Fix several problems that can cause panics on kldload and kldunload.
* kproc_create(fasttrap_pid_cleanup_cb, ...) gets called before
fasttrap_provs.fth_table gets allocated. This can lead to a panic
on module load, because fasttrap_pid_cleanup_cb references
fasttrap_provs.fth_table. Move kproc_create down after the point
that fasttrap_provs.fth_table gets allocated, and modify the error
handling accordingly.
* dtrace_fasttrap_{fork,exec,exit} weren't getting NULLed until
after fasttrap_provs.fth_table got freed. That caused panics on
module unload because fasttrap_exec_exit calls
fasttrap_provider_retire, which references
fasttrap_provs.fth_table. NULL those function pointers earlier.
* There wasn't any code to destroy the
fasttrap_{tpoints,provs,procs}.fth_table mutexes on module unload,
leading to a resource leak when WITNESS is enabled. Destroy those
mutexes during fasttrap_unload().
Reviewed by: markj
Approved by: ken (mentor)
Sponsored by: Spectra Logic
MFC after: 4 weeks
It was not being correctly copied into the kernel on FreeBSD, and as a
result, probes with multiple probe sites were not being created properly.
To fix this, change the ioctl definition so that the fasttrap ioctl handler
is responsible for copying in userland data.
Submitted by: Prashanth Kumar <pra_udupi@yahoo.co.in>
MFC after: 1 month
VM subsystem twice for every written record.
Tests on 24-core system show double reduction of CPU time spent on copying
single large well-compressed file.
This patch is not really needed on illumos (while not harm either) since
their memory allocator by default uses caching for all requests up to 128K.
Reviewed by: Saso Kiselkov <skiselkov.ml@gmail.com>
emulation of the call instruction caused by reversing the uaddr and kaddr
arguments when copying data out to userland: the suword* functions take the
uaddr as the first argument whereas copyout(9) takes the kaddr as the first
argument. This also partially undoes the fixes from r257143.
Submitted by: Prashanth Kumar <pra_udupi@yahoo.co.in> (original version)
MFC after: 1 month
the code was trying to save the stack pointer rather than the frame pointer,
and the arguments to copyout(9) were reversed, so nothing ended up being
saved on the stack. This would cause process crashes when the pid provider
was being used to instrument calls of a function starting with this
instruction.
Reported by: symbolics@gmx.com
Tested by: symbolics@gmx.com (earlier version)
MFC after: 2 weeks
information.
The existing algorithm selects a preferred leaf vdev based on offset of the zio
request modulo the number of members in the mirror. It assumes the devices are
of equal performance and that spreading the requests randomly over both drives
will be sufficient to saturate them. In practice this results in the leaf vdevs
being under utilized.
The new algorithm takes into the following additional factors:
* Load of the vdevs (number outstanding I/O requests)
* The locality of last queued I/O vs the new I/O request.
Within the locality calculation additional knowledge about the underlying vdev
is considered such as; is the device backing the vdev a rotating media device.
This results in performance increases across the board as well as significant
increases for predominantly streaming loads and for configurations which don't
have evenly performing devices.
The following are results from a setup with 3 Way Mirror with 2 x HD's and
1 x SSD from a basic test running multiple parrallel dd's.
With pre-fetch disabled (vfs.zfs.prefetch_disable=1):
== Stripe Balanced (default) ==
Read 15360MB using bs: 1048576, readers: 3, took 161 seconds @ 95 MB/s
== Load Balanced (zfslinux) ==
Read 15360MB using bs: 1048576, readers: 3, took 297 seconds @ 51 MB/s
== Load Balanced (locality freebsd) ==
Read 15360MB using bs: 1048576, readers: 3, took 54 seconds @ 284 MB/s
With pre-fetch enabled (vfs.zfs.prefetch_disable=0):
== Stripe Balanced (default) ==
Read 15360MB using bs: 1048576, readers: 3, took 91 seconds @ 168 MB/s
== Load Balanced (zfslinux) ==
Read 15360MB using bs: 1048576, readers: 3, took 108 seconds @ 142 MB/s
== Load Balanced (locality freebsd) ==
Read 15360MB using bs: 1048576, readers: 3, took 48 seconds @ 320 MB/s
In addition to the performance changes the code was also restructured, with
the help of Justin Gibbs, to provide a more logical flow which also ensures
vdevs loads are only calculated from the set of valid candidates.
The following additional sysctls where added to allow the administrator
to tune the behaviour of the load algorithm:
* vfs.zfs.vdev.mirror.rotating_inc
* vfs.zfs.vdev.mirror.rotating_seek_inc
* vfs.zfs.vdev.mirror.rotating_seek_offset
* vfs.zfs.vdev.mirror.non_rotating_inc
* vfs.zfs.vdev.mirror.non_rotating_seek_inc
These changes where based on work started by the zfsonlinux developers:
https://github.com/zfsonlinux/zfs/pull/1487
Reviewed by: gibbs, mav, will
MFC after: 2 weeks
Sponsored by: Multiplay
zio_compress_data(..) when compressing l2arc buffers.
This eliminates l2arc I/O errors, which resulted in very poor performance on
vdev's configured with block size greater than 512b due to compression
assuming a smaller min block size than the vdev supports.
MFC after: 2 days
When safety requirements are met, it allows to avoid passing I/O requests
to GEOM g_up/g_down thread, executing them directly in the caller context.
That allows to avoid CPU bottlenecks in g_up/g_down threads, plus avoid
several context switches per I/O.
The defined now safety requirements are:
- caller should not hold any locks and should be reenterable;
- callee should not depend on GEOM dual-threaded concurency semantics;
- on the way down, if request is unmapped while callee doesn't support it,
the context should be sleepable;
- kernel thread stack usage should be below 50%.
To keep compatibility with GEOM classes not meeting above requirements
new provider and consumer flags added:
- G_CF_DIRECT_SEND -- consumer code meets caller requirements (request);
- G_CF_DIRECT_RECEIVE -- consumer code meets callee requirements (done);
- G_PF_DIRECT_SEND -- provider code meets caller requirements (done);
- G_PF_DIRECT_RECEIVE -- provider code meets callee requirements (request).
Capable GEOM class can set them, allowing direct dispatch in cases where
it is safe. If any of requirements are not met, request is queued to
g_up or g_down thread same as before.
Such GEOM classes were reviewed and updated to support direct dispatch:
CONCAT, DEV, DISK, GATE, MD, MIRROR, MULTIPATH, NOP, PART, RAID, STRIPE,
VFS, ZERO, ZFS::VDEV, ZFS::ZVOL, all classes based on g_slice KPI (LABEL,
MAP, FLASHMAP, etc).
To declare direct completion capability disk(9) KPI got new flag equivalent
to G_PF_DIRECT_SEND -- DISKFLAG_DIRECT_COMPLETION. da(4) and ada(4) disk
drivers got it set now thanks to earlier CAM locking work.
This change more then twice increases peak block storage performance on
systems with manu CPUs, together with earlier CAM locking changes reaching
more then 1 million IOPS (512 byte raw reads from 16 SATA SSDs on 4 HBAs to
256 user-level threads).
Sponsored by: iXsystems, Inc.
MFC after: 2 months
the register based on the argument index rather than relying on the fields
in struct reg to be in the right order. This assumption is incorrect on
FreeBSD and generally led to bogus argument values for the sixth argument
of PID and USDT probes; the first five are passed directly to dtrace_probe()
via the fasttrap trap handler and so were correctly handled.
MFC after: 2 weeks
null-separated strings to a single string. This can be used to print the
full arguments of a process using execsnoop (from the DTrace toolkit) or
with the following one-liner:
dtrace -n 'syscall::execve:return {trace(curpsinfo->pr_psargs);}'
Note that this relies on the process arguments being cached via the struct
proc, which means that it will not work for argvs longer than
kern.ps_arg_cache_limit. However, the following rather non-portable
script can be used to extract any argv at exec time:
fbt::kern_execve:entry
{
printf("%s", memstr(args[1]->begin_argv, ' ',
args[1]->begin_envv - args[1]->begin_argv));
}
The debug.dtrace.memstr_max sysctl limits the maximum argument size to
memstr(). Thanks to Brendan Gregg for helpful comments on freebsd-dtrace.
Tested by: Fabian Keil (earlier version)
MFC after: 2 weeks
illumos change 14172:be36a38bac3d:
illumos ZFS issues:
4082 zfs receive gets EFBIG from dmu_tx_hold_free()
Please note that this change is slightly different from r255257, because
it is merged out of order with other (larger) upstream changes.
PR: kern/182570
Reported by: Keith White <kwhite@site.uottawa.ca>
Tested by: Keith White <kwhite@site.uottawa.ca>
Approved by: re (glebius)
MFC after: 1 week
X-MFC after: r254753
handlers rather than in the dtrace device open/close methods. The current
approach can cause a panic if the device is closed which the taskqueue
thread is active, or if a kernel module containing a provider is unloaded
while retained enablings are present and the dtrace device isn't opened.
Submitted by: gibbs (original version)
Reviewed by: gibbs
Approved by: re (glebius)
MFC after: 2 weeks
Add support of Illumos dumps on zvol over RAID-Z.
Note that this only adds the features. FreeBSD would
still need more work to support dumping on zvols.
Illumos ZFS issues:
2932 support crash dumps to raidz, etc. pools
MFC after: 1 month
Approved by: re (ZFS blanket)
where 'fdvp' is a directory, 'tvp' is an already existing directory
and they have different mount points.
Reported by: avg, pjd
Reviewed by: pjd
Approved by: re (rodrigc)
Illumos ZFS issues:
3582 zfs_delay() should support a variable resolution
3584 DTrace sdt probes for ZFS txg states
Provide a compatibility shim for Solaris's cv_timedwait_hires
to help aid future porting.
Approved by: re (ZFS blanket)
in the future in a backward compatible (API and ABI) way.
The cap_rights_t represents capability rights. We used to use one bit to
represent one right, but we are running out of spare bits. Currently the new
structure provides place for 114 rights (so 50 more than the previous
cap_rights_t), but it is possible to grow the structure to hold at least 285
rights, although we can make it even larger if 285 rights won't be enough.
The structure definition looks like this:
struct cap_rights {
uint64_t cr_rights[CAP_RIGHTS_VERSION + 2];
};
The initial CAP_RIGHTS_VERSION is 0.
The top two bits in the first element of the cr_rights[] array contain total
number of elements in the array - 2. This means if those two bits are equal to
0, we have 2 array elements.
The top two bits in all remaining array elements should be 0.
The next five bits in all array elements contain array index. Only one bit is
used and bit position in this five-bits range defines array index. This means
there can be at most five array elements in the future.
To define new right the CAPRIGHT() macro must be used. The macro takes two
arguments - an array index and a bit to set, eg.
#define CAP_PDKILL CAPRIGHT(1, 0x0000000000000800ULL)
We still support aliases that combine few rights, but the rights have to belong
to the same array element, eg:
#define CAP_LOOKUP CAPRIGHT(0, 0x0000000000000400ULL)
#define CAP_FCHMOD CAPRIGHT(0, 0x0000000000002000ULL)
#define CAP_FCHMODAT (CAP_FCHMOD | CAP_LOOKUP)
There is new API to manage the new cap_rights_t structure:
cap_rights_t *cap_rights_init(cap_rights_t *rights, ...);
void cap_rights_set(cap_rights_t *rights, ...);
void cap_rights_clear(cap_rights_t *rights, ...);
bool cap_rights_is_set(const cap_rights_t *rights, ...);
bool cap_rights_is_valid(const cap_rights_t *rights);
void cap_rights_merge(cap_rights_t *dst, const cap_rights_t *src);
void cap_rights_remove(cap_rights_t *dst, const cap_rights_t *src);
bool cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little);
Capability rights to the cap_rights_init(), cap_rights_set(),
cap_rights_clear() and cap_rights_is_set() functions are provided by
separating them with commas, eg:
cap_rights_t rights;
cap_rights_init(&rights, CAP_READ, CAP_WRITE, CAP_FSTAT);
There is no need to terminate the list of rights, as those functions are
actually macros that take care of the termination, eg:
#define cap_rights_set(rights, ...) \
__cap_rights_set((rights), __VA_ARGS__, 0ULL)
void __cap_rights_set(cap_rights_t *rights, ...);
Thanks to using one bit as an array index we can assert in those functions that
there are no two rights belonging to different array elements provided
together. For example this is illegal and will be detected, because CAP_LOOKUP
belongs to element 0 and CAP_PDKILL to element 1:
cap_rights_init(&rights, CAP_LOOKUP | CAP_PDKILL);
Providing several rights that belongs to the same array's element this way is
correct, but is not advised. It should only be used for aliases definition.
This commit also breaks compatibility with some existing Capsicum system calls,
but I see no other way to do that. This should be fine as Capsicum is still
experimental and this change is not going to 9.x.
Sponsored by: The FreeBSD Foundation
the passed vnode belongs to the same mount point (v_vfsp or also
known as v_mount in FreeBSD). This check prevents the code from
proceeding further on vnodes that do not belong to ZFS, for
instance, on UFS or NULLFS.
The recent change (merged as r254585) on upstream changes the
check of v_vfsp to instead check the znode's z_zfsvfs. On Illumos
this would work because when the vnode comes from lofs, the
VOP_REALVP() would give the right vnode, this is not true on
FreeBSD where our VOP_REALVP is a no-op, and as such tdvp is
not guaranteed to be a ZFS vnode, and will later trigger a
failed assertion when verifying the vnode.
This changeset modifies our local shims (zfs_freebsd_rename and
zfs_freebsd_link) to check if v_mount matches before proceeding
further.
Reported by: many
Diagnostic work by: avg
kld_unload event handler which gets invoked after a linker file has been
successfully unloaded. The kld_unload and kld_load event handlers are now
invoked with the shared linker lock held, while kld_unload_try is invoked
with the lock exclusively held.
Convert hwpmc(4) to use these event handlers instead of having
kern_kldload() and kern_kldunload() invoke hwpmc(4) hooks whenever files are
loaded or unloaded. This has no functional effect, but simplifes the linker
code somewhat.
Reviewed by: jhb
Don't hold dd_lock for long by breaking it when not doing dsl_dir
accounting. It is not necessary to hold the lock while manipulating
the parent's accounting, because there is no interface for userland
to see a consistent picture of both parent and child at the same
time anyway.
Illumos ZFS issues:
4046 dsl_dataset_t ds_dir->dd_lock is highly contended
Fix a panic from dbuf_free_range() from dmu_free_object() while
doing zfs receive. This is a regression from FreeBSD r253821.
Illumos ZFS issues:
4047 panic from dbuf_free_range() from dmu_free_object() while
doing zfs receive
Illumos DTrace issues:
3089 want ::typedef
3094 libctf should support removing a dynamic type
3095 libctf does not validate arrays correctly
3096 libctf does not validate function types correctly
The flag was mandatory since r209792, where vm_page_grab(9) was
changed to only support the alloc retry semantic.
Suggested and reviewed by: alc
Sponsored by: The FreeBSD Foundation
and CIFS file attributes as BSD stat(2) flags.
This work is intended to be compatible with ZFS, the Solaris CIFS
server's interaction with ZFS, somewhat compatible with MacOS X,
and of course compatible with Windows.
The Windows attributes that are implemented were chosen based on
the attributes that ZFS already supports.
The summary of the flags is as follows:
UF_SYSTEM: Command line name: "system" or "usystem"
ZFS name: XAT_SYSTEM, ZFS_SYSTEM
Windows: FILE_ATTRIBUTE_SYSTEM
This flag means that the file is used by the
operating system. FreeBSD does not enforce any
special handling when this flag is set.
UF_SPARSE: Command line name: "sparse" or "usparse"
ZFS name: XAT_SPARSE, ZFS_SPARSE
Windows: FILE_ATTRIBUTE_SPARSE_FILE
This flag means that the file is sparse. Although
ZFS may modify this in some situations, there is
not generally any special handling for this flag.
UF_OFFLINE: Command line name: "offline" or "uoffline"
ZFS name: XAT_OFFLINE, ZFS_OFFLINE
Windows: FILE_ATTRIBUTE_OFFLINE
This flag means that the file has been moved to
offline storage. FreeBSD does not have any special
handling for this flag.
UF_REPARSE: Command line name: "reparse" or "ureparse"
ZFS name: XAT_REPARSE, ZFS_REPARSE
Windows: FILE_ATTRIBUTE_REPARSE_POINT
This flag means that the file is a Windows reparse
point. ZFS has special handling code for reparse
points, but we don't currently have the other
supporting infrastructure for them.
UF_HIDDEN: Command line name: "hidden" or "uhidden"
ZFS name: XAT_HIDDEN, ZFS_HIDDEN
Windows: FILE_ATTRIBUTE_HIDDEN
This flag means that the file may be excluded from
a directory listing if the application honors it.
FreeBSD has no special handling for this flag.
The name and bit definition for UF_HIDDEN are
identical to the definition in MacOS X.
UF_READONLY: Command line name: "urdonly", "rdonly", "readonly"
ZFS name: XAT_READONLY, ZFS_READONLY
Windows: FILE_ATTRIBUTE_READONLY
This flag means that the file may not written or
appended, but its attributes may be changed.
ZFS currently enforces this flag, but Illumos
developers have discussed disabling enforcement.
The behavior of this flag is different than MacOS X.
MacOS X uses UF_IMMUTABLE to represent the DOS
readonly permission, but that flag has a stronger
meaning than the semantics of DOS readonly permissions.
UF_ARCHIVE: Command line name: "uarch", "uarchive"
ZFS_NAME: XAT_ARCHIVE, ZFS_ARCHIVE
Windows name: FILE_ATTRIBUTE_ARCHIVE
The UF_ARCHIVED flag means that the file has changed and
needs to be archived. The meaning is same as
the Windows FILE_ATTRIBUTE_ARCHIVE attribute, and
the ZFS XAT_ARCHIVE and ZFS_ARCHIVE attribute.
msdosfs and ZFS have special handling for this flag.
i.e. they will set it when the file changes.
sys/param.h: Bump __FreeBSD_version to 1000047 for the
addition of new stat(2) flags.
chflags.1: Document the new command line flag names
(e.g. "system", "hidden") available to the
user.
ls.1: Reference chflags(1) for a list of file flags
and their meanings.
strtofflags.c: Implement the mapping between the new
command line flag names and new stat(2)
flags.
chflags.2: Document all of the new stat(2) flags, and
explain the intended behavior in a little
more detail. Explain how they map to
Windows file attributes.
Different filesystems behave differently
with respect to flags, so warn the
application developer to take care when
using them.
zfs_vnops.c: Add support for getting and setting the
UF_ARCHIVE, UF_READONLY, UF_SYSTEM, UF_HIDDEN,
UF_REPARSE, UF_OFFLINE, and UF_SPARSE flags.
All of these flags are implemented using
attributes that ZFS already supports, so
the on-disk format has not changed.
ZFS currently doesn't allow setting the
UF_REPARSE flag, and we don't really have
the other infrastructure to support reparse
points.
msdosfs_denode.c,
msdosfs_vnops.c: Add support for getting and setting
UF_HIDDEN, UF_SYSTEM and UF_READONLY
in MSDOSFS.
It supported SF_ARCHIVED, but this has been
changed to be UF_ARCHIVE, which has the same
semantics as the DOS archive attribute instead
of inverse semantics like SF_ARCHIVED.
After discussion with Bruce Evans, change
several things in the msdosfs behavior:
Use UF_READONLY to indicate whether a file
is writeable instead of file permissions, but
don't actually enforce it.
Refuse to change attributes on the root
directory, because it is special in FAT
filesystems, but allow most other attribute
changes on directories.
Don't set the archive attribute on a directory
when its modification time is updated.
Windows and DOS don't set the archive attribute
in that scenario, so we are now bug-for-bug
compatible.
smbfs_node.c,
smbfs_vnops.c: Add support for UF_HIDDEN, UF_SYSTEM,
UF_READONLY and UF_ARCHIVE in SMBFS.
This is similar to changes that Apple has
made in their version of SMBFS (as of
smb-583.8, posted on opensource.apple.com),
but not quite the same.
We map SMB_FA_READONLY to UF_READONLY,
because UF_READONLY is intended to match
the semantics of the DOS readonly flag.
The MacOS X code maps both UF_IMMUTABLE
and SF_IMMUTABLE to SMB_FA_READONLY, but
the immutable flags have stronger meaning
than the DOS readonly bit.
stat.h: Add definitions for UF_SYSTEM, UF_SPARSE,
UF_OFFLINE, UF_REPARSE, UF_ARCHIVE, UF_READONLY
and UF_HIDDEN.
The definition of UF_HIDDEN is the same as
the MacOS X definition.
Add commented-out definitions of
UF_COMPRESSED and UF_TRACKED. They are
defined in MacOS X (as of 10.8.2), but we
do not implement them (yet).
ufs_vnops.c: Add support for getting and setting
UF_ARCHIVE, UF_HIDDEN, UF_OFFLINE, UF_READONLY,
UF_REPARSE, UF_SPARSE, and UF_SYSTEM in UFS.
Alphabetize the flags that are supported.
These new flags are only stored, UFS does
not take any action if the flag is set.
Sponsored by: Spectra Logic
Reviewed by: bde (earlier version)
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/sys/zio_compress.h:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zio_compress.c:
Add module lifetime functions to allocate and teardown
state data.
Report:
- Compression attempts.
- Buffers found to be empty.
- Compression calls that are skipped because
the data length is already less than or
equal to the minimum block length.
- Compression attempts that fail to yield a 12.5%
compression ratio.
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dmu.c:
Add calls to the zio_compress.c module's init and fini
functions.
Sponosred by: Spectra Logic Corporation
MFC after: 2 weeks
minimum allocation size for devices. Use this information to
automatically increase ZFS's minimum allocation size for new top-level
vdevs to a value that more closely matches the optimum device
allocation size.
Use GEOM's stripesize attribute, if set, as the physical sector
size of the GEOM.
Calculate the minimum blocksize of each metaslab class. Use the
calculated value instead of SPA_MINBLOCKSIZE (512b) when determining
the likelyhood of compression yeilding a reduction in physical space
usage.
Report devices with sub-optimal block size configuration in "zpool
status". Also properly fail attempts to attach devices with a
logical block size greater than 8kB, since this will cause corruption
to ZFS's label area.
Sponsored by: Spectra Logic Corporaion
MFC after: 2 weeks
Background
==========
Many modern devices use physical allocation units that are much
larger than the minimum logical allocation size accessible by
external commands. Two prevalent examples of this are 512e disk
drives (512b logical sector, 4K physical sector) and flash devices
(512b logical sector, 4K or larger allocation block size, and 128k
or larger erase block size). Operations that modify less than the
physical sector size result in a costly read-modify-write or garbage
collection sequence on these devices.
Simply exporting the true physical sector of the device to ZFS would
yield optimal performance, but has two serious drawbacks:
1) Existing pools created with devices that have different logical
and physical block sizes, but were configured to use the logical
block size (e.g. because the OS version used for pool construction
reported the logical block size instead of the physical block
size) will suddenly find that the vdev allocation size has
increased. This can be easily tolerated for active members of
the array, but ZFS would prevent replacement of a vdev with
another identical device because it now appears that the smaller
allocation size required by the pool is not supported by the new
device.
2) The device's physical block size may be too large to be supported
by ZFS. The optimal allocation size for the vdev may be quite
large. For example, a RAID controller may export a vdev that
requires read-modify-write cycles unless accessed using 64k
aligned/sized requests. ZFS currently has an 8k minimum block
size limit.
Reporting both the logical and physical allocation sizes for vdevs
solves these problems. A device may be used so long as the logical
block size is compatible with the configuration. By comparing the
logical and physical block sizes, new configurations can be optimized
and administrators can be notified of any existing pools that are
sub-optimal.
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/sys/spa.h:
Add the SPA_ASHIFT constant. ZFS currently has a hard upper
limit of 13 (8k) for ashift and this constant is used to
both document and enforce this limit.
sys/cddl/contrib/opensolaris/uts/common/sys/fs/zfs.h:
Add the VDEV_AUX_ASHIFT_TOO_BIG error code.
Add fields for exporting the configured, logical, and
physical ashift to the vdev_stat_t structure.
Add VDEV_STAT_VALID() macro which can be used to verify the
presence of required vdev_stat_t fields in nvlist data.
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/vdev.c:
Provide a SYSCTL_PROC handler for "max_auto_ashift". Since
the limit is only referenced long after boot when a create
operation occurs, there's no compelling need for it to be
a boot time configurable tunable. This also allows the
validation code for the max_auto_ashift value to be contained
within the sysctl handler.
Populate the new fields in the vdev_stat_t structure.
Fail vdev opens if the vdev reports an ashift larger than
SPA_MAXASHIFT.
Propogate vdev_logical_ashift and vdev_physical_ashift between
child and parent vdevs as is done for vdev_ashift.
In vdev_open(), restore code that fails opens for devices
where vdev_ashift grows. This can only happen now if the
device's logical ashift grows, which means it really isn't
safe to use the device.
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/sys/vdev_impl.h:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/vdev.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/vdev_file.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/vdev_geom.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/vdev_mirror.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/vdev_missing.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/vdev_raidz.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/vdev_root.c:
Update the vdev_open() API so that both logical (what was
just ashift before) and physical ashift are reported.
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/sys/vdev_impl.h:
Add two new fields, vdev_physical_ashift and vdev_logical_ashift,
to vdev_t.
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/vdev.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/spa_config.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/spa.c:
Add vdev_ashift_optimize(). Call it anytime a new top-level
vdev is allocated.
cddl/contrib/opensolaris/cmd/zpool/zpool_main.c:
Add text for the VDEV_AUX_ASHIFT_TOO_BIG error.
For each sub-optimally configured leaf vdev, report configured
and native block sizes.
cddl/contrib/opensolaris/cmd/zpool/zpool_main.c:
cddl/contrib/opensolaris/lib/libzfs/common/libzfs.h:
cddl/contrib/opensolaris/lib/libzfs/common/libzfs_status.c:
Introduce a new zpool status: ZPOOL_STATUS_NON_NATIVE_ASHIFT.
This status is reported on healthy pools containing vdevs
configured to use a block size smaller than their reported
physical block size.
cddl/contrib/opensolaris/lib/libzfs/common/libzfs_status.c:
Update find_vdev_problem() and supporting functions to
provide the full vdev_stat_t structure to problem checking
routines, and to allow decent into replacing vdevs.
Add a vdev_non_native_ashift() validator which is used on
the full vdev tree to check for ZPOOL_STATUS_NON_NATIVE_ASHIFT.
cddl/contrib/opensolaris/lib/libzpool/common/kernel.c:
cddl/contrib/opensolaris/lib/libzpool/common/sys/zfs_context.h:
Enhance sysctl userland stubs now that a SYSCTL_PROC handler
is used in vdev.c.
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/metaslab.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/sys/metaslab_impl.h:
When the group membership of a metaslab class changes (i.e.
when a vdev is added or removed from a pool), walk the group
list to determine the smallest block size currently available
and record this in the metaslab class.
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/sys/metaslab.h:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/metaslab.c:
Add the metaslab_class_get_minblocksize() accessor.
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/sys/zio_compress.h:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zio_compress.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c:
In zio_compress_data(), take the minimum blocksize as an
input parameter instead of assuming SPA_MINBLOCKSIZE.
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c:
In l2arc_compress_buf(), pass SPA_MINBLOCKSIZE as the minimum
blocksize of the device. The l2arc code performs has it's own
code for deciding if compression is worth while, so this
effectively disables zio_compress_data() from second guessing
the original decision.
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zio.c:
In zio_write_bp_init(), use the minimum blocksize of the
normal metaslab class when compressing data.
probes declared in a kernel module when that module is unloaded. In
particular,
* Unloading a module with active SDT probes will cause a panic. [1]
* A module's (FBT/SDT) probes aren't destroyed when the module is unloaded;
trying to use them after the fact will generally cause a panic.
This change fixes both problems by porting the DTrace module load/unload
handlers from illumos and registering them with the corresponding
EVENTHANDLER(9) handlers. This allows the DTrace framework to destroy all
probes defined in a module when that module is unloaded, and to prevent a
module unload from proceeding if some of its probes are active. The latter
problem has already been fixed for FBT probes by checking lf->nenabled in
kern_kldunload(), but moving the check into the DTrace framework generalizes
it to all kernel providers and also fixes a race in the current
implementation (since a probe may be activated between the check and the
call to linker_file_unload()).
Additionally, the SDT implementation has been reworked to define SDT
providers/probes/argtypes in linker sets rather than using SYSINIT/SYSUNINIT
to create and destroy SDT probes when a module is loaded or unloaded. This
simplifies things quite a bit since it means that pretty much all of the SDT
code can live in sdt.ko, and since it becomes easier to integrate SDT with
the DTrace framework. Furthermore, this allows FreeBSD to be quite flexible
in that SDT providers spanning multiple modules can be created on the fly
when a module is loaded; at the moment it looks like illumos' SDT
implementation requires all SDT probes to be statically defined in a single
kernel table.
PR: 166927, 166926, 166928
Reported by: davide [1]
Reviewed by: avg, trociny (earlier version)
MFC after: 1 month
In the future, we'll need to come up with new proc_*() functions that accept
locked processes. For now, this prevents postgresql + DTrace from crashing the
system.
MFC after: 1 month
Unify the 2 concept into a real, minimal, sxlock where the shared
acquisition represent the soft busy and the exclusive acquisition
represent the hard busy.
The old VPO_WANTED mechanism becames the hard-path for this new lock
and it becomes per-page rather than per-object.
The vm_object lock becames an interlock for this functionality:
it can be held in both read or write mode.
However, if the vm_object lock is held in read mode while acquiring
or releasing the busy state, the thread owner cannot make any
assumption on the busy state unless it is also busying it.
Also:
- Add a new flag to directly shared busy pages while vm_page_alloc
and vm_page_grab are being executed. This will be very helpful
once these functions happen under a read object lock.
- Move the swapping sleep into its own per-object flag
The KPI is heavilly changed this is why the version is bumped.
It is very likely that some VM ports users will need to change
their own code.
Sponsored by: EMC / Isilon storage division
Discussed with: alc
Reviewed by: jeff, kib
Tested by: gavin, bapt (older version)
Tested by: pho, scottl
Illumos ZFS issues:
3957 ztest should update the cachefile before killing itself
3958 multiple scans can lead to partial resilvering
3959 ddt entries are not always resilvered
3960 dsl_scan can skip over dedup-ed blocks if
physical birth != logical birth
3961 freed gang blocks are not resilvered and can cause pool to suspend
3962 ztest should print out zfs debug buffer before exiting
Fix a regression introduced by fix for Illumos bug #3834. Quote from
Matthew Ahrens on the Illumos issue:
ztest fails this assertion because ztest_dmu_read_write() does
dmu_tx_hold_free(tx, bigobj, bigoff, bigsize);
and then
dmu_object_set_checksum(os, bigobj,
(enum zio_checksum)ztest_random_dsl_prop(ZFS_PROP_CHECKSUM), tx);
If the region to free is past the end of the file, the DMU assumes that there
will be nothing to do for this object. However, ztest does set_checksum(),
which must modify the dnode. The fix is for ztest to also call
dmu_tx_hold_bonus(tx, bigobj);
so we can account for the dirty data associated with setting the checksum
Illumos ZFS issues:
3955 ztest failure: assertion refcount_count(&tx->tx_space_written)
+ delta <= tx->tx_space_towrite
Merge vendor bugfix for ZFS test suite that triggers false positives.
Illumos ZFS issues:
3949 ztest fault injection should avoid resilvering devices
3950 ztest: deadman fires when we're doing a scan
3951 ztest hang when running dedup test
3952 ztest: ztest_reguid test and ztest_fault_inject don't place nice together
Before executing any subcommand, zpool tool fetches pools configuration from
the kernel. Before features support was added, kernel was regenerating that
configuration based on data always present in memory. Unfortunately, pool
features list and activity counters are not such. They are stored in ZAP,
that normally resides in ARC, but under heavy memory pressure may be swapped
out. If pool is suspended at this point, there is no way to recover it back
since any zpool command will stuck.
This change has one predictable flaw: `zpool upgrade` always wish to upgrade
suspended pools, but fortunately it can't do it due to the suspension.
disable TRIM otherwise.
r252840 (illumos bug 3836) is based on assumption that zio_free_sync() has
no lock dependencies and should complete immediately. Unfortunately, with our
TRIM implementation that is not true due to ZIO_STAGE_VDEV_IO_START added
to the ZIO_FREE_PIPELINE, which, while not really accessing devices, still
acquires SCL_ZIO lock for read to be sure devices won't disappear.
When TRIM is disabled, this patch enables direct free execution from r252840
and removes ZIO_STAGE_VDEV_IO_START and ZIO_STAGE_VDEV_IO_ASSESS stages from
the pipeline to avoid lock acquisition. Otherwise it queues free request as
it was before r252840.
Existing async thread is running only on successfull spa_sync() completion,
that is impossible in case of pool loosing required (last) disk(s). That
indefinite delay of SPA_ASYNC_REMOVE processing made ZFS to not close the
lost disks, preventing GEOM/CAM from destroying devices and reusing names
on later disk reattach.
In earlier version of the patch I've tried to just run existing thread
immediately, unrelated to spa_sync() completion, but that exposed number
of situations where it could stuck due to locks held by stuck spa_sync(),
that are required for other kinds of async events.
Experiments with OpenIndiana snapshot confirmed that they also have this
issue with lost disks reattach.
We cannot busy a page before doing pagefaults.
Infact, it can deadlock against vnode lock, as it tries to vget().
Other functions, right now, have an opposite lock ordering, like
vm_object_sync(), which acquires the vnode lock first and then
sleeps on the busy mechanism.
Before this patch is reinserted we need to break this ordering.
Sponsored by: EMC / Isilon storage division
Reported by: kib
- It does not let pages respect the LRU policy
- It bloats the active/inactive queues of few pages
Try to avoid it as much as possible with the long-term target to
completely remove it.
Use the soft-busy mechanism to protect page content accesses during
short-term operations (like uiomove_fromphys()).
After this change only vm_fault_quick_hold_pages() is still using the
hold mechanism for page content access.
There is an additional complexity there as the quick path cannot
immediately access the page object to busy the page and the slow path
cannot however busy more than one page a time (to avoid deadlocks).
Fixing such primitive can bring to complete removal of the page hold
mechanism.
Sponsored by: EMC / Isilon storage division
Discussed with: alc
Reviewed by: jeff
Tested by: pho
Skip eviction step of processing free records when doing ZFS
receive to avoid the expensive search operation of non-existent
dbufs in dn_dbufs.
Illumos ZFS issues:
3834 incremental replication of 'holey' file systems is slow
MFC after: 2 weeks
To quote Illumos issue #3888:
When 'zfs recv -F' is used with an incremental recv it rolls
back any changes made since the last snapshot in case new
changes were made to the file system while the recv is in
progress (without -F the recv would fail when it does it's
final check to commit the recv-ed data as the recv-ed data
conflicts with the newly written data).
However, if there is a snapshot taken after the recv began
rolling back to the 'latest' snapshot will not help and the
recv will still fail. 'zfs recv -F' should be extended to
destroy any snapshots created since the source snapshot when
finishing the recv (effectively rolling back through all
snapshots, instead of just to the latest snapshot).
Illumos ZFS issues:
3888 zfs recv -F should destroy any snapshots created since the
incremental source
MFC after: 2 weeks
To quote Illumos #3875:
The problem here is that if we ever end up in the error
path, we drop the locks protecting access to the zfsvfs_t
prior to forcibly unmounting the filesystem. Because z_os
is NULL, any thread that had already picked up the zfsvfs_t
and was sitting in ZFS_ENTER() when we dropped our locks
in zfs_resume_fs() will now acquire the lock, attempt to
use z_os, and panic.
Illumos ZFS issues:
3875 panic in zfs_root() after failed rollback
MFC after: 2 weeks
existed before IOCTL code refactoring merged change 4445fffb from illumos
at r248571.
This change allows `zpool clear` to be used again to recover suspended pool.
It seems the only was supposed by the code to restore pool operation after
reconnecting lost disks that were required for data completeness. There
are still cases where `zpool clear` command can just safely stuck due to
deadlocks inside ZFS kernel part, but probably that is better then having
no chances to recover at all.
- move init and fini code into separate functions (like it is done upstream)
- invoke fini code via shutdown_post_sync event hook
This should make zfs close its underlying devices during shutdown,
which may be important for their drivers.
MFC after: 20 days
All other places where a znode is allocated do not need z_vnode at all.
These are:
- zfs_create_share_dir
- zfs_create_fs
This chnage ensures two things:
- VN_LOCK_ASHARE is not erroneously called for VFIFO vnodes
- vn_lock is called on a fully constructed vnode with correct v_ops
The change also allows to make zfs_znode_cache_constructor a normal
kmem_cache constructor again (as it is in upstream).
This allows to avoid a problem where zfs_znode_cache_destructor
may be called on un-constructed znodes.
MFC after: 17 days
Unconditionally freeing a page is not good, especially if it is the page
that was wired by the caller. The checks are picked up from
kern_sendfile.
MFC after: 3 weeks
Quoting illumos issue #3836:
Currently zio_free() always puts the zio on a list for subsequent
processing by zio_free_sync(). This is only necessary for frees that
might need to issue reads (gang and dedup blocks).
By processing the majority of the frees as we encounter them, we reduce
the amount of time that the spa_sync() thread spends burning CPU and
not doing any i/o, thus increasing the overall write throughput of the
system.
Illumos ZFS issues:
3836 zio_free() can be processed immediately in the common case
MFC after: 1 week
I missed to register zfs_ioc_jail and zfs_ioc_unjail as legacy ioctl's
with the new zfs_ioctl_register_legacy() function.
These operations do not modify pools or datasets so there is no need to
log them to pool history.
Reported by: Alexander Leidinger <ale@FreeBSD.org> and others on current@
MFC after: 3 days
This could happen if a thread doing a page-in loses a ZFS range lock
race to a thread writing to the same range
This fixes "panic: vm_page_alloc: pindex already allocated" in
http://docs.FreeBSD.org/cgi/mid.cgi?1372165971.96049.42.camel
Submitted by: avg
MFC after: 1 week
Restore a previous behavior before r251646, where when destructing
ZFS snapshot, the ioctl would return ENOENT when it hit any of
them in the errlist (the new behavior was only return ENOENT when
all returns error).
Illumos ZFS issues:
3829 fix for 3740 changed behavior of zfs destroy/hold/release ioctl
MFC after: 1 week
ZFS event processing should work on R/O root filesystems
Illumos ZFS issues:
3749 zfs event processing should work on R/O root filesystems
MFC after: 2 weeks
* Illumos ZFS issue #3805 arc shouldn't cache freed blocks
Quote from the Illumos issue:
ZFS should proactively evict freed blocks from the cache.
Even though these freed blocks will never be used again, and thus
will eventually be evicted, this causes us to use memory
inefficiently for 2 reasons:
1. A block that is freed has no chance of being accessed again, but
will be kept in memory preferentially to a block that was accessed
before it (and is thus older) but has not been freed and thus has
at least some chance of being accessed again.
2. We partition the ARC into several buckets:
user data that has been accessed only once (MRU)
metadata that has been accessed only once (MRU)
user data that has been accessed more than once (MFU)
metadata that has been accessed more than once (MFU)
The user data vs metadata split is somewhat arbitrary, and the
primary control on how much memory is used to cache data vs metadata
is to simply try to keep the proportion the same as it has been in the
past (each bucket "evicts against" itself). The secondary control is
to evict data before evicting metadata.
Because of this bucketing, we may end up with one bucket mostly
containing freed blocks that are very old, while another bucket has
more recently accessed, still-allocated blocks. Data in the useful
bucket (with still-allocated blocks) may be evicted in preference to
data in the useless bucket (with old, freed blocks).
On dcenter, we saw that the MFU metadata bucket was 230MB, while the
MFU data bucket was 27GB and the MRU metadata bucket was 256GB.
However, the vast majority of data in the MRU metadata bucket (256GB)
was freed blocks, and thus useless. Meanwhile, the MFU metadata bucket
(230MB) was constantly evicting useful blocks that will be soon needed.
The problem of cache segmentation is a larger problem that needs more
investigation. However, if we stop caching freed blocks, it should
reduce the impact of this more fundamental issue.
MFC after: 2 weeks
* Illumos zfs issue #3137 L2ARC compression
Whether or not to compress buffers entering the L2ARC is
controlled by "compression" setting on the dataset, when
compression is not "off", L2ARC compression is enabled.
The compress method is always LZ4 for L2ARC when enabled
because it works best for the scenario.
MFC after: 2 weeks
USDT probes exits. This was previously done with a callout; however, it is
possible to sleep while holding the DTrace mutexes, so a panic will occur
on INVARIANTS kernels if the callout handler can't immediately acquire one
of these mutexes. This panic will be frequently triggered on systems where
a USDT-enabled program (perl, for instance) is often run.
This revision changes the fasttrap cleanup mechanism so that a dedicated
thread is used instead of a callout. The old behaviour is otherwise
preserved.
Reviewed by: rpaulo
MFC after: 1 month
users to guarantee that the output of DTrace scripts will be time-ordered.
This option is enabled by adding the line
#pragma D option temporal
to the beginning of a script, or by adding '-x temporal' to the arguments of
dtrace(1).
This change fixes a bug in the original port of the temporal option. This
bug was causing some assertions to fail, so they had been disabled; in this
revision the assertions are working properly and are enabled.
The DTrace version number has been bumped from 1.9.0 to 1.9.1 to reflect
the language change that's being introduced.
This change corresponds to part of illumos-gate commit e5803b76927480:
3021 option for time-ordered output from dtrace(1M)
Reviewed by: pfg
Obtained from: illumos
MFC after: 1 month
Merge vendor bugfix for a possible deadlock related to async destroy
and improve write performance by introducing a new lock protecting
tx_open_txg.
Illumos ZFS issues:
3642 dsl_scan_active() should not issue I/O to determine if async
destroying is active
3643 txg_delay should not hold the tc_lock
MFC after: 1 week
impossible to set quota and reservation on pools lower than version 22.
Problem has been reported and a solution discussed with vendor.
Illumos ZFS issues:
3739 cannot set zfs quota or reservation on pool version < 22
Reviewed by: Matthew Ahrens <mahrens@delphix.com>
Reported by: Steve Wills <swills@FreeBSD.org>
MFC after: 3 days
The following change from illumos brought caused DTrace to
pause in an interactive environment:
3026 libdtrace should set LD_NOLAZYLOAD=1 to help the pid provider
This was not detected during testing because it doesn't
affect scripts.
We shouldn't be changing the environment, especially since the
LD_NOLAZYLOAD option doesn't apply to our (GNU) ld.
Unfortunately the change from upstream was made in such a way
that it is very difficult to separate this change from the
others so, at least for now, it's better to just revert
everything.
Reference:
https://www.illumos.org/issues/3026
Reported by: Navdeep Parhar and Mark Johnston
Merge changes from illumos:
3021 option for time-ordered output from dtrace(1M)
3022 DTrace: keys should not affect the sort order when sorting by value
3023 it should be possible to dereference dynamic variables
3024 D integer narrowing needs some work
3025 register leak in D code generation
3026 libdtrace should set LD_NOLAZYLOAD=1 to help the pid provider
This brings yet another feature implemented in upstream DTrace.
A complete description is available here:
http://dtrace.org/blogs/ahl/2012/07/28/my-new-dtrace-favorite/
This change bumps the DT_VERS_* number to 1.9.1 in
accordance to what is done in illumos.
This change was somewhat complicated because upstream is mixed many
changes in an individual commit and some of the tests don't really
apply to us.
There are also appear to be differences in timestamping with Solaris
so we had to workaround some assertions making sure no regression
happened.
Special thanks to Fabian Keil for changes and testing.
Illumos Revisions: 13758:23432da34147
Reference:
https://www.illumos.org/issues/3021https://www.illumos.org/issues/3022https://www.illumos.org/issues/3023https://www.illumos.org/issues/3024https://www.illumos.org/issues/3025https://www.illumos.org/issues/1694
Tested by: Fabian Keil
Obtained from: Illumos
MFC after: 1 months
Merge bugfixes accepted and integrated by vendor. Underlying problems
have been reported by us and fixed in r240942 and r249196.
Illumos ZFS issues:
3645 dmu_send_impl: possibilty of pool hold leak
3692 Panic on zfs receive of a recursive deduplicated stream
MFC after: 8 days
doesn't copyout in this case.
To solve this issue a new struct zfs_iocparm_t is introduced consisting of:
- zfs_ioctl_version (future backwards compatibility purposes)
- user space pointer to zfs_cmd_t (copyin and copyout)
- size of zfs_cmd_t (verification purposes)
The copyin and copyout of zfs_cmd_t is now done the illumos (vendor) way
what makes porting of new changes easier and ensures correct behavior if
returning an error.
MFC after: 10 days
Do not list read-only pools in zpool.cache
Reduce diff against vendor in unused vdev_disk.c
Illumos ZFS issues:
3639 zpool.cache should skip over readonly pools
3640 want automatic devid updates
MFC after: 1 week
Merge change from vendor to reduce diff only.
ZFS dtrace probes are not supported on FreeBSD yet.
Illumos ZFS issues:
3598 want to dtrace when errors are generated in zfs
MFC after: 3 weeks
Import vendor change to reduce diff, no effect on FreeBSD.
Illumos ZFS issues:
3517 importing pool with autoreplace=on and "hole" vdevs crashes syseventd
Prior to r248571 spa_open was always called with a bare pool name,
but now it is called with a dataset name instead (spa_lookup handles
that).
So, when a ZFS root is mounted spa_open is called with a name of a root
dataset, which can very well be different from the pool name.
But zvol_create_minors should be called with the pool name, because it
performs a recursive traversal of all datasets under the name to find
all those that are volumes.
MFC after: 7 days
