zts: add spdx license tags to default_quota tests
Signed-off-by: Ameer Hamza <ahamza at ixsystems.com>
Reviewed-by: Alexander Motin <mav at FreeBSD.org>
Reviewed-by: Tony Hutter <hutter2 at llnl.gov>
Add tests for defaultprojectquota
Extend project quota test coverage to verify defaultprojectquota
behavior. These build on existing project quota tests with additional
cases specific to defaultprojectquota functionality.
Signed-off-by: Ameer Hamza <ahamza at ixsystems.com>
Reviewed-by: Alexander Motin <mav at FreeBSD.org>
Reviewed-by: Tony Hutter <hutter2 at llnl.gov>
Add tests for default user/group quota functionality
Extend test coverage to verify default user and group quota
functionality. These build on existing user/group quota tests with
additional cases specific to default quotas functionality.
Added on top of: https://github.com/openzfs/zfs/pull/16283/commits/e08cd97
Signed-off-by: Todd Seidelmann <seidelma at wharton.upenn.edu>
Signed-off-by: Ameer Hamza <ahamza at ixsystems.com>
Reviewed-by: Alexander Motin <mav at FreeBSD.org>
Reviewed-by: Tony Hutter <hutter2 at llnl.gov>
zfsprops.7 manpage changes for default quotas
Signed-off-by: Ameer Hamza <ahamza at ixsystems.com>
Reviewed-by: Alexander Motin <mav at FreeBSD.org>
Reviewed-by: Tony Hutter <hutter2 at llnl.gov>
Show default quotas in zfs userspace tools
Update zfs userspace, groupspace, and projectspace to display the
default quotas when no per-ID specific quota is configured. This
ensures tool outputs align with enforced limits.
Signed-off-by: Ameer Hamza <ahamza at ixsystems.com>
Reviewed-by: Alexander Motin <mav at FreeBSD.org>
Reviewed-by: Tony Hutter <hutter2 at llnl.gov>
Report default quotas via kernel interfaces
Ensure default user/group/project quotas are visible through quota
tools and filesystem stats when no per-ID quota is configured. This
maintains consistency between quota visibility and configured defaults.
Signed-off-by: Ameer Hamza <ahamza at ixsystems.com>
Reviewed-by: Alexander Motin <mav at FreeBSD.org>
Reviewed-by: Tony Hutter <hutter2 at llnl.gov>
Enforce default quotas when no per-ID quota is set
Update zfs_id_overobjquota() and zfs_id_overblockquota() to enforce
default user/group/project quotas (block and object-based) when no
per-user, per-group, or per-project quota exists. If a specific quota
is not configured for an ID, the default quota value is applied.
Signed-off-by: Ameer Hamza <ahamza at ixsystems.com>
Reviewed-by: Alexander Motin <mav at FreeBSD.org>
Reviewed-by: Tony Hutter <hutter2 at llnl.gov>
Add default user/group/project quota properties
This adds default userquota, groupquota, and projectquota properties to
MASTER_NODE_OBJ to make them accessible during zfsvfs_init() (regular
DSL properties require dsl_config_lock, which cannot be safely acquired
in this context). The zfs_fill_zplprops_impl() logic is updated to read
these default properties directly from MASTER_NODE_OBJ.
Signed-off-by: Ameer Hamza <ahamza at ixsystems.com>
Reviewed-by: Alexander Motin <mav at FreeBSD.org>
Reviewed-by: Tony Hutter <hutter2 at llnl.gov>
Fix nonrot property being incorrectly unset (#17206)
When opening a vdev and setting the nonrot property, we used to wait for
each child to be opened before examining its nonrot property. When the
change was made to open vdevs asynchronously, we didn't move the nonrot
check out of the main loop. As a result, the nonrot property is almost
always set to false, regardless of the actual type of the underlying
disks. The fix is simply to move the nonrot check to a separate loop
after the taskq has been waited for.
Sponsored-by: Klara, Inc.
Sponsored-by: Eshtek, Inc.
Signed-off-by: Paul Dagnelie <paul.dagnelie at klarasystems.com>
Co-authored-by: Paul Dagnelie <paul.dagnelie at klarasystems.com>
Reviewed-by: Tony Hutter <hutter2 at llnl.gov>
Reviewed-by: Alexander Motin <mav at FreeBSD.org>
contrib/initramfs: use LVM autoactivation for activating VGs (#17125)
Currently, the zfs initramfs-tools boot script under local-top calls
`vgchange -ay`, which unconditionally activates all logical volumes
(LVs) in all discovered volume groups (VGs). This causes all LVs to be
active after boot. However, users may prefer to not activate certain
VGs/LVs on boot. They might normally use the `--setautoactivation n`
VG/LV flag or the `auto_activation_volume_list` LVM config option to
achieve this, but since the script unconditionally activates all LVs,
neither has an effect.
To fix this, call `vgchange -aay` instead. This triggers LVM
autoactivation, which honors autoactivation settings such as the
`--setautoactivation` flag. It is also more in line with the LVM
documentation, which says autoactivation is "meant to be used by
activation commands that are run automatically by the system" [1].
Note that this change might break misconfigured setups that have ZFS
on top of an LV for which autoactivation is disabled.
[7 lines not shown]
Multiple printf() size fixes (#17199)
cmd/zinject/zinject.c:
- use PRIu64 when printing uint64_t
tests/zfs-tests/cmd/clonefile.c:
- use an unsigned long long to store result from strtoull()
- use %jd for printing off_t, %zu for size_t, %zd for ssize_t
tests/zfs-tests/tests/functional/vdev_disk/page_alignment.c:
- use %zx to print size_t
Discovered when compiling on FreeBSD i386.
Signed-off-by: Martin Matuska <mm at FreeBSD.org>
Reviewed-by: Rob Norris <robn at despairlabs.com>
Reviewed-by: Alexander Motin <mav at FreeBSD.org>
Reviewed-by: Allan Jude <allan at klarasystems.com>
Reviewed-by: @ImAwsumm
Fix lock reversal on device removal cancel
FreeBSD kernel's WITNESS code detected lock ordering violation in
spa_vdev_remove_cancel_sync(). It took svr_lock while holding
ms_lock, which is opposite to other places. I was thinking to
resolve it similar to #17145, but looking closer I don't think
we even need svr_lock at that point, since we already asserted
svr_allocd_segs is empty, and we don't need to add there segments
we are going to call free_mapped_segment_cb for.
Reviewed-by: Tony Hutter <hutter2 at llnl.gov>
Reviewed-by: Allan Jude <allan at klarasystems.com>
Signed-off-by: Alexander Motin <mav at FreeBSD.org>
Sponsored by: iXsystems, Inc.
Closes #17164
Fix dspace underflow bug
Since spa_dspace accounts only normal allocation class space,
spa_nonallocating_dspace should do the same. Otherwise we may get
negative overflow or respective assertion spa_update_dspace() if
removed special/dedup vdev is bigger than all normal class space.
Reviewed-by: Alexander Motin <mav at FreeBSD.org>
Reviewed-by: Allan Jude <allan at klarasystems.com>
Signed-off-by: Paul Dagnelie <paul.dagnelie at klarasystems.com>
Closes #17183
simd_powerpc.h: enable FPU on FreeBSD
FreeBSD nowadays supports FPU in the kernel on powerpc*, so enable it.
Reviewed-by: Alexander Motin <mav at FreeBSD.org>
Reviewed-by: Tony Hutter <hutter2 at llnl.gov>
Reviewed-by: Brian Behlendorf <behlendorf1 at llnl.gov>
Signed-off-by: Piotr Kubaj <pkubaj at FreeBSD.org>
Closes #17191
kstat: silence "maybe uninitialized" warnings
Firmly in the "shouldn't happen" camp, but at least GCC 7.4 (Ubuntu
18.04) complained about them, and it's easy to shut up, so do so.
Sponsored-by: Klara, Inc.
Sponsored-by: Wasabi Technology, Inc.
Reviewed-by: Alexander Motin <mav at FreeBSD.org>
Signed-off-by: Rob Norris <rob.norris at klarasystems.com>
Closes #17189
Remove mg_allocators (#17192)
Previous code allowed each metaslab group to have different number
of allocators. But in practice it worked only for embedded SLOGs,
relying on a number of conditions and creating a significant mine
field if any of those change. I just stepped on one myself.
This change makes all groups to have spa_alloc_count allocators.
It may cost us extra 192 bytes of memory per normal top-level vdev
on large systems, but I find it a small price for cleaner and more
reliable code.
Signed-off-by: Alexander Motin <mav at FreeBSD.org>
Sponsored by: iXsystems, Inc.
Fixes #17188
Reviewed-by: Tony Hutter <hutter2 at llnl.gov>
Reviewed-by: Paul Dagnelie <pcd at delphix.com>
zed: Ensure spare activation after kernel-initiated device removal
In addition to hotplug events, the kernel may also mark a failing vdev
as REMOVED. This was observed in a customer report and reproduced by
forcing the NVMe host driver to disable the device after a failed reset
due to command timeout. In such cases, the spare was not activated
because the device had already transitioned to a REMOVED state before
zed processed the event.
To address this, explicitly attempt hot spare activation when the
kernel marks a device as REMOVED.
Reviewed-by: Alexander Motin <mav at FreeBSD.org>
Reviewed-by: Tony Hutter <hutter2 at llnl.gov>
Signed-off-by: Ameer Hamza <ahamza at ixsystems.com>
Closes #17187
config: cache results of kernel checks (#17106)
Kernel checks are the heaviest part of the configure checks. This allows
the results to be cached through the normal autoconf cache.
Since we don't want to reuse cached values for different kernels, but
don't want to discard the entire cache on every kernel, we instead add a
short checksum to kernel config cache keys, based on the version and
path, so the cache can hold results for multiple different kernels.
Sponsored-by: https://despairlabs.com/sponsor/
Signed-off-by: Rob Norris <robn at despairlabs.com>
Reviewed-by: Tony Hutter <hutter2 at llnl.gov>
Block remap for cloned blocks on device removal
When after device removal we handle block pointers remap, skip blocks
that might be cloned. BRTs are indexed by vdev id and offset from
block pointer's DVA[0]. So if we start addressing the same block by
some different DVA, we won't get the proper reference counter. As
result, we might either remap the block twice, that may result in
assertion during indirect mapping condense, or free it prematurely,
that may result in data overwrite, or free it twice, that may result
in assertion in spacemap code.
Reviewed-by: Ameer Hamza <ahamza at ixsystems.com>
Reviewed-by: Paul Dagnelie <pcd at delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1 at llnl.gov>
Signed-off-by: Alexander Motin <mav at FreeBSD.org>
Sponsored by: iXsystems, Inc.
Closes #15604
Closes #17180
runners: Fix tarball build for zfs-qemu-packages workflow (#17158)
The initial tarballs we built for for zfs-2.3.1 were incorrect since
they did not have a ./configure script, and their files were not
in a top level zfs-2.3.1/ directory. This commit copies the way we
built them on buildbot so the tarballs are created as expected.
Signed-off-by: Tony Hutter <hutter2 at llnl.gov>
Reviewed-by: Tino Reichardt <milky-zfs at mcmilk.de>
Reviewed-by: Brian Behlendorf <behlendorf1 at llnl.gov>
runners: Fix zfs-release RPM creation (#17173)
The zfs-qemu-packages workflow was incorrectly copying the built
zfs-release RPMs to ~/zfsonlinux.github.com rather than ~/zfs. This
meant that the RPMs were not being correctly picked in the artifacts
files. This fixes the issue.
Signed-off-by: Tony Hutter <hutter2 at llnl.gov>
Reviewed-by: Brian Behlendorf <behlendorf1 at llnl.gov>
Reviewed-by: @ImAwsumm
Reviewed-by: Tino Reichardt <milky-zfs at mcmilk.de>
Linux: Fix zfs_prune panics v2 (#17121)
It turns out that approach taken in the original version of the patch
was wrong. So now, we're taking approach in-line with how kernel
actually does it - when sb is being torn down, access to it
is serialized via sb->s_umount rwsem, only when that lock is taken
is it okay to work with s_flags - and the other mistake I was doing
was trying to make SB_ACTIVE work, but apparently the kernel checks
the negative variant - not SB_DYING and not SB_BORN.
Kernels pre-6.6 don't have SB_DYING, but check if sb is hashed
instead.
Signed-off-by: Pavel Snajdr <snajpa at snajpa.net>
Reviewed-by: Tony Hutter <hutter2 at llnl.gov>
Reviewed-by: Brian Behlendorf <behlendorf1 at llnl.gov>
Linux 6.14 compat: META (#17098) (#17172)
Update the META file to reflect compatibility with the 6.14
kernel.
Signed-off-by: Tony Hutter <hutter2 at llnl.gov>
Reviewed-by: Rob Norris <robn at despairlabs.com>
Reviewed-by: Brian Behlendorf <behlendorf1 at llnl.gov>
Reviewed-by: @ImAwsumm
ZTS: Fix zpool_status_features_001_pos local test (#17174)
Update 'zfs-helpers.sh -i' to install the compatibility.d/ file
symlinks. These are need to run the zpool_status_features_001_pos test
from a local workspace (as opposed to running ZTS from a formal
'make install' or install from RPMs, which are unaffected).
Signed-off-by: Tony Hutter <hutter2 at llnl.gov>
Reviewed-by: @ImAwsumm
Reviewed-by: Alexander Motin <mav at FreeBSD.org>
Disambiguate reference to kibibytes, not kilobytes
A minor nitpick that is kind of obvious based on the surrounding context
and reference to powers of two. It's better to be explicit, though.
Signed-off-by: Simon Howard <fraggle at gmail.com>
Reviewed-by: Tony Hutter <hutter2 at llnl.gov>
Reviewed-by: Alexander Motin <mav at FreeBSD.org>
Fix spelling errors
Unlike some of my other fixes which are more subtle, these are
unambigously spelling errors.
Signed-off-by: Simon Howard <fraggle at gmail.com>
Reviewed-by: Tony Hutter <hutter2 at llnl.gov>
Reviewed-by: Alexander Motin <mav at FreeBSD.org>
Correct "umount" to "unmount" in a couple of places
This is admittedly a nitpicky change, but `umount` is the command that
performs an *unmount*. So if we are talking about unmounting something
we should phrase it that way.
Signed-off-by: Simon Howard <fraggle at gmail.com>
Reviewed-by: Tony Hutter <hutter2 at llnl.gov>
Reviewed-by: Alexander Motin <mav at FreeBSD.org>
Capitalize in various places where appropriate
These are mostly acronyms (CPUs; ZILs) but also proper nouns such as
"Unix" and "Unicode" which should also be capitalized.
Signed-off-by: Simon Howard <fraggle at gmail.com>
Reviewed-by: Tony Hutter <hutter2 at llnl.gov>
Reviewed-by: Alexander Motin <mav at FreeBSD.org>
