Add missing getopt specifier for `zdb -V` verbatim option and
set flag with correct bitwise operator.
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
When casesensitivity=insensitive is set for the
file system, we can deadlock in a rename if the user uses different case
for each path. For example rename("A/some-file.txt", "a/some-file.txt").
The simple test for this is:
1. mkdir some-dir in a ZFS file system
2. touch some-dir/some-file.txt
3. mv Some-dir/some-file.txt some-dir/some-other-file.txt
This last request deadlocks trying to relock the i_mutex on the inode for
the parent directory.
The solution is to use d_add_ci in zpl_lookup if we are on a file system
that has the casesensitivity=insensitive attribute set.
This patch checks if we are working on a case insensitive file system and if
so, allocates storage for the case insensitive name and passes it to
zfs_lookup and then calls d_add_ci instead of d_splice_alias.
The performance impact seems to be minimal even though we have introduced a
kmalloc and kfree in the lookup path.
The problem was found when running Microsoft's FSCT against Samba on top of
ZFS On Linux.
Signed-off-by: Richard Sharpe <realrichardsharpe@gmail.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#4136
To make arc_summary.py and dbufstat.py compatible with python3
some minor fixes were required, this was done automatically by
`2to3 -w arc_summary.py` and `2to3 -w dbufstat.py`.
Signed-off-by: Hajo Möller <dasjoe@gmail.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Reviewed-by: Richard Laager <rlaager@wiktel.com>
3139 zdb dies when it tries to determine path of unlinked file
Reviewed by: Matt Ahrens <matthew.ahrens@delphix.com>
Reviewed by: Christopher Siden <chris.siden@delphix.com>
Reviewed by: Eric Schrock <eric.schrock@delphix.com>
Approved by: Dan McDonald <danmcd@nexenta.com>
References:
https://github.com/illumos/illumos-gate/commit/1ce39b5https://www.illumos.org/issues/3139
Ported-by: kernelOfTruth kerneloftruth@gmail.com
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
2077 lots of unreachable breaks in illumos gate
Reviewed by: Dan McDonald <danmcd@nexenta.com>
Reviewed by: Garrett D'Amore <garrett@damore.org>
Approved by: Richard Lowe <richlowe@richlowe.net>
References:
https://www.illumos.org/issues/2077https://github.com/illumos/illumos-gate/commit/33f5ff1
Porting notes:
- Only one file of the original patch applied to ZFS
- Minor formating change to align copyright block with upstream
Ported-by: Brian Behlendorf <behlendorf1@llnl.gov>
The function sa_update() accepts a 32-bit length parameter and
assigns it to a 16-bit field in sa_bulk_attr_t, potentially
truncating the passed-in value. This could lead to corrupt system
attribute (SA) records getting written to the pool. Add a VERIFY to
sa_update() to detect cases where overflow would occur. The SA length
is limited to 16-bit values by the on-disk format defined by
sa_hdr_phys_t.
The function zfs_sa_set_xattr() is vulnerable to this bug if the
unpacked nvlist of xattrs is less than 64k in size but the packed
size is greater than 64k. Fix this by appropriately checking the
size of the packed nvlist before calling sa_update(). Add error
handling to zpl_xattr_set_sa() to keep the cached list of SA-based
xattrs consistent with the data on disk.
Lastly, zfs_sa_set_xattr() calls dmu_tx_abort() on an assigned
transaction if sa_update() returns an error, but the DMU only allows
unassigned transactions to be aborted. Wrap the sa_update() call in a
VERIFY0, remove the transaction abort, and call dmu_tx_commit()
unconditionally. This is consistent practice with other callers
of sa_update().
Signed-off-by: Ned Bass <bass6@llnl.gov>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Richard Yao <ryao@gentoo.org>
Closes#4150
5745 zfs set allows only one dataset property to be set at a time
Reviewed by: Christopher Siden <christopher.siden@delphix.com>
Reviewed by: George Wilson <george@delphix.com>
Reviewed by: Matthew Ahrens <mahrens@delphix.com>
Reviewed by: Bayard Bell <buffer.g.overflow@gmail.com>
Reviewed by: Richard PALO <richard@NetBSD.org>
Reviewed by: Steven Hartland <killing@multiplay.co.uk>
Approved by: Rich Lowe <richlowe@richlowe.net>
References:
https://www.illumos.org/issues/5745https://github.com/illumos/illumos-gate/commit/3092556
Porting notes:
- Fix the missing braces around initializer, zfs_cmd_t zc = {"\0"};
- Remove extra format argument in zfs_do_set()
- Declare at the top:
- zfs_prop_t prop;
- nvpair_t *elem;
- nvpair_t *next;
- int i;
- Additionally initialize:
- int added_resv = 0;
- zfs_prop_t prop = 0;
- Assign 0 install of NULL for uint64_t types.
- zc->zc_nvlist_conf = '\0';
- zc->zc_nvlist_src = '\0';
- zc->zc_nvlist_dst = '\0';
Ported-by: kernelOfTruth kerneloftruth@gmail.com
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#3574
We need truncate and remove be in the same tx when doing zfs_rmnode on xattr
dir. Otherwise, if we truncate and crash, we'll end up with inconsistent zap
object on the delete queue. We do this by skipping dmu_free_long_range and let
zfs_znode_delete to do the work.
Signed-off-by: Chunwei Chen <david.chen@osnexus.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Issue #4114
Issue #4052
Issue #4006
Issue #3018
Issue #2861
During zfs_rmnode on a xattr dir, if the system crash just after
dmu_free_long_range, we would get empty xattr dir in delete queue. This would
cause blkid=0 be passed into zap_get_leaf_byblk when doing zfs_purgedir during
mount, and would try to do rw_enter on a wrong structure and cause system
lockup.
We fix this by returning ENOENT when blkid is zero in zap_get_leaf_byblk.
Signed-off-by: Chunwei Chen <david.chen@osnexus.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#4114Closes#4052Closes#4006Closes#3018Closes#2861
There exists a lock inversion between the z_xattr_lock and the
z_teardown_lock. Resolve this by taking the z_teardown_lock in
all registered xattr callbacks prior to taking the z_xattr_lock.
This ensures the locks are always taken is the same order thus
preventing a deadlock. Note the z_teardown_lock is taken again
in zfs_lookup() and this is safe because the z_teardown lock is
a re-entrant read reader/writer lock.
* process-1
zpl_xattr_get -> Takes zp->z_xattr_lock
__zpl_xattr_get
zfs_lookup -> Takes zsb->z_teardown_lock in ZFS_ENTER macro
* process-2
zfs_ioc_recv -> Takes zsb->z_teardown_lock in zfs_suspend_fs()
zfs_resume_fs
zfs_rezget -> Takes zp->z_xattr_lock
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Chunwei Chen <tuxoko@gmail.com>
Closes#3943Closes#3969Closes#4121
Commit efc412b updated spa_config_write() for Linux 4.2 kernels to
truncate and overwrite rather than rename the cache file. This is
the correct fix but it should have only been applied for the kernel
build. In user space rename(2) is needed because ztest depends on
the cache file.
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#4129
When running a kernel with CONFIG_LOCKDEP=y, lockdep reports possible
recursive locking in some cases and possible circular locking dependency
in others, within the SPL and ZFS modules.
This patch uses a mutex type defined in SPL, MUTEX_NOLOCKDEP, to mark
such mutexes when they are initialized. This mutex type causes
attempts to take or release those locks to be wrapped in lockdep_off()
and lockdep_on() calls to silence the dependency checker and allow the
use of lock_stats to examine contention.
For RW locks, it uses an analogous lock type, RW_NOLOCKDEP.
The goal is that these locks are ultimately changed back to type
MUTEX_DEFAULT or RW_DEFAULT, after the locks are annotated to reflect
their relationship (e.g. z_name_lock below) or any real problem with the
lock dependencies are fixed.
Some of the affected locks are:
tc_open_lock:
=============
This is an array of locks, all with same name, which txg_quiesce must
take all of in order to move txg to next state. All default to the same
lockdep class, and so to lockdep appears recursive.
zp->z_name_lock:
================
In zfs_rmdir,
dzp = znode for the directory (input to zfs_dirent_lock)
zp = znode for the entry being removed (output of zfs_dirent_lock)
zfs_rmdir()->zfs_dirent_lock() takes z_name_lock in dzp
zfs_rmdir() takes z_name_lock in zp
Since both dzp and zp are type znode_t, the locks have the same default
class, and lockdep considers it a possible recursive lock attempt.
l->l_rwlock:
============
zap_expand_leaf() sometimes creates two new zap leaf structures, via
these call paths:
zap_deref_leaf()->zap_get_leaf_byblk()->zap_leaf_open()
zap_expand_leaf()->zap_create_leaf()->zap_expand_leaf()->zap_create_leaf()
Because both zap_leaf_open() and zap_create_leaf() initialize
l->l_rwlock in their (separate) leaf structures, the lockdep class is
the same, and the linux kernel believes these might both be the same
lock, and emits a possible recursive lock warning.
Signed-off-by: Olaf Faaland <faaland1@llnl.gov>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#3895
Both lock types were introduced in SPL to allow some locks to be
taken/released with linux lockdep turned off. See SPL commit for
details.
Add the new lock types to zfs_context.h to allow user space compilation.
Depends on SPL commit 692ae8d
SPL pull request refs/pull/480/head
Signed-off-by: Olaf Faaland <faaland1@llnl.gov>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#3895
Adds zio_taskq_batch_pct as an exported module parameter,
allowing users to modify it at module load time.
Signed-off-by: DHE <git@dehacked.net>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#4110
Original-patch-by: @jgoerzen
Signed-off-by: Benjamin Albrecht <git@albrecht.io>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes zfsonlinux/pkg-zfs#102
Closes#4029
Correct some misspelled words and grammatical errors, and remove
trailing white space in the man pages.
Signed-off-by: Ned Bass <bass6@llnl.gov>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#4115
Update the bounds checking for zfs_vdev_aggregation_limit so that
it has a floor of zero and a maximum value of the supported block
size for the pool.
Additionally add an early return when zfs_vdev_aggregation_limit
equals zero to disable aggregation. For very fast solid state or
memory devices it may be more expensive to perform the aggregation
than to issue the IO immediately.
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
This deadlock may manifest itself in slightly different ways but
at the core it is caused by a memory allocation blocking on file-
system reclaim in the zio pipeline. This is normally impossible
because zio_execute() disables filesystem reclaim by setting
PF_FSTRANS on the thread. However, kmem cache allocations may
still indirectly block on file system reclaim while holding the
critical vq->vq_lock as shown below.
To resolve this issue zio_buf_alloc_flags() is introduced which
allocation flags to be passed. This can then be used in
vdev_queue_aggregate() with KM_NOSLEEP when allocating the
aggregate IO buffer. Since aggregating the IO is purely a
performance optimization we want this to either succeed or fail
quickly. Trying too hard to allocate this memory under the
vq->vq_lock can negatively impact performance and result in
this deadlock.
* z_wr_iss
zio_vdev_io_start
vdev_queue_io -> Takes vq->vq_lock
vdev_queue_io_to_issue
vdev_queue_aggregate
zio_buf_alloc -> Waiting on spl_kmem_cache process
* z_wr_int
zio_vdev_io_done
vdev_queue_io_done
mutex_lock -> Waiting on vq->vq_lock held by z_wr_iss
* txg_sync
spa_sync
dsl_pool_sync
zio_wait -> Waiting on zio being handled by z_wr_int
* spl_kmem_cache
spl_cache_grow_work
kv_alloc
spl_vmalloc
...
evict
zpl_evict_inode
zfs_inactive
dmu_tx_wait
txg_wait_open -> Waiting on txg_sync
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Chunwei Chen <david.chen@osnexus.com>
Signed-off-by: Tim Chase <tim@chase2k.com>
Closes#3808Closes#3867
There exists a lock inversion between the z_xattr_lock and the
z_teardown_lock. Detect this case and return EBUSY so zfs_resume_fs()
will mark the inode stale and it can be safely revalidated on next
access.
* process-1
zpl_xattr_get -> Takes zp->z_xattr_lock
__zpl_xattr_get
zfs_lookup -> Takes zsb->z_teardown_lock in ZFS_ENTER macro
* process-2
zfs_ioc_recv -> Takes zsb->z_teardown_lock in zfs_suspend_fs()
zfs_resume_fs
zfs_rezget -> Takes zp->z_xattr_lock
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Chunwei Chen <david.chen@osnexus.com>
Closes#3969
Since uio now supports bvec, we can convert bio into uio and reuse
dmu_{read,write}_uio. This way, we can remove some duplicate code.
Signed-off-by: Chunwei Chen <david.chen@osnexus.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#4078
Userspace can freely pass in whatever iovec it feels like, and it's perfectly
legal to pass an iovec which contains a zero length segment. In the current
implementation, uio_prefaultpages would touch an out of bound byte in the
"last byte" logic. While this probably wouldn't cause any critical error, we
would like uio_prefaultpages to be able to continue gracefully.
Signed-off-by: Chunwei Chen <david.chen@osnexus.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#4078
If a bit were cleared in `bp->blk_birth` such that the txg birth
was now lower than any other txg_birth in the deadlist, then there
will be no entry before this in the tree.
This should be impossible but regardless error handling code has
been added for this case. By default this is left as a fatal case
and the blk_birth is logged. However, setting `zfs_recover=1` will
cause the bp to be placed at the start of the deadlist even though
it contains an invalid blk_birth.
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Chunwei Chen <tuxoko@gmail.com>
Closes#4086Closes#4089
Commit 5f6d0b6 was originally added to gracefully handle block
pointers with a damaged logical size. However, it incorrectly
assumed that all passed arc_done_func_t could handle a NULL
arc_buf_t.
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#4069Closes#4080
Commit ca0bf58d to address arcs_mtx contention removed column "index"
from the output of kstats/dbuf.
dbufstat.py was not updated to reflect this, which causes it to crash
when run with -bx
This removes "index" from hardcoded lists of columns.
Signed-off-by: Olaf Faaland <faaland1@llnl.gov>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#4096
This reverts commit 2026196230.
It is no longer necessary now that we pass -DDEBUG unconditionally.
Signed-off-by: Richard Yao <ryao@gentoo.org>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Issue #4095
Illumos unconditionally builds zdb and ztest with -DDEBUG. This helps
catch bugs and eliminates the need for commits like
2026196230, which changed ASSERTs to
VERIFYs. The following files in the illumos tree show this:
usr/src/cmd/zdb/Makefile.com
usr/src/cmd/ztest/Makefile.com
Given the usefulness of having early failure in these tools, we should
do it too.
Signed-off-by: Richard Yao <ryao@gentoo.org>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Issue #4095
While exceptionally unlikely to cause a problem the zfs_snapentry_t
hold should be taken before the dispatch to prevent any possibility
of the task being processed before the hold.
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Chunwei Chen <david.chen@osnexus.com>
When a concorrent mount finishes just before calling to
zfsctl_snapshot_ismounted, if we return EISDIR, the VFS will return
with EREMOTE. We should instead just return 0, so VFS may retry and
would likely notice the dentry is alreadly mounted. This will be
inline with when usermode helper return EBUSY.
Signed-off-by: Chunwei Chen <david.chen@osnexus.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
By changing the zfs_snapshot_lock from a mutex to a rw lock the
zfsctl_lookup_objset() function can be allowed to run concurrently.
This should reduce the latency of fh_to_dentry lookups in ZFS
snapshots which are being accessed over NFS.
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Chunwei Chen <david.chen@osnexus.com>
The zfsctl_snapshot_unmount_delay() function must not be called
from zfsctl_lookup_objset() while it is currently holding the
zfs_snapshot_lock. This will result in a deadlock. It is safe
to call zfsctl_snapshot_unmount_delay_impl() directly because the
function already has a reference on the zfs_snapentry_t.
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Chunwei Chen <david.chen@osnexus.com>
Closes#3997
There are cases where it's desirable that auto-mounted snapshots
not expire after a fixed duration. They should be unmounted only
when the filesystem they are a snapshot of is unmounted.
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Chunwei Chen <david.chen@osnexus.com>
For some arm, powerpc, and sparc platforms it was possible that
neither _ILP32 of _LP64 would be defined. Update the isa_defs.h
header to explicitly set these macros and generate a compile error
in the case neither are defined.
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Chunwei Chen <tuxoko@gmail.com>
Closes#4048
objsetid is not unique across pool, so using it solely as key would cause
panic when automounting two snapshot on different pools with the same
objsetid. We fix this by adding spa pointer as additional key.
Signed-off-by: Chunwei Chen <david.chen@osnexus.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Richard Yao <ryao@gentoo.org>
Issue #3948
Issue #3786
Issue #3887
While stack size will vary by architecture it has historically defaulted to
8K on x86_64 systems. However, as of Linux 3.15 the default thread stack
size was increased to 16K. These kernels are now the default in most non-
enterprise distributions which means we no longer need to assume 8K stacks.
This patch takes advantage of that fact by appropriately reverting stack
conservation changes which were made to ensure stability. Changes which
may have had a negative impact on performance for certain workloads. This
also has the side effect of bringing the code slightly more in line with
upstream.
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Richard Yao <ryao@gentoo.org>
Closes#4059
Running arcstat.py -x currently throws KeyError due to rmis being
absent, it was removed in commit ca0bf58.
Signed-off-by: cable2999 <cable2999@users.noreply.github.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#3931
Continuations should be indented four spaces.
Signed-off-by: ilovezfs <ilovezfs@icloud.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#4062
5959 clean up per-dataset feature count code
Reviewed by: Toomas Soome <tsoome@me.com>
Reviewed by: George Wilson <george@delphix.com>
Reviewed by: Alex Reece <alex@delphix.com>
Approved by: Richard Lowe <richlowe@richlowe.net>
References:
https://www.illumos.org/issues/5959https://github.com/illumos/illumos-gate/commit/ca0cc39
Porting notes:
illumos code doesn't check for feature_get_refcount() returning
ENOTSUP (which means feature is disabled) in zdb. zfsonlinux added
a check in https://github.com/zfsonlinux/zfs/commit/784652c
due to #3468. The check was reintroduced here.
Ported-by: Witaut Bajaryn <vitaut.bayaryn@gmail.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#3965
Provide a generic interface to prefetch ZAP entries by name. This
functionality is being added for external consumers such as Lustre.
It is based of the existing zap_prefetch_uint64() version which is
used by the deduplication code.
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Richard Yao <ryao@gentoo.org>
Closes#4061
Adding additional entries to the efi conversion array will help prevent
the overwriting of the GPTs of disks with in-use file systems in more
cases. Most notably, this adds partition type 8300 "Linux filesystem"
(0FC63DAF-8483-4772-8E79-3D69D8477DE4), which is often used for ext4 and
btrfs, among others.
This commit itself does nothing to address the underlying problematic
behavior that check_slice() isn't called on partitions of an
unrecognized type, even when they contain a currently mounted file
system.
The additional entries were derived from these two resources:
https://en.wikipedia.org/wiki/GUID_Partition_Tablehttp://sourceforge.net/p/gptfdisk/code/ci/master/tree/parttypes.cc
Signed-off-by: ilovezfs <ilovezfs@icloud.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Issue #4016
Currently, the SET_ERROR tracepoint triggers regardless of whether there
is an error or not. On Illumos, SET_ERROR only triggers on an actual
error, which is avoids irrelevant noise. Linux 2.6.38 added support for
conditional tracepoints, so we modify SET_ERROR to use them when they
are avaliable for functionality equivalent to the Illumos functionality.
Signed-off-by: Richard Yao <ryao@gentoo.org>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#4043
When dumping a block on a little endian system the data must be
byte swapped to display correctly. Example incorrect output:
$ echo 0123456789abcdef > aaa
$ zdb -eR pp 3:1ee00:200
3:1ee00:200
0 1 2 3 4 5 6 7 8 9 a b c d e f 0123456789abcdef
000000: 3736353433323130 6665646362613938 0123456789abcdef
000010: 000000000000000a 0000000000000000 ................
Signed-off-by: Chunwei Chen <david.chen@osnexus.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#4020
The current zdb calling behaviour is really fragile, and is guaranteed to
segfault if ztest is not installed in either /sbin or /usr/sbin. With this
patch, the ztest will try to call zdb in the following order.
1. Use environmental variable ZDB_PATH if provided.
2. If ztest resides in build tree, guess the in tree zdb path.
3. Just pass zdb to popen and let it search it in PATH.
Signed-off-by: Chunwei Chen <david.chen@osnexus.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#3126
This was originally in fe0ed8f910, but somehow
was changed and not working anymore. And it will cause the following error:
modprobe: ERROR: ../libkmod/libkmod.c:506 lookup_builtin_file() could not open builtin file '/lib/modules/4.2.0-18-generic/modules.builtin.bin'
Signed-off-by: Chunwei Chen <david.chen@osnexus.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#4027
Adding VPATH support, commit 47a4a6f, required that a `src`
and `obj` line be added to the top of the Makefiles. They
must be removed from the Makefiles when builtin.
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Issue zfsonlinux/spl#481
Issue zfsonlinux/spl#498