static int
spa_check_special_degraded(spa_t *spa)
{
metaslab_class_t *mc;
metaslab_group_t *mg;
vdev_t *vd;
if (!spa_has_special(spa))
return (0);
mc = spa_special_class(spa);
/*
* Must hold one of the spa_config locks.
*/
ASSERT(spa_config_held(mc->mc_spa, SCL_ALL, RW_READER) ||
spa_config_held(mc->mc_spa, SCL_ALL, RW_WRITER));
if ((mg = mc->mc_rotor) == NULL)
return (0);
do {
vd = mg->mg_vd;
if (vd->vdev_state == VDEV_STATE_DEGRADED ||
vd->vdev_state == VDEV_STATE_FAULTED)
return (1);
} while ((mg = mg->mg_next) != mc->mc_rotor);
return (0);
}
/*
* Remove any holes in the child array.
*/
void
vdev_compact_children(vdev_t *pvd)
{
vdev_t **newchild, *cvd;
int oldc = pvd->vdev_children;
int newc, c;
ASSERT(spa_config_held(pvd->vdev_spa, RW_WRITER));
for (c = newc = 0; c < oldc; c++)
if (pvd->vdev_child[c])
newc++;
newchild = kmem_alloc(newc * sizeof (vdev_t *), KM_SLEEP);
for (c = newc = 0; c < oldc; c++) {
if ((cvd = pvd->vdev_child[c]) != NULL) {
newchild[newc] = cvd;
cvd->vdev_id = newc++;
}
}
kmem_free(pvd->vdev_child, oldc * sizeof (vdev_t *));
pvd->vdev_child = newchild;
pvd->vdev_children = newc;
}
/*
* Add a mirror/replacing vdev above an existing vdev.
*/
vdev_t *
vdev_add_parent(vdev_t *cvd, vdev_ops_t *ops)
{
spa_t *spa = cvd->vdev_spa;
vdev_t *pvd = cvd->vdev_parent;
vdev_t *mvd;
ASSERT(spa_config_held(spa, RW_WRITER));
mvd = vdev_alloc_common(spa, cvd->vdev_id, 0, ops);
mvd->vdev_asize = cvd->vdev_asize;
mvd->vdev_ashift = cvd->vdev_ashift;
mvd->vdev_state = cvd->vdev_state;
vdev_remove_child(pvd, cvd);
vdev_add_child(pvd, mvd);
cvd->vdev_id = mvd->vdev_children;
vdev_add_child(mvd, cvd);
vdev_top_update(cvd->vdev_top, cvd->vdev_top);
if (mvd == mvd->vdev_top)
vdev_top_transfer(cvd, mvd);
return (mvd);
}
static void
vdev_initialize_load(vdev_t *vd)
{
ASSERT(spa_config_held(vd->vdev_spa, SCL_CONFIG, RW_READER) ||
spa_config_held(vd->vdev_spa, SCL_CONFIG, RW_WRITER));
ASSERT(vd->vdev_leaf_zap != 0);
if (vd->vdev_initialize_state == VDEV_INITIALIZE_ACTIVE ||
vd->vdev_initialize_state == VDEV_INITIALIZE_SUSPENDED) {
int err = zap_lookup(vd->vdev_spa->spa_meta_objset,
vd->vdev_leaf_zap, VDEV_LEAF_ZAP_INITIALIZE_LAST_OFFSET,
sizeof (vd->vdev_initialize_last_offset), 1,
&vd->vdev_initialize_last_offset);
ASSERT(err == 0 || err == ENOENT);
}
vdev_initialize_calculate_progress(vd);
}
static void
vdev_disk_rele(vdev_t *vd)
{
ASSERT(spa_config_held(vd->vdev_spa, SCL_STATE, RW_WRITER));
if (vd->vdev_name_vp) {
VN_RELE_ASYNC(vd->vdev_name_vp,
dsl_pool_vnrele_taskq(vd->vdev_spa->spa_dsl_pool));
vd->vdev_name_vp = NULL;
}
if (vd->vdev_devid_vp) {
VN_RELE_ASYNC(vd->vdev_devid_vp,
dsl_pool_vnrele_taskq(vd->vdev_spa->spa_dsl_pool));
vd->vdev_devid_vp = NULL;
}
}
/*
* Generate the pool's configuration based on the current in-core state.
* We infer whether to generate a complete config or just one top-level config
* based on whether vd is the root vdev.
*/
nvlist_t *
spa_config_generate(spa_t *spa, vdev_t *vd, uint64_t txg, int getstats)
{
nvlist_t *config, *nvroot;
vdev_t *rvd = spa->spa_root_vdev;
ASSERT(spa_config_held(spa, RW_READER));
if (vd == NULL)
vd = rvd;
/*
* If txg is -1, report the current value of spa->spa_config_txg.
*/
if (txg == -1ULL)
txg = spa->spa_config_txg;
VERIFY(nvlist_alloc(&config, NV_UNIQUE_NAME, KM_SLEEP) == 0);
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_VERSION,
spa_version(spa)) == 0);
VERIFY(nvlist_add_string(config, ZPOOL_CONFIG_POOL_NAME,
spa_name(spa)) == 0);
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_POOL_STATE,
spa_state(spa)) == 0);
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_POOL_TXG,
txg) == 0);
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_POOL_GUID,
spa_guid(spa)) == 0);
if (vd != rvd) {
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_TOP_GUID,
vd->vdev_top->vdev_guid) == 0);
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_GUID,
vd->vdev_guid) == 0);
if (vd->vdev_isspare)
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_IS_SPARE,
1ULL) == 0);
vd = vd->vdev_top; /* label contains top config */
}
nvroot = vdev_config_generate(spa, vd, getstats, B_FALSE);
VERIFY(nvlist_add_nvlist(config, ZPOOL_CONFIG_VDEV_TREE, nvroot) == 0);
nvlist_free(nvroot);
return (config);
}
void
vdev_add_child(vdev_t *pvd, vdev_t *cvd)
{
size_t oldsize, newsize;
uint64_t id = cvd->vdev_id;
vdev_t **newchild;
ASSERT(spa_config_held(cvd->vdev_spa, RW_WRITER));
ASSERT(cvd->vdev_parent == NULL);
cvd->vdev_parent = pvd;
if (pvd == NULL)
return;
ASSERT(id >= pvd->vdev_children || pvd->vdev_child[id] == NULL);
oldsize = pvd->vdev_children * sizeof (vdev_t *);
pvd->vdev_children = MAX(pvd->vdev_children, id + 1);
newsize = pvd->vdev_children * sizeof (vdev_t *);
newchild = kmem_zalloc(newsize, KM_SLEEP);
if (pvd->vdev_child != NULL) {
bcopy(pvd->vdev_child, newchild, oldsize);
kmem_free(pvd->vdev_child, oldsize);
}
pvd->vdev_child = newchild;
pvd->vdev_child[id] = cvd;
cvd->vdev_top = (pvd->vdev_top ? pvd->vdev_top: cvd);
ASSERT(cvd->vdev_top->vdev_parent->vdev_parent == NULL);
/*
* Walk up all ancestors to update guid sum.
*/
for (; pvd != NULL; pvd = pvd->vdev_parent)
pvd->vdev_guid_sum += cvd->vdev_guid_sum;
if (cvd->vdev_ops->vdev_op_leaf)
cvd->vdev_spa->spa_scrub_maxinflight += zfs_scrub_limit;
}
static void
vdev_disk_hold(vdev_t *vd)
{
ddi_devid_t devid;
char *minor;
ASSERT(spa_config_held(vd->vdev_spa, SCL_STATE, RW_WRITER));
/*
* We must have a pathname, and it must be absolute.
*/
if (vd->vdev_path == NULL || vd->vdev_path[0] != '/')
return;
/*
* Only prefetch path and devid info if the device has
* never been opened.
*/
if (vd->vdev_tsd != NULL)
return;
if (vd->vdev_wholedisk == -1ULL) {
size_t len = strlen(vd->vdev_path) + 3;
char *buf = kmem_alloc(len, KM_SLEEP);
(void) snprintf(buf, len, "%ss0", vd->vdev_path);
(void) ldi_vp_from_name(buf, &vd->vdev_name_vp);
kmem_free(buf, len);
}
if (vd->vdev_name_vp == NULL)
(void) ldi_vp_from_name(vd->vdev_path, &vd->vdev_name_vp);
if (vd->vdev_devid != NULL &&
ddi_devid_str_decode(vd->vdev_devid, &devid, &minor) == 0) {
(void) ldi_vp_from_devid(devid, minor, &vd->vdev_devid_vp);
ddi_devid_str_free(minor);
ddi_devid_free(devid);
}
}
/*
* Remove a 1-way mirror/replacing vdev from the tree.
*/
void
vdev_remove_parent(vdev_t *cvd)
{
vdev_t *mvd = cvd->vdev_parent;
vdev_t *pvd = mvd->vdev_parent;
ASSERT(spa_config_held(cvd->vdev_spa, RW_WRITER));
ASSERT(mvd->vdev_children == 1);
ASSERT(mvd->vdev_ops == &vdev_mirror_ops ||
mvd->vdev_ops == &vdev_replacing_ops ||
mvd->vdev_ops == &vdev_spare_ops);
cvd->vdev_ashift = mvd->vdev_ashift;
vdev_remove_child(mvd, cvd);
vdev_remove_child(pvd, mvd);
cvd->vdev_id = mvd->vdev_id;
vdev_add_child(pvd, cvd);
/*
* If we created a new toplevel vdev, then we need to change the child's
* vdev GUID to match the old toplevel vdev. Otherwise, we could have
* detached an offline device, and when we go to import the pool we'll
* think we have two toplevel vdevs, instead of a different version of
* the same toplevel vdev.
*/
if (cvd->vdev_top == cvd) {
pvd->vdev_guid_sum -= cvd->vdev_guid;
cvd->vdev_guid_sum -= cvd->vdev_guid;
cvd->vdev_guid = mvd->vdev_guid;
cvd->vdev_guid_sum += mvd->vdev_guid;
pvd->vdev_guid_sum += cvd->vdev_guid;
}
vdev_top_update(cvd->vdev_top, cvd->vdev_top);
if (cvd == cvd->vdev_top)
vdev_top_transfer(mvd, cvd);
ASSERT(mvd->vdev_children == 0);
vdev_free(mvd);
}
/*
* Generate the pool's configuration based on the current in-core state.
* We infer whether to generate a complete config or just one top-level config
* based on whether vd is the root vdev.
*/
nvlist_t *
spa_config_generate(spa_t *spa, vdev_t *vd, uint64_t txg, int getstats)
{
nvlist_t *config, *nvroot;
vdev_t *rvd = spa->spa_root_vdev;
unsigned long hostid = 0;
#ifdef __APPLE__
boolean_t skipdevpaths = (txg != -1ULL && txg != 0ULL);
#endif
ASSERT(spa_config_held(spa, RW_READER) ||
spa_config_held(spa, RW_WRITER));
if (vd == NULL)
vd = rvd;
/*
* If txg is -1, report the current value of spa->spa_config_txg.
*/
if (txg == -1ULL)
txg = spa->spa_config_txg;
VERIFY(nvlist_alloc(&config, NV_UNIQUE_NAME, KM_SLEEP) == 0);
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_VERSION,
spa_version(spa)) == 0);
VERIFY(nvlist_add_string(config, ZPOOL_CONFIG_POOL_NAME,
spa_name(spa)) == 0);
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_POOL_STATE,
spa_state(spa)) == 0);
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_POOL_TXG,
txg) == 0);
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_POOL_GUID,
spa_guid(spa)) == 0);
#ifndef __APPLE__
(void) ddi_strtoul(hw_serial, NULL, 10, &hostid);
#endif
if (hostid != 0) {
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_HOSTID,
hostid) == 0);
}
/*XXX NOEL: If host id identification is a feature we are
* interested in having (primarily for poor man's cluster
* support), then query for the hostid here and add it to the
* nvlist
*/
#ifndef __APPLE__
VERIFY(nvlist_add_string(config, ZPOOL_CONFIG_HOSTNAME,
utsname.nodename) == 0);
#endif
if (vd != rvd) {
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_TOP_GUID,
vd->vdev_top->vdev_guid) == 0);
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_GUID,
vd->vdev_guid) == 0);
if (vd->vdev_isspare)
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_IS_SPARE,
1ULL) == 0);
if (vd->vdev_islog)
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_IS_LOG,
1ULL) == 0);
vd = vd->vdev_top; /* label contains top config */
}
#ifdef __APPLE__
nvroot = vdev_config_generate(spa, vd, getstats, B_FALSE, B_FALSE,
skipdevpaths);
#else
nvroot = vdev_config_generate(spa, vd, getstats, B_FALSE, B_FALSE);
#endif
VERIFY(nvlist_add_nvlist(config, ZPOOL_CONFIG_VDEV_TREE, nvroot) == 0);
nvlist_free(nvroot);
return (config);
}
{
int l;
if (offset >= psize - VDEV_LABEL_END_SIZE) {
offset -= psize - VDEV_LABEL_END_SIZE;
offset += (VDEV_LABELS / 2) * sizeof (vdev_label_t);
}
l = offset / sizeof (vdev_label_t);
return (l < VDEV_LABELS ? l : -1);
}
static void
vdev_label_read(zio_t *zio, vdev_t *vd, int l, void *buf, uint64_t offset,
uint64_t size, zio_done_func_t *done, void *private, int flags)
{
ASSERT(spa_config_held(zio->io_spa, SCL_STATE_ALL, RW_WRITER) ==
SCL_STATE_ALL);
ASSERT(flags & ZIO_FLAG_CONFIG_WRITER);
zio_nowait(zio_read_phys(zio, vd,
vdev_label_offset(vd->vdev_psize, l, offset),
size, buf, ZIO_CHECKSUM_LABEL, done, private,
ZIO_PRIORITY_SYNC_READ, flags, B_TRUE));
}
static void
vdev_label_write(zio_t *zio, vdev_t *vd, int l, void *buf, uint64_t offset,
uint64_t size, zio_done_func_t *done, void *private, int flags)
{
ASSERT(spa_config_held(zio->io_spa, SCL_ALL, RW_WRITER) == SCL_ALL ||
(spa_config_held(zio->io_spa, SCL_CONFIG | SCL_STATE, RW_READER) ==
(SCL_CONFIG | SCL_STATE) &&
/*
* Generate the pool's configuration based on the current in-core state.
*
* We infer whether to generate a complete config or just one top-level config
* based on whether vd is the root vdev.
*/
nvlist_t *
spa_config_generate(spa_t *spa, vdev_t *vd, uint64_t txg, int getstats)
{
nvlist_t *config, *nvroot;
vdev_t *rvd = spa->spa_root_vdev;
unsigned long hostid = 0;
boolean_t locked = B_FALSE;
uint64_t split_guid;
if (vd == NULL) {
vd = rvd;
locked = B_TRUE;
spa_config_enter(spa, SCL_CONFIG | SCL_STATE, FTAG, RW_READER);
}
ASSERT(spa_config_held(spa, SCL_CONFIG | SCL_STATE, RW_READER) ==
(SCL_CONFIG | SCL_STATE));
/*
* If txg is -1, report the current value of spa->spa_config_txg.
*/
if (txg == -1ULL)
txg = spa->spa_config_txg;
VERIFY(nvlist_alloc(&config, NV_UNIQUE_NAME, KM_SLEEP) == 0);
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_VERSION,
spa_version(spa)) == 0);
VERIFY(nvlist_add_string(config, ZPOOL_CONFIG_POOL_NAME,
spa_name(spa)) == 0);
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_POOL_STATE,
spa_state(spa)) == 0);
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_POOL_TXG,
txg) == 0);
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_POOL_GUID,
spa_guid(spa)) == 0);
VERIFY(spa->spa_comment == NULL || nvlist_add_string(config,
ZPOOL_CONFIG_COMMENT, spa->spa_comment) == 0);
#ifdef _KERNEL
hostid = zone_get_hostid(NULL);
#else /* _KERNEL */
/*
* We're emulating the system's hostid in userland, so we can't use
* zone_get_hostid().
*/
(void) ddi_strtoul(hw_serial, NULL, 10, &hostid);
#endif /* _KERNEL */
if (hostid != 0) {
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_HOSTID,
hostid) == 0);
}
VERIFY(nvlist_add_string(config, ZPOOL_CONFIG_HOSTNAME,
utsname.nodename) == 0);
if (vd != rvd) {
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_TOP_GUID,
vd->vdev_top->vdev_guid) == 0);
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_GUID,
vd->vdev_guid) == 0);
if (vd->vdev_isspare)
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_IS_SPARE,
1ULL) == 0);
if (vd->vdev_islog)
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_IS_LOG,
1ULL) == 0);
vd = vd->vdev_top; /* label contains top config */
} else {
/*
* Only add the (potentially large) split information
* in the mos config, and not in the vdev labels
*/
if (spa->spa_config_splitting != NULL)
VERIFY(nvlist_add_nvlist(config, ZPOOL_CONFIG_SPLIT,
spa->spa_config_splitting) == 0);
}
/*
* Add the top-level config. We even add this on pools which
* don't support holes in the namespace.
*/
vdev_top_config_generate(spa, config);
/*
* If we're splitting, record the original pool's guid.
*/
if (spa->spa_config_splitting != NULL &&
nvlist_lookup_uint64(spa->spa_config_splitting,
ZPOOL_CONFIG_SPLIT_GUID, &split_guid) == 0) {
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_SPLIT_GUID,
split_guid) == 0);
}
nvroot = vdev_config_generate(spa, vd, getstats, 0);
VERIFY(nvlist_add_nvlist(config, ZPOOL_CONFIG_VDEV_TREE, nvroot) == 0);
nvlist_free(nvroot);
/*
* Store what's necessary for reading the MOS in the label.
*/
VERIFY(nvlist_add_nvlist(config, ZPOOL_CONFIG_FEATURES_FOR_READ,
spa->spa_label_features) == 0);
if (getstats && spa_load_state(spa) == SPA_LOAD_NONE) {
ddt_histogram_t *ddh;
ddt_stat_t *dds;
ddt_object_t *ddo;
ddh = kmem_zalloc(sizeof (ddt_histogram_t), KM_SLEEP);
ddt_get_dedup_histogram(spa, ddh);
VERIFY(nvlist_add_uint64_array(config,
ZPOOL_CONFIG_DDT_HISTOGRAM,
(uint64_t *)ddh, sizeof (*ddh) / sizeof (uint64_t)) == 0);
kmem_free(ddh, sizeof (ddt_histogram_t));
ddo = kmem_zalloc(sizeof (ddt_object_t), KM_SLEEP);
ddt_get_dedup_object_stats(spa, ddo);
VERIFY(nvlist_add_uint64_array(config,
ZPOOL_CONFIG_DDT_OBJ_STATS,
(uint64_t *)ddo, sizeof (*ddo) / sizeof (uint64_t)) == 0);
kmem_free(ddo, sizeof (ddt_object_t));
dds = kmem_zalloc(sizeof (ddt_stat_t), KM_SLEEP);
ddt_get_dedup_stats(spa, dds);
VERIFY(nvlist_add_uint64_array(config,
ZPOOL_CONFIG_DDT_STATS,
(uint64_t *)dds, sizeof (*dds) / sizeof (uint64_t)) == 0);
kmem_free(dds, sizeof (ddt_stat_t));
}
if (locked)
spa_config_exit(spa, SCL_CONFIG | SCL_STATE, FTAG);
return (config);
}
/*
* Generate the pool's configuration based on the current in-core state.
*
* We infer whether to generate a complete config or just one top-level config
* based on whether vd is the root vdev.
*/
nvlist_t *
spa_config_generate(spa_t *spa, vdev_t *vd, uint64_t txg, int getstats)
{
nvlist_t *config, *nvroot;
vdev_t *rvd = spa->spa_root_vdev;
unsigned long hostid = 0;
boolean_t locked = B_FALSE;
uint64_t split_guid;
char *pool_name;
if (vd == NULL) {
vd = rvd;
locked = B_TRUE;
spa_config_enter(spa, SCL_CONFIG | SCL_STATE, FTAG, RW_READER);
}
ASSERT(spa_config_held(spa, SCL_CONFIG | SCL_STATE, RW_READER) ==
(SCL_CONFIG | SCL_STATE));
/*
* If txg is -1, report the current value of spa->spa_config_txg.
*/
if (txg == -1ULL)
txg = spa->spa_config_txg;
/*
* Originally, users had to handle spa namespace collisions by either
* exporting the already imported pool or by specifying a new name for
* the pool with a conflicting name. In the case of root pools from
* virtual guests, neither approach to collision resolution is
* reasonable. This is addressed by extending the new name syntax with
* an option to specify that the new name is temporary. When specified,
* ZFS_IMPORT_TEMP_NAME will be set in spa->spa_import_flags to tell us
* to use the previous name, which we do below.
*/
if (spa->spa_import_flags & ZFS_IMPORT_TEMP_NAME) {
VERIFY0(nvlist_lookup_string(spa->spa_config,
ZPOOL_CONFIG_POOL_NAME, &pool_name));
} else
pool_name = spa_name(spa);
VERIFY(nvlist_alloc(&config, NV_UNIQUE_NAME, KM_SLEEP) == 0);
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_VERSION,
spa_version(spa)) == 0);
VERIFY(nvlist_add_string(config, ZPOOL_CONFIG_POOL_NAME,
pool_name) == 0);
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_POOL_STATE,
spa_state(spa)) == 0);
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_POOL_TXG,
txg) == 0);
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_POOL_GUID,
spa_guid(spa)) == 0);
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_ERRATA,
spa->spa_errata) == 0);
VERIFY(spa->spa_comment == NULL || nvlist_add_string(config,
ZPOOL_CONFIG_COMMENT, spa->spa_comment) == 0);
#ifdef _KERNEL
hostid = zone_get_hostid(NULL);
#else /* _KERNEL */
/*
* We're emulating the system's hostid in userland, so we can't use
* zone_get_hostid().
*/
(void) ddi_strtoul(hw_serial, NULL, 10, &hostid);
#endif /* _KERNEL */
if (hostid != 0) {
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_HOSTID,
hostid) == 0);
}
VERIFY0(nvlist_add_string(config, ZPOOL_CONFIG_HOSTNAME,
utsname()->nodename));
if (vd != rvd) {
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_TOP_GUID,
vd->vdev_top->vdev_guid) == 0);
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_GUID,
vd->vdev_guid) == 0);
if (vd->vdev_isspare)
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_IS_SPARE,
1ULL) == 0);
if (vd->vdev_islog)
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_IS_LOG,
1ULL) == 0);
vd = vd->vdev_top; /* label contains top config */
} else {
/*
* Only add the (potentially large) split information
* in the mos config, and not in the vdev labels
*/
if (spa->spa_config_splitting != NULL)
VERIFY(nvlist_add_nvlist(config, ZPOOL_CONFIG_SPLIT,
spa->spa_config_splitting) == 0);
}
/*
* Add the top-level config. We even add this on pools which
* don't support holes in the namespace.
*/
vdev_top_config_generate(spa, config);
/*
* If we're splitting, record the original pool's guid.
*/
if (spa->spa_config_splitting != NULL &&
nvlist_lookup_uint64(spa->spa_config_splitting,
ZPOOL_CONFIG_SPLIT_GUID, &split_guid) == 0) {
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_SPLIT_GUID,
split_guid) == 0);
}
nvroot = vdev_config_generate(spa, vd, getstats, 0);
VERIFY(nvlist_add_nvlist(config, ZPOOL_CONFIG_VDEV_TREE, nvroot) == 0);
nvlist_free(nvroot);
/*
* Store what's necessary for reading the MOS in the label.
*/
VERIFY(nvlist_add_nvlist(config, ZPOOL_CONFIG_FEATURES_FOR_READ,
spa->spa_label_features) == 0);
if (getstats && spa_load_state(spa) == SPA_LOAD_NONE) {
ddt_histogram_t *ddh;
ddt_stat_t *dds;
ddt_object_t *ddo;
ddh = kmem_zalloc(sizeof (ddt_histogram_t), KM_SLEEP);
ddt_get_dedup_histogram(spa, ddh);
VERIFY(nvlist_add_uint64_array(config,
ZPOOL_CONFIG_DDT_HISTOGRAM,
(uint64_t *)ddh, sizeof (*ddh) / sizeof (uint64_t)) == 0);
kmem_free(ddh, sizeof (ddt_histogram_t));
ddo = kmem_zalloc(sizeof (ddt_object_t), KM_SLEEP);
ddt_get_dedup_object_stats(spa, ddo);
VERIFY(nvlist_add_uint64_array(config,
ZPOOL_CONFIG_DDT_OBJ_STATS,
(uint64_t *)ddo, sizeof (*ddo) / sizeof (uint64_t)) == 0);
kmem_free(ddo, sizeof (ddt_object_t));
dds = kmem_zalloc(sizeof (ddt_stat_t), KM_SLEEP);
ddt_get_dedup_stats(spa, dds);
VERIFY(nvlist_add_uint64_array(config,
ZPOOL_CONFIG_DDT_STATS,
(uint64_t *)dds, sizeof (*dds) / sizeof (uint64_t)) == 0);
kmem_free(dds, sizeof (ddt_stat_t));
}
if (locked)
spa_config_exit(spa, SCL_CONFIG | SCL_STATE, FTAG);
return (config);
}
static void
vdev_initialize_calculate_progress(vdev_t *vd)
{
ASSERT(spa_config_held(vd->vdev_spa, SCL_CONFIG, RW_READER) ||
spa_config_held(vd->vdev_spa, SCL_CONFIG, RW_WRITER));
ASSERT(vd->vdev_leaf_zap != 0);
vd->vdev_initialize_bytes_est = 0;
vd->vdev_initialize_bytes_done = 0;
for (uint64_t i = 0; i < vd->vdev_top->vdev_ms_count; i++) {
metaslab_t *msp = vd->vdev_top->vdev_ms[i];
mutex_enter(&msp->ms_lock);
uint64_t ms_free = msp->ms_size -
metaslab_allocated_space(msp);
if (vd->vdev_top->vdev_ops == &vdev_raidz_ops)
ms_free /= vd->vdev_top->vdev_children;
/*
* Convert the metaslab range to a physical range
* on our vdev. We use this to determine if we are
* in the middle of this metaslab range.
*/
range_seg_t logical_rs, physical_rs;
logical_rs.rs_start = msp->ms_start;
logical_rs.rs_end = msp->ms_start + msp->ms_size;
vdev_xlate(vd, &logical_rs, &physical_rs);
if (vd->vdev_initialize_last_offset <= physical_rs.rs_start) {
vd->vdev_initialize_bytes_est += ms_free;
mutex_exit(&msp->ms_lock);
continue;
} else if (vd->vdev_initialize_last_offset >
physical_rs.rs_end) {
vd->vdev_initialize_bytes_done += ms_free;
vd->vdev_initialize_bytes_est += ms_free;
mutex_exit(&msp->ms_lock);
continue;
}
/*
* If we get here, we're in the middle of initializing this
* metaslab. Load it and walk the free tree for more accurate
* progress estimation.
*/
VERIFY0(metaslab_load(msp));
for (range_seg_t *rs = avl_first(&msp->ms_allocatable->rt_root);
rs; rs = AVL_NEXT(&msp->ms_allocatable->rt_root, rs)) {
logical_rs.rs_start = rs->rs_start;
logical_rs.rs_end = rs->rs_end;
vdev_xlate(vd, &logical_rs, &physical_rs);
uint64_t size = physical_rs.rs_end -
physical_rs.rs_start;
vd->vdev_initialize_bytes_est += size;
if (vd->vdev_initialize_last_offset >
physical_rs.rs_end) {
vd->vdev_initialize_bytes_done += size;
} else if (vd->vdev_initialize_last_offset >
physical_rs.rs_start &&
vd->vdev_initialize_last_offset <
physical_rs.rs_end) {
vd->vdev_initialize_bytes_done +=
vd->vdev_initialize_last_offset -
physical_rs.rs_start;
}
}
mutex_exit(&msp->ms_lock);
}
}
/*
* Allocate a new vdev. The 'alloctype' is used to control whether we are
* creating a new vdev or loading an existing one - the behavior is slightly
* different for each case.
*/
int
vdev_alloc(spa_t *spa, vdev_t **vdp, nvlist_t *nv, vdev_t *parent, uint_t id,
int alloctype)
{
vdev_ops_t *ops;
char *type;
uint64_t guid = 0;
vdev_t *vd;
ASSERT(spa_config_held(spa, RW_WRITER));
if (nvlist_lookup_string(nv, ZPOOL_CONFIG_TYPE, &type) != 0)
return (EINVAL);
if ((ops = vdev_getops(type)) == NULL)
return (EINVAL);
/*
* If this is a load, get the vdev guid from the nvlist.
* Otherwise, vdev_alloc_common() will generate one for us.
*/
if (alloctype == VDEV_ALLOC_LOAD) {
uint64_t label_id;
if (nvlist_lookup_uint64(nv, ZPOOL_CONFIG_ID, &label_id) ||
label_id != id)
return (EINVAL);
if (nvlist_lookup_uint64(nv, ZPOOL_CONFIG_GUID, &guid) != 0)
return (EINVAL);
} else if (alloctype == VDEV_ALLOC_SPARE) {
if (nvlist_lookup_uint64(nv, ZPOOL_CONFIG_GUID, &guid) != 0)
return (EINVAL);
}
/*
* The first allocated vdev must be of type 'root'.
*/
if (ops != &vdev_root_ops && spa->spa_root_vdev == NULL)
return (EINVAL);
vd = vdev_alloc_common(spa, id, guid, ops);
if (nvlist_lookup_string(nv, ZPOOL_CONFIG_PATH, &vd->vdev_path) == 0)
vd->vdev_path = spa_strdup(vd->vdev_path);
if (nvlist_lookup_string(nv, ZPOOL_CONFIG_DEVID, &vd->vdev_devid) == 0)
vd->vdev_devid = spa_strdup(vd->vdev_devid);
/*
* Set the nparity propery for RAID-Z vdevs.
*/
if (ops == &vdev_raidz_ops) {
if (nvlist_lookup_uint64(nv, ZPOOL_CONFIG_NPARITY,
&vd->vdev_nparity) == 0) {
/*
* Currently, we can only support 2 parity devices.
*/
if (vd->vdev_nparity > 2)
return (EINVAL);
/*
* Older versions can only support 1 parity device.
*/
if (vd->vdev_nparity == 2 &&
spa_version(spa) < ZFS_VERSION_RAID6)
return (ENOTSUP);
} else {
/*
* We require the parity to be specified for SPAs that
* support multiple parity levels.
*/
if (spa_version(spa) >= ZFS_VERSION_RAID6)
return (EINVAL);
/*
* Otherwise, we default to 1 parity device for RAID-Z.
*/
vd->vdev_nparity = 1;
}
} else {
vd->vdev_nparity = 0;
}
/*
* Set the whole_disk property. If it's not specified, leave the value
* as -1.
*/
if (nvlist_lookup_uint64(nv, ZPOOL_CONFIG_WHOLE_DISK,
&vd->vdev_wholedisk) != 0)
vd->vdev_wholedisk = -1ULL;
/*
* Look for the 'not present' flag. This will only be set if the device
* was not present at the time of import.
*/
(void) nvlist_lookup_uint64(nv, ZPOOL_CONFIG_NOT_PRESENT,
&vd->vdev_not_present);
/*
* Get the alignment requirement.
*/
(void) nvlist_lookup_uint64(nv, ZPOOL_CONFIG_ASHIFT, &vd->vdev_ashift);
/*
* If we're a top-level vdev, try to load the allocation parameters.
*/
if (parent && !parent->vdev_parent && alloctype == VDEV_ALLOC_LOAD) {
(void) nvlist_lookup_uint64(nv, ZPOOL_CONFIG_METASLAB_ARRAY,
&vd->vdev_ms_array);
(void) nvlist_lookup_uint64(nv, ZPOOL_CONFIG_METASLAB_SHIFT,
&vd->vdev_ms_shift);
(void) nvlist_lookup_uint64(nv, ZPOOL_CONFIG_ASIZE,
&vd->vdev_asize);
}
/*
* If we're a leaf vdev, try to load the DTL object and offline state.
*/
if (vd->vdev_ops->vdev_op_leaf && alloctype == VDEV_ALLOC_LOAD) {
(void) nvlist_lookup_uint64(nv, ZPOOL_CONFIG_DTL,
&vd->vdev_dtl.smo_object);
(void) nvlist_lookup_uint64(nv, ZPOOL_CONFIG_OFFLINE,
&vd->vdev_offline);
}
/*
* Add ourselves to the parent's list of children.
*/
vdev_add_child(parent, vd);
*vdp = vd;
return (0);
}
/*
* Generate the pool's configuration based on the current in-core state.
* We infer whether to generate a complete config or just one top-level config
* based on whether vd is the root vdev.
*/
nvlist_t *
spa_config_generate(spa_t *spa, vdev_t *vd, uint64_t txg, int getstats)
{
nvlist_t *config, *nvroot;
vdev_t *rvd = spa->spa_root_vdev;
unsigned long hostid = 0;
boolean_t locked = B_FALSE;
if (vd == NULL) {
vd = rvd;
locked = B_TRUE;
spa_config_enter(spa, SCL_CONFIG | SCL_STATE, FTAG, RW_READER);
}
ASSERT(spa_config_held(spa, SCL_CONFIG | SCL_STATE, RW_READER) ==
(SCL_CONFIG | SCL_STATE));
/*
* If txg is -1, report the current value of spa->spa_config_txg.
*/
if (txg == -1ULL)
txg = spa->spa_config_txg;
VERIFY(nvlist_alloc(&config, NV_UNIQUE_NAME, KM_SLEEP) == 0);
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_VERSION,
spa_version(spa)) == 0);
VERIFY(nvlist_add_string(config, ZPOOL_CONFIG_POOL_NAME,
spa_name(spa)) == 0);
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_POOL_STATE,
spa_state(spa)) == 0);
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_POOL_TXG,
txg) == 0);
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_POOL_GUID,
spa_guid(spa)) == 0);
#ifdef _KERNEL
hostid = zone_get_hostid(NULL);
#else /* _KERNEL */
/*
* We're emulating the system's hostid in userland, so we can't use
* zone_get_hostid().
*/
(void) ddi_strtoul(hw_serial, NULL, 10, &hostid);
#endif /* _KERNEL */
if (hostid != 0) {
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_HOSTID,
hostid) == 0);
}
VERIFY(nvlist_add_string(config, ZPOOL_CONFIG_HOSTNAME,
utsname.nodename) == 0);
if (vd != rvd) {
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_TOP_GUID,
vd->vdev_top->vdev_guid) == 0);
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_GUID,
vd->vdev_guid) == 0);
if (vd->vdev_isspare)
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_IS_SPARE,
1ULL) == 0);
if (vd->vdev_islog)
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_IS_LOG,
1ULL) == 0);
vd = vd->vdev_top; /* label contains top config */
}
nvroot = vdev_config_generate(spa, vd, getstats, B_FALSE, B_FALSE);
VERIFY(nvlist_add_nvlist(config, ZPOOL_CONFIG_VDEV_TREE, nvroot) == 0);
nvlist_free(nvroot);
if (locked)
spa_config_exit(spa, SCL_CONFIG | SCL_STATE, FTAG);
return (config);
}
/*
* Generate the pool's configuration based on the current in-core state.
* We infer whether to generate a complete config or just one top-level config
* based on whether vd is the root vdev.
*/
nvlist_t *
spa_config_generate(spa_t *spa, vdev_t *vd, uint64_t txg, int getstats)
{
nvlist_t *config, *nvroot;
vdev_t *rvd = spa->spa_root_vdev;
unsigned long hostid = 0;
ASSERT(spa_config_held(spa, RW_READER) ||
spa_config_held(spa, RW_WRITER));
if (vd == NULL)
vd = rvd;
/*
* If txg is -1, report the current value of spa->spa_config_txg.
*/
if (txg == -1ULL)
txg = spa->spa_config_txg;
VERIFY(nvlist_alloc(&config, NV_UNIQUE_NAME, KM_SLEEP) == 0);
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_VERSION,
spa_version(spa)) == 0);
VERIFY(nvlist_add_string(config, ZPOOL_CONFIG_POOL_NAME,
spa_name(spa)) == 0);
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_POOL_STATE,
spa_state(spa)) == 0);
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_POOL_TXG,
txg) == 0);
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_POOL_GUID,
spa_guid(spa)) == 0);
(void) ddi_strtoul(hw_serial, NULL, 10, &hostid);
if (hostid != 0) {
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_HOSTID,
hostid) == 0);
}
VERIFY(nvlist_add_string(config, ZPOOL_CONFIG_HOSTNAME,
utsname.nodename) == 0);
if (vd != rvd) {
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_TOP_GUID,
vd->vdev_top->vdev_guid) == 0);
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_GUID,
vd->vdev_guid) == 0);
if (vd->vdev_isspare)
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_IS_SPARE,
1ULL) == 0);
if (vd->vdev_islog)
VERIFY(nvlist_add_uint64(config, ZPOOL_CONFIG_IS_LOG,
1ULL) == 0);
vd = vd->vdev_top; /* label contains top config */
}
nvroot = vdev_config_generate(spa, vd, getstats, B_FALSE, B_FALSE);
VERIFY(nvlist_add_nvlist(config, ZPOOL_CONFIG_VDEV_TREE, nvroot) == 0);
nvlist_free(nvroot);
#ifdef DEBUG
if ( config ) {
printf("spa_config_generate::config=0x%x\n", config);
}
#endif
return (config);
}
int l;
if (offset >= psize - VDEV_LABEL_END_SIZE) {
offset -= psize - VDEV_LABEL_END_SIZE;
offset += (VDEV_LABELS / 2) * sizeof (vdev_label_t);
}
l = offset / sizeof (vdev_label_t);
return (l < VDEV_LABELS ? l : -1);
}
static void
vdev_label_read(zio_t *zio, vdev_t *vd, int l, abd_t *buf, uint64_t offset,
uint64_t size, zio_done_func_t *done, void *private, int flags)
{
ASSERT(
spa_config_held(zio->io_spa, SCL_STATE, RW_READER) == SCL_STATE ||
spa_config_held(zio->io_spa, SCL_STATE, RW_WRITER) == SCL_STATE);
ASSERT(flags & ZIO_FLAG_CONFIG_WRITER);
zio_nowait(zio_read_phys(zio, vd,
vdev_label_offset(vd->vdev_psize, l, offset),
size, buf, ZIO_CHECKSUM_LABEL, done, private,
ZIO_PRIORITY_SYNC_READ, flags, B_TRUE));
}
void
vdev_label_write(zio_t *zio, vdev_t *vd, int l, abd_t *buf, uint64_t offset,
uint64_t size, zio_done_func_t *done, void *private, int flags)
{
ASSERT(
spa_config_held(zio->io_spa, SCL_STATE, RW_READER) == SCL_STATE ||
