Avoid checking the whole array of objects each time by removing the self
organized memory reaping. this can be managed by the global memory reap
callback which is called every 60 seconds. this will reduce the use if
locking operations significant.
Reviewed-by: Kjeld Schouten <kjeld@schouten-lebbing.nl> Reviewed-by: Mateusz Guzik <mjguzik@gmail.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Sebastian Gottschall <s.gottschall@dd-wrt.com>
Closes #11126
Ryan Moeller [Sat, 31 Oct 2020 16:37:56 +0000 (12:37 -0400)]
Remove duplicate cond_resched() definition
Reviewed-by: Alexander Motin <mav@FreeBSD.org> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Ryan Moeller <ryan@iXsystems.com>
Closes #11131
Ryan Moeller [Wed, 21 Oct 2020 18:09:17 +0000 (18:09 +0000)]
zvol_os: Fix handling of zvol private data
zvol private data is supposed to be nulled by zvol_clear_private before
zvol_free is called as an indicator that the zvol is going away.
Implement zvol_clear_private for volmode=dev.
Assert that zvol_clear_private has been called before zvol_free.
Check that zvol_clear_private has not been called when updating
volsize. If it has, fail with ENXIO.
Reviewed-by: Alexander Motin <mav@FreeBSD.org> Reviewed-by: Matt Macy <mmacy@FreeBSD.org> Signed-off-by: Ryan Moeller <ryan@iXsystems.com>
Closes #11117
Ryan Moeller [Wed, 21 Oct 2020 18:05:18 +0000 (18:05 +0000)]
zvol_os: Don't leak doi in cdev error path
Make sure to free doi in zvol_create_minor impl when make_dev_s fails.
Reviewed-by: Alexander Motin <mav@FreeBSD.org> Reviewed-by: Matt Macy <mmacy@FreeBSD.org> Signed-off-by: Ryan Moeller <ryan@iXsystems.com>
Closes #11117
Ryan Moeller [Wed, 21 Oct 2020 17:59:15 +0000 (17:59 +0000)]
zvol_os: Properly ignore error in volmode lookup
We fall back to a default volmode and continue when looking up a zvol's
volmode property fails. After this we should set the error to 0 to
ensure we take the success paths in the out section.
While here, make sure we only log that the zvol was created on success.
Reviewed-by: Alexander Motin <mav@FreeBSD.org> Reviewed-by: Matt Macy <mmacy@FreeBSD.org> Signed-off-by: Ryan Moeller <ryan@iXsystems.com>
Closes #11117
Ryan Moeller [Wed, 21 Oct 2020 17:54:38 +0000 (17:54 +0000)]
zvol_os: Code cleanup in zvol_create_minor_impl
Nonfunctional changes for readability and consistency.
Reviewed-by: Alexander Motin <mav@FreeBSD.org> Reviewed-by: Matt Macy <mmacy@FreeBSD.org> Signed-off-by: Ryan Moeller <ryan@iXsystems.com>
Closes #11117
Ryan Moeller [Wed, 21 Oct 2020 17:46:48 +0000 (17:46 +0000)]
zvol_os: Keep better track of open count in close
zvol_geom_close gets a count of the number of close operations to do.
Make sure we're always using this count to check if this will be the
last close operation performed on the zvol.
Reviewed-by: Alexander Motin <mav@FreeBSD.org> Reviewed-by: Matt Macy <mmacy@FreeBSD.org> Signed-off-by: Ryan Moeller <ryan@iXsystems.com>
Closes #11117
Ryan Moeller [Wed, 21 Oct 2020 17:23:08 +0000 (17:23 +0000)]
zvol_os: Tidy up asserts
Using more specific assert variants gives better messages on failure.
No functional change.
Reviewed-by: Alexander Motin <mav@FreeBSD.org> Reviewed-by: Matt Macy <mmacy@FreeBSD.org> Signed-off-by: Ryan Moeller <ryan@iXsystems.com>
Closes #11117
Mateusz Guzik [Fri, 30 Oct 2020 22:26:10 +0000 (23:26 +0100)]
zstd: track allocator statistics
Note that this only tracks sizes as requested by the caller.
Actual allocated space will almost always be bigger (e.g., rounded up to
the next power of 2 or page size). Additionally the allocated buffer may
be holding other areas hostage. Nonetheless, this is a starting point
for tracking memory usage in zstd.
Reviewed-by: Allan Jude <allan@klarasystems.com> Reviewed-by: Ryan Moeller <ryan@ixsystems.com> Reviewed-by: Kjeld Schouten <kjeld@schouten-lebbing.nl> Signed-off-by: Mateusz Guzik <mjguzik@gmail.com>
Closes #11129
Attila Fülöp [Fri, 30 Oct 2020 22:24:21 +0000 (23:24 +0100)]
ICP: gcm: Allocate hash subkey table separately
While evaluating other assembler implementations it turns out that
the precomputed hash subkey tables vary in size, from 8*16 bytes
(avx2/avx512) up to 48*16 bytes (avx512-vaes), depending on the
implementation.
To be able to handle the size differences later, allocate
`gcm_Htable` dynamically rather then having a fixed size array, and
adapt consumers.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Attila Fülöp <attila@fueloep.org>
Closes #11102
Ryan Moeller [Fri, 30 Oct 2020 22:19:02 +0000 (18:19 -0400)]
FreeBSD: Fix 12.2-STABLE after AT_BENEATH MFC
AT_BENEATH was merged to stable/12, where kern_unlinkat takes a
non-const path. DECONST the path passed to kern_unlinkat in the
case where AT_BENEATH is defined.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Ryan Moeller <ryan@iXsystems.com>
Closes #11139
Alexander Motin [Fri, 30 Oct 2020 15:57:54 +0000 (11:57 -0400)]
Yield periodically when rebuilding L2ARC
L2ARC devices of several terabytes filled with 4KB blocks may take 15
minutes to rebuild. Due to the way L2ARC log reading is implemented
it is quite likely that for all that time rebuild thread will never
sleep. At least on FreeBSD kernel threads have absolute priority and
can not be preempted by threads with lower priorities. If some thread
is also bound to that specific CPU it may not get any CPU time for all
the 15 minutes.
Reviewed-by: Cedric Berger <cedric@precidata.com> Reviewed-by: Ryan Moeller <freqlabs@FreeBSD.org> Reviewed-by: George Amanakis <gamanakis@gmail.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Closes #11116
Ryan Moeller [Fri, 30 Oct 2020 15:55:59 +0000 (11:55 -0400)]
Update references to nonexistent man pages in code
Refer to the correct section or alternative for FreeBSD and Linux.
Reviewed-by: George Melikov <mail@gmelikov.ru> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Ryan Moeller <ryan@iXsystems.com>
Closes #11132
Alexander Motin [Fri, 30 Oct 2020 15:50:57 +0000 (11:50 -0400)]
FreeBSD: Remove BIO_ORDERED flag from BIO_FLUSH
ZFS always waits for the write completion before flushing the cache.
That is why it does not require explicit ordering fences around it,
which are pretty difficult to implement for NVMe, since one has no
internal concept of strict request ordering.
This was already removed from FreeBSD once, but got resurrected
by mistake during OpenZFS merge.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Ryan Moeller <ryan@iXsystems.com> Signed-off-by: Alexander Motin <mav@FreeBSD.org>
Closes #11130
Tony Hutter [Fri, 30 Oct 2020 15:47:42 +0000 (08:47 -0700)]
ZTS: Fix xattr_004_pos failure, don't use tmpfs
Previously, xattr_004_pos would create files with xattrs on both
tmpfs and ext2, and then copy them to zfs to verify that their
xattrs were preserved. However tmpfs doesn't support xattrs.
This was never noticed until Fedora 33. In Fedora 32 and older,
/tmp was on the root partition (like ext4), whereas on Fedora 33
/tmp is actually tmpfs. That caused this test to fail on Fedora 33.
This fix updates the test to only create the file on ext2, not tmpfs.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Tony Hutter <hutter2@llnl.gov>
Closes #11133
Mateusz Guzik [Thu, 29 Oct 2020 16:54:20 +0000 (17:54 +0100)]
Linux: g/c leftover fence in zfs_znode_alloc
The port removed provisions for zfs_znode_move but the cleanup missed
this bit. To quote the original:
[snip]
list_insert_tail(&zfsvfs->z_all_znodes, zp);
membar_producer();
/*
* Everything else must be valid before assigning z_zfsvfs makes the
* znode eligible for zfs_znode_move().
*/
zp->z_zfsvfs = zfsvfs;
[/snip]
In the current code it is immediately followed by unlock which issues
the same fence, thus plays no role in correctness.
Reviewed-by: Matt Macy <mmacy@FreeBSD.org> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Mateusz Guzik <mjguzik@gmail.com>
Closes #11115
Brian Behlendorf [Tue, 27 Oct 2020 16:43:36 +0000 (09:43 -0700)]
Use known license string for zlua
The Linux kernel MODULE_LICENSE macro only recognizes a handful of
license strings and "MIT" is not one of the them. Update the macro
to use "Dual MIT/GPL" which is recognized and what the kernel expects
MIT licensed modules to use.
Reviewed-by: George Melikov <mail@gmelikov.ru> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes #11112
Closes #11113
Reviewed-by: Matt Macy <mmacy@FreeBSD.org> Reviewed-by: Toomas Soome <tsoome@me.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed off by: Cy Schubert <cy@FreeBSD.org>
Closes #11088
xtouqh [Thu, 22 Oct 2020 18:28:10 +0000 (21:28 +0300)]
Properly format NAME subsection of zfs/zpool subcommands
Use proper names (i.e. zfs-allow and zpool-add) in NAME subsections
of zfs/zpool subcommands instead of current "pretty-printed" ones as
makewhatis utilities (or some implementations of it, namely the one
from mandoc suite used in FreeBSD) look not only at the document title
but also in NAME subsection, adding zfs(8)/zpool(8) to search results
which is not correct. (Common sense and other utilities splitting
subcommands in multiple man pages, e.g. git, do the same.)
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: xtouqh <xtouqh@hotmail.com>
Closes #11086
Ryan Moeller [Thu, 22 Oct 2020 17:18:26 +0000 (13:18 -0400)]
Add missing zfs_arc_evict_batch_limit tunable
It's even documented already.
Reviewed-by: Alexander Motin <mav@FreeBSD.org> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Ryan Moeller <ryan@iXsystems.com>
Closes #11094
Ryan Moeller [Wed, 21 Oct 2020 21:09:14 +0000 (17:09 -0400)]
arcstat: Add -a and -p options from FreeNAS
Added -a option to automatically print all valid statistics.
Added -p option to suppress scaling of printed data.
Reviewed-by: George Melikov <mail@gmelikov.ru> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Authored by: Nick Principe <32284693+powernap@users.noreply.github.com> Ported-by: Ryan Moeller <ryan@iXsystems.com> Signed-off-by: Ryan Moeller <ryan@iXsystems.com>
Closes #11090
Kyle Evans [Tue, 20 Oct 2020 18:34:59 +0000 (13:34 -0500)]
Makefile.bsd: remove directory that no longer exists
This was removed in a reorganization of directories preparing for the
merge of FreeBSD support, 006e9a408824 by mmacy. While llvm is perfectly
happy with the nonexistent -I directory, the gcc6 and gcc9 we can elect
to use as cross-toolchains both trip over it.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Ryan Moeller <ryan@iXsystems.com> Signed-off-by: Kyle Evans <kevans@FreeBSD.org>
Closes #11077
Matthew Macy [Tue, 20 Oct 2020 15:53:16 +0000 (08:53 -0700)]
FreeBSD: delete unreferenced file
zfs_onexit_os.c was not deleted when it was removed from the build
Reviewed-by: Matt Ahrens <matt@delphix.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Ryan Moeller <ryan@iXsystems.com> Signed-off-by: Matt Macy <mmacy@FreeBSD.org>
Closes #11079
Don Brady [Mon, 19 Oct 2020 18:01:00 +0000 (12:01 -0600)]
zed syslog entries drop important info
ZED will log zevents summaries to the syslog, however the log entries
tend to drop event details that can be useful for diagnosis. This is
especially true for ereport events, like io, checksum, and delay.
Update the all-syslog.sh script to log additional event information.
Add an optional config option, ZED_SYSLOG_DISPLAY_GUIDS, to zed.rc
for choosing GUIDs over names for pool and vdev.
Change the default ZED_SYSLOG_SUBCLASS_EXCLUDE to exclude history_event
events. These events tend to be frequent, convey no meaningful info,
and are already logged in the zpool history.
Reviewed-by: John Kennedy <john.kennedy@delphix.com> Reviewed-by: Pavel Zakharov <pavel.zakharov@delphix.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Don Brady <don.brady@delphix.com>
Closes #10967
Don Brady [Thu, 15 Oct 2020 21:45:16 +0000 (15:45 -0600)]
Ignore special vdev ashift for spa ashift min/max
The removal of a vdev in the normal class would fail if there was a
special or deup vdev that had a different ashift than the vdevs in
the normal class.
Moved the initialization of spa_min_ashift / spa_max_ashift from
vdev_open so that it occurs after the vdev allocation bias was
initialized (i.e. after vdev_load).
Caveat -- In order to remove a special/dedup vdev it must have the
same ashift as the normal pool vdevs. This could perhaps be lifted
in the future (i.e. for the case where there is ample space in any
surviving special class vdevs)
Reviewed-by: Matthew Ahrens <mahrens@delphix.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Don Brady <don.brady@delphix.com>
Closes #9363
Closes #9364
Closes #11053
Fix crash caused by invalid snapshot names in redactnvl
This is a follow up fix for commit 0fdd6106bb. The VERIFY is
only true when we haven't hit an error code path. See added
test case for a reproducer.
Reviewed-by: Matthew Ahrens <mahrens@delphix.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Christian Schwarz <me@cschwarz.com>
Closes #11048
Paul Dagnelie [Wed, 14 Oct 2020 15:59:54 +0000 (08:59 -0700)]
Fix incorrect deletion order in range_tree_add_impl gap case
After a side-effectful call like add or remove, references to range
segs stored in btrees can no longer be used safely. We move the
remove call to just before the reinsertion call so that the seg
remains valid for as long as we need it.
Reviewed-by: Matthew Ahrens <mahrens@delphix.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Paul Dagnelie <pcd@delphix.com>
Closes #11044
Closes #11056
Mateusz Guzik [Wed, 14 Oct 2020 15:57:03 +0000 (17:57 +0200)]
FreeBSD: fix panic due to tqid overflow
The 32-bit counter eventually wraps to 0 which is a sentinel for invalid
id.
Make it 64-bit on LP64 platforms and 0-check otherwise.
Note: Linux counterpart uses id stored per queue instead of a global.
I did not check going that way is feasible with the goal being the
minimal fix doing the job.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Ryan Moeller <ryan@iXsystems.com> Signed-off-by: Mateusz Guzik <mjguzik@gmail.com>
Closes #11059
Ryan Moeller [Wed, 14 Oct 2020 04:25:48 +0000 (00:25 -0400)]
Cross-platform acltype
The acltype property is currently hidden on FreeBSD and does not
reflect the NFSv4 style ZFS ACLs used on the platform. This makes it
difficult to observe that a pool imported from FreeBSD on Linux has a
different type of ACL that is being ignored, and vice versa.
Add an nfsv4 acltype and expose the property on FreeBSD.
Make the default acltype nfsv4 on FreeBSD.
Setting acltype to an unhanded style is treated the same as setting
it to off. The ACLs will not be removed, but they will be ignored.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Ryan Moeller <ryan@iXsystems.com>
Closes #10520
Warner Losh [Wed, 14 Oct 2020 04:05:49 +0000 (22:05 -0600)]
FreeBSD: make adjustments for the standalone environment
In FreeBSD, there are three compile environments that are supported:
user land, the kernel and the bootloader / standalone. Adjust the
headers to compile in the standalone environment. Limit kernel-only
items from view when _STANDALONE is defined.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Ryan Moeller <ryan@iXsystems.com> Signed-off-by: Warner Losh <imp@FreeBSD.org>
Closes #10998
Warner Losh [Wed, 14 Oct 2020 04:01:40 +0000 (22:01 -0600)]
aarch64: Use proper guards for NEON instructions
The zstd code assumes that if you are on aarch64, you have NEON
instructions. This is not necessarily true. In a boot loader, where
you might not have the VFP properly initialized, these instructions
may not be available. It's also an error to include arm_neon.h when
the NEON insturctions aren't enabled. Change the guards for using the
NEON instructions from __aarch64__ to __ARM_NEON which is the standard
symbol for knowing if they are available.
__ARM_NEON is the proper symbol, defined in ARM C Language Extensions
Release 2.1 (https://developer.arm.com/documentation/ihi0053/d/). Some
sources suggest __ARM_NEON__, but that's the obsolete spelling from
prior versions of the standard.
Updated based on zstd pull request https://github.com/facebook/zstd/pull/2356
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Warner Losh <imp@bsdimp.com>
Closes #11055
dmu.h: remove stale declaration dmu_objset_snapshot_tmp
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Adam Moss <c@yotes.com> Signed-off-by: Christian Schwarz <me@cschwarz.com>
Closes #11047
Ryan Moeller [Tue, 13 Oct 2020 16:38:40 +0000 (12:38 -0400)]
FreeBSD: Improve libzfs_error_init messages
It is a common mistake to have failed to autoload the module due to
permission issues when running a ZFS command as a user. "Operation
not permitted" is an unhelpfully vague error message.
Use a thread-local message buffer to format a nicer error message.
We can infer that loading the kernel module failed if the module is
not loaded. This can be extended with heuristics for other errors
in the future.
While looking at this stuff, remove an unused thread-local message
buffer found in libspl and remove some inaccurate verbiage from the
comment on libzfs_load_module.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Ryan Moeller <ryan@iXsystems.com>
Closes #11033
Ryan Moeller [Tue, 13 Oct 2020 16:32:34 +0000 (12:32 -0400)]
Expose zfetch_max_idistance tunable
FreeBSD had this value tunable before the switch to the new OpenZFS.
The tunable name has changed, breaking legacy compat.
Restore legacy compat for this tunable, properly expose the tunable
with the new name on all platforms, and document it in
zfs-module-parameters(5).
While here, clean up the documentation for zfetch_max_distance a bit.
Reviewed-by: Alexander Motin <mav@FreeBSD.org> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Ryan Moeller <ryan@iXsystems.com>
Closes #11038
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Ryan Moeller <ryan@iXsystems.com> Co-authored-by: Ryan Moeller <ryan@freqlabs.com> Signed-off-by: Christian Schwarz <me@cschwarz.com>
Closes #11020
Ryan Moeller [Fri, 9 Oct 2020 16:27:14 +0000 (12:27 -0400)]
Linux: Initialize zp in zfs_setattr_dir
The value of zp is used without having been initialized under some
conditions. Initialize the pointer to NULL.
Add a regression test case using chown in acl/posix. However, this is
not enough because the setup sets xattr=sa, which means zfs_setattr_dir
will not be called. Create a second group of acl tests in acl/posix-sa
duplicating the acl/posix tests with symlinks, and remove xattr=sa from
the original acl/posix tests. This provides more coverage for the
default xattr=on code.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Ryan Moeller <ryan@iXsystems.com>
Closes #10043
Closes #11025
This change updates the documentation to refer to the project
as OpenZFS instead ZFS on Linux. Web links have been updated
to refer to https://github.com/openzfs/zfs. The extraneous
zfsonlinux.org web links in the ZED and SPL sources have been
dropped.
Reviewed-by: George Melikov <mail@gmelikov.ru> Reviewed-by: Richard Laager <rlaager@wiktel.com> Reviewed-by: Ryan Moeller <ryan@iXsystems.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes #11007
On gcc (GCC) 9.2.1 20190827 (Red Hat 9.2.1-1), the code removed
caused `-Wmaybe-uninitialized` errors.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Ryan Moeller <ryan@iXsystems.com> Signed-off-by: Christian Schwarz <me@cschwarz.com>
Closes #11021
Toomas Soome [Mon, 5 Oct 2020 21:05:28 +0000 (00:05 +0300)]
zdb should not output binary data on terminal
The zdb is interpreting byte array as textual string in dump_zap,
but there are also binary arrays and we should not output binary
data on terminal.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Igor Kozhukhov <igor@dilos.org> Signed-off-by: Toomas Soome <tsoome@me.com>
External-issue: https://www.illumos.org/issues/12012
External-issue: https://www.illumos.org/issues/11713
Closes #11006
Ryan Moeller [Sat, 3 Oct 2020 00:48:45 +0000 (20:48 -0400)]
FreeBSD: Sort out kernel FPU headers for 12.1-REL
We were missing an include for kernel FPU functions, breaking the build
on FreeBSD 12.1-RELEASE. This was apparently being pulled in from
elsewhere on stable/12 and head.
Sorted the other includes in these files while here.
Reviewed-by: Alexander Motin <mav@FreeBSD.org> Signed-off-by: Ryan Moeller <ryan@iXsystems.com>
Closes #11005
Alan Somers [Sat, 3 Oct 2020 00:47:09 +0000 (18:47 -0600)]
Fix EIO after resuming receive of new dataset over an existing one
When resuming an interrupted ZFS send stream that creates a new dataset
with the same name as an existing dataset, if the existing dataset is
accessed after the failed receive, then after the subsequent successful
receive it will return EIO. This happens because nothing mounts the new
dataset, leaving the old, no longer valid dataset still mounted.
This commit fixes zfs receive to always unmount and remount the
destination, regardless of whether the stream is a new stream or a
resumed stream.
Sponsored by: Axcient Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Ryan Moeller <ryan@iXsystems.com> Signed-off-by: Alan Somers <asomers@gmail.com>
External-issue: https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=249579
Closes #10995
Closes #10999
John Poduska [Sat, 3 Oct 2020 00:40:46 +0000 (20:40 -0400)]
Mismatched nvlist names in zfs_keys_send_space
This causes "zfs send -vt ..." to fail with:
cannot resume send: Unknown error 1030
It turns out that some of the name/value pairs in the verification
list for zfs_ioc_send_space(), zfs_keys_send_space, had the wrong
name, so the ioctl got kicked out in zfs_check_input_nvpairs().
Update the names accordingly.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Ryan Moeller <ryan@iXsystems.com> Signed-off-by: John Poduska <jpoduska@datto.com>
Closes #10978
Brian Behlendorf [Wed, 30 Sep 2020 20:27:51 +0000 (13:27 -0700)]
Fix buggy procfs_list_seq_next warning
The kernel seq_read() helper function expects ->next() to update
the passed position even there are no more entries. Failure to
do so results in the following warning being logged.
seq_file: buggy .next function procfs_list_seq_next [spl]
did not update position index
Functionally there is no issue with the way procfs_list_seq_next()
is implemented and the warning is harmless. However, we want to
silence this some what scary incorrect warning. This commit
updates the Linux procfs code to advance the position even for
the last entry.
Reviewed-by: Tony Hutter <hutter2@llnl.gov> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes #10984
Closes #10996
Ryan Moeller [Wed, 30 Sep 2020 20:25:50 +0000 (16:25 -0400)]
FreeBSD: Fix legacy compat for platform IOCs
The request number is out of bounds of the platform table.
Subtract the starting offset to get the correct subscript.
Reviewed-by: Alexander Motin <mav@FreeBSD.org> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Ryan Moeller <ryan@iXsystems.com>
Closes #10994
Matthew Macy [Wed, 30 Sep 2020 20:24:38 +0000 (13:24 -0700)]
Eliminate gratuitous bzeroing in dbuf_stats_hash_table_data
`dbuf_stats_hash_table_data` can take much longer than it needs to
by repeatedly bzeroing its buffer when in fact the buffer only needs
to be NULL terminated.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Ryan Moeller <ryan@iXsystems.com> Signed-off-by: Matt Macy <mmacy@FreeBSD.org>
Closes #10993
do a cyclic seek for unused memory objects in pool
In non regular use cases allocated memory might stay persistent in memory
pool. This small patch checks every minute if there are old objects which
can be released from memory pool.
Right now with regular use, the pool is checked for old objects on each
allocation attempt from this pool. so basically polling by its use. Now
consider what happens if someone writes a lot of files and stops use of
the volume or even unmounts it. So the code will no longer check if
objects can be released from the pool. Already allocated objects will
still stay in pool cache. this is no big issue for common use. But
someone discovered this issue while doing tests. personally i know this
behavior and I'm aware of it. Its no big issue. just a enhancement
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Kjeld Schouten-Lebbing <kjeld@schouten-lebbing.nl> Signed-off-by: Sebastian Gottschall <s.gottschall@dd-wrt.com>
Closes #10938
Closes #10969
Ryan Moeller [Wed, 30 Sep 2020 20:19:49 +0000 (16:19 -0400)]
Drop references when skipping dmu_send due to EXDEV
When an invalid incremental send is requested where the "to" ds is
before the "from" ds, make sure to drop the reference to the pool
and the dataset before returning the error.
Add an assert on FreeBSD to make sure we don't hold any locks after
returning from an ioctl.
Add some test coverage.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Ryan Moeller <ryan@iXsystems.com>
Closes #10919
Brian Behlendorf [Tue, 29 Sep 2020 01:43:27 +0000 (18:43 -0700)]
Use known license string for zzstd
The Linux kernel MODULE_LICENSE macro only recognizes a handful of
license strings and "BSD" is not one of the them. Update the macro
to use "Dual BSD/GPL" which is recognized and what the kernel expects
BSD licensed module to use.
Brian Behlendorf [Mon, 28 Sep 2020 23:42:54 +0000 (16:42 -0700)]
Fix CONFIG_DEBUG_LOCK_ALLOC configure check
This check was accidentally broken when the kABI checks were updated
to run in parallel, commit 608f874. The check must be for the
config_debug_lock_alloc_license name to determine if the symbol
is license compatible.
Reviewed-by: Tony Hutter <hutter2@llnl.gov> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes #10991
Brian Behlendorf [Mon, 28 Sep 2020 23:40:50 +0000 (16:40 -0700)]
Fix objtool configure check
The m4 objtool configure check can incorrectly fail because of a
missing header in the test. This appears to be the result of a
recent kernel change and was observed on the Fedora 5.8.11-200
kernel.
In file included from /home/fedora/zfs/build/objtool/objtool.c:75:
./arch/x86/include/asm/frame.h:100:57: error: 'struct pt_regs'
declared inside parameter list will not be visible outside
of this definition or declaration [-Werror]
The consequence of this is that the "stack_frame_non_standard"
check is never run and HAVE_STACK_FRAME_NON_STANDARD is set
incorrectly which results in a build failure. This change adds
the appropriate header to the "objtool" check so it now behaves
as intended.
Reviewed-by: Kjeld Schouten <kjeld@schouten-lebbing.nl> Reviewed-by: Tony Hutter <hutter2@llnl.gov> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes #10990
The error returned by `zpool remove` when the encryption keys aren't
loaded isn't very helpful. Furthermore, the man pages make no
mention that the keys need to be loaded. This change doesn't resolve
the error message but it does update the man page to mention this
requirement.
Authored-by: grodik <pat@litke.dev> Reviewed-by: Matthew Ahrens <mahrens@delphix.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes #10939
Closes #10948
Allan Jude [Fri, 25 Sep 2020 21:37:10 +0000 (17:37 -0400)]
zfs userspace: use zfs_path_to_zhandle so argument can be a path
Change zfs userspace subcommand to use zfs_path_to_zhandle() so that
the provided dataset can be a path (/usr) or a dataset (rpool/usr).
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Ryan Moeller <ryan@iXsystems.com> Signed-off-by: Allan Jude <allan@klarasystems.com>
Closes #8915
Adam D. Moss [Fri, 25 Sep 2020 20:49:22 +0000 (13:49 -0700)]
Add DB_RF_NOPREFETCH to dbuf_read()s in dnode.c
Prefetching of dnodes in dbuf_read() can cause significant mutex
contention for some workloads and isn't very helpful. This is
because we already get 32 dnodes for each block read, and when
iterating over a directory we prefetch the dnodes in the directory.
Disable this prefetching to prevent the lock contention.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Submitted-by: Adam Moss <c@yotes.com> Submitted-by: Matthew Ahrens <mahrens@delphix.com> Signed-off-by: Adam Moss <c@yotes.com>
Closes #10877
Closes #10953
Brian Behlendorf [Fri, 25 Sep 2020 20:28:35 +0000 (13:28 -0700)]
Fix PREEMPTION=y and BLK_CGROUP=y config on arm64
With PREEMPTION=y and BLK_CGROUP=y preempt_schedule_notrace() is being
used on arm64 which is a GPL-only function and hence the build of the
DKMS kernel module fails.
Fix that by redefining preempt_schedule_notrace() to preempt_schedule()
which should be safe as long as tracing is not used.
Ryan Moeller [Thu, 3 Sep 2020 21:15:32 +0000 (21:15 +0000)]
FreeBSD: Code cleanup in zio_crypt
Address some unused value and control flow issues flagged by Coverity.
Unreachable code is pruned and unused values are avoided.
Some scattered sections are reordered for coherence.
We can assume kmem_alloc(n, KM_SLEEP) doesn't fail, so there is no need
to check if it returned NULL. The allocated memory doesn't need to be
zeroed, other than the last iovec (the MAC).
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Ryan Moeller <ryan@iXsystems.com>
Closes #10884
George Wilson [Fri, 25 Sep 2020 20:09:40 +0000 (15:09 -0500)]
zpool command complains about /etc/exports.d
If the /etc/exports.d directory does not exist, then we should only
create it when we're performing an action which already requires root
privileges.
This commit moves the directory creation to the enable/disable code
path which ensures that we have the appropriate privileges.
Reviewed-by: Richard Elling <Richard.Elling@RichardElling.com> Reviewed-by: Matthew Ahrens <mahrens@delphix.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: George Wilson <gwilson@delphix.com>
Closes #10785
Closes #10934
zfs_log_write: simplify data copying code for WR_COPIED records
lr_write_t records that are WR_COPIED have the record data directly
appended to them (see lr_write_t type definition).
The data is copied from the debuf using dmu_read_by_dnode.
This function was called, only for WR_COPIED records, as part of a
short-circuiting if-statement's if-expression.
I found this side-effectful call to dmu_read_by_dnode pretty
hard to spot.
This patch improves readability by moving the call to its own line.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: George Wilson <gwilson@delphix.com> Signed-off-by: Christian Schwarz <me@cschwarz.com>
Closes #10956
Matthew Macy [Wed, 23 Sep 2020 23:43:51 +0000 (16:43 -0700)]
FreeBSD: Add support for procfs_list
The procfs_list interface is required by several kstats. Implement
this functionality for FreeBSD to provide access to these kstats.
Reviewed-by: Allan Jude <allan@klarasystems.com> Reviewed-by: Ryan Moeller <ryan@ixsystems.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Matt Macy <mmacy@FreeBSD.org>
Closes #10890
Matthew Macy [Wed, 23 Sep 2020 18:09:48 +0000 (11:09 -0700)]
FreeBSD: Don't save user FPU context in kernel threads
Reviewed-by: Alexander Motin <mav@FreeBSD.org> Reviewed-by: Ryan Moeller <freqlabs@FreeBSD.org> Signed-off-by: Matt Macy <mmacy@FreeBSD.org>
Closes #10899
Mark Johnston [Tue, 22 Sep 2020 23:05:52 +0000 (19:05 -0400)]
Fix a logic bug in the FreeBSD getpages VOP
In commit cd32b4f5b79c ("Fix a deadlock in the FreeBSD getpages VOP") I
introduced a bug while porting the patch originally committed to
FreeBSD: the rangelock pointer may be NULL if the try operation failed,
so we must avoid calling zfs_rangelock_unlock() in that case.
Reviewed-by: Allan Jude <allan@klarasystems.com> Reviewed-by: Ryan Moeller <ryan@ixsystems.com> Reviewed-by: Matt Macy <mmacy@FreeBSD.org> Reported-by: Steve Wills <swills@FreeBSD.org> Signed-off-by: Mark Johnston <markj@FreeBSD.org>
Closes #10519
Closes #10960
Mark Johnston [Fri, 18 Sep 2020 12:45:54 +0000 (08:45 -0400)]
Annontate FreeBSD sysctls with CTLFLAG_MPSAFE
Without this, the sysctl system calls will acquire a global lock before
invoking the handler. This is noticeable in some situations when
running top(1). The global lock is mostly vestigal but continues to see
some use and so contention is still a problem; until the default sense
of the MPSAFE flag changes, we have to annotate each and every handler.
Reviewed-by: Allan Jude <allan@klarasystems.com> Reviewed-by: Ryan Moeller <ryan@ixsystems.com> Signed-off-by: Mark Johnston <markj@FreeBSD.org>
Closes #10836
Mark Johnston [Fri, 18 Sep 2020 12:41:28 +0000 (08:41 -0400)]
Fix switch statement indentation in the FreeBSD kstat code
This is in preparation for some functional changes.
Reviewed-by: Allan Jude <allan@klarasystems.com> Reviewed-by: Ryan Moeller <ryan@ixsystems.com> Signed-off-by: Mark Johnston <markj@FreeBSD.org>
Closes #10950
George Wilson [Fri, 18 Sep 2020 19:13:47 +0000 (14:13 -0500)]
vdev_ashift should only be set once
== Motivation and Context
The new vdev ashift optimization prevents the removal of devices when
a zfs configuration is comprised of disks which have different logical
and physical block sizes. This is caused because we set 'spa_min_ashift'
in vdev_open and then later call 'vdev_ashift_optimize'. This would
result in an inconsistency between spa's ashift calculations and that
of the top-level vdev.
In addition, the optimization logical ignores the overridden ashift
value that would be provided by '-o ashift=<val>'.
== Description
This change reworks the vdev ashift optimization so that it's only
set the first time the device is configured. It still allows the
physical and logical ahsift values to be set every time the device
is opened but those values are only consulted on first open.
Reviewed-by: Matthew Ahrens <mahrens@delphix.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Cedric Berger <cedric@precidata.com> Signed-off-by: George Wilson <gwilson@delphix.com>
External-Issue: DLPX-71831
Closes #10932
George Wilson [Fri, 18 Sep 2020 03:03:10 +0000 (22:03 -0500)]
pool may become suspended during device expansion
When expanding a device zfs needs to rescan the partition table to
get the correct size. This can only happen when we're in the kernel
and requires the device to be closed. As part of the rescan, udev is
notified and the device links are removed and recreated. This leave a
window where the vdev code may try to reopen the device before udev
has recreated the link. If that happens, then the pool may end up in
a suspended state.
To correct this, we leverage the BLKPG_RESIZE_PARTITION ioctl which
allows the partition information to be modified even while it's in use.
This ioctl also does not remove the device link associated with the zfs
data partition so it eliminates the race condition that can occur in
the kernel.
Reviewed-by: Pavel Zakharov <pavel.zakharov@delphix.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: George Wilson <gwilson@delphix.com>
Closes #10897
Matthew Ahrens [Thu, 17 Sep 2020 17:55:30 +0000 (10:55 -0700)]
zdb leak detection fails with in-progress device removal
When a device removal is in progress, there are 2 locations for the data
that's already been moved: the original location, on the device that's
being removed; and the new location, which is pointed to by the indirect
mapping. When doing leak detection, zdb needs to know about both
locations. To determine what's already been copied, we load the
spacemaps of the removing vdev, omit the blocks that are yet to be
copied, and then use the vdev's remap op to find the new location.
The problem is with an optimization to the spacemap-loading code in zdb.
When processing the log spacemaps, we ignore entries that are not
relevant because they are past the point that's been copied. However,
entries which span the point that's been copied (i.e. they are partly
relevant and partly irrelevant) are processed normally. This can lead
to an illegal spacemap operation, for example if offsets up to 100KB
have been copied, and the spacemap log has the following entries:
Because the entirely irrelevant entry was ignored, its space remains in
the spacemap. When the last entry is processed, we attempt to add it to
the spacemap, but it partially overlaps with the 100-150KB entry that
was left over.
This problem was discovered by ztest/zloop.
One solution would be to also ignore the irrelevant parts of
partially-irrelevant entries (i.e. when processing the ALLOC 50-150, to
only add 50-100 to the spacemap). However, this commit implements a
simpler solution, which is to remove this optimization entirely. I.e.
to process the entire spacemap log, without regard for the point that's
been copied. After reconstructing the entire allocatable range tree,
there's already code to remove the parts that have not yet been copied.
Reviewed-by: Serapheim Dimitropoulos <serapheim@delphix.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Matthew Ahrens <mahrens@delphix.com>
External-issue: DLPX-71820
Closes #10920
John Poduska [Thu, 17 Sep 2020 17:53:02 +0000 (13:53 -0400)]
Need a long hold in zpl_mount_impl
In zpl_mount_impl, there is:
dmu_objset_hold ; returns with pool & ds held
dsl_pool_rele
sget
dsl_dataset_rele
As spelled out in the "DSL Pool Configuration Lock" in dsl_pool.c,
this requires a long hold.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Paul Zuchowski <pzuchowski@datto.com> Signed-off-by: John Poduska <jpoduska@datto.com>
Closes #10936
cmd/zgenhostid: replace with simple c implementation
It was discovered that dracut scripts and zgenhostid
always generate little-endian /etc/hostid.
This commit provides simple endianess-aware binary
and updates the scripts to use it.
New features include:
-f flag to force overwrite.
-o flag to write to different file (for dracut)
accepting both 0x01234567 and 01234567 values as input
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Reviewed-by: Olaf Faaland <faaland1@llnl.gov> Signed-off-by: Georgy Yakovlev <gyakovlev@gentoo.org>
Closes #10887
Closes #10925