illumos/illumos-gate@42b141117242b1411172https://www.illumos.org/issues/8648
I'm opening this bug to track integration of the following ZFS on Linux
commit into illumos:
commit f763c3d1df
Author: LOLi <loli10K@users.noreply.github.com>
Date: Mon Aug 21 17:59:48 2017 +0200
Fix range locking in ZIL commit codepath
Since OpenZFS 7578 (1b7c1e5) if we have a ZVOL with logbias=throughput
we will force WR_INDIRECT itxs in zvol_log_write() setting itx->itx_lr
offset and length to the offset and length of the BIO from
zvol_write()->zvol_log_write(): these offset and length are later used
to take a range lock in zillog->zl_get_data function: zvol_get_data().
Now suppose we have a ZVOL with blocksize=8K and push 4K writes to
offset 0: we will only be range-locking 0-4096. This means the
ASSERTion we make in dbuf_unoverride() is no longer valid because now
dmu_sync() is called from zilog's get_data functions holding a partial
lock on the dbuf.
Fix this by taking a range lock on the whole block in zvol_get_data().
Reviewed-by: Chunwei Chen <tuxoko@gmail.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: loli10K <ezomori.nozomu@gmail.com>
Reviewed by: Igor Kozhukhov <igor@dilos.org>
Reviewed by: Matt Ahrens <mahrens@delphix.com>
Reviewed by: Andriy Gapon <avg@FreeBSD.org>
Reviewed by: Alexander Motin <mav@FreeBSD.org>
Approved by: Robert Mustacchi <rm@joyent.com>
Author: LOLi <loli10K@users.noreply.github.com>
MFC after: 10 days
illumos/illumos-gate@bd9d3f9046bd9d3f9046https://www.illumos.org/issues/8661
The "zil-cw1" dtrace probe was previously removed in 8558, and the "zil-cw2"
probe should have been removed in that patch as well. Unfortunately, the "zil-
cw2" was not removed in 8558, so this bug is to track it's removal.
Reviewed by: Serapheim Dimitropoulos <serapheim@delphix.com>
Reviewed by: Matthew Ahrens <mahrens@delphix.com>
Reviewed by: Igor Kozhukhov <igor@dilos.org>
Approved by: Robert Mustacchi <rm@joyent.com>
Author: Prakash Surya <prakash.surya@delphix.com>
MFC after: 1 week
illumos/illumos-gate@554675eee7554675eee7https://www.illumos.org/issues/8473
Scrubbing is supposed to detect and repair all errors in the pool. However,
it wrongly ignores active spare devices. The problem can easily be
reproduced in OpenZFS at git rev 0ef125d with these commands:
truncate -s 64m /tmp/a /tmp/b /tmp/c
sudo zpool create testpool mirror /tmp/a /tmp/b spare /tmp/c
sudo zpool replace testpool /tmp/a /tmp/c
/bin/dd if=/dev/zero bs=1024k count=63 oseek=1 conv=notrunc of=/tmp/c
sync
sudo zpool scrub testpool
zpool status testpool # Will show 0 errors, which is wrong
sudo zpool offline testpool /tmp/a
sudo zpool scrub testpool
zpool status testpool # Will show errors on /tmp/c,
# which should've already been fixed
FreeBSD head is partially affected: the first scrub will detect some errors, but the second scrub will detect more.
Reviewed by: Andy Stormont <astormont@racktopsystems.com>
Reviewed by: Matt Ahrens <mahrens@delphix.com>
Reviewed by: George Wilson <george.wilson@delphix.com>
Approved by: Richard Lowe <richlowe@richlowe.net>
MFC after: 1 week
Sponsored by: Spectra Logic Corp
It is reported that the default value of 4KB results in a substantial
memory use overhead (at least, on some configurations). Using 1KB seems
to reduce the overhead significantly.
PR: 222377
Reported by: Sean Chittenden <sean@chittenden.org>
MFC after: 1 week
I overlooked the fact that that ZIO_IOCTL_PIPELINE does not include
ZIO_STAGE_VDEV_IO_DONE stage. We do allocate a struct bio for an ioctl
zio (a disk cache flush), but we never freed it.
This change splits bio handling into two groups, one for normal
read/write i/o that passes data around and, thus, needs the abd data
tranform; the other group is for "data-less" i/o such as trim and cache
flush.
PR: 222288
Reported by: Dan Nelson <dnelson@allantgroup.com>
Tested by: Borja Marcos <borjam@sarenet.es>
MFC after: 10 days
illumos/illumos-gate@2bcb5458542bcb545854https://www.illumos.org/issues/8602
When I landed the fix for 8558, I incorrectly added the "dp_early_sync_tasks"
field to the "dsl_pool" structure. This field is used in DelphixOS, but not in
illumos. It was incorrectly pulled into illumos, so this bug is to remove it
from the structure.
Reviewed by: Serapheim Dimitropoulos <serapheim@delphix.com>
Reviewed by: Matthew Ahrens <mahrens@delphix.com>
Approved by: Robert Mustacchi <rm@joyent.com>
Author: Prakash Surya <prakash.surya@delphix.com>
MFC after: 1 week
As long as mnt_ref is not zero there can be a consumer that might try
to access mnt_vnodecovered. For this reason the covered vnode must not
be freed until mnt_ref goes to zero.
So, move the release of the covered vnode to vfs_mount_destroy.
Reviewed by: kib
MFC after: 3 weeks
Differential Revision: https://reviews.freebsd.org/D12329
The zfsonlinux feature large_dnode is not yet supported by the loader.
Reviewed by: avg, allanjude
Differential Revision: https://reviews.freebsd.org/D12288
The uncovered vnode is possible because there is no guarantee that
its hold count would go to zero (and it would be inactivated and reclaimed)
immediately after a covering filesystem is unmounted.
So, such a vnode should be expected and it is possible to re-use it
without any trouble.
MFC after: 3 weeks
Sponsored by: Panzura
The only consumer of zfs_get_vfs, zfs_unmount_snap, does not need
the filesystem to be busy, it just need a reference that it can pass
to dounmount.
Also, previously the code was racy as it unbusied the filesystem
before taking a reference on it.
Now the code should be simpler and safer.
MFC after: 2 weeks
Sponsored by: Panzura
illumos/illumos-gate@37e84ab74e37e84ab74ehttps://www.illumos.org/issues/8569
C [C99] has peculiar rules for inline functions that are different from the
C++ rules. Unlike C++ where inline is "fire and forget", in C a programmer
must pay attention to the function's storage class / visibility. The main
problem is with the case where a compiler decides to not inline a call to the
function declared as inline.
Some relevant links:
- http://infocenter.arm.com/help/index.jsp?topic=/com.arm.doc.faqs/ka15831.html
- http://www.drdobbs.com/the-new-c-inline-functions/184401540
The summary is that either the inline functions should be declared 'static
inline' or one of the compilation units (.c files) must provide a callable
externally visible function definition. In the former case, the compiler would
automatically create a local non-inlined function instance in every compilation
unit where it's needed. In the latter case the single external definition is
used to satisfy any non-inlined calls in all compilation units. As things
stand right now, we can get an undefined reference error under certain
combinations of compilers and compiler options. For example, this is what I
get on FreeBSD when compiling with clang 4.0.0 and -O1:
In function `abd_free': /usr/src/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/abd.c:385:
undefined reference to `abd_is_linear'
Reviewed by: Matt Ahrens <mahrens@delphix.com>
Approved by: Robert Mustacchi <rm@joyent.com>
Author: Andriy Gapon <avg@FreeBSD.org>
MFC after: 1 week
illumos/illumos-gate@216d7723a1216d7723a1https://www.illumos.org/issues/8558
On a system with more than 80K ZFS filesystems, we've seen cases where
lwp_create() will start to fail by returning EAGAIN. The problem being,
for each of those 80K ZFS filesystems, a taskq will be created for each
dataset as part of the ZIL for each dataset.
For each of these taskq's, a kernel thread will be created which results
in 24KB being allocated for each thread. With enough of these 24KB
allocations, we eventually exhaust the memory region set aside for these
allocations. Currently, segkpsize is set to a value of 2GB, which means
we can only support about 80K filesystems; 2GB / 24KB = ~80K.
The lwp_create() failure comes into play due to the fact that LWP
creation also allocates 24KB from this same region of memory. Thus, if
we've exhausted this region of memory due to the number of ZIL taskq's,
there won't be any memory avaible to allow the call to lwp_create() to
succeed.
FreeBSD note: I haven't created sysctl-s for the new ZIL clean
parameters. Let's add them if anyone requires to tune them.
Reviewed by: George Wilson <george.wilson@delphix.com>
Reviewed by: Sebastien Roy <sebastien.roy@delphix.com>
Approved by: Robert Mustacchi <rm@joyent.com>
Author: Prakash Surya <prakash.surya@delphix.com>
MFC after: 3 weeks
illumos/illumos-gate@1702cce7511702cce751
FreeBSD note: rather than merging the zpool.8 update I copied the zpool
scrub section from the illumos zpool.1m to FreeBSD zpool.8 almost
verbatim. Now that the illumos page uses the mdoc format, it was an
easier option. Perhaps the change is not in perfect compliance with the
FreeBSD style, but I think that it is acceptible.
https://www.illumos.org/issues/8414
This issue tracks the port of scrub pause from ZoL: https://github.com/zfsonlinux/zfs/pull/6167
Currently, there is no way to pause a scrub. Pausing may be useful when
the pool is busy with other I/O to preserve bandwidth.
Description
This patch adds the ability to pause and resume scrubbing. This is achieved
by maintaining a persistent on-disk scrub state. While the state is 'paused'
we do not scrub any more blocks. We do however perform regular scan
housekeeping such as freeing async destroyed and deadlist blocks while paused.
Motivation and Context
Scrub pausing can be an I/O intensive operation and people have been asking
for the ability to pause a scrub for a while. This allows one to preserve scrub
progress while freeing up bandwidth for other I/O.
Reviewed by: George Melikov <mail@gmelikov.ru>
Reviewed by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed by: Brad Lewis <brad.lewis@delphix.com>
Reviewed by: Serapheim Dimitropoulos <serapheim@delphix.com>
Reviewed by: Matt Ahrens <mahrens@delphix.com>
Approved by: Dan McDonald <danmcd@joyent.com>
Author: Alek Pinchuk <apinchuk@datto.com>
MFC after: 2 weeks
Turn on the required options in the ERL config file, and ensure
that the fbt module is listed as a dependency for mips in
the modules/dtrace/dtraceall/dtraceall.c file.
PR: 220346
Reviewed by: gnn, markj
MFC after: 2 weeks
Differential Revision: https://reviews.freebsd.org/D12227
"zfs mount -o" passes a list of mount options directly to nmount(2) after
sanity checking them. In particular, zfs(8) will refuse to mount an already
existing file system unless "remount" is specified in the option list.
However, the "remount" option only exists in Illumos. FreeBSD's equivalent is
"update".
PR: 221985
Reviewed by: avg
MFC after: 3 weeks
Sponsored by: Spectra Logic Corp
Differential Revision: https://reviews.freebsd.org/D12233
The default value for arc_no_grow_shift may not be optimal when using
several GiB ARC. Expose it via sysctl allows users to tune it easily.
Also expose arc_grow_retry via sysctl for the same reason. The default value of
60s might, in case of intensive load, be too long.
Submitted by: Nikita Kozlov <nikita.kozlov@blade-group.com>
Reviewed by: mav, manu, bapt
MFC after: 2 weeks
Sponsored by: blade
Differential Revision: https://reviews.freebsd.org/D12144
illumos 4185 ("add new cryptographic checksums to ZFS: SHA-512,
Skein, Edon-R") was intentionally merged only partially in r289422,
without adding support for skein, sha512 and edonr on FreeBSD.
Support for skein and sha512 was added later on, but edonr is still not
implemented in FreeBSD.
Prior to this commit zfs(8) correctly rejected edonr, but with an error
message that claimed support:
fk@r500 ~ $zfs set checksum=edonr tank
cannot set property for 'tank': 'checksum' must be one of 'on | off | fletcher2 | fletcher4 | sha256 | sha512 | skein | edonr'
PR: 204055
Submitted by: Fabian Keil
Approved by: allanjude
Obtained from: ElectroBSD
MFC after: 1 week
When built with -fno-inline-functions zfs.ko contains undefined references
to these functions if they are only marked inline.
Reviewed by: avg (earlier version)
MFC after: 1 week
Sponsored by: Chelsio Communications
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/vdev_geom.c
Be more careful about the use of provider names vs vdev names in
ZFS_LOG statements.
MFC after: 3 weeks
Sponsored by: Spectra Logic Corp
illumos/illumos-gate@d28671a3b0d28671a3b0https://www.illumos.org/issues/8373
The code that writes ZIL blocks uses dmu_tx_assign(TXG_WAIT) to assign
a transaction to a transaction group. That seems to be logically
incorrect as writing of the ZIL block does not introduce any new dirty
data. Also, when there is a lot of dirty data, the call can introduce
significant delays into the ZIL commit path, thus affecting all
synchronous writes. Additionally, ARC throttling may affect the ZIL
writing.
Reviewed by: Matthew Ahrens <mahrens@delphix.com>
Reviewed by: Prakash Surya <prakash.surya@delphix.com>
Approved by: Dan McDonald <danmcd@joyent.com>
Author: Andriy Gapon <avg@FreeBSD.org>
MFC after: 2 weeks
illumos/illumos-gate@79c2b812ee79c2b812eehttps://www.illumos.org/issues/8491
The zpool checkpoint feature in DxOS added a new field in the uberblock.
The Multi-Modifier Protection Pull Request from ZoL adds two new fields in the
uberblock (Reference: https://github.com/zfsonlinux/zfs/pull/6279).
As these two changes come from two different sources and once upstreamed and
deployed will introduce an incompatibility with each other we want
to upstream a change that will reserve the padding for both of them so
integration goes smoothly and everyone gets both features.
Reviewed by: Matthew Ahrens <mahrens@delphix.com>
Reviewed by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed by: Olaf Faaland <faaland1@llnl.gov>
Approved by: Gordon Ross <gwr@nexenta.com>
Author: Serapheim Dimitropoulos <serapheim@delphix.com>
MFC after: 3 weeks
illumos/illumos-gate@267ae6c3a8267ae6c3a8https://www.illumos.org/issues/7915
l2arc_evict() is strictly serialized with respect to
l2arc_write_buffers() and l2arc_write_done(). Normally, l2arc_evict()
and l2arc_write_buffers() are called from the same thread, so they can
not be concurrent. Also, l2arc_write_buffers() uses zio_wait() on the
parent zio of all cache zio-s. That ensures that l2arc_write_done()
is completed before l2arc_write_buffers() returns. Finally, if a
cache device is removed, then l2arc_evict() is called under SCL_ALL in
the exclusive mode. That ensures that it can not be concurrent with
the normal L2ARC accesses to the device (including writing and
evicting buffers). Given the above, some checks and actions in
l2arc_evict() do not make sense. For instance, it must never
encounter the write head header let alone remove it from the buffer
list.
Reviewed by: Dan Kimmel <dan.kimmel@delphix.com>
Reviewed by: Prakash Surya <prakash.surya@delphix.com>
Approved by: Matthew Ahrens <mahrens@delphix.com>
Author: Andriy Gapon <avg@FreeBSD.org>
MFC after: 2 weeks
illumos/illumos-gate@dcb6872c56dcb6872c56https://www.illumos.org/issues/8126
The sync thread is concurrently modifying dn_phys->dn_nlevels
while dbuf_dirty() is trying to assert something about it, without
holding the necessary lock. We need to move this assertion further down
in the function, after we have acquired the dn_struct_rwlock.
Reviewed by: Pavel Zakharov <pavel.zakharov@delphix.com>
Reviewed by: Serapheim Dimitropoulos <serapheim@delphix.com>
Approved by: Robert Mustacchi <rm@joyent.com>
Author: Matthew Ahrens <mahrens@delphix.com>
MFC after: 2 weeks
illumos/illumos-gate@9b195260e29b195260e2https://www.illumos.org/issues/8426
abd_copy_from_buf and abd_cmp_buf do not modify their void *buf arguments, so
qualify them with const.
abd_copy_from_buf_off and abd_cmp_buf_off already had that type for the
corresponding arguments.
Reviewed by: Matt Ahrens <mahrens@delphix.com>
Approved by: Robert Mustacchi <rm@joyent.com>
Author: Andriy Gapon <avg@FreeBSD.org>
MFC after: 2 weeks
illumos/illumos-gate@77b171372e77b171372ehttps://www.illumos.org/issues/7600
At present, the kernel side code seems to blindly rollback to whatever happens
to be the latest snapshot at the time when the rollback task is processed.
The expected target's name should be passed to the kernel driver and the sync
task should validate that the target exists and that it is the latest snapshot
indeed.
Reviewed by: Matthew Ahrens <mahrens@delphix.com>
Reviewed by: Pavel Zakharov <pavel.zakharov@delphix.com>
Approved by: Robert Mustacchi <rm@joyent.com>
Author: Andriy Gapon <avg@FreeBSD.org>
MFC after: 3 weeks
illumos/illumos-gate@42418f9e7342418f9e73https://www.illumos.org/issues/8377
The problem is that when dsl_bookmark_destroy_check() is executed from open
context (the pre-check), it fills in dbda_success based on the existence of the
bookmark.
But the bookmark (or containing filesystem as in this case) can be destroyed
before we get to syncing context. When we re-run dsl_bookmark_destroy_check()
in syncing
context, it will not add the deleted bookmark to dbda_success, intending for
dsl_bookmark_destroy_sync() to not process it. But because the bookmark is
still in dbda_success
from the open-context call, we do try to destroy it.
The fix is that dsl_bookmark_destroy_check() should not modify dbda_success
when called from open context.
Reviewed by: Paul Dagnelie <pcd@delphix.com>
Reviewed by: Pavel Zakharov <pavel.zakharov@delphix.com>
Reviewed by: George Wilson <george.wilson@delphix.com>
Approved by: Robert Mustacchi <rm@joyent.com>
Author: Matthew Ahrens <mahrens@delphix.com>
MFC after: 2 weeks
illumos/illumos-gate@b7edcb9408b7edcb9408https://www.illumos.org/issues/8378
The problem is that zfs_get_data() supplies a stale zgd_bp to dmu_sync(), which
we then nopwrite against.
zfs_get_data() doesn't hold any DMU-related locks, so after it copies db_blkptr
to zgd_bp, dbuf_write_ready()
could change db_blkptr, and dbuf_write_done() could remove the dirty record.
dmu_sync() then sees the stale
BP and that the dbuf it not dirty, so it is eligible for nop-writing.
The fix is for dmu_sync() to copy db_blkptr to zgd_bp after acquiring the
db_mtx. We could still see a stale
db_blkptr, but if it is stale then the dirty record will still exist and thus
we won't attempt to nopwrite.
Reviewed by: Prakash Surya <prakash.surya@delphix.com>
Reviewed by: George Wilson <george.wilson@delphix.com>
Approved by: Robert Mustacchi <rm@joyent.com>
Author: Matthew Ahrens <mahrens@delphix.com>
MFC after: 2 weeks
FreeBD note: the essence of this change was committed to FreeBSD in
r314274. This commit catches up with differences between what was
committed to FreeBSD and what was committed to OpenZFS, mainly more
logical variable names.
illumos/illumos-gate@16a7e5ac1116a7e5ac11https://www.illumos.org/issues/7910
It seems that the change in issue #6950 resurrected the problem that was
earlier fixed by the change in issue #5219.
Please also see the following FreeBSD bug report:
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=216178
Reviewed by: George Wilson <george.wilson@delphix.com>
Reviewed by: Dan Kimmel <dan.kimmel@delphix.com>
Approved by: Robert Mustacchi <rm@joyent.com>
Author: Andriy Gapon <avg@FreeBSD.org>
MFC after: 2 weeks
o Replace __riscv64 with (__riscv && __riscv_xlen == 64)
This is required to support new GCC 7.1 compiler.
This is compatible with current GCC 6.1 compiler.
RISC-V is extensible ISA and the idea here is to have built-in define
per each extension, so together with __riscv we will have some subset
of these as well (depending on -march string passed to compiler):
__riscv_compressed
__riscv_atomic
__riscv_mul
__riscv_div
__riscv_muldiv
__riscv_fdiv
__riscv_fsqrt
__riscv_float_abi_soft
__riscv_float_abi_single
__riscv_float_abi_double
__riscv_cmodel_medlow
__riscv_cmodel_medany
__riscv_cmodel_pic
__riscv_xlen
Reviewed by: ngie
Sponsored by: DARPA, AFRL
Differential Revision: https://reviews.freebsd.org/D11901
That is required to support reboot -r with a new root filesystem being
on an already imported pool.
PR: 210721
Reported by: Jan Bramkamp <crest_maintainer@rlwinm.de>
MFC after: 2 weeks
The description is FreeBSD-specific and was added in r266497
to fix PR189865.
PR: 220825
Submitted by: Fabian Keil
Obtained from: ElectroBSD
MFC after: 1 week
I overlooked the fact that vdev_op_io_done hook is called even if the
actual I/O is skipped, for example, in the case of a missing vdev.
Arguably, this could be considered an issue in the zio pipeline engine,
but for now I am adding defensive code to check for io_bp being NULL
along with assertions that that happens only when it can be really
expected.
PR: 220691
Reported by: peter, cy
Tested by: cy
MFC after: 1 week
X-MFC with: r320156, r320452
ZPL_VERSION is unsigned long long, not an int. With this change, a zpool can be
created on a 32-bit system (tested on powerpcspe) and mounted correctly.
Reviewed by: allanjude
The implementation of ZFS refcount_t uses the emulated illumos mutex
(the sx lock) and the waiting memory allocation when ZFS_DEBUG is
enabled. This makes refcount_t unsuitable for use in GEOM g_up
thread where sleeping is prohibited.
When importing the ABD change I modified vdev_geom using illumos
vdev_disk as an example. As a result, I added a call to abd_return_buf
in vdev_geom_io_intr. The latter is called on g_up thread while the
former uses refcount_t.
This change fixes the problem by deferring the abd_return_buf call to
the previously unused vdev_geom_io_done that is called on a ZFS zio
taskqueue thread where sleeping is allowed.
A side bonus of this change is that now a vdev zio has a pointer
to its corresponding bio while the zio is active.
Reported by: Shawn Webb <shawn.webb@hardenedbsd.org>
Tested by: Shawn Webb <shawn.webb@hardenedbsd.org>
MFC after: 1 week
X-MFC with: r320156
3306 zdb should be able to issue reads in parallel
illumos/illumos-gate/31d7e8fa33fae995f558673adb22641b5aa8b6e1
https://www.illumos.org/issues/3306
The upstream change was made before we started to import upstream commits
individually. It was imported into the illumos vendor area as r242733.
That commit was MFV-ed in r260138, but as the commit message says
vdev_file.c was left intact.
This commit actually implements the parallel I/O for vdev_file using a
taskqueue with multiple thread. This implementation does not depend on
the illumos or FreeBSD bio interface at all, but uses zio_t to pass
around all the relevent data. So, the code looks a bit different from
the upstream.
This commit also incorporates ZoL commit
zfsonlinux/zfs/bc25c9325b0e5ced897b9820dad239539d561ec9 that fixed
https://github.com/zfsonlinux/zfs/issues/2270
We need to use a dedicated taskqueue for exactly the same reason as ZoL
as we do not implement TASKQ_DYNAMIC.
Obtained from: illumos, ZFS on Linux
MFC after: 2 weeks
FreeBSD note: the actual change has been in FreeBSD since r297848. This
commit accounts for integration of that change with subsequent changes,
especially r320156 (MFV of r318946) and r314274.
illumos/illumos-gate@403a8da73c403a8da73chttps://www.illumos.org/issues/5220
There are disk devices that have logical sector size larger than 512B, for
example 4KB. That is, their physical sector size is larger than 512B and they
do not provide emulation for 512B sector sizes. For such devices both a data
offset and a data size must be properly aligned. L2ARC should arrange that
because it uses physical I/O.
zio_vdev_io_start() performs a necessary transformation if io_size is not
aligned to vdev_ashift, but that is done only for logical I/O. Something
similar should be done in L2ARC code.
* a temporary write buffer should be allocated if the original buffer is
not going to be compressed and its size is not aligned
* size of a temporary compression buffer should be ashift aligned
* for the reads, if a size of a target buffer is not sufficiently large and
it is not aligned then a temporary read buffer should be allocated
Reviewed by: George Wilson <george.wilson@delphix.com>
Reviewed by: Dan Kimmel <dan.kimmel@delphix.com>
Reviewed by: Saso Kiselkov <saso.kiselkov@nexenta.com>
Approved by: Dan McDonald <danmcd@joyent.com>
Author: Andriy Gapon <avg@FreeBSD.org>
MFC after: 3 weeks
illumos/illumos-gate@0255edcc850255edcc85https://www.illumos.org/issues/8056
The send size estimate for a zvol can be too low, if the size of the record
headers (dmu_replay_record_t's) is a significant portion of the size.
This is typically the case when the data is highly compressible, especially
with embedded blocks.
The problem is that dmu_adjust_send_estimate_for_indirects() assumes that
blocks are the size of the "recordsize" property (128KB).
However, for zvols, the blocks are the size of the "volblocksize" property
(8KB). Therefore, we estimate that there will be 16x less record headers than
there really will be.
Reviewed by: Matthew Ahrens <mahrens@delphix.com>
Reviewed by: Pavel Zakharov <pavel.zakharov@delphix.com>
Approved by: Robert Mustacchi <rm@joyent.com>
Author: Paul Dagnelie <pcd@delphix.com>
MFC after: 3 weeks
FreeBSD note: this commit removes small differences between what mav
committed to FreeBSD in r308782 and what ended up committed to illumos
after addressing all review comments.
illumos/illumos-gate@c5ee46810fc5ee46810fhttps://www.illumos.org/issues/7578
After some ZIL changes 6 years ago zil_slog_limit got partially broken
due to zl_itx_list_sz not updated when async itx'es upgraded to sync.
Actually because of other changes about that time zl_itx_list_sz is not
really required to implement the functionality, so this patch removes
some unneeded broken code and variables.
Original idea of zil_slog_limit was to reduce chance of SLOG abuse by
single heavy logger, that increased latency for other (more latency critical)
loggers, by pushing heavy log out into the main pool instead of SLOG. Beside
huge latency increase for heavy writers, this implementation caused double
write of all data, since the log records were explicitly prepared for SLOG.
Since we now have I/O scheduler, I've found it can be much more efficient
to reduce priority of heavy logger SLOG writes from ZIO_PRIORITY_SYNC_WRITE
to ZIO_PRIORITY_ASYNC_WRITE, while still leave them on SLOG.
Existing ZIL implementation had problem with space efficiency when it
has to write large chunks of data into log blocks of limited size. In some
cases efficiency stopped to almost as low as 50%. In case of ZIL stored on
spinning rust, that also reduced log write speed in half, since head had to
uselessly fly over allocated but not written areas. This change improves
the situation by offloading problematic operations from z*_log_write() to
zil_lwb_commit(), which knows real situation of log blocks allocation and
can split large requests into pieces much more efficiently. Also as side
effect it removes one of two data copy operations done by ZIL code WR_COPIED
case.
While there, untangle and unify code of z*_log_write() functions.
Also zfs_log_write() alike to zvol_log_write() can now handle writes crossing
block boundary, that may also improve efficiency if ZPL is made to do that.
Reviewed by: Matthew Ahrens <mahrens@delphix.com>
Reviewed by: Prakash Surya <prakash.surya@delphix.com>
Reviewed by: Andriy Gapon <avg@FreeBSD.org>
Reviewed by: Steven Hartland <steven.hartland@multiplay.co.uk>
Reviewed by: Brad Lewis <brad.lewis@delphix.com>
Reviewed by: Richard Elling <Richard.Elling@RichardElling.com>
Approved by: Robert Mustacchi <rm@joyent.com>
Author: Alexander Motin <mav@FreeBSD.org>
MFC after: 3 weeks
All of the problems were related to the FreeBSD-only features.
One was caused by a mismerge in the zfsbootcfg support code.
All others were in the TRIM support code.
MFC after: 1 week
X-MFC with: r320156
All of the problems were related to the FreeBSD-only features.
One was caused by a mismerge in the zfsbootcfg support code.
All others were in the TRIM support code.
Reported by: ken,
O. Hartmann <ohartmann@walstatt.org>,
Trond Endrestøl <Trond.Endrestol@fagskolen.gjovik.no>
MFC after: 1 week
X-MFC with: r320156
illumos/illumos-gate@770499e185770499e185https://www.illumos.org/issues/8021
The ARC buf data project (known simply as "ABD" since its genesis in the ZoL
community) changes the way the ARC allocates `b_pdata` memory from using linear
`void *` buffers to using scatter/gather lists of fixed-size 1KB chunks. This
improves ZFS's performance by helping to defragment the address space occupied
by the ARC, in particular for cases where compressed ARC is enabled. It could
also ease future work to allocate pages directly from `segkpm` for minimal-
overhead memory allocations, bypassing the `kmem` subsystem.
This is essentially the same change as the one which recently landed in ZFS on
Linux, although they made some platform-specific changes while adapting this
work to their codebase:
1. Implemented the equivalent of the `segkpm` suggestion for future work
mentioned above to bypass issues that they've had with the Linux kernel memory
allocator.
2. Changed the internal representation of the ABD's scatter/gather list so it
could be used to pass I/O directly into Linux block device drivers. (This
feature is not available in the illumos block device interface yet.)
FreeBSD notes:
- the actual (default) chunk size is 4KB (despite the text above saying 1KB)
- we can try to reimplement ABDs, so that they are not permanently
mapped into the KVA unless explicitly requested, especially on
platforms with scarce KVA
- we can try to use unmapped I/O and avoid intermediate allocation of a
linear, virtual memory mapped buffer
- we can try to avoid extra data copying by referring to chunks / pages
in the original ABD
Reviewed by: Matthew Ahrens <mahrens@delphix.com>
Reviewed by: George Wilson <george.wilson@delphix.com>
Reviewed by: Paul Dagnelie <pcd@delphix.com>
Reviewed by: John Kennedy <john.kennedy@delphix.com>
Reviewed by: Prakash Surya <prakash.surya@delphix.com>
Reviewed by: Prashanth Sreenivasa <pks@delphix.com>
Reviewed by: Pavel Zakharov <pavel.zakharov@delphix.com>
Reviewed by: Chris Williamson <chris.williamson@delphix.com>
Approved by: Richard Lowe <richlowe@richlowe.net>
Author: Dan Kimmel <dan.kimmel@delphix.com>
MFC after: 3 weeks
