diff --git a/include/sys/zfs_znode.h b/include/sys/zfs_znode.h
index 4e0b611bd9..b028e2eb2d 100644
--- a/include/sys/zfs_znode.h
+++ b/include/sys/zfs_znode.h
@@ -323,7 +323,7 @@ extern void	zfs_zinactive(znode_t *);
 extern void	zfs_znode_delete(znode_t *, dmu_tx_t *);
 extern void	zfs_remove_op_tables(void);
 extern int	zfs_create_op_tables(void);
-extern int	zfs_sync(zfs_sb_t *, short, cred_t *);
+extern int	zfs_sync(struct super_block *, int, cred_t *);
 extern dev_t	zfs_cmpldev(uint64_t);
 extern int	zfs_get_zplprop(objset_t *os, zfs_prop_t prop, uint64_t *value);
 extern int	zfs_get_stats(objset_t *os, nvlist_t *nv);
diff --git a/module/zfs/zfs_vfsops.c b/module/zfs/zfs_vfsops.c
index ea7e9cec93..1e8e6af9ab 100644
--- a/module/zfs/zfs_vfsops.c
+++ b/module/zfs/zfs_vfsops.c
@@ -69,8 +69,10 @@
 
 /*ARGSUSED*/
 int
-zfs_sync(zfs_sb_t *zsb, short flag, cred_t *cr)
+zfs_sync(struct super_block *sb, int wait, cred_t *cr)
 {
+	zfs_sb_t *zsb = sb->s_fs_info;
+
 	/*
 	 * Data integrity is job one.  We don't want a compromised kernel
 	 * writing to the storage pool, so we never sync during panic.
@@ -78,6 +80,13 @@ zfs_sync(zfs_sb_t *zsb, short flag, cred_t *cr)
 	if (unlikely(oops_in_progress))
 		return (0);
 
+	/*
+	 * Semantically, the only requirement is that the sync be initiated.
+	 * The DMU syncs out txgs frequently, so there's nothing to do.
+	 */
+	if (!wait)
+		return (0);
+
 	if (zsb != NULL) {
 		/*
 		 * Sync a specific filesystem.
@@ -87,19 +96,14 @@ zfs_sync(zfs_sb_t *zsb, short flag, cred_t *cr)
 		ZFS_ENTER(zsb);
 		dp = dmu_objset_pool(zsb->z_os);
 
-#ifdef HAVE_SHUTDOWN
 		/*
 		 * If the system is shutting down, then skip any
 		 * filesystems which may exist on a suspended pool.
-		 *
-		 * XXX: This can be implemented using the Linux reboot
-		 *      notifiers: {un}register_reboot_notifier().
 		 */
-		if (sys_shutdown && spa_suspended(dp->dp_spa)) {
+		if (spa_suspended(dp->dp_spa)) {
 			ZFS_EXIT(zsb);
 			return (0);
 		}
-#endif /* HAVE_SHUTDOWN */
 
 		if (zsb->z_log != NULL)
 			zil_commit(zsb->z_log, 0);
diff --git a/module/zfs/zpl_super.c b/module/zfs/zpl_super.c
index 0417ce2ff7..73e5268f51 100644
--- a/module/zfs/zpl_super.c
+++ b/module/zfs/zpl_super.c
@@ -104,6 +104,20 @@ zpl_put_super(struct super_block *sb)
 	ASSERT3S(error, <=, 0);
 }
 
+static int
+zpl_sync_fs(struct super_block *sb, int wait)
+{
+	cred_t *cr;
+	int error;
+
+	cr = (cred_t *)get_current_cred();
+	error = -zfs_sync(sb, wait, cr);
+	put_cred(cr);
+	ASSERT3S(error, <=, 0);
+
+	return (error);
+}
+
 static int
 zpl_statfs(struct dentry *dentry, struct kstatfs *statp)
 {
@@ -179,7 +193,7 @@ const struct super_operations zpl_super_operations = {
 #endif /* HAVE_EVICT_INODE */
 	.put_super	= zpl_put_super,
 	.write_super	= NULL,
-	.sync_fs	= NULL,
+	.sync_fs	= zpl_sync_fs,
 	.freeze_fs	= NULL,
 	.unfreeze_fs	= NULL,
 	.statfs		= zpl_statfs,