Add support for asynchronous zvol minor operations

zfsonlinux issue #2217 - zvol minor operations: check snapdev
property before traversing snapshots of a dataset

zfsonlinux issue #3681 - lock order inversion between zvol_open()
and dsl_pool_sync()...zvol_rename_minors()

Create a per-pool zvol taskq for asynchronous zvol tasks.
There are a few key design decisions to be aware of.

* Each taskq must be single threaded to ensure tasks are always
  processed in the order in which they were dispatched.

* There is a taskq per-pool in order to keep the pools independent.
  This way if one pool is suspended it will not impact another.

* The preferred location to dispatch a zvol minor task is a sync
  task.  In this context there is easy access to the spa_t and
  minimal error handling is required because the sync task must
  succeed.

Support for asynchronous zvol minor operations address issue #3681.

Signed-off-by: Boris Protopopov <[email protected]>
Signed-off-by: Brian Behlendorf <[email protected]>
Closes #2217
Closes #3678
Closes #3681

1 files changed, 30 insertions, 7 deletions

diff --git a/module/zfs/spa.c b/module/zfs/spa.c
index 34a317fbe..01048bfe9 100644
--- a/module/zfs/spa.c
+++ b/module/zfs/spa.c
@@ -24,6 +24,7 @@
  * Copyright (c) 2013 by Delphix. All rights reserved.
  * Copyright (c) 2013, 2014, Nexenta Systems, Inc.  All rights reserved.
  * Copyright (c) 2014 Spectra Logic Corporation, All rights reserved.
+ * Copyright (c) 2016 Actifio, Inc. All rights reserved.
  */
 
 /*
@@ -1136,6 +1137,24 @@ spa_activate(spa_t *spa, int mode)
 	avl_create(&spa->spa_errlist_last,
 	    spa_error_entry_compare, sizeof (spa_error_entry_t),
 	    offsetof(spa_error_entry_t, se_avl));
+
+	/*
+	 * This taskq is used to perform zvol-minor-related tasks
+	 * asynchronously. This has several advantages, including easy
+	 * resolution of various deadlocks (zfsonlinux bug #3681).
+	 *
+	 * The taskq must be single threaded to ensure tasks are always
+	 * processed in the order in which they were dispatched.
+	 *
+	 * A taskq per pool allows one to keep the pools independent.
+	 * This way if one pool is suspended, it will not impact another.
+	 *
+	 * The preferred location to dispatch a zvol minor task is a sync
+	 * task. In this context, there is easy access to the spa_t and minimal
+	 * error handling is required because the sync task must succeed.
+	 */
+	spa->spa_zvol_taskq = taskq_create("z_zvol", 1, defclsyspri,
+	    1, INT_MAX, 0);
 }
 
 /*
@@ -1154,6 +1173,11 @@ spa_deactivate(spa_t *spa)
 
 	spa_evicting_os_wait(spa);
 
+	if (spa->spa_zvol_taskq) {
+		taskq_destroy(spa->spa_zvol_taskq);
+		spa->spa_zvol_taskq = NULL;
+	}
+
 	txg_list_destroy(&spa->spa_vdev_txg_list);
 
 	list_destroy(&spa->spa_config_dirty_list);
@@ -3088,10 +3112,8 @@ spa_open_common(const char *pool, spa_t **spapp, void *tag, nvlist_t *nvpolicy,
 		mutex_exit(&spa_namespace_lock);
 	}
 
-#ifdef _KERNEL
 	if (firstopen)
-		zvol_create_minors(spa->spa_name);
-#endif
+		zvol_create_minors(spa, spa_name(spa), B_TRUE);
 
 	*spapp = spa;
 
@@ -4211,10 +4233,7 @@ spa_import(char *pool, nvlist_t *config, nvlist_t *props, uint64_t flags)
 
 	mutex_exit(&spa_namespace_lock);
 	spa_history_log_version(spa, "import");
-
-#ifdef _KERNEL
-	zvol_create_minors(pool);
-#endif
+	zvol_create_minors(spa, pool, B_TRUE);
 
 	return (0);
 }
@@ -4349,6 +4368,10 @@ spa_export_common(char *pool, int new_state, nvlist_t **oldconfig,
 	spa_open_ref(spa, FTAG);
 	mutex_exit(&spa_namespace_lock);
 	spa_async_suspend(spa);
+	if (spa->spa_zvol_taskq) {
+		zvol_remove_minors(spa, spa_name(spa), B_TRUE);
+		taskq_wait(spa->spa_zvol_taskq);
+	}
 	mutex_enter(&spa_namespace_lock);
 	spa_close(spa, FTAG);