Illumos openzfs#2619 and openzfs#2747

2619 asynchronous destruction of ZFS file systems
2747 SPA versioning with zfs feature flags
Reviewed by: Matt Ahrens <mahrens@delphix.com>
Reviewed by: George Wilson <gwilson@delphix.com>
Reviewed by: Richard Lowe <richlowe@richlowe.net>
Reviewed by: Dan Kruchinin <dan.kruchinin@gmail.com>
Approved by: Eric Schrock <Eric.Schrock@delphix.com>

References:
  illumos/illumos-gate@53089ab
  illumos/illumos-gate@ad135b5
  illumos changeset: 13700:2889e2596bd6
  https://www.illumos.org/issues/2619
  https://www.illumos.org/issues/2747

NOTE: The grub specific changes were not ported.  This change
must be made to the Linux grub packages.

Ported-by: Brian Behlendorf <behlendorf1@llnl.gov>

cmd/Makefile.am

@@ -1,2 +1,2 @@
-SUBDIRS  = zfs zpool zdb zinject zstreamdump ztest zpios mount_zfs
+SUBDIRS  = zfs zpool zdb zhack zinject zstreamdump ztest zpios mount_zfs
 SUBDIRS += zpool_layout zvol_id zpool_id vdev_id

cmd/zdb/zdb.c

@@ -18,8 +18,10 @@
  *
  * CDDL HEADER END
  */
+
 /*
  * Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
+ * Copyright (c) 2012 by Delphix. All rights reserved.
  */
 
 #include <stdio.h>
@@ -54,6 +56,7 @@
 #include <sys/zfs_fuid.h>
 #include <sys/arc.h>
 #include <sys/ddt.h>
+#include <sys/zfeature.h>
 #undef ZFS_MAXNAMELEN
 #include <libzfs.h>
 
@@ -62,7 +65,8 @@
 #define	ZDB_CHECKSUM_NAME(idx) ((idx) < ZIO_CHECKSUM_FUNCTIONS ? \
     zio_checksum_table[(idx)].ci_name : "UNKNOWN")
 #define	ZDB_OT_NAME(idx) ((idx) < DMU_OT_NUMTYPES ? \
-    dmu_ot[(idx)].ot_name : "UNKNOWN")
+    dmu_ot[(idx)].ot_name : DMU_OT_IS_VALID(idx) ? \
+    dmu_ot_byteswap[DMU_OT_BYTESWAP(idx)].ob_name : "UNKNOWN")
 #define	ZDB_OT_TYPE(idx) ((idx) < DMU_OT_NUMTYPES ? (idx) : DMU_OT_NUMTYPES)
 
 #ifndef lint
@@ -1099,7 +1103,7 @@ dump_dsl_dataset(objset_t *os, uint64_t object, void *data, size_t size)
 
 	ASSERT(size == sizeof (*ds));
 	crtime = ds->ds_creation_time;
-	zdb_nicenum(ds->ds_used_bytes, used);
+	zdb_nicenum(ds->ds_referenced_bytes, used);
 	zdb_nicenum(ds->ds_compressed_bytes, compressed);
 	zdb_nicenum(ds->ds_uncompressed_bytes, uncompressed);
 	zdb_nicenum(ds->ds_unique_bytes, unique);
@@ -1141,6 +1145,44 @@ dump_dsl_dataset(objset_t *os, uint64_t object, void *data, size_t size)
 	(void) printf("\t\tbp = %s\n", blkbuf);
 }
 
+/* ARGSUSED */
+static int
+dump_bptree_cb(void *arg, const blkptr_t *bp, dmu_tx_t *tx)
+{
+	char blkbuf[BP_SPRINTF_LEN];
+
+	if (bp->blk_birth != 0) {
+		sprintf_blkptr(blkbuf, bp);
+		(void) printf("\t%s\n", blkbuf);
+	}
+	return (0);
+}
+
+static void
+dump_bptree(objset_t *os, uint64_t obj, char *name)
+{
+	char bytes[32];
+	bptree_phys_t *bt;
+	dmu_buf_t *db;
+
+	if (dump_opt['d'] < 3)
+		return;
+
+	VERIFY3U(0, ==, dmu_bonus_hold(os, obj, FTAG, &db));
+	bt = db->db_data;
+	zdb_nicenum(bt->bt_bytes, bytes);
+	(void) printf("\n    %s: %llu datasets, %s\n",
+	    name, (unsigned long long)(bt->bt_end - bt->bt_begin), bytes);
+	dmu_buf_rele(db, FTAG);
+
+	if (dump_opt['d'] < 5)
+		return;
+
+	(void) printf("\n");
+
+	(void) bptree_iterate(os, obj, B_FALSE, dump_bptree_cb, NULL, NULL);
+}
+
 /* ARGSUSED */
 static int
 dump_bpobj_cb(void *arg, const blkptr_t *bp, dmu_tx_t *tx)
@@ -1888,11 +1930,13 @@ typedef struct zdb_blkstats {
  */
 #define	ZDB_OT_DEFERRED	(DMU_OT_NUMTYPES + 0)
 #define	ZDB_OT_DITTO	(DMU_OT_NUMTYPES + 1)
-#define	ZDB_OT_TOTAL	(DMU_OT_NUMTYPES + 2)
+#define	ZDB_OT_OTHER	(DMU_OT_NUMTYPES + 2)
+#define	ZDB_OT_TOTAL	(DMU_OT_NUMTYPES + 3)
 
 static char *zdb_ot_extname[] = {
 	"deferred free",
 	"dedup ditto",
+	"other",
 	"Total",
 };
 
@@ -1974,9 +2018,10 @@ zdb_blkptr_cb(spa_t *spa, zilog_t *zilog, const blkptr_t *bp, arc_buf_t *pbuf,
 
 	type = BP_GET_TYPE(bp);
 
-	zdb_count_block(zcb, zilog, bp, type);
+	zdb_count_block(zcb, zilog, bp,
+	    (type & DMU_OT_NEWTYPE) ? ZDB_OT_OTHER : type);
 
-	is_metadata = (BP_GET_LEVEL(bp) != 0 || dmu_ot[type].ot_metadata);
+	is_metadata = (BP_GET_LEVEL(bp) != 0 || DMU_OT_IS_METADATA(type));
 
 	if (dump_opt['c'] > 1 || (dump_opt['c'] && is_metadata)) {
 		int ioerr;
@@ -2207,6 +2252,12 @@ dump_block_stats(spa_t *spa)
 	    count_block_cb, &zcb, NULL);
 	(void) bpobj_iterate_nofree(&spa->spa_dsl_pool->dp_free_bpobj,
 	    count_block_cb, &zcb, NULL);
+	if (spa_feature_is_active(spa,
+	    &spa_feature_table[SPA_FEATURE_ASYNC_DESTROY])) {
+		VERIFY3U(0, ==, bptree_iterate(spa->spa_meta_objset,
+		    spa->spa_dsl_pool->dp_bptree_obj, B_FALSE, count_block_cb,
+		    &zcb, NULL));
+	}
 
 	if (dump_opt['c'] > 1)
 		flags |= TRAVERSE_PREFETCH_DATA;
@@ -2383,7 +2434,7 @@ zdb_ddt_add_cb(spa_t *spa, zilog_t *zilog, const blkptr_t *bp,
 	}
 
 	if (BP_IS_HOLE(bp) || BP_GET_CHECKSUM(bp) == ZIO_CHECKSUM_OFF ||
-	    BP_GET_LEVEL(bp) > 0 || dmu_ot[BP_GET_TYPE(bp)].ot_metadata)
+	    BP_GET_LEVEL(bp) > 0 || DMU_OT_IS_METADATA(BP_GET_TYPE(bp)))
 		return (0);
 
 	ddt_key_fill(&zdde_search.zdde_key, bp);
@@ -2491,7 +2542,14 @@ dump_zpool(spa_t *spa)
 			dump_bpobj(&spa->spa_deferred_bpobj, "Deferred frees");
 			if (spa_version(spa) >= SPA_VERSION_DEADLISTS) {
 				dump_bpobj(&spa->spa_dsl_pool->dp_free_bpobj,
-				    "Pool frees");
+				    "Pool snapshot frees");
+			}
+
+			if (spa_feature_is_active(spa,
+			    &spa_feature_table[SPA_FEATURE_ASYNC_DESTROY])) {
+				dump_bptree(spa->spa_meta_objset,
+				    spa->spa_dsl_pool->dp_bptree_obj,
+				    "Pool dataset frees");
 			}
 			dump_dtl(spa->spa_root_vdev, 0);
 		}

cmd/zhack/.gitignore

@@ -0,0 +1 @@
+/zhack

cmd/zhack/Makefile.am

@@ -0,0 +1,18 @@
+include $(top_srcdir)/config/Rules.am
+
+DEFAULT_INCLUDES += \
+	-I$(top_srcdir)/include \
+	-I$(top_srcdir)/lib/libspl/include
+
+sbin_PROGRAMS = zhack
+
+zhack_SOURCES = \
+	$(top_srcdir)/cmd/zhack/zhack.c
+
+zhack_LDADD = \
+	$(top_builddir)/lib/libnvpair/libnvpair.la \
+	$(top_builddir)/lib/libuutil/libuutil.la \
+	$(top_builddir)/lib/libzpool/libzpool.la \
+	$(top_builddir)/lib/libzfs/libzfs.la
+
+zhack_LDFLAGS = -pthread -lm $(ZLIB) -lrt -ldl $(LIBUUID) $(LIBBLKID)