blockdev.c 100.1 KB
Newer Older
1 2 3 4 5 6 7
/*
 * QEMU host block devices
 *
 * Copyright (c) 2003-2008 Fabrice Bellard
 *
 * This work is licensed under the terms of the GNU GPL, version 2 or
 * later.  See the COPYING file in the top-level directory.
8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
 *
 * This file incorporates work covered by the following copyright and
 * permission notice:
 *
 * Copyright (c) 2003-2008 Fabrice Bellard
 *
 * 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.
31 32
 */

M
Markus Armbruster 已提交
33
#include "sysemu/block-backend.h"
34
#include "sysemu/blockdev.h"
P
Paolo Bonzini 已提交
35
#include "hw/block/block.h"
36
#include "block/blockjob.h"
37
#include "block/throttle-groups.h"
38
#include "monitor/monitor.h"
39
#include "qemu/error-report.h"
40 41
#include "qemu/option.h"
#include "qemu/config-file.h"
42
#include "qapi/qmp/types.h"
K
Kevin Wolf 已提交
43
#include "qapi-visit.h"
44
#include "qapi/qmp/qerror.h"
K
Kevin Wolf 已提交
45
#include "qapi/qmp-output-visitor.h"
46
#include "qapi/util.h"
47
#include "sysemu/sysemu.h"
48
#include "block/block_int.h"
L
Luiz Capitulino 已提交
49
#include "qmp-commands.h"
S
Stefan Hajnoczi 已提交
50
#include "trace.h"
51
#include "sysemu/arch_init.h"
52

53 54 55 56 57 58 59 60 61 62 63 64
static const char *const if_name[IF_COUNT] = {
    [IF_NONE] = "none",
    [IF_IDE] = "ide",
    [IF_SCSI] = "scsi",
    [IF_FLOPPY] = "floppy",
    [IF_PFLASH] = "pflash",
    [IF_MTD] = "mtd",
    [IF_SD] = "sd",
    [IF_VIRTIO] = "virtio",
    [IF_XEN] = "xen",
};

65
static int if_max_devs[IF_COUNT] = {
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81
    /*
     * Do not change these numbers!  They govern how drive option
     * index maps to unit and bus.  That mapping is ABI.
     *
     * All controllers used to imlement if=T drives need to support
     * if_max_devs[T] units, for any T with if_max_devs[T] != 0.
     * Otherwise, some index values map to "impossible" bus, unit
     * values.
     *
     * For instance, if you change [IF_SCSI] to 255, -drive
     * if=scsi,index=12 no longer means bus=1,unit=5, but
     * bus=0,unit=12.  With an lsi53c895a controller (7 units max),
     * the drive can't be set up.  Regression.
     */
    [IF_IDE] = 2,
    [IF_SCSI] = 7,
82 83
};

84 85 86 87 88 89
/**
 * Boards may call this to offer board-by-board overrides
 * of the default, global values.
 */
void override_max_devs(BlockInterfaceType type, int max_devs)
{
90
    BlockBackend *blk;
91 92 93 94 95 96
    DriveInfo *dinfo;

    if (max_devs <= 0) {
        return;
    }

97 98
    for (blk = blk_next(NULL); blk; blk = blk_next(blk)) {
        dinfo = blk_legacy_dinfo(blk);
99 100 101 102 103 104 105 106 107 108 109
        if (dinfo->type == type) {
            fprintf(stderr, "Cannot override units-per-bus property of"
                    " the %s interface, because a drive of that type has"
                    " already been added.\n", if_name[type]);
            g_assert_not_reached();
        }
    }

    if_max_devs[type] = max_devs;
}

110 111 112 113 114 115 116
/*
 * We automatically delete the drive when a device using it gets
 * unplugged.  Questionable feature, but we can't just drop it.
 * Device models call blockdev_mark_auto_del() to schedule the
 * automatic deletion, and generic qdev code calls blockdev_auto_del()
 * when deletion is actually safe.
 */
117
void blockdev_mark_auto_del(BlockBackend *blk)
118
{
119
    DriveInfo *dinfo = blk_legacy_dinfo(blk);
120
    BlockDriverState *bs = blk_bs(blk);
121
    AioContext *aio_context;
122

123
    if (!dinfo) {
124 125 126
        return;
    }

M
Max Reitz 已提交
127 128 129
    if (bs) {
        aio_context = bdrv_get_aio_context(bs);
        aio_context_acquire(aio_context);
130

M
Max Reitz 已提交
131 132 133
        if (bs->job) {
            block_job_cancel(bs->job);
        }
134

M
Max Reitz 已提交
135 136
        aio_context_release(aio_context);
    }
137

138
    dinfo->auto_del = 1;
139 140
}

141
void blockdev_auto_del(BlockBackend *blk)
142
{
143
    DriveInfo *dinfo = blk_legacy_dinfo(blk);
144

145
    if (dinfo && dinfo->auto_del) {
146
        blk_unref(blk);
147 148 149
    }
}

150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166
/**
 * Returns the current mapping of how many units per bus
 * a particular interface can support.
 *
 *  A positive integer indicates n units per bus.
 *  0 implies the mapping has not been established.
 * -1 indicates an invalid BlockInterfaceType was given.
 */
int drive_get_max_devs(BlockInterfaceType type)
{
    if (type >= IF_IDE && type < IF_COUNT) {
        return if_max_devs[type];
    }

    return -1;
}

167 168 169 170 171 172 173 174 175 176 177 178
static int drive_index_to_bus_id(BlockInterfaceType type, int index)
{
    int max_devs = if_max_devs[type];
    return max_devs ? index / max_devs : 0;
}

static int drive_index_to_unit_id(BlockInterfaceType type, int index)
{
    int max_devs = if_max_devs[type];
    return max_devs ? index % max_devs : index;
}

179 180
QemuOpts *drive_def(const char *optstr)
{
181
    return qemu_opts_parse_noisily(qemu_find_opts("drive"), optstr, false);
182 183 184
}

QemuOpts *drive_add(BlockInterfaceType type, int index, const char *file,
185
                    const char *optstr)
186 187 188
{
    QemuOpts *opts;

189
    opts = drive_def(optstr);
190 191 192
    if (!opts) {
        return NULL;
    }
193
    if (type != IF_DEFAULT) {
194
        qemu_opt_set(opts, "if", if_name[type], &error_abort);
195 196
    }
    if (index >= 0) {
197
        qemu_opt_set_number(opts, "index", index, &error_abort);
198
    }
199
    if (file)
200
        qemu_opt_set(opts, "file", file, &error_abort);
201 202 203 204 205
    return opts;
}

DriveInfo *drive_get(BlockInterfaceType type, int bus, int unit)
{
206
    BlockBackend *blk;
207 208
    DriveInfo *dinfo;

209 210 211 212
    for (blk = blk_next(NULL); blk; blk = blk_next(blk)) {
        dinfo = blk_legacy_dinfo(blk);
        if (dinfo && dinfo->type == type
            && dinfo->bus == bus && dinfo->unit == unit) {
213
            return dinfo;
214
        }
215 216 217 218 219
    }

    return NULL;
}

J
John Snow 已提交
220 221
bool drive_check_orphaned(void)
{
222
    BlockBackend *blk;
J
John Snow 已提交
223 224 225
    DriveInfo *dinfo;
    bool rs = false;

226 227
    for (blk = blk_next(NULL); blk; blk = blk_next(blk)) {
        dinfo = blk_legacy_dinfo(blk);
J
John Snow 已提交
228 229
        /* If dinfo->bdrv->dev is NULL, it has no device attached. */
        /* Unless this is a default drive, this may be an oversight. */
230
        if (!blk_get_attached_dev(blk) && !dinfo->is_default &&
J
John Snow 已提交
231 232 233
            dinfo->type != IF_NONE) {
            fprintf(stderr, "Warning: Orphaned drive without device: "
                    "id=%s,file=%s,if=%s,bus=%d,unit=%d\n",
M
Max Reitz 已提交
234 235
                    blk_name(blk), blk_bs(blk) ? blk_bs(blk)->filename : "",
                    if_name[dinfo->type], dinfo->bus, dinfo->unit);
J
John Snow 已提交
236 237 238 239 240 241 242
            rs = true;
        }
    }

    return rs;
}

243 244 245 246 247 248 249
DriveInfo *drive_get_by_index(BlockInterfaceType type, int index)
{
    return drive_get(type,
                     drive_index_to_bus_id(type, index),
                     drive_index_to_unit_id(type, index));
}

250 251 252
int drive_get_max_bus(BlockInterfaceType type)
{
    int max_bus;
253
    BlockBackend *blk;
254 255 256
    DriveInfo *dinfo;

    max_bus = -1;
257 258 259
    for (blk = blk_next(NULL); blk; blk = blk_next(blk)) {
        dinfo = blk_legacy_dinfo(blk);
        if (dinfo && dinfo->type == type && dinfo->bus > max_bus) {
260
            max_bus = dinfo->bus;
261
        }
262 263 264 265
    }
    return max_bus;
}

266 267 268 269 270 271 272 273 274 275
/* Get a block device.  This should only be used for single-drive devices
   (e.g. SD/Floppy/MTD).  Multi-disk devices (scsi/ide) should use the
   appropriate bus.  */
DriveInfo *drive_get_next(BlockInterfaceType type)
{
    static int next_block_unit[IF_COUNT];

    return drive_get(type, 0, next_block_unit[type]++);
}

276 277
static void bdrv_format_print(void *opaque, const char *name)
{
278
    error_printf(" %s", name);
279 280
}

281 282
typedef struct {
    QEMUBH *bh;
F
Fam Zheng 已提交
283 284
    BlockDriverState *bs;
} BDRVPutRefBH;
285

F
Fam Zheng 已提交
286
static void bdrv_put_ref_bh(void *opaque)
287
{
F
Fam Zheng 已提交
288
    BDRVPutRefBH *s = opaque;
289

F
Fam Zheng 已提交
290
    bdrv_unref(s->bs);
291 292 293 294 295
    qemu_bh_delete(s->bh);
    g_free(s);
}

/*
F
Fam Zheng 已提交
296
 * Release a BDS reference in a BH
297
 *
F
Fam Zheng 已提交
298 299 300
 * It is not safe to use bdrv_unref() from a callback function when the callers
 * still need the BlockDriverState.  In such cases we schedule a BH to release
 * the reference.
301
 */
F
Fam Zheng 已提交
302
static void bdrv_put_ref_bh_schedule(BlockDriverState *bs)
303
{
F
Fam Zheng 已提交
304
    BDRVPutRefBH *s;
305

F
Fam Zheng 已提交
306 307 308
    s = g_new(BDRVPutRefBH, 1);
    s->bh = qemu_bh_new(bdrv_put_ref_bh, s);
    s->bs = bs;
309 310 311
    qemu_bh_schedule(s->bh);
}

312
static int parse_block_error_action(const char *buf, bool is_read, Error **errp)
313 314
{
    if (!strcmp(buf, "ignore")) {
315
        return BLOCKDEV_ON_ERROR_IGNORE;
316
    } else if (!is_read && !strcmp(buf, "enospc")) {
317
        return BLOCKDEV_ON_ERROR_ENOSPC;
318
    } else if (!strcmp(buf, "stop")) {
319
        return BLOCKDEV_ON_ERROR_STOP;
320
    } else if (!strcmp(buf, "report")) {
321
        return BLOCKDEV_ON_ERROR_REPORT;
322
    } else {
323 324
        error_setg(errp, "'%s' invalid %s error action",
                   buf, is_read ? "read" : "write");
325 326 327 328
        return -1;
    }
}

329
static bool check_throttle_config(ThrottleConfig *cfg, Error **errp)
330
{
331 332 333
    if (throttle_conflicting(cfg)) {
        error_setg(errp, "bps/iops/max total values and read/write values"
                         " cannot be used at the same time");
334 335 336
        return false;
    }

337 338
    if (!throttle_is_valid(cfg)) {
        error_setg(errp, "bps/iops/maxs values must be 0 or greater");
339 340 341
        return false;
    }

342 343 344 345 346 347
    if (throttle_max_is_missing_limit(cfg)) {
        error_setg(errp, "bps_max/iops_max require corresponding"
                         " bps/iops values");
        return false;
    }

348 349 350
    return true;
}

351 352
typedef enum { MEDIA_DISK, MEDIA_CDROM } DriveMediaType;

353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462
/* All parameters but @opts are optional and may be set to NULL. */
static void extract_common_blockdev_options(QemuOpts *opts, int *bdrv_flags,
    const char **throttling_group, ThrottleConfig *throttle_cfg,
    BlockdevDetectZeroesOptions *detect_zeroes, Error **errp)
{
    const char *discard;
    Error *local_error = NULL;
    const char *aio;

    if (bdrv_flags) {
        if (!qemu_opt_get_bool(opts, "read-only", false)) {
            *bdrv_flags |= BDRV_O_RDWR;
        }
        if (qemu_opt_get_bool(opts, "copy-on-read", false)) {
            *bdrv_flags |= BDRV_O_COPY_ON_READ;
        }

        if ((discard = qemu_opt_get(opts, "discard")) != NULL) {
            if (bdrv_parse_discard_flags(discard, bdrv_flags) != 0) {
                error_setg(errp, "Invalid discard option");
                return;
            }
        }

        if (qemu_opt_get_bool(opts, BDRV_OPT_CACHE_WB, true)) {
            *bdrv_flags |= BDRV_O_CACHE_WB;
        }
        if (qemu_opt_get_bool(opts, BDRV_OPT_CACHE_DIRECT, false)) {
            *bdrv_flags |= BDRV_O_NOCACHE;
        }
        if (qemu_opt_get_bool(opts, BDRV_OPT_CACHE_NO_FLUSH, false)) {
            *bdrv_flags |= BDRV_O_NO_FLUSH;
        }

        if ((aio = qemu_opt_get(opts, "aio")) != NULL) {
            if (!strcmp(aio, "native")) {
                *bdrv_flags |= BDRV_O_NATIVE_AIO;
            } else if (!strcmp(aio, "threads")) {
                /* this is the default */
            } else {
               error_setg(errp, "invalid aio option");
               return;
            }
        }
    }

    /* disk I/O throttling */
    if (throttling_group) {
        *throttling_group = qemu_opt_get(opts, "throttling.group");
    }

    if (throttle_cfg) {
        memset(throttle_cfg, 0, sizeof(*throttle_cfg));
        throttle_cfg->buckets[THROTTLE_BPS_TOTAL].avg =
            qemu_opt_get_number(opts, "throttling.bps-total", 0);
        throttle_cfg->buckets[THROTTLE_BPS_READ].avg  =
            qemu_opt_get_number(opts, "throttling.bps-read", 0);
        throttle_cfg->buckets[THROTTLE_BPS_WRITE].avg =
            qemu_opt_get_number(opts, "throttling.bps-write", 0);
        throttle_cfg->buckets[THROTTLE_OPS_TOTAL].avg =
            qemu_opt_get_number(opts, "throttling.iops-total", 0);
        throttle_cfg->buckets[THROTTLE_OPS_READ].avg =
            qemu_opt_get_number(opts, "throttling.iops-read", 0);
        throttle_cfg->buckets[THROTTLE_OPS_WRITE].avg =
            qemu_opt_get_number(opts, "throttling.iops-write", 0);

        throttle_cfg->buckets[THROTTLE_BPS_TOTAL].max =
            qemu_opt_get_number(opts, "throttling.bps-total-max", 0);
        throttle_cfg->buckets[THROTTLE_BPS_READ].max  =
            qemu_opt_get_number(opts, "throttling.bps-read-max", 0);
        throttle_cfg->buckets[THROTTLE_BPS_WRITE].max =
            qemu_opt_get_number(opts, "throttling.bps-write-max", 0);
        throttle_cfg->buckets[THROTTLE_OPS_TOTAL].max =
            qemu_opt_get_number(opts, "throttling.iops-total-max", 0);
        throttle_cfg->buckets[THROTTLE_OPS_READ].max =
            qemu_opt_get_number(opts, "throttling.iops-read-max", 0);
        throttle_cfg->buckets[THROTTLE_OPS_WRITE].max =
            qemu_opt_get_number(opts, "throttling.iops-write-max", 0);

        throttle_cfg->op_size =
            qemu_opt_get_number(opts, "throttling.iops-size", 0);

        if (!check_throttle_config(throttle_cfg, errp)) {
            return;
        }
    }

    if (detect_zeroes) {
        *detect_zeroes =
            qapi_enum_parse(BlockdevDetectZeroesOptions_lookup,
                            qemu_opt_get(opts, "detect-zeroes"),
                            BLOCKDEV_DETECT_ZEROES_OPTIONS_MAX,
                            BLOCKDEV_DETECT_ZEROES_OPTIONS_OFF,
                            &local_error);
        if (local_error) {
            error_propagate(errp, local_error);
            return;
        }

        if (bdrv_flags &&
            *detect_zeroes == BLOCKDEV_DETECT_ZEROES_OPTIONS_UNMAP &&
            !(*bdrv_flags & BDRV_O_UNMAP))
        {
            error_setg(errp, "setting detect-zeroes to unmap is not allowed "
                             "without setting discard operation to unmap");
            return;
        }
    }
}

463
/* Takes the ownership of bs_opts */
464 465
static BlockBackend *blockdev_init(const char *file, QDict *bs_opts,
                                   Error **errp)
466 467 468 469
{
    const char *buf;
    int bdrv_flags = 0;
    int on_read_error, on_write_error;
M
Markus Armbruster 已提交
470
    BlockBackend *blk;
471
    BlockDriverState *bs;
472
    ThrottleConfig cfg;
473
    int snapshot = 0;
474
    Error *error = NULL;
475 476
    QemuOpts *opts;
    const char *id;
477
    bool has_driver_specific_opts;
478 479 480
    BlockdevDetectZeroesOptions detect_zeroes =
        BLOCKDEV_DETECT_ZEROES_OPTIONS_OFF;
    const char *throttling_group = NULL;
481

482 483 484
    /* Check common options by copying from bs_opts to opts, all other options
     * stay in bs_opts for processing by bdrv_open(). */
    id = qdict_get_try_str(bs_opts, "id");
485
    opts = qemu_opts_create(&qemu_common_drive_opts, id, 1, &error);
486
    if (error) {
487
        error_propagate(errp, error);
488
        goto err_no_opts;
489 490 491
    }

    qemu_opts_absorb_qdict(opts, bs_opts, &error);
492
    if (error) {
493
        error_propagate(errp, error);
494
        goto early_err;
495 496 497 498 499 500
    }

    if (id) {
        qdict_del(bs_opts, "id");
    }

501 502
    has_driver_specific_opts = !!qdict_size(bs_opts);

503 504 505
    /* extract parameters */
    snapshot = qemu_opt_get_bool(opts, "snapshot", 0);

506 507 508 509 510
    extract_common_blockdev_options(opts, &bdrv_flags, &throttling_group, &cfg,
                                    &detect_zeroes, &error);
    if (error) {
        error_propagate(errp, error);
        goto early_err;
511 512 513
    }

    if ((buf = qemu_opt_get(opts, "format")) != NULL) {
514 515 516 517
        if (is_help_option(buf)) {
            error_printf("Supported formats:");
            bdrv_iterate_format(bdrv_format_print, NULL);
            error_printf("\n");
518
            goto early_err;
519
        }
520

521 522
        if (qdict_haskey(bs_opts, "driver")) {
            error_setg(errp, "Cannot specify both 'driver' and 'format'");
523
            goto early_err;
524
        }
525
        qdict_put(bs_opts, "driver", qstring_from_str(buf));
526 527
    }

528
    on_write_error = BLOCKDEV_ON_ERROR_ENOSPC;
529
    if ((buf = qemu_opt_get(opts, "werror")) != NULL) {
530
        on_write_error = parse_block_error_action(buf, 0, &error);
531
        if (error) {
532
            error_propagate(errp, error);
533
            goto early_err;
534 535 536
        }
    }

537
    on_read_error = BLOCKDEV_ON_ERROR_REPORT;
538
    if ((buf = qemu_opt_get(opts, "rerror")) != NULL) {
539
        on_read_error = parse_block_error_action(buf, 1, &error);
540
        if (error) {
541
            error_propagate(errp, error);
542
            goto early_err;
543 544 545
        }
    }

546 547 548 549 550 551
    if (snapshot) {
        /* always use cache=unsafe with snapshot */
        bdrv_flags &= ~BDRV_O_CACHE_MASK;
        bdrv_flags |= (BDRV_O_SNAPSHOT|BDRV_O_CACHE_WB|BDRV_O_NO_FLUSH);
    }

552
    /* init */
553
    if ((!file || !*file) && !has_driver_specific_opts) {
554 555 556
        BlockBackendRootState *blk_rs;

        blk = blk_new(qemu_opts_id(opts), errp);
557 558 559
        if (!blk) {
            goto early_err;
        }
560

561 562
        blk_rs = blk_get_root_state(blk);
        blk_rs->open_flags    = bdrv_flags;
563
        blk_rs->read_only     = !(bdrv_flags & BDRV_O_RDWR);
564 565 566 567 568 569 570 571 572 573
        blk_rs->detect_zeroes = detect_zeroes;

        if (throttle_enabled(&cfg)) {
            if (!throttling_group) {
                throttling_group = blk_name(blk);
            }
            blk_rs->throttle_group = g_strdup(throttling_group);
            blk_rs->throttle_state = throttle_group_incref(throttling_group);
            blk_rs->throttle_state->cfg = cfg;
        }
574

575 576 577
        QDECREF(bs_opts);
    } else {
        if (file && !*file) {
578 579
            file = NULL;
        }
580

581 582 583 584 585 586
        blk = blk_new_open(qemu_opts_id(opts), file, NULL, bs_opts, bdrv_flags,
                           errp);
        if (!blk) {
            goto err_no_bs_opts;
        }
        bs = blk_bs(blk);
587

588
        bs->detect_zeroes = detect_zeroes;
589

590 591 592 593 594 595 596 597
        /* disk I/O throttling */
        if (throttle_enabled(&cfg)) {
            if (!throttling_group) {
                throttling_group = blk_name(blk);
            }
            bdrv_io_limits_enable(bs, throttling_group);
            bdrv_set_io_limits(bs, &cfg);
        }
598

599 600
        if (bdrv_key_required(bs)) {
            autostart = 0;
601
        }
602 603
    }

604
    blk_set_on_error(blk, on_read_error, on_write_error);
605

606
err_no_bs_opts:
607
    qemu_opts_del(opts);
608
    return blk;
609

610 611
early_err:
    qemu_opts_del(opts);
612 613
err_no_opts:
    QDECREF(bs_opts);
614
    return NULL;
615 616
}

617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664
static QemuOptsList qemu_root_bds_opts;

/* Takes the ownership of bs_opts */
static BlockDriverState *bds_tree_init(QDict *bs_opts, Error **errp)
{
    BlockDriverState *bs;
    QemuOpts *opts;
    Error *local_error = NULL;
    BlockdevDetectZeroesOptions detect_zeroes;
    int ret;
    int bdrv_flags = 0;

    opts = qemu_opts_create(&qemu_root_bds_opts, NULL, 1, errp);
    if (!opts) {
        goto fail;
    }

    qemu_opts_absorb_qdict(opts, bs_opts, &local_error);
    if (local_error) {
        error_propagate(errp, local_error);
        goto fail;
    }

    extract_common_blockdev_options(opts, &bdrv_flags, NULL, NULL,
                                    &detect_zeroes, &local_error);
    if (local_error) {
        error_propagate(errp, local_error);
        goto fail;
    }

    bs = NULL;
    ret = bdrv_open(&bs, NULL, NULL, bs_opts, bdrv_flags, errp);
    if (ret < 0) {
        goto fail_no_bs_opts;
    }

    bs->detect_zeroes = detect_zeroes;

fail_no_bs_opts:
    qemu_opts_del(opts);
    return bs;

fail:
    qemu_opts_del(opts);
    QDECREF(bs_opts);
    return NULL;
}

665 666
static void qemu_opt_rename(QemuOpts *opts, const char *from, const char *to,
                            Error **errp)
667 668 669 670 671
{
    const char *value;

    value = qemu_opt_get(opts, from);
    if (value) {
672 673 674 675 676
        if (qemu_opt_find(opts, to)) {
            error_setg(errp, "'%s' and its alias '%s' can't be used at the "
                       "same time", to, from);
            return;
        }
677 678 679 680
    }

    /* rename all items in opts */
    while ((value = qemu_opt_get(opts, from))) {
681
        qemu_opt_set(opts, to, value, &error_abort);
682 683 684 685
        qemu_opt_unset(opts, from);
    }
}

686 687 688 689 690
QemuOptsList qemu_legacy_drive_opts = {
    .name = "drive",
    .head = QTAILQ_HEAD_INITIALIZER(qemu_legacy_drive_opts.head),
    .desc = {
        {
691 692 693 694 695 696 697 698 699 700 701 702
            .name = "bus",
            .type = QEMU_OPT_NUMBER,
            .help = "bus number",
        },{
            .name = "unit",
            .type = QEMU_OPT_NUMBER,
            .help = "unit number (i.e. lun for scsi)",
        },{
            .name = "index",
            .type = QEMU_OPT_NUMBER,
            .help = "index number",
        },{
703 704 705
            .name = "media",
            .type = QEMU_OPT_STRING,
            .help = "media type (disk, cdrom)",
706 707 708 709
        },{
            .name = "if",
            .type = QEMU_OPT_STRING,
            .help = "interface (ide, scsi, sd, mtd, floppy, pflash, virtio)",
710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725
        },{
            .name = "cyls",
            .type = QEMU_OPT_NUMBER,
            .help = "number of cylinders (ide disk geometry)",
        },{
            .name = "heads",
            .type = QEMU_OPT_NUMBER,
            .help = "number of heads (ide disk geometry)",
        },{
            .name = "secs",
            .type = QEMU_OPT_NUMBER,
            .help = "number of sectors (ide disk geometry)",
        },{
            .name = "trans",
            .type = QEMU_OPT_STRING,
            .help = "chs translation (auto, lba, none)",
726 727 728 729
        },{
            .name = "boot",
            .type = QEMU_OPT_BOOL,
            .help = "(deprecated, ignored)",
730 731 732 733
        },{
            .name = "addr",
            .type = QEMU_OPT_STRING,
            .help = "pci address (virtio only)",
734 735 736 737
        },{
            .name = "serial",
            .type = QEMU_OPT_STRING,
            .help = "disk serial number",
M
Max Reitz 已提交
738 739 740 741
        },{
            .name = "file",
            .type = QEMU_OPT_STRING,
            .help = "file name",
742
        },
743 744 745 746 747 748

        /* Options that are passed on, but have special semantics with -drive */
        {
            .name = "read-only",
            .type = QEMU_OPT_BOOL,
            .help = "open drive file as read-only",
749 750 751 752 753 754 755 756
        },{
            .name = "rerror",
            .type = QEMU_OPT_STRING,
            .help = "read error action",
        },{
            .name = "werror",
            .type = QEMU_OPT_STRING,
            .help = "write error action",
757 758 759 760 761 762
        },{
            .name = "copy-on-read",
            .type = QEMU_OPT_BOOL,
            .help = "copy read data from backing file into image file",
        },

763 764 765 766
        { /* end of list */ }
    },
};

767
DriveInfo *drive_new(QemuOpts *all_opts, BlockInterfaceType block_default_type)
768
{
K
Kevin Wolf 已提交
769
    const char *value;
770
    BlockBackend *blk;
771
    DriveInfo *dinfo = NULL;
772
    QDict *bs_opts;
773 774
    QemuOpts *legacy_opts;
    DriveMediaType media = MEDIA_DISK;
775
    BlockInterfaceType type;
776
    int cyls, heads, secs, translation;
777
    int max_devs, bus_id, unit_id, index;
778
    const char *devaddr;
779
    const char *werror, *rerror;
F
Fam Zheng 已提交
780 781
    bool read_only = false;
    bool copy_on_read;
782
    const char *serial;
M
Max Reitz 已提交
783
    const char *filename;
784
    Error *local_err = NULL;
785
    int i;
K
Kevin Wolf 已提交
786

787
    /* Change legacy command line options into QMP ones */
788 789 790 791 792 793 794
    static const struct {
        const char *from;
        const char *to;
    } opt_renames[] = {
        { "iops",           "throttling.iops-total" },
        { "iops_rd",        "throttling.iops-read" },
        { "iops_wr",        "throttling.iops-write" },
795

796 797 798
        { "bps",            "throttling.bps-total" },
        { "bps_rd",         "throttling.bps-read" },
        { "bps_wr",         "throttling.bps-write" },
799

800 801 802
        { "iops_max",       "throttling.iops-total-max" },
        { "iops_rd_max",    "throttling.iops-read-max" },
        { "iops_wr_max",    "throttling.iops-write-max" },
803

804 805 806
        { "bps_max",        "throttling.bps-total-max" },
        { "bps_rd_max",     "throttling.bps-read-max" },
        { "bps_wr_max",     "throttling.bps-write-max" },
807

808
        { "iops_size",      "throttling.iops-size" },
809

810 811
        { "group",          "throttling.group" },

812 813 814 815
        { "readonly",       "read-only" },
    };

    for (i = 0; i < ARRAY_SIZE(opt_renames); i++) {
816 817 818
        qemu_opt_rename(all_opts, opt_renames[i].from, opt_renames[i].to,
                        &local_err);
        if (local_err) {
819
            error_report_err(local_err);
820 821
            return NULL;
        }
822
    }
823

K
Kevin Wolf 已提交
824 825 826 827 828 829 830 831 832 833
    value = qemu_opt_get(all_opts, "cache");
    if (value) {
        int flags = 0;

        if (bdrv_parse_cache_flags(value, &flags) != 0) {
            error_report("invalid cache option");
            return NULL;
        }

        /* Specific options take precedence */
834 835
        if (!qemu_opt_get(all_opts, BDRV_OPT_CACHE_WB)) {
            qemu_opt_set_bool(all_opts, BDRV_OPT_CACHE_WB,
836
                              !!(flags & BDRV_O_CACHE_WB), &error_abort);
K
Kevin Wolf 已提交
837
        }
838 839
        if (!qemu_opt_get(all_opts, BDRV_OPT_CACHE_DIRECT)) {
            qemu_opt_set_bool(all_opts, BDRV_OPT_CACHE_DIRECT,
840
                              !!(flags & BDRV_O_NOCACHE), &error_abort);
K
Kevin Wolf 已提交
841
        }
842 843
        if (!qemu_opt_get(all_opts, BDRV_OPT_CACHE_NO_FLUSH)) {
            qemu_opt_set_bool(all_opts, BDRV_OPT_CACHE_NO_FLUSH,
844
                              !!(flags & BDRV_O_NO_FLUSH), &error_abort);
K
Kevin Wolf 已提交
845 846 847 848
        }
        qemu_opt_unset(all_opts, "cache");
    }

849 850 851 852
    /* Get a QDict for processing the options */
    bs_opts = qdict_new();
    qemu_opts_to_qdict(all_opts, bs_opts);

853 854
    legacy_opts = qemu_opts_create(&qemu_legacy_drive_opts, NULL, 0,
                                   &error_abort);
855
    qemu_opts_absorb_qdict(legacy_opts, bs_opts, &local_err);
856
    if (local_err) {
857
        error_report_err(local_err);
858 859 860
        goto fail;
    }

861 862 863 864 865 866 867
    /* Deprecated option boot=[on|off] */
    if (qemu_opt_get(legacy_opts, "boot") != NULL) {
        fprintf(stderr, "qemu-kvm: boot=on|off is deprecated and will be "
                "ignored. Future versions will reject this parameter. Please "
                "update your scripts.\n");
    }

868 869 870 871 872 873 874
    /* Media type */
    value = qemu_opt_get(legacy_opts, "media");
    if (value) {
        if (!strcmp(value, "disk")) {
            media = MEDIA_DISK;
        } else if (!strcmp(value, "cdrom")) {
            media = MEDIA_CDROM;
F
Fam Zheng 已提交
875
            read_only = true;
876 877 878 879 880 881
        } else {
            error_report("'%s' invalid media", value);
            goto fail;
        }
    }

882
    /* copy-on-read is disabled with a warning for read-only devices */
F
Fam Zheng 已提交
883
    read_only |= qemu_opt_get_bool(legacy_opts, "read-only", false);
884 885 886 887 888 889 890 891 892 893 894 895
    copy_on_read = qemu_opt_get_bool(legacy_opts, "copy-on-read", false);

    if (read_only && copy_on_read) {
        error_report("warning: disabling copy-on-read on read-only drive");
        copy_on_read = false;
    }

    qdict_put(bs_opts, "read-only",
              qstring_from_str(read_only ? "on" : "off"));
    qdict_put(bs_opts, "copy-on-read",
              qstring_from_str(copy_on_read ? "on" :"off"));

896 897 898 899 900 901 902 903 904 905 906 907 908 909 910
    /* Controller type */
    value = qemu_opt_get(legacy_opts, "if");
    if (value) {
        for (type = 0;
             type < IF_COUNT && strcmp(value, if_name[type]);
             type++) {
        }
        if (type == IF_COUNT) {
            error_report("unsupported bus type '%s'", value);
            goto fail;
        }
    } else {
        type = block_default_type;
    }

911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942
    /* Geometry */
    cyls  = qemu_opt_get_number(legacy_opts, "cyls", 0);
    heads = qemu_opt_get_number(legacy_opts, "heads", 0);
    secs  = qemu_opt_get_number(legacy_opts, "secs", 0);

    if (cyls || heads || secs) {
        if (cyls < 1) {
            error_report("invalid physical cyls number");
            goto fail;
        }
        if (heads < 1) {
            error_report("invalid physical heads number");
            goto fail;
        }
        if (secs < 1) {
            error_report("invalid physical secs number");
            goto fail;
        }
    }

    translation = BIOS_ATA_TRANSLATION_AUTO;
    value = qemu_opt_get(legacy_opts, "trans");
    if (value != NULL) {
        if (!cyls) {
            error_report("'%s' trans must be used with cyls, heads and secs",
                         value);
            goto fail;
        }
        if (!strcmp(value, "none")) {
            translation = BIOS_ATA_TRANSLATION_NONE;
        } else if (!strcmp(value, "lba")) {
            translation = BIOS_ATA_TRANSLATION_LBA;
943 944 945 946
        } else if (!strcmp(value, "large")) {
            translation = BIOS_ATA_TRANSLATION_LARGE;
        } else if (!strcmp(value, "rechs")) {
            translation = BIOS_ATA_TRANSLATION_RECHS;
947 948 949 950 951 952 953 954 955 956 957 958 959 960 961
        } else if (!strcmp(value, "auto")) {
            translation = BIOS_ATA_TRANSLATION_AUTO;
        } else {
            error_report("'%s' invalid translation type", value);
            goto fail;
        }
    }

    if (media == MEDIA_CDROM) {
        if (cyls || secs || heads) {
            error_report("CHS can't be set with media=cdrom");
            goto fail;
        }
    }

962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000
    /* Device address specified by bus/unit or index.
     * If none was specified, try to find the first free one. */
    bus_id  = qemu_opt_get_number(legacy_opts, "bus", 0);
    unit_id = qemu_opt_get_number(legacy_opts, "unit", -1);
    index   = qemu_opt_get_number(legacy_opts, "index", -1);

    max_devs = if_max_devs[type];

    if (index != -1) {
        if (bus_id != 0 || unit_id != -1) {
            error_report("index cannot be used with bus and unit");
            goto fail;
        }
        bus_id = drive_index_to_bus_id(type, index);
        unit_id = drive_index_to_unit_id(type, index);
    }

    if (unit_id == -1) {
       unit_id = 0;
       while (drive_get(type, bus_id, unit_id) != NULL) {
           unit_id++;
           if (max_devs && unit_id >= max_devs) {
               unit_id -= max_devs;
               bus_id++;
           }
       }
    }

    if (max_devs && unit_id >= max_devs) {
        error_report("unit %d too big (max is %d)", unit_id, max_devs - 1);
        goto fail;
    }

    if (drive_get(type, bus_id, unit_id) != NULL) {
        error_report("drive with bus=%d, unit=%d (index=%d) exists",
                     bus_id, unit_id, index);
        goto fail;
    }

1001 1002 1003
    /* Serial number */
    serial = qemu_opt_get(legacy_opts, "serial");

1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021
    /* no id supplied -> create one */
    if (qemu_opts_id(all_opts) == NULL) {
        char *new_id;
        const char *mediastr = "";
        if (type == IF_IDE || type == IF_SCSI) {
            mediastr = (media == MEDIA_CDROM) ? "-cd" : "-hd";
        }
        if (max_devs) {
            new_id = g_strdup_printf("%s%i%s%i", if_name[type], bus_id,
                                     mediastr, unit_id);
        } else {
            new_id = g_strdup_printf("%s%s%i", if_name[type],
                                     mediastr, unit_id);
        }
        qdict_put(bs_opts, "id", qstring_from_str(new_id));
        g_free(new_id);
    }

1022 1023 1024 1025 1026 1027 1028 1029 1030
    /* Add virtio block device */
    devaddr = qemu_opt_get(legacy_opts, "addr");
    if (devaddr && type != IF_VIRTIO) {
        error_report("addr is not supported by this bus type");
        goto fail;
    }

    if (type == IF_VIRTIO) {
        QemuOpts *devopts;
1031 1032
        devopts = qemu_opts_create(qemu_find_opts("device"), NULL, 0,
                                   &error_abort);
1033
        if (arch_type == QEMU_ARCH_S390X) {
1034
            qemu_opt_set(devopts, "driver", "virtio-blk-ccw", &error_abort);
1035
        } else {
1036
            qemu_opt_set(devopts, "driver", "virtio-blk-pci", &error_abort);
1037
        }
1038 1039
        qemu_opt_set(devopts, "drive", qdict_get_str(bs_opts, "id"),
                     &error_abort);
1040
        if (devaddr) {
1041
            qemu_opt_set(devopts, "addr", devaddr, &error_abort);
1042 1043 1044
        }
    }

M
Max Reitz 已提交
1045 1046
    filename = qemu_opt_get(legacy_opts, "file");

1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067
    /* Check werror/rerror compatibility with if=... */
    werror = qemu_opt_get(legacy_opts, "werror");
    if (werror != NULL) {
        if (type != IF_IDE && type != IF_SCSI && type != IF_VIRTIO &&
            type != IF_NONE) {
            error_report("werror is not supported by this bus type");
            goto fail;
        }
        qdict_put(bs_opts, "werror", qstring_from_str(werror));
    }

    rerror = qemu_opt_get(legacy_opts, "rerror");
    if (rerror != NULL) {
        if (type != IF_IDE && type != IF_VIRTIO && type != IF_SCSI &&
            type != IF_NONE) {
            error_report("rerror is not supported by this bus type");
            goto fail;
        }
        qdict_put(bs_opts, "rerror", qstring_from_str(rerror));
    }

1068
    /* Actual block device init: Functionality shared with blockdev-add */
1069
    blk = blockdev_init(filename, bs_opts, &local_err);
1070
    bs_opts = NULL;
1071
    if (!blk) {
1072
        if (local_err) {
1073
            error_report_err(local_err);
1074
        }
1075
        goto fail;
1076
    } else {
1077
        assert(!local_err);
1078 1079
    }

1080 1081
    /* Create legacy DriveInfo */
    dinfo = g_malloc0(sizeof(*dinfo));
1082
    dinfo->opts = all_opts;
1083

1084 1085 1086 1087 1088
    dinfo->cyls = cyls;
    dinfo->heads = heads;
    dinfo->secs = secs;
    dinfo->trans = translation;

1089
    dinfo->type = type;
1090 1091
    dinfo->bus = bus_id;
    dinfo->unit = unit_id;
1092
    dinfo->devaddr = devaddr;
1093 1094
    dinfo->serial = g_strdup(serial);

1095 1096
    blk_set_legacy_dinfo(blk, dinfo);

1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107
    switch(type) {
    case IF_IDE:
    case IF_SCSI:
    case IF_XEN:
    case IF_NONE:
        dinfo->media_cd = media == MEDIA_CDROM;
        break;
    default:
        break;
    }

1108
fail:
1109
    qemu_opts_del(legacy_opts);
1110
    QDECREF(bs_opts);
1111
    return dinfo;
1112 1113
}

1114
void hmp_commit(Monitor *mon, const QDict *qdict)
1115 1116
{
    const char *device = qdict_get_str(qdict, "device");
1117
    BlockBackend *blk;
1118
    int ret;
1119

1120
    if (!strcmp(device, "all")) {
1121
        ret = bdrv_commit_all();
1122
    } else {
1123 1124
        blk = blk_by_name(device);
        if (!blk) {
1125
            monitor_printf(mon, "Device '%s' not found\n", device);
1126
            return;
1127
        }
M
Max Reitz 已提交
1128 1129 1130 1131
        if (!blk_is_available(blk)) {
            monitor_printf(mon, "Device '%s' has no medium\n", device);
            return;
        }
1132
        ret = bdrv_commit(blk_bs(blk));
1133 1134 1135 1136
    }
    if (ret < 0) {
        monitor_printf(mon, "'commit' error for '%s': %s\n", device,
                       strerror(-ret));
1137 1138 1139
    }
}

1140 1141
static void blockdev_do_action(int kind, void *data, Error **errp)
{
1142 1143
    TransactionAction action;
    TransactionActionList list;
1144 1145 1146 1147 1148 1149 1150 1151

    action.kind = kind;
    action.data = data;
    list.value = &action;
    list.next = NULL;
    qmp_transaction(&list, errp);
}

1152 1153 1154 1155 1156
void qmp_blockdev_snapshot_sync(bool has_device, const char *device,
                                bool has_node_name, const char *node_name,
                                const char *snapshot_file,
                                bool has_snapshot_node_name,
                                const char *snapshot_node_name,
1157
                                bool has_format, const char *format,
1158
                                bool has_mode, NewImageMode mode, Error **errp)
1159
{
1160
    BlockdevSnapshot snapshot = {
1161
        .has_device = has_device,
1162
        .device = (char *) device,
1163 1164
        .has_node_name = has_node_name,
        .node_name = (char *) node_name,
1165
        .snapshot_file = (char *) snapshot_file,
1166 1167
        .has_snapshot_node_name = has_snapshot_node_name,
        .snapshot_node_name = (char *) snapshot_node_name,
1168 1169 1170 1171 1172
        .has_format = has_format,
        .format = (char *) format,
        .has_mode = has_mode,
        .mode = mode,
    };
1173 1174
    blockdev_do_action(TRANSACTION_ACTION_KIND_BLOCKDEV_SNAPSHOT_SYNC,
                       &snapshot, errp);
1175 1176
}

1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189
void qmp_blockdev_snapshot_internal_sync(const char *device,
                                         const char *name,
                                         Error **errp)
{
    BlockdevSnapshotInternal snapshot = {
        .device = (char *) device,
        .name = (char *) name
    };

    blockdev_do_action(TRANSACTION_ACTION_KIND_BLOCKDEV_SNAPSHOT_INTERNAL_SYNC,
                       &snapshot, errp);
}

1190 1191 1192 1193 1194 1195 1196
SnapshotInfo *qmp_blockdev_snapshot_delete_internal_sync(const char *device,
                                                         bool has_id,
                                                         const char *id,
                                                         bool has_name,
                                                         const char *name,
                                                         Error **errp)
{
1197 1198
    BlockDriverState *bs;
    BlockBackend *blk;
1199
    AioContext *aio_context;
1200 1201 1202 1203 1204
    QEMUSnapshotInfo sn;
    Error *local_err = NULL;
    SnapshotInfo *info = NULL;
    int ret;

1205 1206
    blk = blk_by_name(device);
    if (!blk) {
1207 1208
        error_set(errp, ERROR_CLASS_DEVICE_NOT_FOUND,
                  "Device '%s' not found", device);
1209 1210
        return NULL;
    }
M
Max Reitz 已提交
1211 1212 1213

    aio_context = blk_get_aio_context(blk);
    aio_context_acquire(aio_context);
1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224

    if (!has_id) {
        id = NULL;
    }

    if (!has_name) {
        name = NULL;
    }

    if (!id && !name) {
        error_setg(errp, "Name or id must be provided");
M
Max Reitz 已提交
1225
        goto out_aio_context;
1226 1227
    }

M
Max Reitz 已提交
1228 1229 1230 1231 1232
    if (!blk_is_available(blk)) {
        error_setg(errp, "Device '%s' has no medium", device);
        goto out_aio_context;
    }
    bs = blk_bs(blk);
1233

1234 1235 1236 1237
    if (bdrv_op_is_blocked(bs, BLOCK_OP_TYPE_INTERNAL_SNAPSHOT_DELETE, errp)) {
        goto out_aio_context;
    }

1238
    ret = bdrv_snapshot_find_by_id_and_name(bs, id, name, &sn, &local_err);
1239
    if (local_err) {
1240
        error_propagate(errp, local_err);
1241
        goto out_aio_context;
1242 1243 1244 1245 1246 1247
    }
    if (!ret) {
        error_setg(errp,
                   "Snapshot with id '%s' and name '%s' does not exist on "
                   "device '%s'",
                   STR_OR_NULL(id), STR_OR_NULL(name), device);
1248
        goto out_aio_context;
1249 1250 1251
    }

    bdrv_snapshot_delete(bs, id, name, &local_err);
1252
    if (local_err) {
1253
        error_propagate(errp, local_err);
1254
        goto out_aio_context;
1255 1256
    }

1257 1258
    aio_context_release(aio_context);

1259
    info = g_new0(SnapshotInfo, 1);
1260 1261 1262 1263 1264 1265 1266 1267 1268
    info->id = g_strdup(sn.id_str);
    info->name = g_strdup(sn.name);
    info->date_nsec = sn.date_nsec;
    info->date_sec = sn.date_sec;
    info->vm_state_size = sn.vm_state_size;
    info->vm_clock_nsec = sn.vm_clock_nsec % 1000000000;
    info->vm_clock_sec = sn.vm_clock_nsec / 1000000000;

    return info;
1269 1270 1271 1272

out_aio_context:
    aio_context_release(aio_context);
    return NULL;
1273
}
1274

1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336
/**
 * block_dirty_bitmap_lookup:
 * Return a dirty bitmap (if present), after validating
 * the node reference and bitmap names.
 *
 * @node: The name of the BDS node to search for bitmaps
 * @name: The name of the bitmap to search for
 * @pbs: Output pointer for BDS lookup, if desired. Can be NULL.
 * @paio: Output pointer for aio_context acquisition, if desired. Can be NULL.
 * @errp: Output pointer for error information. Can be NULL.
 *
 * @return: A bitmap object on success, or NULL on failure.
 */
static BdrvDirtyBitmap *block_dirty_bitmap_lookup(const char *node,
                                                  const char *name,
                                                  BlockDriverState **pbs,
                                                  AioContext **paio,
                                                  Error **errp)
{
    BlockDriverState *bs;
    BdrvDirtyBitmap *bitmap;
    AioContext *aio_context;

    if (!node) {
        error_setg(errp, "Node cannot be NULL");
        return NULL;
    }
    if (!name) {
        error_setg(errp, "Bitmap name cannot be NULL");
        return NULL;
    }
    bs = bdrv_lookup_bs(node, node, NULL);
    if (!bs) {
        error_setg(errp, "Node '%s' not found", node);
        return NULL;
    }

    aio_context = bdrv_get_aio_context(bs);
    aio_context_acquire(aio_context);

    bitmap = bdrv_find_dirty_bitmap(bs, name);
    if (!bitmap) {
        error_setg(errp, "Dirty bitmap '%s' not found", name);
        goto fail;
    }

    if (pbs) {
        *pbs = bs;
    }
    if (paio) {
        *paio = aio_context;
    } else {
        aio_context_release(aio_context);
    }

    return bitmap;

 fail:
    aio_context_release(aio_context);
    return NULL;
}

1337
/* New and old BlockDriverState structs for atomic group operations */
1338

1339
typedef struct BlkTransactionState BlkTransactionState;
1340 1341 1342 1343 1344 1345 1346

/* Only prepare() may fail. In a single transaction, only one of commit() or
   abort() will be called, clean() will always be called if it present. */
typedef struct BdrvActionOps {
    /* Size of state struct, in bytes. */
    size_t instance_size;
    /* Prepare the work, must NOT be NULL. */
1347
    void (*prepare)(BlkTransactionState *common, Error **errp);
1348
    /* Commit the changes, can be NULL. */
1349
    void (*commit)(BlkTransactionState *common);
1350
    /* Abort the changes on fail, can be NULL. */
1351
    void (*abort)(BlkTransactionState *common);
1352
    /* Clean up resource in the end, can be NULL. */
1353
    void (*clean)(BlkTransactionState *common);
1354 1355 1356 1357 1358 1359 1360
} BdrvActionOps;

/*
 * This structure must be arranged as first member in child type, assuming
 * that compiler will also arrange it to the same address with parent instance.
 * Later it will be used in free().
 */
1361
struct BlkTransactionState {
1362
    TransactionAction *action;
1363
    const BdrvActionOps *ops;
1364
    QSIMPLEQ_ENTRY(BlkTransactionState) entry;
1365 1366
};

1367 1368 1369 1370
/* internal snapshot private data */
typedef struct InternalSnapshotState {
    BlkTransactionState common;
    BlockDriverState *bs;
1371
    AioContext *aio_context;
1372 1373 1374 1375 1376 1377
    QEMUSnapshotInfo sn;
} InternalSnapshotState;

static void internal_snapshot_prepare(BlkTransactionState *common,
                                      Error **errp)
{
1378
    Error *local_err = NULL;
1379 1380
    const char *device;
    const char *name;
1381
    BlockBackend *blk;
1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399
    BlockDriverState *bs;
    QEMUSnapshotInfo old_sn, *sn;
    bool ret;
    qemu_timeval tv;
    BlockdevSnapshotInternal *internal;
    InternalSnapshotState *state;
    int ret1;

    g_assert(common->action->kind ==
             TRANSACTION_ACTION_KIND_BLOCKDEV_SNAPSHOT_INTERNAL_SYNC);
    internal = common->action->blockdev_snapshot_internal_sync;
    state = DO_UPCAST(InternalSnapshotState, common, common);

    /* 1. parse input */
    device = internal->device;
    name = internal->name;

    /* 2. check for validation */
1400 1401
    blk = blk_by_name(device);
    if (!blk) {
1402 1403
        error_set(errp, ERROR_CLASS_DEVICE_NOT_FOUND,
                  "Device '%s' not found", device);
1404 1405 1406
        return;
    }

1407
    /* AioContext is released in .clean() */
M
Max Reitz 已提交
1408
    state->aio_context = blk_get_aio_context(blk);
1409 1410
    aio_context_acquire(state->aio_context);

M
Max Reitz 已提交
1411
    if (!blk_is_available(blk)) {
1412
        error_setg(errp, QERR_DEVICE_HAS_NO_MEDIUM, device);
1413 1414
        return;
    }
M
Max Reitz 已提交
1415
    bs = blk_bs(blk);
1416

1417 1418 1419 1420
    if (bdrv_op_is_blocked(bs, BLOCK_OP_TYPE_INTERNAL_SNAPSHOT, errp)) {
        return;
    }

1421
    if (bdrv_is_read_only(bs)) {
1422
        error_setg(errp, "Device '%s' is read only", device);
1423 1424 1425 1426
        return;
    }

    if (!bdrv_can_snapshot(bs)) {
1427 1428 1429
        error_setg(errp, "Block format '%s' used by device '%s' "
                   "does not support internal snapshots",
                   bs->drv->format_name, device);
1430 1431 1432 1433 1434 1435 1436 1437 1438
        return;
    }

    if (!strlen(name)) {
        error_setg(errp, "Name is empty");
        return;
    }

    /* check whether a snapshot with name exist */
1439 1440 1441 1442
    ret = bdrv_snapshot_find_by_id_and_name(bs, NULL, name, &old_sn,
                                            &local_err);
    if (local_err) {
        error_propagate(errp, local_err);
1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493
        return;
    } else if (ret) {
        error_setg(errp,
                   "Snapshot with name '%s' already exists on device '%s'",
                   name, device);
        return;
    }

    /* 3. take the snapshot */
    sn = &state->sn;
    pstrcpy(sn->name, sizeof(sn->name), name);
    qemu_gettimeofday(&tv);
    sn->date_sec = tv.tv_sec;
    sn->date_nsec = tv.tv_usec * 1000;
    sn->vm_clock_nsec = qemu_clock_get_ns(QEMU_CLOCK_VIRTUAL);

    ret1 = bdrv_snapshot_create(bs, sn);
    if (ret1 < 0) {
        error_setg_errno(errp, -ret1,
                         "Failed to create snapshot '%s' on device '%s'",
                         name, device);
        return;
    }

    /* 4. succeed, mark a snapshot is created */
    state->bs = bs;
}

static void internal_snapshot_abort(BlkTransactionState *common)
{
    InternalSnapshotState *state =
                             DO_UPCAST(InternalSnapshotState, common, common);
    BlockDriverState *bs = state->bs;
    QEMUSnapshotInfo *sn = &state->sn;
    Error *local_error = NULL;

    if (!bs) {
        return;
    }

    if (bdrv_snapshot_delete(bs, sn->id_str, sn->name, &local_error) < 0) {
        error_report("Failed to delete snapshot with id '%s' and name '%s' on "
                     "device '%s' in abort: %s",
                     sn->id_str,
                     sn->name,
                     bdrv_get_device_name(bs),
                     error_get_pretty(local_error));
        error_free(local_error);
    }
}

1494 1495 1496 1497 1498 1499 1500 1501 1502 1503
static void internal_snapshot_clean(BlkTransactionState *common)
{
    InternalSnapshotState *state = DO_UPCAST(InternalSnapshotState,
                                             common, common);

    if (state->aio_context) {
        aio_context_release(state->aio_context);
    }
}

1504
/* external snapshot private data */
1505 1506
typedef struct ExternalSnapshotState {
    BlkTransactionState common;
1507 1508
    BlockDriverState *old_bs;
    BlockDriverState *new_bs;
1509
    AioContext *aio_context;
1510
} ExternalSnapshotState;
1511

1512
static void external_snapshot_prepare(BlkTransactionState *common,
1513 1514 1515
                                      Error **errp)
{
    int flags, ret;
1516
    QDict *options;
1517
    Error *local_err = NULL;
1518
    bool has_device = false;
1519
    const char *device;
1520 1521 1522 1523
    bool has_node_name = false;
    const char *node_name;
    bool has_snapshot_node_name = false;
    const char *snapshot_node_name;
1524 1525 1526
    const char *new_image_file;
    const char *format = "qcow2";
    enum NewImageMode mode = NEW_IMAGE_MODE_ABSOLUTE_PATHS;
1527 1528
    ExternalSnapshotState *state =
                             DO_UPCAST(ExternalSnapshotState, common, common);
1529
    TransactionAction *action = common->action;
1530

1531
    /* get parameters */
1532
    g_assert(action->kind == TRANSACTION_ACTION_KIND_BLOCKDEV_SNAPSHOT_SYNC);
1533

1534
    has_device = action->blockdev_snapshot_sync->has_device;
1535
    device = action->blockdev_snapshot_sync->device;
1536 1537 1538 1539 1540 1541
    has_node_name = action->blockdev_snapshot_sync->has_node_name;
    node_name = action->blockdev_snapshot_sync->node_name;
    has_snapshot_node_name =
        action->blockdev_snapshot_sync->has_snapshot_node_name;
    snapshot_node_name = action->blockdev_snapshot_sync->snapshot_node_name;

1542 1543 1544 1545 1546 1547 1548 1549 1550
    new_image_file = action->blockdev_snapshot_sync->snapshot_file;
    if (action->blockdev_snapshot_sync->has_format) {
        format = action->blockdev_snapshot_sync->format;
    }
    if (action->blockdev_snapshot_sync->has_mode) {
        mode = action->blockdev_snapshot_sync->mode;
    }

    /* start processing */
1551 1552 1553
    state->old_bs = bdrv_lookup_bs(has_device ? device : NULL,
                                   has_node_name ? node_name : NULL,
                                   &local_err);
1554
    if (local_err) {
1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565
        error_propagate(errp, local_err);
        return;
    }

    if (has_node_name && !has_snapshot_node_name) {
        error_setg(errp, "New snapshot node name missing");
        return;
    }

    if (has_snapshot_node_name && bdrv_find_node(snapshot_node_name)) {
        error_setg(errp, "New snapshot node name already existing");
1566 1567 1568
        return;
    }

1569 1570 1571
    /* Acquire AioContext now so any threads operating on old_bs stop */
    state->aio_context = bdrv_get_aio_context(state->old_bs);
    aio_context_acquire(state->aio_context);
1572
    bdrv_drained_begin(state->old_bs);
1573

1574
    if (!bdrv_is_inserted(state->old_bs)) {
1575
        error_setg(errp, QERR_DEVICE_HAS_NO_MEDIUM, device);
1576 1577 1578
        return;
    }

1579 1580
    if (bdrv_op_is_blocked(state->old_bs,
                           BLOCK_OP_TYPE_EXTERNAL_SNAPSHOT, errp)) {
1581 1582 1583
        return;
    }

1584 1585
    if (!bdrv_is_read_only(state->old_bs)) {
        if (bdrv_flush(state->old_bs)) {
1586
            error_setg(errp, QERR_IO_ERROR);
1587 1588 1589 1590
            return;
        }
    }

1591
    if (!bdrv_is_first_non_filter(state->old_bs)) {
1592
        error_setg(errp, QERR_FEATURE_DISABLED, "snapshot");
1593 1594 1595
        return;
    }

1596
    flags = state->old_bs->open_flags;
1597 1598 1599 1600

    /* create new image w/backing file */
    if (mode != NEW_IMAGE_MODE_EXISTING) {
        bdrv_img_create(new_image_file, format,
1601 1602
                        state->old_bs->filename,
                        state->old_bs->drv->format_name,
1603
                        NULL, -1, flags, &local_err, false);
1604
        if (local_err) {
1605 1606 1607 1608 1609
            error_propagate(errp, local_err);
            return;
        }
    }

1610
    options = qdict_new();
1611 1612 1613 1614
    if (has_snapshot_node_name) {
        qdict_put(options, "node-name",
                  qstring_from_str(snapshot_node_name));
    }
1615
    qdict_put(options, "driver", qstring_from_str(format));
1616

1617 1618
    /* TODO Inherit bs->options or only take explicit options with an
     * extended QMP command? */
1619
    assert(state->new_bs == NULL);
1620
    ret = bdrv_open(&state->new_bs, new_image_file, NULL, options,
1621
                    flags | BDRV_O_NO_BACKING, &local_err);
1622
    /* We will manually add the backing_hd field to the bs later */
1623
    if (ret != 0) {
1624
        error_propagate(errp, local_err);
1625 1626 1627
    }
}

1628
static void external_snapshot_commit(BlkTransactionState *common)
1629
{
1630 1631
    ExternalSnapshotState *state =
                             DO_UPCAST(ExternalSnapshotState, common, common);
1632

1633 1634
    bdrv_set_aio_context(state->new_bs, state->aio_context);

1635 1636
    /* This removes our old bs and adds the new bs */
    bdrv_append(state->new_bs, state->old_bs);
1637 1638 1639
    /* We don't need (or want) to use the transactional
     * bdrv_reopen_multiple() across all the entries at once, because we
     * don't want to abort all of them if one of them fails the reopen */
1640
    bdrv_reopen(state->old_bs, state->old_bs->open_flags & ~BDRV_O_RDWR,
1641 1642 1643
                NULL);
}

1644
static void external_snapshot_abort(BlkTransactionState *common)
1645
{
1646 1647 1648
    ExternalSnapshotState *state =
                             DO_UPCAST(ExternalSnapshotState, common, common);
    if (state->new_bs) {
F
Fam Zheng 已提交
1649
        bdrv_unref(state->new_bs);
1650
    }
1651 1652 1653 1654 1655 1656
}

static void external_snapshot_clean(BlkTransactionState *common)
{
    ExternalSnapshotState *state =
                             DO_UPCAST(ExternalSnapshotState, common, common);
1657
    if (state->aio_context) {
1658
        bdrv_drained_end(state->old_bs);
1659 1660
        aio_context_release(state->aio_context);
    }
1661 1662
}

1663 1664 1665
typedef struct DriveBackupState {
    BlkTransactionState common;
    BlockDriverState *bs;
1666
    AioContext *aio_context;
1667 1668 1669 1670 1671 1672
    BlockJob *job;
} DriveBackupState;

static void drive_backup_prepare(BlkTransactionState *common, Error **errp)
{
    DriveBackupState *state = DO_UPCAST(DriveBackupState, common, common);
1673
    BlockBackend *blk;
1674 1675 1676 1677 1678 1679
    DriveBackup *backup;
    Error *local_err = NULL;

    assert(common->action->kind == TRANSACTION_ACTION_KIND_DRIVE_BACKUP);
    backup = common->action->drive_backup;

1680 1681
    blk = blk_by_name(backup->device);
    if (!blk) {
1682 1683
        error_set(errp, ERROR_CLASS_DEVICE_NOT_FOUND,
                  "Device '%s' not found", backup->device);
1684 1685 1686
        return;
    }

1687 1688 1689 1690 1691
    if (!blk_is_available(blk)) {
        error_setg(errp, QERR_DEVICE_HAS_NO_MEDIUM, backup->device);
        return;
    }

1692
    /* AioContext is released in .clean() */
M
Max Reitz 已提交
1693
    state->aio_context = blk_get_aio_context(blk);
1694
    aio_context_acquire(state->aio_context);
1695 1696
    bdrv_drained_begin(blk_bs(blk));
    state->bs = blk_bs(blk);
1697

1698 1699
    qmp_drive_backup(backup->device, backup->target,
                     backup->has_format, backup->format,
1700
                     backup->sync,
1701 1702
                     backup->has_mode, backup->mode,
                     backup->has_speed, backup->speed,
1703
                     backup->has_bitmap, backup->bitmap,
1704 1705 1706
                     backup->has_on_source_error, backup->on_source_error,
                     backup->has_on_target_error, backup->on_target_error,
                     &local_err);
1707
    if (local_err) {
1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725
        error_propagate(errp, local_err);
        return;
    }

    state->job = state->bs->job;
}

static void drive_backup_abort(BlkTransactionState *common)
{
    DriveBackupState *state = DO_UPCAST(DriveBackupState, common, common);
    BlockDriverState *bs = state->bs;

    /* Only cancel if it's the job we started */
    if (bs && bs->job && bs->job == state->job) {
        block_job_cancel_sync(bs->job);
    }
}

1726 1727 1728 1729 1730
static void drive_backup_clean(BlkTransactionState *common)
{
    DriveBackupState *state = DO_UPCAST(DriveBackupState, common, common);

    if (state->aio_context) {
1731
        bdrv_drained_end(state->bs);
1732 1733 1734 1735
        aio_context_release(state->aio_context);
    }
}

1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746
typedef struct BlockdevBackupState {
    BlkTransactionState common;
    BlockDriverState *bs;
    BlockJob *job;
    AioContext *aio_context;
} BlockdevBackupState;

static void blockdev_backup_prepare(BlkTransactionState *common, Error **errp)
{
    BlockdevBackupState *state = DO_UPCAST(BlockdevBackupState, common, common);
    BlockdevBackup *backup;
M
Max Reitz 已提交
1747
    BlockBackend *blk, *target;
1748 1749 1750 1751 1752
    Error *local_err = NULL;

    assert(common->action->kind == TRANSACTION_ACTION_KIND_BLOCKDEV_BACKUP);
    backup = common->action->blockdev_backup;

1753 1754
    blk = blk_by_name(backup->device);
    if (!blk) {
1755
        error_setg(errp, "Device '%s' not found", backup->device);
1756 1757 1758
        return;
    }

1759 1760 1761 1762 1763
    if (!blk_is_available(blk)) {
        error_setg(errp, QERR_DEVICE_HAS_NO_MEDIUM, backup->device);
        return;
    }

M
Max Reitz 已提交
1764 1765
    target = blk_by_name(backup->target);
    if (!target) {
1766
        error_setg(errp, "Device '%s' not found", backup->target);
1767 1768 1769 1770
        return;
    }

    /* AioContext is released in .clean() */
M
Max Reitz 已提交
1771 1772
    state->aio_context = blk_get_aio_context(blk);
    if (state->aio_context != blk_get_aio_context(target)) {
1773 1774 1775 1776 1777
        state->aio_context = NULL;
        error_setg(errp, "Backup between two IO threads is not implemented");
        return;
    }
    aio_context_acquire(state->aio_context);
1778 1779
    state->bs = blk_bs(blk);
    bdrv_drained_begin(state->bs);
1780 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 1808 1809 1810

    qmp_blockdev_backup(backup->device, backup->target,
                        backup->sync,
                        backup->has_speed, backup->speed,
                        backup->has_on_source_error, backup->on_source_error,
                        backup->has_on_target_error, backup->on_target_error,
                        &local_err);
    if (local_err) {
        error_propagate(errp, local_err);
        return;
    }

    state->job = state->bs->job;
}

static void blockdev_backup_abort(BlkTransactionState *common)
{
    BlockdevBackupState *state = DO_UPCAST(BlockdevBackupState, common, common);
    BlockDriverState *bs = state->bs;

    /* Only cancel if it's the job we started */
    if (bs && bs->job && bs->job == state->job) {
        block_job_cancel_sync(bs->job);
    }
}

static void blockdev_backup_clean(BlkTransactionState *common)
{
    BlockdevBackupState *state = DO_UPCAST(BlockdevBackupState, common, common);

    if (state->aio_context) {
1811
        bdrv_drained_end(state->bs);
1812 1813 1814 1815
        aio_context_release(state->aio_context);
    }
}

1816 1817 1818 1819 1820 1821 1822
static void abort_prepare(BlkTransactionState *common, Error **errp)
{
    error_setg(errp, "Transaction aborted using Abort action");
}

static void abort_commit(BlkTransactionState *common)
{
1823
    g_assert_not_reached(); /* this action never succeeds */
1824 1825
}

1826
static const BdrvActionOps actions[] = {
1827
    [TRANSACTION_ACTION_KIND_BLOCKDEV_SNAPSHOT_SYNC] = {
1828
        .instance_size = sizeof(ExternalSnapshotState),
1829 1830 1831
        .prepare  = external_snapshot_prepare,
        .commit   = external_snapshot_commit,
        .abort = external_snapshot_abort,
1832
        .clean = external_snapshot_clean,
1833
    },
1834 1835 1836 1837
    [TRANSACTION_ACTION_KIND_DRIVE_BACKUP] = {
        .instance_size = sizeof(DriveBackupState),
        .prepare = drive_backup_prepare,
        .abort = drive_backup_abort,
1838
        .clean = drive_backup_clean,
1839
    },
1840 1841 1842 1843 1844 1845
    [TRANSACTION_ACTION_KIND_BLOCKDEV_BACKUP] = {
        .instance_size = sizeof(BlockdevBackupState),
        .prepare = blockdev_backup_prepare,
        .abort = blockdev_backup_abort,
        .clean = blockdev_backup_clean,
    },
1846 1847 1848 1849 1850
    [TRANSACTION_ACTION_KIND_ABORT] = {
        .instance_size = sizeof(BlkTransactionState),
        .prepare = abort_prepare,
        .commit = abort_commit,
    },
1851 1852 1853 1854
    [TRANSACTION_ACTION_KIND_BLOCKDEV_SNAPSHOT_INTERNAL_SYNC] = {
        .instance_size = sizeof(InternalSnapshotState),
        .prepare  = internal_snapshot_prepare,
        .abort = internal_snapshot_abort,
1855
        .clean = internal_snapshot_clean,
1856
    },
1857 1858
};

1859
/*
1860 1861
 * 'Atomic' group operations.  The operations are performed as a set, and if
 * any fail then we roll back all operations in the group.
1862
 */
1863
void qmp_transaction(TransactionActionList *dev_list, Error **errp)
1864
{
1865
    TransactionActionList *dev_entry = dev_list;
1866
    BlkTransactionState *state, *next;
1867
    Error *local_err = NULL;
1868

1869
    QSIMPLEQ_HEAD(snap_bdrv_states, BlkTransactionState) snap_bdrv_states;
1870 1871
    QSIMPLEQ_INIT(&snap_bdrv_states);

1872
    /* drain all i/o before any operations */
1873 1874
    bdrv_drain_all();

1875
    /* We don't do anything in this loop that commits us to the operations */
1876
    while (NULL != dev_entry) {
1877
        TransactionAction *dev_info = NULL;
1878
        const BdrvActionOps *ops;
1879

1880 1881 1882
        dev_info = dev_entry->value;
        dev_entry = dev_entry->next;

1883 1884 1885
        assert(dev_info->kind < ARRAY_SIZE(actions));

        ops = &actions[dev_info->kind];
1886 1887
        assert(ops->instance_size > 0);

1888 1889 1890 1891
        state = g_malloc0(ops->instance_size);
        state->ops = ops;
        state->action = dev_info;
        QSIMPLEQ_INSERT_TAIL(&snap_bdrv_states, state, entry);
1892

1893
        state->ops->prepare(state, &local_err);
1894
        if (local_err) {
1895 1896
            error_propagate(errp, local_err);
            goto delete_and_fail;
1897 1898 1899
        }
    }

1900
    QSIMPLEQ_FOREACH(state, &snap_bdrv_states, entry) {
1901 1902 1903
        if (state->ops->commit) {
            state->ops->commit(state);
        }
1904 1905 1906 1907 1908 1909
    }

    /* success */
    goto exit;

delete_and_fail:
1910
    /* failure, and it is all-or-none; roll back all operations */
1911 1912 1913
    QSIMPLEQ_FOREACH(state, &snap_bdrv_states, entry) {
        if (state->ops->abort) {
            state->ops->abort(state);
1914
        }
1915 1916
    }
exit:
1917 1918 1919
    QSIMPLEQ_FOREACH_SAFE(state, &snap_bdrv_states, entry, next) {
        if (state->ops->clean) {
            state->ops->clean(state);
1920
        }
1921
        g_free(state);
1922 1923 1924 1925
    }
}


1926
static void eject_device(BlockBackend *blk, int force, Error **errp)
1927
{
1928
    BlockDriverState *bs = blk_bs(blk);
1929 1930
    AioContext *aio_context;

M
Max Reitz 已提交
1931 1932 1933 1934 1935
    if (!bs) {
        /* No medium inserted, so there is nothing to do */
        return;
    }

1936 1937
    aio_context = bdrv_get_aio_context(bs);
    aio_context_acquire(aio_context);
1938

1939
    if (bdrv_op_is_blocked(bs, BLOCK_OP_TYPE_EJECT, errp)) {
1940
        goto out;
1941
    }
1942
    if (!blk_dev_has_removable_media(blk)) {
1943 1944
        error_setg(errp, "Device '%s' is not removable",
                   bdrv_get_device_name(bs));
1945
        goto out;
1946
    }
1947

1948 1949
    if (blk_dev_is_medium_locked(blk) && !blk_dev_is_tray_open(blk)) {
        blk_dev_eject_request(blk, force);
P
Paolo Bonzini 已提交
1950
        if (!force) {
1951 1952
            error_setg(errp, "Device '%s' is locked",
                       bdrv_get_device_name(bs));
1953
            goto out;
P
Paolo Bonzini 已提交
1954
        }
1955
    }
1956

1957
    bdrv_close(bs);
1958 1959 1960

out:
    aio_context_release(aio_context);
1961 1962
}

L
Luiz Capitulino 已提交
1963
void qmp_eject(const char *device, bool has_force, bool force, Error **errp)
1964
{
1965
    BlockBackend *blk;
1966

1967 1968
    blk = blk_by_name(device);
    if (!blk) {
1969 1970
        error_set(errp, ERROR_CLASS_DEVICE_NOT_FOUND,
                  "Device '%s' not found", device);
L
Luiz Capitulino 已提交
1971
        return;
1972 1973
    }

1974
    eject_device(blk, force, errp);
1975 1976
}

1977 1978 1979
void qmp_block_passwd(bool has_device, const char *device,
                      bool has_node_name, const char *node_name,
                      const char *password, Error **errp)
1980
{
1981
    Error *local_err = NULL;
1982
    BlockDriverState *bs;
1983
    AioContext *aio_context;
1984

1985 1986 1987
    bs = bdrv_lookup_bs(has_device ? device : NULL,
                        has_node_name ? node_name : NULL,
                        &local_err);
1988
    if (local_err) {
1989
        error_propagate(errp, local_err);
L
Luiz Capitulino 已提交
1990
        return;
1991 1992
    }

1993 1994 1995
    aio_context = bdrv_get_aio_context(bs);
    aio_context_acquire(aio_context);

1996
    bdrv_add_key(bs, password, errp);
1997 1998

    aio_context_release(aio_context);
1999 2000
}

2001
/* Assumes AioContext is held */
M
Max Reitz 已提交
2002 2003
static void qmp_bdrv_open_encrypted(BlockDriverState **pbs,
                                    const char *filename,
2004
                                    int bdrv_flags, const char *format,
L
Luiz Capitulino 已提交
2005 2006
                                    const char *password, Error **errp)
{
2007
    Error *local_err = NULL;
2008
    QDict *options = NULL;
2009 2010
    int ret;

2011 2012 2013 2014 2015
    if (format) {
        options = qdict_new();
        qdict_put(options, "driver", qstring_from_str(format));
    }

M
Max Reitz 已提交
2016
    ret = bdrv_open(pbs, filename, NULL, options, bdrv_flags, &local_err);
2017
    if (ret < 0) {
2018
        error_propagate(errp, local_err);
L
Luiz Capitulino 已提交
2019 2020 2021
        return;
    }

M
Max Reitz 已提交
2022
    bdrv_add_key(*pbs, password, errp);
L
Luiz Capitulino 已提交
2023 2024 2025
}

void qmp_change_blockdev(const char *device, const char *filename,
2026
                         const char *format, Error **errp)
2027
{
2028
    BlockBackend *blk;
2029
    BlockDriverState *bs;
2030
    AioContext *aio_context;
2031
    int bdrv_flags;
M
Max Reitz 已提交
2032
    bool new_bs;
2033
    Error *err = NULL;
2034

2035 2036
    blk = blk_by_name(device);
    if (!blk) {
2037 2038
        error_set(errp, ERROR_CLASS_DEVICE_NOT_FOUND,
                  "Device '%s' not found", device);
L
Luiz Capitulino 已提交
2039
        return;
2040
    }
2041
    bs = blk_bs(blk);
M
Max Reitz 已提交
2042
    new_bs = !bs;
L
Luiz Capitulino 已提交
2043

M
Max Reitz 已提交
2044
    aio_context = blk_get_aio_context(blk);
2045 2046
    aio_context_acquire(aio_context);

2047
    eject_device(blk, 0, &err);
2048
    if (err) {
L
Luiz Capitulino 已提交
2049
        error_propagate(errp, err);
2050
        goto out;
2051
    }
L
Luiz Capitulino 已提交
2052

M
Max Reitz 已提交
2053 2054 2055 2056 2057 2058 2059 2060
    bdrv_flags = blk_is_read_only(blk) ? 0 : BDRV_O_RDWR;
    bdrv_flags |= blk_get_root_state(blk)->open_flags & ~BDRV_O_RDWR;

    qmp_bdrv_open_encrypted(&bs, filename, bdrv_flags, format, NULL, &err);
    if (err) {
        error_propagate(errp, err);
        goto out;
    }
L
Luiz Capitulino 已提交
2061

M
Max Reitz 已提交
2062 2063 2064 2065 2066 2067
    if (new_bs) {
        blk_insert_bs(blk, bs);
        /* Has been sent automatically by bdrv_open() if blk_bs(blk) was not
         * NULL */
        blk_dev_change_media_cb(blk, true);
    }
2068 2069 2070

out:
    aio_context_release(aio_context);
2071
}
2072

Z
Zhi Yong Wu 已提交
2073
/* throttling disk I/O limits */
2074
void qmp_block_set_io_throttle(const char *device, int64_t bps, int64_t bps_rd,
2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089
                               int64_t bps_wr,
                               int64_t iops,
                               int64_t iops_rd,
                               int64_t iops_wr,
                               bool has_bps_max,
                               int64_t bps_max,
                               bool has_bps_rd_max,
                               int64_t bps_rd_max,
                               bool has_bps_wr_max,
                               int64_t bps_wr_max,
                               bool has_iops_max,
                               int64_t iops_max,
                               bool has_iops_rd_max,
                               int64_t iops_rd_max,
                               bool has_iops_wr_max,
2090 2091
                               int64_t iops_wr_max,
                               bool has_iops_size,
2092 2093 2094
                               int64_t iops_size,
                               bool has_group,
                               const char *group, Error **errp)
Z
Zhi Yong Wu 已提交
2095
{
2096
    ThrottleConfig cfg;
Z
Zhi Yong Wu 已提交
2097
    BlockDriverState *bs;
2098
    BlockBackend *blk;
2099
    AioContext *aio_context;
Z
Zhi Yong Wu 已提交
2100

2101 2102
    blk = blk_by_name(device);
    if (!blk) {
2103 2104
        error_set(errp, ERROR_CLASS_DEVICE_NOT_FOUND,
                  "Device '%s' not found", device);
2105
        return;
Z
Zhi Yong Wu 已提交
2106
    }
M
Max Reitz 已提交
2107 2108 2109 2110

    aio_context = blk_get_aio_context(blk);
    aio_context_acquire(aio_context);

2111
    bs = blk_bs(blk);
M
Max Reitz 已提交
2112 2113 2114 2115
    if (!bs) {
        error_setg(errp, "Device '%s' has no medium", device);
        goto out;
    }
Z
Zhi Yong Wu 已提交
2116

2117 2118 2119 2120 2121 2122 2123 2124 2125
    memset(&cfg, 0, sizeof(cfg));
    cfg.buckets[THROTTLE_BPS_TOTAL].avg = bps;
    cfg.buckets[THROTTLE_BPS_READ].avg  = bps_rd;
    cfg.buckets[THROTTLE_BPS_WRITE].avg = bps_wr;

    cfg.buckets[THROTTLE_OPS_TOTAL].avg = iops;
    cfg.buckets[THROTTLE_OPS_READ].avg  = iops_rd;
    cfg.buckets[THROTTLE_OPS_WRITE].avg = iops_wr;

2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143
    if (has_bps_max) {
        cfg.buckets[THROTTLE_BPS_TOTAL].max = bps_max;
    }
    if (has_bps_rd_max) {
        cfg.buckets[THROTTLE_BPS_READ].max = bps_rd_max;
    }
    if (has_bps_wr_max) {
        cfg.buckets[THROTTLE_BPS_WRITE].max = bps_wr_max;
    }
    if (has_iops_max) {
        cfg.buckets[THROTTLE_OPS_TOTAL].max = iops_max;
    }
    if (has_iops_rd_max) {
        cfg.buckets[THROTTLE_OPS_READ].max = iops_rd_max;
    }
    if (has_iops_wr_max) {
        cfg.buckets[THROTTLE_OPS_WRITE].max = iops_wr_max;
    }
Z
Zhi Yong Wu 已提交
2144

2145 2146 2147
    if (has_iops_size) {
        cfg.op_size = iops_size;
    }
2148 2149

    if (!check_throttle_config(&cfg, errp)) {
M
Max Reitz 已提交
2150
        goto out;
Z
Zhi Yong Wu 已提交
2151 2152
    }

2153 2154 2155 2156 2157 2158 2159 2160 2161
    if (throttle_enabled(&cfg)) {
        /* Enable I/O limits if they're not enabled yet, otherwise
         * just update the throttling group. */
        if (!bs->io_limits_enabled) {
            bdrv_io_limits_enable(bs, has_group ? group : device);
        } else if (has_group) {
            bdrv_io_limits_update_group(bs, group);
        }
        /* Set the new throttling configuration */
2162
        bdrv_set_io_limits(bs, &cfg);
2163 2164 2165
    } else if (bs->io_limits_enabled) {
        /* If all throttling settings are set to 0, disable I/O limits */
        bdrv_io_limits_disable(bs);
Z
Zhi Yong Wu 已提交
2166
    }
2167

M
Max Reitz 已提交
2168
out:
2169
    aio_context_release(aio_context);
Z
Zhi Yong Wu 已提交
2170 2171
}

2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220
void qmp_block_dirty_bitmap_add(const char *node, const char *name,
                                bool has_granularity, uint32_t granularity,
                                Error **errp)
{
    AioContext *aio_context;
    BlockDriverState *bs;

    if (!name || name[0] == '\0') {
        error_setg(errp, "Bitmap name cannot be empty");
        return;
    }

    bs = bdrv_lookup_bs(node, node, errp);
    if (!bs) {
        return;
    }

    aio_context = bdrv_get_aio_context(bs);
    aio_context_acquire(aio_context);

    if (has_granularity) {
        if (granularity < 512 || !is_power_of_2(granularity)) {
            error_setg(errp, "Granularity must be power of 2 "
                             "and at least 512");
            goto out;
        }
    } else {
        /* Default to cluster size, if available: */
        granularity = bdrv_get_default_bitmap_granularity(bs);
    }

    bdrv_create_dirty_bitmap(bs, granularity, name, errp);

 out:
    aio_context_release(aio_context);
}

void qmp_block_dirty_bitmap_remove(const char *node, const char *name,
                                   Error **errp)
{
    AioContext *aio_context;
    BlockDriverState *bs;
    BdrvDirtyBitmap *bitmap;

    bitmap = block_dirty_bitmap_lookup(node, name, &bs, &aio_context, errp);
    if (!bitmap || !bs) {
        return;
    }

J
John Snow 已提交
2221 2222 2223 2224 2225 2226
    if (bdrv_dirty_bitmap_frozen(bitmap)) {
        error_setg(errp,
                   "Bitmap '%s' is currently frozen and cannot be removed",
                   name);
        goto out;
    }
2227
    bdrv_dirty_bitmap_make_anon(bitmap);
2228 2229
    bdrv_release_dirty_bitmap(bs, bitmap);

J
John Snow 已提交
2230
 out:
2231 2232 2233
    aio_context_release(aio_context);
}

J
John Snow 已提交
2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267
/**
 * Completely clear a bitmap, for the purposes of synchronizing a bitmap
 * immediately after a full backup operation.
 */
void qmp_block_dirty_bitmap_clear(const char *node, const char *name,
                                  Error **errp)
{
    AioContext *aio_context;
    BdrvDirtyBitmap *bitmap;
    BlockDriverState *bs;

    bitmap = block_dirty_bitmap_lookup(node, name, &bs, &aio_context, errp);
    if (!bitmap || !bs) {
        return;
    }

    if (bdrv_dirty_bitmap_frozen(bitmap)) {
        error_setg(errp,
                   "Bitmap '%s' is currently frozen and cannot be modified",
                   name);
        goto out;
    } else if (!bdrv_dirty_bitmap_enabled(bitmap)) {
        error_setg(errp,
                   "Bitmap '%s' is currently disabled and cannot be cleared",
                   name);
        goto out;
    }

    bdrv_clear_dirty_bitmap(bitmap);

 out:
    aio_context_release(aio_context);
}

2268
void hmp_drive_del(Monitor *mon, const QDict *qdict)
2269 2270
{
    const char *id = qdict_get_str(qdict, "id");
2271
    BlockBackend *blk;
2272
    BlockDriverState *bs;
2273
    AioContext *aio_context;
2274
    Error *local_err = NULL;
2275

2276 2277
    blk = blk_by_name(id);
    if (!blk) {
2278
        error_report("Device '%s' not found", id);
2279
        return;
2280
    }
2281

2282
    if (!blk_legacy_dinfo(blk)) {
2283 2284
        error_report("Deleting device added with blockdev-add"
                     " is not supported");
2285
        return;
2286 2287
    }

M
Max Reitz 已提交
2288
    aio_context = blk_get_aio_context(blk);
2289 2290
    aio_context_acquire(aio_context);

M
Max Reitz 已提交
2291 2292 2293 2294 2295 2296 2297
    bs = blk_bs(blk);
    if (bs) {
        if (bdrv_op_is_blocked(bs, BLOCK_OP_TYPE_DRIVE_DEL, &local_err)) {
            error_report_err(local_err);
            aio_context_release(aio_context);
            return;
        }
2298

M
Max Reitz 已提交
2299 2300
        bdrv_close(bs);
    }
2301

2302
    /* if we have a device attached to this BlockDriverState
2303 2304 2305 2306
     * then we need to make the drive anonymous until the device
     * can be removed.  If this is a drive with no device backing
     * then we can just get rid of the block driver state right here.
     */
2307
    if (blk_get_attached_dev(blk)) {
2308
        blk_hide_on_behalf_of_hmp_drive_del(blk);
2309
        /* Further I/O must not pause the guest */
2310 2311
        blk_set_on_error(blk, BLOCKDEV_ON_ERROR_REPORT,
                         BLOCKDEV_ON_ERROR_REPORT);
2312
    } else {
2313
        blk_unref(blk);
2314 2315
    }

2316
    aio_context_release(aio_context);
2317
}
2318

2319 2320 2321
void qmp_block_resize(bool has_device, const char *device,
                      bool has_node_name, const char *node_name,
                      int64_t size, Error **errp)
2322
{
2323
    Error *local_err = NULL;
2324
    BlockDriverState *bs;
2325
    AioContext *aio_context;
2326
    int ret;
2327

2328 2329 2330
    bs = bdrv_lookup_bs(has_device ? device : NULL,
                        has_node_name ? node_name : NULL,
                        &local_err);
2331
    if (local_err) {
2332 2333 2334 2335
        error_propagate(errp, local_err);
        return;
    }

2336 2337 2338
    aio_context = bdrv_get_aio_context(bs);
    aio_context_acquire(aio_context);

2339
    if (!bdrv_is_first_non_filter(bs)) {
2340
        error_setg(errp, QERR_FEATURE_DISABLED, "resize");
2341
        goto out;
2342 2343 2344
    }

    if (size < 0) {
2345
        error_setg(errp, QERR_INVALID_PARAMETER_VALUE, "size", "a >0 size");
2346
        goto out;
2347 2348
    }

2349
    if (bdrv_op_is_blocked(bs, BLOCK_OP_TYPE_RESIZE, NULL)) {
2350
        error_setg(errp, QERR_DEVICE_IN_USE, device);
2351
        goto out;
2352 2353
    }

2354 2355 2356
    /* complete all in-flight operations before resizing the device */
    bdrv_drain_all();

2357 2358
    ret = bdrv_truncate(bs, size);
    switch (ret) {
2359 2360 2361
    case 0:
        break;
    case -ENOMEDIUM:
2362
        error_setg(errp, QERR_DEVICE_HAS_NO_MEDIUM, device);
2363 2364
        break;
    case -ENOTSUP:
2365
        error_setg(errp, QERR_UNSUPPORTED);
2366 2367
        break;
    case -EACCES:
2368
        error_setg(errp, "Device '%s' is read only", device);
2369 2370
        break;
    case -EBUSY:
2371
        error_setg(errp, QERR_DEVICE_IN_USE, device);
2372 2373
        break;
    default:
2374
        error_setg_errno(errp, -ret, "Could not resize");
2375
        break;
2376
    }
2377 2378 2379

out:
    aio_context_release(aio_context);
2380
}
S
Stefan Hajnoczi 已提交
2381

2382
static void block_job_cb(void *opaque, int ret)
S
Stefan Hajnoczi 已提交
2383
{
2384 2385 2386 2387 2388
    /* Note that this function may be executed from another AioContext besides
     * the QEMU main loop.  If you need to access anything that assumes the
     * QEMU global mutex, use a BH or introduce a mutex.
     */

S
Stefan Hajnoczi 已提交
2389
    BlockDriverState *bs = opaque;
2390
    const char *msg = NULL;
S
Stefan Hajnoczi 已提交
2391

2392
    trace_block_job_cb(bs, bs->job, ret);
S
Stefan Hajnoczi 已提交
2393 2394

    assert(bs->job);
2395

S
Stefan Hajnoczi 已提交
2396
    if (ret < 0) {
2397
        msg = strerror(-ret);
S
Stefan Hajnoczi 已提交
2398 2399
    }

2400
    if (block_job_is_cancelled(bs->job)) {
2401
        block_job_event_cancelled(bs->job);
2402
    } else {
2403
        block_job_event_completed(bs->job, msg);
2404
    }
2405

F
Fam Zheng 已提交
2406
    bdrv_put_ref_bh_schedule(bs);
S
Stefan Hajnoczi 已提交
2407 2408
}

2409 2410 2411 2412
void qmp_block_stream(const char *device,
                      bool has_base, const char *base,
                      bool has_backing_file, const char *backing_file,
                      bool has_speed, int64_t speed,
P
Paolo Bonzini 已提交
2413 2414
                      bool has_on_error, BlockdevOnError on_error,
                      Error **errp)
S
Stefan Hajnoczi 已提交
2415
{
2416
    BlockBackend *blk;
S
Stefan Hajnoczi 已提交
2417
    BlockDriverState *bs;
2418
    BlockDriverState *base_bs = NULL;
2419
    AioContext *aio_context;
2420
    Error *local_err = NULL;
2421
    const char *base_name = NULL;
S
Stefan Hajnoczi 已提交
2422

P
Paolo Bonzini 已提交
2423 2424 2425 2426
    if (!has_on_error) {
        on_error = BLOCKDEV_ON_ERROR_REPORT;
    }

2427 2428
    blk = blk_by_name(device);
    if (!blk) {
2429 2430
        error_set(errp, ERROR_CLASS_DEVICE_NOT_FOUND,
                  "Device '%s' not found", device);
S
Stefan Hajnoczi 已提交
2431 2432 2433
        return;
    }

M
Max Reitz 已提交
2434
    aio_context = blk_get_aio_context(blk);
2435 2436
    aio_context_acquire(aio_context);

M
Max Reitz 已提交
2437 2438 2439 2440 2441 2442
    if (!blk_is_available(blk)) {
        error_setg(errp, "Device '%s' has no medium", device);
        goto out;
    }
    bs = blk_bs(blk);

2443
    if (bdrv_op_is_blocked(bs, BLOCK_OP_TYPE_STREAM, errp)) {
2444
        goto out;
2445 2446
    }

2447
    if (has_base) {
2448 2449
        base_bs = bdrv_find_backing_image(bs, base);
        if (base_bs == NULL) {
2450
            error_setg(errp, QERR_BASE_NOT_FOUND, base);
2451
            goto out;
2452
        }
2453
        assert(bdrv_get_aio_context(base_bs) == aio_context);
2454
        base_name = base;
S
Stefan Hajnoczi 已提交
2455 2456
    }

2457 2458 2459 2460 2461
    /* if we are streaming the entire chain, the result will have no backing
     * file, and specifying one is therefore an error */
    if (base_bs == NULL && has_backing_file) {
        error_setg(errp, "backing file specified, but streaming the "
                         "entire chain");
2462
        goto out;
2463 2464 2465 2466 2467 2468
    }

    /* backing_file string overrides base bs filename */
    base_name = has_backing_file ? backing_file : base_name;

    stream_start(bs, base_bs, base_name, has_speed ? speed : 0,
P
Paolo Bonzini 已提交
2469
                 on_error, block_job_cb, bs, &local_err);
2470
    if (local_err) {
2471
        error_propagate(errp, local_err);
2472
        goto out;
S
Stefan Hajnoczi 已提交
2473 2474 2475
    }

    trace_qmp_block_stream(bs, bs->job);
2476 2477 2478

out:
    aio_context_release(aio_context);
S
Stefan Hajnoczi 已提交
2479
}
2480

2481
void qmp_block_commit(const char *device,
2482 2483
                      bool has_base, const char *base,
                      bool has_top, const char *top,
2484
                      bool has_backing_file, const char *backing_file,
2485 2486 2487
                      bool has_speed, int64_t speed,
                      Error **errp)
{
2488
    BlockBackend *blk;
2489 2490
    BlockDriverState *bs;
    BlockDriverState *base_bs, *top_bs;
2491
    AioContext *aio_context;
2492 2493 2494 2495
    Error *local_err = NULL;
    /* This will be part of the QMP command, if/when the
     * BlockdevOnError change for blkmirror makes it in
     */
2496
    BlockdevOnError on_error = BLOCKDEV_ON_ERROR_REPORT;
2497

2498 2499 2500 2501
    if (!has_speed) {
        speed = 0;
    }

2502 2503 2504 2505 2506
    /* Important Note:
     *  libvirt relies on the DeviceNotFound error class in order to probe for
     *  live commit feature versions; for this to work, we must make sure to
     *  perform the device lookup before any generic errors that may occur in a
     *  scenario in which all optional arguments are omitted. */
2507 2508
    blk = blk_by_name(device);
    if (!blk) {
2509 2510
        error_set(errp, ERROR_CLASS_DEVICE_NOT_FOUND,
                  "Device '%s' not found", device);
2511
        return;
2512 2513
    }

M
Max Reitz 已提交
2514
    aio_context = blk_get_aio_context(blk);
2515 2516
    aio_context_acquire(aio_context);

M
Max Reitz 已提交
2517 2518 2519 2520 2521 2522
    if (!blk_is_available(blk)) {
        error_setg(errp, "Device '%s' has no medium", device);
        goto out;
    }
    bs = blk_bs(blk);

2523
    if (bdrv_op_is_blocked(bs, BLOCK_OP_TYPE_COMMIT_SOURCE, errp)) {
2524
        goto out;
2525 2526 2527 2528 2529
    }

    /* default top_bs is the active layer */
    top_bs = bs;

2530
    if (has_top && top) {
2531 2532 2533 2534 2535 2536 2537
        if (strcmp(bs->filename, top) != 0) {
            top_bs = bdrv_find_backing_image(bs, top);
        }
    }

    if (top_bs == NULL) {
        error_setg(errp, "Top image file %s not found", top ? top : "NULL");
2538
        goto out;
2539 2540
    }

2541 2542
    assert(bdrv_get_aio_context(top_bs) == aio_context);

2543 2544 2545 2546 2547 2548 2549
    if (has_base && base) {
        base_bs = bdrv_find_backing_image(top_bs, base);
    } else {
        base_bs = bdrv_find_base(top_bs);
    }

    if (base_bs == NULL) {
2550
        error_setg(errp, QERR_BASE_NOT_FOUND, base ? base : "NULL");
2551
        goto out;
2552 2553
    }

2554 2555
    assert(bdrv_get_aio_context(base_bs) == aio_context);

2556 2557 2558 2559
    if (bdrv_op_is_blocked(base_bs, BLOCK_OP_TYPE_COMMIT_TARGET, errp)) {
        goto out;
    }

2560 2561 2562
    /* Do not allow attempts to commit an image into itself */
    if (top_bs == base_bs) {
        error_setg(errp, "cannot commit an image into itself");
2563
        goto out;
2564 2565
    }

F
Fam Zheng 已提交
2566
    if (top_bs == bs) {
2567 2568 2569
        if (has_backing_file) {
            error_setg(errp, "'backing-file' specified,"
                             " but 'top' is the active layer");
2570
            goto out;
2571
        }
F
Fam Zheng 已提交
2572 2573 2574 2575
        commit_active_start(bs, base_bs, speed, on_error, block_job_cb,
                            bs, &local_err);
    } else {
        commit_start(bs, base_bs, top_bs, speed, on_error, block_job_cb, bs,
2576
                     has_backing_file ? backing_file : NULL, &local_err);
F
Fam Zheng 已提交
2577
    }
2578 2579
    if (local_err != NULL) {
        error_propagate(errp, local_err);
2580
        goto out;
2581
    }
2582 2583 2584

out:
    aio_context_release(aio_context);
2585 2586
}

2587 2588
void qmp_drive_backup(const char *device, const char *target,
                      bool has_format, const char *format,
2589
                      enum MirrorSyncMode sync,
2590 2591
                      bool has_mode, enum NewImageMode mode,
                      bool has_speed, int64_t speed,
2592
                      bool has_bitmap, const char *bitmap,
2593 2594 2595 2596
                      bool has_on_source_error, BlockdevOnError on_source_error,
                      bool has_on_target_error, BlockdevOnError on_target_error,
                      Error **errp)
{
2597
    BlockBackend *blk;
2598 2599
    BlockDriverState *bs;
    BlockDriverState *target_bs;
2600
    BlockDriverState *source = NULL;
2601
    BdrvDirtyBitmap *bmap = NULL;
2602
    AioContext *aio_context;
2603
    QDict *options = NULL;
2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621
    Error *local_err = NULL;
    int flags;
    int64_t size;
    int ret;

    if (!has_speed) {
        speed = 0;
    }
    if (!has_on_source_error) {
        on_source_error = BLOCKDEV_ON_ERROR_REPORT;
    }
    if (!has_on_target_error) {
        on_target_error = BLOCKDEV_ON_ERROR_REPORT;
    }
    if (!has_mode) {
        mode = NEW_IMAGE_MODE_ABSOLUTE_PATHS;
    }

2622 2623
    blk = blk_by_name(device);
    if (!blk) {
2624 2625
        error_set(errp, ERROR_CLASS_DEVICE_NOT_FOUND,
                  "Device '%s' not found", device);
2626 2627 2628
        return;
    }

M
Max Reitz 已提交
2629
    aio_context = blk_get_aio_context(blk);
2630 2631
    aio_context_acquire(aio_context);

F
Fam Zheng 已提交
2632 2633
    /* Although backup_run has this check too, we need to use bs->drv below, so
     * do an early check redundantly. */
M
Max Reitz 已提交
2634
    if (!blk_is_available(blk)) {
2635
        error_setg(errp, QERR_DEVICE_HAS_NO_MEDIUM, device);
2636
        goto out;
2637
    }
M
Max Reitz 已提交
2638
    bs = blk_bs(blk);
2639 2640 2641 2642 2643

    if (!has_format) {
        format = mode == NEW_IMAGE_MODE_EXISTING ? NULL : bs->drv->format_name;
    }

F
Fam Zheng 已提交
2644
    /* Early check to avoid creating target */
2645
    if (bdrv_op_is_blocked(bs, BLOCK_OP_TYPE_BACKUP_SOURCE, errp)) {
2646
        goto out;
2647 2648 2649 2650
    }

    flags = bs->open_flags | BDRV_O_RDWR;

2651 2652 2653
    /* See if we have a backing HD we can use to create our new image
     * on top of. */
    if (sync == MIRROR_SYNC_MODE_TOP) {
2654
        source = backing_bs(bs);
2655 2656 2657 2658 2659 2660 2661 2662
        if (!source) {
            sync = MIRROR_SYNC_MODE_FULL;
        }
    }
    if (sync == MIRROR_SYNC_MODE_NONE) {
        source = bs;
    }

2663 2664 2665
    size = bdrv_getlength(bs);
    if (size < 0) {
        error_setg_errno(errp, -size, "bdrv_getlength failed");
2666
        goto out;
2667 2668 2669
    }

    if (mode != NEW_IMAGE_MODE_EXISTING) {
2670
        assert(format);
2671 2672 2673 2674 2675 2676 2677 2678
        if (source) {
            bdrv_img_create(target, format, source->filename,
                            source->drv->format_name, NULL,
                            size, flags, &local_err, false);
        } else {
            bdrv_img_create(target, format, NULL, NULL, NULL,
                            size, flags, &local_err, false);
        }
2679 2680
    }

2681
    if (local_err) {
2682
        error_propagate(errp, local_err);
2683
        goto out;
2684 2685
    }

2686 2687 2688 2689 2690
    if (format) {
        options = qdict_new();
        qdict_put(options, "driver", qstring_from_str(format));
    }

2691
    target_bs = NULL;
2692
    ret = bdrv_open(&target_bs, target, NULL, options, flags, &local_err);
2693
    if (ret < 0) {
2694
        error_propagate(errp, local_err);
2695
        goto out;
2696 2697
    }

2698 2699
    bdrv_set_aio_context(target_bs, aio_context);

2700 2701 2702 2703 2704 2705 2706 2707 2708 2709
    if (has_bitmap) {
        bmap = bdrv_find_dirty_bitmap(bs, bitmap);
        if (!bmap) {
            error_setg(errp, "Bitmap '%s' could not be found", bitmap);
            goto out;
        }
    }

    backup_start(bs, target_bs, speed, sync, bmap,
                 on_source_error, on_target_error,
2710 2711
                 block_job_cb, bs, &local_err);
    if (local_err != NULL) {
F
Fam Zheng 已提交
2712
        bdrv_unref(target_bs);
2713
        error_propagate(errp, local_err);
2714
        goto out;
2715
    }
2716 2717 2718

out:
    aio_context_release(aio_context);
2719 2720
}

2721 2722
BlockDeviceInfoList *qmp_query_named_block_nodes(Error **errp)
{
2723
    return bdrv_named_nodes_list(errp);
2724 2725
}

F
Fam Zheng 已提交
2726 2727 2728 2729 2730 2731 2732 2733 2734
void qmp_blockdev_backup(const char *device, const char *target,
                         enum MirrorSyncMode sync,
                         bool has_speed, int64_t speed,
                         bool has_on_source_error,
                         BlockdevOnError on_source_error,
                         bool has_on_target_error,
                         BlockdevOnError on_target_error,
                         Error **errp)
{
M
Max Reitz 已提交
2735
    BlockBackend *blk, *target_blk;
F
Fam Zheng 已提交
2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750
    BlockDriverState *bs;
    BlockDriverState *target_bs;
    Error *local_err = NULL;
    AioContext *aio_context;

    if (!has_speed) {
        speed = 0;
    }
    if (!has_on_source_error) {
        on_source_error = BLOCKDEV_ON_ERROR_REPORT;
    }
    if (!has_on_target_error) {
        on_target_error = BLOCKDEV_ON_ERROR_REPORT;
    }

2751 2752
    blk = blk_by_name(device);
    if (!blk) {
2753
        error_setg(errp, "Device '%s' not found", device);
F
Fam Zheng 已提交
2754 2755 2756
        return;
    }

M
Max Reitz 已提交
2757
    aio_context = blk_get_aio_context(blk);
F
Fam Zheng 已提交
2758 2759
    aio_context_acquire(aio_context);

M
Max Reitz 已提交
2760 2761 2762 2763 2764 2765 2766 2767
    if (!blk_is_available(blk)) {
        error_setg(errp, "Device '%s' has no medium", device);
        goto out;
    }
    bs = blk_bs(blk);

    target_blk = blk_by_name(target);
    if (!target_blk) {
2768
        error_setg(errp, "Device '%s' not found", target);
F
Fam Zheng 已提交
2769 2770
        goto out;
    }
M
Max Reitz 已提交
2771 2772 2773 2774 2775 2776

    if (!blk_is_available(target_blk)) {
        error_setg(errp, "Device '%s' has no medium", target);
        goto out;
    }
    target_bs = blk_bs(target_blk);
F
Fam Zheng 已提交
2777 2778 2779

    bdrv_ref(target_bs);
    bdrv_set_aio_context(target_bs, aio_context);
2780 2781
    backup_start(bs, target_bs, speed, sync, NULL, on_source_error,
                 on_target_error, block_job_cb, bs, &local_err);
F
Fam Zheng 已提交
2782 2783 2784 2785 2786 2787 2788 2789
    if (local_err != NULL) {
        bdrv_unref(target_bs);
        error_propagate(errp, local_err);
    }
out:
    aio_context_release(aio_context);
}

P
Paolo Bonzini 已提交
2790 2791
void qmp_drive_mirror(const char *device, const char *target,
                      bool has_format, const char *format,
2792
                      bool has_node_name, const char *node_name,
2793
                      bool has_replaces, const char *replaces,
P
Paolo Bonzini 已提交
2794 2795
                      enum MirrorSyncMode sync,
                      bool has_mode, enum NewImageMode mode,
2796
                      bool has_speed, int64_t speed,
2797
                      bool has_granularity, uint32_t granularity,
2798
                      bool has_buf_size, int64_t buf_size,
2799 2800
                      bool has_on_source_error, BlockdevOnError on_source_error,
                      bool has_on_target_error, BlockdevOnError on_target_error,
2801
                      bool has_unmap, bool unmap,
2802
                      Error **errp)
P
Paolo Bonzini 已提交
2803
{
2804
    BlockBackend *blk;
P
Paolo Bonzini 已提交
2805 2806
    BlockDriverState *bs;
    BlockDriverState *source, *target_bs;
2807
    AioContext *aio_context;
P
Paolo Bonzini 已提交
2808
    Error *local_err = NULL;
2809
    QDict *options;
P
Paolo Bonzini 已提交
2810
    int flags;
2811
    int64_t size;
P
Paolo Bonzini 已提交
2812 2813 2814 2815 2816
    int ret;

    if (!has_speed) {
        speed = 0;
    }
2817 2818 2819 2820 2821 2822
    if (!has_on_source_error) {
        on_source_error = BLOCKDEV_ON_ERROR_REPORT;
    }
    if (!has_on_target_error) {
        on_target_error = BLOCKDEV_ON_ERROR_REPORT;
    }
P
Paolo Bonzini 已提交
2823 2824 2825
    if (!has_mode) {
        mode = NEW_IMAGE_MODE_ABSOLUTE_PATHS;
    }
2826 2827 2828
    if (!has_granularity) {
        granularity = 0;
    }
2829
    if (!has_buf_size) {
W
Wen Congyang 已提交
2830
        buf_size = 0;
2831
    }
2832 2833 2834
    if (!has_unmap) {
        unmap = true;
    }
2835

2836
    if (granularity != 0 && (granularity < 512 || granularity > 1048576 * 64)) {
2837 2838
        error_setg(errp, QERR_INVALID_PARAMETER_VALUE, "granularity",
                   "a value in range [512B, 64MB]");
2839 2840 2841
        return;
    }
    if (granularity & (granularity - 1)) {
2842 2843
        error_setg(errp, QERR_INVALID_PARAMETER_VALUE, "granularity",
                   "power of 2");
2844 2845
        return;
    }
P
Paolo Bonzini 已提交
2846

2847 2848
    blk = blk_by_name(device);
    if (!blk) {
2849 2850
        error_set(errp, ERROR_CLASS_DEVICE_NOT_FOUND,
                  "Device '%s' not found", device);
P
Paolo Bonzini 已提交
2851 2852 2853
        return;
    }

M
Max Reitz 已提交
2854
    aio_context = blk_get_aio_context(blk);
2855 2856
    aio_context_acquire(aio_context);

M
Max Reitz 已提交
2857
    if (!blk_is_available(blk)) {
2858
        error_setg(errp, QERR_DEVICE_HAS_NO_MEDIUM, device);
2859
        goto out;
P
Paolo Bonzini 已提交
2860
    }
M
Max Reitz 已提交
2861
    bs = blk_bs(blk);
P
Paolo Bonzini 已提交
2862 2863 2864 2865 2866

    if (!has_format) {
        format = mode == NEW_IMAGE_MODE_EXISTING ? NULL : bs->drv->format_name;
    }

2867
    if (bdrv_op_is_blocked(bs, BLOCK_OP_TYPE_MIRROR, errp)) {
2868
        goto out;
P
Paolo Bonzini 已提交
2869 2870 2871
    }

    flags = bs->open_flags | BDRV_O_RDWR;
2872
    source = backing_bs(bs);
P
Paolo Bonzini 已提交
2873 2874 2875
    if (!source && sync == MIRROR_SYNC_MODE_TOP) {
        sync = MIRROR_SYNC_MODE_FULL;
    }
2876 2877 2878
    if (sync == MIRROR_SYNC_MODE_NONE) {
        source = bs;
    }
P
Paolo Bonzini 已提交
2879

2880 2881 2882
    size = bdrv_getlength(bs);
    if (size < 0) {
        error_setg_errno(errp, -size, "bdrv_getlength failed");
2883
        goto out;
2884 2885
    }

2886 2887
    if (has_replaces) {
        BlockDriverState *to_replace_bs;
2888 2889
        AioContext *replace_aio_context;
        int64_t replace_size;
2890 2891 2892 2893

        if (!has_node_name) {
            error_setg(errp, "a node-name must be provided when replacing a"
                             " named node of the graph");
2894
            goto out;
2895 2896
        }

2897
        to_replace_bs = check_to_replace_node(bs, replaces, &local_err);
2898 2899 2900

        if (!to_replace_bs) {
            error_propagate(errp, local_err);
2901
            goto out;
2902 2903
        }

2904 2905 2906 2907 2908 2909
        replace_aio_context = bdrv_get_aio_context(to_replace_bs);
        aio_context_acquire(replace_aio_context);
        replace_size = bdrv_getlength(to_replace_bs);
        aio_context_release(replace_aio_context);

        if (size != replace_size) {
2910 2911
            error_setg(errp, "cannot replace image with a mirror image of "
                             "different size");
2912
            goto out;
2913 2914 2915
        }
    }

2916 2917 2918
    if ((sync == MIRROR_SYNC_MODE_FULL || !source)
        && mode != NEW_IMAGE_MODE_EXISTING)
    {
P
Paolo Bonzini 已提交
2919
        /* create new image w/o backing file */
2920
        assert(format);
2921
        bdrv_img_create(target, format,
2922
                        NULL, NULL, NULL, size, flags, &local_err, false);
P
Paolo Bonzini 已提交
2923 2924 2925 2926 2927 2928
    } else {
        switch (mode) {
        case NEW_IMAGE_MODE_EXISTING:
            break;
        case NEW_IMAGE_MODE_ABSOLUTE_PATHS:
            /* create new image with backing file */
2929 2930 2931
            bdrv_img_create(target, format,
                            source->filename,
                            source->drv->format_name,
2932
                            NULL, size, flags, &local_err, false);
P
Paolo Bonzini 已提交
2933 2934 2935 2936 2937 2938
            break;
        default:
            abort();
        }
    }

2939
    if (local_err) {
2940
        error_propagate(errp, local_err);
2941
        goto out;
P
Paolo Bonzini 已提交
2942 2943
    }

2944
    options = qdict_new();
2945 2946 2947
    if (has_node_name) {
        qdict_put(options, "node-name", qstring_from_str(node_name));
    }
2948 2949 2950
    if (format) {
        qdict_put(options, "driver", qstring_from_str(format));
    }
2951

2952 2953 2954
    /* Mirroring takes care of copy-on-write using the source's backing
     * file.
     */
2955
    target_bs = NULL;
2956
    ret = bdrv_open(&target_bs, target, NULL, options,
2957
                    flags | BDRV_O_NO_BACKING, &local_err);
P
Paolo Bonzini 已提交
2958
    if (ret < 0) {
2959
        error_propagate(errp, local_err);
2960
        goto out;
P
Paolo Bonzini 已提交
2961 2962
    }

2963 2964
    bdrv_set_aio_context(target_bs, aio_context);

2965 2966 2967 2968 2969 2970
    /* pass the node name to replace to mirror start since it's loose coupling
     * and will allow to check whether the node still exist at mirror completion
     */
    mirror_start(bs, target_bs,
                 has_replaces ? replaces : NULL,
                 speed, granularity, buf_size, sync,
2971
                 on_source_error, on_target_error,
2972
                 unmap,
2973
                 block_job_cb, bs, &local_err);
P
Paolo Bonzini 已提交
2974
    if (local_err != NULL) {
F
Fam Zheng 已提交
2975
        bdrv_unref(target_bs);
P
Paolo Bonzini 已提交
2976
        error_propagate(errp, local_err);
2977
        goto out;
P
Paolo Bonzini 已提交
2978
    }
2979 2980 2981

out:
    aio_context_release(aio_context);
P
Paolo Bonzini 已提交
2982 2983
}

2984
/* Get the block job for a given device name and acquire its AioContext */
2985 2986
static BlockJob *find_block_job(const char *device, AioContext **aio_context,
                                Error **errp)
2987
{
2988
    BlockBackend *blk;
2989 2990
    BlockDriverState *bs;

M
Max Reitz 已提交
2991 2992
    *aio_context = NULL;

2993 2994
    blk = blk_by_name(device);
    if (!blk) {
2995 2996 2997
        goto notfound;
    }

M
Max Reitz 已提交
2998
    *aio_context = blk_get_aio_context(blk);
2999 3000
    aio_context_acquire(*aio_context);

M
Max Reitz 已提交
3001 3002 3003 3004 3005
    if (!blk_is_available(blk)) {
        goto notfound;
    }
    bs = blk_bs(blk);

3006 3007
    if (!bs->job) {
        goto notfound;
3008
    }
3009

3010
    return bs->job;
3011 3012

notfound:
3013 3014
    error_set(errp, ERROR_CLASS_DEVICE_NOT_ACTIVE,
              "No active block job on device '%s'", device);
M
Max Reitz 已提交
3015 3016 3017 3018
    if (*aio_context) {
        aio_context_release(*aio_context);
        *aio_context = NULL;
    }
3019
    return NULL;
3020 3021
}

3022
void qmp_block_job_set_speed(const char *device, int64_t speed, Error **errp)
3023
{
3024
    AioContext *aio_context;
3025
    BlockJob *job = find_block_job(device, &aio_context, errp);
3026 3027 3028 3029 3030

    if (!job) {
        return;
    }

3031
    block_job_set_speed(job, speed, errp);
3032
    aio_context_release(aio_context);
3033
}
3034

3035 3036
void qmp_block_job_cancel(const char *device,
                          bool has_force, bool force, Error **errp)
3037
{
3038
    AioContext *aio_context;
3039
    BlockJob *job = find_block_job(device, &aio_context, errp);
3040

3041 3042 3043
    if (!job) {
        return;
    }
3044 3045 3046 3047 3048

    if (!has_force) {
        force = false;
    }

F
Fam Zheng 已提交
3049
    if (job->user_paused && !force) {
3050 3051
        error_setg(errp, "The block job for device '%s' is currently paused",
                   device);
3052
        goto out;
3053
    }
3054 3055 3056

    trace_qmp_block_job_cancel(job);
    block_job_cancel(job);
3057 3058
out:
    aio_context_release(aio_context);
3059
}
S
Stefan Hajnoczi 已提交
3060

3061 3062
void qmp_block_job_pause(const char *device, Error **errp)
{
3063
    AioContext *aio_context;
3064
    BlockJob *job = find_block_job(device, &aio_context, errp);
3065

F
Fam Zheng 已提交
3066
    if (!job || job->user_paused) {
3067 3068 3069
        return;
    }

F
Fam Zheng 已提交
3070
    job->user_paused = true;
3071 3072
    trace_qmp_block_job_pause(job);
    block_job_pause(job);
3073
    aio_context_release(aio_context);
3074 3075 3076 3077
}

void qmp_block_job_resume(const char *device, Error **errp)
{
3078
    AioContext *aio_context;
3079
    BlockJob *job = find_block_job(device, &aio_context, errp);
3080

F
Fam Zheng 已提交
3081
    if (!job || !job->user_paused) {
3082 3083 3084
        return;
    }

F
Fam Zheng 已提交
3085
    job->user_paused = false;
3086 3087
    trace_qmp_block_job_resume(job);
    block_job_resume(job);
3088
    aio_context_release(aio_context);
3089 3090
}

P
Paolo Bonzini 已提交
3091 3092
void qmp_block_job_complete(const char *device, Error **errp)
{
3093
    AioContext *aio_context;
3094
    BlockJob *job = find_block_job(device, &aio_context, errp);
P
Paolo Bonzini 已提交
3095 3096 3097 3098 3099 3100 3101

    if (!job) {
        return;
    }

    trace_qmp_block_job_complete(job);
    block_job_complete(job, errp);
3102
    aio_context_release(aio_context);
P
Paolo Bonzini 已提交
3103 3104
}

3105 3106 3107 3108 3109
void qmp_change_backing_file(const char *device,
                             const char *image_node_name,
                             const char *backing_file,
                             Error **errp)
{
3110
    BlockBackend *blk;
3111
    BlockDriverState *bs = NULL;
3112
    AioContext *aio_context;
3113 3114 3115 3116 3117 3118
    BlockDriverState *image_bs = NULL;
    Error *local_err = NULL;
    bool ro;
    int open_flags;
    int ret;

3119 3120
    blk = blk_by_name(device);
    if (!blk) {
3121 3122
        error_set(errp, ERROR_CLASS_DEVICE_NOT_FOUND,
                  "Device '%s' not found", device);
3123 3124 3125
        return;
    }

M
Max Reitz 已提交
3126
    aio_context = blk_get_aio_context(blk);
3127 3128
    aio_context_acquire(aio_context);

M
Max Reitz 已提交
3129 3130 3131 3132 3133 3134
    if (!blk_is_available(blk)) {
        error_setg(errp, "Device '%s' has no medium", device);
        goto out;
    }
    bs = blk_bs(blk);

3135 3136 3137
    image_bs = bdrv_lookup_bs(NULL, image_node_name, &local_err);
    if (local_err) {
        error_propagate(errp, local_err);
3138
        goto out;
3139 3140 3141 3142
    }

    if (!image_bs) {
        error_setg(errp, "image file not found");
3143
        goto out;
3144 3145 3146 3147 3148
    }

    if (bdrv_find_base(image_bs) == image_bs) {
        error_setg(errp, "not allowing backing file change on an image "
                         "without a backing file");
3149
        goto out;
3150 3151 3152 3153 3154
    }

    /* even though we are not necessarily operating on bs, we need it to
     * determine if block ops are currently prohibited on the chain */
    if (bdrv_op_is_blocked(bs, BLOCK_OP_TYPE_CHANGE, errp)) {
3155
        goto out;
3156 3157 3158 3159 3160 3161
    }

    /* final sanity check */
    if (!bdrv_chain_contains(bs, image_bs)) {
        error_setg(errp, "'%s' and image file are not in the same chain",
                   device);
3162
        goto out;
3163 3164 3165 3166 3167 3168 3169 3170 3171 3172
    }

    /* if not r/w, reopen to make r/w */
    open_flags = image_bs->open_flags;
    ro = bdrv_is_read_only(image_bs);

    if (ro) {
        bdrv_reopen(image_bs, open_flags | BDRV_O_RDWR, &local_err);
        if (local_err) {
            error_propagate(errp, local_err);
3173
            goto out;
3174 3175 3176 3177 3178 3179 3180 3181 3182 3183 3184 3185 3186 3187 3188 3189 3190 3191 3192
        }
    }

    ret = bdrv_change_backing_file(image_bs, backing_file,
                               image_bs->drv ? image_bs->drv->format_name : "");

    if (ret < 0) {
        error_setg_errno(errp, -ret, "Could not change backing file to '%s'",
                         backing_file);
        /* don't exit here, so we can try to restore open flags if
         * appropriate */
    }

    if (ro) {
        bdrv_reopen(image_bs, open_flags, &local_err);
        if (local_err) {
            error_propagate(errp, local_err); /* will preserve prior errp */
        }
    }
3193 3194 3195

out:
    aio_context_release(aio_context);
3196 3197
}

K
Kevin Wolf 已提交
3198 3199 3200
void qmp_blockdev_add(BlockdevOptions *options, Error **errp)
{
    QmpOutputVisitor *ov = qmp_output_visitor_new();
3201 3202
    BlockDriverState *bs;
    BlockBackend *blk = NULL;
K
Kevin Wolf 已提交
3203 3204 3205 3206
    QObject *obj;
    QDict *qdict;
    Error *local_err = NULL;

3207
    /* TODO Sort it out in raw-posix and drive_new(): Reject aio=native with
K
Kevin Wolf 已提交
3208
     * cache.direct=false instead of silently switching to aio=threads, except
3209
     * when called from drive_new().
K
Kevin Wolf 已提交
3210 3211 3212
     *
     * For now, simply forbidding the combination for all drivers will do. */
    if (options->has_aio && options->aio == BLOCKDEV_AIO_OPTIONS_NATIVE) {
3213 3214 3215 3216
        bool direct = options->has_cache &&
                      options->cache->has_direct &&
                      options->cache->direct;
        if (!direct) {
K
Kevin Wolf 已提交
3217 3218 3219 3220 3221 3222 3223
            error_setg(errp, "aio=native requires cache.direct=true");
            goto fail;
        }
    }

    visit_type_BlockdevOptions(qmp_output_get_visitor(ov),
                               &options, NULL, &local_err);
3224
    if (local_err) {
K
Kevin Wolf 已提交
3225 3226 3227 3228 3229 3230 3231 3232 3233
        error_propagate(errp, local_err);
        goto fail;
    }

    obj = qmp_output_get_qobject(ov);
    qdict = qobject_to_qdict(obj);

    qdict_flatten(qdict);

3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 3245 3246 3247 3248
    if (options->has_id) {
        blk = blockdev_init(NULL, qdict, &local_err);
        if (local_err) {
            error_propagate(errp, local_err);
            goto fail;
        }

        bs = blk_bs(blk);
    } else {
        if (!qdict_get_try_str(qdict, "node-name")) {
            error_setg(errp, "'id' and/or 'node-name' need to be specified for "
                       "the root node");
            goto fail;
        }

3249 3250
        bs = bds_tree_init(qdict, errp);
        if (!bs) {
3251 3252
            goto fail;
        }
K
Kevin Wolf 已提交
3253 3254
    }

3255 3256 3257 3258 3259 3260
    if (bs && bdrv_key_required(bs)) {
        if (blk) {
            blk_unref(blk);
        } else {
            bdrv_unref(bs);
        }
3261 3262 3263 3264
        error_setg(errp, "blockdev-add doesn't support encrypted devices");
        goto fail;
    }

K
Kevin Wolf 已提交
3265 3266 3267 3268
fail:
    qmp_output_visitor_cleanup(ov);
}

3269
BlockJobInfoList *qmp_query_block_jobs(Error **errp)
S
Stefan Hajnoczi 已提交
3270
{
3271 3272
    BlockJobInfoList *head = NULL, **p_next = &head;
    BlockDriverState *bs;
S
Stefan Hajnoczi 已提交
3273

3274
    for (bs = bdrv_next(NULL); bs; bs = bdrv_next(bs)) {
3275 3276 3277 3278
        AioContext *aio_context = bdrv_get_aio_context(bs);

        aio_context_acquire(aio_context);

3279 3280 3281 3282 3283 3284
        if (bs->job) {
            BlockJobInfoList *elem = g_new0(BlockJobInfoList, 1);
            elem->value = block_job_query(bs->job);
            *p_next = elem;
            p_next = &elem->next;
        }
3285 3286

        aio_context_release(aio_context);
S
Stefan Hajnoczi 已提交
3287 3288
    }

3289
    return head;
S
Stefan Hajnoczi 已提交
3290
}
3291

3292
QemuOptsList qemu_common_drive_opts = {
3293
    .name = "drive",
3294
    .head = QTAILQ_HEAD_INITIALIZER(qemu_common_drive_opts.head),
3295 3296 3297 3298 3299
    .desc = {
        {
            .name = "snapshot",
            .type = QEMU_OPT_BOOL,
            .help = "enable/disable snapshot mode",
3300 3301 3302 3303
        },{
            .name = "discard",
            .type = QEMU_OPT_STRING,
            .help = "discard operation (ignore/off, unmap/on)",
3304
        },{
3305
            .name = BDRV_OPT_CACHE_WB,
K
Kevin Wolf 已提交
3306 3307 3308
            .type = QEMU_OPT_BOOL,
            .help = "enables writeback mode for any caches",
        },{
3309
            .name = BDRV_OPT_CACHE_DIRECT,
K
Kevin Wolf 已提交
3310 3311 3312
            .type = QEMU_OPT_BOOL,
            .help = "enables use of O_DIRECT (bypass the host page cache)",
        },{
3313
            .name = BDRV_OPT_CACHE_NO_FLUSH,
K
Kevin Wolf 已提交
3314 3315
            .type = QEMU_OPT_BOOL,
            .help = "ignore any flush requests for the device",
3316 3317 3318 3319 3320 3321 3322 3323 3324 3325 3326 3327 3328 3329 3330 3331 3332
        },{
            .name = "aio",
            .type = QEMU_OPT_STRING,
            .help = "host AIO implementation (threads, native)",
        },{
            .name = "format",
            .type = QEMU_OPT_STRING,
            .help = "disk format (raw, qcow2, ...)",
        },{
            .name = "rerror",
            .type = QEMU_OPT_STRING,
            .help = "read error action",
        },{
            .name = "werror",
            .type = QEMU_OPT_STRING,
            .help = "write error action",
        },{
3333
            .name = "read-only",
3334 3335 3336
            .type = QEMU_OPT_BOOL,
            .help = "open drive file as read-only",
        },{
3337
            .name = "throttling.iops-total",
3338 3339 3340
            .type = QEMU_OPT_NUMBER,
            .help = "limit total I/O operations per second",
        },{
3341
            .name = "throttling.iops-read",
3342 3343 3344
            .type = QEMU_OPT_NUMBER,
            .help = "limit read operations per second",
        },{
3345
            .name = "throttling.iops-write",
3346 3347 3348
            .type = QEMU_OPT_NUMBER,
            .help = "limit write operations per second",
        },{
3349
            .name = "throttling.bps-total",
3350 3351 3352
            .type = QEMU_OPT_NUMBER,
            .help = "limit total bytes per second",
        },{
3353
            .name = "throttling.bps-read",
3354 3355 3356
            .type = QEMU_OPT_NUMBER,
            .help = "limit read bytes per second",
        },{
3357
            .name = "throttling.bps-write",
3358 3359
            .type = QEMU_OPT_NUMBER,
            .help = "limit write bytes per second",
3360 3361 3362 3363 3364 3365 3366 3367 3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 3378 3379 3380 3381 3382 3383
        },{
            .name = "throttling.iops-total-max",
            .type = QEMU_OPT_NUMBER,
            .help = "I/O operations burst",
        },{
            .name = "throttling.iops-read-max",
            .type = QEMU_OPT_NUMBER,
            .help = "I/O operations read burst",
        },{
            .name = "throttling.iops-write-max",
            .type = QEMU_OPT_NUMBER,
            .help = "I/O operations write burst",
        },{
            .name = "throttling.bps-total-max",
            .type = QEMU_OPT_NUMBER,
            .help = "total bytes burst",
        },{
            .name = "throttling.bps-read-max",
            .type = QEMU_OPT_NUMBER,
            .help = "total bytes read burst",
        },{
            .name = "throttling.bps-write-max",
            .type = QEMU_OPT_NUMBER,
            .help = "total bytes write burst",
3384 3385 3386 3387
        },{
            .name = "throttling.iops-size",
            .type = QEMU_OPT_NUMBER,
            .help = "when limiting by iops max size of an I/O in bytes",
3388 3389 3390 3391
        },{
            .name = "throttling.group",
            .type = QEMU_OPT_STRING,
            .help = "name of the block throttling group",
3392 3393 3394 3395
        },{
            .name = "copy-on-read",
            .type = QEMU_OPT_BOOL,
            .help = "copy read data from backing file into image file",
3396 3397 3398 3399
        },{
            .name = "detect-zeroes",
            .type = QEMU_OPT_STRING,
            .help = "try to optimize zero writes (off, on, unmap)",
3400 3401 3402 3403
        },
        { /* end of list */ }
    },
};
3404

3405 3406 3407 3408 3409 3410 3411 3412 3413 3414 3415 3416 3417 3418 3419 3420 3421 3422 3423 3424 3425 3426 3427 3428 3429 3430 3431 3432 3433 3434 3435 3436 3437 3438 3439 3440 3441 3442 3443 3444 3445
static QemuOptsList qemu_root_bds_opts = {
    .name = "root-bds",
    .head = QTAILQ_HEAD_INITIALIZER(qemu_common_drive_opts.head),
    .desc = {
        {
            .name = "discard",
            .type = QEMU_OPT_STRING,
            .help = "discard operation (ignore/off, unmap/on)",
        },{
            .name = "cache.writeback",
            .type = QEMU_OPT_BOOL,
            .help = "enables writeback mode for any caches",
        },{
            .name = "cache.direct",
            .type = QEMU_OPT_BOOL,
            .help = "enables use of O_DIRECT (bypass the host page cache)",
        },{
            .name = "cache.no-flush",
            .type = QEMU_OPT_BOOL,
            .help = "ignore any flush requests for the device",
        },{
            .name = "aio",
            .type = QEMU_OPT_STRING,
            .help = "host AIO implementation (threads, native)",
        },{
            .name = "read-only",
            .type = QEMU_OPT_BOOL,
            .help = "open drive file as read-only",
        },{
            .name = "copy-on-read",
            .type = QEMU_OPT_BOOL,
            .help = "copy read data from backing file into image file",
        },{
            .name = "detect-zeroes",
            .type = QEMU_OPT_STRING,
            .help = "try to optimize zero writes (off, on, unmap)",
        },
        { /* end of list */ }
    },
};

3446 3447 3448 3449
QemuOptsList qemu_drive_opts = {
    .name = "drive",
    .head = QTAILQ_HEAD_INITIALIZER(qemu_drive_opts.head),
    .desc = {
3450 3451 3452 3453
        /*
         * no elements => accept any params
         * validation will happen later
         */
3454 3455 3456
        { /* end of list */ }
    },
};