MFC r268075: MFV r267565:

4757 ZFS embedded-data block pointers ("zero block compression")
4913 zfs release should not be subject to space checks

14 files changed, 287 insertions, 56 deletions

diff --git a/cddl/contrib/opensolaris/cmd/zdb/zdb.c b/cddl/contrib/opensolaris/cmd/zdb/zdb.c
index ac4ea04..6346635 100644
--- a/cddl/contrib/opensolaris/cmd/zdb/zdb.c
+++ b/cddl/contrib/opensolaris/cmd/zdb/zdb.c
@@ -1059,8 +1059,17 @@ snprintf_blkptr_compact(char *blkbuf, size_t buflen, const blkptr_t *bp)
 		return;
 	}
 
-	blkbuf[0] = '\0';
+	if (BP_IS_EMBEDDED(bp)) {
+		(void) sprintf(blkbuf,
+		    "EMBEDDED et=%u %llxL/%llxP B=%llu",
+		    (int)BPE_GET_ETYPE(bp),
+		    (u_longlong_t)BPE_GET_LSIZE(bp),
+		    (u_longlong_t)BPE_GET_PSIZE(bp),
+		    (u_longlong_t)bp->blk_birth);
+		return;
+	}
 
+	blkbuf[0] = '\0';
 	for (int i = 0; i < ndvas; i++)
 		(void) snprintf(blkbuf + strlen(blkbuf),
 		    buflen - strlen(blkbuf), "%llu:%llx:%llx ",
@@ -1078,7 +1087,7 @@ snprintf_blkptr_compact(char *blkbuf, size_t buflen, const blkptr_t *bp)
 		    "%llxL/%llxP F=%llu B=%llu/%llu",
 		    (u_longlong_t)BP_GET_LSIZE(bp),
 		    (u_longlong_t)BP_GET_PSIZE(bp),
-		    (u_longlong_t)bp->blk_fill,
+		    (u_longlong_t)BP_GET_FILL(bp),
 		    (u_longlong_t)bp->blk_birth,
 		    (u_longlong_t)BP_PHYSICAL_BIRTH(bp));
 	}
@@ -1091,8 +1100,10 @@ print_indirect(blkptr_t *bp, const zbookmark_t *zb,
 	char blkbuf[BP_SPRINTF_LEN];
 	int l;
 
-	ASSERT3U(BP_GET_TYPE(bp), ==, dnp->dn_type);
-	ASSERT3U(BP_GET_LEVEL(bp), ==, zb->zb_level);
+	if (!BP_IS_EMBEDDED(bp)) {
+		ASSERT3U(BP_GET_TYPE(bp), ==, dnp->dn_type);
+		ASSERT3U(BP_GET_LEVEL(bp), ==, zb->zb_level);
+	}
 
 	(void) printf("%16llx ", (u_longlong_t)blkid2offset(dnp, bp, zb));
 
@@ -1146,10 +1157,10 @@ visit_indirect(spa_t *spa, const dnode_phys_t *dnp,
 			err = visit_indirect(spa, dnp, cbp, &czb);
 			if (err)
 				break;
-			fill += cbp->blk_fill;
+			fill += BP_GET_FILL(cbp);
 		}
 		if (!err)
-			ASSERT3U(fill, ==, bp->blk_fill);
+			ASSERT3U(fill, ==, BP_GET_FILL(bp));
 		(void) arc_buf_remove_ref(buf, &buf);
 	}
 
@@ -1816,14 +1827,14 @@ dump_dir(objset_t *os)
 
 	if (dds.dds_type == DMU_OST_META) {
 		dds.dds_creation_txg = TXG_INITIAL;
-		usedobjs = os->os_rootbp->blk_fill;
+		usedobjs = BP_GET_FILL(os->os_rootbp);
 		refdbytes = os->os_spa->spa_dsl_pool->
 		    dp_mos_dir->dd_phys->dd_used_bytes;
 	} else {
 		dmu_objset_space(os, &refdbytes, &scratch, &usedobjs, &scratch);
 	}
 
-	ASSERT3U(usedobjs, ==, os->os_rootbp->blk_fill);
+	ASSERT3U(usedobjs, ==, BP_GET_FILL(os->os_rootbp));
 
 	zdb_nicenum(refdbytes, numbuf);
 
@@ -2134,6 +2145,9 @@ typedef struct zdb_cb {
 	zdb_blkstats_t	zcb_type[ZB_TOTAL + 1][ZDB_OT_TOTAL + 1];
 	uint64_t	zcb_dedup_asize;
 	uint64_t	zcb_dedup_blocks;
+	uint64_t	zcb_embedded_blocks[NUM_BP_EMBEDDED_TYPES];
+	uint64_t	zcb_embedded_histogram[NUM_BP_EMBEDDED_TYPES]
+	    [BPE_PAYLOAD_SIZE];
 	uint64_t	zcb_start;
 	uint64_t	zcb_lastprint;
 	uint64_t	zcb_totalasize;
@@ -2188,6 +2202,13 @@ zdb_count_block(zdb_cb_t *zcb, zilog_t *zilog, const blkptr_t *bp,
 
 	}
 
+	if (BP_IS_EMBEDDED(bp)) {
+		zcb->zcb_embedded_blocks[BPE_GET_ETYPE(bp)]++;
+		zcb->zcb_embedded_histogram[BPE_GET_ETYPE(bp)]
+		    [BPE_GET_PSIZE(bp)]++;
+		return;
+	}
+
 	if (dump_opt['L'])
 		return;
 
@@ -2287,7 +2308,8 @@ zdb_blkptr_cb(spa_t *spa, zilog_t *zilog, const blkptr_t *bp,
 
 	is_metadata = (BP_GET_LEVEL(bp) != 0 || DMU_OT_IS_METADATA(type));
 
-	if (dump_opt['c'] > 1 || (dump_opt['c'] && is_metadata)) {
+	if (!BP_IS_EMBEDDED(bp) &&
+	    (dump_opt['c'] > 1 || (dump_opt['c'] && is_metadata))) {
 		size_t size = BP_GET_PSIZE(bp);
 		void *data = zio_data_buf_alloc(size);
 		int flags = ZIO_FLAG_CANFAIL | ZIO_FLAG_SCRUB | ZIO_FLAG_RAW;
@@ -2479,7 +2501,7 @@ dump_block_stats(spa_t *spa)
 	zdb_blkstats_t *zb, *tzb;
 	uint64_t norm_alloc, norm_space, total_alloc, total_found;
 	int flags = TRAVERSE_PRE | TRAVERSE_PREFETCH_METADATA | TRAVERSE_HARD;
-	int leaks = 0;
+	boolean_t leaks = B_FALSE;
 
 	(void) printf("\nTraversing all blocks %s%s%s%s%s...\n\n",
 	    (dump_opt['c'] || !dump_opt['L']) ? "to verify " : "",
@@ -2567,7 +2589,7 @@ dump_block_stats(spa_t *spa)
 		    (u_longlong_t)total_alloc,
 		    (dump_opt['L']) ? "unreachable" : "leaked",
 		    (longlong_t)(total_alloc - total_found));
-		leaks = 1;
+		leaks = B_TRUE;
 	}
 
 	if (tzb->zb_count == 0)
@@ -2599,6 +2621,23 @@ dump_block_stats(spa_t *spa)
 	(void) printf("\tSPA allocated: %10llu     used: %5.2f%%\n",
 	    (u_longlong_t)norm_alloc, 100.0 * norm_alloc / norm_space);
 
+	for (bp_embedded_type_t i = 0; i < NUM_BP_EMBEDDED_TYPES; i++) {
+		if (zcb.zcb_embedded_blocks[i] == 0)
+			continue;
+		(void) printf("\n");
+		(void) printf("\tadditional, non-pointer bps of type %u: "
+		    "%10llu\n",
+		    i, (u_longlong_t)zcb.zcb_embedded_blocks[i]);
+
+		if (dump_opt['b'] >= 3) {
+			(void) printf("\t number of (compressed) bytes:  "
+			    "number of bps\n");
+			dump_histogram(zcb.zcb_embedded_histogram[i],
+			    sizeof (zcb.zcb_embedded_histogram[i]) /
+			    sizeof (zcb.zcb_embedded_histogram[i][0]), 0);
+		}
+	}
+
 	if (tzb->zb_ditto_samevdev != 0) {
 		(void) printf("\tDittoed blocks on same vdev: %llu\n",
 		    (longlong_t)tzb->zb_ditto_samevdev);
@@ -2711,14 +2750,14 @@ zdb_ddt_add_cb(spa_t *spa, zilog_t *zilog, const blkptr_t *bp,
 	avl_index_t where;
 	zdb_ddt_entry_t *zdde, zdde_search;
 
-	if (BP_IS_HOLE(bp))
+	if (BP_IS_HOLE(bp) || BP_IS_EMBEDDED(bp))
 		return (0);
 
 	if (dump_opt['S'] > 1 && zb->zb_level == ZB_ROOT_LEVEL) {
 		(void) printf("traversing objset %llu, %llu objects, "
 		    "%lu blocks so far\n",
 		    (u_longlong_t)zb->zb_objset,
-		    (u_longlong_t)bp->blk_fill,
+		    (u_longlong_t)BP_GET_FILL(bp),
 		    avl_numnodes(t));
 	}
 
diff --git a/cddl/contrib/opensolaris/cmd/zfs/zfs.8 b/cddl/contrib/opensolaris/cmd/zfs/zfs.8
index d0f21d3..4ae5fbe 100644
--- a/cddl/contrib/opensolaris/cmd/zfs/zfs.8
+++ b/cddl/contrib/opensolaris/cmd/zfs/zfs.8
@@ -30,7 +30,7 @@
 .\"
 .\" $FreeBSD$
 .\"
-.Dd May 27, 2014
+.Dd June 30, 2014
 .Dt ZFS 8
 .Os
 .Sh NAME
@@ -179,11 +179,12 @@
 .Ar bookmark
 .Nm
 .Cm send
-.Op Fl DnPpRv
+.Op Fl DnPpRve
 .Op Fl i Ar snapshot | Fl I Ar snapshot
 .Ar snapshot
 .Nm
 .Cm send
+.Op Fl e
 .Op Fl i Ar snapshot Ns | Ns bookmark
 .Ar filesystem Ns | Ns Ar volume Ns | Ns Ar snapshot
 .Nm
@@ -2476,7 +2477,7 @@ feature.
 .It Xo
 .Nm
 .Cm send
-.Op Fl DnPpRv
+.Op Fl DnPpRve
 .Op Fl i Ar snapshot | Fl I Ar snapshot
 .Ar snapshot
 .Xc
@@ -2548,6 +2549,29 @@ be used regardless of the dataset's
 property, but performance will be much better if the filesystem uses a
 dedup-capable checksum (eg.
 .Sy sha256 ) .
+.It Fl e
+Generate a more compact stream by using WRITE_EMBEDDED records for blocks
+which are stored more compactly on disk by the
+.Sy embedded_data
+pool
+feature.
+This flag has no effect if the
+.Sy embedded_data
+feature is
+disabled.
+The receiving system must have the
+.Sy embedded_data
+feature
+enabled.
+If the
+.Sy lz4_compress
+feature is active on the sending system,
+then the receiving system must have that feature enabled as well.
+See
+.Xr zpool-features 7
+for details on ZFS feature flags and the
+.Sy embedded_data
+feature.
 .It Fl p
 Include the dataset's properties in the stream. This flag is implicit when
 .Fl R
@@ -2572,6 +2596,7 @@ on future versions of
 .It Xo
 .Nm
 .Cm send
+.Op Fl e
 .Op Fl i Ar snapshot Ns | Ns Ar bookmark
 .Ar filesystem Ns | Ns Ar volume Ns | Ns Ar snapshot
 .Xc
@@ -2597,6 +2622,29 @@ specified as the last component of the name
 If the incremental target is a clone, the incremental source can
 be the origin snapshot, or an earlier snapshot in the origin's filesystem,
 or the origin's origin, etc.
+.It Fl e
+Generate a more compact stream by using WRITE_EMBEDDED records for blocks
+which are stored more compactly on disk by the
+.Sy embedded_data
+pool
+feature.
+This flag has no effect if the
+.Sy embedded_data
+feature is
+disabled.
+The receiving system must have the
+.Sy embedded_data
+feature
+enabled.
+If the
+.Sy lz4_compress
+feature is active on the sending system,
+then the receiving system must have that feature enabled as well.
+See
+.Xr zpool-features 7
+for details on ZFS feature flags and the
+.Sy embedded_data
+feature.
 .El
 .It Xo
 .Nm
diff --git a/cddl/contrib/opensolaris/cmd/zfs/zfs_main.c b/cddl/contrib/opensolaris/cmd/zfs/zfs_main.c
index 4464b6a..1049c39 100644
--- a/cddl/contrib/opensolaris/cmd/zfs/zfs_main.c
+++ b/cddl/contrib/opensolaris/cmd/zfs/zfs_main.c
@@ -274,9 +274,9 @@ get_usage(zfs_help_t idx)
 	case HELP_ROLLBACK:
 		return (gettext("\trollback [-rRf] <snapshot>\n"));
 	case HELP_SEND:
-		return (gettext("\tsend [-DnPpRv] [-[iI] snapshot] "
+		return (gettext("\tsend [-DnPpRve] [-[iI] snapshot] "
 		    "<snapshot>\n"
-		    "\tsend [-i snapshot|bookmark] "
+		    "\tsend [-e] [-i snapshot|bookmark] "
 		    "<filesystem|volume|snapshot>\n"));
 	case HELP_SET:
 		return (gettext("\tset <property=value> "
@@ -590,6 +590,7 @@ finish_progress(char *done)
 	free(pt_header);
 	pt_header = NULL;
 }
+
 /*
  * zfs clone [-p] [-o prop=value] ... <snap> <fs | vol>
  *
@@ -3368,6 +3369,7 @@ rollback_check_dependent(zfs_handle_t *zhp, void *data)
 	zfs_close(zhp);
 	return (0);
 }
+
 /*
  * Report any snapshots more recent than the one specified.  Used when '-r' is
  * not specified.  We reuse this same callback for the snapshot dependents - if
@@ -3707,7 +3709,7 @@ zfs_do_send(int argc, char **argv)
 	boolean_t extraverbose = B_FALSE;
 
 	/* check options */
-	while ((c = getopt(argc, argv, ":i:I:RDpvnP")) != -1) {
+	while ((c = getopt(argc, argv, ":i:I:RDpvnPe")) != -1) {
 		switch (c) {
 		case 'i':
 			if (fromname)
@@ -3742,6 +3744,9 @@ zfs_do_send(int argc, char **argv)
 		case 'n':
 			flags.dryrun = B_TRUE;
 			break;
+		case 'e':
+			flags.embed_data = B_TRUE;
+			break;
 		case ':':
 			(void) fprintf(stderr, gettext("missing argument for "
 			    "'%c' option\n"), optopt);
@@ -3780,6 +3785,7 @@ zfs_do_send(int argc, char **argv)
 	if (strchr(argv[0], '@') == NULL ||
 	    (fromname && strchr(fromname, '#') != NULL)) {
 		char frombuf[ZFS_MAXNAMELEN];
+		enum lzc_send_flags lzc_flags = 0;
 
 		if (flags.replicate || flags.doall || flags.props ||
 		    flags.dedup || flags.dryrun || flags.verbose ||
@@ -3794,6 +3800,9 @@ zfs_do_send(int argc, char **argv)
 		if (zhp == NULL)
 			return (1);
 
+		if (flags.embed_data)
+			lzc_flags |= LZC_SEND_FLAG_EMBED_DATA;
+
 		if (fromname != NULL &&
 		    (fromname[0] == '#' || fromname[0] == '@')) {
 			/*
@@ -3807,7 +3816,7 @@ zfs_do_send(int argc, char **argv)
 			(void) strlcat(frombuf, fromname, sizeof (frombuf));
 			fromname = frombuf;
 		}
-		err = zfs_send_one(zhp, fromname, STDOUT_FILENO);
+		err = zfs_send_one(zhp, fromname, STDOUT_FILENO, lzc_flags);
 		zfs_close(zhp);
 		return (err != 0);
 	}
diff --git a/cddl/contrib/opensolaris/cmd/zpool/zpool-features.7 b/cddl/contrib/opensolaris/cmd/zpool/zpool-features.7
index f9d866b..e2caa64 100644
--- a/cddl/contrib/opensolaris/cmd/zpool/zpool-features.7
+++ b/cddl/contrib/opensolaris/cmd/zpool/zpool-features.7
@@ -23,7 +23,7 @@
 .\"
 .\" $FreeBSD$
 .\"
-.Dd April 23, 2014
+.Dd June 30, 2014
 .Dt ZPOOL-FEATURES 7
 .Os
 .Sh NAME
@@ -396,6 +396,34 @@ This feature becomes
 as soon as it is enabled and will
 never return to being
 .Sy enabled .
+.It Sy embedded_data
+.Bl -column "READ\-ONLY COMPATIBLE" "com.delphix:embedded_data"
+.It GUID Ta com.delphix:embedded_data
+.It READ\-ONLY COMPATIBLE Ta no
+.It DEPENDENCIES Ta none
+.El
+.Pp
+This feature improves the performance and compression ratio of
+highly-compressible blocks.
+Blocks whose contents can compress to 112 bytes
+or smaller can take advantage of this feature.
+.Pp
+When this feature is enabled, the contents of highly-compressible blocks are
+stored in the block "pointer" itself
+.Po a misnomer in this case, as it contains
+the compresseed data, rather than a pointer to its location on disk
+.Pc .
+Thus
+the space of the block
+.Pq one sector, typically 512 bytes or 4KB
+is saved,
+and no additional i/o is needed to read and write the data block.
+.Pp
+This feature becomes
+.Sy active
+as soon as it is enabled and will
+never return to being
+.Sy enabled .
 .El
 .Sh SEE ALSO
 .Xr zpool 8
diff --git a/cddl/contrib/opensolaris/cmd/zstreamdump/zstreamdump.c b/cddl/contrib/opensolaris/cmd/zstreamdump/zstreamdump.c
index f7a4091..dce1cb3 100644
--- a/cddl/contrib/opensolaris/cmd/zstreamdump/zstreamdump.c
+++ b/cddl/contrib/opensolaris/cmd/zstreamdump/zstreamdump.c
@@ -49,7 +49,6 @@
  */
 #define	DUMP_GROUPING	4
 
-uint64_t drr_record_count[DRR_NUMTYPES];
 uint64_t total_write_size = 0;
 uint64_t total_stream_len = 0;
 FILE *send_stream = 0;
@@ -123,7 +122,7 @@ print_block(char *buf, int length)
 	 * Start printing ASCII characters at a constant offset, after
 	 * the hex prints. Leave 3 characters per byte on a line (2 digit
 	 * hex number plus 1 space) plus spaces between characters and
-	 * groupings
+	 * groupings.
 	 */
 	int ascii_start = BYTES_PER_LINE * 3 +
 	    BYTES_PER_LINE / DUMP_GROUPING + 2;
@@ -160,6 +159,8 @@ int
 main(int argc, char *argv[])
 {
 	char *buf = malloc(INITIAL_BUFLEN);
+	uint64_t drr_record_count[DRR_NUMTYPES] = { 0 };
+	uint64_t total_records = 0;
 	dmu_replay_record_t thedrr;
 	dmu_replay_record_t *drr = &thedrr;
 	struct drr_begin *drrb = &thedrr.drr_u.drr_begin;
@@ -170,6 +171,7 @@ main(int argc, char *argv[])
 	struct drr_write_byref *drrwbr = &thedrr.drr_u.drr_write_byref;
 	struct drr_free *drrf = &thedrr.drr_u.drr_free;
 	struct drr_spill *drrs = &thedrr.drr_u.drr_spill;
+	struct drr_write_embedded *drrwe = &thedrr.drr_u.drr_write_embedded;
 	char c;
 	boolean_t verbose = B_FALSE;
 	boolean_t first = B_TRUE;
@@ -264,6 +266,7 @@ main(int argc, char *argv[])
 		}
 
 		drr_record_count[drr->drr_type]++;
+		total_records++;
 
 		switch (drr->drr_type) {
 		case DRR_BEGIN:
@@ -376,8 +379,8 @@ main(int argc, char *argv[])
 				    drro->drr_bonuslen);
 			}
 			if (drro->drr_bonuslen > 0) {
-				(void) ssread(buf, P2ROUNDUP(drro->drr_bonuslen,
-				    8), &zc);
+				(void) ssread(buf,
+				    P2ROUNDUP(drro->drr_bonuslen, 8), &zc);
 				if (dump) {
 					print_block(buf,
 					    P2ROUNDUP(drro->drr_bonuslen, 8));
@@ -506,6 +509,38 @@ main(int argc, char *argv[])
 				print_block(buf, drrs->drr_length);
 			}
 			break;
+		case DRR_WRITE_EMBEDDED:
+			if (do_byteswap) {
+				drrwe->drr_object =
+				    BSWAP_64(drrwe->drr_object);
+				drrwe->drr_offset =
+				    BSWAP_64(drrwe->drr_offset);
+				drrwe->drr_length =
+				    BSWAP_64(drrwe->drr_length);
+				drrwe->drr_toguid =
+				    BSWAP_64(drrwe->drr_toguid);
+				drrwe->drr_lsize =
+				    BSWAP_32(drrwe->drr_lsize);
+				drrwe->drr_psize =
+				    BSWAP_32(drrwe->drr_psize);
+			}
+			if (verbose) {
+				(void) printf("WRITE_EMBEDDED object = %llu "
+				    "offset = %llu length = %llu\n"
+				    "toguid = %llx comp = %u etype = %u "
+				    "lsize = %u psize = %u\n",
+				    (u_longlong_t)drrwe->drr_object,
+				    (u_longlong_t)drrwe->drr_offset,
+				    (u_longlong_t)drrwe->drr_length,
+				    (u_longlong_t)drrwe->drr_toguid,
+				    drrwe->drr_compression,
+				    drrwe->drr_etype,
+				    drrwe->drr_lsize,
+				    drrwe->drr_psize);
+			}
+			(void) ssread(buf,
+			    P2ROUNDUP(drrwe->drr_psize, 8), &zc);
+			break;
 		}
 		pcksum = zc;
 	}
@@ -524,18 +559,16 @@ main(int argc, char *argv[])
 	    (u_longlong_t)drr_record_count[DRR_FREEOBJECTS]);
 	(void) printf("\tTotal DRR_WRITE records = %lld\n",
 	    (u_longlong_t)drr_record_count[DRR_WRITE]);
+	(void) printf("\tTotal DRR_WRITE_BYREF records = %lld\n",
+	    (u_longlong_t)drr_record_count[DRR_WRITE_BYREF]);
+	(void) printf("\tTotal DRR_WRITE_EMBEDDED records = %lld\n",
+	    (u_longlong_t)drr_record_count[DRR_WRITE_EMBEDDED]);
 	(void) printf("\tTotal DRR_FREE records = %lld\n",
 	    (u_longlong_t)drr_record_count[DRR_FREE]);
 	(void) printf("\tTotal DRR_SPILL records = %lld\n",
 	    (u_longlong_t)drr_record_count[DRR_SPILL]);
 	(void) printf("\tTotal records = %lld\n",
-	    (u_longlong_t)(drr_record_count[DRR_BEGIN] +
-	    drr_record_count[DRR_OBJECT] +
-	    drr_record_count[DRR_FREEOBJECTS] +
-	    drr_record_count[DRR_WRITE] +
-	    drr_record_count[DRR_FREE] +
-	    drr_record_count[DRR_SPILL] +
-	    drr_record_count[DRR_END]));
+	    (u_longlong_t)total_records);
 	(void) printf("\tTotal write size = %lld (0x%llx)\n",
 	    (u_longlong_t)total_write_size, (u_longlong_t)total_write_size);
 	(void) printf("\tTotal stream length = %lld (0x%llx)\n",
diff --git a/cddl/contrib/opensolaris/cmd/ztest/ztest.c b/cddl/contrib/opensolaris/cmd/ztest/ztest.c
index 6e028c3..c0ece98 100644
--- a/cddl/contrib/opensolaris/cmd/ztest/ztest.c
+++ b/cddl/contrib/opensolaris/cmd/ztest/ztest.c
@@ -53,7 +53,7 @@
  *     At random times, the child self-immolates with a SIGKILL.
  *     This is the software equivalent of pulling the power cord.
  *     The parent then runs the test again, using the existing
- *     storage pool, as many times as desired. If backwards compatability
+ *     storage pool, as many times as desired. If backwards compatibility
  *     testing is enabled ztest will sometimes run the "older" version
  *     of ztest after a SIGKILL.
  *
@@ -1267,13 +1267,13 @@ static void
 ztest_bt_verify(ztest_block_tag_t *bt, objset_t *os, uint64_t object,
     uint64_t offset, uint64_t gen, uint64_t txg, uint64_t crtxg)
 {
-	ASSERT(bt->bt_magic == BT_MAGIC);
-	ASSERT(bt->bt_objset == dmu_objset_id(os));
-	ASSERT(bt->bt_object == object);
-	ASSERT(bt->bt_offset == offset);
-	ASSERT(bt->bt_gen <= gen);
-	ASSERT(bt->bt_txg <= txg);
-	ASSERT(bt->bt_crtxg == crtxg);
+	ASSERT3U(bt->bt_magic, ==, BT_MAGIC);
+	ASSERT3U(bt->bt_objset, ==, dmu_objset_id(os));
+	ASSERT3U(bt->bt_object, ==, object);
+	ASSERT3U(bt->bt_offset, ==, offset);
+	ASSERT3U(bt->bt_gen, <=, gen);
+	ASSERT3U(bt->bt_txg, <=, txg);
+	ASSERT3U(bt->bt_crtxg, ==, crtxg);
 }
 
 static ztest_block_tag_t *
@@ -3472,6 +3472,11 @@ ztest_dsl_dataset_promote_busy(ztest_ds_t *zd, uint64_t id)
 	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);
+		ztest_record_enospc(FTAG);
+		goto out;
+	}
 	if (error != EBUSY)
 		fatal(0, "dsl_dataset_promote(%s), %d, not EBUSY", clone2name,
 		    error);
@@ -3627,11 +3632,19 @@ ztest_dmu_read_write(ztest_ds_t *zd, uint64_t id)
 		return;
 	}
 
-	dmu_object_set_checksum(os, bigobj,
-	    (enum zio_checksum)ztest_random_dsl_prop(ZFS_PROP_CHECKSUM), tx);
+	enum zio_checksum cksum;
+	do {
+		cksum = (enum zio_checksum)
+		    ztest_random_dsl_prop(ZFS_PROP_CHECKSUM);
+	} while (cksum >= ZIO_CHECKSUM_LEGACY_FUNCTIONS);
+	dmu_object_set_checksum(os, bigobj, cksum, tx);
 
-	dmu_object_set_compress(os, bigobj,
-	    (enum zio_compress)ztest_random_dsl_prop(ZFS_PROP_COMPRESSION), tx);
+	enum zio_compress comp;
+	do {
+		comp = (enum zio_compress)
+		    ztest_random_dsl_prop(ZFS_PROP_COMPRESSION);
+	} while (comp >= ZIO_COMPRESS_LEGACY_FUNCTIONS);
+	dmu_object_set_compress(os, bigobj, comp, tx);
 
 	/*
 	 * For each index from n to n + s, verify that the existing bufwad
@@ -4711,8 +4724,13 @@ ztest_dmu_snapshot_hold(ztest_ds_t *zd, uint64_t id)
 	error = dsl_dataset_user_hold(holds, 0, NULL);
 	fnvlist_free(holds);
 
-	if (error)
-		fatal(0, "dsl_dataset_user_hold(%s)", fullname, tag);
+	if (error == ENOSPC) {
+		ztest_record_enospc("dsl_dataset_user_hold");
+		goto out;
+	} else if (error) {
+		fatal(0, "dsl_dataset_user_hold(%s, %s) = %u",
+		    fullname, tag, error);
+	}
 
 	error = dsl_destroy_snapshot(fullname, B_FALSE);
 	if (error != EBUSY) {
@@ -5165,7 +5183,7 @@ ztest_run_zdb(char *pool)
 	isa = strdup(isa);
 	/* LINTED */
 	(void) sprintf(bin,
-	    "/usr/sbin%.*s/zdb -bcc%s%s -U %s %s",
+	    "/usr/sbin%.*s/zdb -bcc%s%s -d -U %s %s",
 	    isalen,
 	    isa,
 	    ztest_opts.zo_verbose >= 3 ? "s" : "",
diff --git a/cddl/contrib/opensolaris/lib/libzfs/common/libzfs.h b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs.h
index 24131d6..ef18b45 100644
--- a/cddl/contrib/opensolaris/lib/libzfs/common/libzfs.h
+++ b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs.h
@@ -42,6 +42,7 @@
 #include <sys/fs/zfs.h>
 #include <sys/avl.h>
 #include <sys/zfs_ioctl.h>
+#include <libzfs_core.h>
 
 #ifdef	__cplusplus
 extern "C" {
@@ -607,13 +608,16 @@ typedef struct sendflags {
 
 	/* show progress (ie. -v) */
 	boolean_t progress;
+
+	/* WRITE_EMBEDDED records of type DATA are permitted */
+	boolean_t embed_data;
 } sendflags_t;
 
 typedef boolean_t (snapfilter_cb_t)(zfs_handle_t *, void *);
 
 extern int zfs_send(zfs_handle_t *, const char *, const char *,
     sendflags_t *, int, snapfilter_cb_t, void *, nvlist_t **);
-extern int zfs_send_one(zfs_handle_t *, const char *, int);
+extern int zfs_send_one(zfs_handle_t *, const char *, int, enum lzc_send_flags);
 
 extern int zfs_promote(zfs_handle_t *);
 extern int zfs_hold(zfs_handle_t *, const char *, const char *,
diff --git a/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c
index feddb69..97f18d7 100644
--- a/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c
+++ b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c
@@ -21,7 +21,7 @@
 
 /*
  * Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
- * Copyright (c) 2013 by Delphix. All rights reserved.
+ * Copyright (c) 2011, 2014 by Delphix. All rights reserved.
  * Copyright (c) 2012, Joyent, Inc. All rights reserved.
  * Copyright (c) 2012 Pawel Jakub Dawidek <pawel@dawidek.net>.
  * All rights reserved.
@@ -45,6 +45,7 @@
 #include <time.h>
 
 #include <libzfs.h>
+#include <libzfs_core.h>
 
 #include "zfs_namecheck.h"
 #include "zfs_prop.h"
@@ -222,6 +223,7 @@ cksummer(void *arg)
 	struct drr_object *drro = &thedrr.drr_u.drr_object;
 	struct drr_write *drrw = &thedrr.drr_u.drr_write;
 	struct drr_spill *drrs = &thedrr.drr_u.drr_spill;
+	struct drr_write_embedded *drrwe = &thedrr.drr_u.drr_write_embedded;
 	FILE *ofp;
 	int outfd;
 	dmu_replay_record_t wbr_drr = {0};
@@ -418,6 +420,20 @@ cksummer(void *arg)
 			break;
 		}
 
+		case DRR_WRITE_EMBEDDED:
+		{
+			if (cksum_and_write(drr, sizeof (dmu_replay_record_t),
+			    &stream_cksum, outfd) == -1)
+				goto out;
+			(void) ssread(buf,
+			    P2ROUNDUP((uint64_t)drrwe->drr_psize, 8), ofp);
+			if (cksum_and_write(buf,
+			    P2ROUNDUP((uint64_t)drrwe->drr_psize, 8),
+			    &stream_cksum, outfd) == -1)
+				goto out;
+			break;
+		}
+
 		case DRR_FREE:
 		{
 			if (cksum_and_write(drr, sizeof (dmu_replay_record_t),
@@ -799,7 +815,7 @@ typedef struct send_dump_data {
 	char prevsnap[ZFS_MAXNAMELEN];
 	uint64_t prevsnap_obj;
 	boolean_t seenfrom, seento, replicate, doall, fromorigin;
-	boolean_t verbose, dryrun, parsable, progress;
+	boolean_t verbose, dryrun, parsable, progress, embed_data;
 	int outfd;
 	boolean_t err;
 	nvlist_t *fss;
@@ -878,7 +894,8 @@ estimate_ioctl(zfs_handle_t *zhp, uint64_t fromsnap_obj,
  */
 static int
 dump_ioctl(zfs_handle_t *zhp, const char *fromsnap, uint64_t fromsnap_obj,
-    boolean_t fromorigin, int outfd, nvlist_t *debugnv)
+    boolean_t fromorigin, int outfd, enum lzc_send_flags flags,
+    nvlist_t *debugnv)
 {
 	zfs_cmd_t zc = { 0 };
 	libzfs_handle_t *hdl = zhp->zfs_hdl;
@@ -892,6 +909,7 @@ dump_ioctl(zfs_handle_t *zhp, const char *fromsnap, uint64_t fromsnap_obj,
 	zc.zc_obj = fromorigin;
 	zc.zc_sendobj = zfs_prop_get_int(zhp, ZFS_PROP_OBJSETID);
 	zc.zc_fromobj = fromsnap_obj;
+	zc.zc_flags = flags;
 
 	VERIFY(0 == nvlist_alloc(&thisdbg, NV_UNIQUE_NAME, 0));
 	if (fromsnap && fromsnap[0] != '\0') {
@@ -1144,8 +1162,12 @@ dump_snapshot(zfs_handle_t *zhp, void *arg)
 			}
 		}
 
+		enum lzc_send_flags flags = 0;
+		if (sdd->embed_data)
+			flags |= LZC_SEND_FLAG_EMBED_DATA;
+
 		err = dump_ioctl(zhp, sdd->prevsnap, sdd->prevsnap_obj,
-		    fromorigin, sdd->outfd, sdd->debugnv);
+		    fromorigin, sdd->outfd, flags, sdd->debugnv);
 
 		if (sdd->progress) {
 			(void) pthread_cancel(tid);
@@ -1489,6 +1511,7 @@ zfs_send(zfs_handle_t *zhp, const char *fromsnap, const char *tosnap,
 	sdd.parsable = flags->parsable;
 	sdd.progress = flags->progress;
 	sdd.dryrun = flags->dryrun;
+	sdd.embed_data = flags->embed_data;
 	sdd.filter_cb = filter_func;
 	sdd.filter_cb_arg = cb_arg;
 	if (debugnvp)
@@ -1620,7 +1643,8 @@ err_out:
 }
 
 int
-zfs_send_one(zfs_handle_t *zhp, const char *from, int fd)
+zfs_send_one(zfs_handle_t *zhp, const char *from, int fd,
+    enum lzc_send_flags flags)
 {
 	int err;
 	libzfs_handle_t *hdl = zhp->zfs_hdl;
@@ -1629,7 +1653,7 @@ zfs_send_one(zfs_handle_t *zhp, const char *from, int fd)
 	(void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN,
 	    "warning: cannot send '%s'"), zhp->zfs_name);
 
-	err = lzc_send(zhp->zfs_name, from, fd);
+	err = lzc_send(zhp->zfs_name, from, fd, flags);
 	if (err != 0) {
 		switch (errno) {
 		case EXDEV:
@@ -2576,6 +2600,16 @@ recv_skip(libzfs_handle_t *hdl, int fd, boolean_t byteswap)
 			(void) recv_read(hdl, fd, buf,
 			    drr->drr_u.drr_spill.drr_length, B_FALSE, NULL);
 			break;
+		case DRR_WRITE_EMBEDDED:
+			if (byteswap) {
+				drr->drr_u.drr_write_embedded.drr_psize =
+				    BSWAP_32(drr->drr_u.drr_write_embedded.
+				    drr_psize);
+			}
+			(void) recv_read(hdl, fd, buf,
+			    P2ROUNDUP(drr->drr_u.drr_write_embedded.drr_psize,
+			    8), B_FALSE, NULL);
+			break;
 		case DRR_WRITE_BYREF:
 		case DRR_FREEOBJECTS:
 		case DRR_FREE:
diff --git a/cddl/contrib/opensolaris/lib/libzfs_core/common/libzfs_core.c b/cddl/contrib/opensolaris/lib/libzfs_core/common/libzfs_core.c
index 1c87223..cb38dc2 100644
--- a/cddl/contrib/opensolaris/lib/libzfs_core/common/libzfs_core.c
+++ b/cddl/contrib/opensolaris/lib/libzfs_core/common/libzfs_core.c
@@ -486,6 +486,8 @@ lzc_get_holds(const char *snapname, nvlist_t **holdsp)
 }
 
 /*
+ * Generate a zfs send stream for the specified snapshot and write it to
+ * the specified file descriptor.
  *
  * "snapname" is the full name of the snapshot to send (e.g. "pool/fs@snap")
  *
@@ -499,9 +501,15 @@ lzc_get_holds(const char *snapname, nvlist_t **holdsp)
  * snapshot in the origin, etc.
  *
  * "fd" is the file descriptor to write the send stream to.
+ *
+ * If "flags" contains LZC_SEND_FLAG_EMBED_DATA, the stream is permitted
+ * to contain DRR_WRITE_EMBEDDED records with drr_etype==BP_EMBEDDED_TYPE_DATA,
+ * which the receiving system must support (as indicated by support
+ * for the "embedded_data" feature).
  */
 int
-lzc_send(const char *snapname, const char *from, int fd)
+lzc_send(const char *snapname, const char *from, int fd,
+    enum lzc_send_flags flags)
 {
 	nvlist_t *args;
 	int err;
@@ -510,6 +518,8 @@ lzc_send(const char *snapname, const char *from, int fd)
 	fnvlist_add_int32(args, "fd", fd);
 	if (from != NULL)
 		fnvlist_add_string(args, "fromsnap", from);
+	if (flags & LZC_SEND_FLAG_EMBED_DATA)
+		fnvlist_add_boolean(args, "embedok");
 	err = lzc_ioctl(ZFS_IOC_SEND_NEW, snapname, args, NULL);
 	nvlist_free(args);
 	return (err);
diff --git a/cddl/contrib/opensolaris/lib/libzfs_core/common/libzfs_core.h b/cddl/contrib/opensolaris/lib/libzfs_core/common/libzfs_core.h
index 380560f..99883fe 100644
--- a/cddl/contrib/opensolaris/lib/libzfs_core/common/libzfs_core.h
+++ b/cddl/contrib/opensolaris/lib/libzfs_core/common/libzfs_core.h
@@ -53,7 +53,11 @@ int lzc_hold(nvlist_t *, int, nvlist_t **);
 int lzc_release(nvlist_t *, nvlist_t **);
 int lzc_get_holds(const char *, nvlist_t **);
 
-int lzc_send(const char *, const char *, int);
+enum lzc_send_flags {
+	LZC_SEND_FLAG_EMBED_DATA = 1 << 0
+};
+
+int lzc_send(const char *, const char *, int, enum lzc_send_flags);
 int lzc_receive(const char *, nvlist_t *, const char *, boolean_t, int);
 int lzc_send_space(const char *, const char *, uint64_t *);
 
diff --git a/cddl/sbin/zpool/Makefile b/cddl/sbin/zpool/Makefile
index 1884d24..bfcb017 100644
--- a/cddl/sbin/zpool/Makefile
+++ b/cddl/sbin/zpool/Makefile
@@ -18,6 +18,7 @@ CFLAGS+= -I${.CURDIR}/../../../cddl/contrib/opensolaris/head
 CFLAGS+= -I${.CURDIR}/../../../cddl/contrib/opensolaris/lib/libuutil/common
 CFLAGS+= -I${.CURDIR}/../../../cddl/contrib/opensolaris/lib/libumem/common
 CFLAGS+= -I${.CURDIR}/../../../cddl/contrib/opensolaris/lib/libzfs/common
+CFLAGS+= -I${.CURDIR}/../../../cddl/contrib/opensolaris/lib/libzfs_core/common
 CFLAGS+= -I${.CURDIR}/../../../cddl/contrib/opensolaris/lib/libnvpair
 CFLAGS+= -I${.CURDIR}/../../../sys/cddl/contrib/opensolaris/common/zfs
 CFLAGS+= -I${.CURDIR}/../../../sys/cddl/contrib/opensolaris/uts/common
diff --git a/cddl/usr.bin/zinject/Makefile b/cddl/usr.bin/zinject/Makefile
index 5625149..20d4932 100644
--- a/cddl/usr.bin/zinject/Makefile
+++ b/cddl/usr.bin/zinject/Makefile
@@ -11,6 +11,7 @@ CFLAGS+= -I${.CURDIR}/../../../sys/cddl/compat/opensolaris
 CFLAGS+= -I${.CURDIR}/../../compat/opensolaris/include
 CFLAGS+= -I${.CURDIR}/../../compat/opensolaris/lib/libumem
 CFLAGS+= -I${.CURDIR}/../../contrib/opensolaris/lib/libzfs/common
+CFLAGS+= -I${.CURDIR}/../../contrib/opensolaris/lib/libzfs_core/common
 CFLAGS+= -I${.CURDIR}/../../contrib/opensolaris/lib/libzpool/common
 CFLAGS+= -I${.CURDIR}/../../contrib/opensolaris/lib/libnvpair
 CFLAGS+= -I${.CURDIR}/../../../sys/cddl/contrib/opensolaris/uts/common/fs/zfs
diff --git a/cddl/usr.sbin/zdb/Makefile b/cddl/usr.sbin/zdb/Makefile
index fa9ab98..0e2ec8f 100644
--- a/cddl/usr.sbin/zdb/Makefile
+++ b/cddl/usr.sbin/zdb/Makefile
@@ -15,6 +15,7 @@ CFLAGS+= -I${.CURDIR}/../../../cddl/compat/opensolaris/lib/libumem
 CFLAGS+= -I${.CURDIR}/../../../cddl/contrib/opensolaris/lib/libnvpair
 CFLAGS+= -I${.CURDIR}/../../../cddl/contrib/opensolaris/lib/libuutil/common
 CFLAGS+= -I${.CURDIR}/../../../cddl/contrib/opensolaris/lib/libzfs/common
+CFLAGS+= -I${.CURDIR}/../../../cddl/contrib/opensolaris/lib/libzfs_core/common
 CFLAGS+= -I${.CURDIR}/../../../cddl/contrib/opensolaris/lib/libzpool/common
 CFLAGS+= -I${.CURDIR}/../../../sys/cddl/contrib/opensolaris/uts/common/fs/zfs
 CFLAGS+= -I${.CURDIR}/../../../sys/cddl/contrib/opensolaris/uts/common
diff --git a/cddl/usr.sbin/zhack/Makefile b/cddl/usr.sbin/zhack/Makefile
index f09d2d8..b8ad1b4 100644
--- a/cddl/usr.sbin/zhack/Makefile
+++ b/cddl/usr.sbin/zhack/Makefile
@@ -14,6 +14,7 @@ CFLAGS+= -I${.CURDIR}/../../../cddl/compat/opensolaris/lib/libumem
 CFLAGS+= -I${.CURDIR}/../../../cddl/contrib/opensolaris/lib/libnvpair
 CFLAGS+= -I${.CURDIR}/../../../cddl/contrib/opensolaris/lib/libuutil/common
 CFLAGS+= -I${.CURDIR}/../../../cddl/contrib/opensolaris/lib/libzfs/common
+CFLAGS+= -I${.CURDIR}/../../../cddl/contrib/opensolaris/lib/libzfs_core/common
 CFLAGS+= -I${.CURDIR}/../../../cddl/contrib/opensolaris/lib/libzpool/common
 CFLAGS+= -I${.CURDIR}/../../../sys/cddl/contrib/opensolaris/uts/common/fs/zfs
 CFLAGS+= -I${.CURDIR}/../../../sys/cddl/contrib/opensolaris/uts/common