Native Encryption for ZFS on Linux

This change incorporates three major pieces:

The first change is a keystore that manages wrapping
and encryption keys for encrypted datasets. These
commands mostly involve manipulating the new
DSL Crypto Key ZAP Objects that live in the MOS. Each
encrypted dataset has its own DSL Crypto Key that is
protected with a user's key. This level of indirection
allows users to change their keys without re-encrypting
their entire datasets. The change implements the new
subcommands "zfs load-key", "zfs unload-key" and
"zfs change-key" which allow the user to manage their
encryption keys and settings. In addition, several new
flags and properties have been added to allow dataset
creation and to make mounting and unmounting more
convenient.

The second piece of this patch provides the ability to
encrypt, decyrpt, and authenticate protected datasets.
Each object set maintains a Merkel tree of Message
Authentication Codes that protect the lower layers,
similarly to how checksums are maintained. This part
impacts the zio layer, which handles the actual
encryption and generation of MACs, as well as the ARC
and DMU, which need to be able to handle encrypted
buffers and protected data.

The last addition is the ability to do raw, encrypted
sends and receives. The idea here is to send raw
encrypted and compressed data and receive it exactly
as is on a backup system. This means that the dataset
on the receiving system is protected using the same
user key that is in use on the sending side. By doing
so, datasets can be efficiently backed up to an
untrusted system without fear of data being
compromised.

Reviewed by: Matthew Ahrens <[email protected]>
Reviewed-by: Brian Behlendorf <[email protected]>
Reviewed-by: Jorgen Lundman <[email protected]>
Signed-off-by: Tom Caputi <[email protected]>
Closes #494 
Closes #5769 

1 files changed, 27 insertions, 22 deletions

diff --git a/cmd/ztest/ztest.c b/cmd/ztest/ztest.c
index 277782db0..3acfaecaf 100644
--- a/cmd/ztest/ztest.c
+++ b/cmd/ztest/ztest.c
@@ -2636,7 +2636,7 @@ ztest_spa_create_destroy(ztest_ds_t *zd, uint64_t id)
 	 */
 	nvroot = make_vdev_root("/dev/bogus", NULL, NULL, 0, 0, 0, 0, 0, 1);
 	VERIFY3U(ENOENT, ==,
-	    spa_create("ztest_bad_file", nvroot, NULL, NULL));
+	    spa_create("ztest_bad_file", nvroot, NULL, NULL, NULL));
 	nvlist_free(nvroot);
 
 	/*
@@ -2644,7 +2644,7 @@ ztest_spa_create_destroy(ztest_ds_t *zd, uint64_t id)
 	 */
 	nvroot = make_vdev_root("/dev/bogus", NULL, NULL, 0, 0, 0, 0, 2, 1);
 	VERIFY3U(ENOENT, ==,
-	    spa_create("ztest_bad_mirror", nvroot, NULL, NULL));
+	    spa_create("ztest_bad_mirror", nvroot, NULL, NULL, NULL));
 	nvlist_free(nvroot);
 
 	/*
@@ -2653,7 +2653,8 @@ ztest_spa_create_destroy(ztest_ds_t *zd, uint64_t id)
 	 */
 	(void) rw_rdlock(&ztest_name_lock);
 	nvroot = make_vdev_root("/dev/bogus", NULL, NULL, 0, 0, 0, 0, 0, 1);
-	VERIFY3U(EEXIST, ==, spa_create(zo->zo_pool, nvroot, NULL, NULL));
+	VERIFY3U(EEXIST, ==,
+	    spa_create(zo->zo_pool, nvroot, NULL, NULL, NULL));
 	nvlist_free(nvroot);
 	VERIFY3U(0, ==, spa_open(zo->zo_pool, &spa, FTAG));
 	VERIFY3U(EBUSY, ==, spa_destroy(zo->zo_pool));
@@ -2755,7 +2756,7 @@ ztest_spa_upgrade(ztest_ds_t *zd, uint64_t id)
 	props = fnvlist_alloc();
 	fnvlist_add_uint64(props,
 	    zpool_prop_to_name(ZPOOL_PROP_VERSION), version);
-	VERIFY3S(spa_create(name, nvroot, props, NULL), ==, 0);
+	VERIFY3S(spa_create(name, nvroot, props, NULL, NULL), ==, 0);
 	fnvlist_free(nvroot);
 	fnvlist_free(props);
 
@@ -3530,7 +3531,7 @@ static int
 ztest_dataset_create(char *dsname)
 {
 	uint64_t zilset = ztest_random(100);
-	int err = dmu_objset_create(dsname, DMU_OST_OTHER, 0,
+	int err = dmu_objset_create(dsname, DMU_OST_OTHER, 0, NULL,
 	    ztest_objset_create_cb, NULL);
 
 	if (err || zilset < 80)
@@ -3553,7 +3554,7 @@ ztest_objset_destroy_cb(const char *name, void *arg)
 	/*
 	 * Verify that the dataset contains a directory object.
 	 */
-	VERIFY0(dmu_objset_own(name, DMU_OST_OTHER, B_TRUE, FTAG, &os));
+	VERIFY0(dmu_objset_own(name, DMU_OST_OTHER, B_TRUE, B_TRUE, FTAG, &os));
 	error = dmu_object_info(os, ZTEST_DIROBJ, &doi);
 	if (error != ENOENT) {
 		/* We could have crashed in the middle of destroying it */
@@ -3561,7 +3562,7 @@ ztest_objset_destroy_cb(const char *name, void *arg)
 		ASSERT3U(doi.doi_type, ==, DMU_OT_ZAP_OTHER);
 		ASSERT3S(doi.doi_physical_blocks_512, >=, 0);
 	}
-	dmu_objset_disown(os, FTAG);
+	dmu_objset_disown(os, B_TRUE, FTAG);
 
 	/*
 	 * Destroy the dataset.
@@ -3637,11 +3638,12 @@ ztest_dmu_objset_create_destroy(ztest_ds_t *zd, uint64_t id)
 	 * (invoked from ztest_objset_destroy_cb()) should just throw it away.
 	 */
 	if (ztest_random(2) == 0 &&
-	    dmu_objset_own(name, DMU_OST_OTHER, B_FALSE, FTAG, &os) == 0) {
+	    dmu_objset_own(name, DMU_OST_OTHER, B_FALSE,
+	    B_TRUE, FTAG, &os) == 0) {
 		ztest_zd_init(zdtmp, NULL, os);
 		zil_replay(os, zdtmp, ztest_replay_vector);
 		ztest_zd_fini(zdtmp);
-		dmu_objset_disown(os, FTAG);
+		dmu_objset_disown(os, B_TRUE, FTAG);
 	}
 
 	/*
@@ -3655,7 +3657,7 @@ ztest_dmu_objset_create_destroy(ztest_ds_t *zd, uint64_t id)
 	/*
 	 * Verify that the destroyed dataset is no longer in the namespace.
 	 */
-	VERIFY3U(ENOENT, ==, dmu_objset_own(name, DMU_OST_OTHER, B_TRUE,
+	VERIFY3U(ENOENT, ==, dmu_objset_own(name, DMU_OST_OTHER, B_TRUE, B_TRUE,
 	    FTAG, &os));
 
 	/*
@@ -3670,7 +3672,8 @@ ztest_dmu_objset_create_destroy(ztest_ds_t *zd, uint64_t id)
 		fatal(0, "dmu_objset_create(%s) = %d", name, error);
 	}
 
-	VERIFY0(dmu_objset_own(name, DMU_OST_OTHER, B_FALSE, FTAG, &os));
+	VERIFY0(dmu_objset_own(name, DMU_OST_OTHER, B_FALSE, B_TRUE,
+	    FTAG, &os));
 
 	ztest_zd_init(zdtmp, NULL, os);
 
@@ -3694,7 +3697,7 @@ ztest_dmu_objset_create_destroy(ztest_ds_t *zd, uint64_t id)
 	 * Verify that we cannot create an existing dataset.
 	 */
 	VERIFY3U(EEXIST, ==,
-	    dmu_objset_create(name, DMU_OST_OTHER, 0, NULL, NULL));
+	    dmu_objset_create(name, DMU_OST_OTHER, 0, NULL, NULL, NULL));
 
 	/*
 	 * Verify that we can hold an objset that is also owned.
@@ -3706,10 +3709,10 @@ ztest_dmu_objset_create_destroy(ztest_ds_t *zd, uint64_t id)
 	 * Verify that we cannot own an objset that is already owned.
 	 */
 	VERIFY3U(EBUSY, ==,
-	    dmu_objset_own(name, DMU_OST_OTHER, B_FALSE, FTAG, &os2));
+	    dmu_objset_own(name, DMU_OST_OTHER, B_FALSE, B_TRUE, FTAG, &os2));
 
 	zil_close(zilog);
-	dmu_objset_disown(os, FTAG);
+	dmu_objset_disown(os, B_TRUE, FTAG);
 	ztest_zd_fini(zdtmp);
 out:
 	(void) rw_unlock(&ztest_name_lock);
@@ -3863,19 +3866,20 @@ ztest_dsl_dataset_promote_busy(ztest_ds_t *zd, uint64_t id)
 		fatal(0, "dmu_objset_create(%s) = %d", clone2name, error);
 	}
 
-	error = dmu_objset_own(snap2name, DMU_OST_ANY, B_TRUE, FTAG, &os);
+	error = dmu_objset_own(snap2name, DMU_OST_ANY, B_TRUE, B_TRUE,
+	    FTAG, &os);
 	if (error)
 		fatal(0, "dmu_objset_own(%s) = %d", snap2name, error);
 	error = dsl_dataset_promote(clone2name, NULL);
 	if (error == ENOSPC) {
-		dmu_objset_disown(os, FTAG);
+		dmu_objset_disown(os, B_TRUE, FTAG);
 		ztest_record_enospc(FTAG);
 		goto out;
 	}
 	if (error != EBUSY)
 		fatal(0, "dsl_dataset_promote(%s), %d, not EBUSY", clone2name,
 		    error);
-	dmu_objset_disown(os, FTAG);
+	dmu_objset_disown(os, B_TRUE, FTAG);
 
 out:
 	ztest_dsl_dataset_cleanup(osname, id);
@@ -6253,7 +6257,7 @@ ztest_dataset_open(int d)
 	}
 	ASSERT(error == 0 || error == EEXIST);
 
-	VERIFY0(dmu_objset_own(name, DMU_OST_OTHER, B_FALSE, zd, &os));
+	VERIFY0(dmu_objset_own(name, DMU_OST_OTHER, B_FALSE, B_TRUE, zd, &os));
 	(void) rw_unlock(&ztest_name_lock);
 
 	ztest_zd_init(zd, ZTEST_GET_SHARED_DS(d), os);
@@ -6294,7 +6298,7 @@ ztest_dataset_close(int d)
 	ztest_ds_t *zd = &ztest_ds[d];
 
 	zil_close(zd->zd_zilog);
-	dmu_objset_disown(zd->zd_os, zd);
+	dmu_objset_disown(zd->zd_os, B_TRUE, zd);
 
 	ztest_zd_fini(zd);
 }
@@ -6347,12 +6351,12 @@ ztest_run(ztest_shared_t *zs)
 
 	dmu_objset_stats_t dds;
 	VERIFY0(dmu_objset_own(ztest_opts.zo_pool,
-	    DMU_OST_ANY, B_TRUE, FTAG, &os));
+	    DMU_OST_ANY, B_TRUE, B_TRUE, FTAG, &os));
 	dsl_pool_config_enter(dmu_objset_pool(os), FTAG);
 	dmu_objset_fast_stat(os, &dds);
 	dsl_pool_config_exit(dmu_objset_pool(os), FTAG);
 	zs->zs_guid = dds.dds_guid;
-	dmu_objset_disown(os, FTAG);
+	dmu_objset_disown(os, B_TRUE, FTAG);
 
 	spa->spa_dedup_ditto = 2 * ZIO_DEDUPDITTO_MIN;
 
@@ -6705,7 +6709,8 @@ ztest_init(ztest_shared_t *zs)
 		VERIFY3U(0, ==, nvlist_add_uint64(props, buf, 0));
 		free(buf);
 	}
-	VERIFY3U(0, ==, spa_create(ztest_opts.zo_pool, nvroot, props, NULL));
+	VERIFY3U(0, ==,
+	    spa_create(ztest_opts.zo_pool, nvroot, props, NULL, NULL));
 	nvlist_free(nvroot);
 	nvlist_free(props);