block.c 88.4 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 "qjson.h"
31
#include "qemu-coroutine.h"
L
Luiz Capitulino 已提交
32
#include "qmp-commands.h"
B
bellard 已提交
33

J
Juan Quintela 已提交
34
#ifdef CONFIG_BSD
B
bellard 已提交
35 36 37
#include <sys/types.h>
#include <sys/stat.h>
#include <sys/ioctl.h>
B
Blue Swirl 已提交
38
#include <sys/queue.h>
39
#ifndef __DragonFly__
B
bellard 已提交
40 41
#include <sys/disk.h>
#endif
42
#endif
B
bellard 已提交
43

44 45 46 47
#ifdef _WIN32
#include <windows.h>
#endif

48 49
#define NOT_DONE 0x7fffffff /* used while emulated sync operation in progress */

50
static void bdrv_dev_change_media_cb(BlockDriverState *bs, bool load);
51 52
static BlockDriverAIOCB *bdrv_aio_readv_em(BlockDriverState *bs,
        int64_t sector_num, QEMUIOVector *qiov, int nb_sectors,
53
        BlockDriverCompletionFunc *cb, void *opaque);
54 55
static BlockDriverAIOCB *bdrv_aio_writev_em(BlockDriverState *bs,
        int64_t sector_num, QEMUIOVector *qiov, int nb_sectors,
56
        BlockDriverCompletionFunc *cb, void *opaque);
57 58 59 60 61 62
static int coroutine_fn bdrv_co_readv_em(BlockDriverState *bs,
                                         int64_t sector_num, int nb_sectors,
                                         QEMUIOVector *iov);
static int coroutine_fn bdrv_co_writev_em(BlockDriverState *bs,
                                         int64_t sector_num, int nb_sectors,
                                         QEMUIOVector *iov);
63 64
static int coroutine_fn bdrv_co_do_readv(BlockDriverState *bs,
    int64_t sector_num, int nb_sectors, QEMUIOVector *qiov);
65 66
static int coroutine_fn bdrv_co_do_writev(BlockDriverState *bs,
    int64_t sector_num, int nb_sectors, QEMUIOVector *qiov);
67 68 69 70 71 72
static BlockDriverAIOCB *bdrv_co_aio_rw_vector(BlockDriverState *bs,
                                               int64_t sector_num,
                                               QEMUIOVector *qiov,
                                               int nb_sectors,
                                               BlockDriverCompletionFunc *cb,
                                               void *opaque,
73
                                               bool is_write);
74
static void coroutine_fn bdrv_co_do_rw(void *opaque);
B
bellard 已提交
75

76 77
static QTAILQ_HEAD(, BlockDriverState) bdrv_states =
    QTAILQ_HEAD_INITIALIZER(bdrv_states);
78

79 80
static QLIST_HEAD(, BlockDriver) bdrv_drivers =
    QLIST_HEAD_INITIALIZER(bdrv_drivers);
B
bellard 已提交
81

82 83 84
/* The device to use for VM snapshots */
static BlockDriverState *bs_snapshots;

85 86 87
/* If non-zero, use only whitelisted block drivers */
static int use_bdrv_whitelist;

88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120
#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
121
int path_is_absolute(const char *path)
122
{
B
bellard 已提交
123
    const char *p;
B
bellard 已提交
124 125 126 127 128
#ifdef _WIN32
    /* specific case for names like: "\\.\d:" */
    if (*path == '/' || *path == '\\')
        return 1;
#endif
B
bellard 已提交
129 130 131 132 133
    p = strchr(path, ':');
    if (p)
        p++;
    else
        p = path;
134 135 136 137 138
#ifdef _WIN32
    return (*p == '/' || *p == '\\');
#else
    return (*p == '/');
#endif
139 140
}

B
bellard 已提交
141 142 143 144 145 146
/* 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)
147
{
B
bellard 已提交
148 149 150 151 152 153 154 155 156 157 158 159 160
    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;
161 162 163 164 165 166 167 168 169
        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
170 171 172 173 174 175 176 177 178 179 180 181
        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);
182 183 184
    }
}

185
void bdrv_register(BlockDriver *bdrv)
B
bellard 已提交
186
{
187 188
    /* Block drivers without coroutine functions need emulation */
    if (!bdrv->bdrv_co_readv) {
189 190 191
        bdrv->bdrv_co_readv = bdrv_co_readv_em;
        bdrv->bdrv_co_writev = bdrv_co_writev_em;

192 193 194
        /* bdrv_co_readv_em()/brdv_co_writev_em() work in terms of aio, so if
         * the block driver lacks aio we need to emulate that too.
         */
195 196 197 198 199
        if (!bdrv->bdrv_aio_readv) {
            /* add AIO emulation layer */
            bdrv->bdrv_aio_readv = bdrv_aio_readv_em;
            bdrv->bdrv_aio_writev = bdrv_aio_writev_em;
        }
B
bellard 已提交
200
    }
201

202
    QLIST_INSERT_HEAD(&bdrv_drivers, bdrv, list);
B
bellard 已提交
203
}
B
bellard 已提交
204 205 206 207

/* create a new block device (by default it is empty) */
BlockDriverState *bdrv_new(const char *device_name)
{
208
    BlockDriverState *bs;
B
bellard 已提交
209

210
    bs = g_malloc0(sizeof(BlockDriverState));
B
bellard 已提交
211
    pstrcpy(bs->device_name, sizeof(bs->device_name), device_name);
B
bellard 已提交
212
    if (device_name[0] != '\0') {
213
        QTAILQ_INSERT_TAIL(&bdrv_states, bs, list);
B
bellard 已提交
214
    }
215
    bdrv_iostatus_disable(bs);
B
bellard 已提交
216 217 218
    return bs;
}

B
bellard 已提交
219 220 221
BlockDriver *bdrv_find_format(const char *format_name)
{
    BlockDriver *drv1;
222 223
    QLIST_FOREACH(drv1, &bdrv_drivers, list) {
        if (!strcmp(drv1->format_name, format_name)) {
B
bellard 已提交
224
            return drv1;
225
        }
B
bellard 已提交
226 227 228 229
    }
    return NULL;
}

230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253
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;
}

254 255
int bdrv_create(BlockDriver *drv, const char* filename,
    QEMUOptionParameter *options)
B
bellard 已提交
256 257 258
{
    if (!drv->bdrv_create)
        return -ENOTSUP;
259 260

    return drv->bdrv_create(filename, options);
B
bellard 已提交
261 262
}

263 264 265 266
int bdrv_create_file(const char* filename, QEMUOptionParameter *options)
{
    BlockDriver *drv;

267
    drv = bdrv_find_protocol(filename);
268
    if (drv == NULL) {
269
        return -ENOENT;
270 271 272 273 274
    }

    return bdrv_create(drv, filename, options);
}

B
bellard 已提交
275
#ifdef _WIN32
276
void get_tmp_filename(char *filename, int size)
B
bellard 已提交
277
{
278
    char temp_dir[MAX_PATH];
279

280 281
    GetTempPath(MAX_PATH, temp_dir);
    GetTempFileName(temp_dir, "qem", 0, filename);
B
bellard 已提交
282 283
}
#else
284
void get_tmp_filename(char *filename, int size)
B
bellard 已提交
285
{
B
bellard 已提交
286
    int fd;
287
    const char *tmpdir;
B
bellard 已提交
288
    /* XXX: race condition possible */
A
aurel32 已提交
289 290 291 292
    tmpdir = getenv("TMPDIR");
    if (!tmpdir)
        tmpdir = "/tmp";
    snprintf(filename, size, "%s/vl.XXXXXX", tmpdir);
B
bellard 已提交
293 294 295
    fd = mkstemp(filename);
    close(fd);
}
B
bellard 已提交
296
#endif
B
bellard 已提交
297

298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319
/*
 * 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;
}

320
BlockDriver *bdrv_find_protocol(const char *filename)
B
bellard 已提交
321 322 323
{
    BlockDriver *drv1;
    char protocol[128];
324
    int len;
B
bellard 已提交
325
    const char *p;
B
bellard 已提交
326

327 328
    /* TODO Drivers without bdrv_file_open must be specified explicitly */

329 330 331 332 333 334 335 336 337 338 339 340
    /*
     * 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;
    }

341
    if (!path_has_protocol(filename)) {
342
        return bdrv_find_format("file");
343
    }
344 345
    p = strchr(filename, ':');
    assert(p != NULL);
346 347 348 349 350
    len = p - filename;
    if (len > sizeof(protocol) - 1)
        len = sizeof(protocol) - 1;
    memcpy(protocol, filename, len);
    protocol[len] = '\0';
351
    QLIST_FOREACH(drv1, &bdrv_drivers, list) {
352
        if (drv1->protocol_name &&
353
            !strcmp(drv1->protocol_name, protocol)) {
B
bellard 已提交
354
            return drv1;
355
        }
B
bellard 已提交
356 357 358 359
    }
    return NULL;
}

360
static int find_image_format(const char *filename, BlockDriver **pdrv)
361 362 363 364 365 366
{
    int ret, score, score_max;
    BlockDriver *drv1, *drv;
    uint8_t buf[2048];
    BlockDriverState *bs;

367
    ret = bdrv_file_open(&bs, filename, 0);
368 369 370 371
    if (ret < 0) {
        *pdrv = NULL;
        return ret;
    }
372

373 374
    /* Return the raw BlockDriver * to scsi-generic devices or empty drives */
    if (bs->sg || !bdrv_is_inserted(bs)) {
375
        bdrv_delete(bs);
376 377 378 379 380 381
        drv = bdrv_find_format("raw");
        if (!drv) {
            ret = -ENOENT;
        }
        *pdrv = drv;
        return ret;
382
    }
383

B
bellard 已提交
384 385 386
    ret = bdrv_pread(bs, 0, buf, sizeof(buf));
    bdrv_delete(bs);
    if (ret < 0) {
387 388
        *pdrv = NULL;
        return ret;
B
bellard 已提交
389 390
    }

B
bellard 已提交
391
    score_max = 0;
392
    drv = NULL;
393
    QLIST_FOREACH(drv1, &bdrv_drivers, list) {
B
bellard 已提交
394 395 396 397 398 399
        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
400
        }
B
bellard 已提交
401
    }
402 403 404 405 406
    if (!drv) {
        ret = -ENOENT;
    }
    *pdrv = drv;
    return ret;
B
bellard 已提交
407 408
}

409 410 411 412 413 414 415
/**
 * Set the current 'total_sectors' value
 */
static int refresh_total_sectors(BlockDriverState *bs, int64_t hint)
{
    BlockDriver *drv = bs->drv;

416 417 418 419
    /* Do not attempt drv->bdrv_getlength() on scsi-generic devices */
    if (bs->sg)
        return 0;

420 421 422 423 424 425 426 427 428 429 430 431 432
    /* 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;
}

433 434 435 436 437 438 439 440 441 442 443
/**
 * Set open flags for a given cache mode
 *
 * Return 0 on success, -1 if the cache mode was invalid.
 */
int bdrv_parse_cache_flags(const char *mode, int *flags)
{
    *flags &= ~BDRV_O_CACHE_MASK;

    if (!strcmp(mode, "off") || !strcmp(mode, "none")) {
        *flags |= BDRV_O_NOCACHE | BDRV_O_CACHE_WB;
444 445
    } else if (!strcmp(mode, "directsync")) {
        *flags |= BDRV_O_NOCACHE;
446 447 448 449 450 451 452 453 454 455 456 457 458 459
    } else if (!strcmp(mode, "writeback")) {
        *flags |= BDRV_O_CACHE_WB;
    } else if (!strcmp(mode, "unsafe")) {
        *flags |= BDRV_O_CACHE_WB;
        *flags |= BDRV_O_NO_FLUSH;
    } else if (!strcmp(mode, "writethrough")) {
        /* this is the default */
    } else {
        return -1;
    }

    return 0;
}

460 461 462 463 464 465 466 467 468 469
/*
 * 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);

S
Stefan Hajnoczi 已提交
470 471
    trace_bdrv_open_common(bs, filename, flags, drv->format_name);

472
    bs->file = NULL;
473
    bs->total_sectors = 0;
474 475
    bs->encrypted = 0;
    bs->valid_key = 0;
476
    bs->sg = 0;
477
    bs->open_flags = flags;
478
    bs->growable = 0;
479 480 481
    bs->buffer_alignment = 512;

    pstrcpy(bs->filename, sizeof(bs->filename), filename);
482
    bs->backing_file[0] = '\0';
483 484 485 486 487 488

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

    bs->drv = drv;
489
    bs->opaque = g_malloc0(drv->instance_size);
490

491
    bs->enable_write_cache = !!(flags & BDRV_O_CACHE_WB);
492 493 494 495 496 497 498 499

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

    /*
S
Stefan Weil 已提交
500
     * Snapshots should be writable.
501 502 503 504 505
     */
    if (bs->is_temporary) {
        open_flags |= BDRV_O_RDWR;
    }

506 507
    bs->keep_read_only = bs->read_only = !(open_flags & BDRV_O_RDWR);

508 509 510 511 512 513 514 515 516 517
    /* 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);
        }
    }

518 519 520 521
    if (ret < 0) {
        goto free_and_fail;
    }

522 523 524
    ret = refresh_total_sectors(bs, bs->total_sectors);
    if (ret < 0) {
        goto free_and_fail;
525
    }
526

527 528 529 530 531 532 533 534
#ifndef _WIN32
    if (bs->is_temporary) {
        unlink(filename);
    }
#endif
    return 0;

free_and_fail:
535 536 537 538
    if (bs->file) {
        bdrv_delete(bs->file);
        bs->file = NULL;
    }
539
    g_free(bs->opaque);
540 541 542 543 544
    bs->opaque = NULL;
    bs->drv = NULL;
    return ret;
}

K
Kevin Wolf 已提交
545 546 547
/*
 * Opens a file using a protocol (file, host_device, nbd, ...)
 */
B
bellard 已提交
548
int bdrv_file_open(BlockDriverState **pbs, const char *filename, int flags)
B
bellard 已提交
549
{
B
bellard 已提交
550
    BlockDriverState *bs;
551
    BlockDriver *drv;
B
bellard 已提交
552 553
    int ret;

554
    drv = bdrv_find_protocol(filename);
555 556 557 558
    if (!drv) {
        return -ENOENT;
    }

B
bellard 已提交
559
    bs = bdrv_new("");
K
Kevin Wolf 已提交
560
    ret = bdrv_open_common(bs, filename, flags, drv);
B
bellard 已提交
561 562 563
    if (ret < 0) {
        bdrv_delete(bs);
        return ret;
564
    }
565
    bs->growable = 1;
B
bellard 已提交
566 567 568 569
    *pbs = bs;
    return 0;
}

K
Kevin Wolf 已提交
570 571 572
/*
 * Opens a disk image (raw, qcow2, vmdk, ...)
 */
K
Kevin Wolf 已提交
573 574
int bdrv_open(BlockDriverState *bs, const char *filename, int flags,
              BlockDriver *drv)
B
bellard 已提交
575
{
K
Kevin Wolf 已提交
576
    int ret;
K
Kevin Wolf 已提交
577
    char tmp_filename[PATH_MAX];
B
bellard 已提交
578

B
bellard 已提交
579
    if (flags & BDRV_O_SNAPSHOT) {
B
bellard 已提交
580 581
        BlockDriverState *bs1;
        int64_t total_size;
A
aliguori 已提交
582
        int is_protocol = 0;
K
Kevin Wolf 已提交
583 584
        BlockDriver *bdrv_qcow2;
        QEMUOptionParameter *options;
K
Kevin Wolf 已提交
585
        char backing_filename[PATH_MAX];
586

B
bellard 已提交
587 588
        /* if snapshot, we create a temporary backing file and open it
           instead of opening 'filename' directly */
589

B
bellard 已提交
590 591
        /* if there is a backing file, use it */
        bs1 = bdrv_new("");
K
Kevin Wolf 已提交
592
        ret = bdrv_open(bs1, filename, 0, drv);
593
        if (ret < 0) {
B
bellard 已提交
594
            bdrv_delete(bs1);
595
            return ret;
B
bellard 已提交
596
        }
597
        total_size = bdrv_getlength(bs1) & BDRV_SECTOR_MASK;
A
aliguori 已提交
598 599 600 601

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

B
bellard 已提交
602
        bdrv_delete(bs1);
603

B
bellard 已提交
604
        get_tmp_filename(tmp_filename, sizeof(tmp_filename));
A
aliguori 已提交
605 606 607 608 609

        /* Real path is meaningless for protocols */
        if (is_protocol)
            snprintf(backing_filename, sizeof(backing_filename),
                     "%s", filename);
610 611
        else if (!realpath(filename, backing_filename))
            return -errno;
A
aliguori 已提交
612

K
Kevin Wolf 已提交
613 614 615
        bdrv_qcow2 = bdrv_find_format("qcow2");
        options = parse_option_parameters("", bdrv_qcow2->create_options, NULL);

616
        set_option_parameter_int(options, BLOCK_OPT_SIZE, total_size);
K
Kevin Wolf 已提交
617 618 619 620 621 622 623
        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);
624
        free_option_parameters(options);
625 626
        if (ret < 0) {
            return ret;
B
bellard 已提交
627
        }
K
Kevin Wolf 已提交
628

B
bellard 已提交
629
        filename = tmp_filename;
K
Kevin Wolf 已提交
630
        drv = bdrv_qcow2;
B
bellard 已提交
631 632
        bs->is_temporary = 1;
    }
B
bellard 已提交
633

K
Kevin Wolf 已提交
634
    /* Find the right image format driver */
635
    if (!drv) {
636
        ret = find_image_format(filename, &drv);
637
    }
638

639 640
    if (!drv) {
        goto unlink_and_fail;
B
bellard 已提交
641
    }
K
Kevin Wolf 已提交
642 643 644 645

    /* Open the image */
    ret = bdrv_open_common(bs, filename, flags, drv);
    if (ret < 0) {
646 647 648
        goto unlink_and_fail;
    }

K
Kevin Wolf 已提交
649 650 651 652 653 654 655
    /* 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("");
656 657 658 659 660 661 662 663 664 665

        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
666
            back_drv = bdrv_find_format(bs->backing_format);
667
        }
K
Kevin Wolf 已提交
668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686

        /* 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)) {
687
        bdrv_dev_change_media_cb(bs, true);
K
Kevin Wolf 已提交
688 689 690 691 692 693 694 695 696 697 698
    }

    return 0;

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

B
bellard 已提交
699 700
void bdrv_close(BlockDriverState *bs)
{
B
bellard 已提交
701
    if (bs->drv) {
702 703 704
        if (bs == bs_snapshots) {
            bs_snapshots = NULL;
        }
705
        if (bs->backing_hd) {
B
bellard 已提交
706
            bdrv_delete(bs->backing_hd);
707 708
            bs->backing_hd = NULL;
        }
B
bellard 已提交
709
        bs->drv->bdrv_close(bs);
710
        g_free(bs->opaque);
B
bellard 已提交
711 712 713 714
#ifdef _WIN32
        if (bs->is_temporary) {
            unlink(bs->filename);
        }
B
bellard 已提交
715
#endif
B
bellard 已提交
716 717
        bs->opaque = NULL;
        bs->drv = NULL;
B
bellard 已提交
718

719 720 721 722
        if (bs->file != NULL) {
            bdrv_close(bs->file);
        }

723
        bdrv_dev_change_media_cb(bs, false);
B
bellard 已提交
724 725 726
    }
}

727 728 729 730 731 732 733 734 735
void bdrv_close_all(void)
{
    BlockDriverState *bs;

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

736 737 738 739 740 741 742 743 744 745
/* make a BlockDriverState anonymous by removing from bdrv_state list.
   Also, NULL terminate the device_name to prevent double remove */
void bdrv_make_anon(BlockDriverState *bs)
{
    if (bs->device_name[0] != '\0') {
        QTAILQ_REMOVE(&bdrv_states, bs, list);
    }
    bs->device_name[0] = '\0';
}

B
bellard 已提交
746 747
void bdrv_delete(BlockDriverState *bs)
{
748
    assert(!bs->dev);
749

750
    /* remove from list, if necessary */
751
    bdrv_make_anon(bs);
752

B
bellard 已提交
753
    bdrv_close(bs);
754 755 756 757
    if (bs->file != NULL) {
        bdrv_delete(bs->file);
    }

758
    assert(bs != bs_snapshots);
759
    g_free(bs);
B
bellard 已提交
760 761
}

762 763
int bdrv_attach_dev(BlockDriverState *bs, void *dev)
/* TODO change to DeviceState *dev when all users are qdevified */
764
{
765
    if (bs->dev) {
766 767
        return -EBUSY;
    }
768
    bs->dev = dev;
769
    bdrv_iostatus_reset(bs);
770 771 772
    return 0;
}

773 774
/* TODO qdevified devices don't use this, remove when devices are qdevified */
void bdrv_attach_dev_nofail(BlockDriverState *bs, void *dev)
775
{
776 777 778 779 780 781 782 783 784 785
    if (bdrv_attach_dev(bs, dev) < 0) {
        abort();
    }
}

void bdrv_detach_dev(BlockDriverState *bs, void *dev)
/* TODO change to DeviceState *dev when all users are qdevified */
{
    assert(bs->dev == dev);
    bs->dev = NULL;
786 787
    bs->dev_ops = NULL;
    bs->dev_opaque = NULL;
788
    bs->buffer_alignment = 512;
789 790
}

791 792
/* TODO change to return DeviceState * when all users are qdevified */
void *bdrv_get_attached_dev(BlockDriverState *bs)
793
{
794
    return bs->dev;
795 796
}

797 798 799 800 801
void bdrv_set_dev_ops(BlockDriverState *bs, const BlockDevOps *ops,
                      void *opaque)
{
    bs->dev_ops = ops;
    bs->dev_opaque = opaque;
802 803 804
    if (bdrv_dev_has_removable_media(bs) && bs == bs_snapshots) {
        bs_snapshots = NULL;
    }
805 806
}

807
static void bdrv_dev_change_media_cb(BlockDriverState *bs, bool load)
808
{
809
    if (bs->dev_ops && bs->dev_ops->change_media_cb) {
810
        bs->dev_ops->change_media_cb(bs->dev_opaque, load);
811 812 813
    }
}

814 815 816 817 818
bool bdrv_dev_has_removable_media(BlockDriverState *bs)
{
    return !bs->dev || (bs->dev_ops && bs->dev_ops->change_media_cb);
}

P
Paolo Bonzini 已提交
819 820 821 822 823 824 825
void bdrv_dev_eject_request(BlockDriverState *bs, bool force)
{
    if (bs->dev_ops && bs->dev_ops->eject_request_cb) {
        bs->dev_ops->eject_request_cb(bs->dev_opaque, force);
    }
}

826 827 828 829 830 831 832 833
bool bdrv_dev_is_tray_open(BlockDriverState *bs)
{
    if (bs->dev_ops && bs->dev_ops->is_tray_open) {
        return bs->dev_ops->is_tray_open(bs->dev_opaque);
    }
    return false;
}

834 835 836 837
static void bdrv_dev_resize_cb(BlockDriverState *bs)
{
    if (bs->dev_ops && bs->dev_ops->resize_cb) {
        bs->dev_ops->resize_cb(bs->dev_opaque);
838 839 840
    }
}

841 842 843 844 845 846 847 848
bool bdrv_dev_is_medium_locked(BlockDriverState *bs)
{
    if (bs->dev_ops && bs->dev_ops->is_medium_locked) {
        return bs->dev_ops->is_medium_locked(bs->dev_opaque);
    }
    return false;
}

A
aliguori 已提交
849 850 851
/*
 * Run consistency checks on an image
 *
852
 * Returns 0 if the check could be completed (it doesn't mean that the image is
853
 * free of errors) or -errno when an internal error occurred. The results of the
854
 * check are stored in res.
A
aliguori 已提交
855
 */
856
int bdrv_check(BlockDriverState *bs, BdrvCheckResult *res)
A
aliguori 已提交
857 858 859 860 861
{
    if (bs->drv->bdrv_check == NULL) {
        return -ENOTSUP;
    }

862
    memset(res, 0, sizeof(*res));
863
    return bs->drv->bdrv_check(bs, res);
A
aliguori 已提交
864 865
}

866 867
#define COMMIT_BUF_SECTORS 2048

868 869 870
/* commit COW file into the raw image */
int bdrv_commit(BlockDriverState *bs)
{
B
bellard 已提交
871
    BlockDriver *drv = bs->drv;
872
    BlockDriver *backing_drv;
873 874
    int64_t sector, total_sectors;
    int n, ro, open_flags;
875
    int ret = 0, rw_ret = 0;
876
    uint8_t *buf;
877 878
    char filename[1024];
    BlockDriverState *bs_rw, *bs_ro;
879

B
bellard 已提交
880 881
    if (!drv)
        return -ENOMEDIUM;
882 883 884
    
    if (!bs->backing_hd) {
        return -ENOTSUP;
885 886
    }

887 888 889
    if (bs->backing_hd->keep_read_only) {
        return -EACCES;
    }
890 891

    backing_drv = bs->backing_hd->drv;
892 893 894 895 896 897 898 899 900
    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("");
901 902
        rw_ret = bdrv_open(bs_rw, filename, open_flags | BDRV_O_RDWR,
            backing_drv);
903 904 905 906
        if (rw_ret < 0) {
            bdrv_delete(bs_rw);
            /* try to re-open read-only */
            bs_ro = bdrv_new("");
907 908
            ret = bdrv_open(bs_ro, filename, open_flags & ~BDRV_O_RDWR,
                backing_drv);
909 910 911 912 913 914 915 916 917 918
            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
919
    }
920

J
Jan Kiszka 已提交
921
    total_sectors = bdrv_getlength(bs) >> BDRV_SECTOR_BITS;
922
    buf = g_malloc(COMMIT_BUF_SECTORS * BDRV_SECTOR_SIZE);
923 924 925 926 927 928 929 930 931 932 933 934 935

    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
936
        }
937
    }
938

939 940 941 942
    if (drv->bdrv_make_empty) {
        ret = drv->bdrv_make_empty(bs);
        bdrv_flush(bs);
    }
943

944 945 946 947 948 949
    /*
     * Make sure all data we wrote to the backing device is actually
     * stable on disk.
     */
    if (bs->backing_hd)
        bdrv_flush(bs->backing_hd);
950 951

ro_cleanup:
952
    g_free(buf);
953 954 955 956 957 958

    if (ro) {
        /* re-open as RO */
        bdrv_delete(bs->backing_hd);
        bs->backing_hd = NULL;
        bs_ro = bdrv_new("");
959 960
        ret = bdrv_open(bs_ro, filename, open_flags & ~BDRV_O_RDWR,
            backing_drv);
961 962 963 964 965 966 967 968 969 970
        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;
    }

971
    return ret;
972 973
}

974 975 976 977 978 979 980 981 982
void bdrv_commit_all(void)
{
    BlockDriverState *bs;

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

K
Kevin Wolf 已提交
983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002
/*
 * 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;
    }
}

1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015
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
1016 1017 1018 1019
    if (offset < 0)
        return -EIO;

    if ((offset > len) || (len - offset < size))
1020 1021 1022 1023 1024 1025 1026 1027
        return -EIO;

    return 0;
}

static int bdrv_check_request(BlockDriverState *bs, int64_t sector_num,
                              int nb_sectors)
{
1028 1029
    return bdrv_check_byte_request(bs, sector_num * BDRV_SECTOR_SIZE,
                                   nb_sectors * BDRV_SECTOR_SIZE);
1030 1031
}

1032 1033 1034 1035 1036 1037 1038 1039 1040 1041
typedef struct RwCo {
    BlockDriverState *bs;
    int64_t sector_num;
    int nb_sectors;
    QEMUIOVector *qiov;
    bool is_write;
    int ret;
} RwCo;

static void coroutine_fn bdrv_rw_co_entry(void *opaque)
B
bellard 已提交
1042
{
1043
    RwCo *rwco = opaque;
B
bellard 已提交
1044

1045 1046 1047 1048 1049 1050 1051 1052
    if (!rwco->is_write) {
        rwco->ret = bdrv_co_do_readv(rwco->bs, rwco->sector_num,
                                     rwco->nb_sectors, rwco->qiov);
    } else {
        rwco->ret = bdrv_co_do_writev(rwco->bs, rwco->sector_num,
                                      rwco->nb_sectors, rwco->qiov);
    }
}
1053

1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073
/*
 * Process a synchronous request using coroutines
 */
static int bdrv_rw_co(BlockDriverState *bs, int64_t sector_num, uint8_t *buf,
                      int nb_sectors, bool is_write)
{
    QEMUIOVector qiov;
    struct iovec iov = {
        .iov_base = (void *)buf,
        .iov_len = nb_sectors * BDRV_SECTOR_SIZE,
    };
    Coroutine *co;
    RwCo rwco = {
        .bs = bs,
        .sector_num = sector_num,
        .nb_sectors = nb_sectors,
        .qiov = &qiov,
        .is_write = is_write,
        .ret = NOT_DONE,
    };
1074

1075
    qemu_iovec_init_external(&qiov, &iov, 1);
1076

1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088
    if (qemu_in_coroutine()) {
        /* Fast-path if already in coroutine context */
        bdrv_rw_co_entry(&rwco);
    } else {
        co = qemu_coroutine_create(bdrv_rw_co_entry);
        qemu_coroutine_enter(co, &rwco);
        while (rwco.ret == NOT_DONE) {
            qemu_aio_wait();
        }
    }
    return rwco.ret;
}
B
bellard 已提交
1089

1090 1091 1092 1093 1094
/* return < 0 if error. See bdrv_write() for the return codes */
int bdrv_read(BlockDriverState *bs, int64_t sector_num,
              uint8_t *buf, int nb_sectors)
{
    return bdrv_rw_co(bs, sector_num, buf, nb_sectors, false);
B
bellard 已提交
1095 1096
}

1097
static void set_dirty_bitmap(BlockDriverState *bs, int64_t sector_num,
1098
                             int nb_sectors, int dirty)
1099 1100
{
    int64_t start, end;
1101
    unsigned long val, idx, bit;
1102

J
Jan Kiszka 已提交
1103
    start = sector_num / BDRV_SECTORS_PER_DIRTY_CHUNK;
1104
    end = (sector_num + nb_sectors - 1) / BDRV_SECTORS_PER_DIRTY_CHUNK;
1105 1106

    for (; start <= end; start++) {
1107 1108 1109 1110
        idx = start / (sizeof(unsigned long) * 8);
        bit = start % (sizeof(unsigned long) * 8);
        val = bs->dirty_bitmap[idx];
        if (dirty) {
1111
            if (!(val & (1UL << bit))) {
1112
                bs->dirty_count++;
1113
                val |= 1UL << bit;
1114
            }
1115
        } else {
1116
            if (val & (1UL << bit)) {
1117
                bs->dirty_count--;
1118
                val &= ~(1UL << bit);
1119
            }
1120 1121
        }
        bs->dirty_bitmap[idx] = val;
1122 1123 1124
    }
}

1125
/* Return < 0 if error. Important errors are:
B
bellard 已提交
1126 1127 1128 1129 1130
  -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
*/
1131
int bdrv_write(BlockDriverState *bs, int64_t sector_num,
B
bellard 已提交
1132 1133
               const uint8_t *buf, int nb_sectors)
{
1134
    return bdrv_rw_co(bs, sector_num, (uint8_t *)buf, nb_sectors, true);
B
bellard 已提交
1135 1136
}

1137 1138
int bdrv_pread(BlockDriverState *bs, int64_t offset,
               void *buf, int count1)
B
bellard 已提交
1139
{
J
Jan Kiszka 已提交
1140
    uint8_t tmp_buf[BDRV_SECTOR_SIZE];
B
bellard 已提交
1141 1142
    int len, nb_sectors, count;
    int64_t sector_num;
1143
    int ret;
B
bellard 已提交
1144 1145 1146

    count = count1;
    /* first read to align to sector start */
J
Jan Kiszka 已提交
1147
    len = (BDRV_SECTOR_SIZE - offset) & (BDRV_SECTOR_SIZE - 1);
B
bellard 已提交
1148 1149
    if (len > count)
        len = count;
J
Jan Kiszka 已提交
1150
    sector_num = offset >> BDRV_SECTOR_BITS;
B
bellard 已提交
1151
    if (len > 0) {
1152 1153
        if ((ret = bdrv_read(bs, sector_num, tmp_buf, 1)) < 0)
            return ret;
J
Jan Kiszka 已提交
1154
        memcpy(buf, tmp_buf + (offset & (BDRV_SECTOR_SIZE - 1)), len);
B
bellard 已提交
1155 1156 1157 1158 1159 1160 1161 1162
        count -= len;
        if (count == 0)
            return count1;
        sector_num++;
        buf += len;
    }

    /* read the sectors "in place" */
J
Jan Kiszka 已提交
1163
    nb_sectors = count >> BDRV_SECTOR_BITS;
B
bellard 已提交
1164
    if (nb_sectors > 0) {
1165 1166
        if ((ret = bdrv_read(bs, sector_num, buf, nb_sectors)) < 0)
            return ret;
B
bellard 已提交
1167
        sector_num += nb_sectors;
J
Jan Kiszka 已提交
1168
        len = nb_sectors << BDRV_SECTOR_BITS;
B
bellard 已提交
1169 1170 1171 1172 1173 1174
        buf += len;
        count -= len;
    }

    /* add data from the last sector */
    if (count > 0) {
1175 1176
        if ((ret = bdrv_read(bs, sector_num, tmp_buf, 1)) < 0)
            return ret;
B
bellard 已提交
1177 1178 1179 1180 1181
        memcpy(buf, tmp_buf, count);
    }
    return count1;
}

1182 1183
int bdrv_pwrite(BlockDriverState *bs, int64_t offset,
                const void *buf, int count1)
B
bellard 已提交
1184
{
J
Jan Kiszka 已提交
1185
    uint8_t tmp_buf[BDRV_SECTOR_SIZE];
B
bellard 已提交
1186 1187
    int len, nb_sectors, count;
    int64_t sector_num;
1188
    int ret;
B
bellard 已提交
1189 1190 1191

    count = count1;
    /* first write to align to sector start */
J
Jan Kiszka 已提交
1192
    len = (BDRV_SECTOR_SIZE - offset) & (BDRV_SECTOR_SIZE - 1);
B
bellard 已提交
1193 1194
    if (len > count)
        len = count;
J
Jan Kiszka 已提交
1195
    sector_num = offset >> BDRV_SECTOR_BITS;
B
bellard 已提交
1196
    if (len > 0) {
1197 1198
        if ((ret = bdrv_read(bs, sector_num, tmp_buf, 1)) < 0)
            return ret;
J
Jan Kiszka 已提交
1199
        memcpy(tmp_buf + (offset & (BDRV_SECTOR_SIZE - 1)), buf, len);
1200 1201
        if ((ret = bdrv_write(bs, sector_num, tmp_buf, 1)) < 0)
            return ret;
B
bellard 已提交
1202 1203 1204 1205 1206 1207 1208 1209
        count -= len;
        if (count == 0)
            return count1;
        sector_num++;
        buf += len;
    }

    /* write the sectors "in place" */
J
Jan Kiszka 已提交
1210
    nb_sectors = count >> BDRV_SECTOR_BITS;
B
bellard 已提交
1211
    if (nb_sectors > 0) {
1212 1213
        if ((ret = bdrv_write(bs, sector_num, buf, nb_sectors)) < 0)
            return ret;
B
bellard 已提交
1214
        sector_num += nb_sectors;
J
Jan Kiszka 已提交
1215
        len = nb_sectors << BDRV_SECTOR_BITS;
B
bellard 已提交
1216 1217 1218 1219 1220 1221
        buf += len;
        count -= len;
    }

    /* add data from the last sector */
    if (count > 0) {
1222 1223
        if ((ret = bdrv_read(bs, sector_num, tmp_buf, 1)) < 0)
            return ret;
B
bellard 已提交
1224
        memcpy(tmp_buf, buf, count);
1225 1226
        if ((ret = bdrv_write(bs, sector_num, tmp_buf, 1)) < 0)
            return ret;
B
bellard 已提交
1227 1228 1229 1230
    }
    return count1;
}

K
Kevin Wolf 已提交
1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246
/*
 * 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;
    }

1247 1248
    /* No flush needed for cache modes that use O_DSYNC */
    if ((bs->open_flags & BDRV_O_CACHE_WB) != 0) {
K
Kevin Wolf 已提交
1249 1250 1251 1252 1253 1254
        bdrv_flush(bs);
    }

    return 0;
}

1255 1256 1257 1258 1259
/*
 * Handle a read request in coroutine context
 */
static int coroutine_fn bdrv_co_do_readv(BlockDriverState *bs,
    int64_t sector_num, int nb_sectors, QEMUIOVector *qiov)
K
Kevin Wolf 已提交
1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272
{
    BlockDriver *drv = bs->drv;

    if (!drv) {
        return -ENOMEDIUM;
    }
    if (bdrv_check_request(bs, sector_num, nb_sectors)) {
        return -EIO;
    }

    return drv->bdrv_co_readv(bs, sector_num, nb_sectors, qiov);
}

1273
int coroutine_fn bdrv_co_readv(BlockDriverState *bs, int64_t sector_num,
K
Kevin Wolf 已提交
1274 1275
    int nb_sectors, QEMUIOVector *qiov)
{
1276
    trace_bdrv_co_readv(bs, sector_num, nb_sectors);
K
Kevin Wolf 已提交
1277

1278 1279 1280 1281 1282 1283 1284 1285 1286 1287
    return bdrv_co_do_readv(bs, sector_num, nb_sectors, qiov);
}

/*
 * Handle a write request in coroutine context
 */
static int coroutine_fn bdrv_co_do_writev(BlockDriverState *bs,
    int64_t sector_num, int nb_sectors, QEMUIOVector *qiov)
{
    BlockDriver *drv = bs->drv;
1288
    int ret;
K
Kevin Wolf 已提交
1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299

    if (!bs->drv) {
        return -ENOMEDIUM;
    }
    if (bs->read_only) {
        return -EACCES;
    }
    if (bdrv_check_request(bs, sector_num, nb_sectors)) {
        return -EIO;
    }

1300 1301
    ret = drv->bdrv_co_writev(bs, sector_num, nb_sectors, qiov);

K
Kevin Wolf 已提交
1302 1303 1304 1305 1306 1307 1308 1309
    if (bs->dirty_bitmap) {
        set_dirty_bitmap(bs, sector_num, nb_sectors, 1);
    }

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

1310
    return ret;
K
Kevin Wolf 已提交
1311 1312
}

1313 1314 1315 1316 1317 1318 1319 1320
int coroutine_fn bdrv_co_writev(BlockDriverState *bs, int64_t sector_num,
    int nb_sectors, QEMUIOVector *qiov)
{
    trace_bdrv_co_writev(bs, sector_num, nb_sectors);

    return bdrv_co_do_writev(bs, sector_num, nb_sectors, qiov);
}

B
bellard 已提交
1321 1322 1323 1324 1325 1326
/**
 * Truncate file to 'offset' bytes (needed only for file protocols)
 */
int bdrv_truncate(BlockDriverState *bs, int64_t offset)
{
    BlockDriver *drv = bs->drv;
1327
    int ret;
B
bellard 已提交
1328
    if (!drv)
B
bellard 已提交
1329
        return -ENOMEDIUM;
B
bellard 已提交
1330 1331
    if (!drv->bdrv_truncate)
        return -ENOTSUP;
1332 1333
    if (bs->read_only)
        return -EACCES;
M
Marcelo Tosatti 已提交
1334 1335
    if (bdrv_in_use(bs))
        return -EBUSY;
1336 1337 1338
    ret = drv->bdrv_truncate(bs, offset);
    if (ret == 0) {
        ret = refresh_total_sectors(bs, offset >> BDRV_SECTOR_BITS);
1339
        bdrv_dev_resize_cb(bs);
1340 1341
    }
    return ret;
B
bellard 已提交
1342 1343
}

1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362
/**
 * Length of a allocated file in bytes. Sparse files are counted by actual
 * allocated space. Return < 0 if error or unknown.
 */
int64_t bdrv_get_allocated_file_size(BlockDriverState *bs)
{
    BlockDriver *drv = bs->drv;
    if (!drv) {
        return -ENOMEDIUM;
    }
    if (drv->bdrv_get_allocated_file_size) {
        return drv->bdrv_get_allocated_file_size(bs);
    }
    if (bs->file) {
        return bdrv_get_allocated_file_size(bs->file);
    }
    return -ENOTSUP;
}

B
bellard 已提交
1363 1364 1365 1366 1367 1368 1369
/**
 * 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
1370
        return -ENOMEDIUM;
1371

1372
    if (bs->growable || bdrv_dev_has_removable_media(bs)) {
1373 1374 1375
        if (drv->bdrv_getlength) {
            return drv->bdrv_getlength(bs);
        }
B
bellard 已提交
1376
    }
1377
    return bs->total_sectors * BDRV_SECTOR_SIZE;
B
bellard 已提交
1378 1379
}

B
bellard 已提交
1380
/* return 0 as number of sectors if no device present or error */
1381
void bdrv_get_geometry(BlockDriverState *bs, uint64_t *nb_sectors_ptr)
B
bellard 已提交
1382
{
B
bellard 已提交
1383 1384 1385 1386 1387
    int64_t length;
    length = bdrv_getlength(bs);
    if (length < 0)
        length = 0;
    else
J
Jan Kiszka 已提交
1388
        length = length >> BDRV_SECTOR_BITS;
B
bellard 已提交
1389
    *nb_sectors_ptr = length;
B
bellard 已提交
1390
}
B
bellard 已提交
1391

1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402
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 */
1403
} QEMU_PACKED;
1404 1405 1406 1407 1408

/* 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)
{
1409
    uint8_t buf[BDRV_SECTOR_SIZE];
1410 1411 1412
    int ret, i, heads, sectors, cylinders;
    struct partition *p;
    uint32_t nr_sects;
B
blueswir1 已提交
1413
    uint64_t nb_sectors;
1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452

    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
1453
    uint64_t nb_sectors;
1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507

    /* 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);
    }
}

1508
void bdrv_set_geometry_hint(BlockDriverState *bs,
B
bellard 已提交
1509 1510 1511 1512 1513 1514 1515
                            int cyls, int heads, int secs)
{
    bs->cyls = cyls;
    bs->heads = heads;
    bs->secs = secs;
}

B
bellard 已提交
1516 1517 1518 1519 1520
void bdrv_set_translation_hint(BlockDriverState *bs, int translation)
{
    bs->translation = translation;
}

1521
void bdrv_get_geometry_hint(BlockDriverState *bs,
B
bellard 已提交
1522 1523 1524 1525 1526 1527 1528
                            int *pcyls, int *pheads, int *psecs)
{
    *pcyls = bs->cyls;
    *pheads = bs->heads;
    *psecs = bs->secs;
}

1529 1530 1531 1532 1533 1534 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 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631
/* Recognize floppy formats */
typedef struct FDFormat {
    FDriveType drive;
    uint8_t last_sect;
    uint8_t max_track;
    uint8_t max_head;
} FDFormat;

static const FDFormat fd_formats[] = {
    /* First entry is default format */
    /* 1.44 MB 3"1/2 floppy disks */
    { FDRIVE_DRV_144, 18, 80, 1, },
    { FDRIVE_DRV_144, 20, 80, 1, },
    { FDRIVE_DRV_144, 21, 80, 1, },
    { FDRIVE_DRV_144, 21, 82, 1, },
    { FDRIVE_DRV_144, 21, 83, 1, },
    { FDRIVE_DRV_144, 22, 80, 1, },
    { FDRIVE_DRV_144, 23, 80, 1, },
    { FDRIVE_DRV_144, 24, 80, 1, },
    /* 2.88 MB 3"1/2 floppy disks */
    { FDRIVE_DRV_288, 36, 80, 1, },
    { FDRIVE_DRV_288, 39, 80, 1, },
    { FDRIVE_DRV_288, 40, 80, 1, },
    { FDRIVE_DRV_288, 44, 80, 1, },
    { FDRIVE_DRV_288, 48, 80, 1, },
    /* 720 kB 3"1/2 floppy disks */
    { FDRIVE_DRV_144,  9, 80, 1, },
    { FDRIVE_DRV_144, 10, 80, 1, },
    { FDRIVE_DRV_144, 10, 82, 1, },
    { FDRIVE_DRV_144, 10, 83, 1, },
    { FDRIVE_DRV_144, 13, 80, 1, },
    { FDRIVE_DRV_144, 14, 80, 1, },
    /* 1.2 MB 5"1/4 floppy disks */
    { FDRIVE_DRV_120, 15, 80, 1, },
    { FDRIVE_DRV_120, 18, 80, 1, },
    { FDRIVE_DRV_120, 18, 82, 1, },
    { FDRIVE_DRV_120, 18, 83, 1, },
    { FDRIVE_DRV_120, 20, 80, 1, },
    /* 720 kB 5"1/4 floppy disks */
    { FDRIVE_DRV_120,  9, 80, 1, },
    { FDRIVE_DRV_120, 11, 80, 1, },
    /* 360 kB 5"1/4 floppy disks */
    { FDRIVE_DRV_120,  9, 40, 1, },
    { FDRIVE_DRV_120,  9, 40, 0, },
    { FDRIVE_DRV_120, 10, 41, 1, },
    { FDRIVE_DRV_120, 10, 42, 1, },
    /* 320 kB 5"1/4 floppy disks */
    { FDRIVE_DRV_120,  8, 40, 1, },
    { FDRIVE_DRV_120,  8, 40, 0, },
    /* 360 kB must match 5"1/4 better than 3"1/2... */
    { FDRIVE_DRV_144,  9, 80, 0, },
    /* end */
    { FDRIVE_DRV_NONE, -1, -1, 0, },
};

void bdrv_get_floppy_geometry_hint(BlockDriverState *bs, int *nb_heads,
                                   int *max_track, int *last_sect,
                                   FDriveType drive_in, FDriveType *drive)
{
    const FDFormat *parse;
    uint64_t nb_sectors, size;
    int i, first_match, match;

    bdrv_get_geometry_hint(bs, nb_heads, max_track, last_sect);
    if (*nb_heads != 0 && *max_track != 0 && *last_sect != 0) {
        /* User defined disk */
    } else {
        bdrv_get_geometry(bs, &nb_sectors);
        match = -1;
        first_match = -1;
        for (i = 0; ; i++) {
            parse = &fd_formats[i];
            if (parse->drive == FDRIVE_DRV_NONE) {
                break;
            }
            if (drive_in == parse->drive ||
                drive_in == FDRIVE_DRV_NONE) {
                size = (parse->max_head + 1) * parse->max_track *
                    parse->last_sect;
                if (nb_sectors == size) {
                    match = i;
                    break;
                }
                if (first_match == -1) {
                    first_match = i;
                }
            }
        }
        if (match == -1) {
            if (first_match == -1) {
                match = 1;
            } else {
                match = first_match;
            }
            parse = &fd_formats[match];
        }
        *nb_heads = parse->max_head + 1;
        *max_track = parse->max_track;
        *last_sect = parse->last_sect;
        *drive = parse->drive;
    }
}

B
bellard 已提交
1632 1633 1634 1635 1636
int bdrv_get_translation_hint(BlockDriverState *bs)
{
    return bs->translation;
}

1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648
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;
}

B
bellard 已提交
1649 1650 1651 1652 1653
int bdrv_is_read_only(BlockDriverState *bs)
{
    return bs->read_only;
}

1654 1655 1656 1657 1658
int bdrv_is_sg(BlockDriverState *bs)
{
    return bs->sg;
}

1659 1660 1661 1662 1663
int bdrv_enable_write_cache(BlockDriverState *bs)
{
    return bs->enable_write_cache;
}

B
bellard 已提交
1664 1665 1666 1667 1668 1669 1670
int bdrv_is_encrypted(BlockDriverState *bs)
{
    if (bs->backing_hd && bs->backing_hd->encrypted)
        return 1;
    return bs->encrypted;
}

1671 1672 1673 1674 1675 1676 1677 1678 1679
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
1680 1681 1682 1683 1684 1685 1686 1687 1688 1689
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;
    }
1690 1691 1692 1693 1694
    if (!bs->encrypted) {
        return -EINVAL;
    } else if (!bs->drv || !bs->drv->bdrv_set_key) {
        return -ENOMEDIUM;
    }
1695
    ret = bs->drv->bdrv_set_key(bs, key);
1696 1697 1698 1699 1700
    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 */
1701
        bdrv_dev_change_media_cb(bs, true);
1702
    }
1703
    return ret;
B
bellard 已提交
1704 1705 1706 1707
}

void bdrv_get_format(BlockDriverState *bs, char *buf, int buf_size)
{
B
bellard 已提交
1708
    if (!bs->drv) {
B
bellard 已提交
1709 1710 1711 1712 1713 1714
        buf[0] = '\0';
    } else {
        pstrcpy(buf, buf_size, bs->drv->format_name);
    }
}

1715
void bdrv_iterate_format(void (*it)(void *opaque, const char *name),
B
bellard 已提交
1716 1717 1718 1719
                         void *opaque)
{
    BlockDriver *drv;

1720
    QLIST_FOREACH(drv, &bdrv_drivers, list) {
B
bellard 已提交
1721 1722 1723 1724
        it(opaque, drv->format_name);
    }
}

B
bellard 已提交
1725 1726 1727 1728
BlockDriverState *bdrv_find(const char *name)
{
    BlockDriverState *bs;

1729 1730
    QTAILQ_FOREACH(bs, &bdrv_states, list) {
        if (!strcmp(name, bs->device_name)) {
B
bellard 已提交
1731
            return bs;
1732
        }
B
bellard 已提交
1733 1734 1735 1736
    }
    return NULL;
}

M
Markus Armbruster 已提交
1737 1738 1739 1740 1741 1742 1743 1744
BlockDriverState *bdrv_next(BlockDriverState *bs)
{
    if (!bs) {
        return QTAILQ_FIRST(&bdrv_states);
    }
    return QTAILQ_NEXT(bs, list);
}

1745
void bdrv_iterate(void (*it)(void *opaque, BlockDriverState *bs), void *opaque)
B
bellard 已提交
1746 1747 1748
{
    BlockDriverState *bs;

1749
    QTAILQ_FOREACH(bs, &bdrv_states, list) {
1750
        it(opaque, bs);
B
bellard 已提交
1751 1752 1753
    }
}

B
bellard 已提交
1754 1755 1756 1757 1758
const char *bdrv_get_device_name(BlockDriverState *bs)
{
    return bs->device_name;
}

A
aliguori 已提交
1759 1760 1761 1762
void bdrv_flush_all(void)
{
    BlockDriverState *bs;

1763
    QTAILQ_FOREACH(bs, &bdrv_states, list) {
1764
        if (!bdrv_is_read_only(bs) && bdrv_is_inserted(bs)) {
A
aliguori 已提交
1765
            bdrv_flush(bs);
1766 1767
        }
    }
A
aliguori 已提交
1768 1769
}

K
Kevin Wolf 已提交
1770 1771 1772 1773
int bdrv_has_zero_init(BlockDriverState *bs)
{
    assert(bs->drv);

K
Kevin Wolf 已提交
1774 1775
    if (bs->drv->bdrv_has_zero_init) {
        return bs->drv->bdrv_has_zero_init(bs);
K
Kevin Wolf 已提交
1776 1777 1778 1779 1780
    }

    return 1;
}

1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807
/*
 * 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);
}

1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836
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);
}

L
Luiz Capitulino 已提交
1837
BlockInfoList *qmp_query_block(Error **errp)
B
bellard 已提交
1838
{
L
Luiz Capitulino 已提交
1839
    BlockInfoList *head = NULL, *cur_item = NULL;
B
bellard 已提交
1840 1841
    BlockDriverState *bs;

1842
    QTAILQ_FOREACH(bs, &bdrv_states, list) {
L
Luiz Capitulino 已提交
1843
        BlockInfoList *info = g_malloc0(sizeof(*info));
1844

L
Luiz Capitulino 已提交
1845 1846 1847 1848 1849
        info->value = g_malloc0(sizeof(*info->value));
        info->value->device = g_strdup(bs->device_name);
        info->value->type = g_strdup("unknown");
        info->value->locked = bdrv_dev_is_medium_locked(bs);
        info->value->removable = bdrv_dev_has_removable_media(bs);
1850

1851
        if (bdrv_dev_has_removable_media(bs)) {
L
Luiz Capitulino 已提交
1852 1853
            info->value->has_tray_open = true;
            info->value->tray_open = bdrv_dev_is_tray_open(bs);
1854
        }
1855 1856

        if (bdrv_iostatus_is_enabled(bs)) {
L
Luiz Capitulino 已提交
1857 1858
            info->value->has_io_status = true;
            info->value->io_status = bs->iostatus;
1859 1860
        }

B
bellard 已提交
1861
        if (bs->drv) {
L
Luiz Capitulino 已提交
1862 1863 1864 1865 1866 1867 1868 1869 1870
            info->value->has_inserted = true;
            info->value->inserted = g_malloc0(sizeof(*info->value->inserted));
            info->value->inserted->file = g_strdup(bs->filename);
            info->value->inserted->ro = bs->read_only;
            info->value->inserted->drv = g_strdup(bs->drv->format_name);
            info->value->inserted->encrypted = bs->encrypted;
            if (bs->backing_file[0]) {
                info->value->inserted->has_backing_file = true;
                info->value->inserted->backing_file = g_strdup(bs->backing_file);
A
aliguori 已提交
1871
            }
L
Luiz Capitulino 已提交
1872
        }
1873

L
Luiz Capitulino 已提交
1874 1875 1876 1877 1878 1879
        /* XXX: waiting for the qapi to support GSList */
        if (!cur_item) {
            head = cur_item = info;
        } else {
            cur_item->next = info;
            cur_item = info;
B
bellard 已提交
1880 1881
        }
    }
1882

L
Luiz Capitulino 已提交
1883
    return head;
B
bellard 已提交
1884
}
1885

L
Luiz Capitulino 已提交
1886 1887 1888 1889 1890 1891 1892 1893 1894 1895
/* Consider exposing this as a full fledged QMP command */
static BlockStats *qmp_query_blockstat(const BlockDriverState *bs, Error **errp)
{
    BlockStats *s;

    s = g_malloc0(sizeof(*s));

    if (bs->device_name[0]) {
        s->has_device = true;
        s->device = g_strdup(bs->device_name);
1896 1897
    }

L
Luiz Capitulino 已提交
1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908
    s->stats = g_malloc0(sizeof(*s->stats));
    s->stats->rd_bytes = bs->nr_bytes[BDRV_ACCT_READ];
    s->stats->wr_bytes = bs->nr_bytes[BDRV_ACCT_WRITE];
    s->stats->rd_operations = bs->nr_ops[BDRV_ACCT_READ];
    s->stats->wr_operations = bs->nr_ops[BDRV_ACCT_WRITE];
    s->stats->wr_highest_offset = bs->wr_highest_sector * BDRV_SECTOR_SIZE;
    s->stats->flush_operations = bs->nr_ops[BDRV_ACCT_FLUSH];
    s->stats->wr_total_time_ns = bs->total_time_ns[BDRV_ACCT_WRITE];
    s->stats->rd_total_time_ns = bs->total_time_ns[BDRV_ACCT_READ];
    s->stats->flush_total_time_ns = bs->total_time_ns[BDRV_ACCT_FLUSH];

1909
    if (bs->file) {
L
Luiz Capitulino 已提交
1910 1911
        s->has_parent = true;
        s->parent = qmp_query_blockstat(bs->file, NULL);
1912 1913
    }

L
Luiz Capitulino 已提交
1914
    return s;
1915 1916
}

L
Luiz Capitulino 已提交
1917
BlockStatsList *qmp_query_blockstats(Error **errp)
1918
{
L
Luiz Capitulino 已提交
1919
    BlockStatsList *head = NULL, *cur_item = NULL;
1920 1921
    BlockDriverState *bs;

1922
    QTAILQ_FOREACH(bs, &bdrv_states, list) {
L
Luiz Capitulino 已提交
1923 1924 1925 1926 1927 1928 1929 1930 1931 1932
        BlockStatsList *info = g_malloc0(sizeof(*info));
        info->value = qmp_query_blockstat(bs, NULL);

        /* XXX: waiting for the qapi to support GSList */
        if (!cur_item) {
            head = cur_item = info;
        } else {
            cur_item->next = info;
            cur_item = info;
        }
1933
    }
1934

L
Luiz Capitulino 已提交
1935
    return head;
1936
}
B
bellard 已提交
1937

1938 1939 1940 1941 1942 1943 1944 1945 1946 1947
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;
}

1948
void bdrv_get_backing_filename(BlockDriverState *bs,
B
bellard 已提交
1949 1950
                               char *filename, int filename_size)
{
K
Kevin Wolf 已提交
1951
    pstrcpy(filename, filename_size, bs->backing_file);
B
bellard 已提交
1952 1953
}

1954
int bdrv_write_compressed(BlockDriverState *bs, int64_t sector_num,
B
bellard 已提交
1955 1956 1957 1958
                          const uint8_t *buf, int nb_sectors)
{
    BlockDriver *drv = bs->drv;
    if (!drv)
B
bellard 已提交
1959
        return -ENOMEDIUM;
B
bellard 已提交
1960 1961
    if (!drv->bdrv_write_compressed)
        return -ENOTSUP;
K
Kevin Wolf 已提交
1962 1963
    if (bdrv_check_request(bs, sector_num, nb_sectors))
        return -EIO;
1964

1965
    if (bs->dirty_bitmap) {
1966 1967
        set_dirty_bitmap(bs, sector_num, nb_sectors, 1);
    }
1968

B
bellard 已提交
1969 1970
    return drv->bdrv_write_compressed(bs, sector_num, buf, nb_sectors);
}
1971

B
bellard 已提交
1972 1973 1974 1975
int bdrv_get_info(BlockDriverState *bs, BlockDriverInfo *bdi)
{
    BlockDriver *drv = bs->drv;
    if (!drv)
B
bellard 已提交
1976
        return -ENOMEDIUM;
B
bellard 已提交
1977 1978 1979 1980 1981 1982
    if (!drv->bdrv_get_info)
        return -ENOTSUP;
    memset(bdi, 0, sizeof(*bdi));
    return drv->bdrv_get_info(bs, bdi);
}

1983 1984
int bdrv_save_vmstate(BlockDriverState *bs, const uint8_t *buf,
                      int64_t pos, int size)
1985 1986 1987 1988
{
    BlockDriver *drv = bs->drv;
    if (!drv)
        return -ENOMEDIUM;
1989 1990 1991 1992 1993
    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;
1994 1995
}

1996 1997
int bdrv_load_vmstate(BlockDriverState *bs, uint8_t *buf,
                      int64_t pos, int size)
1998 1999 2000 2001
{
    BlockDriver *drv = bs->drv;
    if (!drv)
        return -ENOMEDIUM;
2002 2003 2004 2005 2006
    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;
2007 2008
}

K
Kevin Wolf 已提交
2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020
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
2021 2022 2023
/**************************************************************/
/* handling of snapshots */

2024 2025 2026
int bdrv_can_snapshot(BlockDriverState *bs)
{
    BlockDriver *drv = bs->drv;
2027
    if (!drv || !bdrv_is_inserted(bs) || bdrv_is_read_only(bs)) {
2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040
        return 0;
    }

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

    return 1;
}

2041 2042 2043 2044 2045
int bdrv_is_snapshot(BlockDriverState *bs)
{
    return !!(bs->open_flags & BDRV_O_SNAPSHOT);
}

2046 2047 2048 2049
BlockDriverState *bdrv_snapshots(void)
{
    BlockDriverState *bs;

2050
    if (bs_snapshots) {
2051
        return bs_snapshots;
2052
    }
2053 2054 2055 2056

    bs = NULL;
    while ((bs = bdrv_next(bs))) {
        if (bdrv_can_snapshot(bs)) {
2057 2058
            bs_snapshots = bs;
            return bs;
2059 2060 2061 2062 2063
        }
    }
    return NULL;
}

2064
int bdrv_snapshot_create(BlockDriverState *bs,
B
bellard 已提交
2065 2066 2067 2068
                         QEMUSnapshotInfo *sn_info)
{
    BlockDriver *drv = bs->drv;
    if (!drv)
B
bellard 已提交
2069
        return -ENOMEDIUM;
2070 2071 2072 2073 2074
    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
2075 2076
}

2077
int bdrv_snapshot_goto(BlockDriverState *bs,
B
bellard 已提交
2078 2079 2080
                       const char *snapshot_id)
{
    BlockDriver *drv = bs->drv;
2081 2082
    int ret, open_ret;

B
bellard 已提交
2083
    if (!drv)
B
bellard 已提交
2084
        return -ENOMEDIUM;
2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100
    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
2101 2102 2103 2104 2105 2106
}

int bdrv_snapshot_delete(BlockDriverState *bs, const char *snapshot_id)
{
    BlockDriver *drv = bs->drv;
    if (!drv)
B
bellard 已提交
2107
        return -ENOMEDIUM;
2108 2109 2110 2111 2112
    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
2113 2114
}

2115
int bdrv_snapshot_list(BlockDriverState *bs,
B
bellard 已提交
2116 2117 2118 2119
                       QEMUSnapshotInfo **psn_info)
{
    BlockDriver *drv = bs->drv;
    if (!drv)
B
bellard 已提交
2120
        return -ENOMEDIUM;
2121 2122 2123 2124 2125
    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
2126 2127
}

E
edison 已提交
2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143
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
2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157
#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)) {
2158
                snprintf(buf, buf_size, "%0.1f%c",
B
bellard 已提交
2159 2160 2161 2162
                         (double)size / base,
                         suffixes[i]);
                break;
            } else if (size < (1000 * base) || i == (NB_SUFFIXES - 1)) {
2163
                snprintf(buf, buf_size, "%" PRId64 "%c",
B
bellard 已提交
2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176
                         ((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];
2177 2178 2179
#ifdef _WIN32
    struct tm *ptm;
#else
B
bellard 已提交
2180
    struct tm tm;
2181
#endif
B
bellard 已提交
2182 2183 2184 2185
    time_t ti;
    int64_t secs;

    if (!sn) {
2186 2187
        snprintf(buf, buf_size,
                 "%-10s%-20s%7s%20s%15s",
B
bellard 已提交
2188 2189 2190
                 "ID", "TAG", "VM SIZE", "DATE", "VM CLOCK");
    } else {
        ti = sn->date_sec;
2191 2192 2193 2194 2195
#ifdef _WIN32
        ptm = localtime(&ti);
        strftime(date_buf, sizeof(date_buf),
                 "%Y-%m-%d %H:%M:%S", ptm);
#else
B
bellard 已提交
2196 2197 2198
        localtime_r(&ti, &tm);
        strftime(date_buf, sizeof(date_buf),
                 "%Y-%m-%d %H:%M:%S", &tm);
2199
#endif
B
bellard 已提交
2200 2201 2202 2203 2204
        secs = sn->vm_clock_nsec / 1000000000;
        snprintf(clock_buf, sizeof(clock_buf),
                 "%02d:%02d:%02d.%03d",
                 (int)(secs / 3600),
                 (int)((secs / 60) % 60),
2205
                 (int)(secs % 60),
B
bellard 已提交
2206 2207
                 (int)((sn->vm_clock_nsec / 1000000) % 1000));
        snprintf(buf, buf_size,
2208
                 "%-10s%-20s%7s%20s%15s",
B
bellard 已提交
2209 2210 2211 2212 2213 2214 2215 2216
                 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
2217
/**************************************************************/
B
bellard 已提交
2218
/* async I/Os */
B
bellard 已提交
2219

2220
BlockDriverAIOCB *bdrv_aio_readv(BlockDriverState *bs, int64_t sector_num,
2221
                                 QEMUIOVector *qiov, int nb_sectors,
2222
                                 BlockDriverCompletionFunc *cb, void *opaque)
B
bellard 已提交
2223
{
2224 2225
    trace_bdrv_aio_readv(bs, sector_num, nb_sectors, opaque);

2226
    return bdrv_co_aio_rw_vector(bs, sector_num, qiov, nb_sectors,
2227
                                 cb, opaque, false);
B
bellard 已提交
2228 2229
}

2230 2231 2232
BlockDriverAIOCB *bdrv_aio_writev(BlockDriverState *bs, int64_t sector_num,
                                  QEMUIOVector *qiov, int nb_sectors,
                                  BlockDriverCompletionFunc *cb, void *opaque)
B
bellard 已提交
2233
{
2234 2235
    trace_bdrv_aio_writev(bs, sector_num, nb_sectors, opaque);

2236
    return bdrv_co_aio_rw_vector(bs, sector_num, qiov, nb_sectors,
2237
                                 cb, opaque, true);
B
bellard 已提交
2238 2239
}

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

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);
2259 2260 2261
        if (mcb->callbacks[i].free_qiov) {
            qemu_iovec_destroy(mcb->callbacks[i].free_qiov);
        }
2262
        g_free(mcb->callbacks[i].free_qiov);
2263
        qemu_vfree(mcb->callbacks[i].free_buf);
K
Kevin Wolf 已提交
2264 2265 2266 2267 2268 2269 2270
    }
}

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

2271 2272
    trace_multiwrite_cb(mcb, ret);

2273
    if (ret < 0 && !mcb->error) {
K
Kevin Wolf 已提交
2274 2275 2276 2277 2278
        mcb->error = ret;
    }

    mcb->num_requests--;
    if (mcb->num_requests == 0) {
2279
        multiwrite_user_cb(mcb);
2280
        g_free(mcb);
K
Kevin Wolf 已提交
2281 2282 2283 2284 2285
    }
}

static int multiwrite_req_compare(const void *a, const void *b)
{
2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298
    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
2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333
}

/*
 * 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]);
        }

2334 2335 2336 2337
        if (reqs[outidx].qiov->niov + reqs[i].qiov->niov + 1 > IOV_MAX) {
            merge = 0;
        }

K
Kevin Wolf 已提交
2338 2339
        if (merge) {
            size_t size;
2340
            QEMUIOVector *qiov = g_malloc0(sizeof(*qiov));
K
Kevin Wolf 已提交
2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360
            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);

2361
            reqs[outidx].nb_sectors = qiov->size >> 9;
K
Kevin Wolf 已提交
2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395
            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;

2396 2397 2398 2399 2400 2401 2402 2403
    /* don't submit writes if we don't have a medium */
    if (bs->drv == NULL) {
        for (i = 0; i < num_reqs; i++) {
            reqs[i].error = -ENOMEDIUM;
        }
        return -1;
    }

K
Kevin Wolf 已提交
2404 2405 2406 2407 2408
    if (num_reqs == 0) {
        return 0;
    }

    // Create MultiwriteCB structure
2409
    mcb = g_malloc0(sizeof(*mcb) + num_reqs * sizeof(*mcb->callbacks));
K
Kevin Wolf 已提交
2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420
    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);

2421 2422
    trace_bdrv_aio_multiwrite(mcb, mcb->num_callbacks, num_reqs);

2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443
    /*
     * 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;

2444
    // Run the aio requests
K
Kevin Wolf 已提交
2445
    for (i = 0; i < num_reqs; i++) {
2446
        mcb->num_requests++;
K
Kevin Wolf 已提交
2447 2448 2449 2450 2451 2452 2453
        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.
2454
            if (i == 0) {
2455
                trace_bdrv_aio_multiwrite_earlyfail(mcb);
K
Kevin Wolf 已提交
2456 2457
                goto fail;
            } else {
2458
                trace_bdrv_aio_multiwrite_latefail(mcb, i);
2459
                multiwrite_cb(mcb, -EIO);
K
Kevin Wolf 已提交
2460 2461 2462 2463 2464
                break;
            }
        }
    }

2465 2466 2467
    /* Complete the dummy request */
    multiwrite_cb(mcb, 0);

K
Kevin Wolf 已提交
2468 2469 2470
    return 0;

fail:
2471 2472 2473
    for (i = 0; i < mcb->num_callbacks; i++) {
        reqs[i].error = -EIO;
    }
2474
    g_free(mcb);
K
Kevin Wolf 已提交
2475 2476 2477
    return -1;
}

B
bellard 已提交
2478 2479
void bdrv_aio_cancel(BlockDriverAIOCB *acb)
{
2480
    acb->pool->cancel(acb);
B
bellard 已提交
2481 2482
}

2483

B
bellard 已提交
2484 2485 2486
/**************************************************************/
/* async block device emulation */

2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498
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)
{
2499 2500
    BlockDriverAIOCBSync *acb =
        container_of(blockacb, BlockDriverAIOCBSync, common);
2501
    qemu_bh_delete(acb->bh);
A
Avi Kivity 已提交
2502
    acb->bh = NULL;
2503 2504 2505 2506 2507 2508 2509 2510
    qemu_aio_release(acb);
}

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

2511
static void bdrv_aio_bh_cb(void *opaque)
B
bellard 已提交
2512
{
2513
    BlockDriverAIOCBSync *acb = opaque;
2514 2515 2516

    if (!acb->is_write)
        qemu_iovec_from_buffer(acb->qiov, acb->bounce, acb->qiov->size);
2517
    qemu_vfree(acb->bounce);
2518
    acb->common.cb(acb->common.opaque, acb->ret);
2519
    qemu_bh_delete(acb->bh);
A
Avi Kivity 已提交
2520
    acb->bh = NULL;
2521
    qemu_aio_release(acb);
B
bellard 已提交
2522
}
2523

2524 2525 2526 2527 2528 2529 2530 2531
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
2532
{
2533 2534
    BlockDriverAIOCBSync *acb;

2535
    acb = qemu_aio_get(&bdrv_em_aio_pool, bs, cb, opaque);
2536 2537
    acb->is_write = is_write;
    acb->qiov = qiov;
2538
    acb->bounce = qemu_blockalign(bs, qiov->size);
2539

2540 2541
    if (!acb->bh)
        acb->bh = qemu_bh_new(bdrv_aio_bh_cb, acb);
2542 2543 2544

    if (is_write) {
        qemu_iovec_to_buffer(acb->qiov, acb->bounce);
2545
        acb->ret = bs->drv->bdrv_write(bs, sector_num, acb->bounce, nb_sectors);
2546
    } else {
2547
        acb->ret = bs->drv->bdrv_read(bs, sector_num, acb->bounce, nb_sectors);
2548 2549
    }

2550
    qemu_bh_schedule(acb->bh);
2551

2552
    return &acb->common;
2553 2554
}

2555 2556
static BlockDriverAIOCB *bdrv_aio_readv_em(BlockDriverState *bs,
        int64_t sector_num, QEMUIOVector *qiov, int nb_sectors,
2557
        BlockDriverCompletionFunc *cb, void *opaque)
2558
{
2559 2560
    return bdrv_aio_rw_vector(bs, sector_num, qiov, nb_sectors, cb, opaque, 0);
}
B
bellard 已提交
2561

2562 2563 2564 2565 2566
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);
2567 2568
}

2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 2586

typedef struct BlockDriverAIOCBCoroutine {
    BlockDriverAIOCB common;
    BlockRequest req;
    bool is_write;
    QEMUBH* bh;
} BlockDriverAIOCBCoroutine;

static void bdrv_aio_co_cancel_em(BlockDriverAIOCB *blockacb)
{
    qemu_aio_flush();
}

static AIOPool bdrv_em_co_aio_pool = {
    .aiocb_size         = sizeof(BlockDriverAIOCBCoroutine),
    .cancel             = bdrv_aio_co_cancel_em,
};

P
Paolo Bonzini 已提交
2587
static void bdrv_co_em_bh(void *opaque)
2588 2589 2590 2591 2592 2593 2594 2595
{
    BlockDriverAIOCBCoroutine *acb = opaque;

    acb->common.cb(acb->common.opaque, acb->req.error);
    qemu_bh_delete(acb->bh);
    qemu_aio_release(acb);
}

2596 2597 2598 2599 2600 2601 2602 2603 2604 2605 2606 2607 2608 2609
/* Invoke bdrv_co_do_readv/bdrv_co_do_writev */
static void coroutine_fn bdrv_co_do_rw(void *opaque)
{
    BlockDriverAIOCBCoroutine *acb = opaque;
    BlockDriverState *bs = acb->common.bs;

    if (!acb->is_write) {
        acb->req.error = bdrv_co_do_readv(bs, acb->req.sector,
            acb->req.nb_sectors, acb->req.qiov);
    } else {
        acb->req.error = bdrv_co_do_writev(bs, acb->req.sector,
            acb->req.nb_sectors, acb->req.qiov);
    }

P
Paolo Bonzini 已提交
2610
    acb->bh = qemu_bh_new(bdrv_co_em_bh, acb);
2611 2612 2613
    qemu_bh_schedule(acb->bh);
}

2614 2615 2616 2617 2618 2619
static BlockDriverAIOCB *bdrv_co_aio_rw_vector(BlockDriverState *bs,
                                               int64_t sector_num,
                                               QEMUIOVector *qiov,
                                               int nb_sectors,
                                               BlockDriverCompletionFunc *cb,
                                               void *opaque,
2620
                                               bool is_write)
2621 2622 2623 2624 2625 2626 2627 2628 2629 2630
{
    Coroutine *co;
    BlockDriverAIOCBCoroutine *acb;

    acb = qemu_aio_get(&bdrv_em_co_aio_pool, bs, cb, opaque);
    acb->req.sector = sector_num;
    acb->req.nb_sectors = nb_sectors;
    acb->req.qiov = qiov;
    acb->is_write = is_write;

2631
    co = qemu_coroutine_create(bdrv_co_do_rw);
2632 2633 2634 2635 2636
    qemu_coroutine_enter(co, acb);

    return &acb->common;
}

P
Paolo Bonzini 已提交
2637
static void coroutine_fn bdrv_aio_flush_co_entry(void *opaque)
2638
{
P
Paolo Bonzini 已提交
2639 2640
    BlockDriverAIOCBCoroutine *acb = opaque;
    BlockDriverState *bs = acb->common.bs;
2641

P
Paolo Bonzini 已提交
2642 2643
    acb->req.error = bdrv_co_flush(bs);
    acb->bh = qemu_bh_new(bdrv_co_em_bh, acb);
2644 2645 2646
    qemu_bh_schedule(acb->bh);
}

P
Paolo Bonzini 已提交
2647
BlockDriverAIOCB *bdrv_aio_flush(BlockDriverState *bs,
2648 2649
        BlockDriverCompletionFunc *cb, void *opaque)
{
P
Paolo Bonzini 已提交
2650
    trace_bdrv_aio_flush(bs, opaque);
2651

P
Paolo Bonzini 已提交
2652 2653
    Coroutine *co;
    BlockDriverAIOCBCoroutine *acb;
2654

P
Paolo Bonzini 已提交
2655 2656 2657
    acb = qemu_aio_get(&bdrv_em_co_aio_pool, bs, cb, opaque);
    co = qemu_coroutine_create(bdrv_aio_flush_co_entry);
    qemu_coroutine_enter(co, acb);
2658 2659 2660 2661

    return &acb->common;
}

2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689
static void coroutine_fn bdrv_aio_discard_co_entry(void *opaque)
{
    BlockDriverAIOCBCoroutine *acb = opaque;
    BlockDriverState *bs = acb->common.bs;

    acb->req.error = bdrv_co_discard(bs, acb->req.sector, acb->req.nb_sectors);
    acb->bh = qemu_bh_new(bdrv_co_em_bh, acb);
    qemu_bh_schedule(acb->bh);
}

BlockDriverAIOCB *bdrv_aio_discard(BlockDriverState *bs,
        int64_t sector_num, int nb_sectors,
        BlockDriverCompletionFunc *cb, void *opaque)
{
    Coroutine *co;
    BlockDriverAIOCBCoroutine *acb;

    trace_bdrv_aio_discard(bs, sector_num, nb_sectors, opaque);

    acb = qemu_aio_get(&bdrv_em_co_aio_pool, bs, cb, opaque);
    acb->req.sector = sector_num;
    acb->req.nb_sectors = nb_sectors;
    co = qemu_coroutine_create(bdrv_aio_discard_co_entry);
    qemu_coroutine_enter(co, acb);

    return &acb->common;
}

B
bellard 已提交
2690 2691
void bdrv_init(void)
{
2692
    module_call_init(MODULE_INIT_BLOCK);
B
bellard 已提交
2693
}
2694

2695 2696 2697 2698 2699 2700
void bdrv_init_with_whitelist(void)
{
    use_bdrv_whitelist = 1;
    bdrv_init();
}

2701 2702
void *qemu_aio_get(AIOPool *pool, BlockDriverState *bs,
                   BlockDriverCompletionFunc *cb, void *opaque)
2703 2704 2705
{
    BlockDriverAIOCB *acb;

2706 2707 2708
    if (pool->free_aiocb) {
        acb = pool->free_aiocb;
        pool->free_aiocb = acb->next;
2709
    } else {
2710
        acb = g_malloc0(pool->aiocb_size);
2711
        acb->pool = pool;
2712 2713 2714 2715 2716 2717 2718 2719 2720
    }
    acb->bs = bs;
    acb->cb = cb;
    acb->opaque = opaque;
    return acb;
}

void qemu_aio_release(void *p)
{
2721 2722 2723 2724
    BlockDriverAIOCB *acb = (BlockDriverAIOCB *)p;
    AIOPool *pool = acb->pool;
    acb->next = pool->free_aiocb;
    pool->free_aiocb = acb;
2725
}
B
bellard 已提交
2726

2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751 2752
/**************************************************************/
/* Coroutine block device emulation */

typedef struct CoroutineIOCompletion {
    Coroutine *coroutine;
    int ret;
} CoroutineIOCompletion;

static void bdrv_co_io_em_complete(void *opaque, int ret)
{
    CoroutineIOCompletion *co = opaque;

    co->ret = ret;
    qemu_coroutine_enter(co->coroutine, NULL);
}

static int coroutine_fn bdrv_co_io_em(BlockDriverState *bs, int64_t sector_num,
                                      int nb_sectors, QEMUIOVector *iov,
                                      bool is_write)
{
    CoroutineIOCompletion co = {
        .coroutine = qemu_coroutine_self(),
    };
    BlockDriverAIOCB *acb;

    if (is_write) {
2753 2754
        acb = bs->drv->bdrv_aio_writev(bs, sector_num, iov, nb_sectors,
                                       bdrv_co_io_em_complete, &co);
2755
    } else {
2756 2757
        acb = bs->drv->bdrv_aio_readv(bs, sector_num, iov, nb_sectors,
                                      bdrv_co_io_em_complete, &co);
2758 2759
    }

2760
    trace_bdrv_co_io_em(bs, sector_num, nb_sectors, is_write, acb);
2761 2762 2763 2764 2765 2766 2767 2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779 2780 2781 2782
    if (!acb) {
        return -EIO;
    }
    qemu_coroutine_yield();

    return co.ret;
}

static int coroutine_fn bdrv_co_readv_em(BlockDriverState *bs,
                                         int64_t sector_num, int nb_sectors,
                                         QEMUIOVector *iov)
{
    return bdrv_co_io_em(bs, sector_num, nb_sectors, iov, false);
}

static int coroutine_fn bdrv_co_writev_em(BlockDriverState *bs,
                                         int64_t sector_num, int nb_sectors,
                                         QEMUIOVector *iov)
{
    return bdrv_co_io_em(bs, sector_num, nb_sectors, iov, true);
}

P
Paolo Bonzini 已提交
2783
static void coroutine_fn bdrv_flush_co_entry(void *opaque)
2784
{
P
Paolo Bonzini 已提交
2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 2795 2796 2797 2798 2799 2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811 2812 2813 2814 2815 2816 2817 2818 2819 2820 2821 2822 2823 2824 2825 2826 2827 2828 2829 2830 2831 2832
    RwCo *rwco = opaque;

    rwco->ret = bdrv_co_flush(rwco->bs);
}

int coroutine_fn bdrv_co_flush(BlockDriverState *bs)
{
    if (bs->open_flags & BDRV_O_NO_FLUSH) {
        return 0;
    } else if (!bs->drv) {
        return 0;
    } else if (bs->drv->bdrv_co_flush) {
        return bs->drv->bdrv_co_flush(bs);
    } else if (bs->drv->bdrv_aio_flush) {
        BlockDriverAIOCB *acb;
        CoroutineIOCompletion co = {
            .coroutine = qemu_coroutine_self(),
        };

        acb = bs->drv->bdrv_aio_flush(bs, bdrv_co_io_em_complete, &co);
        if (acb == NULL) {
            return -EIO;
        } else {
            qemu_coroutine_yield();
            return co.ret;
        }
    } else {
        /*
         * 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;
    }
}

int bdrv_flush(BlockDriverState *bs)
{
    Coroutine *co;
    RwCo rwco = {
        .bs = bs,
        .ret = NOT_DONE,
2833 2834
    };

P
Paolo Bonzini 已提交
2835 2836 2837 2838 2839 2840 2841 2842 2843
    if (qemu_in_coroutine()) {
        /* Fast-path if already in coroutine context */
        bdrv_flush_co_entry(&rwco);
    } else {
        co = qemu_coroutine_create(bdrv_flush_co_entry);
        qemu_coroutine_enter(co, &rwco);
        while (rwco.ret == NOT_DONE) {
            qemu_aio_wait();
        }
2844
    }
P
Paolo Bonzini 已提交
2845 2846

    return rwco.ret;
2847 2848
}

2849 2850 2851 2852 2853 2854 2855 2856 2857 2858 2859 2860 2861 2862 2863 2864 2865 2866 2867 2868 2869 2870 2871 2872 2873 2874 2875 2876 2877 2878 2879 2880 2881 2882 2883 2884 2885 2886 2887 2888 2889 2890 2891 2892 2893 2894 2895 2896 2897 2898 2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909
static void coroutine_fn bdrv_discard_co_entry(void *opaque)
{
    RwCo *rwco = opaque;

    rwco->ret = bdrv_co_discard(rwco->bs, rwco->sector_num, rwco->nb_sectors);
}

int coroutine_fn bdrv_co_discard(BlockDriverState *bs, int64_t sector_num,
                                 int nb_sectors)
{
    if (!bs->drv) {
        return -ENOMEDIUM;
    } else if (bdrv_check_request(bs, sector_num, nb_sectors)) {
        return -EIO;
    } else if (bs->read_only) {
        return -EROFS;
    } else if (bs->drv->bdrv_co_discard) {
        return bs->drv->bdrv_co_discard(bs, sector_num, nb_sectors);
    } else if (bs->drv->bdrv_aio_discard) {
        BlockDriverAIOCB *acb;
        CoroutineIOCompletion co = {
            .coroutine = qemu_coroutine_self(),
        };

        acb = bs->drv->bdrv_aio_discard(bs, sector_num, nb_sectors,
                                        bdrv_co_io_em_complete, &co);
        if (acb == NULL) {
            return -EIO;
        } else {
            qemu_coroutine_yield();
            return co.ret;
        }
    } else {
        return 0;
    }
}

int bdrv_discard(BlockDriverState *bs, int64_t sector_num, int nb_sectors)
{
    Coroutine *co;
    RwCo rwco = {
        .bs = bs,
        .sector_num = sector_num,
        .nb_sectors = nb_sectors,
        .ret = NOT_DONE,
    };

    if (qemu_in_coroutine()) {
        /* Fast-path if already in coroutine context */
        bdrv_discard_co_entry(&rwco);
    } else {
        co = qemu_coroutine_create(bdrv_discard_co_entry);
        qemu_coroutine_enter(co, &rwco);
        while (rwco.ret == NOT_DONE) {
            qemu_aio_wait();
        }
    }

    return rwco.ret;
}

B
bellard 已提交
2910 2911 2912 2913 2914 2915 2916 2917 2918
/**************************************************************/
/* removable device support */

/**
 * Return TRUE if the media is present
 */
int bdrv_is_inserted(BlockDriverState *bs)
{
    BlockDriver *drv = bs->drv;
2919

B
bellard 已提交
2920 2921 2922
    if (!drv)
        return 0;
    if (!drv->bdrv_is_inserted)
2923 2924
        return 1;
    return drv->bdrv_is_inserted(bs);
B
bellard 已提交
2925 2926 2927
}

/**
2928 2929
 * Return whether the media changed since the last call to this
 * function, or -ENOTSUP if we don't know.  Most drivers don't know.
B
bellard 已提交
2930 2931 2932 2933 2934
 */
int bdrv_media_changed(BlockDriverState *bs)
{
    BlockDriver *drv = bs->drv;

2935 2936 2937 2938
    if (drv && drv->bdrv_media_changed) {
        return drv->bdrv_media_changed(bs);
    }
    return -ENOTSUP;
B
bellard 已提交
2939 2940 2941 2942 2943
}

/**
 * If eject_flag is TRUE, eject the media. Otherwise, close the tray
 */
2944
void bdrv_eject(BlockDriverState *bs, int eject_flag)
B
bellard 已提交
2945 2946 2947
{
    BlockDriver *drv = bs->drv;

2948 2949
    if (drv && drv->bdrv_eject) {
        drv->bdrv_eject(bs, eject_flag);
B
bellard 已提交
2950 2951 2952 2953 2954 2955 2956
    }
}

/**
 * Lock or unlock the media (if it is locked, the user won't be able
 * to eject it manually).
 */
2957
void bdrv_lock_medium(BlockDriverState *bs, bool locked)
B
bellard 已提交
2958 2959 2960
{
    BlockDriver *drv = bs->drv;

2961
    trace_bdrv_lock_medium(bs, locked);
S
Stefan Hajnoczi 已提交
2962

2963 2964
    if (drv && drv->bdrv_lock_medium) {
        drv->bdrv_lock_medium(bs, locked);
B
bellard 已提交
2965 2966
    }
}
2967 2968 2969 2970 2971 2972 2973 2974 2975 2976 2977

/* 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;
}
2978

2979 2980 2981
BlockDriverAIOCB *bdrv_aio_ioctl(BlockDriverState *bs,
        unsigned long int req, void *buf,
        BlockDriverCompletionFunc *cb, void *opaque)
2982
{
2983
    BlockDriver *drv = bs->drv;
2984

2985 2986 2987
    if (drv && drv->bdrv_aio_ioctl)
        return drv->bdrv_aio_ioctl(bs, req, buf, cb, opaque);
    return NULL;
2988
}
2989

2990 2991 2992 2993
void bdrv_set_buffer_alignment(BlockDriverState *bs, int align)
{
    bs->buffer_alignment = align;
}
2994

2995 2996 2997 2998
void *qemu_blockalign(BlockDriverState *bs, size_t size)
{
    return qemu_memalign((bs && bs->buffer_alignment) ? bs->buffer_alignment : 512, size);
}
2999 3000 3001 3002

void bdrv_set_dirty_tracking(BlockDriverState *bs, int enable)
{
    int64_t bitmap_size;
3003

3004
    bs->dirty_count = 0;
3005
    if (enable) {
3006 3007 3008 3009
        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;
3010

3011
            bs->dirty_bitmap = g_malloc0(bitmap_size);
3012
        }
3013
    } else {
3014
        if (bs->dirty_bitmap) {
3015
            g_free(bs->dirty_bitmap);
3016
            bs->dirty_bitmap = NULL;
3017
        }
3018 3019 3020 3021 3022
    }
}

int bdrv_get_dirty(BlockDriverState *bs, int64_t sector)
{
J
Jan Kiszka 已提交
3023
    int64_t chunk = sector / (int64_t)BDRV_SECTORS_PER_DIRTY_CHUNK;
3024

3025 3026
    if (bs->dirty_bitmap &&
        (sector << BDRV_SECTOR_BITS) < bdrv_getlength(bs)) {
3027 3028
        return !!(bs->dirty_bitmap[chunk / (sizeof(unsigned long) * 8)] &
            (1UL << (chunk % (sizeof(unsigned long) * 8))));
3029 3030 3031 3032 3033
    } else {
        return 0;
    }
}

3034 3035
void bdrv_reset_dirty(BlockDriverState *bs, int64_t cur_sector,
                      int nr_sectors)
3036 3037 3038
{
    set_dirty_bitmap(bs, cur_sector, nr_sectors, 0);
}
3039 3040 3041 3042 3043

int64_t bdrv_get_dirty_count(BlockDriverState *bs)
{
    return bs->dirty_count;
}
J
Jes Sorensen 已提交
3044

3045 3046 3047 3048 3049 3050 3051 3052 3053 3054 3055
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;
}

3056 3057
void bdrv_iostatus_enable(BlockDriverState *bs)
{
3058
    bs->iostatus_enabled = true;
3059
    bs->iostatus = BLOCK_DEVICE_IO_STATUS_OK;
3060 3061 3062 3063 3064 3065
}

/* The I/O status is only enabled if the drive explicitly
 * enables it _and_ the VM is configured to stop on errors */
bool bdrv_iostatus_is_enabled(const BlockDriverState *bs)
{
3066
    return (bs->iostatus_enabled &&
3067 3068 3069 3070 3071 3072 3073
           (bs->on_write_error == BLOCK_ERR_STOP_ENOSPC ||
            bs->on_write_error == BLOCK_ERR_STOP_ANY    ||
            bs->on_read_error == BLOCK_ERR_STOP_ANY));
}

void bdrv_iostatus_disable(BlockDriverState *bs)
{
3074
    bs->iostatus_enabled = false;
3075 3076 3077 3078 3079
}

void bdrv_iostatus_reset(BlockDriverState *bs)
{
    if (bdrv_iostatus_is_enabled(bs)) {
3080
        bs->iostatus = BLOCK_DEVICE_IO_STATUS_OK;
3081 3082 3083 3084 3085 3086 3087 3088
    }
}

/* XXX: Today this is set by device models because it makes the implementation
   quite simple. However, the block layer knows about the error, so it's
   possible to implement this without device models being involved */
void bdrv_iostatus_set_err(BlockDriverState *bs, int error)
{
3089 3090
    if (bdrv_iostatus_is_enabled(bs) &&
        bs->iostatus == BLOCK_DEVICE_IO_STATUS_OK) {
3091
        assert(error >= 0);
3092 3093
        bs->iostatus = error == ENOSPC ? BLOCK_DEVICE_IO_STATUS_NOSPACE :
                                         BLOCK_DEVICE_IO_STATUS_FAILED;
3094 3095 3096
    }
}

3097 3098 3099 3100 3101 3102 3103
void
bdrv_acct_start(BlockDriverState *bs, BlockAcctCookie *cookie, int64_t bytes,
        enum BlockAcctType type)
{
    assert(type < BDRV_MAX_IOTYPE);

    cookie->bytes = bytes;
C
Christoph Hellwig 已提交
3104
    cookie->start_time_ns = get_clock();
3105 3106 3107 3108 3109 3110 3111 3112 3113 3114
    cookie->type = type;
}

void
bdrv_acct_done(BlockDriverState *bs, BlockAcctCookie *cookie)
{
    assert(cookie->type < BDRV_MAX_IOTYPE);

    bs->nr_bytes[cookie->type] += cookie->bytes;
    bs->nr_ops[cookie->type]++;
C
Christoph Hellwig 已提交
3115
    bs->total_time_ns[cookie->type] += get_clock() - cookie->start_time_ns;
3116 3117
}

J
Jes Sorensen 已提交
3118 3119 3120 3121 3122
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;
K
Kevin Wolf 已提交
3123
    QEMUOptionParameter *backing_fmt, *backing_file, *size;
J
Jes Sorensen 已提交
3124 3125
    BlockDriverState *bs = NULL;
    BlockDriver *drv, *proto_drv;
3126
    BlockDriver *backing_drv = NULL;
J
Jes Sorensen 已提交
3127 3128 3129 3130 3131 3132
    int ret = 0;

    /* Find driver and parse its options */
    drv = bdrv_find_format(fmt);
    if (!drv) {
        error_report("Unknown file format '%s'", fmt);
3133
        ret = -EINVAL;
J
Jes Sorensen 已提交
3134 3135 3136 3137 3138 3139
        goto out;
    }

    proto_drv = bdrv_find_protocol(filename);
    if (!proto_drv) {
        error_report("Unknown protocol '%s'", filename);
3140
        ret = -EINVAL;
J
Jes Sorensen 已提交
3141 3142 3143 3144 3145 3146 3147 3148 3149 3150 3151 3152 3153 3154 3155 3156 3157 3158
        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);
3159
            ret = -EINVAL;
J
Jes Sorensen 已提交
3160 3161 3162 3163 3164 3165 3166 3167 3168
            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);
3169
            ret = -EINVAL;
J
Jes Sorensen 已提交
3170 3171 3172 3173 3174 3175 3176 3177
            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);
3178
            ret = -EINVAL;
J
Jes Sorensen 已提交
3179 3180 3181 3182
            goto out;
        }
    }

3183 3184 3185 3186 3187
    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");
3188
            ret = -EINVAL;
3189 3190 3191 3192
            goto out;
        }
    }

J
Jes Sorensen 已提交
3193 3194
    backing_fmt = get_option_parameter(param, BLOCK_OPT_BACKING_FMT);
    if (backing_fmt && backing_fmt->value.s) {
3195 3196
        backing_drv = bdrv_find_format(backing_fmt->value.s);
        if (!backing_drv) {
J
Jes Sorensen 已提交
3197 3198
            error_report("Unknown backing file format '%s'",
                         backing_fmt->value.s);
3199
            ret = -EINVAL;
J
Jes Sorensen 已提交
3200 3201 3202 3203 3204 3205
            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
K
Kevin Wolf 已提交
3206 3207
    size = get_option_parameter(param, BLOCK_OPT_SIZE);
    if (size && size->value.n == -1) {
J
Jes Sorensen 已提交
3208 3209 3210 3211 3212 3213
        if (backing_file && backing_file->value.s) {
            uint64_t size;
            char buf[32];

            bs = bdrv_new("");

3214
            ret = bdrv_open(bs, backing_file->value.s, flags, backing_drv);
J
Jes Sorensen 已提交
3215
            if (ret < 0) {
3216
                error_report("Could not open '%s'", backing_file->value.s);
J
Jes Sorensen 已提交
3217 3218 3219 3220 3221 3222 3223 3224 3225
                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");
3226
            ret = -EINVAL;
J
Jes Sorensen 已提交
3227 3228 3229 3230 3231 3232 3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 3245 3246 3247 3248 3249 3250 3251 3252 3253 3254 3255 3256
            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);
    }
3257 3258

    return ret;
J
Jes Sorensen 已提交
3259
}