zfs: merge openzfs/zfs@66b81b349

Notable upstream pull request merges:
 #15290 54b1b1d import: require force when cachefile hostid doesn't
                  match on-disk
 #15319 342357c Reduce number of metaslab preload taskq threads
 #15340 2a6c621 ARC: Remove b_cv from struct l1arc_buf_hdr
 #15347 75a2eb7 ARC: Drop different size headers for crypto
 #15350 96b9cf4 ARC: Remove b_bufcnt/b_ebufcnt from ARC headers
 #15353 66b81b3 ZIL: Reduce maximum size of WR_COPIED to 7.5K
 #15362 5b8688e zfsconcepts: add description of block cloning

Obtained from:	OpenZFS
OpenZFS commit:	66b81b3

sys/contrib/openzfs/.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

sys/contrib/openzfs/.gitignore

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

sys/contrib/openzfs/cmd/zpool/zpool_main.c

@@ -3122,12 +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);
@@ -3198,15 +3207,21 @@ 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);
 
 			if (nvlist_exists(config, ZPOOL_CONFIG_TIMESTAMP))
 				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);
 

sys/contrib/openzfs/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)'

sys/contrib/openzfs/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,

sys/contrib/openzfs/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];

sys/contrib/openzfs/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;
 

sys/contrib/openzfs/include/sys/spa_impl.h

@@ -424,7 +424,9 @@ 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 */
@@ -448,8 +450,6 @@ 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;

sys/contrib/openzfs/man/man4/zfs.4

@@ -402,6 +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,
@@ -2144,6 +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.

sys/contrib/openzfs/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.

sys/contrib/openzfs/module/os/freebsd/zfs/sysctl_os.c

@@ -614,28 +614,6 @@ 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