[illumos-gate merge]

commit cd485b49201b16c079663125308af274b6299e96
    5630 stale bonus buffer in recycled dnode_t leads to data corruption

4 files changed, 62 insertions, 9 deletions

diff --git a/usr/src/uts/common/fs/zfs/dbuf.c b/usr/src/uts/common/fs/zfs/dbuf.c
index e569cde5f1..4b644f7479 100644
--- a/usr/src/uts/common/fs/zfs/dbuf.c
+++ b/usr/src/uts/common/fs/zfs/dbuf.c
@@ -2212,21 +2212,60 @@ dbuf_rele_and_unlock(dmu_buf_impl_t *db, void *tag)
 
 	if (holds == 0) {
 		if (db->db_blkid == DMU_BONUS_BLKID) {
-			mutex_exit(&db->db_mtx);
+			dnode_t *dn;
 
 			/*
-			 * If the dnode moves here, we cannot cross this barrier
-			 * until the move completes.
+			 * If the dnode moves here, we cannot cross this
+			 * barrier until the move completes.
 			 */
 			DB_DNODE_ENTER(db);
-			atomic_dec_32(&DB_DNODE(db)->dn_dbufs_count);
+
+			dn = DB_DNODE(db);
+			atomic_dec_32(&dn->dn_dbufs_count);
+
+			/*
+			 * Decrementing the dbuf count means that the bonus
+			 * buffer's dnode hold is no longer discounted in
+			 * dnode_move(). The dnode cannot move until after
+			 * the dnode_rele_and_unlock() below.
+			 */
 			DB_DNODE_EXIT(db);
+
+			/*
+			 * Do not reference db after its lock is dropped.
+			 * Another thread may evict it.
+			 */
+			mutex_exit(&db->db_mtx);
+
 			/*
-			 * The bonus buffer's dnode hold is no longer discounted
-			 * in dnode_move(). The dnode cannot move until after
-			 * the dnode_rele().
+			 * If the dnode has been freed, evict the bonus
+			 * buffer immediately.	The data in the bonus
+			 * buffer is no longer relevant and this prevents
+			 * a stale bonus buffer from being associated
+			 * with this dnode_t should the dnode_t be reused
+			 * prior to being destroyed.
 			 */
-			dnode_rele(DB_DNODE(db), db);
+			mutex_enter(&dn->dn_mtx);
+			if (dn->dn_type == DMU_OT_NONE ||
+			    dn->dn_free_txg != 0) {
+				/*
+				 * Drop dn_mtx.  It is a leaf lock and
+				 * cannot be held when dnode_evict_bonus()
+				 * acquires other locks in order to
+				 * perform the eviction.
+				 *
+				 * Freed dnodes cannot be reused until the
+				 * last hold is released.  Since this bonus
+				 * buffer has a hold, the dnode will remain
+				 * in the free state, even without dn_mtx
+				 * held, until the dnode_rele_and_unlock()
+				 * below.
+				 */
+				mutex_exit(&dn->dn_mtx);
+				dnode_evict_bonus(dn);
+				mutex_enter(&dn->dn_mtx);
+			}
+			dnode_rele_and_unlock(dn, db);
 		} else if (db->db_buf == NULL) {
 			/*
 			 * This is a special case: we never associated this
diff --git a/usr/src/uts/common/fs/zfs/dnode.c b/usr/src/uts/common/fs/zfs/dnode.c
index 4e376fc4d9..12d8db76fa 100644
--- a/usr/src/uts/common/fs/zfs/dnode.c
+++ b/usr/src/uts/common/fs/zfs/dnode.c
@@ -1223,12 +1223,18 @@ dnode_add_ref(dnode_t *dn, void *tag)
 void
 dnode_rele(dnode_t *dn, void *tag)
 {
+	mutex_enter(&dn->dn_mtx);
+	dnode_rele_and_unlock(dn, tag);
+}
+
+void
+dnode_rele_and_unlock(dnode_t *dn, void *tag)
+{
 	uint64_t refs;
 	/* Get while the hold prevents the dnode from moving. */
 	dmu_buf_impl_t *db = dn->dn_dbuf;
 	dnode_handle_t *dnh = dn->dn_handle;
 
-	mutex_enter(&dn->dn_mtx);
 	refs = refcount_remove(&dn->dn_holds, tag);
 	mutex_exit(&dn->dn_mtx);
 
diff --git a/usr/src/uts/common/fs/zfs/dnode_sync.c b/usr/src/uts/common/fs/zfs/dnode_sync.c
index 6a0f8cc3bf..bb18718bed 100644
--- a/usr/src/uts/common/fs/zfs/dnode_sync.c
+++ b/usr/src/uts/common/fs/zfs/dnode_sync.c
@@ -430,6 +430,12 @@ dnode_evict_dbufs(dnode_t *dn)
 	}
 	mutex_exit(&dn->dn_dbufs_mtx);
 
+	dnode_evict_bonus(dn);
+}
+
+void
+dnode_evict_bonus(dnode_t *dn)
+{
 	rw_enter(&dn->dn_struct_rwlock, RW_WRITER);
 	if (dn->dn_bonus && refcount_is_zero(&dn->dn_bonus->db_holds)) {
 		mutex_enter(&dn->dn_bonus->db_mtx);
diff --git a/usr/src/uts/common/fs/zfs/sys/dnode.h b/usr/src/uts/common/fs/zfs/sys/dnode.h
index 406954a653..69cc54dc27 100644
--- a/usr/src/uts/common/fs/zfs/sys/dnode.h
+++ b/usr/src/uts/common/fs/zfs/sys/dnode.h
@@ -281,6 +281,7 @@ int dnode_hold_impl(struct objset *dd, uint64_t object, int flag,
     void *ref, dnode_t **dnp);
 boolean_t dnode_add_ref(dnode_t *dn, void *ref);
 void dnode_rele(dnode_t *dn, void *ref);
+void dnode_rele_and_unlock(dnode_t *dn, void *tag);
 void dnode_setdirty(dnode_t *dn, dmu_tx_t *tx);
 void dnode_sync(dnode_t *dn, dmu_tx_t *tx);
 void dnode_allocate(dnode_t *dn, dmu_object_type_t ot, int blocksize, int ibs,
@@ -302,6 +303,7 @@ void dnode_fini(void);
 int dnode_next_offset(dnode_t *dn, int flags, uint64_t *off,
     int minlvl, uint64_t blkfill, uint64_t txg);
 void dnode_evict_dbufs(dnode_t *dn);
+void dnode_evict_bonus(dnode_t *dn);
 
 #ifdef ZFS_DEBUG