MFV r288063: make dataset property de-registration operation O(1)

A change to a property on a dataset must be propagated to its descendants
in case that property is inherited. For datasets whose information is
not currently loaded into memory (e.g. a snapshot that isn't currently
mounted), there is nothing to do; the property change will take effect
the next time that dataset is loaded. To handle updates to datasets that
are in-core, ZFS registers a callback entry for each property of each
loaded dataset with the dsl directory that holds that dataset. There
is a dsl directory associated with each live dataset that references
both the live dataset and any snapshots of the live dataset. A property
change is effected by doing a traversal of the tree of dsl directories
for a pool, starting at the directory sourcing the change, and invoking
these callbacks.

The current implementation both registers and de-registers properties
individually for each loaded dataset. While registration for a property is
O(1) (insert into a list), de-registration is O(n) (search list and then
remove). The 'n' for de-registration, however, is not limited to the size
(number of snapshots + 1) of the dsl directory. The eviction portion
of the life cycle for the in core state of datasets is asynchronous,
which allows multiple copies of the dataset information to be in-core
at once. Only one of these copies is active at any time with the rest
going through tear down processing, but all copies contribute to the
cost of performing a dsl_prop_unregister().

One way to create multiple, in-flight copies of dataset information
is by performing "zfs list" operations from multiple threads
concurrently. In-core dataset information is loaded on demand and then
evicted when reference counts drops to zero. For datasets that are not
mounted, there is no persistent reference count to keep them resident.
So, a list operation will load them, compute the information required to
do the list operation, and then evict them. When performing this operation
from multiple threads it is possible that some of the in-core dataset
information will be reused, but also possible to lose the race and load
the dataset again, even while the same information is being torn down.

Compounding the performance issue further is a change made for illumos
issue 5056 which made dataset eviction single threaded. In environments
using automation to manage ZFS datasets, it is now possible to create
enough of a backlog of dataset evictions to consume excessive amounts
of kernel memory and to bog down the system.

The fix employed here is to make property de-registration O(1). With this
change in place, it is hoped that a single thread is more than sufficient
to handle eviction processing. If it isn't, the problem can be solved
by increasing the number of threads devoted to the eviction taskq.

sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dsl_dataset.c
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dsl_dir.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dsl_prop.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/sys/dsl_dataset.h:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/sys/dsl_dir.h:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/sys/dsl_prop.h:
    Associate dsl property callback records with both the
    dsl directory and the dsl dataset that is registering the
    callback. Both connections are protected by the dsl directory's
    "dd_lock".

    When linking callbacks into a dsl directory, group them by
    the property type. This helps reduce the space penalty for the
    double association (the property name pointer is stored once
    per dsl_dir instead of in each record) and reduces the number of
    strcmp() calls required to do callback processing when updating
    a single property. Property types are stored in a linked list
    since currently ZFS registers a maximum of 10 property types
    for each dataset.

    Note that the property buckets/records associated with a dsl
    directory are created on demand, but only freed when the dsl
    directory is freed. Given the static nature of property types
    and their small number, there is no benefit to freeing the few
    bytes of memory used to represent the property record earlier.
    When a property record becomes empty, the dsl directory is either
    going to become unreferenced a little later in this thread of
    execution, or there is a high chance that another dataset is
    going to be loaded that would recreate the bucket anyway.

    Replace dsl_prop_unregister() with dsl_prop_unregister_all().
    All callers of dsl_prop_unregister() are trying to remove
    all property registrations for a given dsl dataset anyway. By
    changing the API, we can avoid doing any lookups of callbacks
    by property type and just traverse the list of all callbacks
    for the dataset and free each one.

sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dmu_objset.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_vfsops.c:
    Replace use of dsl_prop_unregister() with the new
    dsl_prop_unregister_all() API.

illumos/illumos-gate@03bad06fbb
    Author: Justin Gibbs <gibbs@scsiguy.com>
    Reviewed by: Matthew Ahrens <mahrens@delphix.com>
    Reviewed by: Prakash Surya <prakash.surya@delphix.com>
    Approved by: Dan McDonald <danmcd@omniti.com>

Illumos issue:
    6171 dsl_prop_unregister() slows down dataset eviction
    https://www.illumos.org/issues/6171

MFC after:	2 weeks

sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dmu_objset.c

@@ -680,40 +680,8 @@ dmu_objset_evict(objset_t *os)
 	for (int t = 0; t < TXG_SIZE; t++)
 		ASSERT(!dmu_objset_is_dirty(os, t));
 
-	if (ds) {
-		if (!ds->ds_is_snapshot) {
-			VERIFY0(dsl_prop_unregister(ds,
-			    zfs_prop_to_name(ZFS_PROP_CHECKSUM),
-			    checksum_changed_cb, os));
-			VERIFY0(dsl_prop_unregister(ds,
-			    zfs_prop_to_name(ZFS_PROP_COMPRESSION),
-			    compression_changed_cb, os));
-			VERIFY0(dsl_prop_unregister(ds,
-			    zfs_prop_to_name(ZFS_PROP_COPIES),
-			    copies_changed_cb, os));
-			VERIFY0(dsl_prop_unregister(ds,
-			    zfs_prop_to_name(ZFS_PROP_DEDUP),
-			    dedup_changed_cb, os));
-			VERIFY0(dsl_prop_unregister(ds,
-			    zfs_prop_to_name(ZFS_PROP_LOGBIAS),
-			    logbias_changed_cb, os));
-			VERIFY0(dsl_prop_unregister(ds,
-			    zfs_prop_to_name(ZFS_PROP_SYNC),
-			    sync_changed_cb, os));
-			VERIFY0(dsl_prop_unregister(ds,
-			    zfs_prop_to_name(ZFS_PROP_REDUNDANT_METADATA),
-			    redundant_metadata_changed_cb, os));
-			VERIFY0(dsl_prop_unregister(ds,
-			    zfs_prop_to_name(ZFS_PROP_RECORDSIZE),
-			    recordsize_changed_cb, os));
-		}
-		VERIFY0(dsl_prop_unregister(ds,
-		    zfs_prop_to_name(ZFS_PROP_PRIMARYCACHE),
-		    primary_cache_changed_cb, os));
-		VERIFY0(dsl_prop_unregister(ds,
-		    zfs_prop_to_name(ZFS_PROP_SECONDARYCACHE),
-		    secondary_cache_changed_cb, os));
-	}
+	if (ds)
+		dsl_prop_unregister_all(ds, os);
 
 	if (os->os_sa)
 		sa_tear_down(os);

sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dsl_dataset.c

@@ -288,6 +288,7 @@ dsl_dataset_evict(void *dbu)
 
 	ASSERT(!list_link_active(&ds->ds_synced_link));
 
+	list_destroy(&ds->ds_prop_cbs);
 	if (mutex_owned(&ds->ds_lock))
 		mutex_exit(&ds->ds_lock);
 	mutex_destroy(&ds->ds_lock);
@@ -434,6 +435,9 @@ dsl_dataset_hold_obj(dsl_pool_t *dp, uint64_t dsobj, void *tag,
 		list_create(&ds->ds_sendstreams, sizeof (dmu_sendarg_t),
 		    offsetof(dmu_sendarg_t, dsa_link));
 
+		list_create(&ds->ds_prop_cbs, sizeof (dsl_prop_cb_record_t),
+		    offsetof(dsl_prop_cb_record_t, cbr_ds_node));
+
 		if (doi.doi_type == DMU_OTN_ZAP_METADATA) {
 			for (spa_feature_t f = 0; f < SPA_FEATURES; f++) {
 				if (!(spa_feature_table[f].fi_flags &

sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dsl_dir.c

@@ -152,11 +152,7 @@ dsl_dir_evict(void *dbu)
 
 	spa_async_close(dd->dd_pool->dp_spa, dd);
 
-	/*
-	 * The props callback list should have been cleaned up by
-	 * objset_evict().
-	 */
-	list_destroy(&dd->dd_prop_cbs);
+	dsl_prop_fini(dd);
 	mutex_destroy(&dd->dd_lock);
 	kmem_free(dd, sizeof (dsl_dir_t));
 }
@@ -191,9 +187,7 @@ dsl_dir_hold_obj(dsl_pool_t *dp, uint64_t ddobj,
 		dd->dd_dbuf = dbuf;
 		dd->dd_pool = dp;
 		mutex_init(&dd->dd_lock, NULL, MUTEX_DEFAULT, NULL);
-
-		list_create(&dd->dd_prop_cbs, sizeof (dsl_prop_cb_record_t),
-		    offsetof(dsl_prop_cb_record_t, cbr_node));
+		dsl_prop_init(dd);
 
 		dsl_dir_snap_cmtime_update(dd);
 
@@ -251,6 +245,7 @@ dsl_dir_hold_obj(dsl_pool_t *dp, uint64_t ddobj,
 		if (winner != NULL) {
 			if (dd->dd_parent)
 				dsl_dir_rele(dd->dd_parent, dd);
+			dsl_prop_fini(dd);
 			mutex_destroy(&dd->dd_lock);
 			kmem_free(dd, sizeof (dsl_dir_t));
 			dd = winner;
@@ -278,6 +273,7 @@ dsl_dir_hold_obj(dsl_pool_t *dp, uint64_t ddobj,
 errout:
 	if (dd->dd_parent)
 		dsl_dir_rele(dd->dd_parent, dd);
+	dsl_prop_fini(dd);
 	mutex_destroy(&dd->dd_lock);
 	kmem_free(dd, sizeof (dsl_dir_t));
 	dmu_buf_rele(dbuf, tag);

sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dsl_prop.c

@@ -215,6 +215,58 @@ dsl_prop_get_ds(dsl_dataset_t *ds, const char *propname,
 	    intsz, numints, buf, setpoint, ds->ds_is_snapshot));
 }
 
+static dsl_prop_record_t *
+dsl_prop_record_find(dsl_dir_t *dd, const char *propname)
+{
+	dsl_prop_record_t *pr = NULL;
+
+	ASSERT(MUTEX_HELD(&dd->dd_lock));
+
+	for (pr = list_head(&dd->dd_props);
+	    pr != NULL; pr = list_next(&dd->dd_props, pr)) {
+		if (strcmp(pr->pr_propname, propname) == 0)
+			break;
+	}
+
+	return (pr);
+}
+
+static dsl_prop_record_t *
+dsl_prop_record_create(dsl_dir_t *dd, const char *propname)
+{
+	dsl_prop_record_t *pr;
+
+	ASSERT(MUTEX_HELD(&dd->dd_lock));
+
+	pr = kmem_alloc(sizeof (dsl_prop_record_t), KM_SLEEP);
+	pr->pr_propname = spa_strdup(propname);
+	list_create(&pr->pr_cbs, sizeof (dsl_prop_cb_record_t),
+	    offsetof(dsl_prop_cb_record_t, cbr_pr_node));
+	list_insert_head(&dd->dd_props, pr);
+
+	return (pr);
+}
+
+void
+dsl_prop_init(dsl_dir_t *dd)
+{
+	list_create(&dd->dd_props, sizeof (dsl_prop_record_t),
+	    offsetof(dsl_prop_record_t, pr_node));
+}
+
+void
+dsl_prop_fini(dsl_dir_t *dd)
+{
+	dsl_prop_record_t *pr;
+
+	while ((pr = list_remove_head(&dd->dd_props)) != NULL) {
+		list_destroy(&pr->pr_cbs);
+		strfree((char *)pr->pr_propname);
+		kmem_free(pr, sizeof (dsl_prop_record_t));
+	}
+	list_destroy(&dd->dd_props);
+}
+
 /*
  * Register interest in the named property.  We'll call the callback
  * once to notify it of the current property value, and again each time
@@ -229,6 +281,7 @@ dsl_prop_register(dsl_dataset_t *ds, const char *propname,
 	dsl_dir_t *dd = ds->ds_dir;
 	dsl_pool_t *dp = dd->dd_pool;
 	uint64_t value;
+	dsl_prop_record_t *pr;
 	dsl_prop_cb_record_t *cbr;
 	int err;
 
@@ -240,12 +293,16 @@ dsl_prop_register(dsl_dataset_t *ds, const char *propname,
 
 	cbr = kmem_alloc(sizeof (dsl_prop_cb_record_t), KM_SLEEP);
 	cbr->cbr_ds = ds;
-	cbr->cbr_propname = kmem_alloc(strlen(propname)+1, KM_SLEEP);
-	(void) strcpy((char *)cbr->cbr_propname, propname);
 	cbr->cbr_func = callback;
 	cbr->cbr_arg = cbarg;
+
 	mutex_enter(&dd->dd_lock);
-	list_insert_head(&dd->dd_prop_cbs, cbr);
+	pr = dsl_prop_record_find(dd, propname);
+	if (pr == NULL)
+		pr = dsl_prop_record_create(dd, propname);
+	cbr->cbr_pr = pr;
+	list_insert_head(&pr->pr_cbs, cbr);
+	list_insert_head(&ds->ds_prop_cbs, cbr);
 	mutex_exit(&dd->dd_lock);
 
 	cbr->cbr_func(cbr->cbr_arg, value);
@@ -376,56 +433,34 @@ dsl_prop_predict(dsl_dir_t *dd, const char *propname,
 }
 
 /*
- * Unregister this callback.  Return 0 on success, ENOENT if ddname is
- * invalid, or ENOMSG if no matching callback registered.
+ * Unregister all callbacks that are registered with the
+ * given callback argument.
  */
-int
-dsl_prop_unregister(dsl_dataset_t *ds, const char *propname,
-    dsl_prop_changed_cb_t *callback, void *cbarg)
+void
+dsl_prop_unregister_all(dsl_dataset_t *ds, void *cbarg)
 {
+	dsl_prop_cb_record_t *cbr, *next_cbr;
+
 	dsl_dir_t *dd = ds->ds_dir;
-	dsl_prop_cb_record_t *cbr;
 
 	mutex_enter(&dd->dd_lock);
-	for (cbr = list_head(&dd->dd_prop_cbs);
-	    cbr; cbr = list_next(&dd->dd_prop_cbs, cbr)) {
-		if (cbr->cbr_ds == ds &&
-		    cbr->cbr_func == callback &&
-		    cbr->cbr_arg == cbarg &&
-		    strcmp(cbr->cbr_propname, propname) == 0)
-			break;
+	next_cbr = list_head(&ds->ds_prop_cbs);
+	while (next_cbr != NULL) {
+		cbr = next_cbr;
+		next_cbr = list_next(&ds->ds_prop_cbs, cbr);
+		if (cbr->cbr_arg == cbarg) {
+			list_remove(&ds->ds_prop_cbs, cbr);
+			list_remove(&cbr->cbr_pr->pr_cbs, cbr);
+			kmem_free(cbr, sizeof (dsl_prop_cb_record_t));
+		}
 	}
-
-	if (cbr == NULL) {
-		mutex_exit(&dd->dd_lock);
-		return (SET_ERROR(ENOMSG));
-	}
-
-	list_remove(&dd->dd_prop_cbs, cbr);
 	mutex_exit(&dd->dd_lock);
-	kmem_free((void*)cbr->cbr_propname, strlen(cbr->cbr_propname)+1);
-	kmem_free(cbr, sizeof (dsl_prop_cb_record_t));
-
-	return (0);
 }
 
 boolean_t
 dsl_prop_hascb(dsl_dataset_t *ds)
 {
-	dsl_dir_t *dd = ds->ds_dir;
-	boolean_t rv = B_FALSE;
-	dsl_prop_cb_record_t *cbr;
-
-	mutex_enter(&dd->dd_lock);
-	for (cbr = list_head(&dd->dd_prop_cbs); cbr;
-	    cbr = list_next(&dd->dd_prop_cbs, cbr)) {
-		if (cbr->cbr_ds == ds) {
-			rv = B_TRUE;
-			break;
-		}
-	}
-	mutex_exit(&dd->dd_lock);
-	return (rv);
+	return (!list_is_empty(&ds->ds_prop_cbs));
 }
 
 /* ARGSUSED */
@@ -433,38 +468,50 @@ static int
 dsl_prop_notify_all_cb(dsl_pool_t *dp, dsl_dataset_t *ds, void *arg)
 {
 	dsl_dir_t *dd = ds->ds_dir;
+	dsl_prop_record_t *pr;
 	dsl_prop_cb_record_t *cbr;
 
 	mutex_enter(&dd->dd_lock);
-	for (cbr = list_head(&dd->dd_prop_cbs); cbr;
-	    cbr = list_next(&dd->dd_prop_cbs, cbr)) {
-		uint64_t value;
+	for (pr = list_head(&dd->dd_props);
+	    pr; pr = list_next(&dd->dd_props, pr)) {
+		for (cbr = list_head(&pr->pr_cbs); cbr;
+		    cbr = list_next(&pr->pr_cbs, cbr)) {
+			uint64_t value;
 
-		/*
-		 * Callback entries do not have holds on their datasets
-		 * so that datasets with registered callbacks are still
-		 * eligible for eviction.  Unlike operations on callbacks
-		 * for a single dataset, we are performing a recursive
-		 * descent of related datasets and the calling context
-		 * for this iteration only has a dataset hold on the root.
-		 * Without a hold, the callback's pointer to the dataset
-		 * could be invalidated by eviction at any time.
-		 *
-		 * Use dsl_dataset_try_add_ref() to verify that the
-		 * dataset has not begun eviction processing and to
-		 * prevent eviction from occurring for the duration
-		 * of the callback.  If the hold attempt fails, this
-		 * object is already being evicted and the callback can
-		 * be safely ignored.
-		 */
-		if (!dsl_dataset_try_add_ref(dp, cbr->cbr_ds, FTAG))
-			continue;
+			/*
+			 * Callback entries do not have holds on their
+			 * datasets so that datasets with registered
+			 * callbacks are still eligible for eviction.
+			 * Unlike operations to update properties on a
+			 * single dataset, we are performing a recursive
+			 * descent of related head datasets.  The caller
+			 * of this function only has a dataset hold on
+			 * the passed in head dataset, not the snapshots
+			 * associated with this dataset.  Without a hold,
+			 * the dataset pointer within callback records
+			 * for snapshots can be invalidated by eviction
+			 * at any time.
+			 *
+			 * Use dsl_dataset_try_add_ref() to verify
+			 * that the dataset for a snapshot has not
+			 * begun eviction processing and to prevent
+			 * eviction from occurring for the duration of
+			 * the callback.  If the hold attempt fails,
+			 * this object is already being evicted and the
+			 * callback can be safely ignored.
+			 */
+			if (ds != cbr->cbr_ds &&
+			    !dsl_dataset_try_add_ref(dp, cbr->cbr_ds, FTAG))
+				continue;
 
-		if (dsl_prop_get_ds(cbr->cbr_ds, cbr->cbr_propname,
-		    sizeof (value), 1, &value, NULL) == 0)
-			cbr->cbr_func(cbr->cbr_arg, value);
+			if (dsl_prop_get_ds(cbr->cbr_ds,
+			    cbr->cbr_pr->pr_propname, sizeof (value), 1,
+			    &value, NULL) == 0)
+				cbr->cbr_func(cbr->cbr_arg, value);
 
-		dsl_dataset_rele(cbr->cbr_ds, FTAG);
+			if (ds != cbr->cbr_ds)
+				dsl_dataset_rele(cbr->cbr_ds, FTAG);
+		}
 	}
 	mutex_exit(&dd->dd_lock);
 
@@ -489,6 +536,7 @@ dsl_prop_changed_notify(dsl_pool_t *dp, uint64_t ddobj,
     const char *propname, uint64_t value, int first)
 {
 	dsl_dir_t *dd;
+	dsl_prop_record_t *pr;
 	dsl_prop_cb_record_t *cbr;
 	objset_t *mos = dp->dp_meta_objset;
 	zap_cursor_t zc;
@@ -515,30 +563,33 @@ dsl_prop_changed_notify(dsl_pool_t *dp, uint64_t ddobj,
 	}
 
 	mutex_enter(&dd->dd_lock);
-	for (cbr = list_head(&dd->dd_prop_cbs); cbr;
-	    cbr = list_next(&dd->dd_prop_cbs, cbr)) {
-		uint64_t propobj;
+	pr = dsl_prop_record_find(dd, propname);
+	if (pr != NULL) {
+		for (cbr = list_head(&pr->pr_cbs); cbr;
+		    cbr = list_next(&pr->pr_cbs, cbr)) {
+			uint64_t propobj;
 
-		/*
-		 * cbr->cbf_ds may be invalidated due to eviction,
-		 * requiring the use of dsl_dataset_try_add_ref().
-		 * See comment block in dsl_prop_notify_all_cb()
-		 * for details.
-		 */
-		if (strcmp(cbr->cbr_propname, propname) != 0 ||
-		    !dsl_dataset_try_add_ref(dp, cbr->cbr_ds, FTAG))
-			continue;
+			/*
+			 * cbr->cbr_ds may be invalidated due to eviction,
+			 * requiring the use of dsl_dataset_try_add_ref().
+			 * See comment block in dsl_prop_notify_all_cb()
+			 * for details.
+			 */
+			if (!dsl_dataset_try_add_ref(dp, cbr->cbr_ds, FTAG))
+				continue;
 
-		propobj = dsl_dataset_phys(cbr->cbr_ds)->ds_props_obj;
+			propobj = dsl_dataset_phys(cbr->cbr_ds)->ds_props_obj;
 
-		/*
-		 * If the property is not set on this ds, then it is
-		 * inherited here; call the callback.
-		 */
-		if (propobj == 0 || zap_contains(mos, propobj, propname) != 0)
-			cbr->cbr_func(cbr->cbr_arg, value);
+			/*
+			 * If the property is not set on this ds, then it is
+			 * inherited here; call the callback.
+			 */
+			if (propobj == 0 ||
+			    zap_contains(mos, propobj, propname) != 0)
+				cbr->cbr_func(cbr->cbr_arg, value);
 
-		dsl_dataset_rele(cbr->cbr_ds, FTAG);
+			dsl_dataset_rele(cbr->cbr_ds, FTAG);
+		}
 	}
 	mutex_exit(&dd->dd_lock);
 
@@ -678,10 +729,10 @@ dsl_prop_set_sync_impl(dsl_dataset_t *ds, const char *propname,
 			 * ds here.
 			 */
 			mutex_enter(&ds->ds_dir->dd_lock);
-			for (cbr = list_head(&ds->ds_dir->dd_prop_cbs); cbr;
-			    cbr = list_next(&ds->ds_dir->dd_prop_cbs, cbr)) {
-				if (cbr->cbr_ds == ds &&
-				    strcmp(cbr->cbr_propname, propname) == 0)
+			for (cbr = list_head(&ds->ds_prop_cbs); cbr;
+			    cbr = list_next(&ds->ds_prop_cbs, cbr)) {
+				if (strcmp(cbr->cbr_pr->pr_propname,
+				    propname) == 0)
 					cbr->cbr_func(cbr->cbr_arg, intval);
 			}
 			mutex_exit(&ds->ds_dir->dd_lock);

sys/cddl/contrib/opensolaris/uts/common/fs/zfs/sys/dsl_dataset.h

@@ -184,6 +184,9 @@ typedef struct dsl_dataset {
 	kmutex_t ds_sendstream_lock;
 	list_t ds_sendstreams;
 
+	/* Protected by our dsl_dir's dd_lock */
+	list_t ds_prop_cbs;
+
 	/*
 	 * For ZFEATURE_FLAG_PER_DATASET features, set if this dataset
 	 * uses this feature.

sys/cddl/contrib/opensolaris/uts/common/fs/zfs/sys/dsl_dir.h

@@ -102,7 +102,7 @@ struct dsl_dir {
 
 	/* Protected by dd_lock */
 	kmutex_t dd_lock;
-	list_t dd_prop_cbs; /* list of dsl_prop_cb_record_t's */
+	list_t dd_props; /* list of dsl_prop_record_t's */
 	timestruc_t dd_snap_cmtime; /* last time snapshot namespace changed */
 	uint64_t dd_origin_txg;
 

sys/cddl/contrib/opensolaris/uts/common/fs/zfs/sys/dsl_prop.h

@@ -41,10 +41,17 @@ struct dsl_dir;
 /* The callback func may not call into the DMU or DSL! */
 typedef void (dsl_prop_changed_cb_t)(void *arg, uint64_t newval);
 
+typedef struct dsl_prop_record {
+	list_node_t pr_node; /* link on dd_props */
+	const char *pr_propname;
+	list_t pr_cbs;
+} dsl_prop_record_t;
+
 typedef struct dsl_prop_cb_record {
-	list_node_t cbr_node; /* link on dd_prop_cbs */
+	list_node_t cbr_pr_node; /* link on pr_cbs */
+	list_node_t cbr_ds_node; /* link on ds_prop_cbs */
+	dsl_prop_record_t *cbr_pr;
 	struct dsl_dataset *cbr_ds;
-	const char *cbr_propname;
 	dsl_prop_changed_cb_t *cbr_func;
 	void *cbr_arg;
 } dsl_prop_cb_record_t;
@@ -54,10 +61,11 @@ typedef struct dsl_props_arg {
 	zprop_source_t pa_source;
 } dsl_props_arg_t;
 
+void dsl_prop_init(dsl_dir_t *dd);
+void dsl_prop_fini(dsl_dir_t *dd);
 int dsl_prop_register(struct dsl_dataset *ds, const char *propname,
     dsl_prop_changed_cb_t *callback, void *cbarg);
-int dsl_prop_unregister(struct dsl_dataset *ds, const char *propname,
-    dsl_prop_changed_cb_t *callback, void *cbarg);
+void dsl_prop_unregister_all(struct dsl_dataset *ds, void *cbarg);
 void dsl_prop_notify_all(struct dsl_dir *dd);
 boolean_t dsl_prop_hascb(struct dsl_dataset *ds);
 

sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_vfsops.c

@@ -555,35 +555,7 @@ zfs_register_callbacks(vfs_t *vfsp)
 	return (0);
 
 unregister:
-	/*
-	 * We may attempt to unregister some callbacks that are not
-	 * registered, but this is OK; it will simply return ENOMSG,
-	 * which we will ignore.
-	 */
-	(void) dsl_prop_unregister(ds, zfs_prop_to_name(ZFS_PROP_ATIME),
-	    atime_changed_cb, zfsvfs);
-	(void) dsl_prop_unregister(ds, zfs_prop_to_name(ZFS_PROP_XATTR),
-	    xattr_changed_cb, zfsvfs);
-	(void) dsl_prop_unregister(ds, zfs_prop_to_name(ZFS_PROP_RECORDSIZE),
-	    blksz_changed_cb, zfsvfs);
-	(void) dsl_prop_unregister(ds, zfs_prop_to_name(ZFS_PROP_READONLY),
-	    readonly_changed_cb, zfsvfs);
-#ifdef illumos
-	(void) dsl_prop_unregister(ds, zfs_prop_to_name(ZFS_PROP_DEVICES),
-	    devices_changed_cb, zfsvfs);
-#endif
-	(void) dsl_prop_unregister(ds, zfs_prop_to_name(ZFS_PROP_SETUID),
-	    setuid_changed_cb, zfsvfs);
-	(void) dsl_prop_unregister(ds, zfs_prop_to_name(ZFS_PROP_EXEC),
-	    exec_changed_cb, zfsvfs);
-	(void) dsl_prop_unregister(ds, zfs_prop_to_name(ZFS_PROP_SNAPDIR),
-	    snapdir_changed_cb, zfsvfs);
-	(void) dsl_prop_unregister(ds, zfs_prop_to_name(ZFS_PROP_ACLMODE),
-	    acl_mode_changed_cb, zfsvfs);
-	(void) dsl_prop_unregister(ds, zfs_prop_to_name(ZFS_PROP_ACLINHERIT),
-	    acl_inherit_changed_cb, zfsvfs);
-	(void) dsl_prop_unregister(ds, zfs_prop_to_name(ZFS_PROP_VSCAN),
-	    vscan_changed_cb, zfsvfs);
+	dsl_prop_unregister_all(ds, zfsvfs);
 	return (error);
 }
 
@@ -1263,43 +1235,9 @@ void
 zfs_unregister_callbacks(zfsvfs_t *zfsvfs)
 {
 	objset_t *os = zfsvfs->z_os;
-	struct dsl_dataset *ds;
 
-	/*
-	 * Unregister properties.
-	 */
-	if (!dmu_objset_is_snapshot(os)) {
-		ds = dmu_objset_ds(os);
-		VERIFY(dsl_prop_unregister(ds, "atime", atime_changed_cb,
-		    zfsvfs) == 0);
-
-		VERIFY(dsl_prop_unregister(ds, "xattr", xattr_changed_cb,
-		    zfsvfs) == 0);
-
-		VERIFY(dsl_prop_unregister(ds, "recordsize", blksz_changed_cb,
-		    zfsvfs) == 0);
-
-		VERIFY(dsl_prop_unregister(ds, "readonly", readonly_changed_cb,
-		    zfsvfs) == 0);
-
-		VERIFY(dsl_prop_unregister(ds, "setuid", setuid_changed_cb,
-		    zfsvfs) == 0);
-
-		VERIFY(dsl_prop_unregister(ds, "exec", exec_changed_cb,
-		    zfsvfs) == 0);
-
-		VERIFY(dsl_prop_unregister(ds, "snapdir", snapdir_changed_cb,
-		    zfsvfs) == 0);
-
-		VERIFY(dsl_prop_unregister(ds, "aclmode", acl_mode_changed_cb,
-		    zfsvfs) == 0);
-
-		VERIFY(dsl_prop_unregister(ds, "aclinherit",
-		    acl_inherit_changed_cb, zfsvfs) == 0);
-
-		VERIFY(dsl_prop_unregister(ds, "vscan",
-		    vscan_changed_cb, zfsvfs) == 0);
-	}
+	if (!dmu_objset_is_snapshot(os))
+		dsl_prop_unregister_all(dmu_objset_ds(os), zfsvfs);
 }
 
 #ifdef SECLABEL