X-Git-Url: https://repo.jachan.dev/qemu.git/blobdiff_plain/52e94ea5de3ed9d7ddf1b0e5fc6ff7c2807ae711..a3a683c33d38fb29c7a78903e88dda12b84cc88d:/block/replication.c diff --git a/block/replication.c b/block/replication.c index d300c15475..3d4dedddfc 100644 --- a/block/replication.c +++ b/block/replication.c @@ -13,37 +13,38 @@ */ #include "qemu/osdep.h" -#include "qemu-common.h" +#include "qemu/option.h" #include "block/nbd.h" #include "block/blockjob.h" #include "block/block_int.h" #include "block/block_backup.h" #include "sysemu/block-backend.h" #include "qapi/error.h" +#include "qapi/qmp/qdict.h" #include "replication.h" +typedef enum { + BLOCK_REPLICATION_NONE, /* block replication is not started */ + BLOCK_REPLICATION_RUNNING, /* block replication is running */ + BLOCK_REPLICATION_FAILOVER, /* failover is running in background */ + BLOCK_REPLICATION_FAILOVER_FAILED, /* failover failed */ + BLOCK_REPLICATION_DONE, /* block replication is done */ +} ReplicationStage; + typedef struct BDRVReplicationState { ReplicationMode mode; - int replication_state; + ReplicationStage stage; BdrvChild *active_disk; BdrvChild *hidden_disk; BdrvChild *secondary_disk; char *top_id; ReplicationState *rs; Error *blocker; - int orig_hidden_flags; - int orig_secondary_flags; + bool orig_hidden_read_only; + bool orig_secondary_read_only; int error; } BDRVReplicationState; -enum { - BLOCK_REPLICATION_NONE, /* block replication is not started */ - BLOCK_REPLICATION_RUNNING, /* block replication is running */ - BLOCK_REPLICATION_FAILOVER, /* failover is running in background */ - BLOCK_REPLICATION_FAILOVER_FAILED, /* failover failed */ - BLOCK_REPLICATION_DONE, /* block replication is done */ -}; - static void replication_start(ReplicationState *rs, ReplicationMode mode, Error **errp); static void replication_do_checkpoint(ReplicationState *rs, Error **errp); @@ -141,11 +142,11 @@ static void replication_close(BlockDriverState *bs) { BDRVReplicationState *s = bs->opaque; - if (s->replication_state == BLOCK_REPLICATION_RUNNING) { + if (s->stage == BLOCK_REPLICATION_RUNNING) { replication_stop(s->rs, false, NULL); } - if (s->replication_state == BLOCK_REPLICATION_FAILOVER) { - block_job_cancel_sync(s->active_disk->bs->job); + if (s->stage == BLOCK_REPLICATION_FAILOVER) { + job_cancel_sync(&s->active_disk->bs->job->job); } if (s->mode == REPLICATION_MODE_SECONDARY) { @@ -157,13 +158,17 @@ static void replication_close(BlockDriverState *bs) static void replication_child_perm(BlockDriverState *bs, BdrvChild *c, const BdrvChildRole *role, + BlockReopenQueue *reopen_queue, uint64_t perm, uint64_t shared, uint64_t *nperm, uint64_t *nshared) { - *nperm = *nshared = BLK_PERM_CONSISTENT_READ \ - | BLK_PERM_WRITE \ - | BLK_PERM_WRITE_UNCHANGED; - + *nperm = BLK_PERM_CONSISTENT_READ; + if ((bs->open_flags & (BDRV_O_INACTIVE | BDRV_O_RDWR)) == BDRV_O_RDWR) { + *nperm |= BLK_PERM_WRITE; + } + *nshared = BLK_PERM_CONSISTENT_READ \ + | BLK_PERM_WRITE \ + | BLK_PERM_WRITE_UNCHANGED; return; } @@ -174,7 +179,7 @@ static int64_t replication_getlength(BlockDriverState *bs) static int replication_get_io_status(BDRVReplicationState *s) { - switch (s->replication_state) { + switch (s->stage) { case BLOCK_REPLICATION_NONE: return -EIO; case BLOCK_REPLICATION_RUNNING: @@ -214,9 +219,6 @@ static coroutine_fn int replication_co_readv(BlockDriverState *bs, QEMUIOVector *qiov) { BDRVReplicationState *s = bs->opaque; - BdrvChild *child = s->secondary_disk; - BlockJob *job = NULL; - CowRequest req; int ret; if (s->mode == REPLICATION_MODE_PRIMARY) { @@ -229,30 +231,17 @@ static coroutine_fn int replication_co_readv(BlockDriverState *bs, return ret; } - if (child && child->bs) { - job = child->bs->job; - } - - if (job) { - backup_wait_for_overlapping_requests(child->bs->job, sector_num, - remaining_sectors); - backup_cow_request_begin(&req, child->bs->job, sector_num, - remaining_sectors); - ret = bdrv_co_readv(bs->file, sector_num, remaining_sectors, - qiov); - backup_cow_request_end(&req); - goto out; - } + ret = bdrv_co_preadv(bs->file, sector_num * BDRV_SECTOR_SIZE, + remaining_sectors * BDRV_SECTOR_SIZE, qiov, 0); - ret = bdrv_co_readv(bs->file, sector_num, remaining_sectors, qiov); -out: return replication_return_value(s, ret); } static coroutine_fn int replication_co_writev(BlockDriverState *bs, int64_t sector_num, int remaining_sectors, - QEMUIOVector *qiov) + QEMUIOVector *qiov, + int flags) { BDRVReplicationState *s = bs->opaque; QEMUIOVector hd_qiov; @@ -260,16 +249,18 @@ static coroutine_fn int replication_co_writev(BlockDriverState *bs, BdrvChild *top = bs->file; BdrvChild *base = s->secondary_disk; BdrvChild *target; - int ret, n; + int ret; + int64_t n; + assert(!flags); ret = replication_get_io_status(s); if (ret < 0) { goto out; } if (ret == 0) { - ret = bdrv_co_writev(top, sector_num, - remaining_sectors, qiov); + ret = bdrv_co_pwritev(top, sector_num * BDRV_SECTOR_SIZE, + remaining_sectors * BDRV_SECTOR_SIZE, qiov, 0); return replication_return_value(s, ret); } @@ -279,24 +270,31 @@ static coroutine_fn int replication_co_writev(BlockDriverState *bs, */ qemu_iovec_init(&hd_qiov, qiov->niov); while (remaining_sectors > 0) { - ret = bdrv_is_allocated_above(top->bs, base->bs, sector_num, - remaining_sectors, &n); + int64_t count; + + ret = bdrv_is_allocated_above(top->bs, base->bs, + sector_num * BDRV_SECTOR_SIZE, + remaining_sectors * BDRV_SECTOR_SIZE, + &count); if (ret < 0) { goto out1; } + assert(QEMU_IS_ALIGNED(count, BDRV_SECTOR_SIZE)); + n = count >> BDRV_SECTOR_BITS; qemu_iovec_reset(&hd_qiov); - qemu_iovec_concat(&hd_qiov, qiov, bytes_done, n * BDRV_SECTOR_SIZE); + qemu_iovec_concat(&hd_qiov, qiov, bytes_done, count); target = ret ? top : base; - ret = bdrv_co_writev(target, sector_num, n, &hd_qiov); + ret = bdrv_co_pwritev(target, sector_num * BDRV_SECTOR_SIZE, + n * BDRV_SECTOR_SIZE, &hd_qiov, 0); if (ret < 0) { goto out1; } remaining_sectors -= n; sector_num += n; - bytes_done += n * BDRV_SECTOR_SIZE; + bytes_done += count; } out1: @@ -327,12 +325,24 @@ static void secondary_do_checkpoint(BDRVReplicationState *s, Error **errp) return; } + if (!s->active_disk->bs->drv) { + error_setg(errp, "Active disk %s is ejected", + s->active_disk->bs->node_name); + return; + } + ret = s->active_disk->bs->drv->bdrv_make_empty(s->active_disk->bs); if (ret < 0) { error_setg(errp, "Cannot make active disk empty"); return; } + if (!s->hidden_disk->bs->drv) { + error_setg(errp, "Hidden disk %s is ejected", + s->hidden_disk->bs->node_name); + return; + } + ret = s->hidden_disk->bs->drv->bdrv_make_empty(s->hidden_disk->bs); if (ret < 0) { error_setg(errp, "Cannot make hidden disk empty"); @@ -340,48 +350,47 @@ static void secondary_do_checkpoint(BDRVReplicationState *s, Error **errp) } } +/* This function is supposed to be called twice: + * first with writable = true, then with writable = false. + * The first call puts s->hidden_disk and s->secondary_disk in + * r/w mode, and the second puts them back in their original state. + */ static void reopen_backing_file(BlockDriverState *bs, bool writable, Error **errp) { BDRVReplicationState *s = bs->opaque; BlockReopenQueue *reopen_queue = NULL; - int orig_hidden_flags, orig_secondary_flags; - int new_hidden_flags, new_secondary_flags; Error *local_err = NULL; if (writable) { - orig_hidden_flags = s->orig_hidden_flags = - bdrv_get_flags(s->hidden_disk->bs); - new_hidden_flags = (orig_hidden_flags | BDRV_O_RDWR) & - ~BDRV_O_INACTIVE; - orig_secondary_flags = s->orig_secondary_flags = - bdrv_get_flags(s->secondary_disk->bs); - new_secondary_flags = (orig_secondary_flags | BDRV_O_RDWR) & - ~BDRV_O_INACTIVE; - } else { - orig_hidden_flags = (s->orig_hidden_flags | BDRV_O_RDWR) & - ~BDRV_O_INACTIVE; - new_hidden_flags = s->orig_hidden_flags; - orig_secondary_flags = (s->orig_secondary_flags | BDRV_O_RDWR) & - ~BDRV_O_INACTIVE; - new_secondary_flags = s->orig_secondary_flags; + s->orig_hidden_read_only = bdrv_is_read_only(s->hidden_disk->bs); + s->orig_secondary_read_only = bdrv_is_read_only(s->secondary_disk->bs); } - if (orig_hidden_flags != new_hidden_flags) { - reopen_queue = bdrv_reopen_queue(reopen_queue, s->hidden_disk->bs, NULL, - new_hidden_flags); + bdrv_subtree_drained_begin(s->hidden_disk->bs); + bdrv_subtree_drained_begin(s->secondary_disk->bs); + + if (s->orig_hidden_read_only) { + QDict *opts = qdict_new(); + qdict_put_bool(opts, BDRV_OPT_READ_ONLY, !writable); + reopen_queue = bdrv_reopen_queue(reopen_queue, s->hidden_disk->bs, + opts, true); } - if (!(orig_secondary_flags & BDRV_O_RDWR)) { + if (s->orig_secondary_read_only) { + QDict *opts = qdict_new(); + qdict_put_bool(opts, BDRV_OPT_READ_ONLY, !writable); reopen_queue = bdrv_reopen_queue(reopen_queue, s->secondary_disk->bs, - NULL, new_secondary_flags); + opts, true); } if (reopen_queue) { - bdrv_reopen_multiple(bdrv_get_aio_context(bs), - reopen_queue, &local_err); + bdrv_reopen_multiple(reopen_queue, &local_err); error_propagate(errp, local_err); } + + bdrv_subtree_drained_end(s->hidden_disk->bs); + bdrv_subtree_drained_end(s->secondary_disk->bs); } static void backup_job_cleanup(BlockDriverState *bs) @@ -403,7 +412,7 @@ static void backup_job_completed(void *opaque, int ret) BlockDriverState *bs = opaque; BDRVReplicationState *s = bs->opaque; - if (s->replication_state != BLOCK_REPLICATION_FAILOVER) { + if (s->stage != BLOCK_REPLICATION_FAILOVER) { /* The backup job is cancelled unexpectedly */ s->error = -EIO; } @@ -445,7 +454,7 @@ static void replication_start(ReplicationState *rs, ReplicationMode mode, aio_context_acquire(aio_context); s = bs->opaque; - if (s->replication_state != BLOCK_REPLICATION_NONE) { + if (s->stage != BLOCK_REPLICATION_NONE) { error_setg(errp, "Block replication is running or done"); aio_context_release(aio_context); return; @@ -496,6 +505,9 @@ static void replication_start(ReplicationState *rs, ReplicationMode mode, return; } + /* Must be true, or the bdrv_getlength() calls would have failed */ + assert(s->active_disk->bs->drv && s->hidden_disk->bs->drv); + if (!s->active_disk->bs->drv->bdrv_make_empty || !s->hidden_disk->bs->drv->bdrv_make_empty) { error_setg(errp, @@ -530,7 +542,7 @@ static void replication_start(ReplicationState *rs, ReplicationMode mode, job = backup_job_create(NULL, s->secondary_disk->bs, s->hidden_disk->bs, 0, MIRROR_SYNC_MODE_NONE, NULL, false, BLOCKDEV_ON_ERROR_REPORT, - BLOCKDEV_ON_ERROR_REPORT, BLOCK_JOB_INTERNAL, + BLOCKDEV_ON_ERROR_REPORT, JOB_INTERNAL, backup_job_completed, bs, NULL, &local_err); if (local_err) { error_propagate(errp, local_err); @@ -538,14 +550,14 @@ static void replication_start(ReplicationState *rs, ReplicationMode mode, aio_context_release(aio_context); return; } - block_job_start(job); + job_start(&job->job); break; default: aio_context_release(aio_context); abort(); } - s->replication_state = BLOCK_REPLICATION_RUNNING; + s->stage = BLOCK_REPLICATION_RUNNING; if (s->mode == REPLICATION_MODE_SECONDARY) { secondary_do_checkpoint(s, errp); @@ -581,7 +593,7 @@ static void replication_get_error(ReplicationState *rs, Error **errp) aio_context_acquire(aio_context); s = bs->opaque; - if (s->replication_state != BLOCK_REPLICATION_RUNNING) { + if (s->stage != BLOCK_REPLICATION_RUNNING) { error_setg(errp, "Block replication is not running"); aio_context_release(aio_context); return; @@ -601,16 +613,14 @@ static void replication_done(void *opaque, int ret) BDRVReplicationState *s = bs->opaque; if (ret == 0) { - s->replication_state = BLOCK_REPLICATION_DONE; + s->stage = BLOCK_REPLICATION_DONE; - /* refresh top bs's filename */ - bdrv_refresh_filename(bs); s->active_disk = NULL; s->secondary_disk = NULL; s->hidden_disk = NULL; s->error = 0; } else { - s->replication_state = BLOCK_REPLICATION_FAILOVER_FAILED; + s->stage = BLOCK_REPLICATION_FAILOVER_FAILED; s->error = -EIO; } } @@ -625,7 +635,7 @@ static void replication_stop(ReplicationState *rs, bool failover, Error **errp) aio_context_acquire(aio_context); s = bs->opaque; - if (s->replication_state != BLOCK_REPLICATION_RUNNING) { + if (s->stage != BLOCK_REPLICATION_RUNNING) { error_setg(errp, "Block replication is not running"); aio_context_release(aio_context); return; @@ -633,7 +643,7 @@ static void replication_stop(ReplicationState *rs, bool failover, Error **errp) switch (s->mode) { case REPLICATION_MODE_PRIMARY: - s->replication_state = BLOCK_REPLICATION_DONE; + s->stage = BLOCK_REPLICATION_DONE; s->error = 0; break; case REPLICATION_MODE_SECONDARY: @@ -643,19 +653,19 @@ static void replication_stop(ReplicationState *rs, bool failover, Error **errp) * disk, secondary disk in backup_job_completed(). */ if (s->secondary_disk->bs->job) { - block_job_cancel_sync(s->secondary_disk->bs->job); + job_cancel_sync(&s->secondary_disk->bs->job->job); } if (!failover) { secondary_do_checkpoint(s, errp); - s->replication_state = BLOCK_REPLICATION_DONE; + s->stage = BLOCK_REPLICATION_DONE; aio_context_release(aio_context); return; } - s->replication_state = BLOCK_REPLICATION_FAILOVER; + s->stage = BLOCK_REPLICATION_FAILOVER; commit_active_start(NULL, s->active_disk->bs, s->secondary_disk->bs, - BLOCK_JOB_INTERNAL, 0, BLOCKDEV_ON_ERROR_REPORT, + JOB_INTERNAL, 0, BLOCKDEV_ON_ERROR_REPORT, NULL, replication_done, bs, true, errp); break; default: @@ -665,9 +675,15 @@ static void replication_stop(ReplicationState *rs, bool failover, Error **errp) aio_context_release(aio_context); } -BlockDriver bdrv_replication = { +static const char *const replication_strong_runtime_opts[] = { + REPLICATION_MODE, + REPLICATION_TOP_ID, + + NULL +}; + +static BlockDriver bdrv_replication = { .format_name = "replication", - .protocol_name = "replication", .instance_size = sizeof(BDRVReplicationState), .bdrv_open = replication_open, @@ -682,6 +698,7 @@ BlockDriver bdrv_replication = { .bdrv_recurse_is_first_non_filter = replication_recurse_is_first_non_filter, .has_variable_length = true, + .strong_runtime_opts = replication_strong_runtime_opts, }; static void bdrv_replication_init(void)