;

    if (drv->bdrv_change_backing_file != NULL) {
        return drv->bdrv_change_backing_file(bs, backing_file, backing_fmt);
    } else {
        return -ENOTSUP;
    }
}

892 893 894 895 896 897 898 899 900 901 902 903 904
static int bdrv_check_byte_request(BlockDriverState *bs, int64_t offset,
                                   size_t size)
{
    int64_t len;

    if (!bdrv_is_inserted(bs))
        return -ENOMEDIUM;

    if (bs->growable)
        return 0;

    len = bdrv_getlength(bs);

K
Kevin Wolf 已提交
905 906 907 908
    if (offset < 0)
        return -EIO;

    if ((offset > len) || (len - offset < size))
909 910 911 912 913 914 915 916
        return -EIO;

    return 0;
}

static int bdrv_check_request(BlockDriverState *bs, int64_t sector_num,
                              int nb_sectors)
{
917 918
    return bdrv_check_byte_request(bs, sector_num * BDRV_SECTOR_SIZE,
                                   nb_sectors * BDRV_SECTOR_SIZE);
919 920
}

B
bellard 已提交
921
/* return < 0 if error. See bdrv_write() for the return codes */
922
int bdrv_read(BlockDriverState *bs, int64_t sector_num,
B
bellard 已提交
923 924
              uint8_t *buf, int nb_sectors)
{
B
bellard 已提交
925 926
    BlockDriver *drv = bs->drv;

B
bellard 已提交
927 928
    if (!drv)
        return -ENOMEDIUM;
929 930
    if (bdrv_check_request(bs, sector_num, nb_sectors))
        return -EIO;
B
bellard 已提交
931

932
    return drv->bdrv_read(bs, sector_num, buf, nb_sectors);
B
bellard 已提交
933 934
}

935
static void set_dirty_bitmap(BlockDriverState *bs, int64_t sector_num,
936
                             int nb_sectors, int dirty)
937 938
{
    int64_t start, end;
939
    unsigned long val, idx, bit;
940

J
Jan Kiszka 已提交
941
    start = sector_num / BDRV_SECTORS_PER_DIRTY_CHUNK;
942
    end = (sector_num + nb_sectors - 1) / BDRV_SECTORS_PER_DIRTY_CHUNK;
943 944

    for (; start <= end; start++) {
945 946 947 948
        idx = start / (sizeof(unsigned long) * 8);
        bit = start % (sizeof(unsigned long) * 8);
        val = bs->dirty_bitmap[idx];
        if (dirty) {
949
            if (!(val & (1UL << bit))) {
950
                bs->dirty_count++;
951
                val |= 1UL << bit;
952
            }
953
        } else {
954
            if (val & (1UL << bit)) {
955
                bs->dirty_count--;
956
                val &= ~(1UL << bit);
957
            }
958 959
        }
        bs->dirty_bitmap[idx] = val;
960 961 962
    }
}

963
/* Return < 0 if error. Important errors are:
B
bellard 已提交
964 965 966 967 968
  -EIO         generic I/O error (may happen for all errors)
  -ENOMEDIUM   No media inserted.
  -EINVAL      Invalid sector number or nb_sectors
  -EACCES      Trying to write a read-only device
*/
969
int bdrv_write(BlockDriverState *bs, int64_t sector_num,
B
bellard 已提交
970 971
               const uint8_t *buf, int nb_sectors)
{
B
bellard 已提交
972
    BlockDriver *drv = bs->drv;
B
bellard 已提交
973 974
    if (!bs->drv)
        return -ENOMEDIUM;
B
bellard 已提交
975
    if (bs->read_only)
B
bellard 已提交
976
        return -EACCES;
977 978
    if (bdrv_check_request(bs, sector_num, nb_sectors))
        return -EIO;
979

980
    if (bs->dirty_bitmap) {
981 982
        set_dirty_bitmap(bs, sector_num, nb_sectors, 1);
    }
983

984 985 986 987
    if (bs->wr_highest_sector < sector_num + nb_sectors - 1) {
        bs->wr_highest_sector = sector_num + nb_sectors - 1;
    }

988
    return drv->bdrv_write(bs, sector_num, buf, nb_sectors);
B
bellard 已提交
989 990
}

991 992
int bdrv_pread(BlockDriverState *bs, int64_t offset,
               void *buf, int count1)
B
bellard 已提交
993
{
J
Jan Kiszka 已提交
994
    uint8_t tmp_buf[BDRV_SECTOR_SIZE];
B
bellard 已提交
995 996
    int len, nb_sectors, count;
    int64_t sector_num;
997
    int ret;
B
bellard 已提交
998 999 1000

    count = count1;
    /* first read to align to sector start */
J
Jan Kiszka 已提交
1001
    len = (BDRV_SECTOR_SIZE - offset) & (BDRV_SECTOR_SIZE - 1);
B
bellard 已提交
1002 1003
    if (len > count)
        len = count;
J
Jan Kiszka 已提交
1004
    sector_num = offset >> BDRV_SECTOR_BITS;
B
bellard 已提交
1005
    if (len > 0) {
1006 1007
        if ((ret = bdrv_read(bs, sector_num, tmp_buf, 1)) < 0)
            return ret;
J
Jan Kiszka 已提交
1008
        memcpy(buf, tmp_buf + (offset & (BDRV_SECTOR_SIZE - 1)), len);
B
bellard 已提交
1009 1010 1011 1012 1013 1014 1015 1016
        count -= len;
        if (count == 0)
            return count1;
        sector_num++;
        buf += len;
    }

    /* read the sectors "in place" */
J
Jan Kiszka 已提交
1017
    nb_sectors = count >> BDRV_SECTOR_BITS;
B
bellard 已提交
1018
    if (nb_sectors > 0) {
1019 1020
        if ((ret = bdrv_read(bs, sector_num, buf, nb_sectors)) < 0)
            return ret;
B
bellard 已提交
1021
        sector_num += nb_sectors;
J
Jan Kiszka 已提交
1022
        len = nb_sectors << BDRV_SECTOR_BITS;
B
bellard 已提交
1023 1024 1025 1026 1027 1028
        buf += len;
        count -= len;
    }

    /* add data from the last sector */
    if (count > 0) {
1029 1030
        if ((ret = bdrv_read(bs, sector_num, tmp_buf, 1)) < 0)
            return ret;
B
bellard 已提交
1031 1032 1033 1034 1035
        memcpy(buf, tmp_buf, count);
    }
    return count1;
}

1036 1037
int bdrv_pwrite(BlockDriverState *bs, int64_t offset,
                const void *buf, int count1)
B
bellard 已提交
1038
{
J
Jan Kiszka 已提交
1039
    uint8_t tmp_buf[BDRV_SECTOR_SIZE];
B
bellard 已提交
1040 1041
    int len, nb_sectors, count;
    int64_t sector_num;
1042
    int ret;
B
bellard 已提交
1043 1044 1045

    count = count1;
    /* first write to align to sector start */
J
Jan Kiszka 已提交
1046
    len = (BDRV_SECTOR_SIZE - offset) & (BDRV_SECTOR_SIZE - 1);
B
bellard 已提交
1047 1048
    if (len > count)
        len = count;
J
Jan Kiszka 已提交
1049
    sector_num = offset >> BDRV_SECTOR_BITS;
B
bellard 已提交
1050
    if (len > 0) {
1051 1052
        if ((ret = bdrv_read(bs, sector_num, tmp_buf, 1)) < 0)
            return ret;
J
Jan Kiszka 已提交
1053
        memcpy(tmp_buf + (offset & (BDRV_SECTOR_SIZE - 1)), buf, len);
1054 1055
        if ((ret = bdrv_write(bs, sector_num, tmp_buf, 1)) < 0)
            return ret;
B
bellard 已提交
1056 1057 1058 1059 1060 1061 1062 1063
        count -= len;
        if (count == 0)
            return count1;
        sector_num++;
        buf += len;
    }

    /* write the sectors "in place" */
J
Jan Kiszka 已提交
1064
    nb_sectors = count >> BDRV_SECTOR_BITS;
B
bellard 已提交
1065
    if (nb_sectors > 0) {
1066 1067
        if ((ret = bdrv_write(bs, sector_num, buf, nb_sectors)) < 0)
            return ret;
B
bellard 已提交
1068
        sector_num += nb_sectors;
J
Jan Kiszka 已提交
1069
        len = nb_sectors << BDRV_SECTOR_BITS;
B
bellard 已提交
1070 1071 1072 1073 1074 1075
        buf += len;
        count -= len;
    }

    /* add data from the last sector */
    if (count > 0) {
1076 1077
        if ((ret = bdrv_read(bs, sector_num, tmp_buf, 1)) < 0)
            return ret;
B
bellard 已提交
1078
        memcpy(tmp_buf, buf, count);
1079 1080
        if ((ret = bdrv_write(bs, sector_num, tmp_buf, 1)) < 0)
            return ret;
B
bellard 已提交
1081 1082 1083 1084
    }
    return count1;
}

K
Kevin Wolf 已提交
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
/*
 * Writes to the file and ensures that no writes are reordered across this
 * request (acts as a barrier)
 *
 * Returns 0 on success, -errno in error cases.
 */
int bdrv_pwrite_sync(BlockDriverState *bs, int64_t offset,
    const void *buf, int count)
{
    int ret;

    ret = bdrv_pwrite(bs, offset, buf, count);
    if (ret < 0) {
        return ret;
    }

    /* No flush needed for cache=writethrough, it uses O_DSYNC */
    if ((bs->open_flags & BDRV_O_CACHE_MASK) != 0) {
        bdrv_flush(bs);
    }

    return 0;
}

/*
 * Writes to the file and ensures that no writes are reordered across this
 * request (acts as a barrier)
 *
 * Returns 0 on success, -errno in error cases.
 */
int bdrv_write_sync(BlockDriverState *bs, int64_t sector_num,
    const uint8_t *buf, int nb_sectors)
{
    return bdrv_pwrite_sync(bs, BDRV_SECTOR_SIZE * sector_num,
        buf, BDRV_SECTOR_SIZE * nb_sectors);
}

B
bellard 已提交
1122 1123 1124 1125 1126 1127
/**
 * Truncate file to 'offset' bytes (needed only for file protocols)
 */
int bdrv_truncate(BlockDriverState *bs, int64_t offset)
{
    BlockDriver *drv = bs->drv;
1128
    int ret;
B
bellard 已提交
1129
    if (!drv)
B
bellard 已提交
1130
        return -ENOMEDIUM;
B
bellard 已提交
1131 1132
    if (!drv->bdrv_truncate)
        return -ENOTSUP;
1133 1134
    if (bs->read_only)
        return -EACCES;
M
Marcelo Tosatti 已提交
1135 1136
    if (bdrv_in_use(bs))
        return -EBUSY;
1137 1138 1139
    ret = drv->bdrv_truncate(bs, offset);
    if (ret == 0) {
        ret = refresh_total_sectors(bs, offset >> BDRV_SECTOR_BITS);
1140 1141 1142
        if (bs->change_cb) {
            bs->change_cb(bs->change_opaque, CHANGE_SIZE);
        }
1143 1144
    }
    return ret;
B
bellard 已提交
1145 1146 1147 1148 1149 1150 1151 1152 1153
}

/**
 * Length of a file in bytes. Return < 0 if error or unknown.
 */
int64_t bdrv_getlength(BlockDriverState *bs)
{
    BlockDriver *drv = bs->drv;
    if (!drv)
B
bellard 已提交
1154
        return -ENOMEDIUM;
1155 1156 1157 1158 1159 1160

    /* Fixed size devices use the total_sectors value for speed instead of
       issuing a length query (like lseek) on each call.  Also, legacy block
       drivers don't provide a bdrv_getlength function and must use
       total_sectors. */
    if (!bs->growable || !drv->bdrv_getlength) {
J
Jan Kiszka 已提交
1161
        return bs->total_sectors * BDRV_SECTOR_SIZE;
B
bellard 已提交
1162 1163
    }
    return drv->bdrv_getlength(bs);
B
bellard 已提交
1164 1165
}

B
bellard 已提交
1166
/* return 0 as number of sectors if no device present or error */
1167
void bdrv_get_geometry(BlockDriverState *bs, uint64_t *nb_sectors_ptr)
B
bellard 已提交
1168
{
B
bellard 已提交
1169 1170 1171 1172 1173
    int64_t length;
    length = bdrv_getlength(bs);
    if (length < 0)
        length = 0;
    else
J
Jan Kiszka 已提交
1174
        length = length >> BDRV_SECTOR_BITS;
B
bellard 已提交
1175
    *nb_sectors_ptr = length;
B
bellard 已提交
1176
}
B
bellard 已提交
1177

1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194
struct partition {
        uint8_t boot_ind;           /* 0x80 - active */
        uint8_t head;               /* starting head */
        uint8_t sector;             /* starting sector */
        uint8_t cyl;                /* starting cylinder */
        uint8_t sys_ind;            /* What partition type */
        uint8_t end_head;           /* end head */
        uint8_t end_sector;         /* end sector */
        uint8_t end_cyl;            /* end cylinder */
        uint32_t start_sect;        /* starting sector counting from 0 */
        uint32_t nr_sects;          /* nr of sectors in partition */
} __attribute__((packed));

/* try to guess the disk logical geometry from the MSDOS partition table. Return 0 if OK, -1 if could not guess */
static int guess_disk_lchs(BlockDriverState *bs,
                           int *pcylinders, int *pheads, int *psectors)
{
1195
    uint8_t buf[BDRV_SECTOR_SIZE];
1196 1197 1198
    int ret, i, heads, sectors, cylinders;
    struct partition *p;
    uint32_t nr_sects;
B
blueswir1 已提交
1199
    uint64_t nb_sectors;
1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238

    bdrv_get_geometry(bs, &nb_sectors);

    ret = bdrv_read(bs, 0, buf, 1);
    if (ret < 0)
        return -1;
    /* test msdos magic */
    if (buf[510] != 0x55 || buf[511] != 0xaa)
        return -1;
    for(i = 0; i < 4; i++) {
        p = ((struct partition *)(buf + 0x1be)) + i;
        nr_sects = le32_to_cpu(p->nr_sects);
        if (nr_sects && p->end_head) {
            /* We make the assumption that the partition terminates on
               a cylinder boundary */
            heads = p->end_head + 1;
            sectors = p->end_sector & 63;
            if (sectors == 0)
                continue;
            cylinders = nb_sectors / (heads * sectors);
            if (cylinders < 1 || cylinders > 16383)
                continue;
            *pheads = heads;
            *psectors = sectors;
            *pcylinders = cylinders;
#if 0
            printf("guessed geometry: LCHS=%d %d %d\n",
                   cylinders, heads, sectors);
#endif
            return 0;
        }
    }
    return -1;
}

void bdrv_guess_geometry(BlockDriverState *bs, int *pcyls, int *pheads, int *psecs)
{
    int translation, lba_detected = 0;
    int cylinders, heads, secs;
B
blueswir1 已提交
1239
    uint64_t nb_sectors;
1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293

    /* if a geometry hint is available, use it */
    bdrv_get_geometry(bs, &nb_sectors);
    bdrv_get_geometry_hint(bs, &cylinders, &heads, &secs);
    translation = bdrv_get_translation_hint(bs);
    if (cylinders != 0) {
        *pcyls = cylinders;
        *pheads = heads;
        *psecs = secs;
    } else {
        if (guess_disk_lchs(bs, &cylinders, &heads, &secs) == 0) {
            if (heads > 16) {
                /* if heads > 16, it means that a BIOS LBA
                   translation was active, so the default
                   hardware geometry is OK */
                lba_detected = 1;
                goto default_geometry;
            } else {
                *pcyls = cylinders;
                *pheads = heads;
                *psecs = secs;
                /* disable any translation to be in sync with
                   the logical geometry */
                if (translation == BIOS_ATA_TRANSLATION_AUTO) {
                    bdrv_set_translation_hint(bs,
                                              BIOS_ATA_TRANSLATION_NONE);
                }
            }
        } else {
        default_geometry:
            /* if no geometry, use a standard physical disk geometry */
            cylinders = nb_sectors / (16 * 63);

            if (cylinders > 16383)
                cylinders = 16383;
            else if (cylinders < 2)
                cylinders = 2;
            *pcyls = cylinders;
            *pheads = 16;
            *psecs = 63;
            if ((lba_detected == 1) && (translation == BIOS_ATA_TRANSLATION_AUTO)) {
                if ((*pcyls * *pheads) <= 131072) {
                    bdrv_set_translation_hint(bs,
                                              BIOS_ATA_TRANSLATION_LARGE);
                } else {
                    bdrv_set_translation_hint(bs,
                                              BIOS_ATA_TRANSLATION_LBA);
                }
            }
        }
        bdrv_set_geometry_hint(bs, *pcyls, *pheads, *psecs);
    }
}

1294
void bdrv_set_geometry_hint(BlockDriverState *bs,
B
bellard 已提交
1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308
                            int cyls, int heads, int secs)
{
    bs->cyls = cyls;
    bs->heads = heads;
    bs->secs = secs;
}

void bdrv_set_type_hint(BlockDriverState *bs, int type)
{
    bs->type = type;
    bs->removable = ((type == BDRV_TYPE_CDROM ||
                      type == BDRV_TYPE_FLOPPY));
}

B
bellard 已提交
1309 1310 1311 1312 1313
void bdrv_set_translation_hint(BlockDriverState *bs, int translation)
{
    bs->translation = translation;
}

1314
void bdrv_get_geometry_hint(BlockDriverState *bs,
B
bellard 已提交
1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326
                            int *pcyls, int *pheads, int *psecs)
{
    *pcyls = bs->cyls;
    *pheads = bs->heads;
    *psecs = bs->secs;
}

int bdrv_get_type_hint(BlockDriverState *bs)
{
    return bs->type;
}

B
bellard 已提交
1327 1328 1329 1330 1331
int bdrv_get_translation_hint(BlockDriverState *bs)
{
    return bs->translation;
}

1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343
void bdrv_set_on_error(BlockDriverState *bs, BlockErrorAction on_read_error,
                       BlockErrorAction on_write_error)
{
    bs->on_read_error = on_read_error;
    bs->on_write_error = on_write_error;
}

BlockErrorAction bdrv_get_on_error(BlockDriverState *bs, int is_read)
{
    return is_read ? bs->on_read_error : bs->on_write_error;
}

1344 1345 1346 1347 1348 1349 1350 1351
void bdrv_set_removable(BlockDriverState *bs, int removable)
{
    bs->removable = removable;
    if (removable && bs == bs_snapshots) {
        bs_snapshots = NULL;
    }
}

B
bellard 已提交
1352 1353 1354 1355 1356 1357 1358 1359 1360 1361
int bdrv_is_removable(BlockDriverState *bs)
{
    return bs->removable;
}

int bdrv_is_read_only(BlockDriverState *bs)
{
    return bs->read_only;
}

1362 1363 1364 1365 1366
int bdrv_is_sg(BlockDriverState *bs)
{
    return bs->sg;
}

1367 1368 1369 1370 1371
int bdrv_enable_write_cache(BlockDriverState *bs)
{
    return bs->enable_write_cache;
}

B
bellard 已提交
1372
/* XXX: no longer used */
1373
void bdrv_set_change_cb(BlockDriverState *bs,
1374 1375
                        void (*change_cb)(void *opaque, int reason),
                        void *opaque)
B
bellard 已提交
1376 1377 1378 1379 1380
{
    bs->change_cb = change_cb;
    bs->change_opaque = opaque;
}

B
bellard 已提交
1381 1382 1383 1384 1385 1386 1387
int bdrv_is_encrypted(BlockDriverState *bs)
{
    if (bs->backing_hd && bs->backing_hd->encrypted)
        return 1;
    return bs->encrypted;
}

1388 1389 1390 1391 1392 1393 1394 1395 1396
int bdrv_key_required(BlockDriverState *bs)
{
    BlockDriverState *backing_hd = bs->backing_hd;

    if (backing_hd && backing_hd->encrypted && !backing_hd->valid_key)
        return 1;
    return (bs->encrypted && !bs->valid_key);
}

B
bellard 已提交
1397 1398 1399 1400 1401 1402 1403 1404 1405 1406
int bdrv_set_key(BlockDriverState *bs, const char *key)
{
    int ret;
    if (bs->backing_hd && bs->backing_hd->encrypted) {
        ret = bdrv_set_key(bs->backing_hd, key);
        if (ret < 0)
            return ret;
        if (!bs->encrypted)
            return 0;
    }
1407 1408 1409 1410 1411
    if (!bs->encrypted) {
        return -EINVAL;
    } else if (!bs->drv || !bs->drv->bdrv_set_key) {
        return -ENOMEDIUM;
    }
1412
    ret = bs->drv->bdrv_set_key(bs, key);
1413 1414 1415 1416 1417 1418 1419
    if (ret < 0) {
        bs->valid_key = 0;
    } else if (!bs->valid_key) {
        bs->valid_key = 1;
        /* call the change callback now, we skipped it on open */
        bs->media_changed = 1;
        if (bs->change_cb)
1420
            bs->change_cb(bs->change_opaque, CHANGE_MEDIA);
1421
    }
1422
    return ret;
B
bellard 已提交
1423 1424 1425 1426
}

void bdrv_get_format(BlockDriverState *bs, char *buf, int buf_size)
{
B
bellard 已提交
1427
    if (!bs->drv) {
B
bellard 已提交
1428 1429 1430 1431 1432 1433
        buf[0] = '\0';
    } else {
        pstrcpy(buf, buf_size, bs->drv->format_name);
    }
}

1434
void bdrv_iterate_format(void (*it)(void *opaque, const char *name),
B
bellard 已提交
1435 1436 1437 1438
                         void *opaque)
{
    BlockDriver *drv;

1439
    QLIST_FOREACH(drv, &bdrv_drivers, list) {
B
bellard 已提交
1440 1441 1442 1443
        it(opaque, drv->format_name);
    }
}

B
bellard 已提交
1444 1445 1446 1447
BlockDriverState *bdrv_find(const char *name)
{
    BlockDriverState *bs;

1448 1449
    QTAILQ_FOREACH(bs, &bdrv_states, list) {
        if (!strcmp(name, bs->device_name)) {
B
bellard 已提交
1450
            return bs;
1451
        }
B
bellard 已提交
1452 1453 1454 1455
    }
    return NULL;
}

M
Markus Armbruster 已提交
1456 1457 1458 1459 1460 1461 1462 1463
BlockDriverState *bdrv_next(BlockDriverState *bs)
{
    if (!bs) {
        return QTAILQ_FIRST(&bdrv_states);
    }
    return QTAILQ_NEXT(bs, list);
}

1464
void bdrv_iterate(void (*it)(void *opaque, BlockDriverState *bs), void *opaque)
B
bellard 已提交
1465 1466 1467
{
    BlockDriverState *bs;

1468
    QTAILQ_FOREACH(bs, &bdrv_states, list) {
1469
        it(opaque, bs);
B
bellard 已提交
1470 1471 1472
    }
}

B
bellard 已提交
1473 1474 1475 1476 1477
const char *bdrv_get_device_name(BlockDriverState *bs)
{
    return bs->device_name;
}

1478
int bdrv_flush(BlockDriverState *bs)
P
pbrook 已提交
1479
{
1480
    if (bs->open_flags & BDRV_O_NO_FLUSH) {
1481 1482 1483 1484 1485
        return 0;
    }

    if (bs->drv && bs->drv->bdrv_flush) {
        return bs->drv->bdrv_flush(bs);
1486 1487
    }

1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498
    /*
     * Some block drivers always operate in either writethrough or unsafe mode
     * and don't support bdrv_flush therefore. Usually qemu doesn't know how
     * the server works (because the behaviour is hardcoded or depends on
     * server-side configuration), so we can't ensure that everything is safe
     * on disk. Returning an error doesn't work because that would break guests
     * even if the server operates in writethrough mode.
     *
     * Let's hope the user knows what he's doing.
     */
    return 0;
P
pbrook 已提交
1499 1500
}

A
aliguori 已提交
1501 1502 1503 1504
void bdrv_flush_all(void)
{
    BlockDriverState *bs;

1505 1506 1507
    QTAILQ_FOREACH(bs, &bdrv_states, list) {
        if (bs->drv && !bdrv_is_read_only(bs) &&
            (!bdrv_is_removable(bs) || bdrv_is_inserted(bs))) {
A
aliguori 已提交
1508
            bdrv_flush(bs);
1509 1510
        }
    }
A
aliguori 已提交
1511 1512
}

K
Kevin Wolf 已提交
1513 1514 1515 1516
int bdrv_has_zero_init(BlockDriverState *bs)
{
    assert(bs->drv);

K
Kevin Wolf 已提交
1517 1518
    if (bs->drv->bdrv_has_zero_init) {
        return bs->drv->bdrv_has_zero_init(bs);
K
Kevin Wolf 已提交
1519 1520 1521 1522 1523
    }

    return 1;
}

C
Christoph Hellwig 已提交
1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534
int bdrv_discard(BlockDriverState *bs, int64_t sector_num, int nb_sectors)
{
    if (!bs->drv) {
        return -ENOMEDIUM;
    }
    if (!bs->drv->bdrv_discard) {
        return 0;
    }
    return bs->drv->bdrv_discard(bs, sector_num, nb_sectors);
}

1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561
/*
 * Returns true iff the specified sector is present in the disk image. Drivers
 * not implementing the functionality are assumed to not support backing files,
 * hence all their sectors are reported as allocated.
 *
 * 'pnum' is set to the number of sectors (including and immediately following
 * the specified sector) that are known to be in the same
 * allocated/unallocated state.
 *
 * 'nb_sectors' is the max value 'pnum' should be set to.
 */
int bdrv_is_allocated(BlockDriverState *bs, int64_t sector_num, int nb_sectors,
	int *pnum)
{
    int64_t n;
    if (!bs->drv->bdrv_is_allocated) {
        if (sector_num >= bs->total_sectors) {
            *pnum = 0;
            return 0;
        }
        n = bs->total_sectors - sector_num;
        *pnum = (n < nb_sectors) ? (n) : (nb_sectors);
        return 1;
    }
    return bs->drv->bdrv_is_allocated(bs, sector_num, nb_sectors, pnum);
}

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
void bdrv_mon_event(const BlockDriverState *bdrv,
                    BlockMonEventAction action, int is_read)
{
    QObject *data;
    const char *action_str;

    switch (action) {
    case BDRV_ACTION_REPORT:
        action_str = "report";
        break;
    case BDRV_ACTION_IGNORE:
        action_str = "ignore";
        break;
    case BDRV_ACTION_STOP:
        action_str = "stop";
        break;
    default:
        abort();
    }

    data = qobject_from_jsonf("{ 'device': %s, 'action': %s, 'operation': %s }",
                              bdrv->device_name,
                              action_str,
                              is_read ? "read" : "write");
    monitor_protocol_event(QEVENT_BLOCK_IO_ERROR, data);

    qobject_decref(data);
}

1591
static void bdrv_print_dict(QObject *obj, void *opaque)
B
bellard 已提交
1592
{
1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634
    QDict *bs_dict;
    Monitor *mon = opaque;

    bs_dict = qobject_to_qdict(obj);

    monitor_printf(mon, "%s: type=%s removable=%d",
                        qdict_get_str(bs_dict, "device"),
                        qdict_get_str(bs_dict, "type"),
                        qdict_get_bool(bs_dict, "removable"));

    if (qdict_get_bool(bs_dict, "removable")) {
        monitor_printf(mon, " locked=%d", qdict_get_bool(bs_dict, "locked"));
    }

    if (qdict_haskey(bs_dict, "inserted")) {
        QDict *qdict = qobject_to_qdict(qdict_get(bs_dict, "inserted"));

        monitor_printf(mon, " file=");
        monitor_print_filename(mon, qdict_get_str(qdict, "file"));
        if (qdict_haskey(qdict, "backing_file")) {
            monitor_printf(mon, " backing_file=");
            monitor_print_filename(mon, qdict_get_str(qdict, "backing_file"));
        }
        monitor_printf(mon, " ro=%d drv=%s encrypted=%d",
                            qdict_get_bool(qdict, "ro"),
                            qdict_get_str(qdict, "drv"),
                            qdict_get_bool(qdict, "encrypted"));
    } else {
        monitor_printf(mon, " [not inserted]");
    }

    monitor_printf(mon, "\n");
}

void bdrv_info_print(Monitor *mon, const QObject *data)
{
    qlist_iter(qobject_to_qlist(data), bdrv_print_dict, mon);
}

void bdrv_info(Monitor *mon, QObject **ret_data)
{
    QList *bs_list;
B
bellard 已提交
1635 1636
    BlockDriverState *bs;

1637 1638
    bs_list = qlist_new();

1639
    QTAILQ_FOREACH(bs, &bdrv_states, list) {
1640 1641 1642
        QObject *bs_obj;
        const char *type = "unknown";

B
bellard 已提交
1643 1644
        switch(bs->type) {
        case BDRV_TYPE_HD:
1645
            type = "hd";
B
bellard 已提交
1646 1647
            break;
        case BDRV_TYPE_CDROM:
1648
            type = "cdrom";
B
bellard 已提交
1649 1650
            break;
        case BDRV_TYPE_FLOPPY:
1651
            type = "floppy";
B
bellard 已提交
1652 1653
            break;
        }
1654 1655 1656 1657 1658 1659

        bs_obj = qobject_from_jsonf("{ 'device': %s, 'type': %s, "
                                    "'removable': %i, 'locked': %i }",
                                    bs->device_name, type, bs->removable,
                                    bs->locked);

B
bellard 已提交
1660
        if (bs->drv) {
1661 1662 1663 1664 1665 1666 1667 1668
            QObject *obj;
            QDict *bs_dict = qobject_to_qdict(bs_obj);

            obj = qobject_from_jsonf("{ 'file': %s, 'ro': %i, 'drv': %s, "
                                     "'encrypted': %i }",
                                     bs->filename, bs->read_only,
                                     bs->drv->format_name,
                                     bdrv_is_encrypted(bs));
1669
            if (bs->backing_file[0] != '\0') {
1670 1671 1672
                QDict *qdict = qobject_to_qdict(obj);
                qdict_put(qdict, "backing_file",
                          qstring_from_str(bs->backing_file));
A
aliguori 已提交
1673
            }
1674 1675

            qdict_put_obj(bs_dict, "inserted", obj);
B
bellard 已提交
1676
        }
1677
        qlist_append_obj(bs_list, bs_obj);
B
bellard 已提交
1678
    }
1679 1680

    *ret_data = QOBJECT(bs_list);
B
bellard 已提交
1681
}
1682

1683
static void bdrv_stats_iter(QObject *data, void *opaque)
1684
{
1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707
    QDict *qdict;
    Monitor *mon = opaque;

    qdict = qobject_to_qdict(data);
    monitor_printf(mon, "%s:", qdict_get_str(qdict, "device"));

    qdict = qobject_to_qdict(qdict_get(qdict, "stats"));
    monitor_printf(mon, " rd_bytes=%" PRId64
                        " wr_bytes=%" PRId64
                        " rd_operations=%" PRId64
                        " wr_operations=%" PRId64
                        "\n",
                        qdict_get_int(qdict, "rd_bytes"),
                        qdict_get_int(qdict, "wr_bytes"),
                        qdict_get_int(qdict, "rd_operations"),
                        qdict_get_int(qdict, "wr_operations"));
}

void bdrv_stats_print(Monitor *mon, const QObject *data)
{
    qlist_iter(qobject_to_qlist(data), bdrv_stats_iter, mon);
}

1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721
static QObject* bdrv_info_stats_bs(BlockDriverState *bs)
{
    QObject *res;
    QDict *dict;

    res = qobject_from_jsonf("{ 'stats': {"
                             "'rd_bytes': %" PRId64 ","
                             "'wr_bytes': %" PRId64 ","
                             "'rd_operations': %" PRId64 ","
                             "'wr_operations': %" PRId64 ","
                             "'wr_highest_offset': %" PRId64
                             "} }",
                             bs->rd_bytes, bs->wr_bytes,
                             bs->rd_ops, bs->wr_ops,
1722 1723
                             bs->wr_highest_sector *
                             (uint64_t)BDRV_SECTOR_SIZE);
1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737
    dict  = qobject_to_qdict(res);

    if (*bs->device_name) {
        qdict_put(dict, "device", qstring_from_str(bs->device_name));
    }

    if (bs->file) {
        QObject *parent = bdrv_info_stats_bs(bs->file);
        qdict_put_obj(dict, "parent", parent);
    }

    return res;
}

1738 1739 1740 1741
void bdrv_info_stats(Monitor *mon, QObject **ret_data)
{
    QObject *obj;
    QList *devices;
1742 1743
    BlockDriverState *bs;

1744 1745
    devices = qlist_new();

1746
    QTAILQ_FOREACH(bs, &bdrv_states, list) {
1747
        obj = bdrv_info_stats_bs(bs);
1748
        qlist_append_obj(devices, obj);
1749
    }
1750 1751

    *ret_data = QOBJECT(devices);
1752
}
B
bellard 已提交
1753

1754 1755 1756 1757 1758 1759 1760 1761 1762 1763
const char *bdrv_get_encrypted_filename(BlockDriverState *bs)
{
    if (bs->backing_hd && bs->backing_hd->encrypted)
        return bs->backing_file;
    else if (bs->encrypted)
        return bs->filename;
    else
        return NULL;
}

1764
void bdrv_get_backing_filename(BlockDriverState *bs,
B
bellard 已提交
1765 1766
                               char *filename, int filename_size)
{
K
Kevin Wolf 已提交
1767
    if (!bs->backing_file) {
B
bellard 已提交
1768 1769 1770 1771 1772 1773
        pstrcpy(filename, filename_size, "");
    } else {
        pstrcpy(filename, filename_size, bs->backing_file);
    }
}

1774
int bdrv_write_compressed(BlockDriverState *bs, int64_t sector_num,
B
bellard 已提交
1775 1776 1777 1778
                          const uint8_t *buf, int nb_sectors)
{
    BlockDriver *drv = bs->drv;
    if (!drv)
B
bellard 已提交
1779
        return -ENOMEDIUM;
B
bellard 已提交
1780 1781
    if (!drv->bdrv_write_compressed)
        return -ENOTSUP;
K
Kevin Wolf 已提交
1782 1783
    if (bdrv_check_request(bs, sector_num, nb_sectors))
        return -EIO;
1784

1785
    if (bs->dirty_bitmap) {
1786 1787
        set_dirty_bitmap(bs, sector_num, nb_sectors, 1);
    }
1788

B
bellard 已提交
1789 1790
    return drv->bdrv_write_compressed(bs, sector_num, buf, nb_sectors);
}
1791

B
bellard 已提交
1792 1793 1794 1795
int bdrv_get_info(BlockDriverState *bs, BlockDriverInfo *bdi)
{
    BlockDriver *drv = bs->drv;
    if (!drv)
B
bellard 已提交
1796
        return -ENOMEDIUM;
B
bellard 已提交
1797 1798 1799 1800 1801 1802
    if (!drv->bdrv_get_info)
        return -ENOTSUP;
    memset(bdi, 0, sizeof(*bdi));
    return drv->bdrv_get_info(bs, bdi);
}

1803 1804
int bdrv_save_vmstate(BlockDriverState *bs, const uint8_t *buf,
                      int64_t pos, int size)
1805 1806 1807 1808
{
    BlockDriver *drv = bs->drv;
    if (!drv)
        return -ENOMEDIUM;
1809 1810 1811 1812 1813
    if (drv->bdrv_save_vmstate)
        return drv->bdrv_save_vmstate(bs, buf, pos, size);
    if (bs->file)
        return bdrv_save_vmstate(bs->file, buf, pos, size);
    return -ENOTSUP;
1814 1815
}

1816 1817
int bdrv_load_vmstate(BlockDriverState *bs, uint8_t *buf,
                      int64_t pos, int size)
1818 1819 1820 1821
{
    BlockDriver *drv = bs->drv;
    if (!drv)
        return -ENOMEDIUM;
1822 1823 1824 1825 1826
    if (drv->bdrv_load_vmstate)
        return drv->bdrv_load_vmstate(bs, buf, pos, size);
    if (bs->file)
        return bdrv_load_vmstate(bs->file, buf, pos, size);
    return -ENOTSUP;
1827 1828
}

K
Kevin Wolf 已提交
1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840
void bdrv_debug_event(BlockDriverState *bs, BlkDebugEvent event)
{
    BlockDriver *drv = bs->drv;

    if (!drv || !drv->bdrv_debug_event) {
        return;
    }

    return drv->bdrv_debug_event(bs, event);

}

B
bellard 已提交
1841 1842 1843
/**************************************************************/
/* handling of snapshots */

1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860
int bdrv_can_snapshot(BlockDriverState *bs)
{
    BlockDriver *drv = bs->drv;
    if (!drv || bdrv_is_removable(bs) || bdrv_is_read_only(bs)) {
        return 0;
    }

    if (!drv->bdrv_snapshot_create) {
        if (bs->file != NULL) {
            return bdrv_can_snapshot(bs->file);
        }
        return 0;
    }

    return 1;
}

1861 1862 1863 1864 1865
int bdrv_is_snapshot(BlockDriverState *bs)
{
    return !!(bs->open_flags & BDRV_O_SNAPSHOT);
}

1866 1867 1868 1869
BlockDriverState *bdrv_snapshots(void)
{
    BlockDriverState *bs;

1870
    if (bs_snapshots) {
1871
        return bs_snapshots;
1872
    }
1873 1874 1875 1876

    bs = NULL;
    while ((bs = bdrv_next(bs))) {
        if (bdrv_can_snapshot(bs)) {
1877 1878
            bs_snapshots = bs;
            return bs;
1879 1880 1881 1882 1883
        }
    }
    return NULL;
}

1884
int bdrv_snapshot_create(BlockDriverState *bs,
B
bellard 已提交
1885 1886 1887 1888
                         QEMUSnapshotInfo *sn_info)
{
    BlockDriver *drv = bs->drv;
    if (!drv)
B
bellard 已提交
1889
        return -ENOMEDIUM;
1890 1891 1892 1893 1894
    if (drv->bdrv_snapshot_create)
        return drv->bdrv_snapshot_create(bs, sn_info);
    if (bs->file)
        return bdrv_snapshot_create(bs->file, sn_info);
    return -ENOTSUP;
B
bellard 已提交
1895 1896
}

1897
int bdrv_snapshot_goto(BlockDriverState *bs,
B
bellard 已提交
1898 1899 1900
                       const char *snapshot_id)
{
    BlockDriver *drv = bs->drv;
1901 1902
    int ret, open_ret;

B
bellard 已提交
1903
    if (!drv)
B
bellard 已提交
1904
        return -ENOMEDIUM;
1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920
    if (drv->bdrv_snapshot_goto)
        return drv->bdrv_snapshot_goto(bs, snapshot_id);

    if (bs->file) {
        drv->bdrv_close(bs);
        ret = bdrv_snapshot_goto(bs->file, snapshot_id);
        open_ret = drv->bdrv_open(bs, bs->open_flags);
        if (open_ret < 0) {
            bdrv_delete(bs->file);
            bs->drv = NULL;
            return open_ret;
        }
        return ret;
    }

    return -ENOTSUP;
B
bellard 已提交
1921 1922 1923 1924 1925 1926
}

int bdrv_snapshot_delete(BlockDriverState *bs, const char *snapshot_id)
{
    BlockDriver *drv = bs->drv;
    if (!drv)
B
bellard 已提交
1927
        return -ENOMEDIUM;
1928 1929 1930 1931 1932
    if (drv->bdrv_snapshot_delete)
        return drv->bdrv_snapshot_delete(bs, snapshot_id);
    if (bs->file)
        return bdrv_snapshot_delete(bs->file, snapshot_id);
    return -ENOTSUP;
B
bellard 已提交
1933 1934
}

1935
int bdrv_snapshot_list(BlockDriverState *bs,
B
bellard 已提交
1936 1937 1938 1939
                       QEMUSnapshotInfo **psn_info)
{
    BlockDriver *drv = bs->drv;
    if (!drv)
B
bellard 已提交
1940
        return -ENOMEDIUM;
1941 1942 1943 1944 1945
    if (drv->bdrv_snapshot_list)
        return drv->bdrv_snapshot_list(bs, psn_info);
    if (bs->file)
        return bdrv_snapshot_list(bs->file, psn_info);
    return -ENOTSUP;
B
bellard 已提交
1946 1947
}

E
edison 已提交
1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963
int bdrv_snapshot_load_tmp(BlockDriverState *bs,
        const char *snapshot_name)
{
    BlockDriver *drv = bs->drv;
    if (!drv) {
        return -ENOMEDIUM;
    }
    if (!bs->read_only) {
        return -EINVAL;
    }
    if (drv->bdrv_snapshot_load_tmp) {
        return drv->bdrv_snapshot_load_tmp(bs, snapshot_name);
    }
    return -ENOTSUP;
}

B
bellard 已提交
1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977
#define NB_SUFFIXES 4

char *get_human_readable_size(char *buf, int buf_size, int64_t size)
{
    static const char suffixes[NB_SUFFIXES] = "KMGT";
    int64_t base;
    int i;

    if (size <= 999) {
        snprintf(buf, buf_size, "%" PRId64, size);
    } else {
        base = 1024;
        for(i = 0; i < NB_SUFFIXES; i++) {
            if (size < (10 * base)) {
1978
                snprintf(buf, buf_size, "%0.1f%c",
B
bellard 已提交
1979 1980 1981 1982
                         (double)size / base,
                         suffixes[i]);
                break;
            } else if (size < (1000 * base) || i == (NB_SUFFIXES - 1)) {
1983
                snprintf(buf, buf_size, "%" PRId64 "%c",
B
bellard 已提交
1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996
                         ((size + (base >> 1)) / base),
                         suffixes[i]);
                break;
            }
            base = base * 1024;
        }
    }
    return buf;
}

char *bdrv_snapshot_dump(char *buf, int buf_size, QEMUSnapshotInfo *sn)
{
    char buf1[128], date_buf[128], clock_buf[128];
1997 1998 1999
#ifdef _WIN32
    struct tm *ptm;
#else
B
bellard 已提交
2000
    struct tm tm;
2001
#endif
B
bellard 已提交
2002 2003 2004 2005
    time_t ti;
    int64_t secs;

    if (!sn) {
2006 2007
        snprintf(buf, buf_size,
                 "%-10s%-20s%7s%20s%15s",
B
bellard 已提交
2008 2009 2010
                 "ID", "TAG", "VM SIZE", "DATE", "VM CLOCK");
    } else {
        ti = sn->date_sec;
2011 2012 2013 2014 2015
#ifdef _WIN32
        ptm = localtime(&ti);
        strftime(date_buf, sizeof(date_buf),
                 "%Y-%m-%d %H:%M:%S", ptm);
#else
B
bellard 已提交
2016 2017 2018
        localtime_r(&ti, &tm);
        strftime(date_buf, sizeof(date_buf),
                 "%Y-%m-%d %H:%M:%S", &tm);
2019
#endif
B
bellard 已提交
2020 2021 2022 2023 2024
        secs = sn->vm_clock_nsec / 1000000000;
        snprintf(clock_buf, sizeof(clock_buf),
                 "%02d:%02d:%02d.%03d",
                 (int)(secs / 3600),
                 (int)((secs / 60) % 60),
2025
                 (int)(secs % 60),
B
bellard 已提交
2026 2027
                 (int)((sn->vm_clock_nsec / 1000000) % 1000));
        snprintf(buf, buf_size,
2028
                 "%-10s%-20s%7s%20s%15s",
B
bellard 已提交
2029 2030 2031 2032 2033 2034 2035 2036
                 sn->id_str, sn->name,
                 get_human_readable_size(buf1, sizeof(buf1), sn->vm_state_size),
                 date_buf,
                 clock_buf);
    }
    return buf;
}

B
bellard 已提交
2037

B
bellard 已提交
2038
/**************************************************************/
B
bellard 已提交
2039
/* async I/Os */
B
bellard 已提交
2040

2041
BlockDriverAIOCB *bdrv_aio_readv(BlockDriverState *bs, int64_t sector_num,
2042
                                 QEMUIOVector *qiov, int nb_sectors,
2043
                                 BlockDriverCompletionFunc *cb, void *opaque)
B
bellard 已提交
2044 2045
{
    BlockDriver *drv = bs->drv;
2046
    BlockDriverAIOCB *ret;
B
bellard 已提交
2047

2048 2049
    trace_bdrv_aio_readv(bs, sector_num, nb_sectors, opaque);

B
bellard 已提交
2050
    if (!drv)
2051
        return NULL;
2052 2053
    if (bdrv_check_request(bs, sector_num, nb_sectors))
        return NULL;
2054

2055 2056
    ret = drv->bdrv_aio_readv(bs, sector_num, qiov, nb_sectors,
                              cb, opaque);
2057 2058 2059

    if (ret) {
	/* Update stats even though technically transfer has not happened. */
J
Jan Kiszka 已提交
2060
	bs->rd_bytes += (unsigned) nb_sectors * BDRV_SECTOR_SIZE;
2061 2062 2063 2064
	bs->rd_ops ++;
    }

    return ret;
B
bellard 已提交
2065 2066
}

2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102
typedef struct BlockCompleteData {
    BlockDriverCompletionFunc *cb;
    void *opaque;
    BlockDriverState *bs;
    int64_t sector_num;
    int nb_sectors;
} BlockCompleteData;

static void block_complete_cb(void *opaque, int ret)
{
    BlockCompleteData *b = opaque;

    if (b->bs->dirty_bitmap) {
        set_dirty_bitmap(b->bs, b->sector_num, b->nb_sectors, 1);
    }
    b->cb(b->opaque, ret);
    qemu_free(b);
}

static BlockCompleteData *blk_dirty_cb_alloc(BlockDriverState *bs,
                                             int64_t sector_num,
                                             int nb_sectors,
                                             BlockDriverCompletionFunc *cb,
                                             void *opaque)
{
    BlockCompleteData *blkdata = qemu_mallocz(sizeof(BlockCompleteData));

    blkdata->bs = bs;
    blkdata->cb = cb;
    blkdata->opaque = opaque;
    blkdata->sector_num = sector_num;
    blkdata->nb_sectors = nb_sectors;

    return blkdata;
}

2103 2104 2105
BlockDriverAIOCB *bdrv_aio_writev(BlockDriverState *bs, int64_t sector_num,
                                  QEMUIOVector *qiov, int nb_sectors,
                                  BlockDriverCompletionFunc *cb, void *opaque)
B
bellard 已提交
2106
{
B
bellard 已提交
2107
    BlockDriver *drv = bs->drv;
2108
    BlockDriverAIOCB *ret;
2109
    BlockCompleteData *blk_cb_data;
B
bellard 已提交
2110

2111 2112
    trace_bdrv_aio_writev(bs, sector_num, nb_sectors, opaque);

B
bellard 已提交
2113
    if (!drv)
2114
        return NULL;
B
bellard 已提交
2115
    if (bs->read_only)
2116
        return NULL;
2117 2118
    if (bdrv_check_request(bs, sector_num, nb_sectors))
        return NULL;
B
bellard 已提交
2119

2120
    if (bs->dirty_bitmap) {
2121 2122 2123 2124
        blk_cb_data = blk_dirty_cb_alloc(bs, sector_num, nb_sectors, cb,
                                         opaque);
        cb = &block_complete_cb;
        opaque = blk_cb_data;
2125
    }
2126

2127 2128
    ret = drv->bdrv_aio_writev(bs, sector_num, qiov, nb_sectors,
                               cb, opaque);
2129 2130

    if (ret) {
2131 2132 2133 2134 2135 2136
        /* Update stats even though technically transfer has not happened. */
        bs->wr_bytes += (unsigned) nb_sectors * BDRV_SECTOR_SIZE;
        bs->wr_ops ++;
        if (bs->wr_highest_sector < sector_num + nb_sectors - 1) {
            bs->wr_highest_sector = sector_num + nb_sectors - 1;
        }
2137 2138 2139
    }

    return ret;
B
bellard 已提交
2140 2141
}

K
Kevin Wolf 已提交
2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160

typedef struct MultiwriteCB {
    int error;
    int num_requests;
    int num_callbacks;
    struct {
        BlockDriverCompletionFunc *cb;
        void *opaque;
        QEMUIOVector *free_qiov;
        void *free_buf;
    } callbacks[];
} MultiwriteCB;

static void multiwrite_user_cb(MultiwriteCB *mcb)
{
    int i;

    for (i = 0; i < mcb->num_callbacks; i++) {
        mcb->callbacks[i].cb(mcb->callbacks[i].opaque, mcb->error);
2161 2162 2163
        if (mcb->callbacks[i].free_qiov) {
            qemu_iovec_destroy(mcb->callbacks[i].free_qiov);
        }
K
Kevin Wolf 已提交
2164
        qemu_free(mcb->callbacks[i].free_qiov);
2165
        qemu_vfree(mcb->callbacks[i].free_buf);
K
Kevin Wolf 已提交
2166 2167 2168 2169 2170 2171 2172
    }
}

static void multiwrite_cb(void *opaque, int ret)
{
    MultiwriteCB *mcb = opaque;

2173 2174
    trace_multiwrite_cb(mcb, ret);

2175
    if (ret < 0 && !mcb->error) {
K
Kevin Wolf 已提交
2176 2177 2178 2179 2180
        mcb->error = ret;
    }

    mcb->num_requests--;
    if (mcb->num_requests == 0) {
2181
        multiwrite_user_cb(mcb);
K
Kevin Wolf 已提交
2182 2183 2184 2185 2186 2187
        qemu_free(mcb);
    }
}

static int multiwrite_req_compare(const void *a, const void *b)
{
2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200
    const BlockRequest *req1 = a, *req2 = b;

    /*
     * Note that we can't simply subtract req2->sector from req1->sector
     * here as that could overflow the return value.
     */
    if (req1->sector > req2->sector) {
        return 1;
    } else if (req1->sector < req2->sector) {
        return -1;
    } else {
        return 0;
    }
K
Kevin Wolf 已提交
2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235
}

/*
 * Takes a bunch of requests and tries to merge them. Returns the number of
 * requests that remain after merging.
 */
static int multiwrite_merge(BlockDriverState *bs, BlockRequest *reqs,
    int num_reqs, MultiwriteCB *mcb)
{
    int i, outidx;

    // Sort requests by start sector
    qsort(reqs, num_reqs, sizeof(*reqs), &multiwrite_req_compare);

    // Check if adjacent requests touch the same clusters. If so, combine them,
    // filling up gaps with zero sectors.
    outidx = 0;
    for (i = 1; i < num_reqs; i++) {
        int merge = 0;
        int64_t oldreq_last = reqs[outidx].sector + reqs[outidx].nb_sectors;

        // This handles the cases that are valid for all block drivers, namely
        // exactly sequential writes and overlapping writes.
        if (reqs[i].sector <= oldreq_last) {
            merge = 1;
        }

        // The block driver may decide that it makes sense to combine requests
        // even if there is a gap of some sectors between them. In this case,
        // the gap is filled with zeros (therefore only applicable for yet
        // unused space in format like qcow2).
        if (!merge && bs->drv->bdrv_merge_requests) {
            merge = bs->drv->bdrv_merge_requests(bs, &reqs[outidx], &reqs[i]);
        }

2236 2237 2238 2239
        if (reqs[outidx].qiov->niov + reqs[i].qiov->niov + 1 > IOV_MAX) {
            merge = 0;
        }

K
Kevin Wolf 已提交
2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262
        if (merge) {
            size_t size;
            QEMUIOVector *qiov = qemu_mallocz(sizeof(*qiov));
            qemu_iovec_init(qiov,
                reqs[outidx].qiov->niov + reqs[i].qiov->niov + 1);

            // Add the first request to the merged one. If the requests are
            // overlapping, drop the last sectors of the first request.
            size = (reqs[i].sector - reqs[outidx].sector) << 9;
            qemu_iovec_concat(qiov, reqs[outidx].qiov, size);

            // We might need to add some zeros between the two requests
            if (reqs[i].sector > oldreq_last) {
                size_t zero_bytes = (reqs[i].sector - oldreq_last) << 9;
                uint8_t *buf = qemu_blockalign(bs, zero_bytes);
                memset(buf, 0, zero_bytes);
                qemu_iovec_add(qiov, buf, zero_bytes);
                mcb->callbacks[i].free_buf = buf;
            }

            // Add the second request
            qemu_iovec_concat(qiov, reqs[i].qiov, reqs[i].qiov->size);

2263
            reqs[outidx].nb_sectors = qiov->size >> 9;
K
Kevin Wolf 已提交
2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314
            reqs[outidx].qiov = qiov;

            mcb->callbacks[i].free_qiov = reqs[outidx].qiov;
        } else {
            outidx++;
            reqs[outidx].sector     = reqs[i].sector;
            reqs[outidx].nb_sectors = reqs[i].nb_sectors;
            reqs[outidx].qiov       = reqs[i].qiov;
        }
    }

    return outidx + 1;
}

/*
 * Submit multiple AIO write requests at once.
 *
 * On success, the function returns 0 and all requests in the reqs array have
 * been submitted. In error case this function returns -1, and any of the
 * requests may or may not be submitted yet. In particular, this means that the
 * callback will be called for some of the requests, for others it won't. The
 * caller must check the error field of the BlockRequest to wait for the right
 * callbacks (if error != 0, no callback will be called).
 *
 * The implementation may modify the contents of the reqs array, e.g. to merge
 * requests. However, the fields opaque and error are left unmodified as they
 * are used to signal failure for a single request to the caller.
 */
int bdrv_aio_multiwrite(BlockDriverState *bs, BlockRequest *reqs, int num_reqs)
{
    BlockDriverAIOCB *acb;
    MultiwriteCB *mcb;
    int i;

    if (num_reqs == 0) {
        return 0;
    }

    // Create MultiwriteCB structure
    mcb = qemu_mallocz(sizeof(*mcb) + num_reqs * sizeof(*mcb->callbacks));
    mcb->num_requests = 0;
    mcb->num_callbacks = num_reqs;

    for (i = 0; i < num_reqs; i++) {
        mcb->callbacks[i].cb = reqs[i].cb;
        mcb->callbacks[i].opaque = reqs[i].opaque;
    }

    // Check for mergable requests
    num_reqs = multiwrite_merge(bs, reqs, num_reqs, mcb);

2315 2316
    trace_bdrv_aio_multiwrite(mcb, mcb->num_callbacks, num_reqs);

2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337
    /*
     * Run the aio requests. As soon as one request can't be submitted
     * successfully, fail all requests that are not yet submitted (we must
     * return failure for all requests anyway)
     *
     * num_requests cannot be set to the right value immediately: If
     * bdrv_aio_writev fails for some request, num_requests would be too high
     * and therefore multiwrite_cb() would never recognize the multiwrite
     * request as completed. We also cannot use the loop variable i to set it
     * when the first request fails because the callback may already have been
     * called for previously submitted requests. Thus, num_requests must be
     * incremented for each request that is submitted.
     *
     * The problem that callbacks may be called early also means that we need
     * to take care that num_requests doesn't become 0 before all requests are
     * submitted - multiwrite_cb() would consider the multiwrite request
     * completed. A dummy request that is "completed" by a manual call to
     * multiwrite_cb() takes care of this.
     */
    mcb->num_requests = 1;

2338
    // Run the aio requests
K
Kevin Wolf 已提交
2339
    for (i = 0; i < num_reqs; i++) {
2340
        mcb->num_requests++;
K
Kevin Wolf 已提交
2341 2342 2343 2344 2345 2346 2347
        acb = bdrv_aio_writev(bs, reqs[i].sector, reqs[i].qiov,
            reqs[i].nb_sectors, multiwrite_cb, mcb);

        if (acb == NULL) {
            // We can only fail the whole thing if no request has been
            // submitted yet. Otherwise we'll wait for the submitted AIOs to
            // complete and report the error in the callback.
2348
            if (i == 0) {
2349
                trace_bdrv_aio_multiwrite_earlyfail(mcb);
K
Kevin Wolf 已提交
2350 2351
                goto fail;
            } else {
2352
                trace_bdrv_aio_multiwrite_latefail(mcb, i);
2353
                multiwrite_cb(mcb, -EIO);
K
Kevin Wolf 已提交
2354 2355 2356 2357 2358
                break;
            }
        }
    }

2359 2360 2361
    /* Complete the dummy request */
    multiwrite_cb(mcb, 0);

K
Kevin Wolf 已提交
2362 2363 2364
    return 0;

fail:
2365 2366 2367
    for (i = 0; i < mcb->num_callbacks; i++) {
        reqs[i].error = -EIO;
    }
B
Bruce Rogers 已提交
2368
    qemu_free(mcb);
K
Kevin Wolf 已提交
2369 2370 2371
    return -1;
}

2372 2373 2374 2375 2376
BlockDriverAIOCB *bdrv_aio_flush(BlockDriverState *bs,
        BlockDriverCompletionFunc *cb, void *opaque)
{
    BlockDriver *drv = bs->drv;

2377 2378 2379 2380
    if (bs->open_flags & BDRV_O_NO_FLUSH) {
        return bdrv_aio_noop_em(bs, cb, opaque);
    }

2381 2382 2383 2384 2385
    if (!drv)
        return NULL;
    return drv->bdrv_aio_flush(bs, cb, opaque);
}

B
bellard 已提交
2386 2387
void bdrv_aio_cancel(BlockDriverAIOCB *acb)
{
2388
    acb->pool->cancel(acb);
B
bellard 已提交
2389 2390
}

2391

B
bellard 已提交
2392 2393 2394
/**************************************************************/
/* async block device emulation */

2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406
typedef struct BlockDriverAIOCBSync {
    BlockDriverAIOCB common;
    QEMUBH *bh;
    int ret;
    /* vector translation state */
    QEMUIOVector *qiov;
    uint8_t *bounce;
    int is_write;
} BlockDriverAIOCBSync;

static void bdrv_aio_cancel_em(BlockDriverAIOCB *blockacb)
{
2407 2408
    BlockDriverAIOCBSync *acb =
        container_of(blockacb, BlockDriverAIOCBSync, common);
2409
    qemu_bh_delete(acb->bh);
A
Avi Kivity 已提交
2410
    acb->bh = NULL;
2411 2412 2413 2414 2415 2416 2417 2418
    qemu_aio_release(acb);
}

static AIOPool bdrv_em_aio_pool = {
    .aiocb_size         = sizeof(BlockDriverAIOCBSync),
    .cancel             = bdrv_aio_cancel_em,
};

2419
static void bdrv_aio_bh_cb(void *opaque)
B
bellard 已提交
2420
{
2421
    BlockDriverAIOCBSync *acb = opaque;
2422 2423 2424

    if (!acb->is_write)
        qemu_iovec_from_buffer(acb->qiov, acb->bounce, acb->qiov->size);
2425
    qemu_vfree(acb->bounce);
2426
    acb->common.cb(acb->common.opaque, acb->ret);
2427
    qemu_bh_delete(acb->bh);
A
Avi Kivity 已提交
2428
    acb->bh = NULL;
2429
    qemu_aio_release(acb);
B
bellard 已提交
2430
}
2431

2432 2433 2434 2435 2436 2437 2438 2439
static BlockDriverAIOCB *bdrv_aio_rw_vector(BlockDriverState *bs,
                                            int64_t sector_num,
                                            QEMUIOVector *qiov,
                                            int nb_sectors,
                                            BlockDriverCompletionFunc *cb,
                                            void *opaque,
                                            int is_write)

B
bellard 已提交
2440
{
2441 2442
    BlockDriverAIOCBSync *acb;

2443
    acb = qemu_aio_get(&bdrv_em_aio_pool, bs, cb, opaque);
2444 2445
    acb->is_write = is_write;
    acb->qiov = qiov;
2446
    acb->bounce = qemu_blockalign(bs, qiov->size);
2447

2448 2449
    if (!acb->bh)
        acb->bh = qemu_bh_new(bdrv_aio_bh_cb, acb);
2450 2451 2452 2453 2454 2455 2456 2457

    if (is_write) {
        qemu_iovec_to_buffer(acb->qiov, acb->bounce);
        acb->ret = bdrv_write(bs, sector_num, acb->bounce, nb_sectors);
    } else {
        acb->ret = bdrv_read(bs, sector_num, acb->bounce, nb_sectors);
    }

2458
    qemu_bh_schedule(acb->bh);
2459

2460
    return &acb->common;
2461 2462
}

2463 2464
static BlockDriverAIOCB *bdrv_aio_readv_em(BlockDriverState *bs,
        int64_t sector_num, QEMUIOVector *qiov, int nb_sectors,
2465
        BlockDriverCompletionFunc *cb, void *opaque)
2466
{
2467 2468
    return bdrv_aio_rw_vector(bs, sector_num, qiov, nb_sectors, cb, opaque, 0);
}
B
bellard 已提交
2469

2470 2471 2472 2473 2474
static BlockDriverAIOCB *bdrv_aio_writev_em(BlockDriverState *bs,
        int64_t sector_num, QEMUIOVector *qiov, int nb_sectors,
        BlockDriverCompletionFunc *cb, void *opaque)
{
    return bdrv_aio_rw_vector(bs, sector_num, qiov, nb_sectors, cb, opaque, 1);
2475 2476
}

2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495
static BlockDriverAIOCB *bdrv_aio_flush_em(BlockDriverState *bs,
        BlockDriverCompletionFunc *cb, void *opaque)
{
    BlockDriverAIOCBSync *acb;

    acb = qemu_aio_get(&bdrv_em_aio_pool, bs, cb, opaque);
    acb->is_write = 1; /* don't bounce in the completion hadler */
    acb->qiov = NULL;
    acb->bounce = NULL;
    acb->ret = 0;

    if (!acb->bh)
        acb->bh = qemu_bh_new(bdrv_aio_bh_cb, acb);

    bdrv_flush(bs);
    qemu_bh_schedule(acb->bh);
    return &acb->common;
}

2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514
static BlockDriverAIOCB *bdrv_aio_noop_em(BlockDriverState *bs,
        BlockDriverCompletionFunc *cb, void *opaque)
{
    BlockDriverAIOCBSync *acb;

    acb = qemu_aio_get(&bdrv_em_aio_pool, bs, cb, opaque);
    acb->is_write = 1; /* don't bounce in the completion handler */
    acb->qiov = NULL;
    acb->bounce = NULL;
    acb->ret = 0;

    if (!acb->bh) {
        acb->bh = qemu_bh_new(bdrv_aio_bh_cb, acb);
    }

    qemu_bh_schedule(acb->bh);
    return &acb->common;
}

B
bellard 已提交
2515 2516
/**************************************************************/
/* sync block device emulation */
B
bellard 已提交
2517

B
bellard 已提交
2518 2519 2520
static void bdrv_rw_em_cb(void *opaque, int ret)
{
    *(int *)opaque = ret;
B
bellard 已提交
2521 2522
}

B
bellard 已提交
2523 2524
#define NOT_DONE 0x7fffffff

2525
static int bdrv_read_em(BlockDriverState *bs, int64_t sector_num,
B
bellard 已提交
2526
                        uint8_t *buf, int nb_sectors)
P
pbrook 已提交
2527
{
2528 2529
    int async_ret;
    BlockDriverAIOCB *acb;
2530 2531
    struct iovec iov;
    QEMUIOVector qiov;
B
bellard 已提交
2532

2533 2534
    async_context_push();

B
bellard 已提交
2535
    async_ret = NOT_DONE;
2536
    iov.iov_base = (void *)buf;
2537
    iov.iov_len = nb_sectors * BDRV_SECTOR_SIZE;
2538 2539 2540
    qemu_iovec_init_external(&qiov, &iov, 1);
    acb = bdrv_aio_readv(bs, sector_num, &qiov, nb_sectors,
        bdrv_rw_em_cb, &async_ret);
2541 2542 2543 2544
    if (acb == NULL) {
        async_ret = -1;
        goto fail;
    }
2545

B
bellard 已提交
2546 2547 2548
    while (async_ret == NOT_DONE) {
        qemu_aio_wait();
    }
2549

2550 2551 2552

fail:
    async_context_pop();
B
bellard 已提交
2553
    return async_ret;
P
pbrook 已提交
2554 2555
}

B
bellard 已提交
2556 2557 2558
static int bdrv_write_em(BlockDriverState *bs, int64_t sector_num,
                         const uint8_t *buf, int nb_sectors)
{
2559 2560
    int async_ret;
    BlockDriverAIOCB *acb;
2561 2562
    struct iovec iov;
    QEMUIOVector qiov;
B
bellard 已提交
2563

2564 2565
    async_context_push();

B
bellard 已提交
2566
    async_ret = NOT_DONE;
2567
    iov.iov_base = (void *)buf;
2568
    iov.iov_len = nb_sectors * BDRV_SECTOR_SIZE;
2569 2570 2571
    qemu_iovec_init_external(&qiov, &iov, 1);
    acb = bdrv_aio_writev(bs, sector_num, &qiov, nb_sectors,
        bdrv_rw_em_cb, &async_ret);
2572 2573 2574 2575
    if (acb == NULL) {
        async_ret = -1;
        goto fail;
    }
B
bellard 已提交
2576 2577 2578
    while (async_ret == NOT_DONE) {
        qemu_aio_wait();
    }
2579 2580 2581

fail:
    async_context_pop();
B
bellard 已提交
2582 2583
    return async_ret;
}
B
bellard 已提交
2584 2585 2586

void bdrv_init(void)
{
2587
    module_call_init(MODULE_INIT_BLOCK);
B
bellard 已提交
2588
}
2589

2590 2591 2592 2593 2594 2595
void bdrv_init_with_whitelist(void)
{
    use_bdrv_whitelist = 1;
    bdrv_init();
}

2596 2597
void *qemu_aio_get(AIOPool *pool, BlockDriverState *bs,
                   BlockDriverCompletionFunc *cb, void *opaque)
2598 2599 2600
{
    BlockDriverAIOCB *acb;

2601 2602 2603
    if (pool->free_aiocb) {
        acb = pool->free_aiocb;
        pool->free_aiocb = acb->next;
2604
    } else {
2605 2606
        acb = qemu_mallocz(pool->aiocb_size);
        acb->pool = pool;
2607 2608 2609 2610 2611 2612 2613 2614 2615
    }
    acb->bs = bs;
    acb->cb = cb;
    acb->opaque = opaque;
    return acb;
}

void qemu_aio_release(void *p)
{
2616 2617 2618 2619
    BlockDriverAIOCB *acb = (BlockDriverAIOCB *)p;
    AIOPool *pool = acb->pool;
    acb->next = pool->free_aiocb;
    pool->free_aiocb = acb;
2620
}
B
bellard 已提交
2621 2622 2623 2624 2625 2626 2627 2628 2629 2630 2631 2632 2633 2634

/**************************************************************/
/* removable device support */

/**
 * Return TRUE if the media is present
 */
int bdrv_is_inserted(BlockDriverState *bs)
{
    BlockDriver *drv = bs->drv;
    int ret;
    if (!drv)
        return 0;
    if (!drv->bdrv_is_inserted)
2635
        return !bs->tray_open;
B
bellard 已提交
2636 2637 2638 2639 2640 2641
    ret = drv->bdrv_is_inserted(bs);
    return ret;
}

/**
 * Return TRUE if the media changed since the last call to this
2642
 * function. It is currently only used for floppy disks
B
bellard 已提交
2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661
 */
int bdrv_media_changed(BlockDriverState *bs)
{
    BlockDriver *drv = bs->drv;
    int ret;

    if (!drv || !drv->bdrv_media_changed)
        ret = -ENOTSUP;
    else
        ret = drv->bdrv_media_changed(bs);
    if (ret == -ENOTSUP)
        ret = bs->media_changed;
    bs->media_changed = 0;
    return ret;
}

/**
 * If eject_flag is TRUE, eject the media. Otherwise, close the tray
 */
2662
int bdrv_eject(BlockDriverState *bs, int eject_flag)
B
bellard 已提交
2663 2664 2665 2666
{
    BlockDriver *drv = bs->drv;
    int ret;

2667 2668 2669 2670
    if (bs->locked) {
        return -EBUSY;
    }

B
bellard 已提交
2671 2672 2673 2674 2675 2676
    if (!drv || !drv->bdrv_eject) {
        ret = -ENOTSUP;
    } else {
        ret = drv->bdrv_eject(bs, eject_flag);
    }
    if (ret == -ENOTSUP) {
2677
        ret = 0;
B
bellard 已提交
2678
    }
2679 2680 2681
    if (ret >= 0) {
        bs->tray_open = eject_flag;
    }
2682 2683

    return ret;
B
bellard 已提交
2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 2699 2700 2701 2702 2703
}

int bdrv_is_locked(BlockDriverState *bs)
{
    return bs->locked;
}

/**
 * Lock or unlock the media (if it is locked, the user won't be able
 * to eject it manually).
 */
void bdrv_set_locked(BlockDriverState *bs, int locked)
{
    BlockDriver *drv = bs->drv;

    bs->locked = locked;
    if (drv && drv->bdrv_set_locked) {
        drv->bdrv_set_locked(bs, locked);
    }
}
2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714

/* needed for generic scsi interface */

int bdrv_ioctl(BlockDriverState *bs, unsigned long int req, void *buf)
{
    BlockDriver *drv = bs->drv;

    if (drv && drv->bdrv_ioctl)
        return drv->bdrv_ioctl(bs, req, buf);
    return -ENOTSUP;
}
2715

2716 2717 2718
BlockDriverAIOCB *bdrv_aio_ioctl(BlockDriverState *bs,
        unsigned long int req, void *buf,
        BlockDriverCompletionFunc *cb, void *opaque)
2719
{
2720
    BlockDriver *drv = bs->drv;
2721

2722 2723 2724
    if (drv && drv->bdrv_aio_ioctl)
        return drv->bdrv_aio_ioctl(bs, req, buf, cb, opaque);
    return NULL;
2725
}
2726

2727 2728


2729 2730 2731 2732
void *qemu_blockalign(BlockDriverState *bs, size_t size)
{
    return qemu_memalign((bs && bs->buffer_alignment) ? bs->buffer_alignment : 512, size);
}
2733 2734 2735 2736

void bdrv_set_dirty_tracking(BlockDriverState *bs, int enable)
{
    int64_t bitmap_size;
2737

2738
    bs->dirty_count = 0;
2739
    if (enable) {
2740 2741 2742 2743
        if (!bs->dirty_bitmap) {
            bitmap_size = (bdrv_getlength(bs) >> BDRV_SECTOR_BITS) +
                    BDRV_SECTORS_PER_DIRTY_CHUNK * 8 - 1;
            bitmap_size /= BDRV_SECTORS_PER_DIRTY_CHUNK * 8;
2744

2745
            bs->dirty_bitmap = qemu_mallocz(bitmap_size);
2746
        }
2747
    } else {
2748
        if (bs->dirty_bitmap) {
2749
            qemu_free(bs->dirty_bitmap);
2750
            bs->dirty_bitmap = NULL;
2751
        }
2752 2753 2754 2755 2756
    }
}

int bdrv_get_dirty(BlockDriverState *bs, int64_t sector)
{
J
Jan Kiszka 已提交
2757
    int64_t chunk = sector / (int64_t)BDRV_SECTORS_PER_DIRTY_CHUNK;
2758

2759 2760
    if (bs->dirty_bitmap &&
        (sector << BDRV_SECTOR_BITS) < bdrv_getlength(bs)) {
2761 2762
        return !!(bs->dirty_bitmap[chunk / (sizeof(unsigned long) * 8)] &
            (1UL << (chunk % (sizeof(unsigned long) * 8))));
2763 2764 2765 2766 2767
    } else {
        return 0;
    }
}

2768 2769
void bdrv_reset_dirty(BlockDriverState *bs, int64_t cur_sector,
                      int nr_sectors)
2770 2771 2772
{
    set_dirty_bitmap(bs, cur_sector, nr_sectors, 0);
}
2773 2774 2775 2776 2777

int64_t bdrv_get_dirty_count(BlockDriverState *bs)
{
    return bs->dirty_count;
}
J
Jes Sorensen 已提交
2778

2779 2780 2781 2782 2783 2784 2785 2786 2787 2788 2789
void bdrv_set_in_use(BlockDriverState *bs, int in_use)
{
    assert(bs->in_use != in_use);
    bs->in_use = in_use;
}

int bdrv_in_use(BlockDriverState *bs)
{
    return bs->in_use;
}

J
Jes Sorensen 已提交
2790 2791 2792 2793 2794
int bdrv_img_create(const char *filename, const char *fmt,
                    const char *base_filename, const char *base_fmt,
                    char *options, uint64_t img_size, int flags)
{
    QEMUOptionParameter *param = NULL, *create_options = NULL;
2795
    QEMUOptionParameter *backing_fmt, *backing_file;
J
Jes Sorensen 已提交
2796 2797
    BlockDriverState *bs = NULL;
    BlockDriver *drv, *proto_drv;
2798
    BlockDriver *backing_drv = NULL;
J
Jes Sorensen 已提交
2799 2800 2801 2802 2803 2804
    int ret = 0;

    /* Find driver and parse its options */
    drv = bdrv_find_format(fmt);
    if (!drv) {
        error_report("Unknown file format '%s'", fmt);
2805
        ret = -EINVAL;
J
Jes Sorensen 已提交
2806 2807 2808 2809 2810 2811
        goto out;
    }

    proto_drv = bdrv_find_protocol(filename);
    if (!proto_drv) {
        error_report("Unknown protocol '%s'", filename);
2812
        ret = -EINVAL;
J
Jes Sorensen 已提交
2813 2814 2815 2816 2817 2818 2819 2820 2821 2822 2823 2824 2825 2826 2827 2828 2829 2830
        goto out;
    }

    create_options = append_option_parameters(create_options,
                                              drv->create_options);
    create_options = append_option_parameters(create_options,
                                              proto_drv->create_options);

    /* Create parameter list with default values */
    param = parse_option_parameters("", create_options, param);

    set_option_parameter_int(param, BLOCK_OPT_SIZE, img_size);

    /* Parse -o options */
    if (options) {
        param = parse_option_parameters(options, create_options, param);
        if (param == NULL) {
            error_report("Invalid options for file format '%s'.", fmt);
2831
            ret = -EINVAL;
J
Jes Sorensen 已提交
2832 2833 2834 2835 2836 2837 2838 2839 2840
            goto out;
        }
    }

    if (base_filename) {
        if (set_option_parameter(param, BLOCK_OPT_BACKING_FILE,
                                 base_filename)) {
            error_report("Backing file not supported for file format '%s'",
                         fmt);
2841
            ret = -EINVAL;
J
Jes Sorensen 已提交
2842 2843 2844 2845 2846 2847 2848 2849
            goto out;
        }
    }

    if (base_fmt) {
        if (set_option_parameter(param, BLOCK_OPT_BACKING_FMT, base_fmt)) {
            error_report("Backing file format not supported for file "
                         "format '%s'", fmt);
2850
            ret = -EINVAL;
J
Jes Sorensen 已提交
2851 2852 2853 2854
            goto out;
        }
    }

2855 2856 2857 2858 2859
    backing_file = get_option_parameter(param, BLOCK_OPT_BACKING_FILE);
    if (backing_file && backing_file->value.s) {
        if (!strcmp(filename, backing_file->value.s)) {
            error_report("Error: Trying to create an image with the "
                         "same filename as the backing file");
2860
            ret = -EINVAL;
2861 2862 2863 2864
            goto out;
        }
    }

J
Jes Sorensen 已提交
2865 2866
    backing_fmt = get_option_parameter(param, BLOCK_OPT_BACKING_FMT);
    if (backing_fmt && backing_fmt->value.s) {
2867 2868
        backing_drv = bdrv_find_format(backing_fmt->value.s);
        if (!backing_drv) {
J
Jes Sorensen 已提交
2869 2870
            error_report("Unknown backing file format '%s'",
                         backing_fmt->value.s);
2871
            ret = -EINVAL;
J
Jes Sorensen 已提交
2872 2873 2874 2875 2876 2877 2878 2879 2880 2881 2882 2883 2884
            goto out;
        }
    }

    // The size for the image must always be specified, with one exception:
    // If we are using a backing file, we can obtain the size from there
    if (get_option_parameter(param, BLOCK_OPT_SIZE)->value.n == -1) {
        if (backing_file && backing_file->value.s) {
            uint64_t size;
            char buf[32];

            bs = bdrv_new("");

2885
            ret = bdrv_open(bs, backing_file->value.s, flags, backing_drv);
J
Jes Sorensen 已提交
2886
            if (ret < 0) {
2887
                error_report("Could not open '%s'", backing_file->value.s);
J
Jes Sorensen 已提交
2888 2889 2890 2891 2892 2893 2894 2895 2896
                goto out;
            }
            bdrv_get_geometry(bs, &size);
            size *= 512;

            snprintf(buf, sizeof(buf), "%" PRId64, size);
            set_option_parameter(param, BLOCK_OPT_SIZE, buf);
        } else {
            error_report("Image creation needs a size parameter");
2897
            ret = -EINVAL;
J
Jes Sorensen 已提交
2898 2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913 2914 2915 2916 2917 2918 2919 2920 2921 2922 2923 2924 2925 2926 2927
            goto out;
        }
    }

    printf("Formatting '%s', fmt=%s ", filename, fmt);
    print_option_parameters(param);
    puts("");

    ret = bdrv_create(drv, filename, param);

    if (ret < 0) {
        if (ret == -ENOTSUP) {
            error_report("Formatting or formatting option not supported for "
                         "file format '%s'", fmt);
        } else if (ret == -EFBIG) {
            error_report("The image size is too large for file format '%s'",
                         fmt);
        } else {
            error_report("%s: error while creating %s: %s", filename, fmt,
                         strerror(-ret));
        }
    }

out:
    free_option_parameters(create_options);
    free_option_parameters(param);

    if (bs) {
        bdrv_delete(bs);
    }
2928 2929

    return ret;
J
Jes Sorensen 已提交
2930
}