Print this page
10592 misc. metaslab and vdev related ZoL bug fixes
Portions contributed by: Jerry Jelinek <jerry.jelinek@joyent.com>
Reviewed by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed by: Giuseppe Di Natale <guss80@gmail.com>
Reviewed by: George Melikov <mail@gmelikov.ru>
Reviewed by: Paul Dagnelie <pcd@delphix.com>
Reviewed by: Matt Ahrens <mahrens@delphix.com>
Reviewed by: Pavel Zakharov <pavel.zakharov@delphix.com>
Reviewed by: Tony Hutter <hutter2@llnl.gov>
Reviewed by: Kody Kantor <kody.kantor@joyent.com>
Approved by: Dan McDonald <danmcd@joyent.com>
@@ -499,11 +499,10 @@
list_link_init(&vd->vdev_leaf_node);
mutex_init(&vd->vdev_dtl_lock, NULL, MUTEX_DEFAULT, NULL);
mutex_init(&vd->vdev_stat_lock, NULL, MUTEX_DEFAULT, NULL);
mutex_init(&vd->vdev_probe_lock, NULL, MUTEX_DEFAULT, NULL);
- mutex_init(&vd->vdev_queue_lock, NULL, MUTEX_DEFAULT, NULL);
mutex_init(&vd->vdev_initialize_lock, NULL, MUTEX_DEFAULT, NULL);
mutex_init(&vd->vdev_initialize_io_lock, NULL, MUTEX_DEFAULT, NULL);
cv_init(&vd->vdev_initialize_cv, NULL, CV_DEFAULT, NULL);
cv_init(&vd->vdev_initialize_io_cv, NULL, CV_DEFAULT, NULL);
@@ -887,11 +886,10 @@
}
range_tree_destroy(vd->vdev_obsolete_segments);
rw_destroy(&vd->vdev_indirect_rwlock);
mutex_destroy(&vd->vdev_obsolete_lock);
- mutex_destroy(&vd->vdev_queue_lock);
mutex_destroy(&vd->vdev_dtl_lock);
mutex_destroy(&vd->vdev_stat_lock);
mutex_destroy(&vd->vdev_probe_lock);
mutex_destroy(&vd->vdev_initialize_lock);
mutex_destroy(&vd->vdev_initialize_io_lock);
@@ -1249,23 +1247,26 @@
*/
vd->vdev_checkpoint_sm = NULL;
}
if (vd->vdev_ms != NULL) {
- uint64_t count = vd->vdev_ms_count;
+ metaslab_group_t *mg = vd->vdev_mg;
+ metaslab_group_passivate(mg);
- metaslab_group_passivate(vd->vdev_mg);
+ uint64_t count = vd->vdev_ms_count;
for (uint64_t m = 0; m < count; m++) {
metaslab_t *msp = vd->vdev_ms[m];
-
if (msp != NULL)
metaslab_fini(msp);
}
kmem_free(vd->vdev_ms, count * sizeof (metaslab_t *));
vd->vdev_ms = NULL;
vd->vdev_ms_count = 0;
+
+ for (int i = 0; i < RANGE_TREE_HISTOGRAM_SIZE; i++)
+ ASSERT0(mg->mg_histogram[i]);
}
ASSERT0(vd->vdev_ms_count);
}
typedef struct vdev_probe_stats {
@@ -2547,17 +2548,10 @@
if (error)
return (error);
ASSERT(vd->vdev_dtl_sm != NULL);
mutex_enter(&vd->vdev_dtl_lock);
-
- /*
- * Now that we've opened the space_map we need to update
- * the in-core DTL.
- */
- space_map_update(vd->vdev_dtl_sm);
-
error = space_map_load(vd->vdev_dtl_sm,
vd->vdev_dtl[DTL_MISSING], SM_ALLOC);
mutex_exit(&vd->vdev_dtl_lock);
return (error);
@@ -2713,14 +2707,10 @@
(u_longlong_t)space_map_object(vd->vdev_dtl_sm));
vdev_config_dirty(vd->vdev_top);
}
dmu_tx_commit(tx);
-
- mutex_enter(&vd->vdev_dtl_lock);
- space_map_update(vd->vdev_dtl_sm);
- mutex_exit(&vd->vdev_dtl_lock);
}
/*
* Determine whether the specified vdev can be offlined/detached/removed
* without losing data.
@@ -2859,11 +2849,14 @@
VDEV_AUX_CORRUPT_DATA);
vdev_dbgmsg(vd, "vdev_load: invalid size. ashift=%llu, "
"asize=%llu", (u_longlong_t)vd->vdev_ashift,
(u_longlong_t)vd->vdev_asize);
return (SET_ERROR(ENXIO));
- } else if ((error = vdev_metaslab_init(vd, 0)) != 0) {
+ }
+
+ error = vdev_metaslab_init(vd, 0);
+ if (error != 0) {
vdev_dbgmsg(vd, "vdev_load: metaslab_init failed "
"[error=%d]", error);
vdev_set_state(vd, B_FALSE, VDEV_STATE_CANT_OPEN,
VDEV_AUX_CORRUPT_DATA);
return (error);
@@ -2873,29 +2866,30 @@
if (checkpoint_sm_obj != 0) {
objset_t *mos = spa_meta_objset(vd->vdev_spa);
ASSERT(vd->vdev_asize != 0);
ASSERT3P(vd->vdev_checkpoint_sm, ==, NULL);
- if ((error = space_map_open(&vd->vdev_checkpoint_sm,
+ error = space_map_open(&vd->vdev_checkpoint_sm,
mos, checkpoint_sm_obj, 0, vd->vdev_asize,
- vd->vdev_ashift))) {
+ vd->vdev_ashift);
+ if (error != 0) {
vdev_dbgmsg(vd, "vdev_load: space_map_open "
"failed for checkpoint spacemap (obj %llu) "
"[error=%d]",
(u_longlong_t)checkpoint_sm_obj, error);
return (error);
}
ASSERT3P(vd->vdev_checkpoint_sm, !=, NULL);
- space_map_update(vd->vdev_checkpoint_sm);
/*
* Since the checkpoint_sm contains free entries
- * exclusively we can use sm_alloc to indicate the
- * culmulative checkpointed space that has been freed.
+ * exclusively we can use space_map_allocated() to
+ * indicate the cumulative checkpointed space that
+ * has been freed.
*/
vd->vdev_stat.vs_checkpoint_space =
- -vd->vdev_checkpoint_sm->sm_alloc;
+ -space_map_allocated(vd->vdev_checkpoint_sm);
vd->vdev_spa->spa_checkpoint_info.sci_dspace +=
vd->vdev_stat.vs_checkpoint_space;
}
}
@@ -2923,11 +2917,10 @@
vdev_dbgmsg(vd, "vdev_load: space_map_open failed for "
"obsolete spacemap (obj %llu) [error=%d]",
(u_longlong_t)obsolete_sm_object, error);
return (error);
}
- space_map_update(vd->vdev_obsolete_sm);
}
return (0);
}
@@ -3010,51 +3003,10 @@
ASSERT(vd->vdev_islog);
ASSERT(vd == vd->vdev_top);
ASSERT3U(txg, ==, spa_syncing_txg(spa));
- if (vd->vdev_ms != NULL) {
- metaslab_group_t *mg = vd->vdev_mg;
-
- metaslab_group_histogram_verify(mg);
- metaslab_class_histogram_verify(mg->mg_class);
-
- for (int m = 0; m < vd->vdev_ms_count; m++) {
- metaslab_t *msp = vd->vdev_ms[m];
-
- if (msp == NULL || msp->ms_sm == NULL)
- continue;
-
- mutex_enter(&msp->ms_lock);
- /*
- * If the metaslab was not loaded when the vdev
- * was removed then the histogram accounting may
- * not be accurate. Update the histogram information
- * here so that we ensure that the metaslab group
- * and metaslab class are up-to-date.
- */
- metaslab_group_histogram_remove(mg, msp);
-
- VERIFY0(space_map_allocated(msp->ms_sm));
- space_map_close(msp->ms_sm);
- msp->ms_sm = NULL;
- mutex_exit(&msp->ms_lock);
- }
-
- if (vd->vdev_checkpoint_sm != NULL) {
- ASSERT(spa_has_checkpoint(spa));
- space_map_close(vd->vdev_checkpoint_sm);
- vd->vdev_checkpoint_sm = NULL;
- }
-
- metaslab_group_histogram_verify(mg);
- metaslab_class_histogram_verify(mg->mg_class);
-
- for (int i = 0; i < RANGE_TREE_HISTOGRAM_SIZE; i++)
- ASSERT0(mg->mg_histogram[i]);
- }
-
dmu_tx_t *tx = dmu_tx_create_assigned(spa_get_dsl(spa), txg);
vdev_destroy_spacemaps(vd, tx);
if (vd->vdev_top_zap != 0) {
vdev_destroy_unlink_zap(vd, vd->vdev_top_zap, tx);
@@ -3084,29 +3036,27 @@
vdev_sync(vdev_t *vd, uint64_t txg)
{
spa_t *spa = vd->vdev_spa;
vdev_t *lvd;
metaslab_t *msp;
- dmu_tx_t *tx;
+ ASSERT3U(txg, ==, spa->spa_syncing_txg);
+ dmu_tx_t *tx = dmu_tx_create_assigned(spa->spa_dsl_pool, txg);
if (range_tree_space(vd->vdev_obsolete_segments) > 0) {
- dmu_tx_t *tx;
-
ASSERT(vd->vdev_removing ||
vd->vdev_ops == &vdev_indirect_ops);
- tx = dmu_tx_create_assigned(spa->spa_dsl_pool, txg);
vdev_indirect_sync_obsolete(vd, tx);
- dmu_tx_commit(tx);
/*
* If the vdev is indirect, it can't have dirty
* metaslabs or DTLs.
*/
if (vd->vdev_ops == &vdev_indirect_ops) {
ASSERT(txg_list_empty(&vd->vdev_ms_list, txg));
ASSERT(txg_list_empty(&vd->vdev_dtl_list, txg));
+ dmu_tx_commit(tx);
return;
}
}
ASSERT(vdev_is_concrete(vd));
@@ -3113,16 +3063,14 @@
if (vd->vdev_ms_array == 0 && vd->vdev_ms_shift != 0 &&
!vd->vdev_removing) {
ASSERT(vd == vd->vdev_top);
ASSERT0(vd->vdev_indirect_config.vic_mapping_object);
- tx = dmu_tx_create_assigned(spa->spa_dsl_pool, txg);
vd->vdev_ms_array = dmu_object_alloc(spa->spa_meta_objset,
DMU_OT_OBJECT_ARRAY, 0, DMU_OT_NONE, 0, tx);
ASSERT(vd->vdev_ms_array != 0);
vdev_config_dirty(vd);
- dmu_tx_commit(tx);
}
while ((msp = txg_list_remove(&vd->vdev_ms_list, txg)) != NULL) {
metaslab_sync(msp, txg);
(void) txg_list_add(&vd->vdev_ms_list, msp, TXG_CLEAN(txg));
@@ -3137,10 +3085,11 @@
*/
if (vd->vdev_islog && vd->vdev_stat.vs_alloc == 0 && vd->vdev_removing)
vdev_remove_empty_log(vd, txg);
(void) txg_list_add(&spa->spa_vdev_txg_list, vd, TXG_CLEAN(txg));
+ dmu_tx_commit(tx);
}
uint64_t
vdev_psize_to_asize(vdev_t *vd, uint64_t psize)
{
@@ -3366,12 +3315,10 @@
* If the log device was successfully reset but has
* checkpointed data, do not offline it.
*/
if (error == 0 &&
tvd->vdev_checkpoint_sm != NULL) {
- ASSERT3U(tvd->vdev_checkpoint_sm->sm_alloc,
- !=, 0);
error = ZFS_ERR_CHECKPOINT_EXISTS;
}
spa_vdev_state_enter(spa, SCL_ALLOC);