John Baldwin [Tue, 7 Aug 2018 00:10:58 +0000 (00:10 +0000)]
Remove spurious ABI tags from kdump output.
The abidump routine output an ABI tag when -A was specified for records
that were not displayed due to type or pid filtering. To fix, split
the code to lookup the ABI from the code to display the ABI, move the
code to display the ABI into dumpheader(), and move dumpheader() later
in the main loop as a simplification. Previously dumpheader() was
called under a condition that repeated conditions made later in the
main loop.
John Baldwin [Mon, 6 Aug 2018 23:51:08 +0000 (23:51 +0000)]
Make the system C11 atomics headers fully compatible with external GCC.
The <sys/cdefs.h> and <stdatomic.h> headers already included support for
C11 atomics via intrinsincs in modern versions of GCC, but these versions
tried to "hide" atomic variables inside a wrapper structure. This wrapper
is not compatible with GCC's internal <stdatomic.h> header, so that if
GCC's <stdatomic.h> was used together with <sys/cdefs.h>, use of C11
atomics would fail to compile. Fix this by not hiding atomic variables
in a structure for modern versions of GCC. The headers already avoid
using a wrapper structure on clang.
Note that this wrapper was only used if C11 was not enabled (e.g.
via -std=c99), so this also fixes compile failures if a modern version
of GCC was used with -std=c11 but with FreeBSD's <stdatomic.h> instead
of GCC's <stdatomic.h> and this change fixes that case as well.
Reported by: Mark Millard
Reviewed by: kib
Differential Revision: https://reviews.freebsd.org/D16585
Navdeep Parhar [Mon, 6 Aug 2018 23:21:13 +0000 (23:21 +0000)]
cxgbe(4): Allow user-configured and driver-configured traffic classes to
be used simultaneously. Move sysctl_tc and sysctl_tc_params to
t4_sched.c while here.
Navdeep Parhar [Mon, 6 Aug 2018 21:54:51 +0000 (21:54 +0000)]
cxgbe(4): Break up sysctl_bitfield into 8 bit and 16 bit variants. Have
them display the current value of the bitfield rather than the fixed
value that was provided when the sysctl node was created.
Kirk McKusick [Mon, 6 Aug 2018 21:09:11 +0000 (21:09 +0000)]
Put in place the framework for consolodating contiguous blocks into
a smaller number of larger TRIM requests. The hope had been to have
the full TRIM consolodation in place for 12.0, but the algorithms
are still under development and need further testing. With this
framework in place it will be possible to easily add TRIM consolodation
once the optimal strategy has been found.
The only functional change with this patch is the elimination of TRIM
requests for blocks that are freed before they have been likely to
have been written.
Reviewed by: kib
Discussed with: Warner Losh and Chuck Silvers
Sponsored by: Netflix
Colin Percival [Mon, 6 Aug 2018 19:21:32 +0000 (19:21 +0000)]
Add EC2PUBLICSNAP option to EC2 builds; this passes a (recently added)
flag to bsdec2-image-upload instructing it to mark the snapshot of its
root disk as public (which is independent from marking the created AMIs
as public).
Address concerns about CPU usage while doing TCP reassembly.
Currently, the per-queue limit is a function of the receive buffer
size and the MSS. In certain cases (such as connections with large
receive buffers), the per-queue segment limit can be quite large.
Because we process segments as a linked list, large queues may not
perform acceptably.
The better long-term solution is to make the queue more efficient.
But, in the short-term, we can provide a way for a system
administrator to set the maximum queue size.
We set the default queue limit to 100. This is an effort to balance
performance with a sane resource limit. Depending on their
environment, goals, etc., an administrator may choose to modify this
limit in either direction.
Reviewed by: jhb
Approved by: so
Security: FreeBSD-SA-18:08.tcp
Security: CVE-2018-6922
Emmanuel Vadot [Mon, 6 Aug 2018 17:21:20 +0000 (17:21 +0000)]
release: arm: Copy the dtb to the fat partition
When booting via EFI on arm we have no way to know the dtb file to load
and we always use the one provided from the bootloader.
This works in most case but :
U-Boot have some really old DTB for some boards, the sync from Linux isn't done automatically for all boards
Some boards (like TI BeagleBone series) use one u-boot for all the model and it doesn't embed the DTBs
Some boards (like IMX6 based ones), don't embed the DTB
We want u-boot to load and patch the DTB with the mac address or the display
node enabled or not.
Mark Johnston [Mon, 6 Aug 2018 16:22:01 +0000 (16:22 +0000)]
dhclient: Don't chroot if we are in capability mode.
The main dhclient process is Capsicumized but also chroots to
restrict filesystem access. With r322369, pidfile(3) maintains a
directory descriptor for the pidfile, which can cause the chroot
to fail in certain cases. To minimize the problem, only chroot
if we fail to enter capability mode, and store dhclient pidfiles
in a subdirectory of /var/run, thus restricting access via
pidfile(3)'s directory descriptor.
PR: 223327
Reviewed by: cem, oshogbo
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D16584
Emmanuel Vadot [Mon, 6 Aug 2018 05:36:00 +0000 (05:36 +0000)]
aw_thermal: Add nvmem and H5 support
Now that aw_sid expose nvmem interface, use that to read the calibration
data.
Add support for H5 SoC.
Fix the bindings, we used to have non-upstreamed bindings. Switch to the
one that have been sent upstream. They are not stable yet, so we switch
from custom, wrong, bindings to correct, proposed bindings
Emmanuel Vadot [Mon, 6 Aug 2018 05:35:24 +0000 (05:35 +0000)]
aw_sid: Add nvmem interface
Rework aw_sid so it can work with the nvmem interface.
Each SoC expose a set of fuses (for now rootkey/boardid and, if available,
the thermal calibration data). A fuse can be private or public, reading private
fuse needs to be done via some registers instead of reading directly.
Each fuse is exposed as a sysctl.
For now leave the possibility for a driver to read any fuse without using
the nvmem interface as the awg and emac driver use this to generate a mac
address.
Ian Lepore [Sun, 5 Aug 2018 22:24:38 +0000 (22:24 +0000)]
Document 64-bit arm in terms of arch name (aarch64) not machine (arm64).
Other architectures are documented in terms of the name that is displayed by
'uname -p', aka MACHINE_ARCH and TARGET_ARCH in the build system, now
aarch64 matches the rest of them.
Rick Macklem [Sun, 5 Aug 2018 19:21:50 +0000 (19:21 +0000)]
Copy all bits of a file handle in case there is padding in the structure.
At least on x86, fhandle_t is a packed structure, so I believe an
assignment will copy all the bits. However, for some current/future
architectures, there might be padding in the structure that doesn't get
copied via an assignment.
Since NFS assumes a file handle is an opaque blob of bits that can be
compared via memcmp()/bcmp(), all the bits including any padding must be
copied.
This patch replaces the assignments with a call to a byte copy function.
Spotted during code inspection.
Kristof Provost [Sun, 5 Aug 2018 13:54:37 +0000 (13:54 +0000)]
pf: Increase default hash table size
Now that we (by default) limit the number of states to 100.000 it makse sense
to also adjust the default size of the hash table.
Based on the benchmarking results in
https://github.com/ocochard/netbenches/blob/master/Atom_C2758_8Cores-Chelsio_T540-CR/pf-states_hashsize/results/fbsd12-head.r332390/README.md
128K entries offers a good compromise between performance and memory use.
Users may still overrule this setting with the net.pf.states_hashsize and
net.pf.source_nodes_hashsize loader(8) tunables.
Kristof Provost [Sun, 5 Aug 2018 11:15:28 +0000 (11:15 +0000)]
zfsboot: Fix startup crash
On a FreeNAS mini XL, with geli encrypted drives the loader crashed in
geli_read().
When we iterate over the list of disks and allocate the zfsdsk structures we
don’t zero out the gdev pointer. In one case that resulted in geli_read()
(called on the bogus pointer) dividing by zero.
Use calloc() to ensure the zfsdsk structure is always zeroed, so the pointer is
initialised to NULL. As a side benefit it gets rid of one #ifdef
LOADER_GELI_SUPPORT.
Emmanuel Vadot [Sun, 5 Aug 2018 06:15:35 +0000 (06:15 +0000)]
extres: clkdiv: Fix div_with_table
We didn't allowed a divider register value of 0 which can exists and
also didn't wrote the value but the divider, which result of a wrong
frequency to be selected
Emmanuel Vadot [Sun, 5 Aug 2018 06:10:13 +0000 (06:10 +0000)]
arm: allwinner: Disconnect A10/A20 HDMI driver
It doesn't work since 2 years when we stopped patching DTS.
The DTS now have the correct bindings but they are a lot different
from our hacked ones we used to have (and more representative of the
reality).
Emmanuel Vadot [Sun, 5 Aug 2018 06:08:23 +0000 (06:08 +0000)]
arm: allwinner: Remove old unused clocks
Remove the old clocks for allwinner as now all the SoCs have been converted
to clkng.
The only old clock now is the gmac clock which still lives under the /clocks
dts node.
Conrad Meyer [Sat, 4 Aug 2018 22:08:24 +0000 (22:08 +0000)]
settimeofday(2): Remove stale note about timezone
Contrary to the removed comment, the kernel does appear to use the timezone
argument of settimeofday. The comment dates to the BSD4.4 import; I assume it
is just stale.
Kyle Evans [Sat, 4 Aug 2018 21:41:10 +0000 (21:41 +0000)]
efirt: Don't enter EFI context early, convert addrs to KVA instead
efi_enter here was needed because efi_runtime dereference causes a fault
outside of EFI context, due to runtime table living in runtime service
space. This may cause problems early in boot, though, so instead access it
by converting paddr to KVA for access.
While here, remove the other direct PHYS_TO_DMAP calls and the explicit DMAP
requirement from efidev.
Swapped-out process that is WKILLED must be swapped in as soon as
possible. The reason is that such process can be killed by OOM and
its pages can be only freed if the process exits. To exit, the kernel
stack of the process must be mapped.
When allocating pages for the stack of the WKILLED process on swap in,
use VM_ALLOC_SYSTEM requests to increase the chance of the allocation
to succeed.
Add counter of the swapped out processes to avoid unneeded iteration
over the allprocs list when there is no work to do, reducing the
allproc_lock ownership.
Mark Johnston [Sat, 4 Aug 2018 20:29:58 +0000 (20:29 +0000)]
Fix the regression test for PR 181741.
With r337328, the test hangs becase the sendmsg() call will block until
the receive buffer is at least partially drained. Fix the problem by
using a non-blocking socket and allowing short writes. Also assert
that a SCM_CREDS message was received if one was expected.
PR: 181741
MFC after: 1 month
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D16516
Mark Johnston [Sat, 4 Aug 2018 20:26:54 +0000 (20:26 +0000)]
Don't check rcv sockbuf limits when sending on a unix stream socket.
sosend_generic() performs an initial comparison of the amount of data
(including control messages) to be transmitted with the send buffer
size. When transmitting on a unix socket, we then compare the amount
of data being sent with the amount of space in the receive buffer size;
if insufficient space is available, sbappendcontrol() returns an error
and the data is lost. This is easily triggered by sending control
messages together with an amount of data roughly equal to the send
buffer size, since the control message size may change in uipc_send()
as file descriptors are internalized.
Fix the problem by removing the space check in sbappendcontrol(),
whose only consumer is the unix sockets code. The stream sockets code
uses the SB_STOP mechanism to ensure that senders will block if the
receive buffer fills up.
PR: 181741
MFC after: 1 month
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D16515
Dimitry Andric [Sat, 4 Aug 2018 14:57:23 +0000 (14:57 +0000)]
Fix build of hyperv with base gcc on i386
Summary:
Base gcc fails to compile `sys/dev/hyperv/pcib/vmbus_pcib.c` for i386,
with the following -Werror warnings:
cc1: warnings being treated as errors
/usr/src/sys/dev/hyperv/pcib/vmbus_pcib.c: In function 'new_pcichild_device':
/usr/src/sys/dev/hyperv/pcib/vmbus_pcib.c:567: warning: cast from pointer to integer of different size [-Wpointer-to-int-cast]
/usr/src/sys/dev/hyperv/pcib/vmbus_pcib.c: In function 'vmbus_pcib_on_channel_callback':
/usr/src/sys/dev/hyperv/pcib/vmbus_pcib.c:940: warning: cast to pointer from integer of different size [-Wint-to-pointer-cast]
/usr/src/sys/dev/hyperv/pcib/vmbus_pcib.c: In function 'hv_pci_protocol_negotiation':
/usr/src/sys/dev/hyperv/pcib/vmbus_pcib.c:1012: warning: cast from pointer to integer of different size [-Wpointer-to-int-cast]
/usr/src/sys/dev/hyperv/pcib/vmbus_pcib.c: In function 'hv_pci_enter_d0':
/usr/src/sys/dev/hyperv/pcib/vmbus_pcib.c:1073: warning: cast from pointer to integer of different size [-Wpointer-to-int-cast]
/usr/src/sys/dev/hyperv/pcib/vmbus_pcib.c: In function 'hv_send_resources_allocated':
/usr/src/sys/dev/hyperv/pcib/vmbus_pcib.c:1125: warning: cast from pointer to integer of different size [-Wpointer-to-int-cast]
/usr/src/sys/dev/hyperv/pcib/vmbus_pcib.c: In function 'vmbus_pcib_map_msi':
/usr/src/sys/dev/hyperv/pcib/vmbus_pcib.c:1730: warning: cast from pointer to integer of different size [-Wpointer-to-int-cast]
This is because on i386, several casts from `uint64_t` to a pointer
reduce the value from 64 bit to 32 bit.
For gcc, this can be fixed by an intermediate cast to uintptr_t. Note
that I am assuming the incoming values will always fit into 32 bit!
Differential Revision: https://reviews.freebsd.org/D15753
MFC after: 3 days
Xin LI [Sat, 4 Aug 2018 14:13:09 +0000 (14:13 +0000)]
In r337271, we limited the sector number to the lower of calculated
number and CHS based number. However, on some systems, BIOS would
report 0 in CHS fields, making the system to think there is 0 sectors.
Add a check before comparing the calculated total with bd_sectors.
wmt(4): Use internal function to calculate input report size
Usbhid's hid_report_size() calculates integral size of all reports of given
kind found in the HID descriptor rather then exact size of report with given
ID as its userland counterpart does. As all input data processed by the
driver is located within the same report, calculate required driver's buffer
size with userland version, imported in one of the previous commits.
This allows us to skip zeroing of buffer on processing of each report.
if present to enable some devices like WaveShare touchscreens. Unlike
Windows we discard content of the blob. We try mimic Windows driver
behaviour from the USB device point of view.
Kyle Evans [Sat, 4 Aug 2018 06:40:18 +0000 (06:40 +0000)]
efi-autoresizecons: Don't fail the boot w/o GOP or UGA
efi-autoresizecons is currently executed for every boot. If it fails, we
risk failing the boot, and we really shouldn't do that unless we absolutely
must.
Not being able to locate GOP or UGA is not a significant enough failure to
kill the boot. We always have the option to fall back to resizing ConOut to
a higher text mode resolution (if available), so do that.
This was detected by Doug [1] while attempting a bhyve + UEFI + PXE boot.
This patch was effectively also submitted by Doug, but I expanded the
comment he had originally sent me a little bit to indicate why this is an OK
idea.
Emmanuel Vadot [Fri, 3 Aug 2018 22:04:00 +0000 (22:04 +0000)]
dtb: am335x: Remove links and add more dts
The links were to cope with the switch to upstream dts.
We don't need them anymore.
While here add the rest of the beaglebone family dts as u-boot is common
on all those boards and load the dtb based on the product name.
This just miss the pocketbeagle variant as it's not yet in sys/gnu/dts but
will be with the Linux 4.18 dts import.
Justin Hibbits [Fri, 3 Aug 2018 20:04:06 +0000 (20:04 +0000)]
nvme(4): Add bus_dmamap_sync() at the end of the request path
Summary:
Some architectures, in this case powerpc64, need explicit synchronization
barriers vs device accesses.
Prior to this change, when running 'make buildworld -j72' on a 18-core
(72-thread) POWER9, I would see controller resets often. With this change, I
don't see these resets messages, though another tester still does, for yet to be
determined reasons, so this may not be a complete fix. Additionally, I see a
~5-10% speed up in buildworld times, likely due to not needing to reset the
controller.
Bryan Drewery [Fri, 3 Aug 2018 19:24:04 +0000 (19:24 +0000)]
Fix some filemon path logging issues.
- Properly handle snprintf return value for truncation and avoid
overflowing the later write with the bogus length.
- Increase the msgbufr size to handle a rename of 2 full files.
The larger allocation causes a slight performance hit which will be mitigated
in the future. A rewrite with sbufs will likely be done as well.
Cy Schubert [Fri, 3 Aug 2018 19:11:00 +0000 (19:11 +0000)]
Some drives report a geometry that is inconsisetent with the total
number of sectors reported through the BIOS. Cylinders * heads *
sectors may not necessarily be equal to the total number of sectors
reported through int13h function 48h.
An example of this is when a Mediasonic HD3-U2B PATA to USB enclosure
with a 80 GB disk is attached. Loader hangs at line 506 of
stand/i386/libi386/biosdisk.c while attempting to read sectors beyond
the end of the disk, sector 156906855. I discovered that the Mediasonic
enclosure was reporting the disk with 9767 cylinders, 255 heads, 63
sectors/track. That's 156906855 sectors. However camcontrol and
Windows 10 both report report the disk having 156301488 sectors, not
the calculated value. At line 280 biosdisk.c sets the sectors to the
higher of either bd->bd_sectors or the total calculated at line 276
(156906855) instead of the lower and correct value of 156301488 reported
by int 13h 48h.
This was tested on all three of my Mediasonic HD3-U2B PATA to USB
enclosures.
Instead of using the higher of bd_sectors (returned by int13h) or the
calculated value, this patch uses the lower and safer of the values.
John Baldwin [Fri, 3 Aug 2018 18:52:51 +0000 (18:52 +0000)]
Install the 32-bit compat sanitizer libraries.
The lib32 build was already building the i386 version of
the clang sanitizers (libclang_rt) but they were not being
installed. This enables the installation.
MK_TOOLCHAIN=no was originally added to the install make
environment to disable includes so that NO_INCS could be
removed. The MK_TOOLCHAIN in bsd.incs.mk was subsequently
renamed to MK_INCLUDES, but bsd.lib.mk doesn't even include
bsd.incs.mk when LIBRARIES_ONLY is defined which the install
make environment for compat libs now defines. However,
setting MK_TOOLCHAIN=no forced MK_CLANG=no which disabled
libclang_rt during the install32 phase. Remove MK_TOOLCHAIN=no
since LIBRARIES_ONLY is now sufficient.
Since the libcompat environment overrides both LIBDIR and
SHLIBDIR, libclang_rt/Makefile.inc has to set both variables
to force the libraries to be installed to the location
expected by the compiler.
This actually makes the rights requirements for accessing PCI config
space and BARs using /dev/pci same. Since unchanged /dev/pci mode
only allows write open for root, default configuration de-facto limits
the BAR read to root only. In particular, state-changing reads of the
registers are limited to root.
Discussed with: se
Suggested and reviewed by: jhb (kernel part)
Sponsored by: The FreeBSD Foundation
MFC after: 12 days
Differential revision: https://reviews.freebsd.org/D16580
Andriy Gapon [Fri, 3 Aug 2018 14:27:28 +0000 (14:27 +0000)]
safer wait-free iteration of shared interrupt handlers
The code that iterates a list of interrupt handlers for a (shared)
interrupt, whether in the ISR context or in the context of an interrupt
thread, does so in a lock-free fashion. Thus, the routines that modify
the list need to take special steps to ensure that the iterating code
has a consistent view of the list. Previously, those routines tried to
play nice only with the code running in the ithread context. The
iteration in the ISR context was left to a chance.
After commit r336635 atomic operations and memory fences are used to
ensure that ie_handlers list is always safe to navigate with respect to
inserting and removal of list elements.
There is still a question of when it is safe to actually free a removed
element.
The idea of this change is somewhat similar to the idea of the epoch
based reclamation. There are some simplifications comparing to the
general epoch based reclamation. All writers are serialized using a
mutex, so we do not need to worry about concurrent modifications. Also,
all read accesses from the open context are serialized too.
So, we can get away just two epochs / phases. When a thread removes an
element it switches the global phase from the current phase to the other
and then drains the previous phase. Only after the draining the removed
element gets actually freed. The code that iterates the list in the ISR
context takes a snapshot of the global phase and then increments the use
count of that phase before iterating the list. The use count (in the
same phase) is decremented after the iteration. This should ensure that
there should be no iteration over the removed element when its gets
freed.
This commit also simplifies the coordination with the interrupt thread
context. Now we always schedule the interrupt thread when removing one
of handlers for its interrupt. This makes the code both simpler and
safer as the interrupt thread masks the interrupt thus ensuring that
there is no interaction with the ISR context.
P.S. This change matters only for shared interrupts and I realize that
those are becoming a thing of the past (and quickly). I also understand
that the problem that I am trying to solve is extremely rare.
PR: 229106
Reviewed by: cem
Discussed with: Samy Al Bahra
MFC after: 5 weeks
Differential Revision: https://reviews.freebsd.org/D15905
Alexander Motin [Fri, 3 Aug 2018 02:16:45 +0000 (02:16 +0000)]
Reduce taskq and context-switch cost of zio pipe
When doing a read from disk, ZFS creates 3 ZIO's: a zio_null(), the
logical zio_read(), and then a physical zio. Currently, each of these
results in a separate taskq_dispatch(zio_execute).
On high-read-iops workloads, this causes a significant performance
impact. By processing all 3 ZIO's in a single taskq entry, we reduce the
overhead on taskq locking and context switching. We accomplish this by
allowing zio_done() to return a "next zio to execute" to zio_execute().
This results in a ~12% performance increase for random reads, from
96,000 iops to 108,000 iops (with recordsize=8k, on SSD's).
Reviewed by: Pavel Zakharov <pavel.zakharov@delphix.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed by: George Wilson <george.wilson@delphix.com> Signed-off-by: Matthew Ahrens <mahrens@delphix.com>
External-issue: DLPX-59292
Closes #7736
Alan Somers [Fri, 3 Aug 2018 01:37:00 +0000 (01:37 +0000)]
Fix LOCAL_PEERCRED with socketpair(2)
Enable the LOCAL_PEERCRED socket option for unix domain stream sockets
created with socketpair(2). Previously, it only worked with unix domain
stream sockets created with socket(2)/listen(2)/connect(2)/accept(2).
PR: 176419
Reported by: Nicholas Wilson <nicholas@nicholaswilson.me.uk>
MFC after: 2 weeks
Differential Revision: https://reviews.freebsd.org/D16350
Reviewed by: Pavel Zakharov <pavel.zakharov@delphix.com>
Reviewed by: George Wilson <george.wilson@delphix.com>
Reviewed by: Prashanth Sreenivasa <pks@delphix.com>
Reviewed by: Robert Mustacchi <rm@joyent.com>
Approved by: Dan McDonald <danmcd@joyent.com>
Author: Matthew Ahrens <mahrens@delphix.com>
Reviewed by: Pavel Zakharov <pavel.zakharov@delphix.com>
Reviewed by: George Wilson <george.wilson@delphix.com>
Reviewed by: Prashanth Sreenivasa <pks@delphix.com>
Reviewed by: Robert Mustacchi <rm@joyent.com>
Approved by: Dan McDonald <danmcd@joyent.com>
Author: Matthew Ahrens <mahrens@delphix.com>
Reviewed by: Sebastien Roy <sebastien.roy@delphix.com>
Reviewed by: George Wilson <george.wilson@delphix.com>
Reviewed by: Robert Mustacchi <rm@joyent.com>
Approved by: Dan McDonald <danmcd@omniti.com>
Author: Matthew Ahrens <mahrens@delphix.com>
Reviewed by: Sebastien Roy <sebastien.roy@delphix.com>
Reviewed by: George Wilson <george.wilson@delphix.com>
Reviewed by: Robert Mustacchi <rm@joyent.com>
Approved by: Dan McDonald <danmcd@omniti.com>
Author: Matthew Ahrens <mahrens@delphix.com>
Reviewed by: Adam Leventhal <ahl@delphix.com>
Reviewed by: George Wilson <george.wilson@delphix.com>
Reviewed by: Robert Mustacchi <rm@joyent.com>
Approved by: Dan McDonald <danmcd@omniti.com>
Author: Matthew Ahrens <mahrens@delphix.com>
Reviewed by: Adam Leventhal <ahl@delphix.com>
Reviewed by: George Wilson <george.wilson@delphix.com>
Reviewed by: Robert Mustacchi <rm@joyent.com>
Approved by: Dan McDonald <danmcd@omniti.com>
Author: Matthew Ahrens <mahrens@delphix.com>