Merge branch 'behelndorf/zfs-2.2-backports'

Signed-off-by: Ameer Hamza <ahamza@ixsystems.com>

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

@@ -0,0 +1,61 @@
+#!/bin/sh
+#
+# 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
+
+echo "off" | tee "$ZEVENT_VDEV_ENC_SYSFS_PATH/power_status"
+
+# Wait for sysfs for report that the slot is off.  It can take ~400ms on some
+# enclosures.
+for i in $(seq 1 20) ; do
+	if [ "$(cat $ZEVENT_VDEV_ENC_SYSFS_PATH/power_status)" == "off" ] ; then
+		break
+	fi
+	sleep 0.1
+done
+
+if [ "$(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

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

@@ -12,11 +12,12 @@ ExecStart=/bin/sh -c '
     decode_root_args || exit 0;                                                                \
     [ "$root" = "zfs:AUTO" ] && root="$(@sbindir@/zpool list -H -o bootfs | grep -m1 -vFx -)"; \
     rootflags="$(getarg rootflags=)";                                                          \
-    case ",$rootflags," in                                                                     \
+    [ "$(@sbindir@/zfs get -H -o value mountpoint "$root")" = legacy ] ||                      \
-        *,zfsutil,*) ;;                                                                        \
+        case ",$rootflags," in                                                                 \
-        ,,) rootflags=zfsutil ;;                                                               \
+            *,zfsutil,*) ;;                                                                    \
-        *)  rootflags="zfsutil,$rootflags" ;;                                                  \
+            ,,) rootflags=zfsutil ;;                                                           \
-    esac;                                                                                      \
+            *)  rootflags="zfsutil,$rootflags" ;;                                              \
+        esac;                                                                                  \
     exec systemctl set-environment BOOTFS="$root" BOOTFSFLAGS="$rootflags"'
 
 [Install]

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/sys/metaslab_impl.h

@@ -313,7 +313,7 @@ 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 +366,9 @@ 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

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_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);
+				err = 0;
-			return ((void *)(uintptr_t)err);
+			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) &&
+		if (send_progress_thread_exit(zhp->zfs_hdl, tid, &oldmask))
-		    send_progress_thread_exit(zhp->zfs_hdl, tid))
 			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) &&
+	if (send_progress_thread_exit(zhp->zfs_hdl, ptid, &oldmask))
-	    send_progress_thread_exit(zhp->zfs_hdl, ptid))
 		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) &&
+		if (send_progress_thread_exit(hdl, tid, &oldmask)) {
-		    send_progress_thread_exit(hdl, tid)) {
 			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) &&
+	if (send_progress_thread_exit(hdl, ptid, &oldmask))
-	    send_progress_thread_exit(hdl, ptid))
 			return (-1);
 
 	if (err == 0 && (flags->props || flags->holds || flags->backup)) {

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

module/os/freebsd/zfs/zfs_vnops_os.c

@@ -6290,7 +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 || error == EOPNOTSUPP)
+	if (error == EXDEV || error == EAGAIN || error == EINVAL ||
+	    error == EOPNOTSUPP)
 		goto bad_locked_fallback;
 	*ap->a_lenp = (size_t)len;
 out_locked:

module/os/linux/zfs/zpl_file_range.c

@@ -103,7 +103,8 @@ zpl_copy_file_range(struct file *src_file, loff_t src_off,
 	 * 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 == -EXDEV)
+	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
@@ -111,7 +112,7 @@ zpl_copy_file_range(struct file *src_file, loff_t src_off,
 	 * 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 == -EXDEV)
+	if (ret == -EINVAL || ret == -EXDEV || ret == -EAGAIN)
 		ret = -EOPNOTSUPP;
 #endif /* HAVE_VFS_GENERIC_COPY_FILE_RANGE */
 

module/zfs/brt.c

@@ -174,7 +174,7 @@
  *	                size_t len, unsigned int flags);
  *
  * Even though offsets and length represent bytes, they have to be
- * block-aligned or we will return the EXDEV error so the upper layer can
+ * block-aligned or we will return an error so the upper layer can
  * fallback to the generic mechanism that will just copy the data.
  * Using copy_file_range(2) will call OS-independent zfs_clone_range() function.
  * This function was implemented based on zfs_write(), but instead of writing
@@ -192,9 +192,9 @@
  * Some special cases to consider and how we address them:
  * - The block we want to clone may have been created within the same
  *   transaction group that we are trying to clone. Such block has no BP
- *   allocated yet, so cannot be immediately cloned. We return EXDEV.
+ *   allocated yet, so cannot be immediately cloned. We return EAGAIN.
  * - The block we want to clone may have been modified within the same
- *   transaction group. We return EXDEV.
+ *   transaction group. We return EAGAIN.
  * - A block may be cloned multiple times during one transaction group (that's
  *   why pending list is actually a tree and not an append-only list - this
  *   way we can figure out faster if this block is cloned for the first time

module/zfs/metaslab.c

@@ -1292,7 +1292,7 @@ metaslab_group_allocatable(metaslab_group_t *mg, metaslab_group_t *rotor,
 
 		/*
 		 * If this metaslab group is below its qmax or it's
-		 * the only allocatable metasable group, then attempt
+		 * the only allocatable metaslab group, then attempt
 		 * to allocate from it.
 		 */
 		if (qdepth < qmax || mc->mc_alloc_groups == 1)

module/zfs/zfs_vnops.c

@@ -1028,6 +1028,10 @@ zfs_exit_two(zfsvfs_t *zfsvfs1, zfsvfs_t *zfsvfs2, const char *tag)
  *
  * On success, the function return the number of bytes copied in *lenp.
  * Note, it doesn't return how much bytes are left to be copied.
+ * On errors which are caused by any file system limitations or
+ * brt limitations `EINVAL` is returned. In the most cases a user
+ * requested bad parameters, it could be possible to clone the file but
+ * some parameters don't match the requirements.
  */
 int
 zfs_clone_range(znode_t *inzp, uint64_t *inoffp, znode_t *outzp,
@@ -1171,7 +1175,7 @@ zfs_clone_range(znode_t *inzp, uint64_t *inoffp, znode_t *outzp,
 	 * We cannot clone into files with different block size.
 	 */
 	if (inblksz != outzp->z_blksz && outzp->z_size > inblksz) {
-		error = SET_ERROR(EXDEV);
+		error = SET_ERROR(EINVAL);
 		goto unlock;
 	}
 
@@ -1179,7 +1183,7 @@ zfs_clone_range(znode_t *inzp, uint64_t *inoffp, znode_t *outzp,
 	 * Offsets and len must be at block boundries.
 	 */
 	if ((inoff % inblksz) != 0 || (outoff % inblksz) != 0) {
-		error = SET_ERROR(EXDEV);
+		error = SET_ERROR(EINVAL);
 		goto unlock;
 	}
 	/*
@@ -1187,7 +1191,7 @@ zfs_clone_range(znode_t *inzp, uint64_t *inoffp, znode_t *outzp,
 	 */
 	if ((len % inblksz) != 0 &&
 	    (len < inzp->z_size - inoff || len < outzp->z_size - outoff)) {
-		error = SET_ERROR(EXDEV);
+		error = SET_ERROR(EINVAL);
 		goto unlock;
 	}
 
@@ -1242,13 +1246,11 @@ zfs_clone_range(znode_t *inzp, uint64_t *inoffp, znode_t *outzp,
 		    &nbps);
 		if (error != 0) {
 			/*
-			 * If we are tyring to clone a block that was created
+			 * If we are trying to clone a block that was created
-			 * in the current transaction group. Return an error,
+			 * in the current transaction group, error will be
-			 * so the caller can fallback to just copying the data.
+			 * EAGAIN here, which we can just return to the caller
+			 * so it can fallback if it likes.
 			 */
-			if (error == EAGAIN) {
-				error = SET_ERROR(EXDEV);
-			}
 			break;
 		}
 		/*

tests/runfiles/linux.run

@@ -41,7 +41,8 @@ tests = ['block_cloning_copyfilerange', 'block_cloning_copyfilerange_partial',
     'block_cloning_ficlonerange_partial',
     'block_cloning_disabled_copyfilerange', 'block_cloning_disabled_ficlone',
     'block_cloning_disabled_ficlonerange',
-    'block_cloning_copyfilerange_cross_dataset']
+    'block_cloning_copyfilerange_cross_dataset',
+    'block_cloning_copyfilerange_fallback_same_txg']
 tags = ['functional', 'block_cloning']
 
 [tests/functional/chattr:Linux]

tests/test-runner/bin/zts-report.py.in

@@ -304,6 +304,8 @@ elif sys.platform.startswith('linux'):
             ['SKIP', cfr_reason],
         'block_cloning/block_cloning_copyfilerange_cross_dataset':
             ['SKIP', cfr_cross_reason],
+        'block_cloning/block_cloning_copyfilerange_fallback_same_txg':
+            ['SKIP', cfr_cross_reason],
     })
 
 

tests/zfs-tests/tests/Makefile.am

@@ -441,9 +441,10 @@ nobase_dist_datadir_zfs_tests_tests_SCRIPTS += \
 	functional/block_cloning/cleanup.ksh \
 	functional/block_cloning/setup.ksh \
 	functional/block_cloning/block_cloning_copyfilerange_cross_dataset.ksh \
+	functional/block_cloning/block_cloning_copyfilerange_fallback.ksh \
+	functional/block_cloning/block_cloning_copyfilerange_fallback_same_txg.ksh \
 	functional/block_cloning/block_cloning_copyfilerange.ksh \
 	functional/block_cloning/block_cloning_copyfilerange_partial.ksh \
-	functional/block_cloning/block_cloning_copyfilerange_fallback.ksh \
 	functional/block_cloning/block_cloning_disabled_copyfilerange.ksh \
 	functional/block_cloning/block_cloning_disabled_ficlone.ksh \
 	functional/block_cloning/block_cloning_disabled_ficlonerange.ksh \

tests/zfs-tests/tests/functional/block_cloning/block_cloning_copyfilerange_fallback_same_txg.ksh

@@ -0,0 +1,66 @@
+#!/bin/ksh -p
+#
+# 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.
+# Copyright (c) 2023, Rob Norris <robn@despairlabs.com>
+#
+
+. $STF_SUITE/include/libtest.shlib
+. $STF_SUITE/tests/functional/block_cloning/block_cloning.kshlib
+
+verify_runnable "global"
+
+if [[ $(linux_version) -lt $(linux_version "4.5") ]]; then
+  log_unsupported "copy_file_range not available before Linux 4.5"
+fi
+
+claim="copy_file_range will fall back to copy when cloning on same txg"
+
+log_assert $claim
+
+typeset timeout=$(get_tunable TXG_TIMEOUT)
+
+function cleanup
+{
+	datasetexists $TESTPOOL && destroy_pool $TESTPOOL
+	set_tunable64 TXG_TIMEOUT $timeout
+}
+
+log_onexit cleanup
+
+log_must zpool create -o feature@block_cloning=enabled $TESTPOOL $DISKS
+
+log_must set_tunable64 TXG_TIMEOUT 5000
+
+log_must dd if=/dev/urandom of=/$TESTPOOL/file bs=128K count=4
+log_must clonefile -f /$TESTPOOL/file /$TESTPOOL/clone 0 0 524288
+
+log_must sync_pool $TESTPOOL
+
+log_must have_same_content /$TESTPOOL/file /$TESTPOOL/clone
+
+typeset blocks=$(unique_blocks $TESTPOOL file $TESTPOOL clone)
+log_must [ "$blocks" = "" ]
+
+log_pass $claim
+