Commit Graph

1273 Commits

Author SHA1 Message Date
Robert Watson
9dcce6e267 Now that DTrace stack traces handle exception frames better, skip fewer
stack frames for FBT 'entry' probes on ARM.

MFC after:	3 days
Sponsored by:	DARPA, AFRL
2015-03-15 15:19:02 +00:00
Robert Watson
2a1d3dee4f On ARM, unlike some other architectures, saved $pc values from in-kernel
traps do appear in the regular call stack, rather than only in a special
trap frame, so we don't need to inject the trap-frame $pc into a returned
stack trace in DTrace.

MFC after:	3 days
Sponsored by:	DARPA, AFRL
2015-03-15 15:17:34 +00:00
Robert Watson
808d83b01a Replace the completely arbitrary '3' with '9' for the number of frames to
skip using the DTrace 'profile' provider on ARM.  This causes stack traces
to skip various driver-and callout-related things as they do on x86, where
the likewise arbitrary values are '6' (32-bit) and '10' (64-bit) for
similar sorts of reasons.

MFC after:	3 days
Sponsored by:	DARPA, AFRL
2015-03-15 14:12:40 +00:00
Steven Hartland
208264283d Allow zvol_geom_worker to process BIO_DELETE's
If zvol_geom_start is called with a BIO_DELETE from a thread which can
sleep it queues it for later processing by the zvol_geom_worker. The
zvol_geom_worker didn't have a delete case so would simply loose the bio
hence preventing the original caller from every completing. In addition
an other unknown types would suffer the same fate.

Allow zvol_geom_worker to process BIO_DELETE's via zvol_strategy and
return unsupported for all unknown bio types.

MFC after:	2 weeks
Sponsored by:	Multiplay
2015-03-14 17:35:04 +00:00
Alexander Motin
0d45c37cb6 Make DIOCGATTR in device mode handle "GEOM::candelete".
MFC after:	3 days
2015-03-12 16:19:18 +00:00
George V. Neville-Neil
a340dc5348 Add support for walltimestamp to DTrace on ARM. 2015-03-07 04:38:25 +00:00
Andrew Turner
be9bc81174 dtrace_cas32 and dtrace_casptr should retrn the data loaded from target
not the new value.

Sponsored by:	ABT Systems Ltd
2015-03-05 18:03:42 +00:00
Andrew Turner
4a8169d97b Add the MD parts of dtrace needed to use fbt on ARM. For this we need to
emulate the instructions used in function entry and exit.

For function entry ARM will use a push instruction to push up to 16
registers to the stack. While we don't expect all 16 to be used we need to
handle any combination the compiler may generate, even if it doesn't make
sense (e.g. pushing the program counter).

On function return we will either have a pop or branch instruction. The
former is similar to the push instruction, but with care to make sure we
update the stack pointer and program counter correctly in the cases they
are either in the list of registers or not. For branch we need to take the
24-bit offset, sign-extend it, and add that number of 4-byte words to the
program counter. Care needs to be taken as, due to historical reasons, the
address the branch is relative to is not the current instruction, but 8
bytes later.

This allows us to use the following probes on ARM boards:
  dtrace -n 'fbt::malloc:entry { stack() }'
and
  dtrace -n 'fbt:🆓return { stack() }'

Differential Revision:	https://reviews.freebsd.org/D2007
Reviewed by:	gnn, rpaulo
Sponsored by:	ABT Systems Ltd
2015-03-05 17:55:31 +00:00
Nathan Whitehorn
4c790d26d1 Fix build after unifying DAR/DEAR storage in trap frame. 2015-03-05 17:02:22 +00:00
Robert Watson
ca162c0888 Don't all DTrace's FBT on ARM to instrument undefinedinstruction(), as
this would lead to DTrace reentrance.

Sponsored by:	DARPA, AFRL
2015-03-05 07:40:41 +00:00
Andrew Turner
2b6af94bc8 Fix the dtrace ARM atomic compare-and-set functions. These functions are
expected to return the data in the memory location pointed at by target
after the operation. The FreeBSD atomic functions previously used return
either 0 or 1 to indicate if the comparison succeeded or not respectively.

With this change these functions only support ARMv6 and later are supported
by these functions.

Sponsored by:	ABT Systems Ltd
2015-03-01 10:04:14 +00:00
Ryan Stone
296a8144aa Allow Illumos code to co-exist with nv(9)
Differential Revision:		https://reviews.freebsd.org/D1881
Reviewed by:			jfv, will
Suggested by:			pjd
MFC after:			1 month
Sponsored by:			Sandvine Inc
2015-03-01 00:22:45 +00:00
Andrew Turner
aeca5b8bc9 Use the ARM unwinder with dtrace to extract the stack when asked. With this
dtrace is able to display a stack trace similar to the one below.

# dtrace -p 603 -n 'tcp:kernel::receive { stack(); }'
  0     70                         :receive
              kernel`ip_input+0x140
              kernel`netisr_dispatch_src+0xb8
              kernel`ether_demux+0x1c4
              kernel`ether_nh_input+0x3a8
              kernel`netisr_dispatch_src+0xb8
              kernel`ether_input+0x60
              kernel`cpsw_intr_rx+0xac
              kernel`intr_event_execute_handlers+0x128
              kernel`ithread_loop+0xb4
              kernel`fork_exit+0x84
              kernel`swi_exit
              kernel`swi_exit

Tested by:	gnn
Sponsored by:	ABT Systems Ltd
2015-02-19 12:20:21 +00:00
George V. Neville-Neil
a57dc98e3a Clean up machine dependent code for DTrace on ARM.
Submitted by:	markj
2015-02-11 17:27:37 +00:00
George V. Neville-Neil
fcb5606706 Initial version of DTrace on ARM32.
Submitted by:	Howard Su based on work by Oleksandr Tymoshenko
Reviewed by:	ian, andrew, rpaulo, markj
2015-02-10 19:41:30 +00:00
Mark Johnston
3277b9a257 Fix a typo in r278137: make sure to free provider state.
X-MFC-With:     r278136
2015-02-08 03:55:12 +00:00
Pedro F. Giffuni
3ccccdc17d MFV r266995:
4767 dtrace_probe() always has the timestamp

Reference:
https://illumos.org/issues/4767

Obtained from:	Illumos
MFC after:	2 weeks
2015-02-03 20:06:30 +00:00
Pedro F. Giffuni
eadcd0fadf MFV r266993:
4469 DTrace helper tracing should be dynamic

Reference:
https://illumos.org/issues/4469

Obtained from:	Illumos
Phabric:	D1551
Reviewed by:	markj
MFC after:	2 weeks
2015-02-03 19:39:53 +00:00
Mark Johnston
c36bd253fa Continue to handle the case where state is NULL, though this currently
cannot happen on FreeBSD. r278136 overlooked the fact that a destructor
registered with devfs_set_cdevpriv(9) is invoked even in the case of an
error.

X-MFC-With:	r278136
2015-02-03 06:04:16 +00:00
Mark Johnston
ac21b651bf Diff reduction with illumos, in preparation for merging r266993 from the
vendor branch. No functional change.

MFC after:	1 week
2015-02-03 05:38:52 +00:00
Steven Hartland
370a13bfff Prevent inlining txg_quiesce
This allows dtrace to monitor the calls to txg_quiesce which can be really
helpful.

Also standardise __noinline order for arc_kmem_reap_now.

Sponsored by:	Multiplay
2015-02-02 00:17:36 +00:00
Mark Johnston
a70a59ea73 Don't attempt to disable enabled fasttrap probes in an exiting process.
There's no need to do so, and we can't hold an exiting process, so this
race can result in panics.

MFC after:	1 week
2015-01-30 05:03:23 +00:00
Mark Johnston
1eb8ad64ea In fasttrap_sigtrap(), use tdsendsignal() rather than tdksignal() to send
SIGTRAP. The latter requires that its thread argument be non-NULL, but
fasttrap_sigtrap() does not.

PR:		193593
MFC after:	1 week
Reported by:	danilo
2015-01-30 04:51:59 +00:00
Xin LI
63cffd61d1 MFV r255258:
Diff reduction with upstream.  The actual change was merged in r272483
already.

MFC after:	2 weeks
2015-01-28 08:56:48 +00:00
Will Andrews
b4e360d239 When creating or updating a node, use vfs_timestamp() for "now" instead
of gethrestime(), to allow the administrator to decide the appropriate
timestamp precision instead of always using nanosecond precision.
2015-01-24 00:43:02 +00:00
Will Andrews
bd3a7c08c4 Remove commented log messages. 2015-01-21 19:30:01 +00:00
Will Andrews
35b540bfb2 Ignore sync requests from the system syncher, i.e. VFS_SYNC(waitfor=MNT_LAZY).
ZFS already commits outstanding data every zfs_txg_timeout seconds, so these
syncs are unnecessarily intrusive.

Submitted by:	gibbs
Sponsored by:	Spectra Logic
MFSpectraBSD:	1105759 on 2014/12/11
2015-01-21 19:25:57 +00:00
Will Andrews
2a2c1d424a Eliminate an #ifdef illumos for zfs_ioc_rename().
Since allow_mounted is a FreeBSD-specific change, default to B_TRUE, then
locally check for the magic bit.  Unconditionally check allow_mounted below.
Convert the setting of allow_mounted to an explicit boolean.

MFC after:	1 week
Sponsored by:	Spectra Logic
MFSpectraBSD:	672578 (in part) on 2013/07/19
2015-01-21 19:20:36 +00:00
Will Andrews
55ddf051d8 Add vfs.zfs.reference_tracking_enable sysctl/tunable.
This is primarily for developer/debugging use; it enables built-in tagged
tracking of refcounts inside ZFS.  It can only be enabled from the loader,
since it modifies how in-core state is managed.  Default remains disabled.

MFC after:	1 week
Sponsored by:	Spectra Logic
2015-01-21 17:03:11 +00:00
Will Andrews
798cbb7523 Fix arc__shrink DTrace probe's to_free argument.
Remove the unnecessary #ifdef _KERNEL, which did not differ in the true or
false cases.  Actually set the value of to_free before using it.

MFC after:	1 week
Sponsored by:	Spectra Logic
2015-01-20 22:39:10 +00:00
Will Andrews
fe20fb9fb0 Use the "zfs_gfs" tag for GFS vnodes to make them easier to identify.
MFC after:	1 week
Sponsored by:	Spectra Logic
2015-01-20 22:31:26 +00:00
Will Andrews
34abed55f3 NSEC_TO_TICK(usec) -> NSEC_TO_TICK(nsec) 2015-01-20 22:29:27 +00:00
Will Andrews
c9c5e04711 Remove unused strdup() #define. 2015-01-20 22:27:45 +00:00
Alexander Motin
d6245e3d44 Allow skipping dmu_buf_will_dirty() call in dsl_dir_transfer_space().
dsl_dir_transfer_space() is mostly called after dsl_dir_diduse_space(),
which already calls dmu_buf_will_dirty() for the same dbuf and tx, so
its duplicate call in those cases will change nothing, only spend time.

Skipping this call by four times reduces time spent in dbuf_write_done()
and descendants, updating dataset statistics with several congested lock
acquisitions.  When rewriting 8K zvol blocks at 1GB/s rate, this reduces
CPU time spent inside dbuf_write_done(), according to profiling, from 45%
of 683K samples to 18% of 422K.

MFC after:	2 weeks
2015-01-20 13:09:12 +00:00
Steven Hartland
5eab7e5406 Clean ZFS spa config before syncing
A number of entries that can be present in the spa config shouldn't be saved
to disk so add a method to ensure this is case. Without this if the last
caller to vdev_config_generate requested stats then we can end up in the
cache file.

Also only skip a none writable pool in the cache file generation if its
active. This prevents unavailable pools incorrectly getting removed from
cache file.

Tested by:	delphij
MFC after:	2 weeks
Sponsored by:	Multiplay
2015-01-18 23:15:49 +00:00
Steven Hartland
bc96366c86 Mechanically convert cddl sun #ifdef's to illumos
Since the upstream for cddl code is now illumos not sun, mechanically
convert all sun #ifdef's to illumos #ifdef's which have been used in all
newer code for some time.

Also do a manual pass to correct the use if #ifdef comments as per style(9)
as well as few uses of #if defined(__FreeBSD__) vs #ifndef illumos.

MFC after:	1 month
Sponsored by:	Multiplay
2015-01-17 14:44:59 +00:00
Alexander Motin
38feff972b Fix overflow bug from r248577, turning 30s TRIM timeout into ~4s.
MFC after:	2 weeks
2015-01-14 16:22:00 +00:00
Alexander Motin
d4f46a775d Reimplement TRIM throttling added in r248577.
Previous throttling implementation approached problem from the wrong side.
It significantly limited useful delaying of TRIM requests and aggregation
potential, while not so much controlled TRIM burstiness under heavy load.

With this change random 4K write benchmarks (probably the worst case for
TRIM) show me IOPS increase by 20%, average latency reduction by 30%, peak
TRIM bursts reduction by 3 times and same peak TRIM map size (memory usage).

Also the new logic does not force map size down so heavily, really allowing
to keep deleted data for 32 TXG or 30 seconds under moderate load.  It was
practically impossible with old throttling logic, which pushed map down to
only 64 segments.

Reviewed by:	smh
MFC after:	2 weeks
Sponsored by:	iXsystems, Inc.
2015-01-14 09:39:57 +00:00
Alexander Motin
5b3a65823d Skip extra bcopy() when scrubbing vdev without redundancy.
According to profiler, this bcopy() can use about 10% of CPU time.

MFC after:	2 weeks
2015-01-12 22:38:55 +00:00
Alexander Motin
f5b85f6551 When aggregating TRIM segments, move the new one to the list end.
New segment at the list head may block all TRIM requests until txg of that
segment can be processed.  On my random I/O tests this change reduce peak
TRIM list length from 650 to 450 segments.  Hopefully it should reduce TRIM
burstiness when list processing is unblocked.

MFC after:	2 weeks
2015-01-11 16:36:39 +00:00
Alexander Motin
2de874ed23 Add LBA as secondary sort key for synchronous I/O requests.
On FreeBSD gethrtime() implemented via getnanouptime(), that has 1ms (1/hz)
precision.  It makes primary sort key (timestamp) collision very possible.
In such situations sorting by secondary key of LBA is much more reasonable
then by totally meaningless zio pointer value.

With this change on multi-threaded synchronous ZVOL read I've measured 10%
throughput increase and average latency reduction.

MFC after:	2 weeks
2015-01-11 00:26:18 +00:00
Alexander Motin
13ea8106d9 Use new optimized dmu_read_uio_dbuf() for ZVOLs in device mode.
This slightly reduces overhead by avoiding dnode_hold()/dnode_rele() calls.

MFC after:	2 weeks
2015-01-10 18:28:58 +00:00
Steven Hartland
8de799ea3a Correct zpool list displaying invalid EXPANDSZ for unavailable pool vdevs
When pools are unavailable their vdevs are also unavailable which means
that vdev_max_asize remains at the default zero. This default was being
used to calculate vs_esize resulting in a negative number as vdev_asize >
vdev_max_asize, which caused zpool list -v to display 16.0E for EXPANDSZ
of these vdevs.
2014-12-31 04:54:48 +00:00
Mark Johnston
cafe874475 Restore the trap type argument to the DTrace trap hook, removed in r268600.
It's redundant at the moment since it can be obtained from the trapframe
on the architectures where DTrace is supported, but this won't be the case
with ARM.
2014-12-23 15:38:19 +00:00
Steven Hartland
51f529b50b Always sync the global ZFS config cache to reflect the new mosconfig
This fixes out of date zpool.cache for root pools, which can cause issues
such as confusion of zdb etc.

MFC after:	1 month
2014-12-23 09:31:24 +00:00
Steven Hartland
6831cf6aa6 Fix panic when resizing ZFS zvol's
Resizing a ZFS ZVOL with debug enabled would result in a panic due to
recursion on dp_config_rwlock.

The upstream change "3464 zfs synctask code needs restructuring" changed
zvol_set_volsize to avoid the recursion on dp_config_rwlock, but this was
missed when originally merged in by r248571 due to significant differences
in our codebases in this area.

These changes also relied on bring in changes from upstream:
3557 dumpvp_size is not updated correctly when a dump zvol's size is
changed, which where also not present.

In order to help prevent future issues in this area a direct comparison
and diff minimisation from current upstream version (b515258) of zvol.c.

Differential Revision:	https://reviews.freebsd.org/D1302
MFC after:	1 month
X-MFC-With:	r276063 & r276066
Sponsored by:	Multiplay
2014-12-22 18:39:38 +00:00
Steven Hartland
08c0f91ecc Refactor zvol locking to minimise diff with upstream
Use #define zfsdev_state_lock spa_namespace_lock instead of replacing all
zfsdev_state_lock with spa_namespace_lock to minimise changes from upstream.

Differential Revision:	D1302
MFC after:	1 month
X-MFC-With	r276063
Sponsored by:	Multiplay
2014-12-22 17:04:51 +00:00
Steven Hartland
39c478eac4 Standardise on illumos for #ifdef's in zvol.c
Also correct as per style(9) on the use of #ifdef comments.

This is a no-op change as pre-cursor to a full cleanup and merge with
upstream zvol changes.

Sponsored by:	Multiplay
2014-12-22 16:38:29 +00:00
Konstantin Belousov
789bdfdbc6 Handle MAKEENTRY cnp flag in the VOP_CREATE(). Curiously, some
fs, e.g. smbfs, already did it.

Tested by:	pho (previous version)
Sponsored by:	The FreeBSD Foundation
MFC after:	2 weeks
2014-12-21 13:29:33 +00:00
Xin LI
62722f7d5b Add missing continue: we can't proceed further if the
kernel does not panic with zfs_panic_recover.

Illumos issue:
    5438 zfs_blkptr_verify should continue after zfs_panic_recover

Reported by:	Coverity
CID:		1232014
2014-12-19 00:20:29 +00:00
Xin LI
7d073f8411 MFV r275914:
As of r270383, the dbuf_compare comparator compares the dbuf
attributes in the following order:

	db_level (indirect level)
	db_blkid (block number)
	db_state (current state)
	the address of the element

Because db_state is being considered before the element's state,
changing of db_state would affect balancedness of the AVL tree,
even when the address of element compares differently.  For
instance, in dbuf_create, db_state may be altered after the
node is inserted into the AVL tree and may break AVL tree
balancedness.

Instead of using db_state as a comparision critera (introduced
in r270383), consider it only when we are doing a lookup, that
is one of the two dbuf pointers contains DB_SEARCH.

Illumos issue:
    5422 preserve AVL invariants in dn_dbufs

MFC after:	2 weeks
2014-12-18 23:45:26 +00:00
Konstantin Belousov
6c21f6edb8 The VOP_LOOKUP() implementations for CREATE op do not put the name
into namecache, to avoid cache trashing when doing large operations.
E.g., tar archive extraction is not usually followed by access to many
of the files created.

Right now, each VOP_LOOKUP() implementation explicitely knowns about
this quirk and tests for both MAKEENTRY flag presence and op != CREATE
to make the call to cache_enter().  Centralize the handling of the
quirk into VFS, by deciding to cache only by MAKEENTRY flag in VOP.
VFS now sets NOCACHE flag for CREATE namei() calls.

Note that the change in semantic is backward-compatible and could be
merged to the stable branch, and is compatible with non-changed
third-party filesystems which correctly handle MAKEENTRY.

Suggested by:	Chris Torek <torek@pi-coral.com>
Reviewed by:	mckusick
Tested by:	pho
Sponsored by:	The FreeBSD Foundation
MFC after:	2 weeks
2014-12-18 10:01:12 +00:00
Xin LI
a771fba68c MFV r275783:
Convert ARC flags to use enum.  Previously, public flags are defined in
arc.h and private flags are defined in arc.c which can lead to confusion
and programming errors.

Consistently use 'hdr' (when referencing arc_buf_hdr_t) instead of 'buf'
or 'ab' because arc_buf_t are often named 'buf' as well.

Illumos issue:
    5369 arc flags should be an enum
    5370 consistent arc_buf_hdr_t naming scheme

MFC after:	2 weeks
2014-12-15 18:22:45 +00:00
Xin LI
7728b3d28c MFV r275551:
Remove "dbuf phys" db->db_data pointer aliases.

Use function accessors that cast db->db_data to the appropriate
"phys" type, removing the need for clients of the dmu buf user
API to keep properly typed pointer aliases to db->db_data in order
to conveniently access their data.

sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zap_leaf.c:
        In zap_leaf() and zap_leaf_byteswap, now that the pointer alias
        field l_phys has been removed, use the db_data field in an on
        stack dmu_buf_t to point to the leaf's phys data.

sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dbuf.c:
        Remove the db_user_data_ptr_ptr field from dbuf and all logic
        to maintain it.

sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dbuf.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dnode.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/sys/dbuf.h:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/sys/dmu.h:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dsl_dataset.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dsl_dir.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/sa.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zap.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zap_micro.c:
        Modify the DMU buf user API to remove the ability to specify
        a db_data aliasing pointer (db_user_data_ptr_ptr).

cddl/contrib/opensolaris/cmd/zdb/zdb.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dmu_diff.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dmu_objset.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dmu_send.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dmu_traverse.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dmu_tx.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dsl_bookmark.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dsl_dataset.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dsl_deadlist.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dsl_deleg.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dsl_destroy.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dsl_dir.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dsl_pool.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dsl_prop.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dsl_scan.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dsl_synctask.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dsl_userhold.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/sa.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/spa.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/spa_history.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zap.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zap_leaf.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zap_micro.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_ioctl.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/sys/dsl_dataset.h:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/sys/dsl_dir.h:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/sys/zap_impl.h:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/sys/zap_leaf.h:
        Create and use the new "phys data" accessor functions
        dsl_dir_phys(), dsl_dataset_phys(), zap_m_phys(),
        zap_f_phys(), and zap_leaf_phys().

sys/cddl/contrib/opensolaris/uts/common/fs/zfs/sys/dsl_dataset.h:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/sys/dsl_dir.h:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/sys/zap_impl.h:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/sys/zap_leaf.h:
        Remove now unused "phys pointer" aliases to db->db_data
        from clients of the DMU buf user API.

Illumos issue:
    5314 Remove "dbuf phys" db->db_data pointer aliases in ZFS

MFC after:	2 weeks
2014-12-15 07:52:23 +00:00
Xin LI
a676454a98 MFV r275550:
In addition to r273158, make the code in spa_sync() that checks if the
current TXG is a no-op TXG less fragile.

Illumos issue:
    5347 idle pool may run itself out of space

MFC after:	2 weeks
2014-12-15 05:10:55 +00:00
Xin LI
51e527396b MFV r275549:
Add a loader tunable, vfs.zfs.arc_meta_min, which controls how much metadata
ZFS should keep in ARC at minimum.

In arc_evict(), when doing recycle, take more factors into account by
applying the following policy:

  1. If no evictable data, evict metadata;
  2. If no evictable metadata, evict data;
  3. If we hit arc_meta_limit, evict metadata;
  4. If we haven't hit arc_meta_min, evict data;
  5* (Illumos only, not present in new FreeBSD code, yet) evict the oldest
     cached element from data and metadata.
     (FreeBSD) evict the data type specified by caller, which is the
     existing behavior.

Note that because of our splitted locks (implemented in r205231 to improve
scalability by reducing lock contention), implementing the fifth Illumos
behavior will not be cheap, so for now just implement the 1-4 and fall back
to current behavior for 5.

Illumos issue:
    5368 ARC should cache more metadata

MFC after:	2 months (assuming we didn't found better solution)
2014-12-15 04:51:36 +00:00
Xin LI
5571af1d8f MFV r247174:
Expose arc_meta_limit, et al via kstats.

Note that as a result, vfs.zfs.arc_meta_used is removed.
The existing vfs.zfs.arc_meta_limit sysctl/tunable is retained
with a SYSCTL_PROC wrapper.

Illumos ZFS issues:
    3561 arc_meta_limit should be exposed via kstats

Relnotes:	yes
MFC after:	2 weeks
2014-12-13 19:17:28 +00:00
Xin LI
d25a5fd878 MFV r275548:
Verify that the block pointer is structurally valid, before attempting to
read it in.  It can only be invalid in the case of a ZFS bug, but this
change will help identify such bugs in a more transparent way, by
panic'ing with a relevant message, rather than indexing off the end of an
array or something.

Illumos issue:
    5349 verify that block pointer is plausible before reading

MFC after:	2 weeks
2014-12-13 02:08:18 +00:00
Xin LI
ca7354fa07 MFV r275546:
Reduce scrub activities when system there is enough dirty data, namely when
dirty data is more than zfs_vdev_async_write_active_min_dirty_percent (once
we start to increase the number of concurrent async writes).

While there also correct rounding error which would make scrub end up
pausing for (zfs_txg_timeout + 1) seconds instead of the desired
zfs_txg_timeout seconds.

Illumos issue:
    5351 scrub goes for an extra second each txg
    5352 scrub should pause when there is some dirty data

MFC after:	2 weeks
2014-12-13 01:39:24 +00:00
Xin LI
8b2dbf8af3 MFV r275545:
If zio_checksum_error() returns other than ECKSUM (e.g. EINVAL), it does not
fill in the "zio_bad_cksum_t *info" parameter. Caller should not attempt to
use it in this case.

Illumos issue:
    5348 zio_checksum_error() only fills in info if ECKSUM

MFC after:	2 weeks
2014-12-13 01:26:06 +00:00
Xin LI
08fda6cd66 MFV r275544:
Clean up some duplicated code in dnode_sync() around freeing spill blocks.

Illumos issue:
    5350 clean up code in dnode_sync()

MFC after:	2 weeks
2014-12-13 01:18:23 +00:00
Xin LI
48438ce4c4 MFV r275543:
Remove always true tests for ds->ds_phys' presence.

Clean up assertions in dsl_dataset_disown.

Remove unreachable code in dsl_dataset_disown().

Illumos issue:
    5310 Remove always true tests for non-NULL ds->ds_phys

MFC after:	2 weeks
2014-12-13 01:14:59 +00:00
Xin LI
c90d7474a8 MFV r275542:
If a dnode has a spill block and there is an error while accessing
a data block then traverse_dnode() loses information about that error
and returns a status of visiting the spill block.

This issue is discovered by Spectra Logic.

Illumos issue:
    5311 traverse_dnode may report success when it should not

Original author:	gibbs
MFC after:		2 weeks
2014-12-13 01:10:17 +00:00
Xin LI
c2161091ad MFV r275540:
When importing a pool, don't assume that the passed pool configuration
at vdev_load is always vaild.  It's possible that a stale configuration
that comes with extra vdevs, where metaslab_init() would fail because
of lower layer returns error.

Change the code to make metaslab_init() handle and return errors from
lower layer and pass it back to upper layer and handle it there.

Illumos issue:
    5213 panic in metaslab_init due to space_map_open returning ENXIO

MFC after:	2 weeks
2014-12-08 06:04:42 +00:00
Andriy Gapon
036a8c5dac remove opensolaris cyclic code, replace with high-precision callouts
In the old days callout(9) had 1 tick precision and that was inadequate
for some uses, e.g. DTrace profile module, so we had to emulate cyclic
API and behavior.  Now we can directly use callout(9) in the very few
places where cyclic was used.

Differential Revision:	https://reviews.freebsd.org/D1161
Reviewed by:	gnn, jhb, markj
MFC after:	2 weeks
2014-12-07 11:21:41 +00:00
Andrew Turner
f9307cced7 Apply the same fix in r274697 to the ARM case. 2014-12-06 12:03:09 +00:00
Xin LI
54f76dcb4a MFV r275535:
Unexpand ISP2() and MSEC2NSEC().

Illumos issue:
    5255 uts shouldn't open-code ISP2

MFC after:	2 weeks
2014-12-06 09:38:28 +00:00
Xin LI
6054c38913 MFV r275534:
Sync with Illumos.  This have no effect to FreeBSD.

Illumos issue:
    5285 pass in cpu_pause_func via pause_cpus

MFC after:	2 weeks
2014-12-06 09:14:46 +00:00
Xin LI
d4548c2e8e MFC r275533:
Sync with Illumos.  This have no effect to FreeBSD.

Illumos issue:
    5100 sparc build failed after 5004

MFC after:	2 weeks
2014-12-06 09:11:13 +00:00
Xin LI
4603a0aeb2 Use %d instead of %u for error number. This way we see ERESTART as -1
not 4294967295 when doing DTrace.

MFC after:	2 weeks
2014-12-05 22:56:10 +00:00
Xin LI
26f96d922b Fix a regression introduced in r274337 (large block support)
In dsl_dataset_hold_obj() we used zap_contains(.., DS_FIELD_LARGE_BLOCKS)
to determine whether the extensible (zapifyed) dataset have large blocks.
The code expects the result be either 0 (found) or ENOENT (not found),
however reused the variable 'err' which later code expects to be 0.

Fix this by adopting similar code construct that is used later for
DS_FIELD_BOOKMARK_NAMES, which uses a temporary variable zaperr to catch
errors from zap_* rountines.

Reported by:	Peter J. Creath (on FreeNAS; FreeNAS bug #6848)
Illumos issue:	5393 spurious failures from dsl_dataset_hold_obj()
Reviewed by:	mahrens
Sponsored by:	iXsystems, Inc.
X-MFC with:	r274337
2014-12-05 18:29:01 +00:00
Alexander Motin
ef8daf3fed Add GET LBA STATUS command support to CTL.
It is implemented for LUNs backed by ZVOLs in "dev" mode and files.
GEOM has no such API, so for LUNs backed by raw devices all LBAs will
be reported as mapped/unknown.

MFC after:	2 weeks
Sponsored by:	iXsystems, Inc.
2014-12-04 11:34:19 +00:00
Andriy Gapon
782c06dfc8 zfs_putpages: actually update mtime and ctime
Reported by:	Paul Koch <paul.koch@akips.com>
Tested by:	Paul Koch <paul.koch@akips.com>
MFC after:	2 weeks
2014-12-02 11:44:56 +00:00
Xin LI
d10e52d627 Revert r273060 per discussion with avg@ as we need to make L2ARC
aware of 4K devices and this one is not the right fix anyway.
2014-11-26 02:20:25 +00:00
Dimitry Andric
c577699e4b Fix the following -Werror warning from clang 3.5.0, while building cddl/lib/libctf:
In file included from cddl/contrib/opensolaris/common/ctf/ctf_create.c:31:
In file included from sys/cddl/contrib/opensolaris/uts/common/sys/sysmacros.h:34:
sys/cddl/contrib/opensolaris/uts/common/sys/isa_defs.h:334:9: warning: '_ILP32' macro redefined [-Wmacro-redefined]
#define _ILP32
        ^
<built-in>:26:9: note: previous definition is here
#define _ILP32 1
        ^
1 warning generated.

This is because clang 3.5.0 started predefining _ILP32 and __ILP32__ for
the i386 arch.  (Earlier versions already predefined _LP64 and __LP64__
for the x86_64 arch.)

Reviewed by:	emaste, avg, smh, delphij, markj
Differential Revision: https://reviews.freebsd.org/D1187
2014-11-19 07:44:21 +00:00
Xin LI
8c3d6a4ab2 Make vfs.zfs.max_recordsize read-write at runtime.
MFC after:	2 weeks
2014-11-18 22:35:19 +00:00
Xin LI
8efcd876ca Add a tunable for spa_slop_shift which controls how much space we
would reserve by default.  Tuning is not recommended.

MFC after:	2 weeks
2014-11-18 18:52:38 +00:00
Xin LI
18144ab1a3 Allow tuning zfs_max_recordsize via loader tunable. Tuning is NOT
recommended.

Requested by:	Slawa Olhovchenkov <slw zxy spb ru>
MFC after:	2 weeks
2014-11-18 18:40:01 +00:00
Andriy Gapon
2c51c83bc8 l2arc: restore correct rounding up of asize of compressed data
This rounding up was lost in a mismerge of illumos code.
See r268075 MFV r267565.
After that commit zio_compress_data() no longer performs any compressed
size adjustment, so it needs to be done externally.  On FreeBSD we round
up the size using vdev_ashift rather than SPA_MINBLOCKSIZE so that 4KB
devices are properly supported.

Additionally, zero out the buffer tail only if compression succeeds.
The compression is considered successful if the size of compressed
data after rounding up to account for the vdev ashift is less than the
original data size.  It does not make sense to have the data compressed
if all the savings are lost to rounding up.
With the new zio_compress_data() it could have been possible that the
rounded compressed size would be greater than the original size and thus
we could zero beyond the allocated buffer if the zeroing code was kept
at the original place.

Discussed with:	delphij, gibbs
MFC after:	2 weeks
X-MFC with:	r274627
2014-11-17 14:45:42 +00:00
Andriy Gapon
0908b20b7e Revert r269093 which introduced physical zio alignment transform
Size of physical ZIOs must never be implicitly adjusted, it's
a responsibility of a caller to make sure that such a ZIO has proper offset
and size.

Discussed with:	delphij, gibbs
MFC after:	2 weeks
2014-11-17 14:16:02 +00:00
Steven Hartland
a559adfbce Disable TRIM on file backed ZFS vdevs and fix TRIM on init
After r265152 TRIM requests are ZIO_TYPE_FREE instead of ZIO_TYPE_IOCTL
this meant file backed vdevs to attempted to process the ZIO as a write
causing a panic.

We now disable TRIM on file backed vdevs and ASSERT the ZIO types supported
by each vdev type to ensure we explicity support the ZIO type being
processed.

Also ensure that TRIM on init is not procesed for devices which declare they
didn't support TRIM via vdev_notrim.

PR:		195061, 194976, 191573
Sponsored by:	Multiplay
2014-11-17 11:32:10 +00:00
Konstantin Belousov
6e646651d3 Remove the no-at variants of the kern_xx() syscall helpers. E.g., we
have both kern_open() and kern_openat(); change the callers to use
kern_openat().

This removes one (sometimes two) levels of indirection and
consolidates arguments checks.

Reviewed by:	mckusick
Sponsored by:	The FreeBSD Foundation
MFC after:	1 week
2014-11-13 18:01:51 +00:00
Xin LI
8bcd603968 MFV r274273:
ZFS large block support.

Please note that booting from datasets that have recordsize greater
than 128KB is not supported (but it's Okay to enable the feature on
the pool).  This *may* remain unchanged because of memory constraint.

Limited safety belt is provided for mounted root filesystem but use
caution is advised.

Illumos issue:
    5027 zfs large block support

MFC after:	1 month
2014-11-10 08:20:21 +00:00
Xin LI
42350b6bde MFV r274272 and diff reduction with upstream.
Illumos issue:
    5244 zio pipeline callers should explicitly invoke next stage

Tested with:	ztest plus ZFS over GELI configuration
MFC after:	1 month
2014-11-09 07:37:00 +00:00
Xin LI
81f1255e58 MFV r274271:
Improve zdb -b performance:

 - Reduce gethrtime() call to 1/100th of blkptr's;
 - Skip manipulating the size-ordered tree;
 - Issue more (10, previously 3) async reads;
 - Use lighter weight testing in traverse_visitbp();

Illumos issue:
    5243 zdb -b could be much faster

MFC after:	2 weeks
2014-11-08 07:30:40 +00:00
Andriy Gapon
2fd3cc0cb2 fix l2arc compression buffers leak
We have observed that arc_release() can be called concurrently with a
l2arc in-flight write.
Also, we have observed that arc_hdr_destroy() can be called from
arc_write_done() for a zio with ZIO_FLAG_IO_REWRITE flag in similar
circumstances.

Previously the l2arc headers would be freed while leaking their
associated compression buffers.  Now the buffers are placed on
l2arc_free_on_write list for delayed freeing.  This is similar to what
was already done to arc buffers that were supposed to be freed
concurrently with in-flight writes of those buffers.

In addition to fixing the discovered leaks this change also adds some
protective code to assert that a compression buffer associated with a
l2arc header is never leaked.

A new kstat l2_cdata_free_on_write is added.  It keeps a count of
delayed compression buffer frees which previously would have been leaks.

Tested by:	Vitalij Satanivskij <satan@ukr.net> et al
Requested by:	many
MFC after:	2 weeks
Sponsored by:	HybridCluster / ClusterHQ
2014-11-06 11:08:02 +00:00
Alexander Motin
c3e7ba3e6d Add to CTL support for logical block provisioning threshold notifications.
For ZVOL-backed LUNs this allows to inform initiators if storage's used or
available spaces get above/below the configured thresholds.

MFC after:	2 weeks
Sponsored by:	iXsystems, Inc.
2014-11-06 00:48:36 +00:00
Josh Paetzel
14127f5b21 This change addresses 4 bugs in ZFS exposed by Richard Kojedzinszky's
crash.sh script attached to FreeNAS bug 4109:
https://bugs.freenas.org/issues/4109

Three are in the snapshot layer:
a) AVG explains in his notes: https://wiki.freebsd.org/AvgVfsSolarisVsFreeBSD

"VOP_INACTIVE must not do any destructive actions to a vnode
and its filesystem node, nor invalidate them in any way."
gfs_vop_inactive and zfsctl_snapshot_inactive did just that. In
OpenSolaris VOP_INACTIVE is much closer to FreeBSD's VOP_RECLAIM.
Rename & move them to gfs_vop_reclaim and zfsctl_snapshot_reclaim
and merge in the requisite vnode_destroy from zfsctl_common_reclaim.

b) gfs_lookup_dot and various zfsctl functions do not honor the
FreeBSD VFS convention of only locking from the root downward. When
looking up ".." the convention is to drop the current leaf vnode lock before
acquiring the directory vnode and then subsequently re-acquiring the lock on the
leaf vnode. This fixes that in all the places that our exercised by crash.sh.

c) The snapshot may already be unmounted when the directory vnode is reclaimed.
Check for this case and return.

One in the common layer:
d) Callers of traverse expect the reference to the vnode passed in to be
maintained. Don't release it.

This last one may be an unclear contract. There may in fact be some callers that
do expect the reference to be dropped on success in addition to callers that
expect it to be released. In this case a further audit of the callers is needed
and a consensus on the correct behavior.

PR:	184677
Submitted by:	kmacy
Reviewed by:	delphij, will, avg
MFC after:	2 weeks
Sponsored by:	iXsystems
2014-10-25 17:42:44 +00:00
Justin Hibbits
3ff2096995 Whitespace
X-MFC-with:	r273570
MFC after:	1 week
2014-10-24 03:34:21 +00:00
Justin Hibbits
24d5dfb116 Three updates to PowerPC FBT:
* Use a constant to define the number of stack frames in a probe exception.
* Only allow function symbols in powerpc64 ('.' prefixed)
* Set the fbtp_roffset for return probes, so the correct dtrace_probe call is
  made.

MFC after:	1 week
2014-10-24 03:33:01 +00:00
Hans Petter Selasky
f0188618f2 Fix multiple incorrect SYSCTL arguments in the kernel:
- Wrong integer type was specified.

- Wrong or missing "access" specifier. The "access" specifier
sometimes included the SYSCTL type, which it should not, except for
procedural SYSCTL nodes.

- Logical OR where binary OR was expected.

- Properly assert the "access" argument passed to all SYSCTL macros,
using the CTASSERT macro. This applies to both static- and dynamically
created SYSCTLs.

- Properly assert the the data type for both static and dynamic
SYSCTLs. In the case of static SYSCTLs we only assert that the data
pointed to by the SYSCTL data pointer has the correct size, hence
there is no easy way to assert types in the C language outside a
C-function.

- Rewrote some code which doesn't pass a constant "access" specifier
when creating dynamic SYSCTL nodes, which is now a requirement.

- Updated "EXAMPLES" section in SYSCTL manual page.

MFC after:	3 days
Sponsored by:	Mellanox Technologies
2014-10-21 07:31:21 +00:00
Xin LI
78701de4b7 Add tunable vfs.zfs.space_map_blksz for space map's maximum block size.
MFC after:	2 weeks
2014-10-18 22:11:10 +00:00
Davide Italiano
2be111bf7d Follow up to r225617. In order to maximize the re-usability of kernel code
in userland rename in-kernel getenv()/setenv() to kern_setenv()/kern_getenv().
This fixes a namespace collision with libc symbols.

Submitted by:   kmacy
Tested by:      make universe
2014-10-16 18:04:43 +00:00
Steven Hartland
ca6505b818 Prevent ZFS leaking pool free space
When processing async destroys ZFS would leak space every txg timeout
(5 seconds by default), if no writes occurred, until the pool is totally
full. At this point it would be unfixable without a pool recreation.

In addition if the machine was rebooted with the pool in this situation
would fail to import on boot, hanging indefinitely, as the import process
requires the ability to write data to the pool. Any attempts to query the
pool status during the hung import would not return as the import holds
the pool lock.

The only way to import such a pool would be to specify -o readonly=on
to the zpool import.

zdb -bb <pool> can be used to check for "deferred free" size which is where
this lost space will be counted.

MFC after:	3 days
Sponsored by:	Multiplay
2014-10-16 02:23:27 +00:00
Xin LI
ba6e85e0cf Use write_psize instead of write_asize when doing vdev_space_update.
Without this change the accounting of L2ARC usage would be wrong and
give 16EB free space because the number became negative and overflows.

Obtained from:	FreeNAS (issue #6239)
MFC after:	2 weeks
2014-10-13 20:39:51 +00:00
Xin LI
a4f5b8db9f Add a tunable for arc_shrink_shift (vfs.zfs.arc_shrink_shift) that
controls how much fraction, 1/2^arc_shrink_shift, should be reclaimed
when there is memory pressure.

Submitted by:	Richard Kojedzinszky <krichy at tvnetwork.hu>
MFC after:	2 weeks
2014-10-13 05:34:10 +00:00
Xin LI
eba15cf463 MFV r272804:
Refactor the code and stop restore_object from creating two transactions.

Illumos issue:
    3693 restore_object uses at least two transactions to restore an object

MFC after:	2 weeks
2014-10-09 07:52:51 +00:00
Xin LI
ce44f14b41 MFV r272803:
Illumos issue:
    5175 implement dmu_read_uio_dbuf() to improve cached read performance

MFC after:	2 weeks
2014-10-09 07:18:40 +00:00
Andriy Gapon
c3d1d2e104 l2arc_write_buffers: reduce headroom value
FreeBSD has ARC_BUFC_NUMMETADATALISTS metadata lists and ARC_BUFC_NUMDATALISTS
data lists (currently both are 16) while illumos has just a single list
of each kind.

headroom determines how much data is scanned on a single list
during each run of the l2arc feed thread.
Because FreeBSD has more lists we proportionally decrease the limit.

Reviewed by:	Brendan Gregg (earlier version)
MFC after:	2 weeks
Sponsored by:	HybridCluster
2014-10-07 16:08:21 +00:00
Andriy Gapon
9f96723ec5 revert r272702: wrong (earlier) change was committed 2014-10-07 16:06:10 +00:00