Illumos #1748: desire support for reguid in zfs

Reviewed by: George Wilson <[email protected]>
Reviewed by: Igor Kozhukhov <[email protected]>
Reviewed by: Alexander Eremin <[email protected]>
Reviewed by: Alexander Stetsenko <[email protected]>
Approved by: Richard Lowe <[email protected]>

References:
  https://www.illumos.org/issues/1748

This commit modifies the user to kernel space ioctl ABI.  Extra
care should be taken when updating to ensure both the kernel
modules and utilities are updated.  If only the user space
component is updated both the 'zpool events' command and the
'zpool reguid' command will not work until the kernel modules
are updated.

Ported by:     Martin Matuska <[email protected]>
Signed-off-by: Brian Behlendorf <[email protected]>
Closes #665

5 files changed, 83 insertions, 10 deletions

diff --git a/module/zfs/arc.c b/module/zfs/arc.c
index 7aa23ca15..aad37a39b 100644
--- a/module/zfs/arc.c
+++ b/module/zfs/arc.c
@@ -20,6 +20,8 @@
  */
 /*
  * Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
+ * Copyright 2011 Nexenta Systems, Inc.  All rights reserved.
+ * Copyright (c) 2011 by Delphix. All rights reserved.
  */
 
 /*
@@ -1284,7 +1286,7 @@ arc_buf_alloc(spa_t *spa, int size, void *tag, arc_buf_contents_t type)
 	ASSERT(BUF_EMPTY(hdr));
 	hdr->b_size = size;
 	hdr->b_type = type;
-	hdr->b_spa = spa_guid(spa);
+	hdr->b_spa = spa_load_guid(spa);
 	hdr->b_state = arc_anon;
 	hdr->b_arc_access = 0;
 	buf = kmem_cache_alloc(buf_cache, KM_PUSHPAGE);
@@ -2056,7 +2058,7 @@ arc_flush(spa_t *spa)
 	uint64_t guid = 0;
 
 	if (spa)
-		guid = spa_guid(spa);
+		guid = spa_load_guid(spa);
 
 	while (list_head(&arc_mru->arcs_list[ARC_BUFC_DATA])) {
 		(void) arc_evict(arc_mru, guid, -1, FALSE, ARC_BUFC_DATA);
@@ -2887,7 +2889,7 @@ arc_read_nolock(zio_t *pio, spa_t *spa, const blkptr_t *bp,
 	arc_buf_t *buf = NULL;
 	kmutex_t *hash_lock;
 	zio_t *rzio;
-	uint64_t guid = spa_guid(spa);
+	uint64_t guid = spa_load_guid(spa);
 
 top:
 	hdr = buf_hash_find(guid, BP_IDENTITY(bp), BP_PHYSICAL_BIRTH(bp),
@@ -4522,7 +4524,7 @@ l2arc_write_buffers(spa_t *spa, l2arc_dev_t *dev, uint64_t target_sz)
 	boolean_t have_lock, full;
 	l2arc_write_callback_t *cb;
 	zio_t *pio, *wzio;
-	uint64_t guid = spa_guid(spa);
+	uint64_t guid = spa_load_guid(spa);
 	int try;
 
 	ASSERT(dev->l2ad_vdev != NULL);
diff --git a/module/zfs/spa.c b/module/zfs/spa.c
index a43b88338..31ffd453f 100644
--- a/module/zfs/spa.c
+++ b/module/zfs/spa.c
@@ -21,9 +21,8 @@
 
 /*
  * Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
- */
-/*
  * Copyright 2011 Nexenta Systems, Inc. All rights reserved.
+ * Copyright (c) 2011 by Delphix. All rights reserved.
  */
 
 /*
@@ -572,6 +571,43 @@ spa_prop_clear_bootfs(spa_t *spa, uint64_t dsobj, dmu_tx_t *tx)
 }
 
 /*
+ * Change the GUID for the pool.  This is done so that we can later
+ * re-import a pool built from a clone of our own vdevs.  We will modify
+ * the root vdev's guid, our own pool guid, and then mark all of our
+ * vdevs dirty.  Note that we must make sure that all our vdevs are
+ * online when we do this, or else any vdevs that weren't present
+ * would be orphaned from our pool.  We are also going to issue a
+ * sysevent to update any watchers.
+ */
+int
+spa_change_guid(spa_t *spa)
+{
+	uint64_t	oldguid, newguid;
+	uint64_t	txg;
+
+	if (!(spa_mode_global & FWRITE))
+		return (EROFS);
+
+	txg = spa_vdev_enter(spa);
+
+	if (spa->spa_root_vdev->vdev_state != VDEV_STATE_HEALTHY)
+		return (spa_vdev_exit(spa, NULL, txg, ENXIO));
+
+	oldguid = spa_guid(spa);
+	newguid = spa_generate_guid(NULL);
+	ASSERT3U(oldguid, !=, newguid);
+
+	spa->spa_root_vdev->vdev_guid = newguid;
+	spa->spa_root_vdev->vdev_guid_sum += (newguid - oldguid);
+
+	vdev_config_dirty(spa->spa_root_vdev);
+
+	spa_event_notify(spa, NULL, FM_EREPORT_ZFS_POOL_REGUID);
+
+	return (spa_vdev_exit(spa, NULL, txg, 0));
+}
+
+/*
  * ==========================================================================
  * SPA state manipulation (open/create/destroy/import/export)
  * ==========================================================================
@@ -1773,7 +1809,7 @@ spa_load(spa_t *spa, spa_load_state_t state, spa_import_type_t type,
 	    spa_guid_exists(pool_guid, 0)) {
 		error = EEXIST;
 	} else {
-		spa->spa_load_guid = pool_guid;
+		spa->spa_config_guid = pool_guid;
 
 		if (nvlist_lookup_nvlist(config, ZPOOL_CONFIG_SPLIT,
 		    &nvl) == 0) {
diff --git a/module/zfs/spa_misc.c b/module/zfs/spa_misc.c
index 1016ac7be..c82dca6c5 100644
--- a/module/zfs/spa_misc.c
+++ b/module/zfs/spa_misc.c
@@ -21,6 +21,7 @@
 /*
  * Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
  * Copyright (c) 2011 by Delphix. All rights reserved.
+ * Copyright 2011 Nexenta Systems, Inc.  All rights reserved.
  */
 
 #include <sys/zfs_context.h>
@@ -1290,13 +1291,24 @@ spa_guid(spa_t *spa)
 	/*
 	 * If we fail to parse the config during spa_load(), we can go through
 	 * the error path (which posts an ereport) and end up here with no root
-	 * vdev.  We stash the original pool guid in 'spa_load_guid' to handle
+	 * vdev.  We stash the original pool guid in 'spa_config_guid' to handle
 	 * this case.
 	 */
 	if (spa->spa_root_vdev != NULL)
 		return (spa->spa_root_vdev->vdev_guid);
 	else
-		return (spa->spa_load_guid);
+		return (spa->spa_config_guid);
+}
+
+uint64_t
+spa_load_guid(spa_t *spa)
+{
+	/*
+	 * This is a GUID that exists solely as a reference for the
+	 * purposes of the arc.  It is generated at load time, and
+	 * is never written to persistent storage.
+	 */
+	return (spa->spa_load_guid);
 }
 
 uint64_t
diff --git a/module/zfs/vdev.c b/module/zfs/vdev.c
index 1630d2fae..18d2d5ca6 100644
--- a/module/zfs/vdev.c
+++ b/module/zfs/vdev.c
@@ -21,6 +21,8 @@
 
 /*
  * Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
+ * Copyright 2011 Nexenta Systems, Inc.  All rights reserved.
+ * Copyright (c) 2011 by Delphix. All rights reserved.
  */
 
 #include <sys/zfs_context.h>
@@ -291,6 +293,7 @@ vdev_alloc_common(spa_t *spa, uint_t id, uint64_t guid, vdev_ops_t *ops)
 	if (spa->spa_root_vdev == NULL) {
 		ASSERT(ops == &vdev_root_ops);
 		spa->spa_root_vdev = vd;
+		spa->spa_load_guid = spa_generate_guid(NULL);
 	}
 
 	if (guid == 0 && ops != &vdev_hole_ops) {
diff --git a/module/zfs/zfs_ioctl.c b/module/zfs/zfs_ioctl.c
index a232ed0d5..65b0a1975 100644
--- a/module/zfs/zfs_ioctl.c
+++ b/module/zfs/zfs_ioctl.c
@@ -24,6 +24,10 @@
  * Portions Copyright 2012 Pawel Jakub Dawidek <[email protected]>
  * Copyright (c) 2012, Joyent, Inc. All rights reserved.
  */
+/*
+ * Copyright 2011 Nexenta Systems, Inc.  All rights reserved.
+ * Copyright (c) 2011 by Delphix. All rights reserved.
+ */
 
 #include <sys/types.h>
 #include <sys/param.h>
@@ -1428,6 +1432,20 @@ zfs_ioc_pool_get_history(zfs_cmd_t *zc)
 }
 
 static int
+zfs_ioc_pool_reguid(zfs_cmd_t *zc)
+{
+	spa_t *spa;
+	int error;
+
+	error = spa_open(zc->zc_name, &spa, FTAG);
+	if (error == 0) {
+		error = spa_change_guid(spa);
+		spa_close(spa, FTAG);
+	}
+	return (error);
+}
+
+static int
 zfs_ioc_dsobj_to_dsname(zfs_cmd_t *zc)
 {
 	int error;
@@ -4698,10 +4716,12 @@ static zfs_ioc_vec_t zfs_ioc_vec[] = {
 	    B_FALSE, POOL_CHECK_SUSPENDED | POOL_CHECK_READONLY },
 	{ zfs_ioc_obj_to_stats, zfs_secpolicy_diff, DATASET_NAME, B_FALSE,
 	    POOL_CHECK_SUSPENDED },
+        { zfs_ioc_pool_reguid, zfs_secpolicy_config, POOL_NAME, B_TRUE,
+            POOL_CHECK_SUSPENDED | POOL_CHECK_READONLY },
 	{ zfs_ioc_events_next, zfs_secpolicy_config, NO_NAME, B_FALSE,
 	    POOL_CHECK_NONE },
 	{ zfs_ioc_events_clear, zfs_secpolicy_config, NO_NAME, B_FALSE,
-	    POOL_CHECK_NONE },
+	    POOL_CHECK_NONE }
 };
 
 int