When doing a read from disk, ZFS creates 3 ZIO's: a zio_null(), the
logical zio_read(), and then a physical zio. Currently, each of these
results in a separate taskq_dispatch(zio_execute).
On high-read-iops workloads, this causes a significant performance
impact. By processing all 3 ZIO's in a single taskq entry, we reduce the
overhead on taskq locking and context switching. We accomplish this by
allowing zio_done() to return a "next zio to execute" to zio_execute().
This results in a ~12% performance increase for random reads, from
96,000 iops to 108,000 iops (with recordsize=8k, on SSD's).
Reviewed by: Pavel Zakharov <pavel.zakharov@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed by: George Wilson <george.wilson@delphix.com>
Signed-off-by: Matthew Ahrens <mahrens@delphix.com>
External-issue: DLPX-59292
Closes#7736
Currently, dnode_check_slots_free() works by checking dn->dn_type
in the dnode to determine if the dnode is reclaimable. However,
there is a small window of time between dnode_free_sync() in the
first call to dsl_dataset_sync() and when the useraccounting code
is run when the type is set DMU_OT_NONE, but the dnode is not yet
evictable, leading to crashes. This patch adds the ability for
dnodes to track which txg they were last dirtied in and adds a
check for this before performing the reclaim.
This patch also corrects several instances when dn_dirty_link was
treated as a list_node_t when it is technically a multilist_node_t.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Tom Caputi <tcaputi@datto.com>
Closes#7147Closes#7388
This reverts commit 2a16d4cfaf.
The commit was causing a "attempt to access beyond the end
of device" error:
list.zfsonlinux.org/pipermail/zfs-discuss/2018-September/032217.html
As a regular kernel function, kern_path() returns errors as negative
errnos, such as -ELOOP. zfsctl_snapdir_vget() must convert these into
the positive errnos used throughout the ZFS code when it returns them
to other ZFS functions so that the ZFS code properly sees them as
errors.
Reviewed-by: George Melikov <mail@gmelikov.ru>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Chris Siebenmann <cks.git01@cs.toronto.edu>
Closes#7764Closes#7864
fs/zfs/zfs/metaslab.c:1055:2: error: positional initialization of field
in ‘struct’ declared with ‘designated_init’ attribute
[-Werror=designated-init]
metaslab_rt_remove,
Signed-off-by: Georgy Yakovlev <ya@sysdump.net>
Reviewed-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Closes: #7069
Currently, there is a bug where older send streams without the
DMU_BACKUP_FEATURE_LARGE_DNODE flag are not handled correctly.
The code in receive_object() fails to handle cases where
drro->drr_dn_slots is set to 0, which is always the case when the
sending code does not support this feature flag. This patch fixes
the issue by ensuring that that a value of 0 is treated as
DNODE_MIN_SLOTS.
Tested-by: DHE <git@dehacked.net>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Tom Caputi <tcaputi@datto.com>
Closes#7617Closes#7662
Currently, when the receive_object() code wants to reclaim an
object, it always assumes that the dnode is the legacy 512 bytes,
even when the incoming bonus buffer exceeds this length. This
causes a buffer overflow if --enable-debug is not provided and
triggers an ASSERT if it is. This patch resolves this issue and
adds an ASSERT to ensure this can't happen again.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Tom Caputi <tcaputi@datto.com>
Closes#7097Closes#7433
This is a port of 047116ac - Raw sends must be able to decrease nlevels,
to the zfs-0.7-stable branch. It includes the various fixes to the
problem of receiving incremental streams which include reallocated dnodes
in which the number of dnode slots has changed but excludes the parts
which are related to raw streams.
From 047116ac:
Currently, when a raw zfs send file includes a
DRR_OBJECT record that would decrease the number of
levels of an existing object, the object is reallocated
with dmu_object_reclaim() which creates the new dnode
using the old object's nlevels. For non-raw sends this
doesn't really matter, but raw sends require that
nlevels on the receive side match that of the send
side so that the checksum-of-MAC tree can be properly
maintained. This patch corrects the issue by freeing
the object completely before allocating it again in
this case.
This patch also corrects several issues with
dnode_hold_impl() and related functions that prevented
dnodes (particularly multi-slot dnodes) from being
reallocated properly due to the fact that existing
dnodes were not being fully cleaned up when they
were freed.
This patch adds a test to make sure that zfs recv
functions properly with incremental streams containing
dnodes of different sizes.
This also includes a one-liner fix from loli10K to fix a test failure:
https://github.com/zfsonlinux/zfs/pull/7792#discussion_r212769264
Authored-by: Tom Caputi <tcaputi@datto.com>
Reviewed by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Jorgen Lundman <lundman@lundman.net>
Signed-off-by: Tom Caputi <tcaputi@datto.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Tim Chase <tim@chase2k.com>
Ported-by: Tim Chase <tim@chase2k.com>
Closes#6821Closes#6864
NOTE: This is the first of the port of 3 related patches patches to the
zfs-0.7-release branch of ZoL. The other two patches should immediately
follow this one.
This change modifies how 'checksum' and 'dedup' properties are verified
in zfs_check_settable() handling the case where they are explicitly
inherited in the dataset hierarchy when receiving a recursive send
stream.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Tom Caputi <tcaputi@datto.com>
Signed-off-by: loli10K <ezomori.nozomu@gmail.com>
Closes#7755Closes#7576Closes#7757
Porting notes:
* As of grub-2.02 these checksums are not supported. However, as
pointed out in #6501 there are alternatives such as EFISTUB which
work and have no such restriction. A warning was added to the
checksum property section of the zfs.8 man page.
Authored by: Toomas Soome <tsoome@me.com>
Reviewed by: C Fraire <cfraire@me.com>
Reviewed by: Robert Mustacchi <rm@joyent.com>
Reviewed by: Yuri Pankov <yuripv@yuripv.net>
Approved by: Dan McDonald <danmcd@joyent.com>
Ported-by: Brian Behlendorf <behlendorf1@llnl.gov>
OpenZFS-issue: https://illumos.org/issues/8906
OpenZFS-commit: https://github.com/openzfs/openzfs/commit/7dec52fCloses#6501Closes#7714
Commit 93b43af10 inadvertently introduced the following scenario which
can result in a deadlock. This issue was most easily reproduced by
LXD containers using a ZFS storage backend but should be reproducible
under any workload which is frequently mounting and unmounting.
-- THREAD A --
spa_sync()
spa_sync_upgrades()
rrw_enter(&dp->dp_config_rwlock, RW_WRITER, FTAG); <- Waiting on B
-- THREAD B --
mount_fs()
zpl_mount()
zpl_mount_impl()
dmu_objset_hold()
dmu_objset_hold_flags()
dsl_pool_hold()
dsl_pool_config_enter()
rrw_enter(&dp->dp_config_rwlock, RW_READER, tag);
sget()
sget_userns()
grab_super()
down_write(&s->s_umount); <- Waiting on C
-- THREAD C --
cleanup_mnt()
deactivate_super()
down_write(&s->s_umount);
deactivate_locked_super()
zpl_kill_sb()
kill_anon_super()
generic_shutdown_super()
sync_filesystem()
zpl_sync_fs()
zfs_sync()
zil_commit()
txg_wait_synced() <- Waiting on A
Reviewed by: Alek Pinchuk <apinchuk@datto.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#7598Closes#7659Closes#7691Closes#7693
Update the SA_COPY_DATA macro to check if architecture supports
efficient unaligned memory accesses at compile time. Otherwise
fallback to using the sa_copy_data() function.
The kernel provided CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS is
used to determine availability in kernel space. In user space
the x86_64, x86, powerpc, and sometimes arm architectures will
define the HAVE_EFFICIENT_UNALIGNED_ACCESS macro.
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#7642Closes#7684
The blk_queue_stackable() function was replaced in the 4.14 kernel
by queue_is_rq_based(), commit torvalds/linux@5fdee212. This change
resulted in the default elevator being used which can negatively
impact performance.
Rather than adding additional compatibility code to detect the
new interface unconditionally attempt to set the elevator. Since
we expect this to fail for block devices without an elevator the
error message has been moved in to zfs_dbgmsg().
Finally, it was observed that the elevator_change() was removed
from the 4.12 kernel, commit torvalds/linux@c033269. Update the
comment to clearly specify which are expected to export the
elevator_change() symbol.
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Reviewed-by: Tony Hutter <hutter2@llnl.gov>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#7645
1. Add a proc entry to display the pool's state:
$ cat /proc/spl/kstat/zfs/tank/state
ONLINE
This is done without using the spa config locks, so it will
never hang.
2. Fix 'zpool status' and 'zpool list -o health' output to print
"SUSPENDED" instead of "ONLINE" for suspended pools.
Reviewed-by: Olaf Faaland <faaland1@llnl.gov>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed by: Richard Elling <Richard.Elling@RichardElling.com>
Signed-off-by: Tony Hutter <hutter2@llnl.gov>
Closes#7331Closes#7563
Update bdev_capacity to have wholedisk vdevs query the
size of the underlying block device (correcting for the size
of the efi parition and partition alignment) and therefore detect
expanded space.
Correct vdev_get_stats_ex so that the expandsize is aligned
to metaslab size and new space is only reported if it is large
enough for a new metaslab.
Reviewed by: Don Brady <don.brady@delphix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed by: George Wilson <george.wilson@delphix.com>
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Reviewed by: John Wren Kennedy <jwk404@gmail.com>
Signed-off-by: sara hartse <sara.hartse@delphix.com>
External-issue: LX-165
Closes#7546
Issue #7582
Callbacks provided for module parameters are executed both
after the module is loaded, when a user alters it via sysfs, e.g
echo bar > /sys/modules/zfs/parameters/foo
as well as when the module is loaded with an argument, e.g.
modprobe zfs foo=bar
In the latter case, the init functions likely have not run yet,
including spa_init() which initializes the namespace lock so it is safe
to use.
Instead of immediately taking the namespace lock and attemping to
iterate over initialized spa structures, check whether spa_mode_global
is nonzero. This is set by spa_init() after it has initialized the
namespace lock.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Tim Chase <tim@chase2k.com>
Signed-off-by: Olaf Faaland <faaland1@llnl.gov>
Closes#7496Closes#7521
Add a helper function to trim the tailing new line. While we're
here use this new hook to immediately apply the new scheduler.
Reviewed-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#3356Closes#6573
Commit cc63068 caused ENOSPC error when copy a large amount of files
between two directories. The reason is that the patch limits zap leaf
expansion to 2 retries, and return ENOSPC when failed.
The intent for limiting retries is to prevent pointlessly growing table
to max size when adding a block full of entries with same name in
different case in mixed mode. However, it turns out we cannot use any
limit on the retry. When we copy files from one directory in readdir
order, we are copying in hash order, one leaf block at a time. Which
means that if the leaf block in source directory has expanded 6 times,
and you copy those entries in that block, by the time you need to expand
the leaf in destination directory, you need to expand it 6 times in one
go. So any limit on the retry will result in error where it shouldn't.
Note that while we do use different salt for different directories, it
seems that the salt/hash function doesn't provide enough randomization
to the hash distance to prevent this from happening.
Since cc63068 has already been reverted. This patch adds it back and
removes the retry limit.
Also, as it turn out, failing on zap_add() has a serious side effect for
mzap_upgrade(). When upgrading from micro zap to fat zap, it will
call zap_add() to transfer entries one at a time. If it hit any error
halfway through, the remaining entries will be lost, causing those files
to become orphan. This patch add a VERIFY to catch it.
Reviewed-by: Sanjeev Bagewadi <sanjeev.bagewadi@gmail.com>
Reviewed-by: Richard Yao <ryao@gentoo.org>
Reviewed-by: Tony Hutter <hutter2@llnl.gov>
Reviewed-by: Albert Lee <trisk@forkgnu.org>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed by: Matthew Ahrens <mahrens@delphix.com>
Signed-off-by: Chunwei Chen <david.chen@nutanix.com>
Closes#7401Closes#7421
vdev_count_leaves() in the denominator may return 0, caught by Coverity.
Introduced by
* 533ea04 Update mmp_delay on sync or skipped, failed write
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Reviewed-by: George Melikov <mail@gmelikov.ru>
Signed-off-by: Olaf Faaland <faaland1@llnl.gov>
Closes#7391
PROBLEM
=======
When `dmu_tx_assign` is called from `zil_lwb_write_issue`, it's possible
for either `ERESTART` or `EIO` to be returned.
If `ERESTART` is returned, this will cause an assertion to fail directly
in `zil_lwb_write_issue`, where the code assumes the return value is
`EIO` if `dmu_tx_assign` returns a non-zero value. This can occur if the
SPA is suspended when `dmu_tx_assign` is called, and most often occurs
when running `zloop`.
If `EIO` is returned, this can cause assertions to fail elsewhere in the
ZIL code. For example, `zil_commit_waiter_timeout` contains the
following logic:
lwb_t *nlwb = zil_lwb_write_issue(zilog, lwb);
ASSERT3S(lwb->lwb_state, !=, LWB_STATE_OPENED);
In this case, if `dmu_tx_assign` returned `EIO` from within
`zil_lwb_write_issue`, the `lwb` variable passed in will not be issued
to disk. Thus, it's `lwb_state` field will remain `LWB_STATE_OPENED` and
this assertion will fail. `zil_commit_waiter_timeout` assumes that after
it calls `zil_lwb_write_issue`, the `lwb` will be issued to disk, and
doesn't handle the case where this is not true; i.e. it doesn't handle
the case where `dmu_tx_assign` returns `EIO`.
SOLUTION
========
This change modifies the `dmu_tx_assign` function such that `txg_how` is
a bitmask, rather than of the `txg_how_t` enum type. Now, the previous
`TXG_WAITED` semantics can be used via `TXG_NOTHROTTLE`, along with
specifying either `TXG_NOWAIT` or `TXG_WAIT` semantics.
Previously, when `TXG_WAITED` was specified, `TXG_NOWAIT` semantics was
automatically invoked. This was not ideal when using `TXG_WAITED` within
`zil_lwb_write_issued`, leading the problem described above. Rather, we
want to achieve the semantics of `TXG_WAIT`, while also preventing the
`tx` from being penalized via the dirty delay throttling.
With this change, `zil_lwb_write_issued` can acheive the semtantics that
it requires by passing in the value `TXG_WAIT | TXG_NOTHROTTLE` to
`dmu_tx_assign`.
Further, consumers of `dmu_tx_assign` wishing to achieve the old
`TXG_WAITED` semantics can pass in the value `TXG_NOWAIT | TXG_NOTHROTTLE`.
Authored by: Prakash Surya <prakash.surya@delphix.com>
Approved by: Robert Mustacchi <rm@joyent.com>
Reviewed by: Matt Ahrens <mahrens@delphix.com>
Reviewed by: Andriy Gapon <avg@FreeBSD.org>
Ported-by: Brian Behlendorf <behlendorf1@llnl.gov>
Porting Notes:
- Additionally updated `zfs_tmpfile` to use `TXG_NOTHROTTLE`
OpenZFS-issue: https://www.illumos.org/issues/8997
OpenZFS-commit: https://github.com/openzfs/openzfs/commit/19ea6cb0f9Closes#7084
Added support for the bops->check_events() interface which was
added in the 2.6.38 kernel to replace bops->media_changed().
Fully implementing this functionality allows the volume resize
code to rely on revalidate_disk(), which is the preferred
mechanism, and removes the need to use check_disk_size_change().
In order for bops->check_events() to lookup the zvol_state_t
stored in the disk->private_data the zvol_state_lock needs to
be held. Since the check events interface may poll the mutex
has been converted to a rwlock for better concurrently. The
rwlock need only be taken as a writer in the zvol_free() path
when disk->private_data is set to NULL.
The configure checks for the block_device_operations structure
were consolidated in a single kernel-block-device-operations.m4
file.
The ZFS_AC_KERNEL_BDEV_BLOCK_DEVICE_OPERATIONS configure checks
and assoicated dead code was removed. This interface was added
to the 2.6.28 kernel which predates the oldest supported 2.6.32
kernel and will therefore always be available.
Updated maximum Linux version in META file. The 4.17 kernel
was released on 2018-06-03 and ZoL is compatible with the
finalized kernel.
Reviewed-by: Boris Protopopov <boris.protopopov@actifio.com>
Reviewed-by: Sara Hartse <sara.hartse@delphix.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#7611
Commit torvalds/linux@95582b0 changes the inode i_atime, i_mtime,
and i_ctime members form timespec's to timespec64's to make them
2038 safe. As part of this change the current_time() function was
also updated to return the timespec64 type.
Resolve this issue by introducing a new inode_timespec_t type which
is defined to match the timespec type used by the inode. It should
be used when working with inode timestamps to ensure matching types.
The timestruc_t type under Illumos was used in a similar fashion but
was specified to always be a timespec_t. Rather than incorrectly
define this type all timespec_t types have been replaced by the new
inode_timespec_t type.
Finally, the kernel and user space 'sys/time.h' headers were aligned
with each other. They define as appropriate for the context several
constants as macros and include static inline implementation of
gethrestime(), gethrestime_sec(), and gethrtime().
Reviewed-by: Chunwei Chen <tuxoko@gmail.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#7643
Backported-by: Richard Yao <ryao@gentoo.org>
Acquire zv_suspend_lock on first open and last close only.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Boris Protopopov <boris.protopopov@actifio.com>
Closes#6342
As of RHEL 7.5 the mainline fops.iterate() method was added to
the file_operations structure and is correctly detected by the
configure script.
Normally this is what we want, but in order to maintain KABI
compatibility the RHEL change additionally does the following:
* Requires that callers intending to use this extended interface
set the FMODE_KABI_ITERATE flag on the file structure when
opening the directory.
* Adds the fops.iterate() method to the end of the structure,
without removing fops.readdir().
This change updates the configure check to ignore the RHEL 7.5+
variant of fops.iterate() when detected. Instead fallback to
the fops.readdir() interface which will be available.
Finally, add the 'zpl_' prefix to the directory context wrappers
to avoid colliding with the kernel provided symbols when both
the fops.iterate() and fops.readdir() are provided by the kernel.
Reviewed-by: Olaf Faaland <faaland1@llnl.gov>
Reviewed-by: Tony Hutter <hutter2@llnl.gov>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#7460Closes#7463
Currently mounting an already mounted zfs dataset results in an
error, whereas it is typically allowed with other filesystems.
This causes some bad interactions with mount namespaces. Take
this sequence for example:
- Create a dataset
- Create a snapshot of the dataset
- Create a clone of the snapshot
- Create a new mount namespace
- Rename the original dataset
The rename results in unmounting and remounting the clone in the
original mount namespace, however the remount fails because the
dataset is still mounted in the new mount namespace. (Note that
this means the mount in the new mount namespace is never being
unmounted, so perhaps the unmount/remount of the clone isn't
actually necessary.)
The problem here is a result of the way mounting is implemented
in the kernel module. Since it is not mounting block devices it
uses mount_nodev() instead of the usual mount_bdev(). However,
mount_nodev() is written for filesystems for which each mount is
a new instance (i.e. a new super block), and zfs should be able
to detect when a mount request can be satisfied using an existing
super block.
Change zpl_mount() to call sget() directly with it's own test
callback. Passing the objset_t object as the fs data allows
checking if a superblock already exists for the dataset, and in
that case we just need to return a new reference for the sb's
root dentry.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Tom Caputi <tcaputi@datto.com>
Signed-off-by: Alek Pinchuk <apinchuk@datto.com>
Signed-off-by: Seth Forshee <seth.forshee@canonical.com>
Closes#5796Closes#7207
When setting `zfs_arc_max` its minimum value is allowed
to be 64 MiB. There was an off-by-1 error which can matter
on tiny systems.
Reviewed-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Chris Zubrzycki <github@mid-earth.net>
Closes#7417
When using 16MB blocks the send/recv queue's aren't quite big
enough. This change leaves the default 16M queue size which a
good value for most pools. But it additionally ensures that the
queue sizes are at least twice the allowed zfs_max_recordsize.
Reviewed-by: loli10K <ezomori.nozomu@gmail.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#7365Closes#7404
When an MMP write is skipped, or fails, and time since
mts->mmp_last_write is already greater than mts->mmp_delay, increase
mts->mmp_delay. The original code only updated mts->mmp_delay when a
write succeeded, but this results in the write(s) after delays and
failed write(s) reporting an ub_mmp_delay which is too low.
Update mmp_last_write and mmp_delay if a txg sync was successful. At
least one uberblock was written, thus extending the time we can be sure
the pool will not be imported by another host.
Do not allow mmp_delay to go below (MSEC2NSEC(zfs_multihost_interval) /
vdev_count_leaves()) so that a period of frequent successful MMP writes,
e.g. due to frequent txg syncs, does not result in an import activity
check so short it is not reliable based on mmp thread writes alone.
Remove unnecessary local variable, start. We do not use the start time
of the loop iteration.
Add a debug message in spa_activity_check() to allow verification of the
import_delay value and to prove the activity check occurred.
Alter the tests that import pools and attempt to detect an activity
check. Calculate the expected duration of spa_activity_check() based on
module parameters at the time the import is performed, rather than a
fixed time set in mmp.cfg. The fixed time may be wrong. Also, use the
default zfs_multihost_interval value so the activity check is longer and
easier to recognize.
Reviewed-by: Tony Hutter <hutter2@llnl.gov>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Signed-off-by: Olaf Faaland <faaland1@llnl.gov>
Closes#7330
During a receive operation zvol_create_minors_impl() can wait
needlessly for the prefetch thread because both share the same tasks
queue. This results in hung tasks:
<3>INFO: task z_zvol:5541 blocked for more than 120 seconds.
<3> Tainted: P O 3.16.0-4-amd64
<3>"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
The first z_zvol:5541 (zvol_task_cb) is waiting for the long running
traverse_prefetch_thread:260
root@linux:~# cat /proc/spl/taskq
taskq act nthr spwn maxt pri mina
spl_system_taskq/0 1 2 0 64 100 1
active: [260]traverse_prefetch_thread [zfs](0xffff88003347ae40)
wait: 5541
spl_delay_taskq/0 0 1 0 4 100 1
delay: spa_deadman [zfs](0xffff880039924000)
z_zvol/1 1 1 0 1 120 1
active: [5541]zvol_task_cb [zfs](0xffff88001fde6400)
pend: zvol_task_cb [zfs](0xffff88001fde6800)
This change adds a dedicated, per-pool, prefetch taskq to prevent the
traverse code from monopolizing the global (and limited) system_taskq by
inappropriately scheduling long running tasks on it.
Reviewed-by: Albert Lee <trisk@forkgnu.org>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: loli10K <ezomori.nozomu@gmail.com>
Closes#6330Closes#6890Closes#7343
Calling uiomove() in mappedread() under the page lock can result
in a deadlock if the user space page needs to be faulted in.
Resolve the issue by dropping the page lock before the uiomove().
The inode range lock protects against concurrent updates via
zfs_read() and zfs_write().
Reviewed-by: Albert Lee <trisk@forkgnu.org>
Reviewed-by: Chunwei Chen <david.chen@nutanix.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#7335Closes#7339
When the pool is suspended, record whether it was due to an I/O error or
due to MMP writes failing to succeed within the required time.
Change spa_suspended from uint8_t to zio_suspend_reason_t to store the
reason.
When userspace queries pool status via spa_tryimport(), report the
reason the pool was suspended in a new key,
ZPOOL_CONFIG_SUSPENDED_REASON.
In libzfs, when interpreting the returned config nvlist, report
suspension due to MMP with a new pool status enum value,
ZPOOL_STATUS_IO_FAILURE_MMP.
In status_callback(), which generates and emits the message when 'zpool
status' is executed, add a case to print an appropriate message for the
new pool status enum value.
Reviewed-by: George Melikov <mail@gmelikov.ru>
Reviewed-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Tony Hutter <hutter2@llnl.gov>
Signed-off-by: Olaf Faaland <faaland1@llnl.gov>
Closes#7296
When it's set, a DTL range will be cleared even if its scan/scrub had
errors. This allows to work around resilver/scrub upon import when the
pool has errors.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Tim Chase <tim@chase2k.com>
Closes#7293
Once per pass through the MMP thread's loop, the vdev tree is walked to
find a suitable leaf to write the next MMP block to. If no such leaf is
found, the thread sleeps for a while and resumes at the top of the loop.
Add an entry to multihost_history when no leaf can be found, and record
the reason in the error column. The error code for such entries is a
bitfield, displayed in hex:
0x1 At least one vdev (interior or leaf) was not writeable.
0x2 At least one writeable leaf vdev was found, but it had a pending
MMP write.
timestamp = the time in seconds since the epoch when no leaf could be
found originally.
duration = the time (in ns) during which no MMP block was written for
this reason. This does not include the preceeding inter-write period
nor the following inter-write period.
vdev_guid = the number of sequential cycles of the MMP thread looop when
this occurred.
Sample output, truncated to fit:
For records of skipped MMP writes the right-most column, vdev_path, is
reported as "-".
id txg timestamp error duration mmp_delay vdev_guid ...
936 11 1520036441 0 146264 891422313 1740883117838 ...
937 11 1520036441 0 163956 888356657 7320395061548 ...
938 11 1520036442 0 130690 885314969 7320395061548 ...
939 11 1520036442 0 2001068577 882296582 1740883117838 ...
940 11 1520036443 0 161806 882296582 7320395061548 ...
941 11 1520036443 0x2 0 998020546 1 ...
942 11 1520036444 0 136585 998020546 7320395061548 ...
943 11 1520036444 0x2 0 998020257 1 ...
944 11 1520036445 5 2002662964 994160219 1740883117838 ...
945 11 1520036445 0x2 998073118 994160219 3 ...
946 11 1520036447 0 247136 994160219 7320395061548 ...
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Olaf Faaland <faaland1@llnl.gov>
Closes#7212
This reverts commit cc63068e95.
Under certain circumstances this change can result in an ENOSPC
error when adding new files to a directory. See #7401 for full
details.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Tony Hutter <hutter2@llnl.gov>
Issue #7401Closes#7416
When a single pool contains more vdevs than the CONFIG_HZ for
for the kernel the mmp thread will not delay properly. Switch
to using cv_timedwait_sig_hires() to handle higher resolution
delays.
This issue was reported on Arch Linux where HZ defaults to only
100 and this could be fairly easily reproduced with a reasonably
large pool. Most distribution kernels set CONFIG_HZ=250 or
CONFIG_HZ=1000 and thus are unlikely to be impacted.
Reviewed-by: George Melikov <mail@gmelikov.ru>
Reviewed-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Reviewed-by: Olaf Faaland <faaland1@llnl.gov>
Reviewed-by: Tony Hutter <hutter2@llnl.gov>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes#7205Closes#7289
When multihost is disabled on a pool, and the pool is resumed via zpool
clear, within a single cycle of the mmp thread's loop (e.g. while it's
in the cv_timedwait call), both mmp_last_write and mmp_delay should be
updated.
The original code mistakenly treated the two cases as if they could not
occur at the same time.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Tony Hutter <hutter2@llnl.gov>
Signed-off-by: Olaf Faaland <faaland1@llnl.gov>
Closes#7286
Change file related checks to use user namespaces and make
sure involved uids/gids are mappable in the current
namespace.
Note that checks without file ownership information will
still not take user namespaces into account, as some of
these should be handled via 'zfs allow' (otherwise root in a
user namespace could issue commands such as `zpool export`).
This also adds an initial user namespace regression test
for the setgid bit loss, with a user_ns_exec helper usable
in further tests.
Additionally, configure checks for the required user
namespace related features are added for:
* ns_capable
* kuid/kgid_has_mapping()
* user_ns in cred_t
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
Closes#6800Closes#7270
If something holds the config lock as a writer for too long, MMP will
fail to issue MMP writes in a timely manner. This will result either in
the pool being suspended, or in an extreme case, in the pool not being
protected.
If the time to acquire the config lock exceeds 1/10 of the minimum
zfs_multihost_interval, report it in the zfs debug log.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Olaf Faaland <faaland1@llnl.gov>
Closes#7212
As of https://github.com/torvalds/linux/commit/fb6d47a, get_disk()
is now get_disk_and_module(). Add a configure check to determine
if we need to use get_disk_and_module().
Reviewed-by: loli10K <ezomori.nozomu@gmail.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Closes#7264
Change checksum & IO delay ratelimit thresholds from 5/sec to 20/sec.
This allows zed to actually trigger if a bunch of these events arrive in
a short period of time (zed has a threshold of 10 events in 10 sec).
Previously, if you had, say, 100 checksum errors in 1 sec, it would get
ratelimited to 5/sec which wouldn't trigger zed to fault the drive.
Also, convert the checksum and IO delay thresholds to module params for
easy testing.
Reviewed-by: loli10K <ezomori.nozomu@gmail.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Signed-off-by: Tony Hutter <hutter2@llnl.gov>
Closes#7252
In zil_lwb_commit() with TX_WRITE, we copy the log write record (lrw)
into the log write block (lwb) and send it off using zil_lwb_add_txg().
If we also have WR_NEED_COPY, we additionally copy the lwr's data into
the lwb to be sent off. If the lwr + data doesn't fit into the lwb, we
send the lrw and as much data as will fit (dnow bytes), then go back
and do the same with the remaining data.
Each time through this loop we're sending dnow data bytes. I.e.
zil_itx_needcopy_bytes should be incremented by dnow.
Reviewed-by: Richard Elling <Richard.Elling@RichardElling.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Chris Dunlop <chris@onthe.net.au>
Closes#6988Closes#7176
Provide infrastructure to auto-configure to enum and API changes in the
global page stats used for our free memory calculations.
arc_free_memory has been broken since an API change in Linux v3.14:
2016-07-28 v4.8 599d0c95 mm, vmscan: move LRU lists to node
2016-07-28 v4.8 75ef7184 mm, vmstat: add infrastructure for per-node
vmstats
These commits moved some of global_page_state() into
global_node_page_state(). The API change was particularly egregious as,
instead of breaking the old code, it silently did the wrong thing and we
continued using global_page_state() where we should have been using
global_node_page_state(), thus indexing into the wrong array via
NR_SLAB_RECLAIMABLE et al.
There have been further API changes along the way:
2017-07-06 v4.13 385386cf mm: vmstat: move slab statistics from zone to
node counters
2017-09-06 v4.14 c41f012a mm: rename global_page_state to
global_zone_page_state
...and various (incomplete, as it turns out) attempts to accomodate
these changes in ZoL:
2017-08-24 2209e409 Linux 4.8+ compatibility fix for vm stats
2017-09-16 787acae0 Linux 3.14 compat: IO acct, global_page_state, etc
2017-09-19 661907e6 Linux 4.14 compat: IO acct, global_page_state, etc
The config infrastructure provided here resolves these issues going back
to the original API change in v3.14 and is robust against further Linux
changes in this area.
Reviewed-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: George Melikov <mail@gmelikov.ru>
Signed-off-by: Chris Dunlop <chris@onthe.net.au>
Closes#7170
While the pool is suspended on host A, it may be imported on host B.
If host A continued to write MMP blocks, it would be blindly
overwriting MMP blocks written by host B, and the blocks written by
host A would have outdated txg information.
Reviewed-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Olaf Faaland <faaland1@llnl.gov>
Closes#7182
PROBLEM
=======
It's possible for a parent zio to complete even though it has children
which have not completed. This can result in the following panic:
> $C
ffffff01809128c0 vpanic()
ffffff01809128e0 mutex_panic+0x58(fffffffffb94c904, ffffff597dde7f80)
ffffff0180912950 mutex_vector_enter+0x347(ffffff597dde7f80)
ffffff01809129b0 zio_remove_child+0x50(ffffff597dde7c58, ffffff32bd901ac0,
ffffff3373370908)
ffffff0180912a40 zio_done+0x390(ffffff32bd901ac0)
ffffff0180912a70 zio_execute+0x78(ffffff32bd901ac0)
ffffff0180912b30 taskq_thread+0x2d0(ffffff33bae44140)
ffffff0180912b40 thread_start+8()
> ::status
debugging crash dump vmcore.2 (64-bit) from batfs0390
operating system: 5.11 joyent_20170911T171900Z (i86pc)
image uuid: (not set)
panic message: mutex_enter: bad mutex, lp=ffffff597dde7f80
owner=ffffff3c59b39480 thread=ffffff0180912c40
dump content: kernel pages only
The problem is that dbuf_prefetch along with l2arc can create a zio tree
which confuses the parent zio and allows it to complete with while children
still exist. Here's the scenario:
zio tree:
pio
|--- lio
The parent zio, pio, has entered the zio_done stage and begins to check its
children to see there are still some that have not completed. In zio_done(),
the children are checked in the following order:
zio_wait_for_children(zio, ZIO_CHILD_VDEV, ZIO_WAIT_DONE)
zio_wait_for_children(zio, ZIO_CHILD_GANG, ZIO_WAIT_DONE)
zio_wait_for_children(zio, ZIO_CHILD_DDT, ZIO_WAIT_DONE)
zio_wait_for_children(zio, ZIO_CHILD_LOGICAL, ZIO_WAIT_DONE)
If pio, finds any child which has not completed then it stops executing and
goes to sleep. Each call to zio_wait_for_children() will grab the io_lock
while checking the particular child.
In this scenario, the pio has completed the first call to
zio_wait_for_children() to check for any ZIO_CHILD_VDEV children. Since
the only zio in the zio tree right now is the logical zio, lio, then it
completes that call and prepares to check the next child type.
In the meantime, the lio completes and in its callback creates a child vdev
zio, cio. The zio tree looks like this:
zio tree:
pio
|--- lio
|--- cio
The lio then grabs the parent's io_lock and removes itself.
zio tree:
pio
|--- cio
The pio continues to run but has already completed its check for ZIO_CHILD_VDEV
and will erroneously complete. When the child zio, cio, completes it will panic
the system trying to reference the parent zio which has been destroyed.
SOLUTION
========
The fix is to rework the zio_wait_for_children() logic to accept a bitfield
for all the children types that it's interested in checking. The
io_lock will is held the entire time we check all the children types. Since
the function now accepts a bitfield, a simple ZIO_CHILD_BIT() macro is provided
to allow for the conversion between a ZIO_CHILD type and the bitfield used by
the zio_wiat_for_children logic.
Authored by: George Wilson <george.wilson@delphix.com>
Reviewed by: Matthew Ahrens <mahrens@delphix.com>
Reviewed by: Andriy Gapon <avg@FreeBSD.org>
Reviewed by: Youzhong Yang <youzhong@gmail.com>
Reviewed by: Brian Behlendorf <behlendorf1@llnl.gov>
Approved by: Dan McDonald <danmcd@omniti.com>
Ported-by: Giuseppe Di Natale <dinatale2@llnl.gov>
OpenZFS-issue: https://www.illumos.org/issues/8857
OpenZFS-commit: https://github.com/openzfs/openzfs/commit/862ff6d99c
Issue #5918Closes#7168
mmp_write_uberblock() and mmp_write_done() should the same tag
for spa_config_locks.
Reviewed-by: Olaf Faaland <faaland1@llnl.gov>
Reviewed-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Reviewed by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Sanjeev Bagewadi <sanjeev.bagewadi@gmail.com>
Closes#6530Closes#7155
With "casesensitivity=mixed", zap_add() could fail when the number of
files/directories with the same name (varying in case) exceed the
capacity of the leaf node of a Fatzap. This results in a ASSERT()
failure as zfs_link_create() does not expect zap_add() to fail. The fix
is to handle these failures and rollback the transactions.
Reviewed by: Matt Ahrens <mahrens@delphix.com>
Reviewed-by: Chunwei Chen <david.chen@nutanix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Sanjeev Bagewadi <sanjeev.bagewadi@gmail.com>
Closes#7011Closes#7054
If a corruption happens to be on a root block of an objset, zdb -c will
not correctly report the error, and it will not traverse the datasets
that come after. This is because traverse_visitbp, which does the
callback and reset error for TRAVERSE_HARD, is skipped when traversing
zil is failed in traverse_impl.
Here's example of what 'zdb -eLcc' command looks like on a pool with
damaged objset root:
== before patch:
Traversing all blocks to verify checksums ...
Error counts:
errno count
block traversal size 379392 != alloc 33987072 (unreachable 33607680)
bp count: 172
ganged count: 0
bp logical: 1678336 avg: 9757
bp physical: 130560 avg: 759 compression: 12.85
bp allocated: 379392 avg: 2205 compression: 4.42
bp deduped: 0 ref>1: 0 deduplication: 1.00
SPA allocated: 33987072 used: 0.80%
additional, non-pointer bps of type 0: 71
Dittoed blocks on same vdev: 101
== after patch:
Traversing all blocks to verify checksums ...
zdb_blkptr_cb: Got error 52 reading <54, 0, -1, 0> -- skipping
Error counts:
errno count
52 1
block traversal size 33963520 != alloc 33987072 (unreachable 23552)
bp count: 447
ganged count: 0
bp logical: 36093440 avg: 80745
bp physical: 33699840 avg: 75391 compression: 1.07
bp allocated: 33963520 avg: 75981 compression: 1.06
bp deduped: 0 ref>1: 0 deduplication: 1.00
SPA allocated: 33987072 used: 0.80%
additional, non-pointer bps of type 0: 76
Dittoed blocks on same vdev: 115
==
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: loli10K <ezomori.nozomu@gmail.com>
Signed-off-by: Chunwei Chen <david.chen@nutanix.com>
Closes#7099
