src/lib, branch vendor/openzfs/master

Make zpool status dedup table support raw bytes -p output

2026-03-13T16:53:56Z

Check if -p flag is enabled, and if so print dedup table with raw bytes. Restructure the logic in zutil_pool to check if -p flag is enabled before printing either the bytes or raw numbers. Calls to print the data for DDT now all use zfs_nicenum_format(). Increased DDT histogram column buffers to 32 bytes to prevent truncation when -p is enabled. Reviewed-by: Brian Behlendorf Reviewed-by: Alexander Motin Signed-off-by: Adi Gollamudi Closes #11626 Closes #17926

Fix the send --exclude option to work with encryption

2026-03-12T22:10:28Z

When using --exclude, filtering needs to take place in two places: in zfs_main.c via the callback previously added to support the options, and in libzfs_sendrecv.c because it generates the nvlist during a first pass, and that results in it complaining if the excluded dataset is not available for sending. (eg, excluding an encrypted dataset so you don't have to use --raw wouldn't work, because the first pass would look at the dataset and decide you couldn't use it.) Add send --exclude tests, including one that tests excluding an encrypted hierarchy. Reviewed-by: Allan Jude Reviewed-by: Alexander Motin Reviewed-by: Brian Behlendorf Signed-off-by: Sean Eric Fagan Closes #18278

libzfs: scrub: only include start and end nv pairs if needed for scrub

2026-03-12T22:06:23Z

This patch addresses running `zpool scrub ` with ZFS 2.4 userspace while the loaded kernel module is still 2.3, failing with: ``` cannot scrub : the loaded zfs module does not support an option for this operation. A reboot may be required to enable this option. ``` Checking for the source of the message via `strace` showed the scrub ioctl failing and setting errno to ZFS_ERR_IOC_ARG_UNAVAIL[0]. With that and the comments in `module/zfs/zfs_ioctl.c`[1] commit: 894edd084 seemed like a likely cause for the backward incompatibility. The corresponding kernelspace code in `module/zfs/zfs_ioctl.c` defaults to a setting of 0 if either parameter is not set, so not providing the nvpairs in case both are 0 should not make a semantic difference. Tested by: * loading zfs.ko in version 2.3.6 * running `zpool scrub testpool` with zpool from master (error occurs) * running `zpool scrub testpool` with this patch applied (scrub is started) This should help users who are still stuck on an older kernel module, while their distribution ships newer ZFS userspace. This was observed in the Proxmox community forum: https://forum.proxmox.com/threads/.180467/ [0] https://github.com/openzfs/zfs/blob/d35951b18d6e12afeb0d5b0539ff2467ab4bfbcf/include/sys/fs/zfs.h#L1762 [1] https://github.com/openzfs/zfs/blob/d35951b18d6e12afeb0d5b0539ff2467ab4bfbcf/module/zfs/zfs_ioctl.c#L7799 Fixes: 894edd084 ("Add TXG timestamp database") Reviewed-by: Alexander Motin Reviewed-by: Brian Behlendorf Signed-off-by: Stoiko Ivanov Co-authored-by: Stoiko Ivanov Closes #18314

zpool clear: remove undocumented rewind flags

2026-03-11T22:15:45Z

Remove the -F, -n, and -X flags from zpool clear. These flags were inherited from OpenSolaris but are not applicable in this context. Unlike zpool import, where the pool is not yet loaded and a specific TXG can be selected, zpool clear operates on an already imported pool whose in-memory state is ahead of what is on disk. Rewinding transactions would require force-exporting the pool first. The rewind policy passed to zpool_clear() is now always ZPOOL_NO_REWIND. Tested on FreeBSD 16.0-CURRENT (amd64). Verified that -F, -n, and -X are properly rejected as invalid options and that the usage output reflects the change. Reviewed-by: Alexander Motin Reviewed-by: Brian Behlendorf Signed-off-by: Christos Longros Closes #13825 Closes #18300

abi: updates for mnttab cleanup

2026-03-10T20:07:07Z

Sponsored-by: TrueNAS Reviewed-by: Ameer Hamza Reviewed-by: Brian Behlendorf Signed-off-by: Rob Norris Closes #18296

libspl/mnttab: remove struct extmnttab

2026-03-10T20:07:07Z

The two additional fields are never used by calling code, and we can replace their sole internal use with an extra stack param. Sponsored-by: TrueNAS Reviewed-by: Ameer Hamza Reviewed-by: Brian Behlendorf Signed-off-by: Rob Norris Closes #18296

libspl/mnttab: remove getmntany()

2026-03-10T20:07:07Z

Only used for when the mount cache was disabled, but since its always enabled now, we don't need it. Sponsored-by: TrueNAS Reviewed-by: Ameer Hamza Reviewed-by: Brian Behlendorf Signed-off-by: Rob Norris Closes #18296

libspl/mnttab: make mnttab source filenames consistent

2026-03-10T20:07:07Z

FreeBSD's getextmntent.c is only separate because it has a different license to mnttab.c, otherwise it would go there too. Sponsored-by: TrueNAS Reviewed-by: Ameer Hamza Reviewed-by: Brian Behlendorf Signed-off-by: Rob Norris Closes #18296

libzfs/mnttab: shorten names, reorg a bit

2026-03-10T20:07:07Z

We can't change the public interface, but internally we don't need so much redundant naming. Sponsored-by: TrueNAS Reviewed-by: Ameer Hamza Reviewed-by: Brian Behlendorf Signed-off-by: Rob Norris Closes #18296

libzfs/mnttab: lift node alloc/free

2026-03-10T20:07:07Z

Sponsored-by: TrueNAS Reviewed-by: Ameer Hamza Reviewed-by: Brian Behlendorf Signed-off-by: Rob Norris Closes #18296