Tag 2.2.0-rc5

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

.cirrus.yml

@@ -0,0 +1,21 @@
+env:
+  CIRRUS_CLONE_DEPTH: 1
+  ARCH: amd64
+
+build_task:
+  matrix:
+    freebsd_instance:
+      image_family: freebsd-12-4
+    freebsd_instance:
+      image_family: freebsd-13-2
+    freebsd_instance:
+      image_family: freebsd-14-0-snap
+  prepare_script:
+    - pkg install -y autoconf automake libtool gettext-runtime gmake ksh93 py39-packaging py39-cffi py39-sysctl
+  configure_script:
+    - env MAKE=gmake ./autogen.sh
+    - env MAKE=gmake ./configure --with-config="user" --with-python=3.9
+  build_script:
+    - gmake -j `sysctl -n kern.smp.cpus`
+  install_script:
+    - gmake install

.gitignore

@@ -42,6 +42,7 @@
 !udev/**
 
 !.editorconfig
+!.cirrus.yml
 !.gitignore
 !.gitmodules
 !AUTHORS
@@ -60,7 +61,6 @@
 !TEST
 !zfs.release.in
 
-
 #
 # Normal rules
 #

META

@@ -2,9 +2,9 @@ Meta:          1
 Name:          zfs
 Branch:        1.0
 Version:       2.2.0
-Release:       rc1
+Release:       rc5
 Release-Tags:  relext
 License:       CDDL
 Author:        OpenZFS
-Linux-Maximum: 6.3
+Linux-Maximum: 6.5
 Linux-Minimum: 3.10

cmd/zdb/zdb.c

@@ -79,6 +79,7 @@
 #include <sys/dsl_crypt.h>
 #include <sys/dsl_scan.h>
 #include <sys/btree.h>
+#include <sys/brt.h>
 #include <zfs_comutil.h>
 #include <sys/zstd/zstd.h>
 
@@ -5178,7 +5179,7 @@ dump_label(const char *dev)
 			if (nvlist_size(config, &size, NV_ENCODE_XDR) != 0)
 				size = buflen;
 
-			/* If the device is a cache device clear the header. */
+			/* If the device is a cache device read the header. */
 			if (!read_l2arc_header) {
 				if (nvlist_lookup_uint64(config,
 				    ZPOOL_CONFIG_POOL_STATE, &l2cache) == 0 &&
@@ -5342,12 +5343,20 @@ static const char *zdb_ot_extname[] = {
 #define	ZB_TOTAL	DN_MAX_LEVELS
 #define	SPA_MAX_FOR_16M	(SPA_MAXBLOCKSHIFT+1)
 
+typedef struct zdb_brt_entry {
+	dva_t		zbre_dva;
+	uint64_t	zbre_refcount;
+	avl_node_t	zbre_node;
+} zdb_brt_entry_t;
+
 typedef struct zdb_cb {
 	zdb_blkstats_t	zcb_type[ZB_TOTAL + 1][ZDB_OT_TOTAL + 1];
 	uint64_t	zcb_removing_size;
 	uint64_t	zcb_checkpoint_size;
 	uint64_t	zcb_dedup_asize;
 	uint64_t	zcb_dedup_blocks;
+	uint64_t	zcb_clone_asize;
+	uint64_t	zcb_clone_blocks;
 	uint64_t	zcb_psize_count[SPA_MAX_FOR_16M];
 	uint64_t	zcb_lsize_count[SPA_MAX_FOR_16M];
 	uint64_t	zcb_asize_count[SPA_MAX_FOR_16M];
@@ -5368,6 +5377,8 @@ typedef struct zdb_cb {
 	int		zcb_haderrors;
 	spa_t		*zcb_spa;
 	uint32_t	**zcb_vd_obsolete_counts;
+	avl_tree_t	zcb_brt;
+	boolean_t	zcb_brt_is_active;
 } zdb_cb_t;
 
 /* test if two DVA offsets from same vdev are within the same metaslab */
@@ -5662,6 +5673,45 @@ zdb_count_block(zdb_cb_t *zcb, zilog_t *zilog, const blkptr_t *bp,
 	zcb->zcb_asize_len[bin] += BP_GET_ASIZE(bp);
 	zcb->zcb_asize_total += BP_GET_ASIZE(bp);
 
+	if (zcb->zcb_brt_is_active && brt_maybe_exists(zcb->zcb_spa, bp)) {
+		/*
+		 * Cloned blocks are special. We need to count them, so we can
+		 * later uncount them when reporting leaked space, and we must
+		 * only claim them them once.
+		 *
+		 * To do this, we keep our own in-memory BRT. For each block
+		 * we haven't seen before, we look it up in the real BRT and
+		 * if its there, we note it and its refcount then proceed as
+		 * normal. If we see the block again, we count it as a clone
+		 * and then give it no further consideration.
+		 */
+		zdb_brt_entry_t zbre_search, *zbre;
+		avl_index_t where;
+
+		zbre_search.zbre_dva = bp->blk_dva[0];
+		zbre = avl_find(&zcb->zcb_brt, &zbre_search, &where);
+		if (zbre != NULL) {
+			zcb->zcb_clone_asize += BP_GET_ASIZE(bp);
+			zcb->zcb_clone_blocks++;
+
+			zbre->zbre_refcount--;
+			if (zbre->zbre_refcount == 0) {
+				avl_remove(&zcb->zcb_brt, zbre);
+				umem_free(zbre, sizeof (zdb_brt_entry_t));
+			}
+			return;
+		}
+
+		uint64_t crefcnt = brt_entry_get_refcount(zcb->zcb_spa, bp);
+		if (crefcnt > 0) {
+			zbre = umem_zalloc(sizeof (zdb_brt_entry_t),
+			    UMEM_NOFAIL);
+			zbre->zbre_dva = bp->blk_dva[0];
+			zbre->zbre_refcount = crefcnt;
+			avl_insert(&zcb->zcb_brt, zbre, where);
+		}
+	}
+
 	if (dump_opt['L'])
 		return;
 
@@ -6664,6 +6714,20 @@ deleted_livelists_dump_mos(spa_t *spa)
 	iterate_deleted_livelists(spa, dump_livelist_cb, NULL);
 }
 
+static int
+zdb_brt_entry_compare(const void *zcn1, const void *zcn2)
+{
+	const dva_t *dva1 = &((const zdb_brt_entry_t *)zcn1)->zbre_dva;
+	const dva_t *dva2 = &((const zdb_brt_entry_t *)zcn2)->zbre_dva;
+	int cmp;
+
+	cmp = TREE_CMP(DVA_GET_VDEV(dva1), DVA_GET_VDEV(dva2));
+	if (cmp == 0)
+		cmp = TREE_CMP(DVA_GET_OFFSET(dva1), DVA_GET_OFFSET(dva2));
+
+	return (cmp);
+}
+
 static int
 dump_block_stats(spa_t *spa)
 {
@@ -6678,6 +6742,13 @@ dump_block_stats(spa_t *spa)
 
 	zcb = umem_zalloc(sizeof (zdb_cb_t), UMEM_NOFAIL);
 
+	if (spa_feature_is_active(spa, SPA_FEATURE_BLOCK_CLONING)) {
+		avl_create(&zcb->zcb_brt, zdb_brt_entry_compare,
+		    sizeof (zdb_brt_entry_t),
+		    offsetof(zdb_brt_entry_t, zbre_node));
+		zcb->zcb_brt_is_active = B_TRUE;
+	}
+
 	(void) printf("\nTraversing all blocks %s%s%s%s%s...\n\n",
 	    (dump_opt['c'] || !dump_opt['L']) ? "to verify " : "",
 	    (dump_opt['c'] == 1) ? "metadata " : "",
@@ -6779,7 +6850,8 @@ dump_block_stats(spa_t *spa)
 	    metaslab_class_get_alloc(spa_special_class(spa)) +
 	    metaslab_class_get_alloc(spa_dedup_class(spa)) +
 	    get_unflushed_alloc_space(spa);
-	total_found = tzb->zb_asize - zcb->zcb_dedup_asize +
+	total_found =
+	    tzb->zb_asize - zcb->zcb_dedup_asize - zcb->zcb_clone_asize +
 	    zcb->zcb_removing_size + zcb->zcb_checkpoint_size;
 
 	if (total_found == total_alloc && !dump_opt['L']) {
@@ -6820,6 +6892,9 @@ dump_block_stats(spa_t *spa)
 	    "bp deduped:", (u_longlong_t)zcb->zcb_dedup_asize,
 	    (u_longlong_t)zcb->zcb_dedup_blocks,
 	    (double)zcb->zcb_dedup_asize / tzb->zb_asize + 1.0);
+	(void) printf("\t%-16s %14llu    count: %6llu\n",
+	    "bp cloned:", (u_longlong_t)zcb->zcb_clone_asize,
+	    (u_longlong_t)zcb->zcb_clone_blocks);
 	(void) printf("\t%-16s %14llu     used: %5.2f%%\n", "Normal class:",
 	    (u_longlong_t)norm_alloc, 100.0 * norm_alloc / norm_space);
 

cmd/zed/agents/zfs_mod.c

@@ -372,6 +372,7 @@ zfs_process_add(zpool_handle_t *zhp, nvlist_t *vdev, boolean_t labeled)
 	/* Only autoreplace bad disks */
 	if ((vs->vs_state != VDEV_STATE_DEGRADED) &&
 	    (vs->vs_state != VDEV_STATE_FAULTED) &&
+	    (vs->vs_state != VDEV_STATE_REMOVED) &&
 	    (vs->vs_state != VDEV_STATE_CANT_OPEN)) {
 		zed_log_msg(LOG_INFO, "  not autoreplacing since disk isn't in "
 		    "a bad state (currently %llu)", vs->vs_state);
@@ -607,8 +608,6 @@ zfs_iter_vdev(zpool_handle_t *zhp, nvlist_t *nvl, void *data)
 		 */
 		if (nvlist_lookup_string(nvl, dp->dd_prop, &path) != 0 ||
 		    strcmp(dp->dd_compare, path) != 0) {
-			zed_log_msg(LOG_INFO, "  %s: no match (%s != vdev %s)",
-			    __func__, dp->dd_compare, path);
 			return;
 		}
 		if (dp->dd_new_vdev_guid != 0 && dp->dd_new_vdev_guid != guid) {

cmd/zed/agents/zfs_retire.c

@@ -416,6 +416,11 @@ zfs_retire_recv(fmd_hdl_t *hdl, fmd_event_t *ep, nvlist_t *nvl,
 		    FM_EREPORT_PAYLOAD_ZFS_VDEV_GUID, &vdev_guid) != 0)
 			return;
 
+		if (vdev_guid == 0) {
+			fmd_hdl_debug(hdl, "Got a zero GUID");
+			return;
+		}
+
 		if (spare) {
 			int nspares = find_and_remove_spares(zhdl, vdev_guid);
 			fmd_hdl_debug(hdl, "%d spares removed", nspares);

cmd/zed/zed.d/Makefile.am

@@ -16,6 +16,7 @@ dist_zedexec_SCRIPTS = \
 	%D%/scrub_finish-notify.sh \
 	%D%/statechange-led.sh \
 	%D%/statechange-notify.sh \
+	%D%/statechange-slot_off.sh \
 	%D%/trim_finish-notify.sh \
 	%D%/vdev_attach-led.sh \
 	%D%/vdev_clear-led.sh
@@ -35,6 +36,7 @@ zedconfdefaults = \
 	scrub_finish-notify.sh \
 	statechange-led.sh \
 	statechange-notify.sh \
+	statechange-slot_off.sh \
 	vdev_attach-led.sh \
 	vdev_clear-led.sh
 

cmd/zed/zed.d/statechange-led.sh

@@ -121,7 +121,7 @@ state_to_val()
 {
 	state="$1"
 	case "$state" in
-		FAULTED|DEGRADED|UNAVAIL)
+		FAULTED|DEGRADED|UNAVAIL|REMOVED)
 			echo 1
 			;;
 		ONLINE)

cmd/zed/zed.d/statechange-slot_off.sh

@@ -0,0 +1,64 @@
+#!/bin/sh
+# shellcheck disable=SC3014,SC2154,SC2086,SC2034
+#
+# Turn off disk's enclosure slot if it becomes FAULTED.
+#
+# Bad SCSI disks can often "disappear and reappear" causing all sorts of chaos
+# as they flip between FAULTED and ONLINE.  If
+# ZED_POWER_OFF_ENCLOUSRE_SLOT_ON_FAULT is set in zed.rc, and the disk gets
+# FAULTED, then power down the slot via sysfs:
+#
+# /sys/class/enclosure/<enclosure>/<slot>/power_status
+#
+# We assume the user will be responsible for turning the slot back on again.
+#
+# Note that this script requires that your enclosure be supported by the
+# Linux SCSI Enclosure services (SES) driver.  The script will do nothing
+# if you have no enclosure, or if your enclosure isn't supported.
+#
+# Exit codes:
+#   0: slot successfully powered off
+#   1: enclosure not available
+#   2: ZED_POWER_OFF_ENCLOUSRE_SLOT_ON_FAULT disabled
+#   3: vdev was not FAULTED
+#   4: The enclosure sysfs path passed from ZFS does not exist
+#   5: Enclosure slot didn't actually turn off after we told it to
+
+[ -f "${ZED_ZEDLET_DIR}/zed.rc" ] && . "${ZED_ZEDLET_DIR}/zed.rc"
+. "${ZED_ZEDLET_DIR}/zed-functions.sh"
+
+if [ ! -d /sys/class/enclosure ] ; then
+	# No JBOD enclosure or NVMe slots
+	exit 1
+fi
+
+if [ "${ZED_POWER_OFF_ENCLOUSRE_SLOT_ON_FAULT}" != "1" ] ; then
+	exit 2
+fi
+
+if [ "$ZEVENT_VDEV_STATE_STR" != "FAULTED" ] ; then
+	exit 3
+fi
+
+if [ ! -f "$ZEVENT_VDEV_ENC_SYSFS_PATH/power_status" ] ; then
+	exit 4
+fi
+
+# Turn off the slot and wait for sysfs to report that the slot is off.
+# It can take ~400ms on some enclosures and multiple retries may be needed.
+for i in $(seq 1 20) ; do
+	echo "off" | tee "$ZEVENT_VDEV_ENC_SYSFS_PATH/power_status"
+
+	for j in $(seq 1 5) ; do
+		if [ "$(cat $ZEVENT_VDEV_ENC_SYSFS_PATH/power_status)" == "off" ] ; then
+			break 2
+		fi
+		sleep 0.1
+	done
+done
+
+if [ "$(cat $ZEVENT_VDEV_ENC_SYSFS_PATH/power_status)" != "off" ] ; then
+	exit 5
+fi
+
+zed_log_msg "powered down slot $ZEVENT_VDEV_ENC_SYSFS_PATH for $ZEVENT_VDEV_PATH"

cmd/zed/zed.d/zed.rc

@@ -142,3 +142,8 @@ ZED_SYSLOG_SUBCLASS_EXCLUDE="history_event"
 # Disabled by default, 1 to enable and 0 to disable.
 #ZED_SYSLOG_DISPLAY_GUIDS=1
 
+##
+# Power off the drive's slot in the enclosure if it becomes FAULTED.  This can
+# help silence misbehaving drives.  This assumes your drive enclosure fully
+# supports slot power control via sysfs.
+#ZED_POWER_OFF_ENCLOUSRE_SLOT_ON_FAULT=1

cmd/zfs/zfs_main.c

@@ -132,6 +132,8 @@ static int zfs_do_zone(int argc, char **argv);
 static int zfs_do_unzone(int argc, char **argv);
 #endif
 
+static int zfs_do_help(int argc, char **argv);
+
 /*
  * Enable a reasonable set of defaults for libumem debugging on DEBUG builds.
  */
@@ -337,7 +339,7 @@ get_usage(zfs_help_t idx)
 		    "\tsend [-nVvPe] -t <receive_resume_token>\n"
 		    "\tsend [-PnVv] --saved filesystem\n"));
 	case HELP_SET:
-		return (gettext("\tset <property=value> ... "
+		return (gettext("\tset [-u] <property=value> ... "
 		    "<filesystem|volume|snapshot> ...\n"));
 	case HELP_SHARE:
 		return (gettext("\tshare [-l] <-a [nfs|smb] | filesystem>\n"));
@@ -606,6 +608,9 @@ usage(boolean_t requested)
 		(void) fprintf(fp,
 		    gettext("\nFor the delegated permission list, run: %s\n"),
 		    "zfs allow|unallow");
+		(void) fprintf(fp,
+		    gettext("\nFor further help on a command or topic, "
+		    "run: %s\n"), "zfs help [<topic>]");
 	}
 
 	/*
@@ -4197,9 +4202,10 @@ out:
 static int
 set_callback(zfs_handle_t *zhp, void *data)
 {
-	nvlist_t *props = data;
+	zprop_set_cbdata_t *cb = data;
+	int ret = zfs_prop_set_list_flags(zhp, cb->cb_proplist, cb->cb_flags);
 
-	if (zfs_prop_set_list(zhp, props) != 0) {
+	if (ret != 0 || libzfs_errno(g_zfs) != EZFS_SUCCESS) {
 		switch (libzfs_errno(g_zfs)) {
 		case EZFS_MOUNTFAILED:
 			(void) fprintf(stderr, gettext("property may be set "
@@ -4210,33 +4216,42 @@ set_callback(zfs_handle_t *zhp, void *data)
 			    "but unable to reshare filesystem\n"));
 			break;
 		}
-		return (1);
 	}
-	return (0);
+	return (ret);
 }
 
 static int
 zfs_do_set(int argc, char **argv)
 {
-	nvlist_t *props = NULL;
+	zprop_set_cbdata_t cb = { 0 };
 	int ds_start = -1; /* argv idx of first dataset arg */
 	int ret = 0;
-	int i;
+	int i, c;
 
-	/* check for options */
-	if (argc > 1 && argv[1][0] == '-') {
+	/* check options */
+	while ((c = getopt(argc, argv, "u")) != -1) {
+		switch (c) {
+		case 'u':
+			cb.cb_flags |= ZFS_SET_NOMOUNT;
+			break;
+		case '?':
+		default:
 			(void) fprintf(stderr, gettext("invalid option '%c'\n"),
-		    argv[1][1]);
+			    optopt);
 			usage(B_FALSE);
 		}
+	}
+
+	argc -= optind;
+	argv += optind;
 
 	/* check number of arguments */
-	if (argc < 2) {
+	if (argc < 1) {
 		(void) fprintf(stderr, gettext("missing arguments\n"));
 		usage(B_FALSE);
 	}
-	if (argc < 3) {
-		if (strchr(argv[1], '=') == NULL) {
+	if (argc < 2) {
+		if (strchr(argv[0], '=') == NULL) {
 			(void) fprintf(stderr, gettext("missing property=value "
 			    "argument(s)\n"));
 		} else {
@@ -4247,7 +4262,7 @@ zfs_do_set(int argc, char **argv)
 	}
 
 	/* validate argument order:  prop=val args followed by dataset args */
-	for (i = 1; i < argc; i++) {
+	for (i = 0; i < argc; i++) {
 		if (strchr(argv[i], '=') != NULL) {
 			if (ds_start > 0) {
 				/* out-of-order prop=val argument */
@@ -4265,20 +4280,20 @@ zfs_do_set(int argc, char **argv)
 	}
 
 	/* Populate a list of property settings */
-	if (nvlist_alloc(&props, NV_UNIQUE_NAME, 0) != 0)
+	if (nvlist_alloc(&cb.cb_proplist, NV_UNIQUE_NAME, 0) != 0)
 		nomem();
-	for (i = 1; i < ds_start; i++) {
-		if (!parseprop(props, argv[i])) {
+	for (i = 0; i < ds_start; i++) {
+		if (!parseprop(cb.cb_proplist, argv[i])) {
 			ret = -1;
 			goto error;
 		}
 	}
 
 	ret = zfs_for_each(argc - ds_start, argv + ds_start, 0,
-	    ZFS_TYPE_DATASET, NULL, NULL, 0, set_callback, props);
+	    ZFS_TYPE_DATASET, NULL, NULL, 0, set_callback, &cb);
 
 error:
-	nvlist_free(props);
+	nvlist_free(cb.cb_proplist);
 	return (ret);
 }
 
@@ -8726,6 +8741,25 @@ zfs_do_version(int argc, char **argv)
 	return (zfs_version_print() != 0);
 }
 
+/* Display documentation */
+static int
+zfs_do_help(int argc, char **argv)
+{
+	char page[MAXNAMELEN];
+	if (argc < 3 || strcmp(argv[2], "zfs") == 0)
+		strcpy(page, "zfs");
+	else if (strcmp(argv[2], "concepts") == 0 ||
+	    strcmp(argv[2], "props") == 0)
+		snprintf(page, sizeof (page), "zfs%s", argv[2]);
+	else
+		snprintf(page, sizeof (page), "zfs-%s", argv[2]);
+
+	execlp("man", "man", page, NULL);
+
+	fprintf(stderr, "couldn't run man program: %s", strerror(errno));
+	return (-1);
+}
+
 int
 main(int argc, char **argv)
 {
@@ -8781,6 +8815,12 @@ main(int argc, char **argv)
 	if ((strcmp(cmdname, "-V") == 0) || (strcmp(cmdname, "--version") == 0))
 		return (zfs_do_version(argc, argv));
 
+	/*
+	 * Special case 'help'
+	 */
+	if (strcmp(cmdname, "help") == 0)
+		return (zfs_do_help(argc, argv));
+
 	if ((g_zfs = libzfs_init()) == NULL) {
 		(void) fprintf(stderr, "%s\n", libzfs_error_init(errno));
 		return (1);

cmd/zpool/zpool_main.c

@@ -126,6 +126,8 @@ static int zpool_do_version(int, char **);
 
 static int zpool_do_wait(int, char **);
 
+static int zpool_do_help(int argc, char **argv);
+
 static zpool_compat_status_t zpool_do_load_compat(
     const char *, boolean_t *);
 
@@ -538,6 +540,10 @@ usage(boolean_t requested)
 				(void) fprintf(fp, "%s",
 				    get_usage(command_table[i].usage));
 		}
+
+		(void) fprintf(fp,
+		    gettext("\nFor further help on a command or topic, "
+		    "run: %s\n"), "zpool help [<topic>]");
 	} else {
 		(void) fprintf(fp, gettext("usage:\n"));
 		(void) fprintf(fp, "%s", get_usage(current_command->usage));
@@ -3116,12 +3122,21 @@ zfs_force_import_required(nvlist_t *config)
 	nvlist_t *nvinfo;
 
 	state = fnvlist_lookup_uint64(config, ZPOOL_CONFIG_POOL_STATE);
-	(void) nvlist_lookup_uint64(config, ZPOOL_CONFIG_HOSTID, &hostid);
+	nvinfo = fnvlist_lookup_nvlist(config, ZPOOL_CONFIG_LOAD_INFO);
+
+	/*
+	 * The hostid on LOAD_INFO comes from the MOS label via
+	 * spa_tryimport(). If its not there then we're likely talking to an
+	 * older kernel, so use the top one, which will be from the label
+	 * discovered in zpool_find_import(), or if a cachefile is in use, the
+	 * local hostid.
+	 */
+	if (nvlist_lookup_uint64(nvinfo, ZPOOL_CONFIG_HOSTID, &hostid) != 0)
+		nvlist_lookup_uint64(config, ZPOOL_CONFIG_HOSTID, &hostid);
 
 	if (state != POOL_STATE_EXPORTED && hostid != get_system_hostid())
 		return (B_TRUE);
 
-	nvinfo = fnvlist_lookup_nvlist(config, ZPOOL_CONFIG_LOAD_INFO);
 	if (nvlist_exists(nvinfo, ZPOOL_CONFIG_MMP_STATE)) {
 		mmp_state_t mmp_state = fnvlist_lookup_uint64(nvinfo,
 		    ZPOOL_CONFIG_MMP_STATE);
@@ -3143,6 +3158,7 @@ do_import(nvlist_t *config, const char *newname, const char *mntopts,
     nvlist_t *props, int flags)
 {
 	int ret = 0;
+	int ms_status = 0;
 	zpool_handle_t *zhp;
 	const char *name;
 	uint64_t version;
@@ -3191,7 +3207,10 @@ do_import(nvlist_t *config, const char *newname, const char *mntopts,
 			time_t timestamp = 0;
 			uint64_t hostid = 0;
 
-			if (nvlist_exists(config, ZPOOL_CONFIG_HOSTNAME))
+			if (nvlist_exists(nvinfo, ZPOOL_CONFIG_HOSTNAME))
+				hostname = fnvlist_lookup_string(nvinfo,
+				    ZPOOL_CONFIG_HOSTNAME);
+			else if (nvlist_exists(config, ZPOOL_CONFIG_HOSTNAME))
 				hostname = fnvlist_lookup_string(config,
 				    ZPOOL_CONFIG_HOSTNAME);
 
@@ -3199,7 +3218,10 @@ do_import(nvlist_t *config, const char *newname, const char *mntopts,
 				timestamp = fnvlist_lookup_uint64(config,
 				    ZPOOL_CONFIG_TIMESTAMP);
 
-			if (nvlist_exists(config, ZPOOL_CONFIG_HOSTID))
+			if (nvlist_exists(nvinfo, ZPOOL_CONFIG_HOSTID))
+				hostid = fnvlist_lookup_uint64(nvinfo,
+				    ZPOOL_CONFIG_HOSTID);
+			else if (nvlist_exists(config, ZPOOL_CONFIG_HOSTID))
 				hostid = fnvlist_lookup_uint64(config,
 				    ZPOOL_CONFIG_HOSTID);
 
@@ -3232,10 +3254,15 @@ do_import(nvlist_t *config, const char *newname, const char *mntopts,
 			ret = 1;
 
 	if (zpool_get_state(zhp) != POOL_STATE_UNAVAIL &&
-	    !(flags & ZFS_IMPORT_ONLY) &&
-	    zpool_enable_datasets(zhp, mntopts, 0) != 0) {
-		zpool_close(zhp);
-		return (1);
+	    !(flags & ZFS_IMPORT_ONLY)) {
+		ms_status = zpool_enable_datasets(zhp, mntopts, 0);
+		if (ms_status == EZFS_SHAREFAILED) {
+			(void) fprintf(stderr, gettext("Import was "
+			    "successful, but unable to share some datasets"));
+		} else if (ms_status == EZFS_MOUNTFAILED) {
+			(void) fprintf(stderr, gettext("Import was "
+			    "successful, but unable to mount some datasets"));
+		}
 	}
 
 	zpool_close(zhp);
@@ -6755,6 +6782,7 @@ zpool_do_split(int argc, char **argv)
 	char *mntopts = NULL;
 	splitflags_t flags;
 	int c, ret = 0;
+	int ms_status = 0;
 	boolean_t loadkeys = B_FALSE;
 	zpool_handle_t *zhp;
 	nvlist_t *config, *props = NULL;
@@ -6891,14 +6919,19 @@ zpool_do_split(int argc, char **argv)
 			ret = 1;
 	}
 
-	if (zpool_get_state(zhp) != POOL_STATE_UNAVAIL &&
-	    zpool_enable_datasets(zhp, mntopts, 0) != 0) {
-		ret = 1;
-		(void) fprintf(stderr, gettext("Split was successful, but "
-		    "the datasets could not all be mounted\n"));
+	if (zpool_get_state(zhp) != POOL_STATE_UNAVAIL) {
+		ms_status = zpool_enable_datasets(zhp, mntopts, 0);
+		if (ms_status == EZFS_SHAREFAILED) {
+			(void) fprintf(stderr, gettext("Split was successful, "
+			    "datasets are mounted but sharing of some datasets "
+			    "has failed\n"));
+		} else if (ms_status == EZFS_MOUNTFAILED) {
+			(void) fprintf(stderr, gettext("Split was successful"
+			    ", but some datasets could not be mounted\n"));
 			(void) fprintf(stderr, gettext("Try doing '%s' with a "
 			    "different altroot\n"), "zpool import");
 		}
+	}
 	zpool_close(zhp);
 	nvlist_free(config);
 	nvlist_free(props);
@@ -11039,6 +11072,25 @@ zpool_do_version(int argc, char **argv)
 	return (zfs_version_print() != 0);
 }
 
+/* Display documentation */
+static int
+zpool_do_help(int argc, char **argv)
+{
+	char page[MAXNAMELEN];
+	if (argc < 3 || strcmp(argv[2], "zpool") == 0)
+		strcpy(page, "zpool");
+	else if (strcmp(argv[2], "concepts") == 0 ||
+	    strcmp(argv[2], "props") == 0)
+		snprintf(page, sizeof (page), "zpool%s", argv[2]);
+	else
+		snprintf(page, sizeof (page), "zpool-%s", argv[2]);
+
+	execlp("man", "man", page, NULL);
+
+	fprintf(stderr, "couldn't run man program: %s", strerror(errno));
+	return (-1);
+}
+
 /*
  * Do zpool_load_compat() and print error message on failure
  */
@@ -11106,6 +11158,12 @@ main(int argc, char **argv)
 	if ((strcmp(cmdname, "-V") == 0) || (strcmp(cmdname, "--version") == 0))
 		return (zpool_do_version(argc, argv));
 
+	/*
+	 * Special case 'help'
+	 */
+	if (strcmp(cmdname, "help") == 0)
+		return (zpool_do_help(argc, argv));
+
 	if ((g_zfs = libzfs_init()) == NULL) {
 		(void) fprintf(stderr, "%s\n", libzfs_error_init(errno));
 		return (1);

cmd/ztest.c

@@ -2412,7 +2412,6 @@ ztest_get_data(void *arg, uint64_t arg2, lr_write_t *lr, char *buf,
 	int error;
 
 	ASSERT3P(lwb, !=, NULL);
-	ASSERT3P(zio, !=, NULL);
 	ASSERT3U(size, !=, 0);
 
 	ztest_object_lock(zd, object, RL_READER);
@@ -2446,6 +2445,7 @@ ztest_get_data(void *arg, uint64_t arg2, lr_write_t *lr, char *buf,
 		    DMU_READ_NO_PREFETCH);
 		ASSERT0(error);
 	} else {
+		ASSERT3P(zio, !=, NULL);
 		size = doi.doi_data_block_size;
 		if (ISP2(size)) {
 			offset = P2ALIGN(offset, size);
@@ -2457,8 +2457,7 @@ ztest_get_data(void *arg, uint64_t arg2, lr_write_t *lr, char *buf,
 		zgd->zgd_lr = (struct zfs_locked_range *)ztest_range_lock(zd,
 		    object, offset, size, RL_READER);
 
-		error = dmu_buf_hold(os, object, offset, zgd, &db,
-		    DMU_READ_NO_PREFETCH);
+		error = dmu_buf_hold_noread(os, object, offset, zgd, &db);
 
 		if (error == 0) {
 			blkptr_t *bp = &lr->lr_blkptr;
@@ -3767,7 +3766,7 @@ ztest_vdev_attach_detach(ztest_ds_t *zd, uint64_t id)
 	else if (ashift > oldvd->vdev_top->vdev_ashift)
 		expected_error = EDOM;
 	else if (newvd_is_dspare && pvd != vdev_draid_spare_get_parent(newvd))
-		expected_error = ENOTSUP;
+		expected_error = EINVAL;
 	else
 		expected_error = 0;
 
@@ -6379,6 +6378,7 @@ ztest_reguid(ztest_ds_t *zd, uint64_t id)
 	spa_t *spa = ztest_spa;
 	uint64_t orig, load;
 	int error;
+	ztest_shared_t *zs = ztest_shared;
 
 	if (ztest_opts.zo_mmp_test)
 		return;
@@ -6388,6 +6388,7 @@ ztest_reguid(ztest_ds_t *zd, uint64_t id)
 
 	(void) pthread_rwlock_wrlock(&ztest_name_lock);
 	error = spa_change_guid(spa);
+	zs->zs_guid = spa_guid(spa);
 	(void) pthread_rwlock_unlock(&ztest_name_lock);
 
 	if (error != 0)
@@ -6917,7 +6918,7 @@ ztest_trim(ztest_ds_t *zd, uint64_t id)
  * Verify pool integrity by running zdb.
  */
 static void
-ztest_run_zdb(const char *pool)
+ztest_run_zdb(uint64_t guid)
 {
 	int status;
 	char *bin;
@@ -6941,13 +6942,13 @@ ztest_run_zdb(const char *pool)
 	free(set_gvars_args);
 
 	size_t would = snprintf(zdb, len,
-	    "%s -bcc%s%s -G -d -Y -e -y %s -p %s %s",
+	    "%s -bcc%s%s -G -d -Y -e -y %s -p %s %"PRIu64,
 	    bin,
 	    ztest_opts.zo_verbose >= 3 ? "s" : "",
 	    ztest_opts.zo_verbose >= 4 ? "v" : "",
 	    set_gvars_args_joined,
 	    ztest_opts.zo_dir,
-	    pool);
+	    guid);
 	ASSERT3U(would, <, len);
 
 	umem_free(set_gvars_args_joined, strlen(set_gvars_args_joined) + 1);
@@ -7525,14 +7526,15 @@ ztest_import(ztest_shared_t *zs)
 	VERIFY0(spa_open(ztest_opts.zo_pool, &spa, FTAG));
 	zs->zs_metaslab_sz =
 	    1ULL << spa->spa_root_vdev->vdev_child[0]->vdev_ms_shift;
+	zs->zs_guid = spa_guid(spa);
 	spa_close(spa, FTAG);
 
 	kernel_fini();
 
 	if (!ztest_opts.zo_mmp_test) {
-		ztest_run_zdb(ztest_opts.zo_pool);
+		ztest_run_zdb(zs->zs_guid);
 		ztest_freeze();
-		ztest_run_zdb(ztest_opts.zo_pool);
+		ztest_run_zdb(zs->zs_guid);
 	}
 
 	(void) pthread_rwlock_destroy(&ztest_name_lock);
@@ -7603,7 +7605,6 @@ ztest_run(ztest_shared_t *zs)
 	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, B_TRUE, FTAG);
 
 	/*
@@ -7874,14 +7875,15 @@ ztest_init(ztest_shared_t *zs)
 	VERIFY0(spa_open(ztest_opts.zo_pool, &spa, FTAG));
 	zs->zs_metaslab_sz =
 	    1ULL << spa->spa_root_vdev->vdev_child[0]->vdev_ms_shift;
+	zs->zs_guid = spa_guid(spa);
 	spa_close(spa, FTAG);
 
 	kernel_fini();
 
 	if (!ztest_opts.zo_mmp_test) {
-		ztest_run_zdb(ztest_opts.zo_pool);
+		ztest_run_zdb(zs->zs_guid);
 		ztest_freeze();
-		ztest_run_zdb(ztest_opts.zo_pool);
+		ztest_run_zdb(zs->zs_guid);
 	}
 
 	(void) pthread_rwlock_destroy(&ztest_name_lock);
@@ -8304,7 +8306,7 @@ main(int argc, char **argv)
 		}
 
 		if (!ztest_opts.zo_mmp_test)
-			ztest_run_zdb(ztest_opts.zo_pool);
+			ztest_run_zdb(zs->zs_guid);
 	}
 
 	if (ztest_opts.zo_verbose >= 1) {

config/Shellcheck.am

@@ -4,6 +4,7 @@
 # Not following: a was not specified as input (see shellcheck -x). [SC1091]
 # Prefer putting braces around variable references even when not strictly required. [SC2250]
 # Consider invoking this command separately to avoid masking its return value (or use '|| true' to ignore). [SC2312]
+# Command appears to be unreachable. Check usage (or ignore if invoked indirectly). [SC2317]
 # In POSIX sh, 'local' is undefined. [SC2039] # older ShellCheck versions
 # In POSIX sh, 'local' is undefined. [SC3043] # newer ShellCheck versions
 
@@ -18,7 +19,7 @@ PHONY += shellcheck
 _STGT = $(subst ^,/,$(subst shellcheck-here-,,$@))
 shellcheck-here-%:
 if HAVE_SHELLCHECK
-	shellcheck --format=gcc --enable=all --exclude=SC1090,SC1091,SC2039,SC2250,SC2312,SC3043 $$([ -n "$(SHELLCHECK_SHELL)" ] && echo "--shell=$(SHELLCHECK_SHELL)") "$$([ -e "$(_STGT)" ] || echo "$(srcdir)/")$(_STGT)"
+	shellcheck --format=gcc --enable=all --exclude=SC1090,SC1091,SC2039,SC2250,SC2312,SC2317,SC3043 $$([ -n "$(SHELLCHECK_SHELL)" ] && echo "--shell=$(SHELLCHECK_SHELL)") "$$([ -e "$(_STGT)" ] || echo "$(srcdir)/")$(_STGT)"
 else
 	@echo "skipping shellcheck of" $(_STGT) "because shellcheck is not installed"
 endif

config/kernel-blkdev.m4

@@ -16,14 +16,65 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_BLKDEV_GET_BY_PATH], [
 	])
 ])
 
+dnl #
+dnl # 6.5.x API change,
+dnl # blkdev_get_by_path() takes 4 args
+dnl #
+AC_DEFUN([ZFS_AC_KERNEL_SRC_BLKDEV_GET_BY_PATH_4ARG], [
+	ZFS_LINUX_TEST_SRC([blkdev_get_by_path_4arg], [
+		#include <linux/fs.h>
+		#include <linux/blkdev.h>
+	], [
+		struct block_device *bdev __attribute__ ((unused)) = NULL;
+		const char *path = "path";
+		fmode_t mode = 0;
+		void *holder = NULL;
+		struct blk_holder_ops h;
+
+		bdev = blkdev_get_by_path(path, mode, holder, &h);
+	])
+])
+
 AC_DEFUN([ZFS_AC_KERNEL_BLKDEV_GET_BY_PATH], [
-	AC_MSG_CHECKING([whether blkdev_get_by_path() exists])
+	AC_MSG_CHECKING([whether blkdev_get_by_path() exists and takes 3 args])
 	ZFS_LINUX_TEST_RESULT([blkdev_get_by_path], [
 		AC_MSG_RESULT(yes)
+	], [
+		AC_MSG_RESULT(no)
+		AC_MSG_CHECKING([whether blkdev_get_by_path() exists and takes 4 args])
+		ZFS_LINUX_TEST_RESULT([blkdev_get_by_path_4arg], [
+			AC_DEFINE(HAVE_BLKDEV_GET_BY_PATH_4ARG, 1,
+				[blkdev_get_by_path() exists and takes 4 args])
+			AC_MSG_RESULT(yes)
 		], [
 			ZFS_LINUX_TEST_ERROR([blkdev_get_by_path()])
 		])
 	])
+])
+
+dnl #
+dnl # 6.5.x API change
+dnl # blk_mode_t was added as a type to supercede some places where fmode_t
+dnl # is used
+dnl #
+AC_DEFUN([ZFS_AC_KERNEL_SRC_BLKDEV_BLK_MODE_T], [
+	ZFS_LINUX_TEST_SRC([blk_mode_t], [
+		#include <linux/fs.h>
+		#include <linux/blkdev.h>
+	], [
+		blk_mode_t m __attribute((unused)) = (blk_mode_t)0;
+	])
+])
+
+AC_DEFUN([ZFS_AC_KERNEL_BLKDEV_BLK_MODE_T], [
+	AC_MSG_CHECKING([whether blk_mode_t is defined])
+	ZFS_LINUX_TEST_RESULT([blk_mode_t], [
+		AC_MSG_RESULT(yes)
+		AC_DEFINE(HAVE_BLK_MODE_T, 1, [blk_mode_t is defined])
+	], [
+		AC_MSG_RESULT(no)
+	])
+])
 
 dnl #
 dnl # 2.6.38 API change,
@@ -41,14 +92,37 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_BLKDEV_PUT], [
 	])
 ])
 
+dnl #
+dnl # 6.5.x API change.
+dnl # blkdev_put() takes (void* holder) as arg 2
+dnl #
+AC_DEFUN([ZFS_AC_KERNEL_SRC_BLKDEV_PUT_HOLDER], [
+	ZFS_LINUX_TEST_SRC([blkdev_put_holder], [
+		#include <linux/fs.h>
+		#include <linux/blkdev.h>
+	], [
+		struct block_device *bdev = NULL;
+		void *holder = NULL;
+
+		blkdev_put(bdev, holder);
+	])
+])
+
 AC_DEFUN([ZFS_AC_KERNEL_BLKDEV_PUT], [
 	AC_MSG_CHECKING([whether blkdev_put() exists])
 	ZFS_LINUX_TEST_RESULT([blkdev_put], [
 		AC_MSG_RESULT(yes)
+	], [
+		AC_MSG_CHECKING([whether blkdev_put() accepts void* as arg 2])
+		ZFS_LINUX_TEST_RESULT([blkdev_put_holder], [
+			AC_MSG_RESULT(yes)
+			AC_DEFINE(HAVE_BLKDEV_PUT_HOLDER, 1,
+				[blkdev_put() accepts void* as arg 2])
 		], [
 			ZFS_LINUX_TEST_ERROR([blkdev_put()])
 		])
 	])
+])
 
 dnl #
 dnl # 4.1 API, exported blkdev_reread_part() symbol, back ported to the
@@ -103,6 +177,33 @@ AC_DEFUN([ZFS_AC_KERNEL_BLKDEV_CHECK_DISK_CHANGE], [
 	])
 ])
 
+dnl #
+dnl # 6.5.x API change
+dnl # disk_check_media_change() was added
+dnl #
+AC_DEFUN([ZFS_AC_KERNEL_SRC_BLKDEV_DISK_CHECK_MEDIA_CHANGE], [
+	ZFS_LINUX_TEST_SRC([disk_check_media_change], [
+		#include <linux/fs.h>
+		#include <linux/blkdev.h>
+	], [
+		struct block_device *bdev = NULL;
+		bool error;
+
+		error = disk_check_media_change(bdev->bd_disk);
+	])
+])
+
+AC_DEFUN([ZFS_AC_KERNEL_BLKDEV_DISK_CHECK_MEDIA_CHANGE], [
+	AC_MSG_CHECKING([whether disk_check_media_change() exists])
+	ZFS_LINUX_TEST_RESULT([disk_check_media_change], [
+		AC_MSG_RESULT(yes)
+		AC_DEFINE(HAVE_DISK_CHECK_MEDIA_CHANGE, 1,
+		    [disk_check_media_change() exists])
+	], [
+		AC_MSG_RESULT(no)
+	])
+])
+
 dnl #
 dnl # bdev_kobj() is introduced from 5.12
 dnl #
@@ -443,9 +544,34 @@ AC_DEFUN([ZFS_AC_KERNEL_BLKDEV_GET_ERESTARTSYS], [
 	])
 ])
 
+dnl #
+dnl # 6.5.x API change
+dnl # BLK_STS_NEXUS replaced with BLK_STS_RESV_CONFLICT
+dnl #
+AC_DEFUN([ZFS_AC_KERNEL_SRC_BLKDEV_BLK_STS_RESV_CONFLICT], [
+	ZFS_LINUX_TEST_SRC([blk_sts_resv_conflict], [
+		#include <linux/blkdev.h>
+	],[
+		blk_status_t s __attribute__ ((unused)) = BLK_STS_RESV_CONFLICT;
+	])
+])
+
+AC_DEFUN([ZFS_AC_KERNEL_BLKDEV_BLK_STS_RESV_CONFLICT], [
+	AC_MSG_CHECKING([whether BLK_STS_RESV_CONFLICT is defined])
+		ZFS_LINUX_TEST_RESULT([blk_sts_resv_conflict], [
+			AC_DEFINE(HAVE_BLK_STS_RESV_CONFLICT, 1, [BLK_STS_RESV_CONFLICT is defined])
+			AC_MSG_RESULT(yes)
+		], [
+			AC_MSG_RESULT(no)
+		])
+	])
+])
+
 AC_DEFUN([ZFS_AC_KERNEL_SRC_BLKDEV], [
 	ZFS_AC_KERNEL_SRC_BLKDEV_GET_BY_PATH
+	ZFS_AC_KERNEL_SRC_BLKDEV_GET_BY_PATH_4ARG
 	ZFS_AC_KERNEL_SRC_BLKDEV_PUT
+	ZFS_AC_KERNEL_SRC_BLKDEV_PUT_HOLDER
 	ZFS_AC_KERNEL_SRC_BLKDEV_REREAD_PART
 	ZFS_AC_KERNEL_SRC_BLKDEV_INVALIDATE_BDEV
 	ZFS_AC_KERNEL_SRC_BLKDEV_LOOKUP_BDEV
@@ -458,6 +584,9 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_BLKDEV], [
 	ZFS_AC_KERNEL_SRC_BLKDEV_ISSUE_SECURE_ERASE
 	ZFS_AC_KERNEL_SRC_BLKDEV_BDEV_KOBJ
 	ZFS_AC_KERNEL_SRC_BLKDEV_PART_TO_DEV
+	ZFS_AC_KERNEL_SRC_BLKDEV_DISK_CHECK_MEDIA_CHANGE
+	ZFS_AC_KERNEL_SRC_BLKDEV_BLK_STS_RESV_CONFLICT
+	ZFS_AC_KERNEL_SRC_BLKDEV_BLK_MODE_T
 ])
 
 AC_DEFUN([ZFS_AC_KERNEL_BLKDEV], [
@@ -476,4 +605,7 @@ AC_DEFUN([ZFS_AC_KERNEL_BLKDEV], [
 	ZFS_AC_KERNEL_BLKDEV_ISSUE_SECURE_ERASE
 	ZFS_AC_KERNEL_BLKDEV_BDEV_KOBJ
 	ZFS_AC_KERNEL_BLKDEV_PART_TO_DEV
+	ZFS_AC_KERNEL_BLKDEV_DISK_CHECK_MEDIA_CHANGE
+	ZFS_AC_KERNEL_BLKDEV_BLK_STS_RESV_CONFLICT
+	ZFS_AC_KERNEL_BLKDEV_BLK_MODE_T
 ])

config/kernel-block-device-operations.m4

@@ -49,14 +49,44 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_BLOCK_DEVICE_OPERATIONS_RELEASE_VOID], [
 	], [], [])
 ])
 
+dnl #
+dnl # 5.9.x API change
+dnl #
+AC_DEFUN([ZFS_AC_KERNEL_SRC_BLOCK_DEVICE_OPERATIONS_RELEASE_1ARG], [
+	ZFS_LINUX_TEST_SRC([block_device_operations_release_void_1arg], [
+		#include <linux/blkdev.h>
+
+		void blk_release(struct gendisk *g) {
+			(void) g;
+			return;
+		}
+
+		static const struct block_device_operations
+		    bops __attribute__ ((unused)) = {
+			.open		= NULL,
+			.release	= blk_release,
+			.ioctl		= NULL,
+			.compat_ioctl	= NULL,
+		};
+	], [], [])
+])
+
 AC_DEFUN([ZFS_AC_KERNEL_BLOCK_DEVICE_OPERATIONS_RELEASE_VOID], [
-	AC_MSG_CHECKING([whether bops->release() is void])
+	AC_MSG_CHECKING([whether bops->release() is void and takes 2 args])
 	ZFS_LINUX_TEST_RESULT([block_device_operations_release_void], [
 		AC_MSG_RESULT(yes)
+	],[
+		AC_MSG_RESULT(no)
+		AC_MSG_CHECKING([whether bops->release() is void and takes 1 arg])
+		ZFS_LINUX_TEST_RESULT([block_device_operations_release_void_1arg], [
+			AC_MSG_RESULT(yes)
+			AC_DEFINE([HAVE_BLOCK_DEVICE_OPERATIONS_RELEASE_1ARG], [1],
+				[Define if release() in block_device_operations takes 1 arg])
 		],[
 			ZFS_LINUX_TEST_ERROR([bops->release()])
 		])
 	])
+])
 
 dnl #
 dnl # 5.13 API change
@@ -92,6 +122,7 @@ AC_DEFUN([ZFS_AC_KERNEL_BLOCK_DEVICE_OPERATIONS_REVALIDATE_DISK], [
 AC_DEFUN([ZFS_AC_KERNEL_SRC_BLOCK_DEVICE_OPERATIONS], [
 	ZFS_AC_KERNEL_SRC_BLOCK_DEVICE_OPERATIONS_CHECK_EVENTS
 	ZFS_AC_KERNEL_SRC_BLOCK_DEVICE_OPERATIONS_RELEASE_VOID
+	ZFS_AC_KERNEL_SRC_BLOCK_DEVICE_OPERATIONS_RELEASE_1ARG
 	ZFS_AC_KERNEL_SRC_BLOCK_DEVICE_OPERATIONS_REVALIDATE_DISK
 ])
 

config/kernel-filemap-splice-read.m4

@@ -0,0 +1,25 @@
+AC_DEFUN([ZFS_AC_KERNEL_SRC_COPY_SPLICE_READ], [
+	dnl #
+	dnl # Kernel 6.5 - generic_file_splice_read was removed in favor
+	dnl # of copy_splice_read for the .splice_read member of the
+	dnl # file_operations struct.
+	dnl #
+	ZFS_LINUX_TEST_SRC([has_copy_splice_read], [
+		#include <linux/fs.h>
+
+		struct file_operations fops __attribute__((unused)) = {
+			.splice_read = copy_splice_read,
+		};
+	],[])
+])
+
+AC_DEFUN([ZFS_AC_KERNEL_COPY_SPLICE_READ], [
+	AC_MSG_CHECKING([whether copy_splice_read() exists])
+	ZFS_LINUX_TEST_RESULT([has_copy_splice_read], [
+		AC_MSG_RESULT(yes)
+		AC_DEFINE(HAVE_COPY_SPLICE_READ, 1,
+		    [copy_splice_read exists])
+	],[
+		AC_MSG_RESULT(no)
+	])
+])

config/kernel-register_sysctl_table.m4

@@ -0,0 +1,27 @@
+dnl #
+dnl # Linux 6.5 removes register_sysctl_table
+dnl #
+AC_DEFUN([ZFS_AC_KERNEL_SRC_REGISTER_SYSCTL_TABLE], [
+	ZFS_LINUX_TEST_SRC([has_register_sysctl_table], [
+		#include <linux/sysctl.h>
+
+		static struct ctl_table dummy_table[] = {
+			{}
+		};
+
+    ],[
+		struct ctl_table_header *h
+			__attribute((unused)) = register_sysctl_table(dummy_table);
+    ])
+])
+
+AC_DEFUN([ZFS_AC_KERNEL_REGISTER_SYSCTL_TABLE], [
+	AC_MSG_CHECKING([whether register_sysctl_table exists])
+	ZFS_LINUX_TEST_RESULT([has_register_sysctl_table], [
+		AC_MSG_RESULT([yes])
+		AC_DEFINE(HAVE_REGISTER_SYSCTL_TABLE, 1,
+			[register_sysctl_table exists])
+	],[
+		AC_MSG_RESULT([no])
+	])
+])

config/kernel-vfs-extended-file_range.m4

@@ -0,0 +1,50 @@
+dnl #
+dnl # EL7 have backported copy_file_range and clone_file_range and
+dnl # added them to an "extended" file_operations struct.
+dnl #
+dnl # We're testing for both functions in one here, because they will only
+dnl # ever appear together and we don't want to match a similar method in
+dnl # some future vendor kernel.
+dnl #
+AC_DEFUN([ZFS_AC_KERNEL_SRC_VFS_FILE_OPERATIONS_EXTEND], [
+	ZFS_LINUX_TEST_SRC([vfs_file_operations_extend], [
+		#include <linux/fs.h>
+
+		static ssize_t test_copy_file_range(struct file *src_file,
+		    loff_t src_off, struct file *dst_file, loff_t dst_off,
+		    size_t len, unsigned int flags) {
+			(void) src_file; (void) src_off;
+			(void) dst_file; (void) dst_off;
+			(void) len; (void) flags;
+			return (0);
+		}
+
+		static int test_clone_file_range(struct file *src_file,
+		    loff_t src_off, struct file *dst_file, loff_t dst_off,
+		    u64 len) {
+			(void) src_file; (void) src_off;
+			(void) dst_file; (void) dst_off;
+			(void) len;
+			return (0);
+		}
+
+		static const struct file_operations_extend
+		    fops __attribute__ ((unused)) = {
+			.kabi_fops = {},
+			.copy_file_range = test_copy_file_range,
+			.clone_file_range = test_clone_file_range,
+		};
+	],[])
+])
+AC_DEFUN([ZFS_AC_KERNEL_VFS_FILE_OPERATIONS_EXTEND], [
+	AC_MSG_CHECKING([whether file_operations_extend takes \
+.copy_file_range() and .clone_file_range()])
+	ZFS_LINUX_TEST_RESULT([vfs_file_operations_extend], [
+		AC_MSG_RESULT([yes])
+		AC_DEFINE(HAVE_VFS_FILE_OPERATIONS_EXTEND, 1,
+		    [file_operations_extend takes .copy_file_range()
+		    and .clone_file_range()])
+	],[
+		AC_MSG_RESULT([no])
+	])
+])

config/kernel-vfs-file_range.m4

@@ -0,0 +1,164 @@
+dnl #
+dnl # The *_file_range APIs have a long history:
+dnl #
+dnl # 2.6.29: BTRFS_IOC_CLONE and BTRFS_IOC_CLONE_RANGE ioctl introduced
+dnl # 3.12: BTRFS_IOC_FILE_EXTENT_SAME ioctl introduced
+dnl #
+dnl # 4.5: copy_file_range() syscall introduced, added to VFS
+dnl # 4.5: BTRFS_IOC_CLONE and BTRFS_IOC_CLONE_RANGE renamed to FICLONE ands
+dnl #      FICLONERANGE, added to VFS as clone_file_range()
+dnl # 4.5: BTRFS_IOC_FILE_EXTENT_SAME renamed to FIDEDUPERANGE, added to VFS
+dnl #      as dedupe_file_range()
+dnl #
+dnl # 4.20: VFS clone_file_range() and dedupe_file_range() replaced by
+dnl #       remap_file_range()
+dnl #
+dnl # 5.3: VFS copy_file_range() expected to do its own fallback,
+dnl #      generic_copy_file_range() added to support it
+dnl #
+AC_DEFUN([ZFS_AC_KERNEL_SRC_VFS_COPY_FILE_RANGE], [
+	ZFS_LINUX_TEST_SRC([vfs_copy_file_range], [
+		#include <linux/fs.h>
+
+		static ssize_t test_copy_file_range(struct file *src_file,
+		    loff_t src_off, struct file *dst_file, loff_t dst_off,
+		    size_t len, unsigned int flags) {
+			(void) src_file; (void) src_off;
+			(void) dst_file; (void) dst_off;
+			(void) len; (void) flags;
+			return (0);
+		}
+
+		static const struct file_operations
+		    fops __attribute__ ((unused)) = {
+			.copy_file_range	= test_copy_file_range,
+		};
+	],[])
+])
+AC_DEFUN([ZFS_AC_KERNEL_VFS_COPY_FILE_RANGE], [
+	AC_MSG_CHECKING([whether fops->copy_file_range() is available])
+	ZFS_LINUX_TEST_RESULT([vfs_copy_file_range], [
+		AC_MSG_RESULT([yes])
+		AC_DEFINE(HAVE_VFS_COPY_FILE_RANGE, 1,
+		    [fops->copy_file_range() is available])
+	],[
+		AC_MSG_RESULT([no])
+	])
+])
+
+AC_DEFUN([ZFS_AC_KERNEL_SRC_VFS_GENERIC_COPY_FILE_RANGE], [
+	ZFS_LINUX_TEST_SRC([generic_copy_file_range], [
+		#include <linux/fs.h>
+	], [
+		struct file *src_file __attribute__ ((unused)) = NULL;
+		loff_t src_off __attribute__ ((unused)) = 0;
+		struct file *dst_file __attribute__ ((unused)) = NULL;
+		loff_t dst_off __attribute__ ((unused)) = 0;
+		size_t len __attribute__ ((unused)) = 0;
+		unsigned int flags __attribute__ ((unused)) = 0;
+		generic_copy_file_range(src_file, src_off, dst_file, dst_off,
+		    len, flags);
+	])
+])
+AC_DEFUN([ZFS_AC_KERNEL_VFS_GENERIC_COPY_FILE_RANGE], [
+	AC_MSG_CHECKING([whether generic_copy_file_range() is available])
+	ZFS_LINUX_TEST_RESULT_SYMBOL([generic_copy_file_range],
+	[generic_copy_file_range], [fs/read_write.c], [
+		AC_MSG_RESULT(yes)
+		AC_DEFINE(HAVE_VFS_GENERIC_COPY_FILE_RANGE, 1,
+		    [generic_copy_file_range() is available])
+	],[
+		AC_MSG_RESULT(no)
+	])
+])
+
+AC_DEFUN([ZFS_AC_KERNEL_SRC_VFS_CLONE_FILE_RANGE], [
+	ZFS_LINUX_TEST_SRC([vfs_clone_file_range], [
+		#include <linux/fs.h>
+
+		static int test_clone_file_range(struct file *src_file,
+		    loff_t src_off, struct file *dst_file, loff_t dst_off,
+		    u64 len) {
+			(void) src_file; (void) src_off;
+			(void) dst_file; (void) dst_off;
+			(void) len;
+			return (0);
+		}
+
+		static const struct file_operations
+		    fops __attribute__ ((unused)) = {
+			.clone_file_range	= test_clone_file_range,
+		};
+	],[])
+])
+AC_DEFUN([ZFS_AC_KERNEL_VFS_CLONE_FILE_RANGE], [
+	AC_MSG_CHECKING([whether fops->clone_file_range() is available])
+	ZFS_LINUX_TEST_RESULT([vfs_clone_file_range], [
+		AC_MSG_RESULT([yes])
+		AC_DEFINE(HAVE_VFS_CLONE_FILE_RANGE, 1,
+		    [fops->clone_file_range() is available])
+	],[
+		AC_MSG_RESULT([no])
+	])
+])
+
+AC_DEFUN([ZFS_AC_KERNEL_SRC_VFS_DEDUPE_FILE_RANGE], [
+	ZFS_LINUX_TEST_SRC([vfs_dedupe_file_range], [
+		#include <linux/fs.h>
+
+		static int test_dedupe_file_range(struct file *src_file,
+		    loff_t src_off, struct file *dst_file, loff_t dst_off,
+		    u64 len) {
+			(void) src_file; (void) src_off;
+			(void) dst_file; (void) dst_off;
+			(void) len;
+			return (0);
+		}
+
+		static const struct file_operations
+		    fops __attribute__ ((unused)) = {
+                .dedupe_file_range	= test_dedupe_file_range,
+		};
+	],[])
+])
+AC_DEFUN([ZFS_AC_KERNEL_VFS_DEDUPE_FILE_RANGE], [
+	AC_MSG_CHECKING([whether fops->dedupe_file_range() is available])
+	ZFS_LINUX_TEST_RESULT([vfs_dedupe_file_range], [
+		AC_MSG_RESULT([yes])
+		AC_DEFINE(HAVE_VFS_DEDUPE_FILE_RANGE, 1,
+		    [fops->dedupe_file_range() is available])
+	],[
+		AC_MSG_RESULT([no])
+	])
+])
+
+AC_DEFUN([ZFS_AC_KERNEL_SRC_VFS_REMAP_FILE_RANGE], [
+	ZFS_LINUX_TEST_SRC([vfs_remap_file_range], [
+		#include <linux/fs.h>
+
+		static loff_t test_remap_file_range(struct file *src_file,
+		    loff_t src_off, struct file *dst_file, loff_t dst_off,
+		    loff_t len, unsigned int flags) {
+			(void) src_file; (void) src_off;
+			(void) dst_file; (void) dst_off;
+			(void) len; (void) flags;
+			return (0);
+		}
+
+		static const struct file_operations
+		    fops __attribute__ ((unused)) = {
+			.remap_file_range	= test_remap_file_range,
+		};
+	],[])
+])
+
+AC_DEFUN([ZFS_AC_KERNEL_VFS_REMAP_FILE_RANGE], [
+	AC_MSG_CHECKING([whether fops->remap_file_range() is available])
+	ZFS_LINUX_TEST_RESULT([vfs_remap_file_range], [
+		AC_MSG_RESULT([yes])
+		AC_DEFINE(HAVE_VFS_REMAP_FILE_RANGE, 1,
+		    [fops->remap_file_range() is available])
+	],[
+		AC_MSG_RESULT([no])
+	])
+])

config/kernel-vfs-iov_iter.m4

@@ -6,8 +6,7 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_VFS_IOV_ITER], [
 		#include <linux/fs.h>
 		#include <linux/uio.h>
 	],[
-		int type __attribute__ ((unused)) =
-		    ITER_IOVEC | ITER_KVEC | ITER_BVEC | ITER_PIPE;
+		int type __attribute__ ((unused)) = ITER_KVEC;
 	])
 
 	ZFS_LINUX_TEST_SRC([iov_iter_advance], [
@@ -93,6 +92,14 @@ AC_DEFUN([ZFS_AC_KERNEL_SRC_VFS_IOV_ITER], [
 		struct iov_iter iter = { 0 };
 		__attribute__((unused)) enum iter_type i = iov_iter_type(&iter);
 	])
+
+	ZFS_LINUX_TEST_SRC([iter_iov], [
+		#include <linux/fs.h>
+		#include <linux/uio.h>
+	],[
+		struct iov_iter iter = { 0 };
+		__attribute__((unused)) const struct iovec *iov = iter_iov(&iter);
+	])
 ])
 
 AC_DEFUN([ZFS_AC_KERNEL_VFS_IOV_ITER], [
@@ -201,4 +208,19 @@ AC_DEFUN([ZFS_AC_KERNEL_VFS_IOV_ITER], [
 		AC_DEFINE(HAVE_VFS_IOV_ITER, 1,
 		    [All required iov_iter interfaces are available])
 	])
+
+	dnl #
+	dnl # Kernel 6.5 introduces the iter_iov() function that returns the
+	dnl # __iov member of an iov_iter*. The iov member was renamed to this
+	dnl # __iov member, and is intended to be accessed via the helper
+	dnl # function now.
+	dnl #
+	AC_MSG_CHECKING([whether iter_iov() is available])
+	ZFS_LINUX_TEST_RESULT([iter_iov], [
+		AC_MSG_RESULT(yes)
+		AC_DEFINE(HAVE_ITER_IOV, 1,
+		    [iter_iov() is available])
+	],[
+		AC_MSG_RESULT(no)
+	])
 ])

config/kernel.m4

@@ -116,6 +116,12 @@ AC_DEFUN([ZFS_AC_KERNEL_TEST_SRC], [
 	ZFS_AC_KERNEL_SRC_VFS_RW_ITERATE
 	ZFS_AC_KERNEL_SRC_VFS_GENERIC_WRITE_CHECKS
 	ZFS_AC_KERNEL_SRC_VFS_IOV_ITER
+	ZFS_AC_KERNEL_SRC_VFS_COPY_FILE_RANGE
+	ZFS_AC_KERNEL_SRC_VFS_GENERIC_COPY_FILE_RANGE
+	ZFS_AC_KERNEL_SRC_VFS_REMAP_FILE_RANGE
+	ZFS_AC_KERNEL_SRC_VFS_CLONE_FILE_RANGE
+	ZFS_AC_KERNEL_SRC_VFS_DEDUPE_FILE_RANGE
+	ZFS_AC_KERNEL_SRC_VFS_FILE_OPERATIONS_EXTEND
 	ZFS_AC_KERNEL_SRC_KMAP_ATOMIC_ARGS
 	ZFS_AC_KERNEL_SRC_FOLLOW_DOWN_ONE
 	ZFS_AC_KERNEL_SRC_MAKE_REQUEST_FN
@@ -154,6 +160,8 @@ AC_DEFUN([ZFS_AC_KERNEL_TEST_SRC], [
 	ZFS_AC_KERNEL_SRC_FILEMAP
 	ZFS_AC_KERNEL_SRC_WRITEPAGE_T
 	ZFS_AC_KERNEL_SRC_RECLAIMED
+	ZFS_AC_KERNEL_SRC_REGISTER_SYSCTL_TABLE
+	ZFS_AC_KERNEL_SRC_COPY_SPLICE_READ
 	case "$host_cpu" in
 		powerpc*)
 			ZFS_AC_KERNEL_SRC_CPU_HAS_FEATURE
@@ -249,6 +257,12 @@ AC_DEFUN([ZFS_AC_KERNEL_TEST_RESULT], [
 	ZFS_AC_KERNEL_VFS_RW_ITERATE
 	ZFS_AC_KERNEL_VFS_GENERIC_WRITE_CHECKS
 	ZFS_AC_KERNEL_VFS_IOV_ITER
+	ZFS_AC_KERNEL_VFS_COPY_FILE_RANGE
+	ZFS_AC_KERNEL_VFS_GENERIC_COPY_FILE_RANGE
+	ZFS_AC_KERNEL_VFS_REMAP_FILE_RANGE
+	ZFS_AC_KERNEL_VFS_CLONE_FILE_RANGE
+	ZFS_AC_KERNEL_VFS_DEDUPE_FILE_RANGE
+	ZFS_AC_KERNEL_VFS_FILE_OPERATIONS_EXTEND
 	ZFS_AC_KERNEL_KMAP_ATOMIC_ARGS
 	ZFS_AC_KERNEL_FOLLOW_DOWN_ONE
 	ZFS_AC_KERNEL_MAKE_REQUEST_FN
@@ -287,6 +301,8 @@ AC_DEFUN([ZFS_AC_KERNEL_TEST_RESULT], [
 	ZFS_AC_KERNEL_FILEMAP
 	ZFS_AC_KERNEL_WRITEPAGE_T
 	ZFS_AC_KERNEL_RECLAIMED
+	ZFS_AC_KERNEL_REGISTER_SYSCTL_TABLE
+	ZFS_AC_KERNEL_COPY_SPLICE_READ
 	case "$host_cpu" in
 		powerpc*)
 			ZFS_AC_KERNEL_CPU_HAS_FEATURE

config/zfs-build.m4

@@ -358,6 +358,9 @@ AC_DEFUN([ZFS_AC_RPM], [
 	AS_IF([test -n "$udevruledir" ], [
 		RPM_DEFINE_UTIL=${RPM_DEFINE_UTIL}' --define "_udevruledir $(udevruledir)"'
 	])
+	AS_IF([test -n "$bashcompletiondir" ], [
+		RPM_DEFINE_UTIL=${RPM_DEFINE_UTIL}' --define "_bashcompletiondir $(bashcompletiondir)"'
+	])
 	RPM_DEFINE_UTIL=${RPM_DEFINE_UTIL}' $(DEFINE_SYSTEMD)'
 	RPM_DEFINE_UTIL=${RPM_DEFINE_UTIL}' $(DEFINE_PYZFS)'
 	RPM_DEFINE_UTIL=${RPM_DEFINE_UTIL}' $(DEFINE_PAM)'
@@ -617,6 +620,17 @@ AC_DEFUN([ZFS_AC_DEFAULT_PACKAGE], [
 		AC_MSG_RESULT([no])
 	fi
 	AC_SUBST(RPM_DEFINE_INITRAMFS)
+
+	AC_MSG_CHECKING([default bash completion directory])
+	case "$VENDOR" in
+		ubuntu)     bashcompletiondir=/usr/share/bash-completion/completions   ;;
+		debian)     bashcompletiondir=/usr/share/bash-completion/completions   ;;
+		freebsd)    bashcompletiondir=$sysconfdir/bash_completion.d;;
+		*)          bashcompletiondir=/etc/bash_completion.d   ;;
+	esac
+	AC_MSG_RESULT([$bashcompletiondir])
+	AC_SUBST(bashcompletiondir)
+
 ])
 
 dnl #

contrib/bash_completion.d/Makefile.am

@@ -1,5 +1,3 @@
-bashcompletiondir = $(sysconfdir)/bash_completion.d
-
 nodist_bashcompletion_DATA  = %D%/zfs
 SUBSTFILES                 += $(nodist_bashcompletion_DATA)
 

contrib/debian/changelog

@@ -1,3 +1,9 @@
+openzfs-linux (2.2.0-0) unstable; urgency=low
+
+  * OpenZFS 2.2.0 is tagged.
+
+ -- Umer Saleem <usaleem@ixsystems.com>  Tue, 25 Jul 2023 15:00:00 +0500
+
 openzfs-linux (2.1.99-1) unstable; urgency=low
 
   * Integrate minimally modified Debian packaging from ZFS on Linux

contrib/debian/control

@@ -4,7 +4,7 @@ Priority: optional
 Maintainer: ZFS on Linux specific mailing list <zfs-discuss@list.zfsonlinux.org>
 Build-Depends: debhelper-compat (= 12),
                dh-python,
-               dkms (>> 2.1.1.2-5),
+               dh-sequence-dkms | dkms (>> 2.1.1.2-5),
                libaio-dev,
                libblkid-dev,
                libcurl4-openssl-dev,

contrib/debian/openzfs-zfs-test.install

@@ -1,10 +1,8 @@
-sbin/zinject
 sbin/ztest
 usr/bin/raidz_test
 usr/share/man/man1/raidz_test.1
 usr/share/man/man1/test-runner.1
 usr/share/man/man1/ztest.1
-usr/share/man/man8/zinject.8
 usr/share/zfs/common.sh
 usr/share/zfs/runfiles/
 usr/share/zfs/test-runner

contrib/debian/openzfs-zfsutils.install

@@ -1,7 +1,6 @@
 etc/default/zfs
 etc/zfs/zfs-functions
 etc/zfs/zpool.d/
-etc/bash_completion.d/zfs
 lib/systemd/system-generators/
 lib/systemd/system-preset/
 lib/systemd/system/zfs-import-cache.service
@@ -27,6 +26,7 @@ sbin/zfs
 sbin/zfs_ids_to_path
 sbin/zgenhostid
 sbin/zhack
+sbin/zinject
 sbin/zpool
 sbin/zstream
 sbin/zstreamdump
@@ -59,7 +59,6 @@ usr/share/man/man8/zfs-get.8
 usr/share/man/man8/zfs-groupspace.8
 usr/share/man/man8/zfs-hold.8
 usr/share/man/man8/zfs-inherit.8
-usr/share/man/man8/zfs-jail.8
 usr/share/man/man8/zfs-list.8
 usr/share/man/man8/zfs-load-key.8
 usr/share/man/man8/zfs-mount-generator.8
@@ -79,7 +78,6 @@ usr/share/man/man8/zfs-set.8
 usr/share/man/man8/zfs-share.8
 usr/share/man/man8/zfs-snapshot.8
 usr/share/man/man8/zfs-unallow.8
-usr/share/man/man8/zfs-unjail.8
 usr/share/man/man8/zfs-unload-key.8
 usr/share/man/man8/zfs-unmount.8
 usr/share/man/man8/zfs-unzone.8
@@ -92,6 +90,7 @@ usr/share/man/man8/zfs_ids_to_path.8
 usr/share/man/man7/zfsconcepts.7
 usr/share/man/man7/zfsprops.7
 usr/share/man/man8/zgenhostid.8
+usr/share/man/man8/zinject.8
 usr/share/man/man8/zpool-add.8
 usr/share/man/man8/zpool-attach.8
 usr/share/man/man8/zpool-checkpoint.8

contrib/debian/rules.in

@@ -71,10 +71,6 @@ override_dh_auto_install:
 	@# Install the utilities.
 	$(MAKE) install DESTDIR='$(CURDIR)/debian/tmp'
 
-	# Use upstream's bash completion
-	install -D -t '$(CURDIR)/debian/tmp/usr/share/bash-completion/completions/' \
-		'$(CURDIR)/contrib/bash_completion.d/zfs'
-
 	# Move from bin_dir to /usr/sbin
 	# Remove suffix (.py) as per policy 10.4 - Scripts
 	# https://www.debian.org/doc/debian-policy/ch-files.html#s-scripts
@@ -136,7 +132,6 @@ override_dh_auto_install:
 
 	chmod a-x '$(CURDIR)/debian/tmp/etc/zfs/zfs-functions'
 	chmod a-x '$(CURDIR)/debian/tmp/etc/default/zfs'
-	chmod a-x '$(CURDIR)/debian/tmp/usr/share/bash-completion/completions/zfs'
 
 override_dh_python3:
 	dh_python3 -p openzfs-python3-pyzfs

contrib/dracut/90zfs/zfs-env-bootfs.service.in

@@ -12,6 +12,7 @@ ExecStart=/bin/sh -c '
     decode_root_args || exit 0;                                                                \
     [ "$root" = "zfs:AUTO" ] && root="$(@sbindir@/zpool list -H -o bootfs | grep -m1 -vFx -)"; \
     rootflags="$(getarg rootflags=)";                                                          \
+    [ "$(@sbindir@/zfs get -H -o value mountpoint "$root")" = legacy ] ||                      \
         case ",$rootflags," in                                                                 \
             *,zfsutil,*) ;;                                                                    \
             ,,) rootflags=zfsutil ;;                                                           \

contrib/dracut/90zfs/zfs-rollback-bootfs.service.in

@@ -2,7 +2,7 @@
 Description=Rollback bootfs just before it is mounted
 Requisite=zfs-import.target
 After=zfs-import.target dracut-pre-mount.service zfs-snapshot-bootfs.service
-Before=dracut-mount.service
+Before=dracut-mount.service sysroot.mount
 DefaultDependencies=no
 ConditionKernelCommandLine=bootfs.rollback
 ConditionEnvironment=BOOTFS

include/libzfs.h

@@ -156,6 +156,7 @@ typedef enum zfs_error {
 	EZFS_NOT_USER_NAMESPACE,	/* a file is not a user namespace */
 	EZFS_CKSUM,		/* insufficient replicas */
 	EZFS_RESUME_EXISTS,	/* Resume on existing dataset without force */
+	EZFS_SHAREFAILED,	/* filesystem share failed */
 	EZFS_UNKNOWN
 } zfs_error_t;
 
@@ -522,6 +523,7 @@ _LIBZFS_H nvlist_t *zfs_valid_proplist(libzfs_handle_t *, zfs_type_t,
 _LIBZFS_H const char *zfs_prop_to_name(zfs_prop_t);
 _LIBZFS_H int zfs_prop_set(zfs_handle_t *, const char *, const char *);
 _LIBZFS_H int zfs_prop_set_list(zfs_handle_t *, nvlist_t *);
+_LIBZFS_H int zfs_prop_set_list_flags(zfs_handle_t *, nvlist_t *, int);
 _LIBZFS_H int zfs_prop_get(zfs_handle_t *, zfs_prop_t, char *, size_t,
     zprop_source_t *, char *, size_t, boolean_t);
 _LIBZFS_H int zfs_prop_get_recvd(zfs_handle_t *, const char *, char *, size_t,
@@ -644,6 +646,13 @@ typedef struct zprop_get_cbdata {
 	vdev_cbdata_t cb_vdevs;
 } zprop_get_cbdata_t;
 
+#define	ZFS_SET_NOMOUNT		1
+
+typedef struct zprop_set_cbdata {
+	int cb_flags;
+	nvlist_t *cb_proplist;
+} zprop_set_cbdata_t;
+
 _LIBZFS_H void zprop_print_one_property(const char *, zprop_get_cbdata_t *,
     const char *, const char *, zprop_source_t, const char *,
     const char *);

include/os/freebsd/spl/sys/vnode.h

@@ -36,7 +36,11 @@ struct xucred;
 typedef struct flock	flock64_t;
 typedef	struct vnode	vnode_t;
 typedef	struct vattr	vattr_t;
+#if __FreeBSD_version < 1400093
 typedef enum vtype vtype_t;
+#else
+#define	vtype_t __enum_uint8(vtype)
+#endif
 
 #include <sys/types.h>
 #include <sys/queue.h>

include/os/freebsd/zfs/sys/zfs_vfsops_os.h

@@ -93,7 +93,6 @@ struct zfsvfs {
 	zfs_teardown_lock_t z_teardown_lock;
 	zfs_teardown_inactive_lock_t z_teardown_inactive_lock;
 	list_t		z_all_znodes;	/* all vnodes in the fs */
-	uint64_t	z_nr_znodes;	/* number of znodes in the fs */
 	kmutex_t	z_znodes_lock;	/* lock for z_all_znodes */
 	struct zfsctl_root	*z_ctldir;	/* .zfs directory pointer */
 	boolean_t	z_show_ctldir;	/* expose .zfs in the root dir */

include/os/linux/kernel/linux/blkdev_compat.h

@@ -181,7 +181,11 @@ bi_status_to_errno(blk_status_t status)
 		return (ENOLINK);
 	case BLK_STS_TARGET:
 		return (EREMOTEIO);
+#ifdef HAVE_BLK_STS_RESV_CONFLICT
+	case BLK_STS_RESV_CONFLICT:
+#else
 	case BLK_STS_NEXUS:
+#endif
 		return (EBADE);
 	case BLK_STS_MEDIUM:
 		return (ENODATA);
@@ -215,7 +219,11 @@ errno_to_bi_status(int error)
 	case EREMOTEIO:
 		return (BLK_STS_TARGET);
 	case EBADE:
+#ifdef HAVE_BLK_STS_RESV_CONFLICT
+		return (BLK_STS_RESV_CONFLICT);
+#else
 		return (BLK_STS_NEXUS);
+#endif
 	case ENODATA:
 		return (BLK_STS_MEDIUM);
 	case EILSEQ:
@@ -337,6 +345,9 @@ zfs_check_media_change(struct block_device *bdev)
 	return (0);
 }
 #define	vdev_bdev_reread_part(bdev)	zfs_check_media_change(bdev)
+#elif defined(HAVE_DISK_CHECK_MEDIA_CHANGE)
+#define	vdev_bdev_reread_part(bdev)	disk_check_media_change(bdev->bd_disk)
+#define	zfs_check_media_change(bdev)	disk_check_media_change(bdev->bd_disk)
 #else
 /*
  * This is encountered if check_disk_change() and bdev_check_media_change()
@@ -387,6 +398,12 @@ vdev_lookup_bdev(const char *path, dev_t *dev)
 #endif
 }
 
+#if defined(HAVE_BLK_MODE_T)
+#define	blk_mode_is_open_write(flag)	((flag) & BLK_OPEN_WRITE)
+#else
+#define	blk_mode_is_open_write(flag)	((flag) & FMODE_WRITE)
+#endif
+
 /*
  * Kernels without bio_set_op_attrs use bi_rw for the bio flags.
  */

include/os/linux/kernel/linux/simd_x86.h

@@ -147,6 +147,15 @@
 #error "Toolchain needs to support the XSAVE assembler instruction"
 #endif
 
+#ifndef XFEATURE_MASK_XTILE
+/*
+ * For kernels where this doesn't exist yet, we still don't want to break
+ * by save/restoring this broken nonsense.
+ * See issue #14989 or Intel errata SPR4 for why
+ */
+#define	XFEATURE_MASK_XTILE	0x60000
+#endif
+
 #include <linux/mm.h>
 #include <linux/slab.h>
 
@@ -315,18 +324,18 @@ kfpu_begin(void)
 	uint8_t *state = zfs_kfpu_fpregs[smp_processor_id()];
 #if defined(HAVE_XSAVES)
 	if (static_cpu_has(X86_FEATURE_XSAVES)) {
-		kfpu_do_xsave("xsaves", state, ~0);
+		kfpu_do_xsave("xsaves", state, ~XFEATURE_MASK_XTILE);
 		return;
 	}
 #endif
 #if defined(HAVE_XSAVEOPT)
 	if (static_cpu_has(X86_FEATURE_XSAVEOPT)) {
-		kfpu_do_xsave("xsaveopt", state, ~0);
+		kfpu_do_xsave("xsaveopt", state, ~XFEATURE_MASK_XTILE);
 		return;
 	}
 #endif
 	if (static_cpu_has(X86_FEATURE_XSAVE)) {
-		kfpu_do_xsave("xsave", state, ~0);
+		kfpu_do_xsave("xsave", state, ~XFEATURE_MASK_XTILE);
 	} else if (static_cpu_has(X86_FEATURE_FXSR)) {
 		kfpu_save_fxsr(state);
 	} else {
@@ -376,12 +385,12 @@ kfpu_end(void)
 	uint8_t  *state = zfs_kfpu_fpregs[smp_processor_id()];
 #if defined(HAVE_XSAVES)
 	if (static_cpu_has(X86_FEATURE_XSAVES)) {
-		kfpu_do_xrstor("xrstors", state, ~0);
+		kfpu_do_xrstor("xrstors", state, ~XFEATURE_MASK_XTILE);
 		goto out;
 	}
 #endif
 	if (static_cpu_has(X86_FEATURE_XSAVE)) {
-		kfpu_do_xrstor("xrstor", state, ~0);
+		kfpu_do_xrstor("xrstor", state, ~XFEATURE_MASK_XTILE);
 	} else if (static_cpu_has(X86_FEATURE_FXSR)) {
 		kfpu_restore_fxsr(state);
 	} else {

include/os/linux/spl/sys/kmem_cache.h

@@ -198,6 +198,14 @@ extern uint64_t spl_kmem_cache_entry_size(kmem_cache_t *cache);
     spl_kmem_cache_create(name, size, align, ctor, dtor, rclm, priv, vmp, fl)
 #define	kmem_cache_set_move(skc, move)	spl_kmem_cache_set_move(skc, move)
 #define	kmem_cache_destroy(skc)		spl_kmem_cache_destroy(skc)
+/*
+ * This is necessary to be compatible with other kernel modules
+ * or in-tree filesystem that may define kmem_cache_alloc,
+ * like bcachefs does it now.
+ */
+#ifdef kmem_cache_alloc
+#undef kmem_cache_alloc
+#endif
 #define	kmem_cache_alloc(skc, flags)	spl_kmem_cache_alloc(skc, flags)
 #define	kmem_cache_free(skc, obj)	spl_kmem_cache_free(skc, obj)
 #define	kmem_cache_reap_now(skc)	spl_kmem_cache_reap_now(skc)

include/os/linux/spl/sys/types.h

@@ -38,7 +38,7 @@ typedef unsigned long		ulong_t;
 typedef unsigned long long	u_longlong_t;
 typedef long long		longlong_t;
 
-typedef unsigned long		intptr_t;
+typedef long			intptr_t;
 typedef unsigned long long	rlim64_t;
 
 typedef struct task_struct	kthread_t;

include/os/linux/spl/sys/uio.h

@@ -173,4 +173,16 @@ zfs_uio_iov_iter_init(zfs_uio_t *uio, struct iov_iter *iter, offset_t offset,
 }
 #endif
 
+#if defined(HAVE_ITER_IOV)
+#define	zfs_uio_iter_iov(iter)	iter_iov((iter))
+#else
+#define	zfs_uio_iter_iov(iter)	(iter)->iov
+#endif
+
+#if defined(HAVE_IOV_ITER_TYPE)
+#define	zfs_uio_iov_iter_type(iter)	iov_iter_type((iter))
+#else
+#define	zfs_uio_iov_iter_type(iter)	(iter)->type
+#endif
+
 #endif /* SPL_UIO_H */

include/os/linux/zfs/sys/trace_arc.h

@@ -51,7 +51,6 @@ DECLARE_EVENT_CLASS(zfs_arc_buf_hdr_class,
 	    __array(uint64_t,		hdr_dva_word, 2)
 	    __field(uint64_t,		hdr_birth)
 	    __field(uint32_t,		hdr_flags)
-	    __field(uint32_t,		hdr_bufcnt)
 	    __field(arc_buf_contents_t,	hdr_type)
 	    __field(uint16_t,		hdr_psize)
 	    __field(uint16_t,		hdr_lsize)
@@ -70,7 +69,6 @@ DECLARE_EVENT_CLASS(zfs_arc_buf_hdr_class,
 	    __entry->hdr_dva_word[1]	= ab->b_dva.dva_word[1];
 	    __entry->hdr_birth		= ab->b_birth;
 	    __entry->hdr_flags		= ab->b_flags;
-	    __entry->hdr_bufcnt	= ab->b_l1hdr.b_bufcnt;
 	    __entry->hdr_psize		= ab->b_psize;
 	    __entry->hdr_lsize		= ab->b_lsize;
 	    __entry->hdr_spa		= ab->b_spa;
@@ -84,12 +82,12 @@ DECLARE_EVENT_CLASS(zfs_arc_buf_hdr_class,
 	    __entry->hdr_refcount	= ab->b_l1hdr.b_refcnt.rc_count;
 	),
 	TP_printk("hdr { dva 0x%llx:0x%llx birth %llu "
-	    "flags 0x%x bufcnt %u type %u psize %u lsize %u spa %llu "
+	    "flags 0x%x type %u psize %u lsize %u spa %llu "
 	    "state_type %u access %lu mru_hits %u mru_ghost_hits %u "
 	    "mfu_hits %u mfu_ghost_hits %u l2_hits %u refcount %lli }",
 	    __entry->hdr_dva_word[0], __entry->hdr_dva_word[1],
 	    __entry->hdr_birth, __entry->hdr_flags,
-	    __entry->hdr_bufcnt, __entry->hdr_type, __entry->hdr_psize,
+	    __entry->hdr_type, __entry->hdr_psize,
 	    __entry->hdr_lsize, __entry->hdr_spa, __entry->hdr_state_type,
 	    __entry->hdr_access, __entry->hdr_mru_hits,
 	    __entry->hdr_mru_ghost_hits, __entry->hdr_mfu_hits,
@@ -192,7 +190,6 @@ DECLARE_EVENT_CLASS(zfs_arc_miss_class,
 	    __array(uint64_t,		hdr_dva_word, 2)
 	    __field(uint64_t,		hdr_birth)
 	    __field(uint32_t,		hdr_flags)
-	    __field(uint32_t,		hdr_bufcnt)
 	    __field(arc_buf_contents_t,	hdr_type)
 	    __field(uint16_t,		hdr_psize)
 	    __field(uint16_t,		hdr_lsize)
@@ -223,7 +220,6 @@ DECLARE_EVENT_CLASS(zfs_arc_miss_class,
 	    __entry->hdr_dva_word[1]	= hdr->b_dva.dva_word[1];
 	    __entry->hdr_birth		= hdr->b_birth;
 	    __entry->hdr_flags		= hdr->b_flags;
-	    __entry->hdr_bufcnt		= hdr->b_l1hdr.b_bufcnt;
 	    __entry->hdr_psize		= hdr->b_psize;
 	    __entry->hdr_lsize		= hdr->b_lsize;
 	    __entry->hdr_spa		= hdr->b_spa;
@@ -255,7 +251,7 @@ DECLARE_EVENT_CLASS(zfs_arc_miss_class,
 	    __entry->zb_blkid		= zb->zb_blkid;
 	),
 	TP_printk("hdr { dva 0x%llx:0x%llx birth %llu "
-	    "flags 0x%x bufcnt %u psize %u lsize %u spa %llu state_type %u "
+	    "flags 0x%x psize %u lsize %u spa %llu state_type %u "
 	    "access %lu mru_hits %u mru_ghost_hits %u mfu_hits %u "
 	    "mfu_ghost_hits %u l2_hits %u refcount %lli } "
 	    "bp { dva0 0x%llx:0x%llx dva1 0x%llx:0x%llx dva2 "
@@ -264,7 +260,7 @@ DECLARE_EVENT_CLASS(zfs_arc_miss_class,
 	    "blkid %llu }",
 	    __entry->hdr_dva_word[0], __entry->hdr_dva_word[1],
 	    __entry->hdr_birth, __entry->hdr_flags,
-	    __entry->hdr_bufcnt, __entry->hdr_psize, __entry->hdr_lsize,
+	    __entry->hdr_psize, __entry->hdr_lsize,
 	    __entry->hdr_spa, __entry->hdr_state_type, __entry->hdr_access,
 	    __entry->hdr_mru_hits, __entry->hdr_mru_ghost_hits,
 	    __entry->hdr_mfu_hits, __entry->hdr_mfu_ghost_hits,

include/os/linux/zfs/sys/trace_dbuf.h

@@ -60,8 +60,12 @@
 
 #define	DBUF_TP_FAST_ASSIGN						\
 	if (db != NULL) {						\
+		if (POINTER_IS_VALID(DB_DNODE(db)->dn_objset)) {	\
 			__assign_str(os_spa,				\
 			spa_name(DB_DNODE(db)->dn_objset->os_spa));	\
+		} else {						\
+			__assign_str(os_spa, "NULL");			\
+		}							\
 									\
 		__entry->ds_object = db->db_objset->os_dsl_dataset ?	\
 		db->db_objset->os_dsl_dataset->ds_object : 0;		\

include/os/linux/zfs/sys/zfs_vfsops_os.h

@@ -105,7 +105,6 @@ struct zfsvfs {
 	rrmlock_t	z_teardown_lock;
 	krwlock_t	z_teardown_inactive_lock;
 	list_t		z_all_znodes;	/* all znodes in the fs */
-	uint64_t	z_nr_znodes;	/* number of znodes in the fs */
 	unsigned long	z_rollback_time; /* last online rollback time */
 	unsigned long	z_snap_defer_time; /* last snapshot unmount deferral */
 	kmutex_t	z_znodes_lock;	/* lock for z_all_znodes */

include/os/linux/zfs/sys/zpl.h

@@ -52,7 +52,11 @@ extern const struct inode_operations zpl_special_inode_operations;
 
 /* zpl_file.c */
 extern const struct address_space_operations zpl_address_space_operations;
+#ifdef HAVE_VFS_FILE_OPERATIONS_EXTEND
+extern const struct file_operations_extend zpl_file_operations;
+#else
 extern const struct file_operations zpl_file_operations;
+#endif
 extern const struct file_operations zpl_dir_file_operations;
 
 /* zpl_super.c */
@@ -180,6 +184,55 @@ zpl_dir_emit_dots(struct file *file, zpl_dir_context_t *ctx)
 }
 #endif /* HAVE_VFS_ITERATE */
 
+
+/* zpl_file_range.c */
+
+/* handlers for file_operations of the same name */
+extern ssize_t zpl_copy_file_range(struct file *src_file, loff_t src_off,
+    struct file *dst_file, loff_t dst_off, size_t len, unsigned int flags);
+extern loff_t zpl_remap_file_range(struct file *src_file, loff_t src_off,
+    struct file *dst_file, loff_t dst_off, loff_t len, unsigned int flags);
+extern int zpl_clone_file_range(struct file *src_file, loff_t src_off,
+    struct file *dst_file, loff_t dst_off, uint64_t len);
+extern int zpl_dedupe_file_range(struct file *src_file, loff_t src_off,
+    struct file *dst_file, loff_t dst_off, uint64_t len);
+
+/* compat for FICLONE/FICLONERANGE/FIDEDUPERANGE ioctls */
+typedef struct {
+	int64_t		fcr_src_fd;
+	uint64_t	fcr_src_offset;
+	uint64_t	fcr_src_length;
+	uint64_t	fcr_dest_offset;
+} zfs_ioc_compat_file_clone_range_t;
+
+typedef struct {
+	int64_t		fdri_dest_fd;
+	uint64_t	fdri_dest_offset;
+	uint64_t	fdri_bytes_deduped;
+	int32_t		fdri_status;
+	uint32_t	fdri_reserved;
+} zfs_ioc_compat_dedupe_range_info_t;
+
+typedef struct {
+	uint64_t	fdr_src_offset;
+	uint64_t	fdr_src_length;
+	uint16_t	fdr_dest_count;
+	uint16_t	fdr_reserved1;
+	uint32_t	fdr_reserved2;
+	zfs_ioc_compat_dedupe_range_info_t	fdr_info[];
+} zfs_ioc_compat_dedupe_range_t;
+
+#define	ZFS_IOC_COMPAT_FICLONE		_IOW(0x94, 9, int)
+#define	ZFS_IOC_COMPAT_FICLONERANGE \
+    _IOW(0x94, 13, zfs_ioc_compat_file_clone_range_t)
+#define	ZFS_IOC_COMPAT_FIDEDUPERANGE \
+    _IOWR(0x94, 54, zfs_ioc_compat_dedupe_range_t)
+
+extern long zpl_ioctl_ficlone(struct file *filp, void *arg);
+extern long zpl_ioctl_ficlonerange(struct file *filp, void *arg);
+extern long zpl_ioctl_fideduperange(struct file *filp, void *arg);
+
+
 #if defined(HAVE_INODE_TIMESTAMP_TRUNCATE)
 #define	zpl_inode_timestamp_truncate(ts, ip)	timestamp_truncate(ts, ip)
 #elif defined(HAVE_INODE_TIMESPEC64_TIMES)

include/sys/arc_impl.h

@@ -159,10 +159,6 @@ struct arc_write_callback {
  * these two allocation states.
  */
 typedef struct l1arc_buf_hdr {
-	/* for waiting on reads to complete */
-	kcondvar_t		b_cv;
-	uint8_t			b_byteswap;
-
 	/* protected by arc state mutex */
 	arc_state_t		*b_state;
 	multilist_node_t	b_arc_node;
@@ -173,7 +169,7 @@ typedef struct l1arc_buf_hdr {
 	uint32_t		b_mru_ghost_hits;
 	uint32_t		b_mfu_hits;
 	uint32_t		b_mfu_ghost_hits;
-	uint32_t		b_bufcnt;
+	uint8_t			b_byteswap;
 	arc_buf_t		*b_buf;
 
 	/* self protecting */
@@ -436,12 +432,12 @@ typedef struct l2arc_dev {
  */
 typedef struct arc_buf_hdr_crypt {
 	abd_t			*b_rabd;	/* raw encrypted data */
-	dmu_object_type_t	b_ot;		/* object type */
-	uint32_t		b_ebufcnt;	/* count of encrypted buffers */
 
 	/* dsobj for looking up encryption key for l2arc encryption */
 	uint64_t		b_dsobj;
 
+	dmu_object_type_t	b_ot;		/* object type */
+
 	/* encryption parameters */
 	uint8_t			b_salt[ZIO_DATA_SALT_LEN];
 	uint8_t			b_iv[ZIO_DATA_IV_LEN];

include/sys/bpobj.h

@@ -60,7 +60,7 @@ typedef struct bpobj {
 	kmutex_t	bpo_lock;
 	objset_t	*bpo_os;
 	uint64_t	bpo_object;
-	int		bpo_epb;
+	uint32_t	bpo_epb;
 	uint8_t		bpo_havecomp;
 	uint8_t		bpo_havesubobj;
 	uint8_t		bpo_havefreed;

include/sys/brt.h

@@ -36,6 +36,7 @@ extern "C" {
 #endif
 
 extern boolean_t brt_entry_decref(spa_t *spa, const blkptr_t *bp);
+extern uint64_t brt_entry_get_refcount(spa_t *spa, const blkptr_t *bp);
 
 extern uint64_t brt_get_dspace(spa_t *spa);
 extern uint64_t brt_get_used(spa_t *spa);

include/sys/dmu.h

@@ -572,11 +572,15 @@ int dmu_buf_hold(objset_t *os, uint64_t object, uint64_t offset,
 int dmu_buf_hold_array(objset_t *os, uint64_t object, uint64_t offset,
     uint64_t length, int read, const void *tag, int *numbufsp,
     dmu_buf_t ***dbpp);
+int dmu_buf_hold_noread(objset_t *os, uint64_t object, uint64_t offset,
+    const void *tag, dmu_buf_t **dbp);
 int dmu_buf_hold_by_dnode(dnode_t *dn, uint64_t offset,
     const void *tag, dmu_buf_t **dbp, int flags);
 int dmu_buf_hold_array_by_dnode(dnode_t *dn, uint64_t offset,
     uint64_t length, boolean_t read, const void *tag, int *numbufsp,
     dmu_buf_t ***dbpp, uint32_t flags);
+int dmu_buf_hold_noread_by_dnode(dnode_t *dn, uint64_t offset, const void *tag,
+    dmu_buf_t **dbp);
 /*
  * Add a reference to a dmu buffer that has already been held via
  * dmu_buf_hold() in the current context.

include/sys/dmu_impl.h

@@ -247,8 +247,6 @@ typedef struct dmu_sendstatus {
 
 void dmu_object_zapify(objset_t *, uint64_t, dmu_object_type_t, dmu_tx_t *);
 void dmu_object_free_zapified(objset_t *, uint64_t, dmu_tx_t *);
-int dmu_buf_hold_noread(objset_t *, uint64_t, uint64_t,
-    const void *, dmu_buf_t **);
 
 #ifdef	__cplusplus
 }

include/sys/dmu_zfetch.h

@@ -36,8 +36,6 @@
 extern "C" {
 #endif
 
-extern uint64_t	zfetch_array_rd_sz;
-
 struct dnode;				/* so we can reference dnode */
 
 typedef struct zfetch {

include/sys/fm/fs/zfs.h

@@ -102,8 +102,6 @@ extern "C" {
 #define	FM_EREPORT_PAYLOAD_ZFS_ZIO_TIMESTAMP	"zio_timestamp"
 #define	FM_EREPORT_PAYLOAD_ZFS_ZIO_DELTA	"zio_delta"
 #define	FM_EREPORT_PAYLOAD_ZFS_PREV_STATE	"prev_state"
-#define	FM_EREPORT_PAYLOAD_ZFS_CKSUM_EXPECTED	"cksum_expected"
-#define	FM_EREPORT_PAYLOAD_ZFS_CKSUM_ACTUAL	"cksum_actual"
 #define	FM_EREPORT_PAYLOAD_ZFS_CKSUM_ALGO	"cksum_algorithm"
 #define	FM_EREPORT_PAYLOAD_ZFS_CKSUM_BYTESWAP	"cksum_byteswap"
 #define	FM_EREPORT_PAYLOAD_ZFS_BAD_OFFSET_RANGES "bad_ranges"
@@ -112,8 +110,6 @@ extern "C" {
 #define	FM_EREPORT_PAYLOAD_ZFS_BAD_RANGE_CLEARS	"bad_range_clears"
 #define	FM_EREPORT_PAYLOAD_ZFS_BAD_SET_BITS	"bad_set_bits"
 #define	FM_EREPORT_PAYLOAD_ZFS_BAD_CLEARED_BITS	"bad_cleared_bits"
-#define	FM_EREPORT_PAYLOAD_ZFS_BAD_SET_HISTOGRAM "bad_set_histogram"
-#define	FM_EREPORT_PAYLOAD_ZFS_BAD_CLEARED_HISTOGRAM "bad_cleared_histogram"
 #define	FM_EREPORT_PAYLOAD_ZFS_SNAPSHOT_NAME	"snapshot_name"
 #define	FM_EREPORT_PAYLOAD_ZFS_DEVICE_NAME	"device_name"
 #define	FM_EREPORT_PAYLOAD_ZFS_RAW_DEVICE_NAME	"raw_name"

include/sys/metaslab.h

@@ -80,7 +80,6 @@ uint64_t metaslab_largest_allocatable(metaslab_t *);
 #define	METASLAB_ASYNC_ALLOC		0x8
 #define	METASLAB_DONT_THROTTLE		0x10
 #define	METASLAB_MUST_RESERVE		0x20
-#define	METASLAB_FASTWRITE		0x40
 #define	METASLAB_ZIL			0x80
 
 int metaslab_alloc(spa_t *, metaslab_class_t *, uint64_t,
@@ -96,8 +95,6 @@ void metaslab_unalloc_dva(spa_t *, const dva_t *, uint64_t);
 int metaslab_claim(spa_t *, const blkptr_t *, uint64_t);
 int metaslab_claim_impl(vdev_t *, uint64_t, uint64_t, uint64_t);
 void metaslab_check_free(spa_t *, const blkptr_t *);
-void metaslab_fastwrite_mark(spa_t *, const blkptr_t *);
-void metaslab_fastwrite_unmark(spa_t *, const blkptr_t *);
 
 void metaslab_stat_init(void);
 void metaslab_stat_fini(void);

include/sys/metaslab_impl.h

@@ -250,7 +250,6 @@ struct metaslab_group {
 	int64_t			mg_activation_count;
 	metaslab_class_t	*mg_class;
 	vdev_t			*mg_vd;
-	taskq_t			*mg_taskq;
 	metaslab_group_t	*mg_prev;
 	metaslab_group_t	*mg_next;
 
@@ -313,7 +312,7 @@ struct metaslab_group {
  * Each metaslab maintains a set of in-core trees to track metaslab
  * operations.  The in-core free tree (ms_allocatable) contains the list of
  * free segments which are eligible for allocation.  As blocks are
- * allocated, the allocated segment are removed from the ms_allocatable and
+ * allocated, the allocated segments are removed from the ms_allocatable and
  * added to a per txg allocation tree (ms_allocating).  As blocks are
  * freed, they are added to the free tree (ms_freeing).  These trees
  * allow us to process all allocations and frees in syncing context
@@ -366,9 +365,9 @@ struct metaslab_group {
 struct metaslab {
 	/*
 	 * This is the main lock of the metaslab and its purpose is to
-	 * coordinate our allocations and frees [e.g metaslab_block_alloc(),
+	 * coordinate our allocations and frees [e.g., metaslab_block_alloc(),
 	 * metaslab_free_concrete(), ..etc] with our various syncing
-	 * procedures [e.g. metaslab_sync(), metaslab_sync_done(), ..etc].
+	 * procedures [e.g., metaslab_sync(), metaslab_sync_done(), ..etc].
 	 *
 	 * The lock is also used during some miscellaneous operations like
 	 * using the metaslab's histogram for the metaslab group's histogram

include/sys/spa.h

@@ -723,16 +723,10 @@ typedef enum spa_mode {
  * Send TRIM commands in-line during normal pool operation while deleting.
  *	OFF: no
  *	ON: yes
- * NB: IN_FREEBSD_BASE is defined within the FreeBSD sources.
  */
 typedef enum {
 	SPA_AUTOTRIM_OFF = 0,	/* default */
 	SPA_AUTOTRIM_ON,
-#ifdef IN_FREEBSD_BASE
-	SPA_AUTOTRIM_DEFAULT = SPA_AUTOTRIM_ON,
-#else
-	SPA_AUTOTRIM_DEFAULT = SPA_AUTOTRIM_OFF,
-#endif
 } spa_autotrim_t;
 
 /*

include/sys/spa_impl.h

@@ -250,6 +250,7 @@ struct spa {
 	uint64_t	spa_min_ashift;		/* of vdevs in normal class */
 	uint64_t	spa_max_ashift;		/* of vdevs in normal class */
 	uint64_t	spa_min_alloc;		/* of vdevs in normal class */
+	uint64_t	spa_gcd_alloc;		/* of vdevs in normal class */
 	uint64_t	spa_config_guid;	/* config pool guid */
 	uint64_t	spa_load_guid;		/* spa_load initialized guid */
 	uint64_t	spa_last_synced_guid;	/* last synced guid */
@@ -422,7 +423,9 @@ struct spa {
 
 	hrtime_t	spa_ccw_fail_time;	/* Conf cache write fail time */
 	taskq_t		*spa_zvol_taskq;	/* Taskq for minor management */
+	taskq_t		*spa_metaslab_taskq;	/* Taskq for metaslab preload */
 	taskq_t		*spa_prefetch_taskq;	/* Taskq for prefetch threads */
+	taskq_t		*spa_upgrade_taskq;	/* Taskq for upgrade jobs */
 	uint64_t	spa_multihost;		/* multihost aware (mmp) */
 	mmp_thread_t	spa_mmp;		/* multihost mmp thread */
 	list_t		spa_leaf_list;		/* list of leaf vdevs */
@@ -446,8 +449,6 @@ struct spa {
 	 */
 	spa_config_lock_t spa_config_lock[SCL_LOCKS]; /* config changes */
 	zfs_refcount_t	spa_refcount;		/* number of opens */
-
-	taskq_t		*spa_upgrade_taskq;	/* taskq for upgrade jobs */
 };
 
 extern char *spa_config_path;

include/sys/vdev_impl.h

@@ -266,7 +266,6 @@ struct vdev {
 	metaslab_group_t *vdev_mg;	/* metaslab group		*/
 	metaslab_group_t *vdev_log_mg;	/* embedded slog metaslab group	*/
 	metaslab_t	**vdev_ms;	/* metaslab array		*/
-	uint64_t	vdev_pending_fastwrite; /* allocated fastwrites */
 	txg_list_t	vdev_ms_list;	/* per-txg dirty metaslab lists	*/
 	txg_list_t	vdev_dtl_list;	/* per-txg dirty DTL lists	*/
 	txg_node_t	vdev_txg_node;	/* per-txg dirty vdev linkage	*/
@@ -420,6 +419,7 @@ struct vdev {
 	boolean_t	vdev_copy_uberblocks;  /* post expand copy uberblocks */
 	boolean_t	vdev_resilver_deferred;  /* resilver deferred */
 	boolean_t	vdev_kobj_flag; /* kobj event record */
+	boolean_t	vdev_attaching; /* vdev attach ashift handling */
 	vdev_queue_t	vdev_queue;	/* I/O deadline schedule queue	*/
 	spa_aux_vdev_t	*vdev_aux;	/* for l2cache and spares vdevs	*/
 	zio_t		*vdev_probe_zio; /* root of current probe	*/

include/sys/zil_impl.h

@@ -38,14 +38,22 @@ extern "C" {
 /*
  * Possible states for a given lwb structure.
  *
- * An lwb will start out in the "closed" state, and then transition to
- * the "opened" state via a call to zil_lwb_write_open(). When
- * transitioning from "closed" to "opened" the zilog's "zl_issuer_lock"
- * must be held.
+ * An lwb will start out in the "new" state, and transition to the "opened"
+ * state via a call to zil_lwb_write_open() on first itx assignment.  When
+ * transitioning from "new" to "opened" the zilog's "zl_issuer_lock" must be
+ * held.
  *
- * After the lwb is "opened", it can transition into the "issued" state
- * via zil_lwb_write_close(). Again, the zilog's "zl_issuer_lock" must
- * be held when making this transition.
+ * After the lwb is "opened", it can be assigned number of itxs and transition
+ * into the "closed" state via zil_lwb_write_close() when full or on timeout.
+ * When transitioning from "opened" to "closed" the zilog's "zl_issuer_lock"
+ * must be held.  New lwb allocation also takes "zl_lock" to protect the list.
+ *
+ * After the lwb is "closed", it can transition into the "ready" state via
+ * zil_lwb_write_issue().  "zl_lock" must be held when making this transition.
+ * Since it is done by the same thread, "zl_issuer_lock" is not needed.
+ *
+ * When lwb in "ready" state receives its block pointer, it can transition to
+ * "issued". "zl_lock" must be held when making this transition.
  *
  * After the lwb's write zio completes, it transitions into the "write
  * done" state via zil_lwb_write_done(); and then into the "flush done"
@@ -62,17 +70,20 @@ extern "C" {
  *
  * Additionally, correctness when reading an lwb's state is often
  * achieved by exploiting the fact that these state transitions occur in
- * this specific order; i.e. "closed" to "opened" to "issued" to "done".
+ * this specific order; i.e. "new" to "opened" to "closed" to "ready" to
+ * "issued" to "write_done" and finally "flush_done".
  *
- * Thus, if an lwb is in the "closed" or "opened" state, holding the
+ * Thus, if an lwb is in the "new" or "opened" state, holding the
  * "zl_issuer_lock" will prevent a concurrent thread from transitioning
- * that lwb to the "issued" state. Likewise, if an lwb is already in the
- * "issued" state, holding the "zl_lock" will prevent a concurrent
- * thread from transitioning that lwb to the "write done" state.
+ * that lwb to the "closed" state. Likewise, if an lwb is already in the
+ * "ready" state, holding the "zl_lock" will prevent a concurrent thread
+ * from transitioning that lwb to the "issued" state.
  */
 typedef enum {
-    LWB_STATE_CLOSED,
+    LWB_STATE_NEW,
     LWB_STATE_OPENED,
+    LWB_STATE_CLOSED,
+    LWB_STATE_READY,
     LWB_STATE_ISSUED,
     LWB_STATE_WRITE_DONE,
     LWB_STATE_FLUSH_DONE,
@@ -91,18 +102,21 @@ typedef enum {
 typedef struct lwb {
 	zilog_t		*lwb_zilog;	/* back pointer to log struct */
 	blkptr_t	lwb_blk;	/* on disk address of this log blk */
-	boolean_t	lwb_fastwrite;	/* is blk marked for fastwrite? */
+	boolean_t	lwb_slim;	/* log block has slim format */
 	boolean_t	lwb_slog;	/* lwb_blk is on SLOG device */
-	boolean_t	lwb_indirect;	/* do not postpone zil_lwb_commit() */
+	int		lwb_error;	/* log block allocation error */
+	int		lwb_nmax;	/* max bytes in the buffer */
 	int		lwb_nused;	/* # used bytes in buffer */
 	int		lwb_nfilled;	/* # filled bytes in buffer */
 	int		lwb_sz;		/* size of block and buffer */
 	lwb_state_t	lwb_state;	/* the state of this lwb */
 	char		*lwb_buf;	/* log write buffer */
+	zio_t		*lwb_child_zio;	/* parent zio for children */
 	zio_t		*lwb_write_zio;	/* zio for the lwb buffer */
 	zio_t		*lwb_root_zio;	/* root zio for lwb write and flushes */
 	hrtime_t	lwb_issued_timestamp; /* when was the lwb issued? */
 	uint64_t	lwb_issued_txg;	/* the txg when the write is issued */
+	uint64_t	lwb_alloc_txg;	/* the txg when lwb_blk is allocated */
 	uint64_t	lwb_max_txg;	/* highest txg in this lwb */
 	list_node_t	lwb_node;	/* zilog->zl_lwb_list linkage */
 	list_node_t	lwb_issue_node;	/* linkage of lwbs ready for issue */

include/sys/zio.h

@@ -222,7 +222,6 @@ typedef uint64_t zio_flag_t;
 #define	ZIO_FLAG_NOPWRITE	(1ULL << 28)
 #define	ZIO_FLAG_REEXECUTED	(1ULL << 29)
 #define	ZIO_FLAG_DELEGATED	(1ULL << 30)
-#define	ZIO_FLAG_FASTWRITE	(1ULL << 31)
 
 #define	ZIO_FLAG_MUSTSUCCEED		0
 #define	ZIO_FLAG_RAW	(ZIO_FLAG_RAW_COMPRESS | ZIO_FLAG_RAW_ENCRYPT)

include/sys/zio_checksum.h

@@ -94,8 +94,6 @@ typedef const struct zio_checksum_info {
 } zio_checksum_info_t;
 
 typedef struct zio_bad_cksum {
-	zio_cksum_t		zbc_expected;
-	zio_cksum_t		zbc_actual;
 	const char		*zbc_checksum_name;
 	uint8_t			zbc_byteswapped;
 	uint8_t			zbc_injected;

lib/libshare/os/freebsd/nfs.c

@@ -161,7 +161,8 @@ nfs_is_shared(sa_share_impl_t impl_share)
 static int
 nfs_validate_shareopts(const char *shareopts)
 {
-	(void) shareopts;
+	if (strlen(shareopts) == 0)
+		return (SA_SYNTAX_ERR);
 	return (SA_OK);
 }
 

lib/libshare/os/linux/nfs.c

@@ -319,12 +319,49 @@ get_linux_shareopts_cb(const char *key, const char *value, void *cookie)
 	    "wdelay" };
 
 	char **plinux_opts = (char **)cookie;
+	char *host, *val_dup, *literal, *next;
 
-	/* host-specific options, these are taken care of elsewhere */
-	if (strcmp(key, "ro") == 0 || strcmp(key, "rw") == 0 ||
-	    strcmp(key, "sec") == 0)
+	if (strcmp(key, "sec") == 0)
 		return (SA_OK);
 
+	if (strcmp(key, "ro") == 0 || strcmp(key, "rw") == 0) {
+		if (value == NULL || strlen(value) == 0)
+			return (SA_OK);
+		val_dup = strdup(value);
+		host = val_dup;
+		if (host == NULL)
+			return (SA_NO_MEMORY);
+		do {
+			if (*host == '[') {
+				host++;
+				literal = strchr(host, ']');
+				if (literal == NULL) {
+					free(val_dup);
+					return (SA_SYNTAX_ERR);
+				}
+				if (literal[1] == '\0')
+					next = NULL;
+				else if (literal[1] == '/') {
+					next = strchr(literal + 2, ':');
+					if (next != NULL)
+						++next;
+				} else if (literal[1] == ':')
+					next = literal + 2;
+				else {
+					free(val_dup);
+					return (SA_SYNTAX_ERR);
+				}
+			} else {
+				next = strchr(host, ':');
+				if (next != NULL)
+					++next;
+			}
+			host = next;
+		} while (host != NULL);
+		free(val_dup);
+		return (SA_OK);
+	}
+
 	if (strcmp(key, "anon") == 0)
 		key = "anonuid";
 
@@ -472,6 +509,10 @@ static int
 nfs_validate_shareopts(const char *shareopts)
 {
 	char *linux_opts = NULL;
+
+	if (strlen(shareopts) == 0)
+		return (SA_SYNTAX_ERR);
+
 	int error = get_linux_shareopts(shareopts, &linux_opts);
 	if (error != SA_OK)
 		return (error);

lib/libzfs/Makefile.am

@@ -57,7 +57,7 @@ libzfs_la_LIBADD = \
 	libzutil.la \
 	libuutil.la
 
-libzfs_la_LIBADD += -lm $(LIBCRYPTO_LIBS) $(ZLIB_LIBS) $(LIBFETCH_LIBS) $(LTLIBINTL)
+libzfs_la_LIBADD += -lrt -lm $(LIBCRYPTO_LIBS) $(ZLIB_LIBS) $(LIBFETCH_LIBS) $(LTLIBINTL)
 
 libzfs_la_LDFLAGS = -pthread
 

lib/libzfs/libzfs.abi

@@ -396,6 +396,7 @@
     <elf-symbol name='zfs_prop_readonly' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
     <elf-symbol name='zfs_prop_set' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
     <elf-symbol name='zfs_prop_set_list' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
+    <elf-symbol name='zfs_prop_set_list_flags' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
     <elf-symbol name='zfs_prop_setonce' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
     <elf-symbol name='zfs_prop_string_to_index' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
     <elf-symbol name='zfs_prop_to_name' type='func-type' binding='global-binding' visibility='default-visibility' is-defined='yes'/>
@@ -4424,6 +4425,12 @@
       <parameter type-id='5ce45b60' name='props'/>
       <return type-id='95e97e5e'/>
     </function-decl>
+    <function-decl name='zfs_prop_set_list_flags' mangled-name='zfs_prop_set_list_flags' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='zfs_prop_set_list_flags'>
+      <parameter type-id='9200a744' name='zhp'/>
+      <parameter type-id='5ce45b60' name='props'/>
+      <parameter type-id='95e97e5e' name='flags'/>
+      <return type-id='95e97e5e'/>
+    </function-decl>
     <function-decl name='zfs_prop_inherit' mangled-name='zfs_prop_inherit' visibility='default' binding='global' size-in-bits='64' elf-symbol-id='zfs_prop_inherit'>
       <parameter type-id='9200a744' name='zhp'/>
       <parameter type-id='80f4b756' name='propname'/>

lib/libzfs/libzfs_changelist.c

@@ -105,6 +105,15 @@ changelist_prefix(prop_changelist_t *clp)
 	    clp->cl_prop != ZFS_PROP_SHARESMB)
 		return (0);
 
+	/*
+	 * If CL_GATHER_DONT_UNMOUNT is set, don't want to unmount/unshare and
+	 * later (re)mount/(re)share the filesystem in postfix phase, so we
+	 * return from here. If filesystem is mounted or unmounted, leave it
+	 * as it is.
+	 */
+	if (clp->cl_gflags & CL_GATHER_DONT_UNMOUNT)
+		return (0);
+
 	if ((walk = uu_avl_walk_start(clp->cl_tree, UU_WALK_ROBUST)) == NULL)
 		return (-1);
 
@@ -129,8 +138,6 @@ changelist_prefix(prop_changelist_t *clp)
 			 */
 			switch (clp->cl_prop) {
 			case ZFS_PROP_MOUNTPOINT:
-				if (clp->cl_gflags & CL_GATHER_DONT_UNMOUNT)
-					break;
 				if (zfs_unmount(cn->cn_handle, NULL,
 				    clp->cl_mflags) != 0) {
 					ret = -1;
@@ -164,9 +171,8 @@ changelist_prefix(prop_changelist_t *clp)
  * reshare the filesystems as necessary.  In changelist_gather() we recorded
  * whether the filesystem was previously shared or mounted.  The action we take
  * depends on the previous state, and whether the value was previously 'legacy'.
- * For non-legacy properties, we only remount/reshare the filesystem if it was
- * previously mounted/shared.  Otherwise, we always remount/reshare the
- * filesystem.
+ * For non-legacy properties, we always remount/reshare the filesystem,
+ * if CL_GATHER_DONT_UNMOUNT is not set.
  */
 int
 changelist_postfix(prop_changelist_t *clp)
@@ -174,10 +180,17 @@ changelist_postfix(prop_changelist_t *clp)
 	prop_changenode_t *cn;
 	uu_avl_walk_t *walk;
 	char shareopts[ZFS_MAXPROPLEN];
-	int errors = 0;
 	boolean_t commit_smb_shares = B_FALSE;
 	boolean_t commit_nfs_shares = B_FALSE;
 
+	/*
+	 * If CL_GATHER_DONT_UNMOUNT is set, it means we don't want to (un)mount
+	 * or (re/un)share the filesystem, so we return from here. If filesystem
+	 * is mounted or unmounted, leave it as it is.
+	 */
+	if (clp->cl_gflags & CL_GATHER_DONT_UNMOUNT)
+		return (0);
+
 	/*
 	 * If we're changing the mountpoint, attempt to destroy the underlying
 	 * mountpoint.  All other datasets will have inherited from this dataset
@@ -240,17 +253,16 @@ changelist_postfix(prop_changelist_t *clp)
 		needs_key = (zfs_prop_get_int(cn->cn_handle,
 		    ZFS_PROP_KEYSTATUS) == ZFS_KEYSTATUS_UNAVAILABLE);
 
-		mounted = (clp->cl_gflags & CL_GATHER_DONT_UNMOUNT) ||
-		    zfs_is_mounted(cn->cn_handle, NULL);
+		mounted = zfs_is_mounted(cn->cn_handle, NULL);
 
 		if (!mounted && !needs_key && (cn->cn_mounted ||
-		    ((sharenfs || sharesmb || clp->cl_waslegacy) &&
+		    (((clp->cl_prop == ZFS_PROP_MOUNTPOINT &&
+		    clp->cl_prop == clp->cl_realprop) ||
+		    sharenfs || sharesmb || clp->cl_waslegacy) &&
 		    (zfs_prop_get_int(cn->cn_handle,
 		    ZFS_PROP_CANMOUNT) == ZFS_CANMOUNT_ON)))) {
 
-			if (zfs_mount(cn->cn_handle, NULL, 0) != 0)
-				errors++;
-			else
+			if (zfs_mount(cn->cn_handle, NULL, 0) == 0)
 				mounted = TRUE;
 		}
 
@@ -262,19 +274,19 @@ changelist_postfix(prop_changelist_t *clp)
 		const enum sa_protocol nfs[] =
 		    {SA_PROTOCOL_NFS, SA_NO_PROTOCOL};
 		if (sharenfs && mounted) {
-			errors += zfs_share(cn->cn_handle, nfs);
+			zfs_share(cn->cn_handle, nfs);
 			commit_nfs_shares = B_TRUE;
 		} else if (cn->cn_shared || clp->cl_waslegacy) {
-			errors += zfs_unshare(cn->cn_handle, NULL, nfs);
+			zfs_unshare(cn->cn_handle, NULL, nfs);
 			commit_nfs_shares = B_TRUE;
 		}
 		const enum sa_protocol smb[] =
 		    {SA_PROTOCOL_SMB, SA_NO_PROTOCOL};
 		if (sharesmb && mounted) {
-			errors += zfs_share(cn->cn_handle, smb);
+			zfs_share(cn->cn_handle, smb);
 			commit_smb_shares = B_TRUE;
 		} else if (cn->cn_shared || clp->cl_waslegacy) {
-			errors += zfs_unshare(cn->cn_handle, NULL, smb);
+			zfs_unshare(cn->cn_handle, NULL, smb);
 			commit_smb_shares = B_TRUE;
 		}
 	}
@@ -288,7 +300,7 @@ changelist_postfix(prop_changelist_t *clp)
 	zfs_commit_shares(proto);
 	uu_avl_walk_end(walk);
 
-	return (errors ? -1 : 0);
+	return (0);
 }
 
 /*

lib/libzfs/libzfs_dataset.c

@@ -1771,14 +1771,24 @@ error:
 	return (ret);
 }
 
-
-
 /*
  * Given an nvlist of property names and values, set the properties for the
  * given dataset.
  */
 int
 zfs_prop_set_list(zfs_handle_t *zhp, nvlist_t *props)
+{
+	return (zfs_prop_set_list_flags(zhp, props, 0));
+}
+
+/*
+ * Given an nvlist of property names, values and flags, set the properties
+ * for the given dataset. If ZFS_SET_NOMOUNT is set, it allows to update
+ * mountpoint, sharenfs and sharesmb properties without (un/re)mounting
+ * and (un/re)sharing the dataset.
+ */
+int
+zfs_prop_set_list_flags(zfs_handle_t *zhp, nvlist_t *props, int flags)
 {
 	zfs_cmd_t zc = {"\0"};
 	int ret = -1;
@@ -1848,7 +1858,9 @@ zfs_prop_set_list(zfs_handle_t *zhp, nvlist_t *props)
 		if (prop != ZFS_PROP_CANMOUNT ||
 		    (fnvpair_value_uint64(elem) == ZFS_CANMOUNT_OFF &&
 		    zfs_is_mounted(zhp, NULL))) {
-			cls[cl_idx] = changelist_gather(zhp, prop, 0, 0);
+			cls[cl_idx] = changelist_gather(zhp, prop,
+			    ((flags & ZFS_SET_NOMOUNT) ?
+			    CL_GATHER_DONT_UNMOUNT : 0), 0);
 			if (cls[cl_idx] == NULL)
 				goto error;
 		}

lib/libzfs/libzfs_mount.c

@@ -1300,7 +1300,7 @@ zpool_enable_datasets(zpool_handle_t *zhp, const char *mntopts, int flags)
 	zfs_foreach_mountpoint(zhp->zpool_hdl, cb.cb_handles, cb.cb_used,
 	    zfs_mount_one, &ms, B_TRUE);
 	if (ms.ms_mntstatus != 0)
-		ret = ms.ms_mntstatus;
+		ret = EZFS_MOUNTFAILED;
 
 	/*
 	 * Share all filesystems that need to be shared. This needs to be
@@ -1311,7 +1311,7 @@ zpool_enable_datasets(zpool_handle_t *zhp, const char *mntopts, int flags)
 	zfs_foreach_mountpoint(zhp->zpool_hdl, cb.cb_handles, cb.cb_used,
 	    zfs_share_one, &ms, B_FALSE);
 	if (ms.ms_mntstatus != 0)
-		ret = ms.ms_mntstatus;
+		ret = EZFS_SHAREFAILED;
 	else
 		zfs_commit_shares(NULL);
 

lib/libzfs/libzfs_pool.c

@@ -29,7 +29,7 @@
  * Copyright (c) 2017, Intel Corporation.
  * Copyright (c) 2018, loli10K <ezomori.nozomu@gmail.com>
  * Copyright (c) 2021, Colm Buckley <colm@tuatha.org>
- * Copyright (c) 2021, Klara Inc.
+ * Copyright (c) 2021, 2023, Klara Inc.
  */
 
 #include <errno.h>
@@ -255,6 +255,7 @@ zpool_get_state_str(zpool_handle_t *zhp)
 	if (zpool_get_state(zhp) == POOL_STATE_UNAVAIL) {
 		str = gettext("FAULTED");
 	} else if (status == ZPOOL_STATUS_IO_FAILURE_WAIT ||
+	    status == ZPOOL_STATUS_IO_FAILURE_CONTINUE ||
 	    status == ZPOOL_STATUS_IO_FAILURE_MMP) {
 		str = gettext("SUSPENDED");
 	} else {
@@ -3926,6 +3927,12 @@ zpool_vdev_remove(zpool_handle_t *zhp, const char *path)
 
 	switch (errno) {
 
+	case EALREADY:
+		zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
+		    "removal for this vdev is already in progress."));
+		(void) zfs_error(hdl, EZFS_BUSY, errbuf);
+		break;
+
 	case EINVAL:
 		zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
 		    "invalid config; all top-level vdevs must "

lib/libzfs/libzfs_sendrecv.c

@@ -928,6 +928,39 @@ zfs_send_progress(zfs_handle_t *zhp, int fd, uint64_t *bytes_written,
 	return (0);
 }
 
+static volatile boolean_t send_progress_thread_signal_duetotimer;
+static void
+send_progress_thread_act(int sig, siginfo_t *info, void *ucontext)
+{
+	(void) sig, (void) ucontext;
+	send_progress_thread_signal_duetotimer = info->si_code == SI_TIMER;
+}
+
+struct timer_desirability {
+	timer_t timer;
+	boolean_t desired;
+};
+static void
+timer_delete_cleanup(void *timer)
+{
+	struct timer_desirability *td = timer;
+	if (td->desired)
+		timer_delete(td->timer);
+}
+
+#ifdef SIGINFO
+#define	SEND_PROGRESS_THREAD_PARENT_BLOCK_SIGINFO sigaddset(&new, SIGINFO)
+#else
+#define	SEND_PROGRESS_THREAD_PARENT_BLOCK_SIGINFO
+#endif
+#define	SEND_PROGRESS_THREAD_PARENT_BLOCK(old) { \
+	sigset_t new; \
+	sigemptyset(&new); \
+	sigaddset(&new, SIGUSR1); \
+	SEND_PROGRESS_THREAD_PARENT_BLOCK_SIGINFO; \
+	pthread_sigmask(SIG_BLOCK, &new, old); \
+}
+
 static void *
 send_progress_thread(void *arg)
 {
@@ -941,6 +974,26 @@ send_progress_thread(void *arg)
 	struct tm tm;
 	int err;
 
+	const struct sigaction signal_action =
+	    {.sa_sigaction = send_progress_thread_act, .sa_flags = SA_SIGINFO};
+	struct sigevent timer_cfg =
+	    {.sigev_notify = SIGEV_SIGNAL, .sigev_signo = SIGUSR1};
+	const struct itimerspec timer_time =
+	    {.it_value = {.tv_sec = 1}, .it_interval = {.tv_sec = 1}};
+	struct timer_desirability timer = {};
+
+	sigaction(SIGUSR1, &signal_action, NULL);
+#ifdef SIGINFO
+	sigaction(SIGINFO, &signal_action, NULL);
+#endif
+
+	if ((timer.desired = pa->pa_progress || pa->pa_astitle)) {
+		if (timer_create(CLOCK_MONOTONIC, &timer_cfg, &timer.timer))
+			return ((void *)(uintptr_t)errno);
+		(void) timer_settime(timer.timer, 0, &timer_time, NULL);
+	}
+	pthread_cleanup_push(timer_delete_cleanup, &timer);
+
 	if (!pa->pa_parsable && pa->pa_progress) {
 		(void) fprintf(stderr,
 		    "TIME       %s   %sSNAPSHOT %s\n",
@@ -953,12 +1006,12 @@ send_progress_thread(void *arg)
 	 * Print the progress from ZFS_IOC_SEND_PROGRESS every second.
 	 */
 	for (;;) {
-		(void) sleep(1);
+		pause();
 		if ((err = zfs_send_progress(zhp, pa->pa_fd, &bytes,
 		    &blocks)) != 0) {
 			if (err == EINTR || err == ENOENT)
-				return ((void *)0);
-			return ((void *)(uintptr_t)err);
+				err = 0;
+			pthread_exit(((void *)(uintptr_t)err));
 		}
 
 		(void) time(&t);
@@ -991,21 +1044,25 @@ send_progress_thread(void *arg)
 			(void) fprintf(stderr, "%02d:%02d:%02d\t%llu\t%s\n",
 			    tm.tm_hour, tm.tm_min, tm.tm_sec,
 			    (u_longlong_t)bytes, zhp->zfs_name);
-		} else if (pa->pa_progress) {
+		} else if (pa->pa_progress ||
+		    !send_progress_thread_signal_duetotimer) {
 			zfs_nicebytes(bytes, buf, sizeof (buf));
 			(void) fprintf(stderr, "%02d:%02d:%02d   %5s   %s\n",
 			    tm.tm_hour, tm.tm_min, tm.tm_sec,
 			    buf, zhp->zfs_name);
 		}
 	}
+	pthread_cleanup_pop(B_TRUE);
 }
 
 static boolean_t
-send_progress_thread_exit(libzfs_handle_t *hdl, pthread_t ptid)
+send_progress_thread_exit(
+    libzfs_handle_t *hdl, pthread_t ptid, sigset_t *oldmask)
 {
 	void *status = NULL;
 	(void) pthread_cancel(ptid);
 	(void) pthread_join(ptid, &status);
+	pthread_sigmask(SIG_SETMASK, oldmask, NULL);
 	int error = (int)(uintptr_t)status;
 	if (error != 0 && status != PTHREAD_CANCELED)
 		return (zfs_standard_error(hdl, error,
@@ -1199,7 +1256,8 @@ dump_snapshot(zfs_handle_t *zhp, void *arg)
 		 * If progress reporting is requested, spawn a new thread to
 		 * poll ZFS_IOC_SEND_PROGRESS at a regular interval.
 		 */
-		if (sdd->progress || sdd->progressastitle) {
+		sigset_t oldmask;
+		{
 			pa.pa_zhp = zhp;
 			pa.pa_fd = sdd->outfd;
 			pa.pa_parsable = sdd->parsable;
@@ -1214,13 +1272,13 @@ dump_snapshot(zfs_handle_t *zhp, void *arg)
 				zfs_close(zhp);
 				return (err);
 			}
+			SEND_PROGRESS_THREAD_PARENT_BLOCK(&oldmask);
 		}
 
 		err = dump_ioctl(zhp, sdd->prevsnap, sdd->prevsnap_obj,
 		    fromorigin, sdd->outfd, flags, sdd->debugnv);
 
-		if ((sdd->progress || sdd->progressastitle) &&
-		    send_progress_thread_exit(zhp->zfs_hdl, tid))
+		if (send_progress_thread_exit(zhp->zfs_hdl, tid, &oldmask))
 			return (-1);
 	}
 
@@ -1562,8 +1620,9 @@ estimate_size(zfs_handle_t *zhp, const char *from, int fd, sendflags_t *flags,
 	progress_arg_t pa = { 0 };
 	int err = 0;
 	pthread_t ptid;
+	sigset_t oldmask;
 
-	if (flags->progress || flags->progressastitle) {
+	{
 		pa.pa_zhp = zhp;
 		pa.pa_fd = fd;
 		pa.pa_parsable = flags->parsable;
@@ -1577,6 +1636,7 @@ estimate_size(zfs_handle_t *zhp, const char *from, int fd, sendflags_t *flags,
 			return (zfs_error(zhp->zfs_hdl,
 			    EZFS_THREADCREATEFAILED, errbuf));
 		}
+		SEND_PROGRESS_THREAD_PARENT_BLOCK(&oldmask);
 	}
 
 	err = lzc_send_space_resume_redacted(zhp->zfs_name, from,
@@ -1584,8 +1644,7 @@ estimate_size(zfs_handle_t *zhp, const char *from, int fd, sendflags_t *flags,
 	    redactbook, fd, &size);
 	*sizep = size;
 
-	if ((flags->progress || flags->progressastitle) &&
-	    send_progress_thread_exit(zhp->zfs_hdl, ptid))
+	if (send_progress_thread_exit(zhp->zfs_hdl, ptid, &oldmask))
 		return (-1);
 
 	if (!flags->progress && !flags->parsable)
@@ -1876,11 +1935,12 @@ zfs_send_resume_impl_cb_impl(libzfs_handle_t *hdl, sendflags_t *flags,
 	if (!flags->dryrun) {
 		progress_arg_t pa = { 0 };
 		pthread_t tid;
+		sigset_t oldmask;
 		/*
 		 * If progress reporting is requested, spawn a new thread to
 		 * poll ZFS_IOC_SEND_PROGRESS at a regular interval.
 		 */
-		if (flags->progress || flags->progressastitle) {
+		{
 			pa.pa_zhp = zhp;
 			pa.pa_fd = outfd;
 			pa.pa_parsable = flags->parsable;
@@ -1898,6 +1958,7 @@ zfs_send_resume_impl_cb_impl(libzfs_handle_t *hdl, sendflags_t *flags,
 				zfs_close(zhp);
 				return (error);
 			}
+			SEND_PROGRESS_THREAD_PARENT_BLOCK(&oldmask);
 		}
 
 		error = lzc_send_resume_redacted(zhp->zfs_name, fromname, outfd,
@@ -1905,8 +1966,7 @@ zfs_send_resume_impl_cb_impl(libzfs_handle_t *hdl, sendflags_t *flags,
 		if (redact_book != NULL)
 			free(redact_book);
 
-		if ((flags->progressastitle || flags->progress) &&
-		    send_progress_thread_exit(hdl, tid)) {
+		if (send_progress_thread_exit(hdl, tid, &oldmask)) {
 			zfs_close(zhp);
 			return (-1);
 		}
@@ -2691,7 +2751,8 @@ zfs_send_one_cb_impl(zfs_handle_t *zhp, const char *from, int fd,
 	 * If progress reporting is requested, spawn a new thread to poll
 	 * ZFS_IOC_SEND_PROGRESS at a regular interval.
 	 */
-	if (flags->progress || flags->progressastitle) {
+	sigset_t oldmask;
+	{
 		pa.pa_zhp = zhp;
 		pa.pa_fd = fd;
 		pa.pa_parsable = flags->parsable;
@@ -2708,13 +2769,13 @@ zfs_send_one_cb_impl(zfs_handle_t *zhp, const char *from, int fd,
 			return (zfs_error(zhp->zfs_hdl,
 			    EZFS_THREADCREATEFAILED, errbuf));
 		}
+		SEND_PROGRESS_THREAD_PARENT_BLOCK(&oldmask);
 	}
 
 	err = lzc_send_redacted(name, from, fd,
 	    lzc_flags_from_sendflags(flags), redactbook);
 
-	if ((flags->progress || flags->progressastitle) &&
-	    send_progress_thread_exit(hdl, ptid))
+	if (send_progress_thread_exit(hdl, ptid, &oldmask))
 			return (-1);
 
 	if (err == 0 && (flags->props || flags->holds || flags->backup)) {

lib/libzfs_core/libzfs_core.c

@@ -650,10 +650,12 @@ send_worker(void *arg)
 	unsigned int bufsiz = max_pipe_buffer(ctx->from);
 	ssize_t rd;
 
-	while ((rd = splice(ctx->from, NULL, ctx->to, NULL, bufsiz,
-	    SPLICE_F_MOVE | SPLICE_F_MORE)) > 0)
-		;
-
+	for (;;) {
+		rd = splice(ctx->from, NULL, ctx->to, NULL, bufsiz,
+		    SPLICE_F_MOVE | SPLICE_F_MORE);
+		if ((rd == -1 && errno != EINTR) || rd == 0)
+			break;
+	}
 	int err = (rd == -1) ? errno : 0;
 	close(ctx->from);
 	return ((void *)(uintptr_t)err);

man/Makefile.am

@@ -38,7 +38,6 @@ dist_man_MANS = \
 	%D%/man8/zfs-groupspace.8 \
 	%D%/man8/zfs-hold.8 \
 	%D%/man8/zfs-inherit.8 \
-	%D%/man8/zfs-jail.8 \
 	%D%/man8/zfs-list.8 \
 	%D%/man8/zfs-load-key.8 \
 	%D%/man8/zfs-mount.8 \
@@ -57,14 +56,11 @@ dist_man_MANS = \
 	%D%/man8/zfs-share.8 \
 	%D%/man8/zfs-snapshot.8 \
 	%D%/man8/zfs-unallow.8 \
-	%D%/man8/zfs-unjail.8 \
 	%D%/man8/zfs-unload-key.8 \
 	%D%/man8/zfs-unmount.8 \
-	%D%/man8/zfs-unzone.8 \
 	%D%/man8/zfs-upgrade.8 \
 	%D%/man8/zfs-userspace.8 \
 	%D%/man8/zfs-wait.8 \
-	%D%/man8/zfs-zone.8 \
 	%D%/man8/zfs_ids_to_path.8 \
 	%D%/man8/zgenhostid.8 \
 	%D%/man8/zinject.8 \
@@ -104,6 +100,18 @@ dist_man_MANS = \
 	%D%/man8/zstreamdump.8 \
 	%D%/man8/zpool_influxdb.8
 
+if BUILD_FREEBSD
+dist_man_MANS += \
+	%D%/man8/zfs-jail.8 \
+	%D%/man8/zfs-unjail.8
+endif
+
+if BUILD_LINUX
+dist_man_MANS += \
+	%D%/man8/zfs-unzone.8 \
+	%D%/man8/zfs-zone.8
+endif
+
 nodist_man_MANS = \
 	%D%/man8/zed.8 \
 	%D%/man8/zfs-mount-generator.8

man/man4/zfs.4

@@ -15,7 +15,7 @@
 .\" own identifying information:
 .\" Portions Copyright [yyyy] [name of copyright owner]
 .\"
-.Dd January 10, 2023
+.Dd July 21, 2023
 .Dt ZFS 4
 .Os
 .
@@ -239,6 +239,11 @@ relative to the pool.
 Make some blocks above a certain size be gang blocks.
 This option is used by the test suite to facilitate testing.
 .
+.It Sy metaslab_force_ganging_pct Ns = Ns Sy 3 Ns % Pq uint
+For blocks that could be forced to be a gang block (due to
+.Sy metaslab_force_ganging ) ,
+force this many of them to be gang blocks.
+.
 .It Sy zfs_ddt_zap_default_bs Ns = Ns Sy 15 Po 32 KiB Pc Pq int
 Default DDT ZAP data block size as a power of 2. Note that changing this after
 creating a DDT on the pool will not affect existing DDTs, only newly created
@@ -397,6 +402,12 @@ Practical upper limit of total metaslabs per top-level vdev.
 .It Sy metaslab_preload_enabled Ns = Ns Sy 1 Ns | Ns 0 Pq int
 Enable metaslab group preloading.
 .
+.It Sy metaslab_preload_limit Ns = Ns Sy 10 Pq uint
+Maximum number of metaslabs per group to preload
+.
+.It Sy metaslab_preload_pct Ns = Ns Sy 50 Pq uint
+Percentage of CPUs to run a metaslab preload taskq
+.
 .It Sy metaslab_lba_weighting_enabled Ns = Ns Sy 1 Ns | Ns 0 Pq int
 Give more weight to metaslabs with lower LBAs,
 assuming they have greater bandwidth,
@@ -519,9 +530,6 @@ However, this is limited by
 Maximum micro ZAP size.
 A micro ZAP is upgraded to a fat ZAP, once it grows beyond the specified size.
 .
-.It Sy zfetch_array_rd_sz Ns = Ns Sy 1048576 Ns B Po 1 MiB Pc Pq u64
-If prefetching is enabled, disable prefetching for reads larger than this size.
-.
 .It Sy zfetch_min_distance Ns = Ns Sy 4194304 Ns B Po 4 MiB Pc Pq uint
 Min bytes to prefetch per stream.
 Prefetch distance starts from the demand access size and quickly grows to
@@ -2142,6 +2150,11 @@ On very fragmented pools, lowering this
 .Pq typically to Sy 36 KiB
 can improve performance.
 .
+.It Sy zil_maxcopied Ns = Ns Sy 7680 Ns B Po 7.5 KiB Pc Pq uint
+This sets the maximum number of write bytes logged via WR_COPIED.
+It tunes a tradeoff between additional memory copy and possibly worse log
+space efficiency vs additional range lock/unlock.
+.
 .It Sy zil_min_commit_timeout Ns = Ns Sy 5000 Pq u64
 This sets the minimum delay in nanoseconds ZIL care to delay block commit,
 waiting for more records.

man/man7/zfsconcepts.7

@@ -28,8 +28,9 @@
 .\" Copyright 2019 Richard Laager. All rights reserved.
 .\" Copyright 2018 Nexenta Systems, Inc.
 .\" Copyright 2019 Joyent, Inc.
+.\" Copyright 2023 Klara, Inc.
 .\"
-.Dd June 30, 2019
+.Dd October 6, 2023
 .Dt ZFSCONCEPTS 7
 .Os
 .
@@ -205,3 +206,40 @@ practices, such as regular backups.
 Consider using the
 .Sy compression
 property as a less resource-intensive alternative.
+.Ss Block cloning
+Block cloning is a facility that allows a file (or parts of a file) to be
+.Qq cloned ,
+that is, a shallow copy made where the existing data blocks are referenced
+rather than copied.
+Later modifications to the data will cause a copy of the data block to be taken
+and that copy modified.
+This facility is used to implement
+.Qq reflinks
+or
+.Qq file-level copy-on-write .
+.Pp
+Cloned blocks are tracked in a special on-disk structure called the Block
+Reference Table
+.Po BRT
+.Pc .
+Unlike deduplication, this table has minimal overhead, so can be enabled at all
+times.
+.Pp
+Also unlike deduplication, cloning must be requested by a user program.
+Many common file copying programs, including newer versions of
+.Nm /bin/cp ,
+will try to create clones automatically.
+Look for
+.Qq clone ,
+.Qq dedupe
+or
+.Qq reflink
+in the documentation for more information.
+.Pp
+There are some limitations to block cloning.
+Only whole blocks can be cloned, and blocks can not be cloned if they are not
+yet written to disk, or if they are encrypted, or the source and destination
+.Sy recordsize
+properties differ.
+The OS may add additional restrictions;
+for example, most versions of Linux will not allow clones across datasets.

man/man7/zfsprops.7

@@ -38,7 +38,7 @@
 .\" Copyright (c) 2019, Kjeld Schouten-Lebbing
 .\" Copyright (c) 2022 Hewlett Packard Enterprise Development LP.
 .\"
-.Dd April 18, 2023
+.Dd August 8, 2023
 .Dt ZFSPROPS 7
 .Os
 .
@@ -1248,10 +1248,18 @@ Otherwise, they are automatically remounted in the new location if the property
 was previously
 .Sy legacy
 or
-.Sy none ,
-or if they were mounted before the property was changed.
+.Sy none .
 In addition, any shared file systems are unshared and shared in the new
 location.
+.Pp
+When the
+.Sy mountpoint
+property is set with
+.Nm zfs Cm set Fl u
+, the
+.Sy mountpoint
+property is updated but dataset is not mounted or unmounted and remains
+as it was before.
 .It Sy nbmand Ns = Ns Sy on Ns | Ns Sy off
 Controls whether the file system should be mounted with
 .Sy nbmand
@@ -1656,6 +1664,13 @@ by default.
 This means that any additional access control
 (disallow specific user specific access etc) must be done on the underlying file
 system.
+.Pp
+When the
+.Sy sharesmb
+property is updated with
+.Nm zfs Cm set Fl u
+, the property is set to desired value, but the operation to share, reshare
+or unshare the the dataset is not performed.
 .It Sy sharenfs Ns = Ns Sy on Ns | Ns Sy off Ns | Ns Ar opts
 Controls whether the file system is shared via NFS, and what options are to be
 used.
@@ -1699,6 +1714,13 @@ or if they were shared before the property was changed.
 If the new property is
 .Sy off ,
 the file systems are unshared.
+.Pp
+When the
+.Sy sharenfs
+property is updated with
+.Nm zfs Cm set Fl u
+, the property is set to desired value, but the operation to share, reshare
+or unshare the the dataset is not performed.
 .It Sy logbias Ns = Ns Sy latency Ns | Ns Sy throughput
 Provide a hint to ZFS about handling of synchronous requests in this dataset.
 If
@@ -1916,13 +1938,15 @@ See
 for more information.
 Jails are a
 .Fx
-feature and are not relevant on other platforms.
-The default value is
-.Sy off .
-.It Sy zoned Ns = Ns Sy on Ns | Ns Sy off
+feature and this property is not available on other platforms.
+.It Sy zoned Ns = Ns Sy off Ns | Ns Sy on
 Controls whether the dataset is managed from a non-global zone or namespace.
-The default value is
-.Sy off .
+See
+.Xr zfs-zone 8
+for more information.
+Zoning is a
+Linux
+feature and this property is not available on other platforms.
 .El
 .Pp
 The following three properties cannot be changed after the file system is

man/man7/zpoolconcepts.7

@@ -203,11 +203,9 @@ For more information, see the
 section.
 .El
 .Pp
-Virtual devices cannot be nested, so a mirror or raidz virtual device can only
-contain files or disks.
-Mirrors of mirrors
-.Pq or other combinations
-are not allowed.
+Virtual devices cannot be nested arbitrarily.
+A mirror, raidz or draid virtual device can only be created with files or disks.
+Mirrors of mirrors or other such combinations are not allowed.
 .Pp
 A pool can have any number of virtual devices at the top of the configuration
 .Po known as

man/man8/zfs-send.8

@@ -29,7 +29,7 @@
 .\" Copyright 2018 Nexenta Systems, Inc.
 .\" Copyright 2019 Joyent, Inc.
 .\"
-.Dd January 12, 2023
+.Dd July 27, 2023
 .Dt ZFS-SEND 8
 .Os
 .
@@ -297,6 +297,12 @@ This flag can only be used in conjunction with
 .It Fl v , -verbose
 Print verbose information about the stream package generated.
 This information includes a per-second report of how much data has been sent.
+The same report can be requested by sending
+.Dv SIGINFO
+or
+.Dv SIGUSR1 ,
+regardless of
+.Fl v .
 .Pp
 The format of the stream is committed.
 You will be able to receive your streams on future versions of ZFS.
@@ -433,6 +439,12 @@ and the verbose output goes to standard error
 .It Fl v , -verbose
 Print verbose information about the stream package generated.
 This information includes a per-second report of how much data has been sent.
+The same report can be requested by sending
+.Dv SIGINFO
+or
+.Dv SIGUSR1 ,
+regardless of
+.Fl v .
 .El
 .It Xo
 .Nm zfs
@@ -669,6 +681,10 @@ ones on the source, and are ready to be used, while the parent snapshot on the
 target contains none of the username and password data present on the source,
 because it was removed by the redacted send operation.
 .
+.Sh SIGNALS
+See
+.Fl v .
+.
 .Sh EXAMPLES
 .\" These are, respectively, examples 12, 13 from zfs.8
 .\" Make sure to update them bidirectionally

man/man8/zfs-set.8

@@ -39,6 +39,7 @@
 .Sh SYNOPSIS
 .Nm zfs
 .Cm set
+.Op Fl u
 .Ar property Ns = Ns Ar value Oo Ar property Ns = Ns Ar value Oc Ns …
 .Ar filesystem Ns | Ns Ar volume Ns | Ns Ar snapshot Ns …
 .Nm zfs
@@ -60,6 +61,7 @@
 .It Xo
 .Nm zfs
 .Cm set
+.Op Fl u
 .Ar property Ns = Ns Ar value Oo Ar property Ns = Ns Ar value Oc Ns …
 .Ar filesystem Ns | Ns Ar volume Ns | Ns Ar snapshot Ns …
 .Xc
@@ -79,6 +81,11 @@ For more information, see the
 .Em User Properties
 section of
 .Xr zfsprops 7 .
+.Bl -tag -width "-u"
+.It Fl u
+Update mountpoint, sharenfs, sharesmb property but do not mount or share the
+dataset.
+.El
 .It Xo
 .Nm zfs
 .Cm get

man/man8/zpool-events.8

@@ -26,7 +26,7 @@
 .\" Copyright 2017 Nexenta Systems, Inc.
 .\" Copyright (c) 2017 Open-E, Inc. All Rights Reserved.
 .\"
-.Dd May 27, 2021
+.Dd July 11, 2023
 .Dt ZPOOL-EVENTS 8
 .Os
 .
@@ -305,10 +305,6 @@ The time when a given I/O request was submitted.
 The time required to service a given I/O request.
 .It Sy prev_state
 The previous state of the vdev.
-.It Sy cksum_expected
-The expected checksum value for the block.
-.It Sy cksum_actual
-The actual checksum value for an errant block.
 .It Sy cksum_algorithm
 Checksum algorithm used.
 See
@@ -362,23 +358,6 @@ Like
 but contains
 .Pq Ar good data No & ~( Ns Ar bad data ) ;
 that is, the bits set in the good data which are cleared in the bad data.
-.It Sy bad_set_histogram
-If this field exists, it is an array of counters.
-Each entry counts bits set in a particular bit of a big-endian uint64 type.
-The first entry counts bits
-set in the high-order bit of the first byte, the 9th byte, etc, and the last
-entry counts bits set of the low-order bit of the 8th byte, the 16th byte, etc.
-This information is useful for observing a stuck bit in a parallel data path,
-such as IDE or parallel SCSI.
-.It Sy bad_cleared_histogram
-If this field exists, it is an array of counters.
-Each entry counts bit clears in a particular bit of a big-endian uint64 type.
-The first entry counts bits
-clears of the high-order bit of the first byte, the 9th byte, etc, and the
-last entry counts clears of the low-order bit of the 8th byte, the 16th byte,
-etc.
-This information is useful for observing a stuck bit in a parallel data
-path, such as IDE or parallel SCSI.
 .El
 .
 .Sh I/O STAGES

man/man8/zpool.8

@@ -110,9 +110,10 @@ Removes ZFS label information from the specified
 .It Xo
 .Xr zpool-attach 8 Ns / Ns Xr zpool-detach 8
 .Xc
-Increases or decreases redundancy by
-.Cm attach Ns ing or
-.Cm detach Ns ing a device on an existing vdev (virtual device).
+Converts a non-redundant disk into a mirror, or increases
+the redundancy level of an existing mirror
+.Cm ( attach Ns ), or performs the inverse operation (
+.Cm detach Ns ).
 .It Xo
 .Xr zpool-add 8 Ns / Ns Xr zpool-remove 8
 .Xc
@@ -233,16 +234,16 @@ Invalid command line options were specified.
 .El
 .
 .Sh EXAMPLES
-.\" Examples 1, 2, 3, 4, 11, 12 are shared with zpool-create.8.
-.\" Examples 5, 13 are shared with zpool-add.8.
-.\" Examples 6, 15 are shared with zpool-list.8.
-.\" Examples 7 are shared with zpool-destroy.8.
-.\" Examples 8 are shared with zpool-export.8.
-.\" Examples 9 are shared with zpool-import.8.
-.\" Examples 10 are shared with zpool-upgrade.8.
-.\" Examples 14 are shared with zpool-remove.8.
-.\" Examples 16 are shared with zpool-status.8.
-.\" Examples 13, 16 are also shared with zpool-iostat.8.
+.\" Examples 1, 2, 3, 4, 12, 13 are shared with zpool-create.8.
+.\" Examples 6, 14 are shared with zpool-add.8.
+.\" Examples 7, 16 are shared with zpool-list.8.
+.\" Examples 8 are shared with zpool-destroy.8.
+.\" Examples 9 are shared with zpool-export.8.
+.\" Examples 10 are shared with zpool-import.8.
+.\" Examples 11 are shared with zpool-upgrade.8.
+.\" Examples 15 are shared with zpool-remove.8.
+.\" Examples 17 are shared with zpool-status.8.
+.\" Examples 14, 17 are also shared with zpool-iostat.8.
 .\" Make sure to update them omnidirectionally
 .Ss Example 1 : No Creating a RAID-Z Storage Pool
 The following command creates a pool with a single raidz root vdev that
@@ -264,14 +265,21 @@ While not recommended, a pool based on files can be useful for experimental
 purposes.
 .Dl # Nm zpool Cm create Ar tank Pa /path/to/file/a /path/to/file/b
 .
-.Ss Example 5 : No Adding a Mirror to a ZFS Storage Pool
+.Ss Example 5 : No Making a non-mirrored ZFS Storage Pool mirrored
+The following command converts an existing single device
+.Ar sda
+into a mirror by attaching a second device to it,
+.Ar sdb .
+.Dl # Nm zpool Cm attach Ar tank Pa sda sdb
+.
+.Ss Example 6 : No Adding a Mirror to a ZFS Storage Pool
 The following command adds two mirrored disks to the pool
 .Ar tank ,
 assuming the pool is already made up of two-way mirrors.
 The additional space is immediately available to any datasets within the pool.
 .Dl # Nm zpool Cm add Ar tank Sy mirror Pa sda sdb
 .
-.Ss Example 6 : No Listing Available ZFS Storage Pools
+.Ss Example 7 : No Listing Available ZFS Storage Pools
 The following command lists all available pools on the system.
 In this case, the pool
 .Ar zion
@@ -285,19 +293,19 @@ tank   61.5G  20.0G  41.5G         -    48%    32%  1.00x  ONLINE  -
 zion       -      -      -         -      -      -      -  FAULTED -
 .Ed
 .
-.Ss Example 7 : No Destroying a ZFS Storage Pool
+.Ss Example 8 : No Destroying a ZFS Storage Pool
 The following command destroys the pool
 .Ar tank
 and any datasets contained within:
 .Dl # Nm zpool Cm destroy Fl f Ar tank
 .
-.Ss Example 8 : No Exporting a ZFS Storage Pool
+.Ss Example 9 : No Exporting a ZFS Storage Pool
 The following command exports the devices in pool
 .Ar tank
 so that they can be relocated or later imported:
 .Dl # Nm zpool Cm export Ar tank
 .
-.Ss Example 9 : No Importing a ZFS Storage Pool
+.Ss Example 10 : No Importing a ZFS Storage Pool
 The following command displays available pools, and then imports the pool
 .Ar tank
 for use on the system.
@@ -318,7 +326,7 @@ config:
 .No # Nm zpool Cm import Ar tank
 .Ed
 .
-.Ss Example 10 : No Upgrading All ZFS Storage Pools to the Current Version
+.Ss Example 11 : No Upgrading All ZFS Storage Pools to the Current Version
 The following command upgrades all ZFS Storage pools to the current version of
 the software:
 .Bd -literal -compact -offset Ds
@@ -326,7 +334,7 @@ the software:
 This system is currently running ZFS version 2.
 .Ed
 .
-.Ss Example 11 : No Managing Hot Spares
+.Ss Example 12 : No Managing Hot Spares
 The following command creates a new pool with an available hot spare:
 .Dl # Nm zpool Cm create Ar tank Sy mirror Pa sda sdb Sy spare Pa sdc
 .Pp
@@ -341,12 +349,12 @@ The hot spare can be permanently removed from the pool using the following
 command:
 .Dl # Nm zpool Cm remove Ar tank Pa sdc
 .
-.Ss Example 12 : No Creating a ZFS Pool with Mirrored Separate Intent Logs
+.Ss Example 13 : No Creating a ZFS Pool with Mirrored Separate Intent Logs
 The following command creates a ZFS storage pool consisting of two, two-way
 mirrors and mirrored log devices:
 .Dl # Nm zpool Cm create Ar pool Sy mirror Pa sda sdb Sy mirror Pa sdc sdd Sy log mirror Pa sde sdf
 .
-.Ss Example 13 : No Adding Cache Devices to a ZFS Pool
+.Ss Example 14 : No Adding Cache Devices to a ZFS Pool
 The following command adds two disks for use as cache devices to a ZFS storage
 pool:
 .Dl # Nm zpool Cm add Ar pool Sy cache Pa sdc sdd
@@ -359,7 +367,7 @@ Capacity and reads can be monitored using the
 subcommand as follows:
 .Dl # Nm zpool Cm iostat Fl v Ar pool 5
 .
-.Ss Example 14 : No Removing a Mirrored top-level (Log or Data) Device
+.Ss Example 15 : No Removing a Mirrored top-level (Log or Data) Device
 The following commands remove the mirrored log device
 .Sy mirror-2
 and mirrored top-level data device
@@ -394,7 +402,7 @@ The command to remove the mirrored data
 .Ar mirror-1 No is :
 .Dl # Nm zpool Cm remove Ar tank mirror-1
 .
-.Ss Example 15 : No Displaying expanded space on a device
+.Ss Example 16 : No Displaying expanded space on a device
 The following command displays the detailed information for the pool
 .Ar data .
 This pool is comprised of a single raidz vdev where one of its devices
@@ -411,7 +419,7 @@ data        23.9G  14.6G  9.30G         -    48%    61%  1.00x  ONLINE  -
     sdc         -      -      -         -      -
 .Ed
 .
-.Ss Example 16 : No Adding output columns
+.Ss Example 17 : No Adding output columns
 Additional columns can be added to the
 .Nm zpool Cm status No and Nm zpool Cm iostat No output with Fl c .
 .Bd -literal -compact -offset Ds

module/Kbuild.in

@@ -461,6 +461,7 @@ ZFS_OBJS_OS := \
 	zpl_ctldir.o \
 	zpl_export.o \
 	zpl_file.o \
+	zpl_file_range.o \
 	zpl_inode.o \
 	zpl_super.o \
 	zpl_xattr.o \

module/Makefile.in

@@ -168,4 +168,4 @@ gen-zstd-symbols:
 	for obj in $(addprefix zstd/,$(ZSTD_UPSTREAM_OBJS)); do echo; echo "/* $${obj#zstd/}: */"; @OBJDUMP@ -t $$obj | awk '$$2 == "g" && !/ zfs_/ {print "#define\t" $$6 " zfs_" $$6}' | sort; done >> zstd/include/zstd_compat_wrapper.h
 
 check-zstd-symbols:
-	@OBJDUMP@ -t $(addprefix zstd/,$(ZSTD_UPSTREAM_OBJS)) | awk '/file format/ {print}  $$2 == "g" && !/ zfs_/ {++ret; print}  END {exit ret}'
+	@OBJDUMP@ -t $(addprefix zstd/,$(ZSTD_UPSTREAM_OBJS)) | awk '/file format/ {print}  $$2 == "g" && (!/ zfs_/ && !/ __pfx_zfs_/) {++ret; print}  END {exit ret}'

module/icp/asm-aarch64/sha2/sha256-armv8.S

@@ -49,6 +49,7 @@
 .type	zfs_sha256_block_armv7,%function
 .align	6
 zfs_sha256_block_armv7:
+	hint	#34					// bti c
 	stp	x29,x30,[sp,#-128]!
 	add	x29,sp,#0
 
@@ -1015,6 +1016,7 @@ zfs_sha256_block_armv7:
 .type	zfs_sha256_block_armv8,%function
 .align	6
 zfs_sha256_block_armv8:
+	hint		#34				// bti c
 .Lv8_entry:
 	stp		x29,x30,[sp,#-16]!
 	add		x29,sp,#0
@@ -1155,6 +1157,7 @@ zfs_sha256_block_armv8:
 .type	zfs_sha256_block_neon,%function
 .align	4
 zfs_sha256_block_neon:
+	hint	#34					// bti c
 .Lneon_entry:
 	stp	x29, x30, [sp, #-16]!
 	mov	x29, sp

module/icp/asm-aarch64/sha2/sha512-armv8.S

@@ -73,6 +73,7 @@
 .type	zfs_sha512_block_armv7,%function
 .align	6
 zfs_sha512_block_armv7:
+	hint	#34					// bti c
 	stp	x29,x30,[sp,#-128]!
 	add	x29,sp,#0
 
@@ -1040,6 +1041,7 @@ zfs_sha512_block_armv7:
 .type	zfs_sha512_block_armv8,%function
 .align	6
 zfs_sha512_block_armv8:
+	hint		#34				// bti c
 .Lv8_entry:
 	// Armv8.3-A PAuth: even though x30 is pushed to stack it is not popped later
 	stp		x29,x30,[sp,#-16]!

module/os/freebsd/zfs/sysctl_os.c

@@ -596,28 +596,6 @@ SYSCTL_UINT(_vfs_zfs_metaslab, OID_AUTO, df_free_pct,
 	" space map to continue allocations in a first-fit fashion");
 /* END CSTYLED */
 
-/*
- * Percentage of all cpus that can be used by the metaslab taskq.
- */
-extern int metaslab_load_pct;
-
-/* BEGIN CSTYLED */
-SYSCTL_INT(_vfs_zfs_metaslab, OID_AUTO, load_pct,
-	CTLFLAG_RWTUN, &metaslab_load_pct, 0,
-	"Percentage of cpus that can be used by the metaslab taskq");
-/* END CSTYLED */
-
-/*
- * Max number of metaslabs per group to preload.
- */
-extern uint_t metaslab_preload_limit;
-
-/* BEGIN CSTYLED */
-SYSCTL_UINT(_vfs_zfs_metaslab, OID_AUTO, preload_limit,
-	CTLFLAG_RWTUN, &metaslab_preload_limit, 0,
-	"Max number of metaslabs per group to preload");
-/* END CSTYLED */
-
 /* mmp.c */
 
 int

module/os/freebsd/zfs/zfs_vfsops.c

@@ -1154,7 +1154,6 @@ zfsvfs_free(zfsvfs_t *zfsvfs)
 
 	mutex_destroy(&zfsvfs->z_znodes_lock);
 	mutex_destroy(&zfsvfs->z_lock);
-	ASSERT3U(zfsvfs->z_nr_znodes, ==, 0);
 	list_destroy(&zfsvfs->z_all_znodes);
 	ZFS_TEARDOWN_DESTROY(zfsvfs);
 	ZFS_TEARDOWN_INACTIVE_DESTROY(zfsvfs);
@@ -1558,12 +1557,11 @@ zfsvfs_teardown(zfsvfs_t *zfsvfs, boolean_t unmounting)
 		 * may add the parents of dir-based xattrs to the taskq
 		 * so we want to wait for these.
 		 *
-		 * We can safely read z_nr_znodes without locking because the
-		 * VFS has already blocked operations which add to the
-		 * z_all_znodes list and thus increment z_nr_znodes.
+		 * We can safely check z_all_znodes for being empty because the
+		 * VFS has already blocked operations which add to it.
 		 */
 		int round = 0;
-		while (zfsvfs->z_nr_znodes > 0) {
+		while (!list_is_empty(&zfsvfs->z_all_znodes)) {
 			taskq_wait_outstanding(dsl_pool_zrele_taskq(
 			    dmu_objset_pool(zfsvfs->z_os)), 0);
 			if (++round > 1 && !unmounting)

module/os/freebsd/zfs/zfs_vnops_os.c

@@ -6263,7 +6263,8 @@ zfs_freebsd_copy_file_range(struct vop_copy_file_range_args *ap)
 			goto bad_write_fallback;
 		}
 	} else {
-#if __FreeBSD_version >= 1400086
+#if (__FreeBSD_version >= 1302506 && __FreeBSD_version < 1400000) || \
+	__FreeBSD_version >= 1400086
 		vn_lock_pair(invp, false, LK_EXCLUSIVE, outvp, false,
 		    LK_EXCLUSIVE);
 #else
@@ -6289,7 +6290,8 @@ zfs_freebsd_copy_file_range(struct vop_copy_file_range_args *ap)
 
 	error = zfs_clone_range(VTOZ(invp), ap->a_inoffp, VTOZ(outvp),
 	    ap->a_outoffp, &len, ap->a_outcred);
-	if (error == EXDEV)
+	if (error == EXDEV || error == EAGAIN || error == EINVAL ||
+	    error == EOPNOTSUPP)
 		goto bad_locked_fallback;
 	*ap->a_lenp = (size_t)len;
 out_locked:

module/os/freebsd/zfs/zfs_znode.c

@@ -537,7 +537,6 @@ zfs_znode_alloc(zfsvfs_t *zfsvfs, dmu_buf_t *db, int blksz,
 
 	mutex_enter(&zfsvfs->z_znodes_lock);
 	list_insert_tail(&zfsvfs->z_all_znodes, zp);
-	zfsvfs->z_nr_znodes++;
 	zp->z_zfsvfs = zfsvfs;
 	mutex_exit(&zfsvfs->z_znodes_lock);
 
@@ -1286,7 +1285,6 @@ zfs_znode_free(znode_t *zp)
 	mutex_enter(&zfsvfs->z_znodes_lock);
 	POINTER_INVALIDATE(&zp->z_zfsvfs);
 	list_remove(&zfsvfs->z_all_znodes, zp);
-	zfsvfs->z_nr_znodes--;
 	mutex_exit(&zfsvfs->z_znodes_lock);
 
 #if __FreeBSD_version >= 1300139

module/os/linux/spl/spl-proc.c

@@ -47,6 +47,10 @@ static unsigned long table_min = 0;
 static unsigned long table_max = ~0;
 
 static struct ctl_table_header *spl_header = NULL;
+#ifndef HAVE_REGISTER_SYSCTL_TABLE
+static struct ctl_table_header *spl_kmem = NULL;
+static struct ctl_table_header *spl_kstat = NULL;
+#endif
 static struct proc_dir_entry *proc_spl = NULL;
 static struct proc_dir_entry *proc_spl_kmem = NULL;
 static struct proc_dir_entry *proc_spl_kmem_slab = NULL;
@@ -624,6 +628,7 @@ static struct ctl_table spl_table[] = {
 		.mode		= 0644,
 		.proc_handler	= &proc_dohostid,
 	},
+#ifdef HAVE_REGISTER_SYSCTL_TABLE
 	{
 		.procname	= "kmem",
 		.mode		= 0555,
@@ -634,9 +639,11 @@ static struct ctl_table spl_table[] = {
 		.mode		= 0555,
 		.child		= spl_kstat_table,
 	},
+#endif
 	{},
 };
 
+#ifdef HAVE_REGISTER_SYSCTL_TABLE
 static struct ctl_table spl_dir[] = {
 	{
 		.procname	= "spl",
@@ -654,15 +661,58 @@ static struct ctl_table spl_root[] = {
 	},
 	{}
 };
+#endif
+
+static void spl_proc_cleanup(void)
+{
+	remove_proc_entry("kstat", proc_spl);
+	remove_proc_entry("slab", proc_spl_kmem);
+	remove_proc_entry("kmem", proc_spl);
+	remove_proc_entry("taskq-all", proc_spl);
+	remove_proc_entry("taskq", proc_spl);
+	remove_proc_entry("spl", NULL);
+
+#ifndef HAVE_REGISTER_SYSCTL_TABLE
+	if (spl_kstat) {
+		unregister_sysctl_table(spl_kstat);
+		spl_kstat = NULL;
+	}
+	if (spl_kmem) {
+		unregister_sysctl_table(spl_kmem);
+		spl_kmem = NULL;
+	}
+#endif
+	if (spl_header) {
+		unregister_sysctl_table(spl_header);
+		spl_header = NULL;
+	}
+}
 
 int
 spl_proc_init(void)
 {
 	int rc = 0;
 
+#ifdef HAVE_REGISTER_SYSCTL_TABLE
 	spl_header = register_sysctl_table(spl_root);
 	if (spl_header == NULL)
 		return (-EUNATCH);
+#else
+	spl_header = register_sysctl("kernel/spl", spl_table);
+	if (spl_header == NULL)
+		return (-EUNATCH);
+
+	spl_kmem = register_sysctl("kernel/spl/kmem", spl_kmem_table);
+	if (spl_kmem == NULL) {
+		rc = -EUNATCH;
+		goto out;
+	}
+	spl_kstat = register_sysctl("kernel/spl/kstat", spl_kstat_table);
+	if (spl_kstat == NULL) {
+		rc = -EUNATCH;
+		goto out;
+	}
+#endif
 
 	proc_spl = proc_mkdir("spl", NULL);
 	if (proc_spl == NULL) {
@@ -703,15 +753,8 @@ spl_proc_init(void)
 		goto out;
 	}
 out:
-	if (rc) {
-		remove_proc_entry("kstat", proc_spl);
-		remove_proc_entry("slab", proc_spl_kmem);
-		remove_proc_entry("kmem", proc_spl);
-		remove_proc_entry("taskq-all", proc_spl);
-		remove_proc_entry("taskq", proc_spl);
-		remove_proc_entry("spl", NULL);
-		unregister_sysctl_table(spl_header);
-	}
+	if (rc)
+		spl_proc_cleanup();
 
 	return (rc);
 }
@@ -719,13 +762,5 @@ out:
 void
 spl_proc_fini(void)
 {
-	remove_proc_entry("kstat", proc_spl);
-	remove_proc_entry("slab", proc_spl_kmem);
-	remove_proc_entry("kmem", proc_spl);
-	remove_proc_entry("taskq-all", proc_spl);
-	remove_proc_entry("taskq", proc_spl);
-	remove_proc_entry("spl", NULL);
-
-	ASSERT(spl_header != NULL);
-	unregister_sysctl_table(spl_header);
+	spl_proc_cleanup();
 }

module/os/linux/zfs/qat_compress.c

@@ -193,7 +193,9 @@ qat_dc_init(void)
 		sd.huffType = CPA_DC_HT_FULL_DYNAMIC;
 		sd.sessDirection = CPA_DC_DIR_COMBINED;
 		sd.sessState = CPA_DC_STATELESS;
+#if (CPA_DC_API_VERSION_NUM_MAJOR == 1 && CPA_DC_API_VERSION_NUM_MINOR < 6)
 		sd.deflateWindowSize = 7;
+#endif
 		sd.checksum = CPA_DC_ADLER32;
 		status = cpaDcGetSessionSize(dc_inst_handles[i],
 		    &sd, &sess_size, &ctx_size);

module/os/linux/zfs/vdev_disk.c

@@ -80,9 +80,22 @@ typedef struct dio_request {
 
 static unsigned int zfs_vdev_failfast_mask = 1;
 
+#ifdef HAVE_BLK_MODE_T
+static blk_mode_t
+#else
 static fmode_t
+#endif
 vdev_bdev_mode(spa_mode_t spa_mode)
 {
+#ifdef HAVE_BLK_MODE_T
+	blk_mode_t mode = 0;
+
+	if (spa_mode & SPA_MODE_READ)
+		mode |= BLK_OPEN_READ;
+
+	if (spa_mode & SPA_MODE_WRITE)
+		mode |= BLK_OPEN_WRITE;
+#else
 	fmode_t mode = 0;
 
 	if (spa_mode & SPA_MODE_READ)
@@ -90,6 +103,7 @@ vdev_bdev_mode(spa_mode_t spa_mode)
 
 	if (spa_mode & SPA_MODE_WRITE)
 		mode |= FMODE_WRITE;
+#endif
 
 	return (mode);
 }
@@ -197,12 +211,47 @@ vdev_disk_kobj_evt_post(vdev_t *v)
 	}
 }
 
+#if !defined(HAVE_BLKDEV_GET_BY_PATH_4ARG)
+/*
+ * Define a dummy struct blk_holder_ops for kernel versions
+ * prior to 6.5.
+ */
+struct blk_holder_ops {};
+#endif
+
+static struct block_device *
+vdev_blkdev_get_by_path(const char *path, spa_mode_t mode, void *holder,
+    const struct blk_holder_ops *hops)
+{
+#ifdef HAVE_BLKDEV_GET_BY_PATH_4ARG
+	return (blkdev_get_by_path(path,
+	    vdev_bdev_mode(mode) | BLK_OPEN_EXCL, holder, hops));
+#else
+	return (blkdev_get_by_path(path,
+	    vdev_bdev_mode(mode) | FMODE_EXCL, holder));
+#endif
+}
+
+static void
+vdev_blkdev_put(struct block_device *bdev, spa_mode_t mode, void *holder)
+{
+#ifdef HAVE_BLKDEV_PUT_HOLDER
+	return (blkdev_put(bdev, holder));
+#else
+	return (blkdev_put(bdev, vdev_bdev_mode(mode) | FMODE_EXCL));
+#endif
+}
+
 static int
 vdev_disk_open(vdev_t *v, uint64_t *psize, uint64_t *max_psize,
     uint64_t *logical_ashift, uint64_t *physical_ashift)
 {
 	struct block_device *bdev;
+#ifdef HAVE_BLK_MODE_T
+	blk_mode_t mode = vdev_bdev_mode(spa_mode(v->vdev_spa));
+#else
 	fmode_t mode = vdev_bdev_mode(spa_mode(v->vdev_spa));
+#endif
 	hrtime_t timeout = MSEC2NSEC(zfs_vdev_open_timeout_ms);
 	vdev_disk_t *vd;
 
@@ -252,15 +301,15 @@ vdev_disk_open(vdev_t *v, uint64_t *psize, uint64_t *max_psize,
 					reread_part = B_TRUE;
 			}
 
-			blkdev_put(bdev, mode | FMODE_EXCL);
+			vdev_blkdev_put(bdev, mode, zfs_vdev_holder);
 		}
 
 		if (reread_part) {
-			bdev = blkdev_get_by_path(disk_name, mode | FMODE_EXCL,
-			    zfs_vdev_holder);
+			bdev = vdev_blkdev_get_by_path(disk_name, mode,
+			    zfs_vdev_holder, NULL);
 			if (!IS_ERR(bdev)) {
 				int error = vdev_bdev_reread_part(bdev);
-				blkdev_put(bdev, mode | FMODE_EXCL);
+				vdev_blkdev_put(bdev, mode, zfs_vdev_holder);
 				if (error == 0) {
 					timeout = MSEC2NSEC(
 					    zfs_vdev_open_timeout_ms * 2);
@@ -305,8 +354,8 @@ vdev_disk_open(vdev_t *v, uint64_t *psize, uint64_t *max_psize,
 	hrtime_t start = gethrtime();
 	bdev = ERR_PTR(-ENXIO);
 	while (IS_ERR(bdev) && ((gethrtime() - start) < timeout)) {
-		bdev = blkdev_get_by_path(v->vdev_path, mode | FMODE_EXCL,
-		    zfs_vdev_holder);
+		bdev = vdev_blkdev_get_by_path(v->vdev_path, mode,
+		    zfs_vdev_holder, NULL);
 		if (unlikely(PTR_ERR(bdev) == -ENOENT)) {
 			/*
 			 * There is no point of waiting since device is removed
@@ -382,8 +431,8 @@ vdev_disk_close(vdev_t *v)
 		return;
 
 	if (vd->vd_bdev != NULL) {
-		blkdev_put(vd->vd_bdev,
-		    vdev_bdev_mode(spa_mode(v->vdev_spa)) | FMODE_EXCL);
+		vdev_blkdev_put(vd->vd_bdev, spa_mode(v->vdev_spa),
+		    zfs_vdev_holder);
 	}
 
 	rw_destroy(&vd->vd_lock);

module/os/linux/zfs/zfs_ctldir.c

@@ -478,16 +478,18 @@ zfsctl_is_snapdir(struct inode *ip)
  */
 static struct inode *
 zfsctl_inode_alloc(zfsvfs_t *zfsvfs, uint64_t id,
-    const struct file_operations *fops, const struct inode_operations *ops)
+    const struct file_operations *fops, const struct inode_operations *ops,
+    uint64_t creation)
 {
-	inode_timespec_t now;
 	struct inode *ip;
 	znode_t *zp;
+	inode_timespec_t now = {.tv_sec = creation};
 
 	ip = new_inode(zfsvfs->z_sb);
 	if (ip == NULL)
 		return (NULL);
 
+	if (!creation)
 		now = current_time(ip);
 	zp = ITOZ(ip);
 	ASSERT3P(zp->z_dirlocks, ==, NULL);
@@ -535,7 +537,6 @@ zfsctl_inode_alloc(zfsvfs_t *zfsvfs, uint64_t id,
 
 	mutex_enter(&zfsvfs->z_znodes_lock);
 	list_insert_tail(&zfsvfs->z_all_znodes, zp);
-	zfsvfs->z_nr_znodes++;
 	membar_producer();
 	mutex_exit(&zfsvfs->z_znodes_lock);
 
@@ -552,14 +553,28 @@ zfsctl_inode_lookup(zfsvfs_t *zfsvfs, uint64_t id,
     const struct file_operations *fops, const struct inode_operations *ops)
 {
 	struct inode *ip = NULL;
+	uint64_t creation = 0;
+	dsl_dataset_t *snap_ds;
+	dsl_pool_t *pool;
 
 	while (ip == NULL) {
 		ip = ilookup(zfsvfs->z_sb, (unsigned long)id);
 		if (ip)
 			break;
 
+		if (id <= ZFSCTL_INO_SNAPDIRS && !creation) {
+			pool = dmu_objset_pool(zfsvfs->z_os);
+			dsl_pool_config_enter(pool, FTAG);
+			if (!dsl_dataset_hold_obj(pool,
+			    ZFSCTL_INO_SNAPDIRS - id, FTAG, &snap_ds)) {
+				creation = dsl_get_creation(snap_ds);
+				dsl_dataset_rele(snap_ds, FTAG);
+			}
+			dsl_pool_config_exit(pool, FTAG);
+		}
+
 		/* May fail due to concurrent zfsctl_inode_alloc() */
-		ip = zfsctl_inode_alloc(zfsvfs, id, fops, ops);
+		ip = zfsctl_inode_alloc(zfsvfs, id, fops, ops, creation);
 	}
 
 	return (ip);
@@ -581,7 +596,7 @@ zfsctl_create(zfsvfs_t *zfsvfs)
 	ASSERT(zfsvfs->z_ctldir == NULL);
 
 	zfsvfs->z_ctldir = zfsctl_inode_alloc(zfsvfs, ZFSCTL_INO_ROOT,
-	    &zpl_fops_root, &zpl_ops_root);
+	    &zpl_fops_root, &zpl_ops_root, 0);
 	if (zfsvfs->z_ctldir == NULL)
 		return (SET_ERROR(ENOENT));
 

module/os/linux/zfs/zfs_vfsops.c

@@ -1330,12 +1330,11 @@ zfsvfs_teardown(zfsvfs_t *zfsvfs, boolean_t unmounting)
 		 * may add the parents of dir-based xattrs to the taskq
 		 * so we want to wait for these.
 		 *
-		 * We can safely read z_nr_znodes without locking because the
-		 * VFS has already blocked operations which add to the
-		 * z_all_znodes list and thus increment z_nr_znodes.
+		 * We can safely check z_all_znodes for being empty because the
+		 * VFS has already blocked operations which add to it.
 		 */
 		int round = 0;
-		while (zfsvfs->z_nr_znodes > 0) {
+		while (!list_is_empty(&zfsvfs->z_all_znodes)) {
 			taskq_wait_outstanding(dsl_pool_zrele_taskq(
 			    dmu_objset_pool(zfsvfs->z_os)), 0);
 			if (++round > 1 && !unmounting)
@@ -1662,6 +1661,7 @@ zfs_umount(struct super_block *sb)
 	}
 
 	zfsvfs_free(zfsvfs);
+	sb->s_fs_info = NULL;
 	return (0);
 }
 
@@ -2091,6 +2091,9 @@ zfs_init(void)
 	zfs_znode_init();
 	dmu_objset_register_type(DMU_OST_ZFS, zpl_get_file_info);
 	register_filesystem(&zpl_fs_type);
+#ifdef HAVE_VFS_FILE_OPERATIONS_EXTEND
+	register_fo_extend(&zpl_file_operations);
+#endif
 }
 
 void
@@ -2101,6 +2104,9 @@ zfs_fini(void)
 	 */
 	taskq_wait(system_delay_taskq);
 	taskq_wait(system_taskq);
+#ifdef HAVE_VFS_FILE_OPERATIONS_EXTEND
+	unregister_fo_extend(&zpl_file_operations);
+#endif
 	unregister_filesystem(&zpl_fs_type);
 	zfs_znode_fini();
 	zfsctl_fini();

module/os/linux/zfs/zfs_vnops_os.c

@@ -186,7 +186,7 @@ zfs_open(struct inode *ip, int mode, int flag, cred_t *cr)
 		return (error);
 
 	/* Honor ZFS_APPENDONLY file attribute */
-	if ((mode & FMODE_WRITE) && (zp->z_pflags & ZFS_APPENDONLY) &&
+	if (blk_mode_is_open_write(mode) && (zp->z_pflags & ZFS_APPENDONLY) &&
 	    ((flag & O_APPEND) == 0)) {
 		zfs_exit(zfsvfs, FTAG);
 		return (SET_ERROR(EPERM));

module/os/linux/zfs/zfs_znode.c

@@ -390,7 +390,6 @@ zfs_inode_destroy(struct inode *ip)
 	mutex_enter(&zfsvfs->z_znodes_lock);
 	if (list_link_active(&zp->z_link_node)) {
 		list_remove(&zfsvfs->z_all_znodes, zp);
-		zfsvfs->z_nr_znodes--;
 	}
 	mutex_exit(&zfsvfs->z_znodes_lock);
 
@@ -415,7 +414,11 @@ zfs_inode_set_ops(zfsvfs_t *zfsvfs, struct inode *ip)
 	switch (ip->i_mode & S_IFMT) {
 	case S_IFREG:
 		ip->i_op = &zpl_inode_operations;
+#ifdef HAVE_VFS_FILE_OPERATIONS_EXTEND
+		ip->i_fop = &zpl_file_operations.kabi_fops;
+#else
 		ip->i_fop = &zpl_file_operations;
+#endif
 		ip->i_mapping->a_ops = &zpl_address_space_operations;
 		break;
 
@@ -455,7 +458,11 @@ zfs_inode_set_ops(zfsvfs_t *zfsvfs, struct inode *ip)
 		/* Assume the inode is a file and attempt to continue */
 		ip->i_mode = S_IFREG | 0644;
 		ip->i_op = &zpl_inode_operations;
+#ifdef HAVE_VFS_FILE_OPERATIONS_EXTEND
+		ip->i_fop = &zpl_file_operations.kabi_fops;
+#else
 		ip->i_fop = &zpl_file_operations;
+#endif
 		ip->i_mapping->a_ops = &zpl_address_space_operations;
 		break;
 	}
@@ -633,7 +640,6 @@ zfs_znode_alloc(zfsvfs_t *zfsvfs, dmu_buf_t *db, int blksz,
 
 	mutex_enter(&zfsvfs->z_znodes_lock);
 	list_insert_tail(&zfsvfs->z_all_znodes, zp);
-	zfsvfs->z_nr_znodes++;
 	mutex_exit(&zfsvfs->z_znodes_lock);
 
 	if (links > 0)

module/os/linux/zfs/zpl_ctldir.c

@@ -42,7 +42,7 @@
 static int
 zpl_common_open(struct inode *ip, struct file *filp)
 {
-	if (filp->f_mode & FMODE_WRITE)
+	if (blk_mode_is_open_write(filp->f_mode))
 		return (-EACCES);
 
 	return (generic_file_open(ip, filp));

module/os/linux/zfs/zpl_file.c

@@ -301,15 +301,10 @@ zpl_uio_init(zfs_uio_t *uio, struct kiocb *kiocb, struct iov_iter *to,
 #if defined(HAVE_VFS_IOV_ITER)
 	zfs_uio_iov_iter_init(uio, to, pos, count, skip);
 #else
-#ifdef HAVE_IOV_ITER_TYPE
-	zfs_uio_iovec_init(uio, to->iov, to->nr_segs, pos,
-	    iov_iter_type(to) & ITER_KVEC ? UIO_SYSSPACE : UIO_USERSPACE,
+	zfs_uio_iovec_init(uio, zfs_uio_iter_iov(to), to->nr_segs, pos,
+	    zfs_uio_iov_iter_type(to) & ITER_KVEC ?
+	    UIO_SYSSPACE : UIO_USERSPACE,
 	    count, skip);
-#else
-	zfs_uio_iovec_init(uio, to->iov, to->nr_segs, pos,
-	    to->type & ITER_KVEC ? UIO_SYSSPACE : UIO_USERSPACE,
-	    count, skip);
-#endif
 #endif
 }
 
@@ -1257,6 +1252,12 @@ zpl_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 		return (zpl_ioctl_getdosflags(filp, (void *)arg));
 	case ZFS_IOC_SETDOSFLAGS:
 		return (zpl_ioctl_setdosflags(filp, (void *)arg));
+	case ZFS_IOC_COMPAT_FICLONE:
+		return (zpl_ioctl_ficlone(filp, (void *)arg));
+	case ZFS_IOC_COMPAT_FICLONERANGE:
+		return (zpl_ioctl_ficlonerange(filp, (void *)arg));
+	case ZFS_IOC_COMPAT_FIDEDUPERANGE:
+		return (zpl_ioctl_fideduperange(filp, (void *)arg));
 	default:
 		return (-ENOTTY);
 	}
@@ -1283,7 +1284,6 @@ zpl_compat_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 }
 #endif /* CONFIG_COMPAT */
 
-
 const struct address_space_operations zpl_address_space_operations = {
 #ifdef HAVE_VFS_READPAGES
 	.readpages	= zpl_readpages,
@@ -1306,7 +1306,12 @@ const struct address_space_operations zpl_address_space_operations = {
 #endif
 };
 
+#ifdef HAVE_VFS_FILE_OPERATIONS_EXTEND
+const struct file_operations_extend zpl_file_operations = {
+	.kabi_fops = {
+#else
 const struct file_operations zpl_file_operations = {
+#endif
 	.open		= zpl_open,
 	.release	= zpl_release,
 	.llseek		= zpl_llseek,
@@ -1318,7 +1323,11 @@ const struct file_operations zpl_file_operations = {
 	.read_iter	= zpl_iter_read,
 	.write_iter	= zpl_iter_write,
 #ifdef HAVE_VFS_IOV_ITER
+#ifdef HAVE_COPY_SPLICE_READ
+	.splice_read	= copy_splice_read,
+#else
 	.splice_read	= generic_file_splice_read,
+#endif
 	.splice_write	= iter_file_splice_write,
 #endif
 #else
@@ -1333,6 +1342,18 @@ const struct file_operations zpl_file_operations = {
 	.aio_fsync	= zpl_aio_fsync,
 #endif
 	.fallocate	= zpl_fallocate,
+#ifdef HAVE_VFS_COPY_FILE_RANGE
+	.copy_file_range	= zpl_copy_file_range,
+#endif
+#ifdef HAVE_VFS_CLONE_FILE_RANGE
+	.clone_file_range	= zpl_clone_file_range,
+#endif
+#ifdef HAVE_VFS_REMAP_FILE_RANGE
+	.remap_file_range	= zpl_remap_file_range,
+#endif
+#ifdef HAVE_VFS_DEDUPE_FILE_RANGE
+	.dedupe_file_range	= zpl_dedupe_file_range,
+#endif
 #ifdef HAVE_FILE_FADVISE
 	.fadvise	= zpl_fadvise,
 #endif
@@ -1340,6 +1361,11 @@ const struct file_operations zpl_file_operations = {
 #ifdef CONFIG_COMPAT
 	.compat_ioctl	= zpl_compat_ioctl,
 #endif
+#ifdef HAVE_VFS_FILE_OPERATIONS_EXTEND
+	}, /* kabi_fops */
+	.copy_file_range	= zpl_copy_file_range,
+	.clone_file_range	= zpl_clone_file_range,
+#endif
 };
 
 const struct file_operations zpl_dir_file_operations = {

module/os/linux/zfs/zpl_file_range.c

@@ -0,0 +1,272 @@
+/*
+ * CDDL HEADER START
+ *
+ * The contents of this file are subject to the terms of the
+ * Common Development and Distribution License (the "License").
+ * You may not use this file except in compliance with the License.
+ *
+ * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+ * or https://opensource.org/licenses/CDDL-1.0.
+ * See the License for the specific language governing permissions
+ * and limitations under the License.
+ *
+ * When distributing Covered Code, include this CDDL HEADER in each
+ * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+ * If applicable, add the following below this CDDL HEADER, with the
+ * fields enclosed by brackets "[]" replaced with your own identifying
+ * information: Portions Copyright [yyyy] [name of copyright owner]
+ *
+ * CDDL HEADER END
+ */
+/*
+ * Copyright (c) 2023, Klara Inc.
+ */
+
+#ifdef CONFIG_COMPAT
+#include <linux/compat.h>
+#endif
+#include <linux/fs.h>
+#include <sys/file.h>
+#include <sys/zfs_znode.h>
+#include <sys/zfs_vnops.h>
+#include <sys/zfeature.h>
+
+/*
+ * Clone part of a file via block cloning.
+ *
+ * Note that we are not required to update file offsets; the kernel will take
+ * care of that depending on how it was called.
+ */
+static ssize_t
+__zpl_clone_file_range(struct file *src_file, loff_t src_off,
+    struct file *dst_file, loff_t dst_off, size_t len)
+{
+	struct inode *src_i = file_inode(src_file);
+	struct inode *dst_i = file_inode(dst_file);
+	uint64_t src_off_o = (uint64_t)src_off;
+	uint64_t dst_off_o = (uint64_t)dst_off;
+	uint64_t len_o = (uint64_t)len;
+	cred_t *cr = CRED();
+	fstrans_cookie_t cookie;
+	int err;
+
+	if (!spa_feature_is_enabled(
+	    dmu_objset_spa(ITOZSB(dst_i)->z_os), SPA_FEATURE_BLOCK_CLONING))
+		return (-EOPNOTSUPP);
+
+	if (src_i != dst_i)
+		spl_inode_lock_shared(src_i);
+	spl_inode_lock(dst_i);
+
+	crhold(cr);
+	cookie = spl_fstrans_mark();
+
+	err = -zfs_clone_range(ITOZ(src_i), &src_off_o, ITOZ(dst_i),
+	    &dst_off_o, &len_o, cr);
+
+	spl_fstrans_unmark(cookie);
+	crfree(cr);
+
+	spl_inode_unlock(dst_i);
+	if (src_i != dst_i)
+		spl_inode_unlock_shared(src_i);
+
+	if (err < 0)
+		return (err);
+
+	return ((ssize_t)len_o);
+}
+
+#if defined(HAVE_VFS_COPY_FILE_RANGE) || \
+    defined(HAVE_VFS_FILE_OPERATIONS_EXTEND)
+/*
+ * Entry point for copy_file_range(). Copy len bytes from src_off in src_file
+ * to dst_off in dst_file. We are permitted to do this however we like, so we
+ * try to just clone the blocks, and if we can't support it, fall back to the
+ * kernel's generic byte copy function.
+ */
+ssize_t
+zpl_copy_file_range(struct file *src_file, loff_t src_off,
+    struct file *dst_file, loff_t dst_off, size_t len, unsigned int flags)
+{
+	ssize_t ret;
+
+	if (flags != 0)
+		return (-EINVAL);
+
+	/* Try to do it via zfs_clone_range() */
+	ret = __zpl_clone_file_range(src_file, src_off,
+	    dst_file, dst_off, len);
+
+#ifdef HAVE_VFS_GENERIC_COPY_FILE_RANGE
+	/*
+	 * Since Linux 5.3 the filesystem driver is responsible for executing
+	 * an appropriate fallback, and a generic fallback function is provided.
+	 */
+	if (ret == -EOPNOTSUPP || ret == -EINVAL || ret == -EXDEV ||
+	    ret == -EAGAIN)
+		ret = generic_copy_file_range(src_file, src_off, dst_file,
+		    dst_off, len, flags);
+#else
+	/*
+	 * Before Linux 5.3 the filesystem has to return -EOPNOTSUPP to signal
+	 * to the kernel that it should fallback to a content copy.
+	 */
+	if (ret == -EINVAL || ret == -EXDEV || ret == -EAGAIN)
+		ret = -EOPNOTSUPP;
+#endif /* HAVE_VFS_GENERIC_COPY_FILE_RANGE */
+
+	return (ret);
+}
+#endif /* HAVE_VFS_COPY_FILE_RANGE || HAVE_VFS_FILE_OPERATIONS_EXTEND */
+
+#ifdef HAVE_VFS_REMAP_FILE_RANGE
+/*
+ * Entry point for FICLONE/FICLONERANGE/FIDEDUPERANGE.
+ *
+ * FICLONE and FICLONERANGE are basically the same as copy_file_range(), except
+ * that they must clone - they cannot fall back to copying. FICLONE is exactly
+ * FICLONERANGE, for the entire file. We don't need to try to tell them apart;
+ * the kernel will sort that out for us.
+ *
+ * FIDEDUPERANGE is for turning a non-clone into a clone, that is, compare the
+ * range in both files and if they're the same, arrange for them to be backed
+ * by the same storage.
+ */
+loff_t
+zpl_remap_file_range(struct file *src_file, loff_t src_off,
+    struct file *dst_file, loff_t dst_off, loff_t len, unsigned int flags)
+{
+	if (flags & ~(REMAP_FILE_DEDUP | REMAP_FILE_CAN_SHORTEN))
+		return (-EINVAL);
+
+	/*
+	 * REMAP_FILE_CAN_SHORTEN lets us know we can clone less than the given
+	 * range if we want. Its designed for filesystems that make data past
+	 * EOF available, and don't want it to be visible in both files. ZFS
+	 * doesn't do that, so we just turn the flag off.
+	 */
+	flags &= ~REMAP_FILE_CAN_SHORTEN;
+
+	if (flags & REMAP_FILE_DEDUP)
+		/* No support for dedup yet */
+		return (-EOPNOTSUPP);
+
+	/* Zero length means to clone everything to the end of the file */
+	if (len == 0)
+		len = i_size_read(file_inode(src_file)) - src_off;
+
+	return (__zpl_clone_file_range(src_file, src_off,
+	    dst_file, dst_off, len));
+}
+#endif /* HAVE_VFS_REMAP_FILE_RANGE */
+
+#if defined(HAVE_VFS_CLONE_FILE_RANGE) || \
+    defined(HAVE_VFS_FILE_OPERATIONS_EXTEND)
+/*
+ * Entry point for FICLONE and FICLONERANGE, before Linux 4.20.
+ */
+int
+zpl_clone_file_range(struct file *src_file, loff_t src_off,
+    struct file *dst_file, loff_t dst_off, uint64_t len)
+{
+	/* Zero length means to clone everything to the end of the file */
+	if (len == 0)
+		len = i_size_read(file_inode(src_file)) - src_off;
+
+	return (__zpl_clone_file_range(src_file, src_off,
+	    dst_file, dst_off, len));
+}
+#endif /* HAVE_VFS_CLONE_FILE_RANGE || HAVE_VFS_FILE_OPERATIONS_EXTEND */
+
+#ifdef HAVE_VFS_DEDUPE_FILE_RANGE
+/*
+ * Entry point for FIDEDUPERANGE, before Linux 4.20.
+ */
+int
+zpl_dedupe_file_range(struct file *src_file, loff_t src_off,
+    struct file *dst_file, loff_t dst_off, uint64_t len)
+{
+	/* No support for dedup yet */
+	return (-EOPNOTSUPP);
+}
+#endif /* HAVE_VFS_DEDUPE_FILE_RANGE */
+
+/* Entry point for FICLONE, before Linux 4.5. */
+long
+zpl_ioctl_ficlone(struct file *dst_file, void *arg)
+{
+	unsigned long sfd = (unsigned long)arg;
+
+	struct file *src_file = fget(sfd);
+	if (src_file == NULL)
+		return (-EBADF);
+
+	if (dst_file->f_op != src_file->f_op)
+		return (-EXDEV);
+
+	size_t len = i_size_read(file_inode(src_file));
+
+	ssize_t ret =
+	    __zpl_clone_file_range(src_file, 0, dst_file, 0, len);
+
+	fput(src_file);
+
+	if (ret < 0) {
+		if (ret == -EOPNOTSUPP)
+			return (-ENOTTY);
+		return (ret);
+	}
+
+	if (ret != len)
+		return (-EINVAL);
+
+	return (0);
+}
+
+/* Entry point for FICLONERANGE, before Linux 4.5. */
+long
+zpl_ioctl_ficlonerange(struct file *dst_file, void __user *arg)
+{
+	zfs_ioc_compat_file_clone_range_t fcr;
+
+	if (copy_from_user(&fcr, arg, sizeof (fcr)))
+		return (-EFAULT);
+
+	struct file *src_file = fget(fcr.fcr_src_fd);
+	if (src_file == NULL)
+		return (-EBADF);
+
+	if (dst_file->f_op != src_file->f_op)
+		return (-EXDEV);
+
+	size_t len = fcr.fcr_src_length;
+	if (len == 0)
+		len = i_size_read(file_inode(src_file)) - fcr.fcr_src_offset;
+
+	ssize_t ret = __zpl_clone_file_range(src_file, fcr.fcr_src_offset,
+	    dst_file, fcr.fcr_dest_offset, len);
+
+	fput(src_file);
+
+	if (ret < 0) {
+		if (ret == -EOPNOTSUPP)
+			return (-ENOTTY);
+		return (ret);
+	}
+
+	if (ret != len)
+		return (-EINVAL);
+
+	return (0);
+}
+
+/* Entry point for FIDEDUPERANGE, before Linux 4.5. */
+long
+zpl_ioctl_fideduperange(struct file *filp, void *arg)
+{
+	(void) arg;
+
+	/* No support for dedup yet */
+	return (-ENOTTY);
+}

module/os/linux/zfs/zpl_super.c

@@ -277,8 +277,6 @@ zpl_test_super(struct super_block *s, void *data)
 {
 	zfsvfs_t *zfsvfs = s->s_fs_info;
 	objset_t *os = data;
-	int match;
-
 	/*
 	 * If the os doesn't match the z_os in the super_block, assume it is
 	 * not a match. Matching would imply a multimount of a dataset. It is
@@ -286,19 +284,7 @@ zpl_test_super(struct super_block *s, void *data)
 	 * that changes the z_os, e.g., rollback, where the match will be
 	 * missed, but in that case the user will get an EBUSY.
 	 */
-	if (zfsvfs == NULL || os != zfsvfs->z_os)
-		return (0);
-
-	/*
-	 * If they do match, recheck with the lock held to prevent mounting the
-	 * wrong dataset since z_os can be stale when the teardown lock is held.
-	 */
-	if (zpl_enter(zfsvfs, FTAG) != 0)
-		return (0);
-	match = (os == zfsvfs->z_os);
-	zpl_exit(zfsvfs, FTAG);
-
-	return (match);
+	return (zfsvfs != NULL && os == zfsvfs->z_os);
 }
 
 static struct super_block *
@@ -324,12 +310,35 @@ zpl_mount_impl(struct file_system_type *fs_type, int flags, zfs_mnt_t *zm)
 
 	s = sget(fs_type, zpl_test_super, set_anon_super, flags, os);
 
+	/*
+	 * Recheck with the lock held to prevent mounting the wrong dataset
+	 * since z_os can be stale when the teardown lock is held.
+	 *
+	 * We can't do this in zpl_test_super in since it's under spinlock and
+	 * also s_umount lock is not held there so it would race with
+	 * zfs_umount and zfsvfs can be freed.
+	 */
+	if (!IS_ERR(s) && s->s_fs_info != NULL) {
+		zfsvfs_t *zfsvfs = s->s_fs_info;
+		if (zpl_enter(zfsvfs, FTAG) == 0) {
+			if (os != zfsvfs->z_os)
+				err = -SET_ERROR(EBUSY);
+			zpl_exit(zfsvfs, FTAG);
+		} else {
+			err = -SET_ERROR(EBUSY);
+		}
+	}
 	dsl_dataset_long_rele(dmu_objset_ds(os), FTAG);
 	dsl_dataset_rele(dmu_objset_ds(os), FTAG);
 
 	if (IS_ERR(s))
 		return (ERR_CAST(s));
 
+	if (err) {
+		deactivate_locked_super(s);
+		return (ERR_PTR(err));
+	}
+
 	if (s->s_root == NULL) {
 		err = zpl_fill_super(s, zm, flags & SB_SILENT ? 1 : 0);
 		if (err) {

module/os/linux/zfs/zvol_os.c

@@ -671,7 +671,11 @@ zvol_request(struct request_queue *q, struct bio *bio)
 }
 
 static int
+#ifdef HAVE_BLK_MODE_T
+zvol_open(struct gendisk *disk, blk_mode_t flag)
+#else
 zvol_open(struct block_device *bdev, fmode_t flag)
+#endif
 {
 	zvol_state_t *zv;
 	int error = 0;
@@ -686,10 +690,14 @@ retry:
 	/*
 	 * Obtain a copy of private_data under the zvol_state_lock to make
 	 * sure that either the result of zvol free code path setting
-	 * bdev->bd_disk->private_data to NULL is observed, or zvol_os_free()
+	 * disk->private_data to NULL is observed, or zvol_os_free()
 	 * is not called on this zv because of the positive zv_open_count.
 	 */
+#ifdef HAVE_BLK_MODE_T
+	zv = disk->private_data;
+#else
 	zv = bdev->bd_disk->private_data;
+#endif
 	if (zv == NULL) {
 		rw_exit(&zvol_state_lock);
 		return (SET_ERROR(-ENXIO));
@@ -769,14 +777,15 @@ retry:
 			}
 		}
 
-		error = -zvol_first_open(zv, !(flag & FMODE_WRITE));
+		error = -zvol_first_open(zv, !(blk_mode_is_open_write(flag)));
 
 		if (drop_namespace)
 			mutex_exit(&spa_namespace_lock);
 	}
 
 	if (error == 0) {
-		if ((flag & FMODE_WRITE) && (zv->zv_flags & ZVOL_RDONLY)) {
+		if ((blk_mode_is_open_write(flag)) &&
+		    (zv->zv_flags & ZVOL_RDONLY)) {
 			if (zv->zv_open_count == 0)
 				zvol_last_close(zv);
 
@@ -791,14 +800,25 @@ retry:
 		rw_exit(&zv->zv_suspend_lock);
 
 	if (error == 0)
+#ifdef HAVE_BLK_MODE_T
+		disk_check_media_change(disk);
+#else
 		zfs_check_media_change(bdev);
+#endif
 
 	return (error);
 }
 
 static void
-zvol_release(struct gendisk *disk, fmode_t mode)
+#ifdef HAVE_BLOCK_DEVICE_OPERATIONS_RELEASE_1ARG
+zvol_release(struct gendisk *disk)
+#else
+zvol_release(struct gendisk *disk, fmode_t unused)
+#endif
 {
+#if !defined(HAVE_BLOCK_DEVICE_OPERATIONS_RELEASE_1ARG)
+	(void) unused;
+#endif
 	zvol_state_t *zv;
 	boolean_t drop_suspend = B_TRUE;
 

module/zcommon/zpool_prop.c

@@ -160,7 +160,7 @@ zpool_prop_init(void)
 	    "wait | continue | panic", "FAILMODE", failuremode_table,
 	    sfeatures);
 	zprop_register_index(ZPOOL_PROP_AUTOTRIM, "autotrim",
-	    SPA_AUTOTRIM_DEFAULT, PROP_DEFAULT, ZFS_TYPE_POOL,
+	    SPA_AUTOTRIM_OFF, PROP_DEFAULT, ZFS_TYPE_POOL,
 	    "on | off", "AUTOTRIM", boolean_table, sfeatures);
 
 	/* hidden properties */

module/zfs/arc.c

@@ -748,8 +748,7 @@ taskq_t *arc_prune_taskq;
  * Other sizes
  */
 
-#define	HDR_FULL_CRYPT_SIZE ((int64_t)sizeof (arc_buf_hdr_t))
-#define	HDR_FULL_SIZE ((int64_t)offsetof(arc_buf_hdr_t, b_crypt_hdr))
+#define	HDR_FULL_SIZE ((int64_t)sizeof (arc_buf_hdr_t))
 #define	HDR_L2ONLY_SIZE ((int64_t)offsetof(arc_buf_hdr_t, b_l1hdr))
 
 /*
@@ -1113,7 +1112,6 @@ buf_hash_remove(arc_buf_hdr_t *hdr)
  */
 
 static kmem_cache_t *hdr_full_cache;
-static kmem_cache_t *hdr_full_crypt_cache;
 static kmem_cache_t *hdr_l2only_cache;
 static kmem_cache_t *buf_cache;
 
@@ -1134,7 +1132,6 @@ buf_fini(void)
 	for (int i = 0; i < BUF_LOCKS; i++)
 		mutex_destroy(BUF_HASH_LOCK(i));
 	kmem_cache_destroy(hdr_full_cache);
-	kmem_cache_destroy(hdr_full_crypt_cache);
 	kmem_cache_destroy(hdr_l2only_cache);
 	kmem_cache_destroy(buf_cache);
 }
@@ -1151,7 +1148,6 @@ hdr_full_cons(void *vbuf, void *unused, int kmflag)
 
 	memset(hdr, 0, HDR_FULL_SIZE);
 	hdr->b_l1hdr.b_byteswap = DMU_BSWAP_NUMFUNCS;
-	cv_init(&hdr->b_l1hdr.b_cv, NULL, CV_DEFAULT, NULL);
 	zfs_refcount_create(&hdr->b_l1hdr.b_refcnt);
 #ifdef ZFS_DEBUG
 	mutex_init(&hdr->b_l1hdr.b_freeze_lock, NULL, MUTEX_DEFAULT, NULL);
@@ -1163,19 +1159,6 @@ hdr_full_cons(void *vbuf, void *unused, int kmflag)
 	return (0);
 }
 
-static int
-hdr_full_crypt_cons(void *vbuf, void *unused, int kmflag)
-{
-	(void) unused;
-	arc_buf_hdr_t *hdr = vbuf;
-
-	hdr_full_cons(vbuf, unused, kmflag);
-	memset(&hdr->b_crypt_hdr, 0, sizeof (hdr->b_crypt_hdr));
-	arc_space_consume(sizeof (hdr->b_crypt_hdr), ARC_SPACE_HDRS);
-
-	return (0);
-}
-
 static int
 hdr_l2only_cons(void *vbuf, void *unused, int kmflag)
 {
@@ -1211,7 +1194,6 @@ hdr_full_dest(void *vbuf, void *unused)
 	arc_buf_hdr_t *hdr = vbuf;
 
 	ASSERT(HDR_EMPTY(hdr));
-	cv_destroy(&hdr->b_l1hdr.b_cv);
 	zfs_refcount_destroy(&hdr->b_l1hdr.b_refcnt);
 #ifdef ZFS_DEBUG
 	mutex_destroy(&hdr->b_l1hdr.b_freeze_lock);
@@ -1220,16 +1202,6 @@ hdr_full_dest(void *vbuf, void *unused)
 	arc_space_return(HDR_FULL_SIZE, ARC_SPACE_HDRS);
 }
 
-static void
-hdr_full_crypt_dest(void *vbuf, void *unused)
-{
-	(void) vbuf, (void) unused;
-
-	hdr_full_dest(vbuf, unused);
-	arc_space_return(sizeof (((arc_buf_hdr_t *)NULL)->b_crypt_hdr),
-	    ARC_SPACE_HDRS);
-}
-
 static void
 hdr_l2only_dest(void *vbuf, void *unused)
 {
@@ -1285,9 +1257,6 @@ retry:
 
 	hdr_full_cache = kmem_cache_create("arc_buf_hdr_t_full", HDR_FULL_SIZE,
 	    0, hdr_full_cons, hdr_full_dest, NULL, NULL, NULL, 0);
-	hdr_full_crypt_cache = kmem_cache_create("arc_buf_hdr_t_full_crypt",
-	    HDR_FULL_CRYPT_SIZE, 0, hdr_full_crypt_cons, hdr_full_crypt_dest,
-	    NULL, NULL, NULL, 0);
 	hdr_l2only_cache = kmem_cache_create("arc_buf_hdr_t_l2only",
 	    HDR_L2ONLY_SIZE, 0, hdr_l2only_cons, hdr_l2only_dest, NULL,
 	    NULL, NULL, 0);
@@ -1995,7 +1964,6 @@ arc_buf_untransform_in_place(arc_buf_t *buf)
 	    arc_buf_size(buf));
 	buf->b_flags &= ~ARC_BUF_FLAG_ENCRYPTED;
 	buf->b_flags &= ~ARC_BUF_FLAG_COMPRESSED;
-	hdr->b_crypt_hdr.b_ebufcnt -= 1;
 }
 
 /*
@@ -2230,7 +2198,6 @@ arc_evictable_space_increment(arc_buf_hdr_t *hdr, arc_state_t *state)
 	ASSERT(HDR_HAS_L1HDR(hdr));
 
 	if (GHOST_STATE(state)) {
-		ASSERT0(hdr->b_l1hdr.b_bufcnt);
 		ASSERT3P(hdr->b_l1hdr.b_buf, ==, NULL);
 		ASSERT3P(hdr->b_l1hdr.b_pabd, ==, NULL);
 		ASSERT(!HDR_HAS_RABD(hdr));
@@ -2270,7 +2237,6 @@ arc_evictable_space_decrement(arc_buf_hdr_t *hdr, arc_state_t *state)
 	ASSERT(HDR_HAS_L1HDR(hdr));
 
 	if (GHOST_STATE(state)) {
-		ASSERT0(hdr->b_l1hdr.b_bufcnt);
 		ASSERT3P(hdr->b_l1hdr.b_buf, ==, NULL);
 		ASSERT3P(hdr->b_l1hdr.b_pabd, ==, NULL);
 		ASSERT(!HDR_HAS_RABD(hdr));
@@ -2386,7 +2352,9 @@ arc_buf_info(arc_buf_t *ab, arc_buf_info_t *abi, int state_index)
 		l2hdr = &hdr->b_l2hdr;
 
 	if (l1hdr) {
-		abi->abi_bufcnt = l1hdr->b_bufcnt;
+		abi->abi_bufcnt = 0;
+		for (arc_buf_t *buf = l1hdr->b_buf; buf; buf = buf->b_next)
+			abi->abi_bufcnt++;
 		abi->abi_access = l1hdr->b_arc_access;
 		abi->abi_mru_hits = l1hdr->b_mru_hits;
 		abi->abi_mru_ghost_hits = l1hdr->b_mru_ghost_hits;
@@ -2414,7 +2382,6 @@ arc_change_state(arc_state_t *new_state, arc_buf_hdr_t *hdr)
 {
 	arc_state_t *old_state;
 	int64_t refcnt;
-	uint32_t bufcnt;
 	boolean_t update_old, update_new;
 	arc_buf_contents_t type = arc_buf_type(hdr);
 
@@ -2428,19 +2395,16 @@ arc_change_state(arc_state_t *new_state, arc_buf_hdr_t *hdr)
 	if (HDR_HAS_L1HDR(hdr)) {
 		old_state = hdr->b_l1hdr.b_state;
 		refcnt = zfs_refcount_count(&hdr->b_l1hdr.b_refcnt);
-		bufcnt = hdr->b_l1hdr.b_bufcnt;
-		update_old = (bufcnt > 0 || hdr->b_l1hdr.b_pabd != NULL ||
-		    HDR_HAS_RABD(hdr));
+		update_old = (hdr->b_l1hdr.b_buf != NULL ||
+		    hdr->b_l1hdr.b_pabd != NULL || HDR_HAS_RABD(hdr));
 
-		IMPLY(GHOST_STATE(old_state), bufcnt == 0);
-		IMPLY(GHOST_STATE(new_state), bufcnt == 0);
 		IMPLY(GHOST_STATE(old_state), hdr->b_l1hdr.b_buf == NULL);
 		IMPLY(GHOST_STATE(new_state), hdr->b_l1hdr.b_buf == NULL);
-		IMPLY(old_state == arc_anon, bufcnt <= 1);
+		IMPLY(old_state == arc_anon, hdr->b_l1hdr.b_buf == NULL ||
+		    ARC_BUF_LAST(hdr->b_l1hdr.b_buf));
 	} else {
 		old_state = arc_l2c_only;
 		refcnt = 0;
-		bufcnt = 0;
 		update_old = B_FALSE;
 	}
 	update_new = update_old;
@@ -2488,14 +2452,12 @@ arc_change_state(arc_state_t *new_state, arc_buf_hdr_t *hdr)
 	if (update_new && new_state != arc_l2c_only) {
 		ASSERT(HDR_HAS_L1HDR(hdr));
 		if (GHOST_STATE(new_state)) {
-			ASSERT0(bufcnt);
 
 			/*
 			 * When moving a header to a ghost state, we first
-			 * remove all arc buffers. Thus, we'll have a
-			 * bufcnt of zero, and no arc buffer to use for
-			 * the reference. As a result, we use the arc
-			 * header pointer for the reference.
+			 * remove all arc buffers. Thus, we'll have no arc
+			 * buffer to use for the reference. As a result, we
+			 * use the arc header pointer for the reference.
 			 */
 			(void) zfs_refcount_add_many(
 			    &new_state->arcs_size[type],
@@ -2503,7 +2465,6 @@ arc_change_state(arc_state_t *new_state, arc_buf_hdr_t *hdr)
 			ASSERT3P(hdr->b_l1hdr.b_pabd, ==, NULL);
 			ASSERT(!HDR_HAS_RABD(hdr));
 		} else {
-			uint32_t buffers = 0;
 
 			/*
 			 * Each individual buffer holds a unique reference,
@@ -2512,8 +2473,6 @@ arc_change_state(arc_state_t *new_state, arc_buf_hdr_t *hdr)
 			 */
 			for (arc_buf_t *buf = hdr->b_l1hdr.b_buf; buf != NULL;
 			    buf = buf->b_next) {
-				ASSERT3U(bufcnt, !=, 0);
-				buffers++;
 
 				/*
 				 * When the arc_buf_t is sharing the data
@@ -2529,7 +2488,6 @@ arc_change_state(arc_state_t *new_state, arc_buf_hdr_t *hdr)
 				    &new_state->arcs_size[type],
 				    arc_buf_size(buf), buf);
 			}
-			ASSERT3U(bufcnt, ==, buffers);
 
 			if (hdr->b_l1hdr.b_pabd != NULL) {
 				(void) zfs_refcount_add_many(
@@ -2548,7 +2506,6 @@ arc_change_state(arc_state_t *new_state, arc_buf_hdr_t *hdr)
 	if (update_old && old_state != arc_l2c_only) {
 		ASSERT(HDR_HAS_L1HDR(hdr));
 		if (GHOST_STATE(old_state)) {
-			ASSERT0(bufcnt);
 			ASSERT3P(hdr->b_l1hdr.b_pabd, ==, NULL);
 			ASSERT(!HDR_HAS_RABD(hdr));
 
@@ -2564,7 +2521,6 @@ arc_change_state(arc_state_t *new_state, arc_buf_hdr_t *hdr)
 			    &old_state->arcs_size[type],
 			    HDR_GET_LSIZE(hdr), hdr);
 		} else {
-			uint32_t buffers = 0;
 
 			/*
 			 * Each individual buffer holds a unique reference,
@@ -2573,8 +2529,6 @@ arc_change_state(arc_state_t *new_state, arc_buf_hdr_t *hdr)
 			 */
 			for (arc_buf_t *buf = hdr->b_l1hdr.b_buf; buf != NULL;
 			    buf = buf->b_next) {
-				ASSERT3U(bufcnt, !=, 0);
-				buffers++;
 
 				/*
 				 * When the arc_buf_t is sharing the data
@@ -2590,7 +2544,6 @@ arc_change_state(arc_state_t *new_state, arc_buf_hdr_t *hdr)
 				    &old_state->arcs_size[type],
 				    arc_buf_size(buf), buf);
 			}
-			ASSERT3U(bufcnt, ==, buffers);
 			ASSERT(hdr->b_l1hdr.b_pabd != NULL ||
 			    HDR_HAS_RABD(hdr));
 
@@ -2838,9 +2791,6 @@ arc_buf_alloc_impl(arc_buf_hdr_t *hdr, spa_t *spa, const zbookmark_phys_t *zb,
 	VERIFY3P(buf->b_data, !=, NULL);
 
 	hdr->b_l1hdr.b_buf = buf;
-	hdr->b_l1hdr.b_bufcnt += 1;
-	if (encrypted)
-		hdr->b_crypt_hdr.b_ebufcnt += 1;
 
 	/*
 	 * If the user wants the data from the hdr, we need to either copy or
@@ -3082,8 +3032,6 @@ arc_buf_remove(arc_buf_hdr_t *hdr, arc_buf_t *buf)
 	}
 	buf->b_next = NULL;
 	ASSERT3P(lastbuf, !=, buf);
-	IMPLY(hdr->b_l1hdr.b_bufcnt > 0, lastbuf != NULL);
-	IMPLY(hdr->b_l1hdr.b_bufcnt > 0, hdr->b_l1hdr.b_buf != NULL);
 	IMPLY(lastbuf != NULL, ARC_BUF_LAST(lastbuf));
 
 	return (lastbuf);
@@ -3122,22 +3070,20 @@ arc_buf_destroy_impl(arc_buf_t *buf)
 		}
 		buf->b_data = NULL;
 
-		ASSERT(hdr->b_l1hdr.b_bufcnt > 0);
-		hdr->b_l1hdr.b_bufcnt -= 1;
-
-		if (ARC_BUF_ENCRYPTED(buf)) {
-			hdr->b_crypt_hdr.b_ebufcnt -= 1;
-
 		/*
-			 * If we have no more encrypted buffers and we've
-			 * already gotten a copy of the decrypted data we can
-			 * free b_rabd to save some space.
+		 * If we have no more encrypted buffers and we've already
+		 * gotten a copy of the decrypted data we can free b_rabd
+		 * to save some space.
 		 */
-			if (hdr->b_crypt_hdr.b_ebufcnt == 0 &&
-			    HDR_HAS_RABD(hdr) && hdr->b_l1hdr.b_pabd != NULL &&
-			    !HDR_IO_IN_PROGRESS(hdr)) {
-				arc_hdr_free_abd(hdr, B_TRUE);
+		if (ARC_BUF_ENCRYPTED(buf) && HDR_HAS_RABD(hdr) &&
+		    hdr->b_l1hdr.b_pabd != NULL && !HDR_IO_IN_PROGRESS(hdr)) {
+			arc_buf_t *b;
+			for (b = hdr->b_l1hdr.b_buf; b; b = b->b_next) {
+				if (b != buf && ARC_BUF_ENCRYPTED(b))
+					break;
 			}
+			if (b == NULL)
+				arc_hdr_free_abd(hdr, B_TRUE);
 		}
 	}
 
@@ -3298,11 +3244,7 @@ arc_hdr_alloc(uint64_t spa, int32_t psize, int32_t lsize,
 	arc_buf_hdr_t *hdr;
 
 	VERIFY(type == ARC_BUFC_DATA || type == ARC_BUFC_METADATA);
-	if (protected) {
-		hdr = kmem_cache_alloc(hdr_full_crypt_cache, KM_PUSHPAGE);
-	} else {
 	hdr = kmem_cache_alloc(hdr_full_cache, KM_PUSHPAGE);
-	}
 
 	ASSERT(HDR_EMPTY(hdr));
 #ifdef ZFS_DEBUG
@@ -3325,7 +3267,6 @@ arc_hdr_alloc(uint64_t spa, int32_t psize, int32_t lsize,
 	hdr->b_l1hdr.b_mru_ghost_hits = 0;
 	hdr->b_l1hdr.b_mfu_hits = 0;
 	hdr->b_l1hdr.b_mfu_ghost_hits = 0;
-	hdr->b_l1hdr.b_bufcnt = 0;
 	hdr->b_l1hdr.b_buf = NULL;
 
 	ASSERT(zfs_refcount_is_zero(&hdr->b_l1hdr.b_refcnt));
@@ -3351,16 +3292,6 @@ arc_hdr_realloc(arc_buf_hdr_t *hdr, kmem_cache_t *old, kmem_cache_t *new)
 	ASSERT((old == hdr_full_cache && new == hdr_l2only_cache) ||
 	    (old == hdr_l2only_cache && new == hdr_full_cache));
 
-	/*
-	 * if the caller wanted a new full header and the header is to be
-	 * encrypted we will actually allocate the header from the full crypt
-	 * cache instead. The same applies to freeing from the old cache.
-	 */
-	if (HDR_PROTECTED(hdr) && new == hdr_full_cache)
-		new = hdr_full_crypt_cache;
-	if (HDR_PROTECTED(hdr) && old == hdr_full_cache)
-		old = hdr_full_crypt_cache;
-
 	nhdr = kmem_cache_alloc(new, KM_PUSHPAGE);
 
 	ASSERT(MUTEX_HELD(HDR_LOCK(hdr)));
@@ -3368,7 +3299,7 @@ arc_hdr_realloc(arc_buf_hdr_t *hdr, kmem_cache_t *old, kmem_cache_t *new)
 
 	memcpy(nhdr, hdr, HDR_L2ONLY_SIZE);
 
-	if (new == hdr_full_cache || new == hdr_full_crypt_cache) {
+	if (new == hdr_full_cache) {
 		arc_hdr_set_flags(nhdr, ARC_FLAG_HAS_L1HDR);
 		/*
 		 * arc_access and arc_change_state need to be aware that a
@@ -3382,7 +3313,6 @@ arc_hdr_realloc(arc_buf_hdr_t *hdr, kmem_cache_t *old, kmem_cache_t *new)
 		ASSERT(!HDR_HAS_RABD(hdr));
 	} else {
 		ASSERT3P(hdr->b_l1hdr.b_buf, ==, NULL);
-		ASSERT0(hdr->b_l1hdr.b_bufcnt);
 #ifdef ZFS_DEBUG
 		ASSERT3P(hdr->b_l1hdr.b_freeze_cksum, ==, NULL);
 #endif
@@ -3448,126 +3378,6 @@ arc_hdr_realloc(arc_buf_hdr_t *hdr, kmem_cache_t *old, kmem_cache_t *new)
 	return (nhdr);
 }
 
-/*
- * This function allows an L1 header to be reallocated as a crypt
- * header and vice versa. If we are going to a crypt header, the
- * new fields will be zeroed out.
- */
-static arc_buf_hdr_t *
-arc_hdr_realloc_crypt(arc_buf_hdr_t *hdr, boolean_t need_crypt)
-{
-	arc_buf_hdr_t *nhdr;
-	arc_buf_t *buf;
-	kmem_cache_t *ncache, *ocache;
-
-	/*
-	 * This function requires that hdr is in the arc_anon state.
-	 * Therefore it won't have any L2ARC data for us to worry
-	 * about copying.
-	 */
-	ASSERT(HDR_HAS_L1HDR(hdr));
-	ASSERT(!HDR_HAS_L2HDR(hdr));
-	ASSERT3U(!!HDR_PROTECTED(hdr), !=, need_crypt);
-	ASSERT3P(hdr->b_l1hdr.b_state, ==, arc_anon);
-	ASSERT(!multilist_link_active(&hdr->b_l1hdr.b_arc_node));
-	ASSERT(!list_link_active(&hdr->b_l2hdr.b_l2node));
-	ASSERT3P(hdr->b_hash_next, ==, NULL);
-
-	if (need_crypt) {
-		ncache = hdr_full_crypt_cache;
-		ocache = hdr_full_cache;
-	} else {
-		ncache = hdr_full_cache;
-		ocache = hdr_full_crypt_cache;
-	}
-
-	nhdr = kmem_cache_alloc(ncache, KM_PUSHPAGE);
-
-	/*
-	 * Copy all members that aren't locks or condvars to the new header.
-	 * No lists are pointing to us (as we asserted above), so we don't
-	 * need to worry about the list nodes.
-	 */
-	nhdr->b_dva = hdr->b_dva;
-	nhdr->b_birth = hdr->b_birth;
-	nhdr->b_type = hdr->b_type;
-	nhdr->b_flags = hdr->b_flags;
-	nhdr->b_psize = hdr->b_psize;
-	nhdr->b_lsize = hdr->b_lsize;
-	nhdr->b_spa = hdr->b_spa;
-#ifdef ZFS_DEBUG
-	nhdr->b_l1hdr.b_freeze_cksum = hdr->b_l1hdr.b_freeze_cksum;
-#endif
-	nhdr->b_l1hdr.b_bufcnt = hdr->b_l1hdr.b_bufcnt;
-	nhdr->b_l1hdr.b_byteswap = hdr->b_l1hdr.b_byteswap;
-	nhdr->b_l1hdr.b_state = hdr->b_l1hdr.b_state;
-	nhdr->b_l1hdr.b_arc_access = hdr->b_l1hdr.b_arc_access;
-	nhdr->b_l1hdr.b_mru_hits = hdr->b_l1hdr.b_mru_hits;
-	nhdr->b_l1hdr.b_mru_ghost_hits = hdr->b_l1hdr.b_mru_ghost_hits;
-	nhdr->b_l1hdr.b_mfu_hits = hdr->b_l1hdr.b_mfu_hits;
-	nhdr->b_l1hdr.b_mfu_ghost_hits = hdr->b_l1hdr.b_mfu_ghost_hits;
-	nhdr->b_l1hdr.b_acb = hdr->b_l1hdr.b_acb;
-	nhdr->b_l1hdr.b_pabd = hdr->b_l1hdr.b_pabd;
-
-	/*
-	 * This zfs_refcount_add() exists only to ensure that the individual
-	 * arc buffers always point to a header that is referenced, avoiding
-	 * a small race condition that could trigger ASSERTs.
-	 */
-	(void) zfs_refcount_add(&nhdr->b_l1hdr.b_refcnt, FTAG);
-	nhdr->b_l1hdr.b_buf = hdr->b_l1hdr.b_buf;
-	for (buf = nhdr->b_l1hdr.b_buf; buf != NULL; buf = buf->b_next)
-		buf->b_hdr = nhdr;
-
-	zfs_refcount_transfer(&nhdr->b_l1hdr.b_refcnt, &hdr->b_l1hdr.b_refcnt);
-	(void) zfs_refcount_remove(&nhdr->b_l1hdr.b_refcnt, FTAG);
-	ASSERT0(zfs_refcount_count(&hdr->b_l1hdr.b_refcnt));
-
-	if (need_crypt) {
-		arc_hdr_set_flags(nhdr, ARC_FLAG_PROTECTED);
-	} else {
-		arc_hdr_clear_flags(nhdr, ARC_FLAG_PROTECTED);
-	}
-
-	/* unset all members of the original hdr */
-	memset(&hdr->b_dva, 0, sizeof (dva_t));
-	hdr->b_birth = 0;
-	hdr->b_type = 0;
-	hdr->b_flags = 0;
-	hdr->b_psize = 0;
-	hdr->b_lsize = 0;
-	hdr->b_spa = 0;
-#ifdef ZFS_DEBUG
-	hdr->b_l1hdr.b_freeze_cksum = NULL;
-#endif
-	hdr->b_l1hdr.b_buf = NULL;
-	hdr->b_l1hdr.b_bufcnt = 0;
-	hdr->b_l1hdr.b_byteswap = 0;
-	hdr->b_l1hdr.b_state = NULL;
-	hdr->b_l1hdr.b_arc_access = 0;
-	hdr->b_l1hdr.b_mru_hits = 0;
-	hdr->b_l1hdr.b_mru_ghost_hits = 0;
-	hdr->b_l1hdr.b_mfu_hits = 0;
-	hdr->b_l1hdr.b_mfu_ghost_hits = 0;
-	hdr->b_l1hdr.b_acb = NULL;
-	hdr->b_l1hdr.b_pabd = NULL;
-
-	if (ocache == hdr_full_crypt_cache) {
-		ASSERT(!HDR_HAS_RABD(hdr));
-		hdr->b_crypt_hdr.b_ot = DMU_OT_NONE;
-		hdr->b_crypt_hdr.b_ebufcnt = 0;
-		hdr->b_crypt_hdr.b_dsobj = 0;
-		memset(hdr->b_crypt_hdr.b_salt, 0, ZIO_DATA_SALT_LEN);
-		memset(hdr->b_crypt_hdr.b_iv, 0, ZIO_DATA_IV_LEN);
-		memset(hdr->b_crypt_hdr.b_mac, 0, ZIO_DATA_MAC_LEN);
-	}
-
-	buf_discard_identity(hdr);
-	kmem_cache_free(ocache, hdr);
-
-	return (nhdr);
-}
-
 /*
  * This function is used by the send / receive code to convert a newly
  * allocated arc_buf_t to one that is suitable for a raw encrypted write. It
@@ -3587,8 +3397,7 @@ arc_convert_to_raw(arc_buf_t *buf, uint64_t dsobj, boolean_t byteorder,
 	ASSERT3P(hdr->b_l1hdr.b_state, ==, arc_anon);
 
 	buf->b_flags |= (ARC_BUF_FLAG_COMPRESSED | ARC_BUF_FLAG_ENCRYPTED);
-	if (!HDR_PROTECTED(hdr))
-		hdr = arc_hdr_realloc_crypt(hdr, B_TRUE);
+	arc_hdr_set_flags(hdr, ARC_FLAG_PROTECTED);
 	hdr->b_crypt_hdr.b_dsobj = dsobj;
 	hdr->b_crypt_hdr.b_ot = ot;
 	hdr->b_l1hdr.b_byteswap = (byteorder == ZFS_HOST_BYTEORDER) ?
@@ -3789,8 +3598,6 @@ static void
 arc_hdr_destroy(arc_buf_hdr_t *hdr)
 {
 	if (HDR_HAS_L1HDR(hdr)) {
-		ASSERT(hdr->b_l1hdr.b_buf == NULL ||
-		    hdr->b_l1hdr.b_bufcnt > 0);
 		ASSERT(zfs_refcount_is_zero(&hdr->b_l1hdr.b_refcnt));
 		ASSERT3P(hdr->b_l1hdr.b_state, ==, arc_anon);
 	}
@@ -3854,12 +3661,7 @@ arc_hdr_destroy(arc_buf_hdr_t *hdr)
 #ifdef ZFS_DEBUG
 		ASSERT3P(hdr->b_l1hdr.b_freeze_cksum, ==, NULL);
 #endif
-
-		if (!HDR_PROTECTED(hdr)) {
 		kmem_cache_free(hdr_full_cache, hdr);
-		} else {
-			kmem_cache_free(hdr_full_crypt_cache, hdr);
-		}
 	} else {
 		kmem_cache_free(hdr_l2only_cache, hdr);
 	}
@@ -3871,7 +3673,8 @@ arc_buf_destroy(arc_buf_t *buf, const void *tag)
 	arc_buf_hdr_t *hdr = buf->b_hdr;
 
 	if (hdr->b_l1hdr.b_state == arc_anon) {
-		ASSERT3U(hdr->b_l1hdr.b_bufcnt, ==, 1);
+		ASSERT3P(hdr->b_l1hdr.b_buf, ==, buf);
+		ASSERT(ARC_BUF_LAST(buf));
 		ASSERT(!HDR_IO_IN_PROGRESS(hdr));
 		VERIFY0(remove_reference(hdr, tag));
 		return;
@@ -3881,7 +3684,7 @@ arc_buf_destroy(arc_buf_t *buf, const void *tag)
 	mutex_enter(hash_lock);
 
 	ASSERT3P(hdr, ==, buf->b_hdr);
-	ASSERT(hdr->b_l1hdr.b_bufcnt > 0);
+	ASSERT3P(hdr->b_l1hdr.b_buf, !=, NULL);
 	ASSERT3P(hash_lock, ==, HDR_LOCK(hdr));
 	ASSERT3P(hdr->b_l1hdr.b_state, !=, arc_anon);
 	ASSERT3P(buf->b_data, !=, NULL);
@@ -3924,7 +3727,6 @@ arc_evict_hdr(arc_buf_hdr_t *hdr, uint64_t *real_evicted)
 	ASSERT(MUTEX_HELD(HDR_LOCK(hdr)));
 	ASSERT(HDR_HAS_L1HDR(hdr));
 	ASSERT(!HDR_IO_IN_PROGRESS(hdr));
-	ASSERT0(hdr->b_l1hdr.b_bufcnt);
 	ASSERT3P(hdr->b_l1hdr.b_buf, ==, NULL);
 	ASSERT0(zfs_refcount_count(&hdr->b_l1hdr.b_refcnt));
 
@@ -5586,13 +5388,6 @@ arc_read_done(zio_t *zio)
 			buf_hash_remove(hdr);
 	}
 
-	/*
-	 * Broadcast before we drop the hash_lock to avoid the possibility
-	 * that the hdr (and hence the cv) might be freed before we get to
-	 * the cv_broadcast().
-	 */
-	cv_broadcast(&hdr->b_l1hdr.b_cv);
-
 	arc_hdr_clear_flags(hdr, ARC_FLAG_IO_IN_PROGRESS);
 	(void) remove_reference(hdr, hdr);
 
@@ -5787,7 +5582,6 @@ top:
 				}
 				acb->acb_zio_head = head_zio;
 				acb->acb_next = hdr->b_l1hdr.b_acb;
-				if (hdr->b_l1hdr.b_acb)
 				hdr->b_l1hdr.b_acb->acb_prev = acb;
 				hdr->b_l1hdr.b_acb = acb;
 			}
@@ -5928,8 +5722,28 @@ top:
 				 * and so the performance impact shouldn't
 				 * matter.
 				 */
-				cv_wait(&hdr->b_l1hdr.b_cv, hash_lock);
+				arc_callback_t *acb = kmem_zalloc(
+				    sizeof (arc_callback_t), KM_SLEEP);
+				acb->acb_wait = B_TRUE;
+				mutex_init(&acb->acb_wait_lock, NULL,
+				    MUTEX_DEFAULT, NULL);
+				cv_init(&acb->acb_wait_cv, NULL, CV_DEFAULT,
+				    NULL);
+				acb->acb_zio_head =
+				    hdr->b_l1hdr.b_acb->acb_zio_head;
+				acb->acb_next = hdr->b_l1hdr.b_acb;
+				hdr->b_l1hdr.b_acb->acb_prev = acb;
+				hdr->b_l1hdr.b_acb = acb;
 				mutex_exit(hash_lock);
+				mutex_enter(&acb->acb_wait_lock);
+				while (acb->acb_wait) {
+					cv_wait(&acb->acb_wait_cv,
+					    &acb->acb_wait_lock);
+				}
+				mutex_exit(&acb->acb_wait_lock);
+				mutex_destroy(&acb->acb_wait_lock);
+				cv_destroy(&acb->acb_wait_cv);
+				kmem_free(acb, sizeof (arc_callback_t));
 				goto top;
 			}
 		}
@@ -6310,7 +6124,8 @@ arc_release(arc_buf_t *buf, const void *tag)
 		ASSERT(!HDR_IN_HASH_TABLE(hdr));
 		ASSERT(!HDR_HAS_L2HDR(hdr));
 
-		ASSERT3U(hdr->b_l1hdr.b_bufcnt, ==, 1);
+		ASSERT3P(hdr->b_l1hdr.b_buf, ==, buf);
+		ASSERT(ARC_BUF_LAST(buf));
 		ASSERT3S(zfs_refcount_count(&hdr->b_l1hdr.b_refcnt), ==, 1);
 		ASSERT(!multilist_link_active(&hdr->b_l1hdr.b_arc_node));
 
@@ -6361,7 +6176,7 @@ arc_release(arc_buf_t *buf, const void *tag)
 	/*
 	 * Do we have more than one buf?
 	 */
-	if (hdr->b_l1hdr.b_bufcnt > 1) {
+	if (hdr->b_l1hdr.b_buf != buf || !ARC_BUF_LAST(buf)) {
 		arc_buf_hdr_t *nhdr;
 		uint64_t spa = hdr->b_spa;
 		uint64_t psize = HDR_GET_PSIZE(hdr);
@@ -6442,10 +6257,6 @@ arc_release(arc_buf_t *buf, const void *tag)
 			    arc_buf_size(buf), buf);
 		}
 
-		hdr->b_l1hdr.b_bufcnt -= 1;
-		if (ARC_BUF_ENCRYPTED(buf))
-			hdr->b_crypt_hdr.b_ebufcnt -= 1;
-
 		arc_cksum_verify(buf);
 		arc_buf_unwatch(buf);
 
@@ -6458,15 +6269,11 @@ arc_release(arc_buf_t *buf, const void *tag)
 		nhdr = arc_hdr_alloc(spa, psize, lsize, protected,
 		    compress, hdr->b_complevel, type);
 		ASSERT3P(nhdr->b_l1hdr.b_buf, ==, NULL);
-		ASSERT0(nhdr->b_l1hdr.b_bufcnt);
 		ASSERT0(zfs_refcount_count(&nhdr->b_l1hdr.b_refcnt));
 		VERIFY3U(nhdr->b_type, ==, type);
 		ASSERT(!HDR_SHARED_DATA(nhdr));
 
 		nhdr->b_l1hdr.b_buf = buf;
-		nhdr->b_l1hdr.b_bufcnt = 1;
-		if (ARC_BUF_ENCRYPTED(buf))
-			nhdr->b_crypt_hdr.b_ebufcnt = 1;
 		(void) zfs_refcount_add(&nhdr->b_l1hdr.b_refcnt, tag);
 		buf->b_hdr = nhdr;
 
@@ -6517,7 +6324,7 @@ arc_write_ready(zio_t *zio)
 
 	ASSERT(HDR_HAS_L1HDR(hdr));
 	ASSERT(!zfs_refcount_is_zero(&buf->b_hdr->b_l1hdr.b_refcnt));
-	ASSERT(hdr->b_l1hdr.b_bufcnt > 0);
+	ASSERT3P(hdr->b_l1hdr.b_buf, !=, NULL);
 
 	/*
 	 * If we're reexecuting this zio because the pool suspended, then
@@ -6552,13 +6359,9 @@ arc_write_ready(zio_t *zio)
 		add_reference(hdr, hdr); /* For IO_IN_PROGRESS. */
 	}
 
-	if (BP_IS_PROTECTED(bp) != !!HDR_PROTECTED(hdr))
-		hdr = arc_hdr_realloc_crypt(hdr, BP_IS_PROTECTED(bp));
-
 	if (BP_IS_PROTECTED(bp)) {
 		/* ZIL blocks are written through zio_rewrite */
 		ASSERT3U(BP_GET_TYPE(bp), !=, DMU_OT_INTENT_LOG);
-		ASSERT(HDR_PROTECTED(hdr));
 
 		if (BP_SHOULD_BYTESWAP(bp)) {
 			if (BP_GET_LEVEL(bp) > 0) {
@@ -6571,11 +6374,14 @@ arc_write_ready(zio_t *zio)
 			hdr->b_l1hdr.b_byteswap = DMU_BSWAP_NUMFUNCS;
 		}
 
+		arc_hdr_set_flags(hdr, ARC_FLAG_PROTECTED);
 		hdr->b_crypt_hdr.b_ot = BP_GET_TYPE(bp);
 		hdr->b_crypt_hdr.b_dsobj = zio->io_bookmark.zb_objset;
 		zio_crypt_decode_params_bp(bp, hdr->b_crypt_hdr.b_salt,
 		    hdr->b_crypt_hdr.b_iv);
 		zio_crypt_decode_mac_bp(bp, hdr->b_crypt_hdr.b_mac);
+	} else {
+		arc_hdr_clear_flags(hdr, ARC_FLAG_PROTECTED);
 	}
 
 	/*
@@ -6656,7 +6462,8 @@ arc_write_ready(zio_t *zio)
 	} else {
 		ASSERT3P(buf->b_data, ==, abd_to_buf(zio->io_orig_abd));
 		ASSERT3U(zio->io_orig_size, ==, arc_buf_size(buf));
-		ASSERT3U(hdr->b_l1hdr.b_bufcnt, ==, 1);
+		ASSERT3P(hdr->b_l1hdr.b_buf, ==, buf);
+		ASSERT(ARC_BUF_LAST(buf));
 
 		arc_share_buf(hdr, buf);
 	}
@@ -6737,7 +6544,8 @@ arc_write_done(zio_t *zio)
 					    (void *)hdr, (void *)exists);
 			} else {
 				/* Dedup */
-				ASSERT(hdr->b_l1hdr.b_bufcnt == 1);
+				ASSERT3P(hdr->b_l1hdr.b_buf, !=, NULL);
+				ASSERT(ARC_BUF_LAST(hdr->b_l1hdr.b_buf));
 				ASSERT(hdr->b_l1hdr.b_state == arc_anon);
 				ASSERT(BP_GET_DEDUP(zio->io_bp));
 				ASSERT(BP_GET_LEVEL(zio->io_bp) == 0);
@@ -6778,7 +6586,7 @@ arc_write(zio_t *pio, spa_t *spa, uint64_t txg,
 	ASSERT(!HDR_IO_ERROR(hdr));
 	ASSERT(!HDR_IO_IN_PROGRESS(hdr));
 	ASSERT3P(hdr->b_l1hdr.b_acb, ==, NULL);
-	ASSERT3U(hdr->b_l1hdr.b_bufcnt, >, 0);
+	ASSERT3P(hdr->b_l1hdr.b_buf, !=, NULL);
 	if (uncached)
 		arc_hdr_set_flags(hdr, ARC_FLAG_UNCACHED);
 	else if (l2arc)
@@ -9092,15 +8900,16 @@ l2arc_apply_transforms(spa_t *spa, arc_buf_hdr_t *hdr, uint64_t asize,
 		 * write things before deciding to fail compression in nearly
 		 * every case.)
 		 */
-		cabd = abd_alloc_for_io(size, ismd);
-		tmp = abd_borrow_buf(cabd, size);
+		uint64_t bufsize = MAX(size, asize);
+		cabd = abd_alloc_for_io(bufsize, ismd);
+		tmp = abd_borrow_buf(cabd, bufsize);
 
 		psize = zio_compress_data(compress, to_write, &tmp, size,
 		    hdr->b_complevel);
 
 		if (psize >= asize) {
 			psize = HDR_GET_PSIZE(hdr);
-			abd_return_buf_copy(cabd, tmp, size);
+			abd_return_buf_copy(cabd, tmp, bufsize);
 			HDR_SET_COMPRESS(hdr, ZIO_COMPRESS_OFF);
 			to_write = cabd;
 			abd_copy(to_write, hdr->b_l1hdr.b_pabd, psize);
@@ -9110,9 +8919,9 @@ l2arc_apply_transforms(spa_t *spa, arc_buf_hdr_t *hdr, uint64_t asize,
 		}
 		ASSERT3U(psize, <=, HDR_GET_PSIZE(hdr));
 		if (psize < asize)
-			memset((char *)tmp + psize, 0, asize - psize);
+			memset((char *)tmp + psize, 0, bufsize - psize);
 		psize = HDR_GET_PSIZE(hdr);
-		abd_return_buf_copy(cabd, tmp, size);
+		abd_return_buf_copy(cabd, tmp, bufsize);
 		to_write = cabd;
 	}
 

module/zfs/bpobj.c

@@ -284,7 +284,17 @@ bpobj_iterate_blkptrs(bpobj_info_t *bpi, bpobj_itor_t func, void *arg,
 	dmu_buf_t *dbuf = NULL;
 	bpobj_t *bpo = bpi->bpi_bpo;
 
-	for (int64_t i = bpo->bpo_phys->bpo_num_blkptrs - 1; i >= start; i--) {
+	int64_t i = bpo->bpo_phys->bpo_num_blkptrs - 1;
+	uint64_t pe = P2ALIGN_TYPED(i, bpo->bpo_epb, uint64_t) *
+	    sizeof (blkptr_t);
+	uint64_t ps = start * sizeof (blkptr_t);
+	uint64_t pb = MAX((pe > dmu_prefetch_max) ? pe - dmu_prefetch_max : 0,
+	    ps);
+	if (pe > pb) {
+		dmu_prefetch(bpo->bpo_os, bpo->bpo_object, 0, pb, pe - pb,
+		    ZIO_PRIORITY_ASYNC_READ);
+	}
+	for (; i >= start; i--) {
 		uint64_t offset = i * sizeof (blkptr_t);
 		uint64_t blkoff = P2PHASE(i, bpo->bpo_epb);
 
@@ -292,9 +302,16 @@ bpobj_iterate_blkptrs(bpobj_info_t *bpi, bpobj_itor_t func, void *arg,
 			if (dbuf)
 				dmu_buf_rele(dbuf, FTAG);
 			err = dmu_buf_hold(bpo->bpo_os, bpo->bpo_object,
-			    offset, FTAG, &dbuf, 0);
+			    offset, FTAG, &dbuf, DMU_READ_NO_PREFETCH);
 			if (err)
 				break;
+			pe = pb;
+			pb = MAX((dbuf->db_offset > dmu_prefetch_max) ?
+			    dbuf->db_offset - dmu_prefetch_max : 0, ps);
+			if (pe > pb) {
+				dmu_prefetch(bpo->bpo_os, bpo->bpo_object, 0,
+				    pb, pe - pb, ZIO_PRIORITY_ASYNC_READ);
+			}
 		}
 
 		ASSERT3U(offset, >=, dbuf->db_offset);
@@ -466,22 +483,30 @@ bpobj_iterate_impl(bpobj_t *initial_bpo, bpobj_itor_t func, void *arg,
 			int64_t i = bpi->bpi_unprocessed_subobjs - 1;
 			uint64_t offset = i * sizeof (uint64_t);
 
-			uint64_t obj_from_sublist;
+			uint64_t subobj;
 			err = dmu_read(bpo->bpo_os, bpo->bpo_phys->bpo_subobjs,
-			    offset, sizeof (uint64_t), &obj_from_sublist,
-			    DMU_READ_PREFETCH);
+			    offset, sizeof (uint64_t), &subobj,
+			    DMU_READ_NO_PREFETCH);
 			if (err)
 				break;
-			bpobj_t *sublist = kmem_alloc(sizeof (bpobj_t),
+
+			bpobj_t *subbpo = kmem_alloc(sizeof (bpobj_t),
 			    KM_SLEEP);
-
-			err = bpobj_open(sublist, bpo->bpo_os,
-			    obj_from_sublist);
-			if (err)
+			err = bpobj_open(subbpo, bpo->bpo_os, subobj);
+			if (err) {
+				kmem_free(subbpo, sizeof (bpobj_t));
 				break;
+			}
 
-			list_insert_head(&stack, bpi_alloc(sublist, bpi, i));
-			mutex_enter(&sublist->bpo_lock);
+			if (subbpo->bpo_havesubobj &&
+			    subbpo->bpo_phys->bpo_subobjs != 0) {
+				dmu_prefetch(subbpo->bpo_os,
+				    subbpo->bpo_phys->bpo_subobjs, 0, 0, 0,
+				    ZIO_PRIORITY_ASYNC_READ);
+			}
+
+			list_insert_head(&stack, bpi_alloc(subbpo, bpi, i));
+			mutex_enter(&subbpo->bpo_lock);
 			bpi->bpi_unprocessed_subobjs--;
 		}
 	}